Abstract
γδ T cells express adaptive antigen receptors encoded by rearranging genes. Their diversity is highest in the small region of TCR V–J junctions, especially in the δ chain, which should enable the γδ TCRs to distinguish differences in small epitopes. Indeed, recognition of small molecules, and of an epitope on a larger protein has been reported. Responses to small non-peptides known as phospho-antigens are multi-clonal yet limited to a single γδ T cell subset in humans and non-human primates. Responses to small peptides are multi-clonal or oligo-clonal, include more than one subset of γδ T cells, and occur in rodents and primates. However, less effort has been devoted to investigate the peptide responses. To settle the questions of whether peptides can be ligands for the γδ TCRs, and whether responses to small peptides might occur normally, peptide binding will have to be demonstrated, and natural peptide ligands identified.
Keywords: γδ T cells, T cell receptor, Peptide, Ligand, Antigen recognition
Introduction
The adaptive immune system of jawed vertebrates includes lymphocytes expressing B cell receptors (BCR), αβ T cell receptors (αβ TCR) and γδ T cell receptors (γδ TCR). All of these receptors are encoded by rearranging genes, which combine variable and constant gene segments to generate diverse receptor molecules. The mechanisms that come into play in building this diversity, which is often random and can be nearly unlimited, have been extensively reviewed elsewhere [1]. It is thought that random receptor diversity is the evolutionary answer to a need to recognize a myriad of antigens expressed by the countless and rapidly changing pathogens that threaten vertebrate survival. The lymphocytes expressing BCRs (B cells) and αβ TCRs (αβ T cells) have been much studied, and they are considered key players in the adaptive immune responses. Thus, B cells produce antibodies, which opsonize and neutralize pathogens and their antigens, and αβ T cells either help B cells to develop into antibody-producing plasma cells, or they target infected or otherwise injured cells directly, and eliminate them by virtue of their cytolytic activity. In contrast, it is less well known what γδ T cells do, and whether they use their adaptive receptors to support adaptive immune responses or perhaps other functions of the immune system. Most investigators in the field believe, however, that knowing the ligands recognized by the γδ TCRs will also provide answers to the current questions about the role of γδ T cells in immune responses.
Lack of a paradigm in γδ ligand recognition
Among the earliest studies concerned with γδ TCR ligand recognition were those that compared αβ and γδ T cells for their ability to recognize MHC differences. The results are confusing because, although γδ T cells did not specifically respond to allo-antigens and did not show MHC-restricted specificities involving conventional antigens, γδ T cells capable of recognizing classical MHC molecules and related cell-surface proteins, including the CD1 molecules, have been isolated. This collection could be mined by crystallographers interested in the problem of ligand recognition by γδ T cells, and the fact that the structure of the MHC-ligand might be already available should represent an advantage in resolving the structure of TCR–ligand complex. However, this collection has hardly been touched so far, and in fact, only a single crystal structure involving a murine γδ TCR in complex with the T10/22 molecules has been solved [2]. This structure showed an interesting molecular interaction very different from those of most MHC-restricted αβ TCRs and their MHC-ligands. Further structures of this type will be needed before generalizations about γδ ligand recognition can be made.
Based on structural comparisons between γδ TCRs and immunoglobulins, it has been suggested that γδ ligand recognition should be Ig-like [3]. Like the immunoglobulins, where the complementarity determining region 3 (CDR3) of the IgH chain makes a major contribution to receptor diversity and specificity, the TCR-δ CDR3 in the γδ TCRs contributes the lion’s share to receptor diversity. Not surprisingly, then, TCR-δ CDR3 is solely implicated in the recognition of the T10/22 molecule, both structurally and functionally [2, 4, 5], and other studies provided further evidence for the importance of this particular portion of the γδ TCR in the recognition of additional ligands [6]. However, specific immunoglobulins can be raised by immunization with almost any antigen, whereas most antigens fail to elicit specific γδ T cells. Therefore, although physical γδ TCR-ligand interactions may be Ig-like, this attractive concept alone does not suffice in predicting what might be a γδ TCR ligand. Recent evidence indicates that receptor selection plays a major role in determining “allowed” specificities in peripheral γδ T cells [7]. Since γδ T cells develop in the thymus, such selection might occur here, and indeed, some studies provided strong evidence for thymic selection of these cells. The evidence is limited, however. So far, there are only two well-studied examples; the thymic selection of cells specific for T10/22 and of the TCR-invariant dendritic epidermal T cells (DETC) [8–10]. The T10/22-specific cells share a CDR3δ motif in their TCRs, which is for the most part contributed by a germline DNA segment (Dδ, reading frame 3) and thus shared by a relatively large number of clones. However, thymic selection was found to affect the functional development of these cells, such that TCR–ligand interactions were needed in the thymus for them to become producers of IFN-γ, rather than producers of IL-17A. The invariant TCR of the thymic precursors of DETC also appears to be selected. In this case, the putative ligand SKINT-1 [10], a molecule that belongs to the immunoglobulin superfamily of receptors [11], remains to be confirmed as a true ligand. Combining the notions of Ig-like receptor–ligand interactions and thymic selection one might speculate that γδ TCRs resemble natural Igs by having an inherent or selected specificity for auto-antigen, but such a concept requires more testing.
The structure of the γδ TCR genes predicts a high degree of adaptivity in the receptors [1] and indeed, numerous studies have documented peripheral TCR selection. These include human diseases such as polymyositis [12] and multiple sclerosis [13] as well as animal models of diseases including collagen-induced arthritis (CIA) [14], and what appears to be natural selection based on endogenous ligands in healthy individuals [15, 16]. However, how the selected TCRs might be functionally different is not clear, because their ligands remain unknown, and their binding properties and ability to activate the cells that express them cannot be assessed. In the mouse model of CIA, cells expressing selected TCRs (Vγ4Vδ4) also expressed the cytokine IL-17A at increased frequencies [14], suggesting a functional connection, but even prior to the immunization, the Vγ4+ subset contained IL-17A+ cells at slightly higher frequency than other subsets [14].
Because many of the putative ligands for the γδ TCRs are non-repetitive and too small to mediate TCR cross-linking, they must be anchored in some way to enable polyvalent interactions with the TCRs. Indeed, evidence for cellular presentation of the so-called phospho-antigens has been obtained [17, 18], although it is still unknown if distinct presenting molecules are involved. Furthermore, it has been reported that human γδ T cells recognize phospholipids when CD1d is present [19, 20], suggesting that CD1d might be the presenting molecule in this particular example. Cellular presentation of conventional peptide antigens, common in the specific responses of αβ T cells, has been suggested for γδ T cells [21]. Moreover, some data suggest that peptides having covalently attached non-peptidic ligands can, when bound to MHC, be recognized by γδ T cells [22]. If polyvalency is the only requirement for stimulatory ligands, small molecular moieties also might be bound to non-cellular structures and still be stimulatory.
Many of the observed γδ T cell responses are oligo-clonal, involving cells with shared germline elements in their TCRs, several of the antigens recognized are very small and require presentation, and the γδ TCRs are selectable and highly adaptive. To rationalize these observations, we suggested that the γδ TCR might be antigen-restricted instead of being restricted by a presenting molecule, and that its variability might serve to adjust to the variable cell surface context of an antigen that can be presented in many different ways [23]. However, non-MHC-restricted presentation of small molecules on the cell surface is not likely to be random, and specific molecular associations still might exist that are recognizable by the γδ TCRs. Thus, a general paradigm that adequately describes TCR–ligand interactions with γδ T cells remains to be established.
Proteins among the molecular moieties that stimulate γδ T cells
Although currently there is a strong focus on human γδ T cell responses to the non-peptidic phospho-antigens [24], proteins still might turn out to represent the bulk of the antigens recognized by γδ T cells. Indeed, tetanus toxoid, a strong immunogen derived from a protein, the tetanospasmin of Clostridium tetani, might have been the first defined antigen reported capable of stimulating specific γδ T cell responses [25, 26]. Others that followed include Ig λ light chain [27], certain heat shock proteins [28, 29], and the proteins staphylococcal enterotoxin A (SEA) [30], herpes simplex glycoprotein I [31] and—perhaps—insulin [32]. Furthermore, the less well-defined mycobacterial purified protein derivative (PPD) has been shown to elicit γδ TCR-dependent responses [28, 33], but it was not conclusively determined if its stimulatory moiety is really a protein. More recently, the defined mycobacterial protein ESAT-6 was found to stimulate γδ T cells [34], however, and this may not be the only mycobacterial protein recognized by γδ T cells [35]. Finally, in addition to natural proteins, a 100-amino acids-long synthetic protein, poly GT (a random hetero-copolymer of roughly 50 glutamic acids and 50 tyrosines), was found to stimulate strong γδ TCR-dependent responses in mice, with other related synthetic proteins being less stimulatory (poly GAT) or non-stimulatory (poly G, poly T) [36–39]. Because poly GT lacks a defined secondary structure, it resembles peptide antigens perhaps more than globular proteins.
In addition to soluble proteins, several cell surface-expressed proteinaceous ligands have been identified, including the classical MHC I and II molecules as well as the non-classical MHC I-related molecules T10/T22 in mice [40–42], and the more distant MHC I-like molecules MICA/B in humans [43, 44] as well as CD1c in humans [45] and CD1d in both mice and humans [20, 46]. Although the degree to which these various responses have been characterized varies greatly, the list in its entirety makes a good case that the γδ TCRs in principle are capable of interacting with proteinaceous ligands. This raises the question of whether small peptides might also be recognized.
TCR-dependent responses to small peptide antigens
Based on the concentrated diversity in the V–J junction of δ and the other TCR genes, together with a relative absence of diversity elsewhere in the TCRs, Davis and Bjorkman suggested that the γδ TCRs might be capable of recognizing “diverse small molecules (such as peptides) embedded in much less diverse and larger MHC molecules” [1], but so far only a small number of studies have actually examined responses of γδ T cells to defined peptide antigens (Table 1). Encouraged by the observation that purified recombinant HSP-60 could stimulate murine γδ T cell hybridomas of a select set [28], as well as findings of others implicating HSP-60 as a possible antigen recognized by γδ T cells in humans [29, 47], we decided to examine synthetic peptides representing portions of the HSP-60 protein for stimulatory activity. We quickly identified one particular sequence within mycobacterial HSP-60 represented by peptide 180–196, which stimulated strong TCR-dependent cytokine responses in these hybridomas [48]. All of the reactive hybridomas expressed Vγ1, most frequently associated with Vδ6 or Vδ4, but other Vδs were also seen [49]. Because the same hybridomas also produced small amounts of cytokine in the absence of the added peptide [28], and because some of the hybridomas responded as well to a matched peptide representing the homologous mammalian HSP-60 protein [48], it seemed possible that the spontaneous response is auto-reactive in nature, but this was not formally shown. Responsiveness to this peptide could be transferred with the rearranged TCR γ and δ genes of a peptide-reactive hybridoma [50], thus establishing TCR-dependence. To determine if the mycobacterial HSP-60 peptide also elicits a response of γδ T cells in vivo, we immunized mice with it and subsequently examined the TCRs and peptide responses of Vγ1+ γδ T cells from the immunized animals. An increase in the relative frequency of peptide-reactive cells and minor changes in their TCR-repertoire and structure were found, indicating that this small peptide indeed triggers select γδ T cells in vivo [51].
Table 1.
Responses of γδ T cells to specific peptides
γδ T cell source | Protein source | Peptide origin, name (epitope) | MHC restriction | TCR-dependent | Referencea |
---|---|---|---|---|---|
Mouse | HSP-60 | Mycobacterial (180–196) | No | Yes | [48–52] |
E. coli (181–197) | |||||
Mammalian (205–221) | |||||
Undefined | Several, corresponding to HSP-70 BiP motif | No | Yes | [55], unpubl. | |
None | EY nonamer | No | Yes | [39], unpubl. | |
Insulin B chain | B:9-23 [9–23] | No | Yes | [57] | |
Rat | Retinal S-antigen | PDSAg (341-354) | ? | ? | [65, 66] |
HLA-B | B27PD (125-138) | ? | ? | [65, 66] | |
Human | Tetanus toxin | C. tetani (1235-1246) | HLA-DRw53 | Yes | [21] |
Ig λ light chain | Processed peptide | No | Yes | [67] | |
Listeriolysin O | L. monocytogenes (470-508) | No | Yes | [68] |
It was not determined in what form this peptide is recognized. Hybridomas responded to it in the absence of added antigen-presenting cells (APCs) [49], suggesting that either APCs are not required at all, or that the hybridoma cell themselves can present the peptide. We observed the same with a TCR-deficient mouse T cell line (58α-β-) expressing a transgenic γδ TCR of the peptide-reactive type [50], indicating that no other γδ lineage-specific molecules are required for this response. However, others failed to find the response when human JURKAT cells were transfected with such a TCR [52]. This is reminiscent of recent studies with the response of primate γδ T cells to phosphoantigens, which requires cellular presentation of the phosphoantigens, but only primate-derived cells can act as presenters [53, 54]. Apparently, this reflects the specificity of the γδ TCR itself, because the same requirement was observed in binding studies with soluble γδ TCRs [18]. These findings suggest that the γδ TCRs have some intrinsic ability to identify a rudimentary “self”, akin perhaps to the intrinsic ability of αβ TCRs to recognize MHC molecules.
In further experiments, we explored the limits of this particular peptide sequence and derivatives in terms of their ability to elicit γδ T cell responses. The shortest HSP-60-derived peptides to retain stimulatory activity were seven amino acids in length, covering the sequence FGLQLEL [50]. A single amino acid substitution in this peptide (to FALQLEL) increased the stimulatory activity [50]. The biological significance of this peptide response is unclear. Because the FGLQLEL peptide roughly matched a motif shared by unfolded proteins, which bind to the molecular chaperone HSP-70 BiP [55], we tested unrelated peptides having this motif for their stimulatory activity as well, and found several that triggered hybridoma responses, albeit none as strong as those to the HSP-60 peptide (unpublished). It is thus conceivable that some γδ T cells recognize epitopes in unfolded proteins.
We also re-examined poly GT (in the current nomenclature for amino acids, this random hetero-copolymer would instead be designated poly E50Y50), and found that the same hybridomas that had shown reactivity with the HSP-60 peptides responded to it, as well as to several derivatives (e.g., poly GAT), and even a nonamer peptide with a non-random sequence of alternating glutamic acids and tyrosines [39]. Taken together, these studies established that short peptide molecules can elicit responses via γδ TCRs. Because the hybridomas examined were derived from non-immunized animals, their responses might reflect some natural reactivity lacking a high degree of specificity. Since these responses are polyclonal, peptide antigens of a wide variety might meet the criteria required for stimulation. A biological role for these peptide responses was not determined, however.
Recently, a different type of peptide response involving murine γδ TCRs was discovered. When examining the TCR-β chain repertoire of insulin peptide B:9-23-specific αβ T cells in NOD mouse model of type I autoimmune diabetes [56], one in a substantial collection of peptide reactive hybridomas was found to express γδ TCR. In contrast to the above-described peptide reactive γδ hybridomas, this clone (named SP9D11) expressed a TCR consisting of Vγ4 paired with Vδ10, and no auto-reactivity was seen with this cell [57]. In addition to the B:9-23 peptide, the SP9D11 hybridoma also responded to purified islets, but did not recognize intact or denatured insulin, or cells from another tissue used as control. Transfer of the TCR of this clone to another cell confirmed TCR-dependence of the peptide response. The hybridoma response was specific for the B:9-23 peptide. However, truncations and amino acid substitutions revealed that the carboxy-terminal amino acids 21 and 22 were not necessary for the response whereas the amino-terminal ones could not be changed. As has been seen with the αβ T cell responses to this peptide [58], substitution of the tyrosine in position 16 with alanine abrogated the γδ response. Interestingly, the same happened when the cysteine in position 19 was substituted with alanine [57], a change that does not affect the αβ T cells [58]. This could indicate that the cysteine itself is recognized by the γδ TCR, or that its ability to dimerize the peptide is essential for stimulation. A dimeric peptide might be able to cross-link γδ TCRs directly. Another interesting finding of this study was that individual isolated hybridoma cells could respond to the B:9-23 peptide [57]. This suggested that the peptide might be recognized without any cellular presentation, either in soluble form or perhaps bound to the surface of the culture plate, or that the responding cells are able to auto-present. To resolve this, direct binding studies will be required.
The SP9D11 hybridoma was derived from a mouse immunized with the B:9-23 peptide. In subsequent experiments, we used non-immunized NOD mice to generate further γδ T cell hybridomas. Several of these responded to the B:9-23 peptide as well [57]. This indicated that γδ T cells with this particular specificity also develop spontaneously, at least in NOD mice. However, these cells mainly expressed Vγ1+ TCRs in contrast to the Vγ4Vδ10 TCR of hybridoma SP9D11, which was derived from a peptide-immunized mouse [57]. Furthermore, their peptide specificity is shared with αβ T cells and B cells [56, 59, 60]. However, unlike αβ T cells and B cells, the γδ T cell response to insulin might be focused on this epitope alone. Importantly, there is now convincing evidence that insulin B:9-23 and closely related peptides occur naturally, contained within intra-islet dendritic cells, which engulf insulin granules and produce and present the peptides [61]. Thus, the B:9-23 peptide is a promising candidate as a natural peptide ligand for γδ T cells.
Perhaps the B:9-23-specific γδ T cells have some regulatory role or can become, like the αβ T cells and B cells with this specificity, involved in the autoimmune response that leads to type I diabetes in this mouse strain. Some time ago, it was reported that inhaled insulin induces γδ T cells capable of suppressing diabetes development in NOD mice [32], and we recently found that γδ T cells, including those expressing Vγ1, infiltrate the islets of prediabetic NOD mice (unpublished data). However, in neither case was the specificity of the γδ T cells determined.
Protein- or peptide-specific reactivity of γδ T cells also has been reported in animal models of mucosal tolerance. Thus, mice and rats that inhaled or ingested the protein ovalbumin (OVA) developed tolerance to this antigen, which could be transferred with γδ T cells [62–64]. The suppressive effect of the transferred γδ T cells was limited to responses against OVA [62]. Orally induced γδ T cells in rats suppressed experimental autoimmune uveitis (EAU) [65, 66]. These cells were induced by feeding the rats with short peptide antigens, including one derived from the retinal soluble antigen (S–Ag) and one derived from the HLA-B27 molecule (B27PD), which was also tolerogenic in this model. The in vivo induced γδ T cells also specifically proliferated in response to the peptides in vitro, albeit only in the presence of irradiated αβ T cells with the same specificity. These responses could be inhibited with antibodies against CD8 and MHC II molecules. It was not determined if the γδ T cells actually recognize the peptide antigens via their TCR, however.
Specific recognition of short peptides by human γδ T cells was also reported. One study involved cloned γδ T cells from the synovial fluid of a patient with rheumatoid arthritis. These clones specifically responded to tetanus toxin-derived peptides, and they expressed Vγ9Vδ2 while lacking CD4 and CD8 expression, the predominant phenotype of γδ T cells in human peripheral blood [21]. Unlike the polyclonal MHC non-restricted reactivity of Vγ9δ2+ γδ T cells to mycobacterial antigens, peptide responsiveness was limited to individual clones. Epitope mapping defined the specificity of one of these clones to a portion of the toxin represented by an 12mer peptide covering amino acids 1235–1246. The response to these peptides required APCs, and a screening of partially HLA-matched donors revealed that one γδ T cell clone (1.4) required a match in HLA-DRw53. HLA-DRw53 is a non-polymorphic class II MHC molecule, which might serve as a presenting element in the peptide response. The clones that responded to the toxin peptide in this study shared the broader reactivity of Vγ9 Vδ2+ cells with mycobacterial antigens [21]. In contrast, the earlier described γδ T cells with specificity for tetanus toxoid were CD8+ and expressed a different TCR (Vγ1+) [25]. The response of these cells also appeared to be MHC II-restricted, but it was not determined if they could recognize short peptide antigens.
γδ CTL clones from a Burkitt’s lymphoma patient, which recognized a B cell lymphoma Ig λ light chain idiotype [27], also responded to cells transfected with a signal sequence-deficient mutant of the SUP-B17 Ig λ chain. Target cells transfected with this construct expressed the truncated protein only intracellularly. Nevertheless, the Ig λ-specific γδ CTL clones lysed the transfected cells as well as others expressing the full-length construct on their surface [67]. Since purified tumor Ig and peptides corresponding to the hypervariable regions of the SUP-B17l chain failed to sensitize target cells to lysis by the γδ clones, the data suggested that antigen processing is required to generate a peptide that can be effectively recognized by the γδ T cells.
Finally, in cultures of human peripheral blood mononuclear cells (PBMC) stimulated in vitro with live Listeria monocytogenes bacteria, γδ T cells developed that were capable of recognizing listeriolysin O (LLO) and a peptide covering LLO amino acids 470–508 [68]. The specific response to this peptide antigen by the γδ T cells required peptide pulsed autologous PBMC, whereas HLA-DR4 matched PBMC, which supported αβ T cells responding to the same peptide, failed to support the γδ response. A specific restricting element was not identified. As with the murine responses to insulin [57], it is interesting to note that LLO-reactive human γδ and αβ T cells seem to share an epitope specificity with regard to this bacterial toxin [68]. However, while the γδ response appears to be focused on this particular epitope, αβ T cells also recognize other LLO epitopes. As noted by the authors, LLO 470–508 is highly homologous to other toxins of Gram-positive bacteria, perhaps indicating that the focus of the γδ T cell response includes an entire spectrum of related toxins.
Features of the peptide responses by γδ T cells
Interactions of peptides with proteins are often found in cell biology. A pentapeptide unit has emerged as a common minimal amino acid sequence in peptide protein interactions and in immune recognition [69]. The number of reported peptide responses of γδ T cells is small, but some features are already evident. First, peptides recognized by γδ T cells can be derived from both autologous and heterologous sources [48]. Hence, they might play roles in health and disease, in sterile and in infectious inflammation, and in the regulation of auto-reactivity and the detection of pathogens. Such a diverse range of peptide antigens fits well with the nearly unlimited potential of structural diversity of the γδ TCRs. Second, there might be few constraints with regard to the size of the recognized peptides. Peptides representing natural amino acid sequences as large as 38 amino acids [68], and as short as seven amino acids in length were found to stimulate γδ TCR-dependent responses [50]. Of note, in the studies with larger peptides, secondary peptide processing was not excluded. Unlimited peptide size would be incompatible with presentation of the larger peptides by MHC class I molecules, but presentation by MHC class II molecules and other molecules that do not restrict peptide size remain possibilities. Third, amino acid composition and sequence of the peptides determine responses. In all reported examples, only certain peptides representing an individual epitope within a larger protein sequence stimulated the γδ response. As mentioned earlier, with the HSP-60-derived peptides, we found that only certain truncations and amino acid substitutions were tolerated, whereas others reduced or completely abrogated the γδ response [50]. Similarly, the γδ TCR-dependent response to the insulin peptides B:9-23 was sensitive to amino-terminal truncations but permissive of carboxy-terminal truncations (up to amino acid 21), and substituting amino acids in positions 16 or 19 with alanine rendered the peptide non-stimulatory [57]. Some of our data suggested that amphipathic properties of a peptide matter, and we found that sequence un-related peptides still might stimulate certain γδ hybridomas within the larger set of HSP-60 peptide-reactive cells, as long as they contained hydrophobic or aromatic amino acids in every second position. The same hybridomas also responded to poly GT, and a nonamer peptide with alternating glutamic acids and tyrosines, EY9 (unpublished data). In sum, these observations indicate that the responses are peptide-specific, albeit with a range of specificities from polyclonal, and including entire groups of similar peptides, to clonal, and limited to one or a few related peptides. Finally, reports vary greatly with regard to antigen presentation requirements. Similar to the originally reported response of human γδ T cells to tetanus toxoid [25], the response of human γδ T cells to a peptide antigen derived from tetanus toxin required APCs [21]. It was also found to be HLA-DR4-restricted, suggesting a mechanism of antigen presentation involving MHC II [21]. In contrast, the response of human γδ T cells to a peptide derived from listeriolysin O did not appear to be MHC-restricted [68]. Similarly, there was no APC-requirement or MHC-restriction in the responses to HSP-60-derived [70] and insulin-derived peptides by murine γδ T cell hybridomas [57]. Here, it remains possible that the hybridoma cells themselves present peptides to one another. However, in the reactivity of one particular hybridoma (SP9D11) with the insulin peptide B:9-23, it was shown that isolated individual cells responded to the peptide, eliminating a requirement for intercellular presentation [57]. Notably, this peptide might represent a special case as it can dimerize by forming a disulfide bridge between cysteines in position 19. Suggestively, a peptide in which the cysteine was substituted by an alanine failed to stimulate the hybridoma [57]. Thus, available data might be best explained by a model in which peptide antigen presentation by professional APCs and MHC molecules is optional and limited to certain peptides only. Others might be presented by cells in an MHC non-restricted fashion, either involving other specialized cell-surface molecules or perhaps merely non-specific binding interactions between reactive peptides and cells [23], and multivalent peptides might even be able to stimulate directly, by cross-linking the TCR in the absence of any kind of presentation [39, 57].
Significance of the peptide responses
Presently, it is difficult to assess the significance of peptide responses by γδ T cells because it has not yet been demonstrated that small peptides bind to the γδ TCR itself or form a complex with a presenting molecule, and it remains to be seen if naturally occurring small peptides can be recognized. The first requirement pertains to the mechanism of peptide recognition as well as the utility of peptides in eliciting specific responses of γδ T cells. Thus, it is important to determine the affinity of binding interactions between γδ TCRs and small peptides as well as possible additional requirements such as the molecular context in which peptides might be recognized. It must also be determined if peptides indeed bind to the variable portion of the γδ TCRs, and if such interaction depends on the diverse V–J junctions. If peptides bind to the γδ TCR and elicit clonal or subset-specific responses, they might be used for immune modulation and other therapeutic purposes, via the functions of the targeted γδ T cells. Determining if γδ TCRs have evolved to recognize small peptide antigens is a more difficult task and will require the isolation of naturally occurring peptide ligands. As with αβ T cells, these might include break down products of larger proteins, but it is also conceivable that small proteins such as peptide hormones or defensins can be recognized.
Summary and conclusions
There is no compelling reason why γδ T cells should not recognize small peptide antigens. Small peptide antigens representing epitopes of a variety of proteins produced antigen-specific γδ TCR-dependent responses. The peptides are mostly recognized in the context of the eukaryotic cell surface, which might provide species-specificity. However, the precise mechanism(s) and the biological significance of peptide recognition via the γδ TCR remain to be determined.
Acknowledgments
The authors would like to thank Dr. George Eisenbarth for reading the manuscript and for helpful suggestions.
Note added in proof
Recently, a systematic approach to the identification of peptide antigens recognized by human gammadelta TCRs was reported as well as a stimulatory epitope on the oxidative stress response regulatory protein (OXYS), expressed by BCG.
Xi, X., Zhang, X., Wang, B., Wang, J., Huang, H., Cui, L., Han, X., Li, L., He, W., Zhao, Z. (2011) A novel strategy to screen Bacillus Calmette-Guerin protein antigen recognized by gammadelta TCR. PLoS ONE 6(4)e18809
References
- 1.Davis MM, Bjorkman PJ. T cell antigen receptor genes and T cell recognition. Nature. 1988;334:395–402. doi: 10.1038/334395a0. [DOI] [PubMed] [Google Scholar]
- 2.Adams EJ, Chien Y-H, Garcia KC. Structure of a gd T cell receptor in complex with the nonclassical MHC T22. Science. 2005;308:227–231. doi: 10.1126/science.1106885. [DOI] [PubMed] [Google Scholar]
- 3.Rock EP, Sibbald PR, Davis MM, Chien Y-H. CDR3 length in antigen-specific immune receptors. J Exp Med. 1994;179:323–328. doi: 10.1084/jem.179.1.323. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Shin S, El-Diwany R, Schaffert S, Adams EJ, Garcia KC, Pereira P, Chien Y-H. Antigen recognition determinants of gd T cell receptors. Science. 2005;308:252–255. doi: 10.1126/science.1106480. [DOI] [PubMed] [Google Scholar]
- 5.Adams EJ, Strop P, Shin S, Chien YH, Garcia KC. An autonomous CDR3delta is sufficient for recognition of the nonclassical MHC class I molecules T10 and T22 by gammadelta T cells. Nat Immunol. 2008;9:777–784. doi: 10.1038/ni.1620. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Chen H, He X, Wang Z, Wu D, Zhang H, Xu C, He H, Cui L, Ba D, He W. Identification of human T cell receptor gammadelta-recognized epitopes/proteins via CDR3delta peptide-based immunobiochemical strategy. J Biol Chem. 2008;283:12528–12537. doi: 10.1074/jbc.M708067200. [DOI] [PubMed] [Google Scholar]
- 7.Bonneville M, O’Brien RL, Born WK. Gammadelta T cell effector functions: a blend of innate programming and acquired plasticity. Nature Rev Immunol. 2010;10:467–478. doi: 10.1038/nri2781. [DOI] [PubMed] [Google Scholar]
- 8.Jensen KDC, Su X, Shin S, Li L, Youssef S, Yamasaki S, Steinman L, Saito T, Locksley RM, Davis MM, Baumgarth N, Chien Y-H. Thymic selection determines gammadelta T cell effector fate: antigen-naive cells make interleukin-17 and antigen-experienced cells make interferon gamma. Immunity. 2008;29:90–100. doi: 10.1016/j.immuni.2008.04.022. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Lewis J, Girardi M, Roberts SJ, Barbee SD, Hayday AC, Tigelaar RE. Selection of the cutaneous intraepithelial gd + T cell repertoire by a thymic stromal element. Nat Immunol. 2006;7:843–850. doi: 10.1038/ni1363. [DOI] [PubMed] [Google Scholar]
- 10.Boyden LM, Lewis JM, Barbee SD, Bas A, Girardi M, Hayday AC, Tigelaar RE, Lifton RP. Skint1, the prototype of a newly identified immunoglobulin superfamily gene cluster, positively selects epidermal gammadelta T cells. Nat Genetics. 2008;40:656–662. doi: 10.1038/ng.108. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Hayday AC. Gammadelta T cells and the lymphoid stress-surveillance response. Immunity. 2009;31:184–196. doi: 10.1016/j.immuni.2009.08.006. [DOI] [PubMed] [Google Scholar]
- 12.Pluschke G, Rüegg D, Hohlfeld R, Engel AG. Autoaggressive myocytotoxic T lymphocytes expressing an unusual γ/δ T cell receptor. J Exp Med. 1992;176:1785–1789. doi: 10.1084/jem.176.6.1785. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Shimonkevitz R, Colburn C, Burnham JA, Murray RS, Kotzin BL. Clonal expansions of activated γ/δ T cells in recent-onset multiple sclerosis. Proc Natl Acad Sci USA. 1993;90:923–927. doi: 10.1073/pnas.90.3.923. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Roark CL, French JD, Taylor MA, Bendele AM, Born WK, O’Brien RL. Exacerbation of collagen-induced arthritis by oligoclonal, iL-17-producing gd T cells. J Immunol. 2007;179:5576–5583. doi: 10.4049/jimmunol.179.8.5576. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Sim G-K, Augustin A. Extrathymic positive selection of γδ T cells. Vγ4Jγ1 rearrangements with “GxYS” junctions. J Immunol. 1991;146:2439–2445. [PubMed] [Google Scholar]
- 16.Sim G-K, Augustin A. Dominant expression of the T cell receptor BALB invariant δ (BID) chain in resident pulmonary lymphocytes is due to selection. Eur J Immunol. 1991;21:859–861. doi: 10.1002/eji.1830210352. [DOI] [PubMed] [Google Scholar]
- 17.Morita CT, Beckman EM, Bukowski JF, Tanaka Y, Band H, Bloom BR, Golan DE, Brenner MB. Direct presentation of nonpeptide prenyl pyrophosphate antigens to human γδ T cells. Immunity. 1995;3:495–507. doi: 10.1016/1074-7613(95)90178-7. [DOI] [PubMed] [Google Scholar]
- 18.Wei H, Huang D, Lai X, Chen M, Zhong W, Wang R, Chen ZW. Definition of APC presentation of phosphoantigen (E)-4-hydroxy-3-methyl-but-2-enyl pyrophosphate to Vgamma2Vdelta2 TCR. J Immunol. 2008;181:4798–4806. doi: 10.4049/jimmunol.181.7.4798. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Russano AM, Agea E, Corazzi L, Postle AD, De Libero G, Porcelli SA, De Benedictis F, Spinozzi F. Recognition of pollen-derived phosphatidyl-ethanolamine by human CD1d-restricted gamma delta T cells. J Allergy Clin Immunol. 2006;117:1178–1184. doi: 10.1016/j.jaci.2006.01.001. [DOI] [PubMed] [Google Scholar]
- 20.Agea E, Russano A, Bistoni O, Mannucci R, Nicoletti I, Corazzi L, Postle AD, De Libero G, Porcelli SA, Spinozzi F. Human CD1-restricted T cell recognition of lipids from pollens. J Exp Med. 2005;202:295–308. doi: 10.1084/jem.20050773. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Holoshitz J, Vila LM, Keroack BJ, McKinley DR, Bayne NK. Dual antigenic recognition by cloned human γδ T cells. J Clin Invest. 1992;89:308–314. doi: 10.1172/JCI115577. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Speir JA, Abdel-Motal UM, Jondal M, Wilson IA. Crystal structure of an MHC class I presented glycopeptide that generates carbohydrate-specific CTL. Immunity. 1999;10:51–61. doi: 10.1016/S1074-7613(00)80006-0. [DOI] [PubMed] [Google Scholar]
- 23.Born WK, O’Brien RL. Antigen-restricted gammadelta T-cell receptors? Arch Immunol Ther Exp (Warsz) 2009;57:129–135. doi: 10.1007/s00005-009-0017-x. [DOI] [PubMed] [Google Scholar]
- 24.Morita CT, Lee HK, Wang H, Li H, Mariuzza RA, Tanaka Y. Structural features of nonpeptide prenyl pyrophosphates that determine their antigenicity for human gamma delta T cells. J Immunol. 2001;167:36–41. doi: 10.4049/jimmunol.167.1.36. [DOI] [PubMed] [Google Scholar]
- 25.Kozbor D, Trinchieri G, Monos DS, Isobe M, Russo G, Haney JA, Zmijewski C, Croce CM. Human TCR-γ+/δ+, CD8+ T lymphocytes recognize tetanus toxoid in an MHC-restricted fashion. J Exp Med. 1989;169:1847–1851. doi: 10.1084/jem.169.5.1847. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Kozbor D, Cassatella MA, Lessin S, Kagan J, Finver S, Faust J, Trinchieri G, Croce CM. Expression and function of gammadelta- and alphabeta-T cell receptor heterodimers on human somatic T cell hybrids. J Immunol. 1990;144:3677–3683. [PubMed] [Google Scholar]
- 27.Wright A, Lee JE, Link MP, Smith SD, Carroll W, Levy R, Clayberger C, Krensky AM. Cytotoxic T lymphocytes specific for self tumor immunoglobulin express T cell receptor δ chain. J Exp Med. 1989;169:1557–1564. doi: 10.1084/jem.169.5.1557. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.O’Brien RL, Happ MP, Dallas A, Palmer E, Kubo R, Born WK. Stimulation of a major subset of lymphocytes expressing T cell receptor γδ by an antigen derived from Mycobacterium tuberculosis . Cell. 1989;57:667–674. doi: 10.1016/0092-8674(89)90135-9. [DOI] [PubMed] [Google Scholar]
- 29.Holoshitz J, Koning F, Coligan JE, De Bruyn J, Strober S. Isolation of CD4− CD8− mycobacteria-reactive T lymphocyte clones from rheumatoid arthritis synovial fluid. Nature. 1989;339:226–229. doi: 10.1038/339226a0. [DOI] [PubMed] [Google Scholar]
- 30.Rust CJJ, Verreck F, Vietor H, Koning F. Specific recognition of staphylococcal enterotoxin A by human T cells bearing receptors with the Vγ9 region. Nature. 1990;346:572–574. doi: 10.1038/346572a0. [DOI] [PubMed] [Google Scholar]
- 31.Johnson RM, Lancki DW, Sperling AI, Dick RF, Spear PG, Fitch FW, Bluestone JA. A murine CD4−, CD8− T cell receptor-γδ T lymphocyte clone specific for Herpes simplex virus glycoprotein I. J Immunol. 1992;148:983–988. [PubMed] [Google Scholar]
- 32.Harrison LC, Dempsey-Collier M, Kramer DR, Takahashi K. Aerosol insulin induces regulatory CD8 gammadelta T cells that prevent murine insulin-dependent diabetes. J Exp Med. 1996;184:2167–2174. doi: 10.1084/jem.184.6.2167. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Happ MP, Kubo RT, Palmer E, Born WK, O’Brien RL. Limited receptor repertoire in a mycobacteria-reactive subset of γδ T lymphocytes. Nature. 1989;342:696–698. doi: 10.1038/342696a0. [DOI] [PubMed] [Google Scholar]
- 34.Li L, Wu C-Y. CD4+ CD25+ Treg cells inhibit human memory gammadelta T cells to produce IFN-g in response to M. tuberculosis antigen ESAT-6. Blood. 2008;111:5629–5636. doi: 10.1182/blood-2008-02-139899. [DOI] [PubMed] [Google Scholar]
- 35.Brodin P, Rosenkrands I, Andersen P, Cole ST, Brosch R. ESAT-6 proteins: protective antigens and virulence factors? Trends Microbiol. 2004;12:500–508. doi: 10.1016/j.tim.2004.09.007. [DOI] [PubMed] [Google Scholar]
- 36.Vidovic D, Roglic M, McKune K, Guerder S, MacKay C, Dembic Z. Qa-1 restricted recognition of foreign antigen by a γδ T-cell hybridoma. Nature. 1989;340:646–650. doi: 10.1038/340646a0. [DOI] [PubMed] [Google Scholar]
- 37.Dembic Z, Vidovic D. Specificity and function of Qa-1 restricted γδ T cells. Res Immunol. 1990;141:600–602. doi: 10.1016/0923-2494(90)90064-6. [DOI] [PubMed] [Google Scholar]
- 38.Vidovic D, Dembic Z. Qa-1 restricted gamma delta T cells can help B cells. Curr Top Microbiol Immunol. 1991;173:239–244. [PubMed] [Google Scholar]
- 39.Cady CT, Lahn M, Vollmer M, Tsuji M, Seo SJ, Reardon CL, O’Brien RL, Born WK. Response of murine γδ T cells to the synthetic polypeptide poly-Glu50Tyr50 . J Immunol. 2000;165:1790–1798. doi: 10.4049/jimmunol.165.4.1790. [DOI] [PubMed] [Google Scholar]
- 40.Matis LA, Fry AM, Cron RQ, Cotterman MM, Dick RF, Bluestone JA. Structure and specificity of a class II alloreactive γδ T cell receptor heterodimer. Science. 1989;245:746–749. doi: 10.1126/science.2528206. [DOI] [PubMed] [Google Scholar]
- 41.Matis LA, Cron R, Bluestone JA. Major histocompatibility complex-linked specificity of γδ receptor-bearing T lymphocytes. Nature. 1987;330:262–264. doi: 10.1038/330262a0. [DOI] [PubMed] [Google Scholar]
- 42.Bonneville M, Ito K, Krecko EG, Itohara S, Kappes D, Ishida I, Kanagawa O, Janeway JCA, Murphy DB, Tonegawa S. Recognition of a self major histocompatibility complex TL region product by γδ T cell receptors. Proc Natl Acad Sci USA. 1989;86:5928–5932. doi: 10.1073/pnas.86.15.5928. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 43.Groh V, Steinle A, Bauer S, Spies T. Recognition of stress-induced MHC molecules by intestinal epithelial γδ T cells. Science. 1998;279:1737–1740. doi: 10.1126/science.279.5357.1737. [DOI] [PubMed] [Google Scholar]
- 44.Wu J, Groh V, Spies T. T cell antigen receptor engagement and specificity in the recognition of stress-inducible MHC class I-related chains by human epithelial γδ T cells. J Immunol. 2002;169:1236–1240. doi: 10.4049/jimmunol.169.3.1236. [DOI] [PubMed] [Google Scholar]
- 45.Porcelli S, Brenner MS, Greenstein JK, Balk SP, Terhorst C, Bleicher P. Recognition of cluster of differentiational antigens by human CD4− CD8− cytolytic T lymphocytes. Nature. 1989;341:447–450. doi: 10.1038/341447a0. [DOI] [PubMed] [Google Scholar]
- 46.Huber S, Sartini D, Exley M. Role of CD1d in Coxsackievirus B3-induced myocarditis. J Immunol. 2003;170:3147–3153. doi: 10.4049/jimmunol.170.6.3147. [DOI] [PubMed] [Google Scholar]
- 47.Fisch P, Malkovsky M, Klein BS, Morrissey LW, Carper SW, Welch WJ, Sondel PM. Human Vγ9/Vδ2 T cells recognize a groEL homolog on Daudi Burkitt’s lymphoma cells. Science. 1990;250:1269–1273. doi: 10.1126/science.1978758. [DOI] [PubMed] [Google Scholar]
- 48.Born W, Hall L, Dallas A, Boymel J, Shinnick T, Young D, Brennan P, O’Brien R. Recognition of a peptide antigen by heat shock reactive γδ T lymphocytes. Science. 1990;249:67–69. doi: 10.1126/science.1695022. [DOI] [PubMed] [Google Scholar]
- 49.O’Brien RL, Fu Y-X, Cranfill R, Dallas A, Reardon C, Lang J, Carding SR, Kubo R, Born W. Heat shock protein Hsp-60 reactive γδ cells: A large, diversified T lymphocyte subset with highly focused specificity. Proc Natl Acad Sci USA. 1992;89:4348–4352. doi: 10.1073/pnas.89.10.4348. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 50.Fu Y-X, Kersh G, Vollmer M, Kalataradi H, Heyborne K, Reardon C, Miles C, O’Brien R, Born W. Structural requirements for peptides that stimulate a subset of γδ T cells. J Immunol. 1994;152:1578–1588. [PubMed] [Google Scholar]
- 51.Fu Y-X, Cranfill R, Vollmer M, van der Zee R, O’Brien RL, Born W. In vivo response of murine γδ T cells to a heat shock protein-derived peptide. Proc Natl Acad Sci USA. 1993;90:322–326. doi: 10.1073/pnas.90.1.322. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 52.Kikuchi GE, Roberts K, Shevach EM, Coligan JE. Gene transfer demonstrates that the Vγ1.1Cγ4Vδ6Cδ T cell receptor is essential for autoreactivity. J Immunol. 1992;148:1302–1307. [PubMed] [Google Scholar]
- 53.Wang H, Lee HK, Bukowski JF, Li H, Mariuzza RA, Chen ZW, Nam K-H, Morita CT. Conservation of nonpeptide antigen recognition by rhesus monkey Vgamma2Vdelta2 T cells. J Immunol. 2003;170:3696–3706. doi: 10.4049/jimmunol.170.7.3696. [DOI] [PubMed] [Google Scholar]
- 54.Sarikonda G, Wang H, Puan KJ, Liu XH, Lee HK, Song Y, Distefano MD, Oldfield E, Prestwich GD, Morita CT. Photoaffinity antigens for human gamma delta T cells. J Immunol. 2008;181:7738–7750. doi: 10.4049/jimmunol.181.11.7738. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 55.Blond-Elguindi S, Cwirla SE, Dower WJ, Lipshutz RJ, Sprang SR, Sambrook JF, Gething M-JH. Affinity panning of a library of peptides displayed on bacteriophages reveals the binding specificity of BiP. Cell. 1993;75:717–728. doi: 10.1016/0092-8674(93)90492-9. [DOI] [PubMed] [Google Scholar]
- 56.Zhang L, Jasinski JM, Kobayashi M, Davenport B, Johnson K, Davidson H, Nakayama M, Haskins K, Eisenbarth GS. Analysis of T cell receptor beta chains that combine with dominant conserved TRAV5D*04 anti-insulin B:9-23 alpha chains. J Autoimmun. 2009;33:42–49. doi: 10.1016/j.jaut.2009.02.003. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 57.Zhang L, Jin N, Nakayama M, O’Brien RL, Eisenbarth GS, Born WK. Gamma delta T cell receptors confer autonomous responsiveness to the insulin-peptide B:9-23. J Autoimmun. 2010;34:478–484. doi: 10.1016/j.jaut.2009.12.008. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 58.Abiru N, Wegmann D, Kawasaki E, Gottlieb P, Simone E, Eisenbarth GS. Dual overlapping peptides recognized by insulin peptide B:9–23 T cell receptor AV13S3 T cell clones of the NOD mouse. J Autoimmun. 2000;14:231–237. doi: 10.1006/jaut.2000.0369. [DOI] [PubMed] [Google Scholar]
- 59.Simone E, Daniel D, Schloot N, Gottlieb P, Babu S, Kawasaki E, Wegmann D, Eisenbarth GS. T cell receptor restriction of diabetogenic autoimmune NOD T cells. Proc Natl Acad Sci USA. 1997;94:2518–2521. doi: 10.1073/pnas.94.6.2518. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 60.Nakayama M, Abiru N, Moriyama H, Babaya N, Liu E, Miao D, Yu L, Wegmann DR, Hutton JC, Elliott JF, Eisenbarth GS. Prime role for an insulin epitope in the development of type 1 diabetes in NOD mice. Nature. 2005;435:220–223. doi: 10.1038/nature03523. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 61.Mohan JF, Levisetti MG, Calderon B, Herzog JW, Petzold SJ, Unanue ER. Unique autoreactive T cells recognize insulin peptides generated within the islets of Langerhans in autoimmune diabetes. Nat Immunol. 2010;4:350–354. doi: 10.1038/ni.1850. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 62.McMenamin C, Pimm C, McKersey M, Holt PG. Regulation of IgE responses to inhaled antigen in mice by antigen-specific γδ T cells. Science. 1994;265:1869–1871. doi: 10.1126/science.7916481. [DOI] [PubMed] [Google Scholar]
- 63.McMenamin C, McKersey M, Kühnlein P, Hünig T, Holt PG. γδ T cells down-regulate primary IgE responses in rats to inhaled soluble protein antigens. J Immunol. 1995;154:4390–4394. [PubMed] [Google Scholar]
- 64.Ke Y, Pearce K, Lake JP, Ziegler HK, Kapp JA. Gamma delta T lymphocytes regulate the induction and maintenance of oral tolerance. J Immunol. 1997;158:3610–3618. [PubMed] [Google Scholar]
- 65.Wildner G, Hünig T, Thurau SR. Orally induced, peptide-specific γ/δ TCR+ cells suppress experimental autoimmune uveitis. Eur J Immunol. 1996;26:2140–2148. doi: 10.1002/eji.1830260927. [DOI] [PubMed] [Google Scholar]
- 66.Wildner G, Thurau SR, Diedrichs-Mohring M. Gamma–delta T cells as orally induced suppressor cells in rats: in vitro characterization. Ann NY Acad Sci. 2004;1029:416–422. doi: 10.1196/annals.1309.050. [DOI] [PubMed] [Google Scholar]
- 67.Kim HT, Nelson EL, Clayberger C, Sanjanwala M, Sklar J, Krensky AM. γδ T cell recognition of tumor Ig peptide. J Immunol. 1995;154:1614–1623. [PubMed] [Google Scholar]
- 68.Guo Y, Ziegler HK, Safley SA, Niesel DW, Vaidya S, Klimpel GR. Human T-cell recognition of Listeria monocytogenes: Recognition of listeriolysin O by TcRαβ+ and TcRγδ+ T cells. Infect Immun. 1995;63:2288–2294. doi: 10.1128/iai.63.6.2288-2294.1995. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 69.Luccese G, Stufano A, Trost B, Kusalik A, Kanduc D. Peptidology: short amino acid modules in cell biology and immunology. Amino Acids. 2007;33:703–707. doi: 10.1007/s00726-006-0458-z. [DOI] [PubMed] [Google Scholar]
- 70.Born W, Cady C, Jones-Carson J, Mukasa A, Lahn M, O’Brien R. Immunoregulatory functions of gammadelta T cells. Adv Immunol. 1999;71:77–144. doi: 10.1016/S0065-2776(08)60400-9. [DOI] [PubMed] [Google Scholar]