Neisseria gonorrhoeae-induced transactivation of EGFR enhances gonococcal invasion.
Journal: 2011/September - Cellular Microbiology
ISSN: 1462-5822
Abstract:
Neisseria gonorrhoeae, the causative agent of the sexually transmitted infection gonorrhoea, adheres to and invades into genital epithelial cells. Here, we investigate host components that are used by the bacteria for their entry into epithelial cells. We found that gonococcal microcolony formation on the surface of HEC-1-B cells disrupted the polarized, basolateral distribution of both epidermal growth factor receptor (EGFR) and ErbB2, a related family member, and induced their accumulation under the microcolonies at the apical membrane. Gonococcal infection increased EGFR and ErbB2 phosphorylation. The EGFR kinase inhibitor, AG1478, reduced gonococcal invasion by 80%, but had no effect on adherence or the recruitment of EGFR and ErbB2 to the microcolonies. Gonococcal inoculation upregulated the mRNA levels of several ligands of EGFR. Prevention of EGFR ligand shedding by blocking matrix metalloproteinase activation reduced gonococcal invasion without altering their adherence, while the addition of the EGFR ligand, HB-EGF, was able to restore invasion to 66% of control levels. These data indicate that N. gonorrhoeae modulates the activity and cellular distribution of host EGFR, facilitating their invasion. EGFR activation does not appear to be due to direct gonococcal binding to EGFR, but instead by its transactivation by gonococcal induced increases in EGFR ligands.
Relations:
Content
Citations
(14)
References
(36)
Chemicals
(5)
Genes
(1)
Organisms
(2)
Processes
(2)
Anatomy
(2)
Affiliates
(1)
Similar articles
Articles by the same authors
Discussion board
Cell Microbiol 13(7): 1078-1090

<em>Neisseria gonorrhoeae</em>-induced transactivation of EGFR enhances gonococcal invasion

Introduction

Neisseria gonorrhoeae is an obligate pathogen of humans and has no other natural host. Infection of genital epithelial cells by N. gonorrhoeae is a multi-step process, consisting of adherence, invasion, intracellular survival and exocytosis. These events are initiated and mediated by multiple interactions of gonococcal surface molecules with genital epithelial cells. The interactions of gonococcal surface molecules with the epithelial cell surface activate signalling cascades in the host cells and trigger the reorganization of the actin cytoskeleton, which is required for the entry of the bacteria into host cells and transmigration across the host epithelium (Grassme et al., 1996). Pilus retraction from adherent gonococci on the epithelial cell surface activates Ca flux (Ayala et al., 2005), the PI3K/ Akt pathway (Lee et al., 2005) and the MAP kinase ERK (Howie et al., 2008). The interaction of opacity protein (Opa) with heparin sulfate proteoglycans (HSPG) activates phosphatidylcholine-specific phospholipase C (PLC) and the acid sphingomyelinase (Grassme et al., 1997). Opa also can trigger integrin-mediated protein kinase C (PKC) activation through binding to the serum-derived extracellular matrix proteins, fibronectin and vitronectin (Dehio et al., 1998). Adherence of N. gonorrhoeae to A431 cells, an epidermoid carcinoma cell line that expresses a high level of epidermal growth factor receptor (EGFR), induces co-clustering of EGFR, CD44v3, ICAM-1 and F-actin under bacterial microcolonies (Merz et al., 1999). EGFR, one of the common surface receptors that are essential for epithelial cell survival and proliferation, can activate signalling cascades, including PI3K, PLCγ, Ca flux, PKC and MAP kinases, all of which can potentially lead to actin rearrangement. However, whether EGFR has a role in gonococcal infection remains to be elucidated.

Bacterial pathogens are known to hijack host cell-signalling pathways and use them to their advantage in order to invade into and survive within host cells. Recent studies have shown that EGFR is a common signalling receptor that is manipulated by pathogens for their survival in their host. Both Pseudomonas aeruginosa and Helicobacter pylori activate EGFR in order to prevent epithelial cell apoptosis (Zhang et al., 2004; Yan et al., 2009). Haemophilus influenzae activation of EGFR negatively regulates TLR2 expression in infected host cells (Mikami et al., 2005). Pasteurella multocida activation of EGFR stimulates proliferation of fibroblasts (Seo et al., 2000). Additionally, activation of ErbB2 (HER2), a related family member of EGFR, has been shown to be crucial for Neisseria meningitidis invasion of endothelial cells (Hoffmann et al., 2001).

EGFR (ErbB1) belongs to the ErbB family of four closely related receptor tyrosine kinases (Yarden and Sliwkowski, 2001; Citri and Yarden, 2006). The ErbB receptors are found in the plasma membrane as inactive dimers that are activated by ligand binding (Tao and Maruyama, 2008). These receptors bind differentially to 13 peptide ligands (Harris et al., 2003; Kochupurakkal et al., 2005). All of the ligands initially are expressed at the plasma membrane as transmembrane proteins. These precursor proteins are shed from the plasma membrane by proteolytic cleavage that is mediated by members of the matrix metalloproteinase (MMP) family and/or ADAM (a disintegrin and metalloproteinase) family (Higashiyama et al., 2008). While ErbB2 does not bind to any ligand, it is able to heterodimerize and form active dimers with the other family members and is their preferred dimerization partner (Tzahar et al., 1996; Graus-Porta et al., 1997). The cytoplasmic tail of each of the ErbBs contains a tyrosine kinase that trans-autophosphorylates the cytoplasmic tail of its dimerization partner. The phosphotyrosines then serve as docking sites for Src homology 2 (SH2) and phosphotyrosine binding domains containing molecules and induce signalling cascades. Depending on the concentration of ligands, the density of the receptor on the cell surface, or the nature of the dimers formed, activation of the ErbB family of receptors results in diverse outcomes, including cell proliferation, survival, migration and/or differentiation.

In this study, we examine the role of ErbB family receptors in gonococcal adherence to and invasion into genital epithelial cells. Our results show that the interaction of gonococci with genital epithelial cells induces both the tyrosine phosphorylation of EGFR and ErbB2 and their recruitment to the sites of gonococcal attachment and microcolony formation. The kinase activity of EGFR is necessary for efficient invasion of gonococci into epithelial cells. Our results further suggest that the activation of EGFR is induced by transactivation via stimulating the gene expression and surface cleavage of EGFR ligands, but not by direct interaction of the gonococci with the receptor.

Results

EGFR and ErbB2 are recruited to the sites of gonococcal adherence in genital epithelial cells

Previous studies have shown that N. gonorrhoeae and N. meningitidis recruit EGFR (Merz et al., 1999) and ErbB2 (Hoffmann et al., 2001), respectively, to the sites of bacterial adherence in epithelial or endothelial cells. To examine how these two receptors are involved in the interaction of gonococci with genital epithelial cells, we compared the cellular distribution of EGFR and ErbB2 in human endometrial epithelial cells (HEC-1-B) and human cervical epithelial cells (ME180) before and after incubation with live or killed N. gonorrhoeae strain MS11 expressing both pili and Opa (Pil Opa). The epithelial cells grown on glass coverslips were inoculated with gonococci for 5 h, followed by fixation and staining for EGFR, ErbB2 and the bacteria without permeabilization of the epithelial cells. In uninfected cells, both EGFR and ErbB2 were found evenly distributed on the surface of HEC-1-B and ME180 cells (Fig. 1A–D and data not shown). After incubation with bacteria, both EGFR and ErbB2 accumulated and surrounded the gonococci at the surfaces of HEC-1-B (Fig. 1E–H) and ME180 cells (Fig. 1Q–T). Gentamicin-killed gonococci adhered to the surface of both cell lines as diplococci without forming microcolonies (Fig. 1I and data not shown). Neither EGFR nor ErbB2 accumulated or localized around the adherent gentamicin-killed gonococci (Fig. 1I–L). This result indicates that N. gonorrhoeae induces the recruitment of EGFR and ErbB2 to the site of bacterial adherence, and that this recruitment requires the viability of gonococci. However, because the gonococci do not colocalize with EGFR and ErbB2 to a high degree (Fig. 1E–H and Q–T), this result does not support a direct interaction of gonococci with these two receptors.

An external file that holds a picture, illustration, etc.
Object name is nihms692425f1.jpg

Live, but not killed, N. gonorrhoeae recruits EGFR and ErbB2 at the epithelial cell surface. HEC-1-B (A–P) or ME180 (Q–T) cells were pretreated with (M–P) or without EGFR inhibitor AG1478 and then incubated with live or gentamicin-killed Pil Opa gonococci for 5 h. Uninoculated HEC-1-B cells (A–D) served as a control. The cells then were fixed and stained with gonococcal antiserum, anti-EGFR mAb, anti-ErbB2 mAb and corresponding secondary antibodies. Optical sections were acquired using a confocal microscope (Zeiss LSM 510). Shown are representative images of single optical sections from three independent experiments. Bar, 5 μm.

EGFR and ErbB2 have been reported to localize predominately at the basolateral surface (Westermark et al., 1986; Kuwada et al., 1998), while gonococci initiate infection at the apical surface of polarized epithelial cells. All previous findings that EGFR and ErbB2 are recruited to the sites of gonococcal adherence are based on unpolarized epithelial cells. In order to investigate if basolaterally located EGFR and ErbB2 are recruited to the adherent gonococci on the apical surface, we polarized HEC-1-B cells by growing them on transwell filters. Polarization of the HEC-1-B cells was monitored by transepithelial resistance (TER) and confirmed by distinct cellular distribution of zonula occludens-1 (ZO-1), a marker protein of the tight junction that divides the apical and basolateral plasma membrane (Fig. 2A and B). The distribution of EGFR and ErbB2 was analysed using confocal microscopy. In uninfected polarized HEC-1-B cells, both EGFR and ErbB2 were found predominately at the basolateral surface (Fig. 2C and E, yellow arrows), but not at the apical surface of polarized HEC-1-B cells (Fig. 2C and E, white arrows), as reported previously in other types of polarized epithelial cells. This result further confirms that these HEC-1-B cells were polarized. After incubation with gonococci, EGFR and ErbB2 were found at both the apical and basolateral surfaces of 90% of gonococcal associated epithelial cells, but only 2–3% of neighbouring cells (Fig. 2D, F and G). Particularly, EGFR and ErbB2 were accumulated under gonococci at the apical surface (Fig. 2D and F, white arrows). Gonococci had no significant effect on the TER of polarized HEC-1-B cells (data not shown). This result indicates that gonococci disrupt the polarized distribution of both EGFR and ErbB2 at the basolateral surface and induce their accumulation at the site of adherent gonococci at the apical surface.

An external file that holds a picture, illustration, etc.
Object name is nihms692425f2.jpg

N. gonorrhoeae recruits EGFR and ErbB2 from the basolateral to the apical surface beneath bacterial microcolonies in polarized epithelial cells.

A. A diagram shows the tight junction (red) dividing the apical (black) and basolateral (green) surfaces.

B–F. HEC-1-B cells were grown on transwells until transepithelial resistance peaked. Polarized HEC-1-B cells were incubated with (D and F) or without Pil Opa gonococci (B, C and E) in the apical chamber for 5 h. Cells then were fixed and stained for gonococci, EGFR (C and D), ErbB2 (E and F), and/or the tight junction marker zonula occluden-1 (ZO-1) (B). Series of images in Z-axes of cells were acquired at 1.0 (B) or 0.5 μm (C-F) per optical section using a Zeiss LSM 510 confocal microscope. Shown are representative images of xy, xz and yz optical sections from three independent experiments. Yellow arrows point to the basolateral staining of EGFR and ErbB2. White arrows point to the apical surfaces and EGFR or ErbB2 staining that concentrates beneath the gonococci adhered to the apical surface of the HEC-1-B cells. Bar, 5 μm. G. Immunofluorescence images of polarized HEC-1-B cells with or without gonococci were quantified to determine the percentage of cells that had EGFR or ErbB2 located at either the basolateral membrane alone or both the apical and basolateral membrane and whether the cells were associated with a gonococcal microcolony. Shown are the mean percentages (SD) from three independent experiments.

Kinase inhibitors of ErbB receptors alter N. gonorrhoeae invasion into epithelial cells

In order to investigate the role of EGFR and ErbB2 in the invasion of N. gonorrhoeae into human epithelial cells, we performed gentamicin protection assays using inhibitors specific for the tyrosine kinases of the two ErbB family receptors. HEC-1-B or ME180 cells were treated with AG1478, an EGFR kinase inhibitor, or AG825, an ErbB2 kinase inhibitor, before their incubation with Pil Opa gonococci. As shown in Figure 3, inhibition of EGFR kinase activity with AG1478 significantly reduced invasion of gonococci into both HEC-1-B (Fig. 3A) and ME180 cells (Fig. 3C). In ME180 cells there was a dose-dependent decrease in the invasion level (80% at 5 μM) (Fig. 3C), whereas the invasion of gonococci into HEC-1-B cells was more sensitive to the inhibitor and showed significant reductions at nanomolar concentrations (Fig. 3A). In contrast, inhibition of ErbB2 kinase activity with AG825 dramatically increased the invasion, up to sixfold, in HEC-1-B cells (Fig. 3B), but had no significant effect on gonococcal invasion into ME180 cells (Fig. 3D). In contrast to their effects on gonococcal invasion, both inhibitors had no significant effect on adherence of gonococci to HEC-1-B and ME180 cells (Fig. 3E and F) and the recruitment of EGFR and ErbB2 to the site of bacterial attachment (Fig. 1M–P and data not shown). Furthermore, there were no detectable effects of the inhibitors on bacterial viability and growth (data not shown). Different sensitivities of gonococcal invasion to the inhibitors in two different cell lines implicate differential expression levels of EGFR and ErbB2 in HEC-1-B and ME180 cells. We found that ME180 cells, which required a much higher concentration of the EGFR inhibitor to reduce gonococcal invasion, expressed a 17-fold higher protein level of EGFR and a fivefold higher protein level of ErbB2 than HEC-1-B cells (Fig. 3G). These results suggest that the kinase activity of EGFR is required for efficient invasion of gonococci into endometrial and cervical epithelial cells, but the kinase activity of ErbB2 has a negative regulatory role in gonococcal invasion. Neither EGFR nor ErbB2 kinase activity is essential for gonococcal adherence to epithelial cells.

An external file that holds a picture, illustration, etc.
Object name is nihms692425f3.jpg

ErbB kinase inhibitors alter gonococcal invasion.

A–D. HEC-1-B (A and B) or ME180 (C and D) were preincubated with either the EGFR kinase inhibitor AG1478 (A and C) or the ErbB2 kinase inhibitor AG825 (B and D) before the addition of Pil Opa gonococci and incubated for 6 h. The number of gentamicin resistant bacteria was determined.

E and F. To quantify the number of adherent bacteria, HEC-1-B (E) and ME180 (F) cells were pretreated with the inhibitors (5 μM) and incubated with gonococci for 2 h. Cells were washed and lysed to quantify the bacteria. The data are plotted as percentages of the gonococci invaded into (A–D) or adhered to (E and F) untreated cells. Shown are the mean percentages (±SD) from three independent experiments with six replicates per experiment. *P < 0.05 (as compared with no inhibitor).

G. Equal amounts of lysates generated from uninoculated ME180 and HEC-1-B cells were analysed by SDS-PAGE and Western blot, probing for EGFR and ErbB2. The blots were stripped and probed for β-tubulin as normalization controls. The blots were quantified by densitometry.

Shown are representative blots and ratios of EGFR and ErbB2 expression between ME180 (M) and HEC-1-B cells (H).

N. gonorrhoeae induces the phosphorylation of EGFR and ErbB2

The effect of EGFR and ErbB2 kinase inhibitors on the invasion of gonococci into epithelial cells suggests the involvement of these receptors in the gonococcal invasion process. Therefore, we examined whether gonococci induce activation of the two ErbB family receptors. The activation of EGFR and ErbB2 was monitored by their tyrosine phosphorylation. Lysates were prepared from HEC-1-B cells that had been incubated with N. gonorrhoeae for up to 6 h and were subjected to immunoprecipitation with a monoclonal antibody (mAb) specific for phosphotyrosine. Phosphorylated EGFR and ErbB2 in the immunoprecipitates were analysed using Western blot. As shown in Figure 4, gonococci induced the tyrosine phosphorylation of both EGFR and ErbB2. The increase in the phosphorylated EGFR peaked at 4 h post inoculation, reaching a 3.5-fold increase over uninfected levels, and remained elevated at 6 h post inoculation. The phosphorylation levels of ErbB2 followed the same temporal pattern as those of EGFR, but its increase was more subtle. Phosphorylation levels of ErbB2 reached their maximum level at 4 h, but returned to near control levels by 6 h post inoculation (Fig. 4).

An external file that holds a picture, illustration, etc.
Object name is nihms692425f4.jpg

N. gonorrhoeae induces the tyrosine phosphorylation of EGFR and ErbB2. HEC-1-B cells were incubated with Pil Opa gonococci for up to 6 h. The cells were lysed and subjected to immunoprecipitation with anti-phosphotyrosine mAb 4G10. The cell lysates were analysed by SDS-PAGE and Western blot, probing for EGFR. ErbB2 was probed after stripping. β-tubulin in cell lysates was analysed as normalization controls. The blots were quantified by densitometry, and the data are plotted as percentages of uninoculated epithelial cell controls. Shown are representative blots (A) from three independent experiments and the densitometric percentage from the blot shown (B).

Blocking EGFR ligand binding by anti-EGFR mAb inhibits N. gonorrhoeae invasion, but not their adherence to epithelial cells

Our findings that gonococci induce the activation of EGFR and ErbB2 tyrosine kinases and that the inhibitor of EGFR kinase reduces gonococcal invasion indicate that the activation of EGFR kinase is important for gonococcal invasion of epithelial cells. We hypothesized that gonococci either activate these receptors by binding directly to EGFR and/or ErbB2, or by increasing the levels of the receptor’s ligands. In order to investigate these possibilities, we preincubated ME180 cells with a mAb specific for the ligand binding domain of EGFR, which blocks the binding of ligands to EGFR (Sato et al., 1983). The effect of this incubation on gonococcal adherence and invasion was determined. The treatment of anti-EGFR mAb had no influence on the ability of gonococci to adhere to ME180 cells (Fig. 5A). Microscopic studies revealed no differences in the size or number of gonococcal clusters formed on the surface of ME180 cells that were treated with anti-EGFR mAb or a control mAb (Fig. 5C). Anti-EGFR mAb at a concentration of 10 μg ml, however, inhibited 50% of the gonococcal invasion as compared with a control mAb (Fig. 5B). These results suggest that the binding of ligand(s) to EGFR is involved in gonococcal invasion, but not in gonococcal adherence.

An external file that holds a picture, illustration, etc.
Object name is nihms692425f5.jpg

Interference with EGFR ligand binding reduces gonococcal invasion.

A and B. ME180 cells were pretreated with either 10 μg ml anti-EGFR mAb that prevents EGFR ligands from binding to EGFR or an isotype control mAb and incubated with Pil Opa gonococci for 6 h. The number of epithelial cell-associated bacteria (A) or gentamicin-resistant invaded bacteria (B) was determined. The results are plotted as a percent of the bacteria adhered to and invaded into cells treated with the control mAb. Shown are mean values (SD) from three independent experiments with replicates of six per experiment. *P < 0.05 (as compared with untreated).

C. ME180 cells that were incubated with anti-EGFR mAb and the bacteria as described above were fixed and stained with DAPI for visualization of nucleic acids and anti-ErbB2 mAb. Images were acquired using a fluorescence microscope. Arrows point to gonococcal clusters. Bar, 5 μm.

N. gonorrhoeae induces the upregulation of a subset of EGFR ligand transcripts

To test whether gonococci can transactivate EGFR and ErbB2 by increasing the expression of the ligands, we quantified mRNA levels of all six ligands that EGFR binds in HEC-1-B cells using real-time PCR. The transcription levels were normalized against the mRNA level of actin. The mRNA levels for heparin binding epidermal growth factor-like growth factor (HB-EGF) and amphiregulin dramatically increased after gonococcal inoculation (Fig. 6). HB-EGF mRNA transcripts had the largest increase, reaching 30-fold that of the levels in uninoculated cells (Fig. 6). Amphiregulin transcripts increased 4.5-fold as compared with uninoculated control cells. TGF-α transcripts steadily increased over time and had doubled that of uninoculated cells by 8 h. The remaining three ligands, EGF, epiregulin and betacellulin were either downregulated or only marginally increased after inoculation, in comparison with the levels in uninoculated cells. This result indicates that gonococci induce the upregulation of a subset of EGFR ligand transcripts.

An external file that holds a picture, illustration, etc.
Object name is nihms692425f6.jpg

N. gonorrhoeae increases a subset of EGFR ligand transcripts. HEC-1-B cells were incubated with Pil Opa gonococci at an MOI of 5 for up to 8 h. Total RNA was extracted and reverse transcribed. The mRNA levels of six EGFR ligands were quantified by real-time PCR. β-actin was used as an internal control for normalization. Shown are representative results from two independent experiments.

Inhibition of EGFR ligand cleavage inhibits N. gonorrhoeae invasion without altering their adherence

The ligands for EGFR are expressed initially as transmembrane precursors and are shed from the plasma membrane after proteolytic cleavage by members of the MMP and ADAM families when needed (Higashiyama et al., 2008). In order to investigate whether the cleavage of EGFR ligands by MMPs is important for gonococcal invasion, we used two methods to prevent the activation of MMPs. First, we used the specific MMP inhibitor, batimastat, and examined its effects on gonococcal adherence and invasion. Batimastat was able to significantly inhibit invasion of gonococci into both HEC-1-B cells and ME180 cells (Fig. 7A and B). Batimastat had no effect on the adherence of gonococci to either cell line (Fig. 7A and B). Our second approach of preventing MMP activation was by heparin washes. Many MMPs, including MMP-1, -2, -7, -9 and -13, bind to the heparan sulfate moieties that decorate cell surface HSPG and/or are in the extracellular matrix (Yu et al., 2002). MMP-7 can be removed from rat uterus tissue by washing with heparin. It is presumed that other heparan sulfate bound MMPs also can be removed in this manner (Yu et al., 2002), suggesting that heparin washes can be used to deplete heparan sulfate bound MMPs at the cell surface in order to prevent the cleavage of HB-EGF and other ErbB ligands. Previous studies have shown that including heparin during the incubation of gonococci with epithelial cells inhibits gonococcal adherence, because of inhibition of Opa binding to HSPG on the surface of epithelial cells (Chen et al., 1995). We have made a similar observation (data not shown). In order to determine the effect of the heparin washes on gonococcal invasion independent of adherence, free heparin was removed by extensive washes with serum-free media after the heparin washes and before adherence and invasion analyses. Heparin washes followed by the removal of free heparin had no significant effect on the adherence of gonococci to either HEC-1-B or ME180 cells (Fig. 7C and D). The heparin washes, however, inhibited the invasion of gonococci into both HEC-1-B and ME180 cells by greater than 75% (Fig. 7C and D). This result suggests that the cleavage of EGFR ligands is important for gonococcal invasion into epithelial cells.

An external file that holds a picture, illustration, etc.
Object name is nihms692425f7.jpg

Inhibition of EGFR ligand cleavage from epithelial cells by MMP inhibition reduces gonococcal invasion.

A–D. HEC-1-B (A) or ME180 (B) cells were preincubated with the MMP inhibitor batimastat at the indicated concentrations before the addition of Pil Opa gonococci. HEC-1-B (C) and ME180 (D) cells were washed with serum-free media alone or with 5 mg ml heparin followed by the serum-free media. Epithelial cells were incubated with Pil Opa gonococci for 6 h. The number of epithelial cell-associated (adherence) and gentamicin-resistant (invasion) bacteria was determined. The results are plotted as a percent of the untreated (A–B) or media washed controls (C–D). Shown are the mean values (±SD) generated from three independent experiments with replicates of six per experiment.

E and F. Lysates were prepared from media washed or heparin washed HEC-1-B cells that had been incubated with Pil Opa gonococci for up to 6 h. The lysates were subjected to Western blot, probing for HB-EGF. The blots were stripped and reblotted for β-tubulin, as a loading control. The blots were quantified by densitometry. The data were plotted as a percentage of the maximal amount of sHB-EGF in the cells exposed to the bacteria. Shown are representative blots (E) and the mean values (SD) of three independent experiments (F).

G. Heparin washed or media washed HEC-1-B cells were incubated with Pil Opa gonococci for 3.5 h and with sHB-EGF at the indicated concentrations for an additional 1.5 h. Gentamicin-resistant bacteria were determined. *P < 0.05 (as compared with media washed).

To confirm that the heparin washes removed heparin sulfate-associated MMPs, consequently preventing the shedding of EGFR ligands, we determined the levels of cleaved, soluble HB-EGF (sHB-EGF). HEC-1-B cells that were subjected to either the heparin wash or medium wash were incubated with gonococci for varying lengths of time. The HEC-1-B/gonococci co-culture media was analysed for sHB-EGF by ELISA. There was no detectable sHB-EGF in the co-culture media from HEC-1-B cells that were subjected to either the medium wash or the heparin wash (data not shown). Because the sHB-EGF is often found associated with heparin sulfate moieties on the cell surface but not found in the supernatant (Xu et al., 2004), we looked for cell-associated sHB-EGF. HEC-1-B cells that were subjected to either the heparin wash or medium wash were incubated with gonococci for varying lengths of time and lysed. The cell lysates were analysed using non-reducing SDS-PAGE, and sHB-EGF was detected by Western blotting using a biotinylated anti-sHB-EGF antibody. In the lysates generated from cells subjected to the medium wash, the presence of gonococci dramatically increased the amount of sHB-EGF. The levels of sHB-EGF increased with time and peaked at 4 h (Fig. 7E and F). This result is consistent with our finding that gonococcal interaction with host cells increases the levels of HB-EGF transcripts (Fig. 6). Importantly, the heparin wash significantly reduced gonococci-induced production of sHB-EGF (Fig. 7E and F). This finding further supports our hypothesis that gonococci activate EGFR by inducing the production of a subset of EGFR ligands and the cleavage of these ligands by MMPs.

In order to confirm the involvement of MMP-cleaved EGFR ligands in gonococcal invasion, we investigated whether the addition of exogenous sHB-EGF to HEC-1-B cells that were depleted of MMPs could rescue gonococcal invasion. HEC-1-B cells that had been washed with either heparin or media were inoculated with gonococci for 3.5 h, which allowed the bacteria to form microcolonies on the epithelial cell surface, followed by the addition of sHB-EGF. In media washed HEC-1-B cells, the addition of sHB-EGF caused a dose-dependent decrease in gonococcal invasion. In heparin washed HEC-1-B cells where gonococcal invasion was inhibited greater than 90%, the addition of exogenous sHB-EGF at the lower concentrations induced a dose-dependent increase in invasion and was able to restore the invasion to 66% of control levels (no addition of sHB-EGF) (Fig. 7G). The addition of higher concentrations of sHB-EGF caused a dose-dependent decrease in invasion that was similar to the HB-EGF inhibition seen in the media washed cells. This result suggests that MMP-cleaved HB-EGF can facilitate gonococcal invasion.

The expression of either pili or Opa is necessary for gonococcal induced EGFR activation and EGFR-dependent gonococcal invasion

Pili and Opa are major surface molecules expressed in gonococci that are important for bacterial adherence and invasion. In order to investigate the individual role of pili and Opa in EGFR activation and EGFR-dependent gonococcal invasion, we selected variants and mutants in strain MS11 that expressed pili and Opa together, either surface component individually, or not at all. HEC-1-B cells were incubated with these variants and mutants for 4 h. The EGFR phosphorylation levels were determined by immunoprecipitation and Western blot and the cellular distribution of EGFR by immunofluorescence microscopy. Live gonococci that expressed pili and Opa, either individually or together, were able to increase the phosphorylation level of EGFR to similar levels (Fig. 8A and B). Killed Pil Opa and live Pil Δopa gonococci, however, were not able to increase the phosphorylation level of EGFR. Similar to Pil Opa gonococci, Pil Opa and Pil Δopa gonococci induced the accumulation of EGFR at the sites of gonococcal adherence (Fig. 8D). We next investigated whether AG1478, the EGFR kinase inhibitor, was able to alter the invasion of gonococcal variants into HEC-1-B cells. As shown in Figure 8C, the invasion of gonococci that express either pili or Opa individually was inhibited to a significant and equivalent degree by AG1478 (50%). This result suggests that the live gonococci that express either of the gonococcal surface components, pili or Opa, is sufficient to induce the activation of EGFR and that this activation is important for gonococcal invasion.

An external file that holds a picture, illustration, etc.
Object name is nihms692425f8.jpg

Coexpression of Pili and Opa is not required for induction of EGFR activation.

A and B. HEC-1-B cells were incubated with Pil Opa, Pil ΔOpa, Pil Opa or Pil ΔOpa gonococci for 4 h. Killed Pil Opa gonococci and media (4 h no GC) served as controls. The cells were lysed and subjected to immunoprecipitation with anti-phosphotyrosine mAb. The cell lysates were analysed by SDS-PAGE and Western blot, probing for EGFR. β-tubulin in cell lysates was analysed as normalization controls. The blots were quantified by densitometry, and the data are plotted as percentages of uninoculated epithelial cell controls. Shown are representative blots (A) and the average percentages (±SD) (B) from three independent experiments. *P < 0.01 (as compared with 0 h uninoculated).

C. HEC-1-B cells were preincubated with AG1478 for 2 h before incubation with Pil ΔOpa or Pil Opa gonococci for 6 h. The number of gentamicin-resistant bacteria was determined, and the results are plotted as a percent of the untreated. Shown are the mean values (±SD) generated from three independent experiments with replicates of six per experiment. *P < 0.01 (as compared with untreated for each gonococcal variant or mutant).

D. HEC-1-B cells were incubated with Pil Δopa or Pil Opa gonococci for 5 h. The cells were fixed and stained with gonococcal antiserum, anti-EGFR mAb, anti-ErbB2 mAb and the corresponding secondary antibodies. Images were acquired using a confocal microscope. Shown are representative images of single optical sections from three independent experiments. Bar, 5 μm.

EGFR and ErbB2 are recruited to the sites of gonococcal adherence in genital epithelial cells

Previous studies have shown that N. gonorrhoeae and N. meningitidis recruit EGFR (Merz et al., 1999) and ErbB2 (Hoffmann et al., 2001), respectively, to the sites of bacterial adherence in epithelial or endothelial cells. To examine how these two receptors are involved in the interaction of gonococci with genital epithelial cells, we compared the cellular distribution of EGFR and ErbB2 in human endometrial epithelial cells (HEC-1-B) and human cervical epithelial cells (ME180) before and after incubation with live or killed N. gonorrhoeae strain MS11 expressing both pili and Opa (Pil Opa). The epithelial cells grown on glass coverslips were inoculated with gonococci for 5 h, followed by fixation and staining for EGFR, ErbB2 and the bacteria without permeabilization of the epithelial cells. In uninfected cells, both EGFR and ErbB2 were found evenly distributed on the surface of HEC-1-B and ME180 cells (Fig. 1A–D and data not shown). After incubation with bacteria, both EGFR and ErbB2 accumulated and surrounded the gonococci at the surfaces of HEC-1-B (Fig. 1E–H) and ME180 cells (Fig. 1Q–T). Gentamicin-killed gonococci adhered to the surface of both cell lines as diplococci without forming microcolonies (Fig. 1I and data not shown). Neither EGFR nor ErbB2 accumulated or localized around the adherent gentamicin-killed gonococci (Fig. 1I–L). This result indicates that N. gonorrhoeae induces the recruitment of EGFR and ErbB2 to the site of bacterial adherence, and that this recruitment requires the viability of gonococci. However, because the gonococci do not colocalize with EGFR and ErbB2 to a high degree (Fig. 1E–H and Q–T), this result does not support a direct interaction of gonococci with these two receptors.

An external file that holds a picture, illustration, etc.
Object name is nihms692425f1.jpg

Live, but not killed, N. gonorrhoeae recruits EGFR and ErbB2 at the epithelial cell surface. HEC-1-B (A–P) or ME180 (Q–T) cells were pretreated with (M–P) or without EGFR inhibitor AG1478 and then incubated with live or gentamicin-killed Pil Opa gonococci for 5 h. Uninoculated HEC-1-B cells (A–D) served as a control. The cells then were fixed and stained with gonococcal antiserum, anti-EGFR mAb, anti-ErbB2 mAb and corresponding secondary antibodies. Optical sections were acquired using a confocal microscope (Zeiss LSM 510). Shown are representative images of single optical sections from three independent experiments. Bar, 5 μm.

EGFR and ErbB2 have been reported to localize predominately at the basolateral surface (Westermark et al., 1986; Kuwada et al., 1998), while gonococci initiate infection at the apical surface of polarized epithelial cells. All previous findings that EGFR and ErbB2 are recruited to the sites of gonococcal adherence are based on unpolarized epithelial cells. In order to investigate if basolaterally located EGFR and ErbB2 are recruited to the adherent gonococci on the apical surface, we polarized HEC-1-B cells by growing them on transwell filters. Polarization of the HEC-1-B cells was monitored by transepithelial resistance (TER) and confirmed by distinct cellular distribution of zonula occludens-1 (ZO-1), a marker protein of the tight junction that divides the apical and basolateral plasma membrane (Fig. 2A and B). The distribution of EGFR and ErbB2 was analysed using confocal microscopy. In uninfected polarized HEC-1-B cells, both EGFR and ErbB2 were found predominately at the basolateral surface (Fig. 2C and E, yellow arrows), but not at the apical surface of polarized HEC-1-B cells (Fig. 2C and E, white arrows), as reported previously in other types of polarized epithelial cells. This result further confirms that these HEC-1-B cells were polarized. After incubation with gonococci, EGFR and ErbB2 were found at both the apical and basolateral surfaces of 90% of gonococcal associated epithelial cells, but only 2–3% of neighbouring cells (Fig. 2D, F and G). Particularly, EGFR and ErbB2 were accumulated under gonococci at the apical surface (Fig. 2D and F, white arrows). Gonococci had no significant effect on the TER of polarized HEC-1-B cells (data not shown). This result indicates that gonococci disrupt the polarized distribution of both EGFR and ErbB2 at the basolateral surface and induce their accumulation at the site of adherent gonococci at the apical surface.

An external file that holds a picture, illustration, etc.
Object name is nihms692425f2.jpg

N. gonorrhoeae recruits EGFR and ErbB2 from the basolateral to the apical surface beneath bacterial microcolonies in polarized epithelial cells.

A. A diagram shows the tight junction (red) dividing the apical (black) and basolateral (green) surfaces.

B–F. HEC-1-B cells were grown on transwells until transepithelial resistance peaked. Polarized HEC-1-B cells were incubated with (D and F) or without Pil Opa gonococci (B, C and E) in the apical chamber for 5 h. Cells then were fixed and stained for gonococci, EGFR (C and D), ErbB2 (E and F), and/or the tight junction marker zonula occluden-1 (ZO-1) (B). Series of images in Z-axes of cells were acquired at 1.0 (B) or 0.5 μm (C-F) per optical section using a Zeiss LSM 510 confocal microscope. Shown are representative images of xy, xz and yz optical sections from three independent experiments. Yellow arrows point to the basolateral staining of EGFR and ErbB2. White arrows point to the apical surfaces and EGFR or ErbB2 staining that concentrates beneath the gonococci adhered to the apical surface of the HEC-1-B cells. Bar, 5 μm. G. Immunofluorescence images of polarized HEC-1-B cells with or without gonococci were quantified to determine the percentage of cells that had EGFR or ErbB2 located at either the basolateral membrane alone or both the apical and basolateral membrane and whether the cells were associated with a gonococcal microcolony. Shown are the mean percentages (SD) from three independent experiments.

Kinase inhibitors of ErbB receptors alter N. gonorrhoeae invasion into epithelial cells

In order to investigate the role of EGFR and ErbB2 in the invasion of N. gonorrhoeae into human epithelial cells, we performed gentamicin protection assays using inhibitors specific for the tyrosine kinases of the two ErbB family receptors. HEC-1-B or ME180 cells were treated with AG1478, an EGFR kinase inhibitor, or AG825, an ErbB2 kinase inhibitor, before their incubation with Pil Opa gonococci. As shown in Figure 3, inhibition of EGFR kinase activity with AG1478 significantly reduced invasion of gonococci into both HEC-1-B (Fig. 3A) and ME180 cells (Fig. 3C). In ME180 cells there was a dose-dependent decrease in the invasion level (80% at 5 μM) (Fig. 3C), whereas the invasion of gonococci into HEC-1-B cells was more sensitive to the inhibitor and showed significant reductions at nanomolar concentrations (Fig. 3A). In contrast, inhibition of ErbB2 kinase activity with AG825 dramatically increased the invasion, up to sixfold, in HEC-1-B cells (Fig. 3B), but had no significant effect on gonococcal invasion into ME180 cells (Fig. 3D). In contrast to their effects on gonococcal invasion, both inhibitors had no significant effect on adherence of gonococci to HEC-1-B and ME180 cells (Fig. 3E and F) and the recruitment of EGFR and ErbB2 to the site of bacterial attachment (Fig. 1M–P and data not shown). Furthermore, there were no detectable effects of the inhibitors on bacterial viability and growth (data not shown). Different sensitivities of gonococcal invasion to the inhibitors in two different cell lines implicate differential expression levels of EGFR and ErbB2 in HEC-1-B and ME180 cells. We found that ME180 cells, which required a much higher concentration of the EGFR inhibitor to reduce gonococcal invasion, expressed a 17-fold higher protein level of EGFR and a fivefold higher protein level of ErbB2 than HEC-1-B cells (Fig. 3G). These results suggest that the kinase activity of EGFR is required for efficient invasion of gonococci into endometrial and cervical epithelial cells, but the kinase activity of ErbB2 has a negative regulatory role in gonococcal invasion. Neither EGFR nor ErbB2 kinase activity is essential for gonococcal adherence to epithelial cells.

An external file that holds a picture, illustration, etc.
Object name is nihms692425f3.jpg

ErbB kinase inhibitors alter gonococcal invasion.

A–D. HEC-1-B (A and B) or ME180 (C and D) were preincubated with either the EGFR kinase inhibitor AG1478 (A and C) or the ErbB2 kinase inhibitor AG825 (B and D) before the addition of Pil Opa gonococci and incubated for 6 h. The number of gentamicin resistant bacteria was determined.

E and F. To quantify the number of adherent bacteria, HEC-1-B (E) and ME180 (F) cells were pretreated with the inhibitors (5 μM) and incubated with gonococci for 2 h. Cells were washed and lysed to quantify the bacteria. The data are plotted as percentages of the gonococci invaded into (A–D) or adhered to (E and F) untreated cells. Shown are the mean percentages (±SD) from three independent experiments with six replicates per experiment. *P < 0.05 (as compared with no inhibitor).

G. Equal amounts of lysates generated from uninoculated ME180 and HEC-1-B cells were analysed by SDS-PAGE and Western blot, probing for EGFR and ErbB2. The blots were stripped and probed for β-tubulin as normalization controls. The blots were quantified by densitometry.

Shown are representative blots and ratios of EGFR and ErbB2 expression between ME180 (M) and HEC-1-B cells (H).

N. gonorrhoeae induces the phosphorylation of EGFR and ErbB2

The effect of EGFR and ErbB2 kinase inhibitors on the invasion of gonococci into epithelial cells suggests the involvement of these receptors in the gonococcal invasion process. Therefore, we examined whether gonococci induce activation of the two ErbB family receptors. The activation of EGFR and ErbB2 was monitored by their tyrosine phosphorylation. Lysates were prepared from HEC-1-B cells that had been incubated with N. gonorrhoeae for up to 6 h and were subjected to immunoprecipitation with a monoclonal antibody (mAb) specific for phosphotyrosine. Phosphorylated EGFR and ErbB2 in the immunoprecipitates were analysed using Western blot. As shown in Figure 4, gonococci induced the tyrosine phosphorylation of both EGFR and ErbB2. The increase in the phosphorylated EGFR peaked at 4 h post inoculation, reaching a 3.5-fold increase over uninfected levels, and remained elevated at 6 h post inoculation. The phosphorylation levels of ErbB2 followed the same temporal pattern as those of EGFR, but its increase was more subtle. Phosphorylation levels of ErbB2 reached their maximum level at 4 h, but returned to near control levels by 6 h post inoculation (Fig. 4).

An external file that holds a picture, illustration, etc.
Object name is nihms692425f4.jpg

N. gonorrhoeae induces the tyrosine phosphorylation of EGFR and ErbB2. HEC-1-B cells were incubated with Pil Opa gonococci for up to 6 h. The cells were lysed and subjected to immunoprecipitation with anti-phosphotyrosine mAb 4G10. The cell lysates were analysed by SDS-PAGE and Western blot, probing for EGFR. ErbB2 was probed after stripping. β-tubulin in cell lysates was analysed as normalization controls. The blots were quantified by densitometry, and the data are plotted as percentages of uninoculated epithelial cell controls. Shown are representative blots (A) from three independent experiments and the densitometric percentage from the blot shown (B).

Blocking EGFR ligand binding by anti-EGFR mAb inhibits N. gonorrhoeae invasion, but not their adherence to epithelial cells

Our findings that gonococci induce the activation of EGFR and ErbB2 tyrosine kinases and that the inhibitor of EGFR kinase reduces gonococcal invasion indicate that the activation of EGFR kinase is important for gonococcal invasion of epithelial cells. We hypothesized that gonococci either activate these receptors by binding directly to EGFR and/or ErbB2, or by increasing the levels of the receptor’s ligands. In order to investigate these possibilities, we preincubated ME180 cells with a mAb specific for the ligand binding domain of EGFR, which blocks the binding of ligands to EGFR (Sato et al., 1983). The effect of this incubation on gonococcal adherence and invasion was determined. The treatment of anti-EGFR mAb had no influence on the ability of gonococci to adhere to ME180 cells (Fig. 5A). Microscopic studies revealed no differences in the size or number of gonococcal clusters formed on the surface of ME180 cells that were treated with anti-EGFR mAb or a control mAb (Fig. 5C). Anti-EGFR mAb at a concentration of 10 μg ml, however, inhibited 50% of the gonococcal invasion as compared with a control mAb (Fig. 5B). These results suggest that the binding of ligand(s) to EGFR is involved in gonococcal invasion, but not in gonococcal adherence.

An external file that holds a picture, illustration, etc.
Object name is nihms692425f5.jpg

Interference with EGFR ligand binding reduces gonococcal invasion.

A and B. ME180 cells were pretreated with either 10 μg ml anti-EGFR mAb that prevents EGFR ligands from binding to EGFR or an isotype control mAb and incubated with Pil Opa gonococci for 6 h. The number of epithelial cell-associated bacteria (A) or gentamicin-resistant invaded bacteria (B) was determined. The results are plotted as a percent of the bacteria adhered to and invaded into cells treated with the control mAb. Shown are mean values (SD) from three independent experiments with replicates of six per experiment. *P < 0.05 (as compared with untreated).

C. ME180 cells that were incubated with anti-EGFR mAb and the bacteria as described above were fixed and stained with DAPI for visualization of nucleic acids and anti-ErbB2 mAb. Images were acquired using a fluorescence microscope. Arrows point to gonococcal clusters. Bar, 5 μm.

N. gonorrhoeae induces the upregulation of a subset of EGFR ligand transcripts

To test whether gonococci can transactivate EGFR and ErbB2 by increasing the expression of the ligands, we quantified mRNA levels of all six ligands that EGFR binds in HEC-1-B cells using real-time PCR. The transcription levels were normalized against the mRNA level of actin. The mRNA levels for heparin binding epidermal growth factor-like growth factor (HB-EGF) and amphiregulin dramatically increased after gonococcal inoculation (Fig. 6). HB-EGF mRNA transcripts had the largest increase, reaching 30-fold that of the levels in uninoculated cells (Fig. 6). Amphiregulin transcripts increased 4.5-fold as compared with uninoculated control cells. TGF-α transcripts steadily increased over time and had doubled that of uninoculated cells by 8 h. The remaining three ligands, EGF, epiregulin and betacellulin were either downregulated or only marginally increased after inoculation, in comparison with the levels in uninoculated cells. This result indicates that gonococci induce the upregulation of a subset of EGFR ligand transcripts.

An external file that holds a picture, illustration, etc.
Object name is nihms692425f6.jpg

N. gonorrhoeae increases a subset of EGFR ligand transcripts. HEC-1-B cells were incubated with Pil Opa gonococci at an MOI of 5 for up to 8 h. Total RNA was extracted and reverse transcribed. The mRNA levels of six EGFR ligands were quantified by real-time PCR. β-actin was used as an internal control for normalization. Shown are representative results from two independent experiments.

Inhibition of EGFR ligand cleavage inhibits N. gonorrhoeae invasion without altering their adherence

The ligands for EGFR are expressed initially as transmembrane precursors and are shed from the plasma membrane after proteolytic cleavage by members of the MMP and ADAM families when needed (Higashiyama et al., 2008). In order to investigate whether the cleavage of EGFR ligands by MMPs is important for gonococcal invasion, we used two methods to prevent the activation of MMPs. First, we used the specific MMP inhibitor, batimastat, and examined its effects on gonococcal adherence and invasion. Batimastat was able to significantly inhibit invasion of gonococci into both HEC-1-B cells and ME180 cells (Fig. 7A and B). Batimastat had no effect on the adherence of gonococci to either cell line (Fig. 7A and B). Our second approach of preventing MMP activation was by heparin washes. Many MMPs, including MMP-1, -2, -7, -9 and -13, bind to the heparan sulfate moieties that decorate cell surface HSPG and/or are in the extracellular matrix (Yu et al., 2002). MMP-7 can be removed from rat uterus tissue by washing with heparin. It is presumed that other heparan sulfate bound MMPs also can be removed in this manner (Yu et al., 2002), suggesting that heparin washes can be used to deplete heparan sulfate bound MMPs at the cell surface in order to prevent the cleavage of HB-EGF and other ErbB ligands. Previous studies have shown that including heparin during the incubation of gonococci with epithelial cells inhibits gonococcal adherence, because of inhibition of Opa binding to HSPG on the surface of epithelial cells (Chen et al., 1995). We have made a similar observation (data not shown). In order to determine the effect of the heparin washes on gonococcal invasion independent of adherence, free heparin was removed by extensive washes with serum-free media after the heparin washes and before adherence and invasion analyses. Heparin washes followed by the removal of free heparin had no significant effect on the adherence of gonococci to either HEC-1-B or ME180 cells (Fig. 7C and D). The heparin washes, however, inhibited the invasion of gonococci into both HEC-1-B and ME180 cells by greater than 75% (Fig. 7C and D). This result suggests that the cleavage of EGFR ligands is important for gonococcal invasion into epithelial cells.

An external file that holds a picture, illustration, etc.
Object name is nihms692425f7.jpg

Inhibition of EGFR ligand cleavage from epithelial cells by MMP inhibition reduces gonococcal invasion.

A–D. HEC-1-B (A) or ME180 (B) cells were preincubated with the MMP inhibitor batimastat at the indicated concentrations before the addition of Pil Opa gonococci. HEC-1-B (C) and ME180 (D) cells were washed with serum-free media alone or with 5 mg ml heparin followed by the serum-free media. Epithelial cells were incubated with Pil Opa gonococci for 6 h. The number of epithelial cell-associated (adherence) and gentamicin-resistant (invasion) bacteria was determined. The results are plotted as a percent of the untreated (A–B) or media washed controls (C–D). Shown are the mean values (±SD) generated from three independent experiments with replicates of six per experiment.

E and F. Lysates were prepared from media washed or heparin washed HEC-1-B cells that had been incubated with Pil Opa gonococci for up to 6 h. The lysates were subjected to Western blot, probing for HB-EGF. The blots were stripped and reblotted for β-tubulin, as a loading control. The blots were quantified by densitometry. The data were plotted as a percentage of the maximal amount of sHB-EGF in the cells exposed to the bacteria. Shown are representative blots (E) and the mean values (SD) of three independent experiments (F).

G. Heparin washed or media washed HEC-1-B cells were incubated with Pil Opa gonococci for 3.5 h and with sHB-EGF at the indicated concentrations for an additional 1.5 h. Gentamicin-resistant bacteria were determined. *P < 0.05 (as compared with media washed).

To confirm that the heparin washes removed heparin sulfate-associated MMPs, consequently preventing the shedding of EGFR ligands, we determined the levels of cleaved, soluble HB-EGF (sHB-EGF). HEC-1-B cells that were subjected to either the heparin wash or medium wash were incubated with gonococci for varying lengths of time. The HEC-1-B/gonococci co-culture media was analysed for sHB-EGF by ELISA. There was no detectable sHB-EGF in the co-culture media from HEC-1-B cells that were subjected to either the medium wash or the heparin wash (data not shown). Because the sHB-EGF is often found associated with heparin sulfate moieties on the cell surface but not found in the supernatant (Xu et al., 2004), we looked for cell-associated sHB-EGF. HEC-1-B cells that were subjected to either the heparin wash or medium wash were incubated with gonococci for varying lengths of time and lysed. The cell lysates were analysed using non-reducing SDS-PAGE, and sHB-EGF was detected by Western blotting using a biotinylated anti-sHB-EGF antibody. In the lysates generated from cells subjected to the medium wash, the presence of gonococci dramatically increased the amount of sHB-EGF. The levels of sHB-EGF increased with time and peaked at 4 h (Fig. 7E and F). This result is consistent with our finding that gonococcal interaction with host cells increases the levels of HB-EGF transcripts (Fig. 6). Importantly, the heparin wash significantly reduced gonococci-induced production of sHB-EGF (Fig. 7E and F). This finding further supports our hypothesis that gonococci activate EGFR by inducing the production of a subset of EGFR ligands and the cleavage of these ligands by MMPs.

In order to confirm the involvement of MMP-cleaved EGFR ligands in gonococcal invasion, we investigated whether the addition of exogenous sHB-EGF to HEC-1-B cells that were depleted of MMPs could rescue gonococcal invasion. HEC-1-B cells that had been washed with either heparin or media were inoculated with gonococci for 3.5 h, which allowed the bacteria to form microcolonies on the epithelial cell surface, followed by the addition of sHB-EGF. In media washed HEC-1-B cells, the addition of sHB-EGF caused a dose-dependent decrease in gonococcal invasion. In heparin washed HEC-1-B cells where gonococcal invasion was inhibited greater than 90%, the addition of exogenous sHB-EGF at the lower concentrations induced a dose-dependent increase in invasion and was able to restore the invasion to 66% of control levels (no addition of sHB-EGF) (Fig. 7G). The addition of higher concentrations of sHB-EGF caused a dose-dependent decrease in invasion that was similar to the HB-EGF inhibition seen in the media washed cells. This result suggests that MMP-cleaved HB-EGF can facilitate gonococcal invasion.

The expression of either pili or Opa is necessary for gonococcal induced EGFR activation and EGFR-dependent gonococcal invasion

Pili and Opa are major surface molecules expressed in gonococci that are important for bacterial adherence and invasion. In order to investigate the individual role of pili and Opa in EGFR activation and EGFR-dependent gonococcal invasion, we selected variants and mutants in strain MS11 that expressed pili and Opa together, either surface component individually, or not at all. HEC-1-B cells were incubated with these variants and mutants for 4 h. The EGFR phosphorylation levels were determined by immunoprecipitation and Western blot and the cellular distribution of EGFR by immunofluorescence microscopy. Live gonococci that expressed pili and Opa, either individually or together, were able to increase the phosphorylation level of EGFR to similar levels (Fig. 8A and B). Killed Pil Opa and live Pil Δopa gonococci, however, were not able to increase the phosphorylation level of EGFR. Similar to Pil Opa gonococci, Pil Opa and Pil Δopa gonococci induced the accumulation of EGFR at the sites of gonococcal adherence (Fig. 8D). We next investigated whether AG1478, the EGFR kinase inhibitor, was able to alter the invasion of gonococcal variants into HEC-1-B cells. As shown in Figure 8C, the invasion of gonococci that express either pili or Opa individually was inhibited to a significant and equivalent degree by AG1478 (50%). This result suggests that the live gonococci that express either of the gonococcal surface components, pili or Opa, is sufficient to induce the activation of EGFR and that this activation is important for gonococcal invasion.

An external file that holds a picture, illustration, etc.
Object name is nihms692425f8.jpg

Coexpression of Pili and Opa is not required for induction of EGFR activation.

A and B. HEC-1-B cells were incubated with Pil Opa, Pil ΔOpa, Pil Opa or Pil ΔOpa gonococci for 4 h. Killed Pil Opa gonococci and media (4 h no GC) served as controls. The cells were lysed and subjected to immunoprecipitation with anti-phosphotyrosine mAb. The cell lysates were analysed by SDS-PAGE and Western blot, probing for EGFR. β-tubulin in cell lysates was analysed as normalization controls. The blots were quantified by densitometry, and the data are plotted as percentages of uninoculated epithelial cell controls. Shown are representative blots (A) and the average percentages (±SD) (B) from three independent experiments. *P < 0.01 (as compared with 0 h uninoculated).

C. HEC-1-B cells were preincubated with AG1478 for 2 h before incubation with Pil ΔOpa or Pil Opa gonococci for 6 h. The number of gentamicin-resistant bacteria was determined, and the results are plotted as a percent of the untreated. Shown are the mean values (±SD) generated from three independent experiments with replicates of six per experiment. *P < 0.01 (as compared with untreated for each gonococcal variant or mutant).

D. HEC-1-B cells were incubated with Pil Δopa or Pil Opa gonococci for 5 h. The cells were fixed and stained with gonococcal antiserum, anti-EGFR mAb, anti-ErbB2 mAb and the corresponding secondary antibodies. Images were acquired using a confocal microscope. Shown are representative images of single optical sections from three independent experiments. Bar, 5 μm.

Discussion

In this study we investigated the role of EGFR, a common signalling receptor on the epithelial cell surface, in the invasion of N. gonorrhoeae into genital epithelial cells. Our results demonstrate that N. gonorrhoeae induces the activation of EGFR and that this activation is required for efficient invasion of gonococci into epithelial cells. Our results further demonstrate that gonococci activate EGFR by increasing the shedding of EGFR ligands, rather than interacting directly with EGFR. This work uncovers a mechanism by which N. gonorrhoeae activates EGFR for their invasion.

Gonococci-induced EGFR activation is demonstrated by the tyrosine phosphorylation of EGFR and its dimerization partner ErbB2 and their recruitment to the site of gonococcal adherence. The recruitment of EGFR to gonococcal microcolonies has been observed previously (Merz et al., 1999). However, all earlier studies have used unpolarized epithelial cells, and therefore it was unclear whether such recruitment would occur in vivo, where genital epithelial cells are polarized with distinct apical and basolateral surfaces. N. gonorrhoeae establishes adherence at the apical surface of polarized epithelial cells, while EGFR predominately is expressed at the basolateral surface (Westermark et al., 1986; Kuwada et al., 1998). Here we show that EGFR and ErbB2, which are preferentially expressed on the basolateral surface of polarized HEC-1-B cells, lost their polarized distribution and accumulated beneath the sites of gonococcal adherence at the apical surface. This observation supports the notion that N. gonorrhoeae can recruit EGFR and ErbB2 in both polarized and unpolarized epithelial cells.

In addition, we found that the recruitment of both EGFR and ErbB2 is dependent upon the viability of gonococci, as gentamicin-killed gonococci are unable to induce the recruitment. The mechanistic reason for the inability of gentamicin-killed gonococci to recruit EGFR and ErbB2 is unclear. A major difference between live and killed gonococci is that while the killed bacteria are able to adhere to epithelial cells; they are unable to form microcolonies on or invade into epithelial cells (Bish et al., 2008). The formation of microcolonies may be essential to induce the redistribution of EGFR and ErbB2 simply by the physical impact of the microcolony on the epithelial cell surface, which could induce intracellular signalling. Nonpiliated gonococci that have had all of their opa genes deleted do not form the typical microcolony; instead they attach to the surface of HEC-1-B and ME180 cells primarily as individual or small clusters of diplococci, appearing similar to killed gonococci that express both pili and Opa (data not shown). In the absence of both pili and Opa, gonococci invade about two logs less well than gonococci expressing either structure individually (Swanson et al., 2001). This finding lends credence to the idea that microcolony formation is important for the invasion of gonococci into epithelial cells. A second possibility is killed gonococci are unable to synthesize new proteins and surface molecules, even though initially expressing the same surface structures as the live gonococci. The requirement of newly synthesized proteins and other molecules for gonococcal invasion has been suggested by previous observations that inhibition of bacterial protein synthesis with chloramphenicol inhibits gonococcal invasion (Griffiss et al., 1999), and that preincubation of the bacteria with fixed HEC-1-B cells, which potentially alters the expression of bacterial molecules, increases gonococcal invasion ability (Chen et al., 1991). The fact that gonococci that express either Opa or pili activate EGFR and their invasion is enhanced by EGFR activation suggests that there may be additional bacterial molecules involved in invasion. For example, lipooligosaccharide have been shown to be important for gonococcal invasion and also has the ability to phase vary its surface OS structure. (Song et al., 2000).

Our finding that the EGFR kinase inhibitor significantly reduced gonococcal invasion indicates an important role of EGFR activation in gonococcal invasion. Our data further show that while gonococci induce the activation of both EGFR and ErbB2, blocking EGFR and ErbB2 kinases have different effects on gonococcal invasion. Prevention of EGFR kinase activation inhibits gonococcal invasion into epithelial cells. However, prevention of ErbB2 kinase activity either has no effect (ME180 cells) or significantly increases (HEC-1-B) the invasive ability of gonococci.

A major difference between ME180 and HEC-1-B cells is their expression levels of EGFR and ErbB2. ME180 cells express 17-fold more EGFR and fivefold more ErbB2 than HEC-1-B cells. This difference provides an explanation for the different sensitivities of HEC-1-B and ME180 cells to the EGFR and ErbB2 kinase inhibitors. In addition, the differences in EGFR and ErbB2 expression levels would change the molecular ratios of EGFR to ErbB2, consequently altering the nature of ErbB dimers formed on the surface of the two cell lines in response to ligand binding. HEC-1-B cells have a higher ErbB2:EGFR ratio, thus are expected to generate more ErbB2 : EGFR heterodimers than ME180 cells. ME180 cells on the other hand have a higher EGFR : ErbB2 ratio, and are expected to generate more EGFR : EGFR homodimers than HEC-1-B cells. The two different dimers formed by ligand binding, EGFR : EGFR and EGFR : ErbB2, could activate different signalling cascades because their cytoplasmic tails contain different numbers of tyrosine phosphorylation sites and bind to different signalling molecules (Yarden and Sliwkowski, 2001).

While both EGFR and ErbB2 are essential for epithelial cell survival and proliferation, activation of ErbB2 has been shown to disrupt apical-basal polarity and tight junctions by ErbB2’s direct association with the Par polarity complex (Aranda et al., 2006). Thus, while both EGFR and ErbB2 are activated by gonococci, the two receptors could play roles in different steps of the infection. Our finding of enhanced gonococcal invasion in the absence of ErbB2 kinase suggests that the kinase activity of ErbB2 is not essential and may be inhibitory to gonococcal invasion, but it does not exclude possible roles for ErbB2 in other cellular processes of gonococcal infection, such as disrupting the tight junction for gonococcal transmigration across the epithelium. Furthermore, the expression levels of different ErbB family receptors could be varied at different genital tissue locations, and their expression levels can be further regulated by sex hormones that control the menstrual cycle (McBean et al., 1997; Cotroneo et al., 2005). The ability of N. gonorrhoeae to activate multiple members of ErbB receptors may allow the bacteria to establish infection at different locations of the genital tissue and different stages of the menstrual cycle using different mechanisms.

Our results demonstrate the ability of N. gonorrhoeae to induce the activation of EGFR. Gonococci either may bind directly to EGFR or induce the expression and secretion of EGFR ligands thereby transactivating it. The results from this study argue against a direct interaction of gonococci with EGFR as the mechanism for EGFR activation. First, killed gonococci failed to recruit EGFR and ErbB2, even though they can adhere to the epithelial cell surface. Second, confocal microscopic studies revealed that EGFR and ErbB2 were adjacent to, but did not appear to colocalize with the bacteria extensively. Third, an anti-EGFR mAb binding to the extracellular ligand binding domain had no influence on gonococcal adherence to the epithelial cells. Our finding that gonococci not only induce HB-EGF transcription, but also the shedding of sHB-EGF supports a ligand driven activation of EGFR by N. gonorrhoeae.

EGFR ligands are expressed by many cell types, including epithelial and endothelial cells, and function in an endocrine, paracrine, autocrine or juxtacrine fashion. These ligands, including HB-EGF, amphiregulin and TGFα whose transcription is increased by gonococcal inoculation, are shed from the surface of cells by zinc metalloproteinases, either of the MMP or ADAM families. This study found that inhibition of MMP activation by both the specific inhibitor, batimastat and the removal of several MMPs from the epithelial cell surface by sequential heparin washes inhibited not only the shedding of sHB-EGF, but also the invasion of gonococci into epithelial cells, without affecting adherence and microcolony formation. Because MMPs are responsible for cleavage of EGFR ligands besides HB-EGF, these results cannot preclude the possible contribution of other EGFR ligands, in addition to HB-EGF, in the invasion process. These results demonstrate that gonococci activate EGFR by inducing the expression and surface cleavage of a subset of EGFR ligands, a transactivation mechanism, and that this transactivation is important for gonococcal invasion.

The requirement of gonococcal induced EGFR ligand shedding for gonococcal invasion is further underscored by our finding that exogenous addition of low concentrations of sHB-EGF rescued invasion of the gonococci. This study also found that the addition of sHB-EGF to cells that were not heparin washed or addition of high concentrations of HB-EGF to heparin washed cells decreased gonococcal invasion. This suggests that the cleavage of HB-EGF and HB-EGF-mediated activation of EGFR in response to gonococci is a tightly controlled process. Because HB-EGF is membrane associated both before and after MMP-mediated cleavage, the cleavage event and EGFR activation events induced by gonococci are likely to be localized. The recruitment of EGFR to gonococcal microcolonies supports this hypothesis. The exogenous addition of sHB-EGF to epithelial cells, on the other hand, activates EGFR at the cell surface indiscriminately, which will activate EGFR that are and are not recruited to the bacteria. Activation of EGFR by sHB-EGF induces signalling events, and such indiscriminate activation of EGFR signalling, when it reaches a certain level, may interfere with gonococcal invasion. Whether gonococci induce localized cleavage of HB-EGF and activation of EGFR at the sites of their adherence and how EGFR contribute to the bacterial invasion are the subjects of future investigation.

As an in vitro study, we have used cancer epithelial cell lines that have originated from genital tissue to study the role of EGFR in gonococcal invasion. Over expression of the ErbB family of receptors is common in epithelial cell cancers. The results generated from an epithelial cell line that expresses much higher levels of ErbB receptors than normal epithelial cells may not reflect what occurs in vivo. To address this potential issue, we used two different cell lines that express different levels of EGFR and ErbB2. While HEC-1-B cells express much lower levels of EGFR and ErbB2 than ME180 cells, the two cell lines behaved similarly in most of the analyses. The higher sensitivity of HEC-1-B cells to EGFR and ErbB2 inhibitors in gonococcal invasion than ME180 cells further argues against a significant impact of EGFR and ErbB2 expression levels on our conclusion.

Taken together, the results of this study not only demonstrate that EGFR activation is important for N. gonorrhoeae invasion, but also reveal the mechanism by which N. gonorrhoeae activates EGFR, transactivation by increasing the surface shedding of EGFR ligands. EGFR is a key surface receptor on epithelial cells, and its signal transduction function is essential for epithelial cell survival and proliferation. Multiple studies have shown that other bacterial pathogens transactivate host EGFR for their survival (Seo et al., 2000; Zhang et al., 2004; Mikami et al., 2005; Yan et al., 2009). This study demonstrates that N. gonorrhoeae has the capability to co-opt host signalling through an indispensable receptor for their invasion. Activation of EGFR leads to many different outcomes depending on cell types and their micro-environment. Thus, hijacking the EGFR signalling pathway could be a common mechanism for pathogens to drive their invasion and intracellular survival. Further studies are required to understand how gonococci are able to induce the expression and surface cleavage of EGFR ligands and how EGFR signalling mediates gonococcal invasion.

Experimental procedures

Bacterial strains and epithelial cell lines

Neisseria gonorrhoeae strain MS11 was maintained as a frozen stock of predominately piliated (Pil) and Opa-expressing (Opa) variants. Unless stated otherwise, MS11 Pil Opa variants were used throughout this study. For use in assays, gonococci were grown on gonococcal media base (GCK) plates with 1% Kellogg’s supplement (White and Kellogg, 1965) for 15–18 h. Pil Opa variants were selected by light refracting properties of bacterial colonies using a dissecting light microscope. The MS11 opa deletion mutant (Δopa) was generated by deleting all eleven opa genes. The deletion begins at the first T of the −35 region of the promoter to the A of the stop codon for all 11 opa genes. The correct deletion of all opa genes was verified by PCR analysis, and Southern and Western blotting (Levan et al., manuscript in preparation). The concentration of bacteria in suspension was determined spectrophotometrically and verified by viable plate count. Gonococci were killed by incubation with 100 μg ml gentamicin sulfate at 37°C for 2 h, followed by overnight at 4°C. HEC-1-B cells, a human endometrial adenocarcinoma cell line (ATCC# HTB-113), were maintained in Eagle’s MEM supplemented with 10% fetal bovine serum (FBS). ME180 cells, a human cervical epidermal carcinoma cell line (ATCC# HTB-33), were maintained in RPMI1640 supplemented with 10% FBS. For establishing polarized epithelial cells, HEC-1-B cells were seeded at 4 × 10 into 6.5 mm diameter, 3 μm pore size transwell filters (Corning, Lowell, MA, USA) and incubated at 37°C with 5% CO2, changing the media every other day. Polarization was monitored by transepithelial resistance (TER) readings daily using a Millicell ERS volt-ohm meter (Millipore, Bedford, MA, USA). The cells were allowed to grow for 7–9 days until the TER values reached ~400 Ω, the maximal resistance value that HEC-1-B cells can reach. TER measures the permeability of epithelial cell monolayers grown on transwell filters and reflects the integrity of polarized epithelial cells.

Inhibitors and antibodies

AG1478, an EGFR kinase inhibitor, AG825, an ErbB2 kinase inhibitor and anti-EGFR mAbs that were used for confocal microscopy and prevention of EGFR ligand binding were purchased from Calbiochem (San Diego, CA, USA). Anti-EGFR antibody that was used for Western blot and anti-ErbB2 antibodies were purchased from Santa Cruz Biotechnology (Santa Cruz, CA, USA). Anti-β-tubulin antibody and heparin were purchased from Sigma (St. Louis, MO, USA). Anti-phosphotyrosine mAb (4G10) was purchased from Millipore (Temecula, CA, USA). Anti-zonula occluden-1 was purchased from BD Biosciences (San Jose, CA, USA). Anti-sHB-EGF-biotin conjugate and HB-EGF was purchased from R&amp;D Systems (Minneapolis, MN, USA). Batimastat was purchased from Tocris Bioscience (Ellisville, MO, USA).

Bacterial adherence and invasion assays

Epithelial cells (5 × 10/well) were seeded in 96-well plates and incubated at 37°C in 5% CO2. After 24 h, cells were cultured in serum-free medium overnight. Cells were preincubated with AG1478 and AG825 for 2 h, batimastat for 30 min or anti-EGFR mAb for 30 min. Next, cells were incubated with Pil Opa gono-cocci at an MOI of 5 for 2 h for adherence assays and 6 h for invasion assays at 37°C. For adherence assays, cells were washed with PBS and then lysed in 1% saponin, and appropriate dilutions were plated on GCK medium. For invasion assays, cells were washed and then incubated with 50 μg ml gentamicin for 1.5 h at 37°C. After extensively washing to remove remaining gentamicin, bacteria that had invaded were quantified by lysing the epithelial cells with 1% saponin and serially plating the cell lysates on GCK plates. The significance of differences was assessed using the Student’s t-test for independent population means.

For the heparin wash treatment, the epithelial cells were incubated with 5 mg ml heparin in serum-free media at 37°C two times for 15 min each and three times for 1 min. The cells were washed with serum-free media four times to remove any remaining heparin before proceeding with the adherence or invasion assays.

Immunofluorescence microscopy

Epithelial cells were seeded at 2 × 10 onto coverslips in 24-well dishes, cultured for 24 h and then serum-starved overnight. Cells were incubated with gonococci at an MOI of 5 for 5 h, washed and fixed with 4% paraformaldehyde (PFA) (Electron Microscopy Sciences, Ft. Washington, PA, USA). Then, cells were stained with anti-ErbB2, anti-EGFR and anti-N. gonorrhoeae antibodies (Bish et al., 2008). The polarized cells were fixed before immunostaining using the pH shift method. The cells were first fixed with 4% PFA in 80 mM Pipes, pH 6.5, 150 mM NaCl, 5 mM EGTA and 2 mM MgCl2 for 10 min and then shifted to 4% PFA in 100 mM NaBorate and 150 mM NaCl for 10 min. The cells were permeabilized and blocked in PB solution (DMEM, 10% FBS, 10 mM Hepes pH 7.6, 10 mM glycine, 0.05% saponin) and stained with primary and secondary antibodies in PB. After postfixing with 2% PFA, cells were mounted and analysed using a Zeiss LSM 510 laser scanning confocal microscope. For Z-stack images, a series of images from the top to bottom of the cells were taken at 0.5 μm steps.

Immunoblotting

Epithelial cells were seeded at 1 × 10 in 6-well dishes. After 24 h, the cells were serum-starved overnight and incubated with gonococci at an MOI of 5 for up to 6 h. The cells then were washed with ice-cold PBS and lysed in 75 μl RIPA buffer [1% NP-40, 0.5% deoxycholate, 0.1% SDS, 50 mM Tris-HCl pH 7.4, 150 mM NaCl, 1 mM EGTA, 2 mM EDTA, 1 mM Na3VO4, 50 mM NaF, 10 mM Na4P2O7, 1× proteinase inhibitor cocktail (Sigma)]. Cell lysates were mixed 1:1 with nondenaturing loading buffer for detection of HB-EGF or denaturing loading buffer for all other proteins. Lysates were separated through SDS-polyacrylamide gels (Bio-Rad, Hercules, CA, USA), analysed by Western blot, and visualized using Western Lightning chemiluminescence substrate (Perkin Elmer, Boston, MA, USA). Images were acquired and digitized directly using Fujifilm LAS-3000 (Valhalla, NY, USA) or acquired with X-ray film. Blots were stripped with Restore Western blot stripping solution (Pierce, Rockford, IL, USA) and reprobed with anti-β-tubulin antibody. The blots were quantified by densitometry using MultiGauge software from Fujifilm.

Immunoprecipitation

Epithelial cells were seeded at 1 × 10 in 6-well dishes. After 24 h, the cells were serum-starved overnight and incubated with gonococci at an MOI of 5 for up to 6 h. The cells then were washed with ice-cold PBS and lysed in 1% Triton X-100 lysis buffer [20 mM Tris-HCl pH 7.4, 150 mM NaCl, 1 mM MgCl2, 1 mM EGTA, 2 mM EDTA, 1 mM Na3VO4, 50 mM NaF, 10 mM Na4P2O7, 1× proteinase inhibitor cocktail (Sigma)]. The lysates were sonicated and centrifuged at 4°C. The supernatants were subjected to immunoprecipitation using anti-phosphotyrosine mAb and Protein G conjugated Sepharose beads (GE Health-care, Piscataway, NJ, USA). The immunoprecipitates were analysed using SDS-PAGE and Western blot, probing for EGFR or ErbB2 using specific antibodies.

Real-time PCR

HEC-1-B cells that were grown to ~ 90% confluence were serum starved overnight and incubated with MS11 Pil Opa gonococci at an MOI of 5 for up to 8 h. Total RNA was extracted with Trizol (Invitrogen). RNA was converted to cDNA with Superscript III (Invitrogen) using oligo dT primers according to the manufacturer’s protocol. The cDNA was amplified using 1× SYBR Green master mix (Applied Biosystems, Foster City, CA, USA) using the following conditions: 50°C for 2 min and then 95°C for 10 s, followed by 46 cycles of 95°C for 15 s, 58°C for 15 s and 72°C for 30 s. The products were denatured at 95°C for 15 s, annealed at 58°C for 30 s and then subjected to a slow dissociation by ramping from 58°C to 95°C at 2% of the normal ramp rate in order to insure that only one PCR product was amplified. The primers for the EGFR ligands have been previously published (Sorensen et al., 2004). Beta-actin was amplified as an internal standard using QuantumRNA beta-Actin primers (Ambion, Austin, TX, USA).

Bacterial strains and epithelial cell lines

Neisseria gonorrhoeae strain MS11 was maintained as a frozen stock of predominately piliated (Pil) and Opa-expressing (Opa) variants. Unless stated otherwise, MS11 Pil Opa variants were used throughout this study. For use in assays, gonococci were grown on gonococcal media base (GCK) plates with 1% Kellogg’s supplement (White and Kellogg, 1965) for 15–18 h. Pil Opa variants were selected by light refracting properties of bacterial colonies using a dissecting light microscope. The MS11 opa deletion mutant (Δopa) was generated by deleting all eleven opa genes. The deletion begins at the first T of the −35 region of the promoter to the A of the stop codon for all 11 opa genes. The correct deletion of all opa genes was verified by PCR analysis, and Southern and Western blotting (Levan et al., manuscript in preparation). The concentration of bacteria in suspension was determined spectrophotometrically and verified by viable plate count. Gonococci were killed by incubation with 100 μg ml gentamicin sulfate at 37°C for 2 h, followed by overnight at 4°C. HEC-1-B cells, a human endometrial adenocarcinoma cell line (ATCC# HTB-113), were maintained in Eagle’s MEM supplemented with 10% fetal bovine serum (FBS). ME180 cells, a human cervical epidermal carcinoma cell line (ATCC# HTB-33), were maintained in RPMI1640 supplemented with 10% FBS. For establishing polarized epithelial cells, HEC-1-B cells were seeded at 4 × 10 into 6.5 mm diameter, 3 μm pore size transwell filters (Corning, Lowell, MA, USA) and incubated at 37°C with 5% CO2, changing the media every other day. Polarization was monitored by transepithelial resistance (TER) readings daily using a Millicell ERS volt-ohm meter (Millipore, Bedford, MA, USA). The cells were allowed to grow for 7–9 days until the TER values reached ~400 Ω, the maximal resistance value that HEC-1-B cells can reach. TER measures the permeability of epithelial cell monolayers grown on transwell filters and reflects the integrity of polarized epithelial cells.

Inhibitors and antibodies

AG1478, an EGFR kinase inhibitor, AG825, an ErbB2 kinase inhibitor and anti-EGFR mAbs that were used for confocal microscopy and prevention of EGFR ligand binding were purchased from Calbiochem (San Diego, CA, USA). Anti-EGFR antibody that was used for Western blot and anti-ErbB2 antibodies were purchased from Santa Cruz Biotechnology (Santa Cruz, CA, USA). Anti-β-tubulin antibody and heparin were purchased from Sigma (St. Louis, MO, USA). Anti-phosphotyrosine mAb (4G10) was purchased from Millipore (Temecula, CA, USA). Anti-zonula occluden-1 was purchased from BD Biosciences (San Jose, CA, USA). Anti-sHB-EGF-biotin conjugate and HB-EGF was purchased from R&amp;D Systems (Minneapolis, MN, USA). Batimastat was purchased from Tocris Bioscience (Ellisville, MO, USA).

Bacterial adherence and invasion assays

Epithelial cells (5 × 10/well) were seeded in 96-well plates and incubated at 37°C in 5% CO2. After 24 h, cells were cultured in serum-free medium overnight. Cells were preincubated with AG1478 and AG825 for 2 h, batimastat for 30 min or anti-EGFR mAb for 30 min. Next, cells were incubated with Pil Opa gono-cocci at an MOI of 5 for 2 h for adherence assays and 6 h for invasion assays at 37°C. For adherence assays, cells were washed with PBS and then lysed in 1% saponin, and appropriate dilutions were plated on GCK medium. For invasion assays, cells were washed and then incubated with 50 μg ml gentamicin for 1.5 h at 37°C. After extensively washing to remove remaining gentamicin, bacteria that had invaded were quantified by lysing the epithelial cells with 1% saponin and serially plating the cell lysates on GCK plates. The significance of differences was assessed using the Student’s t-test for independent population means.

For the heparin wash treatment, the epithelial cells were incubated with 5 mg ml heparin in serum-free media at 37°C two times for 15 min each and three times for 1 min. The cells were washed with serum-free media four times to remove any remaining heparin before proceeding with the adherence or invasion assays.

Immunofluorescence microscopy

Epithelial cells were seeded at 2 × 10 onto coverslips in 24-well dishes, cultured for 24 h and then serum-starved overnight. Cells were incubated with gonococci at an MOI of 5 for 5 h, washed and fixed with 4% paraformaldehyde (PFA) (Electron Microscopy Sciences, Ft. Washington, PA, USA). Then, cells were stained with anti-ErbB2, anti-EGFR and anti-N. gonorrhoeae antibodies (Bish et al., 2008). The polarized cells were fixed before immunostaining using the pH shift method. The cells were first fixed with 4% PFA in 80 mM Pipes, pH 6.5, 150 mM NaCl, 5 mM EGTA and 2 mM MgCl2 for 10 min and then shifted to 4% PFA in 100 mM NaBorate and 150 mM NaCl for 10 min. The cells were permeabilized and blocked in PB solution (DMEM, 10% FBS, 10 mM Hepes pH 7.6, 10 mM glycine, 0.05% saponin) and stained with primary and secondary antibodies in PB. After postfixing with 2% PFA, cells were mounted and analysed using a Zeiss LSM 510 laser scanning confocal microscope. For Z-stack images, a series of images from the top to bottom of the cells were taken at 0.5 μm steps.

Immunoblotting

Epithelial cells were seeded at 1 × 10 in 6-well dishes. After 24 h, the cells were serum-starved overnight and incubated with gonococci at an MOI of 5 for up to 6 h. The cells then were washed with ice-cold PBS and lysed in 75 μl RIPA buffer [1% NP-40, 0.5% deoxycholate, 0.1% SDS, 50 mM Tris-HCl pH 7.4, 150 mM NaCl, 1 mM EGTA, 2 mM EDTA, 1 mM Na3VO4, 50 mM NaF, 10 mM Na4P2O7, 1× proteinase inhibitor cocktail (Sigma)]. Cell lysates were mixed 1:1 with nondenaturing loading buffer for detection of HB-EGF or denaturing loading buffer for all other proteins. Lysates were separated through SDS-polyacrylamide gels (Bio-Rad, Hercules, CA, USA), analysed by Western blot, and visualized using Western Lightning chemiluminescence substrate (Perkin Elmer, Boston, MA, USA). Images were acquired and digitized directly using Fujifilm LAS-3000 (Valhalla, NY, USA) or acquired with X-ray film. Blots were stripped with Restore Western blot stripping solution (Pierce, Rockford, IL, USA) and reprobed with anti-β-tubulin antibody. The blots were quantified by densitometry using MultiGauge software from Fujifilm.

Immunoprecipitation

Epithelial cells were seeded at 1 × 10 in 6-well dishes. After 24 h, the cells were serum-starved overnight and incubated with gonococci at an MOI of 5 for up to 6 h. The cells then were washed with ice-cold PBS and lysed in 1% Triton X-100 lysis buffer [20 mM Tris-HCl pH 7.4, 150 mM NaCl, 1 mM MgCl2, 1 mM EGTA, 2 mM EDTA, 1 mM Na3VO4, 50 mM NaF, 10 mM Na4P2O7, 1× proteinase inhibitor cocktail (Sigma)]. The lysates were sonicated and centrifuged at 4°C. The supernatants were subjected to immunoprecipitation using anti-phosphotyrosine mAb and Protein G conjugated Sepharose beads (GE Health-care, Piscataway, NJ, USA). The immunoprecipitates were analysed using SDS-PAGE and Western blot, probing for EGFR or ErbB2 using specific antibodies.

Real-time PCR

HEC-1-B cells that were grown to ~ 90% confluence were serum starved overnight and incubated with MS11 Pil Opa gonococci at an MOI of 5 for up to 8 h. Total RNA was extracted with Trizol (Invitrogen). RNA was converted to cDNA with Superscript III (Invitrogen) using oligo dT primers according to the manufacturer’s protocol. The cDNA was amplified using 1× SYBR Green master mix (Applied Biosystems, Foster City, CA, USA) using the following conditions: 50°C for 2 min and then 95°C for 10 s, followed by 46 cycles of 95°C for 15 s, 58°C for 15 s and 72°C for 30 s. The products were denatured at 95°C for 15 s, annealed at 58°C for 30 s and then subjected to a slow dissociation by ramping from 58°C to 95°C at 2% of the normal ramp rate in order to insure that only one PCR product was amplified. The primers for the EGFR ligands have been previously published (Sorensen et al., 2004). Beta-actin was amplified as an internal standard using QuantumRNA beta-Actin primers (Ambion, Austin, TX, USA).

Acknowledgments

This work was supported by NIH Grants AI68888 (DCS and WS), ARRA Supplement to {"type":"entrez-nucleotide","attrs":{"text":"AI065605","term_id":"30052316","term_text":"AI065605"}}AI065605 (JMcLG), and the Research Service of the Department of Veterans Affairs (JMcLG). We acknowledge the UMCP-CBMG Imaging Core, where all confocal microscopy experiments were performed.

Department of Cell Biology &amp; Molecular Genetics, and Maryland Pathogen Research Institute, University of Maryland, College Park, MD, USA
Department of Laboratory Medicine, VA Medical Center and University of California, San Francisco, CA, USA
For correspondence. ude.dmu@gnosxnew; Tel. (+1) 301 405 7552; Fax (+1) 301 314 9489

Summary

Neisseria gonorrhoeae, the causative agent of the sexually transmitted infection gonorrhoea, adheres to and invades into genital epithelial cells. Here, we investigate host components that are used by the bacteria for their entry into epithelial cells. We found that gonococcal microcolony formation on the surface of HEC-1-B cells disrupted the polarized, basolateral distribution of both epidermal growth factor receptor (EGFR) and ErbB2, a related family member, and induced their accumulation under the microcolonies at the apical membrane. Gonococcal infection increased EGFR and ErbB2 phosphorylation. The EGFR kinase inhibitor, AG1478, reduced gonococcal invasion by 80%, but had no effect on adherence or the recruitment of EGFR and ErbB2 to the microcolonies. Gonococcal inoculation upregulated the mRNA levels of several ligands of EGFR. Prevention of EGFR ligand shedding by blocking matrix metalloproteinase activation reduced gonococcal invasion without altering their adherence, while the addition of the EGFR ligand, HB-EGF, was able to restore invasion to 66% of control levels. These data indicate that N. gonorrhoeae modulates the activity and cellular distribution of host EGFR, facilitating their invasion. EGFR activation does not appear to be due to direct gonococcal binding to EGFR, but instead by its transactivation by gonococcal induced increases in EGFR ligands.

Summary
Collaboration tool especially designed for Life Science professionals.Drag-and-drop any entity to your messages.