Protein Localization In Yeast Research Articles

Democratized image analytics by visual programming through integration of deep models and small-scale machine learning

Analysis of biomedical images requires computational expertize that are uncommon among biomedical scientists. Deep learning approaches for image analysis provide an opportunity to develop user-friendly tools for exploratory data analysis. Here, we use the visual programming toolbox Orange (http://orange.biolab.si) to simplify image analysis by integrating deep-learning embedding, machine learning procedures, and data visualization. Orange supports the construction of data analysis workflows by assembling components for data preprocessing, visualization, and modeling. We equipped Orange with components that use pre-trained deep convolutional networks to profile images with vectors of features. These vectors are used in image clustering and classification in a framework that enables mining of image sets for both novel and experienced users. We demonstrate the utility of the tool in image analysis of progenitor cells in mouse bone healing, identification of developmental competence in mouse oocytes, subcellular protein localization in yeast, and developmental morphology of social amoebae.

Mitochondrial targeting increases specific activity of a heterologous valine assimilation pathway in Saccharomyces cerevisiae

Bio-based isobutantol is a sustainable ‘drop in’ substitute for petroleum-based fuels. However, well-studied production routes, such as the Ehrlich pathway, have yet to be commercialized despite more than a century of research. The more versatile bacterial valine catabolism may be a competitive alternate route producing not only an isobutanol precursor but several carboxylic acids with applications as biomonomers, and building blocks for other advanced biofuels. Here, we transfer the first two committed steps of the pathway from pathogenic Pseudomonas aeruginosa PAO1 to yeast to evaluate their activity in a safer model organism. Genes encoding the heteroligomeric branched chain keto-acid dehydrogenase (BCKAD; bkdA1, bkdA2, bkdB, lpdV), and the homooligomeric acyl-CoA dehydrogenase (ACD; acd1) were tagged with fluorescence epitopes and targeted for expression in either the mitochondria or cytoplasm of S. cerevisiae. We verified the localization of our constructs with confocal fluorescence microscopy before measuring the activity of tag-free constructs. Despite reduced heterologous expression of mitochondria-targeted enzymes, their specific activities were significantly improved with total enzyme activities up to 138% greater than those of enzymes expressed in the cytoplasm. In total, our results demonstrate that the choice of protein localization in yeast has significant impact on heterologous activity, and suggests a new path forward for isobutanol production.

Immuno- and correlative light microscopy-electron tomography methods for 3D protein localization in yeast.

Compartmentalization of eukaryotic cells is created and maintained through membrane rearrangements that include membrane transport and organelle biogenesis. Three-dimensional reconstructions with nanoscale resolution in combination with protein localization are essential for an accurate molecular dissection of these processes. The yeast Saccharomyces cerevisiae is a key model system for identifying genes and characterizing pathways essential for the organization of cellular ultrastructures. Electron microscopy studies of yeast, however, have been hampered by the presence of a cell wall that obstructs penetration of resins and cryoprotectants, and by the protein dense cytoplasm, which obscures the membrane details. Here we present an immuno-electron tomography (IET) method, which allows the determination of protein distribution patterns on reconstructed organelles from yeast. In addition, we extend this IET approach into a correlative light microscopy-electron tomography procedure where structures positive for a specific protein localized through a fluorescent signal are resolved in 3D. These new investigative tools for yeast will help to advance our understanding of the endomembrane system organization in eukaryotic cells.

Predicting Yeast Protein Localization Sites by a New Clustering Algorithm Based on Weighted Feature Ensemble

The analysis of protein localization sites is an important task in bioinformatics. Predicting the yeast protein localization sites is a promising domain among numerous research methods based on the yeast protein measurement data which have multiple indexes/features. In order to reflect the different contributions of those features to predicting tasks, a clustering algorithm based on weighted feature ensemble (WFE) is proposed to predict yeast protein localization sites on the basis of the gathered yeast protein localization data. WFE process firstly assigns different weights to features, and then the results are computed and presented to obtain the best outcome. Experimental results on our algorithm based on WFE and other several clustering algorithms based on the ideas of weighted features have shown that our new algorithm outperformed the other feature weighting type algorithms in accuracy and stability.

Protein networks markedly improve prediction of subcellular localization in multiple eukaryotic species

The function of a protein is intimately tied to its subcellular localization. Although localizations have been measured for many yeast proteins through systematic GFP fusions, similar studies in other branches of life are still forthcoming. In the interim, various machine-learning methods have been proposed to predict localization using physical characteristics of a protein, such as amino acid content, hydrophobicity, side-chain mass and domain composition. However, there has been comparatively little work on predicting localization using protein networks. Here, we predict protein localizations by integrating an extensive set of protein physical characteristics over a protein's extended protein–protein interaction neighborhood, using a classification framework called ‘Divide and Conquer k-Nearest Neighbors’ (DC-kNN). These predictions achieve significantly higher accuracy than two well-known methods for predicting protein localization in yeast. Using new GFP imaging experiments, we show that the network-based approach can extend and revise previous annotations made from high-throughput studies. Finally, we show that our approach remains highly predictive in higher eukaryotes such as fly and human, in which most localizations are unknown and the protein network coverage is less substantial.

Cex1p is a novel cytoplasmic component of the Saccharomyces cerevisiae nuclear tRNA export machinery

The Saccharomyces cerevisiae Yor112wp, which we named Cex1p, was identified using a yeast tRNA three-hybrid interaction approach and an in vivo nuclear tRNA export assay as a cytoplasmic component of the nuclear tRNA export machinery. Cex1p binds tRNA saturably, and associates with the nuclear pore complex by interacting directly with Nup116p. Cex1p co-purifies with the nuclear tRNA export receptors Los1p and Msn5p, the eukaryotic elongation factor eEF-1A, which delivers aminoacylated tRNAs to the ribosome, and the RanGTPase Gsp1p, but not with Cca1p, a tRNA maturation enzyme that facilitates translocation of non-aminoacylated tRNAs across the nuclear pore complex. Depletion of Cex1p and eEF-1A or Los1p significantly reduced the efficiency of nuclear tRNA export. Cex1p interacts with Los1p but not with eEF-1A in vitro. These findings suggest that Cex1p is a component of the nuclear aminoacylation-dependent tRNA export pathway in S. cerevisiae. They also suggest that Cex1p collects aminoacyl-tRNAs from the nuclear export receptors at the cytoplasmic side of the nuclear pore complex, and transfers them to eEF-1A using a channelling mechanism.

MPact: the MIPS protein interaction resource on yeast

In recent years, the Munich Information Center for Protein Sequences (MIPS) yeast protein–protein interaction (PPI) dataset has been used in numerous analyses of protein networks and has been called a gold standard because of its quality and comprehensiveness [H. Yu, N. M. Luscombe, H. X. Lu, X. Zhu, Y. Xia, J. D. Han, N. Bertin, S. Chung, M. Vidal and M. Gerstein (2004) Genome Res., 14, 1107–1118]. MPact and the yeast protein localization catalog provide information related to the proximity of proteins in yeast. Beside the integration of high-throughput data, information about experimental evidence for PPIs in the literature was compiled by experts adding up to 4300 distinct PPIs connecting 1500 proteins in yeast. As the interaction data is a complementary part of CYGD, interactive mapping of data on other integrated data types such as the functional classification catalog [A. Ruepp, A. Zollner, D. Maier, K. Albermann, J. Hani, M. Mokrejs, I. Tetko, U. Güldener, G. Mannhaupt, M. Münsterkötter and H. W. Mewes (2004) Nucleic Acids Res., 32, 5539–5545] is possible. A survey of signaling proteins and comparison with pathway data from KEGG demonstrates that based on these manually annotated data only an extensive overview of the complexity of this functional network can be obtained in yeast. The implementation of a web-based PPI-analysis tool allows analysis and visualization of protein interaction networks and facilitates integration of our curated data with high-throughput datasets. The complete dataset as well as user-defined sub-networks can be retrieved easily in the standardized PSI-MI format. The resource can be accessed through .

The Spatial Organization of Lipid Synthesis in the Yeast Saccharomyces cerevisiae Derived from Large Scale Green Fluorescent Protein Tagging and High Resolution Microscopy

The localization pattern of proteins involved in lipid metabolism in the yeast Saccharomyces cerevisiae was determined using C-terminal green fluorescent protein tagging and high resolution confocal laser scanning microscopy. A list of 493 candidate proteins ( approximately 9% of the yeast proteome) was assembled based on proteins of known function in lipid metabolism, their interacting proteins, proteins defined by genetic interactions, and regulatory factors acting on selected genes or proteins. Overall 400 (81%) transformants yielded a positive green fluorescent protein signal, and of these, 248 (62% of the 400) displayed a localization pattern that was not cytosolic. Observations for many proteins with known localization patterns were consistent with published data derived from cell fractionation or large scale localization approaches. However, in many cases, high resolution microscopy provided additional information that indicated that proteins distributed to multiple subcellular locations. The majority of tagged enzymes localized to the endoplasmic reticulum (91), but others localized to mitochondria (27), peroxisomes (17), lipid droplets (23), and vesicles (53). We assembled enzyme localization patterns for phospholipid, sterol, and sphingolipid biosynthetic pathways and propose a model, based on enzyme localization, for concerted regulation of sterol and sphingolipid metabolism that involves shuttling of key enzymes between endoplasmic reticulum, lipid droplets, vesicles, and Golgi.

YPL.db2: the yeast protein localization database, version 2.0

The Yeast Protein Localization database (YPL.db(2)) is an archive of microscopic image data of protein localization patterns in the yeast Saccharomyces cerevisiae. The current version of YPL.db(2) harbours 500 sets of image data derived from high-resolution microscopic analyses of proteins tagged with the green fluorescent protein (GFP). Major functional improvements in YPL.db(2) over a previous release are a web-based experiment and image submission interface, facilitating standardized data entry by remote users through the Internet. The image display page provides image gallery and image scrolling features. In addition, fluorescence and transmission images can be superimposed, allowing image fading for precise correlation of the protein's localization in the cellular context. The reference structure database displaying 'prototypic' localization patterns was extended, and a feature to display and manipulate 3D-image datasets, using a freely available VRML plug-in, was included. Access to the Yeast Protein Localization database version 2.0 (YPL.db(2)) is available through http://YPL.uni-graz.at.

YPL.db: the Yeast Protein Localization database.

The Yeast Protein Localization database (YPL.db) contains information about the localization patterns of yeast proteins resulting from microscopic analyses. The data and parameters of the experiments to obtain the localization information, together with images from confocal or video microscopy, are stored in a relational database, building an archive of, and the documentation for, all experiments. The database can be queried based on gene name, protein localization, growth conditions and a number of additional parameters. All experiment parameters are selectable from predefined lists to ensure database integrity and conformity across different investigators. The database provides a structure reference resource to allow for better characterization of unknown or ambiguous localization patterns. Links to MIPS, YPD and SGD databases are provided to allow fast access to further information not contained in the localization database itself. YPL.db is available at http://ypl.tugraz.at.