Fig 1.
Expression of EPCR and thrombomodulin by endothelial cells after infection by N. meningitidis.
Human Dermal Microvessels Endothelial Cells (HDMEC) were infected with a wild-type meningococcus (red) or left non-infected (green). Grey: isotype control. EPCR or thrombomodulin membranous expression were assessed by a FACS analysis. (A) Time course of EPCR expression during 4 hours of infection. The % of cells with no EPCR staining (EPCR-neg) is also indicated. EPCR-neg population was gated using non-infected cells. For the quantification of EPCR expression see supplemental data S1 Fig. (B) Thrombomodulin expression after 4 hours of infection. Data are representative of 3 independent experiments.
Fig 2.
Meningococcal adhesion on endothelial cells is required to decrease EPCR expression.
(A) HDMEC cells were left non-infected or were infected for 4h with a WT strain. DNA (cyan), junctionnal VE-Cadherin (red), or membranous EPCR (green) were stained using appropriate antibodies or DAPI. Bacteria were also stained using DAPI. Upper panel: non-infected cells. Medium panel: infected cells. Lower panel is a detailed view of the infected monolayer showing a non-infected cell surrounded by cells with N. meningitidis microcolonies (arrows) on the apical side. Scale bar: 50 μM. (B) HDMEC cells were infected for 4 hours with a WT piliated N. meningitidis strain (red) or a non-adherent non-piliated ΔpilE strain (orange) or left non-infected (green). EPCR expression was assessed by a FACS analysis. Data are representative of 3 independent experiments. (grey: isotype control). For the quantification of EPCR expression see supplemental data S1 Fig.
Fig 3.
Meningococcal type IV pili trigger the decrease of EPCR expression.
HDMEC were infected with either a WT N. meningitidis strain at a Multiplicity Of Infection (MOI) = 25 (red), Opa+ΔSiaD strain at MOI = 25 (black), Opa+ΔSiaDΔpilE strain at MOI = 500 (blue–see Material and Methods) or left uninfected (green, tinted) for 4 hours (grey: isotype control). EPCR expression was assessed by a FACS analysis. Data are representative of 3 independent experiments. (grey: isotype control). For the quantification of EPCR expression see supplemental data S1 Fig.
Fig 4.
Meningococcal adhesion on endothelial cells induces an ADAM-dependant shedding of EPCR.
(A) HDMEC were infected for 4 hours with a WT strain, a non-adherent non-piliated ΔpilE strain, or left non-infected (control). The amount of soluble forms of EPCR (sEPCR) was determined in the supernatant of cells by ELISA after infection. Data are mean +/- SEM of 3 independent experiments. * p = 0.0001 versus control (t-test). (B) HDMEC cells were treated for 1 hour with Tapi-0 (25μM), a pan-ADAM inhibitor, then infected with a WT strain (red) or left non-infected (green). EPCR expression was assessed by a FACS analysis. Data are representative of at least 3 independent experiments. (grey: isotype control). For the quantification of EPCR expression see supplemental data S1 Fig. (C) Amount of soluble EPCR in the supernatant determined by ELISA. Data are mean +/- SEM of 3 independent experiments.
Fig 5.
ADAM17 is not responsible for the meningococcus-induced EPCR shedding.
(A) ADAM17 membranous expression of HDMEC cells transfected with a siRNA against ADAM17 (blue) or control (green). Grey: isotype control. For the quantification of ADAM17 expression see supplemental data S3 Fig. (B) EPCR expression in HDMEC cells transfected with either a control siRNA (left) or ADAM17 siRNA (right) and infected with a WT N. meningitidis strain for 4 hours (red) or left non-infected (green). Grey: isotype control. For the quantification of EPCR expression see supplemental data S3 Fig. A Crispr/Cas9-mediated knockdown of ADAM17 expression was performed as described in Material and Methods in the Human Cerebral Microvessels Endothelial Cells line hCMEC/D3. (C) ADAM17 expression in the resulting cell line (blue) compared to that of the parental cell line (green). (D) EPCR expression in parental hCMEC/D3 cell line (left) or ADAM17-negative-hCMEC/D3 cell line (right) infected with a WT N. meningitidis strain for 4 hours (red) or left non-infected (green). The adhesion of N. meningitidis on hCMEC/D3 is reduced compared to the adhesion on HDMEC explaining the slighty reduced shedding phenotype observed with hCMEC/D3. Data are representative of 3 independent experiments. (grey: isotype control). For the quantification of EPCR expression see supplemental data S5 Fig. (E) Soluble EPCR in the supernatant of ADAM17-negative-hCMEC/D3 cells after infection with a WT strain or non-infected (control) cells. Data are mean (+/- SEM) of 3 independent experiments. *:p = 0.007 (t-test).
Fig 6.
The EPCR shedding induced by meningococcus is mediated by ADAM10.
(A) HDMEC were treated with the ADAM10 inhibitor GI254023X (1μM, right), or DMSO only (left) and were infected for 4h by a wild-type adherent N. meningitidis (red) or left non-infected (green). EPCR expression was assessed by a FACS analysis (grey: isotype control). Data are representative of 3 independent experiments. For the quantification of EPCR expression see supplemental data S1 Fig. (B) ADAM10 membranous expression of HDMEC cells transfected with a siRNA against ADAM10 (blue) or control (green). Grey: isotype control. For the quantification of ADAM10 expression see supplemental data S3 Fig. (C) EPCR expression by HDMEC cells transfected with either a control siRNA (left) or ADAM10 siRNA (right) and infected with a WT N. meningitidis strain for 4 hours (red) or left non-infected (green). Grey: isotype control. For the quantification of EPCR expression see supplemental data S3 Fig. A Crispr/Cas9-mediated knockdown of ADAM10 expression was performed in hCMEC/D3 as described in the Material and Methods section. (D) ADAM10 expression in the resulting cell line (blue) compared to that of the parental cell line (green). (E) EPCR expression in the parental hCMEC/D3 cell line (left) or ADAM10-negative-hCMEC/D3 cell line (right) and infected with a WT N. meningitidis strain for 4 hours (red) or left non-infected (green). Data are representative of 3 independent experiments. (grey: isotype control). For the quantification of EPCR expression see supplemental data S5 Fig. (F) Soluble EPCR in the supernatant of ADAM10-negative-hCMEC/D3 cells after infection with a WT strain or non-infected (control) cells, assessed by ELISA. Data are mean (+/- SEM) of 3 independent experiments. (G) Soluble EPCR in the supernatant of ADAM17-negative or ADAM10-negative hCMEC/D3 cells after 4 hours of treatment with PMA (1 μM). Data are mean (+/- SEM) of 3 independent experiments. *: p < 0.0001.
Fig 7.
aPC generation by endothelial cells is impaired upon infection by N. meningitidis.
After infection or antibody treatment, endothelial cells were washed and incubated for 30 min with Protein C and thrombin. The amount of aPC generated was determined through its proteolytic activity against a chromogenic substrate. The same experimental procedure was applied to an empty well without cells to assess the spontaneous degradation of the chromogenic substrate (no cells bar). (A) HDMEC were infected with either a wild type or ΔpilE strain for 4 hours. Cells were also treated for 10 minutes with an antibody blocking EPCR-PC interactions (clone RCR-252, 10 μg/mL) or a control antibody (directed against tissue factor, 10 μg/mL). Data are means (+/-SEM) of % of aPC activity compared to that of non-infected cells from 3 independent experiments. *:p<0.001 compared to control. (B) HDMEC cells were infected for 4 hours with a WT strain in the presence or absence of Tapi-0 (25μM). Data are means (+/-SEM) of % of aPC activity compared to that of non-infected cells from 3 independent experiments. (C) ADAM17-negative or ADAM10-negative hCMEC/D3 cells were infected for 4 hours with a WT strain. Data are means (+/-SEM) of % of aPC activity compared to that of non-infected cells from 3 independent experiments. *: p<0.0001 compared to control.
Fig 8.
Meningococcal infection suppresses the aPC barrier protective effect.
The barrier permeability of HDMEC monolayers was assessed using the iCELLigence system that continuously measures the electrical impedance. Cells were infected with a wild type meningococcus for 4 hours or left non-infected. After the infection, cell media were replaced, and cells were treated with aPC (50 nM or 100 nM) or left untreated. After 2 hours, thrombin (1 nM) was added and electrical impedance (Cell Index) was measured every minutes. (A) Representative data showing the effect of thrombin on monolayers impedance when HDMEC cells were treated (blue) or not (green) with aPC (100 nM). (B) Representative data showing the effect of thrombin on monolayers impedance when HDMEC cells were treated (orange) or not (red) with aPC (100 nM). (C) Loss of CI induced by thrombin expressed as % of non-treated cells. Data are mean (+/-SEM) from three independent experiments. For each condition (non-infected/infected) an analysis of variance (ANOVA) was done followed by multiple comparisons by t-tests with a correction of Bonferonni. *: p <0.01 compared to non-treated cells. NS: non significant (p>0.05).