Differential Display Methods and Protocols (Methods in Molecular Biology)

1 Differential Display A General Protocol Peng Liang and Arthur B. Pardee 1. Introduction One of the greatest unsolved...

Author: Peng Liang | Arthur B. Pardee

20 downloads 686 Views 12MB Size Report

This content was uploaded by our users and we assume good faith they have the permission to share this book. If you own the copyright to this book and it is wrongfully on our website, we offer a simple DMCA procedure to remove your content from our site. Start by pressing the button below!
Report copyright / DMCA form

DOWNLOAD PDF

1 Differential

Display

A General Protocol Peng Liang and Arthur B. Pardee 1. Introduction One of the greatest unsolved mysteries of life 1show the hundreds of thousands of genes embedded in the genome of an organism are selectively expressed mto the mRNA and protems m a temporally and spatially regulated manner that gives rise to different tissues and organs. The abnormality m this intricate regulatory cn-cuitry IS beheved to be one of the underlmmg causes of a variety of pathological alterations or disease states.The rsolation and characterization of differentially expressed genes becomes one of the first steps toward the understanding of these important biological questions. Differential display (1) and a related RAP-PCR method (2) were developed to more efficiently Identify and isolate these genes. The general strategy for differential display (Fig. 1) IS based on a combmatton of three techniques brought together by a concept: 1 Reverse transcrlptlon of mRNA from anchored primers (see Note l), 2 Choice of arbitrary primers for setting lengths of cDNAs to be amplified by the polymerase chain reaction (PCR), each corresponding to part of a mRNA (tags), 3 Sequencmg gels for high resolution of amplified cDNA

The objective IS to obtain a tag of a few hundred bases,which 1ssufficiently long to uniquely identify a mRNA and yet short enough to be separated from others by size. Given the fact that primers of nearly any length, with or without anchors, can generate cDNA fingerprints sufticrently reproducible to allow differentially expressed genes to be identified, it may be hard to define what should be a standard protocol for differential display. The followmg protocol using one-base anchored primer m combmatron with arbitrary 13-mers (3) IS given as an example to illustrate the methodology. From

Methods m Molecular Bology, Vol 8.5 D/fferenf/a/ Edlted by P Llang and A B Pardee Humana

3

Dsplay Methods and Protocols Press Inc , Totowa, NJ

Liang and Pardee

L

Reverse banswiption

5’.AAGC3’ dNTPs MMLV reverse tfanscnptase .

WAA-A~ GVGAA

4 JI.

(H-TUG)

PcRamplificatloll

AAGCTTGATECC

I

5’-AAGCTTGAITGCC-3’ (H-AP-1 Ptimer) 5’44GC3 (H-TIIG) dNTPs a-(‘“S-dATP] Ampli’lhq DNA poiymelase . GWGAA

A4GCiTGA’lTGCC

. G-GU

IIL

Denaturing polyacrylamide gel

Fig. 1. Schemattc representation of one-base anchored differentral display

2. Materials 1. 5X RT buffer: 125 mA4TrwC1, pH 8 3,188 mMKCl,7.5 dithrothretol. 2 MMLV reverse transcrrptase (100 cl/@,) 3. dNTP (250 @I) 4. 5’-AAGCTTTTTTTTTTTG-3’ (2 /.&I). 5. 5’-AAGCTTTTTTTTTTTA-3’ (2 pA4) 6 5’-AAGCTTTTTTTTTTTC-3’ (2 /&I). 7 Arbitrary 13-mers (2 pM) 8. 10X PCR buffer.

mMMgC12, and 25 ti

Differential 9 10. Il. 12. 13. 14. 15 16. 17 18. 19. 20. 2 1.

Display

5

dNTP (25 pJ4). Glycogen (10 mg/mL). Distilled water (dH,O). DEPC-treated H,O Loading dye AmphTaq DNA polymerase, Perkin-Elmer Corporation (Norwalk, CT). a-[33P]dATP (>2000 Wrnmole) or a-[35S]dATP (>l,OOO Ci/mmole) (see Note 2). RNase-free DNase I (10 U/pL). QIAEXrM DNA extraction kit (Qiagen, Chatsworth, CA). pCR-TRAPTM clonmg system (GenHunter Corporation, Nashville, TN). Thermocycler 6% denaturmg polyacrylamide gel. DNA sequencing apparatus.

Although individual components may be purchased separately from varrous suppliers, most of them can be obtained in kit forms from GenHunter Corporation. 3. Methods 3.1. DNase I Treatment of Total RNA Purification polyadenylated RNAs is neither necessary nor helpful for differential display. The major pitfalls of using the polyadenylated mRNAs are the frequent contammation of the oligo-dT primers, that give high background smearing in the display and the difficulty m assessing the integrity of the mRNAs templates (4). Total cellular RNAs can be easily purified with one-step acid-phenol extraction method (5). However, no matter what methods are used for the total RNA purification, a trace amount of chromosomal DNA contamination m the RNA sample could be amplified along with mRNAs thereby comphcating the pattern of displayed bands. Therefore removal of all contaminating chromosomal DNA from RNA samples is essential before carrying out differential display. 1. Incubate 10-100 pg of total cellular RNA with 10 U of DNase I (RNase free) in 10 mMTris-Cl, pH 8.3, 50 mMKC1, 1.5 mMMgC& for 30 mm at 37°C. 2 Inactivate DNase I by adding an equal volume of phenolchloroform (3: 1) to the

sample 3. Mix by vortexing and leave the sample on ice for 10 mm.

4 Centrifuge the samplefor 5 min at 4’C m an Eppendorf centrifuge. 5. Save the supernatant, and ethanol precipitate the RNA by adding 3 vol of ethanol in the presence of 0.3MNaOAC, and incubate at -80°C for 30 mm. 6. Pellet the RNA by centrifuging at 4°C for 10 min. 7. Rinse the RNA pellet with 0.5 mL of 70% ethanol (made with DEPC-H20) and redissolve the RNA in 20 pL of DEPC-treated HzO. 8. Measure the RNA concentration at ODS6s with a spectrophotometer by diluting 1 pL of the RNA sample in 1 mL of HzO.

Llang and Pardee

6

9 Check the integrity of the RNA samples before and after cleanmg wtth DNase I by runnmg 1-3 ~18of each RNA on a 7% formaldehyde agarose gel 10 Store the RNA sample at a concentratton htgher then 1 pg/$ at -80°C before using for differential dtsplay

3.2. Reverse

Transcription

of mRNA

1 Set up three reverse transcription reactions for each RNA sample in three microfuge tubes (0 5-mL), each contammg one of the three dtfferent anchored ohgo-dT prrmers as follows. For 20 pL final volume 9 4 p.L of dH,O, 4 pI. of 5X RT buffer, 1.6 pL of dNTP (250 I.&‘), 2 pL of DNA-free total RNA (freshly diluted to 0 1 pg/pL wtth DEPC-treated H,O), and 2 pL of AAGCT, ,M (2 $4) (M can be either G, A, or C) 2 Program your thermocycler to* 65°C for 5 mm, 37°C for 60 min, 75’C for 5 mm, 4°C (see Note 3) 3 1 pL MMLV reverse transcrlptase 1s added to each tube 10 mm after at 37°C and mix well quickly by finger tipping 4. Continue mcubation and at the end of the reverse transcription reaction, spm the tube briefly to collect condensation Set tubes on ice for PCR or store at -80°C for later use.

3.3. PCR Amplification 1 Set up PCR reacttons at room temperature as follow* 20 p.L final volume for each primer set combmation 10 pL of dH,O, 2 $ of 10X PCR buffer, 1.6 l.iL of dNTP (25 pA4), 2 pL of arbitrary 13-mer (2 CUM),2 pL of AAGCT, ,M (2 CIM), 2 pL of RT-mix from step 3 2 , 0 2 pL of a-[33P]-dATP (see Note 2), 0 2 p.L of AmpliTaq. Mix well by pipetmg up and down (see Note 4) 2. Add 25 pI.. mineral oil if needed 3 PCR as follows 94°C for 30 s, 40°C for 2 mm, 72°C for 30 s for 40 cycles, 72°C for 5 mm, 4“C (For Perkm-Elmer’s 9600 thermocycler it is recommend that the denaturanon temperature be shortened to 15 s and the rest of parameters kept the same )

3.4. 6% Denaturing 1 2 3 4

Polyacrylamide

Gel Electrophoresis

Prepare a 6% denaturmg polyacrylamide gel m TBE buffer Let it polymertze at least for more than 2 h before usmg Prerun the gel for 30 mm Mix 3.5 pL of each sample with 2 p.L of loading dye and incubate at 80°C for 2 mm immediately before loading onto a 6% DNA sequencmg gel (see Note 5) 5 Electrophorese for about 3 5 h at 60 W constant power (with voltage not to exceed 1700 V) until the xylene dye (the slower movmg dye) reaches the bottom Turn off the power supply and blot the gel onto a piece of 3M Paper Cover the gel with a plastic wrap and dry tt at 80°C for 1 h Do not fix the gel with methanol/ acetic acid (see Note 6) 6. Orient the autoradiogram and dried gel wtth radioacttve mk or needle punches before exposing to a X-ray film. Figure 2 shows a representative differential dts-

7

Differential Display H-T110 H-AP3

H-T110 H-APB

H-TIIA H-API

H-WA H-AP3

H-TIIA

H-TIIC

H-TllC

H-AP3

H-APl

H-AP3

Fig. 2. Differential display using one-base anchored oligo-dT primers (7). Four RNA samples from non-transformed cell line Rat 1 and H-ras transformed cell lines rat 1 (ras), T101-4 and Al-5 (lanes from left to right, respectively) were compared by differential display using three one-base anchored oligo-dT primers, AAGCT, ,G, AAGCT, ,A and AAGCT, ,C in combinations with three arbitrary 13-mers, H-API (AAGCTTGATTGCC), HAP2 (AAGCTTCGACTGT) and HAP3 (AAGC’l’l’l’GGTCAG). The mob-l (ZO) and mob-7 cDNA fragments were marked by the right and let? arrowheads, respectively. play obtained with three one-base anchored oligo-dT primers in combinations with three arbitrary 13-mers (3).

3.5. Reamplification

of cDNA Probe

1. After developing the film (overnight to 72-h exposure), orient the autoradiogram with the gel. 2. Locate bands of interest (see Note 7) either by marking with a clean pencil from underneath the film or punching through the film with a needle at the four corners

Liang and Pardee

3. 4. 5 6. 7. 8 9. 10. 11. 12 13.

14.

15

of each band of interest (Handle the dried gel with gloves and save it between two sheets of clean paper) Cut out the located band with a clean razor blade Soak the gel slice along with the 3M paper in 100 pL dH,O for 10 mm. Boil the tube with tightly closed cap (e.g., with parefilm) for 15 min. Spm for 2 mm to collect condensatron and pellet the gel and paper debris Transfer the supernatant to a new micromge tube. Add 10 pL of 3MNaOAC, 5 pL of glycogen (10 mg/mL) and 450 p.L of 100% EtOH. Let sit for 30 mm on dry ice or in a -8O’C freezer Spm for 10 mm at 4°C to pellet DNA. Remove supernatant and rinse the pellet with 200 pL we-cold 85% EtOH (you will lose your DNA if less concentrated EtOH is used!). Spm briefly and remove the resrdual ethanol. Dissolve the pellet m 10 pL of PCR H,O and use 4 pL for reamplificatron. Save the rest at -20°C in case of mishaps. Reamplification should be done using the same primer set and PCR conditions except the dNTP concentrations are at 20 piV (use 250 @4 dNTP stock) instead of 2-4 pA4 and no rsotopes added. A 40-pL reaction is recommended for each reactron: 20.4 of pL dH,O, 4 pL of 10X PCR buffer, 3.2 pL of dNTP (250 @4), 4 & of arbitrary 13-mer (2 @?), 4 $ AAGCT, ,M (2 @4) (M can be either G, C, or A), 4 pL of cDNA template from step 3.2. and 0.4 pL of AmphTaq (5 U/pL). Run 30 pL of the PCR sample on a 1.5% agarose gel stained with ethidmm bromide (More than 90% probes should be visible on the agarose gel ) Save the remaining PCR samples at -2O’C for subclomng. Check to see rf the size of your reamplified PCR products are consistent with their size on the denaturing polyacrylamide gel.

3.6. Confirmation

of Differential

Gene Expression

1. Extract the reamplified cDNA probe from the agarose gel using QIAEX kit. 2. Use the extracted cDNA as a probe for Northern blot confirmation following the standard protocol (ref. 5; see Note 8; Fig. 3) 3. Clone the cDNA probe using the pCR-TR4PTM cloning system (see Note 9). 4. Confirmation of differentially expressed cDNA probes can be also carried out more efficiently by “Reverse Northern” dot blot or differential screening of cloned cDNA probes by colony hybrtdization (ref. 6; Chapter 8 by H Zhang et al. m this book). 5. Clone the full-length cDNA by screening a cDNA library followmg the standard procedure (5).

4. Notes 1. The initial choice of usmg two-base anchored ohgo-dT primers (1) instead of one-base anchored primers (3) were owing to a historical rather than scienttfic reason. The cloned marine thymidine kmase (TK) cDNA originally used as a

Differential

9

Display

Mob-l

A H-AP2

+ AAgc~~CTGTACAAAGG~~C~~T~A~~AC~~~~~ ATATGTAAGAACGTATGTATCAATGGGTAGITAAAGTlTACATAGG CAAATGClTl-GAATGCTACATAlTACAAGATGTGlTGGATGGlllTCAMATAMAT GTACTGTATTGAATGTAGTATGAGACCAAAAAA GTAATAAAGTAATAATAACTGAC ATGAAAAAAAAAAAGC-IT 4 H-T1 I C

Mob-7

B H-AP2

* AAgcttcGAcTGTAcAAA~GcGGAAcTccfGAATGTATTTT ATAT~AAGAAClTGTGTGGTAAGTATGTATGTAfCAATGGGTAGlTAAA~ACATAGG CAAATGCllTGAATGCTACATATTACAAGATGElTGGATGGlllTCAAAATAAAAT GTACCCAAAAAAGTAATAAAGTAATAATAACTGAC ATGAAATGCAAAAAAAAAAAGCTT 4 H-T1 I G

C

1234

Mob-7

rRNA

Fig. 3. Nucleotide sequences of mob-l (A) and mob-7 (B) cDNA fragments cloned by differential display. The flanking primers are marked by arrow bars and the polyadenylation site is underlined. Mob-7 differs from mob-l only by 6 base addition at the 3’ end of the cDNA (see Note 10). Northern blot analysis with mob-7 cDNA probe (C). The 253 bp mob-7 cDNA was used as a probe to confirm the differential expression of the gene using 20 pg of total RNA from Rat 1 and three transformed derivatives Rat 1 @as), T101-4 and Al-5 cells (lanes 1 to 4, respectively). The lower panel is ethidium bromide staining of ribosomal RNAs as a control for equal sample loading.

10

2

3.

4

5 6

7.

8.

9.

Lang and Pardee control cDNA template had only 11 As m its poly(A) tall It was found that onebase anchored prrmer Tl 1C fatled to amplify the TK 3’ termmus m combmatton wtth an upstream primer specific to TK. Extension of one more base from the 3’ end instead of the 5’ end of the anchored primer was a logical. Interestingly, Tl 1CA started to work successfully m PCR to amplify the expected TK cDNA template (1) Later, longer one-base anchored primers that had mismatches at the 5’ ends of the prtmers were shown to be much more efficient for differential display m subdividing the mRNA populations mto three groups (3) One-base anchored primers have significant advantage over the two-base anchored primers m that the former cuts down the redundancy of priming, elimmates the high background smearing problem for two-base anchored pnmers ending with the 3’ “T” and reduce the number of reverse transcription reactions from 12 to 3 per RNA sample. It has been observed that 35S labeled nucleottde origmally used for differential display would leak through PCR reaction tubes (espectally when thin-walled tubes are used) and 33P labeled nucleotide was recommended as the best alternative (9). 33P is not only safer to use but also gives better sensmvtty compared to 35S. For the reverse transcrtption reaction, the mmal 65°C incubation is intended to denature the RNA secondary structure. The final incubation at 75°C for 5 minis to inactivate the reverse transcrlptase without denaturing the cDNA/mRNA duplexes Therefore “hot start “PCR is neither necessary, nor helpful for the subsequent PCR reactions using cDNAs as templates Make core mixes as much as possible to avoid ptpetmg errors (e g , aliquot RT-mix and AP-primer mdrvidually) Otherwise it would be difficult to pipet 0.2 pL of AmpliTaq. Mix well by pipetmg up and down It is crucial that the urea in the wells be completely flushed right before loading your samples For best resolution, flush every 4-6 wells each time during sample loading while trying not to disturb the samples that have been already loaded. DNA is acid labile, especially at high temperature when the gel is dried. This will affect the subsequent PCR during the reamplificatton of the cDNA fragments to be analyzed further First tentatively identify those bands that appear to be differentially expressed on the initial display gel. Then repeat the RT step and the PCR reactions for these lanes and see if these differences are reproducible before pursumg further It 1s recommended that bands bigger than 100 bp be selected. It has been generally observed that shorter cDNA probes have higher probability of failing to detect any signals on the Northern blot. It IS recommended that the standard prehybrrdrzatron and hybridrzation condttton at 42°C be used. Wash with 1X SSC, 0.1% SDS at room temperature for 15 min twice followed by washing with 0.25X SSC, 0 1% SDS at 55-6O”C for 15-30 mm. Do not go over 60°C Expose with intensifying screen at -80°C for overnight to 1 wk. pCR-TRAP cloning system is by far the most efficient cloning method for PCR products that we have tested The pCR-TRAP clonmg system utilizes the third generation cloning vector that features postttve-selection for DNA mserts Only

Different/al Display

II

the recombinant plasmtds confer the antibiotic resistance The prmclple of thts unique clonmg system 1sbased on that the phage Lambda repressor gene c1 cloned on the pCR-TRAP vector codes for a repressor protein. The repressor protein binds to the Lambda right operators Or1 to Or3 of the cro gene, thereby turning off the promoter that drives the TetR gene on the plasmtd Therefore, cloning of the PCR products dtrectly, without any post-PCR purification, into the c1 gene leads to the inactivation of the repressor gene, thus turnmg on the TetR gene The cloned PCR insert can then be readily sequenced or retrieved as a probe by PCR using primers flanking the cloning site of the vector. 10. It 1sknown that the poly(A) tail of a rnRNA is not always added at a fixed position downstream of the AATAAA polyadenylatton signal This 1s why both mob-l and mob-7 correspondmg to the same mRNA were detected by the same arbitrary primer m combmatton with different anchored primers

Acknowledgment We thank GenHunter Corporation for the permtsslon of adapting its protocols for Message CleanTMkit and RNAlmage TMkit for differential display. The work was supported m part by a Natlonal Institute of Health grant CA61232 awarded to Arthur B. Pardee and Peng Llang.

References 1 Liang, P. and Pardee, A B. (1992) Dtfferentlal display of eukaryotic messenger RNA by means of the polymerase cham reaction. Science 257, 967-97 1. 2 Welsh, J , Chada, K , Dalal, S S , Cheng, R , Ralph, D., and McClelland, M (1992) Arbitrarily primed PCR fingerprmtmg of RNA Nuclezc Aczds Res 20, 4965-4970 3 Liang, P. Zhu, W , Zhang, X., Guo, Z , O’Connell, R P., Averboukh, L , Wang, F , and Pardee A B (1994) Differenttal display usmg one-base anchored ohgo dT primers. Nucleic Acids Res 22, 5763,5764. 4 Ltang, P. Averboukh, L., and Pardee, A B (1993) Dtstrlbutton and cloning of eukaryottc mRNAs by means of differential display* refinements and optimization Nuclezc Acids Res 21, 3269-3275. 5 Ausubel, F , Brent, R , Kingston, R. E , Moore, D. D., Seidman, J. G , Smith, J A, and Struhl, K (1988) Current Protocols In Molecular Biology, Greene and Wiley-Interscience, New York 6. Zhang, H , Zhang, R., and Ltang, P. (1996) Differential screening of gene expression difference enriched by differential display. Nuclezc Acids Res 24,2454-2456. 7 Trentmann, S M , Knaap, E., Kende, H., Ltang, P., and Pardee A. B (1995) Alternatives to 35S as a label for the differential display of eukaryottc messenger RNA. Sczence 267,1186,1187

Fingerprinting

by Arbitrarily

Primed PCR

Michael McClelland, Rhonda Honeycutt, Francoise Mathieu-Daude, Thomas Vogt, and John Welsh 1. Introduction PCR using primers of arbitrary sequence can generate a reproducible tingerprint of products from DNA (1,2). Differences m the fingerprint of products generated from DNA of related organism are a result of polymorphisms. These polymorphisms proved useful markers for genetic mappmg (3-7), population biology (8-13), epidemiology (Z&19), and even the discovery of the mutator phenotype m cancer (20). When applied to RNA the method is also capable of detecting polymorphisms in expressed transcripts (21). However, more importantly, the method can detect a sample of differentially expressed genes (21,22). This chapter will concentrate on recent efforts in our laboratory to extend and improve the method and ascertain its limitations using relatively “lowtech” solutions available to most laboratories. There are three vital steps common to all RNA tingerprmting experiments: (1) the arbitrarily primed PCR amphfication of a sample of transcripts, (2) the isolation and characterization of differentially amplified products, and (3) the confirmation of differentially expressed products in the system of interest. We will discuss issues pertaining to each of these steps. When performed with the correct controls, with efficient analysis protocols, and taking into account what it cannot do efficiently, the method is suitable for identifying a sample of regulated genes m a wide variety of experimental systems. It should be noted that our protocol (RNA arbitrarily primed PCR [RAPPCR]) differs from the Liang and Pardee Differential Display protocol m that we use an arbitrary primer m both steps of the PCR reaction, rather than an anchored oligo(dT) in the first step. The use of arbitrary primers to define both ends of fingerprmt products allows internal RNA fragments to be sampled, From

Methods m Molecular Bfology, Vol 85 D/fferent/a/ Edited by P Llang and A B Pardee Humana

13

Display Methods and Protocols Press Inc , Totowa, NJ

14

McClelland et al.

including open reading frames. In addition, mRNAs that are not polyadenylated can be sampled, such as some bacterial RNAs (23). We cannot be certain that all the properties of these two protocols are the same,nevertheless, there 1sreason to believe that the issues we discuss here are equally applicable

to both protocols.

2. Materials 1. 2 3. 4 5 6 7 8. 9 10 I 1. 12 13 14 15 16 17. 18 19. 20 21. 22.

Multipipetor for 5- to 200~pL volumes RNeasy total RNA purificatton kit (QIAGEN, Chatsworth, CA) DNase stock (10 U/uL) (Boehrmger Mannhelm Biochemicals, Indianapolis, IN) RNasm RNase mhibitor, 40 II/& (BMB). 2X RT Buffer. 100 tiTris, pH 8 3, 100 mA4KC1, 8 mMMgC12 MuLV-reverse transcrtptase 1200 U/pL (Promega, Madison, WI) Stocks of all four dNTPs (5 mM) Stocks of primers (100 w (Genosys, Woodlands, TX) [a-32P] dCTP (3000 Ci/mmol; ICN, Costa Mesa, CA). 2X Tuq polymerase mixture 20 mM Tris, pH 8 3; 20 mM or 100 mA4 KCl; 8 n-&f MgC12 AmphTaq polymerase, 5 U/pL. (Perkm-Elmer-Cents, Norwalk, CT) AmphTuq polymerase Stoffel fragment, 10 U/pL (Perkm-Elmer-Cetus). GeneAmp PCR System 9600 thermocycler (Perkm-Elmer-Cetus) Formamrde dye solution. 96% formamide, 0.1% bromophenol blue, 0 1% xylene cyanol, 10 mA4 EDTA Acrylamrde stock solutrons (40% 19.1 acrylamrde.bzs-acrylamide) Urea Ammonium persulfate (fresh 10% solution) TEMED. 10X TBE buffer: 90 mM Tris-Borate, 20 mMNa2EDTA, pH 8.3. Hydrolmk MDE gel solutions (J T Baker Inc , Phillipsburg, NJ). NaOH dye solution. 96% formamide, 0 1% bromophenol blue, 0.1% xylene cyanol, 10 mM NaOH. Fluroimager or Aquasol scintillation fluid and a scinhllation counter.

3. Methods 3.1. Fingerprinting 1. Total RNA is purified usmg the RNeasy total RNA purification kit (QIAGEN, Chatsworth, CA) Typically, lo6 mammalian cells from cell culture yield 5 pg of RNA m 50 $ For bacterta, a hot phenol extraction may be used (23) This is treated with 0 08 U/pL DNase (plus 0 24 U/$ of RNasm, an RNase inhibitor) at 37°C for 40 mm m 1X RT buffer The RNA is repurtfied using the RNeasy kit The yield is estimated by spectrophotometry and the RNA is diluted to 200 ng/pL in water If sufficient RNA is available, the quality and concentration of the RNA is checked by agarose gel electrophoresis before bemg stored at -80°C (see Note 1 for comments on experrmental destgn)

Fmgerprin

ting

15

Reverse transcription 1s performed on total RNA at three concentrations per sample (500, 250, and 125 ng per reactton) using an ohgonucleotrde primer of arbitrary sequence of 10-20 nt m length (see Note 2) 5 ,LILof each RNA is mixed with the same volume of RT reaction mrxture for a 10 pL final reaction contammg 50 mA4 Tris pH 8.3, 50 mA4 KCI, 4 mA4 MgCl,, 10 n&f DTT, 0 2 mM of each dNTP, 2 @4 of first primer, and 16 U of MuLV-reverse transcrrptase (see Notes 3-6). The first strand cDNA synthesis reaction IS then ramped from room temperature to 37°C over 5 min, held at 37’C for 1 h, then heated to 94°C for 5 mm to stop the reaction Finally, the resultant first strand cDNA IS diluted fourfold by the addttion of 3 vol of water. For second strand synthesis, the diluted cDNA (10 pL) 1s mixed with the same volume of PCR mixture for a 20-pL final reaction contammg 10 mMTrts pH 8.3, 10 rnA4KC1,4 mA4MgC12, 0.2 mA4of each dNTP, 4 pA4of a second primer, 1 &I [u-~~P] dCTP, and 4 U of AmphZrq polymerase Stoffel fragment (see Note 7) Thermocyclmg 1sperformed using: 30 cycles of 94°C for 30 s, 35°C for 1 mm, 1 mm ramp to 72°C and 72°C for 2 mm. Amphficatron products (5 pL) are mixed with 15 ,rrL of formamide dye solution, denatured at 68°C for 10 mm, and 2 2 pL IS loaded onto a 5% acrylamrde-50% urea gel, prepared m 1X TBE buffer. Electrophoresis 1s performed using a sequencing apparatus at 58 W constant power (about 1500 V) until the xylene cyan01 tracking dye reaches the bottom of the gel (approx 4 h). After electrophoresrs the gel is transferred to Whatman 3MM paper, and dried under vacuum The drred gel 1s autoradrographed using Bromax film (Kodak) for 12 h-4 d. An mtensrfymg screen 1snot used because this tends to blur the fine details of closely packed PCR products. An example IS shown in Fig. 1,

3.2. Isolation

of Differentially Amplified PCR Products One of the major bottlenecks m using RAP-PCR or Differential Display is the need to isolate and characterize the PCR fragments representmg differentially amplified PCR products. The initial step employed has usually been the same as that used for AP-PCR of DNA (4), namely, cutting the band from the denaturing polyacrylamide gel and reamplifymg the resultmg product. However, there is quite often a problem with this approach. The “Cot effect” (see Note 6) preferentially amplifies the other products of a stmrlar srze that are copurified with the band of interest. Although this IS not a problem in many caseswhere the band of interest vastly predominates even after partial normalrzatton during PCR, it can lead to a lot of wasted effort for a subset of products where contaminants predominate. We have used two approaches to address this problem. One way to increase the probability of cloning the correct band from the mixture is to simply mimmrze the number of cycles of PCR, thereby limiting the mass of DNA made and thus the Cot effect. Alternatively, the

16

McClelland et al. tel

di

mes

met

622 bp

527 bp

404 bp

307 bp ’

Fig. 1. Differential gene expression in the mouse brain. The following parts of the mouse brain were dissected at 12 d postconception by Kiran Chada, UMDNJ: telencephalon, diencephalon, mesencephalon, and metencephalon. We prepared RNA from these samples and 1000, 500, and 250 ng of each were reverse transcribed with the primer Tryp 1- (S-GTGGCGTTGAT). One-fourth of the first strand synthesis was used in a second strand cDNA reaction after the addition of the primer OPN25 (S- GGGGCACCAG) and Stoffel fragment of Tuq polymerase. After PCR the products were resolved on a 5% denaturing acrylamide gel. Arrows indicate two differentially expressed genes.

fingerprinting

17

reamplified material can be separated by single-strand conformation polymorphisms (SSCP) resolved on a native polyacrylamlde gel (24). This method resolves the denatured reampllfied PCR product by virtue of secondary structure. Because the product of interest and the contaminants are of entirely different sequence, they migrate to different points on the gel. The level of contammatton can be assessed from these gels allowing problematic reampiificatlon mixtures to be rejected for further study. Finally, if enough PCR product is loaded in the initial fingerprmtmg gel then SSCP can be performed on the eluted product of interest without reamplification. For this procedure we have found a 2-4 d autoradiography with an intensifying screen is needed to vlsuahze the strands. However, this strategy entirely avoids the problems associated with reampllficatlon of the initial PCR product until the product of interest is highly purified on the SSCP gel. After removal from the SSCP gel the product(s) are reamplified for 20 cycles or less and either used directly for sequencing, or for cloning then sequencing. This latter protocol is presented here. 1. When products of Interest are identified the RAP-PCR reactions of interest and suitable control RAP-PCR reactions are loaded in multiple adJacent lanes following the fingerprinting protocol m Section 3.1 , steps 6-8. The purpose of this preparative gel is to resolve large quantities of the PCR products of interest (see Note 8). RadIoactive ink or fluorescent markers are attached to the dried gel to allow reorlentatlon of the film with the gel. 2. The band of interest IS excised from this preparative gel, and the gel re-exposed to X-ray film to reveal a clear swathe where the band was excrsed. 3 The piece of dried acrylamide containing the band of Interest IS placed in 100 & TE, and heated to 68°C for an hour then left overnight at room temperature to allow the PCR product to diffuse out. 4. The eluted material IS ethanol precipitated. The incompletely polymerized acrylamide acts as an effective carrier 5. The radioactive pellet is directly dissolved m 4 pL 10 mA4 NaOH dye solution, denatured for 2 min at 94”C, placed on ice for 5 mm, then loaded on an MDE gel for resolution of single strand conformation polymorphisms 6 After electrophoresls, the SSCP gel IS dried (see steps 6, and 7 in Section 3.1,) and autoradlographed using an intenslfymg screen. Because the band of interest IS of higher intensity than background bands, the darkest band on the SSCP gel corresponds to the desired product Frequently, this band will resolve mto two strands. Background bands, of relatively low intensity and of an entirely dlfferent sequence, will resolve to many positions on the gel, and are often not visible after short autoradiographic exposure Usually double stranded product runs much faster than the single stranded products of interest An example is shown in Frg 2 7. Once resolved by SSCP, the product can be excised and eluted from the gel, PCR amplified for 20 cycles or less, and used for direct sequencing or for cloning

18

McClelland et al.

SlS2S3S4S5S6S *rl

Fig. 2. SSCP of products isolated from a RAP-PCR gel. Products of interest were cut from the dried fingerprinting gel shown above. The products were eluted overnight, ethanol precipitated, and the entire pellet was mixed with 10 mM NaOH buffer, denatured, and loaded on a 1X MDE native acrylamide gel (JT Baker Inc., Phillipsburg, NJ). A denatured MspI digest of pBR322 was used as a marker. After electrophoresis the gel was dried and exposed to X-ray film using an intensifying screen at -80°C. Subsequently each of the single stranded products was isolated, reamplitied, cloned, sequenced, and deposited in the GenBank database.

19

Fingerprinting

3.3. Confirmation

of Differential

Expression

Not all products that are characterized from RNA fingerprmtmg gels are actually differentially expressed. The sources of false positives m&de “sporadic” PCR products generated during fingerprmtmg, although these are usually ehmmated by comparmg two RNA concentrations per sample. In addition, the wrong product may reampltfy efficiently from the excised fragment of interest, although this IS usually elrmmated by performing SSCP before reamplrficatton. Some independent method such as Northern blotting or RT-PCR must be used to confirm the product arises from a differentially expressed transcript We have developed streamlined ways to confirm differential expression One method that has proved both effective and relatively simple IS based on RT-PCR A patr of spectfic primers with melting temperatures of 60°C or more (typically 18 bases or longer) are derived from the sequence of the RAP-PCR product of interest The spectftc prtmers are both used at a concentration of 2 l&J for first strand reverse transcrtptton for each of the RNAs where differential expression needs to be examined (usmg the buffer mcubatton condtttons m Section 3 l., steps 2 and 3) PCR is performed using low stringency (35°C) annealing, m a manner similar to RAP-PCR, at two template concentrattons (typically, 100 and 200 ng). The buffer condtttons and amphficatton profile m Section 3 1 , steps 4 and 5 are used except that the primers are at a concentratton of 1 5 cul/%PCR 1sperformed m separate tubes for 15,20, and 25 cycles to ensure that the spectfic product of interest can be easrly detected, regardless of mtttal abundance. Electrophorests and autoradtography are as m Section 3.1 , steps 6-8 Thts protocol generates a product of the expected size for the transcript of mterest. However, it also generates other products from arbttrartly prtmed PCR Most of these arbitrary products are derived from RNAs that are not dtfferenttally expressed. Such products act as a convenient internal control for the level of amphfication and the quality of the PCR reaction The desired PCR products and a swathe of control products are each cut from each lane of the gel and counted m scmtillation fluid If a fluortmager is available this would be the preferred tool. When the mass of the PCR product from the gene of interest is normahzed against the mternal controls an estimate of the relative expresston of the gene m the different samples is obtained One caveat m this method IS that the Cot effect (see Note 6) causes differences observed between products to underestimate the true differences between the startmg RNA samples However, this discrepancy is reduced at lower cycle numbers so the amplificatton IS best sampled at a number of PCR cycles or at a number of starting RNA concentrations The most useful data is then derived for the lowest number of cycles or lowest RNA concentratton that 1scompatible wtth the method employed for measuring product formation.

20

McClellana et al.

4. Notes 1. Experimental design: There are multiple steps needed to detect and characterize a differentially expressed transcript and sampling is biased towards more abundant transcripts (see Note 4). Thus, for many questions regarding differential gene expression there may be other methods that are better suited, such as differential screening or subtractive hybridization. However, unlike these methods, RAP-PCR allows many RNA samples to be compared in parallel (25). If eight different RNA samples are compared and the levels of each transcript is unchanged or up- or downregulated in each RNA sample then there are almost 3* (more than 6000) possible permutations of gene expression that are surveyed. This calculation does not take into account the further division of expression profiles to account for large versus small increase or decreases in gene expression. Most of the vast number of observable regulatory categories will not exist but, nevertheless, examples of genes that fall into rather sophisticated regulatory categories can be searched for. Furthermore, genes that fall into totally unexpected categories may be found. 2. Longer primers can be used. For an 1%mer primer the annealing temperature during PCR is changed to 45°C and Tuq polymerase Stoffel fragment is replaced by AmpliTaq. One advantage of longer primers is that they can easily accommodate restriction sites for subsequent cloning or can encode “motif’ sequences that may direct priming to transcripts encoding conserved amino acid sequences (26). Even 10-mer primers can encode short motifs. An example is the Tryp 1- primer (5’-GTGGCGTTGAT) in Fig. I that is a conceptual translation of a tyrosine phosphatase conserved amino acid motif. 3. Fingerprinting multiple RNA concentrations: As is true for DNA fingerprints, a few products in each RNA fingerprint can be sensitive to RNA concentration and quality. For this reason we perform RNA fingerprints using at least two RNA concentrations that differ by twofold. This strategy allows those PCR products that are not reproducible at both RNA concentrations to be eliminated from further consideration, Fingerprinting duplicates of each RNA from each experimental condition at the same concentration is less effective because this does not necessarily control for variation that is concentration- and quality-dependent. Fingerprinting separate RNA preparations for each sample condition is also less effective because differences in RNA quality between different experimental conditions are not controlled. 4. Sampling efficiency: One of the primary limitations of arbitrarily primed RNA fingerprinting methods is the fact that the probability of being able to visualize a PCR product derived from a particular transcript is a function of both the quality of the match of the primer with two sites in the template and the abundance of the transcript. For example, if two templates have the same match and are equally efficiently amplified but differ by loo-fold in abundance, the ratio of these products will remain 100: 1 during most of the reaction. As a consequence, less abundant transcripts will be more difficult to see on a fingerprint or may not be visible at all. Calculations that assume each transcript can be sampled with equal effi-

Fingerprinting ctency do not take into account the fact that rarer transcripts will be on average much more difficult to vtsualize. Thus, a calculation of the number of fingerprints needed to sample, say, 95% of the transcripts that assumed normalized sampling and equal visibility of products derived from rare transcripts would be a considerable underestimate of the actual number of fingerprints needed One logical response to these considerations is to ensure that any expertment that is to be performed using this method does not require efficient coverage of rarer transcripts. 5. One method we have used in an attempt to improve sampling of rarer transcripts is to reamplify the RNA fingerprint with a primer that contains an extra arbitrarily selected base at the 5’ end (or more than one such base) (26). The Idea is to select a subset of the initial fingerprint for reamplification, Including PCR products, which are not visible on the initial fingerprint because they originate from the complex class of rare transcripts. When primers of 18 bases m length are used for the initial fingerprinting and nestmgs, this method can generate a new fingerprint when the “nesting” includes up to three bases However, the fingerprint becomes less reliable as the nesting length is increased, presumably because products of extension from poorer and less efficient matches are also amplified from the background of the fingerprint. Interestingly, lo-mer primers are too short for nesting presumably because annealing must take place at so low a temperature that extension occurs from all of the initial fingerprint products, regardless of the 3’ match The method is also useful if the RNA used for fingerprmtmg is so limited that the generation of further fingerprints from the first fingerprint would be desirable to conserve precious RNA. 6. The “Cot effect”: In later cycles of PCR the concentration of the more abundant PCR products is sufficiently high that there is considerable product self-annealing that occurs in the lower temperature steps of the PCR cycles. This has the effect of preferentially slowing the amplification efficiency of the more abundant PCR products. Two consequences can be expected. First, less abundant products have a chance to gain some ground and become visible. This advantage of rarer products may slightly mitigate against the fact that rarer products are otherwise less easily seen in fingerprints. Second, differences visible between different RNA samples may be partially erased, particularly for the more abundant products. Thus, surprismgly, differences visible between lanes may actually underestimate the true differences in abundances of the transcripts. We have called this phenomenon of product self-annealing the Cot effect in reference to the dependence of annealing on the initial concentration and time. It is interesting to note that intentionally increasing the Cot effect could be desirable in some circumstances. For example, if one is interested in “normahzing” a mtxture of PCR products. Possible application would include sampling the entire complexity of genotypes for an environmental sample. A class of genes of interest would be PCR amplified under conditions where homologs would rehybndize, allowing other amplified genes from rarer organisms m the sample to become relatively enriched in the sample. Another example might be the nor-

McClelland et al.

22

maltzatton of a cDNA library, although crossover PCR might be a concern In both cases the PCR reaction would be held at 60-85°C for many mmutes, or perhaps even hours, at each cycle to block amphficatton of the more abundant products 7 About 0 5 l&f of the first primer 1scarried over into the second reaction Adding 4 @4 of a second primer for the second strand syntheses results prtmartly m PCR products that have the first primer at the 3’ end of the sense strand of the transcript and the second primer at the 5’ end Thus, the sense ortentatton of the products 1sgenerally known Thts phenomenon is accentuated by the observatron that PCR products that have the same primer at each end seem to be out-competed by products that have different primers at each end (27) The use of an arbitrary primer m the RT reaction allows PCR products to be derived from internal parts of a transcript, mcludmg the protein coding region. This can help later when determinmg the nature of the transcript Also, transcripts that are not polyadenylated, such as bactertal RNAs, can be sampled by thus method. 8 If the origmal fingerprmt IS no longer radioactive then It can be made radioactive again by reamphficatron of 1 pL of the reaction for a further five PCR cycles m 10 & of fresh RAP-PCR amplification mixture

Acknowledgments Thts work was supported in part by grants A134829, NS33377, and CA68822 from the National Institutes of Health and by a generous gift from Sidney Kmnnel.

References Welsh, J and McClelland, M. (1990) Fingerprintmg genomes using PCR wtth arbttrary primers Nuclezc Aczds Res 18, 7213-72 18 Wtlltams, J G , Kubehk, A R , Ltvak, K J , Rafalskt, J. A , and Tmgey, S. V (1990) DNA polymorphtsms amplified by arbttrary primers are useful as genettc markers Nuclezc Acids Res 18, 653 l-6535 Rerter, R. S , Willrams, J. G., Feldmann, K A., Rafalskt, J A , Tingey, S. V , and Scolmk, P A. (1992) Global and local genome mappmg m Arabzdopszs thalzana by using recombinant inbred lines and random amphfied polymorphic DNAs Proc Nat1 Acad Sa USA 89,1477-1481 Welsh, J , Petersen, C , and McClelland, M (1991) Polymorphtsms generated by arbttrartly prrmed PCR in the mouse. application to strain rdenttficatton and genetic mappmg Nuclezc Aczds Res 19, 303-306 Al Janabt, S M , Honeycutt, R J , McClelland, M., and Sobral, B W. (1993) A genetic linkage map of Saccharum spontaneum L ‘SES 208’ Genetics 134, 1249-1260 Birkenmeter, E. H , Schneider, U , and Thurston, S J (1992) Fingerprmtmg genomes by use of PCR with prrmers that encode protem mottfs or contam sequences that regulate gene expression [published erratum appears m Mamm Genome 1993,4(2) 1331. Mamm Genome 3,537-545

Fingerprinting

23

7 Mlchelmore, R. W , Paran, I , and Kesseh, R V (1991) Identtficatlon of markers lmked to disease-resistance genes by bulked segregant analysts* a rapid method to detect markers m specific genomic regions by usmg segregating populattons Proc Natl Acad. Scz. USA 88,9828-9832.

8 Mathteu-Daude, F , Stevens, J., Welsh, J., Tibayrenc, M , and McClelland, M. (1995) Genetic diversity and population structure of Trypanosoma brucet clonality versus sexuality. Mol Blochem Parasltol 72, 89-101 9. O’Rourke, M. and Sprat& B. G (1994) Further evidence for the non-clonal population structure of Nezsserza gonorrhoeae extensive genettc dtverstty within ISOlates of the same electrophoretlc type Mzcrobzology 140, 1285-1290 10 Fukatsu, T and Ishtkawa, H (1994) Dtfferenttatlon of aphid clones by arbttrarlly primed polymerase cham reaction (AP-PCR) DNA fingerprmtmg Mel Ecol 3, 187-192 11. Levttan, D R. and Grosberg, R K (1993) The analysts of paternity and mater-t-my m the marme hydrozoan Hydvactznla symbzolonglcarpus usmg randomly amplltied polymorphic DNA (RAPD) markers MoZ Ecol 2,3 15-326 12 Tamate, H B., Shlbata, K , Tsuchtya, T , and Ohtalshi, N (1995) Assessment of genettc variations within populations of Sika deer m Japan by analysts of randomly amplified polymorphic DNA (RAPD) Zoolog Scz 12,669-673 13 Chapco, W., Ashton, N W , Martel, R K , Antomshyn, N., and Crosby, W L (1992) A feaslblllty study of the use of random amplified polymorphtc DNA m the population genetics and systematics of grasshoppers Genome 35, 569-574. 14 Fang, F C , McClelland, M , Gumey, D G , Jackson, M M , Hartstem, A I, Morthland, V H , Davis, C E , McPherson, D C , and Welsh, J (1993) Value of molecular epldemlologrc analysts in a nosocomtal methlclllm-resistant Staphylococcus aureus outbreak [see comments] JAMA 270, 1323-1328. 15 van Belkum, A , van Leeuwen, W , Kluytmans, J., and Verbrugh, H (1995) Molecular nosocomtal epidemiology* high speed typmg of mlcroblal pathogens by arbitrary primed polymerase chain reaction assays Infect Control Hasp Epldemzol

16,658466

16 Tang, Y. J., Houston, S T , Gumerlock, P. H , Mulhgan, M E , Gerdmg, D. N , Johnson, S , Fekety, F R., and Silva, J J. (1995) Comparison of arbitrarily primed PCR with restrmtton endonuclease and lmmunoblot analyses for typmg Clostrldlum d&f?clle isolates J Clan Mlcroblol 33, 3169-3173 17. Coelho, A., Vicente, A. C , Baptista, M A , Momen, H , Santos, F A , and Salles, C A. (1995) The distinction of pathogenic Vtbrto cholerae groups using arbltrarlly primed PCR fingerprints. Res Mlcroblol. 146, 67 l-683 18. van Belkum, A , Kluytmans, J , van Leeuwen, W., Bax, R , Qumt, W., Peters, E , Fluit, A , Vandenbroucke Grauls, C , van den Brule, A , Koeleman, H , et al (1995) Multtcenter evaluation of arbitrarily primed PCR for typmg of Staphylococcus aureus strains J Clwz. Mlcroblol 33, 1537-1547 19. Madico, G., Akopyants, N. S , and Berg, D E (1995) Arbitrarily primed PCR DNA fingerprmtmg of Escherzchza co110157.H7 strams by using templates from boiled cultures J Clrn Mzcroblol 33, 1534-1536

McClelland et al.

24

20. Perucho, M., Welsh, J., Peinado, M. A., Ionov, Y., and McClelland, M. (1995) Fingerprinting of DNA and RNA by arbitrarily primed polymerase chain reaction: apphcatlons m cancer research Methods Enzymol 254,275-290 21. Welsh, J., Chada, K., Dalal, S S., Cheng, R., Ralph, D., and McClelland, M. (1992) Arbitrarily primed PCR tingerprmting of RNA. Nucleic Acids Res 20, 4965-4970. 22 Liang, P. and Pardee, A. B. (1992) Differential drsplay of eukaryotic messenger RNA by means of the polymerase chain reaction [see comments] Science 257, 967-971. 23. Wong, K. K. and McClelland, M. (1994) Stress-inducible gene of Salmonella typhlmurlum identified by arbitrarily primed PCR of RNA Proc Natl. Acad SCL USA 91,639-643

24. Hayashi, K. and Yandell, D W. (1993) How sensitive is PCR-SSCP? Hum Mutat 2,338-346

25. McClelland, M., Ralph, D., Cheng, R., and Welsh, J. (1994) Interactions among regulators of RNA abundance characterized using RNA tingerprintmg by arbitrarily primed PCR. Nuclezc Acids Res 22,44 19-443 1 26. Ralph, D., McClelland, M., and Welsh, J. (1993) RNA fingerprmtmg usmg arbltrarlly primed PCR identifies drfferentially regulated RNAs in mink lung (Mv 1Lu) cells growth arrested by transforming growth factor beta 1 Proc Nut1 Acad Sci. USA 90, 10,71O-10,714 27. Welsh, J. and McClelland, M. (1991) Genomic tingerprmting using arbitrarily primed PCR and a matrix of painvlse combinations of primers. Nuclex Acids Res 19,5275-5279.

3 Differential Display Using Random Hexamer-Primed cDNA, Motif Primers, and Agarose Gel Electrophoresis Patrick J. Donohue, Debbie K. W. Hsu, and Jeffrey A. Winkles I, Introduction Polypeptide growth factors stimulate cellular proliferation by bmdmg to the extracellular domam of transmembrane receptors and thereby activating mtracellular signal transduction pathways. One cellular response to mitogenic stimulation is the sequential transcriptional induction of specific nuclear genes encoding proteins of diverse functions (reviewed in refs. 1 and 2). Many of these proteins are likely to be required for DNA replication and cellular division. As an approach to identify novel gene products involved in polypeptide growth factor signaling, we are studying tibroblast growth factor (FGF)- 1-inducible gene expression m murine NIH-3T3 cells. FGF-1 is a member of a family of structurally-related mitogens than can promote cellular proliferation, migration and differentiation (reviewed in refs. 3’ and 4). Its biological effects are mediated by protein tyrosine kinase cell surface receptors present on most cell types. FGF-1 is likely to be involved in the pathogenesis of several human diseases, including atherosclerosis and cancer. We reported in 1993 that a differential display approach using agarose gel electrophoresis could be used successfully to identify genes expressed following FGF-1 treatment of serum-starved NEH-3T3 fibroblasts (5). This approach is conceptually similar to the mRNA differential display and RNA fingerprmtmg methods described by other groups (reviewed in refs. 6 and 7). However, in comparison to these approaches, in our method (1) cDNA is synthesized usmg random hexamer primers; (2) polymerase chain reaction (PCR) assaysare performed using sense and antisense oligonucleotide primers, usually degenerate From

Methods m Molecular Bology, Vol 85 D/fferentral Edlted by P Llang and A B Pardee Humana

25

D/splay Methods and Protocols Press Inc , Totowa, NJ

Donohue, Hsu, and Winkles

26

m sequence, which are designed to amplify cDNA templates encoding proteins with particular structural motifs; and (3) ampllficatlon products are displayed using agarose gel electrophoresls and ethldmm bromide staining (reviewed m ref. 8). This method does not require the use of radlolsotopes nor the potent neurotoxms, acrylamlde and bu-acrylamlde. However, It is hmlted by the relatively low resolution of agarose gels and the Inability of fluorescent dye staining to detect amplification products derived from relatively rare cDNA templates. In our mltlal series of experiments, 30 cDNA fragments were isolated and 25 of these were successfully reamphfied and cloned mto a plasmld vector. When used as probes m Northern blot hybridization experiments, 15 of the 25 cDNAs detected transcripts that were expressed at an increased level m FGF- 1-stimulated cells. DNA sequence analysis revealed that 13 of the 15 cDNAs were unique and that four of the 13 cDNAs were amplified when a single ollgonucleotlde functioned as both a sense and antisense primer. Furthermore, although our initial goal was to use the motif primers to enrich for differentially expressed members of particular gene families, the majority of the FGF- l-inducible genes characterized to date do not encode protems with the targeted motifs ($9-11). This 1s because, under the PCR condltlons used, many of the motif primers were able to anneal to and prime cDNA templates with a relatively low degree of sequence identity. However, m at least one case, the targeting aspect of the approach descrtbed here was successful (12). In this report, we outline our basic differential display technique and note vanatlons performed by us and also described by others.

2. Materials 2. I. RNA Isolation

from Tissue Culture Cells

1 RNA STAT-60 (Tel-Test “B,” Friendswood, TX). This solution should be stored at 4°C up to 9 mo and 1slight-sensitive. It contains phenol and guamdlmum thlocyanate and should be handled wearing gloves and a lab coat. Avold breathmg vapor. 2 Chloroform. This should also be handled using the precautions described m item 1 3. Isopropanol. 4. Ethanol

2.2. cDNA Synthesis 1. Moloney murme leukemia virus (M-MLV) reverse transcrlptase (Life Technologies, Galthersburg, MD), 200 U/pi. Store at -20°C (not m a frost-free freezer) 2. 5X Reverse transcrlptase buffer: 250 mM Tns-HCl, pH 8.3, 375 mM KCl, 15 mMMgC1, 3. Dlthiothreltol, O.lM. 4 dNTP mix: 1 25 mM of each dNTP (Boehringer Mannhelm, Indianapolis, IN) 5 RNasm ribonuclease inhibitor (Promega, Madison, WI), 33 U/@ Store at -2O’C (not In a frost-free freezer). 6 Random hexamer (pd[N]& primers (Boehringer Mannhelm), 50 ng/&

Differential D/splay Vauatlons Table 1 Motif Oligonucleotide Motif Protein tyrosme kmase Zmc linger Leucme zipper

Src homology-2

Primers

Ammo acid sequence IHRDL DVWSFG GQKPYEC HQRIHTG LEEKATQL LEEKATQL LEEKATQL LEEKATQL FLVRESET VKHYKIR FLVRESET VKHYKIR

27 Used for PCR Amplification DNA strand

Sense Antisense Sense Antisense Sense Sense Antisense Antisense Sense Sense Antisense Antisense

Primer sequencea b CGGATCCACMGNGAYYT GGAATTCCAWAGGACCASACRTC

GGNGAGAARCCCTWYGARTG CCHGTGTGARTCCTCTGRTG CTGGAGGAGAAGGYGRCCCAGCT CTGGARGMNVAGRHSRMSMMGCT AGCTGGGYCRCCTTCTCCTCCAG AGCKKSKYSDYCTBNKCYTCCAG TTCCTGGTGCGGGAGTCTGAGACC GTGAAGCACTACAAGATCCGG GGTCTCAGACTCCCGCACCAGGAA CCGGATCTTGTAGTGCTTCAC

aPrlmer sequences are 5’ to 3’, addltlonal nucleotldes used for restrxtlon enzyme recogmtlon are m bold 6Degenerate bases m the primers are abbreviated as recommended by a nomenclature committee (21)

2.3. PCR 1. Taq DNA polymerase (Boehrmger Mannhelm), 5 U/pL. Store at -20°C (not m a frost-free freezer) 2 10X PCR buffer 100 mMTris-HCl, pH 8 3,500 mM KCI, 15 mMMgC1, 3. dNTP mix. 1 25 mA4 of each dNTP (Boehrmger Mannhelm) 4 Motif ohgonucleotide primers (Table 1); 0 5 ug/pL

2.4. Agarose

Gel EIectrophoresis

1 Agarose (Life Technologies) 2. 1OX Tris-acetate (TAE) buffer 400 mA4 Tris-acetate, 10 mM EDTA 3. 10X DNA gel loading buffer 50% glycerol, 0 2% bromophenol blue, 0.2% xylene cyan01 4. Ethtdmm bromide (Sigma, St. Louis, MO), 10 mg/mL This fluorescent dye is stored at 4°C and is light-sensitive. It is a mutagen and may be carcmogemc/ teratogemc, therefore, it should be handled wearing gloves and ethidmm bromide-contammg solutions should be disposed of properly

2.5. PCR Product Isolation,

Reamplification,

and Cloning

1 PCR and agarose gel electrophoresis reagents described above m Sections 2 3 and 2.4 2 Plasmid pCRI1 (Invitrogen, San Diego, CA), 25 ng/pL

28


3 1OX Ligation buffer 600 mM Trts-HCI, pH 7 5,600 mA4 MgCI,, 500 mM NaCf , 10 mg/mL bovine serum albumin, 700 mM P-mercaptoethanol, 10 mM ATP, 200 mM dtthtothrettol, 100 mA4 spermidme 4. T4 DNA ligase (Invitrogen), 4 UIuL Store at -20°C (not in a frost-free freezer) 5 E colz DH5a competent cells (Life Technologies) Store in ahquots at -70°C 6 SOC medra. 2% bacto-tryptone, 0 5% bacto-yeast extract, 10 mA4NaCl. 2 5 mM KCl, 10 mM MgCl,, 10 mM MgSO,, 20 mM glucose This media is prepared by first combmmg the tryptone, yeast extract, NaCl and KC1 and then autoclavmg. A glucose stock solutton (2M) and a Mg2+ stock solution (2M, comprised of 1M MgCl, and 1M MgSO,) are then prepared and sterilized by filtration through a 22-pm membrane. Finally, the media, glucose and Mg2+ are combined and sterilized by filtration. 7. Luria broth* 1% bacto-tryptone, 0 5% bacto-yeast extract, 1% NaCl Sterilize by autoclaving. 8 Bacto-agar (Dtfco, Detroit, MI) 9 Ampicillm (Sigma), 50 mg/mL. 10. 5-Bromo-4-chloro-3-mdolyl-p-n-galactopyranostde (X-gal, Boehrmger Mannhelm), 40 mg/mL in NJ-dtmethylformamtde

3. Methods 3.1. RNA Isolation

from Tissue Culture Cells

1 Lyse frozen cell pellets (-5 x lo6 cells; see Note 1) m 1 mL of RNA STAT-60 by repetmve plpeting. Incubate at room temperature for 5 mm 2 Add 0 2 mL of chloroform, mix and place samples at room temperature for 5 mm 3 Centrifuge the homogenate at 15,OOOg for 15 mm 4. Transfer 0.5 mL of the upper, colorless aqueous phase to a new tube, add 0 5 mL tsopropanol, vortex, and place samples at room temperature for 10 min 5. Centrifuge at 15,OOOg for 10 min and discard supematant 6. Wash RNA precipitate once with 1 mL 70% ethanol by vortexmg Centrifuge at 15,OOOg for 10 mm. Discard supematant. 7 Dry RNA pellet briefly for l-2 mm m Savant Speed-Vat Concentrator. 8. Dissolve RNA m 20 & autoclaved distilled H20. An mcubatlon at 55°C for 15 min may be required for complete resuspenston. 9 Determine the RNA concentration by measurmg the A260 of an ahquot using a UV spectrophotometer (1 A260 U = 40 pg/mL) 10 Store RNA samples at -70°C (see Note 2).

3.2. cDNA Synthesis 1. Combme 1 pg of total RNA and H20 to a final volume of 14.5 &. Heat at 65°C for 5 min and then cool on ice. 2. Set up the following 50 pL reactton by adding the reagents m this order 14.5 p.L RNA, 10.0 pI. 5X reverse transcriptase buffer, 5.0 pL dithlothreitol, 16.0 pL dNTP mix, 2 0 uL random hexamer primers (see Note 3), 0 5 p,L RNasm, and 2.0 pL M-MLV reverse transcriptase

Different/al Display Variations

29

3 Incubate at 37’C for 1 h. 4 Terminate reactton by heating at 95°C for 10 mm 5 Add 200 pL of dtstilled H,O and store at -20°C (see Note 4)

3.3. PCR 1 Set up the followmg 50 ~JLreaction by adding the reagents in this order: 5.0 pL cDNA, 5 0 p.L 10X PCR buffer, 8.0 pL dNTP mix, 30.5 l.tL H,O, 0.5 pL sense motif primer (see Notes 5-7), 0 5 l.tL antisense motif pnmer, and 0.5 pL Tuq DNA polymerase. 2. PCR 1sperformed using the following condittons (see Note 8) and a Perkm-Elmer Model 9600 thermocycler: 94”C, 5 mm Imtial Denaturation Stage I(8 cycles) 94°C 30 s denaturation 46°C 30 s annealing 68’C, 30 s extension Stage II (24 cycles) 94”C, 30 s denaturation 58“C, 30 s annealing 72”C, 30 s extension Final primer extension 72”C, 10 mm 3. Store samples at 4°C.

3.4. Agarose

Gel Hectrophoresis

1. Add 1.8 g agarose per 100 mL 1X TAE buffer (see Note 9). Dtssolve by heating m mmrowave oven, cool, and pour mto gel tray. 2. Prepare 1X TAE running buffer containing 1.5 pg/mL ethidium bromide. 3. An aliquot (18 pL) of each amplification mixture is then combined with 2 $ of 1OX DNA gel loading buffer and carefully pipeted into a gel lane. 4. Gel electrophorests is performed at 50-60 V constant voltage, usually for 0.5-2 h. 5. The gel is then briefly destamed by soaking in H,O with agitation and differenttal display products are visualized by UV light illumination (see Notes 10 and 11).

3.5. PCR Product Isolation,

Reamplification

and Cloning

1. DNA fragments of interest are excised from the gel using a razor blade, placed m 1.5-mL Eppendorf tubes and frozen at -7O’C for 15 min 2. Centrifuge at 15,000g for 15 mm Collect supematant and place in new tube. 3. A PCR product reampllfication step is then performed to generate a sufficient amount of DNA for subsequent cloning. The reaction is set up as described m Section 3.3 using 5 pL of the gel-recovered DNA solution and the primer pair used in the imttal PCR PCR IS performed using the same conditions as m the initial amplificatton except Stage I is omitted 4. The DNA (18 pL) is then subjected to agarose gel electrophorests and recovered as described prevtously m steps 1 and 2 (see Note 12). 5. Set up the following IO-@., reaction by adding the reagents m this order: 6.0 p.L PCR product, 2.0 pL Plasmtd (pCRI1) DNA, I 0 uL 10X ligation buffer, and 1 .O pL T4 DNA ligase.

30


6 Incubate at 12°C overnight 7 Set up the bacterial transformatton by combmmg 2-3 pL of the ligation reactton wtth 50 & of competent cells 8 Incubate on tee for 30 mm Heat shock at 42°C for 1 mm Place on ice 9 Add 450 pL of SOC media to cells Incubate at 37’C with agitation for 1 h. 10. Prepare four bactertal plates (100 x 15 mm) per transformation by combmmg 100 mL Lurla broth with 1.5 g Bacto-agar and autoclavmg 11 Place agar solution m 60°C water bath until cooled Add 150 p.L amptctllm and 100 p.L X-gal, mix and pour 25 mL per plate. Place at room temperature unttl sohdrfied 12 Spread 125 & of transformed cells per plate. Incubate at 37°C overnight 13 Blue and white colomes should be visible. Pick several white colonies, grow m Lurta broth/ampicillin media and prepare plasmid DNA by standard techniques (see Note 13)

4. Notes 1 In our specific case, we were attemptmg to tdenttfy genes that were expressed followmg FGF-1 sttmulatton of quiescent murme NIH-3T3 fibroblasts Thus, RNA was Isolated from quiescent cells as well as from cells treated with FGF- 1 for either 2 or 12 h 2 It 1simportant to confirm that the RNA is undegraded prior to its use An altquot (5-10 pg) should be subJected to electrophorests on a 1 2% agarose gel contammg 2.2M formaldehyde and then 28s and 18s rRNA integrity examined by ethtdium bromide stammg/UV light tllummatton usmg standard procedures 3. Reverse transcription with random hexamers will generate cDNA molecuies representing several regions ofmost mRNA species. However, unless poly(A)+ RNA 1sused as the template, most of the cDNA populatton wtll be copied from rRNA, which comprises the majority of the cellular RNA mass. In some cases, we have also made cDNA using the RACE ohgo d(T)-double adapter prtmer described by Frohman (13) 4. We always determme whether the cDNA synthesis reaction was successful by performmg PCR using 2% of the product and either p-actin or FGF receptor-l oligonucleotrde primers as descrtbed (14) 5 Twelve motif oltgonucleottde primers are listed m Table 1. The degenerate protem tyrosme kmase domain primers are identical to those described by Walks (15) The degenerate zmc finger and leucme zipper domam prtmers were destgned by alignmg various mouse DNA sequences and establishmg a consensus sequence The SK homology-2 domam primers were designed usmg the ammo acid sequence altgnment figure presented in Koch et al (16) and codon utiltzatton data (27) PCR assays were performed usmg one pan of primers at a time (sense and antisense) and all of the possible combmattons. As menttoned in Secton 1 , often these motif primers behaved as arbttrary primers under the PCR condtttons used in our experiments. Several examples of the primer cDNA mteractions that we have noted are shown m Ftg. 1,

Differential

Display

31

Varia Cons

1. Low sequence IdentIty between pnmer and cDNA ~'-GTGAAGC~~T~~~~I;A;~IC:~~;~~' SH2 sense FR-9 TCCTTCTACCACAAGCTCCGG 2245 PTt(antIsense 5'-G~IpATTfCAAAGF;eF~e~A~e?~3' FR-28 TGAGACCAGGGAGACCAGCCATC

2265

1349

1371

2

High sequence Identtty, but sense primer acttng as antisense primer (or we versa) i!F sense 5'-GG?F+Q?AGCCCTTCGAGTG3' 11111,1111,1 FR-1 ACAGAGATGCCCTTCGAGTG 687

666

3. Sequence Identity to mRNA 5’- or 3’-UT region. P~antlSenSe 5'-GGf\ATTC$!$ipAyGACCAGACATC-3' I I , I I I I I FR-1 TCATTAACAAGGACACAGACATC 17

39

Fig. 1. Nucleotlde sequence identity between several of the motif primers and the correspondmg regions of the FR cDNA clones DNA sequence analysis has Indicated that the majority ofthe genes that were identified using the motif primers do not encode proteins contammg the targeted structural domains Examples illustrating several of the most common explanations for this finding are shown here. In general, under the PCR conditions employed, the selectivity of the motif primers was poor and thus they were actually functlonmg m a similar manner as long arbitrary primers 6 Motif primers have also been used by two other groups attemptmg to identify differentially expressed members of known gene families Stone and Wharton (18) used several different degenerate protein kmase primers or zmc finger primers in combmation with degenerate arbitrary primers in their PCR assays They were able to identify cDNAs encodmg a novel casein kmase I lsoform and several zinc finger domain-contammg proteins. More recently, Bayarsalhan et al. (19) used inosine-containing SPChomology-2 and homeobox domain primers m combination with an arbitrary pnmer m their PCR assays.They isolated several cDNA clones but did not report whether they encoded proteins with the targeted structural mottfs 7. In those cases when cDNA samples were prepared using the RACE oligo d(T)-double adaptor primer (see Note 3), individual motif sense primers are used m combination with an antisense RACE outer adaptor primer (I 3) One of the 13 cDNAs that we ldentlfied by differential display was amplified using this RACE primer approach (5) 8 Some of our PCR assays have been performed using the followmg Stage I and II condltlons. 94”C, 30 s denaturation Stage I (8 cycles) 4O”C, 30 s annealing 68”C, 30 s extension Stage II (3@-40 cycles) 94”C, 30 s denaturatlon .54”C, 30 s annealing 72”C, 30 s extension


32 FGFM

0

’

2h

1 12h

’

ZF/PTK

Primers

- .-

281/271 234194 -

Fig. 2. Identification of an FGF- 1-inducible mRNA by differential display. RNA isolated from quiescent or FGF- 1-stimulated NIB-3T3 fibroblasts was converted to cDNA using random hexamer primers. PCR was then performed using a degenerate sense zinc finger (ZF) primer and a degenerate antisense protein tyrosine kinase (PTK) primer (top panel). As a control for the cDNA synthesis, PCR was also performed using sense and antisense FGF receptor (FGFR)- 1 primers (bottom panel). Amplification products were displayed using agarose gel electrophoresis and ethidium bromide staining. DNA size markers (M; in bp) are shown on the left. The pattern of amplified cDNAs obtained using RNA from quiescent or FGF-l-stimulated cells was similar except for an approx 700~bp DNA fragment, noted with an arrow in the top panel, that was present in the 12-h poststimulation lane. This fragment, termed FR- 1, was excised from the gel, reamplified, and cloned. 9. We have also displayed PCR amplification products on either 2% agarose gels or 4% agarose gels prepared using a 3:l ratio of NuSieve GTG low melt agarose (FMC Bioproducts, Rockland, ME) and standard agarose. 10. A photograph depicting a typical differential display gel (1.8% agarose) is shown in Fig. 2. 11. Two other laboratories have also successfully identified differentially expressed genes by differential display using agarose gel electrophoresis and ethidium bromide staining (Z9,20). 12. We have had poor success using this DNA as either a template for DNA sequence analysis or as a probe for Northern blot hybridization analysis. Therefore, we

Differential

Display Variations

33

FR-1

rRNA

Fig. 3. Northern blot hybridization analysis of FR- 1 mRNA expression in FGF- lstimulated flbroblasts. Northern blot analysis was performed to confirm the differential display results shown in Fig. 2 indicating that FR-1 mRNA levels were elevated in FGF- 1-stimulated cells. Quiescent NIH-3T3 cells were either left untreated (NT, no treatment) or treated with FGF- 1, FGF- 1 and actinomycin D (Act.D), or actinomycin D alone for 8 h. RNA was isolated and equivalent amounts of each sample, analyzed by Northern blot hybridization. The bottom panel is a photograph illustrating the relative amounts of 28s rRNA in each gel lane. FGF-1 induced FR-1 mRNA expression; furthermore, this effect is due, at least in part, to increased FR-1 gene transcription, since it does not occur if RNA synthesis is inhibited by actinomycin D.

routinely clone cDNAs isolated by differential display using the TA Cloning System (Invitrogen). 13. We always perform Northern blot hybridization analysis with purified cDNA insert as our first experiment in order to confirm that the gene under study is in fact differentially expressed. Our Northern blot hybridization protocol is described in Hsu et al. (I 1) and a representative result is shown in Fig. 3.

Acknowledgments This work was supported in part by National Institutes of Health Research Grants HL-39727 and HL-547 10. P. J. D. performed this work in partial fulfillment of the requirements for the degree of Doctor of Philosophy from the Graduate Genetics Program, George Washington University, Washington, DC. D. K. W. H. was supported in part by National Institutes of Health Training Grant HL-07698. We thank W. H. Burgess for providing the FGF-1, K. Wawzinski and D. Weber for excellent secretarial assistance, and G. F. Alberts for reviewing the manuscript.

34


References 1. Willtams, G. T., Abler, A S., and Lau, L. F (1992) Regulation of gene expression by serum growth factors, m Molecular and Cellular Approaches to the Control of Prolzferation and Dtfferentzatton (Stem, G S. and Lian, J. B , eds ) Academic, Orlando, FL, pp. 115-161 2 Muller, R , Mumberg, D , and Lucibello, F C (1993) Signals and genes m the control of cell-cycle progression Brochtm Bzophys Acta 1155, 15 l-1 79 3. Burgess, W H. and Wmkles, J A (1996) The fibroblast growth factor famtly multifunctional regulators of cell proliferation, in Cell Proltferatton zn Cancer Regulatory Mechanisms ofNeoplasttc Cell Growth (Pusztai, L , Lewis, C E , and Yap, E , eds ), Oxford University Press, Oxford, pp 154-2 17. 4 Fernig, D. G. and Gallagher, J T (1994) Fibroblast growth factors and their receptors: An information network controllmg tissue growth, morphogenests and repair Prog Growth Factor Rex 5, 353-317 5 Hsu, D K W , Donohue, P J., Alberts, G. F , and Winkles, J. A (1993) Fibroblast growth factor-l induces phosphofructokmase, fatty acid synthase and Ca2+-ATPase mRNA expression in NIH 3T3 cells Bzochem Bzophys. Res Commun 197,1483-1491 6 McClelland, M., Mathieu-Daude, F , and Welsh, J (1995) RNA fingerprmtmg and differential display using arbitrarily primed PCR. Trends Genet. 11,24 l-246. 7 Liang, P , and Pardee, A B. (1995) Recent advances m differential display Curr Open Immunol 7,274-280. 8. Winkles, J. A, Donohue, P J , Hsu, D K. W , Guo, Y , Alberts, G. F , andpeifley, K A. (1995) Identification of FGF-1 -mducible genes by differential display, in Cardtovascular Dtsease 2 Cellular and Molecular Mechanums, Preventton and Treatment (Gallo, L. L , ed ) Plenum, New York, pp 109-120 9. Donohue, P. J., Alberts, G. F., Hampton, B. S., and Winkles, J. A (1994) A delayed-early gene activated by tibroblast growth factor-l encodes a protein related to aldose reductase. J Biol Chem. 269,8604-8609 10 Hsu, D K W., Guo, Y., Alberts, G F., Petfley, K A , and Winkles, J A (1996) Fibroblast growth factor-l -inducible gene FR-17 encodes a nonmuscle a-actmm isoform. J Cell Physiol 167,261-268 11. Hsu, D K. W , Guo, Y., Alberts, G. F., Copeland, N G., Gilbert, D J., Jenkins, N A., Peifley, K A, and Winkles, J. A, (1996) Identification of a murme TEF1-related gene expressed after mitogemc stimulation of quiescent fibroblasts and during myogemc dtfferentiation J. Btol. Chem 271, 13,786-13,795 12 Donohue, P. J., Alberts, G F., Guo, Y , and Winkles, J A (1995) Identification by targeted differential display of an immediate-early gene encoding a putative serinelthreonine kinase J B1o1 Chem 270, 10,351-10,357. 13 Frohman, M A. (1993) Rapid amphfication of cDNA ends (RACE)* User-friendly cDNA cloning Amplzficatzons 5, 1l-l 5 14 Brogi, E., Winkles, J. A., Underwood, R , Clmton, S. K., Alberts, G F., and Libby, P (1993) Distinct patterns of expression of fibroblast growth factors and their receptors m human atheroma and non-atherosclerotic arteries. Assoctatton of

Differential

15 16

17. 18

19

Display Variations

35

acidic FGF with plaque microvessels and macrophages J Clin Invest 92, 2408-2418 Walks, A. F. (1989) Two putative protein-tyrosme kmases Identified by applrcanon of the polymerase chain reaction. Proc Nat1 Acad Scl USA 86,1603-1607 Koch, C. A., Anderson, D , Moran, M F., Ellis, C., and Pawson, T. (1991) SH2 and SH3 domains: Elements that control mteracttons of cytoplasmtc signaling protems Science 252, 668-674. Lathe, R. (1985) Synthetic oligonucleotide probes deduced from ammo acid sequence data* Theoretical and practical considerations. J Mel Bzol 183, 1-12. Stone, B and Wharton, W (1994) Targeted RNA fingerprintmg* The cloning of differentially-expressed cDNA fragments enriched for members of the zmc finger gene family. Nuclezc Acids Res 22,2612-2618 Bayarsarhan, D., Enkhmandakh, B., Farrell, C , and Lukens, L (1996) Rapid tdentrticatton of a novel chondrocyte-specific gene by RNA differential display Blochem Blophys Res Commun 220,449-452

20 Sokolov, B. P and Prockop, D J (1994) A rapid and simple PCR-based method for tsolatton of cDNAs from differentially expressed genes Nuclex Acids Res 22,4009-40 15 21 Dixon, H B F , Blelka, H , Cantor, C R , Llebecq, C , Sharon, N , Velrck, S F , and Vhegenthart, J F G (1986) Nomenclature for mcompletely specified bases m nucleic acid sequences J Bzol. Chem 261, 13-17

4 Fluorescent

Differential

Takashi Ito and Yoshiyuki

Display

Sakaki

1. Introduction Since their first mtroduction in 1992 (1,2), differential display (DD) and its relatives have been used extensively in various fields of biology where the identification of differentially expressed messagesare of particular interest and importance (for review, see ref. 3). These new techniques have several unique advantages over the conventtonal methods such as differential and/or subtractive hybrtdization-based ones. They can compare a number of samples m parallel to reveal transcripts of various behaviors, unlike subtractive hybridization techniques that essentially compare two samples unidirectionally. They can detect transcripts of low abundance as well as those showing subtle changes, that would be overlooked by conventional hybridization-based methods. Also, the techniques require only a tiny amount of RNAs to start the analysis so that they can be applied to biological samples that are difficult or impossible to prepare in large amounts. Despite these attractive features, the methods have several drawbacks. For instance, since they are essentially random sampling approaches, considerable numbers of reactions (i.e., primer combinations) have to be tested to statistically cover the complex transcript population. If one assume the sampling 1sa completely stochasticevent, the number of reactions required (N is calculated as N=ln(l

-P)lln(l

-B/E)

where P is the probability desired, E is the number of expressed transcript species, and B is the number of bands per each reaction. Therefore, if each DD reaction amplifies 100 cDNAs (B = loo), one has to perform about 450 reaction to cover all the transcript in typical mammalian cells (E = 15,000) with 95% probability (P = 0.95). From

Methods m Molecular B/ology, Vol 85 D/fferent/al E&ted by P Llang and A B Pardee Humana

37


Ito and Sakakl

38

To run this scale of analyses with multiple samples, one has to Increase the speed of each analysis, m parttcular, that of electrophoresis, postrun gel processing and signal detection Furthermore, the safety problem inherent to radioactive DD would not be negligible m that scale (4). To overcome these drawbacks, we established two DD protocols compatible with fluorescent detection, termed fluorescent differential display (FDD) protocols S and L (S-7) Protocol S uses short arbitrary primers (1 0-mer) and low stringency PCR condition as the origmal DD protocol, but the design of anchor primers were modified so that one can obtain signals with enough mtensmes as well as satisfactory reproducibility The anchor primers currently used m protocol S are GT,,N (N = A, C, or G). They contam additional dGs at then S-ends, that are necessary for the generation of signals intense enough for fluorescent detection. (Addition of further nucleotides did not improve the signal anymore.) The protocol L uses primers of usual lengths (-20-mer) as upstream arbitrary primers. The anchor primers for protocol L are CCCGGATCCTt5N (N = A, C, or G). In the mittal cycle of PCR in protocol L, a low strmgency condition was used so that even a longer primer can nutlate second strand synthesis on multiple cDNAs. In the second and later cycles, the molecules tagged at their both ends with anchor and arbitrary primers are specifically amplified by means of high stringency PCR. The sequences added to the S-ends of the anchor primers are to increase their T, to be used m the high stringency PCR steps. Here we describe both FDD protocols using fluorescence image scanner, that ensure highly reliable DD analysis with much Improved speed and safety as demonstrated so far (5-7).

2. Materials 2.1. Reverse

Transcription

1 The anchor primers. FITC-GT,,N (N = A, C or G) for protocol S and FITCCCCGGATCCT1sN (N = A, C or G) for protocol L, were synthesized on ABI 392A DNA synthesizer using FluorPrime (Pharmacia, Sweden) The synthesized primers were dissolved m dtethylpyrocarbonate (DEPC)-treated water at the concentration of 50 @4. Anchor primers for PCR steps may be further purified by HPLC or PAGE

2. Preamplification

Kit (BRL, ME)* This kit contains Superscript 11reverse tran-

scriptase and other soluttons for reverse transcription

3. Thermal cycler: Since reverse transcription step uses vanous temperatures, we are using tube-type thermal cycler for the mcubatton 4 TE buffer 10 mA4 Tris-HCl, pH 8 0, and 1 mM EDTA

2.2. Polymerase 1 Arbitrary

Chain Reaction

primers: Arbitrary

lo-mers were obtained from Operon (Alameda, CA)

or synthesized in our laboratory (see Note 4), whereas various primers of 20 nt

Fiuorescent

Different/al

Display

39

long, that had been origmally designed for mdivtdual purposes m our lab, were recruited as arbitrary primers for protocol L They were adJusted to 10 fl and stored at -20°C 2. Taq DNA polymerase* We use GeneTaq DNA polymerase (Nippon Gene, Toyama, Japan), an N-terminally truncated Tuq DNA polymerase supplied with 1OX buffer and dNTP solution. To improve ampltfication of htgher mol-wt species, conventional Taq DNA polymerase from Perkm-Elmer or BRL 1sused with GeneTaq 3 Thermal cycler For a large number of PCR, those with 96-well microtiter plate format are convenient We routmely use Techne PHC-3 thermal cycler (Techne, UK) for PCR

2.3. Gel Electrophoresis 1 Gel solutton’ 6% polyacrylamide or LongRanger (FMC BtoProducts, ME) contaming 8 3M urea and 1X TBE buffer 2. Gel plates: The glass plates of low mtrmstc fluorescence are recommended. For the assembly of gel slab, we use 0 3.5-mm thick spacer and shark tooth comb 3 Image scanner We are usmg Vistra FluorImager SI (Molecular Dynamics, CA) for the scanning

3. Methods 3.7. Reverse

Transcription

1 Heat a thermal cycler to 70°C 2 Mix 8 0 l.tL of DEPC-treated water, 1 0 pL of Anchor Primer (50 CUM)and 1 0 pL of total RNA solution (2 5 pg/uL) m a 0 5-mL tube (see Note 1). 3. Place the tube m the thermocycler at 70°C for 5-10 mm 4 Durmg the mcubatton, prepare 2X RT solution by mixmg 2.0 pL of DEPC-treated water, 2.0 &of 10X PCR buffer, 2 0 pL of25 mMMgCl,, 2.0 &of 100 mMDTT, 1 0 cls, of 10 mM dNTP and 1 0 clr, of Superscript II (200 II/$) using a BRL preampltfication kit (see Note 2). Prepare enough amount, taking the pipeting loss mto account Followmg thorough but gentle mixing, place the tube on Ice 5 Place the heated tubes containing RNA and primers mto ice-water bath for a few minutes, and cool the thermal cycler to 25°C Following the brief spin, add 10 $ of 2X RT solutton mto each tube and mtx thoroughly by gentle pipetmg 6 Place the tubes m the thermal cycler, and incubate them for 10 mm at 25’C, 50 mm at 42°C and 15 mm at 70°C. 7. Following the above mcubatton, spm the tube briefly and add 80 & of TE. Store at -20°C until use

3.2. Polymerase

Chain Reaction

1 Mtx 13.0 (x N) $ ofdistilled water, 2 0 (x N) & of 10X GeneTaq buffer, 1.6 (x N) & of 2.5 mA4 dNTP, 0.2 (x N) pL o f anchor primer and 2 0 (x N) $ of cDNA solution (PCR mtx I). Calculate iV, taking into account the loss during repeated ptpetmg

Ito and Sakaki

40

2. Dispense 1.OpL of arbitrary primers ( 10 pmol) m each well of 96-well thermal plate 3. Make PCR mix II by mixmg 18.8 (x N) pL of PCR mix I, 0.1(x N) pL of GeneTaq DNA polymerase and 0 1 (x N) pL of AmphTaq DNA polymerase (see Note 3) 4. Dispense 19.0 pL of PCR mix II to each well containing 1.0 pL of arbitrary primers. 5 Overlay each well with a drop of mineral oil 6 Place the 96-well plate in thermal cycler, and run the followmg program (see Note 4). For protocol S: (94’C, 3 min + 4O”C, 5 min + 72”C, 5 mm) for 1 cycle + (95°C 15 s + 40°C, 2 mm + 72’C, 1 mm) for 20-24 cycles + 72°C 5 mm. For protocol L: (94’C, 3 min + 37°C 5 mm + 72”C, 5 min) for 1 cycle + (95°C 15 s + 55’C, 1 mm + 72’C, 1 mm) for 20-24 cycles + 72°C 5 mm.

3.3. Gel Electrophoresis 1 Prepare 6% polyacrylamide (or LongRanger) gel (200 x 330 x 0.35 mm) 2. Followmg polymerization, prerun the gel for 1 h at 1000 V. 3 Mix PCR products with isovolume of formamide dye solution, and heat at 90°C for 3 mm. 4. Load 5-6 pL of sample to each well of shark-tooth comb and run the gel at 1500 V for 1 h until bromophenol blue dye reaches to the bottom of the gel. 5 Remove the gel slab from the electrophoretic chamber, clean the glass plate and scan it by FluorImager SI using the high sensitivity mode This scan is to obtain images focusing on lower mol-wt bands (see Note 5 and Fig. 1). 6. Following the first scanning, place the gel slab to the electrophoretic chamber again and run for an additional hour until xylenecyanol dye migrates near the bottom of the gel. Scan the gel again to visualize higher mol-wt bands.

3.4. Molecular

Cloning of the Bands of Interest

1 Repeat the experiment using RNA batches different from those used m the first experiments to confirm the reproducibility of the behavior of the bands of your interest (see Note 6). 2. Run the gel of appropriate concentration and for an optimized duration to maximize the separation of the bands of interest Remove the upper glass plate, and scan the gel on FluorImager 3. Prmt the image usmg “actual size mode” and place the gel precisely onto the prmtout 4 Excise the band of your mterest using a razor blade. Following the excision, scan the gel again to see how precisely the band was excised. 5 Rinse the excised gel piece with distilled water. 6. Put the half of the gel piece mto a PCR tube, and cut tt into several smaller pieces using a flat loading tip just like a blade. 7. Add 100 pL of the PCR reaction mix II as described above except for the lower primer concentration (0.25 @4each) and subjected to the following thermal cycling

Fluorescent Differential

Display

Fig. 1. Identification of a transcript induced during the Alzheimer’s beta-peptideinduced apoptosis of neuroblastoma cells by the two step scanning procedure. Total RNAs were isolated from neuroblastoma LA-N-5 cells (lane 1) and those treated with Alzheimer’s beta-peptide for 0, 1, 6, 12,24, 30, and 48 h in lanes 2-8, respectively. The anchor primer used was GT&, and the arbitrary primers used from left to right were CTCACCGTCC, AAGCCTCGTC, GACGGATCAG and TTCCCCCCAG, respectively. While the left image was obtained following -1 h of gel running to focus on lower mol-wt species, the right one was taken from the same gel after an additional running for -1 h to put emphasis on higher mol-wt species. Note that the band shown by the arrowhead, that had been poorly resolved at the first scanning, was clearly separated in the right panel.

For protocol S: 94’C, 3 min + (94”C, 30 s + 4O”C, 2 min + 72°C 1 min) for -20 cycles + 72’C, 5 min. For protocol L: 94’C, 3 min + (94’C, 30 s + 55’C, 1 min + 72’C, 1 min) for -20 cycles + 72”C, 5 min. 8. Run a portion of reamplified product on a polyacrylamide gel to confirm the amount and purity of the product. 9. Following ethanol precipitation, clone the PCR product into the so-called T-vector.

/to and Sakakr

42 3.5. Selection

of Correct Clones

1 Suspend colomes of the transformants m 40 pL of L-broth 2. Use 1 0 pL of above suspenston m 20 pL of PCR using the same conditton described m Section 3 4 , step 7. 3 Run 0 l-l pL of the product m parallel with the ortgmal FDD reactton (“comigration test”). Select clones bearing inserts that precisely comtgrate wrth the band of Interest 4 Sequence the selected clones 5. Search restrtctron enzyme sites m the nucleottde sequences of the candidate clones. Digest the amplified mserts of the clones and the origmal FDD reaction with these restrtction enzymes, and run them m parallel to compare the digestion pattern of the target band and the mserts (“restrtctton test”) Select clones that are digested stmilarly to the band m FDD reaction for further analysrs 6 Confirm the expression pattern of the correspondmg transcrtpt by Northern blot hybndrzation, RNase protectton or quantitative RT-PCR assays 4. Notes Total RNAs prepared by various methods can be used for FDD We routinely use total RNAs prepared by a modified actd-guamdmmm-phenol-chloroform method usmg TRIzol reagents (BRL) accordmg to the supplier’s recommendatton As the contaminatron of genomtc DNA has severe effects on the DD pattern, m particular, m protocol S, we usually treated total RNA with RNase-free DNase (Promega) m the presence of placental RNase inhibttor or vanadyl-ribonuceoside complex to remove the residual contammatmg genomtc DNAs Although most RNA samples prepared by this method can be used for FDD, some may requrre further treatment. For Instance, we experrenced that RNAs from Xenopus early embryo requrres LiCl precipitation step for successful FDD (6) It should be noted that the use of RNAs of similar qualities is crtttcal for FDD We recommend to discard low quality RNA samples and re-prepare RNAs for the FDD We are using a kit from BRL for the reverse transcrtptton step, but others will give comparable results. Smce the first strand syntheses reqmres various temperatures, tt IS convement to use thermal cycler rather than preparing several water baths. Anchor primers used in protocols S and L are GT,,N and CCCGGATCCTlsN (N = A, C or G), respectively As they are incubated with RNA, special attention not to contammate with RNase has to be paid when preparing these primers We routinely use GeneTaq DNA polymerase (NIpponGene, Japan) for FDD analysis. This Tuq polymerase has a large deletron m its N-termmal portion and gives much stronger signals in lower mol-wt size range but weaker signals m higher mol-wt range than conventtonal Taq DNA polymerase. Thus, we use a cocktail of this enzyme and usual full-length Tuq polymerase to cover wider size range with higher signal mtensities We prefer 96-well plate-type thermocycler for large-number of PCR. As the fingerprmtmg pattern IS affected by the make of thermal cycler, one must not compare results obtamed by drfferent themal cyclers.

Fluorescent Dlfferentlal Display

43

4 In addition to arbitrarily chosen lo-mers, we are using novel IO-mers to mmimaze the transcripts that escape the analysis. Based on the observation that more than 90% of mammalian mRNAs seem to have at least one Mb01 (GATC) site in their 3’-end porttons (8), we prepared 32 primers that should cover all the possible Mb01 sttes as follows GGNXYZGATC N = mixture of A, C, G, and T X = R (purme) or Y (pyrimidme) Y = oneofA,C,G,orT Z = one of A, C, G, or T The positions 1 and 2 are fixed to G’s, since our data on mismatches between the arbitrary primes and correspondmg sequences on the transcripts showed less importance of these sites m annealing and priming We found that the primers with degenerate bases m posittons 3 and 4 can detect more bands than those carrying single nucleotrdes at these posmons Since we observed G-T mismatches m more 3’-end positions at the annealmg, we assume that these primers would cover not only all the possible Mb01 sites but other sequences, thereby further reducing the fraction escaping the survey. 5 The two-step scannmg procedure described allowed us to scan wider size range than conventional DD using autoradiographtc detection. Note that this 1s another advantage of FDD, mmimizmg the possibihty of mrssmg interesting bands An example of effective two-step scanning procedure is shown m Fig 1, where the band indicated by the arrow-head poorly resolved m the first scan was clearly separated m the second scanning Although we are currently using Vistra FluorImager SI, we confirmed that the other commercially available imager FMBIO-100 (Takara, Kyoto, Japan) gave comparable results when using with Texas-red or rhodamme X-labeled primers. Also, the detail for scannmg by DNA sequencer was described elsewhere (.5,7). 6. Molecular clonmg of the bands of interest is the other most crucial step for successful DD experiments Although the excised bands are occasionally pure enough to be directly used m cycle sequencing analysis, they are often contaminated with neighbormg bands and sometimes contain cDNAs that comigrate precisely with but distinct from the target species Therefore, one should pay every attention to confirm that the band of your interest, but not the comigrating ones, is really cloned. Described m Sectton 3 5 is our routine procedure to discrimmate true clones from false ones, but various other alternatives to the same purpose would be plausible, mcludmg differential screening of the clones with DD products or probing the Southern blot of the fingerprmts with candidate clones

Acknowledgments We thank A. Shlbata for providing data shown m the Fig. 1, This work was partly supported from grants from Ministry of Education, Science, Sports and Culture of Japan and those from Science and Technology

Agency of Japan.

/to and Sakaki

44

References 1 Ltang, P and Pardee, A. B (1992) Differential display of eukaryottc messenger RNA by means of the polymerase chain reactton Sczence 257,967-971 2. Welsh, J., Chada, K., Dalai, S S., Cheng, R., Ralph, D , and McClelland, M (1992) Arbitrarily primed PCR fingerprintmg of RNA Nuclezc Acid Res. 20, 4965-4970 3. McClelland

4

5.

6

7

M , Mathteu-Daude F , and Welsh, J. (1995) RNA fingerprintmg and differential display using arbitrarily primed PCR. Trends Genet 11, 242-246 Trentmann, S. M., van der Knaap, E., and Kende, H., Liang, P., and Pardee, A B. (1995) Alternattve to 35Sas a label for the differential drsplay of eukaryotic messenger RNA. Sczence 267, 1186,1187 Ito, T., Kito, K , Adati, N., Mttsui, Y., Hagtwara, H., and Sakaki, Y. (1994) Fluorescent differential display: arbitrarily primed RT-PCR tingerprmting on an automated DNA sequencer FEBS Lett 351,23 l-236 Ada&N., Ito, T., Koga, C., Kito, K., Sakaki, Y., and Shtokawa, K (1995) Differential display analysts of gene expression in developing embryos of Xenopus laevw. Blochim Blophys Acta 1262,43-5 1. Ito, T and Sakaki, Y (1996) Fluorescent differential display method for htghspeed scanning of tissue- or cell-specific transcripts Methods 4401 Genet 8,

229-245. 8 Ivanova, N B. and Belyavsky, A. V. (1995) Identification of differentially expressed

genes by restriction endonuclease-based gene expresston fingerprinting. Acids Res 23,2954--2958.

Nucleic

5 Cloning Differentially Expressed by Using Differential Display and Subtractive Hybridization

Genes

Jackson S. Wan and Mark G. Erlander 1. Introduction Differential gene expression occurs in all phases of life, includmg development, maintenance, injury, and death of an organism. Being able to identify these genes will help understand not only gene function but also the underlying molecular mechanisms of a particular biological system. Differential display (DD) (I,21 and subtractive hybridization (SH) (review see refs. 3-5) are by far the most common methods currently used by investigators for this purpose. However, do these techniques identify both abundant and rare mRNAs? Previous kinetic studies (6,7) indrcate that approx 98% of all mRNA species within the cell have a prevalence ranging from l/10,000 to l/100,000, and thus are considered rare. Furthermore, the remammg approx 2% of mRNA species (about 200-500) contribute to over 50% of the total mRNA mass (6,7). Therefore, it is relatively easy to identify a few very abundant differentially expressed mRNAs (i.e., using plus/minus screening); but if the quest is to identify the majority of differentially expressed mRNA species, then a method that is not sensitive to mRNA abundance is required. To evaluate how well DD and SH can identify abundant as well as rare differentially expressed mRNAs, we used both of these methods to find differentially expressed mRNAs within HeLa cells in response to interferon-y. Described below is a summary of our results (see ref. 8 for original report), followed by detailed protocols of DD, SH, and reverse Northern. Since the introductions of DD and SH, hundreds of reports have been published that document either successful applications or method improvements. From

Methods in Molecular Bology, Vol 85’ D/fferent/a/ Ed&d by P Llang and A B Pardee Humana

45


46

Wan and Et-lander

However, some users of these technologies have experienced frustration and failure m implementmg these methods. For DD, most of the criticisms are: 1 2 3 4.

A high false posltwe rate Questioned ability of DD to identify both abundant and rare mRNAs Coding regions of mRNAs are usually not cloned Verlticatlon process IS time consummg and usually requires a fair amount of RNA (9,lO)

For SH, the most common complamt is the technical difficulty of the method; not only does one have to complete numerous tasks (2-3 wk of experiments) before one knows whether the subtraction has been successful but also if the subtractlon fails it is sometimes difficult to sort out which step went wrong. To identify differentially expressed genes m HeLa cells in response to IFNy, we surveyed via DD approx 1300 mRNA species with 72 primer sets and via SH sequenced 1000 cDNA clones from a subtracted library containing cDNAs made from IFN-y induced HeLa cells versus untreated Hela cells. In total, 33 and 23 differentially expressed genes were identified by SH and DD respectively; two of these were found by both methods. The low overlap suggest that many more differenttally expressed genes are yet to be found m this expenmental system and also that these two methods may identify different mRNA populations. The median mRNA abundance of the genes identified by both techniques was similar, approx l/20,000 and ranged from l/1000 to l/200,000 Thus, both methods were able to identify abundant and rare mRNAs. The htt rate (true positives) of DD (1 in 2) is slightly better than that of SH (1 m 5.6). However, hit rates of both methods can vary depending on the criteria for selecting clones/bands for evaluation and the biology being studied. For SH, cDNAs found multiple times m the subtracted library have a much higher hit rate than those found only once. Therefore, if numerous abundant differentially expressed mRNAs are present within the biological system under study then the hit rate ~111be high. With DD, we have observed that bands that are intense, sharp, and appear in a all-or-none fashion are more likely to be true positives. Redundancy (i.e., clonmg the same cDNA more than once) was observed m both methods. As indicated above, for SH, redundancy is mRNA abundancedependent. Consistent with this was our observation that for SH, abundant differentially expressed mRNAs were more readily identified than rare differentially expressed mRNAs. For DD, redundancy is primer sequence-dependent and can be observed m at least four ways. First, we as well as others, have shown that the four most 3’ bases of the lo-mer are the most critical m determinmg annealing specificity (8, II) By comparmg the sequence of a IO-mer with its site of annealmg on

Gene Cloning and Differential Display

47

the cDNAs, we have found that in 9 out of 10 cases,the mismatches occurred within the six bases at the 5’ end of the IO-mer and not in the four most 3’ bases. However, IO-mers identical at the last four bases do not necessarily produce similar band patterns (Fig. 1, lanes 2 and 3), indicating that although the four most 3’ bases determme annealing specificity, the 5’ end of the primer probably plays a role in stabilizmg or destabilizing a particular primer-cDNA complex. Second, redundancy can be caused by “weak” IO-mers. We define “weak” primers here as 10-mers that produce low intensity bands or blank lanes when used alone m a DD PCR (no TlzVX primer, i.e., anchor primer, is used m the reaction; see Fig. 1, lanes 6 and 9). With a weak 10-mer, one observes the same pattern of amplified cDNAs whether one uses IO-mer + anchor primer or anchor primer only (see Fig. 1, and compare lanes 1 and 4 with 5). Thus, weak lo-mers do not contribute to the amplltication process. We recommend that each anchor primer be used as a single primer m a DD PCR so that the anchor only pattern can be compared with anchor + IO-mer patterns for the process of eliminating weak primers from the lo-mer repertoire. Third, redundancy is contributed by the 3’ anchor primers as well. We have observed that the four anchor prtmers (Tt2VG, T12VA, T,,VT, TtZVC) do not completely separate the mRNA population mto four subpopulations when these anchor prtmers are used as first-strand synthesis primers for cDNA synthesis. Thus 1sillustrated m Fig. 2 (lanes 5, 10, 15, and 20), which demonstrates that a IO-mer used alone m a DD PCR amplifies stmilar cDNAs/patterns from all four first-strand synthesis preparations. Finally, the molar ratio between the lo-mer and the anchor primer plays a role m redundancy. A high lo-mer/anchor primer ratio will produce patterns simtlar to that of lo-mer alone whereas a low ratio will produce patterns similar to that of using only the anchor primer (Fig. 2). Therefore, to mmimize all types of observed redundancy we recommend the followmg: 1 The last four bases of the lo-mer should be as diverse as possrble

2, The anchorprimers shouldbe usedalone asa control to eliminate weak IO-mers. 3. A different set of IO-mers should be used for each anchor primer. 4. An equal molar ratio of lo-mers and anchor primers should be used

There are advantages and dtsadvantagesto usmg either SH or DD (see Table 1). If one has a two-way comparison in which l-5 ug of poly (A) RNA are available, we recommend doing both DD and SH because most likely different differentially expressed mRNAs will be found by the two methods. In addrtion, although SH up front is more labor intensive, once the subtractive library is made, the screening (see Section 3.3.) is straight forward and false positives are quite minimal. However, we prefer DD overall because of its versatility and sensitivity. Specifically, SH requires l-5 pg of poly(A) selected RNA, whereas DD only needs lO,OOOcDNAs; DD

48

Wan and Erlander Anchor Primer:

T12VC

F g

10 mer only

5

6

-lolanes:

1

2

nnnn-nnnn

3

4

7

6

9

Fig. 1. Differential display redundancy may be caused by “weak” 1O-mers. The reactions were pedormed in duplicates and loaded adjacent to each other. TlzVC was used during cDNA synthesis. For the PCR, 1.25 pA4TlzVC and of each IO-mers [ 15 (TACGCAGTAC), 143 (TAACGCGTAC), 271 (AGTAGGGTAC), 399 (GATCGTGTAC)] wasused in lanes l-4, respectively; 1.25 pM T,*VC alone was used in lane 5; 1.25 pit4of each IO-mer alone was used in lanes 6-9.

Gene Cloning and Differential

WV Tl,VA

Anchor:'20125051,25 lamer: lanes:

051252.0

1 2 m-.-.-r--

3

Display

49

Tlw

Tl,VG

0'2.01.25051.25

0

1.25

4

5

051252.0

6

7

0

6

0'

'201.2505125

1.25

051252.0

9 --.-~-----10 11

12

T12W 012012505125

0

13

14

0'4

1250512520

15

16

17

16

0

1253

19 r

20:

Fig. 2. Using the same IO-mer in combination with the four anchor primer may cause redundancy in differential display. The reactions were performed in duplicates and loaded adjacent to each other. Tt2VA, Tt2VC, TIZVG, and Tt2VT was used for cDNA synthesis for lanes 1-5, 6-10, 11-15, 16-20, respectively. IO-mer #7 (CATTCAGCAC) was used in all reactions except for lanes 4, 9, 14, and 19. The primers and concentrations used for the PCR are marked above each lane. TlzVA (lane 4) and T,,VT (lane 19) alone consistently produce patterns with low intensity.

50

Wan and Erlander

Table 1 Comparison

of Differential

Display

and Subtractive

Hybridization

Differential display RNA required Prevalence of mRNA surveyed Type of differences Type of redundancy found Optimal application

Consideratrons

Subtractive hybrrdtzatron

5 pg total RNA Abundant and rare

l-5 pg poly (A) RNA Abundant and rare

2 2-fold Primer sequence dependent

All or none mRNA abundance dependent Targeting mRNAs with large (> 1OX) changes in concentration

Seeking rapid output and/or simultaneous evaluation of several experimental condrtrons Usually clone short 3’ untranslated region

Only one-way comparisons, technically dtfficult

offers the advantage of identifying both increases and decreases (22-fold) of a partrcular mRNA vs SH; and the fraction of differentially expressed mRNAs m aparticular experiment can be quickly assessedwith a few gels using DD. This is a valuable diagnostic tool for the investigator because both too many changes (>5% of cDNAs) and the failure to identify any differentially expressed mRNAs may indicate a need to redesign the current comparisons Because cDNAs obtained from DD are usually short and at the noncodmg region (3’ untranslated region), we found that the Merck EST database can be helpful

for identification

of human mRNAs

because each cDNA m the Merck

EST database is sequenced from both 5’ and 3’ orientation. Thus, a match at the 3’ end can often lead to coding information at the 5’ end. (Thus problem should eventually go away when full length cDNA contig databasesare available when all human EST sequencing have been completed.) Finally, to alleviate the problem with a limiting RNA source for verification, we have found that amplified RNA (aRNA) (1Z,12) is representative of the orrgmal mRNA population and thus can be used as template for screening putatives from sources where mRNA is limiting 2. Materials

2.1. Total RNA Purification

from Tissues or Cells

1. RNeasy Total RNA kit (Qtagen, Chatsworth, CA; cat. no. 74104). RNeasy Spin Columns 1 5-mL collection tubes.


2 3 4. 5

51

2-mL collectlon tubes Lysls buffer RLT (P-mercaptoethanol [P-ME] at 10 pL/mL RLT must be added before use) Wash buffer RW 1 Wash buffer RPE (supphed as a concentrate, 4 vol of ethanol [96-lOO%] must be added before use to obtain a working solution) DEPC-H,O P-mercaptoethanol (P-ME) (Sigma, St. LOUIS, MO, cat no M3 148). Ethanol (100%) and ethanol (70% in water). DNase I (amphficatlon grade) (Gibco BRL, Gaithersburg, MD, cat no 18068-015). QIAshredder (Qiagen, Chatsworth, CA, cat no 79653)

2.2. Differential Display 2.2.1. cDNA Synthesis 1 Superscript Preampllficatlon System (Gibco BRL; cat no 18089-011) 10X PCR buffer 200 mM Tris-HCI, pH 8.4, 500 mA4 KC1 25 mk’ MgCl, 10 mM dNTP mix 10 n&f each dATP, dGTP, dCTP, dTTP 0 1M DTT. SuperScrIpt II RT (200 U/$) RNaseH (2 U/pL) H,O (DEPC treated) 2. Degenerate anchor primers (1 pg/&) T,,VG, T,,VA, T,,VT, and T,,VC where V represents an equal molar mixture of G, A, and C 3 RNA template* mRNA (100 ng/pL) or total RNA (1 pg/pL)

2.2.2. Differential Display PCR 1 Degenerate anchor primers (25 @4) (see cDNA synthesis) 2. 5’ primer (lo-mer) (25 CrM) (see Note 1). 3. 10X PCR buffer 100 mMTris-HCl, pH 8 3,500 mMKC1, 15 mMMgCl,, 0.01% w/v autoclaved gelatin (Perkm Elmer, Foster City, CA, cat no N808-0006) 4. 20 pA4 dNTP mix 20 pA4 each of dATP, dGTP, dCTP, dTTP 5 200 @4 dNTP mix 200 cuz/ieach of dATP, dGTP, dCTP, dTTP 6 [cz-~~S] dATP (1000 Cl/mmol) (Amersham, Arlington Heights, IL). 7 Ampli Taq 5 U/pL (Perkm Elmer; cat no N801-0060) 8. Sequencing loading dye: 95% Formamide, 20 mM EDTA, 0 05% bromophenol blue, 0 05% xylene cyan01 FF 9. X-ray film: Kodak XAR 5. 10 TA cloning kit (Invitrogen, Carlsbad, CA, cat. no K2000-0 1): 10X ligation buffer. 60 mMTris-HCl, pH 7 5,60 mMMgC12, 50 mMNaC1, 1 mg/mL bovine serum albumin, 70 mA4 P-mercaptoethanol, 1 mM ATP, 20 mM dlthiothreltol, 10 mM spermldine. pCRTM2.1 vector (25 ng/pL). Sterile water. T4 DNA llgase (4 0 Weiss U/pL)

52

Wan and Et-lander

SOC medium 2% tryptone, 0 5% yeast extract, 10 mMNaCI2.5 mA4 KCl, 10 & MgCI,, 10 mM MgS04, 20 mM glucose (dextrose) 0 5M b-mercaptoethanol INVaF’ cells 11 LB plates containing 50 pg/mL ampictllm and spread wtth 40 pL of 40 mg/mL X-Gal

2.3. Reverse Northern 2.3 1, cDNA Probe Synthesis 1 SuperScrIpt Preampltficatton System (Gtbco BRL; cat no 18089-011) Ok0 dT,,-1, (0 5 i-&cLL) 10 X PCR buffer 200 mM Trts-HCl, pH 8 4,500 mM KC1 25 mA4 MgCl, 0 1MDTT Superscript II RT (200 U/pL) H,O (DEPC treated) 2 Random hexamers 7 4 mg/mL (Pharmacta, Alameda, CA). 3. RNA template 0 4 pg/pL mRNA or 0 4 pg/$ amplified RNA (aRNA) (13). 4 d(GAT)TP mix. 20 mA4 each of dGTP, dATP, dTTP 5 dCTP (120 l&!) 6 [u~~P] dCTP (3000 Ct/mmol) (Amersham) 7 3NNaOH (make fresh) 8 1M Trts-HCl, pH 7.4 9. 2NHCl. 10 G-50 sephadex column (Boehremger Mannhelm, Indtanapolts, IN)

2.3.2. Dot Blot 1 Nylon membranes Maximum strength nytran, 0 45 mm (Schletcher & Schuell, Keene, NH) 2. Whatman chromatography paper 3 Denaturing solutton 1 5MNaC1, 0 5MNaOH 4 Neutralizing solution. 1 5MNaC1, lMTru+HCl, pH 7 4 5 2xssc. 6 UV-Stratalmker 2400 (Stratagene, La Jolla, CA) 7 Bellco rotating hybrtdizatton oven 8 2X Southern prehybrtdizatton buffer (5 prtme- 3 pnme Inc cat no. 5302~820800)* 10X SSC, pH 7 0, 10X Denhardt’s solutton, 0 1M sodium phosphate, pH 6 8, 0 2% SDS, and 0 OlM EDTA. 9. 2X Southern hybridization buffer (5 prime- 3 prime Inc cat no. 5302-500197) 10X SSC, pH 7 0, 2X Denhardt’s solutton, 0 04M sodium phosphate, pH 6 8, 0 4% SDS, and 0.0 IM EDTA 10 Sheared salmon sperm DNA (10 mg/mL) 11. Formamide 12 Prehybrtdization solution 1 part 2X Southern prehybridtzatton buffer, 1 part formamtde, and 0 1 mg/mL salmon sperm DNA boiled for 5 mm


53

13. Hybridization solution: 1 part 2X Southern hybridization buffer, formamide, and 0.1 mg/mL salmon sperm DNA boiled for 5 mm. 14. Wash solution I* 2X SSC, 0.1% SDS. 15. Wash solution II* 0.2X SSC, 0 1% SDS. 16. Wash solution III: 0.1X SSC, 0 1% SDS. 17 PhosphorImager 445SI (Molecular Dynamics)

1 part

2.4. Subtractive Hybridization 2.4.1. Construction of Unidirectional cDNA Libraries 1 cDNA synthesis usmg Timesaver kit (Pharmacia; cat. no. 27-9262-01). 2. Directional cloning kit (Pharmacla; cat. no. 27-9274-01). 3. pT7T3D vector NotIIEcoRI digested and BAP treated (Pharmacia; cat no 274987-01). 4. pGEMl1Zf vector (Promega, Madison, WI; cat. no. P2421). 5. Calf-intestinal alkalme phosphatase or CIAP (Boehringer Mannheim, Indianapolis, IN, cat. no. 713023) 6 Electrocompetent cells (Stratagene; SURE cells, cat no 200227)

2.4.2. cRNA Synthesis 1. 2. 3. 4. 5. 6. 7. 8.

MEGAscript T7 (Ambron, Austm, TX; cat. no. 1334). CsCl purified unidirectional cDNA libraries. Not1

Protemase K (20 clg/pL). Phenol/chloroform (neutral pH-buffered). DNase I (RNase free). 7.5M Ammoma acetate 10X digest buffer: 100 mM NaCl, 50 mA4 Tris-HCl, dithiothreitol (DTT), pH 7.9, at 25°C 9. 100% ETOH and 80% ETOH. 10. RNase-free HsO.

10 mM MgC&,

1 mM

2.4.3. First Strand Synthesis 1. RNase-free H,O. 2. MMLV-reverse transcriptase (Superscript II (RNaseH negative); BRL, cat. no. 18064-o 14) plus 5X buffer. 3. Not1 dT primer (Pharmacia; cat. no. 27-9274-01). 4 0 1MDTT 5 10 mM dNTP mix: 10 mM each of dGTP, dATP, dTTP, dCTP. 6. [a32P] dCTP (3000 Cl/mmole; 10 mCi/mL). 7. 0.5M EDTA, pH 8.0. 8. 10% SDS. 9. 3NNaOH. 10. lMTris-HCl, pH 7.4. 11 2NHCl

Wan and Erlander

54 12 G-50 spun column. 13 5X buffer: 250 mMTrts-HCl,

pH 8.3, 375 mMKC1,

15 mMMgC&

2.4.4. Target and Driver Annealing 1 2 3 4 5 6 7 8 9.

Phenol/chloroform (neutral pH buffered) RNase-free Hz0 7 5M Ammoma acetate. Annealing buffer: 125 mM HEPES, pH 7.6,5 mMEDTA, 5M NaCl. Mineral 011. 100% ETOH and 80% ETOH. Chloroform TE/NaCl. TE, pH 7.5, with 0 5MNaCl.

2.4.5. Separation with Hydroxyapatlte I. 2 3 4. 5 6. 7 8. 9 10 11, 12 13. 14 15 16 17 18 19. 20. 2 1.

pH 8.0, 0 5% SDS

(HAP)

Chloroform TEMaCl. TE, pH 7 5, with 0 5MNaCl. 1Mstock solutions of monophosphate (138 g monobasrc NaH,PO, H,O/L H20) 1M stock solution of dlbaslc phosphate (268 g Na2HP04 * 7H20/L HZO) 1M phosphate buffer (mix equal volumes of 1M stock solutrons of mono and dibasrc phosphate; should be approxrmately pH 6.5) 50 mMphosphate buffer contams 0 2% SDS. 150 mMphosphate buffer contams 0 2% SDS. 400 &phosphate buffer contains 0.2% SDS. Hydroxyapatrte (Bra-Rad, Hercules, CA, DNA Grade Blo-Gel HTP Gel cat no. 130-0520). Jacketed column (Bra-Rad Econo Column, cat. no 737-6116). Centrrcon 30s (Amicon, Beverly, MA; cat. no 4220) Fractron collector Peristaltic pump. Phenol/chloroform (neutral pH-buffered). RNase-free H20. 0 5MEDTA, pH 8.0 10% SDS 3N NaOH. lMTns-HCl, pH 7 5. 2NHCl. G-50 spun column

2.4 6. Constructron of Subtracted Library 1 10X PCR buffer 0 167M (NH&S04, 0 67M Trrs-HCl, pH 8 8, 0.1 % gelatin, 0.067M MgCl*. 2. 25 mM dNTP mix: 25 mM each of dGTP, dATP, dTTP, dCTP. 3 T7T3D 5’ primer (5’-CGAGGCCGAATTCGGCACGA-3’)


55

4. Not3P 3’ primer (5’-AACTGGAAGAATTCGCGG-3’). 5 Size Sep 400 column (Sepharose CL-4B, Pharmacia; cat. no 27-5105-01). 6. 1X lrgase buffer, 66 mA4Tris-HCl, pH 7.6,O.l mM sperrmdme, 6 6 mM MgCl, 10 mMDTT, 150 mMNaC1 7. Not1 (10 U/pL) and EcoRI (10 U/pL) 8. Amp11 Taq 5 U/pL (Perkin Elmer; cat no N801-0060) 9 Phenol/chloroform (neutral pH-buffered).

3. Methods 3.1. Total RNA Isolation 3.1.1. Total RNA Purification from Tissues The followmg

IS a modtfication

of the Qiagen RNeasy Total RNA kit

1. Use one RNeasy spur column per 350 pL RLT + /3ME lysis buffer per 30 mg tissue. Prepare up to 100 pg of total RNA per RNeasy spm column 2 Homogenize tissue with a Power Gen 35 Homogenizer or a baked siliconized glass tissue grinder m 350 p.L RLT + PME lysis buffer If lysate IS too thick, more lysis buffer may be added and still be able to load onto 1 RNeasy spm column m step no 5. 3. To avoid cloggmg the spm column, brtefly spm the lysate at 15,000g and discard the pellet. 4. Transfer the homogenized lysate to a new tube, add 1 volume of 70% ethanol and mix by pipetmg 5 Load onto one RNeasy spm column (about 700 pL). Microfuge at SOOOgfor 15 s. Discard flow-through. If lysate does not pass through, the speed may be increased to 15,000g. If the volume of the total lysate/ethanol exceeds 700 $, load the mixture successively onto the RNeasy spm column and microfuge as above 6 Wash with 700 pL wash buffer RWl, microfuge, and dtscard flow-through as above 7. Place spin column m a new 2-mL collection tube Wash with 500 pL wash buffer RPE Microfuge and discard flow through as above 8 Wash with 500 pL wash buffer RPE and microfuge spur column at 15,OOOg for 2 mm to dry membrane. 9. Transfer spin column to a new 1.5~mL collectton tube and elute the total RNA with 30 pL of DEPC-treated water. Ptpet the water directly onto the spm column membrane Mtcrofuge SOOOgfor 1 min and then 15,OOOgfor 15 s. If the expected yteld is above 30 ug, a second elution with 30 pL of DEPC-treated water will be needed. 10 To remove the DNA from the eluate, add 1 U DNase I/2 ,ug total RNA m a final reaction volume of 100 pL. Digest at room temperature for 15 min 11. To purify RNA after DNase digestton, add 350 pL RLT + /3ME lys~s buffer to the sample, and mix by pipeting. 12. Add 250 pL 100% ethanol to the sample and mix by ptpetmg. 13 Load sample onto one RNeasy spin column. Mtcrofuge 8000g for 15 s Discard flow-through.

Wan and &lander

56

14 Follow steps 7-9 15. Check concentration by a 260/280 OD reading on a spectrophotometer 16. Run 1 pg total RNA sample on a 1 5% TBE agarose gel to check for quality of RNA.

3.1.2. Total RNA Purification from Cells The following

is a modification

of the Qiagen RNeasy Total RNA kit.

1. Use one RNeasy spm column per 350 pL RLT + PME lysls buffer per 5 x lo6 cells. Prepare up to 100 ccg of total RNA per RNeasy spin column. 2. Homogenize cells usmg Qlagen QIAshredder. Add 350 pL, RLT + PME lysls buffer to 5 x lo6 cells, pipet cell lysate directly mto one QIAshredder column, and microfuge at 15,OOOgfor 2 mm. If lysate is too thick, more lysls buffer may be added and put through the QIAshredder 3 Follow procedure from tissue purification steps 4-16

3.2. Differential Display 3.2.1. cDNA Synthesis For each RNA condition to be compared, the followmg reaction m duplicates is performed for each of the four degenerate anchor primers. In addition, a control reaction (without Superscript II RT) is performed. Thus there 1s a total of 12 reactions per RNA condition. If more than a few bands are seen on the differential display gel from the reaction without Superscript II RT, then your RNA may be contaminated by chromosomal DNA. You may want to treat the samples with DNase I before cDNA synthesis. 1 Mix 1p.L of RNA template with 1 $ of an anchor primer and 10 & HZ0 in a 1.5-n& tube, Incubate at 70°C for 10 min, then place on ice for 1 mm. 2. Add 2 pL 10X PCR buffer, 2 pL MgCl,, 1 pL dNTP mix, and 2 pL DTT, incubate at 42’C for 5 mm to anneal, add 1 pL SuperScnpt II RT; incubate at 42°C for 50 min for cDNA synthesis, terminate the reaction at 70°C for 15 min, then place on ice 3. Add 1 & RNaseH and incubate at 37’C for 20 min. 4. The cDNAs can be store at -20°C until use. 5. Before use, dilute the cDNA to 1:20 with HZ0 (You may need to dilute more if you see a smear on the sequencing gel after differential display PCR).

3.2.2. Differential Display PCR To minimize the redundancy associated with differential display, each of the 4 degenerate anchor primers T 12VG, T12VA, T 12VT, and T 12VC is paired with only l/4 of the 256 possible lo-mers (see Note 1). 1 Mix 1 clr, cDNA template (from cDNA synthesis), 1 @ degenerate anchor primer (use the corresponding primer that was used during cDNA synthesis), 1 pL 5’ primer, 2 pL 10X PCR buffer, 2 pL 20 @4dNTP mix, 1 & [u-~~S] dATP, 0.5 p.L

Gene Cloning and Differential Display Anchor: 10 l?ler:

57

T,,‘JA OP.DDRT3

OP.DDRTI

OP-DDRTS OP-DDRTG OP-DDRT7 OP-DDRT8

OP.DDATS

hours: lanes:

t7lObp

+-630bp

+491bp

Fig. 3. An example of a differential display gel. HeLa cell was untreated (0 h) or treated with IFN-y for 48 h prior to mRNA isolation and PCR as described in the text. The reactions were performed in duplicates and loaded adjacent to each other. T,,VA was used during cDNA synthesis and in all DD PCRs. The lo-mers used are marked above the lanes. The asterisks mark some of the bands representing mRNAs with putative increase or decrease in expression after treatment. Ampli Taq, and 11.5 pL HZ0 in a 0.5-mL thin wall PCR tube. Perform PCR using the following program: (94°C for 20 s, 42°C for 20 s, 72°C for 30 s) for 40 cycles at 72°C for 10 min (we use the GeneAmp PCR system 9600 machine from Perkin Elmer). 2. Mix 1 pL PCR product with 1 p.L sequencing loading dye, heat to 80°C for Smin and load on a 6% acrylamidelXA4 urea sequencing gel (load the duplicate reactions for each RNA condition in adjacent lanes) (see Note 2). Run at 55 W for about 3.5 h. Stop the gel after the top dye (Xylene Cyanol) runs off the bottom. To determine the size of the bands, also load a molecular size marker (such as the M 13 DNA sequencing reaction control from the Sequenase DNA sequencing kit [USB; cat. no. US707701) in a separate lane (see Fig. 3 for example of DD gel).

Wan and Blander

58

paper, dry wlthout fixmg in a 3 Transfer gel onto Whatman Chromatography vacuum gel dryer for 2 h, expose to X-ray film from overmght to two mghts (be sure to use some kmd of fluorescent sticker on the gel so that you can align It to the film) 4 Align the film to the gel, use tape to secure the position, excise out the dlfferentlally expressed bands by cutting through the film, select only bands that are present m both lanes derived from the duplicate cDNA reactions (you may want to expose agam after this step to see If the correct bands were excised) 5 Add 100 $ HZ0 to the excised band m a 1 5-mL tube, sit at room temperature for 10 mm, boll for 15 mm, spin down, and transfer the liquid to a new tube, use directly as template for PCR amplification below 6. MIX 2 & template (from previous step) (don’t use more than 3 &, the urea may inhibit the reaction), 2 $ degenerate anchor primer, 2 & 5’ primer, 4 pL 10X PCR buffer, 25 6 JJL H,O, 4 & 200 WdNTP, and 0 4 & Amp11 Taq in a 0 5-mL thm wall PCR tube Perform PCR using the following program (20 s at 94’C, 20 s at 43”C, and 30 s at 72°C 7 Check the size of the PCR product by checkmg 5 pL of the reaction on a 1.5% agarose gel, then use 2 pL. of the reaction directly for cloning via the TA clonmg kit (Invitrogen; cat. no K2000-01) 8 MIX 2 4 of PCR product, 1 & 1OX llgatlon buffer, 2 PL pCRTM2.1 vector, 4 & sterile water, and 1 pL T4 DNA ligase m a 1.5-mL tube, incubate at 14°C overnight 9 Transform into INVaF’ cell followmg the Invltrogen protocol or mto equivalent cell, spread onto LB plate, incubate at 37°C overnight IO Pick and sequence 4 to 6 white colonies for each band (you need to sequence more than one because the band may be composed of multiple DNA species)

3.3. Reverse Northern Reverse Northern IS a quick and easy way to verify induced mRNAs ldentlfied by differential display or subtractive hybrldlzatlon However, the major difficulty with this technique IS the detection of low abundant mRNAs We have shown that by usmg PCR amplified fragments of cloned cDNAs instead of mmiprep DNA, mRNAs with abundance level as low as 1 m 40,000 can be detected Furthermore, amplified RNAs (aRNA) (13) can be used for probe synthesis, thus reducmg the amount of RNAs required for the verification procedure. 3.3.1. cDNA Probe Synthesis Repeat the followmg for each RNA condition to be compared: 1 Mix 5 pL mRNA and 2 $ ohgo dT,,-,s (or 5 pL aRNA and 740 ng of random hexamers in a total of 7 pL H,O) m a 1 5-mL tube, incubate at 70°C for IO mm, then place on ice for at least 1 min


59

2. Add 2.5 & 10X PCR buffer, 2 5 $ MgCl,, 2 5 @. DTT, 1 pL d(GAT)TP mix, 1 & dCTP (cold), and 7 5 r.lr, [a32P] dCTP (the [a32P] dCTP 1s reduced from 50 pL [500 pCt] to 7.5 pL by drymg m a speedvac Do not let tt dry completely because it will stick to the plastic side) Incubate at 42°C for 5 mm 3. Add 1 pL Superscript II RT, Incubate at 42°C for 50 mm 4. Add 3 p.L NaOH, incubate at 68°C for 30 min to remove the RNA 5. Add a mixture of 10 p.L Trts-HCl, 3 pL HCl, and 9 p.L H,O to neutralize 6 Take 1 pL out to measure counts per minute 7 Spm through a G-50 sephadex column to remove the unmcorporated nucleotrdes 8 Take l-)..tL from the flow-through to measure count per minute and compare to that from step 6 (you should get 2&O% mcorporation, on average, 10s cpm are obtained from 2 pg of mRNA)

3.3.2. Dot Blot 1 Isolate mmiprep DNA from a 1-mL overmght culture for each clone to be tested (alternatively, PCR amplify the insert from each of the clones) 2 Spot 0.5 pL DNA (l/20 of mmiprep approx 1 pg) from each clone (or 1 pg of each PCR product) in a grid-like pattern onto duplicate nylon membranes (one membrane for each RNA condition) (see Fig. 4) 3 Air dry for 2630 mm, denature for 5 min at room temperature by placing membrane (DNA side up) on Whatman paper presoaked with denaturmg solution, wash twice in neutrabzmg solution for 5 mm at room temperature, rinse in 2X SSC. 4. UV crosslmk once (Stratalmker auto crosslmk setting) 5 Place the dupltcate membranes each m a separate hybrtdtzatton tubes, add 10 mL prehybndization solutton (for a 15 x 15 cm membrane) and incubate with rotation m a Bellco rotating hybridization oven at 42°C overnight 6. For each RNA condition, replace the prehybrtdizatton solutton with a mixture of 10 mL hybridization solution and equal amount (cpm) of boiled (5 mm) probe (1 x 10’ to 2 x 10’ cpm per mL of hybridization solutton), incubate with rotation at 42°C overnight 7 Wash membrane twtce m wash solutton I for 5 mm at room temperature, wash twice m wash solution II for 15 mm at 65°C expose membrane on X-ray film (keep membrane moist) 8. You may want to wash the membrane again m wash solution III for 15-30 mm at 65’C to reduce background, then expose again

3.4. Subtractive Hybridization 3.4. I. Construction of Unidirectional cDNA Libraries A detailed protocol for library construction is not given below because many kits are available to the investigator for the lsolatlon of poly (A) RNA as well as for the construction of umdlrectlonal cDNA hbraries. However, to successfully use this subtractive protocol the mvestlgator must construct the unidrrecttonal

libraries

using the vectors described

below or make

the necessary

60

Wan and Erlander A

Fig. 4. Reverse Northern showing cDNAs induced by IFN-y treatment. 7 1 cDNA clones were spotted onto duplicate nylon membranes (A) and (B) at the same relative positions. The membranes were then probed with labeled cDNA from untreated HeLa cells (A) and from HeLa cells treated with IF%y for 48 h (B). Clones in boxes were confirmed by Northems to be IFN-y induced.

changes in the different PCR primers required to amplify the subtracted library (see Fig. 5 for flowchart of subtractive hybridization protocol). 1. Extract at least 1 pg poly(A) selected RNA from the two sources of interest (treated = TARGET and untreated or control = DRIVER). 2. Unidirectional libraries are to be constructed by using Directional Cloning kit that contains the NotIloligo dT primer, with the Timesaver cDNA synthesis. Ligate TARGET cDNA intoNotIIEcoRI/BAR pT7T3D vector and ligate DRIVER into NotI/ EcoRXLAp treated vector pGEMl1Zf and electroporate into electrocompetent cells. Both libraries need to be 21 x 1O6recombinants (if unequal, better to have DRIVER recombinant number >TARGET recombinant number to avoid false positives). Rut-itication of cDNA libraries by CsCI gradient centrifugation is recommended.


E(cL-iiJ’

67

E(TzJ

I

Linearize

both templates Noti dtgestlon

wtth I

Complete cRNA synthesis on both with V RNA polymerase

I Complete 2 - 1st strand syntheses with 2pg cRNA with 32p-adCTP I

t Remove GINA from cDNA with NaOH

250 fold molar excess

( -25bg

cRNA)

/I Save for HAP calibration T/ Hybndize “antlsense” target cDNA with 25pg (sense) cRNA Isolate smgle +stranded CDNA with HAP column

4

Repept

Once

+ Construct subtractive cDNA library + Screen

clones

Fig. 5. Flowchart of subtractive hybridization

protocol

3.4.2. cRNA Synthesis 1. Linearize both Target and Driver cDNA libraries by incubating 30 H of each with 100 U ofNot& 30 & 10X digest buffer m a total volume of 300 $ at 37OC for 2 h. 2. Run sample (10 pL) of digest on 0.8% agarose gel with molecular weight markers to make sure all vector-cDNA is linearized. 3. Meanwhile to digest add 0.75 pL of Proteinase K and continue to incubate at 37°C for 30 min. 4. Extract twice wtth phenol/chloroform and precipitate linearized DNA with 0.5 vol 7 SMammonia acetate and 2.5 vol of 100% ethanol. Wash pellet with 80% ethanol.

62

Wan and Erlander

5. Dry DNAs m speedvac and resuspend each DNA pellet m 20 $L of RNase-free H,O Determme OD,,, for transcription reactions 6 Set up cRNA transcrtption reactions (we use Ambton MEGAscrtpt) as described below 7. For Drover cDNA do a 3X reaction (60 pL vol) and for Target cDNA do a 1X reactron (20 pL vol) 8 For a 1X reachon add the followmg at room temperature 2 pL. 1OX transcription buffer for T7 RNA polymerase, 8 $ of 75 mMrNTP mix (ATP, CTP, GTP, and UTP), 1 c(g of lmearize cDNA, 2 & T7 RNA polymerase mix, and RNase-free H,O to 20 $I-,. Incubate at 37°C for 4 h 9 To each transcription reaction add an equal vol of 1X transcrtptton buffer (we find tt helps to dilute cRNA prior to DNase I addition for complete digestton) 10 Remove a 5-pL aliquot from each transcription for gel analysis (save and run with samples obtamed after DNase I digest) 11 Add 4 U DNase I/20 $ reaction and incubate at 37’C for 15 mm After digest remove 5 & from each reaction for gel analysis 12 Separate the cRNA reactions on a 0 8% agarose gel (before and after DNase I treatment) to confirm entire removal of cDNA template 13 Extract cRNAs once with phenol/chloroform followed by two subsequent prectpttations with 0.5 vol 7 5M ammonia acetate and 2.5 vol 100% ethanol to remove free nucleotides Wash pellet with 80% ethanol Resuspend m RNase-free H,O 14 Determine OD,,, of Target and Driver cRNAs You need at least 4 pg of target cRNA and 50 pg of driver cRNA for the followmg procedures Store cRNAs at -20°C

3.4.3. First Strand Synthesis 1 Complete two first strand syntheses of Target cRNA by domg the following m duplrcate add 2 pg of Target cRNA, 0 5 pg Not1 dT prtmer to RNase-free H,O for a total volume of 14.80 &, heat for 10 minutes at 65°C and quench on Ice. Then add 6.6 p.L of 5X buffer, 3 3 pL DTT, 1.65 p.L dNTP mtx, 5 pL [cz~~P] dCTP MIX up and down with pipet, then add 1.65 pL MMLV-reverse transcriptase Incubate at 37°C for 60 mm 2 Save one of the reactions for the calibration of the HAP column (see Sectton 3 4 5.) To the other add 1 65 & EDTA and 1.65 pL SDS Mix and then add 5 p.L NaOH and Incubate at 65’C for 60 mm to remove cRNA, then let It cool to room temperature. 3 Add 16 5 & Tris-HCl followed by 5 pL HCI. Mix up and down 4. Remove unmcorporated [a32P] dCTP from both first strand reactions with G-50 spun column. Check with Geiger counter to make sure eluent has at least 1O-fold over background m cpm Store both first strand cDNAs at -20°C

3.4.4. Target and Driver Annealmg 1. Ptpet about half of the NaOH treated cDNA into a screw-cap Eppendorf tube. Extract once with phenoVchloroform and take aqueous phase to a fresh screw-cap tube


63

2. Add the equivalent of 25 pg of cRNA drover Measure total vol, add 0 5 vol of 7.5 ammonia acetate and then 2.5 vol 100% ETOH Precipitate, wash (80% ethanol), and allow to dry at room temperature 3 Resuspend cDNA/cRNA pellet m 5 pL RNase-free H20, 4 p.L annealing buffer, 1 pL 5M NaCI (RNase-free) Prpet up and down to make sure cDNA/cRNA 1s totally resuspended 4. Overlay cDNA/cRNA with 50 pL of mineral 011and boil mixture for 90 s 5 Incubate at 65-68’C for 20-24 h. 6 Add 100 pL chloroform and 90 pL TE/NaCl to the mineral orl/cDNA/cRNA and vortex Save aqueous phase and add to the organic phase an addmonal 100 pL of TE/NaCl. Repeat extraction. 7. Extract the combined aqueous ahquots once with phenol/chloroform. Add 2 5 vol 100% ethanol (do not add ammonia acetate), precipitate at -7O’C for 10 min, and wash subsequent pellet with 80% ethanol 8. You can store pellets overnight at-20°C ifneeded before proceeding with separation

3.4 5. Separation on HAP Column 3.4 5.1

PREPARATION OF THE HAP COLUMN

1. Weigh out about 1 g of hydroxyapatrte (HAP) mto a 50-mL conical tube. This should be about 7 5 mL of dry HAP Add 50 mM phosphate buffer for a total volume of 45 mL and swirl gently Let HAP settle for 10 mm, then pipet off the cloudy upper layer containing the “fines,” and repeat this procedure until the buffer above the settled HAP does not appear as cloudy 2 After asprratmg off the buffer for the last time, then add an equal volume of 50 mM phosphate buffer to the HAP slurry. 3 Set up Jacketed column, peristaltic pump and fraction collector as shown on Fig 6 Circulate the Jacket with 60°C Hz0 and eqmhbrate all phosphate buffers at 60°C 4 With the set-up complete, add 2-3 mL of 50 mMphosphate buffer to an empty column and then add HAP slurry and let rt settle by gravity Ideal HAP height is about 1 5 cm above the frit after packing and this is achieved by trial and error (use long Pasteur prpet to remove HAP). It is very important to gently resuspend the HAP before removing or adding HAP so that the matrix settles evenly. Ideal flow rate 1s 0.5 ml/mm, collectmg fractions of 0.5-mL/tube 5 Once the column is packed, wash with 5 mL of 60°C 50 mA4phosphate buffer by gently adding buffer down the mstde side of the column and then turnmg on the pump Allow the memscus of the buffer to dram untilJust (approx 0.2 cm) above the HAP and then mm off the pump. Do not allow the HAP to dry’! Now you are ready to calibrate the column 3.452.

CALIBRATION OF HAP COLUMN

1 Unfortunately each batch (lot) of HAP is different and therefore we recommend you spend an afternoon calibrating your partrcular HAP with your Jacketed column for determining the optimal concentration of phosphate buffer for eluting single stranded and double stranded nucleic acid

Wan and Erlander

64 Jacketed HAP Column

k-m Peristaltic

Pump

-wEI

Fig. 6. Diagram of experimental set-up for HAP column separation. 2. Take an aliquot of the first-strand synthesis of the DRIVER that was NOT hydrolyzed with NaOH as your double stranded nucleic acid (RNA-DNA heteroduplex). Use about 2-5 x lo5 cpm of it so you can easily follow it. Followmg the identical directions as outlined below for your actual sample (1.e , add 6 mL of 50 M, then 6 mL 150 mM, then 6 mL 400 mM phosphate buffers) make sure that the majority (>80-90%) of your counts (i e., RNA-DNA) comes off when 400 mM phosphate is added. 3. Next, take an aliquot of your single stranded nucleic acid (use NaOH treated first-strand synthesis) and again add 50 mA4,150 mM, followed by 400 mMphosphate buffer; the majority (>80-90%) should come off at 50 mMand 150 mM. If not, then adjust accordingly (e.g , if 50% is elutmg at 400 mM, then you need to increase the concentration of phosphate buffer used to elute single strand nucleic acid). Our experience is that single stranded elutes at 120-150 mA4 and double stranded always at 400 m&f.

3.453. SEPARATING YOURSAMPLES ON HAP COLUMN 1. To load and run the column, first be sure that all three buffers are equilibrated to 60°C. Resuspend sample pellet m 500 pL of 50 mM phosphate buffer. Let the column buffer dram until the meniscus is just above the HAP and turn off the pump. 2 Very carefully load the sample along the side of the column wtth a Pasteur pipet, being careful not to disturb the HAP, and turn on the pump just long enough to pump as much of the sample into the matrix as possible without letting the HAP column dry out. 3. Turn off the pump, gently load 6 mL of 50 mM phosphate buffer, and turn on both the pump and the fraction collector. When the buffer is just above the HAP,


65

HAP Column Fractionation

r

300000

250000

200000 E 3

150000

100000

50000

0

Fig. 7. Actual example of collected fractions after HAP column separation Here Target = HeLa cells treated for 48 h with IFN-y and Driver = untreated HeLa cells Fractions 1-12 were eluted with 50 &phosphate buffer (contains 0 2% SDS), fractions 13-24 were eluted with 150 mA4 phosphate buffer (contains 0.2% SDS) and fractions 25-40 were eluted with 400 Mphosphate buffer (contams 0 2% SDS). In this experiment, fractions 17,18,19, and 20 (containing single-stranded subtracted cDNA) were pooled and subsequently processed.

4.

5.

6.

7

turn off the pump, load 6 mL of preheated 150 mMphosphate buffer and elute as before. It IS helpful to advance the collecting rack 1-2 spaces between changes of buffer (easier to keep track of tubes later). Follow the 150 mA4phosphate buffer with 6 mL of 400 mM phosphate buffer. Determine which fractions contam eluted nucleic acid, by measuring Cherenkov counts of each fraction. For each separation, select the four fractions with highest counts for each of the three eluttons (i e , 50, 150, and 400 mM) (see Fig. 7 for example). You should see three peak-the second, eluted with 150 mA4 phosphate buffer, contains the single-stranded nucleic acid. Pool the four fractions from the single-stranded peak with the highest cpm and extract twice with phenol/chloroform (this may be done m 14-mL polypropylene tubes at 3000 rpm for 5 min) Apply aqueous (upper) phase to a Centrlcon 30 and centrifuge at 5OOOg for 25 min with a fixed angle rotor. Wash three times with 1 mL of RNase Free HZ0 (same centrifuge conditions) and measure volume--lt should be about 100 +. At this point, the sample should be rehybridlzed with driver. First, do one phenol/chloroform to make sure it IS free of RNase--take supernatant and pre-

Wan and Et-lander

66 cipitate with an additional tion process

25 pg driver and repeat annealing and HAP separa-

3 4.5 4. AFTER SECOND HAP COLUMN 1 Bring up or dry down the washed eiuent (subtracted cDNA) from the Centricon 30 to a total volume of 100 pL. To the 100 pL of target cDNA add 5 pL EDTA, 5 uL SDS, 15.1 pL NaOH. Incubate at 65°C for 30 min Cool to room temperature and add 50 Ilr, Tris-HCl and 15 1 pL HCI 2 Do one phenol/chloroform extraction, bring up to 100 pL with water and apply to a G-50 spun column Store eluent at -20°C

3.4.6. Construction of Subtracted cDNA Library 1 Amplify

2 3

4

5 6 7. 8

9. 10

subtracted cDNA m a total volume of 100 u,L by addmg 10 clr, 1OX PCR buffer, 0 8 pL 25 mM dNTPs, 200 ng T7T3D5 primer, 200 ng Not3P primer and 1 pL of HAP eluent (subtracted cDNA) Do a HOT START using 0 5 p.L of Taq polymerase and complete 30 cycles (Perkm-Elmer model 9600 conditions are 94°C for 15 s, 60°C for 15 s, and 72’C for 1 mm) Run 10 p.L of the PCR products on an agarose gel Expected result is a smear from 300 to 1 kb Equilibrate three times m a Stze Sep 400 column wtth 1X hgase buffer (This IS done by drammg column but not dry, then add 2 mL of 1X hgase buffer, cap column and mvert/mix sepharose with 1X hgase buffer, dram and then repeat process two more times ) Do one phenol/chloroform extraction with amplified subtracted cDNA, then bring up to 100 p.L with 1X ligase buffer, and apply to the spun column (to prepare spun column, take 1X ligase buffer-equilibrated column and centrifuge for 2 min in swinging bucket rotor at 4OOg, then very carefully apply sample to middle of dried sepharose, then spm again at 400g for 2 min to collect eluent) Add 1 5 pL of Not1 (10 II/@,) to eluent and incubate at 37°C for 1 h Heat mactivate at 65°C for 10 min and cool to room temperature Add 1.5 pL of EcoRI and incubate 37°C for 1 h Heat macttvate as above, extract once with phenol/chloroform, bring up to 100 pL with 1X hgase buffer rf needed, and apply to a pre-equilibrated (1 X ligase buffer) Size Sep 400 spun column (as described m step 3). Ligate mto pT7T3D vector, electroporate/transform mto SURE cells, and plate 1, 0 1, and 0.01 pL onto 150 mm LB-ampicillm agar plates Should have at least 1O6recombmants Do maxiprep (of lo6 recombmants) with the remannng broth culture and store cDNA library at -20°C

3.4.7. Verification of Putative Positives Randomly pick 96 clones from your subtracted cDNA library and grow overnight followed by mml-preps to isolate DNA. Spot (by hand or with 96-well


67

dot blot manifold) approx 1 pg of DNA on duphcate filters and follow the reverse Northern protocol as described in Section 3.3. 4. Notes 1 These are arbitrary lo-mers with 50% GC content. Since the site of annealmg is highly depended on the 3’ end of the IO-mer, the last four bases of the oligo should be as different as possible The IO-mer performs best when the most 3’ base 1s a G or C, however, to increase the diversity, we have also used A or T with the caveat that the band intensity decreases slightly. Therefore, a total of 256 different lo-mers can be used 2. Dtfferentlal display IS a technically demanding method. In order to minimize false positlves, It IS essential that the bands are sharp and well separated on the gel. We found that standard flat well comb produces better bands than sharkstooth comb Also, the DNA should be loaded as close to the bottom of the well as possible; this ~111 produce a sharp flat band and elimmate streaks appearing m the center of the lane.

References 1 Llang, P. and Pardee, A B (1992) Dlfferentlal display of eukaryotlc messenger RNA by means of the polymerase chain reactlon. Science 257, 967-971 2 Welsh, J , Chada, K , Dalal, S S , Cheng, R , Ralph, D , and McClelland, M. (1992) Arbltrarlly primed PCR fingerprinting of RNA Nuclezc Acids Res 20,4965-4970. 3 Travis, G H , Milner, R. J , and Sutchffe, J. G. (1989) Preparation and use of subtractive cDNA hybridlzatlon probes for cDNA cloning, in Neuromethods, vol. 16, Molecular NeurobloEogzcaZ Technzques (Boulton, A A., Baker, G. B., and Campagnom, A. T., eds ) Humana, Chfton, NJ, pp. 49-78. 4 Watson, J B and Margu!ies, J. E. (1993) Differential cDNA screening strategies to Identify novel stage-specific protems m the developmg mammalian brain Developmental

Neuroscl

15,77-86.

5. Sambrook, J , Fntsch, E. F , and Maniatls, T (1989) Molecular Cloning A Laboratory Manual, Cold Spring Harbor Laboratory, Cold Sprmg Harbor, NY 6. Axel, R., Felgelson, P , and Schutz, G. (1976) Analysis of the complexity and diversity of mRNA from chlcken liver and oviduct Cell 7,247-254. 7. Bishop, J. O., Morton, J G , Rosbash, M , and Richardson, M. (1974) Three abundance classes m HeLa cell messenger RNA. Nature 250, 199-204 8. Wan, J. S., Sharp, S J., Polrier, G. M.-C., Wagaman, P C , Chambers, J., Pyati, J., Horn, Y.-L., Galmdo, J E., Huvar, A., Peterson, P. A., Jackson, M. R., and Erlander, M. G. (1996) Cloning differentially expressed mRNAs Nature Bzotechnol 14, 1685-1691. 9 Bertioh, D. J , Schhchter, U. H , Adams, M J , Burrows, P R., Steinblss, H. H., and Antomw, J. F. (1995) An analysis of differential display shows a strong bias towards high copy number mRNAs Nucleic Acids Res 23,4520-4523 10 Debouck, C (1995) Differential display or differential dismay? Cur-r Opinzon Bzotechnol 6, 597-599.

68

Wan and Erlander

11. Bauer, D., Muller, H , Retch, J , Rtedel, H., Ahrenktel, V , Warthoe, P., and Strauss, M. (1993) Identtficatton of dtfferenttally expressed mRNA spectes by an improved display technique (DDRT-PCR) ~Vucletc Aczds Res 21,4272-4280. 12 Van Gelder, R N , von Zastrow, M. E , Yool, A, Dement, W. C , Barchas, J D , and Eberwine, J. H (1990) Amplified RNA synthesized from hmrted quantttres of heterogeneous cDNA Proc Nat1 Acad Scz USA 87, 1663-1667 13. Ponier, G M -C , Pyatt, J., Wan, J S., and Erlander, M G (1997) Screening differentially expressed cDNA clones obtained by differential display usmg amplified RNA Nuclerc Aczds Res 25,9 13,9 14.

Direct Sequencing of Differential Display PCR Products Xinkang Wang and Giora 2. Feuerstein 1. Introduction The polymerase chain reactron (PCR)-based mRNA differential display (I) has become an increasingly popular alternative technique for isolating genes of interest m a variety of m vitro and m VIVOsystems (for review, see ref. 2), as compared to such conventional techruques as differential screening and subtractive hybridization. The method of mRNA drfferenttal display consists of two basic steps: (1) reverse transcrtptron (RT) using a set of 3’-anchored primers, T,,MN (M = G, A or C; N = G, A, T or C); and (2) PCR amplification of cDNA fragments using arbitrary lo-mer primers (upstream) and anchored downstream primers. One critical feature of this technique is to display most of the mRNA population on a sequencing gel after PCR. A similar technique referred to as RNA fingerprinting by arbitrary primed PCR (RAP-PCR) was introduced to identify differentially expressed genes (3). RAP-PCR uses longer primers, e.g., 18-20 mers, for both RT and PCR, that allow the annealing temperature of PCR reaction increase up to 60°C after several mitral cycles of annealing at low temperature (3,4). The use of longer primers in the RAP-PCR method not only increased the reproducibility of band display pattern but also made rt possible for direct sequencing of the PCR products. However, a potential problem of RAP-PCR IS displaying fewer bands than mRNA differential display. Since the initial introductron of mRNA differential display, rt has been extensively refined to reduce the redundancy of anchored primers and decrease the number of RT reactions, to reduce the incidence of false positives, and increase the efficiency of the technique (2). For example, the modtficatrons made based on the combination of RAP-PCR and differential display applied From

Methods In Molecular B/o/ogy, Vol 85 D/fferent/a/ Edlied by P Ltang and A B Pardee Humana

69


Wang and Feuerstein

70

longer primers for differential display (5-7), which not only Increase the reproducibility of dlfferentlal display but also are possible for direct sequencmg after PCR amphficatlon. DNA sequencing analysis 1s one of the critical steps following mRNA dlfferentlal display. In the mltlal protocol (l), short primers (lo-mers arbitrary primers and T,*MN anchored primers) are used for differential display, but they are too short to prime sequencing reactions. Therefore, subclonmg is often required after differential display m order to analyze the sequence of particular genes. This chapter focuses on direct sequencing analysis after differential display using the typical (IO-mer) primers. Slmllarly, the same protocol 1s also feasible for the analysis of differential display PCR products usmg longer primers, as well as the PCR products obtained by other techniques such as RAP-PCR.

2. Materials 2.7. Differential

Display Using Short Primers

1. RNA* LPS-stimulated vs control. Stored at -70°C 2. Pruners: downstream, T&L4 primers (M = G, A, or C); upstream, S-GACCGCTTGT3’. Stored at -20°C 3 RNAmap kit (GenHunter [Nashville, TN]). Stored at -2O’C. 4. 10X PCR buffer: 100 mM Tns-HCl, pH 8.4, 500 mM KCl, 15 mM MgClz and 0 0 I % gelatin Stored at -20°C. 5 [35S]ol-dATP or [33P]a-dATP (1000-3000 Wmmol, Amersham, Arlington Hetghts, IL). 6 AmpllTaq DNA polymerase (5 U/pL, Perkm-Elmer, Norwalk, CT). Stored at -20°C 7. 6% polyacrylamldel8M urea sequencmg gel

2.2. Differential

Display Using Elongated

Primers

1 Upstream primer, S-GCGAATTCCGACCGCTTGT-3’, downstream anchored primer 5’-CGGATCCCTTTTTTTTTTTTMA-3’ These elongated primers are generated by adding extra bases (usually a restriction enzyme site) at the 5’ end Stored at -2O’C. 2 Others as hsted m Section 2 1

2.3. Preparation

of Differential

Display PCR Products

1 Six percent polyacrylamrde gel 2. DNA extraction solution (for acrylamlde gel): 500 mM NH4Ac, 1 mM EDTA. (Stored at room temperature )

2.4. Northern

Analysis

1 RNA samples. 2 GeneScreen Plus membrane (DuPont-New England Nuclear, Boston, MA) 3 Northern hybrldlzation buffer: 5X SSPE (750 mMNaC1,50 mMNaH2P04, pH 7.6, 5 mA4EDTA), 50% deionized formamide (stored at-70°C), 5X Denhardt’s solu-

Direct Sequencing

71

non (50X stock solution stored at 4”C), 2% SDS, 100 pg/ml-. polyA and 200 pg/mL boiled salmon sperm DNA Freshly make the solution prior to use 4 Random-priming DNA labelmg ktt (Boehrmger Mannhelm, Indianapolls, IN) 5 [33P]a-dATP (3000 C~/mmol, Amersham). 6. Washing solution 2X SSPE, 2% SDS. Stored at room temperature

2.5. DNA Sequencing Analysis 1, Double strand-DNA cycle sequencing system (Gibco BRL, Grand Island, NY) Stored at -2O’C. 2. [32P]y-ATP (5000 C~/mmol, Amersham) 3 PNK: T4 polynucleotide kmase at stock concentration of 10 U/pL Stored at -20°C. 4. 10X PNK buffer 700 mA4 Trts-HCl, pH 7 6, 100 mM MgC12, 50 mM dithiothrettol Stored at -2O’C 5. Blo-Spm 6-column (Bto-Rad, Hercules, CA). 6. 6% polyactylamtde/SM urea sequencmg gel

2.6. Computer Database Search Computer network to accessgenomic databases, e.g., GenBank 3. Methods

3. I. Differential Display Using Short Primers 1 Perform reverse transcription reactions in a volume of 20 j,tL, contammg 0 2 pg total cellular RNA (e g , isolated from LPS-stimulated or untreated aorta), 2 pL of 2 pA4Ti2MA primers, 4 pL of 5X RT buffer (125 flTris-HCl, pH 8 3, 188 mM KCl, 7.5 mM MgC12 and 25 mA4 DTT), 1 6 p.L of 250 @4 dNTP, 0.5 ).tL RNase inhibitor, and 1 pL MMLV reverse transcriptase Heat RNA at 65°C for 5 mm and place on me prtor to the addition of enzyme and RNase mhtbttor Incubate the RT reaction mixture at 37°C for 60 mm, and then heat at 95°C for 5 mm and place on me for PCR, or store at -20°C for later use 2. Prepare PCR reaction mixture containing 2 pL of 2 w 5’ decamer arbitrary primer (5’-GACCGCTTGT-3’) and 3’ T12MA primer, 1 pL [3sS]a-dATP (or 0.25 pL [33P]a-dATP), 2 pL 10X PCR buffer, 1 6 pL dNTP (25 @4), 2 pL RT products, and 0 2 pL Ampliraq DNA polymerase (5 U/pL). Perform PCR for 40 cycles as follows 94°C for 30 s (for denaturing), 40°C for 2 mm (for annealing), 72’C for 30 s (for extension), followed by 1 cycle for extension at 70°C for 10 min. 3. Resolve the PCR products through an 8M urea, 6% polyacrylamtde DNA sequencing gel and analyze the gene expresston by autoradiography (see Fig. 1A)

3.2. Differential Display Using Elongated Primers Following the same protocol described in Section 3.1.) but with the following exceptlons (see Note 1): 1. RT reaction is carried out with either the anchored primers of TlzMA extended prtmers, e.g , S-CGGATCCCTTTTTTTTTTTTMA-3’

or the

Wang and Feuerstein

72

A

12

3 4

4

LPS-7

B 12

34

LPS7

Fig. 1. Differential display analysis of upregulated gene expression in cultured rat aorta stimulated with LPS. (A) The differential display was carried out using RNA samples isolated from cultured rat aorta in the presence or absence of LPS. PCR products were resolved in an 8M urea, 6% polyacrylamide sequencing gel: lane 1, unstimulated; lane 2, stimulated aorta from spontaneously hypertensive rats; lane 3, unstimulated; and lane 4, stimulated aorta from Wistar-Kyoto rats. The band(s) indicated with an arrow (designated as LPS-7) shows a marked induction in response to LPS stimulation, which was generated by an upstream primer (S-GACCGCTTGT-3’) and a downstream primer (T&4). (B) Northern analysis to confirm the LPS-7 mRNA expression in cultured aorta in response to LPS stimulation. Total RNA (10 pg/lane) was loaded in the same order as in A, and resolved by electrophoresis, transferred to a nylon membrane, and hybridized to LPS-7 and ribosomal protein L32 (rpL32; a loading control) cDNA probes sequentially. The mRNA size was indicated on the right. 2. PCR is carried out using an upstream primer, 5’-GCGAATTCCGACCGCTTGT3’; and downstream primers, 5’-CGGATCCC~TTTTTTTTMA-3’. The cycle conditions are as follows: the first l-4 cycles at 94°C for 30 s, 40°C for 2 min, and 72°C for 1 min; and the following 35 cycles at 94°C for 30 s, 60°C for 2 min, and 72’C for 1 min; followed by 72°C for 7 min.

3.3. Preparation of Differential Display PCR Products 1. Localize the band of interest, cut out the band with a razor, release the DNA by boiling in 100 pL TE ( 10 nut4 Tris, pH 8.0,l mMEDTA) for 10 min, and precipi-

Direct Sequencing

73

A

Fig. 2, Comparison of the reamplitied LPS-7 band using short differential display primers with the elongated primers. (A) Schematic graph showing the reamplification of the differential display bands with initial differential display primers (primer A and B) or with the elongated primers (primer A’ and B’). (B) Electrophoresis of the reamplified DNA fragments using the initial differential display primers and the elongated primers as illustrated in A. PCR samples were resolved in a 6% polyacrylamide gel and stained with ethidium bromide. Lane 1, the amplification of LPS-7 with primers A and B; lane 2, the amplification with the modified primers A’ and B’; lane 3, DNA markers.

2.

3.

4. 5.

6.

tate with 80% ethanol plus 5 pL glycogen (10 mg/mL) for 30 min (to overnight) at -70°C. Microcentrifuge for 10 min at 4’C to pellet DNA, wash the pellet with 80% icecold ethanol by microcentrifuge at 4OC for l-2 min. Dry the pellet with speedvacuum and dissolve in 10 pL water. Reamplify the DNA using 4 pL resuspended sample, 4 pL of each elongated primers, and 3.2 pL dNTP (250 CLM)in a final volume of 40 pL. All other conditions are essentially the same as for the initial amplification. Resolve DNA bands on a 6% polyacrylamide gel (see Note 2), stain the gel with ethidium bromide (see Fig. 2). Cut out the DNA band under long UV wavelength, mince the gel slice into small pieces, and extract the DNA in 3-vol of 500 mA4N&Ac, 1 mMEDTA overnight at 37°C. Microcentrifuge the samples for 3 min at room temperature, collect the supernatant (repeat several times to remove the gel debris if necessary), extract the

Wang and Feuerstein

74

supernatant once with phenol/chloroform, and preclpltate with ethanol Resuspend m 10-20 pL TE and store at -20°C The DNA IS now ready for either direct sequencmg or probe-labeling

3.4. Northern

Analysis

Apply standard protocols

for Northern

analysis (see Note 3).

1 Resolve RNA samples (10 E/lane) m a formaldehyde-agarose slab gel and transfer to a GeneScreen Plus membrane 2. Prehybrldlze the membrane for 4 h, and add lo6 cpm/mL probe for overnight hybridization at 42°C 3. Wash the membrane m 2X SSPE at room temperature for 15 mm, followed by 2X SSPE and 2% SDS at 65OC for 30 min with 2-3 times dependmg on slgnal intensities. 4 Expose the membrane on X-ray film at -7O’C with a Cronex Lightning-Plus intensifying screen (see Fig. 1B)

3.5. DNA Sequencing

Analysis

Cycle DNA sequencing method is recommended for this analysis since it only requires a small amount of template (see Note 4). 1 Label the elongated primers (either the upstream primer or the downstream primer; see Fig 3) by mcubatlon of 100 ng ohgonucleotlde, 3 $ [32P]y-ATP, 2 $, 1OX kmase buffer, and 1 pL PNK m a final volume of 20 pL at 37°C for 60 mm Apply Blo-Spin 6 column to remove free [32P]ATP 2. Prepare a prereaction mixture containing 0.1 ~18DNA, lo-40 ng of [32P]-end labeled primer, 4 5 pL 10X Tag sequencmg buffer, and 0 5 & Taq DNA polymerase (2 5 U) in a final volume of 36 & MIX well and add 8 PL prereaction mixture to each tube contammg 2 pL of termination mix G, A, T, and C, respectively 3. Heat thermal cycler to 94’C Place reaction tubes in wells and start the follows cycle reactions. initial cycle, 94°C for 4 min, 42°C for 30 s, and 70°C for 1 mm; followed by 29 cycles of 94°C for 30 s, 42°C for 30 seconds, and 70°C for 1 min. 4. Terminate the reactlon by adding 5 pL of stop solution to each tube, and resolve the samples (2 &/well) in an 8M urea, 8% polyacrylamlde sequencing gel Dry the gel and subject to autoradlograph (see Fig. 3).

3.6. Computer

Database Search

Apply Genetxs Computer display gene products.

Group (GCG) program to analyze the dlfferentlal

1 Use SeqEd to edit the sequence (enter at least 50 nucleotldes of unambiguous sequence) 2. Use FastA for the search of nucleotlde sequence identity (see Note 5) against computer databases, e g., GenBank

Direct Sequencing

75 A’ primer /,,TC\

B’primer

* ,

Fig. 3. Direct sequencing of LPS-7 DNA amplified with the elongated primers. LPS-7 DNA was reamplified with the elongated primers (A’ and B’) and isolated from 6% polyacrylamide gel. Sequencing reaction was carried out using a cycle sequencing method in the presence of 32P-labeled primer A’ or B’. The samples were resolved by an 8M urea, 8% polyacrylamide gel, dried, and subjected to autoradiograph.

4. Notes 1. The-elongated primers are generated from the initial differential display primers (lo-mers) by adding extra-nucleotides at the S-end of the primers. These extended bases usually include a restriction enzyme site to facilitate subcloning if needed. The elongated primers can be directly used for differential display ($6) or used during the reamplification following original differential display method (8). In addition, the elongated primers are able to directly prime the sequencing reaction without subcloning of the differential display PCR products. 2. If DNA fragments are smaller than 500 bases, acrylamide gel electrophoresis is recommended to resolve and prepare the reamplified products for its advantage to resolve small nucleotides. However, if longer fragments are generated, agarose gel (152%) electrophoresis is recommended and corresponding methods for DNA isolation should be applied.

Wang and Feuerstein

76

3, Northern analysts IS a critical step that not only confirms the differential

gene expression (see Fig 1B), but may also mdrcate the purity of DNA since heterogeneous hybrldrzatlon patterns usually mdrcate the presence of multiple genes Similarly, other methods such as RNase protection assay, nuclear run-on analySISmay also be used. The posstbthty of dtrect DNA sequencing of dtfferential display products depends on the purrty of the Isolated DNA bands. The better resolutton of the sequencing gel will no doubt generate the purer DNA products If heterogeneous DNA species exist m some particular cases, these bands have to be subcloned prior to sequencing analysts The presence of a restriction enzyme site m the elongated primers will no doubt facihtate thts subclonmg process Because the anchored PCR primers of differential display are located next to the poly(A) tall, the majority of differential display products correspond to 3’ untranslated region (UTR) Therefore, a computer database search on the nucleotlde level IS recommended

References 1 Liang, P. and Pardee, A B (1992) Dtfferential display of eukaryottc messenger RNA by means of the polymerase chain reaction. Sczence 257,967-971 2. Llang, P and Pardee, A B (1995) Recent advances in differential display Curr Oplnlon Immunoi 7,274-280. 3 Welsh, J , Chada, K., Daldal, S. S , Cheng, R, Ralph, D , and McClelland, M (1992) Arbitrarily primed PCR fingerprintmg of RNA Nuclezc Acids Res 20, 4965-4970. 4. McClelland, M., Mathteu-Daude, F., and Welsh, J. (1995) RNA fingerprmtmg and drfferentral drsplay usmg arbitrarily primed PCR Trends Genet 11,242-246 5. Zhao, S., 001, S L., and Pardee, A B (1995) New primer strategy improves precision of differential display. BzoTechnzques 18, 842-850. 6 Drachenko, L B , Ledesma, J., Chenchlk, A A, and Srebert, P D (1996) Combming the techmque of RNA fingerprintmg and differentral display to obtam dlfferentrally expressed mRNA Blochem Bzophys Res Comm 219,824-828. 7. Yoshlkawa, T., Xmg, G. Q., and Detera-Wadlergh, S. D. (1995) Detection, stmultaneous display and dn-ect sequencing of multiple nuclear hormone receptor genes using btlaterally targeted RNA fingerprmtmg. Bzochzm Bzophys Acta 1264,63-7 1 8 Wang, X. K and Feuerstem, G. Z (1995) Direct sequencing of DNA isolated from mRNA differential dtsplay. BzoTechnzques 18,448-452

A Direct-Sequencing-Based Strategy for Identifying and Cloning cDNAs from Differential Display Gels Katherine

J. Martin, Chi-Pong

Kwan, and Ruth Sager

1. Introduction Thus report describes an approach to identifying and clomng messages from differentral display (DD) gels that has markedly reduced the occurrence of false positives as well as the time required for the process. Most importantly, the use of Northern blots with potentially contaminated probes is avoided and a streamlined direct sequencing protocol is used as an imtral step. The individual methods presented here are not new, but have been put together from a variety of sources in a revised order. Methods for the DD reverse transcrrptron-polymerase chain reactions (RT-PCR) and electrophoresrs conditions, for which we have used published methods (2) and the GenomyxLR DNA sequencer (see Note 1 and ref. 2), are not covered here. Conventronal strategies of identifymg and clonmg mRNAs from DD gels involve first verifying differential expression by Northern blot analysis using a probe made by PCR-amplifying cDNA cut out from the DD gel. During the PCR process, a significant amount of contaminating DNAs of the same size as the band of interest may be produced and/or amplified that cannot be removed by gel purificatron. Even what may appear to be low levels of such contaminants can cause problems for Northern blots if they hybridize efficiently to highly represented RNAs. As a result, signals from differentially expressed mRNAs may be obscured and the DD band scored as a false posmve. Even if such a mixture is not abandoned at this point, it can be time-consuming and labor-intensive to locate the differentially expressed message using conventional strategies. From

Methods m Molecular Brology, Vol 85 Dffferentral Edlted by P Llang and A B Pardee Humana

77


Martin, Kwan, and Sager

78

The strategy described here, which has resulted m a low rate of false posltives (see Note 2), avolds Northern blotting prior to the availability of a pure probe. The method 1sbased on the imtlal direct sequencing of DD bands. Unlike Northern blots, DNA sequencmg is not sensitive to hybridization efficiency or cellular message representation level. In most cases,direct sequencing rapidly provides positive identification of a DD band. Followmg database verification, sequence information can be used to design ohgonucleotides with which to generate probes for Northern analysis using either of two methods: (1) small (approx 20 bp) oligonucleotide pairs allow the generation of long (>150 bp) probes by RT-PCR from total cellular RNA; and (2) a smgle oligonucleotlde 40 bp in length can be end-labeled and used directly as a probe (3) For situations where faint or crowded DD bands do not produce a legible DNA sequence or where the direct sequence cannot be verified by a database match (see Notes 3 and 4), a strategy for screening subclones 1sdescribed that employs single-base sequencing to ldenttfy unique subclones (4) followed by reverse Northern slot-blottmg (5) to Identify subclones that are differentially expressed. The overall strategy/Is shown (Fig. 1). This strategy can rapidly generate a large number of clones or tags for differentially expressed messages m a partlcular system of mvestlgation. Such mformatlon 1spotentially invaluable for the production of arrays that can be blotted by different sources of cellular RNA, yielding information on patterns of gene expression and regulation. 2. Materials 2.1. DNA Elution from DD Gel, PCR Amplification and DNA Sequencing 1 TE. 10 mM TrwCl, pH 7.4, 1 n-&? EDTA. 2 Glycogen. 10 mg/mL stock solution m dHzO. 3. DD primers: three 21 bp anchor primers LH-T,

,N, twelve 18 bp LH-AP andten

20-2 1 bp extendedOPA arbitrary primers (I). 4. Thermal cycle PCR machme

5. AmpliTaq@DNA polymerase(Perkm Elmer, Branchburg, NJ). 6 DeoxyNTPs:

stock solutions of 2 5 mM and 250 @I m dH,O

7 Taq PCR buffer 1OX stock solution* 100 m&I Tns-Cl, pH 8 4,500 mM KCl, 15 n&f MgCl*, and 0 01% gelatin 8 Mineral oil. 9. 4MNH,OAc. 10 Isopropanol

11 70% ethanol. 12 VentR@(exe-) DNA polymerase(New England BioLabs, Beverly, MA) 13. a33P-dATP,2000 Cilmrnol (DuPont NEN, Boston,MA).

A Direct-Sequencing-Based

Strategy for DD Gels Oifferentiai

display

Cut out differentially PCR-amplify

79

gel

efipressed

bands

(28 cycles)

Sequence

Unreadable

Readable

I

Known genes

EST match

No match

1 1 Use sequence to make probe for Northern or slot-blot for reuerse Northern I I

+

t Northern

~

Clone Screen subclones by reuerse Northern slot-blotting

or reuerse

Northern

Obtain full length clones for selected new genes

Fig. 1. Flowchart showing overall strategy based on mitral direct DNA sequencing for identifying and clonmg cDNAs from DD gels. 14. 15 16. 17

CircumVentTM Thermal Cycle DNA Sequencing Kit (New England BroLabs) GenomyxLR DNA sequencer (Genomyx, Foster City, CA) Access to DNA databases. Synthetic 40 bp ohgonucleotrdes complementary to appropriate region of genes of interest, which can be end-labeled and used to probe Northern blots (3) or synthetic 20-25 bp ohgonucleotrde pairs with matching melting temperatures complementary to appropriate region of genes of interest to PCR-amplify Northern probes from total cellular RNA.

80


2.2. Cloning, Single-Track Sequencing and Reverse Northern Slot-Blotting 1 Method of cloning PCR fragments, e g., Origmal TA cloning Kit (InVitrogen, San Diego, CA) 2 Mmi plasmid preparation method, e.g , QIAprep Spin Plasmtd Mimprep Kit (Qtagen, Chatsworth, CA) 3 Filter unit, 0.45 pm (Ultrafree-MC, Mtlhpore, Bedford, MA). 4. Nylon membrane, e g , ZetaProbe Blotting Membrane (Bto-Rad, Hercules, CA). 5 Slot-blot manifold (Schlelcher & Schuell, Keene, NH) 6. Method to synthesize 32P-labeled cDNA from total cellular RNA, e.g., SuperScrtpt Preamphfication System (Gibco BRL. Gatthersbury, MD). 7 a32P-dCTP, 3000 Ci/rnmol (NEN, Boston, MA) 8 150 x 15-mm tissue culture dishes 9 X-ray film of high sensitivny, e g., Kodak XAR 10 20X SSC: 3MNaC1,O 3MNa3citrate 2H20, pH 7.0

3. Methods 3.7. DNA Elution from DD Gel, PCR Amplification and DNA Sequencing 1, Elute DNA from cut-out DD band by boiling gel slice 10-20 mm in 100 pL TE Centrifuge briefly, transfer supernatant to a clean tube and precipitate DNA with ethanol using 3 pL glycogen as a carrier. Redissolve DNA m 20 pL ddHzO (see Note 5) 2. PCR-amplify using as few cycles as posstble (see Fig. 2). Use the same two primers as those used m the DD procedure to produce the band being sequenced (see Note 6), each at a final concentration of 400 ti. Mix primers with l-3 pL redtssolved DNA template, 0.5 U Taq polymerase and 20 @4 each dNTP m standard Tuq PCR buffer, final volume 20 pL Cover with mineral 011and perform approx 20 PCR cycles: 94°C for 45 s, 60°C for 2 min, and 72’C for 1 mm Elongate an additional 5 min at 72’C. 3 Remove primers and dNTPs by selective alcohol precipltatton (7). All reagents and incubations for this step should be at room temperature. Transfer 20 pL of the completed PCR reaction to a clean tube and add 20 pL 4MNH40Ac and 40 pL isopropanol Incubate 10 mm, then mtcrocentrifuge at 14K rpm for 10 min Remove the supernatant with a drawn-out glass ptpet. The pellet will generally not be visible. Add 100 pL 70% EtOH to the pellet and centrifuge briefly Remove supernatant with the drawn-out pipet, dry the precipitate 10-15 min m a SpeedVac Concentrator (Savant) and redissolve in 10 pL TE (see Note 8). 4. Sequence DNA using anchor or arbitrary DD primer (or extended primer; see Note 7 and 8), which should be at least approx 18 bp in length. Cycle sequence with Vent (exe-) DNA polymerase, a33P-dATP and 2 pL redissolved DNA for each band being analyzed. Perform annealing step at the same temperature as that used for DD An example of the results is shown (see Notes 1 and 3; Fig 2).


Strategy for DD Gels

Cycles of PCR:

15

25

Neu,

C/EBP 6

al ACH 48

15

25

81

49

15

25

E 48

8

Fig. 2. Effect on DNA sequencing tracks of increasing the number of PCR cycles. Examples, with identities indicated at top, were generated from DD bands that were seen when RT-PCR reactions were performed using RNA from the nontransformed breast cell line 76N, but not with RNA from the breast tumor cell line MDA-MB-435S (see Note 1). al ACH, a-l antichymotrypsin; C/EBP 6, CAAT-box enhancer binding protein. 5. Read DNA sequence and compare to sequences available in GenBank, EMBL and any additional DNA databases available database (e.g., BLAST network service at NCBI) (see Note 10). Also compare with sequences entered in EST database.


82

6. Followmg database verification of sequence (see Note 4), 20-25 bp primers with matching melting temperatures can be destgned and synthesized to generate by PCR a probe (approx 100 to 500 bp) for standard Northern analysis or reverse Northern slot-blotting (see Section 3 2 , step 5) Alternatively, a smgle synthetic ohgonucleotide of approx 40 bp can be made, end-labeled with kmase and used to probe a Northern blot (3)

3.2. Cloning, Single-Track Sequencing and Reverse Northern Slot-Blotting This sectron describes methods used to clone DD band cDNAs and select subclones representing differentially expressed messages.We have applied this strategy to DD bands that (1) drd not yield good quality direct sequences or (2) dtd not match any current database entrres (includmg ESTs). This has comprised about 30% of the DD bands that we Isolated (see Notes 3 and 4). 1 PCR-amplify 1 pL precipitated DNA from step Section 3.1 , step 1 using DD primers and PCR conditions described (l), but reduce the number of PCR cycles to 25 2 Clone PCR products by standard methods. Select about four colonies for analysis m cases where the direct sequence 1s readable or eight colonies where the sequence appears highly contaminated or no sequencing signal is detected. Isolate plasmids using mini-preparation methods 3 Perform single-track (G, A, T, or C) sequencing reactions of the selected plasmids using standard sequencing methods and compare the patterns generated (see Note 1). Restriction analysis is reported to be a useful alternative to smgletracking (4) To avoid sequence patterns that differ depending on the orientation of the insert, use a primer that anneals to the cloned DNA or a umdnectional clonmg strategy. 4 PCR-amplify inserts from the selected subclones and extract from 1% agarose gels using a spin column 5 Perform reverse Northern slot-blotting Prepare membrane for slot-blottmg according to standard methods (8) Apply 5 pL denatured PCR product onto each of two nylon membranes using a slot-blot manifold Include controls known to be differentially and similarly expressed Synthesize 32P-labeled cDNA probes from 10 pg total RNA used for DD by standard methods with 50 $1 a3*P-dCTP Prehybridize and hybridize membranes using standard Northern blotting conditions m two 150 x 15 mm tissue culture dishes sealed with parafilm. Hybridize overnight at 42°C then wash twice at moderate strmgency (2X SSC with 0.1% SDS at 50°C 15 min each). Expose membranes to film for about 5 d (see Notes 11 and 12; Fig 3)

4. Notes 1 The extended format of the programmable

GenomyxLR

sequencer is useful for

band separationand gives reproducible results both for the DD gels (2) and the single- and four-track sequencing reactions

A Direct-Sequencing-Based 76N

83

Strategy for DO Gels

MDA-MB-435 laspin :omplement

component

C3

iST A Votectin

K059)

la lb - PAIIC

!a !b !C !d - Laminin

p3

!e

Fig. 3. Reverse Northern slot-blotting used to screen subclones for differential expression. DNA fragments corresponding to the genes or subclones indicated at the right were slot-blotted in duplicate and probed with mixtures made by reverse transcribing, then 32P-labeling, total cellular RNA from the nontransformed breast cell line 76N or the breast tumor cell line MDA-MB-435s (see Note 6). Four genes known to be down-regulated in breast tumor cells lines are shown at top as controls: maspin, complement component C3, GST a, and protectin (CD59). Examples of subclones obtained from two differentially expressed DD bands are shown at bottom. Subclones la, lb, and lc were obtained from a single DD band identified by direct sequencing as plasminogen activator inhibitor 2 (PAI-2); subclone 1b was identified as differentially expressed and sequencing verified its identification as PAI-2. Similarly, subclones 2a-2e were identified as laminin 83 and the differentially expressed subclone 2d confirmed the identification. (Subclones 2a and 2c are also laminin l33; la, lc, 2b, and 2e were not examined.) 2. We have confirmed differential expression by Northern blotting or reverse Northern slot-blotting for all DD bands we have tested to date (manuscript in preparation). 3. Using the methods described, we have found that 89 (87%) of 102 differentially expressed DD bands yielded good quality DNA sequence information (i.e., 80-99% of bases could be read) allowing database searches for matching sequences. Bands that did not produce good sequencing tracks tended to be faint on the DD gel. 4. In our experience, 80% (44/55) of sequences could be verified by matching to an entry in the GenBank, EMBL, and EST DNA databases.

84

Mart/n, Kwan, and Sager

5 Though the PCR reaction of Section 3 1 , step 2 can be performed with the actual

6

7 8.

9.

10

I1 12

gel band m the tube, rather than botlmg the gel band and precipitating the eluted DNA as in Section 3 1 , step 1, these procedures provide one with a reusable DNA solutton and a higher yield of PCR product per cut-out band RNAs from two cultured breast cell lines were used, mcludmg the normal cell lme 76N, whtch was isolated from a reduction mammoplasty spectmen (9), and the transformed cell lme MDA-MB-435S, originally isolated from a metastattc ductal adenocarcmoma of the breast and obtained from the American Type Culture Collection (Rockvtlle, MD) If short primers are used for DD, extended primer sets may be used at Section 3 1 , step 2 to produce PCR products that can be directly sequenced (6) If some 011 was transferred from PCR reaction of Section 3 1 , step 2, into the prectpttatton tube It ~111 remam m the tube during drying, making the completed sample appear wet Thts ~111 not be detrimental to subsequent steps of the sequencing procedure We routinely use one of three 2 1-bp DD anchor prrmers, LH-T, ,N (4), as a primer for cycle sequencing reactions Ninety-four DD bands from which we obtained readable sequences (dnectly or from differenttally expressed subclones), represented 55 dtfferent genes AdJacent bands wtth tdenttcal expression patterns on the DD gels usually gave the same DNA sequence, apparently resulting from strands of DNA that separated on the denaturing DD gels A small number of genes have also been found many (up to 20) times with the same arbttrary prtmer 32P-labeled cDNA probes can be saved and reused to blot addmonal filters. After hybndrzmg the membrane to identify the differentially expressed subclones, both filters can be strrpped and reblotted in parallel wtth a 32P-labeled cDNA probe mtxture made from a smgle cell line, e g., the MDA-MB-435s tumor cell Ime, to verify equal loadmg of subclone DNAs onto the filters

Acknowledgments We thank S. Douglas and A Amsowlcz for helpful comments on the manuscript, and M. Erlander for assistance with the reverse Northern protocol.

References 1. Zhao, S , Ooi, S L , and Pardee, A B (1995) New primer strategy Improves the precision of differential dtsplay BloTechnrques 18, 842-850. 2. Averboukh, L., Douglas, S A, Zhao, S., Lowe, K., Maher, J., and Pardee, A B (1996) Better gel resolution and longer cDNAs increase the precision of differential display. BzoTechnzques 20,9 18-92 1 3 Lmskens, M H K , Feng, J., Andrews, W. H , Enlow, B. E., Saatt, S M., Tonkm, L. A., Funk, W D., and Vtlleponteau, B. (1995) Cataloging altered gene expression m young and senescent cells using enhanced differential display. Nuclezc Acids Res 23,3244-325

1.


Strategy for DD Gels

85

4. Zhao, S . 001 S L , Yang, F -C., and Pardee, A. B (1996) Three methods for identtficatton of the positive cloned cDNA fragments m differential display. BzoTechnzques 20,400-402.

5 Mou, L., Miller, H , Wang, E., and Chalifour, J. (1994) Improvements to the differential display method for gene analysis. Blochem. Bzophys Res. Commun. 199, 564-569. 6 Wang, X. and Feuerstem, G. Z. (1995) Direct sequencing of DNA isolated from mRNA differential display BloTechnzques l&448-452. 7. Brow, M. A. D. (1990) Sequencing with Taq DNA polymerase, m PCR Protocols. A Guide to Methods and Applicatzons (Inms, M A., Gelfand, D H., Sninsky, J. J , and White, T. J., eds.), Academtc, San Diego, CA, pp. 189-196. 8. Brown, T (1995) Dot and slot blotting of DNA, m Current Protocols zn Molecular Biology (Ausubel, F M , Brent, R., Kmgston, R. E., Moore, D D., Seidman, J. G., Smith, J. A., and Struhl, K., eds.), Greene Publishing Associates, and John Wiley and Sons, pp. 2.9.15-2.9.20. 9 Band, V and Sager, R (1989) Distinctive tracts of normal and tumor-derived human mammary epithelial cells expressed in a medium that supports long-term growth of both cell types Proc Nat1 Acad Set USA 738, 103-123

Differential Screening of Differential Display cDNA Products by Reverse Northern Hong Zhang, Rong Zhang, and Peng Liang 1. Introduction Differential Display (DD) is a widely used methodology for cloning differentially expressed genes. Since its inception in 1992 (I), much effort has been made to improve and optimize the technique (2,3). One major bottle-neck remains, though, in the screening step for the verification of the cDNA fragments tentatively identified by differential display. In the standard protocol, Northern blot verification was recommended, with the reamplified cDNA as a probe (4). An alternative approach is to clone the reamplified cDNA and then use it as a probe for Northern analysis. Both approaches have their drawbacks. It is known that the reamplified cDNA probes from differential display sometimes contain a mixture rather than a single cDNA species, due to the contamination of comigrating cDNA fragments. As a result, direct use of the reamplified cDNA as probes might detect multiple messenger RNA species, making the identification of the truly differentially expressed gene difficult. On the other hand, if the cDNA fragment is cloned before being used as a probe for Northern blot, the choice of which clone to use becomes a matter of luck, depending on the extent of cDNA contamination. Compounding the problem is the laborious screening by Northern blot for each and every cDNA fragment identified. It is especially difficult under circumstances where the amount of RNA samples is limited. To this end, various approaches have been tried to improve the method. Mou et al. (5) described the slot-blot approach using random labeled composite cDNA probes to screen the reamplified cDNA fragments from differential display. Although increasing the through-put of the screening, this method could characterize many cDNA From:

Methods

in Molecular

Biology,

Edited

by: P. Liang

and A. B. Pardee

Vol. 85: Differential

87

Humana

Display Methods Press

Inc., Totowa,

and Protocols NJ

Zhang, Zhang, and hang

88

fragments as false posrtives If the reamplitied cDNA consrsts of more than one species, especially when the contaminating cDNAs correspond to abundant messages. Here we describe the use of differential screening by Reverse Northern dotblot as well as colony hybridization to screen for cDNA fragments that truly represent differentially expressed mRNAs. The method takes advantage of the high cloning efficiency of the pCR-TRAP cloning system, which is based on posmve-selection for insert. After cloning the reamplified cDNA fragments of interest, cDNAs labeled by reverse transcriptron are used as probes to screen for colonies harboring differentially expressed genes. This method has the followmg advantages: (1) many filters representing different cDNA fragments can be screened at the same time and with the same set of probes; (2) the masking

effect of false positive

clones is resolved; and (3)

once a cDNA clone IS confirmed to be posmve, rt IS ready to be sequenced, checked against the databaseand, if necessary,used as a probe to screen libraries. 2. Materials 2. I. Labeling 1, 2. 3, 4. 5 6.

2.2. Subcloning 1 2. 3 4 5. 6. 7. 8. 9 10. 11. 12.

of cDNA Probes

5X RT buffer: 125 mMTns-Cl, pH 8.3,188 mMKCl,7.5 mMMgClz and 25 mA4DTT dNTP(-C) mix* 500 lJ4 of each stock (without dCTP) mixed m equal volume. Oligo-dT primer: 10 w Tzo primer (GenHunter, Nashville, TN). a[32P]dCTP* 3000 Cl/mm01 (NEN, Boston, MA) MMLV reverse transcriptase, 100 U/pL (GenHunter) Quick Spin Column (Sephadex G-50, Boehringer Mannhelm, Indtanapolis, IN).

and Reverse Northern

Dot-Blot

T4 DNA ligase: 200 U/pL (GenHunter). pCR-TRAP vector. 150 ng/pL (GenHunter). GH competent cells (GenHunter) Tetracycline plates: LB plates with 10 pg/mL tetracycline (stored in dark at 4’C) 10X PCR buffer: 100 mMTris-HCl, pH 8 4,500 mM KCl, 15 rnA4MgC12, and 0.01% gelatin. 250 l&f dNTP (GenHunter). 2 pA4 Rgh and Lgh primers: primers flanking the cloning site of the pCR-TRAP vector (GenHunter). AmpliTaq DNA polymerase: 5 U/clr, (Perkin-Elmer, Foster Ctty, CA). 2N sodmm hydroxyde solution. 3Msodium acetate, pH 5.0. 20X SSC: 3MNaCl and 0.3M sodium citrate, pH 7.2. Prehybndization solutton: 50% formamtde, 0.75MNaCJ50 mMsodmm phosphate, pH 7.4, 5 mM EDTA, 0.1% Ficoll-400, 0 1% BSA, 0.1% polyvmylpyrohdone, and 0 1% SDS Stored at -2O’C. Add freshly boiled salmon sperm DNA to 20 &nL right before use.

Differential

Screening by Reverse Northern

89

2.3. Colony Hybrid&a tion 1. Denaturing solution: 0 5N NaOH/1.5M NaCl 2 Neutralizing solution 0.5M Tns-HCl, pH 7.0, 1 5M NaCl

3. Methods 3.1. Labeling

of cDNA Probes

1. Prepare 32P-labeled cDNA Probe by reverse transcrlptlon m a 50-pL reaction To a mlcrofuge tube, add 10 pL of 5X RT buffer, 5 PL ohgo-dT primer, 8 PL dNTP(-C) mix, 5 p.L a[32P]dCTP, 10 pg RNA, and dH20 to make up to 45 @ 2. Mix well and incubate the sample at 65°C for 5 min Then shift it to 37°C for 10 mm. Add 5 & MMLV reverse transcrlptase and continue to incubate at 37 “C for 1 h 3. Use a Quick Spin column to remove the unmcorperated 32P (as instructed by the manufacturer) Count 1 pL of the probe to check the radloactlvlty, which 1stypltally m the range of lo’-10s total cpm

3.2. Subcloning

and Reverse Northern

Dot-Blot

Perform the differential display using the RNAlmage kit (GenHunter) essentially as mstructed (see also Chapter 1). Retrieve the cDNA fragments of interest (Fig. 1A) from the denaturing polyacrylamlde gel and reamphfy them with the same sets of primers as in the Initial differential display PCR reactions (Fig. 1B). 1. Ligate 5 @. of the reamphficatlon product to 2 pL of the pCR-TRAP vector, with 2 pL 10X ligation buffer, 10 p,L dH20 and 1 pL T4 DNA ligase Carry out the ligation at 16’C overnight. 2. Transform 100 pL GH competent cells with 10 pL of the ligation mix Plate onefifth of the cells onto tetracycline plates After an overnight Incubation at 37”C, 50-100 colonies are typically obtained 3. Randomly pick four colonies from each plate and stick each mto 50 mL lysis buffer. Boll for 5 mm After spm for 5 mm, save the lysates. 4. Perform the colony-PCR m 30 cycles (94°C for 15 s, 52°C 40 s, and 72’C 1 mm) in 40-K reactions, with 4-pL 10X PCR buffer, 4-a Rgh primer, 4 @, Lgh primer, 4 pL of the colony lysates, 19.6 & dH,O, and 0.4 pL Taq polymerase m each tube Followmg a 5-min chain extension at 72”C, check the size of the inserts with 5 pL of the PCR products on a 1 5% agarose gel (Fig. 2A). 5. To prepare for blotting, mrx 30 & of each colony-PCR product with 5 pL of 2N sodium hydroxide Boll the mixture for 5 min to denature the DNA, then neutralize it with 5 pL of 3M sodium acetate, pH 5.0 After bringing the total volume to 105 & with dH20, dot-blot 50 pL of each sample onto duplicate Nylon membranes using the Blo-Dot mlcrofiltration system (see Note 2) UV cross-link the membrane and rinse it m 6X SSC before prehybndization. 6 Shake and incubate the blot m 10 mL of prehybrldization solution at 42’C After 4 or more hours, boll the cDNA probes for 5 mm and add equal counts (1 06-1 0’ cpm) to the respective bottles. Hybridize overnight at 42°C.

Zhang, Zhang, and Liang

90

A Control

Mob-36

Mob-37

Mob-38

Mob-39

B MW

CaWol

Mob-36

Mob-37

Mob-36

Mob-39

Fig. 1. (A) Differential display of the messenger RNA from Rat1 cells (left lanes) and T 10 l-4 cells (right lanes) (see Note 1). cDNA fragments of interest were indicated by arrows. H-T, ,C was used as the anchor pruner in combinations with arbitrary primers H-AP18, H-AP18, H-AP29, H-AP3 1 and H-AP34 (GenHunter) to amplify the negative control, Mob-36, Mob-37, Mob-38 and Mob-39, respectively. (B) 1.5% agarose gel electrophoresis of the PCR products from the retrieved fragments. 7. Wash the blots twice at room temperature, for 15 min each, with 1X SSC and 0.1% SDS, then once at 60°C for 15 min with 0.25X SSC and 0.1% SDS. 8. Expose an X-ray film against the blot overnight at -80 “C with an intensifying screen (Fig. 2B; see Note 3).

3.3. Screening

by Colony Hybridization

(see Note 4)

1. Transfer the tetracycline-resistant colonies (obtained from Section 3.2.) by replica-plating onto duplicate nitrocellulose membrane filters laid on top of the tetracycline plates. 2. After an overnight incubation, float each filter in denaturing solution for 2 min to lyse the cells and denature the DNA, then float it in neutralizing solution for 2 min. 3. Wash the filters in 6X SSC/O.S% SDS at room temperature for 1 hr, with vigorous shaking to remove the cell debris from the tilters. 4. After crosslinking the filters by W irradiation for 2 min, hybridize them with the same set of 32P labeled cDNA probes as for the dot-blots (Fig. 3; see Note 5).

91

Differential Screening by Reverse Northern

B

Rat1 abc

TlOl-4 d

a

b

c

d

Mob-5 & Control Mob-38

Mob-37 Mob-38 Mob-39

Fig. 2. (A) 1.5% agarose gel electrophoresis of the colony-PCR products from the cloned cDNA. Two randomly picked colonies for the positive control, Mob-5 and the negative control, four randomly picked colonies each for Mob-36, Mob-37, Mob-38, and Mob-39 were amplified by PCR with a pair of primers flanking the cloning site of the pCR-TRAP cloning vector (GenHunter). (B) Reverse Northern dot-blot of the colony PCR products from Fig. 2A. The left two dots on the top row of the duplicate blots are the positive control Mob-5, while the right two dots are the negative control, which indicates the equal labeling of the two cDNA probes from the Rat1 cells (left blot) and Tl 01-4 cells (right blot). (C) Northern blot analysis of Mob-5, showing the differential expression of the gene in H-ras transformed TlO l-4 cells. The 28s and 18s rRNAs are shown at the bottom as loading control.

4. Notes 1. As part of our ongoing effort to apply differential display to identify genes that are regulated by the rus signaling pathway, the immortalized rat embryo fibroblast cell line Rat 1 and its H-rus transformed derivative, T 10 l-4 (6) were compared. The genes that are expressed at a higher level in the T 10 l-4 cell line are of great interest. 2. We recommend UV cross-linking the membrane directly after blotting the DNA, since rinsing the wells with 6X SSC would create rings instead of round spots on the blot.

Zhang, Zhang, and Liang

92 Rat1

TlOl-4

Mob-37

Rat1 ..-

TlOl-4 --

Mob-38

Fig. 3. Colony hybridization of the duplicate filters with the same set of cDNA probes used for the dot-blot in Fig. 2B. (A) Mob-37, the arrow denotes a false positive clone, later confirmed by sequencing. (B) Mob-38, arrows indicate some of the positive colonies. 3. Mob-5, a previously cloned gene expressed only in the transformed cells (Liang et al., unpublished data), was used as a positive control. As a negative control, two clones from a band equally expressed from the differential display (Fig. 1A). Except for the false positive Mob-36d and the no-signal clones Mob-37d and Mob-38b and c, all other clones represent truly differentially expressed genes (Mob-36 through Mob-39). 4. Reverse Northern dot-blotting allows simultaneous screening of multiple cDNA fragments. However, two pitfalls needs special attention. First, the method is still cumbersome since the cloned cDNA has to be either amplified or isolated before blotting. Second, only a limited number of colonies for each cDNA fragment could be analyzed, so a truly differentially expressed gene might elude detection if the contamination of false positive genes is severe. The colony hybridization method described here should be able to overcome the aforementioned pitfalls.

Differential Screening by Reverse Northern

93

5 We have smce sequenced all four positive clones. Database search results mdicate that, while three of the cloned cDNA fragments represent potentially novel genes, Mob-36 corresponds to a well-characterized gene, osteopontin, which was known to be induced by oncogenic ras and involved in many tumors (7) 6. Much of this work has been published previously (8).

References 1, Liang, P. and Pardee, A. B. (1992) Differential display of eukaryotrc mRNA by means of the polymerase chain reaction. Science 257,967-97 1. 2. Liang, P. and Pardee, A. B. (1995) Recent advances in differential display. Curr Opmion Immunol

7,2742%X

3. Liang, P., Zhu, W., Zhang, X., Guo, Z., O’Connell, R., Averboukh, L., Wang, F., and Pardee, A. B. (1994) Differential display using one-base anchored ohgo-dT primer. Nucleic Acids Res 22,5763-5764. 4. Liang, P., Averboukh, L., and Pardee, A B. (1993) Distribution and clonmg of eukaryotic mRNAs by means of differential display refinements and optimization. Nucleic Aczds Res. 21,3269-3275. 5. Mou, L., Miller, H., Li, J., Wang, E., and Chalifour, L. (1994) Improvements to the differential display method for gene analysis. Biochem Bzophys Res Comm 199,564-569

6. Liang, P., Averboukh, L., Zhu, W., and Pardee, A. B. (1994) Ras activation of novel genes Mob-l as a model Proc Nat1 Acad. Scz. USA 91, 12,515-12,519. 7. Craig, A. M , Nemn, M., MukherJee, B. B., Chambers, A F., and Denhardt, D. T. (1988) Identification of the maJor phosphoprotein secreted by many rodent cell lines as 2ar/osteopontm* enhanced expression m H-ras-transformed 3T3 cells Blochem. Blophys. Res Comm 157, 166-173. 8. Zhang, H., Zhang, R., and Liang, P. (1996) Differential screening of gene expression difference enriched by differential display Nucleic Acids Res. 24,2454-2456

9 Screening for Positive Clones Generated by Differential Display Regina V8gelbLange, Niels Biirckert, Thomas Boiler, and Andres Wiemken 1. Introduction Differential display 1sa powerful method to identify genes that are differentially expressed m different tissues or m the same tissue exposed to different treatments (1,2). The method can be performed wtth relatively little biological material, and the resultmg partial cDNAs can be cloned, sequenced, and used for subsequent expression studies. cDNA fragments derived from downregulated as well as upregulated transcripts can be detected simultaneously, and cDNAs corresponding to messages of high and low abundance are detected with similar probabtlities. One drawback of this method, however, is the relatively high number of false positives, i.e., of cloned cDNAs that give no differential hybridization pattern on Northern blots. In part, this could be attributed to exogenous factors such as bad PCR tubes (3), degraded RNA or contammatmg DNA. In addition, however, there is an intrinsic problem of the method itself. There is increasing evidence in the literature that single bands with differential appearance on gels, here called differentials, represent a mixture of several cDNA fragments of identical size but different sequence. Thus the cDNA clones obtained from one particular differential may be derived from several different mRNAs, includmg both differentially and constitutively expressed ones. It 1s of interest to select for further characterization only the clones that represent truly differentially expressed genes. For this purpose, Lt et al. (4) proposed to screen several cDNAs from one differential by Northern blot affinity capturing. This method is applicable only when large quanttties of mRNA are available. An alternattve approach consists of spotting plasmtd DNA of several individual clones From

Methods m Molecular Bology, Vol 85 D/fferenbal Edlted by P Llang and A 8 Pardee Humana

95

Display Methods and Protocols Press Inc , Totowa. NJ

Vligeli-Lange

96

et al.

obtained from one differential onto a membrane that IS then hybridized to a portion of the orrginal differential eluted from the polyacrylamide gel (5). This method may work well if the bulk of the radioactivity m the ortgmal differential actually represents a single differentially expressed gene. However, rt will fail rf the differential represents mainly a constitutively expressed gene or rf rt is composed of several different cDNAs. To circumvent this problem, we hybridize the original 33P-labeled differential display products from both the control RNA and the RNA source of interest to duplicate samples of several individual clones obtained from one differential (6). By comparing the hybridization patterns, we can easily select clones representing sequences that were differentially amplified from the two sources of RNA. Positive clones identrtied by our new screening method (6) are then promising candidates for truly differentially expressed genes and can be used for further analysts.

2. Materials All solutions are prepared with autoclaved double-distilled or filter-purified water and should be stored at room temperature unless stated otherwise.

2.7. Differential Display and CIoning of Differen tia//y Displayed c DNA Fragments 1. 2 3 4. 5. 6.

7.

8. 9. 10 11. 12. 13. 14.

TE buffer: 10 mM Tns-HCl, pH 7.4, 1 mM EDTA. Autoclave. 3M Sodium acetate, pH 4.8-5 2 Autoclave

Glycogen (10 mg/mL). Store at -20°C. 100% Ethanol 85% (v/v) Ethanol. Store at -20°C H-T1 tM and H-AP primers (GenHunter Corporation, Nashville, TN) for reamphfication of eluted drfferentials: 2-w stock solutions. Store at -20°C Use the same set of primers that has been used to generate the differentral (see Chapter 1 for details) dNTP stock: 250 @4 each of dATP, dCTP, dGTP, dTTP (Pharmacia, Uppsala, Sweden). Prepare by mrxmg equal amounts of 1-mI4 stock solutions of each of the four dNTPs. Store at -20°C DNA polymerase (5 U/pL), e g , AmphTuq (Perkm Elmer). Store at -20°C. lOXPCRbuffer* lOOmMTris-HCl,pH8 3,500mA4KCl, 15 mA4MgClB,0.01% (w/v) gelatin. Store at -2O’C Cloning vector for PCR products, e.g., pCR-TRAP (GenHunter Corporation). Store at -20°C. T4 DNA ligase (200 U/&) Store at -20°C. Competent E coli cells Store at -7O’C. LB medium: Add 10 g tryptone, 5 g yeast extract, and 9 g NaCl to 1 L HzO. Autoclave; store at 4’C. LB plates, 20 pg/mL tetracycline: LB medium supplemented with 15 g agar Autoclave, when cooled down to 60°C add 1 mL tetracycline solution (20 mg/mL in 95% ethanol). Pour plates, store at 4°C in the dark.

Screening for Positwe Clones

97

I5 Colony lys~s buffer TE buffer with 0 1% Tween-20 Store at -20°C 16 Primers flanking the cloning site, 2 M each. Store at -20°C

2.2. Slot Blot Analysis of Cloned Differentials 1 2 3 4

20X SSC stock solution 3MNaCl,0.3MNa3-citrate, pH 7 0 Autoclave 0.5M NaOH, 1 SM NaCl Autoclave. 0 5MTris-HCl, pH 7 5, 1 5MNaCl Autoclave 50X Denhardt’s stock solution Dissolve 5 g Flcoll (Type 400, Pharmacla), and 5 g polyvmylpyrrolldone, 5 g bovine serum albumin m a total volume of 500 mL H,O Store in ahquots at -2O’C 5 10% SDS. 6 Salmon sperm DNA stock 10 mg/mL m TE buffer, shear DNA by somcatlon Store in aliquots at -20°C.

3. Method 3.1. Differential Display and Cloning of Differentially Displayed cDNA Fragments The basic method for differential display 1sdescribed m detail m Chapter 1 (see also refs. I and 2) We perform differential display with one-base anchored primers flanked at the 5’-end by an Hind111 restrlctlon site (6,7) (see Note 1). For elution and cloning of bands with differential appearance on polyacrylamide gels, the differentials, we employ the modifications described below: 1 Cut out band of interest from dried gel with a new razor blade and place m a 1 5-mL Eppendorf tube (see Note 2) To check whether the appropriate band has been removed from the polyacrylamlde gel, re-expose gel to X-ray film. 2 Add 100 pL of TE buffer, seal tube with parafilm Incubate for 2 h at room temperature followed by 1 h at 37°C with occaslonal shaking. Then Incubate at 4°C for 1 h up to overmght with gentle agltatlon. 3. Centrifuge at 12,OOOg for 2 min, and transfer supernatant to a fresh tube (see Note 3) Precipitate eluted DNA by adding 10 PL 3M sodium acetate buffer, pH 4 8-5.2,5 & glycogen (10 mg/mL), and 450 $ 100% ethanol Place tube for 40 mm in a -70°C freezer 4. Centrifuge for 10 mm at 12,000g and 4°C to pellet the DNA. The glycogen carrier will produce a white pellet at the bottom of the tube Remove the supernatant and rinse the pellet with 200 pL Ice-cold 85% (v/v) ethanol. Centrifuge for 2 mm at room temperature and remove remaming llquld with a mlcroplpet tip Air dry the pellet briefly and then dissolve in 10 $ HZ0 Store at -20°C If not used dn-ectly for reamphficatlon 5 Reampllfy the eluted cDNA fragments by using 3 & of the eluate m a 30-& PCR reaction with the same set of primers used to generate the differential. Final concentrations of the primers are 0 2 cln/r,dNTPs are used each at 20 w with 0 5 U AmpllTaq polymerase (Perkm Elmer). PCR IS performed m 30 cycles of 94°C for 30 s, 40°C for 2 mm, and 72°C for 30 s with a final extension of 5 min Run

98

Vtigeli-Lange et al.

15-20 p.L of the PCR reaction mixture on a 1 3-l 5% agarose gel containing ethldmm bromide and compare the size of the reamphfied PCR products to the size of the correspondmg differential (see Note 2). For ligation, use only PCR products that give a smgle band of expected size If several bands of different size are detected on the agarose gel, prick the expected band once with a fine mlcropipet tip and use this as moculum for a fresh PCR reaction This often helps to produce a single band of expected size 6 Ligate PCR products directly mto an appropriate vector We use the pCR-TRAP clonmg vector (GenHunter Corporation) that has the advantage that only plasmlds with inserts confer tetracycline resistance to transformed E co11 cells, thus allowing direct positive selection for cDNA inserts For ligation, mix 10 & H,O, 2 pL Insert-ready pCR-TRAP vector, 2 pL 1OX hgatlon buffer, 5 & PCR product, and 1 & T4 DNA hgase (200 U/pL) Mix gently and incubate overnight at 16°C 7 Streak out all the cells from a transformation mix with 10 pL of the ligation mix and 100 pL competent E colz cells onto 1 to 2 LB plates supplemented with tetracycline (20 pg/mL) (see Note 4) 8 For each differential, screen approx 10 E colz colonies for plasmlds with inserts of correct size by colony PCR or by restrlctlon enzyme analysis of plasmld DNA For colony PCR, place a small amount of E colz cells mto a 1 5-mL Eppendorf tube contammg 50 pL of TE buffer supplemented with 0.1% Tween-20 Incubate the tubes at 95°C for 10 mm. Spm for 2 mm at 12,OOOg,and transfer 35 pL of the supernatant to a fresh tube Use the supernatant directly for PCR with primers flanking the clonmg site, or store at -20°C for f?uther use (see Note 5)

3.2. Sot Blot Analysis of Cloned Differentials A single band in the drfferential display polyacrylamide gel 1s often composed of several cDNA fragments of identical size but different sequence Therefore, It IS of importance to dlstmgulsh between cDNA fragments that are derived from truly differentially expressed genes and those derived from contaminating constitutively expressed genes. We do this by reverse RNA slot blot analysis using the original 33P-labeled PCR products as hybridization probes. When several candidate positive clones are found, DNA slot blot analysis can be employed

m a second step to determine

whether the different clones

are homologous. For each differential to be analyzed, three replicate membranes are prepared containing the cDNA inserts of several independent clones obtained by transformation of E coZi with that particular differential. Two of the replicate membranes, membranes A and B, are then used for reverse RNA slot blot analysis, the third one, membrane C, can be used subsequently for DNA slot blot analysis if desired. Our general protocol for slot blot analysis of cloned differentials IS as follows* 1 Using a primer pair flanking the cloning site, PCR amplify cDNA from 6-10 independent E co11colonies contaming inserts of correct size obtained by trans-

Screening

2.

3

4. 5

6.

7.

8

for Positive

Clones

99

formation with the differential to be tested For a 20-k PCR reaction, use 2 pL of colony lysate as DNA template, with the final primer concentratton at 0.2 pJ4 and the final concentratton of dNTPs each at 20 @4 PCR products can be used directly for slot blot analysis or stored at -20°C untrl used For slot blot analysts, float a membrane (e.g., NY 13 N, Schleicher and Schuell, Dassel, Germany) first on water, then on 6X SSC Assemble slot or dot blot machine (e.g., Gtbco-BRL). Wash each slot with 0 5 mL 1OX SSC. For each sample to be analyzed, place 19 pL PCR product into a 1 5 mL Eppendorf tube, heat for 5 min at 95°C chill on ice, and quick spm to collect condensate Add 475 pL 10X SSC, mix well. Apply this mixture m three ahquots of 156 pL to three slots or dots of the filtration manifold. Wash each slot twice with 0 5 mL 1OX SSC Disassemble filtration manifold, and float membrane for 7 mm on a solution conststmg of 0 SMNaOH, 1.SMNaCl. This is followed by 2 floatatton steps on 0.5M Tris-HCl, pH 7 5, 1.SMNaCl for 3 mm each Wash membrane briefly m 2X SSC and, with a sharp razor blade, cut mto three identical sets, A, B, and C, each containing the cDNA mserts of 6-10 E coEi colonies obtained by transformation with the differential under investigation Bake membranes for l-2 h at 80°C m a vacuum oven. Membranes A and B then can be used directly for hybridization or, wrapped in aluminum foil, stored together with membrane C at room temperature until used Use membranes A and B for reverse RNA blot analysis Prehybrrdize membranes for 14 h at 60°C in 5 mL of 6X SSC, 5X Denhardt’s solution, 0 5% SDS, and 100 pg/mL denatured salmon sperm DNA. Decant prehybridization solution, replace with 4 mL of fresh prehybridtzatton solution To membrane A, add 15 pL of the original 33P-labeled PCR products from the control tissue denatured for 5 mm at 95°C and to membrane B the corresponding 33P-labeled PCR products from the tissue under mvestigation (see Note 6). Hybridize overnight at 60°C. Remove hybridization solution, and wash membranes twice for 15 mm each at 60°C m 10 mL of 0 2X SSC, 0 2% SDS. Dry membranes briefly on Whatman 3MM paper, wrap m Saran fotl and expose for 6-48 h to a PhosphoImager high sensttivtty screen (BroRad) or, if unavailable, to X-ray film. Compare hybridizatton patterns of blot A and B (see Fig 1) cDNA mserts representing sequences that are differentially amplified, and therefore probably derived from truly up- or downregulated genes, are easily recognized by their differential hybridization pattern, whereas cDNA mserts representing sequences that are equally amplified, and therefore likely derived from contammatmg constitutlvely expressed genes, show identical hybridization patterns with membrane A and B (see Note 7). If several cDNA inserts with differential hybridization patterns are obtained from a single differential, PCR amplify one of these inserts from colony lysate usmg the ortgmal differential display primers and the PCR conditions for reamphficatton of differentials (see Note 8). Use half of the PCR product obtained from a 20 pL reaction to prepare a 33P- or 32P-labeled probe by random primed labeling

Rigeli-Lange

100

A

A

7

9

et al.

Fig. 1. Selection of clones with cDNA inserts from differentially expressed genes by reverse Northern blot analysis utilizing differential display PCR products as hybridization probes. One hundred nanograms of PCR amplified inserts of ten individual E. coli colonies obtained by transformation with the differential Myc 4 (H-T, ,C and H-AP3) were immobilized in duplicate sets onto nylon membranes and hybridized with the original 33P PCR product from control roots (A) or from mycorrhizal roots (B). The order of inserts is from left to right starting with number 1 and ending with number 10. Clones with differentially expressed cDNAs are indicated by arrows. Used with permission from ref. 6.

and, using the conditions described in step (6), hybridize the probe to membrane C. An example of this where all positive clones represent homologous sequences is shown in Fig. 2A (see Note 9). Putative positive clones then can be sequenced. To verify differential expression of putative positive clones, the other half of the PCR product can be used to prepare a 3ZP-labeled hybridization probe for RNA gel blot analysis (see Fig. 2B). So far, we have tested this approach with 50 cloned differentials and for 10 of them RNA gel blot analyses were performed. In five of these cases, differential expression was confirmed by RNA gel blot analysis, the other five probes gave no detectable signals on Northern blots with 10 ~18of total RNA. However, none of the probes tested hybridized to both RNA from control tissue and RNA from treated tissue. This indicates that our screening method is useful to eliminate false positives. The probes that produced no detectable signal on Northern blots most likely are derived from rare messages.More sensitive methods for RNA analysis, such asribonuclease protection assays(81or poly(A) RNA gel blot analysis may help detect differential expression of such messages.

4. Notes 1. A convenient way to get started on differential display is to use one of the kits distributed by GenHunter Corporation. As radioactive nucleotide we prefer 33P over 35S since it is less volatile and produces sharper bands on polyactylamide gels (9).

101

Screening for Positive Clones

Clone Number

f

1

3

5

7

9

11

control roots

13

15 17

I

3

5

7

9

11 13 15 17

3.6 kb

days

mycorrhizal roots

Fig. 2. (A) Classification of clones with cDNA inserts from differentially expressed genes. H-T1 ,C and H-AP3 PCR product of clone pMvc 4.7 was labeled with a[33P]dATP and hybridized to a membrane containing the cDNA inserts of clones piVyc 4.1 to pMyc 4.10. Hybridization signals were quantified using a Bio-Rad GS-250 Molecular Imager. (B) Northern blot analysis of the Myc 4.7 probe. Two membranes, one containing 10 pg/lane RNA from nonmycorrhizal control roots, the other one 10 mg/lane RNA from mycorrhizal roots, were hybridized simultaneously with a[32P]dATP labeled Myc 4.7 (1 x lo6 cpm/mL.). Days indicate time after inoculation with fungus or corresponding control treatment. Used with permission from ref. 6. 2. Since differential display predominantly generates cDNA fragments from the 3’-end of messenger RNAs, it is advantageous to select bands that are more than 250 bp in size. Shorter fragments often correspond to the 3’-untranslated region of a gene, and thus are less helpful for homology searches in the database. A 33P end-labeled DNA marker can be prepared from any DNA ladder with 5’ overhang in a till in reaction with the Klenow fragment of DNA polymerase. We use a Klenow fill-in kit (Stratagene) to label the @X174/ HintI DNA marker (Stratagene) resulting in a ladder of labeled bands ranging from 726 to 24 bp. 3. Store the pellet at -20°C in case of mishaps. Also, if reamplification was unsuccessful, reextracting the excised band often is more helpful than diluting the products of the first round PCR into a fresh PCR reaction.

102

Vtigeli-Lange et al.

4. Basically any tetracycline sensitive E. coli strain can be used for transformation. The strain provided with the GenHunter pCR-TRAP cloning system is not further specified. In addition to the GenHunter strain, we have been using E. coli strain DHSa. What seems however important, is that prior to plating on selective medium, the recovery period of the E. coli cells in LB medium should be extended from 1 to at least 2 h. This increases the number of tetracyline resistant colonies obtained approximately by a factor of five without affecting positive selection. 5. We do 15-pL PCR reactions with 1.5p.L colony lysate and run the entire PCR reaction mixture on an agarose gel. On average, three out of four colonies screened obtain inserts of correct size. We store the colony lysate at -20°C in 0.5-mL tubes to safe freezer space, Colony lysate is also a useful source for the amplification of cDNA to produce hybridization probes. 6. For best results, we combine PCR products from two or three reactions derived from the same primer combination. Therefore, we either run the differential display with several replicates, or we use RNA extracted from a time-course experiment and select the time points containing the differential of interest. If several differentials are obtained with the same primer combination, they can all be screened simultaneously in the same hybridization solution. 7. The whole procedure, including differential display and screening for positive clones, can be completed in a period of two weeks. When fresh 33P(200 Ci/mmol) is used in the differential display, up to 4-wk-old PCR reactions can be used as hybridization probes for the screening step. Often, cDNAs from different clones representing the same differential hybridize with different intensities to the 33P PCR products of one particular reaction (see Fig. 1A or B). As the cDNAs were amplified from colony lysate, where the input of DNA template is not strictly controlled, differential hybridization intensities within the same blot cannot be taken as an indication for the presence of clones with different sequence. 8. In this PCR reaction, do not use primers flanking the cloning site, as these will produce short stretches of DNA that will hybridize to all of the PCR products to be tested. 9. As an alternative to the DNA slot blot step, select several of the cDNAs with differential hybridization pattern directly for DNA sequencing.

Acknowledgments We thank Oxford University Press for the permission to reproduce Figs. 1 and 2 from ref. 6. References 1. Liang, P. and Pardee, A. B. (1992) Differential display of eukaryotic messenger RNA by means of the polymerase chain reaction. Science 257,967-97 1. 2. Liang, P., Averboukh, L., and Pardee A. B. (1993) Distribution and cloning of mRNAs by means of differential display: refinements and optimization. Nucleic Acids Res. 21,3269-3275.

Screening

for Positive

Clones

103

3. Chen, Z., Swisshelm, K., and Sager, R. (1994) A cautionary note on the reaction tubes for differential display and cDNA amplification in thermal cycling. BioTechniques 16, 1003-1006. 4. Li, F., Barnathan, E. S., and Kariko, K. (1994) Rapid method for screening and cloning cDNAs generated in differential mRNA display: application of Northern blot for affinity capturing of cDNAs. Nucleic Acids Rex 22, 1764,1765. 5. Callard, D., Lescure, B., and Mazzolini, L. (1994) A method for the elimination of false positives generated by the mRNA differential display technique. BioTechniques 16, 1096-l 103. 6. Vogeli-Lange, R., Btirckert, N., Boiler, T., and Wiemken, A. (1996) Rapid selection and classification of positive clones generated by mRNA differential display. Nucleic Acids Res. 24, 1385,1386. 7. Liang, P., Zhu, W., Zhang, X., Guo, Z., O’Connell, R. P., Averboukh, L., Wang, F., and Pardee, A. B (1994) Differential display using one-base anchored oligodT primers. Nucleic Acids Rex 22, 5763,5764. 8. Yeatman, T. J. and Mao, W. (1995) Identification of a differentially-expressed message associated with colon cancer liver metastasis using an improved method of differential display. Nucleic Acids Rex 23,4007,4008. 9. Trentmann, S. M., van der Knaap, E., and Kende, H. (1995) Alternatives to 35S as a label for the differential display of eukaryotic messenger RNA. Science 267, 1186.

Cloning of the 3’ Noncoding Regions from Several Members of Heat Shock Protein Gene Families by Differential Display Chandrashekhar P. Joshi and Henry T. Nguyen 1. Introduction Several structural and metabolically important proteins m eukaryotes are encoded by multigene families. For understanding the function of each member of a gene family, it is essential to isolate the representative cDNAs encodmg the protein of interest. In the past, this has been especially difficult owing to highly similar protein coding region of the cDNAs from a gene family with minute variations in then ammo acid composition. The 3’ noncoding regions from different members of a gene family, however, show sufficient heterogeneity that can be harnessed for the isolation and characterization of different members of a gene family. Higher plant genes offer a special advantage because the 3’ noncoding regions of their cDNAs are generally shorter than 300 bp m length (1). Plant cDNAs are, therefore, easily amenable to dtfferential display analysis. We are interested m understanding the functions of different heat shock proteins that belong to multigene families m crop plants (2). In response to heat stress,several heat shock protem (HSP) genes are actively transcribed and preferentially translated (3). HSPs belong to two major groups namely high-molwt HSPs (60-100 kDa) and low-mol-wt HSPs (15-30 kDa) that are further subclassified mto several HSP families depending on their observed/expected molecular masses. In most cases, each subclass of HSPs is encoded by a multigene family of 3-15 members. In general, these HSPs are proposed to perform the protective functions as molecular chaperones before, during and after heat shock treatments (4). However, the function of each of these HSP members belonging to multigene families is largely unknown From

Methods m Molecular Bology, Vol 85 D/fferenbat Edlted by P Llang and A B Pardee Humana

107


Joshi and Nguyen

108 mRNA population

Tl2VA ----

T12VC T12VG Tl2VT

CH

CHCHCH I

Reverae Transcription I

(dNTPs, MMLV RT, Tl2VN

Flmt Strand

cDNA

primer)

I

I I

synthdc

I lssm

N, HSP16.9primer)

40 cycles

Loading on DNA Sequencing gel and electrophoresis *

t Autoradiography

Fig. 1. Schematicdiagram of modified differential display.

We successfully applied a simple modification of differential display technique for the isolation of 3’ noncoding regions from several members of gene families encoding HSP16.9, HSP26.6 and HSP70 in wheat (5-7). We replaced the arbitrary 5’ primer in conventional differential display (8,9) with a conserved HSP gene-family specific primer and have isolated cDNAs representing several HSP family members (see Fig. 1). In this chapter, we describe the detailed procedure used in our laboratory to isolate these cDNAs that includes: (1) growth of wheat seedlings and heat shock treatment; (2) isolation of total RNA; (3) DNase treatment of total RNA; (4) reverse transcription of m.RNAs; (5) synthesis of gene family specific primer; (6) differential display by radioactive PCR and polyacrylamide gel electrophoresis; (7) reamplification of cDNA probes; (8) cloning of cDNA probes; and (9) characterization of cDNA clones. Although the current application describes the use of modified differential display for HSP gene families in wheat, it can also be easily applied to clorie any multigene family members from other plant species (Joshi, unpublished observations). 2. Materials 2.1. Plant Material 1. Wheat (Triticum aestivum L.) cultivar “Mustang” seeds. 2. Controlled growth chamber.

3’ Noncoding Regions

109

2.2. RNA Extraction 1 Guamdme buffer (GB): 8M guamdme hydrochloride, 20 rnJ4 MES (2-[NMorpholino] ethanesulfonic acid), 20 mM EDTA, 50 mM mercaptoethanol (BME), pH 7.0 (see Note 1). Make fresh. 2 Phenol/chloroform/isoamyl alcohol (PCI) mixture (25/24/l v/v), store at 4°C 3. Chloroform/isoamyl alcohol mixture (24: 1 v/v), store at 4°C. 4. 100% Ethanol, store at -20°C . 5. IA4 Acetic acid, store at 4°C 6 3M Sodmm acetate pH 5.2, store at room temperature. 7. 70% Ethanol, store at 4°C 8. Diethyl pyrocarbonate (DEPC) treated sterile H20.

2.3. DNase Treatment All the components required for this experiment are available as a single kit (Messageclean, M60 1) from GenHunter Corporation, Nashville, TN. Individual components of the kit stored at -20°C are listed below: 1. 10X reaction buffer: 100 mM Trts-HCl, pH 8.4, 500 mA4 KCl, 15 mM MgC12, 0.01% gelatin. 2. DNase I* 10 U/pL In addition,

the following

materials

are also required.

3. 95% Ethanol store at -20°C. 4. Phenol: chloroform mixture (3: 1) store at 4°C. 5. 3M Sodium acetate, pH 5 2, store at room temperature

2.4. Reverse

Transcription

All components required for differential display experiment are available as a single kit @NAtnap, M50 1) from GenHunter Corporation, Nashville, TN, and should be stored at -20°C. Individual components of the kit in thts section are listed below: 1. 5X RT buffer: 250 mMTris-HCl, pH 8.3,375 mMKCl,l5 mMMgCl,, 25 mMDTT. 2. dNTP 250 pA4. 3. T,,VN primers (10 l.tA4)mcluding T12VA, T,,VC, T,,VG, T,,VT 4. MMLV Reverse transcrtptase (RT): 100 U/pL.

2.5. Gene Family-Specific Gene family-specific

2.6. Differential

primers:

Primer Synthesis 4-5 l..uW.Store at -20°C

until use.

Display

Individual components of the kit stored at -2O”C, RNAmap (M50 1, GenHunter) used in this section are listed below: 1. 10X PCR buffer: 100 mM Trts-HCl, pH 8.4, 500 mM KCl, 15 mM MgCl*, 0.01% gelatm.

Josh/ and Nguyen

110 2 dNTP (25 /.&!).

3. Gene family-specific primer (4 w 4. T,,VN primers (2 j.uV) In addition, the following materials are also required. 5 6. 7. 8

35S dATP (1200 Wmmole) AmpliTaq polymerase (Perkin Elmer) Mineral oil DNA loading dye 95% formamide, 10 WEDTA, pH 8.0,0.09% xylene cyanol, 0 09% bromophenol blue 9 DNA sequencing gel and electrophoresls apparatus, gel dryer 10 X-ray film

2.7. Reamplification of cDNAs Individual components of the kit, RNAmap M50 1 (GenHunter Corporation) used m this section are listed as follows* 1. 2. 3 4

Materials listed m Section 2.6. Glycogen 10 mg/mL. 3M Sodium acetate. Ice-cold ethanol.

2.8. Cloning of cDNAs PCR-Trap kit (P404) from GenHunter used for this se&on. The individual components are listed below: 1. pCR-Trap vector precut with restriction enzyme that produces blunt ends. 2 1OX Ligation buffer: 500 mA4 Tns-HCl, pH 7 8, 100 mA4 MgCl,, 100 mM DTT, 10 mMATP, 500 pg/mL BSA. 3 T4 DNA hgase (200 U/pL). 4. Competent bacterial cells. 5. LB liquid medium (10 g bacto tryptone, 5 g yeast extract, 9 g NaCI, and 15 g bactoAgar per liter) and LB-tetracycline solid medium plates 6 Colony lysis buffer TE buffer with 0.1% Tween-20 7 10X PCR buffer 100 mM Tris-HCl, pH 8 4, 500 mA4 KCl, 15 mM MgC&, 0 0 1% gelatin 8 250 MdNTP.

9 Lgh primer (2 @4), CGACAACACCGATAATC 10 Rgh primer (2 CIM), GACGCGAACGAAGCAAC 11 AmpliZ’uq Polymerase (Perkin Elmer, Foster City, CA).

2.9. Characterization

of cDNA C/ones

1 Sequenase Version 2 0 kit from United States Blochemlcals (Cleveland, OH) 2. AldSeq Kit C (cat. no P203) from GenHunter including Lseq (ATCACG-

AGGCCCTTTCG) and Rseq (GATACTGGACGCGAGCC) primers (1 @!)


111

3. Methods 3.7. Heat Shock Treatment

of Wheat Seedlings

1. Grow seedlmgs of winter wheat cultivar “Mustang” for 10 d m 50-mL pots of motst vermiculite at 22 k 2°C under a fluorescent light bank provtdmg 300 pm01 PPFD m-* s-i with a 12-h photopertod Relative humidtty should be about 30% on an average 2. Expose IO-d-old seedlings to heat stress at 37°C for l-2 h m a controlled growth chamber with 100% relative humidity. High humidity should be maintained m order to limit transptrattonal coolmg of the leaves. 3 After heat shock, excise the leaf tissue and freeze it immediately m hqutd N, and keep at -75°C for total RNA tsolatton. Leaf tissues of 10 d seedlmgs grown as described above but not heat shocked should be used as control samples

3.2. Total RNA Isolation from Control and Heat Shocked Wheat Seedlings Isolate total RNA from 5-g control and heat-shocked leaf tissues by the Guamdme

hydrochloride

extraction method described by Logematm

et al. (10)

with minor modifications as described below although other methods of RNA extraction are found to be equally successful (Joshi, unpublished observation). 1. Freeze seedling tissue of interest (approx 5 0 g) in liquid N2 and grind frozen tissue to a fine powder using precooled mortar and pestle Grind the tissue two more times with extra liquid mtrogen. 2 Add 2 vol of GB (2 mL for every gram of tissue) (approx 10 mL ) to the ground ttssue and mix it well with a sterile spatula. 3. Add 1.Ovol(2 mL for every gram of tissue) of PC1 mixture (approx 10 mL) to the mortar and mtx it well with a sterile spatula Carefully transfer the contents (green slurry) to a 50-mL Nalgene tube ( or any chloroform resistant centrifuge tube) 4. Centrifuge for 20 min at 10,OOOg 5 Collect aqueous yellowish phase and put it m a clean 50-mL Nalgene tube Use miracloth to filter the aqueous phase if floating debris is present 6. Add 1 vol of chloroform/isoamyl alcohol (24: 1 v/v) mixture and mix by inversion 7 Centrifuge the tubes for 10 min at 10,OOOg 8. Collect the aqueous phase (approx 15 mL) and place tt in a 30-mL glass tube. 9 Add 0 2 vol of 1M acetic acid (approx 3 mL) and mix the contents by inversion 10 Add 0.7 vol of precooled ethanol (approx 10.5 mL). 11 Cover the tubes with paratilm and gently mix by inverston. Total RNA 1sprecipitated and DNA/proteins stay m solution. 12. Store for 1 h or more at -70°C. If any DNA fibers are observed floating m the completely thawed solution, spool them out on a sterile glass loop and discard the DNA. 13. Pellet total RNA by centrifuging for 10 min at 10,OOOg. Thts should result in a large and white pellet. Discard the supernatant 14. Wash the RNA pellet twice with 5 mL of sterile 3M sodium acetate (pH 5.2) at room

temperature

to remove

low molecular

weight

RNAs

and contaminatmg

112

15 16 17 18

Josh/ and Nguyen polysacchartdes. Subsequently, wash the RNA pellet once wtth 70% ethanol to remove the salts After every wash, tubes must be centrifuged for 5 mm at 10,OOOg and supernatant 1s discarded. An dry the RNA pellet for approx 1O-l 5 mm at room temperature. Dissolve the pellet m stenle DEPC treated H,O (approx 400-500 pL, target 4 pg/pL) Transfer the contents to sterile 1 5-mL Eppendorf tube Check the RNA concentratron and quality by spectrophotometer. Also run a denaturing Agarose gel to check the mtegrtty and quality of RNA A Northern blot analysts may be performed followmg the method described by Joshi et al (6) to confirm that the total RNA isolated is of a best quality and free of any impurities

3.3. Removal of Genomic DNA Contamination from the Total RNA With any method of total RNA extraction there is always some possibility that traces of genomic DNA contaminants are prectpttated along wtth total RNA isolated. This genomic DNA could act as template for reverse transcrtp-

tion independent amphfication of DNA bands that may or may not be dtfferentially expressed. This contamination also increases proportions of RAPDs (random

amplified

polymorphtc

DNAs)

that are undesirable

in differential

display experiments (II). It is therefore, essential to remove this genomic DNA by using good quality

of DNase that is not contaminated

with RNase

This is

achieved in this step. (Use Message Clean kit [cat. no. M601] [GenHunter Corporation] .) 1. For DNase I Dtgestion m a 0.5-mL tube, add the following m the same order Total RNA 50 0 /.lL (-50 pg) 10X Reaction buffer 57ccc DNase I (10 U/pL) 1ociL Mix well and incubate for 30 min at 37°C. Longer incubattons are not advisable. 2 Phenol/chloroform (3 1) extraction is performed to remove proteins, Add 40 pL of phenol/chloroform mixture. Vortex for 30 s. Let the mixture sit for 10 mm at room temperature. Spin tubes in Eppendorf mtcrocentrifuge at high speed (14,000 rpm) for 5 mm Collect the top aqueous phase 3 Add 5 pL 3M sodmm acetate and 200 pL 95% ethanol (dtluted with DEPC H20) to the aqueous phase from step 2. A turbid solution 1s immediately formed. Let the mixture sit for >l h at -80°C 4 Spin tubes for 10 mm at 4°C. Remove supernatant A white pellet 1s clearly visible. Wash RNA pellet with 0 5 mL 70% ethanol (made with DEPC treated water) Spm tubes for 5 mm Remove ethanol An-dry the pellet for 10 min 5. Redtssolve RNA m 50 pL DEPC- treated H20. The pellet dissolves easily 6 Quantttate RNA at ODpGOand adjust an aliquot to 0.1 pg/pL with DEPC-Hz0 Use 2 pL (0.2 ug) of DNA-free total RNA for reverse transcrtptron. Use the diluted RNA solutton only once and discard the remaming solutton. Do not dilute more RNA than requned.

173


3.4. Reverse Transcription of mRNA Dlfferenttally expressed cDNAs from control and heat stressedRNA samples should be visualized on sequencing gels (6% denaturing polyacrylamrde) followmg the methods published by Lrang and Pardee(8), Liang et al. (9), and Josh1et al. (6) RNAmap kits from GenHunter Corporation (Brooklme, MA) can be used for this purpose followmg the manufacturer’s suggestions.The entire procedure used IS briefly described below (seeNotes 2 and 3). (Use RNAmap no. M501 kit.) Thaw out the following components and set on ice. Set up four reverse transcription reactions for each RNA sample m four PCR tubes (0.5~mL size).Each tube contains one of the four different Tt2VN asfollows (V = A/C/G; N = A/C/G/T): For 20 p.L final volume. dH,O

94$

5X RT buffer dNTP (250 pA4) Total RNA (DNA free)

40&

T12m

(10 44

Total

16cLL 2.0 Ilc 2oN-

(0 1 !%W)

19op.L

1 Program your thermocycler 65°C for 5 min, 37°C for 60 mm, 95°C for 5 mm (store at 4°C until use)

2. 1 $ MMLV reverse transcriptaseIS addedto eachtube 10mm after at 37°C 3 Mix well quickly by finger tipping 4. At the end of the reverse transcription, spm the tube bnefly to collect condensatron 5. Set tubes on ice for PCR or store at -20°C for later use

3.5. Design and Synthesis of the HSP Gene Family-Specific Primer This is the most important step in this modified application. First, all available protein sequences encoded by a gene family from same or different plants should be assembled using amino acid sequence alignment programs such as ‘pileup’ from Genetics Computer Genetics sequence analysis package allowmg possible gaps (e.g., Fig. 2). A highly conserved region near the carboxyl terminal end of the protein IS selected including a stretch of 1O-l 5 ammo acids and cDNA sequences encoding that amino acid stretch should be collected and aligned using “ptleup” routine This will clarify if any particular region has third base degeneracy or different codons coding for the same amino acids are present in this region. A DNA sequence of 14-18 bp is selected for the primer synthesis on the basis of the following criteria: 1. GC content of the potential primer sequence should be 6&70%. 2 Degeneracy 1s limtted to thud base of codon although attempt should be made to reduce the degeneracy as much as possrble. Degenerate primers also work very

114

Josh/ and Nguyen 201

Tahsp266b Tahsp266a Mzehsp26x Pshsp21 Gmhsp22 Phhsp21 Athsp21

GGDGWWKERS GGDGWWKERS DGDGWWKQRS GGEDCWSRKS GGDDSWSSRT GKDDSWG RN .SDDSWSGRS

Tahsp266b Tahsp266a Mzehsp26x Pshsp21 Gmhsp22 Phhsp21 Athsp21

251 262 VIDVQVQ* VIDVQVQ* VIDVQVQ* VIDVQIQ* VIDVQVQ* VTDVEIK* VIDVQIQ*

250 LSSYDMRLAL VSSYDMRLAL VSSYDMRLAL YSCYDTRLKL YSSYDTRLKL YSSYDTRLSL VSSYGTRLQL

PDECDKSQVR PDECDKSQVR PDECDKSKVR PDNCEKEKVK PDNCEKDKVK PDNVDKDKVK PDNCEKDKIK

AELKNGVLLV AELKNGVLLV AELKNGVLLV AELKDGVLYI AELKNGVLYI AELKNGVLLI AELKNGVLFI

SVPKRETERK SVPKRETERK TVPKTEVERX TIPKTKIERT TIPKTKVERK SIPKTKVEKK TIPKTKVERK

Fig. 2. Multiple sequence ahgnment of the carboxyl terminal region from the avatlable plasttd-localized HSPs from various higher plants. The ammo acid sequence RKVID (printed in bold letters) was selected as a possible conserved ammo acid domain to be used for destgnmggene family-specific primer * denotesthe translation stop codon Tahsp266b. wheat HSP26,6b, EMBL accession no. x67328; Tahsp266a. wheat HSP26 6a, EMBL accessionno x58280; Mzehsp26x maize HSP26, EMBL accessionno L287 12, Pshsp21 pea HSP2 1, EMBL accessionno x07 187, Gmhsp22 soybean HSP22, EMBL accession no x07188, Phhsp21 petunia HSP2 1, EMBL accessionno x54 103 and Athsp2 1 Arabidopsts HSP2 1, EMBL accessionno x.54102

well but do produce higher percentage of false positives (6) Replacement of N (any base)by Inosme residue ISalso advisable 3. Conserved DNA sequence follows the potential pnmer region that could assist m quick tdentitication ofputative sequences of interest on performmg DNA sequencing 4. The distance between primer site and putative polyadenylatton site should not generally exceed 500 bp Special modificattons may be requtred for the amp&ication of the larger size of the differentially expressedcDNA

5. Other usual constraints (e g , nonpalindromtc

sequence) that are normal for

primer selection should also be apphed

Frg. 2 shows the example of the primer selection for the plasttd-localtzed low molecular HSPs in wheat that we have used m our laboratory. Fn-st all the available HSPs from this type of gene famtltes were collected and an ammo

acid sequence RKVID (highlighted) was selected that 1spresent about 9 ammo acids upstream of the translation stop codon. Also the 4 ammo acids followmg RKVID sequence (I.e., VQVQ) were also highly conserved at DNA and ammo

3’ Noncoding

Regions

115

acid sequence levels among all plastid-localized HSPs of higher plants and could be used as a signature of this gene family. This ammo acid sequence, RKVID also fulfilled all the conditions listed above and a primer sequence of 15 bp namely, SCGC AAG GTC ATC GAC 3’ (GC 60%) was synthesized m the Core Facility of the Institute for Biotechnology, Texas Tech University. Similarly a primer was designed that was a 15-mer of S’ACC GTC ACC GTG CCC 3’ sequence (GC content 73%) for targeting the HSP16.9 family of wheat plants (5) and a 14-mer degenerate primer 5’ GGN CCN AAR ATT GA 3’ was used for cytoplasmtc HSP70 family from wheat (6) It must be mentioned that only one gene family-specific primer in combmation with appropriate Tr*VN primer should be used in each radtoacttve PCR and reamplification reactions as described below. 3.6. Radioactive PCR with HSP Gene-Family Specific Primers Use of radioactivity and PCR in this step assiststhe easy detection of amplified product that could be present in very low concentration in the original RNA samples. Primer concentration of HSP gene-family specific primer is doubled to ensure the enough amount of primer 1savailable and the annealing temperature m PCR is reduced to 37°C instead of 40°C as used in the original method of differential display (81. It also must be remembered that HSP mRNAs are abundantly expressed m response to heat shock and differential expression of cDNAs expressing HSPs is almost guaranteed. Thaw out components listed below and set them on ice. Set up the PCR reactions at room temperature asfollows: Use 20 pL of final volume for each gene family-specific primer and anchored ohgo-dT primer combination (see Note 4) dHaO 10X PCR buffer dNTP (25 p~I4) Gene-family specific primer (4 CUM) T12w (10 $0 (Must be the same T,,VN used for RT) RT-mtx from Step I 35S-dATP (1200 Ci/mole) AmpliTaq (Perkm Elmer)

92& 20/J16& 2.0 pL 20/-J-

Total 1. MIX well by pipeting up and down. 2 Add 25 pL mineral 011. 3. Program the PCR machme as follows: 94°C for 30 s, 37’C for 2 mm, 72°C for 30 s (repeat these cycles 40 times), one cycle at 72’C for 5 mm; and then store at 4°C until use. 4 Use 3 5 $ of samples plus 2 pL of loadmg dye

Joshi and Nguyen

116

5 Incubate at 80°C for 2 mm tmmedrately before loading onto a 6% DNA sequencing gel 6 Electrophorese for about 3 h at 1500 constant voltage until the xylene cyan01 dye is 5-7 cm from the bottom 7 Expose gel to X-ray film after drying. 8. The DNA sequencmg gel should be dried without fixmg with methanol/acettc acid 9. Orient the autoradiogram and dried gel with radroacttve mk or fluorescent markers.


of cDNA Probe

1 After developmg the film (overmght exposure), orient the autoradiogram with the gel by aligning the radloacttve mk marks or fluorescent markers spots and staple them together Locate bands of Interest by markmg wtth a clean pencil underneath the film and by cuttmg through the film with a razor blade Handle the dried gel with gloves and save It between two sheets of clean paper Cut out the located band with a clean razor blade every time and pick tt up wtth a clean forceps 2 Soak the gel slice along with the Whatman 3M paper m 100 & dHzO for 10 mm 3 Boil the tube with tightly closed cap for 15 mm 4 Spm 2 mm at highest speed m a mlcrocentrlfuge to collect condensation and pellet the gel and paper debris 5 Transfer the supernatant to a new mtcrofuge tube Add m 10 pL of 3M NaOAC, 5 pL of glycogen (10 mg/mL), and 450 & of 100% ethanol Let the mixture sit for 30 mm m a -80°C freezer 6. Spin for 10 mm at 4°C to pellet DNA A small but dtstmct white pellet should be visible at this stage Remove supernatant and rinse the pellet with 200 pL tcecold 85% ethanol Spin briefly and remove the residual ethanol Air-dry the pellet for 5-10 mm. DNA will be lost if less concentrated ethanol is used 7 Dissolve the pellet m 10 pL of dHzO and use 4 pL for reamphfication. 8 Reamphficatlon should be done using the same primer set and PCR conditions except the dNTP concentrations are at 20 w (use 250 fl dNTP stock) instead of 24 wand no tsotopes added 40 pL reactton IS recommended. Use 40 pL final Volume for each primer set combmatton dHzO 204$ 1OX PCR buffer 4.0 pL dNTP (250 @4) 3 2 l.lL Gene family specific primer (4 l&Q 40& T,zm (10 W 40$ cDNA template from step 7 4.0 /AL AmphTaq (Perkm-Elmer) 04@Total 40.0 pL 9 Check to see if the size of your reamphfied PCR products 1s consistent wrth its size on the DNA sequencing gel (see Note 5). 10 Verify the probe by Northern Blot analysis as descrtbed m Josh1 et al (6,22)


117

11 Clone the cDNA probe using the pCR-TRAPTM clonmg system from GenHunter Corporation (cat. no P404) 12. Verify the cloned cDNA probe by Northern Blot and sequence the cloned cDNA 13, Clone the full length cDNA by screening a cDNA library followmg the standard procedure (see Note 6).

3.8. Cloning

of cDNA Probe

We have found that cloning of the small cDNA fragments with pCR Trap vector (GenHunter Corporation) IS very simple to do and success rate IS remarkably high. Use: pCRTrap no. P404 kit (GenHunter). 1 Set up Ligation reaction as follows: In a OS-mL tube add the following m the same order for a 20-pL ligation reaction. dHzO 10.0 pL Insert-ready pCR-TRAP vector 2ocLL 1OX ligation buffer 2.0 pL PCR product from step 8 (Section 3.7.) 50@T4 DNA hgase (200 II/&) 1.0 pL

2.

3 4 5 6. 7. 8. 9.

Total 20.0 pL Ligate the vector and PCR product for overnight at 16°C. For transformation, thaw competent cells on ice for 10 min Aliquot 100 pL of competent cell to 1.5-mL microfuge tubes and set on ice. Immediately freeze at -8O’C the unused tubes with competent cells for future transformatton. Add 10 pL of ligation mtxture to the tube containing the competent cell and mcubate on ice for 45 min with intermittent finger ttpping Heat shock the cells for 2 mm at 42’C and then set the tube back on ice Add 0.4 mL of LB medium and incubate the cell at 37°C for 1 or 2 h Plate 200 pL of cells on each LB plate containing tetracycline (20 pg/mL) Wait until the plate surface is dry, and incubate the plate upside down overnight at 37°C. Score the TetR colonies and save the plate at 4OC before further analysis For checking the insert size, use the colony lysis PCR method. With a marker pen, number each TetR colony (4-8 colonies per clone) to be analyzed on the plate. Ahquot 50 pL of colony lysis buffer into each numbered microfuge tube.

10. 11. Pick eachcolony with a cleanpipet tip, andtransfer the cells into the colony lysls buffer m the corresponding numbered tube. Take very little bacteria on the tip avoiding touching any other colony. Mix by careful pipeting. Spin at high speed for 2 mm and vortex on low setting for 2 s. 12. Incubate the tube at 95°C for 10 mm. 13 Spm the tube for 2 mm to pellet the cell debris 14 Transfer the supernatant into a clean tube. 15. Use it directly for PCR analysis or store at -20°C for future amplification (can be stored for at least 6-10 mo).

118

Josh/ and Nguyen

16 Perform PCR Reaction as follows For each colony lysate. dH,O 10 X PCR buffer dNTP (250 luz’I Lgh primer (2 ClM) Rgh primer (2 pA4) Colony lysate AmphTaq (Perkm Elmer)

17

18.

19. 20

(see Note 7) 102 l.lL 2occL [email protected]

2.0 pL 20@-

2.0 pL

Total 20 0 pL Mix well, add 30 p.L mineral oil PCR parameters used are as follows 94’C for 30 s, 52°C for 40 s, 72°C for 60 s (repeat these cycles 30 times), one cycle at 72°C for 5 mm; and then store at 4°C until use. Analyze 15-20 pL of PCR product on a 1 5% agarose gel with ethidmm bromide stammg The plasmid with insert should easily produce visible bands that are bigger than 100 bp Verify the insert size by comparmg the molecular weight of the PCR product before and after cloning The PCR product after colony-PCR should be 100 bp larger than the original PCR insert before cloning owing to the flanking vector sequence being amphfied. Purify each band from agarose gel and use it as a probe for Northern blot hybridization or screening a library. To avoid false positives perform southern blotting using a known probe from the members of the gene family as described m Josh1 et al. (6) (see Note 8).

3.9. Sequencing

the Cloned PCR Products

1 After a plasmid has been checked to contain an appropriate sized insert, the corresponding TetR colony should be restreaked to single colonies on a new LB-Tet plate This can be done by locating on the plate the colony marked with number and pickmg it carefully with a clean pipet tip Streak the cells on a new LB-Tet plate. Change to another clean tip and streak again on the new plate m order to obtain smgle colonies 2. Incubate the plate overnight at 37°C 3 Inoculate a single TetR colony mto a 5-mL LB culture and use 3 mL for plasmid mimprep The remammg can be saved as glycerol (50%) stock at -80°C. 4. The mmipreped plasmid can be sequenced using the usual procedure with Sequenase Version 2.0 kit (USB) according to manufacturers suggestions The sequencing primers Lseq (ATCACGAGGCCCTTTCG) and Rseq (GATACTGGACGCGAGCC) (GenHunter) are to be used for this process

3.10. Examples of Use Using this modified differential display technique for the identlficatton of several members of a specific HSP gene family, we have isolated eight mdivldual members of wheat HSP 16.9 gene family that has about 12 members (5).

3’ Noncoding Regions T12VA T12"CT12"G 123456789

119

TIZW

Fig. 3. Differential display pattern of the total RNA from control (lanes 1,3,5, and 7) and heat shocked (lanes 2,4, 6, and 8) wheat seedlings using a HSP 16.9 genefamily-specific 5’ primer (ACCGTCACCGTGCCC) and four 3’ anchored oligo dT primers namely, T,,VA, T,,VC, T,,VG, and T,*VT. Note the specific amplification of radioactive cDNA fragments only in the heat shocked samples. Lane 9 was a positive control where plasmid DNA from HSP16.9a was used under the identical conditions as described above. An amplified cDNA fragment of 270 bp is shown by an arrow. Reprinted with permission from Kluwer Academic Publishers.

Figure 3 shows one example of using the modified differential display technique for the study of differentially expressed cDNAs belonging to wheat HSP16.9 gene family. There is almost no amplification of cDNAs in control reaction indicating the successof using this approach in HSP research. We have also identified a gene-specific probe belonging to plastid-localized HSP26.6 gene family that is differentially expressed during heat shock between thermotolerant and

120

Joshi and Nguyen

therrnosusceptrble wheat genotypes. We have fkrther proved that this member 1s also genetically linked to heat tolerance m wheat by doing the cosegregation analysts (7). Using a degenerate primer for the HSP70 gene family, we have tdentified three putative members of a cytoplasmtc HSP70 multigene family from wheat (6).

4. Notes 1. Warning: Guamdme hydrochloride is toxic if inhaled; avoid breathing dust Potential health hazard. 2. The success of the Differential Display technique depends on the integrity of RNA. It must be free of chromosomal DNA contamination, Total RNA is preferred over poly (A) RNA due to the cleaner background signal, and easy purification and integrity verification 3. It is recommended that core mix without RNA template be made for each anchored ohgo-dT primer if several RNA samples are to be compared. This will mmimize pipetmg errors and keep the comparison as close as possible A control reaction using all other components of reverse transcription without addition of Reverse transcriptase for all the reactions is mandatory to confirm the reversetranscription dependent amplification of mRNAs 4. Make as much of core mixes as possible to avoid pipeting errors. Make 10 times the volume of the PCR core mix (without oligodT primer, gene family-specific primer and RT samples; add these reagents separately in each tube) for each of the pair of control and heat shock treatments (2 x 4 reactions per gene familyspecific primer and minimum one negative control reaction). Otherwise, it would be difficult to pipet 0.2 pL of AmpliTuq m each reaction. 5. 20 pL of PCR samples are run on a 1.5% agarose gel and stained with Ethidmm bromide. About 95% of the probes should be visible on the agarose gel after this round of PCR Save the remainmg PCR samples (20 pL,) at -20°C for clonmg. 6 Depending on the primer design the number of false positives varies For example, HSP26.6 and HSP70 primer design gave us more false positives than HSP16.9. Rarely, RAPDs with same gene family-specific primer on both sides of the cDNA resulted. One band could have more than one sequence and same sequence may be represented at different sites but with different polyadenylation site selection, the feature that is very common m plants (I). Sometimes, 5’ and 3’ ends of the cDNA sequence had appropriate primer sequences but the in between sequence did not match with expected gene family sequence Rarely, Northern blots did not show differenttal expression of selected cDNA bands. We believe that differential display is extremely sensitive technique that brings out the minute differences existmg m the mRNA population by magnifying the difference and northern blots detect only gross quantitative changes Other sensitive means of detection should be employed to differentiate between a real false positive and a putative positive signal. 7 It is recommended that core mix containing everything except colony lysate be made so that many colony lysates can be analyzed at one time 8. Sometimes tetracyclin resistant colonies have a smaller than expected insert incorporation because PCR product purification is not done before cloning. But

3’ Noncodmg Regions

121

band purlficatlon usmg Qmaqmck column (Qutagen, Chatsworth, CA) sign% cantly affects transformation efficiency

Acknowledgments We wish to thank Jeff D. Ray for insplrmg us to use dlfferentlal display for our heat shock protem research. Our colleagues Natalya Klueva, S Kumar, and Rama Joshi are profusely thanked for the discussions, encouragement and assistance. The financral support for this work was partially

provided by a grant

from US Department of Agriculture National research Initiative Grants Program (92-37 100-7903). References 1 Josht, C P (1987) Putative polyadenylation signals m nuclear genes of higher plants: compilatton and analysis Nucleic Aczds Res 15, 9627-9640 2 Nguyen, H. T. and Joshl, C P. (1994) Molecular genetic approaches to improving heat and drought stress tolerance m crop plants, m Biochemical and Cellular Mechanzsms of Stress Tolerance zn Plants (Cherry, J. H , ed ), Springer Verlag, Berlm, pp. 279-289 3 Josht, C. P. and Nguyen, H. T. (1995) 5’ Untranslated leader sequences from eukaryotic mRNAs encoding heat shock proteins. Nucleic Acids Res 23,541-549 4 Vterlmg, E. (1990) The roles of heat shock proteins m plants. Annu Rev PI Physiol and Pl Mol Biol 42, 579420. 5. Joshi, C P and Nguyen, H. T (1996) Differential display mediated rapid tdentificatton of different members of a multigene family, HSP16 9 m wheat Plant Mol Bzol. 31,575-584 6. Joshl, C P , Kumar, S , and Nguyen, H T (1996a) Application of modified differential display technique for the clonmg and sequencing of the 3’ regions from three putative members of wheat HSP70 gene family Plant Mol Blol 30, 641-646. 7. Joshl, C. P., Klueva, N, Morrow, K. J , and Nguyen, H T (1996b) Expression of a umque plastid-localtzed heat shock protem m genetically linked to acqmred thermotolerance m wheat, m preparation. 8. Liang, P. and Pardee, A. B. (1992) Differential display of eukaryotic mRNA by means of polymerase chain reaction. Sczence 25’7,967-97 1. 9. Ltang, P, Averboukh, L., and Pardee, A. B (1993) Dtstributton and cloning of eukaryotic mRNAs by means of differential display: refinements and optimization. Nucleic Acids Res 21,3269-3275. 10. Logemann, J., Schell, J , and Willmitzer, L. (1987) Improved method for the tsolation of RNA from plant tissues. Anal Blochem 163, 16-20 11 Williams, J. G. K , Kubehk, A R , Llvak, K J., Raflaski, J A , and Tmgy, S V (1990) DNA polymorphtsms amplified by arbitrary primers are useful as genetic markers Nucleic Actds Res 18, 1585-1588 12. Josht, C. P., King, S. W., and Nguyen, H. T. (1992) Molecular cloning of cDNA encoding water stress protem (WSP23) from wheat roots Plant SCL 86,7 l-82.

RC4D-Restriction Fragment Length Polymorphism-Coupled Domain-Directed Differential Display Giinter TheiBen and Achim Fischer 1. Introduction Since many life processes, such as development and reproduction, depend on differential gene expression, the spatiotemporal pattern of gene expression is of prime biological interest Accordmgly, the concepts of the mRNA differential display (DD) and RNA fingerprintmg techniques (1,2) promised a breakthrough in molecular biology, because they circumvented some severe limitations of traditional methods, such as differential screening and subtractive hybridrzatton, during the analysts of differential gene expression. However, it is not always necessary to detect the differential expression of potentially every gene in the genome, as is intended by RNA fingerprmtmg and DD. In some cases,interests may focus on the expression of genes sharing a certain conserved sequence motif, as is often the case with members of multigene families. To allow efficient expression analysis of such genes, a technique called “RFLP-coupled domain-directed differential display” (RC4D) was developed (3). To target amplification towards all members of a gene family, RC4D uses primers directed against a conserved region, henceforth called the family specific domain (FSD). In contrast to RNA fingerprintmg and DD, in the course of an RC4D experiment generation of amplicons of gene specific size does not depend on bmdmg of short arbitrary primers, but is accomplished by mtroducing an RFLP step. Therefore, long primers can be employed, avoiding deleterious effects to reproducibility due to the bmdmg behavior of very short primers. A second advantage of employing long PCR primers is the possibility to directly sequence reamplified bands followmg the cycle sequencing protocol, From

Methods m Molecular Biology, Vof 85 Dlfferentraf Edited by P Llang and A 6 Pardee Humana

123


124

The/pen and Fischer

using one of the PCR primers as sequencing prrmer, whrch obviates the need of a clonmg step. In contrast to differential display, RC4D turned out to be hrghly reproducrble, whrch makes repetrtron of experiments drspensable. Another important feature of RC4D 1sIts semrquantrtatrve character: as opposed to DD, the difference m mtensrty of a particular band between two lanes on the autoradrograph IS strongly correlated wrth the srgnal strengths seen on Northem blots (3). The working strategy of RC4D can be seen m Frg. 1 Selectrvrty for a gene family of choice IS provrded by amphfymg cDNA pools from the tissues to be compared with PCR primers directed against the respective conserved family specific domain (FSD) and against the non-T-extension of the cDNA prrmer. For the mtroduction of RFLP, the mixture of amphfied family member cDNAs is then Incubated wrth a surtable frequent cutting restrrctton enzyme. To restore a downstream primer bmdmg site for reamplrfrcatron, double-stranded linker molecules are ligated to the truncated amplrcons To get rid of nonltgated fragments and fragments not carrying a FSD, a second round of amplrficatton 1sperformed. After labelmg, the preparations are run on a sequencmg gel for srze separation. Differential bands are cut out, reamplrfied, and analyzed further. We have applied RC4D to the analysis of MADS-box gene expresslon m maize (3), but the procedure should be also quote surtable m other cases, Including expressron analyses of MADS-box genes m other species (for review, see ref 4) and of other multrgene families n-rall sorts of orgamsms, e.g., the famrlres of homeobox, steroid receptor (5), and heat-shock factor or ETS domain genes (6) in plants and ammals. Possrble applrcatrons of RC4D Include mutant analysis, whenever a certam gene family 1ssuspected to be mvolved m the observed mutant phenotype, as well as the lsolatron of new members of a known gene family Aberrant splrce products can also be detected by RC4D (our unpublished results). Another attractrve apphcatron of RC4D m developmental biology would be to record the expression pattern of a family of regulatory genes (e.g., transcrrption factors) m time. Doing so could unvetl a temporal network of gene actrvrtres, whrch might be useful to define certain stages of development more precisely by molecular events than 1s possrble by only observmg morphologrcal alterations. 2. Materials 2. I. cDNA Preparation 1 10X reverse transcrrptasebuffer (7) 500 mA4 Tns-HCl, pH 8.15 at 42”C, 60 mA4 MgClz, 400 mA4 KCI, 10 mA4 DTT, each dNTP at 10 mA4 2. RNasm RNase mhlbltor (Promega, Madison, WI, 40 U/&) 3 RQl RNase free DNase (Promega; 1 U/&)

125

Domain- Directed Differential Display lIlRNA q 1st strand cDNA

KKIAAA Reversetranscription

cDNA pool

*+ l +

TTTTTT***** TTTTTT*+*** TTTTTT*****

FSD

PCR with adapter primer (* * * * *) and FSD primer (W) 4 AAAAAA*****

FSD Mixture of amplified family member cDNAs

1. Digestion with frequent cutter (arrows) 2. Ligation to double strand4 linkers 3. PCR with linker primer and nested FSD primer Link=

1

FSD Truncated family member cDNAs

1

Linear PCR with labeled nested FSDprimcr ( l )

Labeled fragments

d Autoradiograph

Electrophoresis

1. Cutting out differential bands 2. Reamplification. direct sequencing and tiuther analysis

Fig. 1. Flowchart representing the working strategy of RC4D. With permission from ref. 3. 4. 5. 6. 7. 8.

Phenol, saturated with TE (TE buffer is 10 mMTris-HCl, ChloroformIisoamylalcohol(24: 1). 3M sodium acetate, pH 5.2. Ethanol (absolute; E. Merck, Darmstadt, Germany). 1 mMEDTA, pH 7.5.

pH 8.0,l mMEDTA).

The&m and Fischer

126 9 10 11 12 13

Ohgo(Dynabeads plus set of buffers (Dynal, Oslo, Norway). Dtethylpyrocarbonate treated water (DEPC-water) Reverse Transcrlptase M-MuLV (Boehrmger Mannherm, Mannhelm, Germany) Hybrid dT,,-adapter primer (7) S-GAC TCG AGT CGA CAT CGA T,,-3’ TE buffer: 10 mM Trts-HCl, pH 8 0, 1 n&’ EDTA

2.2, The RC4D Protocol 1 Lmker ohgonucleotides* synthesized on an Apphed Biosystems (Foster City, CA) 392 DNA/RNA synthestzer, deprotected overmght at 50°C and purified on NAP columns (Pharmacia, Uppsala, Sweden). BLl: 5’-ACT CGA TTC TCA ACC CGA AAG TAT AGA TCC CA-3’ BL2. S-TGG GAT CTA TAC TTT CAA-3’. 2 2X Trts/magnesium 100 mM Tris-HCl, pH 7 5,20 mM MgCl, 3 10X PCR buffer 670 mMTris-HCl, pH 8.8, 170 mM(NH4)S04, 1% Tween-20, 8 mkf MgCl* 4 1 mM dNTPs: Solutton of dATP, dCTP, dGTP, and dTTP each at 1 mA4 5 Outer FSD prtmer (or mixture of FSD primers), biotinylated at tts 5’-end 6. Paraffin oil. 7 ?‘aq DNA polymerase (Boehrmger Mannheim) 8 Adapter primer (7). Y-GAC TCG AGT CGA CAT CG-3’ 9 Streptavtdm coated magnetic beads plus magnetic separator for mlcrofuge tubes (Dynal) 10. Frequent cuttmg restriction enzyme: e.g , MseI (New England Biolabs, Beverly, MA), Hi&I (Boehrmger Mannheim). 11 Phenol/chloroform/tsoamyl alcohol (25 24.1) 12 Ethanol (absolute, E Merck). 13 TE buffer 10 mM Tris-HCl, pH 8 0, 1 mA4 EDTA 14 10X end-filhng buffer 500 mMTris-HCl, pH 8.0,60 mMMgCl,, 500 mMNaC1, 100 mA4 dithioerythntol, 400 pg/mL BSA. 15 T4 DNA polymerase (Boehrmger Mannheim) 16 T4 DNA Ligase (Boehrmger Mannheim). 17. 10X Ltgatton buffer I 250 mJ4 Trts-HCl, pH 7.5, 50 mA4 MgC12, 25% (w/v) polyethylene glycol8000,5 mM DTT, 4 mA4 ATP (8). 18 QiaQuick spm columns, accompamed by PB buffer (Qtagen, Hilden, Germany) 19. Inner (nested) FSD primer (or mixture of primers) 20 T4 Polynucleottde kinase plus 1OX PNK buffer (Promega) 2 1 y[32P]dATP, 5000 Wmmol (Amersham, Braunschwetg, Germany) 22. Stop solutton (self-made [ref. 91 or from Promega’s fmol DNA sequencmg system) 23. Length standard; preferred range: approx 50 bp to > 500 bp (32P-labeled) 24 6% sequencing gel: 04 urea, acrylamide. btsacrylamide, 19.1. 25 1% acetic actd 26. X-ray film (e.g , Kodak X-Omat).

Domaln-Directed 2.3. Analysis

Differential D/splay

727

of RC4D-Products

1 Cycle sequencing mgredrents, e.g., thefmol DNA sequencing system (Promega). 2 10X PCR buffer. 670 mA4Trrs-HCl, pH 8.8, 170 mM(NH4)S04, 1% Tween-20, 8 mA4 M&l, 3. 1OX nucleotides for labeling (dATP, dGTP, and dTTP at 20 mM each) 4. Taq DNA polymerase (Boehrmger Mannhelm). 5. o[32P]dCTP, 3000 Wmmol (Amersham) 6 NICK columns (Pharmacta)

3. Methods 3.1. cDNA Preparation

(see Note 1)

Prepare total RNA from the tissues to be analyzed according to the guamdmmm chlorrde tsolatron method (9) Other methods may work equally well Determme the amount of isolated RNA spectrophotometrtcally. Check the integrity of RNA by hybrtdtzmg a Northern blot with a suitable probe hke, e g , GAPDH or Actm Dissolve 50 pg of ethanol precipitated RNA at 0.5 pg/pL m 1X reverse transcrtptase buffer For removal of residual genomrc DNA, add 5 pL RNase free DNase and 1 pL RNasm and incubate 15 mm at 37°C Extract samples once wtth TE-saturated phenol and once with chloroform/rsoamyl alcohol. Spin l-2 mm at 20,OOOg and transfer aqueous layers into fresh mtcrofuge tubes Add l/10 vol of 3M sodium acetate, pH 5 2, and 2 5 vol absolute ethanol, let stand overnight on ice For recovery, centrrfuge 30 mm at O°C and 20,OOOg Resuspend m 100 p.L 1 mMEDTA,pH75 Prepare mRNA by use of ohgo(Dynabeads (Dynal, Norway), accordmg to the manufacturer’s recommendatrons. For 50 pg of total RNA, use 200 pL of bead suspension (1 mg beads) After purrficatton, elutron of mRNA bound to the beads is done m 5 p.L Dynal elutron buffer Dilute 5 uT., of purified mRNA with 12 25 uL drethylpyrocarbonate (DEPC) treated water Secondary structures are resolved by denaturing for 3 mm at 65°C. Chill denatured mRNA on ice and add rt to 4.75 $ of an ice-cold reverse transcrtptase mixture contammg 2.2 pL 10X reverse transcrtptase buffer, 10 U RNasm, 12 U reverse transcrrptase (see Note 2) and 35 pmol of hybrid dT,7adapter primer MIX carefully and incubate the reactions at 42°C for 1 h, followed by 30 min incubatron at 52°C. Dilute cDNA preparations to 1 mL wrth TE buffer and store as “cDNA pools” at 4°C or at -20°C

3.2. The RC4D Protocol 1. For preparation of blunt ended linker molecules, combme 1 25 nmol of ohgonucleotrdes BLl and BL2 with 40 uL 2X Trrs/magnesmm in a total volume of 80 JJL and heat to 90°C for 3 min Cool down slowly to 4°C and store double stranded linkers at -20°C 2. For each cDNA preparatton, set up a linear amphficatton reaction using either a single primer bmdmg to the family specific domain (FSD), or employing a mix-

128

3

4.

5.

6 7.

8

Theipen and Ftscher ture of primers bmdmg to the same regton withm the FSD (see Note 3) Mix 5 pL 10X PCR buffer, 1 pL 1 mM dNTPs, 1 5 pmol of outer FSD primer and 1 p.L “cDNA pool” in a total volume of 50 & Overlay reactions with paraffin oil, put tubes mto a thermoblock preheated to 95”C, and add 2.5 U Taq DNA polymerase (see Note 4). After initial denaturation 3 mm at 95°C run 30 cycles with the followmg parameters’ 30 s denaturation at 95”C, 100 s annealing at a temperature suitable for the employed primers, 2 mm extensron at 72’C. For the second, exponential amphfication phase, cool down reactions to 65°C and add 50 pL of exponential reaction mix prewarmed to 65°C Exponential reaction mix is prepared as above, but contains 5 pL 1 mM dNTPs, 2 5 U of Taq DNA polymerase, and 15 pmol of FSD primer(s) and the adapter primer each per 50 l.L Cycling is done with the same parameters as during the first phase of linear amplification (see Note 5) Try different cycle numbers and run reactions on an agarose gel, no more cycles than necessary for obtammg a visible signal should be run In most cases, the cycle number will be between 20 and 30 After amplification, purify PCR products using streptavidm coated beads as recommended by the manufacturer Digest one fifth of purified PCR product with 15 U of a suitable frequent cutting restriction enzyme for 2-3 h in a volume of 50 pL (see Notes 6 and 7). Destroy the biotin-streptavidm mteraction by heating the reactions for 30 mm to 65°C rn 50% (v/v) phenol, as recommended by Dynal. Remove beads Precipitate restriction fragments with ethanol, recover DNA by spinning 30 min at 2O,OOOg, and dissolve pellet m 5 I.& of TE buffer Combme 0.5 pL of resuspended restriction fragments with 1 pL 10X end-fillmg buffer, 1 pL 1 mMdNTPs and 0.5 U T4 DNA polymerase m a total volume of 10 pL. After incubation for 1 h at 37”C, precipitate end-filled restriction fragments with ethanol (see Note 7a). For linker ligation, resuspend end-filled fragments m 4 pL of a mixture contammg 0.5 pL linker DNA (see Note 8) from step 1 (approx 50-fold molar excess), 0.4 pL 10X ligation buffer I and 0 4 U DNA hgase. For resuspension, let stand on ice for 1 h. Then ligate overmght at 16°C. After ligation, mix with 50 pL PB buffer and purify on QiaQmck spur columns. Elute m a volume of 50 pL TE Using 5 pL of purified ligations, the linker primer and a second inner (“nested”) FSD primer (or set of primers), set up a reamplification under the conditions described in step 2. Reduce the extension time to 1 mm and run 15-20 cycles Prepare endlabeled nested FSD primer by mixing 10 pmol ohgonucleotide, 1 pL 10X PNK-buffer, 10 pmol [y-32P]dATP (see Note 9) and 5 U polynucleotrde kmase in a total volume of 10 pL and incubate 30 mm at 37°C. To remove enzymatic activity, stop labeling reaction by heatmg to 90°C for 2 mm Combine 0.5 pL of reamplification reactions, 0 5 pL 1 mM dNTPs, 1 pL 10X PCR buffer, 1 U Taq DNA polymerase and 0.4 JL of labeled FSD primer m a volume of 10 pL Overlay with paraffin oil, put mto a thermoblock preheated to 94°C and run a primer extension reaction comprismg 2 mm mitral denaturation at 94°C then 30 cycles of 30 s denaturation at 95”C, 30 s annealing, and 1 mm extension at 72°C. After cycling, add 5 pL of stop solution.

Domain-Directed

Different/al D/splay

729

9 Prepare a 32P-labeled length standard (see Note 10) After labeling, add 0 5 vol stop solution. 10 Prepare a standard 6% denaturing sequencmg gel 11 For denaturatton, heat primer extension reactions (from step 8) and marker (from step 9) by heating to 70°C for 2 mm. Cool down on ice and load 2 pL of each reaction on the sequencing gel. A short and a long run IS recommended to efficiently resolve fragments from below 100 bases up to 1 kb 12 After removal of the upper glass plate, immerse the gel m 1% acetic acid Agitate for 10 mm, then repeat two times with fresh acetic acid solution An-dry the gel (see Note 11) and expose overnight to autoradiography film (see Note 12).

3.3. Recovery

of RC4D-Products

1. Label bands of interest on the autoradiograph Punch small (0 2 mm) holes at both ends of each band mto the film, using a sterile hollow punch (see Note 13) Superimpose the autoradiograph to the dried gel, fix it with adhesive tape and mark the positions of the holes on the gel using a thm felt-tipped pen Remove the autoradiograph 2 Lay the gel on a bright support (e g , a white sheet of paper) and, using a sterile scalpel, cut out the labeled bands from the gel (see Notes 14 and 15) To minimize carryover of neighboring bands and background smear, the cut out regions should not exceed the bands’ sizes. Spread 1 pL TE on the surface of a band, wait until the band has become rehydrated (approx 1 mm; see Note 16), pick up gel slice with the scalpel and transfer it into a microfuge tube containing 50 pL TE (see Note 17). 3 Elute DNA either overnight at 37’C or at 80°C for 2 h (see Note 18). 4 Reamphfy one tenth of the eluate of each band in 50-h PCR reactions as described m Section 3.2 , step 6, applying 30-35 cycles. Check reactions on an agarose gel

3.4. Analysis of RC4D-Products 1. To check homogeneity of the reamphficatlon products (see Note 19), use 0.5 pL of the PCR reaction (Sectton 3 3., step 4) as template for a primer extension reactron as described above (Section 3.2 , step 8). Run reamphficattons side by side to their original RC4D reactions on a sequencing gel. When purity is high enough, contmue as described below If strong contammattons should remam, run another cycle of cutting out and reamplrfymg the band from the second sequencmg gel 2. Determine the sequence of the recovered bands following the cycle sequencmg protocol. If long enough to allow annealmg temperatures of 6&7O”C, both PCR primers can be employed for sequencing Otherwise, the long lmker primer should be used preferentrally (see Note 20). 3 To confirm differential expression, prepare hybridization probes suitable for probmg Northern blots (see Note 21) Set up IO-& labeling reactions contammg 10-50 ng reamphfication product, 10 pmol linker primer, 1 pL 10X PCR buffer,

Theipen and Fischer

130

1 pL 10X nucleotldes for labeling, 2.5 U Tuq DNA polymerase, and 5 pL [a32P]dCTP. After addmg a drop of paraffin 011,run a cyclmg program as descrtbed above for prtmer extennons, but reducing extenston time to 30 s Remove umncorporated radtonucleottdes using NICK columns Prehybrtdtze and hybridtze blots as usual.

4. Notes 1 For all RNA work, the usual precautions should be taken. Use RNase free reagents and plasttcware and wear clean gloves at all stages of the procedure 2 To confirm successful DNase digest m Section 3 l., step 2, perform also a control reaction wtthout addition of reverse transcrtptase 3 Great care should be bestowed on FSD primer destgn First, compile all relevant known ammo acid sequences of the chosen type of domain Some computer programs (e g , PILEUP of the GCG package [Genetics Computer Group, Madison, WI]) are excellent for this purpose Determine the regions of maximum conservatton, comprismg at least SIX ammo acids. For bmdmg of outer and inner (nested) primers two such regions or one region of at least seven ammo acids are required If necessary owing to the lack of overall conservation of sequence motifs, subdtvide the gene family mto several subfamilies according to the variations of the identified motifs In this latter case, primers are designed for each of the subfamtlies The next step then is to backtranslate the identified ammo acid motifs into nucleotide sequences. For thts step tt 1squote useful to know the codon usage of the investigated orgamsm. Codon usages of many specres are available online at http.//www dna.affrc go.Jp/-nakamura/CUTG.html The degeneracy, necessary to cover all possible ammo acids at a given position as well as to consider the different codons coding for a single ammo acid, should be kept as low as possible. Thts holds especially true for the last 2-3 codons towards the primer’s 3’-end In many cases tt will be advisable to define the 3’-most base of a primer by the first or second base of a codon, as they tend to be more conserved than the codons’ thud bases. Incorporatton of inosme may be considered at positions with three or four possible bases with similar chances to appear When outer and inner (nested) primers overlap, allow at least 3 addtttonal bases at the inner (nested) prtmer’s end. The optimal length of primers 1s 18-24 bases 4 To allow equal dispersal of Taq polymerase when doing a “hot start”, dilute the enzyme to 5 pL with 1X PCR buffer After addition to the reaction preheated to 9YC, mix gently by ptpeting 5 It may pay off to determme optimal annealing temperatures and magnesium concentrations for each primer pair separately. In general, high annealing temperatures close to the primers melting temperature and low magnesium concentrations (e.g., 0.8-l mA4) are advantageous for obtaining maximum spectfity of primer bmdmg Besides, avotd overamphfrcatton (which may lead to over- or underrepresentation of certain amplification products owing to “plateaumg”) by testing different cycle numbers. We routinely apply so many cycles as are necessary for obtaining a visible signal when runnmg 10 @., of a PCR on an agarose

Domain- Directed Differential Display

6

7.

7a.

8

9

10

11

131

gel. Besides, make sure that the primers of each primer pair are compatible (no drmer formatton, similar G/C-contents, melting temperatures not devtatmg more than 24°C from each other) If owing to a short (e.g , 18 bp) nested FSD pnmer a short linker primer has to be chosen for reamplification of ligation products, the linker primer should be directed against the Imker’s outermost part. Domg so maintains full linker sequence mformation and allows to use a longer, high melting prrmer for sequencing (see Note 20). Posstbly owmg to imperfect strand reannealing after the last denaturatton step, digestron of PCR products m many cases 1s less efticrent compared to “natural” substrates Because partial digests would result in more than one band per transcrrpt, they should be strrctly avorded. Therefore, quite a high amount of restrrction enzyme and prolonged mcubatton times should be applied. Choosmg (a) suitable restrtctton enzyme(s) 1s crucial Some enzymes show star actlvtty even under opttmal mcubatton conditions (e.g , Sau3aI), others (dependmg on codon usage and G/C content) cut too often or not often enough. In our experiments with maize cDNAs, Hz&I, MboI, and MeI yielded fragments m the range of several hundred base pans, as desired, whereas Ah1 and RsaI sites were found to be much too abundant. An endtillmg step should be performed even when a restrtctron endonuclease has been chosen that generates blunt ends. Followmg this suggestion ensures that fragments without an internal recognitron site for the respective enzyme will not be lost during reamplificatron, which otherwise could be the case due to a single terminal A residue added by Tugpolymerase interfering with successful ligation of blunt end linkers Generally, to ensure the linker molecules to be ligated m a defined ortentation, oniy one end should tit to the restrrctron fragments’ staggered or blunt ends, whereas the other end should be tailored to be incapable for hgatron by mtroducing a single stranded overhang or termmal mismatch several bases long Preparing one universal blunt ended linker m combmatton with end-fillmg of staggered restrrctlon fragment ends, as applred here, allows trying different restriction enzymes wrthout the need for one linker per enzyme. To allow annealing temperatures optrmal for sequencing (see Note 20), the lmkers should be at least 24 bp long. It has been shown that when using 35Sin PCR reactions, volatile 35Scompounds may diffuse out of reaction tubes and contaminate the thermocycler (20). Therefore, 32P or 33P are the preferred radioisotopes for radioactive PCRs For the size range up to 500 bp, the “Sequamark” IO-bp ladder (Research Genetics, Huntsville, AL) 1san excellent marker. Owing to its high resolution, even different gel runs can be easrly compared. Moreover, overexposrtron yields a faint 1 bp background ladder, which allows exact determination of fragment sizes To avoid damaging the DNA, we do not recommend heating the gels for drying. Instead, we prefer an-drying them m the air flow of a fume hood after carefully removing any droplets of fixing agent This is best done by briefly laying a sheet

Theipen and Fischer

132

12

13 14

15 16 17 18 19

20

21

of Whatman 3MM paper onto the gel and Immediately and gently peeling it off again Try first at one comer of the gel, because sometimes the paper tends to firmly stick to the gel surface Following these suggestions, a 0 3-mm gel will be dry after 1 h Localization of bands m the gel requires exact superimposition of autoradiograph and gel. For this purpose, we spot old (up to 6 mo) 35S-sequencmg reactions onto adhesive paper labels (1 & per label is sufficient) The labels are attached on pieces of adhesive film, which m turn are then applied to the dried polyacrylamide gel. After cutting out the desired bands, the adhesive film can be removed again, and the labels can be reused Punching holes in autoradiography films is easily done using a hollow punch, which is driven through the film by a hammer blow Keep m mmd that the gel IS radioactive and-especially when labelmg IS done with 32P--irradiates high doses of P-particles’ Note: While manipulating the dried gel, always cover all regions of the gel that are not currently processed with clean glass plates to provide an effective shield Note: When cuttmg the dried gel, take extreme care to avoid mhalation of radioactive gel splinters! Waiting about 1 mm after wetting a gel shce is optimal If this time is greatly exceeded, the slice is drymg again and becomes too sticky for easy removal from the supporting glass plate After excision of each band the scalpel blade should be briefly sterilized m a flame. If elution is done at elevated temperatures (25O”Q add some paraffin oil to avoid buffer evaporation It IS advisable to check homogeneity of reamphfication products. Minor contaminations ~111 not mterfere with subsequent steps, but sometimes strong bands of different size are copurifred with the desired product and prevent successful sequencing As the cycle sequencing protocol is somewhat susceptible to low (~65°C) primer annealing temperatures, PCR primers thought to be used for sequencmg should not be too short. To sequence m an “upstream” orientation, one of the oligonucleotides (BLl) synthesized to prepare the linker can be used as primer. In this case, annealing and extension can be done m one step at 72°C RT-PCR procedures are generally very sensitive and therefore often yield amphfication products from rare transcripts To allow identification of these transcripts by Northern blotting, load enough (3-10 pg) of poly(A)+ RNA per lane. Blots prepared with total RNA may be not sensitive enough An alternative to Northern blotting for confirmation of differential expression may be the recently described “Northern ELISA” method (Boehrmger Mannhetm), which is claimed to be at least lo-fold more sensitive than standard Northern blotting

Acknowledgment We thank Prof. Heinz Saedler for generously

supportmg

our work.

Domain-Directed Differential Display

733

References 1 Liang, P. and Pardee, A. B. (1992) Differential display of eukaryottc messenger RNA by means of the polymerase chant reaction Sczence 251,967-97 1 2. Welsh, J , Chada, K., Dalal, S S , Cheng, R., Ralph, D , and McClelland, M. (1992) Arbrtrartly primed PCR fingerprinting of RNA. Nucleic Aczds Res. 20, 4965-4970. 3. Fischer, A., Saedler, H., and Thetssen, G. (1995) Restrtctton fragment length polymorphism-coupled domain-directed differential display: A highly efficient technique for expression analysts of multtgene families. Proc Nut1 Acad Scz USA 92,533 l-5335. 4. TheiDen, G. and Saedler, H. (1995) MADS-box genes m plant ontogeny and phylogeny: Haeckel’s btogenettc law revisited. Curr. Opwuon Genet Dev 5,628-639 5 Pabo, C. 0. and Sauer, R T. (1992) Transcriptton factors: structural families and principles of DNA recogmtton. Annu. Rev Bzochem. 61, 1053-1095 6. Landsman, D. and Wolffe, A. P. (1995) Common sequence and structural features in the heat-shock factor and Ets families of DNA-binding domains. Trends Biochem Sci 20,225-226. 7. Frohman, M A , Dush, M K , and Martm, G R. (1988) Raptd production of fulllength cDNAs from rare transcripts* Amplification using a single gene-specific oligonucleottde primer Proc Nat1 Acad. Sci. USA 85,8998-9002. 8. Cobianchi, F. and Wilson, S. H. (1987) Enzymes for modifying and labeling DNA and RNA, in Methods in Enzymology, vol. 152, Academic, Orlando, FL, pp 94-l 10 9. Sambrook, J., Frttsch, E F., and Maniatis, T. (1989) Molecular CZonzng* A Laboratory Manual, 2nd ed , Cold Spring Harbor Laboratory, Plamview, NY. 10 Trentmann, S M., van der Knaap, E., and Kende, H. (1995) Alternatives to 35S as a label for the differential display of eukaryotlc messenger RNA. Sczence267, 1186.

12 Identification of Immediate-Early Gene Targets of the Raf-1 SerinelThreonine Protein Kinase Using an Estradiol-Dependent Fusion Protein, ARaf-1 :ER Sean A. McCarthy, Natasha Aziz, and Martin McMahon 1. Introduction The Raf-1 serine/threomne protein kinase is a central component of the Rasl Raf/MEWMAP kinase cascade,a highly conserved srgnaling pathway responsible for the transmission of signals emanating from cell surface growth factor, cytokine, and hormone receptors into the nucleus (J-4). Receptor-dependent activation of the Ras GTPase results in recruitment of Raf-I to the plasma membrane whereupon Raf- 1 kmase is activated by a mechamsm that is not yet fully elucidated Subsequently, Raf-1 phosphorylates and activates the dual specificity kinase, map kmase kmase (MEKMKK), which m turn phosphorylates and activates the mltogen activated protein (MAP/ERK) kmases. MAP kmases then translocate to the nucleus, where they effect the phosphorylation of a variety of transcription factors, thereby evoking changes m gene expression. Much interest has focused on the Ras/Raf/MEK/MAP kmase cascade m view of its important role m regulating cell growth and differentiation m response to polypeptide growth factors and also because constituttve activation of this pathway at level of Ras, Raf, or MEK can promote oncogemc transformation of mammalian cells (2,3) Since it is widely perceived that phenotypic responses to Ras/Raf/MEWMAP kinase signaling involve changes m gene expression, we have employed a conditional form of the Raf-1 protein kinase (5) m conjunctron with differential display PCR (6,7) to identify tmmedtateearly target genes of Raf-1 signaling. Given the complextttes of growth factor signaling, it is becoming mcreasmgly Important that systems are established for the study of responses to mdtFrom

Methods E&ted

m Molecular

Bology,

by

and

P Llang

Vol 85 Dffferenbal

A B Pardee

137

Humana

D/splay Press

Methods Inc , Totowa,

and Protocols NJ

138

McCarthy, A.&, and McMahon

c-RAF-1

ARAF-1 :ER Fig. 1. Schematicrepresentationof The ARaf- 1:ER fusion protein. ARaf- 1:ER is composedof the catalytic (CR3) domain of human Raf-1 and the hormone binding domain of the human estrogenreceptor (ER HBD). Although the construct depicted here and used most extensively in our laboratory contains the ER domain as a C-terminal fusion (5), this and other fusion proteins also demonstratehormonal regulation asN-terminal fusions (M. M. unpublished).

vidual signaling molecules. This requires the development of inducible and/or conditional systems that can be applied to wide ranges of signaling proteins. The hormone binding domains of steroid hormone receptors have been employed to regulate the activities of a number of heterologous proteins including transcription factors and signaling molecules (for review, see ref. 8), rendering the activities of these proteins conditional on the appropriate steroid hormone. The hormone binding domain of the greatest utility to date has been that of the human estrogen receptor in view of the restricted expression of this receptor in cell lines and the consequent lack of concern over activation of endogenous receptors (8). We therefore generated a fusion protein composed of the kinase (CR3) domain of human Raf-1 fused to the hormone binding domain of the human estrogen receptor (hbER) ($9) (Fig. 1). The CR3 domain of Raf-1 is a potent transforming oncogene in NIH-3T3 fibroblasts. The ER domain, however, subjects oncogenic Raf-1 to exquisite regulation by estrogen and its analogues. When the ARaf-1 :ER fusion protein is expressed in NIH3T3 cells, treatment of cells with 17P-estradiol results in activation, within minutes, of the kinase activity of the fusion protein and also of the downstream signaling proteins, MEK and MAP kinase (5). Activation of this signaling pathway persists in the continued presence of hormone and results, within 16-24 h, in complete morphological transformation of the cells (5). Transformation is fully reversible if the hormone is removed from the cells by thorough washing. Thus, ARaf- 1:ER provides an ideal conditional system for the identification of Raf-1 target genes by differential display. Our strategy for identification of immediate-early target genes of ARaf- 1:ER involves the following steps: 1. Identify candidate target genesby differential display PCR, comparing rnRNA isolatedfrom NIH-3T3 cells expressingARaf- 1:ER before and after activation of the fusion protein.

Raf-Regulated

139

Genes

2 Validate that candrdate genes are induced/repressed using RNase protectton analysis of mRNA expression 3. Test the sensitivity of mRNA mductron/represston to mhibttron of protein synthests to determine whether gene regulation IS immediate-early. 4. Perform nuclear run-on and mRNA stability experiments to determine whether mRNA regulatron occurs at the transcrrptronal level.

2. Materials 1 2 3 4 5 6 7 8 9 10. 11 12 13 14 15 16. 17. 18 19. 20. 21. 22

Qiashredder columns (Qragen, Chatsworth, CA) RNeasy RNA Isolation kit (Qragen). Message Clean Kit (GenHunter, Nashville, TN). RNAmap kit (GenHunter). TA Cloning Kit (Invrtrogen, San Diego, CA). Sephadex G-50 Spm Columns (Boehrmger, Indianapolls, IN) Cyclohextmtde (Sigma, St Louis, MO) Actmomycin-D (Sigma) 17a-estradiol (Sigma). Trans(Z)4-hydroxytamoxrfen (Research Brochemtcals, cat no. T- 156 [Natuck, MA) ICI 164,384 and ICI 182,780 (Dr Alan Wakelmg, Zeneca Pharmaceuticals, Macclesfield, UK) RNase protection hybrtdlzatton buffer (400 mM NaCl, 1 mM EDTA, 80% formamide, 40 mM PIPES pH 6 4). RNase digestion solution (300 mMNaC1, 5 mM EDTA, 10 mA4 Trts, pH 7 5, 40 &mL RNase A, 2 pg/mL RNase Tl). ExpressHyb solution (Clontech, Palo Alto, CA) RNasm (Promega, Madison, WI) 100 mM dithrothrettol (Promega) 1OX transcription buffer (Promega) RNase free DNase 1 (Boehringer) Proteinase K (Boehrmger) 100 mA4NTP stock solutions (Boehringer). CL~~P-CTP, a3*P-UTP (Amersham International). Sample loading buffer (80% v/v formamide, 1 rnA4 EDTA, 0 1% bromophenol blue, 0.1% xylene cyanol)

3. Methods 3.7. Identification

Of Candidate

ARaf-1:ER

Inducible

Genes

1. NIH-3T3 cells stably expressmg ARaf-1 *ER (C2 cells; see Note 1) are plated on to lo-cm tissue culture dishes and grown to confluence 2 Confluent C2 cell monolayers are stimulated with either ethanol (0 1%) as a solvent control or with ICI 164,384 (ICI, 1 @J’) ICI 1san estradrol receptor antagomst that activates the ARaf-l.ER fusion protein as efficiently as 17P-estradiol (see Note 2) The duration of the cell stimulation can be vaned depending on the class of regulated genes of the most mterest In our studies, a 4-h stimulation was chosen (11).

140

McCarthy, AZ/Z, and McMahon

3. Total RNA 1sextracted from the cells using RNeasy spin columns from Qtagen, after passing cell lysates over Qiashredder columns to ensure complete cell lysis, according to the manufacturer’s mstructtons (see Note 3). 4. Total RNA samples are DNase treated using the Message Clean Kit from GenHunter This step 1scructal to the success of the differential dtsplay reactions and IS particularly important when RNA has been prepared using RNeasy columns 5 Using reagents from the GenHunter RNAmap Ktt, first strand cDNAs are prepared from the total RNAs to be compared, using single base anchored T,,N primers (see Note 4) We typtcally scale the first strand synthesis up IO-fold to provide suffictent maternal for amphficattons wtth multiple forward primers 6 PCR amplifications are performed on the resulting cDNAs, using the GenHunter forward primers AP l-AP20 m combination with the appropriate T 12N primers. Since dtfferential dtsplay mvolves semtrandom PCR amplification, vartation between reactions is mevitable and must therefore be controlled for carefully by performing amphfications m duplicate or tripkate. As cDNA synthesis IS a lmear process whereas PCR amphtication is logarithmic, the greatest source of variation lies in the PCR reactions (see Note 5). We therefore recommend that secondary screenmg of posittve primer combmattons IS performed by repeating PCR amplicattons on the same cDNA preparattons used m the first round screen. 7 Analyze PCR products by separation on 6% polyacrylamide/urea sequencmg gels Figure 2 shows a typical set of results from differential display PCR amplification of RNAs isolated from C2 cells treated for 4 h with ICI or solvent control. As expected, the vast malorrty of bands show no change between the condrtions compared However, with the primer combmations shown four reproducible differences In gene expression were apparent representing 3 potential ARaf-1 :ER mducrble mRNAs (referred to as 8G, 14G1, and 14G2) and one ARaf- 1 :ER repressed mRNA (I OG).

4. Verification

of Candidate

ARaf-1 :ER Regulated mRNAs Our method of choice for testing whether cDNAs cloned by differential display tn fact represent regulated genes is RNase protection (16), rather than northern analysis, in view of the greater specifictty and sensitivity of protection assays. Furthermore, differential display PCR products are typrcally 100-600 bp m size, which 1s ideal for riboprobe synthesis.

4.7. Cloning of cDNAs From Differentially

Represented

Bands

1. Excise bands of interest from gels and elute and reampllfy as described in the GenHunter RNAmap manual

Fig. 2. (oppositepage) Differenttal dtsplay of mRNAs in NIH-3T3 cells expressmg ARaf- 1*ER PCR display gel showmg the results of secondary amplifications with five

Ra f- Regulated Genes

primer combinations that displayed differential gene expression in a first round screen. Duplicate amplifications were performed from cDNAs prepared from C2 cells stimulated with either solvent control (-) or ICI 164,384 (+) for 4 h. The arrows indicate the bands that were cut, reamplified and sequenced. AP-8G-induced band in AP-8 lanes, AP- 10G - repressed band in AP-10 lanes, 14Gl-upper induced band in AP- 14 lanes, 14G2-lower induced band in AP-14 lanes. Full-length cDNAs for clones 8G and 10G cDNAs are currently being isolated (see Note 8). Band 14G2 was not studied further in view of the fact that sequencing revealed the reamplified cDNA to be composed of at least 10 distinct cDNAs (see Note 6). Band 14G1, described in further detail in the text, encoded a cDNA identical to the 3’ end of the mRNA encoding HBEGF (20, II).


142

2 Ligate the reamphfied PCR products into the pCRI1 TA-clonmg vector from Invitrogen and sequence the cDNA mserts to determine whether the reamphfied cDNA 1shomogeneous (see Note 6)

4.2. Preparation

of Riboprobes

1 For each cloned cDNA, prepare two lmearized plasmrd templates by dlgestmg the pCRII-cDNA constructs with one enzyme that cuts 5’ of the insert and another that cuts 3’ of the Insert Do not use restriction enzymes that leave 3’ overhangs srnce these can serve as transcription mitiatlon sites and result in extremely high background m the hybrtdtzatlons. 2 Heat inactivate the restrlctlon enzyme, phenol-chloroform extract the DNA samples, ethanol precipitate and resuspend to a final concentration of 0 5 mg/mL in RNase free water. These DNA preparations serve as templates for nboprobe synthesis 3 Set up transcription reactions contaming. 2 pL 10X transcription buffer 2 ).IL 100 mM dlthtothreitol 2 p.L NTP mix (4 mM ATP, UTP, GTP) 1 pL RNase mhlbltor 8 5 pL RNase free H,O 2 5 pL (50 @I) a-32P-CTP (800 Wmmol) 1 clr, linear DNA template (0 5 mg/mL) 1 pL T7 or SP6 RNA polymerase (Total = 20 $) For each clone to be tested, both lmeartzed templates should be transcribed m separate reactions and the resulting nboprobes tested m RNase protection assays Although the orientation of the cDNAs is readily predicted m cases m which the T,,N primer has annealed to the polyA tall, anomalous priming events can occur such that the onentation of the cDNA can be unclear (see Note 6). For known cDNAs, the sense nboprobe serves as a good control for spurtous probe protectron (16) 4. Incubate at 37°C for 20-45 mm. 5. Add 2 @ RNase-free DNasel. Incubate for a further 20 mm at 37°C 6. Purify rtboprobes by passing over Sephadex G-50 spin columns, followmg the manufacturer’s mstructlons This step separates unmcorporated label from the probe and gives an indicatton of the success of the labelmg reaction. Between 10’ and lo8 total cpmprobe should be expected. 7 In some cases it IS necessary to subject probes to gel purification tf unpurified probes give high background. Probes are electrophoresed on 6% polyacrylamide/ urea gels and gel shces contammg labeled probes are excused from gels after exposure to X-ray film to locate the major labeled species. Probes are eluted m elution buffer (0.5M ammonium acetate, 10 mM magnestum acetate, 1 mA4 EDTA, 0.1% SDS) at 37°C for 6 h and ethanol preclprtated. Resuspend probes m 100 )J.Lof RNase protection hybridization buffer. 8 Count 1 $ of each punfled riboprobe m a scmttllatton counter to determme yield

Raf-Regulated

Genes

143

4.3. Hybrid&a tions 1 Dilute the required amount of probe to 5 x lo6 cpm/mL m 1X hybrtdizatton buffer A control probe that permits data quantitation should also be prepared and mcluded m the hybridizations (e.g., actm, GAPDH). We typtcally use a 120 bp GAPDH probe (see Note 7) 2. To 10 ~18samples of total RNA m 10 pL volume or less, add 50 p.L of diluted probe. Heat samples to 95-100°C for 5 mm, transfer to a 45°C water bath, and incubate overnight

4.4. RNase Digestion 1. Remove samples from the 4YC water bath and add 400 pL RNase dtgestton buffer Incubate at room temperature for 30 mm 2 Add 10 pL 20% SDS and 2.5 p.L of 20 mg/mL protemase K to each sample Incubate for 20 mm at room temperature 3. Extract samples with 500 uL phenol by thorough vortex mixing. Centrifuge, transfer aqueous phase to a fresh tube and precipitate with exactly 2 vol of ethanol. No additional salt 1s required owing to the high salt concentration in the hybridization buffer 4. Freeze samples on dry me, centrifuge, and rinse pellets thoroughly m 70% ethanol by vortex mixing to remove salt 5 Resuspend pellets in 15 pL sample loading buffer Vortex thoroughly to resuspend the samples. Heat samples to 95-100°C for 5 mm, transfer to ice for 10 min and analyze by electrophorests on 6% polyacryIamrde/urea sequencmg gels Figure 3 shows RNase protection analysis of a ARaf- 1:ER inducible mRNA identified m the screen shown m Fig. 2. A cDNA reamplified from band 14Gl was sequenced and found to represent the 3’ end of the murine mRNA encoding Heparin binding epidermal growth factor (HBEGF). The RNase protection assay shown used a riboprobe complementary to the coding region of mHBEGF (see Note 7). Clearly, HBEGF mRNA 1s rapidly induced followmg acttvation of ARaf- 1 :ER and other related ARaf:ER fusion protems (12). Similar results were obtained using a 3’ probe complementary to the 14Gl cDNA isolated by differential display (II). No mduction of HBEGF mRNA was observed when parental NIH-3T3 cells were stimulated with either 17P-estradiol or ICI indicating that this response requires the presence of the ARaf-1 :ER fusion protein (2 I).

5. Classification Of HBEGF mRNA Induction by ARaf-1 :ER as an Immediate Early Transcriptional 5.1. Sensitivity to Cyc/oheximide

Response

To test whether mductiorkepressron of mRNAs by ARaf-1:ER can be assigned to the immediate-early response class, the sensitivity of the response to the protein synthesis mhtbitor, cycloheximtde, should be tested.

144

McCarthy, Aziz, and McMahon A

3T3 020’1

ARaf-1 :ER 6

2 4 824

+

C

3T3

AB-Raf:ER

0 20’ 1 2 4 8 24

ICI (hrs)

GAPDH -+

3T3 AA-Raf:ER

Rl D

0 20’ 1 2 4 8 24

a ARaf-1 :ER

0 20’ 1 2 4

8 24

ICI (hrs)

+HB-EGF+

+GAPDH

+

Fig. 3. RNase protection analysis of HBEGF mRNA expression following activation of Raf-ER tksion proteins. A cDNA complementary to the coding region of the murine HBEGF cDNA (ZO) was cloned and used to generate antisense riboprobes for RNase protection analysis of HBEGF mRNA expression (see Note 7) following activation of: (A) ARaf-1:ER in NIH-3T3 cells; (B) AB-RafzER in NIH-3T3 cells; (C) AA-Rat ER in NIH-3T3 cells; and (D) ARaf-1 :ER in Rat 1a cells. The fusion proteins were activated with ICI 164, 384 (1 pM). GAPDH mRNA was quantitated simultaneously as a loading control. (Reproduced with permission from ref. 1 I .) 1. C2 cells and NIH-3T3 cells expressing the ARaf-1:ER fusion protein are made quiescent by overnight incubation in medium containing 0.5% fetal calf serum. 2. Cells are treated with cycloheximide (25 pg/mL) for 1 h to inhibit de nova protein synthesis prior to activation of ARaf-1 :ER with ICI (4 h) or challenge with 20% serum (1 h). 3. Total mRNA is extracted from the cells as above and RNase protection assays are performed to quantitate the mRNAs of interest.

Figure 4A shows the effect of cycloheximide pretreatment on induction of HBEGF mRNA expression by ARaf- 1:ER and serum. In the presence of cycloheximide, superinduction of HBEGF mRNA expression occurred in both cell lines in response to both ICI and serum. Thus, induction of HBEGF mRNA expression by ARaf-1 :ER, AB-Raf:ER or serum stimulation (which results in endogenous Raf kinase activation) does not require de nova protein synthesis and can therefore be classified

as an immediate-early

response.

Raf-Regulated

145

Genes

5.2. Nuclear Run-On Transcription

Analysis

To determine whether changes m mRNA expression following activation of ARaf- 1 :ER result from increased gene transcription, nuclear run-on transcription assays should be performed (16). 1 Cells are made quiescent by overnight incubation in medium contaming 0 5% fetal calf serum then treated with either 0 1% (v/v) ethanol as a solvent control or 1 cuz/IICI for the required time 2. Rinse cells in ice-cold phosphate buffered salme (PBS), scrape from plates (on ice) and collect by centrifugation (lOOOg, 5 mm). Wash cell pellets m cold PBS and recentrifuge (1 OOOg,5 mm). 3 Resuspend cells m 5 mL of buffer RSB (10 mMTrts, pH 7 4,10 mMNaC1, 10 mA4 MgCl& and incubate on ice for 2 mm. 4. Add 5 mL of buffer RSB supplemented wtth 1% NP-40 and incubate on me for a further 2 min. Collect nuclei by centrifugation (2OOOg, 10 min) and resuspended m buffer NSB (20 mM Tris, pH 8.0, 75 mM NaCl, 0 5 mM EDTA, 0.85 mA4 dithiothreitol, 0.125 mM PMSF, 50% glycerol). 5. Recentrifuge (2OOOg, IO min) and resuspended m one pellet volume of buffer NSB. 6 Set up run-on transcription reactions containing* 50 Ils, nuclei (from 2 x 10’ to 1 x lo8 cells) 40 pL 5X transcription buffer 43 & 70% glycerol 2 pL 100mMATP 2 $ 100mMGTP 2 /JL 1OOmMCTP 25 pL a32P-UTP (800 Cmnmol, 0 5 mCi) 36 clr, H,O (Total = 200 &) Nuclei should be added last and the reaction mixed gently Incubate at 37°C for 30 min. 8 Extract labeled RNA transcripts from nuclei using Qiagen RNeasy columns by lysmg nuclei m 1 mL of buffer RLT and processing each nuclear lysate on two columns Elute transcripts in 100 p.L,RNase free water and count a I-pL sample in a hqmd scmttllation counter Approximately 10’ cpm should be obtained from 5 x 10’ cells. 9 Prepare hybridization filters. Five-microgram samples of plasmid DNAs contammg cDNAs of interest are slot blotted onto Hybond-N+ membrane, using a Schleicher and Schuell manifold, UV crosslinked in a Stratalmker, and prehybridtzed for 3 h at 65°C in ExpressHyb hybridization buffer. Labeled transcripts are denatured by boiling then hybridized to the immobilized plasmid DNAs overnight at 65°C 10. Wash filters to a final stringency of 0.5X SSC/O 2% SDS at 65’C. 11 Treat filters with 10 pg/mL RNase A m 2X SSC, 0.1% SDS then rinse in 2X SSC and expose to film.

146

McCarthy, Aziz, and McMahon

A 1 2 3 4 5 6 7 6 9101112 1 ARaf-1:ER

1

AB-Raf:ER ICI SERUM CHX

HB-EGF

GAPDH

HB-EGF BS NE0 C-FOS Fig. 4. Immediate-early induction of HBEGF gene transcription by ARaf:ER. (A) Sensitivity to cycloheximide of HBEGF mRNA induction by ARaf- 1:ER, AB-RafiER, and serum. Quiescent NIH-3T3 cells expressing the indicated RafiER fusion proteins were stimulated with either ICI (4 h) or 20% v/v serum (1 h) following either no treatment or pretreatment for 1 h with cycloheximide (25 pg/mL). HBEGF mRNA levels were then quantitated by RNase protection. (B) Nuclear run-on analysis of gene transcription following activation of AB-RaEER in NM-3T3 cells. NIH-3T3 cells expressing AB-RafiER were stimulated for 3 h with either solvent control or ICI and transcription of the HBEGF, c-fos, and neo genes measured as described in the text. The neo gene is expressed from the retroviral LTR driving expression of the AB-RafzER fusion protein and serves as a positive control.

Raf- Regula ted Genes

147

C 12345676 ICI ActD (hrs)

HB-EGF

GAPDH

Fig. 4. (continued) (C) HBEGF mRNA decay in the absence and presence of ARaf- 1:ER activity. C2 cells were treated with solvent control (-) or ICI for 6 h after which actinomycin-n was added for the indicated times. HBEGF mRNA levels were then quantitatedby RNaseprotection.The HJ3EGFmRNA half-life was calculatedto be 2 h in both the presenceandabsenceof ICI. (Reproducedwith permissionfrom ref. II.)

Figure 4B shows the results of nuclear run-on analysis of HBEGF gene transcription following activation of DB-RafiER in NIH-3T3 cells. Compared to the appropriate controls, HBEGF gene transcription is stimulated four- to fivefold within 3 h of activation of AB-Raf:ER indicating that elevation of HBEGF mRNA arises, at least in part, from increased gene transcription. 5.3. Measurement

of mRNA Stability

In addition to changes in gene transcription, changes in mRNA stability might also contribute to the induction or repression of mRNAs following ARafzER activation. Measurement of mRNA decay following inhibition of the transcriptional machinery with actinomycin-n permits changes in mRNA halflife to be determined, indicating whether mRNA stability is altered. 1. C2 cells aretreatedwith either solventcontrol,or ICI to activate ARaf-I :ER for 6 h. 2. After the required time, cells are treated with the RNA polymerase II inhibitor, actinomycin-D, for 0, 1,3, or 6 h. 3. Total RNA is prepared asabove and expressionof the mRNA of interest is quantitated by RNaseprotection.

148


Figure 4C shows RNase protection analysis of HB-EGF mRNA expression m C2 cells followmg mhibition of transcription with actmomycin D In either the presence or absence of ICI, HBEGF mRNA levels decayed with a half-ltfe of 2 h mdicatmg that m thts case changes m mRNA stability are unlikely to contribute to elevation of HBEGF mRNA levels by ARaf-1 :ER, despite the fact that the 3’ untranslated region of the HBEGF mRNA contains multtple mRNA destabilizmg motifs (20). Thus, maintained elevation of HB-EGF mRNA expression m response to ARaf-1 :ER activation requires conrmued gene transcrtption, as previously determmed by run-on transcription experiments.

6. Conclusion This chapter details our approach to the identification of direct target genes of an estradiol-regulated form of the Raf-1 protem kmase, ARaf-1 .ER, m NIH-3T3 cells. The ARaf- 1 :ER fusion protein has also recently been expressed and characterized m cultured cardiac myocytes (18) and in macrophage (19), pheochromocytoma (D. Woods and M. M., unpublished observations) and hippocampal (20) cell lines Moreover, estradiol-regulated forms of the A-Raf and B-Raf kinases have been constructed (12). In addition to Raf kmases, the ER domain has been used to regulate the oncogemc tyrosme kmases V-SK and v-e&B and the dual specificity kmase, map kinase kinase (M. M., unpublished observations) together with several transcription factors mcludmg E 1a and Myc (for review, see ref. 8). The combined application of these and other conditional systems with the powerful technique of differential display PCR should improve our understanding of the mechanisms through which gene expression patterns are modulated dynamically by Intracellular signalmg pathways m multiple cell types

7. Notes 1 The ARaf-I ER fusion protein was introduced into NIH-3T3 cells using a rephcation defective ecotropic retrovnus and stable cell lines expressing the fusion protein were isolated by drug selection (5) C2 refers to one of several clones that were characterized Smce phenol red has weak estrogemc acttvtty whtch can lead to low level activation of ER fusion proteins m cultured cells, cells should be cultured m phenol red-free media Estrogemc compounds in fetal calf serum can also result m basal acttvation of ER fusion proteins but can be removed by charcoal stripping tf necessary (8) although mamtenance m phenol red free medium is usually sufficient 2 In addition to being activated by 17@estradiol, ARaf ER fusion proteins contammg the human ER domain can be activated by the ER antagonists Trans(Z)4hydroxytamoxtfen (4-HT); ICI 164,384; and ICI 182,780 (5) Using these compounds to activate ER fusion proteins has the advantage of ensuring that endogenous estrogen receptors, which may be expressed m some cultured cells, are not activated. Recently we have constructed a ARaf- 1:ER fusion using a sec-

Raf-Regulated

3

4.

5

6

Genes

ond generation ER domain, ER* (D. Woods and M.M., unpublished). ER* IS a modified form of the murme ER hormone bmdmg domam that cannot be activated by estradiol but is activated by 4-HT, ICI 164,384; and ICI 182,780 (13,14) ER* also lacks the TAF-2 hormone-dependent transactivation function, which resides m the hormone-binding domam of the human ER (13,14). Although less of a concern for ER-regulated cytoplasmic signaling molecules, such as Raf-1, TAF-2 activity can interfere with ER-transcription factor fusions Another advantage of ER* is that is opens the possibility of studymg ER fusion protems within the context of transgemc animals smce the ER* domam is resistant to activation by endogenous mouse estrogens. ER* also obviates the requirement for phenol red free media and charcoal stripped fetal calf serum m cell culture (13,14). RNeasy columns from Qtagen provide a rapid and effective means of isolation of total RNA from cultured cell lines that does not mvolve large volumes of phenol or require ultracentrifugatton The RNA is of high quality and performs well m differential display reactions after DNase treatment. One major benefit of this isolation method is the ability to process multiple samples simultaneously, permittmg extensive timecourse studies to be performed Our experience with the 2-base anchored T,,MN primers from Genhunter led us to focus on use of T,,MG and T,*MC since T,,MT often gave high background and T,,MA often gave very poor amplification. These problems have since been circumvented by the use of single base anchored T,,N primers, which work equally well regardless of the nucleottde present at the N position (ZS) Our preliminary experiments were performed usmg duplicate cDNA preparations for each of the two conditions compared, followed by duplicate PCR ampbfications. However, we encountered significant problems with reproducibility such that it was difficult to identify consistently regulated bands. We therefore decided to prepare cDNA once, perform PCR amplifications m duphcate and then on a different day repeat the PCR ampliticattons a second time using the identical cDNA preparations used in the first round amplifications cDNA synthesis reacttons were therefore scaled up 5 to 1O-fold to provide sufficient material for repeated amplification. In experiments where criteria for selecting bands were relaxed to the point that bands were cut after only one PCR amplification, our success rate fell from 75%, as in the screen described here, to less than 10% (McCarthy and Aziz, unpublished observations). We and others have found that reamphfied differential display products are frequently heterogeneous when sequenced, presumably because of amplification of cDNA species that comigrate with the band of interest. This problem can often be avoided by runmng separation gels longer and taking extra care whde cuttmg bands from the gels If reamplified PCR products continue to be heterogeneous, differenttal screening of these clones with radiolabeled cDNAs prepared from the conditions under study can be used to Identify induced/repressed clones. Further complication can arise from priming by T,,N primers on polyA stretches located within cDNAs (17), makmg it important that riboprobes complementary to both strands of cloned cDNAs are tested in RNase protection assays.

McCarthy,

150

AZ/Z, and McMahon

7. The HB-EGF rtboprobe was prepared by SPG-transcrrptron of 627-bp template comprising the entire codmg region of the murme HB-EGF cDNA, which was obtained by PCR amplificatton from first strand cDNA prepared from 3T3 cells using the ohgonucleottde primers; S-ATGAAGCTGCTGCCGTCGGT-3’ and 5’GCGTGGCTAGCTCCCACTGA-3’ The GAPDH rtboprobe was prepared by T3 transcription of a 120-bp fragment of human GAPDH amphfied wtth the ohgonucleotrde primers; 5’-GACTGAATTCGACAACAGCCTCAAGATCAT-3’ and 5’-GACTGGATCCGGCATGGACTGTGGTCATG-3’ The human GAPDH probe crossreacts wtth GAPDH mRNA m mouse, rat, and hamster mRNAs 8 We have found the Genetrapper cDNA Posmve Selection System from Gtbco to be very useful for clonmg of full-length cDNAs represented by dlfferenttal display PCR products Genetrapper mvolves cDNA capture using a btotmylated ohgonucleottde complementary to the dtsplay clone and capture of specific ohgonucleottdecDNA hybrids on magnetic streptavtdin-coated beads Thts techmque 1srapid and permits the processmg of multtple display clones simultaneously However, we have found Genetrapper to be extremely ohgonucleotide dependent and recommend that at least two ohgonucleotrdes per display clone are tried and that close attentron is paid to meltmg temperature smce the repair step 1sperformed at 70°C

References 1 Karm, M. (1994) Signal transductron from the cell surface to the nucleus through the phosphorylatton of transcription factors. Curr Opuuon Cell Btol 6, 415-424 2. Marshall, C J (1994) MAP kmase kmase kinase, MAP kmase kinase and MAP kmase Curr Opmon Genet Dev 4,82-89. 3. Marshall, C. J. (1995) Specifictty of receptor tyrosme kinase signaling: transient versus sustained extracellular signal-regulated kinase activatron Cell 80, 179-185 4 Hill, C S. and Treisman, R (1995) Transcrrptronal regulation by extracellular signals. mechanisms and spectficny. Cell 80, 199-2 11. 5 Samuels, M. L., Weber, M J., Bishop, J M , and McMahon, M (1993) Condttronal transformation of cells and rapid acttvatron of the mttogen-activated protem kmase cascade by an estradrol-dependent human raf-1 protem kmase MOE Cell Blol

13, 6241-6252

6. Lrang, P and Pardee, A. B. (1992) Differential display of eukaryotic messenger RNA by means of the polymerase cham reaction Science 257,967-97 1 7. Ltang, P., Averboukh, L , and Pardee, A. B (1993) Dtstributton and cloning of eukaryotic mRNAs by means of differential display: refinements and optimization Nuclerc Aclds Res 21,3269-3275. 8 Mattioni, T , Louvion, J F , and Picard, D. (1994) Regulation of protein activities by fusion to steroid binding domains Methods Cell Blol 43, 335-352. 9 Kumar, V , Green, S., Staub, A, and Chambon, P (1986) Locahsatron of the oestradtol-bmdmg and putative DNA-bindmg domains of the human oestrogen receptor. EMBO J 5,223 I-2236

Raf-Regulated

Genes

151

10. Htgashtyama, S , Abraham, J A, Miller, J , Ftddes, J C , and Klagsbrun, M (199 1) A heparm-binding growth factor secreted by macrophage-hke cells that 1s related to EGF Sczence 251,936-939 11 McCarthy, S A , Samuels, M L , Pritchard, C A , Abraham, J A., and McMahon, M (1995) Rapid mductton of heparm-bindmg epidermal growth factor/dlphthena toxin receptor expression by Raf and Ras oncogenes. Genes Dev 9, 1953-l 964 12. Pritchard, C. A , Samuels, M. L., Bosch, E , and McMahon, M. (1995) Condttronally oncogemc forms of the A-Raf and B-Raf protein kmases display different biologrcal and biochemtcal properties m NIH 3T3 cells MoZ Cell Bzol 15, 6430-6442. 13. Danielian, P. S , White., R , Hoare, S A., Fawell, S E , and Parker, M. G (1993) Identification of residues m the estrogen receptor that confer differential senstttvity to estrogen and hydroxytamoxifen Mol Endocnnol 7,232-240 14 Littlewood, T D , Hancock, D,C., Danielian, P S., Parker, M G , and Evan, G I (1995) A modified oestrogen receptor hgand-bmdmg domam as an improved switch for the regulation of heterologous proteins. Nuclezc Acids Res 23,1686-l 690 15 Lrang, P., Zhu, W , Zhang, X , Guo, Z , O’Connell R P., Averboukh, L., Wang, F., and Pardee, A. B (1994) Differential display using one-base anchored ohgodT primers NucEelc Acids Res 22, 5763,5764. 16. Ausubel, F. M , Brent, R., Kingston, R E , Moore, D. D , Serdman, J G., Smith, J A., and Struhl, K (1987) Current Protocols zn Molecular Biology, John Wiley, New York. 17 Guimaraes, M J, Lee, F , Zlotmk, A., and McClanahan, T (1995) Differential display by PCR. novel findmgs and applications. Nuclezc Acids Res 23, 1832-l 833 18 Thorburn, J , McMahon, M., and Thorburn, A (1994) Raf-1 kmase actrvity 1s necessary and sufficient for gene expression changes but not sufficient for cellular morphology changes associated with cardiac myocyte hypertrophy. J Blol Chem 269,30,58@-30,586 19. Hambleton, J., McMahon, M , and De France, A. L (1995) Acttvatton of Raf-1 and mttogen-activated protem kinase in murme macrophages partrally mimtcs hpopolysaccharide-induced signaling events. J Exp Med 182, 147-154. 20. Kuo, W., Abe, M., Rhee, J., Eves, E M., McCarthy, S. A., Yan, M , Templeton, D. J., McMahon, M., and Rich Rosner, M. (1996) Raf, but not MEK or ERK, 1s sufficrent for differentiation of hippocampal neuronal cells Mol Cell Btol 16, 1458-1470

13 Isolation of Cytokine-Inducible Genes from Hematopoietic Cells by Differential

Display

Yuan Zhu, Tom Hahn, and Alan D. D’Andrea 1. Introduction Hematopoietic stem cells have the capacity to divide and differentiate mto several different hematopoietic cell lineages. Hematopoietrc growth factors (cytokmes) promote both the growth and differentratron of stem cells and committed progenitor cells. Cytokmes bmd and activate discrete receptors expressed on the surface of stem cells and thereby influence cell fate decisions. Polypeptide receptors, such as the receptor for mterleukm-3 (IL-3) and erythropoietm (EPO), are members of the cytokine receptor superfamily (Z). Followmg ligand strmulatron, these receptors activate several downstream slgnal transduction processes. For instance, both the IL-3R and the EPO-R activate the RAS/Raf-l/MAP kmase signaling pathway (2) and the JAIUSTAT signaling pathway (3,4). How cytokme-specific signals are mduced by these growth-factor regulators remains unknown. Presumably, cytokme-specific signaling pathways also exist downstream of each receptor. The identrfrcatron of drfferentratron-specific pathways downstream of cytokme receptors has been hampered by the absence of cell lines that respond to one growth factor or another. Recently, we have characterized one cell line, Ba/F3-EPO-R, that responds to two different growth factors (Fig. 1). Ba/F3EPO-R cells proliferate m response to IL-3 but proliferate and differentiate into hemoglobin-bearing cells m response to EPO. In the current study, we designed a strategy of differential display to clone cDNAs that are specifically Induced in Ba/F3-EPO-R cells by either IL-3 or EPO. We cloned several specific cDNAs that are induced as either immediateearly genes or delayed-response genes, m response to these cytokmes. The identification of these cDNAs supports a model of discrete signaling pathways From

Methods m Molecular Bology, Vol 85 Dfferentral Edited by P Llang and A E Pardee Humana

153

Dqlay Methods and Profoco/s Press Inc , Totowa, NJ

Zhu, Hahn, and D’Andrea

154 IL3 ,-e

Proliferation

L

~$Z%ferentiation

EPO

Fig. 1. Schematic representation of Ba/F3-EPO-R cells. Ba/F3-EPO-R cells have functional receptors for interleukin3 (IL-3) and erythropoietin (EPO). IL-3 promotes cellular proliferation. EPO promotes proliferation and partial erythroid differentiation (7).

downstream of each cytokine receptor. Moreover, the 5’ cis regulatory elements of these inducible genes should help identify signature DNA sequences upstream of cytokine inducible genes and identify novel transcriptional activators that bind and activate cytokine-specific gene expression. 2. Materials 2.1. Cell Lines and Cell Culture 1. Ba/F3: IL3-dependent murine pro-B cell line (5). 2. Ba/F3-EPO-R: Derived from Ba/F3 by stable transfection with the cDNA encoding the murine EPO-R (6). Ba/F3-EPO-R cells grow in either murine IL-3 or human EPO (7,8). 3. Medium: RPMI- 1640 (Cellgro, Herndon, VA) supplemented with 10% fetal calf serum (FCS) and 10% conditioned medium from WEHI-3B cells as a source of murine IL-3. 4. Murine IL-3 (Kirin, Japan). 5. Human EPO (Genetics Institute, Cambridge, MA): 1 pM= 10 mU/mL. 6. Phosphate-buffered saline (PBS).

2.2. Preparation

of Total Cellular RNA

1. GTC buffer: 4Mguanidinium isothicyanate, 25 mM sodium citrate, pH 7.0,O. 1M P-mecaptoethonal (add freshly). Filter to remove GTC debris. Store at room temperature. 2. CsCl solution: 5.7Mcesium chloride, 25 mM sodium acetate, pH 5.1. Treat with DEPC (0.1%) for 30 min, and then autoclave. Store at room temperature. 3. Backman L8-M Ultracentrifuge. 4. RNase free DNase I (Gibco-BRL, Grand Island, NY).

2.3. Reagents

for Differential

Display

1. 5X First-strand buffer: 250 mMTris-HCl (Gibco-BRL, supplied with enzyme).

pH 8.3,375 mMKC1,

15 mA4 MgClz.

Isolation

of Cytokine-Inducible

Genes

155

2 100 mMdithtothreno1 (Gtbco-BRL, supplied with enzyme). 3. 10X dNTP mix: 0.2 mMdATP, 0.2 mMdCTP, 0.2 mA4 dGTP, 0.2 mMdTTP 4 Ohgo primers (see Note 1). Synthesized on a DNA synthesizer Prepare a lo-pA4 solutton m distilled water. Store at -20°C. 5 Random primers (Operon, Alameda, CA) (see Note 1). IO-mer. Prepare a 2-w solution in distilled water Store at -2O’C 6. M-MLV reverse transcriptase (Gibco-BRL) 7 10X PCR buffer 100 mA4Tris-HCl, pH 8.3,500 mA4KC1, 15 mA4MgC1,O 01% gelatin (Perkin Elmer, supplied with enzyme). 8. AmplrTaq (Perkm-Elmer, Branchburg, NJ) 9 35S-dATP: specific activity 1200 Ci/mmol. 10 QIAEX II Kit (Qtagen, Chatsworth, CA)

2.4. Northern 1. 2. 3 4.

Analysis

Duralon-UV membranes (Stratagene, La Jolla, CA) QuikHyb (Stratagene) 20X SSC: 3MNaC1, 0 3M sodmm citrate, pH 7 0 10% SDS.

2.5. Cloning of PCR Fragments 1. TA Clomng Kit (Invttrogen, San Diego, CA). 2. DNA Sequencing Kit (USB, Lake Placid, NY)

3. Methods 3.1. Preparation

of Cells

1. Collect lo* cells and wash three times in PBS. 2. Incubate the cells in plain medium (RPM1 + 10% fetal calf serum) at 37’C for 8 h 3. Split cells into 3 equal portions Collect the first portion as starved sample. Stimulate the second portion with IL-3 (10 PM), and the third portion with human EPO (5 U/mL) Incubate the cells at 37°C for 3 h and then harvest them. 4 Wash with PBS.

3.2. Preparation 1 2. 3. 4. 5 6. 7.

of RNA Samples

Lyse cells m 3.5 mL GTC buffer with fresh P-mecaptoethanol. Shear DNA by passing through a 21-gage needle. Overlay the lysate on a 1 5 mL of CsCl using SW 55 polyalomer tubes. Centrifuge at 35,000g for 16 h at 20°C Suspend the pellets in 0.3 mL DEPC water. Add NaOAc pH 4 to 0.3M, precipitate with 2 vol of 100% ethanol for 1 h at -70°C. Take 50 mg of each RNA sample. Add 1 mL (10 U) of RNase Inhibitor and 10 U of DNase I Incubate for 30 mm at 37°C. 8. Phenol/chloroform extraction 9. Precipitate RNA as in step 6. 10. Quantitate by ODZ6s and adjust to 0.1 pg/mL in DEPC water. Store at -7O’C

156 3.3. Differential

Zhu, Hahn, and D’Andrea Display

1 Apply 2 & (0 2 pg) of RNA m each reverse transcription reaction Add 4 p.L of 5X first-strand buffer, 2 & of DTT, 2 uL of dNTP mix, 2 & of oligo(dT) primer, and 7 Ils, of DEPC water Heat the mix for 5 mm at 65°C and incubate for 10 mm at 37°C for primer annealmg Add 1 pL (200 U) of M-MLV Reverse Transcriptase. Incubate for 1 h at 37°C 2 Stop the reaction by heating at 9S’C for 5 mm. Store at -20°C 3 Use 2 pL of reverse transcrtptton mix for each PCR m 20 pL final volume Add 2 pL of 10X PCR buffer, 2 pL of 1’ 10 diluted dNTP (20 @4), 1 pL 35S-dATP, 2 & of random primer, 2 pL of ohgo primer, 8 8 pL of distilled water, and 0 2 pL of AmphTaq Cover the solution with 25 pL mineral oil Perform PCR at 94’C for 30 s, 40°C for 2 mm, 72°C for 30 s (40 cycles), and then extend at 72°C for 5 mm 4 Mix 6 pL of PCR reaction with 4 pL of sequencmg loading dye Incubate at 80°C for 2 mm Load onto a 6% DNA sequencing gel Run the gel until the xylene cyanole FF reaches 2 m above the bottom. 5 Directly dry the gel Put a fluorescent labeled Stratagene Glue (Stratagene) on both the top and the bottom of the gel Expose it to X-ray film overnight with intensive screen 6 Locate the specific PCR fragment. Cut the gel through the film and recover the DNA by boilmg the piece of the gel m dtstilled water 7 Reamphfy the DNA as step 3 m 50 pL of total volume 8 Run 20 pL on a 2% agarose gel 9 Elute the DNA fragment using QIAEX II Kit (Qtagen) and use half as probe for Northern analysts (see Note 2) and save the other half for cloning

3.4. Northern Analysis to Confirm the Specificity of Isolated DNA Fragments 1 Run 20 ug of the RNA samples that were applied m differential display experiment on denaturing formaldehyde gels and blot onto Duralon-UV membranes. 2 Label the DNA fragment using random prtmer labeling ktt (Stratagene) 3. Hybridize the membrane with labeled probe for 1 h at 68°C and wash the membrane at room temperature m 0.1X SSC and 0.1% sodium dodecyl sulfate. 4. Expose the membrane to X-ray film overnight at -70°C

3.5. C/one Positive

PCR Fragment

1 For these DNA fragments that show positive on the Northern blots, take a aliquot of the purified DNA (Section 3 3., step 9) and clone mto pCRI1 using TA Cloning Kit. 2. Sequence 3-5 independent clones using DNA sequencing Kit (see Note 3)

3.6. Examples

of Application

We have utiltzed the strategy of dtfferenttal and EPO inducible genes from the Ba/F3-EPO-R

dtsplay (9,10) to isolate IL-3 cells, which have functional

Isolation of Cytokine-Inducible

Genes

157

receptors for both IL-3 and EPO. Imtially, we isolated a partial cDNA that was specifically induced by IL-3 but not by EPO (Fig. 2). For these studies, total RNA was isolated from Ba/F3 cells that were starved and restimulated wrth either murme IL-3 (Fig. 2A, lane 1) or recombinant human EPO (lane 2-4). A spectfic, partial cDNA was identified m IL-3 stimulated cells but not m EPO stimulated cells. In order to confirm differential expression of the mRNA, a Northern blot was performed using total RNA samples prepared from Ba/F3EPO-R grown in the same conditions (Fig. 2B). A 3-kb mRNA was identified that was specifically expressed in IL-3 stimulated cells but not m EPO stimulated cells The IL-3 inducible cDNA was isolated by screening a cDNA library prepared from IL-3 stimulated Ba/F3 cells. We have subsequently identified this mRNA as DUB- 1, an IL-3 mducible immediate-early gene that encodes a growth regulatory deubrquitmating enzyme (I 1). The isolated cDNA was used to clone a full length gene. In this way, we identified a specific c&acting enhancer element that regulated the IL-3 inducible DUB-1 expression (22). Interestingly, cells stimulated with EPO for longer periods of time expressed a larger mRNA (5 kb) that crosshybridrzed with the probe generated by drfferential display (Fig. 2C). The cDNA containing the open reading frame of the DUB-1 gene did not hybridize with this 5-kb mRNA (data not shown) suggesting that the sequencehomology between the IL-3 induced DUB- 1mRNA and the EPO induced 5-kb mRNA was restricted to the 3’ untranslated regions. By screening a cDNA library prepared from EPO-induced Ba/F3-EPO-R cells, a full-length cDNA encoding coproporphorynogen oxrdase (CO) was identified (Fig. 2D) Interestingly, CO has previously been identified as an erythroid specific cDNA that is upregulated in response to EPO-induced erythroid differentiation (13). The region of homology between the DUB-l cDNA and the CO cDNA is found to be a Sl repeat located m the 3’ untranslated regions, as predicted by the hybridization data. The strategy of differential display was next used to identify additional mRNAs that are induced by either IL-3 or EPO (Fig. 3). Two partial cDNAs were identified that are specifically induced in response to IL-3 (Fig. 3A,B). Two partial cDNAs were identified that are specifically induced m response to EPO (Fig. 3C,D). Characteristics of the IL-3 or EPO induced cDNA clones are summarized m Table 1. Some general conclusions can be drawn from these observations. First, IL-3 but not EPO induces the expression of G15. This mRNA is highly related to a known interferon inducible cDNA (14,151. These results suggestthat members of this gene family are cytokine-mducible and may therefore play a regulatory role in cellular proliferation. Another IL-3 inducible clone (A9) encodes mast cell protease-5. MP-5 has previously been identified as a protein expressed in IL-3 dependent bone marrow mast cells (16). The cloning of a


158

A

IL3

hours

B

EPO 81224

DUB-l

IL3 EPO -ii 8h 12h 24h

-b

1

C

I”

1234

2

3

-

28s

-

18s

4

IL3 EPO L -

28s

-

18s

1

1

2

Fig. 2. Identification of DUB-l, an IL-3 inducible immediate-early gene with deubiquitinating activity. (A) Total RNA from Ba/F3-EPO-R cells treated with IL-3 or EPO was subjected to differential display analysis as described in the text. A single band (indicated by arrow) was isolated and subcloned into pCRI1. The partial cDNA was sequenced and shown to contain the PCR primer pair. (B) The cloned partial cDNA was labeled by the random primed labeling procedure and used to probe a Northern blot with the indicated Ba/P3-EPO-R total RNA. (C) The same labeled cDNA used in (B) was used to probe a Northern blot with RNA from (3-d) IL-3 or EPO stimulated Ba/P3-EPO-R.

known IL-3 inducible cDNA in our strategy further confirms the fidelity of our system. Two EPO inducible genes have also been isolated (AS and G6). Clone A5 has sequence similarity to acetyl coenxyme A synthetase (27). Northern blot analysis of multiple hematopoietic cell lines suggests that the A5 mRNA has preferential expression in cells with erythroid phenotype (data not shown). Clone G6 is an orphan cDNA clone with no known homologies in

Isolation of Cytokine-Inducible D

159

Genes

+1

1581

I

I

DUB-l

Coproporphyrinogen

Oxidase

Fig. 2. (continued) @) The cDNAs correspondingto the IL-3-induced 3-kb mRNA (DUB-l) and the EPO-induced 5-kb mRNA (coprophyrinogen oxidase) are shown schematically.

B

A (3 i! 2u.l

Clone G9

(*O Hi

Clone G15

C

C90 r4

Clone A5

D

C90 A4

Clone G6

Fig. 3. Identification of four differentially expressedcDNAS induced by cytokines in Ba/F3-EPO-Rcells.Eachpanel (A-D) showsadifferentially expressedcDNA either by differential display analysis or by Northern blot analysis. Panels A and B are examples of cDNAs selectively induced by IL-3. Panels C and D are examples of cDNAs selectively induced by EPO.

the gene bank. The possible function of these cDNAs in erythroid differentiation is unknown. In conclusion, we have used differential display to identify cytokine inducible genes in hematopoietic cells. By using a cell line that responds to two discrete cytokines, IL-3 and EPO, we were able to identify cytokine-specific genes. Interestingly, the enhancers of these genes have provided us with tools for identifying cytokine-specific signaling pathways. Finally, the cloned cDNAs can now be functionally assessedby their reintroduction into Ba/F3 cells using constitutive and inducible eukaryotic expression systems. In this way, we can demonstrate the role of the cytokine-inducible genes in cell growth and differentiation. 4. Notes

1. The oligo(dT) primersusedare5‘W3’,5’WC 3’, 5’ TTTTTTTTTTTTMG 3’ and 5’ TTTTTTTTTTTTMT

3’ (M = A, C, and G


160 Table 1 Example of cDNAs That Are Selectively Induced by Either IL-3 or EPO in the Same Cell Type mRNA, kb

cDNA cloned, bp

G15(IL-3)

10

763

A9(IL-3) T14(IL-3)

09 3.0

385 2676

AS(EP0)

4.0

2439

G6(EPO) T14(EPO)

30 35

1000

Clone

Gene identity

Sequence homology IFN-regulated family

gene

Mast cell proteaseDeubiquitmatmg enzyme family Ace@-coenzyme A synthetase Unknown

800 Coproporphyrmogen oxidase

mix). Some of the random primers (lo-mer) from Operon gave very few PCR fragments Preselection of good prtmers is recommended 2. Using the re-PCR product directly as a probe for Northern confirmation was successful, at least m our hands. We found that about 20% of the probes were true posttive. Owing to the posstble contammation m re-PCR reaction, we often found multiple mRNA bands on a Northern blot If any of the mRNA species were inducible, we cloned the PCR product and identified the specific DNA by performing Northern analysis respectively. We also found that about 50% of the probes were false positive, and 20-30% of the probes did not give any signal, perhaps owmg to the low copy number of the mRNA. 3. The re-PCR products are likely to be contaminated with nonspecific DNA due to contaminated templates from the gel or nonspecific PCR reactions After cloning, it is necessary to evaluate multiple clones (3-5) and to confirm the 5’ and 3’ primers by DNA sequencing If there is more than one clone with the particular primer pair, test each clone on Northern blots separately One clone may be differentially expressed, the other may not.

References 1 Cosman, D., Lyman, S. D., Idzerda, R L , and Beckmann, M P (1990) A New cytokine receptor superfamily. Trends Biochem Scz 15,265-270 2 Sato, N., Sakamaki, K, Terada, N., Arai, K., and Miyajima, A (1993) signal transduction by the high affinity GM-CSF receptor: two distinct cytoplasmic regions of the common p subunit responsible for differentiation. EMBO J 12, 41814189. 3. Damell, J. E., Kerr, I. M., and Stark, G. R. (1994) Jak-STAT pathways and transcriptional activation m response to IFNs and other extracellular signalling proteins. Sczence 264, 14 15-142 1

/so/at/on of Cytokme-lnduclble

Genes

161

4 Ihle, J , Witthuhn, B. A , Quelle, F W , Yamamoto, K , Thierfelder, W E Kreider, B , and Silvennomen, 0 ( 1994) Signallmg by the cytokme receptor superfamily* JAKs and STATS Trends Bzochem Scz 19,222-227 5 Palacios, R and Stemmetz, M (1985) IL-3 dependent mouseclones that express B220 surface antigen, contam Ig genes m germ-line configuration, and generate B lymphocytes Cell 41,727-734 6 D’Andrea, A D., Lodtsh, H F , and Wong, G G (1989) Expresston cloning of the murme erythropoietm receptor. Cell 57,277 7. Carroll, M., Zhu, Y , and D’Andrea, A D. (1995) Erythropoietm-mduced cellular differentiation requires prolongation of the Gl phase of the cell cycle Proc Nat1

Acad Scz USA 92,2869-2873 8 Liboi, E , Carroll, M , D’Andrea, A D , and Mathey-Prevot, B (1993) The erythropoietm receptor signals both proliferation and erythroid-specific differentiation. Proc Nat1 Acad Scz USA 90,11,351-l 1,355 9 Liang, P and Pardee, A B (1992) Differential display of eukaryotic mRNA by means of the polymerase chain reaction Sczence 257,967-97 1. 10 Liang, P , Averboukh, L , and Pardee, A B (1993) Distribution and clonmg of eukaryotic mRNAs by means of differential display. refinements and optlmization Nucleic Aczds Res 21,3269-3275 11 Zhu, Y., Carroll, M., Papa, F R., Hochstrasser, M , and D’Andrea, A D (1996) DUB, a novel deubiqultmatmg enzyme wtth growth-suppressmg activity PNAS 93,3275-3279 12 Zhu, Y , Pless, M , Inhorn, R , Mathey-Prevot, B , and D’ Andrea, A D The murme DUB-l gene is specifically induced by the PC subunit of the mterleukn-3 (IL-3) receptor Mel Cell Bzol , in press 13 Kohno, H , Furukawa, T , Yoshmaga, T., Tokunaga, R , and Taketam, S (1993) Coproporphyrinogen oxidase. purrfication, molecular cloning, and mduction of mRNA during erythroid differentiation J Bzol Chem 268, 2 1,359-2 1,363 14 Hayzer, D J , Brmson, E., and Runge, M S (1992) A rat P-mterteron-induced mRNA* sequence characterrzation. Gene 117,277-278. 15 Lewm, A. R , Reid, L E , McMahon, M , Stark, G. R., and Kerr, I M. (1991) Molecular analysis of a human interferon-inducible gene family. Eur J Bzochem 199,4 17-423 16 McNeil, H. P., Austen, K F , Somerville, L L , Gurish, M F , and Stevens, R. L (1991) Molecular cloning of the mouse mast cell proteasegene J Bzol Chem 266, 20,316-20,322 17 Priefert, H. and Stembuchel, A (1992) Identification and molecular characterization of the acetyl coenzyme A synthetase gene (acoE) ofAlcalzgenes eutrophus J Bacterzol 174,6590-6599

Hormone-Inducible

Genes in Prostate Cells

Lidia Averboukh, Peng Liang, Stephen A. Douglas, and Arthur B. Pardee 1. Introduction Prostate gland development, growth, function and pathology are strongly influenced by steroid hormones, Androgens greatly mfluence the prostate, but the physiological maintenance of the gland can also depend on the presence of estrogens and other steroid hormones, retinoids and vitamin D (I) Steroid hormones transduce their biological effects via specific receptors that act as transcription factors. The androgen hgand-receptor complex, working as part of a transcription-activating complex, regulates the expression of a large subset of genes, including oncogenes (myc) (21, growth factors (TGF-a, TGF-P, and EGF) (3,4) and their receptors (EGF receptor) (3), apoptosis controllmg genes (M-2) (5), and other genes including those coding for proteases (prostate specific antigen) (6) and cell-cell adhesion molecules (E-cadherin) (7). This chapter discusses the identification of androgen-regulated genes m prostate cancer. An important process during the etiology of prostate cancer mvolves the transition from an early stage, when prostate cells grow m an androgen-dependent manner, to a later stage when cell growth becomes androgen-independent. The mechanisms of androgen-independent growth and the genes that are regulated under such conditions are largely unknown. The identification of genes whose functron or dysfunction is responsible or at least correlated with hormone insensitivity in advanced stage prostate cancers may help to elucidate the molecular mechanisms underlymg prostate cancer progression. In the past, steroid-regulated genes have been isolated by differential and subtractive hybridization techniques. Although both techmques are considered to be useful, they are time consuming and limited in their analytical power. From

Methods m Molecular Bology, Vol 85 Dlfferentral Edlted by P Llang and A B Pardee Humana

163


764

Averboukh eta/.

Neither technique allows compartson of gene expression under more than two different condittons or from more then two cell types The most recent method for tdentilicatton of such genes 1sthe differential display technique, which can be used to compare multiple RNA samples (8). Investtgatton of sterotd-regulated gene expression is an ideal application for differential display since addition of steroids causes only relatively few changes m gene expression, changes are relatively rapid and likely to occur at the transcrtptional level. Indeed, differential display has been applied to compare RNA from androgen-responsive prostate cancer cells with androgen-independent cell lmes (9) and prostate androgen-responsive cancer cells under various androgen conditions (I 0,Il). In one study, RNA isolated from normal adult Sprague-Dawley rat ventral prostates was compared wtth that isolated from castrated rats m the presence or absence of exogenous testosterone (10). One androgen-regulated cDNA, named RVP2, was identtfied and shown to be expressed only m normal rat ventral prostate and in prostate from castrated ammals treated with testosterone. A GenBank database search revealed that this cDNA fragment corresponds to the 3’ end of the rat prostattc spermme-bmdmg protein, a secretory protein, which has already been shown to be regulated by androgen in rats. Another study of androgen-regulated genes utilized the human androgenresponsive prostate cancer cell line LNCaP-FGC (2I). In this study cells were subjected to androgen-deprivatton conditions for different numbers of hours. RNA samples, collected under normal conditions (m the presence of androgen), were compared with those isolated at different time points followmg androgen-deprivation. Using this procedure, a gene encoding the S1OOPcalcium bmdmg protein was identified. The expression of this gene was abohshed after 30 h of androgen deprtvatton of LNCaP-FGC cells. This chapter describes an approach aimed at tdenttfymg androgen-regulated genes where LNCaP-FGC cells are deprived of androgen and then resttmulated by addition of synthetic androgen to the culture medium. This cell lme is one of the few prostate cell lines whose growth 1s responsive to androgen; androgen withdrawal greatly decreases the growth rate of these cells, addmon of the synthetic androgen R1881 (Methyltrtenolone) readily reverses this condition (12,13). Androgen deprivation is achieved by culturmg cells for 4 d with fetal bovine serum stripped of steroid hormones by dextran-charcoal treatment (24) mstead of regular serum. The cells are then restimulated by nonhydrolyzable synthetic androgen R188 1. RNA samples are collected at varying times up to 75 h. This approach enables identtfication of genes whose expression is controlled by androgens. Discovery of such genes will allow the development of new strategtes for prostate cancer therapy. These genes can also serve as potential markers for androgen-independent cell growth.

Hormone-Inducible

Genes

165

2. Materials and Equipment 1. 2. 3. 4. 5. 6 7. 8. 9. 10 11. 12. 13 14. 15. 16 17 18 19. 20. 21 22.

23. 24. 25. 26. 27 28.

LNCaP-FGC cell line (American Type Culture Collection, Rockville, MD). RPMI-1640 cell culture medium without phenol red (Mediatech, Herndon, VA), L-Glutamme (Sigma, St. Louis, MO). Fetal bovine serum (FBS) (Stgma). Dextran-coated charcoal (DCC)-treated fetal bovine serum (HyClone, Logan, UT). Synthetic androgen R188 1 (New England Nuclear, Boston, MA). RNAzol B, RNA isolation system (Biotecx, Houston, TX) Dtethylpyrocarbonate (DEPC)-treated dH20 Phenol:chloroform mixture (3: 1). 3M Sodium acetate (NaOAC) 100,85, and 70% Ethanol (EtOH). RNase-free DNase I (10 U/pL). 5X RT buffer: 125 mMTris-HCl, pH 8 3, 188 mA4KC1, 7.5 mA4MgC12, and 25 mA4 dithiothreitol (DTT) MMLV reverse transcriptase (100 U/uL). dNTP (250 @4). AAGCT,,G (2 @4) AAGCT,,C (2 w AAGCT,,A (2 l.&) Arbitrary 13-mers (2 @4). 10X PCR buffer 100 mM Tris-HCl, pH 8.4, 500 mA4 KCl, 15 mM MgCl,, and 0 01% gelatin. dNTP (25 mM). Glycogen (10 mg/mL) dH,O. DNA loading dye: 95% formamide, 10 mM EDTA, pH 8 0, 0.09% xylene cyanole, and 0.09% bromophenol blue. AmpliTaq DNA polymerase (5 U/a) (Perkin-Elmer Corporation). a[35S]dATP (>lOOO Ci/mrnol), or a[33P]dATP (>2000 Ci/mmol). Thermocycler Conventtonal DNA sequencer/Programmable GenomyxLR sequencer with independent temperature/voltage control and glass plates 62 x 33 cm.

All reagents can be purchased separately from different suppliers. Recommended suppliers: GenHunter Corporation (Nashville, TN) for Differential Display “RNAimage kit,” and Genomyx Corporation (Foster City, CA) for programmable DNA sequencer, sequencing gels and buffers. 3. Methods

3.1. Cell Culture 1. SeedLNCaP-FGCcellsat 2 x 1O4cells/cm2in RPMI- 1640medium,2 mM L-glutamine, 10% FBS, without phenol red.

Averboukh et al.

166

2. Grow cells under androgen deprivation condltlons for 4 d by replacing 10% regular FBS m the culture medium with 10% DCC-treated serum (see Note 1) 3. Perform the androgen mduction by adding synthetic androgen R188 1 (I I) dtssolved m ethanol to the medium containing 10% DCC-treated serum, to a final concentration of 1 nM (see Note 2) See Notes 3 and 4 for further suggested control experiments appropriate models to study androgen-responsive genes.

3.2. RNA Isolation

and advice on

and DNase I Treatment

1. Collect cellular RNA samples after 0, 1, 5, 10, 25, 50, and 75 h of synthetic androgen treatment 2 Purify RNA via the one-step acid-phenol extraction method using RNAzol B RNA isolation system (Btotecx, Houston, TX) 3. To remove contaminating chromosomal DNA m the RNA samples Incubate 50 pg oftotal cellular RNA with 10 U of DNase I (RNase free) m 10 mMTns-HCl, pH 8 3,50 mM KCl, 1 5 mMMgC12 for 30 mm at 37°C 4. Add an equal volume of phenol chloroform (3 1) to each sample, vortex, and incubate on ice for 10 min Centrifuge for 5 mm at 4°C m an Eppendorf centrifuge (high speed) 5 Transfer the supernatant to a new microcentrifuge tube Precipitate the RNA by adding 2.5 vol of 100% ethanol m the presence of 0 3M NaOAC and mcubatmg at -80°C for 30 mm Centrtfuge at 4°C for 10 mm to obtain the RNA pellet Rinse the pellet with 0 5 mL of 70% ethanol Dissolve the RNA m 10-20 pL of DEPC-treated H,O. 6 The RNA concentration is measured on a spectrophotometer at OD260,280 by diluting 1 pL of the RNA sample m 1 mL of HZ0 7. Check RNA integrity by runnmg l-3 pg of each sample (before and after DNase I treatment) on a 1 1% agarose gel containing 7% formaldehyde 8 Store the RNA sample at -80°C. (The DNase I treatment protocol is essentially as described m the protocol for MessageClean kit [GenHuter Corporation] )

3.3. Reverse

Transcription

of mRNA

1 For each RNA sample perform three reverse transcrtptton reactions m three microfuge tubes, each containing one of the three different anchored ollgo-dT primers (1.5). Each 20-pL reverse transcriptton reaction contams dH20 94cLL 5X RT buffer 4.0 pL dNTP (250 m 16$ Total RNA (DNA-free) 2.0 pL (0.1 l?lg/J.lL) AAGCT, ,M (2 clM> 2 0 pL (M stands for G, A, C) 2 Heat RNA samples at 65°C for 5 mm to lmeartze the template, then cool to 37°C for 10 min, then add 1 p.L MMLV reverse transcrlptase to each tube and continue the mcubation at 37°C for an addtttonal 50 mm

Hormone-Inducible

Genes

167

3 Inactive the enzyme by mcubatmg tubes for 5 mm at 75°C. 4 Keep the obtamed cDNA pools either on ice for the followmg PCR step or store them at -20°C

3.4. PCR 1 Each 20-r.lr, PCR reaction contains dH,O 1OX PCR buffer dNTP (25 luM) Arbitrary 13-mer (2 @4) AAGCT, ,M (2 /rM) RT-mix from step I

10.0 pL 2.0 pL lb@2olJ2oclr.

2 0 pL (use the same AAGCT, ,M for PCR as used m reverse transcription step) a[33P]-dATP (2000 Ci/mmol) 0.2 j.lL AmpliTaq (5 U/pL) (Perkm-Elmer) 0.2 pL 2 PCR conditions. 94°C for 30 s, 40°C for 2 mm, 72°C for 30 s (40 cycles), followed by extension at 72’C for 5 min (The Differential Display protocol IS essentially as described in the protocol for RNAimage kit [GenHunter Corporation].)

3.5. Denaturing

Polyacrylamide

Gel Electrophoresis

1 Prepare a 6% denaturing polyacrylamide gel m TBE buffer (conventional sequencer) or use a premade HR- 1000 6% denaturing gel if using the GenomyxLR Sequencer (16) A 4.5% gel can be used for better separation of cDNA fragments of 300 bp in size and greater. Conventional polyacrylamlde gel require at least 2 5 h for full polymerization and prerun for 0.5 h HR-1000 gels polymerize within 1 h and do not need to be prerun 2 Flush urea and gel debris out of the wells before loading the samples 3. Add 2 pL of loading dye to 4 pL of each sample, incubate at 80°C for 2 mm before loading. 4 Electrophorese for 3 5 h at 60 W constant power (not to exceed 1700 V) until the xylene dye reaches the bottom on a conventional sequencer (30-cm glass plate) or for 3 h at 100 W, 3000 V, 50°C on GenomyxLR sequencer (62-cm glass plate). 5 After the run, if using a conventional sequencer, blot the gel on a piece of Whatman 3 MM chromatography paper, cover the gel with a plastic wrap, and dry it at 80°C for 1 h (do not fix the gel with methanol/acetic acid) Gels run with a Genomyx sequencer can be directly dried on a glass plate after rmsmg the gel several times with water to remove urea. 6 Orient the autoradiogram and dried gel with radioactive mk or needle punches before exposing to X-ray film for 12-72 h. An example of a differenttal display gel performed on the LNCaP-FGC cells subjected to androgen stlmulatlon with an identified differentially expressed cDNA is shown m Fig 1

Averboukh

168

et al.

1234567

Fig. 1. LNCaP-FGC prostate cancer cells were subjected to 4 days of hormone deprivation followed by addition of synthetic androgen R1881 to the medium to the final concentration of 1 r&f. Total RNA samples were collected after 0, 1, 5, 10,25, 50, and 75 h of androgen treatment (lanes l-7 respectively). Differential display was performed with the following primers: H-T, ,G (5’ -AAGCTTTTT’TTTTTTG3’) and HAP-3 (5’ -AAGCTTTGGTCAG3’) (GenHunter), where H represents a Hind111 recognition site (AAGCTT). PCR conditions: 94°C for 30 s, 40°C for 2 min, 72°C for 30 s. DNA polymerase: AmpliTuq, 5 U/pL (Perkin-Elmer); Isotope: u[33P]dATP, 2000 Ci/mmol (New England Nuclear); PCR machine: GeneAmp PCR System 9600 (Perkin-Elmer). 6% HR- 1000TM denaturing gel (Genomyx) run on GenomyxLR DNA Sequencer (Genomyx). Run conditions: 5O”C, 3000 V, 100 W, 3 h. Arrowhead indicates differentially expressed cDNA fragment.

3.6. Reampllflcation of cDNA Fragment and Northern Blot Analysis 1. Align the autoradiogram with the gel. 2. Mark the bands of interest with a pencil underneath the film for a gel dried on 3MM paper. If the gel was dried on a glass plate, then place the X-ray film under the plate and mark the bands of interest with a pencil/marker. 3. Cut the bands with a clean razor blade if the gel was transferred on paper. Scrape a piece of gel with a slightly wet razor blade if the gel was dried directly on the glass plate. 4. Transfer the gel slice into a microcentifuge tube containing 100 pL dHzO and soak for 10 min.

Hormone-Inducible

169

Genes 1234567

18s

-

1234567 28s

-14,

*iii

x

m

9

Fig. 2. Northern blot analysis of differentially expressed cDNA (upper panel) in the LNCaP-FGC cells total RNA samples originally used for differential display. Lanes 1-7 are cells androgen treated for 0, 1, 5, 10,25, 50, and 75 h. Twenty micrograms of total RNA from each time-point were analyzed. Positions of 28s and 18s RNA are indicated. Ethidium bromide staining of total RNA samples as controls for equal loading is shown (lower panel). 5. Boil the tube containing the gel slice for 15 min. 6. Centrifuge for 2 min to pellet gel and paper debris. 7. Transfer the supernatant to a new tube. Precipitate DNA by adding 10 pL of 3M NaOAC, 5 pL of glycogen (10 mg/mL) and 450 pL+of 100% EtOH. Incubate for 30 min at -80°C. Centrifuge for 10 min at 4’C to pellet the DNA. Discard the supernatant and rinse the pellet with 200 pL cold 85% EtOH. Airdry the pellet and dissolve it in 10 pL of PCR HzO. Use 4 uL for reamplitication by PCR. 8. Reamplification is done using the same primer set and PCR conditions as for the initial differential display PCR step, except the final dNTP concentration is 20 pit4 (250 pA4 dNTP stock) instead of 2 @4 and no isotope is added. Each 40-pL PCR reamplitication reaction contains: dHzO 20.4 pL 1OX PCR buffer 4.0 pL dNTP (250 @4) 3.2 & Arbitrary 13-mer (2 ClM) 4.0 pL 4.0 pL AAGCTl ,M (2 l&) 4.0 pL cDNA template from RT 0.4 pL AmpliTaq (5 U1p.L)

170

Averboukh 12

4

-

28s 18s

3

et al.

-

1254

18s

-

Fig. 3. Northern blot analysis of differentially expressed cDNA (upper panel) in androgen-independent prostate cancer cell lines. PC 3 (lane 1 and 2) or DU 145 (lane 3 and 4) cells grown either in the presence of 10% FBS (lanes 1 and 3) or in the presence of 10% DCC-treated serum for 48 h (lanes 2 and 4). Twenty micrograms of total RNA from each time-point were analyzed. Positions of 28s and 18s RNA are indicated. Ethidium bromide staining of total RNA samples as controls for equal loading is shown (lower panel).

9. To check the reamplification of the cDNA fragments run 30 pL of the PCR sample on a 1.5% agarose gel stained with ethidium bromide. 10. Extract the reamplified cDNA probe from the agarose gel using a QIAEX kit (QIAGEN). Il. Save the remaining 10 pL of PCR product at -2O’C for subcloning. 12. Perform Northern blot analysis or RNase protection assay to confirm differential expression of the cDNA (27). Northern blot confirmation of the differentially expressed cDNA (see Fig. 1) in LNCaP cells is shown in Fig. 2. 13. Clone the cDNA fragment using a standard PCR product cloning system such as the pCR-TRAPTM (GenHunter Corporation) or the TA cloning kit (Invitrogen). 14. After a cDNA fragment is cloned, its differential expression should be again reconfirmed by Northern blot and sequenced. Sequence information can be used to search DNA databases. A cDNA library can be screened to obtain a full-length cDNA.

4. Notes 1. Dextran-coated charcoal-treated fetal bovine serum can be either prepared following the standard protocols (14) or can be purchased from HyClone.

Hormone-lnduable

Genes

171

Commercially avallable DCC serum IS supplied with a quality control sheet which details exactly how much of testosterone has been removed. Usually the concentration of the testosterone 1s reduced by an order of magnitude (from l-0.1 nM) (12) 2 The most popular synthetic androgen IS R 188 1, which is not metabolized by cells (12) It 1s commercially avallable from New England Nuclear (Boston, MA) Other androgens used for cell culture include naturally occurring hormones Sa-dlhydrotestosterone (DHT) or testosterone (T) available from Sigma 3 If androgen-responsive prostate cell lines are used, it 1simportant to check how well the absence of androgen slows the growth of the cells If androgens are used, their optimal cell growth stimulatory concentration should be determined 4. In addition to the androgen-responsive human prostate cancer cell lme LNCaP there are androgen-unresponsive human prostate cancer cell lines such as DU 145 and PC 3. Northern blot analysis of cDNA expression identified in LNCaP cells (see Fig. 1) m androgen-independent human prostate cancer cell lines is shown m Fig. 3 Most frequently used ammal models Include the Dunning rat androgen-responsive and unresponsive tumors (18) and castrated animals (10).

Acknowledgment The authors would like to thank Christoph the manuscript.

M Ahlers for careful review of

References 1 Wilding, G (1995) Endocrine control of prostate cancer Cancer Surveys 23,43-62 2. Kokontls, J , Takakura, K , Hay, N , and Llao, S. (1994) Increased androgen receptor activity and altered c-myc expression in prostate cancer cells after longterm androgen deprlvatlon Cancer Res 54, 1566-1573 3. Lm, X.-H , Wiley, H S , and Melkle, A W. (1993) Androgen regulates prollferation of human prostate cancer cells in culture by increasmg transformmg growth factor-a (TGF-a) and epldermal growth factor (EGF)/TGF-c~ receptor. J Clan. Endo Met. 77, 1472-1478 4 Knabbe, C , Klein, H., Zugmaier, G , and Voigt, K. D (1993) Hormonal regulation of transformmg growth factor p-2 expression m human prostate cancer. J Steroid Blochem A401 B~ol 47, 137-142 5. McDonnell, T J., Troncoso, P., Brisbay, S. M , Logothetis, C , Chung, L W. K , Hsleh, J.-T., Tu, S -M , and Campbell, M L. (1992) Expression of the protooncogene bcl-2 m the prostate and its association with emergence of androgenindependent prostate cancer. Cancer Res 52,6940-6944 6 Montgomery, B. T , Young C Y.-F , Bilhartz, D. L , Andrews, P E , Prescott, J L , Thompson, N F , and Tmdall D. J (1992) Hormonal regulation of prostatespecific antigen (PSA) glycoprotem m the human prostatic adenocarcmoma cell line, LNCaP Prostate 21,63-73. 7 Carruba, G., Miceh, D’Amico, D., Farruggio, R , Comlto, L., Montesantl, M , Polito, L., and Castagnetta, L A. M. (1995) Sex steroids up-regulate E-cadherm

172

8 9.

10.

11 12 13.

14

15

16. 17. 18

Averboukh

et al.

expression m hormone-responsive LNCaP human prostate cancer cells Bzochem Blophys Res Commun 212,624-63 1 Ltang, P and Pardee, A B (1992) Differential display of eukaryotic messenger RNA by means of polymerase chain reaction. Sczence 257,967-97 1 Blok, L J., Kumar, M. V , and Tmdall, D. J (1995) Isolation of cDNAs that are dtfferentially expressed between androgen-dependent and androgen-independent prostate carcinoma cells using differential dtsplay Prostate 26,2 13-224. Chapman, M. S., Qu, N , Pascoe, S , Chen, W.-X., Apostol, C , Gordon, D , and Miesfeld, R L (1995) Isolation of differentially expressed sequence tags from steroid-responstve cells using mRNA dtfferenttal display Mol Cell Endocrznol 108, RI-R7 Averboukh, L., Liang, P., Kantoff, P W , and Pardee, A. B (1996) Regulation of SlOOP expression by androgen. Prostate 29,350-355 Bonne, C. and Raynaud, J.-P (1975) Methyltrienolone, a specific hgand for cellular androgen receptors. Steroids 26,227-232. van Steenbrugge, G. J., van Uffelen, C. J. C., Bolt, J , and Schroder, F. H. (1991) The human prostate cancer cell lme LNCaP and Its derived sublines. an zn vztro model for the study of androgen sensitivity J Steroid Blochem Mol B1o1 40, 207-2 14. ZaJchowski, D A and Sager, R (1991) Induction of estrogen-regulated genes differs m immortal and tumorigenic human mammary epithelial cells expressing a recombinant estrogen receptor. Mol Endo 5, 16 13-l 623 Liang, P., Zhu, W., Zhang, X., Guo, Z., O’Connell, R P , Averboukh, L., Wang, F , and Pardee A B. (1994) Differential display using one-base anchored ohgodT primers. Nucleic Aczds Res 22, 5763-5764 Averboukh, L., Douglas, S. A , Zhao, S., Lowe, K , Maher, J., and Pardee, A B (1996) Better gel resolution and longer cDNAs increase the prectsion of differential display. Bzotechniques 20,918-92 1. Ausubel, F., Brent, R., Kingston, R. E., Moore, D. D , Seidman, J. G., Smtth, J A, and Struhl, K. (1988) Current Protocols zn Molecular Bzology, Green and Wiley-Interscience, New York. Isaac& J. T , Isaacs W B , Feitz, W. F J., and Scheres, J (1986) Estabhshment and charactertzatton of seven Dunning rat prostatic cancer cell lines and their use in developing methods for predtctmg metastattc abthties of prostattc cancers. Prostate 9, 26 l-28 1,

15 Application of Differential Display in Studying Developmental Processes M. Jorge Guimaraes,

Terrill McClanahan,

and J. Fernando

Bazan

1. Introduction Developmental processes are intrtcate networks of events by which a fertilized egg proliferates and differentiates mto a fully functional multicellular orgamsm (1). Mechamsms are used by dividmg cells to create asymmetries m then mformation content that then allow the emergence of heterogeneous cell types through dtfferenttal gene expression (2,3). Most developmental processes share the following characteristics: 1. Consecutrvestepsof commrtmentand specialization are taken by dividmg cells, 2 Time- and tissue-regulatton of gene expressron; and 3 Migration of specialized cells from their sues of origin to different locations

Differential display has a number of characteristrcs surtable for its successful application to studying developmental processes (4-7): 1. It allows simultaneous comparison of gene expression patterns among multiple mRNA populations, 2. Requires minute amounts of starting total RNA material; 3 Is gene class independent, and 4. Detects genes expressed at both high and low levels, represented by both short and long mRNA transcripts.

Hematopoiesis, or blood cell formation, IS one example of a developmental process. Hematopoiesis IS first detected in the murme yolk sac at d 7.5 of gestation. Following activation of cnculation at the 7 somtte stage (d 8.0 of gestation), hematopoiesrs ISconsecutrvely detected m the AGM (aorta-gonadamesonephros) region (8), liver, spleen, and, finally, bone marrow (9) Srmrlarly to cells found m eptdermrs, mtestme, and testis, a small mmority of From

Methods III Molecular B/o/ogy, Vol 85 LNferentral Edlted by P Llang and A !3 Pardee Humana

175


176

GuimarZies, McClanahan, and Bazan

hematoporettc cells show self-renewal (“stem”) properties (IO). These so-called hematopotetic stem cells (HSCs) are capable of extensive prohferatton, generating progressively more differentiated cells that eventually become erythrocytes, neutrophils, monocytes, basophrls, eosmophils, lymphocytes, or platelets, while maintaining a restricted population of undifferentiated, plurtpotent, stem cells ES cells are derived from the inner cell mass of blastocysts and appear to resemble the prtmrttve ectoderm of the postimplantatton embryo (11-13). These cells have the capacity to participate m the for-matron of an embryo and to form functional germ cells when injected into a blastocyst, generating germhne chimeras (14). Culture systems of ES cells that allow their differentiation m vitro mto embryotd bodies (EBs) contammg hematopoiettc activity have been described (1.5-23) and plurrpotent hematopoiettc cell lines that can be grown m vitro have also been developed (24,25) A number of genes encoding molecules involved in different steps of signalmg transduction pathways have been identified that are expressed m and/or active on hematoporettc progemtors (26). However, almost nothing 1sknown about the molecular events responsible for the formatron of HSCs from undrfferentiated mesoderm precursors, hematopoietrc commrtment, or on the molecular nature of their self-renewal mechanisms. Here we describe the successfulapphcatron of differential display to searching for genes potentially involved in hematopoietic commrtment and hematopoiesrs. Our approach consisted of comparmg geneexpression patterns among 12 samples chosen to represent different levels of hematopoietic activity. Namely, we used the following RNA populatrons dertved from the mouse embryo, EBs and cell lines chosen to represent different levels of hematopotetic activity (27): 1 Samplesrepresentinghematoporettcactrvrty* a d-8 5 yolk sac, b. Early d-8.5 embryo proper (d-8 5 embryoswith 7 or more somites); c Late d-8 5 embryo proper (d-8.5 embryoswith srgnsof active circulatron); d. d-6 EBs; e. d-9 EBs, f. Plurrpotent hematopotetrccell hne FDCPmrxA4 2 Sampleswith no hematoporeticactivity: a. ES cells; b Headprrmordium of d 8 0 embryos(3-6 somrtestageembryos,at which stage circulation 1snot yet active); c. Ftbroblast cell line STO; d Neuronal cell lme N2a 3 Samplesthat may have hematopolettcpotential: a. d-3 EBs; b Posterior regron of the 3-6 somite stageembryo proper.

Differential

Display

and Developmental

Processes

We searched for genes which would agree to all of the followmg

177 crrterta:

1 Preferential expression m the d-8 5 yolk sac relative to the d-8.5 embryo proper; 2. Not expressed m the head prtmordmm but possibly expressed m the posterior region of d-S.0 embryos; 3 Not expressed in ES cells and upregulated m d-3 or d-6 EBs; 4. Expressed m the pluripotent hematopotetic cell line FDCPmixA4 but not m the neuronal cell lme N2a or m the fibroblast cell line ST0 The application of differential display to the simultaneous comparrson of all these RNA populattons illustrates the usefulness of this technique to studying complex brologtcal questions mvolvmg differenttal gene expresston (7,27-31). This approach can be used to msptre similar models to study other developmental systems (2 7).

2. Materials 2.1. Animals, 1. 2. 3 4. 5 6. 7 8 9 10. 11 12 13. 14 15. 16 17 18. 19 20 2 1. 22

Cells, and Reagents

Timed pregnant mice (e g , ICR female) Binocular dtssectmg microscope with incident hght sources. 37”C, 5% CO* humidified incubator 70-pm Nylon cell strainer (Becton Dtckmson, Franklin Lakes, NJ). ES cells (e g , ES cell line CCE [15/) ST0 cells. American 7’ype Cell Culture-ATCC CRL 1503 N2a cells American Type Cell CultureATCC CC1 13 1 FDCPmixA4 cells (I Z) Tissue culture dishes (e.g , FALCON, Becton Dickinson Labware, Lincoln Park, NJ). Bacterial Petri dishes (Baxter, McGraw Park, IL). Phosphate buffered salme (PBS) (Gtbco-BRL, Grand Island, NY). 0 25% Trypsin in PBS with 1 mA4IL EDTA (e g , Gibco-BRL) 4 mol/L L-glutamme (1000X) (JRH Btoscrences, Lenexa, KS). 100 U/L pemcillm and 100 mg/L streptomycm (1000X) (Gibco-BRL). Dulbecco’s modified Eagle’s media (DMEM) (Glbco-BRL). Iscove’s medium (Gibco-BRL) Horse serum (Gtbco-BRL) Fetal calf serum (FCS) (Sigma, St. LOUIS, MO) Fetal calf serum (FCS) (Gemini Bioproducts, Calabasas, CA) (see Note 1) 2-mercaptoethanol(2-ME) (Sigma). Mouse IL-3 (DNAX Research Institute, Palo Alto, CA) 0.1% Gelatin (Sigma) coated tissue culture dishes (plate and leave at room temperature for 30 min; suck off gelatin and store at room temperature).

2.2. Culture Media 1. ES medium* 80% DME + 20% FCS (Gemini Btoproducts) (see Note l), supplemented with 4 m&i/L t.-glutamme, 100 U/mL penicillin, 100 pg/mL streptomycin,

178

Guimara’es, McClanahan, and Bazan

0 I &IL 2-ME plus 0.1% (v/v) condrtroned medium from cos-7 cells transrently transfected with pSRaLIF human leukemia mhibnory factor (LIF) cDNA Store at 4°C and use for 2 wk. 2. EBs medium. same as ES media but without LIF Store at 4°C and use for 2 wk 3. N2a and ST0 media: DMEM + 10% FCS (Sigma), 1% L-glutamme, 100 U/mL penicillin, and 100 mg/mL streptomycin Store at 4°C and use for 2 wk 4. FDCPmixA4 cells media. Iscove medium + 20% Horse serum, 1% L-glutamme, 100 U/mL pemcillm, and 100 mg/mL streptomycm, 10 rig/L Mouse IL-3. Store at 4°C for up to 2 wk

2.3. RNA Extraction

7 8 9 10 11 12

Homogenizer (e g., Polytron PT-MR 3000, Kinematica AG, Littau, Switzerland). RNAzol solution (Tel-test, Inc, Friendswood, TX) Glycogen (10 mg/mL) Isopropanol (Fisher Scientific, Fair Lawn, NJ) 75 and 100% ethanol 3.1 Phenol (Amresco, Solon, Ohio) Chloroform (Mallmckrodt Specialty Chemicals Co , Paris, Kentucky) Keep protected from hght Store at 4°C. Prepare a fresh mixture after approx 1 mo RNase free water (BioWhittaker, Walkersville, MD) Plugged pipet tips (e.g., ART, Molecular Bio-products, San Diego, CA). Formamide (Sigma) Formaldehyde (Sigma); 37% solution (Formalm) 10X MOPS-EDTA-Sodium acetate buffer (Sigma) Ethidium bromide (10 mg/mL) (Sigma).

2.4. Differential 1 2. 3 4 5 6 7 8. 9. 10 11 12. 13

Display

MessageClean Kit (GenHunter Corporation, Nashville, TN) RNAmap kit (GenHunter Corporation) AmpliTuq DNA polymerase (Perkm Elmer-Cetus, Norwalk, CT). [35S]dATP (Amersham, Arlington Heights, IL) Paraffin oil (Sigma). Perkm Elmer-Cetus thermocycler (see Note 2) PCR tubes (PGC Scientific, Frederick, MD) Gel-Mix 6 solution (Gtbco-BRL) for preparing polyacrylamide gels 1% ammomum persulfate (Gibco-BRL) Store the solution at 4°C for up to one week Ultra pure TBE (100 mMTris, 90 mMBoric Acid, 1 0 mMEDTA) (Gtbco-BRL) Square-toothed 32-well combs for sequencing gels (Gibco-BRL) (see Note 3). Sequencing gel apparatus model S2 (Gibco-BRL) DNA Marker for molecular weights ~1 kb (e.g., DNA Marker V, Boehrmger Mannheim) 14. Chemrlummescent markers (Stratagene). 15. Kodak O-mat film (Eastman Kodak, Rochester, NY) 16 3MM Whatman paper (Whatman International, Maidstone, UK)

Differential D/splay and Developmental 2.5. Full-Length 1 2 3 4 5 6 7. 8. 9 10 11 12 13 14 15. 16. 17.

Processes

179

cDNA Cloning

3% NuSieve 3 1 Agarose (FMC Bioproducts, Rockland, ME). Gel extraction kit (e g , QIAEX II extraction kit, QIAGEN, Chatsworth, CA) TA Clomng kit (Invnrogen, San Diego, CA) DNA mmtprep kit (e g , RPM kit, BIO 101, Vista, CA) TE (0 OlMTrts, 0 OOlMEDTA Na2 buffer) with 50 pg/mL RNase (DNase free) Appropriate restriction enzymes and buffers 70750 Reagent Kit For Sequencmg With Sequenase T7 DNA Polymerase and 7-deaza-dGTP (Amersham. Cleveland, 0) lMNaOH, 1 mA4EDTA 2MNH40AC, pH 5 4 mRNA extraction kit (e.g., Oligotex-dT mRNA kit, QIAGEN) Superscript Plasmid System (Gibco-BRL) NotISaZI Arms (Gibco-BRL). h Packagmg System (Gibco-BRL) Hybridization filters (e.g , Hybond-N+, Amersham) Hybrtdtzatton tubes (e g , Robbms Sctenttfic, Sunnyvale, CA) Hybridization oven (e g , Robbms Scientific) Sequencing trays (Stratagene).

2.6. Northern

Blot Analysis

1 2 3 4 5 6. 7

QIAGEN Plasmid Maxi Kit (QIAGEN) QIAEX II gel extraction kit (QIAGEN) 32P-dCTP (Amersham) Prime-It II kit (Stratagene, La Jolla, CA) Nytran membranes (Schleicher & Schuell, Keene, NH) 20X SSC (3 OM NaCl, 0 3M sodium citrate) (Boehrmger Mannheim). Hybridization solution, 0 5M NaHP04, pH 7 2, 7% sodium dodecyl sulfate (SDS), 0 5M EDTA, pH 8 0 Store at room temperature. 8 Wash solutton 1: 2X SSC, 0.1% SDS. 9 Wash solution 2. 0 2X SSC, 0 1% SDS 10 Phosphorimager (Molecular Dynamics, Sunnyvale, CA)

3. Methods 3.1. Cell Culture 1. Keep ES cells in gelatin-coated tissue culture dishes m ES media. Change medium on the second day after passage. Pass cells every 3 d by trypsmization m 0.25% trypsm m PBS with 1 mmol/L EDTA. Disperse the trypsinized cells by ptpetmg up and down and dilute mto medmm. Centrifuge at 1000 rpm for 5 mm and resuspend m fresh medium at 2 x lo6 cells per 60-mm dish. 2. For embryoid body preparation, trypsmize ES cells as above for routme passage, centrifuge for 5 min at 1000 rpm, and resuspend m ES medium without LIF at 1 to 1 5 x lo5 cells/ml m 10 mL m lo-cm bacterial Petri dishes This will generate

180

Guimaraes, McClanahan, and Bazan

300 or 400 embryold bodies per dish. Maintain the embryoid bodies in a humidified 5% CO, atmosphere at 37°C for up to 25 d. Feed the cultures every 3 d by allowing embryoid bodies to settle m a tube, replace medium, and gently pipet with a wide-bore pipet mto fresh Petri dishes 3 Trypsinize ST0 and N2a cells for passage but not FDCPmixA4, which are nonadherent cells 4 The morning of the d when the vaginal plug is found is counted as d 0 Dissect embryomc tissues as quickly as posstble keeping all materials m PBS + 5% FCS on ice Samples are collected by centrifugation and, if necessary, pellets are quickly frozen m dry ice for 10 mm before storage at -80°C 5. To obtam bone marrow cells flush the interior of mouse femurs with PBS + 5% FCS usmg a 2 l-gage syringe. Pass the cells though a 70-pm nylon cell strainer to separate debris and bone fragments, centrifuge at 1000 rpm and store the pellet, if necessary, at -8O’C

3.2. RNA Extraction 1 Remove frozen samples containmg embryonic tissues or cell pellets from -80°C and quickly add 10 mL RNAzol B solution (2 mL/lOO mg tissue or 0.2 mL/lO cells) or add the solution directly to centrifuged pellets or adherent cells. Homogenize the embryonic tissues Solubihze all lysates by passing them through a ptpet Shred all lysates using a 16-gage followed by a 21-gage syringe. Add 0.2 mL chloroform per 2 mL lysate, shake vigorously for 30 s, and let sit on me for 15 mm. Centrifuge at 12,OOOg (4°C) for 15 mm To avoid genomtc DNA contammation remove only the top 2/3 of the upper phase for isopropanol precipitation. If you expect a low RNA yteld, use 1.5 $ glycogen (10 mg/mL) as carrier before adding l/l vol isopropanol Mix well and store at -2O’C for 1 h or longer. Spin for 20 mm at 12,000 g (4°C). Wash the pellet with 75% ethanol by ptpetmg up and down and centrifuge at 12,000g for 10 mm Spm again to remove residual hquid, au-dry the pellet and dissolve it in RNase free water Store the RNA samples at -80°C. 2 DNase treat all samples using the MessageClean Kit Add m order: RNA sample and water, 5 7 & 1OX reaction buffer and 1 @ DNase I (10 U/l.tL) to a final volume of 57 pL Mix well pipetmg up and down and incubate at 37°C for 30 mm. Add 40 mL PhenolCHCl3, vortex for 30 s and let sit on ice for 10 mm. Spur at 14,000 rpm and remove supernatant. Ethanol precipitate using glycogen as carrier add 1 0 mL glycogen, l/l2 (v/v) 3M sodium acetate and 2 5 vol 100% ethanol, mix well and place at -70°C for 1 h or longer. Centrifuge at 7500g (4°C) for 20 min. Wash the pellet with 75% ethanol. Centrifuge as above. Au-dry the pellet and resuspend m water at >O 1 pg/pL for accurate quantitatton 3 Carefully quantitate all RNA samples by absorbance at 260 nm (see Note 4) Prepare a formaldehyde gel to evaluate the quahty of the RNA samples before proceedmg to cDNA synthesis: add 1.2 g agarose, 85 mL water, and 10 mL 10X MOPS buffer. Boil to dissolve in a microwave. Cool to 60°C. Add 5.5 mL formaldehyde m a hood. Mix well Add 2 5 pL ethidmm bromide Mix well Pour m

Differential Display and Developmental

Processes

181

castmg tray and leave m a hood at room temperature for 1 h. Prepare the sample by mixing first 10 pL formamtde, 3.5 pL formaldehyde and 2 clr, 10X MOPS buffer. Mix well and add 4.5 p.L sample (>l pg total RNA, if possible) Mtx well, leave at 55°C for 15 min, and chill on me Add 2 pL loadmg dye. Mix well and apply to the gel. Run at 80 V until the front dye reaches 2/3 of the gel. Assess quality of RNA (see Note 5) Make small 0.1 pg/pL RNA ahquots for cDNA synthesis. Store all RNA samples at -80°C.

3.3. Differential

Display

1. Make a duplicate reverse transcription reactton for each RNA sample (see Note 6 and Fig. 1). Add m order. 9.4 pL water, 4 mL 5X reactton buffer, 1 6 pL 250 mM dNTPs, 2 pL total RNA (0.1 p&L), 2 & T,*MN. Program the thermocycler to 65°C for 5 mm, 37°C for 70 mm, 95‘C for 5 min, soak at 4°C. Add 1 pL MMLV reverse transcriptase to each tube after 10 mm at 37°C and mix well by quickly stirring with the pipet tip and pipeting up and down Ten extra minutes have been dedicated to thrs step when programmmg the thermocycler but tt 1s better to do only a few samples at a time Use a representative sample to which no enzyme was added as control for genomtc DNA contammation. Store the cDNAs at -20°C. 2. Use both cDNA duphcates for PCR. Conduct all preparations on ice using the RNAmap ktt and be careful to avoid contammation. Use core mixes. Add per reaction. 9.2 Ccs,water, 2 pL 10X PCR buffer, 1.6 pL dNTP (25 CIM), 2 mL APprimer (2 l&!), 2 pL T&N (10 CIM), 2 mL RT-mix from step 1,l mL 35S-dATP (1200 Ct/mmol), 0 2 p.L AmpliTaq (Perkin-Elmer). Use a reaction m which cDNA is replaced by water as control for contamination of the reagents Mix the reactions well by pipeting up and down, add 40 & of mineral oil if necessary Program the thermocycler to 94°C for 30 s, 40°C for 2 min, and 72” C for 30 s for 40 cycles (see Note 7 and Fig. 2). Introduce the tubes in the thermocycler after prewarming the block to 94°C (see Note 8 and Fig. 2). 3. Prepare a radtolabeled DNA ladder using T4 DNA Polymerase: add 1 p.L 10X universal buffer (lMKOac, 250 mMtru+acetate pH 7.6, 100 mMMgOAc, 5 mM 2-mercaptoethanol, 100 Clg/mL BSA), 10 pL DNA marker V, 7 p.L water, 2 pL T4 DNA polymerase. Place at 37’C for 2 min. Chill on ice. Add 1.8 pL 0.1 mM dCTP, 5 pL 2 mA4 dATP, 5 & 2 mA4 dTTP, 5 pL 2 mA4 dGTP, 1.5 mL 10X universal buffer, 5 mL a32-PdCTP (3000 Ci/mmol), 6.7 pI. water. Leave at 37°C for 30 min. Separate the counts using a G-50 spm column or equivalent. Use 2000 counts per gel run 4. Prepare a 6% polyacrylamide gel (see Note 3). Set the power source for 60 W constant power and prewarm the gel to 50°C. Use the PCR samples derived from duplicate cDNAs to perform two gel runs per primer combination m one gel The shorter run excludes PCR products smaller than 50 bp, whtle the longer run can be used to increase the resolution of PCR products m the size range of 250-600 bp and confirm the reproducibility of the results (Fig. 1). Take 3 5 pL of each sample (including genomic DNA contamination control, PCR reagents contaminatton control, and radiolabeled DNA ladder) and add 2.0 PI of loadmg dye. Incubate

I

3~,: ..a

‘_

Fig. 1. Duplicate PCR using cDNAs generated from independent reactions were displayed using two gel runs. The T12MC/AP-4 primer combination was used. The arrow indicates a differentially expressed band. The arrow heads on the duplicate reaction run on the right indicate bands that were both found to represent the same cDNA sequence displayed on the left run (see Fig. 3).

Differential

Display and Developmental

Processes

183

Fig. 2. The pattern of bands given by each primer combination depends on the concentration of the pruners, the annealing temperature and the use of a cold or hot start (6). (A) Band seen preferentially using an annealing temperature of 4O’C; (B) band preferentially amplified using a 32°C annealing temperature; (C) band preferentially observed using a cold start and 32°C annealing temperature; (D) band seen under all conditions tested. PCR tubes were placed in the therrnocycler without prewarming the block (cold start) or after prewarming to 94OC (hot start). 32 and 40°C annealing temperatures. SNKIT- AP-primer synthesized for this experiment to match the AP-primer provided (SKIT) at 2 pM.

Guimades,

184

McClanahan, and Bazan

at 80°C for 2 mm Spin briefly to bring down condensation and load the gel after carefully flushing the wells to remove urea. Run at 60 W until the xylene cyan01 dye IS about 8 cm (short run) to 14 cm (long run) from the bottom. Dry the gel without fixation at 80°C for 80 min. Film and gel are oriented by punching the margins or using chemilummescent markers. Expose the gel to Kodak O-mat film at room temperature for 2-5 d or unttl the desired exposure is obtained Gels can be stored m between clean Whatman paper for longer than one year at room temperature. 5 After identifying differentially expressed bands (see Note 9), dectde if you need further separation of your band(s) of interest before extraction from the gel. This can be achieved making long electrophoretic runs of the desired sample side by side with a control. Use the chemiluminescent markers to orient film and dried gel and punch with a needle around the band of interest Use new razors to extract each different band In a screw-cap tube, soak the gel and paper slice in water (100 pL) for 15 mm Boil the tube for 15 min Spin for 2 min to collect condensation Transfer the supernatant to a new tube and ethanol precipitate using 5 pL glycogen. Dissolve the pellet in 10 $ of water and use 4 pL for reamphfication using the same set of primers and PCR conditions except using 250 pA4 dNTP stock. For a 40-pL final volume reaction, use. 20.4 pL water, 4 pL 10X reaction buffer, 3.2 pL dNTP (250 )1M), 4 0 pL AP-primer, 4 pL T&IN pnmer, 4 pL DNA, 0.4 pL AmpliTaq (see Note 10). Run the sample in a 3% NuSieve 3:l agarose gel. Check size of the PCR product and gel extract for cloning (see Note 11)

3.4. Full-Length

cDNA Cloning

1. Clone the gel extracted PCR products usmg a 3.1 insert:pCR II vector molar ratio. Ligate overnight at 14’C using T4 DNA hgase m a 1 I-& vol reactton. 2 To exclude the possibility of more than one PCR product being represented as a single band in the polyacrylamide gel, sequence three independent clones derived from each polyacrylamide gel slice using the 70750 Reagent Kit for sequencmg with sequenase T7 DNA polymerase and 7-deaza-dGTP (see Notes 12 and 13 and Fig. 3). use 2 pg of DNA and add 4 pL of IMNaOH; 1 mA4 EDTA and brmg up with water to total volume of 20 pL. Incubate at room temperature for 5 min Neutralize by adding 2 p.L of 2M NH40Ac, adjusted to pH 5.4 with acettc acid. Place on ice quickly and add 60 pL of 100% ethanol Precipttate and wash the pellet with 70% ethanol. Air-dry the pellet and use immediately. Resuspend pellet in 7 pL water. Add 2 pL of sequencing buffer. Add 1 pL of primer (2-5 ng/pL). Heat at 65°C for 2 min and then allow the tube to cool down slowly to room temperature (15-30 min). To the annealed template-primer add: 3 mL 1:5 diluted labeling mix (2 p.L concentrated labelmg mix, 8 pL water, 5 & DTT), 1 pL 35S-dATP and 2 pL 1.8 diluted sequenase in sequenase dilution buffer. Incubate at room temperature for 5-10 mm. Label four tubes or use sequencing trays. Place 2.5 pL of the ddGTP termination mix in the labeled tubes or wells. Add 3.5 pL of the labeling reaction to each and leave at 37°C for a minimum of 5 mm and up to 30 mm. Add 4 pL of stop solution. Run on a sequencing gel after heating samples to 75°C for 2 mm.

Differential


Processes

185

Fig. 3. Here we document that multiple bands (arrows) can result from the amplification of a cDNA by a single primer in a reproducible manner. These bands were resolved by long electrophoretic runs and found to represent the same cDNA sequence amplified by the T,,MC primer (see Note 15). YS, d-8.5 yolk sac; d3- d-3 EBs; FDCP, FDCPmixA4 cell line; STO, cell line STP; ES-ES cells. C/API, T12MC/AP1 primer combination; ClAP2, T,zMCIAPIL primer combination; ClAP3, T,,MC/AP3 primer combination.

186


3. To construct a cDNA ltbrary, use 5 pg of poly(A) RNA using the Superscrtpt Plasmrd System for cDNA synthesis Ligate h No&Sal1 Arms to the cDNA and package using the h Packaging System 4 Plate the library at a density of 30,000 plaque forming units (pjii) per plate and screen using the probe dertved from the cloned PCR products Hybridize HybondN+ filters at 65°C m 0 SMNaHPO, pH 7 2,7% SDS, and 0 5MEDTA pH 8 0 for 24 h, then wash m 2X SSC 0 1% SDS at 65°C for 30 mm with one change of buffer, followed by a wash m 0 2X SSC, 0 1% SDS at 65°C for 30 mm with one change of buffer, and expose to film at -80°C (see Note 14) This type of library allows the convenient subclonmg of mserts mto plasmtds using an mfectton step with DH 1OB cells grown m amptcrllm plates

3.5. Northern

Blot Analysis

1 Make a large preparation of plasmid DNA and cut the plasmid with appropriate restrictron enzymes to obtain the Insert Run a 1% agarose gel and gel extract the Insert Random prime 200 ng of insert using 32PdCTP Use 1 pL of the T,,MN primer used for PCR m addmon to the random primers provided Store the probe at -20°C m a proper container for up to 10 d. 2 Wet the membrane to be hybrtdtzed m water Use a 5-mL prpet to role the blot around it and introduce the blot mto the hybrtdtzatton tube Take off all excesstve water. Prehybrtdtze the blot for 1 h in hybrtdrzatton solution at 65°C Aliquot mto a screw-cap Eppendorf tube 1 x IO6 counts of probe per mL of hybridtzatron buffer. Boil the probe for 5 mm and put tt on ice Add the probe to the hybrtdtzanon tubes and leave overmght at 65°C 3. Poor off the hybrrdtzatton solution and wash at 65°C m 2X SSC, 0.1% SDS for 15 + 15 mm wtth one change of buffer Next, wash at 65’C m 0 2X SSC, 0 1% SDS for 15 + 15 mm with one change of buffer 4 Expose the blot to film or to a Phosphortmager, rf possible (see Note 17)

3.6. Example of Use: A Search for Genes Involved in Hema topoie tic Development 1 To approach hematoporetrc development at the molecular level we explored the appbcatron of dtfferenttal dtsplay to analyzmg gene expression patterns among a vartety of RNA populations After uttltzmg 20 prtmer combmatrons (representing 1000 bands or l/15 of the estimated 15,000 mRNA repertoire expressed on each cell at any given time), we observed a number of differentially expressed bands (e.g., Fig 1) preferentially expressed m the yolk sac, which also complred with criteria 2 and 3 of our selection method (see Section 1 ) It became clear, however, that we could not identify bands preferentially expressed m FDCPmrxA4 that did not have some level of expression m ST0 and N2a; in other words, bands preferentially expressed m the yolk sac, upregulated m d-3 to d-6 EBs and preferentially expressed m FDCPmtxA4 were also expressed m ST0 and, m some cases, also m N2a. We decided to study thts group of bands (Group 1 of genes m Table 1) In addmon, we selected bands that were only detectable m the yolk sac

Dlfferentlal D/splay and Developmental Table 1 Genes Identified of the Estimated

Processes

After Exploring l/l5 Repertoire of Expressed

187

mRNAs

Group l-yolk sac genes expressed in the hematopoiettc cell lme FDCPmtxA4 and upregulated during EBs development Clone 165 Novel lysosomal staltc acid O-acetylesterase (Zse) (28) Novel unknown (full-length cDNAs not yet available) Clone 260 Clone 305 Novel hematopotettc BUB2-lake protein 1 (hblpl) (30) Clone 560 Coproporphyrmogen oxtdase Clone 1000 Novel selenophosphate synthetase 2 (sps2) (29) Group 2-yolk sac genes not detectable until d 9 of EBs development. Clone 240 Novel yolk sac permease-like molecule 1 &spll) (27) Clone 320 Calbmdm-D9k Clone 460 Novel* unknown (full-length cDNAs not yet available)

but not m the other samples We observed later that thts last group of bands (Group 2 of genes m Table 1) show no expression m mtraembryomc sites of hematopoiesis (27) and likely represent yolk sac genes unrelated to hematopoiesis 2 We confirmed by Northern blot analysts the differential expression of all bands using the same RNA samples used for differenttal display and found that, m most cases, the expression patterns were quite reproductble between the two techniques (2 7) (see Note 18). 3 We have also explored the expressron of Group 1 genes m a diversity of fetal and adult tissues (see Notes 15 and 16) and looked spectfically at then expression m hematopotettc ttssues (yolk sac, AGM region, fetal liver, spleen, and bone marrow) relative to other regions of the embryo and adult tissues (Fig 4) All genes were differentially expressed throughout development and showed preferential expression in different adult tissues All were preferentially expressed in the yolk sac (at different time points m addition to d-8 5 [27-291) and also m the d-l 1 5 AGM region However, they were not abundantly expressed m fetal liver, another mtraembryomc site of hematopotests Clone 165 and Clone 260 were not preferentially expressed m adult bone marrow whereas Clone 305/3 10 and, to a certain extent, Clone 1000 showed preferential expression m this tissue Interestingly, all genes were abundantly expressed in testis, a common finding in genes involved m hematopotests (32) On the other hand, yspll, a novel gene isolated from Group 2, is not detectable m mtraembryomc sites of hematopoiests and appears to represent a potential marker of yolk sac development (Ftg 5) (27) 4. Next, we went on to isolate the full length cDNAs representing four different genes (see Note 14). Differential display 1sparttcularly useful for the identification of novel genes since tt appears to be biased towards low abundance messages. A careful analysts of the nucleic and deduced protein sequence was critical for identifying the genes isolated (Table 1) This was critical for dnectmg our

Guimaraes, McClanahan,

188 d 11.5 FETAL

d 15 FETAL

ADULT

ADULT

and Bazan CELL LINES

i i: e2.3

kb

Fig. 4. Differential expression of the Group 1 genes (see Table 1) in a diversity of fetal and adult tissues. YS, yolk sac; AGM, AGM region; FL, fetal liver; H, head of the embryo, Plac, placenta, BM, bone marrow; S. Muscle, skeletal muscle; A. Fat, perivisceral abdominal fat; FDCP, FDCPmixA4 cell line, ST0 and N2a as described; 28S, ribosomal RNA.

studies by helping us to decide which genes to pursue and in planning further experiments (Fig. 6). These included chromosomal mapping to search for mutations revealed by phenotypes that might relate to the genes under study and/or to defects in hematopoiesis (see Note 19). 5. A radical approach to obtaining functional data on novel genes rests in the knockout methodology using homologous recombination techniques and/or in overexpression models using transgenic mice (Fig. 6). We used a mouse model of selenium deficiency to study the relevance of selenium metabolism, which is activated by sps2 (29), in hematopoiesis (31).

4. Notes 1, It is critical to use this source of FCS for ES cells and Embryoid Bodies (EBs). 2. Keep the thermocycler in a hood. Aerosols are created during PCR which can result in considerable problems of radioactive contamination. If appropriate, place a carton of activated carbon particles under the lid of the thermocycler to limit contamination. Decontaminate the machine periodically and restrict its use, if

Differential


Processes

189

a-FETOPROTEIN YSPL-1 HPRT,

;:. :..&

HPRT

C 1234 YSPL-1

5 6 19

0101112131415191719192021

*

HPRT

Fig. 5. PCR analysis of the expression of&l during the in vitro development of EBs and during in vivo mouse development. (A) Differently fi-om a-fetoprotein, which is tiequently used as a marker of yolk sac tissue, yspll is not significantly expressed in EBs, as shown in this 30 cycle PCR analysis. EBs are shown to develop normally, expressing a mesoderm marker (bruchyury) at d 3 and an hematopoietic marker (PHl globin) at d 6. (B) y&l is not upregulated in EBs developed in serum-free conditions (CDM), in the presence of fetal calf serum (FCS), or with CDM plus specific growth factors, such as Activin-A (Act.) (which are capable of triggering mesoderm development) or Bone Morphogenetic Protein-4 (BMP-4) (which is capable of triggering both mesodetm and hematopoietic development) (22). The signal detected in early d-8.5 embryos (W) may be owing to contamination of this sample by small amotmts of d-8.5 yolk sac (YS). H,O- control for PCR contamination; G- genomic DNA. C- preferential expression ofyspll in the yolk sac relatively to a variety of other fetal and adult tissues. 1, d- 11.5 yolk sac; 2, d- 11.5 AGM region; 3, d-l 1.5 fetal liver; 4, d-l 1.5 head primordium; >16 are adult tissues: 5, placenta; 6, brain; 7, liver; 8, bone marrow; 9, thymus; 10, spleen; 11, skeletal muscle; 12, abdominal fat; 13, kidney; 14, lung; 15, heart; 16, testis; 17, N2a; 18, STO; 19, FDCPmixA4. 20-2 1, controls: 20, reverse transcriptase control (d 11.5 yolk sac RNA was used in the cDNA synthesis reaction to which no reverse transcriptase was added), 2 1, control for contamination of the reagents (water volume was used to replace RT sample).


190

Identtflcatlon of genes Involved hematopoletlc development

Need to compare gene expression among mulhple samples

Selectmn

based

m

patterns

on hssue dsmbbuhon gene class

and

Immunohlstochen-ustry

Fig. 6. Dtagram of the strategy followed development using differential display

to approach the study of hematopotettc

possible, to dtfferenttal display. Remember that the PCR tubes are contaminated on the outside with radtoactlvtty The use of a comb with rectangular wells Increases stgmficantly the definmon of bands and helps to avoid crosswell contammatton. The reproductbtltty of the results IS dependent on a careful quantttatton of all RNA samples Hugh quality RNA IS imperative for success of subsequent steps. Only samples with distinct 28 and 18s rtbosomal RNA bands and no signs of degradatton should be considered for further study. ReJeCt, tf possible, all samples wtth mmlma1 signs of degradatton. This allows a test of the reproductbtltty of PCR product patterns and Increases the chances of tdentlfymg genes expressed at very low levels which may not be detectable m a smgle reaction (5) (Fig. 1). Reducing the number of cycles can increase quantttattve differences between bands that show small differences at 40 cycles A semlquantrtattve analysts of differentially expressed bands constrtutes a valuable tool when choosmg bands for further study (7)

Differential Display and Developmental

Processes

191

8 Using these condmons we observed an average of 45-55 bands/lane mdtcatmg the need to perform a mmtmum of 300 prtmer combmatlons before an estimated 15,000 mRNA spectes can be vtsuahzed as bands An average of 66 bands/lane can be detected, without stgmficantly mcreasmg the background, by mcreasmg the concentratton of the 5’ primer from 2 to 30 @I, decreasmg the annealing temperature from 40 to 32°C using a cold start and 40 PCR cycles (Ftg 2) Thts represents a 25% decrease m the total number of primer combmattons required to vtsualtze 15,000 bands (6), but the cold start may deteriorate reacttons using the T,,MT primer. We feel, however, that unless one aims to explore the enttre mRNA repertoire by differential display, it may not be necessary to change the mttral condmons for PCR 9 Differential dtsplay IS not absolutely reproductble A dtfferenttally expressed band m one reaction that IS equally expressed among different populations in the duplicate reaction may not be worthy of pursuing On the other hand, a dtfferenttally expressed band m one reaction that is not present m the duplicate reaction may indicate a low abundance gene A large number of populations under comparative study allows one to confidently identify differentially expressed bands, since the maJortty of bands are constttutive m expression We had more success repeating thts first reamphficatron step when tt was unsuc10. cessful than using a fraction of its volume for a second round of 40 cycles of PCR 11 The use of gel extracted products IS crucial to avoid cloning primers and primer-dtmers 12 One band generally represents only one PCR product We found only two cases where one of the three cloned products was not identical to the other two (n = 15) However, these products did not contain the sequence of the prtmers used m the PCR, suggesting that they were clonmg artifacts. Also, they varied slightly m size when examined on the agarose gel after clonmg. 13 A significant proportion of bands results from the ampltftcatlon of cDNA sequences by a single primer (6) These bands are also reproducible and represent potentially regulated genes (Fig 3). This results from the annealing of the T,*MN primer to T stretches found m the 3’ UTR of many genes, resulting m the amplificatton of the cDNA segment between that region and the poly(A) tall Additionally, a IO-bp random sequence and its inverse-complement can exist m the same cDNA, resulting m amphficatton by the AP-primer Consequently, the longer the cDNA the more likely tt IS that tt will be pnmed by a particular prtmer combtnatton This may explain m part why we found a bias toward long (‘3 kb) messenger RNAs (all but one representing transcrtpts 22 1 kb, and half 23 6 kb [n = 71) This indicates that, whereas differenttal display theoretically IS biased to the 3’ end, m practice It can be used to isolate fragments located anywhere within a gene (6) 14. Most of the genes that we isolated were of very low frequency The majority of the genes had a frequency tf I l/15,000 and almost half of them had a representation as low as l/60,000 (n = 7) (6) The capacity of differential display to allow detection of differentially expressed genes expressed at very low levels 1s a powerful advantage over techniques such as differenttal and subtractive hybrtdtza-

192

15.

16.

17.

19.

Guimariles, McClanahan, and Bazan tion, m which the hkehhood of findmg a particular gene IS generally determined by its relative abundance. The isolation of full-length cDNAs IS likely to represent the most time- and effort-consuming part of the study Multiple differenttally expressed bands may have resulted from the expression of a single gene owing to the existence of multiple mRNA splice variants. In addition to this possibihty, we found two cases m which differently stzed bands represented the amplification of the same cDNA sequence In one case, two different bands that migrated about 5 bp apart were found upon cloning to have the same sequence, except that the longer product contained four additional base pairs at one end In a second case, five bands were found representmg PCR products of about 2000 bp (Fig. 3), upon isolation from long electrophoretic runs and clonmg, they represented only one species of 430 bp. This fragment contained the 3’ primer sequence at both ends and the groupmg of bands may have resulted from the multimerization of an A-T rich region located at the end of this fragment The isolation of the full length cDNA for this band allowed us to confirm the existence of a 2 1-T stretch m the cDNA at the expected position m the 3’ UTR. One gene may have different mnctions m different tissues or be perhaps more relevant m one tissue than others. It can also be expressed at dtfferent times m different tissues. mRNA and protein expression do not always correlate The development of specific antisera or monoclonal antibodies IS helpful to this regard and also for locatmg the protein at the subcellular level Not all mRNAs encode proteins (33-35). You may need to expose the blot to film for more than 2 wk (or 1 wk on a Phosphorimager) in order to detect a signal. 18. In two out of ten cases we could not detect a signal by Northern analysis after two wk of exposure We feel, however, that this should not constitute the sole crtterium for exclusion of bands for further study if reproducible differential expression is found by differential dtsplay In these cases, more sensitive methods such as RNase protection assays should be tried to confirm differential expression A significant proportion of mutations with recognizable phenotypes result from gross genormc abnormahties that may be identifiable by genomic Southern blot analysis.

Acknowledgments We gratefully acknowledge A. Zlotnik, T. Kinoshita, S. McCarthy, A. Vicari, E. Ching, and G. Hardiman for stimulating discussions, J. M. Pina Cabral for guidance and F. Lee who prompted these studies after suggesting

the use of differential display to approach hematopoietic development. M. Jorge Guimaraes is supported by a grant from JNICT (CI~NCIA/BD/2685/93). DNAX Research Institute of Molecular and Cellular Biology, Incorporated, is supported by the Schering-Plough Corporation. References 1. Gilbert, S. F., ed. (1991) Developmental Biology, Smauer, Sunderland, MA. 2 Horvitz, H. R and Herskowitz, I (1992) Mechanisms of asymmetric cell division: two Bs or not two Bs, that is the question Cell 68,237-255

Dlfferentlal Display and Developmental

Processes

793

3. Amon, A (1996) Mother and daughter cell are doing fine asymmetric cell dlvlslon m yeast. Cell 84,65 l-654 4. Llang, P. and Pardee, A. B. (1992) Dlfferentlal display of eukaryotlc messenger RNA by means of the polymerase cham reaction Sczence 257,967-97 1 5. Llang, P., Averboukh, L , Keyomarsl, K , Sager, R., and Pardee, A B (1992) Differential Display and clonmg of messenger RNAs from human breast cancer versus mammary eplthellal cells Cancer Res 52,696&6968 6 Llang, P., Averboukh, L., and Pardee, A B (1993) Dlstrlbutlon and cloning of eukaryotlc mRNAs by means of differential display. refinements and optlmlzation. Nucleic Acids Res 18, 3269-3275 7. Gulmar&es, M J., Lee, F , Zlotmk, A., and McClanahan, T (1995) Differential Display by PCR* novel findmgs and applications Nucleic AczdsRes 23,1832-l 833 8 Medvinsky, A. L., Samoyhna, N. L , Muller, A. M , and Dzlerzack, E. A. (1993) An early pre-liver mtra-embryonic source of CFU-S in the developing mouse Nature 364,64 9 Delassus, S. and Cumano, A (1996) Clrculatlon of hematopoletlc progenitors m the mouse embryo. Immunzty 4,97-106. 10. Potten, C. S., ed. (1983) Stem Cells, Thezr Identljkatlon and Characterzzatzon Churchill Livingstone, Edmbourgh, London, Melbourne and New York. 11 Evans, M J and Kaufman M H (198 1) Establishment in culture of plurlpotentlal cells from mouse embryos Nature 292, 154 12. Martin, G R. (198 1) Isolation of a pluripotent cell line from early mouse embryos cultured in medmm condltloned by teratocarcmoma stem cells Proc. Natl Acad Sa USA 78,7634-7638 13 Robertson, E J , ed (1987) Teratocarcznomas and Embryonzc Stem Cells, a Practzcal Approach IRL, Oxford, Washington DC 14 Bradley, A., Evans, M , Kaufman, M. H., and Robertson, E. (1984) Formation of germline chnneras from embryo-denved teratocarcinoma cell lines. Nature 309,255-256. 15. Doetschman, T. C , Elstetter, H , Katz, M., Schmidt, W , and Kemler, R (1985) The in vztro development of blastocyst-derived embryonic stem cell lines: Formation of visceral yolk sac, blood Islands, and myocardmm J. Embryo1 Exp Morph01 87,27-45. 16 Lindenbaum, M H. and Grosveld, F. (1990) An zn vztro globin gene switching model based on differentiated embryonic stem cells Genes Dev 4, 2075 17 Wiles, M V. and Keller, G (199 1) Multiple hematopoletlc lineages develop from embryonic stem (ES) cells m culture. Development 111,259. 18. Chen, U , Kosco, M., and Staerz, U. (1992) Estabhshment and charactenzatlon of lymphold and myelold mixed-cell populations from mouse late embryold bodies, “embryonic-stem-cell fetuses” Proc Nat1 Acad Scz USA 89,2541-2545 19 Gutlerrez-Ramos, J. C and Palaclos, R. (1992) In vitro differentiation of embryonic stem cells into lymphocyte precursors able to generate T and B lymphocytes m vitro. Proc Natl. Acad. Scz USA 89,9 17 l-9 175 20. Keller, G , Kennedy, M., Papayannopoulou, T , and Wiles, M V (1993) Hematopoletic commitment durmg embryonic stem cell differentlatlon m culture Mel Cell Biol 13,473486

194

Guimardes, McClanahan, and Bazan

2 1 McClanahan, T , Dalrymple, S , Barkett, M , and Lee, F (1993) Hematopotetic growth factor receptor genes as markers of lmeage commttment during zn vztro development of hematopoiettc cells Blood 81, 2903-2915 22 Johansson,B and Wiles, M V (1995) Evidence for the involvement of acttvm-A and BMP4 m mammalianmesodermand hematopotetic development Mel Cell Bzol 15, 141-15 1 23 Nakano, T., Kodama, H , and Homo, T (1994) Generation of lymphohematopotettc cells from embryomc stem cells m culture Sczence 265, 1098-l 101 24. Ford, A. M., Healy, L H , Bennet, C. A., Navarro, E., Spooncer, E , and Greaves, M F (1992) Multtlineage phenotypes of Interleukm-3-dependent progenitor cells Blood 79, 1962-1971 25 Wong, P M C , Han, X.-D , Ruscetti, F W , and Chung, S -W (1994) Immortalized hematopotettc cells with stem cells properties. Zmmunzty 1, 571-583 26 Orkm, S. H (1995) Hematoporests:how doesrt happen?Curr Opzn Cell Bzol 7, 87&X77. 27 Gutmaraes, M J , Bazan, J F , Zlotnik, A , Wiles, M. V , Grtmaldt, J C , Lee, F , and McClanahan, T (1995) A new approachto the study of hematoporettc development m the yolk sac and embryoid bodies Development 121, 3335-3346 28 Guimaraes, M J , Bazan, J F , Castagnola,J , Dtaz, S , Copeland, N G , Gilbert, D J , Jenkins, N A, Varkt, A, and Zlotnik, A (1996) Molecular cloning and characterizatron of lysosomal stahc-acid 0-acetylesterase J Bzol Chem 271, 13,697-13,706 29 Gutmaraes,M J , Peterson,D , Vicari, A , Cocks, B G , Copeland, N G , Gilbert, D. J., Jenkins, N. A , Ferrick, D. A., Kastelem, R A , Bazan, J. F , and Zlotmk, A. (1996) Identification of a novel selD homolog from eukaryotes, bacteria and archaea* is there an autoregulatory mechanism m selenocysteme metabolism? Proc Nat1 Acad Scz USA 93, 15,08615,091 30 Gmmaraes,M. J and Bazau, J. F (1997) The chessprotein motif redefines relationships among genesinvolved m cell signaling, cell cycle regulation, chromatm function and cell divtsion In preparation 31. Gutmaraes, M J , Hudak, S , Vicari, A., Rosst, D , Leach, M , Hill, K E , Burk, R. F , Pma-Cabral, J M , Renmck, D , and Bazan, J. F. (1997) Thymtc atruphy and delayed kinetics of hematopoiettc repopulation m selenmmdeficient mice In preparation 32 Ito, E , Tokt, T., Ishrhara, H., Ohtam, H ,Gu, L , Yokoyama, M , Engel, J D , and Yamamoto, M. (1993) Erythroid transcription factor GATA- 1 1sabundantly transcribed in mousetestis Nature 362,466. 33 Brannan, C. I , Dees, E. C , Ingram, R S., and Tilghman, S M (1990) The product of the HI9 gene may function as an RNA Mel Cell BzoE 10,28-36 34 McCarrey, J R and Drlworth, D. D. (1992) Expression of Xrst m mouse germ cells correlates with X-chromosome mactivation Nature Genet 2, 20&203 35. Swalla, B J. and Jeffery, W. R (1995) A maternal RNA localized m the yellow crescent 1ssegregatedto the larval musclecells durmg asctdrandevelopment. Dev Bzol 170,353-364

Identification of Novel Genes Involved in Adipose Differentiation by Differential

Display

Erding Hu and Bruce M. Spiegelman 1. Introduction In vertebrate animals, adipose tissue functions primartly to store excess energy m the form of triglycerides m periods of nutritional abundance and mobilize it in response to fasting (I,Z). It also plays important roles in fatty acid metabolism and glucose homeostasts (3). Recently, adipose tissue has been shown to have endocrine functions by synthesizing and secreting a group of stgnalmg molecules to regulate various body functions (4-6). This 1s rllustrated by the identification of ob gene product (leptm). Leptm is secreted exclusively from fat mto the circulation and acts to regulate body weight, via a receptor (db gene product) m the cerebroventricular region of the brain (69). Thus the ability to interact and regulate diverse functions m fat and other tissues represents a new facet of adipose tissue physiology. Adipose tissue development m vivo can be mimicked in tissue culture using several well known preadipocyte cell lines. Among them are 3T3-F442A and 3T3-Ll cells. These cells are derived from mouse embryonic fibroblasts. Upon proper hormonal stimulation, these preadipocytes undergo a dramatic morphological conversion, changing from flat fibroblasts mto round spherical-shaped adipocytes containmg lipid droplets. At a molecular level, many genes expressed m vivo are also expressed m these adipocytes generated in vitro. They includes enzymes involved m lipid synthesis (lipoprotein ltpase, fatty acid synthases, mahc enzyme) (Z&12), lipid and glucose transport proteins such as Glut4, UP-2 (13,14), and many hormones receptors (Insulin receptor, adrenergic receptors) (2), as well as several secreted signalmg molecules, including IGF-I, TNF-a, and leptm (4-6). The authenttcity of the in vitro model of adipose differentiation was further validated by transplantation From

Methods m Molecular B/ology, Vol 85 Dlfferentlal Edited by P Llang and A B Pardee Humana

195


Hu and Spiegelman

196

experiment in which preadipocyte 3T3-F442A cells were transplanted mto nude mice. Subsequent exammatlons revealed that the adipose tissue developed m the nude mice contained the cells derived from 3T3-F442A cells (1.5). The avaliabihty of these cells enabled investigators to study the molecular mechanisms governing adipose differentiation and isolate genes regulated m this process. The ability to analyze and isolate genes specifically expressed in adipose tissue has pivotal importance since these genes provide important insight mto the physiological functions of the tissue. Previous efforts to isolate genes that are expressed in a tissue-specific or differentiation-dependent manner involve techniques such as differential screening of cDNA library or generating subtractive libraries (24, I6,17). These methods, although effective m ldentlfymg important molecules (aP-2, myoD, or pref-l), are technically complex or require large amount of mRNA Differential display techniques provide a new approach that can be utilized to analyze and characterize mRNAs expressed m adipose tissue. It 1s a relatively simple procedure and requires small amount of RNA. We have used this technique to isolate a novel gene termed adipoQ (18) AdlpoQ encodes a secreted molecule with a signal peptide m the N-terminal region of the protein. The expression of adlpoQ is highly regulated during the adipose differentiation process and it 1sexpressed exclusively m adipose tissue m vivo. In addition, it 1sslgnlficantly downregulated in the obese state. Thus adlpoQ 1s a secreted molecule that may have signaling functions and may play a role in the pathogenesis of obesity and diabetes. Here we describe in detail the differential display protocol used to isolate adlpoQ. This procedure can be easily adapted to other models of differentiation for isolating regulatory genes. 2. Materials All glassware used for lsolatlon of RNA 1s treated with 0.05% dlethylpyrocarbonate (DEPC) overnight and autoclaved for 1 h before use. The agents used are also treated similarly by either autoclaving or filter stenllzatlon. 2.1. Isolation of Total Cellular RNA from Fibroblasfs, Preadipocyfe, and Differentiated Adipocytes 1 PBS buffer 137 mMNaC1,2.7 mMKC1,4.3 mMNa2HP04, and 1 4 rnMKH2P04 Autoclave and store at room temperature 2. Cell lysis buffer. 4Mguanidmm thlocyanate, 25 mMsodmm citrate pH 7 0,O 5% sodium N-lauroyl sarcosme Store at room temperature (19). Immediately before use, P-mercaptoethanol is added to a final concentration of 0. 1M. 3. CsCl solution (5 7A4): 1 g/mL of CsCl, 25 mM sodium citrate pH 7 0, 0 1M

EDTA. Store at room temperature

Novel Genes in Adipose Differentiation

197

4. 70% Ethanol. Dilute 200 proof 100% ethanol with distilled water to 70%. Keep at -20°C. 5 TE buffer: 10 mM Tns-HCl, pH 8.0, and 1 rr&! EDTA, autoclave, and store at room temperature 6 Na,Ac (3M): 3M Sodium acetate is prepared using dlstilled water and filtered Store at room temperature

2.2. Treatment

of RNA and Reverse

Transcription

1. 10X PCR buffer: 100 mM Tns-HCl, pH 8.3, 500 mA4 KCI, 15 mM MgCI,, and 0 0 1% (w/v) gelatin Autoclaved and store frozen at -20°C. 2. RNase-free DNase I: purchased from BRL at 10 U/& Store at -20°C. 3. RNase mhibltor: purchased from Promega Corporation (Madison, WI) at 20 U/pL Store frozen. 4. MMLV Reverse transcnptase: purchased from BRL at a concentration of 200 U/pL. 5 Reverse transcription (RT) buffer (5X): 250 mMTns-HCl, pH 8.3,30 mMMgCl,, and 375 mM KCI. Store at -20°C. 6. 250 w dNTP: diluted from 100-a stock obtained from Pharmacia (Piscataway, NJ), and stored at -20°C. 7 Two-base pair anchored 3’ primer: four two-base anchored primers are TIZMA, T12MT, Tt2MG, T,*MC. M indicates a mixture of G, A, C bases. Primers are at a concentration of 10 w. Store at -20°C

2.3. Differential

Display PCR Reaction

1. 10 m dNTP: Dilute 100 mM dNTP stock (Pharmacla) solution with distrlled water and store at -20°C. 2. Tuq polymerase. obtained from Perkm Emer at 5 U/pL 3 5’ arbitrary primer (AP): lo-base-long arbitrary primers are deslgned by keeping G-C content approx 50% and avoidmg any palmdromlc sequences. 4. a[35S]-dATP: purchased from DuPont-NEN (Boston, MA) with a specific activity of 1000 Ci/mmol. 5. Loading dye: 95% formamide, 20 m&f EDTA, 0.05% bromophenol blue, and 0.05% xylene cyan01 FF

2.4. Cloning of Differentially and Further Analysis

Amplified

PCR Fragments

1. Glycogene purchased from Boehringer Mannhelm (Indianapolis, in distilled water to a final concentration of 1 pg/pL. 2. TA cloning vector: purchased from Invitrogen (Carlsbad, CA).

IN) Dissolved

3. Methods 3.7. Isolation of Total Cellular RNA from Fibroblasts, Preadipocyfes and Differentiated Adipocytes (see Note 7) 1. Fibroblasts (3T3-C2) and preadipocytes (3T3-F442A) are cultured in 100 mm culture dishes with DMEM supplemented with 10% fetal calf serum and antlbl-

198

Hu and Spiegelman

ot1cs (pen-strep) to confluence For tibroblasts and preadlpocytes, isolation of RNA 1sperformed when cells reach confluence For adipose dlfferentlatlon, 3T3F442A cells are treated with the same medium and 5 pg/mL of insulin for 10 d or till adipose dlfferentlatlon reaches at least 80% More details of ad1pocytes d1fferentiatlon are described 1n ref 14 2 Rinse loo-mm dishes with 5 mL of PBS, and add 1 mL of RNA lys1s buffer. Scrap the cell extract to 50 mL Falcon 2098 tube Usually 5-10 dishes (-5 x lo7 cells) are needed for subsequent ces1um chloride centrifugatlon step Collect lysed cell extract from all dishes and pool to the Falcon tubes for flbroblasts, preadlpocytes and ad1pocytes respectively. 3. Use a lo-mL syringe and a 22-gage needle to shear the chromosomal DNA by plpeting the cell lysates up and down several times Then to a Beckman heat sealable tube (Beckman 342413), add 6 5 mL of 5 7A4 CsCl solution and then fill the tube with cell lysate (approx 6 5 mL) The tubes are labeled and sealed with heat sealer Spin the Beckman 342413 tube 1n Beckman T170 1 rotor at 200,OOOg for 20 h at 25°C Other tubes can also be used with swing bucket rotor (e.g., SW41) without using heat seahng system 4. At the end of centrifugation, remove the tubes from centrifuge and cut open the top of the sealed tube LJs1ng an autoclaved glass plpet, slowly remove the solution. The RNA pellets should be visible Wash the RNA pellets with 70% Icecold ethanol Use a new glass p1pet to scoop the pellets to an Eppendorf tube containing 500 pL of 70% ethanol, spin down the RNA pellets, and dissolve RNA 1n 100 pL of DEPC-treated water Determine the RNA concentration by OD measurements (1 ODlhO= 40 pg/mL of RNA). The final concentration of RNA should be approx 1 pg/pL Store the RNA samples frozen at -70°C

3.2. Treatment of Total RNA with RNase-Free DNase and Reverse Transcription of Total RNA (see Note 2) 1 To 50 pg of RNA (50 pL), add 1n order 5 7 pL of 10X PCR buffer (see Section 2 ), 1 pL of RNase Inhibitor (20 U/pL), 2 ccl, of RNase-free DNase (10 U/I&,) M1x well and Incubate for 30 m1n at 37OC 2 Add 50-60 p.L of phenol/CHCls (3. I), vortex for 30 s and spin 1n Eppendorf centrifuge for 5 min. Transfer supernatants to new tubes. 3. Add 6 pL of 3MNa2Ac and 150 pL of 95% EtOH Freeze tubes for 2 h at -80°C. Spin for 15 mm at 4°C. Remove supernatants and wash RNA pellets with 500 &L of 70% EtOH Then redissolved RNA 1n 25 pL of DEPC-treated water. Quantltate RNA by ODZ6c and adJUst to 1 pg/pL Use lmmedlately or store the treated RNA at -80°C 4 To four Eppendorf tubes on Ice, add 8 4 pL of water, 4 pL of 5X reverse transcrlptlon (RT) buffer, 2 pI. of 100 mM DTT, 1 6 pL of dNTP (250 p&f), 2 pL of treated RNA (2 pg) and 2 pL of each of the four 3’ primer (T,,MX, 10 w 5 Incubate the Eppendorftube at 65°C for 5 mm, then 37’C for 10 min. Add 1 pL of MMLV reverse transcnptase (200 U) Continue incubation at 37*C for 60 min before heating the tubes to 95°C for 5 m1n. Keep the tube on 1ce or freeze for later use

Novel

Genes

3.3. Differential

in Adipose

Differentiation

Disp/ay PCR Reactions

199

(see Note 3 and Fig. 1)

1, PCR reactions are performed m 20 pL, of reaction volume contammg. 9 2 & of water, 2 & of 10X PCR buffer, 1 6 pL of dNTP (25 I-1M), 1 $ of a[35S]-dATP (1000 CYmmol, NEN), 2 pL, of 5’ arbitrary primer (10 c1M), 2 pL of 3’ anchored primer (10 CUM),2 pL of reverse transcription mixture and 0 2 p.L of Taq polymerase Add all components to PCR tubes on ice. 2. Mix well by plpetmg up and down. Add 25 pL of mineral 011 PCR at 95°C for 30 s, then 40°C for 2 mm, and 72°C for 30 s (for 40 cycles). Fmally mcubate the tubes at 72°C for 5 mm. 3 Take 3 5 pL of PCR sample and add 2 p,L of loading dye Incubate at 8O’C for 2 mm before loading onto a 6% DNA sequencmg gel For sequencing gel, use flat regular teeth comb (0.4 mm thick) instead of shark teeth combs to obtam a better band resolution

3.4. Cloning of Differentially and Further Analysis

Amplified

PCR Fragments

1 After finishmg sequencing gel, directly transfer gel to a 3MM whatman paper and dry the gel under vacuum at 80°C for 2 h Expose the dried gel to X-ray film (Kodak) overnight to vlsuahze the dlfferentlally amphfied DNA fragments To precisely match the gel to the exposed film, use radioactive ink at the four corners of the gel 2 Once a differentially amplified band 1s located, use a razor blade to cut the band (mcludmg dried gel and whatman paper), then carefully remove the paper from the dried gel using surgical blades 3 Put the dried gel slices mto 100 @ of TE buffer Boll for 30 min Spin down the msoluble material and move the supernatants to new tubes. To the supernatants, add 10 ug of glycogen, 10 pL of NazAc (3w and 250 & of 100% ethanol (lcecold) Keep the tubes at -2O’C for at least 1 h Then spm at 4°C for 15 mm and wash the pellets with 70% ethanol. The DNA pellets are dissolved m 10 pL of TE buffer or water for further amphficatlon. 4. One mlcrohter of extracted DNA m TE or water 1sused for further amphficatlon m a 50-pL reaction volume using the same primer pair that gives the dlfferentlal amplification and the same PCR reactlon parameters. However, no a[35S]-dATP 1s used and the dNTP concentration 1sincreased to 20 pA4 (final concentration) After the PCR reaction, 2% agarose gel is used to visualize the amplified band If DNA bands are vlslble, TA clonmg vector 1s used to clone the amphfied fragment as descried by the manufacturer’s instruction (Invltrogen) 5. To verify the expression pattern of the amplified DNA fragments (see Note 4 and 5), Northern blot analysis 1sperformed using either PCR-amplified fragments, or cloned fragments as probes (Figs 2 and 3).

4. Notes 1. The quality and integrity of the RNA samples are of pivotal importance for the differential display. We have observed that CsCl centrlfugatlon give RNAs of better quality than a quicker acid-phenol extraction procedure (19) The RNA

Hu and Spiegelman

Fig. 1. Differential display gel using T&M as 3’ primer and 9 different 5’ arbitrary primers (AP). RNA from three types of cells are analyzed. C2,3T3-C2 fibroblasts; P, 3T3-442A preadipocytes; A, 3T3-adipocytes. RNA isolation, reverse transcription and differential PCR reactions are performed as described in the text. Notice that the expression of majority of mRNA species is not altered in adipose differentiation. However, there are a few mRNA that are specifically induced or decreased during adipose differentiation. The bands with a arrow indicate the amplified DNA bands representing adiopoQ mRNA.

Novel Genes in Adipose Differentiation F442A PA

c2

201 -

Ll PA

AdipoQ

-

28s

-

18s

1

@Actin

EtBr

Stain

1

2

3

4

5

Fig. 2. Expression of adipoQ mRNA in 3T3-C2 fibroblasts (C2), 3T3-F442A preadipocytes (F442A, P), adipocytes (F442A, A), 3T3-L 1 preadipocytes (L 1, P), and adipocytes (Ll, A). Northern analysis is performed as a standard procedures using PCR amplified DNA fragment as a probe. AdipoQ mRNA appears as an abundant message of approx 1.2 kb in length. p-actin and EtBr staining are used to normalize RNA loading. The expression of adipoQ in Northern matches very well with that observed in differential display gel. (With permission from ref. 18.) isolated using acid-phenol extraction is prone to genomic DNA contamination. To ensure the RNA isolated is of high quality, it is advisable that RNA samples before and after DNAase treatment are examined by running a 1% agarose gel. 28s and 18 S rRNA band should be prominent and proportional with little or no trailing. It is also important to make duplicate reactions from independently prepared RNA samples. 2. DMSO and glycerol have been reported to increase the fidelity and specificity of PCR reaction by minimizing DNA damage such as depurination and deamination (20). A range of 5-l 0% DMSO and 10% glycerol could be included in initial differential PCR and subsequently amplification step. The precise concentration

202

Hu and Spiegelman BFHIKLMPS -28s -18s

AdipoQ

aP2

EtBr Stain

1

23456769

Fig. 3. Expression of adipoQ mRNA in various tissues from mouse. RNAs from several different tissues (B, brain; F, fat; H, heart; I, intestine; K, kidney; L, liver; M, muscle; P, pancreas; and S, spleen) from mouse are isolated and analyzed for adipoQ mRNA expression by Northern blots. The same blot is blotted with another adipose-specific gene UP-~. The RNA loading is visualized by ethidium bromide staining. (With permission from ref. 18.) of DMSO or glycerol should be determined empirically. However, the band pattern may be different when DMSO or glycerol is added. This may reflect the effects of DMSO and glycerol on the annealing process of the primers to nonperfectly matched templates. 3. We have also successfully used one-base anchored 3’ primer for reverse transcription and differential display PCR reaction under similar conditions (ZZ). 4. It is empirical that the cloned PCR fragment be checked with Northern blots to verify the expression pattern observed in differential display. We have noticed, in several cases, that the PCR amplification of the excised DNA bands from sequencing gel yields multiple DNA fragments of the same length, so that a Northern signal observed using the uncloned PCR mixture as probe is “lost” after cloning. In this case, picking multiple clones will be necessary to isolate the “correct” fragment that gives signals on Northern blot. 5. Since the reamplified DNA fragments are usually small (200-400 bp), it is essential to obtain labeled probe with high 32P specific activity if this DNA fragment is to be used directly for Northern analysis. We have used the 10 base AP oligo itself as primers in a labeling reaction using either klenow or T7 DNA polymerase.

References 1. Spiegelman, B. M., Choy, L., Hotamisligil, G. S., Graves, R. A., and Tontonoz, P. (1993) Regulation of adipocyte gene expression in differentiation and syndromes of obesity/diabetes. J. Biol. Chem. 268,6823-6826.

Novel

Genes

In Adipose

Differentiation

203

2 Cornelius, P , MacDougald, 0. A , and Lane, M. D (1994) Regulation of adipocyte development Annu Rev Nutr 14,99-129. 3 Sptegelman, B M. and Hotamtshgd, G S (1993) Through thick and thm wastmg, obesity, and TNF alpha Cell 73,625-627 4 Dogho, A., Dana, C., Fredrickson, G., Grimaldt, P , and Athaud, G. (1987) Acute regulation of insulin-like growth factor-I gene expression by growth hormone during adipose cell differentiation EA4BO J 6,40 1l-40 16 5 Hotamtsligil, G S , Shargill, N. S., and Spiegelman, B M (1993) Adipose expression of tumor necrosis factor-alpha. direct role m obestty-linked insulin resistance. Sczence 259, 87-91 6 Zhang, Y , Proenca, R , Maffet, M , Barone, M , Leopold, L , and Friedman, J M (1994) Posmonal clomng of the mouse obese gene and its human homologue Nature 372,42-T-432

7 Halaas, J L , GaJiwala, K. S , Maffei, M , Cohen, S L , Chait, B T , Rabmowltz, D., Lallone, R L., Burley, S K., and Friedman, J. M (1995) Weight-reducing effects of the plasma protein encoded by the obese gene Sczence 269,543-546 8 Tartagha, L A , Dembskt, M., Weng, X., Deng, N , Culpepper, J , Devos, R , Richards, G J , Campfield, L A , Clark, F T , Deeds, J , et al. (1996) Identification and expression cloning of a leptm receptor, OB-R Cell 83, 1263-127 1 9 Lee, G H., Proenca, R., Montez, J M., Carroll, K M , Darvtshzadeh, J G , Lee, J I , and Frtedman, J M (1996) Abnormal sphcmg of the leptm receptor m dtabettc mice Nature 379, 632-635 10. Cornelms, P , Enerback, S , BJursell, G , Olivecrona, T , and Pekala, P H (1988) Regulation of hpoprotem hpase mRNA content in 3T3-Ll cells by tumour necrosis factor Blochem J 249,765-769 11. Moustatd, N. and Sul, H. S. (1991) Identification of an insulin response element m the fatty acid synthase promoter. J BEOI Chem 266, 18,550-l 8,554 12. Wise, L. S., Sul, H. S., and Rubm, C S (1984) Coordinate regulation of the btosynthesis of ATP-citrate lyase and mahc enzyme during adipocyte dtfferenttatton Studies on 3T3-Ll cells J. Brol Chem 259,4827-4832 13 Kaestner, K H., Christy, R J., McLemthan, J C , Bratterman, L T , Cornelius, P , Pekala, P H , and Lane, M D (1989) Sequence, tissue distributton, and differential expression of mRNA for a putative insulin-responsive glucose transporter m mouse 3T3-L 1 adipocytes Proc Nat1 Acad Scr USA 86,3 150-3 154 14 Sptegelman, B M , Frank, M , and Green, H. (1983) Molecular cloning of mRNA from 3T3 adipocytes J Blol Chem 258, 10,083-10,089 15. Green, H and Kehmde, 0. (1979) Formation of normally differentiated subcutaneous fat pads by an established preadipose cell line. J Cell Physzol 101, 169-172 16 Smas, C M and Sul, H S (1993) Pref-1, a protein containing EGF-like repeats, mhibits adipocyte differentiation. Cell 73,725-734 17 Davis, R L , Wemtraub, H , and Lassar, A. B. (1987) Expression of a single transfected cDNA converts tibroblasts to myoblasts Cell 51,987-1000 18 Hu, E , Ltang, P , and Sptegelman, B M (1996) AdtpoQ 1sa novel adipose-specific gene dysregulated m obesity J BloI Chem 271, 10,697-10,703

204

Hu and Spiegelman

19 Chomczynski, C. and Saccht, N. (1987) Single-step method of RNA tsolatton by acid guamdmmm thtocyanate-phenol-chloroform extraction Anal. Biochem. 162, 156159. 20. Cheng, S., Fockler, C., Barnes, W. M., and Htguchi, R. (1994) Effective amphfication of long targets from cloned mserts and human genomrc DNA. Proc NatZ Acad. Scl USA 91,5695-5699.

21 Peng, L., Zhu, W., Zhang, X , Guo, Z., O’Connell, R P , Averboukh, L , Wang, F., and Pardee, A B (1994) Differential display using one-base anchored oligodT primers Nucleic Acids Res 22, 5763,5764

Isolation of Song-Regulated Genes in the Brain of Songbirds Claudio V. Mello, Erich D. Jarvis, Natalia Denisenko, and Miriam Rivas 1. Introduction We have applied the differential display (DD) technique (1) to isolate genes whose expression is regulated in the brain of songbirds when they hear song of their own specres. Song is known to cause a marked increase m mRNA levels of two immediate early genes, ZENK and c-~un, in the auditory forebrain of songbirds, the most pronounced induction occurring in the caudomedral neostriatum, or NCM (2-4). Several studies suggest that this gene regulatory event may be related to song processing, discrimination, and the formation of song auditory memories. For example, ZENK induction in NCM is highest for songs of the same species, Iower for songs of other species, and lowest or absent for nonsong auditory stimuli (2). ZENK induction habituates to repeated presentations of the same song, but can be re-elicited by a novel song (5). Srmrlarly, the electrophysiological response of NCM neurons habituates when the same song is presented repeatedly, but is high again when a novel song IS then presented (6); this song-specific electrophysiological habituation is long-lasting and its long-term maintenance depends on local protein and RNA synthesis (6). Finally, behavioral studies indicate that ZENK induction in NCM correlates with associative learning when song is used as a stimulus (7). The studies above suggest that a cascade of gene regulatory events triggered by song m NCM is involved in the formation of song auditory memories. The ZENK and c-jun genes encode transcripttonal regulators (8,9) and could play a role in coordinating this cascade. To begin to decipher such a cascade, we have used DD in the isolation of a cohort of genes induced by song in NCM. We have studied two time-points, From

Methods m Molecular B/ology, Vat 85 Dffferenbaal D/splay Methods and Protocols Edlted by P Llang and A B Pardee Humana Press Inc , Totowa, NJ

205

206

Mello et al

30 mm and 2 h after the start of song presentation. The first time aimed at tdenttfymg genes coregulated with ZONK and c-jun. The second represents a time when ZENK protein levels are high (20) and downstream genes are presumably being regulated. We descrtbe here the strategy we developed and the optlmtzatlon of the methods for secondary screening necessary to apply DD to our experimental system. 2. Materials For items 2 2-2 4,2.8, and 2 9, use RNase-free salts and reagents and DEPCtreated water (add 0.1% DEPC to disttlled water, mcubate for 12 h at 37°C and autoclave for 15 min), and work under RNase-free condmons (wear gloves and use sterile, disposable plasticware, see Note 1). Store solutrons at-20°C unless otherwtse stated. 2.1. Song Stimulation Animals. adult male and female zebra finches (Taenlopygmag&rata). 2.2. RNA Extraction 1 Solutton D 4M guamdmium thiocyanate, 25 mM sodmm citrate, pH 7 0, 0 5% sarcosyl, and 0.72% /$mercaptoethanol (P-ME), stable for 1 mo at 4°C 2. 2MNaOAc, pH 4 0, store at room temperature 3 Phenol: saturated with distilled water (wrthout any buffer), store at 4°C. 4. Chloroform-isoamyl alcohol (49.1).

2.3. DNase Treatment 1. RNase-free DNase I (Boehrmger-Mannhelm, Indianapolis, IN), 20 U/pL, and 1OX DNase buffer 2 RNase inhibitor (RNasm, Promega, Madison, WI), 40 UlpL 3 Phenol*chloroform isoamyl alcohol (25 24: 1). saturated with 100 mM Tris-HCl, pH 8.0, stable for 1 mo at 4°C protected from light 4 3MNaOAc, pH 5.2, store at room temperature

2.4. cDNA Synthesis 1 Superscript II reverse transcnptase (RT; Gibco-BRL, Grand Island, NY), 200 U/pL, and 5X RT buffer 2 1OOmMDTT 3 dNTP Stock* 250 pA4 each of dATP, dCTP, dGTP, and dTTP 4 10 pMT,,MN primers separate stocks ofthe followmg* T,,MA, T,,MC, T,,MG, and TlzMT (where M = A, C, or G), T,,GC and T&G.

2.5. PCR Amplification 1. 2 pMStocksof arbrtrary 10-mersof randomsequence(AP-primersl-20, GenHunter, Brooklme, MA)

Song-Regulated

207

Genes

2 AmphTaq DNA polymerase (Perkin Elmer, Norwalk, CT), 5 U/pL, and 1OX PCR buffer I. 3 25 wdNTP* 1.10 dllutlon of stock described m Section 2.4 , item 3 4 35S-dATP (NEN, Boston, MA) 5 Loading buffer 98% deionized formamide, 10 mM EDTA, pH 8 0, and 0 025% each of xylene cyan01 FF and bromophenol blue. 6. 5X TBE buffer: combme 54 g of Trls base, 27.5 g of boric acid and 20 mL of 0 5MEDTA, pH 8 0, and add water to 1 L; store at room temperature 7 6% Polyacrylamlde gel mix. to 230 g of urea add 75 mL of 40% polyacrylamlde mix (Fisher), 100 mL of 5X TBE and water to 500 mL; filter (0.45-p pore) and store for several weeks at 4°C 8 lOO-bp DNA ladder (Glbco-BRL). 9 Klenow DNA polymerase (New England BloLabs, Beverly, MA), 2 U/@, and Klenow buffer.

2.6. Fragment

Excision and Reamplification

1 2% Nusleve (Sigma, St Louis, MO) agarose gels (see Note 1) 2 Qiaex II gel extraction kit (Qlagen, Chatsworth, CA) 3 10 mg/mL glycogen

2.7. Cloning into PCR Plasmid Vectors 1 pCRScript (Stratagene, La Jolla, CA) or TA pCR II (Invitrogen, San Diego, CA) clonmg kits 2 EcoRI and BssHII restriction enzymes (Boehrmger Mannhelm), 10 U/pL 3 Random primer labeling kit (Boehrmger Mannhelm) 4 32P-dCTP (NEN).

2.8. Secondary

Screening

by Northern

Hybridization

1 10X MOPS buffer to 20 9 g of MOPS (3-[N-morpholinolpropanesulfomc acid) add 300 mL of water, 50 mL of 0 8MNaOAc, and 50 mL of 0 5M EDTA; adJust the pH to 7 0 with NaOH, add water to 500 mL, filter (0.22-w pore), and store at 4°C protected from hght 2 1% MOPS/formaldehyde agarose gel. 1% agarose in 1X MOPS buffer contaming 2.2% formaldehyde; run m freshly diluted 1X MOPS buffer 3. MOPS sample buffer: combme 4.0 $ of the RNA m water, 2 & of 10X MOPS buffer, 10 & of formamide, 3 7 & of formaldehyde, 2 pL of dye mix (50% glycerol, 1 mM EDTA, pH 8 0, 0.25% bromophenol blue and 0 25% xylene cyan01 FF) and 0 025% ethldmm bromide Prepare fresh and heat to 65°C for 10 min before Ioadmg 4. 20X SSPE: dissolve 175 3 g ofNaC1,27 6 g of NaH2P04 Hz0 and 7.4 g of EDTA in 800 mL of water, adJust pH to 7.4 with NaOH and add water to 1 L. Store at room temperature. 5 Nylon membranes (NEN’s GeneScreen Plus or Amersham’s Hybond Plus). 6. 32P-UTP (NEN)

Mello et al.

208

7 5X Transcription buffer* 2 5 mM ATP, CTP and GTP each, 60 pA4UTP, 50 mM DTT, 200 mA4Tris, pH 7 5,30 mMMgC12, 10 mM spermidme and 50 mMNaC1. 8. 20 mg/mL RNase-free BSA 9 T7, T3, and SP6 RNA polymerases (Promega); 20 U/pL (T7 and T3); and 80 U/pL (SP6). 10 BarnHI, XhoI, and HzndIII restnctlon enzymes (Boehnnger Mannhelm), 10 U/pi. 11. Sephadex G-50 spun-columns (see Note 1). 12 Hybrldizatlon solution. 50% formamide, 10% PEG-8000, 0 25M sodium phosphate buffer, pH 7.2,25 mA4NaC1,l mMEDTA, 20 pg/mL polyA, and 7% SDS Prepare fresh from stocks kept at room temperature, except polyA (-20°C) and formamide (4*C) Mix all other components and heat to 55°C before adding the SDS 13. Washing solution I. 1X SSPE and 0 1% SDS; store at room temperature 14 Washing solution II. 0.1X SSPE and 0.1% SDS; store at room temperature.

2.9. Secondary

Screening

by In Situ Hybridization

1. TlssueTek (Miles, Elkhart, IN) 2 Slide preparation: dip precleaned slides for 20 mm m lMHC1, nnse in water, dehydrate for 20 mm m 100% ethanol, and an-dry. Dip slides m freshly prepared 2% TESPA (3-ammopropyl triethoxysilane; Aldrich, Milwaukee, WI) m dry acetone for 10 s, wash twice m dry acetone, and once in water, and air-dry To mmlmlze organic waste, prepare as many slides as practical m one batch and store coated slides for at least 6 mo. 3 Fixative: 3% paraformaldehyde m O.lM sodium phosphate buffer (PB), pH 7.4 Prepare fresh as follows* to 30 g of paraformaldehyde add 400 mL of water, heat to 55°C for 5 mm under agitation, and add a few drops of 1ON NaOH. After solution clears add 500 mL of 0.2M PB, pH 7 4 and water to 1 L 4 Washing buffer fresh O.lM PB with 0.14M NaCl and 5 mA4 KCl, pH 7 2. 5. Acetylatlon solution: 1 4% triethanolamme and 0 3% acetic anhydride m water. Very unstable use within 10 s of mixing components 6 35S- or 33P-UTP (NEN). 7. Hybridization solution* 50% formamide, 2X SSPE, 2 pg/mL tRNA, 1 clg/mL BSA, 400 ng/mL polyA, and 100 mM DTT; prepare fresh from concentrated stocks kept at -20°C 8 Decovershppmg solution. 2X SSPE and 0.1% P-ME (added just before use). 9 Washing solution I: 50% formamide, 2X SSPE and 0 1% P-ME (added after heating to washing temperature and Just before use) 10. Washing solution II. 0.1X SSPE

2.10. Sequencing 1 Circumvent

Thennocycler Sequencing kit (New England BioLabs).

3. Methods 3.1. Song Stimulation 1 Isolate nine birds (see Note 2) acoustically for 1 d to minimize of mduclble genes.

basal expression

209

Song- Regulated Genes

2 Play tape-recorded song to SIX birds for 30 mm; decapitate three birds immedtately and three birds 2 h after start of the playback; use the three other birds that do not hear song as unstimulated controls. 3 Treat another nme birds as above (three per group) and use their brains for zn situ hybridization (Section 3 9.)

3.2. Tissue Dissection

and RNA Extraction

We base our protocol on Chomczynski modifications, as below:

and Sacchi (see Note 3), with minor

1 Dissect out NCM brain regions quickly and freeze. 2. Homogenize tissues m solution D (1 mL/30-40 mg of tissue) m an all-glass homogenizer at 4°C. 3 Add 0.1 mL of 2MNaOAc, 1 mL of water-saturated phenol and 0.2 mL of chloroform-isoamyl alcohol (49 I), with vortexing after each reagent. 4 Keep mixture on ice for 15 min and centrifuge at 10,OOOgfor 20 min at 4°C 5 Transfer the aqueous phase to a new tube and add 1 mL of isopropanol 6 Precipitate for 1 h at -2O’C and centrifuge at 10,OOOgfor 20 mm at 4OC 7. Dissolve pellet in 0 3 mL of solution D and reprecipitate for 1 h at -2O’C with 1 vol of isopropanol. 8. Wash the final pellet with 70% ethanol, dry m a speedvac, and resuspend m 10 pL of water. Typical yields per bird (right and left NCMs combmed) are 3-5 pg of total RNA, as estimated by spectrophotometry (see Note 1).

3.3. DNase Treatment 1. Incubate RNAs in 50-pL reactions with 1X DNase buffer, 20 U of DNase I and 20 U of RNasm (see Note 4) for 30 min at 37°C 2. Inactivate DNase by heating samples to 65°C for 10 mm and extract RNAs once with 1 vol of phenol:chloroform. 3. Precipitate RNAs at -20°C for 30 mm after adding l/10 vol of 3MNaOAc and 2 vol of 100% ethanol to the aqueous phase. 4. Pellet RNAs by centrifugation, wash in 70% ethanol, dry, and resuspend m 10 p.L of water.

3.4. cDNA Synthesis Incubate RNAs in 20-& reactions containmg 1X RT buffer, 5 @4 DTT, 20 l1A4 dNTPs, 1 @t4 of one TlzMN oligonucleotide primer (see Note 5), 100 ng of DNase-treated RNA (preheated at 65OC for 10 min), 20 U of RNasin and 100 U of Superscript II RT (see Note 6) at 37°C for 30 min (see Notes 7 and 8 for further comments). Run control reactions without adding RT.

3.5. PCt? Amplification 1. Amplify each cDNA in a IO-& reaction containing 1X PCR buffer I, 2 pA4 dNTPs, 0.2 pA4 of one AP-primer, 1 @4 of the T&IN primer used m Section 3 4., 1 pL of the cDNA synthesis reaction, 5 uCi of 35S-dATP, and 0 4 U of

Mello et al.

210

Fig. 1. DD group). RNAs PCR-amplified is high in two

gel comparing silent controls and song-stimulated zebra finches (30-min were isolated from NCM brain regions and cDNAs synthesized and using T,,MC and AP-3 primers. The band indicated by the arrow (ZF9) out of three stimulated birds and in none of the controls.

AmpliTaq (see Note 8). Program the thermocycler for 40 cycles of 94°C for 30 s, 40°C for 2 min and 72°C for 30 s (see Note 9), followed by 72°C for 5 min, and then hold at 4’C. Run control reactions without adding cDNAs. 2. Add loading buffer (5 pL per reaction) and run 7 pL of each final mix on a 6% polyacrylamide gel (see Note 1). 3. In parallel, prepare a radiolabeled DNA ladder in a 20-pL reaction with 25 ng of the lOO-bp DNA ladder, 0.5 mA4dCTP, dGTP, and dTTP each, 1 pCi 35S-dATP, 1X Klenow buffer and 1 U Klenow DNA polymerase, incubated at room temperature for 10 min. Add 10 pL of loading buffer and run 3 pL of the mix on the DD gel for size determination. 4. Dry the gel without fixing and expose to X-ray film for l-3 d.

3.6. Fragment

Excision and Reamplification

1. Cut from gel the PCR fragments present in at least two of three birds of one songstimulated group and absent in all three birds of the control group, or vice-versa (see Fig. 1 and Note 10). 2. Elute by boiling fragments in 100 p.L of water for 15 min and transfer supematant to a new tube. 3. Precipitate at-70°C for 30 min after adding 10 pL of 3MNaOAc, 5 pL of 10 mg/mL glycogen, and 450 pL of 100% ethanol. 4. Pellet fragments by centrifugation, wash in 85% ethanol, dry, and resuspend in 10 j.iL of water.

Song-Regulated

Genes

211

CM

CM 28 18-

2% 18-

_

5.8 5.8 . Fig. 2. Northern analysis of a differentially expressed PCR fragment. Singlestranded DNA probe (see Note 13) synthesized from a candidate song-induced fragment was used to hybridize a blot containing total brain RNA from birds killed 30 min after injection of metrazole (M; see Note 12), or from untreated controls (C). Differential expression of the upper band (left panel, arrow) confirms its activation by neuronal depolarization; a second, nonregulated, band illustrates the issue of heterogeneity of DD fragments. The experiment was ~repeated and, after a more careful isolation of the fragment from the DD gel, the contaminating band could be eliminated (right panel). 5. Reamplify 4 pL of eluted fragments as in Section 3.5., step 1, but scale up the reaction to 40 pL+ use 20 @4 of dNTPs, and add no isotope or cDNA. 6. Use half of the reaction for plasmid cloning (see Section 3.7.). Analyze the other half on a 2% Nusieve agarose gel using the lOO-bp ladder for size determination. 7. Cut fragments from the gel, extract with the Qiaex II kit and use for generating probes to confirm cloning into plasmid vectors.

3.7. Cloning

info PCR Plasmid Vectors

1. Clone reamplified PCR products of the correct size into pCRScript or TA pCR II vectors (see Note 11). 2. Check transformants for insert size by running plasmids restricted with EcoRI or BssHII, respectively, on 1% agarose gels next to the lOO-bp ladder. 3. Blot the gels (see Note 1) and confirm identity of transfonuants by hybridizing blots with 32P-labeled random-primed probes generated from the corresponding Qiaex II-purified PCR reamplitication products (see Note 11).

3.8. Secondary

Screening

by Northern

Hybridization

1. Run 10 pg of total RNA from control and experimental birds on 1% MOPS/ formaldehyde agarose gels (see Notes 1 and 12). 2. Soak gels for 1 hr in 20X SSPE and blot onto nylon membranes. 3. Synthesize sense and antisense riboprobes (or single-stranded DNA probes; see Note 13 and Fig. 2) in a 10-pL reaction containing 50 pCi 32P-UTP, 1X transcrip-

212

Mello et al.

tton buffer, 12 U of RNasm, 5-10 pg/mL of BSA, 0 5 to 1 ug of restricted plasmid DNA (from transformants obtained in Section 3.7., step 1) and 20 U of the appropriate RNA polymerase (T7 for BarnHI- and SP6 for XhoI-restricted TA plasmrds, and T7 for BarnHI- and T3 for HindIII-restricted pCRScrrpt plasmids), Incubated at 42°C for 1 h (see Note 14) 4 Prehybridize filters for 1.5mm at 55°C in hybridtzatton solutron (2 mL/lO cm*) 5 Add rlboprobes purified through G-50 columns (1 06counts/mm/mL) and hybridlze overnight at 55°C 6 Wash filters twtce for 5 mm at room temperature m washing solutton I, twice for 30 mm at 55’C m washing solutron II, and expose to X-ray film for 1 to several days

3.9. Secondary

Screening

by In Situ Hybridization

(see Note 72)

1 Dissect out brains, place m plastic molds with TissueTek and freeze on dry me 2 Cut 10+-n sections on a cryostat and mount onto TESPA-coated slides 3 Incubate sections m fixattve for 5 min at room temperature, rinse twice in washing buffer, dehydrate m an ethanol series (70,95, and lOO%, 2 mm each), an-dry, and store at -70°C 4 On the day of hybrrdlzatron, incubate sections for 10 mm m acetylatton solutton, rinse three times m 2X SSPE, dehydrate m ethanol, and an-dry 5 For each cloned fragment, make rtboprobes m both orientations as described for Nor-therm (Section 3 8 , step 3), except use 35S- or 33P-UTP (see Note 15), purify probes through G-50 columns. 6. Cover acetylated sectrons with 16 pL of hybridization solution containing rtboprobe (0.5-l x lo6 counts/mm), add cover slips, and incubate under mineral 011m a 5&55’C water bath for 3 h. 7 Remove the oil with two washes in chloroform and the covershps by dipping in “decovershppmg” solutron 8. Wash slides for 1 h at room temperature m fresh decovershpping solutton, followed by 1 h at 50-55°C m washing solutron I and 2 x 30 mm at 50-55°C m washing solutron II 9. Dehydrate slides m the ethanol series, au-dry, and expose to X-ray films for at least a week (see Note 16)

An example of in sztu hybridization of a differentially expressed clone IS shown m Fig. 3. Fragments that show no differential expression (no difference in signal between song-induced and control brains) for either rtboprobe strand should be drscarded

as false positives.

For fragments

that show

no signal for

either strand, or similar signal for both strands, repeat hybridization under different conditions (see Note 17) 3.70. Sequencing Analyze clones verified to be differentially expressed by sequencmg and subject results to homology searchesagainst GenBank database (see Note 18).

Song-Regulated

Genes

213 Silence

Song ;

-

,.

Fig. 3. In situ hybridization analysis of a song-induced DD fragment. Parasagittal brain sections from song-stimulated (right column) or unstimulated control (left column) finches were hybridized with 35S-labeled riboprobes and exposed to X-ray film. (A) The ZENK gene is induced by song with a restricted pattern, in particular in NCM (oval area shown by arrow); (B) the candidate ZF9 DD fragment is induced by song in a less restricted fashion than ZEIVK, (C) pCF-2, a cDNA that is not regulated by neuronal depolarization, gives the same signal in both groups and serves as a negative control. Only antisense strands are shown. Orientation: rostra1 is to the right and dorsal to the top. Bar = 1.5 mm.

4. Notes 1. See Sambrook et al. (11) for details on basic techniques, such as work under RNase-free conditions, spectrophotometry, preparation of agarose and polyacrylamide gels, DNA blots, and G-50 spun columns. 2. Interanimal difference is a main source of PCR variability and false positives in DD. To address it, we use multiple animals per condition and pick fragments that are consistent across animals (see Note 10). We recommend three animals per condition as a minimum; a higher number may be disadvantageous, considering the work involved in performing several primer combinations. 3. Chomczynski and Sacchi’s method (12) gives, in our hands, reliable preparations of intact brain RNA. However, resuspension of the final RNA pellet in 0.5% SDS is not recommended, since traces of SDS can partially inhibit RT and DNase enzymes and increase PCR variability. For complete resuspension of RNAs in water, heat samples to 80°C for 10 min. For tissue samples with very high RNase content, or when a long time is required for the precise dissection of small brain nuclei from fresh slices, partial degradation of RNAs may be unavoidable. Two

214

4.

5

6 7

8.

9

10

Mello et al. studies used the protocols described m this chapter successfully under such conditions* the isolation of cDNAs enriched m song control nucleus HVC, or high vocal center (13), and the use of human skm biopsies for the tsolation of cDNAs differentially expressed in psoriasis (14). Partially degraded samples can yield differentially expressed bands, but signals are weaker and fragments of ~400 bp tend to be lost RNasm helps prevent RNA degradation owmg to contaminating protem carried over from the RNA isolation. If too little material is available (11 pg of total RNA per sample), it may be wise to sktp the DNase step m order to prevent RNA loss during the subsequent prectpttation In our hands, using T,,GC or T,,MC (where M = A, G, or C) for cDNA synthesis and PCR amphfication results m similar number of PCR fragments Thus, the number of potential fragments is underesttmated when the degenerate primer IS used We suggest that each one of the 12 possible T,,MN primers (where M = A, G, or C and N= A, G, C, or T) be used separately Preheating of RNAs, addition of RNasm and the use of Superscript II RT are recommended m order to maximize yield of long transcripts Reducing the amount of RT enzyme (to 5 U) and/or duration of the RT reaction (10 mm) tmproves the resolution of PCR fragments m DD gels m case of smearmg In our experience, use of >500 ng of total RNA for cDNA synthesis also leads to loss of resolution of fragments in DD gels; 5&200 ng of total RNA are sufficient to generate reliable DD products and is the range we recommend As m Liang et al (15), we find that there can be some variability among repeated PCR amphfications with the same sample This could be owing to factors such as contammatton of tubes or pipet tips, differences m pipetmg, false primmg, and so on We also find that shght variations m salt concentration or spurious DNA contammations of different solutions or preparations of DEPC-treated water can affect the resultmg display pattern In addition to using multiple independent samples (see Notes 2 and lo), it is essential to process all samples simultaneously and consistently, with the same stock soluttons and reagents. It is also advisable to perform pipetmg by poolmg common reagents and then dividing them mto mdividual tubes, before adding the specific RNAs and primers to each reaction. If using Stratagene’s robocycler or a similar PCR machme, it is advisable to adjust the cycles to 94°C for 45 s, 40°C for 3 mm, and 72°C for 45 s to compensate for faster temperature transitions When performing repeated DD reactions on the same set of multiple independent samples we noticed that some PCR products consistently appear m some mdividual animals, but their occurrence does not correlate with any particular group. Such fragments could represent genetic differences among mdtvtduals, or differences m gene expression levels m the brain due to variables dtfficult to control for, such as stress, arousal state, hormone levels, or differences m the precise anatomical dissection To avoid spurious fragments due to these factors and to those discussed m Note 8, we use multiple animals per group and pick fragments that are consistent across a given group

Song-Regulated

Genes

11 Protocols as provided by supphers The TA pCR II vector 1s stralghtforward to use after Tuq amphficatlon, but its SP6 promoter IS not very efflclent for rlboprobe synthesis Stratagene’s pCRScrlpt IS our vector of choice, although treatment with pfu enzyme (as described by supplier) IS necessary to clone fragments amplified with Taq Fragments obtamed in Section 3 6 are sometlmes heterogeneous, due to contammatlon with adjacent bands from DD gels (Fig 2) or to mtrmslc heterogeneity of DD bands (I 6) It 1sthus advisable to pick several Independent transformants for sequence and expresslon analysis 12 Northern blots have hmlted use when tissue avallablhty 1s restricted Furthermore, for tissues with great diversity of cell types and mRNA species as the brain, dlfferentlal expression m a discrete cell population might go unnoticed on blots To address tissue availability in our system, we have used bram RNAs from bn-ds treated with metrazole, a potent depolanzmg agent that leads to widespread Induction of activity-dependent genes m the brain (Fig 2), mcluding c-gun (3) and ZENK (I 7) In sztu hybndlzatlon, however, IS our method of choice for secondary screening, smce small amounts of tissue are required and expression easily detected at the single cell level. 13 As m Llang et al. (15), only 40-&O% of our DD fragments result m clear slgnal when random-pnmed DNA probes are used for Northern blots. In contrast, when we use riboprobes and the protocol in Section 3.8 , modified from Clayton et al (18), >90% of the fragments give detectable signal (even fragments of - 100 bp) AlternatIvely, smgle-stranded DNA probes are also satisfactory and avoid the need to clone PCR products Sense and antlsense DNA strands are synthesized each m a 20 @, reactlon contaming 1 pL of reamphfied PCR product, 1 pL of one of the primers used in Sectlons 3 4 and 3.5 (T12MN for one strand or APprimer for the other strand), 25 n-&f dATP, dGTP, and dTTP each, 50 $1 32PdCTP, 1X Klenow buffer and 1 U of Klenow DNA polymerase, incubated at 37°C for 1 h After punficatlon through G-50, probes are hybndlzed (m 50% fonnamide, 2X SSPE, 1% SDS, and 10% dextran sulfate) to Northern gels blotted onto nylon membranes, the recommended temperature for hybrldlzatlon and washes (m 0.2X SSPE and 1% SDS) 1s42’C Figure 2 illustrates the use of such probes and the issue of heterogeneity of amplified fragments 14. To maximize nboprobe synthesis, add 0.5 pL of the appropriate RNA polymerase at the start of incubation and 0 5 $ after 30 mm For SP6, the total amount added per reaction corresponds to 80 U 15. When using 35S-UTP, it 1s essential to keep probes reduced to mmnnlze background This IS done by addmg 0 5 cls, of IA4 DTT to the riboprobe syntheses reaction, in addition to the DTT m the buffer stock. Also add 1 pL of 1M DTT to each probe after G-50 purlficatlon and after each freeze/thaw cycle, If stormg probes for multiple experiments We have lately turned to 33P-UTP, which reduces exposure time and avoids probe oxldatlon Use 5 p.L of a 1 5 dllutlon of 33P-UTP (as provided by supplier) m the 10-K nboprobe synthesis reaction. 16 If accessible, a Phosphorimagmg system cuts down on m sztu exposure times slgmficantly Its resolution 1snot nearly as good as that of X-ray film autoradiography, but it

216

Mello et al.

allows a fast quality assessment of the hybridization and an mmal evaluation of the differential expression of probes tested. For tine analysis at the cellular level, dip slides m autoradiographic emulsion (NTB-2, Kodak) and counterstain with cresyl violet (4,). 17. The optimal conditions for zn sztu vary from probe to probe, owing to differences in size and GC content. We have modified our basic in situ protocol (I 7) by lowermg the temperature of hybridization and washes to compensate for the small size of DD fragments relative to the large riboprobes (l-2 kb) for which the protocol was originally designed A tight control over the temperature and timing of washes is essential, since a small difference (2-3”) can vastly affect the ability to detect signal Washing solutions should be preheated and times counted after temperatures (measured inside washing solutions and after addition of slides) reach the appropriate level. For probes in the 250-500-bp range, we recommend 55OC for hybridization and washes. For smaller probes, 45-55’C is recommended When no signal is detected for either strand, repeat the in sztu lowermg the temperature m 5” steps, until signal appears for only one strand. The differentially expressed clone shown in Fig. 3 (19), for example, showed no signal with either strand m the 6065°C range, specific hybridization (signal with only one strand) at 55”C, and background (hybridization with both strands) at 50°C. Fragments with no detectable signal even at low stringency are not processed further. When signal is equally detected for both strands, repeat hybridization at higher temperatures, until signal remains for only one strand. We try to avoid RNase treatment because of the poor resulting histology; however, an alternative protocol using RNase is very effective at eliminating residual background that may persist after high temperature washes (3). 18. GenBank homology searcheswith sequences from DD fragments obtained m zebra finches have not been very informative. This is particularly the case because DD fragments are typically (but not necessarily) representative of 3’ untranslated regions. When at least a partial open reading frame is present, homologs from other species are more readily identified (13). In most cases,though, estabhshmg the identity of the differentially expressedPCR fragments will require screening cDNA libraries for larger clones.

Acknowledgments The birds used in this study came from the colonies kept at the Rockefeller University’s Field Research Center, in Millbrook, NYS. Financial support provided by the Mary Flagler Cary Charitable Trust, B. Altman, PHS grant MH 18343 and NIMH’s training grant MH15 125- 17Al.

References 1. Liang, P. and Pardee, A B. (1992) Differential display of eukaryotic messenger RNA by means of the polymerase chain reaction. Science 257,967-97 1 2. Mello, C. V., Vicario, D. S., and Clayton, D F. (1992) Song presentation mduces gene expression m the songbird forebrain Proc Nat1 Acad Scz USA 89,68 18-6822. 3. Nastmk, K. L., Mello, C. V., George, J. M , and Clayton, D. F. (1994) Immediateearly gene responses in the avian song control system: clonmg and expression analysis of the canary c-jun cDNA. Mel Brain Res 27,299-309.

Song-Regulated

Genes

217

4 Mello, C. V. and Clayton, D F (1994) Song-mduced ZENK gene expressron m auditory pathways of songbird brain and its relatron to the song control system J Neurosci 14,6652-6666. 5 Mello, C. V , Nottebohm, F , and Clayton, D F. (1995) Repeated exposure to one song leads to a rapid and persistent decline m an immediate early gene’s response to that song m the zebra finch telencephalon J Neuroscz 15,6919-6925 6. Chew, S. J., Mello, C.V , Nottebohm, F , Jarvls, E and Vicarlo, D. (1995) Decrements m auditory responses to a repeated conspecific song are long-lasting and require two perrods of protein synthesis in the songbird forebrain. Proc Nat1 Acad Scz. USA 92,3406-3410 7 Jarvrs, E. D., Mello, C. V., and Nottebohm, F (1995) Associative learning and strmulus novelty influence the song-induced expression of an munedrate early gene m the canary forebrain. Learnzng and Memory 2,62-80 8. Milbrandt, J (1987) A nerve growth factor-induced gene encodes a possible transcrrptronal regulatory factor Science 238,797-799 9. Nrshimura, T. and Vogt, P (1988) The avlan homolog of the oncogene Jun Oncogene 3,659-663 10. Mello, C. V. (1995) Immunocytochemlcal analysis of ZENK gene mductton by song stimulation in the zebra finch brain. Sot Neuroscz Abstr 21(2), 959 11. Sambrook, J., Frrtsch, E. F , and Mamatrs, T (1989) Molecular Clonzng A Laboratory Manual, 2nd ed. Cold Spring Harbor Laboratory, Cold Spring Harbor, NY 12 Chomczynski, P. and Sacchi, N (1987) Single-step method of RNA rsolatron by acid guanidmmm thiocyanate-phenol-chloroform extraction. Anal Bzochem 162, 156-159. 13 Denisenko, N., Nottebobm, F. and Mello, C (1995) PCR-based mRNA differential display reveals enrichment of aldehyde dehydrogenase in the high vocal center and m two other nuclei of the song system of songbirds. Sot Neuroscz Abstr 21(2), 960 14. Rrvas, M. V., Jarvrs, E D , Mortsakr, S., Carbonaro, H., Gottlieb, A B , and Krueger, J. (1997) Identification of aberrantly regulated genes m diseased skm using t&e cDNA differential display technique. J Invest Derm 108(2), 188-194 15. Lrang P L., Averboukh, L., and Pardee, A. B. (1993) Distribution and clonmg of eukaryotic mRNAs by means of differential display. refinements and optrmrzation. Nucleic Acid Res. 21,3269-3275 16. Shoham, N. G., Arad, T., Rosin-Abersfeld, R., Mashrah, P., Gazrt, A., and Yamv, A (1996) Differential display assay and analysis BioTechnzques 20, 182-184 17. Mello, C. V. and Clayton, D F. (1995) Differential induction of the ZENK gene within the avian forebrain and song control circuit after metrazole-induced depolarization. J Neurobzol 26, 145-161 18 Clayton, D. F , Huecas, M. E , Sinclair-Thompson, E Y , Nastiuk, K L , and Nottebohm, F (1988) Probes for rare mRNAs reveal distributed cell subsets m canary bram. Neuron 1,249-261. 19 Mello, C. V , Jarvis, E., Demsenko, N., Rivas, M , and Bamea, A (1994) Isolation of song-induced genes m zebra finches by mRNA differential display. Sot. Neuroscz. Abstr 20(l), 164.

Identification of Vertebrate Circadian Clock-Regulated Genes by Differential Display Carla B. Green and Joseph C. Besharse 1. Introduction Differential display is ideally suited for the identification of genes involved m complex physiological events. Unlike subtractive hybridization techniques, differential display requires no previous knowledge about the dynamics of “important” gene products Since the only limitation is the number of samples that can be run on a smgle sequencing gel, comparison of many subtly different conditions can be assessed,and mRNAs that vary m interesting ways can be identified. An example of such a use 1sthe identification of gene products that vary m levels of expression accordmg to the time of day, under the control of a circadian clock. Circadian clocks are internal timekeeping mechanisms that regulate diverse physiological events known as circadian rhythms (2). Circadian rhythms have periods of about 24 h and are sustained even m constant conditions. Circadian rhythms include behavioral rhythms (sleep/wake cycles, locomotion), biochemical rhythms, rhythms in gene transcription, and are found m organisms ranging from bacteria to humans (2,3). Although circadian rhythms are widely studied m many species, little IS known about the molecular mechanism of the underlying clock. However, one of the ways that the clock regulates rhythmic physiological events is through control of rhythmic gene expression (2). Identification of mRNAs that are expressed rhythmically, under the control of the clock, is an important approach m uncovering the molecular mechanisms of circadian control. One excellent system with which to address this question is the Xenopus Zaevis eye. The photoreceptor layer within the retma contains an endogenous circadian clock that controls many aspects of retinal physiology (4-6). MoreFrom

Methods m Molecular Bology, Vol 85 D/fferenbal Ed&d by P Liang and A 6 Pardee Humana

219

D/splay Methods and Profccols Press Inc , Totowa, NJ

220

Green and Besharse

over, the Xenopus eyes can be removed and cultured m vitro, mamtammg many aspectsof normal physiology, including clock function and the rhythmic events that the clock regulates (4,6). We have used differential display to Identify mRNAs that are expressed rhythmically m these cultured eyecups in order to gain mslght mto the molecular mechanism of this circadian clock (7,8) 1.1. Experimental Design Considerations Before begmning, It ISImportant to decide on the specific goal of the expenment. In our case, the goal was to identify mRNAs that are regulated by the circadian clock m the retina. Therefore we used cultured eyecups so there would be no influence of an extraocular clock, and we cultured the eyecups m constant darkness throughout the experiment so that we would know that rhythmlczty was owmg to the clock and not to an acute effect of light or darkness, An additional benefit of this design is that messagesidentified should mamtam robust rhythms m vitro. This 1simportant because it makes them suitable for future study m our circadian culture system. Because one of the inherent problems with differential display IS the high rate of false positives that are time-consummg to track down (9,lO), extra attention at this step can save much time at later stages. In our experiment, the starting material for producmg the RNA for the differential display 1s retinal tissue taken from many different animals. This produces the problem of ammal-to-animal variability that may produce different bands on the display gels that are superimposed on differences resulting from circadian time This problem ~111always be present when asking questions about complex systems, which can not be done using cell lines To cut down on this vanablllty, we used many retinas for RNA Isolation for each time-point. In addition, at each time-pomt retinas from an equal number of animals were used, while ensuring that the two eyes from an individual animal were always included m different groups. Criteria for selection of messages of interest must be clearly defined and sufficiently discriminating to rule out false positives. For example, if we were interested in light-induced messages we could compare RNA isolated from eyecups exposed to a range of different light intensities, or a range of times after light onset. Dlfferentlal displays on all the samples could be run at once with the expectation of a graded effect of intensity or time. However, it 1s important to have some mechanism to allow one to disregard artlfactual bands, such as including RNA samples from duplicate, but independent experiments. In the experiment described here, our approach was to compare displays of RNA from time-points on two consecutive days. This resulted m a particularly strmgent selection m favor of rhythmic messages and against spurious dlfferences m the RNA samples or PCR reactions.

Clock-Regulated

Genes

221

2. Materials 1. Xenopus eyecup culture medium: This medium is made fresh before each use from the followmg stock solutions’ a. 10X Salts: 825 mM NaCl, 20 mM KCl, 350 mM NaHCO,, 50 mM glucose, 10 mMNaH2P0, Filter sterilize and store at 4°C for l-2 mo b. 10X Amino acids 4 9 mMarginine, 0.9 mMcystine, 8.4 nnVglycme, 1.6 mM histtdine, 3.0 mM isoleucine, 3 0 mM leucine, 3.3 mM lysme, 0.8 mM methionme, 1 6 mM phenylalanme, 3.3 mM threonine, 0.4 mM tryptophan, 3.3 mMvaline, 1 6 mMtyrosine in water (dissolve the cystine and the tyrosine m a small amount of 0. 1NNaOH and then add to the rest). Filter sterilize and store at 4°C for l-2 mo c 10X Divalents: 18 mA4CaCl,, 10 mMMgC1, Filter sterilize and store at 4’C for l-2 mo d. 100X 5-HTrp/ascorbate: Dissolve 0.0176 g ascorbate and 0.0256 g 5-hydroxytryptophan (5-HTrp) in 10 mL water. Make this fresh each time. We usually weigh out the chemicals into dry tubes and store them at 4°C At the time of medium preparation, 10 mL of sterile water can be added to quickly make the 100X solution (see Note 1) To make 1 L of medmm. Combine 679 mL of sterile water, 1 mL of 0.5% phenol red, 100 mL 10X salts, and 100 mL 10X ammo acids. Equilibrate with 5%CO,/ 95% 0, by bubbling gas through the medtum unttl the color changes to an orange/ red color (pH 7 4). After the medium is equihbrated to the correct pH, add 100 mL of the 10X dtvalents (see Note 2), 10 mL of 100X pen/strep solution (100X is 10,000 U/mL penicillin G sodium, 10,000 pg/mL streptomycm sulfate), and 10 mL of 100X 5-HTrp/ascorbate. 2. Culture medium + 10X anttbiottcs: identical to the Xenopus eyecup culture medium, except wtth 1O-fold more pen/strep added ( 10 mL of 100X pen/strep per 100 mL of medmm). 3. TRIZOL reagent: Available from Gibco-BRL (Grand Island, NY). 4. DEPC water: Rid water of RNAse by treating with drethylpyrocarbonate (DEPC) at a final concentration of 0 1% for 30 min at 37°C Remove DEPC by autoclaving 5. Message clean kit Available from GenHunter (Nashvtlle, TN) 6. RNAmap ktt: Available from GenHunter 7. Primers: We used primers from GenHunter and also designed and used some custom made primers, followmg the same general primer design (9). 8. pCRI1 vector/TA cloning kit* Available from Invitrogen (Carlsbad, CA) 9. Random pnmers DNA labeling system: Available from Gibco-BRL. IO. RNA Transcription Kit from Stratagene: Available from Stratagene (LaJolla, CA) 11. NucTrap gel filtratton columns: Avatlable from Stratagene.

3. Methods

3.1. Experimental Set Up Our experimental design for the identification of mRNAs regulated by a circadian clock is shown in Fig. 1A (7). Eyecups were placed in culture in light

Green and Besharse

222

dissect and prepare eyecups

harvest retinas and collect culture media at 6 hour intervals under Infrared light

melatonin assay to verify rhythmwty

o-! 0

/

6

12

18

Isolate RNA for differential display

0

6

12

18

ZeltgeberTme

Fig. 1. Experimental paradigm for high-strmgency dlfferentlal display for isolation of clock controlled mRNAs (A) Eyecups (mciudmg the retina, pigment eplthelmm, chorold, and sclera) were prepared (see Fig. 2) late m the afternoon and cultured m constant darkness, begmnmg at the time of normal dark onset (ZT 12). Begmning the next mornmg (ZT 0), retinas were harvested and culture medium was collected under infrared light, at 6-h Intervals, for 2 d Harvested retinas were frozen quickly on dry ice for subsequent RNA lsolatlon and differential display analysis White bar mdlcates light period (normal daytlme) Hatched bars indicate SubJective daytime (in constant darkness), while black bars indicate subjective night Numbers above bars indicate time of harvest All times are specified accordmg to the animals normal hght cycle, where light onset was defined as Zeltgeber time 0 (ZT 0) and dark onset was ZT 12.

Clock-Regulated

Genes

223

on the afternoon before the onset of the experiment. At the time of normal dark onset m the animal holding facility, eyecups were transferred to darkness. They remained in constant darkness at a constant temperature (2 1“C) throughout the remainder of the expenment. At the time of normal light onset (the next day), the first sample of retinas was removed. Retinas and samples of culture medmm were collected at subsequent ttme-points every 6 h for 2 d. Protocol: 1 Prepare dishes (one per time-point) of culture medium (10 mL of medmm m 60-mm tissue culture dishes) and place m an incubator with an atmosphere of 95%0,/ 5%C02 so that the medium remains gassed Begin the dissections m the afternoon, leaving enough ttme to fimsh all the dissections before the time of normal dark onset for the ammals 2 Sacrifice and pith an entrained adult Xenopus Zuevzs(see Note 3). Remove the eyes and transfer to a 30-mm dish containing 4 mL of culture medium + 10X antibiotics and wash briefly Transfer the eyes to large drops of fresh culture medium + 10X antibiotics on a wax dissecting surface (see Note 4) 3 Under a dissecting microscope, using microdissectmg sctssors and forceps, carefully remove all fragments of skm and tissue remammg on the eyes. Transfer again to medmm + 10X antibiotics for a final wash Transfer to fresh drops of normal culture medmm (1X antibiotics). 4. Insert the point of the tine dissection scissors mto the eye at a pomtlust outside the edge of the iris and cut around the outside of the iris, removing the anterior portion of the eye, mcludmg the lens as illustrated m Fig 2 All these mampulations are done under a dissectmg mtcroscope. The result of this dissection IS an eyecup including the retma, pigment epithelmm, choroid, and sclera. Transfer the eyecups mto culture medium on a rotary shaker (60 rpm) in an atmosphere of 95%0,/5%CO, m a constant temperature incubator at 21 + 0 1°C (m bright hght) 5. Repeat steps 24 with all the other animals, randomizmg the eyecups mto different dishes so that two eyes from a pair are never m the same group 6. At the time of normal dark onset, transfer the dishes of eyecups mto black boxes and return to the Incubator (see Note 5). The eyecups will be incubated m constant darkness throughout the remainder of the experiment

(B) Melatonin release from the cultured eyecups is rhythmic over the entire culture period. The times mdicated on the graph represent the time of collection of the media samples; therefore the melatonin measurement actually reflects an average for the 6 h preceding that time. Thus, compared to continuous plots of melatonm release (16), the rhythm is shifted to the right. Each lme on the graph represents media samples from one set of 10 eyecups. Because one set was harvested at each time-point, each subsequent time-point contains one less group Reprinted from ref 7 with permission of Elsevier Science

224

Green and Besharse

#

I

EYE

RETINA cEYECUP

Fig. 2. Preparation of Xenopus eyecups. Xenopus laews that have been entrained to a light/dark cycle are sacrificed by decapitation, followed by pithing, and the eyes are removed. Fme dissection scissors are used to puncture the eye Just below the edge of the iris and the eye is then cut around the outside of the iris, such that the anterior portion of the eye (mcludmg the cornea, iris, and lens) is removed. The remammg “eyecup” containing the retma can then be cultured. 7 The first time-point is taken 12 h later, at the time of normal light onset At this time, collect (and save) the culture medium in each dish, replace with fresh culture medium, and return all dishes (except for dish 1) to the black boxes m the incubator Dissect the retinas from the eyecups in dish 1 and transfer to a microfuge tube on dry ice for rapid freezing. All manipulations, medium changes, and dissections should be done in the dark under infrared hght (see Note 6). Store frozen retinas at -80°C and medium samples at 4’C 8 Repeat step 7 (both the collection and replacement of medium and the isolation of retinas) every 6 h for the next 2 d. Be very careful that the eyecups/retinas are not exposed to light at any time during this period. 9. At the conclusion of the experiment, the medium samples can be assayed for melatonin as a way to verify that the eyecups were alive and contained a functional clock throughout the culture period (Fig 1B; see Note 7).

3.2. Extraction

of R/VA from Xenopus

Retinas

1. Extract RNA from each group of retinas using TRIZOL reagent (see Note 8). 2. At the end of the extraction protocol, dry the final RNA pellet briefly and redissolve m 50 ~.JLof DEPC water by warming at 65’C for a few min. 3. Treat the RNA samples with 10 U DNAse I at 37°C for 30 min to remove any contaminating DNA using the Message Clean kit from GenHunter. Following the DNAse treatment, extract with phenol/chloroform and ethanol precipitate as described in the kit. 4 Dissolve the pellets m 25 pL of DEPC water at 65°C

Clock-Regulated

Genes

225

5 Determine the concentration of each RNA sample by dilutmg a small ahquot and measuring the Absorbance at 260 nm Adjust the volume wtth DEPC water so that all are at the same concentratron (see Note 9)

3.3. Differential

Display

1 Perform mRNA differential display reactions usmg the RNAmap ktt from GenHunter (see Note lo), using 0 2 pg of retinal RNA for each reverse transcnptton reaction. 2 Include one set of reactions that are identrcal, but that do not receive any reverse transcrtptase This will verify that differential display bands are not the result of amphfication of contaminating genomtc DNA No bands should be seen m these lanes if the RNA 1s free of genomtc DNA. 3 Run a 3 5-p-L sample of each PCR reactton on a 7% sequencmg gel. Dry the gel and expose to X-ray film wtth mtenstfymg screens for l-5 d.

3.4. Analysis

of Results

1 Compare the bands in all eight lanes and look for those that show the same temporal pattern on d 1 as on d 2 of the trme-course 2. Excise the candidate bands from the gel and reamphfy as descrtbed m the RNAmap kit 3. Clone the resulting reamphfied PCR products into pCRI1 vector using the TA clonmg kit from Invitrogen.

3.5. Is the Candidate

Band Interesting?

Once a candidate band 1s identified by differential display, the real work starts. Now the band must be cloned, sequenced, and verified by some tndependent tests. These steps can be much more time-consumtng than the ortgmal display. However, there is a high chance of success in this case because of the high stringency of the initial screen. Because of the experimental design, we begin with evidence that the band 1s differentially expressed, rhythmic in constant darkness, sustained in culture, and controlled by a endogenous retinal clock. For our purposes, we immediately test a candidate band three ways: 1. We use the cloned PCR product to make probes for Northern blot analysis of RNA samples from a similar, but independent experiment (see Note 11) This allows confirmatron of rhythmrctty and also gives mformatton about the size and abundance of the transcripts 2. Because we are parttcularly mterested in the cucadtan clock that resides m the photoreceptors and how this clock controls photoreceptor physiology, we chose to use localization as a secondary screen. For these studies, we use the cloned PCR product as a probe for zn sztu hybridization (ISH) to determine whether the candidate IS expressed in photoreceptors (7) 3 If tests 1 and 2 indicate that the candidate 1s of interest, we use the dtfferentlal display product as a probe to screen a retinal cDNA library for isolation of fulllength clones for sequencing and further analysis

Green and Besharse

226

0

6 1218

0

6 12 18

T71

Fig. 3. Candidates for clock-controlled mRNAs exhibit the same differential display pattern on both days in culture. Differential display analysis was performed simultaneously on 8 RNA samples prepared from retinas harvested at 6-h intervals over 2 d in constant darkness (see Fig. 1) and separated on a 7% sequencing gel. Sections of a differential display that contains a candidate clock-controlled mRNA is shown. The arrow marks the position of the candidate band. The size of the T71 band is 445 bp. This band fits the criterion for further analysis based on its consistent temporal pattern. The primers used to generate this display were 5’-CCGAAGGAAT-3’ and 5’-(T)12MT-3’. The position marked by M is degenerate, consisting of A, C, and G. The numbers above the lanes correspond to times of harvest (Zeitgeber time) as in Fig. 1. Dark bars represent subjective night; light hatched bars represent subjective day. Reprinted from ref. 7 with permission of Elsevier Science.

3.6. Example of a Result: Nocturnin Was identified by Differential Display Of the more than 2000 bands examined to date, we have identified four that show a consistent temporal pattern of expression on both days (7). One of these candidate differential display bands (Fig. 3) is present at the time of normal dark onset (ZT12) on both days and is not detectable at the other times. This band (originally named T71) was subsequently excised, reamplified, cloned, and used to generate radiolabeled probes for Northern blot analysis. An example of a representative northern blot is shown in Fig. 4. This probe recognizes a pair of mRNAs (3.8 and 2.2 kb) that are expressed at peak levels at dark onset, the same temporal pattern predicted by the original display gel. Further analysis and quantitation of message levels, under various conditions, showed that these messages are expressed for only about 4 h in early night in eyecups cultured in cyclic light or in constant conditions, indicating that this expression is controlled by the circadian clock (7,8). This result confirms that the differential display procedure was successful in identifying a bona fide clock-controlled gene. We named this gene product “noctumin” for night factor.

227

Clock- Regulated Genes

I 0

6 12

18 0

6 12 18

1-3.8 kb

nocturnin

pactin

1-2.2 kb

llvmu

Fig. 4. The expression pattern of the candidate band is confirmed by Northern blot analysis. A northern blot of retinal RNA isolated at 6-h intervals over 2 d in cyclic light was probed with a riboprobe generated from the cloned differential display product T7 1. This probe identifies a pair of bands that is expressed during the early night, but not during the late night or day. The pair of bands (marked by the arrows) were named nocturnin. Reprobing of the blot with a chicken actin riboprobe verifies consistent loading. Each lane in this blot contains 1 pg total retinal RNA, isolated at the times indicated from eyecups cultured in cyclic light. Numbers to the right of the blot indicate the approximate size of the mRNAs in kilobases. Bars above the blot indicate the lighting conditions: black bars indicate night (dark) and white bars indicate day (light). Numbers above the lanes denote Zeitgeber Time (ZT) in hours as in Fig. 1. This figure is from ref. 8 with permission.

Localization of the nocturnin message by Northern blots and by ISH (7) revealed that this message was a photoreceptor-specific transcript. This is significant because these are the same cells that contain the circadian clock. Subsequent nuclear run on experiments verified that the changes in nocturnin mRNA levels are owing to changes at the level of transcription (8). Because the expression patterns and localization of the nocturnin message fulfilled our criteria for being an “interesting” clock controlled photoreceptor gene, we used the differential display product to screen a retinal cDNA library (8). Sequence analysis of the resulting cDNA clones revealed a 388 amino acid open reading frame that encoded a novel protein that contained a leucine zipper-like protein dimerization motif. This sequence also showed significant similarity to a yeast transcription factor called CCR4. Sequences similar to nocturnin have since been isolated from several other species, including chicken, bovine, mouse, and humans. Although we do not yet know the func-

228

Green and Besharse

tion of nocturmn, the sequence data suggests it may act as a photoreceptorspecific transcription factor that IS expressed only m early night. 4. Notes 1 5-hydroxytryptophan is added to mcrease the amount of melatonm made by the eyecups in culture (I I) and can be omitted if melatonm measurements are not being done. 2. The 1OX divalents will precipitate out of solution if they are added before the pH of the medium IS lowered by equlhbratron with CO2 3. Xenopus Zaevzs(5-6 5 cm) are purchased from Nasco (Fort Atkinson, WI) and maintamed m a strict 12-h light/l2-h dark light cycle Animals for use m these circadian experiments should be entrained m these lighting conditions at least 2 wk prior to use. 4. Washes in medium containing 10X antibiotics are critical for long-term culture of the eyecups. This treatment removes bacteria present on the skin of the frogs and without the washes, the eyecups will often become contaminated after 1-2 d in culture. We also routinely soak our dissecting tools m 70% ethanol at various times during the dissection to avoid carry-over of any contaminatron 5. We use black film boxes for the dark mcubations The boxes must not allow any light to reach the eyecups, but must allow gas exchange to occur The culture dishes are only removed from the black boxes m condttrons of complete darkness (with the exception of infrared light). 6. We use photographic safelights with infrared filters as the light source and wear helmets equipped with Infrared viewers (FJW Optical, Palatine, IL) Note that the medium IS replaced even on the dish being dissected. This 1s important, because dissectron releases large amounts of melatonm mto the medium, which would cause artificially high levels to be measured m the melatonm assay The medium m which the dissection is done is discarded We measure melatonm from eyecup cultures by radrotmmunoassay (RIA) of unextracted culture medium using the antiserum of Rollag and Nlswender (12). This assay was previously validated for measurement of melatonm m our culture medium (I I). Alternatively, kits are available commercially for measurement of melatonm (Elras USA, Osceola, WI). Generally, we use 1.O mL of the TRIZOL reagent for each group of 10 retinas and follow the manufacturer’s protocol We have successfully used many different RNA extraction protocols (see, for example, ref 13) based on the orrgmal Chomzynski and Sacchi protocol (14) Regardless of the RNA rsolation protocol used, all standard precautions for working with RNA must be followed, mcludmg wearing gloves and using RNAse-free solutions and glassware At this point it is advisable to check the integrity of the RNA This can be done by running an ahquot on a formaldehyde gel and staining tt or by domg a Northern blot We generally do Northern blots and use a probe for the rhythmic retinal gene product tryptophan hydroxylase (15) This not only confirms that the RNA is intact, but also verifies rhythmicity.

Clock- Regulated Genes

229

10. Although we generally used the GenHunter primers, we also synthesize some of our own upstream pnmers for specific controls. For example, we synthesized primers that should amplify the known retinal mRNA for tryptophan hydroxylase. This provided a posltlve control to show that the PCR reactions were working correctly. In addition, since we knaw that TPH mRNA 1srhythmic with a low amplitude, this also allowed us to verify that we could detect a two- to threefold change m abundance by this method 11. Random primed probes are prepared from the reamplified PCR products using the Random Primers DNA Labeling System from Gibco-BRL, except that 1 pL of the appropriate downstream anchored primer (10 pA4) is added to the reaction. Antrsense rlboprobes are synthesized using the RNA Transcription Kit from Stratagene. Both types of probes are purified with NucTrap gel filtration columns.

Acknowledgment We would like to thank Sandra Parsons for the production of Fig. 2.

References 1. Pittendrigh,

C. S. (198 1) Circadian systems: general perspective, in Handbook of vol 4 Blologlcal Rhythms, (Aschoff, J , ed. ), Plenum, New York, pp. 57-80. 2. Takahashi, J S. (1993) Circadian-clock regulation of gene expression. Curr Open Behavioral Neurobiology,

Genet. Dev 3,301-309.

3. Florez, J. C and Takahashl, J. S. (1995) The clrcadlan clock: from molecules to behaviour. Ann Med. 27,48 l-490. 4. Besharse, J. C and Iuvone, P. M. (1983) Circadian clock m Xenopus eye controlling retinal serotonin N-acetyltransferase. Nature 305, 133-135. 5. Cahill, G. M. and Besharse, J. C. (1993) Cn-cadlan clock functions localized in Xenopus retinal photoreceptors Neuron 10,573-577 6 Cahill, G M. and Besharse, J. C. (1995) Circadian rhythmicity m vertebrate retinas: regulation by a photoreceptor oscillator. Prog. Ret Eye Res 14,267-291. 7. Green, C. B. and Besharse, J. C. (1996) Use of a high stringency differential display screen for identification of retinal mRNAs that are regulated by a circadian clock. Mol. Brain Res 37, 157-165. 8. Green, C. B. and Besharse, J. C. (1996) Identification of a novel vertebrate circadian clock-regulated gene encoding the protein nocturnin. Proc Nat1 Acad. SCL USA 93, 14,884-14,888. 9. Liang, P., Averboukh, L., and Pardee, A. B. (1993) Distribution and cloning of eukaryotlc mRNAs by means of differential display. refinements and optimization. Nucleic Acids Res. 21, 3269-3275. 10. Llang, P., Zhu, W., Zhang, X., Guo, Z., O’Connell, R. P , Averboukh, L., Wang, F., and Pardee, A. B. (1994) Differential display using one-base anchored oligodT primers. Nuclezc Acids Res 22,5763-5764 11, Cahill, G. M. and Besharse, J. C. (1990) Circadian regulation of melatonin in the retina of Xenopus laevis: limitation by serotonin avallablhty. J. Neurochem. 54,7 16-7 19.

230

Green and Besharse

12 Rollag, M. D and Niswender, G. D. (1976) Radrotmmunoassay of serum concentrations of melatonm m sheep exposed to drfferent ltghtmg regrmens. Endocrznology 98,482-489. 13 Green, C B., Cahill, G M , and Besharse, J C (1995) Regulatton of tryptophan

hydroxylase expression by a retmal circadian osctllator zn vztro Brazn Res 677, 283-290.

14 Chomczynskr, P and Sacchr, N (1987) Single step method of RNA rsolatron by acid guamdmmm throcyanate-phenol-chloroform extractton Anal Blochem 162, 156-159 15. Green, C. B and Besharse, J C. (1994) Tryptophan hydroxylase expression IS regulated by a crrcadtan clock m Xenopus laews retina. J Neurochem 62, 2420-2428.

16. Cahill, G M and Besharse, J C. (1991) Resetting the crrcadran clock m cultured Xenopus eyecups regulation of retinal melatonm rhythms by hght and D2 dopamme receptors J Neuroscz 11,2959-297 1

19 Differential

mRNA Display

Adaption for In Vivo Studies of Diseased Tissues Mary E. Russell, Anne Rtiis&nen-Sokolowski,

and Ulrike Utans

Introduction This section focuses on the application of differential display in gaming instght into cDNAs upregulated in assoctation with a disease state by usmg direct m vivo samples. Although the majority of differential display studies have focused on in vitro models, a growing number of laboratories have used the technique on in vivo models of tissues with success (1-3). With in vitro systems concerns might be raised about potential artifacts related to models using tissue culture conditions, transformed cells, single-cell populations, and exogenous stimulants that may not mimic pathophysiologic condrtions. The advantage of using diseased ttssuesdirectly is that the pathophysiologic condttions including the multiple interacting cells are preserved dtrectly m the regulatory environment m which the disease arises. To best exploit the power of differential mRNA display, one must have an experimental system where there are substantial and measurable biologic dtfferences between cells, tissues or organisms of similar background. First, there should be evidence that some of these differences are mediated by alterations m gene regulation. Second, one should also select a system where there is a reproducible or renewable source of samples for the investtgations required to confirm the differential regulation pattern as well as the subsequent clomng and charactertzatton. For this PCR technique to be used effecttvely, the relative differences between the two samples should be both specific and significant. The samples being compared should be as closely matched as possible. Ideally, the only difference between the populations being studied should arise from the condt1.

From

Methods m Molecular Ecology, Vol 85 Dfferenbal Edltsd by P Llang and A B Pardee Humana

233

Dfsplay Methods and Protocols Press Inc , Totowa, NJ

234

hssell,

RBishen-Sokolowski,

and Utans

non or state under mvestigation. Hence, the species, genetic background, tissue or cell types should be the same and only the stimulation or condition should differ. If this is not the case, one will identify differences between the populations that are not relevant to the condition being studied and m a sense would be considered “true” false positives (i.e., real differences not relevant to the question under mvestigation). In viva tissues have higher variability simply owmg to differences between mdividual animals of the same strain The magnitude of the differences being studied should be large and reproducible. Of course, each mvestigator must define what qualifies as a significant difference m the context of their own expertmental system. Bear in mmd that subtle differences (Le., increases of lO-50%) on differential display gels or in followup evaluation of steady state transcript levels are much harder to study than “all or none” patterns. Dtfferenttal display was spectfically selected for our studies involving chronic rejection, which can only be produced m vwo and develops selectively m donor trssue m the rodent heart transplant models we have studied. Chronic rejection in these models is characterized by early and sustained mflammatory cell infiltration with development of vascular thickening only in the donor vessels (4), whereas vessels m recipient or in control transplants between identical strains are relatively unaffected. Previous studies from our laboratory and others provide evidence that the allomnnune response localized within the donor transplant heart induces the expression of a series of known T-cell and macrophage mediators including chemokmes, cytokmes, and growth factors (5). Macrophages are exqmsitely sensitive to microenvironment changes and can be activated simply by isolation, plating, or culturmg. One of our goals was to study macrophage activation in situ within the transplanted heart. Hence, we have compared a series of transplant allograft hearts undergoing chronic rejection (where the donor and recipient are mismatched) with control syngraft transplants (where the strains are matched). The overall scheme employed for these studies is outlined in the flowchart in Fig. 1. Using this approach, cDNAs were identified m chronically rejecting rat hearts (I), which were reproducibly up- or downregulated, full-length clones isolated, cells expressing the factors identified by in situ hybridization or immunohistochemistry, organ and cell specific mRNA and protem expression patterns established, and homology demonstrated m human transplants (67). 2. Materials 2.1. cDNA Synthesis 1 Total RNA. 2 Tissue culture grade H20. 3. 3’ Primer (H-T, ,M primer, where H = AAGCTT, and M = G, A, or C)

235

Differential m RNA Display

Rat Cardiac Transplantations

degenerate oligo dT primer)

Harvest PCR Band 1

Northern Analysis -Clone

PCR Fragment

1 Screen Cardiac Alloaraft Library

Northern Analysis

Recombinant Protein In Situ RT-PCR Hybridization

Fig. 1. Outline of differential display RT-PCR process.

Russell, R&&en-Sokolowsk,

236 4 5 6 7. 8 9 10 11 12

5’ Primer (arbitrary 10-13-mer ohgonucleotldes) 10X Buffer 200 n-J4 Tris-HCI, pH 8.4, 500 nut4 KC1 25 mA4 MgCl, 250 fl dNTP. M-MLV RT enzyme (SuperScrtptII-RTTM, Gtbco-BRL, RNAase (E. colz RNaseH, Gibco-BRL). Pipets (dedicated to RNA work). Plugged ptpet ups. 200~pL. Thin-walled tubes

2.2. Polymerase

and Utans

MD).

Chain Reaction

1, cDNA (synthesized with 3’ primer). 2, dNTP (10 mM, Perkin Elmer, CA) 3 10X PCR buffer (500 mA4KC1, 100 mMTris-HCl, 0 01% (w/v) gelatm; Perkin Elmer). 4 AmpliTuq DNA Polymerase (Perkm Elmer). 5. a-33P-dATP (2000 mCl/mol) (NEN-DuPont, PA) 6. PCR thermal cycler. 7. Prpets (dedtcated to PCR). 8. Plugged ptpet ttps 9. PCR tubes (200~$,-thin wall).

2.3. Polyacrylamide

pH 8.3, 15 mM M&l,,

and

Gels

1. 2. 3. 4 5.

Ammomum persulfate (Sigma, St. Louts, MO). TEMED (Sigma) 10X TBE buffer. Polysiloxane (SigmacoatTM, Sigma). Sequencing apparatus and power supply: glass plates (Gtbco-BRL), 4-mm spacer wtth sillcon foams (Gtbco-BRL), electrtcal tape (3M), square tooth comb, and bookbinder clamps 6. Whatman 3MM paper. 7. Radiosensitive mk. 8, Film, Kodak Btomax MR 35 x 43 cm (Kodak) Reagents for 4% polyacrylamide gel: (Commerical reagents: PAGE II System (Boehrtnger Mannhelm, IN) or Gel-Mix 4% (Gibco-BRL, MD) or SequaGel,

National Diagnostxs, GA). 2.4. Reconfirmation 1. 2. 3. 4. 5.

by Northern

Analysis

RNA blots (containing 20 &lane of total RNA) Reampified differential display fragment. 32P d-CTP (3000 Ci/mol) Random primmg ktt (Boehrmger Mannhelm). Heating block or bath.

Differential

mRNA

Display

237

6. 7. 8. 9. 10.

Quick spinTMG50 Sephadexcolumn (Boehrmger Mannhelm). 15-mL tube with cap. Saturatedbromophenol blue crystals(Sigma) Carrier DNA (Herrmg spermor calf thymus) (1 mg/mL). Hybridization solution (most standard protocols or commercially available reagentswork well). 11. Hybridization setup (water bath, oven, incubator). 12. 20X SSC:detail. 13. 20% SDS. 2.5. 1. 2. 3. 4.

Cjoning Reamplified differential display fragment. TA cloning kits (Invitrogen, CA) LB plateswith Ampicillm. X-gal (40 mg/mL).

3. Methods 3.1. Defining the Problem The most important theoretrcal consrderatron m the design of a drfferentral display study is the selection of the populations to be analyzed in the cDNA panels. The problem under investigation should be clearly defined and have reproducible and measurable characteristics (i.e., temporal onset, phenotype, locatron). The samples selected should have well-matched backgrounds such that the only difference is the stimuli or condition under investigation. 3.2. Quantitation of RNA Samples This is, by far, the most important practical parameter for successful differential display. Most routine methods can be used for RNA extraction. However, it is essential that selected RNA is of high quality and accurately quantitated by both spectrophotometry and gel electrophoresrs immediately prior to cDNA synthesis (see Note 1). 1. Measurethe RNA concentratronby spectrophotometryat 260 and 280 mn. 2. Evaluate all RNA samplesto beusedfor cDNA synthesis(2 &lane) on the same 5-mm denaturing agarosegel. 3. Confirm that all sampleshave similar intensitiesof 28s and 18s ribosomal RNAs and that the ratio is greaterthan 2: 1. 4. Dilute the RNA samplesto 0.4 pg/pL storing on ice. 5 Proceedimmediately to cDNA synthesis. 3.3, cDNA Synthesis Success in identification of regulated cDNAs from in vivo samples is improved if the analysis involves at least two separate samples of each condi-

Russell, Riis&en-Sokolowski,

238

and Utans

non. It 1simperative that the cDNAs for the analysts are prepared m the same experimental set to ensure that the condmons and reagents are uniform. We recommend larger scale protocols, which reduce prpetmg errors that arose at low concentration and sample volumes. Furthermore, master mixes provide a simpler format for multiple samples and ensure homogeneity m the results. The followmg protocol 1s for one cDNA reaction. However, mcreasmg the quantity of cDNA synthesized permits more extensive rephcate analysts with larger number of primers. 1 Setup PCRthermalcycler(seeNote 2). 65°C for 10s, 37°C for 60 min (Pauseafter 10 mm for later addition of reagents), 75°C for 2 mm, 37°C for 20 mm, and 4°C mdefimtely 2. Thaw reagents at 42°C except Superscript-RT, thaw RNA on ice 3 Set up RNA m PCR tubes RNA (diluted to 0 4 &pL) 1 0 & + dH20 10 4 & (place on we until ready to heat)

4 Make master mix (n. number of RNA samples, include negative control m whtch Superscript-RT has been omitted) 1OX PCR buffer 2Ox(n+l)@ MgCl 2.0 x (n + 1) pL dNTP (250 ~LV) 16x(n+l)uL H-T, ,M primer 2.0 x (n + 1) $ 5 Place the PCR tubes contammg the RNA m the machme and start. 6 Remove Superscript-RT from freezer, add to the master mix and vortex Superscript-RT 1.0 x (n + 1) p.L 7. After 10 mm at 37°C add 8.6 pL master mix, and continue incubation 8 After denaturing step (72°C 2 mm), add RNAseH 1 pL/reaction. Incubate for 20 mm at 37°C. 9 Subahquot mto 20-pL sets, wrap wtth parafilm, and use one set at a time to mmrmize freeze/thaws and danger of loss with spill. 10 Proceed to PCR step or store at -20°C

3.4. Differential Display PCR Given that we recommend the use of larger cDNA panels, it 1sadvisable that preliminary studies be performed using two cDNAs representing two experimental extremes (positive and negative). These studies should be used to confirm the conditions and reagents to be used m larger scale defimtive studies Parameters to be optimtzed include type and amount of isotope (see Fig. 2), annealing temperature, and number of cycles. Once these are establtshed, the mvestigator

can move ahead with confidence

to the larger scale study (see

Notes 3 and 4). 1 Thaw out the reagents and warm up at 42’C (except a[33P]dATP and AmpllTaq DNA polymerase). 2 Make master mix-1 (n, number of samples), extra for negative control

Differential

239

m RNA Display

B

A

Sample

36

32

Cycle number

A

B

AB

40

AB

32P

=v

Fig. 2. Optimization for differential display RT-PCR is an important step. (A) The effect of cycles number on differential expression of the bands. Even in 40 cycles, which is high in regular PCR, the differential expression is preserved. (B) Shows with sharper bands and lower background in 33P isotope. Originally, 35S was used in DD-RT-PCR (Z,ZO) but it was later shown to be volatile and thus hazardous (II).

3. 4.

5.

6.

1OX PCR buffer 2.0 x (n + 2) pL dNTP (25 @4) 1.6x(n+2)pL H-T, ,M primer 2.0 x (n + 2) pL H-AP primer 2.0 x (n x 2) pL Pipet 8.6 pL of master mix-l into PCR tubes, add 2.0 pL of cDNA straight into liquid: 2.0 pL cDNA Make master mix-2 (n, number of cDNA samples). One extra for negative control. dH,O 9.7 x (n + 2) pL u[~~P]~ATP 0.5 x (n + 2) pL AmpliTaq DNA polymerase 0.2 x (n + 2) pL Set up PCR machine as follows (a-e). a. 94’C for 5 s. b. Hot start: (See Note 5) SO’C; pause (make and add master mix-2). c. Low stringency cycle: 94°C for 15 s, 40-45’C for 60 s (1 cycle), 72°C for 20 s (see Note 6). d. High stringency cycle: 94’C for 15 s, 45-50°C for 60 s (39 cycles), 72’C for 20 s (autoextension: 2 s/cycle). e. Final extension: 72% for 5 min, and 4°C. Proceed to PAGE gel analysis or store -20°C for up to 5 d.

3.5. Differential Display Gels Critical decisions about the levels of amplified PCR products are made based on the relative intensity of these bands in radiograms from polyacrylamide gels used for size separation. Hence, gels should be of high quality to obtain maximum resolution and sample loading is consistent and uniform. Load samples within the same group consecutively to assist in the comparison of fainter banding patterns and later harvesting of desired bands.

240

Russell, FGistinen-Sokolowski,

and Utans

1 Prepare and pour a 4% polyacryamide gel mto preassembled plates (m which the short plate has been treated with polysiloxane (see Note 7). 2. Insert the square tooth comb to wlthm 2 mm from the edge of short glass plate into the gel solution. Clamp the sides of the gel plates usmg 2 or 3 bookbinder clamps per side corresponding to the posltlon of the clamps on the sequencing apparatus 3 Allow the gel to polymerize at least 2 h (see Note 8) 4 Remove the bookbinder clamps and cut the tapes at the bottom. 5. Remove the comb straight and gently from gel to preserve the integrity of the wells Rinse the top of the gel with 1X TBE buffer usmg a syringe with needle to remove urea 6 Add loading dye 10 & directly mto each PCR sample tubes 7. Preheat the gel, and denature PCR samples by heating to 100°C for 2 min and store on ice 8. Rmse wells at least 5 times/laneJust prior to loading to remove urea (see Note 9) 9 Load 4 pL (or maxlmum reproducible amount) of each sample per well Avoid bubbles. Mark the sample lanes outslde of the gel plates (see Note 10). 10. Run the gel until the second dye almost reaches the bottom. 11. Dismantle the plates, and lift the gel off usmg Whatman 3MM filter paper 12 Cover with plastic wrap and dry the gel by heating and vacuuming for 1 h at 8O”C, and then vacuum only for 30 mm 13 Mark the gel with radiosensitive ink in all four corners to allow later alignment of the gel and the film for band harvests. Develop the autoradlograph at room temperature for 24 h.

3.6. Selection

and Confirmation

of Candidate

cDNA Fragments

Band selection IS the next critical step. Differential display IS based on ldentifying those comigrating bands whose intensities are higher or lower in one study group compared with another study group. Examples of candidate bands

(Q, R, S) selected in our laboratory are shown m Fig. 3, whereas the Northern blot confirmmg these differences for Band R is shown m Fig. 4. Parameters consider in band selection include:

to

1. The Intensity of the amplified bands relative to other amplified bands comlgratmg the same distance should be dramatically different. A series of exposures may prove helpful in evaluatmg the relative differences in the intensity of the bands. 2. The differential regulation pattern should be highly reproducible such that a candidate band is present in all samples from one study subgroup compared with all of the samples m the reference subgroups (Fig. 3) even with repeat cDNAs sets, and PCR amplifications 3. Larger amplified products (top l/3 of polyacrylamlde gel) are ideal for further analysis. They have a higher likelihood of being reproducible and representing regulated cDNAs They are easier to amplify and clone After sequencing they are more mformatlve regarding identity, structure, and homology

241

Differential mRNA Display PCR cDNA

Run 1 set 1

Run 2 set 2

Band Q _ Band RI’ Band S Days

0 0714

0 0 714

075

014

Fig. 3. This shows the reproducibility of differential regulation pattern in three different sets of cDNA in two separate PCR analysis. It demonstrates downregulation of bands Q, R, and S in various degrees in transplanted rat hearts 7 or 14 d after transplantation. Reproducing the pattern is the most important step before proceeding to next step.

Band

QQRRSST

T

Markers

Fig. 4. Reamplification of harvested bands. Ethidium stained agarose gel shows PCR products amplified from duplicate harvests of bands (Q through T). 4. The background intensity in the lane can be used as a marker of even loading of the gel or within the PCR reaction. Overloading a lane can result in false positives.

3.7. Recovery of Selected Differential Display Fragments This is a fairly straight forward procedure. Harvesting a series of corn&rating bands at the same time increases the concentration of template DNA eluted and allows one to reamplify the candidate cDNAs with a smaller number of cycles (300 bp) Overnight polymerization may give more homogeneous results and reduces frownmg associated with uneven heat dissipation Rmsmg the wells extensively to remove urea Just prior to loading is critical for generating sharp bands m polyacrylamide gels Inclusion of a set of DNA markers or sequencmg reaction as a size indicator can be helpful Prior to labelmg by random priming one might elect to semtpurtfy agarose gel bands containmg the amplified PCR band from low-melt gels and recover the DNA by extraction, digestion with Gelase, or electroelution Confirmation of differential mRNA display patterns can also be completed by RT-PCR if the amplified fragment is >200 bp and clonmg is completed first to provide sequence mformation. We more commonly use RT-PCR to confirm a regulation pattern using primers designed from a full-length sequence that had homology to our differential display fragment (6,9), The major advantage IS that RT-PCR requires ten times less RNA than Northern blots and the cDNA generated can be used m replicate studies or to evaluate any number of factors from the same cDNA panel This requires design of specific primers internal (or nested within) the origmal set used for differential mRNA display reactions. Standard guidelmes for design of specific primers (size from 18 to 22-mer with GC content of 50%) should be adhered primer primer mteractions Most of the commercial primer design software programs are very helpful. After demonstration that a single PCR product can be amphfied from a positive sample, relative differences can be measured usmg any of number of semi-quantitative assays. To ensure that the amplified product accurately reflects the starting RNA template, cycle and RNA or cDNA dilutton studies should be completed to avoid saturating PCR conditions Dramatic differences that appear black and white on Northern blots are most desirable and easy to confirm, whereas more subtle differences mvolvmg reductions of ~50% are troublesome at best. Because human beings tend to over estimate differences, quantitation of the relative mtensities can be helpful. Select bands for further studies where the difference is at least threefold. Consider carefully the type of samples to be loaded on your blot and their order. Once they are on the blot you cannot change anythmg. Remember to include the positive and negative controls Ten or more blots are required for the initial studies attempting to confirm the dtfferential regulation pattern and even more than that to confirm the identify of partial and complete cDNA clones. Successwill be more rapid if the methodology for confirmation is estabhshed prior to the differential display screen In my laboratory, we estabhsh a bank of relevant Northern blots prtor to embarking on differential display RT-PCR.

Differential

mRNA

Display

247

To gam complete mformation on the structure and identity of a given factor, full-length cDNAs must be isolated Hence, sufficient RNA must be available to produce a cDNA library from the population under mvestigatlon, or an alternative library must be available. 16. If the Northern completed with the reamphfied PCR fragments identifies transcripts of more than one size. A number of clones may have to be screened to identify one that corresponds to each transcript (1). In order to isolate unique clones, one can perform Southern transfer of the gels containing the mmlprep digests of various clones or plaque lifts of the colonies after transformation One can then identify those clones that do not hybndlze with the clone known to be associated with a given transcript.

References 1. Utans, U , Llang, P., Wyner, L. R., Kamovsky, M. J , and Russell, M. E. (1994) Chronic cardiac reJection* ldentlficatlon of five upregulated genes m transplanted hearts by differential mRNA display Proc Natl. Acad Scz. USA 91(14), 6463-6467. 2. Dalal, S. S., Welsh, J , Tkachenko, A., Ralph, D , DlClcco-Bloom, E., Bordas, L , McClelland, M., and Chada K. (1994) Rapid isolation of tissue-specific and developmentally regulated brain cDNAs using RNA arbitrarily primed PCR (RAP-PCR). J Mol. Neuroscz 5(2), 93-104 3. Autlen, M. V., Feuerstem, G Z., Yue, T. L , Ohlstem, E H , and Douglas, S A (1995) Use of differential display to identify differentially expressed mRNAs induced by rat carotid artery balloon angloplasty. Lab Invest 72(6), 65-61, 4 Adams, D. H , Wyner, L. R , and Karnovsky, M J. (1993) Experimental graft arteriosclerosis. II Immunocytochemlcal analysis of lesion development Transplantation 56(4), 794-799. 5. Russell, M. E., Wallace, A F., Hancock, W. W., Sayegh, M H., Adams, D H., Slbmga, N E. S., Wyner, L R., and Kamovsky, M. J. (1995) Upregulatlon of cytokines associated with macrophage activation in the Lewis to F344 rat chronic cardiac reJection model Transplantation 59(4), 572-578 6. Russell, M. E., Utans, U., Wallace, A. F., Llang, P., Arcecl, R J., Kamovsky, M. J., Wyner, L. R., Yamashita, Y , and Tam, C (1994) Identification and upregulation of galactose/N-acetylgalactosamine macrophage lectm in rat cardiac allografts with arteriosclerosis. J Clm. Invest. 94(2), 722-730. 7 Utans, U., Quest, W., Wilson, J , McManus, B., and Russell, M E (1996) Allograft inflammatory factor-l. A cytokine-responsive macrophage molecule expressed m transplanted humant hearts. Transpluntatron 61, 1387-1392 8. Russell, M. E Differential mRNA display: power and pitfalls Am. J Clzn Nutr , in press. 9. Utans, U., Arceci, R. J., Yamashlta, Y., and Russell, M. E (1995) Cloning and charactenzation of allografi inflammatory factor-l. a novel macrophage factor identified in rat cardiac allografis with chronic rejection. J Clin Invest 95(6), 2954-2962 10. Liang, P. and Pardee, A. B. (1992) Differential display of eukaryotlc messenger RNA by means of the polymerase chain reaction Sczence 257(5072), 967-971 11 Llang, P. and Pardee, A B. (1995) Recent advances m differential display. Curr Opmzon Immunol 7(2), 274-280

Differential Expression of TIMP-3 During Neoplastic Progression in the Mouse JB6 Model System William D. Pennie, Yi Sun, Hyungtae

Kim, and Nancy H. Colburn

1. Introduction The differential display technique has proved to be of particular value m our laboratory in identifying genes whose transcriptlon is deregulated as part of the carcinogenesis

process. The power of this technique,

when applied to biologi-

cally relevant model systems for cancer development, will surely assist in the identification of the genetic events involved in cancer and potentially have a profound impact on diagnosis, prognostics, and treatment. I. I. Multistage Carcinogenesis Carcinogenesls is a multistage process. The hypothesis of multistage carcinogenesis has been verified m vivo, notably in colorectal carcinoma (I), and by many workers using in vitro models (2,3). By studying animal models such as mouse skin tumorigenesis, the carcinogenesis process has been subdivided into at least three sequential steps: initiation, promotion and progression (4). The carcinogenesls process has been studied in this systemby the use of chemical initiators and tumor promoters. The initiator is an agent, typically 7,12 dimethylbenz(a)anthracene (DMBA), which gives rise to genetic damage, whereas the promoter compound, typically 12Otetradecanoylphorbol13-acetate (TPA), effects are mitlally reversible suggesting eplgenetic events (5). Thus the initiation step gives rise to mutations, rearrangements and/or amplification of cellular DNA; then chronic exposure to tumor promoters gives rise to the progressive emergence of new phenotypes. Selection among these phenotypes yields benign lesions that, with further genetic or eplgenetic changes, may then progress to carcinoma. From

Methods ~1 Molecular Bology, Vol 85 D/ffefeentfa/ Ofsplay Methods and Protocols EdIted by P Llang and A 6 Pardee Humana Press Inc , Totowa, NJ

249

Pennie et al.

250 Table 1 TPA-Induced

Neoplastic

Not required

Possiblerequired Probably required

aSee refs

Transformation

in the JB6 Model=

Mltogenic stlmulatlon EGF receptor binding Stlmulatlon of hexosebinding Mutations in H-ras, c-jun, orp53 Actlvatlon of receptor kinesesPKC or EGFR Decreasedcollagen levels DecreasedTIS geneor&- 1 expression Decreasedganglloslde (trlslaloganglioslde) synthesis Elevation of superoxide AP-1 activation

7-14

1.2. JB6 Model System The ldentlficatlon of candidate oncogenes or tumor suppressor genes whose expression patterns are altered during neoplastlc progression 1sclearly fundamental to understanding the mechanisms of carcmogenesis. One valuable model system m the search for such candidate genes m our laboratory and others 1sthe murme JB6 cell culture model system (6). The JB6 cell lines were originally established from untreated BALB/C primary epidermal cell cultures and consist of three phenotypically drstmct cell variants. P-, which are resistant to tumor promoter-induced neoplastic transformation, P+, which are sensitive to such promoter action; and TX, which are neoplastic derivatives of P+. The phenotypes of these variant lines are stable through many passages allowing the analysis of both induction and maintenance of neoplastic transformation. This system recapitulates early to late stage tumor promotion and progression, and is therefore of great value m the study of molecular events along the pathway of multistage carcinogenesls. A number of factors have been considered as possibly being involved in the neoplastlc transformation (see Table 1). The mechanism underlying the neoplastic progression m the JB6 model 1sonly partially understood. Mutations m the H-ras, c-fun, or p53 genes do not seem to occur during progression (8,9). We have therefore used the differential display mapping technique (1.5) m an attempt to identify those genes responsible for tumor promotlon/progresslon in the JB6 model system m our laboratory. By comparison of the RNA from P-, P+, and TX cells, this sophisticated technique has allowed for the characterlzatlon of transcripts with expression patterns restricted to specific stages of neoplastic progression.

251

Neoplastlc Progress/on 2. Materials 2.1. Differenfial

Display

The differential display technique was performed using RNAmap Kits A and B (Gene-Hunter, Brooklme, MA) with slight modlficatlons to the recommended protocols. [35S]dATP 1sobtained from Amersham (Arlington Heights, IL). DNA 1s punfied from dried gels using the Qiaex kit (Qiagen, Chatworth, CA).

2.2. RNA lsolafion

and Northern

Blotting

RNA isolation utilizes RNazol solution, obtained from TEl-Test (Fnendswood, TX), and is performed as per manufacturer’s protocols. Northern blots are performed usmg Zetabind membrane (Cuno, Meriden, CT). [32P]dCTP radiolabeled probes are generated by “random priming” technique using the Redlprime kit (Amersham, Arlington Heights, IL).

2.3. Cell Lines JB6 cells were grown in Eagle’s minimal fetal calf serum (Gibco, Gaithersburg, MD).

3. Methods 3.1. Differential

essential medium

containing

5%

Display

1 200 ng of total RNA from four JB6 cell lines of preneoplastic and neoplasttc origins were reverse transcribed with T,,MG, T,,MA, T,,MT, or T12MC primers (where M = dG, dA, or dC) 2 Amplify the cDNA by PCR mcorporatmg [35S]dATP usmg the correspondmg downstream T&N primer and one of the arbitrary 10-mers (API - 10) from the manufacturer’s kit (one set of primers per reaction). 3. The amplified fragment sets from each original RNA sample are resolved next to each other on a 6% sequencing gel and the patterns obtamed from the different JB6 variants compared (see ref 16) 4. Following ldentlfication of differentially expressed products, the bands of mterest are excised from the dried gels, purified and reampllfied by PCR using the appropriate PCR primer pair 5. The amplified products are subcloned and used as probes to confirm the dlfferenteal display results by Northern analysis and as probes in subsequent cDNA library screening (for more details see ref. 16)

3.2. Northern

Analysis

1. 15-1.18of total RNA is denatured by heating to 65°C for 10 min, quenched on ice for 5 mm, and then size fractionated on a standard denaturing formaldehyde agarose (1.2%) gel. Be sure to include RNA size markers 2. Rinse with distilled water several times to remove formaldehyde then remove the water and rock the gel gently in 250 mL of 20X SSPE for 45 mm

Pennie et al.

252 P+

TIMP-3

2.6 -c 2.3 > .m nrr’, /” u

- &actin

Fig. 1. Northern blot analysis demonstrates mTIMP-3 is expressed in preneoplastic (P-, P+) but not in neoplastic (TX) JB6 cells. RNA was prepared from various independent clones of P-, P+, and TX JB6 cells and used for Northern blot analysis using standard protocols (see ref. 32). The blot was hybridized to a g-actin probe as an internal standard. All neoplastic JB6 lines analyzed had no detectable mTMP-3 expression suggesting that this gene is specifically downregulated during neoplastic progression. 3. Photograph the gel including a ruler for scale. Transfer the RNA to a nylon membrane overnight by capillary transfer using 10X SSPE as transfer buffer. 4. Air-dry in a vacuum oven at 60-80°C between two sheets of 3M paper then UV crosslink by exposing the membrane to a UV light box for l-l.5 min. Preprehybridize for 1 h at 65OC in 0. 1X SSPE/l% SDS then prehybridize in Church hybridization buffer (l%BSA, 7%SDS, 0.25MNaH2P04, pH 7.2, 1 mM EDTA, and 100 pg/mL. denatured salmon sperm DNA). 5. Add radiolabeled probe to the hybridization buffer (require at least 1 x lo6 counts/mitt/n& of hybridization buffer) and incubate overnight at 65°C. 6. Wash twice in wash buffer A (O.S%BSA, S%SDS, 20 mA4 NaHzP04, 1 mM EDTA) at 65°C then three times in wash buffer B (l%SDS, 20 mM NaH2P04, 1 mM EDTA). Blot-dry, wrap in plastic wrap, and expose to film overnight.

3.3. Example of Use Using this approach, we successfully isolated five differentially expressed clones. One differentially expressed transcript identified in this manner, specific for preneoplastic lineages, was used as a probe in cDNA library screening and in computer database searches. The longest open reading frame of the identified gene was found to have 96% amino acid identity with human tissue inhibitor of metalloproteinases type 3 (I 7). The expression of this gene appears to be completely lost in neoplastic JB6 cells as shown by the Northern analysis in Fig. 1.

Neoplastlc Progress/on

253

3.3.7. TIMP Gene Farm/y The TIMP genes are a family of related but distinct genes that are natural inhibitors of metalloproteinase activity (18). Although all TIMPs can inhibit MMP activity, individual TIMPs show some substrate specificity (19-21). All TIMP proteins have a shared characteristic loop structure maintained by 12 conserved cysteine residues forming six mtramolecular dlsulfide bonds (22). TIMP-I and TIMP-2 are secretory proteins, whereas TIMP-3 shows an extracellular matrix localization. Recently, a new TIMP family member has been described, TIMP-4; like types 1 and 2, it appears to be a secretory protein (Eric Shi, personal communication). The matrix metalloproteinase genes (MMPs) and their inhibitors have been implicated in a number of animal studies as being involved in tumor oncogemcity (231, tumor growth rates (24,25), and invasion/metastasis (25,26), whereas imbalance between the MMP and TIMP proteins has been suggested to contribute to arthritis and tumor metastasisin humans (27). Besides defined roles in the inhibition of MMP activity, other biological activities such as growth factor-like activity in promoting cell proliferation have been described for TIMP genes (28-30). In this regard, TIMP-3 has been shown to be Involved in cell cycle regulation, differentiation, and senescence(31) and was therefore an excellent candidate for having a role m the neoplastlc transformation process as modeled m the JB6 system. 3.32. Characterization of the Structure and Expression of the marine TIMP-3 Gene The murme TIMP-3 (mTIMP-3) gene was cloned and analyzed, and its expression examined both in vivo and in JB6 cell lineages (32). Transcripts of mTIMP-3 can be detected by Northern blot analysis in a number of murine tissues, especially kidney and lung (see Fig. 2). The till-length DNA sequence is 459 1 bp, consisting of a 316-bp 5’ untranslated region, a 636-bp open readmg frame, followed by a 3639-bp 3’ untranslated region. An analysis of the promoter region for transcription factor-binding sites revealed six potential binding sites for A&l, nine for PEA3, for c-fos-SRE, two NFkB sites, twop53 DNA binding motifs (imperfect), three Sp 1 sites, one CM~Csite, three TATA boxes (-10, -80, and -165), and a GC box (Fig. 3). The importance of these sites for the transcriptional regulation of mTZMP-3 is currently being mvestlgated. The putative AP-1 and NF-H3 sites appear to be functional in viva; a subset of the AP-1 sites contribute significantly to the transcriptional activity of the gene and bind AP-1 sites in in vitro gel retardation, DNA footprinting and transcriptional activation assays(H. K. et al., submitted) and indeed TIMP-3 expression 1sinducible m preneoplastic cells by AP- 1 or NF-kB activators (32). The human TIMP-3 gene is not yet sufficiently well characterized to compare

Pennie et al.

254

3

p-actin

Fig. 2. Northern blot analysis of mTIMP-3 expression in murine tissues. Following Northern blotting, RNA from various murine tissues was probed with a radiolabeled probe specific for TZMP-3 message. The three mTZMP-3 transcripts are indicated by arrows (4.6,2.8, and 2.3 kb). The blot was hybridized to a p-actin probe as an internal standard.

+58

I

PEA3

P53

c-myc

c-fos-SRE

AP-1 @

1

GC box

NF-KB

Fig. 3. Identified putative transcription factor binding sites upstream of the transcription initiation site in mTIA4P-3.

upstream regulatory sequences remote from the promoter with those of mTIMP-3, but potential binding sites for the transcription factor Sp 1, NFl, and C\EBP have been identified in the human gene (33). Despite the lack of a canonical TATA box or TATA like sequence, human TIMP-3 transcriptional initiation appears to occur at one major site. This differs from the pattern of the murine gene where three transcript sizes of mTIMP-3 message, (4.6,2.8, and 2.3 kb), are observed in JB6 cells. All three transcripts are polyadenylated and are detectable in P- and P+ cells, but not in neoplastic TX cells (Fig. 1). The specific lack of expression of mTIMP-3 is in contrast to the expression of the other

Neoplas

255

tic Progression

genes m marine cells; there is no evidence that mThfP- 1 or mTIMP-2 dtfferentrally regulated in different cell types of the JB6 model system mTIMP

1s

3.3.3. Mechanism for Differential Regulation of mTlMP -3 in Preneoplastic versus Neoplastic JB6 Cells The loss of mTIMP-3 expression in neoplasttc JB6 cells does not seem to take place by mutation of promoter sequences(32). Gross gene rearrangements, as analyzed by Southern blotting, do not appear to be responsible either, although higher order chromatm structure changes could be involved and are being investigated m our lab. All evidence to date suggeststhe deregulation takes place via subtle eptgenetic events. Whereas the endogenous TIMP-3 is inactive m neoplastrc cells, a transfected construct consisting of the TIMP-3 promoter drrving a lucrferase reporter gene shows both basal and mductble activity m cells where the endogenous gene IS silenced, suggesting that the neoplasttc cells have the full repertoire of transcription factors required for TIMP-3 expressron 3.3.4.

The Role of Methylation

m TI M P-3 Transcriptional

Regulation

One possible mechanism of repression of mTIMP-3 m neoplasttc cells 1s differential methylation. Abnormal methylation has been observed m cancer cells (e.g., during the progression of colon carcmoma [34,35]) and IS known to be an important regulatory mechanism for transcription of many genes The methylation state of the mTZMP-3 promoter has been studied by cleaving genomrc DNA from various JB6 lineages with a methylation sensitive restrrction enzyme, and subsequent analysis of the digestion products by Southern blotting. By this assaymTIMP-3 has the same methylation profile in P+ and Pcells, but of two TX lines analyzed, one was hyper and one hypomethylated with respect to the preneoplasttc lines (32) Treatment of these neoplasttc cells with a methylase inhibitor (5azacytrdine) stimulated expression in the hypermethylated (but not the hypomethylated) cell lure suggesting that methylation events may have played a critrcal role in repressmg expression of mTIMP-3 in this case (Fig. 4). The ability of specific methylation events to silence mTIMP-3 expression IS further demonstrated by the observation that m vitro methylation of mTZMP-3 reporter constructs at distinct sites causes dramatic reductions m the activity of the TIMP-3 promoter in JB6 cells (W. D. P., unpublished results). Mapping which specific CpG residues are methylated in the endogenous mTIMP-3 gene m preneoplastic and neoplastic JB6 cells may help determine the role of methylation m the silencing of this gene. 3.3.5. Overexpression

of TIMP-3 in Tumorigenic Cell Lines

To further examine the role of TIMP-3 m the JB6 neoplastlc progression model system, the effects of overexpresston of mTIMP-3 in neoplasttc lines was

256

Pennie et al.

4.6-

2.6 + 2.3-

Aza

-

+ +

+

+ + +

II Post

Fig. 4.

Aza

2h

26 h

-

+ II

+ + 2h

+ +

+

26 h

Methylase inhibition by S-azacytidinetreatment causesreexpression of

mTZMP3 in low passageRTlOl cells. The methylation stateof the mTZMP-3 gene is

hyperrnethylated in LRTlOl and hypomethylated in HRTlOl neoplastic JB6 cells. Cells were treatedwith Sazacytidinethen immediately or 24 h later treated with either TPA or TNF-a (both treatments are potential activators of TZMP-3 expression), as indicated. Total RNA was isolated and subjectedto Northern analysisas before using mTZMP-3 specific probe. The 5-azacytidinetreatmentrestoresexpressionof mTZMP3 only in the hypermethylatedLRTlOl cells. investigated (36). A mTIMP-3 cDNA construct under control of metallothionein promoter was introduced into both preneoplastic and neoplastic JB6 cells. Transfected cells were G418 selected, and colonies picked and checked for expression of mT’M.P-3 by Northern and Western analysis. Expressing clones were then analyzed for possible phenotypic changes using assays such as soft agar growth, nude mouse tumorigenicity and invasiveness (matigel assay). Expression of mTIMP-3 did not change any of these characteristics when mTIMP-3 expressing neoplastic cells were compared to the vector controls. These experiments indicated that TIMP-3 downregulation in TX lines is not the only event that causes the transformed phenotype, at least in this model. Reversion may require combinatorial changes in the expression of a number of differentially expressed genes, with mTIA4P-3 expression alone being insufficient. In this respect, the differential display protocol may help us identify tt other deregulation events required to operate in concert with loss of mTMP-3 expression to allow neoplastic transformation. Although overexpression of mTIMP-3 in neoplastic murine cells is insufficient to cause a reversion of phenotype, profound changes in tumorigenicity, growth rate, and serum responsiveness are observed when hTIMP-3 is overexpressed in DLD- 1 human colon cancer cells, that normally have no endogenous hTIA4P-3 expression (36). The

Neoplastic Progression

257

extent of these phenotypic changes correlatmg directly with amount of expresston of the stably transfected hTIMP-3 construct. Additionally, flow cytometry analysis suggests that the expression of TIMP-3 in these cells has a stgmficant impact on cell cycle perhaps by arresting cells m Gl or blocking entry to S-phase. Interestingly, it has recently been demonstrated that hTIMP-3 expression levels are regulated throughout the cell cycle, with an expression peak m the middle of Gl and a dramatic decreasewhen the cells enter S phase (31). The extracellular-matrix localization of TIMP-3 suggests that it acts by regulating the expression of effector genes possibly through some membrane associated signal transductton proteins. The identification of TZMP-3 associated proteins in the ECM or cell membrane may deny or confirm this hypothesis It is also conceivable that this postulated posttranslational effector pathway may have been deregulated in the neoplastic mouse JB6 cells explaining why overexpression of mTlMP-3 in this case has no phenotypic effect. The ongoing application of the differential display technique to the JB6 model system m our laboratory should identtfy other genes with possible oncogenie or tumor-suppressing effects in this system. The observation that restoration of TIMP-3 expression to neoplastic JB6 cells does not affect phenotype sigmficantly is a clear mdication that other genes are deregulated m these cells, and differential display may help us m finding these unknown genes. The identification of TIMP-3 as a possible player m the carcinogenic process has posed interesting questions as to the mechanism of the transcriptional stlencmg of this particular gene m neoplastic cells. We are currently attempting to characterize key transcription factors interacting with the TIMP-3 promoter to give clues as to possible targets for the deregulatory mechanisms. As prevtously discussed, the deregulation of TIMP-3 expression in neoplasttc lines is not due to differences in DNA sequence, gross chromosomal changes or transcription factor profiles between preneoplastic and neoplastic JB6 cells, suggesting that the mechamsm of deregulation is epigenetic m nature. Indeed, the methylation state of TIMP-3 is different m neoplastic lines from that in preneoplastic lines. In the two cell lines in which this was observed, the change in methylatton pattern was not consistent, being either hypomethylated or hypermethylated with respect to the preneoplastic state. It is clear that we need a more precise understanding of the methylation state of the TIMP-3 gene in all JB6 cell lmeages. Indeed, by mapping the precise nucleottde residues that are differentially methylated in preneoplastic and neoplastic JB6 cells, we hope to tdenttfy critical methylation events and demonstrate their consequenceson transcription. Whereas methylation clearly can play a role in the loss of mTIMP-3 expression in some, if not all, neoplastic JB6 lines, there may be an impact of other epigenetic mechanisms of transcriptional regulation. Nucleoprotem structure has been shown to have a profound effect on the transcriptional regulation of a

258

Pennie et al.

large (and growing) number of genes (37-39). One way in whtch chromatm structure can impose an additional level of regulation IS by hmttmg the access of transcription factors to their sites on DNA. For example, mouse mammary tumor virus (MMTV) adopts a precise chromatm structure in vivo, which limits the access of several transcription factors until the chromatm has undergone a structural transition effected by activated glucocorticoid receptor (40). It is interesting to consider the possibility that TIMP-3 is in a different chromatm architecture in preneoplasttc and neoplastlc JB6 cells, possibly owing to particular methylation differences of the DNA, which contributes to the differential activity of the gene. We are currently investigating this as another possible regulatory mechanism.

4. Notes 1 Northern blots can be stripped for reprobing by boiling the membrane in distilled water (RNA side face down) Monitor for probe removal by using a Gerger counter and boil again if required Expose to film overnight to check that the blot is clean 2 Do not allow blots that have been hybridized to become completely dry Store damp, wrapped in plastic wrap m an empty film cassette.

References 1. Fearon, E. R. and Vogelstein, B. (1990) A genetic model for colorectal tumorigenesis. Cell 61, 759-767 2. Land, H., Parada, L F , and Wemberg, R A (1983) Tumongemc conversion of primary embryo fibroblasts requires at least two cooperating genes. Nature 304,596-602. 3 Newbold, R F. and Overall, R W (1983) Fibroblast nmnortality is a prerequisite for transformation by EJ c-Ha-ras oncogene. Nature 304,648-65 1 4. Hecker, E , Fusemg, N E , Kunz, W., Marks F , and Thielmann, H. W. (eds ) (1982) Carclnogenesw-A Comprehenswe Survey, vol 7, Raven, New York 5. Yuspa, S H. and Pomer, M C. (1988) Chemical carcmogenesis from animal models to molecular models m one decade Adv Cancer Res 50,25-70 6. Colburn, N H., Kohler, B. A , and Nelson, K J. (1980) A cell culture assay for tumor promoter dependent progression toward neoplastic phenotype* Detection of tumor promoters and promotion mhibitors. Teratogeneszs Carcmog Mutagen 1,87-96 7. Colburn, N. H., Lerman, M I., Hegamyer, G A , Wendel, E., and Gmdhart, T D (1984) Genetic determinants of tumor promotion. Studies with promoter resistant variants of JB6 cells, in Genes and Cancer, Liss, New York, pp 137-155. 8. Sun, Y., Nakamura, K., Hegamyer, G , Dong, Z , and Colburn, N. H. (1993) No point mutation of Ha-ras or ~53 genes expressed m preneoplastic to neoplastic progression as modeled m mouse JB6 cell variants. Mol. Carcinog 8,49-57 9. Sun, Y., Hegamyer, G. A , and Colburn, N H. (1993) PCR-direct sequencing of a GC-rich region by mclusion of 10% DMSO application to mouse c-dun, Blotechnlques 15, 372-374.

Neoplas

tic Progression

259

10. Bernstein, L. R., and Colburn, N. H. (1989) AP-l/Jun is differentially induced m promotion-sensitive and resistant JB6 cells Science 244,566-569 11 Ben-Ari, E., Bernstem, L. R., and Colburn, N. H. (1992) Differential c-Jun expression m response to tumor-promoters in JB6 cells sensitive or resistant to neoplastic transformation Mol Carcznog ,5,62-74. 12. Bernstein, L R , Bravo, R , and Colburn, N. H. (1992) 12-O-tetradecanoylphorbol- 13-acetate induced levels of AP- 1 protems: a 42-kDa protein immunoprecrpitated by anti-fra-1 and induced m promotion-resistant but not promotion-sensitive 1B6 cells Mol Carcrnog 6, 22 l-229 13 Cmarik, J L , Hers&man, H , and Colburn, N H. (1994) Preferential primary response gene expression m promotion-resistant versus promotion-sensitive JB6 Cells Mol Carcznog 11, 115-124 Carcznogenszs 6,221-229 14. Dong, Z., Bnrer, M J., Watts, R G., Matrisian, L. M , and Colburn, N H. (1994) Blocking of tumor promoter-induced AP- 1 activity Inhibits induced transformation in JB6 mouse epidermal cells. Proc. Natl. Acad Scz USA 91,609-613 15 Llang, P. and Pardee, A. B (1992) Differential display of eukaryotlc messenger RNA by means of the polymerase chain reaction. Sczence 257, 967-97 1 16. Sun, Y., Hegamyer, G , and Colburn, N H. (1994) Molecular clonmg of five mRNAs differentrally expressed m preneoplastic or neoplastlc mouse JB6 eplderma1 cells* one is homologous to human mhlbitor of metalloproteinase-3. Cancer Res 54,1139-l I44 17 Apte, S. S , Mattei, M G., and Olsen, B. (1994) Clonmg ofthe cDNA encodmg human tissue mhibitor of metalloproteinases-3 (TIMP-3) and mappmg of the TIMP-3 gene to chromosome 22 Genomxx 19,86-90 18 Khokha, R. and Denhardt, D. T. (1989) Matrix Metalloproteinases and tissue inhibitor of metalloproteinases a review of their role in tumorigenesis and tissue invasion. Invasion Metastasis 9,391-405. 19 Pavloff, N , Staskus, P. W , Kishnam, N. S , and Hawkes, S P. (1992) A new inhibitor of metalloprotemases from chicken: ChTIMP-3 a thud member of the TIMP family. J Bzol Chem 267, 17,321-17,326 20. Leco, K J., Khokha R., Pavloff, N., Hawkes, S. P , and Edwards, D. R (1994) Tissue inhibitor of metalloprotemases-3 (TIMP-3) is an extracellular matrixassociated protein with a distmctive pattern of expression m mouse cells and ttssues. J Blol. Chem 269,9352-9360. 21. Denhardt, D. T , Feng, B., Edwards, D R., Cocuzzi, E. T , and Malyankar, U M (1993) Tissue inhibitor of metalloprotemases (TIMP, aka EPA). structure, control of expression and biological functions Pharmac Ther. 59, 329-341. 22. Williamson, R A., Marston, F A. O., Angal, S., Koklitis, P., Pamco, M., Morris, H. R., Came, A F., Smith, B. J , Harris T. J R., and Freedman, R B. (1990) Disulphide bond assignment in human tissue inhibitor of metalloproteinases (TIMP) Biochem J. 268,267-274 23. Khokha, R., Waterhouse, P., Yagel, S , Lala, P., Overall, C., Norton, G., and Denhardt, D. T. (1989) Antisense RNA-induced reduction in murine TIMP levels confers oncogemcity on Swiss 3T3 cells Sctence 244, 947-950

Pennie et al.

260

24. Koop, S , Khokha, R., Schmidt, B. B , MacDonald, I C , Moms, V. L , Chambers, A P , and Groom, A C (1994) Overexpresslon of metalloprotemase mhlbltor m B 16F10 cells does not affect extravasatlon but reduces tumor growth Cancer Res 54,479 14797 25. Montgomery, A. M P., Mueller, B. M., Relsfeld, R. A, Taylor, S M , and DeClerck, Y A (1994) Effect of tissue inhibitor of matrix metalloprotemases-2 expression on the growth and spontaneous metastasis of a human melanoma cell lme Cancer Res 54,5467-5473 26 Witty, J P , MacDonnell, S , Newell, K., Cannon, P , Navre, M , Tressler, R , and Matrislan, L M. (1994) Modulation of matrllysm levels m colon carcinoma cell lines affects tumorlgemclty m VIVO. Cancer Res 54,4805-48 12 27 Blrkedal-Hansen, H., Moore, W G. I., Bodden, M K , Windsor, L J , BlrkedalHansen, B., DeCarlo, A , and Engler, J A. (1992) Matrix metalloprotemases A review. Crzt Rew Oral Bzol Med 4, 191-250 28. Murphy, A. N., Unsworth, E. J., and Stetler-Stevenson, W. G (1993) Tissue inhibitor of metalloproteinases-2 inhibits bFGF-induced human mlcrovascular endothehal cell proliferation J Cell Physzol 157, 351-358. 29. Corcoran, M. L. and Stetler-Stevenson, W. G (1995) Tissue mhlbltor of metalloproteinases-2 stimulates fibroblast prollferatlon via a CAMP-dependent mechamsm. J. Bzol Chem 270, 13,453-13,459 30. Hyakawa, T , Yamashita, K., Tanzawa, K , UchiJima, E., and Iwata, K. (1992) Growth promoting activity oftlssue mhlbltor of metalloprotemases- 1 (TIMP- 1) for a wide range of cells. a possible new growth factor m serum. FEBS Lett 298,29-32

31. Wick, M., Burger, C., Brusselbach, S , Luclbello, F , and Muller, R. (1994) A novel member of the tissue inhibitor of metalloprotemases (TIMP) gene family 1s regulated during Gl progression, mltogemc stlmultlon, differentiation and senescence J Bzol Chem 269, 18,953-18,960. 32. Sun, Y., Hegamyer, G , Kim, H., Slthanandam, K., Li, H , Watts, R., and Colburn, N H (1995) Molecular cloning of mouse tissue inhibitor of metalloprotemases-3 (mTIMP-3) and its promoter specific lack of expression m neoplastlc JB6 cells may reflect altered gene methylation J Bzol Chem 270, 19,3 12-19,3 19 33. Wick, M , Haronen, R , Mumberg, D , Burger, C., Olsen, B R., Budarf, M L , Apte, S. S., and Muller, R. (1995) Structure of the human TIMP-3 gene and its cell cycle-regulated promoter Bzochem J 311, 54%554 34. Femberg, A. P., Gehrke, C. W , Kuo, K. C., and Ehrhch, M (1988) Reduced genomlc Smethylcytosme content m human colomc neoplasla. Cancer Res 48, 1159-1161 35. Goeltz, S. E , Vogelstem, B., Hamilton, S R., et al (1985) Hypomethylatlon of DNA from benign and malignant human colon neoplasmas. Science 228,187-l 90. 36. Sun, Y., Kim, H., Parker, M., Stetler Stevenson, W. G., and Colburn, N. H. (1996) Lack of suppression of Tumor Cell Phenotype by Overexpresslon of TIMP-3 m Mouse JB6 Tumor Cells: Identification of a Transfectant with Increased Tumorigemcity and Invasiveness. Antzcancer Res 16, 1-8

Neoplastic Progression

261

37. Felsenfeld, G (1992) Chromatin an essential part of the transcrrption apparatus. Nature 355,2 19-224 38. Grunstein, M., Durrin, L. K, Mann, R. K., Fisher-Adams, G , and Johnson, L M. (1992) Histones regulators of transcription m yeast, in Transcriptmnal Regulatzon (McKnight, S and Yamamoto, K., eds), Cold Spring Harbor Laboratory, Cold Sprmg Harbor, NY, pp, 1295-1315 39. Tsukiyama, T., Becker, P. B , and Wu, C (1994) ATP-dependent nucleosome disruption at a heat-shock promoter mediated by binding of GAGA transcription factor. Nature 367, 525-532. 40. Archer, T. K , Lefebvre, P , Wolford, R. G., and Hager, G. L. (1992) Transcription factor loading on the MMTV promoter: a bimodal mechanism for promoter activation. Science 255, 1573-l 576

21 Identification of Genes Associated to Photodynamic Therapy-Mediated

with Resistance Oxidative Stress

Charles J. Gomer, Marian C. Luna, and Angela Ferrario 1. Introduction Photodynamic therapy (PDT) is an experimental treatment for malignant diseases (1,2). This procedure involves the systemic administration of a tumorlocahzmg photosensitizer and subsequent exposure of the malignant lesion to tissue penetrating red light (3). Properties of photosensitizer localization m tumor tissue and photochemtcal generation of reactive oxygen species arc combined with precise delivery of laser generated light to produce a procedure offermg effective local tumoricidal activity (3). Tumors of the bronchus, bladder, esophagus, head and neck, brain and skm are being treated with PDT m clinical trials (4). The clmlcal results are encouraging and PDT recently received FDA approval for the treatment of advanced esophageal carcinoma m the United States.Additional regulatory approval of PDT has been obtained m Canada, the Netherlands, and Japan (5). Photochemical generation of reactive oxygen species during PDT can produce damage to lipids, proteins, and nucleic acids (6). Injury to cellular membranes, organelles, enzymes, and DNA has been documented m numerous studies. In addition, in VIVOstudies indicate that direct tumor cell and host cell cytotoxicity as well as damage to the exposed mtcrovasculature all can play a role in the rapid tumoricidal response observed followmg PDT (1,2,6). However, informatton 1slimited regarding actual mechanisms or targets involved m PDT induced toxicity at the cellular or tissue level. Murine radiation induced fibrosarcoma cells (RIF- 1) exhibiting a stable PDT resistant phenotype (Fig. 1) have been generated m an attempt to gain further insights regarding PDT mechanisms of action (7). Using the technique of mRNA differential display, we have detected a transcript that 1sconsistently present in parental cells but From

Mefhods in Molecular Bology, Vol 85 D~fferenbal D/splay Methods and Protocols E&ted by P Liang and A B Pardee Humana Press Inc , Totowa, NJ

263

264

Gomer, Luna, and Ferrario

0.ooo1~ 0

250

Light

I 500

I ‘Is0

1 1000

Dose (Jou1edsq.m.)

Fig. 1. Survival curves for RIF-1 parental cells (Cl) and PDT resistant cells CL-S (0) and CL- l(0). Cells were incubated with Phototin at 25 pg/mL for 16 h prior to light treatment Points, average of three experiments; bars, SE.

absentin our resistantcells (Fig. 2) (8). Repeateduse of mRNA differential display confirmed the altered transcription expression. The transcript was then isolated, subcloned, sequenced and identified as ct-2 macroglobulin receptor/low density lipoprotein receptor related protein (a-2 MR/LRP). Northern and Western immunoblot analysis confirmed that a-2 MR/LRP was present in the parental cell line but not in the PDT resistant clones. This receptor may play a role in PDT sensitivity by modulating photosensitlzer uptake and/or subcellular locallzatlon. 2. Materials 2.1. Cell Culture and Photosensitization

Reactions

1. Parental mouse radiation induced fibrosarcoma cells (RIF) and PDT resistant

RIF clones(7,s). 2. RPMI-1640 culture media supplemented with 15% fetal calf serum, 100 U/mL penicillin, and 100 pg/mL streptomycm. 3 Phosphate buffered saline (PBS): O.l37MNaC1,2 7 mMKCl,4 3 mMNa2HP04, 1.47 mM KH2P04, pH 7 2. 4. Trypsin solution: 0.05% (w/v) trypsin, 0.53 mA4EDTA, dissolved in PBS, pH 7.2. 5 Photofiin: Quadra Loglcs Technology, Vancouver, Bntish Columbq Canada. The

photosensitizer is dissolvedin 5%dextrosein waterat2.5 mg/ti andstoredat-20°C. 6. Light source for cell treatments. visible red light (570-650 nm, 0.35 mW/cm2) gener-

atedby a parallel seriesof 30-W fluorescentbulbs filtered with ared mylar film

identification

of Stress-Related

265

Genes

123 _*I W-id+!+ - .. -..

A

2

3

a-tubulln

1

2

3

C

kDa a,MR/LRP

(heavy

+

chain) -266

-116 -

97

Fig. 2. (A) Expression of PCR fragments following differential mRNA display. Total RNA was isolated from RIF- 1 cells (I), CL-1 cells (2) or CL-8 cells (3), reverse transcribed and PCR amplified in the presence of [3sS]dATP. The fragments were displayed on a 6% acrylamide sequencing gel. Arrow shows fragment present only in parental RIF-1 cells. (B) Northern analysis of total RNA isolated from RIF- 1 cells (I), CL- 1 cells (2) or CL-8 cells (3). RNA was hybridized with the radiolabeled cDNA fragment identified in 2 (A). Blots were rehybridized with a-tubulin to analyze RNA loading. (C) Western immunoblot of a-2 MRI LRP in RIF-1 cells (I), CL- 1 cells (2) or CL-8 cells (reproduced with permission from ref. 8).

2.2. RNA Isolation 1. 2M Sodium acetate. 2. Water saturated phenol. 3. Denaturing solution: 4M guanidinium thiocyanate, 25 rnM sodium citrate, pH 7.0, 0.5% sarcosyl, O.lM2-mercaptoethanol (see Note 1).

266 4. 5. 6. 7. 8 9.

Gomer, Luna, and Ferrario Chlorofonntsoamyl alcohol mtxture (49.1) 15mL Polypropylene tubes (Falcon cat no 2059). Isopropanol 75% Ethanol 1OX DNase buffer: 200 mA4 Tris-HCl, pH 8 4,20 mM MgCl,, 500 mM KC1 0 1% Dtethyl pyrocarbonate (DEPC)-treated water

2.3. Reverse

Transcription

PCR and mRNA Differential

Display

1, Degenerate 3’ anchoring primers, 25 w T&IN (specifically Tt2 MA for detectmg a-2 MR/LRP) (synthesized m Chtldrens Hospttal Btotech Laboratory). 2 [a-35S]dATP. spectfic activity 1200 Cilmmol (DuPont-NEN, Wilmmgton, DE) 3. pBluescrtpt subclonmg TA vector, (Stratagene, La Jolla, CA) 4. Mouse Maloney Leukemia Virus (MMLV) reverse transcriptase 50 U/& (stored at -2O’C) 5 Thermocycler Perkm Elmer Model 9600 6. 25 mA4MgC1,. 7 10X PCR Buffer II: 100 mMTri.s, pH 8.3, 500 mA4KCl 8 200 w Nucleottde trtphosphate (NTP) stock (made from mdtvidual 10-w stocks supplied m Gene Amp RNA PCR kit). 9 RNase Inhtbttor: 20 U/pL (Boehrmger Mannheim, Indtanapohs, IN). 10 AmpliTaq DNA Polymerase. 5 U/pL (Perkm Elmer, Norwalk, CT) 11. 5’ Random primers, lo-mers, 5.0 pA4(specifically ACCGGCTCGC for detecting a-2 MRLRP) 12 1OX TBE buffer 1.6M Tris base, 1M Boric acid, 40 mA4 Na,EDTA, pH 7 2 13. 6% Acrylamide, 8MUrea solution: 47 mL H20, 48 g Urea, 5 7 g acrylamide, 0.3 g btsacrylamtde, stir until dtssolved and then add 10 mL 10X TBE buffer Filter by vacuum through a 0 22-pm filter to remove particles. Place solution m a Erlenmeyer vacuum flask and remove gas bubbles under vacuum. 14 Stgmacote (Sigma, St Louis, MO) 15 Tetramethylethylenedlamlne (TEMED). 16. 10% Ammonmm persulfate (store for 2-3 d m dark at 4’C). 17 Loading solution. 95% formamtde, 20 mM EDTA, 0 05% bromophenol blue, 0.05% xylene cyan01 FF

2.4. cDNA Recovery 1. 2. 3 4. 5 6.

and Reamplification

Glycogen, 10 mg/mL. Ethanol (lOO%, 85%). 3M Sodmm acetate. 25 mMMgC12. 10X PCR Buffer II: 100 mMTrts, pH 8.3, 500 mMKC1 200 l.&f NTP stock (made from mdividual 10-d stocks supplied m Gene Amp RNA PCR kit) 7. 3’ Anchoring primers, 25 flT&N (specifically T12MA for detecting a-2 MR&RP)

Identification of Stress-Related

Genes

8. AmpliTaq DNA Polymerase. 5 U/l& (Perkm Elmer). 9 5’ Random primers, lo-mers, 5 0 @(specifically ACCGGCTCGC a-2 MR/LRP)

267 for detectmg

2.5. Subcloning Isolated cDNAs 1, T-4 DNA hgase. 200 U/$ (Gibco-BRL, Grand Island, NY) 2 10X Ligase buffer. 300 mA4 Tris, pH 7 4, 100 mA4 dithiothreitol (DTT), 100 mM M&l, 10 miV ATP 3 Insert ready TA-vector: pBluescrtpt (Stratagene) (see Note 2) 4 Competent E colz for pBluescript* XL-l-Blue MRF’, Stratagene. 5 Colony lysis buffer 0 1% Tween m TE. 6 25 mlMMgC1, 7 1OX PCR Buffer II: 100 mM Tris pH 8 3,500 mM KC1 8 200 @t4NTP stock Made from individual 10 n-u!4 stocks supplied in Gene Amp RNA-PCR kit (2 pL each dATP, dCTP, dGTP, and dTTP m 92 pL water) 9 25 pA4 T,,MN (synthesized m Childrens Hospital Biotech Laboratory). 10 AmphTaq DNA Polymerase (Perkm Elmer). 11. 5 0 M Primers used in differential display (see Section 2 3.) 12 TE 10 mMTris-HCl, pH 8.0, 1 mMEDTA 13. Plasmid miniprep buffers. a. Buffer P 1. Resuspension buffer-50 mA4 Tris-HCl, pH 8 0, 10 mM EDTA; add 100 pg/mL RNase A before use. b Buffer P2: Lysis buffer-200 mA4NaOH, 1% SDS c. Buffer P3* Neutralization buffer-3 OMpotassmm acetate, pH 5 5 d Buffer QBT Equihbration buffer-750 mA4 CaCl, 50 mM MOPS, pH 7 0, 15% ethanol, 0 15 % Trnon X-100 e. Buffer QC Wash buffer-l.OM NaCl, 50 mM MOPS (3-[N-morpholino] propanesulfomc acid), pH 7 0, 15% ethanol f. Buffer QF: Elution buffer-l 25MNaCl,50 mMTris-HCl, pH 8 5,15% ethanol

2.6. Northern Analysis 1 Ohgo labeling buffer (OLB). made from solutions A.B:C m a ratio of 100.250*150 (store OLB at -8O’C). a Solution A 1 mL (1 25M Tns-HCl, 0 125M MgC&, pH 8.0) plus 8 pL 2-mercaptoethanol, 5 pL dATP, 5 pL dCTP, 5 pL dGTP, 5 pL dTTP (each triphosphate previously dissolved in TE, pH 7.0 at a concentration of 0 1M) b. Solution B. 2M HEPES, pH 6.6 c Solution C Hexadeoxynbonucleotides m TE at 90 U/mL 2. Bovine serum albumin (BSA): 10 mg/mL. 3. 32P-dCTP, specific activity 3000 Ci/mmol (DuPont-NEN, Wilmington, DE). 4 Klenow DNA polymerase I* 2 U/pL (Boehringer Mannhelm) 5. Stopping solutron: 20 mM NaCl, 20 mM Trrs-HCl, pH 7.5, 2 mM EDTA, 0.25% sodium dodecyl sulfate (SDS), 1 $4 dCTP

Gomer, Luna, and ferrario

268

6 10X MOPS (3-[N-morpholmo]propanesulfomc actd) Buffer: 0 2MMOPS, 0 05M sodium acetate, O.OlM EDTA. 7 37% Formaldehyde solutron. 8 Loading buffer 0 72 mL formamtde, 0 16 mL 10X MOPS buffer, 0.26 mL 37% formaldehyde, 0.18 mL H,O, 0 1 mL 80% glycerol, 0 08 mL saturated bromopheno1 blue solutton 9. 1% denaturing agarose gel: 1 g agarose, 10 mL 10X MOPS buffer, 85 mL H,O. (Heat in microwave to dissolve agarose, cool to 50°C, add 5 4 mL 37% formaldehyde, swirl to mix, positron comb and pour gel [pour gel m hood to mmtmtze exposure to formaldehyde fumes]) 10. Nylon filter; Nytran Plus, Schletcher and Schuell Keene, NH. 11. Hybridization buffer: lMNaC1, 50 nnI4 Tris, pH 7.5, 10% dextran sulfate, 1% SDS, 0.2 mg/mL salmon sperm DNA (boiled) and 50% formamtde 12. Wash solutions: IX SSC (150 mMNaC1, 15 n-&sodium citrate), 0 1% SDS

2.7. Sequencing

cDNA

1 1OX TBE buffer: 16M Tns base, 1M Boric acid, 40 mMNa,EDTA * 2H20, pH 7 2 2. 8% Acrylamide 8M Urea solution: 48 g Urea, 7 6 g acrylamtde, 0.4 g btsacrylamtde, 10 mL 10X TBE buffer, dissolve by stirring and then adJUst volume to 100 mL wtth H20 When ready to pour gel, add 1 mL of 10% ammonium persulfate and 25-& TEMED. Filter by vacuum through a 0.22~pm filter, place solutton in a Erlenmeyer vacuum flask and remove gas bubbles under vacuum. 3 Stgmacote. 4 Tetramethylethylenedtamme (TEMED) 5. 10% Ammomum persulfate (store for 2-3 d m dark at 4’C). 6. Loading solution* 95% Formamtde, 20 mA4 EDTA, 0.05% bromophenol blue, 0 05% xylene cyan01 FF. 7. Plasmrd DNA. 8 2MNaOH. 9. 20 mMEDTA. 10 3M Sodium acetate, pH 4 5. Il. Ethanol: 100 and 70%. 12. c+~~S dATP, specific activity 1000-1500 Ct/mmol (DuPont-NEN). 13 5X Sequenase buffer: 200 mMTns-HCl, pH 7.5,lOO mA4MgCl,, 250 mMNaC1 14 100 n&I Dithtothrertol. 15. T-7 Primer (-40), 5’-GTTTTCCCAGTCACGAC-3’, 0.5 pmol/& 16. 5 X Labeling mtx: 7.5 ).tA4dGTP, 7.5 WdCTP, 7.5 @4dTTP. 17. ddG Termination mix* 80 w dGTP, 80 @dATP, 80 @4 dCTP, 80 @4dTTP, 8 WddGTP, 50 mA4NaCl. 18 ddA Terminatton mtx* 80 fl dGTP, 80 @dATP, 80 @4 dCTP, 80 WdTTP, 8 @4ddATP, 50 mMNaC1. 19. ddT Termination mix* 80 NdGTP, 80 WdATP, 80 @4dCTP, 80 @4dTTP, 8 @I ddTTP, 50 rmI4 NaCl.


Genes

269

20 ddC Termination mix: 80 @!dGTP, 80 @dATP, 80 @4dCTP, 80 @4dTTP, 8 @4 ddCTP, 50 mMNaC1. 21. Enzyme dilution buffer 10 mM Trls-HCI, pH 7 5,5 mMDTT, 0.5 mg/mL BSA. 22 T7 DNA Polymerase (Sequenase Vernon 2.0): 13 U/pL in 20 mMKPO,, pH 7 4, 1 mM DTT, 0.1 mM EDTA, 50% glycerol.

2.8. Western Blot Analysis 1. 2. 3 4 5. 6 7 8. 9. 10. 11. 12. 13.

4-l 2% Precast polyacrylamide gradient gel (Jule, New Haven, CT) 10% Glycerol 10X Electrode running buffer 0 25 Tns, pH 8.6, 2Mglycine, 1% SDS. 2X Sample lysing buffer: 0.125M Tns, pH 6 8, 4% SDS, 10% glycerol, 0 02% bromophenol blue, 4% P-mercaptoethanol 50% Methanol. 10% Acetic acid. Transfer blotting buffer. 0.03MTris , pH 8.3, 150 Wglycine, 20% methanol. Transfer nitrocellulose paper, BA-S NC, Schleicher and Schuell. Protein transfer umt, Hoefer Scientific Instruments. Polyclonal antibody, R-777 against a-2 macroglobulm/low-density lipoprotem receptor-related protein (a-2 MIULRP) Alkaline phosphatase H-linked avidin-biotin detection system (Vectastain, Vector Laboratories, Burlingame, CA). TBS (Tris buffered saline) buffer; 100 mA4Tns, 0 9% NaCl, pH 7.5. TTBS (Tween Tris buffered saline); 0 1% Tween, TBS

3. Methods 3.7. Cell Culture Procedures

and Cell Treatment

Protocols

1. RIF cells and PDT resistant RIF clones are maintained as a monolayer culture in RPMI-1640 media containmg 15% FCS and antibiotics (complete growth media). 2. Cytotoxicty following PDT is analyzed by measuring clonogenic survival. 3. Exponentially growing cells are removed from T-75 plastrc tissue culture flasks by trypsinization (remove culture media, rinse once with PBS, add 2.0 mL of trypsin solution, incubate at 37°C for 5 min, inactivate trypsin by adding 10 mL of complete growth media). 4. Pipet cell mixture vigorously to obtain a single cell suspension and count cells with a Coulter counter. 5. Cells (10~50,000) are replated in 60-mm plastic Petri dishes, incubated for 16 h m media containing Photofrm (25 pg/mL) and 5% FCS, rinsed with PBS, and then incubated for 30 mm m fresh growth media supplemented with 10% FCS. 6. Dishes are then exposed to graded doses of vlslble red light. 7. Cell survival following PDT is determined by incubating treated dishes at 37°C for 7 d in complete growth media and countmg the number of colonies that arise from surviving single cells. 8. Cytotoxicity following a 24-h PEA incubation at 37’C in complete growth media was also determined by clonogenic assay.


270

3.2. Isolation of Total RNA from Mouse Cells (Adapted from the Procedure by Chomeznski and Sacchi [9]) 1 2 3 4 5 6 7. 8. 9 10

Culture parental and PDT resistant RIF cells in T-75 plastic culture flasks Rinse cells (3-5 x 107) with PBS and add 2 mL of denaturmg solution Swtrl until the cells are completely dissolved Scrape flask and place solution of denatured cells in 15 mL polypropylene tubes. Add sequentially; 0.2 mL 2A4 sodium acetate pH 4 0,2 mL phenol, 0.4 mL chloroform-isoamyl alcohol (mtx after each addmon) Shake mixture vtgorously for 10 s and cool on ice for 15 mm. Centrifuge samples at 10,OOOgfor 20 min at 4’C Transfer the aqueous phase (upper layer) to a new tube and add 2 mL rsopropanol and mix. Precipttate RNA by incubating at -20°C for 1 h Centrifuge

sample at 10,OOOg for 20 mm at 4T

to obtain RNA pellet

11 Resuspend the RNA pellet in 0.3 mL denaturing solution and transfer to a 1.5-r& Eppendorf tube 12 Precipitate RNA by adding 1 vol of tsopropanol and incubating at -20°C for 1 h 13 Spm down RNA m a microfuge for 10 min at 4°C. 14 Rinse RNA pellet with 75% ethanol and au-dry 1.5. Resuspend RNA m DEPC treated water for Northern analysts 16 Before using RNA for differential display, remove DNA contamination by mcubating 50 pg RNA in a 100~pL solution containing 10 $ 1OX DNase buffer, 10 U DNase I and 10 U rtbonuclease inhibitor for 30 mm at 37°C 17 Extract the RNA with phenol/chloroform (3:l) 18. Transfer aqueous phase (top layer) to new microtube and add l/10 volume of 3M sodium acetate and then 2.5 vol of ethanol. 19. Precrpttate RNA at -20°C for 1 h. 20 Centrifuge at 4OC for 15 mm 2 1, Resuspend RNA pellet in DEPC-treated water.

3.3. Reverse Transcription-PCR and mRNA Differential (adapted from the procedure by Liang and Pardee [lo])

Display

1. Initiate the reverse transcription reaction by combinmg 4 pL 25 mA4 MgC&, 2 pL 10X PCR buffer II, 2 pL 200 WNTP stock, 2 pL T&N anchoring pnmer, 0.2 pg RNA and bring final volume up to 20 @., with DEPC-treated water 2. Heat mixture to 65°C for 5 mm and then cool to 35°C for 10 mm. 3. Add 1 ,uL RNase inhibitor and 1 ,tiL MMLV reverse transcrlptase 4. Mix sample by ptpetmg and then incubate at 35°C for 50 mm. 5. Heat mactrvate the sample at 95°C for 5 mm. 6. Centrifuge tube briefly to collect condensatton and place sample on ice for PCR. 7 To initiate PCR amphficatton, combme 2 l.tL from reverse transcrtptase reaction with 18 pL PCR mix (1.8 pL, 10X PCR buffer II, 1.8 pL 20 @4dNTP, 0.9 pL 25 mA4 MgC12, 1 0 $ Tr,MN anchoring primer, 1.O pL 5’ random 10-mer primer, 1.O p.L [cG~~S] dATP, 10.3 pL water and 0.2 @-.Ampli’Taq DNA Polymerase)


Genes

271

8 Mix and add a top layer of mmeral oil (50 &) to reaction tube to avoid evaporation during cycling. 9. Denature the reaction mixture at 94°C for 2 mm before amplifying. 10 PCR parameters include denaturation at 94°C for 30 s, annealing at 40°C for 2 mm, and elongation at 72OC for 30 s for 40 cycles, followed by elongation for 5 mm at 72°C. 11. Combme 4 pL of resulting PCR sample with 2 clr, of sequencing loading solution for separation on a 6% sequencing gel. 12 Prepare glass plates for sequencing gel (clean, dry, use slgmacote to silamze the top plate, and clamp plates with spacers). 13. Add 20 uL TEMED and 1 mL of 10% ammomum persulfate to 6% acrylamlde solution, swirl and pour gel Place desired comb in gel and allow gel to polymerize for l-2 h. 14. Load samples (6 a) and run gel using at 60 W constant power (voltage not to exceed 1700) for 3-4 h (until xylene cyan01 is at the bottom of the gel) using 1X TBE buffer. 15. Blot the differential display gel onto Whatman 3MM paper, cover with plastrc wrap and dry on a gel dryer under vacuum at 80°C for 1 h without methanolacidic fixing. 16 Remove plastic wrap from the dried gel and expose to X-ray film (either Kodak XAR-5 or BMR-2) for l-3 d Mark gel with either radioactive ink, tape, or needle holes for orientation after developing 17 Confirm that differences in mRNA display are reproducible by repeatmg the reaction.

3.4. cDNA Recovery

and Reamplification

1. Line up differential display autoradiography film with gel. 2. Mark bands of interest on dried gel using a needle or pencil and then cut out the marked areas on the gel with a scalpel 3 Incubate the gel slice, along with the 3MM paper, m 100 l.iL HZ0 for 10 mm. 4. Diffuse the cDNA out of the gel by boiling the gel slice for 15 min m a capped and paratilm wrapped test tube. 5. Transfer the cDNA solution to a new tube. Add l/l0 volume 3MNaOAC, 5 pL glycogen as a carrier and 450 pL ethanol, place at -80°C for 30 mm and recover cDNA by centrifuging at 16,OOOgfor 15 mm at 4°C. Rinse pellet with 100 pL ice cold 85% ethanol, centrifuge briefly and remove residual ethanol 6. Dissolve recovered cDNA in 10 pL Hz0 and use 4 uL to reamplify the cDNA fragment. 7. Reamphfy cDNA in a 40 pL reaction using the primers that produced the initial differential display, 20 pA4 NTPs, no radioisotope, and the same PCR cyclmg conditions used to amplify the cDNA for the initial differential display (4 pL recovered cDNA, 2 @ 25 mM MgCl,. 4 pL 1OX PCR buffer II, 4 )iI, 200 @4 NTP stock, 2 clr, Ti,MN primer, 2 pL random lo-mer, 21 6 pL HZ0 and 0 4 pL AmpliTaq DNA polymerase)

272


8 Run 30 uL of the PCR sample on a 1.5% agarose gel stained with ethidium bromtde. Check the size of your fragment to see if it agrees with the size of the fragment on the differential display gel Store the remaining 10 pL for subcloning. 9 Isolate reamplified cDNA for use as a probe m Northern analysis by cuttmg bands of amplified cDNA fragments out of agarose gel. 10 With low-meltmg agarose and TAE buffer, weigh gel slice, dilute 1:3 with water, boll for 7 mm and label dnectly by random prime labelmg (see Section 3.5.). One can also extract the reamphfied cDNA from an agarose gel using a QIAEX kit from Qiagen (Santa Clartta, CA).

3.5. Subcloning

isolated cDNA

1 Ligate PCR cDNA product into insert ready vector by combmmg 10 uL H20, 2 pL vector, 2 pL 10X hgase buffer, 5 uL PCR product from reamplification step 3.4 and 1 pL T-4 DNA ligase 2 MIX well by ptpetmg up and down gently 3 Ligate overnight at 16°C 4. Transform competent cells. Place 10 pL ligation mixture in 15mL round bottom tube. Thaw competent E colz on ice and munedrately add 100 p.L of E colz to ligation mixture and place on ice for 10 min Heat shock E coli for 2 min at 42°C add 1 mL LB medium and incubate for 1 h at 37°C in an environmental shaker or roller drum at 250 rpm. Plate ahquots of transformation culture on LB plates contammg ampictllm or the appropriate antrbiotrc for selectton. 5. Select positive colomes (number the colomes of interest on the back of each plate) 6 Ahquot 50 pL colony lysts buffer into labeled tubes 7 Pick each colony with a clean pipet tip and transfer to the corresponding tube. (Store plates at 4°C for future use.) 8 Incubate tubes at 95’C for 10 mm. Spin the tube at 16,000g for 2 mm to pellet the cell debris Transfer the supernatant to a clean tube and either use it directly for PCR analysis or store at -20°C for future use 9. Check insert for correct size using PCR with original primers and amplification condittons. For each colony lysate mix 2 pL of the lysate wtth 1 pL 25 mMMgC&, 2 pL 10X PCR buffer II, 2 pL 200 @4NTP stock, 1 pL T&IN, 1 pL random lo-mer, 11.8 pL HZ0 and 0 2 pL AmphTaq. Perform PCR using previously described cycling parameters 10 Run the PCR products on a 1.5% agarose gel stained with ethtdium bromide. 11 Purify each band out of the agarose gel as before (see Section 3.4.). Recheck specificity by labeling the insert and using it to probe a Northern blot (see Section 3.6 ) 12. Select colonies with inserts whose size and hybridization pattern correspond to original differential display. 13. Restreak colonies on LB plus ampicillin (or approprtate antibiotic) plates. Incubate overnight at 37’C. Select a single colony and grow in 25 mL LB plus appropriate antibiotic. Use 23 mL for plasmid preparation (several commerctal plasmtd columns/ktts are available, we use Qiagen plasmid mtdi Kit with Qiagen-tip 100). Save the remammg 2 mL of the LB culture as a 50% glycerol stock at -8O’C.

Identification

of Stress-Related

Genes

14 Harvest the bacterial cells by centnmgation at 6000g for 15 min at 4°C. Remove all traces of supernatant by mvertmg the open centrifuge tube until the entire medmm has been drained. Proceed directly with purification protocol or freeze at -20°C for storage 15. Before usmg the Kit for the first time, add lyophilized RNase A provided to buffer Pl This solution is stable for 6 m at 4°C Check buffer P2 for SDS preciprtation due to low storage temperatures. If necessary dtssolve the SDS by warming. Prechill buffer P3 16. Resuspend the bacterial pellet m 4 mL of buffer P 1 17 Add 4 mL of buffer P2, mrx gently, and Incubate at room temperature for 5 mm. 18 Add 4 mL of chilled buffer P3, mix immediately but gently, and incubate on ice for 15 min 19. Centrifuge at >2O,OOOg for 15 min at 4°C Remove supernatant promptly and place m fresh tube 20 Recentrifuge at >2O,OOOg for 15 min at 4°C. Remove supernatant and combine with supernatant from step 19. 2 1 Equilibrate an Qiagen-tip 100 by applymg 4 mL buffer QBT, and allow the column to empty by gravity flow. 22. Apply the supernatant from step 20 to the Qiagen-tip and allow the flmd to enter the resin by gravity flow 23. Wash the Qiagen-tip with 2 x 10 mL buffer QC 24. Elute DNA with 5 mL buffer QF. 25 Prectpitate DNA with 0.7 volumes of room-temperature tsopropanol. Centrifuge immediately at 215,OOOg for 30 min at 4°C and carefully remove the supernatant. 26. Wash DNA with 2 mL of 70% ethanol, remove alcohol carefully and completely, am-dry for 5 mm and dissolve m a 50 pL of TE, pH 8 0 27. To determine yield, measure DNA concentration in a UV spectrophotometer at 260 nm. The mmiprep plasmid is now ready to sequence

3.6. Northern

Analysis

1. Denature RNA by heatmg at 60°C for 5 min in RNA loading buffer 2 Load 6 pg RNA (2-4 pL RNA plus 14 pL RNA loading buffer) on a denaturmg 1% agarose gel contammg 0.66M formaldehyde, 1X MOPS buffer 3. Separate RNA using 100 V until bromophenol blue dye migrates three-fourths of the way down the gel 4 Transfer separated RNA to a nylon filter. 5. Crosslmk RNA to nylon filter by exposure to UV light or by baking in a vacuum oven at 80°C for 2 h. 6 Set up radioisotope labeling reaction for cDNA probe with 10 p.L OLB, 2 pL BSA, cDNA probe (up to 3 1.5 pL and between 20 and 200 ng), 5 pL [32 P]dCTP, 1 pL of appropriate T,&4N primer, water to bring the final reaction volume to 50 pL and 1 pL Klenow. 7. Run labeling reaction at room temperature overnight. 8. Add 200 pL stop solution when reaction is complete. The probe is ready to use on Northern blot analysis


274

9 Boil probe for 10 min to denature 32P labeled cDNA prior to hybrldlzatlon 10 Hybridize labeled probe with filter in 5 mL hybridization buffer at 42’C overnight 11 Wash filter at room temperature m 1X SSC, 0 1% SDS for 15 min, followed by 2 42°C washes m 0 2X SSC, 0 1% SDS for 15-30 mm An additional wash at 55°C m 0.2X SSC, 0.1% SDS for 15 min may be necessary if radloactlvlty of filter IS high 12 Expose the filter, using an intensifying screen, to Kodak XAR-5 film at -80°C (overnight to 1 wk). 13 cDNA fragments with Northern hybridization patterns matching differential display patterns are subcloned and sequenced

3.7. Sequencing

cDNA

1 Denature, precipitate and resuspend 3-5 K of the plasmld to be sequenced (see Section 3.5.). Denature plasmld DNA by adding 0 1 vol of 2M NaOH, 2 mA4 EDTA and incubating for 30 mm at 37°C. Neutralize the mixture by adding 0 1 vol of 3M sodium acetate and precipitate the DNA with 3 vol ethanol Place at -80°C for 30 min and centrifuge at 16,000g for 15 mm at 4°C in an Eppendorf centnfuge Wash the pellet in 70% ethanol and dissolve m 7 pL distilled water. 2 Add 2 pL Sequenase reactlon buffer and 1 pL primer Anneal by heating 2 mm at 65”C, then cool slowly to 35°C for 20 niin. Centrifuge at 16,000g briefly and chill on ice for use m labeling reaction 3 While annealing mixture IS coolmg, label, fill Eppendorf tubes with 2.5 pL of termination mixtures (G, A, T, and C) 4 Dilute labeling mix fivefold to obtain a working concentration 5 Prewarm 4 termination tubes (G, A, T, and C) m 37°C water bath. 6. Labeling reaction: To ice-cold annealed DNA mixture (10 pL) add 1 pL DTT, 2 $ diluted labeling mix, 0.5 & [35S] dATP, and 2 pL, diluted Sequenase polymerase MIX and incubate at room temperature 2-5 mm 7. Termmatlon reactions Transfer 3 5 pL of labeling reaction to each termmatlon tube (G, A, T, and C), mix and continue incubation of the termmatlon reaction at 37°C for 5 min. 8 Stop the reactions by adding 4 pL of Stop solution 9 Heat samples to 75°C for 2 min, nnmediately before loading onto sequencmg gel. Prepare gel (see Section 3.3.) Load 2-3 pL m each lane 10. Run sequencing gel m 1X TBE buffer at 60 W 11. Followmg electrophoresls, soak gel for 15 mm m 5% acetic acid and 15% methanol to remove the urea. 12. Dry gel at 80°C for 1 h and expose dried gel directly to X-ray film (see Note 3).

3.8. Western lmmunoblot

Analysis

1. Attach 4-12% gradient gel to electrophoresls unit and add 500 mL of 1X electrode buffer to top and bottom chambers 2. Dissolve cells (1 x 10’) m 4% SDS sample lysing buffer and heat at 95°C for 3 min. 3 Load 100 ~18of each protein solution into sample wells of the 4-12% precast denaturmg gradient polyacrylamide gel

Identification

of Stress-Related

Genes

275

4 Run gel at 150-200 V for 4 h at 4°C. 5 Remove the gel from the plates and transfer proteins from gel to mtrocellulose. 6. Fill transfer unit with Blotting Buffer and keep unit at 4°C Soak transfer cassette, sponge, and blotter paper with blotting buffer. Place mtrocellulose paper on top of blotter paper and the polyacrylamtde gel on top of the rntrocellulose paper. Place another piece of blotter paper on top of the gel and place the cassettem the transfer unit. 7 Set the unit to run at 20 V to transfer the protein to the nitrocellulose paper. Run the sample overnight at 4°C (the mtrocellulose paper must be located between the gel and the anode). 8. Wash the nitrocellulose paper three times with TBS and then incubate the mtrocellulose m 10% nonfat dry milk m TBS for 2 h at 37°C 9 Wash the nitrocellulose three times with TTBS and then incubate the nitrocellulose paper with the primary LRP antibody (R777) for 2 h at room temperature using a 1: 1000 ratio of antibody m PBS 10. Wash the mtrocellulose paper three times with TTBS. 11 Incubate the mtrocellulose paper with btotinylated secondary antnabbit IgG antrbody (Vectastain ABC-AP kit) diluted m TTBS for 30 mm at room temperature Wash three ttmes with TTBS and then incubate with avrdin DH-btotmylated alkalme phosphatase diluted m TTBS for 30 mm. 12. Wash three times with TTBS and then incubate the mtrocellulose paper with alkaline phosphatase substrate (2 mg/mL p-nitrophenylphosphate in 100 mM sodium bicarbonate, pH 9.5, 10 mMMgC1) until staining is complete (normally 1 mm) Stop the reaction by washing the mtrocellulose paper water.

4. Notes 1. Storage time 1sapprox 3 mo at room temperature without 2-mercaptoethanol or 1 mo at room temperature with 2-mercaptoethanol. 2 TA cloning kits are commercially available through Invttrogen (San Diego, CA), addrttonally GenHunter (Brookline, MA) recommends their PCR-TRAPTM cloning system). 3. The isolated 174-bp fragment corresponded to nucleotldes 6810-6983 of the 14849 bp cDNA for LRP. The isolated clone had flanking sequences of only the random primer instead of to both the poly (A) tail and the random primer. This type of result has been reported previously and indicates that the random primer served as both the 5’ and 3’ primer m the amplificatron reaction followmg reverse transcription-PCR.

References 1. Gomer, C. J., Rucker, N., Ferrarro, A , and Wong, S. (1989) Prmctpals and appltcations of photodynamic therapy. Radiation Res. 120, l-l 8. 2. Henderson, B. W. (1989) Photodynamtc therapy-commg of age. Photodermatology 6,200-2 11. 3. Pass, H. I (1993) Photodynamtc therapy m oncology, Mechanisms and clmrcal use. J Natl. CancerInst 85, 145-157.

276


4 Fisher, A M R , Murphree, A L , and Gomer, C. J (1995) Cluucal and prechmcal photodynamic therapy. Lasers Surg Med 17,2-3 1 5 Levy, J C (1995) Photodynamic therapy Trends Biotechnol 13, 14-17. 6. Gomer, C J (199 1) Preclmtcal exammatton of first and second generation photosensitizers used m photodynamic therapy. Photochem. Photoblol 54, 1093-L 107. 7 Luna, M C and Gomer, C J (199 1) Isolation and initial characterization of mouse tumor cells resistant to porphyrm-mediated photodynamic therapy. Cancer Res 51,4243-4249 8 Luna, M. C., Ferrario, A., Rucker, N., and Gomer, C J. (1995) Decreased expression and function of a-2 macroglobulm receptor/low density lipoprotein receptorrelated protein in photodynamtc therapy-resistant mouse tumor cells Cancer Res 55, 1820-l 823 9 Chomezynskt, P. and Saccht, N (1987) Single step method of RNA tsolatton by acid guanidmmm thiocyanate phenol-chloroform extraction Anal Blochem 162, 156 10. Ltang, P and Pardee, A B (1992) Dtfferenttal dtsplay of eukaryotic messenger RNA by means of the polymerase chain reaction. Science 257,967-970

22 RAP-PCR Using RNA from Tissue Microdissection Roy A. Jensen, David L. Page, and Jeffrey T. Holt 1. Introduction Cell culture provides powerful models for studying cell biology, but many pathologic states cannot be appropriately studied without tissue-based approaches, We have developed a tissue-based microdtssection method for molecular analysis and have chosen premalignant breast disease for study. This approach requires methods for selective isolation and enrichment of parttcular cell types to prevent trivtal results gained by comparison of tissue samples that differ in percent epithelial cells, presence of inflammatory cells, and so on. The development of breast cancer is presumed to involve a series of genetic alterations that confer increasing growth independence and metastatic capability on somatic cells. Understanding the molecular events involved in this process will ultimately require methods to study gene expression within whole human tissues since it is extremely unlikely that key processes can be modeled using cell culture. Careful epidemiological studies have established the relative risk for breast cancer development for specific histologic lesions (1,2). Invasive breast cancer develops near the previous biopsy site m 2530% of patients following diagnosis of noncomedo ductal carcinoma zn situ (DCIS) (3,4), providing strong evidence that this premalignant lesion is a determinant event in breast cancer progression (5). In order to study molecular events in human premalignancy we developed a method for isolating epithelial cell RNA from histologically identified lesions in human breast biopsies, and used this method to clone and study genes that are differentially expressed in DCIS (6-8). This chapter will describe the methods involved in tissue-based gene comparison including microdissection, RNA purification, cDNA library construction, and RAP-PCR. From* Methods U-I Molecular &logy, Vol 85 Dlfferenhal Edlted by P Llang and A B Pardee Humana

277


Jensen, Page,andHo/t

278 2. Materials 1 5 6A4 Guamdmmm

2. 3 5 6.

lsothtocyanate solutton, make fresh as required Phenol, twice dtsttlled and equilibrated m TE (Trts-EDTA), pH 7.4. Chloroform, reagent grade 4 3 5M Sodmm acetate, pH 5.0. Absolute ethanol DEPC-treated water add 1 mL of dtethylpyrocarbonate to 1 L of double drstllled water Leave overnight and then autoclave

3. Method Freshly obtained breast biopsy, mastectomy, or reduction mammoplasty specimens were evaluated for inclusion m the study. The specimens were serially sectioned and areas of breast parenchyma, exclusive of adipose tissue were submitted for frozen section analysis to identify normal breast epithelial elements, noncomedo ductal carcinoma in sztu, and invasive breast carcinoma. With identification of these entities, a portion of the sample was then submitted for routine paraffin embedding and sectioning, and areas not required for confirmation of the diagnosis were maintained frozen m O.C.T. (Miles Scientific [Elkhart, IN]) at -70°C. 3.1. RNA Preparation 1 Frozen tissue samples were remounted m the cryostat and thinly cut to obtain a scout section to be used to rdenttfy and map eptthehal elements for directed harvesting (see Note 1) 2. A 2-mm punch IS used to obtain a core from the frozen block that IS nnmedtately minced m 800 pL of 5 6M guamdmmm isothtocyanate and 40% phenol Epltheha1 tissues are readily disrupted by this method, whereas stroma and dense connective tissue are msoluble and removed by a simple 5-s centrrfugatton m an Eppendorf tube to remove particulate matter (see Notes 2-4 and Fig 1) 3. The supematant from the rapid spin IS then aspirated through an 1s-gage needle twice and then through a 22-gage needle three times to disrupt DNA and reduce vtscostty 4 200 pL of chloroform 1sthen added and the sample 1splaced on ice and intermittently vortexed for 5 mm. 5. The upper aqueous layer (approx 500 pL) is then saved 50 pL of 3 5M sodium acetate 1s added and 1 mL of ethanol 1s used for ethanol prectpitatton at -20°C overnight 6. RNA IS then washed m 70% ethanol and resuspended m water that 1spretreated with DEPC RNA can then be employed for differential display (9-11) or RAPPCR (12-14)

3.2. cDNA Library

Construction

1. PolyA selectton using oligo-dT should be performed tf sufficient RNA 1sobtained followmg mlcrodtssectton

279

RA P-PCR Con

NLl

NL2

NL3

NL4

112

(18

#4

#8

HO

tlOC

Fig. 1. Molecular characterization of RNA samples. Expression of collagen III mRNA in tissue mRNA samples was analyzed by RNase protection assay by methods we have reported previously (8,). One pg of mRNA was hybridized with 32P labeled, T7 polymerase-generated RNA probes directed against GADP and collagen III. The following RNA samples were probed: NLl-cultured human breast epithelial cells, NLZ-normal breast tissue, NL3-fibrous stromal fraction of breast tissue, NL4-a second sample of normal breast tissue. In addition, samples from DCIS cases 12,6,4, 8, and 10 were probed. Sample 1OC is RNA obtained from a focus of invasive cancer present in a separate area of the block in case 10. Con is a control sample using tRNA. Note that the NL3, the presumed stromal component from normal tissue contains large amounts of collagen III mRNA consistent with its mesenchymal character. In contrast the remaining samples show low levels of expression consistent with their epithelial origin. 2. cDNA libraries are constructed in lambda phage (Lambda ZAP, Stratagene, La Jolla, CA) using polyA-selected mRNA from frozen tissue samples. Each tissue sample is used to produce a separate and unique cDNA library. 3. Analyze cDNA libraries for quality, analyzing percent inserts and the insert size (our unamplified libraries had greater than 50% inserts and contained between 2 x 106and 7 x 10’ phage recombinants with an average insert size varying between 500 and 1000 bp). 4. Rescue the phage cDNA library to obtain plasmid DNA representative of the entire library with suitable helper phage. 5. Amplify the rescued plasmid DNA library and then purify plasmid DNA by cesium chloride gradient centrifugation.

3.3. RA P-PCR 1. Transcribe 0.5 pg of plasmid DNA (from a cDNA library) with 300 U of MMLV reverse transcriptase for 60 min at 35°C along with 2.5 l.& random primer and 20 pi&t deoxynucleotide triphosphates.

280

Jensen, Page, and Ho/t

2. Following heat mactivatton at 95°C for 5 mm, two 25-base random ohgonucleottdes (see Note 5 and Figs. 2 and 3) are added to 5-pA4 concentration and PCR amplification is performed employmg Tug polymerase. 3 The following cycle conditions are used for PCR: 40 cycles performed with denaturation for 1 mm at 94°C annealing for 2 min at 42°C and extension for 1 mm at 72°C 4. Resolve samples on 8% nondenaturing polyacrylamide gels, dry, and autoradtograph. 5. Specific bands may be reamphfied with the same primers used for their generation, purified on 8% nondenaturing polyacrylamide gels, then the gel bands can be extracted, subcloned by standard methods and sequenced.

4. Notes 1. We selected lessons for microlocalization that were relattvely isolated and homogeneous. Thus we did not utthze cases m which mvasive, zn sztu, and normal epithelial components were admixed and could not be selectively harvested as a homogeneous sample by a 2-mm punch 2 To confirm the purely epithelial nature of the DCIS lesions and to Identify genes useful in molecular characterizatton of the samples, formalm-fixed paraffinembedded sections of the tissues were stained with Masson’s trichrome, and also tmmunohistochemically analyzed for the presence of cytokeratin (keratms 8 and 18) and vlmentin with antibodies obtained from Becton Dickinson (San Jose, CA) and Boehrmger Mannheim (Indianapolis, IN), respectively. Antibodies were used at recommended dilutions and immunohtstochemical staining was performed on an automated Ventana Model 320 immunostainer, which utilizes a modified avidm-biotm complex method with diammobenzidme visualtzation. Following immunostaining, all slides were lightly counterstained with hematoxylm (6). 3. The particulate material resistant to guamdinium-phenol extraction is white and fibrous and has been demonstrated to be breast stroma by unmunohistochemistry and molecular analysis (7). This particulate material was sparse m DCIS samples but abundant m samples obtained from reduction marmnoplasty. To obtain RNA samples from these stromal cells, this white particulate material was minced with a tissuemizer, washed with PBS, treated with collagenase at 37°C for 30 mm, somcated, extracted with phenol/chloroform, and ethanol precipitated. Fig. 1 shows a nuclease protection assay using a collagen III probe to demonstrate maxrma1 expression of this connective tissue marker in the NL3 sample from the stromal sample. 4. Multiple punches from individual lesions were needed to obtain sufficient RNA for polyA selection and library construction. Two hundred micrograms of total RNA was obtained by pooling 20 punches from each normal breast tissue sample (reduction mammoplasties) or 5-8 punches from each DCIS lesion, presumably reflecting the greater cellularity of the DCIS samples. Pooling of 2-mm punches was done only on individual lessons from single patients or on normal tissue from individual patients.

281

RAP-PCR a=random(25,0/0,1,4) y-round Ja,O) col(6O)=if(y=4,“T”,O) col(61)=1f(y=3,“G”,0) col(62)=if(y=2,“C”,O) col(63)=if(y=l,“A”,O) col(l)=col(6o)+col(61)+col(62)+col(63) col(76)=if(y=4,2,0) col(77)=lf(y=3,4,0) col(78)=if(y=2,4,0) col(79)=if(y=l,2.0) q=(col(76),col(77),col(78),co1(79)} ql=Total(q) put ql Into cell(l,27) al=random(25,010,1,4) yl=round (al ,O) co1(64)=if(yl=4,‘1” ,O) col(65)=if(y1=3,“G”,O) col(66)=~f(yl=2,“C”,O) col(67)=if(yl=i ,“A”,O) col(2)=cot(64)+col(65)+col[66)+col(67) col(3)=d (col(1 ,1,25)=col(2,1,25),1,“--“) g-count (col(3)) put g into cell (3,26) col(80)=lf(y1=4,2,0) col(61)=lf(yl=3,4,0) col(82)=if(yl=2,4,0) col(83)=lf(yl=l,2,0) q2=(col(8O),col(81),col(82),col(83)) q3=Total(q2) put q3 into cell(2,27) a3=random(25,0/0,1,4) y3=round (a3,O) col(68)=if(y3s4,“T”,O) col(69)=d(y3=3,%“,0) col(7o)=lf(y3=2,“C”,O) co1(71)=if(y3=1 ,“N,O) col(6)=col(68)+col(69)+col(7O)+col(71) col(84)=if(y3=4,2,0) ool(85)=if(y3=3,4,0) col(86)=if(y3=2,4,0) col(87)=if(y3=1,2,0) q9=(col(84),col(85),col(86),col(87)}

q4=Total(q9) put q4 into cel1(5,27) a4=random(25,0/0,1,4) yrl=round (a4,O) col(72)=lf(y4=4~~,0) col(73)=if(y4=3,“G”,O) col(74)=if(y4=2,laI,O) col(75)=rf(y4=1 ,“A”,O) col(6)=col(72)+col(73)+col(74)+col(75 col(7)=if (col(5,i ,25)=col(6,1,25),1 >--“I h-count (col(7)) put h rnto cell (7,26) col(88)=if(y4=4,2,0) co1(89)=if(y4=3,4,0) col(90)=if(y4=2,4,0) col(9l)=if(y4=1,2,0) q5={col(88),col(89),col(9O),col(91)} qbTotal(q5) put q6 into cell(6,27) I,

Fig. 2. Computer program from Sigmaplot, generating random 25mer oligonucleotldes The column numbers (e.g., co1[60]) are arbttrary and other columns could be used 5. We developed a computer program to generate random primers with minimal overlap. This computer program can be run m SrgmaPlot and IS presented in Fig. 2 A serves of 200 primers were generated and screened for homology to

Jensen, Page, and Ho/t

282 CGCCCCCCGCGCGCCGGCGGGCGCC l’s 2's CGCGACGGCCGCGCGTCTGCCAGGG CiXCCCTGCGTTACCCTCCCCGCCG GGATGGCGTCCTGTAACCCGACGCT ACTGGGCTGTCCTGCGGTGGCGGGG CTGAGAGGTAGCCGCGCGGAGGCTG ECCTGGCCGCGACACGGATTACCGC TTAGCGCATGGTGGACCTGGAGACG TGTGGTTACGTCAGCGAAGGTAATA AGTCGCACGCATGTCACGCTCCGCC TATCCAAGCGGCAGGCTACGAGGCC GGCGCGCCCGACGGTCTGGTATCTA GGCATATAGCCCGTTCAAGGCCATC CTCCCTCCCCGGACTCGGGGTTAGT ATGCGGGCGGCTCGGGCCTGGTCGC CGTGAAGCCTATGCCCTCCCTCAAC GTGCCGTCGTAGCCCTTCAGCGATC GCGACACTAGGCTCCCGGAGGAGGG TGCGCCAGGCCTCCGGGCCCGGTAT CCGGAACTGCGATAGCGTCCGTCCC AGCGGACACCTGTTTCCCGAGAGCC AACGGGTGGACATCCGCCTGCCGCC TCGACCACGATGTCAATCGTCCCGA TCATCCCCGCCGAAAGACGCTCGCC ATAGGCTGCGGCACGCGCTGGGACT GACCAGGTGCGCACGAGCATGTACA AGCGTAGTCATCGGCCTTCGCGCCC GGCCCCTAGCCCAGGGTGAAGCCCA CCCAGTGCTACGGGCCGCGCCAAGC CCTTCCTGGGTTACCTGCCCTCGGG ~~CCGACAGCAGCCACGCCAAGGGC ACGCGCTGGTCCAGCGAGGCCTGAT CGATGCAAGGCCAGCAGCACTCGAC CCCCCGGAGCGGACCACCGGACGTG AGCGGGGAGGGATCGGGGGCCAAGC GCCTGGTGTAGGCAGGCAGCTCTTA CCACCCCTGTAGTGCGGGCTGCGAG GGAACCCGACGCCCGTCCAGGGTTC TCGGGCAGCAAGGCCGGGACGCTCC GACGGGGGACGGGCTAGGTGGCTTA CTTGTTGCCGGCGGAGAGGGCTGCC

Fig. 3. List of 40 random ohgonucleottdes with minimal overlap generated by the computer program shown in Fig. 2 DNA sequences shown are listed 5’ to 3’ generate the 40 prtmers that showed the least overlap. A list of these 40 prtmers is presented m Fig 3.

Acknowledgments We thank Cheryl Robinson-Bemon for preparatron of cDNA libraries, Patrice Oberrniller for differential display studies, Sharifah Moore for DNA sequencing. The work was supported by the A. B. Hancock, Jr Memorial Laboratory, Susan G. Komen Foundatron, and the Frances Williams Preston Laboratory of the T. J. Martell Foundation.

RA P-PCR

283

References 1 DuPont, W D and Page, D L (1985) Risk factors for breast cancer m women with proliferative breast disease N Engl J Med 312, 146-15 1 2. London, S. J , Connolly, J L , Schmtt, S J., and Colditz, G. A (1992) A prospectrve study of benign breast disease and risk of breast cancer JAMA 267,941-944 3 Betsill, W. L., Rosen, P P , Lieberman, P H., and Robbins, G F (1978) Intraductal carcinoma. long-term follow-up after treatment by biopsy alone JAMA 239, 1863-I 867 4 Page, D L , DuPont, W D , Rogers, L W., and Landenberger, M (1982) Intraductal carcinoma of the breast follow-up after biopsy. Cancer 49,75 l-758 5 Page, D. L. and DuPont, W D (1990) Anatomrc markers of human premalignancy and risk of breast cancer Cancer 66, 1326-1335. 6. Holt, J T , Jensen, R A., and Page, D. L. (1993) m Cancer Surveys-Advances and Prospects m Clmcal, Epldemiologxal and Laboratory Oncology, vol 18 (Fenttman, I. and Taylor-Papadlmttriou, J., eds ), Cold Spring Harbor Laboratory, Cold Spring Harbor, New York, pp 115-l 33. 7 Jensen, R A , Page, D L , and Holt, J T (1994) Identtficatton of genes expressed m premahgnant breast disease by microscopy-directed clonmg Proc Nat1 Acad Sci USA 91,9257-9261 8 Thompson, M. E , Jensen, M E , Obermrller, P S , Page, D L , and Halt, J T (1995) Decreased expression of BRCAI accelerates growth and 1s often present during sporadic breast cancer progresston Nature Genet 9,444-450 9. Ltang, P., Averboukh, L , Keyomarst, K , Sager, R , and Pardee, A. B. (1992) Differential display and clonmg of messenger RNAs from human breast cancer versus mammary eptthehal cells. Cancer Res 52, 6966-6968. 10 Liang, P and Pardee, A. B. (1992) Differential display of eukaryotic messenger RNA by means of the polymerase chain reaction Sczence 257,967-97 1 11 Lrang, P , Averboukh, L , and Pardee, A. B. (1993) Distribution and cloning of eukaryottc mRNAs by means of differential display refinements and optrmizatlon. Nucleic Acids Res 21, 3269-3275 12. Welsh, J. and McClelland, M (1990) Fingerprmting genomes using PCR with arbitrary prtmers Nucleic Acids Res 18, 72 13-72 18 13. Welsh, J and McClelland, M. (1991) Genomtc fingerprmtmg using arbitrarily primed PCR and a matrix of pan-wise combinattons of primers Nuclezc Aczds Res 19,5275-5279. 14. Welsh, J., Chada, K., Dalal, S S , Cheng, R., Ralph, D., and McClelland, M (1992) Arbitrarily primed PCR fingerprinting of RNA. Nuclezc Aczds Res. 20, 4965-4970

23 The Application of Differential Display to the Brain Adaptations for the Study of Heterogeneous

Tissue

Joseph WI. Babity, Richard A. Newton, Mario E. Guido, and Harold A. Robertson 1. Introduction One of the main advantages of using differential display is the ability to examine simultaneously gene expression in multiple mRNA populations. In other techniques, such as differential screening or subtractive hybridization, only two mRNA populations can be easily examined at the same time. This feature is of particular importance in identifying specific changes in gene expression within a complex biological system. It has been estimated that at least 30% and perhaps as many as 50% of all mammalian genes code for proteins that are uniquely expressed m the brain (1). In addition, with the possible exception of the immune system, the brain is the most heterogeneous tissue m the body. Therefore, the analysis of alterations in gene expression in the brain is complicated by the complextty of the message population and the heterogeneity of tissues within the brain. The differential display techmque as originally described (2) was developed and proven on homogenous cell lines and many of the applications have been specific to homogenous cell lmes. However, a limited amount of work had been done assessingthe utility of differential display analysis in heterogeneous tissues and in particular the analysis of complex physiological changes within an in vivo biologrcal system. Our interest was to examine changes m gene expression m response to neuroplastic changes in the brain. Previous analysis in one model of neuroplastic change, kainic acid-induced seizures, suggested that as many as 1000 changes in mRNA expression might be associated with neuronal plasticity (3). Therefore, the challenge in examining changes in gene expression in the brain was From

Methods m Molecular Bology, Vol 85 Dffferenflal Edtted by P Llang and A B Pardee Humana

285


286

Babity et al.

not to tsolate changes m gene expresston, per se, but to tsolate btologtcally significant changes m gene expresston. In particular, the differential display technique 1suseful m addressing this problem because it allows the comparison of gene expression m multtple RNA populattons. In other techniques, such as differential screening or subtractive hybrtdtzation, only two experimental condtttons can be easily examined at the same time. Therefore, the differential display screenmg technique allowed us to design a much more selective screen for changes m gene expression associated with excttotoxic inJury leading to delayed neuronal degeneration. In addition to the study of delayed cell death we have applied the technique to the detection of genes that may play a role m ctrcadian rhythmtctty and psychotropic drug reward. In each case, isolation of differenttally displayed fragments spectfic to the btological question was facilitated by examining gene expression in discrete brain regions at appropriate time-points. The followmg chapter will focus on the apphcatton of dtfferentral display to the discovery of changes m gene expression associated with delayed cell death 7.7. Experimental

Design: isolation

of a Gene in Hippocampus

Related to Delayed Cell Death

In response to kamate-induced seizures, specific bram regions undergo delayed cell death approx 24 h after seizure acttvity Regions of the brain that are most severely affected include the ptriform cortex and areas within the hippocampal formatton (4,.5). In contrast, other regions of the brain, mcludmg the partetal cortex, remain largely unaffected. By comparmg gene expression m the piriform cortex and htppocampus versus the parietal cortex, rt 1sposstble to identify candidate genes associated with delayed neuronal degeneration. Additional crtterta of our search strategy included an examination of the temporal pattern of gene expresston Prevtous work has shown that immedtate early genes such as C-$X and c-lun are turned on m response to kamtc acid-induced seizure acttvtty (6). Smce these genes are known to be mvolved m the regulation of transcriptional activity, it is reasonable to assumethat downstream genes are bemg regulated at the transcrtpttonal level m response to nnmedtate early gene activity. Immediate early genes are rapidly induced m response to specific stimuli and then are rapidly turned off. Therefore, there 1sa charactertsttc temporal pattern of immediate early gene expresston that occurs m response to kamtc acid-induced seizure activity. Specttically, Fos and Jun gene products are turned on within 30 mm of seizure activity and are turned off by 3 h after seizure acttvtty. Therefore, we reasoned that the most appropriate time to look for changes m gene expression that occur owing to immediate early gene expression 1ssometime between 3 and 6 h after kamic acid admmistration. We were able to compare gene expression m the htppocampus, ptrtform cortex and

Apphcation of Different/al Display

287

partetal cortex at 3,6, and 12 h after kaimc acid admmistration. To our minds, this provided a powerful approach that was particularly suited to differential display screenmg. We examined alterations in gene expression using 25 primer combinations that allowed the display of approximately 100 cDNA fragments per primer combination. Since it has been estimated that there are between 10,000 and 15,000 mRNA transcripts per cell, a reasonable estimate of the number of expressed genes in a specific brain tissue is around 25,000. Therefore, we estimate that we had screened approximately 10% of the messages expressed m the brain regions of interest. On average, several differential display fragments were identified for each primer set. In total, approx 25 differential display fragments correlated with changes m gene expression were observed. This number, when extrapolated, fits well with the estimate of genes that undergo changes m gene expression m responseto kainic acid-induced seizureactivity (3). Among these 25 cDNA fragments that appeared to have altered transcription levels, only a handful of genes fit the criteria outlined above for genes that may be mvolved m delayed neuronal cell death. Genes expressed m all trssues or genes expressed only at 3 h, were ehmmated from further study. These were thought to represent immediate-early genes and a number of these were found Only differentially displayed fragments expressed at later times (6, 9, and 12 h) and m tissues (hippocampus, piriform cortex) where delayed cell death would occur at 24 h were sought. SIX such differential display fragments exemplified by Fig. 1 were identified. Clonmg and sequencmg of the fragment of Fig. 1 showed that tt represented a new member of the synaptotagmm gene family that we call synaptotagmin X (7). The pattern of expression was verified by Northern analysis (Fig. 2) and in sztuhybridtzation. The differential display protocol was based on that origmally described by Liang and Pardee (2) and IS detailed below. We have also recently taken advantage of the significant advances m differenttal display technology (8) that improve reproducibiltty, reduce the number of false positives and faciletate more rapid screening, and these are detailed in the Notes section. 2. Materials 2.1. RNA Preparation 1. Dounce homogenizer(baked, seeNote 1) 2 TRIzol@Reagent(Gibco-BRL) 3 Chloroform (BDH, ACS grade) 4. 5 6. 7.

Isopropanol (BDH, ACS grade) 100% Ethanol 75% Ethanol. 2MNaCl (made RNase-free, see Note 1)

288

Babity

1 C Chippocampus 3 3 612121C

Cpiriform 3 3 6cortex 612121C

et al.

parietal cortex C 3 3 6 612121

Fig. 1. Differential display of mRNA comparing gene expression between hippocampus, piriform cortex, and parietal cortex after kainic acid administration. C, 3,6, and 12 denote control, 3, 6, and 12 h, respectively. Reactions were performed in duplicate using 5’-TTTTTTTTTTTCG-3’ as an anchored primer and 5’-TACAACGAGG-3’ as a random arbitrary primer. The arrow indicates a PCR-amplified cDNA fragment that appears to be induced only in degenerating tissues (hippocampus and piriform cortex) at 6 and 12 h after kainic acid injection. In the hippocampus and piriform cortex, baseline levels of expression are seen in control animals and at the 3 h time-point. By comparison, only baseline levels of expression are observed in the parietal cortex, a tissue in which seizure-induced neuronal degeneration is not observed. 8. DNase I (RNase-free): (RQl TM,Promega) 1 U/pL. Store at -20°C. 9. 3M Sodium acetate. 10. dH,O (see Notes 1 and 2).

2.2. Reverse

Transcription

1. M-MLV Reverse Transcriptase, 200 U/a (Gibco-BRL). 2. 5X Reverse transcriptase buffer (supplied by Gibco-BRL with M-MLV reverse transcriptase): 250 mA4 Tris-HCl, pH 8.3, 375 mi!4 KCl, 15 mA4 MgCl,. 3. O.lM DTT (supplied by Gibco-BRL with M-MLV reverse transcriptase). 4. dH,O. 5. Ribonuclease inhibitor: RNasinTM (Promega) 20 U/k. 6. dNTPs (Pharmacia). 7. Anchored cDNA synthesis primer: 5’-TTTTTTTTTTTMN-3’ where M = C,G, or A and N = T, C, G, or A. (Primers were synthesized by Oligos and the like and were reverse phase-cartridge purified.)

2.3. Differential

Display

1. Arbitrary 10-mer primers (Operon Technologies, Alameda, CA) used in combination with anchored primer (5’-TTTTTTTTTTTMN-3’).

Application of Differential Display

289

parietal cortex piriform cortex hippocampus 1 0 3 6 12 24’ 0 3 6 12 241 0 3 6 12 241 ....I SV

‘:

p-actin Fig. 2. Northern blot analysis of syt X expression. Poly(A)+ RNA from parietal cortex, hippocampus, and piriform cortex at 0, 3, 6, 12, and 24 h after kainic acid-induced seizures was probed with syt X. The probe identifies a single message that is 8.0 kb in length. The blot was also probed with a p-actin oligonucleotide probe to verify that comparable levels of poly(A)+ RNA were present in each lane. 2. Gene Amp PCR Kit (Perkin-Elmer): This includes the 10X PCR buffer (100 mit4 Tris-HCl, pH 8.3, 500 mM KCl, 15 mM MgCl*, dNTPs and AmpliTaq DNA polymerase (5 U/g). 3. a[33P]dATP, 2000 Ci/mmol (NEN DuPont, stable-when-frozen version, see Note 3). 4. PTC- 1OOTMThermal Controller (MJ Research). 5. Formamide loading buffer (95% formamide, 0.2% bromophenol blue. 0.2% xylene cyanol). 6. 6% Polyacryamidel8iU urea solution in 1X TBE. 7. X-Ray film (e.g., Kodak BIOMAXTM MR).


and Subcloning

1. Invitrogen TA vector (Invitrogen). 2. Ampicillin (Sigma, St. Louis, MO). 3. LB media: 10 g bacto-tryptone (Difco), 10 g NaCl, 5 g yeast extract (Difco) per liter. Adjust to pH 7.0. 4. Wizard Minipreps (Promega) plasmid isolation kit.

2.5. Southern 1. 2. 3. 4. 5. 6.

and Northern

Hybridization

Nylon hybridization membrane (Zeta-Probe@GT). Whatman 3MM filter paper. Stratalinker@ UV crosslinker. RibosepTM mRNA Isolation Kit (Collaborative Research). Ready-To-Go DNA labeling kit (AMP) (Pharmacia Biotech). cx[32P]dATP, 3000 Ci/mmol (Amersham).

3. Methods To obtain successfbl differential display results, it is imperative that high quality RNA is used as a starting material. In addition, the highest degree of

290

Babity et al.

reproduclblllty IS obtained when fresh materials are used and the time m between cDNA synthesis and differential display 1skept to a mmlmum. 3.1. RNA Isolation 1 Tissues used for RNA lsolatlon should be quickly dlssected,snap frozen and stored in liquid nitrogen until used (seeNote 4). 2. Frozen tissue may be transferred du-ectly into a dounce homogemzer contammg TRIzol Reagent (see Note 5). As the frozen tissue thaws, homogenize until the sample IS umformly lysed Rapld homogemzatlon ensures mmlmal RNA degradation 3 Add 2/l 0 vol of chloroform directly to the homogenate, shake vigorously for 15 s and place on ice for 5 mm Centrifuge the suspension at 12,OOOgfor 15 mm at 4°C 4 After centnfugatlon, recover the colorless upper aqueous phase taking care to avoid material at the interface Add 0 5 mL of lsopropanol per 1 mL of homogenate, mix and store for 15 mm at 4°C Centrifuge at 12,000g for 15 mm at 4°C Wash the RNA pellet once with 75% ethanol, adding 1 mL 75% ethanol per 1 mL of TRIzo Reagent used for the mltlal Isolation. Vortex the sample and centrifuge at 7500g for 5 mm at 4°C 5 Air dry the RNA pellet for 15 mm on the benchtop Overdrying the pellet can result m a pellet that 1s dlfflcult to resuspend Resuspend the pellet m RNase-free dH,O.

6 In order to further purify the total RNA, add l/10 vol of 2MNaCl to the resuspended RNA and precipitate with two volumes of 100% ethanol for 1 h at -20°C Wash the precipitate with 75% ethanol, air-dry, and resuspend m RNase-free dH,O 7 Treat the RNA sample with an RNase-free DNase to remove any possible DNA contammatlon from the sample. Extract with an equal volume of phenol.chloroform Extract with an equal volume of chloroform. Ethanol preclpltate to recover the RNA (see Note 6).

3.1.1 Use of Differential Display to Discover Genes in Small Brain Nuclei As differential display is based on the polymerase cham reactlon, it is mherently capable of analysis of very small samples. Because we are also interested in the role of gene expression m the regulation of circadian rhythmlclty m mammals, we are forced to work with the very small amount of tissue available from the rat suprachiasmatic nucleus (SCN) (9). For the SCN (or any other small brain region), animals are killed by rapid decapitation and the brain rapIdly removed on Ice, dissected to remove a block containing the hypothalamus and frozen on dry ice. The frozen hypothalamus 1smounted on a cryostat and a 660~w-thick brain section cut at a cryostat temperature of-12 to -14°C. The tissue sample 1spunched out usmg a 16-gage needle that has been flattened and bevelled to produce a punch. For differential display and Northern analysis, SCN tissue 1scollected from 15-20 animals. A tissue punch can be constructed

297

Application of Differential Display from an appropriate size needle and thus this procedure almost any brain region,

3.2. cDNA Synthesis

can be adapted

to

(see Notes 7 and 8)

1 Heat-denature 0 2 pg of DNA-free total RNA m 2 pL of RNase-free dHzO at 65°C for 5 mm, and then chill on ice. 2 Prepare a master mix that contains the followmg components per each reverse transcription reaction performed. 4 pL reverse transcription buffer, 2 pL 0. 1M DTT, 1 @. RNasin, 1 6 pL 250 w dNTPs, 2 pL 10 Cul/iTitMN, 6 4 pL dH,O. 3. Add 17 pL of the master mix to the denatured RNA, incubate 10 min at 37°C then add 1 pL MMLV reverse transcriptase and incubate for a further 50 mm 4 At the end of the reverse transcription reaction, heat mactrvate the reverse transcrtptase by denaturing at 95’C for 5 mm Briefly spin the tubes to collect condensatron. 5 Set the tubes on Ice for PCR or store at -2O’C for later use

3.3. Differential

Display Reaction (see Notes 9-l I)

1. Prepare reactions in duplicate, on ice, in 500 pL micro test-tubes. 2 Prepare a master mrx that contains the following components for each differential display reaction: 2 pL 10X PCR reaction buffer, 10 $ dH20, 1 6 p.L 25 @4 dNTP mix, 0.2 pL a[33P]dATP, and 0 2 PI of AmpliTaq DNA polymerase 3 For each dtfferential display reaction combme 2 pL cDNA synthesis reaction, 2 pL anchored 3’ primer (10 ClM) and 2 pL arbitrary 5’ primer (10 CLM). 4. Add 14 pL of master mix to each differential display reaction tube and mix well by plpetmg up and down. 5. Begin PCR cycling using the following cycling parameters 94°C for 1 mm; followed by forty cycles each of 94°C for 30 s, 40°C for 2 mm, and 72°C for 30 s, followed by 72°C incubation for 7 min, and then a 4°C soak. 6 The differential display reactions may be stored at -2O’C until ready to analyze.

3.4. Gel Electrophoresis Gel electrophoresis methodology varies greatly depending on the particular electrophoresis apparatus used. For this reason, specific mstructions relating to the pouring and running of polyacrylamide gels will not be discussed in detail. Instead instructrons specific to the differential display technique will be addressed. 1. Pour a 6% polyacrylamlde/SM urea sequencmg gel. 2 Mix 4 pL of the differential display reaction with 2 pL of loading buffer and heat denature at 80°C for 3 min. Place samples on Ice unttl ready to load. 3 Remove any excess urea or acrylamide debris by flushing the top of the gel with a Pasteur pipet 4. Position the sharkstooth comb between the glass plates to create sample wells. 5 Load 4 pL of each sample onto a prewarmed gel to ensure denatunng conditions.

292

Babity et al.

6. Electrophorese the samples until the xylene cyan01 dye has Just run off of the gel. 7 Dismantle the sequencing gel apparatus, separate the glass plate from the gel, and blot the gel onto 3MM Whatman filter paper and cover wrth plasttc wrap 8 Dry the gel at 80°C under vacuum for 1.5-2 h. 9. Tape the X-ray film onto the drred gel and align the film to the gel by using an 1s-gage needle to poke holes through the film and gel 10. Usually an overnight exposure IS sufficient to vtsuahze the differential display banding patterns.


and Subcloning

1. Once a candidate fragment has been identified, use a clean razor blade to cut the differentially displayed fragment from the gel 2 Place the excised gel fragment mto a tube containing 40 pL of dHzO and boll for 15 min rn order to elute DNA from the gel The remainder of the eluted gel fragment can be stored at -20°C Centrifuge the tube at 12,000g for 1 min to spin down paper and gel fragments 3 Reamphfy the fragment of interest by combmmg the followmg reagents. 7 pL eluted DNA, 5 pL of 10X PCR buffer, 4 pL 250 w dNTP, 5 pL 10 @4 anchored 3’ primer, 5 pL 10 ~.IIVS’ arbitrary primer, 0.5 pL of AmpliTaq DNA polymerase and 23 5 pL dH20 PCR cycling was performed using the same cyclmg parameters employed m the ortgmal PCR generation of the fragment (see Note 12) 4. Analyze 10 pL of the amplified reaction on a 1% TBE agarose gel The reamphfied cDNA fragment should be the same size as the fragment eluted from the original polyacrylamide gel. 5 For subclonmg, Qtaqutck gel-purified cDNA bands from fresh PCR reactions were prepared (see Note 13) and 3 1 molar ratio of cDNA to TA vector used for ligation overnight at 15°C. Following transformatton of competent cells and blue/ white color selection on amprcrllin plates (100 pg/mL) overlaid with X-GAL, posrttve colonies were grown overnight m LB-amptctllm, and plasmtd DNA ISOlated Those plasmrd preparations possessing an appropriate size Insert, as determined by restriction digestion and agarose gel electrophoresrs, were sequenced using a T7SequencmgTM kit (Pharmacra Brotech). Generally subclonmg was only performed once the cDNA fragment had been confirmed by Northern analysts to be dtfferenttally expressed.

3.6. Southern

Blot Analysis

Southern hybridization provides a postttve control for probe quality m the absence of a signal on a Northern blot and also allows confirmatron that a particular clone isolated from a PCR product is responsible for the differential banding pattern on the gel. 1. Cut a region of the dried gel contammg the band of interest usmg a razor blade 2. Prepare a piece of nylon membrane slightly larger than the gel area of mterest and six pieces of filter paper the same size.

Application of Differential Display

293

3. Cut a piece of filter paper and place on a glass plate such that rt overhangs mto a dish contammg 10X SSC. Soak three sheets of precut filter paper and the dried gel with 3M NaCl, 0.5M Tris-HCl, pH 7.4, and lay, gel uppermost, on the glass plate and flood surface of gel with buffer. 4. Place membrane (previously soaked in H,O then 10X SSC), on top of gel, followed by two pieces of filter paper soaked m 10X SSC and then a dry piece. Stack paper towels such that transfer is through the gel, cover with a plate and a weight (approx 500 g) and leave overnight 5. Wash membrane with 6X SSC and irradiate face up with a UV crosslmker Dry membrane and obtain an autoradiographtc image to ascertain pattern obtamed m the absence of probe

3.7. Northern

Blot Analysis

Confirmation of drfferential expression was generally performed on Poly(A)+ RNA blots containing approx 2 ~18of mRNA/lane. The eluted cDNA fragment or clone can be used as a template for random primed probe synthesis. 1. Use Qiaquick gel purification to purify the cDNA fragment. 2. Label approx 30 ng of denatured cDNA with a[32P]dATP by random prrming. 3 Following 4 h prehybridization, hybridize blots with approx 2 x lo6 cpm/mL in a standard hybridization buffer containmg 50% formamide and 10% dextran sulfate.

4. Notes 1 dHzO and all solutions (not including Tris, nucleotides, or other ammes) used in RNA preparation were treated for 4 h with 0.1% diethylpyrocarbonate (DEPC) then autoclaved. All glassware were treated with 0.5NNaOH for 30 min, rinsed with DEPC-treated HZ0 and then baked at 18O’C for at least 4 h. I 2. dH20 corresponds to NANOpure deionized water (Type I Reagent Grade Water). 3. In our hands, the dye-stabilized radionucleotides do not perform reproducibly under PCR cycling conditions 4 All RNA samples should be prepared by the same isolation procedure in order to ensure reproducibility of the banding pattern. 5. In our hands, using minimal quantities of TRIzol to extract RNA results in low AZ6s/A2s0 ratios and can often lead to poor reproducibility between RT-PCR reactions. 6. Make sure that the RNA is of high quality. Pure RNA should have an A&AZsO ratio between 1.8 and 2.0. We usually examine the DNase-treated total RNA samples on an ethidium bromide stained formaldehyde agarose gel. The ratio of 28s to 18s ribosomal RNA should be roughly two to one. RNA degradation is evident if the 18s ribosomal RNA appears to be more abundant than the 28s ribosomal RNA. Further purification is necessary if the A26s/A2s0ratio is less than 1.8 to 2.0. 7. We have also successfully performed differential display using a single oligo(dT) primer (pd[T],,,s, Pharmacia), rather than one of the 12 anchored primers. The

294

8

9

10

11

12. 13.

Babity et al. resultant cDNA can be used for any primer combmation and provides sufficient template for 100 PCR reactions. 2.0 pg of DNA-free total RNA is combmed with 1 pL of 1 @4ohgo(dT) primer and RNase-free dH,O added to a total volume of 5 pL The mixture is heat-denatured at 65°C for 5 mm and chilled on ice, 5 & of a master mix are added that contam (per reaction) 2 l.tL reverse transcription buffer, 2 p.L 5 mM dNTPs, and 1 pL MMLV reverse transcriptase. Followmg incubation for 1 h at 42”C, the reverse transcriptase is heat inactivated (75’C, 10 mm) and the cDNA samples stored at -70°C The best differential display results are obtained using high quality RNA and fresh cDNA products. The differential display pattern obtained from older cDNA products is noticeably degraded in comparison to fresh products. We normally use a thermal cycler with a heated hd that allows for oil-free PCR amplification. Without a heated hd, a drop of mineral oil is necessary to prevent condensation on the side of the PCR tubes Later work m our laboratory showed that the use of an N-terminal deletion of Taq, such as the AmpliTaq Stoffel fragment, in conjunction with a proofreading enzyme such as Vent produces longer PCR products, m agreement with published reports (8). In addition, the proofreading properties of this enzyme results m a lower error rate m comparison to the native AmphTaq enzyme. This long-distance PCR mixture was used together with longer primers: 30-mer anchored primers and 25-mer arbitrary primers (8) In PCR with extended primers, low stringency annealing at 40°C occurs m the first three cycles of amplification while subsequent cycles occur at 60°C annealing. Therefore randomly primed products generated m the first rounds of PCR are accurately replicated in subsequent cycles and many of the artefacts generated by contmuous low temperature annealing are avoided The use of extended primers also facilitates their use in the direct sequencing of reampltfied cDNA bands In addition, the use of an anchored primer m reamplification that mcorporated an RNA polymerase binding site allowed for the generation of antisense riboprobes Owmg to their single-stranded nature, such probes offer greater sensitivtty than random-primed cDNA probes and hence facilitate the detection of rare messages m Northerns, RNase protection assays and in zn situ hybridization Typically, differential display reactions utthzmg long-distance PCR conditions yields cDNA fragments that range up to 1.5-kb m length. In order to resolve the higher molecular-weight fragments produced by this protocol, we have used a Genomyx LRTM sequencing apparatus. This unit IS a temperature-controlled system that allows better resolution of high-molecular-weight fragments m comparison to a conventional sequencing and therefore reduces the possibihty of multiple band isolation during differential display band excision A high number of PCR cycles during reamplification leads to greater complexity of the final reaction products. The use of gel-purified bands for ligation resulted m a much greater proportion of plasmids from posittve colonies containing inserts of the correct size

Apphca t/on of Differen t/al D/splay

295

References 1 Milner, R J and Sutchffe, J G (1983) Gene expression m the bram Nuclezc Aced Res 11,5497-5520 2 Liang, P and Pardee, A B (1992) Differential display of eukarvotic messenger RNA by means of the polymerase chain reaction Science 257, 967-97 1 3. Nedivi, E , Hevroni, D , Naot, D , Israeli, D , and Curl, Y (1993) Numerous candidate plasticity-related genes revealed by differential cDNA clonmg. Nature 363, 7 18-722 4 Sperk, G. (1993) Kamic acid seizures m the rat. Prog Neurobzol 32, l-32 5. Armstrong, J N , Plumier, J -C , Robertson, H. A., and Currie, R W (1996) The inducible 70-kDa keat shock protein is expressed m the degenerating dentate hilus and piriform cortex after systemic admnnstration of kamic acid m the rat. Neuroscience 74,685%693. 6 Robertson, H. A (1992) Immediate early genes, neuronal plastictty and memory. Blochem Cell Btol. 70,729-737 7 Dtachenko, L. B., Ledesma, J., Chenchtk, A A , and Stebert, P. D. (1996) Combmmg the technique of RNA fingerprintmg and differential display to obtain differentially expressed mRNA Blochem Bzophys Res Comm 219,824-828 8 Guide, M E , Rusak, B., and Robertson, H A (1996) Spontaneous circadian and light-induced expression OfJunB mRNA in the hamster suprachiasmatic nucleus. Bram Res 732,2 15-222

24 Analysis of Gene Expression in Hypothalamus in Obese and Normal Mice Using Differential Display Eleftheria

Maratos-Flier,

Daqing Qu, and Steen Gammeltoft

1. Introduction Obesity may result from increased caloric consumptton, decreased energy expenditure, or both (I) These processes are controlled, at least in part, by the hypothalamus. Hypothalamic dysfunction IS important in the development of obesity and it has long been known that lesions m the ventromedial hypothalamus lead to hyperphagia (2) and obesity, while lesions in the lateral hypothalamus Impair the appetite response and lead to death by starvation. Neuropeptrdes that stimulate appetite such as NPY (3) and galamn, as well as neuropeptides that suppress appetite such as CRF (4), CCK (5), and GLP-1 (6) have been identified. However, understanding of the control of appetite remains incomplete. A key signal from the periphery, leptm, a hormone made m fat that regulates hypothalamic function, was only recently identified (7). While NPY is a potent stimulator of eating behavior, the NPY knockout mouse (8) exhibits a normal feeding phenotype, indicating that other factors are involved in the strmulatron of hunger. One approach to the elucidation of the molecular mechanisms that contribute to hypothalamic dysfunction m obesity is to identify mRNAs that are abnormally regulated. Such studies are difficult because hypothalamic tissue is scarce (a mouse hypothalamus weighs 8-10 mg and yields 8-10 pg of total RNA). We utilized the reverse transcription (RT) polymerase chain reaction (PCR) differential display to attempt to identify abnormally regulated mRNAs. We examined differential expression in the Mob mouse model. These mice are markedly obese as the result of a spontaneous mutation in the leptm gene, which leads to the msertron of a stop codon in the coding sequence. Leptm IS a From

Methock m Molecular Bfology, Vol 85 DffferenM Edited by P Llang and A B Pardee Humana

297


Maratos-Flier,

298

Qu, and Gammeltoft

hormone made and secreted by fat. Homozygous mice make no leptm and, m addition to obesity, have impaired nonshivering thermogenesis, decreased motor activity, and are infertile. Whereas the identification of leptm identified the precise mutation responsible for the ob/ob syndrome, all the physiologic consequences of leptm absence are still not understood. In addttion, the complete spectrum of hypothalamic dysregulation that results from leptm de% ciency is unknown. Recent studies suggest that leptm has a range of actions on the hypothalamus that are important m the neuroendocrme response to starvation (9). Hence identification of differentially expressed transcripts in the obese mice might provide insight into the pathogenesis of the obese state. 2. Materials 2.1. Preparation

of Hypothalamic

Total RNA

1 Preparatton of the mice a Obtain from Jackson laboratortes (Bar Harbor, ME) and house for 4 d after arrtval b Keep on a normal day-night cycle and use m the morning shortly after the start of the light cycle. c Anesthetize with 200 mg/kg sodmm amytal and decapitate 2. Preparation of hypothalamtc RNA a. Remove the brain and excise the hypothalamus b Pool hypothalami from 10 ob/ob and 10 ob/+ mice (weight of each pool -100 mg) c Extract total RNA using RNAzol (Cinna/Btotex Laboratories, Houston, TX) d Assess quality of RNA by agarose gel electrophorests and ethydmm bromide staining

2.2. Primers Obtain primers from a commercial vendor (e g , Ohgos) Perform all reactions with one unique anchored and one unique downstream primer, i.e , degenerate groups of prtmers were not utihzed. 2 Utilize arbitrary primers, 1.e , CAGGCCCTTC, TGCCGACTCTG, AGTCACTCCAC, AATCGGGCTG, GGTCCCTGAC, GAAACGGGTG, GTGACGTAGG,GGGTAACGCC,GTGATCGCCAG,CAATCGCCGT, TCGGCGATAG, CAGCACCCAC, TCTGTGCTGG, TTCCGAACCC, AGCCAGCGAA, GACCGCTTGT, AGGTGACCGT, CAAACGTCGG, GTTGCGATCC (Dr George Kmg suggested the sequence for these primers )

2.3. cDNA Synthesis 1. Synthesize cDNA using M-MLV reverse transcrtptase (Superscript RNAase H reverse Transcrtptase, Gibco-BRL, Grand Island, NY) and one of nme anchored primers.

Gene

Expression

Analysis

299

in Mice

2. Utilize anchored primers T,,VV where V 1s A, C, or G. Generate each set of cDNAs m differential display with 20 upstream arbitrary decamers, for a total of 180 reactlons. 3 Perform reactions in the followmg mixture: 1.O w T,,VV, 20.0 w each dNTP, 0 OIMDTT, 50 mMTrls PH 8.3,75 mA4KC1, 3 mMMgCl*.

3. Methods PCR reactions were performed, with few modifications, as previously described (IO,ll). 3.1. DNAase Digestion

of RNA

1. DNAase digestion mixture (final concentration m 50 J.IL): 0.2 U/pL RNase Inhibitor, 10 mA4Trls pH 8 3,50 mMKC1, 1.5 mMMgCl,, 0 001% gelatin (w/v), 0.48 U/pL DNase (Boehringer-Mannheim, Chlcago, IL) I diluted m 0.1X TE. 2 Add 42 5-pg ahquots of total RNA to the DNase I digestion mixture Incubate the mixture at 37°C for 30 mm. 3. Followmg the digestlon, extract the mixture with phenollCHC13 and precipitate the RNA from the aqueous phase by the addition of 2 vol of absolute ethanol and l/10 vol of sodium acetate. 4 Chill the mixture for 30 mm at -20°C and then centrifuge for 5 mm in a refngerated microfuge at maximum speed.

3.2. cDNA Synthesis 5. Synthesize cDNA by RT by adding total mRNA obtained from the DNAase digestion to the cDNA synthesis mixture at a final concentration of 0.01 pg/p.L RNA 6. Incubate the samples for 5 mm at 65°C and then for 10 mm at 37°C 7 Add 10 U/pL RT to each reaction and samples and incubate for another 60 mm at 37°C followed by a 5-mm incubation at 95°C

3.3. PCR Amplification 8 Amplify the cDNAs generated from the RT reactlon using an AmpliTaq DNA Polymerase (Perkm Elmer, Foster City, CA) or similar equipment 9. Perform 25 cycles consisting of 30 s at 94”C, 2 min at 40°C, and 30 s at 72°C. 10. Label PCR products using 35S-dATP (NEN, Boston, MA) 11. PCR reaction (25 $). 10 mA4 Tris pH 8.3, 50 mA4 KCl, 3 mA4 MgC&, 0.001% gelatin (weight for volume, Sigma), 2 w dNTP, 0.2 @4 upstream primer, 1 0 pA4 T,,VV, 10 % v/v Reverse transcrlptase mix, 0.50 pCl/& S35-dATP, 0.05 U/pL AmphTaq (1 25 U Total)

3.4. Evaluation

and Selection of Differentially

12. Separate resulting DNAs (from the PCR amplification) denaturmg condltlons

Expressed

Bands

on sequencing gels under

Mara tos- Flier, Qu, and Gammeltoft

300

13. Dry gels and expose to Kodak X-OMAT AR film (Eastman Kodak) for 1 to 2 d Compare DNA fragments from ob/ob and ob/+ mice 14 When a primer pan yields bands umque to either mouse, reperform the reacttons and reanalyze using fresh cDNA dertved from a second RT reaction If the unique band appears on the second reaction, excise it from the dried gel and extract the DNA by botlmg m TE buffer and precipitating with ethanol using mussel glycogen (Boehrmger-Mannhelm) as a carrier 15 Reamplify the DNA from each band using the same set of primers and the same thermal cycling conditions. Run the resulting reaction products on an agarose gel, stam with ethydium bromide, and elute from the gel. 16 Ligate the DNA mto the PCR- 1 plasmid (InVnrogen) and sequence the reactions using both Ml 3R and -2 lM13F This determines both sequence and orientation of the insert.

3.5. Confirmation

of Differential

Expression

17. Use Sp6 or T7 promoter to generate a riboprobe (dependmg on nature of band). 18 Confirm altered expression of the band of interest usmg either Northern blot analysts or rtbonuclease protectton assay. 19 Subject 20 pg of total RNA from either ob/+ or ob/ob hypothalamus to agarose gel electrophoresis and transfer to nylon. 20. Prehybridize blots for 4 h at 58’C After addition of probe filters, continue hybrtdization at the same temperature overnight 2 1. Wash filters with 2X SSC, 0.1% SDS at room temperature for 10 mm and then wash once with 0 1X SSC, 0.1% SDS at 60°C for 10 mm 22 Repeat the sequence of washes for a total of three times. An-dry the blots and expose either to Kodak X-OMAT film or analysis using a Molecular Dynamics Phosphorimager

3.6. Examples of Use Display reactions with 180 primer pairs yielded 52 DNA bands that appeared to be differentially expressed on at least two separate reactions. Of these, 35 were further evaluated by either Northern blot analysis or rrbonuclease protection assay. Differential expression was confirmed m only SIX bands, while no difference in expression was seen with 20 bands. No signal could be detected in nine bands. Of the six bands that were differentially expressed, one represented a 453-base fragment of Melanin Concentratmg Hormone (MCH) (Fig. l), which contained the coding region of MCH and two other expressed neuropeptides, N-E1 and N-GE. Thrs fragment was used to confirm altered expression in ob/ob, ob/+, and wild-type +/+ animals (Fig. 2); in addition, differences in expression m fasted and fed animals were examined. Expression of melanin-concentrating hormone was increased twofold in ob/ob ammals when compared to control animals. Expression of MCH mRNA increased approximately threefold after fasting, m

Gene Expression Analysis in Mice

301

Fig. 1. PCR differential display on RNA from the hypothalamus of ob/+ and ob/ob mice. YT,*AC-3’ as the arbitrary primer and S-AATCGGGCTG-3’ as the arbitrary primer. Candidate band, which proved to have the sequence of melanin-concentrating hormone is indicated by the arrow (reprinted with permission from ref. 15).

both normal and obese animals. This difference was similar to differences in the expression of another neuropeptide, NPY. This neuropeptide is known to play an important role in feeding behavior. To confirm that MCH was involved in feeding behavior, we injected peptide (purchased from Bachem, Switzerland) into the lateral ventricles of LongEvans rats. Rats injected with a single bolus of 5 pg of MCH acutely consumed two- to threefold more food than rats injected with artificial CSF only (Fig. 3). Melanin-concentrating hormone was originally isolated from chum salmon pituitaries in 1983 (12) and was found to be a hormone of 17 amino acids that induced aggregation of melanosomes in fish melanophores. Mammalian MCH was identified in 1989 using traditional peptide isolation techniques. Rat MCH was purified by affinity chromatography using anti-salmon-MCH antibodies (13). Rat MCH is a 19 amino acid peptide that is highly homologous to fish

302

Maratos-Flier,

Qu, and Gammeltoft

Fig. 2. Northern blot of hypothalamic RNA from +/+, ob/+ and ob/ob animals in the fed and fasted state. 20 pg of total RNA was loaded onto each band and blots were probed with a riboprobe made against either MCH or NPY mRNA (reprinted with permission from ref. 15).

MCI-I. Subsequently, it was shown that mouse MCH and human MCH are identical to rat MCH. In mammals, MCH is expressed in the lateral hypothalamus and the zona incerta, and does not circulate. Expression outside the CNS, if it occurs, occurs at very low levels. MCH neurons project to the median eminence and directly to frontal cortex. The distribution of neurons projecting to frontal cortex is interesting and suggested that MCH was involved in the regulation of goal-oriented behaviors such as food intake or possibly general arousal (14). The role of MCH in mammalian physiology remained obscure until our observation of increased expression of MCH mRNA in the ob/ob mouse (15). This finding, coupled with the finding of increased expression following fasting in both normal and ob/ob mice, led to the speculation that MCH might be involved in the regulation of eating behavior. This hypothesis was confirmed in experiments in which it was shown that MCH injected into the lateral ventricles of rats induced increased eating. The identification of MCH as a potential regulator of neuroendocrine aspects of obesity resulted from differential display screening. Current experiments are focused on defining other actions relevant to nutritional homeostasis as well as the interaction of MCH and other central regulators of energy balance.

Gene Expression Analysis in Mice

303

y!/r,,#, 0123456

Time (Hours)

Fig. 3. Effect of 5 ug of MCH (gray diamond) given mtracerebroventrtcularly on chow consumptton in rats. Control rats (black squares) recetved an inJection of artificial cerebrospinal fluid *p = 0 0077, **p = 0 012.

4. Notes For PCR amplrfication the reaction vials were kept on ice at 4°C while reactants were added. This was important for obtammg reproducible display gels. In addition, the upstream primer was always added next to last, after the addition of the AmpliTaq

and prtor to the addition

of the S35-dATP.

References 1 Flier, J. S. (1995) The adrpocyte* Storage depot or node on the energy mformatton superhighway. Cell 8, 15-18 2 Brobeck, J. R. (1946) Mechanisms of development of obestty in animals with hypothalamic lesions. Physzol Rev 256,54 1. 3. Stanley, B. G., Anderson, K C , Grayson, M. H , and Lerbowttz, S. F (1989) Repeated hypothalamic stimulation with neuropeptrde Y increases dally carbohydrate and fat intake and body weight gain in female rats. Physzol Behav. 46, 173-177. 4 Hemrrchs, S C , Manzaghr, F , Ptch, E M., Hauger, R. L., and Koob, G. F (1993) Cortrcotropin releasing factor m the paraventrtcular nucleus modulates feeding induced by neuropepttde Y Brain Res 611, 18-24

304

Maratos- Flier, Qu, and Gammeltoft

5 Baile, C. A , McLaughlin, C. L., and Della-Ferra, M. A. (1986) Role of cholecystokinm and oprod peptide m control of food Intake Physrol Rev 66, 172-234. 6 Turton, M D , O’Shea, D., Gunn, I., Beak, S. A , Edwards, C. M B , Meeran, K., Chow, S J , Taylor, G M., Heath, M M , Lambert, P D , Wilding, J. P H., Smith, D M., Ghatel, M A., Herbert, J. M., and Bloom, S. R (1996) A role for glucagon-like pepttde-l m the central regulation of feeding Nature 379, 69-72 7 Zhang, Y ,, Proenca, R , Maffer, M , Barone, M , Leopold, L , and Frredman, J M (1994) Posmonal clonmg of the mouse obese gene and its human homologue Nature 372,425-432 8 Errckson, J. C., Clegg, K E , and Palmrter, R D. (1996) Sensmvrty to leptin and susceptrbllny to seizures of mice lacking neuropeptrde Y Nature 381,4 15-4 18. 9 Ahrma, R S., Prabakaran, D., Mantzoros, C S., Qu, D , Lowell, B B , Maratos-Flier, E , and Flier, J S (1996) Role of leptm m the neuroendocnne response to fastmg Nature 382,250-252. 10 Lrang, P and Pardee, A B (1992) Differential display of eukaryottc messenger RNA by means of the polymerase chain reactton. Sczence 257,967-97 1. 11 Atello, L. P , Robmson. G. S., Lm, Y W , Ntshto, Y., and King, G. L (1994) Identificatton of multiple genes m bovine retinal pertcytes altered by exposure to elevated levels of glucose by using mRNA differential display. Proc Nutl Acad Scl USA 91,623 l-6235. 12 Kawauchi, H , Kawazoe, I., Tsubokawa, M , Klshida, M , and Baker, B I. (1983) Charactertzatron of melanin-concentratmg hormone m chum salmon prtuitartes Nature 305,321-323. 13 Vaughan, J M , Fisher, W H , Hogar, C., Rtvrer, J., and Vale, W. (1989) Characterization of melanin concentrating hormone from rat hypothalamus. Endocnnology 125, 1660-1665. 14 Nahon, J. L. (1994) The melanm concentratmg hormone: From the peptide to the gene. Cut Rev Neurobzol 864,221-262 15 Qu, D., Ludwig, D. S , Gammeltoft, S , Piper, M., Pelleymounter, M. A , Cullen, M. J , Mathes, W F , Przypek, J., Kanarek, R , and Maratos-Flier, E (1996) A role for melanm-concentrating hormone in the central regulatton of feeding behavior. Nature 380,243-247.

Differential Display Methods and Protocols (Methods in Molecular Biology)

Differential Display Methods and Protocols (Methods in Molecular Biology)

Differential Display Methods and Protocols