Salmonella Gallinarum in Small-Scale Commercial Layer Flocks: Occurrence, Molecular Diversity and Antibiogram

Salmonella Gallinarum is one of the most important bacterial pathogens associated with diminished egg production in poultry. The aim of this study was to understand the occurrence, molecular traits and antimicrobial resistance patterns of Salmonella Gallinarum strains isolated from small-scale commercial layer flocks with low level biosecurity standards in Bangladesh. A total of 765 samples, including cloacal swabs (535), visceral organs (50), and droppings (180), were collected from chickens of 12 layer flocks in 11 districts. Salmonella Gallinarum was isolated and characterized through culture-based method, followed by biochemical tests, sero-grouping, PCR assays, sequencing, and antibiogram. The identity of biochemically detected isolates of Salmonella Gallinarum was confirmed via genus-specific 16S rRNA gene based PCR, followed by invA and spvC genes based PCR assays. Occurrence of Salmonella Gallinarum was detected in overall 25.75% (197/765) samples, with a significantly (p < 0.05) higher incidence in visceral organs (42%) in comparison to cloacal swab (24%) and droppings (26%). Sequencing and subsequent phylogenetic analysis of invA and spvC genes in representative strains of Salmonella Gallinarum revealed a close genetic lineage, with a sequence similarity of 98.05–99.21% and 97.51–99.45%, respectively, to previously published sequences of the corresponding genes from the same serogroup strains. Remarkably, 66.5% (131/197) of the isolated strains of Salmonella Gallinarum were found to be resistant to 3 to 6 antimicrobial agents, and interpreted as multidrug resistant (MDR). The findings of this study underscore an inherent need of appropriate control measures to curb the widespread incidence of MDR Salmonella Gallinarum in small-scale commercial layer flocks, thereby, facilitating enhanced egg production and further support to the food security and safety in low resource settings.


Introduction
Eggs and meat from poultry are indispensable protein sources in peoples meals in Bangladesh [1]. However, the advancement in the poultry production is often interrupted by the overwhelming occurrence of infectious diseases in low resource settings of developing countries such as Bangladesh. Occurrence of these diseases, which incur a colossal loss due to less production of a quality product and also from the treatment cost, are attributable to noncompliance of good agriculture practices (GAP), low biosecurity, and inadequate hygienic measures in poultry farms [2,3]. Amongst the infectious pathogenic microbes, multidrug-resistant (MDR) Salmonella spp. has been regarded as a significant problem to the poultry sector in Bangladesh [4]. Salmonella species are rod shaped Gramnegative bacteria, which are predominantly classified as Salmonella enterica and Salmonella

Sample Collection from the Layer Farms
The samples comprised of cloacal swabs (n = 535, 70.39%), poultry drooping (n = 180, 23.52%); and visceral organs (n = 50, 6.35%), which represented the composite samples of liver, spleen, ovary, and ovarian follicle from the dead layer birds. All samples were collected in sterile plastic containers, properly labeled and transferred in an insulated foam box maintaining cool chain at 4-6 • C to the Department of Microbiology and Hygiene laboratory of Bangladesh Agricultural University and processed within 24 h. During each sampling, information on different parameters as shown in the sample collection checklist (Supplementary Table S2) were obtained.

Isolation and Identification via Culture and Biochemical Tests
Isolation of Salmonella Gallinarum was done based on selective enrichment of the samples (cloacal swab, visceral organ, and droppings) in Rappaport Vassiliadis Soya Broth (RVS Broth) (HiMedia, Mumbai, India) following the protocol described earlier [34] with some modifications. A loopful of enriched broth was streaked on xylose lysine deoxycholate (XLD) agar media (HiMedia, Mumbai, India) and incubated at 43 ± 0.2 • C for 24 ± 2 h. Representative colonies of Salmonella cells grown on selective XLD agar media were separated individually by subculture on the same media. Thus, the obtained pure culture of each of the selected isolates were subjected to biochemical tests viz., sugar fermentation test, indole, and MR-VP tests [35]; and additionally, motility test was done using the hanging drop slide technique [36] to ascertain their identity as Salmonella spp.

Serogrouping of Salmonella Isolates
The isolates of Salmonella spp. obtained from the poultry samples were subjected to sero-grouping by rapid serum plate agglutination test (RSPAT) using commercially available Salmonella agglutinating antisera (S & A Reagents Lab Ltd., Bangkok, Thailand), and following the standard method described earlier [37].

Molecular Detection
Selected isolates of Salmonella spp. were subjected to detection by culture-based methods, and further confirmed by molecular tools, including PCR assays and sequencing. Culture lysate and DNA template of S. Gallinarum isolates were prepared using the standard procedure, as described previously [38]. A previously established 16S rRNA gene-based PCR, using primers and conditions shown in Table 1, was performed for confirmation of the genus Salmonella [39]. A reaction mixture of 25 µL volume, including 12.5 µL 2X master mixture, 2 µL genomic DNA, 1 µL each primer, and 8.5-µL nuclease-free water (Thermo Fisher Scientific, Waltham, MA, USA) was prepared and PCR amplification was carried out using a thermocycler (Astec, Fukuoka, Japan) as per the manufactures' protocol. The PCR reaction comprised of an initial denaturation step at 94 • C for 5 min; followed by 30 cycles of DNA amplification, each including denaturation at 94 • C for 30 s, annealing at 50 • C for 30 s, and extension at 72 • C for 30 s; followed by a final extension step at 72 • C for 5 min. The PCR products were subjected to gel (2% agarose) electrophoresis (Invitrogen, Carlsbad, CA, USA), followed by staining of the gel with ethidium bromide (0.5 µg/mL) and de-staining in distilled water, each for 10 min.; and finally, the PCR amplicons in the gel were visualized under UV light, and images were captured using a gel documentation system (Biometra, Göttingen, Germany). Species-specific PCR assays, targeting invA and spvC genes, were employed to confirm the identity of Salmonella spp. isolates [13,18]. Details of the primers used in the PCR assays are listed in Table 1. Initially, a reaction mixture (25 µL) was prepared with template DNA, × 5 PremixTaq™ cradle, deoxynucleotide triphosphates (10 mM each), 25 mM MgCl2, 10 µM of each primer with Premix Taq™ DNA polymerase (Takara Bio Inc., Shiga, Japan), and nuclease-free water in volumes of 3 µL, 5 µL, 0.5 µL, 1.6 µL, 1 µL, 0.3 µL, and 12.6 µL, respectively. PCR amplification, agarose gel electrophoresis, and visualization of the PCR products were done following the standard protocol, as described previously. PCR products of representative isolates, namely BAUSG3 and BAUSG6, were purified using the GeneJET Genomic DNA Purification Kit (Thermo Scientific™, Thermo Fisher Scientific, Waltham, MA, USA); and then sequenced using specific primers partially amplifying the invA and spvC genes (Table 1) in a Genetic Analyzer 3130 (Applied Biosystems™, Thermo Fisher Scientific, Waltham, MA, USA) as per manufacturer's instructions. After initial quality check-up, trimming and editing, as required, the sequenced genes were uploaded in the GenBank. The ClustalW algorithm of the Molecular Evolutionary Genetics Analysis (MEGA) software (version 4.1.0) was applied to align the different sequences of the specific genes, including those obtained in this study and others selected through homology search using BLAST tool (www.ncbi.nlm.nih.gov/BLAST, accessed on 4 December 2020). Afterwards, a phylogenetic tree was constructed following the neighbour-joining (NJ) method [38]. The bootstrap test (100 replicates) was used to evaluate the percentage of replicate trees with related taxa assembled and that appeared as branches, identically [39]. The sequence identity was validated by comparison with the published sequences of the target genes available in the GenBank database (https://www.ncbi.nlm.nih.gov/, accessed on 10 January 2021).

Data Management and Statistical Analysis
All data obtained in this study were recorded in Excel spreadsheets and analyzed for descriptive statistics (frequency, proportion and 95% Confidence Interval [CI]). To estimate the variations in occurrence and antimicrobial resistance pattern of Salmonella Gallinarum, SPSS software (version 22.0, IBM Corp., Armonk, NY, USA) was used. Chi-squared test was performed where applicable to determine the level of significance of difference or association. A p value of <0.05 was considered statistically significant.

Sample Level Occurrence
Among the collected samples (N = 765), 28% (n = 214) were confirmed as positive for Salmonella spp. by conventional culture-based method. Among these isolates (n = 214) of Salmonella, 199 were confirmed to be Salmonella Gallinarum through selective biochemical tests, including carbohydrate fermentation, indole production, methyl red test, Voges-Proskauer reaction, and motility test. Isolates of Salmonella Gallinarum produced positive results in fermentation of glucose, maltose, and dulcitol, all without any gas production, which are typical biochmemical traits of this serovar, and useful to differentiate from closely related serovars (Tables 2 and 3).  Table 3. Results of different biochemical test among culture positive isolates of Salmonella spp. (n = 214).

Biochemical Tests Salmonella Gallinarum Salmonella Typhimurium Salmonella Pullorum Others
Carbohydrate fermentation Glucose + + + -(Acid) (Acid and Gas) (Acid and Gas) - (Acid and Gas) (Acid and Gas) Indole production -- The 16S rRNA gene-based PCR confirmed the genus identity of 205 of 214 isolates, biochemically determined as Salmonella spp. Results of this genus-specific PCR showed that, overall, an estimated 26.80% (205 of 765) samples were contaminated with Salmonella spp., since the representative isolates from the positive samples generated the expected amplicon size of 574 bp (Table 4). However, invA and spvC-gene based PCR assays confirmed that approximately 25.80% (n = 197) samples/isolates were positive for Salmonella Gallinarum. The identity of these genes was further validated via sequencing, while phylogenetic analysis revealing their closest affinity to corresponding gene sequences reported previously from Salmonella Gallinarum strains. Observed variations in the occurrence of Salmonella spp. in different kinds of samples was found to be statistically significant (Tables 2 and 4). Among the four kind of samples, a distinctive higher (p < 0.025) occurrence of Salmonella spp., including Salmonella Gallinarum (mean incidence rate 44% and 42%, respectively), was observed for the tested visceral organs in comparison to cloacal swab and dropping samples (mean incidence rate ca. 25 to 27% and 24 to 26%, respectively). Of 205 isolates of Salmonella spp., when subjected to serogrouping using fur commercially available types of antisera in rapid serum plate agglutination test (RSPAT), the Salmonella spp. was classified into three serogroups, viz, with exclusive dominance of Group D (n = 199, 97%), and minor presence of Group B (n = 4, 2%) and Group C (n = 2, 1%) strains (Table 5). . The observed variations in division-wise occurrence of the bacterium was found to be statistically non-significant (p = 0.97) (Figure 2). One the other hand, of the 12 flocks surveyed in this study, 75% (9/12) were found to be positive for Salmonella Gallinarum (Supplementary Table S1). However, the occurrence of Salmonella Gallinarum in different samples (cloacal swab, droppings, and visceral organ) at individual farms showed more wide-ranging variations, i.e., from 0% to 56.9% of total samples (Figure 2). average occurrence of 25.8% (95% CI: 22.7-29). The observed variations in division-wise occurrence of the bacterium was found to be statistically non-significant (p = 0.97) ( Figure  2). One the other hand, of the 12 flocks surveyed in this study, 75% (9/12) were found to be positive for Salmonella Gallinarum (Supplementary Table S1). However, the occurrence of Salmonella Gallinarum in different samples (cloacal swab, droppings, and visceral organ) at individual farms showed more wide-ranging variations, i.e., from 0% to 56.9% of total samples (Figure 2).

Sequencing and Phylogenetic Analysis
Gene (invA and spvC) based PCR assays, followed by sequencing and phylogenetic analysis of the representative target genes confirmed the identity of Salmonella Gallinarum strains. BLAST analysis confirmed a 100% similarity of invA gene between the isolated strains of Salmonella Gallinarum, namely BAUSG3 and BAUSG6. These nucleotide sequences of invA gene showed a close homology between 99.21% and 98.05% to earlier reported strains isolated from India (Accession number: JQ812057.1), China (EU348366.1), Korea (KF192263.1), Egypt (KM282011.1), UK(AM933173.1), and the USA (CP019035.1) ( Figure 3A). Phylogenetic analysis of invA genes revealed the representative strains clustering to a lineage, which comprised of five previously published sequences of Salmonella Gallinarum strains from different countries. Similarly, nucleotide sequences of the spvC genes of Salmonella Gallinarum strains (BAUSG3 and BAUSG6) were found to be closely related with the previously published sequences of spvC of the same serogroup strains isolated from Korea and the UK ( Figure 3B). According to BLAST analysis, a 97.8% similarity with five mismatches between the nucleotide sequences of the spvC genes was observed for the isolated strains, BAUSG3 and BAUSG6. However, these spvC genes showed between 99.82% and 97.51% homology to previously published sequences of this gene in strains isolated from different countries. The nucleotide sequences generated in this study were submitted in the GenBank and are available under accession numbers: MK801112 and MK801113 (invA1 and invA2, respectively) and MK801114 and MK801115 (spvC1 and spvC2, respectively).
BLAST analysis, a 97.8% similarity with five mismatches between the nucleotide sequences of the spvC genes was observed for the isolated strains, BAUSG3 and BAUSG6. However, these spvC genes showed between 99.82% and 97.51% homology to previously published sequences of this gene in strains isolated from different countries. The nucleotide sequences generated in this study were submitted in the GenBank and are available under accession numbers: MK801112 and MK801113 (invA1 and invA2, respectively) and MK801114 and MK801115 (spvC1 and spvC2, respectively)

Discussion
Salmonellosis is considered as one of the major causes of decreased meat and egg production in commercial poultry. The occurrence of disease has been reported to be connected with inadequate biosecurity measures in poultry farming practices in Bangladesh [20,44]. It has been difficult to evaluate the actual disease burden and adopt appropriate interventions in small-scale commercial poultry farms of Bangladesh due to limited systematic information on decreased egg production in layer flocks associated with salmonellosis, particularly, Salmonella Gallinarum infection. Of the surveyed farms scattered over different districts in five divisions in Bangladesh, a huge majority (75%, 9/12 farms) of the layer flocks were found infected with Salmonella Gallinarum, which could be connected to decreased egg production [18,29,45]. Considering its significance, Salmonella Gallinarum infection in the poultry sector has been included within the important notifiable diseases of the World Organization for Animal Health (OIE), whereas there has been an increased attention to implement a strict control measures regarding the import of birds and eggs [46].
Observations made in this study showed the occurrence of Salmonella spp., detected by biochemical tests and genus-specific 16S rRNA assay, in 26.8% (205/765, 95% CI: 23.7-30.1) samples of layer flocks. Interestingly, an overwhelming occurrence in these samples, 25.8% (197/765,, of Salmonella Gallinarum, confirmed by differential biochemical tests and molecular assays targeting virulence genes, invA and spvC was revealed. The observed large-scale contamination in visceral organs of Salmonella Gallinarum can be related to the typical invasive feature of Fowl Typhoid (FT), which causes lesions in multiple organs, including liver, heart, spleen, ovary, and intestine. In congruence, a higher prevalence of this virulent serogroup in visceral organs (42%, 21/50) in comparison to dropping (26.1%, 47/180) and cloacal swabs (24.1%, 129/535) was notable. A ubiquitous occurrence of invA and spvC genes observed for the isolated strains (n = 197) of Salmonella Gallinarum is in congruence to their prevalence reported previously among the virulent serotypes of S. enterica [10]. The observed similarity in the sequences of virulence invA and spvC genes of Salmonella Gallinarum strains obtained in this study indicates their genetic relatedness to the Salmonella Gallinarum strains isolated from poultry sources in different counties, including India, China, Korea, Egypt, UK, and the USA. Sequencing of invA and spvC in a higher number of Salmonella Gallinarum strains would enrich our knowledge to better understand the evolutionary trend linked to the geospatial prevalence of these virulent genes in Salmonella spp.
The overall occurrence of Salmonella Gallinarum in 25.8% (197/765) of the samples is lower than a previously reported occurrence in 53.5% of the samples of layer flocks in Bangladesh [21]. However, as observed in the present study, this pathogenic serovar can contaminate the majority samples (at least 57%) of the poultry flocks in an individual farm. On the other hand, a variable occurrence, between 11.5% and 24%, of Salmonella spp. reported for poultry samples in different geographical locations, e.g., France, Japan, Tanzania, and United Kingdom, respectively [47][48][49][50], could be attributable to geo-climatic variations. Nonetheless, differential anthropogenic risk factors, namely, age of the birds, flock size, feed, hygienic condition of the farm, and environmental determinants, including transmission from poultry litter, pest, and rodent may influence the preponderance of pathogenic microbes such as Salmonella Gallinarum in small-scale poultry farms [51,52].
Among the potential biosecurity interventions, systematic vaccination programs based on surveillance studies on the pathogenic microbes circulating in a particular region is a key approach to effectively control the infection of virulent strains, including Salmonella serotypes, in poultry flocks [53]. The poultry farms surveyed in this study were selected based on their non-vaccinated status, imposing a higher likelihood of diseases from Salmonella Gallinarum in reared layer flocks [54]. At present, a few imported commercial vaccines are available for immunization of layer poultry flocks in Bangladesh. Notably the SG 9R strain of Salmonella Gallinarum is being used as FT vaccine (viz. Nobilis ® SG 9R, MSD Animal Health). However, this vaccine may not be antigenically well-matched with the circulating strains. Phylogenetic analysis of the nucleotide sequences of a viru-lence gene, spvC, obtained from a couple of Salmonella Gallinarum strains of this study, displayed its close relatedness to that of the SG 9R strain (accession number: HM044661). This kind of locally adapted strain could be considered as a more potential candidate strain to develop antigenically well-matched vaccine for Salmonella Gallinarum strains circulating in Bangladesh.
The rise of MDR among Salmonella spp. is a growing concern worldwide, particularly in developing countries, where multiple antibiotics are indiscriminately used at poultry farms for enhanced production [19,[55][56][57]. The present study showed that 48.7%, 45.2%, and 43.1% isolates were intermediately resistant to Ciprofloxacin, Levofloxacin, and Gentamcin, respectively, however, 79.7%, 61.4%, 49.7%, 47.7%, 43.7%, and 37.6% isolates were fully resistant to Oxytetracycline, Doxycycline, Amoxycillin, Sulfamethoxazole, Enrofloxacin, and Neomycin. These findings are in congruence to the results of earlier studies [25,[57][58][59]. Bangladesh is already included among the countries at high risk of AMR, according to the WHO [60]. Likewise, this study reporting 66.5% (n = 131) Salmonella Gallinarum isolates as resistant to three to six antimicrobials, indicates an alarming consequence to the extensive use of different antibiotic classes in layer flocks.
The low biosecurity standards in majority (>60%) of the small-scale commercial poultry farms, together with no vaccination status, and unscrupulous use of antibiotics might have substantiated the high occurrence of MDR strains of Salmonella Gallinarum in the layer flocks in Bangladesh [51,61]. Apart from further infections caused by this kind of MDR Salmonella Gallinarum strains, reportedly also resistant to prophylactic antibiotics [62], exposure of residual antimicrobials through the food chain is considered to impose a significant hazard to public health [63]. Therefore, training programs to enrich farmers' knowledge on Good Agriculture Practices (GAP) including prudent use of antibiotics, immunization of layer birds and other biosecurity measures, including hygiene, water and waste management [24,64,65], would be vital to lessen the Salmonella Gallinarum infection in poultry farms.
The limited sampling scheme employed for each of the study farms did not capture any temporal variation in the occurrence of Salmonella Gallinaruam in layer flocks. Because of high phenotypic similarity between the closely related serovars, Salmonella Gallinarum and Salmonella Pulloram, a number of biochemical tests and serotyping were required to substantiate molecular detection targeting a couple of genes, including their sequencing and phylogenetic interpretation, of Salmonella Gallinarum isolates under this study. Culturebased biochemical identification methods have some inherent drawbacks, e.g., being labor-intensive and time consuming (requiring five to seven days). Another limitation of the methodological approach was the qualitative estimation of the bacterial occurrence, whereas a quantitative approach could capture more detail of the variations. On the other hand, sequencing and phylogenetic analysis of the target genes were done for only a few isolates. These could be considered as the primary limitations or constrains of this study.

Conclusions
The present study clearly shows a widespread occurrence of Salmonella Gallinarum strains with MDR traits in small-scale commercial layer flocks in all the major divisions of Bangladesh. The updated information on AMR patterns of Salmonella spp. circulating in the poultry flocks will contribute in efforts to lessen the multi-spectrum hazards, including treatment failure and production loss associated with the large-scale infection of MDR Salmonella Gallinarum in the poultry sector. Results obtained from the surveyed farms indicate an indispensable need of promoting biosecurity measures, and farmers' training on GAP, including strict vaccination, and prudent use of antimicrobials. This study will eventually benefit the policy makers in divulging a strategic framework for the improvement of farmers' livelihood, health and food security and safety in the context of the alarming MDR infections in poultry flocks at low resource settings.