Going in the right direction: mating-type switching of Schizosaccharomyces pombe is controlled by judicious expression of two different swi2 transcripts

Abstract

Schizosaccharomyces pombe, the fission yeast, cells alternate between P- and M-mating type, controlled by the alternate alleles of the mating-type locus (mat1). The mat1 switching occurs by replacing mat1 with a copy derived from a silenced "donor locus," mat2P or mat3M. The mechanism of donor choice ensuring that switching occurs primarily and productively to the opposite type, called directionality, is largely unknown. Here we identified the mat1-Mc gene, a mammalian sex-determination gene (SRY) homolog, as the primary gene that dictates directionality in M cells. A previously unrecognized, shorter swi2 mRNA, a truncated form of the swi2, was identified, and its expression requires the mat1-Mc function. We also found that the abp1 gene (human CENPB homolog) controls directionality through swi2 regulation. In addition, we implicated a cis-acting DNA sequence in mat2 utilization. Overall, we showed that switching directionality is controlled by judicious expression of two swi2 transcripts through a cell-type-regulated dual promoter. In this respect, this regulation mechanism resembles that of the Drosophila sex-determination Slx gene.

Figure 1

The directionality phenomenon of mating-type switching in fission yeast. The diagram shows mat1, mat2P, and mat3M genes located from left to right in chromosome 2. The mat1 locus is expressed and it confers cell type, while mat2P and mat3M, located in the silenced region, act only as donors for providing copies of genetic information for mat1 switching. The mat1 can be either P (black zigzag line, representing 1104-bp-long DNA sequence) or M (shaded bar, 1128 bp). Each mat cassette is flanked by the homology regions, boxes H2 (left open box, 135 bp) and H1 (right solid box, 59 bp), which are used for switching-promoted recombination. The imprint site (solid triangle) is located at the junction of the mat1 allele-specific and the H1 box sequences. The donor preference is determined by the mat1 cell type; mat1-P cells choose mat3 and mat1-M cells choose mat2, both with ∼90% preference (thick arrow) over the other, less-preferred (∼10%, thin arrow) donor locus. The HindIII restriction enzyme sites (H) flanking each cassette are shown. Digestion of yeast genomic DNA with the restriction enzyme generates three fragments of indicated sizes, each containing a specific mat locus. The figure is not drawn to scale.

Figure 2

Effect of mat1 mutations on mat1 imprint and on directionality of switching. (A) Imprint level determined by Southern blot analysis. The wild-type (wt, SP976; see Table 2 for complete genotype) was the homothallic h⁹⁰ strain, which contains all three cassettes diagrammed in Figure 1. The stocks with mating-type gene mutations in pc⁻ (SP718), pi⁻ (SP717), mi⁻ (SP716), mc⁻ (SP715) had been deleted for both donor loci; thus bands reflecting mat2P and mat3M genes were lacking in them. Such strains lacking donor loci cannot switch, and therefore stably maintain their mat1 mutations. Genomic DNA was digested with HindIII and the blot was probed with mat1P gene-containing 10.6 kb long probe (Figure 1). Locations of the HindIII sites flanking each cassette are shown in Figure 1. The 5.6-kb and 5.0-kb bands result from the double-strand break (DSB) generated by shearing imprinted DNA during its preparation. The DSB reflects the level of mat1 imprint (Beach and Klar 1984). (B) Effect of mat1 genes mutations on directionality of h⁹⁰ (wt, SP976; mc⁻, CY195; mi⁻, CY200; pi⁻, CY204) and h⁰⁹ (wt, PG19; mc⁻, CY196; mi⁻, CY202; pi⁻, CY207; pc⁻, CY199) strains. Quantitative multiplex PCR analysis of cultures showed M/P ratio of the mat1 locus of strains in which mutations introduced in donor loci have been transposed to mat1 through switching. The PCR reaction included one primer from outside the mat1 allele-specific sequences and another one from the mat1-P or mat1-M allele-specific sequences (see Materials and Methods ). The mat1 allele-specific PCR product bands are indicated. (C) Effect of Δste11 (h⁹⁰, CY146; h⁰⁹, CY148) on directionality. The mat1 M/P ratio of cultures was determined as described above.

Figure 3

Characterization of the swi2L and swi2S products. (A) The swi2 gene, mRNAs, and predicted proteins. Upper drawing shows the swi2 gene’s DNA structure indicated along with its two translation initiation ATG codons. The Northern blot probes, Swi2 5′P, Swi2P, we used are also indicated. The two lower drawings indicate mRNA and ORF length of swi2L and swi2S forms of the swi2 gene. Both swi2L and swi2S predicted proteins share the same open reading frame (ORF) in regions where the genes overlap. The Swi2 protein domains, such as for other proteins interaction, self-interaction, and two AT-hook motifs, are marked (Akamatsu et al. 2003). The Swi2S lacks the self-interaction domain. The GenBank accession nos. for swi2L and swi2S are JQ308182 and JQ308183. (B) Northern blot of swi2 transcripts. The cultures with indicated genotype were grown in PMA media (nitrogen-starvation condition) and YEA media (nitrogen-rich condition). Strains: P, SP713; M, SP714; wt, SP976; mc⁻ in h⁹⁰, CY195; mc⁻ in Δdonors, SP715; Δabp1, spA160. (C) Rapid amplification of swi2 cDNA ends (RACE). Results show agarose gel electrophoresis of RACE products obtained by amplifying 5′- and 3′-end products of swi2 cDNA. (D) Western blot analysis of the Swi2L and Swi2S proteins. Analysis of the swi2HA tagged gene from the endogenous locus in donors-deleted P (CY347) and M (CY348) cells is presented in the left two lanes. The remaining lanes reflect analysis of swi2L-HA and swi2S-HA tagged genes that were expressed from a plasmid vector in h⁹⁰ cells (SP976). The anti-HA antibody was used for Swi2 detection. The h⁹⁰ wild-type culture harboring empty pREP3 vector was used as a negative signal detection control.

Figure 4

Effect of Δswi2L on mating-type switching. (A) Structure of the swi2L::ura4 allele is diagrammed. A 1.8-kb ura4 HindIII fragment was inserted in the translation initiation site (ATG) of the swi2L gene. (B) Northern blot analysis of wild-type (SP976) and the swi2L::ura4 (CY345) cultures. RNAs prepared from cultures grown in PMA medium were analyzed with the Swi2P probe (Figure 3A). (C) Iodine-staining colony phenotype. Numbers below each colony reflect their efficiency of sporulation. (D) The mat1 M/P ratio of cultures was determined with the method described in Figure 2B.

Figure 5

Ectopically expressed swi2S suppresses mc⁻ (CY195) and Δabp1 (CY327) mutants’ switching defect. (A) The mat1 M/P ratio of cultures was determined with the method described in Figure 2B. Analysis of an empty pREP3 vector transformant was included for comparison, as a negative control (wt, SP976). (B) Iodine-staining colony phenotype. Moderate swi2S gene expression was achieved by 0.05 μM thiamine addition to the medium and thiamine was not added to achieve overexpression.

Figure 6

Effect of ΔSRE (switching recombinational enhancer) deletion on mating-type switching. (A) The diagram shows location of SRE1 and SRE2 elements with respect to donor loci. ΔSRE2 (CY212) comprises a 400-bp deletion located distal to the H1 region of mat2P. The SRE1 is defined by a 454-bp deletion of sequences (ΔSRE1, CY294) located distal to H1 region of mat3M (Jia et al. 2004). The iodine staining colony phenotype (B) and the mat1 M/P ratio (C) of indicated strains are presented.

Figure 7

The mechanism directionality of switching. (A) Summary of genetic determinants controlling directionality of mat1 switching. In P cells, because Mc factor is absent, only swi2L form is expressed ON. In M cells, the Mc factor, in combination with the Abp1 factor, activates the swi2S gene. (B) Proposed role of Swi2L and Swi2S in directionality of switching. In P cells, the Swi2L protein promotes localization of the Swi2/Swi5 complex to the mat3 region to make it as the preferred donor. In M cells, perhaps Swi2S and Swi2L cooperate to promote assembly of the Swi2/Swi5 complex across the entire silenced region to make mat2 as the preferred donor. The two SRE elements might facilitate enrichment of Swi2/Swi5 complexes at donor loci in cell-type specific fashion.