SARS-CoV-2 proteome microarray for mapping COVID-19 antibody interactions at amino acid resolution

Hongye Wang, Xin Hou, Xian Wu, Te Liang, Xiaomei Zhang, Dan Wang, Fei Teng, Jiayu Dai, Hu Duan, Shubin Guo, Yongzhe Li, Xiaobo YubioRxiv 2020 · DOI

Abstract

The COVID-19 pandemic created an urgent need for high-resolution tools to understand the human immune response to SARS-CoV-2. While standard serological tests identify antibodies against whole proteins, they lack the granularity to pinpoint specific binding sites, or epitopes. To address this, we developed a high-throughput SARS-CoV-2 proteome microarray capable of mapping antibody interactions at amino acid resolution. By spotting 966 peptides (15 amino acids long with 5-amino acid offsets) representing the full viral proteome, we achieved a platform with a 1.5-hour processing time and a detection limit of 94 pg/mL. Our analysis reveals that existing SARS-CoV-1 antibodies cross-react with SARS-CoV-2, identifies immunodominant epitopes in patient serum, and highlights specific sequences that may neutralize viral entry via the ACE2 receptor.

Introduction

Early in the COVID-19 pandemic, the scientific community faced a significant challenge: while it was clear that patients produced antibodies against the virus, the specific targets of these antibodies across the viral proteome remained largely unknown. Most diagnostic tools focused on large proteins like the Spike (S) or Nucleocapsid (N) proteins. However, identifying the exact segments—epitopes—where antibodies bind is critical for developing precise diagnostics, effective vaccines, and therapeutic monoclonal antibodies.

Before this study, no immuno-proteomics platform existed that could perform such a proteome-wide analysis with high sensitivity and speed. This research introduces a peptide-based microarray designed to provide a high-resolution "atlas" of the antibody landscape at the level of individual amino acids.

Methodology and Platform Design

The core of this study is a custom-designed peptide microarray. To ensure comprehensive coverage of the SARS-CoV-2 genome (Accession no. MN908947.3), the researchers synthesized a library of 966 peptides.

Tiling Strategy and Chemistry

Each peptide is 15 amino acids long. To achieve high resolution, the team used a "tiling" or "offset" strategy where each subsequent peptide overlaps the previous one by 10 amino acids (a 5-amino acid offset). This ensures that no potential binding site is missed. These peptides were labeled with biotin and printed onto 3D-modified glass slides using biotin-streptavidin chemistry, which ensures stable and uniform orientation of the fragments.

Optimization for Speed and Sensitivity

A major technical achievement of this platform is its efficiency. Traditional protein microarray protocols often require up to 18 hours of processing time. By optimizing the workflow, the researchers reduced this to just 1.5 hours. Furthermore, the array demonstrated high sensitivity with a limit of detection (LOD) of 94 pg/mL and exceptional reproducibility, showing an r-correlation of 0.9992 within arrays.

Key Findings

1. Leveraging SARS-CoV-1 Reagents

One of the most immediate practical findings was the cross-reactivity of antibodies. The study tested commercially available rabbit monoclonal and polyclonal antibodies originally designed for the 2003 SARS-CoV-1 virus. The results showed that these reagents could effectively target SARS-CoV-2 proteins. This discovery allowed researchers to use existing, validated tools for COVID-19 research immediately, rather than waiting months for the development of new specific antibodies.

2. Identifying Immunodominant Epitopes

By screening serum samples from 10 COVID-19 patients, the study profiled IgM and IgG antibody responses across the entire proteome. They identified:

  • 32 epitopes on the Orf1ab polyprotein.
  • 8 epitopes on the Spike protein.
  • 8 epitopes on the Nucleocapsid protein.

Four specific epitopes were found to trigger an antibody response in over 80% of the patients, marking them as "hotspots" for the human immune system. These hotspots are primary candidates for the development of highly sensitive diagnostic tests.

3. Neutralization Potential: The FRKSN Sequence

By mapping identified epitopes onto 3D protein structures (using PDB IDs: 6VYO, 6VXX, 6M17), the researchers identified a critical sequence within the Spike protein’s Receptor Binding Domain (RBD). The sequence FRKSN (residues 456-460) sits precisely at the interface where the virus binds to the human ACE2 receptor. Antibodies that target this specific epitope could potentially neutralize the virus by physically blocking its ability to enter host cells.

Discussion and Limitations

The ability to map epitopes onto 3D models also revealed that while many immunogenic sites are located on the surface of the virus, some are buried within the protein structure. This suggests that the immune system may encounter these internal segments during viral degradation or when proteins are in the process of folding.

Despite its high resolution, the platform has limitations:

  • Linear vs. Conformational Epitopes: Because the array uses short, linear peptides, it may not detect antibodies that recognize "conformational epitopes"—shapes formed by the complex folding of a full protein. To mitigate this, the researchers included some full-length proteins on the array.
  • Viral Evolution: The library was based on the initial reference genome. As SARS-CoV-2 continues to mutate, the peptide library must be updated to account for new variants.

Conclusion

The SARS-CoV-2 proteome microarray provides a powerful tool for understanding the humoral immune response in unprecedented detail. By identifying specific immunodominant regions and potential neutralizing targets like the FRKSN epitope, this platform offers invaluable data for the design of next-generation vaccines and precision diagnostics. Future work will involve applying this technology to larger patient cohorts to determine if specific antibody "signatures" can predict disease severity and clinical outcomes.

Summary updated

Abstract

Abstract COVID-19 has quickly become a worldwide pandemic, which has significantly impacted the economy, education, and social interactions. Understanding the humoral antibody response to SARS-CoV-2 proteins may help identify biomarkers that can be used to detect and treat COVID-19 infection. However, no immuno-proteomics platform exists that can perform such proteome-wide analysis. To address this need, we created a SARS-CoV-2 proteome microarray to analyze antibody interactions at amino acid resolution by spotting peptides 15 amino acids long with 5-amino acid offsets representing full-length SARS-CoV-2 proteins. Moreover, the array processing time is short (1.5 hours), the dynamic range is ~2 orders of magnitude, and the lowest limit of detection is 94 pg/mL. Here, the SARS-CoV-2 proteome array reveals that antibodies commercially available for SARS-CoV-1 proteins can also target SARS-CoV-2 proteins. These readily available reagents could be used immediately in COVID-19 research. Second, IgM and IgG immunogenic epitopes of SARS-CoV-2 proteins were profiled in the serum of ten COVID-19 patients. Such epitope biomarkers provide insight into the immune response to COVID-19 and are potential targets for COVID-19 diagnosis and vaccine development. Finally, serological antibodies that may neutralize viral entry into host cells via the ACE2 receptor were identified. Further investigation into whether these antibodies can inhibit the propagation of SARS-CoV-2 is warranted. Antibody and epitope profiling in response to COVID-19 is possible with our peptide-based SARS-COV-2 proteome microarray. The data gleaned from the array could provide invaluable information to the scientific community to understand, detect, and treat COVID-19.
BibTeX
@misc{wang2020sars,
  author = {Hongye Wang and Xin Hou and Xian Wu and Te Liang and Xiaomei Zhang and Dan Wang and Fei Teng and Jiayu Dai and Hu Duan and Shubin Guo and Yongzhe Li and Xiaobo Yu},
  title = {{SARS-CoV-2 proteome microarray for mapping COVID-19 antibody interactions at amino acid resolution}},
  year = {2020},
  doi = {10.1101/2020.03.26.994756},
  howpublished = {bioRxiv},
}
Figure