Sterol-mediated Regulation of Mevalonic Acid Synthesis ACCUMULATION OF 4-CARBOXYSTEROLS AS THE PREDOMINANT STEROLS SYNTHESIZED IN A CHINESE HAMSTER

Chinese hamster ovary-215 (CHO-215) mutant cells are auxotrophic for cholesterol. Berry and Chang (Berry, D. J., and Chang, T. Y. (1982) Biochemistry 21, 573-580) suggested that the metabolic lesion was at the level of 4-methyl sterol oxidation. However, the observed cellular accumulation of lanosterol was not consistent with a defect at this metabolic site. With the use of a novel Silica Sep Pak sterol separation procedure, we demonstrated that 60-80% of the acetonesoluble lipid radioactivity in [5-3H]mevalonate-labeled CHO-215 cells was incorporated into acidic sterols. 7(8),Cholesten-4 beta-methyl,4 alpha-carboxy,3 beta-ol was the dominant end product. In addition to this acidic sterol, 7(8),24-cholestadien,4 beta-methyl,4 alpha-carboxy,3 beta-ol and 7(8),24-cholestadien,4 alpha-carboxy,3 beta-ol were also isolated. Incubation of cell-free extracts with [3H]7(8)-cholesten-4 beta-methyl, 4 alpha-carboxy,3 beta-ol and pyridine nucleotides confirmed that CHO-215 4-carboxysterol decarboxylase activity was less than 1% of that for wild type cells. Thus, a correspondence between decreased 4-carboxysterol decarboxylase activity and the spectrum of accumulated sterol products by intact CHO-215 cells was demonstrated. No detectable cholesterol was synthesized by CHO-215 cells. 3H-Product accumulation studies demonstrated that 7(8),24-cholestadien, 4 beta-methyl,4 alpha-carboxy,3 beta-ol increased prior to its subsequent saturation at the delta 24 carbon. Furthermore, the steady state ratio for delta 24-saturated acidic sterols/unsaturated acidic sterols was dependent on media cholesterol source and amount. Finally, the accumulated acidic sterol(s) were not regulatory signal molecules for the modulation of 3-hydroxy-3-methyl-glutaryl coenzyme. A reductase activity in response to cholesterol availability.

Recently, Panini et al. (9) have reported that the major sterol product synthesized by CHO-215 cells was lanosterol and possibly 14-desmethyl lanosterol.
In this laboratory we wanted a cholesterol auxotroph which could be used to test the oxysterol hypothesis (10-13). That is, to determine directly (without the contribution of endogenously synthesized cholesterol) whether down-regulation of HMG-CoA reductase activity by regulatory levels of exogenously added highly purified cholesterol correlated with the formation of biologically synthesized oxysterol(s). CHO-215 cells were chosen as our model. However, in view of the apparent uncertainty about the CHO-215 cell metabolic lesion(s), it was important to (i) firmly define the sterologenic defect and (ii) to determine whether the cell did/did not synthesize C-27 sterols in general or cholesterol specifically.
In contrast to the conclusions published by Chang et al. (2) Fig. 2) of Sep-Pak fraction IV yielded three distinct peaks with retention times of 6.0, 6.8, and 7.8 min. Each radioactive peak was isolated by repetitive HPLC, checked for purity by TLC, and analyzed by high and low resolution mass spectrometry.

TIME IN MINUTES
Rs values from the TLC analysis of each purified acidic sterol are summarized in Table II (Table II) for TLC of the 7.8-min peak compound in solvent systems used by others (3, 24) to resolve acidic sterols allowed us to tentatively identify it as 7(8)-cholesten-4&methyl,4a-carboxy,3&ol (Fig. 3  The compound with a 6.8-min retention time gave a m+ at m/z 442. Characteristic fragmentation consistent with the  Table II were consistent with the compound with a 6.8-min retention time being 7(8),24-cholestadien,4/3-methyl,4a-carboxyl,3P-o1 (Fig. 3). at the times noted and processed for Sea-Pak fractionation as described under "Materials and Methods."-The acidic sterol fraction was resolved as described in the legend to Fig. 3. Radio-HPLC profiles are presented. Equivalent acidic sterol aliquots/l50cm' flask were injected.
It should be noted that cells incubated in media which contained BZ or lipoprotein poor serum grew at normal growth rates for 36-48 h and began to dislodge from the flasks after 72 h of incubation (( 1) and results not shown).
To ascertain whether acidic sterols were secreted into the medium, CHO-215 cells were prelabeled with [3H]mevalonate (24 h) washed with McCoy's 5A medium and refed with media which contained no mevalonate and either 5% FCS or 5% BP. Media and cells were assayed separately for radioactive sterols over a subsequent 48-h incubation period. Less than 10% of the initial radiolabeled cellular neutral and acidic sterols appeared in the media. Radioactive cellular neutral sterols decreased approximately 90%, and there was a similar increase in the cellular acidic sterol fraction (data not presented). The acidic sterols were localized primarily (94%) in the 100,000 X g particulate fraction of sonicated cell-free extracts. Thus, CHO-215 acidic sterols were not secreted but accumulated in cellular membranes. Steady State Concentration of Acidic SteroLs-In order to determine the steady state concentration of the acidic sterols, we isolated [3H]7(8),cholesten-4&methyl,4a-carboxy,3@-ol by preparative HPLC and determined its mass by acetylation with [1-'Clacetic anhydride of known specific activity (18.7 mCi/mmol). Different aliquot sizes of an isopropyl alcohol solution of the pure unacetylated [3H]4-methyl,4-carboxysterol were resolved by reverse-phase (C,,) HPLC and their peak heights at 210 nm measured. A 4-methyl,4-carboxysterol peak height versus mass standard curve was generated and used to obtain this sterol's steady state concentration level in CHO-215 cells grown for 48 h in medium which contained 10 pg of cholesterol/ml of the CH-PC dispersion. Total steady state acidic sterol content was determined to be 3-4 pg/mg protein. Cellular cholesterol's concentration was 18-25 wg/mg protein.
Non-acidic Sterols-Examination of Sep-Pak fraction II by reverse-phase (Cis) HPLC (Fig. 5) showed that labeled lanosterol and dihydrolanosterol were the dominant mono-oxygenated sterols. However, a radioactive peak which migrated with a retention time similar to that for cholesterol (peak 1) was also detected, ubiquinone was not eluted under these conditions.
Normal-phase radio-HPLC (Si-CN column) analysis of CHO-215 cells' Sep-Pak fraction II showed that 52% of the injected radioactivity was in lanosterol and dihydrolanosterol, 16% in an unidentified isopentenoid which eluted with cholesterol, and the remaining 32% was in ubiquinone (data not presented).
Although Berry and Chang (3) reported (based on TLC migration) that CHO-215 cells synthesizedcholesterol at ~5% of the wild type rate, it was important to determine, with our analytical system, whether this observation was confirmed. The unknown radioactive peak which migrated as cholesterol Both acetylated and unacetylated derivatives of the unknown radioactive compound were resolved from the cholesterol standards by HPTLC plates impregnated with or without AgN03 (Table II). Since our reverse-phase HPLC system would have resolved A7, A5,7-, and A5-cholestenols, it was concluded that the unknown compound was probably not a CZ7 sterol and definitely not cholesterol. It was possible that the unidentified sterol was identical to one of the C&-C& sterols tentatively identified by Berry and Chang (3 6). However, the CHO-215 cell homogenate yielded no detectable neutral sterols. The only detectable product (l-2% of the input radioactvity) with CHO-215 extracts migrated as ["H]7(8)-cholesten,4oc-carboxy,3P-o1.
Thus, the in vitro assay confirmed the intact cell observations (Table II and Fig. 4). CHO-215 decarboxylation of 4/3-methyl,4Lu-carboxysterols was strongly decreased and decarboxylation of 4cr-carboxysterols was not detectable.
HMG-CoA Reductase Activity in CHO-215 Cells-Since CHO-215 cells accumulated primarily acidic oxysterols (Table  II) and not lanosterol as reported previously (2, 3) it was important to determine whether their amount correlated with the modulation of HMG-CoA reductase activity. Therefore, we measured reductase activity and determined 4-carboxysterol amount in extracts from CHO-215 cells which had been maintained in media which contained different cholesterol sources. Our results are summarized in Table III.
Whereas there was a correlation between the availability of media cholesterol and CHO-215 HMG-CoA reductase activity as reported before (1, 3), there was none with total 4-carboxysterol level. Therefore, 4-carboxysterols (acidic oxysterols) were unlikely regulatory molecules responsible for the modulation of reductase activity. DISCUSSION We demonstrated that [3H]mevalonate-labeled CHO-215 mutant cells, accumulated 4-carboxysterols (Figs. 2 and 4) and not lanosterol or 14-desmethyl lanosterol as reported by others (2,3,9). This is the first report of cellular 4-carboxysterol accumulation.
The ratio of A24,4-carboxysterols:24,25-dihydro, 4-carboxysterols was dependent on exogenous cholesterol source. Continuous [3H]mevalonate labeling studies with CHO-215 incubated with a cholesterol source (B2) which did not support long term growth resulted in a steady state ratio of 1:4 (Figs. 2 and 4). However, exogenous cholesterol sources (FCS and CH-PC) which supported continuous CHO-215 growth established steady state ratios of A24,4-carboxysterols:24,25-dihydro, 4-carboxysterols close to 1.0 (Fig. 2). This distribution did not reflect the steady state lanosterol/dihydrolanosterol ratio of 4:l reported for CHO-Kl cells grown in the presence of FCS (25). Thus, it appeared that 24,25-dihydro, 4-carboxysterol accumulation was facilitated by the lack of A24, 4carboxysterol conversion to CZ7 sterols. Berry and Chang (3) reported that approximately 5% of the radioactivity incorporated into CHO-215 cell neutral sterols (Sep-Pak fractions I-III) was in C&-C& sterols and/or cholesterol. We demonstrated that the product which migrated with a HPLC (Fig. 5) retention time similar to cholesterol was not this sterol. Unfortunately, insufficient amounts of the unknown material was available for determination of its identity. However, migration characteristics on HPTLC suggested that it might be a 4-methyl sterol.
Since the end products synthesized by CHO-215 mutant cells were oxidized at C-4, it was possible that they could act as "oxysterol" regulatory signal molecules (11) to modulate HMG-CoA-reductase activity. However, endogenously synthesized 4-carboxysterols accumulated independent of cholesterol availability (Table III, Fig. 2). Therefore, 4-carboxyster-01s were unlikely participants in the regulatory cascade which modulated HMG-CoA reductase activity.
The finding that CHO-215 cells did not synthesize detect-