Network Pharmacology Approach and Experimental Verication of Huashi Dingtong Decoction Against Knee Osteoarthritis

Background: The aim of this study is to clarify the ingredients and targets of HSDTT against KOA by network pharmacology,and to verify the mechanism of HSDTT in treatment of KOA in vivo. Methods: Ingredient-target network for HSDTT and KOA was created to identify the potential targets, protein-protein interaction network was used to nd the key targets of HSDTT in treatment for KOA, GO enrichment and KEGG pathway was conducted to illuminate the pathway related to KOA treat by HSDTT. Rat model of KOA was established by joint injection in papain.The morphology of cartilage were assessed by H&E. ELISA was used to detect the contents of inammation cytokines in synovial uid and synovium. The expression of the pathway protein were assessed by PCR. Results: The results of network pharmacology demonstrate that there are 440 ingredients of HSDTT against knee osteoarthritis by 478 targets.The KEGG enrichment analysis showed that PI3K-Akt signaling pathway, MAPK signaling pathway were the key pathways for HSDTT to treat KOA. Morphology of cartilage was improved in the HSDTT group when compared with the model group. Our experiment show that HSDTT can reduced the expressions of p38 and p53 in cartilage, increased the expression of collagen (cid:0) . The contents of IL-1β and TNF-α in the synovium and COX-2 and PGE-2 in synovial uid were decreased signicantly in the HSDTT group when compared with the model group. Conclusions: Our study indicadites that HSDTT is capable to alleviate inammation and delay the progression of KOA by p38MAPK signaling pathway.


Introduction
Knee osteoarthritis (KOA) is a chronic articular disease which characterized by cartilaginous degeneration, synovial membrane in ammation, hyperosteosis of subchondral bones.Most of KOA patients go to hospital for joint pain, stiffness, swelling, and dysfunction [1,2]. Without timely and effective treatment, it will lead to different degrees of malformation in joint or even disability with the development of this disease [3]. It is reported in an epidemiological study that KOA is the main disease which acount for 7.1% among musculoskeletal disorders [4]. Furthermore, in a multivariate analysis, disability was positively related with cardiovascular events and all-cause mortality among patients with KOA [5]. Therefore, how to treat KOA effectively and delay the progression of it is an urgent problem at the present stage.
In the past few decades, most of the experts have focused on chondropotection when they treated KOA [6]. However, in recent years, many clinical researches show that KOA is not just a disease related to cartilage degradation but a whole jiont disease which involving cartilage, synovial membrane, ligaments, bones, muscles [7]. Furthermore, while KOA has been classi ed as non-in ammatory disease for a long time, the degree of in ammation is now considered as a critical factor in KOA pathology [8]. Several studies indicated that in the early-stage of KOA, many in ammation cytokines like tumor necrosis factor alpha (TNF-α) and interleukin-1 beta ( IL-1β) ,which triggered by factors such as biomechanical stress ,will active the signaling pathways and then in turn accelerate the progression of KOA [9].
Mitogenactivated protein kinase (MAPK) is one of the signaling pathways in cartilage which regulates the expression of proin ammatory cytokines [10]. The cascades of this signaling pathway include p38 MAPK, extracellular regulated kinase (ERK) and c-Jun N-terminal kinase (JNK) [11]. p38 MAPK is considered as a positive regulator in differentiation,in ammation and apoptosis of chondrocyte [12,13].Once the cartilage was damaged, the fragments of it will trigger the expression of the in ammatory cytokines, and further affect the synovial membrane and induce the degradation of cartilage by activiatng the protein kinases of this pathway [10].
Huashi Dingtong decoction(HSDTT) is a traditional Chinese medicine (TCM) formula, which consists of one protein and eight herbs: Squama Manis ,Cinnanmomi Cortex, Rhizoma Dioscoreae, Angelicae Sinensis Radix ,Cortex Acanthopanacis, Radix Angelicae Biseratae ,Notopterygii Rhizoma Et Radix, Atractylodes lancea(Thumb.)DC ,Licorice. HSDTT was presented in the 'Orthopedics Experience' formulary written by Rugao Lin, a famous doctor of Traditional Chinese Osteopathy & Traumatology and was used to treat KOA commonly by TCM doctors.However, the potential mechanism that HSDTT delay the progression of KOA remains unknown. In our study, we try to use network pharmacology method to present the drug-ingredient-target network of HSDTT and analyse the speci c mechanism of the HSDTT effects on KOA. Finally, an animal experiment was conducted to verify whether HSDTT would alleviate in ammation and delay the development of KOA based on the result of the network pharmacology.  [14] and TCMSP database (https://tcmspw.com/index.php) [15]. All the 2D structure of ingredients were download from TCMSP database and pubchem(https://pubchem.ncbi.nlm.nih.gov/) [15,16]. The inclusion criterias of these ingredients are assessed by SwissADME (http://www.swissadme.ch/index.php) [17]. If the uploading ingredient in SwissADME meet the rules that it shows "high" in GI absorption of "Pharmacokinetics " and more than two "yes" in "Druglikeness",we will bring it into our study [18]. KOA-related genes were obtained from TTD database (https://db.idrblab.org/ttd) [19],OMIMdatabase(https://www.omim.org/) [20],Genecardsdatabase(https://www.genecards.org) [21].

Drug-ingredients-targets network
The targets, which "probability">0, of included active ingredients were selected by Swiss TargetPrediction (http://www.swisstargetprediction.ch/) [22]. The drugs-ingredients-targets network was established by Cytoscape 8.0.0.The degree of each node in the network was analyzed by using Network Analyzer Tool in cytoscape [23].

Protein-protein interaction (PPI) network construction and analysis
The selected targets were uploaded in STRING databases(http://string-db.org) to con rm the potential proteins interactions [24]. Then the le named 'interaction' was put into Cytoscape 8.0.0 to construct the PPI network. The degree and combind-score analyzed by 'Network Analyzer Tool' indicated the importance of each target. Gene ontology (GO) enrichment and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analysis were performed in R with the clusterPro ler package [25].

Chemicals and reagents
Huashi Dingtong decoction was purchased from TCM Pharmacy of the third hospital a liated to Fujian University of Traditional Chinese Medicine. Ethyl carbamate and Hematoxylin-Eosin staining kits was obtained from Beijing Solarbio Science & Technology Co., Ltd (Beijing,China). 500 U of type II collagenase from clostridium histolyticum was purchased from Sigma-Aldrich. 0.9% Sodium Chloride Injection was from Jiangxi Kelun Pharmaceutical Co.,Ltd (Jiangxi,China). ELISA kits of IL-1β,TNF-α,COX-2,PGE-2 was provided by Shanghai Westang Bio-tech co.,Ltd (Shanghai,China).

Animals experiments, grouping and model establishment
32 male Sprague-Dawley (SD) rats weighing 180-220g were provided by Shanghai SLAC Laboratory Animal Co.,Ltd (Shanghai, China),the certi cate number is SCXK 2017-0005. Rats housed in SPF facility were randomly divided into 3 groups: HSDTT group, Model group, Control group.There were 12 rats in Control group and Model Group, 8 rats in HSDTT group. Rats in HSDTT group were received HSDTT decoction at a dose of 4.68g/kg/d by intragastric administration for 28 days after model successfully established. The other two groups were recived saline for the same days. The dose of intragastric administration was calculated by the surface area of the body. KOA model was established by intra-articular injection with type II collagenase (0.4mg/ml). Rats in model group and HSDTT group were injected (0.4ml) into the medial side of the right knee joint of the hind limbs in the rst, fourth day. The control group was injected with 0.4mL saline. After the model estabilshed successfully for 1 week, four rats were randomly selected to identi cate and evaluate the model in each group except HSDTT group. The other rats were sacri ced after intragastric administration for 28 days.
The procedures of this reseach complied with the Chinese Animal Welfare Law and approved by Fujian University of Chineses Traditional Medicine (Protocol ID:NO. FUTCM-2019028).

Measurement of the Knee Joint Diameter
The transverse diameters of the right knees were measured at day0, day7,day14,day21,day28,day35 by digimatic caliper (Mitutoyo, Kanagawa, Japan).The knee joints were exed at 90° and measured the distance between the left and right highest points of the knee.

Hematoxylin-eosin staining (H&E)
The right hind knee joint of rats were dissected and xed in 10% formalin for 72h at room temperature.Then tissues were decalci ed in ethylenediaminetetraacetic acid for 6 weeks. Subsequently, knee joints were routinely sliced (5 μm) after dehydration and para n imbedding. Finally, the slices were stained with hematoxylin-eosin staining kit and observed under a light microscope (magni cation, 100×).

Enzyme-linked immunosorbent assays (ELISA)
The contents of IL-1β and TNF-α in the synovium and COX-2 and PGE-2 in synovial uid were measured with ELISA kits. In brief, the procedure to test for IL-1β,TNF-α,COX-2 and PGE-2 was followed by the manufacturer's instructions. Each group was selected 3 rats to assay and all the samples were assessed for three times.

Statistical analysis
All the experimental data are expressed as the mean ± SD. Two-sample t-test or Wilcoxon rank sum test were used to compare the difference between the two groups by software SPSS20.0. The signi cant difference was dicided by p<0.05. . Total 3096 KOA-associated genes were searched from TTD/GeneCards/OMIM databases. As is shown in Fig. 2 and Table S2, 478 genes which were ralated to HSDTT and KOA, were identi ed with Venn digram. Considering the large number of these genes, the rst 200 genes were chosen to construct the drugs-ingredients-targets network with Cytoscape 8.0.0 software. It is composed of 605 nodes, which contain 9 chinese herbs, 403 ingredients, 200 targets. Hexagons represent different ingredients, circles represent different herbs, rhombus represent different targets (Fig.3).

PPI network construction and core target screening
The obtained potential therapeutic targets associated with KOA and HSDTT were input into String database.Then the le about PPI network from String database were input into Cytoscape 8.0.0 software.There were 478 nodes and 11506 edges in network. The larger and darker color nodes indicates the greater degree of targets. The thicker and the darker color lines represent the larger combined-score between targets (Fig. 4).

Enrichment of GO analysis and KEGG pathway analysis
To analyse the enrichment of Gene Ontology (GO) and KEGG pathway of 478 targets, the R with clusterpro ler package was used to make the results more visualized.As is shown in the Fig.5A, the top 3 functions of GO enrichment, which is related to protein binding and biological process, were protein tyrosine kinase activity,protein serine/threonine kinase activity,transmembrane receptor protein kinase activity.The KEGG enrichment, which is associated with signaling pathway, was performed in Fig.5B . It shows that the effects of HSDTT against KOA were connected with MAPK signaling pathway. The biological process that is linked to MAPK signaling pathway was displayed in Fig.6. We further investigated the in uence of HSDTT on partly of this pathway in vivo.

The effect of HSDTT on cartilage tissue of knee in KOA rats
As is shown in the Fig.7A-B,the surface of cartilage was smooth without cracks,the cartilage structure was clearly, such as regularly arranged chondrocyte and the complete tidemark in the HE staining of the control group. But in the staining of the model group, the surface of cartilage was damaged seriously, cells in transition zone arranged disordered with subchondral bone and erosion pannus formation. The makin score in model group is higher than the control group signi cantly. (p<0.05, Fig.7C). It con rms that the model of knee osteoarthritis was successfully established. After KOA rats treated with HSDTT for one month, whatever the arrangement of cells in transition zone and the expression of collagen in HSDTT group were signi cantly improved when compared with the model group. (p<0.05,Fig.7D-G) 3.2.2 The effects of HSDTT on the mRNA expression of p38 and p53 Levels of p38 and p53 were further indicated that the changes of cartilage are related to the mechanism that HSDTT inhibits the p53 signaling pathway of KOA.In comparison with the control group,the mRNA expression of p38 and p53 were increased in model group(p<0.01). However,after intervention with HSDTT, the protein expression of P38 and p53 decreased signi cantly(p<0.01, Fig.7H-I).

The effects of HSDTT against in ammation in KOA rats
To verify the e cacy of HSDTT to KOA rats, knee swelling of the right legs were measured (Fig.8A). On day 7, transverse diameters of the right knees in model group was signi cantly larger than that in control group (p<0.05). After the rats gavaged for 28 days, the knee joint diameter of HSDTT group was signi cantly smaller than that of model group on day 35 (p<0.05). Furthermore, the synovium and synovial uid were used to test the expression of in ammatory cytokine. As is shown in Fig.8B-C, the content of IL-1β and TNF-α in synovium signi cantly increased in model group when compared with control group (p<0.01). Besides, in comparison with control group, rats in model group were observed with a signi cant increase in content of COX-2 and PGE-2 in synovial uid(p<0.01, Fig.8D-E). By contrast, the expressions of IL-1β ,TNF-α,COX-2 and PGE-2 in HSDTT group are signi cantly lower than in model group (p<0.01, Fig.8B-E). It indicated that HSDTT has a good effect in alleviating in ammation of KOA.

Discussion
Nowadays, KOA has been a worldwide public healthy problem with its high disability rate among the people over 50 years old [26]. Therefore, a large number of researchers are devoted themselves to explore an effective method to halt the progression of KOA. As a common prescription of KOA, HSDTT has a good effect in early-stage of KOA, but the mechanisms of it are still unknown. Network pharmacology is a powerful tool which integrates other platforms and technologies to investigate the correlation between the ingredients of TCM and the targets of disease [27].In present study, the mechanism of HSDTT against KOA was analyzed by the network pharmacology and then validated by animal experiment.
In this study, 3096 potential targets of KOA and 1176 ingredients of HSDTT were analyzed through network pharmacology. The results indicated that HSDTT played a critical role in treating KOA by regulating some targets and signaling pathways. As the component of HSDTT, Atractylodes lancea (Thumb.)DC not only reduce the expression of in ammatory cytokines in serum, but also decreases the expression of beclin-1 protein in synovial tissue of rat [28].Moreover, Radix Angelicae pubescentis and licorice all have a great effect on alliviating in ammtory ,and Radix Angelicae pubescentis has been shown to relieve pain by inhibiting the expression of TNF-α and IL-1β [29,30]. Angelicae Sinensis Radix inhibits apoptosis of chondrocyte via suppressing JNK and p38MAPK pathways [31].
While some of components have been researched as a single herb in other studys, this decoction with all these components treating KOA remains to be explored. The results of network pharmacology indicated that the reason why can HSDTT treat KOA effectivly is that this decoction works through multiple components and targets. PPI network showed that the key targets of HSDTT in treatment of KOA were ATK1, GAPDH, IL-6, TP53, ALB, VEGFA, MAPK3, TNF, EGFR, MAPK1.We further revealed by GO enrichment analysis and highlighted that HSDTT delayed the progression of KOA by regulating the activity of protein serine/threonine kinase ,endopeptidase ,protein tyrosine kinase and so on. The KEGG enrichment analysis showed that PI3K-Akt signaling pathway, MAPK signaling pathway were the key pathways for HSDTT to treat KOA.
P38 MAPK is one of pathways in MAPK signaling pathway,which regulated the expression of proin ammatory cytokines such as IL-1β and TNF-α to accelerate cartilage degeneration in KOA [32]. When growth factors and pro-in ammatory cytokines bind to corresponding receptors on the cell membrane, the pathway will be triggered.Then p38 will be phosphorylated and in turn active other in ammatory cytokines in downstream [9]. Recently, researchers have reported that synovial in ammantion is another important factor which can aggravate KOA by osteophytosis, cartilage degeneration and in ammation [33,34]. some sudies have shown that overexpressions of COX-2 and PGE-2 in synovial tissues were probably induced by proin ammatory mediators like IL-1β and TNF-α actived by p38 MAPK signaling pathway [35,36].
We further established a model of KOA in vivo by the injection in type II collagenase and com rmed that HSDTT was effective in anti-in ammation .The results indicate that HSDTT is capable of delaying the progression of knee osteoarthritis, probably via suppressing the mRNA expressions of p38 and p53, lowering in ammatory cytokine (IL-1β and TNF-α ) and then decreasing the release of COX-2 and PGE-2, increasing collagen II.
This study had some limitations. Firstly,the network pharmacology is a new tool analyzed by the existing active ingredients of TCM, but the HSDTT decoction still has undiscovered ingredients. Secondly,we just selected only one signaling pathway identi ed through network pharmacological analysis to con rm the mechanism against KOA in vivo.Therefore, the machanisms that HSDTT treating for KOA analyzed by network pharmacology need more experiments in vivo and in vivo to com rm.

Conclusions
In this study,a network pharmacological analysis was combined with experiment in vivo to clarify the mechanism of HSDTT against KOA. The results of network pharmacology show that HSDTT treats KOA by multiple targets and pathways. Further,via experimental evidence, we show that HSDTT delays the progression of KOA by regulating p38MAPK signaling pathway, thus alleviating in ammation and protecting cartilage.This present research provides evidence to support the clinical use of HSDTT for the treatment of KOA.   Table.S1.) of HSDTT. The circle represents different drugs of HSDTT.

Figure 4
The protein-protein interaction (PPI) network.(A) The PPI network containing 478 nodes and11506 edges.
(B)The PPI network for the candidate targets ranked in the top 100. The PPI network containing 100 nodes and 2980 edges.(C)The PPI network for the candidate targets ranked in the top 30.The PPI network containing 30nodes and 420 edges.   The pharmacological mechanism of HSDTT against KOA. MAPK signaling pathway play an important role in treating KOA with HSDTT.