Yeast cells as tools for target-oriented screening.

Information about biomolecular interaction networks is crucial for understanding cellular functions and the development of disease processes. Many diseases are known to be based on aberrations of DNA sequences encoding proteins with key functions in the cellular metabolism. Alterations in the respective proteins often lead to disturbances in biomolecular interactions caused by unbalanced stoichiometries, and thus result in alterations of molecule fluxes, cell architecture and signalling pathways. Drug discovery programmes have been designed to find promising chemical lead structures with the help of target-oriented bioassay systems. These are, in most cases, based upon the interaction of small molecules to specific macromolecular targets in vivo or in vitro, as exemplified by enzyme assays or small-ligand-based receptor systems. In addition, interactions between large biomolecules, such as proteins or nucleic acids, offer a huge arsenal of potential drug targets that can be addressed by small chemical compounds. This latter approach is gaining considerable attention because many potential target structures are becoming available through genomic research. Funnelling these new targets into high-throughput screening programs represents a major challenge for today's pharmaceutical research. An important outcome of the ongoing genome projects is the fact that the basic cellular structures, pathways and signalling principles show a high degree of conservation. Model organisms that are easily approachable by genetic, biochemical and physiological means can thus play an important role in the design of target-oriented screening systems. They offer the possibility to express individual proteins, nucleic acids or even more complex aggregates of biomolecules such as protein-interaction networks or transcription-initiation complexes, which can be addressed by small effector molecules in vivo. Combining these targets with biological signalling systems is an attractive way of creating robust cellular assay systems.

Functional analysis of 150 deletion mutants in Saccharomyces cerevisiae by a systematic approach.

In a systematic approach to the study of Saccharomyces cerevisiae genes of unknown function, 150 deletion mutants were constructed (1 double, 149 single mutants) and phenotypically analysed. Twenty percent of all genes examined were essential. The viable deletion mutants were subjected to 20 different test systems, ranging from high throughput to highly specific test systems. Phenotypes were obtained for two-thirds of the mutants tested. During the course of this investigation, mutants for 26 of the genes were described by others. For 18 of these the reported data were in accordance with our results. Surprisingly, for seven genes, additional, unexpected phenotypes were found in our tests. This suggests that the type of analysis presented here provides a more complete description of gene function.

Single-read sequence tags of a limited number of genomic DNA fragments provide an inexpensive tool for comparative genome analysis.

Single-read sequences from both ends of 415 3-kb average size genomic DNA fragments of Candida albicans were compared with the complete sequence data of Saccharomyces cerevisiae. Comparison at the protein level, translated DNA against protein sequences, revealed 138 sequence tags with clear similarity to S. cerevisiae proteins or open reading frames. One case of synteny was found for the open reading frames of RAD16 and LYS2, which are adjacent to each other in S. cerevisiae and C. albicans.

Drug-induced phenotypes provide a tool for the functional analysis of yeast genes.

The post-genome sequencing era of Saccharomyces cerevisiae is defined by the analysis of newly discovered open reading frames of unknown function. In this report, we describe a genetic method for the rapid identification and characterisation of genes involved in a given phenotype. This approach is based on the ability of overexpressed genomic DNA fragments to cure an induced phenotype in yeast. To validate this concept, yeast cells carrying a yeast DNA library present on multicopy plasmid vectors were screened for resistance to the antifungal drug ketoconazole. Among 1.2 million colonies 13 clones tested positive, including those expressing the lanosterol C-14 demethylase, known to be a cellular target for azole drugs, and the cytochrome-c oxidase of mitochondria, regulating the respiratory chain electron transport. Several other resistant clones were identified, which code for yeast proteins of so far unknown function. These genes may represent potential candidates for antifungal drug effects. Together with the availability of the entire yeast genome sequence, the described genetic screening method is a powerful tool for the effective functional analysis of yeast genes.

Homologous recombination partly restores the secretion defect of underglycosylated acid phosphatase in yeast.

The majority of secreted acid phosphatase in Saccharomyces cerevisiae is encoded by the PH05 gene. The secretion level of this acid phosphatase is directly determined by its level of glycosylation. Consequently, PHO5-11-encoded acid phosphatase which lacks 11 of 12 glycosylation sites is only poorly secreted. We have isolated and characterized both UV- and EMS-induced variants, which are partly able to restore the secretion of acid phosphatase. Our data indicate that the improved secretion is caused by mitotic intrachromosomal recombination between the PHO5-11 allele and the homologous tandemly repeated PHO3 sequences, resulting in the restoration of glycosylation sites in PHO5-11. Two different recombination mechanisms, unequal sister-chromatid exchange and sister-chromatid gene conversion, are responsible for these alterations of the PHO5-11 locus. Thus, recombination between mutant and wild-type sequences are able to restore the ability of mutant yeast cells to secrete acid phosphatase.

Biochemical properties and excretion behavior of repressible acid phosphatases with altered subunit composition.

Yeast repressible acid phosphatase (rAP) is the oligomeric extracellular enzyme encoded by the three structural genes PH05 (p60), PHO10 (p58) and PHO11 (p56). We examined the ability of acid phosphatases formed by various subunit combinations to be excreted into the medium. Plasmids with repressible acid phosphatase structural genes under control of the yeast glyceraldehyde-phosphate dehydrogenase (GAP) promoter were constructed to obtain constitutive expression of acid phosphatase, and yeast strains with disruptions in PHO5, PHO10 and PHO11, respectively, were used to generate mutants expressing single genes or specific gene combinations. EndoF treatment of acid phosphatases, produced by these strains, followed by SDS-electrophoresis in combination with densitometry techniques revealed that the ratio p60/(p56 + p58) among structural polypeptides in extracellular enzyme is constant and equals to 6.0. A study of acid phosphatases formed by single type subunits was undertaken. Expression products of PHO5, PHO10 and PHO11 genes were isolated from the culture medium. The specific activities of the enzymes were found to be 33, 2 and 2 mM x mg-1 x min-1, respectively. The values of Mr estimated by HPLC chromatography for the enzymes encoded for by the genes PHO5, PHO10 and PHO11 and SDS-polyacrilamide gel electrophoresis data suggested an oligomeric organisation of the enzymes. Isoelectric focusing in polyacrylamide gel with immobilised pH gradient followed by activity staining yielded numerous sharp bands of homopolymeric acid phosphatases forms being different in their pI. The kinetic characterisation of the enzymes revealed differences in Km values, sensitivity to temperature inactivation, inhibition by orthophosphate and the effect of pH on the enzyme activity.

High-level expression of endogenous acid phosphatase inhibits growth and vectorial secretion in Saccharomyces cerevisiae.

The secretion pathway of Saccharomyces cerevisiae was challenged by constitutively overexpressing plasmid-encoded acid phosphatase, a secreted endogenous glycoprotein. A 2-microns-based multicopy plasmid carrying the coding sequence of acid phosphatase under the control of a truncated variant of the strong constitutive glyceraldehyde-3-phosphate dehydrogenase promoter was used for expression. Selection for the promoterless dLEU2 marker leads to a growth arrest. This is not per se due to leucine starvation, but due to intracellular accumulation of highly glycosylated enzymatically active acid phosphatase. Immunofluorescence and cytological analysis indicate that intracellular accumulation of acid phosphatase occurs in a subpopulation of cells. By Ludox-AM density centrifugation, these cells can be enriched on the basis of their higher density. The dense accumulating cells have a higher average plasmid copy number and produce more acid phosphatase than non-accumulating cells of low density. These cells are defective in directed secretion and bud formation, therefore can no longer grow and show dramatic changes in cell morphology. We suggest that the secretion pathway in these cells is overloaded with the high level of acid phosphatase leading to a shutdown in vectorial secretion, subsequently to a standstill in growth and to the intracellular accumulation of further expressed acid phosphatase. We have indications that accumulation of acid phosphatase occurs in the late Golgi, suggesting a limitation of the overall secretion at this stage.

Identification of the yeast ACC1 gene product (acetyl-CoA carboxylase) as the target of the polyketide fungicide soraphen A.

Soraphen A, a polyketide isolated from the myxobacterium Sorangium cellulosum, is a potent inhibitor of fungal growth. We have used a genetic approach to localize the target of this drug, employing Saccharomyces cerevisiae as a model organism. we have isolated soraphen A-resistant mutants and found that all of them map at the same genetic locus and exhibit a broad range of semidominant phenotypes. Data from genetic crosses of soraphen A-resistant clones with an acc1 mutant revealed that ACC1, coding for acetyl-CoA carboxylase (E.C. 6.4.1.2), is tightly linked to soraphen A resistance. Partially-purified enzyme extracts containing acetyl-CoA carboxylase were prepared and assayed for their soraphen A sensitivity. Our experiments showed that the catalytic activity of the wild-type enzyme is inhibited in vitro by soraphen A while the mutant enzyme remains catalytically active. Taken together these data strongly suggest that the ACC1 gene product is the primary target for soraphen A in vivo.

Characterization of the Aspergillus niger pelB gene: structure and regulation of expression.

The nucleotide sequence of pelB, a member of the Aspergillus niger pectin lyase multigene family, has been determined. The pelB gene product, PLB, shares 65% amino acid identity with pectin lyase A (PLA) and 60% with pectin lyase D (PLD). Although growth of pelB multicopy transformants on pectin-containing media results in elevated pelB mRNA levels, pectin lyase B (PLB) is barely detectable. This is probably due to degradation of PLB by acid proteases, since multicopy transformants grown on pectin medium with a high concentration of phosphate, leading to a less rapid decline in pH, secrete detectable amounts of PLB. To produce PLB in high amounts under conditions where few other extracellular enzymes are present, we tried two strategies. Firstly, heterologous expression of the pelB gene in A. nidulans, and secondly, expression of the pelB gene under control of the constitutive A. niger pki promoter.

Temperature sensitivity of the cdc9-1 allele of Saccharomyces cerevisiae DNA ligase is dependent on specific combinations of amino acids in the primary structure of the expressed protein.

In this study we present the characterization of the temperature-sensitive mutant allele cdc9-1 encoding DNA ligase, of Saccharomyces cerevisiae strain A364A by DNA sequencing. Comparison with the published wild-type sequence from strain SK1 revealed 13 nucleotide exchanges between these two sequences, which are derived from non-isogenic genetic backgrounds. Only four of these changes, distributed over the whole coding region, lead to amino acid exchanges in the protein chain. Our analysis of the sequence of the wild-type CDC9 allele from strain A364A revealed differences from the isogenic cdc9-1 allele in only two nucleotides: one silent change and one leading to a single amino acid exchange. The latter is therefore responsible for the temperature-sensitive phenotype. A mosaic protein, in which a region carrying this amino acid exchange has been inserted in place of the corresponding part of CDC9 from the non-isogenic strain SK1, is not temperature sensitive. The exchange of a longer stretch of DNA leading to atteration of three amino acids of the protein compared with the original sequence of SK1 is required to obtain a temperature-sensitive DNA ligase in this strain, while in strain A364A a single amino acid change is sufficient for expression of a temperature-sensitive protein.

Heterologous expression of human microsomal epoxide hydrolase in Saccharomyces cerevisiae. Study of the valpromide-carbamazepine epoxide interaction.

A cDNA of human microsomal epoxide hydrolase (hmEH) was constitutively and inducibly expressed in Saccharomyces cerevisiae. The heterologous enzyme was located mainly in the microsomal fraction of yeast cells. Yeast microsomes containing hmEH exerted styrene oxide hydrolase activity (Km = 300 microM; Vmax = 22 nmol/mg min) as well as carbamazepine epoxide hydrolase activity. The hmEH catalysed exclusively the formation of carbamazepine-10,11-transdihydrodiol, since no carbamazepine-10,11-cisdihydrodiol was detected. Inhibition studies using these microsomes revealed unequivocally hmEH as the target for inhibition by the antiepileptic drug valpromide. A Ki value of 27 microM was determined for the inhibitor valpromide with styrene oxide as substrate. For carbamazepine epoxide, a Ki value of 8.6 microM was obtained, which is well in line with data published for hmEH determined with human liver microsomes. Our results demonstrate the potential of heterologous gene expression in S. cerevisiae and its application to the in vitro study of pharmacological and toxicological problems.

Removal of N-glycosylation sites of the yeast acid phosphatase severely affects protein folding.

The influence of N glycosylation on the production of yeast acid phosphatase was studied. A set of synthetic hypoglycosylation mutants was generated by oligonucleotide-directed mutagenesis of the 12 putative sequons (Asn-X-Ser/Thr). Derepression of the hypoglycosylation mutants and analysis of their molecular sizes showed that all 12 sequons of the wild-type acid phosphatase are glycosylated. Activity measurements in combination with pulse-chase experiments revealed that the specific activity was not impaired by the introduced amino acid exchanges. However, absence of N glycosylation severely affected protein folding. Protein folding was found to be the rate-limiting factor in acid phosphatase secretion, and improper folding resulted in irreversible retention of malfolded acid phosphatase in the endoplasmic reticulum. With a decreasing number of attached glycan chains, less active acid phosphatase was secreted. Efficiency of correct folding was shown to be temperature dependent; i.e., lower temperatures could compensate for the reduction in attached oligosaccharides. In addition, protein folding and stability were shown to depend on both the number and the position of the attached oligosaccharides. N glycosylation was found to occur in a process independent of secondary structures, and thus our data support the model of a cotranslocational mechanism of glycosylation.

A new system for amplifying 2 microns plasmid copy number in Saccharomyces cerevisiae.

The yeast 2 microns plasmid is found in the nucleus of almost all Saccharomyces cerevisiae strains. Its replication is very similar to that of chromosomal DNA. Although the plasmid does not encode essential genes it is stably maintained in the yeast population and exhibits only a small, though detectable, loss rate. This stability is achieved by a plasmid-encoded copy-number control system which ensures constant plasmid levels. For the investigation of 2 microns replication, a yeast strain that is absolutely dependent on this plasmid was constructed. This was achieved by disruption of the chromosomal CDC9 gene, coding for DNA ligase and providing this essential gene on a 2 microns-derived plasmid. This plasmid is absolutely stable under all growth conditions tested. Using the temperature-sensitive mutant allele cdc9-1 we have developed an artificial control system which allows one to change the copy number of 2 microns-derived plasmids solely by changing the incubation temperature.

The yeast phosphatase system.

Yeast cells produce a set of enzymes which are involved in the metabolism of phosphate, and include acid and alkaline phosphatases as well as permeases. Most of these enzymes are synthesized in response to the presence or absence of inorganic phosphate. In the past few years a considerable amount of genetic and molecular evidence has accumulated and a rather precise overall picture emerges which describes the mechanism of phosphate control at the level of gene activation. This mini-review summarizes these data. The main focus lies on the regulatory features associated with the control of transcription of PHO5, a gene coding for most of the regulated acid phosphatase activity produced by yeast cells.

Constitutive and inducible expression of human cytochrome P450IA1 in yeast Saccharomyces cerevisiae: an alternative enzyme source for in vitro studies.

A cDNA of human cytochrome P450IA1 was expressed in yeast Saccharomyces cerevisiae on a multicopy plasmid under the control of the constitutive GAPFL or the inducible PHO5 promoter. Microsomes of transformed yeast contained substantial amounts of the heterologous enzyme as determined by reduced CO-difference spectra (156-68 pmol/mg). Enzyme kinetics with 7-ethoxyresorufin as substrate resulted in a Km value of 92 nM and a Vmax value of 223 pmol/mg/min, which is comparable to data obtained with human liver microsomes. The antimycotic drug ketoconazole (Ki = 22nM) as well as the isozyme specific P450 inhibitor alpha-naphthoflavone (Ki = 1.2 nM) were shown to be strong inhibitors of human P450IA1. Taken together, these data show that heterologous P450 gene expression in yeast is a potent instrument for the study of enzyme specific parameters and might be used to answer further questions with regard to substrate specificity as well as drug interaction in a background with no interfering activities.

The influence of GAP promoter variants on hirudin production, average plasmid copy number and cell growth in Saccharomyces cerevisiae.

The yeast Saccharomyces cerevisiae has been engineered to synthesize and secrete desulfato-hirudin (hirudin), a thrombin inhibitor from the leech Hirudo medicinalis. The synthetic gene coding for hirudin was expressed constitutively under the control of four size-variants of the yeast glyceraldehyde-3-phosphate dehydrogenase promoter (GAP) and cloned into a 2 mu based multicopy yeast vector. The constitutive action of the four promoter variants was confirmed by demonstrating that the expression and secretion of hirudin is growth-related. The different efficiencies of the promoter variants not only affected hirudin expression but also led to changes in several cellular parameters, such as cell growth, average plasmid copy number and plasmid stability. The observed changes show that yeast cells establish a specific equilibrium for each promoter variant. We conclude, that the adjustment of cellular parameters in response to the expression levels of a heterologous protein is regulated by two counteracting selective forces: (1) the need for complementation of the auxotrophic host marker by the plasmid-encoded selection gene which, in the case of dLEU2, requires several plasmid copies; and (2) a selective advantage of cells with a lower copy number enabling them to escape the burden of heterologous protein production.

In vivo translocation of the cell wall acid phosphatase across the yeast endoplasmic reticulum membrane: are there multiple signals for the targeting process?

The repressible Saccharomyces cerevisiae acid phosphatase (APase) coded by the PHO5 gene is a cell wall protein that follows the yeast secretory pathway. We had previously described the in vivo fate of a multicopy plasmid-encoded modified protein, lacking 15 out of 17 signal peptide amino acids. This modified protein accumulates mainly within the cell as an inactive unglycosylated form. However 30% of this precursor is translocated, glycosylated and dispatched to the cell wall. We establish, in the present report, that this phenomenon did not result from an overproduction of the plasmid encoded protein, since it was also observed in a normal single copy situation. The secretion persisted after a deletion including the single hydrophobic segment present in the N-terminus of the mature protein. The entry of both wild type and mutant APase into the ER was inhibited in sec62 mutants suggesting that the SEC62 gene product would not be implicated in signal peptide recognition.