Mechano-Hypoxia Conditioning of Engineered Human Meniscus

Abstract

Meniscus fibrochondrocytes (MFCs) experience simultaneous hypoxia and mechanical loading in the knee joint. Experimental conditions based on these aspects of the native MFC environment may have promising applications in human meniscus tissue engineering. We hypothesized that in vitro "mechano-hypoxia conditioning" with mechanical loading such as dynamic compression (DC) and cyclic hydrostatic pressure (CHP) would enhance development of human meniscus fibrocartilage extracellular matrix in vitro. MFCs from inner human meniscus surgical discards were pre-cultured on porous type I collagen scaffolds with TGF-β3 supplementation to form baseline tissues with newly formed matrix that were used in a series of experiments. First, baseline tissues were treated with DC or CHP under hypoxia (HYP, 3% O₂) for 5 days. DC was the more effective load regime in inducing gene expression changes, and combined HYP/DC enhanced gene expression of fibrocartilage precursors. The individual treatments of DC and HYP regulated thousands of genes, such as chondrogenic markers SOX5/6, in an overwhelmingly additive rather than synergistic manner. Similar baseline tissues were then treated with a short course of DC (5 vs 60 min, 10-20% vs 30-40% strain) with different pre-culture duration (3 vs 6 weeks). The longer course of loading (60 min) had diminishing returns in regulating mechano-sensitive and inflammatory genes such as c-FOS and PTGS2, suggesting that as few as 5 min of DC was adequate. There was a dose-effect in gene regulation by higher DC strains, whereas outcomes were inconsistent for different MFC donors in pre-culture durations. A final set of baseline tissues was then cultured for 3 weeks with mechano-hypoxia conditioning to assess mechanical and protein-level outcomes. There were 1.8-5.1-fold gains in the dynamic modulus relative to baseline in HYP/DC, but matrix outcomes were equal or inferior to static controls. Long-term mechano-hypoxia conditioning was effective in suppressing hypertrophic markers (e.g., COL10A1 10-fold suppression vs static/normoxia). Taken together, these results indicate that appropriately applied mechano-hypoxia conditioning can support meniscus fibrocartilage development in vitro and may be useful as a strategy for developing non-hypertrophic articular cartilage using mesenchymal stem cells.

Keywords: cyclic hydrostatic pressure; dynamic compression; extracellular matrix; human meniscus; hypoxia; mechanical loading; tissue engineering.

FIGURE 1

Experiment outlines. Created with Biorender.com (2021).

FIGURE 2

Experiment I: (A) Representative baseline tissue matrix morphology (3 weeks preculture), and (B) gene expression in response to 5-days of mechano-hypoxia treatment. DC and CHP each had their own VCtrl group since the loading occurred in different chambers. Fold changes are with respect to the appropriate N/VCtrl group within each donor, and values for each gene of interest are normalized to the mean expression of housekeeping genes β-actin, B2M, and YWHAZ. Expression was measured 30 min after the final loading event. Two-way ANOVA was performed twice, once for DC and once for CHP groups. “DC” and “Oxygen” indicate that these factors had a significant main effect. *: p < 0.05, **: p < 0.01. CHP, Cyclic hydrostatic pressure; DC, dynamic compression; H, hypoxia/HYP; N, normoxia/NRX; VCtrl, static vehicle control.

FIGURE 3

Experiment I: Principal component analysis and heatmaps of the RNA-seq dataset from the 5-days mechano-hypoxia treatment. Expression was measured 30 min after the final loading event. DC, dynamic compression; HYP, hypoxia; NRX, normoxia; PC, principal component; VCtrl, static vehicle control.

FIGURE 4

Experiment I: (A) Interaction analysis after the 5-days mechano-hypoxia treatment. The dashed red line indicates the expected bin height for a uniform distribution. (B) Genes with the largest absolute fold changes for each group (q < 0.05, total count >250). Expression was measured 30 min after the final loading event. CI: confidence interval. Neg, negative.

FIGURE 5

Experiment II: (A) The gene expression effects of DC loading incident duration. Fold changes are with respect to the static vehicle control group within each donor, and values for each gene of interest are normalized to the mean expression of housekeeping genes β-actin, B2M, and YWHAZ. Expression was measured 30 min after the loading event. “DC” indicates that a group is significantly different than the static vehicle control. “t” indicates that a DC group is significantly different than 5c. “a” indicates that a 60-min group is significantly different than 60c. ***: p < 0.001, **: p < 0.01, *: p < 0.05, +: p < 0.10. (B) Loading analysis for the 60c group. The arrowheads show the direction of onloading and offloading. The re-contact strains were defined as the points within the compression phase when the force rose above 0 N, indicating contact. The effective strain amplitudes are calculated as 40% minus the strain at the re-contact points.

FIGURE 6

Experiment II: The matrix and gene expression effects of pre-culture duration and strain. (A) Representative strain and stress vs time curves. (B) Mechanical analysis. (C) Representative safranin-O staining after loading in each group. (D) Gene expression post-DC loading by qRT-PCR. Fold changes are respect to the VCtrl group within each time point within each donor, and values for each gene of interest are normalized to the mean expression of housekeeping genes β-actin, B2M, and YWHAZ. Expression was measured 30 min after the loading event. Analysis of variance was restricted to 3-weeks groups because only two donors were available at 6 weeks. *: p < 0.05, **: p < 0.01, ***: p < 0.001.

FIGURE 7

Experiment III: Mechanical and protein outcomes from 3 weeks of mechano-hypoxia conditioning after a 3-weeks baseline pre-culture period (up to a 6-weeks total culture duration). (A) Mechanical performance over time for each donor. The strain offset was increased from 30 to 35% on loading day 8, resulting in a jump in properties. For clarity, properties for only 1 of four loading events is presented per day. (B) Representative tissue matrix phenotypes. (C) Biochemistry. GAG, glycosaminoglycans; WW, wet weight. *: p < 0.05, **: p < 0.01, ***: p < 0.001.

FIGURE 8

Experiment III: Gene expression by qRT-PCR after the 3-weeks mechano-hypoxia treatment. Expression was measured 6 h after the final loading event. Fold changes are with respect to the NRX/static group within each donor, and values for each gene of interest are normalized to the mean expression of housekeeping genes β-actin, B2M, and YWHAZ. *: p < 0.05, **: p < 0.01, ***: p < 0.001.