The American Journal of Pathology, Vol. 187, No.

10, October 2017

ajp.amjpathol.org

TUMORIGENESIS AND NEOPLASTIC PROGRESSION

Gut MicrobeeMediated Suppression of

Inflammation-Associated Colon Carcinogenesis by
Luminal Histamine Production
Chunxu Gao,*y Bhanu Priya Ganesh,*y Zhongcheng Shi,*y Rajesh Rasik Shah,z Robert Fultz,yx Angela Major,y Susan Venable,*y
Monica Lugo,*y Kathleen Hoch,y Xiaowei Chen,{ Anthony Haag,*y Timothy C. Wang,{ and James Versalovic*y

From the Department of Pathology and Immunology,* the Section of Gastroenterology and Hepatology,z Department of Medicine, and the Graduate Program
in Integrative Molecular and Biomedical Sciences,x Baylor College of Medicine, Houston, Texas; the Department of Pathology,y Texas Children’s Hospital,
Houston, Texas; and the Division of Digestive and Liver Diseases,{ Department of Medicine and Irving Cancer Center, Columbia University, New York,
New York

Accepted for publication

June 21, 2017. Microbiome-mediated suppression of carcinogenesis may open new avenues for identification of therapeutic
targets and prevention strategies in oncology. Histidine decarboxylase (HDC) deficiency has been shown to
Address correspondence to
James Versalovic, M.D., Ph.D., promote inflammation-associated colorectal cancer by accumulation of CD11bþGr-1þ immature myeloid
Department of Pathology, cells, indicating a potential antitumorigenic effect of histamine. Here, we demonstrate that administration
Texas Children’s Hospital, of hdcþ Lactobacillus reuteri in the gut resulted in luminal hdc gene expression and histamine production in
Feigin Tower, 1102 Bates St., the intestines of Hdc/ mice. This histamine-producing probiotic decreased the number and size of colon
FC Ste. 830, Houston, tumors and colonic uptake of [18F]-fluorodeoxyglucose by positron emission tomography in Hdc/ mice.
TX 77030. E-mail: jamesv@ Administration of L. reuteri suppressed keratinocyte chemoattractant (KC), Il22, Il6, Tnf, and IL1a gene
bcm.edu. expression in the colonic mucosa and reduced the amounts of proinflammatory, cancer-associated cytokines,
keratinocyte chemoattractant, IL-22, and IL-6, in plasma. Histamine-generating L. reuteri also decreased the
relative numbers of splenic CD11bþGr-1þ immature myeloid cells. Furthermore, an isogenic HDC-deficient
L. reuteri mutant that was unable to generate histamine did not suppress carcinogenesis, indicating a
significant role of the cometabolite, histamine, in suppression of chronic intestinal inflammation and
colorectal tumorigenesis. These findings link luminal conversion of amino acids to biogenic amines by gut
microbes and probiotic-mediated suppression of colorectal neoplasia. (Am J Pathol 2017, 187: 2323e2336;
http://dx.doi.org/10.1016/j.ajpath.2017.06.011)

Colorectal cancer (CRC) is the third most common cancer of CRC.10e12 Manipulation of the gut microbiome by pro-
and the third leading cause of cancer-related mortality.1 biotics could provide new therapeutic strategies for CRC
Population-based cohort studies have shown that patients prevention. Several probiotic strains including Bifidobacte-
with inflammatory bowel disease have an increased lifetime rium longum,13 Lactobacillus acidophilus NCFM,14 and
risk of CRC compared with the general population.2,3 This Lactobacillus rhamnosus GG15 have shown beneficial
risk can be reduced by treatment of colitis with suppression
of intestinal inflammation.4 These observations, in Supported by the NIH (J.V.), including the National Institute of Diabetes,
conjunction with studies showing that immune cells, cyto- Digestive and Kidney Diseases grant R01 DK065075, Texas Medical
kines, and other immunomodulatory agents play a role in Center Digestive Disease Center grant P30 DK56338, National Cancer
CRC development,5 underline the association between CRC Institute grant U01 CA170930, and unrestricted research support from
and colonic inflammation. BioGaia AB (J.V.).
Disclosures: J.V. receives unrestricted research support from Biogaia
The role of the intestinal microbiome in colon cancer AB.
development has recently been investigated.6e9 Specific gut Current address of C.G., Immunology Research, Janssen Pharmaceutical
microbes and their metabolites may contribute to the cause Companies of Johnson & Johnson, Spring House, PA.

Copyright ª 2017 American Society for Investigative Pathology. Published by Elsevier Inc. All rights reserved.
http://dx.doi.org/10.1016/j.ajpath.2017.06.011
Gao et al

effects in different murine models of colon cancer. How- chirayu/proggene/database/index.php, last accessed May
ever, the molecular mechanisms mediating suppression of 12, 2017).
colonic carcinogenesis by these microbes remain unknown.
Lactobacillus reuteri is a commensal intestinal Firmicute
Bacterial Strains and Culture Conditions
and probiotic that is widely prevalent in the gastrointestinal
tracts of diverse avian and mammalian species.16 L. reuteri Bacterial strains, including L. reuteri ATCC PTA 6475 and
has been reported to suppress production of proin- its isogenic hdcA mutant (generated by induction of a stop
flammatory cytokines by intestinal epithelial cells17 and codon in hdcA gene), were described previously.18 Both
monocytes,18 in addition to reducing intestinal inflammation strains were cultured at 37 C in deMan, Rogosa, Sharpe
in different rodent models.17,19e23 A pangenomic study (MRS) media (Difco, Franklin Lakes, NJ) in an anaerobic
showed that human-derived clade II L. reuteri strains con- workstation (MACS MG-500; Microbiology International,
tained a complete chromosomal hdc gene cluster (genes Frederick, MD) supplied with a mixture of 10% CO2, 10%
hdcA, hdcB, hdcP) and the genetic capacity to convert his- H2, and 80% N2.
tidine to histamine.24 The ability to generate histamine by
strains of this clade was correlated with suppression of
human tumor necrosis factor (TNF) production.24 Studies Mouse Care
also showed that histamine is considered to be a primary
candidate immunomodulin produced by L. reuteri.18 Inac- Hdc/ BALB/c mice in which exon 5 of the Hdc gene was
tivation of histidine-to-histamine converting capacity by replaced with a neomycin cassette were originally provided
mutagenesis of the histidine decarboxylase gene (hdcA) or by Timothy C. Wang (Columbia University, New York, NY)
the histidine-histamine antiporter gene (hdcP) diminished and rederived at Baylor College of Medicine (Houston, TX)
the ability of hdc-positive L. reuteri strain ATCC (Mana- by embryo transfer. Four pups, including two male and two
ssas, VA) PTA 6475 to suppress human TNF production.18 female pups, were obtained from rederivation. Mouse tail
Exploration of histidine metabolism, particularly histamine clippings were used for mouse DNA extraction. The Hdc-
production by the gut microbes, deserves attention as a deficient genotype of the rederived mice was confirmed by
possible gateway to deepening our understanding of PCR and DNA gel electrophoresis, with neomycin amplicon
microbiome-mediated intestinal immunomodulation.25,26 primers (forward, 50 -AATGGCCGCTTTTCTGGATTCA-
The lack of functional mammalian histidine decarbox- 30 ; reverse, 50 -GGGAGCGGCGATACCGTAAAG-30 ) and
ylase (HDC), the enzyme converting L-histidine to exon 5 of Hdc gene amplicon primers (forward, 50 -
histamine, yielded increased susceptibility to inflammation- TTAGTCTTTGGGTGTTCCTGGTCA-30 ; reverse, 50 -
associated CRC in adult mice.27 Here, we set out to address CCCTGTTGCTTGTCTTCCTCAATA-30 ). Breeding and
the ability of hdcþ L. reuteri to reduce the frequency maintenance of Hdc/ mice were performed under specific
and severity of inflammation-associated colon cancer in pathogen-free conditions at Texas Children’s Hospital
Hdc/ mice and to investigate whether microbe-generated (Houston, TX). Mice were kept under filter top cages and had
metabolites may suppress inflammation-associated cancer free access to distilled water and PicoLab Rodent 50IF/6F
phenotypes exacerbated by mammalian enzyme diet. All mouse experiments were performed according to an
deficiencies. Institutional Animal Care and Use Committeeeapproved
mouse protocol at Baylor College of Medicine.

Materials and Methods

Bacterial Preparation and Administration to Mice
Association between HDC and H2R Gene Expression
and Overall Survival Rates in Colon Cancer Patients L. reuteri was cultured in MRS media, and bacteria were
harvested at exponential phase. Cells were centrifuged at
To investigate whether HDC and histamine H2 receptor 2500  g for 4 minutes, and the bacterial pellets were
(H2R; symbol: HRH2) expression is associated with resuspended in sterile MRS media for animal feeding.
changes in survival rates of colon cancer patients, the L. reuteri strains were freshly prepared before administra-
PROGgeneV2 database28 was queried by selecting gene tion to mice. Each mouse received 5  109 colony-forming
name HDC or HRH2, cancer type colorectal, and survival units of bacteria in 0.2 mL of MRS media or MRS media
measure death. Samples were divided into high and low only as control by orogastric gavage. The frequency of
gene expression groups, bifurcating at median expression bacteria administration was once per day for 7 days before
value for mRNA expression. Data were plotted and azoxymethane (AOM; Sigma-Aldrich, St. Louis, MO) in-
compared by using the Coxph function to compute hazard jection, followed by administration once per 3 days for 15
ratio estimate and related log-rank test P value according to weeks. L. reuteri was not administered during the two
Goswami and Nakshatri.28 All of the CRC data sets in the cycles of dextran sulfate sodium (DSS; 36,000 to 50,000
database (2113 patient samples in 15 data sets) were molecular weight; MP Biomedicals, Solon, OH) challenge
included in this search (http://watson.compbio.iupui.edu/ in drinking water (6 days per cycle).

2324 ajp.amjpathol.org - The American Journal of Pathology

 L. reuteri Suppresses Colon Cancer

Induction of Colon Cancer in BALB/c Mice (BD Biosciences). The accumulated data were analyzed with
FlowJo software version 10 (FlowJo, Ashland, OR) gated
At 12 weeks of age, mice in the positive control group and using forward scatter and side scatter, followed by
bacteria-treated groups received a single dose of the geno- allophycocyanin-cyanine 7 (Gr-1þ) on the y axis and
toxic colonic carcinogen AOM (12.5 mg/kg body weight) fluorescein isothiocyanate (CD11bþ) on the x axis.
by i.p. injection. These mice were challenged with two
cycles of 2% (w/v) DSS in drinking water for 6 days, with
one cycle immediately after AOM injection, followed by a PET Imaging of Living Mice
recovery period with drinking water for 2 weeks before the
Positron emission tomography (PET)/computed tomography
second cycle. Mice in the negative control group received
(CT) scanning was performed 15 weeks after AOM injec-
one dose of phosphate-buffered saline (PBS) solution
tion, just before euthanizing the mice as described previ-
instead of AOM and drinking water.
ously.23 Briefly, mice were anesthetized with isoflurane and
received 200 mCi of 18F-fluorodeoxyglucose (FDG) by i.p.
Tumor Assessment and Tissue Preparations injection. After 1 hour, these mice received 200 mL of MD-
Gastroview as contrast agent rectally via a 3.5F catheter. CT
Fifteen weeks after AOM injection, mice were euthanized, scan was performed for 10 minutes, followed by a PET scan
and specimens were collected as follows. Blood was for 20 minutes using the Inveon PET/CT Multimodality
collected from sedated mice via cardiac puncture in blood System (Siemens, Munich, Germany). Mouse images were
sample collection tubes with K2EDTA (Becton, Dickinson recorded, and FDG standardized uptake values were
and Company, Franklin Lakes, NJ) and centrifuged at analyzed blindly (R.R.S.) using Inveon Research Workplace
17,000  g for 10 minutes at 4 C to isolate plasma. The software version 1.5 (Siemens). The images of mice were
gastrointestinal tract was carefully removed, and luminal generated by OsiriX Imaging software version 3.9.4 (Pix-
contents of ileum, cecum, and colon were collected and meo, Bernex, Switzerland).
flash-frozen in liquid nitrogen. The mouse colons were
excised and opened longitudinally, and the number and size
of tumors were counted and measured blindly (M.L.). In- Determination of the Relative Abundances of L. reuteri,
testinal mucosa was scraped with an operating knife blade hdc Genes, and mRNA in Vivo
and stored in RNALater (Ambion, Austin, TX) to analyze
cytokine mRNA expression. All of the samples were stored Twelve-week-old male Hdc/ mice were given one dose of
at 80 C until analyzed. L. reuteri 6475 or hdcA mutant or MRS media only as
Mouse intestines were fixed in 10% formalin, embedded described above. After 2 days, the mice were euthanized,
with paraffin, and sectioned at 3 mm by a microtome. and total DNA and RNA from luminal contents in the distal
Hematoxylin and eosin staining was performed. colon were isolated for bacterial DNA and mRNA extrac-
tion. Relative bacterial DNA and mRNA quantities of hdc
genes were evaluated by real-time quantitative PCR and
Flow Cytometric Analysis
normalized to the housekeeping gene rpoB as described
Bone marrow and spleen samples were collected immedi- previously.23 Whole colons were fixed in Carnoy’s solution,
ately after euthanasia of mice. Bone marrowederived cells and fluorescence in situ hybridization was performed with a
from the femurs and tibia were immediately flushed with ice- probe specific to a unique 16S rRNA sequence to L. reuteri
cold Dulbecco’s modified Eagle’s medium (ATCC) con- (Reverse: 50 -GATCCATCGTCAATCAGGTGC-30 ) as
taining 10% fetal bovine serum. This procedure was followed described previously.29 Slides were counterstained with
by the addition of red blood cell lysis buffer to deplete red DAPI (Sigma-Aldrich) and imaged at 400 magnification
blood cells (BD Biosciences, San Jose, CA). Spleen samples with an Eclipse 90i fluorescence microscope (Nikon In-
were stored in ice-cold Dulbecco’s modified Eagle’s medium struments, Melville, NY).
with 10% fetal bovine serum. This step was followed by the
isolation of the spleen cells using sterile glass slides and Histamine Quantification in Hdc/ Mouse Feces
addition of red blood cell lysis buffer to the isolated cells.
Single-cell suspensions were made by filtering the cells Fecal specimens were weighed and dissolved in 200 mL of
through 40-mm filter strains. For flow cytometric analysis, 50/50 PBS/methanol by vortexing and ultrasound disruption
single-cell suspensions were stained with antibodies [1 mL of for 4 minutes with 30-second intervals. After spinning for 5
allophycocyanin-cyanine 7econjugated antieGr-1 (BD minutes, 14,000  g at 4 C, the supernatant fluids were
Biosciences), and 5 mL of fluorescein iso- size-fractionated with Amicon Ultra-0.5 mL centrifugal fil-
thiocyanateeconjugated anti-CD11b (BD Biosciences)] for ters (molecular weight cutoff, 3 kDa; Millipore, Temecula,
30 minutes on ice, in the dark and evaluated by multicolor CA). The histamine quantities in the collected supernatant
flow cytometry using a BD FACSCanto cell analyzer and fluids were determined by liquid chromatography-mass
data collected with FACSDiva software version 6 spectrometry using selected reaction monitoring (SRM).

The American Journal of Pathology - ajp.amjpathol.org 2325

Gao et al

Table 1 Primers Used in qPCR-Based Gene Expression Studies psi; collision gas, High; spray voltage, 4.5 kV; ion source
Gene/ gas 1, 20 psi; ion source gas 20, 2 psi; interface heater
primer temperature, 175 C; Q1 and Q3 resolution, unit; scan time,
name Primer pairs 100 milliseconds; de-clustering potential, 100 V; entrance
Il1a F: 50 -CAGAGAGGGAGTCAACTCATTG-30 potential, 8 V; and collision exit potential, 10 V. The in-
R: 50 -GTTTCTGGCAACTCCTTCAGC-30 strument was calibrated by using Sciex PPG calibration
KC F: 50 -GACTCCAGCCACACTCCAAC-30 standard and tuned to the manufacturer’s specifications.
R: 50 -TGACAGCGCAGCTCATTG-30 SRM transitions monitored for histamine were 112 / 95
Il22 F: 50 -TGACGACCAGAACATCCAGA-30 (20 eV) and 112 / 68 (30 eV). For histamine-d4, the SRM
R: 50 -AATCGCCTTGATCTCTCCAC-30 transitions 116 / 99 (20 eV) and 116 / 72 (30 eV) were
Ifn-g F: 50 -GCCAAGTTTGAGGTCAACAACCC-30 monitored. Data were acquired with Analyst Software
R: 50 -CCGAATCAGCAGCGACTCCT-30 version 1.6.2 (AB Sciex LP, Concord, ON, Canada) and
Tnf F: 50 -CCTCACACTCAGATCATCTTCTC-30
quantification performed using Multiquant Software version
R: 50 -GTCTTTGAGATCCATGCCGT-30
3.0.1 (AB Sciex LP).
Il6 F: 50 -CTCTGCAAGAGACTTCCATCCA-30
R: 50 -TAAGCCTCCGACTTGTGAAGTA-30
Il12 F: 50 -ATGTGTCAATCACGCTACCTCCTC-30 Cytokine Measurement by Multiplex Immunoassay in
R: 50 -GGTCTTCAGCAGGTTTCGGG-30 the Mouse Plasma
Il23 F: 50 -TAGCCTGGAACGCACATGCAC-30
R: 50 -GCAAGCAGAACTGGCTGTTGTA-30 The concentrations of murine interferon (Ifn)-g, Il-1a, Il-1b,
Il17 F: 50 -ACTCTCCACCGCAATGAAGACA-30 Il-4, Il-6, Il-10, Il-12, Il-13, Il-17A, keratinocyte chemo-
R: 50 -CCCTCTTCAGGACCAGGATCTC-30 attractant (KC), Tnf, Il-21, Il-22, Il-23, and epidermal
Gapdh F: 50 -GCCAAAAGGGTCATCATCTC-30 growth factor in the plasma were measured using cytokine
R: 50 -CACACCCATCACAAACATGG-30 multiplex kits (Millipore, Billerica, MA). Quantification of
hdcA F: 50 -GCACTAACGATAACCGTCGTC-30
cytokines was performed using the Luminex system (Aus-
R: 50 -CACCCTTATTAGCACAAACAATGA-30
tin, TX) according to the manufacturer’s instructions.
hdcP F: 50 -TCCCTACGGATACCAAGCAC-30
R: 50 -AGAGGAACGCTAAGACACCAAT-30 Briefly, 25-mL plasma samples collected above from each
rpoB F: 50 -CGTGATACTTCATTACGTGTTCCT-30 mouse were thawed completely and diluted with the same
R: 50 -AGTGAAGACTTTAACATCTTGGATGA-30 amount of Assay Buffer provided in the kits. The assays
were performed in duplicate blindly (S.V.). The reports
KC, keratinocyte chemoattractant; qPCR, real-time quantitative PCR.
automatically generated by MILLIPLEX Analyst software
version 5.1 (Millipore) were reviewed, and only cytokines
For histamine quantification, 30 mL of each sample was that were greater than the lower limit of detection and below
dried in a SpeedVac for 2 hours. The dried samples were the saturation value were considered.
mixed with 30 mL of 20 ng/mL histamine-d4 (histamine-
a,a,b,b-d4; CDN Isotopes, Pointe-Claire, Canada) solution Modulation of Mammalian Cytokine Gene Expression in
by vortexing for 1 minute. The prepared samples were the Colonic Mucosa
loaded into a Shimadzu (Kyoto, Japan) SIL-20ACxr auto-
sampler and separation was achieved using a Shimadzu To quantify the relative mRNA expression levels of Ifn-g,
Nexera-XR HPLC system. Samples (5 mL) were loaded Tnf, Il-6, Il-12, Il-23, Il-17, Il-22, Il-1a, and KC, RNA was
onto a Phenomenex (Torrance, CA) 1 mm  50 mm phe- extracted from colonic mucosa samples using the miRNeasy
nylhexyl reversed phased column equipped with a Phe- mini kit (Qiagen, Hilden, Germany). One microgram of RNA
nomenex phenylhexyl 4 mm  2 mm guard column. The was reverse-transcribed to single-stranded cDNA using the
aqueous mobile phase (A) consisted of water/acetonitrile/ RevertAid H minus First Strand cDNA Synthesis Kit (Ther-
formic acid/perfluoroheptanoic acid (99.3:0.5:0.1:0.1 v/v/v/ moFisher Scientific, Waltham, MA). Real-time RT-PCR was
v), and the organic mobile phase (B) consisted of acetoni- performed using Real-Time PCR system (Stratagene, La
trile/formic acid (99.9:0.1 v/v). Column flow was set at 80 Jolla, CA). The RT-PCR reaction mix (adjusted with water to
mL/minute. The elution gradient was optimized as follows: a total volume of 25 mL) contained 1 mL of template DNA,
started from 20% B for 0.5 minute and increased to 70% B 12.5 mL of Power SYBR Green PCR master mix (ABI, Life
over 5.5 minutes; ramp to 80% B after 0.1 minute and hold Technologies, Carlsbad, CA), and 0.5 mL of the respective
for 1 minute; ramp back to 20% B over 6 seconds and primers (10 mmol/L each). The primers used for IFN-g, IL-12,
maintained at 20% for a total chromatographic run time of IL-17, TNF, IL-6, and IL-23 quantification were
12 minutes to re-equilibrate. described previously,30 and all of the primers are shown in
SRM was performed on a Sciex (Framingham, MA) 6500 Table 1. Relative murine mRNA target gene expression
QTRAP mass spectrometer equipped with a Turbo V ion levels (Ratio Z [(Etarget)dCPtarget (ControlSample)]/
source. The mass spectrometer was operated in the positive [(Eref.)dCPref. (ControlSample)]) were normalized to the house-
ion mode under the following conditions: curtain gas of 20 keeping gene GAPDH and used as a reference. Subsequently,

2326 ajp.amjpathol.org - The American Journal of Pathology

 L. reuteri Suppresses Colon Cancer

TCGA–COAD GSE17537
Survival measure - overall survival Survival measure - overall survival
HDC gene expression HRH2 gene expression
Cohort divided at median of gene expression Cohort divided at median of gene expression

1.0
1.0

0.8
0.8
Overall survival

Overall survival
0.6
0.6

0.4
0.4

0.2
0.2

HR: 0.6(0.36-1)│PVAL: 0.0501286 3 Years HR: 0.02(0-0.56)│PVAL: 0.0211746 3 Years 5 Years

HIGH EXPRESSION HIGH EXPRESSION
0.0

0.0
LOW EXPRESSION LOW EXPRESSION

0 500 1000 1500 0 500 1000 1500 2000 2500 3000

Days Days

Figure 1 Overall survival rates increase in colon cancer patients with elevated human HDC and HRH2 gene expression. The overall survival rates in
colorectal cancer patients based on HDC or HRH2 gene expression categories were plotted using the PROGgeneV2 database. Representative figures with the
lowest P value are shown for HDC (A) and HRH2 (B). HR, hazard ratio; PVAL, P value.

intestinal mucosal cytokine gene expression values of the of HDC gene expression demonstrated improved survival in
control group were set to 1.0 and used as the calibrator to CRC (TCGA_COAD; P Z 0.05) (Figure 1A).
identify the relative mRNA fold difference between the We further investigated the association between H2R
negative control group (MRS/PBS-H2O), positive control gene expression and survival rates in CRC patients because
group (MRS/AOM-DSS), L. reuteri 6475etreated group (L. our previous study23 indicated an H2R-mediated suppres-
reuteri 6475/AOM-DSS), and isogenic L. reuteri hdcA sion of colitis by histamine. Elevated patterns of H2R gene
mutantetreated group (hdcA mutant/AOM-DSS). (HRH2) expression was correlated with enhanced survival
among patients with CRC (GSE17537; P Z 0.02)
Statistical Analysis (Figure 1B), indicating a potential antitumorigenic effect by
H2R up-regulation.
Statistical analysis was performed using GraphPad Prism These findings based on human cancer gene expression
software version 5 (GraphPad Inc., La Jolla, CA). For profiling set the stage to explore whether microbial Hdc
numeric variables that fit normal distribution (determined by expression and histamine generation could complement
Kolmogorov-Smirnov test), data were presented as histamine deficiency in Hdc/ mice and provide mecha-
means  SD, and different groups were compared with the nistic insights in murine models of CRC.
t-test (two groups) or one-way analysis of variance (more
than two groups). Otherwise, data were presented as box
and whiskers plots showing the median values and 10th and L. reuteri 6475 Administration Increases hdc Gene
90th percentiles or scatter plots showing the median values. Expression and Histamine Production in the Intestines
Different groups were compared by a nonparametric U-test of Hdc/ Mice
(two groups) or Kruskal-Wallis test.
To explore whether oral administration of L. reuteri
Results increased the relative abundance of this bacterial species in
the intestines of Hdc-deficient mice, fluorescence in situ
Elevated Human HDC and H2R Gene Expression Are hybridization was performed in the colons of Hdc/ mice 2
Associated with Improved Survival Outcomes in Colon days after L. reuteri administration. Orogastric gavage with
Cancer Patients wild-type L. reuteri 6475 and HdcA-deficient L. reuteri
increased the relative abundances of L. reuteri in the colonic
HDC is the key enzyme involved in histidine-to-histamine lumen visualized by fluorescence microscopy compared
conversion and histamine generation in mammalian and with control mice that did not receive exogenous bacteria
bacterial cells. To investigate the association between HDC (Figure 2A). In addition, the relative abundances of hdcA
gene expression and survival rates in CRC patients, a DNA and hdcA and hdcP mRNA were significantly
large-scale human cancer gene expression database named increased, demonstrating active gene expression in the in-
PROGgene28 was explored. Patients with elevated patterns testinal lumen by viable bacteria (Figure 2, BeD).

The American Journal of Pathology - ajp.amjpathol.org 2327

Gao et al

Figure 2 Increased abundance of bacterial histidine decarboxylase (Hdc) mRNA and histamine in the mouse intestine after Lactobacillus reuteri
supplementation. A: Representative figures showing the microscopic visualization of L. reuteri in the colon of Hdc/ mice. Media only, L. reuteri 6475, or
HdcA-deficient L. reuteri were administered to mice orally, followed by L. reuteri speciesespecific fluorescence in situ hybridization after 2 days. Red indicates
cyanine 3-conjugated probe to label L. reuteri; blue, DAPI-labeled host cell nuclei. B: Relative abundance of hdcA gene in mouse gut microbiome in different
groups. C and D: Relative quantities of hdcA (C) and hdcP (D) mRNA in colonic luminal contents of mice. E: Histamine quantities in mouse feces determined by
liquid chromatography-mass spectrometry. n Z 5 for each group (AeD); n Z 15 to 20 for each group (E). *P < 0.05, ***P < 0.001. MRS, deMan, Rogosa,
Sharpe.

Moreover, L. reuteri 6475 administration significantly control mice that received PBS and drinking water, instead
increased histamine quantities in mouse feces, in contrast to of AOM and 2% DSS, did not develop tumors (Figure 3,
the lack of increased amounts of histamine after adminis- BeE). Positive control mice that were challenged with
tration of L. reuteri defective in HDC function (hdcA AOM/DSS and gavaged with MRS media, but did not
mutant) (Figure 2E). These studies indicate that i) L. reuteri receive exogenous bacteria, developed colonic tumors.
administration increased its relative abundance in the guts of Administration of exponential phase L. reuteri 6475 with a
Hdc/ mice, and the presence of L. reuteri in the intestine wild-type allele of the HDC gene significantly reduced the
was not affected by the presence or absence of an intact number and size of colonic tumors in male mice compared
HCD gene in L. reuteri; ii) L. reuteri can express the HCD with the positive control group. However, oral administra-
and histidine/histamine antiporter genes in the mammalian tion of Hdc-deficient L. reuteri did not suppress colonic
intestines of Hdc/ mice; iii) hdc-positive L. reuteri is able carcinogenesis. The relative abundances of hdcA DNA and
to generate histamine in the gut, providing a mechanistic hdcA and hdcP mRNA in the feces of the experimental mice
underpinning for histamine-mediated immunomodulation were also determined (Supplemental Figure S1), and these
by the mammalian gut microbes. bacterial genes showed a similar pattern as luminal contents
(Figure 2, BeD), indicating that AOM and DSS did not
HDC Gene Is Necessary for L. reuteri 6475 to Attenuate affect L-histidine metabolism by L. reuteri.
Colonic Carcinogenesis in Vivo
Abdominal Imaging Yields Evidence of Suppression of
To probe the ability of histamine-generating L. reuteri 6475 Colon Tumorigenesis by L. reuteri
to suppress CRC associated with chronic intestinal inflam-
mation, the combination of AOM and DSS was used to To further confirm the antitumorigenic effects of L. reuteri
induce colitis-associated colon cancer in 12-week-old male in the AOM/DSS-induced Hdc-deficient mouse model of
Hdc/ BALB/c mice (Figure 3A). We found that negative colon cancer, PET imaging was applied to evaluate the

2328 ajp.amjpathol.org - The American Journal of Pathology

 L. reuteri Suppresses Colon Cancer

Figure 3 Lactobacillus reuteri 6475 but not the hdcA mutant attenuates colonic carcinogenesis in vivo. A: Procedures for mouse experiments.
B: Representative colon images of mice in the negative control group [deMan, Rogosa, Sharpe/phosphate-buffered salineewater (MRS/PBS-H2O)], positive
control group [MRS/azoxymethane-dextran sulfate sodium (AOM-DSS)], L. reuteri 6475etreated group (L. reuteri 6475/AOM-DSS), and isogenic L. reuteri hdcA
mutantetreated group (hdcA mutant/AOM-DSS). C: Representative microscopic colon images (hemoxylin and eosin stained) from mice in different groups.
D and E: Numbers of both large (>3 mm) and small (<3 mm) colonic tumors from mice in different groups. n Z 8 to 10 for each group (D and E). **P < 0.01,
***P < 0.001. Scale bars Z 500 mm (C).

process of colonic carcinogenesis and suppression of compared with the wild-type strain and consistent with
colonic neoplasia in living mice before euthanasia unfettered colon carcinogenesis (Figure 4, B and C, and
(Figure 4A). Trace amounts of FDG signal were evident in Supplemental Video S4). Together, these results showed
the colon and surrounding abdominal sites, indicating low that histamine-generating lactobacilli could suppress carci-
baseline glucose uptake in the colons of healthy mice nogenesis by interkingdom complementation, whereas
(Figure 4, B and C, and Supplemental Video S1). In positive isogenic strains incapable of generating histamine did not
control mice, hot spots in the colon were observed, and the suppress colonic carcinogenesis.
relative FDG intensity in the whole mouse colon was
significantly increased compared with the negative control Histamine-Generating L. reuteri Suppresses
group (Figure 4, B and C, and Supplemental Video S2), Proinflammatory Cytokine Gene Expression in the
indicating the presence of colonic tumors and increased Colonic Mucosa
glucose uptake in the colons of mice that received AOM
plus DSS. L. reuteri 6475etreated mice showed reduced Specific proinflammatory cytokines have been reported to
numbers of hot spots in the colon and significantly contribute to the development of colon tumorigenesis by
decreased abdominal FDG intensities compared with posi- promoting the formation of a tumor-supportive microenvi-
tive control mice, indicating that histamine-generating strain ronment.31 By analyzing cytokine gene expression of selected
L. reuteri 6475 could suppress metabolic activity associated cytokines in the colonic mucosa (Table 1), we showed that
with colon carcinogenesis in live animals (Figure 4, B AOM/DSS challenge induced expression of proinflammatory
and C, and Supplemental Video S3). Meanwhile, the cytokine genes encoding KC, Il-22, Il-6, Tnf, and Il-1a
isogenic hdcA mutant strain deficient in HDC activity did compared with healthy Hdc/ mice (Figure 5). Oral treatment
not yield similar effects and showed increased numbers of of AOM/DSS-challenged mice with hdc-positive L. reuteri
hot spots and increased abdominal FDG intensities 6475 reduced the relative gene expression of these cytokines,

The American Journal of Pathology - ajp.amjpathol.org 2329

Gao et al

Figure 4 Attenuation of colonic tumorigenesis as shown by positron emission tomography (PET) imaging. A: A schematic diagram shows the PET imaging
procedures with live mice. B: Representative mouse images captured by PET-computed tomography (CT) scanning in each group. The color code bar represents
fluorodeoxyglucose (FDG) signal intensity. The abdominal regions are indicated with red boxes. C: Standardized uptake values of FDG signal in the whole
mouse colon in different groups. n Z 5 to 6 for each group (B and C). *P < 0.05. AOM, azoxymethane; DSS, dextran sulfate sodium; L. reuteri, Lactobacillus
reuteri; MRS, deMan, Rogosa, Sharpe; PBS, phosphate-buffered saline.

whereas the isogenic L. reuteri strain lacking functional HdcA Il-22, and Il-6 in plasma, compared with that of control
did not inhibit gene expression of these proinflammatory cy- mice (Figure 6). L. reuteri 6475 administration diminished
tokines. Detection of other cytokine mRNAs by real-time concentrations of these three cytokines (KC, Il-22, Il-6),
quantitative PCR yielded either undetectable results (IL-17) whereas administration of isogenic L. reuteri hdcA mutant
or no significant differences (Il-12, Il-23, and Ifn-g) among the lacking histamine-producing capacity did not reduce these
groups (ie, cytokines that this probiotic intervention did not same cytokine concentrations in AOM/DSS-treated Hdc/

repress), indicating that histamine-generating L. reuteri sup- mice. The results with KC, Il-22, and Il-6 are consistent
pressed a subset of proinflammatory murine cytokine genes in in terms of plasma protein quantitation and mucosal gene
the colonic mucosa. expression. Other cytokines measured in plasma were
either not detectable (Il-1b, Il-21, Il-23, epidermal growth
Circulating Cytokine Concentrations in Plasma Are factor) or showed no significant changes (Il-4, Il-17, Ifn,
Reduced by Histamine-Generating L. reuteri Il-1a, Il-12, Tnf, Il-10, Il-13) among groups, indicating
selective reduction of circulating cytokines in plasma by
To further investigate the association between cytokines histamine-generating L. reuteri. Together with cytokine
and CRC, we performed multiplex analyses of 16 cyto- gene expression in the colonic mucosa, histamine gener-
kines (Supplemental Table S1) in the plasma by using a ation by lactobacilli appeared to be important for
Luminex system (Millipore). AOM/DSS treatment in suppression of particular proinflammatory and cancer-
Hdc/ mice increased the relative concentrations of KC, associated cytokines.

2330 ajp.amjpathol.org - The American Journal of Pathology

 L. reuteri Suppresses Colon Cancer

Figure 5 Histamine-generating Lactobacillus reuteri suppresses mammalian cytokine gene expression in colonic mucosa. Wild-type L. reuteri 6475
administration significantly decreases proinflammatory cytokine keratinocyte chemoattractant (KC) (A), IL-22 (B), IL-6 (C), tumor necrosis factor (Tnf) (D),
and Il-1a (E) expression in Hdc/ mouse colonic mucosa in mRNA level determined by quantitative RT-PCR, whereas the isogenic L. reuteri hdcA mutant which
lacks the histamine-producing capacity lacks such effects. *P < 0.05, **P < 0.01, and ***P < 0.001. AOM, azoxymethane; DSS, dextran sulfate sodium; MRS,
deMan, Rogosa, Sharpe; PBS, phosphate-buffered saline.

Abundance of Splenic Immature Myeloid Cells Is proinflammatory cytokines produced by relatively mature
Suppressed by hdcþ L. reuteri myeloid cells.

The absence of endogenous histamine leads to markedly

increased numbers of immature myeloid cells (IMCs; Discussion
CD11bþ Gr-1þ) after AOM/DSS treatment, and this
finding is associated with cancer progression in mam- Histamine generated by intestinal microbes may supplement
mals.27 Here, we found that the percentage of CD11bþGr- histamine-generating capacity by mammalian cells and may
1þ IMCs in the spleen was significantly increased in AOM/ offer new possibilities for microbiome-mediated gene ther-
DSS-challenged mice compared with healthy control mice apy. Our prior studies showed that microbial histamine may
(Figure 7, B and C), consistent with a previous report.27 L. suppress intestinal inflammation in an acute colitis model.23
reuteri 6475 administration in AOM/DSS-challenged mice From the studies by Yang et al27 that showed an increased
significantly decreased the relative abundance of susceptibility of Hdc/ mice to inflammation-associated
CD11bþGr-1þ IMCs compared with mice that did not colonic cancer induced by AOM/DSS challenge, we
receive exogenous lactobacilli. A similar, although less addressed the ability of histamine-producing gut microbes
dramatic pattern, was observed in the bone marrow such as L. reuteri to complement the histamine deficiency in
(Figure 7, A and C). These observations are consistent with Hdc/ mice by interkingdom complementation. Our results
the conclusion that histamine-generating probiotic L. reu- indicate that histamine-producing L. reuteri 6475 protected
teri 6475 may attenuate AOM/DSS-induced colon carci- Hdc/ mice from AOM/DSS-induced inflammation-
nogenesis, at least in part, via enhanced maturation of associated colonic cancer. The bacterial enzyme, HDC,
circulating myeloid cells and concomitant reduction of must be present in this microbe to generate histamine as the

The American Journal of Pathology - ajp.amjpathol.org 2331

Gao et al

Figure 6 Plasma cytokine concentrations are reduced by histidine decarboxylase (Hdc)A-positive Lactobacillus reuteri. Wild-type L. reuteri 6475 admin-
istration significantly decreases the abundances of proinflammatory cytokines keratinocyte chemoattractant (KC) (A), IL-22 (B), and IL-6 (C) in Hdc/ mouse
plasma determined by Luminex assays, whereas the isogenic L. reuteri hdcA mutant that lacks the histamine-producing capacity lacks such effects. *P < 0.05,
**P < 0.01, and ***P < 0.001. AOM, azoxymethane; DSS, dextran sulfate sodium; MRS, deMan, Rogosa, Sharpe; PBS, phosphate-buffered saline.

bioactive compound in the intestine, yielding anti- luminal gut microbes.33 Our prior studies demonstrated
inflammatory and antitumorigenic effects (Figure 8). that gut microbes such as L. reuteri suppressed intestinal
Histamine receptors on immune cells and the intestinal inflammation via H2R.23 Gut microbes such as L. reuteri
epithelium are key factors driving the responsiveness of were capable of activating intracellular mammalian protein
mammalian cells to histamine generated by intestinal mi- kinase A and inhibiting extracellular signal-regulated ki-
crobes. H2R is present on the human intestinal epithe- nase signaling via histamine receptors.18 Past studies
lium,32 and H2R has been proposed to play a key role in documented H2R-mediated signaling via cAMP production
responses by mammalian cells to histamine generated by and protein kinase A activation, followed by inhibition of

Figure 7 Immature myeloid cells (IMCs) in bone marrow and spleen are reduced by Lactobacillus reuteri 6475. A and B: Percentages of CD11bþGr-1þ IMCs
acquired by flow cytometric analysis of the bone marrow (A) and spleen (B) samples of Hdc/ mice. C: Fluorescence-activated cell sorting analysis shows the
relative abundance of CD11bþ Gr-1þ IMCs in the bone marrow and spleens in Hdc/ mice gated using forward scatter and side scatter, followed by
allophycocyanin-cyanine 7 (Gr-1þ) on y axis and fluorescein isothiocyanate (CD11bþ) on x axis. n Z 3 to 4 for each group (A and B). ***P < 0.001. AOM,
azoxymethane; DSS, dextran sulfate sodium; MRS, deMan, Rogosa, Sharpe; PBS, phosphate-buffered saline.

2332 ajp.amjpathol.org - The American Journal of Pathology

 L. reuteri Suppresses Colon Cancer

Figure 8 Proposed mechanism describing gut

microbeemediated suppression of inflammation-
associated colon carcinogenesis by bacterial his-
tamine production. azoxymethane (AOM)/dextran
sulfate sodium (DSS) treatment induces keratino-
cyte chemoattractant (KC), Il6, Tnf, Il22, and Il1a
gene expression in the colon, increases IL-6,
IL-22, and KC protein concentrations in mouse
plasma, and increases abundance of immature
myeloid cells (IMCs) in the spleen. These effects
may all contribute to colon carcinogenesis. When
Hdc/ mice receive hdcþ Lactobacillus reuteri,
histamine is produced in the intestinal lumen via
histidine decarboxylase (Hdc) protein machinery
in L. reuteri: L-histidine is converted to histamine
by HdcA and histamine is exported to the gut
lumen by histidine-histamine antiporter (HdcP).
L. reuteriederived histamine may activate hista-
mine H2 receptors (H2Rs) on the intestinal
epithelium and trigger anticarcinogenic pathways
as indicated by suppression of mammalian cyto-
kine gene expression in the colonic mucosa and
reduction of IMCs in the spleen. TNF, tumor
necrosis factor.

c-Raf and mitogen-activated protein kinase kinase/ Myeloid cell populations are altered in Hdc/ mice chal-
extracellular signal-regulated kinase signaling.18,34e36 Be- lenged with AOM/DSS. A marked accumulation of CD11bþ
sides allergic diseases, other disease phenotypes may be Gr-1þ IMCs in the spleen suggests that myeloid cell matura-
affected by histamine signaling. For example, tion programs may be affected by the absence of histamine.
H2R-mediated signaling may offer protection against CD11bþGr-1þ IMC populations have expanded in terms of
nonalcoholic fatty liver disease,37 and histamine type 1 cell numbers in the spleens of cancer-bearing mice,49 with
receptoremediated signaling may contribute to irritable greatly increased IMC populations observed in Hdc-deficient
bowel syndrome.38 The opposing proinflammatory (hista- mice.27 Our present study shows that oral administration of
mine type 1 receptor) and anti-inflammatory (H2R) effects histamine-producing bacteria reduced the numbers of
of histamine receptors may depend on differences in CD11bþGr-1þ IMCs in the bone marrow and spleen, consis-
relative tissue distributions of histamine receptors. Hista- tent with previous studies.27 These findings support the notion
mine generated by microbial or mammalian cells may that exogenous histamine from the gut microbes can stimulate
contribute to tissue homeostasis or pathologic processes maturation or proliferation of myeloid cells at distant sites.
based on relative local concentrations of histamine and its CD11bþGr-1þ IMCs are known to produce IL-6 and other
receptors. Gut microbes may represent an important source proinflammatory cytokines and may help explain how myeloid
of histamine in mammals with implications for intestinal cell lineages may contribute to changes in cytokine quantities
inflammation and colorectal neoplasia. and inflammation-associated carcinogenesis.
Histamine-producing L. reuteri suppressed gene expression Possible effects of histamine on cancer progression remain
and production of several cytokines, in contrast to the mutant controversial.50,51 Significantly increased HDC activity was
strain lacking intact bacterial HDC. KC shares many functional found in tumor specimens in a series of 10 surgical patients
properties with IL-839 and has been reported to promote colon with colorectal carcinoma, suggesting possible dysregulation
cancer growth, progression, and metastasis.40,41 IL-22 was also due to genetic abnormalities (epiphenomenon) or a role for
shown to promote gastric cancer cell invasion42,43 and colon HDC enzymatic activity in oncogenesis.52 However, a defi-
cancer stemness.44 IL-6 has been considered as a key regulator ciency of endogenous histamine was shown to promote
of CRC development,45,46 and increased quantities of plasma inflammation-associated colonic cancer by promoting the
IL-6 were correlated with a poor prognosis in a variety of accumulation of IL-6 and CD11bþGr-1þ IMCs.27 With
cancers, including colon cancer.47 Cytokine signaling in CRC respect to human cancers, Lactobacillus was shown to be
provides opportunities for therapeutic intervention, and some diminished in CRC patients compared with healthy control
clinical trials targeting selected cytokines have been completed participants,53,54 and some Lactobacillus species have been
or are ongoing.48 Our findings on gut microbe-mediated sup- negatively associated with the risk of colon cancer.55,56
pression of cancer-associated cytokines may open new ave- Although the correlation between Lactobacillus HDC activ-
nues for preventive and therapeutic strategies in oncology. ity and CRC risk has not been reported, emerging evidence

The American Journal of Pathology - ajp.amjpathol.org 2333

Gao et al

indicates that allergy and atopy are inversely associated with analyses; R.R.S. performed PET imaging analysis; R.F.
glioma development and CRC in patients.57e59 Histamine- contributed to cell isolation and flow cytometry; A.M. per-
generating microbes were recently demonstrated to be formed histologic examination and immunohistochemical
increased in the intestines of asthma patients.60 These findings staining; S.V. performed the Luminex assays for multiplex
suggest that excessive histamine and other mediators in pa- protein profiling; M.L. helped with mouse sample collections;
tients with allergic disease may protect individuals from can- K.H. and A.H. performed histamine quantification; X.C. and
cer. Our present study showed that patients with elevated T.C.W. provided the Hdc/ mice and helpful suggestions for
patterns of HDC and H2R gene expression demonstrated our study and writing of the manuscript; J.V. provided guid-
improved survival rates in human CRC. Future investigations ance, helped design the experiments, co-wrote the manuscript,
may clarify whether the antitumorigenic effects in the present and provided funding.
study by L. reuteri are mediated via H2R and whether dif-
ferences in H2R distribution affect cancer phenotypes. Supplemental Data
This report highlights the potential importance of luminal
conversion of L-histidine to histamine by intestinal microbes as Supplemental material for this article can be found at
a factor in susceptibility to colitis and colon cancer risk. His- http://dx.doi.org/10.1016/j.ajpath.2017.06.011.
tamine represents one of the many microbial metabolites that
may have profound consequences for mammalian biology,
including but not limited to intestinal immunology and
References
epithelial cell biology. Another important implication is that 1. Siegel R, Desantis C, Jemal A: Colorectal cancer statistics, 2014. CA
the mammalian microbiome may modify disease risk by Cancer J Clin 2014, 64:104e117
complementing or supplementing deficient functions. Genetic 2. Jess T, Simonsen J, Jorgensen KT, Pedersen BV, Nielsen NM,
deficiencies in specific mammalian enzymes, as demonstrated Frisch M: Decreasing risk of colorectal cancer in patients with in-
by the lack of HDC in this mouse model, may be com- flammatory bowel disease over 30 years. Gastroenterology 2012, 143:
375e381.e1. quiz e13ee14
plemented by supplying missing genes and enzymatic func- 3. Herrinton LJ, Liu L, Levin TR, Allison JE, Lewis JD, Velayos F: Incidence
tions via the mammalian gut microbiome. Biogenic amines, and mortality of colorectal adenocarcinoma in persons with inflammatory
such as histamine, may provide microbe-derived signals that bowel disease from 1998 to 2010. Gastroenterology 2012, 143:382e389
augment or alter the functions of immune cells at remote sites, 4. Beaugerie L, Svrcek M, Seksik P, Bouvier AM, Simon T, Allez M,
and these findings indicate that microbial metabolites may Brixi H, Gornet JM, Altwegg R, Beau P, Duclos B, Bourreille A,
Faivre J, Peyrin-Biroulet L, Flejou JF, Carrat F; CESAME Study
have an impact on mucosal biology, immune function, and Group: Risk of colorectal high-grade dysplasia and cancer in a pro-
cancer susceptibility. This study summarized findings with a spective observational cohort of patients with inflammatory bowel
specific microbe (L. reuteri) in the gut, microbial cometabolite disease. Gastroenterology 2013, 145:166e175.e8
(histamine), and the phenotype of inflammation-associated 5. Terzic J, Grivennikov S, Karin E, Karin M: Inflammation and colon
cancer to gain insights into probiotic-mediated modulation of cancer. Gastroenterology 2010, 138:2101e2114.e5
6. Louis P, Hold GL, Flint HJ: The gut microbiota, bacterial metabolites
oncogenesis. The importance of an amino acid, L-histidine, as and colorectal cancer. Nat Rev Microbiol 2014, 12:661e672
the substrate for luminal conversion suggests a key link be- 7. Sears CL, Garrett WS: Microbes, microbiota, and colon cancer. Cell
tween diet and probiotic functions in vivo. Known mammalian Host Microbe 2014, 15:317e328
drug targets such as histamine receptors may be reconsidered 8. Schwabe RF, Jobin C: The microbiome and cancer. Nat Rev Cancer
in new disease applications as a result of investigations in 2013, 13:800e812
9. Tjalsma H, Boleij A, Marchesi JR, Dutilh BE: A bacterial driver-
mammalian microbiology. These findings may point the way passenger model for colorectal cancer: beyond the usual suspects.
to a deeper understanding of colorectal neoplasia and Nat Rev Microbiol 2012, 10:575e582
probiotic-mediated prevention or suppression of epithelial- 10. Arthur JC, Perez-Chanona E, Muhlbauer M, Tomkovich S,
derived cancers. Uronis JM, Fan TJ, Campbell BJ, Abujamel T, Dogan B, Rogers AB,
Rhodes JM, Stintzi A, Simpson KW, Hansen JJ, Keku TO,
Fodor AA, Jobin C: Intestinal inflammation targets cancer-inducing
activity of the microbiota. Science 2012, 338:120e123
Acknowledgments 11. Arthur JC, Gharaibeh RZ, Muhlbauer M, Perez-Chanona E,
Uronis JM, McCafferty J, Fodor AA, Jobin C: Microbial genomic
We thank Eamonn Connolly and Stefan Roos (BioGaia AB, analysis reveals the essential role of inflammation in bacteria-induced
Stockholm) for providing the Lactobacillus reuteri strains colorectal cancer. Nat Commun 2014, 5:4724
12. Goodwin AC, Destefano Shields CE, Wu S, Huso DL, Wu X,
and sharing pertinent expertise, and the staff of the Small Murray-Stewart TR, Hacker-Prietz A, Rabizadeh S, Woster PM,
Animal Imaging facility at Texas Children’s Hospital, Sears CL, Casero RA Jr: Polyamine catabolism contributes to en-
especially to David A Rendon and M. Waleed Gaber, for terotoxigenic Bacteroides fragilis-induced colon tumorigenesis. Proc
positron emission tomography (PET) imaging. Natl Acad Sci U S A 2011, 108:15354e15359
C.G. designed the study, performed experiments, analyzed 13. Rowland IR, Rumney CJ, Coutts JT, Lievense LC: Effect of Bifido-
bacterium longum and inulin on gut bacterial metabolism and carcinogen-
and interpreted data, and wrote the manuscript; B.P.G. per- induced aberrant crypt foci in rats. Carcinogenesis 1998, 19:281e285
formed flow cytometry and gene expression studies, analyzed 14. Chen CC, Lin WC, Kong MS, Shi HN, Walker WA, Lin CY,
data, and wrote the manuscript; Z.S. performed histologic Huang CT, Lin YC, Jung SM, Lin TY: Oral inoculation of probiotics

2334 ajp.amjpathol.org - The American Journal of Pathology

 L. reuteri Suppresses Colon Cancer

Lactobacillus acidophilus NCFM suppresses tumour growth both in of histamine receptors H1R, H2R, and H4R, but not H3R, in the
segmental orthotopic colon cancer and extra-intestinal tissue. Br J human intestinal tract. Gut 2006, 55:498e504
Nutr 2012, 107:1623e1634 33. Ferstl R, Frei R, Schiavi E, Konieczna P, Barcik W, Ziegler M,
15. Verma A, Shukla G: Probiotics Lactobacillus rhamnosus GG, Lauener RP, Chassard C, Lacroix C, Akdis CA, O’Mahony L: Histamine
Lactobacillus acidophilus suppresses DMH-induced procarcinogenic receptor 2 is a key influence in immune responses to intestinal histamine-
fecal enzymes and preneoplastic aberrant crypt foci in early colon secreting microbes. J Allergy Clin Immunol 2014, 134:744e746.e3
carcinogenesis in Sprague Dawley rats. Nutr Cancer 2013, 65:84e91 34. Funaki C, Hodges RR, Dartt DA: Identification of the Raf-1 signaling
16. Casas IA, Dobrogosz WJ: Validation of the probiotic concept: Lacto- pathway used by cAMP to inhibit p42/p44 MAPK in rat lacrimal
bacillus reuteri confers broad-spectrum protection against disease in gland acini: role in potentiation of protein secretion. Invest Oph-
humans and animals. Microb Ecol Health Dis 2000, 12:247e285 thalmol Vis Sci 2010, 51:6321e6328
17. Liu Y, Fatheree NY, Mangalat N, Rhoads JM: Human-derived probiotic 35. Hershenson MB, Chao TS, Abe MK, Gomes I, Kelleher MD,
Lactobacillus reuteri strains differentially reduce intestinal inflammation. Solway J, Rosner MR: Histamine antagonizes serotonin and growth
Am J Physiol Gastrointest Liver Physiol 2010, 299:G1087eG1096 factor-induced mitogen-activated protein kinase activation in bovine
18. Thomas CM, Hong T, van Pijkeren JP, Hemarajata P, Trinh DV, tracheal smooth muscle cells. J Biol Chem 1995, 270:19908e19913
Hu W, Britton RA, Kalkum M, Versalovic J: Histamine derived from 36. Waltereit R, Weller M: Signaling from cAMP/PKA to MAPK and
probiotic Lactobacillus reuteri suppresses TNF via modulation of synaptic plasticity. Mol Neurobiol 2003, 27:99e106
PKA and ERK signaling. PLoS One 2012, 7:e31951 37. Yamada S, Guo X, Wang KY, Tanimoto A, Sasaguri Y: Novel
19. Madsen KL, Doyle JS, Jewell LD, Tavernini MM, Fedorak RN: function of histamine signaling via histamine receptors in cholesterol
Lactobacillus species prevents colitis in interleukin 10 gene-deficient and bile acid metabolism: histamine H2 receptor protects against
mice. Gastroenterology 1999, 116:1107e1114 nonalcoholic fatty liver disease. Pathol Int 2016, 66:376e385
20. Pena JA, Li SY, Wilson PH, Thibodeau SA, Szary AJ, Versalovic J: 38. Wouters MM, Balemans D, Van Wanrooy S, Dooley J, Cibert-Goton V,
Genotypic and phenotypic studies of murine intestinal lactobacilli: Alpizar YA, Valdez-Morales EE, Nasser Y, Van Veldhoven PP,
species differences in mice with and without colitis. Appl Environ Vanbrabant W, Van der Merwe S, Mols R, Ghesquiere B, Cirillo C,
Microbiol 2004, 70:558e568 Kortekaas I, Carmeliet P, Peetermans WE, Vermeire S, Rutgeerts P,
21. Schreiber O, Petersson J, Phillipson M, Perry M, Roos S, Holm L: Augustijns P, Hellings PW, Belmans A, Vanner S, Bulmer DC,
Lactobacillus reuteri prevents colitis by reducing P-selectin-associ- Talavera K, Vanden Berghe P, Liston A, Boeckxstaens GE: Histamine
ated leukocyte- and platelet-endothelial cell interactions. Am J receptor H1-mediated sensitization of TRPV1 mediates visceral hyper-
Physiol Gastrointest Liver Physiol 2009, 296:G534eG542 sensitivity and symptoms in patients with irritable Bowel syndrome.
22. Preidis GA, Saulnier DM, Blutt SE, Mistretta TA, Riehle KP, Gastroenterology 2016, 150:875e887.e9
Major AM, Venable SF, Barrish JP, Finegold MJ, Petrosino JF, 39. Oquendo P, Alberta J, Wen DZ, Graycar JL, Derynck R, Stiles CD:
Guerrant RL, Conner ME, Versalovic J: Host response to probiotics The platelet-derived growth factor-inducible KC gene encodes a
determined by nutritional status of rotavirus-infected neonatal mice. J secretory protein related to platelet alpha-granule proteins. J Biol
Pediatr Gastroenterol Nutr 2012, 55:299e307 Chem 1989, 264:4133e4137
23. Gao C, Major A, Rendon D, Lugo M, Jackson V, Shi Z, Mori-Akiyama Y, 40. Asfaha S, Dubeykovskiy AN, Tomita H, Yang X, Stokes S,
Versalovic J: Histamine H2 receptor-mediated suppression of intestinal Shibata W, Friedman RA, Ariyama H, Dubeykovskaya ZA,
inflammation by probiotic Lactobacillus reuteri. MBio 2015, 6:e01358-15 Muthupalani S, Ericksen R, Frucht H, Fox JG, Wang TC: Mice that
24. Spinler J, Sontakke A, Hollister E, Venable S, Oh LP, Balderas M, express human interleukin-8 have increased mobilization of immature
Saulnier DMA, Mistretta TA, Devaraj S, Walter J, Versalovic J, myeloid cells, which exacerbates inflammation and accelerates colon
Highlander SK: From prediction to function using evolutionary ge- carcinogenesis. Gastroenterology 2013, 144:155e166
nomics: human-specific ecotypes of Lactobacillus reuteri have diverse 41. Lee YS, Choi I, Ning Y, Kim NY, Khatchadourian V, Yang D,
probiotic functions. Genome Biol Evol 2014, 6:1772e1789 Chung HK, Choi D, LaBonte MJ, Ladner RD, Nagulapalli
25. Preidis GA, Versalovic J: Targeting the human microbiome with Venkata KC, Rosenberg DO, Petasis NA, Lenz HJ, Hong YK:
antibiotics, probiotics, and prebiotics: gastroenterology enters the Interleukin-8 and its receptor CXCR2 in the tumour microenviron-
metagenomics era. Gastroenterology 2009, 136:2015e2031 ment promote colon cancer growth, progression and metastasis. Br J
26. Hemarajata P, Gao C, Pflughoeft KJ, Thomas CM, Saulnier DM, Cancer 2012, 106:1833e1841
Spinler JK, Versalovic J: Lactobacillus reuteri-specific immunoreg- 42. Fukui H, Zhang X, Sun C, Hara K, Kikuchi S, Yamasaki T, Kondo T,
ulatory gene rsiR modulates histamine production and immunomo- Tomita T, Oshima T, Watari J, Imura J, Fujimori T, Sasako M,
dulation by Lactobacillus reuteri. J Bacteriol 2013, 195:5567e5576 Miwa H: IL-22 produced by cancer-associated fibroblasts promotes
27. Yang XD, Ai W, Asfaha S, Bhagat G, Friedman RA, Jin G, Park H, gastric cancer cell invasion via STAT3 and ERK signaling. Br J
Shykind B, Diacovo TG, Falus A, Wang TC: Histamine deficiency Cancer 2014, 111:763e771
promotes inflammation-associated carcinogenesis through reduced 43. Ji Y, Yang X, Li J, Lu Z, Li X, Yu J, Li N: IL-22 promotes the
myeloid maturation and accumulation of CD11bþLy6Gþ immature migration and invasion of gastric cancer cells via IL-
myeloid cells. Nat Med 2011, 17:87e95 22R1/AKT/MMP-9 signaling. Int J Clin Exp Pathol 2014, 7:
28. Goswami CP, Nakshatri H: PROGgeneV2: enhancements on the 3694e3703
existing database. BMC Cancer 2014, 14:970 44. Kryczek I, Lin Y, Nagarsheth N, Peng D, Zhao L, Zhao E, Vatan L,
29. Preidis GA, Saulnier DM, Blutt SE, Mistretta TA, Riehle KP, Szeliga W, Dou Y, Owens S, Zgodzinski W, Majewski M,
Major AM, Venable SF, Finegold MJ, Petrosino JF, Conner ME, Wallner G, Fang J, Huang E, Zou W: IL-22(þ)CD4(þ) T cells
Versalovic J: Probiotics stimulate enterocyte migration and microbial promote colorectal cancer stemness via STAT3 transcription factor
diversity in the neonatal mouse intestine. FASEB J 2012, 26:1960e1969 activation and induction of the methyltransferase DOT1L. Immunity
30. Ganesh BP, Richter JF, Blaut M, Loh G: Enterococcus faecium 2014, 40:772e784
NCIMB 10415 does not protect interleukin-10 knock-out mice from 45. Rokavec M, Oner MG, Li H, Jackstadt R, Jiang L, Lodygin D,
chronic gut inflammation. Benef Microbes 2012, 3:43e50 Kaller M, Horst D, Ziegler PK, Schwitalla S, Slotta-Huspenina J,
31. Landskron G, De la Fuente M, Thuwajit P, Thuwajit C, Bader FG, Greten FR, Hermeking H: IL-6R/STAT3/miR-34a feed-
Hermoso MA: Chronic inflammation and cytokines in the tumor back loop promotes EMT-mediated colorectal cancer invasion and
microenvironment. J Immunol Res 2014, 2014:149185 metastasis. J Clin Invest 2014, 124:1853e1867
32. Sander LE, Lorentz A, Sellge G, Coeffier M, Neipp M, Veres T, 46. Waldner MJ, Foersch S, Neurath MF: Interleukin-6ea key regulator
Frieling T, Meier PN, Manns MP, Bischoff SC: Selective expression of colorectal cancer development. Int J Biol Sci 2012, 8:1248e1253

The American Journal of Pathology - ajp.amjpathol.org 2335

Gao et al

47. Nagasaki T, Hara M, Nakanishi H, Takahashi H, Sato M, 55. O’Keefe SJ, Chung D, Mahmoud N, Sepulveda AR, Manafe M,
Takeyama H: Interleukin-6 released by colon cancer-associated fi- Arch J, Adada H, van der Merwe T: Why do African Americans get
broblasts is critical for tumour angiogenesis: anti-interleukin-6 re- more colon cancer than Native Africans? J Nutr 2007, 137:
ceptor antibody suppressed angiogenesis and inhibited tumour-stroma 175Se182S
interaction. Br J Cancer 2014, 110:469e478 56. Moore WE, Moore LH: Intestinal floras of populations that have a
48. West NR, McCuaig S, Franchini F, Powrie F: Emerging cytokine high risk of colon cancer. Appl Environ Microbiol 1995, 61:
networks in colorectal cancer. Nat Rev Immunol 2015, 15:615e629 3202e3207
49. Watanabe S, Deguchi K, Zheng R, Tamai H, Wang LX, Cohen PA, Shu S: 57. Amirian ES, Zhou R, Wrensch MR, Olson SH, Scheurer ME,
Tumor-induced CD11bþGr-1þ myeloid cells suppress T cell sensitiza- Il’yasova D, Lachance D, Armstrong GN, McCoy LS, Lau CC,
tion in tumor-draining lymph nodes. J Immunol 2008, 181:3291e3300 Claus EB, Barnholtz-Sloan JS, Schildkraut J, Ali-Osman F,
50. Kennedy L, Hodges K, Meng F, Alpini G, Francis H: Histamine and Sadetzki S, Johansen C, Houlston RS, Jenkins RB, Bernstein JL,
histamine receptor regulation of gastrointestinal cancers. Transl Merrell RT, Davis FG, Lai R, Shete S, Amos CI, Melin BS,
Gastrointest Cancer 2012, 1:215e227 Bondy ML: Approaching a scientific consensus on the association
51. Cianchi F, Vinci MC, Masini E: Histamine in cancer: the dual faces between allergies and glioma risk: a report from the Glioma Inter-
of the coin. Cancer Biol Ther 2008, 7:36e37 national Case-Control Study. Cancer Epidemiol Biomarkers Prev
52. Garcia-Caballero M, Neugebauer E, Campos R, Nunez de Castro I, 2016, 25:282e290
Vara-Thorbeck C: Increased histidine decarboxylase (HDC) activity 58. Negri E, Bosetti C, La Vecchia C, Levi F, Tomei F, Franceschi S:
in human colorectal cancer: results of a study on ten patients. Agents Allergy and other selected diseases and risk of colorectal cancer. Eur
Actions 1988, 23:357e360 J Cancer 1999, 35:1838e1841
53. Borges-Canha M, Portela-Cidade JP, Dinis-Ribeiro M, Leite- 59. Prizment AE, Folsom AR, Cerhan JR, Flood A, Ross JA,
Moreira AF, Pimentel-Nunes P: Role of colonic microbiota in colo- Anderson KE: History of allergy and reduced incidence of colorectal
rectal carcinogenesis: a systematic review. Rev Esp Enferm Dig 2015, cancer, Iowa Women’s Health Study. Cancer Epidemiol Biomarkers
107:659e671 Prev 2007, 16:2357e2362
54. Mira-Pascual L, Cabrera-Rubio R, Ocon S, Costales P, Parra A, 60. Barcik W, Pugin B, Westermann P, Perez NR, Ferstl R,
Suarez A, Moris F, Rodrigo L, Mira A, Collado MC: Microbial Wawrzyniak M, Smolinska S, Jutel M, Hessel EM, Michalovich D,
mucosal colonic shifts associated with the development of colorectal Akdis CA, Frei R, O’Mahony L: Histamine-secreting microbes are
cancer reveal the presence of different bacterial and archaeal bio- increased in the gut of adult asthma patients. J Allergy Clin Immunol
markers. J Gastroenterol 2015, 50:167e179 2016, 138:1491e1494.e7

2336 ajp.amjpathol.org - The American Journal of Pathology