Fig. 7: PSMC5 and TUBB3 are directly SUMO1ylated by PIAS2b, and retained at the spindle at mitotic onset. PIAS2b-dsRNAi activates the proteasome too early in mitosis with catastrophic results.

PSMC5 and TUBB3 PIAS2b-dependent SUMOylation at mitosis, using non-denaturing extracts with SUMO-protease inhibitors (a), or 10xHis-SUMO1 transfected denaturing extracts (b). From top-to bottom re-stripped westerns presented. a GFP-Trap pull-downs of EGFP or EGFP-PSMC5 in asynchronous (As), DT/R-0 h, DT/R-6 h (Left); or treated with ns-dsRNAi or PIAS2b-dsRNAi at DT/R-5 h (Right). EGFP-PSMC5 and endogenous PSMC5 enhance SUMO1ylation at mitosis, lost with PIAS2b silencing. Endogenous PSMC5 associates to exogenous EGFP-PSMC5. PIAS2b is associated to- and SUMO1ylated in the GFP pull-downs. b 10xHis-SUMO1 Ni-NTA pull-downs in denatured extracts reveal covalent SUMO1ylated – or poly-SUMOylated, PSMC5, TUBB3 and PIAS2b at DT/R-6 h, that are lost with PIAS2b silencing. mSUMO1 and His-Tag patterns show enhanced SUMO1ylation at DT/R-6h, and reduced SUMO1ylation in PIAS2b absence. Unspecific bands (-u) in Ni-NTA pull-downs of double consecutively transfected, 10xHis-Tag-SUMO1+dsRNAi, required longer gel running (right). (Controls in Supplementary Fig. 7b). Western blots (anti-SUMO1 and anti-His-Tag) in Eluted lanes show similar patterns due to detection of identical SUMOylated proteins, while some variation in Input lanes reflects both SUMO1ylation and overexpressed 10xHis-tagged SUMO1ylation. However, PIAS2b depletion accumulates free mono-His-SUMO1. c Imaris quantification using 3D super-resolution microscopy of EGFP-PSMC5 or rPSMC5, mPIAS2 or rPIAS2b, TUBB-647 or SUMO1-647 combinations presented as violin plots of double (green, orange) and triple (blue, pink) colocalization. d Surface rendering of progressive mitotic phases with increasing association at the spindle (planes in Supplementary Fig. 7c). e–g Nuclear accumulation of PSMC5 is promoted by PIAS2b-dsRNAi silencing at DT/R-5 h, with earlier abnormal proteasome activity increase and reduced pSer10-HH3. h Mitotic perturbations at the kinetochore and centrosome at DT/R-5 h in PIAS2b-dsRNAi cells (pictures in Supplementary Fig. 7d–f): gTubulin + CREST colocalizing nuclear spots increase to 23, coincident with kinetochores, while the few in ns-dsRNAI are side-to-side with CREST; increased nuclear gTubulin colocalized with BUB3 + CREST (kinetochore); duplicated PLK1 (mitosis onset) and PCNT do not colocalize; gTubulin and pericentrin (PCNT) colocalization reveals qualitative and quantitative centrosome alterations. a, b n = 3 independent experiments. c, d Violin plots with cells per condition indicated, from n = 5 independent experiments; two-sided one-way ANOVA with Holm–Sidak’s multiple comparison test. f n = 2 independent experiments. g n = 6 independent experiments, two-sided unpaired T-test. h (upper part) at least 100 cells analyzed per condition, and h (bottom) PLK1 + PCNT n = 4, Centrosomes analysis n = 8 independent experiments; two-sided Mann–Whitney. Bar indicate means ± SEM; exact p value is indicated in the Figure. Source Data are provided.