Genomic insights into zoonotic transmission and antimicrobial resistance in Campylobacter jejuni from farm to fork: a one health perspective

Compared to other gastrointestinal pathogens, Campylobacter is highly infectious with an infective dose of 500 to 800 organisms for C. jejuni [13]. Although most of the infections caused by C. jejuni to humans are observed as isolated cases, outbreaks can take place [14]. Campylobacter infection varies from a self-limiting disease to serious extraintestinal infection. The most frequent clinical symptoms of campylobacteriosis are unspecific including fever, abdominal cramps, general malaise, muscle pain, diarrhea, and acute uncomplicated enterocolitis. Chronic gastrointestinal complications of Campylobacter infection include irritable bowel syndrome (IBS), inflammatory bowel disease, functional dyspepsia, and colitis [15]. While the disease is typically mild in healthy adults, severe and extended course of the disease, including bacteriemia, is a potential threat to young children, elderly adults, and immunocompromised patients [16].

Campylobacter infection and post-infection complications are quite rare. Nevertheless, many of these complications have a worse prognosis than the acute disease itself. The intensity of symptoms is thought to be affected by co-infection with another foodborne bacteria [17]. The most important extraintestinal complications are bacteremia, meningitis, hepatitis, endocarditis, and pulmonary infection [9]. The evolution of Campylobacter infections to a critical systemic illness, resulting in sepsis and death, is remarkably uncommon with a case-fatality rate of 0.05 in every 1000 infections [18]. Guillain–Barre syndrome (GBS) is a rare autoimmune neurological disorder in which peripheral nerves are demyelinated [19]. C. jejuni infection is the most common preceding infection and was reported in about 30% of GBS cases [19]. Specific C. jejuni serotypes have been associated with an increased risk of GBS (capsular types HS19, HS2, HS41, HS1/44c, HS4c, HS23/36c) [19]. Not all patients with cross-reactive antibodies develop neurologic manifestations. This can be explained by host determinants of post-Campylobacter GBS, particularly human lymphocyte antigen type [20]. The clinical isolate 81–176 of C. jejuni may be involved in development of colorectal cancer due to the cytolethal distending toxin [21].

Campylobacter is primarily a zoonotic disease-causing bacterium [13]. Poultry are the main natural reservoirs, especially for C. jejuni, with a cecal content of up to 1 × 10⁸ CFU/g [22]. By contaminating the carcass and surviving processing in slaughterhouses, C. jejuni can be transferred to humans via undercooked chicken meat or through cross-contamination of other foodstuffs at the kitchen [23]. Contaminated water, milk, and dairy products are other sources of infection [24]. In addition, C. jejuni and other Campylobacter species are frequently found in other animal reservoirs, in up to 90% of cattle, 85% of pigs, and 17.5% of sheep and goats [9]. Cattle isolates are usually clustered in C. jejuni clonal complexes (CC) CC-21, CC-45, CC-48, CC-42, CC-61 and CC-206 [25].

Although cross contamination of food represents the most common source to contact Campylobacter, different reservoirs of human campylobacteriosis may potentially play a role in the epidemiology and transmission of Campylobacter. Other unconventional routes of transmission of Campylobacter spp. are surface water, pets, and wild birds [26]. Surface water accounts for a significant number of human cases. Contamination of surface water with wild animal feces and agricultural waste makes water a collection vessel of different Campylobacter strains from various hosts [26]. Pets were found to cause considerable number of human cases where the transmission of Campylobacter can be bi-directional from owners to pets and vice versa. Pets may acquire infection in parallel with their owners from a common source [26]. While a genomic characterization and profiling of C. jejuni showed a partial overlap between isolates from livestock, pets, and clinical cases, isolates from pets showed specific genomic profiles. Thus, pets can be a potential reservoir for C. jejuni [27]. Wild birds acquire C. jejuni from contaminated water, refuse dumps, and waste from animal farms, pets, and humans [28]. Their body temperatures, in addition to foraging and breeding habits, enable wild birds to be potential reservoirs and spreading routes of Campylobacter [28]. Some studies detected antibiotic resistant C. jejuni in wild birds in many geographical locations, which poses a concern in urban areas and agricultural farms with increased wild birds population [29]. Other studies showed that food and human C. jejuni isolates differed from those of wild birds [30]. Other environmental habitats such as soil are directly or indirectly implicated in human campylobacteriosis [31]. Intriguingly, Campylobacter can be also transmitted through flies to chicken flocks and possibly to humans [32].

Although it has strenuous growing conditions in the lab, C. jejuni acquired resistance to a plethora of stressors, such as low pH, temperature variability, oxidative stress, osmotic pressure, and antimicrobials [33]. These mechanisms enable C. jejuni to survive and transmit between diverse hosts [33]. Tolerance to environmental stressors is frequent among disease-causing lineages and can result in more adapted isolates, with an impact in the general epidemiology of C. jejuni [33]. Among the resistance mechanisms to environmental stressors are the transformation into a viable non culturable state (VBNC), biofilm formation, and mutations specific to certain lineages [33, 34]. Some of the host-generalist and host-specific strains of C. jejuni were proved to survive in aerobic conditions and under oxidative stress [35].

Pan-genome is a term used to describe the entire gene collection identified in a species. The term encompasses two classes: core genome and accessory genome [36]. The core genome represents the set of genes present in every isolate of the species and carries out the necessary cellular functions [36]. The accessory genome constitutes the variable dispensable genome acquired for adaptation and is present only in a few strains or even unique to one strain. Pan-genomic studies highlight the marked variations of bacterial genomes between different genera and species and even between different strains of the same species. These variations can be referred to as genomic plasticity [37].

Almost all bacterial genomes have mosaic structures that are assembled from different DNA segments during evolution and adaptation [38]. The exchange of DNA between bacterial cells occurs via horizontal gene transfer (HGT) either by conjugation, transformation, or transduction [38]. While plasmids and conjugative transposons mediate conjugation, phages that infect bacterial cells mediate transduction [38]. Transformation, on the other hand, occurs when naturally competent bacteria take up extracellular DNA from the environment [38]. The genome of C. jejuni is relatively small, however, it is characterized by high variation even at the strain level [31]. C. jejuni is naturally competent as it uses a DNA uptake system called type II secretion system to transport foreign extracellular DNA to the cytoplasm [5]. To integrate the homologous DNA into the chromosome, C. jejuni uses RecA recombinase, a protein that promotes homologous recombination [5].

A recent study elucidated the role of the chicken gut environment, particularly that of the ceca, in providing suitable conditions for recombination to occur [39]. The results suggested that increased HGT in chicken gut promotes the genetic diversity and hence the adaptability of C. jejuni to the constantly challenging gut environment [39]. The exchange of DNA between bacterial cells contributes to bacterial adaptation to a wide range of environmental conditions and to the colonization of multiple niches. Moreover, it plays an essential role in the evolution of antibiotic resistance and bacterial virulence [38]. When comparing the pattern of genetic variations observed in human pathogenic isolates to that of poultry isolates belonging to the same clonal complex, evidence is further supporting that host-specific mutations develop within certain hosts [5]. Although host specialists are mainly found in only one host species while host generalists are commonly associated with multiple hosts, host specialists have been recently shown to infect more than one definite host [40, 41]. This host adaptation was recently supported by Nennig et al. by implementing different typing schemes based on WGS gene-by-gene approach. They concluded that some C. jejuni lineages have clonally expanded and can colonize or infect multiple hosts as they show adaptation to different niches [42]. Remarkably, a recent study has found that host generalist lineages are better equipped to withstand hostile environmental conditions compared to host specialists, but this needs to be further characterized at the molecular level [33]. To provide better insights into the emergence of generalists, Woodcock et al. investigated the role of genomic plasticity in the coexistence of generalist and specialist Campylobacter lineages. They concluded that the ecological generalism observed in some C. jejuni isolates reflected their genotypic and phenotypic plasticity and resulted in their rapid host adaptation in different host environments [37].

As mentioned in the previous section, certain lineages of C. jejuni are specific to a particular host species that is related to host adaptation [40‐43]. Another interesting example is the recent study conducted by Parker et al. where two strains of C. jejuni colonizing guinea pigs were compared with well characterized Campylobacter strains [44]. They found that isolates from guinea pigs were of novel sequence type, distinct from other known Campylobacter strains, and had genes gain and loss in their genomes. This can further support that genomic divergence occurs as a result of host adaptation mechanisms [44]. This extensive genome variability may play an important role in C. jejuni survival and host adaptation [31].

One application for WGS is studying bacterial genomic diversity accrued by animal colonization and human infection [45‐48]. Golz et al. identified hybrid strains of C. jejuni where extensive gene transfer between the two species interfered with the analysis of species differentiation and multilocus sequence typing (MLST) [49]. These adaptation mechanisms lead to the emergence of host-associated genes or clusters of genes that can be resolved by WGS, which can be of a remarkable use to detect and infer host adaptation mechanisms in C. jejuni [50].

Systematic surveillance of C. jejuni infection is a complex process due to high genomic diversity of the bacteria and interactions between different routes of transmission [31, 43]. In the context of epidemiological surveillance of C. jejuni, bacterial subtyping is crucial to differentiate bacteria sharing certain genomic similarities and link them to the same source [51]. The distinction between epidemiologically related incidents and sporadic cases requires high-resolution detection and typing techniques [10]. This is especially important to track both point source and diffuse outbreaks.

Consistent detection and identification of C. jejuni genotypes is challenging due to their high variability. Additionally, traditional culture-based methods fail to detect bacterial variations [52]. They are time-consuming, low throughput, laborious, of low sensitivity, and may yield false negative results if the bacteria are in VBNC state [52]. Without accurate diagnostic tests to detect the presence of Campylobacter, precise differentiation and diagnosis of enteric illnesses caused by other bacteria including Salmonella, Shigella, and Yersinia can be challenging [53]. Campylobacter species, specifically C. jejuni, possess highly changeable physiology, metabolism, and phenotypic diversity. Consequently, traditional detection methods are inadequate, inaccurate, and not sensitive enough [52]. Thus, research is now driven towards devising more accurate, cost- and time-effective detection methods, especially in the food industry where screening is crucial to prevent transmission [54].

Molecular typing schemes have been previously used for C. jejuni, including in outbreak investigations, host-association, and population structure studies. Restriction fragment length polymorphism, ribotyping, PCR-based methods, pulsed-field gel electrophoresis, and antigen gene sequence typing (as for flaA and porA genes) are considered as robust and reproducible genotyping methods in understanding the biology of C. jejuni. These typing methods can be implemented for different epidemiological purposes, including for Campylobacter subtyping, phylogenetics, identification of outbreak-inducing lineages, and epidemiologic tracking [55]. Nonetheless, the aforementioned subtyping methods have limited discrimination capacity in epidemiological investigations and have several drawbacks such as poor discriminatory power and incompatibility with high throughput applications [55]. MLST has been previously employed over the past decades as the gold standard subtyping method in studying relationships between Campylobacter spp. strains, investigating the evolution, population structure, and the molecular epidemiology of the disease and exploring the potential host reservoirs and host associations [56]. MLST classifies isolates based on polymorphisms present in certain regions of housekeeping genes. Closely related sequence types are grouped under clonal complexes [57]. According to MLST genotyping, strong associations were found between C. jejuni host generalists and some clonal complexes (CC) including: ST-21, ST-48, ST-206, and ST-45 [41]. These lineages are causative sources of human diseases [58]. Despite broad geographical distinction between Campylobacter species, specific STs are found to be associated with infections in specific countries. For instance, ST-22 and ST-4526 were found in Finland and Japan, respectively, while ST-190 and ST-474 emerged in New Zealand [36]. The comparison of geographically distinctive Campylobacter isolates is made possible by molecular typing and WGS. One example is the analysis of Campylobacter genomes in UK and North America [59]. The analysis concluded the clustering of these isolates based on variations of highly recombining genes while the isolates were geographically distant [59]. Another study compared C. jejuni isolates in Egypt and UK. CC21 isolates from the same country shared more accessory genome genes that were lineage-specific; thus, isolates were geographically clustered [60]. Therefore, biogeographical identification of signatures from Campylobacter genomes can help improve campylobacteriosis source attribution and implement reliable intervention strategies.

While MLST is superior to other classic typing methods in studying the population structure for source attribution and the identification of transmission routes in outbreaks, it has several drawbacks [57]. MLST alone may not be sufficient to resolve closely related bacterial strains and, in this case, a specific MLST scheme should be devised [57]. Therefore, new tools and screening techniques were needed for epidemiological surveillance of C. jejuni to address the limitations of the classic typing methods.

WGS technologies are continuously evolving to sequence nucleotides at reasonable speed and low cost. As a result, more and more bacterial genomes are becoming available for analysis and routine surveillance and outbreak tracking are becoming more feasible [61]. Instead of conventional genotyping, which is restricted to only some parts of the genome, WGS provides information on the entire genomic content of isolates. WGS can thus enhance microbial safety surveillance to help control foodborne outbreaks [61]. WGS is characterized by enhanced discriminatory power at the strain level, also enabling the association of specific genotypes with phenotypes that are clinically and epidemiologically relevant [62]. WGS based subtyping demonstrates several advantages over traditional genotyping methods, including in silico prediction of antimicrobial resistance determinants, attribution of transmission sources and routes, and enhanced surveillance of food-borne pathogens [63‐65]. Therefore, WGS serves as an effective measure for controlling and preventing foodborne infections [62]. This is especially demonstrated during infectious disease outbreaks where WGS was used for typing [62]. WGS paves the way to characterize the genomic diversity among Campylobacter isolates; thus, improving decision making and intervention to control outbreaks [10].