Proteome-wide analysis of differentially-expressed SARS-CoV-2 antibodies in early COVID-19 infection

Xiaomei Zhang, Xian Wu, Dan Wang, Minya Lu, Xin Hou, Hongye Wang, Te Liang, Jiayu Dai, Hu Duan, Yingchun Xu, Yongzhe Li, Xiaobo YumedRxiv 2020 · DOI

Abstract

Rapid and accurate serological tests are vital for managing viral outbreaks. This study utilized a custom SARS-CoV-2 proteome microarray to profile IgM and IgG antibody responses in patients during the early stages of COVID-19 infection. By comparing these responses against influenza and non-influenza controls, the researchers identified that the commonly used Nucleocapsid (N) protein is a suboptimal biomarker due to low early-stage immunogenicity. In contrast, specific domains of the Spike (S) protein, particularly the S1-RBD and S1+S2ECD, emerged as superior targets for diagnostic sensitivity. Furthermore, the study identified significant correlations between COVID-19-specific antibodies and clinical markers of acute myocardial injury, providing insight into the systemic nature of the disease.

Introduction

In the initial months of the COVID-19 pandemic, the global healthcare system faced a critical shortage of reliable diagnostic tools. While nucleic acid testing (RT-PCR) remained the gold standard, serological assays were needed to identify past infections and complement viral detection. However, early antibody tests often suffered from high false-negative rates, particularly in the first week of symptom onset. This research, published in May 2020, sought to map the human immune response across the entire SARS-CoV-2 proteome to identify the most effective antigens for early-stage diagnosis.

High-Throughput Proteome Mapping

To achieve a high-resolution view of the immune landscape, the researchers developed a SARS-CoV-2 proteome microarray. This tool allowed for the simultaneous screening of patient serum against 966 tiled peptides covering the entire viral proteome, alongside full-length Nucleocapsid (N) and Envelope (E) proteins and various truncated versions of the Spike (S) protein.

The study analyzed a cohort of 40 patients from Peking Union Medical College Hospital: 15 with confirmed early-stage COVID-19 (median 4 days post-symptom onset), 13 with influenza, and 12 with non-influenza respiratory symptoms. This experimental design was crucial for ensuring specificity, as it allowed the team to distinguish the COVID-19 immune signature from other common respiratory viruses that present with nearly identical symptoms.

The Nucleocapsid Fallacy vs. Spike Superiority

A significant finding of the study was the relative failure of the Nucleocapsid (N) protein as an early diagnostic marker. Many first-generation serology tests targeted the N protein, but the microarray data revealed that IgM antibodies against N could not reliably differentiate COVID-19 patients from those with the flu during the early phase of infection. This explains the high false-negative rates observed in early clinical settings.

Conversely, the Spike (S) protein—the "key" the virus uses to enter human cells—proved to be a far more effective target. The researchers proposed an "inside-out" logic for this: because S proteins are located on the exterior of the virion, they are immediately accessible to the immune system. In contrast, internal proteins like N and Orf1ab are likely only recognized by the immune system after the virus has begun replicating and lysing host cells, delaying the antibody response.

Specifically, the study identified:

  • S1-RBD (Receptor Binding Domain): The optimal antigen for detecting IgM ("first responder") antibodies.
  • S1+S2ECD (Extracellular Domain): The most effective antigen for detecting both IgM and IgG ("long-term memory") antibodies.

Clinical Implications and Cardiac Correlation

Beyond diagnostics, the study uncovered a provocative link between the immune response and clinical severity. The levels of specific anti-SARS-CoV-2 antibodies were found to correlate significantly (p≤0.05) with clinical indices of viral infection and acute myocardial injury, such as NT-proBNP and CKMB-mass. This suggests that the antibody response itself may contribute to the acute myocardial injury observed in severe COVID-19 cases, or at least serves as a proxy for the systemic inflammatory stress caused by the virus.

Diagnostic Performance and Epitope Mapping

The microarray approach demonstrated a clear performance advantage over contemporary commercial tests. By combining multiple IgM and IgG biomarkers, the researchers achieved identification rates of 86.7% and 73.3%, respectively. In comparison, the commercial Gold lateral flow, ELISA, and Chemiluminescence tests available at the time showed positive rates ranging from a mere 7% to 53% on the same patient samples.

Furthermore, the study mapped specific "hotspots" or linear epitopes, such as residues 816-820 on the S protein. These precise amino acid sequences provide a blueprint for developing highly targeted synthetic vaccines and next-generation diagnostic assays.

Conclusion and Limitations

This proteome-wide analysis provided a foundational map of the early immune response to SARS-CoV-2. While the study was limited by a small sample size and a focus on linear rather than conformational epitopes, its findings were instrumental in shifting the focus of diagnostic development toward the Spike protein. By identifying the most immunogenic viral fragments, this work paved the way for more accurate screening and a deeper understanding of the interplay between the immune system and COVID-19 pathology.

Summary updated

Abstract

Abstract Rapid and accurate tests that detect IgM and IgG antibodies to SARS-CoV-2 proteins are essential in slowing the spread of COVID-19 by identifying patients who are infected with COVID-19. Using a SARS-CoV-2 proteome microarray developed in our lab, we comprehensively profiled both IgM and IgG antibodies in forty patients with early-stage COVID-19, influenza, or non-influenza who had similar symptoms. The results revealed that the SARS-CoV-2 N protein is not an ideal biomarker for COVID-19 diagnosis because of its low immunogenicity, thus tests that rely on this marker alone will have a high false negative rate. Our data further suggest that the S protein subunit 1 receptor binding domain (S1-RBD) might be the optimal antigen for IgM antibody detection, while the S protein extracellular domain (S1+S2ECD) would be the optimal antigen for both IgM and IgG antibody detection. Notably, the combination of all IgM and IgG biomarkers can identify 87% and 73.3% COVID-19 patients, respectively. Finally, the COVID-19-specific antibodies are significantly correlated with the clinical indices of viral infection and acute myocardial injury (p≤0.05). Our data may help understand the function of anti-SARS-CoV-2 antibodies and improve serology tests for rapid COVID-19 screening.
BibTeX
@misc{zhang2020proteome,
  author = {Xiaomei Zhang and Xian Wu and Dan Wang and Minya Lu and Xin Hou and Hongye Wang and Te Liang and Jiayu Dai and Hu Duan and Yingchun Xu and Yongzhe Li and Xiaobo Yu},
  title = {{Proteome-wide analysis of differentially-expressed SARS-CoV-2 antibodies in early COVID-19 infection}},
  year = {2020},
  doi = {10.1101/2020.04.14.20064535},
  howpublished = {medRxiv},
}
Figure