Fig 2: Polη and Rad18 act in the same pathway in response to replication stress and this requires Rad18 UBZ but not its ubiquitin ligase activity.(a) MRC5-V1 cells were transfected with siRNAs directed against polη (siPOLH) and/or Rad18 (siRAD18) mRNAs 48 h before exposure to 0.15 μM APH for 24 h. The percentage of anaphases with chromosome fragments was scored (mean±s.d. of three independent experiments, n=50 for each experiment). See Supplementary Fig. 10a for siRNA efficiency. (b) The percentage of anaphases with chromosome fragments was scored in polη-depleted WT or RAD18−/− HCT116 cells 24 h after 0.15 μM APH (mean±s.d. of three independent experiments, n=50 for each experiment). See Supplementary Fig. 10c for siRNA efficiency. (c,d) MRC5-V1 cell populations expressing GFP, GFP-Rad18WT, GFP-Rad18C28F or GFP-Rad18C207F were depleted for endogenous Rad18 using a siRNA directed against the 3′-UTR of the mRNA (siR18 3′-UTR). Cells were then treated for 24 h with 0.15 μM APH and anaphases were analysed in GFP-positive cells. Data are the mean±s.d. of four (c) or three (d) independent experiments. See Supplementary Fig. 10d for siRNA efficiency. (e) The impact of Rad18 or PIAS1 depletion on the chromosome fragments in anaphase was determined in XPV, polηWT and polηSUMO cells 24 h after 0.3 μM APH. Data are the mean±s.d. of three independent experiments (n=50 for each experiment). ns: not significant, *P<0.05, **P<0.01, ***P<0.001 (t-test). (f) Western blot showing the efficiency of the siRNAs used in e.

Fig 3: Polη is SUMOylated in a PIAS1-dependent manner.(a) 293FT cells were transfected with pcDNA-POLH and Flag-PIAS1 or Flag-PIAS4 and co-immunoprecipitations were performed using an anti-polη antibody. (b) Reversed immunoprecipitations were performed with an anti-Flag antibody. (c) 293FT cells were transfected with the indicated siRNAs 24 h before co-transfection of plasmids expressing polη and His-SUMO3. Denaturing Ni pull-down was carried out as described in Fig. 2. (d) WT or K163R (KR) polη was co-expressed in 293FT cells with His-SUMO3 and Flag or Flag-PIAS1 and cells were processed as in c. (e) XPV, polηWT and polηSUMO cells were transfected with non-targeting or PIAS1 siRNAs. Nascent DNA was pulse-labelled with EdU for 5 min and cells were pre-extracted and fixed. Biotin was conjugated to EdU by click chemistry in order to perform an in situ PLA between polη and EdU-biotin. EdU-biotin was then counterstained using a fluorescent secondary antibody. The distribution of the number of PLA spots per EdU-positive cells was assessed in two independent experiments. One representative experiment is shown in box-plot with 10–90 percentile whiskers (n>150, ns: not significant, ***P<0.001, Mann–Whitney test). (f) Western blot showing the efficiency of PIAS1 depletion. (g) Representative images of the PLA experiment quantified in (e) (magnification × 63).

Fig 4: K163R and SUMO polη retain full catalytic activity in vitro.(a) Schematic representations of the polη mutants used in the study. In addition to WT and K163R polη, a constitutively SUMOylated polη (SUMO) was constructed by inserting the sequence of SUMO2 in place of K163. A track of 7 glycines (G) was added before SUMO2 to confer flexibility. The C-terminal di-G motif of SUMO2 was mutated to alanines to prevent cleavage by the SENPs. Two constructs were generated: SUMOa contains the atg of SUMO2, whereas it was mutated to G in SUMOb. A catalytically inactive polη (polDEAD) was used as a negative control. (b) The indicated proteins were produced in rabbit reticulocytes and their catalytic activity was analysed by primer extension on a circular single-stranded undamaged template (representative gel). (c) Polymerase activity is shown as a percentage of the activity measured for WT polη in 10-min reaction (mean±s.d. of three independent experiments). (d) For analysis of TLS efficiency, equal amounts of the in vitro translated proteins were introduced in an XPV cell extract. Bypass of a single TT-CPD (TLS) was assessed as described in Methods. L-1: fragment elongated up to one nucleotide before the lesion.

Fig 5: Abrogation of polη SUMOylation leads to replication defects in response to mild replication stress.(a,b) XPV cells stably expressing polηWT or polηK163R were treated with 0.3 μM APH for 6 or 24 h and processed as described in Fig. 4a–d. The a panel shows representative images after 24 h of APH. (c) XPV, polηWT and polηK163R cells were treated with 0.3 μM APH for 26 h before fixation and were immunostained with 53BP1 and cyclin A antibodies. Representative images are shown (magnification × 63). (d) The number of 53BP1 nuclear bodies (53BP1 NBs) was assessed in at least 100 cyclin A-negative cells (G1 cells, white arrows). Experiment was repeated three times, giving similar results. The distribution of 53BP1 NBs in G1 for one experiment is shown in a box-plot with 5–95 percentile whiskers (see also Supplementary Fig. 6b). ns: not significant; ***P<0.001 (Mann–Whitney test). (e) MRC5-V1 cells were transfected with a siRNA directed against polη mRNA (siPOLH) or a non-targeting control (siNT) 48 h before exposure to 0.15 μM APH for 24 h. Cells were fixed and centromeres were detected by immunostaining of CENPA. DNA was visualized using DAPI. Representative images of a regular anaphase and anaphases presenting lagging chromosome fragments are shown (magnification × 100). Most of the fragments lack CENPA staining (white arrows). CENPA can occasionally be found in lagging fragments (yellow arrow). (f) The percentage of anaphases presenting chromosome fragments was assessed in four independent experiments (mean±s.d., n=50 for each experiment). (g) Western blot confirming the efficiency of polη depletion. (h) The proportion of aberrant anaphases was also assessed in XPV cells and XPV cells stably expressing WT or K163R polη 24 h after treatment with 0.3 μM APH (mean±s.d. of five independent experiments, n=50 for each experiment). (i) XPV, polηWT and polηK163R cells were treated with 0.3 μM APH for 72 h and surviving fraction was expressed as a percentage of mock-treated cells (mean±s.d. of four independent experiments). ns: not significant, *P<0.05, **P<0.01 (t-test).