The green fluorescent protein (GFP) from the jellyfish Aequorea victoria has attracted much attention as a tool to study a number of biological processes. This study describes the use of GFP as a vital reporter molecule for localization and expression studies in Saccharomyces cerevisiae. Construction of GFP expression vectors which allow N‐ or C‐terminal fusion of the gfp gene to a gene of interest allowed the generation of fusion proteins whose subcellular localization was followed by fluorescence microscopy in living yeast cells. Analysis of three unknown open reading frames obtained from the budding yeast chromosome XIV resulted in distinct staining patterns, allowing prediction of the cellular localization of these unknown proteins. Furthermore, GFP was used to construct a gene replacement cassette which, after homologous integration into the genomic locus, placed the gfp gene behind a promoter of interest. The amount of GFP produced from this promoter was then quantified in living yeast cells by flow cytometry. With this novel replacement cassette a gene of interest can be deleted and at the same time its expression level studied under various growth conditions. The experiments presented here suggest that GFP represents a convenient fluorescent marker for localization studies as well as gene expression studies in budding yeast. Systematic studies of a large number of genes should benefit from such assays.

Spinal muscular atrophy (SMA), a frequent neurodegenerative disease, is caused by reduced levels of functional survival of motoneuron (SMN) protein. SMN is involved in multiple pathways, including RNA metabolism and splicing as well as motoneuron development and function. Here we provide evidence for a major contribution of the Rho-kinase (ROCK) pathway in SMA pathogenesis. Using an in vivo protein interaction system based on SUMOylation of proteins, we found that SMN is directly interacting with profilin2a. Profilin2a binds to a stretch of proline residues in SMN, which is heavily impaired by a novel SMN2 missense mutation (S230L) derived from a SMA patient. In different SMA models, we identified differential phosphorylation of the ROCK-downstream targets cofilin, myosin-light chain phosphatase and profilin2a. We suggest that hyper-phosphorylation of profilin2a is the molecular link between SMN and the ROCK pathway repressing neurite outgrowth in neuronal cells. Finally, we found a neuron-specific increase in the F-/G-actin ratio that further support the role of actin dynamics in SMA pathogenesis.

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.

In a systematic approach to the study of Saccharomyces cerevisiae genes of unknown function, 150 deletion mutants were constructed (1 double, 149 single mutants) and phenotypically analysed. Twenty percent of all genes examined were essential. The viable deletion mutants were subjected to 20 different test systems, ranging from high throughput to highly specific test systems. Phenotypes were obtained for two-thirds of the mutants tested. During the course of this investigation, mutants for 26 of the genes were described by others. For 18 of these the reported data were in accordance with our results. Surprisingly, for seven genes, additional, unexpected phenotypes were found in our tests. This suggests that the type of analysis presented here provides a more complete description of gene function.