HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Citation Map Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Pereira, P. Articles by Boucontet, L. Search for Related Content PubMed PubMed Citation Articles by Pereira, P. Articles by Boucontet, L.

Rates of Recombination and Chain Pair Biases Greatly Influence the Primary TCR Repertoire in the Thymus of Adult Mice¹

Unité du Développement des Lymphocytes, Centre National de la Recherche Scientifique, Unité de Recherche Associée 1961, Institut Pasteur, Paris, France

	Abstract

Top Abstract Introduction Materials and Methods Results Discussion References

 
Analyses of the rearrangement status of the TCR and TCR chain loci in progenies of individual thymocytes showed a hierarchy of the different V and V gene segments to participate in a recombination reaction. Moreover, individual TCR chains only pair efficiently with a variable number of TCR chains. Interestingly, these two parameters are inversely correlated such that the TCR and TCR chains that rearrange more often show a higher level of restriction in their pairing capabilities. Our data suggest that these mechanisms, together with a natural variation affecting the expected frequencies at which rearrangement of different V gene segments give raise to functional TCR chains, have coevolved to maximize the diversity of the TCR repertoire minimizing the risk that a T cell will express more than one TCR specificity at the cell surface, despite the fact that multiple TCR rearrangements take place in the same progenitor cell.

	Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

 
Exons encoding the variable regions of lymphocyte receptors for Ag (Ig and TCR) are assembled during lymphocyte ontogeny from clusters of V, (D), and J segments in a process known as V(D)J recombination. Seven Ag receptor loci are known to undergo V(D)J recombination in vivo, ultimately generating or T lymphocytes and B lymphocytes which may express or L chains. The establishment of the primary repertoire of lymphocyte Igs and TCRs results from a combination of stochastic (probabilistic), regulated, and selective events. Stochastic events occur mainly during the exon assembling and are responsible for most of the potential diversity of the primary repertoires of Ig and TCRs. Although not completely random, the choice of the gene segments that participate in the recombination reaction, the modification of the ends of the gene segments before ligation, and the association of the proteins produced by the recombination process are examples of stochastic events. V(D)J recombination in precursor cells giving rise to T cells and B lymphocytes is also regulated on different levels including cell type specificity (e.g., completed V_H(D)J_H and V(D)J in precursors of B and T cells, respectively), order of rearrangement (e.g., D-J before V-DJ), temporal specificity (e.g., IgH before IgL or TCR before TCR), and monoallelic expression of IgH and TCR chains due in part to a feedback regulation mediated by the products of successful recombination at these loci and that shut down further rearrangement (a phenomenon usually referred to as allelic exclusion) (1, 2, 3, 4). Such mechanisms ensure that the appropriate receptor gene is expressed only in cells of the correct lineage, increase the rate of success of developing lymphocytes through the expansion of cells with one successful rearrangement, and contribute to minimize the chance that a lymphocyte express more than one Ag-specific receptor. Finally, selective events based on receptor specificities ensure that mature lymphocytes do not overly react with self-Ags and, in the case of T cells, that they recognize Ag in the context of self-MHC molecules (5).

Much less is known about the mechanisms regulating the development of the primary repertoire of TCRs expressed by mouse T cells, although some evidence indicates that they differ substantially from that of T cells and B cells. Thus, assembly of TCR and TCR V-region genes occurs nearly contemporaneously in precursor cells that give rise to both and T cells (6, 7). Consequently, simultaneous expression of functional TCR and TCR chains seems to be required for differentiation along the T cell lineage (8, 9). Another important feature to consider in the development of T cells relates to the structural complexity of the TCR locus in the mouse (see Fig. 1). Four clusters of V, J, and C regions containing seven V gene segments, four J gene segments and four C regions (hereafter referred to as V1 to V7, J1 to J4, and C1 to C4, respectively, according to the nomenclature of Heilig and Tonegawa (10)) exist in the mouse. Each cluster contains a C region, linked to a single J element and one to four V gene segments, which rearrange preferentially to the J segment present in the same cluster. The V3-J3-C3 cluster has been deleted in most TCR haplotypes and is believed to be nonfunctional in some strains carrying the TCRB haplotype (11), although it appears functional in the C57BL/6 (B6) strain (12).

View larger version (17K): [in this window] [in a new window]   FIGURE 1. Strategy to analyze the extent of TCR rearrangements in progenies of individual thymocytes. Top, Schematic representation of the genomic organization of the mouse TCR locus. The map is not drawn to scale. Arrows indicate transcriptional orientation. Bottom, Primers used for the analyses of the rearrangement status of the TCR locus in progenies of individual T cells. Primers inside a box are meant to imply that they are used together in the same PCR, whereas isolated primers denote individual PCRs. Sequences of the primers are shown in Table I.

Except the V5 and V6 genes which rearrange almost exclusively during fetal life (10, 14, 17), all other V genes are known to rearrange in the adult thymus and T cells expressing the V1, V4, or V7 chains at the cell surface can be found among thymocytes, albeit at different frequencies (20, 21, 22). Whether V-defined subsets in the adult thymus arise from different progenitors with distinct capacities to rearrange different V and V genes is not known. However, if a progenitor cell is given the chance to rearrange all their TCR genes, and in the absence of specific mechanisms preventing this, a T cell could bear as many as six to eight functional TCR chains. Given the fact that assembly of TCR genes does not exhibit properties of allelic exclusion (23), a mouse T cell could theoretically express up to 16 different TCRs at the cell surface.

To identify mechanisms controlling the development of the primary TCR repertoire expressed by adult thymocytes and to quantify the extent of allelic and isotypic inclusion in T cells, we have developed an experimental system that allows the determination of the rearrangement status of the TCR and TCR loci in progenies of individual T cells with high efficiency. The comparison of the repertoires of rearrangements at the TCR and TCR loci between cells expressing different TCR chains at the cell surface isolated from normal mice allows conclusions for the mechanisms that govern the selection of T cells bearing different TCRs into the mature T cell pool, provides a number of mechanisms that restrict the number of different TCRs that can be expressed by a single T cell and help to explain a number of paradoxes previously observed in studies analyzing the frequencies of productive TCR and TCR rearrangements in lineage cells.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

 
Animals

C57BL/6JIco (B6) mice were obtained from Iffa Credo (L’Abresle, France). B6 mice deficient in the TCR C region (TCR-KO) (18) were maintained in our animal facilities. Adult Armenian hamsters were obtained from Cytogen Research (West Roxbury, MA).

Abs, flow cytometry, and cell sorting

Anti-C (3A10), anti-C4 (2.11), anti V4 (49.2; (22), anti-V7 (F2.64; (22)), and 4B2.9 were purified from culture supernatant by affinity chromatography on protein G-Sepharose (Pharmacia, Uppsala, Sweden), and FITC-labeled or biotinylated in our laboratory by standard procedures. PE-labeled anti-C (GL3) and allophycocyanin-labeled anti-CD3 were purchased from BD Pharmingen (San Diego, CA). Cell surface mAb labeling was performed as described (24). Cells were analyzed in a FACSCalibur Cytometer (BD Biosciences, Franklin Lakes, NJ). For FACS sorting, CD4^–CD8^– thymocytes, prepared by complement-mediated lysis as described (24), were incubated with FITC-labeled anti-V Abs, PE-labeled anti-C, and allophycocyanin-labeled anti-CD3 for 30 min on ice, washed, and sorted directly into 96-well culture plates in a MoFlow (Cytometrix, Fort Collins, CO).

Generation and characterization of the 4B2.9 mAb

The 4B2.9 mAb was obtained as the 2.11 mAb (25). Briefly, Armenian hamsters were immunized i.p. three times at 3-wk intervals with 2 x 10⁶–10⁷ irradiated, V1/V6 hybridoma cells, which constitutively secrete IL-2 when cultured in vitro, resuspended in saline. Three days after the last injection, spleen cells were fused with the murine myeloma SP2/0 at a ratio of 10:1 (spleen cell-myeloma) in 1 ml of 50% polyethylene glycol as described (25). The cells were then distributed in 96-well flat-bottom plates with hypoxanthine aminopterin thymidine medium. Culture supernatants from growth-positive wells were tested for their ability to inhibit the constitutive IL-2 production of the immunizing hybridoma. Staining analyses of ⁺ hybridoma cells with known V/V use showed that the 4B2.9 mAb recognizes every V1⁺ hybridoma cell tested regardless of their V use and does not bind to hybridoma cells expressing V4, V5, or V7 chains. Furthermore, the 4B2.9 mAb does not bind to TCR-loss variants of the same V1⁺ hybridomas and SDS-PAGE analysis of immunoprecipitates from lysate of surface-labeled V1⁺ hybridoma cells revealed an apparently identical pattern of two bands for the 4B2.9 mAb and the pan--specific mAb, 3A10 (not shown). The V2 specificity of the 4B2.9 mAb is described in detail in Results.

Cell cultures

Individual thymocytes were directly sorted in 96-well round-bottom plates previously coated with anti- TCR mAbs (5–10 µg/ml) and cultured in 100 µl of RPMI 1640 with Glutamax-I medium (Invitrogen Life Technologies, Gaithersburg, MD) supplemented with sodium pyruvate, 5 x 10^–5M 2-ME, nonessential amino acids, HEPES, and antibiotics (all from Invitrogen Life Technologies), 10% FCS (Boehringer Mannheim, Mannheim, Germany), and 100 U/ml rIL-2. Cells expressing distinct V chains were activated in plates coated with the sane anti-TCR mAb used for sorting, thus providing a second control for the TCR chain expressed at the cell surface by the sorted cells. V1^–V4^–V7^{– +} thymocytes were activated in plates coated with anti-C mAb.

Single clone PCR for TCR and TCR rearrangements, cloning, and sequencing

After 1 wk of expansion, T cell clones were washed once in PBS, pelleted, frozen at –80°C, thawed, resuspended in 20 µl of PCR buffer II of the AccuPrime TaqDNA polymerase (Invitrogen Life Technologies, San Diego, CA) containing 10 µg of proteinase K (Eurobio, Paris, France) and individual clones were transferred to individual wells of a 96-well PCR plate. The plates were incubated at 56°C for 60 min and at 95°C for 20 min. Aliquots of 5 µl were used as a template for the amplifications of the TCR and TCR rearrangements present in individual clones.

The rearrangement status of the TCR locus was analyzed in a two-step PCR. In the first PCR, template DNA was amplified with a universal reverse J primer (PanJG) and a reverse primer specific for the 3' region of the V1 gene segment (VG1GLext) together with a set of forward primers recognizing the seven V genes (VG1 + 2 + 3, VG4, VG5, VG6, VG7) and a primer specific for a region upstream of the J1, J2, and J3 gene segments (JG1 + 2 + 3GL). Five microliters of DNA template were added to 10 µl of PCR buffer containing 0.5 U of AccuPrime TaqDNA polymerase and 10 pmol of each primer and a first PCR was run for 35 cycles with an initial 94°C denaturation step for 4 min and also a final extension at 72°C for 4 min. Each PCR cycle consisted of incubations at 94°C for 30 s, followed by 60°C for 30 s, and 72°C for 45 s.

Specific V-J rearrangements were assayed in seven seminested or nested PCRs, each one using the same or an internal primer specific of each individual V gene segment together with two reverse J primers (one specific for the J4 segment and the other recognizing the J1, J2, and J3 gene segments; primers JG4 and JG1 + 2 + 3, respectively), allowing the detection of all possible V-J rearrangements. Given the high level of sequence identity between the V1, V2, and V3 gene segments it was not unusual to amplify rearrangements involving any of these genes with any of the three primers in the conditions described above and the identity of the gene segments involved was obtained by sequencing the PCR products and confirmed in PCRs performed with primers recognizing exclusively the V and the J segments involved in the rearrangement (primers VG1, VG2, VG3, JG4, JG2, and JG3). Germline configuration of the J1, J2, and J3 regions or of the V1 gene segments were assessed in four nested PCRs with nested reverse primers specific for each J segment (JG1, JG2, and JG3) or for the 3' flanking region of the V1 gene segment (VG1GLint) together with forward primers specific for the upstream region of each J segment (JG1GLint, JG2GLint, and JG3GLint) or with the same V1-specific primer (VG1). All these nested or seminested PCR were run for 35 cycles using 0.25 U of EurobioTaq (Eurobio), 10pmol of each primer, 1 mM MgCl₂, and one-hundredth of each first amplification reaction as template in a final volume of 25 µl. Each PCR cycle consisted of the same incubations as described above, except that the extension time at 72°C was 30 s.

For the analysis of TCR rearrangements, the first PCR was performed in the same conditions as above except that 1 pmol of each primer (forward primers: VA2, VA10, VD2, VD4, VD5, VD6Aext, VD6Bext, DD1GLext, DD2GLext, JD1GLext, and reverse primer: JD1ext) was added. Complete or incomplete rearrangements at the TCR locus were analyzed in three nested or seminested PCRs, all containing an internal reverse primer specific for the J1 region (JD1) together with either primers VA2, VA10, VD4, and VD6Bint; VD5, DD1GLint, DD2GLint, and JD1GLint, or VD6Aext and VD2 as forward primers. These primers were designed to allow an easy identification of complete and incomplete rearrangements and of the V involved in the rearrangement by the size of the amplification products in a 2% agarose gel stained with ethidium bromide (Eurobio).

All PCR products were directly sequenced as described (26), using the specific V or V primers (for complete rearrangements) or the JD1 primer for incomplete TCR rearrangements. When two different junctions appeared in the sequence, the amplification products were cloned using the TOPO TA Cloning kit for sequencing (Invitrogen Life Technologies) following the manufacturer’s instructions. Plasmid DNA from 10 individual colonies was amplified with the same primers and sequenced as above. All PCR and sequencing reactions were performed using a GeneAmp PCR System 9700 (PerkinElmer/Cetus, Norwalk, CT). Sequences of the primers used are shown in Table I.

Real-time PCRs were performed in a GeneAmp 5700 Sequence Detection System (Applied Biosystems, Warrington, U.K.) using SYBR Green PCR master mix (Applied Biosystems) and 10 pmol of each primer, in a final volume of 25 µl according to the manufacturer’s instructions. PCR cycles were as follows: 1 cycle of 94°C for 10 min; 45 cycles of 94°C for 15 s, 60°C for 30 s, and 72°C for 1 min. A dissociation protocol was performed at the end of the last cycle to control each sample. Serial dilutions of DNA samples from TCR KO thymocytes were analyzed for hypoxanthine phosphoribosyltransferase (HPRT)³ (as a control; primers GACTGAAAGACTTGCTCGAG and CCAGCAAGCTTGCAACCTTAACCA) and for V1-J4,V2-J2, V4-J1, and V7-J1 using the same primers as above. Serial dilutions of DNA from a hybridoma cell line containing one V1-J4, two V2-J2, one V4-J1, and one V7-J1 rearrangement and one copy of the HPRT gene were amplified with the same primers and used for quantification.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

 
Experimental approach to detect all TCR rearrangements in single T cells

Individually sorted thymocytes were expanded in vitro for 5–7 days in plates coated with anti- TCR mAbs and rIL-2. In these conditions, 20–60% of the seeded T cells expanded to produce clones of 50–1000 cells. DNA from each individual clone was then used as template in a two-step PCR. The rearrangement status of the TCR locus was analyzed using a universal reverse J primer and a primer located 3' of the V1 gene segment together with a set of forward primers recognizing the seven V genes and the regions upstream of three J regions in the first step PCR (Fig. 1). Specific V-J rearrangements were assayed in seven seminested or nested PCRs, each one using the same or an internal primer for each individual V gene segment together with two reverse J primers (one specific for the J4 and the other recognizing the J1, J2, and J3 gene segments), allowing the detection of all possible V-J rearrangements (Fig. 1). Germline configuration of the J1, J2, and J3 regions or of the V1 gene segment were assessed by nested PCRs with reverse primers specific for each J gene segment and forward primers specific for the upstream region of each J gene segment or by a seminested PCR using the same V1 primer and an internal primer located 3' of the V1 gene segment, respectively (Fig. 1).

Distinct J and V gene segments rearrange at different frequencies

TCR rearrangements were analyzed in 57 thymocytes sorted according to their surface expression of the V4 chain (23 cells), the V7 chain (12 cells), or the V1 chain (22 cells) (Fig. 2). Cells expressing any of these three TCR chains represent 90% of the thymocytes in B6 mice (Ref. 20 and Fig. 2). Prior separation of the cells by the surface expression of defined V chains allowed comparisons between different subsets of thymocytes and provided an internal control for the efficiencies of the PCR amplifications because the chain that is mainly, if not exclusively, expressed at the cell surface can be unequivocally defined in cells that may contain more than one functionally rearranged TCR gene.

View larger version (47K): [in this window] [in a new window]   FIGURE 2. Rearrangement status of the four J regions in progenies of individual thymocytes. A, Schematic representation of the genomic organization of the mouse TCR locus where each V gene segment is denoted by a different color. B–D, Rearrangement status of the four J gene segments in progenies of individually sorted V4+, V1+, and V7+ thymocytes (B), V1–V4–V7– thymocytes (C), and 2.11–4B2.9+ thymocytes (D). Numbers indicate the fraction of thymocytes bearing the particular V chain in B6 mice. Each of the two boxes under every J gene segment represents one of the alleles without any intentional order. The colors denote the V gene involved in the rearrangement. Absence of a box denotes deletion of a particular allele due to recombination of V and J segments present in different clusters. Filled boxes denote productive rearrangements. Hatched boxes denote unproductive rearrangements. Empty boxes denote lack of rearrangement at that allele. Light colored boxes denote that the rearrangement status of these alleles could not be determined unambiguously.

All but two of the thymocyte clones analyzed contained multiple rearrangements at the TCR locus, ranging from two to six TCR rearrangements per cell and with a maximum frequency of three, four, and five rearrangements per cell in V7⁺, V4⁺, and V1⁺ cells, respectively. Except two V3-J4 rearrangements, all other rearrangements occurred between V and J segments located in the same cluster.

The frequencies at which each of the four J segments was found rearranged in the studied T cell clones were different. Rearrangements involving J1 were found in every T cell tested and about half of the cells (26 of 57) contained the J1 segment rearranged in both chromosomes. In V1⁺ cells, 94% (31 of 33) of the rearrangements involving J1 also involved the V4 gene segment whereas the remaining 6% (2 of 33) of V-J1 rearrangements involved the V7 gene segment. As expected, all V4⁺ and V7⁺ cells contained at least one V4-J1 or one V7-J1 rearrangement, respectively. Consistent with their almost exclusive use during fetal life (10, 14, 17), no rearrangements involving the V5 or the V6 gene were found in these analyses.

In contrast with the rearrangements involving J1, rearrangements involving J4 were present in every V1⁺ cell (as expected by their selection as V1⁺ cells) but they were rare in V4⁺ and V7⁺ cells. In fact, only 4 of the 35 V4⁺ or V7⁺ clones contained a V1-J4 rearrangement and one contained a V3-J4 rearrangement. Consequently, only 14% of the V4⁺ or V7⁺ cells have rearranged a J4 segment. This frequency was not significantly different from that of V1⁺ cells that have rearranged their J4 segment in both chromosomes (9%, or 2 of the 22 V1⁺ clones).

Rearrangements involving the J2 segment were frequent in all T cell populations. Except one V7⁺ and three V4⁺ clones that maintained both J2 segments in germline configuration and one V1⁺ clone that contained one J2 segment in germline configuration and deleted the other J2 segment, all other clones had rearranged at least one J2 segment. Furthermore, 39 of the 57 clones analyzed (68.5%) had rearranged the J2 segment in both chromosomes. All rearrangements involving J2 also involved the V2 gene segment. Finally, rearrangements involving the J3 segment were present in 12% of the T cells studied and no clone contained two J3 rearrangements.

Taken together, these data are most consistent with the existence in adult mice of a unique type of T cell progenitor which has the potential for rearranging most, if not all, V and J gene segments but in which the different V and J gene segments rearrange with different probabilities. A more precise quantification performed on the most commonly rearranging V gene segments showed that V4 and V2 rearrange roughly at similar frequencies and 10–12 and 16–20 times more often than V1 and V7 genes, respectively (Fig. 3).

View larger version (14K): [in this window] [in a new window]   FIGURE 3. Quantitative analysis of different V-J rearrangements. Serial dilutions of DNA from TCR-KO thymocytes or from a T cell hybridoma were amplified by real-time PCR with primers specific for V2 and J2, V4 and J1, V1 and J4, V7 and J1, and HPRT as described in Materials and Methods. Data are shown as the number of the indicated rearrangements in 100 cells (normalized as to 100 copies of HPRT) and represent the mean ± SD of six dilutions from one experiment representative of two.

About half of the thymocytes contain two functional TCR rearrangements, one of which involves the V2 and J2 gene segments

To assess the functionality of the rearrangements, their amplification products were sequenced. As expected, each V4⁺ cell contained a functional V4-J1 rearrangement and each V7⁺ cell contained a functional V7-J1 rearrangement. Surprisingly, one of the V1⁺ clones lacked a V1-J4 rearrangement but contained a functional V3-J4 rearrangement suggesting that the 2.11 mAb recognizes the C4 region rather than V1. None of the clones studied contained two functional rearrangements involving either the two J1 or the two J4 alleles suggesting that, at least in what concerns these J regions, most T cells are allelically excluded. Furthermore, concerning the same J segments, only 2 of the 57 clones analyzed contained two functional rearrangements (one clone contained a V1-J4 and a V4-J1, and the other clone contained a V1-J4 and a V7-J1) indicating that isotypic inclusion involving the J1 and J4 segments exist in the thymocyte population albeit at a relatively low frequency. A more precise quantification of the extent of allelic and isotypic inclusion for the functionally most important J1 and J4 rearrangements requires the analyses of a larger number of clones and, together with studies on their possible mechanisms, will be presented elsewhere.

The scenario was radically different when the functionality of the rearrangements involving the J2 and J3 gene segments was analyzed. Thus, about one-third of the V2-J2 and V3-J3 rearrangements were functional (29 of 91 V2-J2 and 3 of 7 V3-J3 rearrangements produced in-frame joints), and that independently of whether the cell expressed the V1, the V4, or the V7 chain at the cell surface. Consequently, about half of the V1⁺, V4⁺, and V7⁺ cells also contained a functionally rearranged V2 or V3 chain. These frequencies of functional V2 and V3 rearrangements are very close to the frequency expected by random rearrangement in the absence of cellular selection (assuming reading frame errors as the major source of nonfunctional rearrangements, one-third of all rearrangements are expected to produce a functional chain). This strongly suggests that the functional expression of the V2 or the V3 chains is irrelevant for the fate of the V1⁺, V4⁺, and V7⁺ thymocytes, as it has been previously shown for T cells and other T cell populations (12, 27).

Do V2 surface-positive cells exist?

The lack of cellular selection for functional V2 chains could reflect their inability to be expressed as part of a selectable TCR. However, V2⁺ hybridomas have been reported (21) and 8% of the thymocytes in B6 mice express TCR chains other than the V1, V4, or V7 chains. Given the extremely low frequency of V5 and V6 rearrangements in adult thymocytes it is likely that a large fraction of these cells express the V2 or the V3 chains. To analyze the TCR chains expressed by the V1^–V4^–V7^{– +} thymocytes and the possible constraints imposed to the expression of their functionally rearranged TCR chains, 22 such cells were cloned and their TCR rearrangements were analyzed (Fig. 2C).

Two clones contained multiple functional rearrangements involving different V gene segments and, therefore, the identity of the TCR chain expressed at the cell surface could not be unambiguously defined. Two clones contained a functional V6-J1 rearrangement and two others contained functional V3-J3 rearrangements suggesting that a minor proportion of thymocytes (1%) bear a V3 or a V6 chain as part of their TCR. The other 16 clones contained one or two functionally rearranged TCR chains involving the V2 and J2 segments, consistent with the possibility that a large fraction of the V1^–V4^–V7^– thymocytes express the V2 chain as part of the TCR.

Presence of a functionally rearranged V1, V4, or V7 chain precludes surface expression of functionally rearranged V2 chains

Because V2 chains appear to be expressed as part of a functional TCR at the cell surface of a fraction of the thymocytes, the question arises as to whether T cells containing a functional V2-J2 rearrangement together with either a functional V1-J4, V4-J1, or a V7-J1 rearrangement (which represents about one half of the V1⁺, V4⁺, and V7⁺ cells) will coexpress both TCR chains at the cell surface. This issue could only be analyzed if we had an Ab specifically recognizing the V2 chain. In an attempt to produce such Ab we obtained a hamster mAb, termed 4B2.9, which recognized virtually all T cells recognized by the anti-C4 mAb (2.11) and a small but sizeable fraction of thymocytes staining negative with the same mAb (Fig. 4A). TCR rearrangement analyses of progenies of single-sorted 2.11^–4B2.9⁺ thymocytes expanded in vitro in 4B2.9-coated plates are shown in Fig. 2D. Of the 19 clones analyzed, 17 contained a single functional rearrangement that involved the V2 and J2 segments in 16 clones and the V1 and J4 rearrangements in one clone. This later clone could represent a V1⁺ cell falling into the sorting gate possibly because its surface expression of the V1 chain was low. The remaining two clones contained a functional V2-J2 rearrangement together with a second functional rearrangement involving either the V1 and J4 segments or the V3 and J3 segments, respectively. Altogether, these data indicate that 6–8% of the thymocytes in adult B6 mice express a V2 chain at the cell surface and that these cells can be identified by their staining with the 4B2.9 mAb and their absence of reactivity to the 2.11 mAb. These data also indicate that the 4B2.9 mAb specifically recognizes T cells bearing the V1 or the V2 chains at the cell surface. Whether this Ab also recognizes the rare T cells bearing the V3 chain cannot be ascertained from these analyses, although given the high homology between the V1 and the V3 gene segments it is likely that it does so.

View larger version (15K): [in this window] [in a new window]   FIGURE 4. Surface expression of the V2 chain in thymocytes. CD4–CD8– thymocytes were stained with anti--PE, 4B2.9-biotin, CD3-allophycocyanin, and either 2.11-FITC (A and C) or anti-V4-FITC plus anti-V7-FITC (B) followed by streptavidin-PerCP and analyzed in a FACSCalibur. Data are shown as dot plots of the log10 of fluorescence intensity of the indicated mAbs in electronically gated CD3++ cells (A and B) or CD3+ cells (C). Numbers in A and B denote the fraction of cells in each quadrant.

V2 chains are expressed with a restricted number of TCR chains

The dominance of V1, V4, or V7 chains over V2 chains to be expressed at the cell surface in cells containing two functionally rearranged TCR isotypes may reflect an unequal competition of these chains for available TCR chain(s) or for CD3 components. Alternatively, this dominance may just be apparent and reflect the fact that V2 chains are only expressed at the cell surface together with a limited number of V chains. To directly test this possibility, we analyzed the TCR rearrangements present in V2⁺ cells and compare them with those present in V1⁺, V4⁺, and V7⁺ cells (Fig. 5A). Unlike these latter cells, which contain functional TCR rearrangements involving multiple V gene segments, V2⁺ cells contained functional TCR rearrangements that involved exclusively either the V5 segment (70%) or members of the V10 or V2 gene families (30%), indicating a preferential expression of the V2 chain with these V chains. Consistent with this interpretation, 70% of the V1^–V4^–V7^– thymocytes in B6 mice stained positive with a V5-specific mAb, whereas <10% of these cells were stained with a combination of V4- and V6-specific mAbs (Fig. 5B). Interestingly, previous analysis of three V2⁺ hybridomas showed that they coexpressed a functionally rearranged V5 chain (21).

View larger version (31K): [in this window] [in a new window]   FIGURE 5. The V2 chain pairs with a restricted number of V chain families. A, Representation of different V subfamilies in functional (left) and nonfunctional (right) rearrangements in progenies of individual thymocytes. Analyses of TCR rearrangements were performed by a two-step PCR as indicated in Materials and Methods. Data represent the fraction of cells containing a functional TCR chain of the indicated V subfamilies in thymocytes separated on the basis of the expression of different TCR chains. n = Number of T cell clones analyzed in each subset. B, CD4–CD8– thymocytes were stained with a mixture of anti-V1-FITC, anti-V4-FITC and anti-V7-FITC, anti--PE, CD3-allophycocyanin and the indicated biotinylated anti-V Abs followed by streptavidin-PerCP and analyzed in a FACSCalibur. Data are shown as histograms of the log10 of fluorescence intensity of the indicated V-specific mAbs in electronically gated CD3++ V1–4–7– cells. Numbers denote the fraction of positive cells.

V2 chains appears highly selective in pairing with V chains

Three different mechanisms could explain the selectivity in pairing with the V chain observed in the V2⁺ population. First, V2⁺ cells could originate from progenitor cells that preferentially rearrange the V2 and the V5 gene segments. Second, V2⁺ cells bearing V chains other than V5 or V10 could be specifically deleted. Third, the V2 chain could be highly selective in pairing with V chains.

The extent of rearrangements involving the J1 gene segment found in V2⁺ and in V2^– cells (Fig. 2) as well as the diversity of V gene segments that rearranged nonfunctionally in V2⁺ and in V2^– cells (Fig. 5) excludes the first possibility. Moreover, the fact that the frequency of functional V2-J2 rearrangements present in V2^{– +} T cells is almost identical to the random frequency of productive V2-J2 rearrangements excludes a cellular selection mechanism and indicates that some constraints imposed to a developing T cell to become a V2⁺ cell must be molecular in nature, and likely take place inside the cell before TCR-dependent selection may operate. Therefore, it is very likely that V2 chains are restricted in their ability to pair with different V chains. Indeed, experimental evidence indicating that V6 chains pair with V1 but not with V2 chains have been recently obtained in cotransfection experiments (Y.-H. Chien, personal communication).

The primary sequences of the V2-J2 junctions further restrict the cell surface expression of V2 chains

Comparison of the V5(D)J junctional sequences found in V2⁺ cells and in V2^– cells revealed no evident differences (not shown) indicating that the pairing restrictions show by V2 chains concern mainly the V gene segments and not the CDR3 of the V chain. In contrast, analyses of the V2-J2 junctional sequences revealed evident differences between theV2⁺ and the V2^– populations (Fig. 6). Thus, of 17 V2⁺ clones, 15 (88%) display functional V2-J2 junctional sequences that contained the germline-encoded methionin residue (ATG). In contrast, the same residue was only present in 6 of the 24 (25%) V2^– cells (Fig. 6) and in 16 of 60 (27%) nonfunctional V2-J2 rearrangements (not shown). Although less evident, there is also a relative depletion in V2⁺ cells of V2-J2 junctions containing a tyrosine residue (TAT; possibly formed by P nucleotide insertions at the end of the J2 segment), which is present in 3 of the 17 (18%) and in 15 of the 24 (62%) of the V2⁺ and V2^– clones, respectively (Fig. 6). The putative P nucleotide was present in 30 of the 60 (50%) nonfunctional V2-J2 rearrangements (not shown). Altogether, these data indicate another constraint that further reduces the frequency at which the product of a productive V2-J2 rearrangement will be part of a functional TCR. Considering only the presence or absence of the methionin residue in random V2-J2 rearrangements it can be estimated that about one of four functional V2-J2 rearrangements will produce a selectable V2 chain, although the actual frequency of selectable V2 chains may be lower than that.

View larger version (47K): [in this window] [in a new window]   FIGURE 6. V2-J2 junctional sequences found in V2+ and V2– thymocytes. Nucleotide (left) and amino acid (right) sequences of functional V2-J2 rearrangements from V2+ and V2– thymocytes. Numbers of the clones are as in Fig. 2. The first number in V2– thymocytes denotes the TCR chain expressed at the cell surface.

Distinct V gene segments also rearrange at different frequencies in adult T cell progenitors

Previous analyses of TCR rearrangements in T cell hybridomas have shown that 40% of the T cells contained the second TCR allele either in germline configuration or incompletely rearranged (23). Furthermore, the same study reported that T cells bearing two functional TCR rearrangements represent 30% of the T cells containing two complete TCR rearrangements. Our analyses of TCR rearrangements in progenies of individual thymocytes confirm and extend these observations. Thus, of a total of 85 T cell clones analyzed we found 30 (35.3%) containing the second TCR allele either in germline configuration or incompletely rearranged. Furthermore, of 55 of these clones in which we could clone the two TCR rearrangements, 15 (27.3%) contained two functionally rearranged TCR chains.

An estimation of the relative frequencies at which different V subfamilies rearrange in progenitor cells could also be obtained from these analyses by calculating the relative frequency of TCR rearrangements involving defined V gene subfamilies among all unproductively rearranged TCR genes (Fig. 7). These data define a hierarchy of different V gene segment subfamilies to participate in a recombination reaction similar to that previously shown for TCR gene segments.

View larger version (13K): [in this window] [in a new window]   FIGURE 7. Rearrangement frequencies of different V subfamilies. Data are shown as the frequency of nonproductive rearrangements containing different V gene subfamilies of all nonfunctional TCR rearrangements found in the analyses of V1+, V4+, and V7+ thymocytes. n = Number of total nonproductive TCR sequences.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

View this table: [in this window] [in a new window]   Table II. Opposing forces of recombination and chain pairing biases shape the TCR repertoire

Interestingly, the frequencies at which different V and V gene segments participate in a recombination reaction in adult progenitor cells are inversely correlated with the apparent ability of the resulting TCR or TCR chain to participate in the formation of a functional TCR (Table II). Thus, considering the TCR chains, V2-J2 and V4-J1 rearrangements take place in adult progenitor cells at the highest frequencies, whereas V1-J4 and V7-J1 rearrangements take place at significantly lower frequencies. Rearrangements involving the V3, V5, and V6 genes are too rare in adult progenitors to contribute significantly to the emergent repertoire of TCRs in the adult. Significantly, V2 chains show the highest level of restriction in their ability to pair with TCR chains, its expression at detectable levels at the cell surface being limited to progenitor cells that have rearrange functionally a V5 or a V10 chain. Albeit less pronounced than the V2 chains, V4 chains also appear to be restricted in their ability to pair with different TCR chains. Thus, TCR repertoire analyses of T cell hybridomas (32) and of normal thymocytes with available anti-TCR and TCR mAbs (22) have shown a virtual absence of V4V6⁺ and V4V2⁺ cells. Indirect evidence consistent with the possibility that the virtual absence of V4V6⁺ cells is due to the inefficiency of these two chains to pair is suggested by the overrepresentation of functional V6 chains in V4⁺ thymocytes that contain two functional TCR chains (of 11 such clones, 7 contained a V6 chain; our unpublished observations). In contrast, V1 and V7 chains, which are formed in progenitor cells at considerably lower frequencies, show no evident restriction in their pairing with TCR chains (22). The same inverse correlation applies to TCR chains (Table II). Thus, V6 chains, which are produced at the highest frequency in progenitor cells, will only form part of a TCR if the same precursor cells rearrange functionally either the V1 or the V7 gene segment, which is expected to occur at a relatively low frequency.

Finally, the presence of an in-frame stop codon at the 3' end of the germline V4 gene segment, which lowers substantially the frequency at which a rearrangement involving this gene segment will produce a functional chain (7, 33), also increases the chances that T cells, bearing at the cell surface TCR proteins encoded by the rearrangements of V and J segments occurring at low frequencies, will be produced in substantial numbers. This may be important if, as suggested by experiments performed in several infectious models, T cells bearing different VV TCRs perform different functions and home to different places (reviewed in Ref. 34).

Interestingly, our data provide an explanation to the long-standing paradox brought about by the fact that whereas the frequencies of productive rearrangements involving either the V1 or the V4 gene segments in double-positive thymocytes or mature T cells were significantly lower than the expected frequencies of random rearrangements (7, 13, 27, 33, 35), those involving the V2 gene segment were not different from the expected frequency of random rearrangements in similar cell populations (12, 27). A similar apparent paradox was also observed in the analyses of TCR rearrangements in -lineage cells. Thus, whereas the frequencies of productive rearrangements involving the V4 and the V5 gene segments in different populations of immature -lineage cells or mature T cells were found consistently lower than the 33% expected by random rearrangements (6, 35, 36, 37), those involving the V6 gene segment were consistently found close to the random frequency of productive rearrangements (6, 37). In light of the experiments presented here, these will be the expected results because, for the constraints outlined above, most progenitor cells containing functional V2 or V6 chains will be unable to form a functional TCR and, therefore, will not be diverted from the T cell developmental pathway. Our data are consistent with the notion that development of a T cell, and therefore diversion of the progenitor cell from the -lineage pathway, requires expression of a TCR that can be expressed at detectable levels at the cell surface. Given the number of constraints imposed to different TCR and TCR chains to be expressed as part of a TCR, it would look inappropriate to conclude on the expression of TCR and TCR proteins as part of the TCR, based solely on the presence of functional rearrangements.

The probabilities at which different V and V gene segments participate in a recombination reaction in progenitor cells, together with the constraints imposed to different TCR and TCR chains to be expressed as part of a TCR outlined above, may define the molecular basis dictating the frequencies at which different VV T cell subsets are generated in the adult thymus. However, it is expected that selective mechanisms based on TCR specificities finally shape the available repertoire of TCRs already in this organ. In this study, we have shown evidence indicating that selection of cells containing functional V2 chains into the V2⁺ T cell pool is not only restricted to progenitor cells expressing a limited repertoire of V chains but also to cells displaying restricted junctional V2-J2 sequences. Because only the CDR3 sequence is the target of this selective process, it is possibly the consequence of a cellular selection event more than that of a molecular constraint. Similarly, our data will predict that most V1⁺ cells will also bear a V6 chain, due to the fact that members of the V6 gene subfamily rearrange very often in progenitor cells and progenitor cells expressing functional V6 chains will not be diverted by the functional expression of V2-J2 or V4-J1 chains. However, previous analyses with mAbs (22) and data shown in Fig. 5 showed that the V1⁺V6⁺ and the V1⁺V4⁺ populations are comparable in size. The mechanisms responsible for this apparent loss of V1⁺V6⁺ cells in B6 mice are not known.

Analyses of TCR rearrangements in individual T cells may seem to suggest that assembly of TCR genes does not exhibit properties of allelic exclusion. This is mostly due to the fact that 7% of the T cells (4 of the 57 clones shown in Fig. 2B) contain two functional V2-J2 rearrangements. However, if allelic exclusion at the TCR locus is mediated by a functional TCR, it is expected that progenitor cells that rearrange V genes that do not pair with V2 chains may continue rearranging their TCR genes and appear as allelically included cells at the molecular level although they are allelically excluded at the level of TCR surface expression. Similarly, allelically included B lymphocytes were shown to display allelic inclusion at the level of pre-BCR surface expression (38).

	Acknowledgments

 
We thank Dr. Yueh-Hsiu Chien for sharing results with us before publication and Drs. A. Cumano and P. Vieira for discussion and critical reading of the manuscript.

	Footnotes

 
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

¹ This work was supported by institutional grants and by grants from the "Association pour la Recherche sur le Cancer".

² Address correspondence and reprint requests to Dr. Pablo Pereira, Unité du Développement des Lymphocytes, Centre National de la Recherche Scientifique, Unité de Recherche Associée 1961, Institut Pasteur, 25 Rue du Dr. Roux, 75724 Paris Cedex 15, France. E-mail address: ppereira{at}pasteur.fr

³ Abbreviations used in this paper: HPRT, hypoxanthine phosphoribosyltransferase; RSS, recombination signal sequence.

Received for publication April 16, 2004. Accepted for publication June 22, 2004.

	References

Top Abstract Introduction Materials and Methods Results Discussion References

 

1. Hesslein, D. G., D. G. Shatz. 2001. Factors and forces controlling V(D)J recombination. Adv. Immunol. 78:169.[Medline]
2. Bassing, C. H., W. Swat, F. W. Alt. 2002. The mechanism and regulation of chromosomal V(D)J recombination. Cell 109:(Suppl.):S45.
3. Khor, B., B. P. Sleckman. 2002. Allelic exclusion at the TCR locus. Curr.