Biomedical engineers at Duke University have devised a test to quickly and easily assess how well a person’s neutralizing antibodies fight infection from multiple variants of COVID-19 such as Delta and the newly discovered Omicron variant.
ARS-CoV-2 is a single-stranded RNA virus with four structural proteins: nucleocapsid, membrane, envelope, and spike (S) (1). The S protein – composed of the S1 and S2 domains – is exposed on the viral coat of SARS-CoV-2 and plays an essential role in viral attachment, fusion, entry, and transmission (2).
Specifically, the receptor binding domain (RBD) of S1 binds to the SARS-CoV-2 cellular receptor—angiotensin-converting enzyme-2 (ACE2), which mediates viral entry into cells (2). Because of its critical role in viral entry, the S protein serves as the basis for COVID-19 vaccines and as the target for antibody-based therapeutics (3).
Although coronaviruses have genetic proofreading mechanisms to maintain their genome (4), they are still prone to mutations that can alter viral replication, transmission, and recognition by the host immune response. Of particular concern are mutations within the RBD because of its important role in viral entry. Recent genetic epidemiological surveillance has identified emerging SARS-CoV-2 variants that are circulating globally.
Variant B.1.1.7 (also known as Alpha variant) originated in the United Kingdom (5) and contains nine mutations in the S protein, including one within the RBD—N501Y. This mutation increases the binding affinity to ACE2 (6, 7) and contributes to the increased transmissibility of B.1.1.7 (8–10).
Fortunately, several studies have shown that convalescent and vaccinee sera effectively cross-neutralize B.1.1.7 with only a minimal decrease in potency (11–14). Conversely, variant B.1.351 (Beta), which originated in South Africa (15), and variant P.1 (Gamma), which originated in Brazil (16) and Japan (17), each harbor three mutations within the RBD—K417N (B.1.351)/K417T (P.1), E484K, and N501Y.
Evidence is mounting that these two strains can evade neutralization by monoclonal antibody (mAb) therapies and are more resistant to neutralization by polyclonal antibodies resulting from natural infection or immunization (12–14, 18–21). These variants of concern (VOCs), as well as newly emerging ones, such as the B.1.617 (Delta) lineage identified in India (22), pose new challenges in the effort to contain the spread of the virus.
Assays to detect anti–SARS-CoV-2 antibodies are an important tool to assess natural or vaccine-induced humoral response at the individual patient level and for epidemiological surveillance at the population level. While many antibody binding assays have been developed for COVID-19 serodiagnosis (23–27), these tests are unable to determine the specific fraction of antibodies that can potentially neutralize the SARS-CoV-2 virus and thus confer protection.
However, these assays are labor-intensive, costly, and require highly trained personnel working in biosafety level 3 facilities. Neutralization assays using vesicular stomatitis virus and lentivirus pseudotyped with SARS-CoV-2 S protein have also been reported (31, 32). These assays can be performed in biosafety level 2 facilities; however, they still require live cells and >24 hours to carry out the assay.
To circumvent the need for viruses, permissive cells, and level 2 or 3 containment facilities, enzyme-linked immunosorbent assay (ELISA)–type ACE2-RBD blocking assays have been developed that mimic the virus-host interaction and serve as a surrogate for antibody neutralization activity (33–37).
Specifically, these assays measure the ability of nAbs to block interactions between ACE2 and purified RBD using a competitive binding inhibition format in an ELISA plate. Several studies have demonstrated that this assay format shows high correlation with conventional neutralization tests (33, 34), and thus its readout can serve as a proxy for the protection conferred by antibodies.
With more transmissible and virulent SARS-CoV-2 strains now circulating globally, there is an urgent need for a test that can measure nAbs against several VOCs simultaneously by an easily deployable rapid test. Such a test could be useful to study the impact of RBD mutations on neutralization, to monitor the efficacy of vaccines against circulating VOCs in low-resource settings, to identify individuals who may be susceptible to reinfection or breakthrough infections even after vaccination, and to identify patients with COVID-19 who may benefit from mAb therapies.
As proof of principle, we demonstrate the performance of our assay against four SARS-CoV-2 strains—wild type (WT), B.1.1.7, P.1, and B.1.351. This assay is constructed by inkjet-printing RBD proteins from each variant (Fig. 1B) onto a “nonfouling” poly(oligoethylene glycol methyl ether methacrylate) (POEGMA) coating, as described previously (27, 38).
Nearby, fluorescently labeled human ACE2 is inkjet-printed upon a dissolvable trehalose pad (fig. S1 shows the chip layout). When a sample without nAbs is added, fluorescently labeled ACE2 dissolves from the POEMGA brush and binds to RBD capture sites, leading to a high fluorescence signal.
In the presence of potential nAbs, the RBD-ACE2 interaction can be partially or completely blocked, resulting in a decrease in fluorescence signal (Fig. 1C). We demonstrate the multiplexing capability of CoVariant-SCAN by simultaneously assessing the neutralizing activity against WT, B.1.1.7, P.1, and B.1.351 from a single sample.
We used CoVariant-SCAN to assess the efficacy of known neutralizing therapeutic mAbs, natural immunity from convalescent plasma, and vaccine-induced immunity. During this study, as the B.1.617.2 (Delta) variant emerged in India, we rapidly incorporated this variant into the CoVariant-SCAN assay without needing to reoptimize the assay, which shows that new VOCs can be easily accommodated, as they emerge.
DISCUSSION
The emergence of new SARS-CoV-2 variants with the potential to escape neutralization by therapeutic mAbs and natural or vaccine-induced immunity is of great concern. As sequencing programs continue to identify mutations within the SARS-CoV-2 genome, functional tests are needed to rapidly assess the impact of those mutations, especially those that arise in the S gene, due to its critical role in viral entry, propagation, and transmission.
As of late 2020, several new variants of interest have been identified that are circulating in New York (57), California (58), India (59), and other locations around the world, all of which contain mutations in S. Monitoring the potential escape of these and future variants needs to be a priority to reduce the risk of resurgence in cases, even in highly seropositive populations, as SARS-CoV-2 will likely become endemic (60).
In this study, we developed a rapid test, termed the CoVariant-SCAN, to simultaneously assess the ability of antibodies to block the ACE2-RBD interaction against five SARS-CoV-2 variants: WT, B.1.1.7, P.1, B.1.351, and B.1.617.2. Our work is motivated by the urgent need for a rapid and easy-to-use assay that supplements conventional antibody neutralization tests that are labor-intensive, costly, require highly trained personnel, and are thus inaccessible in many regions around the world. While other assays have been developed that measure the ability of nAbs to block ACE2-RBD binding (33, 34, 37), to the best of our knowledge, CoVariant-SCAN is the first test that can detect nAbs against several SARS-CoV-2 variants simultaneously within 1 hour.
Furthermore, because the CoVariant-SCAN is built upon a “nonfouling” polymer coating that eliminates nearly all nonspecific binding, the assay can be conducted directly from undiluted plasma (38, 61, 62). Although CoVariant-SCAN is not quite as sensitive as the cPass SARS-CoV-2 neutralization antibody detection kit (fig. S9)—developed by Lin-Fa Wang’s group (33) and commercialized by GenScript—our multiplexed assay only requires a single 1-hour incubation, which could be automated with a newly developed microfluidic chip (27) and read with a portable detector (27, 62), making the platform well suited for point-of-care use in resource poor environments.
Using the CoVariant-SCAN, we demonstrated that several mAbs partially or completely lose blocking/neutralizing activity toward the RBD proteins of B.1.351 (K417N, E484K, and N501Y) and P.1 (K417T, E484K, and N501Y) variants. B.1.1.7, which only contains the N501Y mutation within the RBD, is less prone to escape neutralization by mAbs raised against WT RBD. This finding is consistent with other studies that used authentic virus and pseudovirus assays (13, 14, 21). We also demonstrated that CoVariant-SCAN specifically measures nAbs, as our assay measured weak or no blocking activity for a panel of eight convalescent patient-derived mAbs with binding specificity to RBD but no neutralizing capacity.
Our data confirm that the EUA-approved mAb REGN10933 may lose activity against B.1.351 and P.1, but when combined with REGN10987, blocking/neutralization is less severely affected by VOCs, highlighting the importance of identifying potent mAb cocktails and continued monitoring of their effectiveness against emerging VOCs. Comparison of the REGN antibody cocktail with the Pfizer and Moderna vaccines shows that the Regeneron therapeutic antibody cocktail possibly provides greater protection against the highly infectious Delta variant that is currently displacing other variants worldwide.
Given that the individual response of vaccinated individuals against the Delta variant is, on average, up to 50% lower than WT, whereas neutralization of the Delta variant by the antibody cocktail is similar to WT, we suggest that this cocktail may be useful as a prophylactic for vaccinated individuals with a low level of nAbs against the Delta variant who are at increased imminent risk of exposure to this variant, such as because of travel to a region where that VOC is endemic.
We also investigated the ability of polyclonal antibodies from the plasma of naturally infected individuals to neutralize each SARS-CoV-2 variant. We found that individuals with mild, moderate, and severe cases of COVID-19 developed nAbs that could effectively block ACE2-RBD binding for the WT virus. Severe cases that required hospitalization in the ICU resulted in a more robust nAb response that could neutralize B.1.1.7, P.1, and B.1.351, albeit at a diminished level relative to WT.
For mild and moderate cases, ACE2 blocking against P.1 and, B.1.351 variants was more similar compared to prepandemic plasma. These results suggest that individuals with mild or moderate COVID-19 infections may not produce sufficient nAb titers to prevent reinfection from P.1 or B.1.351 variants, which may explain the reinfection cases that have been appearing globally (63–65). These findings also underscore the importance of control measures to avoid resurgence in locations, even where seropositivity is already high, as is happening in Manaus, Brazil (16).
Next, we assessed the ability of CoVariant-SCAN to measure the neutralizing activity of vaccinee plasma. We found that individuals who received the Pfizer and Moderna vaccines developed nAbs that could neutralize all variants relative to prepandemic negative control plasma; however, activity against P.1 and B.1.351 variants were significantly attenuated relative to WT.
Conversely, B.1.1.7 neutralization was relatively unaffected. We also demonstrated that the nAb titer continues to increase after the second dose of the Pfizer and Moderna vaccines and that the second dose yields nAbs that are better at cross-neutralizing VOCs. Our data also show that there is a considerable individual heterogeneity in nAb levels and that some individuals develop robust responses against all variants tested, while others may benefit from variant specific boosters that are currently being developed.
The CoVariant-SCAN is an ideal platform to identify those individuals because it could be conducted at the point of care, is easily manufactured at scale, and can be deployed globally independent of a cold-chain or centralized testing laboratory.
Another key strength of our platform is the ability to rapidly test the impact of S protein mutations on immunity as they arise in newly emerging VOCs.
Our workflow only requires inkjet printing purified RBD proteins as a row of separate capture sites without any changes to the detection reagent, making further multiplexing simple. As SARS-CoV-2 variant sequences are identified and deposited into repositoriessuch as the Global Initiative on Sharing Avian Influenza Data (GISAID) (66), recombinant RBDs from these variants can be quickly expressed, purified, and integrated into our assay, as we demonstrated with the B.1.617.2 variant, while this study was in progress.
Although we focused on mutations within the RBD for this study, we can also use the full S1 protein—which contains additional mutations in VOCs—as the capture antigen on the CoVariant-SCAN (fig. S10). Likewise, we could prospectively assess the impact of modifications at important residues on immunity, as others have done (46, 47), to identify mutations of concern. Therefore, we believe that the bespoke nature of CoVariant-SCAN will be useful to assess the impact of emerging SARS-CoV-2 mutations (67).
There are several potential scenarios where the CoVariant-SCAN could be useful. First, it could be deployed as an epidemiological tool to assess the efficacy of vaccines against circulating or emerging VOCs in specific regions. Second, it could be used to monitor individual patients’ risk for future infection by a circulating VOC based on their nAb profile.
Similarly, the CoVariant-SCAN could be used at the bedside to test patients presenting with acute COVID-19 who are either known to have been infected by a VOC or if there is a high burden of VOC in their community, making it likely that their infection is caused by a VOC. Patients with low neutralizing activity could be treated immediately with the Regeneron cocktail or a similar mAb therapy to reduce the likelihood of severe infection. This approach would be especially useful for patients who are immunocompromised at the time of SARS-CoV-2 infection or vaccination, as they are likely to have a weaker humoral response and therefore are more at risk for reinfection and/or severe disease.
There are several limitations of the CoVariant-SCAN that warrant further discussion. First, our assay uses ACE2-RBD blocking as a proxy for antibody neutralization, which does not account for other antibody-mediated effector functions such as complement activation or antibody-dependent cellular cytotoxicity which can contribute to immunity (68). Furthermore, we only consider mutations within the RBD and therefore do not account for mutations occurring at other locations that could affect neutralization, such as within the N-terminal domain of S, which is an important target of several nAbs (69).
Our assay format also does not consider the role of cellular immunity from memory T cells after primary infection or immunization, which is known to play an important role in preventing SARS-CoV-2 infection (70–72). Despite these limitations, we believe that CoVariant-SCAN is a promising rapid test to provide public health officials with the tools for effective serologic surveillance of SARS-CoV-2 variants and to answer questions being raised about the effectiveness of immunity—both natural and vaccine-induced—against emerging variants of SARS-CoV-2.
More information: Jacob Heggestad et al, Rapid test to assess the escape of SARS-CoV-2 variants of concern, Science Advances (2021). DOI: 10.1126/sciadv.abl7682. www.science.org/doi/10.1126/sciadv.abl7682
