Overcoming Irinotecan Resistance by Targeting Its Downstream Signaling Pathways in Colon Cancer

Abstract

Among the most popular chemotherapeutic agents, irinotecan, regarded as a prodrug belonging to the camptothecin family that inhibits topoisomerase I, is widely used to treat metastatic colorectal cancer (CRC). Although immunotherapy is promising for several cancer types, only microsatellite-instable (~7%) and not microsatellite-stable CRCs are responsive to it. Therefore, it is important to investigate the mechanism of irinotecan function to identify cellular proteins and/or pathways that could be targeted for combination therapy. Here, we have determined the effect of irinotecan treatment on the expression/activation of tumor suppressor genes (including p15^Ink4b, p21^Cip1, p27^Kip1, and p53) and oncogenes (including OPN, IL8, PD-L1, NF-κB, ISG15, Cyclin D1, and c-Myc) using qRT-PCR, Western blotting, immunofluorescence (IF), and RNA sequencing of tumor specimens. We employed stable knockdown, neutralizing antibodies (Abs), and inhibitors of OPN, p53, and NF-κB to establish downstream signaling and sensitivity/resistance to the cytotoxic activities of irinotecan. Suppression of secretory OPN and NF-κB sensitized colon cancer cells to irinotecan. p53 inhibition or knockdown was not sufficient to block or potentiate SN38-regulated signaling, suggesting p53-independent effects. Irinotecan treatment inhibited tumor growth in syngeneic mice. Analyses of allograft tumors from irinotecan-treated mice validated the cell culture results. RNA-seq data suggested that irinotecan-mediated activation of NF-κB signaling modulated immune and inflammatory genes in mice, which may compromise drug efficacy and promote resistance. In sum, these results suggest that, for CRCs, targeting OPN, NF-κB, PD-L1, and/or ISG15 signaling may provide a potential strategy to overcome resistance to irinotecan-based chemotherapy.

Keywords: CDKIs; ISG15; NF-κB; PD-L1; SN38; colorectal cancer; drug resistance; immunomodulation; irinotecan; osteopontin.

Figure 1

SN38 induces expression of several apoptotic-related genes in addition to the conventional p53 pathway. DLD-1, SW480, and FET cells were treated with SN38 at 100 nM concentration for 48 h. (A) Western blots show higher levels of p21^Cip1, p27^Kip1, and Bax and lower levels of c-Myc and CyclinD1 upon SN38 treatment. Bar diagrams show the relative mRNA expression of p15^Ink4b, p21^Cip1, p27^Kip1, and p53 in (B) DLD-1, (C) SW480, and (D) FET cells after SN38 treatment. (E) p53-silenced DLD-1, SW480, and FET cells were treated with SN38 at 50 or 100 nM concentrations for 48 h. Western blots show the p53-independent, SN38-induced differential levels of p21^Cip1, p27^Kip1, and Bax. β-Actin served as a loading control. Statistical analysis of the samples was by a Student’s t-test. p < 0.05 was considered to be significant (* p < 0.05, ** p < 0.01, *** p < 0.001, and **** p < 0.001).

Figure 2

SN38 upregulates pro-oncogenic factors, including survivin, PD-L1, osteopontin, and ISG15. (A) DLD-1, SW480, and FET cells were treated with SN38 at increasing concentrations for 48 h. Conditioned media from samples were concentrated and normalized with cell counts and protein concentrations. Western blots show increased cellular levels of OPN, survivin, and PD-L1 and secreted levels of OPN. (B) p53-silenced DLD-1, SW480, and FET cells were treated with SN38 at 50 or 100 nM concentrations for 48 h. Western blots show the p53-independent, SN38-induced levels of OPN, PD-L1, and survivin. Cells were treated with an OPN Ab (2 μg/mL) in combination with various concentrations of SN38 for 48 h. Graph showing lower cell survival (%) (MTT assay) of (C) SW480 and (D) FET cells. (E) Western blots showing higher levels of ISG15 after 48 h of SN38 treatment of DLD-1, SW480, and FET cells. β-Actin served as a loading control. Statistical analysis of the samples was by a Student’s t-test. p < 0.05 was considered to be significant (* p < 0.05, ** p < 0.01, *** p < 0.001, and **** p < 0.001).

Figure 3

SN38 induces NF-κB activity and its nuclear localization. DLD-1, SW40, and FET cells were pretreated with Bay (1 μM) or QNZ (1 μM) for 3 h followed by in combination with SN38 (100 nM) for 24 h. (A) Western blots of nuclear lysates show elevated nuclear localization of p-p65 (NF-κB), an effect inhibited by QNZ. PARP served as a nuclear protein loading control. IF staining of (B) DLD-1, (C) SW480, and (D) FET cells show increased nuclear localization of p-p65 (NF-κB), an effect inhibited by QNZ (scale bar = 10 μm). IF data were analyzed, and NF-κB nuclear-positive cells were counted. Graph showing increased numbers of NF-κB-positive cells (%) (n = 1000) in (E) DLD-1, (F) SW480, and (G) FET cells, which were increased by SN38 treatment and partially regulated by Bay and effectively regulated by QNZ. (H) DLD-1, SW480, and FET cells were co-transfected with pGL2-NF-κB-luciferase and CMV-β-Gal followed by treatment with Bay or QNZ in combination with SN38 under the above conditions. Luciferase activity was measured and normalized with β-galactosidase activity. Bar diagram showing the increased relative luciferase activity, which was partially regulated by Bay and effectively regulated by QNZ. (I) Bar diagram showing the cell survival (%) (MTT assay) of Bay- or QNZ-treated DLD-1, SW480, and FET cells upon treatment with SN38 for 48 h. Statistical analysis of the samples was by a Student’s t-test. p < 0.05 was considered to be significant (** p < 0.01, *** p < 0.001, and **** p < 0.001).

Figure 4

SN38 induces immunomodulatory molecules through non-canonical NF-κB signaling. (A) OPN-silenced DLD-1 cells were treated with SN38 at 50 or 100 nM concentrations for 48 h. Western blotting shows OPN silencing decreases the level of PD-L1. (B) DLD-1 and SW480 cells were treated with PFTα (200 nM) (a p53 transactivation inhibitor) in combination with SN38 for 48 h. Western blots show no effects of p53 inhibition on PD-L1 levels. (C) DLD-1, (D) SW480, and (E) FET cells were pretreated with Bay (1 μM) or QNZ (1 μM) for 3 h followed by in combination with SN38 (100 nM) for 24 h. Bar diagrams showing increased mRNA expressions of IL8, CCL3, CCL5, and RANKL, an effect reduced by QNZ. Relative mRNA expressions of these proteins upon SN38 treatment relative to untreated cells, and the effects of SN38 treatment were compared with SN38 in combination with Bay or QNZ. (F) DLD-1, SW480, and FET cells were pretreated with Bay (1 μM) or QNZ (1 μM) for 3 h followed by in combination with SN38 (100 nM) for 48 h. Western blots show inhibition of SN38 on OPN, survivin, and ISG15. β-Actin served as a loading control. Statistical analysis of the samples was by a Student’s t-test. p < 0.05 was considered to be significant (* p < 0.05, ** p < 0.01, *** p < 0.001, and **** p < 0.001).

Figure 5

Irinotecan regulates tumor growth by differential regulation of pro- and anti-oncogenic factors. MC38 tumor-bearing mice were treated every third day with irinotecan (5 mg/kg) (n = 5) or irinotecan (15 mg/kg) (n = 5) for 24 days. Tumor volumes were calculated by the equation V = L × W² × 0.5, where L is the length and W is the width of a tumor. (A) Graph showing the kinetics of tumor growth. (B) Western blotting showing high levels of CD44, OPN, PD-L1, p21^Cip1, survivin, and ISG15 in tumor lysates after treatment of tumor-bearing mice with irinotecan (15 mg/kg). IF analysis showing increased (C) OPN, (D) p21^Cip1, (E) p65, (F) survivin, (G) PD-L1, (H) c-Myc, and (I) ISG15 and (J) decreased CyclinD1 in the tumor tissues of mice after treatment with irinotecan (15 mg/kg) (scale bar = 10 μm). β-Actin served as a loading control. Statistical analysis of the samples was by a Student’s t-test. p < 0.05 was considered to be significant (**** p < 0.001).

Figure 6

Irinotecan treatment regulates immune and inflammatory genes. RNA sequencing analysis after irinotecan (15 mg/kg) treatment compared to the vehicle control group (A) showing the expression of 3518 upregulated and 3650 downregulated genes. (B) Venn diagram showing the expression of 590 distinctive genes after irinotecan treatment and 560 distinctive genes in the vehicle control group. (C) GO pathway analysis showing the differential expression of various genes involved in biological processes, cellular components, and metabolic pathways. (D,E) Heatmaps showing high expressions of Sox2, p53, c-Myc, PD-L1, Snail, OPN, Oct4, p15^Ink4b, Bax, survivin, p21^Cip1, and Slug after treatment with irinotecan. Bar diagram showing mRNA expression from RNA seq data (F) increased TLR, (G) CXCR, (H) CXCL, (I) CCL, and (J) interleukins and their receptors (p < 0.01).

Figure 7

Graphical summary: p53-independent effects of irinotecan in colorectal cancer. ↑ Denotes upregulation and ↓ denotes downregulation of protein expression or signaling.