Respiratory Syncytial Virus Uses CX3CR1 as a Receptor on Primary Human Airway Epithelial Cultures

Abstract

Respiratory syncytial virus (RSV) is the most frequent cause of lower respiratory disease in infants, but no vaccine or effective therapy is available. The initiation of RSV infection of immortalized cells is largely dependent on cell surface heparan sulfate (HS), a receptor for the RSV attachment (G) glycoprotein in immortalized cells. However, RSV infects the ciliated cells in primary well differentiated human airway epithelial (HAE) cultures via the apical surface, but HS is not detectable on this surface. Here we show that soluble HS inhibits infection of immortalized cells, but not HAE cultures, confirming that HS is not the receptor on HAE cultures. Conversely, a "non-neutralizing" monoclonal antibody against the G protein that does not block RSV infection of immortalized cells, does inhibit infection of HAE cultures. This antibody was previously shown to block the interaction between the G protein and the chemokine receptor CX3CR1 and we have mapped the binding site for this antibody to the CX3C motif and its surrounding region in the G protein. We show that CX3CR1 is present on the apical surface of ciliated cells in HAE cultures and especially on the cilia. RSV infection of HAE cultures is reduced by an antibody against CX3CR1 and by mutations in the G protein CX3C motif. Additionally, mice lacking CX3CR1 are less susceptible to RSV infection. These findings demonstrate that RSV uses CX3CR1 as a cellular receptor on HAE cultures and highlight the importance of using a physiologically relevant model to study virus entry and antibody neutralization.

Fig 1. Differential neutralization of RSV infection of HeLa and HAE cultures.

rgRSV was incubated with (A) soluble HS, (B) anti-G protein mAb 131-2g or IgG1 (isotype control), or (C) anti-G protein mAb L9 or IgG2a for 30 min prior to inoculation of HeLa and HAE cultures. (D) 2–20 or (E) B1 was incubated with 5 μg/ml mAb 131-2g for 30 min prior to inoculation of HeLa and HAE cultures. Infected cells (fluorescent foci) were counted at 24 hr post inoculation. All data are of one experiment representative of three. Data are expressed as the mean ± s.d. of 3 wells. Two-sided student's t test: *P < 0.05, **P < 0.01, ***P < 0.001.

Fig 2. MAb binding to the RSV G protein.

(A) The G protein, an N-terminally anchored [37] type II glycoprotein, has a central conserved region that overlaps with the cysteine noose held together by two disulfide bonds. The third and fourth cysteines are also part of the CX3C motif. C terminal to the CX3C motif is a highly basic heparin binding domain (HBD). RSV G protein (90 kDa) from (B) parental WT (Long strain) RSV and L9 neutralization escape mutants RSV2 (F165L, F170L, I175T and C186R) and RSV6a (F168S, F170P, C186R and V225A), (C) parental strain A2 RSV and recombinant RSV with a F170A mutation, or (D) parental strain A2 RSV and recombinant RSV with a C186S mutation were detected by immunoblot with the indicated antibodies.

Fig 3. Detection of CX3CR1 on HAE cultures and importance of the RSV G protein interaction with CX3CR1 for infection.

(A) Cross-section of a paraffin embedded HAE cultures stained with polyclonal antibodies to human CX3CR1 (red) and counterstained with DAPI to identify nuclei (blue). Images were taken with a 20X objective. (B) HeLa and HAE cultures were incubated with DMEM, 25 μg/ml isotype control, or 25 μg/ml anti-CX3CR1 for 30 min prior to inoculation with rgRSV. (C) Cells were incubated with 25 μg/ml anti-CX3CR1 or isotype control before, during and after apical inoculation with rgRSV. (D) HeLa and HAE cultures were inoculated with rgRSV or C186S RSV or (E) with rgRSV or CX4C RSV. Infected cells (fluorescent foci) were counted at 24 hr post inoculation. All data are of one experiment representative of three. Data from b-d are expressed as the mean ± s.d. of 3 wells. Two-sided student's t test: *P < 0.05, **P < 0.01, ***P < 0.005.

Fig 4. RSV infection of CHO A745 cells expressing CX3CR1.

CHO A745 cells were transfected with a plasmid encoding (A) CX3CR1 or (B) empty vector and stained 24 hr later with anti-CX3CR1. Image was taken with a 20X objective. 24 hr post-transfection cells were inoculated with rgRSV in the (C) absence or (D) presence of 25 μg/ml anti-CX3CR1 or 5 μg/ml L9, 131-2g, or their isotype controls. Infected cells (fluorescent foci) were counted at 24 hr post inoculation. All data are of one experiment representative of three. Data in C and D are expressed as the mean ± s.d. of 3 wells. Two-sided student's t test: *P < 0.05, **P < 0.01.

Fig 5. RSV infection of WT and CX3CR1-/- mice.

For each experiment, WT and CX3CR1-/- mice (3 mice per group) were intranasally inoculated with (A) 10⁶ PFU rgRSV or (B) 10⁵ PFU rgRSV-F, which lacks the G and SH genes and grows to lower titers than rgRSV. Lungs were harvested at 5 dpi and PFU per lung was determined. Data for three rgRSV experiments were combined and expressed as the mean ± s.d. of 9 mice. The two-sided student's t test was used to determine significance.