Antibody Responses with Fc-Mediated Functions after Vaccination of HIV-Infected Subjects with Trivalent Influenza Vaccine

Abstract

This study seeks to assess the ability of seasonal trivalent inactivated influenza vaccine (TIV) to induce nonneutralizing antibodies (Abs) with Fc-mediated functions in HIV-uninfected and HIV-infected subjects. Functional influenza-specific Ab responses were studied in 30 HIV-negative and 27 HIV-positive subjects immunized against seasonal influenza. All 57 subjects received the 2015 TIV. Fc-mediated antihemagglutinin (anti-HA) Ab activity was measured in plasma before and 4 weeks after vaccination using Fc-receptor-binding assays, NK cell activation assays, and phagocytosis assays. At baseline, the HIV-positive group had detectable but reduced functional Ab responses to both vaccine and nonvaccine influenza antigens. TIV enhanced Fc-mediated Ab responses in both HIV-positive and HIV-negative groups. A larger rise was generally observed in the HIV-positive group, such that there was no difference in functional Ab responses between the two groups after vaccination. The 2015 TIV enhanced functional influenza-specific Ab responses in both HIV-negative and HIV-positive subjects to a range of influenza HA proteins. The increase in functional Ab responses in the HIV-positive group supports recommendations to immunize this at-risk group.

Importance: Infection with HIV is associated with increasing disease severity following influenza infections, and annual influenza vaccinations are recommended for this target group. However, HIV-infected individuals respond relatively poorly to vaccination compared to healthy individuals, particularly if immunodeficient. There is therefore a need to increase our understanding of immunity to influenza in the context of underlying HIV infection. While antibodies can mediate direct virus neutralization, interactions with cellular Fc receptors may be important for anti-influenza immunity in vivo by facilitating antibody-dependent cellular cytotoxicity (ADCC) and/or antibody-dependent phagocytosis (ADP). The ability of seasonal influenza vaccines to induce antibody responses with potent Fc-mediated antiviral activity is currently unclear. Probing the ADCC and ADP responses to influenza vaccination has provided important new information in the quest to improve immunity to influenza.

FIG 1

Study enrollment flowchart.

FIG 2

Hemagglutination inhibition (HI) assay. Serum HI titers against vaccine influenza virus strains A/South Australia/55/2014 (H3N2) (A), B/Phuket/3073/2013 (B), and A/California/7/2009 (H1N1) (C) and a nonvaccine strain, A/New Caledonia/20/1999 (H1N1) (D). Baseline and postvaccine data were compared using the Wilcoxon matched-pair test. Lines and error bars indicate means and SEM, respectively. A cutoff HI titer of ≥40 (dashed line) was used to assess positive responses.

FIG 3

FcγRIIIa dimer-binding assay. (A) Assay setup. Influenza HA protein (50 ng/well) was coated onto a 96-well plate before adding diluted plasma (containing IgGs). Subsequently, biotinylated FcγRIIIa dimers were added, and binding between Abs and FcγRIIIa dimers was measured using HRP-conjugated streptavidin, TMB, and HCl. Absorbance was measured at 450 nm. (B) FcγRIIIa-V158-binding Ab responses to HA from A/Switzerland/9715293/2013 (H3N2) (H3 Switz) and SIV gp120 protein. Responses are normalized to 5 μg/ml of IVIG. Plasma was diluted 20 times before analysis. Baseline and postvaccine data were compared using the Wilcoxon matched-pair test. HIV-negative and HIV-positive groups were compared using the Mann-Whitney test. P values of <0.05 were considered significant. Lines and error bars indicate means and SEM, respectively.

FIG 4

FcγRIIIa dimer-binding activity to vaccine and nonvaccine influenza virus strains. (A) FcγRIIIa-V158-binding Ab responses (left panel) and FcγRIIIa-F158-binding Ab responses (center panel) to the 2015 TIV vaccine-HAs [A/Switzerland/9715293/2013 (H3N2), A/California/7/2009 (H1N1), and B/Phuket/3073/2013] and FcγRIIIa-V158-binding Abs to three nonvaccine HAs (A/Perth/16/2009 (H3N2), A/New Caledonia/20/1999 (H1N1), and B/Florida/4/2006; right panel). Responses are with background subtracted (SIV gp120-coated wells) and normalized to 5 μg/ml of IVIG. Plasma was diluted 20 times before analysis (40 times before analysis with A/New Caledonia/20/1999). Baseline and postvaccine data were compared using Wilcoxon matched-pairs test. HIV-negative and HIV-positive groups were compared using the Mann-Whitney test. P values of <0.05 were considered significant. Lines and error bars indicate means and SEM, respectively. (B) Correlation of the HA-specific Abs binding to the high-affinity (V158) and low-affinity (F158) FcγRIIIa dimer, analyzed with nonparametric Spearman correlation. Data shown for all three vaccine HAs, i.e., A/South Australia/55/2014, A/California/7/2009, and B/Phuket/3073/2013. Results for all subjects (baseline and postvaccine) are shown in the graph. (C) Correlation between HI titers and FcγRIIIa-V158 dimer-binding Abs to the three vaccine antigens, A/Switzerland/9715293/2013 (A/South Australia/55/2014 used in HI assay), A/California/7/2009, and B/Phuket/3073/2013, analyzed with the nonparametric Spearman correlation. Results for all HIV-negative and HIV-positive subjects (baseline and postvaccine) are shown in the graph.

FIG 5

NK cell activation assay with the NK-92 cell line. (A) Assay setup. A 96-well plate was coated with influenza HA protein (600 ng/well) from A/Switzerland/9715293/2013 (H3N2) before diluted plasma (containing IgGs) was added. The plate was then incubated with FcγRIIIa-expressing NK-92 cells (a human NK cell line) for 5 h at 37°C, after which NK-92 cell activation was analyzed with flow cytometry by measuring the percentage of CD107a-expressing NK-92 cells. SIV gp120 protein (600 ng/well) was included in each experiment as a negative control. (B) The NK-92 cell-activating Ab response to A/Switzerland/9715293/2013, a new antigen in the 2015 TIV, and to SIV gp120 protein. NK cell activation is measured as the percentage of CD107a-expressing NK-92 cells. Plasma was diluted 40 times before analysis. Baseline and postvaccine data were compared using the Wilcoxon matched-pair test. HIV-negative and HIV-positive groups were compared using the Mann-Whitney test. P values of <0.05 were considered significant. Lines and error bars indicate means and SEM, respectively. (C) Correlation between FcγRIIIa-V158-binding Abs and NK-92 cell-activating Abs to HA from A/Switzerland/9715293/2013, analyzed with nonparametric Spearman correlation. Background-subtracted data are presented in the graph. Results for all HIV-negative and HIV-positive subjects (baseline and postvaccine) are shown in the graph. (D) NK-92 cell-activating Ab response to HA from A/Switzerland/9715293/2013 in the absence (HI < 40) and presence (HI ≥ 40) of HI Abs to A/South Australia/55/2014, an A/Switzerland/9715293/2013-like virus. Background-subtracted data are presented in the graph.

FIG 6

ADP assay with THP-1 monocytic cell line. (A) Assay setup. Fluorescently labeled beads were coated with both biotinylated HA from A/Switzerland/9715293/2013 (H3N2) and a fluorescent internalization probe and then opsonized with Abs and incubated with a THP-1 monocytic cell line, known to express FcγRIIa. Fluorescence derived from internalization probes on surface-bound beads was quenched with a complementary quencher probe before internalization was measured with flow cytometry. SIV gp120 protein was included in each experiment as a negative control. (B) ADP Ab responses to HA from A/Switzerland/9715293/2013 and to SIV gp120 protein in HIV-negative and HIV-positive subjects. ADP activity is measured as the percentage of monocytes with internalized beads. Baseline and postvaccine data were compared using the Wilcoxon matched-pair test. HIV-negative and HIV-positive groups were compared using the Mann-Whitney test. P values of <0.05 were considered significant. Lines and error bars indicate means and SEM, respectively. (C) FcγRIIa-H131-binding Ab responses to A/Switzerland/9715293/2013 (H3N2) HA. Responses are with background subtracted (SIV gp120-coated wells) and normalized to 5 μg/ml of IVIG. Plasma was diluted 20 times before analysis. Baseline and postvaccine data were compared using the Wilcoxon matched-pair test. HIV-negative and HIV-positive groups were compared using the Mann-Whitney test. P values of <0.05 were considered significant. Lines and error bars indicate means and SEM, respectively. (D) Correlation between FcγRIIa-H131-binding Abs and ADP-stimulating Abs to HA from A/Switzerland/9715293/2013, analyzed with nonparametric Spearman correlation. Background-subtracted data are presented in the graph. Results for all HIV-negative and HIV-positive subjects (baseline and postvaccine) are shown in the graph. (E) Correlation between HI Ab titers and ADP activity to A/Switzerland/9715293/2013 (A/South Australia/55/2014 used in HI assay), analyzed with the nonparametric Spearman correlation. Background-subtracted ADP data are presented in the graph. Results for all HIV-negative and HIV-positive subjects (baseline and postvaccine) are shown in the graph.

FIG 7

Correlations of CD4 T cell counts and HIV infection duration with functional Ab responses. (A) Correlation between CD4 T cell counts and the change in HI titer, FcγRIIIa-V158-dimer binding, NK cell activation, and ADP activity to A/Switzerland/9715293/2013 (postvaccine minus baseline response). (B) Correlation between the duration of HIV infection and the change in HI titer, FcγRIIIa-V158-dimer binding, NK cell activation, and ADP activity to A/Switzerland/9715293/2013 (postvaccine minus baseline response). Results for all 27 HIV-positive subjects are shown in the graphs. Analysis was by nonparametric Spearman's correlation.