Deficiency of Krüppel-like factor 15 activates β-catenin and is involved in the development of osteoarthritis

Abstract

Krüppel-like factor 15 is a metabolism-related transcription factor that has been implicated in the regulation of the Wnt/β-catenin signaling pathway. This study aimed to elucidate its function in cartilage, particularly its interaction with β-catenin, using inducible cartilage-specific knockout mice. Krüppel-like factor 15 knockout mice and control mice underwent destabilization of medial meniscal surgery to induce osteoarthritis. Histological analysis and immunohistochemistry were carried out, along with in vitro co-immunoprecipitation, double immunofluorescence, and subcellular fractionation, and a dual-luciferase reporter assay. The effects of Wnt3A stimulation and treatment with Wnt/β-catenin inhibitors were evaluated. Knockout mice showed significantly more cartilage degeneration at eight weeks after surgery when compared with control mice. Knockout mice also showed elevated Osteoarthritis Research Society International scores and increased expression of β-catenin, matrix metalloprotease-13, and a disintegrin-like metalloprotease with thrombospondin motifs-5, along with decreased expression of SRY-box transcription factor 9. Krüppel-like factor 15 was seen to physically interact and colocalize with β-catenin in chondrocyte nuclei. Its deficiency promoted canonical Wnt/β-catenin pathway activation and nuclear β-catenin accumulation. Consistently, dual-luciferase reporter assays demonstrated increased β-catenin/TCF transcriptional activity in Krüppel-like factor 15-deficient chondrocytes, which was further enhanced by Wnt3A stimulation. Wnt3A stimulation amplified catabolic gene expression and suppressed anabolic factor expression. However, administration of a Wnt/β-catenin signaling inhibitor canceled out these effects. This study indicates that Krüppel-like factor 15 interacts with β-catenin and functions as a negative regulator of Wnt/β-catenin signaling in cartilage, maintaining matrix homeostasis, and suppressing osteoarthritis progression. Deficiency of Krüppel-like factor 15 promotes catabolism and inhibits anabolism mediated by β-catenin in cartilage. These findings suggest that Krüppel-like factor 15 may be a potential therapeutic target for osteoarthritis, either by restoring its function or enhancing its interaction with β-catenin.

Supplementary Information

The online version contains supplementary material available at 10.1186/s13075-026-03838-2.

Background

Osteoarthritis (OA) is a common joint disease that develops due to aging and lifestyle habits, which significantly reduces quality of life. [1, 2] The pathology of OA progresses as a series of complex changes in the entire joint structure, accompanied by degeneration of articular cartilage, synovitis, bone sclerosis, and osteophyte formation. To date, no effective disease-modifying treatment has been established to treat OA. [3, 4] Artificial joint replacement surgery has delivered good long-term results. [5] However, the efficacy of conservative treatments such as analgesics, hyaluronic acid injections, and rehabilitation is limited in their efficacy. [6] Therefore, elucidating the mechanisms underlying OA progression and identifying new therapeutic targets is necessary.

Krüppel-like factors (KLFs) are a family of transcription factors with C2H2-type zinc fingers that are involved in a wide range of biological processes. [7] Currently, 17 members of the KLFs family have been identified, each with different tissue specificities and functions. [8, 9] Recently, the involvement of KLFs in OA has attracted attention. In particular, KLF2 and KLF4, which are involved in the maintenance of chondrocyte homeostasis, are known to be associated with OA pathology. [10] KLF2 inhibits oxidative stress and ferroptosis and protects mitochondrial function in chondrocytes, thereby suppressing OA progression. Specifically, KLF2 inhibits the oxidative stress response by activating Nrf2/ARE signaling. [11] It improves ferroptosis and mitochondrial dysfunction via the SIRT1/GPX4 pathway, thereby alleviating OA pathology. [12] KLF4 promotes chondrocyte differentiation and expression of extracellular matrix (ECM) components, and is considered an important transcription factor in cartilage repair regulation. Furthermore, KLF4 has a protective function in OA by regulating the inflammatory response of synovial cells and chondrocytes and suppressing tissue destruction. [13, 14].

KLF15 is a metabolism-related transcription factor that functions as a key regulator of metabolic homeostasis in various tissues, specifically in lipid metabolism and energy homeostasis. [15–18] In skeletal muscle, KLF15 cooperates with peroxisome proliferator-activated receptor delta (PPARδ) to regulate lipid flux and maintain mitochondrial function, contributing to exercise response and muscle function adaptation. [19] However, the possibility that KLF15 may also be involved in the pathology of OA has recently attracted attention. Disruption of lipid metabolism is considered a new pathological factor involved in the disturbance of cartilage homeostasis and OA progression. [20] Abnormal fatty acid oxidation impairs cartilage homeostasis through SRY-box transcription factor 9 (SOX9) degradation and epigenetic regulation. [21] Such metabolic abnormalities promote cartilage degeneration. Therefore, KLF15 may function as a metabolic regulator even within chondrocytes. KLF15 directly induces SOX9 expression during chondrogenic differentiation of human mesenchymal stem cells (hMSCs) and promotes cartilage differentiation. [22] Additionally, Klf15 expression is significantly reduced in chondrocytes derived from patients with OA compared with healthy individuals. [10] Furthermore, KLF15 plays a role in suppressing the inflammatory response and progression of matrix degradation by suppressing matrix metalloproteinase-13 (MMP-13) transcription in response to tumor necrosis factor-alpha (TNFα) stimulation. [23] Our previous study reported that Klf15 deficiency reduces peroxisome proliferator-activated receptor gamma (PPARγ) expression, which promotes chondrocyte apoptosis, resulting in the worsening of OA pathology. [24] Thus, KLF15 is a multifaceted transcription factor that contributes to the metabolic regulation, differentiation control, and suppression of inflammatory responses in cartilage. Specifically, its role in homeostasis is extremely important. However, the role of KLF15 in cartilage homeostasis and the progression of OA pathology has not yet been fully elucidated, and further research is required.

The Wnt/β-catenin signaling pathway is a highly conserved and fundamental signaling cascade that plays pivotal roles in several biological processes. [25] Canonical Wnt signaling through β-catenin is activated by Wnt ligand binding to the cell surface Frizzled receptor and co-receptor low-density lipoprotein receptor-related protein (LRP) 5/6. In the absence of Wnt ligands, cytoplasmic β-catenin is targeted for proteasomal degradation via a degradation complex, which phosphorylates β-catenin and marks it for ubiquitination. Upon Wnt activation, this degradation machinery is inactivated and β-catenin accumulates in the cytoplasm before translocating to the nucleus. There, Wnt/β-catenin activates the transcription of various target genes involved in cell proliferation, differentiation, and matrix remodeling. [26] The canonical Wnt/β-catenin pathway is essential for proper cartilage tissue development and maintenance under physiological conditions. Its dysregulation has been suggested to be deeply involved in OA pathology. Evidence suggests that abnormal or persistent activation of this pathway promotes the expression of degradative enzymes, such as MMP-13 and disintegrin-like metalloprotease with thrombospondin motifs-5 (ADAMTS-5), which are central mediators of ECM degradation in OA cartilage. [27–29] These enzymes degrade collagen type II alpha 1 (COL2A1) and aggrecan, resulting in the structural destruction of articular cartilage. In particular, overexpression of β-catenin has been reported to be involved in the progression of OA. [30, 31] Interestingly, recent studies suggest that transcription factors, such as KLF15, which regulate metabolic and signaling pathways in other tissues, such as the heart and kidney, may interact with Wnt/β-catenin signaling. [17, 32] These findings suggest that KLF15 may also affect cartilage integrity by regulating the expression of β-catenin in chondrocytes.

Information regarding the crosstalk between KLF15 and β-catenin in OA pathology is currently scarce. Understanding the precise nature of the involvement of KLF15 in the canonical Wnt/β-catenin pathway may help to identify novel transcriptional regulatory mechanisms in OA and aid in identifying future therapeutic targets. In this study, we aimed to evaluate how the interaction between KLF15 and β-catenin regulates OA progression using a Klf15-deficient mouse model.

Materials and methods

Generation of inducible cartilage-specific Klf15 conditional knockout mice

Mice were maintained under a 12-h light/dark cycle with free access to food and water. Klf15^fl/fl mice were generated as described previously. [33] Col2-CreERT2 mice were obtained from Jackson Laboratories (Bar Harbor, Maine, USA). Klf15^fl/fl mice were crossed with Col2-CreERT2 mice to generate inducible cartilage-specific Klf15 conditional knockout mice (Klf15 cKO). Klf15^fl/fl littermates lacking Col2-CreERT2 were used as control mice. Genotyping was performed on tail biopsy samples by polymerase chain reaction (PCR) using a DNA extraction kit (Qiagen, Valencia, California, USA). To induce gene deletion, 6-week-old Klf15 cKO and control mice were intraperitoneally injected with 4-hydroxytamoxifen (40 mg/kg) for 5 consecutive days. In this study, only male mice were used because they develop more severe OA than females. [34] All mice were backcrossed to the C57BL/6 background for more than six generations. Each experimental group included at least six mice.

In vivo OA mouse model

All animal procedures were conducted in strict accordance with the National Institutes of Health guidelines for the care and use of laboratory animals (Bethesda, MD, USA). The Animal Research Committee of Kobe University approved the protocol (approval number: P230606). Ten-week-old male Klf15 cKO and control mice were used. All surgical procedures were performed on the right knee joint under combined anesthesia (medetomidine, 0.375 mg/kg; midazolam, 2.0 mg/kg; butorphanol, 2.5 mg/kg). OA was induced by destabilization of the medial meniscus (DMM) via medial meniscotibial ligament. [35] Sham operations involved exposure without ligament transection. Mice were allowed to bear weight after recovery. Mice were euthanized for histological evaluation at 1 day, 1 week, 4 weeks, and 8 weeks post-operation or at 8 weeks post-sham. Six mice were analyzed per group (total n = 60; Klf15 cKO Sham, Klf15 cKO DMM, Control Sham, Control DMM).

Histological evaluation of cartilage degeneration

Knee joints were fixed in 10% neutral-buffered formalin for 24 h, decalcified with 4% ethylenediaminetetraacetic acid for 7 days, and embedded in paraffin. Coronal sections were cut at 80 μm intervals and stained with Safranin O and Fast Green. OA severity was evaluated using the Osteoarthritis Research Society International (OARSI) cartilage histopathology scoring system.

Immunohistochemistry

Deparaffinized sections were digested for 10 min with proteinase (Dako, Glostrup, Denmark) and treated with 3% hydrogen peroxide (Wako Pure Chemical Industries, Osaka, Japan) to block endogenous peroxidase activity. Sections were incubated overnight at 4 °C with the following primary antibodies: anti-β-catenin (1:5000, bs-0530R, Abcam, Cambridge, UK), anti-ADAMTS-5 (1:50, GTX100332, GeneTex, Irvine, CA, USA), anti-MMP-13 (1:50, GTX100665, GeneTex), and anti-SOX9 (1:200, ab185230, Abcam, Cambridge, UK). Secondary antibody incubation was performed using peroxidase-labeled antirabbit immunoglobulin (Histofine Simple Stain MAX PO; Nichirei Biosciences, Tokyo, Japan) at 22 ± 2℃ for 30 min, followed by staining with 3,3’-diaminobenzidine (DAB; Histofine Simple Stain DAB kit; Nichirei Biosciences, Tokyo, Japan). One coronal section at the center of the most severe OA lesion of each tibial plateau was analyzed. The number of positively stained cells in both superficial and deep zones of cartilage was counted in three high-power fields by three blinded observers. The mean percentage of positive cells to total cells was calculated for each marker per group (n = 6). Cells above the tidemark were considered positive.

Isolation of mouse cartilage and chondrocytes

Cartilage tissues were isolated from fetal rib cages. Klf15^fl/fl Cre mice were mated, and pregnant female mice were intraperitoneally injected with tamoxifen at 12.5 and 13.5 days post coitum (dpc). Fetuses were harvested from these mice by Caesarean section at 18.5 dpc. Genotyping was performed on tail biopsy samples by PCR using a DNA extraction kit (Qiagen, Valencia, California, USA). Primary chondrocytes were isolated from cartilage by digestion with 3 mg/mL collagenase D (Roche, Basel, Switzerland) in Dulbecco’s Modified Eagle’s Medium (DMEM; D5546, Sigma-Aldrich, St. Louis, MO, USA) at 37℃ under 5% CO₂ for 45 min. The digestion was repeated after transferring the cartilage pieces to a new dish. The resulting cells were cultured overnight in a 1:5 mix of culture medium (DMEM supplemented with 2 mM glutamine, 50 U/mL penicillin, 0.05 mg/mL streptomycin) and digestion solution. Non-adherent cells were removed, and adherent cells were further cultured with 10% fetal bovine serum (FBS; 50 mL per 500 mL DMEM) for 5–6 days until confluence. First-passage cells were used in all experiments.

Co-immunoprecipitation

Primary cultured chondrocytes from control fetal costal cartilage were used. Cells were lysed in M-PER cell lysis buffer (#78501, Thermo Fisher Scientific, Waltham, MA, USA) supplemented with protease inhibitors and phosphatase inhibitors (Roche, Basel, Switzerland). β-catenin antibody (1:5000, bs-0530R, Abcam, Cambridge, UK) was added to the cell lysate and incubated for 10 min at room temperature with rotation, followed by the addition of Dynabeads Protein A (10001D, Thermo Fisher Scientific, Waltham, MA, USA) with rotation for 1 h at 4℃ to precipitate the antigen-antibody complexes. After washing the beads twice with washing buffer, the precipitated complex was eluted, separated by SDS-PAGE, and detected by western blot with KLF15 antibody (1:500, ab2647; Abcam, Cambridge, UK). For the negative control (IP: IgG), normal rabbit IgG (same species and isotype as the primary antibody) was added to an identical amount of lysate and processed in parallel with Dynabeads Protein A under the same conditions. Whole cell lysates were also analyzed as input to confirm protein expression.

Double immunofluorescence

To confirm the colocalization of KLF15 and β-catenin in chondrocytes, fluorescent immunostaining was performed in chondrocytes from control mice (n = 5). Sections were incubated with 3% hydrogen peroxide (Wako Pure Chemical Corporation, Osaka, Japan) for 30 min. The sections were incubated overnight at 4℃ with anti-KLF15 antibody (1:500, ab2647; Abcam, Cambridge, UK) and anti-β-catenin antibody (1:5000, bs-0530R; Abcam, Cambridge, UK) as primary antibodies, and then incubated with the corresponding secondary antibodies (Alexa Fluor 488-conjugated goat anti-mouse, 1:500; A-11001, Alexa Fluor 568-conjugated rabbit anti-mouse, 1:500; A-11061, Thermo Fisher Scientific, Waltham, MA, USA) for 2 h at room temperature in the dark. For nuclear staining, 4′,6-diamidino-2-phenylindole (DAPI; Thermo Fisher Scientific, Waltham, MA, USA) was added and incubated at room temperature for 5 min. The cells were imaged using a confocal microscope. Colocalization was quantitatively analyzed using the Coloc2 plugin in ImageJ/Fiji. Pearson’s correlation coefficient and Manders’ overlap coefficients were calculated to assess the degree of colocalization.

Cytoplasmic and nuclear protein extraction

To compare β-catenin expression levels in cytoplasmic and nuclear fractions, western blot analysis was performed following subcellular fractionation. Primary chondrocytes were isolated from fetal costal cartilage of Klf15 cKO and control mice. Cytoplasmic and nuclear proteins were extracted using homemade buffers (Buffer A and Buffer C). Buffer A consisted of 10 mM HEPES (pH 7.9), 10 mM KCl, 0.1 mM EDTA, 0.1 mM EGTA, 1 mM DTT, and 0.5 mM phenylmethylsulfonyl fluoride (PMSF). Buffer C consisted of 20 mM HEPES (pH 7.9), 400 mM NaCl, 1 mM EDTA, 1 mM EGTA, 1 mM DTT, and 1 mM PMSF. Protease and phosphatase inhibitors were added immediately before use. Briefly, chondrocytes were lysed in ice-cold Buffer A for 15 min. IGEPAL CA-630 (CAS No. 9002-93-1) was then added, and the lysates were centrifuged at 12,000 rpm for 30 s at 4 °C. The resulting supernatants were collected as the cytoplasmic fractions. The remaining pellets were resuspended in ice-cold Buffer C, followed by centrifugation at 12,000 rpm for 5 min at 4 °C to obtain the nuclear fractions.

Western blot analysis

Chondrocytes (n = 4) were lysed using the NucleoSpin RNA/Protein kit (Takara Bio, Tokyo, Japan), and the total protein concentration was quantified using the Bradford method and protein assay reagent (Bio-Rad). Anti-β-catenin antibody (1:5000, bs-0530R, Abcam, Cambridge, UK) was used as the primary antibody. Horseradish peroxidase-conjugated goat anti-rabbit immunoglobulin G antibody (IgG Ab, GE Healthcare, Little Chalfont, UK) was used as the secondary antibody, and signals were visualized using ECL Plus reagent (GE Healthcare) on a Chemilumino Analyzer LAS-3000 mini (FujiFilm, Tokyo, Japan). As internal controls, rabbit polyclonal anti-Histone H3 antibody (1:8000; 17168-1-AP, Proteintech) was used for the nuclear fraction, and mouse monoclonal anti-GAPDH antibody (1:50000; 60004-1-Ig, Proteintech, Chicago, IL, USA) was used for the cytoplasmic fraction, and these were used to correct the purity and expression levels of the fractions. Protein expression was determined by semi-quantification of digitally captured images using NIH ImageJ software (http://imagej.nih.gov/ij/).

β-catenin transcriptional activity assay

To evaluate β-catenin/TCF transcriptional activity, a dual-luciferase reporter assay was performed using a TCF reporter plasmid kit (Merck Millipore, Burlington, MA, USA). Primary chondrocytes were transiently transfected with either the TOPFlash reporter plasmid, containing TCF/LEF binding sites, or the FOPFlash plasmid, containing mutated binding sites, together with a Renilla luciferase control vector (pRL-SV40; Promega, Madison, WI, USA) using Lipofectamine 3000 (Thermo Fisher Scientific, Waltham, MA, USA), according to the manufacturer’s instructions.

After 24 h of transfection, cells were stimulated with recombinant mouse Wnt3A for an additional 24 h. Luciferase activity was measured using the Dual-Luciferase^® Reporter Assay System (Promega, Madison, WI, USA). Firefly luciferase activity was normalized to Renilla luciferase activity to control for transfection efficiency. Relative transcriptional activity was expressed as normalized luciferase activity.

Quantitative RT-PCR analysis

Chondrocytes were cultured in 6-well plates for 24 h with or without stimulation by recombinant mouse Wnt3A protein (200 ng/mL, ab81484, Abcam, Cambridge, UK) and/or XAV-939 (200 ng/mL, S1180, Selleck Chemicals, Houston, TX, USA). Total RNA was extracted using QIAshredder and RNeasy Mini Kit (Qiagen, Valencia, CA, USA).

For validation in adult cartilage, primary articular chondrocytes were additionally isolated from the knee joints of adult mice. Cartilage tissues were dissected from the femoral condyles and tibial plateaus, followed by enzymatic digestion to obtain chondrocytes. These cells were cultured and analyzed under the same conditions as described above. Due to the limited yield of adult articular chondrocytes, a limited number of samples were analyzed.

In detail, 1 µg of total RNA was reverse transcribed into first-strand cDNA using 1.25 µM oligo dT primer in 40 µL PCR buffer II containing 2.5 mM MgCl2, 0.5 mM dNTP mix, 0.5 U RNase inhibitor, and 1.25 U MuLV reverse transcriptase (PerkinElmer, Foster City, CA, USA) for 60 min at 42℃. Relative expression of β-catenin, Axin2, Adamts5, Mmp13, Sox9, and the gene Col2a1 was quantified by SYBR Green real-time PCR using an ABI Prism 7500 system (Applied Biosystems, Foster City, CA, USA). The relative expression of the target genes was normalized to that of the Gapdh housekeeping gene using the comparative cycle threshold (Ct) value method. The difference in the mean Ct value of the target gene and the housekeeping gene is shown as ΔCt, and the difference between ΔCt and the Ct value of the calibration sample is shown as ΔΔCt. Log2(ΔΔCt) represented the relative value of gene expression. Primer sequences are listed in Supplementary Table S1.

Statistical analyses

For comparisons among time points within the same group, the Friedman test followed by the Wilcoxon signed-rank test was used. Mann–Whitney U tests were used for comparisons between two groups. Statistical analyses were performed using EZR (Easy R), a graphical user interface for R (The R Foundation for Statistical Computing, Vienna, Austria). Results were expressed as means with 95% confidence intervals (CI). A p-value < 0.05 was considered statistically significant.

Results

Klf15 cKO mice are more susceptible to OA-related changes in vivo

At 8 weeks postoperatively, Klf15 cKO mice exhibited greater cartilage erosion and loss of Safranin O staining compared with control mice (Fig. 1A). According to the OARSI cartilage OA histopathology scoring system, the mean total scores in both Klf15 cKO and control mice were significantly higher at 4 and 8 weeks after DMM surgery compared with sham-operated mice (Fig. 1B; Klf15 cKO mice: p < 0.001 and p < 0.001, respectively; control mice: p < 0.001 and p < 0.001, respectively). The mean total score was significantly higher in Klf15 cKO mice than in control mice at both 4 and 8 weeks postoperatively (p < 0.001 and p < 0.001, respectively).

Fig. 1.

Cartilage-specific Klf15 deficiency accelerates OA progression after DMM surgery. A Knee tissue samples were collected postoperatively from the right tibia of Klf15 cKO and control mice at various time points after DMM and sham surgery. Sections were stained with Safranin O and fast green. Black arrows indicate areas of cartilage erosion. Scale bar: 100 μm. B OARSI scores of Klf15 cKO and control mice at the indicated time points following DMM or sham surgery. Data are presented as box-and-whisker plots, with each dot representing an individual mouse (n = 6 per group at each time point). For comparisons among time points within each genotype, the Friedman test followed by the Wilcoxon signed-rank test was used. Comparisons between genotypes at each time point were performed using the Mann–Whitney U test. *p < 0.05 indicates a significant difference between the indicated groups. Abbreviations: p.o.: Postoperative, Klf15: Krüppel-like factor 15, cKO: Conditional knockout, OA: Osteoarthritis, OARSI: Osteoarthritis Research Society International, DMM: Destabilization of medial meniscus surgery

Klf15 deficiency is associated with increased β-catenin expression and OA-related changes

In Klf15 cKO mice, β-catenin expression was increased compared to control mice (Fig. 2A). The positive cell ratio of β-catenin was significantly increased in Klf15 cKO mice compared with control mice at 4 and 8 weeks after surgery (p = 0.026 and p = 0.026, respectively; Fig. 2B). It was also significantly increased in both Klf15 cKO and control mice at 4 and 8 weeks after DMM surgery compared with sham-operated mice (Fig. 2B, p = 0.009 and p = 0.002, respectively, in Klf15 cKO mice, and p = 0.041 and p = 0.002, respectively, in control mice).

Fig. 2.

β-catenin expression is increased in cartilage of Klf15 cKO mice. A Immunohistochemical staining with anti-β-catenin antibody. Scale bar: 100 μm. B Quantitative data of β-catenin-positive cells in each group. Data are shown as box plots with whiskers. Each circle represents an individual mouse (n = 6 per group). For comparisons among time points within each genotype, the Friedman test followed by the Wilcoxon signed-rank test was used. Comparisons between genotypes at each time point were performed using the Mann–Whitney U test. *p < 0.05 indicates a significant difference between the indicated groups. Abbreviations: p.o.: Postoperative, Klf15: Krüppel-like factor 15, cKO: Conditional knockout

Klf15 cKO mice exhibit increased catabolic factor expression in cartilage

Klf15 cKO mice showed higher MMP-13 and ADAMTS-5 expression compared with control mice (Fig. 3A, B). In Klf15 cKO mice, the positive cell ratio of MMP-13 was significantly increased at 4 and 8 weeks after surgery, and the positive cell ratio of ADAMTS-5 was significantly increased at 1, 4, and 8 weeks after surgery (Fig. 3D, E; MMP-13: p = 0.030 and p = 0.030, respectively; ADAMTS-5: p = 0.004, p = 0.002, and p = 0.002, respectively). Compared to sham mice, the positive cell ratio of MMP-13 was significantly increased in Klf15 cKO mice at 4 and 8 weeks after surgery and in control mice at 8 weeks after surgery (Fig. 3D; Klf15 cKO: p = 0.030 and p = 0.002, respectively; control: p = 0.004). The positive cell ratio of ADAMTS-5 was significantly increased in Klf15 cKO mice at 1, 4, and 8 weeks and was similarly increased in control mice at 8 weeks after DMM surgery compared with sham mice (Fig. 3E; Klf15 cKO: p = 0.027, p < 0.001, and p < 0.001, respectively; control: p = 0.004).

Fig. 3.

Klf15 deficiency increases catabolic factor expression and decreases anabolic factor expression in articular cartilage. A, B, C Immunohistochemical staining with antibodies against MMP-13, ADAMTS-5, and SOX9. Scale bar: 100 μm. D, E, F Quantitative data of MMP-13-, ADAMTS-5-, and SOX9-positive cells in each group. Data are shown as box plots with whiskers. Each circle represents an individual mouse (n = 6 per group). For comparisons among time points within each genotype, the Friedman test followed by the Wilcoxon signed-rank test was used. Comparisons between genotypes at each time point were performed using the Mann–Whitney U test. *p < 0.05 indicates a significant difference between the indicated groups. Abbreviations: p.o.: Postoperative, Klf15: Krüppel-like factor 15, MMP-13: Matrix metalloprotease-13, ADAMTS-5: A disintegrin-like metalloprotease with thrombospondin motifs-5, SOX9: SRY-box transcription factor 9

Klf15 cKO mice exhibit decreased SOX9 expression, contributing to OA progression

SOX9 expression was reduced in Klf15 cKO mice compared with that in control mice (Fig. 3C). The positive cell ratio of SOX9 was significantly decreased in Klf15 cKO mice compared with control mice at all observation time points (Fig. 3F; p = 0.009, p = 0.009, p = 0.026, p = 0.009, and p = 0.015, respectively). The positive cell ratio of SOX9 was significantly decreased in both Klf15 cKO and control mice compared to sham mice at 4 and 8 weeks after DMM surgery (Klf15 cKO: p = 0.001 and p < 0.001, respectively; control: p = 0.002 and p < 0.001, respectively).

KLF15 interacts with β-catenin in the nucleus of chondrocytes, while KLF15 deficiency activates the canonical Wnt/β-catenin pathway

Co-immunoprecipitation was performed in chondrocytes from control mice to examine whether KLF15 interacts directly with β-catenin. Chondrocyte cell lysates were immunoprecipitated with β-catenin, and KLF15 was detected by western blotting. A clear band of the expected size for KLF15 was detected, confirming the interaction between KLF15 and β-catenin (Fig. 4A).

Fig. 4.

KLF15 interacts with β-catenin and regulates its nuclear localization and transcriptional activity in chondrocytes. A Co-immunoprecipitation using chondrocytes from control mice. Chondrocyte lysates were subjected to immunoprecipitation with anti–β-catenin antibody (lane 2, IP: β-catenin) or control IgG (lane 3, IP: IgG), followed by immunoblotting with anti-KLF15 antibody. Lane 1 (Input) represents 5% of the total cell lysate before immunoprecipitation and serves as a positive control for KLF15 expression, whereas the IgG lane serves as a negative control to confirm the specificity of the KLF15–β-catenin interaction. Immunoblotting for β-catenin was also performed to confirm successful immunoprecipitation. B Immunofluorescence staining of KLF15 and β-catenin in chondrocytes. KLF15 was stained with Alexa Fluor 488 (green), β-catenin with Alexa Fluor 594 (red), and nuclei were counterstained with DAPI (blue). Co-localization is indicated by yellow circles in merged images. White arrows indicate the regions shown at higher magnification in the inset images. Scale bar: 100 μm. C Real-time RT-PCR analyses for the mRNA expression of β-catenin and Axin2 in chondrocytes from Klf15 cKO and control mice. D, F Western blotting for β-catenin protein expression separately in the cytoplasm and nucleus. GAPDH and Histone H3 were used for normalization in the cytoplasm and nucleus, respectively. E, G Relative band intensity values are calculated as the ratio of protein of interest to that of GAPDH or Histone H3. H A dual-luciferase reporter assay was performed using TOPFlash and FOPFlash constructs. Firefly luciferase activity was normalized to Renilla luciferase activity and expressed relative to the control Wnt- group (set as 100%) in each assay. Data are presented as box plots with whiskers. Each circle represents an individual trial (RT-PCR: n = 6; western blotting results: n = 4; A dual-luciferase reporter assay: n = 6). The Friedman test followed by the Wilcoxon signed-rank test was performed to compare the ratings for each time point, while the Mann–Whitney U test was performed to compare the groups in each assessment. *p < 0.05 indicates a significant differences between the indicated groups. Abbreviations: KLF15: Krüppel-like factor 15, DAPI: 4′,6-diamidino-2-phenylindole, RT-PCR: Reverse-transcription polymerase chain reaction, cKO: Conditional knockout, GAPDH: Glyceraldehyde-3-phosphate dehydrogenase

Immunofluorescence staining revealed that KLF15 and β-catenin stained more intensely at the same position in the nucleus, confirming that KLF15 and β-catenin colocalize in chondrocyte nuclei (Fig. 4B). Quantitative colocalization analysis demonstrated a Pearson’s correlation coefficient of 0.59, along with Manders’ overlap coefficients of 0.41 and 0.40, indicating moderate nuclear colocalization (Supplementary Table 2).

Real-time RT-PCR analysis showed that β-catenin and Axin2 mRNA expression in chondrocytes derived from Klf15 cKO mice was significantly higher than that in control mice (Fig. 4C, p = 0.003 and p = 0.003, respectively).

Klf15 cKO and control mouse chondrocytes were separated by cell fractionation, and the expression levels of β-catenin protein in the cytoplasm and nucleus were detected by western blotting. There was no significant difference in the expression level of β-catenin in the cytoplasm between Klf15 cKO and control mice (Fig. 4D and E: p = 0.702). The expression level of β-catenin in the nucleus was significantly higher in Klf15 cKO mice than in control mice (Fig. 4F and G: p < 0.001).

To directly assess β-catenin/TCF transcriptional activity, a dual-luciferase reporter assay was performed. Wnt3A stimulation significantly increased TOPFlash activity in both groups (Fig. 4H: control, p = 0.005; Klf15 cKO, p = 0.002), with a greater increase observed in Klf15 cKO mice compared with control mice (Fig. 4H: p = 0.005).

In contrast, FOPFlash activity showed no significant differences among groups, confirming the specificity of β-catenin/TCF-dependent transcriptional activation (Supplemental Fig. 1).

Wnt3A-induced changes in gene expression are enhanced in Klf15 cKO chondrocytes and suppressed by XAV939

Real-time RT-PCR analysis showed that Wnt3A stimulation significantly increased the expression of β-catenin, Axin2, Mmp13, and Adamts5 in both Klf15 cKO and control mice (Fig. 5A-D; β-catenin, p < 0.001 and p < 0.001, respectively; Axin2, p < 0.001 and p < 0.001, respectively; Mmp13, p = 0.003 and p = 0.001, respectively; Adamts5, p < 0.001 and p < 0.001, respectively). Furthermore, the increase in expression of these four genes by Wnt3A stimulation was suppressed by XAV939, a tankyrase inhibitor that stabilizes AXIN and inhibits canonical Wnt/β-catenin signaling, treatment in both Klf15 cKO and control mice, and their expression was significantly decreased (Fig. 5A-D; β-catenin, p < 0.001 and p < 0.001, Axin2, p < 0.001 and p < 0.001; Mmp13, p = 0.005 and p = 0.005, respectively; Adamts5, p < 0.001 and p < 0.001, respectively). Wnt3A stimulation significantly decreased the expression of Sos9 and Col2a1 in both Klf15 cKO and control mice (Fig. 5E, F; all p < 0.001). Furthermore, the decrease in the expression of these two genes by Wnt3A stimulation was suppressed by XAV939 administration in both Klf15 cKO and control mice, and their expression was significantly increased (Fig. 5E, F; Sox9, p < 0.001 and p < 0.001, respectively; Col2a1, p < 0.001 and p < 0.001, respectively). Wnt3A stimulation significantly increased the expression of β-catenin, Axin2, and Adamts5 in Klf15 cKO mice compared to control mice (Fig. 5A, B, D; p < 0.001, p < 0.001, and p < 0.001, respectively). Conversely, the expression of Sox9 and Col2a1 was significantly decreased (Fig. 5E, F; p < 0.001 and p = 0.012, respectively). In addition, under basal conditions without Wnt3A stimulation, XAV939 significantly decreased β-catenin expression and increased Col2a1 expression in Klf15 cKO chondrocytes, suggesting partial reversal of the Klf15-deficient phenotype (Fig. 5A, F; p = 0.001 and p < 0.001, respectively). These findings were consistent with the in vivo results. A similar trend in gene expression was observed in primary chondrocytes isolated from adult articular cartilage (Supplementary Fig. 2), although no statistically significant differences were detected between groups, likely due to the limited sample size.

Fig. 5.

Enhancement of Wnt/β-catenin signaling, catabolic factor amplification, and anabolic factor suppression induced by Wnt3A stimulation were canceled by treatment with a Wnt/β-catenin signaling inhibitor in chondrocytes from Klf15 cKO and control mice. A-F Real-time RT-PCR analyses for the mRNA expression of β-catenin, Axin2, Mmp13, Adamts5, Sox9, and Col2a1 in chondrocytes from Klf15 cKO and control mice treated with or without Wnt3A (200 ng/mL) or canonical Wnt/β-catenin signaling inhibitor (XAV-939, 200 ng/mL). Chondrocytes from control mice not treated with either Wnt3A or canonical Wnt/β-catenin signaling inhibitor were used as controls. Data are presented as box plots with whiskers. Each circle represents an individual trial (RT-PCR: n = 6). The Friedman test followed by the Wilcoxon signed-rank test was performed to compare the ratings for each time point, while the Mann–Whitney U test was performed to compare the groups in each assessment. *p < 0.05 indicates a significant differences between the indicated groups. Abbreviations: Klf15: Krüppel-like factor 15, RT-PCR: Reverse-transcription polymerase chain reaction, cKO: Conditional knockout, Mmp13: Matrix metalloprotease 13, Adamts5: A disintegrin-like metalloprotease with thrombospondin motifs 5, Sox9: SRY-box transcription factor 9, Col2a1: Collagen type 2 alpha 1

Discussion

This study presents the first evidence that KLF15 negatively regulates canonical Wnt/β-catenin signaling in cartilage and that it may play an important role in OA progression. Specifically, KLF15 suppresses β-catenin activity and helps to maintain cartilage matrix homeostasis in both mouse models and fetal costal chondrocytes in vitro.

Wnt ligand binding to Frizzled receptors and LRP5/6 inhibits β-catenin degradation, allowing it to accumulate in the cytoplasm and translocate into the nucleus, activating TCF/LEF-dependent transcription. [36, 37] Excessive activation of the canonical Wnt/β-catenin signaling pathway induces the expression of matrix-degrading enzymes, such as MMP-13 and ADAMTS-5, promoting OA pathology. [30] Experimental β-catenin overactivation in chondrocytes produces OA-like changes including osteophyte formation, cartilage thinning, and joint space narrowing. This suggests that β-catenin activation directly promotes cartilage destruction. [31].

Indeed, the nuclear localization of β-catenin and the expression of its target genes are increased in the articular cartilage of patients with OA. [38] In non-cartilage tissues, KLF15 has been shown to inhibit canonical Wnt/β-catenin signaling by binding to β-catenin and promoting its proteasomal degradation. [39].

KLF15 induction has also been shown to suppress canonical Wnt/β-catenin signaling by binding to active β-catenin and inhibiting its phosphorylation in renal interstitial cells. [40] Furthermore, KLF15 may directly suppress β-catenin transcriptional activity by physically binding to β-catenin and forming a protein complex. This is supported by the increased β-catenin/TCF transcriptional activity observed in Klf15-deficient chondrocytes.

In addition to its direct interaction with β-catenin, KLF15 may regulate cartilage homeostasis through metabolic pathways. KLF15 is a key regulator of cellular metabolism, and metabolic dysregulation has been implicated in osteoarthritis progression. Recent evidence suggests that metabolic status can influence Wnt/β-catenin signaling. [27, 29] Thus, Klf15 deficiency may promote β-catenin activation not only through direct interaction but also by altering the metabolic environment of chondrocytes. This dual mechanism may contribute to the enhanced β-catenin activity observed in Klf15-deficient chondrocytes.

Klf15 deficiency enhanced β-catenin activity and increased expression of MMP-13 and ADAMTS-5 while reducing SOX9 and COL2A1, indicating a shift toward a catabolic state. [41] These changes were further intensified by Wnt3a stimulation and partially rescued by XAV939. Consistent with our findings, previous studies have shown that β-catenin overactivation leads to structural abnormalities, not only in articular cartilage but also in intervertebral disc tissue, further supporting the destructive potential of this pathway in cartilage structure. [42] XAV-939, a well-characterized small molecule inhibitor, thereby forming a degradation complex and promoting the phosphorylation-dependent degradation of β-catenin. [43] In a previous study, intra-articular injection of XAV939 in a DMM mouse model attenuated the severity of OA and demonstrated anti-catabolic effects in chondrocytes. [44].

The results of this study indicate that Klf15 deficiency induces a catabolic environment via excessive activation of canonical Wnt/β-catenin signaling, thereby promoting cartilage destruction. Although the sample size was limited, similar trends were observed in primary chondrocytes derived from adult articular cartilage, supporting the relevance of our findings beyond fetal chondrocytes. Furthermore, inhibition of the canonical Wnt/β-catenin signaling pathway suppresses catabolic factor expression, indicating that OA-like changes associated with Klf15 deficiency occur via this pathway.

In addition, reduced anabolic factor expression in response to Klf15 deficiency indicates that the ability of chondrocytes to maintain differentiation and synthesize the matrix was significantly impaired. This reduction was partially restored by XAV939 administration, revealing that abnormal activation of the canonical Wnt/β-catenin pathway is also involved in the breakdown of the anabolic environment. Therefore, KLF15 may suppress OA progression by interacting with and negatively regulating β-catenin in the cell nucleus, thereby suppressing the expression of cartilage-degrading enzymes, and maintaining an anabolic state (Fig. 6).

Fig. 6.

Schematic model of KLF15-mediated regulation of canonical Wnt/β-catenin signaling in chondrocytes.Under normal conditions, KLF15 interacts with β-catenin in the nucleus and suppresses its transcriptional activity, thereby maintaining cartilage homeostasis by promoting anabolic factors and suppressing catabolic enzymes. In contrast, Klf15 deficiency leads to enhanced nuclear accumulation of β-catenin and activation of Wnt/β-catenin signaling, resulting in increased expression of catabolic factors and decreased expression of anabolic factors. Arrows indicate activation, blunt-ended lines indicate inhibition, red arrows indicate upregulation, and blue arrows indicate downregulation. Abbreviations: KLF15: Krüppel-like factor 15

Limitations

This study has some limitations. First, Klf15 was specifically knocked out in cartilage tissue, so the environment may be different from the real-life disease environment. This is because OA is a complex disease that progresses with degeneration of synovium, bone, and meniscus in addition to cartilage tissue. [45, 46] Second, fetal costal chondrocytes were used in vitro, not knee articular cartilage. Ideally, knee chondrocytes from adult mice should have been used. However, we were unable to obtain the estimated number of knee chondrocytes required for this experiment.

The findings of this study suggest that KLF15 is a potential therapeutic target for OA. If the canonical Wnt/β-catenin pathway can be selectively suppressed by interventions that activate or restore KLF15 function, it may be possible to slow down or even prevent the destruction of articular cartilage and maintain cartilage differentiation. In addition, the development of small-molecule compounds that enhance the nuclear interaction between KLF15 and β-catenin and the restoration of KLF15 function by gene transfer should be considered as future approaches for treatment. However, because KLF15 is also involved in glucose metabolism and adipogenesis, its usefulness as a therapeutic target must be assessed in the context of its systemic effects.

Authors’ contributions

Akira Saitoh (Acquisition of the data, Analysis and interpretation of the data, Drafting of the article, Final approval of the article), Shinya Hayashi (Conception and design, Analysis and interpretation of the data, Drafting of the article, Obtaining of funding, Final approval of the article), Toshiki Kitamura (Analysis and interpretation of the data, Critical revision of the article for important intellectual content, Final approval of the article), Takuma Hayashi (Analysis and interpretation of the data, Critical revision of the article for important intellectual content, Final approval of the article), Kohei Motono (Analysis and interpretation of the data, Critical revision of the article for important intellectual content, Final approval of the article), Shotaro Araki (Analysis and interpretation of the data, Critical revision of the article for important intellectual content, Final approval of the article), Takuma Maeda (Analysis and interpretation of the data, Critical revision of the article for important intellectual content, Final approval of the article), Yoshihito Suda (Analysis and interpretation of the data, Critical revision of the article for important intellectual content, Final approval of the article), Kensuke Wada (Analysis and interpretation of the data, Critical revision of the article for important intellectual content, Final approval of the article), Kemmei Ikuta (Analysis and interpretation of the data, Critical revision of the article for important intellectual content, Final approval of the article), Masanori Tsubosaka (Analysis and interpretation of the data, Critical revision of the article for important intellectual content, Final approval of the article), Tomoyuki Kamenaga (Analysis and interpretation of the data, Critical revision of the article for important intellectual content, Final approval of the article), Yuichi Kuroda (Analysis and interpretation of the data, Critical revision of the article for important intellectual content, Final approval of the article), Naoki Nakano (Analysis and interpretation of the data, Critical revision of the article for important intellectual content, Final approval of the article), Tomoyuki Matsumoto (Analysis and interpretation of the data, Critical revision of the article for important intellectual content, Final approval of the article), Tetsuya Hosooka (Analysis and interpretation of the data, Critical revision of the article for important intellectual content, Final approval of the article), Wataru Ogawa (Analysis and interpretation of the data, Critical revision of the article for important intellectual content, Final approval of the article), Ryosuke Kuroda (Analysis and interpretation of the data, Critical revision of the article for important intellectual content, Final approval of the article).

Funding

This study was funded by the Grant-in-Aid for Scientific Research (21K09278).

Data availability

The datasets generated and/or analyzed during the current study are available from the corresponding author on reasonable request.

Declarations

Ethics approval and consent to participants

All animal experiments were approved by the Animal Experimentation Committee of Kobe University (Approval No. P230606) and conducted in accordance with institutional guidelines and the ARRIVE reporting standards. All efforts were made to minimize animal suffering and to reduce the number of animals used.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.