Analysis of the molecular mechanisms regulating paralogue-specific SUMOylation and its effects on protein function

The first S&barbelow;mall U&barbelow;biquitin related MOdifier (SUMO-1) was identified 12 years ago as a protein that is posttranslationally and covalently conjugated to the Ran G&barbelow;TPase A&barbelow;ctivating P&barbelow;rotein 1 (RanGAP1). To date, three SUMO-1 paralogues (SUMO-2 through SUMO-4) are identified and hundreds of additional proteins serve as substrates for modification. The functions of SUMOylation are diverse and dependent on the target proteins' specific activities. Although SUMOs regulate a variety of cellular processes, it is difficult to know the exact molecular effect of SUMOylation. This is because SUMOylation is a transient and dynamic process. For most proteins, less than 1% is conjugated by SUMO in vivo at steady state. In addition, the lack of substrate-specific SUMO conjugating and deconjugating enzymes makes it difficult to use classical methods to study the effect of SUMO modification on a single protein.;I started my thesis project by developing a novel rapamycin-mediated heterodimerization system that addresses these problems. Using RanGAP1 as a model substrate, I was able to reconstitute the effect of SUMO modification in targeting RanGAP1 to N&barbelow;uclear P&barbelow;ore C&barbelow;omplexes (NPCs) by non-covalently tethering SUMO to RanGAP1 by heterodimerization in mammalian cells. This system later proved to be useful for studying the effects of SUMO modification on other substrates.;Following these studies, I shifted my research focus to understanding the molecular mechanisms regulating paralogue-selective SUMOylation. Increasing evidence indicates that SUMO-1, SUMO-2 and SUMO-3 are preferentially conjugated to different subsets of proteins. Therefore, they may regulate distinct cellular functions. To answer the question of how proteins are preferentially modified by SUMO paralogues, I again used RanGAP1 as a model substrate. In mammalian cells, RanGAP1 is preferentially modified by SUMO-1. A series of biochemistry and cell biology experiments showed that the preferential SUMO-1 modification of RanGAP1 is due to selective binding of SUMO-1 modified RanGAP1 to Nup358 and protection from isopeptidase-mediated deconjugation. Our model can be a general mechanism for paralogue-selective SUMOylation of the other proteins.