The fission yeast ubiquitin-conjugating enzymes UbcP3, Ubc15, and Rhp6 affect transcriptional silencing of the mating-type region

Abstract

Genes transcribed by RNA polymerase II are silenced when introduced near the mat2 or mat3 mating-type loci of the fission yeast Schizosaccharomyces pombe. Silencing is mediated by a number of gene products and cis-acting elements. We report here the finding of novel trans-acting factors identified in a screen for high-copy-number disruptors of silencing. Expression of cDNAs encoding the putative E2 ubiquitin-conjugating enzymes UbcP3, Ubc15 (ubiquitin-conjugating enzyme), or Rhp6 (Rad homolog pombe) from the strong nmt1 promoter derepressed the silent mating-type loci mat2 and mat3 and reporter genes inserted nearby. Deletion of rhp6 slightly derepressed an ade6 reporter gene placed in the mating-type region, whereas disruption of ubcP3 or ubc15 had no obvious effect on silencing. Rhp18 is the S. pombe homolog of Saccharomyces cerevisiae Rad18p, a DNA-binding protein that physically interacts with Rad6p. Rhp18 was not required for the derepression observed when UbcP3, Ubc15, or Rhp6 was overproduced. Overexpressing Rhp6 active-site mutants showed that the ubiquitin-conjugating activity of Rhp6 is essential for disruption of silencing. However, high dosage of UbcP3, Ubc15, or Rhp6 was not suppressed by a mutation in the 26S proteasome, suggesting that loss of silencing is not due to an increased degradation of silencing factors but rather to the posttranslational modification of proteins by ubiquitination. We discuss the implications of these results for the possible modes of action of UbcP3, Ubc15, and Rhp6.

FIG. 1.

Screen for high-copy-number disruptors of silencing. (A) Mating-type region of the tester strain PG1608. (B) Haploid meiosis phenotypes of PG1608 overexpressing ubcP3, rhp6, ubc15, or no cDNA (−). Colonies grown on MSA plus uracil plus adenine were stained with iodine. (C) Silencing of ura4 in transformed PG1608 cells was estimated by plating 10-fold serial dilutions of cell suspensions on medium lacking uracil (AA-thiamine-leucine-uracil [− uracil]) or containing 5-FOA (FOA-thiamine-leucine [+ FOA]) and on complete medium (AA-thiamine-leucine). (D) Multiple-sequence alignment showing the homology between UbcP3, Ubc15, and Rhp6. Black shading indicates identical residues, while gray shading indicates conserved residues.

FIG. 2.

High dosage of UbcP3, Rhp6, or Ubc15 generally affects silencing in the mating-type region. (A) Mating-type region of the switchable (h⁹⁰) strain PG1899. (B) Silencing in PG1899 was assayed by plating serial dilutions of cells transformed with plasmids containing no cDNA (−), ubcP3, ubc15, or rhp6 on the same media as in Fig. 1C. (C) Mating-type region of the nonswitchable strain PG1789. (D) Silencing was assayed in PG1789 by spotting serial dilutions of cells on the same media as in Fig. 1C. (E) Mating-type region of PB141. (F) Silencing of ade6 placed near mat3-M was assayed by plating dilution series of transformed PB141 cells on medium lacking adenine (PM plus uracil) (− adenine) and complete medium (PM plus uracil plus adenine).

FIG. 3.

High dosage of UbcP3, Rhp6, or Ubc15 increases the amount of ura4 transcript originating from the mating-type region. Strains containing the ura4-DS/E allele (1) and a random integration of ura4 (FY340 [R.I.] [1]), the mat3-M (EcoRV)::ura4 allele depicted in Fig. 2A (PG1899 [mat3-M]), the mat2-P (XbaI)::ura4 allele depicted in Fig. 2C (PG1789 [mat2-P]), or the Δ(BglII-BssHII) mat2-P (XbaI)::ura4 allele depicted in Fig. 1A (PG1786 [Δ1.5 kb mat2-P]) were transformed with the indicated cDNAs and used to prepare a Northern blot. “ura4” indicates the full-length ura4 mRNA transcribed from the mating-type region, while “ura4-DS/E” indicates the truncated mRNA transcribed from the ura4-DS/E allele. The ratio of ura4 to ura4-DS/E transcript is indicated below each lane.

FIG. 4.

High dosage of UbcP3, Ubc15, or Rhp6 combined with mutations in cis- and trans-acting silencers. (A) Mating-type region of the switchable strains PG9 (h⁹⁰) and PG595 (h⁹⁰, swi6-115). (B) Sporulation of PG9 (swi6⁺) (+) and PG595 (swi6-115) cells transformed with plasmids containing no cDNA (−), ubcP3, ubc15, or rhp6 was assayed by staining colonies grown on sporulation medium (MSA plus adenine plus uracil) with iodine. (C) Mating-type region of the nonswitchable strains PG1598 (mat1-Msmt-0), PG506 (mat1-Msmt-0, clr1-5), and PG2883 (mat1-Msmt-0, esp1-3). (D) Sporulation of PG1598 (+), PG506 (clr1-5), and PG2883 (esp1-3) transformed with the indicated plasmids was assayed as done for panel B. (E) PG1247 contains an ∼8-kb deletion between mat2-P and mat3-M. (F) Sporulation in PG1247 cells transformed with plasmids containing no cDNA (−), ubcP3, ubc15, or rhp6 was assayed by iodine staining.

FIG. 5.

The Rhp6 active site is required for the antisilencing effect of Rhp6. (A) Mating-type region of PG9. (B) The effects on silencing of overexpressing rhp6, rhp6-C88S, rhp6-C88A, or no cDNA were compared by testing the ability of transformed PG9 strains to grow on complete medium (MSA plus adenine plus uracil), medium lacking uracil (MSA plus adenine) (− uracil), and medium containing 5-FOA (PM plus adenine plus uracil plus FOA) (+ FOA).

FIG. 6.

Deletion of rhp6 slightly derepresses silencing in the mating-type region. Strains containing a random integration of ura4 (R.I.), the mat3-M (EcoRV)::ura4 allele depicted in Fig. 2A (mat3-M), the mat2-P (XbaI)::ura4 allele depicted in Fig. 2C (mat2-P), or the Δ(BglII-BssHII) mat2-P (XbaI)::ura4 allele depicted in Fig. 1A (Δ1.5 kb mat2-P) and a wild-type (+) or null (Δ) allele of rhp6 were used to estimate the amount of ura4 transcript originating from the mating-type region relative to the ura4-DS/E control. ura4-to-ura4-DS/E ratios are reported below the blot. The strains were as follow, from left to right: R.I. +, FY340; R.I. Δ, PG2875; R.I. Δ, PG2876; mat3-M +, PG1899; mat3-M Δ, PG2887; mat3-M Δ, PG2888; mat2-P +, PG1789; mat2-P Δ, PG2889; mat2-P Δ, PG2890; Δ1.5 kb mat2-P +, PG1786; Δ1.5 kb mat2-P Δ, PG2891; and Δ1.5 kb mat2-P Δ, PG2892.

FIG. 7.

Disruption of silencing caused by overexpression of ubcP3, ubc15, or rhp6 does not require rhp18. (A) Mating-type region of Hu52 and PG2895. (B) Silencing of ade6 near mat3-M was tested by plating dilution series of rhp18⁺ (Hu52) or rhp18Δ::ura4 (PG2895) cells transformed with plasmids containing no cDNA (−), ubcP3, ubc15, or rhp6 on medium lacking adenine (AA-thiamine-leucine-adenine) (− adenine - T), on medium poor in adenine (AA-thiamine-leucine + 12 mg of adenine per liter) (Low adenine - T), and on complete medium (AA-acid-thiamine-leucine). (C) A Northern blot prepared with RNA from the strains displayed in panel B was hybridized to a probe for the ade6 open reading frame. “ade6” indicates the full-length ade6 transcript originating from the mating-type region, while “ade6-DN/N” indicates the truncated ade6 transcript originating from the ade6-DN/N allele.

FIG. 8.

The silencing defects caused by overexpression of ubcP3 or rhp6 do not require a functional 26S proteasome. (A) Mating-type region of PI131. (B) PI131 cells overexpressing the indicated proteins were plated on medium lacking uracil (MSA plus adenine) (− uracil) or containing 5-FOA (PM plus adenine plus uranine plus FOA) (+ FOA). The effect of overexpressing ubcP3, ubc15, or rhp6 on the growth of PI131 was tested by plating on noninducing (AA-leucine [complete]) and inducing (MSA plus adenine plus uracil [complete-T]) media. Plate contents were incubated at 30°C.

FIG. 9.

Ubiquitin-conjugating enzymes and silencing: models. (A) Transcriptionally silenced region. Silencing factors, represented by objects of various shapes, are associated with chromatin (bar). (B to E) Models for the disruption of silencing by the overexpression of ubiquitin-conjugating enzymes (Ubc). The models are discussed in the text. Ub, ubiquitin.