Protein tyrosine phosphatase 1B is a mediator of cyclic ADP ribose-induced Ca2+ signaling in ventricular myocytes

Cyclic ADP-ribose (cADPR) releases Ca2+ from ryanodine receptor (RyR)-sensitive calcium pools in various cell types. In cardiac myocytes, the physiological levels of cADPR transiently increase the amplitude and frequency of Ca2+ (that is, a rapid increase and decrease of calcium within one second) during the cardiac action potential. In this study, we demonstrated that cADPR levels higher than physiological levels induce a slow and gradual increase in the resting intracellular Ca2+ ([Ca2+]i) level over 10 min by inhibiting the sarcoendoplasmic reticulum Ca2+ ATPase (SERCA). Higher cADPR levels mediate the tyrosine-dephosphorylation of α-actin by protein tyrosine phosphatase 1B (PTP1B) present in the endoplasmic reticulum. The tyrosine dephosphorylation of α-actin dissociates phospholamban, the key regulator of SERCA, from α-actin and results in SERCA inhibition. The disruption of the integrity of α-actin by cytochalasin B and the inhibition of α-actin tyrosine dephosphorylation by a PTP1B inhibitor block cADPR-mediated Ca2+ increase. Our results suggest that levels of cADPR that are relatively higher than normal physiological levels modify calcium homeostasis through the dephosphorylation of α-actin by PTB1B and the subsequent inhibition of SERCA in cardiac myocytes.


INTRODUCTION
Ca 2+ plays a fundamental role in the cardiac contraction and relaxation cycle by linking the electrical depolarization of cardiomyocytes with contraction (that is, excitationcontraction coupling; EC coupling). 1 Cellular depolarization after the action potential is generated from the sinoatrial node activates voltage-operated Ca 2+ channels, which causes an influx of Ca 2+ across the sarcolemma and into the cytoplasm. 1 The resulting Ca 2+ influx activates ryanodine receptors (RyRs) on the sarcoplasmic reticulum (SR), which causes more Ca 2+ to be released into the cytosol-this phenomenon is known as Ca 2+ -induced Ca 2+ release (CICR). [2][3][4] A transient increase in the free cytosolic calcium concentration ([Ca 2+ ] i ), Ca 2+ transient in cardiac myocytes allows the actin and myosin contractile filaments to engage and slide past each other, resulting in cardiac muscle contraction. 4 Sarcoendoplasmic reticulum Ca 2+ ATPase (SERCA) in sarcoplasmic reticulum (SR) in cardiac myocytes transfers Ca 2+ from the cytosol to the lumen of the SR as a result of ATP hydrolysis during muscle relaxation. 5 The transfer of Ca 2+ by SERCA from the cytosol to the SR is inhibited by unphosphorylated phospholamban (PLB). 6 PLB phosphorylation can relieve the inhibition of the SERCA pump and enhance [Ca 2+ ] i . 6 Cyclic ADP-ribose (cADPR) is synthesized from NAD+ by bifunctional ectoenzymes including CD38 and CD157, and monofunctional ADP ribosyl cyclase from the Aplysia mollusc. [7][8][9] In cardiac myocytes, nanomolar cADPR concentrations increase the amplitude and frequency of Ca 2+ transient through an increased accumulation of Ca 2+ in the SR and the subsequent luminal Ca 2+ -dependent activation of RyRs. 10 In this study, we investigated the mechanism by which micromolar cADPR concentrations affect [Ca 2+ ] i in cardiac myocytes.

MATERIALS AND METHODS Materials
were from Calbiochem (San Diego, CA, USA). Fura 2 AM and Fura 2 lowaff were obtained from Invitrogen (Carlsbad, CA, USA) and TEFLabs (Austin, TX, USA), respectively.

Preparation of cardiac myocytes
This study was approved by the institutional review committee of Chonbuk National University (Reference Number: CBU 2008-0057). New Zealand white rabbits (1.8-2.4 kg) were anesthetized with an intramuscular injection of 50 mg kg − 1 ketamine and 20 mg kg − 1 xylazine hydrochloride. The hearts were removed after deep anesthesia was confirmed by the disappearance of the corneal reflex and the withdrawal of the hindlimb resulting from clamping of the paw. Cardiac myocytes were enzymatically isolated from the ventricle as previously described 11

Fluorimetric determination of [Ca 2+ ] i
Rabbit ventricular myocytes were loaded with fura 2-AM (5 μM) through a 60-min incubation. After they were washed, the cells were seeded in a 200-μl well with a glass coverslip on the bottom and incubated on the stage of an inverted fluorescence microscope (Nikon, Tokyo, Japan) that was continuously perfused at 37°C. The fluorescence was measured at a determined site through a pinhole with alternating excitation wavelengths of 340 and 380 nm and an emission wavelength of 510 nm using a Ca 2+ microspectrofluorometer (PTI). At the end of each recording, the data were calibrated in terms of [Ca 2+ ] i as described by Grynkyewicz et al. based on equation (a) 12 and normalized to calculate the % increase of [Ca 2+ ] i (b) 6,13-15 A K d value of 229 nM was assumed for the binding of Ca 2+ to fura 2-AM. R max , R min , Sf2 and Sb2 were measured in each experimental cell by the addition of 20 mM CaCl 2 (R max ) and 50 mM EGTA (R min ).
At the beginning of each experiment, the cells were washed in a Ca 2+ -free solution containing (mM) 135 NaCl, 5.4 KCl, 1 MgCl 2 , 10 glucose, and 5 HEPES. Patch-pipettes were pulled from borosilicate glass capillary tubules by a micropipette puller (PP-83, Narishige Co. Ltd, Tokyo, Japan) and fire-polished. The patch-pipette tip resistance was between 4-6 MΩ. The intracellular solution for filling each patch pipette contained 110 mM KCl, 5 mM K 2 ATP, 10 mM HEPES (pH 7.2 with KOH). After attachment to the myocytes, the resistance was monitored beginning when the giga seal formation (Axon AxoScope 10). cADPR (100 μM) or IP 3 (100 nM) was applied to the cytosol by rupturing the plasma membrane with a patch-pipette containing each reagent. The reagents 8-Bromo cADPR (Br-cADPR), ryanodine, caffeine, dantrolene, xestospongin, tetracaine, the PTP1B inhibitor, the SHP1/2 PTPase inhibitor or sodium stibogluconate were applied to the bath. Thapsigargin was also applied to the cytosol with a patch pipette.

Immunoprecipitation and immunoblotting
The isolated ventricular myocytes were lysed in lysis buffer containing (in mM) 150 NaCl, 20 HEPES, 1 EDTA, and 0.1 PMSF, as well as 1% Triton X-100 (in μg ml − 1 ), at pH 7.2, and the lysates were centrifuged at 13,000 g for 10 min. The proteins were then immunoprecipitated with rabbit IgG (Sigma), mouse anti-SERCA (Affinity BioReagents), mouse anti-α-actin (Sigma) or mouse anti-PLB antibodies (Affinity BioReagents) (1:100 dilution). The immune complexes were subsequently collected by adding protein A or G beads (1/10 volume, Sigma), fractionated by 10% SDS-PAGE, and transferred to polyvinylidene difluoride membranes. The blots were incubated with anti-phosphotyrosine (Santa Cruz Biotechnology, Inc., Dallas, TX, USA), anti-α-actin, anti-PLB antibodies, anti-phospho-Ser, 16 or antiphospho-Thr 17 PLB antibodies (Badrilla, Leeds, UK) and subsequently incubated with goat anti-mouse or anti-rabbit alkaline phosphataseconjugated secondary antibodies (Santa Cruz Biotechnology). The proteins were visualized using an enhanced chemiluminescence system (Intron, Seongnam, Republic of Korea) and an LAS 3000 imaging system (Fuji, Tokyo, Japan). To detect the total or tyrosinephosphorylated α-actin in SR vesicles, we removed the IgG heavy chain band using ImmunoCruzTM IP/WB Optima E (Santa Cruz Biotechnology, Inc.) according to the manufacturer's instructions.

Fluorimetric measurement of SR Ca 2+ uptake
Rabbit SR vesicles (0.5 mg per 1 ml cuvette) were pre-incubated for 1 min at room temperature in the SR Ca 2+ uptake solution containing (mM) 50 KCl, 20 MOPS, 0.01 CaCl 2 , 5 NaN 3 , 1 KH 2 PO 4 , 5 creatine phosphate, and 10 U creatine phosphokinase. Rabbit SR vesicles (0.5 mg per 1 ml cuvette) suspended in a cuvette were mixed with 1 μM Fura-2 lowAff. The Fura 2 fluorescence outside of the vesicles was monitored over time with alternating excitation wavelengths of 340 and 380 nm and an emission wavelength of 510 nm, using a Ca 2+ microspectrofluorometer (PTI). The extravesicular Ca 2+ concentration was expressed as a fluorescence excitation ratio (R 340nm/380nm ).

Identification of proteins
The dried samples were analyzed by matrix-assisted laser desorption/ionization-time-of flight (MALDI-TOF) mass spectrometry (Voyager-DE PRO) for peptide mass fingerprinting and by electrospray ionization quadrupole time of flight (ESI-Q-TOF) mass spectrometry for peptide sequencing. Database searches were carried out using MS-Fit, accessed via the World Wide Web at http://prospector.uscf.edu.

Measurement of ATPase activity in intact SR vesicles
The ATPase activity in the SR vesicles was determined using an ATPase assay kit (Innova Biosciences, Cambridge, UK) according to the manufacturer's instructions with minor modifications. Briefly, SR vesicles (0.1 μg) in a total volume of 100 μl were incubated for 10 min in the presence of various concentrations of cADPR (0-40 μM) with or without a 10 min pretreatment with various inhibitors. The mixture was incubated for 1 min at 37°C in substrate buffer (0.5 M Tris-HCl, 0.1 M MgCl 2 , 10 mM purified ATP, 5 μM calcein-AM, 5 mM sodium azide, and 1 mM ouabain, pH 7.4) with or without Na-orthovanadate (2-20 μM). After incubation, the reaction was stopped by the addition of 50 μl of Gold Mix from the ATPase assay kit. Two minutes later, 20 μl of stabilizer from the ATPase assay kit was added, and the solution was incubated for 20 min at 37°C in the dark. The enzyme activity was calculated by measuring the Pi-dye complex released via ATP hydrolysis using an ELISA plate reader at 635 nm (Molecular Devices, Sunnyvale, CA, USA).

Statistical analysis
Values are expressed as the mean ± s.e.m. Significant differences were determined by Student's t-test; Po0.05 was considered significant.   Figure 1a). When the cells were pretreated with 10 μM 8-Br-cADPR (a competitive cADPR antagonist) for 10 min, the intracellular cADPR application had no effect on the resting [Ca 2+ ] i (Figure 1b) To prevent the perturbation of Ca 2+ signaling by Fura 2-mediated Ca 2+ buffering and the effect of Ca 2+ influx by calcium channels in the plasma membrane, we investigated whether cADPR induces a Ca 2+ release from isolated SR vesicles using Fura-2 lowAff, a low affinity calcium chelator. The addition of ATP rapidly decreased extravesicular [Ca 2+ ] (Figures 2a and b) due to Ca 2+ chelation followed by Ca 2+ uptake into SR vesicles via an SR Ca 2+ pump and then the maintenance of a steady state of [Ca 2+ ]. When 100 μM cADPR or 10 mM caffeine was added after the steady state in the absence or presence of 10 μM tetracaine, an increase in [Ca 2+ ] outside the vesicles was seen (Figures 2a and b). Tetracaine pretreatment blocked caffeine-but not cADPR-induced increase in extravesicular [Ca 2+ ] (Figures 2a and b). These results support the hypothesis that cADPR-induced increase in resting [Ca 2+ ] i in cardiac myocytes did not involve RyRs.
cADPR induces α-actin tyrosine dephosphorylation and its dissociation from PLB PLB regulates SERCA via a physical interaction. 18 Dephosphorylated PLB inhibits SERCA activity, whereas phosphorylation of PLB by cAMP-dependent protein kinase (PKA) relieves the inhibitory effect on SERCA. 19 PLB is   regulated via phosphorylation/dephosphorylation of Ser16 and Thr17 by PKA, a Ca 2+ /calmodulin-dependent protein kinase and protein phosphatases. 16,17,[20][21][22] It has been reported that cADPR increases SERCA activity in Xenopus oocytes. 10,23 However, we found that cADPR inhibits SR ATPase activity in a concentration-dependent manner, which is completely blocked by 8-Br-cADPR pretreatment (Figure 3a). To determine whether cADPR-induced increase in resting [Ca 2+ ] i involves phosphorylation including PLB phosphorylation, we investigated the effect of phosphatase inhibitors. Okadaic acid (OA), a serine phosphatase inhibitor, had no effect on cADPRinduced increase in resting [Ca 2+ ] i . In contrast, phenylarsine oxide (PAO), a tyrosine phosphatase inhibitor, blocked cADPR-induced increase in resting [Ca 2+ ] i (Figure 3b). However, cADPR and PAO had no effect on phosphorylation of Ser16 residue of PLB (Figure 3c). These data suggest that tyrosine phosphorylation rather than serine/threonine phosphorylation, such as that which characterizes PLB phosphorylation, is required for cADPR-mediated calcium signaling. We identified a 42-kDa tyrosine-phosphorylated protein associated with PLB using an immunoprecipitation assay. cADPR dissociated the protein with PLB, which was blocked by 8-Br-cADPR pretreatment (Figure 3d). MALDI-TOF and ESI-Q-TOF mass spectrometry revealed that the 42-kDa protein was α-actin. cADPR decreased the association of α-actin with PLB by approximately half, which was blocked by PAO pretreatment (Figure 3e). These results indicate that cADPR activates a protein tyrosine phosphatase that dephosphorylates tyrosine residues on α-actin and dissociates α-actin from PLB.
We further investigated the effect of other specific inhibitors of tyrosine phosphatase on cADPR-induced increase in resting [Ca 2+ ] i in isolated rabbit ventricular myocytes and in extravesicular [Ca 2+ ] in SR vesicles (Figures 4a and b). The pharmacological inhibition of PTP1B but not SHP1/2 blocked cADPR-induced increase in resting [Ca 2+ ] i in ventricular myocytes and extravesicular [Ca 2+ ] in SR vesicles. cADPR-mediated decrease of α-actin association with PLB was abolished by PTP1B but not by SHP1/2 inhibition (Figure 4c). These results suggest that α-actin tyrosine dephosphorylation by PTP1B is required for cADPR-induced increase in resting [Ca 2+ ] i in cardiac myocytes.

α-actin integrity is required for cADPR-induced increase in resting [Ca 2+ ] i in isolated rabbit ventricular myocytes
Our results clearly showed that the dissociation of α-actin from PLB plays a critical role in cADPR-induced increase in resting  (Figures 5b and c). These results suggest that a physical interaction between α-actin and PLB is required for cADPR-induced increase in resting [Ca 2+ ] i . cADPR-induced inhibition of SERCA is dependent on α-actin integrity and PTP1B activity in isolated rabbit ventricular myocytes We next investigated the effect of thapsigargin, an SERCA inhibitor, on cADPR-induced increase in resting [Ca 2+ ] i . Direct application of thapsigargin in the cytosol via a patch pipette slowly increased resting [Ca 2+ ] i in a manner similar to that of cADPR, although the effect of thapsigargin was much more potent and slightly more rapid than that of cADPR (Figure 6a). However, the combined application of thapsigargin and cADPR in the cytosol did not have an additive effect on increase in resting [Ca 2+ ] i (Figures 6a and b). Additionally, thapsigargin pretreatment blocked cADPR-induced increase in resting [Ca 2+ ] i (Figure 6c). cADPR inhibited SERCA activity in isolated cardiac SR vesicles in a dose-dependent manner that was blocked by 8-Br-cADPR pretreatment (Figure 3a). cADPR-mediated inhibition of SERCA activity was blocked by α-actin disruption and PTP1B inhibition (Figure 6d). These results suggest that cADPR induces an increase in resting [Ca 2+ ] i by SERCA inhibition via PTP1B -mediated α-actin dephosphorylation.

DISCUSSION
Previous studies have reported that cADPR is a specific agonist of RyR channels and induces Ca 2+ release by sensitizing the RyR to cytosolic Ca 2+ . [24][25][26][27][28][29][30][31][32][33] Lukyanenko et al. 10 reported that cADPR activates SERCA and potentiates a Ca 2+ release in saponin-permeabilized rat ventricular myocytes. However, we demonstrated that cADPR inhibits SERCA in intact rabbit ventricular myocytes. This discrepancy may be due to experimental conditions such as the manner in which nucleotides such as ATP are included in the reaction. The report showed that cADPR activates SERCA by monitoring SR Ca 2+ uptake in the presence of rheutenium red and ATP. Different from that result, we showed that cADPR inhibits SERCA under conditions closer to the actual physiological conditions in the absence of rheutenium red.
Ca 2+ transient is a transient increase of calcium in the cell that peaks and gradually decreases during the cardiac action potential. Ca 2+ transient is exclusively dependent on Ca 2+ influx through L-type calcium channels. Excitationcontraction (EC) coupling in cardiac myocytes is mediated by a mechanism known as calcium-induced calcium release (CICR) where Ca 2+ entry via L-type calcium channels induces calcium release from ryanodine-sensitive Ca 2+ stores. 34,35 In cardiomyocytes, a cytosolic injection of 8-amino-cADPR reduces Ca 2+ transient and contractions, indicating that cADPR increases EC coupling. 29 In addition, photoreleased cADPR induces an increase in the magnitude and frequency of whole cell Ca 2+ transient. In this study, we focused on the effect of cADPR on resting [Ca 2+ ] i but not Ca 2+ transient. We investigated the effect of cADPR on resting [Ca 2+ ] i at higher concentrations than physiological level. We observed that relatively high concentrations of cADPR inhibit rather than activate SERCA activity, resulting in an increase in resting [Ca 2+ ] i in cardiac myocytes. The endogenous level of cADPR in the heart is 1.04-150 pmol mg − 1 , which corresponds to 30 nM-4.5 μM depending on experimental conditions. 36,37 Hypoxia increases cADPR concentration by twofold in the second-order branches of the pulmonary artery and by 10-fold in the third order branches. 38 These results suggest that cADPR level could increase up to approximately 50 μM during specific pathological states in the heart. It has been known that myocardial ischemia, a hypoxic condition induced by blood flow restriction, increases Ca 2+ concentration in cardiomyocytes, and it is associated with abnormal cardiac function. 13 Cardiac ADPR cyclase, a protein distinct from CD38 or the archetypical ADPR cyclase from A. californica, has not been cloned to date. 39 To verify the direct relationships between higher level of cADPR and ischemic injury, the identification of cardiac ADPR cyclase is necessary. The present study suggests that high levels of cADPR under pathophysiological conditions might decrease SR load and inhibit CICR, resulting in the inhibition of excitation-contraction (EC) coupling in cardiac myocytes.
Our results show that α-actin is involved in cADPR-but not IP3-or caffeine-induced increase of [Ca 2+ ] i . The disruption of α-actin blocked cADPR-but not IP 3 -or caffeine-induced increases in resting [Ca 2+ ] i , suggesting that α-actin integrity is required for cADPR action. Interestingly, α-actin was associated with PLB, and their association was decreased by cADPR treatment. The cADPR-mediated dissociation was inhibited by PTP1B inhibition. Our results suggest that cADPR activates PTP1B, which causes tyrosine dephosphorylation of α-actin.
Interestingly, although cADPR and IP 3 act on the same calcium stores, only cADPR-mediated calcium response was sensitive to the disruption of α-actin integrity. IP 3 mediates calcium release by direct action on IP3 receptor in SR vesicles, whereas cADPR mediates it indirectly via PTP1B, possibly resulting in the sensitivity difference to α-actin disruption.