Antimicrobial Resistance of Shigella flexneri Serotype 1b Isolates in China

Shigella flexneri serotype 1b is among the most prominent serotypes in developing countries, followed by serotype 2a. However, only limited data is available on the global phenotypic and genotypic characteristics of S. flexneri 1b. In the present study, 40 S. flexneri 1b isolates from different regions of China were confirmed by serotyping and biochemical characterization. Antimicrobial susceptibility testing showed that 85% of these isolates were multidrug-resistant strains and antibiotic susceptibility profiles varied between geographical locations. Strains from Yunnan were far more resistant than those from Xinjiang, while only one strain from Shanghai was resistant to ceftazidime and aztreonam. Fifteen cephalosporin resistant isolates were identified in this study. ESBL genes (bla SHV, bla TEM, bla OXA, and bla CTX-M) and ampC genes (bla MOX, bla FOX, bla MIR(ACT-1), bla DHA, bla CIT and bla ACC) were subsequently detected among the 15 isolates. The results showed that these strains were positive only for bla TEM, bla OXA, bla CTX-M, intI1, and intI2. Furthermore, pulsed-field gel electrophoresis (PFGE) analysis showed that the 40 isolates formed different profiles, and the PFGE patterns of Xinjiang isolates were distinct from Yunnan and Shanghai isolates by one obvious, large, missing band. In summary, similarities in resistance patterns were observed in strains with the same PFGE pattern. Overall, the results supported the need for more prudent selection and use of antibiotics in China. We suggest that antibiotic susceptibility testing should be performed at the start of an outbreak, and antibiotic use should be restricted to severe Shigella cases, based on resistance pattern variations observed in different regions. The data obtained in the current study might help to develop a strategy for the treatment of infections caused by S. flexneri 1b in China.


Introduction
Shigella, the causative agent of human shigellosis, is a significant public health burden. Worldwide, 164.7 million Shigella episodes per year are estimated, of which 163.2 million are in developing countries. The majority of the infections and deaths caused by Shigella are in children <5 years old [1]. Moreover, a surveillance study in six Asian countries showed the global disease burden caused by shigellosis might be much higher than these estimates [2]. In China, shigellosis is one of the top four notifiable infectious diseases, with nearly half a million cases annually [3].
Pickering LK showed that antimicrobial susceptibility patterns of Shigella species, were influenced by geographic location and other factors (year, classes of antimicrobial agents, pressure exerted by antimicrobial use, source of isolates) in endemic regions [8]. However, comparisons of associations between geographical diversity and antimicrobial resistance of serotype 1b strains are largely lacking. The objective of the present study was to evaluate the antimicrobial resistance profiles of S. flexneri 1b isolates from different regions of China, and guide the selection of the most effective antimicrobial agents for shigellosis treatment.

Materials and Methods
Bacterial isolates, serotyping, and biochemical characterization During our routine surveillance of bacillary dysentery, fecal samples from individual outpatients with diarrhea or dysentery were collected and screened for Shigella species in sentinel hospitals based on a national pathogen monitoring system. Samples were streaked directly onto Salmonella-Shigella (SS) agar and incubated overnight at 37°C. Resultant colonies were picked and streaked directly onto SS agar and again incubated overnight at 37°C. The colonies were subcultured on Luria-Bertani agar plates and incubated overnight at 37°C, and were then sent to our laboratory for further confirmation. This study was approved by the ethics committee of the Academy of Military Medical Sciences (China). Written informed consent was obtained from the patients involved in this study. The isolates were confirmed using API 20E test strips (bioMerieux Vitek, Marcy-l'Etoile, France) following the manufacturer's recommendations. Serotype identification was performed using two serotyping kits: a kit including antisera specific for all type and group-factor antigens (Denka Seiken, Tokyo, Japan), and a monoclonal antibody reagent kit (Reagensia AB, Stockholm, Sweden), specific for all S. flexneri type and group-factor antigens. Serological reactions were performed using the slide agglutination test, as described previously [9].

Pulsed-field gel electrophoresis (PFGE)
The isolates were analyzed by PFGE following digestion with NotI to generate DNA fingerprinting profiles according to the procedures developed by the CDC PulseNet program [10]. Salmonella enterica serotype Braenderup H9812 was used as the molecular size marker. PFGE patterns were interpreted using BioNumerics (Version 6.0) software (Applied Maths, Sint-Martens-Latem, Belgium). A tree indicating relative genetic similarity was constructed based on the unweighted pair group method of averages (UPGMA) and a position tolerance of 1.2%.

PCR amplification of the antibiotic-resistance determinants and integrons
PCR assays were carried out as described previously [11][12][13][14][15][16] to detect ESBL genes (bla SHV , bla-TEM , bla OXA , and bla CTX-M ) and ampC genes (bla MOX , bla FOX , bla MIR(ACT-1) , bla DHA , bla CIT and bla ACC ) which are responsible for resistance to cephalosporins. We also amplified the variable regions of class 1 and 2 integrons using primers listed in Table 1. The reverse primers aadA1, aadA2, aadA5, and cmlA1 were used with forward primer hep58 to amplify the gene cassettes of class 1 integron-positive strains. Forward primer hep74 and reverse primer hep51 were used to amplify the gene cassettes of class 2 integron-positive strains. Resultant PCR products were fully sequenced and analyzed by comparison with sequences in GenBank.

Bacterial isolation and biochemical characterization
We isolated and identified 40 S. flexneri 1b isolates from patients with either diarrhea or dysentery over the period from 2005 to 2013. The strain information is detailed in Table 2. Among them, 28 isolates were collected from Xinjiang from 2005-2012, 11 from Yunnan in 2009, and one from Shanghai in 2013. All S. flexneri serotype 1b strains examined possessed typical S. flexneri biochemical characteristics, and analysis of biochemical reactions indicated the presence of four biotypes among these serotype 1b strains (Table 3).

Molecular analysis of antibiotic-resistance determinants and integrons
In this study, fifteen isolates were found to be resistant to cephalosporin antibiotics, and these strains were tested for antibiotic-resistance determinants and integrons, including the bla SHV , bla TEM , bla OXA , bla CTX-M , bla MOX , bla FOX , bla MIR(ACT-1) , bla DHA , bla CIT , bla ACC , intI1, and intI2 gene regions (Table 6). PCR results showed that all 15 tested isolates were negative for

PFGE analysis
PFGE was performed to determine the genetic diversity of the S. flexneri 1b isolates. The 40 S. flexneri 1b isolates generated 18 PFGE patterns (Fig 1). All isolates could be divided into five main groups (A-E), with approximately 82% similarity. Notably, there was a high level of variation in the PFGE profiles of serotype 1b isolates between different regions of China, which contrasted with a previous study showing that 70 S. flexneri 1b strains isolated in Bangladesh  shared the same PFGE profile [7]. In the present study, isolates from Xinjiang belonged to groups A, D, and E, whereas isolates from Yunnan and Shanghai belonged to groups B and C, respectively. Twenty-eight isolates from Xinjiang showed 14 different PFGE patterns; 26 of the 28 isolates formed a single cluster, with the two remaining isolates forming separate branches. Eleven isolates from Yunnan showed three different PFGE patterns, and the single isolate from Shanghai displayed a distinct PFGE pattern. Interestingly, the PFGE patterns of Xinjiang isolates were distinct from Yunnan and Shanghai isolates by one obvious, large, missing band indicated with a red border in Fig 1. The original number, origin, and isolation year are indicated for each strain. The missing band of Xinjiang isolates was indicated with a red border and 15 cephalosporin resistant isolates were marked with triangles.
Among the isolates in different PFGE clusters, the prevalence of antibiotic resistance in clusters A, D, and E (Xinjiang) was much lower than that of isolates in cluster B (Yunnan). All 15 cephalosporin resistant isolates (marked with triangles in Fig 1) belonged to clusters B  (Yunnan), C (Shanghai), and a small branch of cluster A (Xinjiang). All isolates in cluster B were resistant to cephalosporins, such as CFP, CRO, and CFZ, while the resistance rate of isolates in clusters A, D, and E (Xinjiang) to these antibiotics was 10.7%. Higher resistance rates to SXT, PIP, AMP, TIC, TE, C, and TIM were also observed in isolates from Xinjiang than in those from Yunnan. The isolate in cluster C (Shanghai) was the only isolate showing resistance to CAZ and ATM. In summary, similarities in resistance patterns were observed in strains with the same PFGE pattern.

Discussion
Approximately 85% of the strains in our study showed MDR profiles, and were highly resistant to the older-generation antimicrobials, such as AMP, C, TE, and TIC. So the treatment should be based on susceptibility patterns and antimicrobials with current resistance may turn to be effective in the future. This rapid accumulation of resistance in Shigella is probably the result of inappropriate prescription of, and easy access to, antimicrobials among outpatients in China [18]. Besides, the PFGE dendrogram showed that the S. flexneri type 1b isolates from Yunnan are closely related (95% similarity), indicating that they are possibly derived from a common parental strain. In contrast, the isolates from Xinjiang shared lower degrees of similarity, suggesting that they are likely derived from divergent sources. The antibiotic susceptibility profiles of the isolates also varied between the geographical locations; the antimicrobial resistance situation in Xinjiang does not seem to be as serious as that in Yunnan and Shanghai. Among the 40 isolates, only the isolate from Shanghai was resistant to CAZ and ATM, and isolates from Yunnan were more resistant to antibiotics CFP, CRO, CFZ, PIP, AMP, TIC, TE, C, TIM, and SXT than those from Xinjiang and Shanghai. The diverse antibiotic susceptibility may be attributed to the fact that mainland China covers a large territory and has many nationalities, with more than 1.3 billion people living in diverse areas, hence many varying factors such as customs, climates, economies, geographies may affect the antimicrobial susceptibility patterns.
In the present study, all 15 cephalosporin resistant isolates harbored bla CTX-M genes, most of which were bla CTX-M-14 , followed by bla CTX-M-79 , and bla CTX-M-28 . A previous study indicated that plasmids carrying different bla CTX-M genes were introduced into isolates at different times [18]. The bla CTX-M-14 subtype was identified in 14 of these isolates, which indicated that these genes might have been circulating among Shigella isolates for a long time. In addition, bla TEM was detected in all 15 cephalosporin resistant isolates, and bla OXA was amplified in all such isolates, except the isolate from Shanghai. OXA-type β-lactamases are characterized by high hydrolytic activity against cloxacillin and oxacillin, which confer resistance to cephalothin and ampicillin [19]. Notably, all bla OXA isolates in this study carried bla OXA-1 , suggesting a host specificity for this subtype in S. flexneri, which is consistent with a previous study by Siu et al. [17]. bla OXA-30 and aadA1 were also detected in the gene cassettes of class 1 integrons in five of the 15 cephalosporin resistant isolates in our study. These two gene cassettes appear to be coordinately excised or integrated, as described by a previous study [3]. Pan et al. reported that class 2 integrons were present in 87.9% of S. flexneri isolates [14]. In our study, all of the 15 strains examined harbored class 2 integrons, which followed the dfrA1+sat1+aadA1 gene cassette, conferring resistance to trimethoprim and streptomycin [20]. All of these various resistance genes facilitate the dissemination of resistance determinants and the survival of bacteria under the selective pressure of various antibiotics.
Fluoroquinolones and third-generation cephalosporins were two popular empirical options to treat severe gastrointestinal infections caused by pathogenic bacteria [21]. Ciprofloxacin is currently the first-line antibiotic recommended by the World Health Organization for treating shigellosis [22]. Fortunately, limited resistance to fluoroquinolones was observed in our study. The control of resistance to quinolones could be attributed to a relevant study on quinolone-induced cartilage toxicity [23]. Third-generation cephalosporins are considered alternative drugs for shigellosis treatment [22], but in our study, the isolates from Yunnan and Shanghai were all resistant to CFP, CRO, and CFZ, and the rate of resistance of isolates from Xinjiang to these cephalosporins was 10.7%. According to the literature, strains may acquire resistance to multiple antibiotics if exposed to a resistant gastrointestinal pathogen following antibiotic therapy [24]. Therefore, these ESBL-producing isolates, especially those from Yunnan and Shanghai, would put patients at high risk if third-generation cephalosporins were still used for empirical therapy. Moreover, the isolate from Shanghai also exhibited resistance to ATM and CAZ, which clearly raises the perennial issue of strict control of antimicrobial prescription in China.
In the present study, diverse antimicrobial resistance patterns were observed in different endemic areas, which support the need for prudent selection and use of antibiotics. Therefore, we suggest that antibiotic susceptibility testing should be performed at the start of an outbreak, and antibiotic use should be restricted to severe Shigella cases, based on variations in resistance patterns in different regions. Furthermore, resistance patterns should be recorded timely and locally, and empiric antibiotic therapy for Shigella cases should be changed accordingly. The data that we obtained here might provide a strategy for the treatment of infections caused by S. flexneri serotype 1b strains in China.