INTRODUCTION

Top Abstract Introduction Materials and Methods Results Discussion References

Antigen receptor genes are assembled during B- and T-cell development by a series of site-specific DNA recombination reactions known as V(D)J recombination (58). The lymphocyte-specific gene products RAG-1 and RAG-2 are essential components of the V(D)J recombinase complex (38, 44, 51). Together, they recognize recombination signal sequences which flank rearranging gene segments and introduce double-stranded DNA breaks between these signals and gene-containing DNA segments (16). A null mutation in either gene prevents V(D)J recombination and completely blocks lymphocyte development at the early progenitor stage (38, 51).

The RAG-1 and RAG-2 genes are physically linked in the genomes of all chordates in which they have been studied; they are convergently transcribed and separated by approximately 8 kb of DNA (50). Coupled with the recent observation that the RAG proteins have DNA transposase activity in vitro, these unusual structural features of the RAG locus have led to the suggestion that RAG-1 and RAG-2 were once part of a transposable-element system (1, 26, 56).

Transcription of the RAG-1 and RAG-2 genes is limited to specific stages of B- and T-cell development. Transcription can be detected in the earliest T- and B-cell progenitors and remains high until the complete assembly of the T-cell receptor (TCR)  chain gene or the immunoglobulin (Ig) heavy-chain gene (19, 60, 63). During the midstages of lymphoid development, RAG transcription diminishes significantly, coincident with several rounds of cell division. Pre-T and pre-B cells then exit the cell cycle and increase RAG transcription, leading to rearrangement of the TCR  or Ig light-chain loci (19, 63). RAG transcription continues in T cells until positive selection occurs, at which time expression is extinguished via a TCR-dependent signal (60). The situation in B cells is more complex. RAG transcription generally stops when a complete Ig molecule is expressed on the cell surface (19). However, if that Ig recognizes self-antigen, RAG expression is stimulated, and recombination continues in a process known as receptor editing (15, 35, 57). Thus, regulated RAG expression contributes to self-tolerance of the B-cell repertoire. Finally, the RAG genes are not expressed in mature peripheral T and B cells; however, there are some data which suggest that transcription [as well as V(D)J recombination] can be reactivated in B cells during an antigen-specific immune response (20, 21, 24, 25, 46). Given the complex regulation and critical involvement of the RAG genes in lymphocyte development, efforts have been made to decipher the molecular basis of their transcriptional regulation.

Previous studies performed in our own laboratory and in others have described the general structure of the murine and human RAG-1 and RAG-2 promoters (13, 31, 32, 65). We reported recently that unlike the RAG-1 promoter, the murine RAG-2 promoter displays cell-type specificity in transient-transfection assays (32). A RAG-2 promoter-reporter construct containing the transcription start site and 279 bp of 5' flanking DNA was fully active in both T- and B-cell lines. Surprisingly, we found that a 5' promoter deletion mutant extending to nucleotide 71 with respect to the start site retained full activity in B cells but lost approximately 70% of its activity in T cells. Further deletion of the promoter to position 45 eliminated nearly all activity in both B and T cells (32). The DNA sequences of the murine and human RAG-2 promoters are identical between nucleotides 70 and 50. Within this region, we identified a binding site for the B-cell-specific transcription factor BSAP (Pax-5) and showed that BSAP binds this sequence both in vitro and in vivo. Furthermore, mutations which disrupt BSAP binding greatly diminish B-cell-specific promoter activity. T cells, however, do not express BSAP but still require sequences in the conserved 70 to 50 promoter region for RAG-2 promoter activity (32). Recent reports suggested that c-Myb, a hematopoietic lineage-restricted transcription factor, is preferentially expressed in developing T but not B cells (2, 11) (see below). In the experiments described below, we have gone on to determine that c-Myb binds to the RAG-2 promoter and is critical for RAG-2 promoter activity in T cells.

	MATERIALS AND METHODS

Top Abstract Introduction Materials and Methods Results Discussion References

Cells and cell culture. Thymocytes used in the chromatin immunoprecipitation (ChIP) assay were isolated from 3-week-old C57BL/6J mice (Jackson Laboratory). Jurkat, a mature human T-cell lymphoma line, and 2017 (54), an Abelson virus-transformed immature murine T-cell line, were grown at 37°C and 5% CO₂ in RPMI 1640 (Mediatech) supplemented with 10% fetal calf serum, penicillin-streptomycin, L-glutamine, and 50 µM -mercaptoethanol. 293T, a human embryonic kidney cell line expressing simian virus 40 T antigen, and HeLa (human cervical carcinoma) cells were grown in Dulbecco's modified Eagle's medium (Mediatech) with 1 g of glucose per liter, supplemented with 10% fetal calf serum, L-glutamine, penicillin-streptomycin, and 50 µM -mercaptoethanol.

Plasmid constructs. The RAG-2 promoter constructs used in the various cotransfection assays have been described previously (32). The eukaryotic c-Myb expression construct was created by cloning the murine c-Myb cDNA (a gift from Chi Dang) into the pEFBneo vector (a modified version of pEF-BOS [37]). The basic stable-transfection reporter construct consisted of the PGK-neo drug resistance cassette and the green fluorescent protein (GFP) cDNA, intron, and polyadenylation signal from pGreenLantern (Gibco-BRL) separated by the pBSK polylinker. Tandem copies of the 1.2-kb chicken -globin insulator were excised from pJC13 (10) and cloned into the XbaI site of the polylinker. The RAG-2 promoter from 279 to +123 was cloned in the proper orientation upstream of the GFP cDNA. A 4-kb HindIII fragment containing the E enhancer core was cloned into polylinker sites upstream of the promoter and downstream from the insulator sequences.

Transient-transfection reporter assay. For cotransfection experiments, 10 µg of the 279 to +123 reporter, 2 µg of either empty pEFBneo vector or pEFBneoMyb, and 100 ng of p-CMV--gal were used. 293T cells were transfected with GenePorter reagent (Gene Therapy Systems) and 2017 cells were transfected with Superfect reagent (Qiagen) according to the manufacturers' instructions. The cells were harvested as described previously (32). Luciferase activity was measured with a luminometer (Analytical Luminescence Laboratories). -Galactosidase assays were performed with the Galacto-Light Plus Kit (Tropix) according to the manufacturer's instructions. Luciferase activity for each sample was normalized to the -galactosidase assay control.

Stable-transfection reporter assay. Twenty micrograms of each linearized construct in 40 µl of sterile phosphate-buffered saline was mixed with 10⁷ cells resuspended in 800 µl of culture medium. The cells were in the logarithmic phase of growth. The mixture of cells and DNA was electroporated at 250 V and 960 µF in a 0.4-cm-gap electroporation cuvette (Bio-Rad). The cells were grown in 50 ml of medium for 48 h without G418, and then 3 × 10⁴ cells were seeded into each well of two 24-well plates in the presence of 2 mg of G418 (Life Technologies)/ml. After 12 to 14 days of selection, all wells were positive, and they were pooled. This ensured an adequate number of founder cells, each of which represents an independent integration event.

ChIP. The formaldehyde cross-linking and immunoprecipitation experiments were done as reported by Boyd and colleagues (9). In brief, 37% formaldehyde solution (Fisher Scientific) was added directly to cell culture medium at a final concentration of 1%. In the case of the mouse thymocytes, the solution was added to a single-cell suspension of thymocytes in culture medium. Cross-linking was allowed to occur at room temperature for 10 min. The cells were lysed, and the nuclei were collected and resuspended in sonication buffer. Chromatin was sonicated to an average length of 400 to 800 bp, and the suspension was precleared with blocked Staphylococcus A cells. Chromatin from 4.5 × 10⁷ cells was then incubated with 4 µg of c-Myb-specific rabbit polyclonal antibody (for Jurkat cells, anti-human c-Myb; for mouse thymocytes, anti-mouse c-Myb [Santa Cruz]), affinity-purified rabbit IgG (Jackson Immunoresearch), or no antibody and rotated overnight at 4°C. The immune complexes were precipitated, washed, and eluted. After precipitation, supernatant from the "no-antibody" sample was processed to the cross-link reversal step and analyzed as unfractionated input chromatin. Cross-links were reversed. After proteinase K digestion, samples were phenol extracted, and DNA was ethanol precipitated and resuspended in 30 µl of H₂O. Two microliters of immunoprecipitate or 50 ng of total input DNA was used for 28 cycles of PCR amplification. PCR products were analyzed by electrophoresis on a 2% agarose gel and visualized by ethidium bromide staining or denatured and transferred onto a Hybond-N membrane (Amersham) for Southern blot analysis. The following primers were used in the ChIP assays: human RAG-2 promoter, hR2U (5'GTGAATTGTGTTGCCATTGTTGC) and humR2P-2 (5'TGTGTGCCTACAGATGTTC); human RAG-2 coding exon, hRAG-2exon2/F (5'GTTCTTCTGCTGAAAGTTC) and HRAG-2exon2/B (5'GGTGATGGAAACAACAAAAG); human -actin, hBA1 (5'CATGTGCAAGGCCGGCTTCG) and hBA2 (5'GAAGGTGTGGTGCCAGATTT); mouse RAG-2 promoter, R2F1 (5'CAACCATCACAGGGGTGCAG) and R2R2 (5'GCCTACAGATGTTCCAGTGAG); and mouse Ig kappa locus, Jk1F (5'ACCAGATTCTGGCACTCTCC) and JkREV (5'GAGTAAGATTTTATACATCATTTTTAGACA).

DNase I genomic footprinting. DNase I genomic footprinting was performed as described by Weinmann et al. (62). In brief, Jurkat or HeLa cells were harvested and resuspended in cold NP-40 lysis buffer. Samples were incubated on ice for 5 min, and intact nuclei were pelleted and resuspended in DNase I digestion buffer. Genomic DNA was purified from Jurkat cells by proteinase K treatment, phenol extraction, and ethanol precipitation. DNase I (Boehringer Mannheim Biochemicals) was added to nuclei (from 3 × 10⁶ cells) or pure genomic DNA (from 8 × 10⁶ cells) to various final concentrations (nuclei, 2 to 10 U/sample; genomic DNA, 0.2 to 0.8 U/sample) and incubated on ice for 5 min. Reactions were stopped and processed for DNA purification. DNA was resuspended in H₂O, and its concentration was determined by spectrophotometry. One microgram of DNA was subjected to ligation-mediated PCR. The PCR cycles were as follows: for primer extension, 95°C for 5 min, 59°C for 30 min, and 76°C for 10 min; for amplification of ligated products, 95°C for 4 min, 60°C for 2 min, and 76°C for 3 min for one cycle, 20 cycles of 95°C for 1 min, 60°C for 2 min, and 76°C for 3 min 15 s/cycle, and final extension for 10 min at 76°C; for labeling reactions, 95°C for 4 min, 61°C for 2 min, and 76°C for 10 min for one cycle and two cycles of 95°C for 1 min, 61°C for 2 min, and 76°C for 10 min/cycle. The following locus-specific primers were used: primer extension, 5' TGCCGCTAAACCAGGTATTAA; amplification, 5' TTAATTGTCAGCACTTGGGGA; end labeling, 5'ATTGTCAGCACTTGGGGAAGA; and DNA sequencing, 5' ATCTTTGCCGCTAAACCAGGT. Samples were heated to 90°C for 5 min and analyzed by electrophoresis on a 6% polyacrylamide-7 M urea denaturing gel alongside a DNA sequencing ladder. The gels were dried and exposed to PhosphorImager screens.

	RESULTS

Top Abstract Introduction Materials and Methods Results Discussion References

c-Myb activates the RAG-2 promoter in cotransfected cell lines. During characterization of the BSAP binding site in the 70 to 50 region of the murine RAG-2 promoter, we generated a series of 5-bp substitution mutations across this interval (Fig. 1A). An analysis of these promoter mutations suggested that distinct DNA sequences within this small region were necessary for promoter activity in B- and T-lineage cell lines (32).

View larger version (32K): [in this window] [in a new window]   FIG. 1.   c-Myb transactivates the RAG-2 promoter through a proximal-region binding site. (A) Critical sequences in the murine RAG-2 promoter. The major transcription start site is numbered +1. The boxed regions represent sequences important for promoter activity in transiently transfected Jurkat T cells (32). The human and murine proximal-region DNA sequences are shown, with the BSAP binding site in boldface and a potential c-Myb site indicated by the thin horizontal bar above the sequence. The thick bar denotes a 20-bp region of complete identity between murine and human sequences. The sites and DNA sequences of four substitution mutations (labeled 1, 2, 3, and 4) are shown in boxes below the sequence. (B) Diagram of the transient-transfection assay vector for RAG-2 promoter activity. Each substitution mutation was analyzed in the context of the full 279 to +123 promoter. The arrow indicates the transcription start site. (C) Luciferase assay analysis of transient transfections into 2017 murine pre-T cells. The cells were cotransfected with the RAG-2 promoter-reporter construct and either empty expression vector () or c-Myb expression vector (+). Two 5' promoter deletion mutants (71/+123 and 45/+123) and four 5-nucleotide substitution mutations (labeled 1 to 4) were also tested. The results are shown as the fold increase in luciferase activity with c-Myb cotransfection compared to the activity with empty vector (the actual luminometer unit readings were 2,759 ± 434 for the wild-type promoter cotransfected with an empty expression vector and 194,886 ± 228 for the wild-type promoter cotransfected with the c-Myb expression vector). The data shown are the average (+ standard deviation) of two experiments, each performed in duplicate and adjusted for transfection efficiency using -galactosidase. (D) Luciferase assay analysis of transient transfections into 293T human embryonic kidney cells as described for panel C. The actual luminometer unit readings were 1,723 ± 96 for the wild-type promoter cotransfected with an empty expression vector and 16,573 ± 67 for the wild-type promoter cotransfected with the c-Myb expression vector.

c-Myb binds to the RAG-2 promoter in vivo. Although Jurkat cells were reported to lack V(D)J recombinase activity (33, 49), we could easily detect endogenous RAG-2 mRNA in this cell line by reverse transcription-PCR analysis (Fig. 2) (14). Since both the endogenous RAG-2 gene and the RAG-2 promoter-reporter construct were expressed in Jurkat cells, we proceeded to use this cell line to explore the role of c-Myb in regulating the RAG-2 promoter in vivo.

View larger version (44K): [in this window] [in a new window]   FIG. 2.   Jurkat T cells express a readily detectable amount of RAG-2 mRNA. Total RNA was purified from Jurkat, HeLa, and 293 cells and converted to cDNA by random-primed reverse transcription. The cDNAs were analyzed by multiplex PCR for -actin and RAG-2 transcripts. On top is a digital image of an ethidium bromide (EtBr)-stained gel, and on the bottom is a phosphorimage of the same gel blotted to a membrane and probed with an internal RAG-2-specific radiolabeled oligonucleotide.

View larger version (67K): [in this window] [in a new window]   FIG. 3.   Anti-c-myb antibody specifically enriches RAG-2 promoter DNA sequences in a (ChIP) assay. Chromatin proteins were cross-linked to DNA in intact Jurkat T cells (A) and in freshly purified mouse thymocytes (B) by formaldehyde treatment, and purified nucleoprotein complexes were fractionated using either anti-Myb antibody (lanes 1), nonspecific IgG (lanes 2), or no antibody (lanes 3). The precipitated DNA fractions were analyzed by PCR for the presence of the proximal RAG-2 promoter region, RAG-2 major exon DNA, or a region of the Ig kappa locus. In each case, the input DNA was used as a positive control (lanes 4). Amplification products were analyzed on a 2% agarose gel and visualized by ethidium bromide staining. IP, immunoprecipitate.

View larger version (100K): [in this window] [in a new window]   FIG. 4.   The RAG-2 promoter proximal region is occupied in the vicinity of the c-Myb site in vivo. Purified nuclei from Jurkat (lanes 5 and 6) and HeLa (lanes 1 and 2) cells or purified genomic DNA from Jurkat cells (lanes 3 and 4) were treated with increasing concentrations of DNase I. DNase-treated DNAs were then repurified and subjected to linker ligation and PCR amplification. Radiolabeled extension products were electrophoresed on a denaturing polyacrylamide gel alongside a DNA sequencing ladder, and the dried gel was analyzed by phosphorimaging. This ligation-mediated PCR analysis reveals DNase-sensitive sites on the top strand of the promoter through the use of a radiolabeled bottom-strand primer. The numbers indicate positions with respect to the transcription start site. The arrows point to in vivo DNase I-hypersensitive sites in Jurkat cells, and the position of the c-Myb binding site is indicated (box). The DNA sequencing ladder shows the bottom-strand sequence.

The Myb site is critical for promoter activity and required for recruitment of c-Myb to the RAG-2 promoter in vivo. To probe the functional consequences of the interaction between c-Myb and its binding site in the proximal RAG-2 promoter in a chromosomal context, we analyzed a series of Jurkat cell lines stably transfected with reporter constructs containing the murine RAG-2 promoter.

View larger version (27K): [in this window] [in a new window]   FIG. 5.   The c-Myb binding site is essential for RAG-2 promoter activity in an integrated reporter construct in Jurkat T cells. (A) Reporter construct used in the stable-transfection assay. Linearized constructs contain a GFP cDNA and the wild-type or mutant 3 RAG-2 promoter (279 to +123) in the presence or absence of the TCR  locus enhancer (E). Ins, the 1.2-kb chicken -globin insulator sequence (10), included to prevent the neomycin resistance cassette (PGK-neo) from influencing GFP activity. (B and C) Reporter constructs containing the wild-type or mutant 3 RAG-2 promoter in the presence or absence of E were transfected into Jurkat cells by electroporation. Transfected cells were selected as a pool in G418 and then analyzed by flow cytometry for GFP expression. The fluorescence-activated cell sorter histograms of untransfected (dashed line) and wild-type promoter-transfected cells (bold solid line) are coincident. The thin solid line represents transfectants containing the wild-type promoter and E, and the dotted line represents the mutant 3 promoter and E.

View larger version (59K): [in this window] [in a new window]   FIG. 6.   The c-Myb site in the proximal region is essential for in vivo recruitment of c-Myb to the RAG-2 promoter. Pools of G418-resistant cells containing the wild-type (A) or mutant 3 (B) RAG-2 promoter analyzed in Fig. 5 were subjected to the ChIP assay using anti-c-Myb antibody (Ab) (lanes 1) or nonspecific IgG (lanes 2). Mock (lanes 3) indicates samples processed in the absence of antibody, and input (lanes 4) is DNA prior to immunoprecipitation. Primers specific for either the transfected or endogenous RAG-2 promoter were used in PCR analysis of the various precipitated chromatin fractions. Images of ethidium bromide-stained gels are shown.

	DISCUSSION

Top Abstract Introduction Materials and Methods Results Discussion References

Regulated expression of RAG-1 and RAG-2 is essential for lymphocyte development. Recent transgenic studies have revealed distinct DNA sequence requirements for RAG expression in B and T cells. Reporter constructs containing as little as 2 kb of genomic DNA sequence 5' of the RAG-2 promoter are active in B but not T cells, whereas similar constructs containing 9 kb of upstream sequence are active in both lineages (39). In a separate study, DNA sequences 5' of the RAG-2 promoter were found to be involved in the coordinate expression of RAG-1 and RAG-2 in T and B cells while sequences on the RAG-1 side of the locus were only important for expression of RAG-1. Perhaps surprisingly, no regulatory elements were identified in the intergenic region (64). Our own previous studies revealed distinct requirements for RAG-2 promoter activity in B and T cells (32). In B-lineage cells, BSAP bound to a promoter-proximal DNA sequence and was critical for promoter activity. This factor is not expressed in the T lineage, however. In the present study, we found that the hematopoietic lineage transcription factor c-Myb binds to the RAG-2 promoter in T cells, where it is important for RAG-2 transcription.

We showed that a c-Myb expression vector can transactivate a RAG-2 promoter-reporter construct in a transient-transfection assay (Fig. 1C and D). When serial deletions and substitution mutations of this promoter were tested, we found that c-Myb activates luciferase reporter gene expression through a c-Myb consensus binding site completely conserved between the mouse and human promoters and located in a region critical for promoter activity in the transient-transfection assay. Mutation of a sequence adjacent to the conserved c-Myb site also interferes with RAG-2 promoter activity and its induction by c-Myb. This may indicate that c-Myb binds the RAG-2 promoter cooperatively with a second, adjacently bound factor. We went on to show that c-Myb was bound to this consensus site in vivo in Jurkat T cells by using a ChIP assay combined with in vivo DNase I footprinting. Enrichment of RAG-2 promoter sequences by c-Myb-specific antibody in the ChIP assay provides strong evidence that c-Myb is binding in the vicinity of the promoter (Fig. 3A). We obtained the same result using mouse thymocytes (Fig. 3B), which actively express RAG genes. A chromosomally integrated RAG-2 promoter construct with a mutant c-Myb site was not specifically immunoprecipitated (Fig. 6B). In a complementary set of experiments, in vivo DNase footprinting revealed hypersensitive cleavage adjacent to the c-Myb site in Jurkat cell nuclei but not in non-lymphoid-cell nuclei or in purified genomic DNA (Fig. 4). This is consistent with the c-Myb expression pattern as well as the tissue specificity of the murine RAG-2 promoter.

Using an electrophoretic mobility shift assay and Jurkat cell nuclear extract, we were unable to detect c-Myb binding to a 40-bp oligonucleotide probe spanning the RAG-2 proximal-promoter c-Myb binding site (data not shown). This may be due to the inadequacy of our in vitro binding and analysis conditions. Alternatively, it may indicate that c-Myb must cooperate with various other factors to efficiently bind its site in the RAG-2 promoter. First, as we demonstrated previously, a DNA sequence located at a distal position (between positions 156 and 107 [Fig. 1A]) is required for RAG-2 promoter activity in T-lineage cells (32). Whichever transcription factor binds to this region of the promoter may influence the binding of c-Myb to its site in the proximal region. For example, c-Myb collaborates with NF-M to induce mim-1 gene expression, which is restricted to differentiating granulocytes (43). In cooperation with core binding factor, c-Myb regulates the TCR  enhancer (22, 23). A recent report provided in vitro evidence that GATA-3 may regulate the RAG-2 promoter as a T-cell-specific factor (30). Further studies will be required to identify factors which collaborate with c-Myb in the regulation of the RAG-2 promoter in T cells.

The Myb transcription factor family. The Myb family of transcription factors include three members, A-Myb, B-Myb, and c-Myb (45). Although all three proteins share a highly conserved DNA binding domain and specifically recognize the same hexanucleotide consensus sequence, subtle differences exist in the binding preferences of different members (27, 36). Moreover, studies suggest that each member interacts with distinct collections of cellular factors and performs distinct biological functions (42, 45). A-Myb expression is restricted to male germ cells, ovaries, and germinal-center B lymphocytes and is not found at significant levels in immature hematopoietic lineages, including bone marrow cells and T cells (18). A-Myb homozygous null mutant mice show defects in spermatogenesis and female breast development, as well as growth abnormalities (59). These data lead us to believe that A-Myb is not involved in RAG expression. B-Myb is expressed in dividing cells of a wide variety of tissues, and it is a key regulator of the cell cycle (48, 52). B-Myb has been suggested to play a role in cellular differentiation (45), but its expression closely parallels cell proliferation rather than differentiation status and particular cell type (29, 47, 48). For this reason, B-Myb is less likely to be an activator of RAG expression, since RAG transcription is diminished in proliferating cells. An antagonistic relationship between c-Myb and B-Myb has been noted (12, 61). This raises the possibility that B-Myb plays a role in inhibiting RAG expression, perhaps during active periods of cell division in early lymphoid cell development. Since Jurkat cells do express RAG mRNA, a different model system would be needed to test this hypothesis. c-Myb is prevalently expressed in immature hematopoietic cells, where genetic analyses have shown that it plays an essential role in lymphocyte development (see below). In this regard, it is worth noting that the anti-c-Myb antiserum used in the ChIP experiments reported here (Fig. 3 and 6) lacks cross-reactivity with B-Myb. Thus, we conclude that c-Myb, rather than another Myb family member, is involved in regulating RAG-2 expression.

The Role of c-Myb in T-cell development. The c-Myb gene is primarily expressed in immature lymphoid, erythroid, and myeloid cells (45). This transcription factor is also present in developing airway epithelium, hair follicles, and gastrointestinal crypt epithelial cells, as well as tooth buds and the thyroid primordium during development (11). Within the lymphoid lineage, mature T and B cells can reactivate c-Myb expression following in vitro mitogenic stimulation (17). Many tumor cell lines of T- and B-lymphocyte origin also express high levels of c-Myb (4, 6, 7). Recent studies of normal developing tissue, however, suggested that c-Myb is preferentially expressed in the T, but not B, lineage (2, 11). Primary cells at various stages of development were examined by reverse transcription PCR for expression of different transcription factors (2). While pro- and pre-T cells showed similar significant amounts of c-Myb mRNA, pro- and pre-B cells totally lacked c-Myb mRNA expression. The same expression pattern was shared by GATA-3, a T-cell-specific transcription factor, but not by other factors examined, such as GATA-1 and -2, PU-1, and c/EBP. Other workers have characterized c-Myb expression using in situ hybridization (11). In the developing thymus, high levels of c-Myb RNA were observed in the cortex, the area populated with immature thymocytes expressing RAG-1 and RAG-2 and undergoing V(D)J recombination at TCR loci. When adult bone marrow was examined, a strong c-Myb signal was noted in only a small percentage of blastoid-appearing cells, which are likely precursors of erythroid and myeloid lineages.

/

/