Monoclonal antibodies (MAb) to members of the Small Ubiquitin-like modifier (SUMO) family are essential tools in the study of cellular SUMOylation. However, many reagents are poorly validated, and the choice of which antibody to use for which detection format is without an evidence base. Here we test twenty-four anti-SUMO monoclonal antibodies for sensitivity and specificity to SUMO1-4 in monomeric and polymeric states in dot-blots, immunoblots, immunofluorescence and immunoprecipitation. We find substantial variability between SUMO MAbs for different conjugation states, detection of increased SUMOylation in response to thirteen different stress agents and as enrichment reagents for analysis of SUMOylated RanGAP1 or KAP1. All four anti-SUMO4 monoclonal antibodies tested cross-reacted with SUMO2/3, and several SUMO2/3 monoclonal antibodies cross-reacted with SUMO4. ~10% of tested monoclonal antibodies produced specific results across multiple applications.

Protein conjugation with the Small Ubiquitin-like Modifier SUMO1 or the related SUMO2/3 drive changes to protein behaviour. Many substrates are found modified by both SUMO1 and SUMO2/3, while others are modified by one or the other. How isoform specificity is directed is poorly understood. Here we examine modification of the catalytic component of the human SUMO Activation Enzyme, SAE2. We find that an acetylated K164-SAE2 analogue preferentially activates SUMO2 in competition with SUMO1, and that K164-SAE2 discriminates paralogues through their C-terminal regions. We find that K164-SAE2 is deacetylated during mitosis. Mitotic defects in cells expressing an acetylated K164-SAE2 analogue can be corrected by over-expression of SUMO1, suggesting SUMO1 conjugation driven by the deacetylated enzyme supports mitotic fidelity. These surprising data reveal that modification of the SUMO-activating enzyme can bias SUMO paralogue conjugation.

The small ubiquitin-like modifier (SUMO) family is composed of five members, SUMO1, the highly similar SUMO2/SUMO3, SUMO4 and the tissue-specific SUMO5 (SUMO1P1). Sequence variation in SUMO4 is thought to prevent its maturation, resulting in an un-conjugatable SUMO isoform, and consequently, its functions are poorly understood. Here we show for the first time that SUMO4 promotes DNA double-strand break signalling in a manner distinct from SUMO1 or SUMO2/3. We show that SUMO4 function depends on interaction with partner proteins through SUMO interacting motifs and, on its inability, to be conjugated. We show that SUMO4 promotes the activity of the SUMO protease SENP1. In the absence of SUMO4, reduced SENP1 catalytic activity results in hyperSUMOylation that unbalances the recruitment of several DSB repair factors, including RAP80. These data reveal that SUMO4 acts as a buffer for the SUMOylation system.

Cells lacking several DNA repair proteins, including those promoting homologous recombination (HR), are sensitive to polymerase theta (Polθ) repression. Polθ drives theta-mediated end joining (TMEJ) and suppresses HR but what mediates its synthetic lethal relationships is unclear. Here we examine murine Brca1C61G/ C61G 53bp1-/-cells and find they are largely HR proficient by using RNF168 and RAD52. They exhibit no more TMEJ than 53bp1-/- cells yet are more sensitive to targeting of Polθ. We find that RAD52 recruitment to damaged chromatin is increased following Polθ depletion. RAD52 accumulation and cellular sensitivity to Polθ repression can be curbed by the RAD51-binding regions of BARD1 and BRCA2, and sensitivity of BRCA1/2 depleted cells to Polθ repression is suppressed by RAD52 inhibition. 53bp1-/- cells exhibit a smaller increase in RAD52 recruitment following Polθ repression and also become resistant to Polθ repression following RAD52 inhibition. Thus, RAD52 mediates sensitivity to targeting Polθ in these contexts