What Lies Beneath: Antibody Dependent Natural Killer Cell Activation by Antibodies to Internal Influenza Virus Proteins

Abstract

The conserved internal influenza proteins nucleoprotein (NP) and matrix 1 (M1) are well characterised for T cell immunity, but whether they also elicit functional antibodies capable of activating natural killer (NK) cells has not been explored. We studied NP and M1-specific ADCC activity using biochemical, NK cell activation and killing assays with plasma from healthy and influenza-infected subjects. Healthy adults had antibodies to M1 and NP capable of binding dimeric FcγRIIIa and activating NK cells. Natural symptomatic and experimental influenza infections resulted in a rise in antibody dependent NK cell activation post-infection to the hemagglutinin of the infecting strain, but changes in NK cell activation to M1 and NP were variable. Although antibody dependent killing of target cells infected with vaccinia viruses expressing internal influenza proteins was not detected, opsonising antibodies to NP and M1 likely contribute to an antiviral microenvironment by stimulating innate immune cells to secrete cytokines early in infection. We conclude that effector cell activating antibodies to conserved internal influenza proteins are common in healthy and influenza-infected adults. Given the significance of such antibodies in animal models of heterologous influenza infection, the definition of their importance and mechanism of action in human immunity to influenza is essential.

Keywords: Antibody dependent cellular cytotoxicity; Hemagglutinin; Influenza; Matrix protein 1; Natural killer cells; Nucleoprotein.

Supplementary Fig. 1

Influenza-specific NK cell activation by healthy donor plasma in primary and purified primary NK cells. Gating strategies are shown for primary NK cells in PBMCs (a) and purified primary NK cells (b). For both conditions, CD3 − CD56 + dim primary NK cells were selected for analysis using IFNγ and CD107a as activation markers. PBMCs or purified NK cells were incubated with influenza protein (600 ng/well) in the absence of plasma from a healthy influenza-exposed donor, irrelevant viral protein gp140 (600 ng/well) with plasma from an influenza-exposed donor and influenza proteins (M1 and NP) with plasma from an influenza-exposed donor. c) Primary and purified primary NK cell activation is shown with plasma from a healthy influenza-exposed donor (+) and an influenza-naïve pigtail macaque (−) by IFNγ and/or CD107a expression to HA of A/California/04/2009 (H1pdm09), HA of A/Perth/19/2009 (H3Perth09), M1 of A/Puerto Rico/8/1934 (M1), NP of A/California/07/2009 (NP) and irrelevant viral protein gp140. This assay showed similar results with a second healthy plasma donor and an IVIG preparation (10 mg/ml).

Supplementary Fig. 2

Activation of GFP-CD16 (V176) NK-92 and parental NK-92 cells by M1- and NP-specific Abs in an IVIG preparation. GFP-CD16 (V176) NK-92 and parental NK-92 cells were gated for size and granularity (FSC-A vs SSC-A) ensuring single cells (FSC-A vs FSC-H). CD16-GFP expression was shown for the CD16-GFP (V176) NK-92 cells (top panel labelled “CD16-GFP (V176) NK-92”) and the parental NK-92 cells (top panel labelled “Parental NK-92”). Single CD16-GFP (V176) NK-92 (middle panel) and parental NK-92 (bottom panel) cells were analysed for activation by CD107a expression following incubation with or without IVIG (10 mg/ml) and influenza proteins (M1 and NP at 600 ng/well) or an irrelevant viral protein (gp140 at 600 ng/well).

Supplementary Fig. 3

Surface and total expression of NP for rVV-NP infected CEM cells between 5 and 30 hours post-infection. Unpermeabilized (a) or permeabilized (b) CEM cells were stained with an anti-NP Ab conjugated to FITC and measured by flow cytometry at 5, 8, 24 and 30 hours post-infection (hpi) with no virus, rVV-wt or rVV-NP.

Fig. 1

M1- and NP-specific primary NK cell activation in healthy influenza-exposed adults. a) Lymphocytes were gated on by size and granularity (FSC-A vs SSC-A) ensuring single cells (FSC-A vs FSC-H). CD3 − CD56 + dim primary NK cells were selected for analysis using IFNγ and CD107a as activation markers. PBMCs were incubated with influenza protein (600 ng/well) in the absence of IVIG from influenza-exposed adults, irrelevant viral protein gp140 (600 ng/well) with IVIG from influenza-exposed adults and influenza proteins (M1 and NP) with IVIG from influenza-exposed adults. Primary NK cell activation with plasma from 14 healthy adults (Flu +) and four influenza-naïve pigtail macaques (Flu −) is shown by IFNγ (b) and CD107a (c) expression to HA of A/California/04/2009 (H1pdm09), HA of A/Perth/19/2009 (H3Perth09), M1 of A/Puerto Rico/8/1934 (M1), NP of A/California/07/2009 (NP) and irrelevant viral protein gp140. Values are unsubtracted with gp140 background shown for all samples. For each influenza protein tested Flu + and Flu − groups were compared with a Mann Whitney U test where p < 0.05 was considered significant. A Friedman test followed by a Wilcoxon matched pairs signed-rank test with a Bonferroni correction for multiple comparisons was used to compare the Flu + group incubated with M1 and NP to the Flu + group incubated with irrelevant protein gp140, a corrected p value of < 0.0125 considered significant. *** = p < 0.001.

Fig. 2

Titration of M1- and NP-specific NK activating Abs in healthy adults with NK-92 cells and correlation with primary NK cells. a) NK-92 cells were gated on by size and granularity (FSC-A vs SSC-A) ensuring single cells. CD16-GFP + cells were selected for analysis using CD107a as an activation marker. NK-92 cells were incubated with influenza protein (600 ng/well) in the absence of IVIG from influenza-exposed adults, irrelevant viral protein gp140 (600 ng/well) with IVIG from influenza-exposed adults and influenza proteins (M1 and NP) with IVIG from influenza-exposed adults. 14 healthy donor plasma samples previously screened for primary NK cell activation in the presence of influenza proteins (Fig. 1b, c) were titrated for NK cell activating Abs to M1 (b) in a series of 2-fold plasma dilutions and NP (d) in a series of 4-fold plasma dilutions. All values were background subtracted with wells containing plasma and the irrelevant HIV-1 protein gp140. Correlations between primary NK cell activation (percentage of NK cells expressing IFNγ) with undiluted plasma and NK-92 activation (percentage of NK-92 cells expressing CD107a) with a 1:40 plasma dilution are shown for M1 (c) and NP (e). Spearman correlation was used to determine correlation between primary NK and NK-92 cell activation where p < 0.05 was considered significant.

Fig. 3

NP and M1 opsonised with Abs from healthy influenza-exposed adults bind dimeric rsFcγRIIIa. a) A subset of 9 healthy human donors (closed circle) and one influenza-naïve macaque (open circle) previously screened for NK cell activating Abs to M1 and NP (Fig. 1, Fig. 2) were tested for dimeric rsFcγRIIIa binding to Ab opsonised M1 and NP using a novel ELISA. Three half-log plasma dilutions starting at 1:10 were screened for each donor. All normalised OD values were subtracted with a no antigen control well, containing plasma but no influenza protein. Correlations between primary NK cell activation (percentage of NK cells expressing IFNγ) and dimeric rsFcγRIIIa binding with a plasma dilution of 1:10 for M1 (b) and 1:320 for NP (c) are shown. All percentages of primary NK cells expressing IFNγ are background subtracted with wells containing plasma and the irrelevant HIV-1 protein gp140. Spearman correlation was used to determine correlation between primary NK cell activation and dimeric rsFcγRIIIa binding where p < 0.05 was considered significant.

Fig. 4

Influenza-specific NK cell activation by IVIG preparations and titration of H1pdm09, M1 and NP NK cell activating Abs in five IVIG samples. 18 IVIG preparations (10 mg/ml) were tested for primary NK cell (a) and NK-92 cell (b) activation to influenza proteins H1pdm09, M1 and NP as well as the irrelevant HIV-1 protein gp140. Values are unsubtracted with gp140 background shown for all samples. A Friedman test followed by a Wilcoxon matched pairs signed-rank test with a Bonferroni correction for multiple comparisons was used to compare influenza-specific NK activation by IVIGs to activation by gp140, a corrected p value of < 0.0167 considered significant. Titrations of NK cell activating Abs to H1pdm09 (c), M1 (d) and NP (e) were performed in a series of 2-fold dilutions for five IVIG samples two prepared pre-2009 (broken line, open symbol) and three prepared post-2009 (solid line, closed symbol). For all titrations (c–e) values were background subtracted with wells containing IVIG and gp140. *** = p < 0.001.

Fig. 5

NK-92 activation by pre- and post-seroconversion sera samples from three naturally influenza-infected patients. a) A titration of NK-92 cell activating Abs, measured by percentage of CD107a⁺ cells, was performed with sera from subjects naturally infected with a suspected A/California/07/2009 (H1N1)-like influenza virus. NK cell activating Abs to influenza H1pdm09, M1 and NP were titrated in pre (broken line) and post-seroconversion (solid line) sera by 2-fold serial dilutions starting at a 1:80 dilution. All values were background subtracted with wells containing patient sera and gp140. b) Endpoint titres of NK cell activating Abs to H1pdm09 (closed black circles), M1 (open squares) and NP (grey triangles) are shown in pre- and post-seroconversion sera for the three subjects. Endpoint titres of NK cell activating Abs were defined as the last serum dilution before the percentage of NK-92 cells expressing CD107a fell below the threshold of three times background (calculated with wells containing sera and gp140). c) Correlation of NK cell activation (by CD107a expression) at a 1:80 serum dilution to endpoint titre of NK cell activating Abs. The percentage of CD107a⁺ NK-92 cells and the endpoint titres of NK cell activating Abs are depicted for all influenza proteins tested (H1pdm09, M1 and NP) with pre- and post-seroconversion sera from the three subjects naturally infected with influenza virus. Spearman correlation was used to compare NK-92 activation and endpoint titre of NK cell activating Abs, where p < 0.05 was considered significant.

Fig. 6

Influenza-specific Ab-dependent NK-92 activation in subjects experimentally infected with influenza and correlation with NAbs. The percentage of CD107a⁺ NK-92 cells following incubation with H3 Wsn05 (a), M1 (b) and NP (c) was compared for pre and post (36 days) infection plasma samples at a 1:20 dilution from 11 subjects experimentally infected with A/Wisconsin/67/2005 (H3N2) influenza virus. Black lines (circles) represent individuals that developed moderate/severe disease, grey lines (squares) depict subjects who showed mild disease and broken lines (triangles) are individuals that did not demonstrate any disease symptoms. All values were background subtracted with wells containing patient sera and gp140. Wilcoxon matched pairs signed-rank test was used to test for significant differences between pre- and post-infection plasma samples where p < 0.05 was considered significant. * = p < 0.05, ns = not significant. d) Correlation of HI titre to A/Wisconsin/67/2005 (H3N2) and percentage of NK-92 cells expressing CD107a for post-infection plasma samples. Spearman correlation was used to compare HI titre and NK-92 activation where p < 0.05 was considered significant.

Fig. 7

Breadth of NK cell activating Abs to homosubtypic and heterosubtypic HA proteins in experimentally influenza-infected subjects with moderate/severe disease. NK-92 activation following stimulation with H3 Wsn05 (unbroken black line, closed circle), H3 X-31 (broken grey line, open square), H3 Perth09 (unbroken grey line, open circle) and H1pdm09 (broken black line, open diamond) was compared for pre and post (36 days) infection plasma samples at a 1:20 dilution from four subjects that demonstrated moderate/severe disease following A/Wisconsin/67/2005 (H3N2) infection. All values were background subtracted with wells containing patient sera and gp140. Wilcoxon matched pairs signed- rank test was used to test for significant differences between pre- and post-infection plasma samples where p < 0.05 was considered significant. *** = p < 0.001.

Fig. 8

ADCC mediated killing of targets cells infected with rVVs expressing individual influenza proteins. a) Percent cytotoxicity (measured by LDH release) of rVV infected CEM T-cell targets by PBMC effectors in the presence of 50 μg/ml, 5 μg/ml, 0.5 μg/ml or 0 μg/ml IgG from a healthy influenza-exposed adult. Effector PBMCs were added to targets 20 hour post-infection (hpi) with rVV at an effector to target ratio of 25:1 (solid lines, closed symbols) or 12.5:1 (broken lines, open symbols). b) Percent cytotoxicity of rVV infected lung epithelial A549 targets by effector PBMCs in the presence of healthy influenza-exposed donor IgG at concentrations of 50 μg/ml, 5 μg/ml, 0.5 μg/ml or 0 μg/ml. PBMC effectors were added to targets 20 hpi with rVV at an effector to target ratio of 25:1. c) Percent cytotoxicity of rVV infected CEM T-cell targets by NK-92 effector cells in the presence of healthy influenza-exposed donor IgG at 50 μg/ml, 5 μg/ml, 0.5 μg/ml or 0 μg/ml. Effector NK-92 cells were added to targets 8 hpi with rVV at effector to target ratios of 5:1 (solid lines, closed symbols) and 1:1 (broken lines, open symbols).