AdipoRon Reduces TGFβ1-Mediated Collagen Deposition In Vitro and Alleviates Knee Stiffness In Vivo

Abstract

Arthrofibrosis, which causes joint motion restrictions, is a common complication following total knee arthroplasty (TKA). Key features associated with arthrofibrosis include myofibroblast activation, knee stiffness, and excessive scar tissue formation. We previously demonstrated that adiponectin levels are suppressed within the knee tissues of patients affected by arthrofibrosis and showed that AdipoRon, an adiponectin receptor agonist, exhibited anti-fibrotic properties in human mesenchymal stem cells. In this study, the therapeutic potential of AdipoRon was evaluated on TGFβ1-mediated myofibroblast differentiation of primary human knee fibroblasts and in a mouse model of knee stiffness. Picrosirius red staining revealed that AdipoRon reduced TGFβ1-induced collagen deposition in primary knee fibroblasts derived from patients undergoing primary TKA and revision TKA for arthrofibrosis. AdipoRon also reduced mRNA and protein levels of ACTA2, a key myofibroblast marker. RNA-seq analysis corroborated the anti-myofibrogenic effects of AdipoRon. In our knee stiffness mouse model, six weeks of knee immobilization, to induce a knee contracture, in conjunction with daily vehicle (DMSO) or AdipoRon (1 mg/kg, 5 mg/kg, and 25 mg/kg) via intraperitoneal injections was well tolerated based on animal behavior and weight measurements. Biomechanical testing demonstrated that passive extension angles (PEA) of experimental knees were similar between vehicle and AdipoRon treatment groups in mice evaluated immediately following immobilization. Interestingly, relative to vehicle treated mice, 5 mg/kg AdipoRon therapy improved the PEA of the experimental knees in mice that underwent four weeks of knee remobilization following the immobilization and therapy. Together, these studies revealed that AdipoRon may be an effective therapeutic modality for arthrofibrosis.

1. INTRODUCTION

The current treatment of end-stage, degenerative knee arthritis most commonly involves total knee arthroplasty (TKA) (Losina et al., 2009). Despite modern innovations in surgical techniques, implant design, and perioperative care, roughly 5% of patients undergoing TKA develop acquired idiopathic stiffness, or arthrofibrosis (Owen, Tibbo, et al., 2021a; Tibbo et al., 2019). While the exact mechanisms involved in this disease pathogenesis remain elusive (Dagneaux et al., 2020), restriction in knee function resulting in impediments of daily activities are attributed mostly to periarticular fibrotic tissue deposition within the knee joint (Ibrahim, Nazarian, & Rodriguez, 2020; Tibbo et al., 2019). Established clinical treatments for arthrofibrosis involve physical therapy (PT), manipulation under anesthesia (MUA), surgical lysis of adhesions (LOA), as well as complex revision TKA (Cheuy et al., 2017). While these treatment options may offer partial restoration of knee function, they are also associated with additional cost, discomfort, morbidity, and a high rate of disease recurrence (Bingham et al., 2019).

The cellular and molecular features generally associated with non-musculoskeletal fibrotic conditions are also observed in knee arthrofibrosis (Dagneaux et al., 2020). The key hallmarks of arthrofibrosis include local alterations in inflammatory events, excessive and dysregulated collagen deposition, and myofibroblasts as the effector cells. A key trigger linked to arthrofibrosis is aberrant transforming growth factor β1 (TGFβ1/TGFB1), which is associated with the transition of fibroblasts into myofibroblasts, a specialized cell type characterized by enhanced α-smooth muscle actin (α-SMA/ACTA2) expression (Abdel et al., 2012; Abdul et al., 2015; Unterhauser, Bosch, Zeichen, & Weiler, 2004). Active myofibroblasts are associated with enhanced production of pro-fibrotic collagens (e.g., COL1A1 and COL3A1) that support the production of the aberrant extra-cellular matrix (ECM) that is phenotypically associated with arthrofibrosis (Dagneaux et al., 2020; Frangogiannis, 2020).

Previous studies have shown that the transcript levels of the adiponectin gene (ADIPOQ), an adipokine secreted by adipocytes, were reduced within knee tissues of patients suffering from arthrofibrosis (Bayram, Limberg, et al., 2020; Bayram, Owen, et al., 2020). Of interest, alterations in adiponectin have been observed in other forms of fibrosis (e.g., renal, hepatic, and skin) (Enomoto et al., 2019; Lakota et al., 2012; Zha, Wu, & Gao, 2017). Mechanistically, adiponectin binds to and activates two receptors, ADIPOR1 and ADIPOR2, which are expressed in many tissues, including musculoskeletal tissues (Cao et al., 2022; Maddineni, Metzger, Ocón, Hendricks, & Ramachandran, 2005). This process activates signaling pathways that play a role in maintaining homeostasis of many tissues by regulating systemic processes, including insulin sensitivity, inflammation, and energy metabolism (Fang & Sweeney, 2006; Yamauchi, Iwabu, Okada-Iwabu, & Kadowaki, 2014).

It has been established that adiponectin modulates the fibrotic phenotype of soft tissue fibrosis (e.g., liver, renal, and skin) (Marangoni et al., 2017; Park, Sanz-Garcia, & Nagy, 2015; Yang et al., 2013). However, it is worth noting that adiponectin was deemed anti-fibrotic in some investigations (Marangoni et al., 2017; Park et al., 2015), but pro-fibrotic in other studies (Yang et al., 2013). The idea that adiponectin is anti-fibrotic is supported by studies which establish that AdipoRon, a synthetic small-molecule agonist of ADIPOR1 and ADIPOR2 (i.e., an adiponectin mimetic), suppressed the fibrotic process in animal models of soft tissue fibrosis (Li, Song, Ruan, Xue, & Zhao, 2021; Sha et al., 2020; Yamashita et al., 2018; Zhang et al., 2018). Recently, our group showed that AdipoRon exhibited anti-fibrotic properties when adipose-derived mesenchymal stem/stromal cells (AMSCs) were differentiated into myofibroblasts by the addition of TGFβ1 (Bayram, Owen, et al., 2020).

In this study, we aimed to assess whether AdipoRon impairs TGFβ1-mediated myofibroblast differentiation of human knee fibroblasts derived from patients undergoing primary total knee arthroplasty (PTKA) and patients undergoing revision surgery for arthrofibrosis (RTKA-A). Additionally, we aimed to determine whether daily AdipoRon administration can improve knee extension in a surgically induced mouse model of knee stiffness (Dagneaux et al., 2023). Our studies revealed that AdipoRon impeded TGFβ1-mediated myofibroblast differentiation of PTKA and RTKA-A fibroblasts and that certain doses of AdipoRon increased the passive extension angles (PEA) of knees in mice. To our knowledge, these studies are the first to reveal the anti-fibrotic effects of AdipoRon in primary cell and animal models that are relevant to knee arthrofibrosis.

2. METHODS

2.1. Isolation and culture of primary knee fibroblasts

Following patients’ consent via an approved Institutional Review Board (IRB) protocol, knee tissues from the suprapatellar pouch (SP) were collected from patients undergoing PTKA (i.e., P1–P3) while fibrotic tissue was collected from patients undergoing RTKA-A (i.e., A1–A3) (Suppl. Table 1). At the time of surgery, patients undergoing PTKA exhibited normal knee range of motion, and to date, none of these patients have developed arthrofibrosis. Tissues were excised by the operative surgeons, and the samples were immediately delivered to the laboratory.

Knee tissue samples were then divided into smaller pieces and subsequently digested with 2 mg/mL type I collagenase (Worthington Biochemical Corporation, Lakewood, NJ, USA) in advanced minimum essential medium (Advanced MEM, Thermo Fisher Scientific, Waltham, MA, USA) for 3 hours at 37°C in an orbital shaker (~120 rotations per minute). The cell-containing digestion solution was filtered through 70-μm cell strainers, centrifuged at 1200 rpm for 7 minutes at 4°C, and the collected cells were cultured in Advanced MEM supplemented with 5% human PLTMax (Mill Creek Life Sciences, Rochester, MN, USA), a clinical grade commercial platelet lysate product, 2 U/mL heparin (hospital pharmacy), GlutaMAX^™ (Thermo Fisher Scientific), and antibiotic/antimycotic (Thermo Fisher Scientific) at 37°C, 95% humidity, 21% O2, and 5% CO₂ to establish PTKA and RTKA-A fibroblasts. At approximately 100% cell confluency, passage 0 cells were detached using TrypLE Express (Thermo Fisher Scientific) and banked using CryoStor^R (Sigma-Aldrich, St. Louis, MO, USA). For experimental purposes, cells were retrieved from frozen stocks, expanded in growth medium (same as above), and passage 3–5 cells were used for all experiments. Splitting of cells was done by trypsinization at 80%–100% confluence, and each trypsinization step resulted in an increased passage number.

2.2. Treatment of Cells with AdipoRon

For experiments, cells were plated in 6-well plates at a density of 10,000 cells/cm². At confluence (Day 0), cells were treated with 25 μM AdipoRon (Tocris Bioscience, Bristol, United Kingdom) or its vehicle (DMSO) in the presence of 10 ng/mL TGFβ1 (R&D Systems, Minneapolis, MN, USA) or its vehicle (4 mM HCl and 0.1% Bovine Serum Albumin in Phosphate-Buffered Saline) in growth medium supplemented with 50 μg/mL ascorbic acid (Sigma-Aldrich). Alternatively, at confluence (Day 0), cells were treated with 5 μM AdipoRon or its vehicle (DMSO) in the presence of 10 ng/mL TGFβ1 or its vehicle in growth medium supplemented with 50 μg/mL ascorbic acid. Additional 5 μM final concentration of AdipoRon and equal volume of its vehicle were added on Day 1 and Day 2 without media change. On Day 3, cells were processed and assessed for metabolic activity by MTS activity assay, collagen deposition by picrosirius red staining, RNA expression by RT-qPCR analysis, and protein levels by western blotting.

2.3. MTS Activity Assay

[3-(4,5-dimethylthiazol-2-yl)-5-(3-carboxymethoxyphenyl)-2-(4-sulfophenyl)-2H-tetrazolium] (MTS) activity assay (Promega, Madison, WI, USA) was performed according to the manufacturer’s protocol to assess for cytotoxic effects following single dose (25 μM) or daily AdipoRon treatment. MTS activity was quantified at 490 nm using a SpectraMAX Plus spectrophotometer (Molecular Devices, San Jose, CA), background intensity subtracted from each value, and MTS activity was normalized relative to vehicle treatment.

2.4. Picrosirius Red Staining

Following the MTS activity assay (same tissue culture plates), collagen deposition in cell cultures was quantified as previously described (Thaler et al., 2022). Briefly, cell membranes were dissolved in 0.5% sodium deoxycholate in PBS for 20 min at 4°C, the remaining ECM was washed with PBS, and collagen staining was performed by picrosirius red staining kit (Polysciences Inc., Warrington, PA, USA). Fluorescence quantification of ECM was performed on a multi-plate reader and accompanying Magellan version 7.2 software at 535 nm/633 nm (Tecan, Männedorf, Switzerland). The ECM collagen structure was visualized using a Zeiss LSM 780 laser scanning microscope (561 nm/628 nm) along with Zen blue software version 2.3 (Zeiss, Jenna, Germany). Picrosirius red staining intensity was normalized relative to the vehicle.

2.5. RNA Isolation and Gene Expression Analysis

Gene expression analysis by quantitative reverse transcription polymerase chain reaction (qRT-PCR) was performed as previously described (Dudakovic et al., 2023; Paradise et al., 2022). Briefly, cell lysis and RNA isolation were completed with TRI-Reagent (Zymo Research, Irvine, CA, USA) and Direct-zol RNA isolation kit (Zymo Research), respectively. RNA quality and quantity was assessed using a NanoDrop 2000 spectrophotometer (Thermo Fischer Scientific, Waltham, MA, USA) followed by cDNA reverse transcription with Promega Reverse Transcription kit (Promega). Gene expression was assessed using QuantiTect SYBR Green PCR Kit (Qiagen, Hilden, Germany) and the CFX384 Real-Time System (BioRad, Hercules, CA, USA). Gene transcript levels were quantified using the 2^−ΔΔCt method followed by housekeeping gene normalization (GAPDH). These primer pairs were used: GAPDH - ATGTTCGTCATGGGTGTGAA (forward), TGTGGTCATGAGTCCTTCCA (reverse); ACTA2 AAAAGACAGCTACGTGGGTGA (forward), GCCATGTTCTATCGGGTACTTC (reverse). Following GAPDH normalization, ACTA2 expression was normalized to the vehicle samples.

2.6. Next-generation mRNA-sequencing (RNA-seq)

RNAs from PTKA (P1–3) and RTKA-A (A1–3) treated with vehicle, TGFβ1, and TGFβ1 + 5 μM AdipoRon (three 5 μM administration on Days 0–2) were evaluated by RNA-seq analysis as previously described (Bayram et al., 2022b; Dudakovic et al., 2020; Thaler et al., 2022). RNA from three biological replicates was pooled for each of the three treatment groups for each primary cell line. In total, eighteen samples underwent RNA-seq [three treatment arms (vehicle, TGFβ1, and TGFβ1 + AdipoRon) in six primary cells lines (3 PTKA and 3 RTKA-A)]. RNA sequencing and subsequent bioinformatics analysis were completed in collaboration with the Mayo Clinic RNA sequencing and bioinformatics cores. LibTruSeq Kits were used for indexing to permit multiplex sample loading on the flow cells. Paired-end sequencing reads were generated on the Illumina NextSeqP2 sequencer. Quality control for concentration and library size distribution was performed using an Agilent Bioanalyzer DNA 1000 chip and Qubit fluorometry (Invitrogen, Carlsbad, CA). Sequence alignment of reads and determination of normalized gene counts were performed using the MAP-RSEq. (v.1.2.1) workflow, utilizing TopHat (Kalari et al., 2014) and HTSEq (Anders, Pyl, & Huber, 2015; Putri, Anders, Pyl, Pimanda, & Zanini, 2022). Normalized read counts were expressed as fragments per kilobase of transcript per million mapped reads (FPKM). Tertiary data analysis was done using the online Gene Set Enrichment Analysis platform (GSEA) and DAVID gene ontology analysis. Heat mapping and hierarchical clustering was performed using the online platform Morpheus. PCA analysis and plot representation were done using the R packages stats and ggplot2. The RNA-seq data were deposited in the Gene Expression Omnibus of the National Center for Biotechnology Information (GSE240476).

2.7. Western Blotting Analysis

Western blot analysis was performed using standard western blotting technology as previously described (Bayram, Owen, et al., 2020; Bayram et al., 2022a; Dudakovic et al., 2023; Dudakovic et al., 2015) or using Jess (Bio-Techne, Minneapolis, MN, USA), an automated western blot system, according to instructions provided by the manufacturer. Cells were lysed with radio-immunoprecipitation buffer (Thermo Fischer Scientific) supplemented with protease and phosphatase inhibitor cocktail (Thermo Fischer Scientific). Subsequently, proteins were quantified via the DC Protein Assay Kit II (Bio-Rad, Hercules, CA, USA).

For standard western blotting analysis, the following antibodies were used: pSMAD2 (Ser465/467, 1:1,000, #3108, Cell Signaling), SMAD2 (1:1,000, #5339, Cell Signaling), pAMPKα (Thr172, 1:1,000, #50081, Cell Signaling), AMPKα (1:1,000, #5831, Cell Signaling), GAPDH (1:30,000, ab181602, AbCam), and anti-rabbit secondary (1:10,000, ab205718, AbCam). Following the development of phosphorylated antibodies (i.e., pSMAD2 and pAMPKα), blots were stripped using Restore Western Blot (Thermo Fisher Scientific) and re-probed with antibodies targeting the respective total protein (i.e., SMAD2 and AMPKα).

For Jess western blotting, the following antibodies were used and diluted in the provided antibody diluent solution, as necessary: Mouse anti-α-SMA (1:50, ab7817, Abcam, Cambridge, MA), rabbit anti-GAPDH (1:50, ab181602, Abcam), anti-mouse secondary antibody (undiluted, 042–205, ProteinSimple, Santa Clara, CA), and anti-rabbit secondary (undiluted, 042–206, ProteinSimple). The lysates, antibodies, and designated chemiluminescent reagents were dispensed into the assay plate. The microplate was then transferred to Jess where a 25-capillary chemiluminescent cartridge (12–230 kDa) (SM-FL004, ProteinSimple) drew up the aforementioned solutions. Compass for Simple Western software was used to analyze the results. Protein expression levels for ACTA2 were normalized to GAPDH.

2.8. Animal Welfare

Animal studies were performed according to recommendations provided by the National Institutes of Health and the Institute of Laboratory Animal Resources, National Research Council (National Research Council Committee for the Update of the Guide for the & Use of Laboratory, 2011). These mouse studies were approved by the Institutional Animal Care and Use Committee. Mice were housed in an accredited facility under a 12-hour light/dark cycle and provided water and food ad libitum. Mice were fed either a normal (PicoLab Rodent Diet 20, LabDiet) or a doxycycline-enriched diet (Mod LabDiet^® 5053, LabDiet). Animal experiments were approved and monitored by the veterinary staff at our institution, and all experiments adhered to the Animals in Research: Reporting In Vivo Experiments (ARRIVE) guidelines (Kilkenny, Browne, Cuthill, Emerson, & Altman, 2011).

2.9. Knee Immobilization and AdipoRon Treatment

We recently developed a mouse model of robust knee contracture that reproduces key arthrofibrotic phenotypes (i.e., knee stiffness) (Dagneaux et al., 2023). Our model involves a percutaneous technique to immobilize the experimental knee in 150° of flexion with a non-absorbable suture for a period of time (typically 4 to 8 weeks). Thereafter, a secondary procedure is performed on the experimental limbs to remove the non-absorbable suture prior to the remobilization period (typically 2 to 4 weeks). Following animal sacrifice and limb removal, the level of contracture is phenotyped through biomechanical and/or histologic methods. Alternatively, mice can be sacrificed immediately after the immobilization period (i.e., no remobilization period), suture removed postmortem, and subjected to phenotyping. As described in our original work (Dagneaux et al., 2023), both surgical interventions are performed under general anaesthesia using isoflurane. Cefazolin (30 mg/kg) and buprenorphine ER-LAB (1.0 mg/kg) are administered subcutaneously as antibiotic prophylaxis and preoperative analgesia, respectively. The animals were prepared for surgery with a combination of povidone-iodine and isopropyl alcohol solutions and draped in sterile fashion.

For these studies, right knees of 12-week-old C57BL/6J female mice (n = 64; The Jackson Laboratory, Ben Harbor, ME) were immobilized for six weeks. Mice were randomly divided into four treatment groups (n = 16 per group): vehicle (DMSO), 1 mg/kg AdipoRon, 5 mg/kg AdipoRon, and 25 mg/kg AdipoRon. Each treatment group (n = 16) was then further divided into a no remobilization cohort (i.e., 6 weeks of immobilization and 0 weeks of remobilization, n = 8 per group) and a four week remobilization group (i.e., 6 weeks of immobilization and 4 weeks of remobilization, remobilization cohort, n = 8 per group). As highlighted in the original mouse knee contracture model (Dagneaux et al., 2023), a power analysis revealed that six mice were required in each group to detect an effect difference of 10° with a standard deviation (SD) of 8°. Due to potential complications from daily vehicle and AdipoRon injections, two additional mice were included in each group. The breakdown of the groups is summarized in Figure 6.

Figure 6. Experimental outline of AdipoRon and mouse knee stiffness model.

At 12 weeks of age, female mice underwent an immobilization procedure. Starting on the day immobilization, mice were administered vehicle (DMSO) or one of several doses of AdipoRon (1 mg/kg, 5 mg/kg, and 25 mg/kg) for 6 weeks (weeks 12–18) via daily IP injections. After the 6 week immobilization period, a set of mice were sacrificed (n = 8 per treatment group), immobilization suture was removed from the experimental limb, and the experimental and contralateral limbs were assessed biomechanically (n = 8 per treatment group). This first set of mice is referred to as 6 weeks of immobilization and 0 weeks of remobilization group (6 wks Im + 0 wks Rem group; no remobilization group). A second set of mice underwent a remobilization procedure following the immobilization period. Following a remobilization period of 4 weeks, this second group of mice was sacrificed, and the experimental and contralateral limbs were assessed biomechanically (n = 8 per treatment group). This second set of mice is referred to as 6 weeks of immobilization and 4 weeks of remobilization (6 wks Im + 4 wks Rem group; remobilization group). 12 weeks, 18 weeks, and 22 weeks refers to the mouse age. In depth experimental details are provided in the Methods Section.

Starting on the day of immobilization, AdipoRon and its vehicle (DMSO) were administered during the entire immobilization period (i.e., 6 weeks) via daily intraperitoneal (IP) injections. Injections were not administered for two consecutive days during the treatment period (during week four). AdipoRon dosing (1 mg/kg, 5 mg/kg, and 25 mg/kg) and delivery method were selected based on existing literature (Lee et al., 2021; Nicolas et al., 2018; Nicolas, Rochet, Gautier, Chabry, & Pisani, 2020; Sun et al., 2023; Takenaga, Akimoto, Koshikawa, & Nagase, 2021; Tarkhnishvili et al., 2022; Zheng et al., 2019). AdipoRon and DMSO were combined with 20% Captisol (pH = 4 – 4.5) just prior to the IP injections as previously described (Dudakovic et al., 2016; Dudakovic et al., 2020). Animal weights were measured on a weekly basis and AdipoRon dosing was adjusted, if needed.

Following the immobilization period, the 6-week immobilization and 0-week remobilization groups (i.e., no remobilization group) were sacrificed and knee extension was assessed biomechanically (see below). On the other hand, following the immobilization period, the 6-week immobilization and 4-week remobilization groups (i.e., remobilization group) underwent the suture removal procedure followed by 4 additional weeks of cage activity before sacrifice and biomechanical testing.

2.10. Biomechanical Testing

Biomechanical testing of experimental and contralateral knees was performed as previously described (Dagneaux et al., 2023). A similar approach has also been used to evaluate the arthrofibrotic phenotype in rats and rabbits (Nesterenko et al., 2009; Owen et al., 2022; Trousdale et al., 2022). Experimental and contralateral limbs were disarticulated at the hip and midfoot joints (immobilization suture removed for the no remobilization group), skin and subcutaneous tissues removed, and muscular tissues detached from the proximal third of the femur and tibia. The proximal end of the femur was potted in polymethylmethacrylate (PMMA) bone cement (Stryker, Kalamazoo, MI, USA) and mounted to a bracket on the measuring device. A passive extension motion (from 135° of knee flexion) was applied at a speed of 1°/second. The torque applied to the knee joint was measured by a dedicated torque sensor (Transducer Techniques, USA). The resulting curves were transformed into graphs of passive extension versus torque using Matlab 2016a (MathWorks, Natick, MA, USA). The passive extension angle (PEA) at a specific torque (i.e., 0.4 N-cm) was evaluated for right (experimental) and left (contralateral) knees. Two mice were sacrificed following surgical complications (one in the 5 mg/kg no remobilization group and one in the 25 mg/kg no remobilization group). Due to technical challenges, biomechanical data were not collected on one contralateral knee in the 1 mg/kg AdipoRon no remobilization group and one experimental knee in the 5 mg/kg AdipoRon no remobilization group. In all, biomechanical data was collected from 122 out of the possible 128 experimental/contralateral limbs.

2.11. Statistical Analysis

Graphical data summary and statistical analyses were performed using GraphPad Prism version 9.3.1 (GraphPad Software, San Diego, CA, USA). For statistical analysis, all other groups were compared to TGFβ1 group for in vitro studies and vehicle group for in vivo studies using one-way ANOVA. When only two groups are shown, a student’s t-test was employed. A student’s t-test was also used to compare the two contralateral groups and the two experimental group shown in Figure 9. Where applicable, significance is noted in the figures with a standard asterisk convention (* = p ≤ 0.05, ** = p ≤ 0.01, *** = p ≤ 0.001, **** = p ≤ 0.0001).

Figure 9. AdipoRon reduces knee stiffness following knee remobilization in mouse knee stiffness model.

Passive extension angles (PEA) were evaluated at a specific torque (0.4 N-cm) by biomechanical testing in contralateral (left) and experiment (right) knees in vehicle and AdipoRon treatment groups. Contralateral (left knee) (A) and experimental (right knee) (B) PEA for 6 Weeks Immobilization + 0 Weeks Remobilization vehicle and AdipoRon groups (n = 7 to 8, mean ± SD). Contralateral (left knee) (C) and experimental (right knee) (D) PEA for 6 Weeks Immobilization + 4 Weeks Remobilization vehicle and AdipoRon groups (n = 8, mean ± SD). Significance is noted with a standard asterisk convention between relevant groups (* = p ≤ 0.05, ** = p ≤ 0.01, *** = p ≤ 0.001, **** = p ≤ 0.0001).

3. RESULTS

3.1. AdipoRon reduced TGFβ1-mediated collagen deposition in primary knee fibroblasts

To evaluate the impact of AdipoRon on TGFβ1-mediated myofibroblast differentiation, we first performed a concentration response curve in fibroblasts isolated from patients undergoing PTKA (Suppl. Figure 1). MTS activity assay was used as a proxy for cell viability while picrosirius red staining was employed to evaluate collagen deposition. Repeated administration (Days 0–2) of various AdipoRon concentrations resulted in viable cells (Suppl. Figure 1A) but repeated 5 μM of AdipoRon reduced TGFβ1-induced collagen deposition nearly to the level of unstimulated (i.e., no TGFβ1) cells (Suppl. Figure 1B). Higher concentrations of AdipoRon; however, added once (Day 0) reduced cell viability (Suppl. Figure 1C) and TGFβ1-mediated collagen deposition in a dose-dependent manner. (Suppl. Figure 1B). Notably, the reduction in collagen deposition was more robust when compared to reduction in cell viability at each AdipoRon concentration.

Based on the above observations, AdipoRon-related studies were expanded to three primary PTKA (P1–3) and three primary RTKA-A (A1–3) fibroblast cell lines undergoing TGFβ1-induced myofibroblast differentiation using repeated (3X) 5 μM AdipoRon and single (1X) 25 μM AdipoRon treatment strategies. As anticipated, TGFβ1 stimulated collagen deposition when compared to undifferentiated cell cultures (Figures 1&2). Importantly, both repeated administration of 5 μM AdipoRon (Figure 1) as well as a single bolus administration of 25 μM AdipoRon (Figure 2), impaired TGFβ1-induced collagen deposition. Only a modest reduction in cell viability was observed in the AdipoRon plus TGFβ1 groups when compared to relevant TGFβ1 (more so in the single 25 μM AdipoRon treatment arm) (Suppl. Figure 2), suggesting that the major effect of AdipoRon on collagen deposition was attributed to the inhibition of the fibroblast to myofibroblast differentiation process rather than cell toxicity.

Figure 1. Repeated AdipoRon administration reduces TGFβ1-mediated ECM deposition in primary knee fibroblasts.

Primary knee fibroblasts derived from three patients undergoing primary TKA [PTKA (P1–3)] and three patients undergoing revision TKA for arthrofibrosis [RTKA-A (A1–3)] were undifferentiated or underwent TGFβ1-mediated (10 ng/ml) myofibroblast differentiation in the presence of vehicle (DMSO) or AdipoRon (5 μM) added on three consecutive days (i.e., Days 0–2). Picrosirius red staining to visualize/quantify ECM deposition was performed on Day 3. Picrosirius red staining quantification in PTKA (A) and RTKA-A (B) cells (n = 3, mean ± SD). Representative whole-well picrosirius red staining images (top) and LSM images (bottom) in PTKA (C) and RTKA-A (D) cells. Significance is noted with a standard asterisk convention between relevant groups (* = p ≤ 0.05, ** = p ≤ 0.01, *** = p ≤ 0.001, **** = p ≤ 0.0001).

Figure 2. Bolus AdipoRon administration reduces TGFβ1-mediated ECM deposition in primary knee fibroblasts.

Primary knee fibroblasts derived from three patients undergoing primary TKA [PTKA (P1–3)] and three patients undergoing revision TKA for arthrofibrosis [RTKA-A (A1–3)] were undifferentiated or underwent TGFβ1-mediated (10 ng/ml) myofibroblast differentiation in the presence of vehicle (DMSO) or AdipoRon (25 μM) added on Day 0. Picrosirius red staining to visualize/quantify ECM deposition was performed on Day 3. Picrosirius red staining quantification in PTKA (A) and RTKA-A (B) cells (n = 3, mean ± SD). Example whole-well picrosirius red staining images (top) and LSM images (bottom) in PTKA (C) and RTKA-A (D) cells. Significance is noted with a standard asterisk convention between relevant groups (* = p ≤ 0.05, ** = p ≤ 0.01, *** = p ≤ 0.001, **** = p ≤ 0.0001).

3.2. AdipoRon reduced TGFβ1-mediated ACTA2 expression in primary knee fibroblasts

ACTA2 is a key marker of myofibroblast differentiation. As such, ACTA2 mRNA and protein levels were evaluated in the presence of repeated (3 × 5 μM) and single bolus (1 × 25 μM) AdipoRon treatment upon TGFβ1 induction of fibroblasts into myofibroblasts. As predicted, ACTA2 mRNA and protein levels were induced by TGFβ1 treatment in all three PTKA and all three RTKA-A primary cell lines (Figures 3&4). In general, a robust reduction in TGFβ1-mediated ACTA2 mRNA levels was observed with a single administration of 25 μM AdipoRon (P1 cell line being an exception) (Figure 3 A&B). These trends were also observed with repeated 5 μM AdipoRon treatment, albeit to a lesser extent and significance was only reached in three cell lines (Figure 3 C&D). Similar to mRNA levels, western blotting analysis revealed that the single 25 μM AdipoRon treatment abolished TGFβ1-meadiated ACTA2 protein induction in all PTKA and RTKA-A cells (Figure 4 A&B). Interestingly, the reduction of TGFβ1-induced ACTA2 protein levels was also observed with three administrations of 5 μM AdipoRon across all six cell lines, with significance in four (Figure 4C&D). Together, these data demonstrated that AdipoRon reduces mRNA and protein levels of ACTA2, a key marker that is induced during TGFβ1-mediated differentiation of fibroblasts into myofibroblasts.

Figure 3. AdipoRon reduces TGFβ1-induced ACTA2 mRNA expression.

Primary knee fibroblasts derived from three patients undergoing primary TKA [PTKA (P1–3)] and three patients undergoing revision TKA for arthrofibrosis [RTKA-A (A1–3)] were undifferentiated or underwent TGFβ1-mediated (10 ng/ml) myofibroblast differentiation. Cells were treated either one time (on Day 0) with vehicle (DMSO) or AdipoRon (25 μM) (A and B) or three times (on Days 0–2) with vehicle (DMSO) or AdipoRon (5 μM) (C and D). ACTA2 mRNA expression was quantified relative to GAPDH mRNA levels (A-D) (n = 3, mean ± SD). Significance is noted with a standard asterisk convention between relevant groups (* = p ≤ 0.05, ** = p ≤ 0.01, *** = p ≤ 0.001, **** = p ≤ 0.0001).

Figure 4. AdipoRon reduces TGFβ1-mediated induction of ACTA2 protein.

Primary knee fibroblasts derived from three patients undergoing primary TKA [PTKA (P1–3)] and three patients undergoing revision TKA for arthrofibrosis [RTKA-A (A1–3)] were undifferentiated or underwent TGFβ1-mediated (10 ng/ml) myofibroblast differentiation. Cells were treated either one time (on Day 0) with vehicle (DMSO) or AdipoRon (25 μM) (A and B) or three times (on Days 0–2) with vehicle (DMSO) or AdipoRon (5 μM) (C and D). ACTA2 protein levels were quantified relative to GAPDH protein levels (A-D) (n = 3, mean ± SD). Significance is noted with a standard asterisk convention between relevant groups (* = p ≤ 0.05, ** = p ≤ 0.01, *** = p ≤ 0.001, **** = p ≤ 0.0001).

3.3. Activation of adiponectin axis by AdipoRon did not impact TGFβ1-mediated SMAD2 phosphorylation

Having established the anti-myofibrogenic effects of AdipoRon, the impact of AdipoRon on the activation of adiponectin axis (i.e., AMPKα phosphorylation) and TGFβ1 signaling (i.e., SMAD2 phosphorylation) was evaluated in PTKA and RTKA-A primary cells. One hour treatment with AdipoRon (25 μM) induced robust AMPKα phosphorylation in PTKA (Suppl. Figure 3A) and RTKA-A (Suppl. Figure 3B) cells, which was not significantly altered by simultaneous TGFβ1 administration. As anticipated, administration of TGFβ1 stimulated SMAD2 phosphorylation (maximal induction occurred one hour after treatment) in a time-dependent manner in RTKA-A cells (Suppl. Figure 4A), which was not altered by the presence of AdipoRon. Similarly, quantitative analysis of western blots revealed similar SMAD2 phosphorylation levels with TGFβ1 alone or TGFβ1 plus AdipoRon treatments in RTKA-A (Suppl. Figure 4B) and PTKA (Suppl. Figure 4C) cells treated for one hour. Similar observations were made when RTKA-A and PTKA cells were pre-treated three hours with AdipoRon prior to TGFβ1 administration (Suppl. Figure 4D–F). In these sets of experiments, maximal induction of SMAD2 phosphorylation was achieved ten minutes following TGFβ1 administration, which is most likely attributed to the fresh media addition at the time of AdipoRon addition that occurred three hours prior to TGFβ1 administration. In sum, these data demonstrate that AdipoRon caused an increase in AMPKα phosphorylation, but did not alter TGFβ1-induced SMAD2 phosphorylation in primary knee fibroblasts.

3.4. Total transcriptomic analysis underlined the anti-fibrotic effect of AdipoRon in vitro

To better elucidate the anti-fibrotic effects of AdipoRon on the transcriptome, we performed RNA-seq analysis on three PTKA and three RTKA-A cells each treated with vehicle, TGFβ1, and TGFβ1 + AdipoRon (three 5μM additions on Days 0–2). Principle component analysis (PCA) revealed that the treatment groups (TGFβ1 and TGFβ1 + AdipoRon) clearly separated from control samples (Veh). Of note, PC1 explained over 80% of the variation in the dataset, indicating TGFβ1 as the main driver for the observed results (Figure 5A). As AdipoRon counteracts the pro-myofibrogenic effects of TGFβ1, next we assessed which genes would be neutralized by AdipoRon upon upregulation by TGFβ1 (comparison 1) and which genes would be rescued by AdipoRon upon down-regulation by TGFβ1 (comparison 2). These analyses generated a gene set of 280 genes for the first comparison and a gene group of 226 genes for the second comparison. Importantly, for both comparisons PTKA and RTKA-A cells clearly separated between TGFβ1 and TGFβ1 plus AdipoRon treatments (Figure 5B). To obtain a functional assessment of the generated gene sets, we performed gene set enrichment analysis (GSEA) as well as gene ontology (GO) analysis using DAVID. The obtained results highlight the involvement of the genes in the 280 gene-set in TGFβ1 related pathways as well as ECM related pathways and functions (Suppl. Figure 5A). When the genes in the 226 gene set were analyzed, we found that many of the genes were related to general immune response and to the extra cellular space (Suppl. Figure 5B). However, the significance values for these associations were clearly less pronounced when compared with the 280 gene set (compare Suppl. Figure 5A&B). The top 30 genes associated with the found functional clusters in GSEA and GO are shown (Figure 5C). Altogether, these data reiterated the anti-fibrotic effects of AdipoRon on the molecular level in vitro.

Figure 5. Global transcriptional analysis by RNA-seq reiterates experimental observations at the molecular level.

Primary knee fibroblasts derived from three patients undergoing primary TKA [PTKA (P1–3)] and three patients undergoing revision TKA for arthrofibrosis [RTKA-A (A1–3)] were undifferentiated or underwent TGFβ1-mediated (10 ng/ml) myofibroblast differentiation in the presence of vehicle (DMSO) or AdipoRon (5 μM) added on three consecutive days (Days 0–2). Following RNA isolation, RNA-seq analysis was performed. PCA analysis for knee fibroblasts from PTKA (P) and RTKA-A (A) (A). Hierarchical clustering for top genes up-regulated by TGFβ1 and neutralized by combined treatment of TGFβ1 and AdipoRon (left) and top genes down-regulated by TGFβ1 and rescued by combined treatment of TGFβ1 and AdipoRon (right). For both groups, genes were selected when one or more comparison were p<0.05 (B). Heatmap profiles for top 30 genes linked to GSEA or DAVID-GO gene clusters (C). P = PTKA cells, A = RTKA-A cells, AR = AdipoRon, EC = extracellular, ECM = extracellular matrix.

3.5. AdipoRon treatments were well-tolerated by mice that exhibit knee stiffness

To evaluate the therapeutic potential of AdipoRon on stiffness in vivo, we employed our recently developed knee stiffness mouse model (Dagneaux et al., 2023) in conjunction with AdipoRon therapy at previously established dosing regimens (Lee et al., 2021; Nicolas et al., 2018; Nicolas et al., 2020; Sun et al., 2023; Takenaga et al., 2021; Tarkhnishvili et al., 2022; Zheng et al., 2019). The overall experimental design is outlined in Figure 6.

To assess for potential toxicity of various doses of AdipoRon, weekly weight measurements were obtained during the knee immobilization/treatment time frame (12 to 18 weeks of age) (Suppl. Figure 6). No apparent changes in weekly body weight (Suppl. Figure 6A) and weekly weight change relative to beginning weight (Suppl. Figure 6B) were observed when comparing vehicle to the three AdipoRon doses. Similarly, when comparing the terminal weights of no remobilization cohorts (Figure 7A) and remobilization cohorts (Figure 7B), no significant changes in weight were noted between vehicle and three AdipoRon doses. Together, these data demonstrate that AdipoRon treatment was safe in our knee stiffness mouse model.

Figure 7. Mouse weights at time of sacrifice.

Mouse weights at time of sacrifice for 6 weeks of immobilization and 0 weeks of remobilization group (A) and 6 weeks of immobilization and 4 weeks of remobilization (B) (n = 7 to 8 for each group, mean ± SD).

3.6. AdipoRon reduced knee stiffness following knee remobilization

At sacrifice, biomechanical data were collected on the limbs using a dynamic load cell device that captures the passive extension angle (PEA) at a torque of 0.4 N-cm (Dagneaux et al., 2023). To confirm knee stiffness in our mouse model, the PEAs of experimental (right) and contralateral (left) knees of vehicle-injected groups were compared in the no remobilization cohort (6 weeks immobilization + 0 weeks remobilization) and the remobilization cohort (6 weeks immobilization + 4 weeks remobilization) (Figure 8). A significant reduction in PEA was observed in the experimental knees when compared to contralateral knees in the no remobilization cohort (Figure 8A) and the remobilization cohort (Figure 8B). When comparing the means, a difference of 59° (163° vs 104°) and 90° (161° vs 71°) in PEAs is observed between contralateral and experimental knees for the no remobilization cohort and the remobilization cohort, respectively. In these experiments, this difference between the contralateral and experimental knees is considered a ‘therapeutic window’ as an anti-arthrofibrotic intervention would be anticipated to reduce this gap in PEA. Together, these analyses confirmed knee stiffness in our mouse model of arthrofibrosis.

Figure 8. Confirmation of knee stiffness in mouse knee stiffness model.

Passive extension angles (PEA) were evaluated at a specific torque (0.4 N-cm) by biomechanical testing in contralateral (left) and experiment (right) knees following sacrifice for 6 Weeks Immobilization + 0 Weeks Remobilization vehicle group (A) (n = 8, mean ± SD) and 6 Weeks Immobilization + 6 Weeks Remobilization vehicle group (B) (n = 8, mean ± SD). Models on the right summarize the numerical data and depict the therapeutic window for each experimental set-up (i.e., 59° for 6 Weeks Immobilization + 0 Weeks Remobilization and 90° for 6 Weeks Immobilization + 6 Weeks Remobilization).

We next compared contralateral and experimental knee PEAs of vehicle and AdipoRon treated mice in the no remobilization cohort and the remobilization cohort (Figure 9). The PEAs of contralateral (Figure 9A) and experimental (Figure 9B) knees were similar between vehicle and three AdipoRon treatment groups in the no remobilization cohort. While the PEAs of the contralateral knees were similar between the treatment groups (Figure 9C), a significant improvement (p = 0.0037) in PEA of experimental limbs was observed with 5 mg/kg AdipoRon when compared to the vehicle treatment group (38% increase in PEA) in the remobilization cohort (Figure 9D). While trending (p = 0.0608), 1 mg/kg AdipoRon (25% increase in PEA) did not reach statistical significance when compared to vehicle treatment. A higher dose of AdipoRon (25 mg/kg) did not result in greater PEA improvements relative to vehicle. In sum, these results demonstrated that AdipoRon (5 mg/kg) significantly improved knee motion following knee remobilization in a mouse model of arthrofibrosis.

Our studies revealed a 90° therapeutic window when comparing contralateral and experimental knees of vehicle treated animals in the remobilization cohort (Figure 8B). To further assess knee extension improvements by 5 mg/kg AdipoRon therapy, we compared the observed PEA changes in the contralateral and experimental limbs of vehicle and 5 mg/kg AdipoRon in the remobilization cohort (Figure 10). This analysis revealed that 5 mg/kg AdipoRon therapy restored knee extension by nearly 27°, which accounts for ~30% (27° of possible 90°) of the therapeutic window (Figure 10A&B) in the experimental knees. No significant variations were observed when comparing the change in PEA between vehicle (90°) and 5 mg/kg AdipoRon (96°) in the contralateral limbs.

Figure 10. Visualization of therapeutic effect of AdipoRon in mouse knee stiffness model.

A sub-set of data presented in Figure 9 (parts C and D) are shown to illustrate therapeutic effect of AdipoRon. Contralateral (left knee) and experimental (right knee) PEA for 6 Weeks Immobilization + 4 Weeks Remobilization vehicle and 5 mg/kg AdipoRon groups (A) (n = 8, mean ± SD). Visualization of the average PEA for vehicle treated contralateral limb (left; 161°), vehicle-treated experimental limb (right; 71°), and 5 mg/kg AdipoRon treated experimental limb (right; 97°) (B). Significance is noted with a standard asterisk convention between relevant groups (* = p ≤ 0.05, ** = p ≤ 0.01, *** = p ≤ 0.001, **** = p ≤ 0.0001).

4. DISCUSSION

Our current study demonstrates that AdipoRon treatment reduced collagen deposition and ACTA2 expression, two key markers associated with arthrofibrosis, in primary fibroblasts undergoing TGFβ1-mediated myofibroblast differentiation. Single (1 × 25 μM) and repeated (3 × 5 μM) AdipoRon dosing resulted in robust reduction of collagen deposition; however, the single dosing regimen was more potent at reducing ACTA2 mRNA and protein expression. This observation can be explained by the total load of AdipoRon (25 μM vs 15 μM) or the experimental design in which the bolus AdipoRon was added at the time of TGFβ1 addition while repeated dosing of AdipoRon occurred on the day of and two subsequent days following TGFβ1 administration. These findings are supported by our previous studies which showed that the expression of adiponectin is reduced in human arthrofibrotic tissues (Bayram, Limberg, et al., 2020) and TGFβ1-induced myofibroblast differentiation of AMSCs is impeded with AdipoRon treatment (Bayram, Owen, et al., 2020). In support of the general idea that activation of the adiponectin axis may prevent the fibrotic process, other studies have established a regulatory role for adiponectin (Marangoni et al., 2017; Park et al., 2015; Yang et al., 2013) and therapeutic potential of AdipoRon (Li et al., 2021; Sha et al., 2020; Yamashita et al., 2018; Zhang et al., 2018) in soft tissue fibrosis. As anticipated, AdipoRon activated adiponectin axis signaling (AMPKα phosphorylation) and TGFβ1 stimulated early events associated with the canonical TGFβ1 pathway (SMAD2 phosphorylation). We did not observe any alterations in TGFβ1-mediated SMAD2 phosphorylation by AdipoRon. However, RNA-Seq analysis revealed significant alterations in gene expression related to the TGFβ1 pathway and ECM deposition when AdipoRon is co-administered with TGFβ1. Together, these molecular analyses suggest that adiponectin activation via AdipoRon acts on downstream events regulated by TGFβ1. The novelty of our present in vitro studies is the evidence that AdipoRon impaired myofibroblast differentiation of three primary cell lines derived from patients undergoing PTKA and three primary cell lines isolated from patients undergoing RTKA-A. These findings are directly linked with our observation that adiponectin expression is suppressed within arthrofibrotic knee tissues (Bayram, Limberg, et al., 2020).

While our group has developed arthrofibrosis models in rabbits (Nesterenko et al., 2009), rats (Owen, Dagneaux, et al., 2021), and mice (Dagneaux et al., 2023), the present investigation evaluated the therapeutic potential of AdipoRon in mice due to established therapeutic regimens (i.e., dose, interval, and route of administration) (Lee et al., 2021; Nicolas et al., 2018; Nicolas et al., 2020; Sun et al., 2023; Takenaga et al., 2021; Tarkhnishvili et al., 2022; Zheng et al., 2019). The use of AdipoRon in rats has been reported (Gu et al., 2020; Jenke et al., 2021), but a scarcity of AdipoRon safety data in rats precluded the use of AdipoRon in our established rat model of arthrofibrosis (Owen, Dagneaux, et al., 2021). We have previously reported the safety of AdipoRon in our rabbit model of arthrofibrosis (Salmons et al., 2023). This recent study revealed that six intravenous (marginal ear vein) administrations of up to 5 mg/kg were deemed safe in our well-established surgical model of arthrofibrosis. These observations in rabbits could be harnessed in the future to evaluate the efficacy of AdipoRon in our rabbit model of arthrofibrosis (Nesterenko et al., 2009).

Based on our robust current and previous in vitro findings (Bayram, Owen, et al., 2020), we combined our mouse model of knee stiffness (Dagneaux et al., 2023) with daily IP injections of AdipoRon to evaluate whether this agent could impede knee stiffness in vivo. Due to the wide range of AdipoRon dosing used in mice (Lee et al., 2021; Nicolas et al., 2018; Nicolas et al., 2020; Sun et al., 2023; Takenaga et al., 2021; Tarkhnishvili et al., 2022; Zheng et al., 2019), we selected a broad dosing range (1 mg/kg, 5 mg/kg, and 25 mg/kg) to evaluate its efficacy in vivo. In support of previous AdipoRon studies (Lee et al., 2021; Nicolas et al., 2018; Nicolas et al., 2020; Sun et al., 2023; Takenaga et al., 2021; Tarkhnishvili et al., 2022; Zheng et al., 2019), our studies did not reveal any apparent cytotoxic effects (e.g., weight change and animal behavior) of daily IP AdipoRon (6 weeks) in mice. While previous use of AdipoRon was mostly limited to 3 weeks (Lee et al., 2021; Nicolas et al., 2018; Nicolas et al., 2020; Sun et al., 2023; Takenaga et al., 2021; Tarkhnishvili et al., 2022; Zheng et al., 2019), the present investigation revealed that AdipoRon therapy can be extended to six weeks in mice. While we did not observe weight changes between vehicle and AdipoRon treated mice, it is of interest that all mice failed to gain weight during the immobilization and treatment regimen phase of the experiment (weeks 12 to 18 of age). However, an increase in body weight was observed after drug/vehicle cessation and remobilization (weeks 18 to 22 of age) across the treatment groups. However, it remains elusive whether the increase in overall animal weight between weeks 18 and 22 is attributed to the cessation of daily injections (vehicle or AdipoRon) and/or due to the remobilization procedure as both events occurred simultaneously.

Although we did not observe knee extension improvements in the no remobilization cohort with AdipoRon therapy, our studies revealed significant improvements in PEA with 5 mg/kg AdipoRon in the remobilization cohort. Of interest, the lower dose (1 mg/kg) of AdipoRon also trended towards improved knee extension while the higher dose (25 mg/kg) of AdipoRon resembled vehicle treatment. We acknowledge a dose-dependent effect of AdipoRon between 1 mg/kg and 5 mg/kg doses in the remobilization group; however, this trend does not continue with the 25 mg/kg AdipoRon dose. It is possible that off target effects are activated by higher AdipoRon dosing, which was suggested by a previous study (Lee et al., 2021). It is noteworthy that the data spread (i.e., standard deviation) was more extensive in the experimental limbs of the no remobilization cohort when compared to the remobilization cohort, which is also supported by the findings in our studies aimed at developing a knee stiffness mouse model (Dagneaux et al., 2023). Interestingly, when comparing the vehicle treated groups only between the two cohorts, the PEA of the experimental knees is lower in remobilization cohort when compared to the no remobilization cohort, which was also observed during development of this mouse model (Dagneaux et al., 2023). Conceptually speaking, treatment of mice with AdipoRon (5 mg/kg) during the immobilization period appeared to prevent additional impairments in the PEA following the remobilization procedure.

Our studies revealed a mean improvement of 27° in the PEA in mice treated with 5 mg/kg AdipoRon, which accounts for roughly 30% of the PEA lost (90°) in the experimental limbs of vehicle treated animals. The clinical definition adopted by our group defines arthrofibrosis as an arc of motion ≤90° as measured by goniometer lasting 12+ weeks postoperatively in patients without prior limitation in knee range of motion (Tibbo et al., 2019). As mentioned earlier, MUA is employed to improve knee range of motion in patients affected by arthrofibrosis (Newman et al., 2018; Pitta et al., 2018). A recent analysis of 9,771 patients (12,735 knees) by our group revealed that MUA increased the range of motion by 34° following MUA of arthrofibrotic knees; however, this range of motion was inferior to the range of motion of knees not affected by arthrofibrosis (Owen, Tibbo, et al., 2021b). In the future, it may also be possible to combine AdipoRon therapy with MUA alone or in combination with other potential anti-arthrofibrotic therapies, including the non-steroidal anti-inflammatory drug celecoxib (Limberg et al., 2020; Salib et al., 2019; Trousdale et al., 2022; Trousdale et al., 2018). However, in our rabbit model of arthrofibrosis, celecoxib was superior to surgical capsular release (used in rabbits as a proxy for manipulation under anesthesia), and the combination of celecoxib and a surgical release did not provide additive benefits (Trousdale et al., 2022). While biomechanical testing in mouse and rabbit knees (i.e., PEA at a particular torque) is not directly comparable to the clinical range of motion in patients, the PEA improvement observed in our mouse model of arthrofibrosis in the present studies may provide evidence that activation of the adiponectin axis may be a potential therapeutic option in the prevention and treatment of knee arthrofibrosis, alone or in combination with other therapeutics and manipulation under anesthesia.

While the present investigation highlights the potential use of adiponectin axis small molecule activators in the prevention and/or treatment of knee arthrofibrosis, the present studies are not without limitations. Our in vitro and in vivo investigation revealed inhibitory functions of AdipoRon towards TGFβ1-mediated myofibroblast differentiation and knee stiffness, respectively. However, mechanistic insights that explain these effects are still lacking. Of note, while AdipoRon stimulated AMPKα phosphorylation, the addition of this adiponectin mimetic did not alter TGFβ1-induced SMAD2 phosphorylation. However, the use of genome-wide based studies (e.g., RNA-seq) confirmed the overall phenotype observed upon AdipoRon administration in TGFβ1 treated cells. Thus, additional follow up studies are needed to help define AdipoRon-mediated mechanism(s) that contribute to the anti-arthrofibrotic effects. We also acknowledge that our studies only employed female mice. However, it is important to point out that female patients are 2.5 times more likely to develop arthrofibrosis (Sanders et al., 2017; Tibbo et al., 2019). Beyond these clinical findings, we also have observed an increase in complications following knee immobilization in male mice (data not shown). We recognize that additional studies which incorporate histologic assessment or molecular characterization of knee tissues would also be beneficial. These studies were presently omitted, as biomechanical testing does not allow for tissue preservation required for additional downstream analysis. Future studies aimed at defining histologic and molecular alterations within the knee can be employed to compare vehicle and 5 mg/kg treated AdipoRon mice. As mentioned earlier, we have already established a safety profile of AdipoRon in a rabbit model of arthrofibrosis (Salmons et al., 2023). Future assessment, including histologic and molecular approaches, of AdipoRon could thus be performed in mice or rabbits. While more involved, assessment of AdipoRon in rabbits provides several advantages, including improved knee tissue access and greater tissue quantity.

In summary, our current studies showed that AdipoRon impeded TGFβ1-induced myofibroblast differentiation of primary knee fibroblasts derived from patients undergoing primary total knee arthroplasty and revision total knee arthroplasty for arthrofibrosis. Further, these studies showed that 5 mg/kg AdipoRon increased the PEA in our mouse model. Collectively, our studies suggested that AdipoRon may be a viable therapeutic option for the prevention and/or treatment of knee arthrofibrosis.

Data Availability Statement

Datasets utilized in this study are archived in the Gene Expression Omnibus of the National Institute for Biotechnology Information (GSE240476).