gyrA ser83 mutation among fluoroquinolone-resistant Salmonella enterica serovars from enteric fever patients in tertiary care hospital, Kathmandu

The management of enteric fever through antibiotics is difficult these days due to the emerging resistance of Salmonella to various antimicrobial agents. The development of antimicrobial resistance is associated with multiple factors including mutations in the specific genes. To know the current status of mutation-mediated fluoroquinolone-resistance among Salmonella enterica serovars; Typhi, Paratyphi A, B and C, this study was focused on detecting gyrA ser83 mutation by restriction digestion analysis of gyrA gene using HinfI endonuclease. A total of 948 blood samples were processed for isolation of Salmonella spp. and 3.4% of them were found to be positive for Salmonella growth. Out of the 32 Salmonella isolates, 2.2% were S. Typhi and 1.2% were S. Paratyphi A. More interestingly, we observed less than 5% of isolates were resistant to first-line drugs including chloramphenicol, cotrimoxazole and ampicillin. More than 80% of isolates were resistant to fluoroquinolones accounting for 84.4% to levofloxacin followed by 87.5% to ofloxacin and 100% to ciprofloxacin by disc diffusion methods. However, the minimum inhibitory concentration method using agar dilution showed only 50% of isolates were resistant to ciprofloxacin. A total of 3.1% of isolates were multidrug-resistant. Similarly, 90.6% of the Salmonella isolates showed gyrA ser83 mutation with resistance to nalidixic acid. The increased resistance to fluoroquinolones and nalidixic acid in Salmonella isolates in our study suggests the use of alternative drugs as empirical treatment. Rather, the treatment should focus on prescribing first-line antibiotics since we observed less than 5% of Salmonella isolates were resistant to these drugs.


Background
Enteric fever is an acute, life-threatening febrile infection caused by Salmonella enterica serovars; Typhi, Paratyphi A, B and C. Symptoms may vary from mild to severe and usually begin 6 to 30 days after exposure [1]. These isolates are highly adapted infections to the human population, having no animal and environmental reservoirs. Enteric fever was once considered a major cause of morbidity and mortality throughout the world accounting for approximately 15% case fatality rate. World Health Organization (WHO) estimates the annual global incidence of enteric fever is between 11 and 21 million cases and approximately 128,000 to 161,000 deaths [2]. In addition, the infection is more prevalent in developing countries with poor sanitation including Nepal. As per the annual report published by the Department of Health Service (DoHS), Ministry of Health and Population, Nepal, enteric fever was still one of the top ten causes for inpatient morbidity (8.48%) [3]. Nepal remains endemic to enteric fever for many years. S. Typhi and S. Paratyphi A are the two most predominant bacterial pathogens in the blood cultures of enteric febrile patients in Nepal.
Although enteric fever is endemic in many developing countries including Nepal, effective antimicrobial therapy has significantly reduced morbidity and mortality among the infected cases. S. Typhi and S. Paratyphi A, B and C are susceptible to a variety of antibiotics in vitro testing. However, in vivo responses are not always similar to in vitro susceptibility. It is very difficult to predict accurately because of their predominantly intracellular location within phagocytic cells. These isolates were usually sensitive to the fluoroquinolone antibiotics which were rapidly bactericidal than the third-generation cephalosporins. Unfortunately, the resistance to fluoroquinolones has now developed in S. Typhi and S. Paratyphi isolates from central Asia, southern India and Vietnam [4][5][6][7]. The emerging resistance to fluoroquinolones adds another burden to manage enteric febrile cases. This is because of extensive use of fluoroquinolones for the treatment of infections caused by multidrugresistant Salmonella spp. [8]. S. enterica has developed resistance to fluoroquinolones in many different ways including inactivation of drug, reduced membrane permeability, alteration of the target site and active efflux [9]. Likewise, single point mutations in the "quinolone resistance determining region" of the gyrA gene are usually associated with a mechanism of fluoroquinolone resistance [9]. The mutation may occur in gyrA when there is a change in the amino acid sequence either by addition or deletion leading to the development of resistance. The most commonly identified mutation has been a serine to phenylalanine substitution at position 83. Less common mutations have been reported as aspartate to tyrosine or glycine at position 87 [10,11]. Moreover, the reduced susceptibility to the fluoroquinolones has also been reported in Salmonella isolates without any mutations in gryA gene suggesting an alternative way of resistance. The mutation in gryA gene is not only responsible for the reduced susceptibility to the fluoroquinolones but also increases the minimum inhibitory concentration (MIC) breakpoint values to fluoroquinolone antibiotics [12]. Hence, it is necessary to evaluate MIC break points to those antibiotics regularly. The susceptibility testing of Salmonella isolates to a variety of antibiotics are equally important to know the current antimicrobialresistant status and prescribing potential drugs for disease management. This study was therefore conducted to determine the antibiogram patterns of S. Typhi and S. Paratyphi isolated in blood cultures of enteric febrile patients. Our study also estimated the fluoroquinolones resistance due to mutation of gryA gene at ser83 position using restriction digestion. Finally, we have also recommended some potential antibiotics to manage fluoroquinolones resistant Salmonella infections.

Study design, duration, and site
A hospital-based cross-sectional study was conducted from February to July 2019 among the clinically suspected patients with enteric fever visiting Sukraraj Tropical and Infectious Disease Hospital, Kathmandu, Nepal.

Inclusion and exclusion criteria
The patients suspected of enteric fever from all sexes of greater than five years who gave written consent to participate in the study were included for data and sample collection. The collected samples with visible signs of contamination or insufficient volume were excluded.

Consent from the patients
Written consent was taken from each participant. The patients' information including demographic characteristics, travel history and clinical history were collected using a structured questionnaire.

Sample collection
About 8-10 ml venous blood sample was collected from each adult patient and 3-5 ml from each child patient [13,14]. A total of 948 blood samples from the patients were processed during the study period. 5 ml of blood from each adult patient and 3 ml from each child patient were inoculated in a BACTEC culture bottle and incubated at 37 °C for up to five days. The remaining blood samples were used for routine laboratory diagnosis including hematology, serology and biochemical investigation. The culture bottles showing bacterial growth were sub-cultured on Mac-Conkey agar (MA), Xylose lysine deoxycholate agar (XLD) and Blood agar (BA). The isolated colonies were further identified as per standard microbiological techniques including colony morphology, staining reaction, biochemical characteristics and serotyping method [15,16].

Minimum inhibitory concentration (MIC) of ciprofloxacin
Salmonella enterica isolates were further subjected to the minimum inhibitory concentration of ciprofloxacin by the agar dilution method using the antibiotic concentration of 1280 μg/ml. Salmonella ATCC 35664 was used for quality control. Muller Hinton Agar (MHA) plates were prepared with different concentrations of ciprofloxacin antibiotics (0.002 to 256 μg/ml) as per the recommendation in Andrews and CLSI guidelines [17,19]. Standardized test inoculum was inoculated on antibiotic-MHA plates using a micropipette capable of delivering 2 μl volume with about 10 4 cells per spot. Each test organism was inoculated in triplicate. QC strain was inoculated for the validity of the test for every batch of agar dilution. The plates incubated aerobically at 37 °C for 18 to 20 h, were observed for visible colonies to confirm the growth of the bacterium. Inoculation of the organism in an antibioticfree plate was undertaken as a control [19].

DNA extraction and PCR amplification of gyrA gene
Genomic DNA was extracted from Salmonella isolates by phenol:chloroform method and the presence of DNA was confirmed on 0.8% agarose gel electrophoresis at 120 V for one hour with 0.1 μg/ml ethidium bromide (EtBr) concentration [20]. The PCR amplification of gyrA gene was carried out using a set of specific primers; gyrA-F (5′-CGA GAG AAA TTA CAC CGG TCA-3′) and gyrA-R (5′-AGC CCT TCA ATG CTG ATG TC-3′) from Macrogen Inc., Korea [21]. A total of 25 μl PCR reaction mixture volume containing 21 μl of 1X PCR master mix (Qiagen), 3 μl of template DNA, 0.5 μl of gyrA forward primer and 0.5 μl of gyrA reverse primer was used for PCR amplification. DNA from Salmonella ATCC 35664 as positive control and no template as negative control were run simultaneously in each PCR cycle. The gyrA gene amplification was carried out at an initial denaturation of 95 °C for 15 min followed by 30 cycles of denaturation at 94 °C for 60 s, annealing at 57 °C for 90 s, extension at 72 °C for 60 s and a final extension at 72 °C for 7 min [21]. The amplified gyrA gene was confirmed by its molecular size of 610 bp on 1.5% agarose gel electrophoresis with 0.1 μg/ml EtBr concentration.

Digestion of amplified PCR product of gyrA by HinfI
The amplified PCR product of gyrA gene was further digested by using restriction endonuclease; Fast HinfI from Thermofisher Scientific Inc. A total restriction digestion reaction of 30 μl containing 17 μl of nucleasefree water, 2 μl of fast digest green buffer, 10 μl of amplified gyrA product and 1 μl of fast digest HinfI enzyme was prepared in a tube and mixed properly. The tube was incubated at 37 °C in a water bath for five minutes and then placed into another water bath at 65 °C for 20 min for enzyme inactivation following the manufacture's guidelines. The digested product was run through 2% agarose gel electrophoresis at 120 V for 60 min and visualized under a UV transilluminator. The digested products into two fragments of molecular sizes; 343 bp and 149 bp were considered as the mutation in gyrA gene while the digested products from non-mutated isolates were revealed into three fragments of 244 bp, 149 bp and 118 bp sizes. The mutated and non-mutated isolates were further compared with nalidixic acid-resistant (NAS) and nalidixic acid-sensitive (NAS) groups.

Data analysis
The preliminary data was managed by MS Excel and later transcribed to Statistical Package for Social Science (SPSS) software (version25) for statistical analysis. A scatterplot diagram to determine the breakpoint of ciprofloxacin MIC was generated using WHONET 5.0.

Total blood cultures and growth results
Out of 948 blood cultures, 76 (8.0%) samples were culture positive, in which collectively 32 (3.4%) samples were found to be positive for Salmonella spp. including Salmonella Typhi (n = 21) and Salmonella Paratyphi A (n = 11). For the remaining samples, 17 (1.8%) samples were positive for Staphylococcus aureus, 20 (2.1%) for Coagulase Negative Staphylococci (CONS) and 7 (0.7%) were positive for Escherichia coli. Since our study was focused on enteric febrile cases by Salmonella isolates, the other isolates were not further processed.

Distribution of enteric fever cases based on the age of the patients
The highest rate of Salmonella was isolated from blood specimens of the age groups 5-20 years (1.6%; 15/945) and 21-35 years (1.6%; 15/945) followed by 36

Antimicrobial susceptibility pattern of Salmonella isolates
None of the isolates of Salmonella were resistant to cefixime followed by 3.1% of the isolates were resistant to the antibiotics; ampicillin, chloramphenicol, cotrimoxazole and ceftriaxone. In the case of fluoroquinolones, 100% of the isolates were resistant to ciprofloxacin while 84.4% of Salmonella isolates were resistant to levofloxacin followed by 90.6% resistant to nalidixic acid and 87.5% non-susceptible to ofloxacin ( Table 3). Out of 32 Salmonella isolates, only one S. Typhi was found to be multidrug-resistant.

Minimum inhibitory concentration value of ciprofloxacin
The scatter plot correlating the MIC values and disk diffusion method of ciprofloxacin suggests the reduced susceptibility of Salmonella isolates. Out of nine resistant isolates by disk diffusion (zone of inhibition < 21 mm), eight were resistant to ciprofloxacin by MIC (MIC value ≥1 μg/ml) method and one showed reduced susceptibility (0.12-0.5 μg/ml). Similarly, out of 23 isolates that were intermediately resistant to ciprofloxacin by disk diffusion method, only eight isolates were found to be resistant by MIC, 12 isolates with reduced susceptibility and three were sensitive. A large number of isolates were found in the area of ciprofloxacin intermediate with slightly increased MIC values (Fig. 1).

PCR amplification of gyrA and HinfI restriction digestion analysis of gyrA
GyrA gene of molecular weight of 610 bp was amplified in all 32 Salmonella isolates ( Fig. 2A). Out of 32 isolates, gyrA ser83 mutation was observed in 90.6% [n=22] isolates showing two fragments of size 343 bp and 149 bp by HinfI digestion. 28.1% of mutated isolates were resistant and 62.5% showed reduced susceptibility to ciprofloxacin. All those isolates were resistant to nalidixic acid. In contrast, gyrA ser83 mutation was not detected in 9.4% [n=3] of Salmonella isolates. All three were sensitive to nalidixic acid antibiotic (Table 4, Fig. 2B).

Discussion
Our study revealed 3.4% culture-positive blood samples to Salmonella enterica. This study agreed with a descriptive study done from 2011 to 2013 which reported 3.05% Table 1 Distribution of blood cultures and number of S. enterica isolates based on the age and sex group of patients     [24]. The prevalence of enteric fever in Nepal seems to have fluctuated between 3 to 10% for the last decade without a significant increase or decrease in cases. Among age-wise distribution of enteric fever, the greatest number of cases were observed from the age group of 5-20 years and 21-35 years. Pokharel et al. had reported the highest number of enteric fever from the age group of 21-35 years in a hospital at Kathmandu [25].. Likewise, Bhetwal et al. also reported the highest number of Salmonella infections (S. Typhi and S. Paratyphi A) among the children of age 5-15 years in a community-based teaching hospital in Nepal in 2017 [13]. The higher rate of culture positivity among the age group of 5-35 years might be due to poor food hygiene and their dependency on out-sourcing foods [26]. Similarly, personal hygiene might be the reason for the higher infections among children.

Age group (years) Number of cases (%) Number of S. enterica isolates (%) p-value
We observed none of Salmonella isolates were resistant to cefixime and less than 5% resistance to the four antibiotics namely ampicillin, ceftriaxone, chloramphenicol and cotrimoxazole. The results also agree with the antibiogram report presented by Shrestha and Basnet in 2019 at Patan Hospital, Lalitpur, Nepal [27]. Although cefixime is a preferred drug of choice in developing countries due to the availability of an oral form for uncomplicated enteric fever [28], a study by Pandit et al. reported resistance to cefixime in vivo even if the isolates were susceptible in vitro study [29]. Although third-generation cephalosporins were found to be effective drugs for Salmonella isolates in our study, the increasing rate of extended-spectrum β-lactamase (ESBL) producing strains among other members of the Enterobacteriaceae family have added a huge challenge in empirical therapy [22,[30][31][32]. Those isolates could be the sources of ESBL genes to Salmonella isolates via horizontal transfer. Compared to a study by Maharjan et al. at the same hospital [33], we observed reduced susceptibility to fluoroquinolone antibiotics. The results were also supported by the data of Shrestha and Basnet [27]. The reason might be due to the extensive use of these drugs as therapeutic management of enteric fever. The rate of multidrugresistant Salmonella was consistently low in the last two decades [22,24,34]. We also found only one MDR isolate   [33]. A low rate of MDR was supported by the findings of Britto et al. 2018 [35]. The abstain use of first-line antibiotics as empirical treatment for a longer period might have caused a loss of IncH1 plasmids leading to emerging back to sensitivity towards these drugs [36][37][38]. In addition, immunization could theoretically reduce the number of circulating MDR [35]. In our study, the scatterplot correlating the MIC and disk diffusion values of ciprofloxacin illustrates the reduced ciprofloxacin susceptibility among Salmonella isolates as compared to the studies at the same hospital [33,39]. Compared to fluoroquinolone sensitive isolates, MIC values were commonly two or more dilutions higher (ciprofloxacin MIC 0.12 to 2 μg/ml for resistant strains compared with 0.5 to 16 μg/ml for intermediate resistant strains and 0.06 μg/ml for sensitive strains). When compared with nalidixic acid disc, the resistant strains showed ciprofloxacin MIC value of greater than 0.12 μg/ ml. Nalidixic acid resistance hence serves as a surrogate marker for gyrA mutation associated with diminished fluoroquinolone susceptibility [40]. Out of 90.6% of NAR isolates, 40.6% of the Salmonella isolates had decreased susceptibility to ciprofloxacin with MIC value of 0.12 to 2 μg/ml and 50% isolates were resistant to ciprofloxacin with MIC value of up to 16 μg/ml. A similar study undertaken in India reported 47.5% resistance and 36.2% decreased susceptibility to ciprofloxacin among 97.5% of NAR Salmonella isolates, [41]. Nalidixic acid resistance showed a 100% predictive value for ciprofloxacin resistance as reported by Agrawal et al. [42]. Besides the gyrA gene mutation that occurred usually in the bacterial chromosome, the plasmid-mediated quinolone resistance mechanisms cannot be ignored [43].
Ser-83 mutation is the most commonly occurring point mutations among fluoroquinolone resistance strains. Over 90% of gyrA ser83 mutation was observed in Salmonella strains in our study which were similar to the studies by Khadka [39,44,45]. Complete resistance or reduced susceptibility to fluoroquinolones is due to the mutation at ser-83 or asp-87 position of gyrA [35]. The mutations in the quinolone resistance determining regions of chromosomal genes such as gyrA, gyrB, parC and parE and plasmid-mediated qnr, qepA and aacs genes cause fluoroquinolones not to act and induce resistance in Salmonella spp. [43]. In our study, we observed that the strains with reduced susceptibility or complete resistance to fluoroquinolones retained ser-83 mutation. However, the isolates resistant to nalidixic acid and ciprofloxacin may have two or more mutations in the gyrA, gyrB, parC, or parE gene [46] which were not investigated in this study. In the case of patients with reduced susceptibility to fluoroquinolones for enteric fever, the infections can be managed with a higher dose of fluoroquinolones for a longer duration of time [47].
As a major study limitation, we only focused on Salmonella isolates and didn't process for other bacterial pathogens isolated on the blood culture. Although there are many other mechanisms by which Salmonella may develop resistance to fluoroquinolones, due to limited time and budget, we couldn't analyze all those mechanisms. We could only investigate gyrA ser83 mediated mutation among the isolates using restriction endonuclease HinfI. In addition, our study has shown the recent trend of the antimicrobial-resistant pattern of Salmonella isolates and suggested potential drugs for antimicrobial therapy.

Conclusions
Since fluoroquinolones are the drugs of choice for the management of enteric fever, especially caused by S. Typhi and S. Paratyphi A in most developing countries, the decreased susceptibility to fluoroquinolones along with the accumulation of mutation in the ser83 position of gyrA gene may pose a threat to the disease management. In conclusion, our study suggests reintroducing ampicillin, chloramphenicol and cotrimoxazole drugs as empirical treatment for enteric fever as well as the use of third generation cephalosporins.