Network pharmacology analysis to explore the potential targets of BZD therapy in KOA
Active compounds of BZD
The compounds of Rehmanniae Radix Praeparata, Cornus Officinalis, Dipsaci Radix, Eucommiae Cortex, Eleutherococcus gracilistylus, Angelicae Sinensis Radix, Paeoniae Radix Alba, Poria Cocos, Citri Reticulatae Pericarpium Viride, and Achyranthis Bidentatae Radix in BZD were obtained from the traditional Chinese Medicine Systems Pharmacology Database and Analysis Platform (TCMSP, tcmspw.com/tcmsp.php). Biological properties of drugs including molecular weight, oral bioavailability (OB) and drug similarity (DL) can be accrued from the TCMSP database. OB and DL were used as indicators for screening effective compounds [14]. In this study, the following screening stipulations were used to reap the active compounds of BZD: OB ≥ 30%, DL ≥ 0.18 [15]. And potential targets of BZD can be also collected from the TCMSP platform.
Active compounds and predicted subchondral bone targets of BZD treatment in KOA
Genes for human osteoarthritis were gathered from the following databases: GeneCards (https://www.genecards.org/), Online Mendelian Inheritance in Man (OMIM, https://www.omim.org/), PharmGkb (https://www.pharmgkb.org/), Therapeutic Target Database (https://bidd.nus.edu.sg/bidd-databases/TTD/TTD.asp), and DrugBank (https://www.DrugBank.ca) by using “osteoarthritis” as a keyword [16]. In order to make the results more comprehensive and reliable, these five databases are combined here. Then, this list was combined with subchondral bone targets collected from the GSE51588 dataset, which was downloaded from the Gene Expression Omnibus database (https://www.ncbi.nlm.nih.gov/geo/) with osteoarthritis and subchondral bone as keywords and human was selected as the category [17]. The intersection is the differential gene of subchondral bone in human osteoarthritis. The intersection of genes that overlapped with the potential targets of BZD was considered predicted subchondral bone targets of BZD treatment in KOA. The traditional Chinese medicine botanical agents acting on these predicted targets were considered the active components of BZD.
Protein-protein interaction network construction and pathway enrichment analysis
The STRING (https://cn.string-db.org/) database was used to construct a protein-protein interaction (PPI) network to analyze the functional interactions between proteins [18]. In this study, the PPI protein interaction network removing the discrete targets can be obtained by selecting the discrete points in the hidden diagram in the settings. R software program permits bioinformatics analysis and visualization of results. Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analysis visualizes data by using the three R software installation packages, “enrichplot,” “DOSE,” and “clusterProfiler”. P-value cutoff = 0.05 and q-value cutoff = 0.05 were used to screen the KEGG results with significant differences. Then, the subchondral bone-associated neural signaling pathways of KOA pain were screened using the KEGG database (https://www.kegg.jp/).
Verification of the therapeutic effects of BZD in KOA in vivo
Animals
All animal remedy strategies observed the suggestions of the Animal Care and Use Committee of Fujian University of Traditional Chinese Medicine (Approval Number: 2019062). Forty-five healthful male Sprague–Dawley rats (8 weeks old, weighing 180–220 g each) were purchased from Hangzhou Medical College (Hangzhou, Zhejiang, China). The rats were raised at 23°C room temperature and lights (12:12 h light-dark cycle) with advert libitum food and water. The animals have been allowed adaptive feeding for one week earlier than the begin of the experiments.
Koa Rat Model
Animals were randomly assigned to sham group (n = 15), KOA model group (n = 15), and KOA + BZD treatment group (n = 15). The KOA group adopted the modified Hulth method to establish model [19]. The rats were anesthetized with isoflurane (1.5–2.0%) and disinfected with skin application. Briefly, 30 rats underwent a patellar incision on the medical side of the knee [20]. The medial collateral ligament and anterior cruciate ligament were severed, and the medial meniscus used to be resected. Once no lively bleeding was observed, the incision was sutured layer by way of layer. In the sham group, solely the knee cavity and sutured incision were exposed. Two weeks after surgery, 15 KOA rats were randomly selected for treatment with BZD (10.5 g/kg, once/d, intragastrical.). The intervention dose of BZD was transformed into the intervention dose of rats in accordance to the equivalent dose of human and rat. Sham group and KOA group were also given the equal dose of 0.9% normal saline intragastric administration.
Pain-related Behavioral Assessments
The paw withdrawal threshold (PWT) and paw withdrawal latency (PWL) tests were used as pain indicators in accordance to previous research [21, 22]. PWT was measured using a digital force meter (YLS-3E, Yima Optoelec Co., Ltd., China) with a force range of 0–2000 g and a resolution of 0.1 g. A continuous elevation force was applied to the hind paw until the retraction reflex occurred. The force intensity was recorded as the PWT. The rats were positioned in a glass chamber to acclimate for 15 min earlier than carrying out PWL. PWL measurements were performed on animals using a heat pain stimulator (BME-410C, American Institute of Biomedical Engineering) with a 12 V/10 W halogen lamp. The stimulation temperatures were 45°C–65°C, with a time accuracy of 10 ms and a cutoff time of 20 s. The time of the withdrawal response was recorded as the PWL. Before each test, the heat source was cooled until it returned to room temperature, and the glass chamber was kept dry to prevent urine from affecting the test results.
Hematoxylin & Eosin, Safranin O-fast Green, And Masson Staining
After 12 weeks of intervention, the tibial plateau was fixed in 4% paraformaldehyde solution for 48 h and decalcified with 10% ethylenediamine tetraacetic acid disodium salt (EDTA-2NA) for 8 weeks. After 8 weeks, the extra ligaments and other tissues were pruned and then paraffin embedded. The paraffin-embedded tissue was fixed on a slicer, cut into 8-µm thick slices, and blanched flat in 37°C water. The slices were positioned on slides and dried in an incubator at 55°C. Then, xylene transparent treatment and dewaxing were performed with high to low concentrations of anhydrous ethanol. Hematoxylin & eosin, safranin O-fast green, and Masson staining were carried out in accordance to the directions of the package (Solarbio, China). Light microscopy (DM4000B, Leica, Germany) was used to study cartilage and subchondral bone morphology and structure.
Detection Of Inflammatory Factors In Rat Serum By Suspension Chip Analysis
After intervention, the animals were anesthetized (isoflurane, 1.5–2%), blood was accrued from the inferior vena cava, and the serum was used to prepare in a centrifuge (1000 g, 15 min, 4°C, then 1000 g, 10 min, 4°C) for suspension chip analysis. The serum contents of 23 inflammatory factors were assessed with a multiplex assay package (Bio-Plex Pro Rat Cytokine 23-Plex Kit, Bio-Rad, USA). The results were detected using the corrected Bio-Plex 100, 200, and 3D systems.
Detection Of Neurotransmitters In Rat Serum
After sampling, the venous blood was rested for 1 h at room temperature and centrifuged (3000 g, 10 min, 4°C), and the supernatant was accrued for similarly centrifugation (12000 g, 10 min, 4°C) for UPLC-MS/MS Analysis. UPLC separation was used to carry out using the usage of ExionLC System geared up with a Waters ACQUITY UPLC HSS T3 (100 × 2.1 mm, 1.8 µm). For the mobile phase, phase A was 0.1% formic acid and 1 mM/L ammonium formate in water, and phase B was acetonitrile. The column temperature was 40°C. The auto-sampler temperature was set at 4°C, and the injection volume was 1 µL. The AB Sciex QTrap 6500 + mass spectrometer was used for the assay. Typical ion source parameters were: IonSpray voltage: +5000 V, curtain gas: 35 psi, temperature: 400°C, ion source gas 1: 60 psi, and ion source gas 2: 60 psi. Skyline software was employed for multiple reaction monitoring (MRM) data processing.
Enzyme-linked Immunosorbent Assay
Blood was used to gather from the vena cava, and serum was prepared in a centrifuge (10000 g, 30 min, 4°C). The upper serum was placed in a 0.5-mL Eppendorf tube and a − 80°C fridge for enzyme-linked immunosorbent assay (ELISA). SP, CGRP, NPY, and TH were examined with ELISA kits (TSZ, USA) in accordance to the manufacturer’s instructions. All measurements had been taken by means of skilled technicians who have been blinded to the therapy groups.
Western Blot Analyses
Western blot analyses
Standard western blot analysis was used to detect the levels of related proteins, including collagen II (Col-II), matrix metalloproteinase-3 (MMP-3), MMP-9, and MMP-13 in cartilage and receptor activator of nuclear factor-kappa B ligand (RANKL)/osteoprotegerin (OPG), alkaline phosphatase (ALP), tartrate-resistant acid phosphatase (TRACP), SP, CGRP, NPY, TH, 5-HT3, and VEGFA in the subchondral bone. The primary antibodies were as follows: Col-II, MMP-9, ALP, TH, 5-HT3 and VEGFA (1:1000, Abcam, USA), MMP-3 (1:2000, Abcam), MMP-13 (1:2000, Proteintech, China), RANKL (1:1000, Proteintech), OPG (1:300, Abcam), TRACP (1:4000, Abcam), SP (1:2000, Sigma–Aldrich, USA), CGRP and NPY (1:1000, CST, USA), glyceraldehyde-3-phosphate dehydrogenase (GAPDH) (1:10000, Abcam, USA) and β-actin (1:1000, Abcam, USA). The membrane was incubated with the corresponding secondary antibody (goat anti-rabbit IgG, 1:5000, Bioss, China) conjugated with HRP for 1 h at room temperature. Immunoreactive proteins were visualized the usage of an enhanced chemiluminescence western assay kit (Thermo Fisher Scientific, USA). Image Lab software program was used to quantify the density of each strip.
Exploration Of The Potential Neuromodulation Mechanism Of Bzd With Resting-state Fmri
The rats were anesthetized (with 1.5–2.0% isoflurane) and constant to a headrest with a built-in coil and physique tube. Isoflurane (0.5–1.0%) was delivered in a 30% O2/70% N2 mixture through a nose cone. The whole system was positioned in a magnet. Respiratory rate and rectal temperature were closely monitored throughout the procedure. The animal’s core temperature was maintained at 37°C ± 0.5°C the use of a cycling heated bed. The T2W structural image acquisition was implemented in a 9.4 T MRI scanner (Biospec 9.4, Bruker Medizenk, Germany), and an orthogonal surface coil with a diameter of 35 mm was utilized. T2 weighted imaging was used to swiftly obtained with a relaxation enhancement sequence with the following parameters: repetition time (TR) = 2500 ms, echo time (TE) = 33 ms, field of view (FOV) = 3.2 × 3.2 cm, averages = 4, slice number = 21, slice thickness = 1.0 mm, matrix size = 256 × 256, bandwidth = 326084 Hz, echo spacing (ESP) = 10000 ms, refocusing angle = 180°, excitation angle = 90°, echo train length (ETL) = 8, and k-zero = 3.
All resting-state fMRI statistics used AFNI (National Institutes of Health, USA), the Advanced Normalization Tools (ANTs) (stnava.github.io/ANTs/), and FMRIB's Software Library (FSL) for processing. Processing of the T1-weighted image included: 1) “N4BiasCorrection” and “denoise image” commands to perform field deflection correction and image denoising, respectively; 2) manually removing the skull and scalp tissues of the image; and 3) registering the preprocessed images to the standard sigma template and partitioning the image [23]. The preprocessing of resting-state fMRI included the following steps: first, slice-timing correction and despike were carried out. Then, motion correction was used to limit the artifacts caused by head movement. Rigid registration of the fMRI and T1W images was carried out with the ANTs tool, and then time filtering was carried out at 0.01–0.1 Hz. Finally, the preprocessed resting-state fMRI data were normalized to sigma templates, regional homogeneity (ReHo) indicators were calculated, and amplitude of low-frequency fluctuation (ALFF) and functional connection matrices were calculated after smooth processing. The command “3dttest ++” in AFNI was used to test the ReHo/ALFF/functional connectivity indicators in a two-sample T test between two pairs (P < 0.05, cluster size ≥ 10 voxels).
Statistical Analyses
SPSS software (v.25.0; IBM, USA) was used for statistical analysis of the experimental data, and all values obtained from the experimental results are expressed as mean ± standard deviation. For normally distributed samples, one-way analysis of variance was used to compare groups. When the samples did not follow a normal distribution, a non-parametric test was used to compare the differences between groups. P-values of < 0.05 were considered statistically significant.