Senolytic therapy combining Dasatinib and Quercetin restores the chondrogenic phenotype of human osteoarthritic chondrocytes by the release of pro‐anabolic mediators

Abstract

Cellular senescence is associated with various age‐related disorders and is assumed to play a major role in the pathogenesis of osteoarthritis (OA). Based on this, we tested a senolytic combination therapy using Dasatinib (D) and Quercetin (Q) on aged isolated human articular chondrocytes (hACs), as well as in OA‐affected cartilage tissue (OARSI grade 1–2). Stimulation with D + Q selectively eliminated senescent cells in both, cartilage explants and isolated hAC. Furthermore, the therapy significantly promoted chondroanabolism, as demonstrated by increased gene expression levels of COL2A1, ACAN, and SOX9, as well as elevated collagen type II and glycosaminoglycan biosynthesis. Additionally, D + Q treatment significantly reduced the release of SASP factors (IL6, CXCL1). RNA sequencing analysis revealed an upregulation of the anabolic factors, inter alia, FGF18, IGF1, and TGFB2, as well as inhibitory effects on cytokines and the YAP‐1 signaling pathway, explaining the underlying mechanism of the chondroanabolic promotion upon senolytic treatment. Accordingly, stimulation of untreated hAC with conditioned medium of D + Q‐treated cells similarly induced the expression of chondrogenic markers. Detailed analyses demonstrated that chondroanabolic effects could be mainly attributed to Dasatinib, while monotherapeutical application of Quercetin or Navitoclax did not promote the chondroanabolism. Overall, D + Q therapy restored the chondrogenic phenotype in OA hAC most likely by creating a pro‐chondroanabolic environment through the reduction of SASP factors and upregulation of growth factors. This senolytic approach could therefore be a promising candidate for further testing as a disease‐modifying osteoarthritis drug.

Dasatinib and Quercetin therapy restores the chondrogenic phenotype by (1) selective elimination of senescent chondrocytes and thus decreased SASP factors expression, and (2) via YAP1 suppression and induction of pro‐anabolic mediators.

ABBREVIATIONS

AAT

alpha1‐antitrypsin

ANOVA

analysis of variance

BCL

B‐cell lymphoma

CXCL

C‐X‐C motif chemokine ligand

CDM

chondrogenic differentiation medium

CML

chronic myeloid leukemia

COL2

collagen type II

CTCF

corrected total cell fluorescence

DAMPs

damage‐associated molecular patterns

D

Dasatinib

DMSO

dimethyl sulfoxide

DMOADs

disease‐modifying osteoarthritis drugs

Doxo

Doxorubicin

ELISA

enzyme‐linked immunosorbent assay

ECM

extracellular matrix

FGF

fibroblast growth factor

GAGs

glycosaminoglycans

hAC

human articular chondrocytes

IF

immunofluorescence

IHC

immunohistochemistry

IGF

insulin like growth factor

ITS

Insulin‐Transferrin‐Selenium

IL

interleukin

KEGG

Kyoto Encyclopedia of Genes and Genomes

MMP

matrixmetalloproteinase

MSCs

mesenchymal stem cells

OA

osteoarthritis

OARSI

Osteoarthritis Research Society International

Ph+ALL

philadelphia‐chromosome positive acute lymphoblastic leukemia

PI3K

phosphatidylinositol‐3 kinase

CPII

Procollagen II C‐Propeptide

PG

proteoglycan

Q

Quercetin

RT

room temperature

SASP

senescence‐associated secretory phenotype

SA‐β‐Gal

senescence‐associated β‐galactosidase activity

SDS

sodium dodecyl sulphate

SOX

SRY‐box

SIPS

stress‐induced premature senescence

TMJ

temporomandibular joint

TGF

transforming growth factor

YAP

yes‐associated protein

1. INTRODUCTION

Osteoarthritis (OA) is the most prevalent joint disease worldwide. It presents a complex disorder of the entire joint, including cartilage destruction, sclerosis of subchondral bone, formation of osteophytes, and synovitis (Jeon et al., 2018; Loeser et al., 2012; Riegger & Brenner, 2020). The clinical pattern is characterized by pain and joint stiffness (Zhang & Jordan, 2010). So far, OA is considered as an uncurable disease and the therapy is mainly limited to symptom relief (Collins et al., 2018). In recent years, much research has been conducted in developing disease modifying OA drugs (DMOADs), comprising mainly anti‐inflammatory, anti‐catabolic, and chondroanabolic strategies. Recently, cellular senescence has emerged as a potentially promising target in OA therapy (Karsdal et al., 2016; Oo et al., 2021; Oo & Hunter, 2022). Cellular senescence is characterized by an irreversible cell cycle arrest and was first discovered in fibroblasts by Hayflick and Moorhead in 1961 (Hayflick & Moorhead, 1961). Besides its relevance as a potent anticancer mechanism and its importance during embryogenesis and wound healing, cellular senescence is one of the hallmarks of aging and associated with several age‐related diseases (Childs et al., 2015; Jeon et al., 2018). In line with this, an accumulation of senescent cells was discovered in OA tissue, which might be linked to the pathogenesis of the disease. It is thought that stress factors, such as oxidative stress, release of damage associated molecular patterns (DAMPs), or chronic inflammation after joint trauma or during aging play an important role in the accumulation of senescent cells in OA, consequently referred to as stress‐induced premature senescence (SIPS) (Jeon et al., 2018; Kirsch et al., 2022; Martin et al., 2004). Senescent chondrocytes display an upregulation of antiapoptotic pathways by an increase of B‐cell lymphoma 2 and ‐XL family (BCL‐2, ‐XL), enhanced expression of the cyclin dependent kinase inhibitors p21^Cip1/WAF1 (CDKN1A) and p16^INK4a (CDKN2A), elevated activity of the senescence associated β‐galactosidase (SA‐β‐Gal), and are characterized by a so‐called senescence‐associated secretory phenotype (SASP) (Coryell et al., 2021; Kirsch et al., 2022; Loeser, 2009; McCulloch et al., 2017; Rim et al., 2020). Senescent chondrocytes secrete elevated levels of pro‐inflammatory chemokines and cytokines, such as C‐X‐C motif chemokine ligand 1 (CXCL‐1), interleukin 1 (IL‐1), IL‐6, and IL‐8, but also matrix metalloproteinases like MMP‐1, ‐3, and ‐13. These SASP factors contribute to cartilage degeneration and “inflammaging”, which can be understood as the spreading of senescence to neighboring cells due to chronic inflammation (Coryell et al., 2021; Jeon et al., 2018; Kirsch et al., 2022; McCulloch et al., 2017). Overall, senescent chondrocytes represent a potential therapeutic target. Consequently, senotherapeutics play an increasingly important role in the research field of OA. The senolytic drugs are divided into senomorphics and senolytics. While senomorphics target the secretion and action of SASP factors, senolytics eliminate senescent cells by the selective induction of apoptosis (Coryell et al., 2021; Mylonas & O'Loghlen, 2022).

Among the senolytic drugs, the combination of Dasatinib and Quercetin (D + Q) has turned out to be particularly efficient in the context of musculoskeletal diseases as proven in murine (Zhou et al., 2021) and rat models (Dai et al., 2020) of OA, bone healing (Farr et al., 2017a, 2017b; Wang et al., 2023), and disc degeneration (Novais et al., 2021). Interestingly, D + Q therapy achieved better results in combination as compared to the monotherapeutic approach (Zhu et al., 2015). Dasatinib is an already FDA‐ and EMA‐approved drug (Sprycel®) for the treatment of chronic myeloid leukemia (CML) and philadelphia‐chromosome positive acute lymphoblastic leukemia (Ph‐ + ALL) (Horinkova et al., 2019). The drug exerts its pro‐apoptotic effects through inhibition of several tyrosine kinases (Horinkova et al., 2019; Rivera‐Torres & San Jose, 2019), which contribute to the pro‐survival network in cancer and senescent cells (Kirkland & Tchkonia, 2017; Lee et al., 2017; Novais et al., 2021). Quercetin on the contrary is a flavonoid, found in fruits and vegetables, such as grapes, citrus fruits, onions, etc. and is known for its anti‐inflammatory, antioxidative, anticancer, and antimicrobial effects (Almatroodi et al., 2021; Anand David et al., 2016; Samadi et al., 2022). It exerts its pro‐apoptotic effects in senescent cells by inhibition of the phosphatidylinositide‐3‐kinase (PI3K)/Akt pathway and downregulation of BCL‐2 (Almatroodi et al., 2021; Reyes‐Farias & Carrasco‐Pozo, 2019). Although D + Q efficiently attenuated progression of OA after distraction arthroplasty in rats (Dai et al., 2020) and reduced age‐related temporomandibular joint (TMJ) degeneration in mice (Zhou et al., 2021), none of the previous studies investigated the effects of D + Q on human OA chondrocytes. Moreover, the phenotype of surviving chondrocytes after D + Q treatment has not been addressed so far. To close this gap in research, we investigated the efficacy of the combined application of the senolytic agents D + Q on OA hAC and cartilage tissue to particularly evaluate the chondrogenic characteristics after elimination of senescent cells. Further, we compared the outcome of D + Q treatment with one of the most commonly used senolytic, Navitoclax, which exerts its senolytic properties through the inhibition of BCL‐2/‐XL (Chaib et al., 2022; Soto‐Gamez & Demaria, 2017; Yang et al., 2020).

Our results indicate that the treatment with D + Q selectively eliminates senescent cells in human OA cartilage tissue and in isolated hAC. The therapy decreased the expression of SASP factors and enhanced that of growth factors insulin‐like growth factor 1 (IGF1), fibroblast growth factor 18 (FGF18), and transforming growth factor beta 2 (TGFB2). As a result, D + Q restored the chondrogenic phenotype of OA hAC as demonstrated by a remarkable increase of collagen type II (COL2) and proteoglycan (PG) production. Accordingly, the therapy also resulted in a recovery of the chondrogenic redifferentiation potential of high passage chondrocytes. While the chondroanabolic effects could be mainly attributed to Dasatinib, the senolytic Navitoclax impaired the chondrogenic phenotype of hAC. Moreover, selective elimination of senescent cells and increase of COL2 synthesis was confirmed in OA cartilage tissue. Overall, this is the first study demonstrating that D + Q therapy, and in particularly Dasatinib, stabilizes and possibly even restores the chondrogenic phenotype of human OA chondrocytes. These novel findings provide important insight about the divergent therapeutic efficacy of senolytics in OA and are of particular interest regarding future application of D + Q in clinical studies.

2. METHODS

2.1. Cultivation of human cartilage tissue and isolation of chondrocytes

Human articular cartilage tissue was obtained from donors undergoing knee replacement surgery. Informed consent was received from patients according to the regulations of the Ethics Committee of the University of Ulm (ethical approval number 353/18). Cartilage explants were created using a biopsy punch (⌀ 6 mm) from macroscopically intact tissue (Osteoarthritis Research Society International (OARSI) score ≤2) of human femoral condyles and thereafter cultured in serum‐free medium in an incubator (37°C, 5% CO₂, and 95% humidity) (Appendix S1a) (Riegger et al., 2016). Isolation of chondrocytes was performed by cutting cartilage slices from condyles, digesting with 0.2% protease (Sigma‐Aldrich) for 45 min at 37°C and further digesting overnight in 0.025% collagenase (collagenase from clostridium histolyticum, Sigma‐Aldrich) at 37°C. Afterwards, chondrocytes were filtered (cell strainer 40 μm), centrifuged, resuspended, and cultured in serum‐containing medium (Appendix S1a). Passaging was performed at approximately 80% confluence (Kirsch et al., 2022). Cells were used at low (p ≤ 2) and high passages (p ≥ 4) as indicated in the figure legends.

2.2. Porcine cartilage tissue

Porcine knee joints were obtained from a butcher (age: 4–6 months). Cartilage tissue was harvested from femoral condyles and cultured in serum‐free medium containing high glucose DMEM (Appendix S1a).

2.3. Senolytic treatment

Isolated chondrocytes were cultured in serum reduced medium (5% FBS), consisting of 1:1 serum‐containing medium and serum‐free medium (Appendix S1a), during experiments. Cells were treated either for 3 d with Dasatinib (50–500 nM) (S1021, Seleckchem) and Quercetin (5 μM) (S2391, Seleckchem) or with Navitoclax (5 μM) (S1001, Seleckchem) followed by 4 d without senolytics (d7). Untreated cells served as control. Senolytics were dissolved in dimethyl sulfoxide (DMSO), hence, a DMSO control was included. The DMSO concentration in the experiments with D + Q was 0.008%, which corresponded to the highest final concentration of DMSO in the treatment with D (500 nM) + Q (5 μM). In the experiments with Navi, the DMSO concentration was 0.05%. Medium was changed at least three times a week. Supernatants were collected at d3 and 7 and total RNA was isolated. In contrast to isolated chondrocytes, cartilage explants were cultured in serum‐free medium (Appendix S1a) during the experiments and stimulated according to the same application scheme, including additional analysis at d14.

2.4. Stimulation with SIPS‐ and SASP‐conditioned medium

hAC were treated with 0.1 μM Doxorubicin (Doxo) (Selleckchem) for 7 d to create a SIPS phenotype (Kirsch et al., 2022). Next, the supernatant of untreated and treated (SIPS‐CM) cells was collected 48 h after deprivation of Doxorubicin. Supernatants were used to stimulate hAC, in order to study paracrine effects.

Furthermore, high passage hAC were treated for 3 d with D (250 nM) + Q (5 μM) or cultured without senolytic treatment. Afterwards, cells were cultured for 2 d without the senolytics and medium was collected at d5. Cells were split and cultured with medium from previously untreated cells (SASP‐CM) or with medium from cells prior treated with D + Q (DQ‐CM). Cells cultured with fresh medium (fresh‐M) served as control. Overall, this experiment includes the following stimulation groups: group 1 (5 d w/o treatment, 3 d with fresh‐M), group 2 (5 d w/o treatment, 3 d with SASP‐CM), group 3 (5 d w/o treatment, 3 d with DQ‐CM), group 4 (3 d with D + Q treatment, 2 d w/o treatment, 3 d with fresh‐M), group 5 (3 d with D + Q treatment, 2 d w/o treatment, 3 d with SASP‐CM), and group 6 (3 d with D + Q treatment, 2 d w/o treatment, 3 d with DQ‐CM) (Figure 1h).

FIGURE 1.

Analysis of hACs' characteristic SASP factors and paracrine effects after D + Q treatment. (a, b) mRNA levels of CDKN1A and ‐2A after stimulation with Doxo for 7d (n = 4). (c) alamarBlue assay: Relative FIs after stimulation of hAC with SIPS‐CM (n = 3). Gene expression analysis (qRT‐PCR) of (d) CDKN1A (n = 4), (e) ‐2A (n = 4), (f) ACAN (n = 4), and (g) COL2A1 (n = 4) after stimulation with SIPS‐CM. (h) Experimental setup and (i) relative mRNA levels (qRT‐PCR) of COL2A1 (n = 4) after stimulation of hACs with DQ‐CM or SASP‐CM. (j, k) Quantification of secreted CXCL1 (n = 4), (l, m) IL6 (n = 4) and (n, o) IL8 (n = 4) at d3 and 7 by means of an ELISA. Blue colored box plots = gene expression analysis data; green colored box plots = data of protein analysis. Statistical analysis: (a) + (c) unpaired t test: *p < 0.05, **p < 0.01; (i) + (l, m) one‐way ANOVA with Šidák's multiple comparisons test: *p < 0.05, **p < 0.01, ****p < 0.0001.

2.5. ATDC5 murine chondrogenic progenitor cell line

ATDC5 cells were cultured in serum containing medium (Appendix S1a) (Tschaffon et al., 2022). During the experiments, cells were either cultured in chondrogenic differentiation medium (CDM) (Appendix S1a) for 7 d, or in serum‐reduced medium with D (250 nM) + Q (5 μM) for 3 d followed by 4 d without senolytics (d7). All analyses were performed at d7, and the medium was changed at least three times a week.

2.6. Cytotoxicity/ cell proliferation assay—alamarBlue™

This cell viability assay is based on the ability of healthy cells to metabolize the blue dye resazurin into the red fluorescent resorufin. After treatment with the senolytics, cells were treated with 5% Resazurin solution for 3 h at 37°C. Afterwards, the fluorescence intensity (FI) was measured using a microplate reader (Tecan, Microplate Reader infinite M200 pro) with an excitation (ex) wavelength of 550 nm and emission (em) detection at 590 nm (Kirsch et al., 2022).

2.7. Caspase3/7 activity assay

Caspase‐3 activity as an important initiator of apoptosis was assessed by the Amplite™ Fluorimetric Caspase 3/7 Assay Kit (AAT Bioquest) and performed according to the manufacturer's instruction. Cells were stimulated for 6 h with Navi (2.5–5 μM) and treated for 2 h with the Caspase 3/7 working solution at room temperature (RT) in the dark. Afterwards, the plate was centrifuged and fluorescence was measured with a microplate reader (ex = 350 nm, em = 450 nm) (Kirsch et al., 2022).

2.8. Cell viability assay—Live/dead staining

To determine cell viability in cartilage explants, the LIVE/DEAD™ viability/cytotoxicity kit for mammalian cells (Invitrogen) was used. Live cells were stained green with calcein AM (1 μM), while dead cells were stained red with ethidium homodimer‐1 (2 μM). In short, staining of 0.5 mm thick cartilage slices for 30 min and microscopic image taking by means of a z‐stack (Zeiss, Microscope Axioskop 2 mot plus, Software AxioVision) (Riegger et al., 2016).

2.9. Gene expression analysis—qPCR

Total RNA was isolated from hAC using the RNeasy Mini Kit (Qiagen) and afterwards reverse transcribed to cDNA by means of the Superscript II Reverse Transcriptase (Invitrogen). Gene expression levels were determined by quantitative real‐time polymerase chain reaction (RT‐PCR) (StepOnePlus™ Real‐Time PCR System; Applied Biosystems) with the 2^−ΔΔCT method (Riegger et al., 2016). GADPH, HPRT1 and in case of hAC additionally TMEM199 was used as reference genes. Target genes were analyzed by the SYBR Green/TaqMan™ Gene Expression Assay (Appendix S1b).

2.10. RNA‐sequencing analysis

Total RNA isolated from high passage hAC of three donors w/o or with 250 nM D + 5 μM Q therapy for 3 d was sent to Biomarker Technologies (BMK) GmBH (Münster, Germany) for mRNA sequencing analysis. Bioinformatic analysis and generation of graphics were performed by the company.

2.11. Supernatant analysis—ELISA

Supernatants of cultured hAC and cartilage tissue were harvested and analyzed by means of enzyme‐linked immunosorbent assays (ELISAs). SASP factors: human IL‐8 Uncoated ELISA (88–8086‐88, Invitrogen), human IL‐6 Uncoated ELISA (88–7066‐88, Invitrogen), and human GRO alpha (CXCL1) Uncoated ELISA (88‐52122‐22, Invitrogen). Collagen type II synthesis: Procollagen II C‐Propeptide (CPII) ELISA (60‐1003‐001, IBEX Pharmaceuticals Inc.). All steps were performed according to the manufacturers' instruction (Kirsch et al., 2022; Riegger et al., 2022). As treatment with senolytics eliminates senescent cells, raw ELISA data of the SASP factors were normalized to the alamarBlue values.

2.12. Immunofluorescence (IF) and phalloidin staining

hAC were fixed with 4% formalin and blocked for 1 h at 37°C (Protein‐Block serum free, X0909 Dako Agilent). COL2 antibody (MA5‐13026, Invitrogen) was diluted 1:100 (Antibody Diluent, S0809 Dako Agilent) and cells were incubated for 2 h at RT. Afterwards, secondary biotinylated antibody (biotinylated link, K0690 Dako Agilent) and the corresponding dye iFluor™568‐streptavidin (AAT Bioquest) were used. Nuclei staining was performed using Dapi (0.25 μg/μL). IF staining was completed by mounting the cells (Fluorsave™ Reagent, Merck) and capturing images under a fluorescent microscope (Zeiss, Microscope Axioskop 2 mot plus). FIs were analyzed using the software ImageJ2 (version 2.3.0/1.53q) with a modified corrected total cell fluorescence (CTCF) method, taking into account the cell number (CTCF = (integrated density [image] – (area [image]*mean background fluorescence))/cell number). In case of Ki‐67 staining, cells were after fixation permeabilized with 0.1% triton‐x‐100 for 15 min at RT, blocked for 1 h and the overnight staining was performed with the primary antibody (ab16667, dilution 1:250) at 4°C. Continuation with secondary antibody Alexa Fluor488 (ab15077) for 30 min at RT, and Dapi staining for 15 min at RT. For phalloidin staining, hAC were fixed with paraformaldehyde (15 min, RT), permeabilized with 1% triton‐x‐100 (30 min, RT) and afterwards stained with the phalloidin antibody (abcam, ab235137, 1:1000) and Dapi (15 min, RT).

2.13. Alcian blue staining

Glycosaminoglycans (GAGs) were stained by using alcian blue. hAC were fixed with 4% formalin followed by incubation with 0.1 M HCL (Merck) for 10 min. Subsequently, staining was performed with 1% alcian blue solution (Roth) for 30 min during shaking. Cells were washed three times with 0.1 M HCL, and images were taken by means of a light microscope (Zeiss, Axiovert 35). In case of ATDC5 cells, destaining with 1% sodium dodecyl sulfate (SDS) (Sigma Aldrich) was performed and absorbance was measured at 595 nm.

2.14. SA‐β‐gal activity assay

For measuring the SA‐β‐Gal activity, the 96‐Well Cellular Senescence Assay Kit (Cell Biolabs) was used. In short, 5000 cells were seeded onto the wells. After 24 h, cell lysis was performed using the cell lysis buffer. Subsequently, cell lysates were centrifuged and the supernatant was collected, and incubated with the assay buffer at 37°C for 2 h. Stopping solution was used for termination of the reaction and fluorescence intensity was measured (ex = 360 nm; em = 465 nm) (Riegger & Brenner, 2023).

2.15. Chondrogenic redifferentiation of hAC

High passage hAC were treated with the senolytics for 3 d and subsequently cultured in a 3D pellet culture (3.5 × 10⁵ cells) for 28 d in CDM. During the first 7 d, the growth factors rhBMP‐6 and rhTGF‐β3 were added freshly every day and afterwards with medium change thrice a week (Riegger, Palm, & Brenner, 2018). Chondrogenic differentiation potential was assessed by an adopted scoring system as previously described, including the pellet size, staining intensities of safranin‐O and COL2 (immunohistochemistry (IHC)), cell morphology and the produced matrix (cell–cell distance) (Ruths et al., 2024).

2.16. Histology and immunohistochemistry

Cartilage explants and chondrocyte pellets were fixed with 4% formalin, dehydrated, embedded in paraffine, and cut in 3.5 μm thick sections to perform IHC and safranin‐O staining. Briefly, samples were dewaxed and rehydrated, Weigert's iron hematoxylin was used for nuclear staining, safranin‐O (Chroma‐Gesellschaft) for staining of PGs and fast‐green (Sigma Chemical Company) as a counterstaining. For the COL2‐IHC staining, samples were as well dewaxed, rehydrated and then digested for 40 min at 37°C with pepsin (Sigma Aldrich). In case of Ki67 and CDKN2A, samples were incubated in citrate buffer overnight for antigen‐retrieval. 3% H₂0₂ treatment was performed for blocking of endogenous peroxidase and further protein blocking with Dako‐blocking buffer was performed. Next, samples were incubated overnight with the corresponding primary antibodies (COL2‐AB 1:2000, DSHB II‐II6B3‐s; Ki‐67‐AB 1:200, ab16667; CDKN2A‐AB 1:200, LS‐B1347) at 4°C. Staining was continued using the Dako LSAB2 System‐HRP kit (Dako Agilent) and Mayer's hematoxylin (Merck) solution was used for nuclei staining (Riegger, Palm, & Brenner, 2018).

2.17. Statistical analysis

Data sets were analyzed using GraphPad Prism 10 software. Outliers were quantified for data sets n ≥ 5 by means of a Grubbs' test and excluded from statistical analysis. The comparison of two groups with each other was performed using a t test, whereas a multiple t test, one‐way or two‐way analysis of variance (ANOVA) was performed for the comparison of several groups (significance level α = 0.05). Detailed statistical information can be obtained from the corresponding figure captions.

3. RESULTS

3.1. D + Q therapy exhibits cytotoxic effect on high‐passage hAC

To investigate senolytic effects of D (25–500 nM) and Q (0–50 μM) on OA hAC, we first performed a broad screening of different concentrations of the drugs and determined cytotoxicity by means of an alamarBlue assay. This is a metabolic activity assay measuring the conversion of resazurin into the fluorescent resorufin which is proportional to the cell number. The results demonstrate that higher concentrations of D and Q are associated with a reduction in the FI, which indicates cytotoxic effects of D + Q treatment (Figure 2a). Based on these results, an optimal treatment range of 25–500 nM D in combination with 5 μM Q was identified and thus chosen for the following experiments.

FIGURE 2.

Selective elimination of senescent hAC by the senolytics D + Q. (a) Dosage finding by means of an alamarBlue assay, measuring the cell viability of isolated high passage hAC. Cells were treated with D (25–500 nM) + Q (0–50 μM) for 3 d. Cell viability was approximated by the FI visualized in the heatmap. (b) Relative SA‐β‐Gal activity of isolated hAC (passage 1–4) (n ≥ 3) and (c) mRNA levels of CDKN1A and ‐2A of hAC at low versus high passage rel. to cartilage tissue (n ≥ 8). (d) alamarBlue assay of low versus high passage hAC after D + Q stimulation for 3 d (n ≥ 4). (e) Exemplary live/dead images of human cartilage tissue w/o and with D + Q stimulation for 3 d at d3 and d7 (scale bar = 100 μm). Living cells appear in green, while dead cells appear in red. (f) Corresponding cell viability analysis of human (OARSI grade 1–2) (n ≥ 3) and (g) healthy porcine cartilage (n ≥ 3) after treatment for 3 d with D + Q (analysis at d3, d7 and for human cartilage additionally at d14). Statistical analysis: (c) multiple unpaired t test; (d) two‐way ANOVA with Šidák's multiple comparisons test: *p < 0.05, **p < 0.01; (f) two‐way ANOVA with Dunett's multiple comparisons test: *p < 0.05.

Next, we determined whether cell toxicity of D (25–500 nM) and Q (5 μM) on OA hAC resulted from selective senolysis of senescent cells. For this purpose, we first confirmed the increase of cellular senescence through passaging by means of an SA‐β‐Gal activity assay, which displayed a correlation between passaging and the enzymes' activity (Figure 2b). Moreover, passaging resulted in an altered cell morphology, demonstrated by an elongated cell shape (Figure S2a,b). The accumulation of senescent cells was further confirmed on gene expression level. In the following, OA hAC of passage ≥4 are referred to as “high passage hAC”, while hAC of passage ≤2 as “low passage hAC”. High passage hAC expressed significantly elevated CDKN1A and ‐2A mRNA levels as compared with low passage hAC (Figure 2c). Stimulation of high and low passage hAC with D + Q resulted in a significantly more pronounced reduction of the FI in cells of high passages in case of 250 nM and 500 nM D (Figure 2d). Therefore, these concentrations were used for all subsequent experiments. Moreover, this finding confirms that D + Q might selectively eliminate senescent chondrocytes. To further investigate the selective removal of senescent cells, human OA cartilage exhibiting moderate degeneration (OARSI grade 1–2) and healthy porcine cartilage tissue (OARSI grade 0) was treated with the senolytics. By means of a live/dead staining, it was demonstrated that the senolytic treatment resulted in increased cell death in OA tissue (Figure 2e,f), whereas no cytotoxic effects were observed in healthy porcine cartilage (Figure 2g). Cell viability in human explants recovered after removal of the therapeutics (Figure 2f).

3.2. D + Q treatment reduces cell proliferation in isolated hAC and percentage of p16 positive cells in cartilage tissue

It is assumed that the cell cycle regulator p21^WAF1/Cip1 (CDKN1A) is required for the initiation of cellular senescence and p16^INK4a (CDKN2A) for maintaining this state (Kumari & Jat, 2021). Therefore, their expression was studied in greater depth immediately after senolytic treatment and 4 d after withdrawal of the therapeutics. Immediately after therapy with D + Q (d3), no decrease of CDKN1A mRNA levels was found in high passage hAC (Figure 3a), whereas the expression was significantly downregulated 4 d after withdrawal of the senolytics (d7) (Figure 3b). CDKN2A mRNA levels were not affected by the senolytics in vitro (Figure 3c,d). Treatment of isolated hAC with D + Q resulted in a reduction of Ki‐67 positive cells 2 d after deprivation (Figure 3e). These results are in line with the alamarBlue assay, in which D + Q treated cells exhibited a lower growth rate after withdrawal of the senolytics (Figure 3f). In cartilage tissue, only few cells were positive for Ki‐67 and no significant differences were found at d7 and d14 in D + Q‐treated tissue as compared to the untreated control (Figure 3g). In contrast to the in vitro analysis, D + Q stimulation resulted in a distinct reduction of CDKN2A positive cells in moderately degenerated cartilage tissue from OA patients (OARSI grade 1–2) (Figure 3h,i), indicating the selective clearance of senescent cells.

FIGURE 3.

Influence of D + Q therapy on cell cycle regulators and cell proliferation. (a, b) qRT‐PCR: mRNA levels of CDKN1A (n = 6) and (c, d) CDKN2A (n = 5) of high passage hAC immediately after application (d3) and after deprivation (d7) of D + Q. (e) Percentage of Ki‐67 positive cells in isolated hAC (IF staining) (n = 4) and (f) the measured cell proliferation by means of an alamarBlue assay relative to d4 (n = 4). (g) Ki‐67 positive cells in cartilage tissue (n = 3), (h) quantification of p16 positive cells (n = 4) and (i) the exemplary images (IHC) after D + Q treatment of human cartilage explants (scale bar = 100 μm). Blue colored box plots = gene expression analysis data; green colored box plots = data of protein analysis. Statistical analysis: (b) one‐way ANOVA with Šidák's multiple comparisons test: *p < 0.05; (h) one‐way ANOVA with Dunnett's multiple comparisons test: ****p < 0.0001.

3.3. D + Q treatment promotes chondroanabolism

It is well known that chondrocytes rapidly lose their chondrogenic phenotype during in vitro culturing, including aggrecan and collagen type II expression among other characteristics (Aulthouse et al., 1989; Benya & Shaffer, 1982). For this reason, we have investigated the effects of D + Q on the chondroanabolism of high passage hAC. The gene expression of ACAN was significantly upregulated immediately after therapy and 4 d after withdrawal of the drugs (Figure 4a,b). Interestingly, COL2A1 expression was likewise significantly increased at d3, but even further enhanced, from about 7‐fold (d3) to approximately 30‐fold, after deprivation of the senolytics (Figure 4c,d). Moreover, D + Q treatment resulted in a significant increase of SRY‐box transcription factor 9 (SOX9) mRNA levels immediately after therapy and were still slightly enhanced after withdrawal (Figure 4e,f). Asides from that, higher amounts of GAGs and a significantly increased biosynthesis of COL2 were detected in 2D‐cultured isolated hAC (Figure 4g‐i). In line with the cell culture experiments, we could confirm elevated COL2 production in human cartilage tissue by means of a CPII ELISA (Figure 4j,k) and IHC staining against COL2 (Figure 4l) (Figure S3).

FIGURE 4.

Promotion of the chondroanabolism after elimination of senescent hAC using D + Q. (a, b) Gene expression levels of ACAN (a) n = 6, (b) n = 6, (c, d) COL2A1 (c) n ≥ 5, (d) n = 6 and (e, f) SOX9 immediately after application (d3) and after withdrawal (d7) of the senolytics (e) n = 4, (f) n ≥ 2. (g) GAG staining of isolated hAC using alcian blue (d7), (h) exemplary images of COL2‐IF staining (scale bar = 50 μm) and (i) the corresponding quantification of the FI (n = 3). (j,k) Relative CPII levels after senolytic therapy (d7 + d14) measured by means of an ELISA (j) n = 8, (k) n = 7. (l) Exemplary COL2 stainings (d7) of untreated human cartilage and after treatment with D + Q (scale bar: Overview = 200 μm, superficial zone = 100 μm, deep zone = 100 μm). Blue colored box plots = gene expression analysis data; green colored box plots = data of protein analysis. Statistical analysis: (a–e) one‐way ANOVA with Šidák's multiple comparisons test: *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001; (i) unpaired t test ***p < 0.001; (j, k) one‐way ANOVA with Dunnett's multiple comparisons test: **p < 0.01, ***p < 0.001.

3.4. Single treatment of Dasatinib but not quercetin or navitoclax promotes chondroanabolism in high passage hAC

As mentioned above, chondrocytes undergo phenotypical alteration through in vitro expansion, which is reflected in the loss of cartilage‐typical extracellular matrix production. Therefore, we used high passage hAC, which can be considered as “dedifferentiated”, and studied their “redifferentiation” potential after D + Q treatment by means of an in vitro chondrogenesis assay. For this purpose, we treated hAC with D + Q for 3 d and compared the differentiation potential with the untreated hAC in a 3D pellet culture for 28 d. Based on the safranin‐O and COL2 staining, we observed that elimination of senescent cells using D + Q promoted the chondrogenic differentiation capacities of the surviving hAC as demonstrated by a significant increase in the overall score (Figure 5a,b). Next, we investigated, whether the treatment with D + Q can induce chondrogenic differentiation in ATDC5 cells. The gene expression analysis of COL2A1 (Figure 5c) and ACAN (Figure 5d) demonstrates that both were induced by CDM, while no increase of the chondroanabolic markers was observed after D + Q treatment in basal medium. These results were in line with the alcian blue and COL2‐IF staining (Figure 5e,f). Since chondrogenic differentiation in ATDC5 cells is known to be induced by Insulin‐Transferrin‐Selenium (ITS), D + Q treatment was performed in absence of ITS. To exclude that the presence of ITS was required in D + Q‐related chondroanabolic response of hAC observed above, we cultured hAC in analogue conditions. In hAC, ITS administration displayed no beneficial effects on the gene expression levels of COL2A1 (Figure 5g) and ACAN (Figure 5h) or GAG production (Figure 5i) as compared to ITS‐free conditions, which suggests that the enhancement of the chondroanabolism in hAC can be attributed to D + Q treatment alone. Since the senolytic treatment promoted the chondrogenic phenotype in hAC but not in ATDC5 cells, we concluded that the therapy can be seen as a kind of reprogramming or restoring of the former phenotype rather than a direct inducer of chondrogenesis.

FIGURE 5.

Influences of D + Q treatment on the chondrogenic phenotype of hAC and ATDC5 cells. (a) Overall score of the chondrogenic redifferentiation potential of dedifferentiated hAC (n = 7) and (b) exemplary images of the COL2 and safranin‐O staining (scale bar = 200 μm). (c) Gene expression analysis of COL2A1 (n = 4) and (d) ACAN (n = 4) in ATDC5 cells after culturing in CDM for 7 d or serum‐reduced medium after withdrawal (d7) of the senolytics. (e) Corresponding alcian blue and COL2‐IF staining (scale bar = 50 μm) of ATDC5 cells. (f) Absorbance at 595 nm after destaining alcian blue with 1% SDS (n = 4). (g) qRT‐PCR: Relative mRNA levels of COL2A1 (n = 5) and (h) ACAN (n = 5) after senolytic treatment of hAC with or w/o ITS. (i) Exemplary alcian blue staining of untreated hAC (basal) and after D + Q treatment with and w/o ITS. (j) qRT‐PCR: mRNA levels of SOX9 (d3) (n = 3) and (k) of COL2A1 (d7) (n = 4) after stimulation with D, Q or D + Q. (l) Exemplary alcian blue staining after single and combination therapy with D + Q. (m) qRT‐PCR: Gene expression levels of YAP1 (n = 4) after D, Q or D + Q therapy (d7). (n) Phalloidin staining after D, Q or D + Q therapy (d3) (scale bar = 50 μm). Blue colored box plots = gene expression analysis data; green colored box plots = data of protein analysis. Statistical analysis: (a) unpaired t test: *p < 0.05; (c, d) + (f‐, h) one‐way ANOVA with Dunnett's multiple comparisons test: *p < 0.05, **p < 0.01, ***p < 0.001; (j) one‐way ANOVA with Šidák's multiple comparisons test: *p < 0.05.

Next, we examined if both senolytics (D + Q) are required to achieve the promotion of the chondroanabolism or whether this can be achieved by single therapy of either D or Q. Based on the results obtained after separate application of either D or Q, Dasatinib was found to induce chondroanabolic effects, as shown by a strong increase of SOX9 (d3) (Figure 5j) and COL2A1 (d7) (Figure 5k) gene expression, as well as by an elevated production of GAGs verified by means of an alcian blue staining (Figure 5l). Quercetin did not affect the expression of chondroanabolic markers. One possible underlying mechanism of Dasatinib‐mediated promotion of the chondrogenic phenotype could be through the inhibition of the yes‐associated protein (YAP‐1) (Nie et al., 2019). In our experiments, YAP1 mRNA levels were slightly downregulated by D or Q and the combination (Figure 5m). Furthermore, the phalloidin staining demonstrated distinct changes in the cytoskeleton through therapy with D or D + Q (Figure 5n). As a next step, we clarified whether the clearance of senescent cells per se promotes the chondrogenic phenotype of high passage hAC or if this is related to the unique effects of Dasatinib. For this purpose, senescent cells were eliminated using Navitoclax, a commonly used and well‐described senolytic, which induces apoptosis by targeting antiapoptotic proteins of the BCL‐2 family (Chaib et al., 2022; Soto‐Gamez & Demaria, 2017; Yang et al., 2020). Navitoclax revealed stronger cytotoxic effects in high‐passage (p ≥ 4) hAC compared to low passages (p ≤ 2) (Figure S4a,b) and reduced the number of p16‐positive cells in cartilage tissue (Figure S4c) as similarly demonstrated for D + Q therapy above (Figure 2d and Figure 3h,i). However, in contrast to D + Q, gene expression levels of COL2A1 and ACAN of isolated high passage hAC were not elevated upon Navitoclax treatment (Figure S4d–g). In addition, Navitoclax could not promote the chondrogenic redifferentiation potential of “dedifferentiated” high passage hAC (Figure S4h,i). Overall, the results clearly demonstrate that the elimination of senescent cells alone does not promote chondroanabolism and that this effect is particularly related to Dasatinib.

3.5. Senolytic treatment results in downregulation of SASP factors and creates a pro‐chondrogenic environment

Senescent cells are characterized by a cell type‐specific SASP. In case of chondrocytes, SASP factors contribute to a pro‐inflammatory and anti‐anabolic environment, which provokes a shift in cartilage homeostasis towards catabolism, while suppressing extracellular matrix synthesis. Furthermore, these mediators are assumed to promote increasing senescence in neighboring cells (Coryell et al., 2021; Jeon et al., 2018; McCulloch et al., 2017).

By the following experiment, we tested if the supernatant of senescent cells, containing various SASP factors, was able to affect the phenotype of low passage hAC with regard to senescent and chondroanabolic markers. For this purpose, we induced senescence in hAC by Doxo as previously described (Kirsch et al., 2022). The senescent phenotype of Doxo‐stimulated cells was confirmed on mRNA level (Figure 1a,b). The SASP‐conditioned medium of these cells was then used to stimulate non‐senescent hAC for 48 h. Exposition to SIPS‐conditioned medium reduced proliferation (Figure 1c) and increased gene expression of CDKN1A (Figure 1d), while suppression that of COL2A1 (Figure 1g) in hAC as compared to hAC cultivated in nonconditioned medium. No differences were detected in case of CDKN2A and ACAN mRNA levels (Figure 1e,f).

Similar results on COL2A1 gene expression were also achieved by stimulating cells with medium from high passage hAC (SASP‐CM). Conversely, treatment of hAC with medium derived from cells prior stimulated with D + Q (DQ‐CM) significantly increased the gene expression of COL2A1 (Figure 1i) This indicated that Dasatinib promotes chondroanabolism via creation of a pro‐chondrogenic environment. To confirm our suggestion, the expression of SASP markers was investigated in isolated high passage hAC immediately after therapy (d3) and after deprivation (d7) of D + Q. The senolytic therapy resulted in a downregulation of the CXCL1 and IL6 gene expression (Figure S5a,c), however, this effect was not maintained after deprivation of the senolytics (Figure S5b,d). Similar results could be confirmed on protein level by means of an ELISA (Figure 1j‐m). The senolytic treatment had no influence on the release of IL8 (Figure 1n,o). Since senescent chondrocytes secrete considerable amounts of MMPs, which in turn promote degradation of extracellular matrix (ECM) components (Loeser, 2009), MMP13 expression was examined. The results demonstrate that the mRNA levels were significantly reduced during therapy, which could be sustained even after withdrawal (Figure S5e,f). Although not all experiments reached statistical significance, a clear trend was found regarding the reduction of SASP factors after senolysis.

3.6. RNA‐sequencing reveals downregulation of SASP factors and upregulation of chondroanbolic markers and growth factors

To further investigate the underlying pathways and potential key factors involved in D + Q‐mediated senolytic and chondrogenic effects, transcriptomic analysis was performed. For this purpose, high passage hAC were treated with D + Q for 3 d followed by total RNA isolation. Bulk RNA‐sequencing analysis demonstrated a significant upregulation of 348 genes and a significant downregulation of 426 genes (Figure 6a,b). In detail, increased gene expression was measured for several markers related to hyaline cartilage (COL2A1, COL9A1/2/3, COL11A2, MATN4, SOX9, and SNORC) and chondroanabolic growth factors (FGF18, IGF1, and TGFB2) (Figure 6a,b). The upregulation of FGF18 and IGF1 was further confirmed via qRT‐PCR (Figure 6c,d). At the same time, anti‐anabolic IGF1‐binding proteins, IGFBP1 and 3, were significantly reduced (Figure 6a,b) and the expression of FGF receptors (FGFR)‐2, ‐3, and ‐4 was upregulated (Figure 6a,b). Subsequent qRT‐PCR analysis revealed a decreased FGFR1/FGFR3 expression ratio through senolytic therapy (Figure 6e), implying a shift towards chondroanabolic FGFR‐signaling. In line with this, gene expression of SERPINA1 and SERPINA3, encoding for alpha1‐antitrypsin (AAT), was significantly enhanced (Figure 6a,b).

FIGURE 6.

RNA‐sequencing analysis after D + Q treatment. High passage hAC were treated with D + Q for 3 d and transcriptomic analysis was performed by means of a bulk RNA‐sequencing. (a) DEG‐Plot—significantly up‐ and down‐regulated genes. (b) Volcano Plot—significantly up‐ and down‐regulated genes. Confirmation of RNA‐sequencing data by qRT PCR analysis of (c) IGF1 (n = 4), (d) FGF18 (n = 5), and (e) FGFR1/3 ratio (d3) (n = 4). Results of KEGG pathway enrichment analysis of (f) down‐ and (g) up‐regulated signaling pathways. Statistical analysis: (a, b), (f, g): t test, (d) unpaired t test.

CTGF, BIRC5 (survivin), and CYR61, target genes of YAP‐1 (Hooglugt et al., 2021) were downregulated, thus indicating an inhibition of the YAP‐1 signaling pathway (Figure 6a,b). Additionally, the therapy resulted in an increased gene expression of the extracellular antioxidative enzyme SOD3 (Figure 6a,b). Further anti‐inflammatory properties were displayed by the decreased expression of IL6, IL7R, IL26, and IL32 (Figure 6a,b). Anti‐catabolic effects were demonstrated by downregulation of the aggrecanases ADAMTS‐1 and 5 (Figure 6a,b). Furthermore, the treatment decreased the expression level of RIPK3 (Figure 6a,b), an important mediator during necroptosis (Morgan & Kim, 2022). In addition, the senolytic therapy resulted in the downregulation of several genes related to proliferation (Figure 6a,b).

Overall, the Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analysis revealed a downregulation of genes associated with cell cycle, cellular senescence, DNA replication, and p53 signaling (Figure 6f). An upregulation was detected in case of the PI3K‐Akt‐, and TGF‐β‐signaling pathway (Figure 6g). Moreover, genes related to the FoxO signaling pathway, which plays an important role in chondrocyte function and senescence (Wang et al., 2018), were significantly regulated upon D + Q treatment (Figure 6f,g).

4. DISCUSSION

The relevance of research into new therapies for the treatment of OA becomes particularly clear, given that 654.1 million people worldwide are affected by osteoarthritis of the knee joint, as estimated in 2020 (Cui et al., 2020). Currently, no disease modifying OA drugs (DMOADs) are approved and therefore OA is still an uncurable disease (Arden et al., 2021; Kim et al., 2022). Due to the detrimental effects of cellular senescence on tissue homeostasis, selective elimination of senescent cells received increasing attention in OA as well as other degenerative diseases during the last decades (Chaib et al., 2022; Lagoumtzi & Chondrogianni, 2021). Although the senolytic UBX0101 failed in phase II clinical trials as a potential therapeutic agent in knee OA (Chaib et al., 2022; Lagoumtzi & Chondrogianni, 2021), senolysis is still considered a promising strategy and further senolytics, like fisetin, are currently under investigation in OA patients (NCT04210986).

In our study, we tested the senolytic combination therapy with D + Q for the first time in human cartilage tissue derived from OA patients. Considering our novel findings on D + Q therapy, we concluded that the senolytics could be a suitable therapeutic approach against OA in many ways. D + Q selectively eliminated senescent chondrocytes in cell as well as tissue culture. The treatment resulted in a tremendous promotion of the chondroanabolism in surviving chondrocytes, demonstrated by elevated COL2 and PG levels. Additionally, D + Q treatment of hACs improved the chondrogenic redifferentiation indicating a rejuvenation or cellular reprogramming of the dysfunctional cells. Overall, our results imply that D + Q therapy, and in particularly Dasatinib, restored the chondrogenic phenotype of hAC, while elimination of senescent cells by Navitoclax had no beneficial effect regarding the chondroanabolism. Furthermore, we provide evidence that D + Q‐mediated senolysis created a pro‐chondrogenic environment by suppressing the production of pro‐inflammatory and anti‐anabolic SASP factors, such as CXCL1, IL6, and IGFBPs, and increasing the expression of chondroanabolic growth factors.

After identifying the beneficial effects of D + Q treatment on the chondroanabolism, we investigated if D + Q in combination is required or if this can be achieved by single therapy of one of those. In detailed analyses, we demonstrated that Dasatinib alone promoted the chondrogenic phenotype. Several in vivo studies provide evidence that quercetin may be useful for the treatment of OA, particularly in the prevention of PTOA (Lv et al., 2022; Mok et al., 2020; Wei et al., 2019). Administration of quercetin has been described with anti‐nociceptive (Mok et al., 2020), cell‐ and chondroprotective (Hu et al., 2019; Lv et al., 2022) as well as anti‐inflammatory properties (Hu et al., 2019), thus attenuating surgically‐induced OA progression in rats. Monotherapeutic treatment with Dasatinib, in turn, exhibited pro‐regenerative effects in a rabbit full‐thickness cartilage defect‐model (Nie et al., 2019). In fact, it has been reported that Dasatinib promotes mesenchymal stem cells (MSCs) chondrogenesis, presumably via the inhibition of the Src/Hippo‐YAP signaling pathway. The treatment resulted in a decreased expression and enhanced cytoplasmic degradation of YAP (Nie et al., 2019). The downregulation of YAP is known to be necessary for the chondrogenesis of MSCs (Karystinou et al., 2015; Nie et al., 2019), while high YAP activity, in contrast, is thought to suppress chondrogenic differentiation (Karystinou et al., 2015). Moreover, the YAP‐TEAD complex was found to maintain survival of senescent cells (Anerillas et al., 2023). In our study, we could confirm a significant decrease of the three main signature genes of the YAP1‐signaling pathway (BIRC5, CTGF, and CYR61) upon D + Q treatment by means of an RNA‐sequencing. Furthermore, we observed a negative trend in the transcription of YAP1 itself. In addition, we found remarkable changes in the actin cytoskeleton due to Dasatinib and D + Q treatment. Indeed, the transcriptional coactivators TAZ and YAP are known to play an important role in the dynamic of the cytoskeleton (Mason et al., 2019). It is well‐known that dedifferentiated chondrocytes display an elongated, fibroblast‐like cell morphology and increase in stress fibers (actin polymerization), while actin depolymerization is thought to result in chondrocyte redifferentiation (Lindberg et al., 2024). Thus, our findings are in line with previous reports and further emphasize the importance of YAP1 regulation in chondrocyte maturation. Although, Quercetin had no noticeable beneficial effect regarding the pro‐anabolic properties of Dasatinib, we could not find any adverse effect either. Considering this and the beforementioned positive effects of Quercetin reported in other studies (Hu et al., 2019; Lv et al., 2022; Mok et al., 2020; Wei et al., 2019), we suggest that the combination of D + Q is still reasonable.

Next, we considered whether the elimination of senescent cells alone exerts beneficial effects on the chondroanabolism. Therefore, we treated chondrocytes with Navitoclax and confirmed the selective elimination of senescent cells through this therapy. However, Navitoclax could not support the chondroanabolism but rather impaired the chondrogenic phenotype of hAC. In contrast, Yang et al. reported not only chondroprotective effects in a rat PTOA model, but also an increase in the chondrogenic phenotype of human OA chondroyctes after Navitoclax treatment (Yang et al., 2020). Nevertheless, there are further studies from the musculoskeletal field, which observed anti‐anabolic effects of Navitoclax, confirming our findings. For example, Sharma et al. reported trabecular bone loss and impaired osteoprogenitor cell function in osteoporotic mice after Navitoclax treatment (Sharma et al., 2020). In contrast, D + Q therapy had anabolic and antiresorptive effects on bone, thus preventing bone loss and enhancing the repair potential in murine postmenopausal and age‐related osteoporosis models (Farr et al., 2017a, 2017b; Wang et al., 2023). Overall, these findings imply that senolysis alone does not restore the phenotype of bone or cartilage cells and that senolytics have to be carefully chosen, depending on the target tissue and cell type.

Further, we were able to demonstrate that cells which received medium from previously D + Q stimulated cells expressed higher levels of COL2A1, while medium from untreated high passage cells had suppressing effects. We supposed that this could be due to the creation of a pro‐chondrogenic microenvironment through D + Q therapy. In line with our assumption, the results of the RNA sequencing analysis provided evidence that the chondroanabolic effects of the conditioned medium might be explained by upregulation of relevant growth factors (IGF1, FGF18, and TGFB2) as well as SERPINA1/3 (AAT) and concomitant reduction of antianabolic proteins, including cytokines and IGFBPs. IGF1 and FGF18 are well‐known drivers of the ECM synthesis (Fortier et al., 2011). In a previous study, we could confirm the chondroanabolic effects of IGF‐1 on cartilage tissue by an increased COL2A1 gene expression levels and elevated CPII levels, a biomarker for COL2 synthesis (Riegger, Joos, et al., 2018). FGF‐18 treatment displayed pro‐anabolic features in various animal studies. For example, it enhanced cartilage repair in an ovine model after a chondral defect (Power et al., 2014) and displayed cartilage protective effects in a rat PTOA model (Yao et al., 2019). Additionally, we previously observed cell‐ and chondroprotective effects of FGF18 therapy (Riegger, Joos, et al., 2018). In case of AAT, it has been reported that SERPINA1 and 3 are highly expressed in articular cartilage and during chondrogenesis of MSCs, but decreased during in vitro expansion (dedifferentiation) of hAC and OA progression (Boeuf et al., 2008). Furthermore, AAT exerted chondroprotective and chondroanabolic effects (Kaneva et al., 2021). Altogether, increased expression of markers associated with hyaline cartilage formation (e.g., ACAN, COL2A1, COL9A1/2/3, COL11A2, and MATN4), chondrocyte maturation (e.g., SNORC, SOX9, and SERPINA1/3), and the observed decline in proliferation in D + Q‐treated hAC, imply that the dedifferentiated hAC return to a mature, non‐mitotic state. This is very impressive regarding the previous assumption that serum‐free conditions and a 3D environment are obligatory requirements in redifferentiation of primary chondrocytes (Ahmed et al., 2014; Caron et al., 2012). Beyond the pro‐anabolic effects, the sequencing analysis revealed a downregulation of aggrecanases and pro‐inflammatory factors after D + Q treatment. Altogether, the reduction of SASP factors is of great importance, as they negatively influence the surrounding cells (Zhang et al., 2021). However, it should be noted that our data imply that senolysis per se does not ensure a pro‐chondroanabolic microenvironment, as Navitoclax rather suppressed COL2 synthesis despite elimination of senescent.

The therapy with D + Q is currently investigated in clinical trials regarding its diseases modifying effects in common age‐related disorders such as the chronic kidney disease or the idiopathic pulmonary disease (Chaib et al., 2022; Hickson et al., 2019, 2020; Nambiar et al., 2023). Therapy with D + Q was generally well‐tolerated (Hickson et al., 2019, 2020; Nambiar et al., 2023), which suggests that intraarticular injection of D + Q in OA joints would be similarly innocuous and could—with regard to our meaningful ex vivo results—certainly be a promising therapeutic approach in patients suffering from knee OA. Despite previous in vitro and in vivo investigations on the effects of D + Q in OA, we consider our study as novel and relevant. However, previous studies did not investigate the phenotype of surviving chondrocytes. Moreover, the results derived from animal experiments, mainly performed on young animals, cannot be directly transferred to the clinical situation in which predominately elderly people suffering from OA. Therefore, we used human OA cartilage, that provides important information about the efficacy of D + Q on aged and OA‐affected tissue. This tissue contains senescent cells which have accumulated in the course of OA progression and were not provoked by extrinsic manipulation, such as radiation (Xu et al., 2018) or addition of chemicals (Kirsch et al., 2022), thus perfectly mirroring the clinical appearance. Despite of the advantages of using patient material, our study is limited to the effects on cartilage tissue and does not consider other joint‐related tissues, such as the synovial membrane or subchondral bone, involved in OA progression. Thus, restorage of the chondrogenic phenotype alone might not halt disease progression. However, beneficial effects of D + Q treatment on other joint tissues affected by OA, including subchondral bone and synovial membrane, have been recently reported in aged mice. Efficient senolysis and reduced expression of markers associated with innervation and angiogenesis was confirmed in human OA synoviocytes. Overall, the authors concluded that D + Q might be suitable to reduce synovitis and consequent pain in age‐related OA (Gil et al., 2022). Nevertheless, further studies on the effects of senolytics—and particularly D + Q—on senescent cells in, for example, the infrapatellar fat pad might be very interesting, because of its substantial contribution to joint inflammation and OA progression (Zhou et al., 2022). Moreover, it was not possible to include a healthy control group in our study. To overcome this issue, healthy porcine tissue had to be used.

Considering our results, we could demonstrate that promotion of the chondrogenic phenotype by D + Q after selective elimination of senescent chondrocytes might be based on two mechanisms: first, direct modulation of the YAP‐1 signaling pathway and, second, paracrine effects through reduction of detrimental SASP factors and increased expression of pro‐chondroanabolic mediators. Altogether, this senolytic combination therapy could offer a new therapeutic strategy in OA.

AUTHOR CONTRIBUTIONS

SM: conceptualization, experiment conduction and data analysis, visualization, manuscript writing. JRK: funding acquisition, conceptualization, experiment conduction and data analysis, visualization, manuscript writing. VK + LR: experiment conduction and data analysis, review and editing of manuscript. RB: review and editing of manuscript.

CONFLICT OF INTEREST STATEMENT

The authors declare that they have no competing interest in this publication.

Supporting information

Appendix S1.

Maurer, S. , Kirsch, V. , Ruths, L. , Brenner, R. E. , & Riegger, J. (2025). Senolytic therapy combining Dasatinib and Quercetin restores the chondrogenic phenotype of human osteoarthritic chondrocytes by the release of pro‐anabolic mediators. Aging Cell, 24, e14361. 10.1111/acel.14361

DATA AVAILABILITY STATEMENT

The data that support the findings of this study are available from the corresponding author upon reasonable request.