Fig 2: Identification of Senp5S isoform(A) The deduced amino acid sequences of mouse Senp5L and Senp5S. Identical amino acids are highlighted. Gaps in the alignment are indicated by dashes. Arrowhead denotes the catalytic center Cys-707 necessary for peptidase activity. His-640, Asp-657, and Cys-707 constitute the catalytic triad for cysteine protease. The N-terminal LLKL sequence (110–113) is the SUMO-interacting motif. Only Senp5S lacks the C-terminal catalytic center, although both isoforms have the N-terminal SUMO-interacting motif.(B and C) Developmental changes in Senp5L (B) and Senp5S (C) mRNA expression in the mouse cerebral cortex by qPCR analysis. Gray dots represent three independent experiments normalized to the corresponding ß-actin mRNA. Mean ± SD are also shown.(D) PCNs prepared from mouse embryonic brains were electroporated with Senp5 shRNA or non-targeting control shRNA and cultured for 3 div. Immunoblot with anti-Senp5 or a-tubulin antibody demonstrated the suppression of endogenous protein expression of Senp5L and Senp5S.(E) Developmental changes in Senp5L/5S protein expression in the mouse brain. Total protein extracts were prepared from the whole brains (E12–P28), adult cerebral cortex (ctx), or adult cerebellum (cb) and subjected to immunoblotting with anti-Senp5 antibody (left panel). Protein lysate from primary cultured neural stem/precursor cells (NSPC) or primary cultured cortical neurons (PCN) at 3 div was also analyzed by immunoblot (middle and right panels). The blots were reprobed with anti-a-tubulin antibody (bottom panels) to examine protein loading quantitatively.

Fig 3: Senp5L/5S regulate neurite outgrowth and polarization(A–C) Senp5 shRNAs (shRNA #01 or shRNA #02) or control shRNA was introduced into primary cultured cortical neurons along with a reporter plasmid encoding red fluorescent protein (RFP) (red). (A) Neurofilaments in individual neurons were stained with anti-SMI312 (green) at 3 div. (B) The stacked bar chart shows the percentages of cells having SMI312+ axons (stage 3). Numbers in parentheses indicate the number of cells analyzed. ***, p < 0.001; chi-square tests with Holm-Bonferroni correction. (C) Box and whisker plots summarize the numbers of RFP+ neurites extending from individual neurons. Numbers in parentheses indicate the number of cells measured in three independent experiments. ns, not significant; ***, p < 0.001; Wilcoxon rank sum tests with Holm-Bonferroni correction.(D–F) Overexpression of Senp5L and Senp5S in immature neurons. pIRES2-mCherry-Senp5L, pIRES2-mCherry-Senp5S, or control pIRES2-mCherry was electroporated into primary cultured cortical neurons. Neurofilaments and mCherry were stained with anti-SMI312 (green) and anti-RFP (red) at 3 div. (E) The stacked bar chart shows the percentages of cells having SMI312+ axons (stage 3). Numbers in parentheses indicate the number of cells measured in three independent experiments; ***, p < 0.001; chi-square tests with Holm-Bonferroni correction. (F) Box and whisker plots summarize the numbers of RFP+ neurites extending from individual neurons. Numbers in parentheses indicate the number of cells measured. ns, not significant; ***, p < 0.001; Wilcoxon rank sum tests with Holm-Bonferroni correction. Scale bars, 20 µm.

Fig 4: Senp5S competes with other Senps and promotes SUMOylation(A) Global SUMO3-ylation levels changed by expression of Senp5 isoforms. HEK293T cells were transfected with EGFP, EGFP-Senp3, EGFP-Senp5L, or EGFP-Senp5S along with HA-SUMO3 or HA-SUMO1, followed by immunoblotting with anti-HA (upper panel), anti-GFP (middle panel), or anti-a-tubulin (bottom panel).(B) Effect of Senp5L and Senp5S expression on SUMO3-ylation of Drp1. HEK293T expressing EGFP, EGFP-Senp5L, or EGFP-Senp5S along with Myc-Drp1 and HA-SUMO3 were immunoprecipitated (IP) with anti-HA antibody. Total cell lysates (input) or IP samples were analyzed by immunoblotting with anti-Myc (upper panel) and anti-HA (bottom panel). Immunoblot with an a-tubulin antibody as a loading control. SUMOylated-Drp1 is indicated by the perpendicular line. Heavy chain, immunoglobulin heavy chain of the HA antibody.(C) Domain structures of wild-type Drp1 (Drp1-WT) and a non-SUMOylatable Drp1 mutant (Drp1-4KR). The D-octadecapeptide sequence, comprising 18 amino acid residues (554–571) encompassing in the variable domain, is indicated. Drp1-4KR was created by substituting four SUMO-acceptor lysine residues (red) in the D-octadecapeptide with arginine residues. Oval and rectangle indicate GTPase activity domain and GTPase effector domain (GED), respectively.(D) Drp1 was SUMO3-ylated through the D-octadecapeptide sequence. HEK293 cells were co-transfected with HA-SUMO3, p14-Arf, Ubc9, and GFP, YFP-Drp1-WT or YFP-Drp1-4KR, then immunoprecipitated with anti-HA. Cell lysates (input) or IP samples were analyzed by immunoblotting with anti-GFP (top panel) or anti-HA (bottom panel). Immunoblot with an a-tubulin antibody as a loading control. SUMOylated-Drp1 is indicated by the perpendicular line.

Fig 5: Dysregulation of Senp5L/5S expression or Drp1-SUMOylation causes migration defects of neurons in embryonic cerebral cortex(A–C) Distribution of Senp5 in embryonic cortex. Coronal sections of cerebral cortex at E13.5 (A) and E16.5 (B) were immunostained with anti-pan-Senp5 antibody. (C) Higher magnification of the boxed area of (B). lv, lateral ventricle; CP, cortical plate; IZ, intermediate zone; VZ, ventricular zone.(D and E) Senp5 knockdown (KD) induces accumulation of migrating neurons in the IZ. Non-targeting control shRNA or Senp5 shRNA construct (shRNA #01 or shRNA #02) was electroporated into E14.5 cortices in utero together with TurboRFP. (D) Coronal sections of E17.5 cortex showing the distribution of TurboRFP+ cells expressing shRNA (yellow). Nuclei were counterstained with Hoechst dye (blue). (E) Distribution of TurboRFP+ cells in the indicated areas at E17.5. ***, p < 0.001; two-way ANOVA with Holm-Bonferroni correction.(F and G) Overexpression of Senp5L or Drp1-WT represses neuronal migration. (F) EGFP-Senp5L, EGFP-Senp5S, HA-Drp1-WT (with EGFP), HA-Drp1-4KR (with EGFP), or control EGFP (green) was electroporated in utero at E14.5, and the neocortex was analyzed at E17.5. (G) Distribution of electroporated EGFP+ cells in the indicated areas. ns, not significant; *, p < 0.05; ***, p < 0.001; two-way ANOVA with Holm-Bonferroni correction. Numbers in parentheses indicate the number of embryos analyzed.(H) Magnified view of immature neurons migrating in the IZ of the brain electroporated with HA-Drp1-WT or HA-Drp1-4RK. Sections were immunostained with anti-Tom20 (red) and anti-HA (magenta). Right: Higher magnification of the boxed areas showing the morphology of Tom20+ mitochondria (white). Arrows denote fragmented mitochondria.(I) Box and whisker plots summarize the mitochondrial length data (µm). Forced expression of Drp1-WT altered mitochondrial morphology in migrating neurons. Numbers in parentheses indicate the number of cells analyzed. ns, not significant; **, p < 0.01; Welch's t tests with Holm-Bonferroni correction. Stacked bar charts represent mean ± SEM. Scale bars: 100 µm in (A) and (B), 25 µm in (C), 100 µm in (D) and (F), 5 µm in (H), and 1 µm in the magnified view.