Polydactyly bone marrow-derived mesenchymal stem cell-conditioned medium can prevent cartilage degeneration and alleviate knee pain in a rat model of osteoarthritis

Abstract

Background

Osteoarthritis (OA) is characterized by progressive degeneration of joint cartilage and has substantially increased worldwide. Recently, Mesenchymal stem cells (MSCs) have been explored as a cell-based therapy for the treatment of OA. MSC therapy has not been safely introduced into clinical practice due to immune rejection and the need for highly controlled culture protocols. The therapeutic efficacy of MSC treatment is mediated by the paracrine effect, and conditioned medium (CM) containing MSC-derived secreted factors have potential for useful cell-free therapy. CM can be prepared from various sources, though bone marrow MSCs obtained from infant polydactyly patients may be particularly useful for chondrogenic differentiation and the regulation of cartilage formation. The purpose of this study was to evaluate the effect of polydactyly bone marrow MSC-derived conditioned medium (pBMSC-CM) on the prevention of cartilage degeneration and knee pain using a rat monoiodoacetate (MIA)-induced knee OA model.

Methods

pBMSC-CM was isolated from the bone marrow of an infant polydactyly thumb. In vitro, cell proliferation was evaluated in chondrocytes cultured with pBMSC-CM, and gene expression of cartilage matrix markers was assessed by quantitative reverse transcription PCR. RNA sequencing was conducted to analyze differentially expressed genes in pBMSCs. A knee arthritis model was generated by an intra-articular injection of MIA into the right knee of 18 Sprague-Dawley rats, which were then divided into three groups: (1) No-treatment group, (2) Culture medium only group, and (3) pBMSC-CM group. Each group received intra-articular injections of either saline, alpha-Minimal Essential Medium, or pBMSC-CM one week after the MIA injection. To assess knee joint pain, the struggle threshold of the knee joint extension angle was measured every week. Two weeks after treatment injections, a retrograde neurotracer was injected into both knees. After one week, articular cartilage and synovitis were evaluated based on histological characteristics. The dorsal root ganglions (DRG) were immunostained to evaluate the expression of calcitonin gene-related peptide (CGRP).

Results

pBMSC-CM enhanced chondrocyte proliferation and upregulated cartilage matrix genes. Chondrogenic and neuroprotective genes were highly expressed in pBMSCs. In behavioral tests, the pBMSC-CM group showed significant improvement in struggle threshold compared to the no-treatment and culture medium only groups. Histological evaluation showed significantly less cartilage degeneration in the pBMSC-CM group than in the other two groups. There were no significant differences in the degree of synovitis among the three groups, although CGRP expression in DRG was lower in the pBMSC-CM group.

Conclusion

In this study, pBMSC-CM therapy reduced extracellular matrix degeneration in the articular cartilage and suppressed CGRP expression in the DRG in an MIA-induced rat OA model. Our findings suggest that pBMSC-CM therapy alleviated knee pain not only in the behavioral test but also in CGRP assessment. Overall, this study provides important insights into the potential of pBMSC-CM for the treatment of knee OA and contributes to future clinical trials.

Highlights

• •Therapeutic effects of pBMSC-conditioned medium were evaluated in a rat model of MIA-induced knee osteoarthritis.
• •pBMSC-conditioned medium reduced cartilage matrix degeneration and suppressed CGRP expression in the DRG.
• •Intra-articular injection of pBMSC-conditioned medium prevented cartilage degeneration and alleviated knee pain.

1. Introduction

Osteoarthritis (OA) is the most common chronic arthritis in the world, characterized by progressive degeneration of articular cartilage [1]. The existing conventional treatment approaches for OA are physiotherapy, drug therapy, and surgical operation. The primary objective of pharmacological therapy is to provide temporary pain relief and control inflammation in order to improve patients’ quality of life [2]. Currently, it is not possible to fundamentally slow the progressive degeneration of articular cartilage in OA, and taking this situation into account, a new cell-based therapy for OA has been gaining attention.

In recent years, mesenchymal stem cells (MSCs) have been explored as an innovative cell-based therapy for the treatment of OA. There have been several studies demonstrating the efficacy of MSC treatment in cartilage repair in animal models and even clinical studies [3,4]. The therapeutic effect of MSCs is primarily mediated by the paracrine effect, and trophic factors secreted by MSCs play a vital role in tissue repair [5,6]. However, MSC treatments remain controversial due to the need for highly controlled culture protocols and ethical issues. In addition, concerns about the risk of immune rejection and tumor formation remain barriers to their safe implementation in clinical practice [7,8].

In view of this, a new cell-based treatment for OA has attracted increasing attention. As an alternative to MSC therapy, the use of cell-free conditioned medium (CM) is a safer choice. CM contains a variety of trophic factors and extracellular vesicles (EVs) released by MSCs, and these secreted factors are thought to promote tissue repair even in the absence of cells [[9], [10], [11]]. CM can be prepared from various sources, including bone marrow MSCs (BMSCs), adipose tissue MSCs (AD-MSCs), and umbilical cord MSCs (UC-MSCs) [[12], [13], [14]]. Polydactyly bone marrow is one of the tissues that can be obtained relatively easily from young donors, and polydactyly bone marrow-derived MSCs (pBMSCs) exhibit a superior ability to differentiate into chondrocytes and regulate chondrogenesis [15]. In addition, EVs derived from pBMSCs have been reported to regulate chondrocyte formation through BMP4 signaling and to enhance the expression of chondrogenic markers such as aggrecan and type II collagen [15].Therefore, we have focused on pBMSC-derived conditioned medium (pBMSC-CM) as a valuable treatment for OA.

Although the exact mechanisms of OA-related pain remain unclear, previous studies have highlighted the involvement of calcitonin gene-related peptide (CGRP), a neuropeptide expressed in the dorsal root ganglion (DRG) [16,17]. It has been reported that OA upregulates CGRP expression in DRGs, where it mediates peripheral nervous system inflammation and contributes to pain [18,19]. Given its characteristics, the CGRP may reflect the pain mechanisms associated with arthritis and serve as a valuable tool for evaluating the efficacy of therapeutic interventions. In this study, we hypothesized that pBMSC-CM alleviates OA symptoms by preserving cartilage structure, reducing inflammation, and relieving joint pain.

The purpose of this study was to evaluate the preventive effect of pBMSC-CM treatment on cartilage degeneration and to assess CGRP expression in the DRG using a monoiodoacetate (MIA)-induced knee OA rat model.

2. Materials and methods

2.1. pBMSCs and pBMSC-CM

The polydactyly thumb was harvested by surgical excision from a 1-year-old male patient with congenital polydactyly. The inclusion criteria comprised patients who underwent surgical excision of supernumerary digits for congenital polydactyly, in cases where the discarded tissue was deemed appropriate for research use. Patients with active infections identified via preoperative blood tests were excluded. Written informed consent was obtained from the donor's proxy, and the study was approved by the hospital's ethics committee (permit number R2294). The marrow was scraped with a sharp spoon from the cut surface of the distal and proximal phalanges of the polydactyly thumb. The sample was then suspended in culture medium, alpha-Minimal Essential Medium (α-MEM; Thermo Fisher Scientific, Waltham, MA, USA) containing 15 % fetal bovine serum (FBS; SAFC Biosciences, Lenexa, KS, USA) and 0.25 μg/ml amphotericin B (Cheplapharm, Greifswald, Germany) and cultured at 37 °C in 5 % CO2. The medium was changed every 2–3 days and any cells not adhering to the flask were removed. After culturing for 14 days, the cells were detached using TrypLE™ Select Enzyme (Thermo Fisher Scientific) and subcultured at 3500–5000 cells/cm². The cells were passaged up to P4 and frozen stocks were prepared. To prepare the culture supernatant, the frozen stock was thawed, seeded into a flask at 104 cells/cm², and cultured overnight. The medium was then replaced with α-MEM without FBS or antibiotics and cultured for 3 days. The resulting culture supernatant was stored at −80 °C for up to 2 months and thawed immediately before use in the subsequent experiments. We considered the effects of a single freeze-thaw cycle on the biological activity of the conditioned medium to be minimal.

2.2. Characterization of pBMSCs

2.2.1. Flow cytometry

Immunophenotypes of the pBMSCs (P4) were analyzed by flow cytometry analysis. The cells were incubated with monoclonal antibodies CD90-FITC, CD105-PE, CD73-APC, CD34-BV421, CD45-FITC, CD11b-PE, CD19-BV421, HLA-DR-FITC or the isotype control (BD Biosciences, San Jose, CA, USA) for 30 min at 4 °C and analyzed using FACSLyric ™ and FACSuite software 1.2.1 (BD Biosciences).

2.2.2. Differentiation assays

Multipotency of pBMSCs (P4) was assessed as follows: for osteogenic differentiation, 3 × 10⁴ cells were seeded in 12-well plates and cultured for 4 days with pre-culture medium, α-MEM containing 15 % FBS (SAFC) and 1 % antibiotic-antimycotic (Thermo Fisher Scientific). The medium was subsequently changed to osteogenic differentiation medium, which consisted of pre-culture medium supplemented with 100 μM dexamethasone (Sigma-Aldrich, St. Louis, MO, USA), 10 mM glycerophosphate (Sigma-Aldrich), and 50 mg/ml ascorbic acid-2-phosphate (Sigma-Aldrich) for an additional 20 days. These cells were stained with alizarin red solution. For chondrogenic differentiation, 2.5 × 10⁵ cells were centrifuged at 400 g for 5 min in a 15-ml polypropylene tube. The pellet was cultured in chondrogenic differentiation medium, high glucose Dulbecco's modified Eagle's medium (Thermo Fisher Scientific) supplemented with 1 % antibiotic-antimycotic (Thermo Fisher Scientific), 1 % ITS Premix (Corning, Corning, NY, USA), 100 μM dexamethasone (Sigma-Aldrich), 50 μg/mL ascorbic acid-2-phosphate (FUJIFILM Wako Pure Chemical, Osaka, Japan), 40 μg/ml proline (Sigma-Aldrich), 10 ng/ml transforming growth factor β3 (Wako), and 500 ng/ml bone morphogenetic protein 6 (R&D Systems, Minneapolis, MN, USA) for 21 days. The pellets were fixed in 4 % paraformaldehyde, embedded in paraffin, cut into 5 μm sections, and stained with safranin O. For adipogenic differentiation, 1 × 10⁵ cells were seeded in 12-well plates and cultured in pre-culture medium for 4 days. The medium was then changed to adipogenic differentiation medium, which consisted of pre-culture medium supplemented with 100 μM dexamethasone (Sigma-Aldrich), 500 mM 3-isobutyl-1-methylxanthine (Sigma-Aldrich), and 100 mM indomethacin (Sigma-Aldrich) for an additional 21 days. These cells were fixed in 4 % paraformaldehyde and stained with oil red O solution.

2.3. Cell proliferation assay

To evaluate the effects of CM derived from pBMSCs on chondrocyte proliferation, a micromass culture system was employed using research-grade chondrocytes at passage 3. For comparison, CM derived from UC-MSCs was also applied under the same micromass culture conditions. Chondrocytes were adjusted to a concentration of 1 × 10⁷ cells/mL, and 2.5 × 10⁵ cells were seeded in the center of each well of a 12-well plate. The plates were incubated at 37 °C for 3 h to allow cell attachment. After washing with phosphate buffered saline (PBS), the cells were cultured at 37 °C in 5 % CO₂ using CM derived from each type of MSC, supplemented with the same concentrations of additives as those in standard chondrogenic medium, excluding BMP6. The medium was replaced every 3–4 days, and after 3 weeks of culture, the cell aggregates were digested using a collagenase solution. The number of chondrocytes was then determined by trypan blue exclusion. The proliferation rate was calculated by dividing the final cell number by the initial cell number.

2.4. Quantitative reverse transcription PCR

To evaluate the effect of CM derived from MSCs of various tissue origins on the enhancement of extracellular matrix production in chondrocytes, pellet cultures of chondrocytes were treated with CM, and gene expression related to cartilage matrix production was analyzed by quantitative reverse transcription PCR (qRT-PCR). Passage 3 chondrocytes were suspended in the same culture medium used in the cell proliferation assay and seeded at a density of 2.5 × 10^5 cells per well in non-adherent U-bottom 96-well plates (Sumitomo Bakelite, Tokyo, Japan). The cells were centrifuged at 200×g and cultured at 37 °C in a humidified atmosphere of 5 % CO₂. The medium was changed every 3–4 days, and the cultures were maintained for three weeks. The resulting chondrocyte pellets were homogenized using a QiaShredder (Qiagen, Valencia, CA, USA), and total RNA was extracted using the RNeasy Plus Mini Kit (Qiagen). cDNA was synthesized using the SuperScript VILO Master Mix (Invitrogen, Grand Island, NY, USA). qRT-PCR was performed with PowerUp™ SYBR™ Green Master Mix (Invitrogen) on a CFX96™ Real-Time PCR System (Bio-Rad Laboratories, Hercules, CA, USA). The cycling conditions were as follows: one cycle at 50 °C for 2 min and 95 °C for 2 min, followed by 40 cycles of amplification at 95 °C for 15 s and 60 °C for 30 s. The mRNA expression levels of each target gene were normalized to GAPDH and analyzed using the ΔΔCt method. The sequences of the primers used were as follows: COL2A1 (F): 5′-GGCAATAGCAGGTTCACGTACA-3′, (R): 5′-CGATAACAGTCTTGCCCCACTT-3'; ACAN (F): 5′-TACGAAGACGGCTTCCACCA-3′, (R): 5′-CTCATCCTTGTCTCCATAGC-3'; GAPDH (F): 5′-ATGGGGAAGGTGAAGGTCG-3′, (R): 5′-TAAAAGCAGCCCTGGTGACC-3'. These primers were synthesized by Eurofins Genomics (Tokyo, Japan).

2.5. RNA sequencing

Total RNA was extracted from cultured pBMSCs and UC-MSCs using the RNeasy Mini Kit (Qiagen, Hilden, Germany), according to the manufacturer's instructions. The purified RNA was sent to Takara Bio Inc. (Kusatsu, Shiga, Japan) for library preparation and sequencing. Library preparation was performed using the SMART-Seq v4 Ultra Low Input RNA Kit for Sequencing (Takara Bio USA, San Jose, CA, USA) and the Nextera XT DNA Library Prep Kit (Illumina, San Diego, CA, USA), following the manufacturers' protocols. The quality of the libraries was assessed using the Agilent 4200 TapeStation (Agilent Technologies, Santa Clara, CA, USA), after which the libraries were pooled and sequenced on the Illumina NovaSeq 6000 platform. The obtained reads were mapped to the human reference genome using the DRAGEN Bio-IT Platform v3.6.3 (Illumina) with GENCODE v39 annotation. Gene-level read counts and transcripts per million (TPM) values were calculated. To evaluate transcriptomic differences between pBMSCs and UC-MSCs, an MA plot was generated. Genes with large expression differences were highlighted to illustrate tissue-specific expression characteristics. In addition to the highly differentially expressed genes, other genes of interest potentially related to synovitis pathology were also examined. Furthermore, top genes with distinct expression in pBMSCs were extracted and functionally annotated based on their known biological roles.

2.6. Experimental animals

The experiments were performed using 18 5-week-old male Sprague-Dawley rats (Charles River Laboratories, Yokohama, Japan). They were maintained on an environmentally controlled 12-h light/dark cycle with food and water ad libitum. Prior to the injection, rats were anesthetized with intraperitoneal injection of 0.25 mg/kg of medetomidine (Nippon Zenyaku Kogyo, Fukushima, Japan), 2.0 mg/kg of midazolam (Astellas Pharma, Tokyo, Japan), and 2.5 mg/kg of butorphanol (Meiji Seika Pharma, Tokyo, Japan) following earlier research findings [20]. MIA (0.2mg/50 μL) dissolved in 100 μL saline was injected into the right knee joint using a 28-gauge needle as previously reported [18,21]. As a control, 100 μL saline was injected intra-articularly into the left knee. A week after MIA injection into the right knee, 18 rats were then divided into three experimental groups: (1) No-treatment group: injected with 100 μL saline solution (n = 6); (2) Culture medium only group: injected with 100 μL α-MEM (Thermo Fisher Scientific) (n = 6); and (3) pBMSC-CM group: injected with 100 μL pBMSC-CM (n = 6) (Fig. 1). Two weeks after treatment injection, 1 % Fluoro-gold (FG; Fluorochrome, Denver, CO, USA) retrograde neurotracer was injected into both knees in all groups to confirm retrograde labeling of knee afferent neurons [22,23]. One week after FG injection, all rats were euthanized under anesthesia and both knee joints and DRGs from the fifth lumbar vertebra (L5) were harvested. The knee joints were used for histological evaluation, and the DRGs were used for immunohistochemical staining.

Fig. 1.

(A) The polydactyly tissue was obtained through surgical resection. MSC-conditioned medium (CM) derived from polydactyly bone marrow MSCs (pBMSCs) was injected into rat knees. (B) Flow diagram of animal experiment.

2.7. Behavioral test

All groups were examined for behavioral tests every week. A simple, quantifiable behavioral test was previously reported, using the extension angle struggle threshold of the affected knee joint as an indicator of pain behavior [24,25]. The leg was manually extended from a flexed position to the point at which the rat squeaked or exhibited struggling behavior. The distance traveled by the heel from a flexed position to the extended position was recorded. The angle of extension was then calculated using trigonometric functions based on the distance between the length of the tibia and the heel position. The extension angles of both knees were then measured and the difference between both knees was calculated. All behavioral assessments were performed by the same evaluator to ensure consistency throughout the study. The evaluator adhered to standardized procedures to minimize potential bias.

2.8. Histological assessment of the knee

The knee joints were fixed in 10 % formaldehyde for 7 days, decalcified in 20 % ethylenediaminetetraacetic acid (EDTA) solution for 21 days, and subsequently embedded in paraffin wax. The specimens were sectioned sagittally at a thickness of 5 μm. The samples in each group were stained with hematoxylin and eosin and safranin O. The cartilage degeneration was scored using the Osteoarthritis Research Society International (OARSI) histopathology scoring system on a scale of 0–24 points (normal 0, worst 24) [26]. The average score for the femoral condyle and tibial plateau was used for statistical analysis. To assess the severity of synovial inflammation and structural changes in the IFP, synovitis was evaluated using an established IFP inflammation score [21,27]. The score consisted of two components: (1) cell infiltration at the surface of the IFP, and (2) fibrosis in the body of the IFP, which was assessed on a scale from 0 to 6 points (normal 0, worst 6) according to previously described methods. For each knee, one representative histological section was selected and evaluated. Two independent observers, blinded to each other's evaluations, scored the sections. To assess interobserver reliability, the interclass correlation coefficients (ICC) were calculated for both OARSI and IFP scores.

2.9. Immunohistochemical evaluation of the DRG

Four weeks after MIA injection, under deep anesthesia, the bilateral L5 DRGs in the three groups were resected and prepared for immunohistochemistry. The DRG specimens were immersed in the phosphate-buffered paraformaldehyde overnight at 4 $°C$. After storage in 0.01 M PBS containing 20 % sucrose for 20 h at 4 $°C$, they were frozen in liquid nitrogen. The DRGs were sliced into 10-μm-thick slices using a cryostat (Cryostar NX50, Epredia, Kalamazoo, MI, USA). The DRG sections were mounted on slides and then treated with a nonspecific blocking solution of 0.3 % Triton X-100 (Sigma-Aldrich) mixed with bovine serum albumin, and PBS for 90 min at room temperature. The sections were incubated with anti-CGRP rabbit antibody (1:1000 dilution; 24112; ImmunoStar, Hudson, WI, USA) for 20 h at 4 °C, and then incubated with Alexa Fluor 488-conjugated goat anti-rabbit IgG (for CGRP immunoreactivity, 1:1000; Jackson ImmunoResearch Laboratories, West Grove, PA, USA). Following each stage of the protocol, the tissue sections were rinsed three times in PBS. Subsequently, observation of the immunostained sections was performed using a fluorescence microscope (LSM780, Carl Zeiss, Germany). The number of FG-labeled CGRP positive neurons was counted and the percentage of the total number of FG-labeled neurons in the DRG sample was calculated using ImageJ software (National Institute of Health, USA).

2.10. Statistical analysis

All results are expressed as mean ± standard deviation (SD). Behavioral study data were analyzed using two-way repeated-measures analysis of variance (ANOVA) with post hoc Tukey's honestly significant difference (HSD) test. One-way ANOVA and Tukey's HSD test (post hoc) were used to analyze differences among multiple groups. The ICC of interobserver variability among the two observers in the histological score were 0.900 (OARSI score) and 0.870 (IFP score). A p-value of <0.05 was considered statistically significant. Correlations among the IFP score, OARSI score, and behavioral extension angle difference were assessed using Spearman's rank correlation analysis. All available data were included in the histological analyses. For the DRG analysis, data from one rat was not available due to a technical issue during tissue processing. Statistical analysis was performed using JMP Pro 16 (SAS Institute, Cary, NC, USA).

3. Results

3.1. Characterization of the pBMSC

Most of the pBMSCs expressed MSC positive markers CD90, CD105 and CD73 (98.6 %, 99.3 %, 99.2 %, respectively), and expression of negative markers CD34, CD45, CD11b, CD19 and HLA-DR was hardly observed (0.2 %, 0.2 %, 0.1 %, 0.2 %, 0.0 %, respectively) (Fig. 2A). The pBMSCs showed osteogenesis, chondrogenesis, and adipogenesis (Fig. 2B–D). These data indicate that the pBMSCs fulfill the criteria for MSCs as defined by the International Society for Cellular Therapy [28].

Fig. 2.

Multipotency of pBMSCs. (A) Flow cytometry analysis of surface markers of pBMSCs. (B) Osteogenesis. The pBMSCs were stained with Alizarin Red. (C) Chondrogenesis. Histological sections were stained with Safranin O. (D) Adipogenesis. The pBMSCs were stained with Oil Red-O.

3.2. Chondrocyte proliferation and qRT-PCR

Chondrocyte proliferation was significantly higher in both the pBMSC-CM (2.41 ± 0.05) and UC-MSC-CM (2.33 ± 0.12) groups compared to the control group (0.97 ± 0.13, p < 0.001), while no significant difference was observed between the pBMSC-CM and UC-MSC-CM groups (p = 0.64) (Fig. 3A).

Fig. 3.

Effects of CM from pBMSC and UC-MSC on chondrocyte proliferation and matrix-related gene expression. (A) Chondrocyte proliferation after culture with CM. Relative mRNA expression levels of COL2A1 (B), and ACAN (C) evaluated by RT-qPCR. Data are presented as mean ± SD. ∗p < 0.05, ∗∗∗p < 0.001.

Expression levels of cartilage matrix genes were evaluated in chondrocytes cultured with MSC-CM. COL2A1 expression was 1.00 ± 0.17 in the control group, 2.44 ± 0.71 in the UC-MSC-CM group, and 8.23 ± 0.50 in the pBMSC-CM group. COL2A1 expression levels were significantly higher in the pBMSC-CM group compared to both the control (p < 0.001) and UC-MSC-CM groups (p < 0.001), and the UC-MSC-CM group also showed a modestly higher level than the control group (p < 0.05) (Fig. 3B). ACAN expression was 1.00 ± 0.02 in the control group, 1.35 ± 0.11 in the UC-MSC-CM group, and 3.09 ± 0.30 in the pBMSC-CM group. ACAN expression levels were significantly higher in the pBMSC-CM group compared to both the control (p < 0.001) and UC-MSC-CM groups (p < 0.001), while the difference between the UC-MSC-CM and control groups was not statistically significant (p = 0.13) (Fig. 3C). Statistical analysis was performed using one-way ANOVA with post hoc Tukey's HSD test.

3.3. RNA sequencing

To investigate transcriptomic differences between pBMSC and UC-MSC, an MA plot was generated based on RNA sequencing data. The plot revealed distinct expression patterns between the two MSC sources. Among the differentially expressed transcripts, ten genes were identified with the highest expression levels in pBMSC compared to UC-MSC. These genes were: PENK, LINC01145, ENSG00000276664, ENSG00000283411, SCUBE3, NRN1, DLX5, MIR1307, CD24, and ELN. These top 10 genes were highlighted in red on the MA plot (Fig. 4).

Fig. 4.

MA plot comparing transcript expression between pBMSC and UC-MSC. Each dot represents a transcript detected by RNA sequencing. The x-axis indicates the average log2 CPM (Counts Per Million), and the y-axis shows the log2 fold change in expression between pBMSC and UC-MSC. The top 10 transcripts most upregulated in pBMSC compared to UC-MSC are highlighted in red.

3.4. Behavioral tests

Behavioral testing showed that one week after the MIA injection, all three groups had impaired extension of the affected knee due to inflammation, resulting in an increased difference compared to the contralateral knee. Interaction was detected between group and time (two-way ANOVA; F = 2.29; p = 0.03). At the 3-week mark, the difference in extension angle was lower in the pBMSC-CM group than in the no-treatment group (p = 0.032). At 4 weeks, the difference in extension angle was lower in the pBMSC-CM group than in the no-treatment and culture medium only groups (p = 0.015, p = 0.037, respectively) (Fig. 5).

Fig. 5.

Changes in extension angle difference between the affected and contralateral knees. The difference in extension angle between the affected and contralateral knees was recorded for each group. At the 3-week mark, the difference in extension angle was lower in the pBMSC-CM group than in the no-treatment group (∗p < 0.05). At 4 weeks, the difference in extension angle was lower in the pBMSC-CM group than in the no-treatment and culture medium only groups (∗p < 0.05). Black bar, significantly higher compared to other groups. Results are presented as mean ± SD. Behavioral data were analyzed using two-way repeated-measures ANOVA with post hoc Tukey's HSD test.

3.5. Histopathological assessment of the knee

The pBMSC-CM group showed stronger proteoglycan staining and maintained cartilage matrix, whereas cartilage matrix loss was observed in the no-treatment and culture medium only groups (Fig. 6A). According to quantitative histological scores based on the OARSI scoring system, the pBMSC-CM group (5.5 ± 1.5) had lower scores compared to the no-treatment (13.9 ± 2.9) and culture medium only (14.5 ± 2.2) groups (p < 0.001, p < 0.001, respectively) (Fig. 6C). The IFP was histologically evaluated to assess synovitis. An increase in the thickness of the lining cells at the surface of the IFP, as well as an expansion of the fibrotic area and a reduction in fat cells in the IFP, were observed in all groups (Fig. 6B). There were no significant differences in the IFP score among the three groups (no-treatment, 3.6 ± 1.2; culture medium only, 3.7 ± 1.0; pBMSC-CM, 3.4 ± 0.5) (Fig. 6D). Spearman's rank correlation analysis showed a significant positive correlation between OARSI scores and knee extension angle at week 4 (ρ = 0.548, p = 0.019), suggesting that greater cartilage degeneration was associated with increased knee pain. No significant correlations were found between IFP scores and either OARSI scores (ρ = 0.081, p = 0.751) or knee extension angle (ρ = 0.338, p = 0.169).

Fig. 6.

(A) Histological analysis of cartilage using Safranin O staining. (B) Histological analysis of synovial tissue and the IFP using Hematoxylin and Eosin (HE) staining. (C) Average of OARSI histological scores. (D) Average of IFP inflammation scores. Black bar, significantly higher compared to other groups. ∗∗p < 0.01, ∗∗∗p < 0.001.

3.6. Immunohistochemical assessment of the DRG

The distribution of CGRP-expressing sensory nerve fibers in L5 DRG was evaluated by immunohistochemical staining to assess pain innervation in the knee joint. Higher numbers of CGRP-positive neurons in the DRG were observed after MIA injection, whereas the contralateral side without MIA injection had fewer CGRP-positive neurons (Fig. 7A). To confirm retrograde labeling of knee joint afferent neurons, FG-labeled neurons were observed in the DRG (Fig. 7B). The proportion of CGRP- and FG-labeled neurons among FG-labeled neurons in the pBMSC-CM (37.2 ± 15.6) group was significantly lower than in the no-treatment (56.5 ± 15.0) and culture medium only (53.9 ± 16.8) groups. (p < 0.001, p < 0.001, respectively) (Fig. 7D).

Fig. 7.

Fluorescence photomicrographs of the L5 DRG. CGRP positive DRG neurons are indicated in red (A-1, A-2, A-3, A-4), and FG labeled DRG neurons are indicated in yellow (B-1, B-2, B-3, B-4). Merged images are displayed in the bottom row (C-1, C-2, C-3, C-4). The ratio of FG labeled CGRP positive neurons/FG labeled DRG neurons (D). Black bar, significantly higher compared to other groups. ∗∗∗p < 0.001.

4. Discussion

In this study, pBMSC-CM prevented articular cartilage degeneration in a rat knee OA model. According to our experiments, intra-articular injections of pBMSC-CM improved behavioral tests and histological findings of articular cartilage compared to the no-treatment and culture medium only groups. Furthermore, we found that injection of pBMSC-CM into the knee joint suppressed CGRP expression in the DRG neurons.

MSCs are multipotent stromal cells that can be differentiated into multiple cell lineages, such as chondrocytes, osteoblasts, and adipocytes [29]. In addition to their multilineage differentiation potential, MSCs possess immunomodulatory and anti-inflammatory properties [30]. Recent studies have shown that MSC therapy for OA helps protect articular cartilage and prevents the progression of chronic pain and joint dysfunction due to its paracrine effects [3,31,32]. However, MSC treatments remain controversial due to ethical issues and the need for highly controlled culture protocols. Some negative responses to MSC treatment were also observed, namely increased pain-related behavior and osteophyte formation [33]. Moreover, concerns about the risk of immune rejection and tumor formation remain barriers to their safe implementation in clinical practice [7,8].

In these respects, cell-free treatments such as CM and EVs have gained attention as potentially safer options for OA treatment [34]. CM contains a variety of bioactive substances which include cytokines, chemokines, growth factors, EVs released by MSCs, and these secreted factors are thought to promote tissue repair even in the absence of cells [[9], [10], [11]]. CM can be prepared from various sources such as BMSCs, AD-MSCs, and UC-MSCs [[12], [13], [14],35]. MSCs have immunomodulatory and anti-inflammatory effects, but factors secreted from MSCs also exhibit anti-inflammatory and anti-catabolic effects in osteoarthritic synovium and cartilage [5]. Among the commonly mentioned growth factors in CM are TGF-β, FGF, HGF, and PDGF, which play important roles in tissue regeneration processes [36,37]. TGF-β possess immunomodulatory properties and protect chondrocytes against oxidative stress [38]. Other components of CM include tissue inhibitors of metalloproteinases (TIMP-1 and TIMP-2), which are involved in extracellular matrix turnover. TIMPs regulate matrix metalloproteinase activity and are known for their anti-fibrotic effects [39]. Intra-articular injection of BMSCs-CM was sufficient in attenuating OA by protecting the subchondral bone structure, maintaining cartilage matrix homeostasis, and enhancing autophagy to inhibit chondrocyte apoptosis [40]. Similarly, factors secreted from MSCs in adipose tissue may protect articular cartilage from extracellular matrix loss in rat knee OA [41].

MSCs from young donors have higher immunosuppressive capacity and greater potential to differentiate into bone and fat cells compared to those from older donors [42]. Yeh et al. reported that infant BMSCs exhibited higher expression levels of chondrogenic genes such as SOX9, COL2, and COL10, and superior chondrogenic potential compared to adult BMSCs [43]. In addition, EVs secreted from pBMSCs not only promoted chondrocyte proliferation and migration, but also delayed the progression of OA in a rat model [15]. In this study, pBMSC-CM showed higher expression levels of COL2A1 and ACAN in cultured chondrocytes compared to UC-MSC-CM. Consistently, in our in vivo rat model, pBMSCs-CM preserved the extracellular matrix of articular cartilage and demonstrated a potential protective role against cartilage degeneration. Given the proven effectiveness of polydactyly bone marrow, tissue harvested from polydactyly patients may provide a one-step treatment for cartilage disease.

While osteoarthritis (OA) has been linked to synovitis, the IFP score did not show significant differences among the three groups, suggesting that synovitis was not notably improved in this study. Udo et al. showed that in the MIA-induced rat OA model, cartilage degeneration progresses over approximately 8 weeks, whereas IFP inflammation and degeneration progress rapidly within 7 days [21]. Araya et al. also reported that platelet-rich plasma injected one day after the MIA injection improved synovitis scores compared to the control group [27]. However, it may be explained by the possibility that synovitis was suppressed because the injection was administered prior to the synovitis reaching its peak. Furthermore, this report suggests that multiple doses given at intervals may be effective, similar to the report by Cheng et al. that synovitis improved after two doses of CM in an animal model [41]. In our study, secreted factors from MSCs were administered via a single injection one week after the MIA injection. As a result, the synovitis was likely already advanced by the time treatment began, and since only one injection was given, the synovitis may not have improved. Furthermore, correlation analysis revealed that the behavioral outcome was significantly associated only with the OARSI score, suggesting that preservation of the extracellular matrix structure in articular cartilage may have played a more important role in pain relief. To further investigate the underlying mechanisms, we performed RNA sequencing to explore gene expression profiles in pBMSCs. RNA sequencing is a well-established method for comprehensively analyzing functional changes in MSCs, allowing for the identification of molecular mechanisms that cannot be detected by single-cytokine assays [44]. Through this analysis, we found that the expression of SCUBE3 and DLX5 was upregulated in pBMSCs. SCUBE3 is involved in TGF-β signaling and promotes chondrogenesis, while DLX5 is a transcription factor that enhances chondrogenic differentiation [45,46]. Moreover, pBMSCs also showed increased expression of PENK, which is implicated in nociceptive modulation, and NRN1, which contributes to neuroprotection [47,48]. In particular, the PENK gene encodes enkephalins, which exert analgesic effects by acting on opioid receptors in both peripheral and central neurons [47]. In this study, in addition to the preservation of cartilage structure, the upregulation of gene expression related to endogenous opioid activity may account for the observed improvements in pain behavior and CGRP suppression, despite insufficient suppression of synovitis.

Recently, increasing CGRP expression in DRG neurons has been reported to be associated with pain in the knee osteoarthritis model [49]. Aso et al. investigated the relationship between CGRP expression in the subchondral bone and OA-related pain [16]. Miyamoto et al. also showed that the expression of retrograde neurotracer-labeled CGRP positive neurons in the DRG increased after induced inflammation of the hip [50]. However, these reports investigated the association between pain and CGRP expression but did not evaluate the therapeutic effects on OA. With this in mind, we used a retrograde neurotracer to evaluate the therapeutic effect on knee pain in animal models. The results suggested that factors secreted by MSCs reduce the expression of neurotracer-labeled CGRP in DRG neurons and alleviate knee-derived pain.

There were limitations to this study. First, this study was a small animal experiment, and further experiments using larger animal models and human clinical trials are necessary. Second, the pathological changes in the MIA-induced rat OA model may differ from those in human knee OA. Third, a comprehensive analysis of the components of MSCs-CM and their specific roles and mechanisms in this therapeutic process remains incomplete. Further elucidation is required to guide future research and clinical applications. Fourth, there are various methods for the preparation of MSC-CM, which may result in differences in the content of growth factors, secretomes, EVs, and RNA. Fifth, this study evaluated tissue at a single time point, limiting the observation of changes in cartilage degradation and synovitis changes over time. In this study, pBMSC-CM therapy reduced extracellular matrix degeneration in the articular cartilage and suppressed CGRP expression in the DRG in an MIA-induced rat OA model. Our findings suggest that pBMSC-CM therapy alleviated knee pain not only in the behavioral test but also in CGRP assessment. Overall, this study provides important insights into the potential of pBMSC-CM for the treatment of knee OA and contributes to future clinical trials.

Ethics approval

All animal experiments were performed following the guidelines of the Animal Care and Use Committee of Hyogo Medical University in accordance with the Act on Welfare and Management of Animals and with the approval of the committee under the institutional approval numbers of #23–092A.

Declaration of competing interest

All authors have no conflicts of interest relevant to this article.