Chondrocyte and mesenchymal stem cell derived engineered cartilage exhibits differential sensitivity to pro-inflammatory cytokines

Abstract

Tissue engineering is a promising approach for the repair of articular cartilage defects, with engineered constructs emerging that match native tissue properties. However, the inflammatory environment of the damaged joint might compromise outcomes, and this may be impacted by the choice of cell source in terms of their ability to operate anabolically in an inflamed environment. Here, we compared the response of engineered cartilage derived from native chondrocytes and mesenchymal stem cells (MSCs) to challenge by TNFα and IL-1β in order to determine if either cell type possessed an inherent advantage. Compositional (extracellular matrix) and functional (mechanical) characteristics, as well as the release of catabolic mediators (matrix metalloproteinases [MMPs], nitric oxide [NO]) were assessed to determine cell- and tissue-level changes following exposure to IL-1β or TNF-α. Results demonstrated that MSC-derived constructs were more sensitive to inflammatory mediators than chondrocyte-derived constructs, exhibiting a greater loss of proteoglycans and functional properties at lower cytokine concentrations. While MSCs and chondrocytes both have the capacity to form functional engineered cartilage in vitro, this study suggests that the presence of an inflammatory environment is more likely to impair the in vivo success of MSC-derived cartilage repair. © 2018 Orthopaedic Research Society. Published by Wiley Periodicals, Inc. J Orthop Res 36:2901-2910, 2018.

Keywords: cytokines; inflammation cartilage; matrix degradation cartilage; progenitors and stem cells cartilage; tissue engineering and repair cartilage.

Figure 1.. Mechanical properties of MSC-derived engineered cartilage are more sensitive to cytokine challenge compared to chondrocyte-derived constructs.

Construct equilibrium (E_Y) and dynamic (G*) modulus show a dose-dependent effect for (A-C) IL-1β and (B-D) TNF-α. Cytokine challenge of MSC-derived cartilage shows a near complete loss of mechanical integrity at the highest cytokine concentrations (5 and 10ng/mL). (E-F) Fit of an exponential decay curve indicated an accelerated rate of decline in mechanical properties (K) for MSC- compared to chondrocyte-derived constructs. p<0.05 for * vs. control, + vs. 1ng/mL and control, S vs. MSC-derived constructs (based on fold change values) at the same cytokine concentration and time point.

Figure 2.. Matrix loss from engineered constructs with exposure to IL-1β or TNF-α.

(A-B) GAG content in MSC-derived constructs decreased to a greater extent than chondrocyte-derived constructs, and in a dose-dependent manner. Conversely, collagen content (C-D) was largely unaffected by cytokine exposure. Chondrocyte-derived constructs showed a moderate decrease in response to IL-1β or TNF-α exposure; no significant differences were found for MSCs. p<0.05 for * vs. control, + vs. 1ng/mL and control, S vs. MSC-derived constructs (based on fold change values) at the same cytokine concentration and time point.

Figure 3.. Histological assessment of engineered constructs confirmed biochemical measurements following exposure to IL-1β.

Compared to (A) chondrocyte-derived constructs, IL-1β induced a greater progressive loss of staining intensity for proteoglycans (Alcian blue) and chondroitin sulfate (CS) in (B) MSC-derived constructs, most notably at 5 and 10ng/mL. Collagen Type II staining was not markedly affected by cytokine exposure.

Figure 4.. Histological assessment of engineered constructs confirmed biochemical measurements following exposure to TNF-α.

Staining for proteoglycans (Alcian blue) and chondroitin sulfate (CS) was modestly reduced in (A) chondrocyte-derived constructs as compared to the marked decrease observed in (B) MSC-derived constructs. Collagen Type II staining was minimally affected by cytokine exposure, with changes only observed at the highest cytokine concentration.

Figure 5.. Differential release of catabolic mediators and matrix components by chondrocyte- and MSC-derived engineered cartilage in response to cytokine challenge.

(A-B) Nitric oxide production (nitrite) was higher for chondrocyte-derived constructs compared to MSCs, and on day 3 compared to day 6, for both cytokines and across all doses. (C-D) In response to IL-1β, MMP activity increased from day 3 to day 6, in a dose-dependent manner. Although absolute measurement of activated MMPs was higher for MSC-derived constructs, the relative increase compared to control was larger for chondrocytes due to a lower baseline level. In contrast, TNF-α exposure resulted in only low levels of MMP activity on day 3 for both chondrocytes and MSCs; however, by day 6 MMP activity for MSC-derived constructs increased markedly and matched that resulting from IL-1β exposure. (E-F) GAG release to the media was consistently lower for chondrocyte-derived constructs than MSCs on day 3 for both cytokines. However, by day 6, minimal differences in matrix loss were observed between both cell types. p<0.05 for * vs. control, + vs. 1ng/mL and control, ∂ vs. 5ng/mL, 1ng/mL, and control, # vs. day 6 for the same cell type, S vs. MSC-derived constructs (based on fold change values for activated MMP and released GAG measurements) at the same cytokine concentration and time point.