Injected human umbilical cord-derived mesenchymal stromal cells do not appear to elicit an inflammatory response in a murine model of osteoarthritis

Abstract

Objective: This study investigated the effect of hUC-MSCs on osteoarthritis (OA) progression in a xenogeneic model.

Design: Male, 10 week-old C57BL/6 mice underwent sham surgery (n = 15) or partial medial meniscectomy (PMM; n = 76). 5x10⁵ hUC-MSCs (from 3 donors: D1, D2 and D3) were phenotyped via RT-qPCR and immunoprofiling their response to inflammatory stimuli.They were injected into the mouse joints 3 and 6 weeks post-surgery, harvesting joints at 8 and 12 weeks post-surgery, respectively. A no cell 'control' group was also used (n = 29). All knee joints were assessed via micro-computed tomography (μCT) and histology and 10 plasma markers were analysed at 12 weeks.

Results: PMM resulted in cartilage loss and osteophyte formation resembling human OA at both time-points. Injection of one donor's hUC-MSCs into the joint significantly reduced the loss of joint space at 12 weeks post-operatively compared with the PMM control.This 'effective' population of MSCs up-regulated the genes, IDO and TSG6, when stimulated with inflammatory cytokines, more than those from the other two donors.No evidence of an inflammatory response to the injected cells in any animals, either histologically or with plasma biomarkers, arose.

Conclusion: Beneficial change in a PMM joint was seen with only one hUC-MSC population, perhaps indicating that cell therapy is not appropriate for severely osteoarthritic joints. However, none of the implanted cells appeared to elicit an inflammatory response at the time-points studied. The variability of UC donors suggests some populations may be more therapeutic than others and donor characterisation is essential in developing allogeneic cell therapies.

Keywords: Allogeneic cell therapy; Human MSCs; OA; Preclinical model; Xenogeneic.

Fig. 1

Experimental design: hUC-MSCs were injected into the hind left knee of C57BL/6 mice at either 3 or 6 weeks post-PMM (time points 1 and 2, respectively); mice were culled at 8 and 12 weeks post-PMM for time points 1 and 2 respectively. Sham operated mice and PMM control mice were also included, with neither of these receiving hUC-MSCs.

Fig. 2

Characterisation of hUC-MSCs prior to use in the PMM model. Immunoprofiling of hUC-MSCs assessed via flow cytometry for markers indicative of MSCs, putative chondrogenic markers and immunomodulation on un-stimulated cells (A). Markers indicative of MSCs (CD14, CD73), co-stimulation (CD40, CD80, CD86), Human Leukocyte Antigen-DR (HLA-DR) and immunomodulation (CD106, CD317) on hUC-MSCs cultured in normal media without IFN-γ (unstimulated), following stimulation with 25 ng/ml IFN-γ for 24 h and following stimulation with 25 ng/ml IFN-γ, 50 ng/ml TNF-α 10 ng/ml IL-1β for 24 h (inflammatory cocktail; B). RT-qPCR analysis showing gene expression in stimulated and un-stimulated hUC-MSCs for TSG6 (stimulated with inflammatory cocktail for 24 h; C), IDO (stimulated with IFN- γ for 24 h; D) and IDO (stimulated with inflammatory cocktail for 24 h; E). Data is presented as the mean ± SD. Results are shown from 3 patients' hUC-MSCs, (D1-3).

Fig. 3

Assessment of joint space. Representative two-dimensional coronal μ-CT images at 12 weeks in sham (A) and PMM control group (B). The medial joint space of the different groups analysed by μ-CT at 8 weeks (C) indicated that all PMM groups, other than those injected with hUC-MSCs from donor D2 had a significantly reduced joint space (μm) compared to the sham control; in contrast, at 12 weeks (D) the mice injected with hUC-MSCs from donor D1 had a significantly greater joint space than the PMM control, but it was not significantly reduced compared to the sham control. Data is presented as the mean ± SD. Significance was determined below p < 0.05.

Fig. 4

Quantification of osteophyte development. There were significantly more osteophytes in all cell groups at 8 weeks (A) but only D3 at 12 weeks (B). Total osteophyte volume was only significantly increased in animals that received hUC-MSCs from donor D2 at 8 weeks (C); there were no significant differences in any group at 12 weeks (D). Data is presented as the mean ± SD. Significance was determined below p < 0.05.

Fig. 5

Subchondral bone changes. hUC-MSCs from D1 demonstrated a significantly higher medial/lateral BV than sham mice at 8 weeks (A) but not at 12 weeks (B). When looking at subchondral bone thickness, cells from D3 demonstrated a significantly higher lateral trabecular thickness than the PMM control at 8 weeks (A) but not 12 weeks (D) with no further significant differences observed for any other treatment group or time point. There were no significant differences in any group at 8 or 12 weeks when looking at subchondral bone thickness on the medial side. Data is presented as the mean ± SD. Significance was determined below p < 0.05.

Fig. 6

Assessment of cartilage degradation. The summed joint score was significantly higher than the sham for each cell group at 8 weeks (A) but not 12 weeks (B). The maximum scores for the MFC and MTP were also significantly higher than sham and PMM control for each cell group at 8 weeks but not 12 weeks (C–F). Histological changes with extreme loss of GAG (as seen by loss of safranin O staining) and loss of articular cartilage in the mouse knee can be seen in the medial femoral condyle (MFC) and medial tibial plateau (MTP) in the control PMM knee with ‘no cells’ (G) and to a lesser extent in the PMM knee with donor D1 cells (H). No loss of GAG was seen in the sham knee (I). Data is presented as the mean ± SD. Significance was determined below p < 0.05. All sections were stained with safranin O and fast-green counterstain.

Fig. 7

Assessment of synovial reaction: No significant synovitis was detected at either 8 (A) or 12 (B) weeks post-PMM surgery. Data is presented as the mean ± SD.

Fig. 8

Inflammatory markers. Conditioned media was collected from the following hUC-MSC cell cultures: un-stimulated, stimulated with IFN-γ (25 ng/ml for 24 h) and stimulated with an inflammatory cocktail (25 ng/ml IFN-γ, 50 ng/ml TNF-α 10 ng/ml IL-1β for 24 h) and assessed for the presence of: Monocyte chemoattractant protein 1 (MCP-1; A) and IFN-γ-inducible protein 10 (IP-10; B). MCP-1 and IP-10 were significantly increased following stimulation with an inflammatory cocktail for 24 h, compared with the unstimulated UC-MSCs. Data is presented as the mean ± SD. Significance was determined below p < 0.05. ▪ = donor D1; ▲ = donor D2; ● = donor D3.