Progress in Medicinal Chemistry, Volume 39

Progress in Medicinal Chemistry 39

This Page Intentionally Left Blank

Progress in Medicinal Chemistry 39

Editors: F.D. KING,

BSC.,D.PHIL.. C.CHEM., F.R.S.C.

GlnxoSmithKline New Frontiers Science Park (North) Third Avenue Hurlow, Essex CM19 5 A W United Kingdon?

and A.W. OXFORD,

M.A., D.PHIL.

Consultunt in Medicinal Chemistry P.O. Box IS1 Royston SG8 5 Y Q United Kingdom

2002

ELSEVIER AMSTERDAM-BOSTON-LONDON-NEWYORK*OXFORD-PARlS*SANDIEGO. SAN FRANCISCO.SINGAPORE.SYDNEY.TOKY0

ELSEVIER SCIENCE B.V. Sara Burgerhartstraat 25 P.O. Box 21 I , 1000 AE Amsterdam, The Netherlands

c) 2002 Elsevier Science B.V. All rights reserved. This work is protected under copyright by Elsevier Science, and the following terms and conditions apply to its use: Photocopying Single photocopies of single chapters may be made for personal use as allowed by national copyright laws. Permission of the Publisher and payment of a fee is required for all other photocopying, including multiple or systematic copying, copying for advertising or promotional purposes, resale, and all forms of document delivery. Special rates are available for educational institutions that wish to make photocopies for non-profit educational classroom use.

Permissions may be sought directly from Elsevier Science Global Rights Department, PO Box 800, Oxford OX5 IDX, UK; phone: (+44) 1865 843830, fax: (+44) 1865 853333, e-mail: [email protected]. You may also contact Global Rights directly through Elsevier’s home page (http://www.elsevier.com),by selecting ‘Obtaining Permissions’. In the USA, users may clear permissions and make payments through the Copyright Clearance Center, Inc., 222 Rosewood Drive, Danvers, M A 01923, USA; phone: ( + I ) (978) 7508400, fax: ( + I ) (978) 7504744, and in the UK through the Copyright Licensing Agency Rapid Clearance Service (CLARCS), 90 Tottenhain Court Road, London WIP OLP, UK; phone: (+44) 207 631 5555; fax: (+44) 207 631 5500. Other countries may have a local reprographic rights agency for payments. Derivative Works Tables of contents may be reproduced for internal circulation, but permission of Elsevier Science is required for external resale or distribution of such material. Permission of the Publisher is required for all other derivative works, including compilations and translations. Electronic Storage or Usage Permission of the Publisher is required to store or use electronically any material contained in this work, including any chapter or part of a chapter.

Except as outlined above, no part of this work may be reproduced, stored in a retrieval system or transmitted in any form or by any means, electronic, mechanical, photocopying, recording or otherwise, without prior written permission of the Publisher. Address pennissions requests to: Elsevier Science Global Rights Department, at the mail, fax and e-mail addresses noted above. Notice

No responsibility is assumed by the Publisher for any injury and/or damage to persons or property as a matter of products liability, negligence or otherwise, or from any use or operation of any methods, products, instructions or ideas contained in the material herein. Because of rapid advances in the medical sciences, in particular, independent verification of diagnoses and drug dosages should be made. First edition 2002 British Library Cataloging in Publication Data

1. Pharmaceutical chemistry I. King, F. D. 11. Oxford, A. W. 6 15.1’9 ISBN: 0 444 50959 3

Please refer to card number 62-2712 for this series. ISBN: 0 444 509593 (this volume) ISBN (series): 0 7204 7400 0

@ The paper used in this publication meets the requirements of ANSINIS0 239.48-1992 (Permanence of Paper). Printed in The Netherlands.

Preface Five important topics in medicinal chemistry are reviewed in this volume. Chapter 1 provides a comprehensive account of inhibitors of the caspase family of proteolytic enzymes that represent a new class of anti-inflammatory and antiapoptotic agents of potential value in rheumatoid arthritis, brain ischaemia and other central and peripheral indications. The extent to which the peptidic character of early inhibitors has been reduced is also discussed. It is imperative that novel approaches are urgently pursued to overcome the increasing problems of resistance to antibacterial and antiviral agents, especially to HIV. Hence chapter 2 focuses on recent advances in our understanding of the binding of these agents to ribosomal RNA or RNA/ protein complexes as this should provide a major impetus to the design of sinall molecules. Included also is a well documented survey of semi-synthetic and totally synthetic antibiotics and anti HIV agents and their sites of interaction. A plethora of structural types, including many derived from natural products, have been described as inhibitors of the intracellular enzyme acylCoA: cholesterol O-acyltransferase (ACAT) and these are comprehensively reviewed in chapter 3. Although disappointing in hypercholesterolemia, for which they were originally developed, ACAT inhibitors may prove effective as anti-atherosclerotic agents. Strategies to reduce the peptide character of growth hormone secretagogues leading to drugs with improved oral bioavailability and duration of action are described in chapter 4. Encouraging clinical studies with first generation compounds in growth hormone deficient children and adults suggest these agents will fulfil an unmet medical need. The proteolytic enzyme, hepatitis C virus NS3-4A protease, required for viral replication, is one of the most attractive targets for HCV infections. The problems in designing potent inhibitors is lucidly described in chapter 5 and traces the evolution so far of non-peptide inhibitors from peptides.

vi

PREFACE

We are most grateful to all the authors of this volume for committing so much of their time and effort to assessing the extensive literature of their topics and compiling these reviews. We also thank the staff of the publishers for their continuing support and encouragement to the series.

July 2001

F. D. King A.W. Oxford

vii

List of Contributors Fareed Aboul-Ela Department of Structural Biology, RiboTargets Ltd., Granta Park, Abington, Cambridge CB1 6GB, U.K. Mohammad Afshar Department of Drug Design, RiboTargets Ltd., Granta Park, Abington, Cambridge CB1 6GB, U.K. Michael Ankersen Medicinal Chemistry Research I, Novo Nordisk A/S, Novo Nordisk Park, DK-2760 Denmark Martin J. Drysdale Department of Chemistry, RiboTargets Ltd., Granta Park, Abington, Cambridge CBI 6GB, U.K. Piotr P. Graczyk Department of Medicinal Chemistry, EISAI London Research Laboratories, University College London, Bernard Katz Building, London WClE 6BT, U.K. Brian R. Krause Pfizer Global Research and Development, Ann Arbor Laboratories, 2800 Plymouth Road, Ann Arbor, MI 48105. U.S.A. Ann D. Kwong Vertex Pharmaceuticals, Inc. 130 Waverly Street, Cambridge, MA 02139. U.S.A. Georg Lentzen Department of Drug Discovery, RiboTargets Ltd., Granta Park, Abington, Cambridge CBl 6GB, U.K.

...

Vlll

LIST OF CONTRIBUTORS

Natalia Matassova Department of Drug Discovery, RiboTargets Ltd., Granta Park, Abington, Cambridge CBI 6GB, U.K. Alastair I. H. Murchie Department of Drug Discovery, RiboTargets Ltd., Granta Park, Abington, Cambridge CBI 6GB, U.K Robert B. Perni Vertex Pharmaceuticals Inc., 130 Waverly Street, Cambridge, MA 02139, U.S.A. Joseph A. Picard Pfizer Global Research and Development, Ann Arbor Laboratories, 2800 Plymouth Road, Ann Arbor, MI 48105, U S A . Drago R. Sliskovic Pfizer Global Research and Development, Ann Arbor Laboratories, 2800 Plymouth Road, Ann Arbor, MI 48105, U.S.A.

Contents List of Contributors Preface

vii V

Caspase Inhibitors as Anti-inflammatory and Antiapoptotic Agents Piotr P. Graczyk

1

RNA as a Drug Target Martin J. Drysdale, Georg Lentzen, Natalia Matassova, Alastair I.H. Murchie, Fareed Aboul-Ela and Mohanimad Afshar

73

ACAT Inhibitors: The Search for a Novel and Effective Treatment of Hypercholesterolemia and Atherosclerosis Drago R. Sliskovic, Joseph A. Picard a n d Brian R. Krause

121

Growth Hormone Secretagogues: Discovery of Small Orally Active Molecules by Peptidomimetic Strategies Michael Ankersen

173

Inhibitors of Hepatitis C Virus NS3.4A Protease: An Overdue Line of Therapy Robert B. Perni and Ann D. Kwong

215

Subject Index Author Index (Vols. 1-39) Subject Index (Vols. 1-39)

257 263 269

This Page Intentionally Left Blank

Progress in Medicinal Chemistry - Vol. 39, Edited by F.D. King and A.W. Oxford 0 2002 Elsevier Science B.V. All rights reserved.

1 Caspase Inhibitors as Anti-inflammatory and Ant iapoptotic Agents PIOTR P. GRACZYK Department qf Medicinal Chemistry, EISAI London Research Laboratories, University College London, Bernard Katz Building, London WCIE 6BT. U.K.

INTRODUCTION ROLE OF CASPASES Caspase- I Caspase-2 Caspase-3 Caspases-4 and -5 Caspases-6 and -7 Caspase-8 Caspase-9 Caspases- 10 and - 1 1 Caspases-12, -13. and -14

4 4 6 7 8 8 8 8 9 9

REGULATION OF CASPASES Transcriptional regulation Post translational modification Activation

9 9 II II

INHIBITION OF CASPASES Protein inhibitors P35 I AP CrmA Other serpins: Serp2. P I 9 Non-peptidic small molecules Peptide-derived inhibitors Pharmacophore moiety Non-acidic pharmacophores Peptide backbone (P2-P4) and its simple modifications

14 15 15 15 17 17 18 24 24 30 32

2

CASPASE INHIBITORS

Peptidomimetic modification o / the P2-P4 amino ucids - acyclic strirctiires Peptidomimetic replucement u / the P3 amino acid Peptidomimetic replacemenl of the P2 amino acid Bicyclic peptidomimetic replacements of P2-P3

35 31 39 40

CASPASE INHIBITORS I N VlVO Axotomy/axonal lesions Brain ischaemia Other ischaemia models Brain trauma Excitotoxicity (kainic acid, NMDA, AMPA) Endotoxic shock Amyotrophic lateral sclerosis Multiple sclerosis Parkinson’s disease Huntington’s disease Graft rejection CNS inflammation Peripheral inflammation Hepatitis Shigellosis Meningitis Pneumopathy

43 43 44 45 46 46 41 41 41 48 48 48 48 48 49 49 49 50

CLINICAL STUDIES

50

SUMMARY

51

ACKNOWLEDGEMENT

51

REFERENCES

51

INTRODUCTION Among the enzymes responsible for cleaving a peptide bond (peptidases) exists a group whose proteolytic activity is due to the presence of a cysteine thiol functionality at the active site [I]. Some of them, based on shared sequence homology and a preference for the aspartate residue in the P1 position of their substrates, have been arranged into a family called ‘caspases’ [2]. There are fourteen caspases known at present: caspase- 1 (ICE) [3,4], caspase2 (Nedd2, Ich-1) [5], caspase-3 (CPP32, Yama, apopain) [B], caspase-4 (TX, Ich-2, ICEJI) [7-91, caspase-5 (ICErcl-III, TY) [9], caspase-6 (Mch2) [ 101, caspase-7 (Mch3, ICE-LAP3, CMH-1) [ 11-13], caspase-8 (FLICE, MACH, Mch5) [14-181, caspase-9 (ICE-LAPB, Mchb) [ 19,201, caspase-10 (Mch4) [12, 211, caspase-11 (Ich-3) [22,23], caspase-12 [24], caspase-13 (ERICE) [25], and caspase-14 (MICE) [2628]. The human caspase family members form 3 groups

P.P. GRACZYK

3

based on substrate specificity, structural similarities and function [2,24]. Group I (caspase-1 subfamily) consists of caspases-I, -4, -5, and -13. They function primarily as inflammatory mediators with caspase- 1 being the most extensively studied both in terms of biology [29] and design of its inhibitors [3, 4, 301. The remaining caspases, group I1 (caspases-2, -3, and -7) and group 111 (caspases-6, -8, -9, and -lo), play an important role in apoptosis. Apoptosis is a physiological cell suicide programme which occurs naturally during normal animal development [31, 321. It regulates cell number and morphogenesis. In order to maintain healthy tissues, abnormal cells can also be removed by apoptotic mechanisms. Recently, however, contribution of this type of death to various diseases has become an emerging concept. Within the nervous system, the apoptotic component in slow-onset neurodegenerative diseases such as Alzheimer’s (AD) [33, 341, Huntington’s (HD) [35] and Parkinson’s (PD) [36, 371 disease and amyotrophic lateral sclerosis (ALS) [38] has recently been appreciated [3%43]. An increasing abundance of data suggests that apoptosis may also contribute to acute-onset neurodegeneration which occurs after stroke [40, 441, spinal cord [45-471 and traumatic brain injury [48, 491, and acute bacterial meningitis [50]. Although the rapid neuronal death seen following such insults can be attributed to excitotoxicity [5 I], the apoptotic mechanisms responsible for delayed cell death are supported by characteristic DNA fragmentation in the ischaemic rodent brain [52, 531, studies on the role of apoptosis-regulatory genes in ischaemia [54-571 as well as other histological and biochemical data [58, 591. One may add that certain features of apoptosis may be favoured in the developing versus the adult brain [60]. Apoptotic death may also occur in the course of autoimmune diseases [61] and leukemia [62]. After myocardial ischaemia and reperfusion, cells die by apoptosis, probably due to oxidative stress in the reperfused myocardium [63, 641. Endotoxin-induced liver failure is accompanied by apoptosis of parenchymal cells [65]. The number of apoptotic cells is also increased in renal amyloidosis [66], acute renal failure [67], and HIV-induced nephropathy [68]. Apoptotic mechanisms may contribute to the hair follicle regression [69] and especially hair loss induced by anticancer drugs [70, 711. The exact pathway of cell death, and in particular the relative contribution of apoptosis in the above disorders, still remains a matter of controversy [39, 48, 72-75]. Although alternative, non apoptotic forms of cell death have also been considered [76, 771, the apoptotic machinery has been an attractive target for the design of cell death inhibitors with neuroprotective drugs potentially fulfilling the areas of highest unmet medical needs [7&84]. Since one of the most important hallmarks of apoptosis is thought to be activation of the caspase family of proteases [24, 85481, the search for antiapoptotic caspase inhibitors has become an important area for medicinal chemistry [8%92] and

4

CASPASE INHIBITORS

neuroscience in particular [75, 78, 93-95]. Moreover, new areas of application may emerge in the future as the involvement of caspase-like proteases in apoptotic cell death pathways in plants is also gaining support [96, 971.

ROLE OF CASPASES Studies using knockouts of caspases in mice by homologous recombination have shown that these enzymes play essential roles in development, immune regulation and apoptosis [98]. The role of individual enzymes is discussed below and summarised in Table 1.1 CASPASE-I

The primary role of the prototype enzyme, caspase- 1 (interleukin- 1/?-converting enzyme, ICE) is the control of key steps in inflammatory response and immunity, by activation of the proinflammatory cytokines interleukin- 1p (IL1s) and interleukin-1 8 (IL- 18, formerly interferon-y-inducing factor) [99]. IL- 18 and IL- 1/? are synthesized as inactive precursors which require cleavage into an active molecule. Caspase-1 cleaves the Asp1 16-Alal17 bond in the intracellular inactive precursor 3 1 kDa form of pro-interleukin- 1/? (pro-IL1/?) and releases a 17 kDa active IL-I/? fragment. Mature IL-I/? is then secreted into the extracellular space. One has to note, however, that alternative

Table 1 .I Caspase

SUMMARY DATA FOR CASPASES

Alternutive name

Diseuse

1

ICE

2

Nedd2, Ich-l

3

CPP32, Yarna, apopain

4

TX, Ich-2. ICEreiII ICEJII, TY Mch2 Mch3, ICE-LAP3, CMH-1 FLICE, MACH, Mch5 ICE-LAP6, Mch6 Mch4 Ich-3

Brain trauma, MS, HD, PD, liver injury, AD, ALS, OA, IBD AD, HIV-I infection, Salrnonellosis TBI, heart failure, hepatitis, AD, OA, PD PKD

~

5

6 I 8 9 10

I1 12

13 14

ERICE MICE

~~

HD Hepatitis Hepatitis, HD ALPS MS, ischaemia

P.P. GRACZYK

5

pro-IL-1P processing pathways exist. Extracellular pro-IL- 1 can also be cleaved by cathepsin G, chymotrypsin, elastase, a mast cell chymase, matrix metalloproteinases and granzyme A [ 100, 10I]. Caspase-1 -independent release of IL-IP - induced by Fas ligand has also been described [102]. Some of these alternative pathways can play an important role in vivo since the inflammation induced by turpentine injection in caspase- 1 knockout mice resulted in IL- 1/) expression indistinguishable from that observed in wild type animals [ 1031. As mentioned above, caspase- I is the primary enzyme responsible for cleavage of pro-IL-18 to generate IL-18. This cytokine may contribute to inflammation, gene expression and the synthesis of tumour necrosis factor (TNF), IL-1 and chemokines [104, 1051. However, it has been demonstrated that proteinase-3 is an alternative IL-I 8 -processing enzyme [ 1011. Caspase- 1independent secretion of IL- 18 from macrophages has also been described [ 1061. Despite similarity between IL-18 and IL-1B both in terms of primary amino acid sequence and folding pattern, their functions differ [99, 1011. IL-IP is a key cytokine contributing to inflammation [ 1071 while IL-18 plays an important role in the T-helper cell type 1 (Thl) response, by inducing interferon-y (IFN-y) production in T cells and natural killer (NK) cells. The role of IL-18 modulation in tumours, infections, and autoimmune and inflammatory diseases has recently been reviewed [99, 108, 1091. Although a proinflammatory role of IL-18 in rheumatoid arthritis (RA) [I 101, lung injury or inflammatory bowel disease (IBD) [99] has been suggested, in the opinion of Dayer [ 1 1 I ] blocking IL- 18 production should be approached with caution. Caspase- 1-deficient mice are resistant to lipopolisaccharide (LPS)-induced liver injury [ 1121. Nonetheless, IL-18 release in FasL-induced acute liver injury in Propionibacterium acnes-primed caspase- 1-deficient mice may occur in a caspase- 1-independent manner [ 1 121. Since the secretion of IL- 18 may still be inhibited by caspase inhibitors it has been suggested that some caspases other than caspase-1 may be involved in the processing of IL-18 in FasL-stimulated macrophages [106]. Apart from roles in inflammation and immunoregulation caspase-1 may also be important in the control of apoptosis and neurodegeneration. In 1993, Yuan et al. [ 1 131 recognized a homology between caspase- 1 and Caenorhabditis elegans gene, ced-3, which is necessary for programmed cell death (apoptosis) in the nematode. This finding raised the question as to whether caspase-1 is important for apoptosis in mammals. Initial studies in caspase- 1-deficient mice [ 1 141 and other models [ 1 151 indicated that caspase-1 does not play a major role in apoptotic events. However, subsequent experiments with transgenic mice expressing a dominant-negative mutant of caspase-1 [116, 1171, and with mice deficient in the caspase- 1 gene [ 1 181 have shown that the absence of this enzyme prevents neuronal cell death induced by trophic support withdrawal,

6

CASPASE INHIBITORS

and reduces ischaemic brain injury [ 1 191. Furthermore, attenuated brain damage occurs after moderate hypoxia-ischaemia in caspase- 1 knockout mice [ 1201. While the role of caspase- 1 in neuronal injury due to oxygen deprivation in the developing brain has also received significant support [121], it is not clear, however, whether the mechanism of protection is based on reduced apoptosis and/or inflammation. These results indicate that caspase- 1 activity may play an important role in stroke. Indeed, caspase-1-deficient mice exhibit reduced incidence and severity of experimental autoimmune encephalomyelitis (EAE) [122], suggesting that the protease may be involved in central nervous system (CNS) inflammation and in particular in multiple sclerosis (MS), a chronic inflammatory demyelinating disease of the CNS [ 1231. Caspase-1 activation has been demonstrated in the brains of a transgenic mouse model of HD as well as HD human patients [124]. Elimination of caspase-l activity in the mouse model by expression of a dominant-negative caspase-1 mutant delayed the disease progression and mortality [ 1241. Moreover, these transgenic mice exhibit decreased injury following brain trauma [ 1251. Since the expression of the dominant-negative caspase-1 mutant also results in an increased resistance to MPTP-induced neurotoxicity, caspase inhibitors have been suggested for the treatment of Parkinson’s disease (PD) [ 1261. Thus, the pretreatment of ventral mesencephalon cultures with tetrapeptide inhibitors of the caspase-3-like proteases, Z-VAD-fmk or Ac-DEVD-H, specifically inhibited death of dopamine neurones induced by low concentrations of 1 -methyl-4-phenylpyridinium (MPP ) [ 1271, whereas the caspase- 1 inhibitor, Ac-WAD-H, was without effect [ 1271. Caspase-1 appears to promote osteoarthritis (OA). Its presence was confirmed in human articular cartilage, and an increased cellular level has been found in OA tissue [128]. There is also evidence that caspase-1 may be involved in IBD [ 1291. Macrophage apoptosis plays a pivotal role in the course of bacillary dysentery caused by the enterobacteria Shigellu. One prerequisite gene for cytotoxicity is IpaB which activates caspase-1. The released IL-IP initiates a strong inflammatory response and contributes to the pathogenesis of shigellosis, which is one of the major causes of infant mortality in developing countries [ 1301. Contribution of caspase-1 and IL-lB to the pathology of AD is supported by an elevated caspase-1 activity found in post-mortem brains of AD patients [ 1311. Caspase-l activity is also elevated by an average of 81.5% in the spinal cord of humans with ALS [ 1321. +

CASPASE-2

Studies in caspase-2 deficient mice [ 1331 suggest that it acts both as a positive and negative cell death effector. In particular, the role of caspase-2 depends on

P.P.GRACZYK

7

the type of its isoform [134]. Since neurones from caspase-2 null mice are totally resistant to amyloid beta (AD, -42) toxicity, caspase-2 may be important for the pathology of AD [ 1351. It is required for apoptotic death of PC 12 cells and sympathetic neurones following withdrawal of trophic support [ 136, 1371, and oxyhaemoglobin (0xyHb)-induced apoptosis of brain microvessel endothelial cells [ 1381. It is also activated during apoptosis of human CD4 + T-lymphocytes after infection with HIV-I [ 1391, apoptosis of macrophages induced by Salmonella [ 1401, and sepsis-induced apoptosis of thymocytes [141]. Apart from its nuclear distribution, caspase-2 is localised to the Golgi complex, where it is involved in transducing proapoptotic signals [ 1421. CASPASE-3

Although thymocytes of caspase-3 deficient mouse are sensitive to apoptotic stimuli, programmed cell death in the brain is decreased resulting in abnormal development [ 1431. The importance of caspase-3 for neuronal death [6] has been further corroborated by experiments on the traumatic brain injury (TBI) in the rat [ 1441 and a recent finding of caspase-3 upregulation in human brain after traumatic head injury [ 1451. Cleavage of huntingtin by caspase-3 may be a crucial step in aggregate formation and neurotoxicity in HD [146, 1471. It is also the predominant caspase involved in amyloid precursor protein (APP) processing which results in elevated AP peptide formation [148]. This peptide, in turn, may induce PC 12 cell death mediated by caspases [ 1491. Furthermore, amyloidogenic Afi~5-3~as well as prion protein PrPloG126 can dose-dependently activate caspase-3 in rat cortical neurones [ 1501. Rather unexpectedly, however, in brains of patients with AD, the protein level of caspase-3 (and caspases-8 and -9) is decreased [ 1511. Caspase-3 may contribute to apoptotic death following cytokine deprivation in hematopoietic cells [ 1521, apoptosis of thymocytes [153], cardiac myocytes [ 1541, neurones and astrocytes [ 155, 1561, and tumour cell lines in response to various apoptotic stimuli [ 157-1591. Caspase-3 seems to be involved in apoptotic cell death in ischaemic/reperhsed rat heart as the enzyme levels were substantially elevated after regional myocardial ischaemia in ischaemic/reperfused rat heart [63]. Recent evidence suggests that caspase-3 activation occurs also in human heart failure [160]. The enzymes belonging to the caspase-3 subfamily (group II), rather than the caspase- 1 subfamily (group I), play the dominant role in Fas antibody-induced hepatitis [ 16 I]. Activation of caspase-3-like enzyme was also observed during rat liver regeneration following partial hepatectomy [ 1621. Interestingly, however, ischaemic preconditioning of mouse liver offered dramatic protection against prolonged ischaemia through down-regulation of caspase-3 activity [ 1631.

8

CASPASE INHIBITORS

Activation of caspase-3 in dopaminergic neurones in PD precedes and is not a consequence of apoptotic cell death [ 1641. Caspase-3 levels are also increased following optic nerve transection [165]. This enzyme may also play an important role in osteoarthritis since a positive correlation between the levels of chondrocyte apoptosis and levels of caspase-3 was found in the relevant dog model [ 1661. CASPASES-4 AND -5

The Fas-mediated apoptotic pathway seems to be mediated by a caspase-4-like protease [ 1671, although caspase-4, together with caspases-1 and -5, acts primarily as a mediator of inflammation [85]. Greater than seven-fold upregulation of caspase-4 has been observed in cystic kidneys in a mice model of polycystic kidney disease (PKD) [ 1681. CASPASES-6 AND -7

Caspase-6 may be involved in the cleavage of huntingtin thus contributing to aggregate formation and neuronal apoptosis in HD [146]. It has been found that both caspase-3 and caspase-6 are the two major active caspases present in apoptotic cells [ 1571. Caspase-7 seems to be a potential mediator of lovastatin-induced apoptosis in the prostate cancer cell line LNCaP [169]. CASPASE-8

Caspases-8 [ 17, 1701, -9 [ 171, 1721, and possibly - 10 [24] are critical upstream activators of the caspase cascade in some forms of apoptosis. Fas-mediated apoptosis in mice hepatocytes seems to be dependent on activation of caspase-8 [ 1731 and executor caspases-3, and -7 [ 174, 1751. The active form of caspase-8 was found in neurones as early as 6 h after focal stroke induced in rats by permanent middle cerebral artery occlusion (MCAO) [ 1761. Activated caspase8 is also present in the insoluble fraction of affected brain regions from HD patients [177]. An essential role of caspase-8 in HD is supported by the requirement for caspase-8 for the death of primary rat neurones induced by an expanded polyglutamine repeat (479) [ 1771. In addition, insufficient expression of functional caspase-8 was linked to childhood neuroblastomas [ 1781, although this has been questioned [179]. CASPASE-9

Caspase-9, an important activator of caspase-3, plays a critical role in apoptosis induction in axotomized retinal ganglion cells (RGCs) in vivo and is regulated

P.P. GRACZYK

9

following treatment with growth and survival factors [ 1801. It is also released from the intermembrane space (IMS) of mitochondria in an animal model of stroke [ 18 I]. Activation of caspase-9 was also observed in serum-/glucosedeprived cardiac myocytes [ 1541. CASPASES-I0 AND - I 1

A mutation in caspase-10 leading to defective lymphocyte and dendritic cell apoptosis is considered to be the origin of type I1 Autoimmune Lymphoproliferative Syndrome (ALPS) [182]. Caspase- 1 1 is involved in the activation of caspase- 1 in TNF-induced oligodendrocyte cell death [I831 and may play a crucial role in autoimmunemediated demyelination [ 1841. Furthermore, caspase-1 1 may be the initiator caspase responsible for the activation of caspase-3 as well as caspase-1 under certain pathological conditions, e.g., ischaemic insult [ 1851. CASPASES-12, -13, AND -14

Caspase-12 is involved in apoptosis due to stress in the endoplasmic reticulum [ 186, 1871. Overexpression of caspase- 13 induces apoptosis of 293 human embryonic kidney cells and MCF7 breast carcinoma cells [25]. Caspase- 14 is activated during differentiation of keratinocytes and seems to be involved in the formation of human skin [ 188, 1891. REGULATION OF CASPASES Although caspase-independent pathways in the inflammatory response and cell death in some systems have been found [76, 103, 19&195], the important contribution of caspases to neurodegeneration and inflammation has stimulated dynamic research to delineate the mechanisms involved in the regulation of caspase activity [49, 85, 88, 89, 119, 170, 1711. Currently, it is assumed that caspase regulation may occur via transcription, post translational modification, activation, and inhibition. TRANSCRIPTIONAL REGULATION

During the acute stage of EAE, an animal model of MS, both caspase- 1 and IL18 mRNA levels increase [ 1961. Recent studies on the role of caspase-1 in EAE confirmed this upregulation [ 1221. Furthermore, caspase-1 mRNA levels are significantly increased in peripheral blood mononuclear cells (PBMCs) from patients with MS compared with healthy controls ( p < 0.001) [ 1231. However,

10

CASPASE INHIBITORS

an ischaemic insult did not upregulate caspase-1 in the rat [197] and gerbil [ 1981 brain. Stimulation of U937 and HL60 leukemic cells and HT29 colon carcinoma cells with etoposide results in upregulation of caspase-2 and -3 genes and enhances the synthesis of pro-caspases [ 1991. In some of the animal models of ischaemia [ 197, 198, 200,2011 the levels of caspase-2 mRNA are also elevated (2-fold [198] and 3.8-fold [197]). A lateral fluid percussion brain injury in rats (a model of TBI) is similarly accompanied by an increase in caspase-2 mRNA level 12 h after the injury [202]. After permanent MCAO in rat, caspase-3 mRNA level increases 5.8-fold within 24 h [ 1971. An analogous increase both at the mRNA and protein level occurs in rat hippocampal CA1 pyramidal neurones 8-72 h after transient global ischaemia [201, 2031. In contrast, there were no significant changes in the expression of pro-caspase-3 at the protein level over 24 h after MCAO in mice although immunoreactivity due to pro-caspase-3 increased in the penumbra [204]. During transient focal ischaemia in the rat, an elevation in caspase-3 at the mRNA level was observed only in the resistant dorsomedial cortex at 1 day, and has been attributed to reactive changes in resistant brain areas [205]. A recent report suggests that in gerbil brains subjected to two 10min episodes of unilateral common carotid artery occlusion separated by 5 h, caspase-3 mRNA levels increase with a timecourse dependent on the brain area [206]. In another study of ischaemia in Mongolian gerbil caspase-3 and -4 mRNA levels remained constant over 2 days [ 1981. Upregulation of caspase-3 mRNA was also observed after ischaemiareperfusion injury in rat kidney [200], in motor neurones of both neonatal and adult rats after facial motor neurone axotomy [207], and in the rat spinal cord injury model [208]. Nevertheless, the mechanism by which the caspase-3 gene is induced is still not clear [203]. It has been shown that caspase-1 and caspase-3 mRNA upregulation can be inhibited by minocycline [209]. Using a recently reported polymerase chain reaction system, Lin et al. [2 101 have demonstrated IFN-y-induced transcriptional upregulation of caspase-5 in HT-29 colon carcinoma cells in the absence of upregulation at the protein level. In contrast, both caspase-5 mRNA and caspase-5 protein were induced by LPS in the monocytic cell line THP- 1 [2 lo]. Inhibition of endogenous nitric oxide (NO) synthesis in human melanoma cells by aminoguanidine, a specific, inducible nitric oxide synthetase (iNOS) inhibitor, led to upregulation of caspase- 1, caspase-3, and caspase-6 mRNA and eventually to cell death by apoptosis [21 I]. Pro-caspase-8 is constitutively expressed within spinal cord neurones. However, as early as 1.5 h after transient spinal cord ischaemia, an increase in

P.P. GRACZYK

11

caspase-8 (p 18) and caspase-8 mRNA levels could be observed within neurones in intermediate grey matter and in the medial ventral horn, and was followed by caspase-3 activation [212].

POST TRANSLATIONAL MODIFICATION

Caspases can be directly regulated by protein phosphorylation. In particular, Martins et al. suggested that phosphorylation of an active caspase could have an inhibitory effect on the enzyme activity [213]. Indeed, AKT can phosphorylate pro-caspase-9 which leads to inhibition of its proteolytic activity [2 141. Moreover, cytochrome c-induced proteolytic processing of pro-caspase-9 is defective in cytosolic extracts from cells expressing either active Ras or AKT [214]. These results are in contrast with other studies [215] showing that AKT cannot prevent apoptosis induced by microinjection of cytochrome c. ACTIVATION

Caspase proenzymes are normally present in cells [24, 2161 and can be rapidly activated in response to various stresses including DNA damage, UV radiation, heat shock, oxidative stress, or extracellular stimuli which triggers the caspase cascade [24, 85, 170, 2 17, 2 181. Activation pathways for proinflammatory caspases (group I) have not been studied as much as those for caspases involved in apoptosis. A RIP-like kinase, termed CARDIAK, promotes caspase- 1 activation in vitro by interacting with the caspase-1 recruiting domain (CARD) [2 191. Caspase- 1 might also be activated by caspase- I 1 [23]. Studies in vitro suggest that caspase-1 (and -1 1 ) can be activated by caspase-8 [220]. There is some evidence that, upon receipt of a proinflammatory stimulus, an upstream adaptor, RIP2, binds and oligomerizes caspase- 1 zymogen, promoting its autoactivation [221]. However, in the opinion of Zeuner et a/. [220], it is unlikely that self-processing leading to activation represents the pathway occuring in vivo. The association of caspase-1 with RIP2 can be prevented by a recently discovered protein ICEBERG [22 11. Apoptosis involves initiator (group 111) and executioner (group 11) caspases, and is currently interpreted in terms of two basic pathways: the first via death receptors (‘extrinsic pathway’) and the second involving mitochondria (‘intrinsic pathway’) [ 1701 (Figure 1.1). Caspase-8, and possibly caspase-10 are the apical caspases of a cascade triggered by the death receptors such as TNF-Rl (TNFa receptor l), CD95 (Fas/Apo- 1 ) [223], DR3 (APO-3/TRAMP/Wsl-l /LARD), DR4 (TRAIL-Rl) and DR5 (TRAIL-R2) [224]. For instance, upon assembly of the CD95

12

CASPASE INHIBITORS

Figure I . 1. SimpliJied diagram of extrinsic and intrinsic apoptotic pathways.

(Fas/Apo- 1) receptor death-inducing signalling complex (DISC), pro-caspase8 undergoes recruitment by Fas-associated death domain (FADD), an adapter protein, and oligomerization. According to the induced proximity model [225] this leads to an increase in the local concentration of the proenzyme and concomitant self-activation [226, 2271. In an analogous way, after stimulation

P.P.GRACZYK

13

of TNF-RI by TNFa, the receptor recruits an adapter protein TRADD (TNFR1-associated death domain) which then interacts with FADD and pro-caspase8, leading to its activation. Interestingly, activation of caspase-8 induced by various drugs can be mediated by a CD95/Fas-independent mechanism and preceded by activation of (an)other caspase(s) [22&230]. Active caspase-8 activates caspases-3, -7 [231, 2321, and -13 [25] and probably several more [220] via cleavage of the relevant pro-caspases (stepwise for pro-caspase-3 [233]). There is some confusion as to which downstream/executioner caspases are capable of activation of other caspases. Caspase-3 has been shown to subsequently activate caspase-6 [23 1, 2321. However, during apoptosis of cerebellar granule cells (CGCs) induced by withdrawal of trophic support, caspase-6 activated caspase-3 in cellular extracts from non-apoptotic CGCs, whereas caspase-3 failed to activate caspase-6 [234]. Experiments in vitro show, however, that caspase-3 and -6 cleave each other’s precursors quite well [220]. Similarly, caspases-6 and -7 can activate each other [220]. It has also been reported [I671 that death signal transmission from caspase-8 to caspase-3(-like) proteases may involve a caspase-4(-like) protease. Scaffidi et al. have suggested that there are two types of cells in which two different CD95 (Fas/APO-I) signalling pathways exist [235]. In type I cells activation of caspase-8 occurs rapidly via DISC. In type I1 cells activation of caspases-8 and -3 is delayed and occurs downstream of mitochondria. In these cells, DISC formation is strongly reduced [235]. Signalling through death receptors can be inhibited by viral [236] or cellular [237, 2381 proteins known as FLIP, which contain a fragment similar to that present in caspase-8, and therefore are able to compete with pro-caspase-8 for binding with FADD. A similar mechanism has been considered for ARC, an inhibitor of apoptosis expressed in skeletal muscle and heart [239]. Activity of caspase-3 might also be inhibited indirectly by hepatitis B virus-encoded HBx protein [240]. Caspase-9 is the most upstream protease in the mitochondria1 (intrinsic) pathway in which caspase activation is triggered by cytochrome c and dATP [24 1-2431. During apoptosis cytochrome c is released from the mitochondria into the cytosol where it forms a complex involving Apaf-1 (a CED-4 homologue), then dATP [244] and pro-caspase-9 [245]. Once activated in an autocatalytic manner, caspase-9 then activates caspases-3 and -7 [243, 2461. Recent studies, however, indicate that autoproteolysis of pro-caspase-9 is not required, as the expression ofthe zymogen activity is dependent on cytosolic factors [247]. Caspase-3 may subsequently activate caspases-2 and -6 as shown in Figure 1.1 [243]. Caspase-6 may then activate caspases-8 and -10. In addition, the caspase-9 pathway might be modulated by proteins from the Bcl-2/Ced-9 family [248, 2491 and recently discovered Aven [250]. Moreover, activation of

14

CASPASE INHIBITORS

caspase-9 by Apaf-1 can be blocked by an endogenous, alternatively spliced isoform of caspase-9, named caspase-9b, which functions as an endogenous inhibitor of apoptosis [25 11. Importantly, the extrinsic and intrinsic pathways may interact with each other. In particular the mitochondria1 pathway may amplify the receptormediated apoptotic signal [252]. Caspase-8 can cleave the proapoptotic protein BID, to liberate truncated BID (tBID) which translocates to mitochondria and triggers cytochrome c release [253, 2541. Furthermore, caspase-3-catalyzed cleavage of antiapoptotic Bcl-2 produces a carboxyl-terminal Bcl-2 cleavage product able to induce cell death and further activation of downstream caspases [255]. Activated caspases are capable of cleaving many targets, e.g. poly(ADP-ribose) polymerase (PAM) [256, 2571, lamins [258], cell cycle proteins [259] and many other nuclear [260] and cytoplasmic targets [89] including proteins involved in transduction of survival signals [261]. This proteolytic activity eventually contributes to the characteristic morphological changes such as chromatin condensation and DNA fragmentation. It should be noted, however, that the role of caspases -6 and -7 as important downstream effectors has recently been questioned [262]. The actual apoptotic pathway is most probably much more complex [ 157, 2631. Recent results from this laboratory [264,265] and others [32, 73, 76, 266, 2671 indicate that the mode of cell death and activation of specific caspases depend on the cell type and the death stimulus involved. Additional activation of other non-caspase-like proteases [268], including the 24-kDa apoptotic serine protease (AP24), may also be necessary [269, 2701. Furthermore, Nakagawa et al. have identified an additional, caspase- 12-mediated, apoptotic mechanism which is initiated by stress in the endoplasmic reticulum [ I861 and in which caspase-12 may be activated by calpain [271]. However, inhibition of caspase activity in a cell may not be sufficient to keep it alive [272, 2731. Zeuner et al. pointed out that activation of proapoptotic caspases may even occur in the absence of cell death indicating their possible non apoptotic role [222]. Interestingly, FasL, usually connected with death signalling, may also be involved in other pathways, e.g. inflammation [ 1021.

INHIBITION OF CASPASES Caspase inhibition can be achieved using compounds of both natural and synthetic origin. Some of these compounds are proteins which are derived either from viruses (p35, CrmA) or which may also be endogenously expressed in mammals (IAP).

P.P. GRACZYK

15

PROTEIN INHIBITORS

Due to the importance of caspases in many biological processes, several ways to control activity of these enzymes have developed during evolution. A strategy which is based on generating proteins capable of inhibiting the active caspases involves p35, IAP, and CrmA proteins. P35 Organisms fight against viral infections by killing the infected cells. To counteract this defence mechanism viruses evolved antiapoptotic genes such as p35 and IAP [274]. The product of p35 is a 35-kDa protein initially found to suppress virus-induced apoptosis in host cells of Spodopterufbugiperda [275]. Further studies have shown antiapoptotic activity of p35 in many other cell types including cardiomyocytes [276], cancer [277, 2781, neuronal [27%281], host pupal eye cells of the fly Drosophila melunoguster [282], and C. elegans cells normally programmed to die [283]. p35 is a broad range caspase-specific and active site-directed [284] inhibitor capable of interacting with caspases-I, -2, -3, -4 [285], -6, -7, -8, and -10 with Ki values from 0.1 to 9 nM [284]. Interestingly, the p35 protein does not inhibit caspase-9 activity in a cell-free system of mammalian caspase activation [286]. In the opinion of Vier et al. [286] p35 evolved specifically to inhibit effector rather than initiator caspases. Indeed, the presence of an initiator caspase that is resistant to p35 has been suggested [287]. Inhibition of enzymatic activity is accompanied by cleavage [285] of p35 at Asp-87 [284]. It has also been demonstrated that the insect Spodoptera frugiperdu target of the baculovirus p35 is Sf caspase-1 of which the sequence and specific activity is highly related to human caspase-7 and caspase-3 [288,289]. p35 serves as a substrate of Sf caspase-1, which cleaves p35 to similar sized fragments [289]. Recently, Fischer et ul. have determined the crystal structure of p35 and proposed a multistep mechanism for stoichiometric caspase inhibition involving a reactive site loop analogous to that of serpins [290]. IAP Inhibitor of apoptosis proteins (IAPs) have been found in baculoviruses, higher eukaryotes [291, 2921, and mammals [291,293]. The mammalian IAPs include X-IAP (MIHA, hILP), c-IAP1 (MIHB, inhibitor of apoptosis-1), c-IAP2 (MIHC), neuronal apoptosis inhibitory protein (NAIP), the recently discovered survivin [294], apollon [295], livin (IAP-3) [296, 2971, and ML-IAP [298]. Down-regulation of X-IAP following adenoviral antisense expression may induce apoptosis in some types of cancer cells [299].

16

CASPASE INHIBITORS

IAPs seem to have a much narrower spectrum of activity than p35 [300]. Viral IAP, when expressed in mammalian cells, can block caspase-1 and -2-induced apoptosis, but is ineffective against death due to caspase-7 expression [301] and caspase-induced apoptosis in insect cells [288, 3021. IAP can prevent caspase-mediated cleavage of p35 in vivo and inhibit caspase activity upon viral infection or UV irradiation [302]. This was rationalized by assuming that IAPs inhibit the activation of pro-caspases [288, 302, 3031. Thus, the antiapoptotic properties of NF-KB could be due to the upregulation of IAP genes [304, 3051, and subsequent inhibition of pro-caspase-8 activation [305]. However, it is possible that not all IAPs work by the same mechanism [292]. Caspase-inhibitory function of IAP may require homo-oligomerization [306]. Other studies have shown that the members of the human IAP family specifically bind to caspases-3 and -7 [307-3091 with inhibition constants for X-IAP of 0.7 and 0.2 nM, respectively [3 lo], thus suggesting that IAPs may act as direct caspase inhibitors. X-IAP can also inhibit caspase-9 (but not caspases- 1, -6 or -8) [296, 297, 308, 31 1, 3121. In contrast to p35 protein, no cleavage of IAPs occurs during inhibition of caspases, indicating a different mechanism of action [307]. However, during Fas-induced apoptosis, endogenous X-IAP can be cleaved into two fragments comprising baculoviral inhibitory repeat (BIR) 1 and 2 domains and the BIR3 and RING domains (BIR3-Ring) [313]. Recently, it has been found that only the second (BIR2) of the three BIR domains in X-IAP is necessary and sufficient for inhibiting caspases-3 and -7 with apparent inhibition constants (Ki) of 2-5 nM [314]. The fragment of X-IAP encompassing the third BIR domain (BIR3) and RING domain is responsible for specific inhibition of mammalian caspase-9 [313]. This finding agrees with the apoptosis-suppressing activity of a new Spodoptera frugiperda S'IAP in human cells, which occurs by inhibition of a caspase-9 homologue, and requires both BIR and RING domains [315]. The remaining IAPs, i.e. c-IAP1 and c-IAP2 are less potent inhibitors with K,'s against caspases-3 and -7 in the range from 29 to 108 nM [307]. NAIP does not inhibit any of these enzymes and presumably inhibits apoptosis via other targets [307]. Overexpression of survivin occurs in many common types of cancer [294]. Although survivin is capable of inhibiting caspases-3 and -7 (but not caspase8), and promotes survival of 293 cells exposed to diverse apoptotic stimuli [309], its function may also be related to cytokinesis [316]. Recently, its caspase-3 inhibitory activity has been questioned [3 171. Livin can bind caspases-3 and -7 via its BIR domain [297]. It might also interact with both unprocessed and cleaved forms of caspase-9 [297]. Interestingly, IAPs can be inactivated by binding to endogenous proteins like Smac [318]/Diablo [319], which are released into the cytosol when cells undergo apoptosis [32&322].

P.P. GRACZYK

17

Finally, failed caspase inhibition due to the presence of disfunctional NAIP gene has been linked with spinal muscular atrophy (SMA), a genetic disorder characterized by motor neurone loss in the spinal cord [323]. CrmA

CrmA (cytokine response modifier A) is a 38 kDa protein from cowpox virus which belongs to the serpin superfamily of serine proteinase inhibitors. For several years it has been known as a caspase-1 inhibitor [324, 3251. Its potency may be attributed to the sequence LVAD of the reactive site loop, which is close to YVAD recognized by caspase-1, -3 and -4 [326]. CrmA rapidly (1.7 x 1 0 7 ~ - 1s - ~ ) [327] inhibits caspase-1 forming a tight complex with an equilibrium constant for inhibition (K,) of about 0.004-0.010 nM [325, 327, 3281. It is also active against caspase-4 [329] ( K , 1.1 nM) [327], caspase-5 [329] ( K , < 0.1 nM) [327], caspase-8 [23 I] (K, < 0.34; 0.95 nM) [327, 3281, caspase-9 ( K , 10000 1960 > 10000 3090 21.1 508 330

0.76 362 I63 > 10000 > 10000 > 10000 > 10000 352 970 408

132 205 1710 0.23 1.6 31 0.92 60 12

18

Boc-IETD-H

(8)>>(9)>(7). Correlations between anti-PCP effects and binding to DNA or DNA-binding proteins have been seen [19], suggesting that requirements for RNA binding compared to DNA binding are quite different in this series.

RNA AS A DRUG TARGET

80

0

HNYNH

f-fN"

A series of compounds derived from tri-0-acetybglucal were described in the patent literature by workers from Allelix Biopharmaceuticals [20]. Shown in Table 2.1, the OMe compound with undefined stereochemistry at R' (1 1) and the a-OEt anomer (12) have the best activity with ICso values of 0.051 and 0.046 pM respectively. Extending this side-chain as either a polyether (13) or introducing an additional basic centre (14) results in a loss in activity. Introduction of additional hydroxyl fbnctionality at R2 and R3 to give (15) was detrimental to activity. Additionally, removal of the anomeric substituent completely abrogated activity by at least 1000-fold in the R3 substituted Table 2.1. ACTIVITY OF D-GLUCAL DERIVED GUANIDINE ANALOGUES IN A Tat/TAR ELECTROPHORETIC GEL-MOBILITY SHIFT ASSAY

Compound

R'

(11) (12) (13) (14) (15) (16) (17) (18)

H OMe H a-OEt u - O ( C H ~ ) ~ - O ( C H ~ ) ~ O EH~ u-O(CH~)~NHC(NH)NH~H a-OMe OH H H H H H H

R2

'R

Gel-shqt IC,, (pM)

H H H H OH OH OMe O(CH2)3NHC(NH)NHZ

0.05 1 0.046 0.5 0.17 1.35 55 185 110

M.J. DRYSDALE ET AL.

81

analogues ( 1 6 18). The loss of activity of the tris-guanidino compounds (14) and ( 1 8) compared to the bis-guanidines ( 1 1) and (12), shows that simply populating such structures with more basic centres does not necessarily lead to an increase in binding to RNA. High throughput screening A group at Parke-Davis have described a complete therapeutic programme targeting TAR RNA, where they utilize high throughput screening to identify inhibitors of the Tat/TAR complex [21]. As well as the previously described gel shift assay, they described two other assay types suitable for an HTS campaign towards this target. Firstly, a scintillation proximity assay (SPA) was outlined using '251-Tat12 and biotinylated TAR RNA. Secondly they described a high throughput filtration assay using 32P-labelledTAR RNA and unlabelled Tat peptide. In this assay, the protein adheres to the membrane material and so protein-bound RNA is also captured whilst the free RNA passes through the membrane. Thus in the presence of a competitor ligand the quantitation of counts on the membrane compared to controls give a value for ligand binding. The SPA and filtration assays were utilized in the parallel screening of a 150,000 compound collection, where the compounds were screened as mixtures with each component at 20 yM. A hit rate of 1-2% was claimed, and upon deconvolution, one-third of these had IC5,, values < 50 pM. In a later publication [22] the same authors identified aminoglycosides (exemplified by neomycin (19)), quinoxaline2,3-diones (e.g. (20)) and 2,4-diaminoquinazolines (e.g. (21)) as three classes of hit structures from their HTS campaign. values as measured in the gelshift assay were quoted at 0.92yM, 1.3pM and lOyM for (19), (20) and (21) respectively.

In order to determine whether these compounds bound directly to the RNA component in these assays, electrospray ionisation mass spectrometry

82

RNA AS A DRUG TARGET

(ESI-MS) was used. Molecular weights corresponding to TAR alone and the 1: 1 complexes of TAR and ligands (19), (20) and (21) were observed by ESIMS. Under similar conditions, no complexes with a Tat4o peptide were found for any of these molecules. Further studies using chemical and enzymatic footprinting studies elucidated the region of TAR targeted by each inhibitor. Compound (2 1) was further characterized for cellular activity. HeLa cells constitutively expressing Tat, were transfected by a plasmid (pHIVlacZ) containing a promoter domain from the HIV-1 3'-LTR and a lacZ reporter gene. The HIV-1 LTR promoter containing TAR RNA can be bound and activated by Tat. The lac2 gene is controlled by the LTR promoter and the products of expression, p-galactosidases, can be quantified. In this assay compound (21) has a dose response effect with an IC5,, measured at 19 pM. No apparent cellular toxicity was observed at concentrations below 100pM. In a related cell line, the HIV-1 promoter is replaced by the cytomeglovirus (CMV) promoter to measure TAR independent gene expression. Compound (21) showed no effects on the extent of gene expression driven by the CMV promoter in contrast to that observed with LTR-activated lacZ expression. HIV-1 replication studies were carried out in both OM-10.1 and Ul cell lines. In OM-10.1 cells, (21) has an EC=,o=4pM and a toxic concentration (TC)50 value of 54pM. In the U1 cell line, an ECso of 16pM was measured.

The group at ISIS pharmaceuticals have utilized a related high throughput SPA assay to that described above to identify novel piperazinyl polyazacyclophane scaffolds [23] and polyazadipyridinocyclophanes [24] inhibitors of the Tat/TAR interaction. This group used a solution phase combinatorial approach. In this strategy, the secondary amines in (22) and (24) (R' -R4 = H), are treated with a mixture of alkylating agents to give a mixture of 625 compounds in each case. For the piperazinyl polyazacyclophane analogues, the library depicted by (23) gave an IC50= 0.08 pM, and in the case of the polyazadipyridinocyclophanes (25) and (26), IC50 values of 0.08 pM and 0.09 pM respectively were measured. In each case, related libraries with less polar or

M.J. DRYSDALE ET AL.

83

fewer basic functionalities were less effective at disrupting the Tat/TAR interaction. NH (24) R1-R*=H

0

NC

II

NH

Others

Hoechst 33258 (27), a bisbenzimidazole derivative, was shown by a variety of techniques to bind to TAR [25]. Using a combination of UV absorption studies, thermal denaturation, circular dichroism (CD), electrical linear dichroism (ELD) and RNase A footprinting, this ligand was shown to be a tight binder of TAR with specific contacts in the bulge region of the RNA.

( 2 7 ) Hoechst 33258 Rev/RRE

RRE RNA is a 234-nucleotide RNA sequence located within the env gene of HIV and is regulated by binding to Rev protein [26, 271. Biochemical studies have identified a high affinity Rev binding site in the stem-loop region of stem IIB of RRE [28, 291. Model constructs based on this region of RNA, and the Arg-rich binding domain of the Rev protein (Rev34-50and related peptides) have been used in a number of studies to look at small molecule interactions in this system.

RNA AS A DRUG TARGET

84

Aminoglycoside analogues Using methods such as surface plasmon resonance (SPR) [30], Neomycin B (28), has been shown to be the most effective inhibitor of Rev/RRE binding with an ICso in the region 0.1-1 pM. Closely related analogues such as paromomycin (2) have been shown to be significantly less active by this method. SPR is a useful technique that allows direct observation of ligand-RNA interactions and can be used to test for specificity. 5TRQARRNRRRRWRERQR

ooc 0 OH

H2N

(28) Neornycin B

(29) RhdRev

In a more recent study [31,, a tetramethylrhodamine labellei Rev34-50 peptide, RhdRev (29) was prepared which upon binding to RRE results in an increase in fluorescence anisotropy of RhdRev. Competition by an antagonist results in decreased fluorescence anisotropy and dissociation constants measured. Ten aminoglycosides were studied and their Kd values are shown in Table 2.2. Consistent with previous data, neomycin B is the most potent compound with a Kd = 4.4pM. All the other aminoglycosides had Kd values

Table 2.2. DISSOCIATION CONSTANTS OF AMINOGLYCOSIDES FOR RRE IIB RNA Aminoglycoside

Kd (PM)

Gentamycin Hygromycin B Kanamycin A Kanamycin B Neomycin B (28) Nearnine (30) Parornomycin (2) Ribostamycin Streptomycin Tobramycin

34.6 248 46.0 14.4 4.38 20.8 34.5 87.5 200 15.5

M.J. DRYSDALE ET AL.

85

Table 2.3. DISSOCIATION CONSTANTS OF AROMATIC RING SUBSTITUTED NEAMINE ANALOGUES FOR REZE IIB RNA

Inh ibiiors

Kd (@)!

20.8 141 40.6 3.6 0.240 0.753 0.277

>I0 pM. Of note is the value for neamine (30), Kd = 20.8 pM. Modification of this simplified scaffold by incorporation of groups capable of intercalating with the groove or stacking with bases (3 1-36) results in inhibitors with significantly improved activity (Table 2.3). The loss of activity of (31) and (32) compared to (30) suggests that the benzoyl and 2-naphthoyl groups may not stack well, and that substitution of the primary amine is detrimental to binding. However, introduction of the anthracene-9-carbonyl group (33) results in a compound with six-fold higher activity than the parent neamine (& = 3.6 pM vs. 20.8pM). Introduction of a pyrene unit had even more significant effects. Depending on linker length, Kd values of 0.240, 0.753 and 0.277 pM respectively were measured for (34-36) respectively. In a related communication [32], neomycin B was linked to 9-aminoacridine to give (37). In a gel-shift mobility assay IC50 values for (28) and (37) were measured at 5.9 and 0.65 pM. This communication also described a fluorescence anisotropy assay related to that described above, where a C-terminal fluorescein-labeled Rev34-50 analogue was displaced from RRE by inhibitors. In this assay paradigm, competition experiments gave IC50 values of 0.8 pM and 0.015 pM for (28) and (37) respectively. Additionally enzymatic footprinting studies in this system confirmed direct binding of (37) to RRE RNA.

Diarylamidine analogues A series of diphenylfuran bisamidines related to structures (7-10) were reported to inhibit Rev/= binding [33]. ICS0values were measured in a gelshift assay and compound (38) (Table 2.4), ICsOlOO-fold as shown with (40).

RNA AS A DRUG TARGET

86

(30) R = H

(34)R =

(33)R=

(

'

3

6

)

R / -/ 4

/

Table 2.4 DIPHENYLFURAN BISAMIDINES IN GEL-SHIFT Rev/RRE ASSAY

Compound

R

Gel-shift IC,, (pM) loo0

(@)!

> 1000 750 250 150

780 190 630 800 500

The HCV NS3 P1 specificity pocket differs from that of most serine proteases. Unlike thrombin for example, basic groups are deleterious to binding to NS304A allowing for good selectivity versus the clotting enzymes. The small shallow pocket is defined by the Leu-135, Phe-154, and Ala-157 side-chains [52, 531. Phe-154 is primarily responsible for the observed specificity by interacting with the cysteine thiol found in trans cleavage sites [133]. Table 5.12 presents a brief survey of small replacement groups for the P1 cysteine found in the natural substrate. Interestingly non-polar groups function at least as well as polar substituents. Ethyl (65), and N-propyl (66) derivatives demonstrate the best potencies in this, albeit, limited series. Surprisingly, despite the poor activity of (69), a compound possessing a gemdimethyl substitution, it was found that a cyclopropyl P1 moiety is advantageous [ 1191. This result has been exploited [ 118, 134-1361 with a non-covalent carboxylate warhead. The cyclopropyl PI (73) provides a 3-fold improvement in binding relative to a norvaline P1 group (72). An optimized hexapeptide inhibitor (74) with larger hydrophobic groups at P2 and P4, displays nanomolar potency.

R.B. PERNI AND A.D. KWONG

239

COOH

(72) Ki=150pM COOH

(73) Ki = 54 p FOOH

M

8

s

(74) Ki = 0.013 pM

The cyclopropyl P1 group has also been combined with a boronate warhead [128]. The boromate was only slighly more potent than the corresponding carboxylate. P2 VARIANTS AND SOLVENT SHIELDING

Recent NMR calculations on an a-ketoacid containing peptidomimetic inhibitor (53) show that a leucine side-chain at P2 has a particular stabilizing effect on the catalytic His-57 imidazole to Asp-81 carboxylate hydrogen bond [ 1 121. This stabilization appears to be effected by the shielding of that site from solvent exposure. This effect can be exploited by the use of large hydrophobic substituents at the P2 position and may be part of the reason for the good potency observed for the naphthylmethylether proline derivative (74). Cyclohexylalanine at P2 has been shown to be superior to smaller side-chains such as leucine [ 1211. As peptidomimetic inhibitors have evolved in recent years the cyclohexylalanine group (cha) has given way to 0-substituted 4-hydroxylproline as the P2 residue of choice. The 0-benzyl group is among the most studied substituent ( e g (50)) but compounds incorporating larger groups are displaying excellent inhibitory potencies. Compound (75) with a tricyclic proline

240

HEPATITIS C VIRUS AN OVERDUE LINE OF THERAPY

substituent possesses an ICs0 of less than 500 nM against the NS3-4A [118]. Truncation to tripetide derivatives was effective in this series maintaining significant potencies [ 1341.

P3, P4 SUBSITUENTS

Both P3 and P4 subsites benefit from the presence of hydrophobic substitution in regard to enzyme inhibition. The P4 SAR clearly shows the effects of hydrophobic substitution (Table 5.13) [ 12 11. The Dif (diphenylalanine) containing inhibitor (81) is the most potent but from a practical perspective, is very expensive. Cheaper alternatives are available for substitution at P4 but the trade-off is a reduction in potency. Although a Glu residue at P3 is close to optimal, non-charged side-chains are obviously preferred since charged groups hinder cellular penetration. Small hydrophobic amino acids such as valine or iso-leucine have been shown to be comparable to the Glu at P3. D-amino acids are unacceptable at both P3 and P4 [121].

Table 5.13. P4 OPTIMIZATION

Compound

Ac-Asp-Glu-X-Glu-Cha-Cys

~CSO(FW

X Val Nleu Cha Ileu Leu Dif

0.330 0.224 0.140 0.122 0.118 0.055

R.B. PERNI AND A.D. KWONG

24 1

P3 Indolinyl derivatives A particularly interesting class of inhibitors contains a novel P3 group, a 2-substituted indolyl moiety. These compounds incorporating an a-ketoacid warhead display surprising potency despite the presence, in some cases, of isomeric mixtures [72, 1371. The SAR is summarized in Table 5.14.

Cyclic derivatives

A series of novel macrocyclic peptidomimetics have been studied [ 1381. Compound (89) incorporates a 15-membered ring structure tying PI to P3. Cellular activity is claimed for this series of inhibitors in Huh-7 cells. These cells express Table 5.14. P3 INDOLINYL DERIVATIVES

Y-F

F

Compound

R

ICsu (PM)

Isomer Ratio

50

single

92

1.5:l:l:l

16

1:l:l

45

1:l:I:l

5

single

0.8

>10:1

H

gc?z H

H

I

242

HEPATITIS C VIRUS AN OVERDUE LINE OF THERAPY

a portion of the HCV polyprotein from NS3 through NSSA ending with the first six amino acids of NSSB and utilizes SEAP as the reporter construct [93].

Table 5.15. PI’-REGIONBASED INHIBITORS Compound

Peptide Sequence

Ac-Asp-Glu-Dif-Ile-Cha-Cys-Pro-Nle-Ser-Leu-NH2 Ac-Asp-Glu-Dif-Ile-Cha-Cys-Pro-Nle-Gln-Leu-NH2 Ac-Asp-Glu-Dif-Ile-Cha-Cys-Pro-Nle-Hyp-Leu-NH2 Ac-Asp-Glu-Dif-Ile-Cha-Cys-Pro-Nle-Asp-Leu-NH2 Ac-Asp-Glu-Dif-Ile-Cha-Cys-Pro-Cha-Ser-Leu-NH2 Ac-Asp-Glu-Dif-Ile-C ha-Cys-Pro-Cha-(D)-Trp-Leu-NH2 Ac-Asp-Glu-Dif-lle-C ha-Cys-Pro-Cha-Gln-Leu-NH2 Ac-Asp-GIu-Dif-Ile-Cha-Cys-Pro-Cha-Hyp-Leu-NH2 Ac-Asp-Glu-Dif-lle-C ha-Cys-Pro-ChaAsp-Leu-NH2 Ac-Asp-Glu-Dif-Ile-Cha-Cys-Pro-H~f-Gln-Leu-NH~ Ac-Asp-Glu-Di f-Ile-Cha-C y s-Pro-Hof-Hyp-Leu-NH2 Ac-Asp-Glu-Di f-Ile-Cha-Cy s-Pro-Hof-Asp-Leu-NH2 Ac-Asp-Glu-Dif-lle-Cha-Cys-Pro-Phg-Asp-Leu-NH2

Ic50

(nM)

64 32 26 1.8

23 820 14 11

1.3 18 15 1.8 7

Table 5.16. OPTIMIZED PI’-BASEDINHIBITORS Compound

Peptide Sequence Ac-Glu-A~p-Val-Val- Abu-Cys-Pro-Nle-Ser-NH~

I G n (nM)

8500 876 3100 64 Ac-Asp-Glu-Dif-Ile-Cha-Cys-Pro-Cha-Ser-Leu-NH2 23 Ac-Asp-Glu-Dif-Ile-Cha-Cys-Pro-Cha-Asp-Leu-NH2 1.3 Ac-Asp-D-Glu-Leu-Ile-Cha-Cys-Pro-Cha-Asp-Leu-NH2