INTRODUCTION

Top Abstract Introduction Materials & Methods Results Discussion References

T-lymphocyte development is similar to that of B-cell maturation in that the germ line variable (V),¹ diversity (D), and joining (J) gene segments of the T-cell antigen receptor (TCR) are somatically recombined to form a V-D-J (for -chain) or V-J (for -chain) exon encoding the variable domain portion of the receptor. Studies in B-cells support the idea that to achieve the recombination, the rearranging Ig gene segments must be transcriptionally activated (1-4). If this prerequisite also applies to TCR genes, then the promoter elements that control transcriptional activity of TCR V genes will play a pivotal role in rearrangement. Specific transcriptional factors may bind to distinct DNA sequence motifs in the promoter region and control transcription. These factors can also affect the accessibility of germ line loci to the recombinational machinery, thus regulating gene expression in a tissue- or developmental stage-specific way.

A 10-base pair decamer sequence motif, AGTGAYRTCA, was found to be conserved in the promoter regions of most murine and human TCR V genes (5-7). Elimination of this decamer motif in the murine V8.3 promoter reduced transcriptional activity (8). Binding of thymic factors to this decamer motif was found to be developmentally regulated, and no decamer binding activity was detected in nuclear extracts prepared from thymuses of severe combined immunodeficiency mice, suggesting that the decamer motif plays an important role in the connection between murine TCR V gene transcription and rearrangement (9).

The most highly conserved portion in the decamer motif is an inverted repeat with the sequence TGA-TCA. This palindromic feature links the decamer to other regulatory elements, such as the c-AMP response element, TGACGTCA, and the AP-1 binding site, TGACTCA (10, 11). We have observed some discrete differences in the location and the consensus sequence of the decamer motif between different human TCR V subfamilies (7). Little is known regarding the role of the decamer motif in promoter function or whether differences in the decamer motif may affect promoter activity.

In this study, we have characterized 14 promoters of human TCR V genes, with a specific emphasis on V6.7. By using a transient transfection system, we have shown that human TCR V promoters vary in terms of their strength. Whereas most promoters tested are weak, the promoters of the V6 subfamily, particularly V6.7, are constitutively strong. Sequence analyses have revealed three decamer-like motifs in the promoter region of V6.7, and each of them contributes differently to promoter activity. The one located most 5' to the transcription initiation site has essentially no effect on promoter activity. In contrast, the one located proximal to the initiation site forms the major contribution to promoter strength. A 5' deletion study has shown that a fragment of 147 bp immediately 5' upstream of the transcription initiation site can constitutively drive the reporter luciferase gene expression. Further shortening of this fragment results in a decrease in reporter gene expression. Therefore, this fragment represents the minimal promoter element of V6.7. Inside this fragment there are a CACCC motif, a TATA box, and two decamer-like motifs. Further site-directed mutagenesis analyses have shown that a 12-nucleotide sequence, AGTGATGTCACT, is responsible for the high promoter activity. Therefore, the regulatory element in the promoter of V6.7 should be considered a dodecamer rather than a decamer. The possible contributions to V6.7 promoter activity from additional transcription factor binding sites 3' behind the dodecamer are also discussed.

	MATERIALS AND METHODS

Top Abstract Introduction Materials & Methods Results Discussion References

Cell Culture and Total Cellular RNA Isolation-- Human peripheral blood T-cells were cultured in RPMI 1640 medium with 10% fetal calf serum. CD8⁺ T-cells were depleted following treatment with an anti-CD8 monoclonal antibody (OKT 8) and anti-mouse Ig-coated magnetic beads (Dynal, Great Neck, NY). CD4⁺ T-cells were activated with anti-V6.7-specific monoclonal antibody OT145 (12). After a second retriggering, more than 80% of T-cells were V6.7-positive. Cells were washed once with phosphate-buffered saline, and the pellet was spun down. Total cellular RNA was isolated using an acidified guanidinium/phenol/chloroform isolation kit (RNazol, TEL-TEST, Friendswood, TX) and used to identify the transcription initiation sites by RNase protection and primer extension methods.

RNase Protection and Primer Extension-- To prepare an antisense RNA probe, a double-stranded DNA fragment spanning the first and the second exons plus the 5'-untranslated portion of V6.7 was subcloned into a T/A vector that contains a T7 promoter. The template plasmid DNA was digested upstream at an unique XhoI site. The antisense RNA probe was synthesized with a Riboprobe-T7 system (Promega, Madison, WI) in the presence of T7 RNA polymerase and [-³²P]CTP. The labeled RNA probe was hybridized with total cellular RNA isolated from V6.7⁺ T-cells. The protected products were purified once with phenol/chloroform extraction, denatured, and loaded on a 6% polyacrylamide, 7 M urea sequencing gel. The dideoxyl-terminate sequencing reaction product from a plasmid DNA with a known sequence was loaded alongside as a size marker. After electrophoresis, the gel was dried and exposed to x-ray film at 70 °C for 12-48 h.

Preparation of Constructs and Site-directed Mutagenesis-- The 5'-flanking sequences were PCR-amplified from either genomic DNA or  phage DNA clones. The sequences of 5' sense and 3' antisense oligonucleotide primers used for 14 human TCR V genes were listed in Table I. These PCR amplified fragments were ~500 bp and were inserted 5' to the reporter luciferase gene segment in pGL-2/Basic vector (Promega). The sequences and orientations of these fragments inside pGL-2 were confirmed by sequencing. The plasmid DNAs used for transfection were purified by cesium chloride banding. The site-directed mutagenesis was achieved using the PCR-overlapping method (13).

View this table: [in this window] [in a new window]   Table I The oligonucleotide primers used in PCR to amplify the 5'-flanking sequences of TCR V gene segments

Transient Transfection-- Human Jurkat T-cell line cells were grown to confluence in RPMI 1640 medium plus 10% fetal bovine serum and other supplements. Cells were transfected with plasmid DNA containing promoter-reporter constructs (described above) by a modified DEAE-dextran electroporation method. Briefly, cells were washed twice with serum-free RPMI 1640 medium and resuspended in the same medium at a concentration of 17 × 10⁶/ml. 600 µl of cell suspension was placed in a 0.45-cm electroporation cuvette (Gene Pulser, Bio-Rad) followed by adding 180 µl of DEAE-dextran (100 µg/ml) and 10 µg of promoter-reporter plasmid DNA. To evaluate the transfection efficiency, 5 µg of plasmid DNA containing a -galactosidase gene segment was co-transfected into the cells. The electroporation was carried out at a capacitance of 960 microfarads and 250 V. After electroporation, the cells remained at room temperature for 10 min and then were resuspended in 10 ml of complete RPMI 1640 medium, placed in wells (3 ml/well) of a 12-well plate, and incubated for 48-72 h at 37 °C in 5% CO₂.

Luciferase Assay-- 5× luc stock buffer, which contains 125 mM glycyl glycine, 75 mM MgSO₄, and 20 mM EGTA, was premade and used later in preparation of lysis solution, assay solution, and luciferin substrate solution. The transfected cells were washed twice with 1× phosphate-buffered saline, pelleted, and lysed with 500 µl of lysis solution (1× luc buffer, 1% Triton X-100, 1 mM of dithiothreitol added immediately prior to use). 100 µl of cell lysate was added to a test tube that contained 500 µl of assay solution (1× luc buffer, 1 mM KHPO₄, 1 mM dithiothreitol, and 2 mM of ATP). The tube was loaded into the measurement chamber, and 100 µl of luciferin substrate solution (1× luc buffer, 0.4 mM D-luciferin-sodium salt, and 2 mM dithiothreitol) was injected automatically. The relative light units (RLUs) were measured on a Monolight 2001 Luminometer (Analytical Luminescence Laboratory, San Diego, CA). The background RLUs from cells transfected with vector pGL-2 only were measured in each experiment. The luciferase assay data were calibrated with -galactosidase activity and are presented as averages in triplicate measurements.

-Galactosidase Assay-- The -galactosidase activity was measured by using a Galacto-Light Plus Chemiluminescent Reporter assay kit following the manufacturer's instructions (Tropix, Bedford, MA).

Computer Analysis of 5'-Flanking Region Sequences-- Putative regulatory elements in the 5'-flanking region of human TCR V genes were identified by searching the Genetics Computer Group (version 8.1, Madison, WI) transcription factor binding sites data base (TFSITE) through Rockefeller University Computer Services. The decamer-like motifs were screened by the FINDPATTERNS program.

	RESULTS

Top Abstract Introduction Materials & Methods Results Discussion References

Decamer Motif and TCR V Promoter Activities-- To determine the role of decamer sequence motifs in promoter activity, we have searched for this regulatory element in promoter regions among 54 functional TCR V genes based on their genomic DNA sequences (Hood et al., GenBank accession no. L36092). The consensus sequence for the decamer is AGTGARYTCA. However, we observed some discrete differences in different TCR V gene subfamilies, such as AGTGATGTCA in V6, AGTGACATCA in V5, and TGANNNNTCA in V13 subfamilies (7). Therefore, we have screened two general patterns of the decamer motif, NNTGANNTCA and TGANNNNTCA, in the promoter regions ~500 bp 5' to the translation initiation codon ATG, using the FINDPATTERNS computer program. Among 54 functional human TCR V gene families, 42 V promoters contain these general patterns (Table II). 22 Vs each contain one NNTGANNTCA element. 8 Vs each contain two such elements, and the remaining 5 Vs each contain three such elements. Of these 5 Vs, 4 come from the V6 subfamily. They are V6.3, 6.7, 6.11, and 6.14. All V13 subfamilies, except V13.1, contain the TGANNNNTCA pattern.

View this table: [in this window] [in a new window]   Table II Decamer motifs in human TCR V gene promoters

View larger version (25K): [in this window] [in a new window]   Fig. 1.   Promoter activities from 13 human TCR V genes measured by RLUs in a luciferase assay. Data were calibrated with the -galactosidase activities and are presented here as the average of triplicate measurements.

Allelic Variations in Promoter Region of V6.7-- We previously identified two alleles of the human TCR V6.7 gene, V6.7a and V6.7b (14). They differ at two amino acid positions inside the coding region: V6.7a encodes Ser³⁸ and Gly⁷², whereas V6.7b encodes Arg³⁸ and Glu⁷². An allele-specific monoclonal antibody, OT145, can recognize the product of 6.7a but not 6.7b (15). Later, Vissinga et al. (16) found that the peripheral expression of these two alleles in heterozygous individuals was skewed, indicating that an allelic polymorphism in the coding region can have a significant impact on gene expression in the peripheral repertoire. However, other polymorphisms in the TCR  locus, such as those in the promoter region, may also affect gene expression. To address this possibility, we analyzed the 5'-flanking sequences of V6.7 alleles a and b. First, a 463-bp sequence upstream of V6.7a in phage  DNA clone 5-2 (7) was obtained. The promoter segment of V6.7 allele b was then PCR amplified from the genomic DNA of a b/b homozygous individual. Sequence analyses revealed three point mutations between alleles a and b within the 463-bp 5'-flanking region that were confirmed in a total of 24 plasmid clones derived from three V6.7 a/a and three V6.7 b/b homozygous individuals (Fig. 2). These mutations represent allelic variations in the promoter region of V6.7. We expected that no significant difference in promoter activities would be detected between V6.7 alleles a and b because no destruction of any major regulatory elements was caused by these mutations. Our expectation was confirmed by luciferase reporter assay (data not shown).

View larger version (47K): [in this window] [in a new window]   Fig. 2.   5'-Flanking sequence alignment of human TCR V6.7 allele a versus allele b. Three point mutations in the promoter region were identified and confirmed as allelic variations. The transcription initiation site is marked with an asterisk (*). The three decamer-like motifs are underlined and marked D1, D2, and D3. The coding region sequence of V6.7 is presented in boldface type, and the intron sequence is shown in lowercase letters.

Transcription Initiation Site of V6.7-- To determine the transcription initiation site of V6.7, primer extension analysis was performed. An antisense oligonucleotide primer, 5'-TGTGATCTGCCCCCAGGA-3', was specifically designed. In this primer, 12 nucleotides match the 3'-end sequence of the first exon, and the remaining 6 nucleotides match the 5'-beginning of the second exon of V6.7 (see the coding sequence of V6.7 in Fig. 2). Only cDNA, but not the genomic DNA, can anneal to this primer. In Fig. 3, lanes C, T, A, and G were sequencing products of a plasmid DNA with known sequence, shown here as the size marker. Lane 1 was the primer extension product incubated with RNA isolated from V6.7⁺ T-cells. Lane 2 was the primer extension product incubated with RNA isolated from mouse thymoma cell line BW5147, included here as negative control. Two intense bands were observed in lane 1. The higher one is 63 bp, and the lower one 60 bp. We assumed that the 63-bp band represents the full length extension product. This product aligns with an A 26 bp 5' to the ATG codon. Given that adenine is a favored base for eukaryotic transcription initiation (17), we defined it as the initiation site for human V6.7; it is marked with an asterisk in Fig. 2. This transcription initiation site was further confirmed by RNase protection assay (data not shown) and is consistent with the findings in most murine TCR V genes, in which the cap sites fall into a range of 19-40 bp 5' to the ATG codon (6). The lower band of 60 bp in Fig. 3, lane 1, may represent the partial extension product.

View larger version (83K): [in this window] [in a new window]   Fig. 3.   The transcription initiation site of V6.7 determined by primer extension analysis. Lanes C, T, A, and G represent sequencing products of a plasmid DNA with known sequence as size markers. Lane 1 represents the primer extension products when 32P-labeled primer was annealed to the total RNA isolated from human V6.7+ cells. Lane 2 represents the extension products when 32P-labeled primer was incubated with cellular RNA isolated from a mouse hybridoma cell line BW5147. Two distinct bands of 63 and 60 bp in lane 1 represent the full and partial extension products.

5'-Deletion and Fragmentation Analyses of V6.7-- Three decamer-like motifs were identified in the promoter region of V6.7. They were located at positions 84, 111, and 352 and are designated D1, D2, and D3 (Fig. 2). To evaluate the contribution of each decamer or other transcription factor binding elements to the promoter activity, the 5'-flanking sequence of V6.7 was fragmented. Each fragment was ligated into the luciferase reporter gene construct (Fig. 4, left panel), and then RLUs were measured (Fig. 4, right panel). Fragment I, from 1 to 463, representing the entire 5'-flanking sequence of V6.7, can drive the luciferase activity to ~700-fold higher than the background of pGL-2/Basic vector only. Fragment II, from 463 to 336, which contains one decamer-like motif D3, had essentially no effect on luciferase activity. The same result was obtained with fragment III (336 to 127). In contrast, when fragment IV (147 to 1) was used, the RLUs were even higher than the level induced by the entire 5'-sequence (fragment I). 5'-Deletion further created two small fragments: fragment V (147 to 80), which contains D2 only, and fragment VI (89 to 1), which contains D1 only. As shown in Fig. 4, right panel, the RLUs for fragment V were low, and the RLUs for fragment VI were high. However, fragment VI was unable to reach the same level of activity as that induced by fragment IV. Therefore, the 147-bp of fragment IV represents the basic promoter for V6.7.

View larger version (13K): [in this window] [in a new window]   Fig. 4.   5'-Flanking sequence deletion and fragmentation analyses for V6.7. The left panel indicates the composition of constructs containing the 5'-promoter region of V6.7 that were used in transfection studies. The right panel represents the results of a luciferase assay using Jurkat T-cells transfected with each of these constructs. Results are expressed as the average RLUs of triplicate measurements.

No T-cell Lineage Specificity for Promoter Element in V6.7-- To determine whether the constitutive activity of the promoter element in V6.7 was specific to the T-cell lineage, several cell lines, including human T-cells (Jurkat), Burkitt's lymphoma cells (Ramos), fibroblast cells (cos 7), mouse thymoma cells (BW5147), and human embryonic kidney cells (293) were used. The luciferase reporter constructs containing TCR V6.7 promoter segment and a nonspecific Rous sarcoma virus promoter were used to transfect these different cell lines. When compared with the nonspecific Rous sarcoma virus promoter, the RLUs were relatively high for V6.7 promoter in lymphocyte cell lines (Jurkat, Ramos, and BW5147) but low in fibroblast cell cos 7 and embryonic kidney cell lines (data not shown). The effect of T-cell activation by stimuli that mimic TCR ligation, such as phorbol 12-myristate 13-acetate and ionomycin, on the promoter activity was further examined. As shown in Fig. 5, there were no significant changes in promoter activity when cells were stimulated by phorbol 12-myristate 13-acetate and ionomycin for 2-12 h, whether the basic 147-bp or the entire 463-bp V6.7 promoter fragments were tested. Our data indicate that V6.7 promoter element is lymphocyte-specific but not T-cell lineage-specific.

View larger version (23K): [in this window] [in a new window]   Fig. 5.   Luciferase assay to evaluate the effect of phorbol 12-myristate 13-acetate and ionomycin on V6.7 promoter activity. 1 × 106 cells were transfected with the construct containing the entire 463-bp promoter sequence of V6.7. Phorbol 12-myristate 13-acetate (20 ng/ml) and ionomycin (500 ng/ml) were added 12 h after transfection. Cell lysates were obtained at 2, 4, 6, and 12 h after stimulation and used for luciferase assay.

Dodecamer, Rather Than Decamer, Is Critical to V6.7 Promoter Activity-- To further evaluate the role of decamer D1 and its surrounding nucleotides in promoter activity, site-directed mutagenesis was performed. When D1 (AGTGATGTCA) was mutated to D1 m1 (AGAGATGCCA), in which the two Ts that form the core element of TGA-TCA were substituted, one by an A and the other by a C, the luciferase activity was totally abolished (Fig. 6), suggesting that the palindromic TGA-TCA is critical. We further mutated two nucleotides immediately 3' of D1, which changed AGTGATGTCACT to AGTGATGTCAAA (Fig. 6, D1 m2). This mutation destroys the reverse complementation between the first two nucleotides, AG, to the last two nucleotides, CT, in the original sequence. As shown in Fig. 6, the luciferase activity was decreased to two-thirds that of the original, nonmutated sequence, indicating that these two nucleotides are also important to promoter activity. Therefore, the regulatory motif in V6.7 should be considered a dodecamer (12-bp) rather than a decamer.

View larger version (18K): [in this window] [in a new window]   Fig. 6.   Site-directed mutagenesis of decamer motif D1 in the promoter of V6.7. The left panel shows a schematic diagram of constructs used. The right panel shows the results of a transfection experiment using these constructs. Construct 1, a 147-bp basic promoter element that contains two decamer-like motifs, D1 and D2, and shows the full promoter activity. Construct 2, the same fragment as Construct 1, but inserted into pGL-2 in the opposite orientation. Construct 3, the same fragment as Construct 1, but D1 was mutated to D1 m1. Construct 4, the same fragment as Construct 1, but D1 was mutated to D1 m2. In D1 m1, two Ts that form the core structure of TGA-TCA were replaced, one by A and another by C. In D1 m2, two nucleotides, CT, 3' to the end of the decamer motif, were mutated to AA.

View larger version (18K): [in this window] [in a new window]   Fig. 7.   Augmented strength of the V5.2 promoter fragments containing the V6.7 dodecamer motif. A luciferase assay was performed using the following constructs: pGL-2/Basic, vector only; pVb5.2, the original V5.2 promoter fragment; pVb5.2/*, the V5.2 promoter in which the original imperfect dodecamer was corrected; and pVb5.2/dd, the V5.2 promoter in which the original imperfect dodecamer and its 3' behind sequence was replaced by an 85-bp fragment from 85 to 1 of the V6.7 promoter, which contains the dodecamer motif AGTGATGTCACT.

View larger version (8K): [in this window] [in a new window]   Fig. 8.   Promoter activity comparison between V6 subfamily members. Constructs used in transient transfection and luciferase assays were pVb6.7d1, pVb6.1d1, and pVb6.3d1. These constructs contained the 3'-end promoter sequences from V6.7, 6.1, and 6.3, respectively.

View larger version (31K): [in this window] [in a new window]   Fig. 9.   3'-End sequence alignment for promoters of V6.7, 6.1, and 6.3. All three promoters contain the same dodecamer motif (shown in boldface type). Three additional transcription factor binding sites, BS15 (underlined), SIF_core_RS (marked with * above the line), and Sp1-U2snR.2 (double-underlined), were identified in V6.7 but not in V6.1 or 6.3. The consensus and reverse complementary sequences for these sites are listed in the bottom panel.

	DISCUSSION

Top Abstract Introduction Materials & Methods Results Discussion References

The process of T-lymphocyte development is similar to that of B-cell maturation. Transcriptional activation of rearranging Ag receptor gene segments has been hypothesized to regulate their accessibility to the recombinational machinery. Several pieces of evidence support this prerequisite in both B-cells and T-cells (1-4). Therefore, an understanding of transcriptional regulation may be important in elucidating the mechanisms controlling the lineage-specific patterns of rearrangement of the TCR genes during thymocyte differentiation.

Human TCR V gene promoters themselves, which control the transcription, were reported to be essentially inactive. It was the enhancer, located 3.5-5.0 kilobase pairs 3' to the C2 gene segment, that conferred the transcriptional activity and T-cell lineage specificity to V promoters (18-20). However, in this study, we found that V promoters varied in terms of their activities. Although most of the V promoters that we tested were very weak, some were constitutively strong, such as those from V6 subfamilies. The promoter of V6.7 was found to be the strongest when compared with others. The 5'-flanking sequence analyses revealed that there are three decamer-like sequence motifs in the promoter region of V6.7. There seems to be a correlation between the promoter activity and decamer-like motifs. For example, the strong promoters, such as V6.3 and 6.7, all contain three decamer-like sequences. The most highly conserved portion in the decamer is an inverted repeat with the core sequence TGA-TCA. We identified the decamer as AGTGATGTCA in V6.7, AGTGACATCA in V5.2, and TGANNNNTCA in V13 subfamilies (7). The general idea is that the core regulatory element is composed of TGA-TCA. Those nucleotides in front of TGA, after TCA, or between TGA and TCA may be not so important. However, our results show that it is not so simple. Certainly, TGA-TCA is critical. When this element was destroyed by site-directed mutagenesis, the promoter activity was totally abolished (see Fig. 6, Construct 3). However, TGA-TCA alone is not sufficient to confer promoter activity. For instance, the decamer-like motif D3 in the 5'-distal segment of the V6.7 promoter was found to have essentially no effect on promoter activity (Fig. 4, II). Motif D2 in V6.7 also showed a very low contribution to the promoter activity (Fig. 4, V). The surrounding nucleotides 5' and 3' to TGA-TCA, AG and CT, were also important. When the CT at the 3'-end was mutated to AA, the promoter activity dropped to two-thirds of the original activity (see Fig. 6, Construct 4). These data support our hypothesis that the most critical regulatory element is a 12-nucleotide motif with a core sequence AGTGA-TCACT. Therefore, it is a dodecamer, rather than a decamer. Further evidence to support this idea comes from the study of promoter activities for V5.1 and 5.2. Although these promoters also contain a dodecamer-like motif at relatively the same location as the dodecamer in the V6.7 promoter, their activities were found to be relatively low (Fig. 1). Not surprisingly, the dodecamer-like motif in the promoters of V5.1 and 5.2 has the sequence AGTGACATCACA, which is not a perfect dodecamer as in the V6 genes. When this imperfect dodecamer in V5.2 was corrected by site-directed mutagenesis, the promoter activity was found to be increased. However, only when its 3'-end was replaced by a fragment from 85 to 1 of V6.7 did the V5.2 promoter became very strong, with activity comparable to V6.7 (see Fig. 7).

The palindromic features lead us to propose that the dodecamer may form a stem and loop structure by reverse complementation. Such a stem-loop structure will exert the regulatory function by binding with homo- or heterodimers of transcription factors, such as members of the cyclic AMP response element binding protein or activating transcription factor families (21-23). To test this hypothesis, we simply compared the light outputs for constructs in which the 147 bp of fragment IV were inserted in sense or antisense orientations (Fig. 6, Constructs 1 and 2). No significant difference in luciferase activity was observed between the two constructs, suggesting that this stem-loop model may stand. Together, these results indicate that a perfect dodecamer motif, which can form a stem-loop structure, plays an important role in TCR V promoter activity.

In addition to the dodecamer motif, there are other elements, such as the TATA box (at 123) and CACCC box (at 140), that also contribute to the promoter activity of V6.7. In murine TCR V genes, the decamer motifs were identified in the promoter region 10-40 bp upstream of the TATA box (6). In human TCR V genes, the locations of the decamer motif are different. The decamer in the V8 subfamily, similar to murine TCR Vs, is located ~20 bp upstream of the TATA box. Interestingly, the dodecamer in the promoters of V6.1, 6.3 and 6.7 were located 36 bp downstream of the TATA box. It was unclear whether these different locations might affect the promoter activities. The CACCC box, identified in human V8.1 and mouse TCR  gene silencer (24), was found to be another transcription factor binding site with T-cell lineage specificity (25, 26). We found the CACCC box in the 150-bp basic promoter fragment from V6.7. The fragment that contains the dodecamer only, without TATA and CACCC boxes, showed partial promoter activity (Fig. 4, VI), indicating that TATA and CACCC boxes may play a synergistic role with the dodecamer and contribute to the high constitutive promoter activity of V6 genes. Other motifs identified 3' behind the dodecamer of V6.7, such as BS15, SIF_core_RS, and Sp1-U2snR.2, are also important but have not been characterized. Further analyses will help us to evaluate the contributions of these binding sites.