Characteristics of ESBL-positive Klebsiella pneumoniae isolated from paired children with and without diarrhea

Abstract

Background: Klebsiella pneumoniae producing extended-spectrum β-lactamase (ESBL) colonizing and transmitting in the intestine, especially in children, have significant public health implications. Investigating antibiotic resistance, antibiotic resistance genes (ARGs), virulence factor genes (VFGs), and genetic relationships may help us to explore the characteristics and differences of ESBL-positive K. pneumoniae in children with and without diarrhea.

Methods: After selecting and pairing, 26 pairs of 52 ESBL-positive K. pneumoniae strains were isolated from 323 children with diarrhea and 393 children without diarrhea. Antimicrobial susceptibility test and whole genome sequencing were performed to explore antibiotic resistance, ARGs, and VFGs. The genetic relationship was explored by conducting a maximum likelihood phylogenetic tree and investigating plasmid and sequence type (ST).

Results: All strains showed resistance to cephalosporins, with ESBL-producing genes widely carried (98.1%). Carbapenem-resistant K. pneumoniae (CRKP) were found in both groups. Hypervirulent K. pneumoniae (hvKP) were isolated from children with diarrhea carrying iucA on plasmid. The emergence of ST5670 CRKP and ST2108 hvKP highlighted the necessity for close monitoring of community-acquired K. pneumoniae.

Conclusions: Severe drug resistance was found among ESBL-positive K. pneumoniae strains isolated from children with and without diarrhea. Attention must be paid to ESBL-positive K. pneumoniae colonized in the intestine of children, and pathogen and ARG monitoring in children should be strengthened, even in healthy people.

Keywords: Klebsiella pneumoniae; Children; Diarrhea; ESBL.

Fig. 1

Flowchart for determining two groups of extended-spectrum β-lactamase (ESBL) positive Klebsiella pneumoniae isolates

Fig. 2

The heatmap and cluster analysis of antibiotic resistance (left) and carrying antibiotic resistant genes (ARGs) (right) were conducted, with the bar chart on the top showing the resistance rate (left) and the carrying rate of ARGs (right) in two groups. The statistical difference was marked in red at the top of the bars. The strains primarily aggregated into three primary clades according to antibiotic-resistant distribution (left) and also clustered in three distinct branches by ARGs (right). Clusters of phenotypes were matched to the clusters of ARGs and connected in different color lines [Heatmap of antibiotic resistance for 25 antibiotic agents including cefazolin (CFZ), cefepime (FEP), cefoxitin (FOX), ceftazidime (CAZ), ceftriaxone (CRO), cefuroxime (CXM), amoxicillin-clavulanate (AMC), ampicillin/sulbactam (SAM), piperacillin-tazobactam (TZP), aztreonam (ATM), ertapenem (EIP), imipenem (IPM), meropenem (MEM), amikacin (AMK), gentamicin (GEN), tobramycin (TOB), ciprofloxacin (CIP), levofloxacin (LVX), tetracycline (TET), tigecycline (TGC), minocycline (MIN), colistin (COL), nitrofurantoin (NIT), chloramphenicol (CHL), and trimethoprim-sulfamethoxazole (SXT)]

Fig. 3

The distribution of ARGs of two groups on plasmid and chromosome was showed. (a) The number of ARGs carried by each strain was shown in children with and without diarrhea distributed on plasmid or chromosome, and there was no difference between groups (P > 0.05), but a significant difference can be found on distributing positions (P < 0.05); (b) The distribution of ARGs for each type of antibiotics was illustrated in two groups. Different ARGs had varying preferences for chromosomes or plasmids. Aminoglycoside, macrolide, sulfonamide, tetracycline, chloramphenicol, rifampicin, and trimethoprim resistant genes were predominantly found on plasmids, whereas fosfomycin resistance genes were primarily located on chromosomes. Quinolone and β-lactam resistance genes were distributed on both plasmids and chromosomes, exhibiting distinct subtype variations

Fig. 4

Maximum likelihood (ML) phylogenetic analysis of strains. (a) One strain of ST15 from our study clustered with 142 ST15 K. pneumoniae strains, most of which were carbapenem-resistant. The strain from our study showed a potential genetic relationship with a strain isolated from a patient in Hunan in the same year (SNP = 58); (b) Two ST14 strains from our study were clustered with two carbapenem-resistant K. pneumoniae (CRKP) ST14 strains isolated from China. The high SNP distance (SNP ≥ 1670) indicated no direct genetic relationship among them; (c) One ST25 strain from our study clustered with four K. pneumoniae strains of ST25 from China. The strain isolated in our study and three strains in Genbank were hypervirulent K. pneumoniae (hvKP), with no CRKP identified. The ST25 hvKP strain in our study may be genetically distant from other hvKP ST25 strains (SNP ≥ 349); (d) All strains from our study were clustered in two branches. The heatmap of antibiotic resistance, carried ARGs, and carried vital virulent genes were followed