No title

RESEARCH HIGHLIGHTS HIGHLIGHT ADVISORS URLs CEZMI AKDIS SWISS INSTITUTE OF ALLERGY AND ASTHMA RESEARCH, SWITZERLAND BR...

Author: Lucy Bird

8 downloads 318 Views 4MB 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

RESEARCH

HIGHLIGHTS HIGHLIGHT ADVISORS URLs CEZMI AKDIS SWISS INSTITUTE OF ALLERGY AND ASTHMA RESEARCH, SWITZERLAND BRUCE BEUTLER SCRIPPS RESEARCH INSTITUTE, USA PETER CRESSWELL YALE UNIVERSITY, USA JAMES DI SANTO PASTEUR INSTITUTE, FRANCE GARY KORETZKY UNIVERSITY OF PENNSYLVANIA, USA CHARLES MACKAY GARVAN INSTITUTE OF MEDICAL RESEARCH, AUSTRALIA CORNELIS J. M. MELIEF LEIDEN UNIVERSITY MEDICAL CENTRE, THE NETHERLANDS MICHEL NUSSENZWEIG THE ROCKEFELLER UNIVERSITY, USA RICHARD RANSOHOFF CLEVELAND CLINIC FOUNDATION, USA ALAN SHER NATIONAL INSTITUTE OF ALLERGY AND INFECTIOUS DISEASE, USA ANDREAS STRASSER THE WALTER AND ELIZA HALL INSTITUTE, AUSTRALIA MEGAN SYKES HARVARD MEDICAL SCHOOL, USA ERIC VIVIER CENTRE D’IMMUNOLOGIE DE MARSEILLE-LUMINY, FRANCE MATTHIAS VON HERRATH LA JOLLA INSTITUTE FOR ALLERGY AND IMMUNOLOGY, USA

900 | DECEMBER 2005

I N N AT E I M M U N I T Y

VAMPing up macrophage responses As crucial players in immediate early defence against infection, macrophages are armed with multiple mechanisms to alert the immune system to an intruder. Two such macrophage activities are phagocytosis of the intruding microorganism and rapid secretion of pro-inflammatory cytokines, such as tumour-necrosis factor (TNF). In a recent report by Jennifer Stow and colleagues, these two activities are shown to be linked by a joint trafficking pathway, which allows macrophages simultaneously to release TNF and to expand their plasma membrane for phagocytosis. To identify the proteins that are involved in the TNF-secretion pathway, the authors carried out DNAmicroarray analysis of macrophages activated with lipopolysaccharide. Vesicle-associated membrane protein 3 (VAMP3) was highly expressed by activated macrophages, and its expression correlated with TNF secretion. Consistent with having a role in the trafficking of TNF, VAMP3 could interact with SNARE proteins (which mediate intracellular membrane-fusion events) in the Golgi (such as syntaxin-6) and the plasma membrane (such as syntaxin-4). Moreover, overexpression of VAMP3 by macrophages increased TNF secretion, whereas knockdown of Vamp3 expression, using small interfering RNA, blocked the delivery of TNF to the cell surface, leading to its accumulation in the Golgi. Fluorescently labelled VAMP3 and TNF were both shown to be

localized in recycling endosomes, as indicated by colocalization with internalized transferrin and the recycling-endosome protein RAB11, pointing to an unexpected route for TNF exocytosis. Further analysis of this pathway indicated that two fusion events involving VAMP3 and its SNARE-protein partners occurred: the first between TNF carriers exiting the Golgi and recycling endosomes, and the second between recycling endosomes and the plasma membrane. Because SNARE proteins and VAMP3 have previously been shown to be required for phagocytosis of large microorganisms, such as yeast, the authors next investigated the possibility that the TNF-secretion pathway and phagocytosis converged. Imaging of activated macrophages that were incubated with the yeast Candida albicans indicated that VAMP3 and TNF were present in the phagocytic cup — the initial stage of phagocytosis — but TNF was not

| VOLUME 5

detected after internalization or in mature phagosomes. Addition of an inhibitor of TNF-converting enzyme (TACE) to the cells, to block proteolytic release of cell-surface TNF, led to the accumulation of TNF in the phagocytic cups. So, the delivery of TNF to the macrophage surface, mediated by a series of membrane-fusion events that involve VAMP3 and SNARE proteins, seems to be targeted to sites of phagocytic-cup formation. After it has reached the cell surface, TNF is rapidly cleaved by TACE and released, before closure of the phagocytic cup. This provides the macrophage with an elegant means to release TNF while simultaneously expanding the plasma membrane for formation of the phagosome. Lucy Bird References and links ORIGINAL RESEARCH PAPER Murray, R. Z.,

Kay, J. G., Sangermani, D. G. & Stow, J. L. A role for the phagosome in cytokine secretion. Science 10 Nov 2005 (doi:10.1126/science.1120225)

www.nature.com/reviews/immunol

© 2005 Nature Publishing Group

RESEARCH HIGHLIGHTS

IN BRIEF

B CELLS

Phosphorylation initiates AID response

I M M U N E R E G U L AT I O N

Monocarboxylate transporter MCT1 is a target for immunosuppression. Murray, C. M. et al. Nature Chem. Biol. 30 Oct 2005 (doi:10.1038/nchembio744) 293

Ac t iv at i on - i n du c e d c y t i d i n e deamin ase (AID) is required for antibody diversification through class-switch recombination (CSR) and somatic hypermutation (SHM). AID isolated from activated B cells has been shown to associate with replication protein A (RPA). This interaction requires phosphorylation of AID and is thought to direct AID to CSR and SHM target sequences in activated B cells. New research, published in Nature, now identifies cyclic-AMP-dependent protein kinase (PKA) as the enzyme that is responsible for phosphorylation of AID and regulation of AID-mediated CSR. That AID isolated from activated B cells can interact with RPA, whereas AID from 293 cells (which are nonlymphoid) engineered to express AID (denoted AID293) cannot, has been suggested to be important for the B-cell-specific targeting of AID to sites of CSR and SHM. Because the AID–RPA interaction requires phosphorylation of AID, Basu et al. set out to identify the protein kinase that is responsible for this in activated B cells. Nuclear extracts from activated B cells but not 293 cells could phosphorylate AID in vitro, indicating the presence of an AID protein kinase in nuclear extracts from activated B cells. Detailed analysis of these extracts showed that AIDprotein-kinase activity was associated with the catalytic subunit of PKA (denoted PKAcat) and that purified

PKAcat could phosphorylate AID in vitro. This phosphorylation conferred on AID293 the ability to deaminate double-stranded DNA (dsDNA) efficiently in vitro, a property that is associated with CSR and SHM (but freshly isolated AID 293 lacks this ability). Importantly, treatment with a PKA-specific inhibitor markedly impaired CSR in a B-cell clone that was activated to undergo CSR. Mass-spectrometry analysis of AID from activated B cells showed that Ser38 and Tyr184 were phosphory lated, and Ser38 was found to be in a consensus site for phosphorylation by PKA. Thr27 was identified as an additional potential site for phosphorylation by PKA. AID mutants in which either Ser38 or Thr27 had been mutated to Ala were not phosphorylated by PKA, and they failed to interact with RPA and failed to mediate deamination of dsDNA. By contrast, an AID mutant in which Tyr184 had been substituted with Ala was phosphorylated and could mediate deamination of dsDNA. Consistent with these results, when each of these AID mutants was expressed by AID-deficient B cells induced to undergo CSR, only the AID mutant with the Tyr184Ala substitution could restore CSR. These data indicate that PKA phosphorylates Ser38 of mouse AID, and this allows AID to interact with RPA and be targeted to sites of CSR. Because AID that is present in the cytoplasm of activated B cells is less efficient at deaminating dsDNA than is nuclear AID, the authors suggest that cytoplasmic AID is in an inactive state and that activation of PKA induces phosphorylation and thereby activation of AID such that it can enter the nucleus to mediate CSR. Karen Honey References and links ORIGINAL RESEARCH PAPER Basu, U. et al.

The AID antibody diversification enzyme is regulated by protein kinase A phosphorylation. Nature 26 Oct 2005 (doi:10.1038/nature04255)

Previous studies identified a group of immunomodulatory compounds that were structurally distinct from other immunosuppressive agents. Murray et al. generated analogues of these compounds and showed that one inhibited T-cell proliferation in vivo. Surprisingly, whereas many immunosuppressive drugs (such as cyclosporin and rapamycin) inhibit T-cell cytokine production, this agent inhibited T-cell proliferation. Furthermore, its target was found to be monocarboxylate transporter 1 (MCT1), which is involved in the transport of monocarboxylates such as lactate and pyruvate across the plasma membrane. This agent could inhibit l-lactate efflux and thereby reduce the glycolytic rate of activated T cells, presumably (as the authors suggest) to rates that are insufficient to support rapid T-cell proliferation. I M M U N O LO G I C A L SY N A P S E S

Cytotoxic granule polarization and cytolysis can occur without central supramolecular activation cluster formation in CD8+ effector T cells. O’Keefe, J. P. & Gajewski, T. F. J. Immunol. 175, 5581–5585 (2005)

Cell-surface receptors at the point of contact between an effector CD8+ T cell and a target cell segregate and cluster. Formation of a central supramolecular activation cluster (cSMAC) has been linked to granule exocytosis by, and cytotoxicity of, effector CD8+ T cells. However, data in this report indicate that formation of a cSMAC is not required. First, the number of effector CD8+ T cells that formed a cSMAC after encountering a target cell was fewer than the number of cells in which granules polarized towards the point of cell–cell contact. Second, the number of effector CD8+ T cells that formed a cSMAC was increased by target-cell expression of CD80, but the number of cells in which granules polarized towards the site of cell–cell contact and that mediated target-cell lysis was similar in the presence or absence of CD80. C Y TO K I N E S

Transcriptional repressor DREAM regulates T-lymphocyte proliferation and cytokine gene expression. Savignac, M. et al. EMBO J. 24, 3555–3564 (2005)

Savignac et al. show that downstream-regulatory-element antagonist modulator (DREAM), which is a transcriptional repressor that is released from DNA following Ca2+ signalling, is expressed by T cells and that its expression is downregulated after T-cell receptor (TCR) ligation. T cells isolated from transgenic mice expressing a dominant-active form of DREAM (which remains bound to DNA following Ca2+ signalling) showed decreased proliferation and cytokine production following TCR ligation. Conversely, when DREAM expression was decreased in wild-type splenocytes, an increase in the basal expression of mRNA encoding interleukin-2 (IL-2) and interferon-γ (IFN-γ) was detected. Because DREAM was bound to the promoters of the genes encoding IL-2 and IFN-γ, the authors suggest that DREAM has a role in regulating the basal expression of these cytokines.

NATURE REVIEWS | IMMUNOLOGY

VOLUME 5 | DECEMBER 2005 | 901


RESEARCH HIGHLIGHTS

P H A G O C Y TO S I S

ER fails to make a contribution

When a phagocyte encounters a pathogen, the triggering of cell-surface receptors leads to pathogen uptake in a membrane-bound vacuole (the phagosome). The simplest way to imagine

the process is as an invagination of the plasma membrane, but a paper published three years ago by Desjardins and colleagues (see Further reading) concluded that the endoplasmic reticulum (ER) also contributes membrane to the phagosome. Now, Touret et al. have re-examined this finding and conclude from the results of various independent methods that the ER does not make a significant contribution to phagosome formation. Using glycosylphosphatidylinositol (GPI)linked green fluorescent protein (GFP) as a marker of the plasma membrane in a macrophage cell line phagocytosing latex beads, they showed that the plasma membrane constitutes a large proportion of the early phagosome membrane, limiting any potential contribution of the ER. By contrast, in macrophages expressing the ER marker GFP–KDEL, there was no significant overlap of this marker with phagosome markers. This observation that the ER is not a significant component

B-CELL RESPONSES

What does it take to silence a B cell?

Even though there are mechanisms to eliminate developing B cells that recognize self-antigens in the bone marrow, a proportion of these cells are thought to escape elimination and enter the periphery. With their potential to contribute to autoimmune reactions, these self-reactive B cells must be kept in an anergic or tolerant state in the periphery. So, the authors of a recent report in Nature Immunology asked what might be required from the B-cell receptor (BCR) to maintain B-cell anergy. Using

902 | DECEMBER 2005

a mouse model of B-cell anergy, they showed that continuous BCR occupancy is required to ensure that self-reactive B cells remain unresponsive to encounter with additional antigens. It had previously been suggested that, because some tissue-specific self-antigens might be encountered only rarely by patrolling B cells, a single, transient encounter with cognate self-antigen would be sufficient to induce the anergic state that is then ‘memorized’ for the lifetime of the B cell. But this hypothesis needed direct assessment. To this end, John Cambier and colleagues made use of immunoglobulintransgenic mice in which B cells are specific for the hapten arsonate but crossreact with a self-antigen that induces anergy (probably single-stranded DNA). However, when these arsonate-specific B cells are cultured in the presence of high concentrations of a monovalent form of arsonate (arsonate-tyrosine, denoted ArsTyr), autoantigen is competitively dissociated, and anergy is lost.

| VOLUME 5

of the phagosome was also made for various combinations of ER and phagosome markers and for mouse dendritic cells. Immuno-electron microscopy of macrophage sections that were labelled with calnexin-specific antibodies could not detect the ER-resident protein calnexin on the phagosome membrane, and immunocytochemical staining could not detect the ER marker glucose 6-phosphatase in phagosomes. The possibility that the ER makes a minor contribution to the phagosome membrane by transient fusion was analysed for macrophages that had been stably transfected with the ER marker GFP–KDEL and were exposed to latex beads in the presence of high extracellular concentrations of the dye FM4-64. FM4-64 stained the plasma and phagosome membranes but did not reach the ER membrane (as shown by lack of colocalization with GFP–KDEL), even when phagocytosis was arrested by wortmannin to stabilize any transient connections that might form between the phagosome and the ER. Biochemical assessment of ER–phagosome fusion was carried out using macrophages that had been stably transfected with the soluble ER marker avidin–KDEL and were allowed to ingest biotinylated beads. Immunostaining

To explore the kinetics of this reversal of anergy, the authors looked at intracellular concentrations of free calcium in the anergic B cells treated with or without ArsTyr. Anergic B cells are known to have higher basal concentrations of intracellular calcium than naive B cells, possibly because of either continuous BCR signalling or an altered physiological state that is triggered by a single exposure to self-antigen. Consistent with the former explanation, treatment of arsonatespecific B cells with ArsTyr led to a rapid reduction (within 2–4 minutes) in intracellular calcium concentrations. Washing the cells free of ArsTyr and then culturing them in the absence of ArsTyr rapidly restored intracellular calcium to the original high concentration. These findings indicate that at least some features of anergy are maintained by continued biochemical signals rather than by genetic reprogramming. Treatment with ArsTyr also reversed other features of anergic B cells, including increased basal phosphorylation of extracellular-signal-regulated kinase (ERK) and increased expression of activation markers, such as CD80 and CD95. Importantly, the shortened lifespan that is characteristic of anergic B cells



of isolated phagosomes with avidin-specific antibodies failed to detect association of avidin with the phagocytosed beads. Finally, the pH sensitivity of GFP (and its derivatives such as yellow fluorescent protein, YFP) was used to analyse the association of phagosomes with the ER on the basis that ER components delivered to the acidic phagosome would experience a decrease in pH. Addition of the weak base ammonia increased the fluorescence of GPI– YFP in phagosomes but had no effect on the fluorescence of GFP–KDEL, showing that this ER marker is not exposed to the acidic phagosome environment. All of these experiments fail to provide any evidence in support of fusion of the ER with phagosomes, although the authors point out that they cannot definitively rule out the involvement of the ER in phagocytosis of particles of a different size or in phagocytosis using different cell-surface receptors. Kirsty Minton References and links ORIGINAL RESEARCH PAPER Touret, N. et al.

Quantitative and dynamic assessment of the contribution of the ER to phagosome formation. Cell 123, 157–170 (2005) FURTHER READING Gagnon, E. et al. Endoplasmic reticulum-mediated phagocytosis is a mechanism of entry into macrophages. Cell 110, 119–131 (2002)

was markedly extended by treatment with ArsTyr, and the inability of these cells to upregulate expression of the activation marker CD86 after stimulation with IgM-specific antibody was restored by treatment with ArsTyr. Using pharmacological inhibitors of protein kinases that are known to be involved in BCR signalling, the authors confirmed that maintenance of the anergic B-cell phenotype depends on signals from the BCR. One implication of these observations is that transient loss of self-antigenmediated BCR signalling might lead to a resetting of a threshold trigger such that activation now might contribute to autoimmunity. Indeed, pretreatment of arsonate-specific B cells with ArsTyr for as little as 2–3 minutes allowed these previously unresponsive cells to respond to stimulation with IgM-specific antibody, by mobilizing calcium. Lucy Bird References and links ORIGINAL RESEARCH PAPER Gauld, S. B., Benschop, R. J., Merrell, K. T. & Cambier, J. C. Maintenance of B cell anergy requires constant antigen receptor occupancy and signaling. Nature Immunol. 6, 1160–1167 (2005) FURTHER READING Wienands, J. Unraveling B cell receptor mechanics. Nature Immunol. 6, 1072–1074 (2005)

N AT U R A L K I L L E R T C E L L S

Identifying functional subsets of NKT cells The ability of natural killer T (NKT) cells either to promote or to suppress cellmediated immunity has been shown in various model systems in mice, but the reason for these contrasting effects is not well understood. Now, Nadine Crowe and colleagues show that there are functionally distinct subsets of NKT cells in vivo, which could help to explain the range of effects of NKT cells. In previous studies, the authors showed that NKT cells derived from the liver can promote antitumour immune responses in two model systems: mice injected with the 3-methylcholanthrene-induced sarcoma cell line MCA-1, and mice injected with the melanoma cell line B16F10. Using these models, it was shown that mice that lack T-cell receptor (TCR) α-chains that contain Jα18 (denoted TCR Jα18), which are deficient in NKT cells, are more susceptible to tumour growth. In both tumour models, the ability of the NKT cells to promote antitumour responses was dependent on their production of interferon-γ. Previous reports have shown that there are at least two phenotypically distinct subsets of NKT cells in mice and humans — CD4+ and CD4– NKT cells — and that these subsets show differential cytokine production in vitro. To test the idea that NKT-cell subsets are functionally distinct, as well as phenotypically distinct, the authors isolated NKT cells from the spleen, thymus and liver, then adoptively transferred these cells to TCR Jα18deficient mice that had been injected with MCA-1. Only the liver-derived NKT cells could completely inhibit tumour growth, and this protection was found to be provided mainly by the CD4– population of NKT cells. The inability of thymusderived NKT cells to confer protection was not a consequence of their impaired survival after transfer, because they were easily detectable in the liver and other organs for at least 1 week after transfer.


Because it was possible that liver-derived NKT cells were preferentially activated in the MCA-1 model, the authors then tested various NKT-cell subsets in the B16F10 model. In this model, liver-derived NKT cells transferred to B16F10-inoculated TCR Jα18-deficient mice that were treated with the pan-NKT-cell-activating molecule α-galactosylceramide (α-GalCer) could inhibit the formation of lung metastases. Similar to the MCA-1 model, spleen- and thymus-derived NKT cells were less effective than were liver-derived NKT cells at preventing tumour growth, and the CD4– subset of liver-derived NKT cells was more potent at promoting the antitumour response than was the CD4+ NKT-cell subset. However, these differences did not seem to result from differences in interferon-γ production, because liverderived NKT cells that were isolated from mice deficient in interleukin-4 (IL-4) were considerably better at protecting against the formation of metastases than were their wild-type counterparts, and thymusderived NKT cells from these mice were also protective. This indicates that IL-4 production by NKT cells could antagonize the ability of these cells to mediate tumour rejection. However, because wild-type, liver-derived NKT cells produce similar amounts of IL-4 to wild-type, thymusderived NKT cells, it is not clear why thymus-derived cells cannot confer this protection. This study shows that there are functionally distinct subsets of NKT cells in vivo, and it highlights the importance of addressing this issue in future studies and in clinical trials of α-GalCer-based therapies. Elaine Bell References and links ORIGINAL RESEARCH PAPER Crowe, N. Y. et al. Differential antitumor immunity mediated by NKT cell subsets in vivo. J. Exp. Med. 202, 1279–1288 (2005)



RESEARCH HIGHLIGHTS

I M M U N E R E G U L AT I O N

Take two for Notch Previous studies have indicated that signalling through Notch proteins is involved in the differentiation of CD4+ T cells into T helper (TH) cells, although whether these signals are required for differentiation into TH1 cells, TH2 cells or both was not clear. But now, in a paper published in The Journal of Experimental Medicine, Notch signalling in mature CD4+ T cells is shown to be required for TH2-cell-mediated immunity. Notch signalling that depends on the transcription factor CSL (also known as RBP-Jκ) requires the transcription cofactor mastermind-like 1 (MAML1). So, to study the role of Notch signalling in CD4+ T cells, Tu et al. generated mice (denoted as CCD mice) in which a green-fluorescent-protein-tagged, dominant-negative form of MAML1 was expressed only by CD4+ T cells.

Importantly, T-cell development — in terms of the cellularity of the thymus, the proportion of CD4+ and CD8+ T cells, and the expression of cell-surface markers of activation — was normal in these mice. However, when CD4+ T cells from these mice were cultured under TH2-cellpolarizing conditions, the proportion of cells that produced interleukin-4 (IL-4), and the amount of IL-4 that they produced, was markedly lower than in control-cell cultures. By contrast, normal numbers of interferon-γ-producing cells were generated after culture under TH1-cell-polarizing conditions. Consistent with a role for CSL-dependent Notch signalling in optimal differentiation into TH2 cells, CCD mice failed to generate a protective TH2-cell response after infection with Trichuris muris. By contrast, CCD mice infected with Leishmania major generated

IMMUNE RESPONSES

Shuttling serotonin: not just in our heads

Serotonin is generally thought of as a neurotransmitter that is passed between neurons at neuronal synapses for the regulation of appetite, mood and pain. Now, Peta O’Connell and colleagues report that serotonin might be delivered from dendritic cells (DCs) to

904 | DECEMBER 2005

T cells across the immunological synapse in a manner similar to that which occurs between neurons. They propose that this is a new form of rapid communication between DCs and T cells and that this might have important implications for the regulation of T-cell responses. A role for serotonin in the immune response has previously been reported: serotonin has been shown to be a mediator that is released by mast cells and platelets in response to injury or pro-inflammatory signals. In this study, the authors show that, although DCs cannot themselves synthesize serotonin, they express the serotonin transporter SERT (also known as SLC6A4). Cell-surface expression of SERT by DCs was increased following activation and maturation, and this enabled the cells to take up serotonin from their microenvironment. Presumably, mature DCs could acquire serotonin from platelets or mast cells at inflammatory sites, but the authors show that activated T cells can synthesize serotonin for uptake by mature DCs. Serotonin taken up by activated DCs was stored in vesicular compartments and not

| VOLUME 5

a protective TH1-cell response similar to that generated by control animals. This study indicates that CSL-dependent Notch signalling is required for generation of a protective TH2-cell response in vivo. The authors suggest that targeting this signalling pathway could potentially be used to treat TH2-cell-mediated diseases such as asthma. Karen Honey References and links ORIGINAL RESEARCH PAPER Tu, L. et al. Notch signaling

is an important regulator of type 2 immunity. J. Exp. Med. 202, 1037–1042 (2005)

degraded by the serotonin catabolic enzymes monoamine oxidase A (MAOA) and MAOB, owing to downregulation of expression of these enzymes after DC activation. By loading DCs with radiolabelled serotonin, the authors next showed that stored serotonin is rapidly released, through exocytosis, in the presence of extracellular ATP, which induces the mobilization of intracellular calcium as occurs in DCs on interaction with T cells. The authors then asked what effect serotonin might have on T cells. Treatment of T cells with exogenous serotonin reduced the concentration of the intracellular-signalling molecule cyclic AMP (cAMP), which inhibits T-cell proliferation at high concentrations, indicating that serotonin might terminate the initial increase in cAMP concentration that occurs after T-cell-receptor ligation and therefore promote T-cell activation. Given these data, the authors suggest that DCs shuttle serotonin, possibly synthesized by activated T cells, to the synaptic space between a DC and a naive T cell. This allows localized delivery of a high concentration of this labile molecule, which is required for activating the naive T cell. Lucy Bird References and links ORIGINAL RESEARCH PAPER O’Connell, P. J. et al. A novel

form of immune signaling revealed by transmission of the inflammatory mediator serotonin between dendritic cells and T cells. Blood 13 Oct 2005 (doi:10.1182/blood-2005-07-2903)



REVIEWS IMMUNIZATION WITHOUT NEEDLES Samir Mitragotri Abstract | Most current immunization procedures make use of needles and syringes for vaccine administration. With the increase in the number of immunizations that children around the world routinely receive, health organizations are beginning to look for safer alternatives that reduce the risk of cross-contamination that arises from needle reuse. This article focuses on contemporary developments in needle-free methods of immunization, such as liquid-jet injectors, topical application to the skin, oral pills and nasal sprays. LIVE ATTENUATED VIRUS

A weakened mutant of a wildtype virus that is antigenic but not infectious.

Department of Chemical Engineering, University of California, Santa Barbara, California 93106, USA. e-mail: samir@engineering. ucsb.edu doi:10.1038/nri1728 Published online 20 October 2005

Needles and syringes are the most commonly used method for administering vaccines and protein therapeutics, such as insulin, to humans. The World Health Organization (WHO; see the Online links box) estimates that 12 billion injections are given annually, 5% of which are used for immunizations1. Despite their common use, needle-based immunizations have several limitations. Needle phobia is an important issue for both adults and children2 and makes immunizations stressful3. In addition, accidental needle sticks are a serious problem in both developed and developing countries. The Centers for Disease Control and Prevention (CDC; see National Immunization Program in the Online links box), in the United States, estimates that more than 300,000 needle-stick injuries occur annually in US hospitals4. An estimated 5 accidental needle-stick injuries occur per 100 injections worldwide, posing a considerable risk to health-care providers5. An even greater shortcoming of injections is their improper and unsafe use. This mainly involves the reuse of needles and syringes, which is common in developing countries for reasons of cost1. (A detailed list of unsafe injection practices is given in REF. 1.) The WHO has estimated that as many as one-third of immunization injections are unsafe in four of its six geographical regions5,6. Each year, an overwhelming number of infections with HIV (80,000–160,000), hepatitis C virus (HCV; 2.3–4.7 million) and hepatitis B virus (HBV; 8–16 million) are thought to originate from the reuse of needles and syringes by health-care providers7. The WHO estimates that 32% of HBV infections, 40% of HCV infections and 5% of


HIV infections in developing countries are attributable to unsafe injection practices8. Not surprisingly, the development of needle-free immunization methods has now been identified as an important goal in global health care9. Needle-free immunizations made their first notable appearance almost 50 years ago with the oral polio vaccine, which is still used in developing countries but has been discontinued in the United States since 2000 BOX 1. This vaccine, which contains a LIVE ATTENUATED poliovirus, can infect the gastrointestinal tract and, subsequently, generate adequate immune protection in the host. Several other needle-free vaccines (oral typhoid fever, oral cholera, oral rotavirus and nasal influenza), which also contain live attenuated pathogens, are now available TIMELINE. However, the administration of most vaccines without the use of needles has proved to be challenging, especially for non-living vaccines (that is, killed pathogens, and subunit, toxoid, peptide and DNA vaccines), which offer several advantages BOX 1. Consequently, in developed countries, as well as developing countries, most childhood vaccines — including those against hepatitis B (a subunit vaccine), diphtheria–tetanus–pertussis (toxoid and inactivated bacteria), polio (killed virus), varicella (live attenuated virus), measles–mumps– rubella (live attenuated virus), tuberculosis (live attenuated bacteria) and yellow fever (live attenuated virus) — are administered using needles and syringes. In the past decade, however, there has been a strong step forward in addressing the technological challenges that are associated with immunization without needles10,11.



REVIEWS Current methods of needle-free immunization, either commercially available or under development, can be classified into two broad classes — cutaneous immunization and mucosal immunization — depending on the site of vaccine administration (FIG. 1). Cutaneous methods of immunization include the following: liquid-jet injection, which delivers a highspeed vaccine stream into intradermal, subcutaneous or intramuscular regions; ballistic methods (also known as epidermal powder immunization), which accelerate particulate vaccine material and deposit it in the skin; and topical application methods, which deliver the vaccine into or across the skin through passive diffusion or facilitated transdermal transport (FIG. 2) . Mucosal immunization methods involve delivery of vaccines to a mucosal membrane, such as the ocular, oral, nasal, pulmonary, vaginal or rectal membrane TABLE 1. This article provides an overview of the development of these methods, with an emphasis on the challenges that are associated with the delivery of non-living vaccines. Particular attention is paid to needle-free cutaneous immunization. Detailed discussion of mucosal immunization can be found elsewhere11,12. Liquid-jet injections

Jet injection is the oldest method of needle-free immunization. The origin of jet injections can be traced to the late 1800s, when a technique known as aquapuncture was reported in the medical literature13. This device was used to deliver jets of water and other liquids for applications other than immunization: for example, for the treatment of uncontrolled neuralgia.

However, it was in the early 1950s when jet injections took their place as a needle-free method of delivering medications and vaccines14. A liquid-jet injector uses the kinetic energy of a highvelocity vaccine jet (typically more than 100 m per second) with a diameter that ranges from 76 µm to 360 µm, which is smaller than the outer diameter of a standard hypodermic needle (810 µm for a 21G needle). Liquid jets penetrate the skin and deliver the vaccine into the skin (that is, intradermally), the subcutaneous tissue or the muscle (intramuscularly) (FIG. 2A). Skin is a particularly attractive target for vaccine administration because it forms an integral part of the immune system15,16. The epidermis is enriched with LANGERHANS CELLS, which form a network that allows them to take up antigen efficiently and therefore to carry out immune surveillance17. The Langerhans-cell network is the next line of defence after the physical barrier of the skin has been breached. Langerhans cells initiate specific immune responses by processing and presenting antigen fragments to naive T cells in the lymph nodes18. This promotes the generation of both systemic (IgG and IgM) and mucosal (IgA) humoral immune responses19,20. Targeting the vaccine to the skin promotes its contact with Langerhans cells and reduces the required dose of vaccine21,22, a factor that would become crucial at a time of vaccine shortage, such as the predicted H5N1 influenza-virus pandemic. Vaccines that are delivered by liquid-jet injectors typically spread throughout a larger tissue volume after injection than do vaccines that are administered with needles23, which might allow them to establish better or faster contact with antigen-presenting cells before they are degraded.

Box 1 | Live versus non-living vaccines

LANGERHANS CELL

A type of dendritic cell (which are professional antigenpresenting cells) that is localized in the epidermal layer of the skin.

906 | DECEMBER 2005

Live vaccines contain attenuated pathogens that have minimal virulence but high immunogenicity. These vaccines include natural viruses that are non-transmissible analogues of pathogens (for example, the smallpox vaccine), wild-type viruses that have been passaged in vitro to reduce their pathogenicity (for example, the measles–mumps– rubella vaccine) and cold-adapted viruses (for example, one type of influenza vaccine). Advantages of live vaccines include the generation of strong humoral, as well as cell-mediated, immune responses, the generation of immune responses of lengthy duration and the lack of requirement of an adjuvant. However, live vaccines, in theory, carry a risk of virulence, and occasionally, they have other vaccine-associated effects. For example, the use of oral polio vaccine was discontinued in the United States because of rare cases of vaccine-associated paralytic polio144, and it was replaced by inactivated poliovirus vaccine. Another live vaccine, RotaShield (Wyeth Ayerst), was also recalled, owing to vaccine-related intussusception145. For a detailed review on vaccine safety, see REF. 146. Some live vaccines are also limited by practical constraints, such as difficulties in culturing (for example, hepatitis C virus147) and requirements for advanced storage and handling (for example, control of temperature) to maintain the pathogen. Non-living vaccines encompass several types of vaccine: inactivated whole pathogens (for example, one type of influenza vaccine or the hepatitis A vaccine); subunit vaccines that include the epitope that is recognized by antibodies (for example, the hepatitis B vaccine); toxoids, which are deactivated pathogenic toxins (for example, the tetanus vaccine); synthetic peptides that mimic epitopes from the antigen; and DNA that encodes the antigen. Advantages of non-living vaccines include the absence of virulent pathogens, the ability to manufacture these vaccines to a high level of purity, and their stability under adverse conditions (for example, heat), which facilitates their use in field applications. Non-living vaccines, however, do not typically induce cellular immunity, which is partly because of their method of administration (by injection). They require frequent administration and larger doses than do live vaccines. Non-living vaccines require parenteral administration (except for one non-living cholera vaccine). Unlike live vaccines, which survive the low pH and the enzymatic environment of the stomach and can then infect the gastrointestinal tract and generate an immune response, non-living vaccines (especially subunit, synthetic peptide and DNA vaccines) are destroyed in the gastrointestinal tract. Strong adjuvants are typically required to counteract this vaccine loss and to make up for the lower immunogenicity of non-living vaccines.

| VOLUME 5



REVIEWS

Timeline | Important events in the development of needle-free methods of immunization Oral transgenic-plant heat-labile enterotoxin vaccine tested in humans Vaginal and rectal immunization with whole cell plus subunit cholera vaccine tested Oral microspheres tested in humans

1952

1960

MUNJIs introduced

Oral polio vaccine tested

1961

1975

Aerosol measles immunization initiated Safety concerns with MUNJIs noted (1975–1990s)

1989

Vivotif (Berna Biotech AG), oral typhoid fever vaccine, approved by FDA

1990

1994

DCJIs introduced (early 1990s)

Epidermal powder immunization tested in humans for hepatitis B virus DNA

Oral liposomes tested in humans

1996

Transcutaneous immunization with an adjuvant patch tested in humans

1998

RotaShield (Wyeth Ayerst), oral rotavirus vaccine, approved by FDA; recalled in 1999

1999

FluINsure (ID Biomedical Corporation), nasal influenza vaccine, Phase II trial completed

2000

Orochol (Berna Biotech AG), oral cholera vaccine, approved in Switzerland

2002

Ultrasound device tested in humans for tetanus toxoid

2003

2004

FluMist (Med Immune Vaccines, Inc.), nasal influenza vaccine, approved by FDA Rotarix (GlaxoSmithKline), oral rotavirus vaccine, approved in Mexico

Nasalflu (Berna Biotech AG), nasal influenza vaccine, introduced in Switzerland; discontinued in 2001

2005

RotaTeq (Merck & Co., Inc.), oral rotavirus vaccine, FDA application filed

Dukoral (SBL Vaccin AB), oral cholera vaccine, approved by EMEA

Events below the arrow (red boxes) correspond to methods that have been used in commercial applications. Events above the arrow (green boxes) correspond to methods that are currently under development but that have been clinically tested. DCJI, disposable-cartridge jet injector; EMEA, European Medicines Agency; FDA, Food and Drug Administration (United States); MUNJI, multi-use-nozzle jet injector.

SEROCONVERSION

Development of a detectable concentration of pathogenspecific antibodies in the serum as a result of infection or immunization.

Liquid-jet injections were first popularized by multi-use-nozzle jet injectors (MUNJIs), which allow injection of several doses using the same nozzle and vaccine reservoir at a rate of up to 1,000 immunizations per hour. MUNJIs were successfully used for immunizing humans with live vaccines against measles and smallpox, as well as non-living vaccines against cholera, hepatitis B, influenza and polio24. Liquid-jet injectors offer several advantages in addition to avoiding the use of sharps. They have a long history of use, and they work with existing vaccine formulations that have been developed for needle-based administration. In one example, they resulted in higher SEROCONVERSION rates, but the reasons for this are not clear25. At the same time, liquid-jet injectors have several limitations. In some studies, they were associated with higher levels of pain than were needle-based injections26, especially when using older MUNJI devices. Liquid-jet injectors have also been associated with more-frequent site reactions than have needles, such as soreness, redness and swelling of the injection site25,27. Perhaps the main safety issue that is associated with MUNJIs is the increased risk of subject-to-subject contamination. In the 1980s, the spread of HBV between subjects was linked to liquid-jet injections28. Splashes of small amounts of blood or interstitial fluid on the nozzle of the MUNJI were blamed for this spread of HBV. Systematic studies have shown that MUNJIs can transmit considerable volumes of blood (more than 10 pl) from one subject to another when the MUNJIs are used on multiple subjects. This volume of blood is presumed to be sufficient to transmit HBV infection29. The WHO and CDC recommend that MUNJIs should be used for mass immunization only when the


gains from rapid immunization outweigh the risks of blood-borne disease: for example, during influenza pandemics or bioterrorism attacks30. To minimize the risk of contamination, protective devices that are disposable have been developed to cover the surface of the injector, and studies that have been carried out with these devices showed no risk of contamination31. Disposable-cartridge jet injectors (DCJIs; which are non-disposable injectors to which disposable nozzles are attached for each use) have also been developed to alleviate concerns about contamination. Single-use, pre-filled disposable devices are also under development to alleviate these concerns about contamination, and these devices present a new direction in liquid-jet injectors32. In addition to conventional vaccines, some DCJIs have also been effectively used to carry out DNA vaccination against dengue fever and influenza in animals33–35. Although MUNJIs are no longer used for routine immunizations, DCJIs are used for childhood immunizations at the physician’s discretion. However, at present, the number of immunizations that is carried out with DCJIs is far less than the number that is carried out using needles, possibly because of the cost, the low level of awareness among health-care providers and patients, the potential pain, and the problems that were associated with earlier generations of liquid-jet injectors. Numerous liquid-jet injectors are already on the market and are used for various vaccines, including those against influenza and hepatitis B. Newer, more convenient liquid-jet injectors are continually being developed, mainly by small companies. Although substantial technological advances have been made in



REVIEWS

Cutaneous immunization

Mucosal immunization Ocular immunization (Drops)

Epidermal powder immunization (DNA-coated gold particles or vaccine powders) Liquid-jet injection (Off-the-shelf vaccine formulations) Topical application (Adjuvant patches, colloidal carriers, ultrasound or microneedles)

Nasal immunization (Sprays and drops containing adjuvants plus liquid formulations, liposomes or microspheres) Pulmonary immunization (Aerosols or powders)

Oral immunization (Liquid formulations and pills containing adjuvants plus liposomes, microspheres or bacterial ghosts) Vaginal or rectal immunization (Creams containing adjuvants)

Figure 1 | Schematic representation of various methods of needle-free immunization. Four distinct modes of immunization are discussed in this article: liquid-jet injection, epidermal powder immunization and topical application (which are all forms of cutaneous immunization), and mucosal immunization, which is further classified into ocular, nasal, oral, pulmonary, and vaginal or rectal routes. Ocular immunization can be carried out using eye drops. Nasal immunization is carried out using sprays that comprise liquid formulations, liposomes or microspheres. Vaccines can be delivered orally in the form of liquid doses or pills, both of which can consist of various formulations: for example, microspheres. Vaccines can also be delivered to the vaginal or rectal mucosal membrane, using topical creams, or to the lungs, using aerosols or powders. Liquid-jet injection delivers vaccines to the skin, subcutaneous fat or muscles, depending on the parameters of the injection. Epidermal powder immunization delivers vaccine powders to the superficial layers of the skin. Transdermal patches also deliver vaccines to the superficial layers of the skin. Topical delivery of vaccines is facilitated by adjuvant patches, colloidal carriers, or physical methods, such as microneedles and ultrasound. For a general discussion of the issues that are associated with delivery of molecules (not just vaccines) through these routes, see the following references: ocular drug delivery148,149, epidermal powder delivery150, liquid-jet injection23, transdermal drug delivery51, nasal drug delivery151,152, pulmonary drug delivery153, oral drug delivery86,154 and vaginal drug delivery155.

the past decade, the fundamental science that underlies liquid-jet injections is not well understood. Studies on various basic aspects of liquid-jet injections, such as dynamics of jet fluid, mechanics of jet penetration and dynamics of jet dispersion, have only recently been reported in the literature36,37. It is hoped that this understanding, together with technological advances, will lead to better and cheaper devices. Particle bombardment of the skin

ADJUVANT

An agent that is mixed with an antigen and increases the immune response to that antigen following immunization.

908 | DECEMBER 2005

Particle-based methods (also known as ballistic methods) accelerate powdered vaccines such that they penetrate the outer layer of the skin (that is, the stratum corneum) and are deposited in the epidermis or the superficial layers of the dermis, a method known as epidermal powder immunization (EPI)38 (FIG. 2B). The ballistic technique was first developed in 1986, for the delivery of DNA-coated metal particles of ∼1 µm in diameter into plants to genetically modify them, and it was known as the gene gun. In the early 1990s, the ballistic method was developed into devices for delivering both conventional and DNA vaccines to humans39. Unlike liquid-jet injectors, which routinely deliver the vaccine to the subcutaneous or intramuscular space, ballistic methods mainly deliver the vaccine to the superficial layers of the skin39 and therefore naturally target Langerhans cells.

| VOLUME 5

Several vaccines have been delivered to animals using EPI. Influenza vaccine, when administered together with ADJUVANTS such as cholera toxin (CT) by EPI, elicited augmented serum and mucosal antibody responses in mice compared with those in unimmunized animals40,41. Similar results were reported for diphtheria toxoid (DT). Co-administration of adjuvants increased IgG titres after EPI-based immunization against diphtheria42 and influenza43. EPI has also been used to deliver DNA vaccines to animals. Small (1–3 µm) DNA-coated gold or tungsten particles delivered by EPI directly penetrate epidermal keratinocytes or Langerhans cells, and they induce expression of encoded antigens44. Additional means to capitalize on the immunostimulatory properties of Langerhans cells — for example, co-administering DNA that encodes cytokines (such as interleukin-6) or inhibitors of apoptosis45 — have also been adopted. Apoptosis inhibitors increase the survival of Langerhans cells after particleinduced mechanical trauma, and the expression of cytokines facilitates the migration of Langerhans cells, which is typically promoted by pro-inflammatory mediators. There are fewer reports of EPI of humans. In one study, EPI efficiently delivered influenza vaccine to humans46. For all influenza-virus strains, titres of IgG were equivalent in groups in which EPI was used and in needle-immunized individuals. Clinical studies of immunization with DNA using ballistic methods have also yielded encouraging results47–49. EPI-based DNA immunization against infection with HBV induced high titres of protective antibody, as well as cell-mediated immune responses, in humans. Despite promising results in clinical studies, however, the commercial development of EPI for conventional vaccines seems to be stagnant. Instead, current development efforts in industry are focused entirely on DNA vaccines. EPI offers several advantages as a mode of immunization. The use of powders simplifies handling and storage compared with liquid formulations. EPI also naturally targets Langerhans cells and allows their direct transfection. Initial safety studies of EPI seem to be satisfactory, although occasional bleeding was observed in some cases46. As is the case for liquid-jet injectors, fundamental studies that focus on the mechanics of particle penetration, which will be useful for understanding the mechanisms of EPI, have only recently been initiated38,50. These studies have documented the role in EPI of the properties of the particle (such as density and size), the operating conditions (that is, temperature, humidity and velocity) and the mechanical properties of the skin. Understanding gained from such studies might assist in the design of future EPI devices. Topical application to the skin

The skin has been used for administering medication to treat local conditions (for example, inflammation) for thousands of years. Systemic drug delivery through the skin became prominent with the introduction of



REVIEWS

COLLOIDAL CARRIER

A stable system of small particles of lipids, polymers or any other material that encapsulate a vaccine.

strong barrier to molecular transport54. Increasing the permeability of the stratum corneum without irritating the underlying keratinocytes has been a considerable challenge in the field. Several innovative methodologies are being developed to facilitate antigen delivery into the skin. These include the use of topically applied adjuvants, COLLOIDAL CARRIERS to encapsulate vaccines, and physical methods to increase the permeability of the skin to vaccines (FIG. 2C).

transdermal scopolamine patches for treating motion sickness, in 1979 REF. 51. The use of the skin for the administration of vaccines has an even longer history. Immunization against smallpox was practiced in India more than 1,000 years ago by scratching dry scabs from smallpox lesions onto the skin of healthy individuals. The skin remains the site for immunization against smallpox in the modern era, using the bifurcated needle52. Although the skin has had a historical role in immunization, the use of topical vaccine application as a general mode of immunization has only recently (in the mid-1990s) received attention. The simple topical application of a vaccine does not typically yield an adequate immune response, although rare cases can be found in the literature53. Topical delivery of vaccines into the skin is limited by the low permeability of the stratum corneum, the outer layer of skin, which is 15–20 µm in thickness and consists of cornified keratinocytes embedded in a lipid-rich matrix. The lipids of the stratum corneum are organized into an ordered bilayer structure and, consequently, form a

Topical adjuvants. Topical application of adjuvants such as CT together with the vaccine on the skin generates a strong systemic and mucosal immune response55–57. This is the most studied of all topical-immunization methods. Topical application of CT provides the required activation signal for Langerhans cells to mature and become potent antigen-presenting cells that can prime the immune response to co-administered vaccines58. It is unclear how CT, a relatively large protein (86 kDa), diffuses across the stratum corneum, but hydrationinduced permeabilization of the stratum corneum is one

Colloidal carrier

b Tape stripping

a

d

c

Ultrasound

e

Electroporation

f

Adjuvant patch

g

h Microneedle

Stratum corneum

Micropore

Epidermis

Vaccine powder

Dermis

Langerhans cell

A

B

Liquid-jet injection

Epidermal powder immunization

Hair follicle

C Topical application

Figure 2 | Immunization by cutaneous routes. A | Liquid-jet injection delivers vaccine to muscular, subcutaneous or dermal regions, depending on the parameters of the injection. B | Epidermal powder immunization delivers vaccine powders to the superficial layers of the skin (that is, the epidermis and the superficial layers of the dermis), where they are recognized by Langerhans cells. C | Topical application of vaccines delivers vaccines to the epidermis, where they are recognized and processed by Langerhans cells. Immunization by topical vaccine application is facilitated by several methods. Ca | DNA immunization can be carried out through hair follicles. Cb | Tape stripping removes the stratum corneum and facilitates vaccine absorption. Cc | Thermal or radio-wave-mediated ablation of the stratum corneum creates micropores that increase vaccine delivery. Cd | Colloidal carriers such as microemulsions and transfersomes increase dermal absorption of topically applied vaccines. Ce | Low-frequency ultrasound is an adjuvant for topically applied vaccines, and it also increases vaccine delivery to the skin. Cf | Topically applied adjuvants, such as cholera toxin, can induce potent immune responses. Cg | Electroporation of the stratum corneum increases the delivery of DNA vaccines to the epidermis. Ch | Shallow microneedles that penetrate into the epidermis deliver vaccines effectively. For an overview of the issues that are associated with transdermal delivery of molecules (not just vaccines) by some of these methods, see REFS 56,72,156158.




REVIEWS

Table 1 | Comparison of routes of needle-free delivery of non-living vaccines Method

Advantages

Limitations

Cutaneous immunization Liquid-jet injection

Long history of use, ability to work with existing formulations, and success with many forms of vaccine

Issues associated with cross-contamination when using MUNJIs, high cost of device, and occasional pain and bleeding

Epidermal powder Use of powders facilitates storage, strong data immunization for DNA vaccines, and natural targeting of Langerhans cells

High cost of device, occasional bleeding, limited clinical data for non-DNA vaccines, and limited clinical history

Topical application

Strong adjuvants or permeabilizing agents required, some permeabilization methods require expensive devices, and most delivery methods have limited clinical history

Ease of access, natural targeting of Langerhans cells, generation of mucosal and systemic immunity, and high patient compliance

Mucosal immunization Oral

Ease of administration, high patient compliance, no complex devices necessary in most cases, primary site of infection of many pathogens, and long history of use for live attenuated pathogens

Gastrointestinal deactivation of vaccines, high doses required, variability of response, and mixed clinical data

Nasal

Easier access to mucosal membrane than for oral delivery, low cost, and one of the main sites of infection for airborne pathogens

Short contact time, enzymatic activity of nasal tissue, adjuvants required, safety concerns with earlier nasal vaccine, and limited applicability in patients with upper respiratory-tract infections

Pulmonary

Large surface area, one of the main sites of infection for airborne pathogens, and history of use for measles vaccine

Strong adjuvants required, high cost of some devices, and interference from upper respiratory-tract infections

Vaginal or rectal

High relevance for HIV and causative agents of other sexually transmitted diseases

Poor patient compliance for general applications, and strong adjuvants required

MUNJI, multi-use-nozzle jet injector.

possibility. In recent studies, however, disruption of the stratum corneum using emery paper (which is an abrasive paper) has been used before the application of vaccine and/or adjuvant to achieve immune responses59,60. Additional strategies, which also involve disruption of the stratum corneum, are discussed later. Several other adjuvants that have fewer toxicity issues than CT, such as the B subunit of CT (CTB), also have adjuvant-like properties after topical administration61. The effectiveness of adjuvant-mediated cutaneous immunization has been shown in animals, using vaccines that include tetanus toxoid (TT)19, DT62 or Bacillus anthracis (the causative agent of anthrax)59. Clinical studies have also confirmed the generation of a strong IgG and IgA response in volunteers after topical application of a colonization factor from enterotoxigenic Escherichia coli together with adjuvants 63. Earlier clinical studies that were carried out using E. coli heat-labile enterotoxin (LT) as an adjuvant showed that mucosal antibodies were generated, and these studies confirmed the role of Langerhans cells in immunization by topical vaccine application, as shown by the strong presence of these cells in the skin 24–48 hours after immunization64. MICROEMULSION

A stabilized emulsion (that is, a preparation of two immiscible liquids, in which one is dispersed in the other) in which the dispersed droplets are extremely small.

910 | DECEMBER 2005

Colloidal carriers. Encapsulation of vaccines in colloidal carriers facilitates the generation of an immune response after topical application. Few studies have reported on the use of colloidal carriers for the topical delivery of vaccines, all in animals. Topical application

| VOLUME 5

of TT encapsulated in lipid vesicles, after booster immunization, elicited a specific immune response (IgG) comparable to that produced by intramuscular injections of alum-adsorbed TT65. Another lipid-based system has been used to deliver DNA vaccines to animals66. Topical application of this DNA–lipid vaccine resulted in both antibody responses and cellular responses. Ethanol-in-fluorocarbon MICROEMULSION systems67,68 and cationic nanoparticles coated with DNA vaccines have also been used for topical DNA immunization of animals69. The precise mechanisms by which colloidal carriers penetrate the stratum corneum remain a topic of research. Whether the results obtained from animal studies can be translated to humans also remains to be seen. Physical methods. Physical techniques that use microneedles, tape stripping, ultrasound, microporation or electroporation have also been used to deliver vaccines across the skin. Most of these techniques, although well studied for general drug-delivery applications, have only recently emerged as potential immunization techniques. In the microporation technique, a vaporization process (which involves focused deposition of thermal energy into the skin through an electrically heated element) is used to remove small areas of the stratum corneum, thereby exposing the immunocompetent epidermis. In one study, in hairless mice, application of an adenoviral vector to microporated skin resulted in 10–100-fold greater cellular



REVIEWS

MICROFOLD CELL

(M cell). A specialized type of epithelial cell that delivers antigens from the lumen directly to intraepithelial lymphocytes and to subepithelial lymphoid tissues, using transepithelial vesicular transport. PEYER’S PATCH

A section of the intestinal epithelium that contains microfold cells. These regions form the mucosa-associated lymphoid tissue. LIPOSOME

A lipid vesicle that encapsulates vaccines in a lipid-bilayer membrane and facilitates their delivery.

and humoral immune responses than did application to intact skin70. Microneedles (which are solid and hollow arrays of micrometre-scale silicon projections) have also been used on several occasions to carry out topical immunization with various vaccines71–73. Microprojection arrays have been used to deliver naked plasmid DNA, inducing stronger and less variable immune responses (as judged by serum IgG titres) than those induced by needle-based injections. They also reduced the number of immunizations that was required for full seroconversion73. In another study, microprojection-array patches were used to deliver a model antigen, ovalbumin, to generate a strong immune response71. Ovalbumin that was administered by microprojection array generated an immune response up to 50-fold greater than that observed after the same dose administered subcutaneously or intramuscularly using a needle. A handful of studies have reported the use of tape stripping to facilitate transdermal vaccine absorption in animals74,75. Repeated peeling using tape (for example, Scotch tape) effectively removes the stratum corneum. Application of peptides that represent tumour-derived epitopes to tape-stripped mouse skin primed tumourspecific cytotoxic T cells in the lymph nodes and the spleen, protected mice against a subsequent challenge with the corresponding tumour cells and suppressed the growth of established tumours76. Skin abrasion using a razor and a toothbrush followed by application of adenoviral vectors has yielded promising results in humans77. Ultrasound, at low frequency (20 kHz), has also been shown to deliver a vaccine (consisting of TT) to mice in one study78. The immune response that was generated by ultrasonically delivered vaccine was about tenfold greater per unit dose of vaccine that entered the skin than occurred after subcutaneous injection. (About 1% of the topically applied dose entered the skin78.) Compared with simple topical administration, pretreatment with ultrasound was shown to increase vaccine delivery, thereby allowing enough vaccine to enter the skin to activate the immune response. Ultrasound has been shown to increase skin permeability through disruption of the stratum corneum by acoustic cavitation (which involves the formation and collapse of gaseous cavities)79. Furthermore, application of ultrasound resulted in activation of Langerhans cells, the reasons for which are not clear. In another study, electroporation (which involves the application of highvoltage, short-duration electric pulses) has been used to increase the delivery of DNA vaccines across the skin of mice80. Electroporation has also been shown to induce an effective immune response after transdermal delivery of peptide vaccines81. Additional approaches, including the use of laser-assisted permeabilization of the stratum corneum and various designs of microneedles (other than those published in peer-reviewed literature), are also being pursued by industry. Most of the physical methods for topical immunization have only been tested on animals, specifically on mice or rats. It remains to be seen how many of these methods


can be applied to human skin, which differs substantially in barrier properties from rodent skin. Some of the methods that are discussed in this section are purported to have been tested on humans for immunization purposes; however, these studies have not yet appeared in peer-reviewed literature. Topical vaccine application (including the use of topical adjuvants, colloidal carriers and physical disruption) offers several advantages. Administration to the skin is generally easy to carry out and leads to high compliance of patients82. Topical vaccine application also naturally targets Langerhans cells. However, cost issues must be investigated in depth before these techniques can be adopted for wide-scale applications in humans. Methods that are based on physical techniques such as ultrasound, electroporation and microporation use expensive devices, which might pose constraints on their adoption by developing countries. The use of electric power might limit the widespread use of some of these methods, especially for field applications. Several companies, mostly small-scale businesses, are working to address these challenges. Mucosal administration

Mucosal routes (especially oral and nasal routes) have been used for delivering medication for millennia, predating needles and syringes. Several centuries ago, nasal administration of dried scabs of smallpox lesions and oral administration of fleas from cows with cowpox were practiced in China as a means of immunization against smallpox. It was the Sabin oral polio vaccine, however, that brought mucosal immunization to prominence, in the early 1960s, and that had an important role in the programme for global eradication of polio. Since then, several mucosal vaccines have been marketed TIMELINE. Because many pathogens — for example, HIV and influenza virus — enter the body through mucosal tissues, the development of vaccines that offer mucosal immunity has received considerable attention in the past 20 years12,83. Oral route. Oral delivery of vaccines is an attractive mode of immunization because of its acceptability and its ease of administration84. Orally delivered vaccines, especially particulates, are recognized by MICROFOLD M CELLS (which sample antigen) in the PEYER’S PATCHES of the intestine and by dendritic cells that reside there12 (FIG. 3). At present, few vaccines (those against polio, typhoid fever and cholera) are administered orally, and most of these are based on live attenuated pathogens TIMELINE. Oral delivery of non-living vaccines has proved to be extremely challenging, owing to poor stability of proteins, peptides and DNA in the acidic and enzyme-rich environment of the gastrointestinal tract85. Several strategies, including the use of biodegradable polymeric particles and LIPOSOMES, have been adopted to protect the antigens in the gastrointestinal tract86,87. In addition, strong adjuvants — for example, bacterial enterotoxins such as CT and LT — have also been successfully used for the oral immunization of animals88. However, toxicity of these enterotoxins limits their



REVIEWS

Lumen (intestine or nose) Particulate antigen (liposome, microsphere or bacterial ghost)

Epithelium

Soluble antigen (subunit vaccine)

Epithelial cell

M cell

Dendritic cell

Lymphocyte Subepithelium

Figure 3 | Immunization by mucosal routes. Vaccines that are delivered by mucosal routes (that is, ocular, oral, nasal, pulmonary, vaginal or rectal routes) are recognized by microfold (M) cells and dendritic cells in the mucosaassociated lymphoid tissue. Particulate antigens (such as microspheres, liposomes and bacterial ghosts) are recognized by receptors at the surface of M cells and are presented to lymphocytes and macrophages. Soluble antigens, as well as small pathogens, can permeate the epithelium and be recognized by dendritic cells.

BACTERIAL GHOST

A bacterium, the cytoplasmic contents of which have been removed.

912 | DECEMBER 2005

applications in humans89. To alleviate the toxicity issues, mutants and subunits of LT and CT have been used as adjuvants in many studies of oral immunization of animals90. A detailed review of mucosal adjuvants can be found in REFS 88,89. Encapsulation of antigens in biodegradable polymer microspheres (especially poly(lactide-co-glycolide), PLG) has been successfully used for oral immunization of animals against HBV91, TT92 and other antigens86,93. Several other materials have also been used to encapsulate antigens, but a clear advantage of any particular material is not obvious94–96. In addition to protecting antigens from the hostile gastrointestinal environment, microspheres have been suggested to aid immunization through the sustained release of antigens, which might overcome the need for booster doses, which are typically required for vaccines that are administered by intramuscular injection97. Additional mechanisms, such as direct intracellular delivery of antigens through phagocytosis of particles, have also been proposed to explain the adjuvant activity of polymeric microspheres98. Oral delivery of DNA vaccines that have been encapsulated in PLG or chitosan microspheres has also received considerable attention99–101. Several

| VOLUME 5

strategies that involve the use of antibodies, IgA or lectins have also been proposed to target microspheres to M cells102. Liposomes offer an alternative means of protecting a vaccine103–105. Conventional liposomes are not particularly stable in the gastric environment, so polymerized liposomes have been developed as carriers for oral vaccines106. Modification of liposome composition using lipids from archaebacteria has been attempted to facilitate vaccine uptake by antigen-presenting cells107,108. Both nanoscale lipid particles that consist of lipids, adjuvants and antigens, which are known as immunostimulating complexes (ISCOMs)107, and 109 BACTERIAL GHOSTS , which are bacteria that lack their cytoplasmic contents, have also been successfully used as vaccine carriers in animals. Bacterial ghosts, the surface properties of which resemble those of live bacteria, are highly immunogenic and are therefore strong adjuvants. Despite considerable effort, oral immunization with encapsulated antigens is still limited by several issues that are specific to this route. The effectiveness of oral immunization has been established in several animal studies (mostly in mice), but clinical experience in this field has been mixed105,110–112. Specifically, serum antibody titres after oral delivery of liposome-encapsulated TT or DT to humans were variable and were lower than those observed for animals105. In another clinical study, oral immunization against enterotoxigenic E. coli using a microsphere-encapsulated colonization-factor antigen rendered protection against subsequent challenge in only 30% of patients110. In a more recent study, oral delivery of PLG-encapsulated CS6 antigen from E. coli generated antigen-secreting cells and an IgA response, but the differences between responses that were generated by encapsulated and non-encapsulated antigen were not great112. Attempting to scale up the results of oral immunization from animals to humans is generally problematic. Typically, the doses that are required to elicit an immune response through the oral route are substantially higher (by up to 100-fold) than those that are required when using injection113. This raises the crucial issue of the cost of immunization. Furthermore, oral immunization with non-living vaccines requires the use of carriers and adjuvants, and the safety of exposing the sensitive gastrointestinal tract to these compounds, in addition to the safety of exposure to the vaccine itself, remains to be carefully studied. A completely different solution to this issue has been offered by the use of genetically engineered plants as immunizing agents. This approach has yielded encouraging results in animals114 and humans115,116, but the safety of transgenic-plant vaccines needs to be further evaluated. Nasal route. Intranasal delivery of vaccines using a nasal spray delivered into the nostrils is an attractive mode of immunization. The nose, similar to the mouth, is a practical site for vaccine administration, and nasopharynx-associated lymphoid tissue efficiently induces antigen-specific immune responses in



REVIEWS both mucosal and systemic immune compartments117,118. A detailed review of nasal immunization can be found in REFS 118,119. FluMist (MedImmune Vaccines, Inc.), a live influenza-virus vaccine, has already been approved by the Food and Drug Administration (United States) for intranasal administration. The development of nonliving nasal vaccines has proved to be challenging, but considerable progress has been made in the past decade120. One nasal vaccine, an inactivated influenzavirus product known as Nasalflu (Berna Biotech AG), was introduced in Switzerland in 2000. However, it was withdrawn from the market in 2001 because of the vaccineassociated incidence of Bell’s palsy, which was thought to originate from the use of E. coli LT as an adjuvant121. Efforts to develop other nasal vaccines, however, have continued to make progress. In the past decade, several clinical studies have confirmed the generation of local and systemic immunity after nasal immunization of humans against diphtheria and tetanus122, influenza123 and infection with Streptococcus mutans124. Varying local reactions, ranging from good tolerance to stinging, were reported in response to intranasal administration of vaccine122. A far larger number of studies in mice, pigs and monkeys also confirm the effectiveness of nasal immunization with a variety of vaccines119. Nasal vaccines have been delivered in various physical forms, including aerosolized liquids125, and liposome126 and microsphere formulations127, which can be administered together with various adjuvants88,128,129. Nasal immunization generally requires much lower antigen doses than does oral immunization, owing to lower enzymatic activity in the nasal cavity than in the gastrointestinal tract. Vaginal or rectal route. Vaginal and rectal immunization through topical application of a cream has recently received attention for immunization against sexually transmitted diseases such as HIV/AIDS130. Vaginal immunization with a multicomponent peptide vaccine against HIV infection has been shown to induce local antibody responses in mice when administered together with strong adjuvants such as CTB131,132. DNA-vaccine strategies for preventing HIV infection using vaginal or rectal immunization have also been tested and have been shown to be effective at generating systemic and mucosal immune responses in mice133. Clinical experiments involving vaginal and rectal immunization against cholera, using a vaccine containing both whole cells and CTB, have yielded successful results134. In general, however, vaginal and rectal immunization with non-living vaccines has had limited success, owing to delivery and adjuvanticity issues. Other routes. Additional mucosal routes, including pulmonary, ocular and sublingual administration of vaccines, have also been attempted in several cases. Aerosolized vaccines have been delivered through the pulmonary route, which aims to deliver vaccine at various levels of the bronchial tree, including the alveoli. Pulmonary delivery of vaccines, which targets


bronchus-associated lymphoid tissue, has been effectively used for immunization of humans against measles, using a live attenuated virus135. Animal studies have also shown the effectiveness of non-living pulmonary vaccines, including inactivated influenza virus136. Ocular immunization has been attempted against infection with herpes simplex virus (HSV). This was motivated by the strong need to generate ocular mucosal immunity to HSV, which commonly infects the eye in addition to other sites. Heat-killed HSV, as well as HSV-based subunit vaccines, generated effective mucosal immunity to HSV in pre-clinical animal studies that involved direct administration to the eye in the form of drops137,138. The effectiveness of sublingual immunization has also been shown in several studies in animals139,140. However, ocular and sublingual routes have been less well studied as generalized methods for immunization than have other mucosal routes. A large body of literature has confirmed the merits of all modes of mucosal immunization. Products for mucosal immunization, especially nasal immunization, are under development in several companies. Immunization through mucosal routes (oral, ocular, pulmonary, nasal, vaginal or rectal) generates IgAproducing cells at the site of infection129. However, it should be noted that mucosal administration of antigen is not essential for the generation of antigenspecific IgA-producing plasma cells. Topical application of vaccine has also been shown to generate these cells and to induce mucosal immunity in the host19. Among the methods of mucosal immunization, nasal immunization seems to offer an optimal balance of immunogenicity, dosing and accessibility, as well as patient acceptability. Other mucosal routes are limited as generic modes of immunization. The oral route is limited by the difficulty in accessing M cells in the gastrointestinal tract and by dosing issues. The vaginal and rectal routes are limited by acceptability and immunogenicity issues. The issues of dosing and immunogenicity could be addressed through future research focused on developing strategies for better encapsulation, adjuvanticity and targeting. Fundamental studies focused on the transport of antigens or antigen carriers from the point of administration to the mucosal membrane will also bring new insights to mucosal immunization. These studies should be complemented by research focused on gaining a better understanding of the immunology of mucosal membranes, in particular by identification of specific target cells and receptors for vaccines and by study of the crosstalk between different mucosal compartments141. Conclusions

The shortcomings of injections have led to active research and development of needle-free methods of immunization. The shift from needle-based to needlefree immunization is also catalysed, in part, by the realization that the skin and the mucosal membranes, which cannot be effectively accessed by conventional



REVIEWS needles, are ideal targets for vaccine delivery12,142. Until recently, needle-free methods of immunization were restricted to liquid-jet injection and oral delivery of live attenuated pathogens. Considerable advances have been made in the past decade, especially in transdermal and nasal immunization, but it should be noted that most of the technologies that are discussed here are still at an early stage and lack detailed evaluation in terms of safety, toxicity, reproducibility and economic feasibility. Cost of immunization is an important factor in the acceptability of new methods. According to the WHO, the current cost of administering three doses of the diphtheria–tetanus–pertussis vaccine is US$4–70 per child, depending on the country. Use of needle-free immunization might push this cost even higher, owing to the increased cost of development. The potential higher cost of needle-free immunization needs to be viewed in light of its benefits. Care needs to be taken when carrying out cost–benefit analysis of needle-free

1.

2.

3.

4.

5.

6. 7.

8.

9. 10.

11. 12.

13.

Kermode, M. Unsafe injections in low-income country health settings: need for injection safety promotion to prevent the spread of blood-borne viruses. Health Promot. Int. 19, 95–103 (2004). Nir, Y., Paz, A., Sabo, E. & Potasman, I. Fear of injections in young adults: prevalence and associations. Am. J. Trop. Med. Hyg. 68, 341–344 (2003). Breau, L. M. et al. Facial expression of children receiving immunizations: a principal components analysis of the child facial coding system. Clin. J. Pain 17, 178–186 (2001). Rosenstock, L. Needlestick injuries among healthcare workers. Centers for Disease Control and Prevention [online], (2000). World Health Organization. State of the World’s Vaccines and Immunization (World Health Organization, Geneva, 1996). Miller, M. A. & Pisani, E. The cost of unsafe injections. Bull. World Health Organ. 77, 808–811 (1999). Kane, A. et al. Transmission of hepatitis B, hepatitis C and human immunodeficiency viruses through unsafe injections in the developing world: model-based regional estimates. Bull. World Health Organ. 77, 801–807 (1999). This paper reports alarming estimates of the global and regional prevalence of HBV, HCV and HIV infections that might occur as a result of unsafe injection practices in developing countries. This underscores the motivation for developing needlefree methods of immunization. World Health Organization. Safety of Injections: Global Facts & Figures (World Health Organization, Geneva, 2004). Varmus, H. et al. Grand challenges in global health. Science 302, 398–399 (2003). Levine, M. M. Can needle-free administration of vaccines become the norm in global immunization? Nature Med. 9, 99–103 (2003). O’Hagan, D. T. & Rappuoli, R. Novel approaches to vaccine delivery. Pharm. Res. 21, 1519–1530 (2004). Holmgren, J. & Czerkinsky, C. Mucosal immunity and vaccines. Nature Med. 11, S45–S53 (2005). This Review describes the properties of the mucosal immune system (that is, the mucosa-associated lymphoid tissue) and its mechanisms of antigen presentation. It also discusses advances in the development of mucosal vaccines and provides perspectives on future developments. Weniger, B. G. Needle-free jet injection technology: bibliographic references, device & manufacturer roster, patents list and general/miscellaneous resources. Centers for Disease Control and Prevention [online], (2005). This is an information portal about liquid-jet injectors. A comprehensive list of past and current manufacturers of liquid-jet injectors and of published literature is provided.

914 | DECEMBER 2005

methods, because several benefits of needle-free methods are difficult to quantify. It is hoped that needle-free methods will lower the economic burden that is associated with needle-borne infections143 and will eventually prove to be economically feasible. A comparison of the advantages and limitations of various methods of needle-free immunization TABLE 1 makes it evident that there is no one method that is superior. Each method has advantages that are attractive for immunization. At the same time, all methods have limitations that need to be overcome. Each method is eventually likely to find its niche application, which will depend on the type of vaccine and the site of immunization, and on intellectual-property considerations. Opportunities in needle-free immunization have attracted an array of interdisciplinary researchers and businesses to the field of vaccine development. With the influx of new technologies and talent to this field, needle-free immunization is sure to become a reality.

14. Hingson, R. A. & Figge, F. H. J. A survey of the development of jet injection in parenteral therapy. Curr. Res. Anesth. Analg. 31, 361–366 (1952). 15. Peachman, K. K., Rao, M. & Alving, C. R. Immunization with DNA through the skin. Methods 31, 232–242 (2003). 16. Babiuk, S. et al. Cutaneous vaccination: the skin as an immunologically active tissue and the challenge of antigen delivery. J. Control. Release 66, 199–214 (2000). 17. Bodey, B., Bodey, B. Jr & Kaiser, H. E. Dendritic type, accessory cells within the mammalian thymic microenvironment. Antigen presentation in the dendritic neuro-endocrine–immune cellular network. In Vivo 11, 351–370 (1997). 18. Stoitzner, P. et al. Visualization and characterization of migratory Langerhans cells in murine skin and lymph nodes by antibodies against Langerin/CD207. J. Invest. Dermatol. 120, 266–274 (2003). 19. Gockel, C. M., Bao, S. & Beagley, K. W. Transcutaneous immunization induces mucosal and systemic immunity: a potent method for targeting immunity to the female reproductive tract. Mol. Immunol. 37, 537–544 (2000). 20. Glenn, G. M. et al. Transcutaneous immunization with cholera toxin protects mice against lethal mucosal toxin challenge. J. Immunol. 161, 3211–3214 (1998). 21. Kenney, R. T. et al. Dose sparing with intradermal injection of influenza vaccine. N. Engl. J. Med. 351, 2295–2301 (2004). 22. Ren, S. et al. Low-volume jet injection for intradermal immunization in rabbits. BMC Biotechnol. [online] 2, 10 (2002). 23. Baxter, J. in Fundamental Mechanisms of Drug Delivery by Jet Injection: Basis for the Development of a Painless Microjet Injector. 161 Thesis, Univ. California, Santa Barbara (2004). 24. Weniger, B. G. Jet injection of vaccines: overview and challenges for mass vaccination with jet injections (JIs). United States Department of Health and Human Services [online], (2003). 25. Williams, J. et al. Hepatitis A vaccine administration: comparison between jet-injector and needle injection. Vaccine 18, 1939–1943 (2000). 26. Jackson, L. A. et al. Safety and immunogenicity of varying dosages of trivalent inactivated influenza vaccine administered by needle-free jet injectors. Vaccine 19, 4703–4709 (2001). 27. Mathei, C., Van Damme, P. & Meheus, A. Hepatitis B vaccine administration: comparison between jet-gun and syringe and needle. Vaccine 15, 402–404 (1997). 28. Canter, J. et al. An outbreak of hepatitis-B associated with jet injections in a weight-reduction clinic. Arch. Intern. Med. 150, 1923–1927 (1990). 29. Hoffman, P. N. et al. A model to assess the infection potential of jet injectors used in mass immunisation. Vaccine 19, 4020–4027 (2001).

| VOLUME 5

30. Weniger, B. G. New high-speed jet injectors for mass vaccination: pros and cons of disposable-cartridge jet injectors (DCJIs) versus multi-use-nozzle jet injectors (MUNJIs). World Health Organization [online], (2004). 31. Dimache, G. et al. A clinical, epidemiological and laboratory study on avoiding the risk of transmitting viral hepatitis during vaccinations with the Dermojet protected by an anticontaminant disposable device. Vaccine 15, 1010–1013 (1997). 32. Cooper, J. A., Bromley, L. M., Baranowski, A. P. & Barker, S. G. Evaluation of a needle-free injection system for local anaesthesia prior to venous cannulation. Anaesthesia 55, 247–250 (2000). 33. Epstein, J. E. et al. Safety, tolerability, and lack of antibody responses after administration of a PfCSP DNA malaria vaccine via needle or needle-free jet injection, and comparison of intramuscular and combination intramuscular/intradermal routes. Hum. Gene Ther. 13, 1551–1560 (2002). 34. Haensler, J. et al. Intradermal DNA immunization by using jet-injectors in mice and monkeys. Vaccine 17, 628–638 (1999). 35. Raviprakash, K. et al. Needle-free Biojector injection of a dengue virus type 1 DNA vaccine with human immunostimulatory sequences and the GM-CSF gene increases immunogenicity and protection from virus challenge in Aotus monkeys. Virology 315, 345–352 (2003). 36. Schramm, J. & Mitragotri, S. Transdermal drug delivery by jet injectors: energetics of jet formation and penetration. Pharm. Res. 19, 1673–1679 (2002). 37. Schramm-Baxter, J. & Mitragotri, S. Needle-free jet injections: dependence of jet penetration and dispersion in the skin on jet power. J. Control. Release 97, 527–535 (2004). 38. Kendall, M., Mitchell, T. & Wrighton-Smith, P. Intradermal ballistic delivery of micro-particles into excised human skin for pharmaceutical applications. J. Biomech. 37, 1733–1741 (2004). 39. Chen, D., Maa, Y. F. & Haynes, J. R. Needle-free epidermal powder immunization. Expert Rev. Vaccines 1, 265–276 (2002). 40. Chen, D. et al. Epidermal powder immunization of mice and monkeys with an influenza vaccine. Vaccine 21, 2830–2836 (2003). 41. Chen, D. et al. Serum and mucosal immune responses to an inactivated influenza virus vaccine induced by epidermal powder immunization. J. Virol. 75, 7956–7965 (2001). 42. Chen, D. et al. Adjuvantation of epidermal powder immunization. Vaccine 19, 2908–2917 (2001). 43. Chen, D. et al. Epidermal powder immunization using non-toxic bacterial enterotoxin adjuvants with influenza vaccine augments protective immunity. Vaccine 20, 2671–2679 (2002).



REVIEWS

44. Dean, H. J., Haynes, J. & Schmaljohn, C. The role of particle-mediated DNA vaccines in biodefense preparedness. Adv. Drug Deliv. Rev. 57, 1315–1342 (2005). This is a review of recent developments in particlemediated DNA vaccines. An overview of the principles and applications of EPI with respect to DNA vaccination is provided. 45. Kim, T. W. et al. Enhancing DNA vaccine potency by coadministration of DNA encoding antiapoptotic proteins. J. Clin. Invest. 112, 109–117 (2003). 46. Dean, H. J. & Chen, D. Epidermal powder immunization against influenza. Vaccine 23, 681–686 (2004). 47. Roy, M. J. et al. Induction of antigen-specific CD8+ T cells, T helper cells, and protective levels of antibody in humans by particle-mediated administration of a hepatitis B virus DNA vaccine. Vaccine 19, 764–778 (2000). 48. Tacket, C. O. et al. Phase 1 safety and immune response studies of a DNA vaccine encoding hepatitis B surface antigen delivered by a gene delivery device. Vaccine 17, 2826–2829 (1999). 49. Roberts, L. K. et al. Clinical safety and efficacy of a powdered hepatitis B nucleic acid vaccine delivered to the epidermis by a commercial prototype device. Vaccine 23, 4867–4878 (2005). 50. Kendall, M., Rishworth, S., Carter, F. & Mitchell, T. Effects of relative humidity and ambient temperature on the ballistic delivery of micro-particles to excised porcine skin. J. Invest. Dermatol. 122, 739–746 (2004). 51. Prausnitz, M. R., Mitragotri, S. &. Langer, R. Current status and future potential of transdermal drug delivery. Nature Rev. Drug Discov. 3, 115–124 (2004). This paper reviews transdermal drug delivery. It describes the principles of dermal penetration, its current status and the methods that are under development to increase the permeability of the skin to molecules, including vaccines. These methods have an important role in immunization by topical application. 52. Barquet, N. & Domingo, P. Smallpox: the triumph over the most terrible of the ministers of death. Ann. Intern. Med. 127, 635–642 (1997). 53. Fan, H., Lin, Q., Morrissey, G. R. & Khavari, P. A. Immunization via hair follicles by topical application of naked DNA to normal skin. Nature Biotechnol. 17, 870–872 (1999). 54. Bouwstra, J. et al. New aspects of the skin barrier organization. Skin Pharmacol. Appl. Skin Physiol. 14 (Suppl. 1), 52–62 (2001). 55. Glenn, G. M., Rao, M., Matyas, G. R. & Alving, C. R. Skin immunization made possible by cholera toxin. Nature 391, 851 (1998). 56. Glenn, G. M. et al. Transcutaneous immunization and immunostimulant strategies: capitalizing on the immunocompetence of the skin. Expert Rev. Vaccines 2, 253–267 (2003). 57. Hammond, S. A., Walwender, D., Alving, C. R. & Glenn, G. M. Transcutaneous immunization: T cell responses and boosting of existing immunity. Vaccine 19, 2701–2707 (2001). 58. Belyakov, I. M. et al. Transcutaneous immunization induces mucosal CTLs and protective immunity by migration of primed skin dendritic cells. J. Clin. Invest. 113, 998–1007 (2004). 59. Kenney, R. T. et al. Induction of protective immunity against lethal anthrax challenge with a patch. J. Infect. Dis. 190, 774–782 (2004). 60. Frech, S. A. et al. Improved immune responses to influenza vaccination in the elderly using an immunostimulant patch. Vaccine 23, 946–950 (2005). 61. Scharton-Kersten, T. et al. Transcutaneous immunization with bacterial ADP-ribosylating exotoxins, subunits, and unrelated adjuvants. Infect. Immun. 68, 5306–5313 (2000). 62. Scharton-Kersten, T. et al. Principles of transcutaneous immunization using cholera toxin as an adjuvant. Vaccine 17, S37–S43 (1999). 63. Guerena-Burgueno, F. et al. Safety and immunogenicity of a prototype enterotoxigenic Escherichia coli vaccine administered transcutaneously. Infect. Immun. 70, 1874–1880 (2002). 64. Glenn, G. M. et al. Transcutaneous immunization: a human vaccine delivery strategy using a patch. Nature Med. 6, 1403–1406 (2000). This is one of the pioneering reports of immunization by topical application of vaccines. It describes a clinical study of adjuvant-patch-mediated cutaneous immunization. 65. Gupta, P. N. et al. Tetanus toxoid-loaded transfersomes for topical immunization. J. Pharm. Pharmacol. 57, 295–301 (2005).

66. Baca-Estrada, M. E., Foldvari, M., Babiuk, S. L. & Babiuk, L. A. Vaccine delivery: lipid-based delivery systems. J. Biotechnol. 83, 91–104 (2000). 67. Cui, Z. et al. Novel ethanol-in-fluorocarbon microemulsions for topical genetic immunization. Pharm. Res. 20, 16–23 (2003). 68. Cui, Z. & Mumper, R. J. Topical immunization using nanoengineered genetic vaccines. J. Control. Release 81, 173–184 (2002). 69. Cui, Z. & Mumper, R. J. Chitosan-based nanoparticles for topical genetic immunization. J. Control. Release 75, 409–419 (2001). 70. Bramson, J. et al. Enabling topical immunization via microporation: a novel method for pain-free and needlefree delivery of adenovirus-based vaccines. Gene Ther. 10, 251–260 (2003). 71. Matriano, J. A. et al. Macroflux microprojection array patch technology: a new and efficient approach for intracutaneous immunization. Pharm. Res. 19, 63–70 (2002). 72. Prausnitz, M. R. Microneedles for transdermal drug delivery. Adv. Drug Deliv. Rev. 56, 581–587 (2004). 73. Mikszta, J. A. et al. Improved genetic immunization via micromechanical disruption of skin-barrier function and targeted epidermal delivery. Nature Med. 8, 415–419 (2002). 74. Kahlon, R. et al. Optimization of epicutaneous immunization for the induction of CTL. Vaccine 21, 2890–2899 (2003). 75. Takigawa, M. et al. Percutaneous peptide immunization via corneum barrier-disrupted murine skin for experimental tumor immunoprophylaxis. Ann. NY Acad. Sci. 941, 139–146 (2001). 76. Seo, N. et al. Percutaneous peptide immunization via corneum barrier-disrupted murine skin for experimental tumor immunoprophylaxis. Proc. Natl Acad. Sci. USA 97, 371–376 (2000). 77. Van Kampen, K. R. et al. Safety and immunogenicity of adenovirus-vectored nasal and epicutaneous influenza vaccines in humans. Vaccine 23, 1029–1036 (2005). 78. Tezel, A., Paliwal, S., Shen, Z. & Mitragotri, S. Low-frequency ultrasound as a transcutaneous immunization adjuvant. Vaccine 23, 3800–3807 (2005). 79. Tezel, A. & Mitragotri, S. Interactions of inertial cavitation bubbles with stratum corneum lipid bilayers during low-frequency sonophoresis. Biophys. J. 85, 3502–3512 (2003). 80. Zhang, L., Nolan, E., Kreitschitz, S. & Rabussay, D. P. Enhanced delivery of naked DNA to the skin by noninvasive in vivo electroporation. Biochim. Biophys. Acta 1572, 1–9 (2002). 81. Misra, A., Ganga, S. & Upadhyay, P. Needle-free, non-adjuvanted skin immunization by electroporationenhanced transdermal delivery of diphtheria toxoid and a candidate peptide vaccine against hepatitis B virus. Vaccine 18, 517–523 (1999). 82. Archer, D. F. et al. Assessment of compliance with a weekly contraceptive patch (Ortho Evra/Evra) among North American women. Fertil. Steril. 77, S27–S31 (2002). 83. Yuki, Y. & Kiyono, H. New generation of mucosal adjuvants for the induction of protective immunity. Rev. Med. Virol. 13, 293–310 (2003). 84. Fooks, A. R. Development of oral vaccines for human use. Curr. Opin. Mol. Ther. 2, 80–86 (2000). 85. Brown, W. R. Enteric immunization: promises and challenges. Dig. Dis. 14, 192–200 (1996). 86. Lambkin, I. & Pinilla, C. Targeting approaches to oral drug delivery. Expert Opin. Biol. Ther. 2, 67–73 (2002). 87. Kersten, G. & Hirschberg, H. Antigen delivery systems. Expert Rev. Vaccines 3, 453–462 (2004). 88. Freytag, L. C. & Clements, J. D. Mucosal adjuvants. Vaccine 23, 1804–1813 (2005). 89. Eriksson, K. & Holmgren, J. Recent advances in mucosal vaccines and adjuvants. Curr. Opin. Immunol. 14, 666–672 (2002). 90. Pizza, M. et al. Mucosal vaccines: non toxic derivatives of LT and CT as mucosal adjuvants. Vaccine 19, 2534–2541 (2001). 91. Nellore, R. V., Pande, P. G., Young, D. & Bhagat, H. R. Evaluation of biodegradable microspheres as vaccine adjuvant for hepatitis B surface antigen. J. Parenter. Sci. Technol. 46, 176–180 (1992). 92. Esparza, I. & Kissel, T. Parameters affecting the immunogenicity of microencapsulated tetanus toxoid. Vaccine 10, 714–720 (1992). 93. Challacombe, S. J. et al. Enhanced secretory IgA and systemic IgG antibody responses after oral immunization with biodegradable microparticles containing antigen. Immunology 76, 164–168 (1992).


94. Stertman, L., Strindeliu, L. & Sjoholm, I. Starch microparticles as an adjuvant in immunisation: effect of route of administration on the immune response in mice. Vaccine 22, 2863–2872 (2004). 95. Ren, J. M. et al. PELA microspheres loaded H. pylori lysates and their mucosal immune response. World J. Gastroenterol. 8, 1098–1102 (2002). 96. Wikingsson, L. & Sjoholm, I. Polyacryl starch microparticles as adjuvant in oral immunisation, inducing mucosal and systemic immune responses in mice. Vaccine 20, 3355–3363 (2002). 97. Preis, I. & Langer, R. S. A single-step immunization by sustained antigen release. J. Immunol. Methods 28, 193–197 (1979). 98. Langer, R., Cleland, J. L. & Hanes, J. New advances in microsphere-based single-dose vaccines. Adv. Drug Deliv. Rev. 28, 97–119 (1997). 99. Jones, D. H. et al. Poly(DL-lactide-co-glycolide)encapsulated plasmid DNA elicits systemic and mucosal antibody responses to encoded protein after oral administration. Vaccine 15, 814–817 (1997). 100. Chew, J. L. et al. Chitosan nanoparticles containing plasmid DNA encoding house dust mite allergen, Der p 1 for oral vaccination in mice. Vaccine 21, 2720–2729 (2003). 101. Roy, K., Mao, H. Q., Huang, S. K. & Leong, K. W. Oral gene delivery with chitosan–DNA nanoparticles generates immunologic protection in a murine model of peanut allergy. Nature Med. 5, 387–391 (1999). 102. Ermak, T. H. & Giannasca, P. J. Microparticle targeting to M cells. Adv. Drug Deliv. Rev. 34, 261–283 (1998). 103. Perrie, Y., Obrenovic, M., McCarthy, D. & Gregoriadis, G. Liposome (Lipodine)-mediated DNA vaccination by the oral route. J. Liposome Res. 12, 185–197 (2002). 104. Chen, H. & Langer, R. Oral particulate delivery: status and future trends. Adv. Drug Deliv. Rev. 34, 339–350 (1998). 105. Mirchamsy, H. et al. Stimulating role of toxoids-laden liposomes in oral immunization against diphtheria and tetanus infections. Biologicals 24, 343–350 (1996). 106. Chen, H., Torchilin, V. & Langer, R. Lectin-bearing polymerized liposomes as potential oral vaccine carriers. Pharm. Res. 13, 1378–1383 (1996). 107. Kersten, G. F. & Crommelin, D. J. Liposomes and ISCOMs. Vaccine 21, 915–920 (2003). 108. Patel, G. B., Omri, A., Deschatelets, L. & Sprott, G. D. Safety of archaeosome adjuvants evaluated in a mouse model. J. Liposome Res. 12, 353–372 (2002). 109. Jalava, K., Eko, F. O., Riedmann, E. & Lubitz, W. Bacterial ghosts as carrier and targeting systems for mucosal antigen delivery. Expert Rev. Vaccines 2, 45–51 (2003). 110. Tacket, C. O. et al. Enteral immunization and challenge of volunteers given enterotoxigenic E. coli CFA/II encapsulated in biodegradable microspheres. Vaccine 12, 1270–1274 (1994). 111. Lambert, J. S. et al. A Phase I safety and immunogenicity trial of UBI microparticulate monovalent HIV-1 MN oral peptide immunogen with parenteral boost in HIV-1 seronegative human subjects. Vaccine 19, 3033–3042 (2001). 112. Katz, D. E. et al. Oral immunization of adult volunteers with microencapsulated enterotoxigenic Escherichia coli (ETEC) CS6 antigen. Vaccine 21, 341–346 (2003). 113. Brayden, D. J. Oral vaccination in man using antigens in particles: current status. Eur. J. Pharm. Sci. 14, 183–189 (2001). This paper provides a critical review of the field of oral vaccine delivery. It discusses issues that are associated with the scaling up of animal results to humans. 114. Haq, T. A., Mason, H. S., Clements, J. D. & Arntzen, C. J. Oral immunization with a recombinant bacterial antigen produced in transgenic plants. Science 268, 714–716 (1995). 115. Tacket, C. O. et al. Immunogenicity in humans of a recombinant bacterial antigen delivered in a transgenic potato. Nature Med. 4, 607–609 (1998). This paper reports on studies that involved the immunization of humans with transgenic vegetables. Potatoes expressing LT were used to immunize human volunteers, and the generation of neutralizing antibodies was confirmed. 116. Thanavala, Y. et al. Immunogenicity in humans of an edible vaccine for hepatitis B. Proc. Natl Acad. Sci. USA 102, 3378–3382 (2005). 117. Kiyono, H. & Fukuyama, S. NALT- versus Peyer’s-patchmediated mucosal immunity. Nature Rev. Immunol. 4, 699–710 (2004). 118. Vajdy, M. & O’Hagan, D. T. Microparticles for intranasal immunization. Adv. Drug Deliv. Rev. 51, 127–141 (2001).



REVIEWS

119. Davis, S. S. Nasal vaccines. Adv. Drug Deliv. Rev. 51, 21–42 (2001). This paper provides an overview of nasal immunization. The structure and function of nasopharynx-associated lymphoid tissue and its role in nasal immunization are discussed. 120. Haneberg, B. & Holst, J. Can nonliving nasal vaccines be made to work? Expert Rev. Vaccines 1, 227–232 (2002). 121. Mutsch, M. et al. Use of the inactivated intranasal influenza vaccine and the risk of Bell’s palsy in Switzerland. N. Engl. J. Med. 350, 896–903 (2004). 122. Aggerbeck, H., Gizurarson, S., Wantzin, J. & Heron, I. Intranasal booster vaccination against diphtheria and tetanus in man. Vaccine 15, 307–316 (1997). 123. Gluck, U., Gebbers, J. O. & Gluck, R. Phase 1 evaluation of intranasal virosomal influenza vaccine with and without Escherichia coli heat-labile toxin in adult volunteers. J. Virol. 73, 7780–7786 (1999). 124. Li, F. et al. Intranasal immunization of humans with Streptococcus mutans antigens. Oral Microbiol. Immunol. 18, 271–277 (2003). 125. Roth, Y., Chapnik, J. S. & Cole, P. Feasibility of aerosol vaccination in humans. Ann. Otol. Rhinol. Laryngol. 112, 264–270 (2003). 126. de Jonge, M. I. et al. Intranasal immunisation of mice with liposomes containing recombinant meningococcal OpaB and OpaJ proteins. Vaccine 22, 4021–4028 (2004). 127. Alpar, H. O., Somavarapu, S., Atuah, K. N. & Bramwell, V. W. Biodegradable mucoadhesive particulates for nasal and pulmonary antigen and DNA delivery. Adv. Drug Deliv. Rev. 57, 411–430 (2005). 128. Singh, M. & O’Hagan, D. T. Recent advances in vaccine adjuvants. Pharm. Res. 19, 715–728 (2002). 129. Vajdy, M. et al. Mucosal adjuvants and delivery systems for protein-, DNA- and RNA-based vaccines. Immunol. Cell Biol. 82, 617–627 (2004). 130. Stevceva, L. & Strober, W. Mucosal HIV vaccines: where are we now? Curr. HIV Res. 2, 1–10 (2004). 131. Russell, M. W. Immunization for protection of the reproductive tract: a review. Am. J. Reprod. Immunol. 47, 265–268 (2002). 132. Kato, H. et al. Rectal and vaginal immunization with a macromolecular multicomponent peptide vaccine candidate for HIV-1 infection induces HIV-specific protective immune responses. Vaccine 18, 1151–1160 (2000). 133. Hamajima, K. et al. Systemic and mucosal immune responses in mice after rectal and vaginal immunization with HIV-DNA vaccine. Clin. Immunol. 102, 12–18 (2002). 134. Wassen, L. et al. Local intravaginal vaccination of the female genital tract. Scand. J. Immunol. 44, 408–414 (1996).

916 | DECEMBER 2005

135. Dilraj, A. et al. Response to different measles vaccine strains given by aerosol and subcutaneous routes to schoolchildren: a randomised trial. Lancet 355, 798–803 (2000). 136. Smith, D. J., Bot, S., Dellamary, L. & Bot, A. Evaluation of novel aerosol formulations designed for mucosal vaccination against influenza virus. Vaccine 21, 2805–2812 (2003). 137. Narang, H. K. Efficacy of herpes vaccine and acyclovir (ACV) in a rabbit model following intraocular inoculation of herpes simplex virus. J. Chemother. 7, 210–215 (1995). 138. Nesburn, A. B. et al. Therapeutic periocular vaccination with a subunit vaccine induces higher levels of herpes simplex virus-specific tear secretory immunoglobulin A than systemic vaccination and provides protection against recurrent spontaneous ocular shedding of virus in latently infected rabbits. Virology 252, 200–209 (1998). 139. BenMohamed, L. et al. Systemic immune responses induced by mucosal administration of lipopeptides without adjuvant. Eur. J. Immunol. 32, 2274–2281 (2002). 140. Montgomery, P. C. & Rafferty, D. E. Induction of secretory and serum antibody responses following oral administration of antigen with bioadhesive degradable starch microparticles. Oral Microbiol. Immunol. 13, 139–149 (1998). 141. Bouvet, J. P., Decroix, N. & Pamonsinlapatham, P. Stimulation of local antibody production: parenteral or mucosal vaccination? Trends Immunol. 23, 209–213 (2002). 142. Kupper, T. S. & Fuhlbrigge, R. C. Immune surveillance in the skin: mechanisms and clinical consequences. Nature Rev. Immunol. 4, 211–222 (2004). This is a review of the immune function of the skin, which has an important role in immunization by topical application. Interactions between the innate and adaptive immune systems in the skin and their role in immune surveillance are discussed. 143. Ekwueme, D. U., Weniger, B. G. & Chen, R. T. Model-based estimates of risks of disease transmission and economic costs of seven injection devices in subSaharan Africa. Bull. World Health Organ. 80, 859–870 (2002). 144. Alexander, L. N. et al. Vaccine policy changes and epidemiology of poliomyelitis in the United States. JAMA 292, 1696–1701 (2004). 145. Centers for Disease Control and Prevention. Suspension of rotavirus vaccine after reports of intussusception — United States, 1999. MMWR Morb. Mortal. Wkly Rep. 53, 786–789 (2004). 146. O’Hagan, D. T. & Rappuoli, R. The safety of vaccines. Drug Discov. Today 9, 846–854 (2004). 147. O’Hagan, D. T. Recent developments in vaccine delivery systems. Curr. Drug Targets Infect. Disord. 1, 273–286 (2001).

| VOLUME 5

148. Bourlais, C. L. et al. Ophthalmic drug delivery systems — recent advances. Prog. Retin. Eye Res. 17, 33–58 (1998). 149. Davis, J. L., Gilger, B. C. & Robinson, M. R. Novel approaches to ocular drug delivery. Curr. Opin. Mol. Ther. 6, 195–205 (2004). 150. Burkoth, T. L. et al. Transdermal and transmucosal powdered drug delivery. Crit. Rev. Ther. Drug Carrier Syst. 16, 331–384 (1999). 151. Vidgren, M. T. & Kublik, H. Nasal delivery systems and their effect on deposition and absorption. Adv. Drug Deliv. Rev. 29, 157–177 (1998). 152. Hussain, A. A. Intranasal drug delivery. Adv. Drug Deliv. Rev. 29, 39–49 (1998). 153. Edwards, D. A. & Dunbar, C. Bioengineering of therapeutic aerosols. Annu. Rev. Biomed. Eng. 4, 93–107 (2002). 154. Sastry, S. V., Nyshadham, J. R. & Fix, J. A. Recent technological advances in oral drug delivery — a review. Pharm. Sci. Technol. Today 3, 138–145 (2000). 155. Hussain, A. & Ahsan, F. The vagina as a route for systemic drug delivery. J. Control. Release 103, 301–313 (2005). 156. Mitragotri, S. & Kost, J. Low-frequency sonophoresis: a review. Adv. Drug Deliv. Rev. 56, 589–601 (2004). 157. Denet, A. R., Vanbever, R. & Preat, V. Skin electroporation for transdermal and topical delivery. Adv. Drug Deliv. Rev. 56, 659–674 (2004). 158. Cevc, G. Lipid vesicles and other colloids as drug carriers on the skin. Adv. Drug Deliv. Rev. 56, 675–711 (2004).

Acknowledgements The author acknowledges support from The Whitaker Foundation (United States) and the National Institutes of Health (United States). The author thanks P. Karande, J. Champion, A. Jain and S. Paliwal for assistance during manuscript preparation.

Competing interests statement The author declares competing financial interests: see web version for details

Online links DATABASES The following terms in this article are linked online to: Infectious disease information: http://www.cdc.gov/ncidod/diseases/index.htm cholera | diphtheria | HBV | HCV | HIV | influenza | measles | mumps | pertussis | polio | rotavirus | rubella | tetanus | tuberculosis | typhoid fever | varicella | yellow fever FURTHER INFORMATION Samir Mitragotri’s laboratory: http://drugdelivery.engr.ucsb.edu/ National Immunization Program: http://www.cdc.gov/nip/ World Health Organization: http://www.who.int/en/ Access to this interactive links box is free online.



REVIEWS

IMMUNOPATHOGENESIS OF CORONAVIRUS INFECTIONS: IMPLICATIONS FOR SARS Stanley Perlman*‡§ and Ajai A. Dandekar* Abstract | At the end of 2002, the first cases of severe acute respiratory syndrome (SARS) were reported, and in the following year, SARS resulted in considerable mortality and morbidity worldwide. SARS is caused by a novel species of coronavirus (SARS-CoV) and is the most severe coronavirus-mediated human disease that has been described so far. On the basis of similarities with other coronavirus infections, SARS might, in part, be immune mediated. As discussed in this Review, studies of animals that are infected with other coronaviruses indicate that excessive and sometimes dysregulated responses by macrophages and other pro-inflammatory cells might be particularly important in the pathogenesis of disease that is caused by infection with these viruses. It is hoped that lessons from such studies will help us to understand more about the pathogenesis of SARS in humans and to prevent or control outbreaks of SARS in the future.

*Interdisciplinary Program in Immunology, and Departments of Pediatrics ‡ and Microbiology §, University of Iowa, Iowa City, Iowa 52242, USA. Correspondence to S.P. e-mail: [email protected] doi:10.1038/nri1732

Viral infection of mammals results in certain typical responses by the host immune system. These responses are initiated by the innate immune system, which recognizes ‘molecular patterns’ (such as double-stranded RNA) that are unique to pathogens. The adaptive immune system — which consists of T cells that can kill virus-infected cells and B cells that produce pathogen-specific antibodies — then proceeds to mount a response. Initiation of the adaptive and/or innate immune response results in the production of chemokines and other cytokines that induce a pro-inflammatory response and attract cells, such as neutrophils and macrophages, to sites of infection. These cells, in turn, might release cytotoxic substances, such as matrix metalloproteinases. Although these responses are crucial to clear the infection, all of these processes can cause damage to normal host tissues. Indeed, ‘side-effects’ of the immune response account for many of the signs and symptoms in human infections: for example, during infection with hepatitis B virus, hepatitis C


virus, measles virus or respiratory syncytial virus1–3. Consequently, a ‘normal’ immune response often results in a transient disequilibrium of tissue homeostasis, and this is required for clearance of an infection but can contribute to disease. In this Review, we consider any immune response that results in an increase in clinical disease or tissue destruction to be immunopathological. In many cases, immunopathogenesis is the outcome of immune dysregulation rather than of a normal response TABLE 1. This could occur in one of three ways. First, viral infection might result in an intense inflammatory response that compromises physiological function or results in excessive destruction of host tissue. In this situation, viral infection might interfere with the normal feedback mechanisms that control inflammation, and proinflammatory chemokines or other cytokines might be produced in large amounts or for an excessive period. For example, induction of expression of the pro-inflammatory cytokine interleukin-6 (IL-6) is a



REVIEWS

Table 1 | Mechanisms of immunopathogenesis Mechanism

Description

Coronavirus example

References

Inflammatory storm

Excessive host response to pathogen occurs, resulting in either increased severity of localized disease or systemic disease; excessive response might be non-specific or induced by specific viral proteins; and manifestations might also occur as part of the ‘normal’ immune response required for viral clearance

MHV, FIPV, SARS-CoV

Bystander activation

T cells that are not specific for the pathogen or any host protein that is expressed at the site of inflammation are nevertheless activated (possibly by cytokines), resulting in increased tissue damage

MHV

74,111

Molecular mimicry

Pathogen and host share B- or T-cell epitopes, resulting in an autoimmune reaction in the host tissue that expresses the protein

None

–

Epitope spreading

Ongoing inflammation leads to presentation of selfepitopes, resulting in an autoimmune reaction in the host tissue that expresses the protein

MHV

69

Antibodydependent enhancement

Antibodies specific for cell-surface glycoproteins increase virus uptake by macrophages, through cell-surface Fc receptors, resulting in disease enhancement

FIPV, possibly SARS-CoV

7,8,40,59

40,99,105

Fc receptor, receptor for immunoglobulin; FIPV, feline infectious peritonitis virus; MHV, murine hepatitis virus; SARS-CoV, severeacute-respiratory-syndrome coronavirus.

HAEMOPHAGOCYTOSIS

The phagocytosis of erythrocytes that results from excessive activation of macrophages. This is usually a consequence of uncontrolled activation and proliferation of T cells. MOLECULAR MIMICRY

A mechanism for the induction of autoimmunity in which a pathogen expresses a protein or peptide that is similar to a self-protein. After the induction of a pathogenspecific immune response, a crossreactive response to self results in autoimmune pathology. EPITOPE SPREADING

The de novo activation of autoreactive T cells by selfantigens that have been released after virus-specific T- or B-cellmediated bystander damage.

918 | DECEMBER 2005

consequence of activation of p38 mitogen-activated protein kinase (p38MAPK) by the murine coronavirus, murine hepatitis virus (MHV)4. Excessive production of pro-inflammatory mediators might then result in an unchecked influx of pro-inflammatory cells to the site of infection. Several of these types of cell, most notably neutrophils and macrophages, contribute to inflammation by producing toxic agents, such as reactive oxygen species, that kill both infected and normal cells at sites of infection, which would further exacerbate the response and result in immunopathological changes such as HAEMOPHAGOCYTOSIS5 . Several of the released pro-inflammatory cytokines, such as tumour-necrosis factor (TNF), also induce apoptosis, which would result in increased tissue destruction. In addition, activated T cells that are not specific for the infecting virus or host antigens at the site of infection could traffic to sites of inflammation and contribute to tissue destruction, presumably through the production of chemokines or other cytokines. This has been shown for MHV-infected mice and is known as bystander activation TABLE 1. Second, direct infection of immune cells by a virus might cause increased or dysregulated production of immune mediators distinct from the aberrant production of chemokines and other cytokines previously discussed. For example, infection of mice with MHV strain 3 (MHV-3) results in production of the procoagulant prothrombinase by macrophages, leading to fulminant hepatitis and death6 (discussed in detail later). Third, adaptive immune responses might become directed against host epitopes, and this would result in autoimmune reactions. Pathogen-specific antibodies or T cells might also recognize a host protein or epitope (through a process known as MOLECULAR MIMICRY). In other cases, prolonged infection and the ensuing tissue destruction might result in presentation

| VOLUME 5

of host-protein-derived T- or B-cell epitopes that were previously cryptic (through a process known as EPITOPE SPREADING). The response to these epitopes might prolong the inflammatory response, with consequent tissue destruction, even after virus has been cleared. These autoimmune responses would be limited to a certain tissue or cell type by the specificity of the immune cells involved. Several such mechanisms of autoimmune immunopathogenesis have been described for models of both coronavirus infection and non-coronavirus infection; these are described in TABLE 1. Antibodies might also contribute to immunopathogenesis. With regard to coronaviruses, virus-specific antibody increases the uptake of several viruses by macrophages — including the feline coronavirus feline infectious peritonitis virus (FIPV) — resulting in activation of these macrophages and secretion of chemokines and other cytokines. In several animal models of coronavirus infection, the immune system contributes considerably to the disease process — indeed, the pathology that is seen in several models is wholly, or at least partly, immune mediated. In this Review, we discuss the role of the immune system in the pathology that is seen in animals with coronavirus infections as a window onto the pathological processes that occur in humans infected with the severe acute respiratory syndrome (SARS) coronavirus (SARS-CoV). We focus on murine and feline coronavirus infections, because an immunopathological role in disease has been most clearly documented in these settings. SARS: a severe human coronavirus infection

In the winter of 2002–2003, SARS emerged in China and subsequently spread throughout the world. In the nine months between November 2002 and July 2003, 8,437 cases of this new disease, resulting



REVIEWS

ENZOOTIC VIRUS

A virus that infects animals and is endemic to a geographical locale, with only minimal changes in its incidence over time.

in 813 deaths, were reported to the World Health Organization by 29 countries. The aetiological agent of this disease, a coronavirus, was identified by several research groups in April 2003, and the virus has been named SARS-CoV after the syndrome it causes7,8. Coronaviruses are the largest of all of the RNA viruses, and they have a positive-sense, singlestranded RNA genome of 30–32 kilobases9 (FIG. 1). As a family, the Coronaviridae have a broad host range and cause a wide variety of gastrointestinal, respiratory and systemic diseases in animals, including infectious bronchitis in birds, a fatal disease with multi-organ involvement in felines, and enteritis in pigs, cows, turkeys and dogs. In humans, coronaviruses cause respiratory disease and, to a lesser extent, gastroenteritis. The viruses human coronavirus OC43 (HCoV-OC43) and HCoV-229E have long been known to be causative agents of the common cold, and the more recently identified agents HCoV-HKU1 and HCoV-NL63 cause more severe, although rarely fatal, infections of the upper and lower respiratory tract10–12. SARS-CoV, which causes a much more severe respiratory disease, seems to be an ENZOOTIC VIRUS in Southeast Asia. Several species that might be infected, such as masked palm civets (Paguma larvata), are consumed as food in parts of China, and the ‘wet markets’, at which live animals are bought and sold, are likely venues for the initial crossover event to humans8. The 2002–2003 outbreak of SARS

a SARS-CoV genome 8a 5′

ORF1a

S

3b M 3a E

ORF1b

7a

N

6

3′

7b 8b 9b

b SARS-CoV virion

Spike ssRNA

Nucleocapsid Matrix

Envelope

Figure 1 | The severe-acute-respiratory-syndrome coronavirus genome and virion. a | The severe-acute-respiratory-syndrome coronavirus (SARS-CoV) genome consists of 28 putative open reading frames (ORFs) in 9 mRNA transcripts. ORF1a and ORF1b, which account for about two-thirds of the genome, both encode large polyproteins. ORF1b protein is produced by a –1-base-pair ribosomal frameshift from the reading frame of ORF1a. The SARS-CoV genome encodes four structural proteins: spike (S), envelope (E), matrix (M) and nucleocapsid (N). In non-human isolates, transcription of ORF8a and ORF8b produces a single protein. b | A schematic representation of a SARS-CoV virion is shown. ssRNA, single-stranded RNA.


in humans probably resulted from an interspecific transfer of the virus by aerosols from live, exotic animals that were infected with SARS-CoV to workers in these wet markets. Sera from masked palm civets, raccoon dogs (Nyctereutes procyonoides) and Chinese ferret-badgers (Melogale moschata) were shown to contain neutralizing antibodies that were specific for SARS-CoV, and virus that was nearly identical to the strains that were isolated from infected humans was detected in masked palm civets13–15. However, a specific animal reservoir for the virus has yet to be definitively identified. Infection of humans with SARS-CoV typically causes an influenza-like syndrome of malaise, rigors, fatigue and high fevers. In two-thirds of infected patients, the disease progresses to an atypical pneumonia, with shortness of breath and poor oxygen exchange in the alveoli. Many of these patients also develop watery diarrhoea with active virus shedding, which might increase the transmissibility of the virus. Respiratory insufficiency leading to respiratory failure is the most common cause of death among those infected with SARS-CoV8 . Consistent with these clinical observations, the host cell-surface receptor for SARS-CoV, angiotensin-converting enzyme 2 (ACE2), is detected in the lungs and gastrointestinal tract16,17. Because it binds ACE2, SARS-CoV might contribute to lung-tissue injury by a novel mechanism. ACE2 has a protective role during acute lung injury. By binding ACE2, SARS-CoV leads to the downregulation of ACE2 expression and might therefore negate the protective effect of ACE2. This mechanism of injury was shown in a mouse model using SARS-CoV spike glycoprotein, through which the virus binds ACE2, but it has not yet been confirmed in humans or animals that are infected with SARS-CoV18,19. Severe cases of SARS are associated with lymphopaenia, neutrophilia, mild thrombocytopaenia and coagulation defects20. Haemophagocytosis, which is indicative of cytokine dysregulation, is also detected in some patients with severe disease21,22. Damage to the lungs of patients who are infected with SARS-CoV seems to occur directly, by viral destruction of alveolar and bronchial epithelial cells and macrophages, as well as indirectly, through production of immune mediators, although the exact role of these direct and indirect mechanisms remains controversial. Viral load, as determined from titres in nasopharyngeal aspirate, diminishes 10–15 days after the onset of symptoms, even though clinical disease and alveolar damage worsen, indicating that the host immune response is responsible for some of the pathology in SARS-CoV-infected patients23,24. However, nasopharyngeal viral titres do not necessarily reflect viral loads in the lungs, and high concentrations of virus have been detected in several organs at autopsy, including the lungs, intestine, kidneys and brain25–27. Infection of macrophages and lymphocytes is likely to be a key component in SARS-CoV-induced pathogenesis. SARS-CoV directly infects T cells, contributing



REVIEWS

ABORTIVE INFECTION

An infection in which viral replication is initiated but no infectious virus is produced. SEROSITIS

Inflammation of the membranes that line the lungs (the pleura), the heart (the pericardium), and the abdomen (the peritoneum) and the organs within. PYOGRANULOMATOUS VASCULITIS

A type of vasculitis (that is, inflammation of the blood vessels) that is associated with a chronic inflammatory process in which neutrophils are mixed with components of granulomas. ASCITIC FLUID

Serous fluid that accumulates in the abdominal cavity. It might result from serositis.

920 | DECEMBER 2005

to lymphopaenia and to atrophy of the spleen and lymphoid tissue25. The murine coronavirus MHV-3 also infects and destroys lymphocytes, thereby facilitating viral replication and persistence28. SARS-CoV-infected lymphocytes, similar to FIPV-infected macrophages in domestic cats, might transport the virus to distant organs, resulting in systemic infection25. Macrophages, both infected and uninfected, are detected in large numbers in the lungs of patients who died as a result of SARS22,25,29. Although infected macrophages have been found in vivo, SARS-CoV causes an ABORTIVE INFECTION of these cells in vitro30–32. SARS-CoV also interferes with the initiation of the innate immune response by inhibiting the expression of type I interferons (IFNs) by infected cells, including human monocyte-derived dendritic cells (DCs) and macrophages. IFN production requires the phosphorylation and dimerization of a constitutively expressed protein, IFN-regulatory factor 3 (IRF3). IRF3 is not activated, at least in vitro, after infection with SARS-CoV 30–34 . By contrast, expression of chemokines such as CXC-chemokine ligand 10 (CXCL10), CC-chemokine ligand 2 (CCL2), CCL3, CCL5 and CCL8 is upregulated by abortively infected DCs and macrophages and might contribute to the influx of monocytes and/or macrophages that is observed in infected tissues30,34; CXCL10 and CCL2 expression are also upregulated in the blood of patients with SARS35. Expression of CXCL8 (also known as IL-8), which is an attractant for neutrophils, is also upregulated in the serum of patients with SARS35–37. Consistent with a role for CXCL8 in pathogenesis, severe disease is associated with an increase in the number of neutrophils in the blood20,38. Although these studies indicate that upregulation of expression of pro-inflammatory molecules contributes to the pathogenesis of SARS, increased serum concentrations of two anti-inflammatory molecules — transforming growth factor-β and prostaglandin E2 — were detected in another study 39. Increased concentrations of these anti-inflammatory molecules might impair clearance of virus. The mechanism by which chemokines and other cytokines are regulated in infected patients is not known, but SARS-CoV induces the activation of p38MAPK in monocytes39. The murine coronavirus, MHV, also induces p38MAPK activation, and inhibition of p38MAPK activation results in decreased production of infectious MHV4. Induction of p38MAPK expression by MHV results in production of IL-6 and phosphorylation of eukaryotic translation-initiation factor 4E (EIF4E), which increases cap-dependent (including MHV-specific) protein production. By analogy, it is possible that increased cytokine production by the monocytes of patients with SARS might be an untoward consequence of a mechanism that is used by the virus to increase replication in some cells4. Together, these findings are consistent with models in which the immune system contributes to the SARS disease process. Understanding the immunopathogenesis of SARS could provide new insights into effective treatments for this illness.

| VOLUME 5

Feline coronavirus infections

Macrophage infection and dysfunction. Feline enteric coronaviruses (FECVs) generally cause mild or asymptomatic infections, mainly of domestic cats, although cases have been reported in wild feline populations40. Domestic cats often become persistently infected with FECV. Occasionally, these animals develop the uniformly fatal disease infectious peritonitis, which is caused by a macrophage-tropic variant of FECV that is known as FIPV. Indeed, virulent strains of FIPV replicate more efficiently in feline peritoneal macrophages in vitro than do avirulent strains of FECV41. However, most strains of FIPV are antigenically identical to their avirulent FECV counterparts, and the genetic changes that are responsible for the gain in virulence are not well understood. In an elegant longitudinal study, de Groot and colleagues42 showed that domestic cats that were experimentally infected with FIPV developed a multiphasic disease. Initially, all animals developed fever, weight loss and lymphopaenia but could contain the infection. Total lymphocyte counts recovered with time; however, in most animals, the infection relapsed, as was shown by an increase in viral load. These increased viral burdens resulted in repeated bouts of disease, which again coincided with fever, weight loss and lymphopaenia. FIPV-infected felines develop histological evidence of SEROSITIS and PYOGRANULOMATOUS VASCULITIS. In the more common ‘wet’ form of FIP (also known as the effusive form), yellow ASCITIC FLUID gradually accumulates as the disease progresses; there is also a ‘dry’ form of the disease, which does not involve the accumulation of ascitic fluid. Antigen–antibody-complex formation and complement activation occur in the late stages of disease and might contribute to the production of ascitic fluid in the wet form of disease43. The clinical signs and lymphocyte depletion are postulated to be a direct consequence of the infection of macrophages and DCs by FIPV. In support of this idea, both macrophages and DCs express CD13 (also known as aminopeptidase N), the receptor for FIPV, and infection of macrophages by FIPV has been shown in vitro41,44 (FIG. 2a,b). Infected macrophages traffic throughout the body, resulting in a disseminated infection. In lymph nodes, infection of macrophages and DCs might alter the interaction of these cells with T cells so that they dampen, rather than reinforce, the FIPV-specific T-cell response. This might occur through induction of expression of IL-10 (probably produced by macrophages), which is present in increased amounts in infected lymph nodes45. The production of IL-10 might skew the immune response away from a protective T helper 1 (TH1)-cell response towards a non-protective TH2-cell response, thereby diminishing the ability of immune cells to clear the virus. Regarding the lymphopaenia, lymphocyte apoptosis is commonly observed in infected lymphoid tissues, although lymphocytes are not themselves infected by the virus42,46,47. So, lymphoid-cell depletion is probably mediated by a soluble factor. In support of this idea,



REVIEWS a Macrophage infection and dysfunction Immune response bias to TH2 cells

Antibody-opsonized FIPV virion

Dissemination of infection

T cell

IL-10?

FcγR

Activated, infected macrophage

Infection of macrophages (and possibly DCs)

Apoptosis

TNF?

Interaction of infected macrophage with T cells in lymphoid tissues

b Coronavirus-receptor-mediated entry Macrophage

Spike

FIPV

Nucleus

FIPV receptor (CD13)

c Antibody-dependent enhancement of entry Antigen–antibodycomplex formation

Complement activation C3a

Vasculitis and oedema

C3b

FIPV spikespecific antibody

Figure 2 | Macrophage infection and antibody-dependent enhancement of virus entry in infection with feline infectious peritonitis virus. a | Infection of macrophages and possibly dendritic cells (DCs) results in both dissemination of feline infectious peritonitis virus (FIPV) infection and dysregulation of these cells, leading to lymphocyte apoptosis. b | FIPV usually infects cells through the binding of the spike protein to its cellular receptor, CD13. The virus is then internalized and released into the cytoplasm. c | In antibody-dependent entry, specific antibodies bind the spike protein. The antibody-opsonized FIPV virions then interact with FcγRs (receptors for IgG). Some evidence indicates that this process augments the normal spike–CD13 interaction. After binding of the opsonized virions to FcγRs, the virus is internalized and released into the cytoplasm. Antigen–antibody complexes are also deposited in the vasculature, resulting in complement activation. Activation of complement contributes to the development of vasculitis and oedema, with death of the animal occurring soon after. C3, complement component 3; IL-10, interleukin-10; TH2 cell, T helper 2 cell; TNF, tumour-necrosis factor.


conditioned media from peritoneal or splenic cells that were isolated from infected domestic cats and cultured in vitro induced the apoptosis of T cells46. The specific factor that is responsible for apoptosis has not yet been identified, although TNF, a factor that is implicated in lymphocyte apoptosis, is detected at increased levels in domestic cats with FIP48,49. By contrast, domestic cats that are exposed to FIPV but do not develop disease show lymphoid hyperplasia, which is consistent with the development of a protective T-cell response50. Antibody-dependent enhancement of FIPV infection. As disease progresses, FIPV-infected macrophages are deposited in the endothelium of small blood vessels in organs such as the liver, spleen and kidneys, and GRANULOMAS subsequently form at these sites. Granulomas consist of monocytes and/or macrophages, B cells and CD4+ T cells, with only small amounts of virus detected51,52. These pyogranulomatous lesions are associated with severe damage to the endothelium and are responsible for many of the manifestations of disease, such as liver and renal disease, in infected domestic cats. Another manifestation of FIP is B-cell hyperplasia with associated hypergammaglobulinaemia. Although the aetiology of B-cell hyperplasia is not known, it most probably results from aberrant cytokine production by infected macrophages. Neutralizing antibodies that are raised during FIPV infection and other coronavirus infections are mainly directed against spike protein. In the case of FIPV, however, the presence of these antibodies does not provide sterilizing immunity. Instead, these antibodies opsonize virus particles and facilitate their entry to monocytes and/or macrophages through Fcγ receptors (receptors for IgG)40 (FIG. 2a,c). Furthermore, antigen–antibody complexes are deposited in the blood-vessel walls. These complexes activate complement, leading to vasculitis and oedema, and might thereby contribute to the development of the wet form of FIP40. Antigen–antibody-complex formation and its consequences occur late in the course of disease and are predictive of a poor outcome53. Indeed, the humoral response that develops in FECV-immune domestic cats does not protect animals against FIPV infection and might contribute to a particularly fulminant disease known as early death syndrome. This ‘enhanced’ form of disease has not been documented for animals that are naturally infected but has been shown for domestic cats that are passively or actively immunized against FIPV. Administration of spikeprotein-specific antibodies to uninfected domestic cats or active immunization of domestic cats with recombinant vaccinia virus that expresses spike protein results in an accelerated disease course after infection with FIPV54,55. This syndrome does not develop in FECVinfected domestic cats, possibly reflecting the ability of FIPV to replicate more efficiently in macrophages. This accelerated pathogenesis of FIPV in immunized animals provides strong support for the idea that the immune response is an important contributing factor to disease.



REVIEWS

Murine coronavirus infections

GRANULOMA

A collection of modified macrophages that resemble epithelial cells. This is usually surrounded by a layer of lymphocytes that often includes multinucleated giant cells. Granuloma formation is a chronic inflammatory response that is initiated by various infectious and non-infectious agents.

Infection

Immune-mediated demyelination: the result of an excessive immune response? Several strains of MHV cause acute and chronic neurological diseases in susceptible mice and rats. The JHM and A59 strains of MHV cause demyelination; this infection has been intensively studied because it is a useful model of the human disease multiple sclerosis56–58. In early studies, demyelination was thought to result from virus-mediated lysis of infected cells59,60. However, it is now clear that myelin destruction is largely immune mediated (FIG. 3). Accordingly, mice that receive sub-lethal doses of irradiation or are congenitally immunodeficient (such as mice with severe combined immunodeficiency (SCID) or mice that are deficient in recombination-activating gene 1 (Rag1) activity) do not develop demyelination after infection with MHV-JHM61–64. Both Rag1–/– and SCID mice lack T and B cells but have normal numbers of macrophages and natural killer cells. In one model of infection, an attenuated variant of MHV-JHM (MHV-JHM 2.2-V-1) with a tropism for oligodendrocytes is used to inoculate susceptible mice (usually C57BL/6 or BALB/c mice)61,65–68. These mice develop signs of demyelination, including hind-limb paralysis and gait disturbances, by 7 days after infection. The virus is cleared by 12–14 days after infection, but the neurological deficits persist. MHV-JHM infection of the central nervous system (CNS) results in

Activation of complement and Fcγ-receptor pathways

e

MHV CD66 Opsonized MHV virion

Glial cell

Macrophage

Killing of infected cell by T cell

Peptide–MHC

a

TCR

Production of pro-inflammatory chemokines and other cytokines

d

Overexpression of CCL2

c b

MHV-specific T cell

Activated macrophage

Destruction of oligodendrocytes and myelin, resulting in demyelination

Figure 3 | Mechanisms of immune-mediated demyelination in infection with murine hepatitis virus. In immunocompetent mice, infection of glial cells (that is, astrocytes, microglia and oligodendrocytes) results in migration of T cells into the central nervous system (a). Myelin destruction is mediated by CD4+ and CD8+ T cells, and these cells activate macrophages by the production of cytokines (b) or kill infected cells directly (c), both of which result in demyelination. In recombination-activating gene 1 (Rag1)–/– mice, which lack T and B cells, two additional mechanisms of demyelination have been elucidated. Rag1–/– mice do not develop demyelination when infected with the JHM strain of murine hepatitis virus (MHV-JHM); as occurs in immunocompetent mice, demyelination develops after the adoptive transfer of MHV-JHM-specific T cells. However, demyelination also results if infected Rag1–/– mice are infected with a recombinant MHV-JHM expressing the macrophage attractant CC-chemokine ligand 2 (CCL2) (d), presumably by direct activation of macrophages. Similarly, exogenous delivery of neutralizing MHV-JHM-specific antibody (e) results in macrophage activation and demyelination; this process depends on activation through complement and activating Fcγ receptors (receptors for IgG). TCR, T-cell receptor.

922 | DECEMBER 2005

| VOLUME 5

a large infiltration of B cells and CD4+ and CD8+ T cells. MHV-JHM-specific CD8+ T cells are required for initial clearance of virus, which occurs through perforin- and IFN-γ-dependent pathways, whereas MHV-JHMspecific antibody is required to prevent reactivation of virus61,65–67. However, the influx of these T cells also seems to set up a destructive cascade of cytokine production that results in demyelination. These T cells are mainly directed against MHV-JHM antigens, although in one report, the authors identified myelin-specific CD4+ T cells in the CNS of the infected rat (a possible example of epitope spreading)69. Histological evidence of demyelination is present in the spinal cords of MHV-JHM-infected mice, beginning at ∼5 days after infection, with maximal myelin destruction occurring at 14–21 days after infection. Most tissue destruction occurs when infectious virus has been completely or partially cleared70,71. Viral antigens are not detected in areas of demyelination, whereas they are found in adjacent white matter that appears to be normal, indicating that myelin destruction is a direct consequence of clearance of virus72. Large influxes of activated macrophages and microglia are found in these areas of demyelination70, indicating that these cells have a crucial role in the pathological process. Both CD4+ and CD8+ T cells can mediate disease64,72. Mice that have defective MHC class I or class II expression and therefore lack CD8+ or CD4+ T cells, respectively, develop demyelination when infected with MHV-JHM or MHV-A59 REFS 64,73. Other studies have used adoptive transfer of MHV-JHM-specific immune cells to infected Rag1–/– mice; in the absence of transferred cells, these mice do not develop demyelination after infection with MHV-JHM, but transfer of splenocytes results in robust demyelination. Infected Rag1–/– mice that are reconstituted with either CD4+ or CD8+ T cells from a mouse that is immune to infection with MHV-JHM also develop demyelination. Demyelination can also be induced by CD8+ T cells that do not recognize viral or CNS antigens, if these T cells are sufficiently activated74 (a process that is known as bystander activation) TABLE 1. Accordingly, MHV-JHM-infected Rag1–/– mice that expressed a single T-cell receptor specific for an epitope encoded by another virus (lymphocytic choriomeningitis virus) developed demyelination, but this only occurred when the transgenic T cells were activated by cognate antigen. Although the antiviral T-cell response is crucial for MHV-JHM- or MHV-A59-induced demyelination in normal mice, MHV-JHM-specific antibodies (in the absence of any transferred T cells) can mediate demyelination in infected Rag1–/– mice75. Virus-specific T cells and antibodies activate macrophages and/or microglia, resulting in their migration into the white matter of the CNS and, subsequently, in demyelination. The activation and migration of macrophages and microglia might be the most crucial process for demyelination, because virus-encoded expression of a single macrophage attractant, CCL2, is sufficient to induce demyelination in MHV-JHM-infected Rag1–/– mice in the absence of transferred T cells or antibody76.



REVIEWS macrophages is a cause of the immune dysregulation that is observed in these animals. Instead, a generalized, excessive, but perhaps appropriate, response by both infected and uninfected macrophages seems to be crucial for demyelination.

MHV-3

a

CD66

b Production of nucleocapsid protein

c p38MAPK

Unknown protein

Macrophage Nucleus

d e

HNF4α Fgl2

FGL2 Fibrin deposition in the liver

Figure 4 | Infection with murine hepatitis virus strain 3 results in upregulation of expression of fibrinogen-like protein 2. Murine hepatitis virus strain 3 (MHV-3) becomes internalized after binding macrophages through its receptor, CD66 (a). Subsequent to internalization, virions are uncoated and begin to replicate. As part of the replication process (b), the nucleocapsid protein is synthesized. Subsequently, a signalling pathway involving p38 mitogen-activated protein kinase (p38MAPK) activation, as well as the nucleocapsid protein and other unknown host factors, is initiated (c), resulting, ultimately, in binding of the transcription factor hepatocyte nuclear factor 4α (HNF4α) to the gene that encodes fibrinogen-like protein 2 (FGL2) (d), which is a prothrombinase. The FGL2 protein that is produced then translocates to the cell surface (e), where it induces fibrin deposition and, consequently, acute liver necrosis.

Soluble factors also have a key role in MHV-JHMinduced demyelination. The cytokine IFN-γ is directly involved in clearance of virus from oligodendrocytes67 and is required for demyelination induced by CD8+ T cells but not by CD4+ T cells61,72,77,78. Several studies have shown that certain chemokines contribute to maximal viral clearance and demyelination (but none has been shown to be required for either). Chemokines such as CCL4, CCL5, CXCL9 and CXCL10 are crucial for lymphocyte and macrophage infiltration into the MHV-JHM-infected CNS, and they positively reinforce the milieu in which demyelination takes place79–82. Collectively, these data indicate that myelin destruction is immune mediated. Virus-specific T cells mediate myelin destruction in immunologically intact mice, but in some circumstances, demyelination clearly occurs in their absence. A common feature in all forms of demyelination is the activation and migration of macrophages and/or microglia into the white matter, and this process is sufficient, in the absence of T or B cells, for myelin destruction in MHV-JHM-infected mice. As occurs in FIPV-infected domestic cats, macrophages and microglia can be infected by MHV-JHM71,83,84, but there are no data indicating that direct infection of


MHV-3-mediated lethal hepatitis: induction of expression of a novel prothrombinase. MHV-3 causes various diseases, with the outcome dependent on the strain, age and immune status of the mouse host85. Semisusceptible strains, such as C3H mice and F1 crosses of resistant and susceptible strains, become persistently infected, which manifests mainly as neurological disease: ependymitis, encephalitis and hydrocephalus are hallmarks of this disease59. MHV-3-infected C3H mice also develop a chronic thrombotic vasculitis, with viral antigen detected in endothelia. The pathogenesis of this disease is not well understood, but it is likely to be immune mediated. MHV-3, unlike the neurotropic MHV-JHM and MHV-A59 strains, infects T and B cells. This infection is largely non-productive but results in lymphocyte death, perhaps by apoptosis, with consequent cellular and humoral immunosuppression. This immunosuppression is an important contributory factor to the persistence of MHV-3 REF. 28. Infection of susceptible strains of mice with MHV-3 results in an acute hepatitis, with death occurring a few days after inoculation. The receptor for MHV-3, CD66 (also known as CEACAM1), is expressed by macrophages86, and infection of these cells has a central role in the pathogenesis of the liver failure that is seen in these animals. Macrophages from susceptible mouse strains that are infected with MHV-3 upregulate the production of several pro-inflammatory molecules both in vitro and in vivo, including a transmembrane procoagulant molecule, fibrinogen-like protein 2 (FGL2; also known as fibroleukin)6,87–89 (FIG. 4). Expression of FGL2 results in cleavage of prothrombin to thrombin, which initiates the coagulation cascade that begins with fibrin deposition. This deposition of fibrin throughout the hepatic sinusoids and venous system results in inadequate perfusion or lack of perfusion to the liver and accelerates the necrosis that is caused by the direct cytotoxicity of MHV-3 to hepatocytes. This upregulation of expression of the prothrombinase FGL2 occurs only in macrophages, monocytes and, to a lesser extent, endothelial cells in the liver of susceptible strains of mice, even though MHV-3 can also replicate in the liver-resident macrophages of resistant strains of mice. Upregulation of FGL2 expression by macrophages and monocytes correlates with the severity of liver disease considerably better than do viral titres. Expression of FGL2 depends on activation of p38MAPK, expression of the MHV-3 nucleocapsid protein, and binding of the transcription factor hepatocyte nuclear factor 4α (HNF4α), which is constitutively expressed by macrophages, to the Fgl2 promoter90,91. How the nucleocapsid protein induces HNF4α binding to the promoter of Fgl2 is unknown; similarly, the factors that abrogate upregulation of FGL2 expression in resistant strains of mice are also unknown.



REVIEWS

Table 2 | Does SARS have an immunopathological component? Criterion

Evidence in SARS-CoV infection

Precedent in other coronavirus infections?

Worse disease with decrease in viral load

Controversial; viral titres, measured in nasopharyngeal-aspirate samples, decrease as clinical disease worsens24,36; but high viral loads have been detected in lungs and immune cells after death25–27

MHV-induced demyelination increases as virus is cleared71; MHV-3-induced hepatitis correlates with macrophage activation and not viral load87; and IBVinduced nephritis is detected in chickens with very low viral loads92

Macrophage or DC infection

Infection is abortive but induces expression of pro-inflammatory mediators30,32,34

MHV and FIPV productively infect macrophages40,59

Macrophage infiltration into sites of inflammation

Macrophages are present in large numbers in infected lungs22,29

In MHV infection, macrophages infiltrate the CNS coincident with demyelination (thought to be the final effector cell)62,76; and in FIPV infection, macrophages are the main cell type in granulomas and are crucial for pathogenesis40,43,52

High concentration of pro-inflammatory mediators in serum or at site of infection

Controversial; anti-inflammatory mediators might contribute to delayed viral clearance7,39

MHV-3-induced FGL2 expression is crucial for liver necrosis; in MHV-JHM-infected mice, IFN-γ is required for CD8+ T-cell-mediated responses61,77,89; and in FIPV infection, increased cytokine concentrations are present in blood and tissues during exacerbation of disease49,112,113

Inhibition of type I IFN induction in infected cells

Shown using isolated macrophages, DCs and fibroblasts30,32–34

MHV does not induce type I IFN expression114,115

Lymphopaenia and neutrophilia

Present in most severe cases8,38; and lymphocytic infection has been detected25

In FIPV infection, lymphopaenia is present during clinical relapses42; and in MHV-3 infection, lymphopaenia is present and lymphocytic infection has been detected28

Haemophagocytosis

Present in severe cases21,22

Not reported

CNS, central nervous system; DC, dendritic cell; FGL2, fibrinogen-like protein 2; FIPV, feline infectious peritonitis virus; IBV, avian infectious bronchitis virus; IFN, interferon; MHV, murine hepatitis virus; MHV-JHM, MHV strain JHM; SARS, severe acute respiratory syndrome; SARS-CoV, SARS coronavirus.

Avian coronavirus infections

Role of robust innate immune response in acute respiratory disease. Avian infectious bronchitis virus (IBV) causes marked respiratory disease, especially in young chickens92,93. Similar to SARS-CoV, IBV also infects organs other than the respiratory tract. IBV replicates in the gastrointestinal tract, but infection of the gut does not usually result in clinically evident disease. IBV also infects the kidneys, and some strains of virus cause severe nephritis, which results in a high rate of mortality92,93. At present, there is no evidence that IBV infects macrophages. It is clear that virus-induced cytolytic destruction accounts for many of the pathological changes that are observed in this infection. However, there are indications that an immunopathogenic component contributes to IBV-induced disease. Much of the respiratory disease that is observed in young chickens with severe clinical signs is a result of mucosal thickening and excessive production of a thick mucus in the airways, which asphyxiates the infected host92. Although it has not been proven for IBV-infected chickens, it is probable that this excessive production of mucus is mediated by several cytokines, such as IL-1β, IL-6 and CXCL8, which are secreted by infected epithelial cells in other respiratory-virus infections94. CXCL8 is an attractant for neutrophils, which are one of the main cellular components of the nasal exudates that are found in infected chickens. Neutrophil depletion by

924 | DECEMBER 2005

| VOLUME 5

5-fluorouracil has been shown to reduce the thickness of the nasal exudate, thereby diminishing epithelial-cell damage and cilia loss95. Similarly, the early stages of renal disease are characterized by a massive infiltration of neutrophils96. Collectively, these observations indicate that the cellular innate immune response to the virus is an important factor in the development of severe disease. Chronic lymphocytic nephritis. Under certain conditions (depending on age at time of inoculation, and strain of IBV and of chicken), IBV causes a persistent infection, with an interstitial lymphocytic nephritis in the kidneys92,93,97. Similar to the MHV-infected CNS, lymphocytic infiltration and ongoing renal damage have been shown when viral loads are low or undetectable97. The precise roles of persistent virus and the host immune response in IBV-induced renal disease are not known, but by analogy with other coronavirus infections, it is probable that the chronic nephritis that is observed in these animals is partially immune mediated. An immunopathogenic component in SARS?

Several features of SARS indicate, but do not prove, that the host immune response contributes to disease; these have been described in detail in previous sections and are summarized in TABLE 2. However, as also outlined in this Review, the pathogenesis of



REVIEWS several animal coronaviruses includes an immunemediated component. It is therefore reasonable to suggest that similar mechanisms occur in SARS. Clearly, it is crucial to determine the extent to which the pathological changes that are observed in SARS result from direct destruction by virus compared with immune-mediated elimination of infected cells. In FIPV-infected domestic cats, increased viral replication initiates a cascade of events that leads to injury to the immune system, as well as to several organs42,43, whereas in mice with MHV-mediated demyelination, clinical disease and myelin destruction increase as virus is cleared70,71,76. In patients with SARS, virus is detected in the lungs and in immune cells at the time of death, indicating that virus directly causes pulmonary and immune-system injury25–27. However, the kinetics of viral clearance from sites of infection in individual patients need to be established before the role of the host immune response in the disease process can be fully evaluated. Several features that are common to animals infected with FIPV or MHV and to patients infected with SARS-CoV are consistent with immunopathological disease. These include the propensity of virus to infect macrophages and DCs, and the presence of increased, and perhaps pathological, systemic concentrations of chemokines and other cytokines. In animals infected with MHV or FIPV, activated macrophages are present at sites of inflammation and participate in tissue destruction40,62,70. Activated macrophages are also present in the lungs of SARSCoV-infected individuals. SARS-CoV-infected macrophages and DCs express increased amounts of pro-inflammatory cytokines30,32,34. Consistent with this, increased concentrations of pro-inflammatory chemokines and other cytokines are present in most infected patients35–37,39, and by analogy with other coronavirus infections, as well with ARDS (adult respiratory distress syndrome)98, expression of these pro-inflammatory mediators might contribute to disease. Another immunopathological mechanism, antibody-dependent enhancement of disease, is observed in immunized domestic cats after challenge with FIPV55, and it occurs when tissue-culture cells are exposed to recombinant viral vectors that are coated with the SARS-CoV spike protein99. However, this phenomenon has not been shown in most immunization studies7, and it needs to be confirmed using infectious SARS-CoV.

1.

2.

3.

4.

Griffin, D. E. Immune responses during measles virus infection. Curr. Top. Microbiol. Immunol. 191, 117–134 (1995). Rehermann, B. & Nascimbeni, M. Immunology of hepatitis B virus and hepatitis C virus infection. Nature Rev. Immunol. 5, 215–229 (2005). Mejias, A., Chavez-Bueno, S. & Ramilo, O. Respiratory syncytial virus pneumonia: mechanisms of inflammation and prolonged airway hyperresponsiveness. Curr. Opin. Infect. Dis. 18, 199–204 (2005). Banerjee, S., Narayanan, K., Mizutani, T. & Makino, S. Murine coronavirus replication-induced p38 mitogenactivated protein kinase activation promotes interleukin-6

5.

6.

7.

Future directions

Fortunately, the world has not witnessed a re-emergence of SARS since 2003. To be prepared for any outbreak that might occur in the future, it is crucial to understand the pathogenesis of this disease. In vitro studies will be useful for investigating how SARS-CoV modifies gene expression in primary target cells, such as macrophages, DCs, lymphocytes and pulmonary epithelial cells. However, many of the outstanding issues that have been discussed in this Review will be answered only in the context of SARS-CoV-infected animals or humans. Particularly in the absence of any resurgence of disease in humans, it will be most important to develop an animal model that accurately reproduces the human infection. Current animal models of coronavirus infection are useful for testing vaccines and antiviral drugs, but they do not reproduce the pulmonary or immunesystem disease that is observed in individuals with SARS7. Although SARS-CoV replicates in the lungs of mice, hamsters and domestic cats, these animals remain asymptomatic100–103. Initial reports indicated that cynomolgus macaques (Macaca fascicularis) and ferrets develop clinically evident respiratory disease and would therefore be useful animal models for studying SARS; however, these results have not been reproducible 101,104–108 . Only an animal model will allow investigators to determine the relationship of viral load to disease outcome, as well as to evaluate fully the role of infection and dysfunction of macrophages and lymphocytes in the disease process. In the case of another human coronavirus, HCoV-229E, development of a mouse model of infection required transgenic expression of the human host-cell receptor (CD13), disruption of the innate immune response of the mouse and adaptation of the virus to growth in CD13-expressing mouse cells109. Mouse and rat ACE2 molecules are less-efficient receptors for SARS-CoV than is human ACE2 REF. 110, and the development of a useful murine model of SARS will probably require transgenic expression of human ACE2. However, by analogy with the mouse model of HCoV-229E infection, the development of a transgenic mouse might be only the first step towards developing a useful murine model. The knowledge gained from the study of an animal model will facilitate the development of specific therapies that are designed to minimize pulmonary disease and optimize the anti-SARS-CoV immune response, whether it be excessive (but not necessarily dysregulated), suppressed or both.

production and virus replication in cultured cells. J. Virol. 76, 5937–5948 (2002). Menasche, G., Feldmann, J., Fischer, A. & de Saint Basile, G. Primary hemophagocytic syndromes point to a direct link between lymphocyte cytotoxicity and homeostasis. Immunol. Rev. 203, 165–179 (2005). Parr, R. L. et al. Association of mouse fibrinogen-like protein with murine hepatitis virus-induced prothrombinase activity. J. Virol. 69, 5033–5038 (1995). Peiris, J. S., Guan, Y. & Yuen, K. Y. Severe acute respiratory syndrome. Nature Med. 10, S88–S97 (2004).


8.

Peiris, J. S., Yuen, K. Y., Osterhaus, A. D. & Stohr, K. The severe acute respiratory syndrome. N. Engl. J. Med. 349, 2431–2441 (2003). 9. Lai, M. M. C. & Cavanagh, D. The molecular biology of coronaviruses. Adv. Virus Res. 48, 1–100 (1997). 10. van der Hoek, L. et al. Identification of a new human coronavirus. Nature Med. 10, 368–373 (2004). 11. Fouchier, R. A. et al. A previously undescribed coronavirus associated with respiratory disease in humans. Proc. Natl Acad. Sci. USA 101, 6212–6216 (2004). 12. Woo, P. C. et al. Characterization and complete genome sequence of a novel coronavirus, coronavirus HKU1, from patients with pneumonia. J. Virol. 79, 884–895 (2005).



REVIEWS

13. Guan, Y. et al. Isolation and characterization of viruses related to the SARS coronavirus from animals in southern China. Science 302, 276–278 (2003). 14. Chinese SARS Molecular Epidemiology Consortium. Molecular evolution of the SARS coronavirus during the course of the SARS epidemic in China. Science 303, 1666–1669 (2004). References 13 and 14 trace the evolution of SARS-CoV in human and masked palm-civet populations. 15. Song, H. D. et al. Cross-host evolution of severe acute respiratory syndrome coronavirus in palm civet and human. Proc. Natl Acad. Sci. USA 102, 2430–2435 (2005). 16. Hamming, I. et al. Tissue distribution of ACE2 protein, the functional receptor for SARS coronavirus. A first step in understanding SARS pathogenesis. J. Pathol. 203, 631–637 (2004). 17. Li, W. et al. Angiotensin-converting enzyme 2 is a functional receptor for the SARS coronavirus. Nature 426, 450–454 (2003). 18. Kuba, K. et al. A crucial role of angiotensin converting enzyme 2 (ACE2) in SARS coronavirus-induced lung injury. Nature Med. 11, 875–879 (2005). 19. Imai, Y. et al. Angiotensin-converting enzyme 2 protects from severe acute lung failure. Nature 436, 112–116 (2005). 20. Wong, R. S. et al. Haematological manifestations in patients with severe acute respiratory syndrome: retrospective analysis. BMJ 326, 1358–1362 (2003). 21. Wang, J. T. et al. Clinical manifestations, laboratory findings, and treatment outcomes of SARS patients. Emerg. Infect. Dis. 10, 818–824 (2004). 22. Nicholls, J. M. et al. Lung pathology of fatal severe acute respiratory syndrome. Lancet 361, 1773–1778 (2003). 23. Chan, K. H. et al. Detection of SARS coronavirus in patients with suspected SARS. Emerg. Infect. Dis. 10, 294–299 (2004). 24. Peiris, J. S. et al. Clinical progression and viral load in a community outbreak of coronavirus-associated SARS pneumonia: a prospective study. Lancet 361, 1767–1772 (2003). 25. Gu, J. et al. Multiple organ infection and the pathogenesis of SARS. J. Exp. Med. 202, 415–424 (2005). This is an important histopathological study that indicates that SARS-CoV infection of lymphocytes is a crucial determinant of disease outcome. 26. Farcas, G. A. et al. Fatal severe acute respiratory syndrome is associated with multiorgan involvement by coronavirus. J. Infect. Dis. 191, 193–197 (2005). 27. Mazzulli, T. et al. Severe acute respiratory syndromeassociated coronavirus in lung tissue. Emerg. Infect. Dis. 10, 20–24 (2004). 28. Lamontagne, L., Descoteaux, J. P. & Jolicoeur, P. Mouse hepatitis virus 3 replication in T and B lymphocytes correlate with viral pathogenicity. J. Immunol. 142, 4458–4465 (1989). 29. Franks, T. J. et al. Lung pathology of severe acute respiratory syndrome (SARS): a study of 8 autopsy cases from Singapore. Hum. Pathol. 34, 743–748 (2003). 30. Cheung, C. Y. et al. Cytokine responses in severe acute respiratory syndrome coronavirus-infected macrophages in vitro: possible relevance to pathogenesis. J. Virol. 79, 7819–7826 (2005). 31. Yilla, M. et al. SARS-coronavirus replication in human peripheral monocytes/macrophages. Virus Res. 107, 93–101 (2005). 32. Tseng, C. T., Perrone, L. A., Zhu, H., Makino, S. & Peters, C. J. Severe acute respiratory syndrome and the innate immune responses: modulation of effector cell function without productive infection. J. Immunol. 174, 7977–7985 (2005). 33. Spiegel, M. et al. Inhibition of β interferon induction by severe acute respiratory syndrome coronavirus suggests a two-step model for activation of interferon regulatory factor 3. J. Virol. 79, 2079–2086 (2005). This was the first description of the mechanism of type I IFN suppression of a coronavirus. 34. Law, H. K. et al. Chemokine upregulation in SARS coronavirus infected human monocyte derived dendritic cells. Blood 106, 2366–2374 (2005). 35. Wong, C. K. et al. Plasma inflammatory cytokines and chemokines in severe acute respiratory syndrome. Clin. Exp. Immunol. 136, 95–103 (2004). 36. Wang, W. K. et al. Temporal relationship of viral load, ribavirin, interleukin (IL)-6, IL-8, and clinical progression in patients with severe acute respiratory syndrome. Clin. Infect. Dis. 39, 1071–1075 (2004). 37. Zhang, Y. et al. Analysis of serum cytokines in patients with severe acute respiratory syndrome. Infect. Immun. 72, 4410–4415 (2004).

926 | DECEMBER 2005

38. Tsui, P. T., Kwok, M. L., Yuen, H. & Lai, S. T. Severe acute respiratory syndrome: clinical outcome and prognostic correlates. Emerg. Infect. Dis. 9, 1064–1069 (2003). 39. Lee, C. H. et al. Altered p38 mitogen-activated protein kinase expression in different leukocytes with increment of immunosuppressive mediators in patients with severe acute respiratory syndrome. J. Immunol. 172, 7841–7847 (2004). 40. de Groot, R. J. & Horzinek, M. C. in The Coronaviridae (ed. Siddell, S.) 293–315 (Plenum, New York, 1995). 41. Stoddart, C. A. & Scott, F. W. Intrinsic resistance of feline peritoneal macrophages to coronavirus infection correlates with in vivo virulence. J. Virol. 63, 436–440 (1989). 42. de Groot-Mijnes, J. D., van Dun, J. M., van der Most, R. G. & de Groot, R. J. Natural history of a recurrent feline coronavirus infection and the role of cellular immunity in survival and disease. J. Virol. 79, 1036–1044 (2005). This is a careful study of the clinical and immunological characteristics of FIP. 43. Hoskins, J. D. Coronavirus infection in cats. Vet. Clin. North Am. Small Anim. Pract. 23, 1–16 (1993). 44. Tresnan, D. B., Levis, R. & Holmes, K. V. Feline aminopeptidase N serves as a receptor for feline, canine, porcine, and human coronaviruses in serogroup I. J. Virol. 70, 8669–8674 (1996). 45. Dean, G. A., Olivry, T., Stanton, C. & Pedersen, N. C. In vivo cytokine response to experimental feline infectious peritonitis virus infection. Vet. Microbiol. 97, 1–12 (2003). 46. Haagmans, B., Egberink, H. & Horzinek, M. Apoptosis and T-cell depletion during feline infectious peritonitis. J. Virol. 70, 8977–8983 (1996). 47. Kipar, A. et al. Histopathological alterations of lymphatic tissues in cats without feline infectious peritonitis after long-term exposure to FIP virus. Vet. Microbiol. 69, 131–137 (1999). 48. Zheng, L. et al. Induction of apoptosis in mature T cells by tumour necrosis factor. Nature 377, 348–351 (1995). 49. Kiss, I., Poland, A. M. & Pedersen, N. C. Disease outcome and cytokine responses in cats immunized with an avirulent feline infectious peritonitis virus (FIPV)-UCD1 and challenge-exposed with virulent FIPV-UCD8. J. Feline Med. Surg. 6, 89–97 (2004). 50. Kipar, A., Kohler, K., Leukert, W. & Reinacher, M. A comparison of lymphatic tissues from cats with spontaneous feline infectious peritonitis (FIP), cats with FIP virus infection but no FIP, and cats with no infection. J. Comp. Pathol. 125, 182–191 (2001). 51. Paltrinieri, S., Cammarata, M. P., Cammarata, G. & Comazzi, S. Some aspects of humoral and cellular immunity in naturally occuring feline infectious peritonitis. Vet. Immunol. Immunopathol. 65, 205–220 (1998). 52. Kipar, A., Bellmann, S., Kremendahl, J., Kohler, K. & Reinacher, M. Cellular composition, coronavirus antigen expression and production of specific antibodies in lesions in feline infectious peritonitis. Vet. Immunol. Immunopathol. 65, 243–257 (1998). 53. Jacobse-Geels, H. E., Daha, M. R. & Horzinek, M. C. Antibody, immune complexes, and complement activity fluctuations in kittens with experimentally induced feline infectious peritonitis. Am. J. Vet. Res. 43, 666–670 (1982). 54. Weiss, R. C. & Scott, F. W. Antibody-mediated enhancement of disease in feline infectious peritonitis: comparisons with dengue hemorrhagic fever. Comp. Immunol. Microbiol. Infect. Dis. 4, 175–189 (1981). 55. Vennema, H. et al. Early death after feline infectious peritonitis virus challenge due to recombinant vaccinia virus immunization. J. Virol. 64, 1407–1409 (1990). This work clearly shows that immunization with vaccinia virus expressing FIPV spike protein results in more severe disease after infection with FIPV. 56. Lavi, E., Gilden, D., Wroblewska, Z., Rorke, L. & Weiss, S. Experimental demyelination produced by the A59 strain of MHV. Neurology 34, 597–603 (1984). 57. Lampert, P. W., Sims, J. K. & Kniazeff, A. J. Mechanism of demyelination in JHM virus encephalomyelitis. Acta Neuropathol. 24, 76–85 (1973). 58. Weiner, L. P. Pathogenesis of demyelination induced by a mouse hepatitis virus (JHM virus). Arch. Neurol. 28, 298–303 (1973). 59. Stohlman, S. A., Bergmann, C. C. & Perlman, S. in Persistent Viral Infections (eds Ahmed, R. & Chen, I.) 537–557 (Wiley & Sons, New York, 1998). 60. Perlman, S. Pathogenesis of coronavirus-induced infections: review of pathological and immunological aspects. Adv. Exp. Med. Biol. 440, 503–513 (1998). 61. Bergmann, C. C. et al. Perforin and γ interferon-mediated control of coronavirus central nervous system infection by CD8 T cells in the absence of CD4 T cells. J. Virol. 78, 1739–1750 (2004). 62. Wu, G. F. & Perlman, S. Macrophage infiltration, but not apoptosis, is correlated with immune-mediated

| VOLUME 5

63.

64.

65.

66.

67.

68.

69.

70. 71.

72.

73.

74.

75.

76.

77.

78.

79.

80.

81.

82.

83.

demyelination following murine infection with a neurotropic coronavirus. J. Virol. 73, 8771–8780 (1999). Wang, F., Stohlman, S. A. & Fleming, J. O. Demyelination induced by murine hepatitis virus JHM strain (MHV-4) is immunologically mediated. J. Neuroimmunol. 30, 31–41 (1990). Houtman, J. J. & Fleming, J. O. Dissociation of demyelination and viral clearance in congenitally immunodeficient mice infected with murine coronavirus JHM. J. Neurovirol. 2, 101–110 (1996). References 61–64, together with reference 72, show that MHV-JHM-induced demyelination is mainly immune mediated. Lin, M. T., Hinton, D. R., Marten, N. W., Bergmann, C. C. & Stohlman, S. A. Antibody prevents virus reactivation within the central nervous system. J. Immunol. 162, 7358–7368 (1999). Lin, M. T., Stohlman, S. A. & Hinton, D. R. Mouse hepatitis virus is cleared from the central nervous systems of mice lacking perforin-mediated cytolysis. J. Virol. 71, 383–391 (1997). Parra, B. et al. IFN-γ is required for viral clearance from central nervous system oligodendroglia. J. Immunol. 162, 1641–1647 (1999). Fleming, J. O., Trousdale, M. D., El-Zaatari, F., Stohlman, S. A. & Weiner, L. P. Pathogenicity of antigenic variants of murine coronavirus JHM selected with monoclonal antibodies. J. Virol. 58, 869–875 (1986). Watanabe, R., Wege, H. & ter Meulen, V. Adoptive transfer of EAE-like lesions from rats with coronavirus-induced demyelinating encephalomyelitis. Nature 305, 150–153 (1983). Stohlman, S. A. & Hinton, D. R. Viral induced demyelination. Brain Pathol. 11, 92–106 (2001). Wang, F.-I., Hinton, D., Gilmore, W., Trousdale, M. & Fleming, J. O. Sequential infection of glial cells by the murine hepatitis virus JHM strain (MHV-4) leads to a characteristic distribution of demyelination. Lab. Invest. 66, 744–754 (1992). Wu, G. F., Dandekar, A. A., Pewe, L. & Perlman, S. CD4 and CD8 T cells have redundant but not identical roles in virus-induced demyelination. J. Immunol. 165, 2278–2286 (2000). Gombold, J., Sutherland, R., Lavi, E., Paterson, Y. & Weiss, S. R. Mouse hepatitis virus A59-induced demyelination can occur in the absence of CD8+ T cells. Microb. Pathog. 18, 211–221 (1995). Haring, J. S., Pewe, L. L. & Perlman, S. Bystander CD8 T cell-mediated demyelination after viral infection of the central nervous system. J. Immunol. 169, 1550–1555 (2002). Kim, T. S. & Perlman, S. Virus-specific antibody, in the absence of T cells, mediates demyelination in mice infected with a neurotropic coronavirus. Am. J. Pathol. 166, 801–809 (2005). Kim, T. S. & Perlman, S. Viral expression of CCL2 is sufficient to induce demyelination in RAG1–/– mice infected with a neurotropic coronavirus. J. Virol. 79, 7113–7120 (2005). Pewe, L. L. & Perlman, S. CD8 T cell-mediated demyelination is IFN-γ dependent in mice infected with a neurotropic coronavirus. J. Immunol. 168, 1547–1551 (2002). Pewe, L., Haring, J. & Perlman, S. CD4 T-cell-mediated demyelination is increased in the absence of γ interferon in mice infected with mouse hepatitis virus. J. Virol. 76, 7329–7333 (2002). Chen, B. P., Kuziel, W. A. & Lane, T. E. Lack of CCR2 results in increased mortality and impaired leukocyte activation and trafficking following infection of the central nervous system with a neurotropic coronavirus. J. Immunol. 167, 4585–4592 (2001). Glass, W. G. et al. Antibody targeting of the CC chemokine ligand 5 results in diminished leukocyte infiltration into the central nervous system and reduced neurologic disease in a viral model of multiple sclerosis. J. Immunol. 172, 4018–4025 (2004). Liu, M. T., Armstrong, D., Hamilton, T. A. & Lane, T. E. Expression of MIG (monokine induced by interferon-γ) is important in T lymphocyte recruitment and host defense following viral infection of the central nervous system. J. Immunol. 166, 1790–1795 (2001). Liu, M. T. et al. The T cell chemoattractant IFN-inducible protein 10 is essential in host defense against viral-induced neurologic disease. J. Immunol. 165, 2327–2330 (2000). Knobler, R. L., Tunison, L. A. & Oldstone, M. B. Host genetic control of mouse hepatitis virus type 4 (JHM strain) replication. I. Restriction of virus amplification and spread in macrophages from resistant mice. J. Gen. Virol. 65, 1543–1548 (1984).



REVIEWS

84. Xue, S., Sun, N., van Rooijen, N. & Perlman, S. Depletion of blood-borne macrophages does not reduce demyelination in mice infected with a neurotropic coronavirus. J. Virol. 73, 6327–6334 (1999). 85. Yuwaraj, S., Cattral, M., Pope, M. & Levy, G. Murine hepatitis virus: molecular biology and pathogenesis. Viral Hep. Rev. 2, 125–142 (1996). 86. Coutelier, J. P. et al. B lymphocyte and macrophage expression of carcinoembryonic antigen-related adhesion molecules that serve as receptors for murine coronavirus. Eur. J. Immunol. 24, 1383–1390 (1994). 87. Pope, M. et al. Pattern of disease after murine hepatitis virus strain 3 infection correlates with macrophage activation and not viral replication. J. Virol. 69, 5252–5260 (1995). 88. Ding, J. W. et al. Fulminant hepatic failure in murine hepatitis virus strain 3 infection: tissue-specific expression of a novel fgl2 prothrombinase. J. Virol. 71, 9223–9230 (1997). 89. Marsden, P. A. et al. The Fgl2/fibroleukin prothrombinase contributes to immunologically mediated thrombosis in experimental and human viral hepatitis. J. Clin. Invest. 112, 58–66 (2003). This work uses FGL2-deficient mice to show the crucial role of this prothrombinase in liver necrosis. 90. McGilvray, I. D. et al. Murine hepatitis virus strain 3 induces the macrophage prothrombinase fgl-2 through p38 mitogen-activated protein kinase activation. J. Biol. Chem. 273, 32222–32229 (1998). 91. Ning, Q. et al. Induction of prothrombinase fgl2 by the nucleocapsid protein of virulent mouse hepatitis virus is dependent on host hepatic nuclear factor-4α. J. Biol. Chem. 278, 15541–15549 (2003). 92. Raj, G. D. & Jones, R. C. Infectious bronchitis virus: immunopathogenesis of infection in the chicken. Avian Pathol. 26, 677–706 (1997). 93. Cavanagh, D. Severe acute respiratory syndrome vaccine development: experiences of vaccination against avian infectious bronchitis coronavirus. Avian Pathol. 32, 567–582 (2003). 94. Hendley, J. O. The host response, not the virus, causes the symptoms of the common cold. Clin. Infect. Dis. 26, 847–848 (1998). 95. Raj, G. D., Savage, C. E. & Jones, R. C. Effect of heterophil depletion by 5-fluorouracil on infectious bronchitis virus infection in chickens. Avian Pathol. 26, 427–432 (1997).

96. Chen, B. Y., Hosi, S., Nunoya, T. & Otakura, C. Histopathology and immunochemistry of renal lesions due to infectious bronchitis virus in chickens. Avian Pathol. 25, 269–283 (1996). 97. Lee, C., Brown, C., Hilt, D. A. & Jackwood, M. W. Nephropathogenesis of chickens experimentally infected with various strains of infectious bronchitis virus. Avian Pathol. 66, 835–840 (2004). 98. Ware, L. B. & Matthay, M. A. The acute respiratory distress syndrome. N. Engl. J. Med. 342, 1334–1349 (2000). 99. Yang, Z. Y. et al. Evasion of antibody neutralization in emerging severe acute respiratory syndrome coronaviruses. Proc. Natl Acad. Sci. USA 102, 797–801 (2005). 100. Subbarao, K. et al. Prior infection and passive transfer of neutralizing antibody prevent replication of severe acute respiratory syndrome coronavirus in the respiratory tract of mice. J. Virol. 78, 3572–3577 (2004). 101. Martina, B. E. et al. SARS virus infection of cats and ferrets. Nature 425, 915 (2003). 102. Glass, W. G., Subbarao, K., Murphy, B. & Murphy, P. M. Mechanisms of host defense following severe acute respiratory syndrome-coronavirus (SARS-CoV) pulmonary infection of mice. J. Immunol. 173, 4030–4039 (2004). 103. Roberts, A. et al. Severe acute respiratory syndrome coronavirus infection of Golden Syrian hamsters. J. Virol. 79, 503–511 (2005). 104. Fouchier, R. A. et al. Koch’s postulates fulfilled for SARS virus. Nature 423, 240 (2003). 105. Weingartl, H. et al. Immunization with modified vaccinia virus Ankara-based recombinant vaccine against severe acute respiratory syndrome is associated with enhanced hepatitis in ferrets. J. Virol. 78, 12672–12676 (2004). 106. McAuliffe, J. et al. Replication of SARS coronavirus administered into the respiratory tract of African Green, rhesus and cynomolgus monkeys. Virology 330, 8–15 (2004). 107. Rowe, T. et al. Macaque model for severe acute respiratory syndrome. J. Virol. 78, 11401–11404 (2004). 108. Kuiken, T. et al. Newly discovered coronavirus as the primary cause of severe acute respiratory syndrome. Lancet 362, 263–270 (2003). 109. Lassnig, C. et al. Development of a transgenic mouse model susceptible to human coronavirus 229E. Proc. Natl Acad. Sci. USA 102, 8275–8280 (2005). 110. Li, W. et al. Efficient replication of severe acute respiratory syndrome coronavirus in mouse cells is


111.

112. 113.

114.

115.

limited by murine angiotensin-converting enzyme 2. J. Virol. 78, 11429–11433 (2004). Dandekar, A. A., Anghelina, D. & Perlman, S. Bystander CD8 T-cell-mediated demyelination is interferon-γdependent in a coronavirus model of multiple sclerosis. Am. J. Pathol. 164, 363–369 (2004). Goitsuka, R. et al. IL-6 activity in feline infectious peritonitis. J. Immunol. 144, 2599–2603 (1990). Foley, J. E., Rand, C. & Leutenegger, C. Inflammation and changes in cytokine levels in neurological feline infectious peritonitis. J. Feline Med. Surg. 5, 313–322 (2003). Garlinghouse, L. E. Jr, Smith, A. L. & Holford, T. The biological relationship of mouse hepatitis virus (MHV) strains and interferon: in vitro induction and sensitivities. Arch. Virol. 82, 19–29 (1984). Pewe, L. et al. A severe acute respiratory syndromeassociated coronavirus-specific protein enhances virulence of an attenuated murine coronavirus. J. Virol. 79, 11335– 11342 (2005).

Acknowledgements We thank J. Harty, T. Gallagher, S. Varga and N. Butler for comments. This work was supported by the National Institutes of Health (United States) and the National Multiple Sclerosis Society (United States).

Competing interests statement The authors declare no competing financial interests.

Online links DATABASES The following terms in this article are linked online to: Entrez Gene: http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=gene ACE2 | CCL2 | CD66 | CXCL8 | CXCL10 | FGL2 | IL-6 | IRF3 | p38MAPK | spike Infectious diseases information: http://www.cdc.gov/ncidod/diseases/index.htm SARS FURTHER INFORMATION Stanley Perlman’s homepage: http://immuno.grad.uiowa.edu/faculty/facultydetail.asp?ID=21 Access to this interactive links box is free online.



REVIEWS

SUPERNATURAL T CELLS: GENETIC MODIFICATION OF T CELLS FOR CANCER THERAPY Michael H. Kershaw, Michele W. L. Teng, Mark J. Smyth and Phillip K. Darcy Abstract | Immunotherapy is receiving much attention as a means of treating cancer, but complete, durable responses remain rare for most malignancies. The natural immune system seems to have limitations and deficiencies that might affect its ability to control malignant disease. An alternative to relying on endogenous components in the immune repertoire is to generate lymphocytes with abilities that are greater than those of natural T cells, through genetic modification to produce ‘supernatural’ T cells. This Review describes how such T cells can circumvent many of the barriers that are inherent in the tumour microenvironment while optimizing T-cell specificity, activation, homing and antitumour function.

TUMOURASSOCIATED ANTIGEN

(TAA). An antigen that is expressed by tumour cells. These antigens belong to four main categories: overexpressed antigens that are also expressed in small amounts in some normal tissues; tissuedifferentiation antigens, which are also expressed by nonmalignant cells; mutated or aberrantly expressed molecules; and cancer-testis antigens, which are normally expressed in the testes and occasionally in the placenta.

Peter MacCallum Cancer Centre, Saint Andrews Place, East Melbourne, Victoria 3002, Australia. Correspondence to M.H.K. e-mail: michael.kershaw@ petermac.org doi:10.1038/nri1729 Published online 18 November 2005

928 | DECEMBER 2005

Throughout our lives, our immune system protects us from a formidable onslaught of pathogens, and it does this through the coordinated efforts of many types of immune cell and secreted molecule. For example, when challenged by virus, cells of the innate immune system — including natural killer (NK) cells, macrophages and dendritic cells (DCs) — are quickly recruited to the site of challenge, where they engage in cytolysis of infected cells, secretion of pro-inflammatory mediators and phagocytosis of viral particles. Adding to the innate immune response, cells of the adaptive immune system — including T cells and B cells — become activated and proliferate in lymphoid tissue. Antibodies that can neutralize free virus are also produced, and cytotoxic T lymphocytes (CTLs) are recruited to the site of viral challenge, where they participate in the clearance of infected cells. This process is extremely efficient, and large disease burdens can be cleared quickly with high specificity. It is a different story, however, when it comes to immune reactivity to tumours. Despite evidence for immunosurveillance in controlling the incidence of cancer1, many types of malignancy can escape immune control and threaten our lives. Nevertheless, the association between spontaneous regression of

| VOLUME 5

tumours in patients, which occurs rarely, and immune activation2,3 has fuelled intense interest in developing immunotherapy for cancer. Immunotherapy is moving towards acceptance in the clinic for a range of cancers, including melanoma4, lymphoma 5,6 and virus-induced malignancies 7,8 . Importantly, T cells have been shown to have a crucial role in immunotherapy in mouse models and have been implicated in the responses of patients to treatment. Indeed, tumour-specific T cells can be isolated from patients with melanoma and used in adoptive-transfer regimens that can lead to tumour regression4. In addition, the frequency of circulating T cells that react with TUMOURASSOCIATED ANTIGENS (TAAs) can be increased in patients following administration of cancer vaccine. However, outside a limited range of malignancies that seem to be relatively immunogenic — mainly melanoma and renal-cell carcinoma — there are few reports of effective antitumour immune induction and patient responses9. Indeed, with the exception of melanoma10, attempts to generate tumour-specific T cells using vaccines generally only succeed in producing T cells that react with surrogate tumour target cells. Surrogate targets consist of autologous or HLA-matched cells that are



REVIEWS

CENTRAL TOLERANCE

Self-tolerance that is created at the level of the central lymphoid organs. Developing T cells, in the thymus, and B cells, in the bone marrow, that strongly recognize self-antigen face deletion or marked suppression. PERIPHERAL TOLERANCE

The lack of self-responsiveness of mature lymphocytes in the periphery to specific antigens. These mechanisms control potentially self-reactive lymphocytes that have escaped central tolerance. Peripheral tolerance is associated with suppression of production of self-reactive antibodies by B cells and inhibition of selfreactive effector cells, such as cytotoxic T lymphocytes.

pulsed with defined TAA-derived peptide, which results in large amounts of antigen being presented to T cells, but it is doubtful whether this has relevance to the normal physiological levels of processing and presentation of endogenous antigens by tumour cells. The demonstration of T cells that react with autologous or HLA-matched allogeneic tumour cells is a relatively rare occurrence. It seems that specific T cells of appropriate avidity to mediate an effective immune response to tumours are often lacking in the natural immune repertoire. This is not surprising, because most TAAs are self-antigens, and mechanisms of CENTRAL TOLERANCE and PERIPHERAL TOLERANCE delete or suppress the self-reactive T-cell repertoire. In addition, the natural immune system might fall short of fulfilling any of several other crucial requirements for an effective response to malignant disease, including the ability of diverse cell subsets to proliferate to large numbers and home to the tumour. Tumours also have numerous mechanisms to evade detection, such as downregulation of MHC class I expression and

Box 1 | Why the immune system fails to eliminate cancer An immune response to cancer can fail if the immune system does not meet certain basic intrinsic requirements or if extrinsic evasion strategies are used by tumours. The following summarizes various reasons behind the failure of the immune system to clear tumours. One factor or a combination of several factors might be responsible for this failure.

Intrinsic immune failure • Tumour-specific T cells might be absent from the immune repertoire. Mechanisms of central and peripheral tolerance sculpt the immune system, resulting in the elimination of tumour-reactive T cells or, at best, leaving only weakly tumourreactive T cells. • T cells fail to proliferate and persist in response to tumours. Tumour-associated antigens might not be presented by antigen-presenting cells in a form that is sufficiently antigenic for T cells. • Tumour-reactive immune cells do not localize to the tumour. The tumour might not be perceived as a threat to the body, so inflammatory mechanisms that are normally associated with disease might not be initiated94. • Regulatory T cells inhibit tumour-specific T-cell activity. There is a network of CD4+CD25+ regulatory T cells that can suppress tumour-specific T-cell generation and activity95.

Extrinsic immune evasion As tumours develop and grow, they undergo a process of immunoediting through interaction with the immune system that results in the outgrowth of tumour cells that can evade immune-mediated control. This can involve the following processes. • Downregulation of antigen expression. Antigens that were initially recognized by the immune system might be dowregulated or lost entirely under immuneselection pressure. Antigen loss can be at the level of the peptide itself, or it might involve loss of expression of MHC molecules, peptide transporters or proteasome subunits, all of which are important for the presentation of peptides11. • Secretion of immunosuppressive factors. Tumour cells or stromal elements can secrete various factors — including transforming growth factor-β96, interleukin-10 REF. 97 and vascular endothelial growth factor — that inhibit immune cells or bias the effector type produced (such as T helper 1 or T helper 2 cells). • Downregulation of death-receptor pathways. Death receptors such as CD95 and TRAIL receptor (tumour-necrosis-factor-related apoptosis-inducing ligand receptor), or their downstream signalling molecules, can be mutated or lost entirely.


secretion of immunosuppressive factors that render immune cells non-functional11,12 BOX 1. So, the endogenous immune system seems to have weaknesses that could be so fundamental and profound that they might not be easily overcome by manipulation of traditional vaccines and adjuvants TABLE 1. Nevertheless, considerable effort continues to be expended worldwide to boost the endogenous immune system against cancer, and this might eventually lead to effective therapies. However, it is also worth considering an alternative approach that is not entirely reliant on endogenous immune components: that is, the genetic modification of natural T cells to produce ‘supernatural’ T cells with enhanced abilities. In this Review, we outline the hurdles that face the mounting of an immune response to cancer, and we describe genetic strategies to circumvent these hurdles. Approaches that are aimed at providing T cells with increased tumour specificity, antitumour activity and tumour-homing potential are presented, together with ways of increasing T-cell proliferation, survival and resistance to tumour-derived immune-inhibitory factors. In addition to a description of current, validated approaches, we also present novel, as-yet-untested strategies that might find application in the future. Manipulating T-cell specificity

A crucial requirement for an effective immune response is specific T cells, but how can we produce tumourreactive T cells when most tumours express mainly self-molecules? T cells naturally recognize antigen as peptides that are associated with MHC molecules, and highly self-reactive T cells are deleted or rendered nonresponsive during development. T cells of appropriate antiself or tumour specificity and avidity are therefore often lacking in the natural immune repertoire. Nevertheless, genetic strategies can be used to endow T cells with reactivity to TAAs. Modification using TCR genes. Genes that encode the α- and β-chains of tumour-reactive T-cell receptors (TCRs) can be isolated from the T cells of the rare patients who respond to tumours and then can be used to transduce T cells or their precursors, thereby endowing them with the ability to react to tumour cells. The clinical use of such T cells would involve isolating T cells from the peripheral blood of a patient, followed by genetic modification in vitro and re-infusion to the patient. Such TCR-modified human T cells have been shown to react to tumour cells in vitro13. Genetically modified T cells have also been shown to induce antitumour effects in vivo after genetic modification with TCR genes that are specific for influenza-virus nucleoprotein, which was used as a model TAA in mice14. However, it might not be possible to find endogenous, high-affinity human TCRs for all malignancies, so another approach for isolating tumour-specific TCR genes involves the use of transgenic mice that express human MHC molecules15,16. In this approach, human TAAs are seen as foreign by the mouse immune system, and



REVIEWS high-affinity TCRs can be generated that are specific for peptides associated with human MHC molecules. Subsequent use of these mouse TCR genes in humans requires HUMANIZATION using molecular techniques, to

avoid subsequent rejection after transfer to patients. Such humanization procedures have been carried out frequently for monoclonal antibodies, and partial humanization has been achieved for some mouse

Table 1 | Summary of representative published clinical trials of cancer vaccines Vaccine

Cancer type

Immunological end-point

Reference

Melan-A peptide + IL-12

Melanoma

Lysis of peptide-pulsed T2 cells mediated by T cells

98

MAGE3-peptide-pulsed DCs

Gastrointestinal

Peptide-specific response to autologous DCs

99

MAGE3 delivered as peptide, recombinant virus or peptide-pulsed DCs

Melanoma

Lysis of HLA-matched allogeneic tumour cells

100

Melan-A-peptide-pulsed PBMCs + IL-12

Melanoma

Peptide-specific IFN-γ production in response to peptide-pulsed T2 cells

101

Tyrosinase-peptide-pulsed DCs

Melanoma

Peptide-specific IFN-γ production in response to peptide-pulsed PBMCs

102

PSMA-peptide-pulsed DCs

Prostate

50% reduction in PSMA

103

CEA-peptide-pulsed DCs

Colorectal, breast, ovarian or pancreatic

Response to antigen-loaded, cryopreserved DCs

104

ERBB2-peptide-pulsed DCs

Gastrointestinal

Peptide-specific IFN-γ production in response to peptide-pulsed T2 cells

105

Modified gp100 peptide + IFA

Melanoma

IFN-γ production in response to allogeneic tumour cells

10

Tyrosinase peptide + GM-CSF

Melanoma

Peptide-specific T-cell generation (measured by ELISPOT)

106

Multiple peptides (tyrosinase and gp100 peptides) + adjuvant or pulsed onto DCs

Melanoma

Specific IFN-γ production in response to peptide-pulsed T2 cells (measured by ELISPOT)

107

Multiple peptides (combinations of SART1, SART2, SART3, LCK, ART1, ART4 and CYPB peptides) + adjuvant

Colon

Cytolytic activity against allogeneic tumour cell line, and peptidespecific IgG present in serum

108

PSA encoded by recombinant vaccinia virus (+ radiotherapy)

Prostate

Specific IFN-γ production in response to peptide-pulsed C1R cells, and lysis of allogeneic tumour cells by T cells from one patient

109

TERT-peptide-pulsed DCs

Breast or prostate

Specific IFN-γ production in response to peptide-pulsed PBMCs, and lysis of allogeneic tumour cells by T cells from two patients

110

ERBB2 peptide + GM-CSF

Breast or ovarian

Specific IFN-γ production in response to peptide (measured by ELISPOT)

111

NY-ESO-1 peptide + or – GM-CSF

Melanoma, breast or ovarian

Cytotoxicity against peptide-pulsed T2 cells

112

RAS peptide + adjuvant

Tumours with a mutated RAS gene

Cytotoxicity against HLA-A2-matched tumour cell lines carrying the corresponding mutant but not the wild-type RAS gene

13

PSA encoded by recombinant vaccinia virus

Prostate

Cytotoxicity against HLA-A2-expressing, PSA-peptide-pulsed cell line

114

CEA- and CD80-encoding avipoxvirus

Colon, breast or pancreatic

IFN-γ production in response to peptide-pulsed C1R cells expressing HLA-A2

115

CEA-encoding avipoxvirus

Colon, pancreatic or stomach

Cytotoxicity against HLA-matched allogeneic tumour cells, and lysis of autologous tumour cells by T cells from one patient

116

Melan-A peptide + IFA

Melanoma

IFN-γ production in response to peptide-pulsed PBMCs (measured by ELISPOT)

117

p53-peptide-pulsed DCs

Breast

IFN-γ production in response to peptide-pulsed DCs

118

Multiple peptides (MAGE3, tyrosinase, gp100 and melan-A peptides) pulsed onto DCs

Melanoma

DTH response and IFN-γ production in response to peptide

119

BCR–ABL peptide + adjuvant

Chronic myeloid leukaemia

IFN-γ production in response to peptide-pulsed T2 cells

120

Trials include those using peptide alone, peptide-pulsed autologous dendritic cells (DCs), recombinant DNA or virus. With the exception of melanoma, in most patients, T-cell activity was shown against peptide-pulsed target cells but not autologous or HLA-matched tumour cells. ART, adenocarcinoma antigen recognized by T cells; BCR–ABL, Abelson leukaemia-virus protein (ABL) fused with the breakpoint-cluster region (BCR); C1R, a human B-lymphoblastoid cell line; CEA, carcinoembryonic antigen; CYPB, cyclophilin B; DTH, delayed-type hypersensitivity; ELISPOT, enzyme-linked immunosorbent spot; ERBB2, also known as HER2 or NEU; GM-CSF, granulocyte/macrophage colony-stimulating factor; gp100, glycoprotein 100; IFA, incomplete Freund’s adjuvant; IFN-γ, interferon-γ; IL-12, interleukin-12; MAGE, melanoma-associated antigen; NYESO-1, New York oesophageal squamous-cell carcinoma 1; p53, tumour-suppressor protein p53; PBMC, peripheral-blood mononuclear cell; PSA, prostate-specific antigen; PSMA, prostate-specific membrane antigen; SART, squamous-cell carcinoma antigen recognized by T cells; T2, a human T-cell line; TERT, telomerase reverse transcriptase.

930 | DECEMBER 2005

| VOLUME 5



REVIEWS TCRs15. There are considerable technological issues associated with this approach, because it involves the construction, validation and comparison of multiple molecular formats. However, further development might yield valuable reagents for gene therapy for cancer and other chronic diseases. Although the TCR gene-transfer approach might find application for some malignancies, several concerns could limit its use. Included in these concerns is the tendency of transgenic TCR chains to mispair with endogenous TCR chains in T cells isolated from the patient, thereby reducing the cell-surface density of tumour-reactive TCR. Nevertheless, manipulating transmembrane association domains or changing other specific pairing strategies might overcome this concern17. Of further concern with this approach is the necessity of isolating TCRs that are specific for all MHC haplotypes, because TCRs recognize antigen as peptides presented by MHC molecules, which differ between patients. In addition, the inability of most TCRs to recognize carbohydrate or glycolipid antigens makes it impractical to target this important group of TAAs using this strategy.

HUMANIZATION

A process by which parts of molecules originating from other species are replaced by the homologous domains of human molecules, using recombinant DNA techniques. SINGLECHAIN VARIABLE FRAGMENT

A recombinant molecule that is composed of the variable regions of the heavy and light chains of immunoglobulin joined by a flexible oligopeptide linker. HAPTEN

A molecule that can bind antibody but cannot by itself elicit an immune response. Antibodies that are specific for a hapten can be generated when the hapten is chemically linked to a protein carrier that can elicit a T-cell response.

Modification using antibody genes. An alternative approach for the production of tumour-specific T cells involves the use of genes that encode monoclonal antibody chains. Antibodies that are specific for TAAs can readily be generated from immunized mice, and genes that encode SINGLECHAIN VARIABLE FRAGMENTS (scFvs) of these antibodies can be used to construct chimeric receptors that are effective at redirecting T cells to respond to tumour cells18. Chimeric receptors are typically composed of an scFv that is linked by a hinge region to transmembrane and intracellular signalling domains that are derived from the TCR complex or the high-affinity receptor for IgE (FcεRI) (FIG. 1). An advantage of tumour-specific T cells that are generated in this manner is that they respond to antigen in a non-MHC-restricted manner, and the approach is therefore widely applicable for treating patients of all MHC haplotypes, which is in contrast to using cells modified with TCR chains, recognition by which is restricted to a single type of MHC molecule. Early studies showed that T-cell lines could be redirected to recognize the HAPTEN trinitrophenol in vitro by inserting the gene encoding a chimeric trinitrophenol-specific receptor into T cells19. This proof of principle led to the design of the first chimeric scFv reagent that targeted a bona fide ovarian TAA, folate-binding protein; transgenic expression of this chimeric receptor endowed T cells with the ability to lyse tumour cells and secrete cytokines in vitro in response to folate-binding protein20. Since then, several scFv-based chimeric receptors that target TAAs from various malignancies have been generated TABLE 2 and have shown considerable TAA specificity in vitro. In addition, chimeric-receptor-directed T cells have been observed to induce antitumour effects in vivo; however, this is often limited to early disease, such as small subcutaneous tumours21, or to models of peritoneal, lung or bone metastases22–24, which might be more


accessible to immune cells as a result of their diffuse structure, which lacks large, solid masses. Importantly, the range of antigens that can be targeted by scFv-based chimeric receptors is not restricted to protein antigens (in contrast to conventional TCRs) but extends to TAAs of carbohydrate and glycolipid origin25–27, which might not mutate as rapidly as protein antigens. In essence, any cell-surface TAA is a potential target as long as an antibody can be raised against it. Concerns and issues. It is difficult, however, to produce scFv-based chimeric receptors that can direct T cells towards peptides that are derived from intracellular molecules, although some optimism regarding this process is supported by examples of antibodies that can recognize peptide associated with MHC molecules28. In addition, because scFv-transgenic T cells are not MHC restricted, they cannot receive the assistance of professional antigen-presenting cells (APCs), such as DCs, that contribute to the activation of MHC-restrictedTCR-expressing T cells. A further point to consider regarding non-MHC-restricted effector T cells is the potential effect of soluble antigen on effector function (for example, antigen released by the tumour into the bloodstream). However, genetically modified T cells VL CL

VH CH1

VH

VL

CH2 Tumour-specific antibody CH3 Tumour-associated antigen

T cell

Tumour cell

Signalling domains

Hinge

scFv

Antitumour immune response

Figure 1 | Schematic representation of a typical chimeric receptor. The specificity of monoclonal antibodies can be used to redirect T cells towards tumours. In this process, a chimeric receptor encoded by a gene construct is composed of the single-chain variable fragments (scFvs) of tumour-specific antibody linked through a hinge region to transmembrane and cytoplasmic domains of T-cell signalling molecules. Engagement of antigen by the chimeric receptor at the surface of tumour cells results in a T-cell response to the tumour cell. The signalling portion can be composed of several individual regions, and these can include domains from primary activation molecules (such as the γ-chain of the high-affinity receptor for IgE (FcεRIγ) or the ζ-chain of CD3), from co-stimulatory molecules (including CD28, inducible T-cell co-stimulator (ICOS) and 4-1BB) or from intracellular protein kinases (such as LCK). CH, immunoglobulin heavychain constant region; CL, immunoglobulin light-chain constant region; VH, immunoglobulin heavy-chain variable region; VL, immunoglobulin light-chain variable region.



REVIEWS

Table 2 | Tumour-specific chimeric receptors Cancer type

Target antigen

Signalling domains

References

B cell

CD19

CD3ζ 4-1BB–CD3ζ

24,43 24,43

Colon

CEA

CD3ζ CD28–CD3ζ

37,54 121

Ovarian

FBP

FcεRIγ

22,64

Breast and others

ERBB2

CD3ζ CD28–CD3ζ

Prostate

PSMA

CD28–CD3ζ

Many (neovasculature)

KDR

FcεRIγ

21 40,122 39 123

Adenocarcinoma

TAG72

CD3ζ

58

Many

EGP2

CD3ζ

124

Melanoma

GD3

CD3ζ

125

Neuroblastoma

GD2

CD3ζ

57

Renal-cell carcinoma

CA9

FcεRIγ CD4–FcεRIγ

56 126

CA9, carbonic anhydrase IX; CEA, carcinoembryonic antigen; EGP2, epithelial glycoprotein 2; ERBB2, also known as HER2 or NEU; FBP, folate-binding protein; FcεRIγ, γ-chain of the highaffinity receptor for IgE; GD, ganglioside; KDR, kinase insert domain receptor; PSMA, prostatespecific membrane antigen; TAG72, tumour-associated glycoprotein 72.

transmit inhibitory signals, perhaps through domains derived from NKG2A (NK group 2, member A), CTLA4 (cytotoxic T-lymphocyte antigen 4) or PD1 (programmed death 1). Such genetic strategies could yield ‘smart’ T cells that can ‘proof-read’ the surface of cells, with the decision to destroy a cell or let it live being derived from the balance between activating and inhibitory signals. Alternatively, a cell-surface ‘signature’ of several TAAs that is entirely tumour-specific could be determined. T cells could then be modified to express several receptors, each of which recognizes a different TAA. The expression levels and signalling moieties of the receptors could be tailored, perhaps by manipulation of promoter regions and kinase domains, such that T-cell activation is achieved only when the full TAA signature is recognized. Genetic modification could therefore be used to endow T cells with a specificity that is superior to that found in the endogenous immune repertoire; however, simply supplying specificity is not enough to eliminate a threat as dire as malignant disease. T cells also need to have a potent and varied function. Improving T-cell function

IMMUNOEDITING

The process by which interaction of a heterogeneous population of tumour cells with the immune system generates tumour variants with reduced immunogenicity, which might therefore escape from immune responses.

932 | DECEMBER 2005

have been shown to respond to some membranebound antigens even in the presence of soluble antigen, although it is not known whether this is the case for all antigens29. Another concern is that tumours can lose expression of an individual antigen, and IMMUNOEDITING is one of the many forces that sculpt the heterogeneity of tumours12. So, another potential shortcoming of the specificity of normal T cells is that it is monoclonal. The restriction to one specificity, which resulted from an evolutionary process, is crucial for the regulation of endogenous immune reactivity, because having a single specificity is instrumental in tolerance induction and prevention of autoreactivity. However, genetic strategies present us with the opportunity for the rational design of T cells to endow them with multiple specificities, thereby allowing targeting of the special case of malignancy, which involves self-antigens. So, genetic modification can be used to endow T cells with reactivity to several antigens, thereby reducing the risk of tumour escape30. Improving on the inherent monoclonal nature of T cells could also lead to the control of potential collateral damage through autoimmunity. Following T-celldirected immunotherapies, autoimmune responses have been observed to TAAs that are expressed in normal tissues in mice and humans31–34, and these responses might be expected to become more frequent as therapies become more effective and more widely applicable. Genetic modification could be used to equip T cells with multiple receptors, some that are specific for TAAs and some that recognize cell-surface molecules that are expressed only by specific ‘endangered’ types of normal tissue. In this strategy, tumour-specific receptors would transmit activation signals, whereas receptors that engage specifically threatened normal tissue would

| VOLUME 5

After ligation of microbial antigens, T cells have potent functions, including cytokine secretion and cytotoxicity. For example, interferon-γ that is secreted in response to influenza virus can reach microgram quantities per million T cells in vitro, and high levels of lysis of virus-infected target cells can occur at effector-to-target ratios of less than 1 REF. 35. This extent of response, or an even greater response, is what would be ideal in response to tumours. However, tumour-reactive T cells that are derived from the natural pool of lymphocytes rarely approach this level of reactivity, and more often, they have less than 10% of the activity that is observed against non-self-antigens or are entirely non-responsive36, owing to the poor immunogenicity of many tumours or the induction of tolerance to self-molecules BOX 1. Similarly, T cells that were genetically redirected with the earliest versions of chimeric receptors did not secrete large amounts of cytokines in response to tumour cells37. However, the genetic-modification strategy was easily adapted to link antigen recognition to additional activation and co-stimulatory domains (such as from CD28) in a single ‘super signalling’ molecule format. For example, chimeric receptors incorporating both CD3ζ and CD28 signalling domains in their cytoplasmic regions mediated higher amounts of cytokine secretion than receptors that incorporated either domain alone38–41. Genetically modified T cells linking both activation and co-stimulation through CD28 to antigen recognition have also been shown to mediate increased antitumour effects in mice24,40. Incorporation of domains from other co-stimulatory molecules — including 4-1BB, inducible T-cell costimulator (ICOS) and OX40 — into chimeric receptors has also been shown to endow T cells with improved



REVIEWS

SIGNALMOTIF LIBRARY

A collection of cDNA clones that encodes various intracellular-signalling domains joined together randomly and ligated to an extracellular antigen-recognition domain. Individual cells displaying different signalling constructs are screened for optimal function.

function in vitro42,43. In addition, increased chimericreceptor-mediated cytokine secretion has been observed following the sequential addition of co-stimulatory (CD28), co-receptor (CD4) and intracellular kinase (spleen tyrosine kinase (SYK) and LCK) domains to the basic CD3ζ signalling module44,45 . Various co-stimulatory molecules might differ in the signals that they transmit, resulting in differences in the function that they impart46. For example, phosphatidylinositol 3-kinase (PI3K) is important in signal transduction through CD28, whereas signals from 4-1BB can be independent of PI3K. These receptors can function synergistically or can independently induce differential cytokine secretion by T cells47. So, it might be possible to use particular co-stimulatory domains rationally, depending on the desired response to a tumour. The manipulation of signalling domains provides much scope for optimizing receptor function, as seen by the rational design of receptor signalling using the candidate molecules that are described here. However, given the large range of signalling motifs that could be incorporated, a quicker solution for achieving optimal chimeric-receptor function might, in the future, involve the use of SIGNALMOTIF LIBRARY approaches. In this way, signalling domains from many diverse molecules could be integrated as single or multiple copies at random into chimeric receptors, and T cells could then be screened for optimal function following receptor ligation, to identify the best receptor format. In addition, it is conceivable that genetic modification could provide an opportunity to control the type of antitumour activity by enlisting responses from appropriate T-cell subsets. Individual subsets could be genetically modified with a tumour-specific TCR ex vivo, and/or receptor and signalling domains could be manipulated to favour T helper 1 (TH1)- or TH 2-cell responses by enlisting receptors such as Notch48, or transcription factors such as GATA-binding protein 3 (GATA3)49 (for orientation towards TH2-cell responses) or T-bet50 (to favour TH1-cell responses). Furthermore, if other functions such as cytotoxicity are required, then using signalling molecules such as DAP12 (DNAX activation protein 12) might be beneficial. DAP12 is associated with NKG2D in NK cells and enables NK cells to have potent lytic capacity against target cells displaying ligands for NKG2D51. So, there are many options for specifying a tailor-made immune response, and the choice of strategy depends on which mechanisms might be best suited to help us to eliminate a particular tumour type. Using these strategies, we could shift antitumour immunity towards a type 1 immune response to clear tumour cells that have been shown to be relatively sensitive to CTLs52 or towards a type 2 immune response to clear those tumour cells that are more resistant to CTLs53. Genetically redirected T cells can show cytotoxicity against tumours20,54–58, although the current levels of lysis that can be achieved are still lower than those observed for virus-specific T cells lysing cells displaying cognate antigen, which might be a consequence of the virus-specific endogenous TCR complex enlisting


various additional signalling molecules that are not recruited by chimeric receptors. Indeed, increases in cytotoxic ability have been shown after the inclusion of DNA encoding additional co-stimulatory domains, such as those of CD28, in the chimeric-transgene design42. However, this is not the case for chimeric receptors of all specificities38–40; the reasons for this are not clear but might include structural differences in the receptors or variations in the properties of the TAAs. Other strategies to increase the potency of tumourreactive T cells include transduction with genes that encode members of the tumour-necrosis factor (TNF) superfamily that can deliver apoptotic signals to some tumour cells through death receptors. For example, increased production of TNF has been shown by transduced tumour-infiltrating lymphocytes (TILs)59, and increased cytotoxicity of TNF-transduced TILs has been shown against tumour cells that express death receptors in vitro60. However, TNF expression by the TILs in these experiments was constitutive, which might result in non-specific toxicity in vivo. Inducible expression of TNF, on encounter with TAAs or exposure to drugs, might be a safer option for clinical application. Other molecules that are important in lymphocytemediated cell death include CD95 ligand (also known as FAS ligand), TNF-related apoptosis-inducing ligand (TRAIL), and intracellular granule components, such as perforin. Modification of T cells to express large amounts of these molecules might be worth pursuing to increase the antitumour potency of T cells. However, the feasibility of this approach depends on selecting T cells that lack expression of the ligands for these death-inducing molecules or on placing expression under the control of an inducible promoter such as that responsive to tetracycline. In a recent novel approach, T cells were used as delivery vehicles for a retrovirus encoding a ‘suicide’ gene61. T cells with specificity for the TAA carcinoembryonic antigen were engineered to produce a retrovirus encoding the suicide gene herpes simplex virus (HSV) thymidine kinase (tk) in an inducible manner. Significantly increased survival was shown for tumourbearing mice that received retrovirus-producing T cells. Using imaginative approaches such as these could extend the arsenal of effector mechanisms that are used by T cells to counter the resistance that tumours acquire to natural assaults BOX 1. Improving T-cell proliferative capacity

Of crucial importance in mounting an effective immune response is the ability of T cells to proliferate. With more than 50,000 km of blood vessels in a human adult and with blood velocities of ∼1 cm per second in the periphery, a single leukocyte would take more than 30 years to investigate each section of tissue. The proliferation of T cells to high numbers is therefore an important determinant of in vivo efficacy. The proliferation of T cells in response to non-selfantigens can be considerable. T cells can divide as often as every 10–12 hours, leading to the frequency



REVIEWS of antigen-specific T cells in the circulation reaching more than 1 in 4 REF. 62. In this way, antigen-specific precursors, which are estimated to comprise fewer than 105 cells, can proliferate to eliminate large burdens of infectious disease. As mentioned earlier, however, tumours are poorly immunogenic, and T-cell proliferation to self-antigens is tightly controlled, with self-reactive T cells being relatively non-responsive or completely absent. In addition, co-stimulatory ligands might be absent or have reduced expression on a tumour or at the surface of APCs (if these cells are not fully activated by the poorly immunogenic tumour). Nevertheless, genetic-modification strategies are amenable to linking the recognition of a TAA to proliferation through the incorporation of appropriate signalling domains into the cytoplasmic region of chimeric receptors. For example, the inclusion of CD28 signalling domains in chimeric receptors has been shown to result in a two- to fourfold increase in TAA-induced proliferation39,40. These levels of clonal expansion do not yet approach those that are observed in response to foreign antigen, but further optimization of signalling domains might lead to comparable clonal-expansion rates. Remarkably, the proliferative response that results from the introduction of chimeric receptors, which occurs in a non-MHC-restricted manner, might be an advantage of inducing proliferation at the site of malignancy rather than within the local lymph nodes. T cells that express conventional TCRs generally receive proliferative signals from APCs in lymphoid tissue and then need to make their way to tumours. The non-MHC-restricted process of recognition by chimeric receptors might short-circuit some of the requirements of the T-cell response, thereby helping to avoid some regulatory mechanisms that are present in the lymph node, such as regulatory T cells, and subsequent opportunities for tumour escape. Another strategy to enable tumour-reactive T cells to proliferate to a similar degree to that of antigen-specific T cells in response to foreign antigen involves the use of genetic modification to produce dual-specific T cells that are specific for TAA and a potent immunogen. Clinical application of dual-specific T cells would involve their transfer to patients followed by an immunization regimen using the potent immunogen. Such dual-specific T cells have been shown to respond to both tumour cells and immunogens such as Epstein–Barr virus, alloantigen or influenza virus26,63. Importantly, dual specificity has allowed the clonal expansion of tumourreactive cells in vivo in response to allo-immunization, and it has increased antitumour activity in mice64. The extent of the antitumour response using dual-specific T cells might therefore approach that observed to infectious agents such as influenza virus, for which more than 25% of circulating T cells can be specific62. This is in contrast to the T-cell response to tumours following manipulation of the endogenous immune response by immunization, which at best results in 4% (or more usually less than 1%) of T cells being tumour specific65. Another potential advantage of dual-specific T cells

934 | DECEMBER 2005

| VOLUME 5

is that their endogenous specificity is known, so they should be safer to use than bulk populations of T cells, which might include some autoreactive clonotypes. Another innovative way to generate large numbers of tumour-reactive T cells involves genetic modification of haematopoietic precursor cells with TCR α- and β-chains. In a model system in mice, bone marrow was transduced with a TCR that is specific for ovalbumin (OVA), and this resulted in T-cell production after transfer to recombination-activating gene 1 (Rag1)–/– mice66. These T cells were shown to respond to OVA in vitro and to proliferate in vivo following immunization with OVA. If this approach were extended to the human system, the continual production of tumour-reactive T cells from precursor cells might provide an effective, long-lasting antitumour response and protection from relapse. In another novel approach to improving proliferation, T cells themselves were engineered to present antigen and therefore to allow effective clonal expansion of a population of T cells specific for the TAA CD19 REF. 67. This system takes advantage of the constitutive expression of MHC class II molecules and co-stimulatory molecules by activated T cells, which might be sufficient to activate other T cells, although perhaps not as potently as do professional APCs. Such innovative approaches as this one and strategies involving ‘artificial’ APCs68 could provide large numbers of cells with a heightened stimulatory capacity and reduced natural restraints, which could increase T-cell activation and proliferation. So, there are several ways of producing T cells with an increased proliferative capacity, and these could influence tumour growth simply though the quantity of cells generated. However, a similar result could be achieved with smaller numbers of T cells if these cells could be made to survive longer, in particular after encounter with antigen. Engineering improved T-cell survival

There are constraints on the lifespan of naturally occurring tumour-reactive T cells that could be alleviated through genetic-modification strategies. At the conclusion of a natural adaptive immune response, antigenspecific T cells enter a retraction phase, in which T cells undergo apoptosis. Early induction of apoptosis, following incomplete activation or before exposure to antigen, might limit the extent of an antitumour response, and several genetic strategies are being investigated to improve T-cell survival. Included among these is the modification of tumour-reactive T cells to secrete interleukin-2 (IL-2), which is an important T-cell growth factor. Melanoma-specific T cells secreting IL-2 were shown to survive in vitro in the absence of exogenous IL-2 and to retain their antitumour activity 69. Another example of a strategy to extend the lifespan of T cells can be found in studies using a gene that encodes a chimeric granulocyte/macrophage colony-stimulating factor (GM-CSF)–IL-2 receptor, with which T cells can initiate IL-2-associated survival and proliferation programmes in response



REVIEWS

TELOMERES

Regions of highly repetitive DNA at the ends of linear eukaryotic chromosomes. They protect the ends of chromosomes from shortening on replication. Telomerase reverse transcriptase is a ribonucleoprotein enzyme that maintains the ends of chromosomes by adding a characteristic series of nucleotides to telomeres. SMALL INTERFERING RNA

(siRNA). Synthetic RNA molecules of 19–23 nucleotides that are used to ‘knockdown’ (that is, silence the expression of) a specific gene. This is known as RNA interference (RNAi) and is mediated by the sequence-specific degradation of mRNA. PATTERNRECOGNITION RECEPTOR

(PRR). A host receptor (such as Toll-like receptors) that can sense pathogen-associated molecular patterns and initiate signalling cascades (which involve activation of nuclear factor-κB) that lead to an innate immune response.

to GM-CSF70. Therefore, to expand and prolong the survival of a population of specific T cells, the relatively safe molecule GM-CSF could be used in preference to IL-2, which has considerable dose-limiting toxicity. Longer-lived T cells might reasonably be expected to have greater opportunity to respond to tumour cells. Along similar lines of reasoning, co-stimulatory genes have been used to improve T-cell persistence and activity in attempts to influence disease. A study in which genetic modification was used to achieve expression of the co-stimulatory molecule CD28 by T cells resulted in the restoration of antiviral activity and memory to virus-specific T cells71. An additional factor that is important for determining T-cell lifespan is TELOMERE length, with normal T cells able to undergo up to 50 cell divisions before senescence or cell death. Extensions in T-cell survival time in vitro have been achieved by overexpression of the enzyme telomerase reverse transcriptase by human T cells72, although it still remains to be seen whether it would be possible to modify tumour-specific T cells in this way without compromising their function. More-direct genetic attacks on the cellular apoptotic machinery are also being devised to increase T-cell survival. Overexpression of either of the anti-apoptotic molecules B-cell lymphoma 2 (BCL-2)73,74 or BCL-XL75 resulted in increased survival of T cells following growth-factor withdrawal. In some cases, decreased sensitivity to the important T-cell regulatory pathways involving transforming growth factor-β (TGF-β) and CD95 (also known as FAS) that operate through the apoptotic machinery was also observed. Genetic restriction of blockade of these and similar processes that are inhibitory to tumour-reactive lymphocytes might have an advantage over systemic methods (such as using blocking antibodies) by allowing prolonged blockade without widespread immune dysregulation. In addition to approaches that seek to promote the ability of T cells by increasing the expression of crucial molecules, future strategies could involve the ‘knockdown’ of mRNA encoding negative regulatory proteins such as pro-apoptotic molecules. For example, SMALL INTERFERING RNA technology can be used to reduce the expression of pro-apoptotic molecules such as CD95 that have a role in downregulating T-cell numbers76. Approaches such as these could circumvent premature decreases in T-cell numbers that would reduce antitumour activity. Recent observations regarding the importance of cytokines such as IL-7, IL-12 and IL-15 for the generation of memory T cells might also provide strategies for promoting T-cell persistence and reactivation following tumour recurrence. For example, recombinant cytokines could be administered or could be constitutively produced from transgenes in T cells. In addition, several other mechanisms are also used in the downregulation of immune responses, including the interaction of T-cell-expressed PD1 and CTLA4 with their corresponding ligands (PD1 ligand 1 (PDL1) or PDL2, and CD80 or CD86, respectively) at the cell surface of APCs. There is potential for genetically inhibiting


these pathways to increase T-cell persistence. Indeed, CTLA4 blockade using monoclonal antibodies has been shown to increase immune-mediated rejection of tumours in mice77. So, the effective control of tumours requires that large numbers of potent, specific T cells are generated and maintained. However, of equal importance is the requirement for these T cells to localize to the tumour mass. Improving T-cell tumour-localizing ability

To exert functional effects, large numbers of specific T cells need to traffic to, and penetrate, sites of challenge. In an immune response to challenge with antigen, the expression of chemokines and vascular adhesion molecules is upregulated in challenged tissue in response to pro-inflammatory (danger) signals, and this is often mediated through PATTERNRECOGNITION RECEPTORS (PRRs) at the surface of DCs and NK cells. These signals mediate the trafficking of T cells that express the relevant chemokine receptors and adhesion molecules. In tumours, the appropriate proinflammatory signals, PRR ligands or chemokines might be lacking, and T cells might also lack the appropriate receptors. Intratumoral injection of chemokines such as XC-chemokine ligand 1 (XCL1; also known as lymphotactin) has been shown to increase the localization of T cells to tumours in mice and to increase antitumour responses 78 . However, the difficulty of injecting tumours directly might limit the application of this approach. Nevertheless, many tumour types naturally produce various chemokines, such as CXC-chemokine ligand 8 (CXCL8; also known as IL-8) and CXCL10 (also known as IP10)79. Tumours can also express receptors for chemokines and can use chemokines in this manner to provide autocrine signals for growth or signals for increasing angiogenesis79. The expression of chemokines by tumours, as well as the often growth-promoting properties and the chemotactic qualities of chemokines, makes them an attractive target for directing the trafficking of specific T cells to tumours. However, tumour-reactive T cells do not necessarily express the appropriate receptors for chemokines that are produced at the tumour site. For example, the chemokine CXCL1 (also known as Gro-α) is produced by a large proportion of melanomas80, but its receptor, CXC-chemokine receptor 2 (CXCR2), is expressed by only a small number of T cells81. The feasibility of correcting such deficiencies has been investigated by genetically modifying T cells to express CXCR2 constitutively 82. Following transduction with the gene that encodes CXCR2, T cells acquired the ability to migrate towards CXCL1 in vitro. However, it is not known how the constitutive expression of chemokine receptors achieved using the current retroviral vectors might affect T-cell function or recirculation after T cells have arrived in tumour tissue. Further investigation of this and other aspects of the approach could lead to improved T-cell localization to tumours in vivo.



REVIEWS

CD4+CD25+ REGULATORY T CELL

A specialized type of CD4+ T cell that can suppress the responses of other T cells. These cells provide a crucial mechanism for the maintenance of peripheral self-tolerance and are characterized by expression of CD25 (also known as the α-chain of the interleukin-2 receptor) and the transcription factor forkhead box P3 (FOXP3).

Increasing T-cell resistance to inhibitory factors

The tumour environment can contain not only immuneinhibitory factors, such as TGF-β, but also cytokines, such as IL-10, that promote a TH2-cell response, but this response might not always be sufficient to eliminate the tumour. In addition, chemokines such as CC-chemokine ligand 17 (CCL17; also known as TARC) that attract TH2 cells and regulatory T cells can be present in tumours and can correlate with a poor prognosis83. Strategies that are aimed at countering these suppressive mechanisms should increase T-cell responses to tumours. Indeed, T cells that are transduced with a gene that encodes a dominant-negative form of the TGF-β receptor have increased resistance to TGF-β-mediated inhibition, as

shown by their ability to continue to proliferate and secrete cytokines in response to antigen in the presence of TGF-β84. Furthermore, modification of T cells to produce the TH1 cytokine IL-12 constitutively resulted in the resistance of these T cells to immunosuppressive tumour-derived factors such as TGF-β in vitro85. It is possible that constitutive production of IL-12 by T cells in the tumour in vivo might overcome the inhibitory effects of the environment. As we learn more about the complex interplay between tumour, cytokines, stroma and immune cells, further targets for genetic intervention should become apparent. T cells can also be subject to other forms of inhibition, including by CD4 CD25 REGULATORY T CELLS, which +

+

Table 3 | Summary of T-cell genetic-modification strategies under investigation Approach

Effect

Advantages

Disadvantages

References

Transducing T cells with genes encoding TCR α-chain and β-chain*

Provides specificity

Transduced T cells can recognize intracellular or cell-surface antigens and proliferate in response to immunization

TCRs specific for many HLA types need to be cloned, and mispairing with endogenous TCR chains can occur

13–17

Transducing T cells with a gene encoding a singlechain chimeric receptor*

Provides specificity

Receptors can be relatively easily produced using TAA-specific monoclonal antibodies, and transduced T cells can recognize carbohydrate and glycolipid TAAs

Recognition of antigens derived from intracellular proteins is difficult to achieve

18–27

Transducing T cells with a gene encoding a chimeric receptor that contains co-stimulatory domains*

Increases cytokine production and proliferation

Single transgene fulfils requirements for antigen recognition, cell activation and co-stimulation

Concerns about structural and steric hindrance when combining several domains

38–45

Increasing cytotoxic capabilities of T cells

Increases antitumour activity

Approach does not rely on inherent cytotoxicity of T cells, and circumvents tumour resistance to other mechanisms of destruction

High level of stable expression has not been achieved so far

59,60

Infecting T cells with retrovirus

Delivers suicide gene to tumour

All subsets of T cells can be converted to killers

T cells still need to reach tumour in large numbers, and tumour needs to be receptive to entry of retrovirus

Using dual-specific T cells*

Increases proliferation

Problem of poorly immunogenic TAAs can be circumvented by using a second potent immunogen

Immunogen elicits proliferation of both dual-specific and endogenous mono-specific T cells

Transducing haematopoietic stem cells with TCRs

Provides long-term supply of T cells

Approach counters need to keep transferring T cells

Technically more difficult than manipulation of T cells, and precedent for transformation of haematopoietic stem cells

Using T cells as APCs

Increases proliferation

T cells are easier to produce in large numbers than professional APCs

Not as potent as professional APCs

Transducing T cells with genes encoding growth factors

Increases survival of T cells

T cells can be independent of external growth factors, and fewer T cells are required

Danger of lymphoproliferative disorder or transformation

69,70

Using anti-apoptotic strategies‡

Prolongs survival of T cells

Key molecules common to many inhibitory pathways can be targeted, and fewer T cells are required

Danger of lymphoproliferative disorder or transformation

73–75,128

Transducing T cells with genes encoding chemokine receptors

Provides tumourhoming capacity

Migration of all T-cell types can be redirected more specifically to tumours

Additional pro-inflammatory signals or adhesion molecules might be required

Inducing resistance to tumour-derived inhibitory factors‡

Maintains T-cell activity within tumour

Approach counters some immuneescape mechanisms of tumours

Multiple immune-escape mechanisms could be occurring

Transducing T cells with suicide genes

Provides capacity to eliminate T cells when task is complete

Approach provides safety in the event of autoimmunity or transformation

Current genes cannot eliminate 100% of transduced cells

61

26,63,64

66

127

82

84,85

89,91,129

*Approaches that have been used in early clinical trials. ‡Approaches that are nearing clinical application. (The other approaches, and combinations of approaches, need optimization and further investigation before they can be considered for the treatment of cancer.) APC, antigen-presenting cell; TAA, tumour-associated antigen; TCR, T-cell receptor.

936 | DECEMBER 2005

| VOLUME 5



REVIEWS

Survival and persistence Activation and cytolytic activity

Endothelial cell

BIM siRNA

Erythrocyte Chemokine

Migration T cell Inhibition of activation

Blood vessel

Normal epithelium

Tumour

Chimeric activating receptor

Tumour-associated antigen

Chimeric inhibiting receptor

Specific normal-tissue molecule

Chemokine receptor

Cytokines and lytic molecules

Figure 2 | Schematic representation of a supernatural T cell. In this example of how genetic modification could be used to generate T cells with increased antitumour function, a T cell (green) migrates into a tumour site after ligation of a transgene-encoded chemokine receptor by a tumour-derived chemokine (orange). As the T cell moves towards the tumour, it encounters normal tissue cells that express the targeted tumour-associated antigen. Autoreactivity to normal tissue is prevented by a receptor that is engineered to deliver an inhibitory signal in response to a specific normal-tissue-expressed molecule. On subsequent encounter with tumour-associated antigen at the surface of tumour cells, cytokines and cytolytic molecules are released towards the tumour after triggering through the T-cellexpressed activating receptor alone. Inhibition of T-cell apoptosis is also achieved through transgenic expression of small interfering RNA (siRNA) that blocks expression of the proapoptotic molecule BIM (B-cell lymphoma 2 (BCL-2)-interacting mediator of cell death), leading to prolonged T-cell survival. So, depending on the malignancy, a coordinated strategy of genetic reprogramming of T cells might be devised to satisfy diverse requirements for an effective response to a tumour.

can mediate their effects through secreted factors and contact-dependent mechanisms. As we learn more about the specific receptors and factors that are important in these types of interaction, it might be possible to devise genetic strategies to circumvent inhibition. Indeed, as mentioned earlier, simply redirecting T cells in a non-MHC-restricted manner might bypass some of the regulatory mechanisms that are imposed on naturally occurring T cells. Safety issues

There are several safety issues that are of concern regarding the use of this technology TABLE 3. The generation of highly active T cells with a high proliferative capacity could itself result in pathology. In addition, the potential for T cells to become transformed or gain autoimmune reactivity are concerns. At present, retroviruses are the most efficient type of vector for genetic manipulation of T cells, but these result in genomic insertion of transgenes at locations that are mainly random. Genomic integration of viral vectors can result in dysregulation of growth control in T cells, which could lead to leukaemia or lymphoma. Other genetic-modification methods — including the use of


adeno-associated virus or episomal vectors, which have a restricted integration site or remain unintegrated — might eventually be safer, but these systems await technological advances. So far, leukaemias have been documented to arise in humans from haematopoietic stem cells that have been modified with the gene that encodes the common cytokine-receptor γ-chain (γc)86. In addition, a study involving transduction of mouse haematopoietic stem cells with a nerve-growth-factor reporter gene also resulted in leukaemia in some recipient mice87. However, the conclusions from the mouse haematopoietic-stemcell study were not supported by experimental and clinical evidence that was assembled from 17 other groups of independent investigators88. Nevertheless, it seems prudent to use safety strategies when genetically modifying T cells. Transduction of T cells with genes that encode drug sensitivity is one means to achieve this outcome; in this way, the T cells can be eliminated if adverse reactions occur. HSV tk has been used with moderate success in this way; the drug ganciclovir has been used effectively to deplete HSV-tk-expressing allogeneic lymphocytes following bone-marrow transplantation89. However, depletion is not always complete, and the foreign nature of the HSV protein raises concerns about immunogenicity, because the persistence of transduced T cells might be decreased in immunocompetent hosts. Alternative suicide-gene strategies using human gene products might be non-immunogenic. Indeed, genes that encode chimeric molecules that allow drug-mediated oligomerization of apoptosis-inducing domains have shown potential for eliminating transduced T cells90,91. In these studies, a transgene encoding a drug-inducible oligomerization domain was fused to cytoplasmic portions of death receptors, and T cells were transduced with this construct; elimination of the transduced T cells occurred in the presence of the drug. Concluding remarks

In our quest to develop specific, effective therapies for cancer, genetic-modification strategies present us with the opportunity to harness the immune system against malignant disease. This Review focuses on the genetic modification of T cells; however, it is clear that many cell types can contribute to antitumour immunity, so modification of cells such as NK cells, macrophages and DCs might also result in the production of leukocytes with increased antitumour activity. Combinations of the modifications that are described in this Review might be required to achieve optimal antitumour activity, so transduction of cells with several genes might be required. An example of how several natural limitations might be overcome using a combination of several transgenes is shown in FIG. 2. At present, vector technology limits the number of genes that can be stably inserted into T cells to three or four, but advances in technology will undoubtedly lead to a lifting of this limitation. Indeed, it might soon be possible to use larger genes that contain regulatory



REVIEWS elements, which would allow sequential gene expression, thereby enabling T cells to mediate particular functions at appropriate times. Other technological issues are also of concern for the widespread use of these approaches. At present, genetic modification of T cells is cumbersome, involving ex vivo activation and culture. Not only is this time consuming and expensive (estimated at approximately US$20,000–50,000 per patient), but the T cells that are produced, particularly after longer-term culture, might not be ideally suited to a subsequent life in vivo. However, it might soon be possible to modify T cells

1.

2.

3.

4. 5.

6.

7.

8.

9.

10.

11.

12.

13.

14.

15.

16.

17.

18.

Smyth, M. J., Godfrey, D. I. & Trapani, J. A. A fresh look at tumor immunosurveillance and immunotherapy. Nature Immunol. 2, 293–299 (2001). Rosenberg, S. A., Fox, E. & Churchill, W. H. Spontaneous regression of hepatic metastases from gastric carcinoma. Cancer 29, 472–474 (1972). Cole, W. H. Relationship of causative factors in spontaneous regression of cancer to immunologic factors possibly effective in cancer. J. Surg. Oncol. 8, 391–411 (1976). Rosenberg, S. A. Progress in human tumour immunology and immunotherapy. Nature 411, 380–384 (2001). Interferon α versus chemotherapy for chronic myeloid leukemia: a meta-analysis of seven randomized trials: Chronic Myeloid Leukemia Trialists’ Collaborative Group. J. Natl Cancer Inst. 89, 1616–1620 (1997). Grillo-Lopez, A. J., Hedrick, E., Rashford, M. & Benyunes, M. Rituximab: ongoing and future clinical development. Semin. Oncol. 29, 105–112 (2002). Heslop, H. E. et al. Long-term restoration of immunity against Epstein–Barr virus infection by adoptive transfer of gene-modified virus-specific T lymphocytes. Nature Med. 2, 551–555 (1996). Walter, E. A. et al. Reconstitution of cellular immunity against cytomegalovirus in recipients of allogeneic bone marrow by transfer of T-cell clones from the donor. N. Engl. J. Med. 333, 1038–1044 (1995). Rosenberg, S. A., Yang, J. C. & Restifo, N. P. Cancer immunotherapy: moving beyond current vaccines. Nature Med. 10, 909–915 (2004). Rosenberg, S. A. et al. Immunologic and therapeutic evaluation of a synthetic peptide vaccine for the treatment of patients with metastatic melanoma. Nature Med. 4, 321–327 (1998). Khong, H. T. & Restifo, N. P. Natural selection of tumor variants in the generation of ‘tumor escape’ phenotypes. Nature Immunol. 3, 999–1005 (2002). Dunn, G. P., Old, L. J. & Schreiber, R. D. The immunobiology of cancer immunosurveillance and immunoediting. Immunity 21, 137–148 (2004). Zhao, Y. et al. Primary human lymphocytes transduced with NY-ESO-1 antigen-specific TCR genes recognize and kill diverse human tumor cell lines. J. Immunol. 174, 4415–4423 (2005). Kessels, H. W., Wolkers, M. C., van den Boom, M. D., van der Valk, M. A. & Schumacher, T. N. Immunotherapy through TCR gene transfer. Nature Immunol. 2, 957–961 (2001). Stanislawski, T. et al. Circumventing tolerance to a human MDM2-derived tumor antigen by TCR gene transfer. Nature Immunol. 2, 962–970 (2001). Kuball, J. et al. Cooperation of human tumor-reactive CD4+ and CD8+ T cells after redirection of their specificity by a high-affinity p53A2.1-specific TCR. Immunity 22, 117–129 (2005). Willemsen, R. A. et al. Grafting primary human T lymphocytes with cancer-specific chimeric single chain and two chain TCR. Gene Ther. 7, 1369–1377 (2000). Gross, G., Waks, T. & Eshhar, Z. Expression of immunoglobulin–T-cell receptor chimeric molecules as functional receptors with antibody-type specificity. Proc. Natl Acad. Sci. USA 86, 10024–10028 (1989). This is the original report describing the concept of redirecting T-cell function by genetic modification with chimeric receptors composed of antibody determinants linked to T-cell signalling components. In this study, a mouse T-cell hybridoma line was engineered to react with the hapten trinitrophenol conjugated to tumour cells.

938 | DECEMBER 2005

directly in vivo, as is possible for other tissues92,93, by using vectors that target cell-specific receptors. In summary, genetic modification of T cells can be used to rectify the failings of the immune response to tumours. Elements as diverse as specificity, function and survival can be manipulated to increase the antitumour potential of T cells. With further advances in genetic strategies and technology, it might soon be possible to produce an antitumour immune response with a potency and specificity far beyond that which is possible with the immune system that Mother Nature has provided.

19. Eshhar, Z., Waks, T., Gross, G. & Schindler, D. G. Specific activation and targeting of cytotoxic lymphocytes through chimeric single chains consisting of antibodybinding domains and the γ or ζ subunits of the immunoglobulin and T-cell receptors. Proc. Natl Acad. Sci. USA 90, 720–724 (1993). 20. Hwu, P. et al. Lysis of ovarian cancer cells by human lymphocytes redirected with a chimeric gene composed of an antibody variable region and the Fc receptor γ chain. J. Exp. Med. 178, 361–366 (1993). 21. Altenschmidt, U., Klundt, E. & Groner, B. Adoptive transfer of in vitro-targeted, activated T lymphocytes results in total tumor regression. J. Immunol. 159, 5509–5515 (1997). 22. Hwu, P. et al. In vivo antitumor activity of T cells redirected with chimeric antibody/T-cell receptor genes. Cancer Res. 55, 3369–3373 (1995). This study shows the potential of genetically redirected T cells to influence tumour growth in mice. A gene that encoded a chimeric antibody–Fcγ receptor enabled adoptively transferred mouse T cells to inhibit early intraperitoneal or lung metastases. 23. Pinthus, J. H. et al. Adoptive immunotherapy of prostate cancer bone lesions using redirected effector lymphocytes. J. Clin. Invest. 114, 1774–1781 (2004). 24. Brentjens, R. J. et al. Eradication of systemic B-cell tumors by genetically targeted human T lymphocytes co-stimulated by CD80 and interleukin-15. Nature Med. 9, 279–286 (2003). 25. Mezzanzanica, D. et al. Transfer of chimeric receptor gene made of variable regions of tumor-specific antibody confers anticarbohydrate specificity on T cells. Cancer Gene Ther. 5, 401–407 (1998). 26. Rossig, C., Bollard, C. M., Nuchtern, J. G., Rooney, C. M. & Brenner, M. K. Epstein–Barr virus-specific human T lymphocytes expressing antitumor chimeric T-cell receptors: potential for improved immunotherapy. Blood 99, 2009–2016 (2002). 27. Krause, A. et al. Antigen-dependent CD28 signaling selectively enhances survival and proliferation in genetically modified activated human primary T lymphocytes. J. Exp. Med. 188, 619–626 (1998). 28. Andersen, P. S. et al. A recombinant antibody with the antigen-specific, major histocompatibility complexrestricted specificity of T cells. Proc. Natl Acad. Sci. USA 93, 1820–1824 (1996). 29. Hombach, A. et al. A chimeric receptor that selectively targets membrane-bound carcinoembryonic antigen (mCEA) in the presence of soluble CEA. Gene Ther. 6, 300–304 (1999). 30. Patel, S. D. et al. T-cell killing of heterogenous tumor or viral targets with bispecific chimeric immune receptors. Cancer Gene Ther. 7, 1127–1134 (2000). 31. Yee, C. et al. Melanocyte destruction after antigen-specific immunotherapy of melanoma: direct evidence of T cellmediated vitiligo. J. Exp. Med. 192, 1637–1644 (2000). 32. Overwijk, W. W. et al. Vaccination with a recombinant vaccinia virus encoding a ‘self’ antigen induces autoimmune vitiligo and tumor cell destruction in mice: requirement for CD4+ T lymphocytes. Proc. Natl Acad. Sci. USA 96, 2982–2987 (1999). 33. Phan, G. Q. et al. Cancer regression and autoimmunity induced by cytotoxic T lymphocyte-associated antigen 4 blockade in patients with metastatic melanoma. Proc. Natl Acad. Sci. USA 100, 8372–8377 (2003). 34. Lamers, C. H. et al. Adoptive immuno-gene therapy of cancer with single chain antibody [scFv(Ig)] gene modified T lymphocytes. J. Biol. Regul. Homeost. Agents 18, 134–140 (2004).

| VOLUME 5

35. Chen, W. et al. Reversal in the immunodominance hierarchy in secondary CD8+ T cell responses to influenza A virus: roles for cross-presentation and lysis-independent immunodomination. J. Immunol. 173, 5021–5027 (2004). 36. Lee, P. P. et al. Characterization of circulating T cells specific for tumor-associated antigens in melanoma patients. Nature Med. 5, 677–685 (1999). 37. Haynes, N. M. et al. Redirecting mouse CTL against colon carcinoma: superior signaling efficacy of single-chain variable domain chimeras containing TCR-ζ vs FcεRI-γ. J. Immunol. 166, 182–187 (2001). 38. Hombach, A. et al. Tumor-specific T cell activation by recombinant immunoreceptors: CD3ζ signaling and CD28 costimulation are simultaneously required for efficient IL-2 secretion and can be integrated into one combined CD28/CD3ζ signaling receptor molecule. J. Immunol. 167, 6123–6131 (2001). 39. Maher, J., Brentjens, R. J., Gunset, G., Riviere, I. & Sadelain, M. Human T-lymphocyte cytotoxicity and proliferation directed by a single chimeric TCRζ/CD28 receptor. Nature Biotechnol. 20, 70–75 (2002). 40. Haynes, N. M. et al. Single-chain antigen recognition receptors that costimulate potent rejection of established experimental tumors. Blood 100, 3155–3163 (2002). An important advance in the effectiveness of genetically redirected T cells is described, in which incorporation of a co-stimulatory domain into the cytoplasmic portion of a chimeric gene results in increased antitumour effects and even in the eradication of early subcutaneous tumours or lung metastases in mice. 41. Finney, H. M., Lawson, A. D., Bebbington, C. R. & Weir, A. N. Chimeric receptors providing both primary and costimulatory signaling in T cells from a single gene product. J. Immunol. 161, 2791–2797 (1998). 42. Finney, H. M., Akbar, A. N. & Lawson, A. D. Activation of resting human primary T cells with chimeric receptors: costimulation from CD28, inducible costimulator, CD134, and CD137 in series with signals from the TCRζ chain. J. Immunol. 172, 104–113 (2004). 43. Imai, C. et al. Chimeric receptors with 4-1BB signaling capacity provoke potent cytotoxicity against acute lymphoblastic leukemia. Leukemia 18, 676–684 (2004). 44. Geiger, T. L., Nguyen, P., Leitenberg, D. & Flavell, R. A. Integrated src kinase and costimulatory activity enhances signal transduction through single-chain chimeric receptors in T lymphocytes. Blood 98, 2364–2371 (2001). 45. Fitzer-Attas, C. J., Schindler, D. G., Waks, T. & Eshhar, Z. Harnessing Syk family tyrosine kinases as signaling domains for chimeric single chain of the variable domain receptors: optimal design for T cell activation. J. Immunol. 160, 145–154 (1998). 46. Bertram, E. M., Dawicki, W. & Watts, T. H. Role of T cell costimulation in anti-viral immunity. Semin. Immunol. 16, 185–196 (2004). 47. Vinay, D. S. & Kwon, B. S. Differential expression and costimulatory effect of 4-1BB (CD137) and CD28 molecules on cytokine-induced murine CD8+ TC1 and TC2 cells. Cell. Immunol. 192, 63–71 (1999). 48. Amsen, D. et al. Instruction of distinct CD4 T helper cell fates by different Notch ligands on antigen-presenting cells. Cell 117, 515–526 (2004). 49. Zheng, W. & Flavell, R. A. The transcription factor GATA-3 is necessary and sufficient for TH2 cytokine gene expression in CD4 T cells. Cell 89, 587–596 (1997). 50. Szabo, S. J. et al. A novel transcription factor, T-bet, directs TH1 lineage commitment. Cell 100, 655–669 (2000).



REVIEWS

51. Moretta, A. et al. Activating receptors and coreceptors involved in human natural killer cell-mediated cytolysis. Annu. Rev. Immunol. 19, 197–223 (2001). 52. Ossendorp, F., Mengede, E., Camps, M., Filius, R. & Melief, C. J. Specific T helper cell requirement for optimal induction of cytotoxic T lymphocytes against major histocompatibility complex class II negative tumors. J. Exp. Med. 187, 693–702 (1998). 53. Mattes, J. et al. Immunotherapy of cytotoxic T cellresistant tumors by T helper 2 cells: an eotaxin and STAT6-dependent process. J. Exp. Med. 197, 387–393 (2003). 54. Hombach, A. et al. An entirely humanized CD3 ζ-chain signaling receptor that directs peripheral blood T cells to specific lysis of carcinoembryonic antigen-positive tumor cells. Int. J. Cancer 88, 115–120 (2000). 55. Daly, T. et al. Recognition of human colon cancer by T cells transduced with a chimeric receptor gene. Cancer Gene Ther. 7, 284–291 (2000). 56. Weijtens, M. E., Willemsen, R. A., Valerio, D., Stam, K. & Bolhuis, R. L. Single chain Ig/γ gene-redirected human T lymphocytes produce cytokines, specifically lyse tumor cells, and recycle lytic capacity. J. Immunol. 157, 836–843 (1996). 57. Rossig, C., Bollard, C. M., Nuchtern, J. G., Merchant, D. A. & Brenner, M. K. Targeting of GD2-positive tumor cells by human T lymphocytes engineered to express chimeric T-cell receptor genes. Int. J. Cancer 94, 228–236 (2001). 58. McGuinness, R. P. et al. Anti-tumor activity of human T cells expressing the CC49-ζ chimeric immune receptor. Hum. Gene Ther. 10, 165–173 (1999). 59. Hwu, P. et al. Functional and molecular characterization of tumor-infiltrating lymphocytes transduced with tumor necrosis factor-α cDNA for the gene therapy of cancer in humans. J. Immunol. 150, 4104–4115 (1993). 60. Itoh, Y. et al. Characterization of tumor-necrosis-factorgene-transduced tumor-infiltrating lymphocytes from ascitic fluid of cancer patients: analysis of cytolytic activity, growth rate, adhesion molecule expression and cytokine production. Cancer Immunol. Immunother. 40, 95–102 (1995). 61. Crittenden, M. et al. Pharmacologically regulated production of targeted retrovirus from T cells for systemic antitumor gene therapy. Cancer Res. 63, 3173–3180 (2003). 62. Butz, E. A. & Bevan, M. J. Massive expansion of antigenspecific CD8+ T cells during an acute virus infection. Immunity 8, 167–175 (1998). 63. Heemskerk, M. H. et al. Reprogramming of virus-specific T cells into leukemia-reactive T cells using T cell receptor gene transfer. J. Exp. Med. 199, 885–894 (2004). 64. Kershaw, M. H., Westwood, J. A. & Hwu, P. Dual-specific T cells combine proliferation and antitumor activity. Nature Biotechnol. 20, 1221–1227 (2002). The authors of this report propose a new concept that involves dual-specific T cells to circumvent the poor immunogenicity of many TAAs. By combining tumour reactivity with specificity for a potent immunogen, in this case alloantigen, increased antitumour activity was observed in mice following adoptive transfer of dual-specific cells and allogeneic immunization. 65. Monsurro, V. et al. Kinetics of TCR use in response to repeated epitope-specific immunization. J. Immunol. 166, 5817–5825 (2001). 66. Yang, L., Qin, X. F., Baltimore, D. & Van Parijs, L. Generation of functional antigen-specific T cells in defined genetic backgrounds by retrovirus-mediated expression of TCR cDNAs in hematopoietic precursor cells. Proc. Natl Acad. Sci. USA 99, 6204–6209 (2002). This study shows the feasibility of providing a constant renewable source of T cells that are specific for ovalbumin by genetically modifying haemotopoietic stem cells. The introduction of tumour-specific TCRs into progenitor cells in this way could provide T cells that can affect existing malignant disease, as well as monitor the body for relapse in the long-term. 67. Cooper, L. J. et al. Enhanced antilymphoma efficacy of CD19-redirected influenza MP1-specific CTLs by cotransfer of T cells modified to present influenza MP1. Blood 105, 1622–1631 (2005). 68. Kim, J. V., Latouche, J. B., Riviere, I. & Sadelain, M. The ABCs of artificial antigen presentation. Nature Biotechnol. 22, 403–410 (2004). 69. Liu, K. & Rosenberg, S. A. Transduction of an IL-2 gene into human melanoma-reactive lymphocytes results in their continued growth in the absence of exogenous IL-2 and maintenance of specific antitumor activity. J. Immunol. 167, 6356–6365 (2001).

70. Cheng, L. E., Ohlen, C., Nelson, B. H. & Greenberg, P. D. Enhanced signaling through the IL-2 receptor in CD8+ T cells regulated by antigen recognition results in preferential proliferation and expansion of responding CD8+ T cells rather than promotion of cell death. Proc. Natl Acad. Sci. USA 99, 3001–3006 (2002). 71. Topp, M. S. et al. Restoration of CD28 expression in CD28– CD8+ memory effector T cells reconstitutes antigen-induced IL-2 production. J. Exp. Med. 198, 947–955 (2003). 72. Hooijberg, E. et al. Immortalization of human CD8+ T cell clones by ectopic expression of telomerase reverse transcriptase. J. Immunol. 165, 4239–4245 (2000). 73. Lin, C. M. & Wang, F. H. Selective modification of antigenspecific CD4+ T cells by retroviral-mediated gene transfer and in vitro sensitization with dendritic cells. Clin. Immunol. 104, 58–66 (2002). 74. Charo, J. et al. Bcl-2 overexpression enhances tumorspecific T-cell survival. Cancer Res. 65, 2001–2008 (2005). Increased survival is imparted to tumour-specific T cells by a transgene encoding the anti-apoptotic molecule BCL-2. Such apoptosis-resistant T cells are shown to have increased antitumour capacity in mice. 75. Eaton, D., Gilham, D. E., O’Neill, A. & Hawkins, R. E. Retroviral transduction of human peripheral blood lymphocytes with Bcl-XL promotes in vitro lymphocyte survival in pro-apoptotic conditions. Gene Ther. 9, 527–535 (2002). 76. Dotti, G. et al. Human cytotoxic T lymphocytes with reduced sensitivity to Fas-induced apoptosis. Blood 105, 4677–4684 (2005). 77. Leach, D. R., Krummel, M. F. & Allison, J. P. Enhancement of antitumor immunity by CTLA-4 blockade. Science 271, 1734–1736 (1996). 78. Huang, H., Li, F., Gordon, J. R. & Xiang, J. Synergistic enhancement of antitumor immunity with adoptively transferred tumor-specific CD4+ and CD8+ T cells and intratumoral lymphotactin transgene expression. Cancer Res. 62, 2043–2051 (2002). 79. Homey, B., Muller, A. & Zlotnik, A. Chemokines: agents for the immunotherapy of cancer? Nature Rev. Immunol. 2, 175–184 (2002). 80. Haghnegahdar, H. et al. The tumorigenic and angiogenic effects of MGSA/GRO proteins in melanoma. J. Leukoc. Biol. 67, 53–62 (2000). 81. Chuntharapai, A., Lee, J., Hebert, C. A. & Kim, K. J. Monoclonal antibodies detect different distribution patterns of IL-8 receptor A and IL-8 receptor B on human peripheral blood leukocytes. J. Immunol. 153, 5682–5688 (1994). 82. Kershaw, M. H. et al. Redirecting migration of T cells to chemokine secreted from tumors by genetic modification with CXCR2. Hum. Gene Ther. 13, 1971–1980 (2002). This is the original description of the manipulation of the migration properties of T cells by insertion of a gene that encodes a receptor for the tumourassociated chemokine CXCL1. Although only carried out in vitro, this study raises the possibility of genetically redirecting T cells to penetrate tumours in greater numbers. 83. Iellem, A. et al. Unique chemotactic response profile and specific expression of chemokine receptors CCR4 and CCR8 by CD4+CD25+ regulatory T cells. J. Exp. Med. 194, 847–853 (2001). 84. Bollard, C. M. et al. Adapting a transforming growth factor β-related tumor protection strategy to enhance antitumor immunity. Blood 99, 3179–3187 (2002). A novel method of enabling tumour- or virus-specific T cells to resist the tumour-derived inhibitory factor TGF-β is presented in this in vitro study. Many tumour types secrete TGF-β, so this strategy might allow tumour-specific T cells to maintain their tumour-destructive properties in the face of an otherwise immunosuppressive environment. 85. Wagner, H. J. et al. A strategy for treatment of Epstein– Barr virus-positive Hodgkin’s disease by targeting interleukin 12 to the tumor environment using tumor antigen-specific T cells. Cancer Gene Ther. 11, 81–91 (2004). 86. Hacein-Bey-Abina, S. et al. A serious adverse event after successful gene therapy for X-linked severe combined immunodeficiency. N. Engl. J. Med. 348, 255–256 (2003). 87. Li, Z. et al. Murine leukemia induced by retroviral gene marking. Science 296, 497 (2002). 88. Bonini, C. et al. Safety of retroviral gene marking with a truncated NGF receptor. Nature Med. 9, 367–369 (2003).


89. Bonini, C. et al. HSV-TK gene transfer into donor lymphocytes for control of allogeneic graft-versusleukemia. Science 276, 1719–1724 (1997). This article deals with controlling the persistence and activity of tumour-reactive T cells (after they have been transferred to patients) by insertion of a suicide gene into the T cells. If the genetically modified T cells begin to damage normal tissue or become malignant themselves, a drug can be administered that eliminates genetically modified T cells without harming normal T cells. 90. Straathof, K. C. et al. An inducible caspase 9 safety switch for T-cell therapy. Blood 105, 4247–4254 (2005). 91. Thomis, D. C. et al. A Fas-based suicide switch in human T cells for the treatment of graft-versus-host disease. Blood 97, 1249–1257 (2001). 92. Rettinger, S. D. et al. Liver-directed gene therapy: quantitative evaluation of promoter elements by using in vivo retroviral transduction. Proc. Natl Acad. Sci. USA 91, 1460–1464 (1994). 93. Naldini, L. et al. In vivo gene delivery and stable transduction of nondividing cells by a lentiviral vector. Science 272, 263–267 (1996). 94. Speiser, D. E. et al. Self antigens expressed by solid tumors do not efficiently stimulate naive or activated T cells: implications for immunotherapy. J. Exp. Med. 186, 645–653 (1997). 95. Nomura, T. & Sakaguchi, S. Naturally arising CD25+CD4+ regulatory T cells in tumor immunity. Curr. Top. Microbiol. Immunol. 293, 287–302 (2005). 96. Gorelik, L. & Flavell, R. A. Transforming growth factor-β in T-cell biology. Nature Rev. Immunol. 2, 46–53 (2002). 97. Salazar-Onfray, F. Interleukin-10: a cytokine used by tumors to escape immunosurveillance. Med. Oncol. 16, 86–94 (1999). 98. Cebon, J. et al. Two Phase I studies of low dose recombinant human IL-12 with Melan-A and influenza peptides in subjects with advanced malignant melanoma. Cancer Immun. 3, 7–25 (2003). 99. Sadanaga, N. et al. Dendritic cell vaccination with MAGE peptide is a novel therapeutic approach for gastrointestinal carcinomas. Clin. Cancer Res. 7, 2277–2284 (2001). 100. Lonchay, C. et al. Correlation between tumor regression and T cell responses in melanoma patients vaccinated with a MAGE antigen. Proc. Natl Acad. Sci. USA 101 (Suppl. 2), 14631–14638 (2004). 101. Peterson, A. C., Harlin, H. & Gajewski, T. F. Immunization with Melan-A peptide-pulsed peripheral blood mononuclear cells plus recombinant human interleukin-12 induces clinical activity and T-cell responses in advanced melanoma. J. Clin. Oncol. 21, 2342–2348 (2003). 102. Lau, R. et al. Phase I trial of intravenous peptide-pulsed dendritic cells in patients with metastatic melanoma. J. Immunother. 24, 66–78 (2001). 103. Murphy, G. P. et al. Infusion of dendritic cells pulsed with HLA-A2-specific prostate-specific membrane antigen peptides: a Phase II prostate cancer vaccine trial involving patients with hormone-refractory metastatic disease. Prostate 38, 73–78 (1999). 104. Morse, M. A. et al. A Phase I study of active immunotherapy with carcinoembryonic antigen peptide (CAP-1)-pulsed, autologous human cultured dendritic cells in patients with metastatic malignancies expressing carcinoembryonic antigen. Clin. Cancer Res. 5, 1331–1338 (1999). 105. Kono, K. et al. Dendritic cells pulsed with HER-2/neuderived peptides can induce specific T-cell responses in patients with gastric cancer. Clin. Cancer Res. 8, 3394–3400 (2002). 106. Scheibenbogen, C. et al. Phase 2 trial of vaccination with tyrosinase peptides and granulocyte–macrophage colonystimulating factor in patients with metastatic melanoma. J. Immunother. 23, 275–281 (2000). 107. Slingluff, C. L. Jr et al. Clinical and immunologic results of a randomized Phase II trial of vaccination using four melanoma peptides either administered in granulocyte–macrophage colony-stimulating factor in adjuvant or pulsed on dendritic cells. J. Clin. Oncol. 21, 4016–4026 (2003). 108. Sato, Y. et al. A Phase I trial of cytotoxic T-lymphocyte precursor-oriented peptide vaccines for colorectal carcinoma patients. Br. J. Cancer 90, 1334–1342 (2004). 109. Gulley, J. L. et al. Combining a recombinant cancer vaccine with standard definitive radiotherapy in patients with localized prostate cancer. Clin. Cancer Res. 11, 3353–3362 (2005). 110. Vonderheide, R. H. et al. Vaccination of cancer patients against telomerase induces functional antitumor CD8+ T lymphocytes. Clin. Cancer Res. 10, 828–839 (2004).



REVIEWS

111. Knutson, K. L., Schiffman, K., Cheever, M. A. & Disis, M. L. Immunization of cancer patients with a HER-2/neu, HLA-A2 peptide, p369–377, results in short-lived peptidespecific immunity. Clin. Cancer Res. 8, 1014–1018 (2002). 112. Jager, E. et al. Induction of primary NY-ESO-1 immunity: CD8+ T lymphocyte and antibody responses in peptidevaccinated patients with NY-ESO-1+ cancers. Proc. Natl Acad. Sci. USA 97, 12198–12203 (2000). 113. Khleif, S. N. et al. A Phase I vaccine trial with peptides reflecting ras oncogene mutations of solid tumors. J. Immunother. 22, 155–165 (1999). 114. Gulley, J. et al. Phase I study of a vaccine using recombinant vaccinia virus expressing PSA (rV-PSA) in patients with metastatic androgen-independent prostate cancer. Prostate 53, 109–117 (2002). 115. von Mehren, M. et al. Pilot study of a dual gene recombinant avipox vaccine containing both carcinoembryonic antigen (CEA) and B7.1 transgenes in patients with recurrent CEA-expressing adenocarcinomas. Clin. Cancer Res. 6, 2219–2228 (2000). 116. Zhu, M. Z., Marshall, J., Cole, D., Schlom, J. & Tsang, K. Y. Specific cytolytic T-cell responses to human CEA from patients immunized with recombinant avipox–CEA vaccine. Clin. Cancer Res. 6, 24–33 (2000). 117. Lienard, D. et al. Ex vivo detectable activation of Melan-Aspecific T cells correlating with inflammatory skin reactions in melanoma patients vaccinated with peptides in IFA. Cancer Immun. 4, 4–23 (2004). 118. Svane, I. M. et al. Vaccination with p53-peptide-pulsed dendritic cells, of patients with advanced breast cancer: report from a Phase I study. Cancer Immunol. Immunother. 53, 633–641 (2004).

940 | DECEMBER 2005

119. Hersey, P. et al. Phase I/II study of treatment with dendritic cell vaccines in patients with disseminated melanoma. Cancer Immunol. Immunother. 53, 125–134 (2004). 120. Cathcart, K. et al. A multivalent bcr–abl fusion peptide vaccination trial in patients with chronic myeloid leukemia. Blood 103, 1037–1042 (2004). 121. Haynes, N. M. et al. Rejection of syngeneic colon carcinoma by CTLs expressing single-chain antibody receptors codelivering CD28 costimulation. J. Immunol. 169, 5780–5786 (2002). 122. Teng, M. W., Kershaw, M. H., Moeller, M., Smyth, M. J. & Darcy, P. K. Immunotherapy of cancer using systemically delivered gene-modified human T lymphocytes. Hum. Gene Ther. 15, 699–708 (2004). 123. Kershaw, M. H. et al. Generation of gene-modified T cells reactive against the angiogenic kinase insert domaincontaining receptor (KDR) found on tumor vasculature. Hum. Gene Ther. 11, 2445–2452 (2000). 124. Ren-Heidenreich, L., Hayman, G. T. & Trevor, K. T. Specific targeting of EGP-2+ tumor cells by primary lymphocytes modified with chimeric T cell receptors. Hum. Gene Ther. 11, 9–19 (2000). 125. Yun, C. O., Nolan, K. F., Beecham, E. J., Reisfeld, R. A. & Junghans, R. P. Targeting of T lymphocytes to melanoma cells through chimeric anti-GD3 immunoglobulin T-cell receptors. Neoplasia 2, 449–459 (2000). 126. Lamers, C. H., Willemsen, R. A., Luider, B. A., Debets, R. & Bolhuis, R. L. Protocol for gene transduction and expansion of human T lymphocytes for clinical immunogene therapy of cancer. Cancer Gene Ther. 9, 613–623 (2002).

| VOLUME 5

127. Cooper, L. J. et al. Enhanced anti-lymphoma efficacy of CD19 redirected influenza MP1-specific CTL’s by co-transfer of T-cells modified to present influenza MP1. Blood 105, 1622–1631 (2004). 128. Dotti, G. et al. Human cytotoxic T-lymphocytes with reduced sensitivity to Fas-induced apoptosis. Blood 105, 4677–4684 (2005). 129. Straathof, K. C. et al. An inducible caspase 9 safety switch for T-cell therapy. Blood 105, 4247–4254 (2005).

Acknowledgements We acknowledge support from the National Health and Medical Research Council (Australia), The Cancer Council Victoria (Australia) and the Susan G. Komen Breast Cancer Foundation (United States). We apologize to colleagues whose work might have been cited only indirectly through reviews, owing to space constraints.


Online links DATABASES The following terms in this article are linked online to: Entrez Gene: http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=gene 4-1BB | BCL-2 | BCL-XL | CD3ζ | CD4 | CD28 | CD95 | CD95 ligand | CTLA4 | CXCR2 | DAP12 | GATA3 | ICOS | LCK | NKG2A | OX40 | PD1 | SYK | TGF-β | TNF Access to this interactive links box is free online.



REVIEWS

IMMUNITY BY UBIQUITYLATION: A REVERSIBLE PROCESS OF MODIFICATION Yun-Cai Liu*, Josef Penninger ‡ and Michael Karin§ Abstract | The conjugation of ubiquitin, a 76-amino-acid peptide, to a protein substrate provides a tag that either marks the labelled protein for degradation or modulates its function. The process of protein ubiquitylation — which is catalysed by coordinated enzymatic reactions that are mediated by enzymes known as E1, E2 and E3 — has an important role in the modulation of immune responses. Importantly, protein ubiquitylation is a reversible process, and removal of ubiquitin molecules is mediated by de-ubiquitylating enzymes: for example, A20, which has been implicated in the regulation of immune responses. In addition, the conjugation of ubiquitin-like molecules, such as ISG15 (interferon-stimulated protein of 15 kDa), to proteins is also involved in immune regulation. This Review covers recent progress in our understanding of protein ubiquitylation in the immune system.

*Division of Cell Biology, La Jolla Institute for Allergy and Immunology, San Diego, California 92121, USA. ‡ Institute for Molecular Biotechnology of the Austrian Academy of Sciences, Dr. Bohr-Gasse 3, 1030 Vienna, Austria. § Laboratory of Gene Regulation and Signal Transduction, Department of Pharmacology, School of Medicine, University of California at San Diego, La Jolla, California 92093, USA. Correspondence to Y.-C.L. e-mail: [email protected] doi:10.1038/nri1731

The process of tagging the 76-amino-acid peptide ubiquitin to a protein substrate — known as protein ubiquitylation — was uncovered in the late 1970s and early 1980s and was implicated in the breakdown of harmful cellular proteins in an ATP-dependent manner1. This discovery of a non-lysosomal pathway of protein degradation led to Aaron Ciechanover, Avram Hershko and Irwin Rose being awarded the Nobel Prize in Chemistry for 2004. Following the initial, groundbreaking experiments, scientists then defined the cascade of reactions that mediates ubiquitin conjugation. This includes at least three steps (FIG. 1a): first, the carboxy (C)-terminal glycine residue of ubiquitin is activated by a ubiquitin-activating enzyme, known as enzyme 1 (E1), resulting in the formation of a highenergy THIOESTER BOND between the glycine residue of ubiquitin and the cysteine residue of the active site of E1; second, the activated ubiquitin is transferred to the cysteine residue of the active site of a ubiquitinconjugating enzyme (known as E2); and third, the ubiquitin–E2 complex is recruited to a third enzyme, a ubiquitin–protein ligase (known as E3), which


specifically binds a protein substrate and facilitates the transfer of ubiquitin from E2 to a lysine residue in the substrate2,3. E3 ligases can generally be classified into two families (FIG. 1b): RING (really interesting new gene)type E3 ligases, and HECT (homologous to the E6-associated protein C terminus)-type E3 ligases. The family of RING-type E3 ligases can be further subdivided into two groups: single-protein E3 ligases, such as CBL (Casitas B-lineage lymphoma), in which the RING-finger domain and substrate-recruiting domains are found in one polypeptide; and multisubunit E3 ligases BOX 1. For example, SCF complexes (S-phase kinase-associated protein 1 (SKP1)–cullin-1 (CUL1)–F-box-protein complexes) are multisubunit E3 ligases in which the substrate-recruiting component, the FBOX PROTEIN, associates with the scaffold proteins SKP1 and CUL1 and with the RING-fingerdomain-only protein RBX1 (RING-box-1; also known as ROC1). In these complexes, RBX1 binds ubiquitinloaded E2, and the F-box protein specifically recruits the protein substrate and helps to transfer ubiquitin



REVIEWS from E2 to the substrate. In both single-protein and multisubunit E3 ligases, the protein-recruiting domain or subunit is the crucial determinant of the specificity of the ubiquitylation process, which occurs through protein–protein interactions that are well defined at a biochemical level. A substrate can be tagged with a single ubiquitin molecule, a process that is known as monoubiquitylation, or an elongated ubiquitin chain can be formed Monoubiquitylated target protein

a

ATP Ub

Substrate protein ADP + PPi

Ub

E1 Ub

E2 Ub

E1

+ Ub

DUB

E3 E2 Ub Ub Ub

Ub

Ub

DUB Ub + Ub

Ub

Ub

Polyubiquitylated target protein

b TKB RING

CBL-B

RING Zn Zn Zn Zn

TRAF Single-protein RING-type E3 ligases

RING RING ROQ

GRAIL

ROQ Zn

Roquin

SOCS

Substrate-recruiting component of ECScomplex multisubunit RING-type E3 ligases

OTU Zn Zn Zn Zn

A20

OTU domain is a de-ubiquitylating domain, and zinc-finger domains function as RINGfinger-like domains

C2 WW WW WW WW HECT

ITCH

HECT-type E3 ligase

SH2

SOCS box

Proline-rich sequence

Coiled-coil domain

Transmembrane domain

Figure 1 | The ubiquitin-conjugation cycle and the E3 ligases. a | Ubiquitin is activated by a ubiquitin-activating enzyme (known as enzyme 1, E1) in the presence of ATP and becomes bound to this E1. The activated ubiquitin molecule is then transferred to a ubiquitin-conjugating enzyme (known as E2). Ubiquitin–protein ligases (known as E3 ligases) recruit both the ubiquitin–E2 complex and the substrate protein, and they help to transfer ubiquitin from E2 to the substrate. Substrate-conjugated ubiquitin is removed by de-ubiquitylating enzymes (DUBs). b | A schematic representation of several RING (really interesting new gene)-type and HECT (homologous to the E6-associated protein carboxy terminus)-type E3 ligases is shown. RINGtype E3 ligases can be divided into those that contain both the RING-finger domain and the substrate-recruiting domains (which are known as single-protein RING-type E3 ligases) and those that are multisubunit complexes composed of scaffold proteins, substrate-recruiting proteins and a RING-finger-domain-only protein. For example, suppressor of cytokine signalling (SOCS) proteins function as the substrate-recruiting component of the E3 ligase known as the ECS complex (elongin-B–elongin-C–cullin-5–SOCS-protein complex). A20 functions both as an E3 ligase — through its zinc finger (Zn) domain, which both recruits ubiquitin–E2 complexes and transfers the ubiquitin molecule to the substrate — and as a DUB, through its OTU (ovarian tumour) domain. For further details, see the main text. C2 domain, protein-kinase-C-related domain 2; CBL-B, Casitas B-lineage lymphoma B; GRAIL, gene related to anergy in lymphocytes; ITCH, itchy; PPi, pyrophosphate; ROQ domain, roquin domain; SH2 domain, SRChomology-2 domain; TKB domain, protein-tyrosine-kinase-binding domain; TRAF, tumournecrosis-factor-receptor-associated factor; WW domain, domain that contains two conserved tryptophan residues.

942 | DECEMBER 2005

| VOLUME 5

after the initial conjugation of a ubiquitin molecule, resulting in polyubiquitylation of the substrate 3 . Monoubiquitylation can occur on a single lysine residue or on several lysine residues in a substrate, leading to multiple monoubiquitylation. The ubiquitin molecules in a polyubiquitin chain are generally linked through the lysine residue at position 48 or 63 (and are known as K48- and K63-linked polyubiquitin chains, respectively); however, other lysine residues in a ubiquitin molecule have been shown to participate in linkage. Interestingly, the different types of polyubiquitin chain have different effects on the substrate. Although ubiquitin was originally proposed to deliver a ‘kiss of death’, targeting the tagged protein to the cellular waste-disposal machinery (that is, the 26S proteasome), recent studies have shown that ubiquitylation of proteins has a broad impact on many cellular processes, including modification of protein function, facilitation of cell-surfacereceptor turnover and control of gene transcription4. For example, monoubiquitylation is involved in downmodulation of receptor expression through the endosomal–lysosomal pathway 5, and K63-linked polyubiquitylation modulates protein–protein interactions6. By contrast, K48-linked polyubiquitylation targets substrates for proteasomal degradation. In addition to ubiquitin, ubiquitin-like molecules, such as ISG15 (interferon (IFN)-stimulated protein of 15 kDa), also participate in similar conjugation reactions but with different functional outcomes. Importantly, analogous to protein phosphorylation and dephosphorylation processes, conjugation to ubiquitin or ISG15 is a reversible process, and substrate-linked ubiquitin and ISG15 molecules can be removed by de-ubiquitylating and de-ISGylating enzymes, respectively. From the beginning, ubiquitin has been closely associated with the immune system, because it was originally identified as a lymphocyte-differentiationpromoting factor 7 . However, during the past few years, genetic and biochemical studies have provided evidence that protein ubiquitylation is of fundamental importance in the regulation of both the innate and the adaptive immune system, with roles in the control of immune tolerance, the differentiation of T cells, and the intracellular signal transduction that is induced by antigen, cytokines or Toll-like receptor (TLR) ligands. This Review discusses our current understanding of how conjugation to ubiquitin or the ubiquitin-like molecule ISG15, as well as de-conjugation by specific proteases, affects diverse immunological processes. Ubiquitylation and immune tolerance

The immune system has evolved to mount robust responses to invading pathogens while not destroying (that is, being tolerant of ) self-tissues. T-cell tolerance to self-antigens is generated through both central and peripheral mechanisms8,9. Central tolerance encompasses the processes by which T-cell recognition of self-antigen in the thymus renders the T cells tolerant to self-antigens; these processes



REVIEWS

Box 1 | Multisubunit RING-type E3 ligases Members of the multisubunit RING (really interesting new gene)-type E3 ligase subfamily of E3 ligases consist of several subunits, with a RING-finger-domainonly protein (RING-box-1 (RBX1) or RBX2) as the catalytic core that recruits a ubiquitin–E2 complex. RBX proteins associate with one of the five members of the cullin (CUL) family of scaffold proteins (CUL1, CUL2, CUL3, CUL4 and CUL5). The adaptor protein SKP1 (S-phase kinase-associated protein 1) bridges CUL1 with an F-box protein, which binds the protein substrate for ubiquitylation to form an SCF complex (SKP1–CUL1–F-box-protein complex). By contrast, the adaptor complex elongin-B–elongin-C associates with CUL5 and a SOCS (suppressor of cytokine signalling) protein, which recruits the protein substrate, to form an ECS complex (elongin-B–elongin-C–CUL5–SOCS-protein complex). In both types of multisubunit E3 ligase, the substrate-binding protein (that is, the F-box protein or the SOCS protein) determines the substrate specificity of the ubiquitylation process, which occurs, in most cases, by protein–protein interactions that are chemically well defined. One of the best-characterized SCF complexes is SCFβ-TRCP, in which the F-boxcontaining subunit is β-transducin repeat-containing protein (β-TRCP). The WD40 repeats in β-TRCP enable it to bind phosphorylated IκB (inhibitor of nuclear factor-κB, NF-κB). SCFβ-TRCP induces polyubiquitylation and, subsequently, degradation of IκB. This is an essential step for the release of NF-κB from the IκB–NF-κB complex, and it leads to the nuclear translocation and transcriptional activation of NF-κB. SOCS1 is a subunit of an ECS complex, and it binds Janus kinases (JAKs) through its SRC homology 2 (SH2) domain and the phosphorylated tyrosine residues of the JAKs. The ECSSOCS1 complex promotes polyubiquitylation and degradation of JAKs.

include clonal deletion and the generation of CD4 CD25 REGULATORY T T  CELLS. By contrast, peripheral tolerance encompasses the mechanisms that lead to T-cell tolerance if a T cell first encounters self-antigen outside the thymus; these include ignorance, deletion and anergy induction. +

+

Reg

THIOESTER BOND

A high-energy chemical bridge, such as that between ubiquitin and the active-site cysteine residue of E1. This chemical linkage is initially catalysed by adenylation of the glycine residue at the carboxy terminus of ubiquitin, through the hydrolysis of ATP, and then by linkage of this adenylated residue to the thiol group of the active-site cysteine residue of E1. The activated ubiquitin molecule is then transferred to E2, to which it is attached by a similar thioester bond. FBOX PROTEIN

The F box is a protein module of ∼50 amino acids that interacts with SKP1 (S-phase kinase-associated protein 1) to form an SCF complex (SKP1–cullin-1–F-box-protein complex), which is an E3 ligase. F-box-containing proteins also have protein-interacting domains for recruiting the substrate protein.

Central tolerance. It has been shown that a wide range of peripheral-tissue antigens — such as insulin, thyroglobulin and myelin proteolipid protein — are promiscuously expressed by medullary thymic epithelial cells (mTECs). Thymocytes encountering these antigens at the surface of mTECs are then subject to central-tolerance mechanisms and become tolerant to those antigens10. Genetic studies of a rare human autoimmune disease known as AUTOIMMUNE POLYENDOCRINOPATHYCANDIDIASISECTODERMALDYSTROPHY SYNDROME (APECED) underscore the importance of central tolerance11,12. The gene that is responsible for APECED has been identified as the autoimmune regulator gene (AIRE), and more than 50 linked mutations have been mapped in human patients. AIRE is highly expressed in the thymus, particularly by mTECs13. Targeting of the Aire gene in mice causes the development of spontaneous multi-organ autoimmunity, similar to that of human patients with APECED14,15. Notably, the expression of peripheraltissue-specific proteins such as prepro-insulin and cytochrome p450, which are known targets of autoreactive T cells in human patients with APECED, is decreased in AIRE-deficient mTECs in mice 14. Deficiency in AIRE, however, did not change the number of thymocytes, the proportion of thymocytes that constitute particular subsets or the activation


status of thymocytes, indicating that AIRE does not affect thymocyte development. So, AIRE seems to have an important role in central tolerance through regulating the promiscuous expression of peripheral-tissue self-antigens by mTECs (FIG. 2a). DNA-microarray analysis further indicates that AIRE affects the expression by mTECs of genes that encode molecules involved in intrathymic cell migration, such as MHC molecules and certain chemokines16. AIRE is a 58 kDa protein that, as well as other domains, contains two plant homeodomains (PHDs), which are related to RING-finger domains. AIRE is structurally related to the SP100 SPECKLED PROTEIN OF 100 kDa FAMILY OF TRANSCRIPTIONAL COACTIVATORS, and consistent with this, AIRE contains a nuclear-localization signal and translocates to the nucleus17,18. PHDs have been identified in many proteins, some of which have been shown to have E3-ligase activity. For example, PHD-containing viral proteins are involved in downmodulation of the expression of cell-surface MHC or co-stimulatory molecules19. In vitro biochemical studies have shown that the PHDs of AIRE have E3-ligase activity, and this is abolished by mutation of the conserved cysteine residue in these RING-finger-like motifs20. The transactivating activity of AIRE also depends on the functionality of the PHDs. This indicates that AIRE might regulate central tolerance through its E3-ligase activity. But how this E3-ligase activity is coupled to regulation of gene transcription is still unclear. Furthermore, that AIRE has E3-ligase activity is disputed by an independent research group21. During thymic selection, strong recognition of self-antigen by T cells (that is, with high affinity or high avidity) might trigger the apoptotic pathway and, subsequently, cause cell death22. By contrast, weak engagement of antigen (that is, with low affinity or low avidity) might not be able to trigger survival signals, resulting in death by neglect23. Only T cells that receive signals that engage the survival pathway but do not trigger apoptosis are positively selected and enter the peripheral T-cell pool. Two key regulators in promoting thymic negative selection are the orphan nuclear receptor NUR77 and the pro-apoptotic protein BIM (B-cell lymphoma 2 (BCL-2)-interacting mediator of cell death)24,25. Loss of BIM results in thymocyte resistance to CD3-specific-antibody-induced cell death and to T-cell receptor (TCR)-mediated negative selection25. In a different cell system, BIM was found to be a substrate for CBL26, a well-characterized RINGtype E3 ligase that is implicated in diverse signalling pathways27. CBL is a 120 kDa protein that, as well as other domains, contains a RING-finger domain. There are three mammalian CBL homologues: CBL, CBL-B and CBL3 (also known as CBL-C)27. CBL functions both as an adaptor molecule, forming complexes with several crucial signalling molecules, and as an E3 ligase, promoting the conjugation of ubiquitin to its binding partners. So far, several proteins have been found to be ubiquitylated by the E3-ligase activity of



REVIEWS

a

CUL

(cullin). Cullin-family proteins (CUL1, CUL2, CUL3, CUL4 and CUL5) directly interact with the RING (really interesting new gene)-fingerdomain-only protein RBX1 (RING-box-1). The CUL–RBX1 complex is the core module of a series of multisubunit E3 ligases.

mTEC

γ

ε

δ

TCR ε

ξξ

SP100 FAMILY OF TRANSCRIPTIONAL COACTIVATORS

The nuclear-matrix-associated protein SP100 (speckled protein of 100 kDa) belongs to a family of related proteins that contain nuclear-localization signals, dimerization domains and DNA-binding domains. They interact with other transcription factors to co-activate gene transcription. IONOMYCIN

A divalent calcium ionophore that is widely used as a tool to investigate the role of intracellular calcium in cellular processes. COMPLETE FREUND’S ADJUVANT

(CFA). A mixture of mycobacterial lysate and mineral oil. When animals are immunized with antigen emulsified in CFA, they induce strong immune responses to the antigen.

CD28 Ub

Thymocyte

(APECED). A rare human autoimmune disorder that is inherited in an autosomal recessive manner and is characterized by various endocrine deficiencies, chronic mucocutaneous candidiasis and several ectodermal dystrophies. It is caused by multiple mutations in the gene that encodes autoimmune regulator (AIRE).

944 | DECEMBER 2005

APC Peptide–MHC complex (high affinity for TCR)

(TReg cell). A specialized type of CD4+ T cell that can suppress the responses of other T cells. These cells provide a crucial mechanism for the maintenance of peripheral self-tolerance, and they are characterized by the expression of CD25 (also known as the α-chain of the interleukin-2 receptor) and the transcription factor forkhead box P3 (FOXP3). AUTOIMMUNE POLYENDOCRINOPATHY CANDIDIASISECTODERMAL DYSTROPHY SYNDROME

b

TSAs

AIRE

CD3

CD4+CD25+ REGULATORY T CELL

Nuclear membrane

TSAs

T cell Ca2+

NUR77

CBL-B BIM

ITCH GRAIL

Thymocyte apoptosis

PLC-γ1 Ub PKC-θ Ub

Nucleus Roquin

Anergy

ICOS IL-21 mRNA

Figure 2 | Protein ubiquitylation in T-cell tolerance. a | Autoimmune regulator (AIRE), a RING (really interesting new gene)-type E3 ligase, is implicated in expression of tissue-specific antigens (TSAs) by medullary thymic epithelial cells (mTECs), which present antigens (including TSAs) to developing thymocytes to induce central tolerance to those antigens. In developing thymocytes, signalling through the T-cell receptor (TCR) causes negative selection by inducing thymocyte apoptosis through expression of NUR77 and BIM (B-cell-lymphoma-2-interacting mediator of cell death). The E3 ligase CBL (Casitas B-lineage lymphoma) can function as an E3 ligase for BIM. b | In mature peripheral T cells, ligation of the TCR in the absence of the appropriate co-stimulatory signals results in augmented expression of three E3 ligases — CBL-B, ITCH (itchy) and GRAIL (gene related to anergy in lymphocytes) — and this promotes the conjugation of ubiquitin to phospholipase C-γ1 (PLC-γ1) and protein kinase C-θ (PKC-θ), leading to the induction of T-cell anergy. This pathway seems to define at least one crucial state of T-cell anergy. Another RING-type E3 ligase, roquin, has been implicated in regulating the stability and/or translation of mRNA encoding inducible T-cell co-stimulator (ICOS) and interleukin-21 (IL-21). APC, antigen-presenting cell; Ca2+, calcium ions.

CBL, including the ζ-chain of CD3 (CD3ζ), CRKL and SRC-family kinases 4. Of relevance to thymocyte development, thymocytes from Cbl –/– mice show increased signalling through the TCR 28,29 . Cbl–/– thymocytes also show increased adhesion to intercellular adhesion molecule 1 (ICAM1), probably through increased association between CRKL and C3G (a guanine-nucleotide-exchange factor) and the subsequent activation of RAP1 (a member of the RAS family of small GTPases)30. However, whether CBL functions as an E3 ligase for BIM, or for other molecules that are involved in thymic selection, remains to be investigated. T-cell anergy. For T cells, anergy is a form of peripheral tolerance that is characterized by a lack of proliferation and interleukin-2 (IL-2) production by T cells, even after they have been appropriately stimulated31. It has been known for some time that, during anergy induction, TCR ligation in the absence of co-stimulation results in a block in mobilization of calcium ions (Ca2+), as well as defective activation of RAS and, downstream, impaired phosphorylation of extracellular-signalregulated kinase (ERK)32–34. Recent studies, however, have provided molecular insights that improve our

| VOLUME 5

biochemical understanding of T-cell anergy induction. Treatment with IONOMYCIN or TCR ligation in the absence of co-stimulation leads to upregulation of expression of several E3 ligases, including CBL-B, itchy (ITCH) and gene related to anergy in lymphocytes (GRAIL)35–37. These E3 ligases downmodulate the function of phospholipase C-γ1 (PLC-γ1) and protein kinase C-θ (PKC-θ), signalling molecules that are crucial for T-cell activation; therefore, this blocks activation of downstream signalling pathways that are essential for IL-2 production and T-cell proliferation36 (FIG. 2b). Although all three of these E3 ligases are implicated in T-cell anergy induction in vitro, only CBL-B has been examined in vivo. That CBL-B has a crucial role in T-cell anergy induction was further established using several in vivo mouse models, including adoptive transfer of pre-anergized CD4+ T cells, induction of anergy in viral-peptidespecific CD8+ T cells and induction of anergy using superantigen37. In each case, the anergized state of wildtype T cells that is induced by injection of antigen does not occur for CBL-B-deficient T cells. Furthermore, in a mouse model of autoimmune arthritis (in which disease is induced by immunization with type II collagen and COMPLETE FREUND’S ADJUVANT), it was shown



REVIEWS

WW DOMAIN

that Cblb–/– mice develop disease even in the absence of administration of complete Freund’s adjuvant, indicating that loss of CBL-B uncouples autoaggressive T-cell responses from innate immune signals. More strikingly, in a viral-peptide-induced tolerance model, Cblb–/– mice were found to die quickly following a second injection of antigen, even though wild-type mice survived. So, CBL-B-mediated immune tolerance determines life and death, as a result of a lack of T-cell tolerance, following repeated challenge with antigen. Analysis of naive Cblb–/– T cells has shown that the signalling molecules in these cells seem to be modified differently from those found in naive wild-type T cells. For example, it has previously been shown that, in freshly isolated Cblb–/– T cells, tyrosine phosphorylation of VAV1 and activation of AKT occur to a greater extent than in naive wild-type T cells38–40. Furthermore, the p85 subunit of phosphatidylinositol 3-kinase was identified as a substrate for CBL-B-mediated ubiquitylation in naive T cells40. By contrast, during anergy induction of wild-type T cells, CBL-B activity seems to lead mainly to decreased phosphorylation of PLC-γ1, resulting in decreased Ca2+ mobilization37. More importantly, such a biochemical defect in antigen-receptorinduced Ca2+ signalling was restored, to a large degree, by ablation of CBL-B37. These results indicate that there are both qualitative and quantitative differences in the signal-transduction pathways in naive and tolerized T cells and that these are, in part, because CBL-B has distinct targets in these two groups of cells. It should be noted that the authors of two recent publications about E3 ligases and T-cell anergy came to distinct conclusions about how CBL-B might regulate PLC-γ1, and Ca2+ flux, in anergic T cells: one concluded that CBL-B-mediated ubiquitylation induced the degradation of PLC-γ1 REF. 36, and the other concluded that CBL-B-mediated ubiquitylation reduced phosphorylation and therefore the activity of PLC-γ1 REF. 37. This discrepancy might reflect the different time points that were examined in the two studies: at earlier time points of restimulation, anergic T cells might only show modified PLC-γ1 phosphorylation37; by contrast, at later time points, effects on PLC-γ1 degradation might become evident36. Because early signal transduction is crucial for determining whether T-cell recognition of antigen induces proliferation and IL-2 production, early attenuation of PLC-γ1 phosphorylation could be an essential step for T-cell anergy induction. Whether CBL-B-regulated PLC-γ1 degradation has a role in T-cell anergy induction needs to be established using Cblb mutants that lack E3-ligase activity. Moreover, PLC-γ1 is unlikely to be the only target molecule for the E3-ligase activity of CBL-B in anergic T cells. To gain a clear answer, more-comprehensive proteomics approaches need to be used to identify specific substrates of particular ligases.

A protein–protein interaction module that contains two conserved tryptophan (W) residues ∼20–22 amino acids apart. This domain interacts with proline-rich motifs.

Roquin as a repressor of autoimmunity. The importance of E3 ligases in immune tolerance is further underscored by the recent identification of roquin. Roquin is an E3 ligase, the mutation of which in

SYSTEMICLUPUS ERYTHEMATOSUSLIKE AUTOIMMUNE DISEASE

(SLE-like autoimmune disease). A disease that is characterized by autoantibody production and nephritis and is caused by immune responses to selfproteins. N-ETHYLN-NITROSOUREA INDUCED MUTANT MOUSE

(ENU-induced mutant mouse). ENU is a highly potent mutagen that usually induces single base changes in DNA and is used for efficient large-scale insertional mutagenesis in mice (also known as forward genetic screening). FOLLICULAR B HELPER T CELL

A type of T helper cell that provides help to B cells and is located in the B-cell areas of secondary lymphoid tissues. These cells express the chemokine receptor CXC-chemokine receptor 5 (CXCR5). AGOUTI LOCUS

The agouti locus on mouse chromosome 2 determines the coat colour of a mouse by regulating the synthesis of yellow pigment by hair melanocytes. Mutations in this locus are also linked to the development of obesity and neoplasms.


mice is linked to a severe SYSTEMICLUPUSERYTHEMATOSUS 41 SLELIKE AUTOIMMUNE DISEASE . Systematic screening of NETHYLNNITROSOUREAINDUCED MUTANT MICE led to the identification of a mutant mouse line that had an autoimmune phenotype similar to SLE: that is, high levels of autoantibody production, and nephritis with IgG deposition. The mutation causes intrinsic T-cell dysregulation, including increased numbers of FOLLICULAR B HELPER T CELLS with augmented inducible T-cell co-stimulator (ICOS) expression and IL-21 production. The mutation was mapped to a novel gene and was found to be a T to G point mutation that results in a methionine to arginine amino-acid substitution. The encoded protein, which is known as roquin, contains an amino (N)-terminal RING-finger domain followed by a unique ROQ domain and an RNA-binding zincfinger domain, as well as several other domains. It seems that roquin belongs to a large family of proteins that are highly conserved from invertebrates to humans and have as-yet-unknown functions. Interestingly, the N-terminal RING-finger domain is similar to the RING-finger domain of CBL and of TRAFs (tumournecrosis factor (TNF) receptor (TNFR)-associated factors), indicating that roquin functions as a RING-type E3 ligase. Roquin is localized in the cytoplasm (with a distinct dot-like appearance) together with T-cellinduced antigen 1, a translational silencing factor that is a component of cytoplasmic stress granules, which are implicated in the regulation of mRNA translation and stability 42. This colocalization implies that roquin might function as a regulator of mRNA stability and/or translation. It should be noted that the point mutation that was identified in the mutant mice occurs in the ROQ domain, leading the authors to speculate that this mutation might alter the structure of roquin and therefore affect its biological function. However, whether this particular mutation affects the E3-ligase activity or the function of the RNA-binding zinc-finger domain awaits further investigation. Furthermore, the role of roquin in immune tolerance would be explained by identification of the substrate for its E3-ligase activity, which is unknown at present. ITCH in T-cell differentiation

Genetic studies of the AGOUTI LOCUS and mouse coatcolour alterations uncovered an unusual mutation that causes immunological defects, as well as skin- and ear-scratching (itchy) phenotypes43. The itchy locus, which is disrupted in these mice, encodes a HECTtype E3 ligase known as ITCH, which consists of a C-terminal C2 domain (PKC-related domain 2), four 44 WW DOMAINS and a C-terminal HECT-ligase domain . Although the mutant mice have enlarged spleens and lymph nodes, in vitro studies showed that mutant T cells proliferate and produce IL-2 relatively normally after TCR ligation45. However, further detailed analysis indicated that ITCH might be involved in the production of T helper 2 (TH2) cytokines such as IL-4 and IL-5; this finding is supported by the observation of increased serum concentrations of the TH2-celldependent immunoglobulins IgG1 and IgE in the



REVIEWS mutant mice. The WW domains in ITCH bind PPXY motifs (where X denotes any amino acid) in the JUN PROTEINS JUN and JUNB (but not JUND), and ITCH has been shown to promote degradation of these transcription factors45. Indeed, the amount of JUNB protein is increased in Itch–/– T cells. This could be of importance, because T cells from transgenic mice that overexpress JUNB produce more TH2 cytokines than T cells from wild-type animals46, whereas ablation of JUNB diminishes IL-4 and IL-5 secretion and impairs allergen-induced airway inflammation47. Interestingly, T cells that express a kinase-domain deletion mutant of MEKK1 (mitogen-activated protein kinase (MAPK)/ERK kinase kinase 1) showed similar upregulation of TH2-cytokine production48. Similar to Itch–/– T cells, this mutation in MEKK1 resulted in increased concentrations of JUN and

APC Peptide–MHC complex (high affinity for TCR)

CD80 and/or CD86 CD28

TCR

T cell MEKK1

MAPKK7

JNK

ITCH

P

ITCH

P

JUNB peptide fragments

Nucleus

JUN PROTEINS

JUN proteins comprise a family of cellular DNA-binding proteins. They include JUN, JUNB and JUND. These proteins interact with FOS to form the transcription factor activator protein 1 (AP1), which is essential for the transcription of various cytokine genes.

946 | DECEMBER 2005

JUNB

E2 Ub

JUNB Ub Proteasome

IL-4

Figure 3 | The JNK-signalling pathway in differentiation of CD4+ T cells into TH2 cells. Co-ligation of the T-cell receptor (TCR) and CD28 activates the signalling cascade that involves MEKK1 (mitogen-activated protein kinase (MAPK)/ extracellular-signal-regulated kinase (ERK) kinase kinase 1), MAPK kinase 7 (MAPKK7) and JNK (JUN amino-terminal kinase), and this results in the phosphorylation and activation of the E3 ligase itchy (ITCH). Activated ITCH then targets JUNB, a transcription factor that is required for induction of the gene encoding interleukin-4 (IL-4), for ubiquitylation and degradation. So, this pathway inhibits the expression of IL-4 and the differentiation of CD4+ T cells into T helper 2 (TH2) cells.

| VOLUME 5

JUNB, but not JUND, in T cells. In addition, the concentrations of other transcription factors that are important for the differentiation of CD4+ T cells into TH2 cells, such as nuclear factor of activated T cells 1 (NFAT1; also known as NFATc2), GATA-binding protein 3 (GATA3) and MAF, were reported to be normal. Subsequent biochemical studies showed that MEKK1 functions upstream of ITCH activation, with JUN N-terminal kinase (JNK) functioning downstream of MEKK1 to effect phosphorylation of ITCH and, most importantly, upregulation of its E3-ligase activity 48. This defines a signalling cascade that is initiated by TCR ligation and depends on MEKK1, and that results in JNK activation, ITCH phosphorylation and activation, and JUNB degradation. This cascade has a crucial role in modulating the differentiation of CD4+ T cells into TH2 cells (FIG. 3). In fact, ITCH is the first HECT-type E3 ligase (but probably not the only one) that has been found to have its catalytic activity directly regulated by its phosphorylation. Consistent with this model, studies using mice that are deficient in JNK or MAPK kinase 7 (MAPKK7), a kinase that relays MEKK1 activation to JNK, showed similar phenotypes: that is, T cells that lack either JNK or MAPKK7 produce more TH2 cytokines than wildtype T cells49. Therefore, at least in T cells, contrary to the commonly held idea that the MEKK1–JNKmediated phosphorylation of JUN-family proteins is the main pathway that controls downstream gene transcription, the control of JUN-protein turnover through phosphorylation of an E3 ligase (that is, ITCH) seems more pertinent to the biological functions of MEKK1 and JNK. Another conclusion from this study is that E3ligase-dependent ubiquitylation of a particular substrate is subject to complex regulation. Although the phosphorylation-dependent recognition of particular substrates by E3 ligases has been well studied 50 , MEKK1–JNK-induced phosphorylation of ITCH (and the subsequent activation of ITCH) is a novel example of regulation of protein turnover at the level of an E3 ligase, and phosphorylation of the substrates of ITCH (JUN and JUNB) seems to be irrelevant for their degradation48. It has been argued that the strength of the signal that is received by a CD4+ T cell during antigen recognition regulates its development into a particular type of TH cell, with a weak signal leading to differentiation into TH2 cells and a strong signal favouring development into TH1 cells51. However, the underlying molecular mechanisms have remained largely unexplored. The proposed function of the MEKK1–JNK–ITCH cascade in differentiation into TH2 cells might shed light on this long-standing paradox. It can be speculated that, when T cells receive a weak signal, the MEKK1–JNK pathway, which requires a strong co-stimulatory signal52, is not initiated, so T cells express large amounts of JUNB, which directs TH2-cytokine production. By contrast, when T cells are strongly activated, the MEKK1–JNKsignalling pathway is switched on; this results in the phosphorylation and activation of ITCH, which leads to



REVIEWS the turnover and degradation of JUN proteins and therefore to the dampening of differentiation into TH2 cells. However, this attractive hypothesis needs to be tested: for example, by using mice transgenic for TCRs with varying affinities for peptide–MHC complexes. Given the crucial role for JUN proteins in diverse biological functions, it is not surprising that ITCH is not the only E3 ligase that is involved in the ubiquitylation of JUN proteins. A human homologue of the Arabidopsis thaliana protein de-etiolated 1 (DET1) has been shown to associate with CUL4A and RBX1

to form a multisubunit E3 ligase that can promote ubiquitylation and degradation of JUN in a cell line53. In addition, in neurons, the stability of JUN is regulated by an SCF complex that contains the F-box protein FBW7 REF. 54. In this case, JUN needs to be phosphorylated to be recognized by FBW7, before it can be ubiquitylated and degraded. Whether these multisubunit E3-ligase components also have a role in immune regulation and T-cell tolerance remains to be investigated. Ubiquitylation and nuclear factor-κB signalling

PDK1

TNFR

IL-1R

CARMA1

IRAK

TRADD RIP1

TLR

TRIF

CD28

MyD88

TCR

RIP1

BCL-10 MALT1

co

m

pl

ex

RING-finger-domainonly protein

SC

F

S CU KP1 L1

TRAF6 PKC-θ

UBC13 Ub IKK-γ Ub Ub Ub Ub

IKK-α IKK-β

E2

Ub

P P

F-box protein

IκB p50 p65 NF-κB

K63-linked polyubiquitin chain activates the IKK complex

K48-linked polyubiquitin chain targets IκB to the proteasome P P

Ub Ub Ub Ub

IκB peptide fragments Proteasome

NF-κB target-gene products

Nuclear membrane

NF-κB-binding motif

Figure 4 | Ubiquitin conjugation regulates NF-κB activation. Triggering of various cell-surface receptors leads to assembly of a multimolecular complex that includes the RING (really interesting new gene)-type E3 ligase TRAF6 (tumour-necrosis-factor receptor (TNFR)associated factor 6). TRAF6 recruits the ubiquitin-loaded E2 ubiquitin-conjugating enzyme 13 (UBC13) and promotes K63 (Lys63)-linked polyubiquitylation of IKK-γ (inhibitor of NF-κB (IκB) kinase-γ) and activation of the IKK complex. Phosphorylation of IκB by the IKK complex recruits an SCF complex (S-phase kinase-associated protein 1 (SKP1)–cullin-1 (CUL1)–F-boxprotein complex), which is an E3 ligase that induces K48-linked polyubiquitylation of IκB. This results in the proteasome-dependent degradation of IκB and the release of NF-κB (nuclear factor-κB), which is required for the transactivation of genes by NF-κB. BCL-10, B-cell lymphoma 10; CARMA1, caspase-recruitment domain (CARD)–membrane-associated guanylate kinase (MAGUK) protein 1; IL-1R, interleukin-1 receptor; IRAK, IL-1R-associated kinase; MALT1, mucosa-associated-lymphoid-tissue lymphoma-translocation gene 1; MyD88, myeloid differentiation primary-response protein 88; PDK1, 3-phosphoinositidedependent protein kinase 1; PKC-θ, protein kinase C-θ; RIP1, receptor-interacting protein 1; TCR, T-cell receptor; TLR, Toll-like receptor; TRADD, TNFR-associated via death domain; TRIF, Toll/IL-1R (TIR)-domain-containing adaptor protein inducing interferon-β.


Differential ubiquitin-chain formation. Activation of the transcription factor nuclear factor-κB (NF-κB) is a centrepiece of innate immunity, because it is initiated in response to ligation of TLRs by pathogen-associated molecular patterns (PAMPs)55. NF-κB activation is also important for adaptive immunity and is one of the first immune responses that was found to depend on protein ubiquitylation50. The ligation of TLRs by PAMPs, as well as numerous other stimuli, causes phosphorylation of the NF-κB-bound molecule IκB (inhibitor of NF-κB), which is recognized by the F-box protein β-transducin repeat-containing protein (β-TRCP), a component of an SCF complex50 BOX 1. IκB is polyubiquitylated with a K48-linked polyubiquitin chain and is therefore targeted for proteasome-dependent degradation. This results in the liberation of IκB-bound NF-κB dimers, leading to their translocation to the nucleus and the transactivation of NF-κB target genes (FIG. 4). Phosphorylation of IκB is mediated by the IκB kinase (IKK) complex, which consists of three subunits: IKK-α, IKK-β and IKK-γ (also known as NEMO)56. Recent studies have shown that ubiquitylation is also crucical for activation of the IKK complex. Studies of TNFR and IL-1-receptor signalling led to the discovery that TRAFs, a family of proteins that is characterized by the presence of a RING-finger domain and coiledcoil domains, are important mediators of signal transduction through TNFRs57. One TRAF-family member, TRAF6, recruits the ubiquitin–UBC13 (ubiquitinconjugating enzyme 13)–UEV1 UBIQUITINCONJUGATING ENZYME E2 VARIANT 1 complex through its RING-finger domain and promotes ubiquitin conjugation to itself or to IKK-γ 6,58. However, instead of inducing conventional K48-linked polyubiquitylation, TRAF6 induces the formation of a K63-linked polyubiquitin chain. Importantly, K63-linked polyubiquitylation of IKK-γ does not lead to proteasome-dependent degradation but somehow results in activation of the IKK complex and the subsequent phosphorylation of IκB58. Both genetic and biochemical studies have indicated that a complex consisting of CARMA1 (caspaserecruitment domain (CARD)–membrane-associated guanylate kinase (MAGUK) protein 1), BCL-10 and MALT1 (mucosa-associated-lymphoid-tissue lymphoma-translocation gene 1) has an essential role in antigen-receptor signalling that leads to NF-κB activation 59. This complex interacts with TRAF6, which directly associates with BCL-10–MALT1 REF. 58 . The BCL-10–MALT1 complex induces



REVIEWS oligomerization of TRAF6 and activation of its E3-ligase activity. Strikingly, BCL-10–MALT1 alone can target IKK-γ for K63-linked polyubiquitylation, even in the absence of TRAF6 REF. 60. It has therefore been suggested that the formation of the BCL-10–MALT1 complex leads to a conformational change in the complex, which results in recruitment of ubiquitin-bound UBC13 as an E2. Because BCL-10 and MALT1 also form oligomeric complexes with IKK-γ, the binding of ubiquitin–UBC13 to a BCL-10–MALT1 complex allows the transfer of ubiquitin to IKK-γ for K63linked polyubiquitylation. Therefore, even without a physical RING-finger domain, the BCL-10–MALT1 complex can function as an E3 ligase. Consistent with these biochemical studies, loss of CARMA1, BCL-10 or MALT1 results in defective NF-κB activation in both T cells and B cells61. TCR ligation has also been shown to induce the recruitment of PKC-θ to plasma-membrane lipid rafts and the activation of PKC-θ, and this has been implicated in NF-κB activation (as has the BCL-10–MALT1 complex)62. However, the exact mechanisms by which PKC-θ ‘communicates’ with the IKK complex and induces NF-κB activation have remained unclear. But a recent study has indicated that 3-phosphoinositidedependent protein kinase 1 (PDK1) phosphorylates PKC-θ and thereby allows PKC-θ to recruit the IKK complex to the plasma membrane63. PDK1 also functions as an adaptor that physically associates with CARMA1, which then recruits the BCL-10–MALT1 complex to lipid rafts, forming a super-complex that induces K63-linked polyubiquitylation of IKK-γ (mediated by the BCL-10–MALT1 and, probably, TRAF6 components). UBIQUITINCONJUGATING ENZYME E2 VARIANTS

(UEVs). An E2 normally functions as a single polypeptide. It has a conserved active-site cysteine residue that forms a thioester bond with a ubiquitin molecule bound to an E1 and a domain for interacting with RING (really interesting new gene)-finger domains or HECT (homologous to the E6-associated protein carboxy terminus) domains. However, a UEV resembles an E2 but lacks the canonical cysteine residue. UEVs form a complex with a specific E2, ubiquitinconjugating enzyme 13 (UBC13), to facilitate K63 (Lys63)-linked polyubiquitinchain assembly. OTU DOMAIN

(ovarian-tumour domain). A domain that is found in a large family of proteins characterized by the presence of a putative catalytic triad of cysteine proteases. Several of these proteins are known to function as de-ubiquitylating enzymes.

948 | DECEMBER 2005

De-ubiquitylation. The pleiotropic cytokine TNF elicits a broad range of immune and inflammatory responses by inducing NF-κB activation. Unlike antigen-receptor signalling, which depends on tyrosine phosphorylation, the binding of TNF to TNFR induces the formation of a signalling complex that includes TRADD (TNFR-associated via death domain), RIP1 (receptor-interacting protein 1), TRAF2 and A20 REF. 64. Formation of this complex helps to recruit the IKK complex to the vicinity of the activated receptor, and it is required for activation of the IKK complex. Similar to TRAF6, TRAF2 functions as an E3 ligase and promotes K63-linked polyubiquitylation of molecules such as RIP1 REF. 65. It is conceivable that K63-linked RIP1 might facilitate downstream NF-κB signalling. A20 was originally discovered as a TNF-inducible gene product, and it functions as an inhibitor of TNF-induced signalling 66 . Importantly, ablation of A20 results in augmented responsiveness to lipopolysaccharide and TNF by thymocytes and fibroblasts67. A20 –/– mice develop multi-organ inflammation and die at a young age. Loss of A20 seems to result in increased kinase activity of IKK and in prolonged activation of NF-κB. Structurally, A20 is characterized by the presence of an OTU OVARIAN TUMOUR DOMAIN linked to

| VOLUME 5

multiple zinc-finger domains in the C-terminal portion of the protein. Database searches showed that the OTU domain is characteristic of a superfamily of otubains, a group of putative cysteine proteases68,69. This family of proteins contains conserved ubiquitin-interacting and ubiquitin-associated motifs that are commonly found in de-ubiquitylating enzymes, indicating that these proteins function as de-ubiquitylating enzymes. Indeed, A20 was later found to be a de-ubiquitylating enzyme for K63-linked polyubiquitylated RIP1, thereby leading to termination of TNF-induced NF-κB signalling70. This is not the end of the story. A20 also functions as an E3 ligase and recruits one of the E2 proteins UBCH5A or UBCH7 using its C-terminal zinc-finger domains (which also mediate the transfer of ubiquitin from E2 to the substrate), to promote the ubiquitylation of RIP1 REF. 70. In this reaction, the polyubiquitin chain is K48-linked and targets RIP1 for proteasomal degradation. Importantly, A20-induced ubiquitylation of RIP1 occurs after its de-ubiquitylation. Therefore, A20 is a dual-function enzyme: first, it operates as a de-ubiquitylating enzyme, removing the K63-linked polyubiquitin chain from RIP1; and second, it operates as an E3 ligase, promoting K48-linked polyubiquitylation of RIP1 and thereby targeting it for degradation (FIG. 5). By functioning as a ‘double-edged sword’, A20 prevents the over-reaction of innate immune responses. It remains unclear, however, whether A20 removes and adds ubiquitin molecules on the same lysine residues of RIP1 or on different residues. Further analysis, using A20 –/– macrophages, indicates that A20 also has an important role in the regulation of TLR-signalling pathways, which result in a wide range of responses, including NF-κB activation and pro-inflammatory cytokine production71. Similar to TNFR signalling, TLRs recruit RIPs and TRAFs to the receptor complex72. It seems that A20 also removes K63-linked polyubiquitin chains from TRAF6, thereby terminating NF-κB activation that is induced by TLR ligation71. Because RIP1 is also an essential mediator of TLR-induced NF-κB activation73, A20 might also modulate RIP1 through de-ubiquitylation and ubiquitylation during innate immune responses. A20 is not the only de-ubiquitylating enzyme that is implicated in NF-κB signalling. The cylindromatosis protein (CYLD) — a tumour-suppressor protein, the mutation of which is linked to benign tumours in humans — specifically removes K63-linked polyubiquitin chains from TRAFs and does not affect K48-linked polyubiquitin chains on IκB74–76. Unlike the OTUdomain-containing protein A20, CYLD contains a classical ubiquitin C-terminal hydrolase domain, which confers cysteine-specific protease activity for the cleavage of polyubiquitin chains. It is not yet clear whether CYLD is involved in immune regulation. Ubiquitylation and other cytokines

Ubiquitylation in transforming-growth-factor- β mediated signalling. Transforming growth factor-β (TGF-β) is a multifunctional cytokine that is implicated in the regulation of various aspects of immune



REVIEWS

A

B TLR TNFR

c De-ubiquitylation Ub Ub Ub Ub a K63-linked b polyubiquitin Ub UBC13 chain

RIP1 A20

TRAF2

RIP1

d UBCH5 Ub

Macrophage TRIF

K48-linked polyubiquitin chain RIP1 Ub Ub Ub Ub

TRADD

RIP1 TRAF6

K63-linked polyubiquitin chain Ub Ub Ub Ub

UBC13

De-ubiquitylation

Ub A20 A20

NF-κB signalling A20 induction

NF-κB signalling

RIP1 proteasomal degradation

NF-κB signalling

Figure 5 | A20 as a dual enzyme in NF-κB signalling. A | Stimulation through tumour-necrosis factor (TNF) receptors (TNFRs) causes TNFR-associated factor 2 (TRAF2)-mediated K63 (Lys63)-linked polyubiquitylation of receptor-interacting protein 1 (RIP1) (a), which results in the downstream activation of nuclear factor-κB (NF-κB). A20 then functions first as a de-ubiquitylating enzyme to remove K63-linked polyubiquitin chains (b), which terminates TNF-mediated signalling (c). At this stage, A20 then functions as an E3 ligase through recruitment of the ubiquitin-bound E2 ubiquitin-conjugating enzyme H5 (UBCH5), facilitating the transfer of ubiquitin from UBCH5 to RIP1 and promoting K48-linked polyubiquitylation of RIP1 (d) and subsequent degradation of RIP1. B | A20 also functions as a de-ubiquitylating enzyme to remove K63-linked polyubiquitin chains from TRAF6, a mediator of Toll-like receptor (TLR) signalling. TRADD, TNFR-associated via death domain; TRIF, Toll/interleukin-1 receptor (TIR)-domain-containing adaptor protein inducing interferon-β.

responses. Ablation of TGF-β or expression of a dominant-negative TGF-β receptor results in aberrant lymphocyte proliferation and multi-organ inflammation77,78. TGF-β also has an important role in the inhibition of CD4+ T-cell differentiation into TH1 and TH2 cells, by reducing expression of T-bet and GATA3 or by inhibiting phosphorylation of the TEC-family kinase ITK (IL-2-inducible T-cell kinase)79–81. Recently, TGF-β has also been shown to convert naive CD4+CD25– T cells in the periphery into TReg cells (which are CD4+CD25+) through induction of expression of a crucial transcription factor, forkhead box P3 (FOXP3)82. Collectively, these results indicate that TGF-β is an important cytokine in immune homeostasis and tolerance induction. TGF-β-mediated signalling is initiated by binding of the cytokine to its receptors, which are heterodimeric transmembrane serine/threonine kinases, and this is followed by phosphorylation of a family of SMADs (mothers against decapentaplegic homologues), translocation to the nucleus by these proteins and subsequent regulation of gene transcription83. SMADs can be divided into three groups: receptor-regulated SMADs (R-SMADs; such as SMAD2 or SMAD3), commonmediator SMADs (Co-SMADs; such as SMAD4) and inhibitory SMADs (I-SMADs; such as SMAD7). Many studies have documented that the SMAD-signalling pathway, which is highly conserved, is subjected to regulation by ubiquitylation. At least two factors that regulate the ubiquitylation of SMADs, SMAD ubiquitylation regulatory factor 1 (SMURF1) and SMURF2, have been characterized as HECT-type E3 ligases84,85. Structurally, the WW domains of SMURFs bind PPXY motifs in some SMADs84. However, SMURFs can either inhibit responses to TGF-β by inducing the degradation of


SMADs86 or increase signalling in response to TGF-β by inducing the degradation of SKI-related novel protein N (SnoN), a transcriptional repressor of SMADs that works by forming an inhibitory SMURF–SMAD–SnoN complex87. In addition, SMURF2 can bind the I-SMAD SMAD7, which forms a complex with the TGF-β receptor, thereby targeting the TGF-β receptor for ubiquitylation and degradation88. Given the similarity of the immunological phenotypes of Itch–/– mice and TGF-β-deficient mice, and given that both ITCH and SMURFs are HECT-type E3 ligases, the role of ITCH in TGF-β-mediated signalling was assessed89. The TGF-β-induced arrest in growth of mouse embryonic fibroblasts was not found to occur in ITCH-deficient mice. Surprisingly, ITCH was not found to affect the protein stability of either SMADs or SnoN. Instead, ITCH bound SMAD2 and promoted its ubiquitylation. It seems that ubiquitylation of SMAD2 increases its TGF-βinduced phosphorylation: ITCH facilitates the formation of a complex between the TGF-β receptor and SMAD2 through ubiquitylation of SMAD2, leading to an increase in TGF-β-induced gene transcription. So, ITCH functions as a positive regulator of TGF-β-mediated signalling through proteolysisindependent ubiquitylation. Whether this function of ITCH has a role in immune regulation, such as in T-cell differentiation or TReg-cell generation, awaits further investigation. Suppressor of cytokine signalling proteins as E3 ligases. Suppressor of cytokine signalling (SOCS) proteins comprise a well-characterized family of proteins that inhibit the Janus kinase (JAK)–signal transducer and activator of transcription (STAT)-signalling pathway.



REVIEWS

IFN or IL-6 IFN receptor or IL-6 receptor

JAK

JAK

SOCS1 Elongin B– elongin C CUL5

STAT

Ub E2 P

Ub

Ub

Ub

Ub

ECS complex

RBX2 P

P

P

complex to a particular substrate and facilitates the transfer of ubiquitin from E2 to the substrate. Consistent with predictions from its structure, SOCS1 has been shown to promote the conjugation of ubiquitin to JAK2, VAV1 and the p65 subunit of NF-κB (also known as REL-A)96–98. Interestingly, it has been shown that, in genetically modified mice in which the SOCS-box domain of SOCS1 has been deleted, there are increased responses to IFN-γ, and such mice develop chronic multi-organ inflammation99. So, SOCS proteins are a crucial component of these multi-protein E3 ligases that bind a specific target and help to promote its ubiquitylation. ISGylation and de-ISGylation

Target-gene transcription

Target-gene transcription

Figure 6 | SOCS1 as a component of the E3 ligase the ECS complex. Triggering of interferon (IFN) or interleukin-6 (IL-6) receptors results in the recruitment and activation of Janus kinases (JAKs) and, subsequently, in the phosphorylation of downstream signal transducer and activator of transcription (STAT) proteins. Activated JAKs also associate with suppressor of cytokine signalling 1 (SOCS1) through the SRC homology 2 (SH2) domain of SOCS1. The SOCS-box domain of SOCS1 forms a complex with the elongin-B–elongin-C complex and cullin-5 (CUL5), to form an ECS complex (elongin-B–elongin-C–CUL5–SOCS-protein complex), which is an E3 ligase that promotes polyubiquitylation and degradation of JAKs and inhibits STAT-mediated gene transcription. RBX2, RING (really interesting new gene)-box-2.

SOCSBOX DOMAIN

(suppressor-of-cytokinesignalling-box domain). A protein–protein interaction module that allows binding of the elongin-B–elongin-C complex to form an ECS complex (elongin-B– elongin-C–cullin-5–SOCSprotein complex), which is an E3 ligase. POLYINOSINIC POLYCYTIDYLIC ACID

(poly I:C). A synthetic polymer that resembles the RNA of infectious viruses and is used to stimulate the production of interferon by the immune system through binding Toll-like receptor.

950 | DECEMBER 2005

The SOCS family has eight closely related members — SOCS1, SOCS2, SOCS3, SOCS4, SOCS5, SOCS6, SOCS7 and cytokine-inducible SRC homology 2 (SH2)-containing protein (CIS) — and each contains an SH2 domain and a C-terminal SOCSBOX DOMAIN90. Expression of SOCS proteins is induced after stimulation with cytokine. The SH2 domains of SOCS proteins then associate with phosphorylated tyrosine residues in activated JAKs, and this attenuates the kinase activity of these JAKs. The importance of this family of proteins became further evident following studies of gene-targeted mice. Socs1–/– mice show neonatal lethality, severe lymphopaenia, upregulated lymphocyte apoptosis and multi-organ lymphocyte infiltration91–93. The hyper-responsiveness of Socs1–/– lymphocytes was attributed to heightened IFN-γ-mediated signalling, which results in augmented phosphorylation of STAT1. In early biochemical studies, SOCS proteins were identified as binding partners for the complex that forms between elongin B and elongin C. This multi-protein complex is now known as an ECS complex (elongin-B–elongin-C–CUL5–SOCSprotein complex), and it is an E3 ligase94,95. In these complexes, CUL5 functions as a scaffold that binds elongin-B–elongin-C, a SOCS protein (which directly binds the substrate) and the RING-finger-domainonly protein RBX2 (which recruits ubiquitin–E2 complexes) (FIG. 6). ECS complexes are analogous to SCF complexes, and together they form a subgroup of the multisubunit RING-type E3 ligases BOX 1. In the case of SCF complexes, the F-box protein recruits a specific protein substrate; by contrast, in the case of ECS complexes, the SOCS protein targets the ECS

| VOLUME 5

ISG15 was originally identified as an IFN-stimulated gene. This gene encodes a 15 kDa protein that has two ubiquitin-like domains100,101. In addition to stimulation with type I IFNs, the expression of ISG15 is induced by microbial challenge, by genotoxic stress, during pregnancy and during retinoid-induced cellular differentiation. ISG15 is activated by an E1-like ubiquitin-activating enzyme (UBE1L) and is then transferred to UBCH8, which functions as an E2 for ISGylation of substrates102,103. So far, several signalling molecules, including PLC-γ1, JAK1 and ERK, have been identified as substrates for ISG15 conjugation104. Because UBCH8 is also recruited to E3 ligases, it remains unclear whether the E3 ligases that function in ubiquitinconjugation pathways and ISG15-conjugation pathways are similar or distinct. A recent exciting development is the identification of a de-ISGylating enzyme, ubiquitin-binding protein 43 (UBP43; also known as USP18). Biochemically, UBP43 functions as a protease that specifically removes ISG15 from the proteins to which it is conjugated105. However, ablation of Ubp43 in mice leads to hypersensitivity to POLYINOSINICPOLYCYTIDYLIC ACID (poly I:C), with reduced survival rates and decreased numbers of peripheral-blood cells and bone-marrow cells106. The JAK–STAT-signalling pathway is highly upregulated in Ubp43 –/– cells following stimulation with type I IFNs, and this is accompanied by augmented cell death. In mouse models of viral infection, it was recently found that Ubp43–/– mice are highly resistant to lethal inoculation with lymphocytic choriomeningitis virus or vesicular stomatitis virus107. Viral replication in Ubp43–/– cells is abrogated, presumably as a result of the hyper-responsiveness to type I IFNs. This study shows the importance of a balance between ISGylation and de-ISGylation in the regulation of innate immune responses to viral infection. However, a more recent study showed that deficiency in ISG15 does not affect IFN-induced activation of STATs and immune responses to viral infection108. It should also be noted that UBP43 is itself regulated by ubiquitindependent degradation mediated by the multisubunit E3 ligase SCFSKP2, in which SKP2 is the F-box protein that recruits the substrate109, indicating that there is another layer of control, which involves ISGylation, de-ISGylation and ubiquitylation.



REVIEWS

Concluding remarks

Various E3 ligases and de-conjugating enzymes have been identified as crucial regulators of immune responses. Mouse models of various diseases — such as autoimmune diabetes, arthritis, allergic asthma and viral infections — will provide important tools to study the biological significance of conjugating or de-conjugating molecules. In addition, comprehensive approaches such as proteomics and genomics are needed to identify the entire set of substrates that is recognized by a 1.

2. 3. 4. 5. 6.

7.

8. 9.

10.

11.

12. 13.

14.

15.

16.

17.

18.

19.

20. 21.

22.

23.

Hershko, A., Ciechanover, A., Heller, H., Haas, A. L. & Rose, I. A. Proposed role of ATP in protein breakdown: conjugation of protein with multiple chains of the polypeptide of ATP-dependent proteolysis. Proc. Natl Acad. Sci. USA 77, 1783–1786 (1980). Pickart, C. M. Mechanisms underlying ubiquitination. Annu. Rev. Biochem. 70, 503–533 (2001). Weissman, A. M. Themes and variations on ubiquitylation. Nature Rev. Mol. Cell Biol. 2, 169–178 (2001). Liu, Y. C. Ubiquitin ligases and the immune response. Annu. Rev. Immunol. 22, 81–127 (2004). Hicke, L. Protein regulation by monoubiquitin. Nature Rev. Mol. Cell Biol. 2, 195–201 (2001). Deng, L. et al. Activation of the IκB kinase complex by TRAF6 requires a dimeric ubiquitin-conjugating enzyme complex and a unique polyubiquitin chain. Cell 103, 351–361 (2000). Goldstein, G. et al. Isolation of a polypeptide that has lymphocyte-differentiating properties and is probably represented universally in living cells. Proc. Natl Acad. Sci. USA 72, 11–15 (1975). Mathis, D. & Benoist, C. Back to central tolerance. Immunity 20, 509–516 (2004). Walker, L. S. & Abbas, A. K. The enemy within: keeping self-reactive T cells at bay in the periphery. Nature Rev. Immunol. 2, 11–19 (2002). Derbinski, J., Schulte, A., Kyewski, B. & Klein, L. Promiscuous gene expression in medullary thymic epithelial cells mirrors the peripheral self. Nature Immunol. 2, 1032–1039 (2001). Aaltonen, J. et al. An autoimmune disease, APECED, caused by mutations in a novel gene featuring two PHDtype zinc-finger domains. Nature Genet. 17, 399–403 (1997). Nagamine, K. et al. Positional cloning of the APECED gene. Nature Genet. 17, 393–398 (1997). Heino, M. et al. Autoimmune regulator is expressed in the cells regulating immune tolerance in thymus medulla. Biochem. Biophys. Res. Commun. 257, 821–825 (1999). Anderson, M. S. et al. Projection of an immunological self shadow within the thymus by the Aire protein. Science 298, 1395–1401 (2002). Liston, A., Lesage, S., Wilson, J., Peltonen, L. & Goodnow, C. C. Aire regulates negative selection of organspecific T cells. Nature Immunol. 4, 350–354 (2003). References 14 and 15 provide genetic evidence that AIRE is crucial for central-tolerance induction. Johnnidis, J. B. et al. Chromosomal clustering of genes controlled by the aire transcription factor. Proc. Natl Acad. Sci. USA 102, 7233–7238 (2005). Pitkanen, J. et al. The autoimmune regulator protein has transcriptional transactivating properties and interacts with the common coactivator CREB-binding protein. J. Biol. Chem. 275, 16802–16809 (2000). Kumar, P. G. et al. The autoimmune regulator (AIRE) is a DNA-binding protein. J. Biol. Chem. 276, 41357–41364 (2001). Coscoy, L. & Ganem, D. PHD domains and E3 ubiquitin ligases: viruses make the connection. Trends Cell Biol. 13, 7–12 (2003). Uchida, D. et al. AIRE functions as an E3 ubiquitin ligase. J. Exp. Med. 199, 167–172 (2004). Bottomley, M. J. et al. NMR structure of the first PHD finger of autoimmune regulator protein (AIRE1). Insights into autoimmune polyendocrinopathy-candidiasisectodermal dystrophy (APECED) disease. J. Biol. Chem. 280, 11505–11512 (2005). Winoto, A. Genes involved in T-cell receptor-mediated apoptosis of thymocytes and T-cell hybridomas. Semin. Immunol. 9, 51–58 (1997). Mariathasan, S., Jones, R. G. & Ohashi, P. S. Signals involved in thymocyte positive and negative selection. Semin. Immunol. 11, 263–272 (1999).

given ubiquitylating or de-ubiquitylating enzyme so that we can gain deeper insights into the mechanisms and specific targets by which ubiquitin-mediated or ubiquitin-like-molecule-mediated conjugation pathways participate in immune regulation. It is also anticipated that a better and more detailed understanding of the ubiquitin-conjugation system in immune regulation will ultimately help to develop novel therapeutic interventions for human diseases such as autoimmune diseases, cancer, allergic inflammation and infectious diseases.

24. Calnan, B. J., Szychowski, S., Chan, F. K., Cado, D. & Winoto, A. A role for the orphan steroid receptor Nur77 in apoptosis accompanying antigen-induced negative selection. Immunity 3, 273–282 (1995). 25. Bouillet, P. et al. BH3-only Bcl-2 family member Bim is required for apoptosis of autoreactive thymocytes. Nature 415, 922–926 (2002). 26. Akiyama, T. et al. Regulation of osteoclast apoptosis by ubiquitylation of proapoptotic BH3-only Bcl-2 family member Bim. EMBO J. 22, 6653–6664 (2003). 27. Thien, C. B. & Langdon, W. Y. Cbl: many adaptations to regulate protein tyrosine kinases. Nature Rev. Mol. Cell Biol. 2, 294–307 (2001). 28. Murphy, M. A. et al. Tissue hyperplasia and enhanced T-cell signalling via ZAP-70 in c-Cbl-deficient mice. Mol. Cell. Biol. 18, 4872–4882 (1998). 29. Naramura, M., Kole, H. K., Hu, R.-J. & Gu, H. Altered thymic positive selection and intracellular signals in Cbl-deficient mice. Proc. Natl Acad. Sci. USA 95, 15547–15552 (1998). 30. Shao, Y., Elly, C. & Liu, Y.-C. Negative regulation of Rap1 activation by the Cbl E3 ubiquitin ligase. EMBO Rep. 4, 425–431 (2003). 31. Schwartz, R. H. T cell anergy. Annu. Rev. Immunol. 21, 305–334 (2003). 32. Fields, P. E., Gajewski, T. F. & Fitch, F. W. Blocked Ras activation in anergic CD4+ T cells. Science 271, 1276–1278 (1996). 33. Gajewski, T. F., Qian, D., Fields, P. & Fitch, F. W. Anergic T-lymphocyte clones have altered inositol phosphate, calcium, and tyrosine kinase signaling pathways. Proc. Natl Acad. Sci. USA 91, 38–42 (1994). 34. Li, W., Whaley, C. D., Mondino, A. & Mueller, D. L. Blocked signal transduction to the ERK and JNK protein kinases in anergic CD4+ T cells. Science 271, 1272–1276 (1996). 35. Anandasabapathy, N. et al. GRAIL: an E3 ubiquitin ligase that inhibits cytokine gene transcription is expressed in anergic CD4+ T cells. Immunity 18, 535–547 (2003). 36. Heissmeyer, V. et al. Calcineurin imposes T cell unresponsiveness through targeted proteolysis of signaling proteins. Nature Immunol. 5, 255–265 (2004). 37. Jeon, M. S. et al. Essential role of the E3 ubiquitin ligase Cbl-b in T cell anergy induction. Immunity 21, 167–177 (2004). References 35–37 provide evidence that E3 ligases are involved in T-cell anergy induction. 38. Bachmaier, K. et al. Negative regulation of lymphocyte activation and autoimmunity by the molecular adaptor Cbl-b. Nature 403, 211–216 (2000). 39. Chiang, Y. J. et al. Cbl-b-regulates the CD28 dependence of T-cell activation. Nature 403, 216–220 (2000). 40. Fang, D. & Liu, Y.-C. Proteolysis-independent regulation of phosphatidylinositol 3-kinase by Cbl-b-mediated ubiquitination in T cells. Nature Immunol. 2, 870–875 (2001). 41. Vinuesa, C. G. et al. A RING-type ubiquitin ligase family member required to repress follicular helper T cells and autoimmunity. Nature 435, 452–458 (2005). This paper identifies a mutation in a RING-fingerdomain-containing protein, roquin, that results in autoimmunity. 42. Laroia, G., Cuesta, R., Brewer, G. & Schneider, R. J. Control of mRNA decay by heat shock–ubiquitin– proteasome pathway. Science 284, 499–502 (1999). 43. Hustad, C. M. et al. Molecular genetic characterization of six recessive viable alleles of the mouse agouti locus. Genetics 140, 255–265 (1995). 44. Perry, W. L. et al. The itchy locus encodes a novel ubiquitin protein ligase that is disrupted in a18H mice. Nature Genet. 18, 143–146 (1998). 45. Fang, D. et al. Dysregulation of T lymphocyte function in itchy mice: a role for Itch in TH2 differentiation. Nature Immunol. 3, 281–287 (2002).


46. Li, B., Tournier, C., Davis, R. J. & Flavell, R. A. Regulation of IL-4 expression by the transcription factor JunB during T helper cell differentiation. EMBO J. 18, 420–432 (1999). 47. Hartenstein, B. et al. TH2 cell-specific cytokine expression and allergen-induced airway inflammation depend on JunB. EMBO J. 21, 6321–6329 (2002). 48. Gao, M. et al. Jun turnover is controlled through JNK-dependent phosphorylation of the E3 ligase Itch. Science 306, 271–275 (2004). This paper provides genetic and biochemical evidence that JNK signalling modulates the turnover of JUN proteins by phosphorylation and activation of the E3 ligase ITCH. 49. Dong, C. et al. JNK is required for effector T-cell function but not for T-cell activation. Nature 405, 91–94 (2000). 50. Karin, M. & Ben-Neriah, Y. Phosphorylation meets ubiquitination: the control of NF-κB activity. Annu. Rev. Immunol. 18, 621–663 (2000). 51. Constant, S. L. & Bottomly, K. Induction of TH1 and TH2 CD4+ T cell responses: the alternative approaches. Annu. Rev. Immunol. 15, 297–322 (1997). 52. Su, B. et al. JNK is involved in signal integration during costimulation of T lymphocytes. Cell 77, 727–736 (1994). 53. Wertz, I. E. et al. Human de-etiolated-1 regulates c-Jun by assembling a CUL4A ubiquitin ligase. Science 303, 1371–1374 (2004). 54. Nateri, A. S., Riera-Sans, L., Da Costa, C. & Behrens, A. The ubiquitin ligase SCFFbw7 antagonizes apoptotic JNK signaling. Science 303, 1374–1378 (2004). 55. Silverman, N. & Maniatis, T. NF-κB signaling pathways in mammalian and insect innate immunity. Genes Dev. 15, 2321–2342 (2001). 56. Ghosh, S. & Karin, M. Missing pieces in the NF-κB puzzle. Cell 109, S81–S96 (2002). 57. Hsu, H., Shu, H. B., Pan, M. G. & Goeddel, D. V. TRADD–TRAF2 and TRADD–FADD interactions define two distinct TNF receptor 1 signal transduction pathways. Cell 84, 299–308 (1996). 58. Sun, L., Deng, L., Ea, C. K., Xia, Z. P. & Chen, Z. J. The TRAF6 ubiquitin ligase and TAK1 kinase mediate IKK activation by BCL10 and MALT1 in T lymphocytes. Mol. Cell 14, 289–301 (2004). Reference 58, together with reference 60, shows that K63-linked polyubiquitylation of the IKK complex is crucial for NF-κB activation. 59. Lin, X. & Wang, D. The roles of CARMA1, Bcl10, and MALT1 in antigen receptor signaling. Semin. Immunol. 16, 429–435 (2004). 60. Zhou, H. et al. Bcl10 activates the NF-κB pathway through ubiquitination of NEMO. Nature 427, 167–171 (2004). 61. Ruland, J. et al. Bcl10 is a positive regulator of antigen receptor-induced activation of NF-κB and neural tube closure. Cell 104, 33–42 (2001). 62. Sedwick, C. E. & Altman, A. Perspectives on PKCθ in T cell activation. Mol. Immunol. 41, 675–686 (2004). 63. Lee, K. Y., D’Acquisto, F., Hayden, M. S., Shim, J. H. & Ghosh, S. PDK1 nucleates T cell receptor-induced signaling complex for NF-κB activation. Science 308, 114–118 (2005). This paper shows that PDK1 has a dual role in NF-κB activation through phosphorylation of PKC-θ and through recruitment of the CARMA1–BCL-10–MALT1 complex. 64. Karin, M. & Lin, A. NF-κB at the crossroads of life and death. Nature Immunol. 3, 221–227 (2002). 65. Lee, T. H., Shank, J., Cusson, N. & Kelliher, M. A. The kinase activity of Rip1 is not required for tumor necrosis factor-α-induced IκB kinase or p38 MAP kinase activation or for the ubiquitination of Rip1 by Traf2. J. Biol. Chem. 279, 33185–33191 (2004).



REVIEWS

66. Opipari, A. W. Jr, Boguski, M. S. & Dixit, V. M. The A20 cDNA induced by tumor necrosis factor α encodes a novel type of zinc finger protein. J. Biol. Chem. 265, 14705–14708 (1990). 67. Lee, E. G. et al. Failure to regulate TNF-induced NF-κB and cell death responses in A20-deficient mice. Science 289, 2350–2354 (2000). 68. Balakirev, M. Y., Tcherniuk, S. O., Jaquinod, M. & Chroboczek, J. Otubains: a new family of cysteine proteases in the ubiquitin pathway. EMBO Rep. 4, 517–522 (2003). 69. Makarova, K. S., Aravind, L. & Koonin, E. V. A novel superfamily of predicted cysteine proteases from eukaryotes, viruses and Chlamydia pneumoniae. Trends Biochem. Sci. 25, 50–52 (2000). 70. Wertz, I. E. et al. De-ubiquitination and ubiquitin ligase domains of A20 downregulate NF-κB signalling. Nature 430, 694–699 (2004). 71. Boone, D. L. et al. The ubiquitin-modifying enzyme A20 is required for termination of Toll-like receptor responses. Nature Immunol. 5, 1052–1060 (2004). References 70 and 71 uncover the importance of A20 in NF-κB activation. 72. Barton, G. M. & Medzhitov, R. Toll signaling: RIPping off the TNF pathway. Nature Immunol. 5, 472–474 (2004). 73. Meylan, E. et al. RIP1 is an essential mediator of Toll-like receptor 3-induced NF-κB activation. Nature Immunol. 5, 503–507 (2004). 74. Brummelkamp, T. R., Nijman, S. M., Dirac, A. M. & Bernards, R. Loss of the cylindromatosis tumour suppressor inhibits apoptosis by activating NF-κB. Nature 424, 797–801 (2003). 75. Kovalenko, A. et al. The tumour suppressor CYLD negatively regulates NF-κB signalling by deubiquitination. Nature 424, 801–805 (2003). 76. Trompouki, E. et al. CYLD is a deubiquitinating enzyme that negatively regulates NF-κB activation by TNFR family members. Nature 424, 793–796 (2003). 77. Shull, M. M. et al. Targeted disruption of the mouse transforming growth factor-β1 gene results in multifocal inflammatory disease. Nature 359, 693–699 (1992). 78. Gorelik, L. & Flavell, R. A. Abrogation of TGFβ signaling in T cells leads to spontaneous T cell differentiation and autoimmune disease. Immunity 12, 171–181 (2000). 79. Chen, C. H. et al. Transforming growth factor β blocks Tec kinase phosphorylation, Ca2+ influx, and NFATc translocation causing inhibition of T cell differentiation. J. Exp. Med. 197, 1689–1699 (2003). 80. Gorelik, L., Constant, S. & Flavell, R. A. Mechanism of transforming growth factor β-induced inhibition of T helper type 1 differentiation. J. Exp. Med. 195, 1499–1505 (2002). 81. Gorelik, L., Fields, P. E. & Flavell, R. A. TGF-β inhibits TH type 2 development through inhibition of GATA-3 expression. J. Immunol. 165, 4773–4777 (2000). 82. Chen, W. et al. Conversion of peripheral CD4+CD25– naive T cells to CD4+CD25+ regulatory T cells by TGF-β induction of transcription factor Foxp3. J. Exp. Med. 198, 1875–1886 (2003). 83. Shi, Y. & Massague, J. Mechanisms of TGF-β signaling from cell membrane to the nucleus. Cell 113, 685–700 (2003).

952 | DECEMBER 2005

84. Zhu, H., Kavsak, P., Abdollah, S., Wrana, J. L. & Thomsen, G. H. A SMAD ubiquitin ligase targets the BMP pathway and affects embryonic pattern formation. Nature 400, 687–693 (1999). 85. Lin, X., Liang, M. & Feng, X. H. Smurf2 is a ubiquitin E3 ligase mediating proteasome-dependent degradation of Smad2 in transforming growth factor-β signaling. J. Biol. Chem. 275, 36818–36822 (2000). 86. Zhang, Y., Chang, C., Gehling, D. J., Hemmati-Brivanlou, A. & Derynck, R. Regulation of Smad degradation and activity by Smurf2, an E3 ubiquitin ligase. Proc. Natl Acad. Sci. USA 98, 974–979 (2001). 87. Bonni, S. et al. TGF-β induces assembly of a Smad2– Smurf2 ubiquitin ligase complex that targets SnoN for degradation. Nature Cell Biol. 3, 587–595 (2001). 88. Kavsak, P. et al. Smad7 binds to Smurf2 to form an E3 ubiquitin ligase that targets the TGF β receptor for degradation. Mol. Cell 6, 1365–1375 (2000). 89. Bai, Y., Yang, C., Hu, K., Elly, C. & Liu, Y.-C. Itch E3 ligase-mediated regulation of TGF-β signaling by modulating Smad2 phosphorylation. Mol. Cell 15, 825–831 (2004). This paper shows an unexpected role for ITCH in TGF-β-mediated signalling, through ubiquitylation and the subsequent increase in the association of SMAD2 with the TGF-β β receptor. 90. Alexander, W. S. Suppressors of cytokine signalling (SOCS) in the immune system. Nature Rev. Immunol. 2, 410–416 (2002). 91. Marine, J. C. et al. SOCS1 deficiency causes a lymphocytedependent perinatal lethality. Cell 98, 609–616 (1999). 92. Naka, T. et al. Accelerated apoptosis of lymphocytes by augmented induction of Bax in SSI-1 (STAT-induced STAT inhibitor-1) deficient mice. Proc. Natl Acad. Sci. USA 95, 15577–15582 (1998). 93. Starr, R. et al. Liver degeneration and lymphoid deficiencies in mice lacking suppressor of cytokine signaling-1. Proc. Natl Acad. Sci. USA 95, 14395–14399 (1998). 94. Kamura, T. et al. The elongin BC complex interacts with the conserved SOCS-box motif present in members of the SOCS, ras, WD-40 repeat, and ankyrin repeat families. Genes Dev. 12, 3872–3881 (1998). 95. Zhang, J. G. et al. The conserved SOCS box motif in suppressors of cytokine signaling binds to elongins B and C and may couple bound proteins to proteasomal degradation. Proc. Natl Acad. Sci. USA 96, 2071–2076 (1999). 96. De Sepulveda, P., Ilangumaran, S. & Rottapel, R. Suppressor of cytokine signaling-1 inhibits VAV function through protein degradation. J. Biol. Chem. 275, 14005–14008 (2000). 97. Kamizono, S. et al. The SOCS box of SOCS-1 accelerates ubiquitin-dependent proteolysis of TEL–JAK2. J. Biol. Chem. 276, 12530–12538 (2001). 98. Ryo, A. et al. Regulation of NF-κB signaling by Pin1dependent prolyl isomerization and ubiquitin-mediated proteolysis of p65/RelA. Mol. Cell 12, 1413–1426 (2003). 99. Zhang, J. G. et al. The SOCS box of suppressor of cytokine signaling-1 is important for inhibition of cytokine action in vivo. Proc. Natl Acad. Sci. USA 98, 13261–13265 (2001). 100. Haas, A. L., Ahrens, P., Bright, P. M. & Ankel, H. Interferon induces a 15-kilodalton protein exhibiting marked homology to ubiquitin. J. Biol. Chem. 262, 11315–11323 (1987).

| VOLUME 5

101. Blomstrom, D. C., Fahey, D., Kutny, R., Korant, B. D. & Knight, E. Jr. Molecular characterization of the interferoninduced 15-kDa protein. Molecular cloning and nucleotide and amino acid sequence. J. Biol. Chem. 261, 8811–8816 (1986). 102. Zhao, C. et al. The UbcH8 ubiquitin E2 enzyme is also the E2 enzyme for ISG15, an IFN-α/β-induced ubiquitin-like protein. Proc. Natl Acad. Sci. USA 101, 7578–7582 (2004). 103. Yuan, W. & Krug, R. M. Influenza B virus NS1 protein inhibits conjugation of the interferon (IFN)-induced ubiquitin-like ISG15 protein. EMBO J. 20, 362–371 (2001). 104. Malakhov, M. P. et al. High-throughput immunoblotting. Ubiquitin-like protein ISG15 modifies key regulators of signal transduction. J. Biol. Chem. 278, 16608–16613 (2003). 105. Malakhov, M. P., Malakhova, O. A., Kim, K. I., Ritchie, K. J. & Zhang, D. E. UBP43 (USP18) specifically removes ISG15 from conjugated proteins. J. Biol. Chem. 277, 9976–9981 (2002). 106. Malakhova, O. A. et al. Protein ISGylation modulates the JAK–STAT signaling pathway. Genes Dev. 17, 455–460 (2003). 107. Ritchie, K. J. et al. Role of ISG15 protease UBP43 (USP18) in innate immunity to viral infection. Nature Med. 10, 1374–1378 (2004). This paper provides genetic evidence that a deficiency in de-ISGylation affects innate immune responses. 108. Osiak, A., Utermohlen, O., Niendorf, S., Horak, I. & Knobeloch, K. P. ISG15, an interferon-stimulated ubiquitinlike protein, is not essential for STAT1 signaling and responses against vesicular stomatitis and lymphocytic choriomeningitis virus. Mol. Cell. Biol. 25, 6338–6345 (2005). 109. Tokarz, S. et al. The ISG15 isopeptidase UBP43 is regulated by proteolysis via the SCFSkp2 ubiquitin ligase. J. Biol. Chem. 279, 46424–46430 (2004).

Acknowledgements The authors apologize for not citing other important articles, owing to space limitations. This work was supported by research grants from the National Institutes of Health (United States) to Y.-C.L. and to M.K., who is also an American Cancer Society Research Professor.

Competing interests statement The authors declare competing financial interests: see web version for details.

Online links DATABASES The following terms in this article are linked online to: Entrez Gene: http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=gene A20 | AIRE | CBL-B | CUL1 | GRAIL | ISG15 | ITCH | RBX1 | roquin | SKP1 FURTHER INFORMATION La Jolla Institute for Allergy and Immunology: http://www.liai.org Access to this interactive links box is free online.



REVIEWS

MONOCYTE AND MACROPHAGE HETEROGENEITY Siamon Gordon and Philip R. Taylor Abstract | Heterogeneity of the macrophage lineage has long been recognized and, in part, is a result of the specialization of tissue macrophages in particular microenvironments. Circulating monocytes give rise to mature macrophages and are also heterogeneous themselves, although the physiological relevance of this is not completely understood. However, as we discuss here, recent studies have shown that monocyte heterogeneity is conserved in humans and mice, allowing dissection of its functional relevance: the different monocyte subsets seem to reflect developmental stages with distinct physiological roles, such as recruitment to inflammatory lesions or entry to normal tissues. These advances in our understanding have implications for the development of therapeutic strategies that are targeted to modify particular subpopulations of monocytes.

OSTEOCLAST

A multinucleate cell that resorbs bone.

Sir William Dunn School of Pathology, University of Oxford, South Parks Road, Oxford OX1 3RE, UK. Correspondence to S.G. e-mail: [email protected] doi:10.1038/nri1733

Circulating monocytes give rise to a variety of tissueresident macrophages throughout the body, as well as to specialized cells such as dendritic cells (DCs) and OSTEOCLASTS. Monocytes are known to originate in the bone marrow from a common myeloid progenitor that is shared with neutrophils, and they are then released into the peripheral blood, where they circulate for several days before entering tissues and replenishing the tissue macrophage populations1. The morphology of mature monocytes in the peripheral circulation is heterogeneous, and these cells constitute ∼5–10% of peripheral-blood leukocytes in humans. They vary in size and have different degrees of granularity and varied nuclear morphology, which at the extremes of variation can lead to confusion with granulocytes, lymphocytes, natural killer cells and DCs. The basic features of the mononuclear-phagocyte system (which includes macrophages and their monocyte precursors and lineage-committed bone-marrow precursors, as well as all other cells that are derived from this lineage2) are summarized in BOX 1. As long ago as 1939, Ebert and Florey3 reported the observation that monocytes emigrated from blood vessels and developed into macrophages in the tissues. Pro-inflammatory, metabolic and immune stimuli all


elicit increased recruitment of monocytes to peripheral sites4, where differentiation into macrophages and DCs occurs, contributing to host defence, and tissue remodelling and repair. Studies of the mononuclear-phagocyte system, using monoclonal antibodies specific for various cell-surface receptors and differentiation antigens, have shown that there is substantial heterogeneity of phenotype, which most probably reflects the specialization of individual macrophage populations within their microenvironments. Although it is clear that monocytes are precursors of both macrophage and DC lineages, this development and differentiation pathway is still relatively poorly studied in vivo. However, the identification of heterogeneity among peripheral-blood monocytes — first, in humans, and more recently, in mice — has provided a powerful insight into the nature of myeloid-cell heterogeneity and has provided novel ways to assess cell fate and function in vivo. The cellular and molecular adhesion mechanisms that are used by migrating monocytes have recently been reviewed5 and are not discussed here. Instead, we briefly summarize the current understanding of human monocyte heterogeneity and then describe recent advances in the study of mouse monocyte heterogeneity, to highlight the different physiological



REVIEWS

AFFERENT LYMPHATIC VESSEL

A vessel that carries lymph into a lymph node.

roles of the subsets in vivo, and we discuss the relevance of heterogeneity to the study of human monocyte biology. We also discuss the origins of tissue-resident macrophage populations, and we describe how knowledge

of their origins might be pertinent to tissue homeostasis at times when it could be advantageous to modulate monocyte function as part of a therapeutic strategy. Monocyte heterogeneity in humans

Box 1 | The mononuclear-phagocyte system Yolk sac Precursor

Primitive macrophage

M-CFU

Monoblast

Fetal liver

HSC

Pro-monocyte

Monocyte

Macrophage

GM-CFU Pro-monocyte Ly6C+ M-CFU Monoblast ‘inflammatory’ Mononuclear-phagocyte system monocyte

G-CFU

Neutrophil

Adult bone marrow Macrophage

Ly6C+ ‘inflammatory’ monocyte

Dendritic cell

Adult tissues

Osteoclast

Ly6C– ‘resident’ monocyte Adult peripheral blood

Phagocytic cells were initially classified as the reticuloendothelial system87; however, this classification failed to distinguish between ‘true’ sinusoidal endothelial cells and sinus-lining macrophages. As a consequence, the classification of the mononuclearphagocyte system was altered to include only macrophages and their monocyte precursors and lineage-committed bone-marrow precursors2. However, this classification might require further refinement as the origins of fetal macrophages become clearer and as unanswered questions about the common precursors of macrophages and lymphocytes are addressed. During development, the origins of cells from the yolk sac that have macrophagelike phenotypes might be distinct from the origins of these cells in adults and after haematopoiesis properly begins in the fetal liver88,89 (see figure). Developing macrophages are first found in the yolk sac, as identified by morphological characteristics, as well as by expression of macrophage markers such as FMS, CD11b and the mannose receptor90–92. Studies in zebrafish, in which cell lineage can more easily be followed because of the transparency of the embryos, have shown population of the yolk sac with macrophage precursors; these cells then differentiate and emigrate into the head mesenchyme and its circulation93. Later in development, haematopoiesis in the fetal liver becomes a source of macrophages that resemble those that are present in adults. Initially, haematopoiesis in the liver generates large numbers of macrophages, which are present in most organs. Population of the organs of the embryo with phagocytes is discussed elsewhere88,89. Furthermore, many recent studies (for further details, see the main text) indicate that, although monocytes can be precursors for the replenishment of tissue-resident macrophage populations, many of these macrophages might be derived from local proliferation in the adult rather than from recruited peripheral monocytes. When bone-marrow monocytes are released into the peripheral blood (having a Ly6C+ phenotype), they are thought to differentiate into a phenotypically distinct (Ly6C–) cell subset (for further details, see the main text). G-CFU, granulocyte colony-forming unit; GM-CFU, granulocyte/macrophage colony-forming unit; HSC, haematopoietic stem cell; M-CFU, macrophage colony-forming unit.

954 | DECEMBER 2005

| VOLUME 5

Peripheral-blood monocytes show morphological heterogeneity, such as variability of size, granularity and nuclear morphology. Monocytes were initially identified by their expression of large amounts of CD14 (which is part of the receptor for lipopolysaccharide). However, the subsequent identification of differential expression of antigenic markers showed that monocytes in human peripheral blood are heterogeneous, and this provided the first clues to the differential physiological activities of monocyte subsets. Differential expression of CD14 and CD16 (also known as FcγRIII) allowed monocytes to be divided into two subsets: CD14hiCD16– cells, which are often called classic monocytes, because this phenotype resembles the original description of monocytes; and CD14+CD16+ cells6. It was subsequently shown that the CD14+CD16+ monocytes expressed higher amounts of MHC class II molecules and CD32 (also known as FcγRII), and it was suggested that these cells resemble mature tissue macrophages7. Distinct chemokine-receptor expression profiles were also among the phenotypic differences that were recognized between these subsets: for example, CD14+CD16+ monocytes expressed CC-chemokine receptor 5 (CCR5), whereas CD14hiCD16– monocytes expressed CCR2 REF. 8. A summary of the cell-surfacemarker phenotype of these human monocyte subsets is presented in TABLE 1. When cultured, both human monocyte subsets can differentiate into DCs in the presence of granulocyte/ macrophage colony-stimulating factor (GM-CSF) and interleukin-4 (IL-4)9,10. Furthermore, using an in vitro transendothelial-migration model, in which freshly isolated peripheral-blood mononuclear cells are incubated with human umbilical-vein endothelial-cell monolayers grown on an endotoxin-free collagenous matrix, it has been shown that monocytes can migrate across an endothelial barrier in vitro and differentiate either into macrophages, which remain in the subendothelial matrix, or into DCs, which then migrate back across the endothelial layer11. In this model, the CD14+CD16+ monocyte subset was found to be more likely to become DCs and reverse transmigrate than was the CD14hiCD16– monocyte subset12, indicating that the CD14+CD16+ cells might be precursors of DCs, which can pass through tissues and then migrate to the lymph nodes through the AFFERENT LYMPHATIC VESSELS. However, these observations do not preclude an in vivo role for CD14hiCD16– monocytes in contributing to the DC pool. Importantly, the extent to which monocytes differentiate into macrophages or DCs in this model depends on both the phenotype of the cultured monocyte population and the factors that are present in the culture. For example, transendothelial migration of CD14+CD16+ monocytes can be induced with soluble CX3C-chemokine ligand 1 (CX3CL1; also www.nature.com/reviews/immunol


REVIEWS known as fractalkine) or CXC-chemokine ligand 12 (CXCL12; also known as SDF1α), the receptors for which are preferentially expressed by these cells13. An additional monocyte subset that is defined by the expression of CD14, CD16 and CD64 (also known as FcγRI) has been reported more recently14. These cells

seem to combine characteristics of monocytes and DCs, with high expression of CD86 and HLA-DR and high T-cell-stimulatory activity. Compared with CD14hiCD16– (classic) monocytes (which are also CD64+), these CD14+CD16+CD64+ cells have a similarly high phagocytic activity and produce similarly

Table 1 | Phenotype of the two best-characterized monocyte subsets in various mammals* Antigen

Human CD14hi CD16– ‘inflammatory’ monocytes

Human CD14+CD16+ ‘resident’ monocytes

Mouse CCR2+ CX3CR1low ‘inflammatory’ monocytes

Mouse Rat CD43low CCR2– ‘inflammatory’ CX3CR1hi monocytes ‘resident’ monocytes

Rat CD43hi ‘resident’ monocytes

Pig CD163 – ‘inflammatory’ monocytes

Pig CD163+ ‘resident’ monocytes

Chemokine receptors CCR1

+

–

ND

ND

ND

ND

ND

ND

CCR2‡

+

–

+

–

+

–

ND

ND

CCR4

+

–

ND

ND

ND

ND

ND

ND

CCR5

–

+

ND

ND

ND

ND

ND

ND

CCR7

+

–

ND

ND

+

–

ND

ND

CXCR1

+

–

ND

ND

ND

ND

ND

ND

CXCR2

+

–

ND

ND

ND

ND

ND

ND

CXCR4

+

++

ND

ND

ND

ND

ND

ND

CX3CR1‡

+

++

+

++

–

+

ND

ND

Other receptors CD4

+

+

ND

ND

+/–

++

ND

ND

CD11a

ND

ND

+

++

ND

ND

+

++

CD11b

++

++

++

++

++

++

ND

ND

CD11c‡

++

+++

–

+

–

+

+

++

CD14

+++

+

ND

ND

ND

ND

++

+

CD31

+++

+++

++

+

ND

ND

ND

ND

CD32

+++

+

ND

ND

+++

+

ND

ND

CD33

+++

+

ND

ND

ND

ND

ND

ND

CD43

ND

ND

–

+

–

+

ND

ND

CD49b

ND

ND

+

–

ND

ND

ND

ND

CD62L‡

++

–

+

–

+

–

ND

ND

CD80

ND

ND

ND

ND

ND

ND

+

++

CD86

+

++

ND

ND

ND

ND

+

++

CD115

++

++

++

++

ND

ND

ND

ND

§

§

CD116

++

++

++

++

ND

ND

ND

ND

CD200R

ND

ND

ND

ND

+

–

ND

ND

F4/80

ND

ND

+

+

ND

ND

ND

ND

Ly6C

ND

ND

+

–

ND

ND

ND

ND

7/4

ND

ND

+

–

ND

ND

ND

ND

MHC class II

+

++

–

–

ND

ND

+

++

*‘Inflammatory’ and ‘resident’ nomenclature is based on studies carried out in mice and extrapolated to other species. Data have been assigned arbitrary symbols that represent no expression (–), marginal expression (+/–) and increasing amounts of expression (+, ++, +++). Mouse monocyte-subset phenotypic data are derived from REFS 16,22,25,35,94,95. Rat monocyte data were kindly provided by U. Yrlid (personal communication). Other data are taken from the references cited in the main text. ‡ For a more detailed definition of the cell-surface receptor and functional heterogeneity of human monocytes, see REF. 15. These receptors show good conservation of expression-pattern differences between the subsets of at least three species. §There is no available commercial reagent that recognizes the extracellular domain of mouse CD116 (also known as granulocyte/macrophage colony-stimulating-factor receptor α-chain). 7/4, an unidentified mouse antigen recognized by monoclonal antibody 7/4; CCR, CC-chemokine receptor; CD200R, CD200 receptor; CXCR, CXC-chemokine receptor; CX3CR1, CX3C-chemokine receptor 1; EMR1, epidermalgrowth-factor-module-containing mucin-like hormone receptor 1; F4/80, monoclonal antibody that recognizes the mouse homologue of the human protein EMR1; ND, either differences between the subsets have not been determined or, in some cases, there is no clear species conservation of the antigens.




REVIEWS large amounts of cytokines (such as tumour-necrosis factor (TNF) and IL-6), and these phenotypes are not shared with the CD14 + CD16 + CD64 – subset. However, the CD14+CD16+CD64+ cells share with the CD14+CD16+CD64– subset a greater stimulatory activity in MIXED LEUKOCYTE REACTIONS than CD14hiCD16– monocytes. A summary of the phenotypic properties of human monocyte subsets, such as the reduced phagocytic activity and increased stimulatory capacity in mixed leukocyte reactions has been compiled by Grage-Griebenow et al.15 The authors speculated that, although the origins of this CD14+CD16+CD64+ subset are not known, it could be an immunoregulatory monocyte phenotype or, possibly (because of the similar characteristics of these cells to both monocytes and DCs), an intermediate phenotype between monocytes and DCs14. So, although these observations from the past 20 years have provided insights into the fate and function of human monocyte subsets, the restriction of the study of monocyte heterogeneity to in vitro analyses of human cells and the initial failure to recognize the mouse counterparts of these monocyte subsets hampered determination of the functional roles of the monocyte subsets in a physiological context. Monocyte heterogeneity in mice

MIXED LEUKOCYTE REACTION

A tissue-culture technique for testing T-cell reactivity. The proliferation of one population of T cells, induced by exposure to inactivated MHCmismatched stimulator cells, is determined by measuring the incorporation of 3H-thymidine into the DNA of dividing cells. CLODRONATELOADED LIPOSOME

A liposome that contains the drug dichloromethylene diphosphonate. These liposomes are ingested by macrophages, resulting in cell death. DILLABELLED LIPOSOSOME

A liposome that is labelled with the fluorochrome Dil (1,1′-dioctadecyl-3,3,3′,3′-tetramethylindocarbocyanine perchlorate). These liposomes are internalized by phagocytic cells, rendering the cells fluorescent.

956 | DECEMBER 2005

Antigenic differentiation of two monocyte subsets in mice was first achieved after the observation that monocytes (identified in mice by their F4/80+CD11b+ phenotype) could be subdivided according to their expression of CCR2, CD62L (also known as L-selectin) and CX3C-chemokine receptor 1 (CX3CR1; measured by expression of green fluorescent protein (GFP) in cells from mice in which GFP had been ‘knocked-in’ to one of the Cx3cr1 alleles)16. One monocyte subset expressed CCR2, CD62L and only moderate amounts of CX3CR1, whereas the second did not express CCR2 or CD62L but expressed higher amounts of CX3CR1. The CCR2+ monocyte subset was, as expected, found to migrate towards the CCR2 ligand CC-chemokine ligand 2 (CCL2; also known as MCP1)16. The expression of CCR2 and the capacity to migrate towards CCL2 is consistent with the important role of this chemokine and its receptor in the recruitment of monocytes to inflammatory lesions, so the subset of monocytes that expresses CCR2 in mice is known as the ‘inflammatory’ subset17–21. Dan Littman’s research group extended our knowledge of the phenotypic differences between the two monocyte subsets that have been described for mice and humans, and they paid particular attention to the expression pattern of chemokine receptors TABLE 1. As can be seen from TABLE 1, the receptors that are differentially expressed by the inflammatory-monocyte subset are broadly considered to be chemokine and adhesion receptors that are involved in the recruitment of leukocytes to an inflammatory lesion. In addition, Geissmann et al.22 identified Ly6C (which is part of the epitope of GR1) as an additional marker of CCR2+ monocytes

| VOLUME 5

in mice TABLE 1. These studies indicated that CCR2+ CD62L+CX3CR1lowLy6C+ mouse monocytes correspond to CD14hiCD16– (classic) human monocytes, which are also CCR2 + CX 3 CR1 low and that CCR2 – CD62L–CX3CR1hiLy6C– mouse monocytes correspond to CD14+CD16+CD64 – human monocytes, which also express large amounts of CX3CR1. These observations were the first to indicate that it would be possible to address the in vivo relevance of human monocyte heterogeneity by studying mice. Functional characterization of mouse monocyte subsets. To distinguish functionally between the two mouse monocyte subsets, Geissmann and colleagues22 adoptively transferred GFP-expressing monocytes from the Cx3cr1-knock-in mice and studied cell fate and function in naive and immunologically challenged recipient mice. They found that the CCR2+CX3CR1low monocytes were short-lived after adoptive transfer and were difficult to detect in the tissues of naive recipients. However, as anticipated by their expression of CCR2 and CD62L (molecules that are known to be involved in inflammatory-cell recruitment17–21,23,24), these cells were rapidly recruited to sites of experimentally induced inflammation22 (the reason for the term inflammatory monocyte). After migration into the inflamed site, the CCR2+ monocytes upregulated expression of CD11c and MHC class II molecules, and some CD11c+MHC class II+ cells were recovered from the draining lymph nodes, indicating that they might have differentiated into DCs. Experimental evidence for the ability of CCR2+ monocytes to differentiate into DCs was also obtained by transferring CCR2+ monocytes into MHC-class-I-deficient mice and showing that the transferred cells could prime naive CD8+ T cells22. By contrast, the CCR2 – CX 3 CR1 hi monocyte population was found to persist for longer in animals after adoptive transfer, and it was possible to recover GFP-marked cells from the blood, spleen, lungs, liver and brain of recipients for several days after transfer. Some of the donor cells that were recovered from the spleen had acquired a DC-like phenotype (that is, CD11c+MHC class II+) following transfer. Similar to their human counterparts9,10, both mouse monocyte subsets have been reported to differentiate into DCs in vitro when cultured with GM-CSF and IL-4 REF. 22. However, the observation that the CCR2– monocytes can enter tissues and acquire a DC-like phenotype under steady-state conditions is consistent with the finding that their human counterparts (CD14+CD16+ monocytes) preferentially differentiate into DCs in the in vitro transendothelial-migration model12. It was not clear at which point the division in lineage between the two monocyte subsets occurred. Sunderkotter and colleagues25 attempted to address this question with a series of experiments based on in vivo depletion of all monocytes with CLODRONATELOADED LIPO SOMES and in vivo labelling with DILLABELLED LIPOSOSOMES, followed by study of the repopulation kinetics of both monocyte subsets. After depletion, the first monocytes



REVIEWS

Bone marrow

Peripheral blood

Inflamed tissue Pathogen

Monocyte

Macrophage

CCL2

Ly6C

a CCR2 Ly6C+

d

Draining lymph node

e

Pathogen clearance and wound healing

Ly6C+

DC CD62L

g

b

f d CCL1 CCR7 DC

Normal tissue CCR8 Ly6Cmid

c

Tissueresident DC

CX3CL1 CX3CR1 ?

h ? Ly6C–

Tissue-resident macrophages: splenic macrophages, Kupffer cells, alveolar macrophages, microglia and osteoclasts

Figure 1 | Development and function of monocyte subsets in mice. Ly6C+ bone-marrow monocytes are released into the peripheral blood (a) and are thought to adopt a Ly6Cmid phenotype (b), which is associated with selective expression of CC-chemokine receptor 7 (CCR7) and CCR8 and retention of CCR2, before (under steady-state conditions) they form CCR2–Ly6C– monocytes (c) that are characterized by high CX3C-chemokine receptor 1 (CX3CR1) expression25,28. Both Ly6C+ and Ly6Cmid monocytes respond to pro-inflammatory cues, such as the CCR2 ligand CC-chemokine ligand 2 (CCL2), and are recruited to inflammatory lesions22,28 (d). Most ‘inflammatory’ monocytes are thought to differentiate into macrophages, which are important for clearance of pathogens and for the resolution of inflammation (e). Some monocytes emigrate from the tissues to the draining lymph nodes, a process that uses CCR7 and CCR8 receptor–ligand interactions28 (f). CCR7+CCR8+ monocytes that are present in the tissue must be recruited directly from the peripheral blood to the inflammatory site or must differentiate from Ly6C+ monocytes in situ (or must arise from a combination of both of these mechanisms). The expression of CCR7 and CCR8 by these cells makes them uniquely disposed to emigrate into the lymphatic vessels. In the draining lymph nodes, these monocytes acquire dendritic cell (DC)-like characteristics (g), which they do not obtain if they are retained in the tissue by selective chemokine-receptor deficiency28. In the absence of inflammation, CX3CR1hiLy6C– monocytes enter the tissues and replenish the tissue-resident macrophage and DC populations22 (h). Solid arrows represent pathways that are supported by established data, whereas dashed arrows represent pathways that are indicated from a compilation of more recent data and speculation. CX3CL1, CX3C-chemokine ligand 1.

to reappear in the circulation were the CCR2+ (inflammatory) subset. The phenotype of these cells resembles that of monocytes found in the bone marrow26, and their appearance at this stage is consistent with the hypothesis that these monocytes precede the CCR2– subset in the developmental pathway. Unfortunately, the adoptive transfer of purified CCR2+ monocytes did not formally confirm this hypothesis, although this was attributed to inappropriate manipulation of the monocyte subset during purification, because the transferred cells could not be recovered from the circulation25. In addition, Sunderkotter and colleagues25 identified a rare population of monocytes that were characterized by intermediate expression of Ly6C. Assuming that the proposed pathway of CCR2–Ly6C– monocyte derivation from CCR2+Ly6C+ monocytes is true, this population might be an intermediate phenotypic state between the two subsets (FIG. 1). It had already been shown that inflammatory monocytes recruited to the skin after injection of fluorescent latex beads could ingest the beads, and although most of these cells remained in the tissue as macrophages, others could migrate to the T-cell area


of the draining lymph node and acquire a DC-like phenotype (that is, CD11c+MHC class II+)27. While investigating the possible role of CCR2 and CX3CR1 in regulating the migration of latex-bead-carrying monocytes from the skin to the draining lymph nodes, proliferation of Ly6Cmid monocytes, associated with a relative reduction in the size of the Ly6C+ monocyte subset, was observed in the peripheral blood of CCR2-deficient mice28. Genetic deficiency in CCR2 or CX3CR1 did not reduce the migration of the latex-bead-carrying DC-like cells to the lymph nodes; however, an unexpected increase in migrating latex-bead-carrying cells was observed in CCR2-deficient animals. This was reflected by an increase in the proportion of latex-bead-carrying Ly6Cmid cells in the skin of CCR2-deficient animals after the injection of latex beads, indicating that the recruitment of these cells to the skin did not depend on CCR2 in this model. The authors speculated that the increase in the proportion of these cells in the peripheral blood and inflamed skin of CCR2-deficient mice and in the number of monocyte-derived DC-like cells in the draining lymph node might be linked and that the



REVIEWS

PERITONEUM

Ly6Cmid monocytes could be a subset of monocytes that is predisposed to differentiate into DCs28. Ly6Cmid monocytes have also been suggested to resemble Ly6C– monocytes rather than Ly6C+ monocytes in terms of their capacity to stimulate allogeneic cells28. Reverse-transcription-PCR analysis of the three subsets indicated that Ly6Cmid monocytes express higher amounts of the mRNAs that encode CCR7 and CCR8 than do the other two monocyte subsets but similar amounts of mRNA that encodes CCR2 to that of Ly6C+ monocytes. Furthermore, analysis of mice that are deficient in CCR8 or the ligands of CCR7 showed a reduction in the number of latex-bead-carrying cells with DC-like phenotype that migrated to the draining lymph node, supporting a role for these molecules in this process. Parallel in vitro studies with human monocytes showed that blockade of CCR8 considerably reduced the reverse transmigration of human monocytes without affecting the initial migration across the endothelial layer28. The results of these studies of the heterogeneity of mouse (and human) monocytes are summarized in FIG. 1. Briefly, by extrapolation from the current data, a picture is emerging in which bonemarrow-derived monocytes (with the phenotype CCR2+CX3CR1lowLy6C+) are released into the circulation and, in the absence of inflammation, alter their functional and phenotypic characteristics, passing through an intermediate phenotype (CCR2+CCR7+ CCR8 + Ly6C mid ). Both the bone-marrow-derived monocytes and the monocytes of intermediate phenotype can respond to pro-inflammatory cues, migrate to inflamed tissues and differentiate into macrophages and DCs. The monocytes with an intermediate phenotype might be particularly predisposed to migrate to the draining lymph nodes and differentiate into DCs. In the absence of inflammation, a switch in monocyte phenotype, the mechanism of regulation of which is unknown, generates monocytes that are postulated to enter the tissues and replenish the tissue-resident macrophage and DC populations; these are known as the ‘resident’ monocyte population (which has the phenotype CCR2–CX3CR1hiLy6C–). The similarities between human and mouse monocyte subsets indicate that this is a conserved system, and further analysis of the mouse system and other mammalian systems will be helpful for understanding human monocyte biology.

The membrane that lines the abdominal cavity.

Monocyte heterogeneity in other mammals

STERILE PERITONITIS

Inflammation of the peritoneum that is induced by sterile injection of an irritant, such as thioglycollate broth. This results in the sequential recruitment of granulocytes, monocytes and lymphocytes. It is widely used to study acute inflammation. EFFERENT LYMPHATIC VESSEL

A vessel that carries lymph out of a lymph node.

958 | DECEMBER 2005

The study of monocyte heterogeneity in mice, in the context of transgenic and gene-knockout technology, is a powerful tool to investigate the mechanisms of monocyte function and fate, but it also has limitations. Constraints in working with mice, such as the limited availability of cells and the limited accessibility of physiological systems to physical manipulation and/or intervention, merit the extension of these studies to larger mammalian models. In this context, heterogeneity of monocytes has been reported in rats and in pigs, with notable parallels evident between these species and the observations that have been made for mice and humans TABLE 1.

| VOLUME 5

Rats. Monocyte heterogeneity was first shown in rats by using CD43 as a differential marker29. In rats, CD43hi monocytes express higher amounts of CD4 than do CD43low monocytes, and the CCR2–CX3CR1hi monocyte subset in mice has also been shown to express CD43 REF. 25. Recently, rat monocytes have been further characterized antigenically and functionally (U. Yrlid, personal communication). CD43low monocytes express higher amounts of CCR2, CCR7, CD32 and CD62L than do CD43hi monocytes, and they migrate to the PERITONEUM during experimental STERILE + + PERITONITIS, confirming their similarity to CCR2 Ly6C hi (inflammatory) mouse monocytes. By contrast, CD43 monocytes express higher amounts of CX3CR1 and CD11c than do CD43low monocytes, so this population is analogous to the Ly6C– (resident) monocyte population in mice. The use of rats has facilitated two important experiments that were technically more demanding in mice. First, when adoptively transferred, purified CCR2+CX3CR1low rat monocytes intravenously acquired the phenotype of CCR2–CX3CR1hi monocytes, strongly supporting the lineage development that was previously proposed to occur in mice 25 . Second, a model of mesenteric lymphadenectomy and thoracic-duct cannulation, in which afferent and EFFERENT LYMPHATIC VESSELS fuse, allowed the collection of migrating DCs. Following adoptive transfer of CCR2– CX 3 CR1 hi monocytes, some donor cells acquired the phenotype of intestinal-lymph DCs without the requirement for additional stimulation, indicating that the transferred resident-monocyte population could differentiate into DCs under steady-state conditions. Importantly, both the conversion of CCR2+CX3CR1low monocytes into CCR2–CX3CR1hi monocytes and the differentiation of CCR2–CX3CR1hi monocytes into cells with the phenotype of intestinal-lymph DCs occurred without cell division. Pigs. Monocyte heterogeneity has also been described in pigs30. CD163– monocytes express higher amounts of CD14 than do their CD163+ counterparts, which in turn express higher levels of MHC class II molecules, several adhesion molecules (such as CD11a and CD11c) and the co-stimulatory molecules CD80 and CD86; CD163+ monocytes also have a higher allostimulatory activity and a greater ability to present soluble antigen to T cells than do CD163– monocytes30,31. Both subsets of monocyte can also differentiate into DCs when cultured in the presence of GM-CSF and IL-4 REF. 32. The greater allostimulatory activity and co-stimulatory molecule expression by the CD163+ monocyte subset indicates that these cells might correspond to Ly6C– (resident) monocytes in mice. Origins of macrophages and DCs

Tissue macrophages have a broad role in the maintenance of tissue homeostasis, through the clearance of senescent cells and the remodelling and repair of tissues after inflammation33,34. They are generally considered to be derived from circulating monocytes and show a



REVIEWS

PATTERNRECOGNITION RECEPTOR

A type of receptor that binds conserved molecular structures that are found in pathogens. Examples include the mannose receptor, which binds terminally mannosylated and polymannosylated compounds, and Toll-like receptors, which are activated by various microbial products, such as bacterial lipopolysaccharides, hypomethylated DNA, flagellin and double-stranded RNA. SCAVENGER RECEPTOR

A cell-surface receptor that is involved in the internalization of selected polyanionic ligands, including modified low-density lipoproteins. TINGIBLEBODY MACROPHAGE

A type of macrophage that is present in the splenic white pulp and is involved in the clearance of apoptotic cells. LAMINA PROPRIA

The connective tissue that underlies the epithelium of the gut mucosa. It contains various myeloid and lymphoid cells, including macrophages, dendritic cells, T cells and B cells. LANGERHANS CELL

A professional antigenpresenting dendritic cell that is localized in the epidermal layer of the skin. BONEMARROW CHIMERA

An individual that has received a transplant of bone marrow from another individual. PARABIOTIC MICE

Mice that share a circulatory system as a result of surgical connection. GRANULOCYTE/MACROPHAGE COLONYFORMINGUNIT PRECURSOR

(GM-CFU precursor). A committed precursor in haematopoietic tissues that can form granulocytes and macrophages in the presence of specific growth factors.

high degree of heterogeneity, which has largely been uncovered through studies with monoclonal antibodies35–37. The heterogeneity reflects the specialization of function that is adopted by macrophages in different anatomical locations, including the following: the ability of osteoclasts to remodel bone38; the high expression of PATTERNRECOGNITION RECEPTORS and SCAVENGER RECEPTORS by alveolar macrophages39–41, which are involved in clearing microorganisms, viruses and environmental particles in the lungs; and the positioning of thymic macrophages42 and TINGIBLEBODY MACROPHAGES43 in the germinal centre for clearance of apoptotic lymphocytes that are generated during the development of an acquired immune response. The gut is one of the richest sources of macrophages in the body, and isolation of macrophages from the LAMINA PROPRIA has highlighted a unique macrophage phenotype that is characterized by high phagocytic and bactericidal activity but weak production of pro-inflammatory cytokines. This phenotype can be induced in peripheral-blood-derived macrophages by intestinal stromal-cell products, indicating that the tissue microenvironment can markedly influence the phenotype of tissue-resident macrophages44. In addition to macrophage heterogeneity in different organs, macrophage heterogeneity can be observed in a single organ, and the mouse spleen, which is discussed briefly in this section, is a particularly good example of this. Since the recent unravelling of monocyte heterogeneity, immediate attention has focused on studies of monocyte fate, mainly regarding the question of the potential of monocytes to form DCs in vitro and in vivo and the possible effect of this on the immune response. By contrast, relatively little attention has been given to the perhaps more challenging question of the role of monocyte heterogeneity in the generation of macrophage populations in mice and the nature of the circulating precursors of macrophage populations. For example, it is still not clear whether tissue macrophages are derived from particular lineage-committed precursors or whether they are derived randomly from the monocyte pool. In addition (and particularly in the case of inflammation-elicited macrophages), after cells have ‘differentiated’ in a microenvironment, are the cells then terminally differentiated, or are they functionally flexible and able to alter their phenotype in response to changes in their location? As previously mentioned, most macrophages in the tissues of an adult are considered to be derived from circulating monocytes, which constitutively replenish tissue-resident macrophage populations. However, studies of the origins of many tissue-resident macrophage populations have shown that local proliferation has a considerable role in the renewal and maintenance of many macrophage types (particularly under steady-state conditions), with the recruitment of circulating precursors having little, if any, role in this process in some cases. But inflammatory insults, such as trauma or infection, can lead to an increased dependence on the recruitment of blood-borne precursors to aid repopulation of the tissue-resident populations in many of these


inflamed tissues. Van Furth et al.45 summarized the results of the initial studies in this area, although these questions have been re-investigated more recently, leading to the need to modify some of the older ideas about the cellular origins of macrophages. A good understanding of the origins of these cells, as well as the timing and context of their recruitment, will be essential not only to understand how tissues recover from exposure to pro-inflammatory stimuli but also to determine how these cells help to restore the status quo ante, which is important for diseases in which a loss of tissue homeostasis might result from dysfunction of the tissue-resident macrophage population(s). Examples in which the origins of tissue-resident macrophage populations have been investigated are discussed in this section and are illustrated in FIG. 2. We have concentrated on the adult because of the substantial differences in the development of phagocytes in the adult and the embryo, which are briefly summarized in BOX 1. One aspect that is clear in many of these examples is that the nature of the circulating precursor that is involved in the renewal of tissue-resident populations remains poorly defined. Studies such as those that were outlined in previous sections for determining monocyte fate during inflammation might be useful for studying the potential involvement of individual monocyte subsets in the renewal of tissue-resident cells. Langerhans cells. LANGERHANS CELLS populate the epidermis during ontogeny and were originally thought to be constantly replenished by bone-marrow-derived cells. However, it has recently been shown that Langerhans cells are self-renewing 46. Using BONEMARROW CHIMERAS made with congenic cells (which were followed for 18 months) and PARABIOTIC MICE (which were followed for 6 months), it was shown that Langerhans cells were solely of recipient origin for the duration of these studies46. When Langerhans cells were depleted by irradiation with ultraviolet light, they were replenished from circulating bone-marrow-derived precursors in a CCR2-dependent manner, indicating that a circulating CCR2+ precursor is only utilized when the system is under stress46. Similar observations have been made in a case of human hand-allograft transplantation. At 4.5 years after transplantation, the Langerhans cells were solely of donor origin, supporting the idea that replacement of Langerhans cells by recipient bonemarrow precursors is rare under steady-state conditions47. Whether the CCR2-dependent recruitment of the Langerhans-cell precursor involves recruitment of an inflammatory monocyte or a lineage-committed precursor is unclear. Osteoclasts. Osteoclast precursors are found in the GRANULOCYTE/MACROPHAGE COLONYFORMING UNIT GMCFU PRECURSOR population and can be derived from unfractionated, mature monocytes from peripheral blood48–50. Osteoclast precursors express the receptor for macrophage colony-stimulating factor (M-CSF; also known as CSF1) and depend on it for their development.



REVIEWS (Mcsf op/Mcsf op mice) have a defect in osteoclast development51, and this has been shown to result from a naturally occurring mutation in the gene that encodes M-CSF52, underlining the importance of M-CSF in osteoclast development. Mice with a disrupted Rankl (receptor-activator-of-nuclearfactor-κB ligand) gene also have osteopetrosis53, and this is consistent with the proposed role of RANKL in osteoclast development54. Culture of unfractionated peripheral-blood monocytes with M-CSF and RANKL is sufficient to induce their differentiation OSTEOPETROTIC MICE

b Inflamed epidermis

into osteoclasts, and it has been assumed that osteoclast precursors are monocytes, although this has not been shown in vivo. Alveolar macrophages. Alveolar macrophages have been reported to be derived both from precursors in peripheral blood and from local proliferation of precursors. Early experiments in which radiosensitive precursors of monocytes in mouse bone marrow were transiently depleted with 89Sr — resulting in the loss of peripheral-blood monocytes without affecting the

c Bone Stromal cells or osteoblasts Activated Langerhans cells

Draining lymph

Bone

M-CSF and RANKL

Local proliferation

Mature osteoclast Langerhans cell

a Normal epidermis Marginal zone

White pulp

Marginal-zone macrophage

Metallophilic macrophage

Peripheral blood Circulating monocyte or lineagecommitted precursor

d Alveolar space Alveolar macrophage

White-pulp macrophage

Red pulp

B cell

GM-CSF induces local proliferation

Local proliferation

Marginal sinus Endothelial cell

Red-pulp macrophage

DC Microglial cell

e Central nervous system

f Spleen

OSTEOPETROTIC MICE

(Mcsf op/Mcsf op mice). An inbred strain of mice that suffers from osteopetrosis (stony bones) as a result of deficient function of osteoclasts. The defect has been localized to the gene that encodes macrophage colonystimulating factor (M-CSF; also known as CSF1).

960 | DECEMBER 2005

Figure 2 | Origins of mature cells in the periphery in adult mice. The importance of circulating precursors in the repopulation of tissue-resident macrophage and dendritic cell (DC) populations is not fully understood. Furthermore, the nature of the precursor — that is, whether it is a peripheral-blood monocyte or a circulating lineage-committed precursor — is unclear. In the case of Langerhans cells, local proliferation in the normal epidermis (a) is adequate to maintain Langerhanscell numbers without the requirement for replenishment by circulating precursors; however, after inflammation (b), the CC-chemokine receptor 2 (CCR2)-dependent recruitment of precursors to the epidermis allows repopulation of the tissueresident populations. In bone (c), it is thought that circulating precursors are recruited to the bone surface, where — under the influence of macrophage colony-stimulating factor (M-CSF) and receptor-activator-of-nuclear-factor-κB ligand (RANKL) — they differentiate into mature osteoclasts, which are multinucleate and resorb bone. In the lungs (d), alveolar macrophages can be sustained for long periods by local proliferation; however, bone-marrow-transplantation experiments show that blood-borne precursors can repopulate and divide in the lungs. In the central nervous system (e), microglia are similarly maintained by local proliferation after an initial embryonic phase that populates the central nervous system, but peripheral-blood monocytes also contribute to this pool. In the spleen (f), there is marked heterogeneity of macrophages, with red-pulp, white-pulp, marginal-zone and metallophilic macrophages occupying specific anatomical niches. A simplified schema is shown, which also indicates B cells, DCs and endothelial cells. At present, splenic macrophages are thought to be replenished by a mixture of emigration from the circulating monocyte pool and local proliferation. GM-CSF, granulocyte/ macrophage colony-stimulating factor.

| VOLUME 5



REVIEWS

ALVEOLAR PROTEINOSIS

A disease that is caused by accumulation of surfactant proteins in the alveoli. MICROGLIAL CELL

A type of macrophage that is derived from bone marrow, arborized and present in the parenchyma of the central nervous system. PERIVASCULAR MACROPHAGE

A type of macrophage that lines small blood vessels: for example, near the surface of the brain. MENINGEAL MACROPHAGE

A type of macrophages that is present in the meninges (the three membranes that surround the brain). CHOROIDPLEXUS MACROPHAGE

A type of macrophage that is present at the interface between the blood and the cerebrospinal fluid in the brain. MARGINALZONE MACROPHAGE

A type of macrophage that is present in the splenic marginal zone and is involved in the recognition and clearance of material, such as pathogenderived material, from the splenic circulation. METALLOPHILIC MACROPHAGE

A type of macrophage that surrounds the splenic white pulp, adjacent to the marginal sinus.

number of alveolar macrophages — supported a role for local proliferation; however, these experiments allowed only short-term studies55. Further support for local proliferation of precursors being the main source of alveolar macrophages under normal conditions came from radiation-chimera studies in mice: if the lungs were protected from radiation, then most alveolar macrophages were of recipient origin almost 1 year after treatment56. However, after whole-body irradiation and transfer of GFP-labelled bone marrow, host alveolar macrophages were replaced with cells of donor origin over a long period, indicating that alveolar macrophages can be replenished from the bone marrow57. Studies in humans who have received allogeneic bone-marrow transplants support this argument and indicate that this replenishment occurs by recruitment of precursors, followed by proliferation of these cells in situ58,59. The importance of GM-CSF in lung physiology indicates that this growth factor has a crucial role in the maintenance of alveolar-macrophage populations, the maturation and/or activity of which are impaired in GM-CSF-deficient mice, which suffer from an ALVEOLARPROTEINOSIS-like disease60,61. Although it is assumed that a monocyte subset might be the precursor of alveolar macrophages, this has not been directly tested. Macrophages in the central nervous system. The central nervous system (CNS) contains various macrophage subsets, including MICROGLIA , PERIVASCULAR MACROPHAGES, MENINGEAL MACROPHAGES and CHOROIDPLEXUS MACROPHAGES. Meningeal macrophages are thought to be rapidly replaced by cells of bone-marrow origin, whereas the replacement of perivascular and choroid-plexus macrophages is slower, which indicates that replacement of the individual populations might occur through different mechanisms or might rely on a shared mechanism to differing extents62. For example, microglial cells seem to persist much longer than do other macrophages in the CNS, and their origins have been investigated by a combination of in vivo labelling of dividing cells and bone-marrow-transplantation experiments. Microglial cells can proliferate in situ, and this might be one of the main sources of microglia in adults; however, bonemarrow-derived cells can enter the CNS across the blood–brain barrier and populate the microglial-cell compartment63,64. Monocytes are assumed to be capable of entering the CNS and differentiating into microglia. Splenic macrophages. Aspects of the heterogeneity of splenic macrophages have been summarized elsewhere65,66, but we briefly highlight, in FIG. 2, the discrete anatomical localization of the different macrophage populations. Macrophages in the white pulp include tingible-body macrophages. MARGINALZONE MACROPHAGES are found adjacent to the marginal sinus (through which the circulation passes), and they express pattern-recognition receptors and scavenger receptors, which aid in the clearance of blood-borne pathogens67–69. METALLOPHILIC MACROPHAGES are found adjacent to the white pulp and marginal sinus and


can sample the circulation70; although their function is unknown, they might be important during viral infections71 and other infections. Several studies have attempted to address the question of the origins of these different macrophage populations, both under steady-state conditions and after experimental depletion of the tissue-resident cells72–75, which is important because (as mentioned previously) repopulation after experimental insult might differ from repopulation that occurs in the steady state. It seems that local proliferation does occur under steady-state conditions, at least with respect to some macrophage subsets, such as white-pulp macrophages and metallophilic macrophages. In addition, it seems that circulating precursors also contribute, but the nature of these precursors is unknown. Unlike in humans, the spleen of adult mice is considered to be a haematopoietic organ, and it differs structurally from the human spleen. This might mean that the mouse is not an ideal model for the study of cellular origins for comparison with humans. Studies with the M-CSF-deficient mice (Mcsf op/Mcsf op mice) highlight the different dependencies of certain macrophage subsets on M-CSF for their survival, because these mice lack metallophilic macrophages while maintaining reasonable numbers of most other splenic macrophage subsets76–78. Kupffer cells. Kupffer cells are an important component of the mononuclear-phagocyte system that is present in the liver. The origin of Kupffer cells has been speculated to involve two mechanisms: replenishment by local proliferation, and recruitment of circulating precursors. Twelve hours after administration of 3H-thymidine, ∼1.5% of Kupffer cells incorporate 3 H-thymidine, giving an indication of the low number of cells that are proliferating under steady-state conditions in the adult79. The proportion of Kupffer cells that incorporate 3H-thymidine increases if mice are exposed to whole-body irradiation; however, shielding of the hind legs of the animals during whole-body irradiation indicates that the increase in 3H-thymidine incorporation depends on bone-marrow precursors79. In mouse bone-marrow transplants, donor-derived cells rapidly populate the liver with Kupffer cells (within 3 weeks), and donor Kupffer cells in liver transplants are replaced with similar kinetics80. However, experiments in rats indicate that Kupffer cells might be long-lived81, and temporary depletion of peripheral-blood monocytes with 89Sr had little effect on Kupffer-cell numbers80. So, Kupffer cells, similar to many other macrophage populations, can be replenished by distinct mechanisms, and it is probable that the mechanisms used are affected by inflammation and other factors. Inflammatory-monocyte-derived macrophages. It has long been recognized that inflammatory monocytes (now defined as CCR2+Ly6C+ monocytes) are recruited and differentiate into macrophages at the site of the inflammatory lesion4. A series of in vitro experiments has led to macrophages often being ascribed



REVIEWS

Peripheral blood

Inflamed tissue

Circulating monocyte

Macrophage

Innate activation by TLR ligands (for example, LPS, LTA and PGN) Increased production of pro-inflammatory cytokines, iNOS and ROS Classical activation by IFN-γ and LPS Increased production of pro-inflammatory cytokines, iNOS and ROS; increased expression of MHC class II molecules and CD86; increased antigen presentation; and increased microbicidal activity Alternative activation by IL-4 or IL-13 Increased endocytic activity; increased expression of mannose receptor, dectin-1 and arginase; increased cell growth; increased tissue repair; and increased parasite killing

Deactivation by IL-10, TGF-β, CD200–CD200R, CD47– CD172a or steroids Increased production of IL-10, TGF-β and PGE2; and reduced expression of MHC class II molecules

Figure 3 | Macrophage heterogeneity during inflammation. In inflamed tissues, the precursors of the elicited macrophages — that is, ‘inflammatory’ peripheral-blood monocytes — are now better understood than was previously the case but are still not completely understood. However, the ability of these elicited macrophages to acquire distinct phenotypes and physiological activities has been observed in vitro. For example, when stimulated with interferon-γ (IFN-γ), macrophages show high microbicidal activity and produce reactive oxygen species (ROS). By contrast, when cultured with interleukin-4 (IL-4), IL-10, IL-13 or transforming growth factor-β (TGF-β), a phenotype is generated that promotes tissue repair and suppresses inflammation. Whether these phenotypes are distinct or whether they indicate a continuum of physiological responsiveness remains unclear. CD200R, CD200 receptor; iNOS, inducible nitricoxide synthase; LPS, lipopolysaccharide; LTA, lipoteichoic acid; PGE2, prostaglandin E2; PGN, peptidoglycan; TLR, Toll-like receptor; TNF, tumour-necrosis factor.

activation states 66,82–84 (FIG. 3). These include the following: classical activation, which can be induced by in vitro culture of macrophages with interferon-γ and lipopolysaccharide (which induces TNF production) and is associated with high microbicidal activity, pro-inflammatory cytokine production and cellular immunity; alternative activation, which results from culture in IL-4 or IL-13 and is associated with tissue repair and humoral immunity; innate activation, which is mediated in culture by ligation of receptors such as Toll-like receptors (most of which are expressed by cells of the monocyte–macrophage lineage85) and is associated with microbicidal activity and pro-inflammatory cytokine production; deactivation, which is induced by culture in the presence of cytokines such as IL-10 or transforming growth factor-β, or by ligation of inhibitory receptors such as CD200 receptor or CD172a, and is associated with anti-inflammatory cytokine production and reduced MHC class II expression. Despite these classifications, the extent of plasticity in this system is unclear: that is, it is unclear whether macrophage fate is determined once or whether it is constantly malleable, and it is also unclear whether distinct ‘activation states’ exist in vivo or whether macrophages, instead, show a broad range of phenotypes. It is probable that, in most situations, an inflammatory environment leads to the exposure of macrophages to multiple stimuli, with complex phenotypic consequences. The recent advances in our ability to follow the fate of the monocyte lineage in vivo (such as the identification and adoptive transfer of monocyte subsets),

962 | DECEMBER 2005

| VOLUME 5

in conjunction with single-cell analysis, could be applied to these fundamental questions of macrophage behaviour and fate during inflammation. It would be particularly useful to study situations such as granuloma formation and parasitic infections, in which polarization of macrophage activation has been implicated in disease pathology and in which a strong bias in the nature of the immune response has been observed in vivo. Concluding remarks

The study of animal models has led to a rapid advance in our understanding of the functional consequences of monocyte heterogeneity, and this knowledge can be directly applied to the study of human biology. The ever-expanding pool of genetically manipulated models will accelerate the progression of knowledge in this field. The initial similarities between the species that have been studied reinforce that there is a conserved commonality in the function of these systems, thereby validating their use. Further study will be important to understand how monocytes are recruited to particular inflammatory sites and what determines their differentiation into DCs or macrophages, cell populations that regulate acquired immune responses and/or carry out surveillance of normal and abnormal tissues. Until now, the characterization of monocyte heterogeneity has largely been led by hypothesis-driven investigations and has been restricted to candidate approaches, such as the extensive use of chemokinereceptor-deficient mice in experimental models of inflammation. Advances in the isolation of monocyte subsets, as well as the increasing availability of improved DNA-microarray and proteomic analysis, should provide important and more objective insights into the extent of the cellular diversification of function that is present in the monocyte lineage. Another important issue will be how this current research translates into the advancement of medical knowledge. Techniques for imaging macrophage recruitment and accumulation in humans (for example, in patients with atherosclerosis) have involved transfer of labelled autologous monocytes or in vivo uptake of fluorodeoxyglucose by metabolically active cells, which (although not specific) correlates with macrophage numbers and plaque stability 86. More selective strategies have been adopted in animal models, such as the in vivo administration of radiolabelled chemokines, which would be anticipated to bind selectively to certain subsets of monocytes (among other cells)86. Perhaps a more challenging aspect of these studies is to determine the contribution of monocytes, or alternative, lineage-committed precursors, to the macrophage and DC populations of the body. It is clear that, for the development of appropriate therapeutic interventions, a better understanding of the natural cell lineage of monocytes is required, because long-term modulation of monocyte activities could have implications for the maintenance of tissue populations of macrophages and DCs and therefore could affect homeostasis, immunity and tolerance.



REVIEWS

1.

2.

3.

4.

5.

6.

7.

8.

9.

10.

11.

12.

13.

14.

15.

16.

17.

18.

19.

20.

21.

22.

Volkman, A. & Gowans, J. L. The origin of macrophages from human bone marrow in the rat. Br. J. Exp. Pathol. 46, 62–70 (1965). van Furth, R. & Cohn, Z. A. The origin and kinetics of mononuclear phagocytes. J. Exp. Med. 128, 415–435 (1968). Ebert, R. H. & Florey, H. W. The extravascular development of the monocyte observed in vivo. Brit. J. Exp. Pathol. 20, 342–356 (1939). van Furth, R., Diesselhoff-den Dulk, M. M. & Mattie, H. Quantitative study on the production and kinetics of mononuclear phagocytes during an acute inflammatory reaction. J. Exp. Med. 138, 1314–1330 (1973). Imhof, B. A. & Aurrand-Lions, M. Adhesion mechanisms regulating the migration of monocytes. Nature Rev. Immunol. 4, 432–444 (2004). Passlick, B., Flieger, D. & Ziegler-Heitbrock, H. W. Identification and characterization of a novel monocyte subpopulation in human peripheral blood. Blood 74, 2527–2534 (1989). Characterization of antigenically distinct monocyte subsets in human peripheral blood. Ziegler-Heitbrock, H. W. et al. The novel subset of CD14+/ CD16+ blood monocytes exhibits features of tissue macrophages. Eur. J. Immunol. 23, 2053–2058 (1993). Weber, C. et al. Differential chemokine receptor expression and function in human monocyte subpopulations. J. Leukoc. Biol. 67, 699–704 (2000). Sallusto, F. & Lanzavecchia, A. Efficient presentation of soluble antigen by cultured human dendritic cells is maintained by granulocyte/macrophage colony-stimulating factor plus interleukin 4 and downregulated by tumor necrosis factor α. J. Exp. Med. 179, 1109–1118 (1994). Sanchez-Torres, C., Garcia-Romo, G. S., Cornejo-Cortes, M. A., Rivas-Carvalho, A. & Sanchez-Schmitz, G. CD16+ and CD16– human blood monocyte subsets differentiate in vitro to dendritic cells with different abilities to stimulate CD4+ T cells. Int. Immunol. 13, 1571–1581 (2001). Randolph, G. J., Beaulieu, S., Lebecque, S., Steinman, R. M. & Muller, W. A. Differentiation of monocytes into dendritic cells in a model of transendothelial trafficking. Science 282, 480–483 (1998). Randolph, G. J., Sanchez-Schmitz, G., Liebman, R. M. & Schakel, K. The CD16+ (FcγRIII+) subset of human monocytes preferentially becomes migratory dendritic cells in a model tissue setting. J. Exp. Med. 196, 517–527 (2002). Identification that the CD16+ human peripheral-blood monocyte subset preferentially differentiates into DCs after endothelial reverse transmigration. Ancuta, P. et al. Fractalkine preferentially mediates arrest and migration of CD16+ monocytes. J. Exp. Med. 197, 1701–1707 (2003). Grage-Griebenow, E. et al. Identification of a novel dendritic cell-like subset of CD64+/CD16+ blood monocytes. Eur. J. Immunol. 31, 48–56 (2001). Grage-Griebenow, E., Flad, H. D. & Ernst, M. Heterogeneity of human peripheral blood monocyte subsets. J. Leukoc. Biol. 69, 11–20 (2001). Palframan, R. T. et al. Inflammatory chemokine transport and presentation in HEV: a remote control mechanism for monocyte recruitment to lymph nodes in inflamed tissues. J. Exp. Med. 194, 1361–1373 (2001). Kuziel, W. A. et al. Severe reduction in leukocyte adhesion and monocyte extravasation in mice deficient in CC chemokine receptor 2. Proc. Natl Acad. Sci. USA 94, 12053–12058 (1997). Kurihara, T., Warr, G., Loy, J. & Bravo, R. Defects in macrophage recruitment and host defense in mice lacking the CCR2 chemokine receptor. J. Exp. Med. 186, 1757–1762 (1997). Lu, B. et al. Abnormalities in monocyte recruitment and cytokine expression in monocyte chemoattractant protein 1-deficient mice. J. Exp. Med. 187, 601–608 (1998). Gu, L. et al. Absence of monocyte chemoattractant protein-1 reduces atherosclerosis in low density lipoprotein receptor-deficient mice. Mol. Cell 2, 275–281 (1998). Serbina, N. V., Salazar-Mather, T. P., Biron, C. A., Kuziel, W. A. & Pamer, E. G. TNF/iNOS-producing dendritic cells mediate innate immune defense against bacterial infection. Immunity 19, 59–70 (2003). Geissmann, F., Jung, S. & Littman, D. R. Blood monocytes consist of two principal subsets with distinct migratory properties. Immunity 19, 71–82 (2003). The first modern functional characterization and cell-fate studies of mouse peripheral-blood monocyte subsets during steady-state conditions and inflammation.

23. Tedder, T. F., Steeber, D. A. & Pizcueta, P. L-selectindeficient mice have impaired leukocyte recruitment into inflammatory sites. J. Exp. Med. 181, 2259–2264 (1995). 24. Rosen, S. D. Ligands for L-selectin: homing, inflammation, and beyond. Annu. Rev. Immunol. 22, 129–156 (2004). 25. Sunderkotter, C. et al. Subpopulations of mouse blood monocytes differ in maturation stage and inflammatory response. J. Immunol. 172, 4410–4417 (2004). 26. de Bruijn, M. F. et al. Distinct mouse bone marrow macrophage precursors identified by differential expression of ER-MP12 and ER-MP20 antigens. Eur. J. Immunol. 24, 2279–2284 (1994). 27. Randolph, G. J., Inaba, K., Robbiani, D. F., Steinman, R. M. & Muller, W. A. Differentiation of phagocytic monocytes into lymph node dendritic cells in vivo. Immunity 11, 753–761 (1999). 28. Qu, C. et al. Role of CCR8 and other chemokine pathways in the migration of monocyte-derived dendritic cells to lymph nodes. J. Exp. Med. 200, 1231–1241 (2004). Suggestion that a third, intermediate monocyte phenotype in mouse peripheral blood might be an inflammatory monocyte that is prone to differentiate into DCs. 29. Ahuja, V., Miller, S. E. & Howell, D. N. Identification of two subpopulations of rat monocytes expressing disparate molecular forms and quantities of CD43. Cell. Immunol. 163, 59–69 (1995). 30. Chamorro, S. et al. Phenotypic characterization of monocyte subpopulations in the pig. Immunobiology 202, 82–93 (2000). 31. Chamorro, S. et al. Phenotypic and functional heterogeneity of porcine blood monocytes and its relation with maturation. Immunology 114, 63–71 (2005). 32. Chamorro, S. et al. In vitro differentiation of porcine blood CD163– and CD163+ monocytes into functional dendritic cells. Immunobiology 209, 57–65 (2004). 33. Gordon, S. Biology of the macrophage. J. Cell Sci. Suppl. 4, 267–286 (1986). 34. Gordon, S. The role of the macrophage in immune regulation. Res. Immunol. 149, 685–688 (1998). 35. Austyn, J. M. & Gordon, S. F4/80, a monoclonal antibody directed specifically against the mouse macrophage. Eur. J. Immunol. 11, 805–815 (1981). 36. Dijkstra, C. D., Van Vliet, E., Dopp, E. A., van der Lelij, A. A. & Kraal, G. Marginal zone macrophages identified by a monoclonal antibody: characterization of immuno- and enzyme-histochemical properties and functional capacities. Immunology 55, 23–30 (1985). 37. Kraal, G. & Janse, M. Marginal metallophilic cells of the mouse spleen identified by a monoclonal antibody. Immunology 58, 665–669 (1986). 38. Quinn, J. M. & Gillespie, M. T. Modulation of osteoclast formation. Biochem. Biophys. Res. Commun. 328, 739–745 (2005). 39. McCusker, K. & Hoidal, J. Characterization of scavenger receptor activity in resident human lung macrophages. Exp. Lung Res. 15, 651–661 (1989). 40. Palecanda, A. et al. Role of the scavenger receptor MARCO in alveolar macrophage binding of unopsonized environmental particles. J. Exp. Med. 189, 1497–1506 (1999). 41. Taylor, P. R. et al. The β-glucan receptor, dectin-1, is predominantly expressed on the surface of cells of the monocyte/macrophage and neutrophil lineages. J. Immunol. 169, 3876–3882 (2002). 42. Surh, C. D. & Sprent, J. T-cell apoptosis detected in situ during positive and negative selection in the thymus. Nature 372, 100–103 (1994). 43. Tabe, H., Kawabata, I., Koba, R. & Homma, T. Cell dynamics in the germinal center of the human tonsil. Acta Otolaryngol. Suppl. 523, 64–67 (1996). 44. Smythies, L. E. et al. Human intestinal macrophages display profound inflammatory anergy despite avid phagocytic and bacteriocidal activity. J. Clin. Invest. 115, 66–75 (2005). 45. van Furth, R., Diesselhoff-den Dulk, M. M., Sluiter, W. & Van Dissel, J. T. in Mononuclear Phagocytes: Characteristics, Physiology and Function (ed. van Furth, R.) 201–210 (Martinus Nijhoff, Dordrecht, 1985). 46. Merad, M. et al. Langerhans cells renew in the skin throughout life under steady-state conditions. Nature Immunol. 3, 1135–1141 (2002). Interesting demonstration of the differences in requirement for circulating precursors or local proliferation in the renewal of Langerhans cells during steady-state conditions and inflammation. 47. Kanitakis, J., Petruzzo, P. & Dubernard, J. M. Turnover of epidermal Langerhans’ cells. N. Engl. J. Med. 351, 2661–2662 (2004).


48. Kurihara, N., Chenu, C., Miller, M., Civin, C. & Roodman, G. D. Identification of committed mononuclear precursors for osteoclast-like cells formed in long term human marrow cultures. Endocrinology 126, 2733–2741 (1990). 49. Udagawa, N. et al. Origin of osteoclasts: mature monocytes and macrophages are capable of differentiating into osteoclasts under a suitable microenvironment prepared by bone marrow-derived stromal cells. Proc. Natl Acad. Sci. USA 87, 7260–7264 (1990). 50. Matsuzaki, K. et al. Osteoclast differentiation factor (ODF) induces osteoclast-like cell formation in human peripheral blood mononuclear cell cultures. Biochem. Biophys. Res. Commun. 246, 199–204 (1998). 51. Marks, S. C. Jr & Lane, P. W. Osteopetrosis, a new recessive skeletal mutation on chromosome 12 of the mouse. J. Hered. 67, 11–18 (1976). 52. Yoshida, H. et al. The murine mutation osteopetrosis is in the coding region of the macrophage colony stimulating factor gene. Nature 345, 442–444 (1990). 53. Kong, Y. Y. et al. OPGL is a key regulator of osteoclastogenesis, lymphocyte development and lymph-node organogenesis. Nature 397, 315–323 (1999). 54. Arai, F. et al. Commitment and differentiation of osteoclast precursor cells by the sequential expression of c-Fms and receptor activator of nuclear factor κB (RANK) receptors. J. Exp. Med. 190, 1741–1754 (1999). 55. Sawyer, R. T., Strausbauch, P. H. & Volkman, A. Resident macrophage proliferation in mice depleted of blood monocytes by strontium-89. Lab. Invest. 46, 165–170 (1982). 56. Tarling, J. D., Lin, H. S. & Hsu, S. Self-renewal of pulmonary alveolar macrophages: evidence from radiation chimera studies. J. Leukoc. Biol. 42, 443–446 (1987). 57. Matute-Bello, G. et al. Optimal timing to repopulation of resident alveolar macrophages with donor cells following total body irradiation and bone marrow transplantation in mice. J. Immunol. Methods 292, 25–34 (2004). 58. Thomas, E. D., Ramberg, R. E., Sale, G. E., Sparkes, R. S. & Golde, D. W. Direct evidence for a bone marrow origin of the alveolar macrophage in man. Science 192, 1016–1018 (1976). 59. Nakata, K. et al. Augmented proliferation of human alveolar macrophages after allogeneic bone marrow transplantation. Blood 93, 667–673 (1999). 60. Stanley, E. et al. Granulocyte/macrophage colonystimulating factor-deficient mice show no major perturbation of hematopoiesis but develop a characteristic pulmonary pathology. Proc. Natl Acad. Sci. USA 91, 5592–5596 (1994). 61. Hamilton, J. A. GM-CSF in inflammation and autoimmunity. Trends Immunol. 23, 403–408 (2002). 62. Hickey, W. F. Leukocyte traffic in the central nervous system: the participants and their roles. Semin. Immunol. 11, 125–137 (1999). 63. Lawson, L. J., Perry, V. H. & Gordon, S. Turnover of resident microglia in the normal adult mouse brain. Neuroscience 48, 405–415 (1992). 64. de Groot, C. J., Huppes, W., Sminia, T., Kraal, G. & Dijkstra, C. D. Determination of the origin and nature of brain macrophages and microglial cells in mouse central nervous system, using non-radioactive in situ hybridization and immunoperoxidase techniques. Glia 6, 301–309 (1992). 65. Kraal, G. Cells in the marginal zone of the spleen. Int. Rev. Cytol. 132, 31–74 (1992). 66. Taylor, P. R. et al. Macrophage receptors and immune recognition. Annu. Rev. Immunol. 23, 901–944 (2005). 67. van der Laan, L. J. et al. Regulation and functional involvement of macrophage scavenger receptor MARCO in clearance of bacteria in vivo. J. Immunol. 162, 939–947 (1999). 68. Kang, Y. S. et al. SIGN-R1, a novel C-type lectin expressed by marginal zone macrophages in spleen, mediates uptake of the polysaccharide dextran. Int. Immunol. 15, 177–186 (2003). 69. Geijtenbeek, T. B. et al. Marginal zone macrophages express a murine homologue of DC-SIGN that captures blood-borne antigens in vivo. Blood 100, 2908–2916 (2002). 70. Taylor, P. R. et al. Development of a specific system for targeting protein to metallophilic macrophages. Proc. Natl Acad. Sci. USA 101, 1963–1968 (2004). 71. Eloranta, M. L. & Alm, G. V. Splenic marginal metallophilic macrophages and marginal zone macrophages are the major interferon-α/β producers in mice upon intravenous challenge with herpes simplex virus. Scand. J. Immunol. 49, 391–394 (1999). 72. van Furth, R. & Diesselhoff-den Dulk, M. M. Dual origin of mouse spleen macrophages. J. Exp. Med. 160, 1273–1283 (1984).



REVIEWS

73. van Rooijen, N., Kors, N. & Kraal, G. Macrophage subset repopulation in the spleen: differential kinetics after liposome-mediated elimination. J. Leukoc. Biol. 45, 97–104 (1989). 74. Van Rooijen, N., Kors, N., van de Ende, M. & Dijkstra, C. D. Depletion and repopulation of macrophages in spleen and liver of rat after intravenous treatment with liposome-encapsulated dichloromethylene diphosphonate. Cell Tissue Res. 260, 215–222 (1990). 75. Wijffels, J. F., de Rover, Z., Beelen, R. H., Kraal, G. & van Rooijen, N. Macrophage subpopulations in the mouse spleen renewed by local proliferation. Immunobiology 191, 52–64 (1994). 76. Witmer-Pack, M. D. et al. Identification of macrophages and dendritic cells in the osteopetrotic (op/op) mouse. J. Cell Sci. 104, 1021–1029 (1993). 77. Takahashi, K., Umeda, S., Shultz, L. D., Hayashi, S. & Nishikawa, S. Effects of macrophage colony-stimulating factor (M-CSF) on the development, differentiation, and maturation of marginal metallophilic macrophages and marginal zone macrophages in the spleen of osteopetrosis (op) mutant mice lacking functional M-CSF activity. J. Leukoc. Biol. 55, 581–588 (1994). 78. Cecchini, M. G. et al. Role of colony stimulating factor-1 in the establishment and regulation of tissue macrophages during postnatal development of the mouse. Development 120, 1357–1372 (1994). 79. Crofton, R. W., Diesselhoff-den Dulk, M. M. & van Furth, R. The origin, kinetics, and characteristics of the Kupffer cells in the normal steady state. J. Exp. Med. 148, 1–17 (1978).

964 | DECEMBER 2005

80. Naito, M., Hasegawa, G. & Takahashi, K. Development, differentiation, and maturation of Kupffer cells. Microsc. Res. Tech. 39, 350–364 (1997). 81. Bouwens, L., Baekeland, M., De Zanger, R. & Wisse, E. Quantitation, tissue distribution and proliferation kinetics of Kupffer cells in normal rat liver. Hepatology 6, 718–722 (1986). 82. Gordon, S. Alternative activation of macrophages. Nature Rev. Immunol. 3, 23–35 (2003). 83. Mosser, D. M. The many faces of macrophage activation. J. Leukoc. Biol. 73, 209–212 (2003). 84. Goerdt, S. & Orfanos, C. E. Other functions, other genes: alternative activation of antigen-presenting cells. Immunity 10, 137–142 (1999). 85. Takeda, K., Kaisho, T. & Akira, S. Toll-like receptors. Annu. Rev. Immunol. 21, 335–376 (2003). 86. Davies, J. R., Rudd, J. F., Fryer, T. D. & Weissberg, P. L. Targeting the vulnerable plaque: the evolving role of nuclear imaging. J. Nucl. Cardiol. 12, 234–246 (2005). 87. Aschoff, L. Das Reticulo-endotheliale System. Ergeb. Inn. Med. Kinderheilkd. 26, 1–118 (1924) (in German). 88. Lichanska, A. M. & Hume, D. A. Origins and functions of phagocytes in the embryo. Exp. Hematol. 28, 601–611 (2000). 89. Shepard, J. L. & Zon, L. I. Developmental derivation of embryonic and adult macrophages. Curr. Opin. Hematol. 7, 3–8 (2000). 90. Takahashi, K., Donovan, M. J., Rogers, R. A. & Ezekowitz, R. A. Distribution of murine mannose receptor expression from early embryogenesis through to adulthood. Cell Tissue Res. 292, 311–323 (1998).

| VOLUME 5

91. Hughes, D. A. & Gordon, S. Expression and function of the type 3 complement receptor in tissues of the developing mouse. J. Immunol. 160, 4543–4552 (1998). 92. Hume, D. A., Monkley, S. J. & Wainwright, B. J. Detection of c-fms protooncogene in early mouse embryos by whole mount in situ hybridization indicates roles for macrophages in tissue remodelling. Br. J. Haematol. 90, 939–942 (1995). 93. Herbomel, P., Thisse, B. & Thisse, C. Ontogeny and behaviour of early macrophages in the zebrafish embryo. Development 126, 3735–3745 (1999). 94. Henderson, R. B., Hobbs, J. A., Mathies, M. & Hogg, N. Rapid recruitment of inflammatory monocytes is independent of neutrophil migration. Blood 102, 328–335 (2003). 95. Taylor, P. R., Brown, G. D., Geldhof, A. B., Martinez-Pomares, L. & Gordon, S. Pattern recognition receptors and differentiation antigens define murine myeloid cell heterogeneity ex vivo. Eur. J. Immunol. 33, 2090–2097 (2003).


Online links DATABASES The following terms in this article are linked online to: Entrez Gene: http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=gene CCR2 | CD14 | CD16 | CD43 | CD62L | CD163 | CX3CR1 | Ly6C Access to this interactive links box is free online.



PERSPECTIVES SCIENCE AND SOCIETY

Art and science collaborations in the United Kingdom Stephen Webster Abstract | In this Perspective article, I consider ways in which the contemporary arts and sciences can interact with each other, and I outline the current opportunities for funding in the United Kingdom. I examine the view that, in an art–science collaboration, it is the artist who benefits most, and I cautiously suggest that this is an oversimplification. Evidence from interviews with scientists who have been involved in these collaborations shows that artistic experience and skills are of value in the scientific research process.

In 2004, the Summer Science Exhibition of The Royal Society (United Kingdom) was titled Be Amazed, and it hosted an unusual installation. Down in the basement, away from the crowds, stood four pillars of clear acrylic, each apparently entombing a person. In fact, Family Portrait, by Marilène Oliver, is constructed from magnetic-resonanceimaging scans of her parents, her sister Sophie and herself (FIG. 1). By assembling the scans on top of each other and in the right sequence, Oliver achieves a striking effect. The body sections are composed of scientific images that ordinarily would be used to reveal aspects of a particular organ or structure. In Oliver’s sculpture, these scientific glimpses are recomposed, and the solid form emerges. The Royal Society might be reflecting a trend. It is no surprise these days to find art turning up in science museums, and science in art galleries. Such a crossover seemed more remarkable back in 1999, when the Hayward

Gallery (London) presented Spectacular Bodies, an exhibition that complemented numerous drawings and models from the history of anatomy with a selection of contemporary artistic studies of the body1. Then, in 2001, with the Human Genome Project in the headlines, the National Portrait Gallery (London) unveiled Marc Quinn’s portrait of the geneticist Sir John Sulston. The ensuing publicity focused less on the relative rarity of scientists in the gallery and more on the extraordinary form of the portrait of Sulston, because it turned out to be a quadrat of agar covered in Sulston’s DNA. Mirroring these developments in the capital’s art venues, in 2002, the Science Museum (London) opened a gallery that specifically relied on the work of both artists and scientists. The first exhibition — Head On: Art with the Brain in Mind — had the human mind as its subject and clearly showed how contemporary artists might work with concepts that are most commonly considered to be the province of scientists. One of the many striking installations at Head On was Annie Cattrell’s pair of sculptures Seeing and Hearing, each resembling a mud splash frozen in a large Perspex block but, on examination, revealed to be resin shapes derived from brain-scan data. Cattrell, with help from the neurophysiologist Mark Lythgoe, had used the technique known as rapid prototyping to render into physical form brain-scan data taken from an individual involved in the process of smelling or hearing. The artist then had literally ‘materialized’ for public view the results of a technology that is regarded by


scientists as bringing us closer to the material basis of human thought. Such detailed artistic attention to the techniques of science was again on display in 2005, at White Cube (London), where Marc Quinn held an exhibition titled Chemical Life Support. In a large and nearly empty room, a few naked figures, apparently made of marble or alabaster, relaxed contentedly on the floor. The surprise lay in discovering that each sculpture depicted a real person whose life depends on the insights of immunology. More specifically, the wax polymer that was used to make these sculptures contained the drugs that keep each individual alive. So, Silvia Petretti is a model of an HIV-positive person and was cast with a mix of wax and antiretroviral drugs. Another of the sculptures provided a clue to the artist’s interest in science. For, in the middle of the room, lay

Figure 1 | Family portrait (Sophie, detail), by Marilène Oliver. An image of part of the installation Family portrait, by Marilène Oliver, is shown. Each plate carries a magneticresonance-imaging scan of the artist’s sister, Sophie. Stacked together as a column, the body re-emerges. Image courtesy of Beaux Arts (United Kingdom) and M. Oliver.



PERSPECTIVES Innoscience, a sculpture of Quinn’s baby son Lucas, apparently the embodiment of health and happiness. Yet the wax matrix included the dozen or so amino acids that formed the artificial feed at one time given to him to overcome his lactose intolerance and ensure his well-being. These few examples show artists taking enormous pains to ‘get the science right’. This is not to say that their work is simply illustrative of the power of modern science. On the contrary, Marilène Oliver’s work can be seen as drawing attention to the manner in which bio-imaging can fragment bodies, focusing the medical gaze on segments of the body rather than on the whole person. Regarding Annie Cattrell’s ‘blocks of thought’, these display, in the clearest way, the power of medical technology to localize thought, rather sharply reminding the viewer of the ambitions of modern science to explain consciousness2. These few interpretations are intended simply to show that there is nothing trivial about artists’ current preconceptions of science. Clearly, these artists are enriching their work with a close attention to contemporary science. My focus in this article, however, is on the converse possibility. Can science gain by close attention to the work of artists? More specifically, when scientists do collaborate with artists, where should we look to find the benefits? The cultural position of science

For signs of these benefits, the first place to look might be in science’s dealings with the public. To risk a generalization, artists differ from scientists in being rather expert at exhibiting their work in public, and they often seem untroubled by any resulting controversy. Even though scientists are exhorted3 to discuss their work with a lay audience, they are unlikely to consider their research to be essentially public. Perhaps, then, one dimension in which the arts can enrich and assist the sciences is public relations. This was a remote idea 20 years ago, when Sir Walter Bodmer chaired The Royal Society’s investigation of the public understanding of science4. The ensuing model of science communication took it as straightforward that any problems between science and its public were due to the ignorance of the public and the sensationalism of the news media. Take those distressing factors out of the equation, and the message of science would get through more evenly and to greater public acclaim. Today, the talk is more of dialogue, of a ‘listening science’ and of public engagement. Successful or not, this engagement is becoming remarkably diverse

966 | DECEMBER 2005

in its methods. Science communication is just as likely to be taking place in café meetings, stage performances and gallery events as it is through celebrity lectures or solemn television documentaries. With the field opened wide, it is no surprise to find artists taking up positions. They are experts in the visual, want their work to be public and are unafraid of their work being interpreted in a myriad of ways. Art has a great interest in representing the body, so it is bound to find in modern science some highly intriguing topics: for example, the clutch of controversies and concepts that are associated with human genetics, stemcell technology, and cloning. Given this, it is no wonder that there is an arts influence in our science museums. No wonder too that some artists are knocking on the doors of our scientific institutes, contacting the scientists at these institutes and asking for access to the secrets within. In the spirit of symmetry, the following question therefore seems natural. What is the nature of the scientist’s involvement? Is the scientist simply a fount of knowledge, providing advice and correcting the more painful inaccuracies? Does the scientist get nothing back? We have seen that artists are welcome partners in the communication of science. But when it comes to the ‘core’ activity of science — research — must artists politely withdraw, recognizing that the thoughts and actions of research science form a sealed-off area? Some commentators, sceptical of art–science ventures, have indeed made science’s alleged separateness the basis for an argument that art has little, if anything, to contribute to scientific research. For example, reviewing the Science Museum’s exhibition Head On, the eminent embryologist Sir Lewis Wolpert described himself as being bemused by “the current vogue that art and science should be brought together” and declared that “Although science has had a strong influence on certain artists … art has contributed virtually nothing to science.”5 In making his argument, Wolpert made several key distinctions. He said that science provides explanations rather than viewpoints and, compared with art, requires for its appreciation a “much greater and quite different intellectual effort.”5 Lying behind Wolpert’s words is a tacit philosophy of science that we are all familiar with. It is what philosophers have called ‘the received view’6 or even ‘the Legend’7, and it imputes to scientific research a core method of rigour and reliability. The assuredness with which this method leads you towards the truth depends only on the tenacity with

| VOLUME 5

which you apply it and the skill with which you prevent extraneous factors — such as funding problems, conflicts of interest and competitive animosity — from interfering. In a sense, this philosophy sees in science a great split. On the one side, there are the rules and the methods of science, principles that, when applied, inevitably produce scientific advances. On the other side, there are the scientists, the all-too-human operators of scientific method. Every distraction weighs them down; their ability to carry out good science is threatened constantly by the swirling forces of emotion, by bias, by the demands of students and so on. As Nature Immunology put it in a recent editorial: “Science is objective, but scientists are not.”8 Assessing the validity of this distinction, and largely doubting that it can hold, has been the philosopher’s workhorse for at least 40 years. Much recent scholarship has involved the study of scientists in their workplace, and from this, it has been concluded that it is not possible to find an abstract scientific method that operates separately from the wider culture of scientific life. All of these points have relevance for the way in which we evaluate the art–science collaboration. Let us suppose that scientific research is, indeed, a kind of inner sanctum, separable from the seething environment that is a scientific institute. Then, artists must see their role in relation to science as that of illustrator, critic or commentator. These would be worthwhile tasks whether or not scientists took any notice. But if scientific research is not defined as such (that is, if there is no ‘heart of the matter’), then what would keep artistic collaborators out of the research process? In all of the recent art–science enterprises, it seems implausible that artists’ methods and concepts have never entered the treasure house of science, its research output. In the last section of this article, I give examples in which such an influence might have occurred. Before that, however, I describe how the art–science interface has been attracting funding and institutional support. For those who are interested in starting their own projects, the web sites in the Online links box provide more information. Current activity and funding

How, in practice, do art–science encounters come about? Frequently, it is through artists taking up a residency in a scientific institute. Artists have been sent to war, to Antarctica and to the bottom of the ocean. Why should they not be invited into our laboratories? Many institutes in the United Kingdom have



PERSPECTIVES hosted residencies, including the following: in Cambridge, the Wellcome Trust Sanger Institute, and the Cavendish Laboratory at the University of Cambridge; and in London, the Natural History Museum, University College London Institute of Child Health, and the National Institute for Medical Research (see the Online links box). Intriguingly, the opposite model also exists: that is, the scientist in residence. The Institute of Contemporary Arts (London) (see the Online links box) inaugurated such a programme in 2002, with the appointment of Daniel Glaser, a neurophysiologist at University College London. In addition, for more than ten years, the agency The Arts Catalyst (see the Online links box) has been organizing and supporting collaborations. All such programmes require funding, and there are various models for this funding. Sometimes, the institute itself makes the investment. The National Institute for Medical Research is an enlightened example in that it employs the arts curator Simon Gould to manage both the residencies and the fund raising. Both the Natural History Museum and the Science Museum also have curatorial programmes. But there are plenty of grants that are more generally available. In 1996, the Wellcome Trust set up a consortium called sciart (see the Online links box), pooling not only its own resources but also those of Arts Council England, the National Endowment for Science, Technology and the Arts, and the Calouste Gulbenkian Foundation. An early — and startling — product was Helen Storey’s exhibition titled Primitive Streak. A collaboration between Helen (a fashion designer) and her sister Kate (an embryologist) led to a series of 27 dresses representing key stages in the first 1,000 hours of a human embryo. These were exhibited at the Institute of Contemporary Arts in late 1997 and have since toured internationally, drawing huge audiences and with bookings set to continue. According to the Helen Storey Foundation web site (see the Online links box), even the British Prime Minister, Tony Blair, took notice: “Primitive Streak is an adventurous and highly successful fusion of art and science, and a fine example of British innovation and creativity. I wish the exhibition every success.” BOX 1 lists eight art–science collaborations that were funded or partially funded by the Wellcome Trust. The Arts & Humanities Research Council (see the Online links box) saw the need for some more-formal research support, and working with Arts Council England, they set up a scheme, which was announced in 2002, to fund Arts and Science Research

Fellowships. If an artist could find a supportive scientific institute, and a scientist to work with, then he or she could apply for funding of up to UK£35,000, for a oneyear tenure. When 16 such fellowships were finally announced in September 2003, the press release from Arts Council England stated that the new fellowships “will explore how spending time together in shared research settings will contribute to the store

of knowledge within science and art, and explore how art can contribute to science and science to art in terms of different ways of working and thinking.”9 The scheme ran again in 2005, with the additional support of the Scottish Arts Council; 11 fellowships have just been announced. A scientific research council has also seen the value of giving formal encouragement to such collaborations. The Engineering

Box 1 | Eight art–science projects, 2002–2005

Mapping Perception • Collaborators: film-maker Andrew Kötting and neurophysiologist Mark Lythgoe • Project: scientific and artistic representations of brain function, centred on Kötting’s daughter Eden, who has the rare neurological condition Joubert’s syndrome • Product: film and installation

On the Scent • Collaborators: artists Helen Paris and Leslie Hill, and olfactory scientist Upinder Bhalla • Project: scientific and artistic investigation of the sense of smell, especially its ability to trigger memory and emotions • Product: live performance

Midge Bait • Collaborators: artist Alison Hayes, film-maker David Mackenzie and scientist Alison Blackwell • Project: cinematic investigation of the flight patterns and attacking behaviour of the Highland midge (Culicoides impunctatus) • Product: moving and still images of midge flight for subsequent scientific analysis

How To Live • Collaborators: artist Bobby Baker and psychologist Richard Hallam • Project: study of cognitive behavioural therapy • Product: theatre performance

Medusae • Collaborators: artist Dorothy Cross and marine biologist Tom Cross • Project: research into the life of Irish naturalist Maude Delap (1866–1953) and analysis of the biomechanics of the box jellyfish Chironex fleckeri • Product: film

D’Alembert’s Dream • Collaborators: artist Phoebe von Held and scientists from the National Institute for Medical Research • Project: dramatized comparison of Denis Diderot’s eighteenth-century visions of the future with those of contemporary scientists • Product: film

Magic Forest • Collaborators: artist Andrew Carnie and neurophysiologist Richard Wingate • Project: new representations of the migration of developing brain cells • Product: installation with slow-dissolve photography

Bone Orchestra • Collaborators: film-makers Polly Nash and Jo Cammack, and scientist Michael Horton • Project: study of the public understanding of the impact of osteoporosis • Product: film See the Online links box for further information about these collaborations.




PERSPECTIVES

Figure 2 | A water sculpture by Philip Kilner. An image of a water sculpture by Philip Kilner is shown. Each of the two surfaces receives a continuous stream but delivers surges of outflow as a result of cyclic wave propagation in the contours of the cavity. One surface can be seen in its emptying phase while the other is in its filling phase. This sculpture is located at the main entrance of the Royal Brompton Hospital (London). Image courtesy of P. Kilner.

and Physical Sciences Research Council (see the Online links box) has put in place a £500,000 scheme for supporting ‘research networks’. These comprise groups of as many as 20 scientists and artists, and these groups meet regularly to investigate an art–science research agenda. At present, there are 15 such networks. It will take time for the results to emerge, but the titles of the projects are suggestive of rich interdisciplinary possibilities. They include the following: Art and Science in Motion Perception; EngineeringArt: A Network Dedicated to Exploring the Art and Science of Materials; and Leonardo: Culture, Creativity and Interaction Design. All of these projects come under the Culture & Creativity funding stream and are grounded in the idea that art–science partnerships, when properly developed, can radically improve innovation in engineering. Collaborations in practice

It is clear that art–science collaborations are attracting great interest, both in the research community and among science communicators. In this final section, I describe two recent art–science collaborations, and I search directly for clues to the benefits that science might be gaining. In these particular examples, two factors emerge. The first example, Project Façade (see the Online links box), shows that, perhaps surprisingly, artists can have a direct

968 | DECEMBER 2005

technical influence on biomedical science. The second example, The Fluent Heart (see the Online links box), indicates an even more radical possibility: that artistic intuition can have a role in the development of theory and in the interpretation of data. The first collaboration, Project Façade, is based in the Department of Oral Maxillofacial Medicine and Pathology at the Dental Institute at Guy’s, King’s College and St Thomas’ Hospitals (London). Materials scientist Ian Thompson and artist Paddy Hartley work together to make glass implants for patients who need facial surgery. Thompson’s interest is in bioactive glass and its potential in bone grafting. Hartley is best known for a series of remarkable face corsets, which were recently exhibited both in the Victoria and Albert Museum (London) and in the Science Museum’s exhibition Future Face. Among his skills is an expertise in the lost wax method of casting, gained during his years as a sculptor. The individual needs of patients who are attending the clinic pose a great challenge. Can an implant that is optimally matched to a patient’s face be prepared? Bioactive glass cannot be mechanically ground to the correct shape. It must be cast, a skill that is unusual to the laboratory but familiar in the sculptor’s workshop. The properties of bioactive glass, and the effectiveness of the collaboration between Thompson and Hartley, allow the institute to manufacture glass implants uniquely matched to each patient.

| VOLUME 5

Hartley supplies assistance that is technical rather than conceptual. According to Thompson, the collaboration is an amalgam of skills. Working together, Hartley and Thompson can supply surgeons with a more finely crafted implant, and surgery time is reduced. It is an example of a scientific development — bioactive glass — being ‘finessed’ into the varying niches of individual treatment by the craft skills of an artist. The second collaboration considered here is The Fluent Heart. Its most high-profile product is a ballet; this was choreographed by Wayne McGregor and has music that was composed by Sir John Tavener. For our purposes, what is more remarkable, however, is the involvement of Philip Kilner, a heartimaging specialist at the Royal Brompton Hospital (London). Kilner was in contact with Tavener for clinical reasons: the composer has a heart condition. Tavener was powerfully affected by seeing the movements of his heart in Kilner’s images: “The pumping of the heart’s chambers and the movement of the blood around the arteries — it looked beautiful to me, like a dance.”10 A composition came to mind; Wayne McGregor was contacted; and with the help of a grant from the Wellcome Trust, the ballet Amu was commissioned. It premiered in the Sadler’s Wells Theatre (London) in September 2005. Tavener’s music is known for its mystical orientation. It is already interesting that a scientist’s work could be part of his inspiration. But that seems to leave untouched the question that animates this article: namely, in what manner can artists influence the work of scientists? It emerges that Kilner’s interest in the potentially beneficial exchange between artistic and scientific approaches can be traced to experiences earlier in his career. He spent several years away from medicine and studied sculpture at Emerson College (Forest Row, East Sussex, United Kingdom). It is relevant that this independent college, where artistic and scientific approaches are regarded as complementary, owes much to influences that can be traced to Johann Wolfgang von Goethe (1749–1832), a perceptive scientist, as well as a dramatist and poet11. At Emerson College, Kilner became interested in the movement of water over sculpted surfaces, to which he was introduced by John Wilkes, one of his tutors. He became fascinated by the way in which an appropriately sculpted cavity, with water running continuously into it, can give rise to cyclic patterns of flow and a varying output. In Kilner’s sculptures — one can be seen in the main entrance of the Royal Brompton Hospital (FIG. 2) — it is the relationship



PERSPECTIVES between the shape of the sculpted form and the dynamics of the water that produces the oscillating patterns of flow. It was his training as a sculptor, combined with a creative and interactive engagement with streaming water, that deepened Kilner’s appreciation of the relationships between form and flow. This interest soon led to specific questions about the internal dynamics of the heart. After working as a sculptor and teacher, Kilner re-entered medicine, winning a British Heart Foundation grant to collaborate with a heart surgeon. His project involved experiments with carefully shaped flow models, and it led to the modification12 of an operation that is carried out for patients with one of a group of unusual but severe congenital heart malformations. Wishing to study the living heart, Kilner then moved into the area of magnetic resonance imaging of the heart. His subsequent scientific papers, which include his drawings, are necessarily technical13. But to Kilner himself, it is clear that his years as an artist, working creatively with form and flow, complement his scientific training as the foundation of his work. How might this artistic training influence Kilner’s work in the field of cardiology? According to the scientist himself, his unusual background assists him to apply a set of skills to the heart that range from the reductionist to the holistic. His functional interpretation of the overall form of the heart depends crucially both on a reductionist understanding of anatomy and mechanics and on a more inclusive appreciation of the relationships between interconnected forms and multidirectional streams, for which his experience as a water sculptor equipped him. The dynamics of fluid flow in a heart are truly complex, and they resist easy reduction to hydrodynamic principles. Yet the role of the heart-imaging specialist is to understand quickly what is happening. According to Kilner, his background in sculpting fluid forms gives him an insight — an intuitive understanding — that is a useful complement to his scientific, analytical approach.

Nature columnist — Martin Kemp. He has described the artist’s style of visual insight as “a gate into the system”14, and he suggests that, although scientific enquiry is usefully carried out at many levels, the reductionist mode is often dominant. By contrast, artists are drawn less to nature at the molecular level and more to its bodily, ecological or social forms. If this is true, then we can speculate that, by working with artists, scientists will find themselves reframing their enquiries. For some scientists, it is this possibility that will continue to entice them into these most interesting of collaborations.

11. Seamon, D. & Zajonc, A. Goethe’s Way of Science: a Phenomenology of Nature (State Univ. New York Press, New York, 1998). 12. de Leval, M., Kilner, P. J., Gewillig, M. & Bull, K. Total cavo-pulmonary connection: a logical alternative to atriopulmonary connection for complex Fontan patients. J. Thorac. Cardiovasc. Surg. 96, 682–695 (1988). 13. Kilner, P. J. et al. Asymmetric redirection of flow through the heart. Nature 404, 759–761 (2000). 14. Kemp, M. Visualisations: The Nature Book of Art and Science (Univ. California Press, Los Angeles, 2000).

Stephen Webster is in the Science Communication Group, Humanities Programme, Imperial College London, London SW7 2AZ, UK. e-mail: [email protected]

FURTHER INFORMATION Arts & Humanities Research Council: http://www.ahrb.ac.uk/apply/research/arts_science_research_ fellowships.asp Engineering and Physical Sciences Research Council: http://gow.epsrc.ac.uk/ViewPanel.aspx?PanelID=3961& bannerlink=Panel%20Details Helen Storey Foundation: http://www.helenstoreyfoundation.org How To Live: http://www.bobbybakersdailylife.com Imperial College London Science Communication Group: www.imperial.ac.uk/sciencecommunication Institute of Contemporary Arts: http://www.ica.org.uk Mapping Perception: http://www.mappingperception.org.uk/ Medusae: http://www.sciart.org/site/medusae.html Midge Bait: http://www.sciart.org/site/midgebait.html National Institute for Medical Research: http://www.nimr.mrc.ac.uk/art/ On the Scent: http://www.placelessness.com/essences/on_the_scent.php Project Façade: http://www.projectfacade.com sciart: http://www.sciart.org; http://www.wellcome.ac.uk/en/ sciartprojects; http://www.wellcome.ac.uk/node2530.html The Arts Catalyst: http://www.artscatalyst.org The Fluent Heart: http://www.doc.ic.ac.uk/~gzy/heart/heart/index.htm Access to this interactive links box is free online.

doi:10.1038/nri1730 Published online 18 November 2005 1.

Kemp, M. & Wallace, M. Spectacular Bodies: the Art and Science of the Human Body from Leonardo to Now (Univ. California Press, Berkeley, 2000). 2. Gere, C. Thought in a vat: thinking through Annie Cattrell. Stud. Hist. Philos. Biol. Biomed. Sci. 35, 415–436 (2004). 3. Augenbraun, E. Weapon of mass attraction. Nature 433, 357–358 (2005). 4. Bodmer, W. J. D. Public Understanding of Science (J. D. Bernal lecture) (Birkbeck College, London, 1986). 5. Wolpert, L. Which side are you on? Observer (Lond.) (10 Mar 2002). 6. Suppe, F. (ed.) The Structure of Scientific Theories (Univ. Illinois Press, Urbana, 1974). 7. Kitcher, P. The Advancement of Science (Oxford Univ. Press, Oxford, 1995). 8. Editorial. Impugning conflict declarations. Nature Immunol. 5, 863 (2004). 9. Arts Council England. Arts and Science Research Fellowships scheme [online], (15 Oct 2004). 10. Mackrell, J. Pump it up John. Guardian (Lond.) (8 Sep 2005).

Acknowledgements I acknowledge W. Abbott, K. Arnold, P. Kilner and J. Thompson.

Competing interests statement The author declares no competing financial interests.

Online links

Concluding comments

Whereas Project Façade is an example of an artist supplying technical skills, The Fluent Heart shows that the arts can also make a conceptual contribution to the sciences. It is also evident that the contribution of an artist, or of an artistic training, can be to help a scientist to balance a reductionist approach with a wider, more inclusive gaze. In this regard, it is worth remembering the words of the art historian — and




PERSPECTIVES between the shape of the sculpted form and the dynamics of the water that produces the oscillating patterns of flow. It was his training as a sculptor, combined with a creative and interactive engagement with streaming water, that deepened Kilner’s appreciation of the relationships between form and flow. This interest soon led to specific questions about the internal dynamics of the heart. After working as a sculptor and teacher, Kilner re-entered medicine, winning a British Heart Foundation grant to collaborate with a heart surgeon. His project involved experiments with carefully shaped flow models, and it led to the modification12 of an operation that is carried out for patients with one of a group of unusual but severe congenital heart malformations. Wishing to study the living heart, Kilner then moved into the area of magnetic resonance imaging of the heart. His subsequent scientific papers, which include his drawings, are necessarily technical13. But to Kilner himself, it is clear that his years as an artist, working creatively with form and flow, complement his scientific training as the foundation of his work. How might this artistic training influence Kilner’s work in the field of cardiology? According to the scientist himself, his unusual background assists him to apply a set of skills to the heart that range from the reductionist to the holistic. His functional interpretation of the overall form of the heart depends crucially both on a reductionist understanding of anatomy and mechanics and on a more inclusive appreciation of the relationships between interconnected forms and multidirectional streams, for which his experience as a water sculptor equipped him. The dynamics of fluid flow in a heart are truly complex, and they resist easy reduction to hydrodynamic principles. Yet the role of the heart-imaging specialist is to understand quickly what is happening. According to Kilner, his background in sculpting fluid forms gives him an insight — an intuitive understanding — that is a useful complement to his scientific, analytical approach. Concluding comments

Whereas Project Façade is an example of an artist supplying technical skills, The Fluent Heart shows that the arts can also make a conceptual contribution to the sciences. It is also evident that the contribution of an artist, or of an artistic training, can be to help a scientist to balance a reductionist approach with a wider, more inclusive gaze. In this regard, it is worth remembering the words of the art historian — and

Nature columnist — Martin Kemp. He has described the artist’s style of visual insight as “a gate into the system”14, and he suggests that, although scientific enquiry is usefully carried out at many levels, the reductionist mode is often dominant. By contrast, artists are drawn less to nature at the molecular level and more to its bodily, ecological or social forms. If this is true, then we can speculate that, by working with artists, scientists will find themselves reframing their enquiries. For some scientists, it is this possibility that will continue to entice them into these most interesting of collaborations.

11. Seamon, D. & Zajonc, A. Goethe’s Way of Science: a Phenomenology of Nature (State Univ. New York Press, New York, 1998). 12. de Leval, M., Kilner, P. J., Gewillig, M. & Bull, K. Total cavo-pulmonary connection: a logical alternative to atriopulmonary connection for complex Fontan patients. J. Thorac. Cardiovasc. Surg. 96, 682–695 (1988). 13. Kilner, P. J. et al. Asymmetric redirection of flow through the heart. Nature 404, 759–761 (2000). 14. Kemp, M. Visualisations: The Nature Book of Art and Science (Univ. California Press, Los Angeles, 2000).

Stephen Webster is in the Science Communication Group, Humanities Programme, Imperial College London, London SW7 2AZ, UK. e-mail: [email protected]

FURTHER INFORMATION Arts & Humanities Research Council: http://www.ahrb.ac.uk/apply/research/arts_science_research_ fellowships.asp Engineering and Physical Sciences Research Council: http://gow.epsrc.ac.uk/ViewPanel.aspx?PanelID=3961& bannerlink=Panel%20Details Helen Storey Foundation: http://www.helenstoreyfoundation.org How To Live: http://www.bobbybakersdailylife.com Imperial College London Science Communication Group: www.imperial.ac.uk/sciencecommunication Institute of Contemporary Arts: http://www.ica.org.uk Mapping Perception: http://www.mappingperception.org.uk/ Medusae: http://www.sciart.org/site/medusae.html Midge Bait: http://www.sciart.org/site/midgebait.html National Institute for Medical Research: http://www.nimr.mrc.ac.uk/art/ On the Scent: http://www.placelessness.com/essences/on_the_scent.php Project Façade: http://www.projectfacade.com sciart: http://www.sciart.org; http://www.wellcome.ac.uk/en/ sciartprojects; http://www.wellcome.ac.uk/node2530.html The Arts Catalyst: http://www.artscatalyst.org The Fluent Heart: http://www.doc.ic.ac.uk/~gzy/heart/heart/index.htm Access to this interactive links box is free online.

doi:10.1038/nri1730 Published online 18 November 2005 1.

Kemp, M. & Wallace, M. Spectacular Bodies: the Art and Science of the Human Body from Leonardo to Now (Univ. California Press, Berkeley, 2000). 2. Gere, C. Thought in a vat: thinking through Annie Cattrell. Stud. Hist. Philos. Biol. Biomed. Sci. 35, 415–436 (2004). 3. Augenbraun, E. Weapon of mass attraction. Nature 433, 357–358 (2005). 4. Bodmer, W. J. D. Public Understanding of Science (J. D. Bernal lecture) (Birkbeck College, London, 1986). 5. Wolpert, L. Which side are you on? Observer (Lond.) (10 Mar 2002). 6. Suppe, F. (ed.) The Structure of Scientific Theories (Univ. Illinois Press, Urbana, 1974). 7. Kitcher, P. The Advancement of Science (Oxford Univ. Press, Oxford, 1995). 8. Editorial. Impugning conflict declarations. Nature Immunol. 5, 863 (2004). 9. Arts Council England. Arts and Science Research Fellowships scheme [online], (15 Oct 2004). 10. Mackrell, J. Pump it up John. Guardian (Lond.) (8 Sep 2005).

Acknowledgements I acknowledge W. Abbott, K. Arnold, P. Kilner and J. Thompson.

Competing interests statement The author declares no competing financial interests.

Online links

ERRATUM

NK CELLS IN HIV INFECTION: PARADIGM FOR PROTECTION OR TARGETS FOR AMBUSH Anthony S. Fauci, Domenico Mavilio and Shyam Kottilil Nature Reviews Immunology 5, 835–843 (2005)

When published, the thirteenth row in Table 1 contained incorrect information. The corrected row (together with the column headings) is shown below.

Receptor

Function

Ligand specificity

Effect of HIV viraemia on NK-cell-receptor expression

Effect of ART on NK-cellreceptor expression

Refs

NKp46 (NCR1)

Activating

Influenza-virus haemagglutinin, others unknown?

Decrease

Restoration to normal levels

6,36, 62,64




No title

No title

No title

No title

No title

No title

No title

No title

No title

No title

No title

No title

No title

No title

No title

No title

No title

No title

No title

No title

No title

No title

No title

No title

No title

No title

Recommend Documents