T cell receptor bias and affinity define two compartments of the CD1b-glycolipid specific T cell repertoire

Abstract

Current views emphasize T cell receptor (TCR) diversity as a key feature that differentiates the group 1 (CD1a, CD1b, CD1c) and group 2 (CD1d) CD1 systems. Whereas TCR sequence motifs define CD1d-reactive NKT cells, the available data do not allow a TCR-based organization of the group 1 CD1 repertoire. The observed TCR diversity might result from donor to donor differences in TCR repertoire, as seen for MHC-restricted T cells. Alternatively, diversity might result from differing CD1 isoforms, antigens and methods used to identify TCRs. Using CD1b tetramers to isolate clones recognizing the same glycolipid, we identified a previously unknown pattern of variable (V) gene usage (TRAV17, TRBV4-1) among unrelated human subjects. These TCRs are distinct from those present on NKT cells and germline-encoded mycolyl lipid reactive (GEM) T cells. Instead, they resemble the TCR of LDN5, one of the first known CD1b-reactive clones that was previously thought to illustrate the diversity of the TCR repertoire. Interdonor TCR conservation was observed in vitro and ex vivo, identifying LDN5-like T cells as a distinct T cell type. These data support TCR-based organization of the CD1b repertoire, which consists of at least two compartments that differ in TCR sequence motifs, affinity, and co-receptor expression.

Introduction

Group 1 CD1 proteins (CD1a, CD1b, and CD1c) are thought to play a role in immunity because they present bacterial lipid antigens to human T cells. The identification of lipid antigens and cellular pathways of lipid antigen presentation, as well as proof of principle that CD1 and lipid specific T cells can perform anti-microbial functions, was made possible by a small panel of T cell clones (1-8). The current understanding of the group 1 CD1 TCR repertoire emphasizes TCR sequence diversity. T cell clones that recognize CD1a, CD1b and CD1c express variable (V), diversity (D) and joining (J) genes that are different from one another (9-13). Lacking TCR sequence motifs, there is currently no basis for organizing group 1 CD1-reactive T cells based on TCR structure. This situation stands in sharp contrast to CD1d, which is also known as the group 2 CD1 system. When CD1d-reactive TCRs from unrelated donors are compared, they show shared sequence motifs that derive from use of a limited range of TCR V and J genes with few nontemplated (N) nucleotides. TCR conservation is a hallmark of NKT cells and is used to define widely recognized T cell subtypes of the CD1d-reactive T cell repertoire. Type I NKT cells, also known as invariant NKT cells (iNKT), have a strictly conserved TCR α chain that uses TRAV10 (also known as Vα24) whereas type II NKT cells also show discernable conservation, but do not strictly adhere to sequence motifs.

The apparently differing patterns of diverse or conserved TCRs among group 1 and group 2 CD1 systems, respectively, has been viewed as a fundamental difference between these systems. Based on comparisons to diverse TCRs that recognize MHC proteins, diverse TCRs in the group 1 CD1 system might imply acquired immune function, whereas conserved TCRs on NKT cells formed the basis of early arguments for their innate function (14). Also, in contrast with NKT cells, which are routinely tracked in vivo by staining the defining TCRs with tetramers or monoclonal antibodies (15), there are no widely used equivalent surface staining reagents for group 1 CD1-reactive T cells. The view that the group 1 CD1-specific TCR repertoire is diverse is currently based on a small number of clones that were generated in different laboratories in response to differing antigens. However, the recent validation of CD1b tetramers provides a method for the rapid generation of clones recognizing the same antigen, derived from genetically unrelated donors under similar conditions (16). Also, antigen-loaded CD1b tetramers allows direct analysis of patterns of TCRs present on polyclonal T cells, which largely bypasses biases that might be caused by technical factors related to the generation of T cell clones in vitro.

Recently, CD1b tetramers bound to the mycobacterial lipid glucose 6-O-monomycolate (GMM) were used to provide the first example of a conserved TCR pattern in the group 1 CD1 repertoire (17). These cells were designed germline-encoded mycolyl lipid-specific (GEM) T cells because their TCRs derive from germline sequences encoded by TRAV1-2 and TRAJ9, joined with few N nucleotide additions to create nearly invariant TCR α chains. Based on this finding, we considered whether the apparent TCR diversity among clones studied to date derives from donor to donor differences in TCR repertoire or instead derives from differences in the antigens and methods used to derive clones. Using tetramers to systematically analyze T cells from different donors that recognize the same GMM antigen, we identified a previously unknown pattern of TCR conservation that can be detected in vitro and ex vivo. Thus, one CD1b-antigen complex gives rise to at least two consistently observed TCR patterns in humans, providing a basis for using TCR motifs to organize the human CD1b TCR repertoire into at least two compartments that differ with regard to the variable regions used and their affinity for CD1b-glycolipid.

Materials and methods

Flow cytometry

GMM was purified from R. equi as described (18). CD1b monomers (NIH tetramer facility) were loaded with GMM and assembled into tetramers (16). Tetramers were incubated for 30 minutes at room temperature prior to adding monoclonal antibodies for an additional 30 minutes on ice. Flow cytometry data were pre-gated for lymphocytes based on forward and side scatter. Anti-CD3 and CD4 were from BD bioscience, the IFN-γ ELISPOT antibody pair was from Mabtech, anti-CD8α was produced in house, anti-CD8β and anti-TRBV4-1 were from Beckman Coulter, anti-TRAV1-2 was from Biolegend. Cells were analyzed on a BD FACS Canto II or sorted on an 11-color FACSAria (Becton Dickinson).

T cell cloning and T cell assays

Blood was donated at Massachusetts General Hospital blood bank or from asymptomatic tuberculin positive subjects with no clinical or radiographic evidence of active tuberculosis (Supplementary Table I), as approved by the institutional review boards of the Lemuel Shattuck Hospital and Partners Healthcare. Sorted GMM-loaded tetramer⁺ T cells were stored overnight in medium containing 0.2 ng/ml IL-15 and plated the next day at 1 cells/well in round bottom 96-well plates containing 2 ×10⁵ irradiated allogeneic peripheral blood mononuclear cells, 4 ×10⁴ irradiated Epstein-Barr virus transformed B cells, 30 ng/ml anti CD3 antibody OKT3 and 1 ng/ml of IL-2 which was added on day 2 of the culture. After 3 weeks, the wells with visible growth were restimulated. Clones were tested for binding of GMM loaded tetramer by flow cytometry and for antigen specificity in an ELISPOT assay. For ELISPOT assays, cocultures of APCs and T cell were incubated for 16 h in a Multiscreen-IP filter plate (96 wells; Millipore) coated according to the manufacturer's instructions (Mabtech).

PCR and molecular cloning

RNA was isolated from T cell clones with an RNeasy kit (Qiagen), and cDNA was synthesized with a Quantitect reverse transcription kit (Qiagen), including a genomic DNA removal step. V segment usage was determined by PCR using primerset IPS000029 and IPS000030 as described at www.imgt.org in combination with TCR α constant region reverse primer 5′ GTGGTAGCAGCTTTCACCTCCTTGG 3′ and TCR β constant region reverse primer 5′ GGTGGCAGACAGGACCCCTTGC 3′. Taq polymerase was used in the supplied buffer (Denville) under the following cycling conditions: an initial denaturation of 5 min. at 95°C, followed by 35 cycles of 1 min. at 94°C, 1 min. at 59°C, 1 min. at 72°C, followed by a final elongation step of 7 min. at 72°C.

SKW-3 cell lines

The full-length TCR α and β chains of CD1b- and MR1-specific T cells were cloned into a 2A peptide-linked pMIG vector, which also contains a gene encoding GFP. Vectors were used to transduce the TCR-negative human T cell leukemia line SKW-3. Cell lines were co-transduced with a pMIG vector containing the CD3 subunits ε, ζ, δ and γ separated by 2A peptides (gift by Richard Berry, Monash University). SKW-3.TCR cells (5×10⁴) were stained in 50μl PBS/2%FCS for 30 minutes at 4°C with 7-aminoactinomycin D viability dye (Sigma), CD3-Pacific Blue (UCHT1, BD Biosciences) and PE-conjugated tetramers of GMM-loaded CD1b. Cells were washed twice and analyzed on a BD LSRFortessa.

TCR affinity measurements

Soluble LDN5- and clone 2 TCR proteins were expressed and purified using methodology described previously (19) and floated in increasing concentrations over GMM-loaded or mock-loaded CD1b coupled to research-grade streptavidin-coated chips in a Biacore 3000. BIAevaluation version 3.1 software (Biacore AB) was used to fit the data to the 1:1 Langmuir binding model and the equilibrium data were analyzed with the Prism program for biostatistics, curve fitting and scientific graphing (GraphPad). Affinities of GEM TCRs were published previously (17).

Results

TCR α chain conservation among tetramer^int clones

We used CD1b tetramers loaded with GMM to sort T cells from four subjects (Supplementary Table I) and cloned them at limiting dilution. Clones were screened for GMM-dependent functional responses and tetramer staining as described (17). The clones segregated into two groups based on the intensity of tetramer staining, with intermediate staining clones (tetramer^int), exemplified by clone 83, and high staining (tetramer^high) clones exemplified by T cell clone 1 (Fig. 1 A). Clone 1 and other tetramer^high clones have been defined as GEM T cells because they express nearly invariant TCR α chains derived from germline gene segments TRAV1-2 and TRAJ9 with short N nucleotide regions (17). Tetramer^high and tetramer^int T cells were present among polyclonal T cells at detectable levels in all subjects included in this study (Supplementary Figure 1).

Figure 1. LDN5-like T cell clones show TCR conservation.

(A) CD1b tetramers were mock-treated or GMM-treated prior to staining of T cell clone 83 and GEM T cell clone 1. (B) Nucleotide sequences (lower case) and amino acid sequences (upper case) of CDR3 regions of the TCRs of tetramer^high GEM clone 1 (17), LDN5 (9), and tetramer^int clones (clones 26, 71, 34, 2, and 83). Light gray, V segment-derived nucleotides; dark gray, J segment-derived nucleotides; boxed areas, non-germline-encoded nucleotides in the α-chain. (C) Staining with antibodies against CD4, CD8α, and CD8β shows that tetramer^int clones can be CD4⁺, CD8αβ⁺ or CD4^–CD8αβ^–. For panel A, two experiments were performed for both clones that are shown. Tetramer staining of clone 83 is representative of six comparable clones, and staining of clone 1 is representative of three comparable clones. For panel C, each staining was performed once.

We sequenced clones from four unrelated subjects with chronic tuberculosis infection (A22, C58, A14, C39) in an attempt to identify conserved sequence motifs among the tetramer^int clones. Tetramer^int clones (clones 2, 26, 30, 71, 34, 58, 81 and 83) did not express GEM-defining TCR V and J segments seen in tetramer^high clones (Fig. 1 B). Further, in contrast to the short N nucleotide regions (0-3 nucleotides) generally found in GEM TCR α chains (17), we noted long and diverse N nucleotide regions involving 4 to 10 nucleotides in all tetramer^int clones. Whereas GEM T cell clones have only been detected in the CD4⁺CD8^– pool, tetramer^int clones showed all three patterns of co-receptor expression that are commonly found among mature T cells: CD4⁺CD8β^–, CD4^–CD8β⁺ and CD4^–CD8β^– (Fig. 1 C).

Despite the lack of precise CDR3 α sequence conservation of the type seen in NKT cells and GEM T cells, certain features of TCR α conservation were apparent. All TCR sequences showed the same CDR3 α length, and bias of TRAV genes was observed (Fig. 1 B). Among the first eight clones studied, five used TRAV17, which is expressed by approximately 6.5 percent of human T cells, based on our survey of published sequences (20). The odds of detecting five or more TRAV17 expressing clones among eight randomly picked ones are 0.10 percent. Separately, TRAV17 is also expressed by the CD1b-GMM-reactive clone LDN5, which was derived from a leprosy patient many years prior to developing the tetramer methods described here (2). Thus, CD1b-GMM-specific TCRs from clones demonstrated conserved CDR3 α length and TRAV17 usage among five donors studied, but stringent sequence motifs comparable to those in GEM T cells, NKT cells or mucosal-associated invariant T (MAIT) cells were not observed.

TCRβ conservation

Despite variation in CDR3 β length and sequence, all but one of the tetramer^int clones from tuberculosis patients expressed TRBV4-1, the same V gene expressed by LDN5 (Fig. 1 B). Thus, contrary to the general view that TCRs that recognize group 1 CD1 proteins express diverse αβ TCRs (9-13), sequences of LDN5 and CD1b tetramer^int clones provide evidence for V segment bias in both TCR chains.

The obtained TCR nucleotide and amino acid sequences from unrelated donors emphasizes three distinct aspects of the nature of sequence conservation among these TCRs (Fig. 1B). First, we detected three instances of identical α and β chain nucleotide sequences present in two clones from the same donor (clones 30 and 71; clones 34 and 58; clones 81 and 83). The clones were derived from different blood draws or culture wells. Because independent identical rearrangement events in both chains are unlikely, this pattern suggests clonal T cell expansion in vivo prior to the collection of the blood sample. Second, in donor C58, the GMM reactive T cell clones included two different β chain sequences that contained TRBV4-1 (Fig. 1B). This pattern is consistent with expansion of similar but not identical T cells that recognize the same antigen within one donor (intradonor conservation). Third, the pattern of GMM-reactive T cell clones expressing a TRAV17-containing α chain paired with a TRBV4-1-containing β chain is seen in multiple donors (CC58, A14, C39 and leprosy patient), demonstrating conservation in the TCR α and β chain V gene expression among unrelated donors (interdonor conservation). LDN5 was previously thought to be a unique type of T cell derived from the skin of one leprosy patient, but these data show that GMM-reactive T cells from other patients with mycobacterial infection express similar TCRs. Based on patterns of intradonor and interdonor conservation of TCRs that resemble the TCR expressed by LDN5, we designated GMM-specific, tetramer^int, TRAV4-1⁺ cells as LDN5-like T cells.

LDN5-like T cell populations ex vivo

To determine if LDN5-like T cells were detected ex vivo by functional responses as polyclonal populations, we tested freshly isolated T cells from a tuberculosis patient and healthy donors recruited in a blood-bank (BB). Experiments addressed the potential shared use of TRBV4-1, which was seen in LDN5 and all but one tetramer^int clone (Fig. 1 B). Using a monoclonal antibody that specifically recognizes TCR β chains with TRBV4-1, sorted CD3⁺TRBV4-1⁺ T cells were compared to CD3⁺TRBV4-1^– cells. Because CD3⁺TRBV4-1^– cells might include GEM T cells, which express TRAV1-2, we excluded TRAV1-2⁺ cells, expecting CD1b-GMM-reactive cells to be in the CD3⁺TRBV4-1⁺ gate only. CD3⁺TRBV4-1⁺ cells from a tuberculosis patient (C32) and a blood-bank donor (BB8) were activated by GMM in a CD1b-dependent manner to produce interferon-γ (IFN-γ) (Fig. 2). No response was detected using equal numbers of CD3⁺TRBV4-1^–TRAV1-2^– T cells. T cells from a second blood bank donor (BB40) failed to recognize GMM, indicating that GMM-specific cells, if present, were below the threshold of detection. TRBV4-1 is used by many other T cells, so sorting TRBV4-1⁺ T cells is expected merely to enrich for LDN5-like T cells, a conclusion that is supported by the rate of T cell response (Fig. 2). Nevertheless, CD1b-GMM reactive T cells are detectable ex vivo without the use of tetramers, and they are enriched among TRBV4-1+ T cells (Fig. 1).

Figure 2. TRBV4-1 expressing T cells are enriched for CD1b-GMM specific T cells.

T cells of a latent tuberculosis patient (C32) or random blood bank donors (BB8, BB40) were sorted based on expression of CD3 (gate not shown), TRAV1-2 and TRBV4-1 (left panels). An IFN-γ ELISPOT assay was performed using equal numbers of the indicated sorted T cells that were stimulated directly ex vivo with K562 cells that were transfected with CD1b or CD1c in the presence or absence of GMM. The experiment was performed twice on donor C32 and once on donors BB40 and BB8.

CD1b-specific cells express variable cytokine profiles

We next sought to determine whether LDN5-like T cells and GEM T cells express any characteristic cytokine profile. Supernatants of T cell clones stimulated with phorbol-12-myristate-13-acetate and ionomycin were analyzed in a Luminex multiplex cytokine assay (Fig. 3). Clones were organized into groups that met criteria for LDN5-like T cells (LDN5, clones 2, 26, 34), GEM T cells (clones 1, 42), and clones recognizing CD1a and the mycobacterial lipopeptide dideoxymycobactin (DDM) (clones 32, 15, 5). As controls, we tested clones with irrelevant antigen specificity (clones 50, 101) or unstimulated LDN5 T cells. The assay included cytokines characteristic for the helper T cell subsets Th1, Th17 and Th2, the acute phase and anti-viral cytokines IL-6, TNF, LT-α, IFN-α and IFN-β, the chemokines IP10 (CXCL10), XCL-1, MIP1β (CCL4), I-309 (CCL1) and TARC (CCL17), and the negatively regulating mediators IL-10 and the IL-1 receptor antagonist (IL-1Ra). The data show that all stimulated clones expressed TNF and IFN-γ, a pattern previously observed for GEM T cells that is consistent with a proposed anti-microbial effector function of GMM-specific T cells. LDN5-like T cell clones did not show shared patterns that define them as Th2- or Th17-like T cells, but instead show somewhat varied cytokine profiles, which suggests that LDN5-like T cells may exist as functionally specialized subsets.

Figure 3. Cytokine production by T cell clones.

Luminex assay of cytokines in supernatants of LDN5-like clones (LDN5, clones 2, 26, 34) GEM T cell clones (clones 42, 1), control clones (clones 101, 50), and CD1a specific T cell clones that recognize didehydroxymycobactin (clones 32, 15, 5). T cells were left unstimulated or were stimulated for 16 h with the phorbol ester PMA and ionomycin. Data are from one Luminex experiment using pooled supernatants from triplicate wells. Analytes that reached levels that were above the limit for reliable detection are indicated with an asterisk.

TCR avidity for CD1b-glycolipid

A consistently observed difference between LDN5-like and GEM T cells is that the former show intermediate staining with CD1b-GMM tetramer, and the latter stain brightly (Fig. 1A, Supplementary Fig. 1) (17). Factors that contribute to the tetramer binding by a T cell clone are the density of TCR expression at the cell surface, the affinity of the TCR for CD1b, and contributions by receptors other than the TCR. Therefore, we sought to measure the avidity and affinity of TCRs using quantitative or direct methods. To control TCR expression level, we transduced SKW-3 cells with GEM TCRs (clones 21, 1, 42) or LDN5-like TCRs (LND5, clones 2, 34) and measured tetramer staining within narrow CD3 windows over a broad range of added tetramer, as show in detail for clone 21 (Fig. 4 A, upper panels). In all cases and across broad ranges of tetramer concentration, SKW-3 cells with GEM TCRs stained more intensely with CD1b-GMM tetramer than cells expressing equivalent density of LDN5-like TCRs (Fig. 4 A, lower panel and Fig. 4 B).

Figure 4. Tetramer binding of TCR-transduced cell lines reflects TCR affinity.

(A) CD1b-GMM tetramer titration using 5×10⁴ SKW-3.TCR21 cells in a fixed volume, in the presence of a fixed quantity of anti-CD3. Cells were gated for equivalent level of CD3 expression. The MFI of tetramer staining of gated cells was determined. The raw data are shown for clone 21 only (upper panels), and the data of all six clones are summarized in the lower panel. Results are expressed as fold change of the MFI of tetramer staining over MFI of unstained cells. (B) TCR-transduced cell lines were stained with anti-CD3 and 750 ng CD1b-GMM tetramer and gated for equivalent levels of CD3 and TCRαβ-GFP expression (upper panels). CD1b-GMM tetramer binding by CD1b-specific TCRs was compared to binding by a negative control cell line transduced with a MAIT TCR. The experiment included one mycolic acid-specific GEM TCRs (clone 18), two GMM-specific GEM TCRs (clones 1 and 21), and three LDN5-like TCRs (LDN5 and clone 2 and 34) and was performed once. The experiments in B and C were performed twice. (C) Coomassie-stained SDS-PAGE gel of recombinant leucine zippered TCR of LDN5 and clone 2. MW: molecular weight marker; R: reducing; NR: non reducing conditions. (D) Surface plasmon resonance measurement of the interaction between soluble TCRs of LDN5 and clone 2 and immobilized mock-loaded or GMM-loaded CD1b. (E) Equilibrium dissociation constant (K_d) calculated from measurements in D, and previously published values for GEM TCRs (17). Data are representative of two independent experiments (error bars: SEM).

As a complementary approach with additional negative controls, SKW-3 cells were transduced with green fluorescent protein (GFP)-tagged MAIT TCR (M33.20), CD1b-mycolic acid reactive TCR (clone 18), LDN5-like TCRs and GEM-TCRs. Cells were gated for equivalent GFP and CD3 staining and validated for CD1b-GMM tetramer binding (Fig. 4 B). The amount of tetramer used in this experiment was 750 ng, which was non-saturating (Fig. 4 A). Again, SKW-3 cells expressing TCRs of GEM T cell clones 1, 21, and 42 showed high tetramer staining, while LDN5-like TCRs mediated intermediate tetramer staining. These data control for TCR expression, cell surface effects and tetramer loading with antigen, providing quantitative measurements in support of the conclusion that LDN5-like TCRs show lower avidity for CD1b-GMM compared to GEM TCRs in all cases studied.

TCR affinity for CD1b-glycolipid

For direct measurement of TCR affinity to CD1b-glycolipid, we generated the disulfide-linked, transmembrane region-truncated recombinant TCRs of LDN5 and the LDN5-like clone 2 (19). The proteins were of the expected apparent molecular weight under reducing and non-reducing conditions (Fig. 4 C). Soluble TCR was passed over immobilized, CD1b to determine the affinity values. Both the LDN5 TCR and the clone 2 TCR bound to GMM-treated, but not mock-treated CD1b. We determined a K_d of 39.4 μM ± 8.0 μM and 19.4 μM ± 0.88 μM for LDN5 and clone 2, respectively (Fig. 4 D). These K_d values are comparable to what are considered intermediate affinity conventional peptide-specific TCRs (21), and they are 20- to 40-fold higher than previously reported values for GEM TCRs made using the same method (Fig. 4 E) (17). Thus, compared to GEM T cells, LDN5-like T cells consistently exhibit lower affinity and avidity binding to GMM-loaded CD1b complexes. These data link the less stringent TCR motifs with lower affinity antigen recognition in a pattern that is reminiscent of type II NKT cells (22).

Discussion

Collectively these results identify LDN5-like T cells as cells expressing diverse co-receptors and TRAV17- and TRBV4-1-biased TCRs, which mediate intermediate affinity interactions with CD1b and GMM. Accordingly, LDN5, a T cell clone that was previously considered a single TCR within the diverse T cell repertoire that recognizes group 1 CD1, now appears to be a “generalizable anecdote” that identifies a previously unknown pattern of interdonor TCR conservation. These observations run counter to widespread views that the group 1 CD1 TCR repertoire is so complex that TCRs cannot be used to meaningfully organize the repertoire into subsets. Instead, systematic analysis of T cell clones recognizing one CD1b-antigen complex revealed that the CD1b repertoire consists of at least two definable compartments, GEM T cells and LDN5-like T cells. Sequence conservation among LDN5-like TCRs relies predominantly on the TCR β chain, and α chain conservation is lower than the nearly identical TCR α chains present in GEM TCRs. Also, whereas all known examples of GEM T cells are CD4⁺CD8^–, LDN5-like T cells seem to have more diverse co-receptor expression with examples of CD4 and CD8 single positive, as well as double negative T cell clones present in the panel (Fig. 1 C, Table I).

Table 1. TCRs define two subsets in the CD1b-reactive repertoire.

	GEM	LDN5-like
TCR α chain	TRAV1-2, TRAJ9	TRAV17
Stringency of motif	Invariant, uniform length	Bias, uniform length
TCR β chain	TRBV6-2	TRBV4-1
Stringency of motif	Bias, variable length	Bias, variable length
Coreceptors	CD4	CD4, CD8, or DN
Affinity	High	Intermediate

Thus, when experimental variables related to differing antigens and cloning methods are eliminated, the first two efforts using tetramers to systematically characterize CD1b-reactive TCRs revealed two clear patterns of interdonor conservation. Two different types of TCRs recognize the same GMM antigen, high affinity TCRs that utilize TRAV1-2 and TRAJ9, and intermediate affinity TCRs that utilize TRAV17 and TRBV4-1. Thus, TCR diversity is not an intrinsic and universal characteristic of group 1 CD1-reactive T cells. The high affinity and highly stringent sequence motifs of GEM T cells compare to type I NKT cells, whereas the lower affinity and motif stringency in LDN5-like T cells compares to type II NKT cells. Indeed, questions now arise as whether the interdonor conservation is typical of the repertoire of T cells recognizing the non-polymorphic antigen presenting molecule, CD1b. To date, GMM is the only CD1b antigen studied systematically, so tetramer-based analysis of other known lipid ligands of CD1b, including phosphaditylinositol mannoside, sulfoglycolipid, free mycolate, glycerol mycolate or phosphatidylinositol provide a means to test this new model. If interdonor conservation exists broadly, it might provide a practical means to diagnose infection or other diseases that involve lipid antigens, and it has implications for where CD1b-reactive T cells fall in the spectrum of innate and acquired immunity.