Leucine-rich repeat-containing, G protein-coupled receptor 5 (LGR5) is highly expressed in colorectal cancer and cancer stem cells (CSCs) that play important roles in tumor initiation, progression, and metastasis. Loss of LGR5 has been shown to enhance therapy resistance. However, the molecular mechanisms that mediate this resistance remain elusive. In this study, we demonstrate conversion of LGR5+ colorectal cancer cells to an LGR5 state in response to chemotherapy, LGR5 targeted antibody–drug conjugates (ADCs), or LGR5 gene ablation led to activation of STAT3. Further investigation revealed increased STAT3 activation occurred as a result of increased mesenchymal epithelial transition (MET) factor receptor activity. LGR5 overexpression decreased MET-STAT3 activity and sensitized colorectal cancer cells to therapy. STAT3 inhibition suppressed MET phosphorylation, while constitutively active STAT3 reduced LGR5 levels and increased MET activity, suggesting a potential feedback mechanism. Combination treatment of MET-STAT3 inhibitors with irinotecan or antibody–drug conjugates (ADCs) substantiated synergistic effects in colorectal cancer cells and tumor organoids. In colorectal cancer xenografts, STAT3 inhibition combined with irinotecan enhanced tumor growth suppression and prolonged survival. These findings suggest a mechanism by which drug-resistant LGR5 colorectal cancer cells acquire a survival advantage through activation of MET-STAT3 and provide rationale for new treatment strategies to target colorectal cancer.

Colorectal cancer stem cells (CSCs), or tumor-initiating cells, mediate drug resistance and relapse through their capacity to either self-renew or differentiate into heterogeneous lineages of tumor cells (1, 2). Leucine-rich repeat-containing, G protein-coupled receptor 5 (LGR5), is highly expressed in approximately 60% to 70% of colorectal cancer (3–5) and is a validated marker of normal adult intestinal stem cells and functional CSCs (6–8). Originally identified as a Wnt target gene, LGR5 and related receptors LGR4/6 bind R-spondin (RSPO; refs. 1–4) growth factors responsible for modulating Wnt/β-catenin signaling, an important pathway implicated in colorectal cancer and stem cell maintenance (9).

LGR5+ CSCs can initiate colorectal tumor growth and exhibit a high level of plasticity, or the capacity to shift between CSC and non-CSC states during disease progression and to circumvent therapy. Elimination of LGR5+ CSCs using either genetic ablation approaches or LGR5 targeted antibody–drug conjugates (ADCs) resulted in tumor inhibition or stasis, with relapse after treatment termination (3, 5, 6, 8). LGR5 colorectal cancer cells were able to sustain tumors, yet the reemergence of LGR5+ CSCs was proven necessary for tumor regrowth. More recently, it was reported that the majority of colorectal cancer cells that disseminate from the primary tumor for seeding liver metastases are LGR5 (10). However, conversion of LGR5 colorectal cancer cells to an LGR5+ state was necessary to promote metastatic outgrowth, demonstrating dynamic plasticity exists during disease progression and metastasis. Furthermore, treatment with standard chemotherapies was shown to trigger LGR5+ CSCs to transition to an LGR5, drug-resistant state (11). Correspondingly, we reported knockdown (KD) or knockout (KO) of LGR5 in colorectal cancer cells conferred a more drug-resistant phenotype (12). Although LGR5+ CSC plasticity is well established, the molecular and cellular processes underpinning plasticity and therapy resistance remain relatively unknown.

In this study, we aimed to identify a molecular mechanism involved in mediating therapy resistance in colorectal cancer cells with loss of LGR5. We showed conversion of colorectal cancer cells from LGR5+ to an LGR5 state in response to different therapies or LGR5 gene ablation led to activation of the MET–STAT3 pathway. LGR5 colorectal cancer cells acquired a survival advantage through MET-STAT3 activation to become more resistant to treatment. Combination therapy of MET-STAT3 inhibitors with standard chemotherapy enhanced treatment efficacy and prolonged survival. These findings identify LGR5 as a novel regulator of MET-STAT3 signaling and may lead to new treatment strategies for targeting refractory colorectal cancer cells.

Chemicals and plasmids

The anti-LGR5 ADC (anti–LGR5-mc-vc-PAB-MMAE) with drug–antibody ratio of 4 was generated as described previously (3). Irinotecan and 5-fluorouracil were purchased from Selleck and Acros Organics, respectively. Stattic was purchased from Tocris, cryptotanshinone, gefitinib from Selleck, crizotinib from Cell Signaling Technology, and XAV939 from Cayman Chemical. The plasmid encoding myc-LGR5 was generated previously (9). Constitutively active mouse STAT3 (STAT3-CA) plasmid Stat3-C Flag pRc/CMV was a gift from Jim Darnell (Addgene, 8722).

Cell culture

LS180, HCT116, and DLD-1 cells were purchased from the ATCC. LoVo cells were obtained from Dr. Shao-Cong Sun (MD Anderson Cancer Center, Houston, TX). Cell lines were authenticated utilizing short tandem repeat profiling, routinely tested for Mycoplasma and cultured in RPMI medium supplemented with 10% FBS and penicillin/streptomycin at 37°C with 95% humidity and 5% CO2. Transient transfections were performed using jetPRIME (Polypus Transfection). LS180 stable LGR5 CRISPR/Cas9 KO clonal lines 1.4 and 1.5 were generated using the lenti-CRISPRv2 vector system as described (12). LoVo-stable shRNA knockdown lines pLKO.1 (shCTL), shLGR5–1, and shLGR5–2 were generated as previously reported (3, 13).

qRT-PCR

Total RNA was isolated using TRIzol Reagent (Thermo Fisher Scientific) followed by column purification using the RNeasy Mini Kit (Qiagen). cDNA was synthesized SuperScript IV VILO master mix with ezDNase enzyme (SABiosciences). qPCR analysis was performed using an amfiSure qGreen Master Mix (GenDEPOT) on CFX96 Touch Real-Time PCR Detection System (Bio-Rad). Each sample was run in triplicate. LGR5 mRNA was normalized to levels of GAPDH and relative expression was calculated using the 2(−ΔΔCt) method. Sequences of the primers used for LGR5 were (forward)-CCTGCTTGACTTTGAGGAAGACC and (reverse)-CCAGCCATCAAGCAGGTGTTCA and for GAPDH, (forward)-TCAAGGCTGAGAACGGGAAG and (reverse)- CGCCCCACTTGATTTTGGAG. All experiments were performed three times.

Patient-derived tumor organoids

The three-dimensional (3D) patient-derived tumor organoid (PDO) model was established from a metastatic rectal tumor harvested from patient-derived xenograft (PDX) mouse model TM00849 (The Jackson Laboratory). Tumor tissue was cut into small fragments, washed with ice-cold PBS, and subsequently digested with Liberase (Roche Life Science) for 1 hour at 37°C with vigorous pipetting every 15 minutes. The remaining fragments were treated with TrypLE Express (Invitrogen) at 37°C for 20 minutes to disperse into single cells. Supernatant was collected and centrifuged at 200 × g for 3 minutes at 4°C. The cell pellet was suspended with growth factor–reduced Matrigel (Corning) and dispensed into 48-well culture plates (25 μL Matrigel/well). After Matrigel polymerization, cells were overlaid with complete medium (Advanced DMEM/F12 supplemented with 50 ng/mL EGF, 100 ng/mL noggin, 500 nmol/L A83–01, 10 nmol/L gastrin, 1× B-27, 1 mmol/L N-acetylcysteine, 10 mmol/L HEPES, 2 mmol/L GlutaMAX, and penicillin/streptomycin) that was replenished every 2 to 3 days.

Western blotting

For Western blot analysis, protein extraction was performed using RIPA buffer (Sigma) supplemented with protease/phosphatase inhibitors. Cell lysates were incubated at 37°C for 1 hour in Laemmli SDS sample buffer before loading on SDS-PAGE. Commercial antibodies used in this study were anti-LGR5 (ab75732), anti-EGFR Y1173 (ab32578) from Abcam and anti-p-STAT3 Y705 (9145), anti-STAT3 (9139), anti-p-MET Y1234/1235 (3077), anti-MET (8198), anti-p-EGFR Y1068 (3777), anti-EGFR (54359), anti–non-p-β-catenin (8814), anti–β-catenin (8480), anti-Cyclin D1 (55506), anti-Bcl-xl (2764), anti-flag (14793), and anti–β-actin (3700) from Cell Signaling Technology. Horseradish peroxidase (HRP)-labeled anti-mouse (Cell Signaling Technology, 7076) and anti-rabbit (Cell Signaling Technology, 7074) secondary antibodies were used for detection with the standard enhanced chemiluminescence protocol.

Viability assays

Colorectal cancer cell viability assays were performed as described previously (12). For tumor organoid viability assays, 1,000 cells/well (passage 3) were seeded into a 48-well plate and incubated for 3 days. Serial dilutions of commercial drugs or anti-LGR5 ADCs were added at the indicated concentrations and allowed to incubate at 37°C for another 3 to 4 days. The viability of colorectal cancer cells and organoids was measured using CellTiter-Glo and CellTiter-Glo 3D cell viability assays (Promega), respectively, according to the manufacturer's protocol. IC50s were determined using Prism 5 (GraphPad Software, Inc.). Representative results of at least three independent experiments are shown. Luminescence was measured using EnVision multilabel plate reader (PerkinElmer). Bright-field images of organoids were acquired using an Olympus IX71 microscope.

Clonogenicity assays

For colony-formation assays, cells were seeded in 6- or 12-well plates at 5,000 or 1,000 cells per well, respectively. For soft agarose assays, cells were seeded in culture medium containing low-melting-point agar at a density of 500 cells per well in 12-well plates. After 2 weeks, the colonies were stained with crystal violet. Experiments were performed at least three times and colonies were quantified by manual counting or using ImageJ.

Immunocytochemistry

Cells were seeded into 8-well chamber slides and allowed to adhere overnight. For LGR5 internalization experiments, cells were treated with irinotecan at the indicated time points and incubated with anti-LGR5 rh8F2 mAb (14) at 37°C for 30 minutes. Cells were washed, fixed in 4% formaldehyde, permeabilized with 0.1% saponin (Sigma), and incubated with anti-human Alexa 555 (Invitrogen) for 1 hour at room temperature. For detection of STAT3 translocation, cells were fixed, permeabilized with 0.3% 1× Triton-X, and incubated with anti-STAT3 Ab (Cell Signaling Technology, 9139), followed by anti-mouse Alexa-488 (Invitrogen). Nuclei were counterstained with TO-PRO-3 (Invitrogen). Images were acquired using confocal microscopy (Leica TCS SP5 microscope) with the LAS AF Lite software (Leica Microsystems, Inc.).

Animal studies

In vivo experiments were carried out in accordance with, and with the approval of, the Institutional Animal Care and Use Committee at UTHealth (AWC-20–0144). NSG-PDX model TM00849 was purchased from The Jackson Laboratory. Female 6- to 8-week-old nu/nu mice (Charles River Laboratories or The Jackson Laboratory) were subcutaneously inoculated with 1 × 106 LoVo or 2 × 106 LS180 cells in 1:1 mixture of PBS and Matrigel into lower right flank. Once tumors reached approximately 100 mm3 to 150 mm3, animals were randomized into four treatment groups (vehicle, stattic, irinotecan, or irinotecan + stattic). Irinotecan treatment was initiated on day 0 and stattic treatment was initiated on day 2. Irinotecan (20 mg/kg) or vehicle (2% DMSO in 5% dextrose in sterile water) was administered intraperitoneally every 5 days (LoVo = 3 doses; LS180 = 2 doses). Stattic (10 mg/kg) or vehicle was administered intraperitoneally every other day (LoVo = 6 doses; LS180 = 3 doses). Tumor volumes were measured at least biweekly and estimated by the formula: Tumor volume = (length × width2)/2. Mice were euthanized when tumor volume reached approximately 2,000 mm3. Once the first tumor from each vehicle group reached maximum tumor burden, treatment was terminated, and animals were monitored for survival.

Statistical analysis

Statistical analysis was performed and IC50 values calculated using the Prism 5 (GraphPad Software, Inc.). In vitro experiments were performed at least three times. The levels of significance between samples were determined through an unpaired two-tailed Student t test (mean comparison with one factor) or one-way ANOVA for groups with multiple comparisons. Combination index (CI) values were calculated by the Chou–Talalay method using CompuSyn (15). Statistical significance of differences in tumor growth was determined using one-way ANOVA and Dunnett multiple comparison test, and the log-rank (Mantel–Cox) test was used for survival studies. Data are shown as mean ± SEM or SD as indicated. P values ≤ 0.05 were considered significant.

Data and material availability

All data are contained within the article and materials can be requested upon reasonable request.

Drug treatment in colorectal cancer cells promotes a resistant LGR5 state

We previously showed loss of LGR5 expression increased resistance to chemotherapies and proliferation of colorectal cancer cells (12). Furthermore, treatment with LGR5-targeted ADCs eliminated LGR5+ colorectal cancer tumors; however, a fraction of tumors relapsed with LGR5-low/negative expression (3). Therefore, we questioned how both gene ablation and therapy-induced loss of LGR5 promotes a more drug-resistant state and if similar molecular mechanisms are involved. First, we tested the effect of chemotherapy and targeted ADCs on LGR5 expression to verify LGR5+ colorectal cancer cells convert to an LGR5 state during treatment. LoVo and LS180 colorectal cancer cell lines were selected on the basis of high endogenous expression of LGR5 (12). Colorectal cancer cells were treated with 10 μmol/L irinotecan or 6.5 nmol/L anti-LGR5 ADC (anti-LGR5-mc-vc-PAB-MMAE) at indicated time points, and changes in LGR5 expression were measured by Western blot analysis (Fig. 1A and B). LGR5 was nearly undetectable after 72 hours in both LoVo and LS180 cells. Comparable results were observed after treatment with 10 μmol/L 5-fluoruoracil (Supplementary Fig. S1A). Immunocytochemistry (ICC) staining revealed LGR5 is homogeneously expressed throughout the colorectal cancer cell lines and verified that transition from an LGR5+ to an LGR5 state is due to drug-induced downregulation rather than survival of cells that originated as LGR5. Comparable with Western blot analysis, irinotecan treatment led to an approximate 75% and 85% decrease in LGR5 positivity after 24 hours, for LoVo and LS180 cells, respectively, with near-complete loss in LGR5 expression after 72 hours (Fig. 1C). Consistently, qRT-PCR showed LGR5 mRNA decreased after treatment with irinotecan or ADCs (Supplementary Fig. S1B and S1C). LGR5 colorectal cancer cells transitioned back to an LGR5+ state after irinotecan and ADC washout (Fig. 1D). These data demonstrate therapy induces the conversion of LGR5+ colorectal cancer cells to an LGR5 state and suggests the LGR5 cells may be more resistant due to activation of survival signaling mechanisms.

Figure 1.

Treatment-induced loss or gene ablation of LGR5 converts colorectal cancer cells to a more drug-resistant state. A and B, Western blot analysis showing time course of changes in LGR5 expression in LoVo and LS180 cells treated with irinotecan (10 μmol/L; A) or anti-LGR5 ADC (6.5 nmol/L; B). C, Confocal microscopy images of LGR5 expression in LoVo and LS180 after irinotecan (10 μmol/L) treatment. D, Western blot analysis of LGR5 expression in cells treated with anti-LGR5 ADCs (6.5 nmol/L) or irinotecan (10 μmol/L) for 72 hours and 72 hours after drug washout. E–H, Quantification of colony-formation assays in LoVo parental (P) shRNA control (CTL) and LGR5 KD cells treated with (E) vehicle or (F) irinotecan (10 μmol/L) and LS180 CTL and LGR5 KO cells treated with vehicle (G) or irinotecan (5 μmol/L; H). I, Cytotoxicity of anti-LGR5 ADC after 4 days. J, Quantification of anti-LGR5 ADC dose-dependent effects on colony formation of LoVo cells. Experiments were performed at least three times. Statistical analysis was performed using Student t test (J) or one-way ANOVA. *, P ≤ 0.05; **, P ≤ 0.01, compared with controls. Error bars are SEM.

Figure 1.

Treatment-induced loss or gene ablation of LGR5 converts colorectal cancer cells to a more drug-resistant state. A and B, Western blot analysis showing time course of changes in LGR5 expression in LoVo and LS180 cells treated with irinotecan (10 μmol/L; A) or anti-LGR5 ADC (6.5 nmol/L; B). C, Confocal microscopy images of LGR5 expression in LoVo and LS180 after irinotecan (10 μmol/L) treatment. D, Western blot analysis of LGR5 expression in cells treated with anti-LGR5 ADCs (6.5 nmol/L) or irinotecan (10 μmol/L) for 72 hours and 72 hours after drug washout. E–H, Quantification of colony-formation assays in LoVo parental (P) shRNA control (CTL) and LGR5 KD cells treated with (E) vehicle or (F) irinotecan (10 μmol/L) and LS180 CTL and LGR5 KO cells treated with vehicle (G) or irinotecan (5 μmol/L; H). I, Cytotoxicity of anti-LGR5 ADC after 4 days. J, Quantification of anti-LGR5 ADC dose-dependent effects on colony formation of LoVo cells. Experiments were performed at least three times. Statistical analysis was performed using Student t test (J) or one-way ANOVA. *, P ≤ 0.05; **, P ≤ 0.01, compared with controls. Error bars are SEM.

Close modal

Deletion of LGR5 enhances survival and drug resistance

As altered survival capacity of colorectal cancer cells is a critical determinant of drug resistance and metastatic potential of tumor cells, we examined changes in clonogenicity with loss of LGR5. We previously generated LoVo LGR5 KD cell lines with independent shRNAs (shLGR5–1 and shLGR5–2) and LS180 LGR5 KO cell lines (1.4 and 1.5) using the CRISPR/Cas9 system (12). Clonogenicity assays were performed in the absence and presence of irinotecan (Fig. 1EH). LGR5 KD/KO cells showed increased colony formation compared with parental (P) and vector control (cytotoxic T lymphocyte; CTL) cell lines when seeded at low density, approximately threefold for LoVo LGR5 KD cells and six- to 11-fold for LS180 LGR5 KO cells (Fig. 1E and G; Supplementary Fig. S1D and S1E). The enhanced resistance and clonogenic capacity of LGR5 KD/KO cells was further demonstrated when cells were subjected to irinotecan. Clonogenicity was significantly increased in LoVo LGR5 KD (10-fold) and LS180 LGR5 KO cells (four- to sevenfold) compared with respective controls (Fig. 1F and H; Supplementary Fig. S1D and S1E). Similar findings were also shown for 5-fluoruoracil (Supplementary Fig. S1F). We reported that LoVo cells were sensitive to anti-LGR5 ADCs linked to an MMAE payload; however, other colorectal cancer cell lines are more resistant, rendering these ADCs less effective despite high LGR5 expression. As shown in Fig. 1I, LoVo cells respond to ADC treatment with an approximate IC50 = 2–5 nmol/L, whereas LS180 cells are relatively resistant. This may be attributed to a 10-fold difference in MMAE potency for LS180 compared with LoVo cells (IC50 = 2.3 vs. 0.2 nmol/L, respectively; Supplementary Fig. S1G). ADC efficacy was diminished with LGR5 KD/KO, demonstrating ADC specificity (Fig. 1I). Treatment of LoVo cells with ADC reduced clonogenicity in a dose-dependent manner (Fig. 1J; Supplementary Fig. S1H). As expected, LGR5 KD increased clonogenicity approximately 2.5-fold and cells were resistant to ADC treatment (Fig. 1J; Supplementary Fig. S1H). These results show that loss of LGR5 expression increases survival and clonogenicity in vitro.

LGR5 KD/KO in colorectal cancer cells increases STAT3 activation mediated through MET

Western blot analysis was performed to identify signaling mechanisms that may mediate therapy resistance in LGR5 KD/KO cells. LoVo LGR5 KD cells showed increased levels of active (nonphosphorylated) β-catenin, consistent with our previous report (13). No significant changes in MEK/ERK or Src activation was observed; two pathways shown mediate drug resistance (Supplementary Fig. S1I). Interestingly, we detected increased phosphorylation and nuclear accumulation of STAT3 with LGR5 KD (Fig. 2A and B), suggesting increased STAT3-mediated transcriptional activity. Consistently, we observed changes in expression levels of STAT3-associated survival proteins involved in proliferation and inhibition of apoptosis (16). Specifically, loss of LGR5 resulted in increased Cyclin D1 and Bcl-xL gene and protein expression (Fig. 2A; Supplementary Fig. S1J). These findings show loss of LGR5 increases Wnt/β-catenin and STAT3 signaling in colorectal cancer cells.

Figure 2.

Loss of LGR5 enhances phosphorylation of STAT3 through activation of MET. A, Western blot analysis of changes in β-catenin, STAT3, and associated target proteins in LoVo control and LGR5 KD cells. B, Confocal microscopy images of phosphorylated STAT3 (green) and TO-PRO-3 staining of nuclei (blue) in LoVo cells. Scale bars in magnified images, 5 μm. C, RNA-seq data for LoVo and LS180 cells from CCLE. D, Western blot analysis of EGFR and MET (β-subunit) in LoVo cells. E, Western blot analysis of changes in β-catenin, STAT3, MET, and associated target proteins in LS180 control and LGR5 KO cells. F, Effect of vehicle (0.1% DMSO), MET inhibitor (crizotinib, 10 μmol/L), or STAT3 inhibitor (stattic, 10 μmol/L) on phosphorylation in LoVo LGR5 KD and LS180 LGR5 KO cells after 6 hours. Experiments were performed at least three times.

Figure 2.

Loss of LGR5 enhances phosphorylation of STAT3 through activation of MET. A, Western blot analysis of changes in β-catenin, STAT3, and associated target proteins in LoVo control and LGR5 KD cells. B, Confocal microscopy images of phosphorylated STAT3 (green) and TO-PRO-3 staining of nuclei (blue) in LoVo cells. Scale bars in magnified images, 5 μm. C, RNA-seq data for LoVo and LS180 cells from CCLE. D, Western blot analysis of EGFR and MET (β-subunit) in LoVo cells. E, Western blot analysis of changes in β-catenin, STAT3, MET, and associated target proteins in LS180 control and LGR5 KO cells. F, Effect of vehicle (0.1% DMSO), MET inhibitor (crizotinib, 10 μmol/L), or STAT3 inhibitor (stattic, 10 μmol/L) on phosphorylation in LoVo LGR5 KD and LS180 LGR5 KO cells after 6 hours. Experiments were performed at least three times.

Close modal

To identify which upstream mediators may be responsible for increased STAT3 phosphorylation, we characterized changes in expression and activation of membrane receptors known to mediate STAT3 activation, including IL6R/Gp130, EGFR, and mesenchymal–epithelial transition (MET; refs. 17–19). RNA sequencing (RNA-seq) data retrieved from the Cancer Cell Line Encyclopedia (CCLE) showed LoVo and LS180 colorectal cancer cells do not express IL6R and have minimal expression of IL6ST (Gp130; Fig. 2C). Compared with LGR5 expression, both cell lines express high levels of MET and moderate expression levels of EGFR (Fig. 2C). To determine whether Gp130 was playing a role in STAT3 activation, LGR5 KD cells were treated with Sc144, a small-molecule Gp130 inhibitor (Supplementary Fig. S2A). Inhibition of Gp130 failed to suppress STAT3 activation, suggesting IL6R/Gp130 is not responsible for aberrant STAT3 signaling. We next examined LGR5 KD cells for changes in activation of EGFR and MET by Western blot analysis. As shown in Fig 2D, no changes in EGFR phosphorylation at sites known to mediate STAT3 signaling (Y1068 or Y1173) were observed. Furthermore, LGR5 KD cells treated with EGFR inhibitor, gefitinib, failed to suppress STAT3 activation (Supplementary Fig. S2B). These findings suggest that STAT3 activation in LGR5 KD cells is most likely not mediated through IL6R/Gp130 or EGFR.

MET receptor tyrosine kinase (RTK), initially synthesized as a partially glycosylated 170-kDa single-chain intracellular precursor (pro-Met), undergoes extensive posttranslational modifications to become a functionally mature protein (20). Mature MET is composed of a 45-kDa α-subunit linked by two disulfide bonds to a 145-kDa β-subunit (20). Of note, LoVo cells exhibit a defect in normal processing of pro-MET, resulting in a 190-kDa single-chain protein (21). Interestingly, LGR5 KD cells have increased total and phosphorylated levels of mature MET as indicated by the 145-kDa band (Fig. 2D). Similarly, we detected an approximate twofold increase in phosphorylation of MET and STAT3 and increased levels of active β-catenin in LS180 LGR5 KO and DLD-1 LGR5 KD cells (Fig. 2E; Supplementary Fig. S2C). LS180 LGR5 KO cells also showed increased expression of STAT3-associated target genes Cyclin D1 and Bcl-xL (Fig. 2E). The effect of MET inhibition on STAT3 was investigated by treating LGR5 KD/KO cells with MET inhibitor, crizotinib, which functions through competitive binding within the ATP-binding pocket (Fig. 2F; Supplementary Fig. S2D). Crizotinib has also been shown to inhibit anaplastic lymphoma kinase (ALK). However, based on CCLE RNA-seq data, neither LoVo nor LS180 cells express this gene (i.e., RPKM = 0 and <0.0075, respectively). Western blot analysis showed that crizotinib treatment decreased both phosphorylated MET and STAT3 (Fig. 2F; Supplementary Fig. S2D), suggesting STAT3 activation is mediated through MET. We also observed loss of active MET after treatment of LGR5 KD/KO cells with STAT3 inhibitors stattic and cryptotanshinone (Fig. 2F; Supplementary Fig. S2E and S2F), which interact with the STAT3 SH2 domain to prevent Tyr705 phosphorylation and dimerization (22, 23). Stattic also decreased levels of total STAT3 in LS180 cells. STAT3 inhibition of MET indicates the involvement of a potential feedback mechanism between STAT3 and MET. These data suggest that STAT3 activation in LGR5 KD/KO cells is mediated through MET.

Inverse regulation of LGR5 and STAT3

To further elucidate the relationship between LGR5 expression and MET-STAT3 activation, constitutively active STAT3 (STAT3-CA) was overexpressed in LoVo and LS180 cells by transient transfection. Western blot analysis showed STAT3-CA concomitantly reduced LGR5 expression and increased MET phosphorylation (Fig. 3A). Viability assays showed STAT3-CA cells were more resistant to irinotecan and anti-LGR5 ADCs, supporting a role for STAT3 in mediating resistance in LGR5 KD/KO cells (Fig. 3BD). Average IC50 values increased two- to threefold with STAT3-CA overexpression in both cell lines (Supplementary Table S1). We next tested whether LGR5 overexpression alters MET-STAT3 activation. We selected HCT116 cells, as they are devoid of endogenous LGR5 and have high baseline levels of phosphorylated STAT3 (Supplementary Fig. S2G). LGR5 was overexpressed by transient transfection using increasing amounts of DNA (Fig 3D). Western blot analysis showed gradual loss of MET and STAT3 activation with increasing amounts of LGR5 (Fig. 3E). LGR5 overexpression also sensitized HCT116 cells to irinotecan treatment, reducing the IC50 by approximately 4.5-fold (Fig. 3F; Supplementary Table S1). These findings further support a role for MET-STAT3 activation in driving drug resistance and survival in colorectal cancer cells with loss of LGR5.

Figure 3.

Effects of constitutively active STAT3 and LGR5 overexpression on treatment resistance. A, Western blot analysis of MET and STAT3 phosphorylation in LoVo and LS180 cells transiently transfected with flag-tagged constitutively active STAT3 (STAT3-CA). B and C, Cytotoxicity of irinotecan in LoVo (B) and LS180 (C) cells transfected with vector or STAT3-CA after 4 days. D, Cytotoxicity of anti-LGR5 ADC in colorectal cancer cells transfected with vector or STAT3-CA after 4 days. E, Western blot analysis of HCT116 cells transfected with increasing amounts of LGR5. F, Cytotoxicity of irinotecan in HCT116 cells after 4 days. Western blots were performed three to four times and cytotoxicity assays were performed two to three times in triplicate. Error bars are SEM.

Figure 3.

Effects of constitutively active STAT3 and LGR5 overexpression on treatment resistance. A, Western blot analysis of MET and STAT3 phosphorylation in LoVo and LS180 cells transiently transfected with flag-tagged constitutively active STAT3 (STAT3-CA). B and C, Cytotoxicity of irinotecan in LoVo (B) and LS180 (C) cells transfected with vector or STAT3-CA after 4 days. D, Cytotoxicity of anti-LGR5 ADC in colorectal cancer cells transfected with vector or STAT3-CA after 4 days. E, Western blot analysis of HCT116 cells transfected with increasing amounts of LGR5. F, Cytotoxicity of irinotecan in HCT116 cells after 4 days. Western blots were performed three to four times and cytotoxicity assays were performed two to three times in triplicate. Error bars are SEM.

Close modal

Drug-induced LGR5 colorectal cancer cells have increased MET-STAT3 activation

As gene ablation of LGR5 resulted in increased MET-STAT3 activation, we next examined MET-STAT3 activity in drug-induced LGR5 colorectal cancer cells. Treatment with 10 μmol/L irinotecan or 6.5 nmol/L anti-LGR5 ADCs converted LGR5+ cells to an LGR5 state over time and increased phosphorylation of MET and STAT3 (Figs. 1AC and 4AD). Because LGR5 KD/KO cells have increased levels of active β-catenin, we tested whether drug treatment altered β-catenin activation and/or expression. As shown in Fig. 4E and F, neither irinotecan nor ADC treatment effected β-catenin in LoVo cells, and only a slight reduction in active levels was observed in LS180 cells after 72 hours. These findings demonstrate that drug-induced loss of LGR5 leads to activation of MET-STAT3, comparable with LGR5 KD/KO.

Figure 4.

Treatment-induced LGR5 colorectal cancer cells have increased MET-STAT3 activation. Western blots of MET and STAT3 phosphorylation in LoVo (A) and LS180 (B) cells treated with irinotecan (10 μmol/L) and LoVo (C) and LS180 (D) cells treated with anti-LGR5 ADC (6.5 nmol/L) at different time points. E and F, Western blots of β-catenin levels in LoVo (E) and LS180 (F) cells treated with irinotecan (10 μmol/L) or anti-LGR5 ADC (6.5 nmol/L). Experiments were performed at least three times.

Figure 4.

Treatment-induced LGR5 colorectal cancer cells have increased MET-STAT3 activation. Western blots of MET and STAT3 phosphorylation in LoVo (A) and LS180 (B) cells treated with irinotecan (10 μmol/L) and LoVo (C) and LS180 (D) cells treated with anti-LGR5 ADC (6.5 nmol/L) at different time points. E and F, Western blots of β-catenin levels in LoVo (E) and LS180 (F) cells treated with irinotecan (10 μmol/L) or anti-LGR5 ADC (6.5 nmol/L). Experiments were performed at least three times.

Close modal

LGR5 regulates sensitivity of colorectal cancer cells to MET-STAT3 inhibition

To determine whether LGR5 KD/KO cells were more dependent on MET/STAT3 signaling for survival, we tested the effects of MET and STAT3 inhibitors on viability. We found LGR5 KD/KO cells were more sensitive to crizotinib, stattic, and cryptotanshinone compared with corresponding control cells (Supplementary Fig. S3A–S3F). The IC50 values were approximately four- to eightfold lower for LoVo LGR5 KD cells and 1.5- to sixfold lower for LS180 LGR5 KO cells (Supplementary Table S2). Because loss of LGR5 also increased levels of active β-catenin (Fig. 2A and E), we evaluated the cytotoxic effect of XAV939, a tankyrase inhibitor that stimulates β-catenin degradation through Axin stabilization (24). XAV939 showed no change in efficacy with LGR5 KD/KO (Supplementary Fig. S3SG–S3H). These findings suggest MET-STAT3 activation may be integral for the enhanced survival capacity of LGR5 colorectal cancer cells.

MET-STAT3 inhibitors synergize with irinotecan and ADCs to enhance treatment efficacy in colorectal cancer cells and tumor organoids

Given that LGR5 KD/KO CRC cells are more sensitive to MET-STAT3 inhibitors, we next evaluated the efficacy of combination treatment as a strategy to overcome resistance to irinotecan or ADCs. To quantify synergistic effects of the different combinations, CI values were calculated using the Chou–Talalay method (15), where CI< 1, = 1, or >1 refers to synergistic, additive, or antagonistic activity, respectively (Supplementary Table S3). First, we evaluated MET or STAT3 inhibitors, at concentrations below the IC25 for control cells, in combination with irinotecan. In LoVo cells, we found both crizotinib and stattic synergized with irinotecan (Fig. 5A; Supplementary Table S3). Similar results were observed with combination treatment of cryptotanshinone and irinotecan (Supplementary Fig. S4). Synergism was much greater with LGR5 KD, due to increased MET-STAT3 activation (Fig. 5B; Supplementary Table S3). MET and STAT3 inhibitors also synergized with irinotecan in LS180 control and LGR5 KO cells (Fig. 5C and D; Supplementary Table S3). Synergism was less robust in LS180 LGR5 KO cells compared with LoVo LGR5 KD cells (Fig. 5B and D), which may be partially attributed to higher basal levels of phosphorylated STAT3 in LS180 cells (Supplementary Fig. S2G). We next tested whether MET and STAT3 inhibitors enhance anti-LGR5 ADCs efficacy. As shown in Fig. 5E and F and Supplementary Table S3, both inhibitors synergized with ADC treatment in LoVo and LS180 cells. However, the effect was less in LoVo cells, as they are highly sensitive to anti-LGR5 ADCs. Because we observed a robust synergistic effect with STAT3 inhibitors in combination with irinotecan, we evaluated the therapeutic efficacy of this combination using a more clinically relevant PDO model. PDOs were established from a metastatic rectal cancer with high LGR5 expression as shown by Western blot analysis (Supplementary Fig. S5A) and treated with irinotecan or ADCs alone or in combination with stattic at concentrations near IC25 (Fig. 6A and B; Supplementary Fig. S5B–S5D). Combination treatments significantly inhibited PDO growth; however, only stattic with irinotecan produced a synergistic effect (Fig. 6A and B). Together, these data suggest combination therapy with MET or STAT3 inhibitors enhances efficacy of chemotherapy and LGR5-targeted ADCs in colorectal cancer cells.

Figure 5.

MET-STAT3 inhibitors synergize with irinotecan or ADCs in vitro. Cytotoxicity of irinotecan alone or in combination with crizotinib (Crizo) or stattic in LoVo control (A) and LGR5 KD (B) cells and LS180 control (C) and LGR5 KO (D) cells. Cytotoxicity of anti-LGR5 ADCs alone or in combination with crizotinib or stattic in LoVo (E) and LS180 (F) cells. Experiments were performed two to three times in triplicates. Cytotoxicity was measured 4 days after treatment. Error bars are SEM.

Figure 5.

MET-STAT3 inhibitors synergize with irinotecan or ADCs in vitro. Cytotoxicity of irinotecan alone or in combination with crizotinib (Crizo) or stattic in LoVo control (A) and LGR5 KD (B) cells and LS180 control (C) and LGR5 KO (D) cells. Cytotoxicity of anti-LGR5 ADCs alone or in combination with crizotinib or stattic in LoVo (E) and LS180 (F) cells. Experiments were performed two to three times in triplicates. Cytotoxicity was measured 4 days after treatment. Error bars are SEM.

Close modal
Figure 6.

Combination of STAT3 inhibition with irinotecan enhances suppression of tumor growth and increases survival in colorectal cancer xenograft models. Representative bright-field microscopy images (A) and cytotoxicity assay of PDOs treated with vehicle (DMSO), 1 μmol/L irinotecan, or 65 nmol/L anti-LGR5 ADC alone or in combination with 3 μmol/L of STAT3 inhibitor, stattic (B). Experiments were performed three times. Error bars are SEM. Statistical analysis was performed using one-way ANOVA. C, Tumor growth of LoVo xenografts treated with vehicle (n = 5), 10 mg/kg stattic (n = 4), 20 mg/kg irinotecan (n = 5), or combination (n = 6). D, Tumor growth of LS180 xenografts treated with vehicle (n = 8), 10 mg/kg stattic (n = 6), 20 mg/kg irinotecan (n = 7), or combination (n = 7). Statistical analysis performed using one-way ANOVA and Dunnett multiple comparison test. *, P ≤ 0.05; **, P ≤ 0.01; ***, P ≤ 0.001 compared with vehicle unless otherwise indicated. Error bars are SD. E and F, Kaplan−Meier survival plot for LoVo (E) and LS180 (F) xenografts. G, Diagram showing potential mechanism of therapy-induced downregulation of LGR5 and activation of MET-STAT3 signaling. In the absence of drug, LGR5 mediates activation of a PTP that suppresses MET and/or STAT3 activity. When LGR5 is lost in response to therapy, the PTP is no longer active. MET phosphorylates STAT3 and active STAT3 dimerizes and translocates into the nucleus to drive transcription of genes involved in drug resistance, survival, and activation of MET via crosstalk. STAT3 represses transcription of LGR5 until treatment is terminated, and then colorectal cancer cells can transition back to an LGR5+ state.

Figure 6.

Combination of STAT3 inhibition with irinotecan enhances suppression of tumor growth and increases survival in colorectal cancer xenograft models. Representative bright-field microscopy images (A) and cytotoxicity assay of PDOs treated with vehicle (DMSO), 1 μmol/L irinotecan, or 65 nmol/L anti-LGR5 ADC alone or in combination with 3 μmol/L of STAT3 inhibitor, stattic (B). Experiments were performed three times. Error bars are SEM. Statistical analysis was performed using one-way ANOVA. C, Tumor growth of LoVo xenografts treated with vehicle (n = 5), 10 mg/kg stattic (n = 4), 20 mg/kg irinotecan (n = 5), or combination (n = 6). D, Tumor growth of LS180 xenografts treated with vehicle (n = 8), 10 mg/kg stattic (n = 6), 20 mg/kg irinotecan (n = 7), or combination (n = 7). Statistical analysis performed using one-way ANOVA and Dunnett multiple comparison test. *, P ≤ 0.05; **, P ≤ 0.01; ***, P ≤ 0.001 compared with vehicle unless otherwise indicated. Error bars are SD. E and F, Kaplan−Meier survival plot for LoVo (E) and LS180 (F) xenografts. G, Diagram showing potential mechanism of therapy-induced downregulation of LGR5 and activation of MET-STAT3 signaling. In the absence of drug, LGR5 mediates activation of a PTP that suppresses MET and/or STAT3 activity. When LGR5 is lost in response to therapy, the PTP is no longer active. MET phosphorylates STAT3 and active STAT3 dimerizes and translocates into the nucleus to drive transcription of genes involved in drug resistance, survival, and activation of MET via crosstalk. STAT3 represses transcription of LGR5 until treatment is terminated, and then colorectal cancer cells can transition back to an LGR5+ state.

Close modal

Combination treatment of irinotecan with a STAT3 inhibitor increases tumor growth inhibition and survival

We next tested the combination of a STAT3 inhibitor with irinotecan in vivo. LoVo and LS180 xenografts were each randomized into four groups once tumors were approximately 100 mm3 to 150 mm3. Mice were administered vehicle, 20 mg/kg irinotecan, 10 mg/kg stattic, or both in combination (Fig. 6C and D). Although each monotherapy slowed tumor growth in both models, combination therapy resulted in a more pronounced effect (Fig. 6C and D). Stattic led to a 30% and 15% reduction in tumor growth in LoVo and LS180 xenografts, respectively, and irinotecan resulted in 53% to 55% inhibition (Fig. 6C and D). Moreover, combination treatment significantly improved efficacy in both colorectal cancer models. Tumor growth inhibition was 82% for LoVo and 65% for LS180 xenografts (Fig. 6C and D). Similarly, combination treatment showed improved efficacy in LoVo LGR5 KD xenografts despite having increased tumor growth rate and irinotecan resistance compared with LoVo controls (Supplementary Fig. S6A). Treatment was terminated after the first animal from each vehicle group reached the tumor burden limit and then survival was monitored. Combination treatment significantly prolonged survival compared with vehicle or monotherapy in all models (Fig. 6E and F; Supplementary Fig. S6B). No significant effect on body weight nor overt toxicity was observed during treatment (Supplementary Fig. S6C and S6D). These findings indicate STAT3-targeted therapy may be highly effective when used in combination with standard chemotherapy to overcome resistance in colorectal cancer.

Aberrant activation of STAT3 has been shown in colorectal cancer and a variety of other cancer types (25, 26) and is associated with adverse clinical outcomes (27). STAT3 promotes stem cell–like characteristics and regulates diverse cellular processes including survival, proliferation, immune evasion, metastatic potential, and drug resistance (25). As mutation of STAT3 in solid tumors is infrequent, hyperactivation of STAT3 in various cancers typically occurs downstream of various growth factor and cytokine receptors (28).

In this study, we observed treatment of colorectal cancer cells with irinotecan or LGR5-targted ADC led to loss of LGR5 and increased STAT3 activation (Figs. 1 and 4). Our data suggest loss of LGR5 was due to a combination of receptor degradation and inhibition of gene expression. We previously showed ADC binding to LGR5 antigen downregulates receptor expression through cointernalization and trafficking of the complex to the lysosomes (3). Similarly, another group showed irinotecan also reduces LGR5 protein and mRNA levels (11). LGR5 KD/KO colorectal cancer cells, shown to be more resistant to standard chemotherapies (12), also showed increased active STAT3 and expression of STAT3-associated target genes (Figs. 1 and 2). This suggests LGR5 colorectal cancer cells acquire a survival advantage via the STAT3 pathway to promote a more drug-resistant phenotype (Fig. 6G). In skin cancer cells, activation of STAT3 led to a slight upregulation in LGR5 expression (29), suggesting interplay between LGR5 and STAT3 may be context dependent. The LGR5-related receptor, LGR4, was shown to be a transcriptional target of STAT3 in osteosarcoma cells and STAT3 KD reduced LGR4 expression (30). LGR4 and LGR5 have different affinities for RSPO1–4 ligands and interact with different Wnt/β-catenin coreceptors and intracellular proteins to regulate distinct functions in cancer and adult stem cells and during development (13, 31–33). Therefore, it is likely LGR4 and LGR5 play diverse roles in the regulation of STAT3 and vice versa. STAT3-β–catenin interactions and crosstalk between signaling pathways in colorectal cancer and therapy resistance has been reported (34, 35). As LGR5 KD/KO increased levels of active β-catenin (Fig. 2A and E), this suggests STAT3 may cooperate with β-catenin to promote survival of LGR5 colorectal cancer cells.

MET is an RTK that binds to hepatocyte growth factor (HGF) and is highly expressed in colorectal cancer. MET has been shown to increase at advanced stages of disease and contribute to poor prognosis (36, 37). Our studies suggest increased STAT3 activation in LGR5 colorectal cancer cells occurs downstream of MET (Fig. 2). MET and STAT3 activity has been shown to be negatively regulated by several different protein tyrosine phosphatases (PTPs), including receptor-type PTP (PTPR) and Src homology 2 domain-containing PTP (SHP) family members (17, 38). Like LGR5, PTPs play important roles in modulating Wnt signaling and cell–cell adhesion (39–41). Thus, we propose a model by which LGR5 promotes activation of a PTP that functions to inhibit MET-STAT3 (Fig. 6G, left), potentially through an interacting partner such as a Wnt coreceptor. When LGR5 expression is lost in response to drug treatment or KD/KO, the PTP remains inactive allowing for increased MET-STAT3 phosphorylation (Fig. 6G, right). We further speculate, in addition to driving transcription of genes associated with therapy resistance and survival, active STAT3 may transcriptionally repress LGR5 (Fig. 6G, right). Thus, colorectal cancer cells only transition back to an LGR5+ state once treatment is terminated. Overexpression of STAT3-CA mutant likely shifts the balance in favor of STAT3 activity, leading to downregulation of LGR5. As MET phosphorylation was decreased by treatment of LGR5 KD/KO colorectal cancer cells with STAT3 inhibitors and increased by STAT3-CA (Figs. 2 and 3), this highlights a potential STAT3-mediated positive feedback mechanism. Active STAT3 may induce transcriptional upregulation of a modulator involved in the ligand-independent activation of MET. Integrins, CD44, G protein-coupled receptors, and other RTKs have all been implicated in MET activation through crosstalk without the requirement for HGF (42). Future studies will be required to potentially identify a PTP and elucidate the exact mechanism of how loss of LGR5 leads to MET-STAT3 activation and how STAT3 promotes survival and protects colorectal cancer cells from drug-induced cell death. Furthermore, whether MET-STAT3 plays a role in the dynamic plasticity of LGR5+ CSCs during tumor progression and metastasis remains to be elucidated.

In oncogene-addicted cancer cells, including KRAS-mutant colorectal cancer cells, resistance to MEK inhibitors has also been shown to be mediated by STAT3 feedback activation (43, 44). MEK inhibition was shown to induce MET-STAT3 activity in colorectal cancer cells due to inhibition of a disintegrin and metalloproteinase domain 17 (ADAM17), resulting in decreased shedding of decoy MET (44). MET can undergo ectodomain shedding through the actions of ADAM10, ADAM17, and γ-secretases (45, 46). The ectodomain can function as a decoy receptor to block HGF activity and suppress downstream signaling. We observed both drug treatments and LGR5 KD in LoVo cells led to an increase in the mature form of MET, suggesting loss of LGR5 may lead to activation of a protease(s) that mediates posttranslational processing of MET (Figs. 2D, 4A, and C). Numerous other proteases have been implicated in the proteolytic cleavage of MET, resulting in intracellular fragments shown to have proapoptotic or proinvasive functions (47). LS180 LGR5 KO cells also showed a subtle increase in the ratio of mature MET to pro-MET expression with overall decrease in total levels of MET (Fig. 2E). We reason that because LGR5 KO increases MET phosphorylation, this may also lead to a more rapid turnover of MET compared with control cells. The role of LGR5 in regulating proteolytic cleavage and posttranslational processing of MET and its relevance in colorectal cancer requires further investigation.

Drug combination studies of MET-STAT3 inhibitors with irinotecan or ADCs showed synergism in colorectal cancer cells and PDOs (Figs. 5 and 6). The effect was markedly greater in LoVo cells compared with LS180 cells, likely due to higher levels of activated MET-STAT3 in response to irinotecan, ADCs, or LGR5 KD/KO. The extent of synergism is likely to differ between different cell lines, as other factors may contribute to their differences in intrinsic resistance. In colorectal cancer xenografts, STAT3-targeted therapy in combination with irinotecan significantly enhanced treatment efficacy and overall survival (OS; Fig. 6CF). Other groups have shown the effectiveness of STAT3 inhibitors on suppression of colorectal tumor growth and sensitization to other chemotherapies and radiotherapies (48–50). Although STAT3 inhibitors have yet to receive FDA approval, several are at various stages of clinical trials for colorectal cancer treatment (25). Our findings suggest combination therapies of MET-STAT3 inhibitors with standard chemotherapy regimens may be more effective than chemotherapy alone to overcome resistance mediated by LGR5 colorectal cancer cell populations. Thus, MET-STAT3 inhibition combined with CSC-targeting agents could have promising therapeutic potential and warrants further evaluation.

In summary, we have identified LGR5 as a novel regulator of MET-STAT3 signaling, which may play an intricate role in mediating therapy resistance in colorectal cancer. LGR5+ CSCs with tumor-initiating capacity have been shown to transition to an LGR5 state to drive metastasis and evade treatment. Therapeutic strategies to target LGR5+ CSCs alone have thus far been insufficient in eliminating colorectal cancer due to tumor heterogeneity and plasticity. This implies cotargeting both LGR5+ and LGR5 colorectal cancer cell types may be a more effective treatment. We reveal MET-STAT3 activation is triggered when colorectal cancer cells transition from an LGR5+ to LGR5 state in response to drug treatment or gene ablation. These findings indicate combination of MET–STAT3 pathway inhibition and irinotecan-based chemotherapies or LGR5-targeted ADCs may offer a promising strategy to target-refractory colorectal cancer and improve treatment efficacy and survival.

No disclosures were reported.

T.A. Posey: Conceptualization, data curation, software, formal analysis, validation, investigation, visualization, methodology, writing–original draft, writing–review, and editing. J. Jacob: Data curation, formal analysis, validation, investigation, writing–review, and editing. A. Parkhurst: Data curation, formal analysis, validation, investigation, writing–review, and editing. S. Subramanian: Data curation, validation, and investigation. L.E. Francisco: Data curation and investigation. Z. Liang: Data curation and investigation. K.S. Carmon: Conceptualization, resources, data curation, software, formal analysis, supervision, funding acquisition, validation, investigation, visualization, methodology, writing–original draft, project administration, writing–review and, editing.

This work was supported by funding from the NIH (NCI R01 CA226894, to K.S. Carmon), a fellowship of the Gulf Coast Consortia, on the Training Interdisciplinary Pharmacology Scientists program (T32 GM139801, to J. Jacob), and a Cancer Therapeutics Training Program fellowship (RP210043, to L.E. Francisco). We thank Qingyun Liu and laboratory, Eduardo Vilar-Sanchez, Venkata Lokesh Battula, and Ali Azhdarinia for helpful discussions. Schematic illustration created with BioRender.com.

The publication costs of this article were defrayed in part by the payment of publication fees. Therefore, and solely to indicate this fact, this article is hereby marked “advertisement” in accordance with 18 USC section 1734.

Note: Supplementary data for this article are available at Molecular Cancer Therapeutics Online (http://mct.aacrjournals.org/).

1.
Reya
T
,
Morrison
SJ
,
Clarke
MF
,
Weissman
IL
.
Stem cells, cancer, and cancer stem cells
.
Nature
2001
;
414
:
105
11
.
2.
Ghiaur
G
,
Gerber
J
,
Jones
RJ
.
Concise review: cancer stem cells and minimal residual disease
.
Stem Cells
2012
;
30
:
89
93
.
3.
Gong
X
,
Azhdarinia
A
,
Ghosh
SC
,
Xiong
W
,
An
Z
,
Liu
Q
, et al
.
LGR5-targeted antibody-drug conjugate eradicates gastrointestinal tumors and prevents recurrence
.
Mol Cancer Ther
2016
;
15
:
1580
90
.
4.
Hirsch
D
,
Barker
N
,
McNeil
N
,
Hu
Y
,
Camps
J
,
McKinnon
K
, et al
.
LGR5 positivity defines stem-like cells in colorectal cancer
.
Carcinogenesis
2014
;
35
:
849
58
.
5.
Junttila
MR
,
Mao
W
,
Wang
X
,
Wang
BE
,
Pham
T
,
Flygare
J
, et al
.
Targeting LGR5+ cells with an antibody-drug conjugate for the treatment of colon cancer
.
Sci Transl Med
2015
;
7
:
314ra186
.
6.
Melo
FS
,
Kurtova
AV
,
Harnoss
JM
,
Kljavin
N
,
Hoeck
JD
,
Hung
J
, et al
.
A distinct role for Lgr5+ stem cells in primary and metastatic colon cancer
.
Nature
2017
;
543
:
676
80
.
7.
Schepers
AG
,
Snippert
HJ
,
Stange
DE
,
van den Born
M
,
van Es
JH
,
van de Wetering
M
, et al
.
Lineage tracing reveals Lgr5+ stem cell activity in mouse intestinal adenomas
.
Science
2012
;
337
:
730
5
.
8.
Shimokawa
M
,
Ohta
Y
,
Nishikori
S
,
Matano
M
,
Takano
A
,
Fujii
M
, et al
.
Visualization and targeting of LGR5+ human colon cancer stem cells
.
Nature
2017
;
545
:
187
92
.
9.
Carmon
KS
,
Gong
X
,
Lin
Q
,
Thomas
A
,
Liu
Q
.
R-spondins function as ligands of the orphan receptors LGR4 and LGR5 to regulate Wnt/beta-catenin signaling
.
Proc Natl Acad Sci U S A
2011
;
108
:
11452
7
.
10.
Fumagalli
A
,
Oost
KC
,
Kester
L
,
Morgner
J
,
Bornes
L
,
Bruens
L
, et al
.
Plasticity of Lgr5-negative cancer cells drives metastasis in colorectal cancer
.
Cell Stem Cell
2020
;
26
:
569
78
.
11.
Kobayashi
S
,
Yamada-Okabe
H
,
Suzuki
M
,
Natori
O
,
Kato
A
,
Matsubara
K
, et al
.
LGR5-positive colon cancer stem cells interconvert with drug-resistant LGR5-negative cells and are capable of tumor reconstitution
.
Stem Cells
2012
;
30
:
2631
44
.
12.
Zhang
S
,
Chatterjee
T
,
Godoy
C
,
Wu
L
,
Liu
QJ
,
Carmon
KS
.
GPR56 drives colorectal tumor growth and promotes drug resistance through upregulation of MDR1 expression via a RhoA-mediated mechanism
.
Mol Cancer Res
2019
;
17
:
2196
207
.
13.
Carmon
KS
,
Gong
X
,
Yi
J
,
Wu
L
,
Thomas
A
,
Moore
CM
, et al
.
LGR5 receptor promotes cell-cell adhesion in stem cells and colon cancer cells via the IQGAP1-Rac1 pathway
.
J Biol Chem
2017
;
292
:
14989
5001
.
14.
Azhdarinia
A
,
Voss
J
,
Ghosh
SC
,
Simien
JA
,
Hernandez Vargas
S
,
Cui
J
, et al
.
Evaluation of Anti-LGR5 antibodies by ImmunoPET for imaging colorectal tumors and development of antibody-drug conjugates
.
Mol Pharm
2018
;
15
:
2448
54
.
15.
Chou
TC
,
Martin
N
.
CompuSyn for drug combinations: PC Software and User's Guide: a computer program for quantitation of synergism and antagonism in drug combinations, and the determination of IC50 and ED50 and LD50 values
.
Paramu, NJ
:
ComboSyn
;
2005
.
16.
Carpenter
RL
,
Lo
HW
.
STAT3 target genes relevant to human cancers
.
Cancers (Basel)
2014
;
6
:
897
925
.
17.
Comoglio
PM
,
Trusolino
L
,
Boccaccio
C
.
Known and novel roles of the MET oncogene in cancer: a coherent approach to targeted therapy
.
Nat Rev Cancer
2018
;
18
:
341
58
.
18.
Johnson
DE
,
O'Keefe
RA
,
Grandis
JR
.
Targeting the IL-6/JAK/STAT3 signalling axis in cancer
.
Nat Rev Clin Oncol
2018
;
15
:
234
48
.
19.
Park
OK
,
Schaefer
TS
,
Nathans
D
.
In vitro activation of Stat3 by epidermal growth factor receptor kinase
.
Proc Natl Acad Sci U S A
1996
;
93
:
13704
8
.
20.
Giordano
S
,
Di Renzo
MF
,
Narsimhan
RP
,
Cooper
CS
,
Rosa
C
,
Comoglio
PM
.
Biosynthesis of the protein encoded by the c-met proto-oncogene
.
Oncogene
1989
;
4
:
1383
8
.
21.
Mondino
A
,
Giordano
S
,
Comoglio
PM
.
Defective posttranslational processing activates the tyrosine kinase encoded by the MET proto-oncogene (hepatocyte growth factor receptor)
.
Mol Cell Biol
1991
;
11
:
6084
92
.
22.
Schust
J
,
Sperl
B
,
Hollis
A
,
Mayer
TU
,
Berg
T
.
Stattic: a small-molecule inhibitor of STAT3 activation and dimerization
.
Chem Biol
2006
;
13
:
1235
42
.
23.
Shin
DS
,
Kim
HN
,
Shin
KD
,
Yoon
YJ
,
Kim
SJ
,
Han
DC
, et al
.
Cryptotanshinone inhibits constitutive signal transducer and activator of transcription 3 function through blocking the dimerization in DU145 prostate cancer cells
.
Cancer Res
2009
;
69
:
193
202
.
24.
Huang
SM
,
Mishina
YM
,
Liu
S
,
Cheung
A
,
Stegmeier
F
,
Michaud
GA
, et al
.
Tankyrase inhibition stabilizes axin and antagonizes Wnt signalling
.
Nature
2009
;
461
:
614
20
.
25.
Gargalionis
AN
,
Papavassiliou
KA
,
Papavassiliou
AG
.
Targeting STAT3 signaling pathway in colorectal cancer
.
Biomedicines
2021
;
9
:
1016
.
26.
Bowman
T
,
Garcia
R
,
Turkson
J
,
Jove
R
.
STATs in oncogenesis
.
Oncogene
2000
;
19
:
2474
88
.
27.
Morikawa
T
,
Baba
Y
,
Yamauchi
M
,
Kuchiba
A
,
Nosho
K
,
Shima
K
, et al
.
STAT3 expression, molecular features, inflammation patterns, and prognosis in a database of 724 colorectal cancers
.
Clin Cancer Res
2011
;
17
:
1452
62
.
28.
Huynh
J
,
Chand
A
,
Gough
D
,
Ernst
M
.
Therapeutically exploiting STAT3 activity in cancer - using tissue repair as a road map
.
Nat Rev Cancer
2019
;
19
:
82
96
.
29.
Jia
J
,
Shi
Y
,
Yan
B
,
Xiao
D
,
Lai
W
,
Pan
Y
, et al
.
LGR5 expression is controled by IKKalpha in basal cell carcinoma through activating STAT3 signaling pathway
.
Oncotarget
2016
;
7
:
27280
94
.
30.
Liu
J
,
Wei
W
,
Guo
CA
,
Han
N
,
Pan
JF
,
Fei
T
, et al
.
Stat3 upregulates leucine-rich repeat-containing g protein-coupled receptor 4 expression in osteosarcoma cells
.
Biomed Res Int
2013
;
2013
:
310691
.
31.
Carmon
KS
,
Gong
X
,
Yi
J
,
Thomas
A
,
Liu
Q
.
RSPO-LGR4 functions via IQGAP1 to potentiate Wnt signaling
.
Proc Natl Acad Sci U S A
2014
;
111
:
E1221
9
.
32.
Kinzel
B
,
Pikiolek
M
,
Orsini
V
,
Sprunger
J
,
Isken
A
,
Zietzling
S
, et al
.
Functional roles of Lgr4 and Lgr5 in embryonic gut, kidney and skin development in mice
.
Dev Biol
2014
;
390
:
181
90
.
33.
Park
S
,
Wu
L
,
Tu
J
,
Yu
W
,
Toh
Y
,
Carmon
KS
, et al
.
Unlike LGR4, LGR5 potentiates Wnt-beta-catenin signaling without sequestering E3 ligases
.
Sci Signal
2020
;
13
:
eaaz4051
.
34.
Phesse
TJ
,
Buchert
M
,
Stuart
E
,
Flanagan
DJ
,
Faux
M
,
Afshar-Sterle
S
, et al
.
Partial inhibition of gp130-Jak-Stat3 signaling prevents Wnt-beta-catenin-mediated intestinal tumor growth and regeneration
.
Sci Signal
2014
;
7
:
ra92
.
35.
Sinnberg
T
,
Makino
E
,
Krueger
MA
,
Velic
A
,
Macek
B
,
Rothbauer
U
, et al
.
A nexus consisting of beta-catenin and Stat3 attenuates BRAF inhibitor efficacy and mediates acquired resistance to vemurafenib
.
EBioMedicine
2016
;
8
:
132
49
.
36.
Di Renzo
MF
,
Olivero
M
,
Giacomini
A
,
Porte
H
,
Chastre
E
,
Mirossay
L
, et al
.
Overexpression and amplification of the met/HGF receptor gene during the progression of colorectal cancer
.
Clin Cancer Res
1995
;
1
:
147
54
.
37.
Takeuchi
H
,
Bilchik
A
,
Saha
S
,
Turner
R
,
Wiese
D
,
Tanaka
M
, et al
.
c-MET expression level in primary colon cancer: a predictor of tumor invasion and lymph node metastases
.
Clin Cancer Res
2003
;
9
:
1480
8
.
38.
Kim
M
,
Morales
LD
,
Jang
IS
,
Cho
YY
,
Kim
DJ
.
Protein tyrosine phosphatases as potential regulators of STAT3 signaling
.
Int J Mol Sci
2018
;
19
:
2708
.
39.
Chang
LS
,
Kim
M
,
Glinka
A
,
Reinhard
C
,
Niehrs
C
.
The tumor suppressor PTPRK promotes ZNRF3 internalization and is required for Wnt inhibition in the Spemann organizer
.
Elife
2020
;
9
:
e51248
.
40.
Martin
E
,
Agazie
YM
.
SHP2 potentiates the oncogenic activity of beta-catenin to promote triple-negative breast cancer
.
Mol Cancer Res
2021
;
19
:
1946
56
.
41.
Fuchs
M
,
Muller
T
,
Lerch
MM
,
Ullrich
A
.
Association of human protein-tyrosine phosphatase kappa with members of the armadillo family
.
J Biol Chem
1996
;
271
:
16712
9
.
42.
Lai
AZ
,
Abella
JV
,
Park
M
.
Crosstalk in Met receptor oncogenesis
.
Trends Cell Biol
2009
;
19
:
542
51
.
43.
Lee
HJ
,
Zhuang
G
,
Cao
Y
,
Du
P
,
Kim
HJ
,
Settleman
J
.
Drug resistance via feedback activation of Stat3 in oncogene-addicted cancer cells
.
Cancer Cell
2014
;
26
:
207
21
.
44.
Van Schaeybroeck
S
,
Kalimutho
M
,
Dunne
PD
,
Carson
R
,
Allen
W
,
Jithesh
PV
, et al
.
ADAM17-dependent c-MET-STAT3 signaling mediates resistance to MEK inhibitors in KRAS mutant colorectal cancer
.
Cell Rep
2014
;
7
:
1940
55
.
45.
Kopitz
C
,
Gerg
M
,
Bandapalli
OR
,
Ister
D
,
Pennington
CJ
,
Hauser
S
, et al
.
Tissue inhibitor of metalloproteinases-1 promotes liver metastasis by induction of hepatocyte growth factor signaling
.
Cancer Res
2007
;
67
:
8615
23
.
46.
Schelter
F
,
Kobuch
J
,
Moss
ML
,
Becherer
JD
,
Comoglio
PM
,
Boccaccio
C
, et al
.
A disintegrin and metalloproteinase-10 (ADAM-10) mediates DN30 antibody-induced shedding of the met surface receptor
.
J Biol Chem
2010
;
285
:
26335
40
.
47.
Fernandes
M
,
Duplaquet
L
,
Tulasne
D
.
Proteolytic cleavages of MET: the divide-and-conquer strategy of a receptor tyrosine kinase
.
BMB Rep
2019
;
52
:
239
49
.
48.
Liu
Z
,
Wang
H
,
Guan
L
,
Lai
C
,
Yu
W
,
Lai
M
.
LL1, a novel and highly selective STAT3 inhibitor, displays anti-colorectal cancer activities in vitro and in vivo
.
Br J Pharmacol
2020
;
177
:
298
313
.
49.
Spitzner
M
,
Roesler
B
,
Bielfeld
C
,
Emons
G
,
Gaedcke
J
,
Wolff
HA
, et al
.
STAT3 inhibition sensitizes colorectal cancer to chemoradiotherapy in vitro and in vivo
.
Int J Cancer
2014
;
134
:
997
1007
.
50.
Wei
N
,
Li
J
,
Fang
C
,
Chang
J
,
Xirou
V
,
Syrigos
NK
, et al
.
Targeting colon cancer with the novel STAT3 inhibitor bruceantinol
.
Oncogene
2019
;
38
:
1676
87
.