S. pombe KO library viewer

Proteome-wide effects of single gene perturbations in a eukaryotic cell

Merve Öztürk¹, Anja Freiwald¹, Jasmin Cartano¹, Ramona Schmitt¹, Mario Dejung¹, Katja Luck¹, Sigurd Braun², Michal Levin¹ and Falk Butter¹

¹Institute of Molecular Biology (IMB), 55122 Mainz, Germany
²Department of Physiological Chemistry, Biomedical Center, Ludwig-Maximilians University of Munich, Planegg-Martinsried, Germany

Eukaryotic gene expression is controlled at the transcriptional, translational and protein degradation level. While transcriptional outcomes, which are commonly also used as a proxy for protein abundance, have been investigated on a larger scale, the study of translational output requires large-scale proteomics data. We here determined the individual proteome changes for 3,308 non-essential genes in the yeast S. pombe. By similarity clustering of proteome changes, we infer gene functionality that can be extended to other species such as human or baker’s yeast. We observed that genes with high proteome remodeling are predominantly involved in gene expression regulation, in particular acting as translational regulators. Focusing on the knockout strains with a large number of altered proteins, we performed paired transcriptome/proteome measurements to uncover translational regulators and features of translational regulation.

Five data types can be explored on this website:

1. The effect of the knockdown of single genes on the protein levels of other genes - LINK
2. The levels of single proteins across different knockout strains - LINK
3. Clusters of knockout strains with similar effects on protein levels of target genes - LINK
4. Clusters of genes with similar effects upon knockdown of single genes - LINK
5. Correlations between changes on the mRNA and protein level of target genes upon knockdown of single genes - LINK

Data acquisition: Proteomics screen with thousands of single gene deletion strains

To systematically investigate the effect of individual gene deletions on the proteome, we used the S. pombe haploid knockout collection with 3,308 individually deleted genes [Kim et al., Nat Biotechnol. 2010]. We took advantage of the 96-well array format of the deletion library (Bioneer, version 3.0) by combining eight knockout strains with 2 different controls per mass spectrometry (MS) measurement (Figure 1a) and assessed the proteome of the 3,308 knockout strains by quantitative proteomics using 10plex TMT (tandem mass tag) [Thompson et al., Anal Chem. 2002] amounting to 469 MS runs (panel a). One of the controls served as technical control required for run-to-run normalization and was generated by growing a large batch of S. pombe wild-type cells. The other biological controls were S. pombe wild-type replicates, grown alongside the knockout strains to be able to judge biological protein expression level variation in wild-type cells (see publication for comprehensive information on normalization approach).

Overall we quantified 2,921 proteins (56 % of protein-coding genes), with a mean of 1,596 proteins per strain, ranging from 1,369 to 1,996 proteins (panel b). Of the 2,921 quantified proteins, 985 (34%) could be quantified in at least 90% of the knockout strains and 1,548 (53%) in at least 50% of the knockout strains.

Proteomics screen to study proteome expression in a knockout library. a, Schematics of the experimental design for quantifying protein expression levels in the S. pombe knockout collection by quantitative proteomics using 10plex TMT. Each MS run contained a technical (127C label) and a biological control (130N label). b, Normalized protein expression levels of 3,308 knockout strains and 461 wild-type replicates. The complete dataset contains quantification for 2,921 protein-coding genes of which 513 proteins were quantified across all samples.

An explanation on how to use our web interface can be downloaded here: "How To"

Data access:

Proteome data has been uploaded to proteomeXchange with the dataset identifiers PXD024332 (genome-wide screen) and PXD024383 (94 gene set). Transcriptomics data has been submitted to GEO with the dataset identifier GSE167543 (94 gene set).