Floriana Capuano

Cytosine DNA methylation is found in Drosophila melanogaster but absent in Saccharomyces cerevisiae, Schizosaccharomyces pombe, and other yeast species.

The methylation of cytosine to 5-methylcytosine (5-meC) is an important epigenetic DNA modification in many bacteria, plants, and mammals, but its relevance for important model organisms, including Caenorhabditis elegans and Drosophila melanogaster, is still equivocal. By reporting the presence of 5-meC in a broad variety of wild, laboratory, and industrial yeasts, a recent study also challenged the dogma about the absence of DNA methylation in yeast species. We would like to bring to attention that the protocol used for gas chromatography/mass spectrometry involved hydrolysis of the DNA preparations. As this process separates cytosine and 5-meC from the sugar phosphate backbone, this method is unable to distinguish DNA- from RNA-derived 5-meC. We employed an alternative LC–MS/MS protocol where by targeting 5-methyldeoxycytidine moieties after enzymatic digestion, only 5-meC specifically derived from DNA is quantified. This technique unambiguously identified cytosine DNA methylation in Arabidopsis thaliana (14.0% of cytosines methylated), Mus musculus (7.6%), and Escherichia coli (2.3%). Despite achieving a detection limit at 250 attomoles (corresponding to <0.00002 methylated cytosines per nonmethylated cytosine), we could not confirm any cytosine DNA methylation in laboratory and industrial strains of Saccharomyces cerevisiae, Schizosaccharomyces pombe, Saccharomyces boulardii, Saccharomyces paradoxus, or Pichia pastoris. The protocol however unequivocally confirmed DNA methylation in adult Drosophila melanogaster at a value (0.034%) that is up to 2 orders of magnitude below the detection limit of bisulphite sequencing. Thus, 5-meC is a rare DNA modification in drosophila but absent in yeast.

A prototrophic deletion mutant collection for yeast metabolomics and systems biology

Auxotrophic markers - mutations within genes encoding enzymes in pathways for the biosynthesis of metabolic building blocks, such as an amino acid or nucleotide, are used as selection markers in the vast majority of yeast genetics and genomics experiments 1-3. The nutritional deficiency caused by the mutation (auxotrophy) can be compensated by supplying the required nutrient in the growth medium. This compensation, however, is not necessarily quantitative because such mutations influence a number of physiological parameters and may act in combination 2,4,5 . The construction of a prototrophic derivative of the parent strain of the widely used genome-scale yeast deletion collection1 has confirmed the need to remove auxotrophic markers in order to reduce bias in physiological and metabolic studies 2 . Moreover, flux balance analyses using a genome-wide metabolic model (Yeast 5)6 indicate that the activity status of some 200-300 reactions changes between different auxotrophic strains and the wild type. To alleviate this bias we have constructed a version of the haploid deletant collection restoring prototrophy in the genetic background, such that influence of auxotrophy to the phenotype of a given gene deletion is prevented. This new deletant library facilitates the exploitation of yeast in both functional genomics and quantitative systems biology. The physiological impact of auxotrophy was assessed by monitoring the growth of 16 yeast strains carrying all possible combinations of the markers (histidine (his3Δ1), leucine (leu2Δ), methionine (met15Δ) and uracil (ura3Δ) used in the MATa version of the yeast deletion collection1. All markers and their combinations affected yeast growth, but without altering the adenylate (ATP, ADP, AMP) energy charge (Fig 1 a). As the most critical phenotypic quantity, the maximum specific growth rate (μmax) varied between 0.125 h−1 (leu2Δ) and 0.20 h−1 (his3Δmet15Δura3Δ), rendering quantitative comparisons between these strains impossible (Fig 1 a, Suppl. Table 1). These growth differences were not explained by the different media supplementations, as i) prototrophic yeast exhibited a different and substantially less diverse growth pattern in the 16 minimal media (Fig 1c, left panel; media recipes are given in the Supplementary material); and ii) growth differences where altered, but not abrogated, when other proteogenic amino acids were supplemented as well (synthetic complete (SC) medium; Fig 1b). Importantly, on both types of media, complex interactions between all auxotrophic mutations were observed. For instance, restoring MET15 in leu2Δura3Δhis3Δmet15Δ (0.185 h−1 → 0.164 h−1) or leu2Δura3Δmet15Δ (0.162 h−1 → 0.149 h−1) had a negative effect on μmax, but surprisingly promoted growth in leu2Δhis3Δmet15Δ (0.136 h−1 → 0.173 h−1) (Fig 1a); restoring LEU2 in leu2Δura3Δhis3Δ (0.164 h−1 → 0.185 h−1) or leu2Δura3Δmet15Δ (0.136 h−1 → 0.161 h−1) had a positive effect, but not in leu2Δura3Δhis3Δmet15Δ (0.185 h−1 → 0.186 h−1) (Fig 1a; Suppl. Table 1). Thus, although blocking different pathways, all markers influence each other, indicating that they have a wide-ranging and combinatorial influence on the metabolic network. Figure 1 The combinatorial impact of yeast auxotrophic markers In batch culture experiments, further problems arise from the unequal consumption of amino acid supplements resulting in cultivation phase-dependent starvation. Growth of BY4741 (the auxotrophic parent of the standard yeast gene-deletion collection 1) in SC media depleted nutrients in a way they became first limiting for met15Δ, then for leu2Δ, his3Δ1 and finally for ura3Δ auxotrophic yeast (Fig 1d). This effect could not be compensated by increasing amino acid supplementation(s), as this inhibited cell growth (Fig 1c, right panel). Chronological lifespan (CLS) is a phenotype that is profoundly influenced by both nutrient supplementation and growth rate. Indeed, we observed an increase in stationary phase survival in YPD media upon restoring prototrophy. In a competitive growth experiment, auxotrophic cells had lost their colony-forming capacities within 10 days, but their prototrophic counterparts were perfectly viable for more than 20 days (Fig 1e). Longer CLS of prototrophic versus auxotrophic yeast was also reported for other backgrounds, and in synthetic media nutrient starvation shortened the lifespan of auxotrophic cells 7,8. Restoring protrotrophy is thus, to our knowledge, one of the most powerful genetic modifications for extending CLS. Hence, as auxotrophic markers have substantial and combinatorial influences on fundamental biological parameters such as growth and ageing, auxotrophic genome resources introduce bias for analyzing physiological parameters and even more to quantitative studies addressing the metabolic network. We would thus encourage the yeast community to switch, where possible, to prototrophic yeast for experiments in transcriptomics, proteomics, and metabolomics. To create a prototrophic resource for genome-scale experiments, we re-introduced auxotrophic markers into the MATa versions of the S288c-based deletion collection (5185 strains)1 and the titratable promoter essential collection (839 strains)3. These strains were transformed with a centromere-containing single-copy vector (minichromosome), containing the chromosome VI centromere, the autonomous replication sequence of HHF1 (ARSH4)9, and the marker genes HIS3, URA3, LEU2, and MET15 under control of their endogenous promoter sequences (pHLUM; Suppl Fig 1, Addgene ID 40276). Under non-selective conditions, the vector was transmitted in 99.15% of cell divisions (0.85% segregation mean over 20 generations). After 20 days, all cells were found prototrophic due to their positive selection (Fig 1e), facilitating screens on both selective and non-selective media. Furthermore, pHLUM- transformed BY4741 derivates wild type for HIS3, LEU2, MET15 or URA3 grew similar as BY4741 pHLUM (Suppl. Fig 2), indicating that the minichromosome fully restored prototrophy. The titratable-promoter essential collection3 was exploited to demonstrate screening capacities. By replicating original and prototrophic strains onto doxycycline-containing media, we found that 13 of the 370 lethal phenotypes were compensated (Fig 1 f, Suppl. Table 2). Thus, auxotrophic markers do not only influence physiological parameters, they are also responsible for a number of essential phenotypes. Since all strains possess a native metabolic network, the new library reduces bias from the use of auxotrophic markers in functional genomics and metabolic systems biology. Based on the pHLUM minichromosome, which is counterselectable, the new resource retains full compatibility with the popular S288c knock-out and essential collections 1,3. However, the use of a plasmid will introduce confounding factors to those mutants which have deficits in plasmid stability and segregation. The library is distributed as arrayed on 96-well plates (Euroscarf, Frankfurt), and contains a deep-red coloured and counter-selectable mutant (ade12Δ) on both universal and plate-specific positions, which simplifies plate orientation and identification, and can serve as replicate-control in quantitative metabolomics experiments (Suppl. Fig 3).

The beauty of being (label)-free: sample preparation methods for SWATH-MS and next-generation targeted proteomics

The combination of qualitative analysis with label-free quantification has greatly facilitated the throughput and flexibility of novel proteomic techniques. However, such methods rely heavily on robust and reproducible sample preparation procedures. Here, we benchmark a selection of in gel, on filter, and in solution digestion workflows for their application in label-free proteomics. Each procedure was associated with differing advantages and disadvantages. The in gel methods interrogated were cost effective, but were limited in throughput and digest efficiency. Filter-aided sample preparations facilitated reasonable processing times and yielded a balanced representation of membrane proteins, but led to a high signal variation in quantification experiments. Two in solution digest protocols, however, gave optimal performance for label-free proteomics. A protocol based on the detergent RapiGest led to the highest number of detected proteins at second-best signal stability, while a protocol based on acetonitrile-digestion, RapidACN, scored best in throughput and signal stability but came second in protein identification. In addition, we compared label-free data dependent (DDA) and data independent (SWATH) acquisition on a TripleTOF 5600 instrument. While largely similar in protein detection, SWATH outperformed DDA in quantification, reducing signal variation and markedly increasing the number of precisely quantified peptides.The combination of qualitative analysis with label-free quantification has greatly facilitated the throughput and flexibility of novel proteomic techniques. However, such methods rely heavily on robust and reproducible sample preparation procedures. Here, we benchmark a selection of in gel, on filter, and in solution digestion workflows for their application in label-free proteomics. Each procedure was associated with differing advantages and disadvantages. The in gel methods interrogated were cost effective, but were limited in throughput and digest efficiency. Filter-aided sample preparations facilitated reasonable processing times and yielded a balanced representation of membrane proteins, but led to a high signal variation in quantification experiments. Two in solution digest protocols, however, gave optimal performance for label-free proteomics. A protocol based on the detergent RapiGest led to the highest number of detected proteins at second-best signal stability, while a protocol based on acetonitrile-digestion, RapidACN, scored best in throughput and signal stability but came second in protein identification. In addition, we compared label-free data dependent (DDA) and data independent (SWATH) acquisition on a TripleTOF 5600 instrument. While largely similar in protein detection, SWATH outperformed DDA in quantification, reducing signal variation and markedly increasing the number of precisely quantified peptides.

The metabolic background is a global player in Saccharomyces gene expression epistasis

The regulation of gene expression in response to nutrient availability is fundamental to the genotype–phenotype relationship. The metabolic–genetic make-up of the cell, as reflected in auxotrophy, is hence likely to be a determinant of gene expression. Here, we address the importance of the metabolic–genetic background by monitoring transcriptome, proteome and metabolome in a repertoire of 16 Saccharomyces cerevisiae laboratory backgrounds, combinatorially perturbed in histidine, leucine, methionine and uracil biosynthesis. The metabolic background affected up to 85% of the coding genome. Suggesting widespread confounding, these transcriptional changes show, on average, 83% overlap between unrelated auxotrophs and 35% with previously published transcriptomes generated for non-metabolic gene knockouts. Background-dependent gene expression correlated with metabolic flux and acted, predominantly through masking or suppression, on 88% of transcriptional interactions epistatically. As a consequence, the deletion of the same metabolic gene in a different background could provoke an entirely different transcriptional response. Propagating to the proteome and scaling up at the metabolome, metabolic background dependencies reveal the prevalence of metabolism-dependent epistasis at all regulatory levels. Urging a fundamental change of the prevailing laboratory practice of using auxotrophs and nutrient supplemented media, these results reveal epistatic intertwining of metabolism with gene expression on the genomic scale.

Using the canary genome to decipher the evolution of hormone-sensitive gene regulation in seasonal singing birds

BackgroundWhile the song of all songbirds is controlled by the same neural circuit, the hormone dependence of singing behavior varies greatly between species. For this reason, songbirds are ideal organisms to study ultimate and proximate mechanisms of hormone-dependent behavior and neuronal plasticity.ResultsWe present the high quality assembly and annotation of a female 1.2-Gbp canary genome. Whole genome alignments between the canary and 13 genomes throughout the bird taxa show a much-conserved synteny, whereas at the single-base resolution there are considerable species differences. These differences impact small sequence motifs like transcription factor binding sites such as estrogen response elements and androgen response elements. To relate these species-specific response elements to the hormone-sensitivity of the canary singing behavior, we identify seasonal testosterone-sensitive transcriptomes of major song-related brain regions, HVC and RA, and find the seasonal gene networks related to neuronal differentiation only in the HVC. Testosterone-sensitive up-regulated gene networks of HVC of singing males concerned neuronal differentiation. Among the testosterone-regulated genes of canary HVC, 20% lack estrogen response elements and 4 to 8% lack androgen response elements in orthologous promoters in the zebra finch.ConclusionsThe canary genome sequence and complementary expression analysis reveal intra-regional evolutionary changes in a multi-regional neural circuit controlling seasonal singing behavior and identify gene evolution related to the hormone-sensitivity of this seasonal singing behavior. Such genes that are testosterone- and estrogen-sensitive specifically in the canary and that are involved in rewiring of neurons might be crucial for seasonal re-differentiation of HVC underlying seasonal song patterning.

Tpo1-mediated spermine and spermidine export controls cell cycle delay and times antioxidant protein expression during the oxidative stress response

Cells counteract oxidative stress by altering metabolism, cell cycle and gene expression. However, the mechanisms that coordinate these adaptations are only marginally understood. Here we provide evidence that timing of these responses in yeast requires export of the polyamines spermidine and spermine. We show that during hydrogen peroxide (H2O2) exposure, the polyamine transporter Tpo1 controls spermidine and spermine concentrations and mediates induction of antioxidant proteins, including Hsp70, Hsp90, Hsp104 and Sod1. Moreover, Tpo1 determines a cell cycle delay during adaptation to increased oxidant levels, and affects H2O2 tolerance. Thus, central components of the stress response are timed through Tpo1‐controlled polyamine export.

LC-MS/MS methods for absolute quantification and identification of proteins associated with chimeric plant oil bodies.

Oil bodies (OBs) are plant cell organelles that consist of a lipid core surrounded by a phospholipid monolayer embedded with specialized proteins such as oleosins. Recombinant proteins expressed in plants can be targeted to OBs as fusions with oleosin. This expression strategy is attractive because OBs are easily enriched and purified from other cellular components, based on their unique physicochemical properties. For recombinant OBs to be a potential therapeutic agent in biomedical applications, it is necessary to comprehensively analyze and quantify both endogenous and heterologously expressed OB proteins. In this study, a mass spectrometry (MS)-based method was developed to accurately quantify an OB-targeted heterologously expressed fusion protein that has potential as a therapeutic agent. The effect of the chimeric oleosin expression upon the OB proteome in transgenic plants was also investigated, and the identification of new potential OB residents was pursued through a variety of liquid chromatography (LC)-MS/MS approaches. The results showed that the accumulation of the fusion protein on OBs was low. Moreover, no significant differences in the accumulation of OB proteins were revealed between transgenic and wild-type seeds. The identification of five new putative components of OB proteome was also reported.

Plant-Based Vaccine Delivery Strategies

Subunit vaccine formulations were prepared in the near past by purifying antigenic components known to activate protective immune responses directly from the pathogen. Nowadays, thanks to the development of high performance gene engineering and biochemical procedures, subunit vaccines commence to be formulated with recombinant versions of protective antigens. These molecules can be synthesized using heterologous hosts such as bacteria, yeast, insect and mammalian cells. To this aim, the alternative use of plants is growing out of advances in methods for foreign gene expression. Plants represent an opportunity in the field of vaccine technology in that this expression system ensures rapidity, low costs, easy scaling up and intrinsic bio-safety of the final product. Nonetheless, the exploitation of plants only as mere “biofactories” of antigens thwarts many of their potentialities.

Machine Learning Predicts the Yeast Metabolome from the Quantitative Proteome of Kinase Knockouts

Summary A challenge in solving the genotype-to-phenotype relationship is to predict a cell’s metabolome, believed to correlate poorly with gene expression. Using comparative quantitative proteomics, we found that differential protein expression in 97 Saccharomyces cerevisiae kinase deletion strains is non-redundant and dominated by abundance changes in metabolic enzymes. Associating differential enzyme expression landscapes to corresponding metabolomes using network models provided reasoning for poor proteome-metabolome correlations; differential protein expression redistributes flux control between many enzymes acting in concert, a mechanism not captured by one-to-one correlation statistics. Mapping these regulatory patterns using machine learning enabled the prediction of metabolite concentrations, as well as identification of candidate genes important for the regulation of metabolism. Overall, our study reveals that a large part of metabolism regulation is explained through coordinated enzyme expression changes. Our quantitative data indicate that this mechanism explains more than half of metabolism regulation and underlies the interdependency between enzyme levels and metabolism, which renders the metabolome a predictable phenotype.