The molecular basis of extensively drug-resistant Salmonella Typhi isolates from pediatric septicemia patients

Sepsis is a syndromic response to infections and is becoming an emerging threat to the public health sector, particularly in developing countries. Salmonella Typhi (S. Typhi), the cause of typhoid fever, is one primary cause of pediatric sepsis in typhoid endemic areas. Extensively drug-resistant (XDR) S. Typhi is more common among pediatric patients, which is responsible for over 90% of the reported XDR typhoid cases, but the majority of antibiotic resistance studies available have been carried out using S. Typhi isolates from adult patients. Here, we characterized antibiotic-resistance profiles of XDR S. Typhi isolates from a medium size cohort of pediatric typhoid patients (n = 45, 68.89% male and 31.11% female) and determined antibiotic-resistance-related gene signatures associated with common treatment options to typhoid fever patients of 18 XDR S. Typhi representing all 45 isolates. Their ages were 1–13 years old: toddlers aging 1–2 years old (n = 9, 20%), pre-schoolers aging 3–5 years old (n = 17, 37.78%), school-age children aging 6–12 years old (n = 17, 37.78%), and adolescents aging 13–18 years old (n = 2, 4.44%). Through analyzing blaTEM1, dhfR7, sul1, and catA1genes for multidrug-resistance, qnrS, gyrA, gyrB, parC, and parE for fluoroquinolone-resistance, blaCTX-M-15 for XDR, and macAB and acrAB efflux pump system-associated genes, we showed the phenotype of the XDR S. Typhi isolates matches with their genotypes featured by the acquisitions of the genes blaTEM1, dhfR7, sul1, catA1, qnrS, and blaCTX-M-15 and a point mutation on gyrA. This study informs the molecular basis of antibiotic-resistance among recent S. Typhi isolates from pediatric septicemia patients, therefore providing insights into the development of molecular detection methods and treatment strategies for XDR S. Typhi.

resistance Introduction Sepsis is a syndromic response to infections and is becoming an emerging threat to the public entire cell envelope (24). In typhoidal Salmonella, point mutations at amino acid position 717 74 (R717Q or R717L) on AcrB, the antibiotic-binding subunit of the RND-type AcrAB-TolC efflux pump, have been correlated with resistance to azithromycin in S. Typhi and S. Paratyphi A, Antibiotic-resistant S. Typhi infection is more common among children; more than 90% Results XDR S. Typhi strains isolated from children at various developmental stages patients with 1-13 years of age (68.89% male and 31.11% female) who have visited the Fatima genes. To carry out a series of molecular characterization via PCR and/or PCR amplicon The acquisition of an IncY region harboring qnrS and one or more point mutations on the 142 genes gyrA, gyrB, parC, and/or parE, referred to as the quinolone resistance determining region 143 (QRDR), have been correlated to fluoroquinolone-resistance among typhoidal Salmonella 144 strains. A PCR primer set specific for qnrS was designed and used to investigate whether XDR S.
Typhi isolates encode this fluoroquinolone-resistance-related gene. We found that, unlike antibiotic-susceptible S. Typhi ISP2825, all XDR S. Typhi isolates tested encode qnrS ( Figure  gyrA, gyrB, parC, and/or parE, which exhibits variations depending on geographical locations (10,(12)(13)(14)(15)(16)(17)(18)(19)(20) (Table 4). Specific primer sets for the known mutations on these four genes were designed for PCR and PCR amplicon sequencing (Table 3). Consistent with their essential roles in bacterial cell replication, both antibiotic-susceptible S. Typhi ISP2825 and all XDR S. Typhi isolates resulted in PCR products with expected size for the four topoisomerase genes ( Figure  3B-E). To determine whether XDR S. Typhi isolates resistant to fluoroquinolones have point 154 mutations in these topoisomerase genes, we carried out Sanger sequencing of PCR amplicons.
Consistent with the antibiotic-susceptible phenotype, S. Typhi ISP2825 carries wild-type topoisomerases. In contrast, we found that all the XDR S. Typhi isolates carry a mutant form of gyrA encoding for GyrA Ser83Phe ( Figure 3F-G). GyrA Ser83Phe has been found most commonly 158 among XDR S. Typhi identified from other endemic regions. We also found that these XDR S.

161
In addition to GyrA Ser83Phe , resistance to second-line antibiotics among XDR S. Typhi in Pakistan has been associated with an IncY region carrying the quinolone resistance gene qnrS 165 ( Figure 3A) and extended-spectrum -lactamase resistance gene blaCTX- M-15 (9, 38).

166
Consistent with resistance to second-line antibiotics, fluoroquinolones and cephalosporins, 167 among these XDR S. Typhi isolates from pediatric septicemia patients, we found the acquisition 168 of blaCTX-M-15 among all XDR S. Typhi tested (Figure 4). In contrast, antibiotic-susceptible S.
Typhi ISP2825 did not result in PCR amplicon for blaCTX-M-15, indicating the specificity of the primers used.
antibiotic-resistance/susceptibility profiles across antibiotics tested in Table 1. These results led 175 us to investigate drug efflux pumps in XDR S. Typhi isolates since drug efflux pump systems are associated with resistance/susceptibility to a wide range of antibiotics. The recent worrisome trend among some XDR S. Typhi includes a correlation between efflux pump mutations and 178 azithromycin resistance. For instance, a point mutation(s) on the antibiotic-binding subunit AcrB of the tripartite AcrAB-TolC efflux pumps (e.g., R717Q or R717L) has recently been correlated to azithromycin-resistance among some S. Typhi and S. Paratyphi A clinical isolates (19,25,26).

181
In addition, Salmonella encodes another tripartite efflux pump, ABC-type MacAB-TolC, and three other small efflux pumps, major facilitator superfamily (MFS), multidrug and toxin extrusion (MATE), and small multidrug resistance (SMR), spanning in the inner membrane of 184 the bacteria and therefore need to cooperate with another tripartite efflux pump such as RND-185 type AcrAB-TolC in exporting antibiotics (24). In Neisseria gonorrhoeae that, like S. Typhi, is also a human-adapted gram-negative bacterial pathogen, point mutations on the promoter of 187 macA of the tripartite MacAB-TolC efflux pump have been correlated to azithromycin-resistance 188 (39).

190
To assess whether point mutations on tripartite efflux pumps have occurred among XDR sequences via PCR and PCR amplicon sequencing using specific primer sets summarized in Table 3. We found that all XDR S. Typhi isolates tested carry wild-type -10 promoter sequence 194 in the macA promoter and wild-type Arg at position 717 on the AcrB protein ( Figure 5A-C).

Discussion
XDR S. Typhi is more common among pediatric patients but the majority of antibiotic 216 resistance studies available have been carried out using S. Typhi isolates from adult patients.
to determine antibiotic-resistance-related gene signatures associated with their drug-resistant profiles. This study provides a valuable overview of the recent (2019-2020) populations in the setting of Lahore, Pakistan, among a septic pediatric cohort and provides insights into the 221 development of simple, cost-effective molecular detection methods with point-of-care testing 222 potential.

224
Biochemical method-mediated identification assisted by the streamlined automated  Besides qnrS, mutations on gyrA, gyrB, parC, and parE also contribute to 237 fluoroquinolone-resistance, which exhibits more diverse patterns depending on geographical locations. Our XDR S. Typhi isolates from pediatric patients in Northern Pakistan carry GyrA S83F Table 4). This result indicates that XDR S. Typhi circulating in this geographical location is   (Table 1). In contrast, 5% and 48% of recent S. Typhi

265
Although further investigations are required to understand the consequence of having the AcrR 266 frameshifted variant in antibiotic-resistance, it is intriguing to hypothesize that the AcrR variant 267 is less effective in repressing the expression of acrAB, therefore contributing to antibiotic-268 resistance such as azithromycin. We also found that those S. Typhi strains carrying the AcrR 269 variant carry wild-type RobA and MarA, activators for the acrAB gene expression, and wild-type 270 AcrAB, collectively supporting the hypothesis that possession of the AcrR frameshifted variant 271 is an adaptation/evolution outcome, rather than a stochastic event outcome.

273
The emergence and spread of S. Typhi resistant to macrolides and carbapenems are a 274 serious global health concern, deserving close surveillance for local and global spread. We 275 envision that some of the methods detecting key molecular determinants for S. Typhi antibiotic 276 resistance used in the current study could be developed as a surveillance strategy and point-of-277 care testing strategy. The detection and analysis methods for resistance to front-line and second-

278
line antibiotics described in the study are straightforward. Besides efflux pump related molecular 279 determinants described in the study, the future surveillance strategy could include additional 280 molecular traits predicted to be associated with resistance to macrolides and carbapenems in S.

281
Typhi. For instance, in Enterobacteriaceae, erm genes encoding for methylases to modify target Memorial Hospital Lahore. In addition, informed consent was obtained from a legal guardian of 299 each study participant. Informed consent was read to the person in the language they understood 300 and signed appropriately. They were willing to provide a sample and utilize the isolates for 301 research. They were assured that the samples would be used solely for research purposes and that 302 personal information would be kept confidential. Before samples were transferred to researchers, 303 all XDR S. Typhi samples were de-identified, number-based identification codes were assigned 304 to samples (Tables 1 and 2). The data were analyzed anonymously throughout the study.

S. Typhi isolation from patient specimens and MIC determination
The BACT/ALERT® 3D Microbial Detection System with PF/PF plus culture bottles 308 (bioMérieux, France), an automated bacterial culture and antibiotic-resistance test system 309 capable of incubating, agitating, and continuously monitoring aerobic and anaerobic media 310 inoculated with patient specimens was used in this study. The samples were collected between child based on their age and bodyweight were taken and placed in BacT/ALERT PF/PF plus   following the vendor's recommendation. The PCR primer sequences and reaction conditions was used for pltB, blaTEM1, dhfR7, sul1, catA1, parC, parE, blaCTXM15, macA, acrB, and for 30 sec, annealing (see Table 3), and extension at 72℃ for 1 kb/min (see Table 3 for amplicon 335 size), and final extension at 72℃ for 7 min using a C1000 Touch Thermal Cycle (BIO-RAD).

Sanger sequencing of PCR amplicons
When indicated, PCR amplicons were extracted from agarose gels for sequencing analysis by using the QIAEX Ⅱ gel extraction system (QIAGEN, cat # 20051), followed by standard Sanger sequencing (The Cornell Institute of Biotechnology or Eton Bioscience Inc). The primer 343 sequences used for Sanger sequencing are summarized in Table 3.

345
Whole-genome sequencing analysis for efflux pump-related genes

346
The latest NCBI RefSeq dataset of the fully assembled complete genome of S. Typhi (107 in 347 total; Table 5) was collected on Feb 26, 2021. The 107 complete whole-genome sequences were utilized to analyze the sequence variations for acrR (NP_459472.1) using 'General Feature Formats (gff)' files with a bash script (grep acrR *.gff | grep pseudo=true). Two whole genome  The authors declare no competing interests.        Table 3 for details.
reactions for qnrS (A), gyrA (B), gyrB (C), parC (D), and parE (E). See Table 3 Table 3 for details. C, Summary of PCR 31 amplicon sequencing analysis for macA and acrB. WT, wild type for the known mutations.