Tracking chondrocyte-to-fibroblast transformation via changes in cell electrophysiology

Abstract

Cultured chondrocytes have potential for cartilage regeneration to treat degenerative diseases. However, when explanted chondrocytes are cultured in monolayer, they are known to dedifferentiate over time, adopting a more fibroblastic phenotype. This greatly impacts both research and potential clinical applications; cell-based cartilage repair therapies require significant in vitro expansion to obtain sufficient chondrocyte numbers for reimplantation, as chondrocytes adopt a more fibroblastic phenotype causing up to 70% of patients to develop fibrocartilaginous fill. We used dielectrophoresis (DEP) to observe changes in the electrophysiological properties of primary bovine chondrocytes over time. Using a multi-conductivity approach, we demonstrate that monitoring the cytoplasmic conductivity is a reliable method of observing cell changes over 100 days in culture. Results show statistically significant changes in both membrane capacitance (p = 0.0039) and cytoplasm conductivity (p = < 0.0001) when tested at multiple conductivities. Analysis of cytoplasmic vs. medium conductivity allowed simple tracking of chondrocyte membrane potential, which exhibited transitions between three stable values of V_m which correspond to patch-clamp-derived literature values as they transition from chondrocytes (- 13 to - 18 mV) to proliferating fibroblasts (- 32 to - 43 mV) and ultimately to non-proliferating fibroblasts (- 55 to - 71 mV), transitions occurring around days 40 and 80 respectively.

Keywords: Biomarker; Dedifferentiation; Dielectrophoresis; Electrome; Membrane potential.

Fig. 1

Average DEP spectra of primary chondrocytes (passage 2 of the first biological repeat) in five different σ_med including (a) 8 mS/m (black circles) (b) 21 mS/m (pink squares) (c) 50 mS/m (teal triangles) (d) 150 mS/m (dark purple inverted triangles) (e) 447 mS/m (lilac diamond). (f) The corresponding MATLAB-fitted DEP models of DEP spectra in 8 mS/m (black line), 21 mS/m (pink line), 50 mS/m (teal line), 150 mS/m (dark purple line) and 447 mS/m (lilac line). r² values denote the goodness-of-fit of the DEP models to the DEP spectra P.

Fig. 2

Values of (a) G_eff (b) C_eff (c) σ_cyto (d) cell radius in chondrocytes (n = 14) suspended in σ_med of 21 mS/m (pink squares), 50 mS/m (teal triangles), and 150 mS/m (purple inverted triangles) shown in Table 2. Each datapoint is a separate passage from a biological repeat (n = 4 biological repeats G_eff, C_eff, and σ_cyto), and cell radius (n = 5 biological repeats). Values of G_eff, C_eff, and σ_cyto were obtained from MATLAB-fitted models of DEP spectra with an r² value of 0.8. Changes in values of G_eff, C_eff, and σ_cyto with σ_med were compared using non-parametric Friedman tests followed by Dunn’s multiple comparisons (*p < 0.05, **p < 0.01, ****p < 0.0001). Changes in values of cell radius with σ_med were compared using repeated-measures one-way ANOVA.

Fig. 3

Mean values of (a) G_eff (b) C_eff (c) σ_cyto (d) cell radius of primary bovine chondrocytes suspended in media of different conductivities including 21 mS/m (squares), 50 mS/m (triangles) and 150 mS/m (inverted triangles), against the number of days in cell culture. Values obtained from MATLAB-fitted models of DEP spectra which possessed an r² value of over 0.8. Simple linear regression models fitted to the graphs denote the goodness-of-fit of the trend line to the plotted data. r² values indicate the goodness-of-fit of the linear regression model by determining the variance in each parameter attributable to the number of days in cell culture.

Fig. 4

Calculated values of V_m (from four separate culture flasks) plotted against the number of days in cell culture. The data appears to be groupable into three phases, with transitions at days 40 and 80, with phase 1 having V_m >  − 17 mV; phase 2, − 25 < V_m <  − 42 mV; and phase 3, V_m <  − 54 mV.