• inflammation;
  • colorectal cancer;
  • mouse model;
  • toll-like receptor;
  • tumor microenvironment


  1. Top of page
  2. Abstract

Background: Patients with ulcerative colitis are at increased risk for developing colorectal cancer. We have shown that Toll-like receptor-4 (TLR4) is overexpressed in human colitis-associated cancer (CAC) and that mice deficient in TLR4 are markedly protected against colitis-associated neoplasia. We wished to elucidate the specific contributions of TLR4 signaling by myeloid cells and colonic epithelial cells (CEC) in colitis-associated tumorigenesis.

Methods: TLR4-deficient mice or wildtype littermates (WT) were transplanted with bone marrow (BM) cells: TLR4−/− BM[RIGHTWARDS ARROW]WT mice (TLR4-expressing CEC) and WT BM[RIGHTWARDS ARROW]TLR4−/− mice (TLR4-expressing myeloid cells). Colitis-associated neoplasia was induced by azoxymethane (AOM 7.3 mg/kg) injection and 2 cycles of dextran sodium sulfate (DSS) treatment.

Results: The number and size of dysplastic lesions were greater in TLR4−/− BM[RIGHTWARDS ARROW]WT mice than in WT BM[RIGHTWARDS ARROW]TLR4−/− mice (P < 0.005). Histologically, TLR4−/− BM[RIGHTWARDS ARROW]WT mice had greater numbers of mucosal neutrophils and macrophages compared to WT BM[RIGHTWARDS ARROW]TLR4−/− mice. The chemokines KC and CCL2, important in recruitment of neutrophils and macrophages, respectively, were induced in mice expressing TLR4 in CEC rather than the myeloid compartment. The lamina propria infiltrate of mice expressing TLR4 in CEC was characterized by macrophages expressing Cox-2. Moreover, mice expressing TLR4 in CEC rather than the myeloid compartment had increased production of amphiregulin and EGFR activation.

Conclusions: These findings indicate that TLR4 signaling on CEC is necessary for recruitment and activation of Cox-2-expressing macrophages and increasing the number and size of dysplastic lesions. Our results implicate innate immune signaling on CEC as a key regulator of a tumor-promoting microenvironment.

(Inflamm Bowel Dis 2009)

Inflammation and cancer are linked. This unfortunate connection is detrimental to patients with inflammatory bowel disease (IBD), who are at a substantially increased risk for colorectal cancer. Even when cancer does not arise, the ominous presence of dysplasia often results in a total proctocolectomy, since it is not currently possible to reliably predict who will progress to cancer or to halt the progression to cancer. Patients with ulcerative colitis (UC) have a lifetime relative risk of colorectal cancer that is 5 to 8 times higher than the normal population.1–5 The exact mechanism by which chronic intestinal inflammation causes colorectal cancer remains uncertain.6–9

We have recently demonstrated that Toll-like receptor-4 (TLR4) expression is upregulated in colitis-associated cancer (CAC) lesions from patients with UC but not in the surrounding tissue.10 Moreover, investigators have examined pathologic specimens from patients with CAC and found that, compared to biopsies from UC, patients that did not develop cancer, colon biopsies from CAC patients had a higher degree of acute inflammation prior to the identification of cancer.8, 11 These results suggest that innate immune signaling in the intestine plays a critical role in the pathogenesis of CAC.

Animal models have provided a useful paradigm to study the contribution of intestinal epithelial cells or inflammatory cells in the pathogenesis of CAC. The combination of azoxymethane (AOM), a colonic genotoxic carcinogen, with dextran sulfate sodium (DSS), an inducer of colitis, has been frequently used as a mouse model of CAC.12, 13 The importance of epithelial and myeloid nuclear factor kappaB (NF-κB) signaling in the development of CAC has been demonstrated using this model.14 More recently, a report has shown that tumor necrosis factor alpha (TNF-α) signaling via the p55 TNF receptor in myeloid cells is necessary for tumor development in the same mouse CAC model, whereas epithelial TNF receptor signaling does not affect tumor numbers.15 TNF-α potently induces NF-κB activation but can also be a NF-κB-target gene.16, 17 These data suggest that inflammatory cytokines and NF-κB activation may influence CAC development both at the level of the epithelial cell or the immune cell.

One of the unique aspects of the intestinal epithelium and the development of colorectal cancer is the role of intestinal bacteria. The intestinal epithelium is bathed in microbes and microbial products. Signaling through TLRs on either epithelial cells or lamina propria mononuclear cells activates NF-κB and is necessary for intestinal homeostasis and repair.18–20 Germ-free rodents are protected from colitis-associated neoplasia, supporting a role for bacterial recognition in CAC.21, 22 To delineate the role of TLR signaling in the development of CAC, we previously used TLR4−/− mice and asked whether these animals were protected against colitis-associated neoplasia. We demonstrated that mice deficient in TLR4 are markedly protected against the development of neoplasia.10 By contrast, single immunoglobulin IL-1 receptor-related molecule (SIGIRR)-deficient animals, which demonstrate increased TLR signaling, have increased intestinal inflammation and increased tumorigenesis following treatment with AOM-DSS.23, 24 Restoring SIGIRR expression in the epithelium reduces inflammation and tumorigenesis, suggesting a role for epithelial TLR signaling in tumor development.

Based on our own work showing the importance of TLR4 in the development of colitis-associated neoplasia, we wished to determine whether TLR4 expression on the epithelium, myeloid cells, or both was required for the development of colitis-associated neoplasia. Traditionally, TLR signaling is thought to occur primarily in hematopoietic cells such as macrophages and dendritic cells. Using bone marrow chimeras, we demonstrate here that TLR4 expression by colonic epithelial cells (CECs) plays a greater role in the development of colitis-associated neoplasia than myeloid expression of TLR4. TLR4 expression by CECs was able to direct recruitment of tumor-associated macrophages expressing Cox-2. Our data support targeted inhibition of TLR signaling for the prevention or treatment of CAC.


  1. Top of page
  2. Abstract

Allogeneic Bone Marrow (BM) Transplantation in Mice

TLR4−/− mice were purchased from Oriental Bio Service (Kyoto, Japan). All knockout mice were backcrossed to C57Bl/6J mice at least 8 generations. C57BL/6J mice were obtained from Jackson Laboratory (Bar Harbor, ME) as controls. Mice were kept in specific-pathogen free (SPF) conditions and fed by free access to a standard diet and water. All experiments were done according to Mount Sinai School of Medicine Animal Experimental Ethics Committee guidelines.

Six-week-old C57Bl/6J wildtype (WT) or TLR4−/− recipient mice were irradiated with 10 Gy from a 137Cs source delivered in 2 doses of 5 Gy each, 3 hours apart. BM cells were isolated from 6–8-week-old C57Bl/6J or TLR4−/− donor mice by flushing the bone shafts of the femurs and tibias with RPMI and 2–3 × 106 BM cells intravenously injected into the recipient mice 3 hours after the last irradiation. WT mice were transplanted with TLR4−/− BM (TLR4−/− BM[RIGHTWARDS ARROW]WT) and TLR4−/− mice were transplanted with WT BM (WT BM[RIGHTWARDS ARROW]TLR4−/−). As controls, WT mice were transplanted with WT BM and TLR4−/− mice transplanted with TLR4−/− BM. Using C57BL/6-Tg (ACTB-EGFP) 10sb/J as donor mice, chimerism was confirmed by examining GFP+ cells in splenocytes 8 weeks after transplantation by fluorescence activated cell sorter (FACS) analysis. Using this protocol we have not seen any histological damage of intestinal mucosa induced by irradiation process at 8 weeks of BM transplantation.

BM chimeric mice were injected with 7.4 mg/kg of AOM (Sigma, St. Louis, MO) intraperitoneally (i.p.) 8 weeks after transplantation. After 14 days of AOM injection, mice were treated with 2.5% DSS (MW 36–50 kDa: ICN, Aurora, Ohio) in their drinking water for 7 days.10 This was followed by 14 days of normal water, another 7 days of 2.5% DSS treatment, and then normal water for an additional 14 days. During the DSS treatment and recovery phase, body weights, stool consistency, and stool occult blood were monitored, as described previously.10

Mice were sacrificed on day 77 after AOM, 5 weeks after the second DSS cycle. Colons were removed and opened longitudinally. Cecum, proximal, and distal parts of the colon were fixed in 10% buffered formalin, embedded in paraffin, sectioned, and stained with hematoxylin and eosin. Histological assessment was performed by 2 independent gastrointestinal pathologists (R.X., H.C.) blinded to the mouse genotype and treatment. Severity of mucosal inflammation was graded using a standard scoring system described previously.10 To quantify the microscopic extent of dysplasia, paraffin-embedded colons were cut in 5-μm-thick serial sections and every 20th section was analyzed for dysplasia. Number, size, and the percentage of the mucosal surface area containing dysplasia were determined microscopically. The size of the lesions was calculated using a scale micrometer on the microscope.

Real-time PCR

One μg of total RNA isolated with RNA Bee (Tel-Test, Friendwood, TX) was used as the template for single-strand cDNA synthesis using the Transcriptor First Strand cDNA Synthesis Kit (Roche, Indianapolis, IN) according to the manufacturer's instructions. Quantitative real-time polymerase chain reaction (PCR) was performed for Cox-2, amphiregulin, and β-actin using TaqMan probes. The primers and probes used in this study are as follows (5′ to 3′ direction), for mouse Cox-2: sense primer, AAG GAA CTC AGC ACT GCA TCC, antisense primer, ACA GGG ATT GGA ACA GCA AGG A, and probe, ACC GCC ACC ACT ACT GCC ACC TCC; for mouse amphiregulin: sense primer, TGT CAC TAT CTT TGT CTC TGC CAT, anti-sense primer, AGC CTC CTT CTT TCT TCT GTT TCT, and probe, TCC TCG CAG CTA TTG GCA TCG GCA; for mouse β-actin: sense primer, ATG ACC CAG ATC ATG TTT G, antisense primer, TAC GAC CAG AGG CAT ACA, and probe, CGT AGC CAT CCA GGC TGT GC. All TaqMan probes and primers were designed using Beacon Designer 3.0 software (Premier Biosoft International, Palo Alto, CA). The cDNA was amplified using TaqMan universal PCR Master Mix (Roche, Indianapolis, IN) on an ABI Prism 7900HT sequence detection system (Applied Biosystems, Foster City, CA), programmed for 95°C for 10 minutes, then 40 cycles of: 95°C for 15 seconds, 60°C for 1 minute.

Real-time PCR for CCL2 and KC was performed using SYBR Premix Ex Taq (Takara Bio, Shiga, Japan) with the standard SYBR Green setting of 7900HT. The primers for mouse CCL2 are: sense primer, GCT GGA GCA TCC ACG TGT T, antisense primer, ATC TTG CTG GTG AAT GAG TAG CA; for mouse KC: sense primer, AAT GAG CTG CGC TGT CAG TG, antisense primer, TGA GGG CAA CAC CTT CAA GC. β-Actin primers are the same primers used for TaqMan real-time PCR.

The amplification results were analyzed using SDS 2.2.1 software (Applied Biosystems) and the gene of interest was normalized to the corresponding β-actin results. Data were expressed as fold induction relative to the lowest gene product amplified.

Western Blot Analysis

Mouse colon samples were taken at the time of sacrifice and frozen at −80°C. Tissue lysates were prepared using a lysis buffer containing 50 mM Tris HCl, 50 mM NaF, 1% Triton X-100, 2 mM EDTA, and 100 mM NaCl, with a proteinase inhibitor cocktail (Calbiochem, San Diego, CA). Protein concentration was determined by the Bradford method using Bio-Rad Protein Assay Dye and SmartSpec 3000 (Bio-Rad Laboratories, Hercules, CA). Twenty-five μg of the lysates were subjected to 10% SDS-PAGE and transferred to Immobilon-P membranes (Millipore, Bedford, MA). The membrane was blocked in 5% skim milk and was immunoblotted with anti-phospho-EGFR (Tyr 1173, goat polyclonal; Santa Cruz Biotechnology, Santa Cruz, CA) antibody for 1 hour, followed by incubation with horseradish peroxidase (HRP)-conjugated antigoat IgG (Zymed Laboratories, South San Francisco, CA). After stripping with Restore Western Blot Stripping Buffer (Pierce Biotechnology, Rockford, IL), the membrane was further blotted with anti-EGFR (rabbit polyclonal; Santa Cruz Biotechnology) followed by incubation with HRP-conjugated antirabbit IgG (Jackson ImmunoResearch Laboratories, West Grove, PA). The membrane was exposed on an x-ray film using an enhanced chemiluminescent substrate SuperSignal West Pico Trial Kit (Pierce Biotechnology). Phospho-EGFR band intensity was calculated using NIH Image 1.62 by normalizing with the intensity of the corresponding EGFR band.

Immunofluorescent and Immunohistochemical Studies

Paraffin-embedded sections were incubated in 10% normal goat serum for 1 hour and stained with antimurine Cox-2 antibody (1:200, Cayman, Ann Arbor, MI) overnight at 4°C, followed by FITC-conjugated antirabbit IgG (1:200, Sigma) for 1 hour at room temperature. The specificity of staining was confirmed using Cox-2 blocking peptide (Cayman) according to the manufacturer's instructions or using rabbit isotype control antibody instead of the primary antibody (Zymed Laboratories).

For the double immunofluorescent staining of CD68 and Cox-2, sections were incubated with 0.1% Trypsin (Sigma) CaCl2 dissolved in 0.05 M Tris-HCl, pH 7.6, for 15 minutes at 37°C. Subsequently, sections were blocked in a 5% skim milk for 1 hour and then incubated with the rat anti-CD68 antibody (1:50, MCA1957S, Serotec, Raleigh, NC) overnight at 4°C. After washing in phosphate-buffered saline (PBS), sections were incubated with TRITC-conjugated rabbit antirat IgG (1:200, Sigma) for 1 hour at room temperature. Then sections were reincubated with 5% skim milk followed by Cox-2 staining as described above using FITC-conjugated goat antirabbit IgG (1:200, Sigma). The number of Cox-2-positive cells plus CD68-positive cells infiltrating the tumor tissue or lamina propria in the surrounding mucosa was counted with double-stained slides at 400× magnification.

Assessment of Proliferation

Colonic tissue sections were examined for cell proliferation by bromodeoxyuridine (BrdU) labeling. Mice were injected with 120 mg/kg of BrdU (Sigma) i.p., 90 minutes prior to sacrificing, and colonic tissues were stained for BrdU using a BrdU staining kit (Zymed Laboratories) according to the manufacturer's instructions. The number of BrdU-positive cells per well-oriented crypt was calculated every 3 crypts for each colon segment at high magnification under light microscopy.

Measurement of PGE2

Production of PGE2 in the tissue culture supernatant was determined using a monoclonal EIA kit (Cayman) according to the manufacturer's instructions and as described previously.25, 26 Briefly, colonic samples from TLR4−/− and WT mice were washed in cold PBS containing penicillin, streptomycin, and fungizone (100 U/mL each). Then 100 mg tissue fragments from the distal part of the colon closest to the anus were cultured for 24 hours in 12-well flat-bottom plates in serum-free RPMI 1640. Culture supernatants were harvested for PGE2 measurement.

Statistical Analysis

Data are presented as mean ± SEM. The significance was analyzed by Kruskal–Wallis test followed by multiple comparisons between the groups using GraphPad Prism v. 3.0 (San Diego, CA). Fisher's exact probability test was used to compare tumor incidence. Student's t-test was used to identify significant differences when comparing 2 samples with Microsoft Excel. P-values were considered significant when <0.05.


  1. Top of page
  2. Abstract

Enhanced Development of Inflammation-Induced Colorectal Tumors When Mice Express TLR4 in CECs

We have demonstrated that mice deficient in TLR4 are markedly protected against development of colitis-associated neoplasia in the AOM-DSS model.10 To examine the contribution of TLR4-expressing CEC versus myeloid cells in the development of colitis-associated neoplasia, we used BM chimeras in the AOM-DSS model (Fig. 1A; Table 1). The incidence of dysplastic lesions was significantly reduced when TLR4−/− mice were used as recipients (TLR4 in myeloid compartment) compared to using WT mice (TLR4 in CEC and stroma) as recipients. The number of dysplastic lesions per colon was strikingly different between animals with or without TLR4 expression in CEC. There were no differences in the number of dysplastic lesions between WT mice receiving TLR4−/− BM and WT mice receiving WT BM, suggesting a significant role of TLR4 signaling by CEC in the development of colitis-associated neoplasia. Moreover, the size and extent of dysplasia was also significantly greater when TLR4 was expressed by CEC rather than by myeloid cells (Table 1). TLR4−/− mice receiving WT BM did, however, develop some dysplastic lesions, supporting a contribution from myeloid TLR4 in dysplasia development, whereas TLR4−/− mice receiving TLR4−/− BM did not. These results indicate that the TLR4 signal from myeloid cells may contribute to some extent to the development of colitis-associated neoplasia but the TLR4 signal from CEC is sufficient to recapitulate the neoplasia regardless of the TLR4 signal from myeloid cells.

thumbnail image

Figure 1. Incidence of dysplasia. A: The number of dysplastic lesions was counted per mouse. Data were gathered from 5 independent experiments. Bar represents mean (**P < 0.001, *P < 0.01, P < 0.05). B: Intestinal epithelial proliferation. BrdU labeling index was defined by counting BrdU-positive epithelial cells per crypt. Data are represented as mean ± SEM. TLR4−/− BM[RIGHTWARDS ARROW]WT mice (n = 10) had greater epithelial proliferation than WT BM[RIGHTWARDS ARROW]TLR4−/− mice (n = 16) (*P < 0.001), but less epithelial proliferation than WT BM[RIGHTWARDS ARROW]WT mice (n = 9) (**P < 0.0001). Epithelial proliferation in WT BM[RIGHTWARDS ARROW]TLR4−/− mice was similar to TLR4−/− BM[RIGHTWARDS ARROW]TLR4−/− mice (n = 8).

Download figure to PowerPoint

Table 1. Incidence and Size of Polyps
  • a


  • b


  • c


  • d


Incidence (%)100080.043.80.26a0.006b 0.08c0.03d
Tumors/animal (range)3.1 ± 0.4 (1–4)02.8 ± 0.8 (1–7)0.7 ± 0.2 (0–3)0.37a<0.001b0.002c 0.07d
Tumor size (mm) (range)3.6 ± 0.3 (1–5) 2.8 ± 0.2 (2–5)1.5 ± 0.2 (0.5–3)0.06a<0.001b<0.001c
Percentage of mucosal surface involved with tumor (range)18.9% ± 4.2 (5–40) 8.3% ± 3.3 (5–20)3.0% ± 0.8 (5–10)0.05a<0.001b0.016c

Consistent with our findings above, BrdU labeling demonstrated significantly higher epithelial proliferation in the mice with TLR4 in CEC than the mice with TLR4 in myeloid cells (Fig. 1B). Epithelial proliferation was similar in WT BM[RIGHTWARDS ARROW]TLR4−/− mice compared to TLR4−/− BM[RIGHTWARDS ARROW]TLR4−/− mice, suggesting myeloid expression of TLR4 has no effect on the epithelial proliferation when epithelial expression of TLR4 is absent. However, TLR4−/− BM[RIGHTWARDS ARROW]WT mice continued to have less epithelial proliferation than WT BM[RIGHTWARDS ARROW]WT mice. These data support that epithelial expression of TLR4 drives proliferation in this model.

Characterization of the Tumor Microenvironment in TLR4-Dependent Inflammatory Colorectal Cancer

Having established a system that permits us to study the contribution of epithelial versus myeloid TLR4 signaling in inflammatory colorectal cancer, we examined the role of TLR4 expression on the tumor microenvironment. Neutrophil infiltration of the mucosa in UC correlates with development of CAC.8, 11, 27 We compared the inflammatory infiltrate in our BM chimeras treated with AOM-DSS. Histologically, WT mice transplanted with TLR4−/− BM had a dense infiltrate of neutrophils not seen in TLR4−/− mice receiving WT BM (45.4 ± 46.2 versus 17.5 ± 19.3 in HPF, P < 0.05) (Fig. 2A). We also found a significant difference of neutrophil infiltrate between WT mice receiving WT BM (49.5 ± 26.7 in HPF) and TLR4−/− mice receiving WT BM (P < 0.01). These data demonstrate that CEC expression of TLR4 is necessary and sufficient for recruitment of neutrophils to the lamina propria.

thumbnail image

Figure 2. Characterization of the tumor microenvironment. A: Microscopic findings in TLR4−/− BM[RIGHTWARDS ARROW]WT mice and WT BM[RIGHTWARDS ARROW]TLR4−/−mice mucosa at day 77 of the AOM-DSS model (H&E staining). TLR4−/− BM[RIGHTWARDS ARROW]WT mice (top panel) have inflammatory infiltrate in the mucosa compared with WT BM[RIGHTWARDS ARROW]TLR4−/−mice (bottom panel). B: CD68 positive cells (macrophages) per HPF (400x) were counted in colon samples from AOM-DSS-treated mice. Significantly greater numbers of macrophages were seen in TLR4−/− BM[RIGHTWARDS ARROW]WT mice (n = 10) compared to WT BM[RIGHTWARDS ARROW]TLR4−/−mice (n = 16) in both the tumors and surrounding lamina propria. Three fields per site (cecum, proximal, distal) were examined. Data are represented as mean ± SEM (*P < 0.05).

Download figure to PowerPoint

In addition to neutrophils, tumor-associated macrophages (TAMs) play a central role in the establishment and growth of colorectal cancers.28, 29 We found that WT mice transplanted with TLR4−/− BM have a higher number of tumor-associated macrophages compared to tumors in TLR4−/− mice receiving WT BM (Fig. 2B). There was no statistical difference of the number of tumor-associated macrophages between WT mice transplanted with TLR4−/− BM and WT mice transplanted with WT BM (data not shown). The expression of TLR4 in CEC was also associated with higher numbers of lamina propria macrophages. Thus, in spite of the fact that TLR4 was not expressed on hematopoietic cells, epithelial expression of TLR4 could direct recruitment of neutrophils and macrophages. The differences in chemokine expression will be discussed below.

To determine whether these lamina propria macrophages could function to support tumor development, we examined expression of Cox-2 and production of PGE2. Intestinal inflammation results in increased expression of Cox-2, which is thought to contribute to human CAC in a variety of ways.30, 31 We first asked whether Cox-2 mRNA expression was increased in our BM chimera model. By real-time PCR, Cox-2 expression is significantly higher when TLR4 is expressed by CEC than by myeloid cells (Fig. 3A). Although there was no statistical difference (P = 0.11), there was an almost 50% decrease in Cox-2 expression when WT mice received TLR4−/− BM compared to WT receiving WT BM, suggesting a contribution of myeloid TLR4 signaling in mucosal Cox-2 expression. PGE2 production mirrored expression of Cox-2 but was also seen in TLR4−/− mice receiving WT BM (myeloid TLR4) (Fig. 3B). To determine the source of Cox-2, we used immunofluorescent staining of AOM-DSS-treated BM chimeras. Cox-2-positive cells were found in the lamina propria infiltrate and coincided with CD68-positive cells (Fig. 3C). The number of Cox-2-expressing macrophages was higher in the lamina propria and tumors of TLR4−/− BM[RIGHTWARDS ARROW]WT mice compared with WT BM[RIGHTWARDS ARROW]TLR4−/− mice. These results demonstrate that TLR4 expression by CEC recruits macrophages that express Cox-2 or induces local expression of Cox-2.

thumbnail image

Figure 3. Cox-2 expression from mucosal macrophages is dependent on TLR4 signaling by CEC. A: Cox-2 expression in the colon of AOM-DSS treated mice. Colonic samples were taken at day 77 in the AOM-DSS model. Cox-2 mRNA expression was analyzed by real-time PCR. A significant difference is shown between TLR4−/− BM[RIGHTWARDS ARROW]WT mice (n = 10) and WT BM[RIGHTWARDS ARROW]TLR4−/− mice (n = 16). Cox-2 mRNA expression in WT BM[RIGHTWARDS ARROW]WT mice (n = 9) was also significantly higher than TLR4−/− BM[RIGHTWARDS ARROW]TLR4−/− mice (n = 8) or WT BM[RIGHTWARDS ARROW]TLR4−/− mice. Data are represented as mean ± SEM of relative values of expression (*P < 0.05). B: PGE2 production by colonic mucosa in the AOM-DSS model. There is a numerical difference but not significant difference between TLR4−/− BM[RIGHTWARDS ARROW]WT mice (n = 10) and WT BM[RIGHTWARDS ARROW]TLR4−/− mice (n = 16). Significant differences were found only between WT BM[RIGHTWARDS ARROW]WT mice (n = 9) and TLR4−/− BM[RIGHTWARDS ARROW]TLR4−/− mice (n = 8). Data are shown as mean ± SEM (*P < 0.05). C: Double immunostaining for Cox-2 (green, FITC) and the macrophage marker CD68 (red, TRITC) in TLR4−/− BM[RIGHTWARDS ARROW]WT mice (top panel) and WT BM[RIGHTWARDS ARROW]TLR4−/− mice (bottom panel) mucosa. The majority of Cox-2 expressing cells in the lamina propria are also positive for CD68.

Download figure to PowerPoint

TLR4-Dependent Chemokine Expression in Inflammatory Colorectal Cancer

Given that TLR4 signaling by CEC is associated with an inflammatory infiltrate consisting of neutrophils and macrophages, we hypothesized that chemokines induced by CEC may be responsible for recruitment of these inflammatory cells. CCL2, a monocyte chemotactic protein-1 (MCP-1), has been implicated in recruitment of TAMs in colorectal cancer and CCL2 can be stimulated by TLR4.32–35 KC, a major neutrophil chemoattractant, is also induced by TLR4 and has been implicated in cancer cell growth.32 We examined mucosal expression of CCL2 and KC mRNA by real-time PCR. Mucosal expression of both CCL2 and KC mRNA was higher in TLR4−/− BM[RIGHTWARDS ARROW]WT mice than in WT BM[RIGHTWARDS ARROW]TLR4−/− mice (Fig. 4A,B). Since WT BM[RIGHTWARDS ARROW]TLR4−/− mice had similar expression levels of CCL2 and KC mRNA when compared to TLR4−/− mice transplanted with TLR4−/− BM, we conclude that the CEC-derived TLR4 signal, but not the myeloid-derived TLR4 signal, is responsible for chemokine expression.

thumbnail image

Figure 4. TLR4 signaling by CEC induces the expression of macrophage and neutrophil chemoattractant genes in the AOM-DSS model. A: CCL2 expression in the colonic mucosa measured by real-time PCR. Samples were taken at day 77 in the AOM-DSS model. A significant difference is shown between TLR4−/− BM[RIGHTWARDS ARROW]WT mice (n = 10) and WT BM[RIGHTWARDS ARROW]TLR4−/− mice (n = 16). CCL2 mRNA expression in WT BM[RIGHTWARDS ARROW]WT mice (n = 9) was also significantly higher than TLR4−/− BM[RIGHTWARDS ARROW]TLR4−/− mice (n = 8) or WT BM[RIGHTWARDS ARROW]TLR4−/− mice. Data are represented as mean ± SEM of relative values of expression (*P < 0.05). B: Real-time PCR for KC mRNA expression in the colonic mucosa. A significant difference is shown between TLR4−/− BM[RIGHTWARDS ARROW]WT mice (n = 10) and WT BM[RIGHTWARDS ARROW]TLR4−/− mice (n = 16). KC mRNA expression in WT BM[RIGHTWARDS ARROW]WT mice (n = 9) was also significantly higher than TLR4−/− BM[RIGHTWARDS ARROW]TLR4−/− mice (n = 8) or WT BM[RIGHTWARDS ARROW]TLR4−/− mice. Data are represented as mean ± SEM of relative values of expression (*P < 0.05).

Download figure to PowerPoint

Growth Factor Signaling in TLR4-Dependent Colitis-Associated Neoplasia

We have demonstrated above that TLR4 expression by CEC increases tumor incidence and size using a BM chimera model. EGFR activation occurs in colorectal cancer and inhibition of this pathway is effective for treatment of cancer.36 Amphiregulin (AR) is an EGF-like molecule implicated in colorectal cancer.37 Both intestinal epithelial cells and myofibroblasts have been shown to produce AR.38, 39 Using our current system, we wished to address which TLR4-expressing cell type is responsible for expression of AR. Mucosal AR expression in TLR4−/− BM[RIGHTWARDS ARROW]WT mice was significantly higher than in WT BM[RIGHTWARDS ARROW]TLR4−/− mice and was similar to AR expression in WT mice transplanted with WT BM (Fig. 5A). TLR4−/− mice even when transplanted with TLR4 positive-myeloid cells, did not induce AR. EGFR activation as measured by Western blot mirrored the results of the AR experiments (Fig. 5B). We conclude that TLR4 signaling by CEC is required for AR induction in a model of colorectal neoplasia and results in activation of the EGFR pathway.

thumbnail image

Figure 5. TLR4 signaling by CEC regulates mucosal amphiregulin expression and EGFR activation. A: AR expression in the colon of AOM-DSS model. AR mRNA expression was analyzed by real-time PCR in the samples at day 77 in the AOM-DSS model. A significant difference is shown between TLR4−/− BM[RIGHTWARDS ARROW]WT mice (n = 10) and WT BM[RIGHTWARDS ARROW]TLR4−/− mice (n = 16). AR mRNA expression in WT BM[RIGHTWARDS ARROW]WT mice (n = 9) was also significantly higher than TLR4−/− BM[RIGHTWARDS ARROW]TLR4−/− mice (n = 8) or WT BM[RIGHTWARDS ARROW]TLR4−/− mice. Data are represented as mean ± SEM of relative values of expression (*P < 0.05). B: Western blot analysis of phosphorylated EGFR and total EGFR in the colon. Results from representative samples obtained from 2 mice per condition are shown. 25 μg/lane of protein was loaded per lane. The membrane was sequentially probed for phospho-EGFR and total EGFR. Phospho-EGFR band intensity was quantified by normalizing with the corresponding EGFR band intensity.

Download figure to PowerPoint


  1. Top of page
  2. Abstract

Although inflammation has long been recognized to engender cancer, we have not been able to harness a rational approach to its prevention or treatment. In the case of IBD, less inflammation should mean less cancer and, by corollary, more effective antiinflammatory treatments should mean less cancer. Studies have suggested a modest chemopreventive effect of 5-aminosalicylates.40 Part of the reason we may not see a better chemopreventive effect from treatment may be that previous episodes of severe inflammation cannot be undone. At the root of the problem may be coexistence with intestinal bacteria and persistent innate immune activation.41, 42 Our laboratory has tried to tease apart the contribution of innate immune signaling in colitis-associated cancer using a reductionist approach. Our previous data using human CAC tissue revealed that TLR4 is upregulated in cancer tissue compared with the surrounding mucosa. In an animal model, the complete absence of TLR4 protects the animals from developing inflammation-induced colorectal tumors.10 These studies could not ascertain which TLR4-expressing cell type was responsible for the oncogenic phenotype.

Our study is the first to show that CEC expression of TLR4 is more important for the development of inflammation-induced tumors than myeloid expression of TLR4. This is surprising, given that epithelial expression of TLR4 is thought to be low under normal conditions.43, 44 During acute intestinal inflammation in mice and humans, TLR4 expression is increased—but these studies could not address whether function is increased.45, 46 In particular, previous studies have demonstrated increased TLR4 expression on lamina propria macrophages in IBD but could not address whether this is important for inflammation or inflammation-induced cancer.47

We sought to determine the function of TLR4 in the intestinal mucosa and address an important issue: is TLR signaling on CEC or in the myeloid compartment necessary for inflammation-induced neoplasia? Like the intestine itself, the answer is multilayered. Previous studies by Karin et al14 have shown that NF-κB activation by epithelial cells increased tumor numbers but that myeloid NF-κB activation was more important for tumor size. Although TLR4 activates NF-κB, the Venn diagram of genes induced by NF-κB and TLR4 signaling are distinct and merit their own careful investigation. Using BM chimeras, investigators have found that MyD88-expressing myeloid cells such as macrophages are required for epithelial cell proliferation in DSS-induced mucosal damage.20, 48 Again using bone marrow chimeras, Mizoguchi et al49 have recently shown that TNF receptor 1 on myeloid cells protects the colonic epithelium from DSS-induced damage. Myeloid expression of the TNF receptor 1 is also important for tumorigenesis in the AOM-DSS model.15 These data point to an interrelationship between the epithelium and the lamina propria myeloid compartment for CEC proliferation and ultimately neoplasia. In our work, the TLR4-responding cell type appears to be the CEC but the myeloid compartment, likely through TNF and other factors, promotes tumor growth.

Mechanistically, the CEC TLR4 signal leads to chemokine expression, recruitment of neutrophils, and activation of macrophages, and EGFR activation. It is not necessary in our model for TLR4 to be expressed by the macrophages themselves in order for Cox-2 to be induced. On the other hand, mice with TLR4 expression only in the myeloid cells still develop tumors, but at significantly reduced numbers. These mice had levels of mucosal PGE2 similar to mice expressing TLR4 in CEC. Thus, the myeloid TLR4 signal contributes to the development of neoplasia by inducing mucosal PGE2. In addition, both Cox-1 and Cox-2 contribute to total levels of PGE2 in the intestinal mucosa, which may explain why the PGE2 production is not entirely reflected in differences in Cox-2 expression.

Although the primary role of TLRs in the periphery is to eradicate pathogens, the intestine has made adaptations to coexist with commensal bacteria. In fact, our work and others have highlighted the importance of TLR signaling in epithelial homeostasis.18–20 Proliferation and repair of the epithelium in response to bacterial invasion of the mucosa is intrinsic for the host to survive in the presence of potentially lethal amounts of intestinal bacteria. For patients with IBD, the presence of luminal bacteria results in persistent innate immune activation. Innate immune signaling is thus a double-edged sword. Our work demonstrates that TLR4 signaling in discrete compartments in the intestine has distinct effects as they relate to inflammation-induced cancer. TLR4 signaling by the epithelium shapes the tumor microenvironment to optimally sustain the growth of a tumor. These observations can guide strategies directed at turning down the volume of TLR signaling in the epithelium to decrease the incidence of CAC.


  1. Top of page
  2. Abstract
  • 1
    Gyde SN, Prior P, Allan RN, et al. Colorectal cancer in ulcerative colitis: a cohort study of primary referrals from three centres. Gut. 1988; 29: 206217.
  • 2
    Rhodes JM, Campbell BJ. Inflammation and colorectal cancer: IBD-associated and sporadic cancer compared. Trends Mol Med. 2002; 8: 1016.
  • 3
    Choi PM, Zelig MP. Similarity of colorectal cancer in Crohn's disease and ulcerative colitis: implications for carcinogenesis and prevention. Gut. 1994; 35: 950954.
  • 4
    Eaden JA, Abrams KR, Mayberry JF. The risk of colorectal cancer in ulcerative colitis: a meta-analysis. Gut. 2001; 48: 526535.
  • 5
    Itzkowitz SH, Harpaz N. Diagnosis and management of dysplasia in patients with inflammatory bowel diseases. Gastroenterology. 2004; 126: 16341648.
  • 6
    Karlen P, Kornfeld D, Brostrom O, et al. Is colonoscopic surveillance reducing colorectal cancer mortality in ulcerative colitis? A population based case control study. Gut. 1998; 42: 711714.
  • 7
    Itzkowitz SH. Cancer prevention in patients with inflammatory bowel disease. Gastroenterol Clin North Am. 2002; 31: 11331144.
  • 8
    Rutter M, Saunders B, Wilkinson K, et al. Severity of inflammation is a risk factor for colorectal neoplasia in ulcerative colitis. Gastroenterology. 2004; 126: 451459.
  • 9
    Beg AA, Sha WC, Bronson RT, et al. Embryonic lethality and liver degeneration in mice lacking the RelA component of NF-kappa B. Nature. 1995; 376: 167170.
  • 10
    Fukata M, Chen A, Vamadevan AS, et al. Toll-like receptor-4 promotes the development of colitis-associated colorectal tumors. Gastroenterology. 2007; 133: 18691881.
  • 11
    Gupta RB, Harpaz N, Itzkowitz S, et al. Histologic inflammation is a risk factor for progression to colorectal neoplasia in ulcerative colitis: a cohort study. Gastroenterology. 2007; 133: 10991105; quiz 1340–1341.
  • 12
    Clapper ML, Cooper HS, Chang WC. Dextran sulfate sodium-induced colitis-associated neoplasia: a promising model for the development of chemopreventive interventions. Acta Pharmacol Sin. 2007; 28: 14501459.
  • 13
    Suzuki R, Kohno H, Sugie S, et al. Strain differences in the susceptibility to azoxymethane and dextran sodium sulfate-induced colon carcinogenesis in mice. Carcinogenesis. 2006; 27: 162169.
  • 14
    Greten FR, Eckmann L, Greten TF, et al. IKKbeta links inflammation and tumorigenesis in a mouse model of colitis-associated cancer. Cell. 2004; 118: 285296.
  • 15
    Popivanova BK, Kitamura K, Wu Y, et al. Blocking TNF-alpha in mice reduces colorectal carcinogenesis associated with chronic colitis. J Clin Invest. 2008; 118: 560570.
  • 16
    Karin M. NF-kappaB and cancer: mechanisms and targets. Mol Carcinog. 2006; 45: 355361.
  • 17
    Pikarsky E, Porat RM, Stein I, et al. NF-kappaB functions as a tumour promoter in inflammation-associated cancer. Nature. 2004; 431: 461466.
  • 18
    Rakoff-Nahoum S, Paglino J, Eslami-Varzaneh F, et al. Recognition of commensal microflora by toll-like receptors is required for intestinal homeostasis. Cell. 2004; 118: 229241.
  • 19
    Fukata M, Michelsen KS, Eri R, et al. Toll-like receptor-4 is required for intestinal response to epithelial injury and limiting bacterial translocation in a murine model of acute colitis. Am J Physiol Gastrointest Liver Physiol. 2005; 288: G10551065.
  • 20
    Pull SL, Doherty JM, Mills JC, et al. Activated macrophages are an adaptive element of the colonic epithelial progenitor niche necessary for regenerative responses to injury. Proc Natl Acad Sci U S A. 2005; 102: 99104.
  • 21
    Engle SJ, Ormsby I, Pawlowski S, et al. Elimination of colon cancer in germ-free transforming growth factor beta 1-deficient mice. Cancer Res. 2002; 62: 63626366.
  • 22
    Laqueur GL, Matsumoto H, Yamamoto RS. Comparison of the carcinogenicity of methylazoxymethanol-beta-D-glucosiduronic acid in conventional and germfree Sprague-Dawley rats. J Natl Cancer Inst. 1981; 67: 10531055.
  • 23
    Xiao H, Gulen MF, Qin J, et al. The Toll-interleukin-1 receptor member SIGIRR regulates colonic epithelial homeostasis, inflammation, and tumorigenesis. Immunity. 2007; 26: 461475.
  • 24
    Garlanda C, Riva F, Veliz T, et al. Increased susceptibility to colitis-associated cancer of mice lacking TIR8, an inhibitory member of the interleukin-1 receptor family. Cancer Res. 2007; 67: 60176021.
  • 25
    Fukata M, Chen A, Klepper A, et al. Cox-2 is regulated by Toll-like receptor-4 (TLR4) signaling: role in proliferation and apoptosis in the intestine. Gastroenterology. 2006; 131: 862877.
  • 26
    Morteau O, Morham SG, Sellon R, et al. Impaired mucosal defense to acute colonic injury in mice lacking cyclooxygenase-1 or cyclooxygenase-2. J Clin Invest. 2000; 105: 469478.
  • 27
    Pollard JW. Tumour-educated macrophages promote tumour progression and metastasis. Nat Rev Cancer. 2004; 4: 7178.
  • 28
    Sickert D, Aust DE, Langer S, et al. Characterization of macrophage subpopulations in colon cancer using tissue microarrays. Histopathology. 2005; 46: 515521.
  • 29
    Hauptmann S, Zwadlo-Klarwasser G, Hartung P, et al. Association of different macrophage phenotypes with infiltrating and non-infiltrating areas of tumor-host interface in colorectal carcinoma. Pathol Res Pract. 1994; 190: 159167.
  • 30
    van der Woude CJ, Kleibeuker JH, Jansen PL, et al. Chronic inflammation, apoptosis and (pre-)malignant lesions in the gastro-intestinal tract. Apoptosis. 2004; 9: 123130.
  • 31
    Agoff SN, Brentnall TA, Crispin DA, et al. The role of cyclooxygenase 2 in ulcerative colitis-associated neoplasia. Am J Pathol. 2000; 157: 737745.
  • 32
    Son DS, Parl AK, Rice VM, et al. Keratinocyte chemoattractant (KC)/human growth-regulated oncogene (GRO) chemokines and pro-inflammatory chemokine networks in mouse and human ovarian epithelial cancer cells. Cancer Biol Ther. 2007; 6: 13021312.
  • 33
    Khan MA, Ma C, Knodler LA, et al. Toll-like receptor 4 contributes to colitis development but not to host defense during Citrobacter rodentium infection in mice. Infect Immun. 2006; 74: 25222536.
  • 34
    Walter S, Bottazzi B, Govoni D, et al. Macrophage infiltration and growth of sarcoma clones expressing different amounts of monocyte chemotactic protein/JE. Int J Cancer. 1991; 49: 431435.
  • 35
    Bottazzi B, Walter S, Govoni D, et al. Monocyte chemotactic cytokine gene transfer modulates macrophage infiltration, growth, and susceptibility to IL-2 therapy of a murine melanoma. J Immunol. 1992; 148: 12801285.
  • 36
    Chua YJ, Cunningham D. Recent data with anti-epidermal growth factor receptor antibodies and irinotecan in colon cancer. Clin Colorectal Cancer. 2005; 5 Suppl 2: S8188.
  • 37
    Shao J, Lee SB, Guo H, et al. Prostaglandin E2 stimulates the growth of colon cancer cells via induction of amphiregulin. Cancer Res. 2003; 63: 52185223.
  • 38
    Saeki T, Stromberg K, Qi CF, et al. Differential immunohistochemical detection of amphiregulin and cripto in human normal colon and colorectal tumors. Cancer Res. 1992; 52: 34673473.
  • 39
    Shao J, Sheng GG, Mifflin RC, et al. Roles of myofibroblasts in prostaglandin E2-stimulated intestinal epithelial proliferation and angiogenesis. Cancer Res. 2006; 66: 846855.
  • 40
    Velayos FS, Terdiman JP, Walsh JM. Effect of 5-aminosalicylate use on colorectal cancer and dysplasia risk: a systematic review and metaanalysis of observational studies. Am J Gastroenterol. 2005; 100: 13451353.
    Direct Link:
  • 41
    Eckmann L. Innate immunity and mucosal bacterial interactions in the intestine. Curr Opin Gastroenterol. 2004; 20: 8288.
  • 42
    Collier-Hyams LS, Neish AS. Innate immune relationship between commensal flora and the mammalian intestinal epithelium. Cell Mol Life Sci. 2005; 62: 13391348.
  • 43
    Otte JM, Cario E, Podolsky DK. Mechanisms of cross hyporesponsiveness to Toll-like receptor bacterial ligands in intestinal epithelial cells. Gastroenterology. 2004; 126: 10541070.
  • 44
    Abreu MT, Vora P, Faure E, et al. Decreased expression of Toll-like receptor-4 and MD-2 correlates with intestinal epithelial cell protection against dysregulated proinflammatory gene expression in response to bacterial lipopolysaccharide. J Immunol. 2001; 167: 16091616.
  • 45
    Cario E, Podolsky DK. Differential alteration in intestinal epithelial cell expression of toll-like receptor 3 (TLR3) and TLR4 in inflammatory bowel disease. Infect Immun. 2000; 68: 70107017.
  • 46
    Ohkawara T, Takeda H, Miyashita K, et al. Regulation of Toll-like receptor 4 expression in mouse colon by macrophage migration inhibitory factor. Histochem Cell Biol. 2006; 125: 575582.
  • 47
    Hausmann M, Kiessling S, Mestermann S, et al. Toll-like receptors 2 and 4 are up-regulated during intestinal inflammation. Gastroenterology. 2002; 122: 19872000.
  • 48
    Brown SL, Riehl TE, Walker MR, et al. Myd88-dependent positioning of Ptgs2-expressing stromal cells maintains colonic epithelial proliferation during injury. J Clin Invest. 2007; 117: 258269.
  • 49
    Mizoguchi E, Hachiya Y, Kawada M, et al. TNF receptor type I-dependent activation of innate responses to reduce intestinal damage-associated mortality. Gastroenterology. 2008; 134: 470480.