In Vitro Cell. Dev. Biol. 28A:730-734, November-December 1992 © 1992 Tissue Culture Association 0883-8364/92 $0t.50+0,00
RELATIVE PROMOTER ACTIVITY IN HUMAN MAMMARY EPITHELIAL CELLS ASSAYED BY TRANSIENT EXPRESSION GUDRUN HUPER, JEFFREY R. MARKS, JON R. WIENER, AND J. DIRK IGLEHART1
Department of Surgery, Duke UniversityMedical Center, Box 3873, Durham, North Carolina 27710 (Received 4 March 1992; accepted 13 May 1992)
SUMMARY Chimeric DNA expression vectors containing regulatory sequences proximal to the 5' end of coding sequences for mammalian genes provide valuable tools to study gene expression. Genes coding for easily measured products (reporter genes) can be used to study promoter strength and regulation of gene expression after transient expression of promoter-reporter constructs in mammalian cells. To determine the strength of a variety of mammalian and viral promoter-enhancer sequences in primary cuhures of human mammary epithelial cells (HMEC), these sequences were fused to the bacterial chloramphenicol acetyltransferase (CAT) gene and transfected into HMEC using strontium phosphate. The long terminal repeat (LTR) of the endogenous murine leukemia virus AKR-623 was the most potent promoter of transient CAT expression in HMEC. A number of commonly available promoter sequences displayed a wide range of activities in these cells. The glucocorticoid responsive LTR promoter from the murine mammary tumor virus modulated expression of CAT and was sensitive to the concentration of dexamethasone in the growth media. In a similar fashion, the regulatory sequences from the murine metallothionein-1 gene retained responsiveness to zinc concentration in the growth media.
Key words: human mammary epithelial cells; transfection; promoter; CAT assay. and human origin (1,11,24). The use of extracellular matrix proteins as a substratum results in the appearance of differentiated characteristics such as casein and milk fat protein, duct formation, and myoepithelial cells (6,13,23). These results demonstrate that coordinated expression of normal cellular genes and complex functions are possible in cultured mammary epithelial cells. Transient expression systems can be utilized to investigate the control of gene expression in normal cells and, perhaps, to investigate regulatory defects in breast malignancy. The current study was undertaken to assess the relative strengths of a series of widely used regulatory sequences in primary cultures of human mammary epithelial cells (HMEC). The goal of these experiments was to identify promoter-enhancer elements for eventual use in stable transfections of putative oncogenic sequences. Strontium phosphate-mediated transfection was used to introduce standardized vector constructs containing the CAT reporter gene into HMEC. The ability of inducible promoter elements to function in normal mammary cells was also studied. A considerable heterogeneity of promoter strength was found which appears to be cell-type specific. In addition, two inducible regulatory elements were found to function in a dose-dependent manner.
INTRODUCTION The introduction of cloned genetic material into cells growing in culture is a valuable means to study the function and regulation of gene expression. Calcium phosphate mediated transfection is a popular technique which can be applied to a wide variety of cell types. Because calcium induces a wide-range of physiologic effects in certain human epithelial cell cultures, including differentiation and growth inhibition, an alternative method of gcne transfer into epithehal cells using strontium phosphate precipitates has been developed. Strontium phosphate mediated transfeetions are efficient and lead to both transient expression and stable genomie integration with a minimum of deleterious effects on cultured epithelial cells (2). Recombinant DNA technology allows the study of gene regulatory sequences separated from the coding sequences that they control. The bacterial gene coding for chloramphenicol acetyltransferase (CAT) is a common reporter gene used in chimeric constructs to study gene expression in a variety of mammalian cell lines (9,10). This simplified system can be used to dissect complex regulation of gene expression by placing the CAT coding sequences under the control of promoter and enhancer elements that are able to function in the target cell type chosen for study. Furthermore, stable transfection and expression of other genes may be optimized by choosing the promoter sequence with the best transient activity in the target cell (9). Culture conditions have been described that will support the short-term proliferation of mammary epithelial cells from mouse
MATERIALSAND METHODS
Construction of CATplasmids. All constructs contained the CAT (Tn9) gene with its constituent SV40 early region intron and polyadenylation sequences (10). pSV2-CAT contains the entire SV40 early promoter/enhancer region previously described (10). The followingplasmids were constructed starting with an enhancer and promoter-less CAT containing plasmid designated pCAT. pCAT is the HindlII-BamHI fragment of pSV2-CAT,
1 To whom correspondence should be addressed. 730
731
CAT GENE EXPRESSION IN BREAST CELLS
piECAT
pAKRCAT
pRPCAT
pSV2 CAT -
PTKCAT
pMMTVplCPOCAT CAT
Fit. l. Autoradiograph exhibiting conversion of [~4C]ehloramphenicol to acetylated products by CAT activity in transfected HMEC. Aeetylated products migrate further into the gel and appear above the unmodified chloramphenicol(bottom). Corresponding regions of the gel were cut out and radioactivity counted to calculate the percent conversion (see Table 1). In this experiment, duplicate transfections are run in parallel lanes. Bottom, paired lanes are labeled with the plasmid construct used for transfection. In each case, 20/xg of plasmid DNA was transfeeted per plate.
converted to blunt ends with Klenow was cloned into the EcoRV site of Bluescript KS+ (Stratagene, LaJolla, CA). pIE-CAT contains the murine cytomegalovirus immediate early enhancer/promoter on a 1.2 kb Hpal: PstI segment from an EcoRI fragment of virion DNA (19) cloned into pCAT. pTK-CAT was constructed from the HSV-1 thymidine kinase (tk) gene promoter by cloning the 250 bp Pvull-Bglll fragment upstream of the tk gene into the polylinker of peAT (18). plCPO-CAT contains the EeoRI fragment of plGA65, from the promoter sequences of the infected-cell polypeptide-0 (ICP0) gene of HSV-1 (8), cloned into pCAT and obtained from Dr. Mark Labow (14). pRP-CAT contains the murine ribosomal protein L32 promoter (3) cloned into the Bluescript polylinker upstream of peAT. This promoter region is contained on a 1.1-kb fragment extending to the StuI site in intron 1. pAKR-CAT contains the long terminal repeat (LTR) from the endogenous murine retrovirus AKR-623 (15) inserted upstream of the CAT gene in pCAT (from J. Horowitz, Duke University). pMMT-CAT contains the mouse metallothionein promoter contained on a 1.9-kb EcoRI-BGI II fragment from p341-3 (5) cloned into peAT. pMMTV-CAT, called pMK-2 in the description of its construction, contains the murlne mammary tumor virus LTR and enhancer elements found in the LTR of the Harvey sarcoma virus, both upstream of the CAT gene in pCAT [obtained from M. Ostrowski, Duke University (21)].
Cell culture and transfection of normal human mammary epithelial cells. Human mammary tissue from reduction mammoplasties were digested enzymatically to epithelial organoids exactly as described (25). Cultures were established in DFCI-1 growth media that contains alpha minimal essential media (otMEM) and Ham's F12 (1:1, vol/vol) and supplemented as described (1). These cells seem epithelial, express human epithelial cytokeratins, and are diploid. Murine strain L cells, clone 929, were maintained in Dulbecco's MEM with 5% fetal bovine serum. L cells were obtained from the American Type Culture Collection, Rockville, MD. Growth and incubations were done in a 5% CO2 humidified incubator. Plasmid DNA was grown in Escherichia coli JMI09 and purified by banding in cesium chloride gradients (16). HMEC were plated in 60-ram plastic cell culture dishes at 2 × 10 s cells per dish and used for transfection 24 h later. Strontium phospfiate transfections were done as described by Brash et al. (2). Precipitates of strontium phosphate were monitored by microscopy and added at the stage of fine dust to culture dishes containing DFCI-1 media. Dishes were incubated for 4 h at 37 ° C, washed twice with phosphate buffered saline, treated for 30 s with 11.8% (vol/vol)glycerol in phosphate buffered saline (PBS), washed 3 times with PBS, and refed with complete DFCI-1 medium. Calcium phosphate transfections were done by parallel techniques exemplified by Gorman et al. (10). Unless indicated,
each plasmid DNA was transfected onto three separate dishes in each experiment to monitor plate-to-plate variability: CAT assays. Forty-eight hours after transfection, cells were dispersed, washed in PBS, and cell pellets resuspended in 100 #1 of 0.25 M Tris-HC1 (pH 7.8). The cells were lysed by three freeze-thaw Cycles and clarified by centrifugation at 12 000 × g for 2 min. Tile CAT assay reaction mixture contained 50 #1 of the cell extract; 75 #1 of 1 M Tris-HC1 (pH 7.8) containing 0.025 #Ci of [14C]chloramphenieol (Amersham, Arlington Heights IL). The reaction was initiated by addition of 25 #1 of 100 #g acetyl coenzyme A in 1 M tris, pH 7.8. The reaction mixtures were incubated at 37 ° C for 90 min and stopped by the addition of 0.7 ml of ethyl acetate and vigorous vortexing. The organic layer was dried, resuspended in 20 #1 of ethyl acetate, and spotted at the origin of thin-layer sihca gels. The samples were chromatographed in a mixture of chloroform-methanol (95:5), and the reaction products located by autoradiography. Spots were cut out and counted in scintillation fluid, Data were expressed as the percent of chloramphenicol converted to its major acetylated form pe r microgram of protein in 1 h. Percent conversion for each cell line represents a minimum of three separate transfections and is expressed as the mean plus or minus the standard deviation. Protein concentration was determined using commercial kits (Bio-Rad Protein Assay, Bio-Rad Laboratories, Richmond, CA). In each set of experiments or at each concentration of media additive s, pSV2-CAT was included in three separate dishes to monitor batch-to-batch conditions of the cell cultures and transfection efficieneies. RESULTS
Promoter strength in HMEC and murine fibroblasts.
To determine the relative strength of the various promoter constructs to transiently express CAT, plasmid DNA was transfected into HMEC using the strontium phosphate method. With certain promoter constructs, varying amounts of plasmid DNA were transfected to establish a dose-response relationship. After 48 h, cells were lysed and CAT activity determined. Figure 1 displays qualitative results from a representative series of transfections analyzed by autoradiography. In Table 1, quantitative results from multiple transfections of each promoter/enhancer construct are expressed as the percent conversion of [14C]chloramphenicol to its major acetylated product. The murine LTR (pAKR-CAT), the SV40 early gene promoter (pSV2-CAT), and the immediate early gene promoters from cyto-
732
HUPER ET AL. TABLE 1 RELATIVE PROMOTER STRENGTH IN HMEC AND MURINE FIBROBLAST CAT Activity~
CAT Constructsb
HMEC, Sr
L929. Ca
pCAT pSV2-CAT 10 gg 20 gg pAKR-CAT 5 #g 10 #g 20/.tg pIE-CAT pRP-CAT pTK-CAT plCPO-CAT pMMTV-CAT~ Complete -dex +dex pMMT-CATd Complete +Zn +2
RELATIVE PROMOTER ACTIVITY IN HUMAN MAMMARY EPITHELIAL CELLS ASSAYED BY TRANSIENT EXPRESSION GUDRUN HUPER, JEFFREY R. MARKS, JON R. WIENER, AND J. DIRK IGLEHART1
Department of Surgery, Duke UniversityMedical Center, Box 3873, Durham, North Carolina 27710 (Received 4 March 1992; accepted 13 May 1992)
SUMMARY Chimeric DNA expression vectors containing regulatory sequences proximal to the 5' end of coding sequences for mammalian genes provide valuable tools to study gene expression. Genes coding for easily measured products (reporter genes) can be used to study promoter strength and regulation of gene expression after transient expression of promoter-reporter constructs in mammalian cells. To determine the strength of a variety of mammalian and viral promoter-enhancer sequences in primary cuhures of human mammary epithelial cells (HMEC), these sequences were fused to the bacterial chloramphenicol acetyltransferase (CAT) gene and transfected into HMEC using strontium phosphate. The long terminal repeat (LTR) of the endogenous murine leukemia virus AKR-623 was the most potent promoter of transient CAT expression in HMEC. A number of commonly available promoter sequences displayed a wide range of activities in these cells. The glucocorticoid responsive LTR promoter from the murine mammary tumor virus modulated expression of CAT and was sensitive to the concentration of dexamethasone in the growth media. In a similar fashion, the regulatory sequences from the murine metallothionein-1 gene retained responsiveness to zinc concentration in the growth media.
Key words: human mammary epithelial cells; transfection; promoter; CAT assay. and human origin (1,11,24). The use of extracellular matrix proteins as a substratum results in the appearance of differentiated characteristics such as casein and milk fat protein, duct formation, and myoepithelial cells (6,13,23). These results demonstrate that coordinated expression of normal cellular genes and complex functions are possible in cultured mammary epithelial cells. Transient expression systems can be utilized to investigate the control of gene expression in normal cells and, perhaps, to investigate regulatory defects in breast malignancy. The current study was undertaken to assess the relative strengths of a series of widely used regulatory sequences in primary cultures of human mammary epithelial cells (HMEC). The goal of these experiments was to identify promoter-enhancer elements for eventual use in stable transfections of putative oncogenic sequences. Strontium phosphate-mediated transfection was used to introduce standardized vector constructs containing the CAT reporter gene into HMEC. The ability of inducible promoter elements to function in normal mammary cells was also studied. A considerable heterogeneity of promoter strength was found which appears to be cell-type specific. In addition, two inducible regulatory elements were found to function in a dose-dependent manner.
INTRODUCTION The introduction of cloned genetic material into cells growing in culture is a valuable means to study the function and regulation of gene expression. Calcium phosphate mediated transfection is a popular technique which can be applied to a wide variety of cell types. Because calcium induces a wide-range of physiologic effects in certain human epithelial cell cultures, including differentiation and growth inhibition, an alternative method of gcne transfer into epithehal cells using strontium phosphate precipitates has been developed. Strontium phosphate mediated transfeetions are efficient and lead to both transient expression and stable genomie integration with a minimum of deleterious effects on cultured epithelial cells (2). Recombinant DNA technology allows the study of gene regulatory sequences separated from the coding sequences that they control. The bacterial gene coding for chloramphenicol acetyltransferase (CAT) is a common reporter gene used in chimeric constructs to study gene expression in a variety of mammalian cell lines (9,10). This simplified system can be used to dissect complex regulation of gene expression by placing the CAT coding sequences under the control of promoter and enhancer elements that are able to function in the target cell type chosen for study. Furthermore, stable transfection and expression of other genes may be optimized by choosing the promoter sequence with the best transient activity in the target cell (9). Culture conditions have been described that will support the short-term proliferation of mammary epithelial cells from mouse
MATERIALSAND METHODS
Construction of CATplasmids. All constructs contained the CAT (Tn9) gene with its constituent SV40 early region intron and polyadenylation sequences (10). pSV2-CAT contains the entire SV40 early promoter/enhancer region previously described (10). The followingplasmids were constructed starting with an enhancer and promoter-less CAT containing plasmid designated pCAT. pCAT is the HindlII-BamHI fragment of pSV2-CAT,
1 To whom correspondence should be addressed. 730
731
CAT GENE EXPRESSION IN BREAST CELLS
piECAT
pAKRCAT
pRPCAT
pSV2 CAT -
PTKCAT
pMMTVplCPOCAT CAT
Fit. l. Autoradiograph exhibiting conversion of [~4C]ehloramphenicol to acetylated products by CAT activity in transfected HMEC. Aeetylated products migrate further into the gel and appear above the unmodified chloramphenicol(bottom). Corresponding regions of the gel were cut out and radioactivity counted to calculate the percent conversion (see Table 1). In this experiment, duplicate transfections are run in parallel lanes. Bottom, paired lanes are labeled with the plasmid construct used for transfection. In each case, 20/xg of plasmid DNA was transfeeted per plate.
converted to blunt ends with Klenow was cloned into the EcoRV site of Bluescript KS+ (Stratagene, LaJolla, CA). pIE-CAT contains the murine cytomegalovirus immediate early enhancer/promoter on a 1.2 kb Hpal: PstI segment from an EcoRI fragment of virion DNA (19) cloned into pCAT. pTK-CAT was constructed from the HSV-1 thymidine kinase (tk) gene promoter by cloning the 250 bp Pvull-Bglll fragment upstream of the tk gene into the polylinker of peAT (18). plCPO-CAT contains the EeoRI fragment of plGA65, from the promoter sequences of the infected-cell polypeptide-0 (ICP0) gene of HSV-1 (8), cloned into pCAT and obtained from Dr. Mark Labow (14). pRP-CAT contains the murine ribosomal protein L32 promoter (3) cloned into the Bluescript polylinker upstream of peAT. This promoter region is contained on a 1.1-kb fragment extending to the StuI site in intron 1. pAKR-CAT contains the long terminal repeat (LTR) from the endogenous murine retrovirus AKR-623 (15) inserted upstream of the CAT gene in pCAT (from J. Horowitz, Duke University). pMMT-CAT contains the mouse metallothionein promoter contained on a 1.9-kb EcoRI-BGI II fragment from p341-3 (5) cloned into peAT. pMMTV-CAT, called pMK-2 in the description of its construction, contains the murlne mammary tumor virus LTR and enhancer elements found in the LTR of the Harvey sarcoma virus, both upstream of the CAT gene in pCAT [obtained from M. Ostrowski, Duke University (21)].
Cell culture and transfection of normal human mammary epithelial cells. Human mammary tissue from reduction mammoplasties were digested enzymatically to epithelial organoids exactly as described (25). Cultures were established in DFCI-1 growth media that contains alpha minimal essential media (otMEM) and Ham's F12 (1:1, vol/vol) and supplemented as described (1). These cells seem epithelial, express human epithelial cytokeratins, and are diploid. Murine strain L cells, clone 929, were maintained in Dulbecco's MEM with 5% fetal bovine serum. L cells were obtained from the American Type Culture Collection, Rockville, MD. Growth and incubations were done in a 5% CO2 humidified incubator. Plasmid DNA was grown in Escherichia coli JMI09 and purified by banding in cesium chloride gradients (16). HMEC were plated in 60-ram plastic cell culture dishes at 2 × 10 s cells per dish and used for transfection 24 h later. Strontium phospfiate transfections were done as described by Brash et al. (2). Precipitates of strontium phosphate were monitored by microscopy and added at the stage of fine dust to culture dishes containing DFCI-1 media. Dishes were incubated for 4 h at 37 ° C, washed twice with phosphate buffered saline, treated for 30 s with 11.8% (vol/vol)glycerol in phosphate buffered saline (PBS), washed 3 times with PBS, and refed with complete DFCI-1 medium. Calcium phosphate transfections were done by parallel techniques exemplified by Gorman et al. (10). Unless indicated,
each plasmid DNA was transfected onto three separate dishes in each experiment to monitor plate-to-plate variability: CAT assays. Forty-eight hours after transfection, cells were dispersed, washed in PBS, and cell pellets resuspended in 100 #1 of 0.25 M Tris-HC1 (pH 7.8). The cells were lysed by three freeze-thaw Cycles and clarified by centrifugation at 12 000 × g for 2 min. Tile CAT assay reaction mixture contained 50 #1 of the cell extract; 75 #1 of 1 M Tris-HC1 (pH 7.8) containing 0.025 #Ci of [14C]chloramphenieol (Amersham, Arlington Heights IL). The reaction was initiated by addition of 25 #1 of 100 #g acetyl coenzyme A in 1 M tris, pH 7.8. The reaction mixtures were incubated at 37 ° C for 90 min and stopped by the addition of 0.7 ml of ethyl acetate and vigorous vortexing. The organic layer was dried, resuspended in 20 #1 of ethyl acetate, and spotted at the origin of thin-layer sihca gels. The samples were chromatographed in a mixture of chloroform-methanol (95:5), and the reaction products located by autoradiography. Spots were cut out and counted in scintillation fluid, Data were expressed as the percent of chloramphenicol converted to its major acetylated form pe r microgram of protein in 1 h. Percent conversion for each cell line represents a minimum of three separate transfections and is expressed as the mean plus or minus the standard deviation. Protein concentration was determined using commercial kits (Bio-Rad Protein Assay, Bio-Rad Laboratories, Richmond, CA). In each set of experiments or at each concentration of media additive s, pSV2-CAT was included in three separate dishes to monitor batch-to-batch conditions of the cell cultures and transfection efficieneies. RESULTS
Promoter strength in HMEC and murine fibroblasts.
To determine the relative strength of the various promoter constructs to transiently express CAT, plasmid DNA was transfected into HMEC using the strontium phosphate method. With certain promoter constructs, varying amounts of plasmid DNA were transfected to establish a dose-response relationship. After 48 h, cells were lysed and CAT activity determined. Figure 1 displays qualitative results from a representative series of transfections analyzed by autoradiography. In Table 1, quantitative results from multiple transfections of each promoter/enhancer construct are expressed as the percent conversion of [14C]chloramphenicol to its major acetylated product. The murine LTR (pAKR-CAT), the SV40 early gene promoter (pSV2-CAT), and the immediate early gene promoters from cyto-
732
HUPER ET AL. TABLE 1 RELATIVE PROMOTER STRENGTH IN HMEC AND MURINE FIBROBLAST CAT Activity~
CAT Constructsb
HMEC, Sr
L929. Ca
pCAT pSV2-CAT 10 gg 20 gg pAKR-CAT 5 #g 10 #g 20/.tg pIE-CAT pRP-CAT pTK-CAT plCPO-CAT pMMTV-CAT~ Complete -dex +dex pMMT-CATd Complete +Zn +2
