Insights into immune structure, recognition, and signaling


  • K. Christopher Garcia

    Corresponding author
    • Howard Hughes Medical Institute, Departments of Molecular and Cellular Physiology, and Structural Biology, Stanford University School of Medicine, Stanford, CA, USA
    Search for more papers by this author

Correspondence to:

K. Christopher Garcia

Howard Hughes Medical Institute, Departments of Molecular and Cellular Physiology, and Structural Biology, Stanford University School of Medicine

Beckman B171B

279 Campus Drive, Stanford, CA 94305-5345, USA

Tel.: +1 650 498 7332

Fax: +1 650 725 8021


This article introduces a series of reviews covering Insights from Structure appearing in Volume 250 of Immunological Reviews.

In this volume of Immunological Reviews, the articles focus on the structural biology of immunoreceptors and what structure is telling us about function. The issue is unabashedly ‘receptor-centric’ and the proteins covered range from antigen receptors, innate receptors, and cytokine receptors, thereby providing an unusually comprehensive snapshot of the current state of the art in molecular immunology. Of all the major fields in biomedical research, including neurobiology, developmental biology, immunology, microbiology, and genetics, just to name a subset, perhaps structure in the broadest sense has had more of an impact on immunology and played more of a major role in guiding the evolution and maturation of the field than any other. Here, I use the term ‘structure’ to encompass not only direct structural approaches, such as x-ray crystallography and nuclear magnetic resonance spectroscopy (NMR), but also imaging, biophysical, molecular, and mechanistic approaches to interrogate aspects of molecular recognition, oligomerization, and signaling. Perhaps the fact that immunobiology is centered on questions of recognition and discrimination and is largely controlled by receptor-ligand systems has led to an unusually high level of sophistication about structure among immunologists. Personally speaking, having delved into structural questions in neurobiology and developmental biology, I can state without equivocation that the structural and molecular savvy of immunologists has led to greater mechanistic clarity underlying central tenets of the field compared to other disciplines. Dating back to the days of elucidating the first antibody structures and the relation between hypervariable regions and antigen recognition, to the transformative clarity of the first major histocompatibility complex (MHC) structures in explaining MHC restriction, to more recent biophysical and structural investigations of T-cell receptor (TCR) recognition of both classical and non-classical MHC, innate immune receptors such as natural killer (NK) and Toll-like receptors, and cytokine receptors, molecular science has illuminated the path for the experimentalists and assumed a large and highly visible role in immunobiology as a whole.

The pace of progress in structural and molecular immunology has increased dramatically in the last 5–7 years to the point that the literature is so vast and rapidly evolving that it is easy to lose track of the remaining important questions. The acceleration of progress is in large part due to structural immunologists’ increasing expertise at expressing difficult proteins and biochemically reconstituting protein complexes and signaling systems. Ten years ago, expressing and crystallizing a single TCR was a major challenge for any laboratory, but enough different methods have been reported for many different immunoreceptors that we now have an array of strategies to produce and crystallize complicated, multi-chain, glycosylated molecules, and low-affinity complexes: generally one or more strategy will succeed although it is not at all formulaic. There has also been rapid progress in biophysical methods associated with structure determination and imaging in tandem with the protein biochemistry. Most importantly, my opinion is that investigators simply believe that it is possible and feasible to attempt highly ambitious molecular and biophysical studies. This psychological aspect cannot be underestimated in a field where structural breakthroughs are akin to climbing mountains: one has to believe to have a chance. As a result, astonishing and beautiful science is described in this issue that in the not too distant past would have been considered nearly impossible.

This volume aims to provide a resource for the community that presents the current state of the field of structural and molecular immunology in several major areas together in one issue. However, not every system or important structural contribution could be included in this volume. The issue has a heavy leaning toward articles in TCR structure, recognition, and signaling [1-8]. There are many outstanding groups working in this general area, and reading different ‘spins’ on related topics illustrates the point that while this field is mature, there remain many unanswered questions and a lack of consensus on important issues. Collectively, these articles cover not only pure structural issues but also link the structures to our understanding of how signaling occurs in tandem with recognition. For example, Brian Baker and colleagues [1] highlight the generally underappreciated importance of conformational dynamics and ‘melding’ in TCR/peptide-MHC interactions, and they certainly make a strong case that this is a critical aspect of antigen recognition that is not always evident from crystal structures, but rather requires a range of approaches to fully capture. Mariuzza and colleagues [3] address the role of structure in autoimmunity, where it is becoming abundantly clear that structural aberrations in the manner of TCR recognition of self- and auto-antigens is, in fact, a recurring feature found in crystal structures of such complexes. This insight begins to provide a molecular template for intervention strategies directed at TCR/pMHC interactions underlying autoimmune disease. Marrack and colleagues [4] discuss the long-debated issue of TCR germline bias for MHC in the context of compelling evidence that the TCR Vβ8 appears to form conserved interactions with MHC molecules that likely arose through coevolution. Given the composite nature of the peptide-MHC surface, clear delineation of MHC versus peptide-directed evolution of the TCR repertoire remains an extremely difficult challenge to resolve experimentally, but structures of TCR/pMHC complexes, together with elegant functional studies, are beginning to lead us to a more granular idea of what is going on, and no doubt work will continue along this line. Rossjohn and colleagues [6] access a tremendous wealth of structural information accrued over the years on TCR recognition of HLA-peptide complexes to attempt to glean conclusive generalizations about HLA flexibility, TCR bias, TCR polymorphism, and self-tolerance. The upshot of this analysis is that there are ‘many ways to skin a cat’ and that earlier concepts of what represents conventional T-cell recognition have been overturned by the astonishing revelations of structural diversity uncovered in the series of studies reported by the Rossjohn and McCluskey groups. As one example, the visualization of the ‘bulged’ peptides recognized by a TCR does force one to reconsider how germline recognition of MHC can be achieved when the TCR is literally ‘held away’ from the MHC by the protruding peptide and forms only scant germline contacts. My own contribution to this issue principally written by Michael Birnbaum [2] focuses on how approaches using molecular diversity, in the broadest sense, have contributed to our understanding of TCR cross-reactivity and how recognition is linked to signaling. TCR cross-reactivity remains a divisive issue; on one hand, some think TCRs are highly degenerate, whereas on the other, T-cell recognition appears highly sensitive to mutation of the presented peptide. We summarize results from peptide-MHC combinatorial libraries showing that T-cell signaling can be modulated to a shocking extent by sequences highly divergent from the cognate antigens and that an individual TCR has the capacity to engage entirely unrelated peptide sequences in both agonistic and non-stimulatory docking modes. As the sequences we analyzed are not naturally occurring, the question of whether these observations are strictly an artificial in vitro phenomenon or whether this is relevant to in vivo T-cell biology remains unresolved.

Several additional articles take us away from the pure aspects of TCR/pMHC recognition into understanding the biophysics of TCR signaling mechanisms. For example, Reinherz and Wang [7] explain their mechanotransduction concept of TCR signaling in the context of a model of the entire TCR-CD3 signaling complex on the T-cell surface. Clearly, evidence from this group implicates shear force and the vector of approach guiding the TCR/pMHC binding event as being important factors in triggering signaling and provides a rationale for how engagement by the non-signaling αβ TCR could be relayed to the ‘side-on’-associated CD3 subunits. Many disparate observations about the role of co-receptors in TCR signaling can also be incorporated into this model in a very satisfying way, but much more work will be required to refine and validate this idea. The reality is that we currently only have a very fuzzy idea for the ultrastructure of the TCR-CD3 complex, and Kuhns and colleagues [5] provide an outstanding review of the current evidence underlying structural models of the ectodomain associations. Although we know that the TCR and CD3 transmembrane domains are closely associated, the role of the TCR and CD3 ectodomains in signaling remains mysterious. Nevertheless, several elegant studies have approached this problem through a variety of methodologies and generated enough constraints that topological models of the TCR-CD3 complex can be constructed and serve as a strawman to confirm or refute. Collectively, structural studies of TCR and peptide-MHC interactions are being increasingly focused on providing an answer to the problem of how antigen recognition is structurally coupled to membrane proximal signaling events. Imaging, either through electron microscopy or X-ray crystallography, the entire TCR-CD3 transmembrane complex is ultimately where the field is headed, but overcoming this technical challenge would qualify as one of the greatest feats in structural biology. This is likely not going to happen anytime soon. But without question, many investigators, including those whose opinions are represented in this volume, are hard at work on this problem, and I only hope to be alive and well when it finally happens.

As it takes two to tango, this issue also presents three articles on the other side of the coin: the peptide-MHC molecule. Structures of peptide-MHC appeared earlier than TCR and TCR/pMHC complexes, so our understanding of the molecular properties of MHC is at a highly advanced stage. However, several important issues remain unsolved frontiers in the areas of antigen presentation. One question that has so far eluded a clear structural answer is how MHC molecules are loaded with peptides. Stern and colleagues [9] demonstrate the power of accessing a large database of class II MHC structures to show a recurrent region of flexibility in the 310 helix of the α subunit bordering the peptide-binding groove. This region as well as one other in the β subunit are also implicated by biophysical and mutational studies as being important in the highly complex, transient, and dynamic peptide-loading process by DM. The authors make a very strong case that this region of flexibility is likely the ‘smoking gun’ regulating accessibility to the empty MHC groove by the loading machinery, and one gets the feeling that structural proof of this conjecture is soon to follow. Ostrov and colleagues [10] discuss a fascinating observation that small molecule drugs bind within the MHC cleft and alter the repertoire of peptides presented by HLA. This has been confirmed by two different crystallographic studies showing the small molecule actually bound within the MHC groove, and these structures explain how allelic polymorphisms within the MHC can affect the efficacy of these drugs. These are very important studies because examples of small molecules that impact protein–protein interactions are rare, and this phenomenon suggests that ‘drugging’ MHC proteins is feasible and could lead to antigen-specific therapeutic approaches. Finally, speaking of non-peptide molecules presented by MHC, Zajonc and colleagues [11] provide an important treatise on structural aspects of the non-classical MHC CD1 and its ability to present glycolipid antigens to the semi-invariant αβ TCR of natural killer T (NKT) cells. The discovery of a class of non-peptide presenting MHC-like molecules and subsequent structural elucidation of how CD1 presents lipid antigens to TCRs has been one of the most exciting additions to the field of molecular immunology in recent years. Lipids present great technical difficulties to structural biologists, and advances in this space have not been easy and credit should go to those hardy souls who have braved the challenges and soldiered on. The mode of antigen presentation by CD1 and TCR engagement bears some general similarities to classical TCR/peptide-MHC interaction, but on closer inspection reveals an abundance of exciting and thought provoking structural differences that are central to the unique biology of this system, also providing us with some unique insights into the role of TCR germline bias in recognition of non-classical MHC.

TCRs represent only two of the three lineages (αβ and γδ) of our antigen receptors, with antibodies being the third. Antibody structural biology is at a very advanced stage, with literally hundreds of antibody-antigen complex structures in the PDB, which raises the question, ‘What new insights could there possibly be in antibody recognition?’ The most exciting work in this space over the last five or so years has been in vaccine design. Antibody complexes with viral antigens have been immensely informative in guiding innovative strategies to trick the immune system to generate neutralizing antibodies to highly mutable pathogens. Crystal structures of neutralizing antibody complexes with HIV gp120 and influenza HA have unlocked the secret for how to effectively neutralize these viruses in their most vulnerable places, and informed protein engineering efforts to create vaccines to elicit such antibodies. In this issue, Wilson and colleagues [12] tell us some highlights of this work that is taking us tantalizingly close to the dream of effective vaccines for AIDS and flu. Along the theme of viral evasion of immune responses, Fremont and colleagues [13] discuss a myriad of fascinating ways that viruses engage endogenous immune regulatory proteins using both direct and indirect structural mimicry of cytokines and cytokine receptors. Structural insights from studies of such ‘decoys’ have been absolutely stunning in showing us how viruses repurpose host proteins for subversion and survival using both agonist and antagonist modes of action.

This volume would not be complete without inclusion of structural insights we have gained from innate immune systems. Jie-Oh Lee and colleagues [14] summarize the stunning portfolio of extracellular Toll-like receptor complexes as well as the recent death domain complex containing myeloid differentiation factor 88 (MyD88). While the details of antigen engagement in these complexes are highly diverse, in keeping with the wide range of ligands encountered (e.g. lipids, RNA, proteins), the oligomeric structures of the Toll-like receptor complexes bear similarities that suggest a conserved overall signaling architecture. While these receptors have captured the lion's share of attention recently in the innate receptor field, Sun and colleagues [15] remind us that Toll-like receptors are not the only game in town and that pentraxins represent a powerful and effective branch of the innate immune system that is mediated in a highly novel manner. Pentraxins recognize microbial adducts, like lipids or sugars, and also directly engage Fc receptors, resulting in an opsonization complex by which macrophages destroy invading pathogens. Sun et al. [15] demonstrate that this recognition is optimized through multi-valency and also bears some similarities to Fc receptor interactions with antibodies – in effect bridging the innate and adaptive arms of immunity. Strong and colleagues [16] highlight structural insights into how NK receptors, whose effector functions are governed by balancing opposing signals from inhibitory and activating receptors, balance the twin tasks of cross-reactive recognition of multiple MHC ligands with specific engagement of pathogen-associated ligands. Structural studies have played a major role in delineating the highly complex energetic mechanisms by which NK receptors discriminate their ligands, and new twists continue to develop in this area. Perhaps structural studies of NK receptor complexes have told us more about the incredible diversity of receptor engagement modes that the MHC fold is capable of than any other. Nearly the entire surface of the MHC is accessed by one type of NK receptor, or TCR, in sometimes convergent yet other times divergent modes. Finally, van der Merwe and colleagues [8] provide a very interesting perspective on a large class of important immune receptors termed non-catalytic tyrosine-phosphorylated receptors (NTRs). Grouped in the NTR category are immune receptors that engage extrinsic Src family tyrosine kinases but do not themselves possess kinase activity on their polypeptide chains. Many families of receptors fall under this category, including KIRs, SLAMs, NKG2s, and others. Van der Merwe et al. argue that despite vastly different structural features within the NTR grouping, by virtue of their small ECD size they are able to share a common kinetic segregation signaling mechanism whereby NTRs are shielded from the tyrosine phosphatases with large ectodomains, and this enables NTRs to rapidly evolve and adapt to a changing ‘foreign’ antigen environment.

This volume is not exclusively biased to antigen recognition. Cytokines work in tandem with immunoreceptors to control immune homeostasis, so this issue contains two timely articles on two of our shared immunoregulatory cytokine receptors, common γ chain [17] and common β chain [18], as well as the more innate type I interferons (IFNs) [19]. Parker and colleagues [18] discuss the unusual and unexpected architecture of the common β chain signaling complex and its relevance to the granulocyte-macrophage colony-stimulating factor (GM-CSF)/interleukin-3 (IL-3)/IL-5 family as a whole. The structure of the GM-CSF ectodomain complex beautifully rationalizes how signaling can be achieved through a ‘dodecameric’ oligomerization results in a dimeric common β chain assembly. Walsh and colleagues [17] take us on a tour of the common γ chain structural literature, looking at it through the eyes of the highly informative IL-7 receptor complex. There does appear to be some semblance of structural rules governing recognition of common γ chain-containing cytokines, and Walsh and colleagues [17] also discuss structural evidence for pre-oligomerization of IL-7Ra and its implications for signal regulation. Schreiber, Piehler and colleagues [19] present a thought-provoking article on the mechanism by which Type I IFN receptors differentially respond to multiple IFN ligands. This highly complex issue, illuminated recently by the crystal structure of the first Type I IFN receptor complex, appears to boil down to the important role of extracellular receptor-ligand complex stability in regulating cell surface dynamics, internalization rates, and ultimately signal strength and quality. Here, it is very clear that the chemistry of IFN recognition serves as arbiter of signaling and highlights the important role of structure in clarifying these questions.

I wish to thank all of the contributors to this issue for their diligence and hard work in crafting their articles, which are uniformly thought provoking and comprehensive. The result is an impressive treatise that I suggest is perhaps the current best resource for anyone in the field of immunity and infection to refresh themselves on the state of structure and molecular science in several of the principal macromolecular systems controlling the immune system.