The variable results in clinical trials of adipose tissue-derived stem cells (ASCs) for chondral defects may be due to the different ex vivo culture conditions of the ASCs which are implanted to treat the lesions. We sought to determine the optimal in vitro chondrocyte co-culture condition that promotes infrapatellar fat pad-derived (IFPD) ASC chondrogenic gene expression in a novel co-culture combination.
In our study, we utilized an in vitro autologous co-culture of IFPD ASCs and articular chondrocytes derived from Kellgren–Lawrence Grade III/IV osteoarthritic human knee joints at ASC-to-chondrocyte seeding log ratios of 1:1, 10:1, and 100:1. Gene expression following in vitro co-culture was quantified by RT-qPCR with a panel comprising COL1A1, COL2A1, COL10A1, L-SOX5, SOX6, SOX9, ACAN, HSPG2, and COMP for chondrogenic gene expression.
The chondrogenic gene expression profiles from co-cultures were greater than would be expected from an expression profile modeled from chondrocyte and ASC-only monocultures. Additionally, chondrogenic gene expression decreased with increasing ASC-to-chondrocyte seeding ratios.
These findings provide insight into the mechanisms underlying clinical ASC therapies and signifies that IFPD ASCs pre-conditioned by chondrocyte co-culture may have improved chondrogenic potential for cartilage repair. This model can help further understand IFPD ASCs in chondral and osteochondral repair and the chondrogenic pathways involved.
The online version contains supplementary material available at
Osteoarthritis is characterized by degradation of articular cartilage, influenced by altered mechanical loading and exacerbated by localized inflammation [
While bone marrow-derived MSCs and umbilical cord blood-derived MSCs have been investigated in clinical trials, they are limited by low relative abundance and ease of isolation respectively. Infrapatellar Fat Pad-Derived (IFPD) ASCs on the other hand exhibit a high relative abundance, ease of isolation, and high proliferation potential [
Our study investigates chondrogenic gene expression in ASCs with a novel in vitro autologous co-culture of early passage (p0) IFPD ASCs and chondrocytes from osteoarthritic knee joints using different ASC-to-chondrocyte seeding ratios. The primary objective of this study was to determine whether and to what extent the chondrogenic gene expression of ASCs is stimulated when co-cultured with chondrocytes. Our study aims to provide insight into the mechanisms underlying chondral repair with ASC therapies and demonstrate the potential presence of ASC-chondrocyte crosstalk in inducing chondrogenic gene expression in ASCs [
Human tissue containing articular cartilage and infrapatellar fat pad from total knee arthroplasties for osteoarthritis (IRAS: 247368, REC: 18/NV/0545, July 2018) was obtained following informed consent and handled in accordance with ethical guidelines outlined by the Human Tissue Act 2004 [
Infrapatellar fat pad samples were bathed in PBS and stored within a sterile container at 4 °C and were processed within 3 h of harvest. ASCs were extracted from the entire infrapatellar fat pad sample by mechanical dissection with a sterile scalpel and enzymatic digestion with 20 mL of sterile 0.2% Collagenase A (Sigma Aldrich) solution (passed through a 0.2 μm filter) made with 40 mg of Collagenase A into 20 mL of Dulbecco’s Modified Eagle Medium (DMEM, ThermoFisher), for 90 min at 37 °C with constant agitation. The digest was eluted through a 70 μm filter. An equal volume of basal medium composed of low-glucose Dulbecco’s Modified Eagle Medium (DMEM, ThermoFisher) supplemented by 10% Fetal Bovine Serum (FBS, ThermoFisher) and 1% penicillin/streptomycin (ThermoFisher) was added and the digest centrifuged at 300 g for 5 min. The supernatant was discarded, and the pellet resuspended in 10 mL of PBS. The resuspended cells underwent the ASC characterization steps described below.
The IFPD cells were expanded to 90% confluence or greater at passage 0 (p0), with media changes every two to three days and were then detached by addition of 2 mL of 1× TryplE Reagent (ThermoFisher), centrifuged and resuspended in 10 mL of 0.2% BSA-EDTA in PBS (ThermoFisher), and stained on ice with monoclonal antibody-fluorophores (Additional file
To determine their tri-lineage potential, aliquots of the cell suspension were cultured separately in chondrogenic (1:200 TGFβ1, 1:500 Vitamin C, 1:10,000 Dexamethasone in ITS & Proline supplemented basal medium,, ThermoFisher & Sigma Aldrich), osteogenic (1:10 BGP, 1:100 Vitamin C, 1:10,000 Dexamethasone in basal medium, ThermoFisher & Sigma Aldrich), and adipogenic medium (StemPro Adipogenesis Differentiation Kit). Chondrogenic differentiation was performed with a suspension of approximately 2 × 106 cells a culture in a test-tube that subsequently formed a 3D spheroid pellet in chondrogenic media (Additional file
Chondrocytes were extracted from the articular cartilage on the medial and lateral femoral condyles, tibial plateau, and patellar facets by mechanical and chemical digestion by 20 mL of sterile 0.2% Collagenase A solution for 20 h at 37 °C with constant agitation. The digest was eluted through a 70 μm filter to remove large non-cellular aggregate, centrifuged at 400
The study involved co-culturing donor-matched p0 ASCs with p0 chondrocytes in monolayer directly in T25 flasks, each with 4 mL of basal media comprising of low-glucose Dulbecco’s Modified Eagle Medium (DMEM, ThermoFisher) supplemented by 10% Fetal Bovine Serum (FBS, ThermoFisher) and 1% penicillin/streptomycin (ThermoFisher) and the digest centrifuged at 300
The Quant-iT Picogreen dsDNA assay (ThermoFisher) was performed to assess cell count and proliferation as per standard protocol [
Total DNA as a measure of cell number was quantified at harvest for each seeding ratio using the Picogreen dsDNA spectrophotometric assay by obtaining a calibration curve, which carries the advantage of low signal interference from proteins and single-strand nuclei acid. Culturing on 25 cm2 monolayers over a fixed time period of 18–21 days.
Following coculture, cells were harvested at p0 at 90% confluence after 18–21 days of culture with QIAzol Lysis Reagent (Qiagen). RNA was extracted using the Direct-zol RNA MicroPrep Kit (Zymo Research) [
Quantitative real-time PCR (RT-qPCR) was performed by the addition of SYBR Green premix (ThermoFisher), corresponding 1 μM forward and 1 μM reverse primers, and 10 ng of cDNA per well to a final volume of 5 μl per well on a 96-well plate. The reaction was initiated with a 5-min pre-incubation at 95 °C, and 40 cycles of 10 s at 95 °C, 30 s at 60 °C for denaturation, annealing, and extension. A melt curve analysis was performed for 15 s at 95 °C, 60 s at 60 °C, and 15 s at 95 °C. Copy numbers per gene of interest were determined by cycle threshold (△Cτ) normalized against the housekeeping gene HPRT1. Individual samples or entire batches with ASC or Chondrocyte monocultures showing Cτ [HPRT1] > 30 were rejected. Primers for the chondrogenic RT-qPCR panel are shown in Table Primers designed with existing studies and PrimerBLAST, manufactured by Sigma Aldrich Gene Forward primer Reverse primer HPRT1 5′-TGACACTGGCAAAACAATGCA-3′ 5′-GGTCCTTTTCACCAGCAAGCT-3′ COL1A1 5′-ATGCCTGGTGAACGTGGT-3′ 5′-AGGAGAGCCATCAGCACCT-3′ COL2A1 5′-AACCAGATTGAGAGCATCCGC-3′ 5′-CGATAACAGTCTTGCCCCACTTAC-3′ COL10A1 5′-CACCTTCTGCACTGCTCATC-3′ 5′-GGCAGCATATTCTCAGATGGA-3′ L-SOX5 5′-GAATGTGATGGGACTGCTTATGTAGA-3′ 5′-GCATTTATTTGTACAGGCCCTACAA-3′ SOX6 5′-CACCAGATATCGACAGAGTGGTCTT-3′ 5′-CAGGGTTAAAGGCAAAGGGATAA-3′ SOX9 5′-GCAGGCGGAGGCAGAGGAG-3′ 5′-GGAGGAGGAGTGTGGCGAGTC-3′ ACAN 5′-AGGGCGAGTGGAATGATGTT-3′ 5′-GGTGGCTGTGCCCTTTTTAC-3′ HSPG2 5′-TCAGTCCCTTGTCACCATCCA-3′ 5′-TAAGCTGCCTCCACGCTTAT-3′ COMP 5′-AACAGTGCCCAGGAGGAC-3′ 5′-TTGTCTACCACCTTGTCTGC-3′
Gene expression was determined by the 2−ΔΔCτ method [
All experiments were repeated independently in triplicate. Statistical analysis was performed with GraphPad Prism 8. Depending on the homoscedasticity and normality of data determined by independent F-tests and Shapiro–Wilk tests respectively, Student’s t-test (homoscedastic), Welch’s t-test (heteroscedastic), or Mann–Whitney U test (skewed non-parametric distribution) was adopted in pairwise comparisons. One-way Fisher’s ANOVA with post hoc Tukey’s HSD (homoscedastic) or One-way Welch’s ANOVA with post hoc Dunnett’s test (heteroscedastic) was adopted for multiple comparisons.
As per the ISCT, the ASCs were characterized and demonstrated: 1. Plastic adherence. 2. Expression of CD73, CD90 & CD105, and absent expression of CD14, CD19, CD45 & HLA-DR. The CD34 ex-pression was heterogeneous, with 14.07% of cells showing positive expression Additional file 3. Chondrogenic, osteogenic, and adipogenic differentiation in vitro (Fig. Epitope characterization of IFPD ASCs using the ISCT panel confirms expression of key markers. Top row markers are expressed: CD73, CD90, CD105, and a sub-population of CD34 (Stained (Red) readouts greater than negative (Blue)). Bottom row markers lack expression: CD14, CD19, CD45, and HLA-DR (stained HLA-DR population overlaps onto negative population) Histology demonstrating the trilineage potential of IFPD ASCs.
Cell proliferation was documented from seeding to harvest, and Fig. Photomicrograph of proliferating co-culture intervals and monoculture controls from seeding to harvest showing no difference in cell morphology at confluence. Black bar indicates a distance of 100 μm
The mean volume of IFP per patient knee was 7.9 mL (± 8.0 mL SD, 5.0–37.5 mL) and mean mass of cartilage was 5.12 g (± 1.99 g SD, 2.73–9.20 g). Following culture on 25cm2 monolayers over 14–21 days, measurements of total DNA using the Picogreen dsDNA assay demonstrated that the concentration of dsDNA was directly proportional to the total quantity of nuclear DNA, and hence the number of cells per flask at harvest (Fig. Quantity of DNA (mean ± SD) across different seeding ratios were not significantly different; positive chondrocyte control (+ ve Chondro), negative ASC control (-ve ASC), used in pilot and experimental trials (n = 9)
We found no statistically significant difference in the quantity of DNA across the different co-culture ratios and monoculture controls. Therefore, it suggests that the level of expression of the genes of interest were dependent on factors other than proliferation rate and cell number.
A chondrogenic panel of three collagen genes (COL1A1, COL2A1 and COL10A1), three SOX genes (L-SOX5, SOX6, SOX9), two proteoglycans, and one extracellular matrix (ECM) protein was designed to study the chondrogenic gene expression of ASC, chondrocyte monocultures and co-cultures at week 2–3. All genes except COL2A1 for ASC monoculture and COL10A1 for any interval are expressed, albeit at various levels of expression for the different genes and cell ratios (Fig. -ΔCτ of different genes of interest showing relative gene expression normalized to HPRT1 (n = 5); positive chondrocyte control (+ ve Chondro), negative ASC control (− ve ASC) Expected vs Observed levels of expression of chondrogenic genes of interest were different across various co-culture ratios normalized against the positive chondrocyte control (n = 5; Mean ± SD). *
Although the potential of ASCs for chondral and osteochondral repair is well established in vitro and in animal studies, similar efficacy has not been observed in clinical trials. One reason for this may be due to difficulties in culturing ASCs ex vivo to undergo chondrogenic differentiation prior to implantation into chondral lesions. Therefore, new ways of improving the chondrogenic differentiation of ASCs are needed. The use of IFPD ASCs isolated from fat pad for coculture with chondrocytes, with the aim of improving chondrogenic differentiation is novel and has not previously been investigated. Human IFPD ASCs have been shown to engraft with acellular dermal matrix and express chondrogenic genes in vitro [
In our study, while the IFPD ASC and chondrocyte co-cultures exhibited more variable results than those for monocultures, there was nonetheless evidence demonstrating in-creased chondrogenicity of co-culture than would be expected by ratios derived from monoculture. This is particularly evidenced by the significant increase in expression of COL2A1, L-SOX5, SOX6, SOX9, ACAN, and COMP in a 1:1 IFPD ASC-to-chondrocyte co-culture ratio. The upregulation of these genes is more pronounced in a co-culture ratio of 1:1 IFPD ASC-to-chondrocyte than 10:1 or 100:1. Although it is also seen to a lesser extent in the 10:1 co-culture ratio for SOX9 & ACAN, and in the 100:1 ratio for L-SOX5 & SOX6. This has implications in providing an optimal co-culture ratio near 1:1 when seeding IFPD ASCs to be pre-conditioned by chondrocytes. It may be beneficial for autologous ASCs to be pre-conditioned with high relative ratios of chondrocyte via co-culture, prior to use for the purposes of tissue-engineering and cartilage regeneration.
COMP encodes thrombospondin-5 [
Therefore, by screening gene expression of collagen, HMG-box transcription factors, and proteoglycan/ECM pertaining to chondrogenic cell fate, this study has provided novel insight into the IFPD ASC and chondrocyte crosstalk and the potential use of pre-conditioned ASCs by autologous co-culture in cellular therapies.
Cell-based therapies for cartilage repair centre on the crosstalk between ASCs and en-dogenous chondrocytes. Our data suggests that this may be exploited in vitro to alter the properties of these therapies. Such crosstalk may involve bidirectional signaling influenced by the secretome of both cell types, including cytokines e.g. TGF-β superfamily [
In view of what we know about the gene expression of mature chondrocytes and ASCs, and the relative respective cell ratios used in our experiments, the more likely explanation for the results seen is the effect of chondrocytes on IFPD ASC gene expression. It is however possible that part or all of the results seen may be due to the effect of IFPD ASC on chondrocyte gene expression, therefore, the nature of this interaction remains un-known at present. Furthermore, as we observed no significant difference in chondrogenic gene expression between monocultures, possible de-differentiation of chondrocytes cannot be excluded. Future studies using conditioned media would help better under-stand this paracrine effect. Further assessment of the cellular contents and conditioned media with protein assays may help provide insight into active components of the secretome as well as chondrogenic differentiation. Through cell-sorting techniques, including Fluorescence Activated Cell Sorting (FACS), ASCs and chondrocytes may also be separated following co-culture to allow features such as gene expression and proliferation of each population to be assessed individually. Our experiments nevertheless demonstrate the potential benefit of co-culture on in vitro chondrogenesis.
Different genes are expressed at different levels, and this is a limiting factor for this study. Small increases in gene expression may be overshadowed by the variance of results. This is limited by the power of the study since observed power is inversely proportional to the p-value due to the small sample size of n = 5. In particular, significant variance was ob-served in the results of the proliferation assay, suggesting that greater sample sizes are required to reliably discern any true difference.
In vivo cell phenotype of chondrocytes and in situ IFPD ASCs may be affected by the chronic stress and inflammatory effects of osteoarthritis, leading to reduced proliferative capacity. Indeed, the fat pad has been implicated as a disease-relevant tissue in the pathogenesis of osteoarthritis, possibly contributing to global joint inflammation and sensitization of the joint. In spite of this, human IFPD ASCs isolated from osteoarthritic joints have been shown to exert chondrogenic capacity similar to that of non-osteoarthritic joints [
In vitro conditions may not necessarily be consistent with the in vivo microenvironment as articular cartilage is avascular and hypoxic [
Our study shows that pre-conditioning IFPD ASCs in co-culture with chondrocytes can enhance chondrogenic gene expression in vitro, and this effect is greater when seeding at a lower IFPD ASC-to-chondrocyte ratio. The observed differences in gene expression did not appear to be due to differences in cell number and proliferation rate, therefore we speculate that paracrine effects and juxtacrine signaling pathways between ASCs and chondrocytes are likely to be involved. Our findings suggest a need to investigate alternative culture conditions for ASCs in clinical trials of intraarticular therapy, potentially through co-culture with chondrocytes or incubation in chondrocyte-conditioned media prior to injection of the ASCs into the target site. There is thus a need to further demonstrate the efficacy of pre-conditioned ASCs by co-culture in pre-clinical, animal, and clinical trials.
Not applicable.
CM: conception and design, collection and assembly of data, data analysis and interpretation, manuscript writing. KT: collection of data, interpretation of data. KF: collection of data, administrative support. RB: administrative support, provision of study material. WK: conception and design, provision of tissue samples. All authors drafted the article and revised it critically. All authors read and approved the final manuscript.
Experiments were supported by a Pump-Priming Grant from the Royal College of Surgeons of England and institutional support was provided from Versus Arthritis (Formerly Arthritis Research UK) through Versus Arthritis Tissue Engineering & Re-generative Therapies Centre (Grant 21156).
The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request.
The study was conducted according to the guidelines of the Declaration of Helsinki, and appropriately approved by the North West—Preston Research Ethics Committee, UK (REC: 18/NW/0545, IRAS: 247368). Informed consent was obtained from all subjects involved in the study.
Not applicable.
The authors declare that they have no competing interests.
Cycle threshold
Dulbecco’s modified eagle medium
Extracellular matrix
Fluorescence activated cell sorting
Fetal bovine serum
Infrapatellar fat pad-derived
Mesenchymal stromal cells
Passage 0
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.