Diagnostic TR-FRET assays for detection of antibodies in patient samples

Abstract

Serological assays are important diagnostic tools for surveying exposure to the pathogen, monitoring immune response post vaccination, and managing spread of the infectious agent among the population. Current serological laboratory assays are often limited because they require the use of specialized laboratory technology and/or work with a limited number of sample types. Here, we evaluate an alternative by developing time-resolved Förster resonance energy transfer (TR-FRET) homogeneous assays that exhibited exceptional versatility, scalability, and sensitivity and outperformed or matched currently used strategies in terms of sensitivity, specificity, and precision. We validated the performance of the assays measuring total immunoglobulin G (IgG) levels; antibodies against severe acute respiratory syndrome coronavirus (SARS-CoV) or Middle Eastern respiratory syndrome (MERS)-CoV spike (S) protein; and SARS-CoV-2 S and nucleocapsid (N) proteins and applied it to several large sample sets and real-world applications. We further established a TR-FRET-based ACE2-S competition assay to assess the neutralization propensity of the antibodies. Overall, these TR-FRET-based serological assays can be rapidly extended to other antigens and are compatible with commonly used plate readers.

Keywords: CoraFluor; TR-FRET assays; neutralization assay; serological testing; vaccine efficacy studies.

Graphical abstract

Figure 1

TR-FRET assay setup, flowchart, and validation by CR3022 antibody (A) Principle of TR-FRET assay. Antibodies recognizing human IgG were labeled with BODIPY FL. SARS-CoV-2 S proteins were labeled with CoraFluor-1 (Tb) and were both mixed with serum for isotype-specific antibody detection. The light pulse at 337 nm excites the CoraFluor-1 (Tb)-labeled S protein and emits light at 490 nm, which in turn triggers energy transfer to the BODIPY FL-labeled secondary antibodies found in proximity induced by the analyte generating a TR-FRET signal detected at 520 nm. (B) Flowchart of TR-FRET assay. The serum samples are diluted into microtiter plates and added into reaction mixture. Reaction mixture is added beforehand by an automated dispenser (Multidrop Combi Reagent Dispenser). The diluted serum samples are added into the reaction mixture using a Gryphon (Art Robbins Instruments) or manually by multi-channel pipetting. Plates are read on a TR-FRET-compatible plate reader (e.g., PHERAstar FSX Microplate Reader). (C) Titration of CR3022 IgG/IgM/IgA1 into pre-formed mix of Tb-S protein (7.5 nM final) and BODIPY FL-labeled αIgG/αIgM/αIgA (250 nM final). (D) As in (C) but in the presence of 1:150 dilution of negative serum. Data of (C) and (D) are represented as means ± SD of two technical replicates (n = 2). (E) Titration of positive and negative serum in final assay condition 250 nM BODIPY FL-αIgG and 7.5 nM Tb-S. Data are represented as means ± SD of three technical replicates (n = 3). (F) TR-FRET αIgG-S assay. Titration of CR3022 IgG in the presence and absence of negative serum at 1:100 dilution. Data are represented as means ± SD of three technical replicates (n = 3). The concentration of CR3022 selected for the LoD study is highlighted with a red arrow. (G) TR-FRET αIgG-S. LoD for TR-FRET assay was assessed by comparing 20 replicates of the CR3022 with 20 replicates of buffer control in the presence and absence of negative serum at 1:100 dilution.

Figure 2

Performance of TR-FRET assay: Sensitivity, specificity, and precision (A) Sensitivity and specificity of ELISA IgG assay performed on a cohort of 68 SARS-CoV-2 PCR-positive samples (CoV2+) and 100 pre-pandemic negative samples (healthy). (B) Sensitivity and specificity of TR-FRET αIgG-S performed on the same cohort. (C) Correlation of TR-FRET IgG with serum dilution 1:150 and ELISA IgG at serum dilution 1:100. Data are represented as means ± SD of two technical replicates (n = 2). (D) Comparison between three independent runs performed on different days by three different operators of a TR-FRET αIgG-S assay on a set of positive responders as well as negative control samples (68 total). (E) The calculated average repeatability across operators (CV%) and average intermediate precision (calculated across days and operators) corresponding to data in (D). Data are represented as means ± SD of two technical replicates (n = 2).

Figure 3

TR-FRET assay format is compatible with other antigens (A) Sensitivity and specificity of TR-FRET αIgG-S protein assay performed on MassCPR set including 90 pre-pandemic negative samples and 100 SARS-CoV-2-positive samples. (B) Sensitivity and specificity of TR-FRET αIgG-N protein assay performed on MassCPR. (C) Correlation of αIgG-S titer in TR-FRET assay versus ELISA for MassCPR. Note the “ceiling” of the signal in ELISA and the high dynamic range of TR-FRET. (D) Correlation of IgG titer N protein in TR-FRET assay versus ELISA for MassCPR. (E) Correlation of TR-FRET αIgG-S and TR-FRET αIgG-N assays performed on MassCPR indicates diverse immune response to different antigens. (F) Cross-reactivity between S proteins of SARS-CoV-2 and SARS-CoV measured by TR-FRET IgG titer on MassCPR. (G) As in (F) but for S proteins of SARS-CoV-2 and MERS-CoV. Data are represented as means ± SD of two technical replicates (n = 2).

Figure 4

TR-FRET assay accepts multiple sample types (A) Correlation of ELISA IgG-S response between the matched set of whole dried blood self-collection samples (Neotheryx kit) and serum samples from the same donors collected within 2 weeks of each other at the Dana-Farber Cancer Institute (DFCI; IRB #20-260). (B) As in (A) but for TR-FRET IgG-S assay. (C) Correlation between TR-FRET and ELISA responses in the IgG-S assay on the self-collection samples. (D) The response of ELISA or TR-FRET assay compared between a set of SARS-CoV-2 negative samples. (E) The response of ELISA or TR-FRET assay compared between a set of SARS-CoV-2 positive samples. Data are represented as means ± SD of two technical replicates (n = 2) on a cohort of 140 CoV2+ and 35 CoV2 samples.

Figure 5

TR-FRET can be applied for total IgG amount testing and validation (A) Correlation between TR-FRET and ELISA responses in the IgG-S assay on IMPACT study samples. (B) Correlation between ELISA in the IgG-S assay and total IgG on IMPACT study samples. (C) Correlation between TR-FRET in the IgG-S assay and total IgG on IMPACT study samples. (D) Correlation of ELISA and TR-FRET with total IgG with difference total IgG levels. Total IgG <2,500 mg/dL is labeled as blue. Total IgG in the range 2,500–3,000 mg/dL is labeled as red. Total IgG >2,500 mg/dL is labeled as yellow. (E) Histogram of total IgG levels in the IMPACT study. (F) The principle of TR-FRET total IgG assay. Nanobodies recognizing human IgG are labeled with AF488. Immunoglobulin-binding protein G is labeled with Tb (terbium). The light pulse at 337 nm excites Tb chelate protein G and emits light at 490 nm, which in turn triggers energy transfer to AF488-labeled nanobodies found in proximity induced by the analyte generating a TR-FRET signal detected at 520 nm. (G) Titration of positive and negative serum in final assay condition 25 nM Tb-protein G and 25 nM AF488-nanobody. Data are represented as means ± SD of two technical replicates (n = 2). (H) Correlation of total IgG measured by TR-FRET and ELISA on a set of 39 samples at 1:40,960 dilutions. Data are presented as n = 1 for ELISA and as two technical replicates (n = 2) for TR-FRET.

Figure 6

TR-FRET neutralization assay setup and validation (A) Principle of TR-FRET neutralization assay. Human ACE2 receptor is labeled with CoraFluor-1. SARS-CoV-2 S protein is labeled with BODIPY FL, and both are mixed with serum for neutralization antibody detection. After excitation at 337 nm, emission at 490 (CoraFluor-1) and 520 nm (BODIPY FL) is detected, and the TR-FRET ratio is calculated as a 520/490 signal. Neutralizing antibodies present in the serum will competitively bind the S protein and reduce the TR-FRET signal. (B) Titration of B38, H4, SAD-S35, 40491-MM43, CR3022, and a negative control αFLAG into pre-formed mix of btn-ACE2 (8 nM final), Tb-SA (2 nM final), and BODIPY FL-S (8 nM final). Data represented as means ± SD of two technical replicates (n = 2). (C) As in (B) but performed in the presence of serum. Data represented as means ± SD of three technical replicates (n = 3). (D) Titration of positive and negative serum in final assay condition 8 nM btn-ACE2, 2 nM Tb-SA, and 8 nM BODIPY FL-S. Data represented as means ± SD of two technical replicates (n = 2). (E) Sensitivity and specificity of TR-FRET ACE2-S neutralization assay performed on MassCPR (90 pre-pandemic negative samples and 100 SARS-CoV-2-positive samples). Final serum dilution of 1:50 (v/v). Data represented as means ± SD of two technical replicates (n = 2). (F) Sensitivity and specificity of reported cellular pseudovirus neutralization assay performed on MassCPR (90 pre-pandemic negative samples and 100 SARS-CoV-2-positive samples). Reported is the NT50, the concentration at which 50% of neutralization is observed as calculated from quadruplicate, 7-point dose response. (G) Correlation of cellular neutralization NT50 against TR-FRET ACE2-S inhibition. (H) Receiver-operating characteristic (ROC) curve indicating the performance of detection of IgG levels for S or N using ELISA or TR-FRET and TR-FRET ACE2-S inhibition assays. All TR-FRET data in (E)–(H) are represented as means ± SD of two technical replicates (n = 2). ELISA data in (H) are represented as means ± SD of four technical replicates (n = 4).