This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Services 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 Shekhar, M. P.V. Articles by Gerard, B. Search for Related Content PubMed PubMed Citation Articles by Shekhar, M. P.V. Articles by Gerard, B.

---

DNA Damage and Cellular Stress Responses

Essential Role of T-Cell Factor/ß-Catenin in Regulation of Rad6B: A Potential Mechanism for Rad6B Overexpression in Breast Cancer Cells

¹ Breast Cancer Program, Karmanos Cancer Institute and ² Department of Pathology, Wayne State University School of Medicine, Detroit, Michigan

Requests for reprints: Malathy P.V. Shekhar, Breast Cancer Program, Karmanos Cancer Institute, 110 East Warren Avenue, Detroit, MI 48201. Phone: 313-833-0715, ext. 2326; Fax: 313-831-7518. E-mail: shekharm{at}karmanos.org

	Abstract

Top Notes Abstract Introduction Results Discussion Materials and Methods Acknowledgements References

 
We have previously shown that the postreplication DNA repair gene Rad6B plays a critical role in the maintenance of genomic integrity of human breast cells. Whereas normal breast cells express low levels of Rad6B, increases in Rad6B expression occur in hyperplasia with overexpression in breast carcinomas. Here, we show that the human Rad6B gene is a transcriptional target of T-cell factor (TCF)-4/ß-catenin/p300. Rad6B promoter activity is subject to negative regulation in normal human MCF10A breast cells whereas it is constitutively active in metastatic MDA-MB-231 breast cancer cells. Derepression and activation of Rad6B promoter in MCF10A cells require coexpression of ß-catenin and p300. Using electrophoresis mobility shift assay, Western blot analysis of electrophoresis mobility shift assay, UV cross-linking, and chromatin immunoprecipitation assay, we show that Rad6B transcriptional repression in MCF10A cells is due to paucity of transcriptionally active ß-catenin assembled on the TCF binding sequence in the Rad6B promoter rather than to a deficit/decreased affinity of TCF-4 for the TCF binding element in Rad6B promoter. Three-dimensional epithelial acini generated in vitro from MCF10A cells cotransfected with ß-catenin and p300 showed ß-catenin expression on the membrane, cytoplasm, and/or nuclei with concomitant Rad6 overexpression, whereas control acini showed ß-catenin on the membranes and negligible Rad6 expression. Immunohistochemical analysis of 12 breast carcinomas showed an 80% correlation between Rad6 and ß-catenin expression, and combined nuclear and cytoplasmic staining of ß-catenin and Rad6 was detected in 25% of the breast carcinomas. In vivo implantation of MCF10A-Rad6B cells produced hyperplastic lesions. These data reveal a potentially important role for transcriptionally active ß-catenin in the regulation of Rad6B gene expression, and link aberrant ß-catenin signaling with transcriptional deregulation of Rad6B and breast cancer development. (Mol Cancer Res 2006;4(10):729–45)

	Introduction

Top Notes Abstract Introduction Results Discussion Materials and Methods Acknowledgements References

 
The Rad6 group is involved in postreplication or "error-prone" repair of DNA (1). The Rad6 gene encodes a 17-kDa protein (2) and belongs to a group of ubiquitin conjugating enzymes (3) that covalently add ubiquitin to selected lysine residues and commit them to rapid proteolysis. The Rad6 gene of Saccharomyces cerevisiae is required for a variety of cellular functions including DNA repair, induced mutagenesis, and sporulation (4, 5). Rad6-mutant phenotypic effects include slow growth, severe defects in induced mutagenesis, and hypersensitivity to UV, X-ray, chemical mutagens, and antifolate drug metabolites (6). All of the functions by Rad6 protein seem to result from ubiquitination because replacement of the conserved Cys⁸⁸ with Ser produces a totally null phenotype (7, 8).

Rad6 is highly conserved among eukaryotes. Two closely related human homologues of yeast Rad6, HHR6A and HHR6B (referred as Rad6A and Rad6B), encode ubiquitin conjugating enzymes and complement the DNA repair and UV mutagenesis defects of the S. cerevisiae rad6 mutant (9, 10), but not the sporulation defect, which seems to be dependent on the acidic tail that is found only in the yeast Rad6 (10, 11). Rad6A and Rad6B share 95% identical amino acid identity and are localized on human chromosomes Xq24-q25 and 5q23-q31, respectively (12). Inactivation of the gene encoding Rad6B in mice leads to male sterility (13), whereas male mice lacking Rad6A are fertile (14). Rad6A null female mice fail to produce offspring despite normal ovarian histology and ovulation. The requirement for at least one functional Rad6A or Rad6B allele in all somatic cell types is supported by the fact that male and female mice lacking both homologues and female mice with only one intact Rad6A allele are nonviable (14). Rad6 is associated with centrosomes throughout the cell cycle, implicating its potential role in maintaining fidelity at various phases of cell cycle (15, 16). In this regard, it is interesting to note that constitutive overexpression of Rad6B in normal human breast cells induces centrosome amplification with consequent increase in abnormal mitotic spindle formation, generation of multinucleated cells, aneuploidy, and acquisition of ability for anchorage-independent growth (16). We have previously shown that Rad6B expression level is low in normal human breast tissues; however, detectable increases in Rad6B expression are observed in hyperplastic breast lesions and considerable overexpression is observed in breast carcinomas (16). These data suggest that induction of Rad6B expression occurs early during breast carcinogenesis. However, there is little information on mechanisms regulating Rad6B gene expression in normal and cancerous breast cells.

The canonical Wnt signaling transduction pathway is involved in many developmental processes as well as in tumor development (17-21). These effects are achieved via activation of the canonical Wnt signaling pathway of downstream target genes by the T-cell factor (TCF)/lymphoid enhancer factor (LEF) family of transcription factors and their coactivator, ß-catenin (22). In the absence of Wnt signaling, a multiprotein complex, composed of axin, glycogen synthase kinase 3ß, and tumor suppressor adenomatous polyposis coli, phosphorylates ß-catenin, leading to its ubiquitin-proteasome–mediated degradation. Wnt signaling negatively regulates the function of this complex and stabilizes ß-catenin, which is thereby able to translocate to the nucleus, form a complex with TCF, and induce expression of a variety of TCF target genes (23, 24). TCF/ß-catenin–mediated signaling regulates a diverse array of genes that are responsible for cell proliferation, differentiation, migration, and polarity (25). A number of TCF target genes have been identified to date, and these include genes important for development or tumorigenesis. A list of known Wnt/ß-catenin–regulated target genes is available online.³ All these genes contain one or more TCF binding elements in their promoter region near the transcription start site. TCF binding sequences are composed of a highly conserved consensus sequence, 5'-CTTTG[A/T][A/T]-3', which is recognized by the TCF and LEF family of transcription factors (26, 27). Mutations in ß-catenin or other Wnt pathway components, which result in ß-catenin accumulation, are found in a wide range of human cancers. In contrast, such mutations have been found only rarely in breast cancer. However, there is strong evidence of stabilization of ß-catenin protein in a majority of human breast tumors, implying the existence of an active canonical Wnt signaling pathway in breast carcinomas (28), and studies in mouse model systems show that activated Wnt signaling leads to mammary tumorigenesis (29, 30).

In this study, we show that Rad6B gene expression is subject to negative regulation in normal human MCF10A breast cells whereas it is constitutively active in metastatic MDA-MB-231 breast cancer cells. Derepression and activation of the Rad6B promoter in MCF10A cells require coexpression of ß-catenin and the coactivator p300. Using electrophoresis mobility shift assay (EMSA), UV cross-linking, and chromatin immunoprecipitation assay, we show that Rad6B transcriptional repression in normal breast cells is due to a paucity of ß-catenin/p300 activation complexes assembled on the TCF binding sequence in the Rad6B promoter rather than to a deficit or decreased affinity of TCF-4 for the TCF binding element in the Rad6B promoter. Coexpression of ß-catenin and p300 elicits an inductory effect on Rad6B expression in normal MCF10A breast cells. MCF10A cells engineered to constitutively overexpress Rad6B form persistent hyperplastic lesions when xenografted in vivo in immunodeficient mice. Immunohistochemical analysis of a small number of breast carcinomas showed a direct correlation between Rad6 and ß-catenin expression. Our findings link aberrant ß-catenin signaling, a mitogenic pathway, with transcriptional deregulation of Rad6B and implicate Rad6B as an early contributor to breast carcinogenesis.

	Results

Top Notes Abstract Introduction Results Discussion Materials and Methods Acknowledgements References

 
Overexpression of ß-Catenin and p300 Is Required for Derepression of Rad6B Gene Expression
Promoter prediction analysis identified a putative promoter spanning bases –592 to –10 relative to the first ATG codon in the human Rad6B gene (Fig. 1 ). To characterize sequences in the Rad6B gene that possess promoter activity, a series of 5' deletions were made and subcloned into promoterless pGL3 reporter vector (Fig. 1A and B). Reporter constructs were transiently transfected into normal MCF10A and metastatic MDA-MB-231 breast cells. The full-length 5'-Rad6B fragment containing the putative promoter (–722/+9) conferred minimal ability (4.0-fold higher levels as compared with empty pGL3 vector) to transactivate luciferase reporter gene in MCF10A cells (Fig. 1C). In contrast, similar transfection into MDA-MB-231 cells resulted in 80-fold higher levels of luciferase gene expression as compared with empty vector (Fig. 1C). 5' deletion resulted in progressive increases in promoter activity in MCF10A cells; removal of sequences 5' of –325 (–325/+9 construct; Fig. 1C) resulted in a dramatic induction of luciferase gene expression, which was 60-fold higher as compared with the intact Rad6B gene fragment (Fig. 1C). Similar deletion of sequences 5' of –325 resulted in 2.5-fold increase in luciferase gene expression relative to the full-length Rad6B gene fragment in MDA-MB-231 cells (Fig. 1C). Further removal of sequences resulted in a complete loss of promoter activity in both MCF10A and MDA-MB-31 cells (Fig. 1C). Such mapping identified a region between nucleotides –325 and –208 that conferred maximal transcriptional induction of luciferase gene expression and, more importantly, the presence of negative regulatory elements located between nucleotides –712 and –325. Because deletion of sequences 5' of –325 resulted in a dramatic induction (60-fold) of promoter activity in MCF10A cells, and a modest increase (<3-fold) in MDA-MB-231 cells, our results suggest that Rad6B gene expression is regulated by similar cis elements in both cell lines but it is under significant negative regulation in normal MCF10A cells as compared with metastatic MDA-MB-231 cells. This difference in Rad6B promoter activity is consistent with decreased levels of Rad6B expression in normal breast cells and its up-regulation in breast cancer cells and tissues (16).

View larger version (16K): [in this window] [in a new window]   FIGURE 1. Rad6B promoter is subject to negative regulation. A. Schematic representation of 5' untranslated region of Rad6B gene. The positions of Sp1 (•) and TCF at –341 (*) relative to the ATG translation start site (+1) are indicated. B. Schematic representation of full-length Rad6B promoter (–712) and its various deletion fragments (Rad6B –550, –401, –325, and –208) subcloned into the promoterless pGL3-Basic vector. C. The activity of the full-length and various deletion fragments of Rad6B promoter was quantified in transactivation assays in MCF10A or MDA-MB-231 cells. Relative activity levels were normalized to Renilla luciferase activity and corrected for the empty pGL3-Basic vector. Columns, mean of at least six independent experiments; bars, SD.

View larger version (35K): [in this window] [in a new window]   FIGURE 2. Identification of the TCF binding sequence in the Rad6B promoter. The activity of the full-length Rad6B promoter –712 carrying wild-type TCF binding site (CTTTGAA) and site-directed mutation of the TCF site (GGGGGAA) at –341 (A) or the Rad6B deletion construct –401 containing wild-type or the mutated (GGGGGAA) TCF site at –341 (B) was quantified in transactivation assays in MCF10A cells. ß-Catenin or p300, 0.25 (1x) or 0.5 µg (2x) each, or a combination of p300 and ß-catenin, or their corresponding control plasmids were cotransfected with either full-length (–712/+9 wild-type or mutant TCF) or deletion (–401/+9 wild-type or mutant TCF) Rad6B promoter constructs. DNA concentrations were kept constant with empty vector DNA. Columns, mean of at least six independent experiments; bars, SD. C. The activity of the full-length Rad6B promoter (–712) construct in the presence of expression vectors for wild-type TCF-4, N-TCF-4, p300 + ß-catenin, or p300 + ß-catenin with wild-type TCF-4 or N-TCF-4, or the corresponding control plasmids was measured in MCF10A cells. Columns, mean of four independent experiments; bars, SD. D. Steady-state levels of ß-catenin in MCF10A, MDA-MB-231, and SW480 total cell lysates relative to ß-actin. E. The transcriptional activity of endogenous ß-catenin in MCF10A cells was assessed by transfection of TOP/FLASH or FOP/FLASH reporter vector. Similar transfections were done in SW480 colon carcinoma cells that served as positive control. Activity was normalized against Renilla luciferase activity. Columns, mean of four independent experiments; bars, SD. F. The effects of ß-catenin, ß-catenin + wild-type TCF-4, or ß-catenin + N-TCF-4, or the corresponding control plasmids on the activity of TOP/FLASH vector was measured in MCF10A cells. Columns, mean of four independent experiments; bars, SD. G. The activity of the full-length Rad6B promoter –712 carrying wild-type TCF binding site (CTTTGAA) and site-directed mutation of the TCF site (GGGGGAA) at –341 was tested in MDA-MB-231 cells in the presence of N-TCF-4, wild-type TCF-4, ß-catenin/p300, N-TCF-4 plus ß-catenin/p300, or wild-type TCF-4 plus ß-catenin/p300 and compared with those transfected with corresponding empty vectors. Activity was normalized against Renilla luciferase activity. Columns, mean of three independent experiments; bars, SD.

Analysis of functionality of the Rad6B TCF binding site in metastatic MDA-MB-231 cells showed that the full-length –712/+9 Rad6B promoter is induced 2.0-fold by ectopic expression of ß-catenin and p300 (Fig. 2G) as compared with 18-fold in MCF10A cells (Fig. 2A). Inclusion of N-TCF-4 caused 66% inhibition of basal activity of the full-length Rad6B promoter, and this inhibition was sustained in the presence of exogenous ß-catenin/p300. Coexpression of wild-type TCF-4 did not significantly influence ß-catenin/p300–induced reporter gene expression (Fig. 2G). Transfection of the TCF mutant –712/+9 Rad6B promoter in MDA-MB-231 cells resulted in >2-fold increase in basal activity of the intact Rad6B promoter fragment, which is uninfluenced by ectopic expression of N-TCF-4, ß-catenin/p300, or wild-type TCF-4 (Fig. 2G). These data suggest that the Rad6B promoter activity in MDA-MB-231 cells is also subject to similar regulation by TCF/ß-catenin/p300 as in MCF10A cells; however, unlike in MCF10A cells, the Rad6B promoter is largely derepressed in MDA-MB-231 cells as observed by very small increases in basal reporter activity by exogenous ß-catenin/p300 and greater inhibition by N-TCF-4 (Fig. 2F).

Activation of the Rad6B Promoter by ß-Catenin/p300 Results in Elevated Levels of Rad6B
Because results of promoter assays implicated an important functional role for TCF/ß-catenin/p300 on Rad6B gene expression, we assessed the effects of ectopic ß-catenin and p300 expression on Rad6B gene expression in MCF10A cells. Rad6B mRNA levels detected by semiquantitative reverse transcription-PCR (RT-PCR) analysis showed 2.8-fold higher levels of Rad6B mRNA levels in ß-catenin/p300–transfected MCF10A cells as compared with corresponding vector control. Inclusion of N-TCF4 inhibited ß-catenin/p300–induced increase in Rad6B mRNA (Fig. 3A ). Higher levels of Rad6B mRNA induction are possibly not detected because Rad6B transcripts have a short half-life, consequently failing to accumulate appreciable steady-state levels; however, the translated Rad6B protein is stable (15). Western blot analysis of Rad6B protein levels showed 3- to 5-fold increase in MCF10A cells transfected with a combination of ß-catenin and p300 as compared with corresponding vector control. Contrary to the results from promoter assays (Fig. 2), transfections with either the p300 or ß-catenin expression vector did not significantly induce Rad6B protein expression (Fig. 3B).

View larger version (62K): [in this window] [in a new window]   FIGURE 3. Levels of Rad6B mRNA (A) and protein (B) were measured by semiquantitative RT-PCR or Western blot analysis from total RNA or cell lysates, respectively, prepared from MCF10A cells transfected in parallel with Fig. 2C. B. Because the antibody may not distinguish the two forms of very closely related Rad6A and Rad6B proteins, immunoreactive Rad6 proteins detected with this antibody are referred to as Rad6 rather than Rad6A or Rad6B. C. MCF10A cells transiently transfected with empty vectors (a-c) or ß-catenin/p300 (d-f) were seeded on coverslips to generate monolayers (a and d) or suspended in Matrigel to generate epithelial acini (b, c, e, and f). Acini generated from MCF10A cells transfected with ß-catenin/p300 (e and f) or the corresponding control plasmids (b and c) were fixed in buffered formalin, paraffin embedded, and sections stained with antibodies to ß-catenin (b and e) or Rad6 (c and f). p300 expression/distribution is not shown because the p300 antibody was not suitable for immunostaining. Note that immunofluorescence staining of corresponding monolayers revealed intense ß-catenin immunoreactivity on the cell membranes, ruffling of cell membranes, and perinuclear (short arrow; d) or nuclear (long arrow; d) staining of ß-catenin/p300–transfected MCF10A cells (d).

TCF/ß-Catenin Complex Formation with Rad6B TCF Binding Sequence Is Detected in MDA-MB-231 Cells but not in MCF10A Cells
To determine whether the TCF binding element identified by promoter assays can indeed bind to TCF/ß-catenin heterodimer, we conducted EMSA using the wild-type TCF binding site of the Rad6B promoter and a corresponding sequence containing mutant TCF binding site. Figure 4 shows that nuclear proteins from MCF10A and MDA-MB-231 cells exhibit different patterns of complex formation when incubated with the TCF binding sequence of Rad6B promoter. Specific complexes B to D formed with MCF10A nuclear proteins (Fig. 4A, lane 1) migrate faster as compared with specific complexes 1 to 4 formed by MDA-MB-231 nuclear proteins (Fig. 4A, lane 4). The complexes formed are specific for the wild-type TCF binding sequence as the corresponding double-stranded oligonucleotides with mutation in the TCF binding site are unable to compete (Fig. 4A, lanes 2 and 7; Fig. 4B, lanes 3 and 11), but are readily competed by a 10-fold (Fig. 4A, lanes 3 and 5) or 50-fold (Fig. 4A, lane 6; Fig. 4B, lane 2) molar excess of the unlabeled wild-type probe. In contrast, the slow-migrating complex A formed from MCF10A nuclear proteins is competed less efficiently by the unlabeled wild-type competitor because addition of 50-fold molar excess of the wild-type competitor caused only 50% decrease in complex A but inhibited complexes B to D by >90% (Fig. 4B, compare lanes 1 and 2). Addition of ß-catenin antibody to binding reactions selectively inhibited the formation of all four specific complexes formed from MDA-MB-231 nuclear proteins (Fig. 4B, lane 7), whereas it had no effect on complexes formed from MCF10A nuclear extracts (Fig. 4B, lane 4). Inclusion of TCF-4 antibody inhibited formation of complex 1 in MDA-MB-231 cells (Fig. 4B, lane 13) but had no effect on complexes 3 and 4 (Fig. 4B, lane 13). Because complex 2 is more diffuse, we are unable to confirm the effect of TCF-4 antibody on complex 2. Incubation with preimmune immunoglobulin G (IgG) did not influence formation or stability of the complexes (Fig. 4B, lanes 5, 8, 14, and 16). These results suggest that complexes 1 to 4 formed from MDA-MB-231 nuclear proteins contain ß-catenin whereas TCF-4 is present mainly in complex 1 and is not part of complexes 3 and 4. Thus, other complexes containing ß-catenin probably contain other isoforms of TCF. Addition of TCF-4 antibody to MCF10A binding reactions had no effect on complex A formation, weakly affected complexes C and D, and inhibited complex B formation (Fig. 4, lane 17). The results from EMSAs are consistent with our observation that MCF10A cells possess weak Rad6B promoter activity due to deficit of transcriptionally active ß-catenin assembled on the Rad6B TCF binding sequence despite the presence of significant amounts of ß-catenin in MCF10A whole-cell extracts (Fig. 2D).

View larger version (24K): [in this window] [in a new window]   FIGURE 4. In vitro binding activity of MCF10A and MDA-MB-231 nuclear proteins to human Rad6B promoter. Duplex oligonucleotides containing the putative (–341) TCF binding sequence of the Rad6B promoter were end labeled and incubated with 10 µg of MCF10A or MDA-MB-231 nuclear proteins. The positions of specific retardations are indicated by complexes A to D for MCF10A cells and complexes 1 to 4 for MDA-MB-231 cells. A. Nuclear proteins were incubated with labeled Rad6B probe or with 10-fold (10x) or 50-fold (50x) molar excess of unlabeled Rad6B sequences [containing wild-type TCF (Wt) or mutant TCF (Mut)] as competitors. Lane 8 contained only the radiolabeled probe. B. MCF10A or MDA-MB-231 nuclear proteins were incubated with 0.5 µg of ß-catenin antibody or TCF-4 antibody in the binding reaction mixture before the addition of labeled Rad6B probe. Equivalent amounts of mouse or rabbit preimmune IgG were added as control (lanes 5, 8, 14, and 16). Note that addition of ß-catenin antibody severely inhibited formation of specific complexes 1 to 4 without any effect on nonspecific complexes (NS; compare lanes 6 and 7), whereas inclusion of normal IgG did not affect complexes 1 to 4 (lane 8). Similar addition of anti-ß-catenin antibody had no effect on complexes formed from MCF10A nuclear proteins (compare lanes 1, 4, and 5). Addition of anti-TCF-4 antibody to MDA-MB-231 reactions inhibited complex 1 without any effect on complex 3 or 4, whereas inclusion of the corresponding normal IgG had no effect (compare lanes 9 and 12 with lanes 13 and 14). Lane 12 contains 20 µg of MDA-MB-231 nuclear proteins. Addition of TCF-4 antibody to MCF10A reactions inhibited complex B, weakly affected complexes C and D, and had no effect on complex A (compare lanes 15 and 16 with lane 17).

View larger version (27K): [in this window] [in a new window]   FIGURE 5. Direct Western blot analysis of EMSA of the Rad6B promoter. MCF10A or MDA-MB-231 nuclear extracts were incubated with radiolabeled Rad6B promoter fragment in the presence or absence of a 50-fold molar excess of corresponding unlabeled wild-type competitor. Complexes were separated and either subjected to autoradiography (A) or transferred to nitrocellulose for Western blot analysis (B and C) with anti-ß-catenin (B) or anti-TCF-4 (C) antibody. The positions of specific complexes 1 to 4 formed with MDA-MB-231 nuclear proteins and separated on large gels (a) are shown for comparison of mobilities of corresponding complexes separated on mini gels (A-C). The specific complexes formed between the Rad6B probe and MCF10A nuclear proteins are not seen on EMSA done in mini gels because the electrophoretic conditions were set for attaining optimal separation of slower migrating complexes 1 to 4, which resulted in loss of the faster-migrating specific complexes formed with MCF10A nuclear proteins (Fig. 4). Because ß-catenin and TCF-4 immunoreactivities are seen as a smear (besides the defined complex 1 band) running up to the top of the gel, it is likely that they contain high molecular weight proteins such as p300 or additional proteins.

View larger version (14K): [in this window] [in a new window]   FIGURE 6. Analysis of UV cross-linked EMSA of the Rad6B promoter. MCF10A or MDA-MB-231 nuclear proteins were incubated with radiolabeled Rad6B promoter probe in the presence or absence of unlabeled wild-type Rad6B competitor. UV-irradiated as well as corresponding nonirradiated binding reactions were separated and subjected to autoradiography (A) or Western blot analysis with ß-catenin (B) or TCF-4 (C) antibodies. Lane 1, binding reaction without UV cross-linking; lane 2, binding reaction with UV irradiation; lanes 3 and 4, binding reactions incubated in the presence of a 10- or 50-fold molar excess of unlabeled competitor, respectively. B and C. Lanes 1, 2 and 1', 2', binding reactions with MCF10A and MDA-MB-231 nuclear extracts, respectively. B. A specific >250-kDa UV cross-linked ß-catenin reactive band (arrow, lane 2') was detected only with MDA-MB-231 extracts, which is not detected in nonirradiated reactions (lane 1'). C. Western blot analysis with TCF-4 antibody showed the presence of a specific UV cross-linked band of 75 kDa (arrow) with both MCF10A and MDA-MB-231 nuclear extracts (lanes 2 and 2') whereas this band was not detected in nonirradiated reactions (lanes 1 and 1'). Note the presence of TCF-4 immunoreactive bands >250 kDa (as in B) with MDA-MB-231 nuclear proteins (lane 2', bracket).

View larger version (16K): [in this window] [in a new window]   FIGURE 7. Rad6B promoter is occupied by ß-catenin in MDA-MB-231 cells but not in MCF10A cells. A. Chromatin immunoprecipitation was done on MCF10A, MCF10A cotransfected with p300 and ß-catenin expression plasmids, or MDA-MB-231 cells with anti-TCF-4 antibody or the corresponding normal IgG, followed by Western blot analysis with anti-ß-catenin or anti-p300 antibodies, or PCR amplification (B, c). Before chromatin immunoprecipitation, equal input was verified by immunoblotting with anti-tubulin antibody. B. Chromatin immunoprecipitation done on MCF10A and MDA-MB-231 cells (a and b) or MCF10A cells cotransfected with p300 and ß-catenin (c). C. Chromatin immunoprecipitation assays done with anti-ß-catenin antibody on MCF10A and MDA-MB-231 cells were PCR amplified with primers (bases 880-1,410; accession no. AF511593) encompassing the TCF [1351] site of the human cyclin D1 gene. The TCF [1351] site on cyclin D1 promoter is occupied by ß-catenin in MDA-MB-231 but not in MCF10A cells.

View larger version (99K): [in this window] [in a new window]   FIGURE 8. Constitutive overexpression of Rad6B in MCF10A cells induces hyperplasia. MCF10A cells stably transfected with Rad6B expression vector or their corresponding empty vector (16) were injected subcutaneously into the mammary fatpads (A) or flanks (B and C) of female nude mice. Lesions were harvested at 10 weeks, fixed, and paraffin embedded. Sections were stained with H&E (A-C), anti–proliferating cell nuclear antigen (D), or anti-Rad6 antibody (E and F). C. Magnified version of the outlined box in B. Vector control MCF10A cells fail to grow in vivo and hence are not shown. Magnification, x20 (A, B, and E); x40 (C, D, and F).

View larger version (161K): [in this window] [in a new window]   FIGURE 9. Immunohistochemical localization of Rad6 and ß-catenin in human breast carcinomas. A to G. Rad6 reactivity. A' to G'. Corresponding ß-catenin staining. B to E and B' to E'. Representative cases of breast tumor tissues showing a strong association between Rad6 and ß-catenin. A and A'. Nontumorous ducts showing negligible Rad6 (A) and nuclear ß-catenin expression in myoepithelial (arrow) and luminal epithelial (arrowhead) cells besides intense membrane reactivity. B and B'. Breast tissue with adenosis showing moderate cytoplasmic reactivity to Rad6 and ß-catenin. C, C' and D, D'. Examples of breast carcinomas with ductal carcinoma in situ (double-sided arrow) and infiltrating carcinoma cells displaying cytoplasmic and nuclear Rad6 and ß-catenin staining. E and E'. Invasive cancer cells showing cytoplasmic Rad6 and ß-catenin. F and F'. Breast cancer tissue showing a weak correlation between Rad6 (1+) and ß-catenin (2+) reactivity. G and G'. Breast tumor tissue showing no correlation between Rad6 and ß-catenin. Magnification, x20 (A, A', D, D', and G, G'); x40 (B, B', C, C', E, E', and F, F').

	Discussion

Top Notes Abstract Introduction Results Discussion Materials and Methods Acknowledgements References

 
In this study, we sought to establish the molecular basis for Rad6B overexpression in human breast cancer cells and tissues. The Rad6B gene does not contain a canonical TATA box, is GC-rich, and contains several putative SP1 sites. All of these features are characteristic of housekeeping genes that are constitutively active. However, dramatic increases in Rad6B gene expression are observed between normal and tumor breast tissues; such changes would seem to be incompatible with the notion of Rad6B as a housekeeping gene. Our data from functional promoter assays show that Rad6B promoter activity is negatively influenced by TCF binding. Mutation of sequences containing the TCF binding site (–341) caused a dramatic induction of basal reporter gene expression while abolishing its responsiveness to ß-catenin. These data indicate that TCF-4 plays an important negative regulatory role in control of Rad6B gene expression and activation of Rad6B promoter requires coexpression of both ß-catenin and p300. Results from EMSA and UV cross-linking experiments showed that the inability to form TCF-4/ß-catenin/p300 complexes in MCF10A cells is primarily due to a paucity of transcriptionally active ß-catenin and p300 in MCF10A nuclei and not due to deficiency of TCF-4 or inability of endogenous TCF-4 to bind to the Rad6B promoter. These data are consistent with high levels of TCF-4 expression reported in the epithelium of normal and cancerous human breast tissues (34).

Schlosshauer et al. (35) reported the presence by Western blot analysis of comparable, if not more, steady-state levels of ß-catenin in MCF10A total cell lysates as compared with MDA-MB-231 cells. However, a more systematic study by Bafico et al. (36) has shown up-regulation of uncomplexed transcriptionally active form of wild-type ß-catenin in 25% of breast cancer cells lines examined including MDA-MB-231 cells. Our findings are consistent with this report in that, unlike MDA-MB-231 cells, transcriptionally active ß-catenin levels are very low in MCF10A cells as shown by their diminished ability to bind to the Rad6B TCF binding sequence both in vitro (EMSA and UV cross-linking) and in vivo (chromatin immunoprecipitation), as well as to transactivate in reporter-based (Rad6B promoter and TOP/Flash) assays. TCF/LEF proteins function as transcriptional corepressors by binding to members of the Groucho/TLE family (31, 32, 37-39). Groucho/TLE proteins are general transcriptional corepressors (40-42) and have been shown to interact with histone deacteylases (39, 40), enzymes that maintain chromatin in a transcriptionally repressed state (40, 43). We have not examined whether Rad6B promoter repression observed in MCF10A cells is a result of formation of repression complexes at the TCF binding site; however, our findings suggest that activation of Rad6B promoter involves molecular switch from transcriptional repression to derepression or activation because it requires functional interaction between p300 and ß-catenin. Consistent with this, chromatin immunoprecipitation experiments done with MCF10A cells cotransfected with ß-catenin and p300 showed their recruitment to the Rad6B TCF binding site and induction of Rad6B promoter activity. These observations support a new concept for the pathophysiologic regulation by TCF-4/ß-catenin/p300 of the Rad6B gene, which is involved in normal and aberrant DNA repair and ubiquitin-mediated proteolytic processes. This regulation occurs at the transcriptional level in the presence or absence of coexpressed TCF-4; however, N-TCF-4 or excessive amounts of wild-type TCF-4 inhibit the ß-catenin/p300–induced Rad6B promoter activity. These data suggest that besides levels of ß-catenin and p300, levels of TCF-4 and potentially different isoforms of TCF that contain repressor or coactivator domains can influence the repression activity of TCF and, hence, the Rad6B promoter activity. In this regard, it is interesting to note that DNA-protein complex 1 formed from MDA-MB-231 nuclear proteins contains both TCF-4 and ß-catenin whereas ß-catenin immunoreactive complexes 3 and 4 lack TCF-4 immunoreactivity (Figs. 4 and 5). It is not clear at present whether this lack of TCF-4 immunoreactivity is indicative of ß-catenin complex formation with another TCF isoform. Although there is considerable divergence in the 5' untranslated region sequences of Rad6B genes, it is notable that the sequences corresponding to segments of human promoter (–581 to –418, –187 to –129, –88 to –50, and –28 to –1 relative to first translation start codon), as well as a putative TCF binding site, are conserved in the rat (586 bases upstream of ATG codon at base 275; accession no. U04303) and mouse (594 bases upstream of ATG codon at base 166,642; accession no. AL669920) Rad6B genes, further supporting their potential conserved role in regulation of Rad6B gene expression.

The canonical Wnt signaling pathway regulates various biological processes, including early neoplasia, by increasing the stability and transcriptional activity of the key mediator, ß-catenin (44-46). Studies in mouse model systems clearly show that activation of the canonical Wnt signaling pathway leads to mammary tumorigenesis (47), and a role for ß-catenin signaling in initiation of alveologenesis has been established. The mammary ducts of young virgin mouse mammary tumor virus (MMTV)-N89ß-catenin and MMTV-N90ß-catenin female mice show ductal hyperplasia, a feature consistent with early alveologenesis (48, 49). Up-regulation of ß-catenin target genes, cyclin D1 (50) and c-myc (51), was observed in both hyperplasia and adenocarcinoma, thus implicating ß-catenin signaling in preneoplastic and neoplastic transformation (48, 49). ß-Catenin-stabilizing mutations that activate Wnt signaling are frequently observed in human cancers of other tissues; however, equivalent mutations have not been detected in human breast cancers (52, 53). There is uncertainty in the literature about ß-catenin deregulation in breast tumors because immunohistochemical studies of ß-catenin in breast carcinomas have shown mixed results in terms of subcellular localization and correlation with other markers and patient outcome. Lin et al. (54) have reported an immunohistochemical correlation between ß-catenin stabilization (cytoplasmic/nuclear localization) and cyclin D1 expression in breast cancer tissues.

Our preliminary data, although from a limited number of breast tumors, have shown that 12 of the 12 breast carcinomas examined displayed moderate to intense reactivity to ß-catenin antibody, and 25% of these tumors showed combined cytoplasmic and nuclear staining. Most of the previous studies have been done on archival material whereas our present study was done on recently fixed tissues (acquired within the last 12 months), which may explain the existing uncertainty of ß-catenin expression in the nucleus. It is not clear whether the breast tumor surgical specimens used by Lin et al. (54) were recently fixed tissues because the details were not provided in the article. Although the number of breast tumor tissues analyzed was small, our current study found an immunohistochemical correlation between Rad6 and ß-catenin in breast carcinomas, with moderate to intense expression of both proteins in the cytoplasm and nucleus of ductal carcinoma in situ and invasive cancer cells. These data indicate that transcriptionally active ß-catenin may contribute to the derepression/activation of Rad6B promoter in breast carcinomas, similar to our findings from reporter and DNA binding assays in MDA-MB-231 breast cancer cells. It is interesting to note that nontumorous ducts of a breast carcinoma contained nuclear ß-catenin in the luminal and myoepithelial cells besides strong membrane reactivity. This early presence of nuclear ß-catenin may distinguish an involved duct from a normal duct and implicate its potential transcriptional activity in up-regulation of target genes such as Rad6B. It is worthwhile to mention that we have thus far not encountered high Rad6 and negative ß-catenin (with exclusive localization on the membranes) expression, although we have encountered moderate ß-catenin and weak or no Rad6 expression. We propose that in Wnt-activated breast cancer cells, ß-catenin/TCF-4 levels increase and activate TCF target genes, including Rad6B, which in turn contribute to lesion development in breast. A review of microarray databases failed to identify Rad6B among the Wnt-activated genes.³ A plausible explanation for this may be the short half-life of Rad6B transcripts, which impedes the build-up of adequate steady-state levels of Rad6B mRNA, which is necessary for identification of differentially regulated genes.

Our results from in vivo assays showed that MCF10A cells stably overexpressing Rad6B mainly produce benign hyperplastic (grade 2) lesions, whereas lesions produced by MCF10A cells engineered to express mutant Ha-ras gene (33) progress to atypical ductal hyperplasia (grade 3), ductal carcinoma in situ (grade 4), and invasive carcinoma (grade 5). The strikingly different effects exerted by Rad6B overexpression versus oncogenic Ha-ras on histologic conversion and neoplastic progression of MCF10A cells are consistent with our notion that induction or deregulation of Rad6B gene expression is an early step in human breast carcinogenesis. These data are consistent with our previous study that showed that detectable increases in Rad6B expression occur early in benign hyperplasias with continued overexpression in ductal carcinoma in situ and invasive breast carcinomas (16).

Rad6 is critical for postreplication repair of DNA and maintenance of genomic integrity. However, maintenance of critical levels of Rad6 is essential as imbalances in levels of Rad6B, such as by forced overexpression in normal breast cells, induce centrosome amplification, formation of multipolar mitotic spindles, multinucleation, and aneuploidy (16). Conversely, depletion of Rad6B in normal breast cells triggers defective growth and hypersensitivity to drugs (15). Although Rad6B is a fundamental component of the postreplication repair pathway and its expression is induced by DNA-damaging drugs, our present finding that it is a transcriptional target of TCF/ß-catenin shows that its expression is also regulated by oncogenic/mitogenic pathways, thus providing an alternate but important route to Rad6B transcriptional deregulation. ß-catenin/TCF target genes exhibit differences in their requirement for coactivators. Our findings from the present study show that TCF/ß-catenin–induced Rad6B expression is regulated in a p300-dependent fashion. A recent study showed that survivin transcription via ß-catenin/TCF signaling is dependent on the coactivator cAMP-responsive element binding protein (CREB)–binding protein (CBP), and that switching from CBP to p300 at the survivin promoter is associated with its transcriptional repression (55). Whether similar cooperation of ß-catenin with CBP or other transcription factors modulates Rad6B expression will require further study.

A larger study covering a wider spectrum of breast tumor types and grades is under way to further authenticate the ß-catenin and Rad6B expression profiles. In summary, this is the first report that provides evidence for a potential ß-catenin-mediated mechanism for Rad6B overexpression that may be operative in a subset of breast tumors.

	Materials and Methods

Top Notes Abstract Introduction Results Discussion Materials and Methods Acknowledgements References

 
Reagents
Antibodies were purchased as follows: anti-ß-catenin mouse monoclonal antibody (clone 15B8; epitope not described; Sigma, St. Louis, MO); anti-TCF-4 (13027X; rabbit polyclonal antibody recognizes epitope corresponding to amino acids 486-610 mapping near the COOH terminus of human TCF-4; Santa Cruz Biotechnology, Santa Cruz, CA); anti-p300 (Santa Cruz Biotechnology); anti-ß-actin (Sigma); and control rabbit and mouse antibody (Zymed, San Francisco, CA). pCMV-ß-catenin encoding full-length human ß-catenin, pCDNA3/myc-human TCF4, and pCDNA3-myc-N human TCF-4 (lacking the NH₂-terminal 30 amino acids) were generously provided by Dr. Bert Vogelstein (Johns Hopkins University, Baltimore, MD). pRV-RSV-HA-p300 was supplied by Dr. Richard Goodman (Vollum Institute, Oregon Health Sciences University, Portland, OR). Reporter plasmids pFOP/FLASH and pTOP/FLASH containing the consensus mutant and wild-type TCF binding elements, respectively, were purchased from Upstate Biotechnology (Lake Placid, NY).

Cells and Cultures
MCF10A cells were cultured in DMEM/F-12 supplemented with 5% horse serum, 10 µg/mL insulin, 0.02 µg/mL epidermal growth factor, 0.5 µg/mL hydrocortisone, 0.1 µg/mL cholera toxin, 100 units/mL penicillin, and 100 µg/mL streptomycin (16). MDA-MB-231 and SW480 cells were cultured in DMEM supplemented with 5% fetal bovine serum.

Construction of Rad6B Promoter Reporter
Cells were transfected with full-length (722 bp) and truncated (560, 410, 335, and 218 bp) human Rad6B promoter fragments that were subcloned upstream of a promoterless luciferase reporter construct pGL3-Luc (Promega, Madison, WI). Plasmids are named according to the length of the 5' untranslated region relative to the translation start codon (Fig. 1A). The forward primers used were 5'-gctgttgattctgccatgacctc-3' (–712/–690), 5'-acgcgtcatagggacacgtggttc-3' (–550/–527), 5'-ccccgccgagcctaaactagtg-3' (–401/–380), 5'-tctagtctgaacacagaga-3' (–325/–307), and 5'-cagcactcacactgtggtagcg-3' (–208/–187). The antisense primer was 5'-ggtcgacatgctccgcagctg-3' (+9/–12). The template was genomic DNA from MCF10A cells. The promoter fragments were subcloned into pCRII TA cloning vector (Invitrogen, Carlsbad, CA), digested with XhoI/HindIII, and cloned into the XhoI/HindIII sites of the pGL3-Basic reporter vector (Promega). All constructs were confirmed by sequencing in both directions.

Site-Directed Mutagenesis
The mutant (–712/+9) full-length Rad6B promoter was generated by site-directed mutagenesis of wild-type template DNA (pGL3–722/+9) with a single mutagenic oligonucleotide for PCR amplification followed by digestion with DpnI of template DNA. The conditions used for PCR in a total volume of 50 µL were as follows: 100 ng of template DNA, 20 pmol mutagenic primer, 5 µL of 10x Pfu buffer, 25 mmol/L of each deoxynucleotide triphosphate, and 2.5 units of Pfu polymerase. The conditions for PCR amplification were 95°C, 2 minutes followed by 18 cycles consisting of 1 minute at 95°C, annealing at 48°C for 2 minutes, and extension at 68°C for 10 minutes where the extension step is twice the length of the plasmid in kilobase in minutes. Final extension was done at 68°C for 20 minutes. Sequence integrity and the presence of the correct mutation at the TCF site (CTTTGAAggggGAA) were verified by sequence analysis. Mutant Rad6B promoter (–401/+9) was generated by PCR amplification from pGL3-TCF mutant –712/+9 as template, subcloned into pCRII vector, and subsequently cloned into the XhoI/HindIII sites of pGL3-Basic vector as described above.

Transient Transfections
Transient transfections were done using Metafectene (Biontex, Munich, Germany). Cultured cells at 70% confluence in 35-mm culture dishes were cotransfected with 2 µg of Rad6B promoter reporter constructs or the empty vector and 0.2 µg of the pRL-TK vector, a Renilla luciferase control reporter vector (Promega) used to normalize for transfection efficiency. To evaluate the transcriptional activity of endogenous TCF/ß-catenin, MCF10A or SW480 cells were transfected with 1.0 µg of TOP/FLASH or FOP/FLASH. The ability of endogenous TCF to be transactivated by ectopic ß-catenin was tested in MCF10A cells transfected with TOP/FLASH in the presence of ß-catenin (0.25 µg), and the specificity of ß-catenin stimulated response was tested by cotransfection of equivalent amounts of wild-type or N-TCF plasmids. To determine whether the putative TCF site on the Rad6B promoter is functional, MCF10A or MDA-MB-231 cells were cotransfected with wild-type or mutant TCF Rad6B promoter in the presence of ß-catenin (0.25 and 0.5 µg), p300 (0.25 and 0.5 µg), a combination of ß-catenin and p300, or a combination of ß-catenin, p300, and 2 µg of N-TCF-4 or wild-type TCF-4. For all cotransfection experiments, the total mass of DNA was kept constant by including appropriate amounts of the respective control vectors. At 45 hours after transfection, the cells were lysed and the luciferase activity was measured using the Promega Dual Luciferase Assay system. Absolute promoter firefly luciferase activity was normalized against Renilla luciferase activity to correct for transfection efficiency. Triplicate dishes were assayed for each transfection and at least four to six independent transfection assays were done for each reporter construct. Effects of ectopic ß-catenin, p300, and TCF-4 on Rad6B gene expression were determined by RT-PCR and Western blot analysis.

EMSA
A double-stranded oligonucleotide (5'-gtgtttctgtttcgtggtctttgaatccacaacctctag-3') containing the putative TCF/LEF binding site (underlined) was used as the probe. The double-stranded oligonucleotide was end labeled with [?-³²P]ATP and T4 polynucleotide kinase. Binding reactions were done for 15 minutes at 30°C by incubating 10 µg of MCF10A or MDA-MB-231 nuclear extracts and 0.1 pmol of labeled oligonucleotide in 20 µL of binding buffer [10 mmol/L Tris-HCl (pH 7.5), 50 mmol/L NaCl, 0.5 mmol/L EDTA, 0.5 mmol/L DTT, 2 µg of poly(deoxyinosinic-deoxycytidylic acid)]. Competition analysis was done by adding excess (1-5 pmol) unlabeled probe or with double-stranded oligonlucleotide (5'-gtgtttctgtttcgtggtgggggaatccacaacctctag-3') containing mutation (underlined) in the TCF/LEF binding site. Protein-DNA complexes were separated on 7% polyacrylamide gel. For supershift assays, nuclear extracts were preincubated with 0.5 µg of anti-TCF-4 or anti-ß-catenin antibody, or equivalent amounts of rabbit or mouse preimmune IgG, before the addition of the labeled double-stranded oligonucleotide. The gels were dried and autoradiographed with intensifying screens at –80°C.

To directly analyze and identify protein components binding to the DNA element of gel shift assay, the protein-DNA complexes in EMSAs were separated by PAGE on mini 5% polyacrylamide gels at 150 V in 0.5x Tris-borate EDTA (45 mmol/L Tris/45 mmol/L borate/0.5 mmol/L EDTA, pH 8) at 4°C. Western blots of gel shifts were done on nitrocellulose (Schleicher & Schuell, Keene, NH) membrane at 200 mA for 2.5 hours in 15 mmol/L Tris/120 mmol/L glycine (pH 8.3) at 4°C. Under these conditions, only the proteins are bound to the membranes whereas the double-stranded oligonucleotides are not retained on the filter (confirmed by autoradiography overnight), suggesting that protein-DNA complexes dissociate during the transfer. Proteins bound to membranes were identified by immunoblotting with antibodies to ß-catenin or TCF-4 and enhanced chemiluminescence reaction. Complexes formed between the detected proteins and the Rad6B DNA probe were verified by autoradiography of portions of gel from the same experiment.

UV Cross-Linking
Binding reactions were done as described above for EMSAs except that 100 µg of nuclear protein and 5 ng of radiolabeled probe were used. Following binding reactions in the presence or absence of competitor, the reaction mixtures were exposed to UV radiation (total energy, 0.25 J) in a UV Stratalinker (Stratagene, La Jolla, CA) at room temperature. UV-irradiated reactions and binding reactions not subjected to UV cross-linking were boiled in SDS-sample buffer and subjected to SDS-PAGE. Gels were dried and subjected to autoradiography or to Western blot analysis with anti-ß-catenin or anti-TCF-4 antibodies.

Chromatin Immunoprecipitation
Proteins and chromatin were cross-linked by treating MCF10A (control and transfected with ß-catenin/p300) and MDA-MB-231 cells with 0.5% formaldehyde for 15 minutes at room temperature before sonication in lysis buffer [10 mmol/L EDTA, 50 mmol/L Tris-HCl (pH 8.0), 1% SDS, protease inhibitor cocktail (Roche Biochemicals, Indianapolis, IN)] to achieve cross-linked DNA of 200 to 600 bp in length (56). After centrifugation, the supernatant was diluted 10-fold with dilution buffer [1% Triton X-100, 1.2 mmol/L EDTA, 167 mmol/L NaCl, 16.7 mmol/L Tris-HCl (pH 8.0)]. Nonspecific binding was eliminated via preincubation with nonimmune IgG and protein A/G agarose at 4°C for 2 hours. Agarose beads were removed by centrifugation and 15% of the supernatant was used as input. The supernatant was incubated with indicated antibodies and protein A/G agarose overnight at 4°C. Beads were collected and sequentially washed twice with 1 mL of wash buffers [0.1% SDS, 1% Triton X-100, 2 mmol/L EDTA, 20 mmol/L Tris-HCl (pH 8.0), 150 mmol/L NaCl; same buffer with 500 mmol/L NaCl; and 0.25 mol/L LiCl, 1 mmol/L EDTA, 0.5% NP40, 0.5% sodium deoxycholate, 10 mmol/L Tris-HCl (pH 8.0)]. The immunoprecipitates were reverse cross-linked for PCR or boiled for 5 minutes in SDS-sample buffer for Western blot analysis. The bound and the input DNA were analyzed by PCR with primers (–550 to –147 relative to +1 ATG start codon) that amplify a 404-bp fragment of the human Rad6B promoter. To verify the specificity of the chromatin immunoprecipitation assay, control primers circumventing the TCF binding sequence (–1,490 to –1,180 relative to +1 ATG start codon) in Rad6B promoter were also used for PCR amplification. In vivo occupancy of the cyclin D1 promoter by ß-catenin was verified by PCR with primers (bases 880-1,410; accession no. AF511593) encompassing the TCF (base 1,351) binding site of the human cyclin D1 promoter. PCR conditions were 95°C for 1 minute, 58°C for 2 minutes, and 72°C for 3 minutes for 35 cycles, followed by extension at 72°C for 10 minutes.

RT-PCR Analysis
DNase I–treated total RNA (2 µg) from MCF10A cells transfected with wild-type TCF-4, N-TCF-4, ß-catenin/p300, ß-catenin/p300/wt TCF-4, ß-catenin/p300/N-TCF-4, or control vectors was reverse transcribed using random hexamers and Superscript II (Life Technologies, Inc., Rockville, MD). Relative levels of Rad6B mRNA expression were determined by semiquantitative RT-PCR with +391/+407 (Rad6B specific) and +541/+557 (accession no. NM_003337) as forward and reverse primers, respectively, under predetermined conditions that yield a detectable PCR product within a linear range. For determination of optimal conditions, the reverse-transcribed cDNA was diluted serially in water from 100 to 3 ng/µL and mixed in a final volume of 10 µL with 1 µmol/L primer pairs, 1 unit of Taq DNA polymerase, and 1 µCi of [-³²P]dCTP. Amplification was done for 35 cycles (30 seconds at 95°C, 1 minute at 57°C, 2 minutes at 72°C) with a final extension for 10 minutes at 72°C. RT-PCR analysis of glyceraldehyde-3-phosphate dehydrogenase (GAPDH) cDNA was carried out for each reaction as an internal control for the amount of cDNA tested. The GAPDH-specific primers were forward, 5'-CATTGACCTCAACTACATGGT-3', and reverse, 5'-GGATCTCGCTCCTGGAAGA-3' (accession no. NM_002046). The PCR products were separated by agarose gel electrophoresis, subjected to autoradiography, and the relative intensities of the Rad6B and GAPDH cDNAs quantitated with NIH Imaging software.

Western Blot Analysis
Cell lysates were prepared as previously described (16) in 10 mmol/L Tris-HCl (pH 7.5)/150 mmol/L NaCl/1% Triton X-100/1 mmol/L phenyl methylsulfonyl fluoride, 1 µg/mL each of leupeptin, pepstatin, and antipain, and 1 mmol/L sodium orthovanadate from MCF10A cells transiently transfected with expression vectors for ß-catenin, p300, a combination of ß-catenin and p300, or corresponding control vectors. Proteins (50 µg) from each lysate were separated by SDS-PAGE and transblotted onto Immobilon-P membranes. The blots were stained with anti-Rad6 antibody. Levels of ß-catenin, p300, and TCF-4 were determined with antibodies to ß-catenin, p300, and TCF-4, respectively. Loading of protein was monitored by reprobing stripped membranes with anti-ß-actin antibody. The Rad6 antibody was generated by multiple immunization of New Zealand white rabbits with a synthetic peptide (K plus amino acid residues 131-152; accession no. NP_003328) that is conserved 100% in mouse and human HR6B and 91% in human HR6A (16). Because the human Rad6A and Rad6B proteins share 96% identical amino acid residues, and the synthetic peptide used for generation of Rad6B differs from Rad6A by only two amino acids, levels of Rad6 proteins expressed by Rad6A and Rad6B are not distinguishable. Thus, the 17-kDa Rad6-immunoreactive protein(s) detected in human cells is regarded as Rad6 rather than Rad6A or Rad6B (16). The relative amounts of Rad6 (HR6A/HR6B) protein(s) to ß-actin bands were quantitated with NIH Imaging software. MCF10A, MCF10A cotransfected with ß-catenin and p300 expression plasmids, or MDA-MB-231 cells were subjected to chromatin immunoprecipitation with anti-TCF-4 antibody or the corresponding normal IgG and immunoblotted with anti-ß-catenin or anti-p300 antibodies. Before chromatin immunoprecipitation, equal input was verified by immunoblotting with anti-tubulin antibody.

Three-Dimensional Morphogenesis Assay
To determine the effects of ectopic ß-catenin expression on Rad6B expression, single-cell suspensions of MCF10A cells (1 x 10⁵) in culture medium were transiently transfected with control vector or a mixture of ß-catenin and p300 expression vectors, mixed with an equal volume of 1:2 diluted Matrigel (Collaborative Biomedical Products, Bradford, MA) and seeded onto eight-chambered slides precoated with 40 µL of Matrigel (57). Culture media were replaced every third day, and on day 5 to 6, the cultures were fixed in buffered formalin and paraffin embedded. Sections were stained with antibodies to Rad6 or ß-catenin, followed by biotinylated anti-rabbit IgG or anti-mouse IgG, respectively, and horseradish peroxidase–conjugated streptavidin. Nuclei were counterstained with hematoxylin. The transiently transfected cells were also seeded on coverslips and monolayers assessed for expression/distribution of ß-catenin by immunofluorescence staining. Nuclei were stained with 4',6-diamidino-2-phenylindole.

Immunohistochemistry
Formalin-fixed, paraffin-embedded 5-µm breast carcinoma sections were incubated with anti-Rad6 or anti-ß-catenin antibodies, followed by biotinylated antirabbit or antimouse IgG secondary antibody, respectively, and horseradish peroxidase–conjugated streptavidin. Staining was visualized with 3,3'-diaminobenzidine and nuclei were counterstained with hematoxylin. Slides were also stained in the absence of primary antibody to evaluate nonspecific secondary antibody reactions. The