No abstract

No abstract

Centromeres and several promoters of Saccharomyces cerevisiae contain a highly conserved octanucleotide, RTCACRTG, called CDEI. Using biochemical, genetic and structural analyses, we show that the same protein binds in vivo to CDEI sites in centromeres and in promoters. This protein, called CPF1 for centromere promoter factor, binds DNA as a dimer. Inactivation of the gene is not lethal but leads to a partial loss of the centromere function and to a Met‐ phenotype. Changes of the chromatin structure due to inactivation of CPF1 are seen at centromeres and at several CDEI‐carrying promoters (e.g. MET25, TRP1, GAL2). However promoter activities are affected in diverse ways making it presently difficult to describe a function for CPF1 in gene expression. The sequence of the cloned gene reveals in the carboxy‐terminal part two potential amphipathic helices preceded by a positively charged stretch of amino acids very similar to the helix‐loop‐helix domains recently identified in factors controlling tissue specific transcription in higher eukaryotes. Carboxy‐terminal truncations of CPF1 lacking this domain no longer bind to CDEI. The amino‐terminal half of CPF1 carries two clusters of negatively charged amino acid residues. Surprisingly, deletions of these clusters still render cells Met+ and lead only to a marginal decrease in centromere activity.

Although small ubiquitin-like modifier (SUMO) is conjugated to proteins involved in diverse cellular processes, the functional analysis of SUMOylated proteins is often hampered by low levels of specific SUMOylated proteins in the cell. Here we describe a SUMO-conjugating enzyme (Ubc9) fusion-directed SUMOylation (UFDS) system, which allows efficient and selective in vivo SUMOylation of proteins. Although SUMOylation of overexpressed p53 and STAT1 was difficult to detect in HEK293 cells, up to 40% of p53 and STAT1 were conjugated with endogenous SUMO when fused to Ubc9. We verified the specificity of UFDS using SUMOylation-site mutants and showed that the method is not dependent on SUMO ligases. Using UFDS we demonstrated that SUMOylation of STAT1 inhibits its phosphorylation at Tyr701 and discovered p53 multi-SUMOylation in vivo. We propose that UFDS will be useful for the analysis of function of SUMOylation in protein interactions, subcellular localization as well as enzymatic activity.

No abstract