High-throughput pseudovirion-based neutralization assay for analysis of natural and vaccine-induced antibodies against human papillomaviruses

Abstract

A highly sensitive, automated, purely add-on, high-throughput pseudovirion-based neutralization assay (HT-PBNA) with excellent repeatability and run-to-run reproducibility was developed for human papillomavirus types (HPV) 16, 18, 31, 45, 52, 58 and bovine papillomavirus type 1. Preparation of 384 well assay plates with serially diluted sera and the actual cell-based assay are separated in time, therefore batches of up to one hundred assay plates can be processed sequentially. A mean coefficient of variation (CV) of 13% was obtained for anti-HPV 16 and HPV 18 titers for a standard serum tested in a total of 58 repeats on individual plates in seven independent runs. Natural antibody response was analyzed in 35 sera from patients with HPV 16 DNA positive cervical intraepithelial neoplasia grade 2+ lesions. The new HT-PBNA is based on Gaussia luciferase with increased sensitivity compared to the previously described manual PBNA (manPBNA) based on secreted alkaline phosphatase as reporter. Titers obtained with HT-PBNA were generally higher than titers obtained with the manPBNA. A good linear correlation (R(2) = 0.7) was found between HT-PBNA titers and anti-HPV 16 L1 antibody-levels determined by a Luminex bead-based GST-capture assay for these 35 sera and a Kappa-value of 0.72, with only 3 discordant sera in the low titer range. In addition to natural low titer antibody responses the high sensitivity of the HT-PBNA also allows detection of cross-neutralizing antibodies induced by commercial HPV L1-vaccines and experimental L2-vaccines. When analyzing the WHO international standards for HPV 16 and 18 we determined an analytical sensitivity of 0.864 and 1.105 mIU, respectively.

Figure 1. HT-PBNA protocol.

Assay plate preparation (A) and neutralization assay (B) are separated. In a first step, serial dilutions of serum samples are performed on one dilution plate. Identical assay plates are generated by transferring the dilutions to multiple replica-assay plates which can be stored at −20°C. In the second step, the neutralization assay is carried out in a add-on format using the previously prepared assay plates. A) Assay plate preparation. Sera are transferred from a 96 well storage in SBS standard to a 384 well polypropylene V-bottom plate for serial dilution with a pipetting robot in cell culture medium. Finally the serially diluted sera are transferred with the same pipetting robot to each of 3*n white 384 well cell culture assay plates (n = number of PV-types). The plates are sealed immediately with a cover foil and stored at −20°C until their use in the PBNA. B) Assay assembly and read out. Assay plates are thawed and pseudovirions followed by reporter cells are added with a bulk dispenser. Three identical assay plates originating from the same serum dilution plate are used for each PV- type. After 2 days of incubation the luminescence from the Gaussia reporter is read directly in the assay plates. Inhibition (%) is calculated by normalization of the luminescence to the mean of the negative control wells without serum present on each plate. The median of the triplicate values is used for the calculation of the ED₅₀-value (effective dilution giving 50% inhibition) for each serum according to the four parameter logistic curve fit model y = A+(B−A)/(1+(C/x)^∧D).

Figure 2. Influence of cell density on Gaussia luciferase activity.

HeLaT clone 4 cells were seeded at the indicated densities into white 384 well plates containing HPV 16°C for 2 days substrate for the Gaussia luciferase or ATPLite1step™ was added and the luminescence measured. Columns show the median luminescence of 64 wells with bars indicating the standard deviation.

Figure 3. Titration of different PSV preparations.

Serial dilution of PSV preparations from HPV types 16, 18, 31, 45, 52, 58 and BPV-1 were assayed for Gaussia luciferase activity in the HT-PBNA. For HPV types 16 and 18 a second PSV preparation is indicated by an asterisk (*). Luminescence signals are expressed as relative light units (RLU). The titers for undiluted HPV 16, 18, and BPV-1 PSV were 3.0×10⁹, 3.9×10⁹, 1.1×10¹⁰ genomes per ml, respectively.

Figure 4. Influence of PSV concentration on HT-PBNA analytical sensitivity and robustness.

Neutralization titers of a serum from a Gardasil® immunized individual expressed as ED₅₀ values (open circles and open squares) with the variability (bars indicating the 95% confidence intervals) were determined at different PSV concentrations. The maximal luminescence intensities (RLU) obtained without serum are shown as crosses. An arrow indicates the dilution of the PSV preparation that was used in subsequent neutralization assays.

Figure 5. Effect of PSV-serum premix incubation time on neutralization titer.

A serum from a Gardasil® vaccinated individual was pre-incubated for different times with HPV 16 or HPV 18 PSVs before the neutralization assay was initiated by the addition of reporter cells. ED₅₀ values with 95% confidence intervals are shown.

Figure 6. HT-PBNA inter- and intra-run variability of neutralization titers.

The ED₅₀ values for HPV 16 and HPV 18 PSV of the serum standard were determined in 58 repeats on seven assay days (runs) over a period of 2 months. For six of the 7 assay days, triplicates of the standard serum dilutions were assayed 8 times each, for one assay date the standard serum was assayed 10 times.

Figure 7. Analytical sensitivity and type-specificity of the HPV 16 and HPV 18 HT-PBNA.

Titration of the WHO International Standards for antibodies to HPV 16 (left) and HPV 18 (right) in HT-PBNA using PSV of HPV types 16, 18, 31, 45, 52, 58 and BPV-1.

Figure 8. Detection of neutralizing antibodies as result of natural infection by HT-PBNA.

Thirty-five pre-vaccination sera from a study involving patients with CIN2+ lesions were tested for neutralizing antibodies against PSVs of HPV types 16, 18 and 31. The geometrical mean titer for each HPV type is indicated as a horizontal line and the cut off value (ED₅₀ = 80) as a dashed line.

Figure 9. Comparison of HPV 16 HT-PBNA (Gaussia) with manPBNA (SEAP and Gaussia) and GST-HPV 16 L1 antibody binding assay for natural and vaccine-induced HPV16-specific responses.

HPV 16 HT-PBNA (A–D), manPBNA using SEAP reporter (A and B; titers, ED₅₀) and a bead-based GST-HPV 16 L1 antibody binding assay (C and D; median fluorescent intensity (MFI) at 1∶100 or 1∶2700 serum dilution) were used to determine reactivity of pre- (n = 35; A and C) and post-vaccination (n = 72; B and D) sera. Serum samples analyzed were from women with HPV 16 positive, high-grade cervical intraepithelial neoplasia (base-line sera of the chimeric HPV 16 L1-E7 vaccination trial). Cut off values used for positive/negative classification (broken lines in C) were an ED₅₀ of 80 for the HT-PBNA and 109 MFI at 1∶100 for the GST-L1 antibody binding assay.

Figure 10. Titration of vaccine sera in the HT-PBNA using PSV of HPV types 16, 18, 31, 45, 52, 58 and BPV-1. A

) Standard serum from Gardasil® vaccinated person. B) Cervarix® vaccine serum. C) Serum from a mouse immunized with L2 epitopes (amino acids 17–36) from HPV16 inserted in the capsid of adeno-associated virus 2 particles.