Strain-specific differences in pili formation and the interaction of Corynebacterium diphtheriae with host cells

Abstract

Background: Corynebacterium diphtheriae, the causative agent of diphtheria, is well-investigated in respect to toxin production, while little is known about C. diphtheriae factors crucial for colonization of the host. In this study, we investigated strain-specific differences in adhesion, invasion and intracellular survival and analyzed formation of pili in different isolates.

Results: Adhesion of different C. diphtheriae strains to epithelial cells and invasion of these cells are not strictly coupled processes. Using ultrastructure analyses by atomic force microscopy, significant differences in macromolecular surface structures were found between the investigated C. diphtheriae strains in respect to number and length of pili. Interestingly, adhesion and pili formation are not coupled processes and also no correlation between invasion and pili formation was found. Using RNA hybridization and Western blotting experiments, strain-specific pili expression patterns were observed. None of the studied C. diphtheriae strains had a dramatic detrimental effect on host cell viability as indicated by measurements of transepithelial resistance of Detroit 562 cell monolayers and fluorescence microscopy, leading to the assumption that C. diphtheriae strains might use epithelial cells as an environmental niche supplying protection against antibodies and macrophages.

Conclusions: The results obtained suggest that it is necessary to investigate various isolates on a molecular level to understand and to predict the colonization process of different C. diphtheriae strains.

Figure 1

Adhesion of C. diphtheriae strains to D562 cell layers. D562 cells were infected with different C. diphtheriae strains. Besides DSM43989, which is tox⁺, the isolates are non-toxigenic. The cells were washed with PBS, detached with trypsin solution, lysed with Tween 20, and the number of colony forming units (cfus) was determined. Adhesion is expressed as percentage of the inoculum, showing means and standard deviations of ten independent measurements (biological replicates) with 3 samples each (technical replicates). All strains, except ISS4746 and ISS4749, show statistically significant differences in adhesion rates (students TTEST values below 0.04).

Figure 2

Invasion of epithelial cells by C. diphtheriae strains. D562 cells were infected with different C. diphtheriae strains (DSM43989 tox⁺, all others are non-toxigenic), washed, and incubated 2.0 (A), 8.5 (B) and 18.5 (C) hours with 100 μg ml^-1 gentamicin. Subsequently, cells were washed, detached with trypsin solution, lysed with Tween 20, and the number of intracellular cfus was determined. Invasion is shown as percentage of the inoculum internalized (means and standard deviations of three independent biological replicates with three samples each (technical replicates). Statistically relevant differences between the strains (based on students TTEST values below 0.05) are indicated by letters above columns.

Figure 3

Detection of intracellular C. diphtheriae in Detroit562 cells by immune-fluorescence microscopy. D562 cells were seeded on coverslips 48 h prior to infection and infected with C. diphtheriae (DSM43989 tox⁺, all others are non-toxigenic) for 4 h with at a MOI of 200 as described earlier [26]. Antibodies directed against the surface proteome of C. diphtheriae were used as primary, Alexa Fluor 488 goat anti-rabbit IgGs and Alexa-Fluor 568 goat anti-rabbit IgGs as secondary antibodies (A, D: intact D562, B, E: permeabilized D562, C, F: overlay with blue F-actin stain Phalloidin-Alexa-Fluor 647, A-C: ISS3319, D-F: ISS4060. Green stain in panels A and D indicate extracellular bacteria. Dark red stain in panels B and E indicates internalized C. diphtheriae, while adherent bacteria appear in light red. In the overlay (C, F) extracellular C. diphtheriae appear orange, while internalized bacteria are stained dark red. Scale bars: 20 μm.

Figure 4

Transepithelial resistance of polarized D562 monolayers grown on transwells. (A) Control experiments of cells, which were incubated without bacteria (open circles) and S. enterica serovar Typhimurium (open squares). (B) Incubation with C. diphtheriae strains DSM43989 (tox⁺, open stars), ISS4749 (inverted closed triangles), ISS4746 (closed triangles), ISS4060 closed circles, ISS3319 (closed square), DSM43988 (closed hexagons), and DSM44123 (closed diamonds). Experiments were carried out independently at least thrice and typical results are shown.

Figure 5

Ultrastructural analysis of the cell surface of C. diphtheriae strains. (A) Bacteria were fixed on glass slides by drying using compressed air. Atomic force microscopy was carried out under ambient laboratory conditions and operated in tapping mode. Scale bars: 500 nm. (B) AFM images were analyzed in respect to pili number per bacterium. For each strain pili of at least six bacteria were counted; error bars indicate deviations from mean values.

Figure 6

Strain-specific distribution and expression of pili-encoding genes. (A) Levels of spa gene transcripts in different C. diphtheriae strains. Total RNA was isolated from the indicated C. diphtheriae strains and hybridized with probes monitoring 16SrRNA for control as well as spa gene transcription. (B) PCR detection of spa genes. Chromosomal DNA of the indicated C. diphtheriae strains was used as template for PCR using specific oligonucleotide pairs for the spa genes indicated at the right side of the figure.

Figure 7

PCR detection of spa genes in C. diphtheriae strain NCTC 13129. Chromosomal DNA of C. diphtheriae strain NCTC 13129 was used as template for PCR using specific oligonucleotide pairs for the indicated spa genes. In all cases, DNA fragments of the expected size were amplified.