Mass spectrometry-based proteomic platforms for better understanding of SARS-CoV-2 induced pathogenesis and potential diagnostic approaches

Abstract

While protein-protein interaction is the first step of the SARS-CoV-2 infection, recent comparative proteomic profiling enabled the identification of over 11,000 protein dynamics, thus providing a comprehensive reflection of the molecular mechanisms underlying the cellular system in response to viral infection. Here we summarize and rationalize the results obtained by various mass spectrometry (MS)-based proteomic approaches applied to the functional characterization of proteins and pathways associated with SARS-CoV-2-mediated infections in humans. Comparative analysis of cell-lines versus tissue samples indicates that our knowledge in proteome profile alternation in response to SARS-CoV-2 infection is still incomplete and the tissue-specific response to SARS-CoV-2 infection can probably not be recapitulated efficiently by in vitro experiments. However, regardless of the viral infection period, sample types, and experimental strategies, a thorough cross-comparison of the recently published proteome, phosphoproteome, and interactome datasets led to the identification of a common set of proteins and kinases associated with PI3K-Akt, EGFR, MAPK, Rap1, and AMPK signaling pathways. Ephrin receptor A2 (EPHA2) was identified by 11 studies including all proteomic platforms, suggesting it as a potential future target for SARS-CoV-2 infection mechanisms and the development of new therapeutic strategies. We further discuss the potentials of future proteomics strategies for identifying prognostic SARS-CoV-2 responsive age-, gender-dependent, tissue-specific protein targets.

Keywords: COVID-19; biomarkers; comparative proteomics; kinase-substrate signaling; post-translational modifications; targeted proteomics; top-down proteomics.

FIGURE 1

Proteomic platforms utilized for the analysis of SARS‐CoV‐2 and COVID‐19 samples. Datasets collected from 29 published papers were used in this comparative analysis (see details in Table S2). A, bar diagram of the total proteome, phosphoproteome and/or interactome datasets; number of differentially abundant proteins (DAP) identified in each study are indicated by red circles. B, Venn diagram analysis of unique and overlapped number of DAPs reported in five cell lines, nine tissue‐specific and three extracellular proteomic datasets showed in panel A. C, the common (red circle) and unique protein kinases identified in proteome, phosphoproteome and interactome datasets. D, the KEGG pathway analysis of the 43 kinases shared across the three different proteomic platforms

FIGURE 2

Heat map analysis of over 11,000 proteins identified as differentially abundant proteins (DAPs) and/or SARS‐CoV‐2 interacting proteins from a total of 29 individual proteomic studies. The right panel shows the list of 23 proteins identified in at least in 10 individual studies. Red star shows ephrin receptor A2 (EPHA2), a tyrosine kinase was identified by 11 studies including all three proteomic platforms (comparative, phosphoproteomics and immunoproteomics). See details in Table S2.

FIGURE 3

Phosphorylation of SARS‐CoV‐2 proteins. The heat map shows the phosphorylation sites identified by five large‐scale phosphoproteomic analyses. Proteins identified as phosphorylated in at least two studies were included. Green, blue and orange correspond to pS, pT and pY modifications, respectively. Gray indicates sites were not identified

FIGURE 4

Post‐translational modifications on the surface of two SARS‐CoV‐2 proteins. (A) The N‐terminal domain (NTD) and C‐terminal domain (CTD) of the N protein; and (B) The S protein are depicted without and with posttranslational modifications (gray). Glycans are shown in red, phosphorylation in orange, palmitoylation in green, and methylation in yellow. Methylation and phosphorylation are represented as spheres. The domains of the N protein are built from two structures, PDB ID: 6YI3 and 6ZCO, and glycosylation sites were modeled after Supekar et al. [59]. The NTD and CTD contain 3 and 2 glycans, respectively; the NTD contains a total of 8 modeled phosphorylation sites. The CTD exists as a dimer, which are depicted with different shades of gray. The non‐glycosylated structure of the S protein based on PDB‐ID 6VXX is available through CHARMM‐GUI [60] and Jo et al. [61], and methylation was modeled based on Sun et al. [62]. The S protein contains 23 glycosylation sites, 2 palmitoylation sites, and 5 methylation sites on each chain of the trimer. The phosphate heads of the membrane that the spike is embedded in are shown as spheres

FIGURE 5

Potential bottom‐up proteomics platforms for elucidation of the molecular mechanisms of SARS‐CoV‐2 infection. A, cell‐line based phosphoproteomic approach with or without potential drugs treatments to SARS‐CoV‐2 and mutant variants. B, study of non‐human primate models infected by SARS‐CoV‐2 and mutant variants with or without drug response for large‐scale tissue‐specific proteome and phosphoproteome analysis. C, tissue‐specific proteomic analysis of COVID‐19 patient samples from diverse cohorts. D, identification of SARS‐CoV‐2 mutant proteins interacting partners using complementary methodologies. E, identification and verification of kinase‐specific targets and modification sites using kinase substrate assay followed by mass spectrometry analysis. This figure is created using BioRender.com