147
Molecular and Cellular Endocrinology, 87 (1992) 147-156
0 1992 Elsevier Scientific Publishers Ireland, Ltd. 0303-7207/92/$05.00 MOLCEL 02817
A distal region enhances the prolactin induced promoter activity of the rabbit cusI-casein gene S. Pierre, G. Jolivet, E. Devinoy, M.C. Thkon, R. Mali&ou-N’Gassa, and L.M. Houdebine
C. Puissant
Unite de Diff~re~ciation Ce~iulaire, i~~tjt~t rational de la Recherche ~g~o?lo~~q~e, Jou~-en-Jesus, France
(Received 13 March 1992; accepted 5 June 1992)
Key words: Organoids: Transfection; Transient expression; CHO cells; HCll cells; Chloramphenicol
acetyl transferase
Casein gene expression is induced in the rabbit mammary gland by prolactin (PRL). asl-casein is the major casein secreted into milk. In order to define the position of the DNA sequences involved in the control of rabbit cusl-casein gene regulation by PRL, chimeric genes were constructed between upstream regions of the rabbit asl-casein gene and the chloramphenicol acetyl transferase (CAT) reporter gene. A series of 5’-deleted fusion genes was obtained by nuclease digestion of the cwsl-casein gene upstream region. These gene constructs were transfected into rabbit primary mammary cells, or cotransfected in CHO cells with the plasmid coding for the rabbit mammary receptor (PRL-R). A regulatory region has been located between nt -3768 and -3155. This region enhances the prolactin induced promoter activity of the cusl-casein gene. It might possess or cooperate with prolactin responsive elements located further downstream in the crsl-casein gene.
Introduction
The mammal gland represents an attractive biological system for studying the effects of both polypeptide (prolactin (PRL) and insulin (I)) and steroid hormones (glucocorticoids and progesterone) on the regulation of gene expression. Indeed, during lactation, these lactogenic hormones stimulate the synthesis of milk proteins in a coordinated manner. Among these proteins, caseins are the most abundantly synthesized in almost all species. Their role is fundamental for
Correspondence to: G. Jo&vet, Unite de Diff~renciati~n Celluiaire, lnstitut National de la Recherche Agronomique, 78352 Jouy-en-Josas Cedex, France. Tel. (33-I) 34.65.25.44; Fax (33-l) 34.65.22.13.
the newborn, since they represent an important source of proteins, calcium and phosphorus. They form a small protein gene family, including asl, tys2, p, and K caseins in the ovine (Leveziel et al., 1991). During lactation, milk protein genes are abundantly expressed. The level of casein mRNA increases more than 100-fold comparing virgin and lactating female rats or rabbits (Shuster et al., 1976; Hobbs et al., 1982). This increase results from both the stimulation of the gene transcription and the stabilization of mRNAs. All casein genes are under the same complex control by lactogenic hormones. Several studies have been performed to investigate the mechanism of action of Iactogenic hormones on the regulation of casein gene expression. In these studies, regulatory sequences have been searched in the 5’ flanking
148
region of casein genes from several species including the bovine (Meade et al., 1990; Schmidhauser et al., 19901, the mouse (Yoshimura and Oka, 1990), the rabbit (Buhler et al., 19901, and the rat (Yu-Lee and Rosen, 1988; Lee et al., 1989; Schmitt-Ney et al., 1991). Experiments were performed using whole casein genes (Lee et al., 1989) or fusion genes containing a casein gene fragment linked to different reporter genes. These reporter genes were genes encoding the bacterial chloramphenicol acetyl transferase (CAT) or genes encoding proteins of pharmaceutical interest as human urokinase or interleukin-2. These constructs have been microinjected into mouse or rabbit embryos to raise transgenic animals (Lee et al., 1989; Buhler et al., 1990; Meade et al., 1990), or transfected into cultured cells (Lee et al., 1989; Schmidhauser et al., 1990; Yoshimura and Oka, 1990). These studies have shown the presence of elements responsible for tissue specificity and for PRL sensitivity, in the proximal 5’ flanking region of p-casein genes, within 500 nt upstream from the cap site. The binding of nuclear transcription factors to two proximal regions of the rat p-casein gene has also been described (Schmitt-Ney et al., 1991). These factors are specifically expressed in the mammary gland cells during lactation. Unlike studies concerning p-casein gene regulation, only a few experiments have been performed to identify the regulatory elements of the asl-casein gene. cusl-casein is the most abundant protein secreted in rabbit milk (Grabowsky et al., 1991). The rabbit asl-casein gene has been cloned and sequenced in our laboratory (Jolivet et al., 1992). We observed that the proximal 5’ flanking region, the first intron, and the 5’ non-translated region of casein genes are highly conserved in several mammalian species (Devinoy et al., 1988). In the present experiments, DNA fragments containing the upstream region of the rabbit asl-casein gene located from position - 3768 upstream from the cap site, down to position + 1774 in the second exon, including most of the 5’ untranslated region of the asl-casein mRNA, were linked to the bacterial CAT gene and transfected into rabbit mammary cells derived from primary cultures or cotransfected with the rabbit mammary prolactin receptor (R-PRL) cDNA into
CHO cells. This last model has been previously described and used successfully to study the PRL regulation of another milk protein gene, the ovine @lactoglobulin gene (Lesueur et al., 1990, 1991). Using both cellular systems, a distal 5’ element, located between nt - 3768 and - 3155, was shown to be important for the strong induction of the rabbit cusl-casein gene expression by prolactin. Materials
and methods
Cell cultures, transfections and hormonal inductions Rabbit primary mammary cells were prepared from the mammary glands of lbday pregnant rabbits and cultured on rat tail collagen gels for 4 days. They were then transfected using Lipofectin (Gibco BRL), and submitted to hormonal treatments as previously described (Devinoy et al., 1991). In all cases, the medium contained insulin, 5 kg/ml, plus cortisol, 1.5 PM, in the presence or absence of oPRL 40 nM (+ PRL) (- PRLI. Cells were harvested during transient expression of the transfected genes as previously described (Devinoy et al., 1991). Each hormonal induction was carried out on three separate dishes. The protein content of cell extracts obtained from the individual plates was assayed (Bradford, 1976). CHO-Kl cells were grown as previously described (Lesueur et al., 1991). One day before transfection and during the whole transient expression period, cells were maintained in a medium containing insulin (I, 3 pug/ml) but deprived of fetal calf serum (modified GC3 medium) (Lesueur et al., 1991). Cells were transfccted by the calcium phosphate precipitation procedure with 3 pg of pCHl10 (@galactosidase expression vector, Pharmacia), 3 pg of PER,-3 (rabbit prolactin receptor (PRL-R) expression vector, containing the cDNA of the rabbit PRL-R under the transcriptional control of the SV40 early promoter (Edery et al., 1989)), 1.5 pg of paccaslcasein)CAT or control plasmids (pSVE-CAT) (Gorman et al., 1982), pBLCAT2 (Luckow and Schiitz, 1987) and 2.5 pg of carrier plasmid DNA. In order to minimize the differences in the reporter gene expression resulting from the variability in DNA precipitation, half of each precipi-
149
tate was added to a 60 mm culture dish which will not be treated with prolactin on the following day and the second half was added to a prolactin stimulated dish. The following day, fresh modified GC3 medium containing cortisol (6 X low8 M, control cells) or cortisol and ovine prolactin (16 nM, oPRL) was added (Lesueur et al., 1991). Each hormonal induction was carried out in two separate dishes. Forty-eight hours after addition of hormones, cells were scraped and lysed by repeated freezing-thawing cycles in 200 ~1 Tris 250 mM pH 8.0. After 10 min centrifugation at 15,000 x g, the supernatant was collected. CA T and P-galactosidase assays CHO extracts were assayed for CAT activity using a mixed phase assay (Nielsen et al., 1989). The complete reaction mixture (25 ~1 final volume) contained 15 ~1 of cell extract and 0.1 PCi [ “Hlacetyl-coenzyme A (5 Ci/mM, Amersham), in the presence of 30 PM acetyl-coenzyme A, 1 PM chloramphenicol, 4 mM EDTA and 200 mM Tris pH 7.8. Incubation was performed for 2 h at 37°C. The reaction was terminated by mixing the reaction mixture with 250 ~1 urea (in water) in a 5-ml scintillation vial. Scintillation cocktail (2.5 ml) was added and radioactivity was determined in a scintillation counter as described (Nielsen et al., 1989). Each sample was assayed in duplicate as a complete reaction, in the presence of chloramphenicol. In one sample representative of each transfection experiment on a cell density basis, radioactivity was also measured in the presence of cell extract but in the absence of chloramphenicol (extract blank). Specific CAT activity in the sample was determined by subtracting the value obtained for the extract blank from that obtained for complete reaction. Data from the CAT assay were corrected for transfection-induced variation by referring to P-galactosidase enzyme activity (Herbomel et al., 1984). For each experiment, a standard curve was obtained using increasing amounts of purified bacterial CAT (Boehringer-Mannheim) diluted in 800 Kg bovine serum albumin/ml water. The linearity of each standard curve was checked over a range of CAT activities spanning from 10-j to 1.5 X lop3 units. Experimental data are expressed as enzyme units per dish.
Primary cells extracts were assayed for CAT activity using the [‘4Clchloramphenicol thin-layer chromatography assay (German et al., 1982) as previously described (Devinoy et al., 1991). The quantitation was made by scintillation counting of radioactive spots from the thin layer. Values are expressed as percent conversion of chloramphenico1 per 100 pg of cellular proteins. Northern blot analysis Messenger RNAs for rabbit asl-casein were identified in primary cell cultures by Northern blot analysis performed on total cellular RNA extracted and blotted as previously described (Devinoy et al., 1991). A PstI/PstI fragment excised from a cDNA clone of rabbit asl-casein (Devinoy et al., 1988) and a fragment containing the mouse p-actin cDNA (Alanso et al., 1986) were used as probes. Construction of asl-casein gene CAT plasmid A 5.5-kilobase (kb) PstI/AuaII fragment of the rabbit asl-casein gene was obtained from a A Charon 4A clone (Devinoy et al., 1988). It consisted of 3.7 kb of 5’ flanking sequence, the first non-coding exon (49 nucleotides (nt)), the first intron (1719 nt> and 5 nt of the second exon until an AcaII restriction site (Fig. 5). The initiation of translation of TVE asl-casein gene (ATG codon) was not included, the AvaII restriction site being located 8 nt upstream from it. This Pst I/AcaII fragment possessed an internal Pst I site and was thus subcloned in pPolyII1 (Lathe et al., 1987) in two steps. A first step consisted in subcloning a PstI/AcaII fragment (1012 nt) into the SmaI/ PstI sites of pPolyII1 to create the pH1 plasmid. A second step consisted in subcloning the Pst I/ Pst I fragment (4.5 kb) into the PstI site of pH1, in the same orientation as the already subcloned asl-casein gene fragment to create the plasmid pHla. CAT gene (containing the initiation site of translation) and the polyadenylation signal of SV40 were excised from pD23 (Mohun et al., 1987) after digesting the plasmid with BamHI restriction enzyme. The BamHI fragment containing CAT/SV40 sequences was subcloned into the BamHI site of pHla in the same orientation as the already subcloned cusl-casein gene fragment to create the plasmid pac(- 3768/
+ 1774)CAT. Figures in parentheses indicate the 5’ and 3’ borders of cysl-casein gene sequences according to the previously established numbering (Jolivet et al., 1992). The 5’-deletions were introduced after subcloning a Not1 fragment excised from pac(- 3768,’ + 1774)CAT containing asl-casein and CAT sequences into the Not I site of the pBlueto create pBacscript (SK + , Stratagene) ( - 3768,’ + 17741CAT. Using the exonuclease III/mung bean nuclease technique (Stratagene), various 5’ deletions were created starting from the XbaI site of the Bluescript polylinker. Bluescript sequences were protected by cleaving the Sal1 site of the polylinker and deoxythioderivative filling-in of this Sal1 site by Klenow DNA polymerase. Deletion end points were determined by double-strand DNA sequencing using USB DNA sequencing kit with Sequenase (version 2.0). Three plasmids: pBac( -3155,’ + 1774)CAT; pBac( - 1971,’ + 17741CAT; pBac( - 387,’ + 1774)CAT, were created by gradually deleting the 5’ end of the cusl-casein gene fragment. The X/z01 fragments containing C-3155,’ + 1774)CAT and ( - 387,’ + 1774)CAT sequences were excised from the respective pBac(aslcasein)CAT and reinserted into the Xi?01 site of the pPolyII1 polylinker to create pac( - 3155,’ + 1774)CAT and pact - 387/ + 1774)CAT. The containing (- 1971/ fragment Kpn I /Sac I + 1774)CAT sequences was excised from pBac( - 1971,’ + 1774)CAT and reinserted into the KpnI/SacI sites of the pPolyII1 polylinker to create pac( - 1971,’ + 1774)CAT. For the construction of pac( - 51/ + 1774)CAT encompassing the TATA box sequence, a defined fragment of the asl-casein gene c-51, + 416) was synthesized by the polymerase chain reaction. The plasmid pBac( - 3768,’ + 1774)CAT was used as template. The primers contained a SmaI site at their 5’ ends. The sense primer was 5’TAACCCGGGAGCATTTTACTGATCACTGG3’ and the anti-sense was 5’TAACCCGGGCTACCAAAACAAGCCTTGCT3’. The chain reaction contained 1 ng of template in 100 ~1 of 50 mM KCI, 10 mM Tris HCI pH 9.0, 1.5 mM MgCl,, 0.01% gelatin w/v, 0.1% Triton X-100, 0.2 mM dNTP, 1 mM of each primer and 2.5 U of Taq
DNA polymerase (Promega). A cycle consisted of 30 s at 95°C 30 s at 50°C and 1 min at 72°C. The PCR reaction was run for 27 cycles. The chain reaction fragment was ethanol-precipitated then dissolved in Tris 10 mM pH 8.0, EDTA 1 mM. It was then digested with SmaI and NsiI generating two SmaI/Nsi! fragments, c-51,’ + 180) and (+ 180,’ + 416). The SmaI/NsiI fragment ( - 51,’ + 1801 containing 51 nt of 5’ flanking sequences, the first exon and 131 nt of the first intron was introduced into a XbaI/NsiI fragment excised from pac( - 1971,’ + 1774)CAT, containing (+ 181,’ + 1774ICAT sequences and the whole pPolyII1 sequences after filling the XbaI site with the T4 DNA polymerase. The pBLCAT2 vector (Luckow and Schiitz, 19871 containing the CAT gene and SV40 polyadenylation sequences under control of the herpes simplex virus tk promoter was used to test the activity of the fragment c-3768/ - 3155) of the cwsl-casein gene. An AccI fragment (- 3768,’ - 31131 was excised from pBac( - 3768,’ + 1774)CAT and reintroduced in the same orientation as the tkCAT sequences in the BarnHI site of the polylinker of pBLCAT2 after filling-in the AccI and BamHI sites by T4 DNA polymerase to create p( - 3768,’ - 31551-tk-CAT. Results Induction of asI-casein gene promoter acticity by oPRL. in transfected rabbit primary mammary cells In rabbit primary mammary cells, grown on collagen gels for 4 days, expression of the endogenous cusl-casein gene was induced by oPRL in the presence of insulin and cortisol which amplify prolactin action. cusl-casein mRNA was undetectable when cells were cultured in the presence of insulin and cortisol but in the absence of oPRL (Fig. 1). Transcription was induced after addition of oPRL, whereas the level of p-actin mRNA remained unchanged. These rabbit primary mammary ceils, in which the endogenous cusl-casein gene expression can be induced by PRL, were thus used in our experiments in order to define the presence of PRL responsive regions in cwsl-casein-CAT chimeric genes. For this purpose, rabbit primary mammary cells cultured on collagen gels for 4 days were trans-
151
fected with different chimeric asl-casein-CAT constructs (pac(x/1774)CAT constructs). By the end of transfection, oPRL was introduced into the culture medium and cells were harvested 72 h later, during transient expression of the CAT gene. CAT activity was estimated in cell extracts. When pac( - 3768/ + 1774)CAT construct was transfected into these cells, the percentage of [ “C]chloramphenicol which became acetylated was 14-fold higher after oPRL stimulation (Fig. 2A). This increase was not observed when pOCAT, a construct deprived of asl-casein gene sequence and of any promoter was transfected into these cells. Sequences located between -3768 and + 1774 in the cwsl-casein gene sequence thus allow oPRL to induce CAT gene expression. This induced CAT gene expression could be due to general effects of PRL on the mammary cells and observed even when the CAT gene was transcribed from a prolactin insensitive promoter. SV40 promoter is not sensitive to lactogenic hormones. When pSVECAT, a construct containing the CAT gene under control of this viral promoter, was transfected into these cells, prolactin stimulation of CAT gene expression was not observed (Fig. 2B). CAT gene expression is thus regulated by oPRL, only when upstream sequences of the asl-casein gene, located between nt - 3768 and + 1774, are present in the transfected CAT construct. The 5’ end of the upstream fragment (- 3768,’ + 1774) was progressively deleted using nucleases. For this purpose the c-3768/ + 1774)CAT fragment was cloned into a pBluescript vector, giving rise to the pBac(-3768/ + 1774)CAT plasmid. The 5’ end of the upstream fragment c-3768/ + 1774) was deleted generating the pBac( - 3155/ + 1774)CAT plasmid. Both plasmids pBac(- 3768/ + 1774)CAT and pBacc-3155/ + 1774)CAT were transfected into rabbit primary mammary cells. As previously observed for cells transfected with pac(-3768/ + 1774)CAT, oPRL was able to induce a marked increase in CAT gene expression (6.3-fold, Fig. 2B), when cells were transfected with pBac( - 3768/ + 1774)CAT. After transfection with the pBac(-3155/ + 1774)CAT construct, CAT gene expression was no longer significantly stimu-
lated by the oPRL treatment (Fig. 2B, 1.6fold induction). The 5’ deletion of a 612 nt fragment between located position - 3768 and - 3155 thus reduced oPRL stimulation of CAT gene transcription to a non-significant level. This fragment therefore contributes to the strong induction of the asl-casein gene expression by PRL, in rabbit primary mammary cells grown on a collagen matrix in the presence of insulin and cortisol. PRL responsive elements, or regulatory elements which might interact with PRL responsive elements located further downstream on the asl-casein gene, may be located in this distal region. However, complex hormonal and culture conditions are required to fully induce the casein gene expression by prolactin in cultured primary mammary cells (Emerman and Pitelka, 1977) as well as to induce the maximal CAT gene expression of transfected asl-casein gene construct. The 612 nt fragment might thus also carry mammary specific regulatory regions, matrix dependent regulatory elements, insulin or cortisol responsive elements. In an attempt to define the nature of the elements present in the 612 nt fragment, experiments were pursued using a cellular system which is matrix and insulin independent, the CHO cells.
OPRL A.
alphas1 casein >
-
OPRL +
B.
-
+
c)
Molecular and Cellular Endocrinology, 87 (1992) 147-156
0 1992 Elsevier Scientific Publishers Ireland, Ltd. 0303-7207/92/$05.00 MOLCEL 02817
A distal region enhances the prolactin induced promoter activity of the rabbit cusI-casein gene S. Pierre, G. Jolivet, E. Devinoy, M.C. Thkon, R. Mali&ou-N’Gassa, and L.M. Houdebine
C. Puissant
Unite de Diff~re~ciation Ce~iulaire, i~~tjt~t rational de la Recherche ~g~o?lo~~q~e, Jou~-en-Jesus, France
(Received 13 March 1992; accepted 5 June 1992)
Key words: Organoids: Transfection; Transient expression; CHO cells; HCll cells; Chloramphenicol
acetyl transferase
Casein gene expression is induced in the rabbit mammary gland by prolactin (PRL). asl-casein is the major casein secreted into milk. In order to define the position of the DNA sequences involved in the control of rabbit cusl-casein gene regulation by PRL, chimeric genes were constructed between upstream regions of the rabbit asl-casein gene and the chloramphenicol acetyl transferase (CAT) reporter gene. A series of 5’-deleted fusion genes was obtained by nuclease digestion of the cwsl-casein gene upstream region. These gene constructs were transfected into rabbit primary mammary cells, or cotransfected in CHO cells with the plasmid coding for the rabbit mammary receptor (PRL-R). A regulatory region has been located between nt -3768 and -3155. This region enhances the prolactin induced promoter activity of the cusl-casein gene. It might possess or cooperate with prolactin responsive elements located further downstream in the crsl-casein gene.
Introduction
The mammal gland represents an attractive biological system for studying the effects of both polypeptide (prolactin (PRL) and insulin (I)) and steroid hormones (glucocorticoids and progesterone) on the regulation of gene expression. Indeed, during lactation, these lactogenic hormones stimulate the synthesis of milk proteins in a coordinated manner. Among these proteins, caseins are the most abundantly synthesized in almost all species. Their role is fundamental for
Correspondence to: G. Jo&vet, Unite de Diff~renciati~n Celluiaire, lnstitut National de la Recherche Agronomique, 78352 Jouy-en-Josas Cedex, France. Tel. (33-I) 34.65.25.44; Fax (33-l) 34.65.22.13.
the newborn, since they represent an important source of proteins, calcium and phosphorus. They form a small protein gene family, including asl, tys2, p, and K caseins in the ovine (Leveziel et al., 1991). During lactation, milk protein genes are abundantly expressed. The level of casein mRNA increases more than 100-fold comparing virgin and lactating female rats or rabbits (Shuster et al., 1976; Hobbs et al., 1982). This increase results from both the stimulation of the gene transcription and the stabilization of mRNAs. All casein genes are under the same complex control by lactogenic hormones. Several studies have been performed to investigate the mechanism of action of Iactogenic hormones on the regulation of casein gene expression. In these studies, regulatory sequences have been searched in the 5’ flanking
148
region of casein genes from several species including the bovine (Meade et al., 1990; Schmidhauser et al., 19901, the mouse (Yoshimura and Oka, 1990), the rabbit (Buhler et al., 19901, and the rat (Yu-Lee and Rosen, 1988; Lee et al., 1989; Schmitt-Ney et al., 1991). Experiments were performed using whole casein genes (Lee et al., 1989) or fusion genes containing a casein gene fragment linked to different reporter genes. These reporter genes were genes encoding the bacterial chloramphenicol acetyl transferase (CAT) or genes encoding proteins of pharmaceutical interest as human urokinase or interleukin-2. These constructs have been microinjected into mouse or rabbit embryos to raise transgenic animals (Lee et al., 1989; Buhler et al., 1990; Meade et al., 1990), or transfected into cultured cells (Lee et al., 1989; Schmidhauser et al., 1990; Yoshimura and Oka, 1990). These studies have shown the presence of elements responsible for tissue specificity and for PRL sensitivity, in the proximal 5’ flanking region of p-casein genes, within 500 nt upstream from the cap site. The binding of nuclear transcription factors to two proximal regions of the rat p-casein gene has also been described (Schmitt-Ney et al., 1991). These factors are specifically expressed in the mammary gland cells during lactation. Unlike studies concerning p-casein gene regulation, only a few experiments have been performed to identify the regulatory elements of the asl-casein gene. cusl-casein is the most abundant protein secreted in rabbit milk (Grabowsky et al., 1991). The rabbit asl-casein gene has been cloned and sequenced in our laboratory (Jolivet et al., 1992). We observed that the proximal 5’ flanking region, the first intron, and the 5’ non-translated region of casein genes are highly conserved in several mammalian species (Devinoy et al., 1988). In the present experiments, DNA fragments containing the upstream region of the rabbit asl-casein gene located from position - 3768 upstream from the cap site, down to position + 1774 in the second exon, including most of the 5’ untranslated region of the asl-casein mRNA, were linked to the bacterial CAT gene and transfected into rabbit mammary cells derived from primary cultures or cotransfected with the rabbit mammary prolactin receptor (R-PRL) cDNA into
CHO cells. This last model has been previously described and used successfully to study the PRL regulation of another milk protein gene, the ovine @lactoglobulin gene (Lesueur et al., 1990, 1991). Using both cellular systems, a distal 5’ element, located between nt - 3768 and - 3155, was shown to be important for the strong induction of the rabbit cusl-casein gene expression by prolactin. Materials
and methods
Cell cultures, transfections and hormonal inductions Rabbit primary mammary cells were prepared from the mammary glands of lbday pregnant rabbits and cultured on rat tail collagen gels for 4 days. They were then transfected using Lipofectin (Gibco BRL), and submitted to hormonal treatments as previously described (Devinoy et al., 1991). In all cases, the medium contained insulin, 5 kg/ml, plus cortisol, 1.5 PM, in the presence or absence of oPRL 40 nM (+ PRL) (- PRLI. Cells were harvested during transient expression of the transfected genes as previously described (Devinoy et al., 1991). Each hormonal induction was carried out on three separate dishes. The protein content of cell extracts obtained from the individual plates was assayed (Bradford, 1976). CHO-Kl cells were grown as previously described (Lesueur et al., 1991). One day before transfection and during the whole transient expression period, cells were maintained in a medium containing insulin (I, 3 pug/ml) but deprived of fetal calf serum (modified GC3 medium) (Lesueur et al., 1991). Cells were transfccted by the calcium phosphate precipitation procedure with 3 pg of pCHl10 (@galactosidase expression vector, Pharmacia), 3 pg of PER,-3 (rabbit prolactin receptor (PRL-R) expression vector, containing the cDNA of the rabbit PRL-R under the transcriptional control of the SV40 early promoter (Edery et al., 1989)), 1.5 pg of paccaslcasein)CAT or control plasmids (pSVE-CAT) (Gorman et al., 1982), pBLCAT2 (Luckow and Schiitz, 1987) and 2.5 pg of carrier plasmid DNA. In order to minimize the differences in the reporter gene expression resulting from the variability in DNA precipitation, half of each precipi-
149
tate was added to a 60 mm culture dish which will not be treated with prolactin on the following day and the second half was added to a prolactin stimulated dish. The following day, fresh modified GC3 medium containing cortisol (6 X low8 M, control cells) or cortisol and ovine prolactin (16 nM, oPRL) was added (Lesueur et al., 1991). Each hormonal induction was carried out in two separate dishes. Forty-eight hours after addition of hormones, cells were scraped and lysed by repeated freezing-thawing cycles in 200 ~1 Tris 250 mM pH 8.0. After 10 min centrifugation at 15,000 x g, the supernatant was collected. CA T and P-galactosidase assays CHO extracts were assayed for CAT activity using a mixed phase assay (Nielsen et al., 1989). The complete reaction mixture (25 ~1 final volume) contained 15 ~1 of cell extract and 0.1 PCi [ “Hlacetyl-coenzyme A (5 Ci/mM, Amersham), in the presence of 30 PM acetyl-coenzyme A, 1 PM chloramphenicol, 4 mM EDTA and 200 mM Tris pH 7.8. Incubation was performed for 2 h at 37°C. The reaction was terminated by mixing the reaction mixture with 250 ~1 urea (in water) in a 5-ml scintillation vial. Scintillation cocktail (2.5 ml) was added and radioactivity was determined in a scintillation counter as described (Nielsen et al., 1989). Each sample was assayed in duplicate as a complete reaction, in the presence of chloramphenicol. In one sample representative of each transfection experiment on a cell density basis, radioactivity was also measured in the presence of cell extract but in the absence of chloramphenicol (extract blank). Specific CAT activity in the sample was determined by subtracting the value obtained for the extract blank from that obtained for complete reaction. Data from the CAT assay were corrected for transfection-induced variation by referring to P-galactosidase enzyme activity (Herbomel et al., 1984). For each experiment, a standard curve was obtained using increasing amounts of purified bacterial CAT (Boehringer-Mannheim) diluted in 800 Kg bovine serum albumin/ml water. The linearity of each standard curve was checked over a range of CAT activities spanning from 10-j to 1.5 X lop3 units. Experimental data are expressed as enzyme units per dish.
Primary cells extracts were assayed for CAT activity using the [‘4Clchloramphenicol thin-layer chromatography assay (German et al., 1982) as previously described (Devinoy et al., 1991). The quantitation was made by scintillation counting of radioactive spots from the thin layer. Values are expressed as percent conversion of chloramphenico1 per 100 pg of cellular proteins. Northern blot analysis Messenger RNAs for rabbit asl-casein were identified in primary cell cultures by Northern blot analysis performed on total cellular RNA extracted and blotted as previously described (Devinoy et al., 1991). A PstI/PstI fragment excised from a cDNA clone of rabbit asl-casein (Devinoy et al., 1988) and a fragment containing the mouse p-actin cDNA (Alanso et al., 1986) were used as probes. Construction of asl-casein gene CAT plasmid A 5.5-kilobase (kb) PstI/AuaII fragment of the rabbit asl-casein gene was obtained from a A Charon 4A clone (Devinoy et al., 1988). It consisted of 3.7 kb of 5’ flanking sequence, the first non-coding exon (49 nucleotides (nt)), the first intron (1719 nt> and 5 nt of the second exon until an AcaII restriction site (Fig. 5). The initiation of translation of TVE asl-casein gene (ATG codon) was not included, the AvaII restriction site being located 8 nt upstream from it. This Pst I/AcaII fragment possessed an internal Pst I site and was thus subcloned in pPolyII1 (Lathe et al., 1987) in two steps. A first step consisted in subcloning a PstI/AcaII fragment (1012 nt) into the SmaI/ PstI sites of pPolyII1 to create the pH1 plasmid. A second step consisted in subcloning the Pst I/ Pst I fragment (4.5 kb) into the PstI site of pH1, in the same orientation as the already subcloned asl-casein gene fragment to create the plasmid pHla. CAT gene (containing the initiation site of translation) and the polyadenylation signal of SV40 were excised from pD23 (Mohun et al., 1987) after digesting the plasmid with BamHI restriction enzyme. The BamHI fragment containing CAT/SV40 sequences was subcloned into the BamHI site of pHla in the same orientation as the already subcloned cusl-casein gene fragment to create the plasmid pac(- 3768/
+ 1774)CAT. Figures in parentheses indicate the 5’ and 3’ borders of cysl-casein gene sequences according to the previously established numbering (Jolivet et al., 1992). The 5’-deletions were introduced after subcloning a Not1 fragment excised from pac(- 3768,’ + 1774)CAT containing asl-casein and CAT sequences into the Not I site of the pBlueto create pBacscript (SK + , Stratagene) ( - 3768,’ + 17741CAT. Using the exonuclease III/mung bean nuclease technique (Stratagene), various 5’ deletions were created starting from the XbaI site of the Bluescript polylinker. Bluescript sequences were protected by cleaving the Sal1 site of the polylinker and deoxythioderivative filling-in of this Sal1 site by Klenow DNA polymerase. Deletion end points were determined by double-strand DNA sequencing using USB DNA sequencing kit with Sequenase (version 2.0). Three plasmids: pBac( -3155,’ + 1774)CAT; pBac( - 1971,’ + 17741CAT; pBac( - 387,’ + 1774)CAT, were created by gradually deleting the 5’ end of the cusl-casein gene fragment. The X/z01 fragments containing C-3155,’ + 1774)CAT and ( - 387,’ + 1774)CAT sequences were excised from the respective pBac(aslcasein)CAT and reinserted into the Xi?01 site of the pPolyII1 polylinker to create pac( - 3155,’ + 1774)CAT and pact - 387/ + 1774)CAT. The containing (- 1971/ fragment Kpn I /Sac I + 1774)CAT sequences was excised from pBac( - 1971,’ + 1774)CAT and reinserted into the KpnI/SacI sites of the pPolyII1 polylinker to create pac( - 1971,’ + 1774)CAT. For the construction of pac( - 51/ + 1774)CAT encompassing the TATA box sequence, a defined fragment of the asl-casein gene c-51, + 416) was synthesized by the polymerase chain reaction. The plasmid pBac( - 3768,’ + 1774)CAT was used as template. The primers contained a SmaI site at their 5’ ends. The sense primer was 5’TAACCCGGGAGCATTTTACTGATCACTGG3’ and the anti-sense was 5’TAACCCGGGCTACCAAAACAAGCCTTGCT3’. The chain reaction contained 1 ng of template in 100 ~1 of 50 mM KCI, 10 mM Tris HCI pH 9.0, 1.5 mM MgCl,, 0.01% gelatin w/v, 0.1% Triton X-100, 0.2 mM dNTP, 1 mM of each primer and 2.5 U of Taq
DNA polymerase (Promega). A cycle consisted of 30 s at 95°C 30 s at 50°C and 1 min at 72°C. The PCR reaction was run for 27 cycles. The chain reaction fragment was ethanol-precipitated then dissolved in Tris 10 mM pH 8.0, EDTA 1 mM. It was then digested with SmaI and NsiI generating two SmaI/Nsi! fragments, c-51,’ + 180) and (+ 180,’ + 416). The SmaI/NsiI fragment ( - 51,’ + 1801 containing 51 nt of 5’ flanking sequences, the first exon and 131 nt of the first intron was introduced into a XbaI/NsiI fragment excised from pac( - 1971,’ + 1774)CAT, containing (+ 181,’ + 1774ICAT sequences and the whole pPolyII1 sequences after filling the XbaI site with the T4 DNA polymerase. The pBLCAT2 vector (Luckow and Schiitz, 19871 containing the CAT gene and SV40 polyadenylation sequences under control of the herpes simplex virus tk promoter was used to test the activity of the fragment c-3768/ - 3155) of the cwsl-casein gene. An AccI fragment (- 3768,’ - 31131 was excised from pBac( - 3768,’ + 1774)CAT and reintroduced in the same orientation as the tkCAT sequences in the BarnHI site of the polylinker of pBLCAT2 after filling-in the AccI and BamHI sites by T4 DNA polymerase to create p( - 3768,’ - 31551-tk-CAT. Results Induction of asI-casein gene promoter acticity by oPRL. in transfected rabbit primary mammary cells In rabbit primary mammary cells, grown on collagen gels for 4 days, expression of the endogenous cusl-casein gene was induced by oPRL in the presence of insulin and cortisol which amplify prolactin action. cusl-casein mRNA was undetectable when cells were cultured in the presence of insulin and cortisol but in the absence of oPRL (Fig. 1). Transcription was induced after addition of oPRL, whereas the level of p-actin mRNA remained unchanged. These rabbit primary mammary ceils, in which the endogenous cusl-casein gene expression can be induced by PRL, were thus used in our experiments in order to define the presence of PRL responsive regions in cwsl-casein-CAT chimeric genes. For this purpose, rabbit primary mammary cells cultured on collagen gels for 4 days were trans-
151
fected with different chimeric asl-casein-CAT constructs (pac(x/1774)CAT constructs). By the end of transfection, oPRL was introduced into the culture medium and cells were harvested 72 h later, during transient expression of the CAT gene. CAT activity was estimated in cell extracts. When pac( - 3768/ + 1774)CAT construct was transfected into these cells, the percentage of [ “C]chloramphenicol which became acetylated was 14-fold higher after oPRL stimulation (Fig. 2A). This increase was not observed when pOCAT, a construct deprived of asl-casein gene sequence and of any promoter was transfected into these cells. Sequences located between -3768 and + 1774 in the cwsl-casein gene sequence thus allow oPRL to induce CAT gene expression. This induced CAT gene expression could be due to general effects of PRL on the mammary cells and observed even when the CAT gene was transcribed from a prolactin insensitive promoter. SV40 promoter is not sensitive to lactogenic hormones. When pSVECAT, a construct containing the CAT gene under control of this viral promoter, was transfected into these cells, prolactin stimulation of CAT gene expression was not observed (Fig. 2B). CAT gene expression is thus regulated by oPRL, only when upstream sequences of the asl-casein gene, located between nt - 3768 and + 1774, are present in the transfected CAT construct. The 5’ end of the upstream fragment (- 3768,’ + 1774) was progressively deleted using nucleases. For this purpose the c-3768/ + 1774)CAT fragment was cloned into a pBluescript vector, giving rise to the pBac(-3768/ + 1774)CAT plasmid. The 5’ end of the upstream fragment c-3768/ + 1774) was deleted generating the pBac( - 3155/ + 1774)CAT plasmid. Both plasmids pBac(- 3768/ + 1774)CAT and pBacc-3155/ + 1774)CAT were transfected into rabbit primary mammary cells. As previously observed for cells transfected with pac(-3768/ + 1774)CAT, oPRL was able to induce a marked increase in CAT gene expression (6.3-fold, Fig. 2B), when cells were transfected with pBac( - 3768/ + 1774)CAT. After transfection with the pBac(-3155/ + 1774)CAT construct, CAT gene expression was no longer significantly stimu-
lated by the oPRL treatment (Fig. 2B, 1.6fold induction). The 5’ deletion of a 612 nt fragment between located position - 3768 and - 3155 thus reduced oPRL stimulation of CAT gene transcription to a non-significant level. This fragment therefore contributes to the strong induction of the asl-casein gene expression by PRL, in rabbit primary mammary cells grown on a collagen matrix in the presence of insulin and cortisol. PRL responsive elements, or regulatory elements which might interact with PRL responsive elements located further downstream on the asl-casein gene, may be located in this distal region. However, complex hormonal and culture conditions are required to fully induce the casein gene expression by prolactin in cultured primary mammary cells (Emerman and Pitelka, 1977) as well as to induce the maximal CAT gene expression of transfected asl-casein gene construct. The 612 nt fragment might thus also carry mammary specific regulatory regions, matrix dependent regulatory elements, insulin or cortisol responsive elements. In an attempt to define the nature of the elements present in the 612 nt fragment, experiments were pursued using a cellular system which is matrix and insulin independent, the CHO cells.
OPRL A.
alphas1 casein >
-
OPRL +
B.
-
+
c)
