The combinatorial repertoire of AgRs is established through somatic recombination of V, D, and J gene segments during lymphocyte development. Incorporation of D segments into IgH, TCRβ, and TCRδ chains also contributes to junctional diversification by substantially extending the length of the third CDR. The V, D, and J gene segments are flanked by recombination signals (RS) of 12- or 23-mer spacer length that direct recombination according to the 12/23 rule. D genes in the TCRβ and TCRδ loci are flanked by a 12RS and 23RS, and their incorporation is controlled by mechanisms “beyond the 12/23 rule.” In the TCRβ locus, selective interactions between Rag proteins and the RS flanking the V-D and D-J genes, respectively, are sufficient to enforce D gene usage. In this article, we report that in the TCRδ locus, the Rag proteins are not the major determinant of D gene incorporation. In developing mouse and human thymocytes, the two Dδ genes rearrange predominantly to form D-D coding joints. In contrast, when tested in ex vivo transfection assays in a nonlymphoid cell line, the flanking RS mediate deletion, rather than incorporation, of the two D genes on both exogenous recombination substrates and the endogenous locus. These results suggest that selective Rag-RS interactions are not the sole regulators of D gene segment incorporation, and additional, perhaps lymphocyte-specific, mechanisms exist that allow proper shaping of the primary AgR repertoire.

Somatic assembly of AgR V, D, and J genes through VDJ recombination is mediated by the VDJ recombinase encoded in the two RAGs (RAG-1 and RAG-2; for reviews, see Refs. 1 and 2). The Rag proteins specifically bind to the recombination signal (RS) 3 sequences that flank the 3′ end of V, the 5′ end of J, and both ends of the D gene segments. The RS is composed of highly conserved heptamer and nonamer motifs separated by less conserved spacers of 12- or 23-bp length (3). According to the 12/23 rule, only gene segments flanked by an RS of different length can rearrange efficiently (4). At the molecular level, the 12/23 rule is explained by the requirement for the formation of a stable, paired synaptic complex during the initiation of VDJ recombination. This complex in vitro contains the RAG proteins, a 12RS, and a 23RS (5). DNA cleavage by the RAG proteins in vitro or in vivo will occur only if a 12/23RS synaptic complex is assembled (6, 7). The 12/23 rule prevents rearrangement of V or J genes within their own clusters and ensures the obligatory inclusion of a D segment during IgH gene recombination, because the VH and JH genes are both flanked by 23RS, and the DH genes are flanked by 12RS. However, in the TCRβ and TCRδ loci, where D genes are also found, direct V to J joining would be permitted by the 12/23 rule (see Fig. 1 B). Because V-J joints are rare in either loci (reviewed in Ref.8), it has been postulated that mechanisms beyond the 12/23 rule must enforce D segment incorporation into TCR chains (reviewed in Ref.9). The molecular mechanism of beyond 12/23 regulation has been studied only in the TCRβ locus where it appears to be mediated by unique interactions between the Rag proteins and certain combinations of 12/23RS pairs (10, 11, 12, 13, 14). Functionally, incorporation of the D segment appears to be critical to generate TCRβ chains with normal length of the third CDR (CDR3) (15).

FIGURE 1.

The pattern of endogenous TCRδ D1-D2 rearrangements in mouse thymocytes in vivo. A, D1 to D2 rearrangements are PCR-amplified from sorted mouse TN thymocyte subsets (TN1–4). Undigested and ApaLI-digested products are shown. RAG-2 gene amplification served as a control of DNA quantity and quality. The positions of the diverse D1-D2 coding joints (CJ), the signal joint (SJ), and the ApaLI-digested fragments are shown (left). Numbers (right) indicate m.w. marker in base pairs. B, Schematic representation of the mouse TCRδ locus and the pattern of possible D-D rearrangements. The creation of the ApaLI restriction site (underlined) upon signal joint formation is shown. Open/dotted triangles indicate 12RS, and closed/shaded triangles indicate 23RS. Boxes represent the gene segments, and arrows indicate the position of the primers used for PCR amplification.

FIGURE 1.

The pattern of endogenous TCRδ D1-D2 rearrangements in mouse thymocytes in vivo. A, D1 to D2 rearrangements are PCR-amplified from sorted mouse TN thymocyte subsets (TN1–4). Undigested and ApaLI-digested products are shown. RAG-2 gene amplification served as a control of DNA quantity and quality. The positions of the diverse D1-D2 coding joints (CJ), the signal joint (SJ), and the ApaLI-digested fragments are shown (left). Numbers (right) indicate m.w. marker in base pairs. B, Schematic representation of the mouse TCRδ locus and the pattern of possible D-D rearrangements. The creation of the ApaLI restriction site (underlined) upon signal joint formation is shown. Open/dotted triangles indicate 12RS, and closed/shaded triangles indicate 23RS. Boxes represent the gene segments, and arrows indicate the position of the primers used for PCR amplification.

Close modal

The TCRδ chain is part of the γδTCR, which is expressed and functions exclusively in γδ T cells. The role and recognition characteristics of the γδTCR are not well understood, although it is clearly different from the αβTCR in several aspects (16). Most γδT cell clones are not restricted by MHC Ags and do not require peptide processing and presentation (17). Instead, they frequently recognize nonpeptidic ligands derived from bacterial cell wall components or plants (18, 19). Structurally, the γδTCR is more similar to the Ig than the αβTCR (20, 21), in part, due to the presence of an extended CDR3 in the TCRδ chain (22). This CDR3 is encoded by two D genes, which are joined together in nearly all functionally rearranged TCRδ genes. The organization of the TCRδ D and J gene segments and the orientation of the flanking RS are highly conserved and similar to the TCRβ locus (23). Rearrangement of the two D genes is facilitated by the alternating appearance of 12RS and 23RS (see Fig. 1,B). However, this organization could also allow deletion of one or both D genes during TCRδ recombination, the latter event resulting in the retention of a nonfunctional signal joint on the chromosome (Ref.24 and see Fig. 1 B). Similar, atypical rearrangements were first described in the IgH locus of a pre-B cell line and were called “pseudonormal” recombination (25). Surprisingly, in contrast to the paradigm established with the TCRβ locus, we found that the Rag proteins would preferentially delete, rather than incorporate, the two Dδ genes. Our observations demonstrate that selective D gene incorporation may be not only beyond the 12/23 rule, but also may be beyond the control mediated by direct Rag protein-RS interactions.

The basic structure of the competitive recombination substrates has been described previously (14, 26) and is shown on Fig. 2,D. For these studies, ∼500 bp of the murine Dδ1 and 180 bp of the murine Dδ2 genomic DNA sequences were PCR-amplified and cloned into a CMV promoter-based plasmid. Two substrates were generated; one with the endogenous orientation (substrate 1) (S1) and one with both gene segments inverted (substrate 2) (S2) (see Fig. 2,D). In addition, a custom m.w. marker was created by digesting one of the substrate plasmids with restriction enzymes ApaLI, AseI, BamHI, NcoI, NdeI, and NheI. This combination of digests results in a number of fragments, some of which hybridize to the probe used to detect recombination in the transfection assays, and allow us to measure precisely the size of the PCR-amplified bands of the rearranged substrates (see Fig. 2).

FIGURE 2.

The pattern of Dδ1-Dδ2 rearrangements of recombination substrates ex vivo after transfection into HEK293 cells. A, Southern blot analysis of PCR products generated after cotransfection of the indicated substrates (Subst) and full-length (f) or core (c) Rag proteins. The positions of the predicted D-D coding (CJ) or signal joints (SJ) are shown (right). Numbers (left) indicate m.w. marker in base pairs. The results of two of five independent transfections are shown. B, ApaLI digestion of gel-purified PCR products derived from Fig. 2A. ApaLI-digested (+) and mock-digested (−) samples are shown. Because both unrearranged S1 and S2 contain one endogenous ApaLI site, complete digestion of the unrearranged substrates (first two lanes, see asterisks marking the position of the digested bands) was used to verify the efficiency of the assay. Note that this, germline, ApaLI site is retained in S1, but not in S2, after rearrangement, resulting in two smaller restriction fragments without the de novo ApaLI site (CJ-type rearrangement; see arrowheads) and two more smaller fragments if the de novo ApaLI site is present (SJ-type rearrangement; see bracket). Other markers are the same as in Figs. 1 A and 2A. C, Western blot analysis of transient transfections of HEK293 cells. Polyclonal α-Rag-1 and Rag-2 (top) and monoclonal α-myc-epitope tag (bottom) Abs were used to detect full-length (f) and core (c) Rag proteins, as indicated. Only the core proteins are myc-epitope tagged, and α-Rag-1 recognizes only full-length, but not core, Rag-1. D, Schematic structure of the two recombination substrates with the possible rearranged products and their size are indicated. Arrowhead marks the position of an endogenous, germline ApaLI site. The creation of the de novo ApaLI restriction site (underlined) upon signal joint formation and the lack of ApaLI site upon coding joint formation (dots represent varying number of nucleotides) are illustrated. At the bottom, the position of the predicted restriction fragments of the unrearranged and rearranged substrates (sub) are shown with (+) or without (−) ApaLI digestion.

FIGURE 2.

The pattern of Dδ1-Dδ2 rearrangements of recombination substrates ex vivo after transfection into HEK293 cells. A, Southern blot analysis of PCR products generated after cotransfection of the indicated substrates (Subst) and full-length (f) or core (c) Rag proteins. The positions of the predicted D-D coding (CJ) or signal joints (SJ) are shown (right). Numbers (left) indicate m.w. marker in base pairs. The results of two of five independent transfections are shown. B, ApaLI digestion of gel-purified PCR products derived from Fig. 2A. ApaLI-digested (+) and mock-digested (−) samples are shown. Because both unrearranged S1 and S2 contain one endogenous ApaLI site, complete digestion of the unrearranged substrates (first two lanes, see asterisks marking the position of the digested bands) was used to verify the efficiency of the assay. Note that this, germline, ApaLI site is retained in S1, but not in S2, after rearrangement, resulting in two smaller restriction fragments without the de novo ApaLI site (CJ-type rearrangement; see arrowheads) and two more smaller fragments if the de novo ApaLI site is present (SJ-type rearrangement; see bracket). Other markers are the same as in Figs. 1 A and 2A. C, Western blot analysis of transient transfections of HEK293 cells. Polyclonal α-Rag-1 and Rag-2 (top) and monoclonal α-myc-epitope tag (bottom) Abs were used to detect full-length (f) and core (c) Rag proteins, as indicated. Only the core proteins are myc-epitope tagged, and α-Rag-1 recognizes only full-length, but not core, Rag-1. D, Schematic structure of the two recombination substrates with the possible rearranged products and their size are indicated. Arrowhead marks the position of an endogenous, germline ApaLI site. The creation of the de novo ApaLI restriction site (underlined) upon signal joint formation and the lack of ApaLI site upon coding joint formation (dots represent varying number of nucleotides) are illustrated. At the bottom, the position of the predicted restriction fragments of the unrearranged and rearranged substrates (sub) are shown with (+) or without (−) ApaLI digestion.

Close modal

CD3/CD4/CD8 triple negative (TN) thymocytes were isolated from pools of young, adult C57BL/6 mice, with two rounds of magnetic cell depletion following staining by specific Abs (27). TN cells were further fractionated according to the TN1–4 designation (28) using Abs specific for CD24, CD25, and CD44. All sorted cell subsets were >95–98% pure as determined with flowcytometric reanalysis. Agarose gel-embedded, total genomic DNA was prepared from the purified thymocyte subsets as described previously (29). Human thymus samples were obtained from cardiac surgery patients according to protocols approved by the Oklahoma Medical Research Foundation. CD4+ immature single positive (ISP) cells were prepared by magnetic depletion of CD3+ and CD8+ cells followed by flowcytometric sorting of CD4+ cells. Crude deproteinized extracts (human CD4 ISP) and purified or agarose-embedded DNA were PCR-amplified using locus-specific primers (Table I) for 30–34 cycles. ApaLI digestion was performed on column-purified PCR products (Qiagen) after adjusting the buffer conditions, with (+) or without (−) 5 U of ApaLI restriction endonuclease (New England Biolabs) at 37°C for 3 h. Reactions were separated on sieving agarose gel electrophoresis, hybridized to specific, radiolabeled probes, and analyzed on a PhosphorImager (Molecular Dynamics). For detection of signal end breaks, anchor ligation-mediated PCR was performed using locus-specific and anchor-specific primers after T4 ligase-mediated ligation of a specific double-strand oligonucleotide to genomic DNA, as described previously (30).

Table I.

DNA sequence of the primers used in this study. Artificial restriction enzyme (RE) sites are indicated

NamePrimer SequenceRE SiteReference
5Dd1F1 TATGGATCCAAGTTTGATAACAGGTGA BamHI This study 
3Dd1R1 GCTGGATCCATATTGAGGAATGTGATT BamHI This study 
5Dd1SE TGGGATCCTGAGTTTTAGGACTCT BamHI This study 
3Dd1SE GTGGATCCGCTTGATCAATATTGAGG BamHI This study 
5Dd2F TGTGGATCCACATGCAGAAAACACCTG BamHI This study 
5Jd1R TACGAATTCAGCTGGCTCATCATGACTTAAC EcoRI This study 
5LP23 TATGGATCCTCTGACCTTAAGGCTAC BamHI This study 
3Dd2R CAGGGATCCTGGGAGACGGTTCTTCACCCT BamHI This study 
5Vd4 AGCGGATCCGGCCGTTCTTCCTGT BamHI (59
3Jd1C CGTGGATCCACAGTCACTTGGGTT BamHI This study 
3Vd5F2 TGGGGATCCACAGAGATAAAAACGCTGAC BamHI This study 
5RSS GACGGATCCGCCCCATTGACGCA BamHI (14
3RSS2 GCTATTGCTTTATTTGTAACCA  This study 
hDd2F1 ATCGGATCCACATTGGGAGTGTCAACA BamHI (40
hDd3F1 TCAGGATCCATATAGTGTGAAACCGA BamHI This study 
hDd3R1 GGTGAATTCAACTTCCTGCTATCCCTT EcoRI This study 
hJd1R1 CCAGAATTCACCTCTTCCCAGGAGTCCT EcoRI This study 
NamePrimer SequenceRE SiteReference
5Dd1F1 TATGGATCCAAGTTTGATAACAGGTGA BamHI This study 
3Dd1R1 GCTGGATCCATATTGAGGAATGTGATT BamHI This study 
5Dd1SE TGGGATCCTGAGTTTTAGGACTCT BamHI This study 
3Dd1SE GTGGATCCGCTTGATCAATATTGAGG BamHI This study 
5Dd2F TGTGGATCCACATGCAGAAAACACCTG BamHI This study 
5Jd1R TACGAATTCAGCTGGCTCATCATGACTTAAC EcoRI This study 
5LP23 TATGGATCCTCTGACCTTAAGGCTAC BamHI This study 
3Dd2R CAGGGATCCTGGGAGACGGTTCTTCACCCT BamHI This study 
5Vd4 AGCGGATCCGGCCGTTCTTCCTGT BamHI (59
3Jd1C CGTGGATCCACAGTCACTTGGGTT BamHI This study 
3Vd5F2 TGGGGATCCACAGAGATAAAAACGCTGAC BamHI This study 
5RSS GACGGATCCGCCCCATTGACGCA BamHI (14
3RSS2 GCTATTGCTTTATTTGTAACCA  This study 
hDd2F1 ATCGGATCCACATTGGGAGTGTCAACA BamHI (40
hDd3F1 TCAGGATCCATATAGTGTGAAACCGA BamHI This study 
hDd3R1 GGTGAATTCAACTTCCTGCTATCCCTT EcoRI This study 
hJd1R1 CCAGAATTCACCTCTTCCCAGGAGTCCT EcoRI This study 

Transient transfections of the human embryonic kidney (HEK) 293 cells were conducted with 1 μg of substrate and 5–10 μg of each RAG expression plasmids (31) using the Ca phosphate-precipitation method (32). Both full-length and truncated, core Rag-expression plasmids were used, with or without myc-epitope tag, and gave the same results. The transfected DNA was recovered after 68 h using the alkaline lysis protocol of a commercial plasmid miniprep kit (Sigma-Aldrich). High m.w. genomic DNA was prepared using phenol/chloroform extraction. PCR amplification was performed on one-fortieth of the miniprep DNA of the transfections for 30 cycles using primer pairs specific for the transfected substrate (Table I). Restriction digestion of recombination substrates was performed after gel purification (Qiagen) to separate the unrearranged and rearranged PCR products. After adjusting the buffer conditions, the purified PCR products were incubated with (+) or without (−) 5 U of ApaLI restriction endonuclease (New England Biolabs) at 37°C for 3 h. Reactions were separated on sieving agarose gel electrophoresis, hybridized to a specific, radiolabeled probe, and analyzed on a PhosphorImager. Endogenous TCRδ rearrangements were PCR-amplified on one-twentieth of the total genomic DNA with primer pairs specific for the endogenous human TCRδ locus (Table I) and visualized using a FluorImager (Molecular Dynamics), after staining the gels with SYBR Green dye (Molecular Probes). ApaLI digestion was performed on the entire PCRs, without gel purification, due to the limiting amount of DNA.

The 5 × 104 cell-equivalent lysates were separated through 8% SDS-PAGE, blotted onto polyvinylidene difluoride membrane (Millipore), and incubated with a mixture of polyclonal, rabbit anti-Rag-1 (a gift from Dr. D. Schatz, Yale University, New Haven, CT) and anti-Rag-2 (BD Pharmingen) or monoclonal anti-myc-epitope tag (9E10, Cymbus Biotechnology) Abs. Blots were developed using an ECL detection kit (Amersham Biosciences). Western blots were performed routinely on every transfection experiment to ensure consistent expression of the Rag proteins (see Fig. 2 C).

The 12/23 rule permits two types of rearrangement of the two D genes in the mouse TCRδ locus: coding joint formation directed by the internal 12/23RS pairs or deletion of the two D genes by signal joint formation between the two external RS (Fig. 1,B). The two events can be distinguished by using high-resolution gel electrophoresis and ApaLI restriction digestion that can cleave only the typically precise signal joints but not the coding joints (33). D1 to D2 rearrangements were monitored using PCR in four successive stages of early mouse thymic development (28) of purified TN1–4 thymocyte subsets. D1 to D2 coding joint formation is expected to generate products of varying length of ∼226–246 bp. Precise signal joint between the two external RS is expected to generate a single product of 209-bp length. D1 to D2 rearrangements are weakly detectable in TN1 cells, are most prominent in TN2 cells, and decline in TN3 and TN4 thymocytes (Fig. 1,A), in accordance with previous findings that demonstrated the sequential progression of TCR gene recombination during mouse thymocyte development (34, 35). Using high-resolution sieving agarose electrophoresis, we could detect a predominant, larger, broader, and a slightly smaller, sharper band (Fig. 1,A). ApaLI restriction digestion confirmed that the shorter, but not the longer, products contained an internal ApaLI site, which is diagnostic for the precise fusion between the heptamers of the two external RS (Fig. 1,B). Quantitative analysis shows that in TN2 cells, ∼90% of the entire PCR is ApaLI-resistant and only 10% is ApaLI-sensitive (Fig. 1,A). The relative amount of ApaLI-sensitive products is higher in the TN3 and TN4 populations (24) and even more dominant in TN1 cells (Fig. 1 A and see Discussion). These results demonstrate that in vivo, at the TN2 stage of mouse thymocyte development, where TCRδ recombination becomes prominent (34, 35), D1 to D2 rearrangements form predominantly coding joints. Importantly, only D1-D2 coding joints, but not the signal joints that delete the two D genes, can continue to generate the typical, VDDJδ joints that are found in mature T cells.

To determine whether the Rag proteins are sufficient to enforce the predominant pattern of D1 to D2 rearrangements seen in vivo, we transiently transfected the nonlymphoid HEK293 cell line with expression constructs of the murine RAG genes (Fig. 2,C) along with extrachromosomal recombination substrates, which contain parts of the endogenous Dδ1 and Dδ2 genes, including the flanking RS and the entire coding segments (Fig. 2,D). Transient transfection of the entire germline D1-D2 region is not feasible because it is >8800-bp long. We have created two recombination substrates: S1 contains the two genes in their endogenous configuration, whereas S2 contains them in inverted orientation (Fig. 2,D). These substrates enable us to determine whether either the internal or the external RS pairs show preferential rearrangement and provide intrinsic controls for the potential bias of coding vs signal joint formation on recombination substrates (36). High-resolution agarose gel electrophoresis and ApaLI digestion of the PCR-amplified substrates can distinguish between coding and signal joint products as shown above (Fig. 1). Transfection of either S1 or S2 results in a single, predominant product, which migrates in a position that is consistent with a signal or coding joint rearrangement, respectively (Fig. 2,A). Indeed, >90% of S1 forms an ApaLI-sensitive signal joint, whereas >80% of S2 forms an ApaLI-resistant coding joint (Fig. 2,B). This results in exactly the opposite to that seen in vivo, where the endogenous locus (represented by S1 in the transfections) undergoes predominantly coding joint formation. This effect is not due to preferential signal joint formation on extrachromosomal substrates because the same biases are apparent in S2, which forms mainly coding joints instead of signal joints (Fig. 2,B). It should be noted that S2 does appear to rearrange less efficiently in repeated experiments presumably, due to the reportedly less efficient formation of coding joints on extrachromosomal substrates (36). Nonetheless, even this bias cannot compensate effectively for the dominant use of the 5′D1 and 3′D2 RS pair. We conclude that the Rag proteins, operating on extrachromosomal recombination substrates, mediate preferential recombination between the external 5′D1 and 3′D2 RS pairs and would delete, rather than join, the two D gene segments (see S1 in Fig. 2).

In the human TCRδ locus, the D2/D3 and J1/J3 genes are the orthologs of the mouse D1/D2 and J1/J2 genes, respectively (Ref. 37 and Fig. 3,B). The human D1 and J2/J3 genes are used rarely in typical adult TCRδ rearrangements (38). We performed PCR assays to detect D2 to D3 and D2/D3 to J1 rearrangements in sorted, CD4 ISP and total human thymocyte samples. Human CD4 ISP cells correspond approximately to mouse TN3 thymocytes and are expected to have a significant amount of partial TCRδ locus recombination products (39). We could readily identify partial D2-D3, D3-J1, and D2-J1 rearrangements in both populations (Fig. 3,A). ApaLI digestion of the D2-D3 rearranged products indicates that most of these rearrangements are ApaLI-resistant coding joints, in accordance with the findings in primary mouse thymocytes (compare Figs. 1,A vs 3,A). Low level V2-D3 and D2-D3 rearrangements of the endogenous TCRδ locus were also reported previously after transfection of Rag proteins into BOSC cells, a derivative of the HEK293 cell line (40). This observation prompted us to determine whether the endogenous chromosomal sequences behave similarly to the transfected recombination substrates in nonlymphoid cells. We also have found low level D2-D3 and D3-J1, but not D2-J1, rearrangements after transient overexpression of RAG-1 and RAG-2 alone in HEK293 cells (Fig. 3,C). Importantly, predominantly ApaLI-sensitive D2-D3 PCR products were detected, indicating that the majority of D2-D3 rearrangements were precise, signal joint of the external RS that deleted the two D genes (Fig. 3 D). Similarly, DNA sequencing also demonstrated the presence of only signal joint rearrangements in BOSC cells, irrespective of the cotransfection of E2A or HEB transcription factors (40).

FIGURE 3.

A, The pattern of endogenous TCRδ gene rearrangements in human thymocytes in vivo. Partial D2-D3, D3-J, and D2-J1 rearrangements are PCR-amplified from sorted CD4 ISP or total (tot) human (hu) thymocytes. For D2-D3 rearrangements, both undigested and ApaLI digested products are shown. HEK293 (293) cells served as negative control where only the germline DNA between the D3 and J1 genes can be amplified. The positions of the predicted rearranged and germline fragments are shown (right). Numbers (left) indicate m.w. marker in base pairs. B, Schematic representation of the human TCRδ locus and the pattern of possible D2-D3 coding and signal joint rearrangements are shown. Symbols are as in Fig. 1,B. C, The pattern of endogenous, chromosomal TCRδ gene rearrangements ex vivo after transfection of HEK293 cells. FluorImager analysis of PCR products generated after transfection of full-length (f) or core (c) Rag proteins. Each lane represents an independent transfection from a total of eight independent experiments, using full-length or core Rag proteins, as indicated. Markers are the same as in Fig. 3A. D, ApaLI digestion of the PCR products derived from Fig. 3C. Markers are the same as in Figs. 1,A and 2 A.

FIGURE 3.

A, The pattern of endogenous TCRδ gene rearrangements in human thymocytes in vivo. Partial D2-D3, D3-J, and D2-J1 rearrangements are PCR-amplified from sorted CD4 ISP or total (tot) human (hu) thymocytes. For D2-D3 rearrangements, both undigested and ApaLI digested products are shown. HEK293 (293) cells served as negative control where only the germline DNA between the D3 and J1 genes can be amplified. The positions of the predicted rearranged and germline fragments are shown (right). Numbers (left) indicate m.w. marker in base pairs. B, Schematic representation of the human TCRδ locus and the pattern of possible D2-D3 coding and signal joint rearrangements are shown. Symbols are as in Fig. 1,B. C, The pattern of endogenous, chromosomal TCRδ gene rearrangements ex vivo after transfection of HEK293 cells. FluorImager analysis of PCR products generated after transfection of full-length (f) or core (c) Rag proteins. Each lane represents an independent transfection from a total of eight independent experiments, using full-length or core Rag proteins, as indicated. Markers are the same as in Fig. 3A. D, ApaLI digestion of the PCR products derived from Fig. 3C. Markers are the same as in Figs. 1,A and 2 A.

Close modal

Collectively, these results demonstrate that overexpression of the Rag proteins in a nonlymphoid cell line results in deletion of the two Dδ genes and signal joint formation between the external RS pair, both on extrachromosomal substrates and on the endogenous, chromosomal locus. The data also show that, in contrast to nonlymphoid cells, the endogenous TCRδ locus undergoes predominantly coding joint formation in both murine and human primary thymocytes.

Deletion of the two TCRδ D genes can occur only on chromosomes with germline configuration, because either V-D or D-J partial rearrangements would remove one of the external RS flanking the D genes (see Discussion). Although D to D rearrangement clearly occurs in both mice and humans, if these were only extremely minor products during thymic development, the physiological significance of biased RS usage would be diminished greatly. To assess the relative proportions of the possible partial TCRδ rearrangements during early thymocyte development, we performed PCR assays on purified mouse TN1–4 subsets. As expected (34, 35), few partial or complete rearrangements are found in TN1 cells (Fig. 4). Substantial amounts of partial, D2-J1 and D1-D2 rearrangements are seen in TN2 cells. These partial rearrangements are reduced in later stages at the expense of partial D1-(D2)J1 and complete V4-DDJ1 rearrangements (Fig. 4). We could detect only very few V4-D1 partial rearrangements in TN3 cells, a type of rearrangement thought to occur more frequently during human T cell development (41).

FIGURE 4.

Progression of TCRδ locus recombination during early mouse thymocyte development. Southern blot hybridization and ethidium bromide staining (RAG-2 control) of PCR amplifications of mouse TN1–4 thymocyte and negative control kidney (K) DNA. The predicted positions of the specific rearrangements are shown (left). Numbers (right) indicate m.w. marker in base pairs. Note that for D2-J1, but not for the other rearrangements, the larger germline (gl) DNA also can be amplified.

FIGURE 4.

Progression of TCRδ locus recombination during early mouse thymocyte development. Southern blot hybridization and ethidium bromide staining (RAG-2 control) of PCR amplifications of mouse TN1–4 thymocyte and negative control kidney (K) DNA. The predicted positions of the specific rearrangements are shown (left). Numbers (right) indicate m.w. marker in base pairs. Note that for D2-J1, but not for the other rearrangements, the larger germline (gl) DNA also can be amplified.

Close modal

To assess more precisely the accessibility of TCRδ gene segments for recombination, we monitored the appearance of RS end breaks, an indicator of ongoing gene rearrangement (30, 33), using an anchor ligation-mediated PCR assay (42). We found a significant number of breaks only at the two RS flanking the D2 gene in TN2 cells and increasing amount of breaks of the 3′D1, 5′D1, 5′J1, and 3′V RS in TN3 cells (Fig. 5,A). TN4 cells, many of which are actively undergoing cell division after pre-TCR signaling (43), contain very few breaks of any RS in accordance with their more advanced developmental as well as higher proliferation status (42). These results indicate that the order of TCRδ locus recombination, similarly to other Ig and TCR loci, also exhibits tight developmental control. It appears that, in mouse, the D1 and D2 gene segments become accessible first, as shown by the predominant accumulation of 5′/3′D2 signal end breaks in TN2 cells (Fig. 5) and by the accumulation of partial D1-D2 and D2-J1 rearrangements (Fig. 4). This is followed by D1(D2)-J1 recombination and, later, by complete VDJδ rearrangements as the 5′D1 and 3′V RS become accessible in TN3 cells (Fig. 5 A). The absence of 5′J1 signal end breaks in TN2 cells may be due to the lower efficiency of detection of this break because it generates the largest PCR product of all the ligation-mediated PCR.

FIGURE 5.

A, Accumulation of TCRδ locus signal end (SE) breaks during early mouse thymocyte development. Anchor ligation-mediated PCR amplification of signal end breaks of the D1, D2, J1, and Vδ5 genes. DNA from mouse TN1–4 thymocytes was ligated to a double-strand anchor with T4 ligase, PCR-amplified, and hybridized to specific probes. Amplification of the single copy RAG-2 gene served as a loading control. The 5-fold serial dilutions are shown. The last lane shows a negative control with undiluted, unligated TN3 DNA. The positions of the specific products are marked with arrows (left). B, DNA sequence comparison of the RS of the mouse (m) and human (h) TCRδ D genes. The consensus (Cons) heptamer and nonamer motifs (3 ) are also shown. Dashes indicate nucleotide identity. Human D2 and D3 are the orthologs of mouse D1 and D2, respectively. The arrow indicates the highly unusual variation in the sixth position of the nonamer of the 3′D1 (mouse)/3′D2 (human) 23RS.

FIGURE 5.

A, Accumulation of TCRδ locus signal end (SE) breaks during early mouse thymocyte development. Anchor ligation-mediated PCR amplification of signal end breaks of the D1, D2, J1, and Vδ5 genes. DNA from mouse TN1–4 thymocytes was ligated to a double-strand anchor with T4 ligase, PCR-amplified, and hybridized to specific probes. Amplification of the single copy RAG-2 gene served as a loading control. The 5-fold serial dilutions are shown. The last lane shows a negative control with undiluted, unligated TN3 DNA. The positions of the specific products are marked with arrows (left). B, DNA sequence comparison of the RS of the mouse (m) and human (h) TCRδ D genes. The consensus (Cons) heptamer and nonamer motifs (3 ) are also shown. Dashes indicate nucleotide identity. Human D2 and D3 are the orthologs of mouse D1 and D2, respectively. The arrow indicates the highly unusual variation in the sixth position of the nonamer of the 3′D1 (mouse)/3′D2 (human) 23RS.

Close modal

Several mechanisms shape the primary AgR repertoire during lymphocyte development. Recently, multiple examples have been presented where specific interactions between the Rag proteins and RS motifs dominantly influenced the choice of gene segment usage (12, 13, 14, 26, 44). We hypothesized that similar Rag protein-dependent mechanisms could govern incorporation of the two D gene segments into the TCRδ chain. Surprisingly, however, our transfection experiments demonstrate that the Rag proteins, by themselves, are insufficient to enforce TCRδ D gene incorporation. This observation stands in striking contrast to the TCRβ locus where isolated RS/coding end combinations and their interactions with the Rag proteins were sufficient to reproduce the specific patterns of both V-D and D-J recombination (12, 13, 14, 26).

Two possible mechanisms could ensure preferential incorporation of the mouse D1-D2 (human D2-D3) genes during TCRδ locus recombination. One mechanism is based on the sequential order of TCRδ locus recombination, which we would term “preemptive utilization” of RS. In this case early, preferential D2 to J1 rearrangements would use up the 3′D2 RS, preventing it from participating in the D gene-deleting signal joint formation with the 5′D1 RS. Our analysis on the kinetics of TCRδ locus recombination indicates that such D2-J1 rearrangements do occur at significant levels in immature murine thymocytes (Figs. 4 and 5). In humans (41), although not in mouse (see Fig. 4), early V to D1 partial rearrangements may function also in a similar manner by removing the 5′ D1 RS before it could become a target for signal joint rearrangement. In either case, such sequential events wold ensure that at least one, and possibly both, D genes are incorporated into the rearranged TCRδ locus. Similar observations were made in the IgH locus where preferential accessibility of the 3′DH and 5′JH RS results in the typical deletional, rather than inversional, D-JH rearrangements (45). This bias also coincides with the reported preference of the Rag proteins to the 3′DH, instead of 5′DH, RS (46), which may play an important role in promoting biased recombination of the 3′DH RS over the 5′DH RS. In the TCRδ locus, a similar model would postulate that preferential interactions between the 3′D2 and 5′J1 RS (in mouse) or 3′V and 5′D1 RS (in humans) would result in preemptive rearrangements, and the consequential elimination of the 5′D1/3′D2 RS pair that could potentially delete the two D genes.

The second mechanism operates at the level of D-D gene rearrangement, which also occurs at significant levels in TN2 thymocytes. This mechanism must distinguish between the outside (5′D1/3′D2) and inside (3′D1/5′D2) RS pairs to regulate D gene incorporation. Our most unexpected finding is the discrepancy between the D-D rearrangement pattern observed in vivo (Figs. 1 and 3,A) and ex vivo (Figs. 2 and 3,C and 3,D). We believe that the result of the ex vivo transfection assay is not an artifact because sequence comparison of the conserved heptamer/nonamer motifs shows that the external, 5′D1/3′D2 RS, which are favored by the Rag proteins ex vivo, are more similar to the consensus (3) than the two internal RS (Fig. 5,B). Particularly striking is the sixth position variation in the nonamer of the mouse 3′D1/human 3′D2 RS (Fig. 5,B, arrow). Nonconsensus residues in this position have been shown to severely reduce recombination in various experimental assays (3, 47). The mere chromosomal context also can not explain our findings because the endogenous TCRδ locus in HEK293 cells also shows the same preferences as the extrachromosomal substrates (Fig. 3 and Ref.40). One possible hypothesis is that a lymphoid, perhaps T cell-specific, cofactor alters the targeting preferences of the Rag proteins specifically in favor of the less efficient RS pair. Such targeting could be achieved at the level of direct Rag protein binding to the DNA or by regulating the assembly and/or stability of the Rag-RS synaptic complex. Such a hypothetical cofactor could be involved specifically in the control of Dδ gene selection or in the general control of TCR gene segment usage in T cells. It could be expressed throughout thymic development or only in certain stages. Interestingly, we found significant differences in the relative proportion of D1-D2 coding vs signal joints in distinct subsets of TN thymocytes (Fig. 1,A). A relative decline of D1-D2 coding joints in TN3 and TN4 cells could be explained by the fact that only coding joints, but not the signal joints, can participate in complete VDDJδ rearrangements that appear predominantly from the TN3 stage (35). It is more intriguing to find even higher proportions, albeit low absolute levels, of signal joint products in TN1 cells. Part of this finding could be due to contamination by mature γδT cells, which share some phenotypic features with our TN1 preparation (Fig. 4 and Ref.48). However, the unusually high proportion of signal joint rearrangements could also indicate that D1-D2 recombination may initiate in genuine TN1 cells before or during the transition to TN2 differentiation. An intriguing proposition is that at this earliest stage, the hypothetical cofactor is not present, or not active yet, and rearrangement would proceed according to the preference of the already active Rag proteins (49, 50). This would result in deletion, rather than joining of the two D genes, similarly to what we see at the endogenous TCRδ locus in HEK293 cells (Fig. 3, C–D).

Alternatively, developmental regulation of RS preferences is achieved through the more traditional concept of accessibility control of gene rearrangement (51), which plays an important role in the overall patterning of TCRαδ locus recombination (reviewed in Ref.52). In this case, we would have to assume that chromosomal accessibility is blocked selectively or enhanced to the more or less efficient RS pairs, respectively. Such a fine control of accessibility, within the range of 11–16 bp, i.e., the distance of the 5′ and 3′RS motifs, has been proposed to operate on the DH genes (45), perhaps through remodeling of the nucleosomal phasing that can influence Rag protein-mediated cleavage in vitro (53, 54). The consensus, A-T rich nonamer has been shown to help position the RS onto the nucleosome (55), which typically would inhibit the access of Rag proteins to the RS (56, 57). It is possible that the less efficient 3′D1 RS, with a nonconsensus nonamer (Fig. 5 B), becomes relatively more accessible in vivo due to its less efficient positioning onto the nucleosome, although it must be emphasized that the published studies showed modest differences in accessibility and used RS sequences different from those studied in this article (55). Chromosomal accessibility is under the control of cis-acting elements, such as the TCRδ enhancer, which has been shown to be required for fully efficient TCRδ locus recombination (58). In TCRδ enhancer-deficient mice, relatively more partial D-J and D-D rearrangements accumulate, and it will be interesting to determine whether the proportion of D1-D2 coding vs signal joint rearrangements is altered compared with wild-type cells.

Although the importance of Rag protein-RS interactions in the control of gene segment selection remains significant, our studies highlight a hitherto unrecognized example of selective gene segment usage, where Rag protein preferences to certain RS are not simply modulated but directly opposed. Because the D-Jδ cluster is relatively simple, it could serve as the easiest model for identification of a predicted accessibility factor that overrides experimentally demonstrated Rag protein preferences for the RS.

We thank L. F. Thompson and J. L. Chain (Oklahoma Medical Research Foundation) for their generous gift of purified human thymus DNA, D. N. Patterson for assistance with cell culture, D. G. Schatz (Yale University) for the gift of the anti-Rag-1 Ab, and M. F. Flajnik (University of Maryland) for comments on the manuscript.

The authors have no financial conflict of interest.

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

1

This work was supported in part by funds from the National Institutes of Health (to H.T.P.).

3

Abbreviations used in this paper: RS, recombination signal; CDR3, third CDR; TN, CD3/CD4/CD8 triple negative; ISP, immature single positive; HEK, human embryonic kidney; S1, substrate 1; S2, substrate 2.

1
Fugmann, S. D., A. I. Lee, P. E. Shockett, I. J. Villey, D. G. Schatz.
2000
. The RAG proteins and V(D)J recombination: complexes, ends, and transposition.
Annu. Rev. Immunol.
18
:
495
-527.
2
Gellert, M..
2002
. V(D)J recombination: rag proteins, repair factors, and regulation.
Annu. Rev. Biochem.
71
:
101
-132.
3
Hesse, J. E., M. R. Lieber, K. Mizuuchi, M. Gellert.
1989
. V(D)J recombination: a functional definition of the joining signals.
Genes Dev.
3
:
1053
-1061.
4
Tonegawa, S..
1983
. Somatic generation of antibody diversity.
Nature
302
:
575
-581.
5
Hiom, K., M. Gellert.
1998
. Assembly of a 12/23 paired signal complex: a critical control point in V(D)J recombination.
Mol. Cell
1
:
1011
-1019.
6
Steen, S. B., L. Gomelsky, S. L. Speidel, D. B. Roth.
1997
. Initiation of V(D)J recombination in vivo: role of recombination signal sequences in formation of single and paired double-strand breaks.
EMBO J.
16
:
2656
-2664.
7
van Gent, D. C., D. A. Ramsden, M. Gellert.
1996
. The RAG1 and RAG2 proteins establish the 12/23 rule in V(D)J recombination.
Cell
85
:
107
-113.
8
Davis, M. M., P. J. Bjorkman.
1988
. T-cell antigen receptor genes and T-cell recognition.
Nature
334
:
395
-402.
9
Tillman, R. E., A. L. Wooley, M. M. Hughes, B. Khor, B. P. Sleckman.
2004
. Regulation of T-cell receptor β-chain gene assembly by recombination signals: the beyond 12/23 restriction.
Immunol. Rev.
200
:
36
-43.
10
Sleckman, B. P., C. H. Bassing, M. M. Hughes, A. Okada, M. D’Auteuil, T. D. Wehrly, B. B. Woodman, L. Davidson, J. Chen, F. W. Alt.
2000
. Mechanisms that direct ordered assembly of T cell receptor β locus V, D, and J gene segments.
Proc. Natl. Acad. Sci. USA
97
:
7975
-7980.
11
Bassing, C. H., F. W. Alt, M. M. Hughes, M. D’Auteuil, T. D. Wehrly, B. B. Woodman, F. Gartner, J. M. White, L. Davidson, B. P. Sleckman.
2000
. Recombination signal sequences restrict chromosomal V(D)J recombination beyond the 12/23 rule.
Nature
405
:
583
-586.
12
Jung, D., C. H. Bassing, S. D. Fugmann, H. L. Cheng, D. G. Schatz, F. W. Alt.
2003
. Extrachromosomal recombination substrates recapitulate beyond 12/23 restricted VDJ recombination in nonlymphoid cells.
Immunity
18
:
65
-74.
13
Tillman, R. E., A. L. Wooley, B. Khor, T. D. Wehrly, C. A. Little, B. P. Sleckman.
2003
. Cutting edge: targeting of Vβ to Dβ rearrangement by RSSs can be mediated by the V(D)J recombinase in the absence of additional lymphoid-specific factors.
J. Immunol.
170
:
5
-9.
14
Olaru, A., D. N. Patterson, I. Villey, F. Livak.
2003
. DNA-Rag protein interactions in the control of selective D gene utilization in the TCRβ locus.
J. Immunol.
171
:
3605
-3611.
15
Hughes, M. M., M. Yassai, J. R. Sedy, T. D. Wehrly, C. Y. Huang, O. Kanagawa, J. Gorski, B. P. Sleckman.
2003
. T cell receptor CDR3 loop length repertoire is determined primarily by features of the V(D)J recombination reaction.
Eur. J. Immunol.
33
:
1568
-1575.
16
Hayday, A. C..
2000
. γδ Cells: a right time and a right place for a conserved third way of protection.
Annu. Rev. Immunol.
18
:
975
-1026.
17
Chien, Y. H., R. Jores, M. P. Crowley.
1996
. Recognition by γδ T cells.
Annu. Rev. Immunol.
14
:
511
-532.
18
Constant, P., F. Davodeau, M. A. Peyrat, Y. Poquet, G. Puzo, M. Bonneville, J. J. Fournie.
1994
. Stimulation of human γ δ T cells by nonpeptidic mycobacterial ligands.
Science
264
:
267
-270.
19
Tanaka, Y., C. T. Morita, E. Nieves, M. B. Brenner, B. R. Bloom.
1995
. Natural and synthetic non-peptide antigens recognized by human γ δ T cells.
Nature
375
:
155
-158.
20
Allison, T. J., C. C. Winter, J. J. Fournie, M. Bonneville, D. N. Garboczi.
2001
. Structure of a human γδ T-cell antigen receptor.
Nature
411
:
820
-824.
21
Li, H., M. I. Lebedeva, A. S. Llera, B. A. Fields, M. B. Brenner, R. A. Mariuzza.
1998
. Structure of the Vδ domain of a human γδ T-cell antigen receptor.
Nature
391
:
502
-506.
22
Rock, E. P., P. R. Sibbald, M. M. Davis, Y. H. Chien.
1994
. CDR3 length in antigen-specific immune receptors.
J. Exp. Med.
179
:
323
-328.
23
Glusman, G., L. Rowen, I. Lee, C. Boysen, J. C. Roach, A. F. Smit, K. Wang, B. F. Koop, L. Hood.
2001
. Comparative genomics of the human and mouse T cell receptor loci.
Immunity
15
:
337
-349.
24
Carroll, A. M., J. K. Slack, X. Mu.
1993
. V(D)J recombination generates a high frequency of nonstandard TCR D δ-associated rearrangements in thymocytes.
J. Immunol.
150
:
2222
-2230.
25
Alt, F. W., D. Baltimore.
1982
. Joining of immunoglobulin heavy chain gene segments: implications from a chromosome with evidence of three D-JH fusions.
Proc. Natl. Acad. Sci. USA
79
:
4118
-4122.
26
Olaru, A., D. N. Patterson, H. Cai, F. Livak.
2004
. Recombination signal sequence variations and the mechanism of patterned T-cell receptor-β locus rearrangement.
Mol. Immunol.
40
:
1189
-1201.
27
Tourigny, M. R., S. Mazel, D. B. Burtrum, H. T. Petrie.
1997
. T cell receptor (TCR)-β gene recombination: dissociation from cell cycle regulation and developmental progression during T cell ontogeny.
J. Exp. Med.
185
:
1549
-1556.
28
Godfrey, D. I., J. Kennedy, T. Suda, A. Zlotnik.
1993
. A developmental pathway involving four phenotypically and functionally distinct subsets of CD3-CD4-CD8 triple negative adult mouse thymocytes defined by CD44 and CD25 expression.
J. Immunol.
150
:
4244
-4252.
29
Petrie, H. T., F. Livak, D. Burtrum, S. Mazel.
1995
. T cell receptor gene recombination patterns and mechanisms: cell death, rescue and T cell production.
J. Exp. Med.
182
:
121
-127.
30
Livak, F., D. G. Schatz.
1996
. T-cell receptor α locus V(D)J recombination by-products are abundant in thymocytes and mature T cells.
Mol. Cell. Biol.
16
:
609
-618.
31
Fugmann, S. D., I. J. Villey, L. M. Ptaszek, D. G. Schatz.
2000
. Identification of two catalytic residues in RAG1 that define a single active site within the RAG1/RAG2 protein complex.
Mol. Cell
5
:
97
-107.
32
Chen, C., H. Okayama.
1987
. High-efficiency transformation of mammalian cells by plasmid DNA.
Mol. Cell. Biol.
7
:
2745
-2752.
33
Roth, D. B., P. B. Nakajima, J. P. Menetski, M. J. Bosma, M. Gellert.
1992
. V(D)J recombination in mouse thymocytes: double-stranded breaks near T-cell receptor δ rearrangement signals.
Cell
69
:
41
-53.
34
Capone, M., R. D. Hockett, Jr, A. Zlotnik.
1998
. Kinetics of T cell receptor β, γ, and δ rearrangements during adult thymic development: T cell receptor rearrangements are present in CD44+CD25+ pro-T thymocytes.
Proc. Natl. Acad. Sci. USA
95
:
12522
-12527.
35
Livak, F., M. Tourigny, D. G. Schatz, H. T. Petrie.
1999
. Characterization of TCR gene rearrangements during adult murine T cell development.
J. Immunol.
162
:
2575
-2580.
36
Hesse, J. E., M. R. Lieber, M. Gellert, K. Mizuuchi.
1987
. Extrachromosomal DNA substrates in pre-B cells undergo inversion or deletion at immunoglobulin V-(D)-J joining signals.
Cell
49
:
775
-783.
37
Kubota, T., J. Wang, T. W. Gobel, R. D. Hockett, M. D. Cooper, C. H. Chen.
1999
. Characterization of an avian (Gallus gallus domesticus) TCR αδ gene locus.
J. Immunol.
163
:
3858
-3866.
38
McVay, L. D., S. S. Jaswal, C. Kennedy, A. Hayday, S. R. Carding.
1998
. The generation of human γδ T cell repertoires during fetal development.
J. Immunol.
160
:
5851
-5860.
39
Blom, B., M. C. Verschuren, M. H. Heemskerk, A. Q. Bakker, E. J. van Gastel-Mol, I. L. Wolvers-Tettero, J. J. van Dongen, H. Spits.
1999
. TCR gene rearrangements and expression of the pre-T cell receptor complex during human T-cell differentiation.
Blood
93
:
3033
-3043.
40
Langerak, A. W., I. L. Wolvers-Tettero, E. J. van Gastel-Mol, M. E. Oud, J. J. van Dongen.
2001
. Basic helix-loop-helix proteins E2A and HEB induce immature T-cell receptor rearrangements in nonlymphoid cells.
Blood
98
:
2456
-2465.
41
Lauzurica, P., M. S. Krangel.
1994
. Enhancer-dependent and -independent steps in the rearrangement of a human T cell receptor δ transgene.
J. Exp. Med.
179
:
43
-55.
42
Schlissel, M., A. Constantinescu, T. Morrow, M. Baxter, A. Peng.
1993
. Double-strand signal sequence breaks in V(D)J recombination are blunt, 5′-phosphorylated, RAG-dependent, and cell cycle regulated.
Genes Dev.
7
:
2520
-2532.
43
Hoffman, E. S., L. Passoni, T. Crompton, T. M. J. Leu, D. G. Schatz, A. Koff, M. J. Owen, A. C. Hayday.
1996
. Productive T cell receptor β chain gene rearrangement: coincident regulation of cell cycle and clonality during development in vivo.
Genes Dev.
10
:
948
-962.
44
Yu, K., A. Taghva, M. R. Lieber.
2002
. The cleavage efficiency of the human immunoglobulin heavy chain VH elements by the RAG complex: implications for the immune repertoire.
J. Biol. Chem.
277
:
5040
-5046.
45
Stanhope-Baker, P., K. M. Hudson, A. L. Shaffer, A. Constantinescu, M. S. Schlissel.
1996
. Cell type-specific chromatin structure determines the targeting of V(D)J recombinase activity in vitro.
Cell
85
:
887
-897.
46
Gauss, G. H., M. R. Lieber.
1992
. The basis for the mechanistic bias for deletional over inversional V(D)J recombination.
Genes Dev.
6
:
1553
-1561.
47
Akamatsu, Y., N. Tsurushita, F. Nagawa, M. Matsuoka, K. Okazaki, M. Imai, H. Sakano.
1994
. Essential residues in V(D)J recombination signals.
J. Immunol.
153
:
4520
-4529.
48
Zorbas, M., R. Scollay.
1993
. Development of γδ T cells in the adult murine thymus.
Eur. J. Immunol.
23
:
1655
-1660.
49
Igarashi, H., S. C. Gregory, T. Yokota, N. Sakaguchi, P. W. Kincade.
2002
. Transcription from the RAG1 locus marks the earliest lymphocyte progenitors in bone marrow.
Immunity
17
:
117
-130.
50
Borghesi, L., L. Y. Hsu, J. P. Miller, M. Anderson, L. Herzenberg, M. S. Schlissel, D. Allman, R. M. Gerstein.
2004
. B lineage-specific regulation of V(D)J recombinase activity is established in common lymphoid progenitors.
J. Exp. Med.
199
:
491
-502.
51
Yancopoulos, G. D., F. W. Alt.
1985
. Developmentally controlled and tissue-specific expression of unrearranged VH gene segments.
Cell
40
:
271
-281.
52
Krangel, M. S., J. Carabana, I. Abbarategui, R. Schlimgen, A. Hawwari.
2004
. Enforcing order within a complex locus: current perspectives on the control of V(D)J recombination at the murine T-cell receptor αδ locus.
Immunol. Rev.
200
:
224
-232.
53
Patenge, N., S. K. Elkin, M. A. Oettinger.
2004
. ATP-dependent remodeling by SWI/SNF and ISWI proteins stimulates V(D)J cleavage of 5 S arrays.
J. Biol. Chem.
279
:
35360
-35367.
54
Kwon, J., K. B. Morshead, J. R. Guyon, R. E. Kingston, M. A. Oettinger.
2000
. Histone acetylation and hSWI/SNF remodeling act in concert to stimulate V(D)J cleavage of nucleosomal DNA.
Mol. Cell
6
:
1037
-1048.
55
Baumann, M., A. Mamais, F. McBlane, H. Xiao, J. Boyes.
2003
. Regulation of V(D)J recombination by nucleosome positioning at recombination signal sequences.
EMBO J.
22
:
5197
-5207.
56
Golding, A., S. Chandler, E. Ballestar, A. P. Wolffe, M. S. Schlissel.
1999
. Nucleosome structure completely inhibits in vitro cleavage by the V(D)J recombinase.
EMBO J.
18
:
3712
-3723.
57
Kwon, J., A. N. Imbalzano, A. Matthews, M. A. Oettinger.
1998
. Accessibility of nucleosomal DNA to V(D)J cleavage is modulated by RSS positioning and HMG1.
Mol. Cell
2
:
829
-839.
58
Monroe, R. J., B. P. Sleckman, B. C. Monroe, B. Khor, S. Claypool, R. Ferrini, L. Davidson, F. W. Alt.
1999
. Developmental regulation of TCR δ locus accessibility and expression by the TCR δ enhancer.
Immunity
10
:
503
-513.
59
Livak, F., H. T. Petrie, I. N. Crispe, D. G. Schatz.
1995
. In-frame TCR δ gene rearrangements play a critical role in the αβγδ T cell lineage decision.
Immunity
2
:
617
-627.