De Novo Sequencing Provides Insights Into the Pathogenicity of Foodborne Vibrio parahaemolyticus

Abstract

Vibrio parahaemolyticus is a common pathogenic marine bacterium that causes gastrointestinal infections and other health complications, which could be life-threatening to immunocompromised patients. For the past two decades, the pathogenicity of environmental V. parahaemolyticus has increased greatly, and the genomic change behind this phenomenon still needs an in-depth exploration. To investigate the difference in pathogenicity at the genomic level, three strains with different hemolysin expression and biofilm formation capacity were screened out of 69 environmental V. parahaemolyticus strains. Subsequently, 16S rDNA analysis, de novo sequencing, pathogenicity test, and antibiotic resistance assays were performed. Comparative genome-scale interpretation showed that various functional region differences in pathogenicity of the selected V. parahaemolyticus strains were due to dissimilarities in the distribution of key genetic elements and in the secretory system compositions. Furthermore, the genomic analysis-based hypothesis of distinct pathogenic effects was verified by the survival rate of mouse models infected with different V. parahaemolyticus strains. Antibiotic resistance results also presented the multi-directional evolutionary potential in environmental V. parahaemolyticus, in agreement with the phylogenetic analysis results. Our study provides a theoretical basis for better understanding of the increasing pathogenicity of environmental V. parahaemolyticus at the genome level. Further, it has a key referential value for the exploration of pathogenicity and prevention of environmental V. parahaemolyticus in the future.

Keywords: Vibrio parahaemolyticus; de novo sequencing; mouse model; pathogenicity; virulence.

Figure 1

Phylogenetic analysis and biofilm formation capacity of V. parahaemolyticus. (A) Standardized biofilms formed by 69 V. parahaemolyticus environmental strains. Based on the standardized biofilm values, the biofilm formation (BF) capacity of the strains was divided into four groups, viz., strong, moderate, weak, and no capacity. (B) Maximum likelihood tree of 69 V. parahaemolyticus environmental strains based on the 16S rDNA phylogenetic analysis. The bootstrap percentage value was obtained from 1000 samplings. The 69 strains were divided into 6 subclades. (C) Maximum likelihood tree for V. parahaemolyticus strains Vp. 1474, Vp. 1496, and Vp. 1513 based on the 16S rDNA phylogenetic analysis using type strains in Vibrio genus as reference. The bootstrap percentage value was obtained from 1000 samplings. Red solid circle: selected strains of Vp. 1474, Vp. 1496, and Vp. 1513; black solid triangle: reference Vibrio strains.

Figure 2

General genome components analysis of V. parahaemolyticus strains Vp. 1474, Vp. 1496, and Vp. 1513. (A) Statistical histogram of coding gene length distribution for the three strains. Coding gene zones were measured as 0 to 400 bp, 400 to 1,000 bp, 1,000 to 2,000 bp, and longer than 2000 bp as groups of short genes, medium long genes, long genes, and very long genes, respectively. (B) Percentage accumulative bar diagram of coding gene length distribution for the three strains. Coding gene zones were measured in 100 bp; 0 to 400 bp, 400 to 1,000 bp, 1,000 to 2,000 bp and longer than 2000 bp as groups of short genes, medium long genes, long genes and very long genes, respectively. (C) Accumulative bar diagram of genomic islands (GIs) for the three strains. GIs were divided into long islands and short islands by the length of 15 kb. (D) Statistical histogram of CRISPRs and prophages for the three selected strains. The component numbers of CRISPRs or prophages are shown on the Y-axis. (E) Statistical data of coding gene length distribution for the three strains. (F) Statistical data of GIs for the three strains. (G) Detailed characteristics of the predicted prophages for the three strains. (H) Detailed characteristics of the predicted CRISPRs for the three strains.

Figure 3

Gene ontology (GO) annotation of V. parahaemolyticus strains Vp. 1474, Vp. 1496, and Vp. 1513. (A) Statistical histogram of gene annotation to cellular component class of the three strains. A total of 13 parts were predicted in the genome of the three strains. Gene numbers are shown on the Y-axis. (B) Statistical histogram of gene annotation to biological process class of the three strains. A total of 24 parts were predicted within genome genes of the three strains. Gene numbers are shown on the Y-axis. (C) Statistical histogram of gene annotation to molecular function class of the three strains. A total of 10 parts were predicted within genome genes of the three strains. Gene numbers are shown on the Y-axis. (D) Statistical data of annotated genes based on GO ontology for the three strains.

Figure 4

KEGG pathway annotation of V. parahaemolyticus strains Vp. 1474, Vp. 1496, and Vp. 1513. (A) Statistical bar diagram of genes predicted to cellular processes pathway of the three strains. A total of 4 pathways were predicted within genome genes of the three strains. Gene numbers are shown on the horizontal axis. (B) Statistical bar diagram of genes predicted for environmental information processing pathways in the three strains. A total of 2 pathways were predicted within genome genes of the three strains. Gene numbers are shown on the horizontal axis. (C) Statistical bar diagram of genes predicted for genetic information processing pathway of the three strains. A total of 4 pathways were predicted within genome genes of the three strains. Gene numbers are shown on the horizontal axis. (D) Statistical bar diagram of genes predicted to human diseases pathways of the three strains. A total of 7 pathways were predicted within genome genes of the three strains. Gene numbers are shown on the horizontal axis. (E) Statistical bar diagram of genes predicted to metabolism pathways of the three strains. A total of 11 pathways were predicted within genome genes of the three strains. Gene numbers are shown on the horizontal axis. (F) Statistical bar diagram of genes predicted to organismal systems pathways of the three strains. A total of 8 pathways were predicted within genome genes of the three strains. Gene numbers are shown on the horizontal axis. (G) Statistical data of KEGG pathway annotation for the three strains.

Figure 5

COG function classification of V. parahaemolyticus strains Vp. 1474, Vp. 1496, and Vp. 1513. (A) Statistical histogram of gene classified to cellular processing and signaling cluster of the three strains. A total of 10 sub-clusters were predicted within genome genes of the three strains. Gene numbers are shown on the Y-axis. (B) Statistical histogram of gene classified to information storage and processing cluster of the three strains. A total of 5 sub-clusters were predicted within genome genes of the three strains. Gene numbers are shown on the Y-axis. (C) Statistical histogram of gene classified to metabolism cluster of the three strains. A total of 8 sub-clusters were predicted within genome genes of the three strains. Gene numbers are shown on the Y-axis. (D) Statistical histogram of poorly classified genes of the three strains. Only one sub-cluster was predicted to have general function, and the functions of the remaining genes were unknown. Detailed gene numbers are shown on the Y-axis. (E) Statistical data of annotation based on COG database for the three strains.

Figure 6

Virulence and pathogenicity analysis of V. parahaemolyticus strains Vp. 1474, Vp. 1496, and Vp. 1513. (A) Accumulative bar diagram of protein distribution in Type III secretion system (T3SS) of the three strains. T3SS were predicted and divided into effective and non-effective proteins, and their numbers are shown on the Y-axis. (B) Accumulative bar diagram of protein distribution in major Type N secretion system (TNSS) of the three strains. Four major TNSS were predicted in the three strains, viz., T2SS, T3SS, T4SS, and T6SS, and the number of TNSS-associated proteins are shown on the Y-axis. (C) Accumulative bar diagram of secretory proteins distribution of the three strains. Secretory proteins were divided into 3 groups, namely secreted proteins, transmembrane structural proteins, and signal peptide proteins. Numbers and detailed distribution of proteins are shown on the Y-axis. (D) Statistical data of predicted T3SS and other secretion-associated proteins for the three strains. (E) Statistical histogram of genes that are involved in pathogen-host interaction of the three strains. A total of 8 classes were predicted within genome genes of the three strains. Gene numbers are shown on the Y-axis. (F) Statistical histogram of genes annotated to antibiotic resistance and virulence of the three strains. Genes associated with antibiotic resistance were predicted and aligned with CARD and ARDB databases; genes annotated to virulence were predicted and aligned with VFDB database. Detailed numbers are shown on the Y-axis. (G) Statistical data of gene annotation based on PHI database for the three strains.

Figure 7

Genomic visualization analysis of V. parahaemolyticus strains Vp. 1474, Vp. 1496, and Vp. 1513. Three genome-wide maps represented the combination of the prediction results of the coding genes. Results of gene distribution analysis, COG classification, KEGG pathway, GO annotation, and non-coding RNA are presented from the outermost to innermost rings.

Figure 8

V. parahaemolyticus in vivo infection and data statistics. (A) Operation of in vivo infection and data presentation. Suspension cultures of the selected strains, viz., Vp. 1474, Vp. 1496, and Vp. 1513, were adjusted to optical densities (OD) of 0.1, 0.2, and 0.3, and intraperitoneally injected into each group (9 groups, 8 mice per group). (B) Survival analysis of mice infected with V. parahaemolyticus under low concentration of injection (OD=0.1). (C) Survival analysis of mice infected with V. parahaemolyticus under moderate concentration of injection (OD=0.2). (D) Survival analysis of mice infected with V. parahaemolyticus under high concentration of injection (OD=0.3). Grey patch: the surviving mice that were observed in a poor living condition. The survival rate was updated every hour in acute infection period (≤12 h), and subsequently updated every 12 h in infection recovery period. The experiment continued for 4 days.

Figure 9

Antibiotic resistance analysis of mice infected by V. parahaemolyticus strains Vp. 1474, Vp. 1496, and Vp. 1513. (A) Representative results of antimicrobial susceptibility test using Kirby-Bauer method. Black arrow: paper position of piperacillin (PRL) on the representative plates; blue arrow: paper position of cephalothin (KF) on the representative plates; red arrow: paper position of cephazolin (KZ) on the representative plates; black line: zone diameter of V. parahaemolyticus strains resistant to PRL; blue line: zone diameter of V. parahaemolyticus strains resistant to KF; red line: zone diameter of V. parahaemolyticus strains resistant to KZ. (B) Statistical histogram of antibiotic resistance to PRL, KF, and KZ for the three strains. For PRL, the zone diameters for resistant, intermediately sensitive, and sensitive are ≤1.7 cm, 1.8 to 2.0 cm, and ≥2.1 cm, respectively; for KF and KZ, the zone diameters for resistant, intermediately sensitive, and sensitive are ≤1.4 cm, 1.5 to 1.7 cm, and ≥1.8 cm, respectively.