The Characteristic of Biofilm Formation in ESBL-Producing K. pneumoniae Isolates

Abstract

Klebsiella pneumoniae is a pathogen that commonly causes hospital-acquired infections. Bacterial biofilms are structured bacterial communities that adhere to the surface of objects or biological tissues. In this study, we investigated the genome homology and biofilm formation capacity of ESBL-producing K. pneumoniae. Thirty ESBL-producing K. pneumoniae isolates from 25 inpatients at Ruijin Hospital, Shanghai, were subjected to pulsed-field gel electrophoresis (PFGE) to estimate genomic relatedness. Based on the chromosomal DNA patterns we obtained, we identified 21 PFGE profiles from the 30 isolates, eight of which had high homology indicating that they may have genetic relationships and/or potential clonal advantages within the hospital. Approximately 84% (21/25) of the clinical patients had a history of surgery, urinary tract catheterization, and/or arteriovenous intubation, all of which may have increased the risk for nosocomial infections. Biofilms were observed in 73% (22/30) of the isolates and that strains did not express type 3 fimbriae did not have biofilm formation capacity. Above findings indicated that a high percentage of ESBL-producing K. pneumoniae isolates formed biofilms in vitro and even though two strains with cut-off of PFGE reached 100% similarity, they generated biofilms differently. Besides, the variability in biofilm formation ability may be correlated with the expression of type 3 fimbriae. Thus, we next screened four ESBL-producing K. pneumoniae isolates (Kpn5, Kpn7, Kpn11, and Kpn16) with high homology and significant differences in biofilm formation using PFGE molecular typing, colony morphology, and crystal violet tests. Kpn7 and Kpn16 had stronger biofilm formation abilities compared with Kpn5 and Kpn11. The ability of above four ESBL-producing K. pneumoniae isolates to agglutinate in a mannose-resistant manner or in a mannose-sensitive manner, as well as RNA sequencing-based transcriptome results, showed that type 3 fimbriae play a significant role in biofilm formation. In contrast, type 1 fimbriae were downregulated during biofilm formation. Further research is needed to fully understand the regulatory mechanisms which underlie these processes.

Figure 1

Biofilm formation in ESBL-producing K. pneumoniae strains. Biofilm formation by 30 ESBL-producing K. pneumoniae isolates was monitored with crystal violet staining and quantified with OD₅₇₀ values. OD₅₇₀ > 1 were classified as biofilm-forming strains, and OD₅₇₀ < 1 were classified as nonbiofilm-forming strains.

Figure 2

PFGE cluster analysis results of ESBL-producing K. pneumoniae strains. Dendrogram representing the genetic relationship between 30 ESBL-producing K. pneumoniae isolates after restriction with the XbaI enzyme. The gene cluster was formed on the dendrogram with a similarity cut off between 85% and 100% of which 22 PFGE stripe of ESBL-producing K. pneumoniae contained 6 gene clusters (A–F).

Figure 3

Differential analysis of biofilm formation by Kpn5, Kpn7, Kpn11, and Kpn16. (a) Biofilms formed on the 96-well plate, as demonstrated by crystal violet staining. (b) Quantification of biofilm formation by measuring OD₅₇₀ and OD₆₀₀ values. (c) Biofilm folds formed by the Kpn5, Kpn7, Kpn11, and Kpn16 strains on LB solid medium.

Figure 4

Sheep erythrocytes agglutination assay of ESBL-producing K. pneumoniae isolates. (a) Results of sheep erythrocyte agglutination by Kpn5, Kpn7, Kpn11, and Kpn16. Kpn7 and Kpn16 showed mannose-resistant agglutination reactions. (b) Sheep erythrocytes agglutination titers for Kpn5, Kpn7, Kpn11, and Kpn16.

Figure 5

Guinea pig erythrocyte agglutination assay of ESBL-producing K. pneumoniae isolates. (a) Results of guinea pig erythrocyte agglutination by Kpn5, Kpn7, Kpn11, and Kpn16. Kpn5 and Kpn11 showed mannose-sensitive agglutination reactions. (b) Guinea pig erythrocyte agglutination titers for Kpn5, Kpn7, Kpn11, and Kpn16.

Figure 6

Comparison of transcription levels as determined by RNA-seq and quantitative reverse transcription-PCR. 8 genes were selected and subjected to qRT-PCR. The corresponding log2 values were plotted against RNA-seq data. The R² for the two datasets was more than 0.90. R²: correlation coefficient.