Biochemical interactions between proteins and mat1 cis-acting sequences required for imprinting in fission yeast

Abstract

DNA recombination required for mating type (mat1) switching in Schizosaccharomyces pombe is initiated by mat1 imprinting. The imprinting event is regulated by mat1 cis-acting elements and by several trans-acting factors, including swi1 (for switch), swi3, swi7, and sap1. swi1 and swi3 were previously shown to function in dictating unidirectional mat1 DNA replication by controlling replication fork movement around the mat1 region and, second, by pausing fork progression around the imprint site. With biochemical studies, we investigated whether the trans-acting factors function indirectly or directly by binding to the mat1 cis-acting sequences. First, we report the identification and DNA sequence of the swi3 gene. swi3 is not essential for viability, and, like the other factors, it exerts a stimulatory effect on imprinting. Second, we showed that only Swi1p and Swi3p interact to form a multiprotein complex and that complex formation did not require their binding to a DNA region defined by the smt-0 mutation. Third, we found that the Swi1p-Swi3p complex physically binds to a region around the imprint site where pausing of replication occurs. Fourth, the protein complex also interacted with the mat1-proximal polar terminator of replication (RTS1). These results suggest that the stimulatory effect of swi1 and swi3 on switching and imprinting occurs through interaction of the Swi1p-Swi3p complex with the mat1 regions.

FIG. 1.

DNA sequence of the swi3 gene and mapping of mutations in three mutant alleles. (A) Nucleotide sequence of the swi3 gene. The exons and 5′- and 3′-flanking sequences were determined from a cDNA clone, and the intron sequence was determined from a genomic DNA clone. The exons and the intron are indicated in capital letters, and the flanking regions are indicated in lowercase letters. The deduced amino acid (aa) sequence is shown below the nucleotide sequence. The intron sequence is underlined and indicated in italics. The asterisk indicates the stop codon. (B) The nucleotide changes of three different mutations are indicated, with their positions shown in parentheses. The TIPIN-homologous region is indicated by darker shading. Predicted translated products from the mutant and wild-type ORFs are diagrammed as black bars. The effect of the alleles on sporulation is indicated by the level of sporulation in comparison with that of the wild-type control. The genomic DNA in the pRC14 plasmid, which encodes the N-terminal 98 codons of the swi3 ORF, fully complements the sporulation defect of all three mutant alleles.

FIG. 2.

Δswi3 mutation reduces mating type switching. (A) The swi3 gene was disrupted by inserting a 1.6-kb kanMX6 cassette between nucleotides +58 and +514 of the ORF; +1 and +546 denote the first and last nucleotides of the swi3 ORF, respectively. (B) Comparison of the iodine-staining phenotype of homothallic wild-type swi3 (SP982), the point mutant swi3-146 (E146), the Δswi3::kanMX6 deletion mutant (BSP61), and a nonswitching Msmt-0 mutant (SP903). Efficiency of mating type switching was monitored by iodine staining of colonies sporulated on PMA medium.

FIG. 3.

Swi1p interacts with Swi3p. (A) Detection of Swi1p-GST in the immunoprecipitate (IP) of Swi3p-myc. About 25 μg of total protein extracts loaded in each lane from strains SP976 (swi1 swi3, lane 1), SP982 (swi1 swi3, lane 2), BSP9 (swi1-GST swi3, lane 3), and BSP33 (swi1-GST swi3-myc, lane 4) were fractionated by SDS-8% PAGE. The protein G-agarose bead pellet from each immunoprecipitation was added with SDS sample buffer and boiled. After spinning down the beads, the same volume of supernatant of SP976 (lane 5), SP982 (lane 6), BSP9 (lane 7), and BSP33 (lane 8) was loaded and fractionated on the same gel. The proteins were transferred to an Immobilon P membrane and incubated with anti-GST antibody. We note that the second band below Swi1p-GST in lane 8 was not consistently seen in other experiments and could be a degradation product of the fusion protein. Sizes are shown to the left (in kilodaltons). (B) Detection of Swi3p-myc in the immunoprecipitate of Swi1p-GST. About 25 μg of total protein extracts from strains SP976 (lane 1), SP982 (lane 2), BSP16 (swi1 swi3-myc, lane 3), and BSP33 (lane 4) were fractionated by SDS-8% PAGE; immunoprecipitated complexes from each strain were recovered from the protein A-agarose bead pellet, loaded in lanes 5, 6, 7, and 8, respectively, and analyzed as in panel A but with anti-c-myc antibody. Swi3p-myc was detected by Western analysis as two different bands in the total extract lanes, but it was consistently observed that only the upper band was coimmunoprecipitated with Swi1p-GST. Another band below Swi3p-myc in the Swi1p-IP lanes reflects an unrelated cross-reaction of the antibody with immunoglobulin heavy chains from the polyclonal anti-GST antibody. Sizes are shown to the left (in kilodaltons).

FIG. 4.

Separate forms of Swi1p-Swi3p, Swi7p, and Sap1p in vivo. (A) Analysis of Swi7p in the Swi3p-myc immunoprecipitate (IP). About 25 μg of total protein extracts from BSP9 (swi1-GST swi3, lanes 1 and 3) and BSP33 (swi1-GST swi3-myc, lanes 2 and 4) were loaded in each lane. Immunoprecipitates with anti-c-myc antibody from BSP9 (lane 5) and BSP33 (lane 6) were loaded on the same gel. After blotting the gel onto an Immobilon P membrane, the proteins in lanes 1 and 2 were incubated with preimmune IgY as a negative control, and the proteins in lanes 3 to 6 were incubated with anti-polymerase α (Swi7p) antibody (upper panel). The lower panel is a control for the Swi3p-myc immunoprecipitation. After stripping off previous primary and secondary antibodies, the same blot was reprobed with anti-c-myc antibody to show the presence of immunoprecipitated Swi3p-myc in specific lanes (lower panel). (B) Analysis of Sap1p-myc in the Swi1p-GST immunoprecipitate. Total protein extracts from SP982 (swi1 sap1), BSP34 (swi1 sap1-myc), and BSP60 (swi1-GST sap1-myc) were loaded in lanes 1, 2, and 3, respectively, and immunoprecipitates of Swi1p-GST from BSP34 and BSP60 were loaded in lanes 4 and 5, respectively. The Western blot was probed with anti-c-myc antibody (upper panel) and then reprobed with anti-GST antibody later (lower panel) as described for panel A. (C) Analysis of Swi7p in the Sap1p-myc immunoprecipitate. Total protein extracts from SP982 and BSP34 were loaded in lanes 1 and 3 and lanes 2 and 4, respectively, and immunoprecipitates of Sap1p-myc from SP982 and BSP34 were loaded in lanes 7 and 8, respectively. Lanes 5 and 6 contained mock precipitates with SP982 and BSP34 lysates, respectively. In Western analysis, lanes 1 and 2 were incubated with preimmune IgY, and lanes 3 to 8 were incubated with anti-polymerase α (Swi7p) antibody (upper panel). The blot was reprobed with anti-c-myc antibody as a control for Sap1p-myc immunoprecipitation (lower panel). We note that Swi7p was detected as multiple bands by Western analysis, which was also observed in other studies (1, 41).

FIG. 5.

Interaction of Swi1p with Swi3p in the smt-0 mutant. A schematic representation of the 263-bp deletion of the smt-0 mutation is shown at the top. The deletion includes both SAS1 and SAS2 sequences, shown as black boxes. Immunoprecipitation (IP) was performed as described for Fig. 3 and 4. (A) Total protein extracts from SP982 (swi1 swi3) and BSP50 (swi1-GST swi3-myc) were loaded in lanes 1 and 2, respectively. With the total protein extract of BSP50, a mock precipitate of Swi1p-GST with protein A-agarose beads (lane 3) and the immunoprecipitate of the protein with anti-GST antibody (lane 4) were analyzed for Swi3p-myc by Western analysis. (B) Total protein extracts from SP982 and BSP50 were loaded in lanes 1 and 2, respectively. A mock precipitate of Swi3p-myc with protein G-agarose beads and the immunoprecipitate of Swi3p-myc with anti-c-myc antibody were loaded in lanes 3 and 4, respectively. The gel was subjected to Western analysis to detect Swi1p-GST.

FIG. 6.

Localization of Swi1p and Swi3p to the mat1 locus. (A) A schematic diagram of mat1 and its MPS and RTS1 regions amplified by PCR is shown. Primer sets for PCR (see Materials and Methods) are represented by filled-in arrows and were used to amplify 234-bp and 473-bp fragments of MPS and RTS1, respectively. (B) Chromatin immunoprecipitation analysis. Related SP976 (untagged swi1) and BSP4 (swi1-myc) strains were used to determine the localization of Swi1p-myc to the MPS and RTS1 regions. (C) Chromatin immunoprecipitation analysis of related strains SP982 (untagged swi3) and BSP16 (swi3-myc) was used to determine the localization of Swi3p-myc to the MPS and RTS1 regions. DNA recovered from each chromatin precipitate was analyzed by hot PCR with the indicated primer sets (see panel A) and deoxynucleoside triphosphates, including [α-³²p]dCTP. Inp (input) denotes DNA extracted from the total cell extract that was subjected to immunoprecipitation experiments. It was loaded after manyfold dilution; −, mock precipitation without antibody; +, precipitation with anti-c-myc antibody. The lys11 and trp5 primer sets (see Materials and Methods) were used to amplify fragments of 345 bp for lys11 and 370 bp for trp5 for nonspecific DNA binding controls in both panels B and C.