identification of a campylobacter coli methyltransferase targeting

FEMS Microbiology Letters, 364, 2017, fnw268

doi: 10.1093/femsle/fnw268Advance Access Publication Date: 2 December 2016Research Letter

RESEARCH LETTER –Pathogens & Pathogenicity

Identification of a Campylobacter coli methyltransferasetargeting adenines at GATC sitesVikrant Dutta1, Eric Altermann2,3, Maria D. Crespo1, Jonathan W. Olson4,Robin M. Siletzky1 and Sophia Kathariou1,∗

1Department of Food, Bioprocessing and Nutrition Sciences, North Carolina State University, Raleigh, NC27695, USA, 2Rumen Microbiology Team, Animal Nutrition and Health Group, AgResearch Grasslands,Palmerston North 4442, New Zealand, 3Riddet Institute, Massey University, Palmerston North 4442, NewZealand and 4Department of Biological Sciences, North Carolina State University, Raleigh, NC 27695, USA∗Corresponding author: Department of Food, Bioprocessing and Nutrition Sciences, North Carolina State University, Raleigh, NC 27695, USA.Tel: 919 513-2075; E-mail: [email protected] sentence summary: We identified a gene encoding a N6-adenine DNA methyltransferase in Campylobacter coli rendering DNA resistant to digestionby MboI, and described a predilection of this gene for C. coli from swine.Editor: Craig Winstanley

ABSTRACT

Campylobacter coli can infect humans and colonize multiple other animals, but its host-associated genes or adaptations arepoorly understood. Adenine methylation at GATC sites, resulting in MboI resistance of genomic DNA, was earlier frequentlydetected among C. coli from swine but not among turkey-derived isolates. The underlying genetic basis has remainedunknown. Comparative genome sequence analyses of C. coli 6461, a swine-derived strain with MboI-resistant DNA, revealedtwo chromosomal ORFs, 0059 and 0060, encoding a putative DNA methyltransferase and a conserved hypothetical protein,respectively, which were lacking from the genome of the turkey-derived C. coli strain 11601, which had MboI-susceptibleDNA. To determine whether ORF0059 mediated MboI resistance and hence encoded a putative N6-adenine DNAmethyltransferase, the gene was cloned immediately upstream of a chloramphenicol resistance cassette (cat) and a PCRfragment harboring ORF0059-cat was transformed into C. coli 11601. The transformants had MboI-resistant DNA, suggestinga direct role of this gene in methylation of adenines at GATC sites. In silico analyses suggested that the ORF0059-ORF0060cassette was more frequent among C. coli from swine than certain other sources (e.g. cattle, humans). Potential impacts ofORF0059-mediated methylation on C. coli host preference and other adaptations remain to be elucidated.

Keywords: Campylobacter; C. coli; DNA methylation; MboI resistance; GATC; methyltransferase

INTRODUCTION

Campylobacter jejuni and C. coli are leading agents of human food-borne disease and frequently colonize the intestine of avian andmammalian hosts (Huang et al. 2015; Fitzgerald 2015; Sheppardand Maiden 2015). Even though certain lineages are generalists,able to colonize diverse animal hosts, others exhibit preferencefor certain hosts or environments (Ogden et al. 2009; Sheppard

et al. 2011, 2014; Griekspoor et al. 2013). However, host-associatedgenetic attributes and adaptations of Campylobacter spp. remainpoorly understood.

Previously, a putative host-associated DNAmethylation phe-notype was noted in C. coli. Resistance of genomic DNA to di-gestion by the restriction enzyme MboI was encountered rel-atively frequently (43%) among swine-derived C. coli strains,but not among C. coli from turkeys (Wright et al. 2010). MboI

Received: 21 September 2016; Accepted: 17 November 2016C© FEMS 2016. All rights reserved. For permissions, please e-mail: [email protected]

1

Downloaded from https://academic.oup.com/femsle/article-abstract/364/7/fnw268/2631356by gueston 06 April 2018

http://www.oxfordjournals.org

mailto:[email protected]

mailto:[email protected]

2 FEMS Microbiology Letters, 2017, Vol. 364, No. 7

resistance is mediated by the methylation of adenines at GATCsites via DNA methyltransferases that are frequently encoun-tered as part of restriction-modification (RM) systems (McClel-land, Nelson and Raschke 1994; Palmer and Marinus 1994;Roberts et al. 2007). Methylation is crucial for a number of regu-latory processes, which could in turn influence a range of adap-tive phenotypes (Reisenauer et al. 1999; Casadesus and Low 2006;Marinus and Casadesus 2009).

In swine-derived C. coli, MboI resistance was noted amongstrains of various genotypes and could be present or absent instrains of the same sequence type as defined by multilocus se-quence typing (Wright et al. 2010). Such findings suggest hor-izontal gene transfer-mediated acquisition of a DNA methyl-transferase or RM cassette, or, alternatively, gene decay or lossof the relevant determinant(s) by certain strains. Further studieson this methylation were hampered by the fact that the corre-sponding determinant in MboI-resistant C. coli from swine wasnot identified. In this study, we investigated the role of a puta-tive DNAmethyltransferase in theMboI resistance phenotype ofswine-derived strain C. coli 6461 and analyzed in silico the distri-bution of thismethyltransferase gene in Campylobacter spp. fromvarious hosts and sources.

MATERIALS AND METHODS

Bacterial strains and growth conditions

Campylobacter coli 6461 (MboI-resistant DNA), C. coli 11601 (MboI-susceptible DNA) and the C. coli 11601 transformants VD13B andVD13D (both with MboI-resistant DNA) were routinely grownmicroaerobically at 42◦C in Mueller-Hinton Broth (MHB) or onMueller-Hinton agar (MHA; MHB with 1.2% agar) (Becton, Dick-inson and Co., Sparks, MD, USA) as described (Wright et al. 2010).Campylobacter coli 6461 was derived from swine feces in 2004 andwas resistant to tetracycline, streptomycin and erythromycin,while C. coli 11601 was isolated in 2006 from the cecum ofa turkey and was resistant to tetracycline, streptomycin, ery-thromycin, kanamycin, nalidixic acid and ciprofloxacin (Duttaet al. 2016). Both strains were susceptible to chloramphenicol.Escherichia coli strain DH5α (Invitrogen, Carlsbad, CA, USA) wasroutinely grown on Luria-Bertani broth (LB) or LB agar (1.2%agar, Becton, Dickinson and Co.) for 20–24 h at 37◦C. Antibioticswere used as indicated and included kanamycin, nalidixic acidand ciprofloxacin (Fisher Biotech, Fair Lawn, NJ, USA), as well aserythromycin, tetracycline, chloramphenicol and streptomycin(Sigma-Aldrich, St. Louis, MO, USA).

Recombinant plasmid constructs and transformations

Primers used to construct the recombinant plasmids are listed inTable 1, and their locations are shown in Fig. 1. PCR employed theTakara Ex Taq kit (Takara, Madison, WI, USA) and a T1 thermalcycler (Biometra, Goettingen, Germany). Campylobacter coli ge-nomic DNA was extracted using the DNeasy kit (Qiagen, Valen-cia, CA, USA) and used as template for (1) primers P1/P2, to gen-erate a 1428 bp PCR fragment containing ORF0059 (putative DNAmethyltransferase, 852 bp) and 570 bp upstream and 6 bp down-stream of ORF0059 and (2) P3/P4, to amplify a 611 bp fragmentcontainingORF0061 (conserved hypothetical protein; 423 bp) and188 bp downstream (Fig. 1). PCR products obtained with P1/P2and P3/P4 were digested with BamHI/XmaI and XmaI/EcoRI, re-spectively, gel-purified (Qiagen) and ligated using T4 DNA ligase(Promega, Madison) to each other and to BamHI/EcoRI digestedpUC19 (Yanisch-Perron, Vieira and Messing 1985). The ligation

Table 1. Oligonucleotides used in this study.

P1: GACTGGATCCGCCTTCAAAATATATCAGCGP2: GACTCCCGGGCCGTAATCATAAATAACCTCCP3: GACTCCCGGGATGATTTATCTAGCGCAAACAP4: GACTGAATTCCCACTAGATGCTATCAATGCC1: CTCGGCGGTGTTCCTTTCCAAGC2: CGCCCTTTAGTTCCTAAAGGGTS1: GGCGGAAGAGCTATGCGTGS2: TCACTTAAGCCACTCCTGCAGGCS3: GCTAGAAGTGTCGCAGATGS4: AGTCTTGAAATCGATCCGAG

Underlined sequences correspond to restriction enzyme sites: GGATCC, BamHI;CCCGGG, XmaI; GAATCC, EcoRI.

mixture was electroporated into E. coli DH5α and transformantswere selected on LB agar supplemented with X-gal and ampi-cillin (100 μg ml−1). The resulting recombinant plasmid, p6461-Met (4725 bp), was linearized with XmaI, gel-purified and ligatedto a similarly digested and gel-purified chloramphenicol cas-sette (cat, 842 bp) obtained byXmaI digestion of pJMA-001 (kindlyprovided by Jason Andrus), which harbors cat from pRY111 (Yaoet al. 1993). The resulting plasmid (pMet; 5567 bp) was electropo-rated into E. coli DH5α. Transformants were selected on LB agarwith chloramphenicol (20 μg ml−1) and confirmed using gene-specific primers P1–P4 (Fig. 1) and C1—C2 (cat-specific primers)(Table 1). Using pMET DNA as template, P1/P4 were used to am-plify a 2881 bp amplicon (Met PCR) containing the P1–P2 product(1428 bp), cat (842 bp) and the P3–P4 product (611 bp). Gel-purifiedMet PCR was used for transformation experiments.

For transformations, a loopful of cells from a 24 h cultureof C. coli 11601 (prospective recipient) was suspended in 50 μlMHB along with ca. 1 μg Met PCR. As control, cells from eachstrainwere similarly suspended in 50 μl MHBwithout any addedDNA. Cell suspensions (with and without Met PCR) were spot-ted (25 μl) in duplicate on MHA, spots were dried in a laminarflow biosafety cabinet and the plates incubatedmicroaerobicallyovernight at 42◦C. The confluent bacterial growth at the inocu-lated spot was spread-plated on MHA supplemented with chlo-ramphenicol (20 μg ml−1; MHA-C) and plates were incubated for48 h at 42◦C microaerobically. Putative transformants were pu-rified on MHA-C and confirmed using primers S1–S2 and S3–S4(Table 1, Fig. 1).

Characterization of transformants

MboI and Sau3AI resistance of genomic DNA from transfor-mants was determined as described (Wright et al. 2010). MICsto a panel of antibiotics (tetracycline, streptomycin, kanamycin,erythromycin, nalidixic acid, ciprofloxacin, chloramphenicol)and fla typingwere performed as described (Gu et al. 2009;Wrightet al. 2010).

Distribution of ORFs 0059 and 0060 in publicallyavailable genomes of Campylobacter coli from differenthosts

Campylobacter coli genomes and assemblies were retrieved fromNCBI (http://www.ncbi.nlm.nih.gov/genome/genomes/1145) ina tab-delimited text file and imported directly into MicrosoftExcel. The respective FTP link for each organism was thensaved into a separate text file. A custom Perl script was devel-oped to download associated Genbank files (genomic.gbff) and


http://www.ncbi.nlm.nih.gov/genome/genomes/1145

Dutta et al. 3

Figure 1. Genomic organization of the methyltransferase region in C. coli 6461 (swine isolate, MboI-resistant DNA), C. coli 11601 (turkey isolate, MboI-susceptibleDNA) and a transformant of C. coli 11601 with MboI-resistant DNA. Large arrows indicate direction of transcription of the respective ORFs, with ORF0059 (putativemethyltransferase) in black. The chloramphenicol resistance gene cat in the transformant is also in black. Small arrows indicate primers (Table 1), whereas dashedarrows indicate the expected amplicon size in bp. Grey shaded areas indicate homologous regions with percent nucleotide identity and % GC indicated. Genome

organization and content in C. coli 11601 transformants (VD13B and VD13D) were confirmed with PCR using S1/S2 and S3/S4 as described in the section ‘Materials andMethods’.

Genbank master headers (wgsmaster.gbff) as compressedarchives from the NCBI FTP server. Downloaded archives werethen extracted and saved into a new directory. Subsequently,a second custom Perl script investigated each respective Gen-bank header and ‘source’ feature for metadata such as strainand host/source. Available information on each C. coli strain wassaved as a tab-delimited text file and then directly imported intoMicrosoft Excel for further analysis.

Statistical analysis

The prevalence of ORF0059 in C. coli genomes from differenthosts/sources was analyzed by one-way analysis of variancewith a Tukey’s test at P < 0.05, using SPSS version 22 (IBM Cor-poration Software Group, Somers, NY, USA).

RESULTS AND DISCUSSION

Identification of putative DNA methyltransferasein MboI-resistant Campylobacter coli 6461

Comparative in silico analysis of the genomes of the turkey-derived strain C. coli 11601 (accession no. LKCS00000000) (Duttaet al. 2016), with MboI-susceptible DNA, and the swine-derivedstrain C. coli 6461 (accession no. LKCT00000000) (Dutta et al.2016), with MboI-resistant DNA, was performed to identify ge-nomic regions unique to the latter. One such region included

a putative DNA methyltransferase (ORF0059, 27% GC) with anadjacent ORF0060 (29% GC), annotated as conserved hypothet-ical protein (Fig. 1). ORFs 0059 and 0060 were also absent fromthe genome of C. coli 6067 (accession no. LKCQ00000000) (Duttaet al. 2016), isolated fromwater in a turkey house andwithMboI-susceptible DNA.

The ORFs immediately flanking the ORF0059-0060 cassette,i.e. ORF0058 (36% GC) and ORF0061 (33% GC), were highly con-served among C. coli strains 6461, 11601, 6067 and other se-quenced C. coli genomes, and were annotated as putative car-bamoyl phosphate synthase large subunit and hypothetical pro-tein, respectively (Fig. 1 and data not shown). These ORFs wereadjacent to each other in C. coli 11601 and other C. coli genomeslacking the ORF0059-0060 cassette (Fig. 1 and data not shown).The GC content of the ORF0059-ORF0060 cassette (average, 28%GC) was lower than that of the conserved flanking sequences(33% and 36% GC), suggesting acquisition via horizontal genetransfer from another source.

The ORF0059-deduced polypeptide shared sequence simi-larities and functional domains with DNA methyltransferasefamilies (Pfam N6˙N4 Mtase 01555), specifically N-4 cytosine-specific and N-6 adenine-specific DNA methyltransferases. Pro-tein BLAST analysis of ORF0059 and ORF0060 revealed highlysimilar (98%–100% identity) conserved hypothetical proteinsin several C. coli strains, including C. coli Bfr-CA-9557, whichalso had methylated adenines at GATC sites and was em-ployed for methylome analysis (Zautner et al. 2015). In all



Figure 2.MboI restriction profile of genomic DNA from C. coli 6461 (swine isolate)and C. coli 11601 (turkey isolate) along with those of C. coli 11601 transformants(VD13B and VD13D). Lanes: L, 100–2686 bp DNA molecular marker XIV (Roche);U, undigested genomic DNA; S, genomic DNA subjected to Sau3A1 digestion;

and M, genomic DNA subjected to MboI digestion. Digestions were performedas described (Wright et al. 2010).

cases, genomes harboring ORF0059 also harbored ORF0060, andvice versa; analysis of the sequence database failed to iden-tify C. coli genomes that harbored either ORF0059 or ORF0060alone.

ORF0060 was annotated as conserved hypothetical protein,and routine protein BLAST failed to identify conserved domains.Motifs found in most type II restriction endonucleases, e.g. PD-(D/E)XK and PD-(D/E)XK (Niv et al. 2008), were not detected.However, analysis with the updated TIGRfam database (release15.0) revealed an RNA endonuclease domain (TIGR04202, e-value0.0045). Similar analyses also revealed a 16S rRNA modificationdomain (TIGR00095, e-value 0.0012) in ORF0059, while the con-served ORF0061 (Fig. 1) harbored functional domains implicat-ing a possible role in modification of tRNA (TIGR00057, e-value3.9e-04, COG0009, e-value 2e-74) or 16S rRNA (Pfam01300, e-value 2.3e-05). It is possible that this three-gene cluster (ORF0059–0061) may have originally carried out dual functions:RNA modification and/or processing, and as a classical RMsystem.

Heterologous expression of the putativemethyltransferase (ORF0059) confirms involvementin MboI resistance

Campylobacter coli 11601 (with DNA susceptible to digestion byMboI) was successfully transformed to chloramphenicol resis-tance with MET PCR. PCR characterization of two randomlychosen chloramphenicol-resistant transformants (VD13B andVD13D) from one of the transformation experiments confirmedthat both harboredORF0059 and cat (data not shown). Digestionswith MboI (specific for GATC sites and susceptible to adeninemethylation) and Sau3AI (also specific for GATC but susceptibleto cytosine methylation) revealed that genomic DNA of VD13Band VD13D was resistant to MboI digestion but readily digestedby Sau3AI (Fig. 2). These findings support the role of ORF0059as a putative N6-adenine methyltransferase. VD13B and VD13Dwere indistinguishable from the parental strain C. coli 11601 infla type, while C. coli 6461 had a clearly different fla profile (Fig. 3).Except for chloramphenicol resistance (128 μg ml−1, in contrastto chloramphenicol MIC of 4 μg ml−1 for C. coli 11601, expecteddue to the presence of cat), VD13B and VD13D were resistantto the same antibiotics as C. coli 11601, with the same MICs(tetracycline and nalidixic acid, 128 μg ml−1; ciprofloxacin, 32μg ml−1; streptomycin, erythromycin and kanamycin, >256 μgml−1). This antimicrobial resistance profile was clearly differentfrom that of C. coli 6461, which was resistant only to tetracycline,streptomycin and erythromycin. The transformants were in ad-dition resistant to chloramphenicol (MIC, 128 μg ml−1, in con-

Figure 3. fla-typing PCR indicating similarity of the transformants to the parentalstrain, C. coli 11601. Lanes: M, 100–2686 bp DNA molecular marker XIV (Roche);1, 2, 3 and 4, VD13B, VD13D, C. coli 11601 and C. coli 6461, respectively. fla typingwas performed as described (Gu et al. 2009).

trast to 4 μg ml−1 for C. coli 11601), as expected due to the pres-ence of cat.

In silico analysis of ORF0059 and ORF0060 distributionsin Campylobacter coli reveals a propensity of theORF0059-0060 cassette for swine

In total, 758 C. coli genome sequences were identified in thedatabase (http://www.ncbi.nlm.nih.gov/genome/genomes/1145;last accessed 07/03/2016). Nucleotide BLAST analysis of ORF0059and ORF0060 sequences as separate queries revealed highlyconserved homologs (97%–100% identity at the nt sequencelevel, 100% query coverage) in 102/758 (13.5%) of the C. coligenomes, including Bfr-CA-9557 (chicken hearts at retail), forwhich methylome analysis had provided evidence for a highlyconserved ORF0059 homolog (Zautner et al. 2015). Of the 102C. coli genomes, host/source information was available for 100.A substantial fraction (7/22, 31.8%) of swine-derived strains har-bored such highly conserved ORF0059 and ORF0060 homologs.The cassette was not unique to C. coli from swine, but its fre-quency among swine-derived isolates was significantly higher(P < 0.05) than among human, environmental or bovine iso-lates. The cassette was encountered in 18.9% (46/244) amongC. coli from chickens and other avian sources, 16.4% (30/183)among human isolates, 7.5% (11/147) among environmental iso-lates and only in 3.8% (6/158) among isolates of bovine origin.

The ORF0059-0060 cassette is highly associated withCampylobacter coli

Nucleotide BLAST analysis suggests that the ORF0059-0060 cas-sette is highly associated with C. coli. Less conserved homologsof the cassette (80-81% identity at the nt level, 98% query cov-erage) were only detected in C. lari, e.g. strains NCTC 11845 andRM16701, both from river water (Fig. 4). In these C. lari genomes,the cassette was in the vicinity of genes encoding putative L-asparaginase and carbamoyl phosphate synthase large subunit,which is in the vicinity of the cassette in C. coli as well (Fig. 1).

Detection of solitary ORF0059 in other Campylobacterspp.

As discussed above, solitary ORF0059 homologs, i.e. without de-tectable ORF0060 counterparts, were not detected among C. coligenomes. However, solitary ORF0059 homologs were identi-fied in several other recently sequenced Campylobacter spp., in-cluding C. jejuni subsp. doylei 269.97 from human blood (90%


http://www.ncbi.nlm.nih.gov/genome/genomes/1145

Dutta et al. 5

Figure 4. Genomic organization of regions harboring ORF0059 homologs in Campylobacter spp. other than C. coli. ORF designations for the C. lari genomes correspondto the homologous counterparts in C. coli 6461. Arrows indicate transcription direction for each ORF. Black arrows indicate the putative methyltransferase ORF0059

homologs.

identity, 96% coverage) and other Campylobacter spp. fromwildlife, especially wild birds, such as C. subantarcticus LMG24374 and LMG 24377, from a gentoo penguin and a grey-headedalbatross, respectively; C. peloridis LMG 23910, from shellfish; andC. volucris LMG 24379 from a black-headed gull (82% identity,92%–94% coverage) (Fig. 4). In contrast to these findings, solitaryORF0060 homologs were not identified.

ORF0059 is accompanied by a putative restrictionendonuclease in Helicobacter pylori and otherHelicobacter and Campylobacter spp.

Outside of Campylobacter spp., ORF0059 homologs (77%–83%identity, 82%–85% coverage) were identified in several Helicobac-ter pylori genomes and other Helicobacter spp. It was noteworthythat inH. pylori JP28 theORF0059 homolog (77% identity, 82% cov-erage) was annotated as the methytransferase HpyIIIM of theHpyIII type II RM system. In H. pylori, an alternative gene (hrgA)has been found to frequently replace the restriction endonucle-ase (Ando et al. 2002, 2003). It is conceivable that ORF0060 con-stitutes such an alternative gene in C. coli, replacing a restrictionendonuclease cognate to the ORF0059 DNA methyltransferase.Support for this hypothesis comes from the RNA endonucleasedomain identified in ORF0060 along with the detailed analysisof genomes of Campylobacter spp. harboring solitary ORF0059 ho-mologs, as discussed above. For instance, putative restriction en-donuclease geneswere located adjacent to theORF0059homologin C. subantarcticus LMG 24374 and C. peloridis LMG 23910, consti-tuting a putative RM system (Fig. 4 and data not shown). It istempting to speculate that selection pressures specific to agri-cultural systems have resulted in ORF0060 substituting for anoriginal MboIR-like homolog, while such a restriction endonu-clease was maintained in non-agricultural campylobacters suchas those from wild birds and shellfish.

Reasons mediating persistence of the ORF0059-0060 cassettein C. coli but not in C. jejuni from food animals remain to beelucidated. Campylobacter coli exhibits a predilection for swinewhich tend to be infrequently colonized with C. jejuni (Stern1992; Aarestrup et al. 1997; Saenz et al. 2000), and it is possiblethat the cassette was selected as C. coli became adapted to itsswine host under agricultural conditions, with such strains orthe ORF0059-0060 cassette subsequently disseminating to otherhosts. Further studies are needed to assess the potential role ofthe cassette in animal colonization and also to characterize thepotential functions of ORF0060. In H. pylori, intriguing evidencesuggests that hrgA (which as mentioned frequently replaces therestriction endonuclease in the HpyIII RM system) varies in geo-graphic prevalence among humans and may be associated withgastric cancer in Asian patients (Ando et al. 2002).

As discussed earlier, DNAmethylation can impact epigeneticregulation of numerous processes (Casadesus and Low 2006;Marinus and Casadesus 2009). For instance, in C. jejuni inacti-vation of a putative DNA methyltransferase gene (cj1461) led toaberrant flagella, reduced motility and impaired cell invasion(Kim et al. 2008). Further studies are needed to investigate the po-tential roles of the ORF0059-encoded methyltransferase in reg-ulation of processes mediating certain host-associated adapta-tions, especially in interactions of C. coli with swine.

ACKNOWLEDGEMENTS

We thank Victor Jaeyola and R. L. Thompson for assistance withstatistical analysis and are grateful to other members of ourgroup for their help and encouragement.

FUNDING

This research was supported by a grant from the United StatesDepartment of Agriculture (USDA NIFA 2006-35201-17377) to



S. Kathariou and J. Olson and by a grant from the United StatesDepartment of Agriculture (USDA NIFA 2011-04811-226997) toS. Kathariou.

Conflict of interest. None declared.

REFERENCES

Aarestrup FM,Nielsen EM,MadsenM et al.Antimicrobial suscep-tibility patterns of thermophilic Campylobacter spp. from hu-mans, pigs, cattle, and broilers inDenmark.Antimicrob AgentsCh 1997;41:2244–50.

Ando T, Aras RA, Kusugami K et al. Evolutionary history of hrgA,which replaces the restriction gene hpyIIIR in the hpyIII locusof Helicobacter pylori. J Bacteriol 2003;185:295–301.

Ando T, Wassenaar TM, Peek RM et al. A Helicobacter pylori re-striction endonuclease replacing gene, hrgA, associated withgastric cancer in Asian strains. Cancer Res 2002;62:2385–9.

Casadesus J, Low D. Epigenetic gene regulation in the bacterialworld. Microbiol Mol Biol R 2006;70:830–56.

Dutta V, Altermann E, Olson J et al. Whole genome sequence ofagricultural, host-associated Campylobacter coli and Campy-lobacter jejuni strains. Genome Announc 2016;4:e00833–16.

Fitzgerald C. Campylobacter. Clin Lab Med 2015;35:289–98.Griekspoor P, Colles FM, McCarthy ND et al. Marked host speci-

ficity and lack of phylogeographic population structure ofCampylobacter jejuni in wild birds. Mol Ecol 2013;22:1463–72.

Gu W, Siletzky RM, Wright S et al. Antimicrobial susceptibilityprofiles and strain type diversity of Campylobacter jejuni iso-lates from turkeys in eastern North Carolina.Appl Environ Mi-crob 2009;75:474–82.

Huang H, Brooks BW, Lowman R et al. Campylobacter species inanimal, food, and environmental sources, and relevant test-ing programs in Canada. Can J Microbiol 2015;61:701–21.

Kim JS, Li J, Barnes IH et al. Role of the Campylobacter jejuni Cj1461DNA methyltransferase in regulating virulence characteris-tics. J Bacteriol 2008;190:6524–9.

Marinus MG, Casadesus J. Roles of DNA adenine methylationin host-pathogen interactions: mismatch repair, transcrip-tional regulation, and more. FEMS Microbiol Rev 2009;33:488–503.

McClelland M, Nelson M, Raschke E. Effect of site-specific modi-fication on restriction endonucleases and DNA modificationmethyltransferases. Nucleic Acids Res 1994;22:3640–59.

Niv MY, Skrabanek L, Roberts RJ et al. Identification of GATC-and CCGG-recognizing Type II REases and their putative

specificity-determining positions using Scan2S–a novel mo-tif scan algorithm with optional secondary structure con-straints. Proteins 2008;71:631–40.

Ogden ID, Dallas JF, MacRae M et al. Campylobacter excretedinto the environment by animal sources: prevalence, con-centration shed, and host association. Foodborne Pathog Dis2009;6:1161–70.

Palmer BR, Marinus MG. The dam and dcm strains of Escherichiacoli—a review. Gene 1994;143:1–12.

Reisenauer A, Kahng LS, McCollum S et al. Bacteria DNA methy-lation: a cell cycle regulator? J Bacteriol 1999;181:5135–9.

Roberts RJ, Vincze T, Posfai J et al. REBASE–enzymes and genesfor DNA restriction and modification. Nucleic Acids Res 2007:35 (Database issue):D269–70.

Saenz Y, Zarazaga M, Lantero M et al. Antibiotic resistancein Campylobacter strains isolated from animals, foods,and humans in Spain in 1997-1998. Antimicrob Agents Ch2000;44:267–71.

Sheppard SK, Cheng L, Meric G et al. Cryptic ecology among hostgeneralist Campylobacter jejuni in domestic animals. Mol Ecol2014;23:2442–51.

Sheppard SK, Colles FM, McCarthy ND et al. Niche segrega-tion and genetic structure of Campylobacter jejuni popu-lations from wild and agricultural host species. Mol Ecol2011;20:3484–90.

Sheppard SK, Maiden MC. The evolution of Campylobacter je-juni and Campylobacter coli. Cold Spring Harb Perspect Biol2015;7:a018119.

Stern NJ. Reservoirs for Campylobacter jejuni and approaches forintervention in poultry. In: Nachamkin I, Blaser MJ, Tomp-kins LS (eds). Campylobacter jejuni: Current Status and FutureTrends. Washington, DC: American Society for Microbiology,1992, 49–60.

Wright S,Wilson S, MillerWG et al.Differences inmethylation atGATC sites in genomicDNAofCampylobacter coli from turkeysand swine. Appl Environ Microb 2010;76:7314–7.

Yanisch-Perron C, Vieira J, Messing J. Improved M13 phagecloning vectors and host strains: nucleotide sequences of theM13mp18 and pUC19 vectors. Gene 1985;3:103–19.

Yao R, Alm RA, Trust TJ et al. Construction of new Campylobac-ter cloning vectors and a new mutational cat cassette. Gene1993;130:127–30.

Zautner AE, Goldschmidt AM, Thurmer A et al. SMRT se-quencing of the Campylobacter coli BfR-CA-9557 genome se-quence reveals unique methylation motifs. BMC Genomics2015;16:1088.


identification of a campylobacter coli methyltransferase targeting

Documents