key: cord-0995060-0t9ce5ep
authors: Xu, Xuan; Zhang, Qing-Ye; Chu, Xin-Yi; Quan, Yuan; Lv, Bo-Min; Zhang, Hong-Yu
title: Facilitating Antiviral Drug Discovery Using Genetic and Evolutionary Knowledge
date: 2021-10-20
journal: Viruses
DOI: 10.3390/v13112117
sha: fe6df3f71b7cc386b67e96f5e8cbd6bfa0e9ab6c
doc_id: 995060
cord_uid: 0t9ce5ep

Over the course of human history, billions of people worldwide have been infected by various viruses. Despite rapid progress in the development of biomedical techniques, it is still a significant challenge to find promising new antiviral targets and drugs. In the past, antiviral drugs mainly targeted viral proteins when they were used as part of treatment strategies. Since the virus mutation rate is much faster than that of the host, such drugs feature drug resistance and narrow-spectrum antiviral problems. Therefore, the targeting of host molecules has gradually become an important area of research for the development of antiviral drugs. In recent years, rapid advances in high-throughput sequencing techniques have enabled numerous genetic studies (such as genome-wide association studies (GWAS), clustered regularly interspersed short palindromic repeats (CRISPR) screening, etc.) for human diseases, providing valuable genetic and evolutionary resources. Furthermore, it has been revealed that successful drug targets exhibit similar genetic and evolutionary features, which are of great value in identifying promising drug targets and discovering new drugs. Considering these developments, in this article the authors propose a host-targeted antiviral drug discovery strategy based on knowledge of genetics and evolution. We first comprehensively summarized the genetic, subcellular location, and evolutionary features of the human genes that have been successfully used as antiviral targets. Next, the summarized features were used to screen novel druggable antiviral targets and to find potential antiviral drugs, in an attempt to promote the discovery of new antiviral drugs.

Since the earliest yellow fever virus was discovered in 1901, the number of viruses has continued to increase. On average, three to four new species are found each year. So far, more than two hundred viruses that infect humans have been identified [1] . Human viral infection is the most unpredictable disease among infectious diseases. Currently, there are several viruses with high mortality rates in the world. For instance, Coronavirus disease (COVID-19) caused by SARS-CoV-2 has led to a global pandemic and can result in a series of respiratory diseases, such as pneumonia and lung failure [2] [3] [4] ; Ebola virus causes a range of highly pathogenic symptoms, such as hemorrhagic fever, and is highly contagious, with a mortality rate of 57%-90% [5] ; Severe Acute Respiratory Syndrome (SARS), has affected more than 8000 people and killed 774 people by July 2003 [6] ; Epstein-Barr virus (EBV), a common human herpes DNA virus, can lead to lifelong infection in more than 90 percent of the population, and causes a variety of human malignancies [7] ; Human Immunodeficiency Viruses (HIV/AIDS) causes severe defects in human cellular immune function. As of 2018, 77.3 million people worldwide have been infected with HIV, and 35.4 million people have died from AIDS-related diseases [8] . Thousands of human plagues mutation rate of the host protein is significantly lower than that of the virus, host-targeted antiviral drugs are thought to offer a higher genetic barrier to against mutations than directacting antiviral drugs [35] . Ullah et al. recently discovered that the inhibitor of human RACK1 protein shows similar efficacy to the approved anti-herpes drug acyclovir, and that RACK1 protein is a potential target for the development of broad-spectrum antiviral drugs [36] . Tyrrell et al. selected the endoplasmic reticulum a-glucosidases as a target based on the glycosylation mechanism, and iminosugars were shown to have potential efficacy in difficult-to-treat infections [37] . Up to now, this strategy has also been successful in the treatment of DEN [38] and other viral diseases [39] [40] [41] . Burmeister et al. proposed a treatment method based on ecological evolution systems. The virus microorganisms and the host's internal environment are regarded as an ecosystem, and host and microbial pathogens are part of this complex interaction system [42] . It can be seen that the strategy of targeting host factors circumvents the barrier created by the high-frequency mutation of viruses, and provides a unique opportunity for the development of more broad-spectrum antiviral drugs [42] .

On the other hand, recent studies have also found that successful drug targets often have common genetic and evolutionary features, which are of great help in screening potential drug targets and discovering new drugs [43] [44] [45] . Inspired by the above information, in this study we first summarized the genetic characteristics, cellular sublocation, and evolutionary features of existing approved host-targeted antiviral drug targets. Next, based on the summarized features, potential host-aimed antiviral targets were screened, in an attempt to promote the development of new antiviral drugs.

For viruses, the primary barrier to invasion is the cell's innate immunity. This antiviral response first detects viral components through host cells, then transduces the antiviral signals, transcription and translation of antiviral effectors, and finally establishes the body's antiviral state [46] . Combined with the host-targeted strategy, much progress has been made in identifying specific human genetic variations that help to increase sensitivity or resistance to viral diseases [47] . Currently, the most popular genetic testing methods focus on the discovery of a disease's potentially susceptible genes through GWAS [48] [49] [50] and the detection of genetic polymorphisms related to the phenotype of specific viral diseases through whole-exome sequencing (WES) [51, 52] . For example, Akcay et al. recently reported that HLA-DP/DQ mutations contribute to HBV persistence and resistance to vaccines. Besides, susceptible alleles in the HLA-DP/DQ locus are more common in Asian populations than in Whites, which may explain why chronic HBV infection rates are higher in Asian populations [50] . Additionally, Hashemi et al. declared that host-targeted antiviral drugs may be subject to host gene polymorphisms that alter their ability to inhibit the function of target proteins [53] . Indeed, in the clinical setting, suboptimal response to Alisporivir has been reported in 10-15% of patients, which may be due to the potential influence of single nucleotide polymorphisms (SNPs) [54] . Another study tested whole exon mutations at the individual level and showed that the progression of HIV-1 disease is related to microbial infection and innate immune activation, and the role of microbes in the damaged intestinal mucosa is thought to be related to disease progression. Therefore, microbial translocation and the activation of innate pattern recognition receptors (PPR) may affect disease progression and non-AIDS complications [55] . Gene mutations in EC 004 (TAB2), LTNP 005 (PIK3C2B), LTNP 007 (FGD6), and LTNP 008 (PRKDC) are also particularly significant for the slow-progressing phenotype of HIV [55] . These genes, as well as the immunological and cellular biological mechanisms involved in them, could be the subject of further research on the pathogenesis and disease progression of HIV, and may also become potential new drug targets. In addition, based on a genome-wide loss-of-function CRISPR screen, Daniloski et al. obtained highly ranked genes in host. The dysfunction of these genes significantly reduces the infection rate of SARS-CoV-2. Moreover, nine of these genes are druggable genes corresponding to 26 small molecule inhibitors. The experimental results showed that seven of the inhibitors can reduce the SARS-CoV-2 viral load by more than 100-fold. Interestingly, among these seven effective antiviral agents, four target the same gene, PI3K catalytic subunit type 3 (PI3KC3) [56] . Therefore, the authors contended that CRISPR is an effective genetic method for screening antiviral targets [56] [57] [58] . Furthermore, since GWAS can only identify genes that are significantly associated with the disease, there is an increasing number of studies incorporating MR to identify the causal relationship between gene mutations and disease. By combining GWAS and MR methods, some host genes, such as IFNAR2, TYK2, JAK1, JAK2, TNF, IL6, TMPRSS2, and ILIR1 have been identified as potential therapeutic targets for COVID-19 [34, 59] . These findings provide important clues for a better understanding of the mechanism of COVID-19. Taken together, genetics-identified disease genes are considered a promising source of drug targets to help increase the success rate of antivirus drug development [28] .

In a previous study [28] , we combined the information from eight disease gene databases (including Clinvar [60] , Online Mendelian Inheritance in Man (OMIM) [61] , The Human Gene Mutation Database (HGMD) [62] , Orphanet [63] , GWASdb [64] , INtegrated TaRget gEne PredItion (INTREPID) [65] , Genetic Association Database (GAD) [66] , and DisGeNET [67] ) and used the degree of clinical support to award these genes different quality scores (termed druggability score). Through a comprehensive analysis, it was confirmed that these disease genetics-derived druggability scores can be used to measure the association between genes and diseases, which offers important values for screening potential targets and predicting drug activity [28] . The higher the score, the stronger the association between the gene and the disease, and the easier it is to develop the gene as a drug target.

To analyze the genetic features of antiviral drug targets, firstly, 36 known host targets of approved antiviral drugs were downloaded from DrugBank [68] (Table S1 ). Based on the druggability scores derived from the SCG-Drug database (Quan et al. 2019), 32 targets had relevant score information (accounting for 88.9% of the total 36 targets). Next, the median score of these targets with 34 viral diseases was calculated to measure the strength of the association between genes and pathogenic viruses. The results showed that 13 genes received genetic scores equal to or greater than three. Besides, 27 targets were documented in association with more than one viral disease, as shown in the last column of Table 1 . Genes located on the cell membrane are shown in bold. * Complete candidate targets and diseases association information are listed in Table S1 . Disease names are listed in descending order of SCG scores.

The coding capacity of the viral DNA genome is limited, so it is necessary to use host cell factors to promote the replication of its genome and to produce progeny viruses [69] . A critical initial step in the life cycle of an infectious virus is the identification of and interaction with the host proteins. Therefore, to initiate this cycle, the viral genome needs to be transported from viral particles to the cytoplasm through different viral attachment proteins on their shells [70] . The particles are first attached to the specific proteins on the surface of the host cell. These specific proteins are called receptors. The presence of receptors on the cell surface is a major determinant of viral host selectivity and plays a key regulatory role in host range and viral pathogenesis [71] . Until now, antiviral drug strategies targeting cell surface receptors have made progress in the treatment of different viral infectious diseases [72] . One of the main areas of focus for research is anti-HIV drugs. HIV entry into target cells is a complex process, consisting of multiple biological responses, which is triggered by the interaction between the HIV envelope protein (Env) glycoprotein gp120 and the host cell receptor CD4. This interaction causes a conformational change in gp120, exposing a co-receptor binding site [73] . Subsequently, after binding to the CCR5 or CXCR4 core receptor of the primary HIV isolate, Env undergoes further conformational changes, gp41-mediated membrane fusion, and virion core entry into the cytoplasm [74] . Therefore, the polymorphism of the CCR5 gene can affect the spread of HIV or the development of the disease. Targeting the process of HIV invasion into cells is considered to be an effective method to inhibit HIV replication [11] . The anti-HIV efficacy of small molecule CCR5 antagonists, such as Cenicriviroc [75] and Vicriviroc (SCH-D) [76] , has proven this theory.

Given the above findings, a subcellular localization feature needs to be considered in drug development and protein function development. By analyzing the mechanism of virus invasion, it can be speculated that cell membrane receptors are more likely to become potential targets for antiviral drugs. To test this conjecture, we also began with the 36 known antiviral human targets. The Uniprot database was used to retrieve their subcellular localization information [77] . To improve the accuracy and credibility of the results, the manually annotated and reviewed Swiss-Prot database was chosen for the analysis [77] . A total of 20,233 pieces of human genetic information was collected from the Swiss-Prot database, of which 14,574 were labeled with subcellular localization information. Subcellular location information was available on the database for 32 out of the 36 known antiviral human targets, and 21 were located on the cell membrane (65.6%). It was observed that the antiviral targets were indeed significantly enriched on the cell membrane surface (p < 0.001, hypergeometric test). This characteristic can be used as a key indicator for subsequent antiviral host target screening. It is worth noting that among the 20 scored cell membrane receptors, 11 received a median SCG score of three, accounting for 84.6% (11/13) of the total number of targets with a score of three or more. Therefore, the cell membrane receptors demonstrated a high genetic association with viral diseases.

Viruses initially interact with the host through their capsid or envelope proteins. There is an argument that the primitive capsid protein-encoding genes are recruited from cellular hosts, which implies that the coevolution between virus and host is deeply rooted in the origins of viruses [78] . Furthermore, some studies have shown that the evolution of the virus and parasite host range play an important role in the emergence of infectious diseases [78, 79] . Coevolution between viruses and hosts has also led to larger genetic differences between replication populations, which are often related to the range of hosts that co-evolutionary viruses can infect [80, 81] . For example, health disasters caused by influenza viruses have been common in human history. One of the causes of influenza virus epidemics is its ability to continuously perform antigen transformation through genetic diversity. At the same time, the host's immune defense mechanism is constantly evolving. This process might be compared with the evolutionary race between predators and prey [78, 82] . By studying the mechanism of host resistance to infection, SNPs of the interferon-inducible transmembrane proteins (IFITMs), especially ifitm3 gene, were found to be able to increase or decrease susceptibility to influenza infection [83, 84] , showing the antiviral significance of the host protein's evolution. Synergistic benefits can also be produced by integrating the virus-host evolutionary mechanism and the interaction between pathogens and hosts to improve therapeutic effects [42] . So far, two effective treatment strategies, evolutionary trade-offs and/or competitive effects, have been proposed, based on the evolutionary model in which viruses in the host control or eliminate pathogens [42, 85, 86] . Furthermore, Phillips et al. found that the ability of host chaperone proteins to influence virus evolution is a key factor in determining viral protein mutation fitness, which provides new insights for the design of antiviral treatments [87] .

Tomislav et al. showed that the pathogenicity of disease genes is closely related to their evolutionary origin [88] . The results of research by Wang et al. indicated that successful targets share some common evolutionary characteristics, and suggested that evolutionary information can help to identify the drug targets with the greatest potential for therapeutic development [43, 45] . An in-depth understanding of the antiviral activity and action mechanism of host target genes will help to develop broad-spectrum antiviral drugs that enhance their activity or have similar effects [89] .

There are more than 30 popular algorithms through which to infer the origins of gene families [90] [91] [92] . It is a challenge to obtain accurate gene ages because the age of the gene families depends on the accuracy of the genetic algorithm. To solve this problem, Liebeskind et. al. provided a relatively consistent genetic age data set and mapped the species with orthologous genomes to a reference tree from SwissTree to infer the age of each gene [93] . Based on the rules, human genes are divided into the following eight classes by origin: (1) cellular organism, (2) the common ancestors of eukaryotes and archaea (Euk_Archaea), (3) horizontal gene transfer of bacteria (Euk + Bac), (4) Eukaryota, (5) Opisthokonta, (6) Eumetazoa, (7) vertebrata, (8) mammals. It is believed that this set of data includes minor errors.

Based on the above knowledge, the evolutionary origin of 36 human antiviral targets were analyzed. The UniprotKB data corresponding to 36 targets were obtained by querying the UniprotKB database. The evolution information statistics results indicated that the 36 UniProtKB were mainly distributed in genes originating in class 6 (Eumetazoa) (14/36, p = 0.04, hypergeometric test) (shown in Figure 1a ). As for the 21 cell membrane receptors, 13 of them originated in the Eumetazoan period, and the trend of enrichment was more obvious (p = 0.00034, hypergeometric test) (shown as in Figure 1b) . The results of the gene functional annotation using DAVID showed that these receptors are significantly related to immune response (p =1.4 × 10 −8 ). The results of the evolutionary information statistics showed that the 36 approved targets were distributed mainly in genes belonging to class 6 (Eumetazoa) (14/36, p = 0.04, hypergeometric assay) (a). As for the 21 cell membrane receptors, 13 of them were from the Eumetazoan period, and the tendency for enrichment was more pronounced (p = 0.00034, hypergeometric test) (b).

A previous evolutionary study of 658 successful targets by Quan et al. indicated that these genes are significantly enriched in cellular organism, euk + bac, and Eumetazoa [45] . Our results were partially consistent with their study, suggesting that human genes originating in the Eumetazoan period have greater potential to become antiviral targets.

It is often inadequate to screen targets based solely on a single feature summarized by omics data because they usually depend on large-scale experiments and cannot comprehensively explain the causal relationship between human diseases [45, 94] . Besides, biotechnology continues to identify an increasing number of disease-related genes. Not all of them are potential drug targets. Identifying new and effective targets remains a major priority for modern drug discovery [95] . Based on the analysis of the genetic features, subcellular localization, and evolutionary features of approved antiviral host targets, it is summarized that an effective antiviral target generally has the following three characteristics: (1) a median of genetic score(s) associated with viral disease(s) equal to or greater than three; (2) subcellular localization on the cell membrane; (3) origins in the Eumetazoan period. By using these criteria, we screened out some potential antiviral host targets ( Figure 2 ). Based on the analysis of the genetic features, subcellular localization, and evolutionary features of the approved antiviral host targets, it is summarized that an effective antiviral target generally exhibits the following three characteristics: (1) a median of genetic scores associated with viral disease (s) equal to or greater than three; (2) subcellular localization on the cell membrane; (3) origins in the Eumetazoan period.

As mentioned above, the SCG-Drug database integrates information from eight common genetic disease databases and provides different druggability scores for these genes. It provides us with a huge data set for screening potential antiviral drug targets. We started with 914,190 gene-disease associations in the SCG-Drug database and obtained 19,232 genes by removing repetitions. These genes were used for screening according to the three features described above. First, through genetic screening, 7214 genes were found to be associated with 34 viral diseases. For each gene, the median score associated with viral disease(s) was taken as its genetic score. Next, these genes were divided into two categories based on whether there was relevant information on the three well-known drug databases (Therapeutic Target Database (TTD) [96] , DGIdb [97] , and DrugBank [68] ). Of the 2277 genes with relevant drug information on the databases, we found that 687 are located on the cell membrane and 248 originated in the Eumetazoan period. For the 4937 genes with no drug information, 744 were found to be located on the cell membrane, and 239 originated in the Eumetazoan period. To find candidate genes that are strongly associated with viral disease(s), genes with genetic scores equal to or greater than three were selected. Furthermore, through searching the studies available on NCBI PubMed database, we selected genes with virus-related reports as candidate targets. Finally, 35 candidate targets that met all of the above requirements were obtained (Table 3) . For the 13 candidate targets with drug information, their corresponding drugs may be able to treat viral diseases through drug repositioning. As for the remaining 22 candidate targets without drug information, novel antiviral drugs may be derived from them through structure-based drug design. 

Although we can use various technologies to obtain a large number of potential therapeutic targets, it remains a great challenge to bring these targets into clinical trials. The reported statistics results showed that about 10% of genes in the entire human genome are involved in the development of disease, that is, about 3000 potential targets are suitable for drug treatment [98] . Therefore, predicting whether a protein features a pocket that can bind drug-like molecules with high affinity (druggability [99] ) is of great significance in the target screening stage of drug discovery [100] . In order to improve the efficiency of target discovery, many computational methods have been developed to predict target druggability [101] [102] [103] [104] . PockDrug [101] is a reliable model for predicting the druggability of pockets, which is based on a linear discriminant analysis of 52 molecular descriptors. It selects the most stable and effective model according to different prediction objects. In addition, PockDrug maintained the best combination of molecular descriptors that affect the pharmacological properties of the pocket. With the best combination, PockDrug can achieve an average accuracy of about 87% [101] .

For those targets about which there is no drug information, we need to perform druggability analysis based on protein structures to determine whether it has the potential to be a small molecule drug target. The PDB structure of candidate targets were queried in Swiss-Prot. However, not all of these candidate genes feature known protein structures. To improve the accuracy of the results, we only performed the druggability prediction on the potential targets with reported crystal structures. Therefore, 12 candidate genes with known protein crystal structures were selected for druggability analysis. All the predicted information by PockDrug is summarized in Table 4 and Table S3 . Pockets with druggability scores of 0.5 or more were considered druggable. We used the "predict druggability with proteins" function. Next, we uploaded the corresponding PDB file, selected the Fpocket estimation method (not guided by ligand information), and set the Ligand proximity threshold to the default, 5.5 Å. The complete and detailed prediction results are shown in Table S3 .

Drug repositioning has always been an important way to reduce drug development costs and speed up the research process [106] . At present, the number of approved antiviral drugs targeting host genes is significantly higher than that of drugs directly targeting viral genes [107] . Therefore, it is feasible to adopt a host-centered strategy and to integrate genetic and evolutionary knowledge to increase the success rate of drug repositioning. Among the 35 candidate targets listed in Table 3 , 13 targets can be found with relevant approved drug information in the DrugBank database. It is believed that the existing approved drugs targeting these genes may be used in antiviral therapies through drug repositioning.

By searching the ClinicalTrials databases, the efficacy of some repositioned drugs was supported by the clinical evidence. For instance, IL1R1 (Interleukin-1 receptor type 1) encodes a cytokine receptor belonging to the interleukin-1 receptor family, which participates in the regulation of innate immune and inflammatory processes [108] . Anakinra is a human interleukin-1 receptor antagonist approved for the treatment of adult rheumatoid arthritis (RA) and neonatal multiple inflammatory diseases (NOMID). It was predicted by disease genetics that drugs targeting IL1R1 would be used for Paramyxoviridae infections, such as respiratory syncytial virus infections, influenza, and so on ( Table 3, Table S2 ). According to the records of the ClinicalTrials database, Anakinra is currently being studied for the treatment of COVID-19. Related ClinicalTrials.gov identifiers are NCT04362943, NCT04364009, NCT04330638, NCT02735707, NCT04362111, NCT04339712, NCT04366232, NCT04341584, etc. Some of these studies are already in phase 3 or 4 (NCT02735707, NCT04330638, NCT04364009, NCT04362111).

The gene AGTR1 (Type-1 angiotensin II receptor) is one of the important effect factors in mediating the volume and blood pressure of the cardiovascular system [109] . The genetic data suggest that drugs targeting AGTR1 can be used in the treatment of viral diseases, such as Paramyxoviridae infections, AIDS dementia complex, etc. (Table 3, Table S2 ). Losartan is an angiotensin II receptor blocker used to treat high blood pressure and diabetic nephropathy and to reduce the risk of stroke. Losartan is indeed being investigated for the treatment of respiratory syndrome / coronavirus infection (NCT04312009, NCT04311177, NCT04340557) and HIV (NCT01852942, NCT01529749), which partially coincides with our predictions.

Likewise, Formoterol, targeting ADRB2 (Beta-2 adrenergic receptor), etc., is an inhaled beta-2 agonist that was approved by the FDA in 2001 for the treatment of chronic obstructive pulmonary disease (COPD) and asthma [110] . It was inferred through genetic data that it can be used in the treatment of paramyxoviridae infections, respiratory syncytial virus infections, etc. (Table 3, Table S2 ). Indeed, there is a phase 3 study that uses it for COVID-19 treatment (NCT04331054).

It can be seen from the results of the druggability prediction (Table 4, Table S3 ) that candidate targets including FAS, CD209, LGR5, CRHR2, AGER, BIN1, BSG, IL1RL1, PARD6A, RGS7, RHOU, and TNFSF10 have one or more druggable pockets. These twelve potential antiviral host target genes are divided into two types, according to whether they are known targets or not. One is the recognized target type, including CD209, FAS, LGR5, and CRHR2 genes (however, without drug information); the other is the non-recognized type, including AGER, BIN1, BSG, IL1RL1, PARD6A, RGS7, RHOU, and TNFSF10 genes. The twelve potential antiviral host targets have no approved drug information, therefore they may be used for de novo drug discovery.

The CD209 gene encodes the pathogen-recognition receptor, CD209 antigen, which is expressed on the surface of immature dendritic cells and is involved in the initiation of the primary immune response [111] . Many reports have shown that the CD209 antigen could act as an attachment receptor for Ebolavirus [112] , HIV-1 and HIV-2 [112] , Dengue virus [113] , and others. The encoding protein's crystal structure has been released (PDB id: 2XR6), and the predicted best druggable pocket is shown in Figure 3a . Based on the identified cavity, more novel drug screening work is worth carrying out.

The LGR5 gene encodes the Leucine-rich repeat-containing G-protein coupled receptor 5 for R-spondins that potentiates the canonical Wnt signaling pathway [114] . The LGR5 gene is a proven target and has been used as a marker of adult tissue stem cells in the intestine, stomach, hair follicle, and mammary epithelium [115] . The receptor protein's crystal structure was released and deposited in the Protein Data Bank (PDB id: 4ufr), and the predicted best druggable pocket was the cyan surface, as shown in Figure 3b . The pocket could be used as an important cavity for drug screening.

The CRHR2 gene encodes the corticotropin-releasing factor receptor 2. The protein's crystal structure was released and deposited in the Protein Data Bank (PDB id: 3n93). Its molecular functions include corticotropin-releasing factor receptor activity, corticotropinreleasing hormone receptor activity, and peptide hormone binding [116] . The predicted best druggable pocket was the cyan surface, as shown in Figure 3c .

The FAS gene belongs to the tumor necrosis factor receptor superfamily, and encodes the receptor for TNFSF6/FASLG [117] . The molecular functions of FAS include calmodulin binding, identical protein binding, kinase binding, signaling receptor activity, and transmembrane signaling receptor activity [118, 119] . The FAS gene is a proven target and is involved in autoimmune lymphoproliferative syndrome 1A disease. The encoding protein's crystal structure has been released (PDB id: 3tje), and the predicted best druggable pocket is shown in Figure 3d . Follow-up drug screening can be carried out based on this identified cavity.

The AGER gene encodes the advanced glycosylation end-product-specific receptor that mediates interactions between advanced glycosylation end products (AGE) and acts as a mediator of both acute and chronic vascular inflammation [120] . AGER is a member of the immunoglobulin superfamily of cell surface molecules, and AGE/AGER interaction has been linked to the regulation of the production/expression of TNF-alpha, oxidative stress, cancer, and endothelial dysfunction in type 2 diabetes [121] . The encoding protein's crystal structure has been released (PDB id: 3o3u), and the predicted best druggable pocket is shown in Figure 3e . Based on the identified cavity, more structure-based virtual screening could be performed in the future.

The BIN1 gene encodes the Myc box-dependent-interacting protein 1 that is a membrane deforming protein. It was found to be a tumor repressor due to its interaction with MYC oncoproteins [122] . The BIN1 gene is also involved in DNA repair, cell cycle progression, cytoskeleton regulation, apoptosis, and regulating cellular BACE1 levels [123] . The protein of the BIN1 gene's crystal structure was released and deposited in the Protein Data Bank (PDB id 2fic). Based on the crystal structure, the best druggable pocket was predicted and appeared in the cyan surface (Figure 3f) .

The IL1RL1 gene belongs to the interleukin-1 receptor family of cytokines and encodes the interleukin-1 receptor-like 1 protein, which is the receptor for interleukin-33 (IL-33) and whose signaling requires the association of the coreceptor IL1RAP [124] . The protein's alternative name is the ST2 protein. The signaling of IL-33 is generated by its ligandbinding primary receptor, ST2, and the subsequent recruitment of accessory receptor IL-1RAP, which results in the juxtaposition of intracellular toll/interleukin-1 receptor domains of both receptors, which are necessary and sufficient for the activation of the NF-κB and MAPK pathways in the target cells [125] . The ST2 protein's crystal structure has been released (PDB id: 4kc3), and the predicted best druggable pocket is shown in Figure 3g . This receptor was also described as a negative regulator of Toll-like receptor-IL-1 receptor signaling, and it is worthy of further study as a potential antiviral target. Table 4 and Table S3 . Pockets with druggability scores of 0.5 or more were considered druggable.

The PARD6A gene encodes the partitioning defective 6 homolog alpha protein. Its short name is PAR-6, PAR6-alpha, or PAR-6A; the protein's alternative name is PAR6C, or Tax interaction protein 40 (TIP-40). PAR-6 protein is involved in asymmetrical cell division, cell polarization, and cell transformation processes [126] . The crystal structure of PAR-6 protein has been released and deposited in the database (PDB id: 1wmh), and the predicted best druggable pocket is shown in Figure 3h .

The RGS7 gene encodes the regulator of G-protein signaling 7, which belongs to the regulator of the G-protein signaling superfamily. The combination of RGS7 protein and Gbeta5 constitutes a crucial regulator of G protein-coupled receptor signaling in the visual and nervous systems [127] . RGS7 also plays an important role in synaptic vesicle exocytosis through its interaction with snapin [128] . The RGS7 protein's crystal structure (PDB id:2a72) and predicted best druggable pocket is shown in Figure 3i . RGS7 protein also modulates the activity of potassium channels [127] . As a potential antiviral target, it offers significant research potential.

The RHOU gene encodes the Rho-related GTP-binding protein, RhoU. The protein's alternative name is CDC42-like GTPase 1, GTP-binding protein-like 1, Wnt-1 responsive Cdc42 homolog 1 (Wrch-1), or Ryu GTPase. Active Wrch-1 can stimulate quiescent cells to reenter the cell cycle. Wrch-1 could regulate the actin cytoskeleton, cell morphology, cell proliferation, and migration [129] . The Wrch-1 protein's crystal structure has been determined and deposited in the database (PDB id: 2q3h). The predicted best druggable pocket is shown in Figure 3j . Antiviral drug screening could be performed based on this potential antiviral host target.

The TNFSF10 gene encodes the protein that belongs to tumor necrosis factor ligand superfamily member 10. The protein's crystal structure was released and deposited in the Protein Data Bank (PDB id: 1dg6). The molecular functions include cytokine activity, identical protein binding, signaling receptor binding, tumor necrosis factor receptor binding, and others [130, 131] . The predicted best druggable pocket is the cyan surface shown in Figure 3k . The pocket could be used as a valid cavity for drug screening.

In recent years, new antiviral drugs have continued to appear, but the incidence of viral infections has continued to increase. Genetic and evolutionary mechanisms underlie the outbreak and epidemic of viruses. In particular, the continuous outbreak of highmortality viruses, such as COVID-19, HIV, Ebola virus, DEN, and influenza virus, has created an urgent need for research into and development of antiviral drugs. In general, the challenge for antiviral drug discovery arises from the mutation and drug-resistance of viruses. For instance, most antibodies can be evaded by single mutations of the influenza virus, which leads to simultaneous outbreaks of multiple influenza virus subtypes [132] . Traditional antiviral drugs are prone to developing drug resistance after long-term use due to this characteristic of the virus. Host-targeting antivirals not only provide a higher genetic barrier to drug resistance, but also exhibit broad-spectrum antiviral activities and are likely to be effective against newly emerging viruses [11] . Currently, there are multiple avenues under investigation as therapeutic strategies for host-targeted antivirals, such as certain innate or adaptive immune host activities that respond to viral infections (i.e., TLR agonists [133] , CCR5 antagonists [75, 134, 135] , etc.) and the regulation of inflammatory pathways (i.e., TNF-α-mediated antiviral activity [136] etc.). Meanwhile, the appropriate combinations of host-targeted antivirals may improve efficacy, expand antiviral activity, and reduce the likelihood of drug resistance. For instance, McHutchison et al. reported that in HCV-infected patients who failed to respond to treatment with peginterferon alfa and Ribavirin previously, re-treatment with Telaprevir combined with peginterferon alfa and Ribavirin was proven to be more effective than re-treatment with either drug alone [137, 138] .

Although host-targeting antiviral drug discovery strategies have become a major focus of research, very few human host proteins were explored as antiviral targets. In addition, there are challenges in developing host-directed therapies. For example, emerging evidence suggests that viral resistance against host antivirals does occur [136] . Meanwhile, hosttargeted therapies may produce some expected or unexpected side effects, since they may target host cell functions that are critical for cell survival [139] . Therefore, a more in-depth analysis of possible cytotoxicity or the potential to worsen the infection is needed when treating with host-targeted antivirals.

Genetic and evolutionary knowledge provides new insights into the pathogenesis of diseases by identifying specific genes or pathways associated with diseases, and thus provides opportunities to discover new drug targets. By summarizing the properties of 36 approved antiviral targets in genetics, cell biology and evolutionary biology, we found that the targets feature certain characteristics: (1) median genetic score(s) associated with viral disease(s) equal to or greater than three; (2) subcellular localization on the cell membrane;

(3) origins in the Eumetazoan period. Based on these three principles, we used the gene set from the SCG-Drug database to screen new antiviral targets and obtained 35 host targets. By analyzing these candidate targets, we predicted some potential repositioned drugs and identified some new targets that may be used for de novo drug discovery.

Through the continuous development of biotechnology, significant advances have been made in high-throughput sequencing, omics analysis techniques, and molecular biology. Consequently, an increasing number of genetic, subcellular location, and evolutionary information about various genes will be accumulated. Especially in the context of the COVID-19 pandemic, the rapid screening of drug targets based on the above information will facilitate the discovery of druggable antiviral host targets, so as to promote the development of antiviral drugs. 

The authors have no other relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript, apart from those disclosed.

Human Viruses: Discovery and Emergence

A Familial Cluster of Pneumonia Associated with the 2019 Novel Coronavirus Indicating Person-to-Person Transmission: A Study of a Family Cluster

Clinical Features of Patients Infected with 2019 Novel Coronavirus in

Early Transmission Dynamics in Wuhan, China, of Novel Coronavirus-Infected Pneumonia

Ebola Virus: A Global Public Health Menace: A Narrative Review

Severe Acute Respiratory Syndrome: Responses of the Healthcare System to a Global Epidemic

Epstein-Barr Virus and Cancer

Secular Trends in HIV/AIDS Mortality in China from 1990 to 2016: Gender Disparities

Plague Vaccine: Recent Progress and Prospects. npj Vaccines

Infection: Epidemiology, Pathogenesis, Treatment, and Prevention

Medicinal Chemistry Strategies toward Host Targeting Antiviral Agents

Anti-Viral Therapy for Congenital Cytomegalovirus Infection: Pharmacokinetics, Efficacy and Side Effects

Sofosbuvir and Velpatasvir for the Treatment of Hepatitis C

Anti-Hepatitis C Virus Drugs in Development

Selective Anti-Herpesvirus Agents

Antiviral Drugs Against Alphaherpesvirus

Anti-Influenza Treatment: Drugs Currently Used and Under Development

Approved Antiviral Drugs over the Past 50 Years

Editorial Overview: Antivirals and Resistance: Advances and Challenges Ahead

The Signaling Pathways, and Therapeutic Targets of Antiviral Agents: Focusing on the Antiviral Approaches and Clinical Perspectives of Anthocyanins in the Management of Viral Diseases

Preclinical Drug Development

Diagnosing the Decline in Pharmaceutical R&D Efficiency

Tufts Study Pegs Drug Development, Approval Cost at $2.6B. GEN-Genetic Engineering Biotechnology News

Clinical Development Success Rates for Investigational Drugs

Lessons from 60 Years of Pharmaceutical Innovation

Systems Biology-Embedded Target Validation: Improving Efficacy in Drug Discovery

Systems Chemical Genetics-Based Drug Discovery: Prioritizing Agents Targeting Multiple/Reliable Disease-Associated Genes as Drug Candidates

Enteroviruses: Classification, Diseases They Cause, and Approaches to Development of Antiviral Drugs

Systematic Approaches towards the Development of Host-Directed Antiviral Therapeutics

Systems-Biology Approaches to Discover Anti-Viral Effectors of the Human Innate Immune Response

Human Genetic Susceptibility to Infectious Disease

Drivers of Within-Host Genetic Diversity in Acute Infections of Viruses

Genetic Mechanisms of Critical Illness in COVID-19

Virus-Host Interactions: New Insights and Advances in Drug Development Against Viral Pathogens

Host Targeted Antiviral (HTA): Functional Inhibitor Compounds of Scaffold Protein RACK1 Inhibit Herpes Simplex Virus Proliferation

Iminosugars: Promising Therapeutics for Influenza Infection

Dengue Virus Evolution under a Host-Targeted Antiviral

Host Cell Restriction Factors That Limit Influenza A Infection

Opportunities for Host-Targeted Therapies for Malaria

Host-Directed Drug Therapies for Neglected Tropical Diseases Caused by Protozoan Parasites

Fighting Microbial Pathogens by Integrating Host Ecosystem Interactions and Evolution

Can Medical Genetics and Evolutionary Biology Inspire Drug Target Identification?

Human Genetics as a Foundation for Innovative Drug Development

Evolutionary and Genetic Features of Drug Targets

Antiviral Innate Immunity through the Lens of Systems Biology

Human Genetic Determinants of Viral Diseases

Genetic Predisposition to Infectious Disease

Human Genetic Predisposition to Diseases Caused by Viruses from Flaviviridae Family

Host Genetic Factors Affecting Hepatitis B Infection Outcomes: Insights from Genome-Wide Association Studies

Next Generation Sequencing-Based Emerging Trends in Molecular Biology of Gastric Cancer

Investigational Drugs for Nasopharyngeal Carcinoma

Human Gene Polymorphisms and Their Possible Impact on the Clinical Outcome of SARS-CoV-2 Infection

Modeling Viral Kinetics and Treatment Outcome during Alisporivir Interferon-Free Treatment in Hepatitis C Virus Genotype 2 and 3 Patients

Whole Exome Sequencing of HIV-1 Long-Term Non-Progressors Identifies Rare Variants in Genes Encoding Innate Immune Sensors and Signaling Molecules

Identification of Required Host Factors for SARS-CoV-2 Infection in Human Cells

Using Genetics To Dissect SARS-CoV-2 Infection

Host Factors in Coronavirus Replication

Mendelian Randomization Analysis Identified Genes Pleiotropically Associated with the Risk and Prognosis of COVID-19

Public Archive of Relationships among Sequence Variation and Human Phenotype

Online Mendelian Inheritance in Man (OMIM®), an Online Catalog of Human Genes and Genetic Disorders

The Human Gene Mutation Database: Towards a Comprehensive Repository of Inherited Mutation Data for Medical Research, Genetic Diagnosis and next-Generation Sequencing Studies

Clinical Practice Guidelines for Rare Diseases: The Orphanet Database

GWASdb v2: An Update Database for Human Genetic Variants Identified by Genome-Wide Association Studies

Explaining the Disease Phenotype of Intergenic SNP through Predicted Long Range Regulation

The Genetic Association Database

DisGeNET: A Discovery Platform for the Dynamical Exploration of Human Diseases and Their Genes

Virus DNA Replication and the Host DNA Damage Response

Dynamics of Virus-Receptor Interactions in Virus Binding, Signaling, and Endocytosis. Viruses

Virus-Receptor Interactions: The Key to Cellular Invasion

Subcellular Targeting Strategies for Drug Design and Delivery

HIV-1 Receptors and Cell Tropism

Role of HIV-1 Subtype C Envelope V3 to V5 Regions in Viral Entry, Coreceptor Utilization and Replication Efficiency in Primary T-Lymphocytes and Monocyte-Derived Macrophages

CCR5 Receptor Antagonists in Preclinical to Phase II Clinical Development for Treatment of HIV

a New CC-Chemokine Receptor 5 Inhibitor for Treatment of HIV: Properties, Promises and Challenges

UniProt: A Worldwide Hub of Protein Knowledge

Parasite Host Range and the Evolution of Host Resistance

Estimating Parasite Host Range

Antagonistic Coevolution Accelerates Molecular Evolution

Viruses and Mobile Elements as Drivers of Evolutionary Transitions

Influenza Virus-Host Co-Evolution. A Predator-Prey Relationship?

Interferon-Induced Transmembrane Protein 1 Restricts Replication of Viruses That Enter Cells via the Plasma Membrane

Host Transcriptional Response to Influenza and Other Acute Respiratory Viral Infections-A Prospective Cohort Study

Promises and Pitfalls of In Vivo Evolution to Improve Phage Therapy

Steering Phages to Combat Bacterial Pathogens

Shoulders, M.D. Host Proteostasis Modulates Influenza Evolution

Phylostratigraphic Tracking of Cancer Genes Suggests a Link to the Emergence of Multicellularity in Metazoa

The Broad-Spectrum Antiviral Functions of IFIT and IFITM Proteins

Towards Consensus Gene Ages

No Evidence for Phylostratigraphic Bias Impacting Inferences on Patterns of Gene Emergence and Evolution

Quest for Orthologs consortium Big Data and Other Challenges in the Quest for Orthologs

Quest for Orthologs Entails Quest for Tree of Life: In Search of the Gene Stream

Validating Therapeutic Targets through Human Genetics

Chemical Genetics-Based Target Identification in Drug Discovery

Therapeutic Target Database

DGIdb 3.0: A Redesign and Expansion of the Drug-Gene Interaction Database

How Many Genomics Targets Can a Portfolio Afford?

Genomics: Success or Failure to Deliver Drug Targets?

PockDrug: A Model for Predicting Pocket Druggability That Overcomes Pocket Estimation Uncertainties

Druggability Indices for Protein Targets Derived from NMR-Based Screening Data

Elucidating the Druggability of the Human Proteome with EFindSite

Deep Learning to Therapeutically Target Unreported Complexes

RCSB Protein Data Bank: Biological Macromolecular Structures Enabling Research and Education in Fundamental Biology

Rational Drug Repositioning by Medical Genetics

A Computational Framework of Host-Based Drug Repositioning for Broad-Spectrum Antivirals against RNA Viruses

The Family of the Interleukin-1 Receptors

A Male-Specific Association between AGTR1 Hypermethylation and Coronary Heart Disease

Glycopyrronium/Formoterol: A Review in COPD

The Dendritic Cell-Specific Adhesion Receptor DC-SIGN Internalizes Antigen for Presentation to T Cells

Differential N-Linked Glycosylation of Human Immunodeficiency Virus and Ebola Virus Envelope Glycoproteins Modulates Interactions with DC-SIGN and DC-SIGNR

DC-SIGN (CD209) Mediates Dengue Virus Infection of Human Dendritic Cells

Niehrs, C. LGR4 and LGR5 Are R-Spondin Receptors Mediating Wnt/β-Catenin and Wnt/PCP Signalling

Immunostaining of Lgr5, an Intestinal Stem Cell Marker, in Normal and Premalignant Human Gastrointestinal Tissue

Phylogenetic-Based Propagation of Functional Annotations within the Gene Ontology Consortium

Three Functional Soluble Forms of the Human Apoptosis-Inducing Fas Molecule Are Produced by Alternative Splicing

Structural Insights into the Mechanism of Calmodulin Binding to Death Receptors

Purification and Molecular Cloning of the APO-1 Cell Surface Antigen, a Member of the Tumor Necrosis Factor/Nerve Growth Factor Receptor Superfamily. Sequence Identity with the Fas Antigen

RAGE-Dependent Signaling in Microglia Contributes to Neuroinflammation, Aβ Accumulation, and Impaired Learning/Memory in a Mouse Model of Alzheimer's Disease

Advanced Glycation End Product Recognition by the Receptor for AGEs

Mutations in BIN1 Associated with Centronuclear Myopathy Disrupt Membrane Remodeling by Affecting Protein Density and Oligomerization

BIN1 Regulates BACE1 Intracellular Trafficking and Amyloid-β Production

Structure of IL-33 and Its Interaction with the ST2 and IL-1RAcP Receptors-Insight into Heterotrimeric IL-1 Signaling Complexes

Structural Insights into the Interaction of IL-33 with Its Receptors

A Human Homolog of the C. Elegans Polarity Determinant Par-6 Links Rac and Cdc42 to PKCzeta Signaling and Cell Transformation

Palmitoylation Regulates Plasma Membrane-Nuclear Shuttling of R7BP, a Novel Membrane Anchor for the RGS7 Family

Snapin Interacts with the N-Terminus of Regulator of G Protein Signaling 7

Wrch-1, a Novel Member of the Rho Gene Family That Is Regulated by Wnt-1

Induction of Apoptosis by Apo-2 Ligand, a New Member of the Tumor Necrosis Factor Cytokine Family

The Structure of the Death Receptor 4-TNF-Related Apoptosis-Inducing Ligand (DR4-TRAIL) Complex

How Single Mutations Affect Viral Escape from Broad and Narrow Antibodies to H1 Influenza Hemagglutinin

Novel Drugs Targeting Toll-like Receptors for Antiviral Therapy

How Mechanisms of HIV Entry Govern AIDS Pathogenesis

CCR5/CXCR4 Dual Antagonism for the Improvement of HIV Infection Therapy

Host-Directed Antiviral Therapy

Telaprevir for Previously Treated Chronic HCV Infection

Current and Future Therapies for Hepatitis C Virus Infection: From Viral Proteins to Host Targets

Host-Directed Therapies for Bacterial and Viral Infections