Pregnane X Receptor (PXR) expression in colorectal cancer cells restricts irinotecan chemosensitivity through enhanced SN-38 glucuronidation

Abstract

Background: Clinical efficacy of chemotherapy in colorectal cancer is subjected to broad inter-individual variations leading to the inability to predict outcome and toxicity. The topoisomerase I inhibitor irinotecan (CPT-11) is worldwide approved for the treatment of metastatic colorectal cancer and undergoes extensive peripheral and tumoral metabolism. PXR is a xenoreceptor activated by many drugs and environmental compounds regulating the expression of drug metabolism and transport genes in detoxification organs such as liver and gastrointestinal tract. Considering the metabolic pathway of irinotecan and the tissue distribution of Pregnane x Receptor (PXR), we hypothesized that PXR could play a key role in colon cancer cell response to irinotecan.

Results: PXR mRNA expression was quantified by RT-quantitative PCR in a panel of 14 colon tumor samples and their matched normal tissues. PXR expression was modulated in human colorectal cancer cells LS174T, SW480 and SW620 by transfection and siRNA strategies. Cellular response to irinotecan and its active metabolic SN38 was assessed by cell viability assays, HPLC metabolic profiles and mRNA quantification of PXR target genes. We showed that PXR was strongly expressed in colon tumor samples and displayed a great variability of expression. Expression of hPXR in human colorectal cancer cells led to a marked chemoresistance to the active metabolite SN38 correlated with PXR expression level. Metabolic profiles of SN38 showed a strong enhancement of SN38 glucuronidation to the inactive SN38G metabolite in PXR-expressing cells, correlated with an increase of UDPglucuronosyl transferases UGT1A1, UGT1A9 and UGT1A10 mRNAs. Inhibition of PXR expression by lentivirus-mediated shRNA, led to SN38 chemoresistance reversion concomitantly to a decrease of UGT1A1 expression and SN38 glucuronidation. Similarly, PXR mRNA expression levels correlated to UGT1A subfamily expression in human colon tumor biopsies.

Conclusion: Our results demonstrate that tumoral metabolism of SN38 is affected by PXR and point to potential therapeutic significance of PXR quantification in the prediction of irinotecan response. Furthermore, our observations are pharmacologically relevant since many patients suffering from cancer diseases are often exposed to co-medications, food additives or herbal supplements able to activate PXR. A substantial part of the variability observed among patients might be caused by such interactions.

Figure 1

Evaluation of PXR expression in human colorectal tissues and cell lines. A, PXR mRNA expression measured by quantitative real-time PCR in human livers (n = 17) and colon samples (tumoral and healthy adjacent tissue, n = 14). Bars on graph represent mean PXR mRNA expression and SEM. B, Percentage of PXR expression in colon tumors compared to adjacent healthy tissue; bars, SEM. C, PXR mRNA expression in various colon (LS174T, HT29, HCT116, SW480, SW620) and hepatic (HepG2, HuH7) cell lines compared to livers (FT305 and 310) and to normal (N61) and tumoral (T61) colon tissues. D, Immunocytochemical staining for PXR on liver, normal colon and tumors tissues. Liver was used as positive control for PXR expression (b), photos a, c and d display negative controls without primary antibody.

Figure 2

Characterization of LS174T PXR-transfected cells. A, PXR expression level in parent LS174T, pcDNA3-transfected and stable clones PXR2, PXR3 and PXR6 (top: mRNA expression level; bottom: protein expression level). * p < 0.05, PXR expression of stable clones compared to parent and pcDNA3-transfected cells, assayed by Mann and Whitney test. B, CYP3A4 mRNA expression levels of CTRL and PXR overexpressing clones treated 24 h by solvent (0.1% DMSO) or 10 μM rifampicin in serum-free medium. Results were obtained from three separate experiments; bars, SEM. a, CYP3A4 expression of cells treated by rifampicin compared to the vehicle treated groups, b, CYP3A4 expression of vehicle treated stable clones compared to vehicle treated; assayed by Mann and Whitney test (p < 0.05). C, proliferation rate of cell lines. Cells were seeded in six-well plates at 5 × 10⁵cells/well and counted at the indicated times after seeding (Z1 counter, Beckman Coulter). Data from three separate experiments; bars, SEM.

Figure 3

Increased chemoresistance in PXR overexpressing cells to irinotecan and SN38. For neutral red assays, cells were treated for 72 h by increasing concentrations of irinotecan (A) or SN38 (B). Columns, mean viability as a percentage of control (i.e., cells without chemotherapeutics treatment, 100%) from replicates (n = 6) from six separate experiments; bars, SEM. a, viability percentages of PXR2, PXR3 and PXR6 compared to pcDNA3-transfected cells (p < 0.05) (Mann and Whitney test). b, viability percentages of PXR2, PXR3 and PXR6 compared to each other using Kruskal Wallis test (p < 0.05).

Figure 4

A, cell viability assay on pcDNA3-transfected and PXR2 cells treated with rifampicin (RIF). Cells were cultured 24 h with DMSO 0.1% (solvent) or 10 μM rifampicin in 10% serum containing medium and then exposed to SN38 for 72 h. Columns, mean viability as a percentage of control (i.e., cells without chemotherapeutics treatment, 100%) from replicates (n = 6) from three separate experiments; bars, SEM. B, CYP3A4 mRNA expression level of pcDNA3-transfected or PXR-expressing cells cultured with (10%) or without serum and treated with solvent (0.1% DMSO) or 10 μM rifampicin for 24 h. Results were obtained from three separate experiments; bars, SEM. * p < 0.05 (Mann and Whitney test), CYP3A4 mRNA expression of cells cultured with serum, with or without rifampicin compared to vehicle treated cells. C, cell viability assay on pcDNA3-transfected and PXR2 cells treated with PXR antagonist L-Sulforaphane (SFN). Cells were cultured in presence of solvent (DMSO 0.1%) or 10 μM L-Sulforaphane in serum containing medium for 24 h followed by the 72 h treatment of SN38 with or without SFN co-treatment. Columns, mean viability as a percentage of control (i.e., cells without chemotherapeutics treatment, 100%) from replicates (n = 6) from three separate experiments; bars, SEM. *** p < 0.001, * p < 0.05 (student's t-test).

Figure 5

Reversion of chemoresistance in PXR2 by PXR shRNA. A, quantification of PXR expression by quantitative PCR (top) and western blot (bottom). * p < 0.05 (Mann and Whitney test), PXR expression in PXR2-Mock, PXR2-sh1334 and PXR2-sh2116 cells compared to pcDNA3-Mock cells. B, quantification of CYP3A4 mRNA expression after treatment by solvent (DMSO 0.1%) or 10 μM rifampicin for 24 h in serum-free medium. * p < 0.05 (Mann and Whitney test), CYP3A4 expression of cells treated by rifampicin compared to the vehicle treated groups. C, cell viability assay on pcDNA3-Mock, PXR2-Mock, PXR2-sh1334 and PXR2-sh2116 toward SN38. Columns, mean viability as a percentage of control (i.e., cells without chemotherapeutics treatment, 100%) from replicates (n = 6) from six separate experiments; bars, SEM. Student's t-test where performed between CTRL-mock cells to PXR2-Mock, PXR2-sh1334 and PXR2-sh2116: *** p < 0.001, ** p < 0.01, * p < 0.05.

Figure 6

HPLC quantification of SN38 and SN38G A, Relative percentage of total amounts (intra- and extracellular) of SN38 and SN38G on pcDNA3, PXR2, PXR3 and PXR6 cells after 24 h incubation with 10 μM SN38. Results were obtained from three separate experiments; bars, SEM. B, Intracellular and extracellular SN38G/SN38 ratios on pcDNA3, PXR2, PXR3 and PXR6 cells after 24 h incubation with 10 μM SN38. Results were obtained from three separate experiments; bars, SEM. *** p < 0.001, ** p < 0.01, * p < 0.05 (student's t-test) compared to CTRL cells. C, representative chromatograms of extracellular medium of pcDNA3-Mock, PXR2-Mock, PXR2-sh1334 and PXR2-sh2116 cells exposed to SN38.

Figure 7

A, UGT1As mRNA expression levels in pcDNA3, PXR2, PXR3 and PXR6 cells. Results were obtained from six separate experiments; bars, SEM. ** p < 0.01, * p < 0.05 (student's t-test) compared to CTRL cells. B, UGT1A1 mRNA expression levels in CTRL-Mock, PXR2-Mock, PXR2-sh1334 and PXR2-sh2116 cells. Results were obtained from four separate experiments; bars, SEM. *** p < 0.001, * p < 0.05 (student's t-test) compared to CTRL-Mock cells.

Figure 8

Correlation of PXR and total UGTs mRNA expression in colon tumors (n = 14). Correlation coefficient R² = 0.72, *** p < 0.001. Doted lines: SEM for a 95% CI.