Restriction of endogenous T cell antigen receptor β rearrangements to Vβ14 through selective recombination signal sequence modifications

  1. Cherry Wu,
  2. Sheila Ranganath,
  3. Megan Gleason,
  4. Barbara B. Woodman,
  5. Tiffany M. Borjeson,
  6. Frederick W. Alt*, and
  7. Craig H. Bassing†
  1. Howard Hughes Medical Institute, Children's Hospital, CBR Institute for Biomedical Research, and Department of Genetics, Harvard University Medical School, Boston, MA 02115

  1. Contributed by Frederick W. Alt, January 5, 2007 (received for review December 30, 2006)

 
Next Section

Abstract

T cell antigen receptor (TCR)β V(D)J variable region exon assembly is ordered, with Dβ to Jβ rearrangements occurring before joining of Vβs to a DJβ complex. Germ-line V(D)J segments are flanked by recombination signal (RS) sequences, which consist of heptamers and nonamers separated by a spacer of 12 (12-RS) or 23 (23-RS) bp. V(D)J recombination is restricted by the 12/23 rule; joining occurs only between gene segments flanked by 12-RSs and 23-RSs. Vβ segments have 23-RSs and Jβ segments 12-RSs, which based on the 12/23 rule should allow direct joining. However, Vβ segments rearrange only to DJβ complexes and not Jβ segments, because of restrictions beyond 12/23 (B12/23) that make the Vβ23-RS incompatible with the Jβ12-RS. To determine whether direct Vβ to Jβ joining occurs if flanking RSs are B12/23 compatible, we generated mice whose lymphocytes contained replacement of the Vβ1412-RS with the 3′Dβ112-RS on a TCRβ allele lacking Dβ segments (the Jβ1M6 allele). Mice heterozygous for the Jβ1M6 allele had dramatically increased Vβ14+ thymocyte and T cell numbers and decreased numbers of cells expressing other Vβs. This altered Vβ repertoire resulted from direct Vβ14 to Jβ1 rearrangements on the Jβ1M6 allele. Mice harboring lymphocytes homozygous for Jβ1M6 allele developed normal thymocyte and T cell numbers with all expressing Vβ14. Our findings show that selective RS modifications enforce rearrangement of a specific Vβ gene segment and demonstrate the importance of B12/23 mechanisms for ensuring generation of diverse TCRβ repertoires.

During lymphocyte development, T cell antigen receptor (TCR) and Ig variable region exons are assembled from germ-line V(D)J gene segments. V(D)J recombination is initiated by the lymphocyte-specific recombination activation gene (RAG)1/2 endonuclease, which introduces DNA double-strand breaks (DSBs) between a pair of gene segments and their flanking recombination signal (RS) sequences (1). Subsequently, the broken V, D, or J gene segments are joined by the generally expressed nonhomologous DNA end-joining pathway of DNS DSB repair (1). RSs consist of conserved heptamer and nonamer sequences separated by a nonconserved spacer of either 12 or 23 bp. V(D)J recombination occurs only between gene segments flanked by RSs that, respectively, contain 12- (12-RS) and 23- (23-RS) bp spacers because of a restriction in joining (the 12/23 rule) that is mediated at the level of RAG recognition and cutting (2, 3). The formation of synaptic complexes appears to occur through RAG binding to one RS followed by capture of a second RS (Mundy et al., ref. 4; ref. 5; and Jones and Gellert, ref. 6), and the 12/23 rule appears to be enforced by the initial binding of RAG to a 12-RS (6). This restriction in joining ensures the proper assembly of some variable region exons. For example, in the IgH locus VHs and JHs are flanked by 23-RSs and DHs are flanked on both sides by 12-RSs, which enforces the utilization of a D as an intermediate in assembling a complete VHDJH exon.

In CD4−/CD8− (double-negative, DN) thymocytes, TCRβ gene segments are assembled in an ordered manner with Dβ to Jβ rearrangements occurring on both alleles before joining of Vβ segments to a preexisting DJβ complex (7). In-frame (productive) VβDJβ rearrangements generate TCRβ chains. At this stage, pairing of TCRβ chains with the preexisting pTα chain to form preTCRs generates signals that lead to expansion of DN cells and their further development into CD4+/CD8+ (double-positive, DP) thymocytes and the rearrangement of TCRα genes (8, 9). If the first TCRβ allele that undergoes Vβ to DJβ rearrangement generates a productive TCRβ chain that pairs with pTα, Vβ to DJβ rearrangement on the second allele is thought to be inhibited by a feedback mechanism that may serve to help enforce TCRβ allelic exclusion (8, 9). However, out-of-frame VβDJβ rearrangements, or rearrangements that do not lead to a TCRβ chain that pairs with the preTα chain, on the first allele permit Vβ to DJβ rearrangements on the second allele that, if productive, generate a TCRβ chain that pairs with pTα to signal differentiation. In DP thymocytes, productive assembly of TCRα variable region exons can lead to the generation of TCRα chains that, if they pair with TCRβ chains, form αβ TCRs and signal differentiation to the CD4+/CD8− or CD4−/CD8+ (single-positive, SP) thymocyte stage (8, 9). Developmental stage-specific rearrangement of TCRβ and TCRα genes and TCRβ allelic exclusion are thought to be mediated, at least in part, through modulation of V(D)J recombinational accessibility of participating gene segments to the RAG endonuclease (8, 9).

In the TCRβ locus, Vβs have a 23-RS, Jβs a 12-RS, and Dβs a 5′ 12-RS and a 3′ 23-RS. Despite apparent 12/23 compatibility of Vβ 23-RSs and Jβ1 12-RSs, the 5′Dβ1 12-RS, but not the Jβ1 12-RSs, can target endogenous Vβ rearrangements (10). This “beyond 12/23 (B12/23) restriction” can be recapitulated in transfected plasmid substrates and occurs at the level of RAG cutting of DNA substrates in vitro (11, 12). The latter result demonstrates that functional synapses can occur between the RAG proteins and Vβ/Dβ RSs but not Vβ/Jβ RSs. In contrast, 3′Dβ 23-RSs are not B12/23-restricted, because they recombine with both 5′Dβ and Jβ 12-RSs in plasmid substrates (11, 12). Thus, specific replacement of the endogenous Vβ14 RS with the 3′Dβ1 RS (the Vβ14/3′DβRS mutation) results in Vβ14 rearrangements to both DJβ1 complexes and directly to Jβ1 segments (13). In this context, the predominant rearrangements still are mediated by DJβ intermediates, reflecting the relative recombination efficiencies between Vβ14/3′DβRS and the 5′Dβ1 vs. Jβ1 RSs in transfected plasmid substrates (11, 13). Together, these data suggest the restricted ability of Vβ segments to rearrange to DJβ complexes, but not Jβ segments, is mediated by interactions between TCRβ locus RSs and the RAG proteins, rather than by RS-specific transacting factors that modulate recombinational accessibility. The TCRβ locus B12/23 restriction may have evolved to ensure that the assembly of TCRβ variable region exons occurs through two individual recombination events to increase junctional diversity of TCRβ chains and also provide an additional level of regulatory control by requiring the VβDJβ assembly process to proceed through a DJβ intermediate (10).

We have previously demonstrated that selective modifications of TCRβ RS sequences can be used to restrict endogenous TCRβ rearrangements. Mice with both TCRβ alleles lacking the Dβ1 segment and the Dβ2Jβ2 cluster (the Jβ1M3 allele) exhibit a block in thymocyte development at the DN stage because of an inability of Vβ segments to rearrange directly to Jβ1 segments (10). However, specific replacement of the Jβ1.2 12-RS with the 5′Dβ1 12-RS on the Jβ1M3 allele completely rescued thymocyte development by allowing a diverse repertoire of endogenous Vβ segment rearrangements directly to the modified Jβ1.2 segment (10). To evaluate whether selective VβRS modifications can also be used to generate direct Vβ to Jβ rearrangement and to restrict endogenous TCRβ rearrangements to a particular Vβ segment, we now have generated and analyzed mice containing specific replacement of the Vβ14 RS with the 3′Dβ1 RS on the Jβ1M3 allele.

Previous SectionNext Section

Results

Replacement of the Vβ14 RS with the 3′Dβ1 RS on the Jβ1M3 Allele Promotes a Dramatic Increase in Vβ14+ Thymocytes and Peripheral αβ T Cells.

To evaluate whether specific Vβ RS modifications can be used to direct Vβ to Jβ rearrangement and enforce rearrangement of specific Vβ rearrangements as compared with others, we used gene-targeted mutation to specifically replace the endogenous Vβ14 RS with the 3′Dβ1 RS on the Jβ1M3 allele (which lacks Dβ segments) of Jβ1M3/ω ES cells, creating the Jβ1M6 allele (Fig. 1). The Jβ1ω allele, which lacks only the Dβ2Jβ2 cluster but contains Dβ1, supports TCRβ rearrangements involving all 20 endogenous Vβ segments and contributes effectively to normal αβ T cell development (10). This targeting did not alter any sequences immediately adjacent to the RS replacement but did insert both a novel BglII site to distinguish between Vβ14 rearrangements on the Jβ1M6 and Jβ1ω alleles and a single loxP site 172 bp 5′ of the RS replacement (Fig. 1). We previously demonstrated that BglII and loxP sites inserted at this location affect neither αβ T cell development nor Vβ14 rearrangement (13).

Fig. 1.
View larger version:
Fig. 1.

Generation of Jβ1M6/ω ES cells. (A) Schematic diagrams of the Jβ1M3 TCRβ locus, 14/3RS targeting vector, Jβ1M6Neo, and Jβ1M6 alleles. Open boxes illustrate representative upstream Vβs, the six Jβ1s, and Vβ14. Triangles illustrate RSs, with 12-RSs noted by black circles. Open circles indicate loxP sites. The Vβ14 and 3′Dβ1 23-RS sequences are shown with nucleotide differences between the RSs underlined. The locations of the 5′, 3′, and Δp probes are indicated. Restriction site nomenclature: B, BamHI; N, NdeI; B2, BglII; H, HindIII. (B) Schematic diagrams of the Jβ1ω, Jβ1M3, and Jβ1M6 alleles. The 3′Dβ1 RS is noted by the solid triangle.


We used Jβ1M6/ω ES cells and RAG-2-deficient blastocyst complementation (14) to generate chimeric Jβ1M6/ω mice in which all lymphocytes are derived from the Jβ1M6/ω ES cells. The number of thymocytes in Jβ1M6/ω mice (140 ± 60 × 106; three mice) was comparable to those in control 129SvEv mice (173 ± 62 × 106; three mice). Jβ1M6/ω mice also contained similar numbers of cells in the spleen and lymph nodes as compared with control 129SvEv mice (data not shown). Flow cytometry (FACS) analysis of Jβ1M6/ω thymocytes with anti-CD4 and anti-CD8 antibodies showed a normal distribution of DN, DP, and SP populations (Fig. 2 A). Jβ1M6/ω mice also contained normal populations of DP thymocytes expressing “intermediate” levels of cell surface TCRβ and SP thymocytes expressing “high” levels of cell surface TCRβ (Fig. 2 B). In addition, FACS analysis of spleen and lymph node cells from Jβ1M6/ω mice and control 129SvEv mice revealed a normal distribution of CD4+ and CD8+ peripheral αβ T cells [supporting information (SI) Fig. 5; spleen data not shown], perhaps with a slight skewing toward CD4+ cells in Jβ1M6/ω mice.

Fig. 2.
View larger version:
Fig. 2.

FACS analysis of thymocytes from Jβ1M6/ω and wild-type mice. (A) Shown is a representative CD4-PE and CD8-FITC analysis. Total thymocyte numbers and the percentages of DN, DP, and SP cells are indicated for these representative mice. (B) Histogram plot of cell surface TCRβ expression with the percentage of TCRβ high thymocytes indicated for these mice. (C) Depicted is a representative TCRβ-PE and Vβ14-FITC analysis with the percentage of TCRβ high (SP) and TCRβ intermediate Vβ14+ cells indicated. (D) Bar graph showing the average percentage of Vβ14+, Vβ5+, and Vβ8+ TCRβ high (SP) thymocytes calculated from three mice of each genotype shown.


To evaluate the effect of the 3′Dβ1 RS replacement on Vβ repertoire, we conducted FACS analysis of Jβ1M6/ω thymocytes and peripheral lymphocytes with antibodies specific for the TCRβ chain and Vβ14, Vβ5, or Vβ8. We found that, on average, 42% (42.0 ± 1.7%; three mice) of Jβ1M6/ω TCRβ high (SP) thymocytes were Vβ14+, whereas only 7% (7 ± 0%; three mice) of control 129SvEv TCRβ high (SP) thymocytes expressed Vβ14 on the cell surface (Fig. 2 C and D). In contrast, the average percentage of Vβ5+ TCRβ high (SP) thymocytes was reduced from 6% (6 ± 1%; three mice) in control 129SvEv mice to 4% (4.0 ± 0.3%; three mice) in Jβ1M6/ω mice (Fig. 2 D), whereas the average percentage of Vβ8+ TCRβ high (SP) thymocytes was reduced from 21% (21 ± 0.3%; three mice) in control 129SvEv mice to 13% (13.0 ± 0.3%; three mice) in Jβ1M6/ω mice (Fig. 2 D). FACS analysis of spleens and lymph node cells of Jβ1M6/ω mice revealed a similar dramatic increase in the numbers of Vβ14+ peripheral αβ T lymphocytes with a concomitant decrease in the numbers of Vβ5+ and Vβ8+ cells (SI Fig. 5). Because the number of thymocytes and peripheral αβ T cells was comparable among Jβ1M6/ω mice and control mice, Jβ1M6/ω mice develop 6- to 7-fold more Vβ14+ TCRβ high (SP) thymocytes and αβ T lymphocytes with a corresponding decrease in the percentage of cells expressing other Vβs. Consequently, when attached to the endogenous Vβ14 segment on the Jβ1M3 allele of Jβ1M3/ω cells, the 3′Dβ1 23-RS substantially alters the Vβ repertoire of developing thymocytes and mature αβ T cells.

The 3′Dβ1 23-RS Directs a High Level of Vβ14 Rearrangement Directly to the Jβ1 Segments.

Because the peripheral Vβ repertoire is not substantially altered from that generated in the thymus during normal αβ T cell development (15–17), the 3′Dβ1 RS likely increases Vβ14 utilization in Jβ1M6/ω thymocytes and peripheral αβ T cells by promoting Vβ14 rearrangements directly to Jβ1 segments on the Jβ1M6 allele. To confirm this notion, we first assayed TCRβ rearrangements in a panel of 93 Vβ14+ Jβ1M6/ω αβ T cell hybridomas. Because these cells express Vβ14, they must contain either a productive Vβ14Jβ1 rearrangement on the Jβ1M6 allele or a productive Vβ14Dβ1Jβ1 rearrangement on the Jβ1ω allele. Southern blot analysis of genomic DNA demonstrated that all 93 of these hybridomas contained direct Vβ14 rearrangements to Jβ1 rearrangements involving one of the six Jβ1 segments on the Jβ1M6 allele (Table 1, SI Table 2; data not shown). Thus, the sequence of the RS flanking Vβ14 alone determines whether Vβ14 can rearrange directly to Jβ1 segments. Additional Southern blot analyses demonstrated that 50 of 93 (53%) Vβ14 hybridomas contained VβDβ1Jβ1 rearrangements involving Vβ segments other than Vβ14 on the Jβ1ω allele, whereas 43 of 93 (47%) hybridomas contained only Dβ1Jβ1 rearrangements on the Jβ1ω allele (Table 1, SI Table 2; data not shown). Therefore, the 3′Dβ1 RS can promote Vβ14 rearrangements directly to Jβ1 segments on the Jβ1M6 allele at a level that competes with the rearrangement level of the 20 endogenous Vβ segments to DJβ complexes on the Jβ1ω allele.

View this table:
Table 1.

Analysis of rearrangements in Jβ1M6/ω and Jβ1M6/M6 hybridomas


In developing αβ T cells, productive VβDβJβ1 rearrangements must occur to promote continued differentiation. Because only one-third of VβDβJβ rearrangements occur in-frame, ≈40% of Jβ1ω/ω αβ T cell hybridomas often contain VβDβ1Jβ1 rearrangements on both alleles (one of which is usually nonproductive and presumably came first), with the other ≈60% containing only Dβ1Jβ1 rearrangements on the nonselected alleles that are fixed by feedback from the productive VβDJβ rearrangement (10). Thus, to accurately quantify the level of direct Vβ14 to Jβ1 rearrangements on the Jβ1M6 allele, we analyzed TCRβ rearrangements on the nonproductive alleles in a panel of 119 Jβ1M6/ω αβ T cell hybridomas that expressed a Vβ other than Vβ14 (VβX+) on their cell surface. Southern blot analysis confirmed that each of these VβX+ Jβ1M6/ω αβ T cell hybridomas contained a (presumably) productive VβDβ1Jβ1 rearrangement involving an upstream Vβ, and not Vβ14, on the Jβ1ω allele (Table 1, SI Table 3; data not shown). Additional Southern blot analysis demonstrated that only 29 of 119 (24%) VβX+ Jβ1M6/ω αβ T cell hybridomas contained VβDJβ rearrangements on the Jβ1M6 allele and that all of these involved direct Vβ14 rearrangements to one of the six Jβ1 segments (Table 1; data not shown). Therefore, considering that we would expect ≈40% of the second (nonproductive alleles to be rearranged on a normal allele, the 3′Dβ1 RS promotes Vβ14 rearrangements directly to Jβ1 segments on the Jβ1M6 allele at approximately half (24%) the level (40%) that would be generated by rearrangement of the 20 endogenous Vβ segments to DJβ complexes on the Jβ1ω allele.

A Specific RS Can Restrict the Endogenous Vβ Repertoire to a Single Vβ Sequence.

Because direct Vβ14 to Jβ1 rearrangements occur at a substantial level on the Jβ1M6 allele, we also evaluated whether the 3′Dβ1 RS can restrict endogenous TCRβ rearrangements to only Vβ14. For this purpose, we generated Jβ1M6/M6 ES cells through high G418 selection of Jβ1M6Neo/ω ES cells and used these to make Jβ1M6/M6 mice through RAG-2-deficient blastocyst complementation. The number of thymocytes in Jβ1M6/M6 mice (165 ± 8.7 × 106; three mice) was comparable to those in control 129SvEv mice (147 ± 15 × 106; three mice). FACS analysis of Jβ1M6/M6 thymocytes showed a normal distribution of thymocytes (Fig. 3 A). The total numbers of peripheral αβ T cells and CD4+ and CD8+ cells also appeared normal in Jβ1M6/M6 mice (Fig. 3 C). Strikingly, 100% of Jβ1M6/M6 thymocytes and peripheral αβ T cells were Vβ14+ (Fig. 3 B and D). We next assessed TCRβ rearrangements in a panel of 147 Vβ14+ Jβ1M6/M6 αβ T cell hybridomas. Southern blot analysis demonstrated that 126 of 147 (86%) contained Vβ14Jβ1 rearrangements on only one allele, whereas 21 of 147 (14%) contained Vβ14Jβ1 rearrangements on both alleles (Table 1; data not shown), further demonstrating that direct rearrangements of the single Vβ14 to Jβ1 segments on the Jβ1M6 allele are not as efficient as the rearrangements that can involve any of the 20 endogenous Vβ segments to a DJβ1 complex on the Jβ1ω allele. Overall, our data clearly demonstrate that selective RS modifications can be used to restrict endogenous Vβ repertoire through the isolation of Vβ14 rearrangements.

Fig. 3.
View larger version:
Fig. 3.

FACS analysis of thymocytes and lymph node cells from Jβ1M6/M6 and wild-type mice. (A and B) Shown are representative CD4-PE and CD8-FITC (A) and TCRβ-PE and Vβ14-FITC analysis of thymocytes with total thymocyte numbers for these particular mice (B). (C and D) Shown are representative CD4-PE and CD8-FITC (C) and TCRβ-PE and Vβ14-FITC analysis of lymph node cells from mice of each genotype (D).


Regulation of TCRβ Rearrangement in Jβ1M6/ω and Jβ1M6/Μ6 Lymphocytes.

In normal thymocytes, Dβ to Jβ rearrangement is initiated in DN2 stage thymocytes and Vβ to DJβ rearrangement initiate at the later DN3 stage (18). To determine whether direct Vβ14 to Jβ rearrangements on the Jβ1M6 allele were regulated like Dβ to Jβ or normal Vβ to Jβ rearrangements, we used PCR to analyze Vβ14 to (D)Jβ rearrangement in sort-purified DN2 and DN3 cell populations of Jβ1M6/ω mice. Although we observed PCR products corresponding to Dβ to Jβ rearrangements in both DN2 and DN3 cells, we were able to detect only PCR products corresponding to Vβ14 to (D)Jβ rearrangements in DN3 thymocytes (Fig. 4). Thus, the onset of direct Vβ14 to Jβ rearrangements on the Jβ1M6 allele appears to occur at the normal developmental timing for normal Vβ (to DJβ) rearrangements.

Fig. 4.
View larger version:
Fig. 4.

Developmental stage-specific Vβ14 to (D)Jβ rearrangements in Jβ1M6/ω thymocytes. Genomic DNA (100 ng) from Jβ1M6/ω ES cells (lane 1), sorted DN2 T cells (lane 2), and sorted DN3 T cells (lane 3) was analyzed by PCR by using the Vβ14/P2 and Dβ1/P2 primer sets. PCR products were analyzed by Southern blotting by using the PR2 oligonucleotide probe. Bands corresponding to Vβ14 to (D)Jβ1 and Dβ1 to Jβ1 rearrangements are indicated.


Flow cytometry analysis using staining with anti-Vβ14 and anti-TCRαβ antibodies (data not shown) demonstrated that none of the 119 VβX+ Jβ1M6/ω hybridomas obviously produced both surface Vβ14 and VβX [including the 29 that had Vβ(D)Jβ rearrangements on both alleles], consistent with allelic exclusion of the JβM6 allele by VβX to DJβ rearrangements on the Jbw allele. However, because of the multiplicity of different Vβs, we could not readily do the same type of analyses for Vβ14+ Jβ1M6/ω hybridomas, and it was not possible to check Jβ1M6/M6 hybridomas for double producers of Vβ14, because the products of the two alleles are indistinguishable. Therefore, as a preliminary means of looking for double producers, we sequenced both Vβ14Jβ junctions in 17 clonal VβX+ Jβ1M6/ω αβ T cell hybridomas with Vβ14Jβ rearrangements on the M6 allele and eight clonal Vβ14+ Jβ1M6/M6 hybridomas that had Vβ14Jβ rearrangements on both alleles. Notably, three of 17 clonal VβX+ Jβ1M6/ω αβ T cell hybridomas contained a Vβ14Jβ rearrangement that was in-frame at the junction, in addition to their in-frame VβXDβ1Jβ rearrangement (SI Table 4 and data not shown). However, given that none of these hybridomas appeared to produce Vβ14 on their surface (SI Fig. 6), it would appear that allelic exclusion was maintained and that these three in-frame Vβ14Jβ1 joins either had other sequence alterations that made them nonproductive or failed to pair with preTα/TCRα chains. In the latter context, very limited studies have shown two in-frame rearrangements in 5–10% of wild-type αβ T cells and αβ T cell hybridomas (ref. 16; unpublished studies), but double-surface expression is rare (19). Sequences of both VβDJβ junctions in the eight Jβ1M6/M6 hybridomas demonstrated that all had one in-frame and one out-of-frame Vβ14Jβ1 rearrangement, consistent with feedback regulation (SI Table 5). However, assuming that one of three rearrangements occur in-frame, allelic inclusion would only predict one in five peripheral αβ T cells with two in-frame rearrangements, many more rearrangements would need to be sequenced to confirm whether there truly is feedback regulation in this genotype.

Previous SectionNext Section

Discussion

We demonstrate that selective TCRβ locus modifications can be used to enforce rearrangement of a specific Vβ segment directly to Jβ segments and restrict the Vβ repertoire in vivo. The 3′Dβ1 RS, when attached to Vβ14 on a modified TCRβ allele lacking Dβ segments, leads to chromosomal Vβ14 rearrangements directly to Jβ1 segments in the absence of rearrangement of any of Vβ segment on the modified allele. These data indicate the sequence of the Vβ14 RS is a major factor governing the normal specificity for Vβ14 rearrangement to DJβ complexes and not unrearranged Jβ1 segments, thereby contributing to enforcement of the B12/23 restriction between Vβ14 and Jβ1 segments. Consequently, neither the 5′Dβ1 RS nor putative 5′Dβ1 RS specific transfactors are required to target the rearrangement per se of chromosomal Vβ14 rearrangements. The Vβ14 segment is unique among Vβ segments because of its close proximity to Dβ segments and its rearrangement through inversion rather than deletion. Thus, selective RS modifications involving additional Vβ RSs will be required to evaluate the relative contribution of RS sequences vs. other chromosomal factors in targeting the rearrangement of other Vβ segments. The assembly and expression of a diverse TCRβ repertoire in αβ T lymphocytes is essential for the generation of an effective adaptive immune system. Our current findings suggest that Vβ RSs may have evolved under selective pressure to enforce B12/23 restricted joining between Vβ and Jβ1 segments and, thereby, ensure the utilization of a Dβ segment in TCRβ variable region exons. Because the 3′Dβ1 RS promotes Vβ14 to Jβ1 rearrangements on the Jβ1M6 allele at a level that competes with overall Vβ to DJβ rearrangements on the Jβ1ω allele and dramatically alters Vβ repertoire in Jβ1M6/ω αβ T cells, Vβ RS sequences also may have evolved to ensure the expression of a diverse Vβ repertoire.

Expression of productive VβDJβ rearrangements is thought to inhibit further Vβ to DJβ rearrangements to enforce TCRβ locus allelic exclusion (20). In this context, ordered Dβ to Jβ and Vβ to DJβ rearrangement might be related to mechanisms that allow feedback regulation of Vβ rearrangements. Ordered assembly of TCRβ variable region exons may be mediated either by developmental stage-specific accessibility of Dβ and Jβ segments in DN3 thymocytes vs. Vβ segments in DN3 cells or established by differential efficiency of the Dβ to Jβ vs. Vβ to Dβ rearrangement steps (21). We showed that the 3′Dβ1 RS targets endogenous Dβ1 to Jβ1 rearrangements in DN2 thymocytes; however, we could not detect Vβ14 to (Dβ1)Jβ1 rearrangements in DN2 thymocytes of Jβ1M6/ω mice. Thus, the differential efficiency of 3′Dβ vs. Vβ RSs alone does not direct the ordered assembly of TCRβ gene segments. Consequently, developmental stage-specific modulation of Dβ and Jβ versus Vβ recombinational accessibility is likely the major factor that enforces ordered TCRβ gene rearrangement.

Our FACS analysis of Jβ1M6/ω αβ T cells and hybridomas with Vβ-specific antibodies and our sequence analysis of VβDβJβ rearrangements in Jβ1M6/Μ6 αβ T cells and hybridomas suggest that TCRβ allelic exclusion is maintained. These preliminary findings would imply that neither the assembly of DJβ intermediates nor putative 5′Dβ RS specific transfactors are required to enforce TCRβ locus allelic exclusion. However, although much more extensive analysis will required to confirm whether direct Vβ14 to Jβ1 rearrangements on the M6 allele are subject to normal feedback regulation and generate normal feedback regulation of wild-type alleles, our current findings establish locus modification and RS replacement as a useful approach to address these longstanding questions. In this context, gene-targeted replacement of other Vβ RS(s) with the 3′Dβ1 RS on the Jβ1M3 and Jβ1M6 alleles may also provide novel experimental systems to enforce the rearrangement of specific Vβ gene segment(s) and, thereby, to allow investigations of mechanisms that regulate the initiation and feedback regulation of Vβ rearrangement.

Previous SectionNext Section

Experimental Procedures

Targeting Constructs and Probes.

The 14/3RS and the 14/loxp targeting vectors were previously described (13). The 5′ probe is a 1.4-kb PstI/NdeI fragment. The 3′ probe is a 0.7-kb SphI/HindIII fragment. The Δp probe is a 1.5-kb HindIII/NdeI fragment.

Gene Targeting and Generation of ES Cells.

The 14/3RS and 14/loxp targeting vectors (13) were electroporated into Jβ1M3/ω ES cells (10) as described (22) to generate Jβ1M6Neo/ω and Jβ1M9Neo/ω ES cells, respectively. Targeted clones were identified by Southern blot analysis, and the PGK-neor gene was removed by Cre-mediated recombination as described (13) to generate Jβ1M6/ω and Jβ1M9/ω ES cells. The Vβ14 targetings were linked to the M3 mutation through the appearance of either a 2.3- or 4.3-kb band with the 5′Jβ1 probe (10), which are generated through Cre-mediated deletion between the M3 loxP site and either loxP site of the Vβ14 targeting loxP-Neor cassette. Jβ1M6/M6 ES cells were generated through increased G418 selection of Jβ1M6Neo/ω ES cells to select for homozygous Jβ1M6Neo/M6Neo ES cells, which were then subject to Cre-mediated recombination.

RAG-2-Deficient Blastocyst Complementation and the Generation of Jβ1M6/ω and Jβ1M6/M6 Lymphocytes.

Jβ1M6/ω and Jβ1M6/M6 lymphocytes were generated through RAG-2-deficient blastocyst complementation, as described (14). Three Jβ1M6/ω and four Jβ1M6/M6 mice were generated and analyzed.

FACS Analysis.

Cells from the thymus, spleen, and lymph nodes of chimeric mice were isolated and stained with FITC-conjugated anti-CD8, anti-Vβ14 TCR, anti-TCRβ chain and phycoerythrin-conjugated anti-CD4, anti-Vβ5.1, 5.2 TCR, anti-Vβ8, and antiTCRβ chain antibodies (PharMingen, San Diego, CA). Data acquisition and analysis were performed on a FACScalibur flow cytometer equipped with CellQuest software (Becton Dickinson, Franklin Lakes, NJ).

Hybridoma Analysis.

Hybridoma clones were produced by fusion of Con A and IL-2 stimulated T cells with the thymoma cell line BW-1100.129.237 (23), as described (22). FACS analysis of anti-Vβ14 TCR and anti-TCRβ chain antibodies was conducted to identify hydridomas expressing Vβ14 (Vβ14+) and/or another Vβ (VβX+) on the cell surface. Genomic DNA was isolated and subjected to Southern blotting and PCR. The Southern blot analysis of TCRβ rearrangements was conducted as described (13). Vβ14 to Jβ1 rearrangements were amplified by PCR by using the Vβ14 and P4 primers (10). PCR products were directly sequenced with the Vβ14 primer.

Cell Sorting and PCR.

Cell sorting of DN2 and DN3 thymocytes was performed by using a MoFlo cell sorter (Cytomation, Fort Collins, CO) after staining of CD4-depleted thymocytes (Miltenyi Biotec, Auburn, CA) with FITC-conjugated anti-CD8α, FITC-conjugated anti-CD4, FITC-conjugated anti-TCRβ, FITC-conjugated anti-B220, FITC-conjugated anti-TCRγδ, phycoerythrin-conjugated anti-CD25, and CY5-conjugated anti-CD44. Dβ1-Jβ1.2 and Vβ14-Jβ1.2 primers pairs and conditions were described (24, 25). Amplification conditions were 94°C for 3 min, then 94°C for 45 s, 60°C for 75 s, 72°C for 60 s (25 cycles), and 72°C for 5 min. Products were separated by using a 1% agarose/1× Tris-borate/EDTA gel then transferred. Rearrangement products were detected with oligo probe PR2 (10).

Previous SectionNext Section

Acknowledgments

We thank Eugene Oltz and Barry Sleckman for critical review of this manuscript. C.W. was supported by an Irvington Institute for Immunological Research Postdoctoral Fellowship. S.R. was supported by a Genentech/IDEC Fellowship from the American Cancer Society and is currently supported by the Dana–Farber Cancer Institute Postdoctoral Training Program in Cancer Immunology. C.H.B. was a Lymphoma Research Foundation Fellow and is a Pew Scholar in the Biomedical Sciences and a Lymphoma Research Foundation Fellow. F.W.A is an Investigator of the Howard Hughes Medical Institute. This work was supported by National Institutes of Health Grant A.I.20047 (to F.W.A.).

Previous SectionNext Section

Footnotes

  • *To whom correspondence should be addressed. E-mail: alt{at}enders.tch.harvard.edu

  • Author contributions: C.W., S.R., and C.H.B. contributed equally to this work; C.W., S.R., F.W.A., and C.H.B. designed research; C.W., S.R., M.G., B.B.W., and C.H.B. performed research; T.M.B. contributed new reagents/analytic tools; C.W., S.R., F.W.A., and C.H.B. analyzed data; and C.W., S.R., F.W.A., and C.H.B. wrote the paper.

  • ↵ †Present address: Department of Pathology and Laboratory Medicine, Children's Hospital of Philadelphia, University of Pennsylvania School of Medicine, Abramson Family Cancer Research Institute, Philadelphia, PA 19104.

  • The authors declare no conflict of interest.

  • This article contains supporting information online at www.pnas.org/cgi/content/full/0700081104/DC1.

  • Abbreviations:
    RS,
    recombination signal;
    TCR,
    T cell antigen receptor;
    DN,
    double negative;
    DP,
    double positive;
    SP,
    single positive.
Previous Section
 

References

  1. ↵
    1. Jung D,
    2. Alt FW
    (2004) Cell 116:299–311.
    CrossRefMedlineISI
  2. ↵
    1. Eastman QM,
    2. Leu TM,
    3. Schatz DG
    (1996) Nature 380:85–88.
    CrossRefMedline
  3. ↵
    1. van Gent DC,
    2. Ramsden DA,
    3. Gellert M
    (1996) Cell 85:107–113.
    CrossRefMedlineISI
  4. ↵
    1. Mundy CL,
    2. Patenge N,
    3. Matthews AG,
    4. Oettinger MA
    (2002) Mol Cell Biol 22:69–77.
    Abstract/FREE Full Text
  5. ↵
    1. Curry JD,
    2. Geier JK,
    3. Schlissel MS
    (2005) Nat Immunol 6:1272–1279.
    CrossRefMedlineISI
  6. ↵
    1. Jones JM,
    2. Gellert M
    (2002) EMBO J 21:4162–4171.
    CrossRefMedlineISI
  7. ↵
    1. Born W,
    2. Yague J,
    3. Palmer E,
    4. Kappler J,
    5. Marrack P
    (1985) Proc Natl Acad Sci USA 82:2925–2929.
    Abstract/FREE Full Text
  8. ↵
    1. Jackson AM,
    2. Krangel MS
    (2006) Immunol Rev 209:129–141.
    CrossRefMedlineISI
  9. ↵
    1. Krangel MS
    (2003) Nat Immunol 4:624–630.
    CrossRefMedlineISI
  10. ↵
    1. Bassing CH,
    2. Alt FW,
    3. Hughes MM,
    4. D'Auteuil M,
    5. Wehrly TD,
    6. Woodman BB,
    7. Gartner F,
    8. White JM,
    9. Davidson L,
    10. Sleckman BP
    (2000) Nature 405:583–586.
    CrossRefMedline
  11. ↵
    1. Jung D,
    2. Bassing CH,
    3. Fugmann SD,
    4. Cheng HL,
    5. Schatz DG,
    6. Alt FW
    (2003) Immunity 18:65–74.
    CrossRefMedlineISI
  12. ↵
    1. Tillman RE,
    2. Wooley AL,
    3. Khor B,
    4. Wehrly TD,
    5. Little CA,
    6. Sleckman BP
    (2003) J Immunol 170:5–9.
    Abstract/FREE Full Text
  13. ↵
    1. Wu C,
    2. Bassing CH,
    3. Jung D,
    4. Woodman BB,
    5. Foy D,
    6. Alt FW
    (2003) Immunity 18:75–85.
    CrossRefMedlineISI
  14. ↵
    1. Chen J,
    2. Lansford R,
    3. Stewart V,
    4. Young F,
    5. Alt FW
    (1993) Proc Natl Acad Sci USA 90:4528–4532.
    Abstract/FREE Full Text
  15. ↵
    1. Aifantis I,
    2. Buer J,
    3. von Boehmer H,
    4. Azogui O
    (1997) Immunity 7:601–607.
    CrossRefMedlineISI
  16. ↵
    1. Aifantis I,
    2. Pivniouk VI,
    3. Gartner F,
    4. Feinberg J,
    5. Swat W,
    6. Alt FW,
    7. von Boehmer H,
    8. Geha RS
    (1999) J Exp Med 190:1093–1102.
    Abstract/FREE Full Text
  17. ↵
    1. Wilson A,
    2. Marechal C,
    3. MacDonald HR
    (2001) J Immunol 166:51–57.
    Abstract/FREE Full Text
  18. ↵
    1. Godfrey DI,
    2. Kennedy J,
    3. Mombaerts P,
    4. Tonegawa S,
    5. Zlotnik A
    (1994) J Immunol 152:4783–4792.
    Abstract
  19. ↵
    1. Jackson AM,
    2. Krangel MS
    (2005) J Immunol 175:5186–5191.
    Abstract/FREE Full Text
  20. ↵
    1. Uematsu Y,
    2. Ryser S,
    3. Dembic Z,
    4. Borgulya P,
    5. Krimpenfort P,
    6. Berns A,
    7. von Boehmer H,
    8. Steinmetz M
    (1988) Cell 52:831–841.
    CrossRefMedlineISI
  21. ↵
    1. Bassing CH,
    2. Swat W,
    3. Alt FW
    (2002) Cell 109(Suppl):S45–S55.
    CrossRefMedlineISI
  22. ↵
    1. Sleckman BP,
    2. Bardon CG,
    3. Ferrini R,
    4. Davidson L,
    5. Alt FW
    (1997) Immunity 7:505–515.
    CrossRefMedlineISI
  23. ↵
    1. White J,
    2. Blackman M,
    3. Bill J,
    4. Kappler J,
    5. Marrack P,
    6. Gold DP,
    7. Born W
    (1989) J Immunol 143:1822–1825.
    Abstract
  24. ↵
    1. Dudley DD,
    2. Sekiguchi J,
    3. Zhu C,
    4. Sadofsky MJ,
    5. Whitlow S,
    6. DeVido J,
    7. Monroe RJ,
    8. Bassing CH,
    9. Alt FW
    (2003) J Exp Med 198:1439–1450.
    Abstract/FREE Full Text
  25. ↵
    1. Gartner F,
    2. Alt FW,
    3. Monroe R,
    4. Chu M,
    5. Sleckman BP,
    6. Davidson L,
    7. Swat W
    (1999) Immunity 10:537–546.
    CrossRefMedlineISI
« Previous | Next Article »Table of Contents

This Article

  1. Published online before print February 27, 2007, doi: 10.1073/pnas.0700081104 PNAS March 6, 2007 vol. 104 no. 10 4002-4007
  1. AbstractFree
  2. Figures Only
  3. » Full Text
  4. Full Text (PDF)
  5. Supporting Information

Classifications

Navigate This Article