Reversing detrimental effects of OA-related terminal maturation in high-throughput in vitro and biomimetic ex vivo human models

Abstract

Background

Osteoarthritis (OA) lacks disease-modifying therapies, in part due to limited human-relevant preclinical models. OA pathophysiology is characterized by loss of chondrocyte maturation arrest, producing growth plate–like changes and terminal differentiation. We aimed to establish a scalable human organoid-based platform for efficient therapeutic screening.

Methods

Cartilage organoids were derived from OA patient tissues in a 96-well format and benchmarked against matched osteochondral explants. Terminal maturation was induced using triiodothyronine (T3). Three candidate compounds, all thyroid hormone receptor blockers, were tested across dose ranges and incubation times. The most effective compound was validated in a human biomimetic osteochondral explant model to assess translational relevance.

Results

The high-throughput organoid platform consistently reproduced hypertrophic responses, closely mirroring osteochondral explant behavior, and provided reliable readouts of terminal maturation marker CCDC80, Of the three compounds evaluated, Compound A demonstrated the strongest effects, significantly increasing CCDC80 mRNA and MMP3 protein levels in the organoid system. In the osteochondral explant model, Compound A reversed T3-induced terminal maturation, suppressing markers including CCDC80, PHOSPHO1, and MMP3.

Conclusion

This study introduces a scalable and reproducible preclinical screening strategy using human OA tissues for compound evaluation. The platform enables efficient dose-response testing and identifies thyroid signaling as a promising therapeutic axis for modulating chondrocyte maturation. These findings suggest a novel approach to slow OA progression, preserve joint function, and improve mobility and quality of life in affected individuals.

Supplementary Information

The online version contains supplementary material available at https://doi.org/10.1186/s13075-026-03903-w.

Introduction

Osteoarthritis (OA) is an age-related degenerative joint disease. It is particularly characterized by the degeneration of cartilage which causes pain and disability. Currently, there are no effective treatment options that halt or reverse the OA pathophysiology and available treatment options are pain relief methods or end-stage total joint replacement surgery.

Healthy chondrocytes, reside in articular cartilage in a maturation-arrested state with low proliferative and metabolic activity. However, transcriptome-wide studies of OA affected cartilage highlighted that OA pathophysiology is characterized by active chondrocytes that loose their maturational arrested state, recuperate a detrimental growth plate morphology, and undergo terminal maturation, a process known as endochondral ossification [1, 2]. Relevant in this respect is that active thyroid (T3) has been shown to signal terminal maturation in growth plate chondrocytes. Upregulation of thyroid signaling in articular cartilage was shown to underlie OA onset [2, 3] and to affect the propensity of chondrocytes to enter an OA disease state, particularly upon environmental challenges such as mechanical loading [4].

In the absence of reliable human age-related biomimetic OA models of the interacting osteochondral compartment, we have previously set up and exploited ex vivo osteochondral explants as a human preclinical biomimetic model of joint tissues [1, 5]. In subsequent studies, we showed by transcriptome wide analyses that T3 induced terminal chondrocyte maturation was particularly marked by upregulated gene expression of CCDC80, ANKH, and PHOSPHO1 [1]. Additionally, the reactivation of extracellular matrix remodeling was shown by upregulated COL2A1 expression. Moreover, to accommodate sensitive monitoring of T3 induced hypertrophy and terminal maturation in high throughput, downregulation of MMP3 was recognized as a potential candidate for e.g. high throughput ELISA protein screening [1].

In the current study, we set out to perform a proof of concept study in which we miniaturized and validated our previously established neo-cartilage 3D organoid model [6] to allow a high-throughput dose response screen of different compound formulations that target thyroid signaling. Subsequently we exploited the recently characterized human ex vivo biomimetic OA model of T3 induced terminal chondrocyte maturation as faithful validation tool of the most optimal compound conditions. Together we present a drug screening approach that allows for fast personalized testing of potential new treatment regimens and wide preclinical applications.

Methods

High-throughput neo-cartilage organoids and osteochondral explants

Osteochondral explants and primary articular chondrocytes were obtained from joints included in the Research in Articular Osteoarthritis Cartilage (RAAK) study. The RAAK study was approved by the medical ethics committee of the Leiden University Medical Center (P08.239/P19.013) [7].

For the osteochondral explant model, explants were taken and cultured from the condyles of knee joints described previously [4, 8]. In short, biopsies with a diameter of 8 mm were taken from the macroscopically healthy cartilage from OA patients and washed in PBS. Next, the explants were cultured in serum-free chondrogenic differentiation medium (CDM) consisting of DMEM (Gibson) with pen/strepp (Gibco), ITS+ (Corning) and supplemented with ascorbic acid (Sigma-Aldrich), L-proline (Sigma-Aldrich), sodium pyruvate (Sigma-Aldrich) and dexamethasone (Sigma-Aldrich). A total of 6 donors were used, resulting in 12 individual samples.

For the high-throughput neo-cartilage organoids, primary human articular chondrocytes (hPACs) were collected and cultured as described previously [6]. In short, macroscopically healthy cartilage was digested at 37 °C overnight in DMEM (Gibson) supplemented with 2 mg/ml collagenase type 1 (Worthington). To remove cell and tissue debris, cells were strained with a 40 μm cell stainer (Greiner Bio-one). Subsequently, the high-throughput human in-vitro cartilage organoids were formed using 500 x g centrifugal forces on a 96 wells plate with 5 × 10⁵ cells per well. CDM freshly supplemented with TGF-β1 (peprotech) was used for the culture of the cartilage organoids. This optimal cell number was determined based on measurable gene expression levels and ability to use histology on the samples (data not shown). Different timings of spinning down were tested and 500 x g resulted in the best pellet formation (data not shown). Switching the proliferative media towards chondrogenic differentiation medium with TGF-B supplement results, similar as with the normal neo-cartilage organoids, in the formation of cartilaginous matrix. In total 3–10 donors were used, resulting in 3–25 individual samples per condition.

Induction of hypertrophy and terminal maturation in human 3D in vitro and ex vivo models

To induce hypertrophy and terminal maturation in 3D in vitro organoids and in 3D ex vivo osteochondral explants, organoids were exposed to 100 ng/ml triiodothyronine (T3) for 4 days and explants were exposed to 100 ng/ml T3 for 7 days. These concentrations were extracted from previous publications [8, 9] and here further updated by assessing upregulation of hypertrophy markers.

Compounds

In this study efficacy to reduce T3 induced hypertrophy and terminal maturation was assessed for three compounds G1986552 indicated as compound-A (Cpd-A), G1906307 indicated as compound-B (Cpd-B), and G1906508 indicated as compound-C(Cpd-C). These compounds were developed and supplied by Galapagos as full antagonists of thyroid hormone receptors.

Experimental set up compound testing

The dose response and efficacy of compounds to reverse T3 induced hypertrophy and terminal maturation was tested in first in the set-up human 3D in vitro organoid models of chondrocyte hypertrophy and terminal maturation as outlined in Fig. 1A. Herein, dose response was first established by adding compounds in a concentration of either 0.3, 1.0, 3.0, or 10 µM for 4 or 7 days (Supplementary table S1A). Subsequently organoids were harvested for gene expression measurements and medium was collected for protein measurements.

Fig. 1.

Experimental timeline of the culture with neo-cartilage organoids and osteochondralexplants. A, Timeline of the high-throughput neo-cartilage organoids. T3 was added for 4 days. Subsequently, two incubation times were tested, 4 day incubation time in blue and 7 day incubation time in cyan. Compounds A, B and C were added in concentrations of 0.3, 1, 3, and 10 µM. B, Timeline of the biomimetic osteochondral explants culture time. The explants were incubated with 100 ng/ml T3 for 7 days, on day 10 compounds A and B were added in a concentration of 10/20 µM for 7 days. T3 = triiodothyronine

As outlined in Fig. 1B after induction of hypertrophy and terminal maturation in human osteochondral explants, they were incubated with compound A, in a concentration of 10 and 20 µM, and compound B, in a concentration of 10 µM, for 7 days (Supplementary table S1B). Subsequently on day 18, the cartilage was harvested for gene expression measurements and medium was collected for protein measurements.

Gene expression analysis

To assess the effects of T3 and the compounds, the gene expression was measured of the neo-cartilage organoids and explants using RT-qPCR. They were processed and RNA was extracted as described previously [1]. In short, the explants were pulverized and four neo-cartilage organoids were pooled in TRIzol (Invitrogen), subsequently the RNA was extracted using the RNeasy Mini Kit (Qiagen). The quantity and quality of the RNA was assessed and cDNA synthesis was performed. The gene expression of the neo-cartilage organoids was measured with RT-qPCR, using Fast Start Sybr Green Master mix (Roche Applied Science) with the QuantStudio 6 Real-Time PCR system, while the gene expression of the explants was measured with RT-qPCR using the Biomark™ 96.96 Dynamic Arrays (Fluidigm) according to the manufacturer’s protocol. GAPDH was used as housekeeping gene across all measurements and was used for the -ΔCT (-dCT) calculations.

Protein expression analysis

To assess the effects of T3 and the compounds on protein level, medium was collected during harvest and the protein level of MMP3 was measured using an MMP3 ELISA kit (Quantikine). This kit uses a sandwich ELISA. Medium samples and standards were added to the plate. Next, the plate is washed and MMP3 specific polyclonal antibody was added to the plate. After again washing the substrate solution was added to the wells and color development was stopped using the stop buffer. Subsequently, the color intensity at 450 and 570 nm were measured on a microplate reader (Synergy HT, BioTek). The standard curve was made using a log/log curve of the OD values and concentration of the samples were calculated. Samples that fell outside the standard curve were excluded.

Statistical analysis

To determine the differences between conditions on gene expression and protein expression, unpaired generalized estimation equation (GEE test) were performed using SPSS v.29.0. Moreover, GEE analysis was chosen as the applied statistical method to accounts for intra-donor correlation without requiring strict assumptions about the underlying distribution of the data. Herein, genes were taken as dependent variable and the condition as independent variable, qPCR batch was used as a covariate to correct for differences between qPCR experiments. Control samples were taken as reference.

Results

Establishing 96-well neo-cartilage organoid platform

To enable efficient testing of compounds, we miniaturized our previously established in vitro neo-cartilage organoid model of pelleted human primary chondrocytes to a 96 well neo-cartilage organoid format. Hereto we seeded primary chondrocytes in 96-wells plate and applied centrifugal forces, as seen in Fig. 2, cartilaginous matrix was produced in the neo-cartilage organoids on day 6. As shown in Fig. 1, to allow efficacy testing of compounds, we added T3 to this model for 4 days to create a high-throughput hypertrophic and terminally maturing neo-cartilage organoid model.

Fig. 2.

Schematic overview of the miniaturization of the 3D neo-cartilage organoids. Cells are isolated from the macroscopically healthy part of the joint and used to create the neo-cartilage pellets. Inlays showing representative cartilage matrix deposition of the 3D cartilage organoids, using an Alcian Blue staining. Scale bar is 200 μm

Establishing T3 a induced hypertrophy and terminal maturation in in-vitro and ex-vivo human cartilage models

To explore the induction of T3 induced hypertrophy and terminal maturation in the 96-well neo-cartilage organoid platform, we tested response of CCDC80, ANKH and PHOSPHO1 as markers of terminal maturation, and COL2A1 as marker of reactivation and matrix turnover. Human ex vivo osteochondral explants were used as comparison. As seen in Fig. 3, upon T3 exposure, we observed chondrocyte re-activation, marked by COL2A1 (neo-cartilage organoids: beta = 1.4, P = 0.001. osteochondral explants: beta = 2.3, P = 1.5 × 10^− 05, Supplementary table S2) and upregulation of terminal maturation, marked by ANKH (neo-cartilage organoids: beta = 1.8, P = 0.004, osteochondral explants: beta = 1.2, P = 0.007, Supplementary table S2) and CCDC80 (neo-cartilage organoids: beta = 0.5, P = 0.27, osteochondral explants: beta = 2.7, P = 5.2 × 10^− 06, Supplementary table S2). While ANKH showed similar results in the high-throughput model as in the explant model, CCDC80 showed more nuanced results in the high-throughput model (Fig. 3, Supplementary table S2), potentially due to the difference in level of cartilage maturity of the models.

Fig. 3.

T3 induced hypertrophy and terminal maturation in in-vitro and ex-vivo human cartilage models: CCDC80, ANKH and COL2A1 gene expression responses to the hypertrophic T3 treatment in the high-throughput 96 well neo-cartilage organoid model (T3, 100 ng/ml) and the osteochondral explant model (T3, 100 ng/ml).*P < 0.05, **P < 0.01, ***P < 0.001. For human primary miniaturized organoids (mini pellets), primary chondrocytes were isolated from 3–11 donors, resulting in n = 3–25 neo-cartilage samples. For human osteochondral explants we used N = 6 donors resulting in n = 6 explant samples per experimental group

Similar as in the previously established T3 induced hypertrophy and terminal maturation in our osteochondral explant model, COL10A1, marker for early hypertrophy, showed a downregulation in the high-throughput neo-cartilage model (neo-cartilage organoids: beta=-1.6, P = 0.003, osteochondral explants: beta = 0.3, P = 0.714, Supplementary table S2). These results indicate that the high-throughput neo-cartilage organoid model showed similar effects as in the ex vivo osteochondral explants regarding hypertrophic responses on RNA level. The models showed consistent results on tissue state, particularly in COL2A1 and on terminal maturation, particularly in AKH, and a smaller role of CCDC80 in the high-throughput model. In these models, the prioritized markers of T3 induced hypertrophy and terminal maturation have been evaluated with CCDC80 as a promising mRNA candidate in native tissue.

Testing dose response effects of compounds in the 96-well cartilage organoid model

To determine most effective dose response of the chondroprotective effect and duration of incubation of three different formulations (Cpd-A, Cpd-B, Cpd-C), we employed our 96-well cartilage organoid model (Supplementary table S1A). These compounds were developed to attenuate THR activation by hampering the binding of T3 to its receptor, for each compound, four different concentrations (ranging between 0.1 and 10 µM) and two incubation times (4 days and 7 days) were tested in the neo-cartilage organoid model. As readout, the consistent chondrocyte terminal differentiation marker CCDC80 and the commonly used early chondrocyte hypertrophy marker COL10A1 were measured on RNA level and secreted MMP3 protein levels were measured by ELISA. As shown in Fig. 4 and Supplementary table S3, all compounds demonstrated a dose-dependent reduction in CCDC80 gene expression, with the strongest effect observed at a 10 µM concentration after 4 days of incubation (cpd-A: beta=-1,8 and p = 0,04, cpd-B: beta=-1,3 and p = 0,02, cpd-C: beta=-1,5 and p = 0,01, Supplementary table S3). Extending incubation to 7 days reduced this dose-response effect, suggesting these compounds may reach peak effectiveness within 4 days. This indicates that a 7-day incubation period does not provide additional benefits over 4 days, and that 10 µM for 4 days yields the maximum compound effectiveness in this model. Since increased MMP3 production was seen as a marker of healthy matrix turnover, we explored the secreted MMP3 protein expression in the hypertrophic high-throughput neo-cartilage model by ELISA. As shown in Fig. 4C, a favorable dose-dependent increase in secreted MMP3 was observed after 4 days of incubation with Cpd-A and Cpd-B. For Cpd-C, an increase in MMP3 was shown only with low concentrations (1 and 3 µM). With 7 days incubation time we observed a larger dose responsive effect with Cpd-A. However, Cpd-B and C showed a slight decrease in MMP3 secretion. Together, we concluded that in our high-throughput neo-cartilage organoid model of T3 induced terminal maturation, CpdA at a concentration of 10 µM and an incubation of 4 days, showed the most potent CCDC80 mRNA induction and MMP3 protein excretion response and will be put forward for further validation in our human ex vivo explant model .

Fig. 4.

Dose response effects of compounds in the 96-well cartilage organoid model: Different concentrations of Cpd-A, Cpd-B and Cpd-C on CCDC80 and COL10A1 gene expression with A, 4 days incubation time and B, 7 days incubation time in the 96-well neo-cartilage organoid model. C, Secreted MMP3 protein expression measured with ELISA with 4 days compounds incubation time on the left and 7 days compound incubation times on the right. Compound concentrations are listed below the graphs in µM, -=not added, 0.3 = 0.3µM, 1.0 = 1µM, 3.0 = 3µM, 10 = 10µM, Primary articular chondrocytes were isolated from 1 donor, resulting in n = 1–2 samples, per condition

Validation in the ex vivo osteochondral explant model

Next, we set out to validate the most promising formulations Cpd-A and B, in our previously established human ex vivo biomimetic OA model of T3 induced terminal chondrocyte maturation. Hereto, we incubated the explants with 10 and 20 µM cpd-A and 20 µM B for 7 days. We evaluated the restorative effect of the compounds at the mRNA level using CCDC80, PHOSPHO1 and ANKH as a markers for terminal maturation whereas ATOH8 was measured as marker of pre-hypertrophy, and COL10A1 as a marker of early hypertrophy. Secreted MMP3 protein levels were measured by ELISA as a high-throughput marker of T3 induced terminal chondrocyte maturation. Since the incubation with two concentrations of cpd-A showed no difference (Supplementary Table S4), we combined the two concentrations for further analysis to increase statistical power, visible in Fig. 5. Since cpd-B showed different effects with these doses (data not shown), only 20 µM was used. Similar as in the high-throughput model, after incubation with cpd-A, we observed a recovery back to baseline in terminal maturation markers CCDC80, PHOSPHO1, and ANKH (Fig. 5A and Supplementary Table S5). As expected, we observed no effect in the pre-hypertrophy marker ATOH8 and early hypertrophy marker COL10A1. As seen in Fig. 5A, we observed a similar trend for cpd-B, however less pronounced. Finally, the downregulation of the terminal maturated profile on gene expression level was confirmed on protein level where we observed that treatment with cpd-A also resulted in a recovery of secreted MMP3 protein expression towards baseline (Fig. 5B). Treatment with cpd-B however showed no recovery on protein level. These results demonstrate the promising anti-hypertrophic properties of Compound A in both the high-throughput neo-cartilage organoid model and the biomimetic osteochondral explant model. This highlights the potential of using the high-throughput model as a screening tool for osteoarthritis drug candidates, with the biomimetic osteochondral model serving as a valuable platform for validation.

Fig. 5.

Validation in the ex vivo osteochondral explant model A, Gene expression of hypertrophic genes in the osteochondral explant model after T3 treatment and incubation with cpd-A (10 and 20 µM) and B (20 µM). B, Secreted MMP3 protein expression measured with ELISA. Compound concentrations are listed below the graph in µM, -=not added, 20 = 20µM, ns = not significant, *P < 0.05, **P < 0.01. ***P < 0.001. n = 1–2 samples, N = 2–5 donors

Discussion

In the current study we present a flexible, preclinical testing strategy for potential OA treatments in biomimetic human joint tissues. By miniaturizing the previously established neo-cartilage organoid model, we created a 96-well throughput cartilage organoid model, using primary chondrocytes, that allows for fast initial efficacy dose response testing of multiple drug formulations. For validation of the most optimal concentration with 7 days incubation time, we subsequently exploited the recently characterized human ex vivo biomimetic OA model of T3 induced terminal chondrocyte maturation [1]. In the 96-well throughput cartilage organoid of maturation, the compounds showed a beneficial dose response effect on gene level of hypertrophic gene CCDC80 whereas particularly cpd-A also showed a promising effect on secreted MMP3 protein level. Further testing in the hypertrophic human biomimetic osteochondral explant model of cpd-A and cpd-B confirmed restoration of the T3 induced terminal maturation markers CCDC80, PHOSPHO1 and MMP3.

Previously, it was shown that T3 induces terminal maturation of articular cartilage chondrocytes in human ex vivo osteochondral explants, an important hallmark of OA pathophysiology, that was marked among others by the consistent upregulation of CCDC80 [1, 10]. To address the throughput of the explant model, we here miniaturized our 3D cartilage organoid model from 15 ml tubes to 96 well plates and confirmed T3 induction of terminal maturation by consistent upregulation of CCDC80, albeit with a modest effect size. This modest response of CCDC80 in the high-throughput model compared to the osteochondral explant model is likely due to the newly synthesized cartilage matrix in the neo-cartilage organoids. Nonetheless, next to the MMP3 ELISA other mRNA markers that previously responded to T3 displayed a very similar upregulation in the neo-cartilage organoids as in the osteochondral explants. For instance, ANKH, associated with cartilage calcification and terminal maturation, and COL2A1, a marker of chondrocyte activation, were both upregulated. This indicates that the neo-cartilage pellets reach a terminally matured state in response to T3, similar to the ex vivo osteochondral explants. On the other hand, we acknowledge that validation after throughput screening in human native ex vivo, cartilage explant model is recommended to obtain more complete insight in native tissue e.g. regarding CCDC80 expression.

By applying genetic approaches and functional follow up research, we have previously shown that increasing local bio-availability of active thyroid hormone could indeed induce aspects of the OA pathophysiology pathway, a response that was further modified by hyper physiological mechanical stress [1, 3, 4]. Due to a consistent absence of induction of COL10A1, an early hypertrophy marker, we concluded in these studies that T3 particularly induced late hypertrophy and a terminal maturation chondrocyte phenotype [2–4], a phenomena here confirmed.

We showed that different formulations of compounds targeting thyroid signaling are able to counteract the aberrant effect of T3 induced terminal maturation. Specifically, these compounds were designed to reduce THR activation by inhibiting T3 binding to its receptor. The thyroid hormone receptor (THR) belongs to the nuclear receptor family and is located exclusively in the nucleus [11]. THR is a critical player in the hypothalamic-pituitary-thyroid pathway, which is involved in many processes throughout the body like muscle function and bone remodeling [11]. T3 interacts with THR and subsequently regulates gene expression. The current study indicates that reducing local THR activity in OA chondrocytes should be considered a promising, evidence-based treatment option for osteoarthritis. The varying effectiveness of compounds in inhibiting THR activity found in this study, can be attributed to differences in their chemical structures, which causes alterations in the compound-receptor binding effectiveness. Before applying the compounds as treatments for OA, further research is, however, needed e.g. a more thorough insight is necessary in potential side effects regarding the fact that thyroid activity is important for the entire body. In this respect we envision that administering the drug locally via intra-articular injections is necessary e.g. using nanoparticles as we previously tested [4].

We here introduced MMP3 as a consistent inverse marker of T3 induced hypertrophy and terminal maturation, despite its known catabolic role in OA pathophysiology. A likely explanation is that MMP3 is particularly known to reflect increased (reparative) matrix turnover [12] but is not capable of cleaving intact triple helix of collagen type-II [13]. We therefore, advocate that although MMP3 activity may reflect reactive chondrocyte matrix turnover in (early) OA, its role in late-stage OA is reduced, when extensive breakdown of matrix molecules, including triple helix collagen type-II occurs [14].

Preclinical models are the foundation and key for successful clinical trials. Despite this, only a small fraction of therapeutic discoveries transition from preclinical to clinical trials, and even fewer reach market launch, with an approval rate of just 6–7% between 2011 and 2017 [15]. Currently, animal models are still the standard for preclinical research, which leads to a low predictability in humans and many failed clinical trials [16]. Therefore, greater emphasis should be placed on developing human preclinical tissue models to reduce the reliance on animal testing. Since no single human preclinical model can serve as a complete replacement, we here propose a novel drug testing strategy. This strategy starts with the high-throughput neo-cartilage organoid model as a drug screen model after which the complex biomimetic osteochondral ex vivo osteochondral explants allow for validation of the initial findings.

Before applying this drug testing strategy for various OA-related processes such as senescence, the models should be adjusted according to the disease-relevant triggers, such as mechanical loading. Given that OA affects the entire joint, it is important to assess the impact of including additional tissues, such as synovial tissues. One potential method to simulate synovial tissue interactions is to culture the ex vivo osteochondral explants with synovial fluid or OA-conditioned media. In respect to the output of this model, previously, induction of terminal maturation with T3 in neo-cartilage organoids had a detrimental effect on the cartilaginous matrix [3].

A perceived weakness of our model is that we used gene expression as sensitive readout of the changes in phenotypic chondrocyte state. As such the effect on matrix protein levels i.e. the extracellular matrix is not addressed. This generally because there are inherent difficulties with fierce antigen retrieval in cartilage and more specifically because maintaining the small neo-cartilage organoids throughout the dehydration, embedding, and sectioning protein staining of matrix proteins too challenging and failed. Additionally, we feel that the here implemented quantitative ELISA is more suitable for a high-throughput testing system, than subjective histology or immunohistochemistry. Finally, the neo-cartilage organoids were generated by harvesting primary chondrocytes from preserved articular cartilage of patients undergoing joint replacement surgery. It would therefore be beneficial to repeat the current experiment with more donors. However more important, replacing the primary cells by hiPSC derived chondrocytes [17] as stable ‘off the shelf’ cell source would allow further upscaling and wide clinical application of our model. Notable in this respect is the fact that currently robust hiPSC chondrogenic differentiation protocols exist that result in high quality neo-cartilage constructs [18].

Together, we have shown a potential innovative preclinical testing strategy, using human OA joint tissues. Herein, flexible dose response and duration screening of the compounds were performed in a miniaturized 96-well human cartilage organoid model, whereas validation of most consistent conditions were performed in the (fully characterized) human ex vivo biomimetic OA model of T3 induced terminal chondrocyte maturation. As proof-of-concept study, several formulations were tested that target thyroid signaling, which confirmed that attenuating thyroid signaling could be a possible evidence based treatment option. This ultimately to alleviate the burden of OA thereby increasing mobility, well-being and quality of life particularly among elderly.

Authors’ contributions

Conceptualization: IB, IM, NK, MT, KS, BC. Funding acquisition: IM, BC. Methodology: IB, NK. Resources: IM, RM, RN, BC, KS. Supervision: IM, Visualization: IB. Writing—original draft: IB, IM, Writing—reviewing & editing: IB, IM, NK, KS, RM, RN, BC, MT. All authors read and approved the final manuscript.

Funding

The study was part of Health~Holland, Top Sector Life Sciences & Health TKI grants aiming to stimulate public-private partnerships. Herein, FOACUS research collaboration project was co-funded by Galapagos in a private-public partnership (project number LSHM19036). Additional funding was obtained from VOILA – SMARTage (nr LSHM18093). Overall, data is generated within the scope of the Medical Delta programs Regenerative Medicine 4D: Generating complex tissues with stem cells and printing technology and Improving Mobility with Technology. The funding bodies had no influence on study design, reporting or the decision to submit the manuscript for publication.

Data availability

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

Declarations

Ethics approval and consent to participate

The RAAK study has been approved by the medical ethical committee of the Leiden University Medical Center (P08.239/P19.013). Written informed consent was obtained from all patients, and patients had the right to withdraw at any time.

Consent for publication

Conceptualization: IB, IM, NK, MT, KS, BC.

Funding acquisition: IM, BC.

Methodology: IB, NK.

Resources: IM, RM, RN, BC, KS.

Supervision: IM,

Visualization: IB.

Writing—original draft: IB, IM,

Writing—reviewing & editing: IB, IM, NK, KS, RM, RN, BC, MT.

All authors read and approved the final manuscript..

Competing interests

The authors declare no competing interests.