Cartography of SARS‐CoV‐2 variants based on the susceptibility to therapeutic monoclonal antibodies

Abstract

A comprehensive picture of a phenotypic relationship among severe acute respiratory syndrome coronavirus 2 (SARS‐CoV‐2) variants has been poorly studied. Here, this study presents cartography showing how the wild‐type strain of SARS‐CoV‐2 and 14 variants are alike or different from the perspective of the susceptibility to 12 therapeutic monoclonal antibodies. The Alpha variant is close to the wild‐type strain, whereas the Beta, Gamma, and Delta variants diverge from the wild‐type. The map highlights the very unique property of the Omicron variant. Interestingly, sublineages of the Omicron variants, BA.1, BA.2, and BA.4/5, differ substantially in the cartography.

1. INTRODUCTION

After the emergence of severe acute respiratory syndrome coronavirus 2 (SARS‐CoV‐2), the etiological agent of coronavirus disease 2019 (COVID‐19), the virus has spread worldwide and caused great morbidity and mortality. Several vaccines for COVID‐19 have been developed and proved to effectively prevent infection, hospitalization, and death. ¹ Furthermore, the administration of neutralizing antibodies against the viral spike protein is a promising option for the treatment of COVID‐19 ² ; for example, a cocktail treatment with casirivimab and imdevimab has been shown to ameliorate clinical outcomes. ³

The evolution of SARS‐CoV‐2 has accumulated mutations in the genome, some of which affect the transmissibility, pathogenicity, and antigenicity of the virus. ⁴ When variants with such genetic changes are identified to cause possible threats to public health, public health authorities including the World Health Organization designated the variants as Variants of Interest (VOIs). ⁵ Furthermore, variants are designated as Variants of Concern (VOCs) when evidence from multiple countries demonstrates that the variants are associated with increasing transmissibility, virulence, or decrease in the effectiveness of existing measures including diagnostics, vaccines, and therapeutics at a degree of global public health significance. ⁵

Antigenic changes of such VOCs are especially worrisome. They are reported to decrease the effectiveness of current vaccines. ⁶ , ⁷ The efficacy of antibody‐based remedies also seems to decline for some variants. ⁸ , ⁹ However, a phenotypic relationship among SARS‐CoV‐2 variants remains elusive. The present study developed cartography of the SARS‐CoV‐2 wild‐type strain and 14 variants by analyzing data of neutralization tests to assess and visualize a comprehensive picture of the phenotypic relationship among VOCs and VOIs from the perspective of the susceptibility to 12 therapeutic monoclonal antibodies.

2. MATERIALS AND METHODS

2.1. Data collection and integration

Data from published studies that performed neutralization assays to investigate the effectiveness of therapeutic monoclonal antibodies for different SARS‐CoV‐2 variants were collected. Included studies, used antibodies, tested variants, and performed assays are listed in Supporting Information: Table 1.

To integrate data from multiple studies, the effective concentrations of monoclonal antibodies in neutralization tests for SARS‐CoV‐2 variants were divided by the effective concentration against the wild‐type or D614G strain of the virus for normalization in each study. Then, the geometric mean of the normalized effective concentrations of the same monoclonal antibody for each variant among different studies was calculated.

2.2. Statistical analysis

To summarize the integrated data, imputation of missing values and dimensionality reduction were computed by the singular value decomposition algorithm after unit variance scaling using the pcaMethods package in R. The scores of each SARS‐CoV‐2 variant for the first and second dimensions were plotted to depict cartography.

3. RESULTS

The analysis developed cartography of the SARS‐CoV‐2 wild‐type strain, 5 VOCs, and 6 VOIs that reflect the susceptibility to 12 monoclonal antibodies (Figure 1). Two viral strains of the Iota variant with a unique mutation, either S477N or E484K in the spike protein, and sublineages of the Omicron variant, BA.1, BA.2, and BA.4/5, were analyzed separately. Sufficient data for the Eta and Theta variants were unavailable for analysis in this study.

Figure 1.

Phenotypic cartography of SARS‐CoV‐2 variants based on the susceptibility to therapeutic antibodies. A map shows the results of dimensionality reduction of integrated data of neutralization tests using 12 monoclonal antibodies for the SARS‐CoV‐2 wild‐type strain and 14 variants. The x‐ and y‐axes correspond to the first dimension (65% of the variance is explained by the dimension) and the second dimension (12%), respectively. Variants of concern are in red, and variants of interest are in blue. Because the numbers of tested antibodies were small for the Zeta, Lambda, and Mu variants, their positions on the map are less definitive. Those variants are indicated by open circles.

The cartography shows that the Alpha variant locates close to the wild‐type strain (Figure 1). Other VOCs, such as the Beta, Gamma, and Delta variants, diverge from the wild type. The Omicron variant possesses ~30 mutations in the spike protein. The variant is distantly positioned away from other variants reported so far on the map.

The original Omicron variant (BA.1) and its sister sublineage, BA.2, share 21 amino acid mutations in the spike protein, with BA.1 carrying additional 12 unique amino acid mutations and BA.2 carrying different 6 mutations along with several deletions. BA.4 and BA.5 have identical amino acid sequences in the spike protein and appear to have evolved from BA.2. ¹⁰ Compared with BA.2, BA.4/5 has residues 69 and 70 deleted and contains two additional substitutions in the receptor‐binding domain. Those Omicron sublineages form a cluster in the phenotypic cartography based on the susceptibility to monoclonal antibodies (Figure 1); however, they differ as much as formerly circulating VOCs and VOIs.

4. DISCUSSION

It is worth monitoring how SARS‐CoV‐2 evolves. The ongoing global pandemic of SARS‐CoV‐2 and the massive administration of vaccination could change the fitness landscape of the virus. ¹¹ , ¹² The present study showed that an analysis of data of neutralization assays with monoclonal antibodies succeeded in mapping the phenotypic relationships of SARS‐CoV‐2 variants. The cartography shows how those variants are alike or different from the point of view of the susceptibility to therapeutic monoclonal antibodies.

The Alpha variant disseminated globally and caused a surge of COVID‐19 cases in many countries from late 2020 to the beginning of 2021. The N501Y mutation in the spike protein associated with the variant is considered to be responsible for increased transmissibility, and that could also affect antigenicity. ⁴ Still, the vaccine effectiveness was maintained for the variant. ⁶ , ⁷ Likewise, the cartography based on the susceptibility to therapeutic monoclonal antibodies shows a close relationship between the wild‐type strain and the Alpha variant.

Interestingly, variants with a mutation at position 484 in the spike protein (Beta, Gamma, Zeta, IotaE484K, Kappa, Mu, and Omicron) are farther apart from the wild‐type strain. The efficacy of current vaccines for some of those variants was significantly reduced, ⁶ , ⁷ , ¹³ and the effectiveness of therapeutic antibodies decreased for them as well. ⁸ , ⁹ Those suggest a substantial effect of amino acid residue at the position on the effectiveness of therapeutic antibodies. It should be noted that the cartography developed in the present study was generated solely with phenotypic data; information about genetic sequence or amino acid substitutions was not used.

The results indicate that phylogenetic relationships cannot simply predict phenotypic differences. Moreover, convergent evolution can generate genetically distinct strains with a similar phenotype. For example, the Kappa variant is evolutionary related to the Delta variant. ¹⁴ However, it is close to the Beta and Gamma variants in the phenotypic cartography in this study, probably because of a mutation at position 484 in the spike protein.

The Omicron variant, the latest VOC, was first reported in November 2021. The variant could further reduce the effectiveness of immunity generated by vaccination or prior infection with earlier strains. ¹³ , ¹⁵ From the beginning of 2022, sister clades of the original Omicron variant, BA.2 and then BA.4/5, have prevailed in many countries. ¹⁰

There are a considerable number of mutations in the spike protein of the Omicron variant, and the variant is located away from the other variants in the phenotypic cartography in this study. Even among sublineages of the Omicron variant, each possesses unique mutations in the spike protein that could affect the susceptibility to neutralizing antibodies. ¹⁶ As a result, the Omicron sublineages were clearly distinct from each other on the map.

The antigenic maps of SARS‐CoV‐2 variants were recently reported by using convalescent sera from patients infected with different variants. ¹⁷ , ¹⁸ , ¹⁹ , ²⁰ Surprisingly, the phenotypic cartography developed in this study based on the efficacy of monoclonal antibodies corresponds very well with those antigenic maps. The fact suggests that this kind of cartography made by experimental data of multiple monoclonal antibodies can predict how in vivo humoral immunity works against different variants.

Still, the antigenic maps reported in other studies do not completely match the phenotypic cartography in this study. Compared to polyclonal antibodies from convalescent sera to assess antigenic distances, reactivity with monoclonal antibodies has a short range of dynamics for the characterization of viral variants. That is because when variants show complete resistance to particular monoclonal antibodies, the degree of difference between the variants cannot be measured using the antibodies.

Rapid, systematic assessment of continuously emerging SARS‐CoV‐2 variants is essential. The present study showed how we could assess and visualize the phenotypic characteristics of SARS‐CoV‐2 variants. The approach also has the potential to evaluate antigenic changes, predicting the variants’ ability to escape from immunity generated by prior infection or vaccination.

CONFLICT OF INTEREST

The author declares no conflict of interest.

ETHICS STATEMENT

No human subjects or animals were involved in this study.

Supporting information

Supporting information.

Furuse Y. Cartography of SARS‐CoV‐2 variants based on the susceptibility to therapeutic monoclonal antibodies. J Med Virol. 2022;95:e28275. 10.1002/jmv.28275

DATA AVAILABILITY STATEMENT

All data were available in previously published studies listed in Supporting Information: Table 1.