Differential Regulation of Proximal and Distal Vβ Segments Upstream of a Functional VDJβ1 Rearrangement upon β-Selection

Cellular differentiation involves epigenetic changes that regulate the transcription of genes encoding lineage-specific proteins and pluripotency factors. Lymphocyte development involves similar epigenetic changes that regulate Ag receptor gene assembly. This RAG1/RAG2 (RAG) endonuclease-mediated rearrangement of TCR and Ig genes from V(D)J gene segments drives developmental progression. Despite >25 y of V(D)J recombination and lymphocyte development, the mechanisms by which these processes are regulated and coordinated with each other are not completely understood.

Mouse, and likely human, αβ and γδ T cells develop in the thymus from early thymic progenitors (1). Transcription and rearrangement of germline TCRγ, TCRδ, Dβ, and Jβ gene segments initiates in c-Kit⁺/CD25⁺ stage I double-negative (DN) thymocytes (or early thymic progenitors) and continues as these cells differentiate into CD44⁺(c-Kit⁺)/CD25⁺ stage 2 and then CD44⁻(c-Kit⁻)/CD25⁺ stage 3 DN thymocytes (2, 3). Germline Vβ transcription and Vβ rearrangement to assembled DJβ complexes occur in DN3 cells (2). Assembly and expression of functional TCRγ and TCRδ genes in DN2/3 thymocytes creates TCRγ and TCRδ chains that can pair to form γδ TCR complexes, which drive cellular proliferation and promote differentiation into γδ T cells (4, 5). In contrast, assembly and expression of a functional VDJCβ gene in DN3 cells creates TCRβ-chains that can pair with pre-Tα molecules to form pre-TCRs. These pre-TCR complexes signal β-selection through Lat-dependent intracellular pathways that inhibit Vβ rearrangement, drive cellular proliferation through Ccdn3 activation, silence Rag1/Rag2 transcription, and promote DN–to–double-positive (DP) differentiation through CD44⁻/CD25⁻ stage 4 DN and CD4⁻/CD8^low (immature single-positive [ISP]) thymocyte intermediates (4, 6, 7). Transcription and rearrangement of TCRγ and Vβ segments are predominantly silenced in DP thymocytes (8, 9), whereas transcription and rearrangement of Dβ and Jβ segments can continue in these cells (10). TCRγ silencing requires pre-TCR–driven proliferation (11); a similar role for thymocyte proliferation in the silencing of Vβ segments has not been reported. TCRα transcription and rearrangement initiates in DP cells leading to the deletion and “silencing” of intervening TCRδ gene segments (12). Assembly and expression of a functional TCRα gene generates TCRα-chains that can pair with TCRβ proteins to form αβ TCRs (4). Upon their positive selection, αβ TCRs activate intracellular signals that halt TCRα rearrangements and promote differentiation of DP cells into CD4⁺/CD8⁻ or CD4⁻/CD8⁺ (single-positive [SP]) thymocytes, which leave the thymus as CD4⁺ or CD8⁺ αβ T cells (4).

For years, developmental stage-specific regulation of germline Vβ transcriptional accessibility was thought to be the predominant mechanism that promotes Vβ rearrangements in DN cells and prevents Vβ-to-DJβ recombination in DP cells (8). This view was formed through analyses of unrearranged TCRβ alleles in DN cells of Rag-deficient mice and in DP cells generated from Rag-deficient mice expressing a TCRβ transgene or treated with anti-CD3 Abs. In these experiments, germline Vβs are generally accessible in DN thymocytes (13–17), whereas they are predominantly inaccessible in DP cells (14, 16). A caveat of these studies is that all TCRβ alleles in Rag-deficient thymocytes reside in their germline configuration, unlike the physiological conditions in normal DN and DP cells where significant fractions of TCRβ alleles contain VDJβ rearrangements that could affect germline Vβ transcription (18). For example, Vβ-to-DJβ rearrangement involving any Vβ except Vβ14 occurs through deletion of intervening sequences and moves germline Vβs closer to the TCRβ enhancer (Eβ), which may enable Eβ to increase transcription and rearrangement of these Vβ segments (19). In addition, TCRβ transgenes and anti-CD3 Abs have different effects on downregulation of germline Vβ transcription in DP cells (14–16), raising further questions regarding the physiological relevance of findings gained from one or both of these models (20).

TCRβ loci contain ∼20 functional Vβ segments and two Dβ-Jβ clusters (Dβ1-Jβ1 and Dβ2-Jβ2) with associated Cβ exons (Cβ2 and Cβ2) (21), such that Vβ-to-DJβ2 rearrangements can occur on unrearranged TCRβ alleles and alleles containing an assembled VDJCβ1 gene (22). Enforced expression of the E47 transcription factor maintains Vβ accessibility and promotes Vβ-to-DJβ2 recombination in DP thymocytes (23). Transcriptional accessibility of germline Vβ segments can be sustained upstream of functional VDJCβ genes in mature αβ T cells without their prior rearrangement in DP thymocytes (24), consistent with the notion that mechanism(s) in addition to downregulation of Vβ accessibility also prevent Vβ-to-DJβ2 rearrangements over assembled VDJCβ1 genes in DP cells (24, 25). Yet, neither of these studies could directly assess germline Vβ transcription on VDJβ1-rearranged alleles in DP cells. Our recent analysis of mice with preassembled functional TCRβ genes demonstrated that the position of germline Vβs relative to functional VDJCβ1 genes influences their transcriptional accessibility (22). Vβ10 segments rearranged to DJβ2 complexes when located just upstream of a functional Vβ1DJβ1Cβ1 gene (the Vβ1^NT allele), but not on alleles with Vβ10 located distally from a functional VDJCβ1 gene (22). Vβ4 and Vβ16 segments predominantly did not rearrange on either allele (22). Consistent with the Vβ rearrangement events observed on Vβ1^NT alleles, we found that germline Vβ10 segments were transcribed, but Vβ4 and Vβ16 segments were not transcribed in DP cells of Vβ1^NT/NT mice (22). Recombination of Vβ10, but not Vβ4 and Vβ16 segments, on Vβ1^NT alleles in DN thymocytes may be regulated through pre-TCR–mediated intracellular signaling pathways and/or TCRβ locus configuration changes directly caused by the preassembled functional Vβ1DJβ1Cβ1 gene.

Mice expressing preassembled functional VDJCβ1 genes enable investigation of mechanisms that regulate germline Vβ transcription and recombination on VDJβ1-rearranged alleles. Downregulation of germline Vβ transcription and recombination on unrearranged TCRβ alleles is linked with Lat-dependent pre-TCR–initiated signals that enforce TCRβ feedback inhibition, drive cellular proliferation through Ccnd3 (cyclin D3) expression, and promote DN-to-DP thymocyte differentiation (4, 6, 7). Ccnd3-deficient mice exhibit reduced thymic cellularity because of impaired pre-TCR–driven DN3, DN4, and ISP cell proliferation with otherwise normal DN-to-DP thymocyte differentiation (6); potential roles of Ccnd3 in feedback inhibition or regulation of germline Vβ transcription and rearrangement have not been reported. Lat-deficient mice exhibit a complete block in thymocyte development at the DN3 stage and a lack of TCRβ feedback inhibition (7). To begin to elucidate the molecular mechanisms that regulate germline Vβ transcription and Vβ-to-DJβ2 recombination on VDJβ1-rearranged alleles in DN and DP cells, we characterized the surface expression, rearrangement, and transcription of Vβ segments from Vβ1^NT alleles in wild-type (WT), Ccnd3^−/−, and Lat^−/− thymocytes.

Vβ1^NT/NT (22), Ccnd3^−/− (6), Lat^−/− (7), and Rag1^−/− mice (26) were bred among each other to generate the Vβ1^NT/NT, Vβ1^NT/+, Ccnd3^−/−, Vβ1^NT/NTCcnd3^−/−, Vβ1^NT/+Ccnd3^−/−, Lat^−/−, Vβ1^NT/+Lat^−/−, Vβ1^NT/NTLat^−/−, Vβ1^NT/NTLat^−/−Rag1^−/−, and WT mice used in this study. The background strains of these mice were mixed 129SvEv and C57BL/6. TCRβ loci exhibit the same genomic composition and lack detectable RFLPs between these two genetic backgrounds. Studies were conducted on mice between 4 and 6 wk of age. All experiments in mice were performed in accordance with relevant institutional and national guidelines and regulations, and were approved by the Children’s Hospital of Philadelphia Institutional Animal Care and Use Committee.

Vβ rearrangement was assessed by PCR. DNA was isolated from thymocytes with rapid lysis buffer (0.1 M Tris pH 8.5, 0.2% SDS, 0.005 M EDTA, 0.2 M NaCl, and 250 μg/ml proteinase K). PCRs were conducted on 500 ng DNA, and a final volume of 25 μl 10X PCR buffer (Qiagen), 0.2 mM deoxynucleotide triphosphates (ABI), 0.2 mM each primer, and 5 U Hot Star Taq polymerase (Qiagen). PCR primer sequences were previously described (22). PCR conditions were: 94°C for 3 min; 40 cycles of 94°C for 45 s, 60°C for 1 min 30 s, 72°C for 2 min 30 s; and 72°C for 10 min. The generation and analysis of αβ T cell hybridomas were performed as described previously (22).

Single-cell suspensions of thymocytes were stained with Abs in PBS containing 2% BSA and 1 mM EDTA. Abs were purchased from BD Pharmingen. CD4/CD8 stains were performed on total thymocytes using anti-CD4 (553051), anti-CD8α (553033), and anti-Cβ (553174) Abs. DN stains were performed on lineage-negative thymocytes using anti-CD25 (552880) and anti-CD117 (553356) Abs. Lineage-negative cells were identified using a mixture of anti-TCRβ (553172), anti-B220 (553090), anti-CD19 (553786), anti-CD11b (553311), anti-CD11c (557401), anti-Cδ (553178), anti-NK1.1 (553165), anti-CD8α (553033), and anti-Ter119 (553673) Abs.

For germline Vβ transcription assays, RNA was isolated from 1 × 10⁶ cells with TRIzol according to instructions from the manufacturer. Germline Vβ transcription assays were performed on total thymocytes. cDNA was synthesized with NEB Protoscript II Kit using random hexamers. Real-time PCR was carried out with the indicated primers using conditions previously described (22). Methylation analysis was carried out on total thymocytes for all mice as previously described (22), except for Vβ1^NT/NTLat^−/− and Vβ1^NT/NTLat^−/−Rag1^−/− where enriched DN cells were used. To enrich DN cells, we stained total thymocytes with anti-CD25 (BD Pharmingen) in PBS containing 2% BSA and 1 mM EDTA, washed twice in PBS with 2% BSA and 1 mM EDTA, stained with BioMag goat anti-rat beads (Qiagen), and collected with a magnetic separator. Ligation-mediated-PCR was conducted on sorted DP thymocytes as previously described (24). Primers used to amplify and detect Vβ10 and Jα61 signal ends were: Vβ10 2.4 Kb 5′-GCTCCAGACGGGTATTGTGT-3′, Vβ10 1.4 Kb 5′-CATTTCCCCGGTCTCCA-3′, Vβ10 probe 5′-GTGATGCAATGTGGATGCTC-3′, Ja61 F1 5′-TGCATGCTGCTTGAATTCTC-3′, and Jα61 F2 5′-ATTGCCCAGCAAGAAGAGAA-3′.

To evaluate the potential contribution of pre-TCR–driven cellular proliferation in regulation of germline Vβ transcription and TCRβ-mediated feedback inhibition, we generated and analyzed in parallel WT, Ccnd3^−/−, Vβ1^NT/+, Vβ1^NT/NT, Vβ1^NT/+Ccnd3^−/−, and Vβ1^NT/NTCcnd3^−/− mice. Treatment of Ccnd3^−/− mice with anti-CD3 Ab does not restore thymic cellularity (6); however, the ability of constitutive expression of a preassembled TCRβ transgene/gene to rescue pre-TCR–driven DN-to-DP thymocyte expansion in Ccnd3^−/− mice was not reported. Thus, we first characterized the thymocytes of these mice by cell counting and flow cytometric (FACS) analysis with anti-CD4 and anti-CD8 Abs. We observed reduced thymic cellularity in Ccnd3^−/− mice, as compared with WT mice, because of lower numbers of DP and SP cells (Fig. 1A). Notably, we also detected fewer numbers of DN thymocytes in Ccnd3^−/− mice as compared with WT mice (Fig. 1A). In addition, we noticed by visual observation that Ccnd3^−/− mice were ∼50% the size of Ccnd3^+/− and Ccnd3^−/− littermates (data not shown). Neither of these phenotypes was observed in the original characterization of Ccnd3^−/− mice (6). These discrepancies, which indicate that cyclin D3 contributes to cellular proliferation during embryogenesis and in early-stage thymocytes of adult mice, could be caused by differences in genetic background of the Ccnd3^−/− mice used in each study. We found similar thymic cellularity and numbers of DN, DP, and SP cells between Vβ1^NT/NT and WT mice (Fig. 1A, 1B) and between Vβ1^NT/+ and WT mice, although Vβ1^NT/NT and Vβ1^NT/+ mice exhibit an increased CD4⁺/CD8⁺ ratio (data not shown). In contrast, we observed dramatically reduced thymic cellularity in Vβ1^NT/NTCcnd3^−/− mice, as compared with Vβ1^NT/NT mice, because of lower numbers of DN, DP, and SP cells (Fig. 1A, 1B). We observed similar differences in Vβ1^NT/+Ccnd3^−/− mice as compared with Vβ1^NT/+ mice (data not shown). Collectively, our data show that constitutive expression of a preassembled functional TCRβ gene cannot overcome the impaired pre-TCR–driven proliferation and DN-to-DP expansion of Ccnd3^−/− thymocytes.

Constitutive expression of a preassembled TCRβ transgene/gene increases Ccnd3 expression and accelerates DN2-to-DN4 thymocyte differentiation (22, 27, 28). For this reason and because we observed reduced numbers of DN cells in Ccnd3^−/−, Vβ1^NT/+Ccnd3^−/−, and Vβ1^NT/NTCcnd3^−/− mice, we next evaluated the effects of cyclin D3 deficiency on the numbers of cells within each DN stage. To this aim, we conducted cell counting and FACS analysis of CD4⁻/CD8⁻ thymocytes with anti–c-Kit and anti-CD25 Abs. We observed a substantial reduction in the numbers of DN4 cells in Ccnd3^−/− mice as compared with WT mice (Fig. 1C, 1D). We also detected fewer numbers of DN1, DN2, and DN3 thymocytes in Ccnd3^−/− mice compared with WT mice (Fig. 1C, 1D). Although only the difference in DN1 cell numbers was significant, these data suggest that a requirement for cyclin D3 in proliferation of DN1 or earlier stage cells contributes to the decreased numbers of total DN thymocytes in Ccnd3^−/− mice. We found that Vβ1^NT/+ and Vβ1^NT/NT mice each exhibited lower numbers of DN3 cells and greater numbers of DN4 thymocytes compared with WT mice (Fig. 1C, 1D, data not shown), reflecting the accelerated early thymocyte development in mice expressing preassembled TCRβ transgenes/genes (22, 27, 28). We observed normal numbers of DN3 thymocytes and reduced numbers of DN4 thymocytes in Vβ1^NT/+Ccnd3^−/− and Vβ1^NT/NTCcnd3^−/− mice compared with Vβ1^NT/+ and Vβ1^NT/NT mice, respectively (Fig. 1C, 1D, data not shown); DN3 and DN4 cell numbers were comparable among Ccnd3^−/−, Vβ1^NT/+Ccnd3^−/−, and Vβ1^NT/NTCcnd3^−/− mice. Collectively, our data reveal that the ability of a preassembled functional TCRβ gene to “accelerate” DN2-to-DN4 thymocyte development by increasing the ratio of DN4/DN3 cell numbers is dependent on cyclin D3 expression.

To evaluate the contribution of pre-TCR–driven cellular proliferation in silencing germline Vβ segments, we compared the expression, rearrangement, and transcription of Vβ segments from Vβ1^NT and WT TCRβ alleles between normal and Ccnd3-deficient backgrounds. For this purpose, we first conducted FACS analysis of WT, Ccnd3^−/−, Vβ1^NT/+, Vβ1^NT/+Ccnd3^−/− Vβ1^NT/NT, and Vβ1^NT/NTCcnd3^−/− thymocytes with anti-Vβ Abs specific for particular Vβs and an anti-Cβ Ab. In their germline configurations, the Vβ1^NT allele contains the preassembled Vβ1DJβ1Cβ1 gene and unrearranged Vβ10, Vβ16, Vβ4, and Vβ14 segments, whereas the WT allele contains all Vβs unrearranged including Vβ5 and Vβ8, which are deleted on the Vβ1^NT allele (Fig. 2A); anti-Vβ Abs are available for each of these Vβs except Vβ1 and Vβ16. We found similar percentages of TCRβ⁺ cells expressing Vβ4, Vβ10, Vβ5, Vβ8, Vβ6, or Vβ14 in Ccnd3^−/− and WT thymocytes (Fig. 2B), revealing that Vβ repertoire is largely normal in the absence of Ccnd3-dependent proliferation. We detected small numbers of Vβ4⁺, Vβ10⁺, Vβ5⁺, Vβ8⁺, Vβ6⁺, and Vβ14⁺ cells in Vβ1^NT/+ thymocytes (Fig. 2C), caused by incomplete downregulation of Vβ recombination and expression from WT alleles in Vβ1^NT/+ mice. We observed similar small numbers of Vβ4⁺, Vβ10⁺, Vβ5⁺, Vβ8⁺, Vβ6⁺, and Vβ14⁺ cells in Vβ1^NT/+Ccnd3^−/− thymocytes (Fig. 2C), demonstrating that loss of pre-TCR–signaled cellular proliferation does not increase the frequency at which the expression of Vβs from WT alleles is suppressed in Vβ1^NT/+ thymocytes. We detected only Vβ10⁺ cells above background levels in Vβ1^NT/NT thymocytes (Fig. 2C), indicating that, other than the prerearranged Vβ1 segment, Vβ10 is the only Vβ segment expressed from Vβ1^NT alleles at detectable levels in thymocytes. We found a similar small number of Vβ10⁺ cells in Vβ1^NT/NTCcnd3^−/− thymocytes (Fig. 2D), revealing that loss of Ccnd3-dependent proliferation does not increase the repertoire of Vβ segments expressed from Vβ1^NT alleles. Notably, we did observe lower numbers of Vβ10⁺ cells in Vβ1^NT/NTCcnd3^−/− mice compared with Vβ1^NT/NT mice (Fig. 2D), possibly because of increased reliance of the Vβ10⁺ population on cyclin D3 for expansion. Our FACS data demonstrate that expression of Vβ segments is silenced normally in the absence of Ccnd3-dependent proliferation, at least in mice expressing the Vβ1^NT allele.

We previously demonstrated that postrecombination mechanisms silence the expression of TCRβ genes formed by Vβ-to-DJβ rearrangements that escape pre-TCR–mediated feedback inhibition (29). Therefore, to directly evaluate the contribution of Ccnd3-dependent thymocyte proliferation in TCRβ feedback inhibition of Vβ rearrangements, we next conducted PCR on genomic DNA isolated from total thymocytes of WT, Ccnd3^−/−, Vβ1^NT/+, Vβ1^NT/+Ccnd3^−/−, Vβ1^NT/NT, and Vβ1^NT/NTCcnd3^−/− mice to amplify Vβ-to-DJβ rearrangements. Primers specific for each Vβ can be used individually with primers that anneal downstream of Jβ1.2 or Jβ2.2 to amplify potential Vβ-to-DJβ1 or Vβ-to-DJβ2 rearrangements on either TCRβ allele. Through this assay, we found similar levels and repertoire of Vβ-to-DJβ rearrangements on Vβ1^NT and WT alleles between Ccnd3-sufficient and Ccnd3-deficient backgrounds (Fig. 3A), showing that regulation of Vβ rearrangements is not detectably altered by the absence of cyclin D3. These PCR data demonstrate that feedback inhibition of Vβ rearrangements occurs normally in the absence of Ccnd3-dependent proliferation in mice expressing the Vβ1^NT allele.

Developmental stage-specific TCRγ transcriptional silencing in DP cells is dependent on pre-TCR–driven proliferation during the cellular expansion associated with DN-to-DP differentiation (11). For this reason and because mechanism(s) other than downregulation of Vβ accessibility in DP cells help maintain TCRβ feedback inhibition (24, 25), we also wanted to determine whether pre–TCR-driven proliferation is required for transcriptional silencing of germline Vβs in DP cells. For this purpose, we conducted quantitative PCR (qPCR) on RNA isolated from total thymocytes of Vβ1^NT/+, Vβ1^NT/+Ccnd3^−/−, Vβ1^NT/NT, and Vβ1^NT/NTCcnd3^−/− mice to quantify steady-state transcripts of germline Vβ segments and the rearranged Vβ1DJβ1Cβ1 gene. The use of Rag-deficient backgrounds to block Vβ-to-DJβ rearrangements that might reduce detection of germline Vβ transcripts was not necessary because almost no germline Vβs recombined on Vβ1^NT or WT alleles in these cells. We detected Vβ1DJβ1Cβ1 transcripts and germline Vβ transcripts from only Vβ10 segments in Vβ1^NT/+ and Vβ1^NT/NT thymocytes (Fig. 3B, 3C). We were unable to detect germline Vβ transcripts from any other Vβs in Vβ1^NT/+Ccnd3^−/− and Vβ1^NT/NTCcnd3^−/− thymocytes (Fig. 3B) or in total thymocytes of WT mice (data not shown). These qPCR data demonstrate that transcriptional silencing of germline Vβ segments on DN-to-DP thymocyte differentiation is normal in the absence of Ccnd3-dependent proliferation.

In addition to driving Ccnd3-dependent cellular proliferation, the assembly and expression of functional TCRβ genes inhibit further Vβ-to-DJβ rearrangements in DN3 cells and promote DN3-to-DN4–to–ISP-to-DP thymocyte differentiation (4, 6–8). The study of Lat^−/− mice and Lat^−/− mice expressing a functional TCRβ transgene showed that the Lat adaptor protein is essential for transducing pre-TCR signals that inhibit Vβ rearrangements, drive proliferation, and promote differentiation (7, 30). Although these experiments proved that Lat-dependent pre-TCR signals inhibit primary Vβ rearrangements on unrearranged TCRβ alleles, they could not address whether the same signals also suppress secondary Vβ-to-DJβ2 rearrangements on alleles that had previously assembled VDJCβ1 genes.

To elucidate the contributions of Lat-dependent inhibition of Vβ-to-DJβ2 recombination on VDJβ1-rearranged alleles, we generated and analyzed in parallel WT, Lat^−/−, Vβ1^NT/+, Vβ1^NT/NT, Vβ1^NT/+Lat^−/−, and Vβ1^NT/NTLat^−/− mice. We first characterized the thymocytes of these mice by cell counting and FACS analysis with anti-CD4 and anti-CD8 or anti–c-kit and anti-CD25 Abs. We observed reduced thymic cellularity caused by a block in thymocyte development at the DN3 stage in Lat^−/−, Vβ1^NT/+Lat^−/−, and Vβ1^NT/NTLat^−/− mice as compared with WT, Vβ1^NT/+, and Vβ1^NT/NT mice (Fig. 4A). As is the case for TCRβ transgenic Lat^−/− mice (30), our data indicate that expression of a preassembled functional VDJCβ gene cannot signal DN-to-DP thymocyte differentiation and expansion in the absence of the Lat adaptor protein. We next conducted PCR on genomic DNA isolated from WT, Lat^−/−, Vβ1^NT/+, Vβ1^NT/NT, Vβ1^NT/+Lat^−/−, and Vβ1^NT/NTLat^−/− thymocytes to amplify potential Vβ-to-DJβ recombination events on VDJβ1-rearranged and unrearranged TCRβ alleles. We observed higher levels of Vβ-to-DJβ1 and Vβ-to-DJβ2 recombination involving Vβ16, Vβ10, Vβ8, and Vβ4 in Vβ1^NT/+Lat^−/− cells as compared with Vβ1^NT/+ cells (Fig. 4B). Because Vβ-to-DJβ1 recombination involving these Vβs and Vβ-to-DJβ2 recombination involving Vβ8 can occur only on WT alleles, these data confirm, as expected, that Lat-dependent signals are required for feedback inhibition of Vβ recombination on unrearranged alleles in DN thymocytes of mice expressing a preassembled functional TCRβ gene on the other allele. We also observed ∼15-fold greater levels of Vβ-to-DJβ2 rearrangements involving Vβ4 and Vβ16, and ∼2-fold greater levels of Vβ-to-DJβ2 rearrangements of Vβ10 segments in Vβ1^NT/NTLat^−/− DN cells as compared with Vβ1^NT/NT thymocytes (Fig. 4B, data not shown). These PCR data demonstrate that Lat-dependent signals inhibit Vβ4 and Vβ16 rearrangements over the functional VDJCβ1 gene on Vβ1^NT alleles in DN thymocytes.

In mice and humans, germline Vβ4 and Vβ16 segments reside within a Vβ subcluster that is separated from the germline Vβ10 segment by 13.5 kb of genomic sequences containing transposons and other repetitive DNA elements (Fig. 5A) (21). On the Vβ1^NT allele, the germline Vβ10 segment resides 2.5 kb upstream of the preassembled Vβ1DJβ1Cβ1 gene (Fig. 5A). We previously showed that the Vβ4/Vβ16 subcluster is transcriptionally inaccessible and Vβ10 is transcriptionally accessible in total (DP) thymocytes of Vβ1^NT/NT mice (22). To evaluate the potential role of Lat-dependent regulation of Vβ transcriptional accessibility in controlling Vβ4 and Vβ16 rearrangement on the Vβ1^NT allele, we first analyzed the steady-state levels of germline Vβ transcripts and rearranged Vβ1DJβ1Cβ1 transcripts from these alleles in Lat^−/− DN thymocytes. Because we observed significant levels of Vβ-to-DJβ2 rearrangements in Vβ1^NT/NTLat^−/− DN cells (Fig. 4B), we generated and used Vβ1^NT/NTLat^−/−Rag1^−/− mice for these and subsequent experiments to avoid deletion of germline Vβ segments. We detected similar levels of germline Vβ transcripts from Vβ4, Vβ16, and Vβ10 segments in Vβ1^NT/NTLat^−/−Rag1^−/− DN thymocytes (Fig. 5B), reflecting the relative similar levels of Vβ-to-DJβ2 rearrangements over the VDJCβ1 gene involving these Vβs in Vβ1^NT/NTLat^−/− DN cells (Fig. 4B). However, in Vβ1^NT/NTRag1^−/− total (DP) thymocytes, we observed similar levels of germline Vβ4 and Vβ16 transcripts, but a 4-fold greater level of germline Vβ10 transcripts (Fig. 5B). We also detected a 3-fold greater level of rearranged Vβ1DJβ1Cβ1 transcripts in Vβ1^NT/NTRag1^−/− total (DP) thymocytes as compared with Vβ1^NT/NTLat^−/−Rag1^−/− DN cells (Fig. 5B). Together, these data reveal that Lat-dependent signals associated with DN-to-DP thymocyte differentiation upregulate transcription of germline Vβ10 segments and preassembled Vβ1DJβ1Cβ1 genes, but not germline Vβ4 and Vβ16 segments, on Vβ1^NT alleles. Despite the higher level of germline Vβ10 transcripts from Vβ1^NT alleles in DP thymocytes as compared with DN cells, we were not able to detect RAG cleavage of Vβ10 segments above background in DP cells (Fig. 5C), reinforcing the notion that mechanism(s) other than downregulation of Vβ transcriptional accessibility suppress Vβ rearrangements over assembled functional VDJCβ1 genes in DP thymocytes (24, 25).

The extent of CpG methylation of TCRβ locus gene segments provides a strict correlation with their transcriptional activity (13, 31–37). Because silencing of germline Vβ10 transcription on Vβ1^NT alleles during differentiation of DP cells into mature αβ T cells correlates with increased CpG methylation (22), we investigated whether Lat-dependent decreases in Vβ10 CpG methylation mirrors the Lat-dependent upregulation of germline Vβ10 transcription during DN-to-DP thymocyte development. Unfortunately, the high density of repetitive DNA sequences within this region of the TCRβ locus restricts analysis of CpG methylation to four sites: two within Vβ10 and two upstream of Vβ10 (22). We conducted qPCR of genomic DNA digested with methyl-sensitive restriction enzymes to quantify the amount of CpG methylation at sites within the Vβ4/Vβ16-Vβ10 intergenic region (I1 and I2), the Vβ10 promoter (P), or the Vβ10 coding region (C) (Fig. 5A) in Vβ1^NT/NTLat^−/−Rag1^−/− DN cells and Vβ1^NT/NTRag1^−/− total (DP) thymocytes. We detected similar percentages of CpG methylation at sites I1 and I2 between DN and DP cells (Fig. 5D). In contrast, we found a lower percentage of CpG methylation at sites P and C in DP cells compared with DN thymocytes (Fig. 5D). These data reveal that Lat-dependent signals decrease CpG methylation at sites C and P over Vβ10 segments on Vβ1^NT alleles (Fig. 6), suggesting a causative link between developmental stage-specific regulation of Vβ10 CpG methylation and germline transcription.

In this study, we investigated the contribution of distinct β-selection signals on regulation of Vβ-to-DJβ2 rearrangement and germline Vβ transcription on an endogenous TCRβ allele containing a preassembled functional Vβ1DJCβ1 gene. Germline Vβ10 segments located immediately upstream of this VDJCβ1 gene rearrange and are expressed in thymocytes, whereas germline Vβ4 and Vβ16 segments residing further upstream predominantly are not rearranged or expressed (22). We show in this article that the silencing of Vβ4 and Vβ16 recombination on Vβ1^NT alleles requires Lat-dependent β-selection signals other than Ccnd3-driven thymocyte proliferation. Mice expressing TCRβ transgenes/genes bypass the requirement to assemble functional VDJCβ genes in DN2/DN3 thymocytes, leading to higher numbers of DN4 cells (22, 27, 28). This “accelerated” early thymocyte development has been hypothesized to inhibit Vβ-to-DJβ recombination by preventing the normal accessibility of germline Vβ segments (38). We have shown that Ccnd3^−/− mice exhibit decreased numbers of DN4 cells, and expression of the preassembled Vβ1DJCβ1 gene cannot overcome this impaired DN3-to-DN4 thymocyte transition, such that Vβ1^NT/NTCcnd3^−/− mice exhibit a similar number of DN3 cells as Ccnd3^−/− mice. Together, these findings provide unequivocal proof that Ccnd3-dependent cellular proliferation is not required for the silencing of Vβ4 and Vβ16 recombination on Vβ1^NT alleles after β-selection. Consequently, our data demonstrate that the inhibition of Vβ4 and Vβ16 recombination in Vβ1^NT/NT thymocytes cannot simply be due to decreased Vβ accessibility caused by diminished cellular residence within the DN3 stage.

Our observations that the transcription, rearrangement, and expression of germline Vβ segments are suppressed on unrearranged WT alleles in Vβ1^NT/+Ccnd3^−/− mice has important implications for two long-standing hypotheses regarding mechanisms that control TCRβ feedback inhibition. Ag receptor genes are assembled in the G₀/G₁ phase of developing lymphocytes due, at least in part, to cell cycle-restricted expression of the RAG2 protein (39). Cyclin D proteins complex with cyclin-dependent kinases to drive cells from G₀ into G₁, through early/mid G₁, and into late G₁ phase when other cyclin/cyclin-dependent kinase complexes promote the G₁/S transition (40). It has been proposed that TCRβ feedback inhibition is affected by pre-TCR signals that drive DN3 cells through G₁ phase and into S phase, in part to downregulate RAG activity (41, 42). TCRβ allelic exclusion appears normal in mice expressing mutant Rag2 protein that cannot be inactivated at the G₁/S transition (36). However, no analyses of TCRβ feedback inhibition were conducted to evaluate potential compensation by posttranscriptional mechanisms that contribute to enforcing TCRβ allelic exclusion. In addition, it also has been suggested that cellular progression through multiple cell cycles during DN-to-DP differentiation is necessary to promote chromatin changes that downregulate Vβ accessibility and inhibit Vβ rearrangements in DP cells (41, 42). Our data reveal that Ccnd3-driven G₁/S transition of DN3 cells on β-selection and their subsequent progression through multiple cell cycles during DN-to-DP thymocyte differentiation is not required for TCRβ feedback inhibition, at least in mice expressing a preassembled VDJCβ gene.

Our analysis of Vβ1^NT/NTLat^−/− and Vβ1^NT/NTLat^−/−Rag1^−/− mice unmasked a role of β-selection signals in suppressing the rearrangement of distal Vβ segments upstream of the VDJCβ1 gene. Our finding that Vβ4 and Vβ16 segments recombine at similar levels as Vβ10 segments in Vβ1^NT/NTLat^−/− DN cells indicates that expression of TCRβ-chains from the preassembled VDJCβ1 gene prevents Vβ4 and Vβ16 rearrangements in Vβ1^NT/NT thymocytes, rather than intrinsic properties of the Vβ1^NT allele. Because Vβ10 is the predominant Vβ that recombines in Vβ1^NT/NT and Vβ1^NT/NTCcnd3^−/− thymocytes (22), this observation also reveals that pre-TCR signals other than Ccnd3-driven thymocyte proliferation block Vβ4 and Vβ16 recombination on Vβ1^NT alleles. However, our data cannot distinguish between pre-TCR signals that block Vβ4 and Vβ16 rearrangements by impacting the TCRβ locus directly in DN thymocytes versus indirectly during DN-to-DP differentiation (30, 38). For example, pre-TCR signals degrade E47 protein in DN3 cells and could render Vβ4 and Vβ16 inaccessible before Ccnd3-mediated G₁/S progression and DN3-to-DN4 differentiation. Alternatively, transcriptional silencing of E47 during pre-TCR–signaled DN-to-DP differentiation could render Vβ4 and Vβ16 recombinationally inaccessible in DP cells. The differential rearrangement of Vβ4 and Vβ16 between Vβ1^NT/NTLat^−/− and Vβ1^NT/NTCcnd3^−/− thymocytes reveals that DN residence can influence the repertoire of Vβ usage on alleles containing a functional VDJCβ1 gene. In this context, we would predict equivalent frequencies of Vβ4, Vβ16, and Vβ10 rearrangement on an allele containing a nonfunctional Vβ1DJCβ1 gene. One question that arises from our data is that if Vβ4, Vβ16, and Vβ10 are transcribed at similar levels on Vβ1^NT alleles in DN3 thymocytes, why does only Vβ10 rearrange at a significant level? The relative frequencies at which Vβ segments rearrange to DJβ2 complexes theoretically could be influenced by many other factors including effects of chromosomal distances and surrounding genomic sequences on juxtaposition of Vβs with DJβ2 complexes, and influences of Vβ recombination signal sequences and flanking sequences on RAG binding and cleavage. Generation and analysis of mice containing Vβ1^NT alleles with gene-targeted deletion or replacement of Vβ4/Vβ16-Vβ10 intergenic sequences and/or Vβ10 and Vβ4/Vβ16 sequences is required to answer this question.

This study, together with our prior analysis of Vβ1^NT/NT mice, reveals that developmental stage-specific regulation of germline Vβ transcription is more dynamic than previously appreciated in the field. This general view has been that Vβ segments on unrearranged alleles are transcribed in DN cells and silenced in DP thymocytes and αβ T cells, whereas germline Vβs upstream of functional VDJCβ genes can remain transcribed within these later developmental stages (8). Our analyses demonstrate that germline Vβs upstream of a VDJβ-rearranged allele are transcribed at similar levels in DN thymocytes (Fig. 6). However, Lat-dependent signals associated with DN-to-DP thymocyte differentiation increase the transcription of preassembled Vβ1DJβ1Cβ1 genes and immediately upstream germline Vβ10 segments on Vβ1^NT alleles, but not germline Vβ4 and Vβ16 segments further upstream (Fig. 6). This upregulation of germline Vβ10 transcription correlates with downregulation of Vβ10 CpG methylation (Fig. 6), suggesting a possible causative link. In this context, CpG methylation may function to restrict the level of Vβ10 transcription and rearrangement in DN cells when factors that promote juxtaposition of Vβ10 with DJβ2 complexes are expressed, but it may not be needed to regulate Vβ10 rearrangements in DP thymocytes when such factors are not expressed (20, 25). Vβ promoters drive transcription independent of Eβ in DN thymocytes (13) but require Eβ to maintain expression of VDJCβ genes in DP cells (43). In this context, a consequence of functional interactions between Eβ and the Vβ1 promoter on Vβ1^NT alleles in DP cells may be sustained transcription of Vβ10 segments located ∼1 kb upstream of the Vβ1 promoter. The greater distance of germline Vβ4 and Vβ16 segments from the Vβ1 promoter may prevent their regulation by Eβ on Vβ1^NT alleles in DP cells. Alternatively, the Vβ4/Vβ16-Vβ10 intergenic region that exhibits high CpG methylation throughout αβ T cell development (Fig. 6) may function as a boundary element after Vβ-to-DJβ rearrangement to prevent Eβ from activating the Vβ4 and Vβ16 promoters. In addition, repetitive DNA elements within the Vβ4/Vβ16-Vβ10 intergenic region may target enzymes that nucleate and promote the spread of transcriptional repressive heterochromatin across Vβ4 and Vβ16. Dominant effects of interactions between Eβ and the Vβ1 promoter might prevent the spread of heterochromatin over Vβ10 segments immediately upstream of the functional Vβ1DJCβ1 gene. Silencing of distal Vβ segments upstream of a functional VDJβ1 gene in DP cells might function with other mechanisms to prevent Vβ-to-DJβ2 rearrangements during αβ TCR selection. Considering that Rag1/Rag2 can be re-expressed and direct Vβ-to-DJβ2 rearrangements in peripheral αβ T cells recognizing self-Ags (44), silencing of all germline Vβ segments upstream of a functional VDJβ1 gene mature αβ cells might contribute to the regulation of Vβ rearrangements during TCRβ receptor revision to help maintain self-tolerance.

The authors have no financial conflicts of interest.