N-terminal deletions of SUMO1 proteins, but not SUMO2, greatly increase binding of SIM-containing proteins.

(A) Overlay of 5 selected states of a SUMO1 solution structure (2n1v) using Pymol. Whereas the SUMO core and C-terminus (dark grey) remain stable in all states, the N- terminus (light grey) shows a high degree of flexibility. (B) Partial sequence alignment of human SUMO1, SUMO2, C. elegans SMO-1, and S. cerevisiae Smt3 using the Clustal Omega web tool. The flexible N-terminal extension is indicated by a box, the SIM-binding groove is highlighted grey. (C) and (E) Pulldown of recombinant Usp25, Ubc9, TDP2 or RanBP2 with SUMO1 or SUMO2 analyzed by SDS-PAGE and Coomassie staining. (D) The amounts of Usp25 in the eluates of SUMO pull-downs in (C) were quantified relative to Ubc9 (n = 4). (F) The amounts of Usp25 in the eluates of SUMO pull-downs in (E) were quantified relative to Ubc9 (n = 6). (G) Pulldown of recombinant S. cerevisiae Siz1 or hUbc9 with Smt3 analyzed by SDS-PAGE and Coomassie staining. (H) The amounts of Siz1 were quantified relative to hUbc9 (n = 4). (I) Pulldown of recombinant C. elegans GEI-17 or hUbc9 with Ovalbumin (Ova) or Smo- 1 beads analyzed by SDS-PAGE and Coomassie staining. (J) The amounts of GEI-17 were quantified relative to hUbc9 (n = 3). Error bars represent one standard deviation. Asterisks indicate p-values: n.s.: P > 0.05; **: P ≤ 0.01; ***: P ≤ 0.001

N-terminal deletions of SUMO1 augment binding and SUMOylation of Usp25 by enhancing SUMO-SIM interactions.

(A) If not occupied by the N-terminus, the SIM-binding groove of SUMO1 can aid complex formation between SUMOylated proteins and their SIM-containing potential interaction partners. (B) SIMs in SUMOylation targets can promote SUMO-modification by stabilizing complex formation with SUMO-loaded Ubc9 termed “SIM-dependent SUMOylation”. (C) and (E) Pulldown of recombinant Usp25 wt, Usp25-V91A,I92A (SIM mutant), or RanBP2 with glutathione beads loaded with the indicated SUMO (S)-GST variants, analyzed by SDS-PAGE and Coomassie staining. (D) and (F) The amounts of Usp25 and RanBP2 in the eluates of SUMO pull-downs in (C) and (E) were quantified relative to SUMO-GST (n = 3). (G) - (L) In vitro SUMOylation reactions of recombinant Usp25 wt or -V91A,I92A, and RanGAP1 with the indicated human SUMO1, SUMO2 or C. elegans SMO-1 mutants, analyzed by SDS-PAGE and immunoblotting. Error bars represent one standard deviation. Asterisks indicate p-values: **: P ≤ 0.01; ***: P ≤ 0.001;

NMR experiment and Molecular Dynamics simulations reveal that the N- termini of SUMO1 and SMO-1 transiently occupy the SIM binding groove.

(A) 1H-15N hNOE values for SUMO1. The x-axis shows SUMO1 residues in a non- contiguous fashion; residues not shown were either prolines (residues 8, 58 and 77) or could not be analyzed due to insufficient data quality (residues 1-3, 24, 29, 41 and 97). Colored backgrounds show areas with indicated dynamics. (B) HSQC-TROSY experiments comparing full-length SUMO1 wt and the ΔN19 variants. Histogram of CSPs of (15N)SUMO1 ΔN19 compared to full-length (15N)SUMO1; the resonance for Met19 could not be assigned confidently (shown as off-scale in histogram - marked with *); residues 1-18 are displayed only for clarity with a value of 0; Pro8, Pro58, Pro77 are not shown. The SIM binding groove and 70/80 region are marked (black bars). (C-F) Binding time, averaged over 8 1μs-trajectories, of the SUMO N-terminus to other residues for SUMO1, SUMO2, SMO-1 and yeast Smt3. (G) Average binding time of the SUMO N-terminus to the SIM binding groove for SUMO1, SUMO2, SMO-1 and yeast Smt3. Data points (n=8): average along single 1 μs trajectories; bars: average over all 8 trajectories, errors bars: standard errors of the mean of the 8 averages. Asterisks indicate p-values: *: P ≤ 0.05; **: P ≤ 0.01; ***: P ≤ 0.001. (H) Kinetic rate constants for binding/unbinding processes between the N-terminus of SUMO1, SUMO2, SMO-1 and yeast Smt3 and the SIM binding groove.

Mutating negatively charged residues in the N-terminus of SUMO1 abolishes the inhibitory effect of the N-terminus.

The SIM binding groove, encompassing β2-strand and α-1 helix, and N-terminus were shown as non-transparent. The positively/negatively charged residues were colored in blue/red, respectively. The charged residues in N-terminus and around the SIM binding groove were shown as ball and stick. The residues related with post-translational modifications or mutations were highlighted in the figure with their names. (A) human SUMO1; (B) C. elegans SMO-1; (C) Pulldown of recombinant Usp25, Ubc9 or TDP2 with the indicated SUMO1 beads analyzed by SDS-PAGE and Coomassie staining. (D) The amounts of Usp25 in the eluates of SUMO pull-downs in (C) were quantified relative to Ubc9 (n = 6). (E) Pulldown of recombinant C. elegans GEI-17 or hUbc9 with Ovalbumin (Ova) or indicated SMO-1 beads analyzed by SDS-PAGE and Coomassie staining. (F) The amounts of GEI-17 were quantified relative to hUbc9 (n = 3). (G) and (H) In vitro SUMOylation of recombinant Usp25 and RanGAP1 with human SUMO1 wt or the indicated mutants, analyzed by SDS-PAGE and immunoblotting. Error bars represent one standard deviation. Asterisks indicate p-values: *: P ≤ 0.05; **: P ≤ 0.01;

Negatively charged residues in the N-terminus of SUMO1 mediate the inhibitory effect on SIM-dependent interactions

(A) HSQC-TROSY experiments comparing full-length SUMO1 wt and the ED11,12KK mutant. Histogram of CSPs of (15N)SUMO1-ED11,12KK compared to full-length (15N)SUMO1; the resonances for Thr10, Lys12, Gly14 and Glu20 of the mutant could not be assigned confidently (shown as off-scale in histogram - marked with *); Pro8, Pro58 and Pro77 are not shown. The SIM binding groove (black bars) as well as the folded core of the protein (grey background) are marked. (B-C) Binding time and (D) Average binding time of the SUMO N-terminus to the SIM binding groove for wt (identical to Fig. 3C) and ED11,12KK variant of SUMO1. (E-F) Binding time and (G) Average binding time of the N-terminus of wt (identical to Fig. 3E) and DD3,4KK of SMO-1 to residues of the SUMO core. Data points: average along single 1 μs trajectories; bars: average over 8 trajectories, errors bars: standard errors of the mean of the 8 averages. (H) Pulldown of recombinant Usp25 with GST-tagged SUMO1/SUMO2 chimeras, analyzed by SDS-PAGE and Coomassie staining. (I) The amounts of Usp25 in the eluates of SUMO pull-downs in (H) were quantified relative to the SUMO-GST proteins (n = 3). Asterisks indicate p-values: *: P ≤ 0.05; **: P ≤ 0.01

Post-translation modifications of SUMO’s intrinsically disordered N- termini may regulate SIM-dependent interactions.

(A) Average binding time of the SUMO N-terminus to the SIM binding groove for SUMO1 wt (identical to Fig. 3C) and pS2, pS9, and pS9pT10 SUMO1 variants. (B) and (C) Binding time of the N-terminus of SUMO1 wt (identical to Fig. 3C) and SUMO1-pS9 to individual residues of the SUMO core. (D) Average binding time of the SUMO N- terminus to the SIM binding groove for SUMO2 wt and the pT12, AcK7 and AcK11pT12 SUMO2 variants. (E) and (F) Binding time of the N-terminus of SUMO2 wt (identical to Fig. 3D) and SUMO2-pT12 to individual residues of the SUMO core. Data points: average along single 1 μs trajectories; bars: average over 8 trajectories, errors bars: standard errors of the mean of the 8 averages. (G) Pulldown of recombinant Usp25 or Ubc9 with the indicated SUMO1 beads, analyzed by SDS-PAGE and Coomassie staining. (H) The amounts of Usp25 in the eluates of SUMO pull-downs in (G) were quantified relative to Ubc9 (n = 4). Error bars represent one standard deviation. Asterisks indicate p-values: *: P ≤ 0.05; **: P ≤ 0.01; ***: P ≤ 0.001. (I) Model of the intrinsically disordered region of SUMO serving as a regulatable cis - inhibitor for SIM - dependent interactions through post-translation modification.

Deletion of the flexible N-terminus in C. elegans SMO-1 reduces germ cell survival.

(A) Whole-mount DAPI staining of the wild-type and of SMO-1ΔN12 (smo-1(zh156)) animals 24 hours post L4 stage. (B) Brood size, (C) number of germ cell nuclei at the pachytene stage per gonad arm, and (D) area of the pachytene region in wild-type and SMO-1ΔN12 animals. (E) Average number of pachytene stage germ cell nuclei per area. (F) Apoptotic corpses outlined by CED-1::GFP staining in wild-type and SMO- 1ΔN12 mutants 24 and 48 hours post L4. The engulfed corpses are indicated with white arrowheads. The left panels show the DIC images overlayed with CED-1::GFP signal (average intensity projections of the outer gonad layers) and the right panels the CED- 1::GFP signal alone. (G) Number of CED-1::GFP-positive apoptotic corpses per gonad arm in wild-type and SMO-1ΔN12 animals with or without cep-1(gk148) background 24 and 48 hours post L4 stage. (H) Number of SYTO12 positive germ cells per gonad arm in wild-type and SMO-1ΔN12 animals 29 hours post L4 stage. Bars indicate average values, dots individual values and error bars the standard deviations. The numbers in brackets indicate the numbers of animals scored. If the values were normally distributed (in B - E, H), statistical significance was calculated with an unpaired parametric t-test, or else (in G) a Kruskal-Wallis test was used. Asterisks indicate the p-values as * p≤0.05; ** p≤ 0.01; *** p≤ 0.001; **** p≤0.001, n.s. for not significant. The scale bars in (A & F) represent 20µm.