Drug susceptibility and molecular epidemiology of Escherichia coli in bloodstream infections in Shanxi, China

Objectives We carried out a retrospective study to investigate the drug susceptibility and genetic relationship of clinical Escherichia coli isolates from patients with BSIs in Shanxi, China. Methods E. coli isolates causing BSIs were consecutively collected from June 2019 to March 2020. Antimicrobial susceptibility testing was performed by broth microdilution method. PCR was used to detect antimicrobial resistance genes coding for extended-spectrum β-lactamases (ESBLs), phylogenetic groups and seven housekeeping genes of E. coli. Results A total of 76 E. coli were collected. Antimicrobial susceptibility testing revealed that the top six E. coli resistant antibiotics were ampicillin (90.7%), ciprofloxacin (69.7%), cefazolin (65.7%), levofloxacin (63.1%), ceftriaxone and cefotaxime (56.5%). Among the 76 isolates, 43 produced ESBLs. Molecular analysis showed that CTX-M-14 was the most common ESBLs, followed by CTX-M-15 and CTX-M-55. Phylogenetic group D (42.2%) predominated, followed by group B2 (34.2%), group A (18.4%) and group B1 (5.2%). The most prevalent sequence types (STs) were ST131 (15/76), ST69 (12/76) and ST38 (6/76). Conclusions This study is the first to report the phenotypic and molecular characteristics of E. coli isolated from BSIs in Shanxi, China. Our results indicated a high prevalence of MDR in E. coli strains isolated from BSIs and a serious spread of ESBL genes in Shanxi, especially the epidemiological blaCTX-M. Phylogenetic analysis indicated genetic diversity among E. coli BSIs isolates.


INTRODUCTION
Bloodstream infections (BSIs), one of the most common severe infections, is also an important complication leading to extended hospitalization and increased mortality (Yoon et al., 2018). BSIs is occurred in about 2 million episodes and is responsible for about

Bacterial isolates
We performed a retrospective cross-sectional study of patients with E. coli BSIs between June 2019 and March 2020. Cases of E. coli BSIs were identified from the laboratory database in the Department of Clinical Laboratory, Shanxi Provincial People's Hospital. Each patient was included only once, if multiple blood cultures from the same patient were positive, only the first episode was reviewed and recorded. All isolates were identified by matrix-assisted laser desorption ionization-time of flight mass spectrometer (bioMérieux, Marcy-l'Étoile, France) and stored in broth containing 30% glycerol at −80 • C until further experiments.
Broth dilution test (cefotaxime and cefotaxime-clavulanic acid, ceftazidime and ceftazidime-clavulanic acid) was used as confirmatory test for ESBLs producers following CLSI criteria (CLSI, 2019). Either cefotaxime or cefotadime combined with clavulanic acid, the MIC was decreased by threefold concentration, which can be considered as ESBLs positive. Klebsiella pneumoniae ATCC 700603 was used as positive control for ESBLs production. According to the susceptibility results, the ESBLs genes were further detected in the positive isolates.

Phylogenetic group analysis and multilocus sequence typing
According to the well recognized phylogenetic grouping protocol proposed by Clermont, Bonacorsi & Bingen (2000), four major phylogenetic groups (A, B1, B2 and D) in the E. coli isolates were determined using the method of triple PCR based on three genetic markers, namely chuA, yjaA and TspE4.C2. chuA encodes outer membrane hemin receptor gene that involves in heme transport. yjaA encodes for gene responsible for cellular response to hydrogen peroxide and acid stress and TspE4.C2 DNA encodes for putative lipase esterase gene (Javed, Mirani & Pirzada, 2021). PCR was performed to determine the seven conserved housekeeping genes (adk, fumC, gyrB, icd, mdh, purA, and recA) of E. coli, and the primer pairs were listed in Table 2. The allelic profiles and STs were described by the combination of the seven alleles on line (http://mlst.warwick.ac.uk/mlst/dbs/Ecoli/).

Statistical analysis
The statistical analysis was performed using SPSS 25.0 (IBM, Armonk, NY, USA). When fitting a normal distribution, we presented continuous numerical variables as mean and standard deviation (SD). For categorical variables, results were expressed as the percentages of the groups from which they were derived. The χ 2 test was used to compare categorical variables. All tests were two-sided, and p < 0.05 was considered statistically significant.

Characteristics of total patient population
A total of 76 eligible E. coli isolates were enrolled during the study period. From the total of 76 E. coli causing BSIs they were isolates from females (42/76) and (34/76) from males. The age of patients ranged from 13 to 90 years and the mean age was 62.1 ± 14.9 years. The mean age did not differ significantly in ESBLs-producing isolates compared to non-ESBLs-producing E. coli.

DISCUSSION
The increased consumption of antimicrobial agents, the high prevalence and dissemination of drug resistance genes in pathogens, and the poor prevention and control strategies for infections lead to the increase of antimicrobial resistance (Parajuli et al., 2017). Approximately 700,000 deaths in antimicrobial resistance every year in the world, which is expected to soar to a staggering 10 million in 2050 (Huh et al., 2020). As pathogens'antimicrobial resistance rates and mortality in BSI patients increase, monitoring of microorganisms and antimicrobial resistance has become critical (Mehl et al., 2017).
Our study not only monitored the resistance phenotype of E. coli, but also detected the distribution of drug resistance genes and the genetic relationship of the isolates, which provided the basis for designing strategies for the treatment and prevention of these serious infections.
The present study showed that resistance of several commonly used antibiotics used to treat BSIs in our area, such as penicillins, cefazolin, ceftriaxone, cefotaxime, fluoroquinolones and folate pathway inhibitors, was high (Table 3), which was Table 4 Phylogenetic groups and genotypes in MLST of 76 E. coli isolates.

Sequence Type
Phylogenetic groups  consistent with previous report (Zhao et al., 2015). Fortunately, carbapenems, piperacillintazobactam, amikacin and tigecycline have low drug resistance rates, which should be considered for empirical treatment of E. coli isolates in our region. In this study, MDR accounted for 73.6% in E. coli causing BSIs, the value is lower than the result (85.6%) of another similar study in China by Ma et al. (2017), but much higher than that in E. coli causing pyelonephritis (40%), sepsis (32%) and skin and soft tissue infections (26%) (Ranjan et al., 2017). Our study showed that among the 43 ESBLs-producing E. coli, 97.7% were MDR, while among 33 non-ESBLs-producing, only 42.4% were MDR. Since the ESBLs genes are usually found in large plasmids which also contain other antimicrobial resistant genes, most ESBLs producing organisms are MDR isolates (Ma et al., 2017).

Resistance determinants No. of isolates
In this study, the proportion of ESBLs-producing E. coli is 56.5%, which was consistent with that in Zhejiang (57.6%) (Xiao et al., 2019b), but much higher than in Japan (26.1%), Vietnam (39.3%), Singapore (33%), and Brazil (12.8%) (Hung et al., 2019;Komatsu et al., 2018). Since the 2000s, CTX-M have replaced TEM as the most common type of ESBLs in many countries (Liao et al., 2017;Zou et al., 2019). In the present study, genotypes showed remarkable increase in the CTX-M (40/71) compared to TEM (20/71) in ESBLsproducing E. coli. It is noteworthy that bla CTX-M were strongly associated with ST131 (80% carried bla CTX-M ). We found that bla CTX-M-14 was the predominant bla CTX-M , followed by bla CTX-M-15 and bla CTX-M-55 , while in Germany, bla CTX-M-15 was the most common and bla CTX-M-14 was the second most frequently identified genotype (Rohde et al., 2020). CTX-M-55, a variant of CTX-M-15 that contains a substitution of A80V within the β-lactamase possessing enhanced cephalosporin-hydrolyzing activity, was rarely found in clinical isolates previously (Liao et al., 2017;Zhang et al., 2019). While, in some surveys, the prevalence of bla CTX-M-55 has surpassed bla CTX-M-15 and become the secondary genotype of bla CTX-M (Hu et al., 2018;Zhao et al., 2015). It was observed that the predominant of subtypes bla CTX-M gene may change with region and time. In addition, isolates producing bla OXA-1 also accounted for a larger proportion among the ESBLs-producers and mainly belonged to B2-ST131.
In this study we identified 28 STs among 76 E. coli isolates and MLST showed abundant genetic diversity in the E. coli. Similar to other studies in the world (Van Hout et al., 2020;Yasir et al., 2018), ST131 was the predominant ST. Moreover, we found that the predominant ESBLs type in the B2-ST131 (phylogenetic group-sequence type) isolates was CTX-M-14, rather than CTX-M-15 as reported in Australia (Harris et al., 2018), indicating an association of distinct CTX-M types with different settings due to various modes of transmission. In our study, ST69 was the most prevalent among non-ESBLs-producers, while one study from Netherlands showed that ST73 was the most common sequence type among non-ESBLs-producing isolates (Van Hout et al., 2020). This proved that the distribution of STs of E. coli vary significantly by region. Our study found that males accounted for 60.0% of patients infected with ST131 and females accounted for 83.3% of patients infected with ST69, consistent with a Canadian study (Holland et al., 2020). The reasons for genders differences in E. coli ST types are unclear and require further study.

CONCLUSIONS
In conclusion, data for this article described the drug sensitivity and molecular epidemiology of E. coli in patients with BSIs in Shanxi, China. We found that E. coli were highly resistant to commonly used antibiotics, except for carbapenems, piperacillin-tazobactam, amikacin and tigecycline. The B2-ST131 and D-ST69 clonal groups were the most common clinically relevant genotypes. Phylogenetic analysis showed genetic diversity among E. coli isolates. Better monitoring of the epidemiology of E. coli bacteremia is needed to develop and implement effective prevention strategies.