Prevalence and antimicrobial susceptibility profiles of non‐typhoidal Salmonella isolated from chickens in Rajshahi, Bangladesh

Abstract Salmonellosis in poultry is an important disease that seriously impedes the development of the poultry industry. The increased resistance to antimicrobials against Salmonella has been a major public health concern worldwide. We conducted a study from January to June 2016 in and around the Rajshahi district of Bangladesh on the commercial chicken to isolate, identify and characterize poultry‐specific Salmonella, to assess the potential risk factors and to determine the antimicrobial resistance pattern of the isolates. The overall prevalence of Salmonella enterica was 41% (49/120) [95% CI: 31.95%–50.17%] with 41.7% in broiler chicken (25/60) [95% CI: 29.06%–55.12%] and 40% in layer chicken (24/60, 40%) [95% CI: 27.56%–53.46%]. Samples collected from Rajshahi city (OR = 1.37, 95% CI: 0.50–3.73) and Puthia Upazila (OR = 1.51, 95% CI: 0.56–4.12) were more likely to be positive for Salmonella than Charghat Upazila. Salmonella detection was 1.3 times higher in chicken, providing loose feed than those provided ready feed. All the isolates fermented dextrose, maltose and mannitol with the production of acid and gas, but did not ferment sucrose and lactose. The isolates showed catalase, MR, citrate utilization test and TSI agar test positive, but indole and V‐P tests negative. Salmonella isolates were sensitive to ciprofloxacin (90%), gentamycin (80%), amoxicillin (75%), streptomycin (70%), ampicillin (45%) and sulfamethoxazole‐trimethoprim (45%), whereas highly resistant to penicillin (100%) and nalidixic acid (100%) followed by sulfamethoxazole‐trimethoprim (55%), ampicillin (40%) and amoxicillin (25%). Salmonella enterica is endemic in commercial chicken production in Bangladesh with high prevalence. A considerable proportion of Salmonella isolates was found to be resistant to the majority of the common antimicrobial drugs. A good biosecurity system could be effective for the reduction of Salmonella. It is necessary to obtain universal commitments to establish prudent antibiotic use policies.


| INTRODUC TI ON
Salmonella is an important food-borne pathogen causing an estimated 153 million enteric infections and approximately 57,000 diarrhoeal deaths worldwide every year (Kirk et al., 2015).
Poultry and poultry product are often implicated as a potential risk factor for human salmonellosis (Bryan & Doyle, 1995;Humphrey, 2000). Despite significant advances in technology and hygienic practices at all levels of chicken production, salmonellosis poses an unrelenting threat to human and animal health. It is caused by a large group of bacteria of the genus Salmonella under the family Enterobacteriaceae (OIE, 2018). There are more than 2,600 serotypes of Salmonella broadly categorized into host-restricted, host-adapted and generalist based on their host specificity, virulence, phage typing, etc (Mezal et al., 2014). Among them, Salmonella Gallinarum and Salmonella Pullorum are host-restricted non-motile serovars of chicken. However, chickens commonly harbour other generalist non-typhoidal (NT) serovars of public health significance such as S. Typhimurium, S. Enteritidis, S. Heidelberg and S. Newport (Wray et al., 1996). These non-hosts adapted serovars rarely cause clinical diseases in chickens, but they can be transmitted to humans through consumption of contaminated eggs and/or meat (Wray et al., 1996). The Salmonella serovar Gallinarum may be divided into biovars Gallinarum and Pullorum, which are, respectively, responsible for the fowl typhoid and the pullorum disease of chickens, and are widely distributed throughout the world, especially in developing countries (Barbour et al., 2015). Pullorum disease occurs in chicks during their first few days of life, and fowl typhoid is a disease of mature fowls that drops egg production (OIE, 2018).
Antibiotics have been used in livestock and poultry to treat infections and improve feed efficiency (Hutchinson et al., 1991) as well as to control and prevent infections (Tollefson & Miller, 2000).
Poultry products are one of the most commonly consumed products worldwide, but lots of essential antibiotics are used in many countries during its production, threatening the safety of these products (through antimicrobial residues) and the increased possibility of development and spread of microbial resistance in poultry settings (Agyare et al., 2018). Antimicrobial resistance (AMR) is a burgeoning problem for public health, particularly with the introducing of multi-drug-resistant (MDR) microorganisms. In developing countries like Bangladesh, antimicrobials are used not only for therapeutic purposes but also for growth promotion in the poultry industry. Although S. Gallinarum and S. Pullorum cause diseases only in chicken, the emergence of antimicrobial resistance among these serovars can be horizontally transmitted into other non-typhoidal zoonotic serovars. Antimicrobial-resistant zoonotic bacteria are of particular concern as they may impede effective treatment regimes in humans (Prestinaci et al., 2015). Therefore, determining the nature and extent of AMR found in poultry in Salmonella is essential.
Antibacterial sensitivity tests usually are performed to select the suitable antibacterial agents for the effective therapeutic purpose of salmonellosis; however, due to the recent emergence of MDR Salmonella strain, antibiotic treatment for salmonellosis is getting difficult (Kuehn, 2019;Nair et al., 2018).
Salmonellosis is important as both a cause of clinical disease in commercial poultry that hindered the development of the poultry industry in Bangladesh and as a source of human food-borne zoonotic diseases (Mahmud et al., 2011;Waltman et al., 2008). For proper control and management of salmonellosis, it is necessary to determine its status at the farm level. Isolation, identification and characterization of the particular aetiological agent are essential for a better understanding of a disease situation in a particular area (Ahmed et al., 2008). Prevention and control of salmonellosis require to identify it's antimicrobial resistance pattern. Therefore, the present study was undertaken with the objectives (1) to determine the prevalence of Salmonella; it's isolation and identification from apparently healthy chickens, (2) to determine the antimicrobial resistance pattern of the isolates.

| MATERIAL S AND ME THODS
We conducted the study from January to June 2016 in 24 randomly selected poultry farms of three different study areas, namely Puthia Upazila (sub-districts), Charghat Upazila and Rajshahi City Corporation of the Rajshahi district of Bangladesh ( Figure 1). We drew an experimental design for conducting the study following the different steps ( Figure 2).

| Sampling
We collected a total of 120 cloacal swab samples from the apparently healthy chickens of the selected farms. An equal number of samples (40) were collected from each of the three study areas, and among these, 60 samples from broiler farms and 60 samples from layer farms. More specifically, five samples were collected from each of the 24 farms.
Samples were collected from the mucosa of the cloacal opening of both broiler and layer chickens. A sterile swab stick moistened with sterile normal saline water was inserted into the chicken's cloaca, collected the sample and then placed in sterile vials having Stuart's transport medium in the icebox. The swabs were collected randomly and aseptically then transferred immediately to the laboratory.
We recorded the following data during sample collection: flock size, rearing system, feed type, vaccination, biosecurity, age of birds and type of birds (broiler/layer). We recognized those flocks as larger flock, which had more than 1,000 chicken and smaller flock with less than 1,000 chicken. We termed those feed as a 'loose feed' that were formulated readily in the farm by mixing different feed ingredients and a 'ready feed' that brought from commercially available feed company in the form of mash, crumble or pellets and fed directly to chickens without mixing any ingredients in the farm.

| Cultivation and isolation of non-typhoidal
Salmonella from the cloacal swab

| Cultivation of Salmonella
Each swab was inoculated separately into the freshly prepared nutrient broth and marked appropriately. Then, these were incubated at 37°C for 24 hr aerobically in a bacteriological incubator. The incubated tubes were then examined for bacteria growth. After that, the organism was inoculated into Salmonella-Shigella (SS) agar plate and incubated at 37°C overnight. The colonies on primary culture were subcultured by the streak plate method until the pure culture with homogenous colonies were obtained (Cheesbrough, 1987).

| Isolation of Salmonella
Salmonella inoculum was inoculated in SS agar by streak plate technique to obtain isolated colonies (Cheesbrough, 1987). The method was repeated as many times as necessary to obtain a culture containing singe colonies only and usually at least two or more times to ensure purity.

| Identification of Salmonella
We identified Salmonella based on their cultural characteristics, colony character, morphology, Gram's staining, motility and biochemical test. Shape, size, surface texture, edge, elevation, colour and opacity were observed and recorded after 24 hr of incubation for characterizing colony morphology. The Salmonella colonies were stained using Gram's staining method (Merchant & Packer, 1967).
The motility test was done for the separation of motile and nonmotile Salmonella (Cheesbrough, 1987).

| Characterization of Salmonella
We characterized the isolated Salmonella by using the following  (Cown, 1985).

| Antibiogram study of the isolated Salmonella
We performed an antibiotic susceptibility test of Salmonella isolates against eight antimicrobial agents by disc diffusion methods, as stated by the guidelines of Clinical and Laboratory Standard Institute (CLSI, 2012). A total of 20 samples were used for the antibiogram study. Sensitivity and resistance of the isolates were determined against streptomycin, penicillin, gentamicin, ampicillin, ciprofloxacin, amoxicillin, nalidixic acid and sulfamethoxazole-trimethoprim. The antimicrobial discs were dispensed onto the surface of Muller Hinton agar plates using sterile forceps, keeping a distance of about 1cm apart. Within 30 min after applying the discs, the plates were incubated at 37°C for 18 hr in an inverted position. Three or four different discs were placed on one plate. Each disc was pressed down to ensure complete contact with the agar surface. After incubation, each plate was examined. The diameters of the zone of inhibition were measured using a meter ruler. The zone margin was taken as the area showing no obvious, visible growth that can be detected with the unaided eye. The zone of inhibition was interpreted as sensitive, intermediate and resistant, according to CLSI guideline (CLSI, 2012). Any isolate resistant to at least three classes of antimicrobials were considered as multidrug resistant (Magiorakos et al., 2012). The zone of diameter interpreted as the standard for Salmonella is mentioned in Table 1.

| Maintenance of stock culture
For further study, it was necessary to preserve the Salmonella isolates. For this purpose, pure culture of isolated Salmonella was preserved in 50% sterile buffered glycerine and stored at −20°C.

| Data analysis
We calculated the prevalence of non-typhoidal Salmonella at the farm level by dividing culture-positive samples by the total number of tested samples. We also performed bivariate logistic regression analysis to identify the association between the non-typhoidal Salmonella and the variables of interest. The odds ratio (OR) with 95% confidence interval (CI) at 0.05 significance level was estimated to measure the degree of association. Data collected from a questionnaire survey (from the respective study farm) and laboratory study were entered into a Microsoft Excel sheet and analysed using STATA version 13 (Stata Corp & L., 2013).

| Cultural findings
After cultural examination, we found that the positive samples

| Staining and motility test
Morphological characterization revealed that the isolates were Gram-negative, short, plump, rod-shaped organism, arranged in a single or paired. In the motility test, we found that they were nonmotile ( Figure 4).

| Biochemical tests
After biochemical examination, we observed that all of the isolates fermented dextrose, maltose and mannitol and produced acid and gas but did not ferment sucrose and lactose. Additionally, all the isolates were positive to the methyl red test, catalase, TSI agar slant reaction and Simmon's citrate agar slant reaction, but negative to indole test and Voges-Proskauer test (Table 5 and Figure 5).

| Antibiotic sensitivity test
From the antibiogram study against eight different antibiotics, it was revealed that the resistance patterns for Salmonella isolates were 100% to penicillin and nalidixic acid, 55% to sulfamethoxazoletrimethoprim, 40% to ampicillin, 25% to amoxicillin, 20% to streptomycin and 5% to gentamicin and ciprofloxacin. On the other hand, the sensitivity pattern of the isolates was 90% to ciprofloxacin, 80% to gentamicin, 75% to amoxicillin, 70% to streptomycin and 45% to sulfamethoxazole-trimethoprim and ampicillin (Figures 6 and 7).

| D ISCUSS I ON
In this study, we determined the prevalence of poultry-specific non-typhoidal Salmonella and it's antibiotic susceptibility patterns from apparently healthy chickens collected from different poultry farms of Bangladesh. The overall prevalence of non-typhoidal Salmonella in this study was 41%. This finding was almost in agreement with the report of (Alebachew & Mekonnen, 2013), who reported 41.9% Salmonella infection among chicken flock in Jimma town, Ethiopia (Alebachew & Mekonnen, 2013). However, the present finding was lower than the findings of Parbati et al. (2017) and Naurin et al. (2012), who reported 53.33% and 52% prevalence of Salmonella in chickens, respectively (Naurin et al., 2012;. Our findings were higher than the findings of Bhuyan et al. (2010), who recorded a 16.52% prevalence of Salmonella in poultry. Similarly, Alam et al. (2003) reported a 23.8% prevalence of Salmonella infection in poultry in the Dinajpur district of Bangladesh. The prevalence may vary due to differences in the origin of samples, the technique used or due to different environmental conditions. There was a difference in the prevalence of Salmonella infection in different areas in our study. We found a higher prevalence of non-typhoidal Salmonella in chicken, providing loose feed than those provided with ready feed. Research has shown that changes in feed by modifying ingredients and composition of nutrients have an effect on the sensitivity of chickens to Salmonella infection (Vandeplas et al., 2010).
Similarly, Bhuyan et al. (2010) reported a variation in the prevalence of Salmonella in different areas, such as in Gazipur (20%), Manikgonj (16%) and Saver (15%) of Bangladesh. The prevalence of Salmonella was 41.7% in broiler and 40% in the layer. This finding was supported by other studies where the prevalence of Salmonella in broiler and layer was 41.3% and 46.2%, respectively (Alebachew & Mekonnen, 2013). However, a higher prevalence of Salmonella was found in a study where the prevalence was 71.11% in broiler and 38.8% in layer chickens (Naurin et al., 2012). Flock size also influenced the prevalence of Salmonella infection in our study. We found a higher prevalence in the larger flock (47.7%) compared to the smaller flock (32.7%). This finding is in agreement with the findings of another study in Bangladesh where they reported a higher prevalence (34.2%) of Salmonella in large flocks (≥5,001 birds) and lower prevalence (21.3%) in small flocks (≤1,000 birds) (Hossain et al., 2010). The highest infection rate in larger flocks may be due to the high flock density, which facilitates the easy spread of any infection.
Emerging antimicrobial resistance in the food-borne bacterial isolates is a major public health concern. Over the past 30 years, extensive use of antibiotics in livestock has led to increased antibiotic resistance in various bacterial strains (Mölstad et al., 2017).
We found that the sensitivity pattern for streptomycin was 70%.
Similar findings have also been reported in other studies where the isolates were 80% sensitive to streptomycin ( Ramya & MadhavaraoTirupati, 2013).
The test organisms in our study were Gram-negative short, rodshaped and mostly occurred singly or occasionally paired, which also corresponded to morphological characters of Salmonella as described in other study ( Cheesbrough, 1987). In most instances, we found that the test organisms were non-motile. Salmonella Gallinarum and Salmonella Pullorum are non-motile, whereas other poultry Salmonella spp. are found to be motile (Cheesbrough, 1987;Christensen et al., 1993). We found the organism was grown on a different media where they produced circular, smooth, opaque and translucent colonies on NA; black centred and small round on SS agar; translucent pink colony surrounded by a pink zone on BGA; pale, smooth, transparent and raised colonies on MacConkey agar; large, colourless colonies on EMB agar media and on TSI agar slant, black colony against a yellowish background were produced which was corresponded to the findings of others studies (Buxton & Fraser, 1977;Cheesbrough, 1987). The isolates fermented dextrose, maltose, and mannitol and produced both acid and gas, which was corresponded to the findings of others (Hasan et al., 2010;Merchant & Packer, 1967). Both indole and Voges-Proskauer tests were negative, but methyl red, catalase, TSI agar slant reaction and Simmon's citrate agar slant reaction were positive, which are almost similar to the findings of Buxton & Fraser (1977).

| CON CLUS ION
This study's results evidenced the occurrence of host-specific Salmonella serovars in commercial chicken production in Bangladesh, indicating that even apparently healthy chickens could be an important source of salmonellosis for chickens. Proper hygiene and disinfection practices at the farm-level could be effective in the overall reduction of Salmonella. A considerable proportion of Salmonella isolates was found to be resistant to different classes of antimicrobial drugs that could have a significant impact on public health if the resistance mechanisms are transferred into other serovars of zoonotic significance. Therefore, the regulation of the irrational use of antimicrobials in chickens must be addressed, including the restriction of antimicrobial supply in the illegal market.

ACK N OWLED G M ENTS
We wish to express our sincere gratitude and thanks to Professor

CO N FLI C T S O F I NTE R E S T
The authors declare that they have no conflicts of interest.

Pe e r Rev iew
The peer review history for this article is available at https://publo ns.com/publo n/10.1002/vms3.440.