Introduction

Food poisoning is an important health issue in several countries including Thailand. Studies in Thailand have shown that acute diarrhea in children <12 years old is commonly caused by Campylobacter sp. (28 %), Salmonella sp. (18 %), Shigella sp. (9 %) and E. c oli (6 %) [1]. A study conducted in Chiang Mai, the northern region of Thailand, revealed that C. jejuni was most commonly found in patients [2]. This species can be transmitted from infected poultry farms along the processing chain to consumers [3]. The farm environment is known to be a potential source of the occurrence of this bacterium in food animals [4]. In northern Europe, C. jejuni has been found in the environment, water resources and livestock keeping areas [5]. Since there have been a variety of sources of infection and the outbreaks usually consisted of sporadic cases, the sources and transmission routes of outbreaks are difficult to trace [6]. In order to determine the source of infection, several molecular typing schemes have been developed for characterizing and tracing Campylobacter spp.

The multilocus sequence typing (MLST) technique for molecular fingerprinting of the Campylobacter genome has one important advantage as compared with other molecular techniques that global MLST databases can be used for identifying the potential source of infection in outbreak investigations [7–11]. MLST is also considered to have a higher discriminatory index than other molecular methods [12].

In 2015, the Campylobacter Multi Locus Sequence Typing website (http://pubmlst.org/campylobacter/) provided access to 7797 profiles of C. jejuni isolated globally from humans [13]. The aim of the present study was to investigate the genetic relationship between C. jejuni isolated from animals and humans in Chiang Mai, Thailand by comparing sequence types (STs) of isolates from the same geographic area and period during 2002–2007.

Material and Methods

Nine C. jejuni isolates from broilers and 12 from children with diarrhea symptoms from the Chiang Mai region, Thailand during the period 2002–2007 were included in the study. Briefly, broiler samples were collected at the slaughterhouse and again at the markets while stool samples from children with diarrhea symptoms were asked by directly contacting for participation while the child was hospitalized [2]. All sampling sites were located in the Chiang Mai of northern Thailand. Isolation and identification of these Campylobacter bacteria was performed by selective culture in microaerophilic atmosphere and polymerase chain reaction [2]. All isolates were kept at −70 °C in cryovial tube. Prior to DNA extraction, cultures were removed from −70 °C and allowed to thaw at room temperature. These were then spread out onto a Brucella agar plate (Oxoid®, UK) to produce discrete colonies and incubated at 37 °C for 72 h under microaerobic conditions with a CO2 generator (Anaerocult® C; Merck, Germany) (5). DNA extraction was performed using the cetyltrimethyl ammonium bromide (CTAB) precipitation method, as described previously [14].

MLST was performed using standard primers as described in Table 1 using selective amplification of the following seven genes and determining their sequence: aspartase A (aspA), glutamine synthetase (glnA), citrate synthase (gltA), serine hydroxymethyl transferase (glyA), phosphoglucomutase (pgmA), transketolase (tktA), ATP synthase subunit (uncA) [9, 10, 15]. PCR reactions were carried out in a final volume of 50 μl, each mixture including: 4 μg of Campylobacter DNA, 0.2 μM of forward and reverse primers, 1X PCR buffer, 2 mM MgSO4, 0.2 mM deoxynucleoside triphosphates, and 1 U of High Fidelity Platinum Taq DNA polymerase (Invitrogen). Fluorescent dye-determinator sequencing was performed using the ABI Prism 3730 XL DNA sequencer method (Bio Basic Inc, Canada). The sequences were submitted to the international databases on the Campylobacter Multi Locus Sequence Typing website (http://pubmlst.org/campylobacter). All allelic profiles were recorded and a sequence type (ST) and clonal complexes (CC) were assigned [16]. In addition, other 42 sequence types of Thai isolates were included for analysis.

Table 1 Primers
Full size table

Comparison of allelic profiles and ST was done using the tools available online at MLST public database (http://pubmlst.org/perl/mlstanalyse/mlstanalyse.pl?site=pubmlst). Phylogenetic trees were generated from allelic profile data based on the neighbor-joining and UPGMA method using the tree drawing tool of the Bionumerics® version 4.0 software (Applied Maths BVBA, Belgium). Trees were displayed using the Phylodendron software (http://iubio.bio.indiana.edu/soft/molbio/java/apps/trees/). STs found in this study and STs of 42 additional isolates of C. jejuni available on the public MLST databases were grouped together using the ‘based upon related sequence types’ (BURST) algorithm version 1.00 (http://pubmlst.org/analysis/burst/burst.shtml) with a stringent group definition (4/7 shared alleles). Isolates in the groups defined by BURST were considered to belong to a single clonal complex [13]. BURST uses a model of bacterial evolution in which an ancestral (or founding) genotype increases in frequency in the population, and diversifies to produce a cluster of closely related genotypes that all descend from the founding genotype. The BURST algorithm first identifies mutually exclusive groups of related genotypes in the population such as MLST, and attempts to identify the founding genotype of each group. The algorithm then predicts the descent from the predicted founding genotype to the other genotypes in the group. ST with at least two assigned descendent single locus variants (SLV) is defined as a subgroup founder [13].

Results and Discussion

Amongst 21 C. jejuni isolates from both chicken and children from Chiang Mai, Thailand during 2002–2007, 16 STs were identified in the present study (Table 2). Of these, 13 STs (80 %) were previously found in Thailand. ST-574 was found both in human and in chicken isolate. ST-2276 and ST-2921 previously found only in humans in Thailand were also identified from chickens (Table 3). ST-2276 seemed to be indigenous to broilers in Thailand as reported in previous studies [17, 18]. ST-1232 and ST-1919 previously isolated from humans elsewhere were found in the present study in chickens.

Table 2 Genetic character of C. jejuni using MLST
Full size table
Table 3 Comparison of the samples’ number in each STs found in the present study and the ST percentage from other sources in the database (February 2012; http://pubmlst.org/campylobacter/)
Full size table

Eleven STs can be grouped into 8 CCs including 13 C. jejuni isolates. One out of 8 CCs (12.5 %) included isolates from both chickens and children with diarrhea while 3 out of 8 CCs (37.5 %) included only isolates from chickens. Five unassigned STs included one from chicken and the others from humans. Additionally, possible routes of transmission of C. jejuni among chicken and human isolates have been determined by Levesque et al. [6]. The study suggested that sporadic C. jejuni infections in humans are frequent due to the sources other than chickens.

The phylogenetic tree (Fig. 1) showed that C. jejuni isolates from chickens and children with diarrhea were clustered together. BURST grouping also showed that 13 out of 21 isolates can be grouped into 3 groups with similarity >40 % in groups 1 and 3, and in group 2 more than 70 %. The majority of isolates that did not belong to any ancestral group were isolated from diarrhea children (6/8).

Fig. 1
figure 1

UPGMA Phenogram of C. jejuni from the northern region of Thailand

Full size image

A total of 63 isolates, 21 isolates from Chiang Mai and 42 from Thai isolates, were analyzed (Table 4). MLST sequences were grouped into 55 STs and 3 groups by BURST. Group 1 consisted of 39 STs from 45 isolates and ST-464 was the central ST representing 6 subgroups. Groups isolated from human samples only were ST-464, ST-2284 and ST-2922. In case of ST-2922, which had ST-2274, this ST could be found in chickens from Belgium and Thailand also [17, 18], but ST-2284 was found in Thai people only. Group 1 and 3 results (Table 4) showed that all samples were isolated from humans including those 8 STs which could not be joined with any of these groups. Group 2 samples were isolates from humans in Thailand only.

Table 4 Burst grouping result of all C. jejun i MLST found in Thailand (n = 63)
Full size table

The results showed genetically diverse STs of C. jejuni isolated from both chickens and children from northern Thailand. Amongst 21 samples with 16 different STs from the northern region of Thailand, ST-2921, ST-2276 and ST-2332 were reported from Thailand [9, 19] in the present study. According to the BURST and CC results, the ST-52 and ST-607 complexes were the most common Campylobacter species in Thailand. In New South Wales, Australia, the most common CCs were ST-48, ST-257, ST-345 and ST-21 complexes [20]. The ST-21 complex rarely found in Thailand, included 2 Campylobacter isolates from children in the northern region. The most commonly found STs in avian species in Europe, USA, Canada and Australia were ST-21 and ST-45 [9], which could not be detected in broilers in Thailand [17, 18]. A study conducted in Australia from January 1991 to July 1993 reported that amongst a group of ST-607 complex isolates from human campylobacteriosis patients some individuals had a travelling history to Thailand where this was one of the most commonly occurring complex [20].

According to BURST analysis results, most of the chicken isolates belonged to subgroup ST-2940. Interestingly, the ST-2921 belonged to ST-2940 subgroup among human samples. The ST-2276 and ST-2921 previously found in humans in Thailand were also found in broiler samples in the present study. Furthermore, ST-4035 which belonged to the ST-2941 subgroup was isolated from a Guillain–Barre syndrome patient. The finding of ST-436 both in human and wild starling bird samples indicated that wild birds can be a source of infection (Table 2). Both data from chicken samples in the present study and those from the database indicated that there might be a source of infection in humans from sources other than broilers and chicken products. Gormley et al. [21] found the genetic character of C. jejuni in human which was different from those in broilers. This situation could be because of poor hygiene behavior of consumers. The infections have occurred during the cooking process or during meals from several contamination sources including wild bird feces, and this could explain the increase genetic changes and genetic variation in some human beings [22].

The similarity between C. jejuni found in broilers and humans suggests the occurrence of transmission between broilers and humans. This indicates that both biosecurity on broiler farms and personal hygiene amongst humans needs to be strengthened. Although it would be rare for the children under 1 year of age to come into direct contact with chicken, there may be indirect transmission, for example by nurses, food with chicken ingredients or by C. jejuni contaminated water or sand [6]. C. jejuni has also been found in the intestine of domestic animals [19]. Appropriate levels of hygiene in child nurseries are essential, it is being neglected since infection in adults usually occurs asymptomatically. C. jejuni infection in infants can result in severe clinical signs including bloody diarrhea and intussusception or necrotizing enterocolitis. About 92 % of the under 1-year-old children with campylobacteriosis had bloody diarrhea and at least 5 outbreaks in the nurseries have been reported. The transmission came from nosocomial infection particularly from the use of medical equipment such as thermometers or bathtubs without an appropriate sanitization [23].

The genetic characteristics of C. jejuni isolates in Thailand showed a weak clonal structure. Further studies are needed to discover more chicken C. jejuni isolates, which are the main source of C. jejuni. It can relate all data grouped in CCs with the ungrouped data [19]. Collection of samples from environmental sources e.g. water resources, soil, sand, wild birds and domestic animals [6] would be helpful for determining other sources of C. jejuni. Furthermore, studies about virulence factors of C. jejuni should be conducted to improve the understanding of the virulence and pathogenesis of the C. jejuni in human beings. Although MLST is a relatively expensive method, it has the advantage that it can be performed easily if given availability of genetic material and appropriate PCR equipment [12]. The utility of MLST for molecular epidemiological analysis depends on the availability of the number of gene profiles that can be used for comparative purposes, and it is therefore essential to submit any sample to internationally accessible databases.

Conclusion

There was an evidence of genetic relationship between C. jejuni isolated from human beings and broiler. MLST analyses of C. jejuni suggested that broiler might be one of the source of human infection in Chiang Mai, Thailand.