Multipotential stromal cells in the talus and distal tibia in ankle osteoarthritis - Presence, potency and relationships to subchondral bone changes

Abstract

A large proportion of ankle osteoarthritis (OA) has an early onset and is post-traumatic. Surgical interventions have low patient satisfaction and relatively poor clinical outcome, whereas joint-preserving treatments, which rely on endogenous multipotential stromal cells (MSCs), result in suboptimal repair. This study investigates MSC presence and potency in OA-affected talocrural osteochondral tissue. Bone volume fraction (BV/TV) changes for the loading region trabecular volume and subchondral bone plate (SBP) thickness in OA compared with healthy tissue were investigated using microcomputed tomography. CD271-positive MSC topography was related to bone and cartilage damage in OA tissue, and in vitro MSC potency was compared with control healthy iliac crest (IC) MSCs. A 1.3- to 2.5-fold SBP thickening was found in both OA talus and tibia, whereas BV/TV changes were depth-dependent. MSCs were abundant in OA talus and tibia, with similar colony characteristics. Tibial and talar MSCs were tripotential, but talar MSCs had 10-fold lower adipogenesis and twofold higher chondrogenesis than IC MSCs (P = .01 for both). Cartilage damage in both OA tibia and talus correlated with SBP thickening and CD271+ MSCs was 1.4- to twofold more concentrated near the SBP. This work shows multipotential MSCs are present in OA talocrural subchondral bone, with their topography suggesting ongoing involvement in SBP thickening. Potentially, biomechanical stimulation could augment the chondrogenic differentiation of MSCs for joint-preserving treatments.

Keywords: ankle; multipotential stromal cells; osteoarthritis; regenerative medicine.

Figure 1

Morphological changes of OA visible by microcomputed tomography on proximal talus and distal tibia (n = 3 OA patients compared with n = 3 non‐diseased controls, NDC). A, Talocrural tissues rendered in 3D. Healthy control images approximate analysed regions. A—anterior, P—posterior, M—medial, L—lateral. The OA talus and tibia are far smaller due to the minimal amount of bone removed from surgery. B, Change in subchondral bone plate (SBP) thickness in OA (symbols represent an average 20 measurements per donor). C, Change in bone volume of total volume (BV/TV) in the analysed regions in OA talus/tibia (symbols represent individual donors). D, Change in BV/TV by depth in non‐diseased control talus/tibia. E, Change in BV/TV by depth from the SBP in talus. OA talus samples are measured to 4 mm depth due to thinner samples being removed during surgery. F, Change in BV/TV by depth from SBP in tibia. Symbols and error bars on B/C show medians ± interquartile ranges (IQR) for 3 donors with a minimum of 200 slices. D‐F show medians ± interquartile ranges (IQR) for 3 donors with 100 slices per depth interval ***P < .001, using Wilcoxon signed‐rank test

Figure 2

Properties of MSCs isolated from OA talus and tibia (n = 3 donors) and control Iliac crest (n = 3 donors) isolated by collagenase digestion of bone. A, Percentage of cells isolated which are MSCs, measured by CFU‐F assay. Graphs show median ± interquartile range for 3 repeated experiments from 3 donors. B, Integrated density of all MSC‐initiated colonies from CFU‐F assay measured using ImageJ. C, Area of all MSC‐initiated colonies from CFU‐F assay measured by ImageJ. Symbols on B and C represent individual colonies; lines and error bars represent medians and IQR. ***P < .001, Kruskal‐Wallis test with Dunn's correction for multiple comparisons

Figure 3

Characterization of culture‐expanded MSCs from OA distal tibia and talus (n = 3 donors) and control Iliac crest bone (n = 3 donors). A, Surface phenotype of isolated cells from talus, tibia and IC after culture to passage 2 by flow cytometry. Negative markers reflect absence of haemopoietic lineage cells. Bar graphs represent means and standard deviations (SDs). B, Glycosaminoglycan concentration of cartilage pellets following chondrogenic C/D adipogenic assays: C, Oil red and Nile red staining of adipocyte differentiated MSCs. D, Oil red‐positive area of total area. E/F Osteogenic assays: E, Alkaline phosphatase and Alizarin Red staining to prove osteogenic differentiation of MSCs. F, Calcium deposition of differentiated cells normalized to DNA content. Box plots represent medians and IQR for 3 repeated experiments from 3 donors. *P < .05, Kruskal‐Wallis test with Dunn's correction for multiple comparisons

Figure 4

Ankle tissues morphology and cell topography in end‐stage osteoarthritis A, Safranin O histology of talus or distal tibial osteochondral tissue sections obtained from ankle fusion patients. Red staining shows glycosaminoglycans, and blue shows bone. (i—Examples of grade 3 distal tibia/talus. ii—Example of grade 5 distal tibia/talus. iii—Glycosaminoglycan loss in cartilage. iv—Fibrillation of cartilage. v—Fibrillation and tidemark duplication. vi—Denuded cartilage. vii—Chondrocyte clustering. viii—A subchondral bone cyst. ix—example of grade 6 tissue with osseous repair extending above the previous surface. x—and hypertrophy/clustering of chondrocytes (i scale bar 500 µm, ii‐ix scale bar of 200 µm). B, Comparisons in OARSI cartilage damage score between the tibia and talus. Symbols represent individual patients. C, Change in subchondral bone plate thickness (SBP) with cartilage damage grade. D, Bone area of total area beneath the subchondral bone plate (SBP) between the talus and tibia, compared with the OARSI grade of the overlying cartilage. E, Bone area of total area (BA/TA) comparing different bone regions with the distance from the SBP. F, Comparison of changes in BA/TA of talar bone in grade 3 vs grade 5 damaged regions near the subchondral bone plate vs further away. G, Comparison of changes in BA/TA of tibial bone in grade 3 vs grade 5 damaged regions near the subchondral bone plate vs further away. C‐G, Box plots represent medians and IQR from 3 donors measured from a minimum of 4 individual tiles. *P < .05, **P < .01, ***P < .001, Kruskal‐Wallis test with Dunn's correction for multiple comparisons

Figure 5

Relationships between CD271‐positive cells and bone behaviour (A) High magnification images of immunohistochemistry for CD271, a native marker of MSCs, showing staining: (i) Overview of grade 3 and grade 5 regions from talus and tibia. (ii/iii) High staining in SB areas beneath the tidemark. (iv) Areas of chondrocyte clustering. (v) bone marrow stroma around blood vessels. B, Comparison of total CD271 coverage between talus and tibia. C, Relative CD271 coverage comparing regions near the subchondral bone plate to that further away. Box plots represent medians and IQR from 3 donors measured from a minimum of 4 individual tiles. D, The effect of overlying cartilage damage grade with CD271 coverage of trabecular space underneath. Symbols on B and D show individual measurements; lines and error bars represent medians and IQR. *P .05, **.01, Kruskal‐Wallis test with Dunn's correction for multiple comparisons