Dissection of the role of pili and type 2 and 3 secretion systems in adherence and biofilm formation of an atypical enteropathogenic Escherichia coli strain

Abstract

Atypical enteropathogenic Escherichia coli (aEPEC) strains are diarrheal pathogens that lack bundle-forming pilus production but possess the virulence-associated locus of enterocyte effacement. aEPEC strain 1551-2 produces localized adherence (LA) on HeLa cells; however, its isogenic intimin (eae) mutant produces a diffuse-adherence (DA) pattern. In this study, we aimed to identify the DA-associated adhesin of the 1551-2 eae mutant. Electron microscopy of 1551-2 identified rigid rod-like pili composed of an 18-kDa protein, which was identified as the major pilin subunit of type 1 pilus (T1P) by mass spectrometry analysis. Deletion of fimA in 1551-2 affected biofilm formation but had no effect on adherence properties. Analysis of secreted proteins in supernatants of this strain identified a 150-kDa protein corresponding to SslE, a type 2 secreted protein that was recently reported to be involved in biofilm formation of rabbit and human EPEC strains. However, neither adherence nor biofilm formation was affected in a 1551-2 sslE mutant. We then investigated the role of the EspA filament associated with the type 3 secretion system (T3SS) in DA by generating a double eae espA mutant. This strain was no longer adherent, strongly suggesting that the T3SS translocon is the DA adhesin. In agreement with these results, specific anti-EspA antibodies blocked adherence of the 1551-2 eae mutant. Our data support a role for intimin in LA, for the T3SS translocon in DA, and for T1P in biofilm formation, all of which may act in concert to facilitate host intestinal colonization by aEPEC strains.

Fig 1

Atypical EPEC 1551-2 interactions with HeLa cells. (A, B, and C) Light microscopy images demonstrating the differences in the adherence patterns between 1551-2 (LA), the derivative eae mutant (DA), and the complemented strain (LA), respectively (original magnification, ×1,000). (D) Quantitative adherence assay demonstrating a reduction of approximately 33.7% in adherence of 1551-2 eae::Km compared with the wild-type strain. *, P < 0.05; #, significantly more adherent than 1551-2 eae::Km (P < 0.05). (E and F) SEM images showing the presence or absence of AE lesions (arrows) after 6 h of contact by 1551-2 and 1551-2 eae::Km, respectively. Bar = 2 μm.

Fig 2

Identification and purification of T1P from 1551-2 and analysis of its role in HeLa cell adherence. (A) Micrograph of piliated 1551-2 obtained from the supernatant of infected cells after 6 h of incubation and visualized by TEM. (B) TEM micrograph showing pili purified from 1551-2. (C) Depolymerization of purified pili in a 16% SDS-PAGE gel showing an 18-kDa protein (lane 1) whose mass spectrometry analysis showed identity to the FimA pilin subunit protein. MS, mass standards. (D) Light microscopy micrograph showing the LA pattern exhibited by 1551-2 ΔfimA after 6 h of incubation with HeLa cells (original magnification, ×1,000). (E) Quantification of adhering bacteria showing only minor differences in the level of adherence between the wild type and the T1P mutant (P > 0.05). Bar = 500 nm.

Fig 3

Identification and characterization of the T2SS substrate SslE in supernatants of 1551-2. (A) Heat extract of the 1551-2 strain grown in DMEM (lanes 1, 3, and 5) and LB (lanes 2, 4, and 6) at 37°C with shaking (lanes 1 and 2), at 28°C with shaking (lanes 3 and 4), and at 37°C under a 5% CO₂ atmosphere without shaking (lanes 5 and 6). (B) Secreted protein extract showing the presence of the T2SS SslE in the wild type (lane 1) and absence in the sslE mutant (1551-2 sslE::Km) (lane 2). (C) Light microscopy image showing LA by the sslE mutant after 6 h of incubation with HeLa cells (original magnification, ×1,000). (D) Quantitative comparison of adherence between the wild type and sslE mutant showing no statistical difference (P > 0.05).

Fig 4

Contribution of T3SS to the adherence of aEPEC 1551-2 to HeLa cells. (A and B) Adherence of the T3SS-ATPase-deficient mutant (1551-2 escN::Km) and the complemented strain after 6-h assay (original magnification, ×1,000). (C) Quantitative comparison of adherence between the indicated strains demonstrating that adherence of 1551-2 to HeLa is dependent on a functional T3SS. ***, significantly less adherent than 1551-2 (P < 0.005); ###, significantly more adherent than the 1551-2 escN::Km mutant (P < 0.001).

Fig 5

The DA phenotype is dependent on the T3SS needle complex. (A, B, and C) Light microscopy images showing the adherence patterns observed in 1551-2 eae::Km, the double eae espA mutant, and the complemented strain, respectively (original magnification, ×1,000). (D) Effect of mutations in T3SS-related and fimA genes in the DA phenotype of 1551-2 eae::Km to HeLa cells. While mutations in tir and fimA genes did not affect the adherence ability of the eae mutant, deletion of espA significantly impaired its adherence. ***, significantly less adherent than 1551-2 eae::Km (P < 0.0001); ###, significantly more adherent than 1551-2eae::Km ΔespA (P < 0.005). (E) Quantification of adherence of the 1551-2eae::Km strain in the presence of anti-EspA antibodies. Bacteria were added to HeLa cells in the absence (none) or the presence of three dilutions of the anti-EspA (1:100, 1:50, and 1:10). ***, statistically significant with respect to the values obtained in the absence (none) of the anti-EspA antibody (P < 0.0001).

Fig 6

Biofilm formation by aEPEC 1551-2 and derived mutants. (A) Quantitative determination of crystal violet adsorption upon biofilm formation by the indicated strains. Biofilm formation by 1551-2 is dependent on T1P production but not on T3SS or the T2SS substrate. (B) Pellicle formation at the air-liquid interface upon bacterial growth in glass tubes. This phenotype was dependent on T1P production. (C) Images comparing yeast agglutination by the indicated strains. ***, significantly less adherent than 1551-2 (P < 0.0001); ###, significantly more adherent than 1551-2 ΔfimA (P < 0.005).