Infection with the carcinogenic liver fluke Opisthorchis viverrini modifies intestinal and biliary microbiome

The FASEB Journal • Research Communication

Infection with the carcinogenic liver flukeOpisthorchis viverrini modifies intestinal and biliarymicrobiome

Jordan L. Plieskatt,*,† Raksawan Deenonpoe,‡ Jason P. Mulvenna,§ Lutz Krause,�

Banchob Sripa,‡ Jeffrey M. Bethony,*,†,1 and Paul J. Brindley*,†,1

*Department of Microbiology, Immunology, and Tropical Medicine and †Research Center forNeglected Tropical Diseases of Poverty, School of Medicine and Health Sciences, GeorgeWashington University, Washington, District of Columbia, USA; ‡Tropical Disease ResearchLaboratory, Department of Pathology, Faculty of Medicine, Khon Kaen University, Khon Kaen,Thailand; and §Infectious Disease and Cancer and �Bioinformatics Laboratory, Queensland Instituteof Medical Research, Brisbane, Queensland, Australia

ABSTRACT Opisthorchis viverrini is a fish-borne trem-atode endemic in East Asia. Following ingestion, theflukes locate to the biliary tree¸ where chronic infectionfrequently leads to cholangiocarcinoma (CCA). Themechanisms by which O. viverrini infection culminatesin CCA remain unknown. An unexplored aspect is itsinfluence on the host microbiome. In the hamster,infection with this pathogen reliably leads to CCA.Genomic DNAs of microbiota from colorectal contentsand bile of hamsters and from whole O. viverrini wereexamined in this model of fluke-induced CCA. Micro-bial communities were characterized by high-through-put sequencing of variable regions 7–9 of prokaryotic16S ribosomal DNA. Of �1 million sequences, 536,009with useable reads were assignable to 29,776 opera-tional taxonomy units (OTUs) and, in turn, to 20 phylaand 273 genera of Bacteria or Archaea. Microbialcommunity analyses revealed that fluke infection per-turbed the gastrointestinal tract microbiome, increasingLachnospiraceae, Ruminococcaceae, and Lactobacil-laceae, while decreasing Porphyromonadaceae, Erysip-elotrichaceae, and Eubacteriaceae (P<0.05). More than60 OTUs were detected in the biliary system, whichconfirmed bacteriobilia and a noteworthy communityof microbes associated with the parasites. The fluke-associated microorganisms included potential patho-gens from the Enterobacteriaceae and Listeriaceae andothers, including Cyanobacteria and Deinococci, usu-ally found in external environments. Given that opis-thorchiasis is distinguished from other helminth infec-tions by a robust inflammatory phenotype withconspicuously elevated IL-6, and that inflammation ofthe biliary system leads to periductal fibrosis, which is a

precursor of CCA, the flukes and their microbiota maytogether drive this distinctive immune response.—Plieskatt, J. L., Raksawan, D., Mulvenna, J. P., Krause,L., Sripa, B., Bethony, J. M., Brindley, P. J. Infectionwith the carcinogenic liver fluke Opisthorchis viverrinimodifies intestinal and biliary microbiome. FASEB J. 27,000–000 (2013). www.fasebj.org

Key Words: opisthorchiasis � microbiota � neglected tropicaldisease � infection-related cancer � cholangiocarcinoma

More than 20% of all human cancers in the develop-ing world are associated with infectious disease (1).Links between the hepatitis B and C viruses and hepa-tocellular carcinoma are well known, as are links withsome other viruses and bacteria. However, less wellknown is the strong association between helminth(worm) parasites and cancer. The World Health Orga-nization’s International Agency for Research on Cancerrecognizes 3 helminth infections as definitive causes ofcancer: the liver flukes Opisthorchis viverrini andClonorchis sinensis and the blood fluke Schistosoma hema-tobium (1–3). Liver fluke infection is a major publichealth problem in East Asia and Eastern Europe, wheremore than 40 million people are infected. O. viverriniinfection is endemic in Thailand, Laos, Vietnam, andCambodia (4, 5). A food-borne disease, opisthorchiasishas been extensively studied in Thailand, where esti-mates indicate that 8 million people are infected (4).Humans acquire the infection by eating raw or under-cooked fish infected with the metacercariae of the

1 Correspondence: Department of Microbiology, Immunol-ogy, and Tropical Medicine, School of Medicine and HealthSciences, George Washington University, Washington D.C.,USA. E-mail: P.J.B., [email protected]; J.M.B., [email protected]

doi: 10.1096/fj.13-232751This article includes supplemental data. Please visit http://

www.fasebj.org to obtain this information.

Abbreviations: AMI, Amazon Machine Image; CCA, cholan-giocarcinoma; gDNA, genomic DNA; GI, gastrointestinal;HVR, hypervariable region; NCBI, National Center for Bio-technology Information; PCoA, principal coordinates analy-sis; OTU, operational taxonomy unit; QIIME, quantitativeinsight into microbial ecology

10892-6638/13/0027-0001 © FASEB

The FASEB Journal article fj.13-232751. Published online August 7, 2013.

parasite (5). On ingestion, the larval parasites excyst inthe duodenum, and juvenile flukes migrate into thebiliary tree. In the bile ducts, the parasites mature over6 wk into adult flukes, which graze on the biliaryepithelia. Eggs from the parasites are shed in bile andexit the infected person in the fecal stream. Wheresanitation is less than optimal, the eggs may enterfreshwater ecosystems, where the O. viverrini eggs areingested by freshwater snails. The parasite undergoestransformation within the snail, culminating in therelease of cercariae that seek out and penetrate the skinof a freshwater fish. Infection leads to hepatobiliarydisease, including cholangitis, periductal fibrosis, cho-lecystitis, obstructive jaundice, hepatomegaly, and cho-lelithiasis (6). More problematic, experimental andepidemiologic evidence strongly implicates liver flukeinfection in the etiology of a major subtype of livercancer, cholangiocarcinoma (CCA), also referred to asbile duct cancer (1, 2).

Bile duct cancer is a malignancy with a dismalprognosis. The tumors are slow growing and spreadlongitudinally along the bile ducts, with periductal andmass-forming extensions. The poor prognosis is owingto its silent clinical character, difficulty in early diagno-sis, and limited therapeutic approaches. Median sur-vival is �24 mo. Surgical management is the onlypotentially curative treatment, but it is restricted toearly-stage disease. However, CCA has a worldwidedistribution, beyond East Asia, where patients fre-quently develop CCA de novo without obvious riskfactors. Several conditions are associated with an ele-vated risk of CCA. In the United States, primary scle-rosing cholangitis is the well-known precursor. In Thai-land and elsewhere in East Asia, where infections withliver flukes are definitive risk factors for cancer, the riskfactors share a common determinant of long-standinginflammation and chronic injury of the biliary epithe-lium, including that caused by persistent parasitism bythese fish-borne trematodes (6–11).

The precise mechanism by which O. viverrini infec-tion culminates in CCA is not known, although it islikely to be multifactorial, including a diet rich innitrosamines, local and systemic chronic inflammation,and the secretion of mitogens and other mediators bythe parasite (10, 12). One area not yet studied is theinfluence of this parasite on the microbiome of thehost. Other helminth parasites are capable of influenc-ing the microbiome, potentially modulating gastroin-testinal inflammation in response to the appearance ofpathogenic strains (13–15).

The Syrian (golden) hamster, Mesocricetus auratus,provides an informative laboratory model of liverfluke–induced periductal fibrosis of the biliary tree andbile duct cancer. O. viverrini infection of this rodentreliably leads to bile duct cancer within months (16).Given the changes in the microbiota that may accom-pany the initiation and establishment of these hepato-biliary disorders, we examined the microbiome of thegastrointestinal (GI) tract of this model, including thebiliary system, during O. viverrini infection. In our

study, infection of hamsters with this carcinogenic liverfluke perturbed the microbiome of the GI, in particularby increasing the numbers of the Lachnospiraceae,Ruminococcaceae, Lactobacillaceae, and others, whiledecreasing Porphyromonadaceae, Erysipelotrichaceae,and Eubacteriaceae. We also describe a diverse micro-bial community associated with the liver flukes withinthe mammalian biliary tree (and, thereby, also presentthe first report of the microbiome of any trematode).These perturbations in microbiota, especially the in-volvement of the bacteria and archaeans that are asso-ciated with the flukes, may drive the distinctive inflam-matory response of chronic opisthorchiasis, includingprominently elevated IL-6 (10, 11). This response pro-motes biliary periductal fibrosis in humans that, inturn, is conducive to the emergence of opisthorchiasis-associated bile duct cancer (7, 8, 10, 11).

MATERIALS AND METHODS

Ethics statement

This study involving experimental infection of hamsters withthe liver fluke O. viverrini was approved by the Animal EthicsCommittee of Khon Kaen University, based on the Ethics ofAnimal Experimentation of National Research Council ofThailand.

Infection of hamsters with O. viverrini

Metacercariae of O. viverrini were collected from infected fishobtained from reservoirs in Khon Kaen province, Thailand.The fish were digested with pepsin (17). Four Syrian goldenhamsters, Mesocricetus auratus, were infected through an intra-gastric tube with 50 metacercariae each. Four other hamsterswere included as noninfected controls, but otherwise werehoused in identical conditions in the same room as theinfected rodents. The hamsters were maintained at the ani-mal facility, Faculty of Medicine, Khon Kaen University. All 8hamsters were euthanized at 6 wk after infection, when adultO. viverrini flukes and bile were retrieved from the bile ductsand gall bladder of the infected hamsters, bile from the gallbladder of the control hamsters, and colorectal contents(feces) from each hamster. The bile was aspirated asepticallyfrom the gall bladder with a 27-gauge needle. The bile wasstored in �95% ethanol, flukes were stored in �95% ethanolor PBS (pH 7.2), and the feces were stored without buffer, at�20 to �80°C for 6 mo.

Preparation of microbial genomic DNA (gDNA)

Colorectal feces of hamsters

gDNA was isolated from feces from the colon and rectum of4 O. viverrini-infected and 4 control hamsters. Samples (�250mg) from each hamster were processed individually with theEZNA Stool DNA kit (Omega Biotek, Norcross, GA, USA),according to the manufacturer’s protocol for pathogen de-tection. Concentration, purity, and integrity of the gDNA(and from other samples described later) were determined byspectrophotometry (Nanodrop 1000; NanoDrop Technolo-gies, Wilmington, DE, USA) and 2100 Bioanalyzer/DNA12000 Kit (Agilent Technologies, Palo Alto, CA, USA) (Sup-plemental Fig. S1). The gDNAs were stored at �80°C.

2 Vol. 27 November 2013 PLIESKATT ET AL.The FASEB Journal � www.fasebj.org

Liver flukes

Two methods were used to isolate gDNA from worms pre-served in ethanol or PBS. The worms were pelleted bycentrifugation (4000 g, 10 min), supernatants were removed,and pellets were air dried at 37°C for 10 min. DNase/RNase-free water (100 �l) was added, after which flukes (about 5adult worms) were pulverized with a motorized micropestle(Kontes, Vineland, NJ, USA). gDNAs were isolated with theEZNA Bacterial DNA kit (Omega Biotek), according to themanufacturer’s protocol for difficult-to-lyse bacteria, andstored at �80°C. An alternative method was used with 1sample of worms that had been stored in PBS. These wormswere pulverized, after which the EZNA SQ Tissue DNA kit(Omega Biotek) for purification from animal tissue was used,with one modification: 0.75 mg lysozyme was added duringthe first incubation at 60°C for 20 min. In total, 3 gDNApreparations were analyzed, each from a different infectedhamster.

Bile

At necropsy, bile was collected from the gall bladders of the 4O. viverrini-infected hamsters and pooled (�500 �l). Simi-larly, bile was collected from the control (noninfected)hamsters. The bile was clarified by centrifugation (4000 g, 10min) at room temperature, the supernatant removed, and thepellets resuspended in water. gDNA was extracted from thepelleted material by using the EZNA Bacterial DNA kit(Omega Biotek), according to the kit’s protocol for difficult-to-lyse bacteria, and was stored at �80°C. No microbial DNAwas recovered from the bile of the control, noninfectedhamsters (Supplemental Fig. S1). Given that the bile ofhealthy persons and rodents is sterile, the failure to recovermicrobial DNA was not unexpected (18, 19).

Amplification of 16S rRNA genes and pyrosequencing

Universal primers, 5=-GYAACGAGCGCAACCC [1099; for-ward (F)] and 5=-AAGGAGGTGATCCAGCCGCA [1541; re-verse (R)] were used to amplify hypervariable regions (HVRs)7–9 of the bacterial 16S rRNA gene (expected size of theamplicons, �420 nt; ref. 20). Primers were synthesized anddeployed as described in the manufacturer’s literature for GSFLX Titanium Series Lib-A Chemistry (Roche DiagnosticsGmbH, Mannheim, Germany). The amplicons were purified,equally pooled, and used as templates for emulsion PCR.DNA beads from the emulsion PCRs were subjected to 454pyrosequencing with GS Titanium reagents (Roche Diagnos-tics). 16S rRNA gene sequencing was performed by SeqWright (Houston, TX, USA).

Sequence processing, OTU-based analysis, phylogeny, anddiversity analysis

FASTA sequences of amplified 16S small subunit rRNA geneswere processed through the CloVR-16S pipeline (http://clovr.org; refs. 21–23), using an Amazon Elastic ComputeCloud (EC2) instance, Amazon Machine Image (AMI) iden-tifier ami-31890e58 (Amazon, North Seattle WA, USA). Afterparsing in UCHIME to remove chimeras (24), quantitativeinsight into microbial ecology (QIIME; ref. 23) was used forcomparison and analysis of microbial communities, and theRDP classifier was used to cluster sequences and assign theminto an operational taxonomy unit (OTU) in the NationalCenter for Biotechnology Information (NCBI; Bethesda, MD,USA) taxonomy, with an identity threshold of �95% (25, 26).Subsequently, mothur (27) was used to examine the � and �

diversity of the microbial communities. After taxonomicassignment, the files were also uploaded to Calypso Web V3,(http://bioinfo.qimr.edu.au/calypso/), a data-mining andvisualization tool for 16S rRNA datasets, to further investigatediversity by using rarefaction curves, principal coordinatesanalysis (PCoA), multivariate and regression analyses, andanalysis of variance. Differences in global community compo-sition among the groups were assessed by comparing theintragroup and intergroup Jaccard distances between com-munity profiles (OTU level) by Wilcoxon rank test.

OTU-specific PCRs

OTU-specific PCRs were performed on representative bacte-rial phylotypes, as follows: Aerococcaceae Facklamia [forward(F): 5=-TAACGAGCGCAACCCTTATC, reverse (R): 5=-CCC-AATCATCTATCCCACCTTC], Flavobacteriaceae Flavobacterium(F: 5=-GAGTCAAGTCGGGAACTCTAAC, R: 5=-GGATCATG-GCTGATATCCGATTA), Acetobacteraceae Muricoccus (F: 5=-CGGCTACCTTGTTACGACTT, R: 5=-ATCAGATATCGC-GAGGCAAC), Sphingomonadaceae Novosphingobium (F: 5=-TCTGCCTCTCATTGCTGAGTTA, R: 5=-CACGCGAGTGT-GAGCTAAT), and Comamonadaceae Variovorax (F: 5=-CAAGTCCTCATGGCCCTTATAG), R: 5=-CCGATACGGCTAC-CTTGTTAC). Amplifications were performed with GoTaqpolymerase and reagents (Promega, Madison, WI, USA), withthermal cycling at 94°C, 1 min; 54°C, 45 s; and 72°C, 1 min, for35 cycles. The products were sized by electrophoresis through1–1.5% agarose in Tris acetate-EDTA, stained with ethidiumbromide, and photographed. gDNAs (4 ng/�l) were used astemplates in some cases (as were the microbial 16S rRNA geneamplicons), obtained by using the primers specific for HVRs 7–9(6 ng/�l after the latter amplicons had been sized; �420 nt;GeneRuler 1 kb Plus DNA Ladder; Thermo Scientific, Pitts-burgh, PA, USA), and purified (Wizard SV Gel and PCRClean-Up System; Promega).

Phylogenetic analysis

Phylograms were inferred by using the maximum-likelihoodmethod based on the Kimura 2-parameter model (28). Adiscrete � distribution was used to model evolutionary ratedifferences among the sites. Trees were drawn to scale, withbranch lengths measured in the number of substitutions persite. Positions containing gaps and missing data were elimi-nated. Analyses were conducted in MEGA5 (29).

Database accession

Sequence data from this study are available at the Se-quence Read Archive (SRA), NCBI, U.S. National Libraryof Medicine (Bethesda, MD, USA) under accession num-ber SRA065893.

RESULTS

OTUs from 273 genera of Bacteria and Archaea

Sequences from the colorectal feces, bile, and O. viver-rini worms were amplified with primers targeting HVRs7–9 of the prokaryotic 16S rRNA genes. Of �1 millionsequences obtained, 536,009 with useable reads wereassignable to 29,776 OTUs and, in turn, to 20 phylaand 273 genera of the domains Bacteria and Archaea(Fig. 1, Table 1, and Supplemental Table S1). Ten

3CARCINOGENIC FLUKE PERTURBS GI TRACT MICROBIOME

percent of these reads were from bile, 32% from theliver flukes, 29% from infected hamster colon, and 29%from normal hamster colon. Whereas 98% of the 273OTUs identified were bacterial, the remainder werearchaeal, from at least 2 phyla of the Archaea kingdom(Table 1 and Supplemental Table S1).

The carcinogenic liver fluke O. viverrini modified theGI tract microbiota

Significant differences in the microbiota were seen inO. viverrini-infected hamsters vs. the uninfected, con-trol hamsters. Differences were apparent at all taxo-nomic levels examined, from phylum to genus. Therepresentative taxonomic clusters presented in Fig. 1illustrate the differences at the phylum (Fig. 1A) andgenus (Fig. 1B) levels. At the phylum level, members ofSpirochaetes (P�0.02) and Bacteria_Other (P�0.01)were more prevalent in the colorectal contents of theinfected hamsters than in those of the control hamsters(Fig. 1A and Table 1), whereas at the genus level,numerous other differences were observed (Figs. 1B, 2,and 3). Specifically, the colorectal feces of the infectedhamsters exhibited more Ruminococcaceae, Lachno-spiraceae, and Lactobacillus than did those of the control

animals. By contrast, there were more Porphyromon-adaceae, Erysipelotrichaceae, and Eubacteriaceae inthe colorectal feces of control hamsters than in theinfected ones (Figs. 2 and 3). The microbial commu-nities in the colorectal feces of the infected andcontrol hamsters included a higher prevalence ofOTUs belonging to the orders Clostridaiales andBacteroidales and the families Ruminococcaceae andDesulfovibrionaceae (Fig. 2) than did the microbiotaidentified in the O. viverrini worms and bile of theinfected hamsters. Conversely, increased OTUs wereseen in worms and bile, compared with those in thecolorectal feces—specifically, Lactobacillus, Bacilli_Other, and Bacteria_Other (Figs. 1, 2, and 3B; Table 1;and Supplemental Table S1).

Striking differences in diversity among the commu-nities were evident in the colorectal feces of the controland infected hamsters, in the bile of the infectedhamsters, and in the worms living in the bile within thebile ducts (Figs. 2 and 3; Table 1; and SupplementalTable S1). Rarefaction analysis of community richness(Fig. 4A) revealed greater alpha diversity in microbialcommunities in the colon than in the bile and worms.The microbial communities of the colorectal contents

A B

Figure 1. Taxonomic clusters of microbiomes based on 16S rRNA sequences from colorectal feces from control and O.viverrini-infected hamsters and bile from the infected hamsters and from the O. viverrini worms parasitizing the hamsters. Heatmaps at the phylum (A) and genus (B) levels were created with the Skiff visualization tool of CloVR (21–23). After normalization,the values were transformed to proportions within each sample, after which the logarithm of all proportions was taken.HAMFU1-U4, colon contents (feces) from control (uninfected) hamsters; HAMF1-F4, colon contents from infected hamsters;HAMBIPOOL, bile from infected hamsters; HAMOVG1-G2, O. viverrini worms from hamsters; HAMOVS1, O. viverrini gDNAextracted with lysozyme from hamsters.

4 Vol. 27 November 2013 PLIESKATT ET AL.The FASEB Journal � www.fasebj.org

of the infected hamsters, in turn, were more diversethan those in the control animals (Figs. 2 and 4 andTable 1). The communities associated with the liver

flukes and bile were less diverse than those of thecolorectal contents. As quantified by Shannon index,�-diversity revealed a similar trend, with mean indices

TABLE 1. Bacteria and Archaea phyla and number of reads detected in hamster colon contents, hamster bile, and whole O. viverriniworms by pyrosequencing of amplicons targeting prokaryotic 16S rRNA genes

Phylum

Colon, normal Colon, infected OVBile,

infected1 2 3 4 1 2 3 4 G1 G2 S1

BacteriaActinobacteria 193 353 1,159 130 79 24 53 102 49 74 19 115Bacteroidetesa 19,632 9,291 7,999 18,615 7,284 5,558 5,616 9,011 7 10 1 3Cyanobacteria 0 0 0 0 0 0 0 1 5 0 0 1Deferribacteres 0 0 4 1 16 5 4 1 0 2 1 0Deinococcus_

Thermus0 0 0 0 0 0 0 0 0 15 0 2

Fibrobacteres 45 0 0 7 0 2 0 0 0 0 0 0Firmicutes 18,589 26,748 30,491 10,742 31,456 27,870 29,913 22,168 53,003 61,413 14,934 46,897Fusobacteria 0 4 3 13 0 0 0 0 0 0 0 0Gemmatimonadetes 0 0 0 4 0 0 0 0 1 40 0 0Othera 406 420 427 317 1,004 643 775 945 615 2,732 1,672 6,672Proteobacteria 753 1,789 2,540 6,212 2,383 959 1,571 4,914 788 2,046 31,679 799Spirochaetes 0 0 2 1 24 37 14 72 0 0 0 0TM7 0 0 0 1 12 40 2 2 0 0 0 0Tenericutes 0 0 0 1 0 2 0 0 0 0 0 0Thermomicrobia 0 0 0 0 0 0 0 0 3 0 0 0Verrucomicrobia 14 17 19 25 50 12 5 6 0 0 0 0Root_Othera 7 5 11 14 17 10 18 13 497 1,020 934 10

Total reads 157,004 152,693 171,560 54,499Composition (%) 29 29 32 10

ArchaeaCrenarchaeota 0 0 0 0 0 0 0 1 0 0 0 0Euryarchaeota 0 3 14 94 26 1 0 0 0 0 1 0Other 5 3 2 4 13 4 9 6 0 29 38 0

Total reads 125 60 68 0Composition (%) 49 24 27 0

OV, O. viverrini. aRepresentative phyla (nonexclusive list) where significant differences were apparent among the microbial communitiesin colon contents: infected hamsters, control hamsters, and the biliary tree (worms and bile).

Colon, control1 2 3 4

Colon, infected1 2 3 4

Liver flukes1 2 3

Bile1

Figure 2. Composition of the microbial communities in adult O. viverrini liver flukes, bile, and colorectal feces from control andliver fluke–infected hamsters. Comparison of microbial communities from colorectal feces from control and O. viverrini-infectedhamsters, from O. viverrini worms, and from the bile of infected hamsters. Bubble volumes reveal the number of OTUs of thegenera.Colon, control 1–4: HAMFU1-U4, control hamsters; Colon, infected 1–4: HAMF1-F4, infected hamsters; Liver flukes1–3: HAMOVG1 and -G2 and HAMOVS1, O. viverrini worms (HAMOVS1, gDNA extracted with lysozyme); Bile 1: HAMBIPOOL,bile from infected hamsters. Data are expressed as mean abundance (%).

5CARCINOGENIC FLUKE PERTURBS GI TRACT MICROBIOME

of 2.7, 2.4, 0.8, and 0.6 from infected colon, controlcolon, bile, and O. viverrini, respectively (Fig. 4B).Global community composition diverged significantlyamong these microbiomes, as measured by the Jaccarddistance (P�0.05), with the exception of O. viverriniand bile (P0.65) (not shown). O. viverrini in bileexhibited the shortest pair-wise median Jaccard dis-tance (0.43), followed by the distance of O. viverrinifrom infected feces (0.93) and bile from infected feces(0.94). (Indeed, it was not reasonable or practical toseparate the worm/bile communities, given that theworms resided in the bile.) The greatest differences inglobal community composition were found when com-paring bile with normal colorectal feces (0.97) and O.viverrini infected with control feces (0.98). PCoA (Fig.4C) confirmed the dissimilarity in global communitycomposition among the O. viverrini-infected and thecontrol feces of the hamsters and the similarity betweenthe O. viverrini worms and the bile of the hamsters fromwhich the worms were recovered. As mentioned above,it is not feasible to separate the worms from the bile,and the close identity of the microbial communities ofO. viverrini worms and the bile would reflect ingestedmicrobes contained in the bile by the worms andrelease of endogenous microbes from the flukes intothe bile where they reside. Last, rodent-to-rodent dif-ferences were not significant; each of the 4 controlhamsters exhibited similar microbiota in the colorec-tum. Similarly, individual differences among the 4infected hamsters were not significant (Figs. 1, 2, and 4and Table 1).

More genomic DNA was recovered from 250 mg ofcolorectal feces of the infected hamsters (range, 21.0–99.7 �g gDNA) than from that of control hamsters(range, 3.7–7.9 �g) (P�0.05; Supplemental Fig. S1),

indicating a difference in overall microbial load. Fur-ther, as evidenced by the rarefaction and Shannonindices, microbial diversity was greater in the colorectalfeces of the O. viverrini-infected than in the controlhamsters (Fig. 4A). However, differences in the per-centages of total OTUs were not apparent between theinfected and the control colorectal feces (Table 1 anddata not shown), despite the higher microbiotic mass inthe fluke-infected hamsters.

Noteworthy microbiota was associated with thecarcinogenic liver fluke O. viverrini

Approximately 1000 OTUs were scored in the worm/bile samples, compared with �6000 to �9000 in thecolorectal feces of the control and O. viverrini-infectedhamsters, respectively. The rarefaction plot indicatedthat most of the microorganisms associated with O.viverrini and bile had been identified, given that theslopes of these curves plateaued after �20,000 readswere sampled. By contrast, the rarefaction curves of thecolorectal feces remained steep, indicating that addi-tional reads would be needed to reach saturationcoverage and comprehensive identification of thesemicrobial communities (Fig. 4A).

More than 20 OTUs of Bacteria and Archaea werepresent at higher, often much higher, levels in theworms and bile than in the colorectum (SupplementalTable S1). These included phylotypes from the Propi-onibacterineae, Burkholderia, Brevundimonas, Comamo-nas, Enterobacteriaceae including Citrobacter, Gamma-proteobacteria_Other, Incertae sedis 5_Tepidimonas,Lactobacillus, Acinetobacter, Mesorhizobium, and Pseudomon-adaceae, including Pseudomonas, Flavimonas, and Ther-momonas (Supplemental Table S1). Members of the order

Colon, controlColon, infected

A

Colon, controlColon, infected

WormsBile

B

Figure 3. Significant differences were evident among the microbial communities in the control hamsters and in the hamstersinfected with O. viverrini. Community compositions are presented for families/genera of microorganisms that exhibiteddifferences among groups. A) Bacterial composition (mean abundance) of colorectal feces in the control and O. viverrini-infected hamsters. B) Bacterial composition of the colorectal contents (feces) of the control and O. viverrini-infected hamsters,of the whole O. viverrini worms, and of the bile from infected hamsters. *P � 0.05, **P � 0.01, ***P � 0.001.

6 Vol. 27 November 2013 PLIESKATT ET AL.The FASEB Journal � www.fasebj.org

Lactobacillales were the most abundant OTUs in theworms and bile. Furthermore, �60 OTUs that weredetected in the biliary system were not detected in thecolorectal contents by pyrosequencing (SupplementalTable S1). Not only were these OTUs from numerousphyla of the Bacteria and Archaea domains, they alsoincluded phylotypes from genera that are potential patho-gens in humans, including Bordetella, Brochothrix, Burkhold-eria, Leminorella, Pseudomonas, Serratia, and Sphingomonas,all of which are gram-negative or -positive bacteria. Thisbiliary tract microbiota also included microbes knownusually from the external environment, including fresh-water, soil, and even volcanic springs: Cyanobacteria,Methylobacterium, Mesorhizobium, Flavobacterium, Truepera,and others. The presence of these OTUs was intriguing:for example, Deinococcus from the phylum Deinococcus-Thermus (given the well-known radioresistance and resis-tance to environmental extremes of the Deinococci; ref.30) and halophiles of the order Oceanospirillales (31). Aphylogram of some of the liver fluke–associated microbesis presented in Fig. 5, to display the diversity of thisdistinctive biliary tree microbiome. This representative

sample includes 18 OTUs from 6 phyla of the Archaeaand Bacteria domains.

To further investigate whether these microbeswere indeed present only in the biliary system, weused OTU-specific PCRs targeting 5 of the OTUslisted in Supplemental Table S1, to more thoroughlysearch the colorectal feces. [The rarefaction curvefor the infected colon contents remained steep (Fig.4A), indicating that many species had yet to bedetected.] The 5 OTUs were Aerococcaceae Fackla-mia, Flavobacteriaceae Flavobacterium, Acetobacter-aceae Muricoccus, Sphingomonadaceae Novosphingo-bium, and Comamonadaceae Variovorax. We deployedoligonucleotide primers specific for the 16S rRNAgenes of each of the 5 target microbes and per-formed the PCRs with the original gDNA prepara-tions. Four of the five—the Facklamia, Flavobacterium,Muricoccus, and Variovorax OTUs—were detected alsoin the hamster feces, although the signals weregenerally weaker than those in the worm/bile gDNAand, in general, were stronger in the infected com-pared with the control hamster feces (not shown). By

A

B

C

Colon, controlColon, infected

WormsBile

■ Colon, control▲ Colon, infected♦ Worms● Bile

2.5

2.0

1.5

1.0

shan

non

inde

x

Community Diversity Genus p=0.0321

8000

6000

4000

2000

0

0 10000 20000 30000 40000 50000 60000 70000

Sequence reads sampled

Num

ber o

f OTU

s re

ad

Colon, controlColon, infectedWormsBile

PCoA Genus (jaccard)

0.3

0.2

0.1

0.0

-0.1

-0.2

-0.3

-0.4

-0.2 0.40.0 0.2 0.6

P1

P2

Figure 4. Community diversity in microbiotain O. viverrini-infected hamsters. A) Rarefac-tion plot of � diversity. B) Community diver-sity according to the Shannon index (mean)at genus-level phylotype; significant commu-nity diversity was evident. C) PCoA illustrat-ing dissimilarity among the phylotype struc-tures of the GI tract microbiota of the O.viverrini-infected and control hamsters andthe O. viverrini worms from the bile ducts ofthe infected hamsters.

7CARCINOGENIC FLUKE PERTURBS GI TRACT MICROBIOME

contrast, the Novosphingobium OTU was not detectedin the colorectal feces. Subsequently, the broad-range HVR 7, 8, and 9 primers were used again toamplify the 16S rRNA genes, and these ampliconsserved as templates for OTU-specific PCRs targetingthe 5 phylotypes. This nested, OTU-specific PCRapproach also failed to detect the NovosphingobiumOTU in feces (Supplemental Fig. S2). Together, thefindings of the high-throughput pyrosequencing andOTU-specific PCRs strongly indicated that someOTU level phylotypes were present only in the O.viverrini worms from the bile ducts, presumably trans-ported there with the parasites.

Liver fluke–associated Archaea

Phylotypes assigned to the domain Archaea wereidentified in the microbial communities of the colo-rectal feces in the control and infected hamsters,

even though the 16S rRNA gene-specific primers thatwere deployed specifically target Bacteria (ratherthan Archaea; ref. 20). The archaeal OTUs belongedto the phylum Euryarchaeota; class Methanobacteria;and genera Methanobrevibacter, Methanosphaera, Metha-nothermobacter, and others. These particular microbesare well known from the mammalian digestive tractand the oral cavity; they are methanogens (32).Several others were present in the O. viverrini adultworms, but notably, not all were apparently from theEuryarchaeota (Table 1). The phylogeny includingthe constituent phyla of the Archaea domain is notwell established, because these prokaryotes oftenoccur in extreme geographic environments and of-ten cannot be cultured. Whereas many species areknown among the invertebrates (including termites,corals, and sponges), archaeans have not been de-scribed among the flatworms (33). The pathogenic

Met

hylo

bact

eriu

m

GpIIa O

ther

Brochothrix

Halomonas

Flavobacterium

Psychrobacter

Leminorella

Serratia

Terrimonas

Dein

ococ

cus

Sphin

gobiu

m

Novosphingobium

Devosia

Thermaerobacter

Leptothrix

Oceanospirillaceae Other

Aquicella

Euryarchaeota O

ther

0.5Archaea

Deinocooccus_Thermus

Firmicutes

Cyanobacteria

Proteobacteria

Gamma-Proteobacteria

Gamma-Proteobacteria

Beta-Proteobacteria

Alpha-Proteobacteria

Alpha-Proteobacteria Gamma-Proteobacteria

Bacteroidetes

Flavobacteria

Gamma-Proteobacteria

Bacilli

Gamma-ProteobacteriaSphingobacteria

Bacteroidetes

Proteobacteria

Alpha-Proteobacteria

Deinococci

Figure 5. Phylogram of the diverse community of microbes associated with the O. viverrini worms parasitizing the biliary tree ofhamsters. Of the �60 worm-associated OTUs, 18 are presented on branches with OTU identifications. The phyla and classes towhich OTUs of these representative microorganisms were assigned also are shown. The phylogram was inferred by using themaximum likelihood method. A discrete � distribution was used to model evolutionary rate differences among sites (5categories; G, parameter1.3283). The tree is drawn to scale, with branch lengths measured in the number of substitutionsper site. The analysis involved 18 nucleotide sequences; positions containing gaps and missing data were eliminated, and thedataset included a total of 162 positions.

8 Vol. 27 November 2013 PLIESKATT ET AL.The FASEB Journal � www.fasebj.org

potential of archaeans in human disease is of increas-ing interest (32, 34).

DISCUSSION

Escalating interest in the microbiome is leading to ageneral understanding that perturbations in microbialcommunities are associated with human disease (15).Moreover, microbiomes can provide markers of risk,diagnosis, and prognosis of disease, including inflam-matory bowel disease, diabetes, obesity, and bowelcancer (35). Nematode infection can influence thediversity and ecology of intestinal bacteria. For exam-ple, Lacterobacillaceae increase in abundance in theileum of mice infected with the nematode, Heligmoso-moides polygyrus (36). The clinical consequences of theseshifts in microbiota can be beneficial; for example,therapeutic infection with whipworm ameliorates in-flammatory bowel disease (13, 14). We now report thatparasitism of the biliary tract of hamsters by O. viverrinileads to shifts in the microbiota farther down thealimentary tract, reflected in the colorectal contents,including the likely outgrowth or suppression of sensi-tive species in a fashion similar to that of other inflam-mation-related diseases (15, 35, 37). Although theinfection of hamsters with 50 metacercariae reflects aheavy infection and hence may not model closely theinfection loads of the general population in endemiclocations, it very likely models parasite burdens ofheavily infected persons who are at increased risk ofCCA (38). Moreover, this infection dose is well toler-ated by hamsters and has been used in laboratoryinvestigations of the pathogenesis of liver fluke infec-tion (39).

Our study involved infection of Syrian hamsters withO. viverrini, followed by deep sequencing of 16S rRNAgenes, to investigate the microbiota during parasitismby this fluke. As a control, the microbiota of nonpara-sitized hamsters was examined. Although a comprehen-sive catalog of the microbiome of this rodent has notbeen established, details are becoming available. Thegut microbiome has been characterized in other mam-mals and exhibits a similar microbiota at higher taxo-nomic levels (40–43). The microbiome of hamster fecesis composed primarily by members of the bacterial ordersBacteroidales and Clostridiales, with contributions by theDesulfovibrionales, Fibrobacterales, Anaeroplasmatales,Actinobacteridae, and Gammaproteobacteria (44). Weobserved a similar profile, showing that the OTUs of thephyla Bacteroidetes, Firmicutes, and Proteobacteria wereabundant in the colorectal contents.

However, perturbations in microbial communities wereevident in parasitized hamsters, including increased con-tributions from the Lachnospiraceae, Ruminococcaceae,and Lactobacillaceae and concomitant decreases in Por-phyromonadaceae, Erysipelotrichaceae, and Eubacteri-aceae in the colorectal contents.

A remarkable microbiota was associated with theflukes. First, �60 OTUs were detected in the worms

and bile, either solely or at higher levels than in thecolorectum. These biliary system microbes—those inthe worms, the bile, or both—included Lactobacillus,Burkholderia, Pseudomonas, Flavimonas, Acinetobacter, andEnterobacteriaceae family members including Citrobac-ter. In humans, inflammation of the colon is oftenaccompanied by microbial dysbiosis, characterized bythe decreased representation of obligate anaerobes andincreased abundance of facultative anaerobes of thefamily Enterobacteriaceae (42). Second, many OTUswere identified only in the biliary system, with the O.viverrini worms retrieved from the bile ducts and in bile,definitively documenting bacteriobilia in the fluke-infected hamsters. Bacterial and parasitic cholangitisare frequent complications of opisthorchiasis (45).Infection with O. viverrini can induce biliary stasis that,after long-standing obstruction, leads to infection bymicroorganisms from the intestine, resulting in ascend-ing cholangitis (8). In a similar fashion, biliary blockagedue to CCA also induces ascending cholangitis (9,45–47). We anticipate that the potentially pathogenicmicroorganisms detected in our study obtained entryinto the biliary tree during the migration of the liverfluke from the duodenum; these included Pseudomonas,Serratia, Leminorella, and Bartonella. These bacteria arefacultative pathogens and several, including Sphingomo-nas, which are also widely distributed in water and soil,are gaining increasing attention as nosocomial infec-tions (48–51).

Given that numerous species were not detected inthe feces and, occasionally, not even in the bile of theinfected hamsters and notwithstanding that investiga-tion by deeper sequencing might detect many of theseOTUs in the colon, the presence of this community ofbile duct–localized, worm-associated microbes suggeststhat they were ushered into the biliary tree by the liverflukes. Considering the spectrum of taxonomic assign-ments among this biliary system microbial community,the presence of putatively commensal microbes of theliver flukes is intriguing. The group included archae-ans; bacteria such as Methylobacterium that fix nitrogenin plant roots; microbes that participate in removal ofiron and other metals in waste water (Leptothrix); andspecies of the Rhodobacteraceae, which includesmethylotrophs from aquatic environments. The groupalso included Agrobacterium, a soil phytopathogen thatelicits tumors in host plants, is used in genetic engi-neering, has been shown to transform HeLa cells bytransfer of its tumor-inducing (Ti) plasmid, and haseven been implicated in the oncogenesis of Hodgkinlymphoma (52–55). We speculate that these microbes,including facultative pathogens, hitchhike on the meta-cercariae of O. viverrini during the establishment ofhelminth infection (56).

GI tract microbes are essential in the developmentand growth of (at least some) parasitic worms (57).Moreover, the gut microbiota is necessary for thedevelopment of a healthy immune response (58). In-deed, the lack of exposure to helminth infection out-side low- and middle-income countries appears linked

9CARCINOGENIC FLUKE PERTURBS GI TRACT MICROBIOME

to increasing rates of allergic and other autoimmunediseases (the hygiene hypothesis; refs. 13, 59, 60). Aswith other helminth infections, including hookwormsand schistosomes, infection with O. viverrini drives theimmune response toward a Th2 profile, characterizedby elevated cytokines IL-4, IL-5, and IL-13; total andparasite-specific immunoglobulin (Ig) G4 and IgE; andperipheral blood eosinophilia; and increases in theregulatory cytokine IL-10 and regulatory T cells (60).However, in contrast to the typical Th2 phenotypeestablished in chronic helminth infection, these regu-latory responses fail to modulate the robust local andsystemic inflammation induced by this trematode. Themarked inflammation of chronic opisthorchiasis mani-fests as periductal fibrosis, as determined by abdominalultrasonography (61) and in the elevated, systemic IL-6and cellular responses in vitro (10, 11, 62), which hasbeen linked with development of fibrosis and neoplasiain other infection–related contexts (63). Sripa et al.(10) speculate that this situation establishes a smolder-ing and polarized inflammatory milieu that promotesthe fibrosis and malignancy common to CCA (64).

Chronic inflammation is central to a range of malig-nancies (65). As a cause of inflammation, a dysregu-lated immune response to commensal bacteria in anintestinal bowel disorder is predictive of colon cancer(66). Inflammation due to Helicobacter pylori is a keyindicator for gastric cancers (66, 67). Alterations inmicrobiota have systemic effects on glucose tolerance,body weight, fat mass, oxidative stress, and macrophageinfiltration marker expression in visceral adipose tissue(68). Biliary inflammation has been linked with ulcer-ative colitis and hence O. viverrini-induced perturba-tions of microbial communities could contribute toinflammation by exposing the cholangiocytes to mi-crobes or metabolites of enteric or parasite-associatedmicrobes (7, 8) and recruitment of mucosal lympho-cytes to the biliary system (69). In addition, perturba-tions in the microbiota may be responses to bile con-stitution, since O. viverrini infection affects compositionand concentration of bile acids (70, 71). Bile acidcomposition influences gut microbiota (72) and canpromote pathobiont expansion and inflammation (73).Moreover, the dysbiosis associated with inflammatorybowel diseases leads to bile acid dysmetabolism that, inturn, exacerbates inflammation (74).

Whereas evolution of host–parasite relationships ofhelminths and vertebrates has programmed mammalscharacteristically to react immunologically to helminthparasites with a Th2-dominated response (75), infec-tion with O. viverrini is associated with an idiosyncratic,mixed Th1/Th2 phenotype with prominently elevatedproinflammatory cytokine IL-6 (10, 11, 62). Micro-biome perturbations may underlie or at least contributeto this distinctive phenotype and, more profoundly,underlie why O. viverrini causes cancer, a conspicuouslyunusual and unarguably unwelcome sequela of a hel-minth infection. Divergence in microbial communitiesmay arise through immunologic responses to the liverfluke, through the mediators that it releases (76), or

through the conveyance of microbes by the worm to thebiliary system from the outside world or from the mam-malian GI tract. Whatever the cause, however, themicrobial changes that ensue may contribute to the carcino-genicity of opisthorchiasis. Cholangiocytes, the epithe-lial cells that line bile ducts, and the cells from whichmalignancy develops as CCA, are dynamic cells thatrecognize and react to insult and injury (7, 8, 77). Theyexpress Toll-like receptors and can operate as profes-sional antigen-presenting cells. Cholangiocyte N-Rasmediates LPS-induced IL-6 and proliferation (77–79).We now conjecture that the insult may originate in thefluke-associated microbiota, in addition to (or evenrather than) the parasite. This possibility merits exam-ination, given, for example, the well-characterized in-teractions among the CagA virulence factor of H. pylori,epithelial cells, and gastric cancer (80). However, thisinvestigation may be challenging because of the com-munity diversity of this microbiome, from which arisesmultiple pathogen-activated molecular patterns, thepresence of noncultivable species, the absence of infor-mation from microbial ligand–cholangiocyte receptorcommunications, syntrophic archaeal–bacterial inter-actions, and other obstacles (32).

In summary, infection with O. viverrini led to changesin the microbial communities of the GI tract, includingthe emergence of microbes in the biliary system. Thesenovel findings in an instructive rodent model of opis-thorchiasis provide the impetus to investigate the influ-ence of infection with this pathogen on the humanmicrobiome, including within the biliary tree where theinflammatory and fibrotic responses originate duringopisthorchiasis, and the relationships among live flukeinfection, GI tract microbiota, and CCA. We predictthat superinfection involving microbial pathogens rou-tinely occurs during opisthorchiasis, and we now hy-pothesize that microbes and liver flukes together elicithost inflammatory responses. We also hypothesize thatthe microbes described in the biliary system originatefrom the external environment (i.e., are present withliver fluke metacercariae at initiation of parasite infec-tion, from the alimentary tract of the infected individ-ual, or both). If changes in microbiomes, similar tothose we have reported here, also occur in chronicallyinfected people, the dysbiosis of microbial communi-ties would represent another dimension of liver fluke–induced CCA (6, 8–10). Liver fluke–induced CCA isusually an intrahepatic cancer, and whereas there islimited evidence that cancer caused by O. viverriniinfection is different from other intrahepatic CCAs (1,6, 81, 82), closer scrutiny of this aspect would beinstructive, given the current findings that revealedchanges in the GI tract microbiota associated with O.viverrini infection. In parallel, these findings impeldeeper investigation of the microbiome in the contextof opisthorchiasis, including exploration of fungi andother eukaryotes by 18S rRNA gene–targeted ap-proaches, environmental shotgun sequencing of theentire metagenome (15, 43) and, moreover, investiga-

10 Vol. 27 November 2013 PLIESKATT ET AL.The FASEB Journal � www.fasebj.org

tion of the microbiome of metacercariae and otherdevelopmental stages of this carcinogenic parasite.

The authors thank Dr. Steven Zeichner for discussions thatled to this study and Drs. Victoria Mann and Gabriel Rinaldifor comments on the manuscript. SeqWright, Inc. (Houston,TX, USA) provided bioinformatics advice after performingthe pyrosequencing. This work was partially supported by U.S.National Cancer Institute (NCI) grants R01CA155297 toJ.M.B. and P.J.B and R01CA164719 to J.P.M. and P.J.B and byU.S. National Institute of Allergy and Infectious Disease(NIAID) grant P50AI098639 to B.S., J.M.B., and P.J.B. J.P.M.is the recipient of a fellowship from the National Health andMedical Research Council of Australia (NHMRC), and P.J.B.and J.M.B. were supported by the Dr. Cyrus and Myrtle KatzenCancer Research Center at George Washington University(Washington, DC, USA) The contents of the article are solelythe responsibility of the authors and do not necessarilyrepresent the official views of the NIAID, NCI, U.S. NationalInstitutes of Health, or NHMRC.

REFERENCES

1. De Martel, C., Ferlay, J., Franceschi, S., Vignat, J., Bray, F.,Forman, D., and Plummer, M. (2012) Global burden of cancersattributable to infections in 2008: a review and synthetic analysis.Lancet Oncol. 13, 607–615

2. Bouvard, V., Baan, R., Straif, K., Grosse, Y., Secretan, B., ElGhissassi, F., Benbrahim-Tallaa, L., Guha, N., Freeman, C.,Galichet, L., and Cogliano, V. (2009) A review of humancarcinogens, Part B: biological agents. Lancet Oncol. 10, 321–322

3. (2012) Biological agents: a review of human carcinogens. IARCMonogr. Eval. Carcinog. Risks Hum. 100, 1–441

4. Sithithaworn, P., Andrews, R. H., Nguyen, V. D., Wongsaroj, T.,Sinuon, M., Odermatt, P., Nawa, Y., Liang, S., Brindley, P. J., andSripa, B. (2012) The current status of opisthorchiasis andclonorchiasis in the Mekong Basin. Parasitol. Int. 61, 10–16

5. Sripa, B., Bethony, J. M., Sithithaworn, P., Kaewkes, S., Mairiang,E., Loukas, A., Mulvenna, J., Laha, T., Hotez, P. J., and Brindley,P. J. (2011) Opisthorchiasis and opisthorchis-associated cholan-giocarcinoma in Thailand and Laos. Acta Trop. 120(Suppl. 1),S158–S168

6. Blechacz, B., Komuta, M., Roskams, T., and Gores, G. J. (2011)Clinical diagnosis and staging of cholangiocarcinoma. Nat. Rev.Gastroenterol. Hepatol. 8, 512–522

7. Johnson, C., Han, Y., Hughart, N., McCarra, J., Alpini, G., andMeng, F. (2012) Interleukin-6 and its receptor, key players inhepatobiliary inflammation and cancer. Transl. Gastrointest. Can-cer 1, 58–70

8. O’Hara, S. P., Tabibian, J. H., Splinter, P. L., and Larusso, N. F.(2013) The dynamic biliary epithelia: molecules, pathways, anddisease. J. Hepatol. 58, 575–582

9. Razumilava, N., and Gores, G. J. (2013) Classification, diagnosis,and management of cholangiocarcinoma. Clin. Gastroenterol.Hepatol. 11, 13–21 e11; quiz e13–14

10. Sripa, B., Brindley, P. J., Mulvenna, J., Laha, T., Smout, M. J.,Mairiang, E., Bethony, J. M., and Loukas, A. (2012) Thetumorigenic liver fluke Opisthorchis viverrini: multiple pathwaysto cancer. Trends Parasitol. 28, 395–407

11. Sripa, B., Mairiang, E., Thinkhamrop, B., Laha, T., Kaewkes, S.,Sithithaworn, P., Tessana, S., Loukas, A., Brindley, P. J., andBethony, J. M. (2009) Advanced periductal fibrosis from infec-tion with the carcinogenic human liver fluke Opisthorchis viver-rini correlates with elevated levels of interleukin-6. Hepatology 50,1273–1281

12. Satarug, S., Haswell-Elkins, M. R., Sithithaworn, P., Bartsch, H.,Ohshima, H., Tsuda, M., Mairiang, P., Mairiang, E., Yongvanit,P., Esumi, H., and Elkins, D. B. (1998) Relationships betweenthe synthesis of N-nitrosodimethylamine and immune responsesto chronic infection with the carcinogenic parasite, Opisthorchisviverrini, in men. Carcinogenesis 19, 485–491

13. Weinstock, J. V. (2012) Autoimmunity: the worm returns. Nature491, 183–185

14. Broadhurst, M. J., Ardeshir, A., Kanwar, B., Mirpuri, J., Gundra,U. M., Leung, J. M., Wiens, K. E., Vujkovic-Cvijin, I., Kim, C. C.,Yarovinsky, F., Lerche, N. W., McCune, J. M., and Loke, P.(2012) Therapeutic helminth infection of macaques with idio-pathic chronic diarrhea alters the inflammatory signature andmucosal microbiota of the colon. PLoS Pathog. 8, e1003000

15. Schloissnig, S., Arumugam, M., Sunagawa, S., Mitreva, M., Tap,J., Zhu, A., Waller, A., Mende, D. R., Kultima, J. R., Martin, J.,Kota, K., Sunyaev, S. R., Weinstock, G. M., and Bork, P. (2013)Genomic variation landscape of the human gut microbiome.Nature 493, 45–50

16. Bhamarapravati, N., Thammavit, W., and Vajrasthira, S. (1978)Liver changes in hamsters infected with a liver fluke of man,Opisthorchis viverrini. Am. J. Trop. Med. Hyg. 27, 787–794

17. Laha, T., Pinlaor, P., Mulvenna, J., Sripa, B., Sripa, M., Smout,M. J., Gasser, R. B., Brindley, P. J., and Loukas, A. (2007) Genediscovery for the carcinogenic human liver fluke, Opisthorchisviverrini. BMC Genomics 8, 189

18. Csendes, A., Fernandez, M., and Uribe, P. (1975) Bacteriologyof the gallbladder bile in normal subjects. Am. J. Surg. 129,629–631

19. Valero, M. A., Navarro, M., Garcia-Bodelon, M. A., Marcilla, A.,Morales, M., Hernandez, J. L., Mengual, P., and Mas-Coma, S.(2006) High risk of bacterobilia in advanced experimentalchronic fasciolosis. Acta Trop. 100, 17–23

20. Nossa, C. W., Oberdorf, W. E., Yang, L., Aas, J. A., Paster, B. J.,Desantis, T. Z., Brodie, E. L., Malamud, D., Poles, M. A., and Pei,Z. (2010) Design of 16S rRNA gene primers for 454 pyrose-quencing of the human foregut microbiome. World J. Gastroen-terol. 16, 4135–4144

21. Angiuoli, S. V., Matalka, M., Gussman, A., Galens, K., Vangala,M., Riley, D. R., Arze, C., White, J. R., White, O., and Fricke,W. F. (2011) CloVR: a virtual machine for automated andportable sequence analysis from the desktop using cloud com-puting. BMC Bioinformatics 12, 356

22. White, J. R., Arze, C., Matalka, M, The CloVR Team, White, O.,Angiuoli, S. V., and Fricke, W. F. (2011) Phylogenetic microbialcommunity composition analysis based on 16S ribosomal RNAamplicon sequencing: standard operating procedure, ver-sion1.1. Nat. Preced. doi: 10.1038/npre.2011.5888.3

23. Caporaso, J. G., Kuczynski, J., Stombaugh, J., Bittinger, K.,Bushman, F. D., Costello, E. K., Fierer, N., Pena, A. G., Goo-drich, J. K., Gordon, J. I., Huttley, G. A., Kelley, S. T., Knights,D., Koenig, J. E., Ley, R. E., Lozupone, C. A., McDonald, D.,Muegge, B. D., Pirrung, M., Reeder, J., Sevinsky, J. R., Turn-baugh, P. J., Walters, W. A., Widmann, J., Yatsunenko, T.,Zaneveld, J., and Knight, R. (2010) QIIME allows analysis ofhigh-throughput community sequencing data. Nat Methods 7,335–336

24. Edgar, R. C., Haas, B. J., Clemente, J. C., Quince, C., and Knight,R. (2011) UCHIME improves sensitivity and speed of chimeradetection. Bioinformatics. 27, 2194–2200

25. Wang, Q., Garrity, G. M., Tiedje, J. M., and Cole, J. R. (2007)Naive Bayesian classifier for rapid assignment of rRNA se-quences into the new bacterial taxonomy. Appl. Environ. Micro-biol. 73, 5261–5267

26. Hamady, M., and Knight, R. (2009) Microbial communityprofiling for human microbiome projects: tools, techniques,and challenges. Genome. Res. 19, 1141–1152

27. Schloss, P. D., Westcott, S. L., Ryabin, T., Hall, J. R., Hartmann,M., Hollister, E. B., Lesniewski, R. A., Oakley, B. B., Parks, D. H.,Robinson, C. J., Sahl, J. W., Stres, B., Thallinger, G. G., VanHorn, D. J., and Weber, C. F. (2009) Introducing mothur:open-source, platform-independent, community-supported soft-ware for describing and comparing microbial communities.Appl. Environ. Microbiol. 75, 7537–7541

28. Kimura, M. (1980) A simple method for estimating evolutionaryrate of base substitutions through comparative studies of nucle-otide sequences. J. Mol. Evol. 16, 111–120

29. Tamura, K., Peterson, D., Peterson, N., Stecher, G., Nei, M., andKumar, S. (2011) MEGA5: molecular evolutionary geneticsanalysis using maximum likelihood, evolutionary distance, andmaximum parsimony methods. Mol. Biol. Evol. 28, 2731–2739

30. Cox, M. M., and Battista, J. R. (2005) Deinococcus radiodurans: theconsummate survivor. Nat. Rev. Microbiol. 3, 882–892

11CARCINOGENIC FLUKE PERTURBS GI TRACT MICROBIOME

31. Jensen, S., Duperron, S., Birkeland, N. K., and Hovland, M.(2010) Intracellular Oceanospirillales bacteria inhabit gills ofAcesta bivalves. FEMS Microbiol. Ecol. 74, 523–533

32. Horz, H. P., and Conrads, G. (2011) Methanogenic Archaea andoral infections: ways to unravel the black box. J. Oral Microbiol.doi: 10.3402/jom.v3i0.5940

33. Lange, M., Westermann, P., and Ahring, B. K. (2005) Archaea inprotozoa and metazoa. Appl. Microbiol. Biotechnol. 66, 465–474

34. Eckburg, P. B., Lepp, P. W., and Relman, D. A. (2003) Archaeaand their potential role in human disease. Infect. Immun. 71,591–596

35. Virgin, H. W., and Todd, J. A. (2011) Metagenomics andpersonalized medicine. Cell 147, 44–56

36. Walk, S. T., Blum, A. M., Ewing, S. A., Weinstock, J. V., andYoung, V. B. (2010) Alteration of the murine gut microbiotaduring infection with the parasitic helminth Heligmosomoidespolygyrus. Inflamm. Bowel Dis. 16, 1841–1849

37. Lupp, C., Robertson, M. L., Wickham, M. E., Sekirov, I.,Champion, O. L., Gaynor, E. C., and Finlay, B. B. (2007)Host-mediated inflammation disrupts the intestinal microbiotaand promotes the overgrowth of Enterobacteriaceae. Cell HostMicrobe 2, 204

38. Haswell-Elkins, M. R., Mairiang, E., Mairiang, P., Chaiyakum, J.,Chamadol, N., Loapaiboon, V., Sithithaworn, P., and Elkins,D. B. (1994) Cross-sectional study of Opisthorchis viverrini infec-tion and cholangiocarcinoma in communities within a high-riskarea in northeast Thailand. Int. J. Cancer 59, 505–509

39. Lvova, M. N., Tangkawattana, S., Balthaisong, S., Katokhin,A. V., Mordvinov, V. A., and Sripa, B. (2012) Comparativehistopathology of Opisthorchis felineus and Opisthorchis viverrini ina hamster model: an implication of high pathogenicity of theEuropean liver fluke. Parasitol. Int. 61, 167–172

40. Eckburg, P. B., Bik, E. M., Bernstein, C. N., Purdom, E.,Dethlefsen, L., Sargent, M., Gill, S. R., Nelson, K. E., andRelman, D. A. (2005) Diversity of the human intestinal micro-bial flora. Science 308, 1635–1638

41. Frank, D. N., St. Amand, A. L., Feldman, R. A., Boedeker, E. C.,Harpaz, N., and Pace, N. R. (2007) Molecular-phylogeneticcharacterization of microbial community imbalances in humaninflammatory bowel diseases. Proc. Natl. Acad. Sci. U. S. A. 104,13780–13785

42. Winter, S. E., Winter, M. G., Xavier, M. N., Thiennimitr, P.,Poon, V., Keestra, A. M., Laughlin, R. C., Gomez, G., Wu, J.,Lawhon, S. D., Popova, I. E., Parikh, S. J., Adams, L. G., Tsolis,R. M., Stewart, V. J., and Baumler, A. J. (2013) Host-derivednitrate boosts growth of E. coli in the inflamed gut. Science 339,708–711

43. Maurice, C. F., Haiser, H. J., and Turnbaugh, P. J. (2013)Xenobiotics shape the physiology and gene expression of theactive human gut microbiome. Cell 152, 39–50

44. Peterfreund, G. L., Vandivier, L. E., Sinha, R., Marozsan, A. J.,Olson, W. C., Zhu, J., and Bushman, F. D. (2012) Succession inthe gut microbiome following antibiotic and antibody therapiesfor Clostridium difficile. PLoS One 7, e46966

45. Carpenter, H. A. (1998) Bacterial and parasitic cholangitis.Mayo Clin. Proc. 73, 473–478

46. Tajiri, T., Yoshida, H., Mamada, Y., Taniai, N., Yokomuro, S.,and Mizuguchi, Y. (2008) Diagnosis and initial management ofcholangiocarcinoma with obstructive jaundice. World J. Gastro-enterol. 14, 3000–3005

47. Boonyanugomol, W., Chomvarin, C., Sripa, B., Bhudhisawasdi,V., Khuntikeo, N., Hahnvajanawong, C., and Chamsuwan, A.(2012) Helicobacter pylori in Thai patients with cholangiocar-cinoma and its association with biliary inflammation and prolif-eration. HPB (Oxford). 14, 177–184

48. Mahlen, S. D. (2011) Serratia infections: from military experi-ments to current practice. Clin. Microbiol. Rev. 24, 755–791

49. Blekher, L., Siegman-Igra, Y., Schwartz, D., Berger, S. A., andCarmeli, Y. (2000) Clinical significance and antibiotic resistancepatterns of Leminorella spp., an emerging nosocomial patho-gen. J. Clin. Microbiol. 38, 3036–3038

50. Ryan, M. P., and Adley, C. C. (2010) Sphingomonas paucimobilis:a persistent Gram-negative nosocomial infectious organism. J.Hosp. Infect. 75, 153–157

51. Que, Y. A., and Moreillon, P. (2011) Infective endocarditis. Nat.Rev. Cardiol. 8, 322–336

52. Emerson, D., Fleming, E. J., and McBeth, J. M. (2010) Iron-oxidizing bacteria: an environmental and genomic perspective.Annu. Rev. Microbiol. 64, 561–583

53. Dourado, M. N., Andreote, F. D., Dini-Andreote, F., Conti, R.,Araujo, J. M., and Araujo, W. L. (2012) Analysis of 16S rRNAand mxaF genes revealing insights into Methylobacterium niche-specific plant association. Genet. Mol. Biol. 35, 142–148

54. Kunik, T., Tzfira, T., Kapulnik, Y., Gafni, Y., Dingwall, C., andCitovsky, V. (2001) Genetic transformation of HeLa cells byAgrobacterium. Proc. Natl. Acad. Sci. U. S. A. 98, 1871–1876

55. Sauter, C. (1995) Is Hodgkin’s disease a human counterpart ofbacterially induced crown-gall tumours? Lancet 346, 1433

56. Gendrel, D., Kombila, M., Beaudoin-Leblevec, G., and Richard-Lenoble, D. (1994) Nontyphoidal salmonellal septicemia inGabonese children infected with Schistosoma intercalatum. Clin.Infect. Dis. 18, 103–105

57. Hayes, K. S., Bancroft, A. J., Goldrick, M., Portsmouth, C.,Roberts, I. S., and Grencis, R. K. (2010) Exploitation of theintestinal microflora by the parasitic nematode Trichuris muris.Science 328, 1391–1394

58. Tlaskalova-Hogenova, H., Stepankova, R., Kozakova, H., Hudco-vic, T., Vannucci, L., Tuckova, L., Rossmann, P., Hrncir, T.,Kverka, M., Zakostelska, Z., Klimesova, K., Pribylova, J., Bartova,J., Sanchez, D., Fundova, P., Borovska, D., Srutkova, D., Zidek,Z., Schwarzer, M., Drastich, P., and Funda, D. P. (2011) The roleof gut microbiota (commensal bacteria) and the mucosal bar-rier in the pathogenesis of inflammatory and autoimmunediseases and cancer: contribution of germ-free and gnotobioticanimal models of human diseases. Cell. Mol. Immunol. 8, 110–120

59. Rook, G. A. (2012) Hygiene hypothesis and autoimmune dis-eases. Clin. Rev. Allergy Immunol. 42, 5–15

60. McSorley, H. J., and Maizels, R. M. (2012) Helminth infectionsand host immune regulation. Clin. Microbiol. Rev. 25, 585–608

61. Mairiang, E., Laha, T., Bethony, J. M., Thinkhamrop, B.,Kaewkes, S., Sithithaworn, P., Tesana, S., Loukas, A., Brindley,P. J., and Sripa, B. (2012) Ultrasonography assessment ofhepatobiliary abnormalities in 3359 subjects with Opisthorchisviverrini infection in endemic areas of Thailand. Parasitol. Int.61, 208–211

62. Sripa, B., Thinkhamrop, B., Mairiang, E., Laha, T., Kaewkes, S.,Sithithaworn, P., Periago, M. V., Bhudhisawasdi, V., Yonglitthip-agon, P., Mulvenna, J., Brindley, P. J., Loukas, A., and Bethony,J. M. (2012) Elevated plasma IL-6 associates with increased riskof advanced fibrosis and cholangiocarcinoma in individualsinfected by Opisthorchis viverrini. PLoS Negl. Trop. Dis. 6, e1654

63. Stauffer, J. K., Scarzello, A. J., Jiang, Q., and Wiltrout, R. H.(2012) Chronic inflammation, immune escape, and oncogene-sis in the liver: a unique neighborhood for novel intersections.Hepatology 56, 1567–1574

64. Balkwill, F., Charles, K. A., and Mantovani, A. (2005) Smolder-ing and polarized inflammation in the initiation and promotionof malignant disease. Cancer Cell 7, 211–217

65. Fukata, M., and Abreu, M. T. (2008) Role of Toll-like receptorsin gastrointestinal malignancies. Oncogene 27, 234–243

66. Peek, R. M., Jr., and Blaser, M. J. (2002) Helicobacter pylori andgastrointestinal tract adenocarcinomas. Nat. Rev. Cancer 2, 28–37

67. Houghton, J., and Wang, T. C. (2005) Helicobacter pylori andgastric cancer: a new paradigm for inflammation-associatedepithelial cancers. Gastroenterology. 128, 1567–1578

68. Cani, P. D., Bibiloni, R., Knauf, C., Waget, A., Neyrinck, A. M.,Delzenne, N. M., and Burcelin, R. (2008) Changes in gutmicrobiota control metabolic endotoxemia-induced inflamma-tion in high-fat diet-induced obesity and diabetes in mice.Diabetes 57, 1470–1481

69. Adams, D. H., and Eksteen, B. (2006) Aberrant homing ofmucosal T cells and extra-intestinal manifestations of inflamma-tory bowel disease. Nat. Rev. Immunol. 6, 244–251

70. Changbumrung, S., Tungtrongchitr, R., Migasena, P., andChamroenngan, S. (1990) Serum unconjugated primary andsecondary bile acids in patients with cholangiocarcinoma andhepatocellular carcinoma. J. Med. Assoc. Thai. 73, 81–90

71. Wonkchalee, O., Boonmars, T., Kaewkes, S., Chamgramol, Y.,Aromdee, C., Wu, Z., Juasook, A., Sudsarn, P., Boonjaraspinyo, S.,and Pairojkul, C. (2012) Comparative studies on animal models forOpisthorchis viverrini infection: host interaction through susceptibilityand pathology. Parasitol. Res. 110, 1213–1223

12 Vol. 27 November 2013 PLIESKATT ET AL.The FASEB Journal � www.fasebj.org

72. Islam, K. B., Fukiya, S., Hagio, M., Fujii, N., Ishizuka, S., Ooka,T., Ogura, Y., Hayashi, T., and Yokota, A. (2011) Bile acid is ahost factor that regulates the composition of the cecal microbi-ota in rats. Gastroenterology 141, 1773–1781

73. Devkota, S., Wang, Y., Musch, M. W., Leone, V., Fehlner-Peach,H., Nadimpalli, A., Antonopoulos, D. A., Jabri, B., and Chang,E. B. (2012) Dietary-fat-induced taurocholic acid promotespathobiont expansion and colitis in Il10�/� mice. Nature 487,104–108

74. Duboc, H., Rajca, S., Rainteau, D., Benarous, D., Maubert, M. A.,Quervain, E., Thomas, G., Barbu, V., Humbert, L., Despras, G.,Bridonneau, C., Dumetz, F., Grill, J. P., Masliah, J., Beaugerie,L., Cosnes, J., Chazouilleres, O., Poupon, R., Wolf, C., Mallet,J. M., Langella, P., Trugnan, G., Sokol, H., and Seksik, P. (2013)Connecting dysbiosis, bile-acid dysmetabolism and gut inflam-mation in inflammatory bowel diseases. Gut 62, 531–539

75. Allen, J. E., and Wynn, T. A. (2011) Evolution of Th2 immunity:a rapid repair response to tissue destructive pathogens. PLoSPathog. 7, e1002003

76. Robinson, M. W., Donnelly, S., Hutchinson, A. T., To, J., Taylor,N. L., Norton, R. S., Perugini, M. A., and Dalton, J. P. (2011) Afamily of helminth molecules that modulate innate cell re-sponses via molecular mimicry of host antimicrobial peptides.PLoS Pathog. 7, e1002042

77. Syal, G., Fausther, M., and Dranoff, J. A. (2012) Advances incholangiocyte immunobiology. Am. J. Physiol. Gastrointest. LiverPhysiol. 303, G1077–G1086

78. O’Hara, S. P., Splinter, P. L., Trussoni, C. E., Gajdos, G. B.,Lineswala, P. N., and LaRusso, N. F. (2011) CholangiocyteN-Ras protein mediates lipopolysaccharide-induced interleu-kin 6 secretion and proliferation. J. Biol. Chem. 286, 30352–30360

79. Chen, X. M., O’Hara, S. P., Nelson, J. B., Splinter, P. L.,Small, A. J., Tietz, P. S., Limper, A. H., and LaRusso, N. F.(2005) Multiple TLRs are expressed in human cholangiocytesand mediate host epithelial defense responses to Cryptospo-ridium parvum via activation of NF-kappaB. J. Immunol. 175,7447–7456

80. Wroblewski, L. E., Peek, R. M., Jr., and Wilson, K. T. (2010)Helicobacter pylori and gastric cancer: factors that modulatedisease risk. Clin. Microbiol. Rev. 23, 713–739

81. Jinawath, N., Chamgramol, Y., Furukawa, Y., Obama, K.,Tsunoda, T., Sripa, B., Pairojkul, C., and Nakamura, Y. (2006)Comparison of gene expression profiles between Opisthorchisviverrini and non-Opisthorchis viverrini associated humanintrahepatic cholangiocarcinoma. Hepatology 44, 1025–1038

82. De Martel, C., Plummer, M., and Franceschi, S. (2010) Cholan-giocarcinoma: descriptive epidemiology and risk factors. Gastro-enterol. Clin. Biol. 34, 173–180

Received for publication April 7, 2013.Accepted for publication July 22, 2013.

13CARCINOGENIC FLUKE PERTURBS GI TRACT MICROBIOME

Table S1. Operational taxonomy units (OTUs) associated with Opisthorchis viverrini worms in the biliary tree of hamsters. ♣ Left column: OTUs where numbers were much higher the biliary system (in worms and bile) in comparison to colorectal feces of hamsters (infected and control). ♠ Central and right columns: OTUs detected by pyrosequencing using broad range 16S rRNA gene primers in the biliary system - in O. viverrini worms and bile - but not in the colorectal feces of the hamsters (infected or control).

Actinomycetales_Propionibacterineae♣ Archaea_Other♣ Bacillales_Other♣ Bacilli_Other♣ Bacteria_Other♣ Burkholderiaceae_Cupriavidus♣ Clostridiaceae_"Clostridiaceae 1"♣ Comamonadaceae_Comamonas♣ Comamonadaceae_Hydrogenophaga♣ Comamonadaceae_Other♣ Enterobacteriaceae_Citrobacter♣ Enterobacteriaceae_Other♣ Gammaproteobacteria_Other♣ Gemmatimonadaceae_Gemmatimonas♣ Incertae sedis 5_Other♣ Incertae sedis 5_Tepidimonas♣ Lactobacillaceae_Lactobacillus♣ Lactobacillales_Other♣ Moraxellaceae_Acinetobacter♣ Moraxellaceae_Alkanindiges♣ Planococcaceae_Jeotgalibacillus♣ Planococcaceae_Other♣ Pseudomonadaceae_Chryseomonas♣ Pseudomonadaceae_Flavimonas♣ Pseudomonadaceae_Pseudomonas♣ Root_Other♣ Sphingomonadaceae_Sphingomonas♣

Acetobacteraceae_Muricoccus♠ Acetobacteraceae_Roseomonas♠ Aeromonadales_Other♠ Alcaligenaceae_Bordetella♠ Alcaligenaceae_Pusillimonas♠ Anaplasmataceae_Neorickettsia♠ Bacillaceae_"Bacillaceae 1"♠ Bacillariophyta_Other♠ Bradyrhizobiaceae_Bradyrhizobium♠ Burkholderiaceae_Burkholderia♠ Burkholderiaceae_Other♠ Burkholderiaceae_Ralstonia♠ Caulobacteraceae_Brevundimonas♠ Caulobacteraceae_Other♠ Caulobacteraceae_Phenylobacterium♠ Chromatiaceae_Rheinheimera♠ Comamonadaceae_Caldimonas♠ Comamonadaceae_Diaphorobacter♠ Comamonadaceae_Variovorax♠ Coxiellaceae_Aquicella♠ Crenotrichaceae_Terrimonas♠ Deinococcaceae_Deinococcus♠ Enterobacteriaceae_Leminorella♠ Enterobacteriaceae_Rahnella♠ Enterobacteriaceae_Raoultella♠ Enterobacteriaceae_Salmonella♠ Enterobacteriaceae_Serratia♠ Euryarchaeota_Other♠ Flavobacteriaceae_Cloacibacterium♠ Flavobacteriaceae_Flavobacterium♠ Flexibacteraceae_Niastella♠

GpIIa_Other♠ Halomonadaceae_Halomonas♠ Hyphomicrobiaceae_Devosia♠ Incertae Sedis XII_Exiguobacterium♠ Incertae Sedis XI_Anaerococcus♠ Incertae Sedis XVII_Thermaerobacter♠ Incertae sedis 5_Leptothrix♠ Listeriaceae_Brochothrix♠ Methylobacteriaceae_Methylobacterium♠ Moraxellaceae_Other♠ Moraxellaceae_Psychrobacter♠ Oceanospirillaceae_Other♠ Pasteurellaceae_Lonepinella♠ Phyllobacteriaceae_Mesorhizobium♠ Phyllobacteriaceae_Other♠ Planococcaceae_Planomicrobium♠ Porphyromonadaceae_Porphyromonas♠ Pseudomonadaceae_Cellvibrio♠ Pseudomonadales_Other♠ Rhizobiaceae_Agrobacterium♠ Rhodobacteraceae_Other♠ Rhodobacteraceae_Paracoccus♠ Rhodobacteraceae_Rubellimicrobium♠ Rubrobacterales_Rubrobacterineae♠ Sphingomonadaceae_Novosphingobium♠ Sphingomonadaceae_Sphingobium♠ Trueperaceae_Truepera♠ Xanthomonadaceae_Other♠ Xanthomonadaceae_Pseudoxanthomonas♠ Xanthomonadaceae_Stenotrophomonas♠ Xanthomonadaceae_Thermomonas♠

Figure S1. Concentration, purity and integrity of genomic DNAs from prokaryotes from the colorectal contents and bile of Opisthorchis viverrini infected hamsters and from O. viverrini worms. Genomic DNA concentrations, purity and yield were determined, along with ratios for optical density at 260/ 280 nm and 260/30 nm (top panel). Bottom panel shows the gel-like images for the gDNA samples (Bioanalyzer 2100), with the lanes identified using the ‘Sample ID’ from the top panel; size standards are shown at the margins.

Sample ID treatment ng/µl 260/280 yield, µg Ham-F-u1 feces, control hamster 56 1.80 5.6 Ham-F-u2 feces, control hamster 57 1.81 5.7 Ham-F-u3 feces, control hamster 37 1.80 3.7 Ham-F-u4 feces, control hamster 79 1.81 7.9 Ham-F-i1 feces, OV infected hamster 997 2.04 99.7 Ham-F-i2 feces, OV infected hamster 931 2.05 93.1 Ham-F-i3 feces, OV infected hamster 346 2.01 34.6 Ham-F-i4 feces, OV infected hamster 210 1.85 21.0

Ham-OV-G1 OV worms 52 1.84 5.2 Ham-OV-G2 OV worms 72 1.87 7.2 Ham-OV-S1 OV worms 28 1.83 2.8 Ham-B-iPool bile, OV infected hamsters 4.5 1.84 1.1 Bile, control bile, control hamsters 0.01 0.02 0

Figure S2. OTU-specific PCRs locate the Novosphingobium phylotype in liver fluke/biliary system associated microbiota but not in colon contents. Nested PCRs detected low(er) levels of Facklamia (panel A), Flavobacterium (C), Muricoccus (D) and Variovorax (E) OTUs in infected hamster colon contents. By contrast, the Novosphingobium (panel B) OTU was not detected in colon contents, providing additional support to the hypothesis that microbes were Introduced Into the hamster biliary tree by the Opisthorchis viverrini parasites.

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 10

Facklamia Novosphingobium

202 nt 388 nt

Flavobacterium Muricoccus

238 nt 270 nt

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 10

A B

D C

Molecular standards (lane 1, 1 kb ladder), OV-G1 worms (lane 2) , OV-G2 (3), infected hamster bile (4), control hamster colon (lanes 5, 6, 7), infected hamster colon (8,9, 10). Amplicon sizes indicated in each panel by arrow for the respective OTUs. Arrows indicate expected sizes of the five, targeted OTU-specific amplicons.

1 2 3 4 5 6 7 8 9 10

Variovorax

325 nt

E