Keywords
quercetin - knee osteoarthritis - potential targets - molecular mechanism - network
pharmacology - molecular docking
Introduction
Knee osteoarthritis (KOA) is a chronic osteoarthropathy, which is characterized by
degeneration of knee cartilage, secondary osteophyte formation, subchondral bone remodeling
and synovitis.[1] Its clinical manifestations are mainly pain, swelling, functional limitation, and
joint deformity around the knee, which seriously affect the health and quality of
life of patients.[2] According to the theory of traditional Chinese medicine, KOA is the functional deficiency
of liver and kidney, which involves vessel blocking and joint pain. Therefore, it
is believed that the multidrug and multitarget approaches should be considered in
the treatment of KOA. Epidemiological studies have shown that KOA is more common in
middle-aged and elderly people, with 250 million people worldwide.[3] With the extension of life expectancy and the growth of the elderly population,
it is expected that this number will increase gradually in the future. The occurrence
of KOA may be related to many risk factors, such as age, sex, obesity, etc.,[4] and the specific pathogenesis is still unclear. Therefore, how to control KOA more
effectively and delay its development process is also a hot and difficult problem
in current research.
At present, most guidelines recommend non-steroidal anti-inflammatory drugs as first-line
drugs, aiming at relieving pain and improving clinical symptoms in patients with KOA.[5] However, gastrointestinal symptoms (such as gastrointestinal bleeding), cardiovascular
diseases (such as myocardial infarction), and other adverse reactions cannot be ignored,
which limits the clinical application. KOA can be treated through various approaches
such as oral administration of herbal medicine, topical treatment, physiotherapy,
and so on. Quercetin, a plant-derived flavonoid that is widely found in vegetables
and fruits, has multiple biological activities and pharmacological effects such as
antioxidative stress, anti-inflammatory, and immune regulation.[6] Studies have shown that quercetin could reduce oxidative stress and inhibit the
degradation of cartilage extracellular matrix; up-regulate SOD and TIMP-1; down-regulate
MMP-13 in serum, synovial fluid, and synovial tissue; and improve the degradation
of KOA.[7] Studies have shown that the mechanism of quercetin against KOA may be related to
the inhibition of TLR-4/NF-κB signaling and the decrease of IL-1β and TNF-α levels.[8] It also suggested that quercetin could inhibit the apoptosis of chondrocytes, regulate
the polarization of synovial macrophages to M2 macrophages, create chondrogenic environment
for chondrocytes, and enhance cartilage repair in KOA environment, thus playing a
protective role in cartilage.[9] Studies have shown that quercetin may be a potential candidate drug against KOA.
However, the potential targets and molecular mechanisms of quercetin against KOA have
not been systematically explained in detail.
Network pharmacology explains the occurrence and development of diseases from the
perspective of system biology and biological network balance and understands the interaction
between drugs and organisms from the overall perspective of improving or restoring
the balance of biological network.[10] By constructing a multi-level network of drug, target, pathway, and disease, the
potential molecular mechanism of drug treatment for diseases is predicted systematically.
Molecular docking is a research method that simulates the geometric and energy matching
between small drug molecules and target proteins, studies the interaction between
ligands and receptors, and predicts their binding mode and affinity.[11]
[12] Based on network pharmacology and molecular docking methods, this study will preliminarily
clarify the potential targets, signaling pathways, and molecular mechanisms of quercetin
against KOA, and provide some reference and theoretical basis for future research.
The flow chart of network pharmacology and molecular docking of quercetin against
KOA was drawn, as shown in [Fig. 1].
Fig. 1 Flow chart of network pharmacology and molecular docking.
Methods
Screening Chemical Structure and Potential Targets of Quercetin
The chemical structure of quercetin was obtained from PubChem database (https://pubchem.ncbi.nlm.nih.gov/)[13] and downloaded in the format of “SDF.” Then, the “SDF” file was uploaded to the
Swiss Target Prediction database (http://www.swisstargetprediction.ch/)[14] to predict the targets of quercetin, and the limited possibility was more than 0.
The corresponding targets were supplemented in the DrugBank database (https://go.drugbank.com/).[15] At last, the UniProt database (https://www.uniprot.org/)[16] was used to standardize the targets' names after removing duplicates.
Screening Disease Targets of KOA
The disease targets of KOA were collected in DisGeNET (https://www.disgenet.org/),[17] OMIM (https://omim.org/),[18] and GeneCards database (https://www.genecards.org)[19] with the keywords of “Knee osteoarthritis,” “Osteoarthritis, Knee,” “Osteoarthritis
of Knee,” and “KOA.” The targets obtained in the three databases were merged, and
duplicates were removed.
Screening Drug–Disease Intersection Targets
To clarify the interaction between drug targets and disease targets, the targets of
both were intersected. Venn diagram of drug and disease was drawn by online tools
(http://bioinformatics.psb.ugent.be/webtools/Venn/) to obtain the intersection targets of quercetin against KOA.
Constructing Protein–Protein Interaction Network and Screening Core Targets
The intersection targets were uploaded to the STRING database (https://string-db.org/),[20] the specie was set as “Homo sapiens,” the minimum required interaction score was
0.4, and the disconnected nodes in the network were hidden. The node information was
downloaded in the format of “TSV” and imported into Cytoscape 3.8.2 software to construct
a protein–protein interaction (PPI) network. The core targets of quercetin against
KOA were obtained by using cytoHubba plug-in, which was used for molecular docking.
GO and KEGG Enrichment Analysis
Gene ontology (GO) functions enrichment analysis and Kyoto Encyclopedia of Genes and
Genomes (KEGG) pathway enrichment analysis were performed using DAVID database (https://david.ncifcrf.gov/)[21] with p-value less than 0.05, and the specie was set as “Homo sapiens.” The biological process
(BP), cellular component (CC), and molecular function (MF) of the top 10 GO enrichment
and the signaling pathways of the top 20 KEGG enrichment were analyzed by bioinformatics
platform (http://www.bioinformatics.com.cn/).
Constructing Drug–Target–Pathway–Disease Network
Cytoscape 3.8.2 software was used to construct the drug–target–pathway–disease network.
Nodes represented drug, target, pathway, and disease, while edges represented the
interaction among them.
Molecular Docking
The structures of ligand and protein receptors were downloaded from PubChen dabases
(https://pubchem.ncbi.nlm.nih.gov/) and PDB databases (https://www.rcsb.org/).[22] PyMol 2.4.0 software[23] was used to remove water and small molecular ligands, and then AutoDockTools 1.5.6
software was used to hydrogenate and find protein active pockets. Finally, Vina was
run for molecular docking to verify the interaction between quercetin and core targets.[24]
Results
Chemical Structure and Potential Targets of Quercetin
The PubChem CID, molecular formula, and chemical structure of quercetin were obtained
from PubChem database, as shown in [Table 1]. There were 100 targets in Swiss Target Prediction database and 40 targets in DrugBank
database. After eliminating duplicate targets, 133 drug targets of quercetin were
obtained.
Table 1
Chemical information of quercetin
|
PubChem CID
|
Molecular formula
|
2D Structure
|
3D Conformer
|
|
5280343
|
C15H10O7
|
|
|
Disease Targets of KOA
Searching the databases of DisGeNET, OMIM, and GeneCards, 368 targets, 30 targets,
and 2 084 targets were obtained, respectively. After removing duplicate targets, 2,206
disease targets of KOA were obtained.
Drug–Disease Intersection Targets
There were 49 intersection targets for quercetin and KOA, as shown in [Fig. 2].
Fig. 2 Drug-disease targets Venn diagram.
PPI Network and Core Targets
PPI network of quercetin against KOA was constructed by Cytoscape 3.8.2 software,
as shown in [Fig. 3]. There were 49 nodes and 319 edges in the figure. Nodes represented proteins, while
edges represented the interaction between them. Ten core targets, ALB, ACTB, AKT1,
EGFR, ESR1, MMP9, HSP90AA1, MMP2, PTK2, and kinase insert domain receptor (KDR) were
screened out by cytoHubba plug-in.
Fig. 3 PPI network and core targets.
GO and KEGG Enrichment Analysis
The targets of quercetin against KOA were enriched to 159 GO items (p < 0.05), including 98 BP, 22 CC pathways, and 39 MF. The top 10 items of BP, CC,
and MF were selected to draw the bar graph, as shown in [Fig. 4]. These BP mainly involved the negative regulation of apoptotic process, collagen
catabolic process, and extracellular matrix disassembly. CC mainly involved extracellular
space, plasma membrane, and extracellular region. MF involved protein tyrosine kinase
activity, nitric-oxide synthase regulator activity, and protein binding. At the same
time, 59 signaling pathways were obtained by KEGG enrichment analysis; the bubble
diagram of the top 20 signaling pathways is shown in [Fig. 5]. The signaling pathways of quercetin against KOA mainly involved PI3K-Akt signaling
pathway, Rap 1 signaling pathway, FoxO signaling pathway, Ras signaling pathway, TNF
signaling pathway, ErbB signaling pathway, and so on. GO and KEGG chord diagrams ([Fig. 6]) showed the genes enrichment of each item. Then, the important PI3K-Akt signaling
pathway was drawn, as shown in [Fig. 7], in which red represents the targets of quercetin against KOA enriched in this pathway.
Fig. 4 GO functional enrichment analysis.
Fig. 5 KEGG pathway enrichment analysis.
Fig. 6
GO and KEGG chord diagrams. (A) BP chord diagram; (B) CC chord diagram; (C) MF chord diagram; and (D) KEGG chord diagram.
Fig. 7 PI3K-Akt signaling pathway.
Drug–Target–Pathway–Disease Network
The drug–target–pathway–disease network was constructed by Cytoscape 3.8.2 software,
as shown in [Fig. 8]. In the figure, yellow ellipses represent quercetin, red diamond represents KOA,
purple square represents targets, green triangle represents signaling pathways, and
gray line represents the interaction among them.
Fig. 8 Drug–target–pathway–disease network.
Molecular Docking
The core targets were docked with quercetin in turn, as shown in [Fig. 9]. The binding activity of the two could be evaluated according to the binding energy.
If the value was less than 0, it meant that they could spontaneously combine. The
lower the binding energy, the easier it was to bind. Generally, it was considered
that binding energy less than −5.00 kcal • mol−1 meant good binding activity. The binding energy of molecular docking is shown in
[Table 2], with an average of −8.47 kcal • mol−1.
Fig. 9
Molecular docking. (A) ALB, (B) ACTB, (C) AKT1, (D) EGFR, (E) ESR1, (F) MMP9, (G) HSP90AA1, (H) MMP2, (I) PTK2, and (J) KDR.
Table 2
The binding energy of molecular docking (−kcal • mol−1)
|
ALB
|
ACTB
|
AKT1
|
EGFR
|
ESR1
|
MMP9
|
HSP90AA1
|
MMP2
|
PTK2
|
KDR
|
|
−8.2
|
−9.9
|
−6.4
|
−8.3
|
−7.2
|
−10.7
|
−6.4
|
−8.7
|
−9.1
|
−9.8
|
Discussion
KOA is a common and frequently occurring disease among middle-aged and elderly people.
With the extension of life expectancy and the arrival of an aging society, the morbidity
and disability rate are gradually increasing. Epidemiological studies showed that
the prevalence rate of symptomatic KOA among middle-aged and elderly people in China
was 8.1%,[25] while the incidence rate in rural areas was as high as 16.57% and the incidence
rate of elderly people over 70 years old was obviously increasing (29.25% for women
and 24.71% for men).[26] The World Health Organization pointed out in The Global Burden of Disease that KOA has become the fourth largest disabling disease in the world,[27] which has brought more and more patients physical and mental pain and torture and
brought great mental pressure and heavy economic burden to many families and societies.[28] KOA, as a chronic degenerative disease, will cause joint injury and irreversible
disability if it is not treated properly. However, at present, there is no cure for
its recurrence, and drug treatment has reached a certain bottleneck.
In recent years, quercetin has gradually been used in the treatment of KOA. However,
the potential targets, signaling pathways, and molecular mechanisms of quercetin against
KOA still need further study. The pathophysiological mechanism and complexity of KOA
are generally considered to be a variety of BP and multiple signaling pathways involved
in its damage process. In the experimental study, it is difficult to systematically
and comprehensively analyze the relationship among drugs, target proteins, and signaling
pathways. With the rapid development of molecular biology, bioinformatics, pharmacology,
and computer science, network pharmacology explains the interaction between drugs
and targets from multi-level and multi-angle interaction networks and reveals the
mechanism of drugs under the theoretical framework of drug, target, and disease.[29] Therefore, this study attempts to analyze the potential targets and molecular mechanisms
of quercetin against KOA from a microscopic perspective by means of network pharmacology
and molecular docking, so as to provide new insights and ideas for the pathogenesis
and drug treatment of KOA.
By searching the databases, 133 drug targets and 2 206 disease targets were obtained,
including 49 intersection targets and 10 core targets. According to the results of
molecular docking, the binding energies between ligand and receptors were all less
than −5 kcal • mol−1, which indicates that quercetin can stably bind receptor protein and exert a therapeutic
effect. The binding energy between quercetin and MMP9 was the lowest, and the docking
result was the best. Quercetin can stably bind to the active pocket of MMP-9 protein
structure through hydrogen bond interaction among amino acid residues ALA-189, GLN-227,
and MEF-247. Second, ACTB, KDR, PTK 2, MMP 2, and EGFR also have lower binding energy,
which are important core targets of quercetin against KOA.
MMP is a family of zinc-dependent proteolytic enzyme, which mainly participates in
the degradation of extracellular matrix by decomposing collagen and proteoglycan and
plays an important role in the pathological process of KOA articular cartilage matrix
and the damage of chondrocytes.[30] Under normal circumstances, the synthesis and degradation of articular chondrocytes
and extracellular matrix are in a dynamic equilibrium. When KOA occurs, this state
is broken, extracellular matrix is degraded, and chondrocytes are induced to oxidize
under the stressed cell environment, and finally, chondrocyte apoptosis is induced.
Gelatinases, including gelatinase A (MMP-2) and gelatinase B (MMP-9), are the major
contributors to the degradation of extracellular matrix.[31] It is well known that MMP-2 is produced by stromal cells in the subsynovial layer,
while MMP-9 is secreted by neutrophils, macrophages, and synovial cells. Research
works have shown that the expression of these two matrix metalloproteinases is stronger
in superficial chondrocytes, and the degeneration is more obvious in superficial chondrocyte.[32]
ACTB is a housekeeping gene of KOA synovial membrane gene expression research,[33] which is relatively conservative and mainly involved in cell movement and contraction.
VEGF receptor 2, also known as KDR, has anti-apoptotic effect after activation, which
is mediated by PI3K-Akt signaling pathway.[34] VEGF can also induce the production of pro-inflammatory cytokines IL-6, CXC chemokine
IL-8, and MMP-13.[35] In addition to promoting angiogenesis, VEGF also has a strong pro-inflammatory effect.[36] Focal adhesion kinase (FAK) encoded by PTK is a non-receptor protein tyrosine kinase.
By inhibiting abnormal bone formation induced by H-type vessels and specifically inhibiting
FAK signaling in subchondral bone, subchondral bone degeneration and articular cartilage
degeneration can be alleviated.[37] EGFR is an important regulator of cartilage matrix degradation during cartilage
development, which can delay the progress of KOA by down-regulating the survival of
chondrocytes and the degradation of extracellular matrix.[38]
To further understand the BP and signaling pathways of quercetin against KOA, GO functional
enrichment analysis and KEGG pathway enrichment analysis of intersection targets were
performed. GO enrichment analysis showed that the BP of quercetin against KOA mainly
involved the negative regulation of apoptotic process, collagen catabolic process,
and extracellular matrix disassembly, which were undoubtedly closely related to the
occurrence and development of KOA. KEGG analysis showed that apoptosis, oxidative
stress, and inflammatory reaction enriched more target genes, and PI3K-Akt signaling
pathway was more important.
PI3K-Akt signaling pathway is a classical anti-apoptosis pathway, which is composed
of intracellular phosphatidylinositol kinase PI3K and important downstream target
Akt. Akt is a serine/threonine protein kinase, which is activated by extracellular
factors through PI3K-dependent phosphorylation.[39] After phosphorylation of Akt, it can inhibit the activity of downstream pro-apoptotic
protein Bax and promote the activation of anti-apoptotic protein Bcl-2, thus promoting
chondrocyte proliferation, inhibiting chondrocyte apoptosis, and slowing down the
process of KOA cartilage degeneration.[40] FoxO is a transcription factor inducing apoptosis. With the increase of age, the
expression of FoxO on cartilage surface decreases obviously, which promotes the release
of inflammatory mediators and induces cartilage degradation. The function of FoxO
in cartilage growth is mainly mediated by FoxO1, which is closely related to proliferation,
survival, and differentiation of chondrocytes.[41] When FoxO is phosphorylated by Akt, it will cause FoxO1 to transfer from the nucleus
to the cytoplasm and lose its transcriptional activity, which will help the survival,
growth, and proliferation of cells and avoid apoptosis. In addition, FoxO signaling
pathway is especially important in regulating oxidative stress. It was found that
the down-regulation of FoxO transcription factor in chondrocytes decreased the activity
of cells against oxidative stress, which may be related to the decrease of antioxidant
proteins and autophagy-related proteins.[42] That is to say, chondrocytes with decreased expression of FoxO transcription factor
are more likely to die under oxidative stress.
However, PI3K-Akt signaling pathway is closely related to autophagy, inflammation,
and metabolism of extracellular matrix in KOA pathology.[40] Specifically, activated Akt can directly activate mTOR, which is an important downstream
signaling molecule that regulates autophagy. Studies have found that inhibiting the
PI3K-Akt-mTOR pathway increased the autophagy of rat articular chondrocytes and alleviated
the inflammatory response of arthritis rats.[43] In addition, PI3K-Akt signaling pathway can also act synergistically with the downstream
protein NF-κB. Akt activates IκB kinase (IKKα), which leads to the degradation of
IκB, an inhibitor of NF-κB, and the transfer of NF-κB to nucleus. It is found that
NF-κB can lead to the production of matrix-degrading enzymes, proinflammatory cytokines,
and inflammatory mediators by coordinating multi-layer signal networks, thus accelerating
the occurrence and development of diseases.[44] In other words, simply inhibiting or activating PI3K-Akt signaling pathway will
affect the occurrence and development of KOA. Ras protein and Rap1 protein belong
to the small molecule G protein Ras superfamily. When combined with GTP, they are
activated, and when combined with GDP, they are inactivated. They are very important
to control the proliferation, differentiation, apoptosis, and survival of cells.[45] Once Ras protein is activated, many signal transduction pathways will be activated,
such as MAPK and PI3K.[34] It is well known that PI3K can be directly activated by receptor tyrosine kinase
or G protein-coupled receptor, which makes the Ras signal network very complex.[46] ErbB is a tyrosine kinase receptor. When it binds to the ligand, it activates PI3K-Akt,
a downstream signaling pathway. The research shows that inhibition or inactivation
of ErbB2 receptor leads to up-regulation of Bax expression and down-regulation of
Bcl-2 expression.[47] TNF signaling pathway can mediate KOA inflammatory reaction, promote the production
of inflammatory cytokines, and lead to synovium hyperplasia and inflammation. In addition,
quercetin can regulate chemokines, thus affecting the chemotaxis of leukocytes to
inflammatory sites. Therefore, there may be complex interactions between these pathways,
which cooperatively regulate the proliferation and apoptosis of chondrocytes, degradation
of extracellular matrix, and inflammatory reaction.
There are still some shortcomings in this research. First of all, there may be some
problems in the database itself, such as incomplete drug targets and disease targets,
low accuracy, and possible deviations in the results. Second, the research result
is a virtual prediction result. Although a large number of targets and pathways can
be screened out, only molecular docking verification is performed and experimental
verification is not added. Therefore, our future research will further verify the
mechanism of quercetin against KOA in combination with in vitro and in vivo experiments.
Conclusion
To sum up, a total of 133 drug targets and 2 206 disease targets were obtained, including
49 intersection targets and 10 core targets. Quercetin may directly or indirectly
regulate PI3K-Akt signaling pathway, Rap1 signaling pathway, FoxO signaling pathway,
Ras signaling pathway, TNF signaling pathway, ErbB signaling pathway, and chemokine
signaling pathways by regulating the expression of MMP9, ACTB, KDR, PTK3, MMP2, and
EGFR protein targets, so as to play a role in alleviating chondrocyte apoptosis, reducing
immune-inflammatory response, improving human antioxidant stress response, and strengthening
chondrocyte proliferation. Therefore, the effect of quercetin against KOA is related
to the coordinated regulation of multiple targets, multiple BP, and signaling pathways.
This study can provide a reference and theoretical basis for subsequent experiments
in vitro and in vivo.
