Prenylation of mammalian Ras protein in Xenopus oocytes

Ras protein requires an intermediate of the cholesterol biosynthetic pathway for posttranslational modification and membrane anchorage. This step is necessary for biological activity. Maturation of Xenopus laevis oocytes induced by an oncogenic human Ras protein can be inhibited by lovastatin or compactin, inhibitors of the synthesis of mevalonate, an intermediate of cholesterol biosynthesis. This inhibition can be overcome by mevalonic acid or farnesyl diphosphate, a cholesterol biosynthetic intermediate downstream of mevalonate, but not by squalene, an intermediate after farnesyl pyrophosphate in the pathway. This study supports the idea that in Xenopus oocytes, the Ras protein is modified by a farnesyl moiety or its derivative. Furthermore, an octapeptide with the sequence similar to the C-terminus of the c-H-ras protein inhibits the biological activity of Ras proteins in vivo, suggesting that it competes for the enzyme or enzymes responsible for transferring the isoprenoid moiety (prenylation) in the oocytes. This inhibition of Ras prenylation by the peptide was also observed in vitro, using both Saccharomyces cerevisiae and Xenopus oocyte extracts. These observations show that Xenopus oocytes provide a convenient in vivo system for studies of inhibitors of the posttranslational modification of the Ras protein, especially for inhibitors such as peptides that do not penetrate cell membranes.

Cell-type control of membrane biogenesis induced by HMG-CoA reductase

Quantitative increases in HMG-CoA reductase, the rate-limiting enzyme in sterol biosynthesis, induce membrane biogenesis in both yeast and mammalian cells. The subcellular organization of the resulting membrane differs in the two cell types: mammalian cells generate crystalloid endoplasmic reticulum whereas yeast cells assemble karmellae. We examined the consequences of heterologous expression of HMG-CoA reductase to distinguish features of this response that were cell-type specific from those that were isozyme-specific. This analysis demonstrated that membrane proliferation was induced in both mammalian and yeast cells by HMG-CoA reductase from either organism. However, the morphology of the induced membranes was determined by the cell type rather than the particular isozyme. Thus, both yeast and mammalian HMG-CoA reductase contained functional signals for membrane proliferation that were operational in either cell type, but the qualitative response to those signals was cell-type specific.

Enzymatic coupling of cholesterol intermediates to a mating pheromone precursor and to the ras protein

The post-translational processing of the yeast a-mating pheromone precursor, Ras proteins, nuclear lamins, and some subunits of trimeric G proteins requires a set of complex modifications at their carboxyl termini. This processing includes three steps: prenylation of a cysteine residue, proteolytic processing, and carboxymethylation. In the yeast Saccharomyces cerevisiae, the product of the DPR1-RAM1 gene participates in this type of processing. Through the use of an in vitro assay with peptide substrates modeled after a presumptive a-mating pheromone precursor, it was discovered that mutations in DPR1-RAM1 cause a defect in the prenylation reaction. It was further shown that DPR1-RAM1 encodes an essential and limiting component of a protein prenyltransferase. These studies also implied a fixed order of the three processing steps shared by prenylated proteins: prenylation, proteolysis, then carboxymethylation. Because the yeast protein prenyltransferase could also prenylate human H-ras p21 precursor, the human DPR1-RAM1 analogue may be a useful target for anticancer chemotherapy.

Common modifications of trimeric G proteins and ras protein: involvement of polyisoprenylation

The heterotrimeric guanine nucleotide-binding regulatory proteins act at the inner surface of the plasma membrane to relay information from cell surface receptors to effectors inside the cell. These G proteins are not integral membrane proteins, yet are membrane associated. The processing and function of the gamma subunit of the yeast G protein involved in mating-pheromone signal transduction was found to be affected by the same mutations that block ras processing. The nature of these mutations implied that the gamma subunit was polyisoprenylated and that this modification was necessary for membrane association and biological activity. A microbial screen was developed for pharmacological agents that inhibit polyisoprenylation and that have potential application in cancer therapy.

A role for isoprenoid lipids in the localization and function of an oncoprotein

Intermediates of the cholesterol biosynthetic pathway are covalently attached to a number of eukaryotic proteins, including the Ras oncoprotein. Ras protein is post-translationally processed at its carboxyl terminus in three steps, resulting in a COOH-terminal cysteine residue to which a polyisoprenoid moiety, probably farnesyl, is attached in a thioether linkage. Polyisoprenylation of Ras protein is required for its membrane association and for the oncogenicity of mutant forms of the protein. Inhibition of polyisoprenylation may offer a route by which Ras-mediated tumors can be pharmacologically suppressed. Other proteins that are polyisoprenylated include nuclear lamin B, fungal mating factors, and subunits of trimeric guanine nucleotide-binding proteins. A consensus sequence for polyisoprenylation (Cys-aliphatic-aliphatic-X) has been identified at the COOH-terminus of modified proteins. Recent evidence indicates that proteins can be modified by several different polyisoprenoids.

Positive and negative transcriptional control by heme of genes encoding 3-hydroxy-3-methylglutaryl coenzyme A reductase in Saccharomyces cerevisiae

Responses of the yeast genes encoding 3-hydroxy-3-methylglutaryl coenzyme A reductase, HMG1 and HMG2, to in vivo changes in heme concentrations were investigated. Expression of the genes was determined by direct measurement of the mRNA transcribed from each gene, by direct assay of the enzyme activity encoded by each gene, and by measurement of the expression of lacZ fusions to the control regions of each gene. These studies indicated that expression of HMG1 was stimulated by heme, whereas expression of HMG2 was repressed by heme. The effect of heme on HMG1 expression was mediated by the HAP1 transcriptional regulator and was independent of HAP2. Thus, the genes encoding the 3-hydroxy-3-methylglutaryl coenzyme A reductase isozymes join a growing list of gene pairs that are regulated by heme in opposite ways.

Epigenetic inheritance of transcriptional states in S. cerevisiae

SIR1, one of several genes required for repression of yeast silent mating type loci, has a unique role in repression of the HML alpha locus. Single-cell assays revealed that cells with mutant alleles of SIR1, including presumptive null alleles, existed as populations of genetically identical cells whose members were in one of two different regulatory states. A minority of cells had a repressed HML alpha locus whereas the majority had a derepressed HML alpha locus. The two states were mitotically stable, although rare changes in state were observed during mitotic growth, possibly reflecting heritable changes to the HML alpha locus at or before replication. Analysis of changes in state suggests that SIR1 protein has a role in the establishment but not the maintenance of repression of silent mating type genes, whereas SIR2, SIR3, and SIR4 are required for maintenance.

Genetic and pharmacological suppression of oncogenic mutations in ras genes of yeast and humans

The activity of an oncoprotein and the secretion of a pheromone can be affected by an unusual protein modification. Specifically, posttranslational modification of yeast a-factor and Ras protein requires an intermediate of the cholesterol biosynthetic pathway. This modification is apparently essential for biological activity. Studies of yeast mutants blocked in sterol biosynthesis demonstrated that the membrane association and biological activation of the yeast Ras2 protein require mevalonate, a precursor of sterols and other isoprenes such as farnesyl pyrophosphate. Furthermore, drugs that inhibit mevalonate biosynthesis blocked the in vivo action of oncogenic derivatives of human Ras protein in the Xenopus oocyte assay. The same drugs and mutations also prevented the posttranslational processing and secretion of yeast a-factor, a peptide that is farnesylated. Thus, the mevalonate requirement for Ras activation may indicate that attachment of a mevalonate-derived (isoprenoid) moiety to Ras proteins is necessary for membrane association and biological function. These observations establish a connection between the cholesterol biosynthetic pathway and transformation by the ras oncogene and offer a novel pharmacological approach to investigating, and possibly controlling, ras-mediated malignant transformations.

Genetic and molecular characterization of suppressors of SIR4 mutations in Saccharomyces cerevisiae

In order to learn more about other proteins that may be involved in repression of HML and HMR in Saccharomyces cerevisiae, extragenic suppressor mutations were identified that could restore repression in cells defective in SIR4, a gene required for function of the silencer elements flanking HML and HMR. These suppressor mutations, which define at least three new genes, SAN1, SAN2 and SAN3, arose at the frequency expected for loss-of-function mutations following mutagenesis. All san mutations were recessive. Suppression by san1 was allele-nonspecific, since san1 could suppress two very different alleles of SIR4, and was locus-specific since san1 was unable to suppress a SIR3 mutation or a variety of mutations conferring auxotrophies. The SAN1 gene was cloned, sequenced, and used to construct a null allele. The null allele had the same phenotype as the EMS-induced mutations and exhibited no pleiotropies of its own. Thus, the SAN1 gene was not essential. SAN1-mediated suppression was neither due to compensatory mutations in interacting proteins, nor to translational missense suppression. SAN1 may act posttranslationally to control the stability or activity of the SIR4 protein.

Transmission electron microscopy and immunocytochemical studies of yeast: analysis of HMG-CoA reductase overproduction by electron microscopy

The results and anecdotes presented here are intended only as a general guide to other would-be immunocytochemists, because other proteins will undoubtedly respond at least somewhat differently than does HMG-CoA reductase. Nevertheless, based on these experiences, we offer the following suggestions: 1. Antiserum of high specificity should be raised and affinity-purified. Using this antiserum, immunofluorescence microscopy should be attempted before resorting to electron microscopic localization. In the absence of immunolocalization at the light-microscope level, it may be a waste of time to pursue the problem to higher levels of resolution. 2. Cells should be prefixed in 1% formaldehyde-1% glutaraldehyde. Direct fixation of the growing culture and use of phosphate buffer are recommended. The prefixed sample can then be divided into two or three aliquots. One aliquot should receive no postfixation (for optimal immunoreactivity), while the others can be postfixed in osmium-potassium ferricyanide (for possible immunolocalization) or permanganate (for ultrastructural analysis). Because of its ease of use, Spurr's resin should be tried initially. If immunocytochemistry is successful, no further preparations are necessary. If unsuccessful, LR White resin is recommended, but the sample must be treated to remove the cell wall. Electron microscopy and immunocytochemistry offer views into the molecular arrangement of individual cells, a view not easily obtained by other means. It is satisfying and often enlightening to be able to see the extremes as well as the average. In studies of the organization of karmellae, for example, ultrastructural analysis easily revealed the asymmetric segregation pattern, while immunoblots and cell fractionation could not even demonstrate the existence of this membrane organization. The richness of the information available to those who can avert reductionist tendencies, even for a short time, is remarkable.

Structural and functional conservation between yeast and human 3-hydroxy-3-methylglutaryl coenzyme A reductases, the rate-limiting enzyme of sterol biosynthesis

The pathway of sterol biosynthesis is highly conserved in all eucaryotic cells. We demonstrated structural and functional conservation of the rate-limiting enzyme of the mammalian pathway, 3-hydroxy-3-methyl-glutaryl coenzyme A reductase (HMG-CoA reductase), between the yeast Saccharomyces cerevisiae and humans. The amino acid sequence of the two yeast HMG-CoA reductase isozymes was deduced from DNA sequence analysis of the HMG1 and HMG2 genes. Extensive sequence similarity existed between the region of the mammalian enzyme encoding the active site and the corresponding region of the two yeast isozymes. Moreover, each of the yeast isozymes, like the mammalian enzyme, contained seven potential membrane-spanning domains in the NH2-terminal region of the protein. Expression of cDNA clones encoding either hamster or human HMG-CoA reductase rescued the viability of hmg1 hmg2 yeast cells lacking this enzyme. Thus, mammalian HMG-CoA reductase can provide sufficient catalytic function to replace both yeast isozymes in vivo. The availability of yeast cells whose growth depends on human HMG-CoA reductase may provide a microbial screen to identify new drugs that can modulate cholesterol biosynthesis.

Increased amounts of HMG-CoA reductase induce "karmellae": a proliferation of stacked membrane pairs surrounding the yeast nucleus

Overproduction of the enzyme 3-hydroxy-3-methylglutaryl coenzyme A (HMG-CoA) reductase in yeast resulted in striking morphological effects on the structure of intracellular membranes. Specifically, stacks of paired membranes closely associated with the nuclear envelope were observed in strains that over-produced the HMG1 isozyme, one of two isozymes for HMG-CoA reductase in yeast. These nuclear-associated, paired membranes have been named "karmellae." In strains that overproduced the HMG1 isozyme, HMG-CoA reductase was present in the karmellar layers. At mitosis, karmellae were asymmetrically segregated: the mother cells inherited all of the karmellae and the daughter cells inherited none. A membranous structure of different morphology was occasionally found in cells that overproduced the HMG2 isozyme. These observations further establish the existence of cellular mechanisms that monitor the levels of membrane proteins and compensate for changes in these levels by inducing synthesis of particular types of membrane.

Roles of two DNA-binding factors in replication, segregation and transcriptional repression mediated by a yeast silencer

The HMR E silencer is required for SIR-dependent transcriptional repression of the silent mating-type locus, HMR. The silencer also behaves as an origin of replication (ARS element) and allows plasmids to replicate autonomously in yeast. The replication and segregation properties of these plasmids are also dependent on the four SIR genes. We have previously characterized two DNA-binding factors in yeast extracts that recognize specific sequences at the HMR E silencer. These proteins, called ABFI (ARS-Binding Factor) and GRFI (General Regulatory Factor), are not encoded by any of the SIR genes. To investigate the biological roles of these factors, single-base-pair mutations were constructed in both binding sites at the HMR E silencer that were no longer recognized by the corresponding proteins in vitro. Our results indicate that the GRFI-binding site is required for the efficient segregation of plasmids replicated by the HMR E silencer. SIR-dependent transcriptional repression requires either an intact ABFI-binding site or GRFI-binding site, although the GRFI-binding site appears to be more important. A double-mutant silencer that binds neither ABFI nor GRFI does not mediate transcriptional repression of HMR. The replacement of HMR E with a chromosomal origin of replication (ARS1) allows partial SIR-dependent transcriptional repression of HMR, indicating a role for replication in silencer function. Together, these results suggest that the SIR proteins influence the properties of the HMR E silencer through interactions with other DNA-binding proteins.

Evidence for Drosophila P element transposase activity in mammalian cells and yeast

Drosophila P element transposase expression is limited to the germline by tissue-specific splicing of one of its three introns. Removal of this intron by mutagenesis in vitro has allowed both P element excision and transposition to be detected in Drosophila somatic tissues. In order to determine if P element transposase can function in other organisms, we have expressed modified P elements either lacking one intron or lacking all three introns in mammalian cells and yeast, respectively. Using an assay for P element excision, we have detected apparent excision events in cultured monkey cells. Furthermore, expression of the complete P element cDNA is lethal to Saccharomyces cerevisiae cells carrying a mutation in the RAD52 gene, indicating that double-stranded DNA breaks are generated, presumably by transposase action.

Two DNA-binding factors recognize specific sequences at silencers, upstream activating sequences, autonomously replicating sequences, and telomeres in Saccharomyces cerevisiae

Two DNA-binding factors from Saccharomyces cerevisiae have been characterized, GRFI (general regulatory factor I) and ABFI (ARS-binding factor I), that recognize specific sequences within diverse genetic elements. GRFI bound to sequences at the negative regulatory elements (silencers) of the silent mating type loci HML E and HMR E and to the upstream activating sequence (UAS) required for transcription of the MAT alpha genes. A putative conserved UAS located at genes involved in translation (RPG box) was also recognized by GRFI. In addition, GRFI bound with high affinity to sequences with the (C1-3A)-repeat region at yeast telomeres. Binding sites for GRFI with the highest affinity appeared to be of the form 5'-(A/G)(A/C)ACCCANNCA(T/C)(T/C)-3', where N is any nucleotide. ABFI-binding sites were located next to autonomously replicating sequences (ARSs) at controlling elements of the silent mating type loci HMR E, HMR I, and HML I and were associated with ARS1, ARS2, and the 2 micron plasmid ARS. Two tandem ABFI binding sites were found between the HIS3 and DED1 genes, several kilobase pairs from any ARS, indicating that ABFI-binding sites are not restricted to ARSs. The sequences recognized by ABFI showed partial dyad-symmetry and appeared to be variations of the consensus 5'-TATCATTNNNNACGA-3'. GRFI and ABFI were both abundant DNA-binding factors and did not appear to be encoded by the SIR genes, whose products are required for repression of the silent mating type loci. Together, these results indicate that both GRFI and ABFI play multiple roles within the cell.

Two DNA-binding factors recognize specific sequences at silencers, upstream activating sequences, autonomously replicating sequences, and telomeres in Saccharomyces cerevisiae

Two DNA-binding factors from Saccharomyces cerevisiae have been characterized, GRFI (general regulatory factor I) and ABFI (ARS-binding factor I), that recognize specific sequences within diverse genetic elements. GRFI bound to sequences at the negative regulatory elements (silencers) of the silent mating type loci HML E and HMR E and to the upstream activating sequence (UAS) required for transcription of the MAT alpha genes. A putative conserved UAS located at genes involved in translation (RPG box) was also recognized by GRFI. In addition, GRFI bound with high affinity to sequences with the (C1-3A)-repeat region at yeast telomeres. Binding sites for GRFI with the highest affinity appeared to be of the form 5'-(A/G)(A/C)ACCCANNCA(T/C)(T/C)-3', where N is any nucleotide. ABFI-binding sites were located next to autonomously replicating sequences (ARSs) at controlling elements of the silent mating type loci HMR E, HMR I, and HML I and were associated with ARS1, ARS2, and the 2 micron plasmid ARS. Two tandem ABFI binding sites were found between the HIS3 and DED1 genes, several kilobase pairs from any ARS, indicating that ABFI-binding sites are not restricted to ARSs. The sequences recognized by ABFI showed partial dyad-symmetry and appeared to be variations of the consensus 5'-TATCATTNNNNACGA-3'. GRFI and ABFI were both abundant DNA-binding factors and did not appear to be encoded by the SIR genes, whose products are required for repression of the silent mating type loci. Together, these results indicate that both GRFI and ABFI play multiple roles within the cell.

Replication and segregation of plasmids containing cis-acting regulatory sites of silent mating-type genes in Saccharomyces cerevisiae are controlled by the SIR genes

In Saccharomyces cerevisiae, two cis-acting regulatory sites called E and I flank the silent mating-type gene, HMRa, and mediate SIR-dependent transcriptional repression of the a1-a2 promoters. It has been shown previously that the E and I sites have plasmid replicator (ARS) activity. We show in this report that the ARS activity of the E and I sites is governed by the SIR genotype of the cell. In wild-type cells, a plasmid carrying the E site from HMRa (HMR E) in the vector YIp5 exhibited very high mitotic stability at a copy number of approximately 25 per cell. However, in sir2, sir3, or sir4 mutants, plasmids with HMR E had the low mitotic stability characteristic of plasmids containing ARS1, a SIR-independent replicator. Elevated mitotic stability of plasmids that carry HMR E is due to a segregation mechanism provided by SIR and HMR E. In sir2 and sir4 mutants, the plasmid copy number was significantly lowered, suggesting that these gene products also participate in the replication of plasmids carrying HMR E. The phenotype of point mutations introduced at an 11-base-pair ARS consensus sequence present at HMR E indicated that this sequence is functional but not absolutely required for autonomous replication of the plasmid and that it is not required for SIR-dependent mitotic stabilization. A plasmid carrying both a centromere and HMR E exhibited reduced mitotic stability in wild-type cells. This destabilization appeared to be due to antagonism between the segregation functions provided by the centromere and by HMR E.

Identifying mutations in duplicated functions in Saccharomyces cerevisiae: recessive mutations in HMG-CoA reductase genes

The two yeast genes for 3-hydroxy-3-methylglutaryl-coenzyme A (HMG-CoA) reductase, HMG1 and HMG2, each encode a functional isozyme. Although cells bearing null mutations in both genes are inviable, cells bearing a null mutation in either gene are viable. This paper describes a method of screening for recessive mutations in the HMG1 gene, the gene encoding the majority of HMG-CoA reductase activity in the cell. This method should be applicable to the isolation of mutations in other recovered in HMG1. These mutations exhibited intragenic complementation: one allele is in one complementation group and three alleles are in a second complementation group. Assays of HMG-CoA reductase activity indicated that the point mutations destroy most if not all of the activity encoded by HMG1. Intragenic complementation occurred with partial restoration of enzymatic activity. HMG1 was mapped to the left arm of chromosome XIII near SUP79, and HMG2 was mapped to the right arm of chromosome XII near SST2. A slight deleterious effect of a null mutation in either HMG-CoA reductase gene was detected by a co-cultivation experiment involving the wild-type strain and the two single mutants.

Four genes responsible for a position effect on expression from HML and HMR in Saccharomyces cerevisiae

Mating type interconversion in Saccharomyces cerevisiae occurs by transposition of copies of the a or alpha mating type cassettes from inactive loci, HML and HMR, to an active locus, MAT. The lack of expression of the a and alpha genes at the silent loci results from repression by trans-acting regulators encoded by SIR (Silent Information Regulator) genes. In this paper we present evidence for the existence of four SIR genes. Inactivation of any of these genes leads to expression of cassettes at both HML and HMR. Unusual complementation properties are observed for a number of sir mutations. Specifically, some recessive mutations in different genes fail to complement. The correspondence between SIR1, SIR2, SIR3, SIR4 and other genes with similar roles (MAR, CMT, STE8 and STE9) is presented.

Saccharomyces cerevisiae contains two functional genes encoding 3-hydroxy-3-methylglutaryl-coenzyme A reductase

We have isolated two genes from yeast encoding 3-hydroxy-3-methylglutaryl-coenzyme A reductase [hydroxymethylglutaryl-coenzyme A reductase (NADPH); HMG-CoA reductase; EC 1.1.1.34], the rate-limiting enzyme of sterol biosynthesis. These genes, HMG1 and HMG2, were identified by hybridization to a cDNA clone encoding hamster HMG-CoA reductase. DNA sequence analysis reveals homology between the amino acid sequence of the proteins encoded by the two yeast genes and the carboxyl-terminal half of the hamster protein. Cells containing mutant alleles of both HMG1 and HMG2 are unable to undergo spore germination and vegetative growth. However, cells containing a mutant allele of either HMG1 or HMG2 are viable but are more sensitive to compactin, a competitive inhibitor of HMG-CoA reductase, than are wild-type cells. Assays of HMG-CoA reductase activity in extracts from hmg1- and hmg2- mutants indicate that HMG1 contributes at least 83% of the activity found in wild-type cells.

Precocious meiotic centromere separation of a novel yeast chromosome

In Saccharomyces cerevisiae, a reciprocal translocation between chromosome II and a linear plasmid carrying a centromere (CEN6) has split chromosome II into two fragments: one, approximately 530 kilobase pairs (kbp) in size, has the left arm and part of the right arm of chromosome II; the other, a telocentric fragment approximately 350 kbp in size, has CEN6 and the rest of the right arm of chromosome II. A cross of this yeast strain with a strain containing a complete chromosome II exhibits a high frequency of precocious centromere separation (separation of sister chromatids during meiosis I) of the telocentric fragment. Precocious centromere separation is not due to the position of the centromere per se, since diploids that are homozygous for both fragments of chromosome II segregate the telocentric fragment with normal meiotic behavior. The precocious centromere separation described here differs from previously described examples in that pairing and synapsis of this telocentric chromosome seem to be normal. One model of how centromeres function in meiosis is that replication of the centromere is delayed until the second meiotic division. Data presented in this paper indicate that replication of the centromere is complete before the first meiotic division. The precocious separation of the centromere described here may be due to improper synapsis of sequences flanking the centromere.

A position effect on the expression of a tRNA gene mediated by the SIR genes in Saccharomyces cerevisiae

The SIR genes of Saccharomyces cerevisiae are responsible for the position-dependent regulation of the a and alpha mating-type genes. Previous work by others has shown that the products of the SIR genes prevent the accumulation of stable transcripts of the a and alpha genes at HML and HMR. Results of this study establish that this regulation is a region-specific effect rather than a gene-specific effect since expression of a tRNA gene placed at HMR is repressed by the products of the SIR genes.