3.1 Substrate stiffness affects the RVD response of chondrocytes
Our results showed that the RVD response of chondrocytes was intimately related to substrate stiffness. The soft substrate markedly increased the percentage of RVD response in chondrocytes (p<0.001) (Fig.1A). Substrate stiffness caused distinctive swelling and recovering behaviors of chondrocytes (Fig. 1B). The change of cell diameter d depended on the substrate stiffness during the RVD response (Fig.1C). The trend of non-dimensional diameter d/d0 (d0 is the initial cell diameter) over time remained the same with Fig.1C, and it was independent of the cell’s initial diameter (Fig.1D). We found that stiff substrate significantly prolonged the swelling time TS of chondrocytes (p<0.001) (Supplementary Table I and Fig.1E), and soft substrate significantly increased the percentage increase of diameter in chondrocytes during the cell swelling (p<0.001) (Fig.1F). In contrast, during the cell recovery, the recovery time TR in chondrocytes on the stiff substrate was significantly shorter than that on the soft substrate (p<0.05) (Fig.1G). Furthermore, stiff substrate can obviously enhance the capacity of cell diameter recovery during cell recovering (p<0.01) (Fig.1H). Interestingly, the trend of the swelling and recovering time was completely opposite to that of the percentage increase of diameter in chondrocytes during cell RVD responding. Lastly, the soft substrate significantly increased the mean RVD responding time TRes of chondrocytes (p<0.01) (Fig.1I).
The cell diameter rate was obtained by fitting the cell diameter vs. time curves from substrates of varying stiffness. The cell diameter vs. time curves from all substrates during the cell swelling exhibited an approximately linear relationships (Fig.2A-C). Moreover, the soft substrate markedly increased the cell diameter rate during the cell swelling caused by hypo-osmotic (180 mOsm) medium (p<0.001) (Fig.2D). Then, we fitted the approximately linear part of cell diameter vs. time curves during the cell recovering (Fig. 2E-G). However, stiff substrate significantly increased the recovering cell diameter rate (p<0.001) (Fig.2H). For the first time, our results indicated that hypo-osmotic (180 mOsm) challenge-induced chondrocyte deformation is substrate stiffness-dependent.
3.2 Stiff substrate enhances the mechanical properties of chondrocytes during the cell swelling
AFM experiments were used to measure the mechanical properties of chondrocytes on substrates of varying stiffness in the iso-osmotic (320 mOsm) and hypo-osmotic (180 mOsm) medium (Fig. 3A). When the constant indentation displacement 2 μm was held on chondrocytes, chondrocytes first exhibited typical elastic response consistent with the Hertz model and then exhibited stress relaxation consistent with the viscoelastic theoretical model (Fig.3B). Theoretical model fitting of all experimental data exhibited excellent consistency for elastic and viscoelastic equations respectively (Fig. 3C). Our results indicated that stiff substrate enhances the elastic (Eelastic) and viscoelastic parameters (ER, E0 and m) of chondrocytes during the cell swelling in hypo-osmotic (180 mOsm) medium (p<0.001) (Table I and Fig. 3D-G).
Then we compared mechanical parameters of chondrocytes on variable stiffness substrates between the iso-osmotic (Supplementary Table II) and hypo-osmotic (180 mOsm) medium. Our results showed compared to iso-osmotic medium, chondrocytes on the stiff substrate during the cell swelling in the hypo-osmotic (180 mOsm) showed higher elastic and viscoelastic parameters including Eelastic, ER, E0 and m (p<0.001) (Fig. 3H-K). Moreover, the mean percentage increases in elastic modulus Eelastic of chondrocytes were higher on the soft substrate than that on the stiff substrate during the cell swelling. However, in response to the hypo-osmotic medium (180 mOsm), the mean percentage increases in viscoelastic parameters (ER, E0 and μ) of chondrocytes were higher on the stiff substrate than those on the soft substrate (Supplementary Table III). Our results suggested that substrate stiffness determines regulates the dynamic elastic and viscoelastic mechanical properties of chondrocytes during the hypo-osmotic (180 mOsm) challenge.
3.3 Stiff substrate enhances cytosolic Ca2+ oscillation of chondrocytes in iso-osmotic medium
Chondrocytes on variable stiffness substrates showed cytosolic Ca2+ oscillations in the iso-osmotic medium (320 mOsm) (Fig. 4A). Stiff substrate showed the greatest percentage of Ca2+ oscillations of chondrocytes (p<0.001) (Fig. 4B). We recorded the amplitude and frequency of Ca2+ oscillation in chondrocyte on variable stiffness substrates (Fig. 4C, D). Our results showed that stiff substrate markedly enhanced both the amplitude and frequency of Ca2+ oscillations in chondrocytes (p<0.001) (Fig. 4E, I). To test if TRPV4 channels are involved in substrate stiffness mediating Ca2+ signaling in chondrocytes, we applied TRPV4 channel activator 4aPDD and inhibitor GSK205 when recording Ca2+ oscillations, respectively. TRPV4 activator 4aPDD significantly enhanced the calcium responsive rate, amplitude and frequency of Ca2+ oscillations in chondrocytes on all substrates of varying stiffness while TRPV4 inhibitor GSK205 weakened those Ca2+ effects. (p<0.01) (Fig. 4F-H, J-L, Supplementary Table IV). In addition, when treated with TRPV4 activator 4aPDD, chondrocytes on the stiff substrate showed the highest amplitude and frequency of Ca2+ oscillations (p<0.001) (Fig. 4E, I). They also showed a bigger percentage increase of amplitude and frequency of Ca2+ oscillations than chondrocytes on the soft substrate (Supplementary Table V). On the other hand, when treated with the TRPV4 inhibitor GSK205, chondrocytes on varying substrates showed no significant difference in the amplitude of Ca2+ oscillations (p>0.25) (Fig. 4G), though chondrocytes on the stiff substrate exhibited a higher frequency of Ca2+ oscillations (p<0.05) (Fig. 4J). Chondrocytes showed a bigger percentage decrease of amplitude and frequency of Ca2+ oscillations on the stiff substrate than the soft substrate with the application of GSK205 (Supplementary Table V). Taken together, these results suggested that mechanosensitive TRPV4 channel is involved in chondrocytes sensing substrate stiffness and mediating the calcium signaling.
3.4 Substrate stiffness affects the hypo-osmotic (180 mOsm) challenge-induced Ca2+ oscillations in chondrocytes
Only chondrocytes that exhibited both RVD response and Ca2+ oscillations were used in this analysis. Our results show that soft and medium substrates significantly increased the percentage of chondrocytes exhibited both RVD response and Ca2+ oscillations (42 ± 5% and 39 ± 4%, respectively) than the stiff substrate (31 ± 5%) (p<0.01). We then quantified the amplitude and frequency of Ca2+ oscillation in chondrocytes cultured on variable stiffness substrates during the cell swelling and recovering (Fig. 5A, B). The results showed that soft substrate induced a significantly higher percentage of Ca2+ oscillations (58 ± 7%) in chondrocytes than medium (77 ± 8%) and stiff substrates (79 ± 8%) during the hypo-osmotic (180 mOsm) challenge (p<0.001). Application of TRPV4 activator 4aPDD or the TRPV4 inhibitor GSK205 significantly enhanced or weakened the calcium responsive rate of Ca2+ oscillations in chondrocytes on variable stiffness substrates, respectively (p<0.01) (Supplementary Table VI). During the cell swelling, the soft substrate enhanced both the amplitude and frequency of Ca2+ oscillations (p<0.01) (Fig. 5C, F). With the treatment of TRPV4 activator 4aPDD, both the amplitude and frequency of Ca2+ oscillations in chondrocytes were significantly higher on soft substrate (p<0.005) (Fig. 5D, G). With the treatment of TRPV4 inhibitor GSK205, the amplitude and frequency of Ca2+ oscillations were significantly higher on the stiff substrate (p<0.05) (Fig. 5E, H). Moreover, the relative percentage increase of Ca2+ oscillations (amplitude and frequency) in chondrocytes treated with TRPV4 activator 4αPDD was higher on the soft substrate than the stiff substrate during the cell swelling. On the other hand, the decrease of Ca2+ oscillations induced by GSK205 was also higher on the soft substrate than the stiff substrate. (Supplementary Table VII).
During the cell recovering process, the stiff substrate enhanced both the amplitude and frequency of Ca2+ oscillations in chondrocytes (p<0.001) (Fig. 5I, L). With the treatment of TRPV4 activator 4αPDD, both the amplitude and frequency of Ca2+ oscillations in chondrocytes were significantly higher on the stiff substrate (p<0.01) (Fig. 5J, M). With the treatment of TRPV4 inhibitor GSK205, the amplitude of Ca2+ oscillations in chondrocytes showed no significant difference between varying substrates (p>0.05) (Fig. 5K), but the frequency of Ca2+ oscillations in chondrocytes was significantly higher on the stiff substrate (p<0.05) (Fig. 5N). Moreover, with the treatment of TRPV4 activator 4αPDD or the TRPV4 inhibitor GSK205, stiff substrate induced a higher percentage increase or decrease in Ca2+ oscillation in chondrocytes than the soft substrate (Supplementary Table VIII), respectively. Our results, therefore, suggested that the TRPV4 channel is involved in chondrocyte RVD response, exhibiting a completely different substrate stiffness-dependent mode of Ca2+ oscillations between the cell swelling and recovering progress.