HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Suen, T.-C. Articles by Goss, P. E. Search for Related Content PubMed PubMed Citation Articles by Suen, T.-C. Articles by Goss, P. E. Social Bookmarking                      What's this?

J Biol Chem, Vol. 274, Issue 44, 31297-31304, October 29, 1999

Transcription of BRCA1 Is Dependent on the Formation of a Specific Protein-DNA Complex on the Minimal BRCA1 Bi-directional Promoter^*

Ting-Chung Suen and Paul E. Goss^§¶

From the  Breast Cancer Prevention Program, The Toronto Hospital, Oncology Research Laboratories, Canadian Blood Services and the ^§ Breast Group, Department of Medical Oncology, Princess Margaret Hospital, Toronto, Ontario M5G 2M9, Canada

	ABSTRACT

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

BRCA1 is the first tumor suppressor gene linked to hereditary breast and ovarian cancers. Its involvement in sporadic breast cancer, however, remains unclear. Recent studies showed that a loss or lowered expression of BRCA1 is not uncommon in nonfamilial breast cancers. In addition, there have been cases of inherited BRCA1-linked breast cancer with as yet unidentified mutation. Misregulation of BRCA1 at the transcription level is a possible mechanism for loss of BRCA1 expression. To understand transcriptional regulation of the BRCA1 gene, we cloned and examined the BRCA1 promoter, by both functional reporter gene analyses and protein-DNA complex formation electrophorectic mobility shift assays. A bi-directional promoter could be located within a 229-base pair (bp) intergenic region between BRCA1 and its neighboring gene, NBR2. Deletion analyses further delineated a minimal 56-bp EcoRI-HaeIII fragment, which could drive transcription in the NBR2 gene direction 2-4-fold higher than in the BRCA1 direction in all cell lines tested. Furthermore, transcriptional activity in the BRCA1 direction was undetectable in the muscle cell line C2C12, whereas activity in the NBR2 direction was maintained. These results were consistent with the expression pattern of the respective genes. A specific protein-DNA complex was detected when nuclear extracts from HeLa cells and Caco2, a colon cell line, were incubated with the 56-bp minimal promoter. This protein binding activity was further localized to an 18-bp fragment and might involve a tissue-specific factor, because binding was not detected in the C2C12 cell line. The correlation of the detection of this protein-DNA complex only in those cell lines that expressed the chloramphenicol acetyltransferase reporter gene in the BRCA1 direction suggests a significant positive role of this complex in the transcription of the BRCA1 gene.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

BRCA1 is the first breast cancer susceptibility gene isolated (1), and mutations of this gene are found in the majority of families with an increased incidence of early onset breast and/or ovarian cancer (2-4). The frequent loss of the wild type BRCA1 allele in these families suggests its role as a tumor suppressor gene (5, 6). Tumor suppressor function has also been demonstrated by its ability to inhibit the breast cancer cell line MCF7 to induce tumors in nude mice (7). Antisense-mediated reduction of BRCA1 expression in NIH3T3 fibroblasts results in an increase in cellular proliferation and transformation (8).

BRCA1 encodes a gene product of 1863 amino acid residues (Ref. 1; for reviews, see Refs. 9 and 10). Beyond the initial controversy of its subcellular location, it is now generally accepted that BRCA1 is a 220-kDa phosphoprotein, located in the nucleus (11-13). The nuclear localization and the existence of a Ring-finger domain (1, 14) and an acidic domain support a proposed function as a transcription factor. The C-terminal acidic domain, when fused to a yeast Gal4 DNA-binding domain, is able to function as a transcriptional activator both in vivo (15, 16) and in vitro (17). BRCA1 is also a component of the RNA polymerase holoenzyme (18). Its ability to stimulate p21 expression provides direct evidence of its role as a transcription factor (19).

BRCA1 interacts with Rad51, a protein that is known to be involved in DNA repair. The co-localization of both proteins on asynapsed elements of human synaptonemal complexes during meiosis suggests a role of BRCA1 in the control of genome integrity (20). The involvement of BRCA1 in cell cycle check point control is also apparent because its expression (21) and phosphorylation (22) are induced before DNA synthesis.

Since the cloning of the BRCA1 gene, more than a hundred mutations have been found throughout the entire coding sequence (2-4, 10, 23). However, there remain cases of breast and ovarian cancer families with unknown mutations. Some of these patient samples exhibit a specific allelic loss of transcripts and are thought to have a regulatory mutation (1). One of these early reported families is now known to harbor a deletion in the promoter region (24).

In contrast to familial breast or ovarian cancer, somatic mutations of the BRCA1 gene have been reported in only a very few cases of sporadic ovarian cancer (25, 26) and none in sporadic breast cancer (2). It is hypothesized that mutations in BRCA1 might affect the process of growth regulation at an early stage of development, as is demonstrated by mouse knock-out experiments (27-29). The expression pattern of BRCA1 throughout development also suggests its importance in tissues that undergo rapid proliferation and terminal differentiation (30-32). The generation of double knock-out mice with disruptions in both p53 and BRCA1 raises the possibility that the developmental pathways of these two genes are linked (33, 34). BRCA1 has been shown to physically associate with p53 and co-activate p53-responsive genes (35, 36). Interestingly, p53 mutations seem to occur at a very high frequency in BRCA1-associated familial breast cancer (37). Taken together, these studies suggest that p53 might be an important checkpoint for BRCA1-associated tumorigenesis. It is possible that in most sporadic breast cancers, BRCA1 mutations were not being searched for because those samples might have p53 mutations or other oncogenes or tumor suppressor genes identified. On the other hand, a disruption in regulatory sequences could also be the cause for a decrease in BRCA1 expression during sporadic breast cancer progression (38). The analysis of the promoter or the identification of important regulatory sequences will therefore provide information on where to search for these regulatory mutations.

We report here the isolation and functional analyses of the human BRCA1 promoter. A 229-base pair (bp)¹ intergenic region could serve as a promoter for both the BRCA1 and its neighboring gene, NBR2. Deletion analyses delimited a minimal 56-bp fragment within the intergenic region, which retained the bi-directional promoter activity. A specific protein-DNA complex was detected with an 18-bp element within the minimal promoter. The existence of this complex only in BRCA1-expressing cells correlated with the ability of the minimal promoter to drive transcription in the BRCA1 direction. It is conceivable that mutations found within either the cis-acting element or the trans-acting transcription factor described in this report could be a target of mutagenesis in BRCA1-linked cancers with regulatory mutations or even in sporadic breast cancer where no mutation of BRCA1 has been found (2, 39).

	EXPERIMENTAL PROCEDURES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Enzymes and Reagents-- Restriction enzymes and other DNA-modifying enzymes such as T4 kinase, T4 polymerase, T4 ligase, Klenow fragment, calf intestinal phosphatase, and S1 nuclease were purchased from Life Technologies Inc., New England Biolabs (Mississauga, Ontario, Canada), Roche Molecular Biochemicals, or Amersham Pharmacia Biotech. All isotopes were products from Amersham Pharmacia Biotech. Chemicals used for the S1 nuclease protection, CAT, and -galactosidase assays were purchased from Sigma. Thin layer chromatography (TLC) plates were products of Kodak (Rochester, NY). Cell culture media and reagents were obtained from Life Technologies, Inc.

Plasmids-- The plasmid pCR2.1 (Invitrogen, Carlsbad, CA) was used for cloning PCR-amplified products. The plasmid pBluescript(IIKS) (Stratagene, La Jolla, CA) was used for general subcloning purposes. pMT.IC3 is a plasmid containing multiple cloning sites placed upstream of the chloramphenicol acetyltransferase (CAT) gene (40). Most of the BRCA1 DNA restriction fragments were cloned into pBluescript(IIKS) and were then shuffled into the matching unique restriction sites on the polylinker of pMT.IC3. DNA fragments were blunt-ended with Klenow fragment or T4 polymerase when no appropriate restriction enzymes could be used for directional cloning. In addition, reversed orientation of a subcloned fragment in the pMT.IC3 plasmid could easily be obtained by cutting with HindIII, which flank the polylinker, followed by religation. All of the CAT constructs were named according to the direction of transcription. Therefore the pBR and pNB series of CAT plasmids indicate promoters transcribing toward the BRCA1 and the NBR2 gene directions, respectively. pCMV (CLONTECH, Palo Alto, CA), a plasmid that contains the lacZ gene driven by the cytomegalovirus enhancer (41), was used for monitoring transfection efficiency. Detailed maps of all of the plasmids used in this study will be distributed along with the materials upon request.

Sequencing-- The T7 polymerase sequencing kit was purchased from Amersham Pharmacia Biotech. Dideoxy sequencing of double-stranded plasmids with [-³⁵S]dATP or [-³⁵S]dCTP (Amersham Pharmacia Biotech) was performed according to the manufacturer. Most of the BRCA1 promoter subclones, particularly those with DNA inserts of less than 300-bp in size, were confirmed by sequencing.

Oligonucleotides-- Oligonucleotides, obtained from the Hospital for Sick Children Biotechnology Center, Toronto, were as follows: PCR primer, BRCA1 + 89 (5'-GCAGAGGGTGAAGGCCTCCTG-3') and BRCA1 + 2 (5'-GCTCGCTGAGACTTCCTGGAC-3'); sequencing primer, pMTIC3 (5'-TAGGCGTATCACGAGGCCC-3') and pBluescript(IIKS) (the reverse primer or universal primer was used (Stratagene)).

Cell Culture-- Several cell lines representing different tissues of origin were used in this study and are all available from American Type Culture Collection (ATCC, Manassas, VA). This includes HeLa, a human cervical carcinoma cell line; Caco2, a human colon carcinoma cell line; MCF7, a human mammary carcinoma cell line; and C2C12, a mouse myoblast cell line. All cell lines were cultured in Dulbecco's modified Eagle's-F12 medium (Life Technologies, Inc.) supplemented with 10% fetal calf serum and kept in a humidified, 37 °C, 5% CO₂ incubator.

Cloning of the BRCA1 Genomic Sequence-- A pair of DNA primers located in exon 1 of the BRCA1 cDNA were synthesized (Fig. 1B), and PCR was performed on human lymphocyte DNA. Amplified DNA product of the expected size was gel purified and subcloned into the plasmid pCR2.1. After sequence confirmation, the exon 1 DNA was labeled as a probe to screen a phage artificial chromosome (PAC) library of human genomic DNA. Three clones were obtained, one of which, P103014, has been described by others (42). Repeated rounds of restriction mapping, subcloning, and exon 1 probing allowed the isolation of a 3.8-kb PstI fragment. Sequencing analyses of various subcloned fragments confirmed their identity to the BRCA1 genomic sequence that has been deposited in the GenBank^TM (accession no. U37574) and not to a duplicated pseudogene located at the same region (42, 43). A 2.7-kb PstI-XbaI fragment containing intron 1 and upstream sequences of the BRCA1 gene is shown schematically with respect to the genomic organization (Fig. 1A). The sequence was numbered as in U37574. The sequence of the untranslated exon 1 and the upstream region of BRCA1, including the complete intergenic region between BRCA1 and the neighboring gene NBR2 (44), is shown in Fig. 1B. The focus of this report is restricted to the intergenic region between the EcoRI and the SstI sites (shown in bold).

Mapping of Transcription Start Sites-- The S1 nuclease protection method was used to determine the transcription initiation sites (40). RNAs were isolated from different cell lines with Triazol Reagent (Life Technologies, Inc.). A 728-bp StuI-StuI fragment was dephosphorylated and labeled with [-³²P]ATP by T4 kinase. 20 µg of RNAs from various cell lines, or yeast tRNA (as a control), was co-precipitated with 50,000 cpm of probe in ethanol with the addition of lithium acetate. The precipitate was washed with 70% ethanol, dried, and resuspended in 50 µl of S1 hybridization buffer (80% deionized formamide, 40 mM PIPES, pH 6.8, 400 mM NaCl, 1 mM EDTA), after which it was submerged at 60 °C overnight. 300 µl of S1 digestion buffer (280 mM NaCl, 50 mM sodium acetate, pH4.6, 4.5 mM ZnSO₄, 20 µg/ml ssDNA) containing 200 units/ml S1 nuclease was added to the hybridized mix and incubated at 37 °C for 30 min. 50 µl of stop buffer (4 M ammonium acetate, 0.1 mM EDTA) was added, which was followed by phenol/chloroform extraction. Two volumes of ethanol was added to precipitate the DNA, which was then dried in a SpeedVac. The sample was resuspended in loading dye (48% urea, 1 mM EDTA, 0.1 M NaOH, 0.01% bromophenol blue) and run on a 7.5% denaturing gel. A sequencing ladder of a known plasmid DNA was run alongside as a size marker. The gel was dried and exposed to Kodak XAR-5 film at 80 °C.

Transfections and CAT Assays-- A calcium phosphate precipitation method (45) was used for transfection as modified and described previously (46). Briefly, cells were split at a predetermined ratio into 100-mm tissue culture dishes (Falcon) the day before transfection. Unless otherwise indicated, 1 µg of pCMV and 10 µg of a CAT reporter DNA were co-precipitated in the buffer at room temperature for 25 min, before it was added directly to the cells. Precipitate was left incubated with the cells for 16-20 h, after which the cells were washed three times with phosphate-buffered saline, re-fed with fresh medium, and returned to the 37 °C incubator. Cells were washed and harvested after 20-24 h, and the freeze/thaw cycle method was used to lyse the cells. One-fifth of the cell lysate was used for the -galactosidase assay using O-nitrophenyl--D-galactopyranoside as substrate, and the results were used to adjust the amount of lysate for the CAT assay. The TLC method of CAT assays was performed as described previously (40), except that the standard [¹⁴C]chloramphenicol was replaced with 1-deoxy[dichloro-acetyl-1-¹⁴C]chloramphenicol (Amersham Pharmacia Biotech). Because only one acetyl group could be transferred to this substrate, only one product could be seen on the TLC as opposed to three possible products that are generally observed in the literature. This also improves the quantitative aspect of the CAT assays.

Electrophoretic Mobility Shift Assays (EMSA)-- EMSA was performed as described previously (40). Nuclear extract was isolated from the different cell lines by the Dignam method (47). The DNA fragment was isolated by digesting a plasmid subclone with the appropriate restriction enzymes, gel-purified, and labeled with [-³²P]dATP or [-³²P]dCTP (depending on the restriction site) by Klenow fragment. Reaction mixture was added in the order of H₂O, 10× binding buffer (1×, 10 mM Tris, pH 7.5, 50 mM KCl, 1 mM dithiothreitol, 0.1 mM EDTA, 1 mM MgCl₂, and 5% glycerol), 3 µg of poly(dI·dC)·poly(dI·dC), 10 µg of nuclear extract, an appropriate amount of unlabeled competitor if desired, and finally, 20,000 cpm of probe. The mixture was incubated at room temperature for 25 min, after which it was loaded onto a 6% native polyacrylamide gel. The gel was dried under vacuum in a gel dryer and exposed to a Kodak BioMAX MR film at 80 °C.

	RESULTS

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Consensus Transcription Factor Binding Sites within the Minimal Promoter-- Sequence analysis of the intergenic region between BRCA1 and its neighboring gene with respect to consensus transcription factor binding sites has been discussed extensively (42, 48, 49). Because promoter function was confined to the 56-bp EcoRI-HaeIII fragment (see below), only potential transcription factor binding sites within this sequence are highlighted (Fig. 1B). A CT (or AG)-rich sequence, a possible binding site for the Ets family of transcription factors (50), was found between nucleotides 1380 and 1406. A "TTACGTCA" sequence, located between 1405 and 1412, is almost identical to the consensus cAMP-responsive element binding (CREB) site (TGACGTCA) (51). A "GGGTGG" sequence located at 1427 to 1432 is identical to the GT box, known to bind the Sp1 family of transcription factors (52) (Fig. 1B).

View larger version (49K): [in this window] [in a new window]   Fig. 1.   Genomic organization and intergenic sequences between the BRCA1 and NBR2 genes and the mapping of transcription start sites. A, top line shows the organization of introns and exons (marked by open and striped boxes for BRCA1 and NBR2 genes, respectively). Numbering of the nucleotides in this region is the same as the BRCA1 genomic sequence deposited in GenBankTM (accession no. U37574). Transcription of BRCA1 and its neighboring gene NBR2 proceeds toward the right and left sides, respectively. The approximate position of the translation start site for either gene is marked by arrows pointing in their respective directions. The position and restriction map of a 2.7-kb PstI-XbaI fragment is shown below the genomic region. A 728-bp StuI-StuI fragment is marked as a left-pointing arrow. This DNA was labeled with [-32P]ATP at the StuI site at position 1676 (marked with an asterisk, also in B) and was used as a probe for the S1 nuclease protection experiment in C. The EcoRI and SstI sites are marked in bold, and the sequence between them contains almost the entire intergenic region. B, the sequence extending from the EcoRI site to the end of BRCA1 exon 1. Restriction enzyme recognition sites are underlined. The arrowed lines above the sequence represent the PCR primers pairs used in the initial cloning of exon 1. A closed circle above the "T" at position 1579 marks the transcription start site described in Ref. 49. The closed squares above the nucleotides mark the positions of the transcription initiation sites mapped in this study (data shown in C). Circled lowercase letters underneath the sequences mark positions of discrepancy reported in Ref. 49 (see "Transcription Initiation Sites of the BRCA1 Gene" under "Results"). A CREB-binding homologous sequence is underlined by a closed bar. A potential SP1 biding site is underlined by a striped bar. C, S1 nuclease protection assay. The probe labeled at the StuI site (lane 1) (shown and marked by an asterisk in A and B) was hybridized to 20 µg of yeast tRNA (lane 2), total RNAs isolated from C2C12 (lane 3), HeLa (lane 4), Caco2 (lane 5), and MCF7 (lane 6) cells. 60 units of S1 nuclease was added to all samples except lane 1. A sequencing ladder of a plasmid with known sequence was run alongside as a size marker (lanes GATC), the numbers on the left mark the size of DNA from the end of sequencing primer. Only bands that were found at significantly higher intensity in the expressing cells (lanes 4-6), compared with those that were found in the negative control (lane 2) or non-expressing cells (lane 3), are considered as specific protected fragments (marked by arrows), and their corresponding positions are shown in B as closed squares above nucleotides.

Transcription Initiation Sites of the BRCA1 Gene-- Xu and co-workers (49) mapped the transcription start site to position 1579, and named this nucleotide +1 in their paper (marked by a closed circle above the nucleotide in Fig. 1B). However, upon careful inspection and comparison of our sequence with sequences published by the same authors (42, 48), the dinucleotide "GA" at the +1 and +2 positions (Ref. 49 and Fig. 1B, identified by circled lowercase letters under the sequences), which correspond to nucleotides 1579 and 1580, should be "TT" (Fig. 1B). In addition, a hairpin structure that was hypothesized in the paper (49) might not be possible because the "C" at position +32 (Ref. 49 and Fig. 1B, circled lowercase letter under the sequence), which corresponds to position 1610, was in fact a "T" (also reported in Refs. 42 and 48 and in our sequence (shown in Fig. 1B)). We therefore tried to map the transcription start sites by the S1 nuclease protection method (40). RNAs were isolated from various cell lines representing tissues that are known to express BRCA1 (30-32). These included HeLa (cervical carcinoma), Caco2 (colon carcinoma), and MCF7 (mammary carcinoma). A C2C12 myoblast cell line was used as a negative control because BRCA1 expression was undetectable in muscle tissue. A 728-bp DNA fragment was labeled at the StuI site (fragment marked with an asterisk in Fig. 1A), located within the untranslated exon 1 of BRCA1 (marked with an asterisk in Fig. 1B), and hybridized to either yeast tRNA or RNA samples from the above mentioned cell lines. After S1 nuclease treatment, multiple bands could be detected (Fig. 1C). Only those bands that were significantly higher in intensity in the BRCA1-expressing cells (lanes 4-6), compared with the yeast tRNA (lane 2) and muscle cells (lane 3), were considered as transcription initiation sites (marked by arrows in Fig. 1C). Based on the sequencing size marker, they were located at positions 1614, 1613, 1611, 1610, 1609, 1606, 1605, and 1604 (marked by closed squares over the corresponding nucleotides in Fig. 1B).

Localization of a Functional BRCA1 Promoter That Is Bi-directional-- To localize a functional promoter for the BRCA1 gene, various fragments within the 2.7-kb PstI-XbaI DNA were subcloned upstream of a CAT reporter plasmid; results from transfections into HeLa cells are shown in Fig. 2. The activities of constructs 1-4 suggested that sequences between the EcoRI and SstI sites were important for transcription to proceed in the BRCA1 direction, because constructs that excluded this region (constructs 3 and 4) were totally nonfunctional.

View larger version (47K): [in this window] [in a new window]   Fig. 2.   Localization of a bi-directional promoter between BRCA1 and NBR2 genes. The upper panel shows the relative positions of the different subcloned restriction fragments drawn as arrows pointing toward the right for the BRCA1 direction (constructs 1-5) and left for the NBR2 direction (constructs 6-10), respectively. The relative CAT activities of the correspondingly numbered constructs in HeLa cells are shown in the lower panel. The activities of all constructs are expressed as a relative number to that of pBR(EcoRI-SstI)CAT (construct 5), which was assigned arbitrarily as 100. The single product found in our CAT assays was because of the new chloramphenicol substrate used (see "Experimental Procedures"); this allows more accurate quantitation than the CAT assays commonly performed in the literature. Experiments were repeated three times, and a S.D. <10% was observed.

Because the intergenic region between the BRCA1 gene and its neighboring gene is only about 200 bp, it was suspected that this region could function as promoter for either gene (42, 48). Constructs 6-9 were made to test whether promoter activity in the NBR2 gene direction could be detected. As in the case of BRCA1, sequences between the EcoRI and SstI sites were critical for transcription in the NBR2 direction (constructs 6-9), because constructs that excluded this region were nonfunctional (constructs 6 and 7). A bi-directional promoter activity could in fact be found within the EcoRI-SstI fragment (constructs 5 and 10). These results demonstrated that the 229-bp EcoRI-SstI intergenic region could drive transcription as efficiently as those longer constructs in their respective directions.

Absence of an Alternative BRCA1 Promoter in the Intron-- Xu et al. (49) proposed the existence of an alternative promoter located in intron 1 of the BRCA1 gene based on their primer extension experiment. Although the above results clearly demonstrate the importance of the intergenic region, additional constructs were made to test directly whether sequences within the intron 1 (beginning at position 1702; the approximate position is marked by an asterisk in Fig. 3) did indeed contain promoter activity. Consistent with the results shown in Fig. 2, only constructs that contain the EcoRI-SstI region (constructs 1, 2, and 4) could function as promoter, whereas those that start within intron 1 (constructs 3 and 5) did not express above the background level (Fig. 3). Strong promoter activity toward the NBR2 direction was detected as with constructs shown in Fig. 2 when the EcoRI-SstI region was included (construct 6). As expected, the BRCA1 intron 1 region alone was nonfunctional in the NBR2 direction (construct 7). Our results therefore do not support the existence of an alternative promoter for BRCA1 within its intron 1, at least in those cell lines that we tested.

View larger version (52K): [in this window] [in a new window]   Fig. 3.   Absence of an alternative promoter in the intron 1 of BRCA1. The top panel shows the additional constructs used for testing if an alternative promoter could be found in intron 1 of BRCA1 as proposed (49). The beginning of intron 1 is located at nucleotide 1702 (marked with an asterisk). Constructs 1 and 2 therefore contain sequences from the intergenic region, exon 1, and part of intron 1. Construct 3 contains sequences exclusively from intron 1. Construct 4 contains sequences from the intergenic region, and part of exon 1, but excludes any sequence from intron 1. Construct 5 contains part of the exon 1 and almost the complete intron 1. Constructs 6 and 7 contain the same sequences as constructs 4 and 5, respectively, albeit in the opposite orientation. The activities of the correspondingly numbered constructs are shown in the lower panel. The activity of construct 1 was chosen as the reference and assigned as 100. Note that construct 1 (pBR(SpeI-NruI)CAT) in Figs. 2 and 3 are the same, which would allow comparison of activities among all constructs.

Delineation of a Minimal Bi-directional Promoter and Tissue Specificity-- We next sought to determine whether there were DNA elements located within the 229-bp intergenic region that might have differential effects on the bi-directional promoter. In other words, were there cis-acting elements that might affect transcription in one direction but not the other?

Smaller fragments within the EcoRI-SstI DNA were generated by restriction enzymes and were again subcloned in either orientation upstream of the CAT reporter plasmid (Fig. 4). pBR(EcoRI-SstI)CAT and pNB(SstI-EcoRI)CAT (equivalent to constructs 5 and 10 of Fig. 2, respectively) contain the 229-bp promoter-driving transcription in the direction of BRCA1 and NBR2, respectively. In HeLa cells (Fig. 4B, left panel), deletion of a 56-bp EcoRI-HaeIII fragment had a detrimental effect on promoter activity in either direction (constructs 3 and 4 for the pBR series, constructs 7 and 8 for the pNB series), reducing the activities to a level similar to that of the promoterless control (construct 1). Strikingly, the 56-bp EcoRI-HaeIII fragment itself could function as a bi-directional promoter (pBR(EcoRI-HaeIII)CAT (construct 5) and pNB(HaeIII-EcoRI)CAT) (construct 9)) as efficiently as the 229-bp promoter. Based on the expression patterns of both genes (30-32), a muscle cell line was used to test whether a tissue-specific control of transcription was maintained within this intergenic region. Analysis of the CAT activities of the same set of constructs in the C2C12 myoblast cell line (Fig. 4B, right panel) suggested that transcriptional activity in the BRCA1 direction was negligible (pBR constructs 2-5), because none of the constructs produced activity levels that were significantly over the background level (construct 1). In contrast, the expression profile of the deletion constructs directed toward the NBR2 gene (pNB constructs 6-9) was similar to that detected in HeLa cells. The differential functionality of the promoter detected in the CAT assays was consistent with the expression pattern of the respective genes, i.e. NBR2 but not BRCA1 was expressed in the muscle cells (30-32). These results suggest that information for tissue-specific control of gene expression might be found in this small 56-bp fragment.

View larger version (75K): [in this window] [in a new window]   Fig. 4.   Deletion analysis of the intergenic EcoRI-SstI bi-directional promoter. A, deletion constructs were made by subcloning the different restriction fragments within EcoRI-SstI. Construct 1 is the promoterless control. Constructs 2-5 drive transcription in the BRCA1 direction, whereas constructs 6-9 drive transcription in the NBR2 direction. B, CAT activities of the correspondingly numbered constructs shown in A, in both HeLa cells and C2C12 cells. Activity of pBR(EcoRI-SstI)CAT was assigned as 100 and used as a reference in HeLa cells. Because constructs 2-5 are nonfunctional in C2C12, the activity of pNB(SstI-EcoRI)CAT was chosen as the reference. C, functional bi-directional promoter activities in different cell lines. The activities of pBR(EcoRI-SstI)CAT (lanes 1, 3, and 5) and pNB(SstI-EcoRI)CAT (lanes 2, 4, and 6) were tested in the cell lines indicated. The activities in HeLa (lanes 1 and 2) and Caco2 (lanes 3 and 4) should be comparable because they were transfected with similar efficiency (as measured by -galactosidase assays). The activities in the MCF7 cells (lanes 5 and 6) were lower because of an approximately 40-fold lower transfection efficiency in this cell line. Transcriptional activity in the BRCA1 direction (lanes 1, 3, and 5) was consistently 3-4-fold lower than in the NBR2 direction (lanes 2, 4, and 6) for these three cell lines.

To confirm that the 56-bp EcoRI-HaeIII fragment could function as a divergently transcribed promoter, pBR(EcoRI-HaeIII)CAT and pNB(HaeIII-EcoRI)CAT were transfected into HeLa (lanes 1 and 2), Caco2 (lanes 3 and 4), and MCF7 (lanes 5 and 6) cells (Fig. 4C). Transcriptional activity was consistently 2-4-fold higher in the NBR2 direction (lanes 2, 4, and 6) compared with the BRCA1 direction (lanes 1, 3, and 5). The apparent lower activity levels of the constructs in the MCF7 breast cancer cell line (lanes 5 and 6) was because of a low transfection efficiency for this cell line (based on the results of -galactosidase assays, which showed an approximately 40-fold lower activity in MCF7 than either HeLa (lanes 1 and 2) or Caco2 cells (lanes 3 and 4); data not shown).

Protein Binding to the Minimal Promoter-- An EMSA was performed (46) to determine whether any protein would bind to the minimal promoter. Nuclear extracts were isolated from both HeLa and C2C12 cells and incubated with the ³²P-labeled EcoRI-HaeIII fragment (Fig. 5). Several bands that represent protein-DNA complexes appeared in the presence of the HeLa nuclear extract (Fig. 5A, lanes 2-7). Only one of these bands, however, was competed away in a concentration-dependent manner (its position is marked with an arrow and labeled SC for specific complex) when an increasing concentration of the same unlabeled DNA was added as competitor (Fig. 5A, lanes 3-6). The specificity of SC was further confirmed by the inability of a nonspecific (ns) DNA to compete away its formation (lane 7). When nuclear extract isolated from C2C12 muscle cells was used under the same conditions, only nonspecific binding was detected, because none of the bands detected were competed away by the same unlabeled fragment (lanes 8 and 9).

View larger version (87K): [in this window] [in a new window]   Fig. 5.   EMSAs on the minimal promoter. A, the minimal EcoRI-HaeIII promoter was labeled with 32P (lane 1) and incubated with nuclear extracts isolated from either HeLa cells (lanes 2-7), or C2C12 cells (lanes 8 and 9). Several bands appeared when HeLa nuclear extract was added (lane 2). The specificity of these bands was tested by adding increasing levels (lanes 3-6, 1-, 5-, 25-, and 125-fold, respectively) of the same unlabeled fragment (EcoRI-HaeIII). The band that was competed in a concentration-dependent manner is marked by an arrow and labeled SC (for specific complex). The specificity of SC was further confirmed by the inability of a 100-fold excess of a nonspecific DNA (lane 7) to compete for its formation. In contrast, no bands detected with C2C12 nuclear extract (lane 8) was competed away by a 100-fold excess of the unlabeled EcoRI-HaeIII fragment itself (lane 9). B, localization of protein binding to an 18-bp fragment within the minimal promoter. The labeled EcoRI-HaeIII fragment (lane 1 and 2) was cut with the restriction enzyme MspI, and the resulting DNAs, a 38-bp EcoRI-MspI (lanes 3-6) and an 18-bp MspI-HaeIII (lanes 7-10) fragment (shown schematically at the bottom) were gel-purified and subjected to EMSAs. Only HeLa nuclear extract was used in this experiment. Lanes 1 and 2 show the complex SC formed on the EcoRI-HaeIII fragment as detected in A. Lane 3 is the labeled 38-bp EcoRI-MspI fragment in the absence of nuclear extract. When nuclear extract was added (lane 4), retarded bands could be observed, and competition with a 100-fold excess of unlabeled EcoRI-MspI (EM, lane 5) or its neighboring MspI-HaeIII fragment (MH, lane 6) do not show appreciable differences. The bands detected with this DNA are considered to be nonspecific. Lane 7 shows the labeled 18-bp MspI-HaeIII fragment in the absence of nuclear extract. The free DNA has migrated out of the gel because of its small size. In the presence of nuclear extract, several retarded bands could be detected. Only the top band could be competed away with a 100-fold excess of unlabeled self (MspI-HaeIII) (MH, lane 9), but not by its neighboring EcoRI-MspI fragment (EM, lane 10). This specific band (marked with an asterisk) is likely to be SC as detected with the EcoRI-HaeIII fragment.

Localization of the Specific Binding to an 18-bp Fragment-- We then examined whether the bound protein could be localized to a smaller region, which might suggest its identity because several consensus transcription factor binding sites were observed (see Fig. 1B). The minimal promoter was digested with MspI, and the two resulting 38-bp EcoRI-MspI and 18-bp MspI-HaeIII fragments were purified and subjected to EMSA. Only the nuclear extract from HeLa cells was tested (Fig. 5B). As in the previous experiment, SC was detected with the minimal promoter (Fig. 5B, lanes 1 and 2). A single band appeared when the 38-bp EcoRI-MspI fragment (lanes 3-6) was incubated with nuclear extract (lanes 4-6). The mobility of this band (lane 4), however, suggested that it is not SC. More importantly, competition with a 100-fold excess of unlabeled self-fragment (EM, lane 5) or the neighboring 18-bp fragment (MH, lane 6) did not result in a decrease in the intensity of the band, suggesting a nonspecific complex. On the other hand, when the 18-bp MspI-HaeIII fragment (lane 7) was similarly tested, one of the complexes (marked with an asterisk, lane 8) was competed by itself (MH, lane 9) but not by its neighboring fragment (EM, lane 10). The mobility of this band (minus the size difference of the probe) also matched that of the SC detected with the minimal promoter. We therefore concluded that the MspI-HaeIII fragment could be recognized by a protein complex in the nucleus of the HeLa cell.

Tissue Specificity of the Specific Complex-- Because a GT box was observed within the MspI-HaeIII fragment (Fig. 1C), we tested whether SC was formed by Sp1 or a Sp1 family member (52). We first confirmed the specificity of SC by performing a concentration-dependent competition. As shown in Fig. 6, the intensity of the complex (position marked by an asterisk) continued to decrease with a successive 5-fold increase of unlabeled self-fragment (MH, lanes 3-6). The highest level of the unlabeled adjacent EcoRI-MspI fragment did not compete (EM, lane 7) for the complex. A thymidine kinase promoter fragment, which contains two Sp1 binding sites, was also unable to compete for the formation of this complex (TK, lane 8), suggesting that it was not Sp1; this was further confirmed by the inability of purified Sp1 protein to form a similarly migrating complex (lane 9). Because our CAT assay data (Fig. 4C) demonstrated that the minimal promoter was also functional in Caco2 cells, a colon carcinoma line, we tested whether SC could also be detected with Caco2 nuclear extract. Similar to the minimal promoter (Fig. 5A), the complex was not detected in the muscle cell line C2C12 (lane 10) but was observed with much stronger intensity in the Caco2 cells (lane 11). The correlation of its absence with the lack of transcriptional activity only in the BRCA1 direction in C2C12 cells suggested that this protein-DNA complex might play a positive regulatory role in BRCA1 transcription.

View larger version (113K): [in this window] [in a new window]   Fig. 6.   The specific protein (marked with an asterisk) that binds to MspI-HaeIII fragment is not Sp1 and is found only in cell lines that express BRCA1. The 18-bp MspI-HaeIII fragment was labeled in the absence of nuclear extract (lane 1). HeLa nuclear extract was added (lanes 2-8) without cold competitor (lane 2) and 1-, 5-, 25-, and 125-fold of unlabeled MspI-HaeIII fragment (MH, lanes 3-6, respectively). The neighboring EcoRI-MspI fragment (EM, lane 7) and a promoter fragment from the thymidine kinase promoter (TK, lane 8), which contains two functional Sp1 binding sites, were also used as cold competitor. Commercially available purified Sp1 protein (Sp1, lane 9), nuclear extract isolated from C2C12 (C2, lane 10), or Caco2 (Ca, lane 11) were also tested for binding activity. The position of the SC is marked with an asterisk.

The 18-bp MspI-HaeIII Fragment Is Necessary but Not Sufficient for Transcriptional Activity-- To demonstrate the functional significance of the SC-binding MspI-HaeIII fragment, further deletions of the minimal promoter were generated (Fig. 7A). A deletion from the 3'-end, thus removing the 18-bp fragment, rendered the promoter nonfunctional (compare constructs 1 and 2). However, a similar detrimental effect on the promoter activity was also detected when DNA was deleted from the 5'-end (construct 3). The 18-bp MspI-HaeIII fragment itself was also unable to generate detectable activity (construct 4) over that of the promoterless control (construct 5). The 18-bp protein recognition sequence was therefore necessary but not sufficient for promoter activity. Interestingly, a construct that fused the EcoRI-Alw26I fragment downstream to the Alw26I-HaeIII fragment (Fig. 7B, construct 2) and therefore contained all of the sequence as the minimal promoter (construct 1), but in a rearranged manner, was also highly impaired in transcription. These data firmly established that the 56-bp EcoRI-HaeIII fragment is indeed a minimal promoter. Moreover, the organization of the DNA sequence within the minimal promoter was also important for transcription to occur.

View larger version (27K): [in this window] [in a new window]   Fig. 7.   Disruption of minimal promoter activity by deletion or rearrangement. A, the 56-bp minimal promoter is drawn schematically (construct 1). A gray box represents a TC-rich sequence that may bind the Ets family of transcription factors. Closed box represents the cAMP-responsive element-like sequences. Striped box marks the potential binding sites of the Sp1 family of transcription factors. Fragments were generated by restriction enzymes with the positions of their recognition sites shown. Construct 1, CAT activity of the 56-bp EcoRI-HaeIII minimal promoter; construct 2, 3'-deletion of MspI-HaeIII fragment; construct 3, 5'-deletion of EcoRI-Alw26I sequence; construct 4, only the 18-bp MspI-HaeIII fragment is retained; construct 5, promoterless control. B, rearrangement of the minimal promoter (construct 1) by placing the EcoRI-Alw26I fragment downstream of the Alw26I-HaeIII fragment (construct 2).

	DISCUSSION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

The sequence of the BRCA1 promoter was first described by Brown et al. (48) and a distinct BRCA1 promoter was proposed based on the mapping of transcription initiation sites (49). It is puzzling that these reports by the same authors (42, 48, 49) showed a discrepancy in the sequence located at critical positions (Fig. 1C, and "Transcription Initiation Sites of the BRCA1 Gene" under "Results"). We therefore tried to map the transcription start sites using an S1 nuclease protection method (40). Instead of a single transcription start site as reported (49), we observed a cluster of initiation sites located between positions 1604 and 1614 (Fig. 1). The different results that we obtained could be due to the different methods used (RNA protection against nuclease in this study versus reverse transcription-PCR assay). As the intergenic region of BRCA1 does not contain a consensus TATA box, it was not surprising to detect multiple transcription initiation sites, commonly observed among this class of promoters (Ref. 40 and references therein).

Various fragments from the BRCA1 region were subcloned into a CAT reporter gene plasmid, and transfections into different cell lines allowed us to map a very critical region within the intergenic region with bi-directional promoter activity (Fig. 2).

During the progress of our study, Xu et al. (53) reported the activity of a BRCA1 promoter that they named  promoter, as opposed to the  promoter located in intron 1 of the BRCA1 gene, which was proposed originally in their study on transcription start sites (49). Our data on the EcoRI-SstI fragment is comparable with the data on these authors'  promoter. However, we were not able to confirm the existence of the  promoter, because all constructs that started downstream of the SstI site (pBR(SstI-XbaI)CAT in Fig. 2 pBR(NruI-AvaII)CAT and pBR(StuI-StuI)1CAT in Fig. 3, and data not shown) and therefore exclude the  promoter, did not yield significant activity over the promoterless control. One possible explanation for the differences between their results and ours was that different reporter genes were used. The luciferase assays used in their study are much more sensitive than the CAT assays that we used. In this sense, their two constructs that contained only sequences from the  promoter were only 2-fold more active than their promoterless control. We think it is difficult to interpret or conclude what an activity level of 2-fold above the background might mean in a highly sensitive assay. In addition, the higher activity of a construct that contained sequences from both their  and  promoter (similar to pBR(SpeI-NruI)CAT and pBR(AvaII-NruI)CAT in Fig. 3), was interpreted as increasing the activity of the  promoter by 12-fold. However, this supposed 12-fold increase could be interpreted simply as activity from the  promoter, which by itself (constructs that excluded the  promoter totally in their study, and similar to pBR(PstI-SstI)CAT and pBR(EcoRI-SstI)CAT in Fig. 2 and pBR(StuI-StuI)0CAT in Fig. 3) was much higher than the activity of the putative  promoter alone (pBR(StuI-StuI)1CAT in Fig. 3). Our view of the insignificant functional role of this alternative BRCA1 promoter was recently confirmed by others (54).

Our present study focused on important regulatory regions within the bi-directional promoter. Thus, the effects of deletions on the EcoRI-SstI fragment were examined (Fig. 4). It was striking that the 56-bp DNA EcoRI-HaeIII fragment was able to drive transcription divergently (Fig. 4, constructs 5 and 9). However, given the small intergenic region between the two neighboring genes (229 bp), it might be imperative to utilize the DNA efficiently. Given that NBR2 is expressed in most tissues (44), whereas BRCA1 is not expressed in muscle tissue (30-32), the C2C12 cell line was utilized to determine whether the 56-bp promoter still functioned in concordance with the expression pattern of the respective endogenous genes. The CAT reporter analyses in C2C12 cells indeed demonstrated promoter activity only in the direction of the NBR2 gene (Fig. 4A). Although C2C12 is a mouse cell line, the possibility of a species-specific effect was unlikely, because the promoter constructs were also functional in a mouse mammary gland cell line HC11 (data not shown). The results from the functional CAT assays and protein-DNA-binding EMSAs with the EcoRI-HaeIII fragment correlated in terms of a specific complex being detectable only in a BRCA1-expressing cell line (Fig. 5A, HeLa against C2C12). Sequence analysis of the minimal promoter (Fig. 1B) suggested the presence of potential binding sites for the Ets (50), CREB (51), and Sp1 (52) families of transcription factors. EMSAs with shorter restriction fragments within the minimal promoter localized the specific protein complex to an 18-bp MspI-HaeIII fragment (Fig. 5B). This fragment contains a GT box, which is known to interact with the Sp1 family of transcription factors (52). However, the lack of competition with a promoter containing functional Sp1 binding sites (a thymidine kinase promoter) and the failure of purified Sp1 protein to complex with the 18-bp fragment (Fig. 6) both argued against Sp1 involvement. Because BRCA1 is known to increase the risk of colon cancer (55), a colon carcinoma cell line was also tested in both the CAT assays (Fig. 4C) and the EMSAs (Fig. 6). Similar correlation of protein binding and transcriptional activity was observed as in the case of HeLa cells. The combined data suggest that we have identified a very important regulatory element and an interacting protein (SC in Figs. 5 and 6), which most likely functions as a transcriptional activator of the BRCA1 gene. Its absence did not seem to affect expression in the opposite direction (i.e. transcription of the NBR2 gene) in C2C12 cells.

An alternative mechanism that might affect BRCA1 gene expression is the methylation pattern of the promoter (54, 56-58). Mancini et al. (57) postulated methylation of the CREB site as a potential mechanism affecting BRCA1 transcriptional regulation. When we first identified the minimal promoter, the CREB homologous binding site was a particularly attractive candidate as an important transcriptional control element. We transfected two CAT constructs (pBR(SpeI-SstI)CAT and pBR(EcoRI-SstI)CAT, both of which included the potential CREB binding site) into HeLa cells, followed by treatment with forskolin, an activator of the cAMP pathway. No stimulation of CAT activity was observed with either construct (data not shown). We therefore concluded that this CREB-like binding site on the BRCA1 promoter does not function as a CREB binding site. This conclusion was further supported by our EMSA study, in which no specific formation of protein-DNA complex was detected with the fragment that contained the CREB-like site (Fig. 5B, EcoRI-MspI fragment). Upon close inspection of the oligonucleotides that were used in the EMSAs performed by Mancini et al. (57), there was a 12-bp overlap with our 18-bp (MspI-HaeIII) binding sequence. In addition to the "C" located within the CREB-like site, another "C" located 7-bp downstream and therefore within the MspI-HaeIII fragment was also methylated (57). Therefore, the binding that these authors detected that was affected by methylation might in fact not be related to CREB protein but might be the SC protein that we have detected.

Although protein binding could be found on the MspI-HaeIII fragment, by itself this was not enough to drive transcription in either direction (Fig. 7A and data not shown). The sequence within the EcoRI-MspI fragment must serve some important function for transcription to take place. A construct that could be considered as a rearrangement of the EcoRI-HaeIII fragment (Fig. 7B) was also highly defective in transcription. These data suggest that not only must all of the sequence information be present for the minimal promoter to be functional, but it must be in an ordered alignment, perhaps allowing the binding of SC to coordinate events that lead to BRCA1 transcription.

In summary, our data have defined a 56-bp minimal promoter for the BRCA1 gene. This 56-bp minimal promoter is bi-directional and contains tissue-specific transcriptional activity. Further deletions from either end or rearrangement within this minimal promoter is deleterious to its function. This region therefore represents a potential site where mutations could occur that would lead to total loss of BRCA1 expression, and it may provide an explanation for the originally described "regulatory" mutations (1). The specific protein-DNA complex bound to the minimal promoter is likely to be a positive regulatory factor for BRCA1 transcription. More and more recent studies have shown a defective expression of BRCA1 in sporadic breast cancers (13, 58, 59). Potentially, a mutation in transcription factors that regulate BRCA1 expression, such as the SC we have described in this paper, could result in the promotion of breast cancer development, although a search for mutations in the BRCA1 genomic region would definitely be uninformative.

	ACKNOWLEDGEMENTS

We thank Drs. Lesley Tye, Eldad Zacksenhaus, and Robert Hawley for critical reading of the manuscript.

	FOOTNOTES

^* This work was supported by a grant from the Breast Cancer Prevention Program of The Toronto Hospital (to P. E. G.).The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

^¶ To whom correspondence should be addressed: Head, Breast Group, Dept. of Medical Oncology, Princess Margaret Hospital, Rm. 5-303, 610 University Ave., Toronto, Ontario M5G 2M9, Canada. Tel.: 416-946-4501, Ext. 5103; Fax: 416-946-2983; E-mail: pegoss@interlog.com.

	ABBREVIATIONS

The abbreviations used are: bp, base pair(s); PCR, polymerase chain reaction; CAT, chloramphenicol acetyltransferase; kb, kilobase (pair); PIPES, 1,4-piperazinediethanesulfonic acid; EMSA, electrophorectic mobility shift assay; CREB site, cAMP-responsive element binding site; SC, specific complex.

	REFERENCES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES