A mevalonate requirement for maintenance of fatty acid and protein synthesis during hormonally stimulated development of mammary gland in vitro.

The effect of compactin on hormonally induced lipogenesis and protein synthesis was studied in vitro in explants of mammary gland from mid-pregnant rabbits. Compactin blocks mevalonate synthesis by the specific inhibition of 3-hydroxy-3-methylglutaryl-CoA reductase, and in this system, culture with 10 microM compactin for 24, 48, and 72 h inhibited incorporation of [1-14C]acetate (but not [2-14C]mevalonate) into sterol by 98, 95, and 86%, respectively. Removal of compactin prior to assay rapidly reversed this effect and was associated with increased tissue 3-hydroxy-3-methylglutaryl-CoA reductase activity. Fatty acid synthesis (measured by incorporation of [1-14C]acetate or [4,5-3H]leucine) and protein synthesis (measured by incorporation of [4,5-3H]leucine) were both inhibited by around 50% after culture with compactin. This inhibition was not rapidly reversed by removal of compactin prior to assay, but it was prevented by inclusion of 1 mM mevalonolactone in the culture medium. After removal of compactin and continued culture in its absence for 24 h with hormones, the normal tissue capacity for fatty acid and protein synthesis was restored, indicating no permanent cell damage. The results suggest a specific requirement for mevalonate (or derived products) for the hormonal maintenance of the increased fatty acid and protein synthesis characteristic of the development of the mammary gland.

* This work was supported by the Wellcome Foundation. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "aduertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. teristic increase in fatty acid synthesis (11) and protein synthesis (12) after the hormonal stimulus which initiates the development i n uitro. The present work uses compactin (a specific inhibitor of mevalonate synthesis (13)) to examine the requirement for mevalonate production in the maintenance of these metabolic changes characteristic of hormonally stimulated development of the rabbit mammary gland i n vitro. Preparation and Culture of Mammary Explants-Explants of lobuloalveolar mammary tissue were prepared from mid-pregnant rabbits (16th day of pregnancy) as described by Forsyth and Myres (10). Duplicate groups of explants were cultured with or without insulin, prolactin, and cortisol for up to 72 h as described previously (15). The effects of compactin and mevalonolactone were measured by including these in the culture medium from time 0. Solutions of these two compounds were prepared and stored as previously described (16). In some experiments, explants which had been cultured in the presence of compactin were removed from the culture medium, rinsed in compactin-free medium, and recultured in the original volume (10 ml) of compactin-free medium 199 (containing hormones) for 1 or 24 h a t 37 "C prior to measurement of incorporation of radioactive substrates as below.
Incubation of Explants-Explants (10-20 mg wet weight) were removed from the culture medium after 24, 48, or 72 h, blotted, weighed, and transferred to a scintillation vial containing 1 ml of incubation medium (gassed with 0 2 + COS (95% + 5%)). For measurement of incorporation of ["Clacetate and mevalonate into lipid, the medium consisted of Krebs bicarbonate buffer containing 1 mM glucose and either [l-"C]acetate (5 pCi; specific radioactivity 58 mCi/ mmol) or [2-14C]mevalonolactone (5 pCi; specific radioactivity 5 mCi/ mmol). When leucine was the radioactive precursor, the incubation medium was the culture medium (medium 199 plus hormones) but containing [4,5-3H]leucine (10 pCi; specific radioactivity 76 mCi/ The abbreviation used is: HMC-CoA, 3-hydroxy-3-methylglutaryl coenzyme A. mmol). In experiments where compactin or mevalonolactone was present in the culture medium, these were included in the incubation medium at the same concentration unless otherwise stated. Incubations were conducted at 37 "C for 1-2 h and were stopped by immersion in ice and washed with ice-cold phosphate-buffered saline prior to further processing (below).

Measurement of Lipid Synthesis from [l-'4C]Acetate and [2-"C]
Meualormlactone-Explants were saponified, and digitonin precipitation of the nonsaponifiable lipid fraction was performed as previously described (1) but with [1,2-3H]cholesterol (20,000 dpm) added to all samples to correct for procedural losses. The saponifiable fraction was isolated following acidification of the saponification medium (I), and recovery of this fraction was monitored by inclusion of [9,10-3H]palmitic acid (200,000 dpm). After removal of solvent, the lipid fractions were counted for 14C and 'H radioactivity. hhnsurement of the ZncorDoration of Radioactivity from 14,5-3Hl .kucine into Lip& and Pro&n-Afte; incubation with radioactive leucine, the explants were washed twice with 2 ml of ice-cold phosphate-buffered saline. Washed explants were extracted with chloroform:metbanol (1:2, v/v) to extract lipid (17). The lipid extract was separated into saponifiable and nonsaponifiable fractions as described above before assaying for radioactivity. The insoluble residue after lipid extraction was dissolved in 0.2 ml of formic acid and assayed for radioactivity. This latter fraction measured incorporation of leucine into protein.

Preparation of Microsomal Fractions
and Assay of HMG-CoA Re-du&a.se-Explants (20-40 mg wet weight) were homogenized manually in an all-glass homogenizer in 0.5 ml of 300 mM sucrose, 10 mM EDTA, and 10 mM mercaptoethanol, pH 7.0. The microsomal pellet was then obtained by centrifugation (1) and suspended in 0.15 ml of buffer containing 100 mM potassium phosphate, pH 7.5, 10 mM EDTA, and 5 mM ditbiothreitol and used for the assay of HMG-COA reductase activity. This was carried out as previously described (1) but modified to give a final volume of 0.15 ml and an incubation time of 3 h. Formation of ["Clmevalonate was found to be linear with time up to 4 h and was proportional to protein. After lactonization, the product was isolated by thin-layer chromatography (18) and counted for radioactivity. Product recovery averaged 90%. Blanks were run without microsomes and with microsomes but without the NADPH-generating system. Assay of Enzymes in Explnnt Homogenates-After culture, explants were suspended in 0.5 ml of 250 mM potassium phosphate buffer, pH 8.2, containing 1 mM EDTA and 1 mM dithiothreitol and then homogenized in an all-glass homogenizer for 2 min at maximum rotor speed. After addition of 25 ~1 of 20% (w/v) Triton X-100, the homogenate was centrifuged at 14,000 x g for 6 min at 4 "C in a microcentrifuge. The supernatant fraction was removed for enzyme assays. The following enzymes were determined according to previously published methods: citrate synthase (19), lactate dehydrogenase (20), and fatty acid synthase (21).

All experiments
described below involved the culturing of mammary gland explants with hormones to stimulate development with or without compactin for 24 h or more. The rates of lipogenesis or protein synthesis were assayed at the end of the culture period in a short incubation (l-2 h) with the appropriate radioactive precursor in the presence or absence of compactin as described. In preliminary experiments (data not shown), we have confirmed that 10 PM compactin inhibited [1-YJacetate incorporation into sterol when added directly to the assay system and that the same concentration of compactin had no significant effect upon incorporation of [l-14C]acetate into fatty acids or on [4,5-3H]leucine incorporation into protein when added directly to the assay system. Procedures." Asterisks indicate that a particular value is significantly different from the histogram immediately to its kft at that time point. *, p < 0.05; **, P < 0.01. Results are means + S.E. for five animals. Duplicate groups of 10 explants were cultured in medium 199 in the presence of hormones with or without compactin (10 PM). Digitonin-precipitable sterol was isolated as described under "Experimental Procedures." The dotted line indicates radioactivity in digitonin-precipitable sterol in the presence of hormones. Radioactivity in digitonin-precipitable sterol in the presence of hormone plus compactin is expressed as a percentage of this. The mean incorporation rates (from two independent experiments) of [ When compactin was removed after culture and before the assay of sterol synthesis, no inhibition was found. This is illustrated in Fig. 3 where culture was carried out with only 2 pM compactin to facilitate the removal of the inhibitor. After 72 h of culture with 10 pM compactin, the removal of the latter before assay resulted in a 131% increase in the incorporation rate of [1-14C]acetate to sterol when compared to controls (Table I). This was associated with a 133% in-

TABLE I
Effect of compactin removal on various cellutar processes after culturing in its presence for 72 h Duplicate groups of 10 explants were cultured in medium 199 with hormones (insulin (5 pg. ml-'), prolactin (1 Fg-ml-'), and cortisol (1 pg.ml")) for 72 h in the presence or absence of 10 p~ compactin. After culture, the control group (without compactin) was then incubated with [l-14C]acetate or [4,5-3H]leucine to determine rates of lipogenesis or protein synthesis, respectively. The compactin group was either incubated similarly with compactin in the incubation medium or were first rinsed free of compactin by a 1-h incubation in culture medium alone and then assayed as above. Results are expressed as percentage of control rates of lipogenesis or protein synthesis and are means f S.E. with the number of independent experiments given in parentheses. In the control group, the mean incorporation rates of precursors were for [1-'%]acetate into sterol and fatty acid, 1325 2 359 (5) and 32,795 k 14,447 (5) dpm. h".mg", respectively, and for [4,5-3H]leucine into fatty acid and protein, 3,031 -+ 521 (5) and 23.122 f 4.344 (5) dpm.h"-mg", respectively. % of hormonally stimulated incorporation rate (l-''C]Acetate to [4,5- crease in HMG-CoA reductase activity when assayed in these explants ( Table 11). The increase was found irrespective of whether results were expressed relative to tissue weight or to tissue content of lactate dehydrogenase or citrate synthase (both of which were unchanged in activity after culture with compactin (Table 11)). These results confirm the site of action and specificity of the compactin effect in mammary gland and show that mammary tissue responds to the inhibitor by increasing the HMG-CoA reductase activity exactly as has been described previously for fibroblasts (16), rat liver (22), isolated rat hepatocytes (23), and rabbit intestinal mucosa (24).  (Fig. 4a) as well as [l-"C]acetate (Fig. 4b). Culture with hormones stimulated incorporation of radiolabeled leucine and acetate by 12-and 64-fold, respectively. The lower response to hormones found with leucine as lipid precursor could reflect its major diversion into protein synthesis. The rate of incorporation of [3Hlleucine into fatty acid was around 10% of the rate of incorporation into protein after hormonal stimulation of explants. In the presence of compactin, culture for more than 24 h resulted in a 40-60% inhibition (Fig. 4, a and b) of the rate of incorporation of both [3H]leucine and [14C]acetate into fatty acids. A similar result was observed when protein synthesis was measured. The hormonally stimulated incorporation of [4,5-3H]leucine into protein was inhibited by 40% when explants were cultured with compactin for 48 and 72 h (Fig. 5 ) . As with fatty acid synthesis, little effect was noted after 24 h of culture with compactin. These effects of compactin on hormonally induced fatty acid and protein synthesis were prevented by the inclusion of mevalonolactone (1 mM) in the culture medium (Figs. 4 and 5). By itself, the mevalonolactone had no effect on either fatty acid synthesis or protein synthesis (data not shown), but it was taken up and rapidly metabolized by explants when it gave rise to digitonin-precipitable sterol (Fig. 2) and was associated with inhibition of sterol synthesis from [I-"Clacetate ( Fig. 1).
Removal of compactin by a brief wash just before assay of protein or fatty acid synthesis caused no reversal of inhibition (Table I). This striking contrast with the response of sterol synthesis to removal of inhibitor confirmed that compactin had no direct reversible inhibitory action on protein or fatty acid synthesis and suggested that its effect was indirect and secondary to inhibition of mevalonate production. If this were so, then removal of compactin by washing followed by reculturing with hormones for a further 24 h should restore normal rates of protein and fatty acid synthesis. Fig. 6 shows the result of such an experiment. Fatty acid and protein synthesis rates were restored to 95 and 79%, respectively, of the control   values when compactin was removed after 48 h and culture was continued in its absence for a further 24 h. Associated with these effects, there was considerable stimulation of sterol synthesis from [l-14C]acetate, resulting in an overshoot to 264% of control values. Thus, no permanent cell damage resulted from culture with 1 O p~ compactin.

DISCUSSION
These results show that, following inhibition of mevalonate and sterol synthesis by compactin in hormonally stimulated mammary explants, a significant reduction of protein and fatty acid synthesis occurred which can be prevented by supply of exogenous mevalonate and can be reversed by culture for 24 h (but not 1 h) in the absence of compactin.
Compactin is a powerful reversible inhibitor of HMG-CoA reductase, competing with HMG-CoA for this enzyme but being without effect on the other enzymes of HMG-CoA metabolism (13). In cultured fibroblasts, it has been shown to inhibit sterol synthesis but to be without inhibitory effect on fatty acid or protein synthesis (Refs. 25 and 16, respectively). In fact, its presence stimulates the appearance of new molecules of HMG-CoA reductase in isolated hepatocytes (23), and it causes increased numbers of low density lipoprotein receptors in the cell surface of fibroblasts (16). However, compactin does inhibit cell growth (2, 3), but this effect can be overcome by the supply of exogenous cholesterol, mevalonate, and products derived from mevalonate (3). Resistance to the growth-inhibiting effects of compactin is associated with very high levels of HMG-CoA reductase (26), indicating that these effects of compactin are due to its specific action in depriving the cell of mevalonate.
Recent work has shown that mevalonate and its products play a complex role in the cell being required for the initiation (4,5) and the continuation (27) of DNA synthesis in the cell cycle. Also by blocking mevalonate synthesis, compactin has been shown to inhibit differentiation in sea urchin (7) and mouse embryos (8). In both cases, this has been associated with the inhibition of protein glycosylations, and in the former case, the effect was reversed by the mevalonate-derived isoprene dolichol.
In view of this evidence for the specificity of effect but widespread and manifold repercussions of compactin treatment of cells, we interpret our results as indicating a specific requirement for mevalonate (or derived products) for the hormonal maintenance of the increased fatty acid and protein synthesis characteristic of the development of the mammary gland. It is possible that these effects reflect specific inhibition of the processes responsive to the lactogenic stimulus (synthesis of medium-chain fatty acids (11) and milk proteins).
It is relevant here to note that medium 199 used in our experiments contains 0.5 PM cholesterol which we have found to be metabolized by mammary explants to give cholesterol ester. The inhibitory phenomena described here occurred in the presence of this metabolically available exogenous cholesterol, suggesting that compactin does not act, in this case, by depriving the cell of cholesterol.
We are currently investigating whether the inhibition of hormonally stimulated fatty acid and protein synthesis by compactin is related to a need for dolichol or other iosprenes.