Donor ethnicity, sex, and age impact chondrogenic re-differentiation capacity: a multi-demographic study of human articular chondrocytes in vitro

ABSTRACT

To unravel causes of disparities in osteoarthritis (OA) prevalence and to improve cell-based cartilage repair treatments, there is need to investigate the multifactorial impact of patient demographics on a biophysiological level. In this study, we systematically analyse single- and multi-demographic impact on in vitro chondrogenic re-differentiation capacity of human articular chondrocytes (hACs), specifically of an Aotearoa-New Zealand (NZ) patient cohort which displays unique demographic diversity. HACs were isolated from 14 NZ donors with distinct demographics (ethnicity: indigenous Māori vs European descendant Pākehā; sex; age 18–24 vs 25–30 yrs). In vitro chondrogenic re-differentiation capacity of donor chondrocytes was assessed through quantifications of cartilage matrix deposition (GAG, DNA, GAG/DNA) and by histological visualisation. Isolated chondrocytes ranged from low chondrogenic re-differentiation capacity, characterised by fibrocartilage tissue deposition, to high re-differentiation capacity to deposit GAG-rich tissue. Age-related reduction in GAG/DNA content was detected while no impact of ethnicity or sex was observed. Multi-demographic analysis revealed reduced GAG deposition in Māori male (compared to Māori female), and Māori (18–24 yrs) compared to Māori (25–30 yrs) and Pākehā (18–24 yrs) within our cohort. Multi-demographic analysis is a promising strategy to understand disparities in OA prevalence in patient cohorts and can help guide development of cell-based strategies for diverse patient populations.

Introduction

Osteoarthritis (OA) affects approximately 10% of adults in Aotearoa-New Zealand (NZ), and over 527.8 million patients globally (Deloitte Access Economics 2018; Lao et al. 2019; Long et al. 2022). Natural ageing is hypothesised to cause OA, while patients with untreated chondral injuries (76.2% of the patients in NZ) also pose a higher risk to develop post-traumatic OA (Mora et al. 2018; New Zealand ACL Registry Annual Report 2019; Whittaker et al. 2022). Despite the growing prevalence of OA, treatment options remain scarce. Patients with symptomatic OA in weight-bearing joints often receive first-line treatment options, however, these therapies do not prevent further disease progression (Yu and Hunter 2015; Zhang et al. 2016; Mora et al. 2018; Martín et al. 2019; New Zealand ACL Registry Annual Report 2019). Eventually, patients with end-stage OA receive a joint replacement surgery which does not guarantee full joint function restoration and is often associated with complications (Noble et al. 2005; Lützner et al. 2011).

Cell-based cartilage repair strategies offer promising alternative treatment options. Clinically available cell-based strategies include autologous chondrocyte implantation (ACI), or matrix-induced autologous chondrocyte implantation (MACI) where the patients’ own chondrocytes are isolated, expanded and re-implanted (Barlett et al. 2005; Peterson et al. 2010; Makris et al. 2015). Although ACI and MACI are promising, these procedures are not consistently successful in the clinic. Fibrocartilage-like scar tissue, instead of hyaline articular cartilage, is predominantly formed which reduces its biological function, mechanical properties and subsequently its clinical potential (Barlett et al. 2005; Steinert et al. 2007; Khan et al. 2008; Pareek et al. 2016; Martín et al. 2019). To overcome these challenges, advanced biomaterials are being developed to create bioactive and biomimetic scaffolds, and the potency of various cell types (including mesenchymal stromal cells and induced pluripotent stem cells) is widely explored (Aldrich et al. 2021; Makris et al. 2015; O’Shea et al. 2022; Zhou et al. 2022). However, the success rate of cell-based interventions for OA remains highly dependent on the chondrogenic re-differentiation capacity of chondrocytes (or other cell types) encapsulated in the cartilage defect in both conventional and contemporary cell-based strategies (Steinert et al. 2007). It is believed that demographic factors including ethnicity, sex and age may impact the chondrogenic re-differentiation capacity of chondrocytes. It has previously been shown that chondrogenic re-differentiation capacity reduces with age, however, the impact of other demographics (sex and ethnicity) has been largely overlooked (Degroot et al. 1999; Barbero et al. 2004). Study of ethnicity, in particular indigenous populations, has been inadequate and is only recently an emerging area of research, while variations in OA prevalence in populations with diverse ethnicity have been consistently reported on (Hurd and Barnabe 2017).

The Aotearoa-New Zealand population is an example of a cohort with unique ethnic diversity: 16.5% of New Zealanders identifying as indigenous Māori, 9% Pasifika, 15.1% Asian, and 70.2% European (Pākehā) descendants (EHINZ 2023). Furthermore, the NZ patient population is unique for its variation in OA prevalence across patient demographics, as detailed below. Pākehā are reported to have the highest prevalence of OA in the NZ population (12.5%), compared to Māori (7.1%), Pasifika (4.8%), and Asian (2.5%) populations (Deloitte Access Economics 2018; Lao et al. 2019). With a focus on the two largest ethnic groups in NZ, Māori and Pākehā, the Māori OA patient population is younger than the Pākehā OA patient population (Singleton et al. 2013b; Deloitte Access Economics 2018). In addition, Māori males uniquely displayed a higher prevalence of OA compared to females of all age groups, apart from those aged between 65 and 74 years. This is in contrast with the higher prevalence of OA reported for females compared to males globally, and within the sub-ethnic Pākehā, Asian and Pasifika populations in NZ (Deloitte Access Economics 2018). Considering the NZ specific trends in OA prevalence, it is of interest to undertake studies that disentangle the cause of the observed disparities within the NZ population. The described OA trends reveal that multiple demographics may play a role in the incidence rate of OA in NZ. However, the impact of demographics on cellular performance is often solely evaluated for a single demographic, which we refer to as single-demographic analysis. The evaluation of multiple demographics on chondrogenic re-differentiation capacity simultaneously, which we define as multi-demographic analysis in this study, is thus clinically relevant and is to the best of our knowledge rarely seen in literature.

This study aims to broaden our knowledge on the impact of single- and multi-demographics on chondrogenic re-differentiation capacity of human articular chondrocytes (hACs) in vitro. By systematically studying the influence of ethnicity, sex and age on chondrogenic re-differentiation capacity, we seek to specifically correlate our in vitro outcomes from Aotearoa-New Zealand indigenous patients with the reported prevalence disparity of OA in NZ. We hypothesise that there is a significant difference in the in vitro chondrogenic re-differentiation capacity of articular chondrocytes isolated from healthy cartilage of early adulthood (18–30 years) orthopaedic patients across different ethnic, sex and age demographic groups in New Zealand. This approach also places emphasis on the importance of studying the effects of multiple demographics on chondrogenesis to develop cell-based strategies that cater to a broader patient population. To evaluate chondrogenic re-differentiation capacity in a NZ patient cohort, hACs were isolated from 14 NZ donors with distinct demographics (ethnicity, sex and age) and evaluated for their in vitro chondrogenic re-differentiation capacity. Cartilage microtissues were fabricated using standard centrifugation techniques, cultivated, and evaluated for their cartilage matrix formation, both quantitatively and qualitatively, and analysed using both single- and multi-demographic analysis.

Methods

Chondrocyte isolation

Human articular cartilage tissue was harvested from macroscopically healthy regions of the knee articular surfaces from patients receiving anterior cruciate ligament (ACL) reconstruction (New Zealand Health and Disability Ethics Committees, URB/07/04/014/AM02). The ACL reconstructive surgery was performed by the same surgeon and cartilage was obtained from a similar location, the non-load-bearing intercondylar notch region, to strive for consistent cartilage samples between patients. All patients were undergoing surgery 5–12 months post-injury, and no additional joint disorders, such as osteoarthritis (OA) or meniscal damage, were documented besides ligament damage. Primary chondrocytes were isolated as described previously (Lindberg et al. 2019, 2021). Isolated chondrocytes were seeded at 3,000 cells/cm² in chondrogenic base media (Gibco® DMEM high glucose glutaMAX^TM with 10 mM HEPES, 0.4 mM L-Proline, 0.1 mM NEAA, 0.1 mg/ml streptomycin, 100 U/ml penicillin, supplemented with 10% FBS, and 0.1 mM AsAp) and incubated (37 °C, 5% CO₂) to allow exponential growth in high-density monolayers. The cells were passaged and consistently collected at passage 3 for in vitro characterisation. In this study, 14 donors with different donor demographics (ethnicity, sex and age) were isolated, collected and evaluated for their chondrogenic re-differentiation capacity (Table 1). Donor demographics were obtained from patient self-reports during the consent process for cartilage biopsy approval. Patients who listed more than one ethnicity were not included in this study.

Table 1.

Detailed demographics (A) and the demographic distribution (B) of hAC donors of the Aotearoa-New Zealand patient population.

A
Māori donors			Pākehā donors
Identifier	Age	Sex	Identifier	Age	Sex
1	22	M	8	28	M
2	30	F	9	27	M
3	22	F	10	18	F
4	23	M	11	25	F
5	20	F	12	22	M
6	27	M	13	25	M
7	20	F	14	22	M
(M = male, F = female).

B
Demographic		Number of donors
Ethnicity	Māori	7
	Pākehā	7
Sex	Female	6
	Male	8
Age	18–24	8
	25–30	6

Microtissue fabrication

Microtissues were fabricated through centrifugation, a commonly used method to fabricate in vitro cartilage spheroids (Schon et al. 2012). 250,000 cells were resuspended per well in a v-bottom 96 well-plate and centrifuged at 200 g for 4 min. After incubation overnight (37 °C, 5% CO₂), microtissues were detached from the plates and base media was replaced with chondrogenic differentiation media (Gibco® DMEM high glucose glutaMAX™ with the addition of 0.4 mM L-proline, 10 mM HEPES, 0.1 mM NEAA, 100 U/ml penicillin, 0.1 mg/ml streptomycin, 0.2 mM AsAp, 1x ITS + 1 premix, 1.25 mg/ml BSA, 10 nM dexamethasone, and 10 ng/ml TGF-β1). Microtissues were cultivated for two weeks (37 °C, 5% CO₂), chondrogenic differentiation media was refreshed three times a week.

Microtissue characterisation

Following two weeks of chondrogenic culture, microtissues were harvested for qualitative and quantitative analysis. Brightfield images of the microtissues were made (Olympus M-D E-M10 camera) and processed using built-in functions in ImageJ software (version 1.6.0_24) to determine the average diameter (mm). For qualitative histological analysis, microtissues were washed in PBS (3x), fixed in 10% neutral buffered formalin (1 h), washed in PBS, and stored in sterile PBS at 4 °C prior to downstream use. Collection for quantitative analysis required similar washing in PBS (3x) and microtissues were stored at −80 °C until required for analysis.

Histological analysis was performed to visualise the presence and distribution of glycosaminoglycans (GAGs) and tissue organisation. Formalin-fixed samples were embedded in tissue freeze medium (Tissue-Tek®) and were subsequently cryo-sectioned into 30 µm tissue sections (Leica Biosystems cryostat microtome CM1860). Histological sections were collected onto adhesive microscope slides and stained with haematoxylin, fast green, and Safranin-O to allow the visualisation of cell nuclei, collagen, and GAGs respectively. Visual evaluation of Safranin-O stained sections has been performed using the Bern score guidelines to indicate the uniformity and intensity of Safranin-O stain, cellular spreading, and cellular morphology (Grogan et al. 2006). Three to six representative histology images, obtained from three microtissues, were scored for each donor. Evaluation was performed blindly by three individuals with experience in analysing histology staining, resulting in a total number of 9–18 sections evaluated per donor.

To identify the collagen types I and II content and its localisation within the microtissues, immunohistochemistry (IHC) techniques were used. In brief, IHC sections were incubated in 0.2% hyaluronidase (30 min, 37 °C) to reveal antigen epitopes, washed with PBS, blocked with 2% BSA and 0.3M glycine in PBS (30 min, RT) followed by incubation with primary antibodies for collagen type I (1:200, AHP0022) and collagen type II (1:30, II-II6B3-C) diluted in blocking buffer (3 h, RT). Samples were washed three times in blocking buffer for 10 min, followed by incubation of the secondary antibodies, goat-anti-mouse (Alexa Fluor 488) and donkey-anti-rabbit (Alexa Fluor 594), diluted in blocking buffer (1:400). Additionally, sections were incubated with DAPI diluted in blocking buffer (1:1000, 10 min, RT) to stain the nuclei. Lastly, samples were washed in PBS and left hydrated and protected from light until imaging. All histological and IHC samples were visualised using the Zeiss Axioimager Z1 microscope.

Furthermore, collagen II/I ratio was quantified by quantifying the fluorescent intensity of both collagen type II (green channel) and I (red channel) using Image J software (version 1.6.0_24). A region without any cells was selected and used to subtract background fluorescence. The corrected fluorescence intensity (CFI) was calculated by Equation (1), where ID is the integrated density and A is the area of the microtissue.

$$\mathrm{CFI}={\displaystyle \frac{\mathrm{ID}\mathrm{microtissue}}{A\mathrm{microtissue}}-{\displaystyle \frac{\mathrm{ID}\mathrm{background}}{A\mathrm{background}}}}$$ (1)

The collagen II/I ratio was obtained by dividing the CFI of collagen type II by the CFI of collagen type I.

For quantitative analysis, microtissues were digested in 200 µL proteinase K solution (1 mg/ml dissolved in 10 mM Tris-HCl and 1 mM disodium EDTA solution). Biochemical analysis of the digested samples was conducted using a Dimethylmethylene Blue Assay (DMMB) assay for GAG quantification, and Cyquant Cell proliferation kit for DNA quantification. To assess the GAG content in tissue constructs, 50 µL of the (diluted) digested samples was combined with 200 µL DMMB solution (16 µg/ml, pH 3), followed by absorbance reading at 520 nm using a Multi-Mode Microplate Reader (SpectraMax iD3, Molecular devices). To detect the DNA content, the digested microtissues were pre-treated in an RNAse lysis buffer and incubated with a light-sensitive GR dye followed by a fluorescence reading measured at 520 nm (emission) and 480 nm (excitation) using the Multi-Mode Microplate Reader. Overall, the GAG and DNA present in the tissue constructs was plotted and GAG/DNA ratios were calculated. All biochemical measurements (GAG, DNA and GAG/DNA) were normalised to the GAG, DNA and GAG/DNA content detected in randomly chosen reference microtissues fabricated with chondrocyte donor 3. This additional normalisation step has been performed to allow for the pooling of data across multiple experimental repeats and biological donors. Scaling to a common baseline mitigates traditional challenges with technical variability due to varying potency of media components, e.g. FBS and TGF-β between repeats.

Statistical analysis

For every biological donor (n = 14), microtissues were prepared in duplicates or triplicates and three technical replicates were performed to study variability between tissue samples (n  =  6–9). Data of individual donors within the same category (ethnicity, sex or age) were pooled to identify the impact of donor demographics on chondrogenic re-differentiation capacity (n = 36–72). Quantitative differences in inter-donor chondrocyte re-differentiation capacity were assessed by one-way ANOVA with post-hoc Tukey analysis (GraphPad Prism 9), while differences between two groups within a demographic (for example Māori vs Pākehā) were assessed by performing a t-test. In all cases, datapoints were checked for outliers (ROUT method, Q = 1%) and assumptions of parametric analysis were validated using Shapiro–Wilk (n < 50) and Kolmogorov–Smirnov (n > 50) for normal distribution, and Brown–Forsythe to test for equal variance. Kruskal–Wallis One-Way ANOVA and Mann–Whitney U Tests were adopted for samples populations with non-gaussian distributions, while Brown–Forsythe and Welch ANOVA tests with Dunnett’s Method or t-tests with Welch’s correction were used for all populations with unequal variation. All values are reported as the mean including error bars with standard error of the mean. A selection of significant differences is illustrated in the figures and P-values are presented numerically in additional tables in supplementary information.

Results

In this study, we investigated the effect of three demographic factors (ethnicity, sex and age) on the chondrogenic re-differentiation capacity of hACs obtained from a NZ specific patient population. The 14 donors were equally distributed in the three demographics of interest as depicted in Table 1(B). The ethnicity of the donors was divided into two groups: the indigenous Māori donors and the NZ-European Pākehā donors. The age of the evaluated donors ranged between 18 and 30 and was divided in age groups from 18–24 and 25–30 years. This reflects a relatively young group of patients that have suffered traumatic knee injuries (anterior cruciate ligament tears) and subsequently have an elevated risk to develop OA. It was furthermore confirmed that all isolated cells were mycoplasma-free and thus suitable for in vitro evaluation (Table S1).

Microtissue size

After 14 days of culture, dense, spherical microtissues were obtained with variation in microtissue size observed across the donor pool (Figure 1(A)). Microtissues fabricated from donors 2 (Mā-30-F), 3 (Mā-22-F), 10 (Pā-18-F) and 14 (Pā-22-M) were bigger in size compared to other donors (numerical P-values are presented in Table S2). Sex-based analysis revealed that microtissues fabricated with female donors were bigger in size compared to male microtissues (P < 0.0001), while no variation was seen across ethnicity and age (P = 0.12 and P = 0.67 respectively) (see Figure 1(B–D)). Despite the differences in size, no difference in metabolic activity was observed between microtissues (Figure S1).

Figure 1.

Quantification of the average diameter (in mm) of microtissues (A). The impact of donor demographics on the microtissue size is presented in B (ethnicity), C (sex) and D (age). Error bars represent the mean ± SE of at least 6 samples in A, and at least 36 samples in B, C and D. Significant differences (P < 0.05) are illustrated by displaying an asterisk. In graph A, significant differences compared to reference donor 3 are displayed only.

Chondrogenic re-differentiation capacity and the impact of single donor demographics

To analyse the inter-donor variability in tissue formation across the donors, cartilage matrix deposition was assessed both quantitatively and qualitatively. Representative Safranin-O and fast green stained sections for each individual donor are presented in Figure 2(A). Microtissues fabricated with chondrocytes isolated from donor 3 (Mā-22-F) and 10 (Pā-18-F) displayed homogeneous and positive Safranin-O staining which indicated successful GAG deposition throughout the entire microtissue. Although donors 6 (Mā-27-M), 8 (Pā-28-M) and 14 (Pā-22-M) displayed positive staining for Safranin-O, only localised deposition of GAGs was observed in the core. Microtissues fabricated with the remaining donors lacked positive staining for Safranin-O, a sign of fibrocartilage rather than hyaline cartilage tissue formation. Evaluation of the Safranin-O stained sections, utilising the Bern score guidelines, revealed similar results to the GAG quantification. Donor 3 and donor 10 display the highest Bern scores, with a significant higher score compared to all remaining donors in this patient cohort: see Figure 2(B). Furthermore, collagen type II and I deposition was visualised via IHC staining (see Figure 2(C)). All microtissues deposited both collagen type I and II, with differences in localisation of the collagens. For example, donor 3 shows high collagen type II expression in the core of the tissue, while collagen type I rich matrix is seen at the periphery. However, no clear differences in the overall amount of collagen deposition between donors were observed, with IHC image analysis, as confirmed by the lack of significant differences in quantified collagen II/I ratio (see Figure S2). Chondrogenic re-differentiation capacity of each individual donor was further assessed by quantifying GAG and DNA content, and subsequent GAG/DNA ratio (Figure 3(A–H)). Due to variation in absolute values in quantification across experimental repeats, GAG, DNA and GAG/DNA content was normalised (see Figure S3). Although no significant differences in DNA content were observed between the individual donors, a clear variation within the patient cohort was observed (Figure 3(A)). By collating the data for each category, a higher DNA content was observed in donor cells isolated from patients within the age-range 25–30 compared to 18–24 (P = 0.0042) (Figure 3(B–D)). No significant difference was observed between donors with different ethnicity and sex. Quantification of the GAG content in the microtissues revealed that donor 3 deposited a higher GAG content compared to all other donors, except donor 2 (Mā-30-F), 10 (Pā-18-F) and 14 (Pā-22-M) (Figure 3(B)). Numerical P-values to state significance differences in GAG content are presented in Table S2. When comparing GAG deposition within the donor demographics (Figure 3(F)), it was observed that microtissues fabricated from male donors deposited a lower amount of GAG content compared to female donors (P = 0.0005). The GAG/DNA ratio of microtissues – a common measurement of chondrogenic re-differentiation capacity – was further investigated, revealing a reduced re-differentiation capacity with age (P = 0.0086) (see Figure S4D). In general, donors who displayed low GAG content and/or GAG/DNA ratios deposited fibrocartilage-like tissue as displayed by the lack of, or heterogeneous, Safranin-O staining.

Figure 2.

Visualisation of Safranin-O/fast green stained cryosections of microtissues fabricated via centrifugation with 14 donors after 14 days in chondrogenic media, scalebar = 250 µm (A). Evaluation of the Safranin-O stained sections utilising the Bern score evaluation guidelines. Error bars represent the mean of ± SE 8-16 evaluated images. Donor 3 and donor 10 have a significant higher Bern score (P<0.05) compared to all other donors, indicated by an asterisk (*) (B). Visualisation of collagen type II (green) and collagen type I (red) in microtissues utilising immunohistochemical staining, scalebar = 250 µm (C).

Figure 3.

Quantification of microtissues fabricated via centrifugation with individual donors 1–14 (0.25 × 10⁶ chondrocytes/spheroid) after 14 days in chondrogenic culture. Quantification of DNA (A) and GAG (B) content of microtissues fabricated with each individual donor. DNA content pooled and presented per demographic: ethnicity (C), sex (D) and age (E). GAG content pooled and presented per demographic: ethnicity (G), sex (HI) and age (IJ). Error bars represent the mean ± SE of at least 6 samples (Fig. A&B) or at least 36 samples (Fig. C–H). Significant differences (P < 0.05) are illustrated by displaying an asterisk. In graph A and B, significant differences compared to reference donor 3 are displayed only.

Multi-demographic analysis

Systematic categorisation of donor demographics enabled pairwise multi-demographic analysis, revealing additional correlations in both GAG and DNA content (see Figure 4). In the ‘ethnicity & age’ category, it was identified that younger Māori (18–24 yrs) donors have a lower GAG content in fabricated microtissues compared to older Māori (25–30 yrs) as well as young Pākehā (18–24 yrs) donor chondrocytes (P = 0.0282 and P = 0.015 respectively). Evaluation of the ‘ethnicity & sex’ category displayed a lower GAG content in Māori male donors compared to Māori female donors (P = 0.0025). Lastly, both DNA and GAG content was also impacted by the combined ‘age & sex’ category. A significantly lower DNA content was observed in younger male (18–24 yrs) chondrocyte donors compared to older male (25–30 yrs) donors (P = 0.0036). The GAG content in microtissues fabricated with chondrocytes isolated from younger female (18–24 yrs) donors was significant higher compared to male donors of the same age (P = 0.025). Interestingly, no differences were observed in the overall GAG/DNA ratios.

Figure 4.

Quantitative data for DNA, GAG and GAG/DNA pooled together to identify the effect of multiple demographics: ethnicity & age (A–C), ethnicity & sex (D–F) and age & sex (G–I). Error bars represent the mean ± SE of the 6-9 tissue samples collated from each donor fitting the allocated demographic category. *Indicates significant difference (p < 0.05).

Discussion

Multi-demographic analysis is a promising strategy to understand disparities in OA prevalence in patient cohorts and can help guide development of cell-based cartilage repair strategies. Numerous studies have investigated age-related chondrocyte performance in vitro, however, other demographics such as ethnicity, specifically indigenous patient populations, are often neglected (Degroot et al. 1999; Barbero et al. 2004). Moreover, demographic studies are often limited to single-demographic analysis. In this study, we investigated the impact of three donor demographics (sex, ethnicity, age) on the chondrogenic re-differentiation capacity in vitro, via single- and multi-demographic analysis. Furthermore, we compared in vitro outcomes with in vivo prevalence of OA within an Aotearoa-New Zealand specific cohort to enhance understanding of the demographic-dependent disparity in prevalence in the NZ OA patient population. To the best of our knowledge, this has not been studied from a biophysiological context previously.

Donor-dependent variation in chondrogenic re-differentiation capacity amongst isolated primary human articular chondrocytes was observed across 14 NZ donors. In general, isolated chondrocytes with high chondrogenic re-differentiation capacity are able to deposit cartilage-specific extracellular matrix (ECM) molecules, since their chondrogenic phenotype is maintained (Dreier 2010; Hall 2019). In this study, microtissues fabricated with donors 3 (Mā-22-F), 10 (Pā-18-F) and 14 (Pā-22-M) were classified as donors with high chondrogenic re-differentiation capacity as evidenced by successful GAG deposition, both quantitively and qualitatively. Cells with low chondrogenic re-differentiation capacity are unable to preserve their chondrogenic phenotype and instead de-differentiate towards a fibroblastic or hypertrophic phenotype. These chondrocytes have limited ability to deposit cartilage ECM and will deposit fibrocartilage-like tissue instead (Dreier 2010; Hall 2019). In this study, donor 1 (Mā-22-M), 4 (Mā-23-M), 7 (Mā-20-FM), 12 (Pā-22-M) and 13 (Pā-25-M) were identified as donors with poor chondrogenic re-differentiation capacity.

In addition to the need for chondrocytes to deposit cartilage ECM, cellular survival as well as cellular aggregation and compaction of the donor cells is of high importance since this process replicates the condensation mechanisms during normal cartilage development (Yang et al. 2019). DNA quantification after cultivation can help us identify the success of cellular compaction. In this study, a clear variation in DNA content in microtissues was detected between donors. While chondrocytes may have succeeded in initial cellular aggregation, displayed by successful spheroid formation, they may fail to mature into dense cartilage tissue spheroids. Successful cartilage spheroid formation initially depends on inter-cellular connections facilitated by N-cadherins (Lin and Chang 2008; Gionet-Gonzales and Leach 2018; Kouroupis and Correa 2021). Subsequently, integrins and endogenous ECM molecules produced by chondrocytes will further support cellular adhesion. Aggregated cells that achieve successful inter-cellular adhesion will undergo compaction resulting in the stabilisation and maturation of the cartilaginous microtissue architecture (Lin and Chang 2008; Gionet-Gonzales and Leach 2018; Kouroupis and Correa 2021; Lindberg et al. 2021). Limited N-cadherin expression or limited deposition of integrins and ECM molecules by the donor chondrocytes may cause cellular loss before cellular compaction was initiated. Therefore, microtissues with both a smaller size and a reduced DNA content, such as donor 1 (Mā-22-M), 4 (Mā-23-M) and 6 (Mā-27-M), may be reflective of poor inter-cellular adhesion formation and will limit suitability for cell-based strategies.

Single-demographic analysis displayed a higher DNA and a lower GAG/DNA content in microtissues fabricated with chondrocytes isolated from older patients (25–30 yrs) compared to younger patients (18–24 yrs). Initially, a higher DNA content with age seems to deviate from existing literature since age-related reduction in cell yield, proliferation and/or cartilage tissue synthesis has been previously observed in other global studies (Degroot et al. 1999; Barbero et al. 2004). However, other literature has also shown that chondrocytes in a proliferative state are unable to shift to a differentiated phenotype, which would explain the higher DNA content in the microtissues while having a lower GAG/DNA ratio (Tallheden et al. 2005) in our patient cohort. Besides, age-related changes are not always detected in the form of reduction in chondrogenic re-differentiation capacity. For example, no differences in GAG/DNA ratio or chondrogenic gene expression were reported between donors in different age categories (Barbero et al. 2004; Jeyakumar et al. 2017). The variation in outcomes across studies may be explained by the differences in the passage number of the chondrocytes during microtissue fabrication. Expanding isolated chondrocytes is proven to affect the re-differentiation capacity of donor cells, with chondrocytes in higher passages typically displaying reduced capability to re-differentiate towards a chondrogenic phenotype following microtissue fabrication (Hamilton et al. 2005; Sun-Woong Kang 2007). In our study, we consistently fabricated all microtissues with chondrocytes expanded to passage number 3, which corresponds to a range of 7.5 and 11.7 population doublings, to eliminate variation in expansion as a contributing factor. Passage 3 was chosen specifically to prevent loss of chondrogenic re-differentiation capacity of the isolated cells, while obtaining a high cell yield in parallel. Previous studies have shown that chondrocytes maintain telomerase activity and telomere length when cultured for three passages with 10.97 cell divisions, while exhibiting telomere lengths of 10kbp which is well above what is commonly known to correlate with senescence, which is below 4kbp (López-Alcorocho et al. 2019; Al-Masawa et al. 2020). The ability of chondrocytes to maintain their re-differentiation capacity across the investigated passages and population doublings is further supported by literature reporting that chondrocytes express similar levels of aggrecan and collagen type II expression between passages 2 and 4, with cells isolated from both OA and healthy cartilage tissue (Rikkers et al. 2021). Expansion to passage 3 allowed us to obtain a high enough yield necessary to fabricate a high number of dense cell-based constructs (250,000/microtissue). Besides, obtaining a high yield of autologous chondrocytes through cellular expansion is crucial for in vivo cell-based strategies to create cell-based constructs for clinically relevant, big defect sizes, as already applied in cartilage repair strategies in the clinic such as ACI (Roelofs et al. 2013; Minas et al. 2016).

The second demographic of interest, sex, appears to impact both the microtissue size and GAG secretion. Microtissues from male donors were smaller in size and lower in GAG content compared to female donors, suggesting a better chondrogenic tissue formation capacity in vitro by female donors. While some literature has suggested that male donor cells have more potential in cell-based strategies (Kreuz et al. 2013; Patel et al. 2023), it is important to consider it from a hormone status perspective. Literature describes that endogenous hormones and reproductive factors may play a role in the pathogenesis of osteoarthritis (Hussain et al. 2018). Given the age range investigated in this study (18–30 years), the influence of oestrogen in female donors may be present. Prior research suggests that the presence of oestrogen in women before menopause can protect against cartilage damage (Hughbanks et al. 2021), whereas a reduction in re-differentiation capacity of chondrocytes from female donors has been suggested to occur due to hormonal changes (e.g. lower oestrogen) associated with menopause (Lou et al. 2016; Patel et al. 2023). Our study findings are consistent with these observations, with a higher chondrogenic performance detected in females as compared to men, at the ages of 18–30 years. It should furthermore be noted that the ethnic background of a patient may also play a role in the hormone profile of the individual. Cervantes et al. (2018) reported differences in gut and pancreatic hormone profiles as well as inflammatory cytokine expression between Māori and non-Māori individuals. Specifically, Māori participants displayed a higher expression of inflammatory cytokines (Cervantes et al. 2018). The direct link between hormone profiles and chondrogenic re-differentiation of chondrocytes of the patient has, to the best of our knowledge, not been investigated yet and will be useful to further identify the impact of donor demographic factors on a biophysiological level.

No differences in chondrogenic re-differentiation capacity were observed between Māori and Pākehā chondrocyte donors. Potentially, the shorter life expectancy of the Māori population compared to Pākehā may play a role in the lower prevalence of OA, evidenced by the fact that after age-standardisation the rate of hip and knee replacement in New Zealand is higher in Māori than Pākehā (Lao et al. 2019; Stats New Zealand (2021, April)). Overall, single demographic analysis of in vitro chondrogenesis did show an age-related reduction in chondrogenic re-differentiation capacity in this NZ patient cohort, however, no impact of sex or ethnicity was identified and thus could not explain the lower incidence of OA observed in the Māori patient population as compared to Pākehā.

To further elucidate if two different demographics were inter-linked, we systematically analysed our donors for multi-demographic impact on GAG, DNA and GAG/DNA. Reduced GAG content in microtissues fabricated from Māori-derived cells in the 18–24 age group (compared to Pākehā donors of the same age) and microtissues fabricated from male Māori donor cells (compared to female Māori donors) are both in line with the observed higher prevalence of OA in young Māori and Māori males (Singleton et al. 2013b; Deloitte Access Economics 2018). These findings suggest that chondrocytes isolated from patients with this demographic profile have a lower capability to deposit cartilage ECM molecules. In healthy conditions, chondrocytes maintain their extracellular matrix by low-turnover replacement. It is suggested that chondrocytes with a lower capability to deposit ECM molecules are unsuccessful in maintaining cartilage homeostasis and those individuals would be more susceptible to develop OA, especially when suffering from a cartilage-related injury (Dreier 2010; Hall 2019). Multi-demographic analysis displayed similar demographic trends in in vitro chondrogenic potential of chondrocytes in the NZ patient cohort and OA prevalence in NZ. For example, chondrocytes isolated from patients with demographics characteristic for high prevalence of OA in NZ, such as young Māori and Māori males, also displayed a low GAG deposition capacity in vitro, implying a close interconnection.

While reflection of inequitable access to care and influence of socioeconomic status on health outcomes is commonly reported in NZ, it is of high interest to also identify the biophysiological impact, in the form of chondrogenic ability of the cells, on post-operative outcomes (Hooper et al. 2014; Hurd and Barnabe 2017; Hartnett et al. 2022). As described by Islam et al., both the in vitro chondrogenesis capabilities of patients’ cells and post-operative outcomes of ACI were identified and patient demographics were listed and compared (Islam et al. 2019). Clinical outcomes of cell-based cartilage repair strategies in NZ are available, however, details about patient demographics are non-existing – limiting the opportunity to compare in vitro studies with clinical outcomes of cell-based strategies (Clatworthy 2017). As a comparison, post-operative outcomes for joint replacement in NZ are widely described, including the impact of patient demographics. For example, increased age-standardised rate of publicly funded hip replacement surgeries was observed in young, male Māori patients, who in addition also presented with worse post-operative outcomes compared to other ethnicities (Singleton et al. 2013a; Lao et al. 2019). Reduced post-operative outcomes could be explained by the quality and execution of exercise-based physiotherapy rehabilitation prior to and post-surgery. Although higher rates of joint replacements are performed in older patients, female patients, Māori patients and patients living in urban areas, a national survey held in Aotearoa-New Zealand indicates that there is no evidence of inequities in rehabilitation based on ethnicity and geographical differences (Snell et al. 2020). Furthermore, the success rate of total joint replacements is suggested to be impacted by a disparity in the utilisation and timing of total joint replacement across patients with distinct donor demographics (Novicoff and Saleh 2011). Māori patients do show worse conditions prior to surgical interventions which could thus play a role in the observed worse post-operative outcomes in these patients (Chua et al. 2020). To determine the impact of patient demographics on post-operative outcomes, it would be important to include the evaluation of biophysiological, socioeconomic factors and pre-surgery disease severity, individually and combined, on the success rate of treatment.

One limitation inherent in this study is the relatively small sample size, consisting of 14 donors. Although significant differences were detectable, further studies involving an increased number of patients (e.g. > 30 of each single- and multi-demographic analysis) would be desired to elucidate wider conclusions from multi-demographic analysis with statistical power (Serdar et al. 2021). With a small sample size, there is an increased risk of type II statistical errors, wherein potentially meaningful effects may go undetected. Moreover, the findings derived from the presented dataset may offer insights specific to the 14 donors investigated, but they may not reflect the broader population trends. While additional studies are needed to assess the applicability of our findings to the broader population, our results underscore the ability of employing spheroid models as in vitro tissue analogues to study chondrocyte re-differentiation capacity across a wide range of patient samples in a diverse patient population.

Furthermore, a factor that may have contributed to unforeseen differences in chondrogenic re-differentiation capacity of human articular chondrocytes, is the potential variation between biopsies. Cartilage biopsies taken for this study were collected from macroscopically healthy joint regions from patients undergoing an ACL reconstruction. However, it should be noted that the collected cartilage samples may have been exposed to varying degrees of inflammatory conditions within the joint microenvironment prior to surgery, which is known to differ significantly among patients with knee trauma and disease (Jacobs et al. 2020; Turati et al. 2021; Kingery et al. 2022). The inflammatory response in the joint can significantly affect the metabolism of chondrocytes, referred to as metabolic reprogramming, which is a key contributor to cartilage degeneration (Arra and Abu-Amer 2023). The timeframe between injury and surgery may thus affect the health of the cartilage (Driban et al. 2014). Notably, all patients included in the study underwent surgery within the initial year following injury, indicating a period that likely encompasses an acute inflammatory phase with a potential transition into a subacute or chronic inflammatory phase for many patients (Lieberthal et al. 2015; Jacobs et al. 2020). Furthermore, the thickness of the cartilage excised may vary between patients in this study. This determines the type of chondrocytes isolated due to the different zonal regions present in cartilage (Fox et al. 2009). Thus, future studies should pursue the characterisation of inflammatory markers within patients’ synovial fluid and the cellular phenotype isolated from macroscopically healthy cartilage regions to complement our findings.

In addition to investigating the three key demographics (ethnicity, sex and age), future studies should consider gathering information about other demographic factors, such as the patients’ diet or BMI, which may contribute to a reduced chondrogenic re-differentiation capacity and a higher risk of OA development. As reported by Sekar et al. (2017), rats fed with a diet high in saturated fatty acids were showing signs of metabolic syndrome and onset of OA. Similar effect was observed in bovine cartilage explant cultures in vitro, where a higher tissue degradation rate was seen under influence of saturated fatty acids (Sekar et al. 2017). Current clinical trends are also indicating that individuals who have obesity have a higher prevalence of osteoarthritis (Fu and Griffin 2014; Johnston et al. 2020). Obesity increases the weight on the load-bearing joints, causing an increase in wear and tear of the cartilage. Interestingly, individuals with obesity not only have a higher risk of OA in load-bearing joints but also in non-load-bearing joints, which indicates that other factors, such as metabolic factors and disorders, play a role in the increase in risk of OA for individuals with obesity (Oliveria et al. 1999; Johnston et al. 2020). The role of obesity in the NZ patient population is especially important since it is known that the Māori population has a higher rate and risk for early onset of obesity, which can contribute to pathologies such as osteoarthritis (Howe et al. 2015; Wilson and Abbott 2019). Further implementation of the donor screening method outlined in this paper, incorporating additional demographic variables such as BMI and dietary habits of the patients, could help identify other demographic factors, and to what extent, that may affect chondrocyte performances on a biophysiological level following a knee trauma.

Future donor screening studies, like that presented in this study, would benefit from increasing the number of donor samples to enhance the statistical power. Moreover, we believe future studies should include investigation of a wide range of patient demographic factors to allow the field to gain a better insight into which factors, and to what extent, they affect chondrocyte performance on a biophysiological level. This would ultimately support development of the next era of cellular-based repair strategies targeting the wider patient population.

Conclusions

Variation in chondrogenic re-differentiation capacity was observed amongst a patient cohort of 14 NZ chondrocyte donors. Single demographic analysis revealed age-related reduction in chondrogenic re-differentiation capacity in chondrocytes. There was no variation in in vitro chondrogenic re-differentiation capacity in chondrocytes originating from donors of different ethnicities (Māori or Pākehā) or sex. Multi-demographic analysis displayed similar demographic trends in in vitro chondrogenic potential of chondrocytes, in this patient cohort, and OA prevalence in NZ. Obtaining insight into these trends using multi-demographic analysis may predict and explain the OA prevalence in patient cohorts. These findings specifically highlight the importance of investigating multifactorial demographic impact since single demographic analysis was not sufficient to elucidate complex individual differences between patients. It further highlights the need to evaluate patient demographics in more depth in order to achieve the ultimate goal of developing effective cell-based treatments for the wider patient population. Systematic multi-demographic pre-screening of isolated donors may support the development of next generation cell-based strategies suitable for the wider patient population, independent of donor demographics.

Funding Statement

The authors acknowledge funding by Health Research Council of New Zealand Emerging Researcher First Grants 19/679 (GL), Explorer 21/802 (GL), and the University of Otago Health Science Postdoctoral Fellowship (GL).

Disclosure statement

No potential conflict of interest was reported by the authors.

Author contributions

GL, LV, VL and TW contributed to the conception and design of the article. Experiments, analysis and interpretation of the data were performed by GL, LV and VL. The article was written by LV and VL. All authors critically revised and edited the manuscript draft and approved the final manuscript.

Supplemental Material

Supplemental data for this article can be accessed online at https://doi.org/10.1080/03036758.2024.2340481.