Altered spontaneous calcium signaling of in situ chondrocytes in human osteoarthritic cartilage

Intracellular calcium ([Ca2+]i) signaling is an essential universal secondary messenger in articular chondrocytes. However, little is known about its spatiotemporal features in the context of osteoarthritis (OA). Herein, by examining the cartilage samples collected from patients undergoing knee arthroscopic surgery, we investigated the spatiotemporal features of spontaneous [Ca2+]i signaling in in situ chondrocytes at different OA stages. Our data showed zonal dependent spontaneous [Ca2+]i signaling in healthy cartilage samples under 4 mM calcium environment. This signal was significantly attenuated in healthy cartilage samples but increased in early-degenerated cartilage when cultured in 0 mM calcium environment. No significant difference was found in [Ca2+]i intensity oscillation in chondrocytes located in middle zones among ICRS 1–3 samples under both 4 and 0 mM calcium environments. However, the correlation was found in deep zone chondrocytes incubated in 4 mM calcium environment. In addition, increased protein abundance of Cav3.3 T-type voltage dependent calcium channel and Nfatc2 activity were observed in early-degenerated cartilage samples. The present study exhibited OA severity dependent spatiotemporal features of spontaneous [Ca2+]i oscillations of in situ chondrocytes, which might reflect the zonal specific role of chondrocytes during OA progression and provide new insight in articular cartilage degradation during OA progression.

Articular cartilage is a highly organized tissue, its zonal differences in biochemical content and bioelectrical properties are directly associated with depth-dependent changes in mechanical properties 13,14 . Rolauffs and colleagues demonstrated the significantly increased cell density in the deep zone of synchronized primary metabolic chondrons of articular cartilage, compared with superficial zone 15 . Several studies 16,17 confirmed the variety in cytokine sensitivity of chondrocytes located in different zones, indicating the distinctive role of chondrocytes located in different zones during the pathological process of OA. Previous studies investigating chondrocyte [Ca 2+ ] i signaling were mainly focused on cells removed from their physiological environment, such as 2D cultures or chondrocytes/gel constructs 6,18 . These studies neglected the potential role of calcium channel activity linked to ECM, as well as the specialized peri-cellular matrix (PCM) that affected chondrocyte mechanics 19 .
In the present study, we aimed to investigate the spontaneous [Ca 2+ ] i signaling of human articular chondrocytes located in its native PCM and ECM, and compared the zonal differences in [Ca 2+ ] i signaling of in situ human chondrocytes. By comparing the spatiotemporal features of spontaneous [Ca 2+ ] i signaling in in situ chondrocytes at different OA stages, the correlation between spontaneous [Ca 2+ ] i signaling and OA severity was investigated. In addition, to address the possible mechanism of OA chondrocytes [Ca 2+ ] i signaling alternation and downstream regulation of [Ca 2+ ] i signaling, the protein level of T-type voltage-dependent calcium channels (VDCCs) and Nfatc2 activity were analyzed using total knee arthroplasty (TKA) samples.

Results
Histology evaluation. Based on the International Cartilage Research Society (ICRS) grading system, the severity of each collected human articular cartilage sample was verified by histomorphometry approaches. Later on, the spatiotemporal features of spontaneous [Ca 2+ ] i signaling analyzed from each sample were classified into ICRS grades. Representative hematoxylin and eosin (H&E), alcian blue, and sirius red staining images for ICRS 0-3 grades were shown in Fig. 1. As expected, cartilage samples showed increased wear from top to bottom. As shown by alcian blue staining, gradually decreased proteoglycan concentration was most pronounced at superficial and deep zones in degenerated samples (ICRS 1-3). Sirius red staining suggested dramatically changed collagen composition from type П to type І at ICRS 1 samples, and absence of both types of collagen fibrils at late stage OA samples (ICRS 3).

Propagation of calcium signaling in middle and deep zones of articular cartilage samples.
Spontaneous calcium signaling was observed in chondrocytes within both normal (ICRS 0) and OA (ICRS 1-3) articular cartilages without extraneous stimuli. Typical [Ca 2+ ] i oscillations in ICRS 0 cartilage sample were shown in Fig. 2b. Propagation of [Ca 2+ ] i oscillations were observed in a large portion of chondrons in middle and deep zones of both normal and OA cartilages. As shown in Fig. 2c Fig. 3. In ICRS 0 samples, [Ca 2+ ] i oscillations increased with the depth of cartilage (Fig. 3a). During the 15-min recording period, only a small portion of chondrocytes (13.33 ± 13.33%) located in the superficial zone displayed [Ca 2+ ] i peaks larger than 2 fold increment (calcium peak magnitude: 1.60 ± 0.07) (Fig. 3b), while a significantly larger portion of chondrocytes (47.17 ± 4.01%, P < 0.05) located in single chondrons of deep zone demonstrated repetitive (0.60 ± 0.07 vs. 0.13 ± 0.09 peaks, P < 0.05) and stronger [Ca 2+ ] i peaks (calcium peak magnitude: 2.17 ± 0.08, P < 0.01) during the same recording period (Fig. 3b) Correlation between spontaneous calcium signaling and OA severity. To analyze the possible correlation between spontaneous calcium signaling and OA severity, Spearman's rank correlation coefficient method was used to analyze the relationship of [Ca 2+ ] i signaling parameters of ICRS 0-3 samples to ICRS grading system under 4 mM and 0 mM [Ca 2+ ] e environment, respectively (Table 1). Results suggested that calcium peak magnitude (r = −0.33, P < 0.001) and responsive rate (r = −0.486, P < 0.001) were moderately correlated with ICRS grade when cartilage samples were incubated in 4 mM [Ca 2+ ] e environment. In contrast, when cartilage samples were incubated in 0 mM [Ca 2+ ] e environment, the correlation between parameters of [Ca 2+ ] i signaling and ICRS grading system was very weak.

Increased protein abundance of T-type VDCCs in early-degenerated articular cartilage.
To explore the mechanism of unregulated [Ca 2+ ] i signaling in ICRS 1 samples, non-degenerated (Normal) and early-degenerated (OA) regions of TKA cartilage samples were used for RT-qPCR and Western boltting analyses. Compared with non-degenerated cartilage, chondrocytes in early-degenerated cartilage had significantly higher level of MMP-1, MMP-13, and COL1A1; and lower level of SOX-9 and SOX-11 (Fig. 5a). Western boltting data suggested that chondrocytes located in both regions expressed T-type Ca v 3.1, Ca v 3.3, and α2σ1 proteins, but negatively expressed Ca v 3.2 protein. When compared with the non-degenerated region, significantly increased protein level of Ca v 3.3 was noticed in the early-degenerated region ( Fig. 5b and c).
Increased Nfatc2 activity in early-degenerated articular cartilage. To explore the role of unregulated [Ca 2+ ] i signaling in OA progression, the activity of Nfatc2, downstream transcription factors of [Ca 2+ ] i signaling was analyzed using RT-qPCR and Western boltting in TKA cartilage samples. Results indicated significantly up-regulated Nfatc2 expression (Fig. 6a), and increased protein level of de-phosphorylated Nfatc2 in chondrocytes located in early-degenerated regions ( Fig. 6b and c).

Discussion
The objective of this study was to investigate the spatiotemporal features of spontaneous [Ca 2+ ] i signaling in in situ chondrocytes of human healthy and OA articular cartilage, which might shed light on understanding the molecular mechanism of articular cartilage degradation in OA. Our data showed a significant difference of extracellular calcium dependent spontaneous [Ca 2+ ] i signaling between zones in healthy cartilage, while this zonal difference was not observed in degenerated cartilage. In early-degenerated cartilage, abnormal [Ca 2+ ] i activities in deep zone were seen when samples were cultured in both 0 mM and 4 mM [Ca 2+ ] e environment. Increased Ca v 3.3 protein abundance and Nfatc2 activity were possibly associated with abnormal [Ca 2+ ] i activities in early-degenerated cartilage. Also, the intensity of spontaneous [Ca 2+ ] i signaling was found negatively correlated with OA severity. These data indicated that the zonal specific role of chondrocytes in maintaining ECM integrity in human healthy articular cartilages and reflected alternation of chondrocytes metabolic activity during OA progression.
It is recently discovered that both isolated and in situ animal chondrocytes can display repetitive [Ca 2+ ] i signaling without any extracellular stimuli. Occasional spontaneous [Ca 2+ ] i signaling in rabbit chondrocytes located in native ECM and 2D culture was first observed by Kono et al., 21 , mature chondrocytes are well equipped with rich complement of calcium channels such as voltage-dependent calcium channels (VDCCs) 22 , mechanical sensitive calcium channels (MSCCs) 18 , transient receptor potential vanilloid (TRPV) 7 , and Ca 2+ release-activated Ca 2+ (CRAC) channel 23 . These calcium channels synergistically regulate [Ca 2+ ] i homeostasis by modulating    were relatively inactive during the whole process of OA. In contrast, decreased intensity of [Ca 2+ ] i signaling in deep zones of late-stage OA cartilages was found among grades. Decreased calcium channel activity caused by alternation in channel structure of OA chondrocytes could be one of the reasons 22 . Animal model using combination of orally administered monosodium iodoacetate (MIA) and treadmill over-loading showed decreased protein level of TRPV5, a voltage-independent cation channel in OA chondrocytes 31 . A recent study using OA animal model and human sample also revealed decreased TRPV6 expression in degraded cartilages 32 . In addition, by using RNA-seq analysis, Dunn et al. compared gene expression changes in human OA cartilage samples and found that α2/δ, a subunit of voltage-dependent calcium channels regulating channel properties and increasing the functional channel 33 , was significantly down-regulated 34 . By using the larger volume of cartilage sample acquired during total knee arthroplasty (TKA) surgery, we showed that chondrocytes in the early-degenerated area had significantly increased expression level of MMP-1, MMP-13, COL1A1; and decreased expression level of SOX-9 and SOX-11. In addition, increased protein abundance of Ca v 3.3, a α1 subunit of T-type VDCC in the early-degenerated area was noticed. The transiently active Ca v 3.3 T-type VDCC requires weak membrane depolarization for activation. Increased Ca v 3.3 protein abundance might decrease the membrane potential 35 Figure S1). (c) Compared with non-degenerated cartilage (Normal, black bars), Significantly increased dephosphorylated Nfatc2 was found in early-degenerated (OA, grey bars) regions ( * P < 0.05) (n = 8).
presented; and ultimately activated Nfatc2, a transcription factor has been shown to be crucial in chondrogenesis in tracheal cartilage 36 , and function regulation of adult chondrocytes 37 . Despite the fact that early OA was most pronounced in the superficial and middle zones 38 , our data suggested a unique role of deep zone chondrocytes in the initiation of OA. In early OA, the increased Nfatc2 activity might serve as a compensatory mechanism to withstand the established damage in articular cartilage. There is a growing consensus that OA is a whole-joint disease 39 . Increased turnover of subchondral bone seen in OA patients 40 and animal models 41 indicated an important role of subchondral bone-deep zone cartilage crosstalk during the onset of OA. Our previous studies using human OA samples and destabilization of medial meniscus (DMM) model showed decreased thickness in calcified cartilage zone (CCZ) at early stage 40 , accompanied by increased invading vessels density in this region 41 . These observations suggested elevated paracrine transport of inflammatory factors and cytokines released from subchondral bone could influence the calcium channel activities of deep zone chondrocytes, ultimately altering the ECM components such as proteoglycan concentration and composition of collagen observed in the present study.
The present study has several limitations. Firstly, although human biopsy samples at different OA stages represent the real progression of OA and retain the native PCM and ECM, they also suffer from disadvantages such as small sample volume, limited sample quantity, and restricted location for sample collection. In the current study, only a small volume of cartilage sample was acquired from each patient to prevent further damage to articular cartilage. To exclude the individual variation between patients, larger samples size and paired control sample representing non-degenerated cartilage from the same individual seems necessary for mechanism study. Secondly, the role of elevated spontaneous [Ca 2+ ] i signaling in OA progression remains unclear. More stable sample source such as cartilage collected from the large-animal model is needed for performing [Ca 2+ ] i dependent downstream pathway analysis.
Despite the limitations, the current study presented the direct evidence that in situ chondrocytes resided in both healthy and OA human articular cartilage displayed spontaneous [Ca 2+ ] i signaling, and indicated that the spatiotemporal features of the [Ca 2+ ] i oscillations were both zonal and OA severity dependent. While the exact mechanism needs further investigation, the changes of [Ca 2+ ] i signaling might reflect the zonal specific role of chondrocytes during OA progression and provide new insight into molecular mechanism of articular cartilage degradation during OA progression. Also, our observation here also suggested that spontaneous [Ca 2+ ] i signaling of in situ chondrocytes could serve as a potential biomarker for clinical diagnosis of early OA.

Methods
Cartilage explants. Articular cartilage samples for calcium imaging and histology evaluation were collected from patients receiving knee arthroscopy by one senior surgeon in Center for Joint Surgery, Southwest Hospital, Third Military Medical University, Chongqing, China. Informed consent was obtained from participant patients. During the surgical process, a single 2 mm in diameter cylindrical osteochondral sample was collected using a custom-designed biopsy needle that fits the arthroscopic system. For ICRS 0 group, healthy cartilage samples were collected in the non-load bearing area in knee joints with no sign of cartilage degeneration. For ICRS 1-3 groups, samples representing the most severe wear in each knee were collected from the femoral condyle before chondroplasty. The severity of cartilage degradation in their knee joints was evaluated during surgery based on ICRS Visual Assessment Scale, and confirmed by histomorphometry approaches afterward. Cartilage samples containing defects caused by trauma, tumor, or rheumatoid arthritis (RA) were excluded from this study. 3-4 cartilage samples for each ICRS grade were collected from 15 patients (Age: 42.69 ± 2.5 years; BMI: 23.55 ± 0.64; 5 males, 10 females; Table 2). This study was conducted in accordance with the declaration of Helsinki and with approval from the Ethics Committee of the Southwest Hospital of Third Military Medical University (Chongqing, China).

Calcium imaging of in-situ chondrocytes.
Harvested samples containing full-thickness cartilage and subchondral bone were immediately transferred to laboratory in PBS (Hyclone, Beijing, China). Samples were carefully cut into 2 half-cylinders using a scalpel. [Ca 2+ ] i transients were monitored in in situ chondrocytes using Fluo-8 AM (AAT Bioquest, CA, USA) and fluorescence microscopy (IX71, Olympus, Tokyo, Japan). As shown in Fig. 2, collected samples were stained by 5 µM Fluo-8 AM at 37 °C for 30 min; excess dye was removed by washing with PBS for three times. Stained cartilage samples were then transferred into 96-well plates and kept in standard bath solution contained (in mM): NaCl, 144; NaH 2 PO 4 , 0.33; KCl, 4.0; MgCl 2 , 0.53; glucose, 5.5; HEPES, 5.0; CaCl 2 , 0 or 4 mM (the pH was adjusted to 7.4 with NaOH) at 37 °C for 15 min to remove any agitation during staining process 26 . Fluorescent images of chondrocytes were then recorded using 4 x objectives at room temperature for 15 min while the sample was undisturbed. Image processing. To analyze the [Ca 2+ ] i signaling, recorded image sequences were analyzed using Image J (Version 1.44p). The cartilage samples were separated into superficial, middle and deep zones in the image software according to chondrocyte/chondron size and arrangement. In detail, parallelly-arranged cells in cartilage surface were considered as superficial zone chondrocytes; cells in middle range of cartilage samples, and with the formation of medium-size chondrons in random arrangement were considered as middle zone chondrocytes; hypertrophy cells in deep range, and with the formation of large-size chondrons in vertical arrangement were considered as deep zone chondrocytes. For the superficial zone, 5-10 individual cells were randomly chosen and analyzed in each sample. For the middle and deep zone, cells in 5-10 randomly chosen chondrons were analyzed in each sample. Fluorescence intensity in individual cells was normalized with the mean background intensity obtained in three randomly chosen blank areas 42 . The calcium peak magnitude was reported as the mean fold increment of the peak intensity over the mean baseline fluorescence intensity in all tested chondrocytes. The responsive rate was determined by the number of cells whose calcium peak magnitude was over 2-fold of baseline divided by the number of total cells in the superficial zone or each chondron in middle and deep zone. For analyzing the number of peaks during the 15-min recording period, peaks with increment larger than two were counted as one peak, and the peak number was presented as the total number of peaks during the 15-min recording period.
Histology evaluation. For validating the severity of OA in acquired biopsies, the samples used for calcium imaging were fixed after calcium imaging in 4% paraformaldehyde at 4 °C overnight, followed by decalcification for 1 week with 10% EDTA before embedding in paraffin. Serial sections were cut at up to 5 µm thickness and stained with H&E, alcian blue, and sirius red respectively. Cartilage sections were then evaluated using ICRS grading system as previously reported 43 .
RNA extraction and RT-qPCR. Articular cartilage for RT-qPCR and Western boltting collected from TKA surgery with informed consent obtained from participant patients. Cartilage samples collected form non-degenerated and early-degenerated regions were processed into small fractions with a tissue homogenizer (Precellys 24, Bertin, MD, USA) at 4 °C. Total RNA was extracted using RNeasy ® Mini Kit (Qiagen, Hilden, Germany). The purity of RNA was measured and quantitated on a Nanodrop-1000 spectrophotometer (Thermo Scientific, USA). l μg total RNA was used for cDNA synthesis. GAPDH was used as an internal reference, and target gene primers were presented in the supplementary information (see Supplementary information). RT-qPCR was performed in a reaction volume of 25 μl using QuantiTect SYBR Green PCR kit (Qiagen, Hilden, Germany). Assays were performed in triplicates, and the mRNA levels were normalized to GAPDH using the ∆∆CT method. Statistical analysis. Data were presented as mean ± SEM. For demographic data of patients, the statistical significance of age and BMI was determined by one-way ANOVA with Bonferroni's post hoc test. For spatiotemporal features of spontaneous [Ca 2+ ] i signaling, the statistical significance of the normalized fold increment, the number of peaks, and the responsive rate among all groups were determined by two-way ANOVA with Bonferroni's post hoc test. Spearman's rank correlation coefficient method was used to analyze the relationship between each parameter of [Ca 2+ ] i signaling to ICRS grading system. For RT-qPCR and Western boltting data, the data from OA groups were compared with a hypothetical value of 1.0 (Normalized Normal control value) by using the student's t-test. Statistical significance was defined with p < 0.05.
Data availability statement. The datasets generated during and/or analyzed during the current study are available from the corresponding author on reasonable request.