Non-structural protein 1-specific antibodies directed against Zika virus in humans mediate antibody-dependent cellular cytotoxicity

Samples

PBMCs from 12 ZIKV-immune donors were used in the study (Table 1). All ZIKV PBMCs were collected after written informed consent. The studies were approved by the institutional review boards at the University of Rhode Island, National Institute of Allergy and Infectious Diseases (NIAID) and Oregon Health & Science University (OHSU). Blood samples (Donors 1–8) of travellers were collected at the National Institutes of Health Clinical Center from patients displaying symptoms of a suspected ZIKV infection following return to the United States from areas where ZIKV was known to be circulating. Donors 9–12 were enrolled in a flavivirus immune cohort at Oregon Health & Science University [29]. PBMCs were isolated and cryopreserved prior to use. ZIKV-immune sera or plasma (n = 28) were obtained from BEI resources and OHSU. The ZIKV AccuSet™ plasma performance panel was purchased from SeraCare Life Sciences (Milford, MA).

ZIKV reporter virus particle production and neutralization assay

ZIKV (strain H/PF/2013) reporter virus particles (RVPs) were produced as previously described [30]. Briefly, HEK-293T cells were transfected with plasmids encoding a GFP-expressing WNV subgenomic replicon and the ZIKV structural genes. To determine viral titre, serial dilutions of RVP-containing supernatant were used to infect Raji cells that express the flavivirus attachment factor DC-SIGNR (Raji-DC-SIGNR), and GFP-positive infected cells were quantified by flow cytometry. For neutralization studies, ZIKV RVPs were sufficiently diluted to ensure antibody excess at informative points of the dose–response curve and mixed with serial dilutions of heat-inactivated human ZIKV convalescent sera for 1 h at 37°, followed by infection of Raji-DC-SIGNR cells. GFP-positive infected cells were quantified by flow cytometry, and results were analysed by non-linear regression analysis to estimate the dilution of sera required to inhibit 50% of infection (NT50). The reported NT50 values represent the average of two independent neutralization assays, each performed with duplicate technical replicates.

ELISAs

Approximately 2 × 10⁶ PBMCs were stimulated with 2·5 µg/ml of R848 (Invivogen, San Diego, CA, USA) and 1000 U/ml of IL-2 (PeproTech, Rocky Hill, NJ, USA) in a 48-well plate for seven days at 37° and 5% CO₂. ZIKV NS1 and VLP IgG and the four IgG subclasses (IgG1-4) were detected in the supernatants and sera by ELISA. Briefly, 96-well plates were coated overnight with 20 ng/well of ZIKV virus-like particles (Suriname Z1106033) or 50 ng/ml ZIKV NS1 (Uganda MR766) (Native Antigen Company, Oxfordshire, UK). The plates were blocked for 90 min with 1% BSA. The culture supernatant or diluted sera were added to the wells for 60 min. Plates were washed, and goat anti-human IgG coupled to HRP (A80-104P; Bethyl Laboratories, Inc., Montgomery, TX, USA) was added for total ZIKV-specific IgG detection. HRP-conjugated secondary Abs for IgG subclass detection include goat anti-human IgG1 FC (9054-05), IgG1 hinge (9052-05), IgG2 FC (9060-05), IgG3 hinge (9210-05) and IgG4 FC (9200-05) (Southern Biotechnology Associates, Birmingham, AL, USA). The assay was developed with TMB substrate (34021; Thermo Scientific, MA, USA), stopped with 1 m HCL and read at 450 nm. For each independent experiment, sera or supernatants were assayed in duplicate and the average OD values obtained presented.

FluoroSpot assay

FluoroSpot assays were performed as previously described with some modifications [31]. Briefly, IPFL plates (PUV96; CTL, Cleveland, OH, USA) were treated with 20 µl per well 70% ethanol for 1 min. Ig capture antibody (hB2F; CTL), anti-human IgG (MT91/145; Mabtech, Nacka Strand, Sweden) or anti-human IgG1 hinge (9054-05; Southern Biotech) was added to wells, and plates were incubated overnight at 4°. Following washing and blocking of the plates, MBC cultures (2 × 10⁵) were added to duplicate wells for each condition, then incubated overnight at 37° and 5% CO₂. The plates were washed and blocked with 1% BSA for 1h at 37°. Optimized concentrations of fluorescent (FL) ZIKV NS1 and ZIKV diluted in 100 µl of PBS were then added. The plates were incubated for 1h at room temperature, washed three times with PBS and two times with deionized water, and then rinsed with tap water. The plates were allowed to dry completely. To determine total IgG-secreting cells, 5 × 10⁴ MBC cultures were added to wells coated with human Ig and visualized using a polyclonal PE-labelled IgG detection Ab (hB07; CTL).

Data acquisition

The plates were scanned using an CTL ImmunoSpot S5 Analyzer. Spots representing Abs secreted by antigen-specific MBCs were detected using different filters (520/40 for DL488 ZIKV or 630/60 for DL594 ZIKV NS1). The settings for the sensitivity of spot counting were established and adjusted manually for each plate using antigen-stimulated and negative controls to exclude artefacts or background spots. The number of specific or cross-reactive spots was evaluated by the reader and verified manually [31].

ZIKV E and NS1 expression on ZIKV-infected cells

Zika virus FLR strain (GenBank: KX087102) was obtained from BEI Resources (Catalog No. NR-50183). ZIKV was propagated using the C6/36 cell line. Viral stocks were titrated by ELISPOT using Vero 76 cells. Focus-forming units (FFU/ml) were determined by immunostaining with anti-flavivirus antibody 4G2 (Millipore Catalog No. MAB10216). K562 DC-SIGN cells were infected with ZIKV at a multiplicity of infection (MOI) 0·1 and 1 for 24 h at 37° ± CO₂. The cells were then washed, resuspended in BD Cytofix/Cytoperm™ (554714; BD Biosciences, San Jose, CA) and incubated for 20 min at 4°. The cells were washed with BD Perm/Wash™ (BD 554723) and stained with mouse anti-E protein (D1-4G2-4-15) (4G2) (NBP2-52709) or mouse anti-ZIKV NS1 antibody (B4) (AbZIKVNS1-B4, NAC) for 30 min at 4°. Cells were then washed and stained with AF 488 goat anti-mouse IgG (H + L) (A-11001). The cells were washed with BD Perm/Wash™ and fixed with BD Cytofix™ (BD 554655). Samples were acquired on a BD LSR™ II flow cytometer, and the data were analysed using the FlowJo software.

Stable cell line generation

CEM-NK^R cells were obtained from the NIH AIDS Reagent Program and maintained and passaged in RPMI-1640 supplemented with 10% fetal bovine serum. Cells were transfected with linearized plasmid (pcDNA3.1) expressing a codon-optimized ZIKV (GenBank Acc# KU681081), DENV-1 (WestPac74: U88535.1); DENV-2 (S16681: NC_001474): DENV-3 (CH53849: DQ863638.1); and DENV-4 (Singapore /8976/1995: Q5UCB8.1) NS1 proteins designed for maximal surface expression. Stable transfectants were single-cell-sorted and selected to obtain high-level NS1 surface expressing clones.

Opsonization assay

ZIKV NS1 CEM-NK^R cells (1 × 10⁵) were seeded in a round-bottom 96-well plate and incubated with 1:100 and 1:1000 dilution from ZIKV-immune or ZIKV-naïve sera for 1h at 4° or 30 min at 37°. Cells were washed and stained with goat anti-human IgG-PE (SB 2040-09). The cells were then washed with BD Perm/Wash™ and fixed with BD Cytofix™ (BD 554655). Samples were acquired on a BD LSR™ II flow cytometer, and the data were analysed using the FlowJo software.

Human natural killer target visualization assay

The NK-TVA assay was performed according to the manufacturer's instructions (# NK-TVA; CTL, Cleveland, OH, USA). Effector cells (cryopreserved PBMC) were thawed, washed and rested overnight at 37° ± 5% CO₂. Target cells ZIKV NS1 CEM-NK^R cells (1 × 10⁶/ml) were labelled with 1 µl (per ml of cells) of CTL-TVA fluorescent dye and incubated for 15 min at 37°. After incubation, cells were washed and resuspended, and live cells were counted using the CTL ImmunoSpot S5 Analyzer (CTL). We added 50 µl of target cells (counted with the CTL-LDC cell-counting reagent or trypan blue) to a round-bottom 96-well plate with 50 µl of different dilutions of sera from ZIKV-naïve or ZIKV-immune donors for 1 h at 4° to allow opsonization. Effector cells (PBMC) were counted and added to the culture, and plates were incubated at 37° for 5 h. After incubation, the cells were resuspended, and 50 µl of the cells was transferred in triplicate to a flat-bottom 96-well plate. The wells were counted for fluorescent live target cells. The percentage of killing was calculated using the following formula: average count in control wells minus average count in test wells divided by average count in control wells multiplied by 100.

Intracellular cytokine staining assay

Cryopreserved PBMCs were thawed, washed and rested overnight at 37° ± 5% CO₂. CEM-NK^R-expressing ZIKV NS1 or DC-SIGN (negative control) was left in media or incubated with anti-Zika virus NS1 antibody (B4) or plasma samples at different dilutions for 1 h at 4°. PBMCs were counted, added to the culture and incubated at 37° ± 5% CO₂ in the presence of mouse anti-human CD107a (BD 555800) and CD107b (BD 555804). After 1h, BD GolgiPlug™ (BD 555029) and GolgiStop™ (BD 554724) were added to the wells and incubated at 37° ± 5% CO₂ for another 3 h. Cells were washed and stained with surface Abs anti-CD3, CD16, CD56 and CD69 for 30 min. Cells were washed, fixed with BD Cytofix™ and acquired on a BD LSR™ II flow cytometer or washed and resuspended in BD Cytofix/Cytoperm™ for 20 min at 4°. The cells were then washed with BD Perm/Wash™ and stained with anti-IFN-γ for 30 min at 4°. The cells were washed with BD Perm/Wash™ and fixed with BD Cytofix™. Samples were acquired on a BD LSR™ II flow cytometer, and the data were analysed using the FlowJo software.

Statistical analysis

All results are reported as absolute mean values plus SEM. IgG subclass distribution, against ZIKV NS1 and VLPs, was compared using the non-parametric Mann–Whitney test or Friedman test (three or more matched groups) with Dunn's multiple comparisons test. Percentage of lysis of target cells was compared using the non-parametric Kruskal–Wallis test with Dunn's multiple comparisons test. All statistical analysis was performed using the GraphPad Prism V.9.00 software (San Diego, CA), and a P value <0·05 was considered statistically significant.

ZIKV NS1- and VLP-specific Abs are predominantly of the IgG1 isotype in immune sera

We determined the magnitude of IgG responses and isotype specificity of ZIKV-specific Abs in the sera of 28 ZIKV-immune donors using ZIKV VLPs (to measure gold standard antiviral responses) and recombinant NS1 proteins in ELISAs. We found that sera from all donors had IgG Abs specific to NS1, and to prM and E proteins (present in the VLP preparation). We next used well-characterized secondary Abs (two IgG1 mAbs directed against either the hinge or the Fc portion, IgG2, IgG3 and IgG4; Southern Biotechnology) to detect the IgG isotype of ZIKV NS1- or VLP-specific Abs. The most commonly detected IgG isotype of NS1 and VLP Abs in ZIKV-immune sera was IgG1 (Figure 1a,b). Interestingly, we observed a lack of or poor recognition of ZIKV NS1- or VLP-specific Abs in some sera by one or both IgG1 secondary mAbs even though they were all clearly recognized by the polyclonal IgG secondary Abs. Therefore, we asked in a representative subset of sera whether the lack of recognition was unique to ZIKV-specific Abs or whether Abs to other viruses also followed a similar pattern. We found a poorer recognition of ZIKV NS1 or IAV NP compared with ZIKV VLP Abs in sera from the same individual when the hinge or Fc-specific secondary Ab was used (Table S1). We next determined whether the magnitude of IgG1 responses to NS1 and VLPs differed and found a stronger Ab response to VLPs overall compared with NS1 (Figure 1c,d).

ZIKV NS1-specific memory B cells elicited following natural infection

To determine the specificity and magnitude of ZIKV-specific MBCs, we stimulated PBMC from ZIKV-immune and ZIKV-naïve donors with the TLR7/8 agonist, r848, and recombinant (r)IL-2 in vitro for seven days to convert MBCs into Ab-secreting cells (ASCs). Time-points when PBMCs were collected and ZIKV neutralization titres are provided in Table 1. Stimulated cells were added to plates coated with human IgG or IgG1 (hinge) capture Abs. We used optimal concentrations of DL594-labelled ZIKV NS1 and DL488 infectious ZIKV to identify Abs secreted by ZIKV-specific MBC cultures using a FluoroSpot assay we recently developed [31]. Representative images of wells containing MBC cultures from a ZIKV-immune donor with media or total IgG (top panel), ZIKV (green) or ZIKV NS1 (red) with IgG (middle panel) or IgG1 hinge (lower panel) coating Abs are shown in Figure 2a. The frequency of ZIKV and ZIKV NS1 binding ASCs was significantly higher in ZIKV-immune donors than in ZIKV-naïve donors (Figure 2b and Table S2). Cell culture supernatants collected 7 days post-stimulation from immune PBMC also had detectable ZIKV- and NS1-specific Abs. We found a similar pattern of recognition of ZIKV VLP- and NS1-specific Abs by the hinge or Fc-specific IgG1 secondary Abs in the sera and supernatants from stimulated PBMC in the same donors (Figure 2c,d). We detected no ZIKV VLP or NS1 Abs in supernatants from stimulated cultures of PBMC from naïve donors (data not shown). Taken together, our data indicate robust frequencies of ZIKV NS1-specific MBCs and predominant secretion of IgG1 ZIKV NS1-specific Abs by MBCs in the sera of immune individuals.

ZIKV NS1 monoclonal antibody mediates ADCC

We first asked whether NS1 was expressed on virally infected cells and found expression of E and NS1 on ZIKV-infected cells (Figure 3a). To determine whether NS1 Abs could mediate ADCC, we used the CEM-NK^R cell line, which is resistant to NK cell activation in the absence of ADCC. We generated CEM-NK^R cells stably expressing ZIKV NS1 and established assays using a ZIKV NS1-specific human IgG1 mAb (Native Antigen Company). We first determined whether the ZIKV NS1 mAb could opsonize NS1 on ZIKV NS1-transfected CEM-NK^R cells, a prerequisite step for ADCC. We found a significant shift in fluorescence in the presence of the NS1 Ab (Figure 3b). We next measured NK cell activation by flow cytometry following a 4-h incubation of opsonized target cells with healthy PBMC. We found an increase in CD107 degranulation of NK cells only in the presence of the ZIKV NS1-specific human IgG1 mAb and ZIKV NS1-transfected cells (Figure 3d) but not control CEM-NK^R (Figure 3f) cells. Furthermore, we saw a clear downregulation of CD16 only in the presence of the ZIKV NS1 Ab (Figure 3d). Conditions in the absence of the NS1 mAb were no different between the ZIKV NS1 and parental cell line (Figure 3c,e).

We optimized and evaluated an image cytometry assay where the number of ZIKV NS1 CEM-NK^R target cells in the presence of the ZIKV NS1 mAb and healthy PBMC as a source of NK cells was measured (Figure S1). We varied the number of ZIKV NS1 target cells (Figure S1A) and the number of effector PBMC (Figure S1B) and evaluated several healthy PBMC for their ability to lyse target cells (Figure S1C). Counts of triplicate wells in the presence of control IgG or ZIKV NS1 mAb are provided (Figure S1D). Based on the loss of calcein dye in dead/dying target cells, we used an established formula to measure % killing in the calcein assay using the average count of target cells in triplicate experimental wells and negative control wells [32]. A decrease in the number of target cells in the presence of the ZIKV NS1 mAb compared with a control IgG Ab was consistently detected (Figure S1).

ZIKV-immune sera induces NK cell activation in the presence of ZIKV NS1 target cells

As polyclonal sera contain Abs of varying concentration and function, we next evaluated a performance panel of ZIKV-immune sera (SeraCare technologies) for their ability to opsonize CEM-NK^R ZIKV NS1 and target cells expressing NS1 from the 4 related DENV serotypes. We found a strong shift in fluorescence with CEM-NK^R ZIKV NS1 and a more moderate shift with the DENV 1–4 NS1 cell lines in the presence of different dilutions of immune sera on DENV-transfected cells but not untransfected NK^R cells (Figure S2). We optimized conditions to assess NK cell activation in the presence of ZIKV-immune or ZIKV-naïve sera, healthy PBMC as a source of NK cells and CEM-NK^R ZIKV NS1 target cells. NK cells in PBMC were analysed for CD107a degranulation and IFN-γ secretion. We detected NK cells expressing IFN-γ, CD107a and a smaller percentage of NK cells that co-expressed CD107a and IFN-γ when ZIKV-immune sera were added to the assay (Figure 4). Our data indicate that ZIKV NS1 Abs can induce NK cell activation and degranulation.

ZIKV-immune sera induce lysis of CEM-NKR target cells expressing ZIKV NS1

We finally evaluated the ability of ZIKV-immune sera to opsonize CEM-NK^R ZIKV NS1 cells. A representative subset of immune sera shown in Figure 5a clearly show a shift in fluorescence compared with ZIKV-naïve sera. We evaluated the ability of ZIKV-immune sera to induce target cell lysis using the NK-TVA image cytometry assay. Using an optimized number of target cells, effector PBMC and different dilutions on naïve (n = 5), positive control ZIKV-immune (n = 5) and experimental sera (n = 12), we measured the loss of target cells at the single-cell level and calculated the percentage of lysed target cells. Representative images of wells containing ethanol that lyses all cells (positive control), control Ab and the ZIKV NS1 mAb (Figure 5b) are shown. Our results indicate lysis of target cells predominantly in the presence of ZIKV-immune sera over a range of dilutions (Figure 5c). Taken together, our data demonstrate durable circulating ZIKV NS1 MBC frequencies and serum IgG1 Abs in immune individuals that can utilize ADCC to mediate antiviral function.