Pancreatic Amylase Is an Environmental Signal for Regulation ofBiofilm Formation and Host Interaction in Campylobacter jejuni

Waheed Jowiya,a Katja Brunner,b Sherif Abouelhadid,c Haitham A. Hussain,a Sean P. Nair,a Sohaib Sadiq,d Lisa K. Williams,e

Emma K. Trantham,e Holly Stephenson,b* Brendan W. Wren,c Mona Bajaj-Elliott,b Tristan A. Cogan,e Andrew P. Laws,d Jim Wade,f

Nick Dorrell,c Elaine Allana

Department of Microbial Diseases, UCL Eastman Dental Institute, University College London, London, United Kingdoma; Department of Infection and Immunity, Instituteof Child Health, University College London, London, United Kingdomb; Faculty of Infectious & Tropical Diseases, London School of Hygiene & Tropical Medicine, London,United Kingdomc; School of Applied Sciences, University of Huddersfield, Huddersfield, United Kingdomd; School of Veterinary Sciences, University of Bristol, Bristol,United Kingdome; Medical Microbiology, King’s College Hospital, London, United Kingdomf

Campylobacter jejuni is a commensal bacterium in the intestines of animals and birds and a major cause of food-borne gastroen-teritis in humans worldwide. Here we show that exposure to pancreatic amylase leads to secretion of an �-dextran by C. jejuniand that a secreted protease, Cj0511, is required. Exposure of C. jejuni to pancreatic amylase promotes biofilm formation invitro, increases interaction with human epithelial cell lines, increases virulence in the Galleria mellonella infection model, andpromotes colonization of the chicken ileum. We also show that exposure to pancreatic amylase protects C. jejuni from stressconditions in vitro, suggesting that the induced �-dextran may be important during transmission between hosts. This is the firstevidence that pancreatic amylase functions as an interkingdom signal in an enteric microorganism.

Campylobacter infection is a major cause of food-borne gastro-enteritis in humans worldwide (1). Disease ranges from mild,

noninflammatory, self-limiting diarrhea to prolonged, inflamma-tory diarrhea and occasional serious complications such as Guil-lain-Barré syndrome and reactive arthritis (1). Ninety percent ofhuman disease is attributed to Campylobacter jejuni, while Cam-pylobacter coli accounts for the remainder (2).

In contrast to human infection, C. jejuni establishes a largelyasymptomatic but persistent infection in animals and birds, andthe major risk of human infection is from handling and consump-tion of poultry meat (3–5). Campylobacters primarily inhabit thelower gastrointestinal (GI) tract of poultry, and contamination ofthe flesh of the bird occurs at slaughter. Campylobacter organismsare present in the luminal crypts of the chicken gut as denselypacked communities surrounded by mucus (6), an arrangementsuggestive of a biofilm, a structured community of bacteria en-closed in a self-produced exopolymeric matrix (EPM) composedprimarily of polysaccharides (7). Biofilms have also been ob-served on experimentally infected human ileum ex vivo, sug-gesting that this mode of growth is also important during hu-man infection (8, 9).

Despite several studies of C. jejuni biofilms (10–15), the EPMhas not been characterized. Polysaccharide is a major componentof most bacterial EPMs and provides the structural framework (7,16, 17). C. jejuni produces several surface-associated carbohydratestructures, including lipooligosaccharide (LOS), capsular poly-saccharide (CPS), and both N- and O-linked glycoproteins (18),but there is no evidence that these glycans contribute to biofilmformation (11, 19).

Enteric bacteria use environmental signals present in host en-vironments to regulate the expression of components required forinteraction with the host, and these signals include bile salts, mu-cus, low oxygen, bicarbonate, and neuroendocrine stress hor-mones (20–22). C. jejuni shows increased interaction with intes-tinal epithelial cells (IECs) in response to low oxygen, bile salts,and noradrenaline (23–25). In addition, bile salts have recently

been reported to enhance C. jejuni biofilm formation by causingrelease of extracellular DNA (26).

Here, we show that C. jejuni responds to the presence of pan-creatic amylase by secreting an �-dextran that is a component ofthe biofilm EPM. This is the first definitive characterization of thepolysaccharide component of C. jejuni biofilms. Importantly, pre-exposure to pancreatic amylase results in significant changes inthe interactions of C. jejuni with the host in both in vitro and invivo infection models. Preexposure to pancreatic amylase alsopromotes stress tolerance in vitro, suggesting that the �-dextranmay be important during transmission. This is the first evidencethat pancreatic amylase is used by an enteric microorganism as asignal to regulate interaction with the host.

MATERIALS AND METHODSBacterial strains and growth conditions. Bacterial strains are listed inTable 1. C. jejuni was stored at �70°C in brucella broth (BB; Oxoid)containing 15% glycerol and was grown on Mueller-Hinton agar (MHA;Oxoid), brucella agar (BA; Oxoid), and Columbia blood agar (CBA; Ox-oid) containing 7% defibrinated horse blood (E & O Laboratories) in a

Received 18 August 2015 Returned for modification 7 September 2015Accepted 30 September 2015

Accepted manuscript posted online 5 October 2015

Citation Jowiya W, Brunner K, Abouelhadid S, Hussain HA, Nair SP, Sadiq S,Williams LK, Trantham EK, Stephenson H, Wren BW, Bajaj-Elliott M, Cogan TA, LawsAP, Wade J, Dorrell N, Allan E. 2015. Pancreatic amylase is an environmental signalfor regulation of biofilm formation and host interaction in Campylobacter jejuni.Infect Immun 84:4884 –4895. doi:10.1128/IAI.01064-15.

Editor: S. R. Blanke

Address correspondence to Elaine Allan, e.allan@ucl.ac.uk.

* Present address: Holly Stephenson, Department of Cellular Microbiology, MaxPlanck Institute for Infection Biology, Berlin, Germany.

Supplemental material for this article may be found at http://dx.doi.org/10.1128/IAI.01064-15.

Copyright © 2015, American Society for Microbiology. All Rights Reserved.

4884 iai.asm.org December 2015 Volume 83 Number 12Infection and Immunity

on August 8, 2020 by guest

http://iai.asm.org/

Dow

nloaded from

microaerobic atmosphere generated using CampyPaks (Oxoid) in gas jarsat 37°C. Charcoal (0.4%) was added to BA to improve contrast for pho-tography. For broth cultures, BB or Eagle’s minimal essential mediumalpha (MEM-�; Sigma-Aldrich, no. M0894) were used, and cultures wereincubated with shaking (50 rpm) at 37°C in 5% CO2. Kanamycin (Km; 40�g/ml) and chloramphenicol (Cm; 20 �g/ml) were used for selection.

Sources of pancreatic amylase. Hog pancreatic �-amylase (HPA) waspurchased from Sigma-Aldrich. Polyhistidine-tagged human pancreaticamylase expressed and purified from human cells was obtained from SinoBiologicals. For chicken pancreas extract, 20 organs were homogenized in0.05 M Tris, 0.9% NaCl, 0.05 M CaCl2 (pH 8.0) (27). After centrifugation(1,000 � g, 20 min), amylase activity in the supernatant was quantified byenzymatic assay (28).

EPM purification and characterization. Growth (48 h) removedfrom MHA with or without 100 nM HPA was suspended in phosphate-buffered saline (PBS), and the extracellular material was recovered usingmodifications of a method previously used to isolate exopolysaccharidefrom Vibrio parahaemolyticus (29). Briefly, the bacterial cell suspensionsin PBS were washed on a rotary platform at 200 rpm for 1.5 h at 30°C,vortexed for 15 min, and washed again at 200 rpm for 1.5 h at 20°C. Thesupernatant was recovered (1,800 � g, 15 min) and precipitated overnightwith 4 volumes of cold acetone at 4°C. The precipitate was recovered (450 �g, 5 min), washed in distilled water (dH2O), dried (Speedvac SPD1010;Thermo Scientific), dissolved in dH2O, and stored at �70°C. Total car-bohydrate was measured by phenol-sulfuric acid assay (30) with glucoseas the standard. To determine statistical significance, a one-way analysis ofvariance (ANOVA) was used to compare the amount of secreted carbo-hydrate in response to the seven different doses of pancreatic amylase.Within the one-way ANOVA, we constructed two contrasts; the first com-pared the joint effect of the three lowest doses with the joint effect of thefour highest doses, and the second compared the effect of a dose of 50 nMpancreatic amylase with that of a 100 nM dose. As independent samples ofa bacterial culture were exposed to the different doses of amylase in aclosed system, we assumed independence between the seven experiments.To analyze the amount of carbohydrate secreted by the wild type, thecj0511 mutant, and the complemented strain, a two-way ANOVA was

performed comparing the amount of secreted carbohydrate in response tothe four different doses of pancreatic amylase in the three different strains.Contrasts of the 15 nM concentration jointly with the 50 nM concentra-tion versus the 100 nM concentration jointly with the 500 nM concentra-tion were constructed for C. jejuni 11168H, the cj0511 mutant, and thecomplemented strain. Differences between these three contrasts weretested for statistical significance. The contrasts and comparisons wereplanned a priori, and thus no multiple-testing adjustments were required.

Further characterization of the EPM carbohydrate was performed af-ter dialysis (14,000 molecular weight cutoff [MWCO]) for 72 h at 4°Cagainst three changes of dH2O per day. The EPM was freeze-dried andcharacterized using nuclear magnetic resonance (NMR) spectroscopy(31), monomer analysis (32), and linkage analysis (33) in comparisonwith a commercial �-dextran standard (Sigma-Aldrich).

Genetic complementation of a cj0511 mutant. A cj0511 mutant ob-tained from a collection at the London School of Hygiene and TropicalMedicine (Table 1) was used to construct a complemented strain by in-sertion of cj0511 under the control of the iron-inducible fdxA promoterinto a pseudogene (cj0046) in the cj0511 mutant as previously described(34). The oligonucleotides used are listed in Table 2.

Production of recombinant Cj0511 and proteolysis assay. Codon-optimized cj0511 was synthesized by Celtek, USA, and obtained in pGH.The gene was cut by HindIII (New England BioLabs, United Kingdom),and the released DNA fragment was cloned in pET-28a(�) (Novagen,United Kingdom) to generate pMAR31. Expression of cj0511 wasachieved in Escherichia coli BL21(DE3) (New England BioLabs, UnitedKingdom). To localize the recombinant protein in the periplasm of E. coli,recombinant Cj0511 was modified to have a signal peptide from the E. coliDsbA protein fused to the N terminus, while the C terminus was fused toa 6-His affinity tag. Recombinant Cj0511 was purified using nickel affinitycolumns (Qiagen) and eluted in 500 mM imidazole, 300 mM sodiumphosphate, and 20 mM sodium chloride buffer. The eluted fractions werecombined, concentrated using an Amicon ultrafiltration column with a100-kDa cutoff (Merck Millipore, United Kingdom), and then resus-pended in 10 mM sodium phosphate buffer. One microgram of purifiedr-Cj0511 was mixed with 2 �g polyhistidine-tagged human pancreatic

TABLE 1 Bacterial strains and plasmids

Strain or plasmid Characteristic(s) Reference or source

StrainsC. jejuni 11168H Hypermotile variant of NCTC 11168; good colonizer of chickens 63C. jejuni 81-176 Clinical isolate from human diarrhea sample 64C. jejuni 81116 Waterborne human outbreak strain 65C. jejuni X Clinical isolate from patient with enteritis 66C. jejuni G1 Clinical isolate from patient with Guillain-Barré syndrome 67C. jejuni 11168H kpsM::aphA3 Mutation in gene encoding ABC transporter involved in capsule assembly; no CPS

expression39

C. jejuni 11168H waaC::aphA3 Mutation in gene encoding heptosyltransferase I; expresses truncated LOS CampylobacterResource Facilitya

C. jejuni 11168H waaF::aphA3 Mutation in gene encoding heptosyltransferase II; expresses truncated coreoligosaccharide

CampylobacterResource Facility

C. jejuni 11168H cj0511::aphA3 Mutation in gene encoding secreted carboxyl-terminal protease CampylobacterResource Facility

11168H cj0511::aphA3 pWJ4 Complemented strain containing cj0511 gene and the Camr gene inserted into cj0046 This studyC. rectus NCTC 11489 Type strain, from human periodontal pocket Public Health England

culture collectionsE. coli JM109 Cloning host PromegaE. coli BL21(DE3) Expression host New England Biolabs

PlasmidspJMK30 pUC19 carrying C. coli Kanr gene aphA-3 68pGEM-T-Easy Cloning vector enables TA cloning PromegapC46fdxA Complementation vector containing fdxA promoter and Camr selection marker 34

a http://crf.lshtm.ac.uk/index.htm.

C. jejuni Response to Pancreatic Amylase

December 2015 Volume 83 Number 12 iai.asm.org 4885Infection and Immunity

on August 8, 2020 by guest

http://iai.asm.org/

Dow

nloaded from

amylase (Sino Biologicals) or casein (Sigma-Aldrich) in 20 �l of 50 mMsodium phosphate buffer (pH 7.5) and incubated at 37°C for a total of 2 h.The reactions were stopped using 6 � Laemmli buffer, and the reactionmixtures were boiled for 10 min before being separated by SDS-PAGE(16% polyacrylamide) and stained with Coomassie R-250.

Biofilm assays. A 16-h culture of C. jejuni in BB was diluted to anoptical density at 600 nm (OD600) of 0.1 in BB with or without 100 nMHPA, 100 nM recombinant human pancreatic amylase, or an equivalentactivity of chicken pancreas extract, added to a borosilicate glass tube(VWR International), and incubated without shaking at 37°C in 5% CO2.After 48 h, the culture was decanted, and the biofilms were washed withdH2O, dried and stained for 5 min with 0.1% (wt/vol) crystal violet (CV;Sigma-Aldrich), washed with dH2O, and dried. Bound CV was solubilizedin 80% ethanol–20% acetone and quantified by measuring OD600. Statis-tical significance was assessed with a two-sample t test assuming unequalvariance.

Confocal laser scanning microscopy of biofilm. A 16-h culture of C.jejuni was diluted to an OD600 of 0.1 in BB with or without 100 nM HPAand placed in a 6-well tissue culture plate (Sarstedt). Two glass coverslipswere placed upright in each well and incubated for 48 h at 37°C in 5%CO2. The coverslips were washed in PBS, stained with Live/Dead BacLightstain (Invitrogen) according to the manufacturer’s instructions, and visu-alized with a Bio-Rad Radiance 2100 confocal scanning system attached toan Olympus BX51 upright microscope. Digital images were producedusing Image J software (National Institutes of Health).

Adhesion and invasion of Caco-2 cells. Human colon cancer cells(Caco-2) were cultured as previously described (25), and 108 bacteria,grown on MHA with or without 100 nM HPA, were added at a multi-plicity of infection (MOI) of 100:1 and incubated at 37°C in 5% CO2.Interacting and invading bacteria were enumerated at 3, 6, and 24 h aspreviously described (25), and statistical significance was assessed using atwo-sample t test assuming unequal variance.

Interaction with T84 human colonic epithelial cells. T84 cells weregrown in complete Dulbecco’s modified Eagle’s medium nutrient mixturewith Ham F-12 (Invitrogen) in rat collagen-precoated Transwell dishesuntil the transepithelial electrical resistance (TEER) (35) was �800 �,confirming formation of tight junctions in the monolayer. Coculturestudies were performed at an MOI of 10:1 in DMEM–Ham F-12 mediumcontaining 10% fetal bovine serum for 6 or 24 h. Bacteria were grown onMHA with or without 100 nM HPA, and viable counts were determined.Statistical significance was assessed using a Wilcoxon matched paired test.Interleukin 8 (IL-8) levels were determined by enzyme-linked immu-nosorbent assay (ELISA) (eBioscience) from the supernatant of the apicalsurface at 24 h postinfection. Statistical significance was assessed using aone-way ANOVA.

G. mellonella infection model. Galleria mellonella larvae were ob-tained commercially (Cornish Crispa Co.) and stored in containers withwood chips at 12°C. C. jejuni strains were grown on MHA with or without100 nM HPA for 48 h and harvested in PBS, and 106 CFU were injectedinto the right forelegs of 10 Galleria larvae as previously described (36).Statistical significance was assessed using a two-sample t test, assumingunequal variance.

Infection of chickens. Day-old Ross broiler chickens were obtainedfrom a commercial supplier and housed in biosecure housing. The study

was performed under a project license issued by the UK Home Officeunder the Animals (Scientific Procedures) Act 1986 (PPL 30/2599), andprotocols and euthanasia techniques were approved by the University ofBristol Animal Welfare Ethical Review Board. Birds were fed commercialstarter and grower diets (BOCM Pauls Ltd., Ipswich), had access to foodand water ad libitum, and were exposed to a 12-h light/dark cycle. At 21days old, groups of 30 birds were infected by oral gavage with 105 of C.jejuni 11168H or the cj0511 mutant grown on MHA with or without 100�M HPA for 48 h and resuspended in MH broth. Fifteen birds wereeuthanized by cervical dislocation at 4 and 7 days postinfection, and ileumand liver samples were removed for culture. Ileum samples were seriallydiluted and plated on modified charcoal cefoperazone deoxycholate agar(mCCDA; Oxoid) incubated at 37°C for 48 h under microaerobic condi-tions. Liver samples were homogenized and enriched in modified Exeterbroth (37), incubated with minimal headspace at 37°C for 48 h, and platedon mCCDA. Differences in the number of birds colonized were assessedusing a chi-square test. The distributions of numbers of bacteria found insamples from the birds were checked for normality using a D’Agostinoand Pearson omnibus normality test; as data were not shown to follow aGaussian distribution, differences in the number of bacteria found at eachsite were analyzed using a Kruskal-Wallis test with Dunn’s multiple-com-parison test, using a Benjamini and Hochberg correction to adjust formultiple pairwise comparisons, as a post hoc test.

Resistance to environmental stress. The growth from 48-h MHAplates with and without 100 nM HPA was recovered in PBS (4°C) or MHbroth (20°C), washed, and resuspended to an OD600 of 2.0. The cultureswere incubated with shaking (50 rpm) at 37°C in 5% CO2, and sampleswere removed at intervals for viable counting. The statistical significanceof the difference in viable counts in the presence and absence of HPA wasassessed using two-sample t tests, assuming unequal variance.

RESULTSC. jejuni responds to the presence of pancreatic amylase. Phys-iological concentrations of mammalian pancreatic �-amylase areestimated to be nanomolar (38). When hog pancreatic �-amylase(HPA) was incorporated into agar at various concentrations, weobserved the formation of large mucoid colonies of C. jejuni11168H at a minimum concentration of 100 �M (Fig. 1A). It isimportant to note that mucoid colonies were apparent on MHAand also on BA containing 100 �M HPA; since BA does not con-tain starch, this indicates that the response is independent ofstarch-degrading amylase activity. While 100 �M pancreatic am-ylase seems high compared to the estimated in vivo concentra-tions, it is noteworthy that the number of molecules of amylasethat can interact with the bacteria growing on an agar plate is likelyto be low because of spatial confinement, in contrast to physiolog-ical conditions, where bacteria are bathed in liters of pancreaticjuice per day. Mucoid colonies were also produced by other C.jejuni strains (81-176, 81116, G1, and X) in response to 100 �MHPA but not by the periodontal pathogen Campylobacter rectusNCTC 11489 (data not shown).

Pancreatic amylase promotes C. jejuni growth. As the colo-

TABLE 2 Oligonucleotides

Name Sequence (5=-3=)a Use

WJ10F GTCGTCTCACATGTTGAAAACAAAACG Amplification of cj0511 for cloning in pC46fdxAWJ10R GTCGTCTCACATGTTATTGTCCTTGTTTG Amplification of cj0511 for cloning in pC46fdxAWJ4 GTAAATTTTTGATTATCAAAATTTACATTATTTAAG Orientation of cj0511 with respect to PfdxA in pC46fdxA; anneals in PfdxA

WJ11 GGAACACAGCAGAGCACTTG Confirmation of cj0511 complementing strain; anneals in cjs03 RNA codingsequence located 5= to cj0046

WJ12 CCTGGAGAAGTATTAGATAGTAGCG Confirmation of cj0511 complementing strain; anneals in cj0053c located 3= to cj0046a BsmBI sites are underlined.

Jowiya et al.

4886 iai.asm.org December 2015 Volume 83 Number 12Infection and Immunity

on August 8, 2020 by guest

http://iai.asm.org/

Dow

nloaded from

nies on amylase were larger in addition to being mucoid, the num-ber of bacteria per colony was determined for C. jejuni 11168Hfrom HPA (100 �M)-containing and amylase-free MHA. In threeexperiments, colonies contained a mean of (5.3 3.2) � 107 CFUin the presence of amylase, compared to (3.3 2.3) � 106 CFU inthe absence of amylase, representing a 16-fold increase. InMEM-�, supplementation with physiological concentrations(100 nM) of HPA resulted in an increased growth rate (meangeneration time of 180 0 minutes in the presence of pancreaticamylase, compared to 280 35 minutes in its absence). The dif-ference in growth rate was statistically significant (P 0.001) atboth mid-exponential phase (10 h) and stationary phase (15 h)(Fig. 1B).

Mucoid colonies secrete increased amounts of carbohydrate.As mucoid colonies are suggestive of exo-polysaccharide produc-tion, we attempted to purify the extracellular material from colo-nies of strain 11168H grown in the presence of increasing concen-trations of HPA by washing in PBS. Measurement of thecarbohydrate content of this preparation by phenol sulfuric acidassay showed that exposure to physiological concentrations ofHPA (100 nM) resulted in a 100% increase in the soluble carbo-hydrate secreted by C. jejuni 11168H compared to growth withoutHPA (Fig. 2A). A 11168H kpsM mutant, unable to export CPS(39), and 11168H waaF and waaC mutants lacking the outer andentire core oligosaccharide of LOS, respectively (40, 41), alsoshowed increased carbohydrate secretion of a magnitude compa-rable to that of the wild-type (WT) strain in the presence of HPA,indicating that the secreted carbohydrate is independent of bothCPS and LOS (Fig. 2B). The amount of carbohydrate secreted bythe 11168H kpsM, waaC, and waaF mutants did not differ signif-icantly from that secreted by the wild-type strain in either thepresence or absence of pancreatic amylase.

The secreted carbohydrate is an �-dextran. The soluble ex-tracellular material recovered from C. jejuni 11168H and the kpsMmutant provided 1H- and 13C-NMR spectra identical to that of an�-dextran standard (Fig. 2C) and distinct from that of starch (seeFig. S1 in the supplemental material). A series of two-dimensionalNMR (correlation spectroscopy [COSY], heteronuclear singlequantum coherence [HSQC], and heteronuclear multiple-bondcoherence [HMBC]) spectra were also recorded (data not shown)to confirm the identity of the exopolysaccharide as an �-dextran.Monomer analysis was undertaken using high-performance an-ion-exchange chromatography in combination with pulsed am-

perometric detection. After acid hydrolysis, the extracellular ma-terial recovered from both C. jejuni 11168H and the kpsM mutantgave chromatographs containing a single peak which coelutedwith a glucose reference standard (see Fig. S2 in the supplementalmaterial). Monomer analysis thus confirmed that glucose was theonly monosaccharide present, and linkage analysis showed exclu-sively 1,6-glycosidic links (see Fig. S3 in the supplemental mate-rial).

Cj0511 is essential for the response to pancreatic amylase butdoes not degrade it in vitro. Since HPA improves the growth of C.jejuni, we reasoned that it could be degraded by the bacterium. AsCj0511 is a protease (42) that is known to be secreted in outermembrane vesicles (OMVs) (43) and is abundant in a C. jejunistrain that colonizes chickens efficiently (44), we hypothesizedthat it is involved in the response to pancreatic amylase. Determi-nation of the amount of secreted carbohydrate showed that theincreased secretion seen in the WT strain in response to stimula-tion with HPA was absent in a cj0511 mutant (Fig. 2D). Restora-tion of the response to HPA in the complemented strain (Fig. 2D)showed that the defect in the cj0511 mutant is the result of inacti-vation of cj0511 and not due to a polar effect or spontaneousmutation elsewhere in the genome. To determine if Cj0511 de-grades pancreatic amylase, recombinant Cj0511 (r-Cj0511) wasincubated with recombinant human pancreatic amylase or withcasein (Fig. 3). Partial and complete degradation of casein wasapparent at 30 min and 2 h, respectively. In contrast, there was noapparent degradation of pancreatic amylase after 2 h.

Exposure to pancreatic amylase promotes biofilm forma-tion. To determine if exposure to pancreatic amylase contributedto biofilm formation in C. jejuni, biofilms formed at the air/liquidinterface in glass tubes in the presence or absence of physiologicalconcentrations of pancreatic amylase (100 nM) were measured bycrystal violet staining. With the exception of the cj0511 mutant, allthe strains showed significant increases in biofilm formation inthe presence of HPA (Fig. 4A). The magnitude of the increase inbiofilm in response to HPA in the waaC, waaF, and kpsM mutantswas similar to that in the WT (2.1- to 2.5-fold), demonstratingthat it is independent of LOS and CPS. Although we observedincreased biofilm formation by the cj0511 mutant compared tothe WT in the absence of HPA (P � 0.003), there was no signifi-cant increase in the presence of HPA in this mutant (in contrastwith the complemented strain), demonstrating that a functionalCj0511 protease is required for the response (Fig. 4A). To confirm

FIG 1 Exposure to pancreatic amylase promotes growth and results in a large, mucoid colony. (A) C. jejuni 11168H grown at 37°C for 72 h on BA containing0.4% charcoal without (left) or with (right) 100 �M HPA. The ruler marks 5-mm intervals. (B) C. jejuni 11168H growth with shaking (50 rpm) in 5% CO2 at 37°Cin MEM-� with (squares) or without (diamonds) 100 nM HPA. Data are means and standard deviations (SD) from three independent experiments. Thestatistical significance of the difference in growth in the presence and absence of HPA was assessed at two time points (10 and 15 h) using a two-sample t test,assuming unequal variance (***, P 0.001).

C. jejuni Response to Pancreatic Amylase

December 2015 Volume 83 Number 12 iai.asm.org 4887Infection and Immunity

on August 8, 2020 by guest

http://iai.asm.org/

Dow

nloaded from

that the increased biofilm was specifically due to pancreatic amy-lase, the assay was repeated in the presence of 100 nM recombi-nant human pancreatic amylase, and an increase in biofilm den-sity of similar magnitude (2.2-fold; P � 0.001) to that obtainedwith the preparation of HPA was observed (Fig. 4B). Addition of avolume of chicken pancreas extract containing equivalent amylase

activity also resulted in an increase in biofilm density (2.2-fold;P 0.001) (Fig. 4C).

Confocal laser scanning microscopy (CLSM) was used to visu-alize biofilms of C. jejuni 11168H grown in the presence or absenceof HPA. In the presence of HPA there was typical biofilm structureat 48 h with adherent microcolonies of live bacteria separated by

FIG 2 Physiological concentrations of pancreatic amylase induces �-dextran secretion in C. jejuni and the secreted protease, Cj0511 is required. (A) Totalextracellular carbohydrate in C. jejuni 11168H grown in the presence of increasing concentrations of HPA. Data are means and standard deviations from threeexperiments (P 0.0001 for the contrast between the three lowest doses and the four highest doses of pancreatic amylase; P 0.0001 for the contrast between50 mM and 100 nM pancreatic amylase within one-way ANOVA). (B) Total extracellular carbohydrate in different C. jejuni strains grown in the presence (blackbars) or absence (white bars) of 100 nM HPA. Data are means and standard deviations from three experiments. The kpsM, waaC, and waaF mutants were createdin strain 11168H. Statistically significant increases in the presence of HPA were determined by a two-sample t test (***, P 0.001; **, P 0.01; *, P 0.05). (C)1H-NMR spectra of extracellular material from C. jejuni 11168H, a kpsM mutant and a commercial �-dextran recorded in dH2O at 7°C using acetone as aninternal standard on a Bruker Avance 400 MHz spectrometer. (D) Total extracellular carbohydrate in C. jejuni 11168H (black bars), a Cj0511 mutant (white bars),and a Cj0511 complemented strain (gray bars) grown on MHA with increasing concentrations of HPA. A two-way ANOVA showed that the interaction betweenthe bacterial strain and the dose of pancreatic amylase was highly significant (P 0.0001). The contrasts comparing the pattern of the response between thewild-type strain and the Cj0511 mutant and between the mutant and the complemented strain were both statistically significant (P 0.0001).

FIG 3 Proteolysis of casein and pancreatic amylase by r-Cj0511. Coomassie-stained SDS-polyacrylamide gel showing no detectable proteolysis of polyhistidinetagged human pancreatic amylase. Casein served as a positive control.

Jowiya et al.

4888 iai.asm.org December 2015 Volume 83 Number 12Infection and Immunity

on August 8, 2020 by guest

http://iai.asm.org/

Dow

nloaded from

dark channels, whereas there were few adherent bacteria in theabsence of HPA (Fig. 4D). Examination in the x-z plane showedtypical three-dimensional biofilm structure (maximum height,160 �m) only in the presence of HPA.

Preexposure to pancreatic amylase promotes interactionwith human IECs. To determine if preexposure of C. jejuni toHPA affects the interaction with IECs, strains grown with or with-out 100 nM HPA were cocultured with Caco-2 cells. C. jejuni11168H grown in the presence of HPA showed significant in-creases in the number of interacting bacteria after 3, 6, and 24 h ofcoculture (Fig. 5A). Growth of C. jejuni 81-176 in the presence ofHPA also resulted in significant increases in interaction withCaco-2 cells (data not shown). In contrast, exposure of the C.jejuni 11168H cj0511 mutant to HPA did not result in increasedinteraction, whereas the complemented strain did show signifi-cant increases (Fig. 5A). Preexposure to HPA also resulted in in-creased invasion by strains 11168H (Fig. 5B) and 81-176 (data notshown). There was no increase in invasion by the cj0511 mutantwith preexposure to HPA, whereas the complemented strainshowed a significant increase (Fig. 5B).

To explore the effect of preexposure to HPA on IEC cytokineresponses, we measured interleukin 8 (IL-8) secretion from T84human colon cancer cells in response to bacteria grown with orwithout HPA. At 24 h, we observed a significant increase in IL-8secretion in response to infection but growth on HPA had noeffect (Fig. 5C). A similar lack of cytokine response to HPA pre-treatment was observed in THP-1 macrophages (data not shown).The effect of infection on transepithelial electrical resistance

(TEER) was also investigated. No disruption in tight-junction in-tegrity in response to C. jejuni grown on HPA was noted 6 hpostinfection (data not shown), but there was a significant in-crease in translocation of the bacteria preexposed to HPA com-pared to those grown without HPA (Fig. 5D), a difference nolonger apparent at 24 h (data not shown).

Preexposure to pancreatic amylase results in increased kill-ing of Galleria mellonella larvae. C. jejuni 11168H and 81-176and the cj0511 and kpsM mutants were grown with or withoutHPA and injected into 10 G. mellonella larvae, and the number ofdead larvae was determined at 24 h (Fig. 6). For both WT strains,a mean of 1 or 2 larvae were killed when the bacteria were grownwithout HPA compared to a mean of �9 when bacteria weregrown with HPA. The increased virulence in response to HPA waslost in the cj0511 mutant but restored in the complemented strain(Fig. 6), confirming the essential role of Cj0511 in signal detection.

Preexposure to pancreatic amylase promotes colonization ofbroiler chickens. To determine whether preexposure to pancre-atic amylase affects the ability of C. jejuni to colonize chickens,strain 11168H and the cj0511 mutant grown with or without HPAwere used to infect broiler chickens of 21 days old. At day 4, theWT strain had colonized better than the cj0511 mutant, but nodifference between treatments was apparent (Fig. 7A). By day 7,the number of birds colonized by the WT strain exposed to HPAprior to infection was significantly greater than the number ofbirds colonized by this strain not exposed to HPA (P � 0.005)(Fig. 7B). Differences in the numbers of bacteria present betweeninfection groups were seen at day 4 (P � 0.0006; Kruskal-Wallis

FIG 4 Exposure to pancreatic amylase promotes biofilm formation. (A) Biofilm formation at 48 h by C. jejuni strains on glass tubes in the presence (black bars)or absence (white bars) of 100 nM HPA. Data are means and SD from three independent experiments. ***, P 0.0001. (B) Biofilm formation at 48 h by C. jejuni11168H on glass tubes in the presence (black bars) or absence (white bars) of 100 nM recombinant human amylase. Data are means and SD from threeindependent experiments. ***, P 0.001. (C) Biofilm formation at 48 h by C. jejuni 11168H on glass tubes in the presence (black bars) or absence (white bars)of chicken pancreas extract. Data are means and SD from three independent experiments. ***, P 0.001. (D) CLSM of Live/Dead-stained C. jejuni 11168Hbiofilm at 48 h. Top, confocal images of bacteria in the x-y plane; bottom, digital images in the x-z plane.

C. jejuni Response to Pancreatic Amylase

December 2015 Volume 83 Number 12 iai.asm.org 4889Infection and Immunity

on August 8, 2020 by guest

http://iai.asm.org/

Dow

nloaded from

test) and day 7 (P � 0.0026). At day 4 in chickens infected withstrains grown in the presence of HPA the WT strain was present insignificantly higher numbers in the ileum (13/15 birds infected;median, 3.0 � 104 CFU/g; range, 0 to 1.5 � 105 CFU/g) compared

to the cj0511 mutant (5/15 birds infected; median, 0 CFU/g; range,0 to 1.5 � 104 CFU/g) (Fig. 7C). Growth on HPA also significantlyincreased the numbers of the WT strain in the ileum at day 7(11/15 birds infected; median, 2.0 � 104 CFU/g; range, 0 to 3 �104 CFU/g) compared to the unexposed WT (5/15 birds infected;median, 0 CFU/g; range, 0 to 3 � 104 CFU/g), whereas there wasno such increase detectable in the cj0511 mutant (Fig. 7D).

Preexposure to pancreatic amylase promotes survival instress environments. To assess the role of the �-dextran in sur-vival outside the host, we cultured C. jejuni 11168H in the pres-ence or absence of 100 nM HPA, recovered the bacteria in eitherbroth or PBS without HPA, and determined by viable countingthe numbers of surviving bacteria at room temperature and at 4°Cin air. Under both stress conditions, preexposure to HPA pro-longed survival, suggesting that the �-dextran has a protectiveeffect (Fig. 8).

DISCUSSION

Given that C. jejuni has been described as a “sugary bug” (45), it issomewhat surprising that until the work described herein, no de-finitive identification of a polysaccharide involved in biofilm for-mation had been made. Indeed, it has been suggested that DNA isa major component of the C. jejuni biofilm matrix (26). However,when C. jejuni is freshly isolated from animal or human stools, it ishighly mucoid (E. Allan, unpublished observation), suggesting

FIG 5 Preexposure to pancreatic amylase promotes interaction with human intestinal epithelial cells. (A) Interaction with Caco-2 cells by C. jejuni 11168Hgrown in the presence (black bars) or absence (white bars) of 100 nM HPA. Bacteria were cocultured with Caco-2 cells for 3, 6, and 24 h. (B) Invasion of Caco-2cells by C. jejuni 11168H grown in the presence (black bars) or absence (white bars) of 100 nM HPA. (C) Apical IL-8 levels in T84 cells at 24 h in response to C.jejuni 11168H, the cj0511 mutant, and the complemented strain grown in the presence (black bars) and absence (white bars) of 100 nM HPA. (D) Translocationof C. jejuni in T84 cells at 6 h. C. jejuni 11168H, a cj0511 mutant, and a complemented strain were grown in the presence (black bars) and absence (white bars) of 100nM HPA. In panels A to C, data are means and SD from three independent experiments. For panel C, one-way ANOVA showed that there was a significant increase inIL-8 secretion in response to infection (P 0.05); the construction of contrasts within ANOVA showed that there was no significant increase in the levels of IL-8 inresponse to any of the bacterial strains grown with pancreatic amylase compared to the same bacteria grown in its absence (P � 0.05). In panel D, data are means and SDfrom three independent experiments for C. jejuni 11168H and the cj0511 mutant. For the cj0511-complemented strain, means and SD from only two independentexperiments are reported, as a result of contamination in the third experiment; thus, no statistical analysis is presented. ***, P 0.0001; *, P 0.05.

FIG 6 Preexposure to pancreatic amylase promotes infection of Galleria mel-lonella larvae. Killing of G. mellonella larvae by C. jejuni grown in the presence(black bars) or absence (white bars) of 100 nM HPA is shown. Ten larvae wereinfected with each strain or with PBS alone. Injection with PBS did not kill anylarvae. The kpsM mutant, the cj0511 mutant, and the complemented strainwere derived from C. jejuni 11168H. Data are means and SD from three inde-pendent experiments. **, P � 0.001.

Jowiya et al.

4890 iai.asm.org December 2015 Volume 83 Number 12Infection and Immunity

on August 8, 2020 by guest

http://iai.asm.org/

Dow

nloaded from

that a polysaccharide matrix is produced. This led us to hypothe-size that a host factor was responsible for inducing the polysaccha-ride EPM. Such interkingdom signaling between eukaryotes andprokaryotes was first identified over a century ago (46). Over thelast decade or so, it has been established that interkingdom signal-

ing between bacteria and hosts via “hormones” is widespread (47).One of the first reports of this was the effect of adrenaline andnoradrenaline on E. coli (48). There are a limited number of re-ports on host cell molecules affecting biofilm formation by bacte-ria. For example, bile salts enhance biofilm formation by C. jejuni

FIG 7 Preexposure to pancreatic amylase promotes colonization of broiler chickens. (A and B) Numbers of birds from which Campylobacter was isolated fromthe ileum at 4 and 7 days postinfection, respectively. C. jejuni was grown with (black bars) or without HPA (white bars). n � 15 birds per group. By day 7, exposureof the WT strain to HPA prior to infection resulted in infection in a greater number of chickens (bar, P � 0.005). (C and D) CFU of Campylobacter per gram ofileal contents in birds 4 and 7 days postinfection, respectively. Open symbols show strains grown without HPA; closed symbols show strains grown with HPA.Significant differences were seen between groups on days 4 (C; P � 0.0006) and 7 (D; P � 0.0026) by the Kruskal-Wallis test. Bars indicate groups between whichdifferences occur, as assessed by Dunn’s multiple-comparison test. n � 15 birds per group.

FIG 8 Preexposure to pancreatic amylase promotes survival in stress environments. (A) Survival at 4°C of C. jejuni 11168H grown with (squares) or without(diamonds) HPA. (B). Survival at 20°C in air of C. jejuni 11168H grown with (squares) or without (diamonds) HPA. Data are means and SD from threeindependent experiments. Statistical significance between the counts in the presence and absence of HPA was assessed using a two-sample t test, assumingunequal variance. **, P 0.01; ***, P 0.001.

C. jejuni Response to Pancreatic Amylase

December 2015 Volume 83 Number 12 iai.asm.org 4891Infection and Immunity

on August 8, 2020 by guest

http://iai.asm.org/

Dow

nloaded from

(26), Listeria monocytogenes (49), and Vibrio cholerae (50). In thecase of Streptococcus pneumoniae, a number of host factors affectbiofilm formation and structure, including norepinephrine (51)and extracytoplasmic ATP (52). Our work shows for the first timethat avian and mammalian pancreatic amylase can act as an in-terkingdom signaling molecule and enhance C. jejuni growth andbiofilm formation. It is important to emphasize that a mucoidcolony is produced on brucella agar in response to pancreaticamylase. This medium does not contain starch, indicating that theresponse is directly to the presence of amylase protein and is notdependent on starch-degrading amylase activity. This was con-firmed in the biofilm assay, which was also carried out in mediumlacking starch. To our knowledge, this is only the second exampleof a proteinaceous host component that can enhance bacterialbiofilm formation, the first being the cytokine interleukin 1 ,which enhances biofilm formation by Staphylococcus aureus (53).Interestingly, C. rectus did not appear to detect pancreatic amy-lase, suggesting that the response may be a specific adaptation tohost environments. It will be interesting to investigate in futurework whether the enteric and oral Campylobacter spp. are able todetect the presence of salivary amylase, given its similarity in se-quence and structure to pancreatic amylase.

The polysaccharide secreted by C. jejuni was definitively iden-tified as �-dextran, and the ease with which it was removed fromthe bacterial cell surface, by simply washing in saline, indicates that itis a loosely associated “slime polysaccharide,” as opposed to thecapsular polysaccharides, which are covalently attached to the cellwall (16, 54). Given that C. jejuni is capable of producing complexglycan structures (18), the production of a simple �-dextran isintriguing. Dextrans are produced by oral streptococci and Leu-conostoc spp. from sucrose by using dextransucrases, cell wall-anchored enzymes of the glycoside hydrolase superfamily (55).Searching the translated C. jejuni genome sequences with knowndextransucrase sequences did not identify homologues, suggest-ing that dextran synthesis in C. jejuni occurs by a novel mecha-nism. The source of the glucose monomer, whether it is taken upfrom the environment or synthesized from amino acids, awaitsexperimental investigation.

The fact that HPA enhances the growth of C. jejuni suggestedthat it may be used as a nutrient source, and Cj0511, predicted tobe a serine protease of the C-terminal protease (CTP) family andknown to be secreted in OMVs (43), seemed a likely candidate fordegrading pancreatic amylase. Our data show that a cj0511 mutantdid not display the response to HPA, a phenotype that was re-stored in a complemented strain, confirming the role of Cj0511specifically. However, there was no apparent change in migrationof recombinant human amylase on an SDS-PAGE gel followingincubation with Cj0511, even though partial casein degradationwas apparent in 30 min and complete digestion in 2 h. The recom-binant amylase is tagged with 6 histidine residues at the C termi-nus, and C-terminal proteolysis of even a single amino acid wouldresult in a reduction of �1 kDa in the molecular mass of theprotein, which would be apparent on a 16% polyacrylamide gel. Inthis context, it is important to emphasize that the polyhistidine-tagged human recombinant amylase preparation was active, sincewe showed that it induced biofilm formation in the wild-typestrain. Thus, interference by the polyhistidine tag cannot explainwhy we fail to see cleavage by r-Cj0511 in vitro. Since Cj0511 isN-glycosylated in C. jejuni (56), an absence of glycosylation inr-Cj0511 is another plausible explanation. However, we have

shown that a pglB mutant (i.e., defective for N-glycosylation) (56)still responds to pancreatic amylase (W. Jowiya and E. Allan, un-published data), which discounts this explanation also. Thus, weconclude that Cj0511 does not digest pancreatic amylase, at leastin vitro. Bacterial CTPs are involved in a range of physiologicalprocesses. The Escherichia coli tail-specific protease processes pen-icillin binding proteins and regulates cell morphology (57). CTPswith roles in virulence have also been described: for example,CtpA of Brucella suis is required for virulence in a mouse infectionmodel (58). A recent study by Karlyshev and coworkers showedthat a cj0511 mutant was severely attenuated in its ability to colo-nize chickens (42). Since Cj0511 does not appear to degrade pan-creatic amylase, we conclude that it is involved at a later stage inthe molecular mechanism; elucidation of its precise role requiresfurther research. Previous studies have shown that Cj0511 is ex-pressed in the absence of pancreatic amylase (43), suggesting thatits expression is constitutive. It will be interesting in future work tosee if Cj0511 levels are increased in the presence of pancreaticamylase.

Having established that pancreatic amylase induces secre-tion of an �-dextran, we next addressed the biological conse-quence. Growth on physiological concentrations of HPA re-sulted in increased biofilm formation in two different WTstrains. Moreover, the increase in biofilm formation inducedby HPA was dependent on Cj0511. Since the purity of the com-mercial HPA preparation is not guaranteed, we tested purifiedrecombinant human pancreatic amylase in the biofilm assayand observed an increase in biofilm formation of similar mag-nitude to that induced by HPA, confirming that the response isto pancreatic amylase specifically. Biofilm formation was alsoincreased in response to an extract of chicken pancreas, sug-gesting that the �-dextran promotes biofilm formation in theintestines of both mammalian and avian hosts.

The study next investigated the effect of HPA exposure on theinteraction of C. jejuni with IECs. Growth on HPA resulted in asignificant increase in adhesion and invasion of C. jejuni in Caco-2cells. Because the neutrophil chemoattractant IL-8 has a role in theresponse to C. jejuni (59), we studied production by T84 cells, asthey are more potent cytokine responders than Caco-2 cells (60).Surprisingly, no difference in IL-8 secretion in response to bacte-ria grown with and those without HPA was observed. Preexposureto HPA led to a significant increase in bacterial translocation 6 hpostinfection, though without loss of tight-junction integrity.Collectively, these data suggest that the �-dextran modulates earlyhost-pathogen interactions by promoting rapid intracellulartransfer while avoiding overt host inflammatory responses. It isintriguing to speculate that an �-1,6-dextran evolved as a coloni-zation factor, since it exhibits sufficient similarity to dietary starchto be tolerated by the immune system yet is resistant to the �-1,4-glucan glucanohydrolase activity of pancreatic amylase.

In the G. mellonella infection model, growth of C. jejuni onpancreatic amylase resulted in increased virulence, and this isagain dependent on Cj0511. G. mellonella larvae have been used asa model to study infection by C. jejuni (36) and other entericpathogens, as they possess both humoral and cellular immunesystems (61). The increased insecticidal activity of C. jejuni grownon HPA suggests that biofilm formation may contribute to resis-tance to these immune mechanisms. It is notable that the insecti-cidal activity observed for C. jejuni 11168H and 81-176 in theabsence of pancreatic amylase was lower than that reported by

Jowiya et al.

4892 iai.asm.org December 2015 Volume 83 Number 12Infection and Immunity

on August 8, 2020 by guest

http://iai.asm.org/

Dow

nloaded from

Champion et al. (36) for these strains. We assume that the de-creased virulence observed in our study is a reflection of culturingon Mueller-Hinton agar as opposed to Columbia blood agar,which was used by Champion et al.

In the laboratory, C. jejuni is sensitive to environmentalstresses, such as drying, chilling, and exposure to atmosphericoxygen, yet the bacterium exhibits a remarkable ability to per-sist in food-processing facilities and natural environments(62). We have shown that preexposure of C. jejuni to pancreaticamylase prolongs survival under atmospheric conditions atambient and refrigeration temperatures. This demonstratesthat the �-dextran, which is expressed by a C. jejuni organismthat is expelled into the environment from a host organism, isstable and may be important for the environmental resilienceof the bacterium.

In conclusion, we propose a new model in which C. jejuni useshost pancreatic amylase, encountered as it passes the pancreaticduct, as a signal for secretion of an �-dextran which promotesbiofilm formation, contributes to colonization of the mammalianand avian intestines, and increases resistance to host immunefunctions. We have shown that Cj0511, a secreted serine protease,is required for the response to pancreatic amylase, although itsprecise function in the response is currently unknown. The dem-onstration of a secreted dextran that leads to biofilm formation isan important finding that is expected to enable the design of in-tervention strategies to reduce the burden of C. jejuni in the foodchain and ultimately to reduce the incidence of human entericdisease.

ACKNOWLEDGMENTS

We are grateful to Duncan Gaskin and Arnoud van Vliet for the gift ofpC46fdxA. We thank Nicky Mordan for help with confocal microscopy,David Boniface for help with statistical analyses, and Peter Mullany forreading the manuscript.

REFERENCES1. Allos BM. 2001. Campylobacter jejuni infections: update on emerging

issues and trends. Clin Infect Dis 32:1201–1206. http://dx.doi.org/10.1086/319760.

2. Gillespie IA, O’Brien SJ, Frost JA, Adak GK, Horby P, Swan AV, PainterMJ, Neal KR. 2002. A case-case comparison of Campylobacter coli andCampylobacter jejuni infection: a tool for generating hypotheses. EmergInfect Dis 8:937–942. http://dx.doi.org/10.3201/eid0809.010817.

3. Garin B, Gouali M, Wouafo M, Perchec AM, Pham MT, RavaonindrinaN, Urbès F, Gay M, Diawara A, Leclercq A, Rocourt J, Pouillot R. 2012.Prevalence, quantification and antimicrobial resistance of Campylobacterspp. on chicken neck-skins at points of slaughter in 5 major cities locatedon 4 continents. Int J Food Microbiol 157:102–107. http://dx.doi.org/10.1016/j.ijfoodmicro.2012.04.020.

4. Humphrey T, O’Brien S, Madsen M. 2007. Campylobacters as zoonoticpathogens: a food production perspective. Int J Food Microbiol 117:237–257. http://dx.doi.org/10.1016/j.ijfoodmicro.2007.01.006.

5. Wilson DJ, Gabriel E, Leatherbarrow AJ, Cheesbrough J, Gee S, BoltonE, Fox A, Fearnhead P, Hart CA, Diggle PJ. 2008. Tracing the source ofcampylobacteriosis. PLoS Genet 4:e1000203. http://dx.doi.org/10.1371/journal.pgen.1000203.

6. Beery JT, Hugdahl MB, Doyle MP. 1988. Colonization of gastrointesti-nal tracts of chicks by Campylobacter jejuni. Appl Environ Microbiol 54:2365–2370.

7. Costerton JW, Stewart PS, Greenberg EP. 1999. Bacterial biofilms: acommon cause of persistent infections. Science 284:1318 –1322. http://dx.doi.org/10.1126/science.284.5418.1318.

8. Edwards LA, Nistala K, Mills DC, Stephenson H, Zilbauer M, WrenBW, Dorrell N, Lindley KJ, Wedderburn LR, Bajaj-Elliott M. 2010.Delineation of the innate and adaptive T-cell immune outcome in the

human host in response to Campylobacter jejuni infection. PLoS One5:e15398. http://dx.doi.org/10.1371/journal.pone.0015398.

9. Haddock G, Mullin M, MacCallum A, Sherry A, Tetley L, Watson E,Dagleish M, Smith DG, Everest P. 2010. Campylobacter jejuni 81-176forms distinct microcolonies on in vitro-infected human small intestinaltissue prior to biofilm formation. Microbiol 156:3079 –3084. http://dx.doi.org/10.1099/mic.0.039867-0.

10. Buswell CM, Herlihy YM, Lawrence LM, McGuiggan JT, Marsh PD,Keevil CW, Leach SA. 1998. Extended survival and persistence of Cam-pylobacter spp. in water and aquatic biofilms and their detection by im-munofluorescent-antibody and -rRNA staining. Appl Environ Microbiol64:733–741.

11. Joshua GW, Guthrie-Irons C, Karlyshev AV, Wren BW. 2006. Biofilmformation in Campylobacter jejuni. Microbiol 152:387–396. http://dx.doi.org/10.1099/mic.0.28358-0.

12. Reeser RJ, Medler RT, Billington SJ, Jost BH, Joens LA. 2007. Charac-terization of Campylobacter jejuni biofilms under defined growth condi-tions. Appl Environ Microbiol 73:1908 –1913. http://dx.doi.org/10.1128/AEM.00740-06.

13. Reuter M, Mallett A, Pearson BM, van Vliet AH. 2010. Biofilm forma-tion by Campylobacter jejuni is increased under aerobic conditions. ApplEnviron Microbiol 76:2122–2128. http://dx.doi.org/10.1128/AEM.01878-09.

14. Sanders SQ, Frank JF, Arnold JW. 2008. Temperature and nutrienteffects on Campylobacter jejuni attachment on multispecies biofilms onstainless steel. J Food Prot 71:271–278.

15. Trachoo N, Frank JF, Stern NJ. 2002. Survival of Campylobacter jejuni inbiofilms isolated from chicken houses. J Food Prot 65:1110 –1116.

16. Branda SS, Vik S, Friedman L, Kolter R. 2005. Biofilms: the matrixrevisited. Trends Microbiol 13:20 –26. http://dx.doi.org/10.1016/j.tim.2004.11.006.

17. Sutherland IW. 2001. The biofilm matrix—an immobilized but dynamicmicrobial environment. Trends Microbiol 9:222–227. http://dx.doi.org/10.1016/S0966-842X(01)02012-1.

18. Guerry P, Szymanski CM. 2008. Campylobacter sugars sticking out.Trends Microbiol 16:429 – 435. http://dx.doi.org/10.1016/j.tim.2008.07.002.

19. Bernatchez S, Szymanski CM, Ishiyama N, Li J, Jarrell HC, Lau PC,Berghuis AM, Young NM, Wakarchuk WW. 2005. A single bifunctionalUDP-GlcNAc/Glc 4-epimerase supports the synthesis of three cell surfaceglycoconjugates in Campylobacter jejuni. J Biol Chem 280:4792– 4802.http://dx.doi.org/10.1074/jbc.M407767200.

20. Hamner S, McInnerney K, Williamson K, Franklin MJ, Ford TE. 2013.Bile salts affect expression of Escherichia coli O157:H7 genes for virulenceand iron acquisition, and promote growth under iron limiting conditions.PLoS One 8:e74647. http://dx.doi.org/10.1371/journal.pone.0074647.

21. Karavolos MH, Winzer K, Williams P, Khan CM. 2013. Pathogenespionage: multiple bacterial adrenergic sensors eavesdrop on host com-munications systems. Mol Microbiol 87:455– 465. http://dx.doi.org/10.1111/mmi.12110.

22. Rothenbacher FP, Zhu J. 2014. Efficient responses to host and bacterialsignals during Vibrio cholerae colonization. Gut Microbes 5:120 –128.http://dx.doi.org/10.4161/gmic.26944.

23. Cogan TA, Thomas AO, Rees LE, Taylor AH, Jepson MA, Williams PH,Ketley J, Humphrey TJ. 2007. Norepinephrine increases the pathogenicpotential of Campylobacter jejuni. Gut 56:1060 –1065. http://dx.doi.org/10.1136/gut.2006.114926.

24. Malik-Kale P, Parker CT, Konkel ME. 2008. Culture of Campylobacterjejuni with sodium deoxycholate induces virulence gene expression. J Bac-teriol 190:2286 –2297. http://dx.doi.org/10.1128/JB.01736-07.

25. Mills DC, Gundogdu O, Elmi A, Bajaj-Elliott M, Taylor PWP, WrenBW, Dorrell N. 2012. Increase in Campylobacter jejuni invasion of intes-tinal epithelial cells under low-oxygen coculture conditions that reflect thein vivo environment. Infect Immun 80:1690 –1698. http://dx.doi.org/10.1128/IAI.06176-11.

26. Svensson SL, Pryjma M, Gaynor EC. 2014. Flagella-mediated adhesionand extracellular DNA release contribute to biofilm formation and stresstolerance of Campylobacter jejuni. PLoS One 9:e1060. http://dx.doi.org/10.1371/journal.pone.0106063.

27. Madhusudhan KT, Mokady S, Cogan U. 1987. Chicken pancreatic en-zymes for clinical use: autoactivation of the proteolytic zymogens. J SciFood Agric 41:187–193. http://dx.doi.org/10.1002/jsfa.2740410212.

C. jejuni Response to Pancreatic Amylase

December 2015 Volume 83 Number 12 iai.asm.org 4893Infection and Immunity

on August 8, 2020 by guest

http://iai.asm.org/

Dow

nloaded from

28. Bernfeld P. 1955. Amylases, � and . Methods Enzymol 1:149 –158. http://dx.doi.org/10.1016/0076-6879(55)01021-5.

29. Enos-Berlage JL, McCarter LL. 2000. Relation of capsular polysaccharideproduction and colonial cell organization to colony morphology in Vibrioparahaemolyticus. J Bacteriol 182:5513–5520. http://dx.doi.org/10.1128/JB.182.19.5513-5520.2000.

30. Dubois M, Gilles KA, Hamilton JK, Rebers PA, Smith F. 1956. Color-imetric method for determination of sugars and related substances. AnalChem 28:350 –356. http://dx.doi.org/10.1021/ac60111a017.

31. Laws AP, Chadha MJ, Chacon-Romero M, Marshall VM, Maqsood M.2008. Determination of the structure and molecular weights of the exopo-lysaccharide produced by Lactobacillus acidophilus 5e2 when grown ondifferent carbon feeds. Carbohyd Res 343:301–307. http://dx.doi.org/10.1016/j.carres.2007.10.028.

32. Gerwig GJ, Kamerling JP, Vliegenthart JFG. 1978. Determination of thed and l configuration of neutral monosaccharides by high-resolution cap-illary g.l.c. Carbohyd Res 62:349 –357. http://dx.doi.org/10.1016/S0008-6215(00)80881-2.

33. Stellner K, Saito H, Hakomori SI. 1973. Determination of aminosugarlinkages in glycolipids by methylation. Aminosugar linkages of ceramidepentasaccharides of rabbit erythrocytes and of Forssman antigen. ArchBiochem Biophys 155:464 – 472.

34. Gaskin DJ, van Vliet AHM, Pearson BM. 2007. The Campylobactergenetic toolbox: development of tractable and generally applicable genetictechniques for Campylobacter jejuni. Zoonoses Public Health 54(Suppl1):101.

35. Edwards LA, Bajaj-Elliott M, Klein NJ, Murch SH, Phillips AD. 2011.Bacterial-epithelial contact is a key determinant of host innate immuneresponses to enteropathogenic and enteroaggregative Escherichia coli.PLoS One 6:e27030. http://dx.doi.org/10.1371/journal.pone.0027030.

36. Champion OL, Karlyshev AV, Senior NJ, Woodward M, La Ragione R,Howard SL, Wren BW, Titball RW. 2010. Insect infection model forCampylobacter jejuni reveals that O-methyl phosphoramidate has insecti-cidal activity. J Infect Dis 201:776 –782. http://dx.doi.org/10.1086/650494.

37. Mattick K, Durham K, Domingue G, Jørgensen F, Sen M, SchaffnerDW, Humphrey T. 2003. The survival of foodborne pathogens duringdomestic washing-up and subsequent transfer onto washing-up sponges,kitchen surfaces and food. Int J Food Microbiol 85:213–226. http://dx.doi.org/10.1016/S0168-1605(02)00510-X.

38. Slaughter SL, Ellis PR, Butterworth PJ. 2001. An investigation of theaction of porcine pancreatic �-amylase on native and gelatinised starches.Biochim Biophys Acta 1525:29 –36. http://dx.doi.org/10.1016/S0304-4165(00)00162-8.

39. Karlyshev AV, Linton D, Gregson NA, Lastovica AJ, Wren BW. 2000.Genetic and biochemical evidence of a Campylobacter jejuni capsular poly-saccharide that accounts for Penner serotype specificity. Mol Microbiol35:529 –541.

40. Kanipes MI, Papp-Szabo E, Guerry P, Monteiro MA. 2006. Mutation ofwaaC, encoding heptosyltransferase I in Campylobacter jejuni 81-176, af-fects the structure of both lipooligosaccharide and capsular carbohydrate.J Bacteriol 188:3273–3279. http://dx.doi.org/10.1128/JB.188.9.3273-3279.2006.

41. Naito M, Frirdich E, Fields JA, Pryjma M, Li J, Cameron A, Gilbert M,Thompson SA, Gaynor EC. 2010. Effects of sequential Campylobacterjejuni 81-176 lipooligosaccharide core truncations on biofilm formation,stress survival, and pathogenesis. J Bacteriol 192:2182–2192. http://dx.doi.org/10.1128/JB.01222-09.

42. Karlyshev AV, Thacker G, Jones MA, Clements MO, Wren BW. 2014.Campylobacter jejuni gene cj0511 encodes a serine peptidase essential forcolonisation. FEBS Open Bio 4:468 – 472. http://dx.doi.org/10.1016/j.fob.2014.04.012.

43. Elmi A, Watson E, Sandu P, Gundogdu O, Mills DC, Inglis NF, MansonE, Imrie L, Bajaj-Elliott M, Wren BW, Smith DG, Dorrell N. 2012.Campylobacter jejuni outer membrane vesicles play an important role inbacterial interactions with human intestinal epithelial cells. Infect Immun80:4089 – 4098. http://dx.doi.org/10.1128/IAI.00161-12.

44. Seal BS, Hiett KL, Kuntz RL, Woolsey R, Schegg KM, Ard M, Stintzi A.2007. Proteomic analyses of a robust versus a poor chicken gastrointesti-nal colonizing isolate of Campylobacter jejuni. J Proteome Res 6:4582–4591. http://dx.doi.org/10.1021/pr070356a.

45. Szymanski CM, Gaynor EC. 2012. How a sugary bug gets through theday: recent developments in understanding fundamental processes im-

pacting Campylobacter jejuni pathogenesis. Gut Microbes 3:135–144. http://dx.doi.org/10.4161/gmic.19488.

46. Smith EF, Townsend CO. 1907. A plant tumor of bacterial orgins. Science25:671– 673. http://dx.doi.org/10.1126/science.25.643.671.

47. Hughes DT, Sperandio V. 2008. Inter-kingdom signalling: communica-tion between bacteria and their hosts. Nat Rev Microbiol 6:111–120. http://dx.doi.org/10.1038/nrmicro1836.

48. Sperandio V, Torres AG, Jarvis B, Nataro JP, Kaper JB. 2003. Bacteria-host communication: the language of hormones. Proc Natl Acad Sci U S A100:8951– 8956. http://dx.doi.org/10.1073/pnas.1537100100.

49. Begley M, Kerr C, Hill C. 2009. Exposure to bile influences biofilmformation by Listeria monocytogenes. Gut Pathog 1:11. http://dx.doi.org/10.1186/1757-4749-1-11.

50. Hung DT, Zhu J, Sturtevant D, Mekalanos JJ. 2006. Bile acids stimulatebiofilm formation in Vibrio cholerae. Mol Microbiol 59:193–201. http://dx.doi.org/10.1111/j.1365-2958.2005.04846.x.

51. Sandrini S, Alghofaili F, Freestone P, Yesilkaya H. 2014. Host stresshormone norepinephrine stimulates pneumococcal growth, biofilm for-mation and virulence gene expression. BMC Microbiol 14:180. http://dx.doi.org/10.1186/1471-2180-14-180.

52. Marks LR, Davidson BA, Knight PR, Hakansson AP. 2013. Interking-dom signaling induces Streptococcus pneumoniae biofilm dispersion andtransition from asymptomatic colonization to disease. mBio 4:e00438 –13.http://dx.doi.org/10.1128/mBio.00438-13.

53. McLaughlin RA, Hoogewerf AJ. 2006. Interleukin-1beta-induced growthenhancement of Staphylococcus aureus occurs in biofilm but not plank-tonic cultures. Microb Pathog 41:67–79. http://dx.doi.org/10.1016/j.micpath.2006.04.005.

54. Cuthbertson L, Mainprize IL, Naismith JH, Whitfield C. 2009. Pivotalroles of the outer membrane polysaccharide export and polysaccharidecopolymerase protein families in export of extracellular polysaccharides ingram-negative bacteria. Microbiol Mol Biol Rev 73:155–177. http://dx.doi.org/10.1128/MMBR.00024-08.

55. Rehm BH. 2010. Bacterial polymers: biosynthesis, modifications and ap-plications. Nat Rev Microbiol 8:578 –592. http://dx.doi.org/10.1038/nrmicro2354.

56. Young NM, Brisson JR, Kelly J, Watson DC, Tessier L, Lanthier PH,Jarrell HC, Cadotte N, St Michael F, Aberg E, Szymanski CM. 2002.Structure of the N-linked glycan present on multiple glycoproteins in theGram-negative bacterium, Campylobacter jejuni. J Biol Chem 277:42530 –42539. http://dx.doi.org/10.1074/jbc.M206114200.

57. Hara H, Yamamoto Y, Higashitani A, Suzuki H, Nishimura Y. 1991.Cloning, mapping, and characterization of the Escherichia coli prc gene,which is involved in C-terminal processing of penicillin-binding protein3. J Bacteriol 173:4799 – 4813.

58. Bandara AB, Sriranganathan N, Schurig GG, Boyle SM. 2005. Carboxyl-terminal protease regulates Brucella suis morphology in culture and per-sistence in macrophages and mice. J Bacteriol 187:5767–5775. http://dx.doi.org/10.1128/JB.187.16.5767-5775.2005.

59. Hu L, Hickey TE. 2005. Campylobacter jejuni induces secretion of proin-flammatory chemokines from human intestinal epithelial cells. Infect Im-mun 73:4437– 4440. http://dx.doi.org/10.1128/IAI.73.7.4437-4440.2005.

60. MacCallum AJ, Harris D, Haddock G, Everest PH. 2006. Campylo-bacter jejuni-infected human epithelial cell lines vary in their ability tosecrete interleukin-8 compared to in vitro-infected primary humanintestinal tissue. Microbiology 152:3661–3665. http://dx.doi.org/10.1099/mic.0.29234-0.

61. Mylonakis E, Casadevall A, Ausubel FM. 2007. Exploiting amoeboid andnon-vertebrate animal model systems to study the virulence of humanpathogenic fungi. PLoS Pathog 3:e101. http://dx.doi.org/10.1371/journal.ppat.0030101.

62. Solomon EB, Hoover DG. 1999. Campylobacter jejuni: a bacterial para-dox. J Food Saf 19:121–136. http://dx.doi.org/10.1111/j.1745-4565.1999.tb00239.x.

63. Karlyshev AV, Linton D, Gregson NA, Wren BW. 2002. A novel paralo-gous gene family involved in phase-variable flagella-mediated motility inCampylobacter jejuni. Microbiology 148:473– 480. http://dx.doi.org/10.1099/00221287-148-2-473.

64. Korlath JA, Osterholm MT, Judy LA, Forfang JC, Robinson RA. 1985.A point-source outbreak of campylobacteriosis associated with consump-tion of raw milk. J Infect Dis 152:592–596. http://dx.doi.org/10.1093/infdis/152.3.592.

65. Manning G, Duim B, Wassenaar T, Wagenaar JA, Ridley A, Newell DG.

Jowiya et al.

4894 iai.asm.org December 2015 Volume 83 Number 12Infection and Immunity

on August 8, 2020 by guest

http://iai.asm.org/

Dow

nloaded from

2001. Evidence for a genetically stable strain of Campylobacter jejuni. ApplEnviron Microbiol 67:1185–1189. http://dx.doi.org/10.1128/AEM.67.3.1185-1189.2001.

66. Karlyshev AV, Wren BW. 2001. Detection and initial characterization ofnovel capsular polysaccharide among diverse Campylobacter jejuni strainsusing Alcian blue dye. J Clin Microbiol 39:279 –284. http://dx.doi.org/10.1128/JCM.39.1.279-284.2001.

67. Linton D, Karlyshev AV, Hitchen PG, Morris HR, Dell A, Gregson NA,

Wren BW. 2000. Multiple N-acetyl neuraminic acid synthetase (neuB)genes in Campylobacter jejuni: identification and characterization of thegene involved in sialylation of lipo-oligosaccharide. Mol Microbiol 35:1120 –1134. http://dx.doi.org/10.1046/j.1365-2958.2000.01780.x.

68. Van Vliet AHM, Wood AC, Henderson J, Wooldridge K, Ketley JM.1998. Genetic manipulation of enteric Campylobacter species. MethodsM i c r o b i o l 2 7 : 4 0 7 – 4 1 9 . h t t p : / / d x . d o i . o r g / 1 0 . 1 0 1 6 / S 0 5 8 0-9517(08)70301-5.

C. jejuni Response to Pancreatic Amylase

December 2015 Volume 83 Number 12 iai.asm.org 4895Infection and Immunity

on August 8, 2020 by guest

http://iai.asm.org/

Dow

nloaded from