Cytosolic Proteins That Bind Oxygenated Sterols CELLULAR DISTRIBUTION, SPECIFICITY, AND SOME PROPERTIES*

The differential binding of cholesterol and 25-hy-droxycholesterol to cytosolic proteins in various types of cells was investigated. 25-Hydroxycholesterol taken up by six different established cell culture lines and by mouse spleen cells in primary culture was bound to cytosolic components which, during velocity gradient centrifugation, displayed sedimentation coefficients of approximately 5 S and 8 S. In contrast, cholesterol taken up by the cells was concentrated near the bottom (approximately 20 S) of the sucrose density gradient but was distributed throughout. Results with primary cultures of mouse fetal liver differed from those obtained with other cell cultures in that both sterols appeared principally in a 5 S band. Further character-ization of the binding components from intact L cells indicated that binding of 25-hydroxycholesterol to the 8 S fraction was saturable and reversible, whereas binding to the 5 S band was not saturable. The 8 S 25- hydroxycholesterol complex involved a protein with a relatively long half-life. The complex was essentially stable at 0°C but dissociated slowly at 25°C. Sulfhydryl functions were not required for sterol binding, and formation of the complex was not dependent upon CAMP. Competition studies with intact cells and with isolated cytosol indicated that a number of other oxy- genated sterols bind to 8 S sites occupied by 25-hy-droxycholesterol. Those sterols which are potent sup- pressors of 3-hydroxy-3-methylglutaryl coenzyme A reductase (EC 1.1.1.34) activity competed for these binding sites, while those which do

The differential binding of cholesterol and 25-hydroxycholesterol to cytosolic proteins in various types of cells was investigated. 25-Hydroxycholesterol taken up by six different established cell culture lines and by mouse spleen cells in primary culture was bound to cytosolic components which, during velocity gradient centrifugation, displayed sedimentation coefficients of approximately 5 S and 8 S. In contrast, cholesterol taken up by the cells was concentrated near the bottom (approximately 20 S) of the sucrose density gradient but was distributed throughout. Results with primary cultures of mouse fetal liver differed from those obtained with other cell cultures in that both sterols appeared principally in a 5 S band. Further characterization of the binding components from intact L cells indicated that binding of 25-hydroxycholesterol to the 8 S fraction was saturable and reversible, whereas binding to the 5 S band was not saturable. The 8 S 25hydroxycholesterol complex involved a protein with a relatively long half-life. The complex was essentially stable at 0°C but dissociated slowly at 25°C. Sulfhydryl functions were not required for sterol binding, and formation of the complex was not dependent upon CAMP. Competition studies with intact cells and with isolated cytosol indicated that a number of other oxygenated sterols bind to 8 S sites occupied by 25-hydroxycholesterol. Those sterols which are potent suppressors of 3-hydroxy-3-methylglutaryl coenzyme A reductase (EC 1.1.1.34) activity competed for these binding sites, while those which do not suppress the reductase did not compete for them. These studies suggest that the binding of oxygenated sterols to the 8 S cytosolic component may be in some sense specific, while that to the 5 S component is nonspecific. The kinetics of formation and dissociation of the 8 S 25-hydroxycholesterol*protein complex and of the suppression of 3-hydroxy-3-methylglutaryl-CoA reductase in the presence of the diol was also consistent with a postulated role for the complex in the regulation of the enzyme.

~~~~ ~
Certain oxygenated sterols are potent inhibitors of cholesterol synthesis and of cell growth. They act specifically to lower the cellular level of 3-hydroxy-3-methylglutaryl-CoA reductase, the regulatory enzyme in the cholesterol biosynthesis pathway. How they do so is unknown, but it is known that certain structural configurations of the sterol molecule * This work was supported by Grant CA 02758 awarded by the National Cancer Institute, Department of Health, Education and Welfare. 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. are essential (1,2) and that certain of them are produced as precursors of cholesterol (1,3,4) or as products of its catabolism (1, 2, 5, 6). Inhibitory oxygenated sterols are, therefore, present in cells and may function therein to regulate the rate of cholesterol synthesis.
We have recently described intracellular L cell proteins which bind the inhibitory sterol, 25-hydroxycholesterol, but not cholesterol (7). In the present paper we describe in more detail the cellular distribution, specificities, and binding properties of the proteins. These characteristics of one such protein fraction, that which occupies the 8 S region of sucrose velocity sedimentation gradients, are consistent with the possibility that it plays a role in the regulatory actions of the oxygenated sterols.

Sterols-[26,27-3H]25-Hydroxycholesterol,
[1,2-3H]cholesterol, and [4-14C]cholesterol (New England Nuclear) were purified by thin layer chromatography before use (1,2). The sources of unlabeled sterols and procedures used to purify them have been described (4)(5)(6)8). Sodium tetrathionate dihydrate was from Fluka AG, Switzerland. Cell Cultures-L cell mouse fibroblasts (a subline of NCTC clone 929) were grown either as monolayers in plastic cell culture flasks or in spinner culture in serum-free Waymouth's 752/1 medium modified as described (9). Approximately 1 X lo7 cells were pipetted into each of six 150-cmZ plastic culture bottles (Corning) approximately 24 h before they were used in an experiment. The cells became attached and grew as monolayers under these conditions. Primary cutlures of mouse fetal liver cells and an established line of fetal liver cells (FL83/ B) were grown in serum-free medium as described (5). Chinese hamster lung (Dede) cells and Chinese hamster ovary (CHO) cells were grown in medium containing delipidated fetal bovine serum (4 mg/ml) as described previously (IO). Mouse myeloma (MOPCIOIE) cells were grown in Dulbecco's modified medium containing 5% horse serum (9).
Spleen cells were obtained from six adult C57BL/6J male mice. The spleens were pressed through a fine wire mesh, and the mesh was washed with Waymouth's 752/1 medium. The cells were sedimented by centrifugation at 600 X g at 20°C for 5 min and washed twice with medium. They were then suspended in 18 ml of medium and divided into two equal parts. One part was incubated with [3H]cholesterol, the other with [3H]25-hydroxycholesterol under the conditions described below.
Incubation of Sterols with Cell Cultures-Prior to incubation with a sterol the cultures were refed with 9 ml of fresh medium. Sterol suspended in a solution of 10% (v/v) ethanol and 4.5% (w/v) bovine serum albumin (Sigma, essentially fatty acid free) in culture medium was then added so that the final concentration of albumin was 0.45% and that of ethanol was 1% (5). Routinely the cultures were incubated at 37°C on a gyrating table (40 rpm) for an appropriate period of time after which, while the cells were still adhering as a monolayer, they were washed by flooding the dishes four times with 10-ml volumes of fresh medium free of any additions at 0°C.
In competition experiments with whole cells, after labeling with [3H]25-hydroxycholesterol as described above, the cultures were Cytosolic Sterol-binding Proteins washed four times with 10-ml volumes of medium at 37°C. They were then reincubated with unlabeled sterol after which they were washed three times with 1 0 -d volumes of medium at 0°C. Cells were harvested by scraping them into 13 d of medium and sedimented by centrifugation. All further operations were carried out at temperatures between 0°C and 4°C unless indicated otherwise. To obtain nuclear, microsomal, and soluble fractions, the cells were suspended in hypotonic buffer, homogenized (50 passes) with a glass homogenizer with a Teflon plunger, and fractionated by differential centrifugation as described by Baxter et al. (11). Microscopic examination of the unfractionated homogenate in the presence of trypan blue indicated that approximately 98% of the cells were disrupted. When only the soluble fraction of the cells was required, the cells were homogenized in a hypotonic buffer, pH 7.8 (Buffer A) containing 20 p~ Tris, 1 p~ MgC12,2 p~ CaClz, and 2 m M dithiothreitol. A solution of 30% sucrose (w/w) in Buffer A was then added to give a final concentration of 5% sucrose. Subcellular fractions were dissolved in Soluene (New England Nuclear) and assayed for ' H and "C in toluene base scintillation fluid. Sucrose density gradient centrifugation (5% to 20% sucrose), collection of fractions, and assays of the fractions for 14C and 'H were carried out as described (7), except that the sample was centrifuged with the gradient for either 20 h at 40,000 rpm or 40 h at 35,000 rpm. Sterol concentrations in cells or in subcellular fractions were expressed in terms of protein or, when comparisons were between essentially identical cultures, as amounts derived from one culture.
Incubation of Sterols with Isolated Cytosol-Sterols dissolved in 15 pl of freshly distilled absolute ethanol were placed in the bottom of a 10-ml test tube siliconized by treatment with Siliclad (Clay-Adams). The cytosolic fraction (1 &), isolated from cultured cells as described above, containing from 1.2 to 1.8 mg of protein, was added to the tube, shaken gently, and incubated either in an ice bath or, where indicated, in a water bath at 25°C. The mixture was then fractionated by sucrose density gradient centrifugation as described above.
Other Assays-HMG'-CoA reductase activity in whole cell homogenates was determined as described previously (12). Protein was determined by a modification of the Lowry procedure (13) or by a dye-binding assay (14) using reagents purchased from Bio-Rad. The incorporation of ['Hlleucine into perchloric acid-precipitated proteins was determined by incubating cultures with ~-[4,5-~H]leucine at concentrations of 2 pCi (0.03 mmol)/ml of medium for 20 min. The cultures were then scraped from the flasks, sedimented by centrifugation, and extracted three times with 5-ml volumes of 6% perchloric acid and once with 5 ml of 90% ethanol. The residue was then dissolved in a solution of 0.3 N NaOH containing 1.7% sodium dodecyl sulfate. The mixture was neutralized with HCI, and aliquots were taken for protein and tritium assays. The relationship between the two curves illustrating total cell uptake with increasing time (Fig. 1A) is consistent with that reported previously; the uptake of 25-hydroxycholesterol was greater than that of cholesterol over the entire time period. The relative proportions of the two sterols found at various time intervals in these crude soluble (cytosolic), nuclear, and microsomal fractions did not differ greatly from that in the whole cells, and the data, therefore, do not provide clear evidence for the preferential subcellular localization of either sterol. The concentrations of the two sterols were higher in the nuclear fraction than in the microsomal or cytosolic frac- The abbreviation used is: HMG, 3-hydroxy-3-methylglutaric acid. and 100 PM and the incubation period between 15 min and 2 h. Additional similar experiments with long-term fetal liver cell cultures gave results similar to those obtained with L cell cultures. The bulk of the recovered sterols was found in the particulate fractions, presumably because of their association with lipid-rich membrane elements. A relatively small proportion (1 to 2%) of the sterols taken up by the cells was recovered in the cytosolic fraction. In this fraction, however, was found a class of sterol-binding sites with potentially interesting properties, and it is these sites and their properties with which the remainder of this paper deals.

Subcellular Localization a n d Cell
Differential Binding of Cholesterol a n d 25-Hydroxycholesterol to Cytosolic Proteins in Various Lines of Cultured Cells-Although only a small fraction of each sterol taken up by L cells was found in the cytosol, previous studies ( 7 ) showed that the hydroxysterol and cholesterol were bound to different protein fractions. Fig. 2A shows that the sedimentation patterns for 25-hydroxycholesterol and cholesterol bound to the cytosolic proteins of an established fetal liver cell line were similar to those found with L cells ( 7 ) . Two Cells were homogenized and fractionated as described in the text, and portions of each fraction were assayed for radioactivity.
for the other nonhepatic cells. The banding patterns for cholesterol in cytosolic fractions from nonhepatic cells in Fig.  2B are not shown, but they were generally similar to that shown for the established liver cell in Fig. 2 A and that shown previously for L cell cultures (7). In no case was the banding pattern for cholesterol in the cytoplasm of any of the cultures shown in Fig. 2B similar to that found for 25-hydroxycholesterol.
Only one of the types of cell cultures tested, namely primary cultures of mouse fetal liver cells, exhibited sedimentation patterns for bound sterols which appeared to be qualitatively different from those shown in Fig. 2. Cytosol from primary fetal liver cell cultures showed only one major band (approximately 5 S) with a slight shoulder corresponding roughly to an 8 S component, whether the cells were incubated either with [3H]25-hydroxycholesterol or [3H]cholesterol (Fig. 3A). Primary cultures of fetal liver cells differ from the long-term cell line and from other established cell cultures in many respects, including a lower sensitivity to the inhibitory effects of most oxygenated sterols upon cholesterol synthesis (2, 5, 6), a much greater capacity to take up cholesterol and oxygenated sterols from the medium (2-6), and a greater ability to metabolize the oxygenated sterols, at least in comparison to L cells (4). The differences in sensitivity to, and uptake of, sterols might be accounted for by the presence in the primary fetal liver cell cultures of one or more liver-specific proteins which bind both cholesterol and oxygenated sterols. The presence in liver of several different cytoplasmic proteins, each of which is capable of binding a variety of lipids including sterols, has been reported (16-19). The presence of such proteins could also account for the distinctive binding pattern for 25-hydroxycholesterol in the cytoplasm of the primary liver cell culture shown in Fig. 3A, with little evidence for an 8 S and 5 S component, and for the apparent qualitative similarity between the binding of 25-hydroxycholestero1 and cholesterol. In contrast to the primary fetal liver cell cultures, the banding patterns for the established fetal liver cell line ( Fig. 2A)   6, 15). Evidence that differences in banding of the two sterols between primary fetal liver cell cultures and established cell cultures were not due to adaptational changes of cells to the conditions of long-term culture was provided by the results obtained with mouse spleen cells (Fig. 3B). The banding pattern for 25-hydroxycholestero1 was similar to that of the cultured cells shown in Fig. 2 Fig. 4. Previously (7) we showed that bound 25-hydroxycholestero1 was lost from the 8 S and 5 S bands when the labeled cytosol was incubated with a protease, suggesting that the sterol was bound to protein(s) in those fractions. That sulfhydryl groups were not strongly involved in the binding of the diol to the protein was shown by the following experiments. When dithiothreitol was omitted from the buffer, the recovery of 8 S and 5 S sterol in the cytosolic fraction was diminished about 30%.
However, the addition of p-hydroxymercuribenzoate (1 mM) to the buffer in the absence of dithiothreitol did not further depress the amounts of sterol recovered in these bands. Treatment of the isolated, labeled cytosol with sulfhydryl-oxidiz1ng Proteins  (Table I) bound to the 8 S and 5 S cytosolic components appeared to increase somewhat. Under these conditions, incubation for 30 min with cycloheximide resulted in an 86% decrease in the incorporation of C3H]leucine into perchloric acid-precipitable material. In a single experiment, incubation of L cells for 5 h with actinomycin D (10 pg/ml) diminished HMG-CoA reductase activity to 18% of the control value but did not alter the amount of 25-hydroxycholesterol bound to the 8 S cytosolic component.
Kinetics of 25-Hydroxycholesterol Binding to Cytoplasmic Proteins-The results shown in Fig. 5 provide evidence that the binding of 25-hydroxycholesterol to the 8 S cytosolic protein is a saturable process. The proportion of added diol that was bound to the 8 S protein declined with increasing sterol concentration, whereas binding to the 5 S component was linear within the range of concentrations employed (Fig.  5). The linear relationship between the binding of 25-hydroxycholesterol to the 5 S component and sterol concentration, along with information regarding the kinetics of cellular 25hydroxycholesterol uptake over a much wider range of concentrations ( Fig. 1 and Ref. 14), is evidence that the hyperbolic curve shown for the 8 S band in Fig. 5 reflects the character of the binding reaction and is essentially independent of unrelated uptake processes.
A steady state equilibrium between free 25-hydroxycholesterol and that bound to the 8 S cytosolic protein appeared to be reached soon after the sterol, at a concentration of 0.1 p~, was added to the medium (Fig. 6A). Maximum binding to both the 8 S and 5 S components was attained after approximately 30 min of incubation at 37°C using a single concentration of sterol. Following the removal of exogenous [3H]25hydroxycholesterol and incubation of the labeled cells in protein-free medium, or in medium containing 0.45% albumin, the amount of extractable 8 S complex declined slowly (Fig.  6B). The extent of dissociation of the 8 S complex after 20 min in protein-free medium and in medium containing 0.45% albumin was 19% and 30%, respectively. In contrast, 86% dissociation occurred after 20 min in medium containing 1% albumin. Apparent losses of labeled sterol from the 5 S band of as much as 50% occurred during the first 20 min of incubation in sterol-free medium. However, the observed decline in the 5 S complex appeared to be relatively independent of the presence of albumin in the medium and also of time after the first 20 min of incubation. These results indicated that dissociation of the 8 S complex was related to efflux of sterol   Fig. 7 show that, under conditions wherein reductase activity in L cells has been suppressed by a concentration of 25-hydroxycholesterol adequate to saturate the 8 S protein(s) (12.5 p~) , subsequent dissociation of the 8 S complex following the removal of exogenous sterol is correlated with the recovery of HMG-CoA reductase activity. In Fig. 7A, unlabeled 25-hydroxycholesterol was added to the medium at zero time at a level of 12.5 p f , the sterol was removed after 20 min, and the culture was then incubated in protein-free medium. After various periods of time a tracer concentration of [3H]25-hydroxycholesterol was added to the medium, and the incubation was continued for 20 min. HMG-CoA reductase activity and binding of the labeled sterol to the 8 S component were then determined. In cells previously treated with a saturating concentration of unlabeled 25-hydroxycholesterol, the extent to which the [3H]sterol was bound represented dissociation of the unlabeled sterol.protein complex. Following the removal of the unlabeled sterol, reductase activity continued to fall for at least 1 h, and neither dissociation of the 8 S complex nor recovery of reductase activity was detected over 18 h of incubation. In Fig. 7B, the diol was added at zero time and was removed from the culture after 2 h of incubation in order to allow time for nearly complete suppression of reductase activity. Upon reincubation of the culture in medium containing 1%    cellular interactions between sterols added together to the medium. Unlabeled 20a-hydroxycholesterol (a potent suppressor of HMG-CoA reductase activity) as well as 25-hydroxycholesterol competed for 8 S sites occupied by C3H]25hydroxycholesterol. In contrast, the 25,26,27-nor homologue of 20a-hydroxycholesterol (20-propyl-5-pregnene-3P,20adiol), which does not suppress the reductase (6), did not displace the labeled sterol. The abilities of unlabeled inhibitory sterols to occupy 25-hydroxycholesterol-binding sites on the 8 S protein were also evident when the sequence of the two incubations was reversed (7); results similar to those shown in Fig. 8 were obtained when the cells were incubated with the unlabeled sterols (20 PM) for 20 min, then washed and reincubated with [3H]25-hydroxycholesterol (0.1 mM) for 30 min.

-
A survey of competition by various unlabeled sterols for 8 S sites occupied by [3H]25-hydroxycholesterol in intact L cells is summarized in Table 11. Some correlation between the abilities of the sterols to compete for the 8 S sites and their previously measured abilities to suppress HMG-CoA reductase activity is apparent. The two monofunctional sterols and the 25,26,27-nor diol at the top of the list do not suppress HMG-CoA reductase in cell cultures and showed little ability to compete for 25-hydroxycholesterol 8 S-binding sites. The competitive activities of the following 3P-hydroxy sterols functionalized in position 25, 15, 6, 20, 32, 22, or 7 were roughly consistent with their previously reported activities as inhibitors of HMG-CoA reductase activity. The failure of 5a-lanost-8-ene-3/3,32-diol to compete effectively in contrast to the greater activity of its A7 isomer may be noteworthy. The A8 sterol is converted into cholesterol in L cells more rapidly than the A' isomer, but both sterols give rise to other products as well (4). Thus, competitive ability measured in the test with intact cells may be influenced by metabolism of the sterol being tested. A further uncertainty in tests with intact cells is due to a lack of information regarding the intracellular concentrations of the test sterol that are available for binding to the 8 S protein. It is possible that concentrations of weak competitors cannot be elevated high enough to give observa-ble displacement of [3H]25-hydroxycholesterol. Displacement studies utilizing multiple concentrations of each competitor will be necessary to fully characterize the effectiveness of these compounds in this assay system. Nevertheless, the rough concordance of potency with ability to displace 25-hydroxycholesterol in the survey shown in Table I1 is striking.

Competition for 8 S-binding Sites in Isolated L Cell Cyto-
sol-Problems regarding intracellular sterol concentrations and sterol metabolism would be diminished or eliminated under conditions wherein sterols were incubated with the isolated cytosolic fraction at 0°C. As shown in Fig. 9, binding of [3H]25-hydroxycholesterol to both 8 S and 5 S components was detected under such conditions. However, base-line levels of bound radioactive sterol were variable and much higher than those observed when binding of the sterol occurred in intact cells. Attempts to reduce the background levels of radioactive sterol by adsorption with dextran-coated charcoal at 0°C (20) or at 25°C were ineffective. Brief exposure (5 min) of the labeled cytosol to charcoal at 0°C had only a minimal effect upon the pattern, whereas more prolonged exposure (1 h) at 0°C or a short (20 min) period of exposure at 25°C  considerably diminished 3H-sterol throughout the gradient, including that in the 8 S complex.
Accurate quantitation of the 8 S band was not possible under the conditions shown in Fig. 9. Therefore, additional competition studies to examine the specificity of the 8 Sbinding protein were carried out with cytosol isolated from L cells that had been previously incubated with [3H]25-hydroxycholesterol. Preliminary experiments showed that the 8 S and 5 S complexes were nearly stable for as long as 24 h when the isolated cytosol was incubated at 0-4°C. Furthermore, the presence of unlabeled 25-hydroxycholesterol added to the cytosol as an ethanolic solution (15 y l / d of cytosol) in concentrations as high as 750 nM (approximately a 600-fold excess over the labeled sterol present in the cytosol) had only a minor effect upon the amount of bound 3H-sterol in the 8 S band at these temperatures, decreasing it by only 10% at the highest sterol concentration tested. In contrast, at 25°C dissociation of the 8 S complex appeared to be biphasic. Approximately 10% of the bound r3H]sterol was lost during the first hour after which a slower, linear rate of loss obtained for at least 9 additional h (Fig. 10). The 5 S complex also showed an initial rapid rate of dissociation during the first hour. However, further dissociation with continued incubation was not apparent. The inset in Fig. 10 shows that the addition of unlabeled 25-hydroxycholesterol to the labeled cytosol after it had been incubated for 1 h at 25°C increased the rate at which the E"H]sterol was lost from the 8 S band. Presumably, this effect of the unlabeled sterol was due to its ability to occupy vacant binding sites, thus inhibiting reassociation of released A number of other unlabeled sterols were tested for their abilities to compete with [3H]25-hydroxycholesterol for 8 S sites in isolated cytosol under conditions similar to those in Fig. 10. This procedure for testing competition for binding sites with isolated cytosol is cumbersome and insensitive. The dissociation versus time plot is complex, and it is not known whether or not some part of the observed dissociation is due to denaturation of the binding protein(s) and, therefore, irreversible. For these reasons, the results of competition experiments with isolated cytosol as carried out, would not, by themselves, constitute strong evidence for the specific binding of oxysterols, and they are not presented in detail. The data do, however, c o n f i in general the conclusions drawn from competition experiments with intact cells. Unlabeled sterols were tested in amounts 200-to 2000-fold greater than the amount of [3H]25-hydroxycholesterol present in the incubation mixture as an 8 S complex (approximately 70 fmol/mg of cytosolic protein). Differences between the sterols that were seen included the following. 5a-Cholest-8( 14)-en-3P-ol-15-one and 5a-lanost-7-ene-3P,32-diol appeared to compete as effectively for [3H]25-hydroxycholesterol-binding sites as did unlabeled 25-hydroxycholesterol. However, as in tests with intact cells, 5a-lanost-7-ene-3P,32-diol was a better competitor than the As isomer. The 7-functionalized sterols were relatively poor competitors, and two sterols which did not suppress HMG-CoA reductase (cholesterol and 20-propyl-pregnene-3P,20a-diol) were not effective competitors.

DISCUSSION
The present results extend in several directions our knowledge of cytosolic proteins that bind 25-hydroxycholesterol.
The 8 S and 5 S 25-hydroxycholesterol-binding proteins were present in a variety of established cell culture lines and in a primary culture of mouse spleen cells. These proteins did not bind an appreciable amount of cholesterol, which was largely associated with complexes of greater density. Primary cultures of fetal mouse liver cells differed in that 25-hydroxycholesterol and cholesterol appeared to bind similarly to components of the cytosol which sedimented essentially as a single band of approximately 5 S with only a trace of 8 S material. It seems likely that the binding of both sterols in the same fraction of the cytosol of the primary fetal liver cultures reflects the presence of one or more liver proteins which are able to bind a wide range of lipids (19-22). Binding of the sterols to such protein($ in the primary cultures of fetal liver may have masked any of the more specific 25-hydroxycholestero1.8 S complex formed. If this explanation is correct, some of the proteins which bound 25-hydroxycholesterol nonspecifically were lost during the development of the established line of fetal mouse liver cells, unmasking the presence of the 8 S and 5 S binding proteins. It is also possible that the latter proteins only began to be expressed after the line became established. Preliminary experiments indicate that the 8 S binding protein(s) is not identical with receptors for glucocorticoids or the sex steroids. First, as noted above, the 8 S oxygenated sterolbinding protein(s) is not particularly sensitive to reagents known to promote oxidation of thiol groups, whereas the steroid hormone receptors are exquisitely so (21,22). Second, the sterol-binding protein(s) appears to be less thermolabile than are the steroid hormone receptors (21,22). Third, under some conditions steroid hormone receptors are known to sediment as 8 S components on sucrose density gradients (21,22). However, two types of experiments indicated that steroid hormones do not bind specifically to the oxysterol-binding 8 S component. Direct steroid hormone-binding studies and studies of the inhibition of [3H]25-hydroxycholesterol binding gave no indication that cortisone, dexamethasone, testoster-Cytosolic Sterol-binding Proteins one, or 17P-estradiol bind to the 8 S oxysterol-binding pro-tein@) (data not shown). Furthermore, limited studies suggest that the 8 S binding protein(s) is not identical with the sterolbinding proteins previously described by others. Collaborative comparisons of the sedimentation rates of the sterol-binding proteins investigated by Dr. T. J. Scallen and his associates (18) and by Dr. J. L. Gaylor and his associates (19) indicated their nonidentity with the 8 S protein that binds oxygenated sterols.
The binding of 25-hydroxycholesterol to form 8 S and 5 S complexes involves the participation of proteins (7). The results of studies with cycloheximide suggest that the halflives of the 8 S and 5 S binding proteins are longer than 14 h. The formation of the complexes within the cells did not appear to involve CAMP. Several lines of evidence indicate that binding of oxygenated sterols to the 8 S protein(s) differs qualitatively from their binding to the 5 S band. First, the kinetics of 25-hydroxycholestero1 binding to the 8 S protein in intact cells indicate that it is saturable and reversible. In contrast, binding to the 5 S proteins was not saturable at the same concentrations of sterol. Second, evidence that the 8 S sites to which 25-hydroxycholesterol was bound were specific for oxygenated sterols which can suppress HMG-CoA reductase was obtained through two kinds of competition assays. Strong suppressors of HMG-CoA reductase were generally good competitors for the 8 S, 25-hydroxycholesterol-binding sites in intact L cell cultures. However, 5a-lanost-8-ene-3/3,32diol, which was 50 times less potent than 25-hydroxycholesterol as a suppressor of HMG-CoA reductase activity, showed little or no ability to compete for sites occupied by the labeled diol under these conditions. The failure of the A'-lanostenol derivative to compete effectively in assays with intact cells might be due to its metabolism by the cell to inactive sterols, including cholesterol (4). Results obtained in assays with isolated cytosol were generally similar to those obtained with intact cells, and again the A*-derivative failed to compete well, although it is unlikely that significant metabolism of the sterols occurred in the cytosolic fraction under the conditions of the assay. No evidence for metabolism of either of the lanosterol derivatives was obtained in an experiment' wherein the 2-3Hz-labeled compounds (15,500 dpm for the A7-sterol, 31,000 dpm for the A* isomer) were incubated at a concentration of 2.3 p~ with 1 ml of cytosol (2.2 mg of protein) for 7 h at 25"C, then extracted and analyzed by thin layer chromatography as described previously (4). Clearly, the sensitivities of both competition assays may be limited by the solubilities (unknown in most instances) of the sterols in aqueous solutions and by their propensities to bind nonspecifically to various substances including glass and plastic. It may, therefore, have been impossible in these experiments to achieve sufficiently high concentrations of, or appropriate incubation conditions for, some sterols, which have low affinities for binding sites, to demonstrate competition with 25-hydroxycholesterol.
With the aforementioned exception, oxygenated sterols which suppress HMG-CoA reductase activity in cultured cells were able to compete for 8 S 25-hydroxycholesterol-binding sites in at least one of the two assay systems. Sterols which do not suppress the reductase, including the 25,26,27-nor homologue of the potent inhibitor, 20a-hydroxycho1estero1, did not compete for 8 S sites. Thus, the 8 S protein fraction appears to bind inhibitory sterols relatively specifically, a property consistent with a role in the regulation of HMG-CoA reductase activity. The kinetics of the formation and dissociation of the 25-hydroxycholesterol. 8 S complex were also consistent with A. A. Kandutsch and E. B. Thompson, unpublished data.
its involvement in the regulation of the reductase. Formation of the complex in intact cells was rapid enough to precede any appreciable decline in HMG-CoA reductase activity, and dissociation of the complex preceded any rise in reductase activity from a depressed level.
The actual role, if any, of the 8 S sterol -protein complex in the regulation of HMG-CoA reductase is presently only a matter for conjecture, since it has not been established whether oxygenated sterols suppress the synthesis of HMG-CoA reductase, accelerate its degradation, or inactivate it by some indirect mechanism (reviewed in Ref. 1). 25-Hydroxycholesterol does not inactivate the enzyme directly (5, 6) nor does it reversibly inactivate the enzyme indirectly via a system that appears to involve reversible phosphorylation, which is currently under study in several laboratories (12). Furthermore, L cell microsomal HMG-CoA reductase activity was not affected by preincubation with cytosol containing 25-hydroxycholesterol. protein complexes.' Activity values were not significantly different when the microsomal fraction (850 pg) of protein was preincubated for 20 min at 37°C with cytosol (225 pg of protein) from control L cells or from L cells that had been allowed to take up 25-hydroxycholestero1, present in the medium at concentrations of 40 ng/ml or 10 pg/ml (data not shown). On the other hand, a good deal of evidence indicates that most of the known regulatory fluctuations in HMG-CoA reductase activity that occur in vivo involve changes in the rate of enzyme synthesis (see Refs. 1 and 23 for reviews). In our experience temporal changes in reductase activity following the addition of oxygenated sterols to the medium are consistent with alterations in the synthesis of an enzyme with a half-life of 1 to 2 h (6); TI/' values as low as 1 h for the rat liver enzyme have been reported (24). If further studies demonstrate that the oxygenated sterols affect the rate of reductase synthesis then we expect that models for the action of steroid hormone receptors may be relevant to that of the 8 S oxygenated sterol-binding protein.