Saccharomyces cerevisiae Elongator mutations confer resistance to the Kluyveromyces lactis zymocin

Abstract

Kluyveromyces lactis killer strains secrete a zymocin complex that inhibits proliferation of sensitive yeast genera including Saccharomyces cerevisiae. In search of the putative toxin target (TOT), we used mTn3:: tagging to isolate zymocin-resistant tot mutants from budding yeast. Of these we identified the TOT1, TOT2 and TOT3 genes (isoallelic with ELP1, ELP2 and ELP3, respectively) coding for the histone acetyltransferase (HAT)-associated Elongator complex of RNA polymerase II holoenzyme. Other than the typical elp ts-phenotype, tot phenocopies hypersensitivity towards caffeine and Calcofluor White as well as slow growth and a G(1) cell cycle delay. In addition, TOT4 and TOT5 (isoallelic with KTI12 and IKI1, respectively) code for components that associate with ELONGATOR: Intriguingly, strains lacking non-Elongator HATs (gcn5, hat1, hpa3 and sas3) or non-Elongator transcription elongation factors TFIIS (dst1) and Spt4p (spt4) cannot confer resistance towards the K.lactis zymocin, thus providing evidence that Elongator equals TOT and that Elongator plays an important role in signalling toxicity of the K.lactis zymocin.

Fig. 1. The totΔ mutants are resistant towards endogenously expressed γ-toxin and exogenously applied K.lactis zymocin. (A) γ-toxin assay. Serial dilutions of toxin-sensitive (wt) and -resistant (tot1–5Δ) cells transformed with the GAL1-driven γ-toxin vector pHMS14 were replica spotted on repressing (glucose) and inducing (galactose) rich medium and grown for 2 days at 30°C. Lack of growth indicates γ-toxin sensitivity. (B) Killer eclipse assay. Wild-type, tot1–5Δ strains, dst1Δ and spt4Δ cells were spotted twice on to YPD medium. A K.lactis strain (upper row: killer strain AWJ137; lower row: non-killer strain NK40) was set on to the edge of these spots and incubated for 2 days at 30°C. Zones of inhibition around the zymocin-secreting killer strain (upper row) indicate zymocin sensitivity; lack of growth inhibition equals zymocin resistance.

Fig. 2. Deletion of TOT1–5 genes confers slow growth and thermosensitive phenotypes. (A) Growth curve of tot1–5Δ strains. tot1–5Δ cells show a slow growth phenotype and do not reach the biomass of the wild-type reference in stationary phase. Strains were grown in YPD medium (2% glucose) and growth was measured at OD₆₀₀ over a period of 16 h. (B) Ts phenotype of tot1–5Δ strains. Strains were streaked on YPD and incubated for 3 days at 30 and 39°C.

Fig. 3. The slow growth phenotype of tot1–5Δ strains correlates with a significant delay in the G₁ phase of the cell cycle. FACS analyses of exponential growing tot1–5Δ cells show an extended 1n peak in comparison with the wild-type strain, indicating a delay in the G₁ to S transition.

Fig. 4. More tot phenotypes. (A) Calcofluor White sensitivity. Serial dilutions of strains were replica spotted on YPD plates containing no or 50 µg of Calcofluor White (CW). All totΔ strains show hypersensitivity against Calcofluor White, whereas the positive reference, chs3Δ, displays resistance towards the drug. (B) Caffeine sensitivity. Strains were spotted on YPD plates containing up to 10 mM caffeine and grown for 3 days at 30°C. All totΔ strains show more or less hypersensitivity towards caffeine. (C) 6-AU phenotype. Strains were spotted on SD plates containing no 6-AU, 50 µg/ml 6-AU or50 µg/ml 6-AU plus uracil. Except for the tot2Δ mutant, which is mildly affected by the drug, all other totΔ strains show hypersensi tivity towards 6-AU. The dst1Δ mutant served as a positive 6-AU-hypersensitive control strain.

Fig. 5. Effect of HAT gene deletions on zymocin sensitivity. Strains deleted in the HAT-encoding genes SAS3, HPA3, HAT1, HPA1/ELP3, GCN5 and TOT3/ELP3 were subjected to a killer eclipse assay essentially as described in Figure 1. Deletion of TOT3 confers zymocin resistance, whereas the other HAT gene deletions tested are zymocin sensitive.

Fig. 6. The HAT activity of Elp3p/Tot3p is essential for γ-toxin action. (A) Wild-type and tot3Δ strains containing the GAL1-driven γ-toxin vector pHMS14 were transformed with CEN plasmids carrying no insert (pRS316), the wild-type ELP3 allele (pWT) or mutated elp3 alleles (pY540A and pY541A). Serial dilutions of the resulting transformants were replica spotted on glucose-repressing and galactose-inducing medium. The mutant elp3 alleles fail to complement the tot3Δ-associated γ-toxin resistance phenotype. In the wild-type background, the Y540A allele slightly suppresses γ-toxin sensitivity. (B) When overexpressed from yeast multicopy plasmids, both the mutant elp3 alleles (pLF31: Y540A and pLF30: Y541A) suppress the γ-toxin sensitivity of the wild-type strain co-maintaining pHMS14, whereas vector without insert (YEplac181) and the wild-type ELP3 allele (pLF32) are not able to do so.

Fig. 7. Tot protein detection and co-immunoprecipitation. (A) Identification of TOT1–5 as protein-encoding structural genes. Western analysis: Tot1–5p were tagged individually by one-step in vivo tagging with a triple c-Myc tag. Total protein extracts (50 µg) were separated using 10% SDS–PAGE, electroblotted and immunoprobed with the anti-c-Myc antibody 9E10. The position of each tagged Tot protein is indicated by an arrow. (B) Tot3p interacts with Tot1p and Tot2p. Protein extracts from strains expressing c-Myc-tagged Tot3p together with HA-tagged Tot1p or HA-tagged Tot2p were immunoprecipitated using the anti-c-Myc antibody 9E10. Precipitates were then separated using 10% SDS–PAGE, electroblotted and immunoprobed with the anti-HA antibody 12CA5. The positions of HA-tagged Tot1p and Tot2p are indicated by arrows. (C and D) Tot4p (C) and Tot5p (D) interact with Elp1p (Tot1p), Elp2p (Tot2p) and Elp3p (Tot3p). Protein extracts from strains expressing c-Myc-tagged Tot4p (C) or Tot5p (D) were immunoprecipitated using the anti-c-Myc antibody 9E10, separated on a 10% SDS–PAGE gel, electroblotted and immunoprobed by western analysis using polyclonal anti-Elp1p, anti-Elp2p and anti-Elp3p antibodies. The positions of co-immuno precipitated proteins, Elp1p, Elp2p and Elp3p, are indicated by arrows.

Fig. 8. Gene transcription in response to the K.lactis zymocin. (A) Identical amounts of total RNA isolated from untreated and zymocin-arrested cells were subjected to RT–PCR experiments to study the transcription of RNA polymerase I-dependent (RDN18) and RNA polymerase II-dependent (SIC1, HHT1, ACT1 and CLN3) genes by 1% TBE–agarose gel electrophoresis. (B) Identical amounts of total RNA (10 µg) isolated from untreated cells (lane 1) and cells arrested by zymocin for 3 h (lane 2) and 6 h (lane 3) were subjected to northern blot analysis using probes specific for the RDN18, SIC1, HHT1, ACT1 and CLN3 genes.