Two-hybrid search for proteins that interact with Sad1 and Kms1, two membrane-bound components of the spindle pole body in fission yeast

In interphase cells of fission yeast, the spindle pole body (SPB) is thought to be connected with chromosomal centromeres by an as yet unknown mechanism that spans the nuclear membrane. To elucidate this mechanism, we performed two-hybrid screens for proteins that interact with Kms1 and Sad1, which are constitutive membrane-bound components of the SPB that interact with each other. Seven and 26 genes were identified whose products potentially interact with Kms1 and Sad1, respectively. With the exception of Dlc1 (a homolog of the 14-kDa dynein light chain), all of the Kms1 interactors also interacted with Sad1. Among the genes identified were the previously known genes rhp9 ⁺/ crb2 ⁺, cut6 ⁺, ags1 ⁺/ mok1 ⁺, gst3 ⁺, kms2 ⁺, and sid4 ⁺. The products of kms2 ⁺ and sid4 ⁺ localize to the SPB. The novel genes were characterized by constructing disruption mutations and by localization of the gene products. Two of them, putative homologues of budding yeast UFE1 (which encodes a t-SNARE) and SFH1 (an essential component of a chromatin-remodeling complex), were essential for viability. Two further genes, which were only conditionally essential, genetically interact with sad1 ⁺ . One of these was named sif1 ⁺ (for Sad1-interacting factor) and is required for proper septum formation at high temperature. Cells in which this gene was overexpressed displayed a wee -like phenotype. The product of the other gene, apm1 ⁺, is very similar to the medium chain of an adaptor protein complex in clathrin-coated vesicles. Apm1 appears to be required for SPB separation and spindle formation, and tended to accumulate at the SPB when it was overproduced. It was functionally distinct from its homologues Apm2 and Apm4. Other novel genes identified in this study included one for a nucleoporin and genes encoding novel membrane-bound proteins that were genetically related to Sad1. We found that none of the newly identified genes tested were necessary for centromere/telomere clustering.Communicated by C. P. HollenbergThe first three authors contributed equally to this work