This study employed both in vitro and in vivo experimental designs, with in vitro experiments using human chondrocytes for FGF18 overexpression and knockdown, and in vivo experiments using a DMM mouse model to evaluate FGF18 therapeutic effects. The independent variable was FGF18 expression level (overexpression, knockdown, or control), and dependent variables included chondrocyte proliferation, apoptosis, FOXN2 expression, and PTOA-related markers (MMP13, ADAMTS-5, CTX-II, aggrecan).

The usage of human samples in this study has been approved by the Fuzhou Second General Hospital (Approval No. 2022140), and informed written consent was obtained from all participants prior to surgery for both specimen collection and research participation. All animal experimental procedures were approved by the Experimental Animal Ethics Committee of Fujian Medical University (Approval No. IACUC FJMU 2024-0248).

Based on comprehensive literature review of factors involved in cartilage development, repair and osteoarthritis pathogenesis, FGF18 was identified as a promising candidate gene for investigation in PTOA. Previous studies have established FGF18 as an important regulator of chondrocyte proliferation and cartilage formation (Davidson et al., 2005; Ellsworth et al., 2002), and as a therapeutic target with recombinant FGF18 (sprifermin) in clinical trials for primary OA (Hochberg et al., 2019; Eckstein et al., 2021). However, its specific roles and regulatory mechanisms in post-traumatic osteoarthritis remained poorly understood, warranting further investigation (Buniello et al., 2019; MacArthur et al., 2017; Welter et al., 2014).

Articular cartilage tissue was obtained from knee biopsies of patients undergoing total knee arthroplasty (PTOA group, n = 10), and non-OA specimens were sourced from individuals undergoing surgery due to acute traumatic injuries, including fractures or ligament tears, with no prior history of osteoarthritis, as normal controls (NC group, n = 8). Once obtained, the cartilage was immediately placed in liquid nitrogen. The cartilage tissues were processed for further histological examinations.

Male C57BL/6 mice weighing 200 g (8 weeks old) were purchased from the Experimental Animal Centre of Tongji Medical College (Wuhan, China), a nationally certified SPF facility that provides genetically standardized laboratory animals. After transportation, mice underwent a 7-day acclimatization period at the animal facility of Fujian Medical University before experimental procedures. All animal experiments were conducted at the accredited animal facility of Fujian Medical University under strict adherence to institutional and national ethical guidelines.

For the OA experiment model, the mice were placed under anesthesia using sodium pentobarbital (100 mg/kg) administered intraperitoneally by injection. Then the mice were traversed by the medial collateral ligament and destabilized by the medial meniscus (DMM) in the right knee. Then the mice were injected with replication-deficient lentivirus (Lenti)-FGF18 and saline via intra-articular administration 3 days, 1 week, 2 weeks, and 3 weeks after the DMM surgery. The mice were sacrificed 8 weeks after the operation by the cervical dislocation method, and the knee joints were harvested.

Articular cartilage from OA patients and non-OA individuals as well as from the knees of mice after the construction of the OA model, were isolated as previously described (Hendriks et al., 2006). In short, cartilage was separated from the underlying bone and connective tissue and digested in collagenase type II (0.15% Worthington) in DMEM supplemented with 100 mg/mL of streptomycin and 100 U/mL of penicillin for 20–22 h. Chondrocyte proliferation medium (DMEM supplemented with 10% FBS, 1 × nonessential amino acids, 0.2 mM Ascorbic acid 2-phosphate, 0.4 mM proline, 100 U/mL penicillin, and 100 μg/mL streptomycin) was then used to cultivate the primary chondrocytes.

The FGF18 sequence was synthesized and subcloned into the pcDNA3.1 (Invitrogen, Life Technologies, Carlsbad, CA, United States) vector. Overexpression of FGF18 was achieved via pcDNA3.1-FGF18 transfection, with an empty pcDNA3.1 vector used as a control (Invitrogen). Negative control siRNA or siRNA against FGF18 (Sigma-Aldrich, St. Louis, MO, United States) was transfected into chondrocytes at a concentration of 50 µM using Lipofectamine 3000 Transfection Reagent (Invitrogen). Transfection was performed for 6 h at 37 °C, followed by medium replacement and a 48-h recovery period before subsequent functional analyses.

Human cartilage OA tissues (OA) and non-OA cartilage samples (NC) were fixed in 4% neutral-buffered paraformaldehyde (PFA: Salarbio, Beijing, China), embedded in paraffin and sliced with a rotary microtome into 5 µM slices. The cartilage sections were then stained with hematoxylin and eosin, Safranin O 0.25% v/v (Salarbio) for 5 min with Fast Green 0.01% v/v (f7258: Sigma-Aldrich, St Louis, MO, United States) as a counterstain.

The knee joints of DMM-induced OA mice were fixed in 10% formaldehyde at 4 °C for approximately 24 h. The cartilage tissues were then decalcified in 0.5 M ethylenediaminetetraacetic acid in PBS (pH 7.4) for 2 weeks. Then the specimens were embedded in paraffin, sliced into 5 µM sections, and stained with hematoxylin and eosin, Safranin O, and Fast Green.

For immunohistochemical staining, paraffin-embedded human cartilage sections from OA patients and NC individuals were treated with 3% hydrogen peroxide to block the activity of endogenous peroxidase and then incubated in 10% goat serum in PBS. Thereafter, the sections were incubated with rabbit anti-FGF18 (Invitrogen, Waltham, MA, #PA5-106562) diluted 1:100 in 10% goat serum in PBS overnight, then thoroughly washed and incubated with goat anti-rabbit immunoglobulin G (Vector Laboratories, Burlingame, CA, #BA-1000) diluted in 1:100 PBS. ABC kit (Vector Laboratories, #PK-4000) was used for detection followed by the DAB kit (Vector Laboratories, #SK-4100) according to the manufacturer’s protocol. Quantification of anti-FGF18 positive staining was performed usi ng ImageJ.

For immunofluorescence staining, chondrocytes from patients that were transfected with indicated plasmids, and chondrocytes from DMM-induced OA mice transfected with Lenti-FGF18 and saline (control) were analyzed using immunofluorescence microscopy for the expression of CTX-II (both human and mice chondrocytes), BMP-2 (human chondrocytes only), Aggrecan (mice chondrocytes only), MMP13 (mice chondrocytes only), and ADAMTS-5 (mice chondrocytes only). After washing three times with PBS, cells were fixed with 4% formaldehyde for 10 min at room temperature, followed by blocking with 2.5% BSA for 1 h at room temperature. Cells were subsequently incubated with anti-CTX-II (1:500; ab3092; Abcam; validated for human and mouse), anti-BMP-2 (1:500; ab6285; Abcam; human-specific), anti-Aggrecan (1:500; ab36861; Abcam; mouse-specific), anti-MMP13 (1:500; ab39012; Abcam; mouse-specific), anti-ADAMTS-5 (1:500; ab41037; Abcam; mouse-specific) overnight at 4 °C in DPBS containing 2.5% BSA and 0.1% Tween 20. Samples were then washed three times with Dulbecco’s Phosphate-Buffered Saline (DPBS), followed by probing with Fluorescein isothiocyanate (FITC) or Tetramethylrhodamine (TRITC) conjugate secondary antibody (1:1000; Thermo Fisher Scientific) at room temperature for 1 h. Thereafter, cells were counterstained with DAPI and analyzed using fluorescence microscope.

Total RNA was extracted from chondrocytes using RNA isolater Tatol RNA Extraction Reagent (Vazyme, China). Purified RNA was reverse-transcribed into cDNA with HiScript^® II 1st Strand cDNA Synthesis Kit (Vazyme, China) according to the manufacturer’s protocol. Real-time qRT-PCR was performed using the Applied Biosystems StepOnePlus Real-Time PCR System (Foster City, CA, United States) with SYBR Green PCR Master Mix (Toyobo, Japan). The primers used for qRT-PCR were as follows: FGF18 forward: 5′-CAA​GGA​GTG​CGT​GTT​CAT​TGA​GA-3′, reverse: 5′-CGG​GAT​CGC​TTA​GTA​ACT​GTG​GT-3'; GAPDH forward: 5′-GGC​AAG​TTC​AAC​GGC​ACA​G-3′, reverse: 5′-GCC​AGT​AGA​CTC​CAC​GA-3'. To quantify the relative expression levels of mRNA, the values were normalized against the endogenous reference GAPDH. The 2^(-ΔΔCT) method was used for data analysis. Additionally, qRT-PCR was employed to validate RNA-sequencing results, with independent training (n = 6) and validation sets (n = 6) used to confirm the expression of key differentially expressed genes, such as FOXN2, across datasets.

Total RNA was extracted with TRIzol (Invitrogen) from OA and NC chondrocytes. The purity and quantity of total RNA were measured by Nanodrop. The integrity of RNA was evaluated using the RNA Nano6000 Assay Kit on the Bioanalyzer 2100 system (Agilent Technologies, CA, United States). A total of 1 µg RNA per sample was used as an input for further analysis. Strand-specific RNA-sequencing libraries were generated using the NEBNext Ultra II RNA Library Prep Kit (Illumina, United States). Library quality was evaluated on the Agilent Bioanalyzer 2100 system (Agilent Technologies, CA, United States). Final libraries were sequenced on the Illumina Novaseq 6000 platform by 150 bp paired-end reads. The expression matrix was normalized and differentially expressed genes (DEG) were searched using the R package DeSeq2 (52) (thresholds used: adjusted p-value <0.01 and log2(Fold-Change) > 0.58). Supplementary R packages compiled with R version 4.1.1 were used to generate data visualization: ComplexHeatmap (2.8.0).

For total protein extraction, primary chondrocytes of OA patients (OA) and non-OA individuals (NC) were lysed in RIPA buffer (Fudebio, Hangzhou, China) containing protease inhibitor cocktails and phosphatase inhibitors (Fudebio, Hangzhou, China). A bicinchoninic acid protein assay kit (Thermo Fisher Scientific) measured the protein quantities. Equivalent amounts of protein were separated by SDS-PAGE and transferred onto polyvinylidene fluoride (PVDF) membranes (Merck KGaA). After incubation with anti-FGF18 antibody (1:1000, Proteintech) or anti-GADPH (1:1000, Cell Signaling Technology, United States). After that, the membranes were treated with horseradish peroxidase (HRP)-conjugated secondary antibodies (FDM007 and FDR007, Fudebio, Hangzhou, China). Following washing, an enhanced chemiluminescence kit (FD8030, Fudebio, Hangzhou, China) was utilized to identify the band signals. Furthermore, primary chondrocytes after being transfected with pcDNA3.1, pcDNA3.1-FGF18, control siRNA, and FGF18 siRNA were analyzed for the expression of FOXN2, Col II, MMP13 and ADAMTS-5 using primary antibodies against collagen type II (ab3778; Abcam, Cambridge, United Kingdom), MMP13 (sc-515284; Santa Cruz Biotechnology, Dallas, TX, United States), ADAMTS-5 (MAB2198; R&D Systems, Minneapolis, MN, United States), and FOXN2 (#12941, Cell Signaling Technology, Danvers, MA, United States) following instructions as discussed above.

Human articular chondrocytes transfected with pcDNA3.1, pcDNA3.1-FGF18, control siRNA, and FGF18 siRNA were collected using trypsin without EDTA (Thermo Fisher Scientific, Waltham, MA, United States) and washed with PBS twice. The cells were then stained using Annexin V-fluorescein isothiocyanate/propidium iodide (FITC/PI) Apoptosis kit (BD Biosciences, Franklin Lakes, NJ, United States) according to the manufacturer’s protocol. BD FACS flow cytometer (BD Biosciences) was used to identify cellular apoptosis.

For determining transfection efficiency, trypsin was used to remove chondrocytes from culture vessels. Cells were recovered in PBS supplemented with 2% FBS after 5 min. Then green fluorescent protein (GFP), expressed by the cells was analyzed in chondrocytes utilizing Beckman Coulter’s Kaluza analytical software and flow cytometry (EPICS XL FACS).

EDU (5-ethynyl-20-Deoxyuridine) assay was used to evaluate chondrocyte proliferation abilities after they were transfected with pcDNA3.1, pcDNA3.1-FGF18, control siRNA, and FGF18 siRNA. EDU staining was performed according to the manufacturer’s protocol using EDU detecting kits (Ribobio, Guangzhou, China). Briefly, EDU was added at a final concentration of 10 µM and the cells were incubated for 2 h at 37 °C. After labeling, the cells were fixed with 4% PFA for 15 min and permeabilized with 0.2% Triton X-100 for 10 min. The cells were then stained in the dark with click reaction solution for 30 min before being counterstained with Hoechst 33342. Using fluorescence microscopy (Olympus, Tokyo, Japan). EDU-positive cells were counted and the percentage was reported.

Data are presented as means ± SD unless otherwise specified. For comparisons between two groups, two-tailed Student's t-tests were used. For multiple group comparisons, one-way analysis of variance (ANOVA) with Tukey’s post hoc test was applied. For RNA-sequencing data, normalization and differential expression analysis were conducted using the DESeq2 package with a threshold of adjusted P < 0.05 and |log2FoldChange|>1. Sample sizes were determined based on power calculations (α = 0.05, β = 0.2) from preliminary experiments. All statistical analyses were conducted by investigators blinded to group allocation using GraphPad Prism software (version 9.0). P values less than 0.05 were considered statistically significant.

The expression of FGF18 in the cartilage from PTOA and NC trauma patients was explored to identify its role in osteoarthritis development or progression. First, using histological analysis, the morphology of the cartilage was inspected. The histological results from PTOA patients showed that the cartilage was damaged as opposed to the cartilage from NC patients (Figure 1a). Then, using immunochemical analysis, the expression levels of FGF18 were observed (Figure 1b) and the expression patterns of FGF18 were significantly downregulated in the cartilage of patients with PTOA. The results were quantified using real-time RT-PCR and Western blot analysis (Figures 1c,d). These results illustrate that FGF18 may have a functional role in PTOA development or progression.

FGF18 has been implicated in chondrogenesis and osteogenesis thereby stimulating the formation of cartilage (Jan et al., 2023; Wu et al., 2022). This action implies that FGF18 may prevent or improve PTOA. Thus low expression of FGF18 could potentially be associated with poor cartilage regeneration. To determine this, primary chondrocytes from PTOA patients were transfected with the overexpression of FGF18 or FGF18 knockdown, and their controls. The transfection efficiency was analyzed to validate the success of the transfection (Figure 2a). The expression of FGF18 was elevated when FGF18 was overexpressed but downregulated when FGF18 was silenced (Figure 2b). The level of cellular apoptosis of chondrocytes after transfection with FGF18 siRNA was elevated when compared to its control, while the level of apoptosis was decreased when pcDNA3.1-FGF18 was transfected compared to its control (Figure 2c). Furthermore, the EDU assay demonstrated that chondrocyte proliferation was significantly higher after the transfection of FGF18 overexpression into the chondrocytes but lower after the transfection of FGF18 knockdown (Figure 2d). This implicates FGF18 in governing the proliferation of chondrocytes, which may prevent matrix breakdown.

To comprehensively identify the mechanism of FGF18 in the development or progression of PTOA, we performed RNA-seq analysis in chondrocytes from PTOA patients. It was found that FOXN2 was among the differentially expressed genes (Figure 3a). The RNA-seq analysis was conducted on in vitro human chondrocytes from normal controls (NC, n = 3 biological replicates) and FGF18 knockout (FGF18 KO, n = 3 biological replicates), revealing 19 differentially expressed mRNAs between FGF18 knockout cells and normal controls, as summarized in Table 1 for both the training and validation sets. Among these, FOXN2 showed robust upregulation in FGF18 KO cells (log2FC ≈ 3.60, p = 0.001 in RNA-seq; training set: fold change 2.5, p = 0.006; validation set: 2.55, p = 0.004), ranking among the most significantly altered genes. FOXN2 was prioritized for further validation over other DEGs due to its significant fold-change, statistical significance, consistency across training and validation datasets, and established biological relevance to chondrocyte differentiation and hypertrophy. Then we determined that FOXN2 was upregulated when FGF18 was knocked down, this suggests that FOXN2 is a molecular target of FGF18 in the regulatory process of PTOA. Factors associated with PTOA such as MMP13 and ADAMTS-5 were also observed to be upregulated (Figure 3b). Immunofluorescence staining in Figure 3c,d specifically highlights the expression of BMP-2 (Bone Morphogenetic Protein-2, a key pro-anabolic factor that promotes cartilage matrix synthesis and repair by stimulating collagen production and chondrocyte proliferation) and CTX-II (a biomarker of type II collagen degradation) in chondrocytes under FGF18 knockdown conditions. BMP-2 exhibited reduced fluorescence intensity in FGF18-silenced chondrocytes, while CTX-II was markedly upregulated. The diminished BMP-2 signal suggests impaired cartilage repair capacity, whereas elevated CTX-II reflects enhanced collagen breakdown. These results indicate that FGF18 knockdown disrupts the balance between anabolism (via BMP-2 suppression) and catabolism (via MMP13, ADAMTS-5 and CTX-II upregulation), exacerbating cartilage degeneration in PTOA. From these results, we gathered that FGF18 regulates FOXN2 and prevents or counteracts the pathological onset of PTOA. We surmise that the downregulation of FGF18 in PTOA results in the upregulation of FOXN2 together with MMP13, ADAMTS-5 and BMP-2, factors associated with PTOA.

Next, we sought to determine whether the overexpression of FGF18 could attenuate the progression of PTOA in vivo. Lentiviral particles of FGF18 and saline (control) were injected into the knee joints of mice after DMM-induced PTOA. Using H&E staining and Safranin O-fast green for assessment. We observed that the overexpression of FGF18 substantially inhibited articular matrix breakdown and erosion relative to saline-treated controls (Figure 4a). Immunofluorescence analysis revealed upregulated aggrecan expression and downregulated CTX-II levels, both established biomarkers of cartilage metabolism, suggesting enhanced structural integrity of the articular matrix (Figure 4b). However, MMP13 and ADAMTS-5 were downregulated (Figure 4c), these factors break down type II collagen and aggrecan, respectively, and lead to the advancement of PTOA. From these findings, we can establish that the overexpression of FGF18 attenuated cartilage degradation under PTOA conditions.

Next, we sought to determine whether the overexpression of FGF18 could attenuate the progression of PTOA in vivo. Lentiviral particles of FGF18 and saline (control) were injected into the knee joints of mice after DMM-induced PTOA. Using H&E staining and Safranin O-fast green for assessment. We observed that the overexpression of FGF18 substantially inhibited articular matrix breakdown and erosion relative to saline-treated controls (Figure 4a). Immunofluorescence analysis revealed upregulated aggrecan expression and downregulated CTX-II levels (Figure 4b). Furthermore, MMP13 and ADAMTS-5 were downregulated (Figure 4c), these factors break down type II collagen and aggrecan, respectively, and lead to the advancement of PTOA.