MAbs 78/2 and 69/1 Enhance Lung Pathology Following H3N2 Viral Challenge.

To evaluate the potential of antibodies to mediate ADE of influenza disease in vivo following influenza viral challenge, two H3N2-specific MAbs were evaluated in mice. MAb 78/2 binds to the globular head domain of H3N2 hemagglutinin and neutralizes virus infection at high concentration while increasing the number of virus-infected cells at lower concentrations in vitro. MAb 69/1 binds to a different site on the globular head domain, distant to the receptor binding domain, does not neutralize virus at any dose, and increases H3N2 virus infection of the P388D₁ cell line at lower antibody concentrations (16).

To test the impact of these MAbs on H3N2 virus infection in vivo, BALB/c mice were prophylactically treated with 15, 1.5, or 0.15 mg/kg of MAb 78/2 or 69/1 or mock treated with PBS followed by challenge with 10⁴ TCID₅₀ of A/Hong Kong/1/68 (H3N2) virus intranasally, 24 h after MAb treatment (Fig. 1A). The concentrations of MAbs were measured at 24 h postinjection in the sera and lungs on the day of viral challenge (at day 0) using biotin-labeled antibodies. Antibody concentrations were similar for the two MAbs in sera and lung tissues (SI Appendix, Table S1). Weight loss in the mock PBS control animals on days 2 to 5 following viral challenge was modest (average of 5%), and most animals started to gain weight by day 6. MAbs 78/2 or 69/1 treatment did not impact body weight loss after H3N2 viral challenge compared with the mock PBS-treated control group at any pretreatment dose (Fig. 1 B and C). Animals were killed on days 2 and 6 and plaque assays using Madin-Darby canine kidney (MDCK) cells were performed to measure viral load in the mouse lung homogenates. MAb 69/1 did not significantly impact lung viral load at any dose level compared with the mock-treated group on 2 d postinfection (dpi) (Fig. 1D). At 6 dpi, the average lung viral load in 69/1-treated animals (medium and low doses) was 2- to 3-fold higher compared with the PBS control group, but did not reach statistical significance (Fig. 1E). In contrast, MAb 78/2 had a dose-dependent effect on lung viral loads. Lungs from mice pretreated with the highest dose of MAb 78/2 (15 mg/kg) showed significant reduction in lung viral titers compared with the mock-treated PBS control group on day 2 (P < 0.05) (Fig. 1F) and on 6 dpi (P < 0.0005) (Fig. 1G), while mice treated with the lowest dose of MAb 78/2 (0.15 mg/kg) had significantly higher viral lung titers compared with the mock-treated group on day 6 dpi (P < 0.005) (Fig. 1G). Therefore, MAb 78/2 exhibited a dose-dependent reduction or enhancement of viral load in vivo.

• Download figure
• Open in new tab
• Download powerpoint

Fig. 1.

Experimental plan, mouse weight loss, and lung viral titers following prophylactic treatment with MAb 78/2 or 69/1 and challenge with A/Hong Kong/1/68 (H3N2) virus. (A) Mice were prophylactically treated intraperitoneally (IP) with the indicated antibody at the specified dose 24 h before intranasal challenge with 10⁴/10 µL TCID₅₀ of A/Hong Kong/1/68 (H3N2) virus. Mice were killed, and blood and lungs were collected on days 2 and 6 postchallenge. (B and C) The mice were weighed every day for 6 d. Mouse weights were normalized to day 0 weight (treated as 100%) and plotted as percent weight change relative to day 0 weight. (D–G) Mouse lungs treated with MAb 69/1 (D and E) or MAb 78/2 (F and G) from days 2 (D and F) and 6 (E and G) postchallenge were homogenized and used in a plaque assay to determine viral load in the lungs. Statistical significance among groups was calculated using Abnova, P < 0.001 (***), P < 0.01 (**), P < 0.05 (*). n/a, not applicable.

In addition to viral load measurements, lung immunohistochemistry was performed on day 2 post-H3N2 virus challenge (SI Appendix, Fig. S1). Influenza nucleoprotein (NP) positive cells were observed in bronchial epithelial cells in all infected lungs. Furthermore, influenza NP positive cells were observed at higher frequency in alveolar cells (arrowhead) of animals treated with MAb 69/1 and MAb 78/2 (at 1.5 mg/kg and 0.15 mg/kg), suggesting increased localized viral foci in small airways in the lungs of MAb-pretreated animals (SI Appendix, Fig. S1). Virus-infected cells were stained using rabbit anti-NP antibodies and lung macrophages (CD11c⁺) on 2 dpi (SI Appendix, Fig. S2). The majority of infected cells were bronchioalveolar epithelial cells, and not macrophages (CD11c⁻). However, accumulation of infiltrating macrophages near the virus-infected cells was observed in the MAb-pretreated animals compared with PBS-treated controls, especially for the MAb 69/1 middle dose (1.5 mg/mL) and MAb 78/2 lowest dose (0.15 mg/mL) (SI Appendix, Fig. S2). On both days, mouse H&E-stained lungs were blindly evaluated by a certified veterinary pathologist using 5 parameters: whole lung pathology, bronchioles-hyperplasia, bronchioles-peribronchiolitis (lymphocyte cuffs), perivasculitis, and alveolar inflammation (interstitial pneumonia). On day 2 following viral challenge, the lung pathology was minimal, with lesion scores between 0 and 2 for all groups. However, on day 6, the combined histopathological scores of mice treated with MAb 69/1 at the medium dose (1.5 mg/kg) exhibited enhanced lung pathology compared with the PBS control group (P < 0.05) (Fig. 2). Mice treated with MAb 78/2 demonstrated a dose-dependent impact on lung pathology following H3N2 influenza virus challenge. Lungs from animals treated with the high dose (15 mg/kg) had lower lesion scores, while the mice pretreated with the lowest dose of MAb 78/2 (0.15 mg/kg) had more cellular infiltration and significantly higher lung lesion scores (P < 0.05) compared with the mock-treated group (Fig. 2). Neither of the MAbs 78/2 or 69/1 significantly altered C3 complement levels in the mouse lungs either at day 2 or day 6 following H3N2 challenge at any MAb dosage level (SI Appendix, Fig. S3).

• Download figure
• Open in new tab
• Download powerpoint

Fig. 2.

Histopathology analysis of the lungs of mice treated with MAbs 78/2 or 69/1 before H3N2 challenge. Mice were killed at day 6 post-H3N2 viral challenge and the lungs were collected for histopathological analysis. The lungs were fixed in 10% neutral-buffered formalin, embedded in parafilm, sectioned, and stained with H&E. (A) Representative combined images of H&E-stained lungs infected with H3N2 virus treated with either 69/1 or 78/2 antibodies. Lung treated with 1.5 mg/kg of MAb 69/1 and 0.15 mg/kg MAb 78/2 showed more pathological changes (bronchitis, bronchiolitis, and alveolitis) compared with other groups (scale bar indicates 250 μm). (B) Histopathological evaluation of lung was based on combined scores of 5 parameter of whole lung, bronchioles-hyperplasia, bronchioles-peribronchiolitis (lymphocyte cuffs), perivasculitis (vascular cuffs), and alveolar inflammation (interstitial pneumonia). 0, no lesions; 1, minimal lesions; 2, mild lesions; 3, moderate lesions; 4, severe lesions. Median values are shown in the box and whisker plots for each group. The box extends from the 25th to the 75th percentile and the error bars represent the lowest and highest values. Statistical analyses were performed using the Bonferroni multiple comparison test using ANOVA, P < 0.01 (**), P < 0.05 (*).

Impact of MAb Treatment on Lung Cytokine/Chemokine Levels Post-H3N2 Virus Challenge.

Local inflammatory responses can be a driver of cellular infiltration and alveolar epithelial cell damage in lungs. Therefore, cytokine and chemokine levels were measured in the mouse lung homogenates following H3N2 viral challenge at 2 and 6 dpi using a cytokine/chemokine protein array (SI Appendix, Table S2 and Fig. S4). In lungs from high-dose MAb 69/1-pretreated animals (15 mg/kg), significantly higher levels of IL-3, IL-12 (p70), IL-13, and IFNγ were measured, while eotaxin was elevated only in mice treated with the medium dose of MAb 69/1 (1.5 mg/kg) at 6 dpi. Elevated levels of GM-CSF and TNFα were measured in all MAb 69/1-treated animals, irrespective of dose. Both IL-13 and eotaxin are known chemokines for eosinophils. Interestingly, elevated levels of IL-13 mRNA and proteins were previously shown to be involved in chronic lung disease after influenza infection at sites of active viral RNA remnants (18). Lungs from mice prophylactically treated with the lowest dose of MAb 78/2 (0.15 mg/kg) showed significantly higher levels of the proinflammatory cytokine IL-1β, several Th2 cytokines (IL-5 and IL-10), and two chemokines (MIP-1α and MIP-1β) (SI Appendix, Table S2 and Fig. S4). IL-2 and IL-12 (p40) levels were significantly lower in lungs of animals treated with either MAb compared with the uninfected or mock PBS-treated controls.

Together, the animal studies identified both antibody-specific and dose-dependent patterns of elevated lung cytokine/chemokines and enhanced respiratory disease in animals pretreated with MAbs 78/2 and 69/1 following H3N2 virus challenge.

MAbs 78/2 and 69/1 Increase pH-Dependent HA Trypsin Sensitivity.

Following the findings of enhanced lung pathology after influenza virus challenge of animals pretreated with MAbs 78/2 and 69/1, we evaluated the direct effects of the MAbs on HA conformation and function in vitro. Susceptibility to trypsin cleavage was used as a correlate of pH-induced changes in HA conformation (Fig. 3 A and B and SI Appendix, Fig. S5). Uncleaved H3-HA0 was mixed with MAbs 78/2, 69/1, or CR8020, and incubated at 37 °C for 1 h at pHs ranging from 5.5 to 6.5 (Fig. 3A). At the lowest pH tested (pH 5.5), 80% of H3-HA0 was cleaved by trypsin. Pretreatment with the broadly neutralizing stem-targeting MAb CR8020 (19) blocked the pH-induced conformational change and protected the H3-HA0 from trypsin cleavage (Fig. 3A, Bottom Lane and Fig. 3B purple bars). In contrast, MAbs 78/2 and 69/1 increased the pH sensitivity of H3-HA0, resulting in more trypsin cleavage at higher pH (5.75 and 6.0) compared with no antibody control (Fig. 3 A and B; open vs. green and red bars).

• Download figure
• Open in new tab
• Download powerpoint

Fig. 3.

MAb 69/1 and low dose of 78/2 promote disordering and flexibility of the H3N2 HA domain. (A and B) MAbs 69/1 and 78/2 increase HA pH sensitivity to trypsin digestion. (A) H3 Aichi HA was incubated with or without the indicated antibody at the specified pH at 37 °C for 1 h. The samples were neutralized, and TPCK-trypsin was added at a 1:1 trypsin to HA mass ratio, the samples were incubated at 22 °C for 2 h, trypsin inhibitor was added to stop the reaction. Samples were resolved by nonreducing SDS/PAGE and subjected to Western blot analysis. (B) Densitometry analysis of Western blots of A from 5 independent experiments. Bars indicate percent change between sample at indicated pH with trypsin and sample at pH 6.5 with trypsin. SDs are shown as whiskers. Statistical significance compared with H3 only group was calculated using the Bonferroni multiple comparison method, P < 0.0001 (****), P < 0.01 (**). (C–F) Negative stain EM of MAbs 69/1 and 78/2 bound to HA. Fabs of 69/1 (C) or 78/2 (D) were combined with H3 Aichi HA (H3N2) (A/Aichi/2/1968) at 8 times molar excess of Fab and incubated overnight at 4 °C. The Fab/HA trimer complex was purified with a S200i SEC column and added to 400-mesh copper grids then stained with 2% uranyl formate followed by micrographs using Appion. Particles were selected using DoGPick and 2D classes were produced. (C) Negative stain 2D classes of 69/1 Fab in complex H3 Aichi HA (H3N2) (A/Aichi/2/1968). (D) 2D classes of 78/2 Fab in complex with H3 Aichi HA. Highlighted boxes are featured classes in F. (E) Unliganded H3 Aichi HA from the same imaging session as 78/2. (F) H3 Aichi HA model (PDB 3VUN) illustrating the head and stem domains of HA (Left). Image false coloring was done with Photoshop. Selected 2D class averages of unliganded, 69/1 (blue), and 78/2 (purple) Fabs in complex with H3 (yellow). Comparison of the unbound HA to the antibody-bound HA reveals that the stem region of the antibody-bound HA appears as a lighter, diffuse density. This weaker density indicates increased flexibility or disorder of HA stem.

MAbs 78/2 and 69/1 Binding to HA Globular Head Domain Induces Disordering of HA Stem Domain.

To structurally visualize the impact of MAbs 78/2 and 69/1 on HA conformation, 78/2 and 69/1 Fabs were combined with H3-HA0 and incubated overnight at 4 °C. Negative stain electron microscopy images were acquired as described (20). Both antibodies bind to the HA with 1 or 2 Fabs per HA trimer as observed in the 2D class averages (SI Appendix, Fig. S6). The two antibodies, 69/1 and 78/2, bind to distinct epitopes on the HA and at different angles of approach. MAb 78/2 binds closer to the HA receptor binding site (Fig. 3 D and F shown in pink), while MAb 69/1 Fab binds lower on the HA, at the base of the head and close to the stem domain (Fig. 3 C and F shown in blue). Surprisingly, after binding of the Fab of either MAb the HA stem region becomes less ordered relative to unbound HA trimer, as suggested by the more diffused density in the 2D classes (Fig. 3 C and D vs. Fig. 3E).

MAbs 78/2 and 69/1 Enhance H3N2 Virus Fusion Kinetics.

To further understand the impact of antibody-induced HA conformational changes on influenza virus infection, a virus fusion kinetics assay was developed with pHrodo-labeled A/Hong Kong/1/1968 (H3N2) virions to track the virus in the endocytic pathway of infected cells. This assay follows individual pHrodo-labeled virions in the endosome/lysosome compartment of MDCK cells (SI Appendix, Figs. S7 and S8). The pH sensitivity of the pHrodo fluorescence was demonstrated by using a broad pH range. The increase in fluorescent units was clearly pH dependent and increased significantly between pH 6.0 and pH 4.0 (SI Appendix, Fig. S7A). Furthermore, the addition of increasing concentrations of ammonium chloride resulted in deacidification of endosomes and reduction of fluorescent influenza virus particles (SI Appendix, Fig. S7B). This assay was reproducible with similar kinetics, as measured by number of pHrodo particles, in multiple experiments (SI Appendix, Fig. S8). Importantly, fusion of the virus with host cell membranes in the endocytic compartment resulted in the loss of fluorescence (Movie S1). To confirm the pHrodo signals are associated with virions in the endocytic compartments, the MDCK cells were immunostained with an antiinfluenza NP antibody after infection with pHrodo green-labeled H3N2 virus. The green pHrodo fluorescence signals were associated with the influenza virus as confirmed by anti-NP staining (SI Appendix, Fig. S9). In contrast, in the presence of chloroquine, which blocks endosome acidification, the virions were stained with anti-NP but the green fluorescent signal of the pHrodo-labeled virus was absent due to increase in the endosomal pH (SI Appendix, Fig. S9, Lower). Staining with antibodies against Lamp1, an endosomal/lysosome pathway marker, further confirmed the localization of the internalized virions in the low pH endocytic pathway (SI Appendix, Fig. S10). MAb 78/2 or MAb 69/1 binding to H3N2 virus in the endocytic compartment was confirmed by colocalization of MAbs and labeled virions in the same compartment in infected MDCK cells (SI Appendix, Fig. S11). Addition of MAb 78/2 at 100 μg/mL resulted in a delay in the accumulation of virions in MDCK cells and a reduction in number of virions per cell in agreement with the neutralization activity of this MAb at high doses (SI Appendix, Fig. S12).

To further understand the impact of the MAbs on virus fusion kinetics, over 100 virions per group were imaged over a period of 68 min to determine particle residence time within the endosome/lysosome compartment in absence or presence of the MAbs at 37 °C (Fig. 4). In the absence of antibodies (PBS), the residence time of H3N2 influenza virus within the endosome/lysosome compartment averaged 26 min (range of 11 to 72 min) in MDCK cells (Fig. 4 A and K, black). Residence time of other influenza viral strains, A/California/07/2009 (H1N1pdm09) and B/Brisbane/60/2008 were not significantly different from H3N2 A/HongKong/1/68 (SI Appendix, Fig. S13).

• Download figure
• Open in new tab
• Download powerpoint

Fig. 4.

MAbs 78/2 and 69/1 promote fusion kinetics of A/HongKong/1/68 (H3N2) virus in MDCK cells at 37 °C. (A) Residence time distributions of pHrodo-virus particles in no antibody treatment control. The majority of particles show residence time between 10 and 30 min. (B–D) Residence time distributions of palivizumab-treated viral particles is similar among different doses or compared with PBS control, indicating similar fusion kinetics. (E–G) Residence time distributions of pHrodo-virus particles treated with different doses of MAb 78/2. Low dose of MAb 78/2-treated virus shows shifting of the residency time to <20 min, indicating low dose MAb 78/2 accelerates the virus fusion kinetics. (H–J) Residence time distributions of pHrodo-virus particles treated with different doses of MAb 69/1. All doses tested show significant shifting of the viral residency time to <20 min, indicating MAb 69/1 promotes faster virus fusion kinetics. At least three biological replicates were done for PBS control and experimental groups and the representative graphs are shown. (K) Residency time distribution of all samples. Median residency times are shown in black lines with 95% CI. One-way ANOVA between PBS and each treatment group was calculated. **: P < 0.01; ***: P < 0.001; ****: P < 0.0001.

The anti-RSV F MAb, palivizumab, used as a negative control (21, 22), did not significantly impact virus residence time of the pHrodo-labeled H3N2 virus particles in the endocytic pathway at any concentration tested (Fig. 4 B–D and K, blue). When MAb 78/2 was added at 10, 1.0, and 0.1 μg/mL, the high and medium concentrations did not change the endosome/lysosome virus residence time, but the lowest MAb concentration (0.1 µg/mL) reduced H3N2 virus residence time to an average of 18 min, suggesting faster virus fusion kinetics (Fig. 4 E–G and K, pink). Surprisingly, MAb 69/1 significantly reduced the residence time of H3N2 pHrodo-labeled virus particles in the endocytic pathway at all three antibody concentrations (average 12 to 16 min), suggesting faster virus fusion kinetics in the presence of this MAb (Fig. 4 H–J and K, green). These studies were repeated comparing the kinetics of pHrodo-labeled H3N2 virus particle accumulation and residence time in MDCK cells at 33 °C (representing upper respiratory tract temperature) vs. 37 °C (representing lower respiratory tract temperature). The virion internalization rate in the endosomes was somewhat slower at 33 °C compared with 37 °C (SI Appendix, Fig. S14A), but the average residence time in the endocytic compartment was not significantly different at the two temperatures for H3N2 virus (SI Appendix, Fig. S14B). At 33 °C, both MAb 78/2 and MAb 69/1, at 1 and 0.1 μg/mL, significantly shortened the residence time of the pHrodo-labeled H3N2 virus particles, suggesting enhanced fusion kinetics (SI Appendix, Fig. S15).