Discovery of host-directed modulators of virus infection by probing the SARS-CoV-2-host protein-protein interaction network

Abstract

The ongoing coronavirus disease 2019 (COVID-19) pandemic has highlighted the need to better understand virus-host interactions. We developed a network-based method that expands the severe acute respiratory syndrome coronavirus-2 (SARS-CoV-2)-host protein interaction network and identifies host targets that modulate viral infection. To disrupt the SARS-CoV-2 interactome, we systematically probed for potent compounds that selectively target the identified host proteins with high expression in cells relevant to COVID-19. We experimentally tested seven chemical inhibitors of the identified host proteins for modulation of SARS-CoV-2 infection in human cells that express ACE2 and TMPRSS2. Inhibition of the epigenetic regulators bromodomain-containing protein 4 (BRD4) and histone deacetylase 2 (HDAC2), along with ubiquitin-specific peptidase (USP10), enhanced SARS-CoV-2 infection. Such proviral effect was observed upon treatment with compounds JQ1, vorinostat, romidepsin and spautin-1, when measured by cytopathic effect and validated by viral RNA assays, suggesting that the host proteins HDAC2, BRD4 and USP10 have antiviral functions. We observed marked differences in antiviral effects across cell lines, which may have consequences for identification of selective modulators of viral infection or potential antiviral therapeutics. While network-based approaches enable systematic identification of host targets and selective compounds that may modulate the SARS-CoV-2 interactome, further developments are warranted to increase their accuracy and cell-context specificity.

Keywords: SARS-CoV-2; cell context specificity; host modulators; network prioritization; protein–protein interactions.

Figure 1

Topological network analysis of the SARS-CoV-2 PPI network. (A) The giant component of the SARS-CoV-2–host PPI network that contains 297 VIPs and 3678 non-VIPs. The inset shows a subnetwork of select VIP nodes. When calculating the neighbourhood size from VIP to non-VIP nodes, the connections marked in red show the shortest path distance from the MARK1 VIP node to NOTCH2 node in the network (neighbourhood size of 3). (B) Degree, i.e. the number of connections, for VIP and non-VIP nodes (the bars show mean ± SEM). (C) Betweenness centrality, i.e. the number of shortest paths through a given node, for VIP and non-VIP nodes (mean ± SEM). (D) Operation of the RWR (RWR) algorithm. The red seed nodes are the set of 298 VIPs from which the random walker starts exploring the network (marked by red arrows). After iterating through all nodes in the network, a probability score is assigned to all nodes in the network, ranked from the highest to the lowest probability, which was used to identify 200 top-ranked NIP nodes.

Figure 2

GO biological process and KEGG pathways that are commonly or uniquely enriched among the NIP and VIP targets. The enrichment P-values were corrected for multiple testing with Benjamini–Hochberg test, and the pathways with adjusted P < 0.05 were considered significant. The dots are colour-coded based on their corresponding adjusted P-values and the dot size corresponds to the gene ratio (i.e. genes of interest in the GO term/total number of genes of interest.

Figure 3

Identification of potent compounds that inhibit the target proteins based on ChEMBL database (bioactivity <1000 nM, see Materials and Methods). (A) The number of compounds identified for the 27 VIP targets (left histogram), and the number of VIP targets for each potent compound (right histogram). (B) The number of compounds identified for the 25 NIP targets (left), and the number of NIP targets for each potent compound (right). (C) Clinical development phase of compounds that target VIPs. (D) Clinical development phase of compounds that target NIPs. (E) Anatomical Therapeutic Chemical (ATC) classification of the 101 approved compounds that target select VIPs and NIPs.

Figure 4

Protein expression in cells of upper and lower respiratory tract across the (A) 27 VIP targets and (B) 25 NIP targets. The expression classes originate from the Human Protein Atlas (colour legend).

Figure 5

(A–C) Compounds that inhibit HDACs and USP10/13 enhance CPE during virus infection in 293TAT cells (i.e. appear to have proviral effects). Cells were treated with the indicated concentrations of compounds for 2 h prior to infection with SARS-CoV-2 at a MOI of 0.01. Parallel wells contained cells treated only with compounds to study toxicity. Forty-eight hours post infection, cell viability was assessed using the CellTiter-Glo assay and effects on virus-infected cells (efficacy) and non-infected cells (viability) were calculated as described in Materials and Methods. Data points reflect average and standard deviations of triplicate experiments per condition. (D-F) Confirmatory assays in Calu 3 cells. Cells were treated with the indicated concentrations of compounds for 2 h prior to infection with SARS-CoV-2 at a MOI of 0.1. Parallel wells contained cells treated only with compounds. Ninety-six hours post infection, cell viability was assessed using the CellTiter-Glo and effects on virus infection (efficacy) and viability were calculated as described in Materials and Methods. Data points reflect average and standard deviations of triplicate experiments per condition.

Figure 6

Spautin 1, vorinostat and JQ1 enhance SARS-CoV-2 infection. 293TAT cells were treated with the indicated concentrations of compounds prior to infection with SARS-CoV-2 at a MOI of 0.01. Parallel wells contained cells treated only with compounds. Twenty-four hours post infection, RNA was extracted from one plate and the relative abundance of viral RNA was determined by qRT-PCR (A and C). Forty-eight hours post infection, cell viability was assessed using the CellTiter-Glo assay and effect on virus-infected cells (efficacy) and non-infected cells (viability) were calculated (B and D). Data points in A and B reflect average and standard deviations of triplicate and quadruplicate experiments per condition, respectively. Statistical testing was done with a Student’s t-test (two-tailed).

Figure 7

Network representation of selected targets and their protein–protein and compound–protein interactions with the validated compounds. Targets and their nearest neighbours are shown in the network.