Biotin and choline replace the growth requirement of Madin-Darby canine kidney cells for high-density lipoproteins

MDCK cells maintained on extracellular matrix (ECM)-coated dishes and exposed to Dulbecco's modified Eagle's medium (DME) supplemented with transferrin and either high-density lipoproteins (HDLs) or phosphatidyl choline (PC) liposomes have a growth rate and final cell density similar to those of cultures exposed to serum-supplemented DME. When MDCK cells are exposed to a medium consisting of a mixture (1:1) of DME and F12 medium (D/F), the addition of transferrin (10 μg/ml) alone supports cell growth and the presence of HDLs or PC liposomes is no longer required. MDCK cells exposed to D/F medium supplemented with transferrin can be passaged for more than 50 generations in total absence of serum. The F12 components that support growth in the absence of HDLs or PC liposomes are biotin (which is absent in DME) and choline (which is present in insufficient concentration in DME). Supplementation of DME with transferrin, biotin (3.6 ng/ml), and choline (10 μg/ml) allows optimal growth of MDCK cells and permits serial propagation through more than 50 generations. The growth requirement of MDCK cells for HDLs or PC liposomes can therefore be replaced by adequate concentrations of biotin and choline. The widely observed fact that a combination of DME/F12 medium is more effective than DME alone in supporting cell growth may be due in part to the lack of biotin and suboptimal choline concentration in DME.     

The effect of substrates and competitive inhibitors on the phosphatase-dependent activation of hepatic hydroxymethylglutaryl CoA reductase

Previous studies have demonstrated that the in vitro activation of microsomal hepatic hydroxymethylglutaryl (HMG) CoA reductase by dephosphorylation is inhibited by HMG CoA or NADPH, the substrates of HMG CoA reductase (13). In the present study the effect of three competitive inhibitors of HMG CoA reductase on the activation of HMG CoA reductase was investigated. Adenosine-2'-monophospho-5'-diphosphoribose, a competitive inhibitor for the NADPH binding site, blocked the phosphatase-mediated activation of HMG CoA reductase. By contrast, neither compactin nor mevinolin, competitive inhibitors for the HMG CoA binding site, altered the activation of HMG CoA reductase. Moreover, the HMG CoA-mediated inhibition of the activation of HMG CoA reductase was not blocked even by very high concentrations of either compactin or mevinolin. These observations suggest that HMG CoA can bind to two sites on HMG CoA reductase. One site of HMG CoA binding serves as a catalytic site and is competitively blocked by compactin or mevinolin, and the second binding site is an allosteric site to which only HMG CoA is capable of binding. The binding of HMG CoA to this second site inhibits the activation of HMG CoA reductase by phosphatases.     

Analysis of the coordinate expression of 3-hydroxy-3-methylglutaryl coenzyme A synthase and reductase activities in Chinese hamster ovary fibroblasts

Decreased activities of both 3-hydroxy-3-methylglutaryl coenzyme A (HMG CoA) synthase and HMG CoA reductase are observed in the presence of sterol in the Chinese hamster ovary (CHO) fibroblast. In three different genotypes of CHO cell mutants resistant to 25-hydroxycholesterol both enzyme activities exhibit a decreased response to 25-hydroxycholesterol compared to wild-type cells. Permanently repressed levels of both HMG CoA synthase and HMG CoA reductase activities are observed in another CHO mutant, phenotypically a mevalonate auxotroph. Mevinolin, a competitive inhibitor of HMG CoA reductase, has no effect on HMG CoA synthase activity measured in vitro. Incubation of CHO cells with sublethal concentrations of mevinolin produces an inhibition of the conversion of [14C]acetate to cholesterol and results in elevated levels of both HMG CoA synthase and HMG CoA reductase activities. Studies of CHO cells in sterol-free medium supplemented with cycloheximide indicate that continuous protein synthesis is not required for the maximal expression of HMG CoA synthase activity and provide an explanation for the lack of temporal similarity between HMG CoA synthase and reductase activities after derepression. These results support the hypothesis of a common mode of regulation for HMG CoA synthase and HMG CoA reductase activities in CHO fibroblasts.     

Effect of Serum Lipoproteins on Growth and Sterol Synthesis in Cultured Rat Brain Glial Cells

Cells dissociated from brains of 1-day-old rats were cultured in medium containing either lipoprotein-deficient serum (LPDS) or LPDS plus various lipoprotein fractions. Increases in number of cells and in DNA content served as a measure of cell growth. Cholesterol synthesis was measured from the incorporation of [14C]acetate into total nonsaponifiable lipids and digitonin-precipitable sterols, and from the activity of the enzyme 3-hydroxy-3-methylglutaryl coenzyme A (HMG CoA) reductase. The data indicated that cholesterol biosynthesis from acetate was reduced in cells cultured in medium containing either LPDS plus low-density lipoproteins (LDL), high-density lipoproteins (HDL), or total lipoproteins (LP) and that this reduction was accompanied by a reduction in the activity of the HMG CoA reductase and an increase in the esterified sterol content. The reduction in cholesterol synthesis from acetate was maximal in cells cultured in the presence of HDL, whereas the maximal reduction in the activity of HMG CoA reductase occurred in cells cultured in the presence of LP. The presence of LDL or LP in the culture medium enhanced the cell growth but the presence of HDL did not. Esterified sterol content was highest in cells cultured in the medium containing LPDS plus LP and was not detected in cells cultured in LPDS medium. It is inferred from these data that rat brain glial cells in culture are able to utilize cholesterol in lipoproteins, that the presence of LDL in the medium enhances cell growth, and that reduced cholesterol synthesis in the presence of lipoproteins may occur at the HMG CoA reductase step as well as at some other step(s).     

Regulation of cholesterol synthesis in cultured mouse mammary carcinoma FM3A cells

Mouse mammary carcinoma FM3A cells, which are able to grow in a serum-free medium, have novel characteristics that could be valuable in biochemical and somatic cell genetic studies. In FM3A cells grown in the presence of serum, both sterol synthesis and the activity of 3-hydroxy-3-methylglutaryl coenzyme A (HMG-CoA) reductase, the major rate-limiting enzyme in the cholesterol biosynthetic pathway, were strongly suppressed by human low density lipoprotein (LDL). The addition of LDL (50 micrograms protein/ml) resulted in a 50% decrease in the reductase activity within 3 h and a 95% reduction after 24 h. Similarly, over 90% suppression of the reductase activity was obtained by the addition of LDL or mevalonolactone when the cells were grown on a serum-free medium. ML-236B (compactin), a specific inhibitor of HMG-CoA reductase, inhibited sterol synthesis from [14C]acetate by 80% at 1 microM. Reductase activity in FM3A cells was increased by 2.5- to 5-fold when the cells were treated with ML-236B (at 0.26-2.6 microM for 24 h). Thus, in FM3A cells, HMG-CoA reductase activity responded well to LDL, as is observed in human skin fibroblasts. Along with other novel features of this cell line, the present observations indicate that FM3A cells should be useful in biochemical and somatic cell genetic analysis of cholesterol metabolism, especially as regards the regulation of HMG-CoA reductase activity.     

The effect of factors released from the tumor-transformed cells on DNA synthesis, mitosis, and cellular enlargement in 3T3 fibroblasts

Quiescent serum-starved 3T3 cells can be stimulated to initiate DNA synthesis after addition of conditioned media from spontaneously tumor-transformed 3T3 cells (3T6-cells) or from SV-40-transformed 3T3 cells (SV-3T3 cells). The conditioned media were found to stimulate both the chromosome cycle (i.e., DNA synthesis and cell division) and the growth cycle (i.e., cellular enlargement). Furthermore, addition of conditioned media to quiescent 3T3 cells increased the activity of HMG CoA reductase--an enzyme previously proposed to exercise some control on cell proliferation in 3T3 cells (Larsson and Zetterberg: J. Cell. Physiol. 129:99-102, 1986. The increased activity of HMG CoA reductase after treatment with tumor cell conditioned media was correlated to the stimulatory effects on DNA synthesis. By treating 3T3 cells stimulated to resume proliferation by addition of conditioned media with mevinolin (a competitive inhibitor of HMG CoA reductase) the activity of HMG CoA reductase as well as the DNA synthesis and cell division were efficiently inhibited. In contrast, HMG CoA activity was not coupled to the cellular enlargement. Therefore, it is proposed that one set of factors present in tumor cell conditioned media preferentially stimulates the chromosome cycle by increasing the HMG-CoA reductase activity, whereas another set of factors is responsible for growth in cell size. Both types of factors are required for balanced growth.     

Biogenesis of the crystalloid endoplasmic reticulum in UT-1 cells: evidence that newly formed endoplasmic reticulum emerges from the nuclear envelope

The crystalloid endoplasmic reticulum (ER), a specialized smooth ER of the compactin-resistant UT-1 cell, is composed of multiple membrane tubules packed together in a hexagonal pattern. This membrane contains large amounts of 3-hydroxy-3-methylglutaryl coenzyme A (HMG CoA) reductase, an integral membrane protein that enzymatically regulates endogenous cholesterol biosynthesis. Using morphological and immunocytochemical techniques, we have traced the sequence of events in the biogenesis of this ER when compactin-withdrawn UT-1 cells, which do not have a crystalloid ER, are incubated in the presence of compactin. After 15 h of incubation in the presence of compactin, many cells had profiles of ER cisternae that were juxtaposed to the nuclear envelope and studded with ribosomes on their outer membrane. Both the outer nuclear membrane and the ER membrane contained HMG CoA reductase; however, there was little or no detectable enzyme in rough ER that was free in the cytoplasm. With longer times of incubation in the presence of compactin, these cells had lamellar stacks of smooth ER next to the nuclear envelope that contained HMG CoA reductase. Coordinate with the appearance of the smooth ER, crystalloid ER appeared in the same cell. Often regions of continuity were found between the membrane of the smooth ER and the membrane of the crystalloid ER tubules. These studies suggest that HMG CoA reductase is synthesized along the outer nuclear membrane and in response to increased enzyme synthesis, a membrane emerges from the outer nuclear membrane as smooth ER cisternae, which then transforms into crystalloid ER tubules.     

Effects of compactin, mevalonate and low-density lipoprotein on 3-hydroxy-3-methylglutaryl-coenzyme A reductase activity and low-density-lipoprotein-receptor activity in the human hepatoma cell line Hep G2.

Compactin, an inhibitor of HMG-CoA (3-hydroxy-3-methylglutaryl-CoA) reductase, decreased cholesterol synthesis in intact Hep G2 cells. However, after the inhibitor was washed away, the HMG-CoA-reductase activity determined in the cell homogenate was found to be increased. Also the high-affinity association of LDL (low-density lipoprotein) to Hep G2 cells was elevated after incubation with compactin. Lipoprotein-depleted serum, present in the incubation medium, potentiated the compactin effect compared with incubation in the presence of human serum albumin. Addition of either mevalonate or LDL prevented the compactin-induced rise in activities of both HMG-CoA reductase and LDL receptor in a comparable manner. It is concluded that in this human hepatoma cell line, as in non-transformed cells, both endogenous mevalonate or mevalonate-derived products and exogenous cholesterol are able to modulate the HMG-CoA reductase activity as well as the LDL-receptor activity.     

Regulation of 3-hydroxy-3-methylglutaryl coenzyme A reductase and lipid metabolism in a concanavalin A-resistant Chinese hamster ovary cell line

Lipid metabolism in a concanavalin A-resistant, glycosylation-defective mutant cell line was investigated by comparing growth properties, lipid composition, and lipid biosynthesis in wild-type (WT), mutant (CR-7), and revertant (RCR-7) cells. In contrast to WT and RCR-7, the mutant was auxotrophic for cholesterol, but mevalonolactone did not restore growth on lipoprotein-deficient medium. The use of R-[2-14C]mevalonolactone revealed that CR-7 was deficient in the conversion of lanosterol to cholesterol. Total lipid and phospholipid content and composition were similar in all three cell lines, but CR-7 displayed subnormal content and biosynthesis of cholesterol and unsaturated fatty acids. The mutant was hypersensitive to compactin and was unable to upregulate either 3-hydroxy-3-methylglutaryl coenzyme A (HMG-CoA) reductase activity or the binding and internalization of 125I-labeled low-density lipoprotein (LDL) in response to lipoprotein deprivation. HMG-CoA reductase activity in all three cell lines showed similar kinetics and phosphorylation status, and the binding kinetics and degradation of 125I-LDL were also similar, suggesting that CR-7 possesses kinetically normal reductase and LDL binding sites, but is deficient in their coordinate regulation. Tunicamycin (1-2 micrograms/ml) strongly and reversibly suppressed reductase activity in WT and RCR-7. CR-7 was resistant to this inhibitor. In WT cells this suppressive effect was accompanied by inhibition of 3H-labeled mannose incorporation into cellular protein, but 3H-labeled leucine incorporation was unaffected. Immunotitration of HMG-CoA reductase activity in extracts of WT cells, cultured in the presence and absence of tunicamycin, showed that suppression of reductase activity reflected the presence of reduced amounts of reductase protein, implying that glycosylation plays an important role in the coordinate regulation of HMG-CoA reductase activity and LDL binding.     

Effects of 25-hydroxycholesterol, cholesterol, and isoprenoid derivatives on the G1 progression in Swiss 3T3 cells

The effect of inhibition of 3-Hydroxy-3-methylglutaryl Coenzyme A reductase (HMG CoA reductase) on cell cycle progression in proliferating 3T3 cells was studied. It was found that short transient exposures to the HMG CoA reductase inhibitor 25-hydroxycholesterol temporarily blocked the cell cycle traverse in the postmitotic half of G1 (G1pm), whereas cells in the subsequent cell cycle phases were unaffected. The kinetics of the cell cycle delay, induced by 25-hydroxycholesterol, resembled the kinetics of the delay induced by serum depletion, which also inhibited the activity of HMG CoA reductase. In contrast to the case of serum depletion, platelet derived growth factor (PDGF), which efficiently prevented the decrease of HMG CoA reductase in serum-free medium, was not capable of preventing the growth inhibitory effect following treatment by 25-hydroxycholesterol. However, cholesterol and two isoprenoids, dolichol and coenzyme Q, were effective in this respect. In addition, dolichol counteracted the cell cycle delay following short periods of serum starvation.     

The regulation by low-density lipoproteins of the activation of oxidative enzyme-primed lymphocytes is governed by transferrin

The activation of T lymphocytes was regulated in vitro by low-density lipoproteins (LDL). Not all prereplicative events induced by the oxidative enzymatic mitogens neuraminidase and galactose oxidase (NAGO) were susceptible to inhibition by LDL. The accessory cell-independent early blastogenic response was not suppressed. LDL suppressed accessory cell-dependent responses, and the extent of LDL suppression, depended on the concentration of transferrin. A gradient of transferrin determined the point in the cell cycle at which NAGO-primed lymphocytes were suppressed by LDL. When transferrin was low (0-10 micrograms/ml) and in serum-free medium (SFM), LDL suppressed the expression of cell surface receptors for interleukin-2 (IL-2R) and transferrin (TfR), the late blastogenic response prior to DNA replication (72 hr), and DNA replication. At higher levels of transferrin, about 100 micrograms/ml, the LDL-suppressed cells were IL-2R+, TfR+ and responsive to IL-2, but did not enter S phase. LDL suppression could be ablated by IL-2 and by high levels of transferrin (250-1000 micrograms/ml). In RPMI medium containing serum (FBS), the pattern of LDL suppression was different from that in SFM: fully activated IL-2R+, TfR+ lymphocytes were unresponsive to exogenous IL-2, suggesting that they were blocked at the G1/S boundary. This block was also relieved by transferrin (greater than 100 micrograms/ml). The data suggest that the interplay between transferrin and LDL is a critical factor in the NAGO-induced stimulation of T lymphocytes. LDL and transferrin exert negative and positive control of lymphocyte activation, respectively. In SFM, LDL appear to alter transferrin utilization by accessory cells; in RPMI-FBS, by fully activated T lymphocytes.     

Metabolism of homologous and heterologous lipoproteins by cultured rat and human skin fibroblasts

Rat fibroblasts degraded human low density lipoprotein (LDL) very slowly, one-tenth to one-fortieth the rates observed in human fibroblasts. In rat cells, human LDL caused only very small increases in cell cholesterol content and acylCoA:cholesterol acyltransferase (ACAT) activity and caused only small decreases in beta-hydroxy-beta-methylglutaryl CoA (HMG CoA) reductase activity; in human cells, however, human LDL induced very large changes in all three of these parameters, as expected. The binding of human LDL to rat fibroblasts was not reduced by previous incubation with human LDL or with 25-hydroxycholesterol. Thus, in rat fibroblasts there appear to be few, if any, regulated high-affinity receptors that recognize human LDL. Rat LDL fractions (d 1.02-1.05 g/ml), in contrast, were degraded more rapidly than human LDL by rat fibroblasts, caused a significant increase in cell cholesterol content, an increase in ACAT activity, and a significant decrease in HMG CoA reductase activity. Moreover, the degradation of this rat LDL fraction by rat fibroblasts as a function of concentration was biphasic, i.e., there appeared to be a high-affinity component of degradation. Thus, it appears that rat fibroblasts do have a receptor for homologous lipoproteins. However, because both apoprotein B and apoprotein E are present in these rat lipoprotein fractions, the observed effects may relate to recognition of either or both of these apoproteins. The metabolism and metabolic effects of the conventionally defined high density lipoprotein (HDL) fraction of the rat by rat or human fibroblasts resembled those of human LDL in human fibroblasts. It is suggested that rat HDL may, because of its apo E content and higher concentration in rat plasma relative to that of LDL, play an important role in cholesterol homeostasis in vivo.     

Growth requirements of low-density rabbit costal chondrocyte cultures maintained in serum-free medium

The factors required for the active proliferation of low-density rabbit costal chondrocytes exposed to 9:1 (v/v) mixture of Dulbecco's modified Eagle's medium and Ham's F12 medium have been defined. Low-density primary cultures of rabbit costal chondrocytes proliferated actively when the medium was supplemented with high-density lipoprotein (300 micrograms/ml), transferrin (60 micrograms/ml), fibroblast growth factor (FGF) (1 ng/ml), hydrocortisone (10(-6) M), and epidermal growth factor (EGF) (30 ng/ml). Insulin, although it slightly decreased the final cell density, was required for reexpression of the cartilage phenotype at confluence. Optimal proliferation of low-density chondrocyte cultures was only observed when dishes were coated with an extracellular matrix (ECM) produced by cultured corneal endothelial cells, but not on plastic. Furthermore, serum-free chondrocyte cultures seeded at low density and maintained on ECM-coated dishes gave rise to a homogeneous cartilage-like tissue composed of spherical cells. These chondrocytes therefore seem to provide a good experimental system for analyzing factors involved in supporting proliferation of chondrocytes and their phenotypic expression.     

Increase in membrane cholesterol: A possible trigger for degradation of HMG CoA reductase and crystalloid endoplasmic reticulum in UT-1 cells

The crystalloid endoplasmic reticulum (ER) houses large amounts of HMG CoA reductase, the rate-controlling enzyme in cholesterol synthesis. The crystalloid ER appears in UT-1 cells, a line of Chinese hamster ovary cells that has been chronically starved of cholesterol as a result of growth in the presence of compactin, an inhibitor of reductase. When cholesterol was provided to UT-1 cells in the form of low density lipoprotein (LDL), the reductase and crystalloid ER were destroyed. This destruction was preceded by an increase in the cholesterol content of crystalloid ER membranes, as judged by a 4- to 8-fold increase in their ability to form complexes with filipin, a cholesterol-binding compound that can be visualized in freeze-fracture electron micrographs. Filipin binding to other membranes was unchanged. Thus insertion of cholesterol into the crystalloid ER membrane may trigger the degradation of reductase and the membrane itself.     

Gonadotropin modulation of 3-hydroxy-3-methylglutaryl coenzyme A reductase activity in desensitized luteinized rat ovary

These studies were done to examine the effect of gonadotropin on rat luteal 3-hydroxy-3-methylglutaryl coenzyme A (HMG CoA) reductase activity (the rate-limiting step in cholesterol biosynthesis) in ovaries of pregnant mare's serum gonadotropin (PMSG)-human chorionic gonadotropin (hCG) primed rats. Administration of hCG stimulated HMG CoA reductase activity in a time- and dose-dependent manner: significant increases were noted within 4 h, with maximum effects (30-40-fold increases) seen 24 h after hCG (25 IU) administration. This effect was specific in that only LH, of several hormones tested, was as effective as hCG in stimulating HMG CoA reductase activity, and no change in the activity of either liver microsomal HMG CoA reductase or luteal microsomal NADPH-cytochrome c reductase was seen after hCG. The gonadotropin-induced increase in HMG CoA reductase activity seemed to be due to a net increase in enzyme activity, not to a change in the phosphorylated/dephosphorylated state of the enzyme. Pretreatment of animals with aminoglutethimide, an inhibitor of the conversion of cholesterol to steroid (pregnenolone), prevented the hCG-induced rise in HMG CoA reductase activity, whereas treatment with 4-aminopyrazolo[3,4-d]pyrimidine (4-APP), which depletes cellular cholesterol content, led to striking increases in enzyme activity. However, the combined effects of 4-APP and hCG were additive, suggesting that the stimulating effect of hCG on HMG CoA reductase activity is not entirely due to a depletion of cellular sterol content of luteinized ovaries. Similarly, cholesteryl ester and cholesterol syntheses as measured by [14C]acetate conversion were also increased by hCG and 4-APP treatment.(ABSTRACT TRUNCATED AT 250 WORDS)     

A mechanism for the in vitro stimulation of adrenal cortisol biosynthesis by epidermal growth factor

EGF stimulates adrenal steroidogenesis in ewes and in ovine adrenal slices. In vitro, The stimulation is blocked by the cholesterol synthesis inhibitors compactin and AY 9944. EGF stimulates the incorporation of [14C]acetate into cholesterol. EGF increases the activity of the rate limiting enzyme in cholesterol biosynthesis, HMG CoA reductase. EGF has no effect on the levels of any intermediates involved in the conversion of pregnenolone to cortisol, although ACTH produced changes consistent with 17 alpha-hydroxylase activation. We propose that EGF increases adrenal cortisol synthesis in vitro by a stimulation of cholesterol precursor biosynthesis mediated through activation of HMG CoA reductase.     

Lipid metabolism in cultured cells. XVI. Lipoprotein binding and HMG CoA reductase levels in normal and tumor virus-transformed human fibroblasts.

The loss in feedback control of cholesterol biosynthesis in tumor cells was examined in tissue culture. Human fibroblasts from normal subjects, SV40 tumor virus-transformed cell lines, and homozygous familial hypercholesterolemic cells as reference, were grown in tissue culture. Experiments were conducted to relate the regulatory enzyme for cholesterol biosynthesis, HMG CoA reductase, and the membrane-located binding receptors for low density lipoproteins (LDL) that mediate feedback control in normal cells. Monolayers of virus-transformed tumor cells exhibited specific (125)I-labeled LDL binding of 152 +/- 21 ng/mg cell protein, which was essentially the same as that of normal fibroblasts (135 +/- 20 ng/mg). Binding of LDL by familial hypercholesterolemic cells used as controls was only 8 +/- 3 ng/mg under the same test conditions. Basal levels of HMG CoA reductase in tumor cells of 45.2 +/- 6.5 units/mg cell protein were about twice those of normal cells. However, in contrast to the lack of feedback control of this enzyme observed with tumors in vivo, in both the normal and the transformed cells in vitro, activity of the enzyme decreased about fourfold when serum lipids were added. These findings demonstrate that tumor cells growing in vitro contain a normal complement of the membrane-located binding receptors for low density lipoproteins and, although the basal levels are higher than normal, an effective feedback regulation of the enzyme HMG CoA reductase is retained.     

The promoter of the rat 3-hydroxy-3-methylglutaryl coenzyme A reductase gene contains a tissue-specific estrogen-responsive region.

The isoprenoid metabolic pathway is mainly regulated at the level of conversion of 3-hydroxy-3-methylglutaryl coenzyme A (HMG CoA) to mevalonate, catalyzed by HMG CoA reductase. As estrogens are known to influence cholesterol metabolism, we have explored the potential regulation of the HMG CoA reductase gene promoter by estrogens. The promoter contains an estrogen-responsive element-like sequence at position -93 (termed Red-ERE), which differs from the ERE consensus by one mismatch in each half of the palindrome. A Red-ERE oligonucleotide specifically bound estrogen receptor in vitro and conferred receptor-dependent estrogen responsiveness to a heterologous promoter in all cell lines tested. However, expression of a reporter driven by the rat HMG CoA reductase promoter was induced by estrogen treatment after transient transfection into the breast cancer cell line MCF-7 cells but not in hepatic cell lines expressing estrogen receptor. Estrogen induction in MCF-7 cells was dependent on the Red-ERE and was strongly inhibited by the antiestrogen ICI 164,384. A functional cAMP-responsive element is located immediately upstream of the Red-ERE, but cAMP and estrogens inhibit each other in terms of transactivation of the promoter. Similarly, induction by estrogens was inhibited by micromolar concentrations of cholesterol, likely acting via changes in occupancy of the sterol-responsive element located 70 bp upstream of the Red-ERE. Thus, within its natural context, Red-ERE is able to mediate hormonal regulation of the HMG CoA reductase gene in tissues that respond to estrogens with enhanced cell proliferation, while it is not operative in liver cells. We postulate that this tissue-specific regulation of HMG CoA reductase by estrogens could partially explain the protective effect of estrogens against heart disease.     

The seminiferous growth factor induces proliferation of TM4 cells in serum-free medium

The seminiferous growth factor (SGF) of the mammalian testes induces DNA synthesis and cell proliferation of Balb/c 3T3 cells (Bellvé and Feig, 1984; Rec Prog Hormone Res 40:531-567). In this study, SGF was purified 80,000- to 100,000-fold from calf testes and used to examine the growth of TM4 cells in a chemically defined medium. Cells were seeded sparsely in Dulbecco's Modified Eagles/Ham's F12 medium (1:1;v:v) (DME/F12 degrees), containing epidermal growth factor (EGF; 1 ng/ml), insulin (1; 10 micrograms/ml), and transferrin (Tr; 5 micrograms/ml) (DME/F12). After 24 h, the medium was replaced with DME/F12 degrees supplemented with SGF, EGF, 1, or Tr, in two-, three- or four-way combinations. Cell numbers were quantified after another 48 h of culture. EGF, I, and Tr, alone or in two-way combinations, were not mitogenic for TM4 cells. By contrast, SGF (1 U) alone, or with any two of these factors, stimulated TM4 cell proliferation to commensurate levels, and to twofold greater numbers than occurred with the combination of EGF, I, and Tr. Synergisms or inhibitions were not measurable. Follicle-stimulating hormone, luteinizing hormone, prolactin, acidic fibroblast growth factor, or basic fibroblast growth factor was weakly or not mitogenic for TM4 cells. The effect of SGF on cell proliferation was inhibited by 1 microM - 1 nM retinoic acid, but not by retinol or retinyl acetate. SGF was mitogenic for bovine adrenal capillary endothelial cells, an effect that was potentiated by 10 micrograms heparin/ml. Thus, SGF can induce proliferation of TM4 cells and capillary endothelial cells. The former provides a sensitive, and selective, serum-free, bioassay system for SGF activity.