Comparative Study of Genotype and Virulence in CTX-M-Producing and Non-Extended-Spectrum-β-Lactamase-Producing Klebsiella pneumoniae Isolates

TEXT

In parallel with the use of antibiotic drugs, the prevalence of Escherichia coli and Klebsiella pneumoniae strains producing CTX-M-type extended-spectrum β-lactamases (ESBLs) has increased worldwide (1). In addition to CTX-M-15-producing E. coli, CTX-M-15 has increased in prevalence in K. pneumoniae worldwide (2–4). In K. pneumoniae, bla_CTX-M-15 is carried mainly by IncFII-type plasmids (5). Production of the SHV-type ESBLs in K. pneumoniae is associated with an increased tendency to invade epithelial cells and expression of fimbrial adhesions (6). Although highly virulent clones expressing CTX-M-type β-lactamase, such as E. coli ST131, have been reported (7), the association of CTX-M enzyme production with virulence and fitness in K. pneumoniae is not clear. In this study, we compared genotypic and phenotypic characteristics between CTX-M-producing and non-ESBL-producing K. pneumoniae isolates from South Korea. In addition, the fitness cost of carrying the bla_CTX-M-15 gene-bearing plasmid was investigated.

In this study, 98 K. pneumoniae isolates, which were collected from patients with bacteremia from nine South Korean hospitals as a part of multicenter surveillance study from September to December 2008, were included (8). Thirty-three isolates were found to express bla_CTX-M genes (18 bla_CTX-M-14 and 15 bla_CTX-M-15), while 65 isolates did not produce any ESBL. ESBL activity was confirmed using the double-disc method. In vitro antimicrobial susceptibility testing was performed by a broth microdilution method, according to the CLSI guidelines (9). Multilocus sequence typing (MLST) was performed as described previously (http://www.pasteur.fr/recherche/genopole/PF8/mlst/Kpneumoniae.html) with some modifications. Pulsed-field gel electrophoresis (PFGE) was performed for all ST11 K. pneumoniae isolates (10). PCR assays were performed to monitor for the presence of genes previously found to be associated with virulence in K. pneumoniae (11, 12). The string test was used to determine the hypermucoviscosity phenotype (13). α-Hemolysin production was detected using a 5% sheep blood agar plate (14).

The transfer of the plasmid carrying bla_CTX-M-15 was accomplished using an E. coli DH5α strain as described previously (15). The plasmid carrying the bla_CTX-M-15 gene from an ST11 K. pneumoniae isolate, K01-12226, was used (5). The plasmid carrying bla_CTX-M-15 was transferred into five non-ESBL-producing ST11 K. pneumoniae isolates, K01-1054, K01-6053, K01-7096, K01-8102, and K01-8139, from E. coli DH5α as a donor (8). The presence of bla_CTX-M-15 in transconjugants was confirmed by PCR. A serum sensitivity assay was performed on clinical isolates and transconjugants, as previously described (16). Fisher's exact test was used to determine the significance of differences in serum resistance between strains using SPSS version 11.5 (SPSS, Chicago, IL, USA).

The CTX-M-producing K. pneumoniae isolates showed significantly greater resistance to most antibiotics except ampicillin, amikacin, and imipenem than did non-ESBL-producing isolates (P < 0.05) (Table 1). While cf29a, a gene in E. coli encoding adhesin CS31A and associated with diarrhea in humans, was found in only one CTX-M-14-producing, ST11 isolate, it was present in 14 non-ESBL-producing isolates (P = 0.018) (Table 1). Eight (72.7%) of 11 non-ESBL-producing ST11 isolates possessed the cf29a gene, and four ST163 isolates tested positive for it. In addition to cf29a, allS, which encodes an activator of the allantoin regulon, was found more frequently in non-ESBL-producing isolates (P = 0.05).

In MLST analysis, a total of 52 different STs were identified among the 98 K. pneumoniae isolates: 19 STs among the 33 CTX-M-producing isolates and 36 among the 65 non-ESBL-producing isolates (Table 2). Only three STs (ST11, ST15, and ST48) were detected in both CTX-M-producing and non-ESBL-producing isolates. This suggests that most CTX-M-producing K. pneumoniae isolates in South Korea did not occur by transfer of the bla_CTX-M gene into susceptible strains. In PFGE analysis, 16 ST11 isolates did not show exactly the same restriction pattern, but similarities in pattern could be identified. In particular, an approximately 388-kb fragment was found only in CTX-M-15-producing K. pneumoniae isolates. It was revealed to be a plasmid carrying bla_CTX-M-15, by PCR. In addition, PFGE analysis of transconjugants receiving a plasmid showed an additional 388-kb band (not shown). It may be premature to conclude that CTX-M-15-producing ST11 isolates arise by transfer of a bla_CTX-M-15-bearing plasmid into susceptible strains unique to South Korea, since ST11 is distributed among CTX-M-15-producing clones worldwide (17).

Non-ESBL-producing K. pneumoniae isolates showed higher serum resistance than did CTX-M-producing isolates (P = 0.004) (Fig. 1A). In addition, the hypermucoviscosity phenotype was more frequently identified in non-ESBL-producing isolates; only three CTX-M-producing isolates (9.1%) expressed the hypermucoviscosity phenotype compared to 19 non-ESBL-producing isolates (29.2%) (P = 0.038). None of the K. pneumoniae isolates produced α-hemolysin.

• Open in new tab
• Download powerpoint

FIG 1

Results of serum resistance assay are shown here as CFU viability. Error bars indicate standard deviations. (A) Serum resistance assay for all CTX-M-producing and non-ESBL-producing K. pneumoniae isolates. Significance is shown for the difference between CTX-M-producing and non-ESBL-producing K. pneumoniae isolates (*, P < 0.05). (B) Survival rate of each non-ESBL-producing K. pneumoniae isolate and its transconjugant after incubation for 2 h.

To understand the effects of the plasmid on fitness measures such as growth rate and serum resistance, plasmids carrying the bla_CTX-M-15 gene of an ST11 isolate were transferred to non-ESBL-producing ST11 isolates by conjugation. All of these transconjugants showed substantial increases in MICs for cephalosporins, with one exception (cefotaxime for T-6053). The CTX-M-producing, non-ESBL-producing, and transconjugant strains showed no significant differences in growth rates. Although serum resistance did not differ between CTX-M-producing and non-ESBL-producing ST11 K. pneumoniae isolates, transconjugants showed higher survival rates against serum than did their host isolates and donors (Fig. 1B). Comparing the survival rates against serum in pairs of transconjugant and host isolates, transconjugants showed significantly higher serum resistance than did their host isolates, except for the pair that included K01-1054 and T-1054. The traT gene in the plasmid may contribute to serum resistance, which explains the increased survival rates against serum in the transconjugants.

As a whole, non-ESBL-producing K. pneumoniae isolates were assumed to be more virulent than CTX-M-producing isolates. First of all, non-ESBL-producing isolates showed a higher rate of resistance against human serum than did CTX-M-producing K. pneumoniae isolates. Second, the hypermucoviscosity phenotype was more frequently found in non-ESBL-producing isolates. In addition, while CTX-M-producing isolates contained 7.64 virulence factors on average, 8.54 virulence factors, on average, were identified in non-ESBL-producing isolates (P = 0.005).

However, serum resistance did not differ between CTX-M-15-producing and non-ESBL-producing isolates of ST11. More importantly, four out of five transconjugants showed higher serum resistance than did their hosts (Fig. 1B). This suggests that plasmids with the bla_CTX-M-15 gene may confer virulence as well as antimicrobial resistance, although only one kind of plasmid (IncFII) was tested in this study. Thus, CTX-M-15-producing K. pneumoniae isolates may carry more than one positively selective trait, implying that antimicrobial-resistant strains could increase in prevalence even without antimicrobial pressure. Although increased serum resistance is generally associated with decreased virulence and fitness, credible evidence to the contrary has emerged (18). Global dissemination of highly pathogenic and resistant clones would be cause for great concern (7).