Molecular and Cellular Endocrinology,
53
13 (1990) 53-61
Elsevier Scientific Publishers Ireland, Ltd. MOLCEL 02343
Osmotic shock of cultured primary mammary cells amplifies the hormonal induction of casein gene expression Rachel Malienou-Ngassa, Unit.4 de Diff&enciation
Claudine Puissant and Louis-Marie
Cellulaire, Institut National de la Recherche Agronomique,
Houdebine
78350 Jouy-en-Josas,
France
(Received 5 February 1990; accepted 6 July 1990)
Key worrlF: Mammary cell; Osmotic shock; Casein gene expression; (Rabbit)
Primary cells from rabbit mammary gland cultured on floating collagen were transfected with various plasmids in different conditions. Conventional transfection methods using DEAE-dextran or calcium phosphate followed by an osmotic shock with dirnethyl sulphoxide (DMSO), polyethylene glycol (PEG) or glycerol did not prevent lactogenic hormones to induce casein synthesis. On the contrary and unexpectedly, casein synthesis was markedly stimulated by transfection. This amplification was obtained as well with DMSO, PEG and glycerol alone or in the presence of DEAE-dextran, calcium phosphate or DNA. None of these compounds induced casein synthesis in the absence of prolactin. A shock by DMSO also amplified the accumulation of P-casein mRNA in the presence of prolactin. These results show for the first time that primary cultured mammary cells can be efficiently transfected and still keep their capacity to respond to lactogenic hormones. They also indicate that the short osmotic shocks conventionally used in transfection have a potent long-term stimulatory effect on casein gene expression, which is mediated through an unknown mechanism.
Introduction
The induction of milk protein gene expression is under the dependency of several hormones which can be studied in vivo and also in vitro using mammary explants or isolated mammary epithelial cells cultured on floating collagen (Emerman et al., 1977; Houdebine et al., 1985; Bissel and Hall, 1987). Milk protein genes from different species have been isolated but the mechanism of their regulation remains poorly understood. In
Address for correspondence: Louis-Marie Houdebine, Unit6 de Differentiation Cellulaire, Institut National de la Recherche Agronomique, 78350 Jouy-en-Josas, France. 0303-7207/90/$03.50
vitro studies using mammary cell lines COMMA-D and MCF7 have provided so far a limited information on the role of hormones at the gene level (Gordon et al., 1987, 1988; Yu-Lee and Rosen, 1988). Recently, Doppler et al. (1988) have been able to obtain a hormonal regulation of rat /3casein-chloramphenicol acetyltransferase chimaeric gene using the mouse cell line HCll. Although HCll cell line is now available, it may be better to use primary cells which are expected to have kept all the characteristics of functional mammary cells. The work carried out in our laboratory has led us to isolate three of the major genes expressed in the rabbit mammary gland (as,-casein, /3-casein and whey acidic protein) (Devinoy et al., 1988a, b, c; Schaerer et al., 1988). Rabbit mammary cells
0 1990 Elsevier Scientific Publishers Ireland, Ltd.
54
can be cultured and induced by hormones when cultured on floating collagen (Servely et al., 1987) and they can potentially be transfected to study milk protein gene expression. The work reported here indicates for the first time that the conventional methods to transfect cells in culture can be used with primary mammary cells and that unexpectedly the osmotic shock leads to amplification of the stimulation of casein gene expression by lactogenic hormones. Materials and methods Cell culture Mammary cells were cultured essentially as previously described (Servely et al., 1987). In brief, mammary gland from 13- to 20-day pregnant rabbits previously cut into small fragments was digested by collagenase (200 IU/ml) and hyaluronidase (200 IU/ml) for 2 h at 37°C. The undigested material was withdrawn by a filtration. The cells in suspension were centrifuged at 1000 X g for 3 min. The resulting pellet was resuspended in the culture medium and left 15 min for decantation. The supernatant containing fibroblasts and isolated epithelial cells was discarded. The pellet was washed several times in the same conditions. The resulting material containing aggregated epithelial cells was mildly dispersed by pipetting and finally seeded on a gel of collagen extracted from rat tail. The cells were cultured 2-5 days in 35 mm diameter dishes in medium F,, + MEM (50% each) and in the presence of 2% Ultroser G (IBFFrance). In these conditions, growth of fibroblasts remained limited and essentially epithelial cells were present at the time of induction. No differences in the intensity of the response of cells to hormones could be observed when the growth phase did not exceed 1 week. Constant pH was maintained by an atmosphere containing 5% CO, and 95% 0,. After 4 days, Ultroser was removed and the cells were cultured for 2 days more in the presence of insulin (5 pg/rnl) and cortisol (500 ng/ml) to completely deinduce the mammary cells. Transfection and induction by prolactin were then performed. Each dish contained 0.9 ml collagen gel. The number of cells seeded at the beginning of the culture was only roughly estimated ranging from 300,000 to one million per dish according to
the cell preparation. As a matter of fact, freshly prepared cells are essentially aggregates containing variable amounts of cells. Hence in the experiments reported here variable numbers of cells are present in dishes from different cultures but they may be considered as equal in dishes within a given culture. Transjections Transfections were carried out in the culture medium without Ultroser in the presence of DEAE-dextran (200 pg/ml for 5 h) or calcium phosphate (for 5 h). In both cases, DNA was at a concentration of 5 pg/ml. The osmotic shock was done using 10% dimethyl sulphoxide (DMSO) (for 2 mm), 15% or 44% polyethylene glycol 1500 (PEG) (for 4 min) or 10% glycerol (for 2 mm). A new culture medium without Ultroser but containing bovine insulin (5 pg/ml), cortisol(500 ng/ml) with or without ovine prolactin (NIH-PS-13) (1 pg/ml) was added. Collagen was then detached from the dish and culture was pursued for several days as described in the figure legends. Medium was collected and replaced by fresh medium as described. Media were kept frozen until assay of casein or human growth hormone (hGH) content. Cells on collagen were kept frozen until preparation of RNA. A limited number of cells which did not survive after the transfection were removed with the first medium sampling. /3-Casein and hGH measurements /3-Casein in medium and in cells was estimated by a radioimmunoassay as previously described (Servely et al., 1987). The content of ,&casein stored in cells was estimated as follows. After withdrawal of the medium, the cells were lysed by adding 0.5 ml of culture medium containing 0.5% Triton X-100. The solubilized fraction was separated from collagen and cell debris by centrifuging at 3000 x g for 10 min. The supematant containing the intracellular milk proteins was assayed for its content in /3-casein. hGH was estimated using a specific radioimmunoassay. In all cases, results are the mean (f SEM) of three independent dishes. Measurement of P-casein and actin mRNA Total RNA was extracted from cells at the end of the culture using the extraction by phenol and
55
chloroform in the presence of guanidium isothiocyanate (Chomczynski and Sac&i, 1987). RNA (25 pg) was spotted in water after a denaturation in HCHO on 1 cm* nylon membranes (Zeta probe) which were hybridized overnight with a 32Plabelled probe (lo* cpm/pg DNA) (3 X lo6 cpm per ml hybridization medium) in the presence of 0.5 M sodium phosphate (Mahmoudi and Lin, 1989). Results, which are the means (f SEM) of triplicate determinations, are expressed as cpm bound to filters after subtraction of blanks obtained in the absence of RNA. Northern blot of RNA was carried out by conventional method. RNA (20 pg per well) was denatured by HCHO and electrophoresis was carried out in 1.5% agarose in a buffer containing 20 mM sodium phosphate, 5 mM sodium acetate and 10 mM EDTA. A constant voltage of 75 V was maintained for 4-5 h and buffer was permanently renewed to maintain a constant pH. RNA was transferred to Zeta
probe nylon in water and hybridization ried out as described above.
was car-
Chemical products All the products were of reagent grade. The plasmid containing SV40 early gene promoter, hGH, cDNA, an intron from mouse cu-globin and SV40 polyadenylation sequence (pSV518) was kindly given by Dr. I. Lupker (Elf Biorecherches, Toulouse, France). The plasmid containing ovine /?-lactoglobulin gene (pSSltgX) (Simons et al., 1987) was kindly given by Dr. A.J. Clark (AFRc, Edinburgh, U.K.). The plasmid containing actin cDNA was provided by Dr. M. Buckingham (Institut Pasteur, Paris, France). Results
Mammary cells have kept their hormone responsiveness after transfection The plasmid pSV518 was transfected into the
3oc
z \ ,p 2oc
= .
.c f 8 &
E
P
3
=0
5, 100
Fig. 1. A: Fxpression of hGH in the culture medium of transfected mammary cells. Cells were transfected with pSVS18 containing the hGH gene in the presence of DEAE-dextran. Osmotic shock was carried out with DMSO, PEG or glycerol as indicated in Materials and Methods. Medium was changed every 2 days and hGH was measured using a radioimmunoassay. (0 0) control not transfected, cells transfected with pSV518 in the presence of DUE-dextran and shocked with (AA) 10% DMSO, +) 15% PEG, (0 -0) 44% PEG, (W-m) 20% glycerol. B: Induction of fi-casein secretion after transfection. (*Media tested in (A) were assayed for their content in ~-casein. (0 0) control not transfected and cultured without pro&tin, A) control not transfected and cultured with prolactin, (AA) 10% DMSO, (++) 15% PEG, (0 -0) 44% (APEG, (m -m) 20% glycerol.
56
mammary cells using DEAE-dextran followed by an osmotic shock. This plasmid was chosen for several reasons: (1) it contains SV40 early gene promoter which is expected to work in mammary cells, (2) hGH is easily detected with a specific radioimmunoassay, (3) hGH has prolactin-like activity and it may stimulate the transfected epithelial cells through an autocrine action.
2
4
6
8
IO
Results shown in Fig. 1A indicate that hGH was secreted in the culture medium for several days after transfection. Osmotic shock with the different method appeared to have a variable effect. /3-Casein accumulation was induced by the secreted hGH to a significant level (Fig. 1B). This indicates that the epithelial cells were not inactivated by transfection. This also suggests that
Days
2
Fig. 2. A : Induction of /3-casein synthesis in cultured mammary cells after transfection. Cells were cultured for 2 days and they were then subjected to a transfection in the presence of DEAE-dextran followed by an osmotic shock with 44% PEG. 0) untransfected cells cultured in the absence of (Oprolactin, (aA) untransfected cells cultured in the presence of prolactin, (*+) cells after a mock transfection without DNA and cultured in the absence of prolactin, 0) cells after mock transfection without DNA and (Ocultured in the presence of prolactin, (BW) cells after transfection with pSV518 and cultured in the presence of prolactin, (AA) cells after transfection with /?-lactoglobulin gene and cultured in the presence of prolactin. B: Cells were transfected as in (A) but the osmotic shock was carried out with 10% DMSO. Symbols are those of (A). C: Cells were transfected by the calcium phosphate DNA procedure followed by an osmotic shock with 20% glycerol. Symbols are those of (A).
epithelial cells have been successfully transfected. Unexpectedly, the response of the cells after transfection was high in comparison to the hGH concentration present in the medium and in comparison to the action of the only prolactin. This suggested that the transfection process interfered with hormone action. Transfection stimulates casein gene expression In order to evaluate the possible effects of the chemical compounds added with DNA on the induction of casein synthesis, the mammary cells were subjected to the transfection process in the presence or in the absence of DNA. Results reported in Fig. 2A, B and C revealed that after transfection of the cells with hGH or B-lactoglobulin gene, the levels of B-casein secretion were much higher than in the control untransfected cells. This stimulatory effect was observed as well as after the action of DEAE-dextran or of calcium phosphate and after the osmotic shock with PEG, DMSO or glycerol. This effect was also obtained after a mock transfection carried out in the absence of DNA. This indicates that it is the transfection procedures per se and not the foreign DNA which interfere with the hormonal action. In no case, transfections were able to induce casein synthesis in the absence of prolactin indicating that the events which take place during transfection have no prolactin-like effect but only an amplificatory action. DMSO alone stimulates prolactin action All the experiments described above were carried out with the complete mixture used in transfection. In order to discriminate between the possible effects of DEAE-dextran, calcium phosphate, PEG, DMSO or glycerol, the cells were maintained for 2 min in the presence of 10% DMSO with or without a previous treatment with DEAEdextran. The results shown in Fig. 3 clearly demonstrate that the stimulatory effect was due to DMSO itself and not to DEAE-dextran or to a combination of both. Experiments not shown here indicated that PEG and glycerol alone also amplified the prolactin effect on B-casein synthesis. DMSO stimulates jhtrsein mRNA accumulation The experimental data reported above do not indicate if osmotic shock stimulates B-casein
secretion only or if it modifies milk protein gene expression at the mRNA level as well. On the other hand, the effect of the osmotic shock was measured in all cases a relatively long time after withdrawal of the transfection medium. To tentatively know more of the mechanism of action of DMSO, cells were shocked with DMSO and immediately subjected to prolactin stimulation. BCasein mRNA was measured at different periods after the osmotic shock. The results shown in Figs. 4A and 5 indicate that the osmotic shock by DMSO was unable per se to activate casein gene expression but that it amplified B-casein secretion and B-casein mRNA accumulation in a parallel fashion in the presence of prolactin. The effect of DMSO was specific since no induction of actin mRNA concentration could be observed after the osmotic shock (Fig. 4B). The kinetic study also revealed that the stimulatory effect of the osmotic shock by DMSO did not express until after several hours in these experimental conditions (Fig. 4). Measurements of B-casein concentrations in the culture medium and in cells confirmed that Bcasein is accumulated faster in cells than in the medium (Servely et al., 1987). They also indicate that DMSO stimulated simultaneously B-casein synthesis and B-casein mRNA accumulation. Hence, DMSO did not act only by accelerating
I
2
3
4
5 mys
Fig. 3. Induction of b-casein synthesis by prolactin after an osmotic shock with DMSO. Conditions of culture were those untreated cells cultured in the absence of Fig. 2. (o -00) of prolactin, (Dn) untreated cells cultured in the presence of prolactin, (O0) cells treated by 10% DMSO and cultured in the presence of prolactin, (A -A) cells treated by DEAE dextran, shocked by 10% DMSO and cultured in the presence of prolactin.
58
0 P-l
4
s_
x D
E ._
.z 2
z
40.
E
,” -
30..
P n E
.g 20.. B v & IO.
20
30
40
50
HOUfS
PRL
--++++++++++
CMSO
-+-+-+-i--+-+
the .exocytosis of /3-casein (Fig. 5). Moreover, casein synthesis and secretion were somewhat reduced during the early phase of the induction by prolactin in cells previously shocked by DMSO (Fig. 5). This transient inhibition may be ascribed to the stress caused by the osmotic shock.
The data reported here clearly indicate that several compounds commomy used to provoke an
Fig. 4. Effect of an osmotic shock by DMSO on &casein mRNA accumulation. Cells were cultured as described in Materials and Methods. They were then treated or not for 2 min by 10% DMSO and cultured for 2 more days in the presence or in the absence of prolactin. /Kasein and actin mRNA were then measured. RNA were extracted from three pooled 35 mm dishes. A: @-Casein mRNA concentration measured by dot blotting and scintillation counting. Results are the means of triplicates. 10 pg RNA were spotted on each filter and results are the mean of triplicates: (e -a) control; (m-m) i-DMSO. B: ,&Casein and actin mRNA measured by Northern blotting. The RNA fractions were used in (A). 10 pg RNA were added in each well. Two filters were hybridized independently with rabbit @-casein and actin probes.
osmotic shock markedly enhanced the response of mammary cells to prolactin. The fact that mRNA for casein was overaccumulated after treatment by DMSO strongly suggests that this compound amplifies the whole prolactin signal which stimulates milk protein gene expression rather than only the exocytosis of casein. The observed effect is clearly not related to the presence of Ca*+ since it was observed as well after DEAE-dextran and even without a previous treatment with these compounds. In any case, it is known that a massive
59
a
24
48 nours
Fig, 5. Kinetics of /&ea.&n accumulation under prolactin stimulation after an osmotic shock by DMSO. Experimental conditions are those described in the legend to Fig. 4. /3-Casein concentration was measured in the culture medium (1 ml per dish) and in cells. Results are the means of three. independent dishes. (0 -Q control cells; (0 0) cells after DMSO. A: &Casein in cells. 8: j3Xasein in the medium.
60
uptake of Ca2+ by the mammary cell does not alter the response to prolactin (Houdebine, 1981; Devinoy et al., 1988~). On the other hand, the stimulator-y effect of the osmotic shock is obviously not related to the presence of DNA since the same observation was done after mock transfection. It is rather difficult to explain how DMSO and the related compounds exert their action. The most surprising observation is certainly that the stimulatory effect of the brief osmotic shock persists for several days. DMSO is known to be a potent inducer for several genes such as globin (Weidner and Winicov, 1989) and even for casein gene in a rat mammary cell line (Bennett et al., 1978). In these cases, DMSO concentration is low (1.5%) and it must be present permanently. Experiments not shown here revealed that DMSO at 1.5% was totally ineffective in the presence or in the absence of prolactin. On the other hand, it has been reported that several other chemical compounds such as hexamethylene-bis-aceta~de have a stimulatory action similar to DMSO (Weidner and Winicov, 1989). PEG and glycerol obviously do not belong to this category of compounds. It seems therefore reasonable to consider the stimulatory effect of DMSO, PEG and glycerol on prolactin action essentially as a result of the osmotic shock. Experiments carried out with the same biological material have shown that prolactin receptor is spontaneously internalized and degraded and that this receptor is don-reg~ated by prolactin (Djiane et al., 1979, 1980). It is not known if the endocytosed prolactin-receptor complex can still generate the intracellular prolactin signal which stimulates the expression of milk protein gene. Whatever happens, in all the experiments described here, prolactin was added after the withdrawal of DMSO. It is therefore very unlikely that the effect of the osmotic shock consists in enhancing the number of endocytosed prolactinreceptor complexes. In a previous work, it has been shown that prolactin has a persistent effect after having been left for only a few hours in the presence of the mammary cells (Servely et al., 1982). This suggested that prolactin could be maintained in the form of a very stable hormone receptor complex in the periphery of the cell. It is conceivable that
the osmotic shock enhanced the stability of the prolactin-receptor complex or that it favoured the coupling between the receptor and the membrane system which generates the intracellular signal. Results not shown here indicated that the osmotic shock by DMSO increased only very weakly the number of prolactin receptors. The DMSO effect cannot therefore reasonably be explained by a variation of this parameter. Indeed, even a frank accumulation of prolactin receptor induced by the action of lysosomotropic agents proved to be unable to sensitize the explants towards prolactin stimulation (Djiane et al., 1980). The data reported in the present paper thus clearly demonstrate for the first time that a brief osmotic shock unexpectedly induces a long-term sensitization of the mammary epithelial cells towards the prolactin stimulus. After this preliminary observation, it remains difficult to speculate about the mechanism of action of the osmotic shock on prolactin action at the cellular and molecular level. From a practical point of view, the experiments described here indicate that primary epithelial mammary cell can be efficiently transfected to study the hormonal control of milk protein gene expression but that the effect of osmotic shock must be taken into account to properly evaluate the prolactin action on the transferred gene constructs. Acknowledgments This work was carried out with the excellent technical help of Mr. H. Grabowski. It was supported by the financial help of the Biotechnology Action Program of the European Community. References Bennett, D.C., Peachey, L.A., Durbin, H. and Rudiand, P.S. (1978) Cell 15, 283-298. Bissell, M.J. and Hall, H.G. (1987) in The Mammary Gland Development Regulation and Function (M. NeviBe and C. Danills, eds.), pp. 97-146, Plenum Publishing Corp., New York, NY. Chomczynski, P. and Sac&i, N. (1987) Anal. B&hem. 162, 156-159. Devinoy, E., Hubert, C., Schaerer, E., Houdebme, L.M. and Kraehenbuhl, J.P. (1988a) Nucleic Acids Res. 16, 8180. Devinoy, E., Schaerer, E., Jo&et, G., Fontaine, M.L., Rraebenbuhl, J.P. and Houdebine, L.M. (1988b) Nucleic Acids Res. 16, 11813.
61 Devinoy, E., Hubert, C., Jolivet, G., Thepot, D., Clergue, N., Desaleux, M., Dion, M., Servely, J.L. and Houdebine, L.M. (1988~) Reprod. Nutr. Dev. 28, 1145-1164. Djiane, J., Delouis, C. and Kelly, P.A.,(1979) Proc. Sot. Exp. Biol. Med. 162, 342-346. Djiane, J., Kelly, P.A. and Houdebine, L.M. (1980) Mol. Cell. Endocrinol. 18, 87-95. Doppler, W., Groner, B. and Ball, R.K. (1988) Proc. Natl. Acad. Sci. U.S.A. 86,104-108. Emerman, J.T., Enami, J., Pitelka, D.R. and Nandi, S. (1977) Proc. Natl. Acad. Sci. U.S.A. 74, 4466-4470. Gordon, K., Lee, E., Vitale, J.A., Smith, A.E., Westphal, H. and Hetighausen, L. (1987) Biotechnology 5, 1183-1187. Houdebine, L.M. (1981) Biol. Cell 40,129-134. Houdebme, L.M., Djiane, J., Dusanter-Fourt, I., Martel, P.,
Kelly, P.A., Devinoy, E. and Servely, J.L. (1985) J. Dairy Sci. 68,489-500. Mahmoudi, M. and Lin, V.K. (1989) Biotechniques 7,332-333. Schaerer, E., Devinoy, E., Kraehenbuhl, J.P. and Houdebine, L.M. (1988) Nucleic Acids Res. 16, 11814. Servely, J.L., Teyssot, B., Houdebine, L.M., Delouis, C. and Djiane, J. (1982) Biochimie 64, 133-140. Servely, J.L., Gueuens, G.M., Martel, P., Houdebine, L.M. and Debrabander, M. (1987) Biol. Cell. 59, 121-128. Simons, P., McClenaghan, M. and Clark, J. (1987) Nature 328, 530-532. Weidner, D.A. and Winicov, I. (1989) Mol. Cell. B&hem. 90, 175-183. Yu-Lee, L.Y. and Rosen, J.M. (1988) Mol. Endocrinol. 2, 431-443.
53
13 (1990) 53-61
Elsevier Scientific Publishers Ireland, Ltd. MOLCEL 02343
Osmotic shock of cultured primary mammary cells amplifies the hormonal induction of casein gene expression Rachel Malienou-Ngassa, Unit.4 de Diff&enciation
Claudine Puissant and Louis-Marie
Cellulaire, Institut National de la Recherche Agronomique,
Houdebine
78350 Jouy-en-Josas,
France
(Received 5 February 1990; accepted 6 July 1990)
Key worrlF: Mammary cell; Osmotic shock; Casein gene expression; (Rabbit)
Primary cells from rabbit mammary gland cultured on floating collagen were transfected with various plasmids in different conditions. Conventional transfection methods using DEAE-dextran or calcium phosphate followed by an osmotic shock with dirnethyl sulphoxide (DMSO), polyethylene glycol (PEG) or glycerol did not prevent lactogenic hormones to induce casein synthesis. On the contrary and unexpectedly, casein synthesis was markedly stimulated by transfection. This amplification was obtained as well with DMSO, PEG and glycerol alone or in the presence of DEAE-dextran, calcium phosphate or DNA. None of these compounds induced casein synthesis in the absence of prolactin. A shock by DMSO also amplified the accumulation of P-casein mRNA in the presence of prolactin. These results show for the first time that primary cultured mammary cells can be efficiently transfected and still keep their capacity to respond to lactogenic hormones. They also indicate that the short osmotic shocks conventionally used in transfection have a potent long-term stimulatory effect on casein gene expression, which is mediated through an unknown mechanism.
Introduction
The induction of milk protein gene expression is under the dependency of several hormones which can be studied in vivo and also in vitro using mammary explants or isolated mammary epithelial cells cultured on floating collagen (Emerman et al., 1977; Houdebine et al., 1985; Bissel and Hall, 1987). Milk protein genes from different species have been isolated but the mechanism of their regulation remains poorly understood. In
Address for correspondence: Louis-Marie Houdebine, Unit6 de Differentiation Cellulaire, Institut National de la Recherche Agronomique, 78350 Jouy-en-Josas, France. 0303-7207/90/$03.50
vitro studies using mammary cell lines COMMA-D and MCF7 have provided so far a limited information on the role of hormones at the gene level (Gordon et al., 1987, 1988; Yu-Lee and Rosen, 1988). Recently, Doppler et al. (1988) have been able to obtain a hormonal regulation of rat /3casein-chloramphenicol acetyltransferase chimaeric gene using the mouse cell line HCll. Although HCll cell line is now available, it may be better to use primary cells which are expected to have kept all the characteristics of functional mammary cells. The work carried out in our laboratory has led us to isolate three of the major genes expressed in the rabbit mammary gland (as,-casein, /3-casein and whey acidic protein) (Devinoy et al., 1988a, b, c; Schaerer et al., 1988). Rabbit mammary cells
0 1990 Elsevier Scientific Publishers Ireland, Ltd.
54
can be cultured and induced by hormones when cultured on floating collagen (Servely et al., 1987) and they can potentially be transfected to study milk protein gene expression. The work reported here indicates for the first time that the conventional methods to transfect cells in culture can be used with primary mammary cells and that unexpectedly the osmotic shock leads to amplification of the stimulation of casein gene expression by lactogenic hormones. Materials and methods Cell culture Mammary cells were cultured essentially as previously described (Servely et al., 1987). In brief, mammary gland from 13- to 20-day pregnant rabbits previously cut into small fragments was digested by collagenase (200 IU/ml) and hyaluronidase (200 IU/ml) for 2 h at 37°C. The undigested material was withdrawn by a filtration. The cells in suspension were centrifuged at 1000 X g for 3 min. The resulting pellet was resuspended in the culture medium and left 15 min for decantation. The supernatant containing fibroblasts and isolated epithelial cells was discarded. The pellet was washed several times in the same conditions. The resulting material containing aggregated epithelial cells was mildly dispersed by pipetting and finally seeded on a gel of collagen extracted from rat tail. The cells were cultured 2-5 days in 35 mm diameter dishes in medium F,, + MEM (50% each) and in the presence of 2% Ultroser G (IBFFrance). In these conditions, growth of fibroblasts remained limited and essentially epithelial cells were present at the time of induction. No differences in the intensity of the response of cells to hormones could be observed when the growth phase did not exceed 1 week. Constant pH was maintained by an atmosphere containing 5% CO, and 95% 0,. After 4 days, Ultroser was removed and the cells were cultured for 2 days more in the presence of insulin (5 pg/rnl) and cortisol (500 ng/ml) to completely deinduce the mammary cells. Transfection and induction by prolactin were then performed. Each dish contained 0.9 ml collagen gel. The number of cells seeded at the beginning of the culture was only roughly estimated ranging from 300,000 to one million per dish according to
the cell preparation. As a matter of fact, freshly prepared cells are essentially aggregates containing variable amounts of cells. Hence in the experiments reported here variable numbers of cells are present in dishes from different cultures but they may be considered as equal in dishes within a given culture. Transjections Transfections were carried out in the culture medium without Ultroser in the presence of DEAE-dextran (200 pg/ml for 5 h) or calcium phosphate (for 5 h). In both cases, DNA was at a concentration of 5 pg/ml. The osmotic shock was done using 10% dimethyl sulphoxide (DMSO) (for 2 mm), 15% or 44% polyethylene glycol 1500 (PEG) (for 4 min) or 10% glycerol (for 2 mm). A new culture medium without Ultroser but containing bovine insulin (5 pg/ml), cortisol(500 ng/ml) with or without ovine prolactin (NIH-PS-13) (1 pg/ml) was added. Collagen was then detached from the dish and culture was pursued for several days as described in the figure legends. Medium was collected and replaced by fresh medium as described. Media were kept frozen until assay of casein or human growth hormone (hGH) content. Cells on collagen were kept frozen until preparation of RNA. A limited number of cells which did not survive after the transfection were removed with the first medium sampling. /3-Casein and hGH measurements /3-Casein in medium and in cells was estimated by a radioimmunoassay as previously described (Servely et al., 1987). The content of ,&casein stored in cells was estimated as follows. After withdrawal of the medium, the cells were lysed by adding 0.5 ml of culture medium containing 0.5% Triton X-100. The solubilized fraction was separated from collagen and cell debris by centrifuging at 3000 x g for 10 min. The supematant containing the intracellular milk proteins was assayed for its content in /3-casein. hGH was estimated using a specific radioimmunoassay. In all cases, results are the mean (f SEM) of three independent dishes. Measurement of P-casein and actin mRNA Total RNA was extracted from cells at the end of the culture using the extraction by phenol and
55
chloroform in the presence of guanidium isothiocyanate (Chomczynski and Sac&i, 1987). RNA (25 pg) was spotted in water after a denaturation in HCHO on 1 cm* nylon membranes (Zeta probe) which were hybridized overnight with a 32Plabelled probe (lo* cpm/pg DNA) (3 X lo6 cpm per ml hybridization medium) in the presence of 0.5 M sodium phosphate (Mahmoudi and Lin, 1989). Results, which are the means (f SEM) of triplicate determinations, are expressed as cpm bound to filters after subtraction of blanks obtained in the absence of RNA. Northern blot of RNA was carried out by conventional method. RNA (20 pg per well) was denatured by HCHO and electrophoresis was carried out in 1.5% agarose in a buffer containing 20 mM sodium phosphate, 5 mM sodium acetate and 10 mM EDTA. A constant voltage of 75 V was maintained for 4-5 h and buffer was permanently renewed to maintain a constant pH. RNA was transferred to Zeta
probe nylon in water and hybridization ried out as described above.
was car-
Chemical products All the products were of reagent grade. The plasmid containing SV40 early gene promoter, hGH, cDNA, an intron from mouse cu-globin and SV40 polyadenylation sequence (pSV518) was kindly given by Dr. I. Lupker (Elf Biorecherches, Toulouse, France). The plasmid containing ovine /?-lactoglobulin gene (pSSltgX) (Simons et al., 1987) was kindly given by Dr. A.J. Clark (AFRc, Edinburgh, U.K.). The plasmid containing actin cDNA was provided by Dr. M. Buckingham (Institut Pasteur, Paris, France). Results
Mammary cells have kept their hormone responsiveness after transfection The plasmid pSV518 was transfected into the
3oc
z \ ,p 2oc
= .
.c f 8 &
E
P
3
=0
5, 100
Fig. 1. A: Fxpression of hGH in the culture medium of transfected mammary cells. Cells were transfected with pSVS18 containing the hGH gene in the presence of DEAE-dextran. Osmotic shock was carried out with DMSO, PEG or glycerol as indicated in Materials and Methods. Medium was changed every 2 days and hGH was measured using a radioimmunoassay. (0 0) control not transfected, cells transfected with pSV518 in the presence of DUE-dextran and shocked with (AA) 10% DMSO, +) 15% PEG, (0 -0) 44% PEG, (W-m) 20% glycerol. B: Induction of fi-casein secretion after transfection. (*Media tested in (A) were assayed for their content in ~-casein. (0 0) control not transfected and cultured without pro&tin, A) control not transfected and cultured with prolactin, (AA) 10% DMSO, (++) 15% PEG, (0 -0) 44% (APEG, (m -m) 20% glycerol.
56
mammary cells using DEAE-dextran followed by an osmotic shock. This plasmid was chosen for several reasons: (1) it contains SV40 early gene promoter which is expected to work in mammary cells, (2) hGH is easily detected with a specific radioimmunoassay, (3) hGH has prolactin-like activity and it may stimulate the transfected epithelial cells through an autocrine action.
2
4
6
8
IO
Results shown in Fig. 1A indicate that hGH was secreted in the culture medium for several days after transfection. Osmotic shock with the different method appeared to have a variable effect. /3-Casein accumulation was induced by the secreted hGH to a significant level (Fig. 1B). This indicates that the epithelial cells were not inactivated by transfection. This also suggests that
Days
2
Fig. 2. A : Induction of /3-casein synthesis in cultured mammary cells after transfection. Cells were cultured for 2 days and they were then subjected to a transfection in the presence of DEAE-dextran followed by an osmotic shock with 44% PEG. 0) untransfected cells cultured in the absence of (Oprolactin, (aA) untransfected cells cultured in the presence of prolactin, (*+) cells after a mock transfection without DNA and cultured in the absence of prolactin, 0) cells after mock transfection without DNA and (Ocultured in the presence of prolactin, (BW) cells after transfection with pSV518 and cultured in the presence of prolactin, (AA) cells after transfection with /?-lactoglobulin gene and cultured in the presence of prolactin. B: Cells were transfected as in (A) but the osmotic shock was carried out with 10% DMSO. Symbols are those of (A). C: Cells were transfected by the calcium phosphate DNA procedure followed by an osmotic shock with 20% glycerol. Symbols are those of (A).
epithelial cells have been successfully transfected. Unexpectedly, the response of the cells after transfection was high in comparison to the hGH concentration present in the medium and in comparison to the action of the only prolactin. This suggested that the transfection process interfered with hormone action. Transfection stimulates casein gene expression In order to evaluate the possible effects of the chemical compounds added with DNA on the induction of casein synthesis, the mammary cells were subjected to the transfection process in the presence or in the absence of DNA. Results reported in Fig. 2A, B and C revealed that after transfection of the cells with hGH or B-lactoglobulin gene, the levels of B-casein secretion were much higher than in the control untransfected cells. This stimulatory effect was observed as well as after the action of DEAE-dextran or of calcium phosphate and after the osmotic shock with PEG, DMSO or glycerol. This effect was also obtained after a mock transfection carried out in the absence of DNA. This indicates that it is the transfection procedures per se and not the foreign DNA which interfere with the hormonal action. In no case, transfections were able to induce casein synthesis in the absence of prolactin indicating that the events which take place during transfection have no prolactin-like effect but only an amplificatory action. DMSO alone stimulates prolactin action All the experiments described above were carried out with the complete mixture used in transfection. In order to discriminate between the possible effects of DEAE-dextran, calcium phosphate, PEG, DMSO or glycerol, the cells were maintained for 2 min in the presence of 10% DMSO with or without a previous treatment with DEAEdextran. The results shown in Fig. 3 clearly demonstrate that the stimulatory effect was due to DMSO itself and not to DEAE-dextran or to a combination of both. Experiments not shown here indicated that PEG and glycerol alone also amplified the prolactin effect on B-casein synthesis. DMSO stimulates jhtrsein mRNA accumulation The experimental data reported above do not indicate if osmotic shock stimulates B-casein
secretion only or if it modifies milk protein gene expression at the mRNA level as well. On the other hand, the effect of the osmotic shock was measured in all cases a relatively long time after withdrawal of the transfection medium. To tentatively know more of the mechanism of action of DMSO, cells were shocked with DMSO and immediately subjected to prolactin stimulation. BCasein mRNA was measured at different periods after the osmotic shock. The results shown in Figs. 4A and 5 indicate that the osmotic shock by DMSO was unable per se to activate casein gene expression but that it amplified B-casein secretion and B-casein mRNA accumulation in a parallel fashion in the presence of prolactin. The effect of DMSO was specific since no induction of actin mRNA concentration could be observed after the osmotic shock (Fig. 4B). The kinetic study also revealed that the stimulatory effect of the osmotic shock by DMSO did not express until after several hours in these experimental conditions (Fig. 4). Measurements of B-casein concentrations in the culture medium and in cells confirmed that Bcasein is accumulated faster in cells than in the medium (Servely et al., 1987). They also indicate that DMSO stimulated simultaneously B-casein synthesis and B-casein mRNA accumulation. Hence, DMSO did not act only by accelerating
I
2
3
4
5 mys
Fig. 3. Induction of b-casein synthesis by prolactin after an osmotic shock with DMSO. Conditions of culture were those untreated cells cultured in the absence of Fig. 2. (o -00) of prolactin, (Dn) untreated cells cultured in the presence of prolactin, (O0) cells treated by 10% DMSO and cultured in the presence of prolactin, (A -A) cells treated by DEAE dextran, shocked by 10% DMSO and cultured in the presence of prolactin.
58
0 P-l
4
s_
x D
E ._
.z 2
z
40.
E
,” -
30..
P n E
.g 20.. B v & IO.
20
30
40
50
HOUfS
PRL
--++++++++++
CMSO
-+-+-+-i--+-+
the .exocytosis of /3-casein (Fig. 5). Moreover, casein synthesis and secretion were somewhat reduced during the early phase of the induction by prolactin in cells previously shocked by DMSO (Fig. 5). This transient inhibition may be ascribed to the stress caused by the osmotic shock.
The data reported here clearly indicate that several compounds commomy used to provoke an
Fig. 4. Effect of an osmotic shock by DMSO on &casein mRNA accumulation. Cells were cultured as described in Materials and Methods. They were then treated or not for 2 min by 10% DMSO and cultured for 2 more days in the presence or in the absence of prolactin. /Kasein and actin mRNA were then measured. RNA were extracted from three pooled 35 mm dishes. A: @-Casein mRNA concentration measured by dot blotting and scintillation counting. Results are the means of triplicates. 10 pg RNA were spotted on each filter and results are the mean of triplicates: (e -a) control; (m-m) i-DMSO. B: ,&Casein and actin mRNA measured by Northern blotting. The RNA fractions were used in (A). 10 pg RNA were added in each well. Two filters were hybridized independently with rabbit @-casein and actin probes.
osmotic shock markedly enhanced the response of mammary cells to prolactin. The fact that mRNA for casein was overaccumulated after treatment by DMSO strongly suggests that this compound amplifies the whole prolactin signal which stimulates milk protein gene expression rather than only the exocytosis of casein. The observed effect is clearly not related to the presence of Ca*+ since it was observed as well after DEAE-dextran and even without a previous treatment with these compounds. In any case, it is known that a massive
59
a
24
48 nours
Fig, 5. Kinetics of /&ea.&n accumulation under prolactin stimulation after an osmotic shock by DMSO. Experimental conditions are those described in the legend to Fig. 4. /3-Casein concentration was measured in the culture medium (1 ml per dish) and in cells. Results are the means of three. independent dishes. (0 -Q control cells; (0 0) cells after DMSO. A: &Casein in cells. 8: j3Xasein in the medium.
60
uptake of Ca2+ by the mammary cell does not alter the response to prolactin (Houdebine, 1981; Devinoy et al., 1988~). On the other hand, the stimulator-y effect of the osmotic shock is obviously not related to the presence of DNA since the same observation was done after mock transfection. It is rather difficult to explain how DMSO and the related compounds exert their action. The most surprising observation is certainly that the stimulatory effect of the brief osmotic shock persists for several days. DMSO is known to be a potent inducer for several genes such as globin (Weidner and Winicov, 1989) and even for casein gene in a rat mammary cell line (Bennett et al., 1978). In these cases, DMSO concentration is low (1.5%) and it must be present permanently. Experiments not shown here revealed that DMSO at 1.5% was totally ineffective in the presence or in the absence of prolactin. On the other hand, it has been reported that several other chemical compounds such as hexamethylene-bis-aceta~de have a stimulatory action similar to DMSO (Weidner and Winicov, 1989). PEG and glycerol obviously do not belong to this category of compounds. It seems therefore reasonable to consider the stimulatory effect of DMSO, PEG and glycerol on prolactin action essentially as a result of the osmotic shock. Experiments carried out with the same biological material have shown that prolactin receptor is spontaneously internalized and degraded and that this receptor is don-reg~ated by prolactin (Djiane et al., 1979, 1980). It is not known if the endocytosed prolactin-receptor complex can still generate the intracellular prolactin signal which stimulates the expression of milk protein gene. Whatever happens, in all the experiments described here, prolactin was added after the withdrawal of DMSO. It is therefore very unlikely that the effect of the osmotic shock consists in enhancing the number of endocytosed prolactinreceptor complexes. In a previous work, it has been shown that prolactin has a persistent effect after having been left for only a few hours in the presence of the mammary cells (Servely et al., 1982). This suggested that prolactin could be maintained in the form of a very stable hormone receptor complex in the periphery of the cell. It is conceivable that
the osmotic shock enhanced the stability of the prolactin-receptor complex or that it favoured the coupling between the receptor and the membrane system which generates the intracellular signal. Results not shown here indicated that the osmotic shock by DMSO increased only very weakly the number of prolactin receptors. The DMSO effect cannot therefore reasonably be explained by a variation of this parameter. Indeed, even a frank accumulation of prolactin receptor induced by the action of lysosomotropic agents proved to be unable to sensitize the explants towards prolactin stimulation (Djiane et al., 1980). The data reported in the present paper thus clearly demonstrate for the first time that a brief osmotic shock unexpectedly induces a long-term sensitization of the mammary epithelial cells towards the prolactin stimulus. After this preliminary observation, it remains difficult to speculate about the mechanism of action of the osmotic shock on prolactin action at the cellular and molecular level. From a practical point of view, the experiments described here indicate that primary epithelial mammary cell can be efficiently transfected to study the hormonal control of milk protein gene expression but that the effect of osmotic shock must be taken into account to properly evaluate the prolactin action on the transferred gene constructs. Acknowledgments This work was carried out with the excellent technical help of Mr. H. Grabowski. It was supported by the financial help of the Biotechnology Action Program of the European Community. References Bennett, D.C., Peachey, L.A., Durbin, H. and Rudiand, P.S. (1978) Cell 15, 283-298. Bissell, M.J. and Hall, H.G. (1987) in The Mammary Gland Development Regulation and Function (M. NeviBe and C. Danills, eds.), pp. 97-146, Plenum Publishing Corp., New York, NY. Chomczynski, P. and Sac&i, N. (1987) Anal. B&hem. 162, 156-159. Devinoy, E., Hubert, C., Schaerer, E., Houdebme, L.M. and Kraehenbuhl, J.P. (1988a) Nucleic Acids Res. 16, 8180. Devinoy, E., Schaerer, E., Jo&et, G., Fontaine, M.L., Rraebenbuhl, J.P. and Houdebine, L.M. (1988b) Nucleic Acids Res. 16, 11813.
61 Devinoy, E., Hubert, C., Jolivet, G., Thepot, D., Clergue, N., Desaleux, M., Dion, M., Servely, J.L. and Houdebine, L.M. (1988~) Reprod. Nutr. Dev. 28, 1145-1164. Djiane, J., Delouis, C. and Kelly, P.A.,(1979) Proc. Sot. Exp. Biol. Med. 162, 342-346. Djiane, J., Kelly, P.A. and Houdebine, L.M. (1980) Mol. Cell. Endocrinol. 18, 87-95. Doppler, W., Groner, B. and Ball, R.K. (1988) Proc. Natl. Acad. Sci. U.S.A. 86,104-108. Emerman, J.T., Enami, J., Pitelka, D.R. and Nandi, S. (1977) Proc. Natl. Acad. Sci. U.S.A. 74, 4466-4470. Gordon, K., Lee, E., Vitale, J.A., Smith, A.E., Westphal, H. and Hetighausen, L. (1987) Biotechnology 5, 1183-1187. Houdebine, L.M. (1981) Biol. Cell 40,129-134. Houdebme, L.M., Djiane, J., Dusanter-Fourt, I., Martel, P.,
Kelly, P.A., Devinoy, E. and Servely, J.L. (1985) J. Dairy Sci. 68,489-500. Mahmoudi, M. and Lin, V.K. (1989) Biotechniques 7,332-333. Schaerer, E., Devinoy, E., Kraehenbuhl, J.P. and Houdebine, L.M. (1988) Nucleic Acids Res. 16, 11814. Servely, J.L., Teyssot, B., Houdebine, L.M., Delouis, C. and Djiane, J. (1982) Biochimie 64, 133-140. Servely, J.L., Gueuens, G.M., Martel, P., Houdebine, L.M. and Debrabander, M. (1987) Biol. Cell. 59, 121-128. Simons, P., McClenaghan, M. and Clark, J. (1987) Nature 328, 530-532. Weidner, D.A. and Winicov, I. (1989) Mol. Cell. B&hem. 90, 175-183. Yu-Lee, L.Y. and Rosen, J.M. (1988) Mol. Endocrinol. 2, 431-443.
