West Nile virus nonstructural protein NS1 inhibits complement activation by binding the regulatory protein factor H

Abstract

The complement system, by virtue of its dual effector and priming functions, is a major host defense against pathogens. Flavivirus nonstructural protein (NS)-1 has been speculated to have immune evasion activity, because it is a secreted glycoprotein, binds back to cell surfaces, and accumulates to high levels in the serum of infected patients. Herein, we demonstrate an immunomodulatory function of West Nile virus NS1. Soluble and cell-surface-associated NS1 binds to and recruits the complement regulatory protein factor H, resulting in decreased complement activation in solution and attenuated deposition of C3 fragments and C5b-9 membrane attack complexes on cell surfaces. Accordingly, extracellular NS1 may function to minimize immune system targeting of West Nile virus by decreasing complement recognition of infected cells.

Fig. 1.

Copurification of WNV NS1 and bovine fH. Supernatants were harvested from baculovirus-infected SF9 insect cells grown in 10% FBS, and NS1 was purified by nickel-affinity and size-exclusion chromatography. Silver-stained SDS/PAGE of column fractions is shown. WNV NS1 and bovine fH are indicated by arrows. The numbers across the top correspond to the different column fractions. Protein size markers are indicated to the left in kilodaltons.

Fig. 2.

Coimmunoprecipitation of WNV NS1 and human fH. Western blotting with a polyclonal anti-human fH after immunoprecipitation with anti-NS1-mAb-Seph4B, NHS (A and B), and supernatant from WNV replicon-expressing cells (A) or purified NS1 (B). Western blotting was also performed after immunoprecipitation with anti-NS1-mAb-Seph4B (C and D), purified human fH, and supernatant from replicon-expressing cells (C) or purified NS1 (D). BHK cell supernatant or BSA was used as negative control, respectively. fH was identified with a sheep polyclonal anti-human fH antibody. Lanes 1 and 2 of each image represent the starting material without immunoprecipitation. The arrow points to the band that corresponds to the 150-kDa human fH protein.

Fig. 3.

WNV NS1 and human fH binding by ELISA. (A) SDS/PAGE of purified WNV NS1 and human fH. WNV NS1 was purified from baculovirus-infected SF9 cells grown under serum-free conditions (lane 1). Human fH (lane 2) was commercially purchased. The positions of marker proteins are indicated to the left of the gel. (B) fH binds to solid-phase WNV NS1. Microtiter plates were coated overnight with purified NS1, WNV envelope (Ecto-E) protein, or BSA. After blocking, purified fH was added and detected with a polyclonal antibody against human fH. Results are representative of at least three independent experiments performed in duplicate. The error bars indicate standard deviations. (C) NS1 binds to solid-phase human fH. The microtiter plate was coated with human fH or BSA as control, blocked, and then incubated with purified WNV NS1. Subsequently, after washing, NS1 binding was detected with mAb 3NS1 or 22NS1. Experiments were performed in duplicate, and the results are representative of at least three independent experiments. (D) Saturation binding of fH and NS1. Increasing concentrations of fH were added to NS1 on microtiter plates and evaluated for binding by ELISA. The curve is representative of three independent experiments performed in duplicate.

Fig. 4.

WNV NS1 recruits fH in solution to degrade C3b. NS1 or BSA was incubated with 9NS1 or 22NS1 mAb Sepharose overnight at 4°C. The charged or uncharged beads were mixed with 10% NHS (A), purified human fH (B), 10% C57BL/6 mouse serum (MS, C), or 10% congenic fH^−/− MS (designated by triangle, D), washed extensively, and then incubated with biotinylated C3b and fI for 7 or 14 min (A and B) or 20 or 40 min (C and D). In each series of experiments, a buffer negative control (lanes 3 and 7) was included that contained only biotinylated C3b, fI, and anti-NS1-mAb-Sepharose. In E, the NS1-charged beads were added directly to biotinylated C3b and fI. Reactions were stopped with SDS-reducing sample buffer, and C3b fragments were visualized after Western blot. C3b fragments are labeled at the right of the gel, and cofactor activity is established by the appearance of the α1 fragment (≈67 kDa) of the C3b α′-chain. The band marked α represents the small amount of C3 in the preparation that was not cleaved to C3b. Its thioester bond, however, is cleaved, because it is also susceptible to degradation. Under longer exposure, the α2 degradation fragment (≈41 kDa) was also observed (data not shown). Data are representative of at least two to three independent experiments.

Fig. 5.

Inhibition of complement deposition by cell surface-associated NS1. (A) Flow cytometry histograms showing cell surface-associated expression of WNV NS1 on CHO-NS1 stable cell transfectants (green line). As a negative control, CHO-V cells transfected with the parent retroviral vector were stained (purple line). (B) WNV NS1 inhibits C3b deposition on CHO cells. CHO-V and CHO-NS1 cells were preincubated with C7d-HS, sensitized with an anti-CHO antibody, and incubated with C7d-HS in GVB-Mg²⁺-EGTA buffer. After washing, cells were stained with mAb against human C3d and analyzed by flow cytometry. The results are the average of three independent experiments performed in duplicate, and the difference in C3b deposition between CHO-V and CHO-NS1 cells was statistically significant (P < 0.01). (C) fH binding to CHO-V and CHO-NS1 cells. Cells were incubated with medium or 10% NHS for the indicated times (in minutes), washed extensively, and analyzed by Western blot with a sheep anti-human fH antibody. The ≈150 kDa fH is denoted by the arrow. (D) WNV NS1 inhibits C5b–9 deposition on CHO cells expressing WNV NS1. CHO-V and CHO-NS1 cells were preincubated with NHS, sensitized with an anti-CHO antibody, incubated with 10% (Upper) or 25% (Lower) NHS, and then immunostained with murine mAb against human C5b–9 (Right). The results are the average of three independent experiments performed in duplicate, and the difference in C5b–9 deposition between CHO-V and CHO-NS1 cells was statistically significant (P < 0.01). For negative control staining (Left), complement-activated cell lines were stained with only anti-mouse IgG conjugated with FITC. Purple and green lines refer to CHO-V and CHO-NS1 cells, respectively.