Coexistence of tet(A) and blaKPC-2 in the ST11 hypervirulent tigecycline- and carbapenem-resistant Klebsiella pneumoniae isolated from a blood sample

Carbapenem-resistant Klebsiella pneumoniae are distributed worldwide. This study aimed to characterize a hypervirulent tigecycline-resistant and carbapenem-resistant Klebsiella pneumoniae strain, XJ-K2, collected from a patient’s blood. We tested antimicrobial susceptibility, virulence, and whole-genome sequencing (WGS) on strain XJ-K2. WGS data were used to identify virulence and resistance genes and to perform multilocus sequence typing (MLST) and phylogenetic analysis. Three novel plasmids, including a pLVPK-like virulence plasmid (pXJ-K2-p1) and two multiple resistance plasmids (pXJ-K2-KPC-2 and pXJ-K2-p3), were discovered in strain XJ-K2. The IncFII(pCRY) plasmid pXJ-K2-p3 carried the dfrA14, sul2, qnrS1, blaLAP-2, and tet(A) resistance genes. The IncFII(pHN7A8)/IncR plasmid pXJ-K2-KPC-2 also carried a range of resistance elements, containing rmtB, blaKPC-2, blaTEM-1, blaCTX-M-65, and fosA3. MLST analysis revealed that strain XJ-K2 belonged to sequence type 11 (ST11). Seven complete phage sequences and many virulence genes were found in strain XJ-K2. Meanwhile, antimicrobial susceptibility tests and G. mellonella larval infection models confirmed the extensively drug resistance (XDR) and hypervirulence of KJ-K2. To our knowledge, this is the first observation and description of the ST11 hypervirulent tigecycline- and carbapenem-resistant K. pneumoniae strain co-carrying blaKPC-2 and the tet(A) in a patient’s blood in China. Further investigation is needed to understand the resistance and virulence mechanisms of this significant hypervirulent tigecycline- and carbapenem-resistant strain.


Introduction
Klebsiella pneumoniae is a common Gram-negative opportunistic pathogen that causes multiple diseases, including pneumonia, bacteremia, liver abscess, and urinary tract infections [1]. Carbapenem-resistant K. pneumoniae (CRKP) and hypervirulent K. pneumoniae (hvKP) have emerged as the two major types of clinically significant pathogens in China [2]. CRKP is commonly resistant to multiple antibiotics, including carbapenems, quinolones, and aminoglycosides [3]. These widespread CRKP isolates represent an antibiotic resistance threat of the highest priority, limiting therapeutic options for these multidrug resistance strains [4]. In China, the most common clinical CRKP isolates belong to sequence type (ST) 11, considered the most infectious clone responsible for the increasing prevalence of CRKP [5].
HvKP is associated with high morbidity and mortality because it causes life-threatening and community-acquired infections, including liver abscesses, endophthalmitis, pneumonia, and meningitis in healthy individuals [6]. HvKP was first reported in Taiwan in the mid-1980s and 1990s [7]. In the decades since then, hvKP has been reported globally [8]. These hypervirulent strains have a hypermucoviscous phenotype that is detectable as a positive "string test" result and a hypervirulent phenotype that is attributable to the presence of a large pLVPK-like virulence plasmid, harboring two capsular polysaccharide (CPS) regulator genes (rmpA and rmpA2) and several siderophore gene clusters [6].
Most hvKP strains are susceptible to commonly used antibacterial agents except ampicillin [9]. Recent studies have shown that hvKP strains can acquire a carbapenemaseencoding plasmid to transform into carbapenem-resistant hvKP (CR-hvKP) strains, and CRKP strains can evolve to become CR-hvKP strains by acquiring a pLVPK-like virulence plasmid [10]. These strains have made clinical treatment and infection control more difficult because they are simultaneously multidrug-resistant, hypervirulent, and highly transmissible [11]. Due to the emergence of CRKP strains, tigecycline was considered as a last-resort treatment for CRKP infections [12]. In recent years, however, some tigecycline-and carbapenem-resistant K. pneumoniae (TCRKP) have been reported, which has further limited the concern with respect to therapeutic selection [13]. However, until now, the mechanisms underlying K. pneumoniae's resistance to tigecycline have not been fully understood. Previous studies showed that overexpression of the efflux pumps and alteration in the tigecycline target site played important roles in mediating tigecycline resistance [14].
In addition, recent research indicated that deletion of ramR and tet(A) is associated with resistance to tigecycline in K. pneumoniae [13]. Although Gu et al. reported the emergence of bla KPC-2 and the tet(A) in the ST11 hypervirulent TCRKP isolated from a patient's gut, the occurrence of a TCRKP strain with these characteristics in a patient's blood has not previously been reported [15]. This study reports a hypervirulent TCRKP strain, strain XJ-K2, isolated from a patient's blood in China. To our knowledge, this is the first report of the ST11 hypervirulent tigecyclineand carbapenem-resistant K. pneumoniae strain co-carrying bla KPC-2 and tet(A) plasmids that have been detected in patient blood in China.

Bacterial strain
In May 2018, K. pneumoniae strain XJ-K2 was isolated from a blood sample collected from a patient at a teaching hospital in Shanghai, China.

Antimicrobial susceptibility testing
The broth microdilution method was used to determine the minimum inhibitory concentration (MIC) values of 27 antimicrobial agents. The positive control well (without antibiotics) showed significant bacterial growth and the MIC values could be read. The MICs were interpreted following the Clinical and Laboratory Standards Institute (CLSI, 2018). As there are no CLSI breakpoints for tigecycline and colistin, the European Committee on Antimicrobial Susceptibility Testing (EUCAST) followed the MIC interpretation. Escherichia coli ATCC 25,922 was used as quality control strain for antimicrobial susceptibility testing.

Virulence testing for hypervirulent K. pneumoniae strain XJ-K2
A hypermucoviscous phenotype was determined using the string test previously described [6]. The Galleria mellonella (G. mellonella) model was used to evaluate the virulence of K. pneumoniae strains [3]. G. mellonella larvae (~ 300 mg; Tianjin Huiyude Biotech Company, Tianjin, China) were used to test the virulence of the strain XJ-K2. Each group selected ten randomly selected insects weighing 250-350 mg. Overnight cultures of strains XJ-K2, WCHKP030925 (hypervirulent control), and WCHKP13F4 (low virulence control) were washed with phosphate-buffered saline (PBS) and further adjusted with PBS to concentrations of 1 × 10 5 , 1 × 10 6 , and 1 × 10 7 CFU/ml [16]. Ten microliters of inoculum was injected into G. mellonella larvae through the last left pro-leg, followed by incubation at 37 ℃ in darkness [3]. The number of dead larvae was counted at 12-h intervals up to 72 h after the incubation.

Phylogenetic analysis
In this study, Unipro UGENE 1.32 (http:// ugene. net/) was used to perform phylogenetic analysis. We selected the genome sequences of representative K. pneumoniae strains that carry a pLVPK-like virulence plasmid to construct a phylogenetic tree for K. pneumoniae strain XJ-K2.

Isolation of a hypermucoviscous multiresistant K. pneumoniae strain
Klebsiella pneumoniae XJ-K2 was isolated from the Intensive Care Unit (ICU) of Huashan Hospital of Fudan University in Shanghai in May 2018. It was derived from a blood sample of a patient. The patient was admitted to the ICU with cerebral hemorrhage and was treated with indwelling catheterization. The urinary tract infection combined with bacteremia could be a risk factor for the patient's poor prognosis. Antimicrobial susceptibility testing showed that strain XJ-K2 was extensively drug-resistant (XDR). Therefore, the patient was treated with polymyxin B 120 mg/day. After 1-week follow-up, no obvious abnormalities were found in the blood culture, and clinical symptoms were obviously improved. XDR-K. pneumoniae is considered a huge health concern due to the few treatment options available in the clinic. XJ-K2 was determined as a hypermucoviscous strain by the "string test." Hypermucoviscous K. pneumoniae was diagnosed by a positive string test (> 5 mm) [6].
The G. mellonella model was used to evaluate the virulence of K. pneumoniae strains in vivo. G. mellonella was infected by K. pneumoniae strains through exposure to inoculum concentrations of 1 × 10 7 CFU/ml for 48 h, and the survival rates of G. mellonella infected by strains XJ-K2, WCHKP030925, and WCHKP13F4 were 0.0%, 10.0%, and 60%, respectively (Fig. 1). Based on these results, strain XJ-K2 was identified as hypervirulent K. pneumoniae.
The broth microdilution method was used to determine the minimum inhibitory concentration (MIC) of 27 antimicrobial agents. The results showed that strain XJ-K2 was extensively drug-resistant (XDR), and high-level resistance was observed against broad-spectrum cephalosporins, aminoglycosides, fosfomycin, and folate pathway inhibitors (Table 1). Of note, MIC > 2 μg/ml of polymyxin B in K. pneumoniae was regarded as being resistant to polymyxin B according to EUCAST. In this study, the MIC of polymyxin B was 2 μg/ml, close to the critical value of polymyxin B resistance, which might be a potential threat to clinical treatment.

Discussion
CR-hvKP strains have been identified in multiple sequence types, including ST23, ST11, ST36, and ST65 [22]. Gu et al. reported that some ST11 CRKP strains transform into CR-hvKP by obtaining a pLVPK-like virulence plasmid [10]. In addition, Liu et al. reported that ST23 CR-hvKP emerged following its acquisition of carbapenemase plasmids [23], suggesting that the horizontal transfer of virulence plasmids into CRKP strains and carbapenemase plasmids into hvKP strains has substantially contributed to the emergence of multiple sequence types of CR-hvKP. In this study, strain XJ-K2 was resistant to tigecycline, meropenem, and imipenem (Table 1), which suggests that strain XJ-K2 is a TCRKP. The ST11 hypervirulent TCRKP strain XJ-K2 revealed two typical features of hvKP, i.e., a hypermucoviscous phenotype and a pLVPK-like virulence plasmid encoding the virulence genes rmpA, rmpA2, iucABCD, and iutA. Hypervirulent TCRKP in patient blood limits therapeutic options and poses a huge threat to public health. Analysis of its genome sequence revealed that multiple resistance determinants (MDRs) were present in the ST11 hypervirulent TCRKP strain (Table 4), which is the first report of the first time that an ST11 hypervirulent TCRKP strain co-producing bla KPC-2 and the tet(A) has been isolated from patient blood in China. Phylogenetic analysis of strain XJ-K2 revealed that it is closely related to strain XJ-K1 (CP032163, ST11 and Shanghai, China), suggesting that they share a common ancestor. One virulence plasmid (pXJ-K2-p1) and two MDR plasmids (pXJ-K2-KPC-2 and pXJ-K2-p3) were identified in the hypervirulent TCRKP strain XJ-K2. The plasmid pXJ-K2-p1 was highly similar to the typical virulence plasmid pLVPK, which was found in hvKP strain CG43. These two plasmids carried aerobactin-related genes (iucABCD and iutA) and capsular polysaccharide regulator genes (rmpA and rmpA2), which are characteristic features of hvKP [24]. Many virulence genes might have helped this strain to grow and replicate during host infection.
The other MDR plasmid, pXJ-K2-p3, carried an IncFII(pCRY) backbone with five resistance determinants: dfrA14, sul2, qnrS1, bla LAP-2 , and the tet(A). Furthermore, a total of seven complete phage sequences were detected in the strain XJ-K2 chromosome that contributed to the transfer of resistance genes and virulence factors [26]. As previously reported, there was a slight increase in the MIC value of tigatin in isolates carrying tet(A) [27]. Our study integrated the tet(A) was integrated into pXJ-K2-p3 via the insertion sequences IS26 and ΔTnAs1. The high mobility of IS26 could mediate the horizontal transmission of resistance genes or the functional deficiency of partial genes, which could be related to the genetic adaptation of plasmids under antibiotic pressure. Inserted tetA resulted in low levels of tigecycline resistance, which may be related to efflux mechanisms [28]. Interestingly, ramR gene was found on the chromosome of the XJ-K2 strain, which might be Fig. 5 Comparative plasmid map comparing typical virulence plasmid pLVPK (AY378100) with the pLVPKlike virulence plasmid pXJ-K2-p1 from strain XJ-K2. The map was drawn with BLAST Ring Image Generator by using pLVPK as a reference. Sequence alignment revealed that pXJ-K2-p1 showed 90% query coverage and 99.06% identity with pLVPK meaningful regarding increased resistance. Therefore, we speculated that K. pneumoniae carrying ramR and tet(A) tended to evolve tigecycline resistance under selective pressure more easily. However, it remains unknown whether tet(A) played a major role in tigecycline resistance, which needs further study. In191, from the dfrA14 region, IntI1, and dfrA14 were integrated into the plasmid via insertion sequence IS26. In the plasmid pKF3-140 [29], class 1 integron harbored three resistance genes, specifically dfrA17, aadA5, and sul1, which are associated with resistance to trimethoprim, aminoglycoside, and sulfonamide, suggesting that class 1 integron is vital in increasing multidrug resistance among bacterial strains [30].

Conclusion
In conclusion, we have reported an ST11 hypervirulent TCRKP strain, XJ-K2, isolated from patient blood. We characterized the genomic features of strain XJ-K2 via WGS and bioinformatic analysis. To our best knowledge, this is the first report of the ST11 hypervirulent TCRKP strain co-carrying bla KPC-2 and the tet(A) has been isolated from patient blood in China. In addition, we should be devoted to concerning the spread of tet(A) and ramR in K. pneumoniae clinical isolates, which may be potential threats leading to tigecycline resistance. Therefore, active surveillance of this and other hypervirulent TCRKP strains and the development of new drugs are required to control the growth of these strains and prevent their transmission in hospitals.