Surfactant protein A alters endosomal trafficking of influenza A virus in macrophages

Abstract

Influenza A virus infection (IAV) often leads to acute lung injury that impairs breathing and can lead to death, with disproportionate mortality in children and the elderly. Surfactant Protein A (SP-A) is a calcium-dependent opsonin that binds a variety of pathogens to help control pulmonary infections by alveolar macrophages. Alveolar macrophages play critical roles in host resistance and susceptibility to IAV infection. The effect of SP-A on IAV infection and antiviral response of macrophages, however, is not understood. Here, we report that SP-A attenuates IAV infection in a dose-dependent manner at the level of endosomal trafficking, resulting in infection delay in a model macrophage cell line. The ability of SP-A to suppress infection was independent of its glycosylation status. Binding of SP-A to hemagglutinin did not rely on the glycosylation status or sugar binding properties of either protein. Incubation of either macrophages or IAV with SP-A slowed endocytic uptake rate of IAV. SP-A interfered with binding to cell membrane and endosomal exit of the viral genome as indicated by experiments using isolated cell membranes, an antibody recognizing a pH-sensitive conformational epitope on hemagglutinin, and microscopy. Lack of SP-A in mice enhanced IFNβ expression, viral clearance and reduced mortality from IAV infection. These findings support the idea that IAV is an opportunistic pathogen that co-opts SP-A to evade host defense by alveolar macrophages. Our study highlights novel aspects of host-pathogen interactions that may lead to better understanding of the local mechanisms that shape activation of antiviral and inflammatory responses to viral infection in the lung.

Keywords: collectin; influenza A virus; lung; macrophages; surfactant protein A.

Figure 1

SP-A inhibits IAV infection independent of its glycosylation. RAW264.7 cells were cultured overnight at a density of 2 x 10⁵ cells/well in 24-well plates and infected with either IAV PR8 (A-F) or IAV Phil82 (F). Cells were incubated with indicated concentration of virus in 1:1 (v/v) PBS/DMEM and then allowed to proceed in DMEM/10% FBS. At the endpoint of the following experiments cells were harvested, stained with NP antibodies, and analyzed by flow cytometry. (A) SP-A inhibits IAV infection in a concentration-dependent manner. RAW264.7 were pretreated with increasing concentration of 5, 10, 25 and 50 μg/mL SP-A for 5 hours and then infected with IAV PR8 at MOI=2 in the presence of respective concentration of SP-A. The percentage of infected cells was determined by flow cytometry 10-hous after infection. N=4, t-test, **p < 0.05. (B) IAV overcomes SP-A inhibition at high MOI. RAW264.7 cells were incubated in the presence or absence of 50 μg/mL of SP-A for 5 hours, then infected with IAV PR8 at MOI 1, 2, and 5 in 1:1 PBS/DMEM in the presence or absence of 50 μg/mL SP-A and infection was assessed by flow cytometry. N=6, t-test, **p < 0.05 (C) Either pretreatment of macrophages or IAV with SP-A inhibits infection. RAW264.7 cells were incubated with 50 μg/mL SP-A in indicated combinations before or during infection. N=2 independent experiments per condition. (D) De-glycosylation of SP-A was accomplished using EndoF and confirmed by visualization of molecular weight reduction of monomeric and oligomeric forms of SP-A on Western blots using a polyclonal SP-A antibody. (E) The inhibitory effect of SP-A does depend on its glycoconjugate moiety. De-glycosylation enhances the inhibitory effect SP-A on IAV infection. RAW264.7 cells were treated with either native of de-glycosylated (ΔCHO) SP-A and infected with IAV at MOI=2 and infection assessed at 6 and 12 hrs. after infection. N=4, t-test, *p < 0.05; **p < 0.01, NS, not significant. (F) SP-A delays temporal infection with both H1N1 and H3N2 strains of IAV. RAW264.7 were infected with either IAV H1N1 strain PR8 or H3N2 strain Phil82 following a 5-hour treatment with no or either 10 or 50 μg/mL SP-A. Infection was then assessed by flow cytometry at 3, 12, or 24 hours after infection. Representative histograms of n=2 independent experiments per treatment are shown.

Figure 2

SP-A binds HA in a calcium and glycosylation independent manner. SP-A binding to HA was carried out in static solid phase assays (A-C) and under flow using surface plasmon resonance (SPR) (D-F). For static assays, recombinant PR8 HA (1 μg/well) was adsorbed onto 96-well Immobilon-2 plates in 0.05 M bicarbonate buffer (pH 9.6) overnight at 4°C. Coated plates were incubated with albumin control or SP-A at room temperature for 1 hour. Plates were washed in binding buffer and incubated with a polyclonal SP-A antibody. Bound SP-A was visualized colorimetrically using HRP-conjugated anti-rabbit antibody and tetramethylbenzidine. Albumin was used as non-specific control. (A) SP-A binding to HA does not depend on the presence of calcium. HA and albumin control coated plates were washed and incubated with increasing concentration of 10, 25, 50, and 100 μg/mL of SP-A in the presence or absence of 5 mM EDTA. N=4 of 2 independent experiments. (B) De-glycosylation (ΔCHO) of HA does not prevent SP-A binding. Plates coated with ΔCHO recombinant HA, or albumin were washed, and then incubated with increasing concentration of 10, 50, and 100 μg/mL of SP-A. N=2. (C) De-glycosylation (ΔCHO) of SP-A does not prevent binding to HA. Recombinant HA or albumin control coated plates were incubated with increasing concentration of 10, 50, 75, and 100 μg/mL of SP-A or ΔCHO SP-A. N=4 of 2 independent experiments. (D-F) Calcium does not impact kinetic and binding parameters of SP-A binding to HA. For SPR assays, recombinant HA with a carboxy-terminal hexa-histidine tag was immobilized onto Nicoya nitrilotriacetic acid (NTA) sensor chips. Increasing concentration of 10, 20, 40, or 80 μg/mL of SP-A analyte in the presence (D) or absence (E) of 0.5 mM CaCl₂ was injected at 20 μL/min for 300 sec to obtain on-rates and switched to buffer without SP-A for an additional 600 sec to obtain off-rates. Binding sensograms were acquired using a 2-channel Nicoya OpenSPR instrument. Curves were fitted according to 1:1 interaction model to obtain kinetic and binding parameters (F). Data shown are representative of N=3 independent experiments.

Figure 3

SP-A delays IAV infection at endosomal level. (A) SP-A delays synthesis of HA. RAW264.7 cells were treated with SP-A for five hours and then infected with IAV at MOI=15. Cells were harvested 1, 2, 4, 6, 12, and 24-hours after infection to obtain cell extracts for Western blot analysis using a polyclonal rabbit anti-HA antibody and a fluorescent IRDye800CW goat anti-rabbit secondary antibody. Blots were probed with an anti-mouse actin antibody and IRDye680RD goat anti-mouse antibody as loading control. Blots were quantitated by fluorescence densitometry using a LICOR instrument to obtain band intensity graphs and densitometry data for HA band intensity were normalized to actin. N=3, *p < 0.05 and ***p < 0.001 assessed by 2-way ANOVA. (B) To monitor trafficking to acidified endosomes, RAW264.7 cells were treated with SP-A for 5 hours and infected with IAV MOI=15 for 15, 35, 60, 90 and 240 minutes. Cells were harvested, fixed, permeabilized, and stained with an antibody recognizing a pH-dependent epitope that is exposed upon conformational transition of HA to the fusion competent form (pHHA). SP-A suppressed detection of the low pH HA conformation over the first 30 min of infection and delayed synthesis of new HA. N=4, *****p < 0.00005 assessed by tailed t-test. (C, D) SP-A treatment results in reduced infection and retention of NP in a punctate perinuclear compartment. Immunofluorescence microscopy was used to visualize the effect of SP-A on localization of viral NP. Cells were treated with 50 μg/mL SP-A for 5 hours and then infected with IAV for 4 hours. Cells were then fixed and stained with FITC-Cholera Toxin B subunit to stain the cell membrane, antibodies to NP to visualize IAV, and DAPI to stain the nucleus (C). Stained cells were counted across 10 microscopic fields at 63x magnification using a Nikon Eclipse confocal microscope to obtain the distribution of NP between nuclear and cytosolic compartments (D).

Figure 4

Differential reduction in uptake of endocytic cargo and binding of IAV PR8 to cell membrane in the presence of SP-A. Endocytic and binding experiments were performed using RAW264.7 cells (A-E) or isolated RAW264.7 membrane (F). Raw264.7 cells were incubated with Dextran 10,000 or transferrin conjugated with either the pH sensitive dye pHRodo (A, B) or FITC (C, D) in 1:1 DMEM in the presence or absence of SP-A, for 15 minutes at room temperature. The media were replaced with dye-free 1:1 DMEM and cells incubated with PHRodo- and FITC-labeled molecules were chased at 37°C for 10, 20, 30, and 45-minutes and 5, 10, 15, 30, and 45-minutes at 37°C to monitor rates of endosomal acidification and uptake, respectively. (A, B) SP-A does not alter acidification rate for either Dextran (A) or transferrin (B) tracked endosomes, N=6. (C) SP-A reduced level of FITC-Dextran over the first 5-minutes, whereas SP-A appeared to delay uptake of FITC-transferrin between 5-20-minutes reaching a similar plateau after 30-minutes. Differences were significant for FITC-Dextran only. N=6, *p<0.05. (D) and (E) SP-A suppresses uptake rate of Alexa Fluor488-labeled PR8. RAW264.7 cells were incubated with IAV with SP-A throughout infection or pre-treatment only. Cells were infected with Alexa Fluor488-labeled PR8 at MOI=15. The uptake of the labeled virus was monitored for 5, 15, 30, and 45-minutes at 37°C by flow cytometry. N=6 per condition per time point, **p < 0.005 and ****p < 0.001 for vehicle vs SP-A throughout; ++, p < 0.005 and +++, p<0.0005 for vehicle vs SP-A pretreatment; ^##, p < 0.05 for SP-A throughout vs SP-A. (F) To assess differences in membrane binding, immulon-2 flat bottom plates were coated overnight at 4°C with 12.5 μg/ml of Raw264.7 cell membranes and then washed and incubated with PBS, 5 x 10⁶ ffc of PR8, or 50 μg/ml SP-A for 2-hours at room temperature. The membranes were then washed, and incubated for an additional 2-hours with PR8, pre-formed PR8+SP-A complex, or SP-A to evaluate PR8 binding in a combination of conditions shown on the x-axis on panel F graph. Plates were washed and bound PR8 visualized spectrophotometrically using a polyclonal anti-HA antibody and HRP-conjugated anti-rabbit antibody as described in Materials and Methods. N=8, ****p < 0.00001, ***p < 0.001. Statistical differences were assessed by 2-way ANOVA.

Figure 5

Disruption of SP-A improves recovery and clearance of IAV infection in mice. Sftpa^+/+ (n=14), Sftpa^-/- (n=31), Csf2^-/- (n=9) and Sftpa^-/-, Csf2^-/- (n=27) mice were infected with 1000 ffc (1 LD50) of IAV PR8 via the intranasal route. Body weight and survival were monitored daily. Total RNA was extracted from whole lung to measure viral RNA and IFNβ. (A) Lack of SP-A in Sftpa^-/- mice enhances body weight recovery of Sftpa^-/- mice compared to Sftpa^+/+, Csf2^-/- and Sftpa^-/-,Csf2^-/- mice (*p < 0.05, Sftpa^-/- vs. Sftpa^+/+ days 9-14, ***p < 0.001 Sftpa^-/- vs. Csf2^-/- and Sftpa^-/-,Csf2^-/- mice), reduces body weight loss and mortality in the absence of GM-CSF (*p<0.05 Sftpa^-/- vs. Sftpa^+/+ and ***p<0.001, Csf2^-/- vs. Sftpa^-/-, Csf2^-/- ) (A, B), enhances IAV clearance from the lung (***p < 0.001 Sftpa^-/- vs. Sftpa^+/+ on day 7, ^# p < 0.05 Sftpa ^-/- day 7 vs. day 3 (C) and enhances expression of IFNβ as measured on day 3 after infection, n=6 mice, *p < 0.05 (D). The histopathology (E) and histopathology scores (F, G) Sftpa^-/- and Sftpa^+/+ consisting of perivascular cuffing (yellow arrows) and interstitial inflammation 14 days after infection were similar. Representative hematology and eosin stained tissue sections for Sftpa^+/+ (upper panel) and Sftpa^-/- (lower panel) lungs in (E) were captured at 20x magnification. For (F, G) n=6 mice per group. Percent lung affected (F) and inflammation severity (G) scores shown are the mean observation from 5 microscopic fields from each lobe per mouse.

Figure 6

Proposed model of SP-A-IAV interaction in macrophages. Multi-site inhibition of IAV infection of macrophages occurs through indirect inhibition of macropinocytic entry, delayed sorting to acidified endosomes or steric hindrance of endosomal fusion of HA, and reciprocal inhibition of cognate receptor binding. A direct mechanism by the SP-A receptor SP-R210 may mediate endocytic entry and result in permissive infection that may be deployed by different strains of IAV along a gradient of SP-A levels and oligomeric forms towards the lower respiratory tract and over the course of infection.