Introduction
Foodborne diseases persist as a globally prevalent human health concern. According to the Center for Disease Control (CDC), a complex consortium of microorganisms is responsible for 90% of illnesses, wherein Salmonella ranks second.1 This gram-negative bacterium encompasses more than 2,500 serovars, categorized into typhoidal and non-typhoidal Salmonella based on the associated disease syndrome, affecting humans and exhibiting a wide host range.2 Furthermore, Salmonella is deemed a ubiquitous bacterium, demonstrating a heightened prevalence in warm-blooded animals, notably in cattle, pigs, and poultry,3 which function as carriage animals even in the absence of clinical manifestations.4
The Centers for Disease Control and Prevention (CDC) have projected an annual incidence of 1.35 million Salmonella infections in the United States.5 Remarkably, Mexico mirrors this epidemiological landscape. In the year 2020 alone, 64,778 cases were reported, encompassing instances of typhoid fever (25.4%), paratyphoid (84%), and other salmonellosis (66.2%).6 Furthermore, an examination of the past five decades in Mexico reveals a substantial surge from 7,629 to 45,280 infections. Notably, among the various Mexican states, Sinaloa holds the foremost position, exhibiting elevated rates of typhoid fever.7
Historically, epidemiological occurrences of Salmonella outbreaks have been ascribed to direct contact with animals or the ingestion of tainted animal-derived products, including meat, eggs, and milk.8 Consequently, antibiotic administration has traditionally served as the primary therapeutic approach for treating Salmonella infections, albeit with adverse consequences manifested in the gradual emergence of antibiotic resistance mechanisms. Presently, the prevalence of antimicrobial resistance in Salmonella strains has escalated to a critical juncture, posing a significant and pressing challenge.9
Another noteworthy facet pertains to Salmonella pathogenesis, intricately linked to the expression of virulence factor genes situated within distinct pathogenicity islands (SPI).10 A prominent component in this regard is the Type III secretion system (T3SS), recognized for its needle-like structure facilitating protein translocation within epithelial cells. This system is encoded by SPI-1 and SPI-2, playing pivotal roles in inducing cell attachment through cytoskeletal rearrangement, modulating host immune responses, ensuring intracellular survival, and facilitating invasion.11, 12
Within this framework, a One Health approach has been advocated to investigate the interconnection among animals, the environment, and human health as primary focal points.13 Strategic initiatives, such as next-generation sequencing and bioinformatics, have significantly expanded our understanding of the fundamental genomic characteristics of pathogens, enabling timely identification a crucial element in the intervention of foodborne outbreaks.14 In alignment with the overarching goal of One Health, the objective of this study is to scrutinize the genomic population structure to infer certain genotypic traits associated to the Mexican isolates through the utilization of next-generation sequencing (NGS) and a bioinformatic methodology applied to S. enterica isolates obtained from farm animals.
Materials and Methods
Bacterial strains
In this study, a total of 72 S. enterica strains isolated from asymptomatic farm animals’ manure, specifically cattle (n=38), poultry (n=17), and goat (n=17) sourced from the Culiacán Valley in Sinaloa, were employed (Table 1). The strains under investigation were procured from the private collection of the National Food Safety Laboratory (LANIIA) at the Centro de Investigación en Alimentación y Desarrollo (CIAD) in Culiacán, México. DNA extraction was conducted utilizing the DNAeasy Blood & Tissue Culture commercial kit, following the manufacturer's protocols. The concentration of the extracted DNA was quantified using the Qubit dsDNA Broad Range Assay Kit (Thermo Fisher, USA). Subsequently, the Nextera XT DNA sample kit was employed for library preparation, and genome sequencing was performed using the Illumina Miseq platform (Illumina, Inc.) to acquire paired-end reads (2x150 bp).
Reads quality control and assembly
The initial assessment of sequencing-derived reads quality was conducted using FASTQC.15 Subsequently, Trimmomatic V0.3216 was employed to trim sequences of suboptimal quality, defined as those falling below a Phred quality score per base of 20. Furthermore, reads with lengths less than 100 base pairs and adapters were eliminated. The de novo assembly of reads was accomplished through the A5-miseq pipeline17 utilizing the paired-end reads as input. The generated assemblies were uploaded to NCBI under the PRJNA313928 BioProject.
Taxonomy and ST assignation
To validate taxonomy identity and predict sequence type (ST), the pubMLST website (REF) was used to compare Salmonella allelic profiles with the generated draft genomes. For serotype assignment, the stand-alone version of the Salmonella in silico typing resource (SISTR)18 was employed. To visually represent the diversity in ST and locus variant, a spanning tree was generated using the online version of PHYLOViZ.19
Table 1
Metadata associated with the 72 genomes of S. enterica isolated from farm animals in the Culiacán, Sinaloa region
Genomic analysis
Virulence and antibiotic resistance gene annotation were conducted through the mass screening of contigs using ABRIcate.20 This analysis incorporated the Virulence Factor Database (VFDB) and the Resfinder databases, with a selection criterion of 95% coverage and 80% alignment. Plasmid replicon detection utilized the PlasmidFinder21 applying the aforementioned coverage and alignment parameters. The protein secretion system profile for each isolate was investigated using TXSScan with MacSyFinder.22 Bacterial secretion systems were selected according to the fulfillment of mandatory genes for system assembly. Additionally, prophage prediction for S. enterica was executed using PHASTER, 23 with only intact detected prophages considered in the results. To assess genetic relatedness among isolates, a phylogenetic tree was generated based on a core alignment. Parsnp24 employed to create the core alignment, utilizing a random reference from one of the 72 analyzed genomes. For the development of a robust and supported phylogenetic tree, RAxML25 was utilized, employing a time-reversible GTR model with 100 bootstraps for statistical support. The resulted tree was midpoint rooted and subsequently edited using the Interactive Tree of Life iTOL.26
Results
A total of 18 S. enterica serovars were identified among the 72 farm animal isolates (Figure 1, panel A). The three most prevalent serovars were Oranienburg (25%), Give (15.2%), and Saintpaul (12.5%). Furthermore, a primary isolation source was found for each of the most prevalent serovars, such as Oranienburg in goat isolates, and Give and Saintpaul in cattle and poultry isolates, respectively (Figure 1, panel B). The serovars were categorized into 18 sequence types (ST) groups based on allelic profile (Figure 1, panel C). Notably, 94% of the serovars were successfully assigned to a specific ST, except for serovar Newport, designated into ST45 and ST118. STs were assigned for 92% (66 isolates) of the S. enterica genomes. The most frequently occurring STs were ST23 (n = 16), ST654 (n = 11), and ST50 (n = 8), corresponding to the prevalent serovars Oranienburg, Give, and Saintpaul, respectively. In terms of allelic differences, a range of 3 to 7 alleles was observed among detected STs with an average of 6 alleles.
Figure 1
Distribution of Salmonella serovars and isolates. A): Number of isolates from farm animals. B): Distribution of S. enterica serovars across farm animals. C): Spanning tree from the 18 STs detected. Black number inside the circles represents the ST while red numbers the allelic difference among STs. Outer colored ring represents ST proportion by isolation source
Figure 2
S. enterica phylogenetic tree coupled with virulence presence/absence markers. Highlighted labels show serovar. An additional color column showed sample isolation source. Specific clades are colored according serovar. Bootstraps >85 are displayed with blue triangles
Figure 3
Accumulative graph. A): Graph for Salmonella enterica plasmids by serovar. B): Graph for Salmonella enterica phages by serovar. The serovars are shown on the Y axis, while the occurrence percentage is shown on the X axis. Each plasmid and phage are denotated with different color. Total number of plasmid or phage occurrence are shown in bar chart in green
The genomic content linked to virulence markers has been delineated across two figures. Supplementary Figure 1 illustrates the conservative profile of genes detected across all isolates while Figure 2 highlights virulence markers specific to individual serovars. Within the shared genetic framework, genes for thin aggregative fimbriae responsible for biofilm formation (csgA-G), type 1 fimbria, Type III Secretion System (T3SS) encoded by SPI-1 and SPI-2, and the TSSS-1 secreted effectors were identified in all serovars originating from the three animal sources (Figure 1). On the other hand, the related E. coli adhesive fimbriae (faeC-E) were found exclusively for the serovars Anatum, Saintpaul, Minnesota, and Oranienburg. A distinct clade, comprising five different serovars, exhibited the presence of the long polar fimbriae (lpfA-E). Plasmid-encoded fimbriae (pefA-D) were identified in serovar Typhimurium. The immune modulation genes gtrA and gtrB were observed in the Luciana serovar. Interestingly, the sodCI gene, associated with environmental stress, was present in all members of serovars Saintpaul and Weltevreden. Finally, type VI secretion system components tssJ, L, and M were found in Weltevreden and Agona serovars.
Concerning antibiotic resistance, all 72 isolates demonstrated the presence of the aac(6´)-laa gene, indicative of aminoglycoside antibiotic inactivation. Specifically, the foasA7 gene associated with fosfomycin acid antibiotic resistance was found for the two S. Agona genomes. Additionally, floR and tet(A) genes were detected in the singular Montevideo genome, conferring resistance to phenicol and tetracycline, respectively.
A total of 58 replicon plasmids, categorized into 14 distinct types were identified among the isolates (Figure 3, panel A). Notably, the serovar Give exhibited the highest number (23/58) and diversity, encompassing 7 different replicons. The two most observed replicons were IncFII(Prsb107) and IncFII(S), with 20 and 16 occurrences, respectively, constituting 62% of the total observed replicons. The serovars Cayar, Albany, and Montevideo exhibited the lowest replicon number (1/58) and diversity (1/14).
A total of 183 prophage sequences were detected, encompassing 94% of the serovars (17/18). The highest occurrences were associated with serovars Give (46/183), Minnesota (32/183), and Oranienburg (31/183), securing the first, second, and third positions, respectively. The most frequently observed prophages across genomes were Vibrio X29 (21/183), Escher RCS47 (21/183), and Salmon Fels 1 (18/183). In contrast, Gaminara and Albany each exhibited only one prophage insertion. The least observed prophages were Entero DE3, Escher500465, Enteri Sf101, Escher 500465 1, and Entero Mu, each presenting only one sequence.
Discussions
The genomic diversity of S. enterica in farm animals from Sinaloa is composed of a broad range of 18 serovars associated with one ST. The detected serovars have been previously reported in Mexico by other investigations.7 Nevertheless, this is the first report for serovars Thompson and Soahanina in Mexico. Similar serovars have been reported in other studies around the world in farm animals causing diarrhea. 27, 28, 29 Although the three main serovars found in this study are uncommonly related to nontyphoidal salmonellosis, epidemiological outbreaks have been detected in Mexico and the rest of the world.30, 31, 32 Salmonella isolates were classified into 18 STs according to the MLST. This approach consists of a typing tool for isolate comparisons, such as outbreak identification and the recognition of virulent strains. 33 Under this scenario, STs found in Sinaloa have been detected in epidemiological outbreaks in other countries. For example, STs 19 and 26 belonging to S. serovars Typhimurium and Thompson respectively, have been detected in patients with clinical symptoms in the USA. 34 Moreover, epidemiologic studies for serovar Agona ST13 have been related to the French and German outbreaks. 35 The resulting STs for the prevalent serovars Oranienburg (ST23), Give (ST654), and Saintpaul (ST50) support the notion of their wide distribution and prevalence in Mexico. 36, 37
The mandatory genes for T3SS assembly were found for all 72 the examined isolates. These genes facilitate the translocation of protein effectors, leading to their internalization within host cells, thereby enabling intracellular persistence and replication. Conversely, certain isolates exhibited additional virulence markers that may confer a significant advantage in terms of survival and transmission during colonization process,38 particularly noteworthy in generalist serovars, as the predominant observed in this study. An illustrative instance is provided by the fimbria adhesive faeC-E identified in serovars, Oranienburg, Saintpaul, Minnesota, and Anatum. This fimbria notably enhances adherence to epithelial cells. Furthermore, the extensively characterized long polar fimbria may contribute to the binding of M-like cells on the intestinal Peyer patches.39 The found gtr operon is accountable for the structural modification of lipopolysaccharide, potentially augmenting immune invasion by evading surface antigens recognition by the host immune system.40 Notably, the stress adaptation sodCl gene may afford Salmonella protection against phagocytic superoxide during infection. 41 Additionally, it has been elucidated that sodCl plays a pivotal role in the adaptation role in Salmonella survival in non-host environments such river water.42
In contrast to findings reported in other studies,43, 44 our investigation revealed a notably low content of antibiotic resistance gene. A mere of the 95% of the examined genomes exhibited resistance to a single antibiotic, especially aminoglycoside. Conversely, the remaining 5% demonstrated multidrug resistance (MDR), encompassing resistance to tetracycline and phenicol. This discrepancy may be attributed to the fact that the animals analyzed in our study originate from small-scale farms primarily dedicated to self-consumption. Consequently, the utilization of antibiotics on these farms is relatively limited. This limited usage suggests that Salmonella strains in this context do not readily acquire antibiotic resistance, in stark contrast to animals intended for commercialization where the indiscriminatory use of antibiotics has been implicated as a contributing factor for antibiotic resistance development.41
The plasmid replicons IncFII(pRSB107) and IncFII(S) emerged as the most prevalent types among the Salmonella isolates, a consistent finding with observations in the United States45 and China.46 The presence of these replicons may signify a mechanism facilitating horizontal gene transfer within Salmonella genomes, potentially contributing to virulence enhancement. To exemplify, IncFII(pRSB107) has been reported to harbor an aerobactin virulence marker and may be associated with MDR.47 Additionally, the replicon Col(pHAD28) carries the qnrB19 gene, imparting quinolone resistance,48 while IncF(S) is linked to spv, contributing to Salmonella systemic virulence and intramacrophage survival.49 Prophage insertions constitute another pivotal evolutionary mechanism influencing bacterial differentiation and persistence. Notably, associations with immune protection have been established in relation to the prophage VibrioX29. This phenomenon, although infrequently observed in prior studies, was widely detected in our investigation. Furthermore, the identification of the prophages Gifsy-1, Gifsy-2, Fels-1, and Fels-2 in S. Typhimurium has been linked to an adaptative response to stress conditions.50 Notably, studies have established that the Gifsy prophage family may genetically contribute to the presence of the SopE gene, facilitating host entry through membrane folding induced by cytoskeleton rearrangement.
Conclusions
The microcosm population structure of Salmonella isolated from farm animals’ genomes in Sinaloa reveals 18 serovars, with S. Oranienburg, Give, and Saintpaul being the most prevalent. Genomic evidence related to ST outbreaks, the presence of virulence factors, and antibiotic resistance markers highlight a significant public health risk. Profiling these isolates genomically could provide invaluable association studies, aiding in tracking Salmonella sources and implementing timely containment measures. Replicons and prophages, acting as mechanism for horizontal gene transfer, may contribute to Salmonella adaptation by triggering stress responses and expanding its genomic repertoire for virulence. We advocate for an in-depth genomic characterization of replicons and prophages to enhance understanding of Salmonella genomic contribution to its adaptation and survival in various environments, given its ability to transition between environmental reservoirs and host. Such an approach would offer valuable insights into Salmonella broader ecological and adaptive potential, informing public health strategies and interventions effectively.
Article DOI
