Proteomic Analyses Reveals the Mechanism of Acupotomy Intervention on the Treatment of Knee Osteoarthritis in Rabbits

Acupotomy intervention (AI) is an available treatment for knee osteoarthritis (KOA) in China, which is a common health problem over the world. However, the underlying mechanism of AI on the KOA treatment is still unknown. To further understand the mechanism of acupotomy in treating KOA, the morphological observation and TMT proteomic analyses were conducted in rabbits. By using X-ray and MRI, we found that the space of the knee joint was bigger in AI than in KOA. Moreover, the chondrocytes were neatly arranged in AI but disordered in KOA. With proteomic analyses in chondrocytes, 68 differently accumulated proteins (DAPs) were identified in AI vs. KOA and DAPs related to energy metabolism and the TCA cycle were suggested to play a central role in response to AI. Furthermore, AIFM1 was proposed to be an important regulator in controlling the energy production in mitochondrial. Besides, FN1, VIM, COL12A1, COL14A1, MYBPH, and DPYSL3 were suggested to play crucial roles in AI for the treatment of KOA. Our study was systematically elucidating the regulation mechanism of acupotomy intervention in the treatment of KOA.


Introduction
Knee osteoarthritis (KOA) is a chronic degenerative disease mainly characterized by knee ligament relaxation, degeneration of knee cartilage, and peripheral muscle atrophy [1]. KOA is caused by wear and tear subsequently and leads to abnormal remodelling and joint failure [2]. During this process, a series of physiological and molecular changes in multiple tissues inside the joint were observed [3]. As Haartmans et al. (2021) pointed out that a group of molecules rather than one specifc biomarker will provide the possible therapeutic targets for KOA [3]. Terefore, a systematic understanding of the proteomic changes in KOA development was required.
As the common methods of traditional Chinese medicine (TCM) used for the treatment of KOA, acupotomy intervention (AI) was widely used to relieve cartilage tissue pain and restore the movement of the knee joint [4]. In rabbits, acupotomy could alleviate cartilage deterioration by regulating the OPG/RANKL signaling pathway to inhibit subchondral bone resorption [5]. Similarly, Huang et al. (2020) reported that activating the PI3K/Akt signaling pathway to inhibit the chondrocyte apoptosis in KOA rabbits might be a possible mechanism of acupotomy in treating KOA [6]. Moreover, the GSK3 beta-cyclin D1-CDK6 pathway was also regarded as a responsible mechanism of acupotomy to alleviate cartilage degeneration in KOA rabbits [7]. However, the systematic change of acupotomy intervention in treating KOA is still unknown. In the present study, the morphological and proteomic changes in rabbits treated with KOA and AI were analyzed to investigate the underline mechanism of acupotomy intervention in treating KOA.

Animals and Study
Design. Nine male New Zealand white rabbits (6 months old) weighing approximately 2.0 ± 0.5 kg were raised without limitations of food and water in separate cages with 50%-60% humidity and 20°C-25°C temperature. After acclimation for one week, rabbits were randomly divided into 3 groups including the blank control group (CK, n � 3), rabbits with knee osteoarthritis (KOA, n � 3), and KOA rabbits treated with acupotomy intervention (AI, n � 3).

KOA Induction and Acupotomy
Intervention. Te induction of KOA rabbits was performed following the methods described by Wang et al. (2019) [8]. As the KOA model was established for one week, three points located (on left hind leg) at the tendons of quadriceps femoris, vastus medialis, and vastus lateralis were selected as the fxed intervention points. Polyvinylpyrrolidone was used for routine disinfection on the intervention points for three times and the three intervention points located at the tendons of quadriceps femoris, vastus medialis, and vastus lateralis. After disinfection, the blade of the acupotome was vertically inserted into the skin and parallel to the longitudinal axis of the tendons. Te local adhesion to the direction of tendons and bone connection was released by longitudinal dredging and transverse stripping. After the acupotomy intervention, we pressed the intervention points for 1 min. Te acupotomy intervention was performed every 7 d for 4 weeks.

Te Observation of Knee
Joints by using X-Ray, MRI, and a Light Microscope. After KOA induction, the rabbits were anesthetized by intraperitoneal injection with 3% sodium pentobarbital solution (1.5 mL/kg). For X-ray examination, the anteroposterior radiographs were taken in the supine position with hip fexion at 30°, knee extension at 0°, hip abduction at 15°, and keeping the patella in right forward. Te radiating tube was 110 mm away from the knee joint. Te left hind leg was extended and placed in the decubitus position, while the right hind leg was bent at 70°, and the knee was bent at 45°. Te parameters for X-ray examination were set as follows: tube voltage 50 kV, current 250 mA, exposure strength 32 mAs, and exposure time for 128 ms.
Te histomorphological observation of femoral condyle cartilage in rabbits was performed with hematoxylin-eosin (HE) staining and safranin O-fast green staining using standard protocols (Solarbio, Beijing, China).

Protein Extraction.
Tissues used for protein extraction were stripped from tibia. Te complex tissues containing cartilage-subchondral bone were intercepted from the central weight-bearing area of the tibia. Ten, the cartilage tissues were isolated from the cartilage-subchondral bone complex tissues and put into liquid nitrogen directly. Te chondrocytes from each treatment were grinded by liquid nitrogen into cell powder and transferred to a 5 mL centrifuge tube. Four volumes of lysis bufer (8 M urea, 1% protease inhibitor cocktail) were added to the cell powder and sonication on ice three times by using a high intensity ultrasonic processor (Scientz, Ningbo, China). Te mixture was centrifuged at 12,000 g at 4°C for 10 min. After that, the supernatant was collected and the protein concentration was measured with a BCA kit according to the manufacturer's instructions ( Figure 1). Each group had three replicates.

TMT Labeling and LC-MS/MS Analysis.
Te protein solution was reduced with 5 mM dithiothreitol (DTT) at 56°C for 30 min and alkylated with 11 mM iodoacetamide in darkness at room temperature for 15 min. After that, the mixed solution was diluted with TEAB (100 mM) to reduce the urea concentration (less than 2 M). Ten, the protein in the mixed solution was digested overnight by trypsin with a mass ratio (trypsin to protein) of 1 : 50 and for another 4 h with a mass ratio (trypsin to protein) of 1 : 100. After digestion, peptide was desalted and vacuum-dried. Peptide was redissolved in 0.5 M TEAB and labeled by TMT following the manufacturer's instruction.
Te tryptic peptides were dissolved in 0.1% formic acid (solution A) and loaded onto a home-madereversed-phase analytical column (15 cm length, 75 μm i.d.). Te settings for UPLC (EASY-nLC 1000 UPLC, Termo Fisher Scientifc, USA) were shown as follows: (1) gradient increase of solution B (0.1% formic acid in 98% acetonitrile) from 6% to 23% for 26 min; (2) 23% to 35% of solution B in 8 min; (3) 35% to 80% of solution B in 3 min; (4) kept at 80% for 3 min. Te fow rate was kept at 400 nL·min-1, the electrospray voltage was 2.0 kV, the m/z ranged from 350 to 1800, and the resolution of the orbitrap was 70,000. Peptides were selected for MS/MS (Termo Fisher Scientifc, USA) using NCE setting as 28, and the resolution of the orbitrap was 17,500. Automatic gain control was 5E4, and fxed frst mass was 100 m/z. Te MS/MS data were processed by using the MaxQuant search engine (v.1.5.2.8). Tandem mass spectra were searched against the human uniprot database concatenated with a reverse decoy database. Trypsin was specifed as a cleavage enzyme allowing 4 missing cleavages. Carbamidomethyl on Cys and oxidation on Met were specifed as fxed modifcation and variable modifcation, respectively. Te mass tolerance for fragment ions was set as 0.02 Da. FDR was adjusted to <1%, and the minimum score for modifed peptides was 40. Te proteins with more than a 1.3-fold change and those which passed the Student's t-test (p < 0.05) were selected as the diferentially accumulated proteins (DAPs).

Bioinformatic
Analysis. Te Venn diagram was drawn by Venny v2.1, an online software package (https:// bioinfogp.cnb.csic.es/tools/venny/index.html). Hierarchical clustering analysis was carried out using Cluster 3.0 software with the log-transformed fold induction density value in diferent cold acclimation treatments. A complete linkage algorithm was employed to construct the tree diagram by Treeview software (version 1.1.3). Te protein-protein interaction network of DAPs was constructed by using String 11.0 (https://string-db.org) and analyzed by Cytoscape 3.5.1. Te functional modules in the PPI network were identifed by MCODE in Cytoscape.

Statistical Analysis.
For morphology observations and proteomic analyses, three biological replicates were performed in the present study. Statistical signifcance was tested by one-way ANOVA analysis or Duncan's post-test (p < 0.05) for multiple comparisons with SPSS 19.0 (SPSS, Chicago, IL USA). Data in fgures and tables were expressed as means ± SE.

Morphological Observation.
Compared with CK, the space of the knee joint in the KOA group was smaller (Figures 2(a) and 2(b)). However, the joint space in the AI group was bigger than that in the KOA group (Figures 2(b) and 2(c)). Moreover, the MRI results found that more fuid accumulated in the articular cavity and the surface of  Evidence-Based Complementary and Alternative Medicine femoral condyle cartilage was much rougher in the KOA group than that in the CK group. On the contrary, less fuid was observed in the articular cavity and the surface of femoral condyle cartilage was smoother in the AI group than in the KOA group ( Figure 3). Te results of HE staining and safranin O-fast green staining showed that the surface of the cartilage was uneven and the arrangement of chondrocytes was disordered in the KOA group, but the situation was conversed in the AI group ( Figure 4). Besides, several chondrocytes were tightly packed together. Tese results suggested that acupotomy intervention obviously alleviated the damage caused by osteoarthritis in knees.

Protein Identifcation and Diferently Accumulated Protein Analysis.
Te changes of protein profles in three groups were analyzed by the TMT proteome in the present study. A total of 412 diferently accumulated proteins (DAPs) were identifed among the three groups (data not shown). As shown in Figure 5, 132 DAPs, 100 DAPs, and 7 DAPs were specifcally expressed in KOA vs. CK, AI vs. CK, and AI vs. KOA, respectively. In addition, 16 DAPs were shared by the three comparable groups.
To explore the regulatory mechanism of acupotomy intervention which alleviated KOA in rabbits, 68 DAPs in AI vs KOA are listed in Table 1. According to the protein function, these DAPs were divided into 9 groups including energy metabolism, tricarboxylic acid cycle (TCA cycle), amino acid metabolism, lipid metabolism, protein metabolism, transportation, cell development, signaling transduction, and others. Te results of heatmaps of DAPs in diferent function groups showed that most of DAPs related to cell development and signaling transduction were increased in the AI group compared with the KOA group (Figures 6(c) and 6(h)). On the contrary, the abundances of DAPs involved in energy metabolism, the TCA cycle, amino acid metabolism, transportation, lipid metabolism and protein metabolism, and others were almost higher in the KOA group than in the IA group and CK (Figures 6(a), 6(b), 6(d)-6(g)).  Figure 2: Te X-ray images of rabbit's left knee among diferent treatments. (a) CK represents control, (b) KOA represents rabbits with knee osteoarthritis, and (c) AI represents KOA rabbits treated with acupotomy intervention. Compared with CK, the space between knee joints in KOA rabbits was smaller, but the knee joint space in the AI group was similar to that in CK. 4 Evidence-Based Complementary and Alternative Medicine

Te Efect of Acupotomy Intervention on Knee
Osteoarthritis. Knee osteoarthritis (KOA) induces a persistent abnormal load within the joint, and cartilage degeneration has always been the focus of KOA research [9][10][11]. Te morphological observation by Safranin O staining showed that KOA causes the defect of cartilage surface and disorder of the chondrocyte arrangement, which is regarded as symptoms of cartilage degeneration [5]. Similar to the previous study, we also found that the surface of KOA cartilage was rough and the arrangement of chondrocytes was disordered ( Figure 4). Moreover, the space of knee joint was smaller between KOA cartilage, which is prone to cause cartilage wearing [12,13]. Cartilage wearing is one of main reasons for cartilage degeneration during the course of KOA. In the present study, the space of knee joints in KOA rabbits was smaller than CK, which indicated that cartilage wearing was an important reason for KOA (Figures 2 and 3). Accumulating evidences proved that acupotomy therapy was an efective treatment for KOA [14][15][16]. For example, Wang et al. (2019) reported that the application of acupotomy clearly alleviated the symptoms of KOA in rabbits by changing the arrangement of fbrous tissue [8]. Besides, the application of acupotomy in rabbits signifcant prevented both cartilage surface erosion and subchondral bone loss by histological observation [5]. In clinical treatment, Cho et al. (2019) reported that the combined therapeutic efect of miniscalpel acupuncture and splint therapy on osteoarthritis of the hand seems to have positive results [17]. Consistent with the previous study, our results of HE staining and safranin O-fast green staining in the present study also exhibited an orderable arrangement of chondrocytes in AI ( Figure 4). Moreover, several studies on acupotomy therapy revealed that acupotomy could release adhesions to alleviate the damage caused by KOA [18,19]. As shown in Figures 2  and 3, the space of knee joints was bigger in the AI group than that in the KOA group. Tese results were, to some extent, supported the opinion proposed by previous studies. Taken together, the morphological observation presented in our study indicated that acupotomy intervention could be an efective treatment for KOA by altering the morphology of the knee joint and the arrangement of chondrocytes. However, the mechanism of acupotomy on treating KOA was still unknown. Terefore, the proteomic analyses of acupotomy on KOA rabbits were performed in this study.

DAPs Related to Energy Metabolism and TCA Cycle.
As the main subcellular organelles to produce energy in cells, mitochondria control the conversion of energy and primary metabolites by various biological processes, such as glycolysis, the TCA cycle, the oxidative phosphorylation system, and so on [20]. Previous studies found that mitochondrial function was tightly associated with osteoarthritis pathophysiology [21,22]. Moreover, some evidences showed that osteoarthritis induced the energy production in chondrocytes [21,23,24]. In other words, more energy production was a symptom of osteoarthritis in chondrocytes. Similar to previous studies, we also found that DAPs related to energy metabolism were mostly upregulated in KOA compared with CK ( Figure 6(a), Table 1). However, AI treatment reduced the abundances of these DAPs (Figure 6(a), Table 1). For example, the subunits of cytochrome c oxidase (complex IV) and cytochrome b-c 1 complex (complex III), which are the main components of respiratory chain to product energy in mitochondria [22], were frstly up-regulated in KOA vs. CK but down-regulated in AI vs. KOA (Figure 6(a), Table 1). Furthermore, we found that the ATPase inhibitor (ATP5IF1, assession: G1TES2) exhibited a reverse trend compared with DAPs related to cytochrome c oxidase. ATP5IF1 was proved to function as a regulator to avoid the consumption of cellular ATP [25]. Te accumulation of ATP5IF1 in AI vs. KOA but degradation in KOA vs. CK indicated that ATP5IF1 might function as a crucial regulator induced by AI to balance the energy synthesis and consumption in chondrocytes of rabbits. Besides, the TCA cycle was found to play a central role in energy production in osteoarthritis chondrocytes by changing the central metabolism [21,24]. In the present study, 11 DAPs involved in the TCA cycle were totally upregulated in KOA vs. CK but down-regulated in AI vs. KOA (Figure 6(b)). Based on these results, it was suggested that KOA promoted the accumulation of proteins related to energy metabolism and the TCA cycle but down-regulated  Evidence-Based Complementary and Alternative Medicine  of ATP5IF1 to product more energy as an immune response of KOA, while AI alleviated KOA by reducing the accumulation of proteins related to energy metabolism and the TCA cycle but up-regulated ATP5IF1 to limit the energy consumption. Furthermore, our PPI analyses found that most of DAPs in cluster1 and cluster 2, which were the core subnetwork isolated from the whole PPI network, were involved in the TCA cycle and energy metabolism, which implied that DAPs related to the TCA cycle and energy metabolism play a central role in response to KOA and AI (Figures 7 and 8). Figure 6, DAPs related to cell development were totally up-regulated, such as fbronectin (FN1, accession: A0A5F9CPW4), vimentin (VIM, accession: G1SWS9), collagen (COL12A1, COL14A1, accession: A0A5F9D4Y9, A0A5F9DVP3), myosin (MYBPH, accession: G1T0G2), and dihydropyrimidinase like 3 (DPYSL3, accession: A0A5F9CLG7). A recent study reported that fbronectin signifcantly promoted cartilage repair by enhancing the proliferation, migration, and chondrogenic diferentiation of chondrogenic progenitor cells [26]. Te stimulation of the regenerating endogenous cells was considered a novel cartilage repair strategy during KOA treatment [26]. As FN1 was up-regulated after the operation of AI, it was suggested that AI could be an available treatment to alleviate KOA by promoting the synthesis of fbronectin to repair the damaged cartilage tissue. As a main component of chondrocyte cytoskeleton, vimentin played an important role in maintaining the stifness of chondrocytes in osteoarthritis [27]. In addition, dihydropyrimidinase-related protein 1, which is a homologous protein with dihydropyrimidinase-like 3, was proved to have participated in remodelling of the cytoskeleton in human cells [28]. Similarly, collagen is a structural component of many musculoskeletal tissues, and the inhibition of collagen synthesis would result in cartilage damage [29]. Particia et al. (2018) suggested that enhancing collagen synthesis could help the regeneration of cartilage, which could be a treatment of osteoarthritis [30]. In our proteomic data, the accumulation of collagen, vimentin, and dihydropyrimidinase-like 3 in AI vs KOA indicated that AI could alleviate KOA by maintaining the stifness of chondrocytes and promoting the regeneration of cartilage. Besides, muscle weakness is also a major clinic symptom of osteoarthritis. In rats, the application of caicalin could attenuate the dysfunction of muscles in OA rats by promoting the expression of myosin [31]. In other words, the accumulation of myosin might have a protective efect in muscles that sufered from OA. Consistent with the previous study, in the present study, the fact that DAPs related to myosin were up-regulated in AI vs. KOA implied that AI could attenuate KOA by enhancing the expression of myosin. Based on these literature studies and our proteomic results, we suggested that AI-induced FN1, VIM, COL12A1,

DAPs Related to Signaling Transduction.
Protein histidine methylation is a new-typepost-translational modifcation, which is catalyzed by protein-histidineN-methyltransferase [32]. In vitro, histidine N-methyltransferase (SETD3) was found to function in regulating the methylation at beta-actin H73, which afected the synthesis of cellular F-actin and the glycolytic phenotype [32], and actin or the actin pathway in chondrocytes was proved to control both skeletal development and associated diseases such as osteoarthritis [33]. As protein-histidineN-methyltransferase (SETD3, accession: A0A5F9DM36) was up-regulated in KOA rabbits after the operation of AI, we speculated that AI could alleviate KOA in chondrocytes by regulating methylation of actin mediated by SETD3. Cyclophilin B (CyPB), encoded by peptidyl-prolylcistrans isomerase (PPIB), was proposed to indirectly regulate the hydroxylation of collagen and impact collagen glycosylation and fbrillogenesis [34]. Collagen hydroxylation was found to be an important process in cartilage healing [35]. In this study, the abundance of peptidyl-prolylcis-trans isomerase (PPIB, accession: G1T2F2) increased in AI but did not change in KOA which implied that AI-induced PPIB might function as a key regulator in KOA healing by controlling the hydroxylation of collagen and fbrillogenesis.
Apoptosis-inducing factor mitochondrial associated 1 (AIFM1, accession: U3KP10) functions not only as a regulator of apoptosis but also as a regulator of respiratory chain biogenesis [36]. In AIF-defcient cells, the respiratory function was reduced, but it could be reestablished by CHCHD4, which is an AIF-interacting protein to control the import of various metabolites [36]. Respiratory chain is the major biological process to produce energy in mitochondria. In this study, the abundance of AIFM1 was increased in KOA but reduced in AI, which was consistent with the change of DAPs related to energy metabolism and TCA cycle, and this indicated that AIFM1 was an important regulator in mitochondria to control the energy production, such as the TCA cycle.

Conclusion
Te mechanism of acupotomy intervention (AI) on the treatment of knee osteoarthritis (KOA) was investigated in this study. Te results of morphological observations indicated that AI alleviated KOA by altering the morphology of the knee joint and the arrangement of chondrocytes. With proteomic analyses, 68 diferently accumulated proteins (DAPs) were identifed in AI vs. KOA and divided into 9 groups. With proteinprotein interaction analyses, DAPs related to energy metabolism and the TCA cycle were suggested to play a central role in response to AI. Moreover, AIFM1 was proposed to be an important regulator in controlling the energy production in mitochondria. In addition, DAPs involved in cell development, such as FN1, VIM, COL12A1, COL14A1, MYBPH, and DPYSL3, might function as AI-responsive proteins, which were important for the treatment of KOA.

Data Availability
Te data presented in this study are available in the article.

Ethical Approval
Animal experiments have been approved by the Committee on the Ethics of Animal Experiments of the Fujian University of Traditional Chinese Medicine. All the treatments Evidence-Based Complementary and Alternative Medicine on animals strictly obeyed the regulations of the usage and ethics of animals, which was issued by the Ministry of Science and Technology, P. R. C. in 2006.