2.3 Protein-protein interaction (PPI) network uncovering the core anti-HCoVs targets regulated by XCHD
We next constructed a PPI network for the 37 HCoV-association host genes of XHCD, for exploring the functional relationships among interacting proteins and prioritizing the potential anti-HCoV targets. As seen in Fig. 4A, a larger circular diameter and brighter color of the node indicate a bigger number of protein association. There are a total of 95 protein interaction associations, and the average correlation number of each protein is 5.14. Among these proteins, IL-6 and STAT3 are the core proteins with degrees of 17 and 15 respectively (Fig. 4B). IL-6 is a classical proinflammatory cytokine, which is involved in a variety of inflammation related disease states, and plays a major role in B cell maturation [47]. In the lesion site of acute inflammation, IL-6/STAT3 pro-inflammatory signal transduction axis is a part of acute phase response and a non-specific response of innate immune system to pathogen infection [48]. In the secondary lymphoid tissues where an adaptive immune response occurs, IL-6-mediated STAT3 activation promotes the proliferation and survival of T cells and B cell populations [49]. IL-6 also promotes the differentiation of follicular helper T cells through STAT3, effectively linking T and B cell responses [50]. It can be concluded that XCHD may regulate the innate and adaptive immune responses by promoting inflammatory signals, so as to fight against HCoVs.
2.4 Gene enrichment analysis of HCoVs-associated host targets of XCHD.
We gave an insight into the molecular mechanism of XCHD against HCoVs by studying GO annotation and KEGG pathway enrichment of the HCoVs-associated host targets. As the GO annotation result shown in Fig. 5A, the HCoVs host targets acting on XCHD exist in the cell membrane, cytoplasm, mitochondria, Golgi apparatus and cell nucleus, etc. These targets have the binding ability to viruses, sugars, enzymes, transforming growth factor β receptor, and play important roles in enzyme inhibition, protein transport, transcription activation and other activities. They also participate in the viral cell entry, transport, gene replication, immune response function regulation, acute inflammation and other biological processes.
Besides, as presented in Fig. 5B, KEGG pathway enrichment analysis uncovers the biological pathways (p < 0.05) of XCHD against HCoV, such as Toll-like receptor (TLR) signaling pathway (hsa04620, p = 5.56E-04), retinoic acid-inducible gene (RIG)-I-like receptor signaling pathway (hsa04622, p = 0.031), cytosolic DNA-sensing pathway (hsa04623, p = 0.001), focal adhesion (hsa04510, p = 0.042), and apoptosis (hsa04210, p = 0.045). We further constructed a target-pathway network to illustrated the relationships among these HCoVs-associated targets and corresponding enriched pathways (Fig. 5C). The network is composed of 19 target-pathway interactions connecting 9 targets and 5 potential signaling pathways for target intervention by XCHD. TLRs are the earliest identified and most distinctive pattern recognition receptors (PRR), they can recognize molecular patterns related to pathogens such as viruses or bacteria, and identify molecular patterns related to damaged cells, playing an important role in non-specific immune response [51–52]. RIG-1-like receptors also belong to the PRR, and after recognition of the virus, they can pass the mitochondrial antiviral signal MAVS, resulting in the expression of cytokines including type I and type III interferons, thereby limiting the propagation of the virus [53]. Immunization of cytoplasmic DNA is involved in the early activation of the immune system against infection and the activation of adaptive immune responses [54–55]. Focal adhesion kinase is at the intersection of many signaling pathways that control cell adhesion, migration, proliferation, and survival, with a wide range of physiological roles [56]. Apoptosis (programmed cell death) is considered to be an integral part of the normal development and functional processes of the immune system, and pro-apoptotic factors can exert antiviral effects [57–58]. In summary, the gene enrichment analysis indicated that the regulatory targets of XCHD are closely related to immune regulation, suggesting that XCHD may treat HCoVs by regulating the immune function of host cells.
2.5 Integrated pathway of XCHD against HCoVs
As discussed above, XCHD may exert anti-HCoVs effects through regulating pathways directly related to coronavirus. Here, we selected three representative pathways, including RIG-I-like receptor signaling pathway, cytosolic DNA-sensing pathway, and TLR signaling pathway, to construct an integrated pathway for systematically illustrating the potential mechanisms of XCHD on anti-coronavirus treatment (Fig. 6).
RIG-I-like receptor signaling pathway. Upon viral infection, RIG-I-like receptors can detect the presence of virus-associated molecular patterns [59] and trigger the activation of type I interferons (IFN) and inflammatory mediators that eliminate the viral pathogens and infected cells. Previous studies imply that activation of RIG-I-like receptor signaling pathway contributes to inflammatory cascade in MERS-CoV-infected macrophages [60]. Figure 6 shows that compounds in XCHD could target to several key proteins in the RIG-I-like receptor signaling pathway, such as IRF3, IKBKB and CXCL10.
Cytosolic DNA-sensing pathway. DNA-dependent activator of IFN regulatory factors (DAI), a cytosolic DNA sensor, are one of the recently described PRRs, which activates the IRF3- and/or NF-κB-responsive genes, and induces the expression of type I IFNs and proinflammatory cytokines [61–63]. As exhibited in Fig. 6, compounds in XCHD act on multiple targets on cytosolic DNA-sensing pathway (e.g. IL6, IRF3, IKBKB, CXCL10), suggesting the critical role that XCHD may play on cytosolic DNA-sensing module.
Toll-like receptor signaling pathway. TLR signaling pathway proceeds from two pathways: the TRIF-mediated pathway induced by Toll-interleukin-1 receptor (TIR)-domain-containing adaptor, and the myeloid differentiation factor 88 (MyD88)-mediated pathway. The latter activates the transcription factor NF-κB, which contributes to the inflammatory reactions [64]. The key importance of TLR signaling is demonstrated by the host cell recognition by TRIF and MyD88, controlling the progression of respiratory virus infections [65]. There is evidence that MyD88-independent signaling via the TRIF adaptor protein exerts powerful defense against SARS-CoV infection [66]. The compounds in XCHD were found to interact with the key proteins (e.g., IRF3, IL6 and JUN), illustrating that the prescription could also trigger a broad array of cytokines and chemokines through TLR signaling pathway.
As described above, IRF3, IKBKB, JUN, IL6 and CXCL10 were the HCoVs-associated host targets of XCHD playing important roles in the integrated pathway of XCHD against HCoVs. IRF3 and IKBKB are present in all three pathways. IRF3 forms a complex with CREBBP that translocates to the nucleus and activates the transcription of interferons alpha and beta, as well as other interferon-induced genes. The protein encoded by IKBKB can activate NFKB, then translocates into the nucleus and stimulates the expression of genes involved in pro-inflammatory cytokines and chemokines.
JUN in Toll-like receptor signaling pathway is also known as AP1, encoding a protein interacts directly with specific target DNA sequences to regulate the expression of the inflammatory cytokines like IL-6 and CXCL10. The functioning of IL6 is implicated in a wide variety of inflammation-associated diseases, and it has been found that the elevated levels of IL-6 in virus infections including COVID-19. And CXCL10 may be a key regulator in the immune response of 'cytokine storm' to SARS-CoV-2 infection.
2.6 In silico identification of anti-HCoV activity of ingredients from XCHD
Due to the complexity of ingredients and interacted targets, it is always a huge challenge to dig out the precise ingredients in TCM prescription that exert pharmacological effects towards a certain disease. In this section, we try to identify the active components of XCHD with the assistance of in silico approaches. According to the above systems pharmacology analysis, we identified 163 compounds in XCHD that may act on the 37 HCoVs-associated targets. We further applied a machine learning-based model to assess the drug-likeness of these compounds. The predictive model was built by random forest algorithm and the chemical structure was described by MACCS fingerprint [67]. After that, we screened out compounds that have antiviral report in previous pharmacological literature. By comprehensive considering of the chemical structure, content, availability and accessibility, we finally highlighted 30 potential anti-HCoVs candidates from XCHD.
2.7 In vitro validation of anti-HCoV activity of ingredients from XCHD.
Here, we conducted in vitro assays to further validate our predictions. As listed in Table 4, 16 compounds out of the 30 experimental compounds (hit rate = 53.4%) have inhibitory activity against HCoV-229E virus with selection index (SI) value > 1. Among them, betulinic acid derived from Dazao (SI = 46.77), chrysin derived from Huangqin (SI = 7.49), isoliquiritigenin derived from Chaihu. and Zhigancao (SI = 6.98), schisandrin B (SI = 6.7) and (20R)-Ginsenoside Rh1 (SI > 5.20) derived from Renshen. have an SI value > 5. Published studies show that these 5 compounds with anti-HCoV-229E activity have other antiviral or pharmacological activities. Betulinic acid, which exhibited approximate activity (SI = 46.77) compared to positive control drug ribavirin (SI > 49.44 ± 19.59), exerts inhibitory effect on various viruses such as Zika virus, dengue virus, influenza virus and HIV [68–70]. Chrysin is a dihydroxyflavone, has been found to have activities against influenza virus, herpes-virus and enterovirus 71 [71–73]. Isoliquiritigenin, as a member of the class of chalcones, had been confirmed to reduce morbidity and lung inflammation of mice infected with influenza virus [74]. Schisandrin B was identified as new scaffold of novel HIV-1 RT inhibitors [75]. Ginsenoside Rh1 is reported to eliminate the cytoprotective phenotype of human immunodeficiency virus type 1-transduced human macrophages by inhibiting the phosphorylation of pyruvate dehydrogenase lipoamide kinase isozyme 1 [76].
Table 4
The activity evaluation of constituents from XCHD in HCoV-229E induced CPE reduction assay.
Compound
|
Pubchem ID
|
TC50 a
|
IC50 b
|
SI c
|
Source
|
Structure
|
Isoliquiritigenin
|
638278
|
12.93
|
1.85
|
6.98
|
Chaihu; Zhigancao
|
|
Eleutheroside E
|
71312557
|
> 50
|
50
|
> 1.0
|
Renshen
|
|
Ginsenoside Rb1
|
9898279
|
> 50
|
50
|
> 1.0
|
Renshen
|
|
Kaempferide
|
5281666
|
28.87
|
16.67
|
1.73
|
Chaihu;
Renshen
|
|
Citric acid
|
311
|
> 50
|
50
|
> 1.0
|
Chaihu;
Renshen;
Dazao
|
|
Citropten
|
2775
|
28.87
|
16.67
|
1.73
|
Chaihu;
Shengjiang
|
|
6-Shogaol
|
5281794
|
3.21
|
1.44
|
2.23
|
Banxia;
Shengjiang
|
|
(+)-Catechin Hydrate
|
107957
|
> 50
|
38.8
|
> 1.29
|
Dazao
|
|
Betulinic acid
|
64971
|
16.67
|
0.36
|
46.77
|
Dazao
|
|
(20R)-Ginsenoside Rh1
|
21599923
|
> 50
|
9.62
|
> 5.20
|
Renshen
|
|
Methyl salicylate
|
4133
|
> 50
|
34.67
|
> 1.44
|
Chaihu
|
|
Scopoletin
|
5280460
|
> 50
|
50
|
> 1.0
|
Chaihu;
Dazao
|
|
Chrysin
|
5281607
|
34.67
|
4.63
|
7.49
|
Huangqin
|
|
Schisandrin B
|
108130
|
28.87
|
4.31
|
6.7
|
Renshen
|
|
Ginsenoside Rg3
|
9918693
|
> 50
|
16.67
|
> 3.0
|
Renshen
|
|
L-(-)-Fucose
|
3034656
|
> 50
|
28.87
|
> 1.73
|
Renshen
|
|
Paeonol
|
11092
|
> 50
|
50
|
> 1.0
|
Banxia;
Renshen
|
|
Ribavirind
|
37542
|
> 100
|
2.2 ± 0.87
|
> 49.44 ± 19.59
|
-
|
|
aTC50: 50% cytotoxic concentration (µg/ml) |
bIC50: 50% effective concentration (µg/ml) |
cSI: Selection index, TC50/IC50 |
dPositive control drug. |
2.8 Synergistic effects investigation of the five active compounds in XCHD through compound-target subnetwork
As the in vitro experiments shown, betulinic acid, chrysin, isoliquiritigenin, schisandrin B, and (20R)-Ginsenoside Rh1 are the most promising anti-HCoVs candidates of XCHD. To investigate their synergistic effects against HCoVs, we built a C-T subnetwork for the five compounds through extracting the corresponding CTIs from the global H-C-T network. As showed in Fig. 7, the C-T subnetwork is composed of 175 known CTIs and 133 predicted CTIs, connecting 5 compounds with 174 target proteins. The chrysin, isoliquiritigenin, betulinic acid, schisandrin B, and (20R)-Ginsenoside Rh1 are connected to multiple targets with the degree of 98, 60, 39, 38, and 33, respectively. Five proteins, including CYP3A4, LMNA, MAPT, RAB9A and SMN1 are simultaneously targeted by all the 5 ingredients, suggesting that XCHD may mainly exert its anti-HCoVs effect through the synergistic effect of these components. Remarkably, the 5 ingredients in XCHD act on 4 key HCoVs-associated proteins including IL6, STAT3, IKBKB, and TGFB1, indicating its potential multi-target anti- HCoVs mechanism. Taking STAT3 as an example, previous report has revealed that silencing the isoforms of STAT3 could change the expression of ACE2, which has been identified as an entry receptor for SARS-CoV-2 [77]. Besides, as we mentioned in the above analysis, the IL-6/STAT3 pro-inflammatory signal transduction axis is the important regulatory pathway for XCHD to exert antiviral effects. The subnetwork further confirms this molecular mechanism hypothesis.