Quantification of Hepatitis E Virus ORF2 Protein by a Novel Sandwich ELISA

Abstract

Hepatitis E virus (HEV) causes acute hepatitis in humans, which can progress to chronicity in immunosuppressed individuals. Almost all reported HEV infections are caused by Paslahepevirus balayani genotypes 1-4. The structural ORF2 protein is the major antigen detected in the blood of HEV-infected individuals. ELISA assays to detect IgM antibodies to HEV are the first-line diagnostic tests; however, they showed variable performance with frequently discordant results. A qualitative HEV antigen (ORF2) ELISA is currently available for research use. Here, we report a novel quantitative sandwich ELISA to measure HEV ORF2 protein in 3 matrix types. An optimal pair of capture and detection antibodies was selected among 12 unique combinations tested. A sandwich ELISA protocol was developed using these mAbs and biotin-streptavidin technology. The protocol was further optimized to quantify ORF2 antigen in different matrices by interpolating from a standard curve with a linear range of 3.17 to 50.8 femtomoles/mL. Using this method, ORF2 protein was detected in the cell culture medium of Huh7 cells as early as 2-3 days after transfection with HEV genome RNA and in a medium of human hepatocytes infected with HEV. ORF2 antigen was readily detected in the first 2 weeks post-HEV infection in gerbil sera. In immunosuppressed gerbils, ORF2 was detected up to 6 weeks, and the levels were significantly higher between 3 and 6 weeks post-infection. HEV ORF2 antigen levels showed a strong positive correlation with HEV RNA levels in both cell culture medium and gerbil sera. Our novel sandwich ELISA detected at least 7.3 femtomoles/mL ORF2 protein in human plasma spiked with cell culture propagated HEV and detected ORF2 protein in human plasma samples that tested positive for HEV RNA but negative for anti-HEV antibodies. Further, the assay was nonreactive, with negative human plasma, and HBV or HCV-positive human plasma demonstrating specificity. Overall, our ORF2 antigen ELISA will be useful for quantifying ORF2 antigen in cell culture medium, gerbil serum, and human plasma. Further studies are warranted to evaluate its utility in HEV clinical diagnosis.

Keywords: ORF2; antigen detection; biomarker; biotin–streptavidin detection; hepatitis E virus; immunoassay; monoclonal antibodies; quantitative; sandwich ELISA; serology.

Figure 1

Sandwich ELISA development. (A) Identification of optimal capture antibody and detection antibody pair among 12 unique combinations tested in sandwich ELISA. The optimal capture and detection antibody pair is indicated as red symbols with a connecting line. (B) Evaluation of blocking of detection antibody binding to ORF2 p216 antigen by prior capture antibody binding to the antigen. Four same clone combinations of capture and detection antibody were tested. In both unique and same antibody combinations tested, the capture antibody is unconjugated, and the detection antibody is biotinylated to facilitate detection by Streptavidin–HRP system. ‘c’ and ‘d’ prefixes were added before antibody clone names to distinguish the coating and detection antibody clones, respectively. Each data set (indicated with a different symbol) is derived from testing one capture and detection antibody pair in the sandwich ELISA format using 10-fold serially diluted recombinant ORF2 p216 antigen. P/N ratio indicates the ratio of the absorbance (OD450) of the positive sample to that of negative control (blocking buffer). Higher P/N ratios indicate higher levels of nonoverlapping binding of capture and detection antibodies to the recombinant ORF2 p216 antigen in the sandwich format. Data are from at least 3 independent experiments, and ELISA was performed with 2 technical replicates per sample. Error bars represent standard deviation.

Figure 2

Sandwich ELISA optimization for quantification of HEV ORF2 antigen. (A) Standard curve generation with optimized sandwich ELISA protocol using 2-fold serially diluted recombinant ORF2 p216 antigen. Data are representative of at least 3 independent experiments. Error bars indicate standard error of mean. (B) Determination of the linear range of ORF2 protein quantification using the standard curve. A representative linear standard curve (dotted line) was shown, calculated using simple linear regression. The standard curve equation is in form of y = mx ± b (m is the slope and b is the y-intercept), and the goodness of standard curve fit was measured with the coefficient of determination (R²), which is predetermined to be at least 0.99. Due to presence of different ORF2 isoforms and possible degradation products in the biological samples, a common unit of femtomoles (of ORF2 antigen)/milliliter (fmoles/mL) was used to quantify ORF2 antigen.

Figure 3

Detection of ORF2 antigen of major HEV genotypes in optimized sandwich ELISA (A) SDS-PAGE and Coomassie blue staining analysis of purified recombinant ORF2 antigens of HEV-A genotypes 1, 3, 4, and HEV-C1 (rat HEV). NR—Non-reduced, not-boiled; R—Reduced and boiled. The indicated molecular sizes were obtained with pre-stained protein ladder relevant only to completely denatured proteins. (B) Determination of cross-reactivity of the optimized ORF2 sandwich ELISA using 2-fold serially diluted recombinant ORF2 antigens of HEV-A genotypes 1, 3, 4, and HEV-C1 (rat HEV). The nomenclature of each data set includes strain name and molecular size of the corresponding ORF2 subunit antigen following the prefix ‘p’ for protein. P/N ratio indicates the ratio of the absorbance (OD450) of the positive sample to that of negative control (blocking buffer). The data represent mean ± standard deviation obtained from 3 independent experiments.

Figure 4

HEV ORF2 protein quantification in culture supernatants of HEV-infected hepatocytes. (A) ORF2 protein concentrations in the culture supernatants of Huh-7 cells transfected with full-length HEV RNAs (either capped or uncapped p6) or replication-incompetent HEV RNAs (capped p6 GND). HEV RNA concentrations were determined by RT-qPCR. Both ORF2 protein and HEV RNA were measured from same sample every 2 days post-HEV RNA transfection. Blue and red symbols represent data from capped or uncapped HEV RNA transfections, respectively. Grey symbols represent data from capped, replication-incompetent HEV RNAs. The data are mean ± standard error obtained from at least 3 independent experiments. RT-qPCR and ELISA were performed with 2 technical replicates per sample. (B) ORF2 protein concentrations in the culture supernatants of Primary Human Hepatocyte (PHH) cells infected with purified HEV particles. HEV RNA concentrations were determined by RT-qPCR. Both ORF2 protein and HEV RNA were measured from same sample every 2–4 days post-HEV infection. The data are mean ± standard error obtained from 2 independent experiments. RT-qPCR and ELISA were performed with 2 technical replicates per sample. In both figure panels, the solid line indicates HEV ORF2 protein data, and the dotted line indicates HEV RNA data. Left y-axis indicates ORF2 protein concentration expressed as femtomoles (fmoles) per mL of culture supernatant, and right y-axis indicates HEV RNA copies per mL of culture supernatant.

Figure 5

HEV ORF2 protein quantification in sera of immunocompetent and immunosuppressed gerbils infected with HEV. (A) The ORF2 protein quantities in gerbil sera were measured at weekly time points post-HEV infection. To assess the influence of adaptive immune responses on HEV ORF2 protein levels, general immunosuppression was induced in gerbils by controlled release tacrolimus pellet implanted at a subcutaneous neck site 2 weeks prior to HEV infection. Y-axis indicates HEV ORF2 protein expressed as femtomoles (fmoles)/milliliter of serum samples. Data represent mean ± standard error of mean calculated from 5 to 8 gerbils per treatment group. The statistical significance was tested with mixed effects model along with Greenhouse–Geisser correction and Tukey’s multiple comparison test. Asterisk indicates statistical significance at p < 0.05. (B) Simple linear regression analysis of ORF2 protein levels versus HEV RNA levels in sera of immunocompetent and immunosuppressed gerbils infected with HEV. Each data point represents both HEV ORF2 antigen (femtomoles (fmoles)) and HEV RNA (copies) per milliliter of same serum sample. Blue triangles indicate data from immunocompetent gerbils and red diamonds indicate data from immunosuppressed gerbils. The teal line is the regression line fitted with all data points from both immunocompetent and immunosuppressed gerbils.

Figure 6

Quantification of HEV ORF2 protein in human plasma samples spiked with ORF2 antigen and the assay specificity assessment. (A) HEV ORF2 quantification in spiked human plasma samples. Human plasma samples were freshly prepared using blood samples from 4 different blood donors and individually spiked with HEV ORF2 protein from HEV-infected human hepatocyte culture medium at different volume ratios. The spiked human plasma samples were freeze-thawed once before performing ORF2 sandwich ELISA. The data are expressed as femtomoles (fmoles) per milliliter of plasma samples. Error bars indicate mean ± standard deviation. The LOD was determined as 6.3 (mean ± 3 SD), calculated using 10 negative plasma samples and indicated as dotted line. (B) The sandwich ELISA specificity was assessed with negative donor plasma samples (n = 10) as well as with Hepatitis B virus−positive (n = 6) or Hepatitis C virus−positive (n = 9) donor plasma samples. The cutoff was determined as 0.011 (mean ± 3 SD), calculated using 10 negative plasma samples and indicated as dotted line. Y-axis indicates mean OD450 value subtracted with blank OD450 value. Error bars indicate standard deviation. (C) The sandwich ELISA performance was assessed with a commercial HEV panel comprising human plasma samples positive for both HEV RNA and anti-HEV antibodies (panel # 1–6), positive for HEV RNA but negative for anti-HEV antibodies (panel # 7–9) and negative for both HEV RNA and antibodies (panel # 10). The data for anti-HEV IgM and IgG, as well as the HEV RNA copy numbers, are taken from the AccuSet HEV Performance Panel Datasheet provided by the manufacturer (Seracare; 0820-0503). DIA.PRO and Mikrogen are manufacturers of commercial anti-HEV IgM and IgG assays. IU—International Units. “+” and “−” signs represent reactivity and nonreactivity for anti-HEV IgM or IgG in the commercial assays.