Phosphatidylinositol 3,5-Bisphosphate (PI(3,5)P2) Potentiates Cardiac Contractility via Activation of the Ryanodine Receptor*

Phosphatidylinositol 3,5-bisphosphate (PI(3,5)P2) is the most recently identified phosphoinositide, and its functions have yet to be fully elucidated. Recently, members of our muscle group have shown that PI(3,5)P2 plays an important role in skeletal muscle function by altering Ca2+ homeostasis. Therefore, we hypothesized that PI(3,5)P2 may also modulate cardiac muscle contractility by altering intracellular Ca2+ ([Ca2+]i) in cardiac myocytes. We first confirmed that PI(3,5)P2 was present and increased by insulin treatment of cardiomyocytes via immunohistochemistry. To examine the acute effects of PI(3,5)P2 treatment, electrically paced left ventricular muscle strips were incubated with PI(3,5)P2. Treatment with PI(3,5)P2 increased the magnitude of isometric force, the rate of force development, and the area associated with the contractile waveforms. These enhanced contractile responses were also observed in MIP/Mtmr14−/− mouse hearts, which we found to have elevated levels of PI(3,5)P2. In cardiac myocytes loaded with fura-2, PI(3,5)P2 produced a robust elevation in [Ca2+]i. The PI(3,5)P2-induced elevation of [Ca2+]i was not present in conditions free of extracellular Ca2+ and was completely blocked by ryanodine. We investigated whether the phosphoinositide acted directly with the Ca2+ release channels of the sarcoplasmic reticulum (ryanodine receptors; RyR2). PI(3,5)P2 increased [3H]ryanodine binding and increased the open probability (Po) of single RyR2 channels reconstituted in lipid bilayers. This strongly suggests that the phosphoinositide binds directly to the RyR2 channel. Thus, we provide inaugural evidence that PI(3,5)P2 is a powerful activator of sarcoplasmic reticulum Ca2+ release and thereby modulates cardiac contractility.

Phosphatidylinositols are negatively charged amphiphilic phospholipids that are integral members of the lipid bilayer (1). These phosphatidylinositols are concentrated on the cytosolic and sarcoplasmic reticulum (SR) 2 surfaces (2, 3) of cell membranes and are differentially phosphorylated at the 3, 4, and 5 positions of the inositol ring, thereby allowing for the formation of seven distinct phospholipids called phosphoinositides (PIs) (1,4). Phosphoinositides are uniquely able to undergo rapid and reversible phosphorylation, which is an attribute that makes them ideal membrane signaling proteins. The broadness of cellular functions controlled and modulated by PIs seems to be growing. The significance of PIs is underscored by the growing number of diseases associated with altered PI metabolism such as X-linked myotubular myopathy, Charcot-Marie-Tooth disease, diabetes, and others (5).
The most recently identified PI, phosphatidylinositol 3,5bisphosphate (PI (3,5)P2), which is regulated by the enzyme muscle-specific inositol phosphatase/myotubularin-related protein 14 (MIP/Mtmr14), is essential for maintaining Ca 2ϩ homeostasis in skeletal muscle (3). Interestingly, MIP Ϫ/Ϫ mice have elevated levels of PI (3,5)P2 in skeletal muscle and show significant impairment of skeletal muscle function including: reduced tolerance to treadmill running, increased exercise-induced fatigability, and premature muscle atrophy (3). The reduced skeletal muscle function of MIP Ϫ/Ϫ mice was attributed to the chronic leakage of Ca 2ϩ from the sarcoplasmic reticulum due to direct binding of PI (3,5)P2 to the skeletal muscle ryanodine receptor (RyR1). These data strongly suggest that chronic PI (3,5)P2 dysregulation may be associated with dysfunctional Ca 2ϩ handling and impaired contractile performance. However, the role of PI (3,5)P2 in cardiac muscle and its potential contribution to cardiovascular disease are currently unknown.
The regulatory processes that allow for proper functioning of the excitation-contraction coupling (ECC) process in cardiac muscle are highly complex and differ from skeletal muscle in many aspects. In cardiac muscle, ECC regulation is complicated by the fact that Ca 2ϩ entry across the sarcolemma via voltage-gated Ca 2ϩ channels directly triggers Ca 2ϩ release from the SR Ca 2ϩ release channels (ryanodine recep-tors) in a process known as Ca 2ϩ -induced Ca 2ϩ release (CICR). It is precisely this augmentation in Ca 2ϩ release from the cardiac ryanodine receptors (RyR2) that allows for adjustments in cardiac contractility. Despite the significant progress made in the last decade, there is still not a thorough understanding of all the mechanisms that regulate intracellular Ca 2ϩ ([Ca 2ϩ ] i ) and subsequent muscle contraction. We hypothesized that PI (3,5)P2 could serve as a novel mechanism to regulate Ca 2ϩ and cardiac muscle contractility. Importantly, it has been shown that PI (3,5)P2 levels can be acutely increased in the cell by hormonal control (6) and environmental stress (7,8). Few data are available on what effects this increase in PI (3,5)P2 has on cardiomyocytes. It is known that the MIP enzyme is highly expressed in cardiac tissue (3); however, the presence of PI (3,5)P2 in cardiac muscle has remained unreported. Based on the available literature, we wanted to determine whether PI (3,5)P2 was present in cardiomyocytes and explore its role in altering cardiac function. We hypothesized that exogenous PI (3,5)P2 would increase cardiac contractility via direct actions on the cardiac ryanodine receptor (RyR2), which could increase cytosolic Ca 2ϩ release. Moreover, we sought to address the physiological role of elevated levels of endogenous PI(3,5)P2 by using MIP Ϫ/Ϫ mice. We hypothesized that MIP Ϫ/Ϫ animals would also demonstrate altered cardiac function.

EXPERIMENTAL PROCEDURES
Materials-The PIs, the carrier protein histone H1, and the PI(3,5)P2 antibody were purchased from Echelon Biosciences (Salt Lake City, UT). PBS, Hanks' balanced salt solution, Triton X-100, and fura-2 were obtained from Invitrogen. The secondary antibody light goat anti-mouse IgG2b and the AffiniPure Fab fragment goat-anti mouse IgG (HϩL) (used in the blocking buffer) were obtained from Jackson ImmunoResearch (West Grove, PA). The adult rat/mouse cardiomyocyte isolation kit (ac-7018) was acquired from Cellutron Life Technology (Baltimore, MD). 4, 6-Diamidino-2-phenylindole (DAPI) was purchased from Sigma. Ryanodine was obtained from Ascent Scientific (Princeton, NJ). All remaining reagents were purchased from Fisher Scientific.
Statistical Analysis-All graphs were made and statistical procedures were performed using GraphPad Prism 5.0. Data are presented as means Ϯ S.E. Data were compared using either a paired t test or a one-way analysis of variance, and significance was set at the p Ͻ 0.05 level. When necessary, the one-way analysis of variance was followed up by appropriate post hoc testing. Specific details of the statistical tests performed are stated in the corresponding sections below.
Experimental Animals-Sixteen-week-and 1-year-old wild type male C57/BL6 mice (Jackson Laboratories) and 1-yearold MIP Ϫ/Ϫ mice (as described previously) (3) were used in this study. All mice were housed in a temperature-controlled (22 Ϯ 2°C) room with a 12-h:12-h light/dark cycle. Animals were fed ad libitum. All protocols were approved by the Animal Care and Use Committee of the University Missouri-Kansas City School of Medicine. Prior to use, mice were treated with heparin (5000 units/kg of body weight) by intraperitoneal injection 30 min before cervical dislocation.
Phosphoinositides-Aqueous stocks (1 mM) of long chain di-C16 fatty acid PIs were prepared in Ringer's solution with Ca 2ϩ (in mM: 142 NaCl, 5.0 KCl, 2.5 CaCl 2 , 1.8 MgCl 2 , 5 HEPES, 10 glucose, pH 7.4) or without Ca 2ϩ (in mM: 142 NaCl, 5.0 KCl, 0.1 EGTA, 2.0 MgCl 2 , 5 HEPES, 10 glucose, pH 7.4). Because the PIs used are hydrophobic, they were mixed by heating and sonication for 30 min prior to being at stored at Ϫ20°C. These PI stocks were thawed, sonicated, and then vortexed in a (1.0/0.8 ratio) with a carrier solution containing histone H1 to aid in delivering the PIs into the cell. Vehicle experiments were performed using only the histone carrier. The final concentrations of PIs (0.5 M) and histone carrier (0.4 M) for contractility and cell culture experiments were determined following preliminary studies and were based off of previous publications (3,9,10). Lastly cardiomyocytes were exposed to PI(3,5)P2 (0.5 M) for 4 h and then underwent a trypan blue dye exclusion assay. No toxic effect was observed to the cardiomyocytes as a result of PI(3,5)P2 or the histone carrier.
Culture of HL-1 Cardiomyocytes-The HL-1 cell line was a kind gift from Dr. W. C. Claycomb (Louisiana State University Health Science Center, New Orleans, LA). HL-1 cells are currently the only cardiomyocyte cell line available that will continuously divide and spontaneously contract while maintaining a differentiated cardiac phenotype (supplemental Fig.  S1). Extensive characterization of HL-1 cells via microscopy, genetics, electrophysiological, and pharmacological techniques has demonstrated that HL-1 cells behave similarly to mouse primary cardiomyocytes (11), which are difficult to obtain and maintain in prolonged culture. Cells were cultured in T-25 flasks precoated with a 0.00125% fibronectin and 0.02% gelatin solution and maintained in Claycomb medium (JRH Biosciences Ltd.) supplemented with 10% fetal bovine serum, 2 mM L-glutamine, 0.1 mM norepinephrine, 0.3 mM ascorbic acid, and 100 units/ml penicillin, 100 mg/ml streptomycin (full media), at 37°C in a humid atmosphere of 5% CO 2 , 95% air. Media were changed every 24 -48 h. For experiments, cells were seeded at 3500 cells/35-mm dish and allowed to recover for 24 h prior to experimentation. HL-1 cells (supplemental Fig. S1) were tested with PI(3,5)P2 and histone carrier, in the presence of extracellular Ca 2ϩ (1.8 mM), in the absence of extracellular Ca 2ϩ (0.0 mM), and with ryanodine (1 m) in the presence of Ca 2ϩ . This dose of ryanodine was chosen as it inhibited the response of the SR to caffeine (10 mM).
Isolation of Primary Cardiac Myocytes-To confirm the results demonstrated in our stable HL-1 cell line, we repeated essential experiments in adult primary cardiac myocytes. Primary cardiac myocytes were isolated using the non-perfusion adult rat/mouse cardiomyocyte isolation kit, which yields ϳ20% healthy, viable myocytes in our laboratory (supplemental Fig. S1). Briefly, the tissue was cut vertically into small pieces with a pair of scissors. The pieces were then transferred into a 50-ml tube and digested at 37°C on a shaker at the speed of 85-100 rpm. Repeated cell suspensions were collected (4 -6 times), and cells were centrifuged (250 ϫ g) for 2 min, rinsed in wash buffer, and transferred to laminin-treated plates. The isolated cardiac myocytes underwent a 2-h incubation at 37°C to allow the cells to attach to the plate. Pri-mary cells were tested with PI(3,5)P2 and histone carrier, in the presence of Ca 2ϩ (1.8 mM), and with ryanodine (0.1 m). This dose of ryanodine was chosen as it inhibited the response of the SR to caffeine (10 mM) in primary myocytes.
Immunochemistry-Following cervical dislocation, whole hearts were quickly excised, rinsed in PBS, and frozen until use. Hearts were later embedded in tissue freezing medium and sectioned at 8 m using a cryostat. Immunohistochemistry was then performed as described previously (12). Briefly, sections were mounted to a slide and fixed in 4% formaldehyde for 10 min at room temperature. Slides were washed with 0.5% Triton X-100 in TBS followed by blocking with 10% normal goat serum. Sections were incubated overnight at 4°C with either anti-PI(3,5)P 2 mouse monoclonal IgG2b antibody (1:500) in blocking buffer. On the second day, sections were allowed to incubate at room temperature for 1 h followed by three washes in TBS for 5 min each. This was followed by incubation with a fluorescently tagged DyLight goat anti-mouse IgG2b specific secondary antibody (1:5000) in blocking buffer for 1 h at room temperature. Following this, sections were incubated with DAPI, 1 g/ml in blocking buffer for 5 min. Prior to being coverslipped, sections were washed three times in TBS. Sections were dried and coverslipped with mounting medium. Negative control sections (non-immune) were performed by incubating with antibody diluent and DAPI stain and supplementing with a non-immune immunoglobulin of the same isotype as the primary antibody (supplemental Fig. S2).
To test for changes in the levels of endogenous PI(3,5)P2 HL-1 cells were grown on coverslips and serum-starved overnight in Claycomb minimal media (0.5% fetal bovine serum, 2 mM L-glutamine, and 100 units/ml penicillin, 100 mg/ml streptomycin). Cells were then rinsed with TBS and transferred into serum-free DMEM (2 mM L-glutamine, 100 units/ml penicillin, and 100 mg/ml streptomycin) for 4 h. The switch from Claycomb media to DMEM was essential as the Claycomb media contains insulin and other growth factors (11). After 4 h, the cells were given fresh serum-free DMEM and treated with insulin (1 milliunit/ml) for 10 min. Immunocytochemistry was then performed as described above.
Fluorescent images were captured with an Olympus IX51 inverted microscope (Center Valley, PA), Hamamatsu Orca-ERGA CCD camera (Bridgewater, NJ), Semrock BrightLine filter set (Rochester, NY), and X-cite 120 metal halide light source (EXFO, Mississauga, CA). Data were captured and analyzed with the SlideBook ratiometric software (Intelligent Imaging Innovations Inc., Denver, CO). Differences in the relative immunofluorescence between groups was detected using a paired t test. Experiments were repeated three times to confirm results.
Cardiac Contractility Measurements-The mouse hearts used for the muscle strip experiments were quickly excised and placed into an ice-cold cardioprotective Ringer's solution (with Ca 2ϩ ) that included the addition of 2,3-butanedione monoxime (30 mM) and insulin (10 IU/liter) for 30 min, bubbled under 100% O 2 as described previously (13). Briefly, left ventricular muscle strips were prepared (1-2 mm wide by 6 -8 mm long) in the cardioprotective solution. The strips were tied on the proximal and distal ends with a silk thread. The muscle strips were then rinsed three times (5 min each) in Ringer's (with Ca 2ϩ , pH, 7.4) to remove the 2,3-butanedione monoxime. Cutting the ventricles into strips exposes a greater surface area of the heart to the PIs and was essential for PI delivery.
In MIP Ϫ/Ϫ contractility experiments, we utilized intact ventricular tissue. Hearts were isolated and placed into icecold Ringers solution (with Ca 2ϩ ), and atria were removed. Hearts were attached to silk thread using small metallic clips.
The muscle strips and intact hearts were hung vertically and attached to force transducer, between bipolar platinum stimulating electrodes suspended in 25-ml glass tissue chambers, all of which were obtained from AD Instruments (Colorado Springs, CO). Heart muscles were stretched to the length of maximum force development in Ringer's solution (pH, 7.4, without 2,3-butanedione monoxime) and stimulated with pulses of 1 Hz for 5 ms (Grass Technologies stimulator SD9; Quincy, MA) the voltage was set 20% above threshold. The muscles in the chamber were superfused with Ringer's solution (with Ca 2ϩ , pH, 7.4) continuously bubbled with 100% O 2 at room temperature. Muscles were allowed to stabilize for 90 min prior to experimentation and provided with fresh media changes every 30 min. Muscles were paced at 1 Hz to obtain a stable baseline and were drug-treated with either vehicle (histone) or PIPs. The contractile data were recorded and analyzed on the LabChart 6 software; AD Instruments (Colorado Springs, CO). Waveform changes were analyzed in the segments corresponding to peak isometric tension (millinewtons). Slope (millinewtons/s) was analyzed by taking the average slope from 10 to 20 ms after the start of the peak. Area (millinewtons ϫ s) was calculated using the region from 10 to 90% of the peak. (s) was fitted at the baseline using data from 95 to 0% of peak. For exogenous PI experiments, three heart strips were tested in two different PIs and compared with a histone only vehicle control. Each group therefore represents muscle strips from 5 to 6 different animals. Strip experiments were normalized within each condition to baseline levels of contractility and presented as a relative change from baseline contraction data. Differences for isometric tension were analyzed by a one-way analysis of variance followed with a Newman-Keuls multiple comparison post hoc test. Differences for slope, area, and were analyzed via a paired t test.
The intact ventricular experiments of MIP Ϫ/Ϫ animals were compared with age-matched WT controls. The tension data for the intact experiments were taken during the maximal contraction evoked by electrical stimulation at 1 Hz. The isometric tension was normalized to heart weight and is presented as relative force. Differences for isometric tension, slope, area, and were analyzed by a paired t test.
Ca 2ϩ Imaging-Imaging was conducted as reported previously (14). Briefly, all cells were loaded with the Ca 2ϩ indicator fura-2 (2 M) for 30 min at 37°C and allowed to de-esterify for 10 min at 37°C. Plates were washed twice with Hanks' balanced salt solution, and intracellular Ca 2ϩ levels were measured at 37°C using an environmental chamber. Data were captured and analyzed with the SlideBook ratiometric software. Cell culture dishes were pretreated with ryanodine for 10 min prior to treatment with PI(3,5)P2. Primary cells were treated with KCl (80 mM) to determine viability at the end of each test. All diluted drugs were added by dilute bolus injection on the plates. Cells for imaging were selected on the distal side of the plate to allow for drug dilution and diffusion. Nevertheless, because cells may not be equal distances from the drug delivery site, this may contribute to some variability in cell responses. In these experiments, the three treatment conditions (vehicle, PI(3,5)P2 alone, or PI(3,5)P2 ϩ ryanodine) were tested 2-3 times each from isolated cardiac myocytes from a given animal (6 -9 cells). The experiments were repeated in five animals, and therefore, a total of 10 -15 primary cardiac myocytes were tested for each treatment condition. All data are presented at the peak increase in fluorescence. The cells tested in each condition were averaged and then used for data analysis. Data were analyzed by a one-way analysis of variance followed by Newman-Keuls multiple comparison post hoc test.
[ 3 H]Ryanodine Binding Assay-[ 3 H]Ryanodine binding to purified pig cardiac muscle SR was performed as described previously (15,16). The standard incubation medium (0.2 M KCl, 40 mM Na-HEPES at pH 7.2, 7 nM [ 3 H]ryanodine, 1 mM EGTA) contained CaCl 2 in amounts necessary to set free [Ca 2ϩ ] between 10 nM and 10 mM. Samples (0.1 ml) were run in duplicate and incubated at 37°C for 90 min. The Ca 2ϩ / EGTA ratio was calculated by using the online software program MaxChelator. Cardiac muscle SR-enriched microsomes (60 g) were added to the incubation medium together with 10 M PI(3,5)P2. Nonspecific [ 3 H]ryanodine binding was determined in the presence of unlabeled ryanodine (10 M), amounted typically to no more than 10 -15%, and was subtracted from all reported values.
Planar Lipid Bilayer Technique-SR-enriched microsomes from pig heart were reconstituted into Muller-Rudin planar lipid bilayers as described previously (15,16). Control and PI(3,5)P2 were run using the same vesicles; therefore, the density of RyR2 was the same in each condition. Single RyR2 channel data were collected at steady voltages (ϩ35 mV, cis chamber grounded) for 2 min in symmetrical cesium methane-sulfonate (300 mM) and Na-HEPES (10 mM at pH 7.2). The recording solution contained ϳ5 M free Ca 2ϩ as assessed by a calibration curve, which was sufficient to activate the RyR2 channel. EGTA (1 mM), and a calibrated concentration of CaCl 2 was then added to the cis (cytosolic) side of the channel to reach a free [Ca 2ϩ ] in the range of pCa 7 to pCa 4. After recording RyR2 channel activity in the absence of PI(3,5)P2, the PI or vehicle was added to the cytosolic side of the channel. For each condition, single channel data were collected at steady voltages for 2-5 min. The bilayer was "rested" at 0 mV to improve stability while adding reagents or perfusing the cis chamber. Channel activity was recorded with a 16bit VCR-based acquisition and storage system at a 10-kHz sampling rate. Signals were analyzed after filtering with an 8-pole Bessel filter at a sampling frequency of 1.
Ca 2ϩ Imaging of HL-1 Cardiomyocytes-The contractile data indicated that PI(3,5)P2 acts in cardiac muscle by altering intercellular Ca 2ϩ homeostasis. To test whether the source of Ca 2ϩ was from an intracellular or extracellular source, we tested the response of HL-1 cells loaded with the Ca 2ϩ indicator fura-2 to PI(3,5)P2 in the absence and presence of Ca 2ϩ (1.8 mM). Fig. 4A shows a raw trace of the ratiometric change in [Ca 2ϩ ] i measured in HL-1 cardiac myocytes before and after exposure to PI(3,5)P2. Fig. 4B shows the average ratiometric change in intracellular Ca 2ϩ at baseline and following exposure to PI(3,5)P2. Treatment of HL-1 cells tested in extracellular Ca 2ϩ with PI(3,5)P2 elevated the peak cytosolic Ca 2ϩ (1.68 Ϯ 0.33) over vehicle (0.64 Ϯ 0.02). There was no difference in the peak ratiometric response to PI (3,5)P2 between tests performed in the presence (1.68 Ϯ 0.33) or absence (1.61 Ϯ 0.32) of extracellular Ca 2ϩ . To test the involvement of the cardiac RyR2 channel, we pretreated HL-1 cells with ryanodine (Fig. 4B). Following pretreatment with ryanodine, the myocytes were treated with PI(3,5)P2, and the response was significantly reduced. Blockade of the SR by ryanodine inhibited the effect of PI(3,5)P2 (0.65 Ϯ 0.06) and showed no statistical difference when compared with vehicle (0.64 Ϯ 0.02). The baseline data represent the average ratiometric value prior to any treatment. Vehicle treatment was

. PI(3,5)P2 acutely increases contractile force in ex vivo ventricular cardiac muscle strips.
A contains representative tracings of left ventricular muscle contractions at baseline and following treatment with either PI(3,5)P2 or vehicle. B summarizes isometric tension data and shows that PI(3,5)P2 increased tension, whereas PI(4,5)P2 and PI(3)P did not elicit a significant response when compared with vehicle. C shows that PI(3,5)P2 increased the slope and area, but not (n ϭ 4 -5 animals). * denotes statistical significance from vehicle.
included as a comparison with baseline and demonstrates that the histone carrier had no effect on [Ca 2ϩ ] i . The average time for a response to PI(3,5)P2 was 140 Ϯ 41 s for PI (3,5)P2.
Ca 2ϩ Imaging of Isolated Primary Cardiac Myocytes-Next, we wanted to confirm our finding that the effects of PI(3,5)P2 were mediated by the RyR2 channel in primary cardiomyocytes. Fig. 5A shows a raw trace of the ratiometric change in intracellular Ca 2ϩ measured in a primary ventricular cardiac myocyte before and after exposure to PI(3,5)P2. Treatment with PI(3,5)P2 elevated the Ca 2ϩ level, which typically remained elevated for several minutes, eventually causing contraction of the myocyte. To reaffirm that exposure to PI(3,5)P2 acts principally on the RyR2 channel, we pretreated primary cardiac myocytes with ryanodine. Fig. 5B shows that the normal caffeine-evoked ratiometric Ca 2ϩ response is effectively blocked following treatment of primary cardiac myocytes with ryanodine, indicating that the SR Ca 2ϩ pool was disabled. Following pretreatment with ryanodine, the primary cardiac myocytes were treated with PI(3,5)P2, and the response was significantly reduced. Lastly the primary cardiac myocytes were given KCl (80 mM) to ensure viability of the cell. Fig. 5C shows average ratiometric fura-2 responses at baseline, vehicle treatment, PI(3,5)P2 treatment, and pretreatment with ryanodine. Treatment of primary cardiac myocytes tested in extracellular Ca 2ϩ with PI(3,5)P2 elevated the cytosolic Ca 2ϩ (1.53 Ϯ 0.13) when compared with vehicle (0.40 Ϯ 0.06). Blockade of the SR with ryanodine significantly inhibited the effect of PI(3,5)P2 (0.51 Ϯ 0.05) and showed no statistical difference when compared with vehicle. The average time for a response to PI (3,5) (3,5)P2 sensitization of the RyR2 is specific as the effect was significantly inhibited by a PI(3,5)P2 antibody (supplemental Fig. S3) and is concentration-dependent, with maximal stimulation (B max ) ϭ 308 Ϯ 22% and halfmaximal effective dose (ED 50 ) ϭ 6.2 Ϯ 1.2 M (n ϭ 4 experiments) (Fig. 6B). To test whether [ 3 H]ryanodine binding results indeed correlated with RyR2 channel activity, we reconstituted cardiac SR vesicles in planar lipid bilayers and recorded RyR2 channel activity in the absence and presence of PI(3,5)P2. Several biophysical and pharmacological properties of the recorded channel indicated to us that the activity indeed corresponded to RyR2 channels: 1) the unitary conductance corresponded to ϳ700 picosiemens, which is typical for RyR2 channels; 2) Ca 2ϩ (100 nM-10 M) in the cytosolic side of the channel increased channel activity; chelation of Ca 2ϩ with EGTA decreased channel activity; 3) channels entered into a subconductance state (ϳ40% of full conductance) upon the addition of 100 nM ryanodine, which is a signature effect of the alkaloid on RyR channels. Under control conditions and at 100 nM free [Ca 2ϩ ] (pCa 7), RyR2 channel activity was low (open probability, P o , Յ 0.05), as expected, but increased dramatically upon the addition of PI (3,5)P2 to the cytosolic side of the channel (Fig. 7A). The activity increased by 471% with respect to control, with nP o (single channel P o multiplied by the number of observed channels) increasing from 0.03 Ϯ 0.02 in control to 0.18 Ϯ 0.06 after PI(3,5)P2 (n ϭ 6 experiments) (Fig. 7B). A dramatic increase in P o was detected immediately after the addition of PI (3,5)P2 to the cytosolic side of the channel as seen in the P o versus time plot (Fig. 7C). The effect of PI (3,5)P2 lasted more than the mean life of the bilayers, suggesting a ligand-receptor interaction.

DISCUSSION
This is the first study to examine the acute and chronic effects of PI(3,5)P2 in cardiac myocytes. The major findings of A shows a raw tracing of a ventricular muscle contraction from a MIP Ϫ/Ϫ mouse and an age-matched WT mouse. B summarizes isometric tension data and shows that MIP Ϫ/Ϫ mice had increased isometric tension when compared with WT age-matched controls. mN, millinewtons. C presents the averages showing that MIP Ϫ/Ϫ mice have an increased slope and area, but not (n ϭ 4). * denotes statistical significance from wild type age-matched controls. DECEMBER 17, 2010 • VOLUME 285 • NUMBER 51 this study are as follows. 1) PI(3,5)P2 is present and responsive to physiological stress (insulin) in cardiac myocytes; 2) cardiac contractile force, rate of force development, and area (integral) were all increased following exposure to exogenous PI(3,5)P2 and in MIP Ϫ/Ϫ mice; 3) PI(3,5)P2 binds and increases the open probability of RyR2 channels; and 4) PI(3,5)P2 causes [Ca 2ϩ ] i release in cardiac myocytes by acting on RyR2-sensitive Ca 2ϩ stores. These findings have led us to propose that PI(3,5)P2 induces increases in cardiac contractility by binding and opening RyR2, which increases the cytoplasmic Ca 2ϩ levels needed for the ECC process and stimulation of CICR.

PI(3,5)P2, Ca 2؉ Homeostasis, and Cardiac Contractility
PI (3,5)P2 is a newly identified PI, and its presence in various tissues has yet to be defined. Shen et al. (3) have previously shown that the inositol phosphatase, MIP, an enzyme that breaks down PI(3,5)P2, is highly present in both skeletal and cardiac tissue, suggesting that PI(3,5)P2 could potentially be present in the heart. Using immunodetection, we found PI(3,5)P2 expression in longitudinal cross-sections of left ventricular mouse heart. PI (3,5)P2 has previously been shown to be concentrated and localized in the SR of skeletal muscle (3), suggesting that PI(3,5)P2 is in close proximity to, and may modulate, the RyR channels in both cardiac and skeletal muscle.
To investigate the effects of PI(3,5)P2 on cardiac contractility, we compared contractile data in left ventricular muscle strips with those treated with PI(3,5)P2 and vehicle alone. PI(3,5)P2 significantly increased isometric tension, slope, and area of contraction. In addition to PI(3,5)P2, we also tested its isomer, PI(4,5)P2. Recent evidence has suggested that PI(4,5)P2 itself can directly alter ion channel permeability (18). Therefore, we were interested in determining whether this isomer could also induce changes in functional contractile responses. Moreover, we also treated muscle strips with the intermediary product in PI(3,5)P2 synthesis, PI(3)P.  8 mM). B shows the summary of fura-2 ratiometric changes in intracellular Ca 2ϩ response in HL-1 cardiomyocytes after exposure to PI(3,5)P2 with extracellular Ca 2ϩ (1.8 mM), without extracellular Ca 2ϩ (0 mM), and with extracellular Ca 2ϩ ryanodine pretreatment. Increases in intracellular Ca 2ϩ were not affected by extracellular Ca 2ϩ but were fully eliminated by the preaddition of ryanodine (n ϭ 10 -15 cells/condition). Images were captured with a 40ϫ objective. * denotes statistical significance from vehicle, and † denotes statistical significance from PI(3,5)P2.

FIGURE 5. Ryanodine inhibits the elevation of intracellular Ca 2؉ by PI(3,5)P2 in primary cardiac myocytes.
A shows fura-2 ratiometric changes in intracellular Ca 2ϩ in an isolated ventricular adult cardiac myocyte after treatment with PI(3,5)P2, ultimately resulting in contraction. B shows that ryanodine inhibited the release of SR Ca 2ϩ to both caffeine and PI (3,5)P2. C provides a summary of the data. Results in primary myocytes show that increases in intracellular Ca 2ϩ were fully eliminated by the preaddition of ryanodine (n ϭ 10 -15 cells/condition). Images were captured with a 40ϫ objective. * denotes statistical significance from vehicle, and † denotes statistical significance from PI(3,5)P2.
Treatment with PI(4,5)P2 and PI(3)P did not alter the characteristics of the contractile waveforms (force, slope, area, and ) when compared with vehicle. In support of our contractile data, Shen et al. (3) found no changes in [Ca 2ϩ ] i homeostasis in skeletal muscle fibers treated with PI(4,5)P2 and PI(3)P and concluded that PI (3,5)P2 was the principal mediator of functional changes in skeletal muscles from MIP Ϫ/Ϫ mice. In our MIP Ϫ/Ϫ mice, we noted the highly similar changes in cardiac contractility (significantly increased isometric tension, slope, and area of contraction) during maximal stimulation as to those seen with exogenous administration of PI (3,5)P2. These results show that endogenous PI (3,5)P2 has the ability to modulate cardiac contractility. From these contractile data, we hypothesized that the increased contractile force was indicative of an enhanced regulation of [Ca 2ϩ ] i .
During the cardiac ECC process, force is generated on a beat-to-beat basis by a 10-fold increase in cytosolic Ca 2ϩ . As a cardiac myocyte depolarizes, [Ca 2ϩ ] i begins to accumulate principally from the opening of L-type Ca 2ϩ channels, which triggers CICR from RyR2. In adult mouse myocardium, it is estimated that the RyR2 channel contributes upwards of 70% of the Ca 2ϩ ions utilized during contraction and, therefore, it is considered to be the largest determinant of cardiac contrac-tility (19). Because the majority of [Ca 2ϩ ] i accumulation during cardiac muscle contraction is from the SR, and PI(3,5)P2 binds and activates RyR1 (3), we hypothesized that PI (3,5)P2 would principally alter the SR Ca 2ϩ release from RyR2 in cardiomyocytes.
Our data show that elevated PI (3,5)P2 increased the height, slope, and area of contraction; therefore, it is likely that PI(3,5)P2 improves Ca 2ϩ entry either via opening voltagegated Ca 2ϩ channels or via directly altering Ca 2ϩ release from RyR2. Moreover, during relaxation, a return to Ca 2ϩ homeostasis is controlled principally by the sarcoplasmic reticulum Ca 2ϩ -ATPase (sarcoplasmic endoplasmic reticulum Ca 2ϩ -ATPase) and the Na ϩ /Ca 2ϩ exchanger, with a very minor contribution by the plasma membrane Ca-ATPase. If PI (3,5)P2 was increasing cytosolic Ca 2ϩ and the area of contraction by slowing Ca 2ϩ removal, there should be a corresponding increase in (the rate of relaxation following contraction). We found that elevated levels of PI (3,5)P2 had no effect on , demonstrating that it is very unlikely to have an effect on the activity of sarcoplasmic reticulum Ca 2ϩ -ATPase pumps, Na ϩ /Ca 2ϩ exchangers, and/or Ca 2ϩ ATPases.
To test the hypothesis that Ca 2ϩ was being released principally from the SR, we performed Ca 2ϩ imaging on HL-1 cells with or without extracellular Ca 2ϩ . HL-1 cells exposed to PI (3,5) ] (pCa 4 -pCa 2) inactivates the channel. PI(3,5)P2 did not change the bell-shaped form of this curve; however, the ascending limb of the curve was dramatically shifted upward at physiologic levels of Ca 2ϩ (pCa 7 and pCa 6). Therefore, PI(3,5)P2 "sensitizes" RyR2s at resting (100 nM) and contractile (1 M) levels of Ca 2ϩ . The "sensitizing" effects of PI(3,5)P2 were blocked in a concentration-dependent manner by the addition of a PI(3,5)P2 antibody.  4). B is the averaged doseresponse curve of PI(3,5)P2 stimulation of [ 3 H]ryanodine binding to cardiac SR. [Ca 2ϩ ] in the binding medium was "clamped" at pCa 6. Data points were fitted with a Hill function that yielded a maximal stimulation ϭ 308 Ϯ 22% of control (which was determined in the absence of PI(3,5)P2) and halfmaximal effective dose (ED 50 ) ϭ 6.2 Ϯ 1.2 M (n ϭ 4). Nonspecific binding (Ͻ10% of total binding) was determined with 10 M ryanodine and subtracted.
To further investigate the effect of PI(3,5)P2 on RyR2 channel activity, we measured the P o of single RyR2 channels reconstituted in lipid bilayers. The RyR2 channels were activated by PI(3,5)P2 in a concentration-dependent manner, indicating that PI(3,5)P2 directly binds and regulates RyR2. These findings are consistent with those reported for skeletal muscle RyR1 channels (3). Our data indicate that PI(3,5)P2 acts on RyR2 by promoting the release of SR Ca 2ϩ stores and by sensitizing the RyR2 to CICR.
Interestingly, although the mechanism of action of PI (3,5)P2 on RyR appears to be similar to skeletal muscle, the physiological consequence of this interaction seems to differ in cardiac muscle. We found MIP Ϫ/Ϫ -and PI(3,5)P2-treated ventricular muscles to have enhanced contractility. In contrast, Shen et al. (3) reported significant reductions in peak isometric tension and a substantially "prolonged relaxation profile" following maximal contractions of the fast twitch extensor digitorum longus muscles from MIP Ϫ/Ϫ . We believe our responses in cardiac muscle differ from skeletal muscle due to differences in the Ca 2ϩ -handling properties between skeletal and cardiac muscle. In skeletal muscle, the ECC process operates via a mechanical coupling mechanism via the dihydropyridine receptor (DHPR) to trigger RyR1, without extracellular Ca 2ϩ entry; therefore, PI(3,5)P2 likely depletes SR Ca 2ϩ stores while the muscle is at rest. In contrast, we hypothesize that the influx of extracellular Ca 2ϩ during cardiac muscle ECC helps maintain SR Ca 2ϩ stores. Therefore, in cardiac muscle, PI(3,5)P2 may enhance RyR2 sensitivity to Ca 2ϩ and promote greater SR Ca 2ϩ release on a beat-to-beat basis by CICR. Nevertheless, studies specifically designed to directly compare skeletal and cardiac muscle function following PI(3,5)P2 manipulation are warranted.
Further, it remains to be determined whether elevation of PI (3,5)P2 is protective during cardiac disease or promotes the progression of cardiac pathologies. For example, our observed positive inotropic effects of PI(3,5)P2 could arguably help to preserve muscle force during a condition such as heart failure. Alternatively, increased Ca 2ϩ leak from the SR may activate gene expression that contributes to heart failure. Moreover, it remains to be determined whether cardiac PI(3,5)P2 levels in vivo are altered by cardiac disease states or other stressors in the body. Importantly, there is evidence that PI(3,5)P2 levels can be acutely increased in the cell by exposure to environmental stressors such as hyperosmolarity (7) and UV light (8), endogenous agents such as interleukin-2 (8), and insulin (6). We were able to replicate these findings in HL-1 cardiomyo- cytes, using insulin to increase the level of PI(3,5)P2. It is possible that PI(3,5)P2 levels may be altered by acute stressors such as exercise, inflammation, disease, or infection, which could regulate cardiac function.
In summary, our study provides evidence that PI (3,5)P2 is a potent regulator of [Ca 2ϩ ] i homeostasis by directly activating and sensitizing the RyR2 channel. We believe that this action of PI(3,5)P2 on the RyR2 causes additional release of Ca 2ϩ during CICR, which can directly increase the strength of contraction in cardiac muscle.