The globally first case of K. pneumoniae type blaKPC was discovered in the United States in 1996. Over time, carbapenem-resistant bacteria capable of producing blaKPC have spread to most areas in the United States, Israel and southern European countries (especially Greece and Italy), as well as to the South American continent and China[25]. The first blaKPC-2 strain of K. pneumoniae was discovered in Zhejiang Province of China in 2007, and since then blaKPC-2 pneumoniae has spread rapidly to other provinces and cities such as Jiangsu[17], Anhui [17], Fujian[26], Shanghai[27], Chongqing [28], Beijing[29]. Some studies suggested that the plasmids encoding the KPC-2 gene were different in different regions and that the propagation of blaKPC in these CRKPs was due to the horizontal transfer capacity of the plasmids or the mobilization of genetic elements in the plasmids, such as transposons and insertion sequences[17, 30]. In this study, we described the molecular epidemiological trends of CRKP in a tertiary hospital in China from 2016 to 2020, where the incidence of CRKP infections increased during this 5-year period and the overall number of CRKP detections showed a dramatic annual increase. The results highlighted the ability of horizontal transfer of the carbapenemase gene of CRKP to cooperate with clonal transmission, leading to the rapid spread of CRKP within this tertiary care hospital [30, 31]. Type blaKPC-2 of K. pneumoniae was also the main prevalent carbapenemase type isolated from this tertiary care hospital, which was consistent with the molecular epidemiological findings of CRKP conducted in most regions of China[32]. Among the strains collected, however, we found two carbapenemase types blaNDM-1 and blaNDM-5, which were rare in tertiary care hospitals. And some studies suggested that blaNDM-1 CRKP was more likely to cause infection in children [33], blaNDM-1 and blaNDM-5 were considered to be more resistant than blaKPC-2 in previous studies; but in this study blaNDM-1 and blaNDM-5 did not show any significant difference in resistance compared with blaKPC-2. And even more so, a CRKP strain of blaKPC-2 was not sensitive to any of the 30 antimicrobials applied in this study, which seemed to imply that the resistance of blaKPC to antimicrobials in this hospital was gradually evolving [34–36]. The outer membrane protein gene is also an important factor in the regulation of resistance in CRKP strains, and in our study we did not find strains with simultaneous deletion of OmpK35, Ompk36 and OmpK37, which suggested that clinical strains of Klebsiella pneumoniae with simultaneous deletion of the pore protein gene were uncommon [37]. OmpK35 and OmpK37 were detected in all CRKP isolates, and OmpK36 were detected in some CRKP isolates. Some researchers showed that OmpK37 was preferentially expressed in the strains with outer membrane protein deletions, and that all of the OmpK37 detected by us had four mutant sites and OmpK36 had 11 mutant sites, which might be responsible for the increased antimicrobial resistance of Klebsiella pneumoniae in hospitals[19, 38]. It was also shown that the strains with simultaneous deletion of OmpK35 and OmpK36 tended to be more resistant, but the deletion of a single pore protein gene also played a role in the enhancement of bacterial resistance [39, 40]. The OqxAB efflux pump gene was detected in more than half of our CRKP isolates, and OqxAB contributed significantly to the development of resistance to the antimicrobials of quinolones, tetracyclines, and chloramphenicol in Klebsiella pneumoniae [23, 41]. CRKP possessed potent resistance to 20 antimicrobials in the current 31 antimicrobial sensitivity assay (exceeding 90%), including an alarming 100% resistance to 17 antimicrobials, which was consistent with the results of researchers in Zhejiang Province, China [21]. This study showed that β-lactam antimicrobials was no longer resistant to carbapenem-resistant K. pneumoniae, probably because the KPC enzyme was able to catabolize a variety of β-lactamases, including carbapenems, cephalosporins, penicillins, and aminotransim, which were some of the commonly used β-lactam antimicrobials in clinical practice [9]. Whereas aminoglycosides might inhibit CRKP, and aminoglycoside-containing regimens might be an effective treatment option for the infections caused by CRKP strains, the situation is not promising because aminoglycoside-resistant genes are common in these strains [42]. The antimicrobials of Tetracycline and glycopeptide, on the other hand, showed strong inhibitory effects against CRKP, but a small number of CRKP strains in this study were still insensitive to the antimicrobials of tetracycline and glycopeptide and were able to withstand the pressure from the antimicrobials of tetracycline or glycopeptide, suggesting that CRKP was gradually becoming resistant to the antimicrobials of tetracycline and glycopeptide, and perhaps a combination of tetracycline and glycopeptide antimicrobials may be a treatment strategy for patients infected with CRKP resistant to the antimicrobials of tetracycline and/or glycopeptide [43].
The principle behind the MLST scheme is to use 7 housekeeping genes (internal nucleotide sequences of about 400 to 500 bases). Random integers are assigned to the unique sequences (alleles) of 7 housekeeping genes, and these random integers can combine into the unique alleles of each locus, the "allelic signature", to obtain the multi-locus sequence type (ST). To this day, MLST is considered the "gold standard" for species typing[44]. While in the United States and European countries, K. pneumoniae ST258 contributes significantly to the spread of carbapenem resistance, in the Chinese region ST11 dominates; and ST11 is a single motif variant (TonB) of ST258 and they are closely related. It was reported that ST11 and ST258 belonged to the clonal complex of CC258, which included ST11, ST258 and five other STSs (ST270, ST340, ST379, ST407 and ST418) [17, 32]. The CRKP isolated from this tertiary care hospital had a predominant ST11 type, and the number of ST11 type of CRKP in this tertiary care hospital showed a significant annual increase over the five-year period from 2016 to 2020. This is consistent with the results of other Chinese researchers [45, 46]. All CRKPs could be clearly divided into 5 taxa in the PFGE clustering analysis, and taxon A was full of ST15 clonal CRKP but had a low number of 3 strains compared to the results of a study in a medical center in northeastern China [47]. CRKP of taxa B, C, D and E were all dominated by ST11, but there were still some differences even among the same ST11 clonotype, which was probably because multiple incompatibility between plasmids could lead to different resistance genes carried on the plasmids of each isolate, thus resulting in individual differences [48].
It was found in the virulence gene test results that virulence genes were more abundant in some CRKPs compared with the results of other researchers, especially the detection rate of rmpA, magA, and iutA genes was significantly higher than that of isolates from a tertiary hospital in Chongqing area [49], but the results were closer to those of a tertiary care hospital in the Zhejiang area [21]. The rmpA, magA, and iutA virulence genes tend to be common in HVKP, which may be a sign that the CRKP within this hospital is evolving toward high virulence. The virulence plasmids which were considered non-conjugated in most previous studies, are present only in HVKP. Now there has been a worldwide spread of virulence plasmids from HVKP into CRKP, in which the plasmids are increasingly associated with the spread of virulence factors, leading to the spread of important pathogenic characteristics, with CRKP possessing highly virulent plasmids being considerably more pathogenic than before [50–52]. Interestingly, a blaNDM−1 carbapenemase-type CRKP strain had capsular serotype K54, which was the first time in the region that a CRKP isolate was found to have capsular serotype K54, however, according to related reports K54 was one of the supervirulent capsular serotypes of K. Pneumoniae [53]. The virulence gene combination of this blaNDM−1 isolate was WcaG-urea-wabG-fimH-entB-KfuB-alls-uge-ycf with two additional virulence genes of KfuB-alls compared with the virulence gene combinations of other isolates; and the results of the correlation study concluded that there was a strong correlation between kfuB, alls and K1 capsular serotype isolates, and that all K1 strains were considered positive for kfuB and alls [12]. However, the podosome serotype K54 of the CRKP strain isolated in this study possessed the combination of KfuB-alls, which is a very special finding. The virulence plasmid expressing K54 serotype might be accidentally obtained during the transmission of this isolate, which meant that this isolate had both high resistance of ST11 and high virulence of K54. Subsequent studies may be conducted to verify the plasmid horizontal transfer ability of this strain and to obtain the complete genetic information of its plasmid to determine the cause of this phenomenon [54].
The resistance genotypes of CRKP collected during our study were mainly the prevalent genotype of blaKPC-2, with blaNDM-1 and blaNDM-5 resistance genotypes accounting for very few proportions. The types of outer membrane proteins were dominated by OmpK35 and OmpK37. The vast majority of isolates contained both OqxA and OqxB efflux pump genes. ST types were mainly ST11, with ST15, ST273, ST626, and ST340 accounting for a few proportions. The strains with blaKPC and blaNDM resistant genotypes showed very strong resistance to common antimicrobials, and there were also individuals with abundant virulence genes in these CRKPs, which means that some CRKP strains may gradually become novel K. pneumoniae with mixed characteristics of super-resistance and high virulence, so it is a very necessary task to timely monitor, prevent and control CRKP epidemic disease in hospitals.