Characterization of a potential Listeria monocytogenes virulence factor associated with attachment to fresh produce

Abstract

A study to determine the attachment of L. monocytogenes serotype 4b strain F2365 on vegetables and fruits was conducted. In an initial study, we screened 32 genes encoding surface proteins and lipases of the strain to find highly expressed genes on lettuce leaves. The results showed that transcription levels of LMOf2365_0413, LMOf2365_0498, LMOf2365_0859, LMOf2365_2052, and LMOf2365_2812 were significantly upregulated on lettuce leaves. In silico analysis showed that LMOf2365_0859 contains a putative cellulose binding domain. Thus, we hypothesized that this gene may be involved in an attachment to vegetables, and named it lcp (gene encoding Listeria cellulose binding protein [LCP]). lcp mutant (Δlcp) and lcp complement (F2365::pMAD::cat::lcp) strains were generated by homologous recombination. The abilities of a wild-type (WT) strain, the Δlcp strain, and the complemented strain to attach to lettuce leaves were evaluated, which indicated that the attachment of the Δlcp strain to lettuce was significantly less than that of the WT and the complemented strains. Similar results were observed for baby spinach and cantaloupe. Fluorescence microscopy and field emission scanning microscopy analysis further supported these findings. The binding of L. monocytogenes to cellulose was determined using cellulose acetate-coated plates. The results showed that a binding ability of the Δlcp strain was significantly lower than that of the wild type. Combined, these results strongly suggest that LCP plays an important role in an attachment to vegetables and fruits.

Fig 1

Transcription levels of genes encoding listerial surface proteins and lipases. The expression levels of 32 genes encoding surface proteins and lipases in L. monocytogenes colonized on lettuce leaves were measured at 8 and 16 h after incubation. LMOf2365_0413, LMOf2365_0498, LMOf2365_2052, LMOf2365_0859, and LMOf2365_2812 were upregulated at these time points. gap was used as a control gene. Transcription levels were expressed as log₂ values. Data were obtained from three independent experiments using triplicate RNA samples per experiment (n = 9). Data were analyzed by Student's t test. Error bars represent standard errors of the means (SEM).

Fig 2

In silico analysis of the upregulated genes of L. monocytogenes on lettuce leaves. The JCVI annotation file, KEGG database, and NCBI CBLAST module were used to select a candidate gene for the listerial attachment on lettuce leaves. Data obtained from CBLAST showed that LCP (2,027 aa) contains a putative CBD (at aa 20 to 144), 7 bacterial Ig-like (Big_3) domains, and an LPXTG motif (top). The surface protein has amino acid sequences similar to those of a CBD of endoglucanase D from Clostridium cellulovorans. ClustalW2 and ESPript 2.2 were used to generate the sequence alignment (bottom).

Fig 3

Construction of an in-frame deletion lcp mutant. The PCR products from sequence adjacent the 5′ and 3′ flanking regions of LMOf2365_0859 were amplified. The generated up- and downstream DNA fragments were digested with BamHI and SalI and with EcoRI and BglII, respectively. Each DNA fragment with pMAD::cat digested with the same restriction enzymes was ligated. pMAD::cat::lcp was transformed into L. monocytogenes F2365. The recombinant plasmid was incorporated into the chromosome of F2365 by a first homologous recombination at 43°C. F2365 retaining the chromosome with pMAD::cat::lcp incorporated was subcultured at 30°C to select a deletion mutant, mediated through a second homologous recombination. The complementation of F2365::pMAD::cat::lcp was generated after the first incorporation of the recombinant plasmid into the chromosome of F2365.

Fig 4

Confirmation of lcp deletion mutant and bacterial cell growth. (A) Wild-type, Δlcp, and complemented strains were confirmed by PCR using primers (0859F/R) designed from a region of the deleted gene. The size of PCR product is 180 bp. (B) Wild-type, Δlcp, and F2365::pMAD::cat::lcp strains were grown in BHI broth for 24 h at 37°C and 180 rpm. The growth kinetics for bacterial strains were measured at 2, 4, 8, 16, and 24 h by standard plate count. Data were obtained from three independent experiments using duplicate bacterial samples per experiment (n = 6). Data were analyzed by ANOVA. Error bars represent SEM.

Fig 5

Attachment assay and detection for all strains on lettuce leaves. (A) Attached wild-type, Δlcp, and F2365::pMAD::cat::lcp strains on lettuce leaves were homogenized, and the homogenates were plated on BHI agar plates. Data were obtained from three independent experiments using triplicate bacterial samples per experiment (n = 9). The difference in the percentage of attached bacteria relative to total bacterial numbers inoculated on lettuce leaves was analyzed by ANOVA. Error bars represent SEM. There was a significant difference (P < 0.05) between the wild-type and complemented strains and the Δlcp strain. (B) Fluorescence microscopy (Nikon, Tokyo, Japan) of all strains labeled with CFSE (5 nM) at a magnification of ×20 magnification with an FITC filter. (C) Attached WT, Δlcp, and F2365::pMAD::cat::lcp strains were observed under a JEOL JSM-6500F scanning electron microscope (JEOL USA, Peabody, MA) at 5 kV.

Fig 6

Attachment of L. monocytogenes strains on spinach and cantaloupe. Baby spinach leaves and cantaloupe skins inoculated with the wild-type, Δlcp, or F2365::pMAD::cat::lcp strain were homogenized, and the homogenates were plated on BHI agar plates. Data were obtained from three independent experiments using triplicate bacterial samples per experiment (n = 9). The difference in the percentage of attached bacteria relative to total bacterial numbers inoculated on spinach leaves (A) and cantaloupe skins (B) was analyzed by ANOVA. Error bars represent SEM. There was a significant difference (P < 0.05) between the wild-type and complemented strains and the Δlcp strain.

Fig 7

Comparison of the abilities of the wild-type, Δlcp, and F2365::pMAD::cat::lcp strains to bind to cellulose. Ninety-six-well plates were coated with 1% (wt/vol) cellulose acetate. The adhesion of L. monocytogenes to cellulose acetate using 0.5% (wt/vol) crystal violet was measured using a microplate reader at 590 nm. Three independent experiments with triplicate bacterial samples per experiment were used. The difference in OD values was analyzed by ANOVA. *, significant difference (P < 0.05) between the wild-type and complemented strains and the Δlcp strain.