Inhibition of laminin-5 production in breast epithelial cells by overexpression of p300.

The transcriptional coactivator p300 is essential for normal embryonic development and cellular differentiation. We have been studying the role of p300 in the transcription of a variety of genes, and we became interested in the role of this coactivator in the transcription of genes important in breast epithelial cell biology. From MCF-10A cells (spontaneously immortalized, nontransformed human breast epithelial cells), we developed cell lines that stably overexpress p300. These p300-overexpressing cells displayed reduced adhesion to culture dishes and were found to secrete an extracellular matrix deficient in laminin-5. Laminin-5 is the major extracellular matrix component produced by breast epithelium. Immunofluorescence studies, as well as experiments using normal matrix, confirmed that the decreased adhesion of p300-overexpressing cells is due to laminin-5-deficient extracellular matrix and not due to loss of laminin-5 receptors. Northern blots revealed markedly decreased levels of expression of two of the genes (designated LAMA3 and LAMC2) encoding the alpha3 and gamma2 chains of the laminin-5 heterotrimer in the cells that overexpress p300, whereas LAMB3 mRNA, encoding the third or beta3 chain of laminin-5, was not markedly reduced. Transient transfection experiments with a vector containing a murine LAMA3 promoter demonstrate that overexpressing p300 down-regulates the LAMA3 promoter. In summary, overexpression of p300 leads to down-regulation of laminin-5 production in breast epithelial cells, resulting in decreased adhesion.

The transcriptional coactivator p300 is essential for normal embryonic development and cellular differentiation. We have been studying the role of p300 in the transcription of a variety of genes, and we became interested in the role of this coactivator in the transcription of genes important in breast epithelial cell biology. From MCF-10A cells (spontaneously immortalized, nontransformed human breast epithelial cells), we developed cell lines that stably overexpress p300. These p300overexpressing cells displayed reduced adhesion to culture dishes and were found to secrete an extracellular matrix deficient in laminin-5. Laminin-5 is the major extracellular matrix component produced by breast epithelium. Immunofluorescence studies, as well as experiments using normal matrix, confirmed that the decreased adhesion of p300-overexpressing cells is due to laminin-5-deficient extracellular matrix and not due to loss of laminin-5 receptors. Northern blots revealed markedly decreased levels of expression of two of the genes (designated LAMA3 and LAMC2) encoding the ␣3 and ␥2 chains of the laminin-5 heterotrimer in the cells that overexpress p300, whereas LAMB3 mRNA, encoding the third or ␤3 chain of laminin-5, was not markedly reduced. Transient transfection experiments with a vector containing a murine LAMA3 promoter demonstrate that overexpressing p300 down-regulates the LAMA3 promoter. In summary, overexpression of p300 leads to down-regulation of laminin-5 production in breast epithelial cells, resulting in decreased adhesion.
A phosphoprotein originally discovered by virtue of its binding to the adenovirus E1A transforming protein, p300 has subsequently been shown to function as a transcriptional coactivator for a very large number of transcription factors, bridging them with the basal transcription complex (1). p300 has enzymatic activity as a histone acetyltransferase that links chromatin remodeling with activation of transcription (2). p300 can mediate cross-talk among separate signaling pathways (1,3), has important roles in several fundamental cellular processes, including differentiation (4), is essential for normal embryonic and fetal development (5), and may function as a tumor suppressor (6). The p300 protein has several distinct domains that interact specifically with diverse proteins. p300 shares many functions with a highly homologous protein, cAMP-response element-binding protein-binding protein (CBP). 1 Whereas p300 and CBP seem to behave interchangeably for many functions, especially in vitro, there are also some specific functions for which one of the proteins cannot replace the other (7). Recently, experiments utilizing targeted gene disruption (knockout) technology have confirmed that p300 function is essential for normal embryonic cellular proliferation, morphogenesis, and development, with double knockouts resulting in 100% embryonic lethality (5). Even the haploinsufficiency of p300, as generated in the heterozygotes, resulted in severe developmental abnormalities and frequent embryonic lethality. Likewise, haploinsufficiency of CBP gives rise to the severe developmental abnormalities characteristic of the Rubinstein-Taybi syndrome, including mental retardation, craniofacial abnormalities, skeletal abnormalities, and increased cancer incidence (8). Thus, normal levels of p300 cannot replace CBP during embryonic development, and normal levels of CBP cannot replace p300 during embryonic development: both proteins are required.
In this paper, we describe the generation of stable cell lines from MCF10A cells that overexpress full-length p300 and the effects of this overexpression on production of laminin-5. Laminin-5 is the major extracellular matrix protein produced by MCF10A cells and is also the major protein in extracellular matrix of breast epithelium in vivo. The laminin-5 protein is a heterotrimeric glycoprotein consisting of the ␣3, ␤3, and ␥2 chains with each chain the product of a separate gene, designated LAMA3, LAMB3, and LAMC2, respectively (9). Not much is known about the transcriptional regulation of these genes to date. The murine promoter for the LAMA3 gene has been cloned (10), and its regulation by transforming growth factor ␤ in keratinocytes was found to involve three AP-1 sites (11). We show that p300 overexpression decreases the mRNA of both the LAMA3 and LAMC2 genes and can down-regulate the murine LAMA3 promoter in transient transfection assays. These results suggest that p300 may play an important role in the transcriptional regulation of the LAMA3 gene.

EXPERIMENTAL PROCEDURES
Cell Culture and Matrix Preparations-The MCF10A (12) normal breast epithelial cell line was obtained from the American Tissue Culture Collection (Manassas, VA) and was maintained in a 1:1 mix of Dulbecco's modified Eagle's medium and Ham's F-12 medium (Life Technologies, Inc.) supplemented with 5% equine serum (Life Technologies, Inc.), 0.01 mg/ml insulin (Sigma), 20 ng/ml epidermal growth factor (Collaborative Biomedical Products, Bedford, MA) 100 ng/ml cholera toxin (Sigma) and 500 ng/ml hydrocortisone (Collaborative Biomedical Products). Stably transfected clones were also supplemented with 50 g/ml hygromycin (Sigma). 8O4G rat bladder carcinoma cells were maintained as described previously (13).
Matrix preparations from MCF10A and 8O4G cells were prepared as described previously (14,15). Matrix still attached to plastic surfaces was either collected in sample buffer and run on an SDS-PAGE gel for Western blotting or labeled with 0.5 mg/ml biotinamidocaproic acid 3-sulfo-N-hydroxysuccinimide ester (Sigma) in PBS and then collected in sample buffer. Matrix preparations were resolved by 7.5% SDS-PAGE and blotted to nitrocellulose. Blots were incubated with streptavidin-horseradish peroxidase (Amersham Pharmacia Biotech) and detected with the ECL reagent (Amersham Pharmacia Biotech). To coat surfaces with matrix for immunofluorescence and adhesion studies, MCF10A cells or 8O4G cells were grown to confluence, treated with 20 mM ammonium hydroxide to destroy the cells but preserve the matrix on the desired surface, and washed well to remove cell debris (15). New test cells for experiments were then plated on this matrix-coated surface.
Plasmids and Transfection-For stable transfections, MCF10A cells were transfected either with a helium-pulse Accell (gene gun) device (Agracetus, Inc., Middleton, WI) or with Superfect (Qiagen, Inc., Valencia, CA) with the pTKHygro vector containing a hygromycin resistance gene under a thymidine kinase promoter and a vector designated pmet300. To generate pmet300, the cDNA for p300 was obtained from the p300CHA vector (16) and inserted 3Ј to the metallothionein promoter in the pMet vector previously described by Morosov et al. (17). Cells were selected for hygromycin resistance in MCF10A medium supplemented with 50 g/ml hygromycin (Sigma), and clones were isolated. Individual clones were tested for incorporation of the pmet300 vector by genomic and reverse transcription-polymerase chain reaction with primers specific for p300 from the vector. Isolated clones positive for p300 were designated MOP1 and MOP2. Hygromycin-resistant MCF10A cells that do not contain the pmet300 vector were designated M-H cells. Protein expression of p300 from the pmet300 vector was confirmed by immunoprecipitation with an antibody to the HA tag: 12CA5 (Roche Molecular Biochemicals).
The following plasmids were used for transient transfections. The plasmid CMVp300CHA, containing a CMV promoter driving expression of full-length p300 with an HA tag, was a kind gift of David Livingston (16). The plasmid pGalA, containing the murine LAMA3A promoter upstream from a ␤-galactosidase reporter, is the kind gift of Daniel Aberdam (10). MCF10A cells were transfected in 60-mm plates with Superfect (Qiagen, Inc.) with 10 g of pGalA and cotransfected with 2.5 g of pCMVp300CHA or empty vector and 2 g of pCMV luciferase vector as an internal control. Cells were harvested in Reporter Lysis Buffer (Promega Corp., Madison, WI) between 24 and 48 h. Cells were washed with PBS and lysed with 1ϫ lysis reagent (Promega), following the manufacturer's instructions. Luciferase activity was assayed by mixing aliquots of cell extracts with luciferin reaction mixture (Promega luciferase assay kit), and emission of light was quantitated with a Microlumat luminometer. ␤-Galactosidase activity was determined using the ␤-galactosidase luminescence kit (CLONTECH Laboratories, Inc., Palo Alto, CA) per the manufacturer's instruction. Transfections were normalized for efficiency by normalizing relative light units from the ␤-galactosidase assay to the luciferase light units for each individual sample.
Immunoprecipitation and Western Analysis-To verify protein expression from the pmet300 vector, exogenous p300 was immunoprecipitated using antibody to the HA tag. For the MOP1 cell line, cells were radiolabeled with [ 35 S]methionine for 24 h, lysed in radioimmune precipitation buffer (50 mM Tris, pH 8.0, 150 mM NaCl, 1% Nonidet P-40, 0.5% deoxycholic acid, 0.1% SDS, 1 mM EDTA, 20 g/ml phenylmethylsulfonyl fluoride, 1 g/ml pepstatin, 1 g/ml leupeptin, 1 g/ml aprotinin), and immunoprecipitated with the 12CA5 antibody (Amersham Pharmacia Biotech) to the HA tag of p300. Immunoprecipitates were analyzed by SDS-PAGE, the gel was fixed and dried, and bands detected by autoradiography. With the MOP2 cell line, whole cell lysates were immunoprecipitated with the 12CA5 antibody, resolved by SDS-PAGE, and blotted with the anti-p300, carboxyl terminus, clone RW128 (Upstate Biotechnology, Inc., Lake Placid, NY).
Analysis of total p300 levels was performed by collecting cell lysates in a borate buffer (50 mM sodium borate, 150 mM NaCl, 0.1 mg/ml phenylmethylsulfonyl fluoride, 1 g/ml aprotinin, 1 g/ml leupeptin, 1 g/ml pepstatin, 1% Nonidet P-40, 0.5% deoxycholic acid). 150 g of protein was loaded on a 7.5% SDS-PAGE gel and transferred to nitrocellulose. This membrane was then blotted with the p300 N15 antibody (Santa Cruz Biotechnology, Inc., Santa Cruz, CA), which does not cross react with CBP, incubated with a secondary goat-anti-mouse antibody conjugated to horseradish peroxidase, and detected by ECL (Amersham Pharmacia Biotech).
Western analysis of laminin-5 subunits was performed on cell lysates and matrix preparations in SDS-urea buffer. Antibody 10B5 to laminin-5 ␣3 (18), mouse monoclonal antibody 3D5 to laminin-5 ␥2, and the commercially available mouse antibody to kalinin B1 (laminin-5 ␤3 subunit), Clone 17 (Transduction Laboratories, Lexington, KY) were used. The anti-vinculin antibody was the mouse monoclonal VIN-11-5 from Sigma. Goat anti-mouse horseradish peroxidase (Bio-Rad) was used as a secondary antibody. The primary antibody to actin (Amersham Pharmacia Biotech) was detected with a secondary goat-anti mouse IgM (Kirkegaard and Perry Laboratories, Inc., Gaithersburg, MD).
Northern Blotting-The probe for ␤3 laminin-5 chain, clone 5-5C, and the EP-1 plasmid containing the laminin-5 ␣3 chain probe were generous gifts of Drs. Robert E. Burgeson and Maureen C. Ryan. The LAMC2 probe was amplified by reverse transcription-polymerase chain reaction with MCF10A RNA using polymerase chain reaction primers specific for the LAMC2 gene (5Ј-AGATGTTGATGGCTGGAAGGCTAGC-3Ј and 5Ј-AGGGTCACATTGTCAATGTA-3Ј). These primers create a fragment that runs from base pair 756 to base pair 1211 of the LAMC2 sequence obtained from GenBank TM . RNA was harvested from cells using the RNEasy Kit (Qiagen, Inc.). Samples of 30 g of total RNA were electrophoresed in denaturing 1% agarose/formaldehyde gels and transferred by capillary action to Hybond-N membranes (Amersham Pharmacia Biotech) in 20ϫ SSC. Blots were fixed and prehybridized at 68°C for 1 h in QuikHyb buffer (Stratagene, La Jolla, CA). 32 P-Labeled probes were prepared by the random primer Redivue method (Amersham Pharmacia Biotech), and 1 ϫ 10 6 cpm/ml of the LAMA3 probe plus an equal amount of glyceraldehyde-3-phosphate dehydrogenase probe as an internal control were added to the blot for 1 h of hybridization at 68°C. The membrane was then washed twice in 2ϫ SSC/0.1% SDS at room temperature for 5 min and once in 0.1ϫ SSC/0.1% SDS for 30 min at 60°C. This same blot was stripped and reprobed similarly with the LAMB3 and LAMC2 probes. Blots were visualized by autoradiography.
Adhesion Assays-Cells were trypsinized, washed with PBS, and resuspended in Dulbecco's modified Eagle's medium-Ham's F-12 medium and plated in 96-well plates at a concentration of 2 ϫ 10 4 cells/ well. Fibronectin-and laminin-1-coated 96-well plates were obtained from Collaborative Biomedical Products. Rat laminin-5 was purified as described previously (19,20), and 100 l was coated in 96-well plates at a concentration of 25 g/ml. Cells were incubated for 3 h. Medium was aspirated from the plates, which were then washed by submersion in PBS. Remaining cells were stained in a solution of 0.05% methylene blue in 25% methanol for 10 min. Plates were then rinsed well with dH 2 O and allowed to dry. Once dry, 0.1 M sodium citrate in 50% ethanol was added to the wells to resuspend the methylene blue, and the absorbance was read at 540 nm. Measurements were normalized to crystal violet staining per cell for each cell type.
Immunofluorescence-M-H, MOP1, and MOP2 cells were grown on glass coverslips and fixed with 3.7% formaldehyde for 5 min, washed with PBS, permeabilized with 0.5% Triton X-100 at 4°C for 7 min, and washed again in PBS. The coverslips were incubated with primary antibody GoH3 (Immunotech, Inc., Westbrook, ME) for ␣ 6 integrin staining, and antibody P1B5 (Life Technologies, Inc.) for ␣ 3 staining, diluted in PBS at 4°C in a humid chamber overnight, washed three times in PBS, and incubated with an appropriate fluorochrome-conjugated secondary antibody for 1 h at 37°C.

RESULTS
Characterization of MCF10A Clones Overexpressing p300 -MCF10A cells were stably transfected with the pmet300 plasmid expressing full-length p300 cDNA from the metallothionein promoter (17). The p300 in this construct was tagged with a hemagglutinin epitope at the carboxyl terminus (16). These cells were simultaneously transfected with a plasmid containing the hygromycin resistance gene as a selectable marker. We chose the metallothionein promoter for two reasons: (a) we felt that high level p300 expression, using a CMV promoter for example, was likely to be toxic to the cells, and (b) the metallothionein promoter was likely to be activated in MCF10A medium containing 4 M ZnSO 4 . Two clones, called MCF10Aoverexpressing p300 1 and 2 (MOP1 and MOP2), were isolated that express p300 in low enough amounts that the cells con-p300 Overexpression Decreases Laminin-5 tinue growing in long term culture. Fig. 1A shows an immunoprecipitation of [ 35 S]methionine-labeled cell lysates with an anti-HA antibody that was analyzed by SDS-PAGE. M-H control cells, which are resistant to hygromycin but do not contain the pmet300 vector, expressed no HA-tagged protein. MOP1 cells expressed a protein that migrated at 300 kDa and was immunoprecipitated with an antibody to the HA tag. In Fig.  1B, we show that cell lysates from MOP2 cells immunoprecipitated with an anti-HA antibody and blotted with an anti-p300 antibody detected HA-tagged p300, which was absent in M-H control cells. This confirms the identity of the 300-kDa band immunoprecipitated by the anti-HA antibody to be p300. Western analysis of whole cell extracts in Fig. 1C demonstrates that there was a 3-fold increase of p300 in MOP1 cells compared with M-H control cells and that MOP2 cells maintained a 5-fold higher level of p300 than M-H cells. The morphology of these cells in tissue culture is shown in Fig. 2. M-H cells maintain a normal epithelial cobblestone morphology ( Fig. 2A). Panels B and C of Fig. 2 show the more stellate morphology of MOP1 and MOP2 cells, respectively. We have observed that the stellate morphology of MOP cells is more pronounced when the cells are at a low density and less so at higher densities.
Through its interaction with p53, p300 has been suggested to play a role in p53-regulated cell growth (21). In muscle cells, p300 is required for Myo-D-dependent cell cycle arrest (22). Based on these findings, we hypothesized that overexpression of p300 in MOP cells may have altered the growth rates of these cells. Fig. 3 shows that M-H cells maintained a faster growth rate than MOP1 and MOP2 cells. Growth of M-H cells is similar to that of parental MCF10A cells (data not shown). Saturation densities at day 8 were 31.  Adhesion experiments were performed in which cells were treated with 20 g/ml cycloheximide at the time of plating for the adhesion assay to inhibit protein synthesis in the cells. Cells were incubated for 3 h and stained as above. Adhesion of control cells was decreased by 72%, and MOP1 and MOP2 adhesion was reduced 100%, to the cell-free baseline. This shows that protein synthesis is necessary for maximal adhesion in all of the cells tested.
Overexpressing p300 Results in a Decrease of Laminin-5 Extracellular Matrix-MCF10A cells secrete abundant extracellular matrix, of which laminin-5, a heterotrimeric protein with a molecular mass of greater than 400 kDa, makes up more than 80%. Laminin-5 consists of three subunits, ␣3, ␤3, and ␥2, which in their processed forms are 160, 140, and 105 kDa, respectively. We analyzed more closely the laminin-5 production in these cells. Northern blot analysis with probes for the human LAMA3, LAMB3, and LAMC2 genes, which code for the ␣3, ␤3, and ␥2 subunits of laminin-5 respectively, were performed. The LAMA3 transcript was virtually absent in MOP cells overexpressing p300 (Fig. 5A) compared with control cells. A longer exposure of the blot shown revealed that low amounts of LAMA3 transcript were visible in MOP cell lanes when MCF10A and M-H lanes were highly overexposed. The LAMB3 transcript normalized levels were not significantly different among the four cell types (Fig. 5B). Densitometry of this LAMB3 blot revealed normalized values of 1.0, 0.9, 0.56, and 1.33 for MCF10A, M-H, MOP1, and MOP2 respectively. This shows that the values for LAMB3 levels may vary somewhat but not by a large amount or in a manner specifically correlating with p300 overexpression. The LAMC2 transcript was dramatically decreased in MOP cells similarly to the LAMA3 transcript (Fig. 5C). Longer exposure of the blot shown revealed low levels of LAMC2 in MOP cells. Levels of glyceraldehyde-3phosphate dehydrogenase are constant in all four cell types, which confirms that the down-regulation of LAMA3 and LAMC2 transcript was not due to a general effect on transcription of overexpressing p300.
On the basis of the RNA data, we examined protein levels of the ␣3, ␤3, and ␥2 subunits by Western analysis of whole cell extracts and matrix preparations. We first collected matrix from confluent cells, which was subsequently biotinylated, electrophoresed, transferred to nitrocellulose, and detected by blot-ting with horseradish peroxidase-linked streptavidin. Fig. 6A shows the analysis of total cellular matrix from M-H, MOP1, and MOP2 cells. Laminin-5 subunits were easily detected in M-H cells but absent in MOP cells. We further analyzed matrix production by M-H, MOP1, and MOP2 cells by Western analysis with specific antibodies to each subunit. The ␣3 subunit was present in cell lysates and matrix from MCF10A and M-H cells in the 190-kDa unprocessed form and the 160-kDa processed forms (Fig. 6B). The ␣3 subunit was not detectable by Western blot in the MOP1 and MOP2 cell lysates or matrix (Fig. 6B). The ␤3 subunit was present in the lysates of all cell types but was not seen in the MOP1 and MOP2 cell matrices (Fig. 6C). We simultaneously blotted the ␤3 blot with an antibody to actin to control for protein loading. Densitometric analysis of the ␤3 blot in Fig. 6B revealed that the amount of ␤3 protein present in the cell lysates from MCF10A and M-H cells contained 3.5-fold the amount present in MOP1 and MOP2 cells. This is different from the levels of RNA, which do not change significantly, but the difference is probably due to decreased ␤3 protein stability in MOP cells due to the lack of the other laminin-5 subunits. The ␥2 subunit has a 155-kDa unprocessed and 105-kDa processed form. The Western analysis for the ␥2 subunit showed mainly the unprocessed form in cell lysates of MCF10A and M-H cells but no ␥2 in MOP1 or MOP2 cell lysates (Fig. 6D). The processed form was more evident in the matrix of MCF10A and M-H cells. The unprocessed form was still present because matrix samples were collected from subconfluent cells. Fig. 6E shows Western analysis of the blot shown in Fig. 6D, with an antibody to vinculin as a loading control for the cell lysates.
p300 Down-regulates the LAMA3A Promoter-Because the laminin-5 ␣3 subunit was dramatically reduced, we considered it possible that this gene is down-regulated at the transcriptional level. We performed experiments using a clone, designated pGalA, containing the promoter of the murine LAMA3A gene linked to the ␤-galactosidase reporter (8,36). The laminin ␣3 chain in mice has two isoforms, designated 3A and 3B, each using its own promoter. This clone contains the promoter of the 3A isoform. Transient transfection experiments placing pGalA into MCF10A cells revealed that the mouse LAMA3A promoter drove transcription of beta galactosidase in these human cells (Fig. 7A). Cotransfection of MCF10A cells with pGalA and a plasmid expressing p300 driven by a CMV promoter resulted in inhibition of pGalA-driven transcription by about 50% (Fig.  7A). Cells were cotransfected with a CMVLuc construct, and results are shown normalized to luciferase activity. There was no evidence of a general repression of transcription in these cells, as the control CMV-luciferase plasmid expression was unchanged from controls that were not transfected with CMVp300 (data not shown). FIG. 5. Northern analysis of the genes for the three laminin-5 chains. A, 30 g of total RNA was electrophoresed on a formaldehyde/ agarose gel, transferred to Hybond-N, and probed with a 32 P-radiolabeled probe to the LAMA3 gene. B, Northern blot for the ␤3 chain C, Northern blot of the ␥2 chain. A glyceraldehyde-3-phosphate dehydrogenase (GAPDH) internal control is included in each blot.

p300 Overexpression Decreases Laminin-5
We next transiently transfected MCF10A, M-H, MOP1, and MOP2 cells with the pGalA vector and compared ␤-galactosidase activity between the different cell lines. We found that the MOP1 and MOP2 cells that overexpress p300 had decreased pGalA promoter activity compared with MCF10A and M-H controls (Fig. 7B).
Organization of Laminin-5 Integrin Receptors-The major receptors for laminin-5 on breast cells are the integrins ␣ 6 ␤ 4 and ␣ 3 ␤ 1 . The ␣ 3 ␤ 1 integrin localizes primarily in and near areas of cell-to-cell contact. M-H, MOP1, and MOP2 cells all displayed similar cell-to-cell contact accumulation of the ␣ 3 integrin (Fig. 8). The ␣ 6 ␤ 4 integrin heterodimer has been localized primarily to the basal surface of the cell, where it forms part of a complex multiprotein adhesive structure called the hemidesmosome (5,11). In vivo, hemidesmosomes are found attaching cells of epithelia to the basement membrane. Immunofluorescence studies of the ␣ 6 integrin revealed mottled patches with a "Swiss cheese" pattern on the basal surface of the cells, as seen in M-H cells on tissue culture (Fig. 9A). However, the p300-overexpressing cells displayed virtually no hemidesmosomal pattern staining of ␣6 in tissue culture (Fig.  9C). The ␣ 6 integrin staining in MOP cells is more diffuse, with accumulation of ␣ 6 staining seen in cell-cell contacts. The normal hemidesmosomal staining pattern was restored when the MOP2 cells were plated on a laminin-5-rich matrix (Fig. 9E). Staining for the ␤ 4 integrin co-localized with the ␣6 staining in all cells (data not shown).
FIG. 6. Laminin-5 protein analysis. A, biotinylated matrix preps were analyzed by SDS-PAGE, transferred to nitrocellulose, incubated with streptavidin-horseradish peroxidase, and detected by ECL. All of the proteins present in the matrix preparations are detected using this method. The proteins at 160, 140, and 105 kDa are labeled as laminin-5 ␣3, ␤3, and ␥2 subunits, respectively. B, whole cell lysates and matrix preparations normalized to cell number were loaded on a 7.5% SDS-PAGE gel, transferred to nitrocellulose, and blotted with antibody to the laminin-5 ␣3 subunit. C, this same blot was stripped and was double labeled with antibody to the laminin-5 ␤3 subunit and ␣-actin as a control for protein loading in the cell lysates. Matrix preparations were normalized to cell number, as were the lysates, prior to loading. D, a 6% SDS-PAGE gel was run with the same samples used in B. The proteins in the gel were transferred to nitrocellulose and blotted with the 3D5 antibody to laminin-5 ␥2 subunit. E, the blot in D was stripped and reprobed with antibody to vinculin as a control for protein loading in the cell lysates. p300 Overexpression Decreases Laminin-5 DISCUSSION Laminins are a family of large heterotrimeric glycoproteins consisting of three related but different chains held together by coiled-coil interactions and disulfide bonds (23,24). There are at least 11 laminin isoforms, resulting from a variety of combinations of ␣, ␤, and ␥ chains (25). To date, five ␣, three ␤, and two ␥ chains have been described. One of these isoforms, laminin-5, is an extracellular matrix protein that is important in the processes of epithelial cell adhesion and migration (26 -28). In breast cells, laminin-5 has been shown to play a role in branching morphogenesis. Function blocking antibodies to laminin-5 inhibited branching morphogenesis of MCF10A cells on Matrigel (29). The role of laminin-5 in tumor progression appears to be tissue type-dependent, with laminin-5 highly expressed at the invasive edge in gliomas, colon carcinomas, gastric carcinomas, and squamous cell carcinomas (30 -34), but it is reduced in basal cell, prostate, and breast carcinomas (35)(36)(37)(38)(39).
Laminin-5 is found primarily in the basement membranes of epithelial cells, which suggests that its expression is highly regulated and tissue-specific. The expression of p300 is ubiquitous and has been shown to activate transcription in a large variety of cell types. Accumulation of p300 has been shown in undifferentiated embryonal carcinoma cells and mediates the transcriptional repression of the SV40 enhancer in these cells (40). There is not much known about the functional activation of p300 itself. Differences in phosphorylation during the cell cycle suggest that phosphorylation of p300 may be important in its activity (41). It is not known at this time what temporal relationship may exist between laminin-5 production and p300 activity.
We show that p300 overexpression leads to a decrease in the laminin-5 present in the extracellular matrix and that this is due to a decrease in the production of the ␣3 and ␥2 subunits but not the ␤3 subunit. This decreased production is shown both at the mRNA and the protein level. The ␣3 laminin subunit is found in the laminin-5, laminin-6 (␣3, ␤1, and ␥1) and laminin-7 (␣3, ␤2, and ␥1) isoforms, whereas the ␥2 laminin subunit is unique to the laminin-5 isoform. Laminin-5 and laminin-6 are both present in the basal lamina of stratified epithelia (42). Expression of the ␤2 subunit, a component of laminin-7, is believed to be restricted to the motor neuron synapse, blood vessels, and the kidney glomerulus (43)(44)(45). The regulation of the ␣3 and ␥2 chains by p300 may also affect the production of these laminin isoforms.
MOP cells exhibit decreases in adhesion to tissue culture plastic compared with control cells. We show that plating MOP cells on a laminin-5 surface improves their adhesion but also improves the attachment of M-H control cells. In fact, the improvement of attachment of M-H cells to laminin-5 is greater than the change seen in MOP1 cells between tissue culture adhesion and laminin-5 adhesion. We believe that the decreased adhesion of MOP cells to tissue culture plastic may in part be due to their decreased laminin production. There may be additional factors that are contributing to this phenomenon.
There are two integrin receptors, ␣ 3 ␤ 1 and ␣ 6 ␤ 4 , for which laminin-5 has been reported to be a ligand. We show that cells that produce less laminin maintain normal ␣ 3 ␤ 1 integrin receptors. The ␣ 6 ␤ 4 integrin has an altered distribution pattern in MOP cells but has normal distribution when in the presence of a laminin-5-rich matrix. The laminin-5 adhesion data do support the hypothesis that MOP cells express the integrin receptors necessary for attachment to laminin-5, which is confirmed by immunofluorescence.
We find that plating all cell types on fibronectin greatly increases their adhesion. Fibronectin is present in very small quantities in the extracellular matrix of MCF10A cells but is found in whole cell extracts of MCF10A, M-H, MOP1, and MOP2 cells (data not shown), and overexpressing p300 does not result in a reduction of fibronectin production. It is not surpris-   ing that plating on fibronectin increases the adhesion of these cells because breast epithelial cells make the ␣ 5 ␤ 1 and the ␣ 3 ␤ 1 integrin fibronectin receptors. Plating the cells on laminin-5 or fibronectin increases the adhesion of all of the cell types. This confirms that MOP cells are able to adhere to a laminin-5 or fibronectin substrate when it is present but are not able to produce laminin-5. All cell types will adhere strongly to fibronectin despite the low amounts of fibronectin found in their extracellular matrix. It is not clear why laminin-1 is inhibitory to adhesion in all of the cell types, but others have also reported that laminin-1 can be antiadhesive for epithelial cells (46).
Decreased levels of laminin-5 matrix in MOP cells may also be playing a role in the decreased growth rates of these cells. Cell signaling through laminin-5 can regulate proliferation of epithelial cells (19). It may also be possible for p300 to be affecting cell cycle through interactions with p53. We see no detectable laminin-5 in either MOP cell line, but the growth rate of these cell lines is different. These data suggest that although it may be possible that laminin-5 is involved in the decreased rate of cell growth, there are probably other factors that are modulating the growth rates of these cells.
We were very interested in the decrease in LAMA3 expression. We are not the first to report repression of a promoter element when p300 is in excess. Repression of the SV40 enhancer was seen with an accumulation of p300 in undifferentiated embryonal carcinoma cells (40). Results from transient transfection experiments (Fig. 7) suggest that overexpressing p300 has a down-regulating effect on the murine LAMA3A promoter activity. The murine LAMA3A promoter has three AP-1 sites that are important in its regulation (11). AP-1 has been shown to interact with p300 (47,48) and may be important in mediating the down-regulation of the LAMA3A gene when p300 is overexpressed possibly by sequestration of AP-1.
It has recently been discovered that CBP acts as a corepressor for TCF-1 in Drosophila (49). This discovery suggests that p300/CBP may not only act as a coactivator but also as a corepressor. Cloning and further analysis of the promoter will be necessary to determine the mechanism by which overexpression of p300 results in repression of the LAMA3A promoter. The decrease in laminin-5 production appears directly related to p300 through a transcriptional mechanism. In future studies, we hope to determine the specific mechanism(s) by which p300 decreases human LAMA3 transcription.