Conservation of T Cell Receptor Conformation in Epidermal γδ Cells with Disrupted Primary Vγ Gene Usage

See allHide authors and affiliations

Science  13 Mar 1998:
Vol. 279, Issue 5357, pp. 1729-1733
DOI: 10.1126/science.279.5357.1729


A feature that distinguishes γδ T cell subsets from most αβ T cells and B cells is the association of expression of single T cell receptor (TCR) γ and δ variable (V) region gene segments with specific anatomic sites. Mice lacking the TCR Vγ5 chain normally expressed by most dendritic epidermal T cells were shown to retain a conformational determinant (idiotype) ordinarily expressed exclusively by such Vγ5+cells. Conservation by shuffled γδ TCR chains of an idiotype associated with a specific anatomic site indicates that for TCRγδ, as for immunoglobulin, conformation is associated to a greater extent with the function or development of lymphocyte repertoires than is the use of particular gene segments.

The efficacy of the adaptive immune system depends on its capacity to recognize pathogens in a highly antigen-specific manner. B cells and αβ T cells recognize antigens through surface immunoglobulin (Ig) and TCRs, respectively. Although γδ cells regulate immune responses to protozoal, bacterial, and viral infection (1, 2), neither their primary physiological functions nor their antigen specificities have been fully clarified.

A characteristic feature of γδ cells is the association of single γ and δ chains with γδ cell subsets in specific anatomic sites. For example, most human peripheral blood γδ cells express Vγ9 and Vδ2 chains of relatively limited diversity (3). More extreme examples occur in murine epithelia. Essentially all reproductive tract γδ cells express a canonical Vγ6-Vδ1 TCR, whereas 60 to >99% of dendritic epidermal T cells (DETCs)—variation depending on strain and age of the mice—express a canonical Vγ5-Vδ1 TCR (4), which can be detected with the monoclonal antibody (mAb) 17D1 (5). Ordinarily, 17D1 does not react with any other γδ cells, including those of the reproductive epithelium that share with DETCs use of the identical Vδ1 chain. Such site-specific homogeneity of antigen receptor expression had not been observed in previous studies of αβ T cells and B cells. To investigate this feature of γδ cells, we examined the effect on DETC development of Vγ5 gene disruption (6) (Fig. 1).

Figure 1

Disruption of the Vγ5 gene. A 7-kb Balb/c genomic clone that contains both Vγ5 and Vγ6 coding regions was used to generate the Vγ5 disruption construct. The Vγ5 gene was disrupted by insertion of a neomycin resistance gene (neo) under the control of a phosphoglycerate kinase gene promoter into an Eco RV site in the Vγ5 coding region. A 3-kb Bgl II fragment was deleted to generate a unique Bgl II site [(B)], into which two herpes simplex virus thymidine kinase genes (TK) were inserted. The 3-kb Bgl II fragment was subsequently used as a probe in Southern blot analysis, detecting a 7-kb Eco RI (R) genomic fragment in the germline configuration and a 6-kb Eco RI fragment from the recombinant allele. Lane 1, Eco RI–digested DNA from targeted 129 embryonic stem cell clone 21.2; lanes 2 to 4, Eco RI–digested tail DNA of Vγ5+/+, Vγ5−/−male, and Vγ5−/− female mice, respectively.

To confirm Vγ5 gene disruption, we stained epidermal sheets (7) from Vγ5−/− and Vγ5+/+ littermates with mAbs to Vγ5, TCRδ, and CD3ɛ (Fig.2). Vγ5+ cells were readily detectable in epidermis from Vγ5+/+ mice, but not in that from Vγ5−/− mice. Nonetheless, dendritic CD3+TCRδ+ DETCs were present in Vγ5−/− mice at densities not significantly different from those in Vγ5+/+controls (153 ± 26 versus 221 ± 79/mm2, respectively; P = 0.54).

Figure 2

Epidermal sheets prepared from Vγ5+/+ (A to D) and Vγ5−/− (E to H) mice and stained with antibodies to Vγ5 (A and E), antibodies to CD3 (B and F), antibodies to TCRδ (C and G), or mAb 17D1 (D and H) (7). Original magnification, ×400. Data are representative of 10 fields of 0.042 mm2 per specimen, two to six specimens per mouse, two mice per experiment, and three experiments.

Because normal DETCs do not readily develop in mice lacking either TCRδ (8) or p72 SYK, a putative transducer of signals from the DETC TCR (9), the development of a DETC repertoire in Vγ5−/− mice implied that other γδ TCRs could substitute for Vγ5-Vδ1. To determine whether those TCRs were similar in structure to the canonical DETC TCR, we examined the “replacement” repertoire with mAb 17D1. This mAb was generated by immunizing Lou/M rats with a DETC line, fusing splenocytes with the SP2/0 myeloma, and screening the resulting mAbs for reactivity with DETCs but not with peripheral T or natural killer cells. From a DETC clone (1D2), both 17D1 and a pan antibody to TCRδ (3A10) immunoprecipitated proteins (10) of sizes similar to those previously described for the 1D2 γδ TCR (11) (Fig.3A). Moreover, pretreatment of lysates with 3A10 removed all 17D1 immunoreactivity, whereas pretreatment with 17D1 removed all 3A10 immunoreactivity (Fig. 3A). These results placed the 17D1 epitope on the TCR. However, 17D1 yielded negligible staining with either reproductive tract γδ cells or five hybridomas that express the same Vδ1-Dδ2-Jδ2 chain as DETCs (12), but paired with Vγ6 rather than with Vγ5 (Table1). Thus, 17D1 reactivity could not be attributed simply to the expression of Vδ1-Dδ2-Jδ2. Indeed, among 19 hybridomas and cell lines expressing different TCR γδ chain combinations (Table 1), 17D1 reacted only with those expressing both Vδ1 and Vγ5, consistent with it defining a characteristic DETC TCR conformation.

Figure 3

17D1 epitope and antikeratinocyte reactivity displayed by Vδ5 DETCs. (A) DETC clone 1D2 was surface-labeled with125I and lysed with 1% NP-40, and the resulting lysate was analyzed by immunoprecipitation, SDS-PAGE, and autoradiography (10). The mAb 3A10 to TCRδ (lane 2) and 17D1 (lane 4) precipitated two proteins of 42 and 50 kD, the sizes expected for the 1D2 TCR (11); control hamster IgG (lane 1) and rat IgM (lane 3) did not. Pretreatment of lysate with 3A10 removed all 17D1 reactivity (lane 5), and pretreatment with 17D1 removed all 3A10 reactivity (lane 6). Molecular size standards (in kilodaltons) are indicated on the right. (B) DETC line 7-17 (derived from an AKR/J mouse) and clones 30B4 and U10E1 (derived from Vγ5−/− mice) were stained and analyzed by FACS as described (13) with 17D1 and antibodies to Vγ5, to Vγ1, to Vγ7, or to Vδ4. (C) DETC lines 2-10-B6 (Vγ5+, 17D1+) and 5-6 (Vγ5, 17D1+) and clone 30B4 (Vγ5, 17D1+) were cultured (20 hours, 37°C) in medium only or in the presence of irradiated PAM2.12 keratinocytes. Culture supernatants were tested for the ability to support the growth of the IL-2–dependent CTLL cell line as measured by [3H]thymidine incorporation (cells were harvested at 48 hours, after a 24-hour pulse). Data are expressed as units of IL-2 per milliliter and are means ± SEM of triplicate wells (28).

Table 1

Analysis of cell lines and hybridomas of known TCR composition by flow cytometry (13) for reactivity with mAb 17D1 and mAbs to TCRγδ (3A10 or GL3).

View this table:

Unexpectedly, however, the same conformation was detected in epidermal sheets from Vγ5−/− mice (Fig. 2). Flow cytometry of epidermal cells (13) from individual mice showed that 33 ± 13% ( n = 6) of the TCRγδ+ DETCs from Vγ5−/−mice were 17D1+, compared with a value of 78 ± 18% ( n = 5) for Vγ5+/+littermates. To investigate the basis for 17D1 epitope expression on Vγ5−/− DETCs, we applied reverse transcription polymerase chain reaction (RT-PCR) analysis (14) to a 17D1+ DETC clone (30B4) derived from Vγ5−/− DETC mice (15). Transcripts of Vγ1-Jγ4 and Vδ1-Jδ2, but not of Vγ4, -5, -6, or -7 or Vδ4 or Vδ6, were detected (16). Sequencing revealed a simple, in-frame Vγ1-Jγ4 join, identical to that of the thymic hybridoma AA37 (17). The join, devoid of non–template-encoded nucleotides, might have been generated by recombination mediated by small stretches of sequence homology, which is common in fetal thymic V(D)J recombination (8,18). Sequencing of the Vδ1-Jδ2 product likewise revealed a simple, in-frame join, identical to the canonical Vδ1-Dδ2-Jδ2 junctions present in day-13 to day-17 fetal thymocytes, in DETCs, in reproductive tract γδ cells, and in a subset of the cell lines and hybridomas listed in Table 1 (4, 12).

Independent evidence for the association of this Vγ1-Vδ1 chain pairing with the 17D1 epitope was provided by PCR and restriction fragment length polymorphism (RFLP) analysis (19) of polyclonal 17D1+ DETCs that were sorted by flow cytometry from a different Vγ5−/− mouse. The analysis detected the canonical Vδ1-Jδ2 rearrangement and only a single Vγ1-Jγ4 rearrangement, which was of the size predicted for the in-frame 30B4 join. Moreover, the predominant sequence obtained from the polyclonal population was identical to that of 30B4, with the exception of a conservative (serine to threonine) switch at the V-J junction (16).

Flow cytometry confirmed that clone 30B4 expressed Vγ1 (Fig. 3B), as did other Vγ5, 17D1+ cells. Vγ1 expression has been previously detected in the epidermis of normal mice and in hybridomas derived therefrom (20), but not in combination with either Vδ1 or the 17D1 epitope. Several 17D1 lines and clones that were likewise isolated from Vγ5−/−epidermis expressed various TCRs, including Vγ7 (Fig.3B), Vγ4, and Vγ1. Vγ7 is often present in the gut paired with Vδ4 (21,22), but has not previously been identified in the epidermis. No Vγ7+ DETC clones stained with antibodies to Vδ4, although a Vδ4+Vγ7 DETC line was isolated. Thus, DETCs in Vγ5−/− mice constitute a distinct repertoire, including Vγ1+ cells that retain the 17D1-defined TCR conformation, ordinarily characteristic of Vγ5-Vδ1 DETCs.

Conventional Vγ5-Vδ1 DETCs secrete interleukin-2 (IL-2) in response to PAM2.12 transformed keratinocytes (23). This TCR-dependent activity was also exhibited by the 17D1+ clone 30B4, and by the 17D1+ line 5-6 isolated from Vγ5−/− mice (Fig. 3C). Similar to normal DETCs, 30B4 and 5-6 also lysed PAM2.12 cells. Thus, 17D1+ cells from Vγ5-deficient mice could respond to keratinocytes in a manner similar to that of such cells derived from normal mice.

Finally, because Vγ5+ cells normally develop before other TCRγδ+ subsets (24), we used flow cytometry and PCR-RFLP to examine whether Vγ5 mutation overtly affected the development of the remaining TCRγδ+ repertoire. No gross disruption was apparent. For example, canonical Vγ6-Jγ1 rearrangements were abundant in the reproductive tract of both Vγ5+/+ and Vγ5−/−mice (16).

Our data show that an outwardly normal DETC population developed in the absence of Vγ5. Among different TCR γ and δ pairings used by Vγ5 DETCs, at least one (Vγ1-Vδ1) retained the conformational epitope defined by mAb 17D1 that is normally restricted to Vγ5+ DETCs. Thus, similar to most B cells and αβ T cells, the epidermal γδ cell subset is associated more with an antigen receptor conformation than with simple linear epitopes encoded by the particular Vγ and Vδ gene segments normally used by DETCs.

Because other TCRs can create the 17D1+ conformation, it remains to be explained why the DETC repertoire is normally dominated by a single Vγ5-Vδ1 combination. We propose that the high frequency of the canonical rearrangement of Vγ5-Jγ1 and Vδ1-Jδ2 simply reflects the most frequent mechanism used to establish the 17D1 epitope. This high frequency appears to be determined largely by short stretches of nucleotide sequence homology between the V(D)J gene segments (8,18). Indeed, canonical in-frame Vγ5-Jγ1 and Vδ1-Jδ2 rearrangements are common even in mice in which mutations prevent the expression of the TCR (8,25). Although this can be interpreted as evidence against selection on the DETC TCR, our observation that the 17D1 epitope is commonly conserved among DETCs, even in the absence of the usual TCR γδ chain pairing, suggests that the functional efficacy of this epitope may have selected for the retention of the short regions of homology, and not vice versa.

Although we have shown that the proportion of 17D1+cells in Vγ5−/− mice is similar to that in Vγ5+/+ mice, clearly there are 17D1 DETCs. Likewise, neonatal DETC repertoires of many inbred strains comprise heterogeneous γδ cells. However, over time, these regularly converge to >95% 17D1+, consistent with 17D1+ cells having a selective functional or developmental advantage in the skin. The basis for this convergence may be resolved by clarification of the ligands for 17D1+ and 17D1 DETCs. However, it cannot be assumed that the 17D1 epitope on the DETC TCR defines the conformation of the complementarity-determining region (CDR) 3, which, by extrapolation from Ig and TCRαβ, is likely to be an important contact site for antigen. Both an anti-clonotypic antibody and superantigens bind to TCRαβ conformations mapping partly or wholly outside CDR3 (26), and both can markedly and selectively activate the relevant αβ T cells in vivo. The 17D1 epitope may similarly manifest an equally important conformation of the DETC TCR.

  • * These authors contributed equally to this report.

  • To whom correspondence should be addressed. E-mail: adrian.hayday{at}


View Abstract

Navigate This Article