T-type Voltage-Sensitive Calcium Channels Mediate Mechanically-Induced Intracellular Calcium Oscillations in Osteocytes by Regulating Endoplasmic Reticulum Calcium Dynamics

Abstract

One of the earliest responses of bone cells to mechanical stimuli is a rise in intracellular calcium (Ca²⁺), and osteocytes in particular exhibit robust oscillations in Ca²⁺ when subjected to loading. Previous studies implicate roles for both the endoplasmic reticulum (ER) and T-Type voltage-sensitive calcium channels (VSCC) in these responses, but their interactions or relative contributions have not been studied. By observing Ca²⁺ dynamics in the cytosol (Ca²⁺_cyt) and the ER (Ca²⁺_ER), the focus of this study was to explore the role of the ER and T-Type channels in Ca²⁺ signaling in bone cells. We demonstrate that inhibition of T-Type VSCC in osteocytes significantly reduces the number of Ca²⁺_cyt responses and affects Ca²⁺_ER depletion dynamics. Simultaneous observation of Ca²⁺ exchange among these spaces revealed high synchrony between rises in Ca²⁺_cyt and depressions in Ca²⁺_ER, and this synchrony was significantly reduced by challenging T-Type VSCC. We further confirmed that this effect was mediated directly through the ER and not through store-operated Ca²⁺ entry (SOCE) pathways. Taken together, our data suggests that T-Type VSCC facilitate the recovery of Ca²⁺_ER in osteocytes to sustain mechanically-induced Ca²⁺ oscillations, uncovering a new mechanism underlying the behavior of osteocytes as mechanosensors.

1. Introduction

Osteocytes are widely regarded as mechanosensors, capable of detecting changes in the mechanical environment of the bone tissue and modifying cellular responses accordingly^1,2. Indeed, an intact osteocyte network is required for bone changes in response to unloading³, and studies have shown that both loading and unloading influence osteocyte expression of proteins that modulate bone turnover⁴, such as sclerostin^5,6 and receptor activator of nuclear factor kappa B ligand (RANKL)^7,8. Still, mechanisms underlying osteocyte mechanotransduction remain unclear. For instance, one of the earliest responses of bone cells to mechanical stimuli is a rise in intracellular, or cytosolic, calcium (Ca²⁺_cyt), but the mechanisms by which osteocytes generate or rely on Ca²⁺ signals to direct bone adaptation are largely unknown.

Osteocytes exhibit robust oscillations in Ca²⁺_cyt in response to mechanical stimulation, a pattern distinct from their osteoblast precursors and attributed to their mechanosensitivity^9,10. Prior in vitro work from our laboratory concluded that Ca²⁺_cyt transients depend on both the extracellular reservoir of Ca²⁺ ions and intracellular storage organelles, in particular the endoplasmic reticulum (ER)⁹. Mechanically-induced ER Ca²⁺ (Ca²⁺_ER) release depends on the purinergic pathway via inositol trisphosphate receptors (IP₃R) on the ER membrane. The release of Ca²⁺ from the ER is critical to fluid flow-induced Ca²⁺ oscillations in osteocytes; treatment with the drug thapsigargin to block ER Ca²⁺ reuptake significantly reduced the number of Ca²⁺_cyt transients from an average of five down to a single response. A similar effect was observed in in situ osteocytes, where thapsigargin treatment abolished multiple Ca²⁺ responses induced by dynamic loading of a murine tibia¹⁰.

Ca²⁺_cyt oscillations in osteocytes are also affected by inhibition of a number of membrane channels involved in Ca²⁺ transport, and targeting channels expressed primarily in osteocytes should clarify some mechanisms underlying this unique behavior. For instance, the expression of voltage-sensitive calcium channel (VSCC) subtypes changes as osteoblasts differentiate into osteocytes¹¹. Osteoblasts express both low threshold T- and high threshold L-type VSCC, whereas osteocytes predominantly express T-Type VSCC¹². Previous studies published from our laboratory explored the effects of VSCC inhibitors when added to the flow medium after shear stimulation9. The addition of the T-Type inhibitor NNC 55-0396 interrupted Ca²⁺_cyt responses in osteocytes, preventing subsequent Ca²⁺ transients, but had little observable effect on osteoblasts. Treatment of in situ osteocytes with the T-Type inhibitor prior to mechanical stimulation also significantly reduced the number of Ca²⁺ responses¹⁰. Interestingly, inhibition of these channels in both systems had similar effects as disruption of ER stores, though no link has been demonstrated between them in osteocytes.

VSCC have been shown to interact with the ER and Ca²⁺_ER release pathways in other cells. A direct association of T-Type VSCC with the ER has been previously demonstrated in arterial smooth muscle, where Ca_V3.2 channels were found to be localized to ER caveolae by transmission electron microscopy and to bind to ryanodine receptors on the ER by a proximity ligation assay¹³. In addition, a number of proteins participate in Ca²⁺ release from ER stores and Ca²⁺ entry triggered by this release – a phenomenon referred to as store-operated calcium entry (SOCE) – such as Stromal Interaction Molecules (STIMs) and Ca²⁺-Release Activated Channels (CRACs)¹⁴. In particular, STIM1 has been shown to interact with L-Type VSCC in rat cortical neurons15 and vascular smooth muscle cells¹⁶ and Ca_V3.1 T-Type VSCC in cardiac myocytes¹⁷. The potential role of SOCE-related proteins in osteocyte Ca²⁺ signaling have not yet been explored.

Though our previous studies implicate roles for both the ER and T-Type VSCC in osteocyte Ca²⁺_cyt responses, their relative contributions or any interactions between the ER and T-Type VSCC remain unknown. This is largely due to an inability to monitor Ca²⁺ localized to the ER separately from Ca²⁺ entering the cell from the extracellular fluid. However, advances in genetically encoded Ca²⁺ biosensors now enable the targeting of these sensors to subcellular organelles, including the ER^18–20.

Thus, the focus of the current study is to observe Ca²⁺_ER dynamics in bone cells to better understand the role of ER stores in the unique Ca²⁺ oscillations in osteocytes. We hypothesized that an ability to refill ER stores results in Ca²⁺_cyt oscillations in osteocytes. We also hypothesized that the predominant expression of T-type channels in osteocytes may contribute to their unique Ca²⁺_cyt patterns and further speculated that T-type VSCC in osteocytes may interact with ER stores.

2. Materials and Methods

2.1 Cell culture

Osteocyte-like MLO-Y4 cells (a gift from Dr. Lynda Bonewald, University of Missouri-Kansas City, Kansas City, MO) were cultured on 0.15 mg/ml collagen (rat tail type I, BD Biosciences, San Jose, CA) coated culture dishes in minimum essential alpha medium (α-MEM, Life Technologies, Carlsbad, CA) supplemented with 5% fetal bovine serum (FBS, Hyclone Laboratories Inc., Logan, UT) and 5% calf serum (CS, Life Technologies, Carlsbad, CA). MC3T3-E1 pre-osteoblasts (ATCC, Manassas, VA) were cultured in α-MEM supplemented with 10% FBS. Cells were maintained at 5% CO₂ and 37°C in a humidified incubator. MLO-Y4 cells were sub-cultured prior to reaching 70–80% confluence in order to maintain an osteocyte-like phenotype.

2.2 Immuno-detection

To detect the presence of L- and T-Type VSCC in osteocytes and osteoblasts, we performed Western blots and immunostaining using antibodies targeted against Ca_V1.2 (L-Type) and Ca_V3.2 (T-Type) subunits (SDIX, Newark, DE). Protein was extracted from monolayer cell cultures using RIPA lysis buffer with protease and phosphatase inhibitors (Roche Diagnostics, Indianapolis, IN). Western blots were performed on MLO-Y4 and MC3T3-E1 lysates of equal total protein as determined by a BCA assay (Pierce, ThermoFisher Scientific, Waltham, MA). For immunostaining, cells were fixed in 4% paraformaldehyde and permeabilized with 2% Triton™ X-100 (Sigma-Aldrich, St. Louis, MO). Following antibody incubations, cells were counterstained with DAPI to label nuclei (Molecular Probes, Eugene, OR).

2.3 Inhibitors

All inhibitors used in this study were purchased from Sigma-Aldrich (St. Louis, MO). Thapsigargin is an inhibitor of the Ca²⁺-ATPase pump on the ER (SERCA) which facilitates the reuptake of Ca²⁺ into the ER, and thapsigargin treatment (1µM) thereby results in ER depletion. Two VSCC inhibitors were used in this study: the T-Type inhibitor NNC 55-0396 (20 µM) and the L-Type specific inhibitor nifedipine (10µM). MLO-Y4 cells were also treated with the following inhibitors related to SOCE: YM58483 (3µM), a CRAC inhibitor²¹; SKF-96365 (50µM), a SOCE inhibitor that inhibits STIM1 (similar effects as STIM1 silencing)²² but exhibits off-target effects on VSCC activity²³; and 2-APB (50µM), a reliable inhibitor of SOCE and IP₃R^23,24. Cells were incubated in the inhibitors for 15 minutes prior to flow exposure, and the inhibitors remained in the flow medium for the duration of the experiment.

2.4 Ca²⁺_cyt indicators and Ca²⁺_ER visualization

To observe Ca²⁺_cyt changes only, MLO-Y4 and MC3T3-E1 cells were stained with Fluo-8 AM (AAT Bioquest, Sunnyvale, CA) dissolved in 20% Pluronic F-127 in DMSO (Invitrogen, Carlsbad, CA). To visualize ER calcium levels in osteocytes and osteoblasts, cells were transiently transfected with the D1ER plasmid (plasmid #36325, Addgene, Cambridge, MA)²⁵ using standard non-liposomal techniques (Fugene 6, Promega Corporation, Madison, WI). D1ER is a second generation cameleon Ca²⁺ fluorescence resonance energy transfer (FRET) sensor targeted to the ER with a retention sequence. To simultaneously visualize Ca²⁺_cyt and Ca²⁺_ER, cells were transfected with D1ER and then stained with the red-shifted Ca²⁺_cyt indicator Fura Red-AM (20µM, Life Technologies, Carlsbad, CA) in DMSO and Kolliphor® EL (Sigma-Aldrich, St. Louis, MO) prior to fluid flow stimulation.

To verify the localization of the D1ER plasmid, transfected cells were stained with ER-Tracker Red (Molecular Probes, Eugene, OR). Cells were rinsed and incubated with a warmed 1µM working solution of ER-Tracker Red for 30 minutes at 37°C. Cells were rinsed with fresh medium and post-incubated for 15 minutes prior to imaging. The FRET biosensor was excited at 430nm, and fluorescence emission of YFP (530nm) was collected as the FRET image. The ER Tracker Red dye was imaged using 568nm excitation/660nm emission.

2.5 Fluid flow stimulation

Prior to staining with appropriate Ca²⁺ indicators, cells were plated onto 10µg/mL fibronectin-coated glass slides at ~80% confluency to establish cell-cell contact. Slides were then stained and assembled into a custom parallel-plate flow chamber with a glass bottom that permits live cell imaging under fluid shear stimulation. The chamber was placed on the stage of an inverted microscope (Olympus, Waltham, MA) and attached to a magnetic gear pump (Scilog, Madison, WI) for the application of steady, laminar, unidirectional flow at a shear stress of 35 dynes/cm², which has been shown to induce multiple Ca²⁺_cyt responses in osteocytes and fewer, weaker responses in osteoblasts in our previous studies⁹. In addition, the Ca²⁺_cyt patterns observed in vitro under this flow profile are consistent with those observed in ex vivo mouse tibia under physiologic dynamic loads¹⁰, while oscillatory flow induces fewer responses²⁶. Baseline fluorescence intensity was captured for 1 minute prior to fluid shear stimulation for 9 minutes.

2.6 Imaging and image analysis

For monitoring ER depletion, time-lapse images were collected 36–48 hours post transfection at 40× magnification. The FRET biosensor was excited at 430nm, and fluorescence emissions of YFP (530nm) and CFP (470nm) were captured simultaneously using a quadview beamsplitter and custom quad-band polychroic²⁷. To monitor ER depletion, cells were imaged for 20 minutes following the addition of thapsigargin (1µM). Images were acquired every 3 seconds to minimize photobleaching. The FRET ratio was calculated on a pixel-by-pixel basis using cross-correlation based image registration of the FRET and donor emissions to obtain relative Ca²⁺_ER levels. The time to depletion was defined as the time at which the normalized FRET ratio fell below 0.8²⁸.

To monitor the effects of VSCC and SOCE inhibitors on Ca²⁺_cyt responses, time-lapse images of cells were collected at 20× magnification for 9 minute periods of fluid flow stimulation. The Fluo-8 Ca²⁺ indicator was excited at 488nm, and fluorescence emissions were collected at 527nm. A Ca²⁺_cyt transient was defined by an increase Fluo-8 intensity at least 4 times the magnitude of noise prior to flow onset²⁹.

To simultaneously monitor Ca²⁺_cyt and Ca²⁺_ER, single cells expressing Fura Red and the D1ER FRET sensor were identified at 60× magnification. Single excitation (430nm) was used to excite the dye and FRET biosensor, and fluorescence emissions of YFP, CFP, and Fura Red (641nm) were captured separately and simultaneously. Images were acquired every 3 seconds. Ca²⁺_cyt time-courses were extracted from the Fura Red image. Photobleaching was corrected in this channel using an exponential fit of the baseline Ca²⁺_cyt levels. The FRET ratio was determined from registered FRET/donor emissions. Regions of interest were selected to capture Ca²⁺ micro-domains within the cell based on the location of the initiation of a Ca²⁺_cyt response³⁰. To investigate the relationship between Ca²⁺ activity in the cytosol and the ER, we defined a percentage synchrony as the number of coincident Ca²⁺_cyt and Ca²⁺_ER transients divided by the total number of transients (peaks and inverse peaks). Ca²⁺_cyt peaks and decreases in Ca²⁺_ER were identified as transient responses (i.e. decrease and recovery) with a magnitude greater than four times the magnitude of the noise in the baseline images.

2.7 Data analysis

The data presented herein analyzes only responsive cells, though responsive percentages were greater than 80% for all experiments, except for SKF-96365 and 2-ABP treatment, which reduced the number of responsive MLO-Y4 cells to ~60% and 50%, respectively in cell populations. A cell was defined as responsive if the magnitude of the first Ca²⁺_cyt response was at least four times higher than fluctuations in intensity during baseline measurements. Student’s t-tests were used to determine significant differences for the number of Ca²⁺_cyt peaks at a specific experimental condition. All data are shown as mean ± standard deviation. Multi-factorial analysis of variance (ANOVA) with Bonferroni’s post hoc analysis was performed to determine statistical differences between mean values of inhibitor treatments on the two cell types. Statistical significance can be observed when p<0.05.

3. Results

3.1 Expression of L- and T-Type VSCC in osteocytes and osteoblasts

MLO-Y4 cells express little to no detectable L-Type protein by Western blot and little protein by immunohistochemistry (Fig. 1A). T-Type expression, on the other hand, is abundant in both protein extracts and fixed cells. MC3T3-E1 cells express both L- and T-Type proteins in similar abundance (Fig. 1B).

Figure 1. The expression of L- and T-Type VSCC in MLO-Y4 osteocytes and MC3T3-E1 osteoblasts.

Immuno-detection of L- and T-Type VSCC proteins from whole cell lysates (Western blot) and in fixed cell populations of (A) MLO-Y4 and (B) MC3T3-E1 cells.

3.2 Effects of L- and T-Type VSCC inhibitors on bone cell Ca²⁺_cyt responses to fluid flow

Osteocytes (n=480 cells, 18 slides) exhibited numerous Ca²⁺_cyt oscillations under fluid flow, and these multiple responses were significantly reduced by treatment with the T-type inhibitor NNC 55-0396 (n= 201 cells, 10 slides) (Fig. 2A). The L-type inhibitor had no effect on the number of Ca²⁺_cyt responses in osteocytes (n=187 cells, 8 slides). Osteoblasts were less responsive than osteocytes, exhibiting less than two Ca²⁺_cyt peaks on average for all groups (n=237, 53, 79 cells; 8, 2, 3 slides for control, L- and T-Type inhibitors, respectively) (Fig. 2B).

Figure 2. The effects of VSCC inhibitors on Ca2+cyt responses in osteocytes and osteoblasts.

Representative Ca²⁺_cyt time courses are shown for (A) MLO-Y4 and (B) MC3T3-E1 cells treated with the L-Type inhibitor nifedipine or the T-Type inhibitor NNC 55-0396 prior to exposure to fluid flow. Ca²⁺_cyt transients are shown as intensity changes of the fluorescent indicator Fluo-8 normalized to intensity at baseline prior to fluid flow exposure.

3.3 Effects VSCC inhibitors on ER dynamics in bone cells

The localization of the D1ER plasmid to the ER was verified by staining D1ER-transfected osteocytes with ER Tracker Red (Fig. 3A). Transfected cells were imaged for 20 minutes following the addition of thapsigargin to deplete the ER. Representative FRET ratios of a single osteocyte treated with thapsigargin is shown in Fig. 3B.

Figure 3. The effects of VSCC inhibitors on Ca²⁺ dynamics in the ER.

(A) Ca²⁺_ER was indicated by transfect of cells with the D1ER plasmid (yellow = FRET), which localizes to the ER. Images were collected at 40× magnification. (B) Thapsigargin was used to induce depletion of the ER over the time course of the experiment. The FRET ratio was calculated by registering and dividing the FRET and donor images and calculating the ratio within the cell. Depletion was determined by a reduction in the normalized FRET ratio to below 0.8. (C) Untreated MLO-Y4 exhibited similar depletion characteristics as those treated with an L-Type inhibitor. Treatment with the T-Type inhibitor resulted in significantly faster depletion, quantified in the bar graph. (D) In MC3T3-E1, no differences were detected in the depletion times among all groups. Error bars are standard deviation. Significance was assessed by one-way ANOVA. *p<0.05 compared to control.

The time to deplete the ER under the outlined conditions is shown in Fig. 3C and D. The scatter plots show representative depletion curves for MLO-Y4 cells (Fig. 3C) and MC3T3-E1 cells (Fig. 3D). The small difference (3%) between untreated MLO-Y4 cells (n=14) and MC3T3-E1 cells (n=6) in the time to depletion did not reach significance. Similarly, the effect of VSCC inhibitors had on the time to depletion in osteoblasts was not significant (Fig. 3D; n=5, 4 cells for L- and T-Type inhibition, respectively), and the L-Type VSCC inhibitor had no effect in osteocytes (Fig. 3C, n=5 cells). Treatment of MLO-Y4 cells with the T-Type VSCC inhibitor prior to the addition of thapsigargin resulted in a significantly faster time to depletion compared to untreated controls (Fig. 3C, n=6 cells). On average, it took ~300s longer for Ca²⁺_ER levels to deplete in control cells than those with disrupted T-Type VSCC.

3.4 Simultaneous imaging of Ca²⁺_cyt and Ca²⁺_ER in bone cells under fluid flow

To simultaneously image Ca²⁺_cyt and Ca²⁺_ER, bone cells were transfected with D1ER, stained with Fura Red-AM, and imaged using a quadview beamsplitter (Fig. 4A). The FRET ratio was calculated from the FRET and donor images, and in some cells, a subregion corresponding to Ca²⁺_cyt spike initiation was selected for analysis (Fig. 4B). Representative time courses for MLO-Y4 and MC3T3-E1 cells under control (untreated), L-Type inhibitor treated, and T-Type inhibitor treated conditions is shown in Fig. 4C. In osteocytes, elevations of Ca²⁺_cyt coincided with depression of Ca²⁺_ER, with subsequent peaks occurring after recovery of Ca²⁺_ER levels (n=10 cells). In osteoblasts, significantly fewer Ca²⁺_cyt responses were observed, and while the ER contributed to these responses, a depression in Ca²⁺_ER was often not recovered in the course of the experiment (n=11 cells). Treatment of either cell type with the L-Type VSCC inhibitor nifedipine had no detectable effect on the number of Ca²⁺_cyt peaks (n=8 osteocytes, 5 osteoblasts), and the T-Type VSCC inhibitor NNC 55-0396 had no detectable effect on responses in osteoblasts (n=6 cells). Treatment of MLO-Y4 cells with the T-Type VSCC inhibitor significantly reduced the number of Ca²⁺_cyt responses (Fig. 4D, n=9 cells).

Figure 4. The effects of VSCC inhibitor treatment on Ca²⁺_cyt responses and Ca²⁺_cyt/Ca²⁺_ER synchrony in MLO-Y4 osteocytes and MC3T3-E1 osteoblasts.

To simultaneously monitor Ca²⁺ in the cytosol and the ER, cells were transfected with D1ER and stained with the red-shifted fluorescent Ca²⁺_cyt indicator Fura Red-AM. Single cells at 60× were imaged using a quadview beamsplitter with single excitation of both indicators to achieve simultaneous, real-time measurements of Ca²⁺_cyt and Ca²⁺_ER. A representative image is shown in (A). The FRET ratio was calculated from the FRET and donor images, and (B) shows the selection of a subregion to determine the changes in Ca²⁺_ER microdomains corresponding to Ca²⁺_cyt spike initiation. (C) Representative time courses from MLO-Y4 and MC3T3-E1 cells imaged at 60× magnification under fluid flow exposure in the presence or absence of L- and T-Type VSCC inhibitors. Flow was started after the collection of 60s baseline. Detected peaks and valleys are indicated by “+”. (D) The number of Ca²⁺_cyt responses in untreated, L-Type inhibitor treated, and T-Type inhibitor treated bone cells. (E) The percentage of synchronous responses in MLO-Y4 and MC3T3-E1 treated with VSCC inhibitors. The percentage of synchronous responses was defined as the number of coincident Ca²⁺_cyt and Ca²⁺_ER transients divided by the total number of transients (peaks and inverse peaks). Error bars are standard deviation. Significance was assessed by multi-factorial ANOVA with Bonferroni’s post hoc analysis. *p<0.05 compared to control.

The percentage of synchronous responses was defined as the number of coincident Ca²⁺_cyt and Ca²⁺_ER transients divided by the total number of transients (peaks and inverse peaks). The synchronous response percentage was high in control osteocytes, with more than 80% of Ca²⁺_cyt peaks corresponding to dynamic changes in Ca²⁺_ER (Fig. 4E). MC3T3-E1 cells exhibited few Ca²⁺_cyt peaks in response to flow (Fig. 4D), and more often than not those peaks were not synchronous with a transient drop in Ca²⁺_ER (Fig. 4E). Inhibition of T-Type VSCC significantly reduced the synchrony in osteocytes to similar levels observed in osteoblasts, with an average of ~20%.

3.5 Effects of SOCE inhibitors on osteocyte Ca²⁺_cyt responses

MLO-Y4 cells were subjected to fluid flow stimulation in the presence of three different SOCE-related inhibitors to inform the mechanism of Ca²⁺ exchange with the ER. The number of Ca²⁺_cyt peaks (Fig. 5A) and the synchronous response percentage (Fig. 5B) were determined for each group. YM58483 had no effect on Ca²⁺_cyt responses in osteocytes (n=3 cells), and though SKF-96365 reduced the number of Ca²⁺_cyt responses, the effect was not significant (n=4 cells). This trend was consistent with population studies (YM58483: n=223 cells, 7 slides; SKF-96365: n=231 cells, 4 slides; data not shown). Neither treatment influenced synchrony. Treatment with 2-APB significantly reduced the number of Ca²⁺_cyt responses and the percent of synchronous responses (n=5 cells). Again, the effect on Ca²⁺_cyt responses was consistent with Fluo-8 experiments on larger cell populations (n=84 cells, 5 slides; data not shown). Representative curves of Ca²⁺_cyt and Ca²⁺_ER in MLO-Y4 cells treated with these inhibitors showed high synchrony in YM58483 and SKF-96365 treated cells (Fig. 5C,D), whereas 2-APB treatment resulted in no recovery of ER levels following a Ca²⁺_cyt transient (Fig. 5E).

Figure 5. Pharmacologic inhibition of SOCE in MLO-Y4 osteocytes.

(A) Number of Ca²⁺_cyt peaks in control and inhibitor treated groups. (B) The percentage of synchronous responses in MLO-Y4 treated with SOCE inhibitors. Simultaneous measurement of Ca²⁺_cyt and Ca²⁺_ER in MLO-Y4 cells exposed to fluid shear pre-treated with (C) YM58483, (D) SKF-96365, and (E) 2-APB. Flow was started after the collection of 60s baseline. Detected peaks and valleys are indicated by “+”. Significance was assessed by one-way ANOVA. Error bars are standard deviations. *p<0.05, ***p<0.001 compared to control.

4. Discussion

The purpose of this study was to uncover mechanisms underlying the generation of Ca²⁺_cyt oscillations in osteocytes. Previous studies have implicated the ER as an important source of Ca²⁺ for transient Ca²⁺ signals in osteocytes responding to mechanical stimuli, however to date no study has precisely identified the mechanisms by which the ER enables Ca²⁺_cyt oscillations in osteocytes. This study represents the first attempt to visualize the dynamics of Ca²⁺ specifically in the ER in osteocytes. Using a FRET biosensor localized to the ER, we were able to simultaneously observe Ca²⁺ exchange between the ER and the cytosol. We found high synchrony between rises in Ca²⁺_cyt and depressions in Ca²⁺_ER, with subsequent Ca²⁺_cyt responses occurring after a recovery of Ca²⁺_ER levels, suggesting that ER dynamics dictate osteocyte Ca²⁺ responses to fluid flow. Additionally, a major goal of this study was to link a phenotypic difference between osteocytes and osteoblasts in the expression of VSCC to the mechanobiology of these cell types. Indeed, we found that challenging T-Type VSCC alters the kinetics of ER depletion in osteocytes and that the predominant expression of T-Type VSCC in osteocytes sustains Ca²⁺_cyt oscillations by facilitating Ca²⁺_cyt/Ca²⁺_ER synchrony. Using inhibitors of various SOCE-related pathways, we determined that this exchange of Ca²⁺ between the cytosol and the ER is likely not a result of STIM proteins or transport of Ca²⁺ through CRACs, but rather through an interaction of the ER with T-Type channels on the osteocyte cell membrane.

Previous studies detected little to no L-Type VSCC expression in MLO-Y4 osteocytes, but found detectable levels of T-Type VSCC¹¹. Furthermore, a pattern of development suggested a loss of L-Type VSCC with the differentiation of osteoblasts to osteocytes. A later study confirmed this finding and additionally reported that T-Type channels mediated mechanically-induced ATP release in osteocytes¹². Most recently, a study in IDG-SW3 osteocytes found that parathyroid hormone treatment increased the expression of L-Type VSCC and decreased T-Type VSCC in osteocyte-stage cells to revert them to a less mature phenotype³¹. Our results are consistent with the role of VSCC in the differentiation of osteoblasts to osteocytes and additionally suggest the predominant expression of T-Type VSCC in osteocytes is important for their distinct role as mechanosensors.

As previous studies resulted in similar effects of VSCC and ER antagonists on osteocyte Ca²⁺_cyt responses, we sought to determine whether there was any interaction between T-Type VSCC and the ER in osteocytes. The time to depletion was very similar (within 3%) in untreated MLO-Y4 and MC3T3-E1, suggesting that ER release kinetics alone cannot account for the Ca²⁺_cyt oscillation behavior in osteocytes. Treatment of MLO-Y4 with a T-Type VSCC inhibitor prior to thapsigargin resulted in a significantly faster time to depletion compared to untreated controls. This suggests that the presence of T-Type VSCC may serve to refill ER stores in osteocytes, which may explain the ability of osteocytes to sustain Ca²⁺_cyt oscillations.

Though inhibitor studies challenging the ER allude to a mechanism involving Ca²⁺_ER release and refilling in sustaining Ca²⁺_cyt oscillations in osteocytes, these conclusions are speculative without the direct measurement of Ca²⁺_ER dynamics under flow. This is also true for measurements of ER depletion, where only a single induction of Ca²⁺_ER release is monitored. In this study, by simultaneously monitoring Ca²⁺ separately in the cytosol and ER in osteocytes, we found that most rises in Ca²⁺_cyt were accompanied by a corresponding decrease and recovery in Ca²⁺_ER. This supports the hypothesis that osteocytes are capable of generating Ca²⁺_cyt oscillations by an ability to refill the ER stores. The effects of T-Type VSCC inhibition implicate T-Type channels in this process.

Though the T-Type inhibitor had a large suppressive effect on Ca²⁺_cyt/Ca²⁺_ER synchrony in osteocytes, there are other possible mechanisms linking the exchange of Ca²⁺ between these two spaces. However, the lack of a pronounced suppression between control cells and those treated with YM58483 on the number of Ca²⁺_cyt responses and the percentage of synchronous responses indicates that CRACs are not involved. While SKF-96365 treatment mildly decreased the number of responses, this reduction is likely due to reported off-target effects on VSCC²³. The only inhibitor that produced similar effects as the T-Type antagonist was the IP₃R inhibitor 2-APB, which blocks IP₃-mediated release of Ca²⁺ from the ER. Thus, the exchange of Ca²⁺ between the cytosol and the ER is likely not a result of Ca²⁺ transport through CRAC or STIM1 involvement, but rather through the IP₃-mediated release of Ca²⁺_ER. Taken together, this data suggests that T-Type VSCC may directly interact with the ER in osteocytes, with the ER possibly relying on T-Type VSCC to replenish Ca²⁺ following Ca²⁺ release into the cytosol. Indeed, in cerebral arteries, the association of T-Type Ca_V3.2 with Ca²⁺_ER transporters drove RyR-mediated Ca²⁺ sparks¹³. Thus, it is possible a similar interaction exists between T-Type VSCC and ER-mediated Ca²⁺ oscillations in osteocytes.

There are a few limitations to the current study. First, the small sample sizes, though consistent with our experiments on large cell populations, mean our study was only designed to detect relatively large effects of the investigated treatments. Future studies looking for more subtle differences will require a larger sample size. In addition, our conclusions are drawn from experiments using small molecule inhibitors, which often have incomplete or off-target effects. Thus, future studies will be aimed at genetic modifications of cells to alter the expression of VSCC to investigate their role in osteocyte Ca²⁺ signaling. Furthermore, the simultaneous observation of Ca²⁺_ER levels in cells challenged with VSCC inhibitors only alludes to their interactions; true association of these structures will need to be verified by a proximity ligation or pull-down assay. Finally, the average synchrony in osteoblasts was also approximately 20%, though standard deviations were somewhat high due to the limited number of responses. As all T-Type treated MC3T3-E1 cells analyzed in this study had only one response with no coincident inverse response from the ER, the synchrony in the T-Type treated osteoblasts was zero, though this value was still not statistically different from MC3T3-E1 untreated controls. Although the present study indicates that the first Ca²⁺_cyt response in both cells can have some contribution of the ER, it is known that extracellular Ca²⁺ is also required, thus this effect is likely due to differences between the first and subsequent Ca²⁺_cyt transients in these cells. Future work will aim to explore these differences.

5. Conclusions

The sustainment of robust oscillations in Ca²⁺_cyt under mechanical loading is a hallmark of osteocytes both in vitro and ex vivo, thus the primary goal of this study was to uncover the mechanism by which osteocytes rely on specific subcellular stores of Ca²⁺ to generate these multiple responses. By observing Ca²⁺ dynamics in both cytosolic and ER spaces, this study demonstrated that osteocytes generate multiple responses through an ability to refill ER stores. Our data suggests that T-type VSCC facilitate the recovery of Ca²⁺_ER in osteocytes to permit this behavior. We uncovered a new mechanism involving T-type VSCC underlying the unique behavior of osteocytes as mechanosensors.

Highlights.

• Ca²⁺_cyt and Ca²⁺_ER are observed simultaneously in single bone cells under flow.

• Ca²⁺ exchange between the ER and the cytosol is highly synchronized in osteocytes.

• Antagonizing T-Type VSCC in osteocytes alters Ca²⁺_ER dynamics and synchrony.

Abbreviations

ER

Endoplasmic reticulum

VSCC

voltage-sensitive calcium channels

Ca²⁺

calcium

SOCE

store-operated calcium entry

IP₃

inositol trisphosphate