Targeting pregnane X receptor with a potent agonist-based PROTAC to delay colon cancer relapse

Abstract

Tumor recurrence is frequently attributed to drug-tolerant cancer cells. We previously demonstrated that downregulation of the Pregnane X Receptor (PXR, NR1I2) reduces chemoresistance and prevents colorectal cancer recurrence in xenograft mouse models. However, there is currently a lack of clinically-suitable PXR antagonists. In this study, we report the design and synthesis of a novel PXR agonist-based PROTAC (JMV7048) which promotes polyubiquitination and degradation of the human PXR protein via E3 CRBN ubiquitin ligase and 26S proteasome pathways. JMV7048 selectively degrades PXR in colon carcinoma, hepatoma, and pancreatic cancer cell lines, with no impact on primary human hepatocytes. Notably, JMV7048 reduces PXR protein expression in drug-tolerant colon cancer cells, sensitizing them to chemotherapy and significantly delaying cancer relapse in xenografted nude mice. These findings suggest that PXR-targeting PROTACs may serve as novel therapeutic agents to enhance the sensitivity of chemo-resistant cancer cells to chemotherapy.

Fig. 1. Transformation of a PXR agonist into PXR-PROTAC.

A JMV6845, Pre-PROTAC JMV6944 and PROTACs structures (JMV7048, JMV7505 and JMV 7506). The right-hand side shows PROTAC molecules with different lengths of methyl linker in parenthesis between the agonist and the PROTAC part. B Close-up view of the interactions between JMV6944 and some residues of the ligand-binding pocket of PXR. The black dashed line denotes a hydrogen bond between the amine moiety of the side chain of JMV6944 and the main chain carbonyl group of cysteine 207. The inset displays a surface representation of PXR ligand-binding domain (in gray) showing how the JMV6944 side chain finds its way to the protein surface. C Close-up view of the interactions between JMV6845 (green) and some residues (orange) of the ligand-binding pocket of PXR (Oxygen=red, nitrogen=blue and sulfur=yellow). Key secondary structure elements (α-helices H2’, H3, H7, H11, H12; β-strands S1, S3, S4) are shown and labeled. The red asterisk denotes the C-H group of the imidazole ring serving as a derivatization site of JMV6845. D TR-FRET competitive assay between a fluorescent agonist and SR12813 (positive control agonist of PXR) or JMV6845 or JMV6944. TR-FRET ratios (520 nm/490 nm x 10⁴) are expressed as mean ± SD (n > 3). E RT-qPCR analyses of CYP3A4 mRNA expression in two primary human hepatocyte cultures (FH497 and FT468) treated 24 h with 5 µM Rifampicin, JMV6845 or JMV6944. Data are expressed as mean ± SD (n = 3). F, G Western blot and quantification of PXR expression levels in LS174T cells treated 24 h with 0.1% DMSO or 5 µM of the indicated compounds. Data are expressed as mean ± SD (n ≥ 3) of PXR/ACTIN ratio. H RT-qPCR analyses of CYP3A4 mRNA expression in LS174T treated 24 hours with 5 µM Rifampicin, JMV6944 or JMV7048. Data are expressed as mean ± SD (n = 3). FC=Fold-change, ***p < 0.0005, **p < 0.005, *p < 0.05. D–H data are normalized to the DMSO condition.

Fig. 2. JMV7048 is a PXR degrader.

A Western blot analysis and quantification of PXR protein expression levels in LS174T cells treated with the indicated JMV7048 concentrations. Data are expressed as mean ± SD (n > 3) and normalized to untreated cells (O). B Parallel analysis of LS174T cell proliferation index (C.I., xCELLigence apparatus) and quantification of PXR mRNA and protein expression levels after treatment with 5 µM JMV7048. Data are expressed as mean ± SD (n = 3). C Cell viability analysis of LS174T cells treated for 72 h with increasing concentrations of JMV7048 or SN38. Data are expressed as mean ± SD (n = 3). D Western blot analysis and quantification of PXR protein expression in LS174T after an initial treatment of 6 h with 500 nM JMV7048 following by its removal and wash-out for different times. Data are expressed as mean ± SD (n = 3). P values are shown compared to the initial condition (B–D) data are normalized to the DMSO condition. ***p < 0.0005, **p < 0.005, *p < 0.05. E JMV7159 structure. F Western blot analysis of PXR, RXRalpha (RXRα), FXR, VDR, and GSPT1 protein expression in LS174T cells, treated 24 h with 0.1% DMSO or 5 µM JMV7048 or JMV7159. In cellulo detection of JMV compounds in live cells by fluorescence imaging (G) or flow cytometry analysis (H). Cells were treated with or without 5 μM JMV7048 or JMV7159 for 24 h.

Fig. 3. JMV7048 is a bona fide PXR PROTAC.

A TR-FRET competitive assay between a fluorescent agonist and JMV6945, JMV6944 or JMV7048. TR-FRET ratios (520 nm/490 nm x 10⁴) are expressed as mean ± SD (n > 3) and normalized to DMSO (%). B Schematic representation of the competitive assays performed in (C) and JMV6945) structure. C Western-blot analysis of PXR expression in LS174T cells treated for 24 h with 5 µM JMV7048 with or without 5 µM JMV6944 or JMV6945. D Schematic representation showing the different actors in involved in the proteasome-mediated PXR degradation and their inhibitors (MLN4924 and Bortezomib, BZ). E Western blot analysis of PXR expression in LS174T cells treated 24 h with 5 µM JMV7048 or JMV7159 with or without CRBN E3 ligase inhibitor (0.5 µM MLN4924). The numbers above the bands represent PXR quantification (PXR/ACTIN vs DMSO). F Western blot analysis of PXR and ubiquitin (UBI) expression in LS174T cells lysates before (Input) and after PXR-immunoprecipitation (IP PXR). Cells were treated for 24 h with 5 µM JMV7048 in presence or absence 100 nM Bortezomib (BZ). PXR pUBI polyubiquitinated PXR. The numbers above the bands represent PXR quantification (PXR/ACTIN vs DMSO).

Fig. 4. JMV7048 degrades PXR in pancreas and hepatoma cell lines but not in human hepatocytes.

A Western blot analysis of PXR expression after 24 h treatment with 0.1% DMSO (control) or 5 µM JMV7159 or JMV7048 in LS174T, ASPC-1 or HepG2 cancer cell lines. B Western blot analysis of PXR expression after 24 h treatment with 0.1% DMSO or 5 µM JMV7048 in primary cultures of human hepatocyte. C Western blot quantification of A, B data are expressed as mean ± SD (n ≥ 3) of PXR/ACTIN ratio and normalized to DMSO (%). D Metabolic analysis (intrinsic clearance values -CL_int- and half-life -T_1/2-) of JMV7048 in human hepatocyte, LS174T or HepG2 cells treated with 1 µM JMV7048 for 5 min, 10 min, 1 h or 24 h (n = 2). E RT-qPCR analyses of MDR1, ABCG2, and CYP3A4 mRNA expression in primary cultures of human hepatocyte, HepG2 and LS174T cells. Data are expressed as mean ± SD (n = 3) and normalized to the primary human hepatocyte condition. FC Fold-change, ***p < 0.0005, **p < 0.005, *p < 0.05.

Fig. 5. JMV7048 decreases PXR expression in colon cancer stem cells and inhibits their population in vitro.

A Western blot analysis of PXR expression in LS174T Aldefluor-positive (ALDHpos) and negative (ALDHneg) sorted cells and then treated for 24 h hours with 5 µM JMV7048. B RT-qPCR analyses of CYP3A4 chemo-resistance gene mRNA expression in LS174T ALDHpos and ALDHneg sorted cells and then treated for 24 h hours with 0.1% DMSO or 5 µM JMV7048. (F.C., Fold Change). Data are expressed as mean ± SD (n = 3) and normalized to DMSO-treated ALDHneg cells. C Cell viability analysis of LS174T ALDHpos and ALDHneg LS174T sorted cells, treated 16 hours with or without 5 mM JMV7048 and then for 48 h with increasing concentrations of Firi, Firi 1X = 50μM 5-FU + 0.5 μM SN38). Data are expressed as mean ± SD (n = 3) and normalized to DMSO (%). D Quantification of Aldefluor-positive cells in LS174T, HT29 or CPP1 and CPP19 patient derived CCR cells after 24 h treatment with 5 µM JMV7048. Data are expressed as mean ± SD (n > 3) and normalized to DMSO (%). E Sphere-Forming Efficiency (SFC) assay and size of spheres (represented as pixel areas) of HT29 cell line incubated with 5 µM JMV7159 or JMV7048. F Log-rank Mantel–Cox Tumor initiation test. CPP1 cells were first treated in vitro for 48 h with 5 μM JMV7048 and were injected in Nude mice (500 or 5000 cells/mouse, n = 5). The graph shows tumor initiation and the table report the frequency of CSC (tumor initiating cells). G Cell viability analysis of HT29 spheroids or CPP1 and CPP19 patient derived CCR tumoroids pre-treated 24 h with 5μM JMV7048 and then co-treated with increasing concentrations of Firi (Firi 1X = 50 μM 5-FU + 0.5 μM SN38). Data are expressed as mean ± SD (n > 3) and normalized to DMSO (%). FC Fold-change, ***p < 0.0005, **p < 0.005, *p < 0.05.

Fig. 6. JMV7048 decreases PXR expression and tumor relapse.

A Pharmacokinetic studies (plasma concentration profiles, area under the plasma concentration versus time -AUCt-, maximum plasma concentration -Cmax-, and time to reach Cmax -Tmax) of JMV7048 in mice after intravenous (IV), intraperitoneal (IP) or oral (PO) administration. B After subcutaneous xenograft in Scid mice of 300.000 LS174T cells, mice were treated with or without JMV7048 (25 mg/kg, I.V., for 4 days, n = 10) when tumor volume reached 100 mm³. Western blot analysis and quantification of PXR protein in resected xenograft tumors. Data (PXR %) represent human-PXR/human-ACTIN ratio and normalized to vehicle treatment. C, D After subcutaneous xenograft of 20,000 cells, from HT29 spheroids, in athymic nude mice, when tumor volume reached 100 mm³, mice where treated (n = 10/groups) with FIRI (50 mg/kg 5-FU and 25 mg/kg irinotecan, IP, twice a week) with or without JMV7048 (IV, 5 days a week) for 4 weeks. Mouse weight (C) and tumor volume (D) analysis were measured twice a week. E After subcutaneous xenograft of 20,000 cells, from CPP1 spheroids, in athymic nude mice, when tumor volume reached 100 mm³, mice where treated (n = 20/groups) with FIRI (50 mg/kg 5-FU and 25 mg/kg irinotecan, IP, twice a week) with or without JMV7048 (IP, 5 days a week) over 3 weeks. Tumor volume (D) analysis were performed twice a week. F Calculated tumor volume doubling times (TDT) during (Treatment) and after (Relapse) treatments in both HT29 and CPP1 xenograft models. FC Fold-change, ***p < 0.0005, **p < 0.005, *p < 0.05.