Genetic-interaction screens uncover novel biological roles and regulators of transcription factors in fission yeast

Abstract

In Schizosaccharomyces pombe, systematic analyses of single transcription factor deletion or overexpression strains have made substantial advances in determining the biological roles and target genes of transcription factors, yet these characteristics are still relatively unknown for over a quarter of them. Moreover, the comprehensive list of proteins that regulate transcription factors remains incomplete. To further characterize Schizosaccharomyces pombe transcription factors, we performed synthetic sick/lethality and synthetic dosage lethality screens by synthetic genetic array. Examination of 2,672 transcription factor double deletion strains revealed a sick/lethality interaction frequency of 1.72%. Phenotypic analysis of these sick/lethality strains revealed potential cell cycle roles for several poorly characterized transcription factors, including SPBC56F2.05, SPCC320.03, and SPAC3C7.04. In addition, we examined synthetic dosage lethality interactions between 14 transcription factors and a miniarray of 279 deletion strains, observing a synthetic dosage lethality frequency of 4.99%, which consisted of known and novel transcription factor regulators. The miniarray contained deletions of genes that encode primarily posttranslational-modifying enzymes to identify putative upstream regulators of the transcription factor query strains. We discovered that ubiquitin ligase Ubr1 and its E2/E3-interacting protein, Mub1, degrade the glucose-responsive transcriptional repressor Scr1. Loss of ubr1+ or mub1+ increased Scr1 protein expression, which resulted in enhanced repression of flocculation through Scr1. The synthetic dosage lethality screen also captured interactions between Scr1 and 2 of its known repressors, Sds23 and Amk2, each affecting flocculation through Scr1 by influencing its nuclear localization. Our study demonstrates that sick/lethality and synthetic dosage lethality screens can be effective in uncovering novel functions and regulators of Schizosaccharomyces pombe transcription factors.

Keywords: Schizosaccharomyces pombe; synthetic dosage lethality; synthetic lethality; transcription factor.

Fig. 1.

The SGA-based screening protocol for identifying SDL interactions in S. pombe. The 279 deletion array strains were arrayed on a single plate at a colony density of 384. The nmt1-driven overexpression query strain was crossed to the deletion miniarray in step 1. The selection of mated spores in step 2 was similar to the SGA protocol outlined by Dixon et al. (2008), with a 3-day incubation on SPAS plates at 25°C followed by another 3-day incubation at 42°C for mating and selection of spores, respectively. This was followed by a 5-day incubation in step 3 on EMM + AU medium supplemented with thiamin to allow for the spore germination and growth of vegetative cells. In step 4, the deletions were selected for on PMG + AU + G418 with thiamin before induction of the nmt1 promoter. The selection of the deletion mutants and the induction of the nmt1 promoter were performed in steps 5 and 6 to detect putative SDL interactions. PMG + AU + G418 was used to select for both the gene deletion and the plasmid while overexpressing the transcription factor target gene. The final colony size was imaged with the spImager-M system (S&P Robotics, Inc.) and scored using SGAtools (Wagih et al. 2013).

Fig. 2.

A heatmap of the genetic interactions between the 38 query and 91 array transcription factor deletion strains. The interaction scores are mapped to colors as indicated by the color bar at the bottom right, with negative scores in cyan are mostly clustered at the bottom and positive scores in yellow are mostly clutered together at the top right. The light gray squares indicate interactions that were omitted due to the possibility of gene linkage. All screens were performed with 3 biological replicates, and each array strain at 3 different locations on the plate per replicate.

Fig. 3.

Transcription factor double mutants that result in a cell elongation phenotype. Cells were examined by differential interference contrast and fluorescence microscopy using DAPI and calcofluor white. a) The transcription factors Res2 and Tos4 have known roles in the cell cycle and the double deletion mutant cells are elongated more than either single deletion. b) The transcription factor Prz1 has been implicated in cell cycle regulation while SPBC56F2.05c is uncharacterized. The double mutant cells are elongated relative to the wild type and the single mutant controls. c) The transcription factors SPAC3C7.04 and SPCC320.03 are uncharacterized. Their function may be related to cell cycle regulation as the double mutant cells are elongated.

Fig. 4.

Correspondence of reduced fitness of transcription factor overexpression strains detected by robotic pinning and microscope visualization of cells/colony from Vachon et al. (2013). a) The relationship between the manual scores of cells/colony and the fitness scores of the transcription factor overexpression strains from SGA screening with significant differences observed between all 4 categories (P-values are from 2-tailed t-tests). The SGA fitness scores were based on 3 biological replicates. b) The colony sizes on a plate with thiamin after 3 days of growth. c) The colony sizes on a plate without thiamin after 3 days of growth (ectopic expression of the transcription factor gene).

Fig. 5.

The S. pombe genetic interactions from SDL screens of transcription factor overexpression query strains and deletion array strains of posttranslational modifying enzymes. The 121 of 279 genes when deleted that showed a genetic interaction score of either >0.5 or <−0.5 with one of the 14 transcription factor overexpression query strains are shown in the heat map. Each column represents an overexpression query strain and each row represents a deletion array strain. Positive genetic interactions are mostly clustered in the bottom part of the figure and indicated in yellow and negative genetic interactions (SDL) are clustered at the top of the figure and indicated in cyan. The screens were performed with 3 biological replicates.

Fig. 6.

SDL interactions of scr1⁺. a) Confirmation of SDL interactions with scr1⁺ by serial dilution. b) Fluorescence microscopy images of Scr1-GFP under either high- or low-glucose conditions in wild-type, Δubr1, and Δmub1 strains. c) The quantification of the Scr1-GFP total corrected cellular fluorescence in the 3 corresponding strains at the 2 different concentrations of glucose. The Scr1-GFP level in low glucose was significantly higher in the Δubr1 and Δmub1 strains than in wild type (P < 0.0001). The Scr1-GFP level in high glucose was significantly higher in the Δubr1 strain than in wild type (P = 0.0072). The total corrected cellular fluorescence values were calculated as described (McCloy et al. 2014) and represent 30 cells over 3 biological replicates. The error bars represent the standard error of the mean. d) Fluorescence and light microscopy of cells expressing Scr1-GFP under control of its native promoter in FIM for 6 h. Scr1-GFP is excluded from the nucleus in wild-type cells but shows nuclear localization in Δsds23 or Δamk2 cells. Scr1-GFP is upregulated in Δubr1 or Δmub1 cells. All fluorescent cell images were acquired using the same exposure time. Bar = 10 µm.

Fig. 7.

Flocculation assays of scr1⁺ and SDL strains. a) Scr1 is a repressor of flocculation. The Δscr1 strain shows constitutive flocculation in liquid EMM while overexpression of scr1⁺ with the nmt1 promoter prevents flocculation in FIM. b) Sds23 is an activator of flocculation. The Δsds23 strain fails to flocculate in FIM while overexpression of sds23⁺ with the nmt1 promoter induces flocculation in liquid EMM. The nonflocculent phenotype of the Δsds23 strain is unable to suppress the constitutive flocculation of Δscr1 cells. c) The SDL strains Δubr1, Δmub1, and Δamk2 fail to flocculate in FIM. Strains were grown in liquid EMM and FIM at 30°C for 2 and 5 days, respectively. Five milliliters of cell cultures were transferred to Petri dishes, placed on an orbital shaker for 10 min to promote floc formation and imaged with a spImager-M system (S&P Robotics, Inc.).