<<

news and views

are found in chimeras in the to give rise to Treg cells increases but never TCR–transgenic mice. The higher frequency range of 0.01–1% of Treg cell TCR-transgenic reaches 100%. At some point, clonal deletion of precursor cells expressing a particular TCR precursors, it seems more likely that TCR com- begins to cull self-reactive TCRs from the clone in these mice masks normal activ- petition is the factor that limits Treg cell devel- repertoire. Between the Treg cell–propensity ity during development. Another limitation opment in this system. If Treg cell development threshold and the clonal-deletion threshold, it of TCR-transgenic models is that they tend were to require access to rare complexes of self is possible for TCRs to appear in both Treg cell to allow earlier and stronger TCR expres- peptide and major histocompatibility complex and naive T cell repertoires, as is observed in sion. If Treg cell development is dependent (MHC), then decreasing the frequency of pro- actual experiments. In addition, with TCRs of on the detection of subtle shifts in TCR affin- genitor cells with that TCR specificity could higher affinity, IL-2 signaling could raise the ity and/or interaction with unique - radically enhance the probability of encoun- effective threshold for clonal deletion, creating presenting cells in a particular environment, ter with rare self peptide–MHC complexes. a window in which a subset of TCRs is found TCR-transgenic mice (whether conventional Further experiments will be needed to sort exclusively in the Treg cell repertoire. TCRs or retrogenic) may not model development this out. in the low to middle regions of the affinity appropriately. However, the need to study Many indirect results suggest that the Treg spectrum (such as the receptors cloned by the activity of cells with fixed TCR specific- cell repertoire consists of self-reactive TCRs. Bautista et al.) would give rise to Treg cells ity remains as crucial as ever. Thus, the field Given that conclusion, it might be predicted with varying frequency. In contrast, TCRs in continues to be in need of more-refined TCR- that mice transgenic for Treg cell–derived the high end of the affinity spectrum (such as transgenic models. TCRs would show clonal deletion in the thy- the receptor cloned before5) would be overtly Overall, this new research highlights the mus. Unexpectedly, Bautista et al. detect no self reactive and thus would be deleted. It will ‘instructive’ nature of the TCR in Treg cell evidence of clonal deletion in these mice1. On be useful to determine if such high-affinity development but introduces a notable proba- a RAG-deficient background, this TCR trans- receptors give rise to Treg cells when present bilistic element. These seemingly ‘fuzzy’ rules + gene gives rise to large populations of CD4 at low frequencies in mixed–bone marrow result in a Treg cell repertoire that, like the rep- T cells in the and the periphery. In chimeras. The fact that TCRs from different ertoire of naive T cells, is diverse and highly contrast, another group has reported efficient ‘affinity windows’ have been cloned in dif- adaptable and continually responds to self and clonal deletion of CD4+ T cells and almost ferent studies might have to do with how the environmental cues about where and how to + no detectable Foxp3 Treg cells in a distinct Treg cells were isolated. In the earlier study, the function. These cues are at least as complicated line of mice transgenic for Treg cell–derived researchers cloned their Treg cell–derived TCR and fascinating as the Treg cell development TCRs on a RAG-deficient background, which from an αβTCR-transgenic strain in which process itself. + indicates that Treg cell–derived TCRs are self- the appearance of Foxp3 Treg cells depends 5 1. Bautista, J. et al. Nat. Immunol. 10, 610–617 reactive. It is possible to imagine a model that on endogenous TCRα-chain rearrangement . (2009). can accommodate both results, in which both Thus, the Treg cell receptor is a ‘second recep- 2. Hsieh, C.S., Zheng, Y., Liang, Y., Fontenot, J.D. & subtly and overtly self-reactive T cells can tor’ on these cells. In this situation, less surface Rudensky, A.Y. Nat. Immunol. 7, 401–410 (2006). 3. Pacholczyk, R. et al. 27, 493–504 give rise to Treg cells (Fig. 1). In this model, a expression of the TCR could have facilitated (2007). certain low threshold of TCR affinity for self the ‘rescue’ of these cells from deletion. 4. Burchill, M.A. et al. Immunity 28, 112–121 (2008). peptide–MHC is required for positive selec- The results of Bautista et al. highlight a 5. DiPaolo, R.J. & Shevach, E.M. Eur. J. Immunol. 39, tion. With increasing affinity, the propensity chief limitation of the use of monoclonal 234–240 (2009). © All rights reserved. 2009 Inc. Nature America, Antigen-processing and presentation pathways select antigenic HIV peptides in the fight against viral evolution

David H Margulies

The evolution of immunodominant in HIV-1 Gag proteins correlates with quantitative measures of several antigen processing events. Thus, peptides recognized by CD8+ cytolytic T cells are selected by their ability to pass through the antigen processing pathway, as well as by their binding to HLA molecules.

he vertebrate uses dif- natural killer cells, B (and their tem like T cells and B cells repeatedly iden- Tferent ‘arms’ that act together to combat protein derivatives, ), T cells and tify specific ; these antigens, called viral infection. The cells and receptors of specialized antigen-presenting cells work ‘immunodominant antigens’, have become innate and adaptive immunity, including together to identify before the viruses one focus of efforts to develop better vac- can even adsorb to cells, to mark infected cination strategies1. In this issue of Nature David H. Margulies is in the Molecular Biology cells for destruction and/or to limit reactiva- , Tenzer et al.2 identify param- Section, Laboratory of Immunology, National tion of persistently or latently infected cells. eters that influence the generation and Institute of and Infectious Diseases, National Each of these activities waxes and wanes presentation of immunodominant human Institutes of Health, Bethesda, Maryland, USA. throughout the course of viral exposure and type 1 (HIV-1) pep- e-mail: [email protected] infection. Adaptive cells of the immune sys- tides to cytolytic CD8+ T lymphocytes.

566 volume 10 number 6 june 2009 nature immunology news and views

Resistance to chronic viral infections such as HIV can be correlated to some degree with Cytosol ongoing activity of cytolytic CD8+ T lympho- cytes that express αβ T cell antigen recep- (4) TAP tors that recognize complexes of peptide and major histocompatibility complex (MHC) ER class I in which the peptides are derived from viral proteins. A major challenge in under- (1) Protein standing the variety of immune responses (2) Proteasomal to viral infection has been to identify the degradation (3) Intermediate parameters, both molecular and cellular, that peptides determine which peptides from which viral proteins are most commonly selected for pre- sentation at the cell surface and to understand + how those peptides are generated and loaded ERp57 CD8 T cell onto MHC class I molecules. However, the scientific challenge is not only to understand the rules that govern the generation of immu- TCRαβ CD8 nodominant peptides but also to elucidate the (5) Peptide- ERAAP options available to a rapidly evolving virus loading able to modify the protein from which such complex peptides derive. Tenzer et al. build on previ- ER (6) pMHCI ous work showing that wild-type sequences encoding an HIV-1 group-associated antigen Cytosol (Gag) T cell undergo natural escape Cell surface mutations that then can ultimately revert back to the wild-type sequence3. Here, the Figure 1 Steps in antigen processing and presentation in the MHC class I pathway. Native, partially authors study the wild-type sequence of the unfolded, or denatured protein in the cytosol (1) is further denatured concomitant with proteolytic Gag protein p17 as well as seven common processing by constitutive or immunoproteasomes (2). In experiments reported in this issue2, 25– natural variants of this sequence to evaluate amino acid peptides are used as substrates for isolated . Intermediate peptides that result the influence of several crucial parameters from degradation (3) then bind TAP (4) and are transported to the peptide-loading complex of the MHC class I antigen-processing and (5), where the sequential and concerted effects of tapasin, ERp57, calreticulin, MHC class I antigen-presentation pathway on the detec- heavy chain, β2-microglobulin and ERAAP1,2 lead to the formation of a stable MHC class I–β2- tion of each variant peptide by CD8+ T cells. microglobulin–peptide complex (pMHCI; 6). This complex is transported via the Golgi to the cell surface, where it is available for binding to αβ T cell antigen receptor (αβTCR) and CD8αβ molecules For the sake of completeness, they also ana- on the surface of CD8+ T cells. ER, . lyze epitopes from the p24 Gag protein, but the discussion here is limited to the results

© All rights reserved. 2009 Inc. Nature America, obtained with p17. Using synthetic 25–amino acid peptides the intermediate peptides for TAP. As TAP- Underlying their approach is the view that with substitutions representing the sequences mediated transport is dependent on TAP the final product of this complex antigen-pro- commonly found in escape variants, the affinity, peptides with a higher affinity for cessing and antigen-presentation pathway— authors first evaluate the relative abundance TAP are expected to be proportionately bet- that is, the cell surface peptide–MHC class I of smaller, or intermediate, peptides gener- ter represented in the pool of intermediates complex that is subsequently recognized by ated by the proteolytic activity of purified that reach the peptide-loading complex in CD8+ T cells—is generated as a result of the proteasome preparations4. These peptides of the endoplasmic reticulum5. After that, they quantitative efficiency of each step of the path- intermediate length are potentially able to measure the relative efficiency with which way and that these efficiencies can be exam- be transported by TAP into the endoplasmic ERAAP1,2, an endoplasmic reticulum–lu- ined by reconstitution of the relevant steps in reticulum lumen and to bind by their car- menal enzyme, trims each intermediate pep- vitro. The main steps of the pathway include boxyl termini to endoplasmic reticulum–res- tide. Finally, they determine the affinity of cytosolic production of peptide precursors, ident MHC class I molecules. Using fragment the presenting HLA molecule for each of the transport of intermediate peptides from the intensity identified by mass spectrometry as trimmed peptides. They find that cleavage cytosol to the peptide-loading complex of the an indicator of the abundance of the inter- ‘preference’, TAP binding, ERAAP trimming endoplasmic reticulum by the transporter mediate peptides produced by proteasomal and HLA affinity all collectively contribute associated with antigen processing (TAP), digestion, the authors find that the wild-type to the relative hierarchy of T cell recognition and maturation and folding of peptide–MHC 25–amino acid peptide generates the pre- of the resulting peptide antigen. class I complexes expressed on the cell sur- cursor for the immunodominant peptide In general, the overall hierarchy of T cell face (Fig. 1). Studying all steps of the pathway presented by HLA-A2 (SLYNTVATL) rather recognition of p17 and p24 epitopes corre- completely for any given virus (or protein, for poorly. In addition, they note that different lates with a calculated function of epitope that matter) would be a next-to-impossible variant 25–amino acid peptide precursors ‘abundance’ rather than with peptide affinity task, so the authors confine themselves to the give rise to distinct peptides of intermedi- for HLA molecules. In addition, the detec- merely overwhelming job of studying several ate length. Next they measure the affinity tion of differences in the responses of HIV steps of the pathway. of synthetic peptides that represent each of patients to peptide variants with a carbox-

nature immunology volume 10 number 6 june 2009 567 news and views

yl-terminal extension of the core epitope is not completely unexpected: the spectrum ing virus that can tolerate sequence variation prompted structural analyses of this elon- of peptides generated from a rather short in an essential protein for many generations gated peptide bound to HLA-A2. The crys- 25–amino acid peptide precursor after pro- sufficient to facilitate evasion of an immune tal structure shows that this elongated p17 teasomal processing is very large, and this response and then ‘bounce back’ by reverting core peptide may bind to HLA-A2 in a way spectrum can be widely influenced by one to the more ‘fit’ wild-type sequence. Thus, that results in the adoption of an unusual to four substitutions in the sequence of the not unlike the solutions to global warming ‘flipped’ orientation by its carboxy-terminal peptide precursor. The finding that differ- or dependence on foreign oil, the proper tyrosine residue. Thus, although HLA affin- ent peptide intermediates are transported immunization for rapidly evolving viruses ity certainly contributes to the hierarchy of with different efficiency is also predictable, cannot be a single T cell vaccine but instead CD8+ T cell recognition, it is not the sole and the finding that different sequences must combine many different modali- predictor of the cytotoxic T are trimmed differently by ERAAP1,2 is ties that, like the ‘cooperative arms’ of the response. Peptide abundance and structure expected6. However, for the second ques- immune system, work in a complementary when bound to HLA molecules probably also tion, the overall analysis emphasizes the way to counter several steps of a pathogen’s influence cytotoxic T lymphocyte recogni- naiveté of believing that researchers could life cycle. tion. look at the sequence of a single wild-type Does the enormous amount of data pro- protein, predict—using bioinformatics—its 1. Yewdell, J.W. & Bennink, J.R. Annu. Rev. Immunol. 17, 2 51–88 (1999). vided by the analysis by Tenzer et al. provide capacity to bind one or several common HLA 2. Tenzer, S. et al. Nat. Immunol. 10, 636–646 (2009). a better sense of the parameters that might molecules, and know with assurance exactly 3. Iversen, A.K. et al. Nat. Immunol. 7, 179–189 which protein sequence to incorporate into (2006). control immunodominance? If so, can a sen- 4. Kloetzel, P.M. Nat. Rev. Mol. Cell Biol. 2, 179–187 sible theoretical course that might guide the a favorite immunogenic vector. (2001). empirical approach to vaccine development Of course, these findings do not provide 5. Abele, R. & Tampe, R. Physiology (Bethesda) 19, 216– 224 (2004). be outlined? In terms of the first question, a simple guideline for choosing the ‘right’ 6. Kanaseki, T., Blanchard, N., Hammer, G.E., Gonzalez, one general take-home message of this study peptide every time; HIV-1 is a rapidly mutat- F. & Shastri, N. Immunity 25, 795–806 (2006).

IL-17A directly inhibits TH1 cells and thereby suppresses development of intestinal

Amit Awasthi & Vijay K Kuchroo

T helper type 1 cells (TH1 cells) serve a dominant function in T cell–mediated colitis. New work reports that interleukin 17A, an effector required for the development of autoimmune tissue inflammation, directly inhibits TH1 development by suppressing the expression of key TH1-associated genes and therefore regulates TH1

© All rights reserved. 2009 Inc. Nature America, cell–mediated colitis.

egulation of intestinal immunity is addition to IL-10 and transforming growth autoimmune disease models such as those Rcritical for maintaining gut immune factor-β, IL-17A is crucial in directly inhib- that mimic multiple sclerosis (for example, homeostasis, as the gut encounters and is iting the TH1 subset and suppressing gut experimental autoimmune encephalomyeli- exposed to a plethora of environmental and inflammation3. tis in mice) and rheumatoid arthritis, IFN-γ food antigens that can lead to severe immune IL-10 is essential for the immunosuppres- is dispensable for the development of auto- reactions. Several regulatory mechanisms sive effects of Foxp3+ regulatory T cells and immune disease and tissue inflammation. have evolved to prevent these inflamma- type 1 regulatory T cells in colitis. IL-10- In these models, IFN-γ-deficient and IFN-γ tory reactions; both Foxp3+ regulatory deficient mice develop spontaneous colitis receptor–deficient mice are not resistant to 4 T cells and type 1 regulatory T cells, which with more TH1 cells in the inflamed gut . and in fact are more susceptible to autoim- produce IL-10, serve a dominant function Moreover, IL-10-deficient Foxp3+ regula- munity and tissue inflammation7. Such find- in maintaining effector T cell responses1. In tory T cells do not protect against colitis ings have raised the possibility that another 5 addition, T helper type 3 cells (TH3 cells), triggered by gut flora . TH1 cells that pro- subset of T cells and other , distinct which produce mainly transforming growth duce interferon-γ (IFN-γ) are important from TH1 cells and IFN-γ, are responsible for factor-β, mediate tolerance to orally admin- effector cells in inducing colitis and wast- inducing autoimmune tissue inflammation istered antigens2. In this issue of Nature ing disease, as blocking IFN-γ at the time in these disease models. Immunology, O’Connor et al. show that in of T cell transfer prevents development of Another subset of helper T cells, called colitis. Consistent with the idea that IFN-γ ‘TH-17’ cells, that produce a panel of cyto­ Amit Awasthi and Vijay K. Kuchroo are with the and TH1 cells are the essential cytokine and kines (IL-17A, IL-17F, IL-21 and IL-22) and Center for Neurologic Diseases, Brigham and effector cells for the induction of colitis in are distinct from TH1 or TH2 cells, have been Women’s Hospital, Harvard Medical School, immunodeficient animals, adoptive trans- linked to the induction of autoimmune dis- Boston, Massachusetts, USA. fer of T cells from IFN-γ-deficient mice eases in mice and humans7. In support of 6 e-mail: [email protected] fails to induce disease . In contrast, in other the idea of involvement of IL-17 and TH-17

568 volume 10 number 6 june 2009 nature immunology