Diversification of the TCR β Locus V_β Repertoire by CTCF

Abstract

The architectural protein CTCF regulates the conformation and recombination of antigen receptor loci. To study the importance of CTCF in Tcrb locus repertoire formation, we created a conditional knockout mouse line that deletes Ctcf early during thymocyte development. We observed an incomplete block in thymocyte development at the double-negative to double-positive transition, resulting in greatly lowered thymic cellularity. The Tcrb repertoire was altered with a decrease in recombination of V_β gene segments in close proximity to a CTCF binding element (CBE), resulting in an overall repertoire that was skewed in favor of V_β gene segments with no nearby CBE. Therefore we show that CTCF functions to diversify the Tcrb repertoire.

INTRODUCTION

The diversity of antigen receptors on T and B lymphocytes is generated by the process of V(D)J recombination, whereby V, D and J gene segments undergo somatic recombination to create antigen receptors of unique specificity. V(D)J recombination is dependent on the activity of the RAG1 and RAG2 proteins, collectively referred to as RAG. RAG-mediated recombination is tightly controlled by a variety of mechanisms, including RAG expression, quality and accessibility of recombination signal sequences, recruitment of RAG proteins, and the subnuclear localization and conformation of antigen receptor loci (1, 2).

CCCTC-binding factor (CTCF)³ is a ubiquitously expressed zinc finger transcription factor with many known functions in regulating chromatin architecture. CTCF was initially characterized as an insulator that blocked promoter-enhancer interactions (3). Numerous subsequent studies have elucidated the role of CTCF as a transcriptional activator, repressor, looping factor and domain organizer (4, 5). CTCF binds to large numbers of sites across antigen receptor loci and plays a critical role in establishing the loop organization of these loci (6–13). Moreover, CTCF is an important regulator of antigen receptor recombination. Numerous studies have characterized the roles of intergenic CTCF binding elements (CBEs) located between antigen receptor locus V gene segments and recombination centers (RCs), containing D and J gene segments. Deletion of intergenic CBEs in the Igh locus (8, 14), the Igk locus (15, 16), and the Tcra/Tcrd locus (11) caused changes in chromatin loop organization, resulting in repertoires that were strongly biased towards the use of RC-proximal V gene segments. However, most CBEs are located in the array of V gene segments (6, 9, 10, 13, 17), and these have not been extensively studied. In this regard, two recent reports demonstrated that CBEs associated with RC-proximal V_H gene segments facilitate rearrangement of those V_H gene segments (18, 19). Only two studies have analyzed the effects of CTCF depletion on antigen receptor locus recombination in vivo by conditional knockout strategies. In the Tcra locus, the deletion of Ctcf resulted in decreased V_α-to-J_α recombination (10). In the Igk locus, Ctcf deletion biased the repertoire towards use of RC-proximal V_κ gene segments, similar to the phenotype observed upon deletion of intergenic CBEs (9).

The Tcrb locus contains CBEs associated with half of the 24 rearranging V_β gene segments (17) as well as three intergenic CBEs located upstream of the DJC_β RC. One of the intergenic CBEs, termed 5’PC, was shown to tether the Tcrb RC to RC-distal V_β gene segments. The other two CBEs, just upstream of Trbd1, act as a barrier that blocks the spread of active chromatin marks from the RC towards 5’PC. Loss of 5’PC function results in diminished use of several RC-distal V_β gene segments and increased use of an RC-proximal V_β gene segment (12). Absence of all three CBEs disrupts Tcrb locus interaction with the nuclear lamina and strongly biases the V_β repertoire towards RC-proximal V_β usage (20).

The factors that influence the preselection V_β repertoire have been evaluated computationally and assembled into a unifying predictive model (17). In this two-step model, a combination of recombination signal sequence quality and a series of parameters associated with active, accessible chromatin were highly predictive of V_β gene segment usage. The presence of a CBE proximal to a V_β gene segment was one factor predictive of increased representation in the V_β repertoire.

Here we created a Ctcf conditional knockout mouse model to experimentally assess the importance of CTCF in formation of the Tcrb repertoire. We observed a partial block in thymocyte development that resulted in greatly reduced thymic cellularity. Although CTCF protein levels were only partially reduced in CD4⁻CD8⁻ double negative (DN) thymocytes undergoing Tcrb recombination, the Tcrb repertoire was measurably altered, such that V_β gene segments possessing a nearby CBE were underrepresented and V_β gene segments lacking a nearby CBE were overrepresented. Thus, we provide experimental evidence in support of prior modeling, and demonstrate that CTCF plays a significant role in shaping and diversifying the Tcrb repertoire.

MATERIALS AND METHODS

Mice

Ctcf ^f/fCd2-Cre mice were generated by crossing Ctcf ^f/fLck-Cre mice (10) with Yy1^f/fCd2-Cre mice (21) followed by intercrossing to obtain the desired genotype. Ctcf ^f/fCd2-Cre mice were also bred with Rag2^−/− mice to generate Rag2^−/−Ctcf ^f/fCd2-Cre mice. The genetic backgrounds of mice were a mixture of C57BL/6 and 129. Mice were housed in a specific-pathogen-free facility managed by the Duke University Division of Laboratory Animal Resources. Mice of both sexes were included in all experiments; no differences on the basis of sex were noted. Mice were generally sacrificed at 4 weeks of age. All mice were handled in accordance with protocols approved by the Duke University Institutional Animal Care and Use Committee.

Flow cytometry and cell sorting

The following monoclonal antibodies and reagents from Biolegend were used for staining: B220 (RA3–6B2), CD4 (GK1.5), CD8α (53–6.7), CD11b (M1/70), CD11c (N418), CD25 (3C7), CD44 (IM7), Gr-1 (RB6–8C5), Ter-119 (TER-119) and 7-amino-actinomycin D (7AAD). Fluorescein di-β-D-galactopyranoside (FDG) staining was performed using the FluoReporter LacZ Flow Cytometry Kit (Life Technologies). Flow cytometric data acquisition was performed on a FACSCanto II (Becton Dickinson). Cell sorting was performed on an Astrios (Beckman Coulter), MoFlo XDP (Beckman Coulter) or FACSDiVa (Becton Dickinson).

Ctcf deletion efficiency

Genomic DNA (gDNA) from DN3 thymocytes was purified by phenol-chloroform extraction and isopropanol precipitation. Quantitative PCR (qPCR) analysis was performed using the QuantiFast SYBR Green PCR kit (Qiagen) on a Roche Lightcycler 480 using the following program: 5 min at 95°C, followed by 45 cycles of 10 s at 95°C and 30 s at 62°C. Experimental values for Ctcf were normalized to Cd14. Ctcf primers were 5’-CTAGGAGTGTAGTTCAGTGAGGCC-3’ and 5’-GCTCTAAAGAAGGTTGTGAGTTC-3’. Cd14 primers were 5’-GCTCAAACTTTCAGAATCTACCGAC-3’ and 5’- AGTCAGTTCCTGGAGGCCGGAAATC-3’.

Western blot

10⁶ DN3 thymocytes were lysed in 50 μL of 25 mM Tris-HCl pH 8.0, 150 mM NaCl, 1% (v/v) Triton X-100, 0.5% (w/v) sodium deoxycholate, 0.1% (w/v) SDS, and protease inhibitor cocktail (Sigma-Aldrich) for 10 min at 4°C. After centrifugation at 18000 g for 10 min at 4°C, the supernatant was transferred to a new tube and 50 μL of 0.125 M Tris-HCl pH 6.8, 20% (v/v) glycerol, 10% (v/v) 2-mercaptoethanol (Sigma-Aldrich), 4% (w/v) SDS and 0.004% (w/v) bromophenol blue (Sigma-Aldrich) was added. The sample was boiled at 95°C for 5 min and subsequently centrifuged. 10 μL of sample was loaded into an SDS-PAGE gel for gel electrophoresis in a running buffer containing 25 mM Tris-HCl pH 6.8, 190 mM glycine, 0.1 (w/v) SDS. The proteins were transferred onto a nitrocellulose membrane in buffer containing 25 mM Tris-HCl pH 6.8, 190 mM glycine, 20% (v/v) methanol using a Trans-Blot Semi-Dry Transfer Cell (Bio-Rad) running at 20 V for 1 h. Blocking was performed by rocking the membrane in 2% (w/v) fish gelatin dissolved in TBST buffer (20 mM Tris-HCl, pH 7.5, 150 mM NaCl, 0.1% (v/v) Tween 20) for 1 h at 20°C. Primary antibody staining was performed by rocking the membrane with CTCF (Millipore 07–729) or LAT (Cell Signaling 9166)-specific rabbit antibodies dissolved in TBST buffer at 1:1000 dilution overnight at 4°C. After washing in TBST buffer, secondary antibody staining was performed by rocking the membrane with goat anti-rabbit antibody conjugated with Alexa Fluor 680 (Invitrogen A21076) at 1:2000 dilution for 1 h at 20°C. After washing with TBST buffer, the membrane was imaged on a LI-COR Odyssey (LI-COR).

Chromatin immunoprecipitation (ChIP)

10⁶ DN3 thymocytes were crosslinked in 1 ml of RPMI 1640 (Gibco) containing 10% (v/v) fetal bovine serum (Gemini Bio-Products) and 1% (w/v) paraformaldehyde (Electron Microscopy Sciences) for 10 min at 20°C. Crosslinking was terminated by the addition of 100 μL 1.25M glycine (Sigma-Aldrich). Fixed cells were pelleted by centrifugation and cytoplasm removed by incubation in 1 mL 5 mM PIPES pH 8.0, 85 mM KCl, 0.5% (v/v) NP-40 for 10 min on ice. Nuclei were pelleted by centrifugation, lysed in 100 μL 50 mM Tris-HCl pH 8.0, 10 mM EDTA, 1% (w/v) SDS, and the volume was increased to 1 mL by addition of 900 μL 10 mM Tris-HCl pH 8.0, 10 mM EDTA. Chromatin was sheared with a Sonicator 3000 (Qsonica) at a power level of 2 for 10 cycles of 15 s on and 30 s off. After centrifugation at 18000 g for 10 min at 4°C, the supernatant was diluted to 1.5 mL with a final concentration of 16.7 mM Tris-HCl pH 8.0, 167 mM NaCl, 1.2 mM EDTA, 1.1% (v/v) Triton X-100, 0.03% (w/v) SDS. Pre-clearing was performed by addition of 100 μl protein A agarose/salmon sperm DNA slurry (Millipore) and mixing for 3 h at 4°C. 500 μL of pre-cleared supernatant was incubated overnight at 4°C with 5 μg of antibody specific for CTCF (Millipore) or control IgG (R&D Systems). Pulldown was performed by adding 75 μl protein A agarose/salmon sperm DNA slurry (Millipore) and mixing for 45 min at 4°C. The slurry was washed twice with 1 mL of each of the following wash buffers: 167 mM NaCl, 16.7 mM Tris-HCl pH 8.0, 1.2 mM EDTA, 1.1% (v/v) Triton X-100, 0.01% (w/v) SDS; 500 mM NaCl, 20 mM Tris-HCl pH 8.0, 2 mM EDTA, 1% (v/v) Triton X-100, 0.1% (w/v) SDS; 100 mM Tris-HCl pH 8.0, 0.5 M LiCl, 1% (v/v) NP-40, 1% (w/v) sodium deoxycholate; 10 mM Tris-HCl pH 8.0, 1 mM EDTA. Elution was performed by resuspending the slurry in 250 μl of 50 mM NaHCO₃, 1% (w/v) SDS and rocking for 15 min at 20°C. Elution was performed twice and eluates combined. Crosslinks were reversed by addition of NaCl to a final concentration of 200 mM and overnight incubation at 65°C. The eluate was treated with RNAse A (Sigma-Aldrich) at a concentration of 400 μg/mL for 30 min at 37°C followed by proteinase K (VWR) at a concentration of 40 μg/mL for 1 h at 65°C. DNA was purified by phenol-chloroform extraction and isopropanol precipitation and resuspended in 10 mM Tris-HCl pH 8.0, 0.1 mM EDTA. In addition to the listed components, all buffers used before elution also contained 100 μM PMSF (Sigma-Aldrich) and 100 μM benzamidine (Sigma-Aldrich). qPCR analysis was performed using the QuantiFast SYBR Green PCR kit (Qiagen) on a Roche Lightcycler 480 using the following program: 5 min at 95°C, followed by 45 cycles of 10 s at 95°C and 30 s at 62°C. Experimental values were normalized to input. Trbv1 primers were 5’-CGCTGTGGAGAATGAAGGGT-3’ and 5’- CAGGTGAGGGAGCAAGGAAG-3’. 5’PC primers were 5’-TTAAACTGCTATGCTTTCGC-3’ and 5’-AGGTTAGGAGGTAACACAGT-3’. E_α primers were 5’-AGGAAGTCGCAGAACCTGAA-3’ and 5’-GAGGGAGAAAGCCTTTTGGT-3’. Dad1 primers were 5’-TACACTTACCTGGGCCTTTG-3’ and 5’-ATGCACCTTTCCTATGCTGG-3’.

Tcrb repertoire analysis

gDNA from DN3 thymocytes was purified by phenol-chloroform extraction and isopropanol precipitation. High throughput sequencing of the Tcrb repertoire was performed by using the Immunoseq service (Adaptive Biotechnologies). 37,092 unique reads were obtained from Ctcf ^f/f thymocytes while 98,503 unique reads were obtained from Ctcf ^f/fCd2-Cre thymocytes.

Statistical methods

Data were analyzed using GraphPad Prism software. When multiple comparisons were made, p values were adjusted to account for multiple comparisons. Differences with adjusted p values of < 0.05 were considered significant.

RESULTS

Conditional knockout of Ctcf in early thymocyte development

Multiple CBEs are found scattered throughout the Tcrb locus (Fig. 1A), and three intergenic CBEs have been implicated in the regulation of locus contraction, insulation and nuclear lamina association (12, 20). To determine the role of CTCF in the regulation of Tcrb recombination, we created a conditional knockout mouse model whereby CTCF is ablated in thymocytes. We crossed mice bearing floxed Ctcf alleles (Ctcf ^f/f) (22) (Fig. 1B) with mice bearing a Cre recombinase transgene driven by a human Cd2 minigene (Cd2-Cre) (23). The Cd2-Cre transgene is active from the early thymic progenitor stage of development, allowing for Ctcf allele deletion prior to the onset of Tcrb locus V_β-to-DJ_β recombination in CD44⁻CD25⁺ DN3 thymocytes.

Figure 1. Effect of CTCF deletion on thymocyte development.

(A) Diagram of the Tcrb locus. Red Trbv gene segments are classified as having nearby CBEs, whereas black Trbv gene segments do not (17). Red circles denote three RC-proximal CBEs. E_β, Tcrb enhancer. (B) Diagram of Ctcf floxed and deleted alleles. LoxP sites are represented by triangles. PGK-puromycin (puro^r) and splice acceptor-lacZ (SA-LacZ) cassettes are indicated. (C) Representative flow cytometry profiles of Ctcf ^f/f and Ctcf ^f/f Cd2-Cre thymocytes. Thymocytes were pre-gated on live cells and negative for other lineage markers. (D, E) Quantification of thymocyte cellularity and subpopulations as shown in (C). DN thymocyte subsets: DN1 (CD44⁺CD25⁻), DN2 (CD44⁺CD25⁺), DN3 (CD44⁻CD25⁺) and DN4 (CD44⁻CD25⁻). Means are indicated by the horizontal lines. *, p<0.05 by Student’s t-test with Holm-Sidak correction for multiple comparisons. (F) Representative flow cytometry profiles of FDG staining in Ctcf ^f/f Cd2-Cre thymocytes. Colored outlines correspond to the identically-colored boxes shown in (C).

We observed a block in thymocyte development in Ctcf ^f/fCd2-Cre mice, with a small increase in the number of DN thymocytes relative to Ctcf ^f/f mice, but substantial reductions in numbers of CD4⁺CD8⁺ double positive (DP) and CD4⁺CD8⁻ single positive thymocytes as well as overall thymocyte cellularity (Figs. 1C-1E). The remaining CD4⁻CD8⁺ thymocytes likely represent CD8⁺ immature single positive thymocytes, consistent with a block between the DN and DP stages. Cre-deleted Ctcf alleles drive expression of a lacZ reporter gene, enabling cells with at least one deleted allele to be detectable by flow cytometry following incubation with FDG, which is hydrolyzed by LacZ to generate fluorescein (Fig. 1B). The fraction of thymocytes staining positive for FDG increased gradually from DN2 to DN4 (Fig. 1F). Hereafter, to enrich for thymocytes with Cre-mediated deletion of Ctcf alleles, we sorted FDG⁺ DN3 thymocytes from Ctcf ^f/fCd2-Cre mice, and used DN3 thymocytes from Ctcf ^f/f mice as a control group.

Assessment of Ctcf deletion efficiency

To determine the efficiency of Ctcf deletion in FDG⁺ thymocytes, we performed qPCRs on gDNA from sorted FDG⁺ DN3 thymocytes from Ctcf ^f/fCd2-Cre mice and control DN3 thymocytes from Ctcf ^f/f mice. Approximately 95% of alleles were deleted (Fig. 2A). We determined the amount of CTCF protein present by performing western blots. CTCF protein levels were consistently reduced in Ctcf ^f/fCd2-Cre FDG⁺ as compared to Ctcf ^f/f thymocytes, but were still detectable (Fig. 2B). This raised the question of how much CTCF remained bound to the Tcrb locus in Ctcf ^f/fCd2-Cre FDG⁺ DN3 thymocytes. To assess this, we introduced Ctcf-floxed and Cd2-Cre alleles onto a Rag2^−/− background and performed ChIP to evaluate CTCF binding. Compared to Rag2^−/−, CTCF binding in Rag2^−/−Ctcf ^f/fCd2-Cre FDG⁺ thymocytes was reduced by 58% and 84% at two Tcrb CBEs and was more modestly reduced at a CBE associated with the Tcra enhancer (Fig. 2C). However, CTCF binding was substantially higher than at the negative control locus Dad1. Therefore, FDG⁺ Ctcf ^f/fCd2-Cre DN3 thymocytes displayed efficient genomic deletion of Ctcf, but expressed residual CTCF protein which remained bound at reduced levels to Tcrb locus CBEs.

Figure 2. Efficiency of CTCF deletion in Ctcf ^f/fCd2-Cre mice.

(A) Deletion of Ctcf alleles in DN3 thymocyte genomic DNA. qPCR values were initially normalized to Cd14, and values for Ctcf ^f/fCd2-Cre were subsequently normalized to Ctcf ^f/f, which was set to 1. The data represent the mean and SE of 4 mice of each genotype. (B) Western blot of CTCF and LAT in DN3 thymocytes of two Ctcf ^f/f and three Ctcf ^f/f Cd2-Cre mice. LAT served as a loading control (C) CTCF ChIP from DN3 thymocytes of Rag2^−/−Ctcf ^f/f and Rag2^−/−Ctcf ^f/f Cd2-Cre mice. ChIP values are expressed as mean and SE of two independent experiments (Trbv1, 5’PC, E_α). Analysis of Dad1 was performed once. Each analysis used DN3 thymocytes pooled from several mice of the appropriate genotype. *, p<0.05 by two-way ANOVA with Holm-Sidak correction for multiple comparisons, with Dad1 excluded from analysis.

Determination of the Tcrb repertoire

To define the role of CTCF in the formation of the pre-selection Tcrb repertoire, we harvested gDNA from pooled DN3 thymocytes isolated from two Ctcf ^f/f mice and from pooled FDG⁺ DN3 thymocytes isolated from five Ctcf ^f/fCd2-Cre mice, and performed high throughput sequencing (Fig. 3A). We observed that 11 of the 12 V_β gene segments located within 1 kb of a CBE (as previously defined by (17) were used less frequently in Ctcf ^f/fCd2-Cre mice (Fig. 3A, 3B). Relative usage of four of these V_β gene segments decreased by at least 1.5-fold (33% decrease; Trbv12–1, Trbv19, Trbv23, Trbv31). In contrast, of the 12 V_β gene segments lacking a nearby CBE, eight of them were used more frequently in Ctcf ^f/fCd2-Cre mice (Fig. 3A, 3B). Relative usage of five of these V_β gene segments increased by at least 1.5-fold (50% increase; Trbv16, Trbv17, Trbv21, Trbv26, Trbv30). To quantify the overall impact of Ctcf deletion on the Tcrb repertoire, we grouped V_β gene segments according to the presence or absence of a nearby CBE. The 12 V_β gene segments bearing a nearby CBE comprised about half of the Tcrb repertoire in control DN3 thymocytes, but only 37% in Ctcf-deleted DN3 thymocytes. Reciprocally, the 12 V_β gene segments lacking a nearby CBE comprised about half of the Tcrb repertoire in controls, but 63% of the repertoire in Ctcf-deleted cells (Fig. 3C). Therefore, CTCF modulates and diversifies the Tcrb repertoire by promoting the use of V_β gene segments found in proximity to a CBE.

Figure 3. Tcrb repertoire in DN3 thymocytes of Ctcf ^f/f and Ctcf ^f/fCd2-Cre mice.

(A) The Tcrb repertoire analyzed by high throughput sequencing of gDNA isolated from pooled DN3 thymocytes of two Ctcf ^f/f and five Ctcf ^f/f Cd2-Cre mice. V_β usage is plotted as the percentage of unique reads obtained. V_β gene segments with a nearby CBE as defined by (17) are indicated in red. Those lacking a nearby CBE are indicated in black. (B) Contingency table comparing usage of V_β gene segments with and without a nearby CBE in Ctcf ^f/f relative to Ctcf ^f/f Cd2-Cre mice. Statistical significance was evaluated by Fisher’s exact test. (C) V_β usage grouped by proximity of V_β segments to a nearby CBE.

DISCUSSION

In this study, we created a mouse model in which Ctcf was knocked out in thymocytes prior to Tcrb recombination. Thymocyte development was blocked, albeit incompletely, at the DN to DP transition. In sorted DN3 thymocytes with efficient deletion of Ctcf alleles, we detected residual levels of CTCF protein. However, in spite of the incomplete ablation of CTCF protein, we detected a striking change in the Tcrb repertoire, with underrepresentation of V_β gene segments located proximal to a CBE and overrepresentation of V_β segments not associated with a CBE. As such, our results offer experimental proof of a positive role for CBEs in V_β recombination as predicted by prior computational modeling (17).

Our Ctcf ^f/fCd2-Cre model was plagued by inefficient Cre-mediated deletion as well as residual CTCF protein even in DN3 thymocytes lacking both Ctcf alleles. Because of this, we presume that our experiments underestimate the impact of CTCF in shaping the V_β repertoire, and that the complete absence of CTCF would even more strongly disadvantage V_β segments with a nearby CBE. Overall thymocyte developmental progression in Ctcf ^f/fCd2-Cre mice was similar to that in Ctcf ^f/fLck-Cre mice, in which Cre-mediated deletion of Ctcf commenced at the DN2 stage (22). That study attributed developmental blockade during the DN to DP transition to a cell-cycle defect that was independent of Tcrb gene rearrangement. We believe that the effect of CTCF depletion on the V_β repertoire is independent of this cell-cycle defect, because V_β-to-DJ_β recombination occurs in non-cycling DN3 thymocytes.

A likely mechanism for the influence of CTCF on the V_β repertoire is its role as an architectural protein involved the formation of chromatin loops. In some examples, CTCF-dependent loops at antigen receptor loci have been shown to segregate gene segments and regulatory influences from one another (8, 11, 14). However, in other instances, CTCF-dependent loops have been shown to bring V gene segments into proximity of the RC (9, 10, 12). In this scenario, CBE-proximal V gene segments would more effectively compete for RC-bound RAG proteins. However, prior experimental and computational work concluded that V_β contact frequencies with the RC were not predictive of relative V_β recombination frequencies (17). CTCF may also facilitate recombination of nearby V segments by effects on local chromatin structure. Our ability to investigate these mechanistic questions was limited by poor yields of Ctcf-deleted DN3 cells.

Prior work on the Tcrb locus showed that the tethering, insulation and nuclear lamina-association functions of the RC-proximal CBEs are of particular importance in setting the balance between usage of proximal and distal V_β gene segments (12, 20). However, the perturbation of the Tcrb repertoire that we describe in Ctcf-deleted DN3 thymocytes appears to be quite distinct, with no evidence of a strong proximal V_β bias. In our experiments, CTCF was depleted not only at RC-proximal CBEs, but also at CBEs scattered across the V_β array. In this regard, recent studies of chromatin configurations in single cells revealed that cohesin, and by implication, CTCF, is not required for long-distance chromatin contacts, which were frequent and heterogeneous among individual nuclei lacking cohesin. Rather, loss of cohesin impacted the population-average of those contacts and eliminated preferred interactions involving CBEs (24). With this in mind, we postulate that in a Tcrb locus fully occupied by CTCF and cohesin, all V_β segments may contact the RC, with CTCF and cohesin fine-tuning these contacts and perhaps other aspects of V_β fitness for recombination to promote repertoire diversification. In the absence of RC-proximal CBE function, distal V_β segments may be outcompeted by proximal V_β segments for contacts with the RC, resulting in an RC-proximal bias to the repertoire (12, 20). However, with CTCF depleted across the entire locus, V_β segments may all be able to contact the RC, with no disadvantage to distal V_β segments, and no disadvantage to V_β segments lacking a nearby CBE, irrespective of their position in the locus. Our results establish that by binding to V_β-associated CBEs, CTCF influences the balance of V_β use across the V_β array, thereby diversifying the Tcrb repertoire.