The promoter sequences of individual murine TCR Vβ segments are dissimilar, but any functional differences between them are masked after productive gene rearrangement by the dominance of the TCRβ 3′ enhancer. However, thymocytes of recombination-activating gene-2 (Rag2)-deficient mice allow the transcriptional activity of Vβ promoters to be studied before rearrangement. Here we report that many Vβ segments are detectably transcribed in Rag2−/− thymocytes and that there are significant differences in expression among different Vβ segments. Primer extension and characterization of cDNA clones from SCID thymocytes suggest that these germline Vβ transcripts generally use the same start sites as those previously determined in mature T cells. The strength of expression before rearrangement does not correlate with proximity to the known enhancer, because members of the most distal Vβ cluster (Vβ2.1, Vβ1.1, Vβ4.1) are relatively strongly expressed and more proximal Vβ segments (Vβ14.1, Vβ3.1, Vβ7.1, Vβ6.1) are only weakly expressed. Different Vβ segments also show different developmental programs of activation in different thymocyte subsets, with the Vβ5.1(L)-8.2(V) spliced transcript expressed earliest as well as most strongly overall. Comparison with Rag+ MHC class I−/− and class II−/− thymocytes confirms that many of these expression differences are leveled by rearrangement and/or by β selection, before MHC-dependent selection. However, the expression pattern of Vβ2.1 is highly distinctive and includes cell types apparently outside the T lineage, suggesting potential acquisition of specialized roles.

The differentiation and maturation of T cells depends on the successful V(D)J rearrangement and expression of TCR. In mature T cells, the expression of TCRβ-chain is controlled by a 3′ Cβ enhancer and the relevant Vβ promoter. In general, the Cβ enhancer is a stronger regulator than the Vβ promoters, because the choice of Vβ segment used for rearrangement does not detectably affect the levels of TCRβ mRNA in a mature T cell (1, 2). Although the promoter sequences of different Vβ segments are quite divergent, there has been little evidence to date that they may also contribute differentially to the activity of these genes. Several regulatory elements have been identified in TCR Vβ promoter regions, but their functional importance is not well understood. This is because the relative strengths of Vβ promoters are usually masked both by TCR gene rearrangement at the single-cell level and by clonal selection at the T cell population level. However, the strengths of individual Vβ promoters can be revealed by analysis of TCR Vβ transcription before rearrangement.

Therefore, we have used thymocytes from recombinase activating gene-2 (Rag2)3-deficient (Rag2−/−) mice to study the regulatory properties of different TCR Vβ promoters. Because of the lack of recombinase, the TCR gene remains in germline configuration in these cells, with the Cβ enhancer hundreds of kilobases away. Although the Rag2−/− thymocytes are arrested as immature cells, they do reach a stage of development when normal thymocytes have become committed to the T cell lineage and are fully capable of transcribing the TCRβ-chain gene (3, 4, 5, 6). In this context, the basal levels of “germline” transcription from unrearranged Vβ segments can be measured sensitively. A few previous reports have indicated that transcripts from certain Vβ segments could be detected in such cells (7, 8). Here we show that in the absence of rearrangement, different Vβ segments are transcribed at distinctive levels and with distinctive fine-scale developmental regulation in the stages of lymphocyte differentiation encompassing T lineage specification.

The mice used in this study were C57BL/6 (B6), (B6, 129) Rag2−/−, B6.129P2-Tcrbtm1MomTcrdtm1Mom (TCR-β−/−δ−/−), B6-SCID, and B6-I-Abβ−/− β2-microglobulin−/− (B6 MHC−/−). Except as indicated, all of the animals were bred in our own facility, and immunodeficient mice were maintained under sterile conditions in isolators, with autoclaved food, bedding, and acidified water. In some cases, C.B-17-SCID mice were purchased from Taconic Laboratories (Germantown, NY). These SCID mice have a higher incidence of leakiness, as measured by the generation of CD4+CD8+ thymocytes, than the SCID animals bred at Caltech (Pasadena, CA).

Thymocytes and bone marrow cells were collected from ∼6- to 8-wk-old mice. For bone marrow samples, RBC were lysed in 1 ml RBC lysis buffer (155 mM NH4Cl, 10 mM KHCO3, 0.1 mM EDTA) by pipetting up and down for 3 min at room temperature. The lysis was stopped by adding 3 ml CBSS/BSA/azide (1.25× HBSS without phenol red, plus 2.5 mg/ml BSA and 0.03% sodium azide), and cells were washed one more time in the same buffer. For cell sorting, cells were incubated with supernatant from the 2.4G2 hybridoma for 20 min on ice to block Fc receptors before staining with fluorescent-conjugated Abs. FACS sorting was performed using a Coulter Epics Elite cell sorter (Coulter, Hialeah, FL). Stained cells were sorted into chilled tubes containing CBSS/BSA/azide with 5% FBS. A portion of each sorted sample was reanalyzed for purity. The rest of the cells were used for RNA extraction. All of the Abs used in this study were from BD PharMingen (San Diego, CA) or BD Biosciences (San Jose, CA).

To analyze fine-scale developmental control of Vβ expression, Rag2−/− thymocytes were stained with Sca-1(Ly-6A/E)-PE, heat stable Ag (HSA; CD24)-FITC, and CD44-CyChrome, and fractionated by FACS sorting as previously described (9, 10, 11, 12). The subpopulations collected were: 1) Sca-1+ HSA, which are stem-like cells that probably retain potential for T, B, NK, and DC lineages; 2) Sca-1 HSA, which are NK-like cells; 3) HSA+ CD44+, which are “pro-T cells” that are specified but not committed to the T lineage yet; and 4) HSA+ CD44, which are cells committed to the T lineage and undergoing TCRβ rearrangement. To obtain B cell precursor and other immature hemopoietic populations, Rag2−/− bone marrow cells were stained with B220-FITC, CD19-CyChrome, and CD43-PE. The subpopulations collected were: 1) B220+ CD43+ CD19+, which are pro-B cells; and 2) and B220 CD43+, which are a large, granular population likely to be enriched for myeloid lineage cells. Fetal thymocytes were collected from B6 mice at 14.5 days postcoitum (E14.5). Fetal liver cells were collected from B6 mice at E15.5.

Total cellular RNA was prepared using RNAzol B (Leedo Medical, Houston, TX), following manufacturer’s instructions. For sorted samples, 25 μg tRNA was used as carrier and added to the samples at the beginning of extraction. RNA samples were first treated with RNase-free DNase I for 30 min at 37°C to remove the remaining genomic DNA. Portions of each sample were removed and used as RNA quality controls for PCR amplification. First-strand cDNA was synthesized from total RNA using Superscript II (Life Technologies, Gaithersburg, MD) and oligo(dT)12–18. PCR amplification was done for 45 s at 94°C, 45 s at 55°C, and 1 min at 72°C. Thirty to 38 cycles were normally used. PCR products were run on 2% agarose gels and visualized by ethidium bromide staining. For quantification, 0.5 μCi [32P]dCTP was added in the PCR mix. Radioactive signals were quantified by PhosphorImager (Molecular Dynamics, Sunnyvale, CA) after electrophoresis and transfer of the DNA to Hybond-N+ membranes (Amersham, Little Chalfont, U.K.). TCR Vβ primers, which were designed for each Vβ segment to span the intron between the L and V exons, are listed in Table I. Genomic DNA of Rag2−/− thymocytes, which has equal numbers of copies of each Vβ segment, was used as a control to normalize PCR efficiency of different primer sets. All cDNA samples were normalized according to their levels of GAPDH RNA by dilution in 1× Tris/EDTA (TE) buffer. To keep all samples in the linear range for the Vβ-specific PCR, B6 and MHC−/− cDNAs were then diluted to 1/50 of the normalized Rag2−/− and SCID cDNA level, except as otherwise indicated.

Table I.

RT-PCR primers and oligonucleotide probes used in this study

NameSequence
PCR primers for rearranged TCR Vβ genes  
Vβ 8 S-CAAAACACATGGAGGCTGCAGTCA 
Vβ 4 S-GCAGGTCCAGTCGACCCGAAAAT 
Cβ A-GGGTGGAGTCACATTTCTCAGATC 
  
PCR primers for TCR Vβ segments (Lβ sense, Vβ antisense)  
Vβ 1.1 S-GGCTTCTCCTCTATGTTTCCCT 
 A-ACTGGGCACCGTCTCATTTCGA 
Vβ 2.1 S-GTGGCAGTTTTGCATTCTGTGC 
 A-CACCGGGAAGAGATTTGACCTC 
Vβ 3.1 S-GGCTACAAGGCTCCTCTGTTAC 
 A-GGCTGCAAGGTGAGTTTGAAGG 
Vβ 4.1 S-TGGGCTCCATTTTCCTCAGTTG 
 A-GGCAGTCTGATTGTCCATAAGT 
Vβ 5.1 S-GTCTAACACTGCCTTCCCTGAC 
 A-GAGTCTGTTGATACCAGGCCAC 
Vβ 5.2 S-ACTGTCCTCGCTGATTCTGCCT 
 A-TAGCAGAGTCCTCCAGTTCCAA 
Vβ 6.1 S-TGAACAAGTGGGTTTTCTGCTG 
 A-ATAGCCTTCAGATAGATCGCCT 
Vβ 7.1 S-TGAGAGTTAGGCTCATCTCTGC 
 A-AGACCCTGTATCCTTTAGGGAT 
Vβ 8.1 S-TGGGCTCCAGACTCTTCTTTGT 
 A-TTTCTCCGTGCTGTCAGCGA 
Vβ 8.2 S-CTCCAGTCTCCTGTGTTCAA 
 A-TCTGAGAGGGGGTAGCCAACTC 
Vβ 8.3 S-AGGCTCTTTCTGCTCTTGAGC 
 A-TCTGAGAGGGAGAAGCCAATTC 
Vβ 14.1 S-GCAGCTTCTCCGTGCTTAGGAT 
 A-CTTTCTCCTGGGCATGTTCTTG 
  
Probes for cDNA library screening and 5′-RACE  
O 5.1 5′-TAAAATGGAGAGAGATAAAGGAAACCTGCCCAGC 
O 8.2 5′-GTTGGCTACCCCCTCTCAGACATCAGTGTAC 
O 4.1 5′-CTCCCTGTGACTGCCACCAGATATTTTGGTTTCTG 
O 14.1 5′-AGTACCAGTAGAGGTTAGGGCTTGATTTCCCCTTT 
NameSequence
PCR primers for rearranged TCR Vβ genes  
Vβ 8 S-CAAAACACATGGAGGCTGCAGTCA 
Vβ 4 S-GCAGGTCCAGTCGACCCGAAAAT 
Cβ A-GGGTGGAGTCACATTTCTCAGATC 
  
PCR primers for TCR Vβ segments (Lβ sense, Vβ antisense)  
Vβ 1.1 S-GGCTTCTCCTCTATGTTTCCCT 
 A-ACTGGGCACCGTCTCATTTCGA 
Vβ 2.1 S-GTGGCAGTTTTGCATTCTGTGC 
 A-CACCGGGAAGAGATTTGACCTC 
Vβ 3.1 S-GGCTACAAGGCTCCTCTGTTAC 
 A-GGCTGCAAGGTGAGTTTGAAGG 
Vβ 4.1 S-TGGGCTCCATTTTCCTCAGTTG 
 A-GGCAGTCTGATTGTCCATAAGT 
Vβ 5.1 S-GTCTAACACTGCCTTCCCTGAC 
 A-GAGTCTGTTGATACCAGGCCAC 
Vβ 5.2 S-ACTGTCCTCGCTGATTCTGCCT 
 A-TAGCAGAGTCCTCCAGTTCCAA 
Vβ 6.1 S-TGAACAAGTGGGTTTTCTGCTG 
 A-ATAGCCTTCAGATAGATCGCCT 
Vβ 7.1 S-TGAGAGTTAGGCTCATCTCTGC 
 A-AGACCCTGTATCCTTTAGGGAT 
Vβ 8.1 S-TGGGCTCCAGACTCTTCTTTGT 
 A-TTTCTCCGTGCTGTCAGCGA 
Vβ 8.2 S-CTCCAGTCTCCTGTGTTCAA 
 A-TCTGAGAGGGGGTAGCCAACTC 
Vβ 8.3 S-AGGCTCTTTCTGCTCTTGAGC 
 A-TCTGAGAGGGAGAAGCCAATTC 
Vβ 14.1 S-GCAGCTTCTCCGTGCTTAGGAT 
 A-CTTTCTCCTGGGCATGTTCTTG 
  
Probes for cDNA library screening and 5′-RACE  
O 5.1 5′-TAAAATGGAGAGAGATAAAGGAAACCTGCCCAGC 
O 8.2 5′-GTTGGCTACCCCCTCTCAGACATCAGTGTAC 
O 4.1 5′-CTCCCTGTGACTGCCACCAGATATTTTGGTTTCTG 
O 14.1 5′-AGTACCAGTAGAGGTTAGGGCTTGATTTCCCCTTT 

The C.B-17-SCID thymocyte random-primed cDNA library was constructed in the pSPORT1 vector by Michele Anderson (12). The library was arrayed and spotted at high density onto Hybond-N+ membrane using the Q-Bot robot (Genetix, Christchurch, U.K.). DNA probes were labeled with [32P]dATP. Prehybridization, hybridization, and wash conditions were as described (12). The positive clones were selected and sequenced using the ABI Prism dye terminator cycle sequencing ready reaction kit (Perkin-Elmer, Foster City, CA) according to the manufacturer’s procedure.

Poly(A)+ RNA samples from SCID and normal B6 thymocytes were used for 5′-RACE analysis using a Marathon cDNA amplification kit (Clontech, Palo Alto, CA). First- and second-strand cDNA synthesis and adaptor ligation were performed following the manufacturer’s instructions. The 5′-Marathon RACE reactions were conducted using adaptor primer 1 (5′-CCATCCTAATACGACTCACTATAGGGC) and the appropriate 3′-gene specific primer as shown in Table I (antisense primers for Vβ4.1, Vβ8.2, and Vβ14.1). cDNA amplification conditions were: 94°C for 1 min, followed by 30 cycles at 94°C for 30 s, 60°C for 30 s, and 68°C for 3 min. The size of the PCR products was determined by running on a 2% agarose/ethidium bromide gel along with appropriate DNA size markers. To increase the sensitivity, PCR products were transferred to Hybond-N+ membrane after electrophoresis and probed with gene-specific radioactive probes ([γ-32P]ATP end-labeled Vβ4 and Vβ14 oligonucleotides, see Table I).

Thymocytes in Rag2−/− and SCID mutant mice are a well-characterized composite of immature T cell subpopulations. They undergo developmental arrest and death at the CD25+CD44 stage due to an inability to carry out TCR gene rearrangement. To progress beyond this “β selection checkpoint,” T cell precursors need to express the translation product of any successfully rearranged β-chain gene, which the mutant thymocytes cannot do. But because the stages of development up to the β selection checkpoint are well represented in these mice (9), they include the stages when TCR gene rearrangement would normally be initiated. Thus they are an excellent source of cells that activate the same set of transcription factors that may normally coincide with TCR gene rearrangement, without being able to move the Vβ genes from the germline configuration. We have used thymocytes from Rag2−/− and SCID mice in this work to investigate the extent to which Vβ genes are transcribed in T cell precursors in the absence of rearrangement.

To confirm that all TCR Vβ gene expression in Rag2−/− and SCID originates from DNA in a germline configuration, RT-PCR analysis was conducted to determine whether there is any transcription of rearranged TCR genes in Rag2−/− and SCID cDNA samples. PCR primer sets Vβ8-Cβ and Vβ4-Cβ were designed with the 5′ primers in the respective Vβ regions and the 3′ primer in the Cβ region. Without rearrangement, Vβ segments are several hundred kilobases away from the Cβ region, and no band will be detected in the PCR product. As expected, Vβ4-Cβ and Vβ8-Cβ bands are only detected in the B6 (wild-type) thymocytes but not in the Rag2−/− thymocyte sample and only at a very low level in the SCID thymocyte sample (Fig. 1). This data is consistent with the number of thymocytes that have undergone V(D)J rearrangement in these mice. Because Vβ8.2 is the gene segment most commonly used for TCRβ rearrangement, this negative result also renders unlikely any possible contamination of Rag2−/− and SCID samples with products of other rearranged Vβ genes.

FIGURE 1.

Detection of TCR Vβ germline transcription in Rag2−/− and SCID thymocytes. Total RNA samples from 1) Rag2−/−, 2) SCID, and 3) B6 thymocytes were used for RT-PCR analysis. PCR primer pairs Vβ4-Cβ and Vβ8-Cβ were used to detect transcription of rearranged TCR genes. Primer pairs Vβ5.1/8.2 and Cβ-Cβ were used to detect transcription of the TCR Vβ 8.2 and TCR Cβ regions, respectively, from both germline and rearranged genes. GAPDH was used as a control to normalize sample variations.

FIGURE 1.

Detection of TCR Vβ germline transcription in Rag2−/− and SCID thymocytes. Total RNA samples from 1) Rag2−/−, 2) SCID, and 3) B6 thymocytes were used for RT-PCR analysis. PCR primer pairs Vβ4-Cβ and Vβ8-Cβ were used to detect transcription of rearranged TCR genes. Primer pairs Vβ5.1/8.2 and Cβ-Cβ were used to detect transcription of the TCR Vβ 8.2 and TCR Cβ regions, respectively, from both germline and rearranged genes. GAPDH was used as a control to normalize sample variations.

Close modal

However, TCRVβ segments are transcribed in Rag2−/− and SCID thymocytes. Fig. 1 demonstrates this by RT-PCR using primer set L5.1 and V8.2 (Table I) to detect the major transcript of the Vβ8.2 segment (see below). As shown in Fig. 1, with these primers a Vβ8.2 band was readily detected in all three samples. Because no products of TCR rearrangement can be detected in Rag2−/− thymocytes and little if any in SCID thymocytes, all Vβ8.2 detected in those samples represents germline transcripts. Our results confirm that although there is no TCR rearrangement in Rag2−/− and SCID thymocytes, there is germline transcription of at least one segment in the Vβ region without rearrangement.

To compare the transcription of different individual Vβ segments before rearrangement, semiquantitative RT-PCR analyses were conducted using primers designed for each Vβ to span the intron between the L (leader) and V exons. Twelve Lβ-Vβ primer sets were made to cover representatives of the entire β locus, including Vβ2.1, which is the most distant from the enhancer, and Vβ14.1, which is the closest to the enhancer (see below). Serial dilutions of Rag2−/− thymocyte genomic DNA, which has the same number of copies of each TCR Vβ segment, were used as controls to normalize the PCR efficiencies of different primer sets (Fig. 2,A, lanes 1–4). Each primer set was tested to compare its ability to amplify aliquots of cDNA from Rag2−/− thymocytes (R lanes, Fig. 2) with its ability to amplify cDNA from unfractionated thymocytes of normal B6 mice (B lanes, Fig. 2), in which all Vβ genes should be rearranged and expressed. Aliquots of the same cDNA samples were used with all primer pairs, and all these primer pairs amplified fragments of the expected size from the control samples of genomic DNA (Fig. 2,A, lanes 1–4) and from the normal thymocyte cDNA, in parallel reactions. Thus both the amplification conditions and the quality of the samples were fully validated in every experiment. The results indicate that PCR products could successfully be amplified, in differing yields, from Rag2−/− immature thymocytes using 10 of the 12 primer sets tested. Similar results were seen with cDNA from SCID thymocytes (S lanes, Fig. 2, B and C). Thus the transcription of Vβ genes in the absence of rearrangement extends to multiple Vβ segments, not just Vβ8.2.

FIGURE 2.

Differential expression of TCR Vβ segments in the absence of rearrangement. A, Quantitative comparison of expression of twelve Vβ segments. Semiquantitative RT-PCR was conducted using RNA prepared from B6, MHC−/−, Rag2−/−, and SCID thymocytes. For each Vβ segment, PCR primer pairs were designed with the 5′ primer in the L region and the 3′ primer in the corresponding V region. Rag2−/− genomic DNA, which has equal numbers of copies of each TCR Vβ segment, was used as a quantitative standard to normalize the PCR efficiencies of different primer sets. This is possible because the introns spanned by the products are small, and products both with and without the intron are amplified with similar efficiency. Lanes 1–4, PCR products from 10-fold serial dilutions of Rag2−/− thymocyte genomic DNA (1 to 0.001 ng). Lanes 5, 6, and 7, PCR products from B6, MHC−/−, and Rag2−/− thymocyte cDNA samples, respectively. To keep all the samples in the linear range of PCR, B6 cDNA samples were diluted to 1/50 of the Rag2−/− cDNA sample levels and MHC−/− cDNA samples were diluted to 1/10 of the Rag2−/− cDNA sample levels before amplification. B, Comparison of expression of Vβ RNAs in Rag2−/− and SCID thymocytes with expression in normal B6 thymocytes. Analysis as in A. Lanes 1 and 2, PCR products from dilutions of B6 cDNA corresponding to 1/50 and 1/500 the amounts of cDNA used in lanes 3 and 4. Lanes 3 and 4, PCR products from Rag2−/− (lane R) and SCID (laneS) cDNA, respectively. Lane 5, negative control using the same input of Rag2−/− RNA as in lane 3, but without reverse transcription before PCR amplification. C, Expression of germline Vβ15.1 despite location in Vβ6.1-Vβ3.1 cluster. Analysis as in B, except that three serial dilutions of B6 thymocyte cDNA were used (laneB) to calibrate the yield of PCR product from primers for the L and V regions of Vβ15.1, as compared with the yield from primers for Vβ2.1 and Vβ6.1. The figure shows that Vβ15.1 is expressed at levels comparable to Vβ2.1, and substantially more abundantly than Vβ6.1.

FIGURE 2.

Differential expression of TCR Vβ segments in the absence of rearrangement. A, Quantitative comparison of expression of twelve Vβ segments. Semiquantitative RT-PCR was conducted using RNA prepared from B6, MHC−/−, Rag2−/−, and SCID thymocytes. For each Vβ segment, PCR primer pairs were designed with the 5′ primer in the L region and the 3′ primer in the corresponding V region. Rag2−/− genomic DNA, which has equal numbers of copies of each TCR Vβ segment, was used as a quantitative standard to normalize the PCR efficiencies of different primer sets. This is possible because the introns spanned by the products are small, and products both with and without the intron are amplified with similar efficiency. Lanes 1–4, PCR products from 10-fold serial dilutions of Rag2−/− thymocyte genomic DNA (1 to 0.001 ng). Lanes 5, 6, and 7, PCR products from B6, MHC−/−, and Rag2−/− thymocyte cDNA samples, respectively. To keep all the samples in the linear range of PCR, B6 cDNA samples were diluted to 1/50 of the Rag2−/− cDNA sample levels and MHC−/− cDNA samples were diluted to 1/10 of the Rag2−/− cDNA sample levels before amplification. B, Comparison of expression of Vβ RNAs in Rag2−/− and SCID thymocytes with expression in normal B6 thymocytes. Analysis as in A. Lanes 1 and 2, PCR products from dilutions of B6 cDNA corresponding to 1/50 and 1/500 the amounts of cDNA used in lanes 3 and 4. Lanes 3 and 4, PCR products from Rag2−/− (lane R) and SCID (laneS) cDNA, respectively. Lane 5, negative control using the same input of Rag2−/− RNA as in lane 3, but without reverse transcription before PCR amplification. C, Expression of germline Vβ15.1 despite location in Vβ6.1-Vβ3.1 cluster. Analysis as in B, except that three serial dilutions of B6 thymocyte cDNA were used (laneB) to calibrate the yield of PCR product from primers for the L and V regions of Vβ15.1, as compared with the yield from primers for Vβ2.1 and Vβ6.1. The figure shows that Vβ15.1 is expressed at levels comparable to Vβ2.1, and substantially more abundantly than Vβ6.1.

Close modal

As expected, the expression of Vβ segments in samples of Rag2−/− and SCID immature thymocytes was much lower than that in samples of wild-type B6 thymocytes (Fig. 2,A, compare lane 5 to lane 7; and Fig. 2,B, compare lanes 1 and 2 to lanes 3 and 4). In most cases, Vβ germline transcription in the immature cells was at least 50 times weaker than the transcription of rearranged TCR genes in the normal thymocytes, as indicated by the 50-fold dilution of the B6 samples needed to bring them into the same range as the immunodeficient samples (Fig. 2,A,lanes 5 vs 7; Fig. 2,B,lanes 1 vs 3 and 4). This reflects the dominant positive regulation of TCR-β expression by the enhancer once rearrangement has occurred (1). More interestingly, the relative expression levels of germline transcripts of different Vβ segments are also changed from the pattern seen when the genes can rearrange. In the B6 sample, most Vβs are expressed at similar levels except for Vβ8.3 and Vβ8.2. In fact, there is even more similarity in expression because the detection of Vβ8.2 is artificially depressed here: in mature T cells most Vβ8.2 transcripts use L5.1 for the promoter and leader regions, whereas the only Vβ8.2 detected here is the fraction that uses L8.2. However, in Rag2−/− and SCID samples (Fig. 2, A and B, lanes R and S) the expression levels of different Vβs were significantly different. These differences are evident both as normalized to quantitation standards in parallel control reactions (Fig. 2,A, lanes 1–4) and as compared with expression in parallel samples from normal B6 thymocytes (Fig. 2,A, lane 5, and Fig. 2 B, lanes 1 and 2,laneB).

Based on their germline expression levels, the Vβ segments can be classified into four groups. First, Vβ8.2 (Lβ8.2-Vβ8.2) and Vβ5.1 (Lβ5.1-Vβ5.1) are the most strongly expressed, at ≥1/50 the level in B6 thymocytes. The product of the Lβ5.1 promoter spliced to the body of the Vβ8.2 is expressed at even higher levels, (≥1/10 the level of B6 thymocytes; see below) indicating an even stronger intrinsic activity of the Vβ5.1 promoter. Vβ8.3, though accumulating to a lower absolute level, is also expressed at about 1/50 its level in B6 thymocytes. Second, Vβ2.1, Vβ4.1, Vβ1.1, Vβ5.2, and Vβ8.1 are expressed at lower levels, from 1/100 to 1/1000 the level seen in B6 thymocytes. Third, the expression of Vβ7.1 and Vβ14.1 is lower yet (<1/1000 the level of B6 thymocytes). Finally, in the same mutant thymocyte samples, expression of Vβ6.1 and Vβ3.1 is generally undetectable. These expression levels are summarized in Table II. While RNA stability effects could contribute to the relative levels of these transcripts, the results strongly imply that different Vβ promoters are differentially active before rearrangement.

Table II.

Correlation between Vβ germline transcription, Vβ promoter region sequences, and V(D)J rearrangement

Germline TranscriptionTCR RearrangementPromoter Region Sequences
Rag2−/−aRag2−/−/B6bNKTcd14 FTdThyePeripheTGANNTCATGANNNNTCA16-merf
2s1 >1 /500 8.60% 10.10% 15% CTGAGGAAGTCACACC 
4s1 1 /200 2.50% 2% CCATCTACACAACAGG 
1s1 1 /500 16 6% 6.50% ATAGTGACTTCACAGA 
5s2 >1 /500 4% 0.50% AACCTGACATCATAGG 
8s3 >1 /50 10.10% 10.20% ACAGTGACATCACTAA 
5s1 ++ >1 /50 5% 2% AATCTGACATCACAGG 
5s1/8s2 +++ >1 /10 72% 10 23% 21%    
8s2 ++ >1 /10     ACAGTGATGTCACTAA 
8s1 <1 /50 7.50% 10% ACAGTGACATCACTAA 
6s1 − 0.90% 4% 5% None 
15s1 <1 /500 N/A N/A ACTGATGTGTCACTAG 
20s1 <1 /500 N/A N/A GGGATGATGTCACAGG 
3s1 – 2% N/A GGTGTCACGTCAGTGT 
7s1 +/– <1 /1000 16.80% 5% 6% ACAGTGACATCATAAG 
14s1 +/– >1 /5000 5% 4% None 
Germline TranscriptionTCR RearrangementPromoter Region Sequences
Rag2−/−aRag2−/−/B6bNKTcd14 FTdThyePeripheTGANNTCATGANNNNTCA16-merf
2s1 >1 /500 8.60% 10.10% 15% CTGAGGAAGTCACACC 
4s1 1 /200 2.50% 2% CCATCTACACAACAGG 
1s1 1 /500 16 6% 6.50% ATAGTGACTTCACAGA 
5s2 >1 /500 4% 0.50% AACCTGACATCATAGG 
8s3 >1 /50 10.10% 10.20% ACAGTGACATCACTAA 
5s1 ++ >1 /50 5% 2% AATCTGACATCACAGG 
5s1/8s2 +++ >1 /10 72% 10 23% 21%    
8s2 ++ >1 /10     ACAGTGATGTCACTAA 
8s1 <1 /50 7.50% 10% ACAGTGACATCACTAA 
6s1 − 0.90% 4% 5% None 
15s1 <1 /500 N/A N/A ACTGATGTGTCACTAG 
20s1 <1 /500 N/A N/A GGGATGATGTCACAGG 
3s1 – 2% N/A GGTGTCACGTCAGTGT 
7s1 +/– <1 /1000 16.80% 5% 6% ACAGTGACATCATAAG 
14s1 +/– >1 /5000 5% 4% None 
a

Germline transcription level of each Vβ segment was based on three separate experiments. “+++” is above the level of 1 ng Rag2−/− DNA; “++” is around the level of 1 ng DNA; “+” represents an expression level greater than or equal to the level of 0.1 ng DNA; “+/–” represents an expression level less than the level of 0.1 ng DNA; and “–” is lower than the detection limit.

b

Germline expression level of each Vβ segment (Rag2−/−) as compared with the expression level of rearranged Vβs in normal thymus (B6). N, Not detected.

c

Usage of individual Vβ in NK-1+ T cells (NKT): numbers indicate percentages of cells using each segment (25 ).

d

Number of different V(D)J rearrangements of each Vβ observed in 53 rearrangements analyzed from day 14 fetal thymus (d14 FT) (24 ).

e

Relative Vβ transcription levels in thymus (Thy) and in peripheral lymphoid tissue (Periph), as measured by an RNase protection assay. The values are percentage of protected counts, which is the amount of counts for a particular Vβ gene segment divided by the sum of counts for all 15 tested Vβ gene segments (not all the 15 Vβs were shown in this table) (27 ). N/A, Not applicable.

f

Promoter sequences were taken from GenBank accession numbers AE000663, AE000664, and AE000665.

For many L-V primer pairs the sizes of the PCR products from SCID and Rag2−/− cDNA indicate unambiguously that they derived from spliced mRNA, providing confirmation that these gene segments are transcribed in the mutant thymocytes. The sizes of PCR products of other primer sets were consistently the same as those amplified from genomic DNA, e.g., Vβ5.2, Vβ8.3, Vβ8.2, and Vβ8.1 (note that in Fig. 2, Vβ8.2 transcripts were amplified using a 5′ primer from the Lβ8.2 promoter segment and not the more distant Lβ5.1). However, this does not mean that they necessarily resulted from contamination of the RNA samples with genomic DNA, because these bands were not detected in the RNA control samples (RNA without reverse transcription, Fig. 2,B, lane N; compare signals in corresponding sample with reverse transcription, lane R). Therefore, the sizes of these bands may be due to incomplete splicing of the germline transcripts. Germline-sized amplification products were also generated occasionally from SCID and Rag2−/− cDNA with the Vβ14.1 and Vβ7.1 primers, but the yields were irreproducible (e.g., Fig. 2,A,lane 7 vs Fig. 2 B,lanes 3 and 4). Because these amplifications give little product overall, the germline-sized bands in these cases could result from small amounts of contaminating DNA in some preparations (data not shown), and the true levels of Vβ14.1 and Vβ7.1 transcription before rearrangement could be even lower.

Several distinct mechanisms could contribute to the difference between the relative efficiencies of germline transcription in these immature, developmentally arrested cells and transcription of rearranged genes in the wild-type thymocyte population: V(D)J rearrangement, β selection, and positive selection. To test whether this difference is due to positive selection, we examined the Vβ expression in MHC−/− thymocytes. Because MHC class I and II products are required for positive selection, most thymocytes from these doubly deficient animals undergo V(D)J rearrangement and β selection normally but arrest in the CD4+CD8+ stage because they cannot undergo positive selection. As shown in Fig. 2 A,lane 6 (lane M), virtually indistinguishable patterns of Vβ expression were observed with RNA from MHC−/− and B6 thymocytes, suggesting that the difference between the Vβ transcription profiles in Rag2−/− and B6 thymocytes is not due to positive selection.

The distinctive levels of expression of different Vβ in the absence of rearrangement do not correlate with the proximity of the Vβ segment to the enhancer (Fig. 3,A). Instead, not only is the strongest expression seen for segments in the middle of the complex (Vβ5.1 and Vβ8.2), but also transcription of the most distal Vβ segments (Vβ2.1, Vβ1.1, and Vβ 4.1) is stronger than that of the most proximal Vβs analyzed in Fig. 2 (Vβ14, Vβ7.1, Vβ3.1, and Vβ6.1). Most dramatic is the poor expression of Vβ14.1, despite its location within 5 kb of the enhancer. This implies that proximity to the Cβ enhancer is unlikely to be the major element controlling germline transcription of Vβ segments at this stage. In agreement with this interpretation, a recent report (13) shows that germline Vβ transcription is not inhibited by deletion of the TCRβ enhancer. One possibility is that the individual Vβ promoters may be the dominant regulators for Vβ germline transcription. Alternatively, if the clustering of Vβs with similar expression profiles is significant, a few local positive or negative regulatory elements scattered throughout the 700-kb complex might affect the activity of Vβ promoters over neighborhoods of ∼50–100 kb.

FIGURE 3.

Structures of the Mus musculus TCR β locus and individual Vβ promoters. A, Genomic organization of the TCRβ locus. Sequence taken from GenBank accession numbers AE000663, AE000664, and AE000665 is diagrammed to show the map positions of the Vβ segments relative to other elements of the murine TCR-β complex. C1 and C2 are constant regions. D and J segments are not shown, but map in the immediate vicinity of the constant regions. E is the Cβ enhancer. Variable regions are labeled by numbers: all except Vβ14 are transcribed in a left to right direction on the map. The scale bar shows a distance of 100 kb. B, Potential transcription factor binding sites in promoter regions of Vβ segments. Promoter regions of the Vβ segments studied here were analyzed to locate potential transcription factor binding sites as predicted by two search programs: MOTIFSearch (DNASIS, Hitachi Software, San Francisco, CA); and TFSEARCH (ver 1.3; Yutaka Akiyama: “TFSEARCH: Searching Transcription Factor Binding Sites”, http://www.rwcp.or.jp/papia/), using the TRANSFAC Matrix Table, Rel. 3.3 (26 ). Sequences were scanned from −600 to +100 with respect to the initiation codons of the Vβ segments. The figure shows the locations of predicted sites for selected factor families of known hemopoietic relevance. In the case of TFSEARCH predictions, a score of 90.0 was used as the cutoff. In each case, the analysis is shown with the protein coding sequence beginning at position zero on the figure. The Vβ14 analysis is presented in the opposite transcriptional orientation from the others, but in the same sense with respect to its orientation in the genome. Ets includes predicted Ets-1 and PEA-3-type sites. In the case of Vβ2.1, a double Ets site is shown 3′ of the AP-1 site based on functional studies (21 ). E (bHLH) includes “bHLH” and “E2A” predicted sites. GATA includes predicted sites for GATA-1, -2, or -3. GATA (ns) sites deviate from the standard WGATAR motif but meet other criteria for binding. Rel includes predicted NF-κB sites. Placement of symbols above or below the lines on the figure is only to aid visualization of closely linked or overlapping motifs.

FIGURE 3.

Structures of the Mus musculus TCR β locus and individual Vβ promoters. A, Genomic organization of the TCRβ locus. Sequence taken from GenBank accession numbers AE000663, AE000664, and AE000665 is diagrammed to show the map positions of the Vβ segments relative to other elements of the murine TCR-β complex. C1 and C2 are constant regions. D and J segments are not shown, but map in the immediate vicinity of the constant regions. E is the Cβ enhancer. Variable regions are labeled by numbers: all except Vβ14 are transcribed in a left to right direction on the map. The scale bar shows a distance of 100 kb. B, Potential transcription factor binding sites in promoter regions of Vβ segments. Promoter regions of the Vβ segments studied here were analyzed to locate potential transcription factor binding sites as predicted by two search programs: MOTIFSearch (DNASIS, Hitachi Software, San Francisco, CA); and TFSEARCH (ver 1.3; Yutaka Akiyama: “TFSEARCH: Searching Transcription Factor Binding Sites”, http://www.rwcp.or.jp/papia/), using the TRANSFAC Matrix Table, Rel. 3.3 (26 ). Sequences were scanned from −600 to +100 with respect to the initiation codons of the Vβ segments. The figure shows the locations of predicted sites for selected factor families of known hemopoietic relevance. In the case of TFSEARCH predictions, a score of 90.0 was used as the cutoff. In each case, the analysis is shown with the protein coding sequence beginning at position zero on the figure. The Vβ14 analysis is presented in the opposite transcriptional orientation from the others, but in the same sense with respect to its orientation in the genome. Ets includes predicted Ets-1 and PEA-3-type sites. In the case of Vβ2.1, a double Ets site is shown 3′ of the AP-1 site based on functional studies (21 ). E (bHLH) includes “bHLH” and “E2A” predicted sites. GATA includes predicted sites for GATA-1, -2, or -3. GATA (ns) sites deviate from the standard WGATAR motif but meet other criteria for binding. Rel includes predicted NF-κB sites. Placement of symbols above or below the lines on the figure is only to aid visualization of closely linked or overlapping motifs.

Close modal

The least-expressed Vβs are all relatively closely linked in the 3′ portions of the Vβ complex, raising the possibility that there could be a local inhibitor of gene expression in this region. Alternatively, it is possible that these Vβs share some common feature of their promoters. The promoters of different Vβ segments are quite divergent, with few if any blocks of conserved sequence. Computer searches for various transcription factor consensus sites (TFSearch and MOTIFSearch) reveal potential binding sites for T cell factor-1 (TCF-1), E26 transformation-specific gene (Ets), GATA, AP-1, basic helix-loop-helix (bHLH; “E box”), myeloblastosis transforming gene, and acute myelogenous leukemia gene-1 (AML1; Runx1) family transcription factors in highly dissimilar clusterings in the regions from −600 to +100 around the different Vβ promoters, as diagrammed in Fig. 3,B. It has been reported that transcription from TCR Vβ promoters depends on a conserved decamer motif similar to the cAMP response element (CRE), often embedded in a conserved 14-mer (CAGTGAYRTCACTG) (2, 14) or “16-mer” sequence (GenBank accession numbers AE000663, AE000664, and AE000665; see annotation). However, the promoters of Vβ6.1 and Vβ3.1 lack the “16-mer” and even the core TGANNTCA motif (see Table II). To test whether the weak expression of these Vβs is due to the presence of a general silencing element in the 3′ portion of the TCRβ complex, we checked the expression of another Vβ segment in the same cluster, Vβ15.1, located between Vβ6.1 and Vβ3.1, which does have a divergent but identifiable 16-mer motif in its promoter region. As shown in Fig. 2 C, Vβ15.1 is clearly expressed before rearrangement. Thus, there is not a general silencing of genes clustered with Vβ6.1 and Vβ3.1. This implies that the differential expression is likely to be controlled, at least partially, by individual Vβ segment promoters.

Structural analyses of the Vβ transcripts were conducted to determine the promoters used and to verify whether they were indeed derived from genes in germline configuration. To obtain representative Vβ cDNA clones, a high-density arrayed library of SCID thymocyte cDNA was screened using [32P]dATP labeled probes for particular Vβ segments. In agreement with the preferential expression of Vβ8.2 (with an Lβ5.1 leader) and Vβ5.1 segments, over 50 Vβ8.2 clones and one Vβ5.1 clone were detected in a screening of 70,000 clones, as illustrated for a part of the library in Fig. 4,A. No other Vβ-positive clone could be detected in this set using either Vβ4.1 or Vβ14.1 probes (data not shown), consistent with the lower abundance for other transcripts indicated by RT-PCR. However, sequence analysis of the clones obtained showed clearly the retention of intact heptamer and nonamer recombination signal sequences at the 3′ ends of the Vβ segments proper (Fig. 4 B, and data not shown), confirming that they represent transcripts of unrearranged genes.

FIGURE 4.

TCR Vβ5.1 and TCR Vβ8.2 clones from a SCID thymus cDNA library. A, Detection of germline Vβ cDNAs in arrayed library. 35P-labeled Vβ5.1 and Vβ8.2 probes were used to screen a SCID thymus cDNA arrayed library. Fifty Vβ8.2-positive clones were detected. The right panel shows part of the library. One Vβ5.1-positive clone was detected, which is indicated by an arrow in the left panel. B, Sequence elements of Vβ8.2 clones 15M16 and 14L12. The Vβ8.2-positive clones from SCID thymus cDNA were sequenced using ABI dye terminator cycle sequencing ready reaction kit. Their positions in TCR β locus are indicated by number. The position of the “16 mer,” leader region 5.1 (L5.1), Vβ8.2 (V8.2), recombination signal sequence (RSS), and 3′ gene-specific primer of Vβ8.2 are also indicated. The sequences at the 3′ end possess the recombination signal sequence demonstrating that the gene transcribed is in germline configuration.

FIGURE 4.

TCR Vβ5.1 and TCR Vβ8.2 clones from a SCID thymus cDNA library. A, Detection of germline Vβ cDNAs in arrayed library. 35P-labeled Vβ5.1 and Vβ8.2 probes were used to screen a SCID thymus cDNA arrayed library. Fifty Vβ8.2-positive clones were detected. The right panel shows part of the library. One Vβ5.1-positive clone was detected, which is indicated by an arrow in the left panel. B, Sequence elements of Vβ8.2 clones 15M16 and 14L12. The Vβ8.2-positive clones from SCID thymus cDNA were sequenced using ABI dye terminator cycle sequencing ready reaction kit. Their positions in TCR β locus are indicated by number. The position of the “16 mer,” leader region 5.1 (L5.1), Vβ8.2 (V8.2), recombination signal sequence (RSS), and 3′ gene-specific primer of Vβ8.2 are also indicated. The sequences at the 3′ end possess the recombination signal sequence demonstrating that the gene transcribed is in germline configuration.

Close modal

Because the relative efficiencies of transcription of different Vβ segments in the immature cells are different from the profiles in thymocytes with rearranged genes, they might use different promoters from those identified in mature T cells. It has been reported that germline transcription of the TCR Dβ region depends on a developmental stage-specific promoter in precursor T cells (15). Distal Vβ promoters have also been found in the Vβ5.1 and Vβ8.1 regions (16). To test whether any unconventional promoters are used for Vβ germline transcription, 5′-RACE PCR analysis was performed comparing poly(A)+ RNA prepared from SCID and normal B6 thymocytes. Three Vβ segments, Vβ4s1 (Vβ4.1), Vβ8s2 (Vβ8.2), and Vβ14s1 (Vβ14.1), were chosen as representatives of the three expression classes, based on their relative locations in the complex and their relative germline transcription levels.

As shown in Fig. 5 B, for all three Vβ segments, the major 5′-RACE products in both B6 (lane B) and SCID (lane S) samples were the same sizes. We first tested whether any distal Vβ promoters of Vβ5.1 may be preferentially used for germline transcription of Vβ8.2. A major band at 400 bp was detected in both B6 and SCID samples, corresponding to a transcription start site of ∼−40 (relative to the initiation ATG) for the proximal promoter. Hybridization with Lβ5.1- and Lβ8.2-specific probes showed that rearranged (B6) and unrearranged (SCID) gene transcripts both preferentially used the Lβ5.1 promoter, and that even the distributions of minor 5′-RACE products, representing upstream start sites, were similar in both (data not shown). In the case of Vβ4.1, the major band was 380 bp in both B6 and SCID, corresponding to a start site at around −143, and for Vβ14.1, the only product seen was 350 bp, indicating a start site at −55. While some differences were seen in the profiles of minor start sites for the Vβ8.2 and Vβ4.1 segments (species from 500 to 1000 bp), these were not reproducible. Thus, the majority of germline Vβ transcripts share the same transcriptional starting site as the transcripts of rearranged genes.

FIGURE 5.

5′ starting sites of Vβ germline transcripts: 5′-RACE of Vβ8.2, Vβ4.1, and Vβ14.1 transcripts from normal and recombination-deficient thymocytes. 5′-RACE analysis were conducted as described in Materials and Methods using poly(A)+ mRNA prepared from B6 (lane B) and SCID (lane S) thymocytes. 3′ gene-specific primers for Vβ8.2, Vβ4.1, and Vβ14.1 were used. The most abundant bands are indicated by an arrow, and the sizes of the bands are shown on the left. Due to the low level of germline transcription of Vβ4.1, and Vβ14.1, 5′-RACE products were transferred to Hybond-N+ membrane after electrophoresis, and 32P-labeled gene-specific probes (O4.1 and O14.1) were used to increase the detection sensitivity. Note that the exposure time for SCID samples were 2 days for Vβ4.1 and 3 days for Vβ14.1, while the exposure time for B6 samples were 3 h for Vβ4.1 and <20 min for Vβ14.1.

FIGURE 5.

5′ starting sites of Vβ germline transcripts: 5′-RACE of Vβ8.2, Vβ4.1, and Vβ14.1 transcripts from normal and recombination-deficient thymocytes. 5′-RACE analysis were conducted as described in Materials and Methods using poly(A)+ mRNA prepared from B6 (lane B) and SCID (lane S) thymocytes. 3′ gene-specific primers for Vβ8.2, Vβ4.1, and Vβ14.1 were used. The most abundant bands are indicated by an arrow, and the sizes of the bands are shown on the left. Due to the low level of germline transcription of Vβ4.1, and Vβ14.1, 5′-RACE products were transferred to Hybond-N+ membrane after electrophoresis, and 32P-labeled gene-specific probes (O4.1 and O14.1) were used to increase the detection sensitivity. Note that the exposure time for SCID samples were 2 days for Vβ4.1 and 3 days for Vβ14.1, while the exposure time for B6 samples were 3 h for Vβ4.1 and <20 min for Vβ14.1.

Close modal

The 5′-sequences of the Vβ8.2-positive clones from the SCID thymus library (Fig. 5,A) provided an independent assay to check the transcription start sites. All of these cDNAs were of the form Lβ5.1-Vβ8.2. In agreement with the 5′-RACE results, four of the five clones sequenced start at ∼−15 (e.g., Fig. 4,B, clone 14L12); only one clone has a longer 5′ end with a start at ∼−380 (Fig. 4 B, clone 15 M16). These results show that the distinctive transcriptional activities of the germline Vβ segments are likely to be controlled by sequence elements associated with the known promoters as defined in mature T cells, and not with novel promoters mapping at unknown sites far upstream.

Because individual Vβs are expressed at different levels, they could also be regulated differentially during the earliest stages of T cell development. To test this hypothesis, we examined expression of a representative range of Vβ segments in thymocytes at different developmental stages. Early T cell precursor subsets were separated from Rag2−/− or TCR-β−/−δ−/− thymocytes by FACS sorting, using criteria established previously to isolate cells in developmentally distinct states (10, 11, 12). Four populations were collected: 1) Sca-1+HSA cells, which are thought to be pluripotent precursor-like cells; 2) Sca-1 HSA cells, which are NK-like cells; 3) HSA+CD44+ cells, which are rapidly proliferating pro-T cells; and 4) HSA+CD44 cells, which are mostly resting, committed T cell progenitors that in normal mice would be undergoing V-DJβ rearrangement (reviewed in Ref. 5). For this analysis, only Vβ transcripts that are predominantly spliced in total Rag2−/− cells were monitored, to avoid any question about low-level DNA contamination in the sorted populations. Data representative of at least three independent experiments are shown in Fig. 6,A, with a tabulation of results of all trials in Fig. 6 B.

FIGURE 6.

TCR Vβ germline expression during T cell development. A, Vβ expression in Rag2−/− and TCRβ−/−δ−/− thymocyte subsets. Four subpopulations were collected by cell sorting based on the expression of Sca-1, HSA, and CD44. Lanes 1 and 5, Sca-1+HSA; lanes 2 and 6, Sca-1HSA; lanes 3 and 7, HSA+CD44+; and lanes 4 and 8, HSA+CD44. Vβ expression level in each sample were determined by RT-PCR. GAPDH level was used as a control to normalize sample variations. B, Tabulation of the results of analyses of Vβ expression in Rag2−/− or TCRβ−/−δ−/− thymocytes as shown in A. Results were compiled from three to five experiments for each sample. In cases where the results differed in separate experiments, the symbol represents the more reproducible results for that Vβ segment. Note that Vβ14.1 expression was barely detectable in Rag2−/− samples, where the Tcrb locus itself is wild type, as well as in TCRβ−/−δ−/− samples, where the Jβ-Cβ deletion could encroach on the regulatory environment of this segment. Transcripts of another poorly expressed Vβ segment, Vβ7.1, could not be detected in any of the mutant thymocyte subsets by this approach (data not shown). C, Relative Vβ expression levels in fetal liver day 15.5 (FL), fetal thymus day14.5 (FT), and two RAG2−/− bone marrow subsets: B220+CD43+CD19+ (ProB) and B220CD43+ (Myl). Vβ expression levels in each populations were determined by RT-PCR with [32P]dCTP in the reaction mix. Samples were equalized on the basis of GAPDH expression before analysis. Radioactive signals were quantified by PhosphorImager after electrophoresis and transfer to Hybond-N+ membrane. The expression level of each Vβ in fetal thymus was used as standard (=1.0).

FIGURE 6.

TCR Vβ germline expression during T cell development. A, Vβ expression in Rag2−/− and TCRβ−/−δ−/− thymocyte subsets. Four subpopulations were collected by cell sorting based on the expression of Sca-1, HSA, and CD44. Lanes 1 and 5, Sca-1+HSA; lanes 2 and 6, Sca-1HSA; lanes 3 and 7, HSA+CD44+; and lanes 4 and 8, HSA+CD44. Vβ expression level in each sample were determined by RT-PCR. GAPDH level was used as a control to normalize sample variations. B, Tabulation of the results of analyses of Vβ expression in Rag2−/− or TCRβ−/−δ−/− thymocytes as shown in A. Results were compiled from three to five experiments for each sample. In cases where the results differed in separate experiments, the symbol represents the more reproducible results for that Vβ segment. Note that Vβ14.1 expression was barely detectable in Rag2−/− samples, where the Tcrb locus itself is wild type, as well as in TCRβ−/−δ−/− samples, where the Jβ-Cβ deletion could encroach on the regulatory environment of this segment. Transcripts of another poorly expressed Vβ segment, Vβ7.1, could not be detected in any of the mutant thymocyte subsets by this approach (data not shown). C, Relative Vβ expression levels in fetal liver day 15.5 (FL), fetal thymus day14.5 (FT), and two RAG2−/− bone marrow subsets: B220+CD43+CD19+ (ProB) and B220CD43+ (Myl). Vβ expression levels in each populations were determined by RT-PCR with [32P]dCTP in the reaction mix. Samples were equalized on the basis of GAPDH expression before analysis. Radioactive signals were quantified by PhosphorImager after electrophoresis and transfer to Hybond-N+ membrane. The expression level of each Vβ in fetal thymus was used as standard (=1.0).

Close modal

As shown in Fig. 6, individual Vβ not only were expressed to different levels, but also showed evidence of differential developmental regulation. At every stage, the “Vβ 5.1” (Lβ5.1) promoter was most active, yielding high levels of Lβ5.1-Vβ5.1 spliced transcripts and even higher levels of Lβ5.1-Vβ8.2 transcripts. Activity of this promoter appeared to increase at the pro-T stage (Fig. 6,A, compare lanes 1 and 2 vs lanes 3 and 4 and lanes 5 and 6 vs lanes 7 and 8) but was readily detectable earlier. The expression of Vβ1.1 was much lower in the precursor-like and NK-like cells, but appeared to be strongly up-regulated at the pro-T stage and maintained at this level thereafter. Activity of the Vβ4.1 promoter appeared to undergo up-regulation at about the same stage as the Vβ1.1 promoter (Fig. 6,A, lanes 7 and 8 are most representative). Vβ14.1 showed a different pattern: while barely detectable, in two experiments the only expression seen was in the precursor-like cells (not shown). Most noteworthy is the case of Vβ2.1. Vβ2.1 RNA was detectable at a higher level in the precursor-like fraction (Fig. 6,A, lanes 1 and 5) than in the definitive T lineage populations (Fig. 6,A, lanes 3, 4, 7, and 8), and at the highest level of all in the NK-like fraction (Fig. 6 A, lanes 2 and 6). Expression of Vβ2.1 continued to be readily detectable in both fresh and IL-2-stimulated peripheral NK cells, along with expression of Vβ5.1–8.2 (data not shown). The pattern of expression of Vβ2.1 is strikingly different from all of the others.

The individual expression patterns of these different Vβ segments in thymocyte subsets led to the question whether some of these promoters might be active prethymically or in any non-T cells. Therefore, we monitored Vβ expression in bone marrow, fetal liver, and fetal thymus. Two subpopulations were separated from Rag2−/− bone marrow cells: pro-B cells, which are B220+CD43+CD19+; and a population enriched for myeloid lineage cells, which are B220 CD43+. To detect what was anticipated to be very low-level expression, RT-PCR assays were conducted in the presence of radioactive nucleotides, and the yield of products of the right size was determined by phosphorimager analysis, using the expression of each Vβ in fetal thymus as a quantitation standard. As shown in Fig. 6 C, the expression of Vβ1.1, Vβ5.1 (Lβ5.1-Vβ5.1), and Vβ8.2 (Lβ5.1-Vβ8.2) were much lower in pro-B cells, fetal liver, and the myeloid-enriched population than in fetal thymus. However, the expression of Vβ2.1 was much stronger in the B220CD43+ myeloid-enriched population. The expression pattern of Vβ4.1 showed highest levels in the fetal thymus but distinctly elevated expression in the B220CD43+ population as well. The identity of the Vβ-expressing cells in the B220CD43+ bone marrow population remains to be determined. There are no mature T or B cells in Rag2−/− bone marrow, but it is possible that there are some T cell or NK cell precursors in this extrathymic population.

These results confirm that the different Vβ promoters respond differentially to developmental transcriptional regulators in early lymphocyte development. They are all expressed at the highest levels during and after T lineage specification (Fig. 6 A, lanes 3 and 4), but specific Vβ promoters such as Vβ5.1, Vβ2.1, and Vβ4.1 also have distinctive patterns of expression in more immature cells and in prethymic or extrathymic cells that may or may not ever enter the T cell lineage.

The results presented here provide the first evidence that the regulatory sequences of individual TCR Vβ segments differ functionally so that they are transcribed at widely differing levels before rearrangement. We have also shown that Vβ expression levels are actively regulated, independently of rearrangement, during T cell development. Most interestingly, different Vβ segments display distinctive, individual patterns of expression during these early stages. This provides strong support for the view that the promoter regions of different Vβ segments do not act as simple basal promoters, but may have a complex structure that mediates tissue-specific and stage-specific regulation of TCR expression (17).

The data presented here provide an in vivo context in which specific motifs that are differentially distributed among Vβ promoters may be tested for functional significance. Until now, lacking evidence for differential function of various TCR Vβ promoters, the emphasis has been on shared elements, and only a few have been identified. As shown in Table II, most of the Vβ promoters that are expressed well, before rearrangement, share the presence of a CRE-like TGANNTCA motif in their 5′-flanking regions. The Vβ segments that are expressed poorly before rearrangement, even though they contribute strongly to the mature T cell pool, generally lack this motif: i.e., Vβ6.1, Vβ3.1, and Vβ14.1. This motif forms the core of the 14-mer, CAGTGAYRTCACNG (2), that has also been suggested as a conserved regulatory element. However, the TGANNTCA motif cannot be essential for germline expression, for it is not found in a consensus form in the promoters of Vβ2.1 or Vβ4.1 (scanning from −600 to +100 with respect to the translational initiation site), which our results show are certainly expressed before rearrangement, whereas it is clearly present in the promoter of Vβ7.1, which is not expressed nearly as well. So this motif is neither strictly necessary nor rate-limiting for expression before rearrangement. The additional cis-regulatory sites that are particularly important and the factors that bind to them have yet to be determined.

The highly divergent Vβ promoters afford more than enough potential regulatory diversity to account for the noncoordinate expression of these gene segments (see Fig. 3 B). However, other kinds of sequence elements may also contribute. Another feature that may affect the levels of expression reported here could be the genomic context itself, particularly the location of repeat sequences with silencing activity. The proximity and type of repeat sequences abutting the Vβ promoters vary greatly, from examples with repeats restricted to over 1 kb away (Vβ4.1, the genes in the Vβ5–8 cluster, and Vβ14) to examples with repeats encroaching within 350 bp of the presumed transcriptional start site (Vβ2.1, Vβ20.1, and Vβ3.1) (see GenBank accession numbers AE000663, AE000664, and AE000665). Finally, the sensitivity with which particular Vβ transcripts can be detected could be greatly affected by cis-acting sequences that affect RNA stability. To resolve the roles of these different kinds of regulatory elements, transfection and transgenesis experiments will have to be done. The distinctive patterns of expression of particular Vβ segments before rearrangement can now provide the assay system to score results of such experiments.

Our data establish not only that the various Vβ promoters are regulated independently in vivo; they also establish that certain of these promoters may be among the first T lineage genes activated in lymphoid precursors. This confirms and extends earlier work that showed some prethymic expression of the Vβ5.1–8.2 mRNA (7). Our studies of this fine-scale regulation are still in an early phase, but two features deserve comment. The first is the coordinate up-regulation of several TCR Vβ gene segments at the transition when precursor cells undergo specification to the T cell lineage. The second is the observation that certain Vβ segments are also expressed well in cell populations that are unlikely to be committed to the T cell lineage.

The Vβ1.1, Vβ4.1, Vβ5.1, and Vβ5.1–8.2 transcripts undergo up-regulation during T lineage specification and commitment (Fig. 6,A, populations 3 and 4). This is probably caused by the global changes in the expression of transcriptional regulators that take place at this stage (12, 18) and may reflect a response to T lineage restricted factors. One candidate for such a factor acting on Vβ promoters could be TCF-1 (19), which has clusters of five predicted binding sites within 200–350 bp of the mRNA cap sites in each of these promoters (see Fig. 2 B). These Vβ segments join a group of genes that are similarly up-regulated or newly induced at this stage, such as preTα, Rag1, and TdT (10, 11). The Vβ1.1 and Vβ4.1 transcripts may be particularly valuable as new lineage-specific indicators of this critical developmental transition, because the Rag1 and TdT genes are expressed in B lineage as well as T lineage precursors.

At least two Vβ segments studied here show other striking regulatory features. The Vβ5.1–8.2 and Vβ2.1 transcripts are distinguished by their expression in the most immature subsets of thymocytes, which are still thought to be pluripotent (3, 5). The expression of Vβ5.1–8.2 is detectable even in a further purified subset of extremely immature thymocytes with a multilineage gene expression pattern (H. Wang and F. Chen, unpublished data). Vβ2.1, in contrast, provides an example of regulation that is qualitatively as well as quantitatively unique. Unlike all other Vβ segments tested, Vβ2.1 is expressed more strongly in an NK-like subset in the thymus than in pro-T cell subsets, and expressed in pro-B cells in bone marrow. Its transcripts continue to be readily detectable in peripheral NK cells. It is also expressed in an enigmatic CD43+CD45R population of cells, possibly myeloid, in Rag2−/− bone marrow, at even higher levels than in fetal thymus. Vβ5.1–8.2 and Vβ4.1 are expressed in this bone marrow population too, but at lower levels. It may be significant that the promoter region of Vβ2.1 (between about −100 and −200 in Fig. 3 B (20)) is relatively deficient in predicted TCF-1 sites, but instead harbors a unique cluster of Ets sites complexed with a functional AP-1 site (21), unlike any other Vβ analyzed here. The Vβ5.1–8.2 and Vβ2.1 segments thus join IL-2 and perforin in a group of T cell associated genes that are expressed in immature cells precociously or outside the T lineage (11).

The highly divergent regulation of Vβ2.1 may be a direct or indirect result of its isolation from the rest of the cluster (see Fig. 3 A). It is not only far upstream of other Vβ segments, but even upstream of the trypsinogen gene cluster at the 5′ end of the whole complex. There is no TCR Vβ segment or pseudogene in a similar position in the human TCRβ complex (2) and no human ortholog of TCR Vβ2 identifiable on the basis of sequence (22), suggesting that Vβ2.1 may have been deleted altogether in the lineage leading to humans at some point since the last common ancestor of humans and mice. It is interesting to speculate that an early translocation may have moved the Vβ2 ancestral gene to its unusual site, where it might then have been particularly vulnerable to deletion without disruption of the locus as a whole. Such a translocation could juxtapose the Vβ2.1 coding sequences with novel regulatory elements directly. Because the amino acid sequence of murine Vβ2.1 is highly divergent from other murine and human Vβ segments, though, it is also possible that its unusual regulatory pattern may have been established through genetic drift. These results raise the possibility that Vβ2.1 in mice may have acquired new biological roles, in addition to its use in the TCR of conventional αβ T cells.

Even once activated in pro-T cells, expression of these Vβ germline transcripts remains very low. Precise figures are not available, but a rough estimate can be made on the basis of the relative RT-PCR product yields for different Vβ segments and the frequency of Vβ cDNA clones in our SCID thymocyte cDNA library. Cells that have undergone T lineage specification (HSA+) constitute about 90% of the population (23) and presumably dominate the library. A typical thymic lymphocyte with 1 pg of total RNA (est. 20 fg of poly(A)+ RNA) is likely to contain about 20,000 molecules of mRNA per cell (taking an average mRNA to contain 1500 nt at 350 g/mol nucleotide, to give about 9 × 10−19 g per mRNA). Thus the 70,000 clone library that was screened represents approximately three cell equivalents. The Vβ5.1–8.2 sequence was recovered in about 50 separate clones in the library, but it is expressed more strongly than most other Vβ segments by several orders of magnitude. This implies that most of the Vβ germline transcripts are present at much less than one copy per average cell even in the HSA+CD25+ population. This raises the interesting possibility that most germline Vβ segments are transcribed only in a minority of pro-T cells. Preliminary studies indicate that these transcripts need not be allelically excluded. In thymocytes expressing a TCR transgene, although rearrangement of the Vβ8.2 segment is inhibited, expression of the Vβ5.1–8.2 germline transcript continues (data not shown). However, further work will be needed to determine whether in normal mice the same rare precursor cells express more than one Vβ segment at a time.

Could such germline transcription of Vβ segments foreshadow or even direct segment choice for V(D)J rearrangement? This question has been extensively debated, but it is posed in an immediate form by the highly biased transcription we see of some segments at the expense of others. Comparison of the Vβ germline expression levels detected in this study with the first wave of TCR β rearrangement frequencies reported in day 14 wild-type fetal thymus (24) shows some parallels (Table II), but our transcription data and the reported rearrangement frequencies are not well correlated overall. One interpretation may be that germline transcription of certain TCR variable regions could be a consequence of chromatin-level events in a larger subregion of the TCRβ locus, rather than a mechanism for targeting these events. Significant differences among Vβ segments in recombination signal sequences could independently skew rearrangement frequencies, perhaps compensating, to some extent, for quantitative biases introduced by differential germline transcription. However, our data leave open the possibility that differential germline transcription plays a more important role in targeting rearrangements in certain compartments of the TCR repertoire. It is striking that the NK1.1+ subset in the Rag2−/− thymus may express only two Vβ segments, Vβ2 and Vβ5.1(L)-8.2(V). Provocatively, these are also two of the most prevalent Vβ segments used to form the TCR of NK1.1+ T cells in the normal T cell population (25).

In summary, the regulatory elements that control expression of individual TCR Vβ segments have patterns of activity in the absence of rearrangement that differ quantitatively, temporally, and in terms of developmental lineage specificity. The unexpected magnitude of expression differences between the individual germline Vβ segments raise a range of novel questions about their regulation and biological functions, but also provide an opportunity to define the mechanisms responsible.

We thank Michele Anderson for the arrayed cDNA library from SCID thymocytes; Hua Wang for samples of sorted mutant thymocyte populations; Rochelle Diamond for expert assistance with the cell separations; Janice Telfer and Mary Yui for samples of fresh and IL-2 stimulated NK cells; Gabriela Hernandez-Hoyos for transgenic T cells expressing the AND-TCR; and members of the Rothenberg laboratory for valuable advice and suggestions. We also thank Ellen Robey (University of California, Berkeley, CA) for the gift of the MHC-deficient mouse breeding stock. This work was also made possible by the Caltech Biopolymer Synthesis Facility; by the Caltech Flow Cytometry and Cell Sorting Facility; and by Dana Miller and, later, employees of the Office of Laboratory Animal Research at Caltech who provided excellent care of the immunodeficient mice.

1

This work was supported by the DNA Sequencer Royalty Fund at Caltech.

3

Abbreviations used in this paper: Rag2, recombinase-activating gene-2; AML1, acute myelogenous leukemia gene-1 (Runt family transcription factor also known as Runx1, core-binding factor α2, etc.); bHLH, basic helix-loop helix (transcription factor class); CRE, cAMP response element; E14.5, embryonic day of gestation 14.5; Ets, E26 transformation-specific gene (a family of winged-helix transcription factors); RACE, rapid amplification of cDNA ends; TCF-1, T cell factor-1 (high mobility group box transcription factor); TE, Tris-EDTA buffer; B6, C57BL/6; HSA, heat stable Ag.

1
Leiden, J. M..
1993
. Transcriptional regulation of T cell receptor genes.
Annu. Rev. Immunol.
11
:
539
2
Rowen, L., B. F. Koop, L. Hood.
1996
. The complete 685-kilobase DNA sequence of the human β T cell receptor locus.
Science
272
:
1755
3
Shortman, K., L. Wu.
1996
. Early T lymphocyte progenitors.
Annu. Rev. Immunol.
14
:
29
4
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
5
Rothenberg, E. V., J. C. Telfer, M. K. Anderson.
1999
. Transcriptional regulation of lymphocyte lineage commitment.
BioEssays
21
:
726
6
Wilson, A., M. Capone, H. R. MacDonald.
1999
. Unexpectedly late expression of intracellular CD3ε and TCR γδ proteins during adult thymus development.
Int. Immunol.
11
:
1641
7
Candéias, S., R. R. Hardy, Y. S. Li, U. D. Staerz.
1994
. T cell receptor Vβ8.2 gene germ-line transcription: an early event of lymphocyte differentiation.
Eur. J. Immunol.
24
:
3073
8
Bouvier, G., F. Watrin, M. Naspetti, C. Verthuy, P. Naquet, P. Ferrier.
1996
. Deletion of the mouse T-cell receptor β gene enhancer blocks αβ T-cell development.
Proc. Natl. Acad. Sci. USA
93
:
7877
9
Diamond, R. A., S. B. Ward, K. Owada-Makabe, H. Wang, E. V. Rothenberg.
1997
. Different developmental arrest points in RAG2−/− and scid thymocytes on two genetic backgrounds: developmental choices and cell death mechanisms before TCR gene rearrangement.
J. Immunol.
158
:
4052
10
Wang, H., R. A. Diamond, E. V. Rothenberg.
1998
. Cross-lineage expression of Ig-β (B29) in thymocytes: positive and negative gene regulation to establish T-cell identity.
Proc. Natl. Acad. Sci. USA
95
:
6831
11
Wang, H., R. A. Diamond, J. A. Yang-Snyder, E. V. Rothenberg.
1998
. Precocious expression of T-cell functional response genes in vivo in primitive thymocytes before T-lineage commitment.
Int. Immunol.
10
:
1623
12
Anderson, M. K., G. Hernandez-Hoyos, R. A. Diamond, E. V. Rothenberg.
1999
. Precise developmental regulation of Ets family transcription factors during specification and commitment to the T cell lineage.
Development
126
:
3131
13
Mathieu, N., W. M. Hempel, S. Spicuglia, C. Verthuy, P. Ferrier.
2000
. Chromatin remodeling by the T cell receptor (TCR)-β gene enhancer during early T cell development. Implications for the control of TCR-β locus recombination.
J. Exp. Med.
192
:
625
14
Anderson, S. J., S. Miyake, D. Y. Loh.
1989
. Transcription from a murine T-cell receptor Vβ promoter depends on a conserved decamer motif similar to the cyclic AMP response element.
Mol. Cell. Biol.
9
:
4835
15
Sikes, M. L., R. J. Gomez, J. Song, E. Oltz.
1998
. A developmental stage-specific promoter directs germline transcription of DβJβ gene segments in precursor T lymphocytes.
J. Immunol.
161
:
1399
16
Sutton, K. A., M. N. Vu, M. F. Wilkinson.
1998
. Distal Vβ promoters transcribe novel T-cell receptor-β transcripts in early development.
Immunology
93
:
213
17
Halle, J. P., P. Haus-Seuffert, C. Woltering, G. Stelzer, M. Meisterernst.
1997
. A conserved tissue-specific structure at a human T-cell receptor β-chain core promoter.
Mol. Cell. Biol.
17
:
4220
18
Rothenberg, E. V..
2000
. Stepwise specification of lymphocyte developmental lineages.
Curr. Opin. Genet. Dev.
10
:
370
19
Verbeek, S., D. Izon, F. Hofhuis, E. Robanus-Maandag, H. te Riele, M. Van de Wetering, M. Oosterwegel, A. Wilson, H. R. MacDonald, H. Clevers.
1995
. An HMG-box-containing T-cell factor required for thymocyte differentiation.
Nature
374
:
70
20
Ratanavongsiri, J., S. Igarashi, S. Mangal, P. Kilgannon, A. Fu, A. Fotedar.
1990
. Transcription of the T cell receptor β-chain gene is controlled by multiple regulatory elements.
J. Immunol.
144
:
1111
21
Messier, H., J. Ratanavongsiri, T. Fuller, S. Mangal, P. Kilgannon, R. Fotedar, A. Fotedar.
1992
. Mapping of an inducible element in the T cell receptor Vβ2 promoter.
J. Immunol.
149
:
1980
22
Clark, S. P., B. Arden, D. Kabelitz, T. W. Mak.
1995
. Comparison of human and mouse T-cell receptor variable gene segment subfamilies.
Immunogenetics
42
:
531
23
Rothenberg, E. V., D. Chen, R. A. Diamond.
1993
. Functional and phenotypic analysis of thymocytes in SCID mice: evidence for functional response transitions before and after the SCID arrest point.
J. Immunol.
151
:
3530
24
Ema, H., A. Cumano, P. Kourilsky.
1997
. TCR-β repertoire development in the mouse embryo.
J. Immunol.
159
:
4227
25
Ohteki, T., H. R. MacDonald.
1996
. Stringent Vβ requirement for the development of NK1.1+ T cell receptor-α/β+ cells in mouse liver.
J. Exp. Med.
183
:
1277
26
Heinemeyer, T., E. Wingender, I. Reuter, H. Hermjakob, A. E. Kel, O. V. Kel, E. V. Ignatieva, E. A. Ananko, O. A. Podkolodnaya, F. A. Kolpakov, N. L. Podkolodny, N. A. Kolchanov.
1998
. Databases on transcriptional regulation: TRANSFAC, TRRD, and COMPEL.
Nucleic Acids Res.
26
:
364
27
Okada, C. Y., I. L. Weissman.
1989
. Relative Vβ transcript levels in thymus and peripheral lymphoid tissues from various mouse strains: inverse correlation of I-E and Mls expression with relative abundance of several Vβ transcripts in peripheral lymphoid tissues.
J. Exp. Med.
169
:
1703