Distinct roles of histamine H1- and H2-receptor signaling pathways in inflammation-associated colonic tumorigenesis

Abstract

Inflammatory bowel disease (IBD) is a well-known risk factor for the development of colorectal cancer. Prior studies have demonstrated that microbial histamine can ameliorate intestinal inflammation in mice. We tested the hypothesis whether microbe-derived luminal histamine suppresses inflammation-associated colon cancer in Apc^min/+ mice. Mice were colonized with the human-derived Lactobacillus reuteri. Chronic inflammation was induced by repeated cycles of low-dose dextran sulfate sodium (DSS). Mice that were given histamine-producing L. reuteri via oral gavage developed fewer colonic tumors, despite the presence of a complex mouse gut microbiome. We further demonstrated that administration of a histamine H1-receptor (H1R) antagonist suppressed tumorigenesis, while administration of histamine H2-receptor (H2R) antagonist significantly increased both tumor number and size. The bimodal functions of histamine include protumorigenic effects through H1R and antitumorigenic effects via H2R, and these results were supported by gene expression profiling studies on tumor specimens of patients with colorectal cancer. Greater ratios of gene expression of H2R (HRH2) vs. H1R (HRH1) were correlated with improved overall survival outcomes in patients with colorectal cancer. Additionally, activation of H2R suppressed phosphorylation of mitogen-activated protein kinases (MAPKs) and inhibited chemokine gene expression induced by H1R activation in colorectal cancer cells. Moreover, the combination of a H1R antagonist and a H2R agonist yielded potent suppression of lipopolysaccharide-induced MAPK signaling in macrophages. Given the impact on intestinal epithelial and immune cells, simultaneous modulation of H1R and H2R signaling pathways may be a promising therapeutic target for the prevention and treatment of inflammation-associated colorectal cancer.

NEW & NOTEWORTHY Histamine-producing Lactobacillus reuteri can suppress development of inflammation-associated colon cancer in an established mouse model. The net effects of histamine may depend on the relative activity of H1R and H2R signaling pathways in the intestinal mucosa. Our findings suggest that treatment with H1R or H2R antagonists could yield opposite effects. However, by harnessing the ability to block H1R signaling while stimulating H2R signaling, novel strategies for suppression of intestinal inflammation and colorectal neoplasia could be developed.

INTRODUCTION

Discovered more than 100 years ago, histamine is an important signaling compound affecting a variety of biological processes, including neuronal activity, endothelial permeability, vascular tone and gastric acid secretion, inflammation, allergy, and development of cancer (14, 37). The pleiotropic effects of histamine are mediated by the potential activation of each of four histamine receptors (H1R, H2R, H3R, and H4R) present on mammalian cells. H1R, H2R, and H4R have been well described in human and mouse immune cells, cholangiocytes, hepatocytes, and endothelial cells (17, 71), whereas H3R is primarily expressed in neurons of the central nervous system (CNS) (21). Histamine influences myeloid cell differentiation (72), and H1R and H4R are considered to be the two histamine receptors involved in allergic inflammation (64). H2R is central in gastric acid production (8). Clearly, histamine is a biogenic amine synthesized endogenously by mammalian cells with important roles in basic physiologic processes.

Previous studies have focused on the contrasting effects of histamine receptors with respect to inflammation and immune cell signaling (35). The effects of H1R and H2R in chronic inflammation and in inflammation-associated colon cancer are not fully understood. H1R activation regulates downstream pathways through intracellular calcium concentration, while H2R signals through cyclic AMP (cAMP) (15). Activation of H1R and H2R has been widely shown to yield opposing effects in multiple biological processes. For example, in human T-cell-mediated immune responses, H1R activation promotes Th1 polarization, while H2R activation suppresses Th1 polarization (46, 61). Distinct effects of H1R and H2R activation were also evident in terms of smooth muscle contraction. H1R and H2R antagonists, respectively, inhibit and exacerbate histamine-induced bronchoconstriction in patients with mild asthma (53). These findings strongly suggest that histamine may have opposing effects depending on the specific histamine receptor that is activated.

In the pathogenesis of inflammatory bowel disease (IBD), activation of MAP kinase (MAPK) and nuclear factor-κB (NF-κB) pathways are key events that enhance the production of proinflammatory cytokines, such as TNF and IL-6 (5, 12). Therefore, IBD medications have been developed that target specific cytokine signaling pathways (26). Histamine was detected in increased concentrations in the intestinal mucosa of IBD (55) and promoted ERK phosphorylation through H1R activation in human epidermal keratinocytes (45) and aortic endothelial cells (25). Although it is not clear whether H1R is involved in MAPK activation in intestinal epithelial cells, H1R antagonists, such as loratadine (56) and ketotifen (32), have been used to alleviate clinical symptoms in IBD. In contrast, studies have shown that administration of the H2R antagonist cimetidine resulted in more rapid death of mice in an intestinal infection model (24) and our previous study showed that H2R activation suppressed trinitrobenzene sulfonate-induced colonic inflammation in mice (20). Interestingly, H2R blockers also increased the frequency of hospitalization or surgery in patients with ulcerative colitis (34) as well as cancer risk, including intestinal cancers (50). Furthermore, the H2R agonist dimaprit was reported to reduce TNF production induced by LPS in mouse models of endotoxin shock and hepatitis (52).

A link between chronic inflammation and colorectal cancer (CRC) is well recognized (4, 51, 66). Yang et al. (19) demonstrated antitumorigenic effects of histamine in histidine decarboxylase gene (Hdc) knockout mice that lacked the ability to synthesize endogenous histamine and yielded increased susceptibility to chemically induced colon cancer (72). Our group demonstrated that Lactobacillus reuteri-derived histamine suppressed tumor formation in the colons of Hdc knockout mice. Activation of H1R signaling promoted proliferation of cholangiocarcinoma-derived cells (16). Abrogation of H1R signaling with antagonists enhanced radiosensitivity, resulting in reduced viability of colon cancer cells. Furthermore, elevated expression of the HRH1 gene has been shown to be associated with poor survival for both lung cancer and B-cell lymphoma patients (70). Several other studies demonstrated that H1R activation suppressed cell proliferation in prostate cancer (67), melanoma (40), and leukemia (31) cells and promoted cell migration of cervical carcinoma cells (57). Altogether, the effects of histamine-receptor activation seem to be tissue specific, depending on relative distributions of histamine receptors in different tissue types (37). In addition to the crucial role in IBD, MAPKs are also critical mediators of signal transduction in cancer development. For example, ERK activation promotes intestinal tumorigenesis in Apc^Min/+ mice (41) and results in increased proliferation of colon cancer cells in vitro (33). By contrast, H2R activation suppressed ERK phosphorylation in human monocytes (63) and promoted ERK phosphorylation in HEK293T cells (10). The precise mechanisms, however, by which histamine affects inflammation associated-intestinal tumorigenesis through H1R and H2R signaling pathways remain to be elucidated.

In this study, we demonstrated that H1R signaling promoted intestinal tumor formation in vivo and proliferation of colorectal cancer-derived intestinal epithelial cells. By contrast, H2R signaling suppressed tumor growth in inflammation-associated colon cancer in mouse models. To explore the implications in human cancer, we found that an increased ratio of HRH2 vs. HRH1 gene expression was associated with improved survival rates of CRC patients. We attempt to delineate molecular mechanisms explaining how histamine regulates chronic intestinal inflammation via different receptors, H1R and H2R, and how this deeper understanding of intestinal histamine signaling may facilitate the development of new preventive strategies and therapeutics in the future. New therapies aimed at blocking H1R signaling while promoting H2R signaling may offer real promise as cancer therapeutic strategies.

MATERIALS AND METHODS

Mice.

Apc^min/+ mice were maintained under specific pathogen-free conditions with a 12-h:12-h light-dark cycle in animal facilities at Baylor College of Medicine. Two-month-old male mice were randomly divided into three groups (n = 8–11 mice per group, 3–4 mice per cage) and were gavaged with 5 × 10⁹ colony forming units of histamine-producing L. reuteri 6475, histamine negative L. reuteri 6475 (inactivated hdcA and unable to convert l-histidine into histamine), or control media (MRS without bacteria), daily for 7 days. Thereafter, mice were given two cycles of 1.5% DSS (36,000 to 50,000 molecular weight; MP Biomedicals, Solon, OH) in drinking water as follows: DSS for 5 consecutive days followed by 17 days of recovery, and a second cycle of DSS for 4 days followed by 18 days of recovery. During DSS treatment and recovery periods, the mice received either histamine-generating L. reuteri 6475, histidine decarboxylase (HdcA)-deficient L. reuteri 6475, or control media once every 3 days. At the end of the second recovery period, mice were euthanized, and tumors were counted and measured under a dissecting microscope by two individuals blinded to the mouse treatment groups. The intestinal tissues were fixed in formalin and embedded in paraffin (FFPE) for histologic evaluation. Dysplasia was counted throughout the colon on hematoxylin-eosin-stained sections.

To study receptor-specific effects of histamine, Apc^min/+ mice were given pyrilamine (H1R antagonist, 50 mg/l) (65), cimetidine (H2R antagonist, 100 mg/l) (1), or omeprazole (proton pump inhibitor, 10 mg/l) (30) in drinking water (all drugs were fully dissolved in water at the designated concentrations) starting 7 days before DSS treatment 1 until completion of the second recovery period post-DSS treatment 2. Mice were then euthanized and processed as mentioned above. All procedures were approved by the Institutional Review Board at Baylor College of Medicine.

Human cell lines.

Human colorectal cancer-derived cell lines, including HCT116, Caco2, DLD1, LS174T and HT29 (all from ATCC, Manassas VA), were grown in DMEM containing 10% fetal bovine serum (GIBCO, Life Technologies, Carlsbad, CA) in a humidified 5% CO₂ atmosphere at 37°C.

L. reuteri strains.

L. reuteri 6475 and the isogenic hdcA mutant strain were prepared as described previously (63). Bacterial cells were cultured in MRS medium at 37°C in anaerobic conditions (20).

Antibodies, plasmids, and chemicals.

Antibodies are listed in Table 1. Recombinant plasmids were generated for ectopic expression of histamine-receptor genes. Full-length open reading frames of the wild-type mouse Hrh1 and Hrh2 genes were subcloned into the mammalian expression vector pcDNA3.1 (Invitrogen, Carlsbad, CA). The constructs were confirmed by DNA sequencing. Transfections were performed with 1 μg of plasmid DNA per well (24-well plate) and 2 μl of FuGENE HD transfection reagent (Roche, Indianapolis, IN). Chemicals including pyrilamine maleate salt and cimetidine were purchased from Sigma-Aldrich (St. Louis, MO). Omeprazole was obtained from TCI America (Portland, OR). Dimaprit dihydrochloride was obtained from Biotrend (Destin, FL), and 2-pyridylethylamine dihydrochloride was purchased from Tocris (Minneapolis, MN).

Table 1.

Antibodies and their applications used in this study

Antibody Name	Company	Clone Number or Catalog Number	Applications and Dilutions
H1R	Santa Cruz Biotechnology	P-20	Fluorescent staining: 1:200; immunoblot: 1:500
H2R	Santa Cruz Biotechnology	M-19	Fluorescent staining: 1:200; immunoblot: 1:500
Human H1R	Elabscience	ESAP13556	Fluorescent staining: 1:400; immunoblot: 1:1,000
PCNA	Santa Cruz Biotechnology	PC10	Fluorescent staining: 1:400
β-Actin	Santa Cruz Biotechnology	11B7	Immunoblot: 1:500
Phospho-ERK	Cell Signaling	D13.14.4E	Fluorescent staining: 1:300; immunoblot: 1:1,000
Phospho-p38	Cell Signaling	D3F9	Immunoblot: 1:1,000
Phospho-JNK	Cell Signaling	81E11	Immunoblot: 1:1,000
Phospho-p65	Cell Signaling	93H1	Immunoblot: 1:1,000
Total ERK	Cell Signaling	137F5	Immunoblot: 1:1,000
BrdU	Accurate	OBT0030	Histochemical staining: 1:200

H1R and H2R, H1 and H2 receptor, respectively; PCNA, proliferating cell nuclear antigen; BrdU, bromodeoxyuridine.

Immunofluorescence and bromodeoxyuridine staining.

Immunofluorescence was performed on 5-µm-thick FFPE intestinal tissue sections using a goat polyclonal anti-H1R antibody (Santa Cruz Biotechnology, Santa Cruz, Dallas, TX). Slides were deparaffinized in xylene and hydrated with a series of washes using graded alcohols, followed by antigen retrieval with Tris-EDTA buffer (pH 8.0, 1 mM EDTA, 10 mM Tris) in a steamer for 30 min. After a wash with TBS-T (50 mM Tris·Cl), slides were blocked with 10% goat serum for 30 min and then incubated with primary antibodies in TBS-T buffer at 4°C overnight. Dilutions for antibodies were listed in Table 1. After being washed with TBS-T, slides were incubated with Alexa Fluor-conjugated secondary antibodies (Invitrogen) for 30 min, followed by DAPI counterstaining to visualize nuclei. For immunofluorescence staining of cultured cells, HCT116 cells were fixed with 4% paraformaldehyde for 10 min followed by 0.5% Triton X-100 incubation for 5 min and then blocked as described above. Immunofluorescence images were generated by fluorescence microscopy (Nikon Eclipse 90 Ni-E). For bromodeoxyuridine (BrdU) labeling, mice were injected with 200 µl of BrdU 2 h before death (Invitrogen) and staining was performed on FFPE sections of colon tumors using a rat anti-BrdU antibody (Accurate, Westbury, NY) and then visualized with the ABC-Elite Kit using DAB (Vector Laboratories, Burlingame, CA).

Immunoblots.

Lysates from HCT116 cells or tissues were obtained with RIPA lysis buffer (Thermo Fisher Scientific, Waltham, MA) containing a protease inhibitor cocktail (Roche, Indianapolis, IN). Supernatants containing proteins were resolved by SDS-PAGE and transferred to PVDF membranes (Millipore, Billerica, MA). Membranes were blocked with 5% dry milk and incubated with primary antibodies overnight at 4°C. After washes with TBST, membranes were incubated with horseradish peroxidase-conjugated secondary antibodies for 1 h at room temperature and developed with ECL substrate (GE Health Care, Buckinghamshire, UK). Densitometric analyses of immunoblots were performed using Band Leader software.

RNA extraction and quantitative RT-PCR analysis.

Total RNA was extracted using TRIzol reagent (Life Technologies, Gaithersburg, MD) according to the manufacturer’s instructions. Reverse transcription was prepared from 1 μg of total RNA using SuperScript II reverse transcription kit (Roche). Quantitative PCR analyses were then performed using Fast SYBR Green (Life Technologies) and amplified on the Applied Biosystems 7900HT instrument. Ribosomal RNA (18S) was used as a control. Primers were listed in Table 2. Relative levels of gene expression were analyzed using the comparative C_T method (^2−∆∆C_T method).

Table 2.

List of primer sequences and their corresponding applications

Gene Name	Forward (5′-3′)	Reverse (5′-3′)	Applications
HRH1	CACACTGAACCCCCTCATCT	ATTTTGTTGCATCCCCTCAG	qPCR
HRH2	ACCAGCAAGGGCAATCATAC	CATGATCAGTAGCGGGAGGT	qPCR
Egr1	GAGCGAACAACCCTATGAGC	TGGGATAACTCGTCTCCACC	qPCR
CCL20	CCAAGAGTTTGCTCCTGGCT	TGCTTGCTGCTTCTGATTCG	qPCR
IL8	CACCGGAAGGAACCATCTCA	GGAAGGCTGCCAAGAGAGC	qPCR
IL6	GTATGAACAACGATGATGCACTTG	ATGGTACTCCAGAAGACCAGAGGA	qPCR
18S	GGGTCGCGTAACTAGTTAGCATG	CTTAGTTGGTGGAGCGATTTGTC	qPCR
TOPO-Hrh1	CTAGGATCCGCCACCATGAGCCTTCCCAACACCTC	CTAGAATTCTTAGGAACGAATGTGCAGAATTTTTTTG	Cloning
TOPO-Hrh2	CTAGGATCCGCCACCATGGAGCCCAATGGCACGGTTC	CTAGAATTCTTAAGCACTGATATGTAGTGATGG	Cloning

qPCR, quantitative PCR.

Cell viability assay.

HCT116 cells (1 × 10³/well) were plated and cultured in a 96-well plate overnight. Cells were treated with dimaprit dihydrochloride (25 μM), 2-pyridylethylamine dihydrochloride (25 μM), or both for 4 days. Cell viability was determined using a Cell Counting Kit-8 assay (Dojindo Molecular Technologies, Gaithersburg, MD), and the absorbance was read at 450 nm using a spectrophotometer. For transfected HCT116 cells, cell viability was determined after treatment with 100 nM histamine for 4 days.

Statistical analyses.

Statistical analyses were performed using GraphPad Prism version 5.04 (GraphPad Software, San Diego, CA). All data were normally distributed and examined using parametric tests. Data for tumor counts were analyzed by one-way ANOVA with a Bonferroni post hoc test. The band densities of immunoblots at different time points were analyzed by two-way repeated measures ANOVA with a Bonferroni post hoc test. The results of quantitative PCR (see Figs. 4 and 5) and cell viability assay were analyzed by one-way ANOVA with a Bonferroni post hoc test. The results of quantitative PCR studies (see Fig. 2) and BrdU-positive cell counts were analyzed using unpaired, two-tailed Student's t-test. Statistical test results are presented as means ± SD. Significance was set at P < 0.05, P < 0.01, and P < 0.001, as indicated in figures.

Fig. 4.

H2-receptor (H2R) signaling suppresses MAPK signaling pathways induced by H1-receptor (H1R) activation. A: HCT116 cells were transfected with empty Vector, TOPO-Hrh1, TOPO-Hrh2, or TOPO-Hrh1 together with TOPO-Hrh2, and treated with histamine (10 µM) for 0, 15, 30, 60 and 120 min. Immunoblotting was performed for phosphorylated JNK (p-JNK), phosphorylated ERK (p-ERK), phosphorylated p38 (p-p38) and β-actin. The results are representative of 3 independent experiments (biological triplicates). B: all bands of immunoblots were quantified using Band Leader software and the density of each band was normalized relative to that of β-actin. The relative density at 0 min in control cells was designated as 1 (y-axis). The relative density was calculated from biological triplicates. C: immunoblotting was performed to verify H1R and H2R overexpression in HCT116 cells. D: double immunostaining of H1R (green) and p-ERK was performed on Hrh1 overexpressing HCT116 cells treated with histamine (10 µM) for 1 h. DAPI was used to visualize nuclei (blue). E: phosphorylated ERK was not detected in Hrh2 overexpressing HCT116 cells treated with histamine (10 µM) for 1 h. F and G: HCT116 cells were transfected with empty vector, TOPO-Hrh1, TOPO-Hrh2, or TOPO-Hrh1 plus TOPO-Hrh2 and were then treated with 10 µM histamine for 6 h. F and G: the relative expression of CCL20 (F) and IL-8 (G) was determined by RT-qPCR. Data of B were analyzed by two-way repeated measures ANOVA with a Bonferroni post hoc test. The experiments in F and G: were performed 3 times, each with 3 technical replicates. Data are presented as means ± SD (n = 3) from 1 representative experiment and analyzed by one-way ANOVA with a Bonferroni post hoc test. *P < 0.05, **P < 0.01, and ***P < 0.001.

Fig. 5.

H1- and H2-receptor (H1R and H2R) signaling pathways differentially affect cytokine gene expression induced by LPS in macrophages. A and B: RAW 264.7 cells (A) or THP-1 cells (B) were treated with 100 ng/ml LPS, LPS together with 10 μM histamine (HIS), and LPS plus histamine together with 100 μM pyrilamine (PY) or 100 μM cimetidine (CIM) for 6 h. Then, IL-6 and TNF expression were determined by quantitative RT-qPCR. C: RAW 264.7 cells were treated with 100 ng/ml LPS, 10 μM histamine, 10 μM pyrilamine, or 10 μM dimaprit (DIM) as well as different combinations for 30 (light +) and 120 (dark +) minutes, and the expression of p-p38, p-ERK, p-p65, and β-actin was determined by immunoblot (top). The signals (immunoblot bands) generated following exposure to LPS, dimaprit and pyrilamine are highlighted in the rectangle with dashed line borders. The experiments in A and B were performed 3 times, each with 3 biological replicates. Data were analyzed by one-way ANOVA with a Bonferroni post hoc test. Data are presented as means ± SD from 1 representative experiment. The relative densities in C, bottom, were calculated from the quantified signals (immunoblot bands) of 3 independent experiments using Band Leader software. Data were analyzed using two-way repeated measures ANOVA with a Bonferroni post hoc test. Data are presented as means ± SD (n = 3). *P < 0.05, **P < 0.01, and ***P < 0.001.

Fig. 2.

H1- and H2-receptor (H1R and H2R) signaling pathways play opposing roles in tumorigenesis of colonic cancer in Apc^Min/+ mouse model. A: study design for all the experiments in B–E. B: Apc^min/+ mice were given 50 µM pyrilamine (n = 6), 100 µM cimetidine (n = 7), 10 µM omeprazole (n = 6) or water control (n = 9) during dextran sulfate sodium (DSS) treatment. No bacteria were given to those mice. C: tumors in cecum and in colon were counted and measured after the second post-DSS recovery period. In Apc^min/+ mice treated with cimetidine, tumors in cecum and in colon were measured under a dissecting microscope. D: bromodeoxyuridine (BrdU) staining was performed on sections of colon tumors from cimetidine-treated and control mice (left). The percentages of BrdU-positive cells were calculated from 6 fields per mouse (n = 3) (right). E, left: proteins were extracted from DSS-induced colon tumors of Apc^min/+ mice treated with or without cimetidine, and ERK phosphorylation was detected by immunoblot. E, right: quantitative RT-PCR was performed using RNA from DSS-induced tumors to determine the relative expression of Egr1. Statistical comparisons were performed using GraphPad Prism software. Data in B were analyzed using one-way ANOVA followed with a Bonferroni post hoc test. Data in C were analyzed by two-way repeated measures ANOVA with a Bonferroni post hoc test. Data in D and E were analyzed using unpaired, two-tailed Student's t-test. Data are presented as means ± SD. *P < 0.05, **P < 0.01.

RESULTS

Microbial histamine suppresses colonic tumorigenesis in Apc^min/+ mouse model.

To test the hypothesis that histamine generating bacteria may suppress inflammation-associated colon cancer, we used a well-accepted mouse model of colitis-associated colon cancer (11, 62). In this model, chronic inflammation is induced by multiple cycles of low-dose DSS treatment in drinking water with recovery phases in Apc^min/+ mice. DSS treatment induced dysplastic lesions and tumors in the mouse colon and cecum. Apc^min/+ mice received wild-type L. reuteri, hdcA mutant L. reuteri (19), or control media via orogastric gavage daily for 1 wk before the initial DSS treatment of 5 days (Fig. 1A). Thereafter, mice were gavaged once per 3 days throughout the study, based on prior results, demonstrating that orally administered L. reuteri remain in the mouse colon for at least 2 days, regardless of histamine status (19). Mice that received histamine-producing L. reuteri developed fewer colonic tumors (P < 0.05) and exhibited fewer dysplastic lesions (low grade) in comparison to those in the MRS media only group, while histamine-negative L. reuteri did not suppress colonic tumor formation (Fig. 1B) or intestinal dysplasia (Fig. 1, C and D). However, mean tumor size was not significantly different between treatment groups (data not shown). Across all groups, DSS-induced tumors were mostly confined to the midcolon rather than the distal colon. Although histologic differences were not observed in the adenomas among different groups, mice that received histamine-producing L. reuteri had fewer dysplastic lesions, indicating that bacterial histamine from the intestinal microbiome is capable of suppressing inflammation-associated tumorigenesis and dysplasia in Apc^Min/+ mice.

Fig. 1.

Lactobacillus reuteri-derived histamine suppresses colonic tumorigenesis in Apc^Min/+ mice. A: study design for the experiments shown in B–D. Blue arrows indicate oral gavage and red arrow indicates the time of euthanasia. B: Apc^min/+ mice were gavaged with histamine-producing L. reuteri 6475 (n = 11), L. reuteri 6475 hdcA (n = 8), or MRS media control (n = 10) during dextran sulfate sodium (DSS)-induced tumorigenesis, and total tumor numbers in colons were counted under a dissecting microscope. C: dysplastic lesions were counted on hematoxylin-eosin (H&E)-stained sections throughout the colons of mice treated with L. reuteri 6475 (n = 5), L. reuteri 6475 hdcA M (n = 6), or MRS media control (n = 6). D: representative images of H&E staining (top: ×40; bottom: ×200) of colon sections from Apc^min/+ mice gavaged with L. reuteri 6475, L. reuteri 6475 hdcA, and MRS medium control. Statistical comparisons were performed using GraphPad Prism software. Data were analyzed using one-way ANOVA followed by with a Bonferroni post hoc test. *P < 0.05.

H2R signaling blockade promotes colonic tumorigenesis in Apc^min/+ mice.

To test the effects of histamine via different histamine receptors, Apc^min/+ mice were treated with DSS and either a H1R or H2R antagonist. To address the potential confounding effects of histamine-receptor signaling on gastric acid production, one group of mice was treated with omeprazole (Fig. 2A). Omeprazole suppresses gastric acid secretion via direct inhibition of proton pumps in the gastric epithelium, leaving histamine-receptor signaling intact (38). Mice treated with pyrilamine, a H1R antagonist, developed fewer tumors both in the cecum and in the colon (Fig. 2B) compared with control mice, while mice treated with the H2R antagonist, cimetidine, yielded a significantly increased tumor burden (Fig. 2B). Treatment with omeprazole did not significantly affect tumor number compared with the untreated control group. The data indicated that H1R activation promotes tumor initiation and growth, while H2R activation suppresses inflammation-associated cancer initiation in this mouse model.

In addition to promoting cancer initiation, H2R signaling blockade with cimetidine also resulted in increased tumor sizes in the cecum and in the colon (Fig. 2C) in DSS treated Apc^Min/+ mice. This finding suggests that H2R activation suppresses tumor growth. More BrdU-positive cells were visible per microscopic high-power field in colonic tumors of cimetidine-treated mice (Fig. 2D), supporting the suppressive role of H2R signaling in colonic tumor cell proliferation. These results strongly suggest that H2R antagonists could potentially aggravate intestinal inflammation in IBD and may promote colonic tumorigenesis.

As mentioned above, ERK activation is required to promote intestinal tumorigenesis in Apc^Min/+ mice (41). To explore whether H2R activation inhibits ERK phosphorylation in vivo, we compared ERK phosphorylation in tumors of Apc^Min/+ mice treated with cimetidine by immunoblot. Cimetidine treatment dramatically increased ERK phosphorylation (Fig. 2E, left), demonstrating that H2R blockade results in activation of the ERK signaling pathway in this model of inflammation-associated colon cancer. Consequently, the increased ERK activity in tumors elevated the expression of ERK-regulated genes, such as Egr1, by quantitative PCR (Fig. 2E, right). Early growth response-1 (EGR1) is increased in human colorectal cancer cell lines as well as in colorectal cancer (29), and greater quantities of EGR1 promote CRC cell survival (43). Our study suggests that cimetidine administration could promote colorectal carcinogenesis via the ERK/EGR1 signaling pathway.

Human H1R signaling promotes viability of colorectal cancer-derived epithelial cells.

To evaluate the presence of H1R in human colorectal cancer cells, dual immunostaining was performed with antibodies against H1R and proliferating cell nuclear antigen, a marker of cell proliferation, in a specimen from a patient with advanced (stage IV) colorectal cancer. Robust H1R signals were observed in human colorectal cancer cells with prominent proliferating cell nuclear antigen staining (Fig. 3A). Transfection of additional copies of the mouse H1R gene in human colorectal cancer-derived epithelial cells (HCT116 cells) resulted in increased cellular viability, and transfection of mouse H2R gene on a plasmid-borne vector did not increase viability of HCT116 cells. Excess H2R signaling was dominant over H1R signaling (Hrh1 + Hrh2) (Fig. 3B). HCT116 cell viability was enhanced by 2-pyridylethylamine (H1RA), a H1R agonist, which can be suppressed by dimaprit (H2RA), a H2R agonist (Fig. 3C), further supporting contrasting roles of H1R and H2R signaling pathways in intestinal epithelial cell proliferation.

Fig. 3.

H1-receptor (H1R) signaling stimulates and H2-receptor (H2R) signaling suppresses cell proliferation. A: double immunostaining was performed on a formalin-fixed, paraffin-embedded section of a colorectal cancer specimen with H1R and proliferating cell nuclear antigen (PCNA) antibodies. CRC, colorectal cancer. B: HCT116 cells were transfected with empty plasmid vector, TOPO-Hrh1 only (Hrh1), TOPO-Hrh2 only (Hrh2), or TOPO-Hrh2 plus TOPO-Hrh1 (Hrh1 + Hrh2). Twenty-four hours post-transfection, cells were treated with 100 nM histamine for 4 days. Cell viability was determined by the Cell Counting Kit-8 (CCK-8) assay and is depicted in the bar graph. Relative abundances of transfected HCT116 cells were pictured after crystal violet staining (right). C: HCT116 cells were treated with 25 μM 2-pyridylethylamine (H1R agonist, designated as H1RA), 25 μM dimaprit (H2R agonist, designated as H2RA), or a combination of agonists. Cell viability was determined by CCK-8 assay after 4-day treatment. Data in B and C resulted from a single experiment in triplicate, and similar results were obtained in 2 independent experiments. Data were analyzed by one-way ANOVA with a Bonferroni post hoc test. Data are presented as means ± SD. *P < 0.05, **P < 0.01, and ***P < 0.001.

H2R signaling inhibits H1R-induced MAPK phosphorylation in human colorectal cancer-derived cells.

To test whether the activation of H1R and H2R may modify ERK phosphorylation in intestinal cancer cells, we again produced plasmid-borne H1R and H2R in HCT116 cells with either or both Hrh1 and Hrh2 plasmids. Upon histamine treatment, H1R-overproducing HCT116 cells showed enhanced phosphorylation of JNK and ERK MAPKs (Fig. 4A). Simultaneous overproduction of H2R suppressed H1R-mediated activation of ERK and JNK at the 30- and 60-min time points, as measured by immunoblot (Fig. 4 A and B). The activity of these constructs was confirmed by immunoblot using H1R and H2R antibodies, respectively (Fig. 4C). Furthermore, when treated with histamine, HCT116 cells showed phosphorylated ERK in nuclei exclusively within cells that produce H1R by immunofluorescence (Fig. 4D), while H2R overproducing cells show only sparse, if any, signals of phosphorylated ERK (Fig. 4E).

Moreover, in H1R-overproducing HCT116 cells, histamine significantly elevated quantities of chemokine mRNA, including CCL20 and IL8, which were suppressed by the presence of H2R (Fig. 4, F and G). These data demonstrated that H2R signaling is capable of suppressing ERK signaling and chemokine gene expression induced by activation of H1R signaling in colon cancer cells.

H2R signaling suppresses LPS-induced IL6 gene expression in monocytes/macrophages.

LPS/Toll-like receptor-4 is a key pathway in colitis-associated colon cancer and histamine signaling in macrophages contributes to cytokine-mediated inflammation in mouse models (42, 72). Histamine suppressed TNF production via H2R activation in both human PBMCs (60, 68) and human monocytoid cells (63). IL-6 plays a key role in the development of inflammation-associated colon cancer (23). Thus we sought to determine whether H1R and H2R signaling pathways affect LPS-induced proinflammatory cytokine gene expression in macrophages. Quantitative PCR experiments showed that histamine (10 µM) significantly suppressed both TNF and IL6 gene expression induced by LPS in both human THP-1 monocytoid cells and mouse RAW 264.7 macrophages (Fig. 5, A and B). Furthermore, cytokine suppression by histamine can be blocked by the H2R antagonist cimetidine (Fig. 5, A and B), demonstrating that histamine suppresses the transcription of these two genes via H2R signaling. Meanwhile, administration of the H1R antagonist pyrilamine reduced the expression of IL6, but not TNF (Fig. 5, A and B). These findings strongly indicate that histamine promotes IL-6 expression via H1R activation and suppresses IL-6 expression via H2R signaling in human and mouse myeloid cells.

We next examined the effects of a combination of H1R antagonist (pyrilamine) and H2R agonist (dimaprit) on LPS-induced MAPK and NF-κB pathways in RAW 264.7 cells to test if suppression of downstream pathways can be maximized by such a combination. LPS stimulation increased phosphorylation of p38, ERK, and p65 after 30 and 120 min (Fig. 5C). In contrast, administration of histamine, dimaprit, or pyrilamine reduced LPS-induced ERK phosphorylation but did not affect p65 phosphorylation. Notably, the combination of pyrilamine and dimaprit resulted in the most potent suppression of LPS induced p38 and ERK phosphorylation (Fig. 5C, top), which was quantified using Band Leader software (Fig. 5C, bottom). Together, we confirmed that simultaneous inhibition and stimulation of H1R and H2R signaling pathways, respectively, can effectively suppress proinflammatory signaling in macrophages.

Expression of HRH1 and HRH2 is correlated with patient outcomes in human colorectal cancer.

To explore the roles of H1R and H2R signaling pathways in human colorectal cancer, we examined if elevated expression of HRH1 or HRH2 in CRC tissue specimens were associated with CRC patient outcomes using the R2 Platform (human gene expression profiling database; https://hgserver1.amc.nl/cgi-bin/r2/main.cgi). Analysis of the microarray data set GSE17538 revealed that CRC patients with elevated HRH1 human gene expression had worse overall survival than those with lower HRH1 expression (Fig. 6A, left). However, elevated HRH2 expression was associated with an improved 10-year survival (Fig. 6A, right), suggesting that H1R and H2R signaling pathways yield contrasting effects on CRC and its progression. These results are consistent with results obtained with mouse models and human cells in this study.

Fig. 6.

Expression of HRH1 and HRH2 genes are inversely correlated with survival of colorectal cancer (CRC) patients. A: the correlations between gene expression patterns of HRH1 or HRH2 in human CRC tissues and the Kaplan-Meier overall survival-based outcomes were analyzed using the R2 platform (database, n = 232, GSE17538). Raw P values are shown. B: correlations of HRH1, HRH2, and the HRH2/HRH1 gene expression ratio with the overall survival of colorectal cancer patients (n = 174, GSE17536) were analyzed using PROGgeneV2 platform (database).

The observation that H2R activation counteracts H1R signaling in colon cancer cells suggests that the ratio of HRH2/HRH1 gene expression may provide insight in terms of patient survival. Another platform, PROGgeneV2 (human gene expression profiling database) (22), allowed us to examine the correlations between the ratios of HRH2 and HRH1 gene expression and the survival of stage-matched colorectal cancer patients. We found that neither HRH1 nor HRH2 gene expression was a robust prognostic marker independently in this data set (GSE17536) but that patients with an elevated ratio of HRH2 to HRH1 gene expression exhibited markedly improved overall survival (P = 0.035, Fig. 6B). A similar tendency was also observed in other data sets including GSE41258, GSE24551, and GSE29621 (data not shown). Therefore, the expression ratio of two human histamine-receptor genes, HRH2/HRH1, can potentially be used as a prognostic marker for CRC patients. These findings were consistent with our observations in the Apc^Min/+ mouse model experiments.

In our studies, H1R promoted tumorigenesis in murine colon tumors and elevated HRH1 expression was correlated with a relatively poor survival of CRC patients. In addition, increased HRH1 expression was also correlated with poor prognosis in pancreatic cancer (GSE21501, data not shown). These data strongly indicate that H1R function is cell and tissue type specific. In addition to colorectal cancer, we also compared the expression patterns of HRH1 and HRH2 across 15 different human cancers using the R2 platform. Different cancers displayed different expression ratios of HRH1 or HRH2. For example, HRH1 mRNA quantities were greater than HRH2 in esophageal cancer, whereas the quantities of HRH2 mRNA were greater than that of HRH1 in chronic lymphocytic leukemia (data not shown), suggesting that histamine-receptor signaling pathways may act similarly in neoplasms of the gastrointestinal tract and differently in other tissues.

DISCUSSION

Gut microbe-derived histamine can suppress colon cancer in a mammalian host, and the impact of microbial histamine on the development of colonic neoplasms depends on the relative balance between H1R and H2R signaling pathways. In our mouse model studies, H2R signaling suppressed inflammation-associated tumorigenesis, whereas H1R signaling appears to promote gut inflammation and colonic carcinogenesis. In human colon cancer-derived cells and immune cells, H2R signaling counteracted H1R-mediated MAPK activation, and MAPK activation contributes to the development of chronic intestinal inflammation and cancer. These findings in human cell culture and mouse models are consistent with patient outcomes based on gene expression profiles in human CRC, whereby elevated HRH1 and reduced HRH2 gene expression were correlated with worse patient outcomes. Together, the data suggest that the ratio of HRH2/HRH1 expression in human tissue could represent a new prognostic biomarker for IBD and CRC and such insight could point the way to new disease preventive and therapeutic strategies that rely on histamine-receptor signaling.

Intestinal epithelial cells secrete proinflammatory chemokines, including IL-8 and CCL20, to recruit immune cells to local sites of antigen presentation or injury and elicit inflammation. We demonstrate that histamine elevates expression of both IL8 and CCL20 through H1R/MAPKs in intestinal epithelial cells, such as ERK and JNK and that these effects were reversed by H2R activation. Prior studies showed that histamine promotes IL-8 expression through H1R signaling in human bronchial epithelial cells (3). We have extended this observation to intestinal epithelial cells, which may include both H1R and H2R signaling pathways to maintain intestinal homeostasis. Tilting toward increased H1R (relative to H2R) signaling may promote chemokine production and induce inflammation. Consistent with our results, R2 database analysis showed elevated HRH1 gene expression in the IBD mucosa (data not shown). In addition, several previous studies reported increased proinflammatory cytokines, such as IL-8 and CCL20, in the IBD mucosa (36, 47).

In the inflamed gut mucosa, infiltrating immune cells are a major source of cytokines, in which IL-6 is a key contributor to the development of IBD and inflammation-associated CRC (6). We show that LPS-induced IL6 expression was elevated via H1R signaling and suppressed via H2R signaling by histamine in human monocytes and mouse macrophages. Our findings agree with those of other studies, where it has been shown that histamine suppresses LPS-driven proinflammatory cytokine secretion (TNF, IL-12, and CXCL10) in dendritic cells (18) and in peritoneal macrophages via H2R signaling in mice (44). Our results suggest that long-term use of H2R blockers might promote colonic tumor initiation in IBD. In contrast, H1R antagonists (conventional antihistamines) and H2R agonists may suppress inflammation by potently suppressing LPS-induced p38 MAPK activation in macrophages. The p38 MAPK pathway is a key signaling pathway in chronic inflammation and cancer development (73). In IBD, activation of p38 signaling results in elevated TNF expression (69) and in mouse models of colon cancer; inactivation of p38 MAPK signaling significantly suppressed colon tumor development (9). Therefore, a combination of a H2R agonist and a H1R antagonist may effectively suppress p38 MAPK and represent a promising new preventative or therapeutic approach for IBD and CRC. In clinical trials, CNI-1493, a nonselective p38 and JNK inhibitor, alleviated severe Crohn's disease in a small cohort study, suggesting that p38 could be a potential target in IBD treatment (28). However, a larger cohort study using a specific p38 inhibitor, BIRB 796, failed to treat active Crohn’s disease, due to serious adverse events (59). The p38 signaling pathway may be targeted with microbiome-based treatment strategies, possibly reducing adverse effects via a luminal treatment strategy.

The role(s) of histamine on the prevention or development of human cancers remains controversial. For example, histamine increased cell proliferation in human hepatocellular carcinoma and melanoma by acting as an autocrine or paracrine growth factor (7, 39), while a different study showed that histamine suppressed cell proliferation in pancreatic cancer (13). In considering histamine’s specific effects, allergic diseases may affect cancer risk. Several studies showed that a history of allergies yielded an inverse correlation with disease risk including particular human cancers (2, 27, 54). In a large-scale prospective study in women, those with a history of allergic disease including allergic rhinitis and asthma had a significantly lower incidence of colorectal cancer (54). A history of allergies or other atopic conditions and glioma susceptibility were found to be inversely correlated (2). In addition, a large-scale population-based case-control study revealed a significantly lower incidence of pancreatic cancer among those who had a prior history of allergies (27). The concomitant use of antihistamines, which suppress H1R signaling, may be useful for treatment of allergic disease and prevention of specific human cancers based on our findings. Meanwhile, allergy may be a risk factor for lymphatic-hematopoietic malignancies, and prostate and breast cancers (48, 49). In different types of allergic diseases, histamine may induce or exacerbate allergic reactions, mainly through H1R and H4R. Locally manipulating the balance of H1R and H2R signaling could be an ideal strategy to leverage the ability to modulate cancer treatment. Furthermore, the net effect of H2R signaling on human colon tumors may depend on the microenvironment of the tumors. A recent study demonstrated that cimetidine treatment promoted tumor growth in orthotopically implanted human colon tumors in nude mice, while cimetidine treatment suppressed tumor growth when implanted subcutaneously (1, 58). This finding sheds light on the role of the tumor microenvironment affecting tumor growth. Therefore, treatment with H1R or H2R antagonists may differentially influence primary and metastatic tumors in cancer patients.

Mammalian microbiome science is providing opportunities for scientists to interrogate known and new signaling pathways for possible contributions to cancer outcomes. Our studies with microbial histamine extend the literature on a metabolite known for more than a century, but a new “microbiome lens” enables investigators to examine histamine-receptor signaling as a communication link between gut microbes and mammals. As a result of these investigations, we can target specific human genes and signaling pathways using thousands of data points from databases linking cancer gene expression profiles and patient survival data (e.g., R2 and PROGgene). By exploring H1R and H2R signaling pathways using this approach, we detected an association between the greater ratio of H2R/H1R gene expression and longer survival in colorectal cancer. Such findings establish a direction for future studies in patients and may contribute to the development of molecular diagnostics using tissue-based gene expression. New therapeutics could be developed by combining strategies for blocking H1R signaling and promoting H2R signaling, in combination with compounds targeting other signaling pathways. Finally, such combination strategies could be used to prevent colorectal cancer in at-risk patient populations.

GRANTS

This work was supported by National Institute of Diabetes and Digestive and Kidney Diseases Grant P30-DK-56338 (to Texas Medical Center Digestive Disease Center), National Cancer Institute Grant U01-CA-170930 (to J. Versalovic), and unrestricted research support from BioGaia AB (to J. Versalovic).

DISCLOSURES

J. Versalovic serves on the scientific advisory boards of Biomica and Seed Health and also receives unrestricted research support from Biogaia, AB.

AUTHOR CONTRIBUTIONS

Z.S., Y.M.-A., and J.V. conceived and designed research; Z.S., R.S.F., M.A.E., C.G., A.H., and A.M. performed experiments; Z.S. and Y.M.-A. analyzed data; Z.S. and Y.M.-A. interpreted results of experiments; Z.S. prepared figures; Z.S. and Y.M.-A. drafted manuscript; Z.S., Y.M.-A., and J.V. edited and revised manuscript; Y.M.-A. and J.V. approved final version of manuscript.