PMS2抑制剂NP1867诱导MMR缺陷与抗PD-1敏感性

Figure 1.  X-ray crystal structure of NP1867 bound to human PMS2-N terminal domain. A and B, NP1867 (pink in A, purple in B) in the ATP-binding site of human PMS2-NTD (gray) depicting key water-mediated interactions with Asp70, Gly74, and Ser46, direct H-bond interactions with Asp70 and Asn45 (magenta dashed lines with water molecules shown as red spheres), and covalent bond formation with Cys73 (S-enantiomer at adducted carbon atom shown in A, R-enantiomer in B: PDB ID: 9S89). C, View of the protein crystal structure of human PMS2-NTD bound to NP1867 illustrating the unresolved flexible region from residues Thr84 to Glu109 (both highlighted in purple); NP1867 shown in green, and protein in gray. D, Chemical structure of NP1867.

Figure 2.  Target engagement of NP1867 to PMS2 monitored via band shift analysis and PRM-MS in human and murine cancer cell lines. A and B, CT26 murine cells were treated with the indicated concentrations of NP1867 or with NP2567 at 1 μmol/L for 2, 4, 8, or 24 hours. Target engagement of NP1867 to PMS2 was monitored by either A, PMS2 band shift with a tailored Western blotting method, or B, PRM-MS quantifying unadducted Cys73 PMS2 peptide relative to unadducted control peptides (DMSO). B, each color corresponds to a different molecular ion of the adducted Cys73-containing PMS2 peptide, n = 1 experiment. C, HAP1 and D, SW620 human cells were treated with NP1867 for 15, 30, 60, and 120 minutes or with NP2567 for 120 minutes at 0.3 μmol/L; PMS2 target engagement was assessed by PMS2 band shift Western blotting.

Figure 3.  PMS2 KO induces an MMR-d phenotype characterized by 6-TG tolerance. A, Expression of MMR proteins MLH1 and PMS2 and tubulin detected by JESS (automated Western Blotting) in HAP1 WT PMS2 expressing cells or PMS2 KO cells. B and C, HAP1 WT expressing cells or PMS2 KO cells were treated with (+) or without (−) 6-TG at 1 μmol/L for 48 hours (B) or 6-TG at 0.8 μmol/L for 72 hours (C). B, Cells with DNA strand breaks were detected by TUNEL; dUTP incorporation at free 3′-OH ends of DNA strand breaks. Percentage of dUTP-positive cells was measured by flow cytometry. n = 75 experiments. Data are shown as individual values obtained per experiment (circle) and mean (horizontal line) ± SEM. Statistical analysis: one-way ANOVA followed by Tukey test. C, Cell-cycle phases were measured by flow cytometry. n = 4 experiments. Mean ± SEM. Statistical analysis: two-tailed t test. ns, not significant; ***, P < 0.001.

Figure 4.  Inhibition of PMS2 induces an MMR-d phenotype, characterized by 6-TG tolerance and TMZ resistance. A–D, HAP1 WT expressing cells or PMS2 KO cells were treated with or without (no 6-TG) 6-TG at 1 μmol/L for 48 hours (A, B, and D) or at 0.8 μmol/L for 72 hours (C). In addition to 6-TG treatment, cells were treated with the indicated compounds (NP1867 or NP2567) added 24 hours prior to 6-TG treatment at the indicated concentrations. A and B, Cells with DNA strand breaks were detected by TUNEL; dUTP incorporation at free 3′-OH ends of DNA strand breaks. The percentage of cells with DNA strand breaks (dUTP positive) was measured by flow cytometry. A, n = 4 experiments. B, In WT cells, 6-TG + NP2567: n = 3 experiments. In PMS2 KO cells, 6-TG + NP2567: n = 1 experiment. In both WT and PMS2 KO cell lines, no 6-TG + NP1867 or + NP2567: n = 2 experiments. Data are shown as individual values obtained per experiment (circle) and mean (horizontal line) ± SEM. C, Cell-cycle phases were measured by flow cytometry. n = 4 experiments. D, HAP1-WT cells were cotreated with 6-TG at 1 μmol/L and NP1867 at various concentrations 72 hours. Annexin V–positive apoptotic cells were detected by Incucyte. n = 3 experiments. E, Human SW620 colorectal cancer cells were cotreated with TMZ at 25 μmol/L and the indicated concentrations of NP1867 for 4 days. Annexin V–positive apoptotic cells were detected by Incucyte. n = 3–5 experiments. F, Murine CT26 colorectal cancer cells were treated with NP1867 alone at 10 μmol/L (no 6-TG) or with 25 nmol/L 6-TG (6-TG) and the indicated concentrations of NP1867, negative control NP2567 at 30 μmol/L, or vehicle control (DMSO) for 7 days. Doubling time was measured by live-cell imaging with Incucyte. n = 3–4 experiments. G and H, Murine CT26 cells were treated with (+) or without (−) 6-TG at 0.5 μmol/L and with DMSO, NP2567, or NP1867 at 1 μmol/L for 7 days. At day 7, cells were fixed and fluorescently labeled with CellTag700. G, Representative pictures of the whole 24-well plate. H, Fluorescence detection performed by Odyssey Imaging System (LI-COR Biosciences) n = 3 experiments. B–H, Data are shown as mean ± SEM, unless otherwise indicated. Statistical analysis: two-tailed t test. ns, not significant; *, P < 0.05; **, P < 0.01;***, P < 0.001.

Figure 5.  Treatment with NP1867 leads to genotypic changes associated with COSMIC MMR-d mutational signatures. Murine CT26 cells were treated for up to 10 cycles (T10) of 4 weeks with DMSO (vehicle control) or NP1867 at 0.5 μmol/L in triplicate. Analysis of 4 (T4), 6 (T6), 8 (T8), and 10 (T10) cycles of 4-weeks treatment, compared with the start of the experiment of (A) the TMB (SNVs + Indels) expressed as the number of muta- tions gained/Mb, (B) the number of unique mutations gained per replicate and per treatment group (DMSO or NP1867) at each time point, and (C), the MSI score analysis. The red dotted line highlights the threshold for MSI-H status (25). A and C, Data are shown as box-and-whisker plots, median IQR minimum and maximum, or B as mean ± SEM. A–C, Statistical analysis: two-tailed t test. ns, not significant; *, P < 0.05; **, P < 0.01;***, P < 0.001. D, Mutational profile of the gained SNVs of one representative replicate of each NP1867- or DMSO-treated arms at T10. E, Mutational signature analysis of NP1867-treated cells showing enrichment over time of MMR-d–associated mutational signatures, COSMIC SBS15 and COSMIC SBS21. A–E, Data shown as mean ± SEM of n = 3 replicates.

Figure 6.  Chronic treatment with NP1867 sensitizes murine CT26 tumor cells to anti–PD-1 therapy in vivo. A–F, Murine CT26 cells were treated with NP1867 at 0.5 μmol/L or DMSO in vitro for 10 cycles of 4 weeks (T10). A and B, Tumor growth curves (n = 20 mice/group) of NP1867-treated (A) and DMSO-treated (B) CT26 T10 cells in immunocompetent BALB/c mice in the presence of isotype control or anti–PD-1 therapy. Black dotted lines annotated with black arrows denote the timing of isotype control or anti–PD-1 treatments. Tumor volume shown as mean ± SEM until 70% of animals per group remain on study. Statistical test: nonparametric Mann–Whitney test with Holm–Šídák correction. C and D, Cell type profiling scores for total TILs (C) and T cells (D) in satellite tumors (n = 6 mice/group) from A and B, based on the expression of specific gene markers for each cell by NanoString PanCancer Immune Profiling panel analysis. Statistical analysis: one-way ANOVA test followed by Tukey multiple compar- ison test. Data are shown as box-and-whisker plots, median IQR minimum and maximum. E, Heat map of the pathway scores for all 29 predefined immune-related gene sets in the NanoString PanCancer Immune Profiling panel comparing pathway activity levels in satellite tumors (n = 6 mice/group) across all four groups from A and B [DMSO-treated T10 cell tumors receiving isotype control or anti–PD-1 (aPD-1) therapy, and NP1867-treated T10 cell tumors receiving isotype control or anti–PD-1 therapy]. F, GSA significance scores of predefined immune-related pathways in the NanoString PanCancer Immune Profiling panel in tumors from satellite animals (n = 6 mice/group) from Supplementary Fig. S8A. GSA significance scores were computed for the following comparisons: (i) NP1867 T10 tumors receiving anti–PD-1 vs. isotype control therapy–baseline–(green dots/line) and (ii) DMSO T10 tumors receiving anti–PD-1 vs. isotype control therapy–baseline–(black dots/line). The top 10 pathways with the highest GSA signifi- cance scores of 29 predefined pathways are shown. ns, not statistically significant; *, P < 0.05; **, P < 0.01; ***, P < 0.001. DC, dendritic cell.