Ethan O. Perlstein

Global nucleosome occupancy in yeast

BackgroundAlthough eukaryotic genomes are generally thought to be entirely chromatin-associated, the activated PHO5 promoter in yeast is largely devoid of nucleosomes. We systematically evaluated nucleosome occupancy in yeast promoters by immunoprecipitating nucleosomal DNA and quantifying enrichment by microarrays.ResultsNucleosome depletion is observed in promoters that regulate active genes and/or contain multiple evolutionarily conserved motifs that recruit transcription factors. The Rap1 consensus was the only binding motif identified in a completely unbiased search of nucleosome-depleted promoters. Nucleosome depletion in the vicinity of Rap1 consensus sites in ribosomal protein gene promoters was also observed by real-time PCR and micrococcal nuclease digestion. Nucleosome occupancy in these regions was increased by the small molecule rapamycin or, in the case of the RPS11B promoter, by removing the Rap1 consensus sites.ConclusionsThe presence of transcription factor-binding motifs is an important determinant of nucleosome depletion. Most motifs are associated with marked depletion only when they appear in combination, consistent with a model in which transcription factors act collaboratively to exclude nucleosomes and gain access to target sites in the DNA. In contrast, Rap1-binding sites cause marked depletion under steady-state conditions. We speculate that nucleosome depletion enables Rap1 to define chromatin domains and alter them in response to environmental cues.

Harnessing gene expression to identify the genetic basis of drug resistance

The advent of cost‐effective genotyping and sequencing methods have recently made it possible to ask questions that address the genetic basis of phenotypic diversity and how natural variants interact with the environment. We developed Camelot (CAusal Modelling with Expression Linkage for cOmplex Traits), a statistical method that integrates genotype, gene expression and phenotype data to automatically build models that both predict complex quantitative phenotypes and identify genes that actively influence these traits. Camelot integrates genotype and gene expression data, both generated under a reference condition, to predict the response to entirely different conditions. We systematically applied our algorithm to data generated from a collection of yeast segregants, using genotype and gene expression data generated under drug‐free conditions to predict the response to 94 drugs and experimentally confirmed 14 novel gene–drug interactions. Our approach is robust, applicable to other phenotypes and species, and has potential for applications in personalized medicine, for example, in predicting how an individual will respond to a previously unseen drug.

The Antidepressant Sertraline Targets Intracellular Vesiculogenic Membranes in Yeast

Numerous studies have shown that the clinical antidepressant sertraline (Zoloft) is biologically active in model systems, including fungi, which do not express its putative protein target, the serotonin/5-HT transporter, thus demonstrating the existence of one or more secondary targets. Here we show that in the absence of its putative protein target, sertraline targets phospholipid membranes that comprise the acidic organelles of the intracellular vesicle transport system by a mechanism consistent with the bilayer couple hypothesis. On the basis of a combination of drug-resistance selection and chemical-genomic screening, we hypothesize that loss of vacuolar ATPase activity reduces uptake of sertraline into cells, whereas dysregulation of clathrin function reduces the affinity of membranes for sertraline. Remarkably, sublethal doses of sertraline stimulate growth of mutants with impaired clathrin function. Ultrastructural studies of sertraline-treated cells revealed a phenotype that resembles phospholipidosis induced by cationic amphiphilic drugs in mammalian cells. Using reconstituted enzyme assays, we also demonstrated that sertraline inhibits phospholipase A1 and phospholipase D, exhibits mixed effects on phospholipase C, and activates phospholipase A2. Overall, our study identifies two evolutionarily conserved membrane-active processes—vacuolar acidification and clathrin-coat formation—as modulators of sertralines action at membranes.

Accumulation of an antidepressant in vesiculogenic membranes of yeast cells triggers autophagy.

Many antidepressants are cationic amphipaths, which spontaneously accumulate in natural or reconstituted membranes in the absence of their specific protein targets. However, the clinical relevance of cellular membrane accumulation by antidepressants in the human brain is unknown and hotly debated. Here we take a novel, evolutionarily informed approach to studying the effects of the selective-serotonin reuptake inhibitor sertraline/Zoloft® on cell physiology in the model eukaryote Saccharomyces cerevisiae (budding yeast), which lacks a serotonin transporter entirely. We biochemically and pharmacologically characterized cellular uptake and subcellular distribution of radiolabeled sertraline, and in parallel performed a quantitative ultrastructural analysis of organellar membrane homeostasis in untreated vs. sertraline-treated cells. These experiments have revealed that sertraline enters yeast cells and then reshapes vesiculogenic membranes by a complex process. Internalization of the neutral species proceeds by simple diffusion, is accelerated by proton motive forces generated by the vacuolar H+-ATPase, but is counteracted by energy-dependent xenobiotic efflux pumps. At equilibrium, a small fraction (10–15%) of reprotonated sertraline is soluble while the bulk (90–85%) partitions into organellar membranes by adsorption to interfacial anionic sites or by intercalation into the hydrophobic phase of the bilayer. Asymmetric accumulation of sertraline in vesiculogenic membranes leads to local membrane curvature stresses that trigger an adaptive autophagic response. In mutants with altered clathrin function, this adaptive response is associated with increased lipid droplet formation. Our data not only support the notion of a serotonin transporter-independent component of antidepressant function, but also enable a conceptual framework for characterizing the physiological states associated with chronic but not acute antidepressant administration in a model eukaryote.

Using Expression and Genotype to Predict Drug Response in Yeast

Personalized, or genomic, medicine entails tailoring pharmacological therapies according to individual genetic variation at genomic loci encoding proteins in drug-response pathways. It has been previously shown that steady-state mRNA expression can be used to predict the drug response (i.e., sensitivity or resistance) of non-genotyped mammalian cancer cell lines to chemotherapeutic agents. In a real-world setting, clinicians would have access to both steady-state expression levels of patient tissue(s) and a patients genotypic profile, and yet the predictive power of transcripts versus markers is not well understood. We have previously shown that a collection of genotyped and expression-profiled yeast strains can provide a model for personalized medicine. Here we compare the predictive power of 6,229 steady-state mRNA transcript levels and 2,894 genotyped markers using a pattern recognition algorithm. We were able to predict with over 70% accuracy the drug sensitivity of 104 individual genotyped yeast strains derived from a cross between a laboratory strain and a wild isolate. We observe that, independently of drug mechanism of action, both transcripts and markers can accurately predict drug response. Marker-based prediction is usually more accurate than transcript-based prediction, likely reflecting the genetic determination of gene expression in this cross.

Evolutionarily Conserved Optimization of Amino Acid Biosynthesis

The “cognate bias hypothesis” states that early in evolutionary history the biosynthetic enzymes for amino acid x gradually lost residues of x, thereby reducing the threshold for deleterious effects of x scarcity. The resulting reduction in cognate amino acid composition of the enzymes comprising a particular amino acid biosynthetic pathway is predicted to confer a selective growth advantage on cells. Bioinformatic evidence from protein-sequence data of two bacterial species previously demonstrated reduced cognate bias in amino acid biosynthetic pathways. Here we show that cognate bias in amino acid biosynthesis is present in the other domains of life—Archaebacteria and Eukaryota. We also observe evolutionarily conserved underrepresentations (e.g., glycine in methionine biosynthesis) and overrepresentations (e.g., tryptophan in asparagine biosynthesis) of amino acids in noncognate biosynthetic pathways, which can be explained by secondary amino acid metabolism. Additionally, we experimentally validate the cognate bias hypothesis using the yeast Saccharomyces cerevisiae. Specifically, we show that the degree to which growth declines following amino acid deprivation is negatively correlated with the degree to which an amino acid is underrepresented in the enzymes that comprise its cognate biosynthetic pathway. Moreover, we demonstrate that cognate fold representation is more predictive of growth advantage than a host of other potential growth-limiting factors, including an amino acid’s metabolic cost or its intracellular concentration and compartmental distribution.

Amino Acid Metabolic Origin as an Evolutionary Influence on Protein Sequence in Yeast

The metabolic cycle of Saccharomyces cerevisiae consists of alternating oxidative (respiration) and reductive (glycolysis) energy-yielding reactions. The intracellular concentrations of amino acid precursors generated by these reactions oscillate accordingly, attaining maximal concentration during the middle of their respective yeast metabolic cycle phases. Typically, the amino acids themselves are most abundant at the end of their precursor’s phase. We show that this metabolic cycling has likely biased the amino acid composition of proteins across the S. cerevisiae genome. In particular, we observed that the metabolic source of amino acids is the single most important source of variation in the amino acid compositions of functionally related proteins and that this signal appears only in (facultative) organisms using both oxidative and reductive metabolism. Periodically expressed proteins are enriched for amino acids generated in the preceding phase of the metabolic cycle. Proteins expressed during the oxidative phase contain more glycolysis-derived amino acids, whereas proteins expressed during the reductive phase contain more respiration-derived amino acids. Rare amino acids (e.g., tryptophan) are greatly overrepresented or underrepresented, relative to the proteomic average, in periodically expressed proteins, whereas common amino acids vary by a few percent. Genome-wide, we infer that 20,000 to 60,000 residues have been modified by this previously unappreciated pressure. This trend is strongest in ancient proteins, suggesting that oscillating endogenous amino acid availability exerted genome-wide selective pressure on protein sequences across evolutionary time.

Quantifying fitness distributions and phenotypic relationships in recombinant yeast populations

Studies of the role of sex in evolution typically involve a longitudinal comparison of a single ancestor to several intermediate descendants and to one terminally evolved descendant after many generations of adaptation under a given selective regime. Here we take a complementary, statistical approach to sex in evolution, by describing the distribution of phenotypic similarity in a population of yeast F1 meiotic recombinants. By applying graph theory to fitness measurements of thousands of Saccharomyces cerevisiae recombinants treated with 10 mechanistically distinct, growth-inhibitory small-molecule perturbagens (SMPs), we show that the network of phenotypic similarity among F1 recombinants exhibits a scale-free degree distribution. F1 recombinants are often phenotypically unique and sometimes exceptional, and their fitness strengths are unevenly distributed across the 10 compound treatments. By contrast, highly phenotypically similar F1 recombinants constitute failing hubs that display below-average fitness across all compound treatments and are candidate substrates for purifying selection. Comparison of the F1 generation with the parental strains reveals that (i) there is a specialist more fit in any given single condition than any of the parents but (ii) only rarely are there generalists that exhibit greater fitness than both parental strains across a majority of conditions. This analysis allows us to evaluate and to gain better theoretical understanding of the costs and benefits of sex in the F1 generation.