Activation of the Mechanosensitive Ion Channels Piezo1 and TRPV4 in Primary Human Healthy and Osteoarthritic Chondrocytes Exhibits Ion Channel Crosstalk and Modulates Gene Expression

Osteoarthritis (OA) is the most common degenerative joint disease causing pain and functional limitations. Physical activity as a clinically relevant, effective intervention alleviates pain and promotes joint function. In chondrocytes, perception and transmission of mechanical signals are controlled by mechanosensitive ion channels, whose dysfunction in OA chondrocytes is leading to disease progression. Signaling of mechanosensitive ion channels Piezo/TRPV4 was analyzed by Yoda1/GSK1016790A application and calcium-imaging of Fura-2-loaded chondrocytes. Expression analysis was determined by qPCR and immunofluorescence in healthy vs. OA chondrocytes. Chondrocytes were mechanically stimulated using the Flexcell™ technique. Yoda1 and GSK1016790A caused an increase in intracellular calcium [Ca2+]i for Yoda1, depending on extracellularly available Ca2+. When used concomitantly, the agonist applied first inhibited the effect of subsequent agonist application, indicating mutual interference between Piezo/TRPV4. Yoda1 increased the expression of metalloproteinases, bone-morphogenic protein, and interleukins in healthy and OA chondrocytes to a different extent. Flexcell™-induced changes in the expression of MMPs and ILs differed from changes induced by Yoda1. We conclude that Piezo1/TRPV4 communicate with each other, an interference that may be impaired in OA chondrocytes. It is important to consider that mechanical stimulation may have different effects on OA depending on its intensity.


Introduction
Osteoarthritis (OA) is the most common degenerative joint disease that causes functional limitations in addition to pain and has become a major global health problem in recent decades. Considering that the number of OA cases worldwide has more than doubled from 247.51 million in 1990 to 527.81 million in 2019, the consequences of disease-related disabilities may lead to a huge socioeconomic burden in the future [1]. Meta-analyses of studies on patients with knee OA led to the conclusion that physical exercise can function as a clinically relevant, effective intervention and non-drug treatment for reducing pain and improving movement [2,3]. Mechanical loading of joints plays a critical role in maintaining articular cartilage health and function by promoting anabolic cartilage-maintaining processes in chondrocytes. On the contrary, excessive non-physiological mechanical loading is detrimental to cartilage and may lead to an increased risk of developing OA [4].
Mechanical properties of articular cartilage depend to a large extent on the microstructure of the pericellular and extracellular matrix (PCM, ECM) [5,6]. At the molecular level, dynamic compressive loading of cartilage promotes ECM biosynthesis, resulting in improved mechanical properties. Abnormal overload can disrupt ECM homeostasis, trigger catabolic response, lead to tissue damage and promote OA [7,8]. The build-up of the ECM of cartilage during development and its maintenance in healthy tissue is carried out by the metabolic activity of chondrocytes [9,10]. The reciprocal relationship between chondrocytes and the PCM/ECM is based on the ability of chondrocytes to sense physical signals and transduce them into biochemical responses, making them highly mechanosensitive. This conversion of mechanical signals into chemical signals is called mechanotransduction and enables chondrocytes to detect changes in ECM properties [11][12][13]. Based on previous findings, the question arises whether the complex process of mechanotransduction under OA conditions might be altered in a way that cartilage-preserving chondrocytes show different sensitivity to mechanical stimuli due to the disease. Mechanical signals that are positively transduced by healthy cartilage tissue may already be deleterious under OA conditions, leading to a misregulation of matrix metalloproteinase (MMP) and proinflammatory interleukins (ILs) expression.
ECM deformation activates a series of ion channels in chondrocytes. As a consequence, changes in intracellular calcium concentration ([Ca 2+ ] i ) via channel activation are among the most fundamental molecular responses to physical stimuli. [Ca 2+ ] i can be increased via Ca 2+ influx from the extracellular environment along the gradient or the Ca 2+ release from intracellular stores [14]. Both mechanisms are involved in regulating Ca 2+ flux in chondrocytes under physical stimulation. In chondrocytes, the perception and transduction of mechanical signals leading to changes in [Ca 2+ ] i is orchestrated by mechanosensitive ion channels such as Piezo1, Piezo2, and the transient receptor potential vanilloid type 4 cation channel (TRPV4) [15]. Piezo1 and Piezo2 are mechanosensitive cation channels that allow Ca 2+ influx also into chondrocytes and act as important mechanotransducers that sense deleterious mechanical stresses (high strains) [16,17]. Regarding chemical channel activation, a small synthetic molecule termed Yoda1 functions as a channel agonist [18] and stabilizes the open Piezo1 channel. It acts as a molecular wedge that facilitates forceinduced conformational changes, effectively lowering the mechanical threshold for channel activation [19]. The Ca 2+ -permeable cation channel TRPV4 senses and transmits mechanical signals and osmotic signals in various tissues of the musculoskeletal system, including cartilage, bone, and synovium. The importance of TRPV4 is apparent in patients harboring TRPV4 mutations, which result in the development of a spectrum of skeletal dysplasias and arthropathies [20,21]. The genetic knockout of TRPV4 results in the development of OA and decreased osteoclast function [22]. Furthermore, chemical activation of TRPV4 by the agonist GSK1016790A in the absence of mechanical loading similarly enhanced anabolic and suppressed catabolic gene expression and blocked IL-1β mediated articular cartilage matrix destruction [23]. These findings support the hypothesis that TRPV4-mediated Ca 2+ signaling plays a central role in the transduction of mechanical signals to support cartilage extracellular matrix maintenance and joint health. Whereas TRPV4-mediated Ca 2+ appears to support ECM gene expression, the effects of the activation of Piezo1 on the expression of ECM remodeling factors remain to be elucidated.
Numerous studies have shown that Ca 2+ regulation in conjunction with mechanical stimulation is crucial for the behavior and function of chondrocytes and the balanced assembly of the ECM. From a therapeutic perspective, exercise therapy is an essential component of the OA treatment strategy. Therefore, it is important to characterize the abnormalities in mechanical signal transduction caused by OA to determine the extent to which mechanical stimulation of cartilage may have a positive or negative effect at the molecular level. The work presented here is intended to contribute to the elucidation of changes related to mechanical signal transduction in healthy and diseased chondrocytes.

Chemical Activation of Piezo1 by Yoda1 in Primary Human OA Chondrocytes
The effect of Yoda1 on primary human OA chondrocytes (pCH-OA) in its function as an agonist of Piezo1 was investigated in terms of channel activation and functionality in the absence of mechanical stimuli. The application of different concentrations of Yoda1 revealed a concentration-dependent increase in [Ca 2+ ] i determined in Fura-2 loaded pCH-OA cells via Ca 2+ imaging (Figure 1a- (Figure 1b). The depletion of stores by Thapsigargin was verified by the response of [Ca 2+ ] i in pCH-OA cells to histamine (Figure 1c). Concentration-dependent responses to Yoda1 were neither affected by pretreatment with Thapsigargin nor by the intermediate omission of Ca 2+ from the extracellular solution ( Figure 1e). The experiments show that the observed Yoda1-induced increase in [Ca 2+ ] i is primarily caused by Ca 2+ influx rather than due to the release of Ca 2+ from intracellular stores. The course of Ca 2+ changes by Yoda1 is comparable between pCH-OA and healthy chondrocytes (HC).
Numerous studies have shown that Ca 2+ regulation in conjunction with mechanical stimulation is crucial for the behavior and function of chondrocytes and the balanced assembly of the ECM. From a therapeutic perspective, exercise therapy is an essential component of the OA treatment strategy. Therefore, it is important to characterize the abnormalities in mechanical signal transduction caused by OA to determine the extent to which mechanical stimulation of cartilage may have a positive or negative effect at the molecular level. The work presented here is intended to contribute to the elucidation of changes related to mechanical signal transduction in healthy and diseased chondrocytes.

Chemical Activation of Piezo1 by Yoda1 in Primary Human OA Chondrocytes
The effect of Yoda1 on primary human OA chondrocytes (pCH-OA) in its function as an agonist of Piezo1 was investigated in terms of channel activation and functionality in the absence of mechanical stimuli. The application of different concentrations of Yoda1 revealed a concentration-dependent increase in [Ca 2+ ]i determined in Fura-2 loaded pCH-OA cells via Ca 2+ imaging (Figure 1a- (Figure 1b). The depletion of stores by Thapsigargin was verified by the response of [Ca 2+ ]i in pCH-OA cells to histamine (Figure 1c). Concentrationdependent responses to Yoda1 were neither affected by pretreatment with Thapsigargin nor by the intermediate omission of Ca 2+ from the extracellular solution ( Figure 1e). The experiments show that the observed Yoda1-induced increase in [Ca 2+ ]i is primarily caused by Ca 2+ influx rather than due to the release of Ca 2+ from intracellular stores. The course of Ca 2+ changes by Yoda1 is comparable between pCH-OA and healthy chondrocytes (HC). Original traces are shown exemplarily from one donor each (a-c). When Ca 2+ was omitted from the perfusion solution (NT without, w/o, Ca 2+ ), [Ca 2+ ] i was reduced (a). The effect of Yoda1 was not changed by depleting intracellular Ca 2+ stores with Thapsigargin (+Thaps), although histamine-induced Ca 2+ release was completely abolished (HIS w/o Ca 2+ ) (b). Same measurements as in (b) but before Thapsigargin application (c). Averaged responses (mean ± SEM) to different concentrations of Yoda1 before and after removal of Ca 2+ from perfusion solution are shown (d). There the different concentrations of Yoda1 correspond to the time course depicted in (a), whereas 1.0 µM Yoda1 was added before (i) and after (ii) the period of removal of extracellular Ca 2+ , demonstrating that Ca 2+ depletion did not influence response to Yoda 1. The corresponding experiments were performed with four different donor cell preparations (preparations of primary OA chondrocytes from four different patients). For each donor cell preparation, the individual treatment conditions were experimentally repeated with different frequencies; this number of experiments is indicated in brackets in the figure in each case. Different treatment options are summarized and compared to each other; no differences except the histamine response in the absence of thapsigargin functioning as a positive control was observed (e). Statistically significant differences were calculated unpaired and non-parametric by the Mann-Whitney test; *: p < 0.05; ***: p < 0.001.

Interaction and Mutual Influence of GSK1016790A and Yoda1 Effects
We were able to show that the Yoda1 effect in pCH-OA is not due to the release of intracellular calcium and that Yoda1 efficacy diminishes in the absence of extracellular calcium. In view of these results, we also attempted to treat the cells with GSK1016790A (GSK), an agonist for TRPV4, before treating the cells with Yoda1 and vice versa (Figure 2a,b). In order to clarify the influence of OA in this context, we also addressed the question of whether mechanosensitive mechanisms in pCH-OA differ from those in HC cells  Correlation of GSK effect compared to Yoda1 effect after GSK (HC, n = 11; pCH-OA, n = 13) (c). Pearson correlation (r) of reversed treatment, Yoda1 first, GSK in addition thereafter for both cell types is given in (d) (HC, n = 9, pCH-OA, n = 7). The boxplot represents the median values (min to max) of the experiments used as a basis for the correlation processes (numbers of replicates are given (n). Significant differences between successive treatments were identified via the Wilcoxon matched-pairs signed rank test (asterisk), whereas for the comparison between the groups, the Mann-Whitney test (hashtag) was used. Significances are given: *: p < 0.05, **: p < 0.01, ###: p < 0.001.
The initial treatment with the respective first agonist inhibited the effect of subsequent treatment with the respective second agonist. For Yoda1, a prior GSK treatment reduced the effect to 12.2% in HC and 24.9% in pCH-OA of the original Yoda1 response ( Figure 2e). Conversely, pretreatment with Yoda1 reduced and partially prevented the response to GSK as shown by the differences in [Ca 2+ ] i in HC (GSK: 486.371 ± 157.844; GSK after Yoda1: −101.875 ± 93.263) and pCH-OA (GSK: 857.000 ± 147.407; GSK after Yoda1: −185.667 ± 241.763), although both cell types responded properly to initial GSK application. From these results, we infer that there may be a close connection between these two types of mechanosensitive ion channels and that this "interaction" may somehow be disrupted under disease conditions.

Yoda1-Induced Increase in [Ca 2+ ] i Does Not Depend on the Intracellular Ca 2+ Level
Based on the above results, we examined whether the differential magnitude of the response to Yoda1 was dependent on the level of Correlation between values after A23187 application (ii) and subtracted Yoda1 effect (iii) is shown (d).

Expression of Mechanosensitive Cation Channels in pCH-OA and HC
Subsequently, attention also turned to TRPV4, because-assuming the Yoda1 effect might not depend on the [Ca 2+ ] i level-a functional dependence between the two channels (i.e., Piezo1 and TRPV4) seems to be possible. For this reason, we first examined the extent to which the cells expressed mechanosensitive cation channels. Immunofluorescence (IF) studies provided insight into the expression pattern of Piezo1, Piezo2, and TRPV4 in response to Yoda1, both in pCH-OA and HC (Figure 4a). No significant Yoda1-induced change in the expression of mechanosensitive ion channels in comparison to mock-treated controls could be observed. Whereas the expression of all three channels investigated seemed to be slightly increased in pCH-OA compared to HC controls, there was a highly significant 5-fold difference of TRPV4 in pCH-OA (2.735 ± 0.151fold change) compared to HC (0.545 ± 0.053-fold change) in the presence of 0.1 µM Yoda1. Representative images are shown (Figure 4b). To gain new insights into the differences between pCH-OA and HC, agonist-induced channel activation was also compared. There was no difference between the two cell types in their response to Yoda1 (Figure 4c), but the activation of TRPV4 by GSK was significantly different. The GSK effect in pCH-OA (857.000 ± 147.407) was 2.08-fold higher than in HC (411.861 ± 150.300) (Figure 4d). The fact that the GSK effect in pCH-OA was more pronounced than in HC corresponds to the different intensities seen with the IF staining. To confirm these observations at the RNA level, qRT-PCRs were performed, which led to the conclusion that Piezo2 and TRPV4 are higher expressed in pCH-OA than in HC, as shown by the decrease in ∆CT levels ( Figure 4e). From these observations, we conclude that there is a difference in the expression of mechanosensitive cation channels between HC and pCH-OA, particularly with respect to TRPV4, where we also found a functional difference between these two cell types.
PCRs were performed, which led to the conclusion that Piezo2 and TRPV4 are higher expressed in pCH-OA than in HC, as shown by the decrease in ΔCT levels ( Figure 4e). From these observations, we conclude that there is a difference in the expression of mechanosensitive cation channels between HC and pCH-OA, particularly with respect to TRPV4, where we also found a functional difference between these two cell types.  The relative gene expression of MMP1, MMP3, and MMP13 was strongly modulated by Yoda1, whereas for MMP3 and MMP13, the increase was found to be slightly concentration-dependent in HC only (Figure 5a 3.0 µM: 254.564 ± 88.707 (HC), 71.371 ± 2.819 (pCH-OA)). Notably, at 3.0 µM Yoda1, the increase in HC for BMP-2 was 3.6-fold higher than in pCH-OA. Although no effects on Col2A1 expression were detected , treatment of cells with Yoda resulted in a decrease in Col1A1 expression (Figure 5f): in the presence of 1 µM Yoda1, the decrease was significant in pCH-OA (pCH-OA: 0.510 ± 0.074), whereas under 3.0 µM Yoda1, it was highly significant in both cell types (HC: 0.141 ± 0.026; pCH-OA: 0.172 ± 0.081).  Yoda1 treatment further influenced factors involved in inflammatory processes in chondrocytes. While the gene expression level of interleukin-4 (IL4), which protects mechanically overloaded chondrocytes from degradation, was decreased by 3.0 µM Yoda1 in HC (0.252 ± 0.092), no similar effect was observed in pCH-OA (Figure 6a). On the contrary, the proinflammatory IL6 and IL8 were increased by Yoda1 in both cell types (Figure 6b,c). A significant increase of IL6 was detected at 1.0 µM Yoda1 (HC: 4.957 ± 2.077; pCH-OA: 3.461 ± 1.819). As for the expression of IL8, a marked increase was induced by Yoda1 in HC (0.3 µM: 3.962 ± 1.407; 1.0 µM: 8.868 ± 2.717; 3.0 µM: 24.690 ± 6.480). Expression of IL8 was much more pronounced by higher Yoda1 concentration in pCH-OA (1.0 µM: 3.607 ± 0.547; 3.0 µM: 148.158 ± 15.472). Our results, therefore, show that under OA conditions, the extent of Yoda1-induced changes is lower than in healthy cells, except for interleukins. This could lead to the interpretation that mechanical activation under OA has a positive effect in the absence of inflammation. pCH-OA: 3.461 ± 1.819). As for the expression of IL8, a marked increase was induced by Yoda1 in HC (0.3 µM: 3.962 ± 1.407; 1.0 µM: 8.868 ± 2.717; 3.0 µM: 24.690 ± 6.480). Expression of IL8 was much more pronounced by higher Yoda1 concentration in pCH-OA (1.0 µM: 3.607 ± 0.547; 3.0 µM: 148.158 ± 15.472). Our results, therefore, show that under OA conditions, the extent of Yoda1-induced changes is lower than in healthy cells, except for interleukins. This could lead to the interpretation that mechanical activation under OA has a positive effect in the absence of inflammation.

Regulators and Transmitters of Mechanical Signals in pCH-OA and HC
To uncover potential differences in the transmission of mechanical signals in diseased compared to healthy cells, relevant targets were analyzed for their fold changes in both cell types (Figure 7a). A 7.0-fold higher expression of integrin α1 as well as a 2.5-and 2.1fold higher expression of integrin α5 and β1, respectively, were detected compared with the expression level of HC. The expression of TIMP1 was significantly lower in pCH-OA (0.512 ± 0.064), whereas receptors and Ras-related proteins involved in integrin activation and mechanical signaling were slightly higher expressed than in HC (interleukin-6receptor, IL6R: 1.324 ± 0.270; Ras-related protein RAP2: 1.877 ± 0.345; Discoidin domain receptor tyrosine kinase 2, DDR2: 2.064 ± 0.445).
Application of mechanical stimulation to pCH-OA using the Flexcell™ tension system for 48 h and testing for ECM-modulating enzymes revealed changes in the expression of MMP1 and MMP3 (Figure 7b). The expression of MMP1 changed in the same direction as in cells stimulated with the Yoda1 agonist (1.740 ± 0.268). In contrast, the expression of MMP3 decreased significantly (0.398 ± 0.021). While MMP13 did not

Regulators and Transmitters of Mechanical Signals in pCH-OA and HC
To uncover potential differences in the transmission of mechanical signals in diseased compared to healthy cells, relevant targets were analyzed for their fold changes in both cell types (Figure 7a). A 7.0-fold higher expression of integrin α1 as well as a 2.5-and 2.1-fold higher expression of integrin α5 and β1, respectively, were detected compared with the expression level of HC. The expression of TIMP1 was significantly lower in pCH-OA (0.512 ± 0.064), whereas receptors and Ras-related proteins involved in integrin activation and mechanical signaling were slightly higher expressed than in HC (interleukin-6-receptor, IL6R: 1.324 ± 0.270; Ras-related protein RAP2: 1.877 ± 0.345; Discoidin domain receptor tyrosine kinase 2, DDR2: 2.064 ± 0.445).  (Figure 7c). While we observed markedly increased expression of components of the apparatus that detects and transmits mechanical signals, stimulation with the Flexcell™ system only partially reflected the changes in expression also observed under Yoda1. These data may support the view that in OA, sensitivity to certain mechanical forces is altered by overexpression of the relevant proteins. Preliminary studies of downstream signaling triggered by Flexcell™ revealed only vague differences, leading to the interpretation that the type, level, and intensity of mechanical stimulation may play a significant role.  For (a,c), statistical analyses were performed by using Student s t-test, and for (b), the Mann-Whitney test was applied. *: p < 0.05, **: p < 0.01, ***: p < 0.001.

Discussion
Articular cartilage enables painless, low-friction movement of synovial joints and contains loosely distributed, highly specialized cells (i.e., chondrocytes) embedded in a matrix that imparts remarkable mechanical properties to articular cartilage. Mechanical stimulation, including fluid shear stress, stretching, compression, and cell swelling, as well as decreased mechanical conditions, can alter cellular membrane potential, thereby regulating the metabolism of articular cartilage [24,25]. In damaged cartilage and  For (a,c), statistical analyses were performed by using Student's t-test, and for (b), the Mann-Whitney test was applied. *: p < 0.05, **: p < 0.01, ***: p < 0.001. Application of mechanical stimulation to pCH-OA using the Flexcell™ tension system for 48 h and testing for ECM-modulating enzymes revealed changes in the expression of MMP1 and MMP3 (Figure 7b). The expression of MMP1 changed in the same direction as in cells stimulated with the Yoda1 agonist (1.740 ± 0.268). In contrast, the expression of MMP3 decreased significantly (0.398 ± 0.021). While MMP13 did not seem to be affected and Col1A1 was only insignificantly increased, IL6 was reduced by half under the influence of mechanical stimulation (0.461 ± 0.027).
GTPase isoforms associated with MAPK/ERK signaling and integrin regulation were examined for their expression behavior for the first time. To translate the results to signal transduction associated with mechanical stimulation and for comparison with HC, integrins, the purinergic receptor P2Y2, and downstream regulators were tested to gain initial insight. Cells were cultured under control conditions and by applying mechanical stimulation. Normalization to unstimulated cells showed differences in regulation in HC and pCH-OA. Under the influence of mechanical stimulation, P2Y2 decreased in HC, while no change was detected in pCH-OA (HC: 0.679 ± 0.055; pCH-OA: 1.087 ± 0.126). Src and H-Ras slightly increased under mechanical stimulation in HC (Src: HC: 1.407 ± 0.511; pCH-OA: 1.132 ± 0.123; H-Ras: HC: 1.219 ± 0.032; pCH-OA: 1.026 ± 0.035), whereas K-Ras increased at the same rate in both cell types (Figure 7c). While we observed markedly increased expression of components of the apparatus that detects and transmits mechanical signals, stimulation with the Flexcell™ system only partially reflected the changes in expression also observed under Yoda1. These data may support the view that in OA, sensitivity to certain mechanical forces is altered by overexpression of the relevant proteins. Preliminary studies of downstream signaling triggered by Flexcell™ revealed only vague differences, leading to the interpretation that the type, level, and intensity of mechanical stimulation may play a significant role.

Discussion
Articular cartilage enables painless, low-friction movement of synovial joints and contains loosely distributed, highly specialized cells (i.e., chondrocytes) embedded in a matrix that imparts remarkable mechanical properties to articular cartilage. Mechanical stimulation, including fluid shear stress, stretching, compression, and cell swelling, as well as decreased mechanical conditions, can alter cellular membrane potential, thereby regulating the metabolism of articular cartilage [24,25]. In damaged cartilage and misdirected chondrocytes (OA), mechanical stimuli can affect cartilage homeostasis and accelerate cartilage degeneration. Although the loss of TRPV4 in chondrocytes did not alter OA pathogenesis after joint destabilization [26], the imbalanced function of mechanosensitive ion channels, such as Piezo1 and TRPV4, has been implicated in causing cartilage damage [27,28]. Related to this, we found that the Ca 2+ increase triggered by Yoda1 is mainly due to calcium influx from the extracellular space and that the activation of the two mechanosensitive ion channels Piezo1 and TRPV4 are linked to each other. TRPV4 itself has been shown to interact with transmembrane proteins such as integrins and cytoskeletal elements but can also relay a range of environmental signals [29]. In pancreatic cells, it was discovered that stimulation by Piezo1 triggers TRPV4 channel opening, leading to intracellular dysfunction [30]. From our observations, we can conclude that Piezo1 activation inhibits subsequent TRPV4 activation to a greater extent than vice versa. By examining the effect of increased intracellular calcium through the application of calcium ionophores, we were able to exclude the possibility that inhibition of channel activation underlies pure changes in intracellular calcium. We, therefore, postulate that this dependence is due to a direct or indirect interaction of the two channel proteins. The correlation between Yoda1 and GSK after Yoda1 was stronger in pCH-OA cells, which could be explained by the increased expression of mechanosensitive ion channels, particularly TRPV4. It seems likely, therefore, that a situation similar to that in endothelial cells exists, where the effects of shear stress have been shown to be triggered by Piezo1 but to require TRPV4 [31].
As a limitation of this study, however, we need to consider that the 2D culturing models used in this study might not reflect the 3D environment of native articular cartilage. In addition, although we found a mutual influence of Yoda1 and GSK channel activation in that one seems to suppress the other, we cannot exclude that this is also influenced by the different modes of action of the two agonists. Since we assume that the two channels influence each other during activation, either by a direct or indirect mechanism, it is essential for a better understanding to further investigate the contribution of the two ion channels to this effect and also to include this in the context of OA in the next planned experiments.
As a perspective from another angle, it is also likely that OA changes the communication between ion channels leading to altered sensitivity to a mechanical signal. Subsequently, such changes may affect the expression profile of, i.e., MMPs, ILs, and BMP2 in chondrocytes and, thus, modulating the maintenance of the ECM. In this regard, stimulation of OA cells by Yoda1 appeared to downregulate the expression of ECM-modifying enzymes, suggesting that exercise may arrest ECM destruction to some extent. It is also interesting that Yoda1 did not induce BMP2 in pCH-OA cells to the same extent as in healthy cells, meaning that activation of Piezo1 in OA cells appears to result in lower production of BMP2. Notably, serum and synovial fluid BMP2 levels were shown to be associated with radiographic and symptomatic severity of knee OA, allowing BMP2 to serve as an alternative biochemical parameter for determining the disease severity of primary knee OA [32,33]. Because it was shown that serum BMP-2 levels were higher in patients with knee OA than in healthy controls, our observation of a decrease in BMP2 expression in pCH-OA cells under Yoda1 could be interpreted as a protective effect in OA cells. On the other hand, Yoda1 increased the expression of proinflammatory cytokines in healthy and OA cells, suggesting that Yoda1-activated Piezo1 increases inflammation under all circumstances, as also demonstrated in cardiac fibroblasts [34]. Furthermore, Lee et al. 2021 observed an IL1α-mediated inflammatory signaling pathway in articular cartilage cells that upregulates Piezo1 expression [35]. From these perspectives, it is important to weigh whether exercise/exercise therapy was beneficial or more counterproductive in the context of OA, where Piezo1 is considered a sensor of mechanical force and a transducer of that force into biological effects [36].
Our data provide preliminary evidence that there is channel interaction in chondrocytes and that this channel interaction is relaxed in OA cells, possibly due to increased TRPV4 levels leading to changes in the expression of ECM-modulating enzymes and interleukins. Evidence suggests that both Piezo1 and TRPV4 contribute to separate but overlapping mechano-electrical transduction pathways in chondrocytes [37]. While channel activation via Yoda1 appears to be quite intense, mechanical tension on chondrocytes showed similar but only minor effects, suggesting that the expression of relevant proteins is strongly dependent on the intensity of mechanical stimulation. It is also interesting to note that components of the integrin complex are more highly expressed in pCH-OA cells. This observation is equivalent to the cells becoming more sensitive in the diseased state or/and, on the other hand, better able to cope with mechanical stress. Elements of intracellular signal transduction showed only subtle changes, e.g., in Ras expression, so that, in view of the massive changes in expression of some other targets, it seems particularly important to find out what the differences in intracellular signal transduction between OA and healthy cells might be.

Cartilage Samples and Cell Culture
This study was carried out at the Medical University of Graz and the Medical University of Vienna, Austria, in accordance with the ethical standards of the responsible committee (Ethics vote number of the Medical University of Graz: 31-133ex18/19; Ethics vote number of the Medical University of Vienna: 1822/2017) and the Helsinki Declaration. Patients (n = 15) aged 73.4 ± 5.8 years who underwent end-stage knee arthroplasty for knee OA were included and gave their written informed consent. Femoral cartilage was collected intraoperatively and digested using 2 mg/mL collagenase B (Gibco Invitrogen, Carlsbad, CA, USA) in a chondrocyte growth medium at 37 • C for 24 h. After filtration using a 40 µm filter and centrifugation, the primary chondrocytes (pCH-OA) were seeded in cell culture flasks. For pCH-OA, passages 1 and 2 were used for experiments. Human healthy chondrocytes (HC), isolated from a 65-year-old Caucasian male purchased from Cell Application, Inc. (San Diego, CA, USA), were used as healthy control cells up to passage 4, based on experimental experience. To prevent and counteract the dedifferentiation processes of chondrocytes, the chondrocyte-specific growth medium consisting of DMEM/F12, supplemented with 10% fetal bovine serum (FBS), 1% Penicillin-Streptomycin (5.000 U/mL), 1% L-Glutamine, 1% Insulin-Transferrin-Selen, 0.01% TGF-β (10 ng/mL), and 0.01% FGF (10 ng/mL; all Gibco Invitrogen) was used throughout all experiments. Cells were kept in a humidified 5% CO 2 atmosphere at 37 • C.  [38,39]. Background subtraction, rationing of images, and calculation of [Ca 2+ ] i were performed offline using the Sigma Plot software version 14.5 (Systat Software Inc., San Jose, CA, USA). Histamine (Sigma Aldrich) at concentrations of 1, 3, 10, 30, or 100 µM in NT Tyrode's solution was applied by a superfusion system with a 7-channel perfusion pipette (List-electronic, Darmstadt, Germany). The system was driven by a valvebank (TSE, Bad Homburg, Germany) with a solution exchange time of less than 500 ms.

Reverse Transcription Polymerase Chain Reaction (RT-PCR)
Total RNA was isolated 24 h after treatment with DMSO (vehicle control, ctrl), 0.1, 0.3, 1, and 3 µM Yoda1 using the RNeasy Mini Kit and DNase-I treatment according to the manufacturer's manual (Qiagen, Hilden, Germany). Two µg RNA quantified and qualified via optical density measurements were reverse transcribed with the iScript-cDNA Synthesis Kit (BioRad Laboratories Inc., Hercules, CA, USA) using a blend of oligo(dT) and hexamer random primers. Amplification was performed with the SsoAdvanced Universal SYBR Green Supermix (Bio-Rad Laboratories Inc., Hercules, CA, USA) using technical triplicates and measured by the CFX96 Touch (BioRad Laboratories Inc.). The primer sequences, respectively, the QuantiTect primer assay (Qiagen) order numbers for the matrix metalloproteases (MMP1, MMP3, MMP13), the tissue inhibitors of metalloproteinases (TIMP2), the bone morphogenetic protein (BMP-2), type I collagen (Col1A1), type II collagen (Col2A1), interleukins (IL4, IL6, and IL8, the integrin subunits α1, α5, and β1, TIMP-1, the interleukin-6-receptor (IL6R), Ras-related protein (RAP2), Discoidin domain receptor tyrosine kinase 2 (DDR2), K-Ras, H-Ras, Src, and P2Y2 are listed in Table 1. The results were analyzed using the CFX manager software for CFX Real-Time PCR Instruments (Bio-Rad Laboratories Inc., version 3.1) software, and the quantification cycle values (Ct) were exported for statistical analysis. The results with Ct values greater than 32 were excluded from the analysis. Relative quantification of the expression levels was obtained by the ∆∆Ct method, based on the geometric mean of the internal controls ribosomal protein, large, P0 (RPL), and TATA box binding protein (TBP), respectively. The expression level of the target gene was normalized to the reference genes (∆Ct), and the ∆Ct of the test sample was normalized to the ∆Ct of the control (∆∆Ct). Finally, the expression ratio was calculated with the 2 −∆∆Ct method.  TRPV4  QT  00077217  MMP1  fw  CTG TTC AGG GAC AGA ATG TGC T  rev  TCG ATA TGC TTC ACA GTT CTA GGG  MMP3  fw  TTT TGG CCA TCT CTT CCT TCA  rev  TGT GGA TGC CTC TTG GGT ATC  MMP13  fw  TCC TCT TCT TGA GCT GGA CTC ATT  rev  CGC TCT GCA AAC TGG AGG TC  TIMP-2  fw  ACA GGC GTT TTG CAA TGC A  rev  GGG TTG CCA TAA ATG TCG TTT C  TIMP-1  QT  00084168  BMP-2  QT  00012544  Col1A1  QT  00037793  Col2A1  QT  00049518  IL6  QT  00083720  IL8  QT  00000322  IL4  fw  ATC TTT GCT GCC TCC AAG AAC AC  rev  GTA GAA CTG CCG GAG CAC AG  N-Ras  fw  CCC GGC TGT GGT CCT AAA TC  rev  TCC AAC CAC CAC CAG TTT GT  K-Ras  fw  TGT TCA CAA AGG TTT TGT TCT C  rev  CCT TAT AAT AGT TTC CAT TGC CTT G  Src  fw  TGT TCG GAG GCT TCA ACT CC  rev  CCA CCA GTC TCC CTC TGT GT  P2Y2 fw CCG CTC GCT GGA CCT CAG CTG rev CTC ACT GCT GCC CAA CAC ATC Integrin α1 QT 00093723 Integrin α5 QT 00080871 Integrin β1 QT 00068124

Mechanical Stimulation of Chondrocytes
Using the Flexcell FX5K™ Tension System (Flexcell International Corp, Burlington, NC, USA), mechanical cyclic tension straining was applied to the chondrocytes. Six wellpronectin-coated BioFlex culture plates with a flexible silicone membrane allowed deformation of the attached cells. After reaching 70-80% confluence, cells were mechanically stimulated with 15% elongation and 0.2 Hz over a period of 48 h. Immediately afterward, the RNA was isolated, and RT-PCR was performed as indicated above (HC: n = 4; pCH-OA: n = 8).

Statistical Analysis
Statistical analysis of the data was performed using the standard Microsoft Excel program and the statistical program GraphPad Prism 9. In addition to descriptive statistics, significant differences were determined at the non-categorical scale levels for normally distributed data using Student's t-test in an unpaired experimental design; data without normal distribution were tested for significance compared to controls using the Mann-Whitney test for unpaired data comparison, for paired analysis the Wilcoxon test was used. As pairs, we considered data that describe the different effects within one and the same experiment and, therefore, within one and the same cells. The relationship between the two variables was examined using Pearson's correlation analysis. Two-sided p-values (p < 0.0001 ****; p < 0.001 ***; p < 0.01 **; p < 0.05 *) were considered statistically significant.

Conclusions
From our studies, we suggest the existence of efficient interchannel communication between Piezo1 and TRPV4, which appears to be altered in OA chondrocytes. This alteration and the mechanisms leading to an increase in [Ca 2+ ] i could be responsible for an altered expression profile of clinically relevant OA marker genes. In any case, it is clear that strong activation of Piezo1 leads to a decrease in ECM-destroying proteins, whereas inflammatory factors increase. Comparison with mechanical tension also suggests that MMP and IL expression is strongly dependent on the intensity of ion channel activation. Thus, it can be concluded that mechanical stimulation can have different consequences in the treatment of OA and that it should be used carefully.