A simple, sensitive, low-cost assay for SARS-CoV-2 spike protein specific antibodies

« A simple, sensitive, low-cost assay for SARS-CoV-2 spike protein-specific antibodies Stephan Menzel1†, Julia Hambach1†, Tobias Stähler1*, Thomas Eden1, Dorte Wendt1, Natalie Tode1, Friedrich Haag1, Eva Tolosa1, Marcus Altfeld1,2, Anahita Fathi3,4,5, Christine Dahlke3,4,5, Marylyn M.

Addo3,4,5, Friedrich Koch-Nolte1* 1 2 Department of Virus Immunology, Heinrich Pette Institute, Leibniz Institute for Experimental Virology, Hamburg, Germany 3 4 Institute of Immunology, University Medical Center Hamburg-Eppendorf, Hamburg, Germany Section Infectious Diseases, I.

Medical Clinic and Polyclinic, University Medical Center Hamburg-Eppendorf, Hamburg, Germany Department of Clinical Immunology of Infectious Diseases, Bernhard Nocht Institute for Tropical Medicine, Hamburg, Germany. 5 German Center for Infection Research, Partner Site Hamburg-Lübeck-Borstel-Riems, Hamburg, Germany. † Equal contribution *correspondence: Prof.

Dr.

Friedrich Koch-Note, [email protected] Running title: detecting antibodies against native CoV-2 S Key words: SARS-CoV-2; COVID-19; Hamburg epidemic; native spike protein; serology; antibodies; immunofluorescence microscopy; flow cytometry Abbreviations: mAb, monoclonal antibody; hcAb, heavy chain antibody; Nb, nanobody; S, spike protein; RBD, receptor-binding domain; ELISA, enzyme linked immunosorbent assay; IFM, immunofluorescence microscopy; FACS, fluorescent activated cell sorter; CHO chinese hamster ovary; HEK, human embryonic kidney Abstract We report a simple assay for detecting antibodies against the native CoV-2 spike protein.

We applied this assay to monitor antibody development in COVID-19 patients, household contacts, and hospital personnel during the recent epidemic in the city state of Hamburg.

All recovered COVID-19 patients showed high levels of CoV-2 S-specific antibodies.

With one exception, all household members that did not develop symptoms also did not develop detectable antibodies. Similarly, lab personnel that worked during the epidemic and followed social distancing guidelines remained antibody-negative.

We conclude that high-titer CoV-2 S-specific antibodies are found in most recovered COVID-19 patients and in symptomatic contacts, but only rarely in asymptomatic contacts.

Currently (early June), the epidemic has largely subsided in Hamburg (less than 5 new cases per day, down from ~ 220 new cases/day in mid-April). However, to date there seems to be only marginal antibody-based immunity in Hamburg. Plasmids and cells required to set up and perform this assay are freely available from our lab. Introduction Hamburg is a city state of ~1.8 Mio inhabitants in Northern Germany.

The COVID-19 epidemic started in Hamburg in early March when families returned from their traditional skiing vacation (1).

The local epidemic peaked in the middle of April with ~220 new cases per day and has subsided to less than 5 new cases per day at the end of May (Fig.

1).

The accumulated total number of cases currently (June 7th, 2020) is ~5.100 (270 cases/100.000 inhabitants) with 250 deaths (5%) and ~4.700 recovered patients (2). Individuals in Hamburg that were tested SARS-CoV-2 positive by PCR-assay for CoV-2 genomic DNA in throat swabs were referred to quarantine in their homes for 14 days together with other members of their household.

Patients that developed severe disease were admitted to the University Medical Center Hamburg-Eppendorf or to another local hospital.

Patients that developed respiratory failure were transferred to intensive care and those who went on to develop acute respiratory distress syndrome received extracorporeal membrane oxygenation. Hamburg had a surplus of ICU-beds and respirators throughout the epidemic.

Recovered patients and their household roommates were released from quarantine 10 days after recovery from symptoms.

The lockdown in Germany is being reverted in a step-wise fashion on the condition that the number of new daily cases remains below 50/100.000 inhabitants.

There is still uncertainty regarding the degree of immunity in the general population (3). Several serological assays have been developed to detect CoV-2-specific antibodies (4-6).

Most of these are based on ELISAs that utilize recombinant proteins, e.g.

the receptor-binding domain (RBD) of the spike protein (S).

Many of these assays are fraught with high costs, low specificity and/or false positives.

While the epidemic seems to be under control in China and most European countries, it is still spreading rapidly in other regions of the world (7). Therefore, there is still an urgent need for serological tests with higher specificity and lower costs. Our lab has expertise in raising antibodies against native membrane proteins by cDNA immunization (8, 9).

We routinely use transiently transfected CHO or HEK293T cells to monitor specific antibody responses in immunized animals (10).

Here we set out to apply this assay for detecting antibodies directed against the native spike protein of SARS-CoV-2 in human serum samples.

The results demonstrate that this simple low-cost assay allows specific detection of CoV-2 S-specific antibodies.

The assay can easily be set up in any research or diagnostic lab equipped with a cell culture facility and a fluorescent microscope or flow cytometer.

The required plasmids and cells can be obtained freely from our lab. Results Validation of IFM and FACS assays to detect SARS-CoV-2 S-specific antibodies Figure 2 illustrates the assay for detecting CoV-2 S-specific antibodies using transfection of cells with a full-length cDNA expression plasmid followed by immunofluorescence microscopy (IFM) or flow cytometry.

We co-transfect the cells with a cDNA expression vector encoding green fluorescent protein (GFP) fused to a nuclear localization signal in order to distinguish co-transfected cells (green-fluorescent nuclei) from untransfected cells.

The latter serve as negative controls and help to detect antibodies directed against irrelevant target proteins.

Cell-bound CoV-2 S-specific antibodies are detected with a fluorochrome-conjugated secondary antibody.

We use a 96-well format to simultaneously analyze many samples at a cost of less than 20 € per plate. To validate the assay, we used the recently described CoV-S-specific nanobody VHH72 (11) in a human IgG1 heavy chain antibody (hcAb) format (12).

Bound hcAb was detected with PEconjugated anti-human Ig(H+L).

The results show that this antibody strongly stains the cell surface of GFP-transfected CHO cells (Fig.

3A) and HEK293T cells (Fig.

3B).

The results also confirm a high degree of co-transfection of the expression plasmids for GFP and CoV-2 S.

The expression construct for nuclear GFP can be substituted with constructs for other suitable fluorescent proteins, e.g.

mitoDSRED or nuclear BFP, the secondary antibodies with other fluorescently labeled secondary antibodies, e.g.

FITC or AlexaFluor647-conjugates (not shown). The vast majority of COVID-19 patients develop moderate to very high levels of CoV-2 Sspecific antibodies. We next set out to monitor the antibody responses of SARS-CoV-2+ individuals that had been hospitalized at the University Medical Center in Hamburg.

We also analyzed serum samples from members of our lab that had continued to work during the epidemic while following the social distancing recommendations (lab work was organized in two shifts with coworkers wearing masks and maintaining a distance of > 1.5 m).

Sera were treated for 30 min at 56°C to inactivate complement components and SARS-CoV-2 virions.

Figure 3C and 3D show representative microscopy and flow cytometry results.

The staining intensities allow a semiquantitative assessment of antibody levels by IFM (indicated by -, +, ++, +++) and a quantitative assessment by flow cytometry (indicated by the mean fluorescence intensity of GFP+ cells).

The results of IFM and FACS analyses of representative 96-well plates are shown in Supplementary Figures 1 and 2, respectively. Table 1 summarizes the results of the samples from hospitalized patients.

In general, the results obtained by immunofluorescence microscopy and flow cytometry concord very well. Antibodies typically become detectable with 8-14 days after disease onset.

Where serial samples of the same patients were available, these generally showed maintenance of a plateau level for the duration of analysis (maximally 44 days after disease onset).

All 30 samples obtained from healthy coworkers at the UMC Hamburg during the epidemic were negative (Supplementary Table 1).

We also analyzed pre-epidemic samples, some of which had consistently tested positive with the EuroImmune ELISA (13).

None of these samples showed any detectable antibodies in our assays (Supplementary Table 2). Several co-quarantined close contacts of COVID-19 patients did not develop any detectable CoV-2 S-specific antibodies. We also analyzed samples from eight households in which at least one member had contracted COVID-19 and had been quarantined with his/her close contacts (Table 2).

Families H1-H5 had returned to Hamburg from skiing vacation in late February and early March.

Patient H1-1, a physician at our hospital, whose case was recently described in detail, was the first known COVID-19+ patient in Hamburg (1).

He acquired SARS-CoV-2 in Trentino, Italy.

After his return to Hamburg on February 23rd he transmitted CoV-2 to two of his 132 contacts, including his wife (individual H1-2) (1). Members of families H2, H3, and H4 reported fever and headaches on March 5th, 8th, and 9th, respectively.

Other family members developed symptoms within the next 10 days.

Both members of H2 showed typical symptoms and were tested PCR positive within 3 days of another, both developed antibodies.

Family H3 returned from skiing in Lech, Austria on March 8th.

H3-1 (50-yr) was the first to develop symptoms on March 9th and was tested PCR+ a few days later.

H3-2 and H3-3 developed COVID-19 symptoms on March 14th and were also tested PCR positive.

All members of family H3 developed antibodies.

It is suspected that H3-4 (19yr) was actually the first one to acquire the disease.

He does not recall any symptoms beyond a mild sore throat for 1 day and was tested SARS-CoV-2 negative by PCR on March 15th and on March 26th.

He did develop CoV-2-S-specific antibodies, suggesting that he may have been the first member of his family to acquire SARS-CoV-2, but had already eliminated the virus by the time of his PCR-tests.

Interestingly, his 21-yr old sibling H3-3 was tested PCR positive and remained PCR+ for more than 4 weeks, despite only weak symptoms.

All symptomatic members of the H4 family developed antibodies with one exception: H4-2 (26-yr) reported symptoms on March 15th but was not tested by PCR.

He did not develop any detectable CoV-2 specific antibodies.

In H5, the mother reported fever and headache on March 9th, but her 4-yr old son remained asymptomatic throughout the quarantine. Family H6 represents two close contacts of an 89-yr old patient that contracted COVID-19 in early March while in the hospital for unrelated disease and that succumbed to COVID-19 related complications in early April.

H6-1 and H6-2 probably got infected while visiting this patient before the hospitals were closed for visitors.

This patient likely was a super-spreader, because seven of the eight relatives that visited him in the hospital contracted COVID-19. H7 and H8 represent households of physicians that acquired SARS-CoV-2 in Hamburg, most likely at work.

H7-1 (37 yr) developed symptoms on March 23rd and was tested positive two days later.

He lives with three roommates, one of whom is his partner (34 yr).

His partner experienced symptoms 11 days later and was tested positive by PCR.

The physician and his partner both developed antibodies against CoV-2 within 11 days after disease onset.

The other roommates (29-yr, 32-yr) showed no symptoms and no detectable antibodies.

Notably, these two roommates adhered to hygiene rules as much as possible.

H8-1 (29-yr) became ill on March 31 and was tested positive by PCR the following day.

She reported high fever, headache, and anosmia for three days, and developed antibodies within 9 days after disease onset.

She lives with five roommates (29-33-yr), including her partner.

Neither her partner nor any of her roommates experienced symptoms, despite sharing bathroom, kitchen and meals without strict adherence to social distancing principles.

None of these contacts developed any detectable CoV-2 S-specific antibodies. Discussion We report a simple, low-cost assay for detecting SARS-CoV-2 spike protein specific antibodies.

The assay may help health care providers to monitor disease progression, to identify health care personnel that likely are resistant to re-infection, and to identify recovered individuals with high antibody titers that may be suitable as plasma and/or antibody donors (1416). High specificity, low costs, simple set up, and incorporated controls to detect false positives, are the main advantages of our assay over established ELISA and peptide-based assays.

Use of purified plasmid DNAs and natively expressed spike protein likely account for the high specificity and reproducibility of our assay.

ELISA assays typically depend on the production and purification of recombinant proteins, which may be contaminated with traces of proteins from the producing cells.

Peptide arrays often miss antibodies directed against conformational epitopes.

Low costs of readily available reagents (~20 €/100 assays) guarantees that the assay can easily be set up in any laboratory equipped with a standard cell culture facility and access to a simple fluorescent microscope or flow cytometer.

And co-transfection of cells with a cDNA encoding a fluorescent protein allows the clear distinction of "false positive" caused by antibodies that recognize irrelevant antigens.

Plasmids, cells and detailed protocols to set up the assay are available from our lab. We applied the assay to samples collected at the University Medical Center HamburgEppendorf during the COVID-19 pandemic from March until May 2020.

Our results revealed that all individuals that had recovered from a symptomatic COVID-19 infection developed moderate to high titers of CoV-2 S-specific antibodies.

Antibody levels in hospitalized patients correlated with disease severity.

Typically, antibodies became detectable 6-12 days after disease onset and reached plateau levels within three weeks.

None of the 30 analyzed laboratory personnel of the University Medical Center showed any detectable SARS-CoV-2 Sspecific antibodies.

Remarkably, many close contacts of COVID-19 patients in household quarantine did not develop any symptoms.

These individuals also did not develop any detectable CoV-2 S-specific antibodies.

These results indicate that these close contacts were not infected and consequently did not develop antibody-mediated immunity even during coquarantine with a symptomatic patient.

This conclusion is in line with reports from other regional outbreaks of COVID-19 (4, 6, 17, 18). We conclude that individuals that have recovered from COVID-19 carry high titers of CoV-2 S-specific antibodies.

These individuals are likely resistant to re-infection with SARS-CoV-2. In contrast, many asymptomatic close contacts are antibody-negative.

These individuals presumably continue to be susceptible to COVID-19 infection or might have a T-cell driven immunity from previous coronavirus infections (19). Materials & Methods: Serum samples and secondary antibodies: Serum or plasma was obtained from patients and healthy volunteers after informed consent according to protocols approved by the local ethics committee (Ethik-Kommission der Ärztekammer Hamburg PV4780 and PV5139 for samples from healthy donors, and PV7298 for samples from COVID-19 patients).

Serum samples were incubated for 30 min at 56°C to inactivate complement and coronaviruses (20) before use. Fluorochrome-conjugated secondary antibodies were obtained from Dianova, Hamburg (Jackson): PE-conjugated anti-human IgG(H+L) F(ab’)2.

All results shown in this paper were obtained with serum samples at a dilution of 1:100 - 1:200..... »