Genomics of human and chicken Salmonella isolates in Senegal: Broilers as a source of antimicrobial resistance and potentially invasive nontyphoidal salmonellosis infections.

Salmonella enterica is the most common foodborne pathogen worldwide. It causes two types of diseases, a self-limiting gastroenteritis and an invasive, more threatening, infection. Salmonella gastroenteritis is caused by several serotypes and is common worldwide. In contrast, invasive salmonellosis is rare in high-income countries (HIC) while frequent in low- and middle-income countries (LMIC), especially in sub-Saharan Africa (sSA). Invasive Nontyphoidal Salmonella (iNTS), corresponding to serotypes other than Typhi and Paratyphi, have emerged in sSA and pose a significant risk to public health. We conducted a whole-genome sequence (WGS) analysis of 72 strains of Salmonella isolated from diarrheic human patients and chicken meat sold in multipurpose markets in Dakar, Senegal. Antimicrobial susceptibility testing combined with WGS data analysis revealed frequent resistance to fluoroquinolones and the sulfamethoxazole-trimethoprim combination that are among the most used treatments for invasive Salmonella. In contrast, resistance to the historical first-line drugs chloramphenicol and ampicillin, and to cephalosporins was rare. Antimicrobial resistance (AMR) was lower in clinical isolates compared to chicken strains pointing to the concern posed by the excessive use of antimicrobials in farming. Phylogenetic analysis suggested possible transmission of the emerging multidrug resistant (MDR) Kentucky ST198 and serotype Schwarzengrund from chicken to human. These results stress the need for active surveillance of Salmonella and AMR in order to address invasive salmonellosis caused by nontyphoidal Salmonella strains and other important bacterial diseases in sSA.

Introduction

The modern human lifestyle characterized by frequent traveling and globalization of trade and exchange of goods, including food products, increases the rapid spread of infectious disease outbreaks. The current Covid-19 coronavirus crisis represents a dramatic example of the threat represented by infectious diseases in a globalized world. Addressing the problems caused by or related to infectious diseases in modern societies requires integrated, collaborative, and multidisciplinary efforts. Surveillance systems based on a One Health approach play an essential role in monitoring pathogenic parasites, viruses, bacteria, and antimicrobial resistance (AMR) determinants relevant to human, animal, and environmental health.

Salmonella enterica is the most common foodborne pathogen worldwide. It comprises over 2600 serotypes [1] classified according to three major surface antigenic determinants [2]. S. enterica causes two types of disease: typhoid fever and nontyphoidal gastroenteritis. Salmonella gastroenteritis is restricted to the intestinal tract of the host [3]. Typhoid fever is more life-threatening and occurs when the bacteria invade the systemic compartment of the organism. Typhoid fever is caused by human specific serovars Typhi and Paratyphi A. As for many infectious diseases, the incidence, prevalence, and outcome of Salmonella infections drastically differ between rich and poor regions of the globe. While Salmonella gastroenteritis is still a concern in both high-income (HIC) and low- and middle-income countries (LMIC), typhoid fever cases are very rare in the former. Additionally, it is known that nontyphoidal Salmonella (NTS) serovars can potentially cause systemic infections termed invasive nontyphoidal salmonellosis (iNTS) [4]. Typhoid fever and iNTS are estimated to cause over 25 million cases and nearly 900,000 deaths annually [5]. In sSA, several multi-country surveillance programs [6] as well as sporadic studies [7-9] have generated a wealth of epidemiological data on the prevalence and incidence of typhoidal and iNTS. Available data indicate that iNTS is a significant cause of invasive disease in sSA, with Typhimurium and Enteritidis being the most frequently isolated serovars [5, 10, 11]. This is of noticeable importance given that these serotypes can infect multiple hosts, and are transmissible to humans through various sources, including food, animals, and the environment. Food animals (poultry, cattle and pigs) and their products (including eggs, milk and pork) constitute important sources of Salmonella contamination [12-15]. These animals are healthy asymptomatic carriers of many Salmonella serotypes, which can cause clinical disease in humans. These carriers serve as reservoirs for dissemination of not only the bacteria, but also of the antimicrobial resistance (AMR) determinants they harbor.

AMR is a global public health threat. Multidrug resistant (MDR) clones of important human pathogens are emerging at an alarming rate, and experts agree that if proper actions are not taken, the number of deaths due to resistant bacteria will increase in the future. Dissemination of AMR is due to several causes and influenced by adaptive pressures instituted from antimicrobial usage. The emergence of antimicrobial resistant Salmonella strains was first reported in the 1950s, soon after the introduction of chloramphenicol for the treatment of salmonellosis [16]. Since then, the prevalence of resistant clones in humans, domestic animals, and other wildlife species have expanded globally [17]. MDR clones of Salmonella, defined as co-resistance to the traditional first-line drugs chloramphenicol, ampicillin, and sulfamethoxazole-trimethoprim [18, 19], started emerging in the 1980s, necessitating the use of the second-line drugs fluoroquinolones and third-generation cephalosporins [20]. The selective pressure exerted by inappropriate use of antimicrobials, particularly as growth-promoting biologics for farm animals in countries where there is no regulation banning this practice, has led to the emergence of extensively drug resistant (XDR) clones defined as non-susceptibility to at least one agent in all but two or fewer antimicrobial categories [21]. For example, the XDR Kentucky serovar ST198 emerged in Egypt ~1989 following the acquisition of a genomic island harboring several AMR genes and spread widely throughout the world, acquiring additional resistance-conferring mutations [22]. Kentucky ST198 was associated with clinical infections in humans in different regions of the globe [22]. Currently, there is an ongoing epidemic of XDR serovar Typhi that emerged in Pakistan, which has been introduced to the USA by individuals returning from this affected region [23]. These emerging MDR and XDR clones of Salmonella represents a serious public health threat, especially in sSA, where there is a lack of laboratories suitably equipped to perform susceptibility testing. Consequently, when available, antimicrobial therapy is often empirical, which could promote the selection of resistant clones. Surveillance systems that monitor AMR are scarce in sSA. Available data indicate that resistance to first-line, cheap antimicrobials emerged in the 1980s [24]. Recent reports described emergence of MDR clones, which express resistance to the most efficient antimicrobials against Salmonella, such as third-generation cephalosporins and fluoroquinolones [20, 24]. In Senegal, strains resistant to antimicrobials used for treatment of salmonellosis, including MDR clones, have been isolated from human, food, and animal samples [14, 15, 25, 26]. In this study, we performed whole genome sequencing of 72 Salmonella isolates recovered from humans (n = 19) and chicken meat (n = 53) during a surveillance project in Dakar, Senegal [14]. The aim of our study was to determine the resistance profile of nontyphoidal Salmonella strains from clinical salmonellosis cases in humans, and the potential transmission of these strains from retail chicken meat. We report on the analysis of serotype distribution, phylogenic relationship, phenotypic and genotypic AMR profiles, and virulence and plasmid composition of isolates from the two sources analyzed.

Materials and methods

Sampling, culture, and identification of Salmonella isolates

Salmonella isolates from humans were recovered from diarrheic stools of patients visiting the clinical microbiology laboratory of the Institut Pasteur de Dakar, Senegal between July 2012 and June 2013. Isolates from chicken were recovered, during the same period, from carcasses sold at multipurpose markets in Dakar, Senegal. Carcasses were placed in sterile bags with ice packs, immediately sent to the laboratory for microbiological analysis, and processed as previously described [14]. Briefly, 25 mg of mixed parts of the chicken carcasses (part of the neck, thighs, anus, and skin) were removed, ground, mixed with 250 ml of buffered peptone water, and incubated at 37°C overnight. Then, 0.1 ml of the sample was used to inoculate 10 ml of Rappaport-Vassiliadis enrichment medium for Salmonella and incubated at 42°C for 24 hours. Dilutions of the cultures were spread onto XLD and Hektoen plates. Salmonella colonies were recovered and confirmed by biochemical testing (Api 20E, ref 20100; Biomerieux, France). Human and chicken isolates were stored in glycerol stocks until used in further biological analysis.

Antimicrobial susceptibility testing

Susceptibility to antimicrobial drugs was tested using a standard disc diffusion method on Mueller–Hinton agar plates (Oxoid) and using diameters defined by the Clinical and Laboratory Standards Institute (CLSI). The following 22 antimicrobials (or combination of antimicrobials) were tested: chloramphenicol (Cm), sulfamethoxazole + trimethoprim (ST), ampicillin (Ap), ticarcillin (Tc), amoxicillin + clavulanic acid (AC), aztreonam (At), cefalothin (Cf), cefoxitin (Cx), cefotaxime (Ct), ceftazidime (Cz), cefepime (Fp), imipenem (Ip), gentamicin (Gm), kanamycin (Kn), tobramycin (Tm), nalidixic acid (Na), ciprofloxacin (Cp), norfloxacin (No), ofloxacin (Ox), tetracycline (Te), colistin (Co), and erythromycin (Er).

Extraction of bacterial genomic DNA, whole genome sequencing, and assembly

Bacterial genomic DNA was extracted using the MasterPure™ Gram Positive DNA Purification Kit according to the recommendation of the manufacturer. This kit was used because, in our hands, it consistently yielded Salmonella genomic DNA of better quality for sequencing on an Illumina platform. The quality and concentration of the extracted DNA were determined using a NanoDropTM 2000/2000c Spectrophotometer (Thermo Fisher Scientific, Waltham, Massachusetts) and a Qubit 3.0 Fluorometer (Thermo Fisher Scientific). DNA libraries were prepared using the Nextera XT DNA Library Preparation Kit (Illumina, San Diego, CA) following the manufacturer’s instructions. The resulting DNA libraries were purified using AMPure XP beads (Beckman Coulter, Sharon Hill, PA) and re-quantified using the Qubit 3.0 Fluorometer (Thermo Fisher Scientific). Sequencing was performed on the MiSeq System using v2 sequencing reagent kits (Illumina).

Whole genome sequence forward and reverse reads were assembled de novo with Shovill v1.1.0 using SPAdes v3.14.1 [27, 28]. De novo assembly of forward and reverse reads was repeated for plasmid detection using plasmidSPAdes genome assembler v3.14.1 [29]. Plasmid contigs (in fasta file format) were further analyzed for antimicrobial resistance genes, virulence factors, and plasmid replicons using Abricate 1.0.1 [30]. Abricate databases include NCBI AMRFinderPlus (doi:10.1128/AAC.00483-19), Comprehensive Antibiotic Resistance Database (CARD) (doi:10.1093/nar/gkw1004), Resfinder (doi:10.1093/jac/dks261), Virulence Factor Database (VFDB) (doi:10.1093/nar/gkv1239), PlasmidFinder (doi:10.1128/AAC.02412-14), and MEGARES 2.00 (doi:10.1093/nar/gkz1010). Genes detected with a minimum 90% coverage and identity were reported.

Serotype prediction, detection of antimicrobial resistance genes, and phylogenetic analysis

Salmonella serotypes were predicted by submitting the contigs from genome assembly to the SeqSero 2 (http://denglab.info/SeqSero2) platform [31]. Sequence types were determined using the MLST 2.0 software [32]. Antimicrobial resistance genes and alleles with a minimum of 90% coverage and identity were detected from assembled contigs using ResFinder 4.1 software [33, 34] at the Center for Genomic Epidemiology (www.genomicepidemiology.org). For phylogenetic analysis, compiled genome assembly SNP alignments were constructed using CSI Phylogeny 1.4 from the Center for Genomic Epidemiology [35] with all parameters set at default. Generated phylogenetic tree was visualized using R Studio software (http://www.rstudio.com/).

Results

Demographic data

We performed WGS of 72 Salmonella isolates that represented a subset of strains collected during a surveillance project between July 2012 and June 2013 in Senegal. The isolates included 19 and 53 strains originating from human and retail chicken meat, respectively (Table 1, see Materials and methods).

Table 1

Serotypes and origin of Salmonella isolates.

Serotypes	Human	Chicken	Total
Brancaster	0	14	14
Kentucky	1	12	13
Hadar	0	11	11
Chester	0	4	4
Schwarzengrund	1	3	4
Senftenberg	0	4	4
Banana	3	0	3
Gaminara	2	0	2
Johannesburg	0	2	2
Isangi	1	0	1
Give	1	0	1
Poona	1	0	1
Corvallis	1	0	1
Somone	1	0	1
Muenster	1	0	1
Baildon	1	0	1
Oranienburg	1	0	1
3,10:e,h:-	1	0	1
Virchow	1	0	1
Rissen	1	0	1
Okerara	1	0	1
Typhimurium	0	1	1
Brandenburg	0	1	1
Vejle	0	1	1
Total	19	53	72

Serotype distribution

Analysis of the WGS data revealed 24 different serotypes, the most represented being Brancaster (14 isolates), Kentucky (13 isolates), Hadar (11 isolates), Chester (four isolates), Schwarzengrund (four isolates), and Senftenberg (four isolates) (Table 1). Serotype distribution within groups showed that isolates from humans were diverse, with 16 serotypes for 19 strains (Table 1). Only serotype Banana (three isolates) and Gaminara (two isolates) were found in more than one human sample. Interestingly, the Banana isolates were recovered in October 2012 (two isolates) and July 2013 (one isolate), and the two Gaminara isolates in September and November 2013, suggesting a possibility of diarrheic outbreaks, likely of low intensity, caused by these serotypes. Additionally, all but two (Kentucky and Schwarzengrund) of the serotypes from humans were not found in chicken. A phylogenetic analysis showed that the serotypes from stools were genetically diverse, confirming the ability of various Salmonella serovars to cause diarrheal disease in humans (Fig 1). Contrary to strains from humans, there was less serotype diversity in isolates from chicken meat, with 10 different serovars for 53 strains (Table 1). Serotypes Brancaster (14 isolates), Kentucky (12 isolates), and Hadar (11 isolates) predominated among these bacteria.

Fig 1

Phylogenetic analysis of Salmonella strains from human and chicken.

SNPs from genome assemblies were compiled and used to build alignment and generate a phylogenetic tree CSI Phylogeny 1.4 with all parameters set at default. The generated tree was visualized using R Studio. Isolates belonging to Kentucky sequence types ST198 and ST314 that formed two distinct clades are shaded in blue.

Phylogenetic analysis of Salmonella isolates from human and chicken

A concatenated SNP-based phylogeny was performed to evaluate the relationship among the 72 strains. Not surprisingly, the isolates belonging to the same serotypes mostly clustered together (Fig 1). The only exception was serotype Kentucky, whose 13 members were split in two distinct clusters of nine and four isolates, respectively (Fig 1). Isolates of each of the two clusters shared the same sequence types (ST) that were ST314 and ST198 for the first and second clades, respectively. Serotype Kentucky of the ST198 represents a lineage that recently emerged following the acquisition of a genomic island harboring multidrug resistant determinant [22] and disseminate throughout the world (see below). Interestingly, there were two phylogenetic clades composed of the four Kentucky ST198 in one hand and the four isolates of serotype Schwarzengrund in the other, which contained each, human and chicken isolates (Fig 1), suggesting a possible transmission to humans of these bacteria, likely through consumption of chicken meat.

Antimicrobial resistance profile of Salmonella isolates

All isolates were tested for susceptibility to 22 antimicrobials belonging to 11 classes (Table 2). Resistance to the historical first line Salmonella treatment was rare for chloramphenicol (1.4%, [1/72]) and ampicillin (4.2%, [3/72]), and frequent for the trimethoprim-sulfamethoxazole anti-folate combination (38.9%, [28/72]), which was mostly isolated from chicken (27/28) with only one human isolate displaying this phenotype (Table 2). Regarding the second-line drugs for Salmonella treatment, resistance to cephems was rare, with only one human isolate resistant to first (cefalothin), second (cefoxitin), and third-generation (cefotaxime and ceftazidime) cephalosporins, while all the tested isolates were sensitive to cefepime that belongs to the fourth-generation of this drug class and to imipenem, a carbapenem. In contrast, resistance to fluoroquinolones was frequent (19.4%, [14/72] for ciprofloxacin; 18.1%, [13/72] for ofloxacin) except for norfloxacin (4.2%, [3/72]) (Table 2). Besides these antimicrobials, the drugs for which the isolates were most frequently resistant were tetracycline (48.6%; [35/72]) and erythromycin (30.6%, [22/72]). No resistance was observed for imipenem, and one isolate displayed resistance to aztreonam (Table 2).

Table 2

Resistance of human and chicken isolates of Salmonella to different classes of antimicrobials.

Classes	Antimicrobials	Origin and number of resistant isolates		Total
		Human	Chicken
Phenicol	Chloramphenicol	1	0	1
Anti-folate	Sulfamethoxazol +Trimethoprim	1	27	28
Penicillin	Ampicillin	2	1	3
	Ticarcillin	2	1	3
	Amoxicillin + Clavulanic Acid	1	0	1
Monobactam	Aztreonam	1	0	1
Cephem	Cefalothin	1	0	1
	Cefoxitin	1	0	1
	Cefotaxime	1	0	1
	Ceftazidime	1	0	1
	Cefepime	0	0	0
Carbapenem	Imipenem	0	0	0
Aminoglycoside	Gentamicin	2	3	5
	Kanamycin	0	2	2
	Tobramycin	1	0	1
Quinolone	Nalidixic Acid	1	14	15
	Ciprofloxacin	1	14	15
	Norfloxacin	1	2	3
	Ofloxacin	1	12	13
Cycline	Tetracycline	1	35	36
Polymixin	Colistin	0	0	0
Macrolide	Erythromycin	2	22	24

Multidrug resistance, defined in Salmonella as co-resistance to the historical first-line drugs including chloramphenicol, ampicillin, and trimethoprim-sulfamethoxazole combination [36] was found in only one strain, a human isolate of serotype Poona that was resistant to 13/22 antimicrobials tested (Table 1). However, 27.8% (20/72) of the isolates displayed resistance to at least three antimicrobial classes, a common definition of MDR phenotype [37].

Interestingly, AMR was more frequent in isolates from chicken. Indeed, 89.5% (17/19) of the human isolates were sensitive to the 22 antimicrobials tested, while only 17% (9/53) of the chicken strains displayed this feature (Table 1). This observation points to the selection pressure exerted by the use of antimicrobials as a growth promoter in chicken farming.

Whole-genome sequencing (WGS) analysis

Twenty-seven known antimicrobial resistance genes (ARG) and four point mutations previously described as conferring resistance to quinolone, including two on parC and two on gyrA genes, were detected from the WGS data (Table 3). The ARGs encoding resistance to different classes of AMR factors, including aminoglycoside inactivating enzymes, anti-folates, β-lactamases, quinolone resistance proteins, rifamycin inactivator, fosfomycin inhibitor, and efflux pumps, were detected.

Table 3

Antimicrobial resistance profile and resistance gene content of 72 human and chicken isolates of Salmonella.

Serotypes	AMR Phenotypic Profile	ARG&	Plasmidic ARG
Brancaster	(Kn) Na Cp Ox Te Er	sul2 tetB aph(3’)-Ib aph(3")-Ib aph(6)-Id qnrB19	sul2 tetB qnrB19 tetD
Brancaster	(Kn) Na Cp Ox Te Er	sul2 tetB aph(3’)-Ib aph(3")-Ib aph(6)-Id qnrB6	qnrB19
Brancaster (X 4)#	ST Te	sul2 dfrA1 tetB aph(3’)-Ib aph(3")-Ib aph(6)-Id
Brancaster (X 4)#	ST Te	sul2 dfrA1 tetB aph(3’)-Ib aph(3")-Ib aph(6)-Id	sul2 dfrA1 aph(3")-Ib aph(6)-Id
Brancaster	Kn (Tm) Te Er	sul2 tetB aph(3’)-Ib aph(3")-Ib aph(6)-Id
Brancaster	ST (Kn) Na Cp Te Er	sul2 dfrA1 dfrA15 tetA tetB aph(3’)-Ib aph(3")-Ib aph(6)-Id qnrB19	sul2 dfrA1 tetA aph(3")-Ib aph(6)-Id qnrB19
Brancaster	Te	sul2 tetB aph(3’)-Ib aph(3")-Ib aph(6)-Id
Brancaster	(Kn) Te Er	sul2 tetB aph(3’)-Ib aph(3")-Ib aph(6)-Id
Kentucky ST198	Ap Tc Gm (Tm) Na Cp No Ox Te Er	aac(3)-Id aadA7 tetA sul1 blaTEM-1b parC-S80I gyrA-S83F, gyrA-D87N
Kentucky ST198	ST Na Cp No Ox Te Er	sul2 dfrA14 tetA aph(6)-Id parC-S80I	sul2 dfrA14 tetA aph(6)-Id
Kentucky ST198	ST Ap Tc AC Gm Na Cp No Ox Te Er	aac(3)-Id aadA7 blaTEM-1B sul1 tetA parC-S80I gyrA-S83F, gyrA-D87N
Kentucky ST198	ST Na Cp Ox Te Er	sul2 dfrA14 tetA aph(6)-Id parC-S80I gyrA-S83F, gyrA-D87N	sul2 dfrA14 tetA aph(6)-Id
Kentucky ST314 (X 2)#	Er	sul1 dfrA15 *
Kentucky ST314	ST Gm Te Er	aac(3)-Id aadA7 sul1 dfrA15 tetA aph(3")-Ib aph(6)-Id
Kentucky ST314 (X 6)#	ST	sul1 dfrA15
Hadar (X 2)#	ST Na Cp (Ox) Te Er	sul2 dfrA1 tetA aph(3")-Ib aph(6)-Id qnrB19	sul2 dfrA1 aph(3")-Ib aph(6)-Id qnrB19
Hadar	Te Er	sul2 dfrA1* tetA aph(3")-Ib aph(6)-Id
Hadar (X 6)#	Te	tetA aph(3")-Ib aph(6)-Id
Hadar	ST Gm (Tm) Na Cp Ox Te Er	sul2 dfrA1 tetA aph(3")-Ib aph(6)-Id qnrB19
Hadar	Te	tetA aph(3")-Ib
Chester	ST (Kn) Na Cp Ox Te Er	sul2 dfrA14 tetA aph(6)-Id qnrB19
Chester	sensitive
Chester	ST Na Cp Ox Te Er	sul2 dfrA14 tetA aph(6)-Id qnrB19	sul2 dfrA1 tetA aph(6)-Id
Chester	ST Na Cp Ox Te Er	sul2 dfrA14 tetA aph(6)-Id qnrB19
Schwarzengrund	sensitive
Schwarzengrund (X 2)#	sensitive
Schwarzengrund	Er
Senftenberg	sensitive
(X 4) #
Banana (X 3)#	sensitive	fosA
Gaminara	sensitive
(X 2) #
Johannesburg	ST Na Cp Ox Te Er	sul1 dfrA7 tetA gyrA-S83F
Johannesburg	ST Kn Na Cp Ox Te Er	sul1 dfrA7 tetA gyrA-S83F
Isangi	sensitive
Give	sensitive
Poona	Cm ST Ap Tc AC At Cf Cx Cz Ct Gm Tm Er	aadA1 aadA2b ant(2")-Ia aph(3’)-Ia aph(3’)-Via arr-2 sul1 sul2 floR blaDHA-1 blaOXA-10 dfrA14	aph(3’)-Ia aph(3’)-VIa sul1 sul2 floR blaDHA-1
Corvallis	sensitive
Somone	sensitive	fosA
Muenster	sensitive
Baildon	sensitive
Oranienburg	sensitive	fosA
3,10:e,h:-	sensitive	fosA
Virchow	sensitive
Rissen	sensitive
Okerara	sensitive
Typhimurium	Ox Te Er	fosA tetA qnrB7	tetA qnrB7
Brandenburg	sensitive
Vejle	sensitive

ARG, antimicrobial resistance gene. Antimicrobials: AC, amoxicillin + clavulanic acid; Ap, ampicillin; At, aztreonam; Cf, cefalothin; Cx, cefoxitin; Ct, cefotaxime; Cz, ceftazidime; Cm, chloramphenicol; Cp, ciprofloxacin; Er, erythromycin; Fp, cefepime; Co, colistin; Gm, gentamicin; Ip, imipenem; Kn, kanamycin; Na, nalidixic acid; No, norfloxacin; Ox, ofloxacin; ST, sulfamethoxazole + trimethoprim; Tc, ticarcillin; Te, tetracycline; Tm, tobramycin.

&, gene aac(6’)-Iaa, not shown in Table 3, was present in all isolates without conferring AMR

*, sul and dfrA genes present in isolates susceptible to sulfamethoxazole + trimethoprim.

, Number of isolates that have the same resistance phenotype and ARG profile.

Isolates from humans are shaded.

Genes encoding aminoglycoside modifying enzymes were the most frequently found ARG with 10 genes detected in 34 isolates. However, four ARG belonging to the phosphotransferase subfamily (aph(3’)-Ia, aph(3’)-VIa, aph(3”)-Ib, aph(6)-Id) were not associated with any resistance. Similarly, one aminoglycoside acetyltransferase ARG (aac(6’)-Iaa) detected in all the isolates without conferring resistance, is a cryptic gene present in most Salmonella isolates [38]. The remaining aminoglycoside modification ARG included the nucleotidyltransferases aadA1, aadA2, and ant(2’)-Ia that were present together in a MDR human isolate of serotype Poona that displayed resistance to gentamycin and tobramycin. Additionally, aac(3)-Id (acetyltransferase) and aadA7 (nucleotidyltransferase) were co-detected in three gentamicin-resistant isolates of serotype Kentucky including two and one of sequence types ST198 and ST314 respectively (Table 3).

Six ARG targeting folate pathways were detected, including sul1 and sul2 that confer resistance to sulfonamides, and four genes of the dfrA family that confer resistance to trimethoprim. Of the 28/72 isolates displaying resistance to the sulfamethoxazole-trimethoprim combination, the presence of at least one member of each of these two anti-folate subfamilies was detected. However, three isolates belonging to serotype Kentucky ST314 (n = 2) and Hadar (n = 1) possessed these genes without displaying resistance to sulfamethoxazole-trimethoprim. Additionally, one isolate of Kentucky ST198 was resistant to sulfamethoxazole-trimethoprim while possessing a sul1 only without a dfrA gene (Table 3).

Three ARGs of the quinolone resistance family and four point mutations in parC and gyrA were detected in the isolates. One parC mutation (ParC T57S) was found in 69/72 isolates but was not associated with phenotypic quinolone resistance. All the 16/72 strains resistant to at least one quinolone possessed at least one qnrB ARG and/or a point mutation in parC and/or gyrA. Regarding other antimicrobials relevant to Salmonella treatment, one human isolate of serotype Poona resistant to chloramphenicol possessed a floR ARG as expected (Table 3). Similarly, four isolates displaying resistance to a β-lactamase harbored a corresponding gene. The extended spectrum β-lactamase (ESBL) blaOXA-10 and blaTEM-1 ARGs conferred resistance to the penicillin family (ampicillin, ticarcillin, amoxicillin+ clavulanic acid), while blaDHA-1 and blaCMY-2 provided resistance to first, second, and third classes cephalosporins (Table 3).

As expected, of the 36/72 tetracycline resistant isolates, 23 and 14 possessed a tetA or a tetB efflux pump respectively, while one harbored both. Noticeably, while we did not test phenotypic resistance to fosfomycin, seven isolates (six human and one chicken) harbored a fosA7 gene conferring resistance to this antimicrobial (Table 3).

Plasmid presence and content in Salmonella isolates

PlasmidSPAdes analysis detected putative plasmids in 55.6% (40/72) of the isolates, including 42.1% (8/19) and 60.4% (32/53) of the human and chicken strains, respectively (S1 Table). Plasmid presence tended to match with serotypes and was more frequent among the serotypes Brancaster (64.3%, [9/14]), Hadar (81.8%, [9/11]), Senftenberg (100%, [4/4]), and Schwarzengrund (75%, [3/4]) (S1 Table). Noticeably, all (4/4) the Kentucky isolates of sequence type ST198 harbored plasmids while those of ST314 (9/9) were devoid of a plasmid. The ColRNAI type was the most frequently found replicon (19 occurrences), followed by Col156 (10 occurrences) and ColpVC (nine occurrences), while different Inc types were found at low frequency (1 to 4) (S1 Table).

Plasmids containing ARGs were found in 15 genomes, including a single clinical strain of serotype Poona and 14 chicken isolates of serotypes Brancaster (n = 7), Kentucky ST198 (n = 2), Hadar (n = 2), Chester (n = 1), Schwarzengrund (n = 1), and Typhimurium (n = 1). Gene sul2 was the most frequent plasmidic ARG, being present in 13/15 isolates with this feature (Table 4). Interestingly, sul2 occurred with a drfA1 or dfrA14 genes in 12/15 cases suggesting a co-selection of determinant conferring resistance to the trimethoprim-sulfonamide anti-folate combination. Of note, the blaDHA1 and blaCMY-2 ARGs conferring resistance to first, second and third class cephalosporins were located on putative IncA/C and IncI1plamids, respectively (Table 4).

Table 4

Antimicrobial resistance and virulence genes found on plasmids of human and clinical isolates of Salmonella.

Serotypes	Plasmid replicons	ARG	Virulence genes
Brancaster	Col440I	sul2 tetB qnrB19 tetD	papI, papB
Brancaster	Col440I	qnrB19
Brancaster	ColRNAI	sul2 dfrA1 aph(3")-Ib aph(6)-Id
Brancaster	ColRNAI / Col440ICol156ColpVC	sul2 dfrA1 tetA aph(3")-Ib aph(6)-Id qnrB19
Brancaster	ColRNAI	sul2 dfrA1 aph(3")-Ib aph(6)-Id
Brancaster	ColRNAI	sul2 dfrA1 aph(3")-Ib aph(6)-Id
Brancaster	ColRNAI	sul2 dfrA1 aph(3")-Ib aph(6)-Id
Kentucky ST198	ColRNAI / Col440II	sul2 dfrA14 tetA aph(6)-Id
Kentucky ST198	ColRNAI / Col440II	sul2 dfrA14 tetA aph(6)-Id
Hadar	ColRNAI / Col440II Col156	sul2 dfrA1 aph(3")-Ib aph(6)-Id qnrB19
Hadar	ColRNAI / Col440I	sul2 dfrA1 aph(3")-Ib aph(6)-Id qnrB19
Chester	ColRNAI / Col440I Col156	sul2 dfrA1 tetA aph(6)-Id
Schwarzengrund	IncI2	sul2 tetA aph(3’)-Ia aph(3")-Ib aph(6)-Id	papI, papB, papH, papDJKEFG
	IncFIB
	IncP1
Typhimurium	IncX3 / IncY	tetA qnrB7
	ColRNAI / Col440
Poona	IncA/C	aph(3’)-Ia aph(3’)-VIa sul1 sul2 floR blaDHA-1
Gaminara	IncI1	ND	pic, astA
Gaminara	IncFII	ND	faeEDC
Johannesburg	ColRNAI / Col440II ColpVC	ND	papDJKEFGC

ARG, antimicrobial resistance gene; pap, pyelonephritis-associated pili; (pic) serine protease precursor [Pic (VF0232)] [Escherichia coli CFT073]; (astA) heat-stable enterotoxin 1 [EAST1 (VF0216)] [Escherichia coli O44:H18 042]; fae, gene encoding a component of the K88 fimbriae.

ND, no ARG detected.

Isolates from humans are shaded.

Besides ARGs, other relevant plasmid content interestingly included virulence genes of pathogenic E. coli (Table 4). One clinical isolate of serotype Gaminara harbored an IncI1 type plasmid encoding the Pic serine protease autotransporter found in enteroaggregative E. coli (EAEC) and in Shigella [39], and the EAST1 toxin found in various pathogenic E. coli and in some Salmonella [40] (Table 4). Another isolate of the same serotype harbored an IncFII type plasmid with the faeC, faeD, and faeE genes that are part of an operon encoding K88 fimbriae associated with virulence in swine of enterotoxigenic E. coli (ETEC) [41] (Table 4). Additionally, plasmids harboring genes of the pyelonephritis-associated pili (PAP) of uropathogenic E. coli (UPEC) [42] were detected in three chicken isolates, including one strain of serotype Schwarzengrund with a cluster of nine pap genes including regulatory and structural determinants, a Brancaster strain with the papI and papB regulatory genes without any structural gene, and a Johannesburg isolate with seven structural and accessory genes of the pap fimbriae but without the regulatory genes (Table 4).

Salmonella pathogenicity islands (SPI) virulence factor distribution

Within S. enterica core genome, SPI-1 and 2 are ubiquitous as they contain key virulence factors necessary for host cell invasion and intercellular survival, respectively. All isolates collected from both human clinical and meat samples contained virulence factors within SPI-1, 2, and 3 to support these functions. SPI-3 genes mgtCB and misL which aid intracellular survival in low nutrient environments by increasing affinity for magnesium and perpetuate intestinal colonization, respectively, were present in 100% of isolates. All isolates also contained the SPI-4 genes pipB and sopB which are suspected to effect SPI-1 and 2 expression for epithelial invasion, enteric salmonellosis, and chicken colonization. Pathogenicity virulence factors on other SPIs associated with invasive salmonellosis varied across several serotypes in this dataset (S2 Table). Typhoid toxin cdtB from SPI-11, proven to be significantly associated with high rates of invasive disease, was identified in 22% (16/72) isolates, 42% (8/19) clinical and 15% (8/53) chicken isolates (S2 Table). SPI-12 sspH2, conferring actin polymerization and enhanced cell-to-cell motility was found in 35% (25/72) isolates, 26% (5/19) clinical and 38% (20/53) chicken. SPI-24 genes shdA and ratB that propagate adhesion and host cell invasion also detected in 26% (19/72) and 21% (15/ 72) respectively (S2 Table).

Discussion

Invasive NTS continues to be an emerging threat to public health, especially for regions throughout sSA where larger populations exist with depressed host immunity either from malnutrition, lack of sanitation, and/or immunosuppression form Malaria, HIV or other chronic illnesses. S. enterica disease severity is attributed to the combination of host immunity and the strain fitness (AMR and virulence factors often horizontally acquired within SPI and/or vertically through plasmids). This dataset shows a wide distribution of AMR and virulence factors supportive of systemic invasive salmonellosis within several different serovars, isolated from both human diarrheal patients and from chicken meat sold at a market in Dakar, Senegal.

Salmonella transmission from chicken

In this study, we applied a genomics approach to analyze 19 human and 53 chicken isolates of Salmonella enterica collected during a surveillance project in Dakar, Senegal. We found a high serotype diversity in human isolates with 16 serovars represented in 19 strains. This observation shows the ability of many Salmonella serotypes to cause clinical infection in humans. Contrary to human strains, only 10 different serotypes were represented in the 53 chicken isolates analyzed with Brancaster (15 isolates), Kentucky (13 isolates), and Hadar (11 isolates) predominating in these strains. Noticeably, only serotypes Kentucky ST198 and Schwarzengrund were represented in both human (one isolate each) and chicken (13 and three isolates, respectively) strains. Chicken is an important source of Salmonella contamination to humans. Chicken products, including eggs, meat, and their derivative, are the cause of foodborne infection worldwide [43]. In Senegal, several studies reported Salmonella contamination of chickens, including retail meat from markets [15, 44], farm birds [45, 46] and cooked food [47, 48]. Many of the serotypes found in these studies can cause clinical infections in humans [15]. They are asymptomatically carried by chickens, which represent an important reservoir and a source of transmission. A few of these serotypes may cause mortality to newly hatched birds or egg loss but are well tolerated by adults [49]. They colonize different organs, including the reproductive tract, without compromising chicken growth and are thus vertically transmitted. Actually, the most threatening Salmonella serovar for poultry is Gallinarum, a chicken-restricted serotype that causes an acute systemic infection resulting in a high mortality rate [50]. To prevent losses, farmers heavily use antimicrobials to protect chicken against Salmonella [51]. In HIC, this concern is properly addressed by legislation and regulations applied to food manufacture and commercialization [52]. In contrast, such regulations do not exist in many LMIC, or when they exist, they are often not enforced. Additionally, poor hygiene is associated with food production in these countries, with street food being an important supply in many cities.

Although many studies have reported Salmonella contamination of chicken in Senegal, transmission from this source to humans has been rarely mentioned. This is primarily due to limited approaches in previous studies focused on species identification, periodic serotyping, and antimicrobial susceptibility testing. One report in 2001 described a new serotype, Keurmassar, that was isolated from chicken meat and a human clinical sample in the same time period suggesting potential transmission through the food chain [15]. The WGS approach used in this study, by revealing the whole genome, is better suited to compare isolates from different sources and detect transmission of Salmonella or other pathogens from food to humans. Whole genome phylogenetic analysis clearly shows one human clinical isolate (serotype Schwarzengrund) that shares the same phylogenetic clade as three isolates procured from chicken meat (Fig 1). Similarly, out of the five Salmonella Kentucky of the ST198 that clustered together in phylogenetic analysis, one was isolated from the stools of a diarrheic patient while the other four originated from meat (Fig 1). Although this does not correspond to a demonstration of direct transmission from chicken to humans, it is a likely possibility. The case of Kentucky ST198 is important since this strain emerged in Egypt in 1989 after acquiring a genomic island that harbors several AMR genes [53]. Kentucky ST198 is disseminated widely in the world and causes clinical infections in humans. MDR phenotypes are concerning for public health.

Animal reservoir of Salmonella and invasive nontyphoidal salmonellosis in sub-Saharan Africa

Besides chicken, Salmonella has been detected in several food animals in Senegal, including small ruminants [54], beef [55], and swine [56]. Additionally, Salmonella is well spread in Senegalese wildlife species and has been detected in vulture [57], bats [58], and turtles [59]. This endemic presence and the diversity of serovars that can cause clinical infection in humans represent an important public health concern in context of the rising frequency of iNTS in sSA [11]. Invasive NTS is a systemic disease with a high mortality rate caused by serotypes other than Typhi and Paratyphi that are restricted to the human host [10, 60]. This severe disease mostly affects children and immunocompromised adults, especially those infected with HIV [10]. Invasive NTS is also frequently associated with malaria [61]. Several reports have declared that Salmonella is becoming a leading cause of bacteremia in sSA [11]. Still, many laboratories in this region are not equipped to perform adequate surveillance and diagnostics; it is likely that many cases of iNTS are not reported. Therefore, there is an urgent need to establish a surveillance of Salmonella, especially invasive strains and trace their source of acquisition by humans.

Spread of antimicrobial resistant determinants in food animals and the need for a surveillance system to prevent transmission to humans

AMR was rare in human isolates with 17/19 strains susceptible to the 22 antimicrobials tested. In contrast, 48/56 of the chicken isolates displayed resistance to at least one antimicrobial. Importantly, our WGS data were consistent with the AMR phenotype of the isolates except in five cases. Three of these cases consisted of chicken isolates harboring a sul1 or sul2 and a dfrA ARG while sensitive to sulfamethoxazole-trimethoprim combination (Table 1). We do not know the explanation of this observation. It might be due to a low level of expression of one or both of these ARGs. Another case corresponded to a Schwarzengrund strain that harbored a tetA efflux pump while sensitive to tetracycline (Table 1). The fifth case was a Kentucky strain that had a sul1 gene only and was resistant to the sulfamethoxazole-trimethoprim combination without a dfrA gene. This strain might harbor a gene that confers resistance to trimethoprim, but that is unknown.

The high frequency of AMR isolates in chicken points to the use of antimicrobials in farming to prevent or treat infectious diseases and promote growth [45]. The extent of improper use of antimicrobial is not known. However, family units of chicken farming for family consumption or small-scale commercialization are widespread in Senegal. This leads to a continuous selection of MDR strains that can enter the food chain. In this study, we found a significant level of resistance to antimicrobials used for the treatment of clinical salmonellosis, including sulfamethoxazole-trimethoprim combination and fluoroquinolones. Fluoroquinolones were introduced to treat Salmonella infections after resistance emerged against the historical first-line molecules chloramphenicol, ampicillin, and sulfamethoxazole-trimethoprim [20]. Earlier studies in Senegal reported a low level of resistance to fluoroquinolones [62, 63]. More recent studies show Salmonella fluoroquinolone resistance is increasing [14, 25, 26] which is consistent with our results. Subsequently, β-lactams have recently become the first-line treatment against severe Salmonella infections in Senegal. Available data indicate that β-lactam resistance is still rare in Senegal, which is also consistent with our findings. A robust and integrated surveillance system is needed to monitor the emergence of AMR pathogens in order to identify potential sources, implement countermeasures, and reduce the transmission of resistance determinants to humans in Senegal. Such a surveillance can use the WGS methodology presented in this study. It can be based on a one health approach in which comparative genomics of Salmonella from human, animal, food and environmental origins will permit to identify sources of contamination and of AMR dissemination. This will contribute to keep active molecules available for treatment of bacterial infections.

Replicons, antimicrobial resistance and virulence genes harbored by human and chicken isolates of Salmonella.

(XLSX)

Click here for additional data file.

Invasive salmonellosis-associated virulence genes harbored by non-ubiquitous Salmonella Pathogenicity Islands in human and chicken isolates.

(XLSX)

Click here for additional data file.

24 Jan 2022

Submitted filename: Review Comments_PONE-D-21-24873.docx

Click here for additional data file.

2 Feb 2022

Responses to reviewers

Dear Editor,

We are submitting a revised version of manuscript PONE-D-24873 entitled “Genomic comparison of clinical human to chicken meat Salmonella isolates in Senegal: broilers as a source of antimicrobial resistance and potentially invasive nontyphoidal salmonellosis infections”. We have made all the revisions requested. We thank both reviewers for their comments and suggestions that significantly improved the quality of our manuscript. Below are the responses to the academic editor and reviewers.

Responses to the academic editor

1. Please ensure that your manuscript meets PLOS ONE's style requirements, including those for file naming. The PLOS ONE style templates can be found at

https://journals.plos.org/plosone/s/file?id=wjVg/PLOSOne_formatting_sample_main_body.pdf and

https://journals.plos.org/plosone/s/file?id=ba62/PLOSOne_formatting_sample_title_authors_affiliations.pdf

Done

2. Please upload a copy of Figure 2, to which you refer in your text on page 19. If the figure is no longer to be included as part of the submission please remove all reference to it within the text.

Figure 2 was mistakenly mentioned on page 19 of the submitted manuscript. It has been changed to “S2 Table”, page 19, lines 327, 329 and 332 of the revised manuscript.

3. Please include captions for your Supporting Information files at the end of your manuscript, and update any in-text citations to match accordingly. Please see our Supporting Information guidelines for more information: http://journals.plos.org/plosone/s/supporting-information.

Captions (SI Table and S2 Table) have been added for Supporting Information files at the end of the Discussion section, page 24, lines 438 and 440 of the revised manuscript.

4. Please review your reference list to ensure that it is complete and correct. If you have cited papers that have been retracted, please include the rationale for doing so in the manuscript text, or remove these references and replace them with relevant current references. Any changes to the reference list should be mentioned in the rebuttal letter that accompanies your revised manuscript. If you need to cite a retracted article, indicate the article’s retracted status in the References list and also include a citation and full reference for the retraction notice.

The reference list has been updated. All the added citations are mentioned in the responses to the reviewers.

5. We have made the additional changes:

• Line 343 of the original manuscript, line 336 of the revised version: “form” changed to “from”.

Responses to Reviewer 1

1. Throughout the manuscript, you have used ‘antibiotics’ and ‘antimicrobials’ interchangeably. Best to use one consistently – I think ‘antimicrobials’ would be more suitable for this type of manuscript that compares clinical (human) and animal isolates

The changes requested by the reviewer have been made lines 77, 96 and 233 of the revised manuscript.

2. Lines 74-75: Please revise sentence to read Food animals (poultry, cattle and pigs) and their products (including eggs, milk and pork) constitute ….

The changes requested by the reviewer have been made lines 67-68 of the revised manuscript.

3. Line 371-372: You could add to that statement on these lines a further statement to read, ‘or where such regulations exist, they are often not enforced’

The changes requested by the reviewer have been made line 365 of the revised manuscript.

Overall verdict: The paper is well written and offers analysis of AMR from a somewhat one-health approach (has human and animal samples). Future similar analyses will be strengthened by incorporating environmental and plant samples in an expanded dataset. I recommend it for publication once the above minor corrections have been addressed.

We thank the reviewer for his valuable suggestions that contributed to improving the quality of our manuscript.

Responses to Reviewer 2

The title and the introductory part (L22-31) in the abstract are too wordy

The title and the introductory part of the abstract have been shortened according to the reviewer’s suggestion (lines 21-27 of the revised manuscript).

Abstract

Line 37: cephalosporins are not historical first-line drugs

The corresponding sentence has been modified line 33 of the revised manuscript, to correct the mistake identified by the reviewer.

Lines 38-39: ‘..pointing to the concern posed by the excessive use of antimicrobials in farming.’: do the authors have data to sustain this statement, at least for Senegal?

There are several publications reporting surveys on antimicrobial use African countries. In the original manuscript, we mention a reference (Ref 41; Ref 45 in the revised version) lines 423-425 in the Discussion section (lines 416-417 of the revised manuscript) that reported a survey in chicken farms. We have further evidences from a survey (unpublished) made in 2020 on behalf of the AMR Working Group of the Permanent Secretariate of the High National Counsel for Global health Security Agenda (to which we are a member). The survey mentioned the following among others:

• Importers of antimicrobials (AM) distribute molecules without compliance to existing recommendations

• There is no control of the distribution

• 37.5% of the veterinary pharmacists provide AM without prescription

• 60% of the veterinary doctors prescribe AM treatment without antimicrobials susceptibility testing

• 50% of the veterinary doctors prescribe human AM for animal treatment

• Only 53% of the veterinary doctors follow guidelines for AM use in animals

• Etc.

Introduction

Lines 53-62: Introducing Salmonella and Samonella infections is too wordy

The introduction on Salmonella has been shorten according to the reviewer’s suggestion (lines 49-54 of the revised manuscript).

Line 65: The term “developing countries” is no longer preferred. Recommend using the World Bank classification of low- and middle-income countries. Same for “developed countries” (high-income countries). + Line 370 and 372

The changes requested by the reviewer have been made lines 24, 25, 57, 363 and 365 of the revised manuscript.

Lines 69-72: some other data are available (see hereafter)

Please rephrase this part (Lines 69-72) according to the supplied information

We thank the reviewer for bringing to our attention these important data. We have reformulated the pointed section, lines 61-65 of the revised manuscript and added a few of the suggested references (refs 6-9 of the revised manuscript).

Line 75: eggs and milk are not food animals but remained important vehicles of Salmonella, especially eggs

The changes requested by the reviewer have been made line 68 of the revised manuscript.

Lines 90-94: Still accurate for some parts of the world but banned in other parts (In 2006, the EU banned the use of antibiotics as feed additives for growth promotion). This sentence has to be nuanced

The sentence has been change lines 83-87 to incorporate the reviewer’s suggestion.

Line 103: which COULD promote

The changes suggested by the reviewer have been made line 96 of the revised manuscript.

M&M

Line 120: Human Salmonella isolates dated from 2012-2013. Same period for the isolates from chicken. Make it clear at line 123. It is clearly enounced at lines 175-176.

The changes suggested by the reviewer have been made line 114 of the revised manuscript.

Lines 132-139: For antibiotics abbreviations try to use some internationally recognized ones such as http://bsacsurv.org/science/antimicrobials/ OR from EUCAST https://www.eucast.org/fileadmin/src/media/PDFs/EUCAST_files/Disk_test_documents/Disk_abbreviations/EUCAST_system_for_antimicrobial_abbreviations.pdf. Interestingly, those abbreviations were only used later in Table 3

We do agree with the reviewer regarding the abbreviations used for antimicrobials. We used a two-letter code instead of the more commonly used three-letter code for a formatting purpose, in order to have all the information fit in Table 3. This Table is quite long and spans across three pages. The two-letter code helped minimize it length while avoid case breaking between pages.

Line 142: MasterPureTM Gram Positive DNA Purification Kit for Gram negative bacteria

This kit was used because it yielded from Salmonella genomic DNA of better quality for sequencing using an Illumina platform. This was added lines 137-139 in the revised manuscript.

Results

Lines 178-180: these are repetitions of what was mentioned in the M&M section

We deleted the sentences mentioned from the Results section.

Line 193: were absent in chicken Not appropriate: were not found in Salmonella isolates from chicken.

The changes suggested by the reviewer have been made line 186 of the revised manuscript.

Line 197: yes, clearly, serotypes Kentucky and Virchow are known to be closely associated with poultry

Agree.

Lines 272-273 + 420-421: Additionally, one isolate of Kentucky ST198 was resistant to sulfamethoxazole-trimethoprim while possessing a sul1 only without a dfrA gene. We made the same observation in S. maltophilia isolates where the dfrA gene was absent and where the dihydrofolate reductase was mutated and was conferring resistance to trimethoprim (not yet published, only in biorxiv)

We are very grateful to the reviewer for sharing this discovery. We would be very interesting to know more about the mentioned biorxix manuscript. We are trying to identify the genetic determinant supporting the resistance to trimethoprim and would be interested to collaborate with the reviewer in this topic.

Lines 274-278: would have be better if MIC to fluoroquinolone antibiotic had been performed

We do agree with the reviewer. Unfortunately, a MIC system is currently not available in our laboratory to perform the suggested analysis.

Lines 296-298: I do not know about the performance of plasmidSPAdes algorithm for assembling large plasmids from whole genome sequencing data, but the most detected plasmids were only from the Col plasmid family (high copy number and small in size <10kb). Could this affect the results for larger plasmids from the IncH or IncI or F? If yes, this should be mentioned in the Limitations section. PlasmidFinder also allows to find replicons from WGS data.

We tested both plasmidSpades and PlasmidFinder and found the former to give results that are more reliable since it permits to get most of the plasmid sequences. In other genomics studies, we recovered sequences of large plasmids (the biggest being +300 kb) using plasmidSpades. We did not find any bias regarding plasmid size in the strains analyzed in this study and in others not published yet. We will be happy to share more on our experience using plasmidSpades.

Discussion

Line 348: a local market in Senegal. Add the city name where chicken meat was sold

The changes suggested by the reviewer have been made line 341 of the revised manuscript.

Lines 388-392: add reference i.e.

A reference (ref 53 of the revised version) has been added line 383 as requested by the reviewer.

Line 401: add doi: 10.1136/bmjgh-2021-005659 to Ref 6

The indicated article has been added as ref. 60 and mentioned line 394 of the revised manuscript.

Line 424-427: Not sure family units use antibiotics as growth promotors. Therefore, the link between improper use of antibiotics (line 425), small-scale farming (426) and continuous selection of MDR strains (427) is unclear to me.

In Senegal, broilers available to consumers come mostly from family units that raise chicken for consumption and for limited commercialization, and from small and mid-size chicken farms. There is no legislation banning the use of antimicrobials as growth promoter in farming, but only recommendations to which farmers are often not aware of. Consequently, empirical use of antimicrobials to promote animal growth is a common practice.

Do data on antibiotic consumption in human and animals exist in Senegal? If yes, authors should refer to those data in the discussion section, especially in lines 432-439.

Data on human and animal consumption of antimicrobial are not systematically collected and made available in Senegal. However, Permanent Secretariate of the High National Counsel for Global health Security Agenda, through its AMR Working Group is currently putting in place recommendations what would lead to regulations and legislations in the near future.

We would like to thank the reviewer for his thorough review and for the interesting suggestions and guidance he provided.

Submitted filename: Response to Reviewers.docx

Click here for additional data file.

14 Mar 2022

Genomics of human and chicken Salmonella isolates in Senegal: broilers as a source of antimicrobial resistance and potentially invasive nontyphoidal salmonellosis infections

PONE-D-21-24873R1

Dear Dr. Dieye,

We’re pleased to inform you that your manuscript has been judged scientifically suitable for publication and will be formally accepted for publication once it meets all outstanding technical requirements.

Within one week, you’ll receive an e-mail detailing the required amendments. When these have been addressed, you’ll receive a formal acceptance letter and your manuscript will be scheduled for publication.

An invoice for payment will follow shortly after the formal acceptance. To ensure an efficient process, please log into Editorial Manager at http://www.editorialmanager.com/pone/, click the 'Update My Information' link at the top of the page, and double check that your user information is up-to-date. If you have any billing related questions, please contact our Author Billing department directly at authorbilling@plos.org.

If your institution or institutions have a press office, please notify them about your upcoming paper to help maximize its impact. If they’ll be preparing press materials, please inform our press team as soon as possible -- no later than 48 hours after receiving the formal acceptance. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information, please contact onepress@plos.org.

Kind regards,

Patrick Butaye, DVM, PhD

Academic Editor

PLOS ONE

Additional Editor Comments (optional):

Reviewers' comments:

16 Mar 2022

PONE-D-21-24873R1

Genomics of human and chicken Salmonella isolates in Senegal: broilers as a source of antimicrobial resistance and potentially invasive nontyphoidal salmonellosis infections

Dear Dr. Dieye:

I'm pleased to inform you that your manuscript has been deemed suitable for publication in PLOS ONE. Congratulations! Your manuscript is now with our production department.

If your institution or institutions have a press office, please let them know about your upcoming paper now to help maximize its impact. If they'll be preparing press materials, please inform our press team within the next 48 hours. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information please contact onepress@plos.org.

If we can help with anything else, please email us at plosone@plos.org.

Thank you for submitting your work to PLOS ONE and supporting open access.

Kind regards,

PLOS ONE Editorial Office Staff

on behalf of

Professor Patrick Butaye

Academic Editor

PLOS ONE