Susceptibilities of Yersinia pestis to Twelve Antimicrobial Agents in China

Objective: Streptomycin is the preferred choice for therapy of plague in China and other countries. However, Yersinia pestis exhibiting plasmid-mediated antimicrobial agent–resistant traits had been reported in Madagascar. In this study, we evaluated the susceptibility of traditional or newer antimicrobial agents used for treatment and/or prophylaxis of plague. Methods: Following Clinical and Laboratory Standards Institute (CLSI) recommendations, the susceptibility of 12 antimicrobial agents was evaluated by the agar microdilution method in 1,012 strains of Y. pestis isolated from 1943 to 2017 in 12 natural plague foci in China. Results: One clinical Y. pestis isolate (S19960127) was found to be highly resistant to streptomycin, while the strain was still sensitive to other 11 antibiotics, that is, ciprofloxacin, ofloxacin, kanamycin, chloramphenicol, ampicillin, ceftriaxone, cefuroxime, trimethoprim-sulfamethoxazole, tetracycline, spectinomycin and moxifloxacin. The remaining 1,011 Y. pestis strains in this study demonstrated susceptibility to the above-mentioned 12 antimicrobial agents. Conclusion: Antimicrobial sensitivity surveillance of Y. pestis isolates, including dynamic monitoring of streptomycin resistance during various clinical plague treatments, should be carried out routinely.

are classified as two Class A notifiable infectious diseases in Chinese Information System for infectious Disease Control and Prevention, and at least 12 plague foci covering more than 1.4 million square kilometers still exist [10]. Different biovars of Y. pestis strains inhabit the natural plague foci across China. In this study, we investigated the susceptibility of Y. pestis strains in China to 12 antimicrobial agents.

Strains in this study
A total of 1012 Y. pestis strains isolated from 1943 to 2017 in 12 natural plague foci in China were included in this study (Table 1), among which 536 strains had been used in previous research [11]. The sources of these strains were as follows: 570 from rodent animals (marmots, rats, mice, chipmunks, etc.); 268 from humans; 157 from fleas; 14 from artiodactyla (Tibetan sheep and goats); and 3 from other animals. These selected strains represented different biovars and genotypes in China. All strains were collected in the National Y. pestis Preservation Center in QIEDC. All experimental activities with high bio-safety risk, such as the culture of Y. pestis and antibiotic susceptibility testing, were performed in the biosafety level-3 laboratory of QIEDC.

Antibiotic resistance evaluation
Susceptibility testing for Y. pestis and corresponding CLSI quality control reference methods were performed [13]. Minimal inhibitory concentrations (MICs) were determined with the agar dilution method, according to National Committee for Clinical Laboratory Standards guidelines [14] and previous studies [15,16]. The MICs of antibiotics for Y. pestis strains were determined on 96-well plates containing cation-adjusted Mueller-Hinton agar (CAMHA) with multipoint inoculators with an inoculum of 10 4 CFU per spot. The cultures were incubated for 48 hours at 37° [14]. Quality control strains (Pseudomonas aeruginosa ATCC27853 and Escherichia coli ATCC 25922) were tested with each batch of Y. pestis isolates to validate the accuracy of the procedure. Corresponding procedures and interpretation followed the CLSI guidelines for rapidly growing gram-negative bacilli and previous studies [14]. The population ranges of antibiotic susceptibility in various originations of Y. pestis strains were evaluated on the basis of MIC50 and MIC90 values.
A total of 12 antimicrobial agents ( Table 2) were obtained from the Chinese formal pharmacy. The stock solutions (5 mg/ml) were prepared in the appropriate solvents, on the basis of current CLSI recommendations [13]. Antibiotics were serially diluted 2-fold in CAMHA. The concentration range was 64-0.004 μg/ml for tetracycline, ciprofloxacin, chloramphenicol, ofloxacin, kanamycin, ceftriaxone, ampicillin, spectinomycin, cefuroxime, trimethoprim-sulfamethoxazole and moxifloxacin in the plates. For streptomycin, the double dilution upper limit was as high as 4096 mg/ml.
Antibiotic resistance genes of 12 antibiotics were assessed by PCR in the 1012 Y. pestis strains. The oligonucleotide primers targeted for identifying resistance to the 12 antimicrobial agents are listed in Supplemental Table 1. The oligonucleotide primers for the strA and strB genes (plasmid-associated streptomycin resistance genes) were used in PCR to identify the conjugative plasmids pIP1202 and pIP1203 [17,18]. PCR was performed with Taq DNA polymerase (Takara) with the following cycling protocol: Denaturation The strains used in this study. The nomenclature of these plague foci is as previously reported [12].
for 5 min at 95°C; 30 amplification cycles at 95°C for 50 s, Tm °C for 50 s and 72°C for 1 min; and a final extension at 72°C for 5 min. For the genes gyrA, gyrB, parC and rrs, the sequences of PCR products were used to identify the corresponding antibiotic resistance mutations.

Susceptibility of Y. pestis to 12 antimicrobial agents
With the exception of a clinical Y. pestis isolate (S19960127), which exhibited resistance to streptomycin (MIC of 4,096 mg/L, the upper limit of the dilution range) [11], all other Y. pestis strains in this study were susceptible in vitro to 12 antibiotic agents (Table 2). These antibiotic agents included those recommended for plague therapy (streptomycin, ciprofloxacin, chloramphenicol and kanamycin) and prophylaxis (sulfonamides and tetracycline) [4], as well as new antibiotics (ceftriaxone, cefuroxime, spectinomycin and moxifloxacin) and others (ampicillin).

MICs of strains isolated from different sources and plague foci in different years
Generally, we observed no differences in MIC50 or MIC90 for the isolates regardless of their source (humans, rodents and fleas) and natural plague foci (Table 2). In addition, only very limited changes in antibiotic susceptibility for Y. pestis isolated in different years was detected (Table 3). But the susceptibility to these antibiotics varied still within sensitive ranges. This observation also reflected plague natural ecological characteristics.

PCR screening for genes associated with resistance to 12 antibiotics
The PCR screening results for genes associated with resistance to 12 antibiotics were all negative. For the gyrA, gyrB and parC gene targeted for ciprofloxacin, ofloxacin, moxifloxacin, as well as the gene rrs targeted for kanamycin, no corresponding mutations associated with ciprofloxacin, ofloxacin, moxifloxacin, or kanamycin resistance were found in the sequences of PCR products. Y. pestis S19960127 was found to be resistant to streptomycin, which has an MIC of 4096 μg/ml [11]. The other 1011 Y. pestis strains in this study remained susceptible to streptomycin in vitro. The results of PCR with primers for the streptomycin resistance genes strA and strB were negative for Y. pestis S19960127 and the other 1011 Y. pestis strains. A novel mechanism of streptomycin resistance in Y. pestis was subsequently identified as mutation of the rpsL gene [11].

DISCUSSION
Y. pestis isolates were uniformly susceptible to the dominant antibiotics against Gram-negative bacteria described in previous studies [9,19]. In China, with the exception of one clinical Y. pestis strain found to be resistant to streptomycin [11], a large collection of Y. pestis strains in this study were generally found to be susceptible to antimicrobial agents, including antibiotics traditionally recommended for the treatment of Y. pestis infections. These results were similar to those of previous investigation in other counties or areas. For instance, no resistance to eight antimicrobial compounds was identified in 392 Y. pestis isolates from 17 countries [9] in North America, South America, Asia and Africa. In 1996, the susceptibility of 100 South African Y. pestis strains to new antimicrobial agents was determined in vitro; among oral antibiotics, two quinolones (ofloxacin and levofloxacin) showed extremely high antibacterial activity against Y. pestis, whereas cefotaxime was the most effective non-parenteral antibiotic [16]. In Madagascar, although multiple-drug or   single-drug resistance to streptomycin conferred by plasmids had been documented in 1995 [17,18], testing of a total of 713 Y. pestis strains in Madagascar revealed no resistance of Y. pestis isolates in humans, rats or fleas in Madagascar after 1995 [20]. Generally, the plague host animals, particularly for those in sylvatic plague foci, were relatively remote from human living environments, and the Y. pestis strains inhabiting these hosts animals or fleas had relatively less exposure to antibiotics or bacteria with antibiotic resistance. Thus, the Y. pestis strains found in nature were generally susceptible to antimicrobial agents. However, a Y. pestis strain presenting multidrug-resistance to eight antimicrobial agents (streptomycin, chloramphenicol, tetracycline, sulfonamides, ampicillin, kanamycin, spectinomycin and minocycline) has been found in Madagascar [17]. A multidrug-resistant Y. pestis strain was isolated from a marmot in Mongolia in 2000, but the genetic characteristics and transferability of the strain were not reported [21]. In 2018, plasmid-mediated doxycycline resistance in a Y. pestis strain was reported in Madagascar (isolated from a rat in 1998) [21]. Still another case of resistance to streptomycin (25 mg/ml) has been observed in one Y. pestis strain isolated from Vietnamese rats [22].
Evaluation of the variations in MIC50 and MIC90 values of Y. pestis strains would provide valuable reference information regarding antibiotic resistance changes in different backgrounds. In this study, we generally observed no differences in MIC50 or MIC90 values among Chinese Y. pestis isolates, regardless of their source (humans, rodents and fleas), natural plague focus (Table 3) or year isolated. Plague generally occurs among wild animals, and human plague occasionally originates from major reservoirs or domestic animals, such as Marmota himalayana, Meriones unguiculatus, Spermophilus dauricus, Marmota baibacina, Spermophilus undulates, Ovis aries [23], cats and dogs, as well as some rodents (Mus musculus, Allactaga sibirica, Microtus oeconomus, Cricetulus migratorius and Ochotona daurica) or wild animals (lynxes, badgers and foxes).
In 1995, an isolate named 17/95, isolated from a patient with plague in Madagascar, exhibited multidrug-resistant traits to eight antimicrobial agents [17]. In addition, another isolate named 16/95, obtained in 1995 in a patient with plague in Madagascar [18], exhibited only streptomycin resistance. The MIC of streptomycin for strain 17/95 (orientalis) was above 2,048 mg/L [17], whereas that for strain 16/95 (orientalis) was 1,024 mg/L [18]. The resistance to streptomycin was conferred by a conjugative plasmid (pIP1202 in Y. pestis strain 17/95; pIP1203 in Y. pestis 16/95 strain), and the high-level resistance was due to the presence of streptomycin phosphotransferase activity [17].
Streptomycin is the preferred therapy for plague in China. A clinical Y. pestis isolate has exhibited resistance to streptomycin [11]. The strain (biovar antiqua) was isolated from a pneumonic plague outbreak in 1996 in China, in the Marmota himalayana Qinghai-Tibet Plateau plague focus. This was the first report of Y. pestis streptomycin resistance in China, and a novel mechanism of streptomycin resistance in Y. pestis was identified: mutation at 128 bp in the rpsL gene [11]. Subsequently, the same streptomycin resistance mechanism was reported in Madagascar. A Y. pestis strain with corresponding resistance was identified and found to be circulating in a pneumonic plague outbreak in the Faratsiho district in Madagascar in 2013. Another plague case with rpsL gene mutation was found in in a region of Madagascar in 1987 [24].
Antimicrobial therapy is the simplest component of the complex therapy required for patients with plague. Besides the streptomycin is considered as one of the most effective antibiotics for the treatment of plague [4]. In China, combination antibiotic therapy had been used in the treatment of human plague patients; for example, using streptomycin in combination with ciprofloxacin, norfloxacin or ceftriaxone sodium can shorten the course of disease and decrease the dosage of streptomycin [6,7]. In 2004, pneumonic plague occurred in Qinghai, and treatment with ceftriaxone sodium and ofloxacin combined with streptomycin and significantly shortened the course of disease [6]. In 2009, a plague outbreak occurred in Xinghai County, Hainan Prefecture, Qinghai province [7]. Treatment with streptomycin and ciprofloxacin in combination cured all patients within 18 days, a course markedly shorter than the general duration of the disease.
In China, various plague foci containing different reservoirs or vectors exist, representing the most widely distributed, most complicated and most active natural plague foci in the world. In recent years, fewer than ten human plague cases per year have been reported in China, all of which have been limited to remote areas with low populations, such as in Inner Mongolia, Gansu, Qinghai Province. In this study, the susceptibility of Y. pestis isolates to 12 antimicrobial agents provided an antibiotic resistance baseline in China. In addition, the emergence of streptomycin resistance in Y. pestis in China is a critical public health problem. This resistance would result in treatment failure; thus, antibiotic monitoring should be performed in real time during treatment of those plague cases. Furthermore, Y. pestis strains resistant to streptomycin may be involved in transmission of pneumonic plague outbreaks, as occurred in Tibet, China in 1996 [11]. A similar phenomenon and mechanism were reported in a pneumonic plague outbreak in 2013 in Madagascar, and these streptomycin resistant strains were believed to have spontaneously arisen in Y. pestis in the absence of antibiotic selective pressure [24]. Thus, antimicrobial sensitivity surveillance of Y. pestis isolates in animal plague epidemics or in human plague cases should be performed routinely. In addition, the causes of streptomycin resistance due to rpsL mutation in Y. pestis must be further studied experimentally.