The fission yeast HIRA proteins Hip1 and Slm9 are members of an evolutionarily conserved family of histone chaperones that are implicated in nucleosome assembly. Here we have used single-step affinity purification and mass spectrometry to identify factors that interact with both Hip1 and Slm9. This analysis identified Hip3, a previously uncharacterized 187-kDa protein, with similarity to S. cerevisiae Hir3. Consistent with this, cells disrupted for hip3 ؉ exhibit a range of growth defects that are similar to those associated with loss of Hip1 and Slm9. These include temperature sensitivity, a cell cycle delay, and synthetic lethality with cdc25-22. Furthermore, genetic analysis also indicates that disruption of hip3 ؉ is epistatic with mutation of hip1 ؉ and slm9 ؉ . Mutation of hip3 ؉ alleviates transcriptional silencing at several heterochromatic loci, including in the outer (otr) centromeric repeats, indicating that Hip3 is required for the integrity of pericentric heterochromatin. As a result, loss of Hip3 function leads to high levels of minichromosome loss and an increased frequency of lagging chromosomes during mitosis. Importantly, the function of Hip1, Slm9, and Hip3 is not restricted to constitutive heterochromatic loci, since these proteins also repress the expression of a number of genes, including the Tf2 retrotransposons.Centromeres play a critical role in the precise segregation of chromosomes, and as a result, defects in centromere function lead to aneuploidy (1). The fission yeast Schizosaccharomyces pombe provides an excellent system for the study of centromeres (2). In contrast to the budding yeast Saccharomyces cerevisiae, which has simple "point" centromeres (3), S. pombe has large complex centromeres that occupy between 35 and 110 kb and are arranged as a central core (cnt) flanked by arrays of repeated (imr and otr) elements (2). In this respect, S. pombe centromeres are reminiscent of the complex regional centromeres of metazoans. Furthermore, ultrastructural studies have revealed that the overall architectural organization of fission yeast centromeres is conserved with their human counterparts (4). Fission yeast centromeres are organized into distinct chromatin domains (2, 5). An inner domain is assembled into specialized chromatin in which core histone H3 is replaced by Cnp1, the fission yeast homologue of CENP-A (6), whereas the outer regions are associated with chromatin that resembles the pericentric heterochromatin of higher cells (2). Marker genes inserted into these outer regions are subject to heritable inactivation (7,8), which is dependent upon the RNA interference machinery, the methylation of histone H3 on lysine 9, and the association of a number of proteins, including Swi6, a homologue of mammalian HP1 (heterochromatin protein 1) (2, 5, 9 -12). In addition, the integrity of pericentric heterochromatin in fission yeast is dependent upon the two HIRA proteins Hip1 and Slm9 (13,14), since loss of either protein alleviates silencing in the otr centromeric repeats and results in...

SummaryTranscriptional control is exerted by the antagonistic activities of activator and repressor proteins. In Saccharomyces cerevisiae, transcription factor complexes containing the MADS box protein Mcm1p are key regulators of cell cycle-dependent transcription at both the G2/M and M/G1 transitions. The homeodomain repressor protein Yox1p acts in a complex with Mcm1p to control the timing of gene expression. Here, we show that Yox1p interacts with Mcm1p through a motif located N terminally to its homeodomain. Yox1p functions as a transcriptional repressor by competing with the forkhead transcription activator protein Fkh2p for binding to Mcm1p through protein-protein interactions at promoters of a subset of Mcm1p-regulated genes. Importantly, this competition is not through binding the same DNA site that is commonly observed. Thus, this study describes a different mechanism for determining the timing of cell cycle-dependent gene expression that involves competition between short peptide motifs in repressor and activator proteins for interaction with a common binding partner.

No abstract