Complete genome sequences and genomic characterization of five plasmids harbored by environmentally persistent Cronobacter sakazakii strains ST83 H322 and ST64 GK1025B obtained from powdered infant formula manufacturing facilities

Abstract

Background: Cronobacter sakazakii is a foodborne pathogen that causes septicemia, meningitis, and necrotizing enterocolitis in neonates and infants. The current research details the full genome sequences of two extremely persistent C. sakazakii strains (H322 and GK1025B) isolated from powdered infant formula (PIF) manufacturing settings. In addition, the genetic attributes associated with five plasmids, pH322_1, pH322_2, pGK1025B_1, pGK1025B_2, and pGK1025B_3 are described.

Materials and methods: Using PacBio single-molecule real-time (SMRT^®) sequencing technology, whole genome sequence (WGS) assemblies of C. sakazakii H322 [Sequence type (ST)83, clonal complex [CC] 83) and GK1025B (ST64, CC64) were generated. Plasmids, also sequenced, were aligned with phylogenetically related episomes to determine, and identify conserved and missing genomic regions.

Results: A truncated ~ 13 Kbp type 6 secretion system (T6SS) gene cluster harbored on virulence plasmids pH322_2 and pGK1025B_2, and a second large deletion (~ 6 Kbp) on pH322_2, which included genes for a tyrosine-type recombinase/integrase, a hypothetical protein, and a phospholipase D was identified. Within the T6SS of pH322_2 and pGK1025B_2, an arsenic resistance operon was identified which is in common with that of plasmids pSP291_1 and pESA3. In addition, PHASTER analysis identified an intact 96.9 Kbp Salmonella SSU5 prophage gene cluster in pH322_1 and pGK1025B_1 and showed that these two plasmids were phylogenetically related to C. sakazakii plasmids: pCS1, pCsa767a, pCsaC757b, pCsaC105731a. Plasmid pGK1025B_3 was identified as a novel conjugative Cronobacter plasmid. Furthermore, WGS analysis identified a ~ 16.4 Kbp type 4 secretion system gene cluster harbored on pGK1025B_3, which contained a phospholipase D gene, a key virulence factor in several host-pathogen diseases.

Conclusion: These data provide high resolution information on C. sakazakii genomes and emphasizes the need for furthering surveillance studies to link genotype to phenotype of strains from previous investigations. These results provide baseline data necessary for future in-depth investigations of C. sakazakii that colonize PIF manufacturing facility settings and genomic analyses of these two C. sakazakii strains and five associated plasmids will contribute to a better understanding of this pathogen's survival and persistence within various "built environments" like PIF manufacturing facilities.

Keywords: Built environment; Complete genomes; Cronobacter sakazakii; PHASTER; Phage-plasmids; Plasmids; Whole genome sequencing.

Fig. 1

A Sequence alignment of C. sakazakii phage-plasmid class members, pH322_1 and GK1025B_1: Annotated genomes of known phage-plasmids were aligned and compared to identify conserved and divergent sequence features. The annotation of each gene is from NCBI. The inner circle represents the sequence clockwise and the scale marks indicate positions of annotated genes. GenBank annotations of the reference pCsaC105731a (100,874 bp, coding DNA sequence (CDS) in black arranged outside ring), pCS1 (110,093 pb, Blue), pCsa767a (109,716 bp, purple), pCsaC757b (109,716 bp, Green), pGK1025B_1 (101,769 bp, Red) and pH322_1 (100,741 bp, Tan) was downloaded as GFF files for the analysis using the default configuration on the PROKSEE server. Across the circular genomes, selected genes or regions of interest are shown as follows: Missing regions identified by the BLAST analysis on the CGView server’s PROKSEE software are shown as ‘gaps’ on each of the circular genomes. The analysis was carried out on PROKSEE Server from the Stothard Research Group (University of Alberta, Canada) that uses BLAST analysis to illustrate conserved and missing genomic sequences (available online: https://beta.proksee.ca/tools). B Sequence Comparison of C. sakazakii virulence plasmid class members pH322_2 and pGK1025B_2: Annotated genomes of some known virulence plasmids were aligned and compared to identify conserved and divergent sequence features. The annotation of each gene is from NCBI. The inner circle represents the sequence clockwise and the scale marks indicate positions of annotated genes. GenBank annotations of the reference pESA3 (131,196 bp, CDS in black arranged outside ring), pSP291_1 (118,136 bp, Green), pH322_2 (118,185 bp, light blue), and pGK1025B_2 (120,182 bp, purple) was downloaded as GFF files for the analysis using the default configuration on the PROKSEE server. Across the circular genomes, selected genes or regions of interest are shown as follows: Franco et al. [31], adapted siderophore loci with Cronobactin gene, Iron ABC transporter genes, (T6SS), parAB genes, and the cpa gene. Missing regions identified by the BLAST analysis on the CGView server’s PROKSEE software are shown as ‘gaps’ on each of the circular genomes. The analysis was on the PROKSEE Server from the Stothard Research Group (University of Alberta, Canada) that uses BLAST analysis to illustrate conserved and missing genomic sequences (available online: https://beta.proksee.ca/tools)

Fig. 2

A Comparison of SSU5 prophage features with known C. sakazakii phage-plasmid class members: Four known and two new plasmid sequences from this study were compared using PROKSEE with the annotations of the Salmonella prophage SSU5. The inner circle represents the sequence clockwise and the scale marks indicate positions of annotated genes. GenBank annotations of the reference phage-plasmid SSU5 (CDS in Black colored ring, arranged outside ring), pCS1 (Green), pCsa767a (purple), pCsaC757b (Tan), pCsaC105731a (Red), pGK1025B_1 (Teal) and pH322_1 (Mauve) were downloaded as GFF files for analysis using the default configuration on the PROKSEE server. Across the circular genomes, selected genes or regions of interest are shown as follows: Missing regions identified by the BLAST analysis on the CGView server’s PROKSEE software are shown as ‘gaps’ (white color) on each of the circular genomes. These plasmids contained a near-complete SSU5 phage. A BLAST analysis of 630 + WGS assemblies of Cronobacter revealed varied coverage of the phage sequences in many plasmids (See Additional file 3: Table S3). The analysis was carried out on the PROKSEE Server from the Stothard Research Group (University of Alberta, CA) that uses BLAST analysis to illustrate conserved and missing genomic sequences (available online: https://beta.proksee.ca/tools). B Mauve alignment of SSU5 illustrates variations in lengths of the phage-sequences in Cronobacter plasmids: Plasmids from C. sakazakii and C. muytjensii were compared using the Mauve progressive alignment tool (http://darlinglab.org/mauve/user-guide/progressivemauve.html, [27, 28]) implemented on Geneious suite 12. pCS1 from C. sakazakii NCTC 8155 was seen to be the largest plasmid with almost 110 kb when compared to pCmuyZ38_1 from C. muytjensii JZ38 and the two new plasmids from this study. A detailed analysis of these plasmids, and their inclusion in plasmid-finding pipelines, would enable the identification of SSU5-like sequences from the growing number of Cronobacter WGS datasets

Fig. 3

Sequence analysis of newly described C. sakazakii conjugative class member pGK1025B_3: Annotated genomes from pGK1025B_3 and other Cronobacter conjugative plasmids were compared using PROKSEE for identifying conserved and unique sequence features. The inner circle represents the sequence clockwise and the scale marks indicate positions of annotated genes. GenBank annotations of the reference pGK1025B_3 (46,528 bp, CDS in Black arranged outside ring), pC16KP0065-1 (168,415 bp, Teal), pC16KP0098-3 (49,279 bp, Purple), pC17KP0040-2 (56,619 bp, Dark Blue), pMG333 (134,435 bp, Tan), and pESA2 (31,208 bp, Green) were downloaded as a GFF file for analysis using the default configuration on the PROKSEE server. Across the circular genomes, selected genes or regions of interest are shown as follows: Missing regions identified by the BLAST analysis on the CGView server’s PROKSEE software are shown as ‘gaps’ on each of the circular genomes. The analysis was carried out on the PROKSEE Server from the Stothard Research Group (University of Alberta, CA) that uses BLAST analysis to illustrate conserved and missing genomic sequences (available online: https://beta.proksee.ca/tools)

Fig. 4

Comparative analysis of rep gene sequences identifies plasmid pGK1025B_3 as a unique category of conjugative plasmid in Enterobacteriaceae: A nearest neighbor joining phylogenetically tree was developed using a BLASTn analysis with the rep gene (gene locus CP078109) from Cronobacter sakazakii strain MOD1-GK1025B plasmid pGK1025B_3, complete sequence. This rep gene from this plasmid grouped separately and uniquely when compared to rest of the rep gene sequences from the other microorganisms. The top cluster contains the query (shown with an asterisk) and the same sequence as the unique hit. The parameters used to develop the tree included the rep (CP078109.1) gene from pGK1025B_3 in a BLASTn analysis against a database represented by Enterobacteriaceae and related endosymbionts (NCBI taxid:91347) using NCBI’s nearest neighbor algorithm. Bar marker represents 0.01 bp substitutions