Further characterization of a somatic cell mutant defective in regulation of 3-hydroxy-3-methylglutaryl coenzyme A reductase

Proto oncogene/eukaryotic translation initiation factor (eIF) 4E attenuates mevalonate-mediated regulation of 3-hydroxy-3-methylglutaryl coenzyme A (HMG-CoA) reductase synthesis

The rate-limiting enzyme for mevalonate synthesis in mammalian cells is 3-hydroxy-3-methylglutaryl coenzyme A (HMG-CoA) reductase. Products of mevalonate synthesis are required for cell cycle progression as well as cell growth and survival. In tumor cells, HMG-CoA reductase is generally elevated because of attenuated sterol-mediated regulation of transcription. However, tumor cell HMG-CoA reductase remains sensitive to post-transcriptional regulation by mevalonate-derived isoprenoid intermediates of cholesterol synthesis. Isoprenoids suppress HMG-CoA reductase synthesis through a mechanism that reduces initiation of translation on HMG-CoA reductase mRNA. Because HMG-CoA reductase mRNA transcripts have 5'-untranslated regions (UTR) that are GC rich and contain stable secondary structure, we tested the hypothesis that overexpression of eIF4E would attenuate isoprenoid-mediated regulation of HMG-CoA reductase. eIF4E is elevated in many tumor cells and behaves as a proto-oncogene by aberrantly translating mRNAs whose translation is normally suppressed by 5-UTRs that are GC rich. A CHO cell line expressing high levels of eIF4E (rb4E) was developed by infecting cells with retroviruses containing a full-length mouse cDNA for eIF4E. Levels of reductase synthesis were elevated fivefold in rb4E cells compared to noninfected CHO cells; HMG-CoA reductase mRNA levels were not increased in rb4E cells compared to normal CHO cells. Total cellular protein synthesis was only increased by approximately 15% in rb4E cells compared to CHO cells. The mTOR inhibitor rapamycin lowered HMG-CoA reductase synthesis by 50 and 60% in rb4E and CHO cells, respectively; no equivalent effect was observed for HMG-CoA reductase mRNA levels with rapamycin treatment. These results indicate that HMG-CoA reductase mRNA is in a class of mRNAs with highly structured 5'-UTRs whose m(7)GpppX cap-dependent translation is closely linked to the rapamycin-sensitive mitogen activated pathway for protein synthesis.

Isolation of a somatic cell mutant resistant to the induction of apoptosis by oxidized low density lipoprotein

Oxidized low density lipoprotein (oxLDL) induces apoptosis in macrophages, smooth muscle cells, and endothelial cells. To elucidate the molecular mechanism of oxLDL-induced cytotoxicity and determine its tissue specificity, we have used Chinese hamster ovary (CHO)-K1 cells expressing human CD36 (CHO/CD36). Expression of CD36 rendered these cells susceptible to killing by oxLDL. This cytotoxicity was due to the induction of apoptosis. Therefore, CD36 expression is the only requirement for oxLDL-induced apoptosis. Oxysterols apparently mediate the cytotoxicity of oxLDL in macrophage foam cells and endothelial cells. 25-Hydroxycholesterol, at concentrations higher than 1 microg/ml, killed CHO-K1 cells, by apoptosis, in medium supplemented with serum as a source of cholesterol. These effects were not seen in a 25-hydroxycholesterol-resistant CHO/CD36 mutant (OX(R)), which was otherwise capable of undergoing apoptosis in response to staurosporine. This mutant was also resistant to killing by oxLDL, suggesting that oxysterols are at least partially responsible for the toxic effects of oxLDL. Oxysterol-induced apoptosis did not involve regulation of sterol regulatory element-binding protein proteolysis or the cholesterol biosynthetic pathway. 25-Hydroxycholesterol stimulated calcium uptake by CHO-K1 cells within 2 min after addition. Treatment of CHO or THP-1 (macrophage) cells with the calcium channel blocker nifedipine prevented 25-hydroxycholesterol induction of apoptosis. OX(R) showed no enhanced calcium uptake in response to 25-hydroxycholesterol.

Down-regulation of 3-hydroxy-3-methylglutaryl coenzyme A reductase mRNA levels and synthesis in syrian hamster C100 cells by the oxidosqualene cyclase inhibitor [4'-(6-allyl-ethyl-amino-hexyloxy)-2'-fluoro-phenyl]-(4-bromophenyl)-me thanone (Ro 48-8071): comparison to simvastatin

In vivo inhibition of 2,3-oxidosqualene:lanosterol cyclase (OSC, E.C. 5.4.99.7)--the enzyme which catalyzes the cyclization of monooxidosqualene to lanosterol--does not result in elevated 3-hydroxy-3-methylglutaryl CoA reductase (HMGR) activity. This trait is attributed to increased levels of oxysterols, produced upon partial inhibition of OSC, that suppress HMGR and other sterol-responsive genes. The OSC inhibitor [4'-(6-allyl-ethyl-amino-hexyloxy)-2'-fluoro-phenyl]-(4-bromopheny l)-methanone (Ro 48-8071) was shown earlier to lower low-density lipoprotein (LDL) cholesterol in hamsters with no increase in hepatic HMGR, in contrast to simvastatin. To delineate the regulatory mechanism(s) by which Ro 48-8071 reduces cholesterol synthesis without raising HMGR levels, Syrian hamster C100 cells were incubated with either Ro 48-8071 or simvastatin, and their effects on cholesterol synthesis and LDL uptake, as well as on HMGR mRNA levels and rates of synthesis, were determined. Using RNase protection and radioimmunoprecipitation assays, we found that, in the absence of LDL in the culture medium, both HMGR mRNA levels and synthesis were reduced with concentrations of Ro 48-8071 inhibiting cholesterol synthesis by 50-75%, whereas LDL uptake was either reduced or unchanged. In contrast, simvastatin, at concentrations inhibiting cholesterol synthesis by the same 50-75%, increased both HMGR mRNA levels and synthesis, as well as LDL uptake. In the presence of LDL, HMGR mRNA levels and synthesis along with LDL uptake were little affected after incubation with Ro 48-8071. Still, simvastatin markedly increased both HMGR mRNA levels and synthesis in cells incubated in the presence of LDL, leaving LDL uptake unaffected. These data suggest that inhibition of OSC by Ro 48-8071 results in an indirect down-regulation of HMGR mRNA levels and synthesis.

A Drosophila homologue of oxysterol binding protein (OSBP)--implications for the role of OSBP

The identification of a Drosophila homologue (OSBP-Dm) of mammalian oxysterol binding protein (OSBP) is reported. OSBP-Dm was identified by its ability to overcome the cell cycle arrest induced by over-expression of Wee1p in fission yeast. OSBP-Dm has an overall sequence identity of 52% with mammalian OSBP, and shows a number of highly conserved regions of functional significance. Insects are unable to biosynthesize the steroid core, relying instead on dietary sterols to satisfy their requirements. It is therefore unlikely that OSBP-Dm is involved in feedback inhibition of the mevalonate pathway, as has previously been suggested for its mammalian homologues.

Mevalonate regulates polysome distribution and blocks translation-dependent suppression of 3-hydroxy-3-methylglutaryl coenzyme A reductase mRNA: relationship to translational control

We reported previously that 3-hydroxy-3-methylglutaryl coenzyme A reductase synthesis is regulated at the translational level by mevalonate. To determine at what stage mevalonate affects reductase synthesis, we examined the distribution of reductase mRNA in polysomes from cells treated with lovastatin alone; lovastatin and 25-hydroxycholesterol; or lovastatin, 25-hydroxycholesterol, and mevalonate. In lovastatin-treated cells, reductase mRNA was primarily associated with heavy polysome fractions. When 25-hydroxycholesterol was added to lovastatin-treated cells, reductase mRNA levels were reduced approximately fourfold in all polysome fractions, with no accompanying redistribution of reductase mRNA into lighter polysome fractions. However, addition of both 25-hydroxycholesterol and mevalonate to lovastatin-treated cells shifted reductase mRNA from heavier to lighter polysome fractions. No change in the distribution of control beta-actin or ribosomal protein S17 mRNA occurred with any of the treatments. These results suggest that mevalonate suppresses reductase synthesis at the level of initiation. When the translation inhibitor cycloheximide was added to all three regimens, reductase mRNA shifted into heavy polysome fractions. Treatment with either lovastatin alone or lovastatin plus 25-hydroxycholesterol resulted in a 50% greater loss of reductase mRNA from the heavy polysome fractions compared to the same fractions from noncycloheximide-treated cells. No loss of reductase mRNA occurred when cycloheximide was added to cells treated with both 25-hydroxycholesterol and mevalonate. beta-Actin mRNA levels and polysome distribution were not significantly changed by cycloheximide under any of these conditions. Translationally mediated suppression of reductase mRNA did not occur when protein synthesis was inhibited with puromycin. Our results indicate that regulation of reductase mRNA levels is translation-dependent and is linked to the rate of elongation.

3'-untranslated sequences mediate post-transcriptional regulation of 3-hydroxy-3-methylglutaryl-CoA reductase mRNA by 25-hydroxycholesterol

In an earlier study [Choi, Lundquist and Peffley (1993) Biochem. J. 296, 859-866], we determined that 25-hydroxycholesterol regulates 3-hydroxy-3-methylglutaryl-CoA (HMG-CoA) reductase mRNA through a post-transcriptional mechanism that requires protein synthesis. To investigate whether 3'-untranslated sequences play a role in 25-hydroxycholesterol-mediated post-transcriptional control, we ligated approx. 1400 bp of the 3'-untranslated region of HMG-CoA reductase cDNA to the coding region of human beta-globin DNA. beta-Globin-3'-untranslated reductase fusion constructs were then transiently expressed in Chinese hamster ovary fibroblasts under conditions known to regulate reductase mRNA. There were no differences in beta-globin RNA levels in transfected cells incubated with or without lovastatin, a competitive inhibitor of reductase. However, in the presence of lovastatin and an oxysterol, 25-hydroxycholesterol, beta-globin RNA levels were decreased approx. 2-fold. Inhibition of protein synthesis with cycloheximide blocked the effects of 25-hydroxycholesterol on beta-globin RNA. Moreover, replacing the 3'-untranslated sequences with 1367 bp of the simian virus 40 enhancer region eliminated the regulatory effect of 25-hydroxycholesterol. Because the fusion construct has no sterol regulatory elements necessary for transcription, our results indicate that the change in beta-globin RNA occurred at a post-transcriptional level. In addition, we have shown that the 3'-untranslated region of HMG-CoA reductase cDNA imparted oxysterol-mediated post-transcriptional regulation to beta-globin RNA, an effect that required protein synthesis.

Sterol-resistant transcription in CHO cells caused by gene rearrangement that truncates SREBP-2

Sterol-resistant CHO cells (SRD-1 cells) fail to repress sterol synthesis and LDL receptor gene transcription when incubated with 25-hydroxycholesterol. Here we trace the defect to a rearrangement in the gene encoding SREBP-2, a membrane-bound transcription factor that regulates cholesterol homeostasis. SREBP-2 is an 1139-amino acid protein that is bound to extranuclear membranes via a carboxy-terminal attachment domain. In sterol-depleted cells a protease liberates the amino-terminal fragment (approximately 480 amino acids). This fragment, which contains the transcriptional activation and bHLH-Zip domains, translocates to the nucleus. 25-Hydroxycholesterol abolishes protease activity and halts transcription. SRD-1 cells produce a soluble, truncated form of SREBP-2 (amino acids 1-460) that lacks the membrane attachment domain and activates transcription directly, bypassing the sterol-regulated proteolytic step. Although SRD-1 cells produce full-length SREBP-2 from the wild-type allele and a related transcription factor, SREBP-1, they fail to cleave both of these precursors, indicating that the truncated form of SREBP-2 down-regulates the protease through a form of end-product feedback inhibition. The current data provide genetic evidence for the previously proposed model in which cholesterol homeostasis is controlled by sterol-regulated proteolysis of a membrane-bound bHLH-Zip transcription factor.

Inhibition of protein synthesis in baby-hamster kidney cells blocks oxysterol-mediated suppression of 3-hydroxy-3-methylglutaryl-CoA reductase mRNA at a post-transcriptional level

The effects of the protein-synthesis inhibitor cycloheximide on 25-hydroxycholesterol-mediated suppression of 3-hydroxy-3-methylglutaryl-CoA (HMG-CoA) reductase mRNA levels were evaluated in the baby-hamster kidney cell line C100. Cells cultured in medium supplemented with delipidized fetal bovine serum and 25 microM lovastatin for 12-24 h had a 5-fold higher level of HMG-CoA reductase mRNA than cells grown in medium supplemented with non-delipidized fetal bovine serum (FBS). The higher level was due to increased transcription, as determined by run-on assays with isolated nuclei. Addition of 25-hydroxycholesterol to lovastatin-treated cells lowered HMG-CoA reductase mRNA levels within 4 h of treatment to those of cells grown in FBS-supplemented medium. This decrease was due in part to a decrease in gene transcription. Cycloheximide added in conjunction with 25-hydroxycholesterol to lovastatin-treated cells blocked the suppression of mRNA levels, but did not block oxysterol-mediated suppression of transcription. In addition, cycloheximide added to cells grown in FBS-supplemented medium rapidly increased mRNA levels by 10-fold relative to untreated cells, with no comparable increase in transcription. No comparable increase in either the mRNA level or rate of transcription for beta-actin was observed under such conditions. These results indicate that cycloheximide specifically stabilizes HMG-CoA reductase mRNA in the presence of oxysterols and suggests that continuous synthesis of a short lived protein regulator is required for oxysterol-mediated suppression of HMG-CoA reductase mRNA at a post-transcriptional level.

Importance of mevalonate-derived products in the control of HMG-CoA reductase activity and growth of human lung adenocarcinoma cell line A549

HMG-CoA reductase catalyzes the synthesis of mevalonate, a crucial intermediate in the biosynthesis of cholesterol and non-sterol isoprenoid compounds essential for cell growth. The HMG-CoA reductase activity of the A549 tumor cell line is higher than that of normal human fibroblasts. This deregulation in mevalonate needs was not due to an alteration in the activated state of the enzyme by short-term regulation. We show that the HMG-CoA reductase in A549 cell line was subject to a multivalent feedback control. A high fraction (40%) of the reductase activity was devoted to non-sterol products. In contrast, normal fibroblasts had only 15-20% of the reductase activity that generated non-sterol products. We also show that cholesterol and at least one of the non-sterol products are necessary for optimal cell growth of A549 cells. Our data strongly suggest that A549 cells produce more non-sterol substances which may be related to increased requirements of mevalonate for upregulated cell growth.

Loss of transcriptional activation of three sterol-regulated genes in mutant hamster cells

Cholesterol biosynthesis and uptake are controlled by a classic end product-feedback mechanism whereby elevated cellular sterol levels suppress transcription of the genes encoding 3-hydroxy-3-methylglutaryl coenzyme A (HMG-CoA) synthase, HMG-CoA reductase, and the low-density lipoprotein receptor. The 5'-flanking region of each gene contains a common cis-acting element, designated the sterol regulatory element (SRE), that is required for transcriptional regulation. In this report, we describe mutant Chinese hamster ovary (CHO) cell lines that lack SRE-dependent transcription. Mutant cell lines were isolated on the basis of their ability to survive treatment with amphotericin B, a polyene antibiotic that kills cells by interacting with cholesterol in the plasma membrane. Four mutant lines (SRD-6A, -B, -C, and -D) were found to be cholesterol auxotrophs and demonstrated constitutively low levels of mRNA for all three sterol-regulated genes even under conditions of sterol deprivation. The mutant cell lines were found to be genetically recessive, and all four lines belonged to the same complementation group. When transfected with a plasmid containing a sterol-regulated promoter fused to a bacterial reporter gene, SRD-6B cells demonstrated constitutively low levels of transcription, in contrast to wild-type CHO cells, which increased transcription under conditions of sterol deprivation. Mutation of the SREs in this plasmid prior to transfection reduced the level of expression in wild-type CHO cells deprived of sterols to the level of expression found in SRD-6B cells. The defect in SRD-6 cells is limited to transcriptional regulation, since posttranscriptional mechanisms of sterol-mediated regulation were intact: the cells retained the ability to posttranscriptionally suppress HMG-CoA reductase activity and to stimulate acyl-CoA:cholesterol acyltransferase activity. These results suggest that SRD-6 cells lack a factor required for SRE-dependent transcriptional activation. We contrast these cells with a previously isolated oxysterol-resistant cell line (SRD-2) that lacks a factor required for SRE-dependent transcriptional suppression and propose a model for the role of these genetically defined factors in sterol-mediated transcriptional regulation.

Loss of transcriptional activation of three sterol-regulated genes in mutant hamster cells

Cholesterol biosynthesis and uptake are controlled by a classic end product-feedback mechanism whereby elevated cellular sterol levels suppress transcription of the genes encoding 3-hydroxy-3-methylglutaryl coenzyme A (HMG-CoA) synthase, HMG-CoA reductase, and the low-density lipoprotein receptor. The 5'-flanking region of each gene contains a common cis-acting element, designated the sterol regulatory element (SRE), that is required for transcriptional regulation. In this report, we describe mutant Chinese hamster ovary (CHO) cell lines that lack SRE-dependent transcription. Mutant cell lines were isolated on the basis of their ability to survive treatment with amphotericin B, a polyene antibiotic that kills cells by interacting with cholesterol in the plasma membrane. Four mutant lines (SRD-6A, -B, -C, and -D) were found to be cholesterol auxotrophs and demonstrated constitutively low levels of mRNA for all three sterol-regulated genes even under conditions of sterol deprivation. The mutant cell lines were found to be genetically recessive, and all four lines belonged to the same complementation group. When transfected with a plasmid containing a sterol-regulated promoter fused to a bacterial reporter gene, SRD-6B cells demonstrated constitutively low levels of transcription, in contrast to wild-type CHO cells, which increased transcription under conditions of sterol deprivation. Mutation of the SREs in this plasmid prior to transfection reduced the level of expression in wild-type CHO cells deprived of sterols to the level of expression found in SRD-6B cells. The defect in SRD-6 cells is limited to transcriptional regulation, since posttranscriptional mechanisms of sterol-mediated regulation were intact: the cells retained the ability to posttranscriptionally suppress HMG-CoA reductase activity and to stimulate acyl-CoA:cholesterol acyltransferase activity. These results suggest that SRD-6 cells lack a factor required for SRE-dependent transcriptional activation. We contrast these cells with a previously isolated oxysterol-resistant cell line (SRD-2) that lacks a factor required for SRE-dependent transcriptional suppression and propose a model for the role of these genetically defined factors in sterol-mediated transcriptional regulation.

Translocation of oxysterol binding protein to Golgi apparatus triggered by ligand binding

A cDNA encoding a cytoplasmic oxysterol binding protein was expressed at high levels by transfection in animal cells. This protein binds oxysterols such as 25-hydroxycholesterol that regulate sterol metabolism by transcriptional and posttranscriptional effects. In the transfected cells, some of the oxysterol binding protein (OSBP) was distributed diffusely in the cytoplasm, and some was bound to small vesicles near the nucleus, as revealed by indirect immunofluorescence. Upon addition of 25-hydroxycholesterol, most of the OSBP became concentrated in large perinuclear structures that stained with lentil lectin, a protein that stains the Golgi apparatus. The structures that contained OSBP were disrupted by brefeldin A, confirming their identification as Golgi. A mutant OSBP lacking the COOH-terminal oxysterol binding domain localized to the Golgi spontaneously, suggesting that this domain normally occludes the domain that binds to the Golgi and that sterols relieve this occlusion. The previously noted potential leucine zipper sequence in OSBP was not required for Golgi localization, nor was it essential for homodimer formation. We conclude that OSBP is triggered to bind extrinsically to Golgi membranes when it binds oxysterols and speculate that this translocation may play a role in the transport, metabolism, or regulatory actions of oxysterols.

Regulation of 3-hydroxy-3-methylglutaryl coenzyme A reductase synthesis in Syrian hamster C100 cells by mevinolin, 25-hydroxycholesterol, and mevalonate: the role of posttranscriptional control

C100 is a baby hamster kidney cell line that expresses high levels of HMG-CoA reductase relative to its parental cell line SV28. In this study the effects of the oxysterol 25-hydroxycholesterol and mevalonate on the mRNA level and rate of synthesis for HMG-CoA reductase were evaluated in C100 cells treated with mevinolin, a competitive inhibitor of HMG-CoA reductase. The addition of 25-hydroxycholesterol to the cell culture medium resulted in a fourfold decrease in both the rate of synthesis and mRNA level for HMG-CoA reductase. Mevalonate at a concentration of 0.4 mM, when added to mevinolin-treated C100 cells, produced no apparent reduction in HMG-CoA reductase mRNA levels and only a small (25%) decline in HMG-CoA synthesis. Mevalonate (0.4 mM) added to 25-hydroxycholesterol-treated cells resulted in no further reduction in the HMG-CoA reductase mRNA level when compared to cells treated with 25-hydroxycholesterol alone, but produced an additional 30-fold decrease in the rate of HMG-CoA reductase synthesis. Degradation of HMG-CoA reductase was rapid in the presence (t1/2 = 1.34 h) or absence (t1/2 = 1.17 h) of mevinolin and was not changed significantly by adding either 25-hydroxycholesterol, alone (t1/2 = 1.30 h) or both 25-hydroxycholesterol and mevalonate (t1/2 = 1.30 h) to mevinolin-treated cells. This study demonstrates that mevalonate and 25-hydroxycholesterol act synergistically in the presence of mevinolin to achieve a greater degree of suppression in the rate of HMG-CoA reductase synthesis than can be accounted for by their individual effects on HMG-CoA reductase mRNA. In addition, the data suggest that mevalonate affects the synthesis of HMG-CoA reductase at a yet unidentified posttranscriptional control site.