key: cord-0789595-fspxrh9h
authors: Yan, Sijia; Ahmad, Khan Zara; Warden, Antony R.; Ke, Yuqing; Maboyi, Nokuzola; Zhi, Xiao; Ding, Xianting
title: One-pot pre-coated interface proximity extension assay for ultrasensitive co-detection of anti-SARS-CoV-2 antibodies and viral RNA
date: 2021-08-04
journal: Biosens Bioelectron
DOI: 10.1016/j.bios.2021.113535
sha: 5a6b49609b3978eb5315ae04840553d6ea0f8456
doc_id: 789595
cord_uid: fspxrh9h

In the field of in vitro diagnostics, detection of nucleic acids and proteins from biological samples is typically performed with independent platforms; however, co-detection remains a major technical challenge. Specifically, during the coronavirus disease 2019 (COVID-19) pandemic, the ability to simultaneously detect viral RNA and human antibodies would prove highly useful for efficient diagnosis and disease course management. Herein, we present a multiplex one-pot pre-coated interface proximity extension (OPIPE) assay that facilitates the simultaneous recognition of antibodies using a pre-coated antigen interface and a pair of anti-antibodies labeled with oligonucleotides. Following anti-antibody-bound nucleic acid chain extension to form templates in proximity, antibody signals can be amplified, together with that of targeted RNA, via a reverse transcription-polymerase chain reaction. Using four-color fluorescent TaqMan probes, we demonstrate the co-detection of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2)-specific antibodies and viral nucleic acids in a single bio-complex sample, including nucleocapsid protein-specific IgG and IgM, and the RNA fragments of RdRp and E genes. The serum detection limit for this platform is 100 fg/mL (0.67 fM) for the anti-SARS-CoV-2 antibody and 10 copies/μL for viral RNA. The OPIPE assay offers a practical and affordable solution for ultrasensitive co-detection of nucleic acids and antibodies from the same trace biological sample without the additional requirement of complicated equipment.

The coronavirus disease 2019 (COVID-19) pandemic, caused by severe acute respiratory 2 syndrome coronavirus 2 (SARS-CoV-2), has spread rapidly and poses a serious threat to human 3 health (Zhou et al., 2020) . In a retrospective single-center study, 249 COVID-19 cases 4 diagnosed at Shanghai Public Health Clinical Center between January 20 and February 6, 2020, 5 were evaluated. The median age was 51 years, 126 (50.6%) were men, and 22 (8.8%) were 6 critically severe patients requiring intensive care unit admission. This report showed that nearly 7 10% of COVID-19 cases in Shanghai, China were severe and critical, highlighting the need for 8 the development of more effective diagnostic and monitoring approaches . 9

Reverse transcription-polymerase chain reaction (RT-PCR) analysis of RNA from 10 nasopharyngeal swabs or lung fluid is considered the gold standard to detect and confirm 11 SARS-CoV-2 infection (Ai et al., 2020; Chan et al., 2020) . Several highly sensitive and specific 12 systems for SARS-CoV-2 RNA detection have been developed (Ma et al., 2020; Park et al., 13 2021 ). Moreover, serum testing for SARS-CoV-2 RNA can be used as an additional detection 14 method, especially for patients with false-negative swab results as well as patients exhibiting 15 severe infection (Zhang et al., 2020b) . Additionally, conventional enzyme-linked 16 immunosorbent assay (ELISA) for SARS-CoV-2-specific IgG and IgM can be used to detect 17 infection and monitor recovery progress. Indeed, a combinatorial approach comprising nucleic 18 acid and antibody quantification has been shown to increase the positive detection rate of 19 suspected patients from 51.9% to 98.6% (Guo et al., 2020) . 20

In addition to ELISA, alternative methods for qualitative/semi-quantitative analysis of 21 SARS-CoV-2-specific antibodies include colloidal gold immunochromatographic analysis and 22 lateral flow immunochromatographic analysis. Although these methods are simple, rapid, and 23 cost-effective, they exhibit limited sensitivity and quantitative analysis capacity (Huang et al., 24 2020; Van Elslande et al., 2020). Currently, fluorescent microsphere immunochromatographic 25 test strips for SARS-CoV-2 report the lowest limit of detection (LOD) for IgM (1 ng/mL) and 26

IgG (30 ng/mL) antibodies, whereas microfluidic chemiluminescent ELISA can achieve an 27 LOD of 10 pg/mL using spiked serum samples (Tan et al., 2020) . According to quantitative test 28 results, the average IgM concentrations in a weak positive sample and a strong positive sample 29 are 20 ng/mL and 110 ng/mL, respectively, whereas the average IgG concentrations are 0.27 30 μg/mL and 1.88 μg/mL, respectively (Zhang et al., 2020a) . 31

Hence, the ultrasensitive co-detection of nucleic acid and protein biomarkers in a single bio-32 complex sample would not only reduce the false-positive and -negative rates but would also 33 provide insights into clinical progression. However, the current methods capable of achieving 34 J o u r n a l P r e -p r o o f such a feat require the implementation of complex and expensive instruments as well as highly 1 skilled technicians proficient in operating multiple platforms (Amanat et al., 2020; Stadlbauer 2 et al., 2020). 3 Therefore, to address these limitations, we aimed to develop a multiplex one-pot pre-coated 4 interface proximity extension (OPIPE) assay for the ultrasensitive co-detection of nucleic acids 5 and proteins to reduce the burden on medical facilities, particularly in resource-limited settings. 6

This assay is based on the recognition of antibodies using a pre-coated antigen interface and 7

anti-antibodies labeled with oligonucleotides, which extend to double-stranded DNA templates 8 in proximity. After adding fluorescent TaqMan probes and virus RNA template, the protein 9 signal in the solid-phase and that of the nucleic acid in the liquid phase are simultaneously 10 amplified via RT-PCR. In collaboration with the Shanghai Public Health Clinical Center, the 11 sensitivity of the OPIPE assay was assessed using clinical serum samples from SARS-CoV-2-12 positive patients and healthy controls. 13

In compliance with biosafety protocols approved by the Institutional Review Board of 16 Shanghai Jiao Tong University, Shanghai, China (Approval No. 202003048), the performance 17 of OPIPE assay for detecting SARS-CoV-2-specific antibody was evaluated in human serum 18 samples. Pooled and inactive serum samples from patients with COVID-19 (age range 24-77 19 years) and healthy subjects (age range 28-65 years) were purchased from Elabscience Company, 20

Ltd.. (Wuhan, Hubei, China). Each participant provided written consent to participate in the 21 study, which was approved by the regional investigational review board and performed 22

according to the Declaration of Helsinki. The severity information for the purchased samples 23 was provided by the phlebotomists at Elabscience during sample collection. The serum samples 24 are from patients who were diagnosed as positive infections and later-on released from hospital 25 after in-house treatment to ensure their serum viral RNA turned into negative. All serum 26 samples were confirmed by RT-PCR to be free of viral nucleic acids before releasing to 27 subsequent laboratory research use.The serum was purified by filtration through a 0.22-μm 28 filter membrane to remove possible pathogens and impurities. The antibody concentration of 29 both nucleocapsid protein-specific IgG and IgM (N-IgG and IgM) in thus purified serum was 30 quantified to be 10 μg/mL based on ELISA measurement comparing with standard substance 31 from the National Sharing Platform for Reference Materials (Beijing, China) (Fig. S1) . 32

The modified oligonucleotide sequences, probes, and primers (Table S1) 

The technique used to produce OPIPE probes and pre-coated antibody tubes were adapted 17 from our previous work (Yan et al., 2021) . Briefly, TCEP·HCl reduced oligonucleotides were 18 coupled with Sulfo-SMCC-activated anti-human IgG and IgM antibodies to form OPIPE probes. 19

The recombinant SARS-CoV-2 N protein was then pre-coated on a polypropylene PCR tube to 20 capture the specific antibody. 21

Positive and negative serum were diluted with a proportional concentration gradient. All 23 details pertaining to the incubation, extension, and washing processes are described in our 24 previous reports (Yan et al., 2021) . Prior to performing RT-PCR, the SARS-CoV-2 RNA 25 fragments, PrimeDirect™ Probe RT-qPCR Mix (Takara, Japan), RNase-free H2O, and four sets 26 of probe primers were simultaneously added to the reaction tubes. The reaction system is shown 27 in Table S4 . All PCR reactions were performed in a 25 µL reaction volume on a LightCycler 28 96 (Roche, Switzerland) under the manufacturer's recommended fast cycling conditions for 40 29 cycles. The instrument was adjusted to simultaneously record signals from all four channels. 30

For each data point, the raw −ΔCt value was normalized by subtracting the Ct value for the 32 extension control reaction from that of the corresponding sample, thereby correcting for 33 technical variation. All linear regression analyses and curve fitting were performed using Origin 9.0. Two-tailed Student's t-tests were performed using Microsoft Excel 2016 to calculate P-1 values. A two-sided P-value < 0.05 was considered statistically significant.The relative standard 2 deviations (RSD) were also calculated for repilcated tests. 3

2.1 Oligonucleotide design, construction, and purification for the OPIPE assay 5

As a proof of concept, we used OPIPE to simultaneously detect two SARS-CoV-2 RNA 6 fragments,which encode from Envelope (E) and RNA-dependent RNA polymerase ( The probe oligonucleotide sequences in the OPIPE assay were designed to contain the unique 17 binding sites for the primers and TaqMan probes. For the 3′-linked probe, we designed a 57-18 mer oligonucleotide that included 28-bases corresponding to the 5′-linked probe and unique 19 binding sites for specific primers used in qPCR. Primers were generated using Primer-BLAST 20 (https://www.ncbi.nlm.nih.gov/tools/primer-blast/). Considering that the TaqMan probe was 21 used in this experiment, we optimized the design scheme of the probe based on the following 22 logic. The 5′-end of the TaqMan probe must be in close proximity with the 3′-end of the forward 23 primer; the length of the probe should be approximately 25 bp, its GC content should be 50-24 70%, and guanine should be avoided as the first base at the 5′-end. Importantly, alignment of 25 the sequences for the synthetic oligonucleotide, human genome, and other coronavirus genomes 26 revealed no homology. Oligonucleotide sequences, primers, and TaqMan probes for the 27 detection of N-IgM and N-IgG are shown in Fig.2C and Table S1 . The probes used in the OPIPE 28 assay were generated through Sulfo-SMCC-driven conjugation of 5′-thiol-modified 29 oligonucleotides to a pair of monoclonal mouse anti-human IgG antibodies (anti-IgG mAb) or 30 polyclonal rabbit anti-human IgM antibodies (anti-IgM pAb) following amino activation. 31

Successful probe synthesis was confirmed via nucleic acid and protein detection using sodium 32 dodecyl sulfate polyacrylamide gel electrophoresis, followed by staining with GelRed and 33

Coomassie Brilliant Blue, respectively ( Fig. 2D and 2E ). Subsequently, the probes were 34 J o u r n a l P r e -p r o o f purified using centrifugal filtration membranes with a 100-kDa cut-off and stored at −20 °C 1 until use. 2

To optimize the assay pipeline, polypropylene tubes used for conventional PCR were treated 5 with an 8% glutaraldehyde solution and pre-coated with either mouse anti-human IgG or rabbit 6 anti-human IgM (Yin et al., 2013) . The tubes were then incubated with 10% and 100% human 7 serum spiked with a 100-fold dilution series of IgG or IgM at concentrations ranging from 100 8 fg/mL to 10 μg/mL. After washing away unbound substances that were not captured by the 9 tube-coated anti-antibody, the probes were incubated with target IgG or IgM. When tube-coated 10 anti-antibodies and probes simultaneously recognized IgG or IgM, the oligonucleotides of the 11 probes in close proximity were subjected to an extension reaction with T4 polymerase. RT-12 PCR was then performed following addition of the TaqMan probes and primers to the tubes; 13 signals were subsequently recorded. The estimated LOD in 10% serum was 100 fg/mL, and the 14 correlation coefficients (R 2 ) calculated by regression of measured −ΔCt values (Ct value of 15 sample minus that of the control) on logarithmic concentrations of IgG or IgM were 0.990 and 16 0.994, respectively, in the linear range between 100 fg/mL and 10 μg/mL. Moreover, although 17 the −ΔCt values in 100% serum are lower than in 10% serum, the estimated LOD and linear fit 18 are almost the same (Fig.S2) . 19

After confirming the effectiveness of the probe for N-IgG and -IgM, the polypropylene tubes 21 were immobilized with SARS-CoV-2-specific N protein following the aforementioned protocol. 22

To improve the efficiency of capturing specific antibodies in COVID-19-positive donor serum, 23 we optimized the N protein concentration at the solid-liquid phase interface in the tubes. -value<0.05) . Notably, for a diluted serum containing 10 pg/mL N-IgG, the 1 mean −ΔCt values increased from 0.35 (in 2 μg/mL N protein pre-coated tubes) to 1.12 (in 8 2 μg/mL N protein pre-coated tubes), corresponding to a more than 3-fold increase (Fig.S3) . 3 Therefore, 8 μg/mL was used as the coating concentration for the tube interface in all 4 subsequent experiments. 5

Next, COVID-19 positive or negative sera were subjected to 100-fold serial dilution and 6 incubated in the 8 μg/mL N protein pre-coated tubes. Probe incubation and extension were 7 performed as described above. The N-IgG and -IgM signals were obtained after PCR 8 amplification using primers and specific TaqMan probes carrying the fluorescent reporters 9 FAM and HEX, respectively (Table S1 ). The estimated LOD in COVID-19 positive serum was 10 100 fg/mL, and the R 2 of the measured −ΔCt value using logarithmic concentrations of N-IgG 11 or -IgM were 0.995 and 0.991, respectively, in the linear range between 10 pg/mL and 10 μg/mL. 12

The RSD was 2.99% and 4.22% for IgG and IgM at the highest concentration (10 μg/mL).In 13 the linear range, the RSD for the detection of IgM and IgG in positive serum was 10.76% ± 14 3.85% and 9.82% ± 4.17%, respectively. Due to the relatively weak signal, the RSD for the 15 lowest concentrations was around 15%. Significant differences (P < 0.05) were observed in 16 −ΔCt values between the positive and negative serum samples at each dilution, indicating that 17 definitive diagnosis could be made using serum samples diluted up to 10,000,000-times, and 18 clinical samples could be adequately preserved (Fig. 3A-F) . We also compared the abilities of 19 the OPIPE assay to detect N-IgM and N-IgG with standard ELISA (Table S5) . Furthermore, 20

we did experiments with narrower dilutions (10-fold) of serum for both N-IgM and N-IgG 21 (Fig.S4 ). Our results indicate that the linearity is not significantly shifted between different 22 dilution factors. To verify the repeatability of the method, we also performed the experiments 23 at different times (Table S6) . 24 [ Fig. 3 ] 25

Due to the strict prevention requirements of COVID-19 in China, we could not obtain viral 27 RNA or serum containing RNA from patients. As an alternative, we added in vitro synthesized 28 RNA fragments to the positive donor serum. To co-detect nucleic acids and specific antibodies 29 in the same bio-complex-containing sample, we added 10 5 copies RNA transcripts of the E and 30

RdRP genes to the donor serum and then diluted them in 10-fold series dilution. A commercially 31 available one-step RT-qPCR mix included in the same setup was used for antibody detection. 32

The primer and probe with ROX and Cy5 fluorescent reporters are listed in Table S1 . Primers 33 and probes were specific to the SARS-CoV-2 genome sequence and were cross-validated to 34 J o u r n a l P r e -p r o o f ensure no homology in sequence alignment with human SARS-CoV and bat-SARS-related 1 CoV genomes (Fig.S5) . The PCR amplification signal of RNA is exhibited with Ct values, 2 which are inversely proportional to the concentration of the nucleic acid fragments. The Ct 3 values of a non-complementary sequence of H1N1 virus is significantly higher than that of the 4

RdRP and E gene fragments at all concentration (Fig. 3G-I) . Results from three different time 5 points are presented in Table S7 . 6 2.4 OPIPE enables ultrasensitive RNA and antibody co-detection in one pot 7

The above experiments were carried out in parallel tubes using the same equipment. To 8 achieve true co-detection in one pot,10 5 copies of RNA transcripts of the E and RdRP genes 9

were spiked into the positive serum. Then we added the positive serum in 10-fold series dilution 10 in the same tube before RT-PCR. All four sets of TaqMan probes and primers required for the 11 one-pot reaction were then added and co-detected using the one-step RT-qPCR Probe Kit. The 12 reaction conditions used for the simultaneous one-pot co-detection of SARS-CoV-2 RNA as 13 well as minor adjustments required for patient antibodies based on separate assays, are detailed 14 in Table S4 . 15 Our results showed that the signal generated by the one-pot reaction differed slightly from 16 those obtained for the individual separate reactions; however, the sensitivity was not 17 significantly affected. Moreover, compared with the results of the separate detection methods, 18

IgG and IgM showed a slightly lower −ΔCt value at a relatively higher RNA concentration, and 19 a higher −ΔCt value at lower RNA concentrations. Additionally, the estimated LOD for N-IgG 20 and -IgM in 10% COVID-19-positive serum was 100 fg/mL, with R 2 values (of measured Ct 21 values on a logarithmic scale) of 0.995 and 0.991, respectively, in the linear range between 10 22 pg and 10 μg/mL. Within the linear range, the RSD of OPIPE assay for IgG and IgM detection 23 in positive serum was 9.41 ± 2.47 and 9.20 ±2.95, respectively. Significant differences (P < 24 0.05) were observed in the −ΔCt values between positive and negative serum samples at each 25 dilution (Fig. 4A-F) . The nucleic acid results showed no change in sensitivity down to 10 26 copies/μL, with a robust linear relationship observed between RdRP (R 2 = 0.995) and E gene 27 detection (R 2 = 0.998). In the one-pot OPIPE assay, the Ct values of a non-complementary 28 sequence is significantly higher than that of the RdRP and E gene fragments at all concentration 29 Fig. 4G-4I) . 30 [ Fig. 4 ] 31

In addition to sensitivity, we further examined the cross-reactivity of the OPIPE assay. Cross-33 reactivity refers to an extraneous signal caused by the binding of a non-target protein similar to 34 J o u r n a l P r e -p r o o f the expected target protein. In this experiment, we added non-SARS-CoV N protein-specific 1 human IgG and IgM to 10% human serum diluted with PBS. In a 100-fold dilution series of 2 IgG or IgM at concentrations ranging from 100 fg/mL to 10 μg/mL, the OPIPE assay showed 3 no significant cross-reactive signals. The −ΔCt values at the highest concentration of 100 ng/mL 4 were still lower than at the lowest concentration of 100 fg/mL in COVID-19-positive donor 5 serum (Fig. 5A) . Moreover, we evaluated the cross-reactivity of TaqMan probe and primer sets 6 by adding two interfering RNAs, namely RNA fragments of hepatitis B virus (HBV) and the 7 complete RNA of influenza A virus (H1N1), to the SARS-CoV-2 RdRP and E gene fragments. 8

Addition of the two interfering RNAs had no significant effect on the detection signal of RdRP 9 and E genes. As a negative control, the amplification of two non-SARS-CoV-2 sequences, HBV 10 and H1N1 viral fragments, exhibited no significant difference compared with the background 11 containing no viral fragment (Fig.5B ) 12

We compare the LOD of our OPIPE with the existing commercial ELISA assay (approved 13 for emergency-use-administration by Food and Drug Administration), the LOD of which ranges 14 1.6-13,500 ng/mL by using real patient samples. Although both OPIPE and ELISA take about 15 2 hours, OPIPE reduces the detection limit from pM to fM concentration range. Moreover, the 16 OPIPE assay exhibited enhanced sensitivity for antibody detection compared to previously 17 reported methods (Table 1) 

Certain limitations were noted in this study. For instance, the OPIPE protocol must be 30 conducted in a biosafety laboratory when handling infectious pathogens since the washing steps 31 in the current protocol require uncapping operations. Furthermore, the incubation and extension 32 of OPIPE probes as well as the addition of RNA template and primers, should be performed in 33 separate areas to maximally avoid cross-contamination. Nevertheless, we report that the OPIPE 34 assay, a one-pot diagnostic procedure, can simultaneously detect nucleic acids and proteins in 1 the same trace clinical sample. This strategy significantly enhances the specific binding of the 2 target protein with immunoglobulins present in patient serum. Moreover, our platform 3 possesses a high sensitivity of 100 fg/mL (0.67 fM) in serum, representing a minimum 15-fold 4 enhancement over that of solid-phase proximity ligation assays (0.01 pM). Moreover, the 5 OPIPE nucleic acid results exhibited high sensitivity (10 copies/mL). Importantly, the OPIPE 6 assay has an ultrasensitive detection limit and does not require any additional sophisticated 7 instrumentation, making it accessible to clinical settings in resource-limited regions. By The authors declare that they have no competing financial interests or personal relationships 9 that could have influenced the work reported in this paper. 10

The authors thank the staff at AEMD SJTU for their support. 12

We gratefully thank the financial support from NSFC Projects Innovation Special Zone Project. Thanks to AEMD SJTU for the support. The funding source 23 played no role in study design; in the collection, analysis, and interpretation of data; in the 24 writing of the report; or the decision to submit the article for publication. 25

Supplementary data to this article can be found online. 27

Ai, T., Yang, Z., Hou, H., Zhan, C., Chen, C., Lv, W., Tao, Q., Sun, Z., Xia, L., 2020. 29

Radiology 296(2), E32-E40. All data are presented as the mean ± standard deviation. Each experiment had five technical replicates.

J o u r n a l P r e -p r o o f

Cross-reactivity of the OPIPE assay with non-SARS-CoV-2 specific nucleic acids and antibodies. (A) The signal change from 100 fg to 10 μg/mL for a series solution of non-SARS-CoV nucleocapsid protein-specific human IgG and IgM standards in 10% serum. (B) At two different concentrations

Similarly, there was no significant difference between the amplification signals (dark blue) of the E gene alone or that with two non-SARS-COV-2 RNA fragments, fragments of HBV and H1N1 (light blue). Comparison of the amplification signal of the two non-SARS-COV-2 RNA segments for HBV and H1N1

 Table. 1 The comparison between the OPIPE assay and other methods of detection of SARS-CoV-2 specific antibody.

The detection method of SARS-CoV-2 specific antibody (IgG or IgM) Orders of magnitude LOD a) Ref. 

 One-pot pre-coated interface proximity extension assay (OPIPE) for co-detection is developed. A 100-fold enhancement in sensitivity was achieved relative to microfluidic ELISA with as little as 1 μL serum. The serum detection limit for the anti-SARS-CoV-2 antibody is 100 fg/mL (0.67 fM) and for viral RNA is 10 copies/μL.  OPIPE assay detect anti-SARS-CoV-2 antibodies and viral RNA simultaneously from one clinical sample with only PCR equipments.