Continued widespread dissemination and increased poultry host fitness of Campylobacter jejuni ST-4526 and ST-4253 in Japan

Authors


Correspondence

Hiroshi Asakura, Division of Biomedical Food Research, National Institute of Health Sciences, Kamiyoga 1-18-1, Setagaya-ku, Tokyo 158-8501, Japan. E-mail: hasakura@nihs.go.jp

Abstract

Aims

Campylobacter jejuni is a major cause of foodborne gastroenteritis. We previously reported the widespread Camp. jejuni sequence type (ST)-4526 in Japan from 2005 to 2006. This study assesses the potential for this genotype to thrive thereafter.

Methods and Results

Fifty human Camp. jejuni isolates collected in 2010–2011 in Osaka, Japan, were genotyped by multilocus sequence typing (MLST). This approach identified 22 STs and 11 clonal complexes (CCs), including four novel STs. A comparative analysis to the previous data set showed the predominance of CC-21, in which ST-4526 and ST-4253 represented 39 and 63% in each of the two time frames, indicating their continued widespread presence. These two STs belong to close evolutionary lineages and are also isolated from chicken meat. The superior abilities of ST-4526/ST-4253 representatives to colonize chicken gut were demonstrated by co-infections with ST-21, ST-50 and ST-8 representatives.

Conclusions

Data provide evidence for the continued widespread of ST-4526/ST-4253 among human clinical isolates in Japan. These STs showed adaptive fitness to chicken.

Significance and Impact of the study

This is the first evidence of the continued thriving of ST-4526/ST-4253 in Japan with their increased in vivo fitness. Our findings suggest that poultry mediates the microevolution of this pathogen, thereby enabling these STs to become widespread.

Introduction

Campylobacter jejuni is one of the leading causes of foodborne gastroenteritis in humans worldwide (Friedman et al. 2000). This Gram-negative, microaerophilic bacterium is widely distributed in a variety of host and hostile environments, of which poultry and cattle are considered to be major reservoirs for human infection (Sanad et al. 2011; Hermans et al. 2012). The sanitary control of these habitats is therefore likely to contribute to the prevention of human campylobacteriosis. However, progress on this front has been limited, partly because of our limited understanding of how this pathogen is disseminated and which bacterial populations are most frequently associated with human infection.

This pathogen shows high genomic plasticity (Lefébure et al. 2010; Sheppard et al. 2011), which could be mediated through horizontal gene transfer (Avrain et al. 2004), natural transformation (Vegge et al. 2012), conjugative transfer (Oyarzabal et al. 2007) and bacteriophage predation (Scott et al. 2007). Many efforts focussing on this genomic diversity have attempted to genotype Camp. jejuni isolates from different sources and to trace the sources or routes of human transmission. The methods used have included amplified fragment length polymorphism (AFLP), flaA short variable region (SVR) sequencing, pulsed-field gel electrophoresis (PFGE) and multilocus sequence typing (MLST) (Ahmed et al. 2012). The MLST approach, which relies on the nucleotide sequence variation of seven Camp. jejuni housekeeping genes, has been one of the most powerful genetic tools for revealing the population structures of this pathogen in many industrialized countries (Dingle et al. 2001; Sails et al. 2003; Lévesque et al. 2008; Korczak et al. 2009).

We recently used the MLST approach to report the population genetics of Camp. jejuni isolated from humans, poultry and cattle in Japan from 2005 to 2006 (Asakura et al. 2012). Comparative genome analyses further provided possible explanations for the wide distribution of ST-4526 in both human and poultry populations (e.g. increased cell adhesion, decreased DNA uptake/recombination and fluoroquinolone/nalidixic acid resistances) (Asakura et al. 2012). However, it was unclear whether this and the related genotype(s) might remain widespread continued in Japan after years 2005–2006, especially among the human population.

In this study, we collected MLST data for human clinical isolates from 2010 to 2011, most of which had associated epidemiological data about the infected individuals (e.g. causative foods). Comparative population genetic analysis between the two time frames revealed that ST-4526 and the related ST-4253 continued to be widespread among the human population in Japan. Through phenotypic and macrogenotypic analyses (serotype, antibiotic resistance, pulsed-field gel electrophoresis patterns and chicken infection assays), we demonstrated the close evolutionary lineage of and increased poultry host adaptation to ST-4253 and ST-4526. Finally, we discussed how these genotypes established such a thriving population in Japan.

Materials and methods

Bacterial strains and media

Human clinical isolates of Campylobacter jejuni (n = 50, designated as H_0101 to H_0150) were randomly selected from 111 Camp. jejuni isolates originated from a total of 46 foodborne infections in Osaka prefecture, Japan, between 2010 and 2011. These isolates were obtained from the patient's faecal samples and were biochemically confirmed to be Camp. jejuni according to the ISO10272 method (International Organization for Standardization 2006). The bacteria were routinely grown on Mueller-Hinton (MH) agar or in MH broth (BD Bioscience, Franklin lakes, NJ, USA) at 37°C under humidified microaerophilic conditions using an AnaeroPack-Microaero gas system (Mitsubishi Gas Chemicals, Tokyo, Japan).

Multilocus sequence typing (MLST) analysis

The bacterial DNA was extracted with a DNA tissue extraction kit (Qiagen, Hilden, Germany) and subjected to MLST analysis as described previously (Asakura et al. 2012). The sequences obtained were aligned with CLC DNA workbench software equipped with the MLST module (CLC Bio, Aarhus, Denmark) according to the manufacturer's instructions. The MLST profiles were deposited into the Campylobacter MLST database (http://www.pubmlst.org/campylobacter) to ascertain sequence types (STs)/clonal complexes (CCs) and to obtain isolate IDs according to the guidelines provided by the website. The seven-locus allele numbers were then used to construct the minimum spanning tree (http://pubmlst.org/perl/mlstanalyse/mlstanalyse.pl?site=pubmlst&page=mst&referrer=pubmlst.org). The FST fixation index of each CC was calculated by the use of the 3309-bp concatenated sequences with DnaSP ver. 5.10.1 program (Rozas et al. 2003) accordingly to estimate the gene flow between the two time frames (2005–2006 vs 2010–2011).

Antibiotic susceptibility test, serotyping and PCR detection of tetO gene

Disk diffusion tests using Sensi-Disk (BD Bioscience) were performed for the antibiotics norfloxacin (NFLX, 10 μg), ofloxacin (OFLX, 5 μg), ciprofloxacin (CPFX, 5 μg), nalidixic acid (NA, 30 μg), erythromycin (EM, 15 μg), tetracycline (TC, 30 μg) and gentamicin (GM, 10 μg) according to the manufacture's instruction. The PCR detection of the tetO gene from total and plasmid DNA prepared with Wizard plus SV minipreps DNA purification systems (Promega, Fitchburg, WI, USA) was performed as previously described (Ng et al. 2001). Penner serotyping was performed as described (Penner and Hennessy 1980) with a commercially available set of antisera (Denka-Seiken Co. Ltd., Tokyo, Japan). Lior serotyping was performed as described previously (Woodward and Rodgers 2002).

Pulsed-field gel electrophoresis (PFGE)

PFGE analysis was performed for all ST-4526 (= 4) and ST-4253 isolates (n = 6) in the data set from 2010 to 2011 using the CHEF Mapper system (Bio-Rad Laboratories, Hercules, CA, USA) as described (Asakura et al. 2012). The obtained gel images were scanned to construct unweighted pair groups (UPGMA) with arithmetic averages using the FingerPrinting II software (Bio-Rad Laboratories).

Chicken colonization assays

The kanamycin (KM)-resistant GFP expression plasmid pWM1007 (Miller et al. 2000) was introduced into the isolates H_0102 (ST-4253) and H_0105 (ST-4526) by electroporation as previously described (Asakura et al. 2007), and these strains were then designated as H_0102KM or H_0105KM, respectively. Two-week-old white leghorns free of specific pathogens (n = 5 per group) obtained from Nisseiken, Tokyo, Japan, were infected orally with approximately 8 log10 of C. jejuni cells in mixed cultures, consisting of almost equal numbers of H_0102KM (ST-4253), H_0112 (ST-21), H_0126 (ST-50) and H_0106 (ST-53) (for group I) or of H_0105KM (ST-4526), H_0112 (ST-21), H_0126 (ST-50) and H_0106 (ST-53) (group II). The H_0112 (ST21), H_0126 (ST-50) and H_0106 (ST-53) isolates were confirmed to be sensitive to KM and other antimicrobials tested in this study (see Table S1 for more detailed information). The two groups were separately fed ad libitum. At 1 and 2 weeks postinfection (p.i.), the animals were sacrificed, and their caeca were aseptically removed, weighed and homogenized in ninth volumes of sterile phosphate-buffered saline (PBS) (Life technologies, Carlsbad, CA, USA). The bacterial suspension and its serial dilutions were then plated onto mCCDA agars (Oxoid, Hampshire, UK) supplemented with or without KM (200 μg ml−1) to count both KM-resistant and total Camp. jejuni (we confirmed that all isolates except H_0102KM and H_0102KM were sensitive to KM). Simultaneously, the agar plates without KM were fluorescently scanned using a Molecular Imager Fx (Bio-Rad Laboratories) to estimate the per cent of the population that was composed of GFP-positive cells. Twenty colonies were randomly selected from the KM-supplemented agar and subjected to MLST profiling to confirm isogenies during the in vivo colonization. The above animal experiments were performed according to the guidelines for animal care and use in National Institute of Health Sciences, Japan.

In vitro growth assay

Cells from the isolates used in the above in vivo infection assays (1·0–1·4 × 105) were independently grown in MH broth at 37°C with shaking (150 rpm) under microaerophilic conditions. The turbidity of the cultures was measured chronometrically at an optical density of 600 nm at 24-h intervals for up to 72 h after incubation. Simultaneously, the same numbers of each isolate were co-cultured as groups I and II and then incubated in MH broth for 72 h, at which point 50-μl aliquots were incubated in 10 ml of fresh MH broth for an additional 72 h (we confirmed that these incubation periods were sufficient for reaching turbidity to plateau ranging from 1·3 to 1·5 of OD595). After the repeated passaging of these cells for a total of 12 days, the growth of the pWM1007-harbouring isolates (H_0102KM or H_0105KM) was measured as a percentage of the population (by plating the co-culture onto MH agars followed by fluorescent scanning as described above).

Statistics

Significant differences for the in vivo colonization or the in vitro growth fitness of ST-4526/4253 representatives were calculated using Student's t-test for at least three independent assays.

Results

Population dynamics of Campylobacter jejuni human isolates in Osaka, Japan between the periods 2005–06 and 2010–11

In this study, we collected 50 representative human clinical isolates in Osaka, Japan, in 2010-2011 (H_0101 to H_0150), which were used to analyse Camp. jejuni population genetics based on MLST. We identified that 44 of which were associated with nine clonal complexes (CCs). There were six singletons, including three novel sequence types (ST) (Tables 1 and S1). CC-21 was present in 36·4% (16/44), CC-22 was present in 27·3%, and CC-42/443/574/61 were present in 6·8% of the isolates, respectively (Fig. 1a). A comparison with our previous data set (n = 50, H_0051 to H_0100, which were human clinical isolates in Osaka, in 2005–2006; see Asakura et al. 2012 for more details) showed that CC-21 remained the most prominent clonal complex (Fig. 1a). The abundance of CC-22 increased from 12·0% (of the 2005–2006 data set), whereas CC-48 decreased from 14·0% (Fig. 1a). The FST values indicated no significant population differences in these CCs between the two time periods (Table 1). Within CC-21, ST-4526 (n = 4) and ST-4253 (n = 6) were predominant (Fig. 1b). The serotype analysis showed that all of the ST-4526/ST-4523 isolates were serotyped to O:2, and this serotype was restricted to within CC-21 (Tables 1 and S1). The second most abundant clonal complex, CC-22, included 11 ST-22 isolates (22%), 9 of which serotyped to O:19 (Tables 1 and S1). Together, these data indicate that ST-4526 has been continuously distributed in the human population in Osaka, Japan, between 2005–2006 and 2010–2011.

Table 1. Summary of Campylobacter jejuni isolates in Osaka, Japan, in 2005–2006 and 2010–2011
CC (% total; F STa)STb2005–20062010–2011
No. isolate% resistancecDominant serotypedNo. isolate% resistancecDominant serotype
NA/FQsTCNA/FQsTC
  1. Frequency of CCs/STs detected in two human clinical data sets obtained in 2005–2006 and 2010–2011 in Japan.

  2. a

    FST values were calculated in each CC that contained >2 isolates between the two time frames.

  3. b

    Novel STs are shown in bold.

  4. c

    % resistance of antimicrobials are shown per STs.

  5. d

    UT represents untypeable. No. of isolates is shown in parentheses.

CC-21 (34%; 0·02696) 18   16   
ST-45266100100O: 2 (6)4100100O: 2 (4)
ST-425311000O: 2 (1)61000O: 2 (6)
ST-5054060O: 2 (3)25050O: 2 (2)
ST-2136733O: 2 (3)100O: 2 (1)
ST-451100O: 2 (1)O: 2 (1)
ST-810100O: 2 (1)O: 2 (1)
ST-4614100O: 2 (1)O: 2 (1)
ST-53200
ST-2931100100
CC-22 (18%; 0·00022) 6   12   
ST-226170O:19 (5)119127O:19 (9)
ST-5800 11000UT (1)
CC-443 (9%; 0·00186) 3   3   
ST-5799  10100:37 (1)
ST-511100100O: 37 (1)1100100O: 37 (1)
ST-4402050O: 37 (2)1100100O: 37 (1)
CC-48 (8%) 7   1   
ST-484025O: 4 (2)100O: 4 (1)
ST-381 10UT (1) 
ST-4531010UT (1) 
ST-91811000UT (1) 
CC-42 (8%; 0·00029) 5    3  
ST-42400O: 23 (2)3033O: 23 (2)
ST-447100  
CC-45 (4%; 0) 2   2   
ST-4525050O: 12(1)/55 (1)25050O: 12(1)/55(1)
CC-574 (3%) 0   3   
ST-30502100100O: 5 (1)
ST-2031011000UT (1)
CC-61 (3%)0  3     
ST-61300O: 4 (2)
CC-52 (1%)1   0    
ST-521100100 
CC-354 (1%) 1   0   
ST-354100 
CC-257 (1%) 0   1   
ST-2571100100O: 11 (1)
Unassigned (13%,0·08306) 7   6   
ST-2328100 333100UT (3)
ST-922300 10100O: 3 (1)
ST-5801 11000UT (1)
ST-5802 10100O: 18 (1)
ST-40725050 
ST-439010100 
Total 503636 506442 
Figure 1.

Population dynamics of human Campylobacter jejuni isolates in Japan over time. (a) Frequency distribution of the Camp. jejuni clonal complexes (CCs) isolated from humans in Osaka, Japan, in 2005–2006 (open bars) and 2010–2011 (closed bars) (n = 50 each), which were classified into 11 clonal complexes (CCs). *UA represents isolates unassigned to any CC. (b) Frequency of sequence types (STs) within CC-21 is shown in panel (a). Novel STs are shown in bold. (image_n/jam12147-gra-0001.png) 2005–2006; (image_n/jam12147-gra-0002.png) 2010–2011.

ST-4526 and ST-4253 stably acquire antimicrobial resistance

There is mounting evidence that the acquisition of certain antibiotic resistance increases the fitness of this pathogen and potentially enhances its ability to colonize hosts (Luo et al. 2005; Lin and Martinez 2006). In particular, fluoroquinolone (FQ) resistance is a matter of concern in the public health field (Perez-Boto et al. 2012). We therefore characterized the antimicrobial resistances of the human clinical isolates with a disk diffusion test. Among them, 12 of 16 CC-21 isolates (75%) exhibited resistance to nalidixic acid (NA) and FQs, and all of the ST-4526 and ST-4253 isolates were included in this group (Table 1). All of the ST-4526 isolates also exhibited tetracycline (TC) resistance. 10 of 11 ST-22 isolates that were consisted of a total of 12 CC-22 isolates also exhibited resistance to NA and FQs, whereas all of the isolates of CC-42 (n = 5) and CC-61 (n = 6) were susceptible to all of the antimicrobials in the data set from 2005 to 2006 (Table 1). Similar trends were observed in the data set on 2005–2006 (Asakura et al. 2012), except for relatively high yield of NA/FQs-susceptible ST-22 isolates (five of six isolates, Table 1). The differences in TC resistance between ST-4526 and ST-4253 corresponded to the presence of the plasmid-mediated tetO gene (Fig. S1). The frequent antimicrobial resistance observed in ST-4526 and ST-4253 led us to focus on their evolutionary lineages.

Evolutionary lineage and macrogenotype of ST-4526 and ST-4253

As ST-4526 and ST-4253 are both widespread, their evolutionary relatedness was investigated using minimum spanning tree analysis based on the MLST allele profiles. This revealed the close lineage of the two STs within our combined data set for the periods 2005–2006 and 2010–2011 (Fig. 2a). To further examine the genetic variety within the ST-4526 and ST-4253 isolates, pulsed-field gel electrophoresis (PFGE) analysis was also conducted on the ST-4526 (n = 4) and ST-4253 isolates (n = 6) collected in this study. This analysis indicated that all of the ST-4526 isolates exhibited identical PFGE patterns, whereas six ST-4253 isolates exhibited strain-dependent variations (Fig. 2b). Thus, these data indicate that there is higher genomic similarity within ST-4526 than within ST-4253.

Figure 2.

Phylogenetic analyses of human Campylobacter jejuni isolates in Japan. (a) Minimum spanning tree of Camp. jejuni MLST profiles from the combined data sets for 2005–2006 and 2010–2011. ST numbers are shown in the circles. (b) Pulsed-field gel electrophoresis (PFGE) analysis of four ST-4526 and six ST-4253 isolates in 2010–2011 and two ST-4526 isolates (H_0060 and H_0097) in 2005–2006. Samples were SmaI-digested and loaded onto a 1% agarose gel. The UPGMA dendrogram divided these isolates into four groups, with a cut-off value of 70% similarity.

Increased fitness of ST-4526 and ST-4253 in the chicken gut

An epidemiological survey (questionnaire by healthcare centres at an average of 5·4 days after infection, which ranged from 1 to 12 days of periods, data not shown) revealed that the infected individuals from whom all of the ST-4526 and ST-4253 strains were isolated ate raw or undercooked chicken meat (Table S1). Considering the high population frequency and close evolutionary lineage of ST-4526 and ST-4253, we hypothesized that the poultry host might amplify those STs in the gut, thereby mediating their dissemination. To confirm this idea, we examined the colonization fitness of the representative isolates of the two STs (H_0102KM (ST-4253) and H_0105KM (ST-4526), in which the KM resistance plasmid pWM1007 was introduced) in chicken intestines by co-infecting the birds with other representative CC-21 isolates, including H_0112 (ST-21), H_0126 (ST-50) and H_0106 (ST-53) (which were sensitive to KM/FQs/NA/TC, data not shown). As shown in Fig. 3a, at 1 week p.i., plate counts showed that the KM-resistant population made up 41% or 27% of the total number of Camp. jejuni (1·25 ± 0·34 × 107 of 3·05 ± 1·02 × 10CFU g−1, or 1·05 ± 0·32 × 107 of 3·94 ± 0·81 × 107 CFU g−1), and this increased to 74% or 78% (4·79 ± 0·96 × 107 of 6·49 ± 1·58 × 107 CFU g−1, or 7·46 ± 0·83 × 107 of 9·52 ± 1·34 × 107 CFU g−1) at 2 weeks p.i. in group I (H_0102KM+other) and group II (H_0105KM+others), respectively (Fig. 3a). Despite the difference in absolute colonization levels between the two groups, in both experiments the fraction of GFP-expressing bacteria (ST-2453 in group 1 and ST-4526 in group II) increased from 25% in the inoculum to over 70% 2 weeks p.i. (Fig. 3a). Fluorescent scanning also showed an increase in the percentage of the GFP-positive population with no significant differences between groups I and II (Fig. 3a). Despite the increased in vivo fitness of the two representatives assayed, all of the tested isolates showed similar growth rates [without significant differences (Fig. 3b)], and the H_0102KM (ST-4253) and H_0105KM (ST-4526) isolates did not significantly out-compete the others under in vitro co-culture conditions (Fig. 3c). Together, these results demonstrated that the ST-4526/ST-4253 representatives showed increased fitness over time in the chicken gut.

Figure 3.

ST-4526 and ST-4253 representatives exhibit increased chicken colonization and transmission. (a) Results of two independent colonization experiments, in which two groups of chickens were colonized with a mixture of four Campylobacter strains. The mixture included either 25% bacteria of a ST-4253 strain (Group I) or of an ST-4526 strain (Group II) that contained a plasmid conferring KM resistance as well as GFP expression. Colonization levels obtained after 1 and 2 week p.i. are shown as absolute counts (total counts in black, KM-resistant counts in white) and as relative proportion of GFP-expressing bacteria (grey). (b) In vitro growth kinetics of each isolate were determined by turbidity assays (OD 600 nm) at 24, 48 and 72 h incubation periods. (c) The extent of growth for H_0102KM (ST-4253) or H_0105KM (ST-4526) during co-incubation in MH broth. Percentage population of GFP-positive cells in each groups (dotted and open bars represent Group I and II, respectively) at every 72 h (3 days) postincubation are shown. (image_n/jam12147-gra-0003.png) H_0102KM (ST-4253); (image_n/jam12147-gra-0004.png) H_0105KM (ST-4526); (image_n/jam12147-gra-0005.png) H_0112 (ST-21); (image_n/jam12147-gra-0006.png) H_0126 (ST-50); (image_n/jam12147-gra-0007.png) H_0106 (ST-53).

Discussion

We analysed the population genetics, antimicrobial resistance and serotypes of 50 human clinical Camp. jejuni isolates collected from 2010 to 2011 in Osaka, Japan. A comparison of the MLST data set to that from 2005 to 2006 confirmed that ST-4526 and ST-4253 continued to thrive. Macrogenotypic and phylogenetic analyses suggest a close evolutionary lineage between ST-4526 and ST-4253. The in vivo colonization assay further demonstrated that these isolates possessed superior abilities to colonize the chicken intestine, outcompeting the ST-21, ST-50 and ST-53 isolates, suggesting that the poultry host might mediate the widespread dissemination of those STs among poultry populations, thereby increasing their occurrence in human isolates from individuals with campylobacteriosis in Japan.

The widespread genotypes ST-4526/ST-4253 showed resistance to NA/FQs. Ten of 11 ST-22 isolates also exhibited resistance to these antimicrobials. Considering that these 3 STs made up 40% of the new MLST collection (20 of 50 isolates), it is likely that this sort of antimicrobial resistance could contribute to the continuous wide dissemination of these STs in Japan. Previous work reporting the enhanced in vivo fitness of FQ-resistant Camp. jejuni to antibiotic selection pressure (Luo et al. 2005) supports this idea.

In previous work on serotype distribution, Saito et al. (2005) reported that the serotype O:2 and O:4 complex was frequently distributed in human, poultry and bovine isolates, while serotypes O: 23, 36, 53 were common in human and bovine isolates in Japan. More recently, Harada et al. (2009) demonstrated the predominance (32·6%) of these serotypes among 601 cattle and poultry isolates. Our data in this study further support the predominance of these serotypes and their association with CC-21 and CC-48 in human isolates in Japan. In particular, evidence for the relatively high yields of ST-4526 and ST-4253 within serotype O:2 (43·8%) of CC-21 of human isolates, which were also distributed among poultry isolates that were obtained from live birds fed at distinct area, Aichi and Yamaguchi prefectures in Japan (Asakura et al. 2012), further suggests that poultry might mediate the widespread of these STs in the human population in Japan.

Camp. jejuni is known to exhibit high genomic plasticity between strains (Hepworth et al. 2007), and this plasticity is likely to be mediated by passage through animal hosts (de Boer et al. 2002). A more recent report from Hänel et al. (2009) added evidence that passage through the chicken gut altered PFGE patterns and flaA-sequences of multiple Camp. jejuni strains, with partial co-incidence of phenotypic alterations such as cell adhesion, motility and morphology. The differential PFGE patterns of the 6 ST-4253 isolates in this study thus indicate their in vivo passage backgrounds. In contrast, the identical patterns of the four ST-4526 isolates demonstrate the stable genomic features of this genotype, a finding that is also supported by comparative genomics (Asakura et al. 2012). Nevertheless, the tetO gene-mediated TC resistance of ST-4526 isolates also suggests that this genotype may acquire this feature during passage through an animal host because this gene could be horizontally transferred between Camp. jejuni in the chicken gut (Avrain et al. 2004). Thus, these results suggest that poultry is important not only as a reservoir for human infection but also as a mediator for the microevolution of this pathogen.

Given the public health concerns and commercial importance of Camp. jejuni as a zoonotic pathogen, it is somewhat surprising how little is known about the routes and dynamics of its colonization of chickens. Although many poultry flocks appear to be dominated by a single strain (Ring et al. 2005), two competing variants (Jacobs-Reitsma et al. 1995; Berndtson et al. 1996) and even greater multiplicities of strains (Thomas et al. 1997; Hiett et al. 2002) have been isolated from the same flock. Considering that this pathogen can be transmitted from bird to bird immediately after infection (Shanker et al. 1990), the superior colonization abilities of ST-4526 and ST-4253 in the chicken gut suggest that poultry hosts might amplify those STs, increasing the odds that they will be transmitted to humans. In addition, a recent study by Bereswill et al. (2012) showed that intestinal microbiota significantly modulate host innate immunity to alter Camp. jejuni colonization in vivo. As little is known about the dynamics of chicken microbiota in response to Camp. jejuni colonization, future work on the interactions among Camp. jejuni, intestinal microbiota and innate immunity in poultry hosts might improve our understanding of how this pathogen can establish longitudinal colonization in reservoirs that can then infect humans.

In summary, this study is the first to show that ST-4526 and the related ST-4253 continue to thrive in the human Camp. jejuni population in Japan. The results of genetic and phenotypic analyses provide possible reasons for the widespread dissemination of ST-4526 and ST-4253. Future study of the molecular basis underlying the superior host (poultry) adaptation of these sequence types will provide further insight into the link between the microevolution and ecological features of this pathogen, thereby helping to prevent transmission to humans.

Acknowledgements

This work was supported in part through funding from a Grant-in-Aid of Scientific Research (22780275) from the Japan Society for the Promotion of Science (JSPS) and grants from the Ministry of Health, Labour and Welfare in Japan (H24-shokuhin-ippan-008, H24-shokuhin-ippan-009). The authors declare no conflicts of interest.

Ancillary