1 Introduction

Since the emergence of the Delta variant of SARS-CoV-2 in May 2021, more than 155 million medically confirmed cases and approximately 3.24 million deaths occurred in more than 180 countries as a result of the COVID-19 pandemic [1]. The SARS-CoV-2/2019-nCoV novel coronavirus strain causes this respiratory disease [2, 3]. Repurposing of drugs has been a promising strategy at the forefront of strategies to address the continually growing number of COVID-19 cases [4,5,6,7,8]. In the current pandemic situation, drug repurposing should be considered a new avenue for the treatment of COVID-19 [5]. The coronavirus virion consists of structural proteins, namely spike (S), envelope (E), membrane (M), nucleocapsid (N) and, for some beta coronaviruses, hemagglutinin esterase [9]. The genomic structure of SARS-CoV-2 contains short untranslated regions and N-3′, E, M, 5′-replicase (rep)-S, identical to the genome structures of other coronaviruses [10]. As in other coronaviruses, the N protein is a crucial structural component of SARS-CoV-2 [11]. SARS-CoV-2 N has 90% sequence identity to the Severe Acute Respiratory Syndrome coronavirus N protein [12].

The N protein, which is naturally found within the virus, is the most conserved structural protein. It is required for viral replication and transcription, as it binds the viral RNA genome [13]. The N protein structure comprises an N-terminal RNA-binding domain (NTD), a C-terminal domain (CTD), and a naturally disordered central Arg/Ser-rich linker. For the SARS-CoV-2 N protein, each NTD molecule adopts a right-handed fist shape. The core sub domain consists of a five-stranded U-shaped antiparallel β-sheet with β4–β2–β3–β1–β5 topology, sandwiched between two short α-helices (α1 before the β2 strand and α2 after the β5 strand), and a protruding β-hairpin (β2′–β3′) is composed of mostly basic amino acid residues. New COVID-19 drug targets have been identified in the SARS-CoV-2-human interactome recently published by Gordon et al. [14]. Virtual screening and molecular docking have led to the identification of various inhibitors of SARS-CoV-2 [15].

Among the 332 interactions previously identified between viral and host proteins, most involve the innate immune signaling pathway [13, 16]. With this knowledge, researchers have identified an anti-N drug chain that shows excellent promise as a treatment for COVID-19 [17]. Through pathway analysis, a series of anti-N drugs with high potential to combat COVID-19 have been identified. Interestingly, some of the drugs target the N protein, which has been suggested to be a viable target for antiviral drug development [14, 16]. Yellow color indicates human drug target interaction with SARS-CoV-2 (Fig. 6A). G3BP1, G3BP2, and LARP1 human proteins interact with N of SARS CoV-2 and are drug response proteins. Host proteins are involved in RNA splicing, viral defense, ribosome biogenesis in eukaryotes, metabolism of RNA, and mRNA catabolic processes (Fig. 6B).

Fig. 6
figure 6

A Interactions between SARS-CoV-2 N and human proteins. B Bar graph of enriched terms across input gene lists, colored by p-values

Full size image

3.6 Identification of Potential Drug–Target Interactions

The STITCH database was used as a model for drug–protein interactions through a statistical approach. STITCH incorporates data from 2031 genomes on more than 5 million interactions between 430,000 chemicals and 9.6 million proteins. To predict protein–drug interactions, it primarily relies on keyword mining of the literature and experimental evidence. The likelihood that the expected relationship occurs is indicated by a confidence score (0–0.9). A confidence score of 0.9 or higher was used to discover the targets. To construct a network based on binding affinities (\({K}_{i}\) of protein–drug interactions with thickness of edges between nodes, increasing as \({K}_{i}\) value increases), we used STITCH, which is based on STRING v10 [38]. Figure 7 shows the STITCH predictions for the drug–gene relationships among the top five strongly binding drugs. Some of these drugs had relatively fewer high affinity binding targets. Prominent proteins included YES1, RET, FGR, and FGFR1/2/3, which are involved in cancer and associated pathways as well as endocytosis, and dovitinib has been found to target some of these proteins (Fig. 7A). Fedratinib interacted with the JAK2/JAK1 proteins, which are part of the JAK-STAT signaling pathway (Fig. 7B).

Fig. 7
figure 7

Drug–gene network constructed with STITCH. The edge width of protein–drug interactions is scaled according to the binding affinity between the drug and the protein. The drug–protein networks of A dovitinib, B fedratinib, C luteolin, D nintedanib, E rapamycin, and F silmitasertib

Full size image

Luteolin, a flavone, had a low affinity for the proteins CASP3, JUN, CDK2, FOS, and MAPK8, which are involved in the TNF signaling pathway and cancer (Fig. 7C). However, certain drugs had several predicted protein interactions and interacted via various pathways. For example, nintedanib targets the Ras signaling pathway, cancer, and cytokine–cytokine receptor interaction proteins MAP3K7, JAK1, PDGFRA/B, LCK, KIT, MELK, FLT3, and KDR (Fig. 7D). Similarly, sirolimus/rapamycin, a kinase inhibitor, interacts with many targets in the mTOR and AMPK signaling pathways, and with proteins involved in immunosuppression, including mTOR, FKBP1A, and FKBP5 (Fig. 7E). Silmitasertib interacts with CSNK2A1, a serine/threonine kinase protein involved in cell cycle progression, apoptosis, transcription, and viral infection (Fig. 7A). Another drug, foretinib, is involved in endocytosis and focal adhesions, and has similar targets to nintedanib and dovitinib.

4 Discussion

As the RNA binding activity of N protein is essential for viral ribonucleoprotein formation and genome replication, blocking the RNA binding of the N-NTD has been demonstrated to be a potential treatment strategy. N protein is an essential RNA-binding protein with crucial roles in replication and transcription of viral RNA. An overall right-handed fold with a β-sheet core is found between loops, as revealed by recently solved crystal structures of the SARS-CoV-2 N-NTD (PDB ID: 6M3M) and C-CTD (PDB ID: 6WJI). The core region of the β-sheet consists of five antiparallel β-strands with a β6–β2–β5–β1–β7 topology flanked by a single short α-helix just before strand β2, and a protruding β-hairpin (β3 and β4) between strands β2 and β5. In addition, an NMR structure of the SARS-CoV-2 N-NTD in complex with RNA (PDB ID: 6YI3) suggested putative RNA binding sites of A51, T58, H60, R93, I95, L105, S106, R108, R150, and Y173. Moreover, the adenosine monophosphate (AMP) binding site has been structurally characterized in HCoV-OC43 N-NTD by Lin et al. [22]. N49, A51, S52, A56, R89, R108, Y110, Y112, and R150 compose the AMP binding site, as indicated by the structural superposition between the SARS-CoV-2 N-NTD and HCoV-OC43 N-NTD-AMP. Therefore, we extended our investigation by using structure-based molecular docking of N-NTD with different drugs to gain insights into the structural and molecular regions’ potential effectiveness in antiviral drug therapy. The investigated drugs included protein biogenesis inhibitors, anticancer compounds, antiinflammatory compounds, mTOR inhibitors, and stress granule modifiers. The docking results showed that 5 of the 20 drugs bound with strong binding affinity. Among N-NTD inhibitors, the key residues involved in binding were near the helix, i.e., Trp53, Ile75, Asn76, and Thr77; in the β3 strand, i.e., Arg93 and Arg94; and at the C-terminal interface, i.e., Ala153, Ala156, Ile158, Val159, and Gln161.

Silmitasertib is an antiviral drug that has been tested against the N protein of SARS-CoV-2, and found to block the CK2 and enhance SGs formation [39], thus inhibiting SARS-CoV-2 proliferation in vitro. Recently, Taiwan-headquartered Senhwa Biosciences Inc and the US National Institutes of Health (NIH) collaborated in analyzing the effectiveness of silmitasertib for the treatment of COVID-19 (https://www.biospectrumasia.com/news/34/15848/senhwa-biosciences-nih-to-co-develop-COVID-19-drug.html). The drug showed promise in controlling the proliferation of this RNA virus in human clinical tests. Silmitasertib was developed by Senhwa Biosciences to treat cancers, such as pediatric brain tumors, medulloblastoma, and bile duct cancer.

The second molecule tested was nintedanib, a tyrosine kinase inhibitor used to treat idiopathic pulmonary fibrosis or interstitial lung disease [40]. Very recently, the safety and efficacy of nintedanib ethanesulfonate have been analyzed in treating pulmonary fibrosis in patients with mild-to-extreme COVID-19. A placebo-controlled, a single-center, randomized study has been initiated and is currently in a phase 2 clinical trial (ClinicalTrials.gov identifier: NCT04338802).

In addition, the viral translation inhibitors ternatin and zotatifin, which is an FDA-approved drug for the treatment of multiple myeloma, have demonstrated the strongest binding affinity [41]. Plitidepsin is structurally similar to ternatin and is currently undergoing a clinical trial in COVID-19. The flavone luteolin, an antiinflammatory molecule, has broad antiviral properties [42, 43]. Previous studies have shown that luteolin inhibits SARS-CoV S protein and 3CL protease [44, 45]. Recently, both luteolin and quercetin have been identified through virtual screening and molecular docking as the best possible SARS-CoV-2 inhibitors [46, 47]. Furthermore, through SUMMIT, the world's most powerful supercomputer, high-throughput screening of small molecules interacting with the SARS-CoV-2 S protein or S protein–human ACE2 interface have recently been reported. Eriodictyol, a structural analog of luteolin, has been found to be a potential inhibitor of SARS-CoV-2 [48]. The last drug with considerable binding affinity toward N protein is fedratinib, an antiinflammatory JAK2 inhibitor. Wu et al. [49] have reported that fedratinib suppresses the expression of IL17, IL 22, and L23 in murine TH17 cells, and suggested that the drug may help mitigate the cytokine storm associated with SARS-CoV-2 infection. Stebbing et al. [50], through in silico artificial intelligence, have predicted significant beneficial effects of the antiinflammatory agents baricitinib, fedratinib, and ruxolitinib in the treatment of COVID-19.

The drug foretinib is an anticancer agent that inhibits vascular endothelial growth factor receptor (VEGFR) and hepatocyte growth factor receptor (HEGFR or MET) receptor. A recent study has reported that foretinib (DB12307) is a strong binder, on the basis of analysis through in silico virtual screening and molecular docking of 8548 ligands on the SARS-CoV-2 endoribonuclease NendoU (PDB ID: 6VWW) [51]. Our findings also have suggested that this drug binds the nucleocapsid protein with high binding affinity. Thus, foretinib may be repurposed as a broad-spectrum drug and tested against COVID-19 in the future.

Intriguingly, the docking of the N-NTD and C-CTD proteins revealed promising results for all five drugs tested. Notably, fedratinib and luteolin bind the ribonucleotide binding site and thus can inhibit RNA binding of the protein. Similarly, silmitasertib and nintedanib are positioned at the interface of two monomers and thus can impair the oligomerization of the protein. We recommend further experimental investigation of these compounds.

5 Conclusion

The highly immunogenic and abundant nature of the N protein makes it a novel target to treat infection of the respiratory system by SARS-CoV2. We extended the investigation of drug efficacy, stimulated by the recent SARS-CoV-2-host interactome and identification of several anti-N drugs, by using computational analysis. In this study, binding modes were chosen, and the most common anti-N drugs were selected. The probable molecular underpinnings of their effectiveness against COVID-19 have been identified. The docking results indicated that 5 of 20 anti-N inhibitors bind with the energetic landscape of a protein–drug complex and have high thermodynamic scores. The identified drugs have been shown to bind the ribonucleotide binding pocket and protein interface of the N-NTD and C-CTD, thereby suggesting mechanisms of action. Thus, the identification of compounds that bind the N-NTD and C-CTD and interfere with NTD–RNA and NTD–NTD interactions may assist in the development of broad-spectrum antiviral therapeutics.