Ornithine decarboxylase (ODC) is the rate-limiting enzyme for polyamine biosynthesis and restricts M1 macrophage activation in gastrointestinal (GI) infections. However, the role of macrophage ODC in colonic epithelial-driven inflammation is unknown. Here, we investigate cell-specific effects of ODC in colitis and colitis-associated carcinogenesis (CAC). Human colonic macrophages expressed increased ODC levels in active ulcerative colitis and Crohn's disease, colitis-associated dysplasia, and CAC. Mice lacking Odc in myeloid cells (OdcΔmye mice) that were treated with dextran sulfate sodium (DSS) exhibited improved survival, body weight, and colon length and reduced histologic injury versus control mice. In contrast, GI epithelial-specific Odc knockout had no effect on clinical parameters. Despite reduced histologic damage, colitis tissues of OdcΔmye mice had increased levels of multiple proinflammatory cytokines and chemokines and enhanced expression of M1, but not M2 markers. In the azoxymethane-DSS model of CAC, OdcΔmye mice had reduced tumor number, burden, and high-grade dysplasia. Tumors from OdcΔmye mice had increased M1, but not M2 macrophages. Increased levels of histone 3, lysine 9 acetylation, a marker of open chromatin, were manifest in tumor macrophages of OdcΔmye mice, consistent with our findings that macrophage ODC affects histone modifications that upregulate M1 gene transcription during GI infections. These findings support the concept that macrophage ODC augments epithelial injury-associated colitis and CAC by impairing the M1 responses that stimulate epithelial repair, antimicrobial defense, and antitumoral immunity. They also suggest that macrophage ODC is an important target for colon cancer chemoprevention.

Significance: Ornithine decarboxylase contributes to the pathogenesis of colitis and associated carcinogenesis by impairing M1 macrophage responses needed for antitumoral immunity; targeting ODC in macrophages may represent a new strategy for chemoprevention. Cancer Res; 78(15); 4303–15. ©2018 AACR.

M1 macrophage activation is essential in the host defense response to microbes and to tumor cells (1, 2). M2 macrophages have a key role in parasite defense, and are associated with wound healing (3). There has been a long interest in the role of M1 versus M2 macrophages in the tumor microenvironment (2, 4), but as yet this has not led to therapeutic interventions.

In inflammatory bowel disease (IBD), there is a strong risk for colitis-associated carcinogenesis (CAC) that results from the chronic inflammatory state (5, 6). Indeed, CAC occurs in 20% of patients with IBD, and mortality rates can exceed 50% (7, 8). However, despite this clinical problem, the mainstay of clinical management is annual surveillance colonoscopy. Accordingly, insight into the pathogenesis of CAC, and new therapeutic strategies, are urgently needed. In particular, the role of the host immune response in CAC, and especially the roles of M1 and M2 macrophages are not understood.

We have investigated the role of macrophage activation patterns, often referred to M1 versus M2 polarization, in host–pathogen interactions in the gastrointestinal (GI) mucosa (1, 9, 10). One key regulatory component of macrophage function revealed by these studies has been the role of polyamines (10, 11). Polyamine biosynthesis begins with the production of the biogenic amine, putrescine, which is produced by the enzyme ornithine decarboxylase (ODC1, hereafter ODC; refs. 12, 13). Putrescine is then converted to the polyamines spermidine and spermine by the action of constitutively active synthases (13). We recently reported that the myeloid-specific deletion of Odc (OdcΔmye) provokes a marked increase in the response of mouse bone marrow-derived macrophages to M1 stimuli including Helicobacter pylori, Citrobacter rodentium, or LPS + IFNγ, and increased expression of M1 genes (10). In addition, in mouse models of gastric inflammation induced by H. pylori infection, and of colon inflammation induced by C. rodentium infection, OdcΔmye mice have increased M1 responses and production of M1 cytokines and chemokines. However, when such bone marrow-derived OdcΔmye macrophages are exposed to M2 stimulation with IL4, there is also an upregulation of M2 gene expression (10). Thus, the ODC-polyamine circuit can harness both M1 and M2 macrophage responses. Finally, these effects were linked to histone modifications (10), where OdcΔmye macrophages have increases in histone marks that are hallmarks of open chromatin (e.g., H3K4 monomethylation and H3K9 acetylation), leading to increased transcription (10, 14).

Macrophage-driven innate immune responses are major contributors to the pathogenesis of human ulcerative colitis (UC; refs. 15–17) and to related mouse models of colitis and CAC (1, 16, 18). Given this, we assessed the role of ODC in UC and CAC. Herein, we report that ODC expression is increased in human macrophages of IBD tissues displaying active colitis, dysplasia, and CAC. Strikingly, in both dextran sulfate sodium (DSS)-induced colitis and the azoxymethane (AOM)-DSS model of CAC, myeloid-selective loss of Odc reduces clinical disease and tumorigenesis, and this is linked to increases in the M1 response. Thus, in both an epithelial injury model and in associated carcinogenesis, M1 macrophage functions are beneficial, and ODC acts to impair this response, providing new insights into strategies for colon cancer prevention.

Animal studies

Animal experiments were approved by the Vanderbilt University Institutional Animal Care and Use Committee and the Research and Development Committee of the Veterans Affairs Tennessee Valley Healthcare System. Odcfl/fl mice, and OdcΔmye mice generated by crossing Odcfl/fl with Lyz2cre/cre mice, were previously described (10). OdcΔepi mice were generated by crossing Odcfl/fl mice with Foxa3cre/+ mice (19, 20). Male mice between the ages of 7 and 12 weeks were used for all studies.

For the acute DSS colitis model, animals were treated with 2.5% DSS in the drinking water for 5 days followed by a 5-day recovery period. Control mice received normal drinking water for 10 days. For the colitis-associated cancer (CAC) model, mice first received one intraperitoneal injection of AOM (12.5 mg/kg) and were then treated three times with 4% DSS for 5 days beginning at days 5, 26, and 47, all as described (20).

Animals were monitored daily and mice that showed extreme distress, became moribund, or lost more than 20% of initial body weight were euthanized. After sacrifice, colons were removed, measured, cut longitudinally, cleaned, weighed, and Swiss-rolled for histology. Two proximal and distal 2 mm pieces were used for RNA and protein extraction. All histologic assessments were made in a blinded manner by a pathologist (M.K.W.). In the AOM-DSS model, tumors were counted and measured under a dissecting microscope; tumor and nontumor tissues were collected for RNA and protein extraction. Colons were then Swiss-rolled for histology.

Isolation of myeloid cells and RNA analyses

RNA from colonic tissues was extracted using the RNeasy Mini Kit (Qiagen; ref. 21). CD11b+, CD11c+, F4/80+, and Gr1+ cells were positively selected (21) using biotinylated antibodies, streptavadin conjugated with magnetic beads, and IMagnet columns (BD Biosciences). RNA from cells was isolated using the 5 PRIME PerfectPure RNA 96 Cell CS Kit (21). Total RNA (1 μg) was reverse transcribed using an iScript cDNA Synthesis Kit (Bio-Rad) and oligo(dT) primers. Each real-time reverse transcription (RT)-PCR was performed with 2 μL of cDNA, iQ SYBR Green Supermix (Bio-Rad), and primers in Supplementary Table S1.

Immunofluorescence and immunohistochemistry

Sections (5 μm) from formalin-fixed paraffin-embedded Swiss-rolled murine tissues were deparaffinized and treated with citrate buffer for antigen retrieval. An IBD-associated cancer tissue microarray (TMA) from Vanderbilt University Medical Center, which contained cases of normal tissue, inactive and active UC and Crohn's disease, dysplasia, and CAC, was used to assess ODC expression. Immunofluorescence for CD68, ODC, nitric oxide synthase 2 (NOS2), arginase 1 (ARG1), and H3K9Ac, and IHC for CD8 and FOXP3 was performed as described (1, 10). Antibodies are listed in Supplementary Table S2. Quantification of immunofluorescence was performed on images generated by the Aperio Versa platform (Leica Biosystems) of the Vanderbilt Digital Histology Shared Resource using CellProfiler software (1, 20).

Measurement of cytokines and chemokines

Colonic tissues were lysed in Cell Lytic Mammalian Tissue Lysis Extraction Reagent (Sigma) and analyzed using the 25-analyte MILLIPLEX MAP Mouse Cytokine/Chemokine Magnetic Bead Panel (EMD Millipore) on a FLEXMAP 3D instrument (Luminex; refs. 21, 22). Data were standardized to tissue protein concentrations measured by the BCA Protein Assay Kit (Pierce). TNFα and IFNγ in tissues was measured using R&D DuoSet ELISA Kits (10).

Flow cytometry

Immune cells from nontumor and tumor areas were isolated using Liberase DL and TL (23). Cells were stained for CD11b (Supplementary Table S2), washed, fixed, and permeabilized with CytoFix/CytoPerm (BD Biosciences). NOS2 or ARG1 antibodies were used (Supplementary Table S2). Data acquisition was performed using FlowJo software (Tree Star).

Macrophage differentiation and measurement of polyamines

Bone-marrow-derived macrophages (BMmacs) were differentiated, placed in DMEM supplemented with 2% BSA, and activated with IFNγ (200 units/mL) and LPS (10 ng/mL) for 24 hours (10). Polyamines were quantified by LC/MS in cell pellets (10).

Statistical analysis

Data are shown as the mean ± SEM. Statistical analyses were performed with Prism version 5.0c (GraphPad Software). Student t test was performed for comparison between two groups. ANOVA with the Student–Newman–Keuls post hoc multiple comparisons test was performed to compare multiple groups. When data were not normally distributed, log transformation was performed. Survival was analyzed by the log-rank (Mantel–Cox) test.

Patients with active UC and CAC have increased ODC expression in macrophages

To assess macrophage infiltration and ODC expression in patients with UC, as well as precancerous and cancerous colon lesions, a TMA from Vanderbilt University Medical Center was utilized. Histopathologic diagnoses were established from clinical case material from surgical resections, and confirmed by hematoxylin and eosin staining of the cores on the TMA. Immunofluorescent staining was performed for the macrophage marker CD68, as well as ODC. Notably, this demonstrated that high levels of ODC are expressed in macrophages found in active colitis, dysplasia, and carcinoma (Fig. 1A). When scored by whole slide scanning and CellProfiler image analysis, there were marked increases in the percentage of CD68+ODC+ cells per core (Fig. 1B), as well as an increase in the percentage of CD68+ODC+ cells per CD68+ cells (Fig. 1C) in the colon of patients with active colitis, dysplasia, and CAC compared with tissues from normal individuals or to tissues from inactive UC. ODC staining was also detected in epithelial cells; however, there was no significant increase in ODC levels in CD68 cells in colitis, dysplasia, or CAC (Supplementary Fig. S1A). CD68+ODC+ cells were also increased in active Crohn's disease tissues (Supplementary Fig. S1B and S1C). Thus, ODC is selectively overexpressed in macrophages during colonic inflammation in IBD, and in patients with precancerous and cancerous lesions.

Figure 1.

Staining of human colon TMA. A, Representative hematoxylin and eosin staining and immunofluorescence for CD68 (red), ODC (green), and nuclei (DAPI, blue) in control and UC tissues from the TMA, depicted in yellow in these merged images. Boxes indicate areas of the images shown in higher magnification. Arrowheads, cells double-positive for CD68 and ODC. Scale bars, 50 μm. B and C, Quantification of CD68+ODC+ cells among the total number of cells in each individual core (B), or among the total number of CD68+ cells in each individual core (C), was determined using CellProfiler software. Normal subjects: 5 tissues were from the uninvolved normal area (open circles) and one tissue (red circle) was from a normal control case. For B and C, **, P < 0.01 and ***, P < 0.001 versus Normal; §, P < 0.05, §§, P < 0.01, and §§§, P < 0.001 compared with inactive colitis.

Figure 1.

Staining of human colon TMA. A, Representative hematoxylin and eosin staining and immunofluorescence for CD68 (red), ODC (green), and nuclei (DAPI, blue) in control and UC tissues from the TMA, depicted in yellow in these merged images. Boxes indicate areas of the images shown in higher magnification. Arrowheads, cells double-positive for CD68 and ODC. Scale bars, 50 μm. B and C, Quantification of CD68+ODC+ cells among the total number of cells in each individual core (B), or among the total number of CD68+ cells in each individual core (C), was determined using CellProfiler software. Normal subjects: 5 tissues were from the uninvolved normal area (open circles) and one tissue (red circle) was from a normal control case. For B and C, **, P < 0.01 and ***, P < 0.001 versus Normal; §, P < 0.05, §§, P < 0.01, and §§§, P < 0.001 compared with inactive colitis.

Close modal

Macrophage ODC contributes to DSS-induced colonic inflammation

To investigate the relative contribution of ODC in macrophages and epithelial cells to UC and CAC, we used validated mouse experimental colitis and CAC models. Odc expression was elevated in colonic tissues of mice with DSS-induced colitis, and AOM-DSS–induced CAC lesions (Supplementary Fig. S2A and S2B). There was a significant increase of Odc mRNA levels in cells positive for the markers CD11b (myeloid cells) and F4/80 (macrophages), but not CD11c (dendritic cells) or Gr1 (neutrophils and myeloid-derived suppressor cells) from the colon of DSS-treated mice (Supplementary Fig. S2C). To further assess the role of macrophage ODC in an epithelial injury and repair model of colitis induced by DSS, we used OdcΔmye (Odcfl/fl;Lyz2cre/cre) mice, in which we demonstrated loss of ODC in macrophages (10). At baseline, colons of OdcΔmye mice are histologically normal, and do not differ from the control Odcfl/fl mice (Supplementary Fig. S3A). There were no changes in ODC expression in the epithelium of OdcΔmye versus Odcfl/fl mice (Supplementary Fig. S3B). Loss of ODC activity in OdcΔmye mice was verified by measuring polyamines in BMmacs. Upon stimulation with IFNγ + LPS, BMmacs from Odcfl/fl, but not OdcΔmye mice, generated putrescine, the direct product of ODC; spermidine and spermine were not modulated by the M1 stimulus (Supplementary Fig. S3C). In mice treated with DSS for 5 days followed by 5 days of water, OdcΔmye mice exhibited a significant improvement in survival (Fig. 2A). Odcfl/fl mice treated with DSS began to die on day 7 and by day 10, only 52% remained alive, whereas 88% of OdcΔmye mice were alive at day 10. Body weight loss of DSS-treated OdcΔmye mice was also significantly improved compared with DSS-treated Odcfl/fl mice (Fig. 2B), and colon length was less reduced in OdcΔmye versus Odcfl/fl mice (Fig. 2C). Finally, histologic colitis was significantly improved in OdcΔmye mice compared with Odcfl/fl mice (Fig. 2D and E).

Figure 2.

Effect of myeloid-specific Odc deletion on DSS-induced colitis. Mice were treated with 2.5% DSS for 5 days and given normal drinking water for the next 5 days. A, Survival was monitored daily. The Kaplan–Meier plot was generated from two separate experiments pooled together; n = 22 Odcfl/fl mice and n = 22 OdcΔmye mice. For this representation of the data, survival is defined as body weight loss of more than 20%. No death was observed in untreated animals. B, Body weights, presented as the percentage of initial body weight. *, P < 0.05 and **, P < 0.01 compared with DSS-treated Odcfl/fl animals. C, Colon length; *, P < 0.05 compared with DSS-treated Odcfl/fl mice. D and E, Colons were Swiss-rolled and scored for histologic injury (D) from tissues stained with hematoxylin and eosin (E); *, P < 0.05 compared with DSS-treated Odcfl/fl mice. Scale bars, 100 μm. F, Immunofluorescence for CD68 (red), ODC (green), and nuclei (blue) in colonic tissues; CD68+ODC+ cells in the merged images are depicted in yellow. Scale bar, 100 μm.

Figure 2.

Effect of myeloid-specific Odc deletion on DSS-induced colitis. Mice were treated with 2.5% DSS for 5 days and given normal drinking water for the next 5 days. A, Survival was monitored daily. The Kaplan–Meier plot was generated from two separate experiments pooled together; n = 22 Odcfl/fl mice and n = 22 OdcΔmye mice. For this representation of the data, survival is defined as body weight loss of more than 20%. No death was observed in untreated animals. B, Body weights, presented as the percentage of initial body weight. *, P < 0.05 and **, P < 0.01 compared with DSS-treated Odcfl/fl animals. C, Colon length; *, P < 0.05 compared with DSS-treated Odcfl/fl mice. D and E, Colons were Swiss-rolled and scored for histologic injury (D) from tissues stained with hematoxylin and eosin (E); *, P < 0.05 compared with DSS-treated Odcfl/fl mice. Scale bars, 100 μm. F, Immunofluorescence for CD68 (red), ODC (green), and nuclei (blue) in colonic tissues; CD68+ODC+ cells in the merged images are depicted in yellow. Scale bar, 100 μm.

Close modal

Immunofluorescence analyses established that ODC protein expression was markedly elevated in DSS-treated Odcfl/fl versus untreated Odcfl/fl mice (Fig. 2F). Further, ODC expression was predominantly in macrophages, as evidenced by colocalization to cells marked by CD68 staining (Fig. 2F, left). As expected, ODC staining of CD68+ cells was absent in the colon tissues of DSS-treated OdcΔmye mice (Fig. 2F, right). There was modest ODC staining of CD68 colonic epithelial cells (Fig. 2F, right), which was not affected by myeloid-specific deletion of Odc. These data verify the specific knockout of ODC in colonic macrophages during DSS colitis.

Because ODC is also expressed in the colonic epithelial cells of patients with UC and CAC, we assessed the contribution of epithelial-derived ODC to the development of DSS-induced colitis, using OdcΔepi mice. Expression of Odc mRNA (Supplementary Fig. S4A) and ODC protein (Supplementary Fig. S4B) was abolished in colonic epithelial cells from OdcΔepi versus Odcfl/fl mice. When subjected to DSS treatment, OdcΔepi mice exhibited no differences in body weight loss, colon length, or histologic injury compared with Odcfl/fl mice (Supplementary Fig. S4C–S4F).

ODC regulates the mucosal proinflammatory response during acute colitis

To gain further insights into the mechanism by which macrophage ODC contributes to the development of colitis, Luminex multiplex arrays were used to assess effects of ODC on cytokines and chemokines produced in the colonic mucosa. Despite the reduced histologic injury scores, in OdcΔmye versus Odcfl/fl colon tissues from DSS-treated mice there were significant increases (Fig. 3A) in the innate cytokines CSF2 (GM-CSF), TNFα, IL-1α, and IL-1β, as well as the prototype Th1 cytokine IFNγ, and the chemokines CXCL2 (MIP2), CXCL9 (MIG), CXCL10 (IP-10), CCL2 (MCP-1), CCL3 (MIP-1α), CCL4 (MIP-1β), and CCL5 (RANTES). Moreover, there was increased mRNA expression of the M1 markers Nos2 and Il1b in colitis tissues from DSS-treated OdcΔmye mice versus DSS-treated Odcfl/fl mice (Fig. 3B). To further investigate the source of cytokines, macrophages were isolated by positive selection and increased mRNA expression of Nos2, Tnfa, Il1b, Il12a, Ccl5, and Cxcl10 was demonstrated in DSS-treated OdcΔmye versus Odcfl/fl mice (Supplementary Fig. S5). In contrast, expression levels of the M2 markers, Arg1 and Chil3, were comparable in colonic tissues of DSS-treated Odcfl/fl and OdcΔmye mice (Fig. 3B).

Figure 3.

Quantification of cytokines and chemokines, as well as gene expression of M1 and M2 markers in murine colonic tissues. A, Concentrations of cytokines and chemokines in colon tissues were determined by Luminex assay. B, Analysis of Nos2, Il1b, Arg1, and Chil3 mRNA expression in colonic tissues. In A and B, *, P < 0.05, **, P < 0.01, and ***, P < 0.001 versus Odcfl/fl mice.

Figure 3.

Quantification of cytokines and chemokines, as well as gene expression of M1 and M2 markers in murine colonic tissues. A, Concentrations of cytokines and chemokines in colon tissues were determined by Luminex assay. B, Analysis of Nos2, Il1b, Arg1, and Chil3 mRNA expression in colonic tissues. In A and B, *, P < 0.05, **, P < 0.01, and ***, P < 0.001 versus Odcfl/fl mice.

Close modal

OdcΔmye mice are protected from CAC

Because OdcΔmye mice are protected in the DSS colitis model and macrophage ODC expression was detected in patients with CAC, we reasoned that macrophage ODC may play a role in colon carcinogenesis. OdcΔmye and Odcfl/fl mice were subjected to the 77-day AOM-DSS model of CAC. The survival rate of OdcΔmye mice was significantly improved compared with the Odcfl/fl animals on the AOM-DSS protocol (Fig. 4A). With the first and second cycles of DSS, Odcfl/fl and OdcΔmye mice showed similar weight loss and recovery with water (Fig. 4B), but with the third cycle of DSS, OdcΔmye mice had less weight loss than Odcfl/fl mice, and better improvement during the recovery period (Fig. 4B). Consistent with this difference in body weights, OdcΔmye mice had a significantly lower number of tumors per mouse (Fig. 4C) and lower overall tumor burden (Fig. 4D) than Odcfl/fl animals. However, there was no difference in histologic injury scores in the nontumor area between both type of mice (Fig. 4E), most likely due to the long-time gap between the last dose of DSS and sacrifice (28 days). Histologic assessment of the tumors revealed less high-grade dysplasia in OdcΔmye versus Odcfl/fl mice (Fig. 4F and G).

Figure 4.

Effect of myeloid-specific Odc deletion on AOM-DSS-induced CAC. A, Survival was monitored daily; the Kaplan–Meier plot was performed from two separate experiments pooled together; n = 14 Odcfl/fl mice and n = 11 OdcΔmye mice. No death was observed in control animals. B, Percentage of initial body weight was determined at the indicated time points. C, Tumor number, determined using a dissecting microscope. D, Tumor burden, determined by the addition of the area of each tumor. E, Histologic colitis. F, Percentage of cases with low-grade dysplasia (LGD) and high-grade dysplasia (HGD). G, Representative hematoxylin and eosin–stained images from AOM-DSS-treated mice. Scale bars, 100 μm. For B–F, *, P < 0.05, **, P < 0.01, and ***, P < 0.001 compared with Odcfl/fl mice treated with AOM-DSS.

Figure 4.

Effect of myeloid-specific Odc deletion on AOM-DSS-induced CAC. A, Survival was monitored daily; the Kaplan–Meier plot was performed from two separate experiments pooled together; n = 14 Odcfl/fl mice and n = 11 OdcΔmye mice. No death was observed in control animals. B, Percentage of initial body weight was determined at the indicated time points. C, Tumor number, determined using a dissecting microscope. D, Tumor burden, determined by the addition of the area of each tumor. E, Histologic colitis. F, Percentage of cases with low-grade dysplasia (LGD) and high-grade dysplasia (HGD). G, Representative hematoxylin and eosin–stained images from AOM-DSS-treated mice. Scale bars, 100 μm. For B–F, *, P < 0.05, **, P < 0.01, and ***, P < 0.001 compared with Odcfl/fl mice treated with AOM-DSS.

Close modal

Loss of Odc enhances macrophage polarization towards M1 during experimental CAC

The treatment of mice with AOM-DSS led to increases in mRNA expression of the M1 markers Nos2, Il1b, Il12a, Il12b, and Tnfa, as well as the Th1 marker Ifng, in both the nontumor and tumor areas (Fig. 5A). Strikingly, all of these markers were significantly more expressed in the tumors of the OdcΔmye versus Odcfl/fl mice (Fig. 5A). TNFα and IFNγ levels were measured and were also increased at the protein level in tumor tissues from OdcΔmye versus Odcfl/fl mice (Fig. 5B). Although mRNA expression of the M2 markers Arg1, Chil3, Tgfb, Vegfa, Il10, and Tnfsf14 were also generally upregulated in the nontumor and tumor tissues, there were no significant differences in their expression between the Odcfl/fl and OdcΔmye tumors (Supplementary Fig. S6).

Figure 5.

M1 profile in nontumor and tumor tissues. A, mRNA levels of M1 markers Nos2, Il1b, Il12a, Il12b, Tnfa, and Ifng assessed by real-time RT-PCR. B, TNFα and IFNγ were also analyzed by ELISA from colonic tissues 77 days post-AOM injection. *, P < 0.05, **, P < 0.01, and **, P < 0.001 compared with untreated Odcfl/fl mice; §, P < 0.05, §§, P < 0.01, and §§§, P < 0.001 compared with the tumors in Odcfl/fl mice.

Figure 5.

M1 profile in nontumor and tumor tissues. A, mRNA levels of M1 markers Nos2, Il1b, Il12a, Il12b, Tnfa, and Ifng assessed by real-time RT-PCR. B, TNFα and IFNγ were also analyzed by ELISA from colonic tissues 77 days post-AOM injection. *, P < 0.05, **, P < 0.01, and **, P < 0.001 compared with untreated Odcfl/fl mice; §, P < 0.05, §§, P < 0.01, and §§§, P < 0.001 compared with the tumors in Odcfl/fl mice.

Close modal

To confirm these findings, we analyzed by flow cytometry the expression of the prototype M1 and M2 proteins, NOS2 and ARG1, respectively, in intratumoral macrophages. The numbers of NOS2+CD11b+ cells were significantly higher in tumors from OdcΔmye mice versus tumors from Odcfl/fl animals (Fig. 6A), whereas no differences were observed in ARG1 expression in CD11b+ cells (Fig. 6A). The composite data from multiple animals demonstrated significant increases in the percentage of NOS2+CD11b+ cells (Fig. 6B), but not ARG1+CD11b+ cells (Fig. 6C), in the tumors of AOM-DSS-treated OdcΔmye mice versus tumors from Odcfl/fl mice. When quantified as number of cells per gram of colon tissue, a similar pattern of increased NOS2+ (Supplementary Fig. S7A), but not ARG1+ cells (Supplementary Fig. S7B) in the CD11b+ cells was demonstrated, resulting in an increased ratio of NOS2/ARG1 in the CD11b cells (Supplementary Fig. S7C). Similarly, by immunofluorescence, the level of ARG1 in CD68+ cells was not affected by Odc loss (Fig. 7A), whereas NOS2 was more abundant in CD68+ cells present in the tumors of OdcΔmye mice than in tumors of Odcfl/fl animals (Fig. 7B). There were no differences in staining for the Treg marker, FOXP3, or the cytotoxic T cell marker, CD8, in tumors of OdcΔmye versus Odcfl/fl mice (Supplementary Fig. S8).

Figure 6.

Flow cytometry analysis of NOS2 and ARG1 protein expression in colon tissues from Odcfl/fl and OdcΔmye mice. A, Representative flow plots showing NOS2 and ARG1 expression in CD11b+ cells within the immune cells isolated from colon tissues of Odcfl/fl and OdcΔmye mice treated with AOM-DSS. B and C, The number of NOS2+CD11b+ (B) and ARG1+CD11b+ (C) cells were counted and are expressed as a percent of CD11b+ cells. *, P < 0.05 versus the expression of NOS2 in tumors from Odcfl/fl mice.

Figure 6.

Flow cytometry analysis of NOS2 and ARG1 protein expression in colon tissues from Odcfl/fl and OdcΔmye mice. A, Representative flow plots showing NOS2 and ARG1 expression in CD11b+ cells within the immune cells isolated from colon tissues of Odcfl/fl and OdcΔmye mice treated with AOM-DSS. B and C, The number of NOS2+CD11b+ (B) and ARG1+CD11b+ (C) cells were counted and are expressed as a percent of CD11b+ cells. *, P < 0.05 versus the expression of NOS2 in tumors from Odcfl/fl mice.

Close modal
Figure 7.

Immunofluorescent detection of ARG1, NOS2, and H3K9ac in colonic tumors from Odcfl/fl and OdcΔmye mice. A tumor area of the colon of AOM-DSS–treated Odcfl/fl and OdcΔmye mice was used to determine the levels of ARG1 (A), NOS2 (B), or H3K9ac (C) in CD68+ cells by immunofluorescence. In each panel, CD68 is depicted in red, the protein of interest in green, and the nuclei in blue. CD68+ARG1+ and CD68+NOS2+ cells are shown in yellow, and the presence of H3K9ac in CD68+ cells is depicted in bright green/white. In AC, the box in the low power image denotes the area of the higher power image. Scale bars, 100 μm. In C, the open arrowheads indicate cells that are positive for CD68 and negative for H3K9ac; solid arrowheads, cells that are positive for both CD68 and H3K9ac. Quantification was performed using ImageJ and shown in the right panels. *, P < 0.05 for OdcΔmye versus Odcfl/fl mice.

Figure 7.

Immunofluorescent detection of ARG1, NOS2, and H3K9ac in colonic tumors from Odcfl/fl and OdcΔmye mice. A tumor area of the colon of AOM-DSS–treated Odcfl/fl and OdcΔmye mice was used to determine the levels of ARG1 (A), NOS2 (B), or H3K9ac (C) in CD68+ cells by immunofluorescence. In each panel, CD68 is depicted in red, the protein of interest in green, and the nuclei in blue. CD68+ARG1+ and CD68+NOS2+ cells are shown in yellow, and the presence of H3K9ac in CD68+ cells is depicted in bright green/white. In AC, the box in the low power image denotes the area of the higher power image. Scale bars, 100 μm. In C, the open arrowheads indicate cells that are positive for CD68 and negative for H3K9ac; solid arrowheads, cells that are positive for both CD68 and H3K9ac. Quantification was performed using ImageJ and shown in the right panels. *, P < 0.05 for OdcΔmye versus Odcfl/fl mice.

Close modal

In GI infection models, Odc loss in macrophages provokes increases in histone marks that are typical of open chromatin and that are associated with the increased expression of genes encoding M1 markers (NOS2 and cytokines), as described (10). Notably, macrophage levels of the euchromatin histone mark H3K9ac, which is associated with M1 gene expression (10), were augmented in the tumors of OdcΔmye mice versus those in Odcfl/fl animals (Fig. 7C).

The findings presented herein establish critical roles for macrophage ODC in colonic inflammation and associated colon tumorigenesis. Specifically, mice with myeloid-specific knockout of Odc have more robust cytokine and chemokine responses, yet are protected in terms of clinical parameters and histologic injury in the DSS injury-repair model of colitis. These suggest that macrophage ODC impairs responses to the injury induced by DSS, and suggest that increased macrophage ODC expression in UC may contribute to epithelial cell injury. In contrast, intestinal epithelial cell knockout of Odc does not result in significant changes in the level of DSS-induced colitis. Thus, polyamines derived from ODC have macrophage-selective effects on colonic injury and repair.

Because polyamines are well known to support cell proliferation, ODC has been a target to limit carcinogenesis (24) and/or to limit tumor progression (25), including in colon cancer prevention (26). Here we found that ODC expression is increased in the colonic macrophages of mice treated with DSS, but is not overexpressed in the epithelium in this model. In colitis tissues, immunofluorescence demonstrated strong double-staining for ODC and macrophages, and positive selection studies showed that macrophages represent the myeloid subpopulation with an increase in Odc expression. These data indicate that macrophages are the main source of ODC during colitis, although basal ODC in other immune cells could have a role. ODC enzymatic activity has been shown to be increased in colon tissues from humans with familial or somatic APC mutations (27) as well as in the Apcmin/+ mouse model of genetically-driven intestinal tumorigenesis (28). However, to our knowledge, ODC has not been immunolocalized in either human colitis or colon cancer. Using a human TMA, we found that ODC protein levels are increased in colonic macrophages in patients with active UC, and with colitis-associated dysplasia or colon cancer, validating the relevance of our animal studies and suggesting the involvement of macrophage ODC in the pathophysiology of colitis and CAC.

In this study, we show that Odc loss in myeloid cells is associated with increased levels of innate immune mediators, including multiple chemokines and cytokines that are hallmarks of M1 macrophages, as well as the Th1 cytokine IFNγ. These results are consistent with our findings that ODC-derived polyamines inhibit macrophage activation in both H. pylori-induced gastritis and in Citrobacter rodentium-induced colitis (10), such that macrophage ODC suppresses innate immune response to infection, through effects on gene transcription, limiting gastritis and colitis in these models. The data reported herein reveal a more complex scenario, where despite increases in proinflammatory chemokines and cytokines that are manifest with myeloid deletion of Odc, the OdcΔmye animals have improved recovery, better preservation of colon structure, and reduced histologic inflammation and epithelial injury. As epithelial deletion of Odc does not provide benefit, these findings support the concept that activation of innate immune responses in M1 macrophages confers protection from colitis. Consistent with this model, depletion of macrophages in mice worsens experimental colitis (29).

Quite strikingly, the increased mucosal innate immune response in OdcΔmye mice impaired tumorigenesis in the colon. Despite this, there was no reduction in histologic colitis and no difference in M1 gene expression in the nontumor areas of the OdcΔmye versus Odcfl/fl mice. This could be due to the long recovery period after the final dose of DSS (28 days), but could also reflect the specificity of the increased M1 response against tumors. The reduction of tumorigenesis in OdcΔmye mice was somewhat unexpected, as the inflammatory response is thought to drive the development of tumors in the AOM-DSS model (20, 30), and disease duration of UC in humans is associated with increased risk for colonic dysplasia and carcinoma (31). Further, macrophage depletion in mice using clodronate-filled liposomes reduces tumor number and burden in the AOM-DSS model (18), whereas macrophage depletion in the DSS model augments colitis (29). Also confounding are findings that Tlr4−/− mice exhibit less mucosal immune response and less tumor development with AOM-DSS treatment (32), yet signaling through MyD88, the main downstream effector of TLR4, reduces DNA damage and tumors in the same CAC model (33). Similarly, although TNFα has been shown to exhibit anticancer properties (34), mice lacking TNF-Rp55 develop less tumors than wild-type animals with AOM-DSS treatment (35).

Given the apparently paradoxical effects of myeloid Odc loss on CAC, we assessed the phenotype of tumor-associated macrophages. Notably, Odc-deficient macrophages display a robust M1 mRNA and protein response, and the M1 phenotype is known to have antitumor activities (36, 37), principally through the production of NOS2-derived nitric oxide (36, 38). Further, M1 cells support the development and the proliferation of Th1 cells that exhibit antitumor activities through the production of IFNγ (39, 40) and inhibit Th2-mediated expression of the genotoxic factor activation-induced cytidine deaminase (41). In accordance with these features, IFNγ expression is increased in the tumors of AOM-DSS–treated OdcΔmye versus Odcfl/fl mice. In contrast, the myeloid Odc deficiency does not increase the development of intratumoral M2 macrophages that promote tumorigenesis (42) by suppressing immunity via production of TGFβ (43, 44) and IL10 (45), and by supporting angiogenesis (44, 46). Although reduced clinical colitis with DSS may be relevant, our findings are consistent with a model where M1 polarization of macrophages is critical for the host to combat against carcinogenesis, and where macrophage ODC restricts M1 polarization. Further, increased M1 polarization by targeting myeloid ODC is mechanistically linked in vivo to changes in histone marks in tumors, and is consistent with the concept that the polyamine putrescine regulates chromatin structure and M1 target gene expression as found in infectious colitis (10). The lack of differences in tumor infiltrating T cells further points to a key role for the effect of ODC on macrophages.

In conclusion, the findings that macrophage ODC dampens the innate immune response, and M1-driven antitumoral immunity suggests that targeting macrophage ODC is an attractive approach to prevent development of CAC. Thus, agents that can specifically suppress ODC and/or polyamine availability in macrophages may show benefit in high-risk patients with UC-associated dysplasia or other forms of precancerous lesions, such as intestinal metaplasia associated with Barrett's esophagus and H. pylori infection, which also can lead to inflammatory carcinogenesis. In summary, ODC not only enhances tumor cell proliferation, but also prevents M1 macrophage-mediated antitumor immunity.

No potential conflicts of interest were disclosed.

Conception and design: K. Singh, A.P. Gobert, K.T. Wilson

Development of methodology: K. Singh, L.A. Coburn

Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): K. Singh, D.P. Barry, C. Shi, M.K. Washington, C. Schneider, A.G. Delgado, M.B. Piazuelo, J.L. Cleveland

Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): K. Singh, J.L. Cleveland, A.P. Gobert, K.T. Wilson

Writing, review, and/or revision of the manuscript: K. Singh, L.A. Coburn, C. Shi, J.L. Cleveland, A.P. Gobert, K.T. Wilson

Administrative, technical, or material support (i.e., reporting or organizing data, constructing databases): M. Asim, M.M. Allaman, P.B. Luis

Study supervision: K.T. Wilson

We thank Joseph T. Roland for technical assistance with the digital image analysis of the TMA. This work was supported by the NIH grants R01AT004821, R01DK053620, P01CA028842, P01CA116087 (subproject 7515), and R01CA190612 to K.T. Wilson; grant I01BX001453 from the Department of Veterans Affairs to K.T. Wilson; the Vanderbilt Ingram Cancer Center Support Grant (P30CA068485) and the Vanderbilt Digestive Research Center (P30DK058404) for mass spectrometry, grant IK2BX002126 from the Department of Veterans Affairs to L.A. Coburn; the Vanderbilt Center for Mucosal Inflammation and Cancer; and the Thomas F. Frist Sr. Endowed Chair to K.T. Wilson. This work was also supported by NIH grant R01AT006896 to C. Schneider, a fellowship award from the American Heart Association (16POST27250138) to P.B. Luis, and the Moffitt Cancer Center Support Grant P30CA076292, and by the Cortner-Couch Endowed Chair for Cancer Research to J.L. Cleveland. The generation of the TMA was supported by the NIH grants P30DK058404 and P50CA095103, and the generation of Odc floxed mice was supported by the NIH grant R01CA100603.

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.
Hardbower
DM
,
Singh
K
,
Asim
M
,
Verriere
TG
,
Olivares-Villagomez
D
,
Barry
DP
, et al
EGFR regulates macrophage activation and function in bacterial infection
.
J Clin Invest
2016
;
126
:
3296
312
.
2.
Isidro
RA
,
Appleyard
CB
. 
Colonic macrophage polarization in homeostasis, inflammation, and cancer
.
Am J Physiol Gastrointest Liver Physiol
2016
;
311
:
G59
73
.
3.
Galvan-Pena
S
,
O'Neill
LA
. 
Metabolic reprograming in macrophage polarization
.
Front Immunol
2014
;
5
:
420
.
doi: 10.3389/fimmu.2014.00420.
4.
Monteleone
G
,
Pallone
F
,
Stolfi
C
. 
The dual role of inflammation in colon carcinogenesis
.
Int J Mol Sci
2012
;
13
:
11071
84
.
5.
Ekbom
A
,
Helmick
C
,
Zack
M
,
Adami
HO
. 
Ulcerative colitis and colorectal cancer. A population-based study
.
N Engl J Med
1990
;
323
:
1228
33
.
6.
Eaden
JA
,
Abrams
KR
,
Mayberry
JF
. 
The risk of colorectal cancer in ulcerative colitis: a meta-analysis
.
Gut
2001
;
48
:
526
35
.
7.
Feagins
LA
,
Souza
RF
,
Spechler
SJ
. 
Carcinogenesis in BD: Potential targets for the prevention of colorectal cancer
.
Nat Rev Gastroenterol Hepatol
2009
;
6
:
297
305
.
8.
Terzic
J
,
Grivennikov
S
,
Karin
E
,
Karin
M
. 
Inflammation and colon cancer
.
Gastroenterology
2010
;
138
:
2101
14
e5
.
9.
Gobert
AP
,
Verriere
T
,
Asim
M
,
Barry
DP
,
Piazuelo
MB
,
de Sablet
T
, et al
Heme oxygenase-1 dysregulates macrophage polarization and the immune response to Helicobacter pylori
.
J Immunol
2014
;
193
:
3013
22
.
10.
Hardbower
DM
,
Asim
M
,
Luis
PB
,
Singh
K
,
Barry
DP
,
Yang
C
, et al
Ornithine decarboxylase regulates M1 macrophage activation and mucosal inflammation via histone modifications
.
Proc Natl Acad Sci U S A
2017
;
114
:
E751
E60
.
11.
Gobert
AP
,
Cheng
Y
,
Wang
JY
,
Boucher
JL
,
Iyer
RK
,
Cederbaum
SD
, et al
Helicobacter pylori induces macrophage apoptosis by activation of arginase II
.
J Immunol
2002
;
168
:
4692
700
.
12.
Pegg
AE
. 
Regulation of ornithine decarboxylase
.
J Biol Chem
2006
;
281
:
14529
32
.
13.
Pegg
AE
. 
Functions of polyamines in mammals
.
J Biol Chem
2016
;
291
:
14904
12
.
14.
Venkatesh
S
,
Workman
JL
. 
Histone exchange, chromatin structure and the regulation of transcription
.
Nat Rev Mol Cell Biol
2015
;
16
:
178
89
.
15.
Baumgart
DC
,
Carding
SR
. 
Inflammatory bowel disease: cause and immunobiology
.
Lancet
2007
;
369
:
1627
40
.
16.
Schenk
M
,
Bouchon
A
,
Seibold
F
,
Mueller
C
. 
TREM-1-expressing intestinal macrophages crucially amplify chronic inflammation in experimental colitis and inflammatory bowel diseases
.
J Clin Invest
2007
;
117
:
3097
106
.
17.
Rahman
FZ
,
Smith
AM
,
Hayee
B
,
Marks
DJ
,
Bloom
SL
,
Segal
AW
. 
Delayed resolution of acute inflammation in ulcerative colitis is associated with elevated cytokine release downstream of TLR4
.
PLoS One
2010
;
5
:
e9891
.
18.
Bader
JE
,
Enos
RT
,
Velazquez
KT
,
Carson
MS
,
Nagarkatti
M
,
Nagarkatti
PS
, et al
Macrophage depletion using clodronate liposomes decreases tumorigenesis and alters gut microbiota in the AOM/DSS mouse model of colon cancer
.
Am J Physiol Gastrointest Liver Physiol
2018
;
314
:
G22
G31
.
19.
Sierra
JC
,
Asim
M
,
Verriere
TG
,
Piazuelo
MB
,
Suarez
G
,
Romero-Gallo
J
, et al
Epidermal growth factor receptor inhibition downregulates Helicobacter pylori-induced epithelial inflammatory responses, DNA damage and gastric carcinogenesis
.
Gut
2017
;
67
:
1247
60
.
20.
Hardbower
DM
,
Coburn
LA
,
Asim
M
,
Singh
K
,
Sierra
JC
,
Barry
DP
, et al
EGFR-mediated macrophage activation promotes colitis-associated tumorigenesis
.
Oncogene
2017
;
36
:
3807
19
.
21.
Singh
K
,
Coburn
LA
,
Barry
DP
,
Asim
M
,
Scull
BP
,
Allaman
MM
, et al
Deletion of cationic amino acid transporter 2 exacerbates dextran sulfate sodium colitis and leads to an IL-17-predominant t cell response
.
Am J Physiol Gastrointest Liver Physiol
2013
;
305
:
G225
40
.
22.
Barry
DP
,
Asim
M
,
Scull
BP
,
Piazuelo
MB
,
de Sablet
T
,
Lewis
ND
, et al
Cationic amino acid transporter 2 enhances innate immunity during Helicobacter pylori infection
.
PLoS One
2011
;
6
:
e29046
.
23.
Cassetta
L
,
Noy
R
,
Swierczak
A
,
Sugano
G
,
Smith
H
,
Wiechmann
L
, et al
Isolation of mouse and human tumor-associated macrophages
.
Adv Exp Med Biol
2016
;
899
:
211
29
.
24.
Alexiou
GA
,
Lianos
GD
,
Ragos
V
,
Galani
V
,
Kyritsis
AP
. 
Difluoromethylornithine in cancer: new advances
.
Future Oncol
2017
;
13
:
809
19
.
25.
Ye
C
,
Geng
Z
,
Dominguez
D
,
Chen
S
,
Fan
J
,
Qin
L
, et al
Targeting ornithine decarboxylase by alpha-difluoromethylornithine inhibits tumor growth by impairing myeloid-derived suppressor cells
.
J Immunol
2016
;
196
:
915
23
.
26.
Meyskens
FL
 Jr
,
McLaren
CE
,
Pelot
D
,
Fujikawa-Brooks
S
,
Carpenter
PM
,
Hawk
E
, et al
Difluoromethylornithine plus sulindac for the prevention of sporadic colorectal adenomas: a randomized placebo-controlled, double-blind trial
.
Cancer Prev Res (Phila)
2008
;
1
:
32
8
.
27.
Giardiello
FM
,
Hamilton
SR
,
Hylind
LM
,
Yang
VW
,
Tamez
P
,
Casero
RA
 Jr
. 
Ornithine decarboxylase and polyamines in familial adenomatous polyposis
.
Cancer Res
1997
;
57
:
199
201
.
28.
Erdman
SH
,
Ignatenko
NA
,
Powell
MB
,
Blohm-Mangone
KA
,
Holubec
H
,
Guillen-Rodriguez
JM
, et al
Apc-dependent changes in expression of genes influencing polyamine metabolism, and consequences for gastrointestinal carcinogenesis, in the min mouse
.
Carcinogenesis
1999
;
20
:
1709
13
.
29.
Qualls
JE
,
Kaplan
AM
,
van Rooijen
N
,
Cohen
DA
. 
Suppression of experimental colitis by intestinal mononuclear phagocytes
.
J Leukoc Biol
2006
;
80
:
802
15
.
30.
Barrett
CW
,
Fingleton
B
,
Williams
A
,
Ning
W
,
Fischer
MA
,
Washington
MK
, et al
MTGR1 is required for tumorigenesis in the murine AOM/DSS colitis-associated carcinoma model
.
Cancer Res
2011
;
71
:
1302
12
.
31.
Beaugerie
L
,
Itzkowitz
SH
. 
Cancers complicating inflammatory bowel disease
.
N Engl J Med
2015
;
372
:
1441
52
.
32.
Fukata
M
,
Shang
L
,
Santaolalla
R
,
Sotolongo
J
,
Pastorini
C
,
Espana
C
, et al
Constitutive activation of epithelial TLR4 augments inflammatory responses to mucosal injury and drives colitis-associated tumorigenesis
.
Inflamm Bowel Dis
2011
;
17
:
1464
73
.
33.
Salcedo
R
,
Worschech
A
,
Cardone
M
,
Jones
Y
,
Gyulai
Z
,
Dai
RM
, et al
MyD88-mediated signaling prevents development of adenocarcinomas of the colon: Role of interleukin 18
.
J Exp Med
2010
;
207
:
1625
36
.
34.
Drutskaya
MS
,
Efimov
GA
,
Kruglov
AA
,
Kuprash
DV
,
Nedospasov
SA
. 
Tumor necrosis factor, lymphotoxin and cancer
.
IUBMB Life
2010
;
62
:
283
9
.
35.
Popivanova
BK
,
Kitamura
K
,
Wu
Y
,
Kondo
T
,
Kagaya
T
,
Kaneko
S
, et al
Blocking TNF-alpha in mice reduces colorectal carcinogenesis associated with chronic colitis
.
J Clin Invest
2008
;
118
:
560
70
.
36.
Edin
S
,
Wikberg
ML
,
Dahlin
AM
,
Rutegard
J
,
Oberg
A
,
Oldenborg
PA
, et al
The distribution of macrophages with a M1 or M2 phenotype in relation to prognosis and the molecular characteristics of colorectal cancer
.
PLoS One
2012
;
7
:
e47045
.
37.
Ryan
AE
,
Colleran
A
,
O'Gorman
A
,
O'Flynn
L
,
Pindjacova
J
,
Lohan
P
, et al
Targeting colon cancer cell NF-kappaB promotes an anti-tumour M1-like macrophage phenotype and inhibits peritoneal metastasis
.
Oncogene
2015
;
34
:
1563
74
.
38.
Scott
DJ
,
Hull
MA
,
Cartwright
EJ
,
Lam
WK
,
Tisbury
A
,
Poulsom
R
, et al
Lack of inducible nitric oxide synthase promotes intestinal tumorigenesis in the Apc(Min/+) mouse
.
Gastroenterology
2001
;
121
:
889
99
.
39.
Dunn
GP
,
Koebel
CM
,
Schreiber
RD
. 
Interferons, immunity and cancer immunoediting
.
Nat Rev Immunol
2006
;
6
:
836
48
.
40.
Osawa
E
,
Nakajima
A
,
Fujisawa
T
,
Kawamura
YI
,
Toyama-Sorimachi
N
,
Nakagama
H
, et al
Predominant T helper type 2-inflammatory responses promote murine colon cancers
.
Int J Cancer
2006
;
118
:
2232
6
.
41.
Endo
Y
,
Marusawa
H
,
Kou
T
,
Nakase
H
,
Fujii
S
,
Fujimori
T
, et al
Activation-induced cytidine deaminase links between inflammation and the development of colitis-associated colorectal cancers
.
Gastroenterology
2008
;
135
:
889
98
,
98 e1–3
.
42.
Wang
W
,
Li
X
,
Zheng
D
,
Zhang
D
,
Peng
X
,
Zhang
X
, et al
Dynamic changes and functions of macrophages and M1/M2 subpopulations during ulcerative colitis-associated carcinogenesis in an AOM/DSS mouse model
.
Mol Med Rep
2015
;
11
:
2397
406
.
43.
Hahm
KB
,
Lee
KM
,
Kim
YB
,
Hong
WS
,
Lee
WH
,
Han
SU
, et al
Conditional loss of TGF-beta signalling leads to increased susceptibility to gastrointestinal carcinogenesis in mice
.
Aliment Pharmacol Ther
2002
;
16
:
115
27
.
44.
Novitskiy
SV
,
Pickup
MW
,
Chytil
A
,
Polosukhina
D
,
Owens
P
,
Moses
HL
. 
Deletion of TGF-beta signaling in myeloid cells enhances their anti-tumorigenic properties
.
J Leukoc Biol
2012
;
92
:
641
51
.
45.
Berg
DJ
,
Davidson
N
,
Kuhn
R
,
Muller
W
,
Menon
S
,
Holland
G
, et al
Enterocolitis and colon cancer in interleukin-10-deficient mice are associated with aberrant cytokine production and CD4(+) Th1-like responses
.
J Clin Invest
1996
;
98
:
1010
20
.
46.
Pickert
G
,
Lim
HY
,
Weigert
A
,
Haussler
A
,
Myrczek
T
,
Waldner
M
, et al
Inhibition of GTP cyclohydrolase attenuates tumor growth by reducing angiogenesis and M2-like polarization of tumor associated macrophages
.
Int J Cancer
2013
;
132
:
591
604
.