About 40% of colorectal cancers have mutations in KRAS accompanied by downstream activation of MAPK signaling, which promotes tumor invasion and progression. Here, we report that MAPK signaling shows strong intratumoral heterogeneity and unexpectedly remains regulated in colorectal cancer irrespective of KRAS mutation status. Using primary colorectal cancer tissues, xenograft models, and MAPK reporter constructs, we showed that tumor cells with high MAPK activity resided specifically at the leading tumor edge, ceased to proliferate, underwent epithelial–mesenchymal transition (EMT), and expressed markers related to colon cancer stem cells. In KRAS-mutant colon cancer, regulation of MAPK signaling was preserved through remaining wild-type RAS isoforms. Moreover, using a lineage tracing strategy, we provide evidence that high MAPK activity marked a progenitor cell compartment of growth-fueling colon cancer cells in vivo. Our results imply that differential MAPK signaling balances EMT, cancer stem cell potential, and tumor growth in colorectal cancer. Cancer Res; 77(7); 1763–74. ©2017 AACR.

Colorectal cancer derives from normal colonic mucosa by stepwise accumulation of mutations that transform normal epithelial cells into malignant tumor cells (1). Aside from APC mutations, which are found in most colorectal cancers, activating mutations in codons 12 or 13 of KRAS are present in about 40% of these tumors, and they occur early within the malignant transformation process (2).

The GTPase KRAS is a component of MAPK signaling that communicates signals from growth factor receptors into the cell nucleus (3). In KRAS wild-type tumor cells, binding of EGF to its receptor (EGFR) causes GTP-loading of KRAS, which activates RAF to phosphorylate MAPK/ERK-activating kinase (MEK), which in turn phosphorylates ERK. Phosphorylated ERK (p-ERK) then causes expression of MAPK target genes through ELK1 and AP1 transcription factors that promote malignant traits of tumor progression including invasion and metastasis (4, 5). However, in contrast to wild-type KRAS, mutated KRAS binds GTP permanently and independently of EGFR stimulation (6). MAPK signaling therefore is thought to be constitutively active in KRAS-mutated colorectal cancer, promoting tumor progression independently of external EGFR stimulation (7, 8). This translates into clinical application, as in contrast to wild-type cases, KRAS-mutated colorectal cancer lacks significant treatment response to upstream MAPK inhibition when targeting the EGFR with antibody drugs, such as cetuximab (9, 10).

Signaling pathways that are affected by activating mutations can remain regulated in colorectal cancer and contribute to distinct phenotypes of tumor cell subpopulations. This is exemplified by heterogeneous WNT activity in colorectal cancer despite inactivating APC mutations in most of these tumors (11, 12). Colon cancer cells with high WNT activity were more invasive and debated to have higher tumor-initiating cancer stem cell potential, when compared with subpopulations with lower WNT activity that showed a more differentiated tumor cell phenotype (13, 14). In this context, we previously demonstrated that MAPK signaling regulates WNT signaling and suggested a contribution of MAPK signaling to phenotypic tumor cell heterogeneity in colorectal cancer (15). However, the extent of differential MAPK activity and its significance in KRAS wild-type and mutant colorectal cancer has remained largely unknown.

Here, we investigate MAPK signaling in colorectal cancers with and without KRAS mutations and exploit its relevance for phenotypically distinct tumor cell subpopulations. We demonstrate that MAPK signaling remains strongly regulated, even in colorectal cancer with activating KRAS mutations. Furthermore, we show that differential MAPK signaling balances invasive and proliferating tumor cell phenotypes. Finally, using a lineage tracing strategy, we provide evidence that high MAPK activity marks a progenitor cell compartment of growth-fueling colon cancer cells in vivo.

Clinical samples and KRAS mutational analysis

Paraffin-embedded colorectal cancer specimens from patients that underwent surgical resection at the University of Munich (LMU; Munich, Germany) were drawn from the archives of the Institute of Pathology. For KRAS mutational testing, tumor tissue was scraped from deparaffinized tissue sections under microscopic control using sterile scalpel blades, and tumor DNA was extracted with QIAamp DNA Micro Kits. KRAS exon 2 was then analyzed by pyrosequencing on a PyroMark Q24 Advanced instrument (Qiagen) with primers 5′-NNNGGCCTGCTGAAAATGACTGAA-3′ and 5′-Biotin-TTAGCTGTATCGTCAAGGCACTCT-3′ for amplification, and 5′-TGTGGTAGTTGGAGCT-3′ for sequencing, as previously described (16). KRAS wild-type and mutated colorectal cancers then were grade- and stage-matched, resulting in a final collection of 160 cases, half of which had KRAS exon 2 mutations and the other half of which were KRAS wild-type (Supplementary Table S1). Specimens and data were anonymized, and the need for consent was waived by the institutional ethics committee of the Medical Faculty of the LMU.

Immunohistochemistry, laser microdissection, and immunofluorescence

For immunohistochemistry, 5-μm tissue sections of colorectal cancer samples were deparaffinized, incubated with primary antibodies listed in Supplementary Table S2, and stained on a Ventana Benchmark XT autostainer with an ultraView Universal DAB Detection Kit (Ventana Medical Systems). For mutational analyses of tumor cell subpopulations, 1000 positive or negative tumor cells were laser-microdissected from p-ERK stained slides with a PALM system (Zeiss). DNA from these tumor cell subpopulations was then separately isolated and subjected to KRAS mutation analysis as described above. For immunofluorescence, sections of colorectal cancer cases and xenografts were deparaffinized and antigens were retrieved in Target Retrieval Solution (Dako) or Epitope Retrieval Solution pH8 (Leica) for 20 minutes in a microwave oven. Spheroid cultures were fixed in 4% paraformaldehyde and 5% sucrose in PBS for 20 minutes, permeabilized in 1% Triton X-100 for 10 minutes, and blocked with 3% BSA in PBS for 30 minutes at room temperature. Sections or spheroids were then incubated with primary antibodies listed in Supplementary Table S2. Secondary Alexa Fluor 488- or 555-conjugated antibodies (Invitrogen) were used for visualization, and nuclei were counterstained with DAPI (Vector Laboratories). Confocal fluorescence images were taken on an LSM 700 laser scanning microscope using the ZEN software (Zeiss). Colocalization of fluorescence signals was measured using Volocity 6.1.1 software (PerkinElmer) and plotted as percentage values of maximum fluorescence intensity.

Lentiviral vectors

All template plasmids were obtained through Addgene (www.addgene.org). pLenti SRE-GFP was constructed by replacing the PGK promoter in pLenti PGK-GFP (pRRLSIN.cPPT.PGKGFP.WPRE, a gift from Didier Trono) with a serum response element optimal promoter cassette (3× SRE), containing synthetic GGATGTCCATATTAGGACATCT-binding sites (17). We then removed the internal ribosomal entry site (IRES) of pBMN-I-GFP (a gift from Garry Nolan) using Not1 and Nco1 restriction enzymes, amplified CreERT2 from pCAG-CreERT2 (18) by PCR, and inserted both into the Sal1 sites of pLenti PGK-GFP and pLenti SRE-GFP, yielding pLenti PGK-GFP-CreERT2 and pLenti SRE-GFP-CreERT2. For the Cre-sensitive recombination vector pLenti lox-mCh-LacZ, we replaced the GFP cassette of pLenti PGK-GFP with a synthetic sequence containing 2 lox2272 and loxP sites (19) as well as EcoRV and Hpa1 restriction sites. PCR-amplified H2BmCherry from PGK-H2BmCherry (20) and LacZ from LV-Lac (21) were then inserted or reversely inserted into EcoRV and Hpa1 sites, respectively. For construction of pLenti CMVTRE3G-caMEK Puro, we amplified constitutively active MEK1 (caMEK) from pCScherryActMEK (22) by PCR and inserted it between BamH1 and Xba1 restriction sites of pLenti CMVTRE3G eGFP Puro (a gift from Eric Campeau), replacing eGFP by caMEK. Modified vector elements were verified by restriction analysis and sequencing.

Cell and spheroid culture, lentiviral transductions, and single-cell subcloning

SW1222 cells were a gift from the Ludwig Institute for Cancer Research (New York, NY), and primary colon cancer cells (P-Tu) were obtained from HTCR (Munich, Germany). Other cell lines were from the ATCC. All cell lines were obtained between 2009 and 2014 and authenticated using short-tandem repeat profiling. Cell lines were cultured in DMEM containing 10% FBS, 100 U/mL penicillin, and 0.1 mg/mL streptomycin (Biochrom). P-Tu colon cancer cells were grown as spheroids in StemPro hESC SFM medium supplemented with 20 ng/mL EGF and 10 ng/mL bFGF (Life Technologies) in ultra-low attachment flasks (Corning). For in vitro MAPK stimulation experiments, cells were starved for 24 hours in serum-free medium, treated with 10 μg/mL cetuximab (Merck Serono) or PBS, and then with 40 ng/mL EGF (Invitrogen) for 10 minutes before protein isolation. To assess phenotypic effects, tumor spheroids of P-Tu colon cancer cells were mixed with 25-μL rat tail collagen I (Santa Cruz), placed in 8-well culture slides (Falcon), incubated for 4 days in serum-free DMEM with or without 100 nmol/L oleoyl-l-alpha-lysophosphatidic acid (LPA; Santa Cruz), and subjected to immunofluorescence as described above. For transductions, lentivirus was produced in HEK293 cells by co-transfection with lentiviral vector, pCMV-dR8.91, and pMD2.G as previously described (15). Virus containing medium was passed through 0.45-μm filters (Millipore), mixed 1:1 with DMEM, and used to infect T84 or P-Tu colon cancer cells in the presence of 8 μg/mL polybrene (Sigma-Aldrich). Transduced cells then were single cell sorted into 96-well plates on a FACSAria III instrument (BD Biosciences) and expanded. For in vitro induction, pLenti CMV rtTA3G Blast (a gift from Dominic Esposito) and pLenti CMVTRE3G-caMEK Puro–transduced T84 and P-Tu colon cancer cells were treated with 10 ng/mL doxycycline in PBS or with PBS alone before protein isolation and immunoblotting.

Immunoblotting and RAS-GTP assays

Immunoblotting was done using whole-cell lysates of colon cancer cells supplemented with protease and phosphatase inhibitors (Roche). For detection of active RAS-GTP, 500 μg of cell lysates was incubated with GST-Raf1-RBD (Thermo Fisher) for 60 minutes at 4°C, washed 3 times with washing buffer (25 mmol/L Tris-HCl, pH 7.2, 150 mmol/L NaCl, 5 mmol/L MgCl2, 1% NP-40, and 5% glycerol), eluted with SDS sample buffer [25 mmol/L Tris-HCl, pH 6.8, 2% glycerol, 4% SDS (w/v), and 0.05% bromophenol blue], and heated for 5 minutes at 95°C. Proteins were separated by SDS-PAGE, transferred onto polyvinylidene difluoride (PVDF) membranes (Merck Millipore), and incubated with primary antibodies listed in Supplementary Table S2. Bands were visualized using HRP‐conjugated secondary mouse (Promega) or rabbit (Sigma) antibodies and chemiluminescent HRP Substrate (Millipore).

Tumor xenografts and in vivo treatments

Mouse experiments were reviewed and approved by Regierung von Oberbayern. Single clone–expanded T84 or P-Tu colon cancer cells, either native or carrying the respective lentiviral constructs described above, were suspended in 100 μL of a 1:1 mixture of PBS and growth factor–depleted Matrigel (Corning) and injected subcutaneously into 6- to 8-week-old NOD/SCID mice (NOD.CB17-Prkdcscid, The Jackson Laboratory) for xenograft formation. In pLenti CMV rtTA3G Blast and pLenti CMVTRE3G-caMEK Puro–transduced T84 colon cancer cell–derived xenografts, caMEK expression was induced with 1 mg doxycycline p.o. For lineage tracing, pLenti lox-mCh-LacZ and pLenti PGK-GFP-CreERT2 or pLenti SRE-GFP-CreERT2–transduced P-Tu colon cancer cell–derived xenografts were treated once with 3 mg tamoxifen (Sigma Aldrich) by intraperitoneal injection. Treatments were done when tumor diameters reached 7 mm. Nontreated tumors were included as controls. Mice were sacrificed and tumors were removed 4 days after caMEK induction, and 2 or 21 days for short- and long-term lineage tracing experiments, respectively. All xenograft tumors were formalin-fixed and paraffin-embedded for further analyses.

MAPK activity is heterogeneous in colorectal cancer

To characterize MAPK activity in colorectal cancer, we stained a collection of 160 cases, half of which were KRAS wild-type and half of which had activating KRAS mutations (Supplementary Table S1), for p-ERK. Cases with detectable p-ERK (59.4%) generally showed a heterogeneous staining pattern and were composed of p-ERK–positive and -negative tumor cell subpopulations. In specific, p-ERK strongly marked colon cancer cells at the infiltrative tumor edge, including tumor cells that invaded the stroma by apparently detaching from the gland forming tumor mass (Fig. 1A). This staining pattern was observed in KRAS wild-type and mutant colon cancer cases (Fig. 1B). To exclude that heterogeneous p-ERK expression in KRAS-mutant cancers was caused by a mixture of KRAS-mutant and wild-type tumor cell subclones, in 3 of these cases, we then microdissected p-ERK–positive and -negative colon cancer cells from tumor edge and tumor center, respectively, and generally found an identical KRAS mutation status in both subpopulations (data not shown). However, as p-ERK staining quality was variable among tumor specimens and appeared to depend on tissue quality (data not shown), we additionally assessed MAPK activity through staining for FRA1, an ERK-dependent component of the AP1 transcription factor complex that typically is upregulated in colorectal cancer (23, 24). All cases with detectable FRA1 expression (75.0%) showed a heterogeneous distribution with clearly predominant expression in infiltrative tumor cells at the leading tumor edge, where it colocalized with p-ERK staining, whereas this distribution again was irrespective of the tumors' KRAS mutation status (Fig. 1A and B). Furthermore, we injected cells of a KRAS wild-type primary colon cancer (P-Tu) or KRAS-mutant T84 colon cancer cells into NOD/SCID mice for xenograft formation. In both cases, adenocarcinomas formed that showed heterogeneous expression of p-ERK and FRA1 with predominant staining at the tumor edge (Fig. 1C). We then constructed a lentiviral MAPK reporter with GFP expression under control of multimerized serum response elements (pLenti SRE-GFP), expanded single-cell clones of transduced P-TuSRE-GFP and T84SRE-GFP colon cancer cells, and subcutaneously injected them into NOD/SCID mice. In the resulting xenograft tumors, GFP expression was strongly heterogeneous and again predominantly localized at the tumor edge (Fig. 1D). These findings demonstrated that MAPK signaling is strongly regulated in colon cancers with and without activating KRAS mutations by locoregional cues, with colon cancer cells at the infiltrative tumor edge displaying highest MAPK activity levels.

Figure 1.

Heterogeneous MAPK activity in colorectal cancer. A, Representative immunostainings on serial tumor sections for p-ERK and FRA1 in human colorectal cancer specimens with indicated KRAS mutation status. WT, KRAS wild-type. Arrowheads, positive staining of tumor cells at the leading tumor edge. Arrows, tumor cells without staining. Scale bars, 100 μm. B, Frequencies of colorectal cancers (n = 160) without detectable staining (NEG) or heterogeneous staining (HET) for p-ERK and FRA1. KRAS mutation status is indicated by distinct colors. WT, KRAS wild-type. C, Immunostainings of primary (P-Tu) or T84 colon cancer xenografts (n ≥ 3) for p-ERK and FRA1. Arrowheads, positive staining. Scale bars, 50 μm. D, Top, lentiviral MAPK reporter pLenti SRE-GFP. LTR, long terminal repeat; PRE, posttranscriptional regulatory element; SRE, serum response element. Middle, expansion of single transduced colon cancer cells and injection into NOD/SCID mice for xenograft formation. Bottom, immunostainings of pLenti SRE-GFP–transduced P-Tu or T84 single cell–derived xenografts (n ≥ 3) for GFP. Arrowheads, positive staining. Scale bars, 50 μm.

Figure 1.

Heterogeneous MAPK activity in colorectal cancer. A, Representative immunostainings on serial tumor sections for p-ERK and FRA1 in human colorectal cancer specimens with indicated KRAS mutation status. WT, KRAS wild-type. Arrowheads, positive staining of tumor cells at the leading tumor edge. Arrows, tumor cells without staining. Scale bars, 100 μm. B, Frequencies of colorectal cancers (n = 160) without detectable staining (NEG) or heterogeneous staining (HET) for p-ERK and FRA1. KRAS mutation status is indicated by distinct colors. WT, KRAS wild-type. C, Immunostainings of primary (P-Tu) or T84 colon cancer xenografts (n ≥ 3) for p-ERK and FRA1. Arrowheads, positive staining. Scale bars, 50 μm. D, Top, lentiviral MAPK reporter pLenti SRE-GFP. LTR, long terminal repeat; PRE, posttranscriptional regulatory element; SRE, serum response element. Middle, expansion of single transduced colon cancer cells and injection into NOD/SCID mice for xenograft formation. Bottom, immunostainings of pLenti SRE-GFP–transduced P-Tu or T84 single cell–derived xenografts (n ≥ 3) for GFP. Arrowheads, positive staining. Scale bars, 50 μm.

Close modal

MAPK signaling is regulated through wild-type RAS isoforms in colorectal cancer

To determine the underlying regulations, we next analyzed MAPK signaling upon stimulation by growth factor receptors in colon cancer cells with and without KRAS mutations in vitro. Within the MAPK pathway, stimulation of EGFR is transduced through the RAS/RAF/MEK/ERK signaling cascade. Expectedly, stimulation of KRAS wild-type colon cancer cells with EGF caused strong phosphorylation of ERK, indicating pathway activation, whereas blocking EGFR with cetuximab prevented this effect (Fig. 2A). Surprisingly, we observed exactly the same response upon EGF treatment of KRAS-mutant colon cancer cell lines that also was abolished by cetuximab (Fig. 2B). Moreover, in both KRAS wild-type and mutant colon cancer cells, stimulation by EGF led to phosphorylation of EGFR and MEK in addition to ERK, indicating full pathway response (Fig. 2C). These findings suggested that MAPK signaling remains responsive to external stimulation of EGFR in KRAS wild-type and mutant colon cancer cells. To shed light on the underlying mechanism, we then used GTP pull-down assays that expectedly showed predominant GTP loading of KRAS and NRAS in KRAS wild-type colon cancer cells (Fig. 2D). In contrast, KRAS-mutated colon cancer cells showed constitutively GTP-bound KRAS, whereas EGF stimulation caused additional GTP loading of either wild-type NRAS or HRAS or both (Fig. 2D). Collectively, these findings demonstrate that remaining wild-type RAS isoforms contribute to sustained regulation of MAPK signaling in KRAS-mutant colon cancers.

Figure 2.

RAS-mediated MAPK pathway regulation in colorectal cancer cells. A–C, Colon cancer cell lines were serum-starved, treated with 10 μg/mL cetuximab or PBS for 2 hours, and then stimulated with 40 ng/mL EGF or not stimulated. Immunoblotting on whole-cell lysates for indicated proteins 10 minutes after stimulation in KRAS wild-type (WT) cell lines (A), KRAS-mutated cell lines (B), and KRAS WT or mutated cell lines (C). D, RAS activity in cell lines with indicated KRAS mutation status was determined by RAS-GTP pull-down assays and immunoblotting for indicated proteins. n ≥ 3.

Figure 2.

RAS-mediated MAPK pathway regulation in colorectal cancer cells. A–C, Colon cancer cell lines were serum-starved, treated with 10 μg/mL cetuximab or PBS for 2 hours, and then stimulated with 40 ng/mL EGF or not stimulated. Immunoblotting on whole-cell lysates for indicated proteins 10 minutes after stimulation in KRAS wild-type (WT) cell lines (A), KRAS-mutated cell lines (B), and KRAS WT or mutated cell lines (C). D, RAS activity in cell lines with indicated KRAS mutation status was determined by RAS-GTP pull-down assays and immunoblotting for indicated proteins. n ≥ 3.

Close modal

Colorectal cancer cells with high MAPK activity have a distinct phenotype

To learn about the relevance of regulated MAPK signaling in colon cancer, we further characterized the phenotype of colon cancer cell subpopulations with low and high pathway activities. We examined colon cancer specimens with and without KRAS mutations by double immunofluorescence. Tumor cells with high p-ERK expression at the leading tumor edge showed concomitant overexpression of nuclear β-catenin (Fig. 3A and B), indicating coincident activation of MAPK and WNT signaling. Furthermore, as indicated by Ki67 staining, proliferation was strongly reduced in colon cancer cells with high p-ERK staining when compared with gland-forming tumor cells with lower or absent p-ERK staining (Fig. 3C and D). At the same time, highly p-ERK–positive tumor cells exhibited a significantly reduced expression of the epithelial cell adhesion molecule E-cadherin (Fig. 3E and F) and increased expression of LAMC2 (Fig. 3G and H), both indicating epithelial–mesenchymal transition (EMT) in colorectal cancer. Taken together, tumor cells with high MAPK activity at the infiltrative tumor edge of colorectal cancer displayed increased WNT signaling, decreased proliferation, and had undergone an EMT.

Figure 3.

Phenotype of colorectal cancer cells with differential MAPK activity. A, C, E, and G, Representative immunofluorescence for indicated proteins in human colorectal cancer specimens. Arrowheads, p-ERK–positive tumor cells at the leading tumor edge. Arrows, p-ERK–negative gland-forming tumor cells. Scale bars, 100 μm. B, D, F, and H, Quantification of coimmunofluorescence signals for indicated proteins. Percentage values of relative fluorescence intensity (% RFI) for individual tumor cells (n ≥ 700) of different colorectal cancer samples (n ≥ 7) are shown. P values are results of linear regression analyses.

Figure 3.

Phenotype of colorectal cancer cells with differential MAPK activity. A, C, E, and G, Representative immunofluorescence for indicated proteins in human colorectal cancer specimens. Arrowheads, p-ERK–positive tumor cells at the leading tumor edge. Arrows, p-ERK–negative gland-forming tumor cells. Scale bars, 100 μm. B, D, F, and H, Quantification of coimmunofluorescence signals for indicated proteins. Percentage values of relative fluorescence intensity (% RFI) for individual tumor cells (n ≥ 700) of different colorectal cancer samples (n ≥ 7) are shown. P values are results of linear regression analyses.

Close modal

MAPK signaling regulates EMT in colorectal cancer

To determine the impact of MAPK signaling on the observed tumor cell phenotype, we transduced KRAS wild-type and mutant colon cancer cells with 2 lentiviral vectors encoding a doxycycline-inducible constitutively active MEK1 (caMEK-Tet-On, Fig. 4A). Treatment of P-TucaMEK-Tet-On and T84caMEK-Tet-On colon cancer cells with doxycycline caused strong upregulation of p-ERK and FRA1, indicating the expected activation of downstream MAPK signaling (Fig. 4B). Importantly, caMEK induction also caused elevated expression of active β-catenin (ABC), indicating increased WNT activity, as well as induction of SNAI1 and a decrease of E-cadherin, both indicating an EMT phenotype. Furthermore, induction of caMEK caused pronounced expression of the putative colon cancer stem cell antigens CD44, ASCL2, and EPHB2. We then additionally transduced T84caMEK-Tet-On colon cancer cells with a pLenti SRE-GFP reporter (Fig. 1D), yielding T84SRE-GFP/caMEK-Tet-On, expanded single-cell clones, injected them into NOD/SCID mice for xenograft formation, and examined the effects of caMEK induction by doxycycline in vivo. Similar to our results in cell culture, doxycycline treatment induced the expression of FRA1 and GFP, indicating MAPK activation, increased β-catenin and SNAI1 expression, and repressed expression of E-cadherin (Fig. 4C). To further evaluate effects of MAPK on EMT, we then stimulated in vitro spheroids of P-TuSRE-GFP colon cancer cells with the MAPK activator LPA. While tumor spheroids without stimulation formed rounded edges, LPA stimulation caused tumor cells to form radial cytoplasmic extensions with spindle cell morphology (Fig. 4D). Moreover, we found an upregulation of FN1 and LAMC2 with concomitant downregulation of E-cadherin upon LPA stimulation (Fig. 4D). Collectively, these findings demonstrated induction of EMT and expression of putative cancer stem cell antigens upon MAPK stimulation in KRAS wild-type and mutant colon cancer cells.

Figure 4.

Effects of MAPK overactivation in colorectal cancer cells. A, Lentiviral vectors for doxycycline-inducible overexpression (Tet-On) of constitutively active MEK (caMEK). BlastR, blasticidin resistance; CMV, cytomegalovirus promoter; LTR, long terminal repeat; PRE, posttranscriptional regulatory element; PuroR, puromycin resistance. B, Immunoblotting of indicated proteins on whole-cell lysates of pLenti CMV rtTA3G Blast and pLenti CMVTRE3G-caMEK Puro primary (P-TucaMEK-Tet-On) and T84caMEK-Tet-On colon cancer cells at indicated time points after doxycycline (DOX) stimulation. n ≥ 3. C, Immunofluorescence for indicated proteins in T84SRE-GFP/caMEK-Tet-On xenografts (n ≥ 3), 4 days after treatment with doxycycline (+DOX) or without doxycycline treatment (−DOX). Scale bars, 50 μm. D, Immunofluorescence for indicated proteins in 3-dimensional spheroids of P-TuSRE-GFP colon cancer cells (n ≥ 3) with or without LPA stimulation. Arrowheads, radial cytoplasmic extensions upon LPA stimulation. Scale bars, 50 μm. E-cad, E-cadherin.

Figure 4.

Effects of MAPK overactivation in colorectal cancer cells. A, Lentiviral vectors for doxycycline-inducible overexpression (Tet-On) of constitutively active MEK (caMEK). BlastR, blasticidin resistance; CMV, cytomegalovirus promoter; LTR, long terminal repeat; PRE, posttranscriptional regulatory element; PuroR, puromycin resistance. B, Immunoblotting of indicated proteins on whole-cell lysates of pLenti CMV rtTA3G Blast and pLenti CMVTRE3G-caMEK Puro primary (P-TucaMEK-Tet-On) and T84caMEK-Tet-On colon cancer cells at indicated time points after doxycycline (DOX) stimulation. n ≥ 3. C, Immunofluorescence for indicated proteins in T84SRE-GFP/caMEK-Tet-On xenografts (n ≥ 3), 4 days after treatment with doxycycline (+DOX) or without doxycycline treatment (−DOX). Scale bars, 50 μm. D, Immunofluorescence for indicated proteins in 3-dimensional spheroids of P-TuSRE-GFP colon cancer cells (n ≥ 3) with or without LPA stimulation. Arrowheads, radial cytoplasmic extensions upon LPA stimulation. Scale bars, 50 μm. E-cad, E-cadherin.

Close modal

Lineage tracing reveals a progenitor cell phenotype of colon cancer cells with high MAPK activity

Ectopic activation of MAPK signaling caused increased WNT signaling and expression of markers that had been previously related to tumor-initiating colon cancer stem cells (14, 25–27). To test whether colon cancer cells with high MAPK activity represent a progenitor cell compartment within growing tumors in vivo, we developed a Cre-lox–based lineage tracing system that allowed to genetically label and follow tumor cells and their progeny over time. We designed 2 lentiviral vectors that expressed GFP and CreERT2 under control of the MAPK-sensitive serum response element (pLenti SRE-GFP-CreERT2) or of a ubiquitously active PGK promoter (pLenti PGK-GFP-CreERT2; Fig. 5A). We then created a second lentiviral vector that upon Cre activation irreversibly recombined with a switch from mCherry to LacZ expression (pLenti lox-mCh-LacZ; Fig. 5B). Next, we transduced primary colon cancer cells with pLenti lox-mCh-LacZ and either pLenti SRE-GFP-CreERT2 (P-TuSRE-lin) or pLenti PGK-GFP-CreERT2 (P-TuPGK-lin), expanded single-cell clones, and injected them into NOD/SCID mice (Fig. 5C). After xenograft growth, a single tamoxifen pulse was given to induce recombination (Fig. 5D) that after 2 days caused labeling of individual tumor cells by LacZ in P-TuSRE-lin and P-TuPGK-lin xenografts (Fig. 5E). Importantly, at this early time point, recombined LacZ-positive tumor cells were located toward the leading tumor edge in P-TuSRE-lin xenografts, whereas in P-TuPGK-lin xenografts, they were randomly distributed throughout the tumor (Fig. 5E). Moreover, when examining tumors 21 days after tamoxifen induction, the number of recombined LacZ-positive tumor cells had significantly increased throughout the tumor in P-TuSRE-lin xenografts, whereas in contrast, P-TuPGK-lin xenografts showed a loss of LacZ-labeled tumor cells (Fig. 5E and F). These findings demonstrated a higher potential of initially labeled tumor cells in P-TuSRE-lin than in P-TuPGK-lin xenografts for maintaining tumor cell lineages in colon cancer.

Figure 5.

Lineage tracing of colon cancer cells with high MAPK activity. A, Lentiviral vectors for GFP and CreERT2 expression under control of a MAPK-responsive promoter (pLenti SRE-GFP-CreERT2) or a ubiquitously active promoter (pLenti PGK-GFP-CreERT2). B, Cre-responsive lentiviral vector with a double-floxed inverted LacZ gene. Upon Cre activation, the LacZ gene will be irreversibly inverted and expressed under control of the CMV promoter. mCherry expression can be used as a transduction marker. LTR, long terminal repeat; PRE posttranscriptional regulatory element; SRE, serum response element. C, P-TuSRE-lin and control P-TuPGK-lin xenografts (n ≥ 3) were derived from single cells that were double transduced with pLenti lox-mCh-LacZ and pLenti SRE-GFP-CreERT2 or pLenti PGK-GFP-CreERT2, respectively. D, Experimental schedule for lineage tracing. E, Immunostaining for LacZ. F, Quantification of LacZ-positive (LacZ+) tumor cells in P-TuSRE-lin and P-TuPGK-lin xenografts (n ≥ 3) at indicated time points after tamoxifen-induced recombination. Scale bars, 50 μm. *, P < 0.05; ***, P < 0.001 by t test.

Figure 5.

Lineage tracing of colon cancer cells with high MAPK activity. A, Lentiviral vectors for GFP and CreERT2 expression under control of a MAPK-responsive promoter (pLenti SRE-GFP-CreERT2) or a ubiquitously active promoter (pLenti PGK-GFP-CreERT2). B, Cre-responsive lentiviral vector with a double-floxed inverted LacZ gene. Upon Cre activation, the LacZ gene will be irreversibly inverted and expressed under control of the CMV promoter. mCherry expression can be used as a transduction marker. LTR, long terminal repeat; PRE posttranscriptional regulatory element; SRE, serum response element. C, P-TuSRE-lin and control P-TuPGK-lin xenografts (n ≥ 3) were derived from single cells that were double transduced with pLenti lox-mCh-LacZ and pLenti SRE-GFP-CreERT2 or pLenti PGK-GFP-CreERT2, respectively. D, Experimental schedule for lineage tracing. E, Immunostaining for LacZ. F, Quantification of LacZ-positive (LacZ+) tumor cells in P-TuSRE-lin and P-TuPGK-lin xenografts (n ≥ 3) at indicated time points after tamoxifen-induced recombination. Scale bars, 50 μm. *, P < 0.05; ***, P < 0.001 by t test.

Close modal

To further characterize recombined colon cancer cells in P-TuSRE-lin xenografts, we analyzed them for colocalization with GFP, indicating high MAPK activity. As expected, 2 days after tamoxifen induction, most recombined LacZ-positive tumor cells also showed high GFP expression, indicating predominant labeling of tumor cells with high MAPK activity (Fig. 6A and B). Importantly, when examining tumors 21 days after recombination, the LacZ label had extended into tumor cell subpopulations with low GFP expression, and thus, lower MAPK activity (Fig. 6A and B). We then tested for colocalization of recombined cells and nuclear β-catenin as a marker for high WNT activity. Two days after recombination, LacZ-labeled tumor cells had significantly higher levels of nuclear β-catenin than LacZ-negative tumor cells (Fig. 6C and D). Similar to our findings for GFP, the label had extended into tumor cell subpopulations without nuclear β-catenin staining 21 days after recombination (Fig. 6C and D). Collectively, these data demonstrated a significant contribution of colon cancer cell subsets with high MAPK activity and concomitantly high WNT activity to lineage persistence in vivo.

Figure 6.

Phenotypic switch of tumor cells during lineage outgrowth. A and C, Immunofluorescence for indicated proteins in P-TuSRE-lin xenografts (n ≥ 3) at indicated time points after tamoxifen-induced recombination. Scale bars, 50 μm. B and D, Quantification of GFP and β-catenin immunofluorescence in LacZ-positive (LacZ+) or -negative (LacZ−) tumor cells at indicated time points after tamoxifen-induced recombination. Percentage values of relative fluorescence intensity (% RFI) of individual tumor cells (n ≥ 200) in ≥3 biologic replicates are shown. **, P < 0.01; ***, P < 0.0001; n.s., nonsignificant by t test.

Figure 6.

Phenotypic switch of tumor cells during lineage outgrowth. A and C, Immunofluorescence for indicated proteins in P-TuSRE-lin xenografts (n ≥ 3) at indicated time points after tamoxifen-induced recombination. Scale bars, 50 μm. B and D, Quantification of GFP and β-catenin immunofluorescence in LacZ-positive (LacZ+) or -negative (LacZ−) tumor cells at indicated time points after tamoxifen-induced recombination. Percentage values of relative fluorescence intensity (% RFI) of individual tumor cells (n ≥ 200) in ≥3 biologic replicates are shown. **, P < 0.01; ***, P < 0.0001; n.s., nonsignificant by t test.

Close modal

Here, we demonstrate strong intratumoral heterogeneity and sustained regulation of MAPK signaling not only in KRAS wild-type but also in KRAS-mutant colorectal cancer. We show that in human colorectal cancer tissues and in colon cancer xenografts with clonal KRAS mutation status, high MAPK activity is consistently restricted to tumor cells at the leading tumor edge, whereas more centrally and glandular differentiated tumor cell subpopulations have lower MAPK activity. This observation at first is unexpected when considering that oncogenic mutations in KRAS are assumed to constitutively activate MAPK signaling in all clonally derived colon cancer cells (7, 8, 28). However, our in vitro data demonstrate that in both KRAS wild-type and mutant colon cancers, the inducibility of MAPK signaling is retained at all levels upon growth factor stimulation. Furthermore, we show that in KRAS-mutated colorectal cancer, regulation of MAPK signaling is maintained through the remaining wild-type RAS isoforms. Our findings thus parallel recent observations in other RAS-mutated cancer cells in which the capacity to activate downstream MAPK signaling is not saturated (29). When additionally considering that growth factor secreting stromal cells surround colon cancers at the leading tumor edge (11, 14), heterogeneous MAPK signaling may likely be caused by differential stimulation of tumor cells by the tumor microenvironment, irrespective of the tumors' KRAS mutation status. Of note, as activating mutations in RAS decrease EGFR sensitivity (29), this may still explain why despite maintained MAPK regulation, KRAS-mutated colorectal cancers respond less to EGFR inhibition therapies by antibody drugs, such as cetuximab (9, 10). However, in light of WNT as another signaling pathway that remains regulated in APC-mutated tumors (15), the paradigm of constitutive pathway activation through oncogenic mutations may generally require reconsideration.

Colon cancer cells with high MAPK activity had a distinct phenotype with decreased E-cadherin expression, indicating loss of epithelial characteristics, and increased expression of the ZEB1 target LAMC2, a marker for EMT (30). Moreover, ectopic activation of MAPK signaling caused loss of E-cadherin and upregulation of SNAI1, a direct inducer of EMT (31). In line with these findings, previous studies demonstrated that ectopic expression of FRA1 or constitutively active MEK1 induced EMT in colon cancer cells and was linked to tumor progression (32–35). Because we show similar effects of MAPK activation in colon cancer cells with and without activating KRAS mutations, we propose that MAPK signaling generally regulates EMT in colorectal cancer. Moreover, our data may provide a rationale for this sustained MAPK regulation. Because we found that colon cancer cells with high MAPK activity undergoing EMT had low proliferation rates, differentially high and low MAPK activity may be required to balance infiltrative tumor cells undergoing EMT and tumor cell proliferation forming new tumor mass. We therefore suggest that the full malignant potential of colorectal cancer depends on differential MAPK signaling, allowing for phenotypic plasticity that generates tumor cell subpopulations with distinct phenotypes fostering tumor growth and progression, respectively. Nevertheless, because we did not detect invasion-associated upregulation of MAPK signaling in some colorectal cancers that were negative for both p-ERK and FRA1, we cannot exclude that MAPK-independent mechanisms of colorectal cancer invasion also may exist.

In addition to an EMT phenotype, colon cancer cells with high MAPK activity showed strong staining for nuclear β-catenin, indicating high WNT activity (13). This observation can be explained, as we and others have previously demonstrated that full activation of WNT signaling requires MAPK in colorectal cancer (15, 36). Also, more recently an LRP6-dependent mechanism for this signaling crosstalk has been revealed (37). In addition, we here observed that activation of MAPK signaling in colon cancer cells caused increased expression of CD44, ASCL2, and EPHB2. Because these markers and high WNT activity have previously been linked to putative colon cancer–initiating cells (14, 25–27), we hypothesized that high MAPK activity may indicate a progenitor cell phenotype in colorectal cancer. Using a lineage tracing strategy, we demonstrate a higher contribution of colon cancer cells with high MAPK activity to persistent tumor cell lineages, when compared to a random tumor cell subpopulation. Moreover, we show lineage outgrowth of colon cancer cells with high MAPK and high WNT activity into cancer cell subpopulations with lower activity for both signaling pathways. Therefore, our findings suggest that high MAPK activity characterizes stem-like tumor cells that continuously give rise to more differentiated tumor cell subpopulations and might hence be regarded as the cancerous equivalent of an organ-based adult tissue stem cell (38). Our approach differs from previous descriptions of colon cancer stem cells that were operationally defined by tumor-initiating potential in limiting dilution xenografts, which had the caveat of low reproducibility (14, 15, 39, 40). Inducing a tumor in mice may rather require robustness of the injected tumor cell but may not directly assess its hierarchical level within the original tumor (reviewed in ref. 41). In contrast to such tumor-initiating studies that capture an event that has to build a new tumor, our approach marks a growth-fueling progenitor cell compartment within the mature tumor architecture. We therefore believe that lineage tracing advances the cancer stem cell field and will improve our understanding on how tumor cell subpopulations contribute to cancer outgrowth and persistence.

A. Jung has received speakers' bureau honoraria from Amgen, Merck-Serono, AstraZeneca, Roche Pharma and is a consultant/advisory board member for Amgen, Biocartis, and AstraZeneca. T. Kirchner has received speakers' bureau honoraria from Merck, Astra Zeneca, Amgen, MSD, and Novartis. No potential conflicts of interest were disclosed by the other authors.

Conception and design: C. Blaj, D. Horst

Development of methodology: C. Blaj, D. Horst

Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): C. Blaj, E.M. Schmidt, S. Lamprecht, H. Hermeking, T. Kirchner, D. Horst

Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): C. Blaj, E.M. Schmidt, S. Lamprecht, A. Jung, T. Kirchner, D. Horst

Writing, review, and/or revision of the manuscript: C. Blaj, E.M. Schmidt, S. Lamprecht, H. Hermeking, A. Jung, T. Kirchner, D. Horst

Administrative, technical, or material support (i.e., reporting or organizing data, constructing databases): S. Lamprecht, A. Jung, D. Horst

Study supervision: D. Horst

We are grateful to Raffaele Conca, Anja Heier, Anne Küchler, Sabine Sagebiel-Kohler, and Andrea Sendelhofert for experimental assistance and to the HTCR for supplying primary colon cancer cells.

This study was supported by grants from the Deutsche Forschungsgemeinschaft (HO4325/4-1) and the Deutsche Krebshilfe (111669) to D. Horst.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

1.
Fearon
ER
,
Vogelstein
B
. 
A genetic model for colorectal tumorigenesis
.
Cell
1990
;
61
:
759
67
.
2.
Vaughn
CP
,
Zobell
SD
,
Furtado
LV
,
Baker
CL
,
Samowitz
WS
. 
Frequency of KRAS, BRAF, and NRAS mutations in colorectal cancer
.
Genes Chromosomes Cancer
2011
;
50
:
307
12
.
3.
Schubbert
S
,
Shannon
K
,
Bollag
G
. 
Hyperactive Ras in developmental disorders and cancer
.
Nat Rev Cancer
2007
;
7
:
295
308
.
4.
Yordy
JS
,
Muise-Helmericks
RC
. 
Signal transduction and the Ets family of transcription factors
.
Oncogene
2000
;
19
:
6503
13
.
5.
Urosevic
J
,
Garcia-Albeniz
X
,
Planet
E
,
Real
S
,
Cespedes
MV
,
Guiu
M
, et al
Colon cancer cells colonize the lung from established liver metastases through p38 MAPK signalling and PTHLH
.
Nat Cell Biol
2014
;
16
:
685
94
.
6.
Scheffzek
K
,
Ahmadian
MR
,
Kabsch
W
,
Wiesmuller
L
,
Lautwein
A
,
Schmitz
F
, et al
The Ras-RasGAP complex: structural basis for GTPase activation and its loss in oncogenic Ras mutants
.
Science
1997
;
277
:
333
8
.
7.
Hatzivassiliou
G
,
Haling
JR
,
Chen
H
,
Song
K
,
Price
S
,
Heald
R
, et al
Mechanism of MEK inhibition determines efficacy in mutant KRAS- versus BRAF-driven cancers
.
Nature
2013
;
501
:
232
6
.
8.
Khambata-Ford
S
,
Garrett
CR
,
Meropol
NJ
,
Basik
M
,
Harbison
CT
,
Wu
S
, et al
Expression of epiregulin and amphiregulin and K-ras mutation status predict disease control in metastatic colorectal cancer patients treated with cetuximab
.
J Clin Oncol
2007
;
25
:
3230
7
.
9.
Karapetis
CS
,
Khambata-Ford
S
,
Jonker
DJ
,
O'Callaghan
CJ
,
Tu
D
,
Tebbutt
NC
, et al
K-ras mutations and benefit from cetuximab in advanced colorectal cancer
.
N Engl J Med
2008
;
359
:
1757
65
.
10.
Lievre
A
,
Bachet
JB
,
Boige
V
,
Cayre
A
,
Le Corre
D
,
Buc
E
, et al
KRAS mutations as an independent prognostic factor in patients with advanced colorectal cancer treated with cetuximab
.
J Clin Oncol
2008
;
26
:
374
9
.
11.
Brabletz
T
,
Jung
A
,
Reu
S
,
Porzner
M
,
Hlubek
F
,
Kunz-Schughart
LA
, et al
Variable beta-catenin expression in colorectal cancers indicates tumor progression driven by the tumor environment
.
Proc Natl Acad Sci U S A
2001
;
98
:
10356
61
.
12.
Horst
D
,
Reu
S
,
Kriegl
L
,
Engel
J
,
Kirchner
T
,
Jung
A
. 
The intratumoral distribution of nuclear beta-catenin is a prognostic marker in colon cancer
.
Cancer
2009
;
115
:
2063
70
.
13.
Fodde
R
,
Brabletz
T
. 
Wnt/beta-catenin signaling in cancer stemness and malignant behavior
.
Curr Opin Cell Biol
2007
;
19
:
150
8
.
14.
Vermeulen
L
,
De Sousa
EMF
,
van der Heijden
M
,
Cameron
K
,
de Jong
JH
,
Borovski
T
, et al
Wnt activity defines colon cancer stem cells and is regulated by the microenvironment
.
Nat Cell Biol
2010
;
12
:
468
76
.
15.
Horst
D
,
Chen
J
,
Morikawa
T
,
Ogino
S
,
Kirchner
T
,
Shivdasani
RA
. 
Differential WNT activity in colorectal cancer confers limited tumorigenic potential and is regulated by MAPK signaling
.
Cancer Res
2012
;
72
:
1547
56
.
16.
Heinemann
V
,
von Weikersthal
LF
,
Decker
T
,
Kiani
A
,
Vehling-Kaiser
U
,
Al-Batran
SE
, et al
FOLFIRI plus cetuximab versus FOLFIRI plus bevacizumab as first-line treatment for patients with metastatic colorectal cancer (FIRE-3): a randomised, open-label, phase 3 trial
.
Lancet Oncol
2014
;
15
:
1065
75
.
17.
Hayes
TE
,
Sengupta
P
,
Cochran
BH
. 
The human c-fos serum response factor and the yeast factors GRM/PRTF have related DNA-binding specificities
.
Genes Dev
1988
;
2
:
1713
22
.
18.
Matsuda
T
,
Cepko
CL
. 
Controlled expression of transgenes introduced by in vivo electroporation
.
Proc Natl Acad Sci U S A
2007
;
104
:
1027
32
.
19.
Zhang
F
,
Gradinaru
V
,
Adamantidis
AR
,
Durand
R
,
Airan
RD
,
de Lecea
L
, et al
Optogenetic interrogation of neural circuits: technology for probing mammalian brain structures
.
Nat Protoc
2010
;
5
:
439
56
.
20.
Kita-Matsuo
H
,
Barcova
M
,
Prigozhina
N
,
Salomonis
N
,
Wei
K
,
Jacot
JG
, et al
Lentiviral vectors and protocols for creation of stable hESC lines for fluorescent tracking and drug resistance selection of cardiomyocytes
.
PLoS One
2009
;
4
:
e5046
.
21.
Pfeifer
A
,
Brandon
EP
,
Kootstra
N
,
Gage
FH
,
Verma
IM
. 
Delivery of the Cre recombinase by a self-deleting lentiviral vector: efficient gene targeting
in vivo.
Proc Natl Acad Sci U S A
2001
;
98
:
11450
5
.
22.
Covassin
LD
,
Siekmann
AF
,
Kacergis
MC
,
Laver
E
,
Moore
JC
,
Villefranc
JA
, et al
A genetic screen for vascular mutants in zebrafish reveals dynamic roles for Vegf/Plcg1 signaling during artery development
.
Dev Biol
2009
;
329
:
212
26
.
23.
Young
MR
,
Colburn
NH
. 
Fra-1 a target for cancer prevention or intervention
.
Gene
2006
;
379
:
1
11
.
24.
Zhang
W
,
Hart
J
,
McLeod
HL
,
Wang
HL
. 
Differential expression of the AP-1 transcription factor family members in human colorectal epithelial and neuroendocrine neoplasms
.
Am J Clin Pathol
2005
;
124
:
11
9
.
25.
Dalerba
P
,
Dylla
SJ
,
Park
IK
,
Liu
R
,
Wang
X
,
Cho
RW
, et al
Phenotypic characterization of human colorectal cancer stem cells
.
Proc Natl Acad Sci U S A
2007
;
104
:
10158
63
.
26.
Merlos-Suarez
A
,
Barriga
FM
,
Jung
P
,
Iglesias
M
,
Cespedes
MV
,
Rossell
D
, et al
The intestinal stem cell signature identifies colorectal cancer stem cells and predicts disease relapse
.
Cell Stem Cell
2011
;
8
:
511
24
.
27.
Stange
DE
,
Engel
F
,
Longerich
T
,
Koo
BK
,
Koch
M
,
Delhomme
N
, et al
Expression of an ASCL2 related stem cell signature and IGF2 in colorectal cancer liver metastases with 11p15.5 gain
.
Gut
2010
;
59
:
1236
44
.
28.
Fearon
ER
,
Hamilton
SR
,
Vogelstein
B
. 
Clonal analysis of human colorectal tumors
.
Science
1987
;
238
:
193
7
.
29.
Young
A
,
Lou
D
,
McCormick
F
. 
Oncogenic and wild-type Ras play divergent roles in the regulation of mitogen-activated protein kinase signaling
.
Cancer Discov
2013
;
3
:
112
23
.
30.
Sanchez-Tillo
E
,
de Barrios
O
,
Siles
L
,
Cuatrecasas
M
,
Castells
A
,
Postigo
A
. 
beta-catenin/TCF4 complex induces the epithelial-to-mesenchymal transition (EMT)-activator ZEB1 to regulate tumor invasiveness
.
Proc Natl Acad Sci U S A
2011
;
108
:
19204
9
.
31.
Peinado
H
,
Olmeda
D
,
Cano
A
. 
Snail, Zeb and bHLH factors in tumour progression: an alliance against the epithelial phenotype?
Nat Rev Cancer
2007
;
7
:
415
28
.
32.
Bakiri
L
,
Macho-Maschler
S
,
Custic
I
,
Niemiec
J
,
Guio-Carrion
A
,
Hasenfuss
SC
, et al
Fra-1/AP-1 induces EMT in mammary epithelial cells by modulating Zeb1/2 and TGFbeta expression
.
Cell Death Differ
2015
;
22
:
336
50
.
33.
Diesch
J
,
Sanij
E
,
Gilan
O
,
Love
C
,
Tran
H
,
Fleming
NI
, et al
Widespread FRA1-dependent control of mesenchymal transdifferentiation programs in colorectal cancer cells
.
PLoS One
2014
;
9
:
e88950
.
34.
Lemieux
E
,
Bergeron
S
,
Durand
V
,
Asselin
C
,
Saucier
C
,
Rivard
N
. 
Constitutively active MEK1 is sufficient to induce epithelial-to-mesenchymal transition in intestinal epithelial cells and to promote tumor invasion and metastasis
.
Int J Cancer
2009
;
125
:
1575
86
.
35.
Liu
H
,
Ren
G
,
Wang
T
,
Chen
Y
,
Gong
C
,
Bai
Y
, et al
Aberrantly expressed Fra-1 by IL-6/STAT3 transactivation promotes colorectal cancer aggressiveness through epithelial-mesenchymal transition
.
Carcinogenesis
2015
;
36
:
459
68
.
36.
Phelps
RA
,
Chidester
S
,
Dehghanizadeh
S
,
Phelps
J
,
Sandoval
IT
,
Rai
K
, et al
A two-step model for colon adenoma initiation and progression caused by APC loss
.
Cell
2009
;
137
:
623
34
.
37.
Lemieux
E
,
Cagnol
S
,
Beaudry
K
,
Carrier
J
,
Rivard
N
. 
Oncogenic KRAS signalling promotes the Wnt/beta-catenin pathway through LRP6 in colorectal cancer
.
Oncogene
2015
;
34
:
4914
27
.
38.
Barker
N
,
van Es
JH
,
Kuipers
J
,
Kujala
P
,
van den Born
M
,
Cozijnsen
M
, et al
Identification of stem cells in small intestine and colon by marker gene Lgr5
.
Nature
2007
;
449
:
1003
7
.
39.
O'Brien
CA
,
Pollett
A
,
Gallinger
S
,
Dick
JE
. 
A human colon cancer cell capable of initiating tumour growth in immunodeficient mice
.
Nature
2007
;
445
:
106
10
.
40.
Shmelkov
SV
,
Butler
JM
,
Hooper
AT
,
Hormigo
A
,
Kushner
J
,
Milde
T
, et al
CD133 expression is not restricted to stem cells, and both CD133+ and CD133- metastatic colon cancer cells initiate tumors
.
J Clin Invest
2008
;
118
:
2111
20
.
41.
Clevers
H
. 
The cancer stem cell: premises, promises and challenges
.
Nat Med
2011
;
17
:
313
9
.