Development of a human papillomavirus (HPV) multiplex immunoassay to profile HPV antibodies

Neutralizing antibodies (NAbs) are considered the primary mechanism of vaccine‐mediated protection against human papillomaviruses (HPV), the causative agent of cervical cancer. However, the minimum level of NAb needed for protection is currently unknown. The HPV pseudovirion‐based neutralization assay (PBNA) is the gold standard method for assessing HPV antibody responses but is time‐consuming and labor‐intensive. With the development of higher valency HPV vaccines, alternative serological assays with the capacity for multiplexing would improve efficiency and output. Here we describe a multiplex bead‐based immunoassay to characterize the antibody responses to the seven oncogenic HPV types (HPV16/18/31/33/45/52/58) contained in the current licensed nonavalent HPV vaccine. This assay can measure antibody isotypes and subclasses (total IgG, IgM, IgA1–2, IgG1–4), and can be adapted to measure other antibody features (e.g., Fc receptors) that contribute to vaccine immunity. When tested with serum samples from unvaccinated and vaccinated individuals, we found high concordance between HPV‐specific IgG using this multiplex assay and NAbs measured with PBNA. Overall, this assay is high‐throughput, sample‐sparing, and time‐saving, providing an alternative to existing assays for the measurement and characterization of HPV antibody responses.

Although a serological correlate of protection has not been established, HPV-specific neutralizing antibodies (NAbs) induced by vaccination are thought to be the main mechanism of protection against HPV infection.Three serological assays have mainly been used to assess HPV antibody responses: pseudovirion-based neutralization assay (PBNA), enzyme-linked immunosorbent assay (ELISA), and competitive Luminex immunoassay (cLIA).The PBNA can measure antibodies that neutralize HPV L1/L2 pseudovirions that express a reporter gene in vitro.Although PBNA is the gold standard for HPV serology, it is low-throughput, laborand time-intensive. 9cLIA involves the use of labeled monoclonal antibodies that compete with serum antibodies for binding to a single conformational epitope on intact L1 HPV virus-like particles (VLPs), and therefore may underestimate the level of HPV-specific antibodies. 10milar to the cLIA, ELISA uses intact VLPs to detect serum antibodies.
However, ELISAs only measure antibodies to one HPV type at a time and can overestimate the HPV antibody response due to detecting both NAb and non-NAbs.All three assays generally correlate well with each other, although comparisons between assays have been complicated by inherent differences in what antibodies are measured (i.e., neutralizing vs. non-neutralizing) and a lack of a uniform serological protective cutoff value. 11re we describe a multiplex bead-based immunoassay that measures HPV-specific total IgG responses to seven oncogenic HPV genotypes covered by the 9vHPV vaccine, specifically HPV16, HPV18, HPV31, HPV33, HPV45, HPV52, and HPV58.This assay allows simultaneous detection of antibodies to different HPV VLPs in a single well with a lower sample volume requirement, providing a highthroughput method of evaluating HPV antibody responses following vaccination.Furthermore, this assay can also be adapted to measure different antibody isotypes (IgG, IgM) and subclasses (IgA1, IgA2, IgG1-4) to comprehensively profile HPV vaccine antibody responses.This assay was validated with serum samples from individuals who were vaccinated with 4vHPV followed by 2vHPV 6 years later, as well as a serum standard from individuals vaccinated with 9vHPV.

| MATERIALS AND METHODS
HPV VLPs were produced, purified, and conjugated to carboxylated magnetic beads for use in the multiplex assay.All reagents and equipment used to set up the multiplex assay are listed in Supporting Information S1: Tables 1 and 2, respectively.

| Production of HPV VLPs
HPV L1/L2 plasmid DNA was extracted from an Escherichia coli stab obtained from Addgene, as described previously. 12 HPV VLPs were purified using iodixanol density gradient ultracentrifugation, yielding 11 fractions.The concentration of each fraction was quantified using a bicinchoninic acid (BCA) protein assay (Thermo Fisher Scientific).Sodium dodecyl-sulfate polyacrylamide gel electrophoresis (SDS-PAGE) was then used to determine the fraction that contained clarified HPV VLPs, indicated by a band at 55-60 kDa.

| Conjugation of HPV VLPs to carboxylated microspheres
Each of the seven different HPV L1/L2 VLPs (HPV16/18/31/33/ 45/52/58) was covalently conjugated to spectrally unique magnetic microspheres using a two-step sulfo-N-hydroxysulfosuccinimide T A B L E 1 List of buffers used for coupling HPV VLP to carboxylated magnetic microspheres.(NHS)/1-ethyl-3-(3-dimethylaminopropyl) carbodiimide (EDC) coupling method, as described in Brown et al. 14 Commercial tetanus toxoid protein (AJ Vaccines) and bovine papillomavirus (BPV) VLP produced in-house were also conjugated to beads and were included as positive and negative controls, respectively.The buffers used in this protocol are described below in Table 1.A total of 6.25 × 10 6 microspheres for each HPV type were activated by resuspension in 400 µL of activation buffer, as well as 50 µL of each of 50 mg/mL sulfo-NHS and EDC for 30 min at room temperature (RT) on a plate shaker at 800-900 rpm.The activated microspheres were washed twice with 2-(N-morpholino) ethanesulfonic acid (MES) coupling buffer and conjugated with 25 µg of HPV VLP for 2-3 h at RT.To block the remaining carboxyl sites, the conjugated microspheres were resuspended in 1 mL of phosphate-buffered saline (PBS)-0.02%Tween-0.1%BSA-0.05%sodium azide blocking buffer for 30 min.For storage, the HPV VLP-conjugated microspheres were resuspended in 625 µL of 1× PBS, 0.05% sodium azide storage buffer and stored in the dark at 2-8°C, for up to 3 months.
Microsphere concentration was confirmed using a haemocytometer before use in the multiplex immunoassay.(Bio-Rad) using Bio-Plex Manager 6.0 software.Doublet discrimination gates were set between 5000 and 25 000 to exclude bead clusters from fluorescence intensity measurements, and 50 beads per bead region were analyzed per sample.

| Study samples
The serum samples used in this manuscript were derived from a prospective cohort study described previously, which includes individuals who were unvaccinated or previously vaccinated with 4vHPV and subsequently boosted with 2vHPV. 15

| Statistical analysis
The abundance of each HPV genotype-specific antibody isotype and subclass in serum was quantified as MFI, averaged across duplicates, and corrected for background fluorescence by subtracting the MFI of blank wells containing assay buffer only.Antibody isotype and subclass responses were reported in MFI corrected for background fluorescence, with the exception of total IgG which was reported in LU/mL.Total IgG responses were assessed for correlation with NAb responses determined previously by PBNA using the Spearman's rank test.All statistical analyses were performed using GraphPad Prism 9.0 (GraphPad Software), with a p < 0.05 considered to be statistically significant.

| Optimization and validation of HPV multiplex immunoassay
We tested two different concentrations of the HPV16-conjugated microspheres (20 or 50 microspheres/μL) as well as two concentrations of the secondary anti-IgG antibody (1.3 or 0.13 μg/mL).A lower microsphere concentration of 20 microspheres/μL was not significantly different to 50 microspheres/μL and hence was selected for this assay (Supporting Information S1: Figure 1).Similarly, a secondary antibody concentration of 1.3 μg/mL resulted in significantly higher IgG, IgA1, and IgA2 MFI (Supporting Information S1: Figure 1).We also verified that multiplexing did not affect immunoreactivity of the majority of conjugated microspheres (Supporting Information S1: Figure 2).
To validate the HPV multiplex assay, we measured HPV antibodies to the seven HPV types using a pooled serum from 9vHPV-vaccinated individuals.The eight-point standard curve for each HPV type showed good linearity over the concentration range of assigned 0.6-10 000 LU/mL (Figure 2).

| Measurement of HPV antibody isotypes and subclasses
To confirm that the HPV multiplex assay can be used to assess different HPV antibody isotypes and subclasses, we measured eight antibody isotypes and subclasses (IgG, IgM, IgA1-2, IgG1-4) to HPV16 as a proof of concept, using samples from unvaccinated and vaccinated individuals (Figure 3).As expected, there was a clear distinction between unvaccinated and vaccinated individuals for all antigen-specific isotypes and subclasses, except IgG4 which is low in general for all samples.The majority of total IgG responses were made up of IgG1 and IgG3, and to a lower extent IgG2 and IgG4.
Moderately strong IgM, IgA1, and IgA2 responses were observed in vaccinated individuals.

| Correlation analysis between HPV multiplex assay and HPV PBNA
We measured the serum IgG antibody levels by the multiplex assay using serum samples from our cohort study (individuals who were unvaccinated or previously vaccinated with 4vHPV and boosted with 2vHPV) (n = 80).The overall coefficient of variation across eight independent assays of the 9vHPV standard at the linear range of the curve ranged from 10.7% to 32.9% across the seven HPV types (Supporting Information S1: Figure 3).

| DISCUSSION
We developed a high-throughput bead-based multiplex immunoassay that is robust and strongly correlates with the gold-standard PBNA for the measurement of HPV vaccine-induced antibody responses.This assay is sample-sparing and less time-consuming compared to PBNA and can be used to measure HPV antibodies to seven oncogenic HPV types covered in the 9vHPV vaccine simultaneously, with the possibility of measuring a much broader range of HPV types (>10 HPV types) if required.The assay we described measures eight antibody subclasses and isotypes, but can be adapted to measure other antibody features that contribute to effector function such as complement C1q binding, antibody avidity, glycosylation patterns, and Fc receptor binding. 19,20th increasing HPV vaccine valency and immunobridging studies to evaluate new HPV vaccines and vaccination schedules, robust and high-throughput serological assays (i.e., multiplexing) will be crucial.A 9vHPV is licensed for use globally and a 14-valent HPV vaccine developed by SinoCellTech is currently being evaluated in clinical trials. 21A multiplex HPV serological assay will help to improve efficiency and output for multi-valency vaccine evaluation. 22though NAbs induced by vaccination are generally considered to be the primary mechanism of protection against infection,  23 Although other HPV serological multiplex assays have been developed previously, none have tested for antibody isotype and subclass measurements and some were based on a different system (i.e., Meso Scale Discovery electrochemiluminescence). 24,25 Furthermore, we were also able to demonstrate moderate to strong correlation of multiplex IgG data with NAbs for up to seven vaccine oncogenic HPV genotypes, verifying the robustness of our assay.
There are different immunological assays to measure HPV antibodies and their use is dependent on the resources available to establish the assays.As such, standardized measurement using serological standards and the reporting of immunogenicity using International Units are crucial, so that data from clinical studies can be compared. 26In addition to the international serum standard for Briefly, bacterial stabs containing the respective HPV plasmid were cultured on antibioticresistant Luria-Bertani (LB) agar overnight.A single colony was then inoculated in the LB broth with antibiotics and incubated overnight at 37°C.HPV plasmid DNA was extracted using the QIAGEN Maxiprep kit and quantified using the NanoDrop 2000 spectrophotometer.HPV L1/L2 VLPs (HPV16/18/31/33/45/52/58) were produced via transfection of a human embryonic kidney cell line (HEK293TT) with the extracted HPV plasmid DNA, using a protocol from the Frederick National Laboratory for Cancer Research. 13HEK293TT cells were seeded at a density of 21 × 10 6 cells per T225 flask overnight.The following day, 11.25 mL of transfection cocktail (11.25 mL Opti-MEM, 247.5 μL Lipofectamine 2000, 112.5 μg plasmid DNA) was added to each T225 flask and incubated for 5-6 h at 37°C, 5% CO 2 .Media was replaced with D10 media (Dulbecco's modified Eagle medium containing 10% heat inactivatedfetal bovine serum, 1% MEM-non-essential amino acid, 1% Gluta-MAX, 1% penicillin-streptomycin-glutamine) and incubated for 48 ± 2 h from the time of transfection.Following the 2-day incubation, HPV VLPs were harvested by detaching HEK293TT cells with 4 mL of 0.25% trypsin/EDTA solution at 37°C, 5% CO 2 .Cells were collected and centrifuged twice for 10 min at 300g, 20°C, resuspending in 5 mL of Dulbecco's phosphate buffered saline (DPBS) each time.To lyse the transfected cells, a lysis buffer (10 mM DPBS-MgCl 2 , 10% Brij 58, Benzonase Nuclease, ATP-dependent DNase, 1 M ammonium sulfate) was added at 1.5 times the final cell pellet volume and incubated in a 37°C water bath for 22-26 h, followed by 175 µL of 5 M NaCl per 1 mL aliquot for long-term storage at −80°C.

2. 3 |
HPV multiplex immunoassay to measure antibody isotypes/subclasses An overview of the multiplex assay is illustrated in Figure1.A total of seven different HPV VLP-coupled microspheres (HPV16/18/31/33/45/ 52/58) and two control protein-coupled microspheres (BPV and tetanus toxoid) were each diluted to a final concentration of 20 microspheres/µL in assay buffer (1× PBS, 0.1% bovine serum albumin).Fifty microliters of the microsphere mixture was added to each well of a 96-well optical bottom plate, along with 50 µL of serum samples diluted 1:50 or 1:100 in assay buffer and assay controls (samples known to have low or high typespecific NAb levels), then incubated overnight at 2-8°C on a plate shaker at 800-900 rpm.To minimize microsphere loss, the plate was centrifuged at 2680g for 3 min at RT the following day and washed twice with assay buffer on a magnetic plate washer.Fifty microliters of R-phycoerythrinconjugated mouse antihuman secondary antibody (total IgG, IgM, or IgA1) diluted to 1.3µg/mL in assay buffer was added to each well and incubated at RT for 2 h.The plate was washed twice followed by resuspension in 100 µL of Luminex sheath fluid on a plate shaker for 10 min.Quantitative analysis was performed on the Bio-Plex 200 F I G U R E 1 Schematic of the human papillomavirus (HPV) virus-like particle (VLP) multiplex immunoassay.HPV VLPs were produced by transfection of HEK293TT cells with HPV genotype-specific L1/L2 plasmids and purified by density gradient ultracentrifugation.The concentration and purity of HPV VLPs was confirmed using a commercial BCA assay and sodium dodecyl-sulfate polyacrylamide gel electrophoresis (SDS-PAGE) (band at 55-60 kDa), respectively.Purified HPV VLPs were conjugated to magnetic carboxylated microspheres through a two-step carbodiimide process for use in the multiplex immunoassay.HPV genotype-specific antibody responses can be measured in multiplex by incubating serum with HPV VLP-coupled microspheres of different specificities overnight, then with an isotype-specific secondary antibody before data collection and analysis on the Bio-Plex ® 200 (Bio-Rad).
These samples were collected from individuals living in the Greater Suva area of Fiji (Suva, Nausori, and Valevu) between February and March 2015.Samples from unvaccinated (n = 1) or vaccinated individuals (n = 6) with a range of HPV16 and HPV18 NAb titers were used for assay optimization and analysis.Additionally, 80 serum samples were used to validate the assay, including correlation of IgG levels with neutralizing titers from PBNA.All samples were identified by a unique study number and measured in a blinded manner.Additionally, a pooled serum from 9vHPV-vaccinated individuals (kindly provided by Troy Kemp, Frederick National Laboratory) was used to validate the multiplex assay and also used as a standard for HPV16, HPV18, HPV31, HPV33, HPV45, HPV52, and HPV58-specific IgG, to allow a readout in Luminex units/milliliter (LU/mL).IgG median fluorescence intensity (MFI) was converted to LU/mL by interpolation from a 5-parameter logistic regression curve with a 70%-130% recovery range.
antibodies are capable of many other functions apart from neutralization.Antibodies can interact with Fc receptors on immune effector cells or other parts of the immune system (e.g., complement F I G U R E 4 Spearman's rank correlations between total IgG antibody levels (LU/mL) determined by human papillomavirus (HPV) multiplex assay and neutralizing antibody titers (NAb) (ED50) determined by HPV pseudovirion-based neutralization assay using serum samples from individuals who were unvaccinated, and individuals vaccinated with 4vHPV followed by 2vHPV HPV vaccine (n = 80).ED50, median effective dose; LU/mL, Luminex units/milliliter; r, correlation coefficient.proteins) to provide protection.Furthermore, there are different antibody isotypes (IgG, IgM, IgA, IgD, IgE) and subclasses (IgG1-4, IgA1-2) that each have different affinities for antigen and binding for Fc receptors.Data on these antibody characteristics following HPV vaccination are limited.Measurement of these antibody characteristics would provide novel insight into potential mechanisms of immunity against HPV.

HPV16 and 18 ,
the WHO collaborated with National Institute for Biological Standards and Control recently established International serological Standards for the other seven HPV types in the 9vHPVvaccine.27There are some considerations in establishing this multiplex assay.First, the concentration of serum antibodies to HPV vaccine types and cross-reactive (nonvaccine) HPV types will differ, and therefore additional dilutions may be needed to accurately capture the full range of antibody titers.Consequently, sample dilutions may need to be adjusted when adapting the assay to measure other antibody isotypes or subclasses that may produce weaker responses than total IgG.Second, it is recommended to include positive and negative controls in each assay and running samples from the same individual (e.g., at different timepoints) on the same plate to minimize inter-assay variation.Lastly, new batches of HPV VLP and HPV-conjugated microspheres should be validated before use with SDS-PAGE and the multiplex assay to account for batch variation.In conclusion, we developed a bead-based HPV VLP multiplex immunoassay that can measure serum HPV antibodies to at least seven HPV oncogenic vaccine types.This assay can be expanded to measure more HPV types and adapted to measure other antibody isotypes/subclasses and other antibody features that would be crucial to understanding the humoral immune responses generated by the highly immunogenic and effective HPV vaccine.This assay offers a high-throughput alternative to existing HPV serological assays including PBNA.AUTHOR CONTRIBUTIONSZheng Q. Toh, Paul V. Licciardi, and Amy W. Chung conceived, the study and critically revised the paper.Chau Quang conducted the experiments, analyzed the data, and drafted the manuscript.Zheng Q. Toh, Paul V. Licciardi, Chau Quang, and Amy W. Chung interpreted the data.Tupou Ratu, Evelyn Tuivaga, and Fiona Russell were investigators on the clinical study where the samples were obtained.Troy J. Kemp provided technical assistance with production and purification of HPV VLPs.All authors have read and approved the final manuscript.