Close this search box.

Comparison studies identify mesenchymal stromal cells with potent regenerative activity in osteoarthritis treatment – npj Regenerative Medicine

Study design

The aim of this study is to compare the therapeutic effects of Prrx1 and Dermo1 lineage MSCs from white adipose tissues and dermal adipose tissues on knee Osteoarthritis (KOA) mouse models to identify MSC subpopulations with the greatest therapeutic effect in OA treatment. We isolated Prrx1– and Dermo1-lineage MSCs by FACS sorting (Tomato+) from iWAT and dermis of Prrx1-Cre; R26tdTomato and Dermo1-Cre; R26tdTomato mice and compared their in vitro proliferation, tri-lineage differentiation potentials, the expression of cell surface markers, and their scRNA-seq signatures, to determine their stemness. The mice were divided to six groups randomly (Sham, OA + PBS, OA + Prrx1-lineage iWAT MSCs, OA + Prrx1-lineage dermal MSCs, OA + Dermo1-lineage iWAT MSCs, and OA + Dermo1-lineage iWAT MSCs, each group had six mice). We directly injected one dose of MSCs (2.5 × 105 cells in 25 μl PBS) into the articular cavity of OA model mice, which are induced by ACLT. The mice used in each experiment was indicated in the figure legends. Sample size was determined on the basis of previous experience.

Mice maintenance

The Prrx1-Cre, Dermo1-Cre, and R26tdTomato mouse lines were purchased from the Jackson Laboratory ( The Prrx1-Cre; R26tdTomato and Dermo1-Cre; R26tdTomato mouse lines are on C57BL/6 background. The normal mice used to create knee arthritis models are also on C57BL/6 background and were all purchased from The Nanjing model animal center. The mice used in this study were all male as most previous studies used male mice47,48. The mice used in the experiment were anesthetized using 40 mg kg−1 sodium pentobarbital by intraperitoneal injection, to obtain tissue sample or euthanized by carbon dioxide inhalation.

Animal experiments were carried out in accordance with recommendations in the National Research Council Guide for Care and Use of Laboratory Animals and in comply with relevant ethical regulations for animal testing and research, with the protocols approved by the Institutional Animal Care and Use Committee of Shanghai, China (SYXK(SH)2011-0112). This study follows the project that “Studies on the mechanism of MSC self-renewal differentiation and regulation of related tissue stem cells”, which was approved by Shanghai Jiao Tong University in 2015 and the approval number is A2015027. In this study, only corresponding authors and Hongshang Chu were aware of the group allocation of the experiments (during the allocation, the conduct of the experiment, the outcome assessment, and the data analysis).

Cell flow cytometry analysis and cell sorting

The dermis, iWAT, synovial tissue, and knee cartilage were taken from the mice. The dermis was digested with elastase and type IV collagenase, while inguinal white adipose tissue was digested with type II collagenase, synovial tissue was digested with type II collagenase. The knee cartilage was digested with type II collagenase. The released cells were passed through 40 μl and 70 μl cell sieves to get single cells. Then, we sorted out Tomato+ cells using Bio-Rad Flow Cytometer.

For cell surface marker expression with flow cytometry, the following antibodies are used: Sca-1-FITC (Biolegend, Cat #: 108105, Clone: D7, 1:50 dilution), CD29-FITC (Biolegend, Cat #: 102205, Clone: HMβ1-1, 1:50 dilution), CD44-APC (Biolegend, Cat #: 103011, Clone: IM7, 1:200 dilution), CD45-FITC (Biolegend, Cat #: 103107, Clone: 30-F11, 1:200 dilution), CD45-PE594 (Biolegend, Cat #: 103145, Clone: 30-F11, 1:200 dilution), CD73-APC (Biolegend, Cat #: 127209, Clone: TY/11.8, 1:200 dilution), CD146-APC (Biolegend, Cat #: 134711, Clone: ME-9F1, 1:200 dilution), CD105-AF488 (Biolegend, Cat #: 120405, Clone: MJ7/18, 1:250 dilution), CD106-FITC (Biolegend, Cat #: 105705, Clone: 429(MVCAM.A), 1:200 dilution), and CD271-FITC (Biolegend, Cat #: 345103, Clone: ME20.4, 1:100 dilution).

In vitro MSC proliferation and differentiation

The Tomato+ cells were cultured in vitro to determine the proliferation ability by quantitating the expression of Ki67 with an immunofluorescence staining kit (Abcam, ab15580) and DAPI (ThermoFisher Scientific). The cell cultures were also used to determine the differentiation ability. They were induced to differentiate into osteogenic, chondrogenic, and adipogenic cells. For osteogenic differentiation, the MSCs were seeded at 5 × 103/well in 12-well plates. The next day, the cells were switched into osteogenic medium (α-MEM medium containing 15% FBS, 10 mM β-glycerol phosphate, and 50 μg/ml ascorbic acid) for 7–10 days, with medium changed every 2 days. The cells were then fixed in 4% paraformaldehyde and stained with ALP. For chondrogenic differentiation, the cells were suspended at a concentration of 1.0 × 107 cells/ml. We transferred the cells into 1.5 ml RNA-free EP tubes for centrifugation using 2.6 g centrifugal force, gathered the cells at the bottom of EP tubes, and discarded the supernatant. After adding α-MEM containing 15% FBS, 100 IU/ml penicillin, and 100 μg/ml streptomycin, the EP tube was placed in a cell incubator in a semi-closed state for culture for 48 h, and after cell agglomeration, the medium was discarded and chondrogenic medium (α-MEM containing 15% FBS, 100 nM dexamethasone, 10 ng/ml TGFβ1, and 1 μM ascorbate-2-phosphate) was added. The cell mass was maintained for 21 days, with the medium changed every 3 days and the cell mass was fixed with paraformaldehyde and sliced, which were lastly stained with Alcian blue. For adipocyte differentiation, the MSCs were seeded at 2 × 104/well in a 12-well plate and cultured in α-MEM containing 15% FBS, 100 nM dexamethasone, and 5 μM insulin for 2 days. Then, the cells were switched into adipocyte maintenance medium (α-MEM containing 15% FBS and 5 μM insulin) for 6 days, with medium changed every3 days. The cells were then fixed and stained with Oil red O solution.

RNA isolation and quantitative PCR

Differentiation and expression were also assessed with quantitative PCR. Total RNA was extracted from the articular cartilage or cells, which had been induced to differentiate into osteogenic, chondrogenic, and adipogenic, using Trizol reagent (Invitrogen) and was reverse transcribed using PrimerScriptTM RT reagent Kit (TaKaRa, RR037A) to obtain cDNA. Quantitative PCR was performed using the Roche Light Cycler 480II Assay system (Roche). The levels of different mRNA species were calculated with the delta‐delta CT method and normalized to GAPDH. The primer sequences are used in Supplementary Table 1.

Surgically-induced osteoarthritis mouse model and MSC injection

All animals were treated according to standard guidelines approved by the Shanghai Jiao Tong University ethics committee. OA was induced by ACLT in 8-week-old C57BL/6 male mice, which gradually developed KOA pathology starting from 4 weeks, as previous reported47,48. Each mouse is treated as the experimental unit. The mice were randomly divided to six groups (Sham, OA + PBS, OA + Prrx1-lineage iWAT MSCs, OA + Prrx1-lineage dermis MSCs, OA + Dermo1-lineage iWAT MSCs, and OA + Dermo1-lineage dermis MSCs). The mice were all maintained in the same room and same access to water and food. They were derived from the same breeder to minimize potential confounders. Intra-articular injections of 1 × PBS or MSCs (2.5 × 105 cells in 25 μl PBS) were carried out at 4 or 8 weeks post-KOA surgery with German Braun Disposable Sterile Insulin Syringe (0.3*8 mm, 1 ml).

Histological analysis, immunohistochemistry, and immunofluorescence staining

The harvested knee joints at 3 weeks or 4 months post-surgery, which were fixed in 4% (vol/vol) neutral buffered formalin for 24 h and decalcified in neutral 10% (wt/vol) EDTA solution for 1 month at room temperature on the oscillator. The samples were then dehydrated, cleared, and embedded in paraffin blocks sequentially, or soaked in 20% sucrose and then 30% sucrose and finally embed with optimal cutting temperature compound (OCT) in liquid nitrogen. The paraffin-embedded tissues were cut 10 μm with a paraffin microtome and the frozen-embedded tissues were cut 8 μm with a cryostat. Paraffin sections were used for safranin-O and H/E staining while frozen sections were used for immunostaining. Mice samples (n = 6) from each group were evaluated by the Osteoarthritis Research Society international (OARSI) scoring system with a score of 0 standing for normal cartilage, 0.5 = loss of proteoglycan with an intact surface, 1 = superficial fibrillation without loss of cartilage, 2 = vertical clefts and loss of surface lamina (any % or joint surface area), 3 = vertical clefts/erosion to the calcified layer lesion for 1%–25% of the quadrant width, 4 = lesion reaches the calcified cartilage for 25%–50% of the quadrant width, 5 = lesion reaches the calcified cartilage for 50%–75% of the quadrant width, and 6 = lesion reaches the calcified cartilage for 75% of the quadrant width. The synovitis score was based on the method in published articles, all defined histopathological qualities are graded from absent (0), slight (1) and moderate (2) to strong (3), with summaries ranging from 0 to 9. 0 to 1 corresponds to no synovitis (inflammatory grade = 0), 2 to 3 to a slight synovitis (inflammatory grade 1), 4 to 6 to a moderate synovitis (inflammatory grade 2), and 7 to 9 to a strong synovitis (inflammatory grade 3)64. The cartilage sections were incubated overnight with polyclonal anti-Col2α1 antibody (abcam, Cat #: ab34712, rabbit, 1:50 dilution), anti-MMP13 antibody (abcam, Cat #: ab39012, rabbit, 1:100 dilution), anti-Col1α1 antibody (abcam, Cat #: ab21286, rabbit, 1:100 dilution), anti-Aggrecan antibody (millipore, Cat #: AB1031, rabbit, 1:50 dilution), anti-Col1α1 antibody (proteintech, Cat #: 67288-1, mouse, 1:100 dilution), and anti-CD45 antibody (abcam, Cat #: ab10558, rabbit, 1:100 dilution). The sections were then incubated with secondary antibodies conjugated with Alexa Fluor IgG(H + L) 488 (invitrogen, Cat #: A11008, goat anti-rabbit, 1:100 dilution) or Alexa Fluor IgG(H + L) 555 (invitrogen, Cat #: A11001, goat anti-mouse, 1:100 dilution). Slides were mounted with anti-fade mounting medium (OriGene, Cat #: Zli-9556, 1:1 dilution) and DAPI (ThermoFisher Scientific, Cat #: 62248, 1:1000 dilution). Images were taken under Olympus DP72 microscope (Olympus Microsystems).

scRNA-seq and analysis

Isolated Tomato+ cells from iWAT and dermis were used for single cell RNA sequencing by10X Genomics. RNA from the barcoded cells was subsequently reverse-transcribed and sequencing libraries constructed with reagents from a Chromium Single Cell 3’ v3 reagent kit (10X Genomics) following the manufacturer’s instructions. Sequencing was performed with Illumina NovaSeq 6000 (Illumina). Raw reads were demultiplexed and mapped to the mouse reference genome by Cell Ranger version 6.0.2 (10X Genomics) pipeline using default parameters. Each tissue were sequenced at a depth of about 70% saturation. The generated gene-cell expression matrices were used for subsequent analysis in R version 4.3.1 using Seurat version “Cells” fit any of the following criteria were excluded: <200 expressed genes, >20% UMIs mapped to mitochondria. Samples from iWAT or dermis were respectively integrated using “FindIntegrationAnchors” and “IntegrateData” functions. Integrated data were undergone standard cell cycle regression process provided by Seurat. Processed data were used for downstream graph-based clustering and t-SNE visualization. “FeaturePlot” function in Seurat was used for the visualization of Specific genes expression.

Bulk RNA-seq and analysis

We harvested Tomato+ articular chondrocytes formed by injected Prrx1-lineage iWAT MSCs from KOA mice and age -and sex-matched Prrx1-Cre; R26tdTomato mice. Three mice were used in either sample. RNA-seq was performed by BGI using the Illumina system. Bulk RNA-Seq data analysis was performed as follows: (1) Data filtering, the raw data obtained from sequencing was filtered using SOAPnuke (v1.5.6) to filter out 1) reads containing adapter (adapter contamination); 2) reads with an unknown base N content greater than 5%, and 3) low-quality reads (reads with a mass value of less than 15 and more than 20% of the total base number of the reads are low-quality reads) to obtain clean data. Follow-up use of Dr. Tom’s Multi-Omics Data Mining ( Department conducts data analysis, mapping and mining. (2) Differential gene analysis was performed, and the clean data was aligned to the reference genome using HISAT2 (v2.1.0) software. Use Bowtie2 (v2.3.4.3) to align the clean data to the reference gene set. Gene expression quantification was performed using RSEM (v1.3.1) software, and clustering heat maps of gene expression in different samples were plotted using pheatmap (v1.0.8). Differential gene testing was performed using DESeq2 (v1.4.5) with Q value ≤ 0.05 or FDR ≤ 0.001. (3) KEGG and GO enrichment analysis, Phyper was used to perform GO ( and KEGG ( enrichment analysis of differential genes, with Qvalue ≤ 0.05 as the threshold, and the definition of meeting this condition was significant enrichment in candidate genes.

Statistical analysis

All the measurements were collected by two authors who do not know the allocations of mice. The FlowJo software was used to analyze and plot the proportion of flow sorted cells. The GraphPad Prism 8.0.2 software was used to analyze and plot the qPCR and immunofluorescence results. All quantitative data are presented as means ± SD unless indicated otherwise. One-way ANOVA (and nonparametric) multiple comparisons and Unpaired two-tailed Student’s t test were applied to evaluate the correlation data and p < 0.05 was considered as statistically significant. *p < 0.05, **p < 0.01, ***p < 0.001, and ****p < 0.0001. We ensure that we have not set criteria and exclusions in this study. For each analysis, we have reported the exact value of n in each experimental group.