Development of in-house, indirect ELISAs for the detection of SARS-CoV-2 spike protein-
associated serology in COVID-19 patients in Panama.
Authors:
Carolina de la Guardia
1,
, Giselle Rangel
2,
, Alcibiades Villarreal
2
, Amador Goodridge
1
, Patricia L
Fernández
1
, Ricardo Lleonart
1,*
Affiliations:
1
Centro de Biología Celular y Molecular de Enfermedades. Instituto de Investigaciones
Científicas y Servicios de Alta Tecnología (INDICASAT-AIP), Ciudad del Saber, Clayton, Apartado
Postal 0843-01103 Panam, Panama.
2
Centro de Neurociencias y Unidad de Investigacin Clnica, Instituto de Investigaciones
Cientficas y Servicios de Alta Tecnologa (INDICASAT AIP), Ciudad del Saber, Clayton, Apartado
Postal 0843-01103 Panam, Panama.
equal contributions
*
Corresponding author. ORCID: 0000-0002-5830-6315. Email: rlleonart@indicasat.org.pa
Abstract
COVID-19 is the name of the acute respiratory disease caused by the new coronavirus SARS-
CoV-2, a close relative of those that caused the severe outbreaks of SARS and MERS several
years ago. Since first appearance on December of 2019, the COVID-19 pandemic has cause
extremely high levels of mortality, morbidity, global economic breakdown and the consequent
human suffering. While vaccination efforts are not extensive and rapid enough, the main tools
to keep the virus under control are still keeping physical distancing, reinforce personal hygiene
measures, using masks and early diagnosis of virus infected persons, either symptomatic or
not. The main diagnostic test for the confirmation of symptomatic individuals is the detection
of viral RNA by reverse transcriptase quantitative real time PCR (RT-PCR). Additionally,
serology techniques, such as ELISA are extremely useful to measure the antibodies generated
in humans after virus contact, as well as the direct presence of viral antigens. In this study we
aim to assemble and evaluate four ELISAs assays to measure the presence of IgG or IgM
specific for the viral Spike protein in COVID-19 patients, using either the full recombinant
SARS-CoV-2 Spike protein or the fragment corresponding to the receptor binding domain. As a
control, we also analyzed a group of prepandemic serum samples obtained before 2017.
Strong reactivity was observed against both antigens. A few prepandemic samples displayed
All rights reserved. No reuse allowed without permission.
(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprintthis version posted May 4, 2021. ; https://doi.org/10.1101/2021.04.30.21256406doi: medRxiv preprint
NOTE: This preprint reports new research that has not been certified by peer review and should not be used to guide clinical practice.
high OD values, suggesting the possibility of some crossreactivity. All four assays show very
good repeatability, both intra- and inter-assay; however, no clear relationship could be
detected between positivity and time of sample collection for serology. Receiver operating
characteristic analysis allowed the definition of cutoffs and evaluation of performance for each
ELISA by estimation of the area under the curve. This performance parameter was high for all
tests (AUC range: 0.98-0.995). Multiple comparisons between tests revealed no significant
difference between each other (P values: 0.24-0.95). Our results show that both antigens are
very effective to reveal both specific IgG and IgM antibodies, with high sensitivity (range 0.929-
0.99) and specificity (range 0.933-0.977). The estimated congruence with the RT-PCR test, as
estimated by Cohen´s Kappa, indicates a high agreement in all cases (range 0.874-0.937). This
test will allow health authorities to have a new tool to estimate seroprevalence, and to
manage and improve the serious sanitary situation created by this virus.
Keywords: SARS-CoV2, COVID-19 diagnosis, serology, IgG, IgM, ELISA, receptor binding
domain, Spike protein
Introduction
The new coronavirus disease 2019 (COVID-19) is a highly infectious disease caused by the
severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) that has become a world-wide
pandemic with huge human and economic losses
1
. Most patients with COVID-19 develop a
mild to severe respiratory illness, while others develop minimally- or asymptomatic infection
2
.
Several investigations have shown that asymptomatic patients can also spread the disease
3
.
Since COVID-19 was declared a global pandemic, the world has suffered more that 130 million
confirmed cases and 2.8 million deaths
4
. The Americas have also been strongly affected, with
more than 58 million confirmed cases and 1.4 million deceased patients (as of April 11,
2021)
5,6
. The last available situation report for Panama, dated April 6, 2021, shows 356,373
confirmed cases and 6,131 deceased
7
.
Virus specific reverse transcriptase polymerase chain reaction (RT-PCR) has become the
assay of choice to rapidly detect viral RNA sequences in acute infection of either symptomatic
subjects or their contacts
8
. For symptomatic individuals, the gold standard for testing is still the
RT-PCR on nasopharyngeal swab samples
9,10
. However, RT-PCR is not devoid of important
All rights reserved. No reuse allowed without permission.
(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprintthis version posted May 4, 2021. ; https://doi.org/10.1101/2021.04.30.21256406doi: medRxiv preprint
limitations, such as false-negative results
11,12,13,14,15
, variability in accuracy for different types of
specimens
10
, hazardous sample collection procedure
16
and sensitivity issues
11,16
.
Besides RT-PCR, the use of serological techniques has also become an important tool for the
management of the COVID-19 pandemic. In symptomatic individuals, serology testing may also
help understand important points, such as the duration of virus specific antibodies
17
. For
asymptomatic individuals, serology testing contributes to answer epidemiological questions,
including virus exposure in general population or in particular high-risk groups, to plan public
health interventions, and to monitor vaccine applications and performance
17
. In fact, we
recently evaluated the performance of a rapid lateral flow immunoassay to detect IgM/IgG
antibodies against SARS-CoV-2
18
. However, this type of rapid tests may present some
limitations, such as not allowing quantitation of antibody titers or suboptimal sensitivity
19
. In
contrast, better results may be obtained by tests based on the enzyme-linked immunosorbent
assay (ELISA), which are easy and convenient to implement
20
.
The humoral response of humans to the SARS-CoV-2 is apparently dominated by the
production of antibodies specific for several structural proteins of the virion, mainly the Spike
(S) and the Nucleocapsid (N) proteins
21
. Here we report the use of recombinant S protein,
either as a full-length antigen, or just the fragment corresponding to the receptor binding
domain (RBD) to assemble in-house, indirect enzyme-linked immunosorbent assays (ELISAs)
for the detection of SARS-CoV-2-specific IgG and IgM antibodies. We also report the
performance of the tests using a group of local COVID-19 patients and prepandemic serum
samples collected before 2017. Our results confirm the usefulness of these antigens for the
ELISA, opening the possibility for conducting new local studies for better knowledge and
management of the disease in the country.
Materials and methods
Patients and samples
The study population represents a sub cohort of a cross-sectional study conducted between
April and December 2020
18
. Clinical serum samples were obtained from 102 hospitalized
patients confirmed to be positive for SARS-CoV-2 viral infection by reverse transcriptase
polymerase chain reaction (RT-PCR, done by a national reference lab) on nasopharyngeal swab
testing. All participants develop moderate clinical symptoms and were treated in public
hospitals located in Panama and Colon cities. For the purpose of negative control samples, 45
prepandemic serum samples were used from de-identified, bio-banked sera stored at
INDICASAT-AIP. These samples correspond to research projects on tuberculosis and
All rights reserved. No reuse allowed without permission.
(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprintthis version posted May 4, 2021. ; https://doi.org/10.1101/2021.04.30.21256406doi: medRxiv preprint
serosurveillance, and included 15 from patients with active pulmonary tuberculosis, 15 from
patients with latent TB infections and 15 healthy individuals from blood bank. All blood
samples were collected following regular procedures and centrifuged at 2000 x g for 10 min to
obtain serum specimens. Serum samples were aliquoted and frozen at -80°C until testing. Prior
to analysis, all serum samples were heat-inactivated at 56°C for one hour.
Ethics statement
Clinical serum samples from COVID-19 patients were collected as part of a cross-sectional
study registered with the Panama Ministry of Health (No.1462) and was approved by the
National Research Bioethics Committee (CNBI; No. EC-CNBI-2020-03-43). Prepandemic serum
samples were taken in Colón province, with ethical clearance No. 1131/CNBI/ICGES/11 and No.
125/CBI/ICGES/14. All the study participants were enrolled after informed consent was given.
Antigens
Recombinant antigens used for ELISA were donated by Dr Florian Krammer, and consisted of 1)
recombinant RBD fragment, corresponding to amino acids 319541 plus a His tail, and 2) the
full spike protein, modified as described for stabilization and multimerization (aa 1-1213)
22
.
Before use, recombinant protein batches were checked for concentration and integrity by SDS-
PAGE (data not shown).
ELISAs
The ELISA assays were implemented to detect SARS-CoV-2-specific IgG and IgM antibodies
specific to S and RBD antigens. Protocol was adapted from Stadlbauer et al (2020)
23
. Briefly,
96-well high binding microtiter plates (Corning Costar no. 3361) were coated with 50 µL of RBD
or S recombinant protein (both at 2 µg/mL) in 1X Phosphate Buffered Saline (PBS), overnight at
4°C. The next day, wells were washed three times with 250 µL of PBST (PBS containing 0.1%
Tween 20) and blocked for 2 h at room temperature (RT) with 200 µL of 1.5% non-fat dry milk
(NFM) in PBST. Plates were then washed as before and serum samples added in 100 µL,
prediluted 1:50 in PBST with 0.5% NFM, and incubated 2 h at RT. After similar washing,
secondary conjugated antibody was added at 1:3,000 dilution in PBST with 0.5% NFM and
incubated for 1 h at RT. Specific IgG and IgM antibodies were detected with anti-human HRP
conjugates: goat anti-human IgG (H+L), Thermo Fisher No. H10307, or goat anti-human IgM
(Heavy chain), Thermo Fisher No. A18835, respectively. Plates were then washed and 100 µL
of substrate TMB ( Tetramethylbenzidine Liquid substrate, Super Sensitive, Sigma)
All rights reserved. No reuse allowed without permission.
(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprintthis version posted May 4, 2021. ; https://doi.org/10.1101/2021.04.30.21256406doi: medRxiv preprint
added and incubated for 5 min. Reaction was stopped with 50 µL of 1M HCl and immediately
read at 450 nm (SynergyHT, BioTek).
Statistical analysis
All statistical analyses were carried out using Prism v9.1.0 (GraphPad, USA) and easyROC
online calculator (http://www.biosoft.hacettepe.edu.tr/easyROC/). Results of data analysis are
presented as estimated statistics and their 95% confidence intervals (CI), or coefficients of
variation (%CV). Normality of the optical density data was tested using Kolmogorov-Smirnov
procedure. Correlation between OD data and evolution time (time elapsed between the
positive RT-PCR result and serum sampling) was assessed using non parametric Spearman's
rank-order correlation test. Receiver-operating characteristic (ROC) analysis was performed on
ELISA OD data to calculate area under the curve (AUC) and optimal cut-off, based on
procedures implemented in online calculator easyROC
24,25
. Pairwise comparisons between the
AUCs of described ELISA tests were done using the multiple comparison procedure
implemented in easyROC. The performance of the tests was assessed by calculating sensitivity,
specificity, and positive- and negative predictive values, associated P values and their 95%
confidence intervals using Fisher exact test as implemented in GraphPad Prism. Based on
cutoff calculated in the ROC analysis for each test, samples were classified as positive or
negative, and agreement between RT-PCR test and ELISAs was assessed using Cohen`s kappa
statistic, calculated as described by Fleiss et al 2003
26
, using an online calculator (GraphPad
QuickCalcs, USA). For all statistical analyses, P values < 0.05 were considered statistically
significant.
Repeatability assessment was done by selecting two positive serum samples, including one
strongly positive (OD 2) and one weakly reactive (OD 0.6), and one negative. With these
samples, ELISAs were repeated, in quadruplicates, over at least five different days. Then,
coefficients of variation were calculated both intra- and inter-assay using OD data.
Results
We assembled and evaluated in-house, indirect ELISAs to detect SARS-CoV-2 specific IgG or
IgM antibodies in serum samples, using two recombinant viral proteins, the receptor binding
domain of spike (RBD), or the full-length trimeric spike as previously reported
22
. For this study,
we used serum samples from two groups, 102 COVID-19 patients and 45 prepandemic serums
obtained from patients enrolled in previous studies (Table 1). COVID-19 patients were enrolled
from May 4 to December 27, 2020 at two hospital wards including: Complejo Hospitalario Dr.
Arnulfo Arias Madrid in Panama City and Hospital Manuel Amador Guerrero, Colon City. All
All rights reserved. No reuse allowed without permission.
(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprintthis version posted May 4, 2021. ; https://doi.org/10.1101/2021.04.30.21256406doi: medRxiv preprint
