Enhancement of Telomere-Plasmid Segregation by the X-Telomere Associated Sequence in Saccharomyces Cerevisiae Involves Sir2, Sir3, Sir4 and Abf1

We have previously shown that circular replicating plasmids that carry yeast telomere repeat sequence (TG(1-3)) tracts segregate efficiently relative to analogous plasmids lacking the TG(1-3) tract and this efficient segregation is dependent upon RAP1. While a long TG(1-3) tract is sufficient to improve plasmid segregation, the segregation efficiency of telomere plasmids (TEL-plasmids) is enhanced when the X-Telomere Associated Sequence (X-TAS) is also included on the plasmids. We now demonstrate that the enhancement of TEL-plasmid segregation by the X-TAS depends on SIR2, SIR3, SIR4 and ABF1 in trans and requires the Abflp-binding site within the X-TAS. Mutation of the Abflp-binding site within the X-TAS results in TEL-plasmids that are no longer affected by mutations in SIR2, SIR3 or SIR4, despite the fact that other Abflp-binding sites are present on the plasmid. Mutation of the ARS consensus sequence within the X-TAS converts the X-TAS from an enhancer element to a negative element that interferes with TEL-plasmid segregation in a SIR-dependent manner. Thus, telomere associated sequences interact with TG(1-3) tracts on the plasmid, suggesting that the TASs have an active role in modulating telomere function.     

The yeast silent information regulator Sir4p anchors and partitions plasmids.

Circular plasmids containing telomeric TG1-3 arrays or the HMR E silencer segregate efficiently between dividing cells of the yeast Saccharomyces cerevisiae. Subtelomeric X repeats augment the TG1-3 partitioning activity by a process that requires the SIR2, SIR3, and SIR4 genes, which are also required for silencer-based partitioning. Here we show that targeting Sir4p to DNA directly via fusion to the bacterial repressor LexA confers efficient mitotic segregation to otherwise unstable plasmids. The Sir4p partitioning activity resides within a 300-amino-acid region (residues 950 to 1262) which precedes the coiled-coil dimerization motif at the extreme carboxy end of the protein. Using a topology-based assay, we demonstrate that the partitioning domain also retards the axial rotation of LexA operators in vivo. The anchoring and partitioning properties of LexA-Sir4p chimeras persist despite the loss of the endogenous SIR genes, indicating that these functions are intrinsic to Sir4p and not to a complex of Sir factors. In contrast, inactivation of the Sir4p-interacting protein Rap1p reduces partitioning by a LexA-Sir4p fusion. The data are consistent with a model in which the partitioning and anchoring domain of Sir4p (PAD4 domain) attaches to a nuclear component that divides symmetrically between cells at mitosis; DNA linked to Sir4p by LexA serves as a reporter of protein movement in these experiments. We infer that the segregation behavior of telomere- and silencer-based plasmids is, in part, a consequence of these Sir4p-mediated interactions. The assays presented herein illustrate two novel approaches to monitor the intracellular dynamics of nuclear proteins.     

Identification of silencer binding proteins from yeast: possible roles in SIR control and DNA replication

The 'silent' yeast mating-type loci (HML and HMR) are repressed by sequences (HMLE and HMRE) located over 1 kb from their promoters which have properties opposite those of enhancers, and are called 'silencers'. Both silencers contain autonomously replicating sequences (ARS). Silencer activity requires four trans-acting genes called SIR (silent information regulator). We have identified two DNA binding factors , SBF-B and SBF-E, which bind to known regulatory elements at HMRE. SBF-B binds to a region involved in both the silencer and ARS functions of HMRE, but doesn not bind to HMLE. This factor also binds to the unlinked ARS1 element. SBF-E recognizes a sequence found at both silencers. These results suggest that the two silencers may be composed of different combinations of regulatory elements at least one of which is common to both. Neither factor appears to be a SIR gene product. Hence the SIR proteins may not directly interact with the silencer control sites.