Am J Physiol Cell Physiol 306: C167–C177, 2014. First published November 6, 2013; doi:10.1152/ajpcell.00145.2013.

PP1␥ functionally augments the alternative splicing of CaMKII␦ through interaction with ASF Chao Huang,1* Wenwen Cao,1* Rujia Liao,1* Jia Wang,1 Yuzhe Wang,1 Lijuan Tong,1 Xiangfan Chen,1 Weizhong Zhu,2† and Wei Zhang1† 1

Department of Pharmacology, School of Medicine, Nantong University, Nantong, China; and 2Cardiovascular Research Center, School of Medicine, Temple University, Philadelphia, Pennsylvania Submitted 21 May 2013; accepted in final form 31 October 2013

Huang C, Cao W, Liao R, Wang J, Wang Y, Tong L, Chen X, Zhu W, Zhang W. PP1␥ functionally augments the alternative splicing of CaMKII␦ through interaction with ASF. Am J Physiol Cell Physiol 306: C167–C177, 2014. First published November 6, 2013; doi:10.1152/ajpcell.00145.2013.—Protein phosphatase 1 (PP1) and Ca2⫹/calmodulin-dependent protein kinase ␦ (CaMKII␦) are upregulated in heart disorders. Alternative splicing factor (ASF), a major splice factor for CaMKII␦ splicing, can be regulated by both protein kinase and phosphatase. Here we determine the role of PP1 isoforms in ASF-mediated splicing of CaMKII␦ in cells. We found that 1) PP1␥, but not ␣ or ␤ isoform, enhanced the splicing of CaMKII␦ in HEK293T cells; 2) PP1␥ promoted the function of ASF, evidenced by the existence of ASF-PP1␥ association as well as the PP1␥ overexpression- or silencing-mediated change in CaMKII␦ splicing in ASFtransfected HEK293T cells; 3) CaMKII␦ splicing was promoted by overexpression of PP1␥ and impaired by application of PP1 inhibitor 1 (I1PP1) or pharmacological inhibitor tautomycetin in primary cardiomyocytes; 4) CaMKII␦ splicing and enhancement of ASF-PP1␥ association induced by oxygen-glucose deprivation followed by reperfusion (OGD/R) were potentiated by overexpression of PP1␥ and suppressed by inhibition of PP1␥ with I1PP1 or tautomycetin in primary cardiomyocytes; 5) functionally, overexpression and inhibition of PP1␥, respectively, potentiated or suppressed the apoptosis and Bax/Bcl-2 ratio, which were associated with the enhanced activity of CaMKII in OGD/R-stimulated cardiomyocytes; and 6) CaMKII was required for the OGD/R induced- and PP1␥ exacerbated-apoptosis of cardiomyocytes, evidenced by a specific inhibitor of CaMKII KN93, but not its structural analog KN92, attenuating the apoptosis and Bax/Bcl-2 ratio in OGD/R and PP1␥-treated cells. In conclusion, our results show that PP1␥ promotes the alternative splicing of CaMKII␦ through its interacting with ASF, exacerbating OGD/R-triggered apoptosis in primary cardiomyocytes. protein phosphatase 1␥; alternative splicing factor; CaMKII␦C; splicing; cardiomyocytes

(PP1) is a major serine/threonine protein phosphatase that can regulate diverse pathological processes such as cardiac ischemia (31) and contractility (23, 27). Cardiac ischemia was found to induce phospholamban (PLB) dephosphorylation via activation of PP1, thereby inducing intracellular Ca2⫹ overload in reperfusion (31). Elevated mRNA level of PP1␥ and PP1 activity has been observed in end-stage human failing hearts (19, 21). Inducible expression of phosphatase-1 inhibitor-1 (I1PP1), a protein inhibitor of PP1, protects cardiomyocytes against ischemia/reperfusion in-

TYPE 1 PROTEIN PHOSPHATASE

* C. Huang, W. Cao, and R. Liao contributed equally to this study. † W. Zhu and W. Zhang contributed equally to this study. Address for reprint requests and other correspondence: W. Zhang, Dept. of Pharmacology, School of Medicine, Nantong Univ., Rm. 613, 19# Qixiu Rd., Nantong 226001, China (e-mail: [email protected]). http://www.ajpcell.org

jury by enhancing the basal cardiac function (24). Even so, the exact mechanism of PP1 in heart disease is still obscure. Mammalian species have three major PP1 isoforms, which are encoded by three independent genes: PP1␣, PP1␤, and PP1␥ (5, 7, 28). Although their homologues are 80% identical in amino sequence, they still exhibit distinct functions. For example, PP1␣ is essential for the formation of synaptic plasticity and memory processes (35); PP1␤ was found to attenuate Ca2⫹ transients in cardiomyocytes (1); and PP1␥ mediates the striatal dopamine depletion-induced hyperphosphorylation of synaptic proteins in brain (3). However, how these isoforms are involved in cardiac dysfunction is not clear. Apart from the change in PP1, cardiac dysfunction is also accompanied by the variation of the other proteins such as Ca2⫹/calmodulin-dependent protein kinase ␦ (CaMKII␦), a major isoform of CaMKII in heart (18). Three primary splicing variants of the CaMKII␦ isoform, ␦A, ␦B, and ␦C, have been cloned (8, 11). These variants differing in the presence of an 11-amino-acid nuclear localization signal (NLS) show different subcellular distributions to either nuclear or cytoplasmic compartments (32, 38, 39). The CaMKII␦B isoform contains NLS that is absent from CaMKII␦C. Hence, CaMKII␦ heteromers comprised primarily of ␦B isoform localize to the nucleus, while those with predominantly ␦C isoform distribute into the cytoplasm (38, 39). Another isoform, CaMKII␦A, is also expressed in neonatal cardiomyocytes and begins to switch off 30 days after birth (36). Functionally, both ␦B and ␦C play important roles in cardiac ischemia via promoting the apoptosis of cardiomyocytes (25, 38), and the ␦A isoform was reported to mediate cardiac hypertrophy by regulating fetal cardiac genes such as atrial natriuretic factor (ANF) and ␤-major histocompatibility complex (␤-MHC) (15). Thus CaMKII␦ is a promising target for cardiac disorders therapies. So far, a well-known way to modulate CaMKII␦ is alterative splicing factor (ASF)-mediated alternative splicing (36). ASF, also named serine/arginine-rich splicing factor 1 (SRSF1), is a prototypical SR protein that participates in both constitutive and alternative splicing of different genes (13). Its activity and locations are highly regulated by both phosphorylation and dephosphorylation. For example, ASF can be phosphorylated by cyclic AMP-dependent protein kinase (PKA), resulting in the enhancing of ASF function as well as the reduction in CaMKII␦A and -␦B production and the increase in -␦C production (10). Accordingly, silencing of ASF induces an opposite result in CaMKII␦ splicing (30). Some investigators also found that PP1 interacts with ASF and promotes its dephosphorylation (16), indicating that ASF might be a bridge between PP1 and its downstream effectors.

0363-6143/14 Copyright © 2014 the American Physiological Society

C167

C168

PP1␥ CONTROL OF CaMKII␦ SPLICING

In the present study, to clarify the molecular mechanism and significance of PP1 in CaMKII␦ splicing, we explored the functional correlations between PP1 isoforms and CaMKII␦ splicing. Our results demonstrate that PP1␥ promotes the splicing of CaMKII␦ in both HEK293T cells and primary cardiomyocytes by augmenting the function of ASF, aggravating the oxygen and glucose deprivation followed by reperfusion (OGD/R)-induced cardiac apoptosis likely by increasing CaMKII␦C production and Bax/Bcl2 ratio. Suppression of PP1␥ protects cardiomyocytes from OGD/R-induced apoptosis. MATERIALS AND METHODS

Materials. HEK293T cells were purchased from ATCC (Manassas, VA). Dulbecco’s modified Eagle’s medium/F12 (DMEM/F12) was the product of Gibco Invitrogen (Carlsbad, CA). Fetal bovine serum (FBS) was purchased from Hyclone (Logan, UT). 1% Penicillinstreptomycin, Hoechst 33342, KN93, and KN92 were purchased from Sigma (St. Louis, MO). Tautomycetin (TC) was purchased from Tocris Bioscience (Ellisville, MO). Antibodies against ASF, PP1␤, and PP1␥ were the products of Proteintech (Chicago, IL). Antibodies against PP1␣ and cAMP response element-binding protein (CREB) were purchased from Bio-world Technology (Minneapolis, MN). Antibodies against phospho-CREB and IgG were purchased from Santa Cruz Biotechnology (Santa Cruz, CA). Antibodies against PLB and phospho-PLB were purchased from Badrilla (Leeds, UK). Antibodies against Bax, Bcl-2, and glyceraldehyde 3-phosphate dehydrogenase (GAPDH) were the products of Cell Signaling Technology (Beverly, MA). IRDye 680 donkey anti-rabbit IgG and anti-mouse IgG secondary antibodies were purchased from LI-COR Biosciences (Lincoln, NE). Protein A/G PLUS-Agarose was the product of Santa Cruz Biotechnology. Other related agents were purchased from commercial suppliers. TC was dissolved in dimethyl sulfoxide (DMSO). The final concentration of DMSO was ⬍ 0.05%. No detectable effects of DMSO were found in our experiments. All drugs were prepared as stock solutions, and stock solutions were stored at ⫺20°C. Cell preparation. HEK293T cells were cultured in DMEM/F12 supplemented with 10% FBS and 1% penicillin-streptomycin (100 U/ml). Ventricular cardiomyocytes were isolated from neonatal rats (days 0 –1). The use of rats was approved by the University Animal Ethics Committee of Nantong University (Permit Number: 2110836). Rat ventricles were excised and washed three times with PBS containing 136.8 mM NaCl, 2.6 mM KCl, 0.88 mM KH2PO4, 9.7 mM Na2HPO4, and 1% penicillin-streptomycin (pH 7.4), to wash blood clots out. Clean ventricles were then dissected carefully into small slivers using a surgical scissors. Slivers were then transferred into a phial to digest with a trypsin solution (0.6 mg/ml, 37°C, 20 min). Next, the supernatants were discarded after centrifugation at 500 g for 2 min, and the precipitates were resuspended in FBS. The above steps were repeated 7–10 times until the ventricles were completely digested. Then cell suspensions were placed in culture dish in the humidified 5% CO2/95% air at 37°C. After 1.5 h, the supernatants were removed and diluted to confluence 1 ⫻ 106/ml, and placed in 6-well tissue culture dishes with fresh DMEM/F12 medium. Bromodeoxyuridine (0.5 ␮g/ml) was added to prevent the proliferation of noncardiomyocytes. All cells were kept in humidified 5% CO2/95% air at 37°C. Plasmids transfection. CaMKII␦ minigene, pCI/CaMKII␦ E13E17, comprising CaMKII␦ exons 13, 14, 15, 16, and 17 and introns 13, 14, 15, and 16, was provided by Dr. Jianhua Shi in Nantong University. pCEP4-ASF-HA, pcDNA3.1-PP1␣, pcDNA3.1-PP1␤, and pcDNA3.1-PP1␥ (human) were gifts from Professor Fei Liu in Nantong University. Lipofectamine 2000 from Invitrogen (Carsbad, CA) was used to transfect HEK293T cells. In 24-well plates, cells in each well were incubated with 1.2 ␮g plasmid DNA and 2.0 ␮l

Lipofectamine 2000 in 500 ␮l total Opti-MEM (Invitrogen) for 5 h at 37°C (5% CO2); 500 ␮l of DMEM/F12 supplemented with 10% FBS was then added for further culture. For cotransfection experiments, 1.2 ␮g of total plasmid containing 0.4 ␮g of pCI-Neo-E13-E17 splicing vector, 0.4 ␮g of ASF vector, or 0.4 ␮g of PP1 vector or their corresponding control vectors were used to incubate cells. RNA interference. To silence PP1␥, the short interfering RNA (siRNA) target sequences 5=-GUGGAUGAAACACUAAUGUTT3= (PPP1CC-homo-1096), 5=-GUGCAGAAGUGGUUGCAAATT-3= (PPP1CC-homo-753), and 5=-CUGCUGUCAUGGAGGUUUATT-3= (PPP1CC-homo-926) from Shanghai GenePharma (Shanghai, China) to human PP1␥ were used in this study. The target sequences 5=-GCCCAGAAGUCCAAGUUAUTT-3= (ASF-homo-2598), 5=-GGAGUUUGUACGGAAAGAATT-3= (ASF-homo-795), and 5=-GGAGCUGGAUCAUUGGAUUTT-3= (ASF-homo-687) from Shanghai GenePharma to ASF were used to silence ASF. HEK293T cells were cultured in 12-well plates were transfected with various amounts of siRNA using Lipofectamine 2000 (Invitrogen). The same amounts of scramble siRNA were used for controls. After 48 h, cells were lysed, and cellular protein and RNA were extracted for Western blot and q-PCR assay. Generation of OGD/R model in rat neonatal primary cardiomyocytes. Lack of oxygen and glucose was used to mimic in vitro ischemia, and recovering to normal growth conditions was used to mimic reperfusion. The growth medium was replaced with D-Hanks buffers, and the cells were transferred to a hypoxic incubator (Thermo) filled with saturated nitrogen to incubate 4 h. Oxygen was adjusted to 1.0% and CO2 to 5.0%. Reoxygenation was initiated by replacing the hypoxia solutions with Hanks buffers including Ca2⫹, Mg2⫹, and glucose. The cells were then cultured in the incubator under an atmosphere of 5% (vol/vol) CO2 in air at 37°C for further 12 h. Nontreated cells were used as controls. Hoechst 33342 staining. For Hoechst 33342 staining, cells in different groups were fixed and treated with Hoechst 33342 staining solutions. Cells with Hoechst 33342 solutions were mixed gently and stained for 20 –30 min at 4°C, then the cells were observed using fluorescent microscope. Cells with condensed bright nuclei were regarded as apoptotic. The apoptosis rate was calculated by the ratio between the numbers of cells with condensed bright nuclei and total cell numbers. The number of apoptotic cells was counted from the resulting four phases for each point with a digital camera and microscope (Olympus, Japan), and then averaged for each experimental condition. The data presented were generated from three separate assays. TUNEL assay. The apoptotic nuclei of cardiomyocytes were detected with one step terminal deoxynucleotidyl transferase dUTP nick end labeling (TUNEL) assay kit from Bi Yuntian Biological Technology Institution (Shanghai, China). The images of TUNEL-positive cells were viewed with a Nikon Eclipse 800 microscope. PP1␥ adenovirus construction and transfection. The AdEasy Adenoviral vector system was used to prepare recombinant adenoviruses. Rat full-length PP1␥ and I1PP1 cDNA from NCBI (NM 022498.1 and AY648296.1, respectively) were subcloned into the pShuttle vector, which contained cDNA for enhanced green fluorescent protein (EGFP) and red fluorescent protein (RFP), respectively, and were under the control of the cytomegalovirus promoter (Ad/PP1␥ and Ad/I1PP1). Recombinant adenovirus plasmids containing EGFP and RFP (Ad/EGFP and Ad/RFP, respectively) were used as controls. For adenoviral packaging, adenoviruses were amplified in HEK293T cells, the particle count was estimated on a spectrophotometer at a wavelength of OD260, and viral titers were determined by plaque assay. For primary cardiomyocytes, cells were cultured in 6-well plates, and the serum-containing medium was then changed with serum-free medium immediately before transfection. Twenty-four hours later, cardiomyocytes were infected with Ad-Ppp1r1a and Ad-Ppp1cc at 200 –500 viral particles/cell for 24 h. Western blot. To extract the total protein, HEK293T cells or cardiomyocytes were lysed on ice for 30 min in lysis buffer (50 mM

AJP-Cell Physiol • doi:10.1152/ajpcell.00145.2013 • www.ajpcell.org

PP1␥ CONTROL OF CaMKII␦ SPLICING

Tris-HCl, pH 7.4, 1 mM EDTA, 100 mM NaCl, 20 mM NaF, 3 mM Na3VO4, 1 mM PMSF, with 1% Nonidet P-40, and protease inhibitor cocktail). The lysates were centrifuged at 12,000 g for 15 min, and the supernatants were recovered. After denaturation by heating at 95°C for 5 min, 30 ␮g of proteins was separated on 10% SDS/PAGE gels and then transferred to nitrocellulose membranes by using a transfer cell system (Bio-Rad). The membranes were then probed with appropriate primary antibodies after incubation with blocking buffer. Membrane-bound primary antibodies were assayed with IRDye 680-labeled secondary antibodies. Immunoblots were visualized by scanning using Odyssey. All assays were performed at least three times. Analysis of protein band was performed with ImageJ software. Quantification of the splicing of CaMKII␦ by reverse transcription reaction (RT-PCR). Total RNA was isolated from HEK293T cells or purified ventricular cardiomyocytes using Trizol reagent (Invitrogen) followed by isopropanol precipitation. RNA concentration was assessed by spectrophotometry, and 2 ␮g RNA were used for cDNA synthesis with the RevertAid First Strand cDNA Synthesis system for RT-PCR kits (Fermentas). Primers used for checking different CaMKII␦ isoforms in HEK293T cells were 5=-GGTGTCCACTCCCAGTTCAA-3= (forward) and 5=-GTCTTCA TCCTCAATGGTGGTG-3= (reverse). Primers used for checking CaMKII␦A, CaMKII␦B, and CaMKII␦C in primary cardiomyocytes were quoted from a previous study (36): CaMKII␦A, 5=-CGAGAAATTTTTCAGCAGCC-3= (forward) and 5=-ACAGTAGTTTGGGGCTCCAG-3= (reverse); CaMKII␦B, 5=-CGAGAAATTTTTCAGCAGCC-3= (forward) and 5=GCTCTCAGTTGACTCCATCATC-3= (reverse); CaMKII␦C, 5=CGAGAAATTTTTCAGCAGCC-3= (forward) and 5=-CTCAGTTGACTCCTTT ACCCC-3= (reverse). Primers used for checking GAPDH were 5=-CCATGTTCGTCATGGGTGTGAACCA-3= (forward) and 5=GCCAGTAGAGGCAGGGATGATGTTC-3= (reverse). Three microliters of the first strand product was used as a template in each 25 ␮l PCR reactions. The cycling parameters consisted of one cycle of 3

C169

min at 94°C, then 30 cycles of 30 s at 94°C, and then 30 s at 55°C. RT-PCR products were electrophoresed on a 2% agarose gel. All assays were performed at least three times. Analysis of DNA band was performed with ImageJ software. The expression levels of CaMKII␦ isoforms were normalized to the total density of the three CaMKII␦ isoforms in the same treatment group and were shown as the percentage of relative density. Coimmunoprecipitation. Protein lysates were extracted from HEK293T cells or primary cardiomyocytes with cell lysis buffer (50 mM Tris-HCl, pH 7.4, 1 mM EDTA, 100 mM NaCl, 20 mM NaF, 3 mM Na3VO4, 1 mM PMSF, with 1% Nonidet P-40, and protease inhibitor cocktail), and centrifuged at 12,000 g for 15 min at 4°C. The PP1␣, PP1␤, and PP1␥ antibodies were added into cell lysates (1 ␮g/␮l) and incubated overnight on a rotary wheel at 4°C. The protein A/G PLUS agarose beads were washed 2 times (2 min per time) using normal washing buffer (50 mM Tris-HCl, pH 7.4, 100 mM NaCl, 1 mM EDTA, with 1% Nonidet P-40, and protease inhibitor cocktail). Then the washed beads were added into the above protein-antibody mixture (0.25– 0.5 ␮g antibody/10 ␮l agarose beads), and incubated for an additional 2 h at room temperature. Beads were washed 1 time with normal washing buffer at first. Then the beads were further washed 5 times with high-salt washing buffer (50 mM Tris-HCl, pH 7.4, 500 mM NaCl, 1 mM EDTA, with 1% Nonidet P-40, and protease inhibitor cocktail). Bead-bound proteins were dissolved in 4⫻SDS loading buffer, and boiled at 95°C for 5 min. Finally, the identity of proteins was determined by Western blot. For checking the association between ASF and PP1␥, bead-bound ASF was normalized to total ASF. For checking the expression of PP1␥ after pathological stimulation, PP1␥ was normalized to GAPDH. Analysis of protein band was performed with ImageJ software. Statistical analysis. All data are expressed as means ⫾ SE. Statistical analysis was performed using one way ANOVA test following

Fig. 1. Protein phosphatase 1␥ (PP1␥), but not PP1␣ or -␤, promotes CaMKII␦ splicing in HEK293T cells. A: schematic diagram of the alternative splicing of exons 14, 15, and 16 of mini-CaMKII␦-genes, pCI/CaMKIIdE13-E17. NLS, nuclear localization signal. B, top: comparable expression of PP1␣, -␤, and -␥ in HEK293T cells. B, bottom: statistical data showing the expression of PP1 isoforms (n ⫽ 3, *P ⬍ 0.05, **P ⬍ 0.01 vs. control). C: overexpression of PP1␥, but not PP1␣ or -␤, altered Ca2⫹/calmodulin-dependent protein kinase ␦ (CaMKII␦) splicing. CaMKII␦ splicing was measured by RTPCR at 48 h after transfection of miniCaMKII␦-genes. D: statistical data showing the change in CaMKII␦ splicing after overexpression of different PP1 isoforms (n ⫽ 3, *P ⬍ 0.05 vs. control). E: knockdown of PP1␥ by short interfering RNA (siRNA) inhibited CaMKII␦ splicing in HEK293T cells. pCI/CaMKIIdE13-E17 was cotransfected with PP1␥ siRNA or scramble siRNA into HEK293T cells. E, top: PP1␥ expression in PP1␥ siRNA-treated cells. E, bottom: CaMKII␦ splicing was measured by RTPCR at 48 h after transfection. Overexpression of PP1␥ was used as the positive control. F: statistical data showing the change in CaMKII␦ splicing after PP1␥ overexpression or PP1␥ siRNA (n ⫽ 3, *P ⬍ 0.05 vs. control). All data are shown as means ⫾ SE.

AJP-Cell Physiol • doi:10.1152/ajpcell.00145.2013 • www.ajpcell.org

C170

PP1␥ CONTROL OF CaMKII␦ SPLICING

by post hoc test. P ⬍ 0.05 or P ⬍ 0.01 was considered statistically significant. RESULTS

PP1␥, but not PP1␣ or -␤, promotes the splicing of CaMKII␦ in HEK293T cells. It has been shown that PP1 and CaMKII␦C activity are simultaneously upregulated in cardiac disorders (23, 31, 38). However, the relationship between PP1 isoforms and CaMKII␦ splicing was unclear. To test if PP1 isoform specifically promotes CaMKII␦ splicing, we tested the splicing of CaMKII␦ in HEK293T cells cotransfected miniCaMKII␦ gene pCI/CaMKIIE13-E1726 (Fig. 1A) and human PP1␣, -␤, and -␥ plasmids using RT-PCR. In comparable level of PP1␣, -␤, and -␥, evidenced by Western blot (Fig. 1B), only PP1␥ significantly reduced the CaMKII␦A and -␦B production, and correspondingly enhanced CaMKII␦C production (Fig. 1,

C and D). To further confirm the involvement of PP1␥ in CaMKII␦ splicing, endogenous PP1␥ was knocked down by PP1␥-specific siRNA in HEK293T cells. As expected, PP1␥ was significantly silenced by siRNA at the concentrations ranging from 20 to 100 nM (Fig. 1E). This silencing of PP1␥ decreased the CaMKII␦C production and simultaneously increased CaMKII␦A and -␦B production (Fig. 1, E and F). The change of CaMKII␦ splicing in PP1␥-overexpressed cells was used as the positive control (Fig. 1, E and F). Taken together, these results demonstrate that PP1␥ promotes the splicing of CaMKII␦ in HEK293T cells. ASF is required for PP1␥-mediated CaMKII␦ splicing in HEK293T cells. Because ASF was reported to regulate the splicing of CaMKII␦ in different cells (16, 40), we further confirmed the requirement of ASF in PP1␥-mediated CaMKII␦ splicing. Endogenous ASF was knocked down by ASF-specific

Fig. 2. PP1␥ potentiates ASF-elicited CaMKII␦ induction in HEK293T cells. A: knockdown of alternative splicing factor (ASF) by siRNA inhibited the splicing of CaMKII␦. pCI/CaMKII␦E13-E17 was cotransfected with ASF siRNA or scramble siRNA into HEK293T cells. The expression of ASF was measured by Western blot (top). The splicing of CaMKII␦ was measured by RT-PCR at 48 h after transfection (bottom). Overexpression of ASF was used as the positive control. B: statistical data showing the change in CaMKII␦ splicing after ASF overexpression or ASF siRNA (n ⫽ 3, *P ⬍ 0.05 vs. control). C: representative images showing that HEK293T cells were cotransfected with 0.1 ␮g of ASF and different concentrations of PP1␥ (0.05, 0.1, 0.2, and 0.4 ␮g, respectively) for 48 h, and the CaMKII␦ splicing was measured by RTPCR. D–F: statistical data showing the change in CaMKII␦C (D), -␦A (E), and -␦B (F) production in ASF- and PP1␥-overexpressed HEK293T cells (n ⫽ 3, *P ⬍ 0.05 vs. control; #P ⬍ 0.05, ##P ⬍ 0.01 vs. ASF group). G: knockdown PP1␥ by siRNA inhibited the splicing of CaMKII␦ in ASF-overexpressed HEK293T cells. pCI/ CaMKII␦E13-E17 was cotransfected with ASF plasmids and PP1␥ siRNA or scramble siRNA into HEK293T cells. The CaMKII␦ splicing was measured by RT-PCR at 48 h after transfection. Overexpression of both ASF and PP1␥ was used as the positive control. H: statistical data showing the change in CaMKII␦ splicing after ASF overexpression and PP1␥ siRNA (n ⫽ 3, *P ⬍ 0.05 vs. ASF alone group). I: representative image showing the interaction between ASF and PP1␥ in HEK293T cells. All data are shown as means ⫾ SE.

AJP-Cell Physiol • doi:10.1152/ajpcell.00145.2013 • www.ajpcell.org

PP1␥ CONTROL OF CaMKII␦ SPLICING

siRNA (Fig. 2A). Downregulation of ASF expression remarkably decreased the CaMKII␦C production and simultaneously increased CaMKII␦A and -␦B production in HEK293T cells (Fig. 2, A and B). Overexpression of ASF exhibited opposite results in CaMKII␦ splicing (Fig. 2, A and B), further supporting the essential role of ASF in CaMKII␦ splicing. Next, using cotransfection of ASF with PP1␥ plasmids into HEK293T cells, we tested whether PP1␥ augments the function of ASF in CaMKII␦ splicing. The dose-dependent results showed that cotransfection of ASF and PP1␥ increased the CaMKII␦C production (Fig. 2, C and D) and decreased CaMKII␦A and -␦B production (Fig. 2, C, E, and F). Both the maximal increase in CaMKII␦C production and the maximal decrease in CaMKII␦A and -␦B production occurred at the concentration of 0.4 ␮g/well of PP1␥ and 0.1 ␮g/well of ASF (Fig. 2, C–F). These data demonstrated that PP1␥ mediates the CaMKII␦ splicing via augmenting the pro-splicing function of ASF. To further establish the role of PP1␥ in ASF-mediated

C171

splicing of CaMKII␦, we knocked down endogenous PP1␥ with PP1␥-specific siRNA in ASF-overexpressed cells. As anticipated, downregulation of PP1␥ expression decreased the CaMKII␦C production and also increased CaMKII␦A and -␦B production in ASF-overexpressed cells (Fig. 2, G and H). The change of CaMKII␦ splicing in both PP1␥ and ASF-overexpressed cells was considered as the positive control (Fig. 2, G and H). Finally, we found that ASF interacted with PP1␥ in HEK293T cells, further supporting the function of ASF in PP1␥-mediated splicing of CaMKII␦ (Fig. 2I). In general, these findings demonstrate that ASF is necessary and sufficient for PP1␥ promoting CaMKII␦ splicing. PP1␥ interacts with ASF to promote CaMKII␦ splicing in primary cardiomyocytes. To test whether PP1␥ affects CaMKII␦ splicing in cultured neonatal rat cardiomyocytes, we assayed the mRNA level of CaMKII␦A, -␦B, and -␦C in cultured cardiomyocytes transfected with an adeno-PP1␥ tagged (nonfused) with enhanced green fluorescent protein (EGFP) (PP1␥-

Fig. 3. Overexpression of PP1␥ promotes the endogenous CaMKII␦ splicing in primary cardiomyocytes. A, top: representative images showing the effect of phosphatase inhibitor tautomycetin (TC) on the phosphorylation level of CREB in primary cardiomyocytes. Peak change in CREB phosphorylation level was achieved at the concentration of 10 ␮M. A, bottom: statistical data showing the effect of TC on the phosphorylation level of CREB in cardiomyocytes (n ⫽ 3, *P ⬍ 0.05, **P ⬍ 0.01 vs. control). B: representative images showing the change in CaMKII␦ splicing in primary cardiomyocytes. Starved cardiomyocytes were transfected with adenovirus encoding PP1␥-EGFP, EGFP, I1PP1-RFP, and REP. After transfection (48 h), RNA was isolated from cardiomyocytes to detect the mRNA of CaMKII␦ isoforms. In subset cells, cardiomyocytes were pretreated with TC (10 ␮M, 30 min). C: statistical data showing the change in CaMKII␦ splicing after PP1␥ and I1PP1 overexpression or TC treatment (n ⫽ 3, *P ⬍ 0.05 PP1␥-EGFP vs. EGFP, I1PP1RFP vs. RFP, TC vs. control). D: representative images showing the change in CaMKII␦ splicing in OGD/R-treated cardiomyocytes with or without PP1␥-EGFP, EGFP, I1PP1-RFP, and REP, or TC treatment (10 ␮M, 30 min). At 48 h after cardiomyocytes were transfected with adenovirus vector, the cells were performed with OGD/R as described in METHODS. RNA was isolated for RT-PCR. In subset cells, cardiomyocytes were pretreated with TC and then insulted with OGD/R. E: statistical data showing the change in CaMKII␦ splicing after PP1␥ and I1PP1 overexpression or TC treatment in OGD/R-treated cells (n ⫽ 3, *P ⬍ 0.05, **P ⬍0.01 vs. control; #P ⬍ 0.05, PP1␥-EGFP-OGD/R vs. EGFPOGD/R, I1PP1-RFP-OGD/R vs. RFPOGD/R, TC-OGD/R vs. OGD/R). All data are shown as means ⫾ SE.

AJP-Cell Physiol • doi:10.1152/ajpcell.00145.2013 • www.ajpcell.org

C172

PP1␥ CONTROL OF CaMKII␦ SPLICING

EGFP), adeno-I1PP1 tagged with RFP, or a control vector encoding EGFP/RFP alone. We also explored the pharmacological approach to inhibit the activation of PP1 with TC, evidenced by the enhanced pS133-CREB phosphorylation (Fig. 3A). In accordance with the effect of PP1␥ on CaMKII␦ splicing in HEK293T cells, PP1␥ also increased the CaMKII␦C production and at the same time decreased CaMKII␦A and -␦B production (Fig. 3, B and C) in primary cardiomyocytes. EGFP or RFP, as a control, alone had no significant influence on CaMKII␦ splicing. In contrast, inhibition of PP1 by transfection of I1PP1 tagged with enhanced red fluorescent protein (RFP) (I1PP1-RFP) reduced CaMKII␦C production and simultaneously increased CaMKII␦A and -␦B production (Fig. 3, B and C), providing evidence for the involvement of PP1␥ in CaMKII␦ splicing in primary cardiomyocytes. Pretreatment of cells with TC (10 ␮M) further confirmed the role of PP1␥ in CaMKII␦ splicing (Fig. 3, B and C). Furthermore, OGD/R also induced a significant splicing of CaMKII␦ in neonatal rat cardiomyocytes, evidenced by an increase in CaMKII␦A and -␦C production as well as a decrease in CaMKII␦B production (Fig. 3, D and E). Overexpression of PP1␥ further increased the CaMKII␦C production and decreased the CaMKII␦A and -␦B production (Fig. 3, D and E). In contrast, inhibition of PP1␥ with cotransfection of I1PP1 or TC (10 ␮M) exhibited an opposite effect on CaMKII␦ splicing induced by OGD/R-treated cardiomyocytes (Fig. 3, D and E). Because earlier studies have confirmed the interaction between PP1 and ASF (16), we next investigated the association of PP1␥ with ASF in primary cardiomyocytes. We did not observe a significant increase in the expression of ASF protein

in OGD/R-treated cardiomyocytes, but OGD/R stress remarkably enhanced the association of PP1␥ with ASF by 190 ⫾ 12% over control (P ⬍ 0.01, n ⫽ 4, Fig. 4, A and B). Overexpression of PP1␥ increased the association of ASF with PP1␥ by 57 ⫾ 5% over the association in OGD/R-stimulated cells (P ⬍ 0.05, n ⫽ 4, Fig. 4, A and B). These effects were PP1␥-specific because there was no interaction between PP1␣ and ASF. The level of PP1␥ protein was increased by 65 ⫾ 3% over control (P ⬍ 0.01, n ⫽ 4, Fig. 4, C and D) in OGD/Rstimulated cardiomyocytes. Of note, the interaction between PP1␤ and ASF or the expression of PP1␣ and PP1␤ protein was confirmed not to be changed by OGD/R stimulation (Fig. 4, E and F). PP1␥ exacerbates OGD/R-induced apoptosis of primary cardiomyocytes associated with CaMKII␦ splicing and Bax/ Bcl-2 expression. The importance of CaMKII␦C in heart ischemia (25, 40) and the possibility of mRNA translation prompted us to determine whether the PP1␥-induced change in CaMKII␦ splicing affects ischemic apoptosis. Apoptotic cells were measured after OGD/R insult using Hoechst 33342 staining and TUNEL assay. As expected, OGD/R itself induced a significant increase in apoptotic numbers of primary cardiomyocytes, and PP1␥-EGFP overexpression further increased it (Fig. 5, A–C). In contrast, either I1PP1-RFP overexpression or TC pretreatment (10 ␮M, 30 min) exhibited an opposite result (Fig. 5, A–C). These results suggest that PP1␥ is sufficient and necessary to prompt the OGD/R-induced cardiomyocytes apoptosis. To test whether the proapoptotic role of PP1␥ might be associated with CaMKII activity, we observed the change in the phosphorylation level of PLB, a specific substrate of

Fig. 4. Overexpression of PP1␥ potentiates the association between ASF and PP1␥ in OGD/R-treated primary cardiomyocytes. A: representative IP-WB results showing the change in ASF-PP1␥ association in OGD/ R- and/or PP1␥-stimulated cardiomyocytes. Cell lysates were incubated with PP1␥ primary antibody overnight (15 h) at 4°C to pull down ASF. The precipitate was measured by Western blot with anti-ASF antibody. B: statistical data showing the variation of ASF in OGD/R- and/or PP1␥-stimulated cardiomyocytes (n ⫽ 4, *P ⬍ 0.05, **P ⬍ 0.01 vs. control; #P ⬍ 0.05 vs. OGD/R). C: representative images showing the change in PP1␥ protein expression in OGD/R- and/or PP1␥-stimulated cardiomyocytes. D: statistical data showing the variation of PP1␥ protein expression in OGD/R- and/or PP1␥-stimulated cardiomyocytes (n ⫽ 4, *P ⬍ 0.05, **P ⬍ 0.01 vs. control; #P ⬍ 0.05 vs. OGD/R). E: representative images showing the change in ASF-PP1␣ or ASFPP1␤ association in OGD/R-stimulated cardiomyocytes (n ⫽ 4). F: representative images showing the change in PP1␣ or PP1␤ in OGD/ R-stimulated cardiomyocytes (n ⫽ 4). All data are shown as means ⫾ SE.

AJP-Cell Physiol • doi:10.1152/ajpcell.00145.2013 • www.ajpcell.org

PP1␥ CONTROL OF CaMKII␦ SPLICING

C173

Fig. 5. Overexpression or inhibition of PP1␥ alters apoptosis and Bax/Bcl-2 ratio induced by OGD/R stress in primary cardiomyocytes. A: representative images showing the change in apoptosis after PP1␥ and I1PP1 overexpression or TC (10 ␮M) treatment in primary cardiomyocytes. Bars: 50 ␮m. B and C: statistical data showing the percentage of apoptotic nuclei in cardiomyocytes (B: Hoechst 33342 staining; C: TUNEL assay; **P ⬍ 0.01 vs. control; #P ⬍ 0.05 vs. OGD/R). Data were from 3 independent experiments. D: representative images showing the change in the phosphorylation level of p-PLB after PP1␥ and I1PP1 overexpression or TC (10 ␮M) treatment in primary cardiomyocytes. E: statistical data showing the change in the phosphorylation level of pPLB in cardiomyocytes (n ⫽ 3, **P ⬍ 0.01 vs. control; #P ⬍ 0.05, ##P ⬍ 0.01 vs. OGD/R). F: representative images showing the change in Bax and Bcl-2 protein expression after PP1␥ and I1PP1-RFP overexpression or TC (10 ␮M) treatment in primary cardiomyocytes. G: statistical data showing the change in Bax/Bcl-2 ratio in cardiomyocytes (n ⫽ 3, *P ⬍ 0.05 vs. control; #P ⬍ 0.05, ##P ⬍ 0.01 vs. OGD/R). All data are shown as means ⫾ SE.

CaMKII (40), in primary cardiomyocytes. As shown in Fig. 5, D and E, OGD/R itself increased the phosphorylation level of PLB in primary cardiomyocytes, which was further increased by PP1␥-EGFP overexpression. In contrast, both overexpression of I1PP1-RFP and pretreatment of TC attenuated the phosphorylation level of PLB (Fig. 5, D and E). Bax and Bcl-2 are the two classical proteins that contribute to the development of apoptosis (6, 9). To determine whether Bax and Bcl-2 participate in the PP1␥-induced apoptosis, we assayed the expression of Bax and Bcl-2 protein after PP1␥ and I1PP1 overexpression or TC treatment. As shown in Fig. 5, F and G, OGD/R itself increased the Bax/Bcl-2 ratio in cardiomyocytes. Overexpression of PP1␥ further increased this ratio in OGD/R-treated cells. In contrast, I1PP1 overexpression or TC treatment reduced Bax/Bcl-2 ratio (Fig. 5, F and G). Neither EGFP nor RFP had a significant effect on Bax/Bcl-2 ratio (Fig. 5, F and G). These findings demonstrate that PP1␥ promotes the apoptosis of primary cardiomyocytes via disturbing the Bax/Bcl-2 balance. Finally, we used KN93, a specific inhibitor of CaMKII, to investigate if CaMKII is required for the PP1␥-mediated promotion of apoptosis in primary cardiomyocytes. We found that KN93 significantly attenuated the OGD/R and PP1␥ cotreat-

ment-induced increase in apoptotic cells (Fig. 6, A and B). KN92, a structural analog of KN93, failed to affect the OGD/R and PP1␥ cotreatment-induced apoptosis (Fig. 6, A and B). Furthermore, KN93, but not KN92, was found to reduce the OGD/R and PP1␥ cotreatment-induced increase in Bax/Bcl-2 ratio, and also reduce the phosphorylation level of PLB (Fig. 6, C and D), suggesting that activation of CaMKII␦C, a cytosollocated CaMKII␦ isoform, is indeed required for PP1␥-mediated apoptosis in primary cardiomyocytes. DISCUSSION

The findings herein indicate that PP1␥ promotes the alternative splicing of CaMKII␦ by augmenting the function of ASF in HEK293T cells and primary cardiomyocytes, leading to an increase in CaMKII␦C production as well as a decrease in CaMKII␦B production (Fig. 7). PP1␥-induced splicing of CaMKII␦ functionally mediates the OGD/R-induced apoptosis in cardiomyocytes. Blocking of CaMKII␦ splicing or activity results in an effect of cardiac protection, exhibited by the decrease in cellular apoptosis and Bax/Bcl-2 ratio. It has been shown that PP1 activation mediates dephosphorylation of PLB in heart ischemia (31). Induction of the expres-

AJP-Cell Physiol • doi:10.1152/ajpcell.00145.2013 • www.ajpcell.org

C174

PP1␥ CONTROL OF CaMKII␦ SPLICING

Fig. 6. Inhibition of CaMKII activity attenuates OGD/R- and PP1␥-mediated apoptosis and Bax/Bcl-2 ratio in primary cardiomyocytes. A: representative images showing the change in apoptosis after KN93 or KN92 (20 ␮M) treatment in OGD/R- and/or PP1␥stimulated primary cardiomyocytes. Bars: 40 ␮m. B: statistical data showing the percentage of TUNEL positive cell ratio in cardiomyocytes (**P ⬍ 0.01 vs. control; #P ⬍ 0.05 vs. OGD/R). Data were from 3 independent experiments. C: representative images showing the change in the phosphorylation level of p-PLB and Bax/Bcl-2 protein expression after KN93 (20 ␮M) or KN92 (20 ␮M) treatment in OGD/R- and/or PP1␥-stimulated primary cardiomyocytes. D: statistical data showing the change in the phosphorylation level of p-PLB and Bax/ Bcl-2 protein expression after KN93 or KN92 treatment in cardiomyocytes (n ⫽ 3, *P ⬍ 0.05, **P ⬍ 0.01 vs. control; #P ⬍ 0.05, ##P ⬍ 0.01 vs. OGD/R). All data are shown as means ⫾ SE.

sion of I1PP1, an inhibitor of PP1, also protects the cardiomyocytes from ischemic injury (24). Consistently, our present studies showed that overexpression of PP1␥ can aggravate the OGD/R-triggered apoptosis in primary cardiomyocytes, while inhibition of PP1␥ ameliorates apoptosis, suggesting that PP1␥ might be crucial for cardiac apoptosis during ischemia. Bax/ Bcl-2 ratio determines whether cells go into apoptosis or not. We found that overexpression or inhibition of PP1␥ results in an opposite effect on OGD/R-induced increase in Bax/Bcl-2 ratio, further supporting the crucial function of PP1␥ in cardiac injury in vitro ischemia. Splicing factor containing arginine-serine repeats (SR) protein is the essential factor that controls the splicing of precursor mRNA by regulating multiple steps in spliceosome development. The prototypical protein of SR ASF contains two NH2terminal RNA recognition motifs and a 50-residue COOHterminal RS domain that can be phosphorylated at numerous serines by SR-specific protein kinase 1 (SRPK1) (2, 12, 17). Recent studies have shown that phosphorylation of about 10 serines in the RS domain by SRPK1 leads to translocation of ASF from the cytoplasm to the nucleus (14, 22). In contrast, Clk/Sty, a member of the CLK family of nuclear protein

kinases, phosphorylates additional serines in the COOH terminus of RS domain (RS2) and causes ASF to disperse within the nucleus (22). While this model presents the subcellular localization data, in vitro splicing studies have shown that both hypo- and hyperphosphorylation inhibit the splicing activity of ASF, indicating that the link between phosphorylation and splicing is complex and partial phosphorylation states may be important (26). Protein kinase A (PKA) was also found to phosphorylate ASF, and this phosphorylation augments the CaMKII␦ splicing (10) and its function in tau exon 10 inclusion (29). However, phosphorylation of ASF at Ser227, Ser234, and Ser238 by Dyrk1A suppresses the function of ASF in the alternative splicing of tau exon 10 (30). Moreover, the RS1 segment has been reported to be dephosphorylated by PP1 (16). We thus speculate that phosphorylation-dephosphorylation differentially affects the activity of splicing factor as well as its cellular localization. Our present results also support this speculation that PP1␥ binds to ASF and promotes ASF function in CaMKII␦ splicing, consequently resulting in an increase in activation of CaMKII␦C and cardiac apoptosis. However, PP1␤ did not affect the function of ASF although it also binds with ASF in cardiomyocytes.

AJP-Cell Physiol • doi:10.1152/ajpcell.00145.2013 • www.ajpcell.org

PP1␥ CONTROL OF CaMKII␦ SPLICING

Fig. 7. Proposed scheme shows how PP1␥ regulates the ASF-mediated splicing of CaMKII␦ and cell apoptosis in cultured rat primary cardiomyocytes. PP1␥ dephosphorylates ASF to promote the alternative splicing of CaMKII␦, resulting in an increase in CaMKII␦C production and a decrease in CaMKII␦B production to disturb the ratio of Bax/Bcl2 and potentiate the OGD/R-triggered apoptosis in primary cardiomyocytes.

ASF plays an important role in the splicing of CaMKII␦ exons, and this splicing is reported to contribute to the functional remodeling of developing heart (36). Knockout of ASF in the mouse heart promotes the inclusion of exons 15 and 16 and suppresses exclusions of other exons, leading to the increase in CaMKII␦A production and the decrease in CaMKII␦B and -␦C production (36). Dysregulation of CaMKII␦ isoforms in ASF deficient mice appears to enhance excitation-contraction coupling and hypercontraction in isolated cardiomyocytes, the phenotype resulting from the targeting of the kinase to the T-tubules where several key Ca2⫹ handling proteins are located. In this study, we observed that overexpression of ASF promotes the exclusion of exons 14, 15, and 16, increasing the production of endogenous CaMKII␦C in cardiomyocytes, which is consistent with ASF KO study. In general, our study demonstrates that ASF is essential and sufficient for the splicing of CaMKII␦ in HEK293 cells, giving evidence to support the functional significance of PP1␥/ASF association in cardiomyocytes. Although PP1 has already been confirmed to dephosphorylate the RS1 segment in ASF in a COOH-to-NH2 direction (16), the functional consequence of this effect is unclear. Our findings provide direct evidence for the change in CaMKII␦ splicing in PP1-overexpressed conditions. We found that only ␥ isoform participates in the splicing of CaMKII␦ among three isoforms of PP1. The reason could be that 1) PP1␣, -␤, and -␥ have differential substrates, or 2) PP1␣, -␤, and -␥ act on the same substrates but produce different effects. In fact, earlier studies have already reported that each PP1 isoform affects different substrates in a variety of processes. For example, PP1␣ was found to stimulate apoptosis of tumor cells by interacting with the retinoblastoma protein (34). Downregulation of PP1␤ ameliorates left ventricular diastolic function by

C175

interrupting with the function of sarcoplasmic reticulum Ca2⫹ ATPase (20). PP1␥ can regulate the development of specialized flagellar structure in mammalian spermatozoa (33). Moreover, our findings ascertained that it is PP1␤ and PP1␥ but not PP1␣ that have mutual interactions with ASF, and only PP1␥ can promote the splicing of CaMKII␦. The reason for the difference is to be investigated. In the present study, the proapoptotic role of PP1␥ in ischemia-stressed cardiomyocytes was confirmed to be associated with CaMKII␦ splicing. Overexpression of PP1␥ increases CaMKII␦C production and simultaneously reduces CaMKII␦B production, whereas inhibition of PP1␥ partially reverses this effect. Suppression of CaMKII was also found to attenuate the PP1␥-mediated promotion of apoptosis and Bax/ Bcl-2 ratio. Although we did not investigate the corresponding change in CaMKII␦ isoform protein because of the nonavailability of commercial antibodies that can recognize distinct ␦ isoforms, some studies in literature can be found to support the function of PP1␥ in CaMKII␦ splicing-mediated cardiomyocyte apoptosis. For example, our previous data show that cardiac CaMKII␦B and -␦C are indeed inversely regulated in response to ischemic injury and oxidative stress (25, 40). Additionally, several recent studies show that CaMKII␦C is a common mediator of cardiac apoptosis induced by diverse death-inducing stimuli such as intracellular high calcium (4) and excessive ␤1-adrenergic receptor stimulation (37, 41). In contrast to CaMKII␦C, CaMKII␦B serves as a potent suppressor of cardiac apoptosis triggered by the induction of iHSP70 (25). In our study, the increase in CaMKII␦B production was accompanied by the decrease in cardiac apoptosis. CaMKII␦A is mainly expressed in neonatal cardiomyocytes and little is detected in adult heart tissues. Therefore, it has received far less attention than CaMKII␦B and -␦C isoforms. Recent studies found that transgenic CaMKII␦A results in pathological cardiac hypertrophy (36) evidenced by the enlargement of cardiomyocytes and reactivation of various hypertrophic cardiac genes (15, 36). This indicates that CaMKII␦A is a potential target for the treatment of cardiac disorders. However, in our study we found that overexpression of PP1␥ reduces CaMKII␦A production in primary cardiomyocytes, whether the cells were OGD/R-treated or not. This finding suggests that in vitro CaMKII␦A might not participate in the OGD/Rinduced apoptosis, and PP1␥-induced reduction in CaMKII␦A production might have other functions in vitro. In conclusion, our findings indicate that PP1␥ is associated with the OGD/R-triggered cardiac apoptosis possibly by increasing CaMKII␦C production and reducing CaMKII␦B production, thus resulting in the disturbance of Bax/Bcl-2 balance. PP1␥ and its function in ASF-mediated CaMKII␦ splicing may provide important insights into the molecular mechanism of heart ischemia. ACKNOWLEDGMENTS We thank Dr. Jianhua Shi for technical assistance and reagents. GRANTS This work was supported by the National Natural Science Foundation of China to W. Zhang (No. 81070197) and C. Huang (No. 81102428) and the Priority Academic Program Development of Jiangsu Higher Education Institutions.

AJP-Cell Physiol • doi:10.1152/ajpcell.00145.2013 • www.ajpcell.org

C176

PP1␥ CONTROL OF CaMKII␦ SPLICING

DISCLOSURES No conflicts of interest, financial or otherwise, are declared by the author(s).

20.

AUTHOR CONTRIBUTIONS Author contributions: C.H., L.T., W. Zhu, and W. Zhang conception and design of research; C.H., W.C., R.L., J.W., Y.W., X.C., and W. Zhu performed experiments; C.H., W.C., R.L., and W. Zhu analyzed data; C.H., W.C., and W. Zhang interpreted results of experiments; C.H., W.C., and R.L. prepared figures; C.H., R.L., L.T., W. Zhu, and W. Zhang drafted manuscript; C.H., Y.W., L.T., X.C., and W. Zhu edited and revised manuscript; L.T. and W. Zhang approved final version of manuscript. REFERENCES

21.

22.

23.

1. Aoyama H, Ikeda Y, Miyazaki Y, Yoshimura K, Nishino S, Yamamoto T, Yano M, Inui M, Aoki H, Matsuzaki M. Isoform-specific roles of protein phosphatase 1 catalytic subunits in sarcoplasmic reticulum-mediated Ca2⫹ cycling. Cardiovasc Res 89: 79 –88, 2011. 2. Aubol BE, Chakrabarti S, Ngo J, Shaffer J, Nolen B, Fu XD, Ghosh G, Adams JA. Processive phosphorylation of alternative splicing factor/ splicing factor 2. Proc Natl Acad Sci USA 100: 12601–12606, 2003. 3. Brown AM, Baucum AJ, Bass MA, Colbran RJ. Association of protein phosphatase 1 gamma 1 with spinophilin suppresses phosphatase activity in a Parkinson disease model. J Biol Chem 283: 14286 –14294, 2008. 4. Chen X, Zhang X, Kubo H, Harris DM, Mills GD, Moyer J, Berretta R, Potts ST, Marsh JD, Houser SR. Ca2⫹ influx-induced sarcoplasmic reticulum Ca2⫹ overload causes mitochondrial-dependent apoptosis in ventricular myocytes. Circ Res 97: 1009 –1017, 2005. 5. Cohen PT. Two isoforms of protein phosphatase 1 may be produced from the same gene. FEBS Lett 232: 17–23, 1988. 6. Czabotar PE, Colman PM, Huang DC. Bax activation by bim? Cell Death Differ 16: 1187–1191, 2009. 7. Dombradi V, Axton JM, Brewis ND, da Cruz e Silva EF, Alphey L, Cohen PT. Drosophila contains three genes that encode distinct isoforms of protein phosphatase 1. Eur J Biochem 194: 739 –745, 1990. 8. Edman CF, Schulman H. Identification and characterization of deltaBCaM kinase and deltaC-CaM kinase from rat heart, two new multifunctional Ca2⫹/calmodulin-dependent protein kinase isoforms. Biochim Biophys Acta 1221: 89 –101, 1994. 9. Gavathiotis E, Suzuki M, Davis ML, Pitter K, Bird GH, Katz SG, Tu HC, Kim H, Cheng EH, Tjandra N, Walensky LD. Bax activation is initiated at a novel interaction site. Nature 455: 1076 –1081, 2008. 10. Gu Q, Jin N, Sheng H, Yin X, Zhu J. Cyclic AMP-dependent protein kinase a regulates the alternative splicing of CaMKII delta. PloS one 6: e25745, 2011. 11. Hagemann D, Bohlender J, Hoch B, Krause EG, Karczewski P. Expression of Ca2⫹/calmodulin-dependent protein kinase II delta-subunit isoforms in rats with hypertensive cardiac hypertrophy. Mol Cell Biochem 220: 69 –76, 2001. 12. Hagopian JC, Ma CT, Meade BR, Albuquerque CP, Ngo JC, Ghosh G, Jennings PA, Fu XD, Adams JA. Adaptable molecular interactions guide phosphorylation of the sr protein asf/sf2 by srpk1. J Mol Biol 382: 894 –909, 2008. 13. Han J, Ding JH, Byeon CW, Kim JH, Hertel KJ, Jeong S, Fu XD. Sr proteins induce alternative exon skipping through their activities on the flanking constitutive exons. Mol Cell Biol 31: 793–802, 2011. 14. Koizumi J, Okamoto Y, Onogi H, Mayeda A, Krainer AR, Hagiwara M. The subcellular localization of sf2/asf is regulated by direct interaction with sr protein kinases (srpks). J Biol Chem 274: 11125–11131, 1999. 15. Li C, Cai X, Sun H, Bai T, Zheng X, Zhou XW, Chen X, Gill DL, Li J, Tang XD. The ␦a isoform of calmodulin kinase II mediates pathological cardiac hypertrophy by interfering with the hdac4-mef2 signaling pathway. Biochem Biophys Res Commun 409: 125–130, 2011. 16. Ma CT, Ghosh G, Fu XD, Adams JA. Mechanism of dephosphorylation of the sr protein asf/sf2 by protein phosphatase 1. J Mol Biol 403: 386 –404, 2010. 17. Ma CT, Hagopian JC, Ghosh G, Fu XD, Adams JA. Regiospecific phosphorylation control of the sr protein asf/sf2 by srpk1. J Mol Biol 390: 618 –634, 2009. 18. Maier LS. Role of CaMKII for signaling and regulation in the heart. Front Biosci 14: 486 –496, 2009. 19. Mishra S, Gupta RC, Tiwari N, Sharov VG, Sabbah HN. Molecular mechanisms of reduced sarcoplasmic reticulum Ca2⫹ uptake in human

24.

25.

26.

27.

28.

29.

30.

31.

32.

33.

34.

35.

36.

37.

failing left ventricular myocardium. J Heart Lung Transplant 21: 366 – 373, 2002. Miyazaki Y, Ikeda Y, Shiraishi K, Fujimoto SN, Aoyama H, Yoshimura K, Inui M, Hoshijima M, Kasahara H, Aoki H, Matsuzaki M. Heart failure-inducible gene therapy targeting protein phosphatase 1 prevents progressive left ventricular remodeling. PloS one 7: e35875, 2012. Neumann J, Eschenhagen T, Jones LR, Linck B, Schmitz W, Scholz H, Zimmermann N. Increased expression of cardiac phosphatases in patients with end-stage heart failure. J Mol Cell Cardiol 29: 265–272, 1997. Ngo JC, Chakrabarti S, Ding JH, Velazquez-Dones A, Nolen B, Aubol BE, Adams JA, Fu XD, Ghosh G. Interplay between srpk and clk/sty kinases in phosphorylation of the splicing factor asf/sf2 is regulated by a docking motif in asf/sf2. Mol Cell 20: 77–89, 2005. Nicolaou P, Kranias EG. Role of PP1 in the regulation of Ca2⫹ cycling in cardiac physiology and pathophysiology. Front Biosci 14: 3571–3585, 2009. Nicolaou P, Rodriguez P, Ren X, Zhou X, Qian J, Sadayappan S, Mitton B, Pathak A, Robbins J, Hajjar RJ, Jones K, Kranias EG. Inducible expression of active protein phosphatase-1 inhibitor-1 enhances basal cardiac function and protects against ischemia/reperfusion injury. Circ Res 104: 1012–1020, 2009. Peng W, Zhang Y, Zheng M, Cheng H, Zhu W, Cao CM, Xiao RP. Cardioprotection by CaMKII-deltaB is mediated by phosphorylation of heat shock factor 1 and subsequent expression of inducible heat shock protein 70. Circ Res 106: 102–110, 2010. Prasad J, Colwill K, Pawson T, Manley JL. The protein kinase clk/sty directly modulates sr protein activity: Both hyper- and hypo-phosphorylation inhibit splicing. Mol Cell Biol 19: 6991–7000, 1999. Qian J, Vafiadaki E, Florea SM, Singh VP, Song W, Lam CK, Wang Y, Yuan Q, Pritchard TJ, Cai W, Haghighi K, Rodriguez P, Wang HS, Sanoudou D, Fan GC, Kranias EG. Small heat shock protein 20 interacts with protein phosphatase-1 and enhances sarcoplasmic reticulum calcium cycling. Circ Res 108: 1429 –1438, 2011. Sasaki K, Shima H, Kitagawa Y, Irino S, Sugimura T, Nagao M. Identification of members of the protein phosphatase 1 gene family in the rat and enhanced expression of protein phosphatase 1 alpha gene in rat hepatocellular carcinomas. Jpn J Cancer Res 81: 1272–1280, 1990. Shi J, Qian W, Yin X, Iqbal K, Grundke-Iqbal I, Gu X, Ding F, Gong CX, Liu F. Cyclic AMP-dependent protein kinase regulates the alternative splicing of tau exon 10: a mechanism involved in tau pathology of Alzheimer’s disease. J Biol Chem 286: 14639 –14648, 2011. Shi J, Zhang T, Zhou C, Chohan MO, Gu X, Wegiel J, Zhou J, Hwang YW, Iqbal K, Grundke-Iqbal I, Gong CX, Liu F. Increased dosage of dyrk1a alters alternative splicing factor (asf)-regulated alternative splicing of tau in down syndrome. J Biol Chem 283: 28660 –28669, 2008. Shintani-Ishida K, Yoshida K. Ischemia induces phospholamban dephosphorylation via activation of calcineurin, pkc-alpha, and protein phosphatase 1, thereby inducing calcium overload in reperfusion. Biochim Biophys Acta 1812: 743–751, 2011. Srinivasan M, Edman CF, Schulman H. Alternative splicing introduces a nuclear localization signal that targets multifunctional cam kinase to the nucleus. J Cell Biol 126: 839 –852, 1994. Wang R, Kaul A, Sperry AO. Tlrr (lrrc67) interacts with PP1 and is associated with a cytoskeletal complex in the testis. Biol Cell 102: 173–189, 2010. Wang RH, Liu CW, Avramis VI, Berndt N. Protein phosphatase 1 alpha-mediated stimulation of apoptosis is associated with dephosphorylation of the retinoblastoma protein. Oncogene 20: 6111–6122, 2001. Winding C, Sun Y, Hoger H, Bubna-Littitz H, Pollak A, Schmidt P, Lubec G. Serine/threonine-protein phosphatase 1 alpha levels are paralleling olfactory memory formation in the cd1 mouse. Electrophoresis 32: 1675–1683, 2011. Xu X, Yang D, Ding JH, Wang W, Chu PH, Dalton ND, Wang HY, Bermingham JR Jr, Ye Z, Liu F, Rosenfeld MG, Manley JL, Ross J Jr, Chen J, Xiao RP, Cheng H, Fu XD. Asf/sf2-regulated camkiidelta alternative splicing temporally reprograms excitation-contraction coupling in cardiac muscle. Cell 120: 59 –72, 2005. Yang Y, Zhu WZ, Joiner ML, Zhang R, Oddis CV, Hou Y, Yang J, Price EE, Gleaves L, Eren M, Ni G, Vaughan DE, Xiao RP, Anderson ME. Calmodulin kinase II inhibition protects against myocardial cell apoptosis in vivo. Am J Physiol Heart Circ Physiol 291: H3065–H3075, 2006.

AJP-Cell Physiol • doi:10.1152/ajpcell.00145.2013 • www.ajpcell.org

PP1␥ CONTROL OF CaMKII␦ SPLICING 38. Zhang T, Johnson EN, Gu Y, Morissette MR, Sah VP, Gigena MS, Belke DD, Dillmann WH, Rogers TB, Schulman H, Ross J Jr, Brown JH. The cardiac-specific nuclear deltaB isoform of Ca2⫹/calmodulin-dependent protein kinase II induces hypertrophy and dilated cardiomyopathy associated with increased protein phosphatase 2a activity. J Biol Chem 277: 1261–1267, 2002. 39. Zhang T, Maier LS, Dalton ND, Miyamoto S, Ross J Jr, Bers DM, Brown JH. The deltac isoform of CaMKII is activated in cardiac hypertrophy and induces dilated cardiomyopathy and heart failure. Circ Res 92: 912–919, 2003.

C177

40. Zhu W, Woo AY, Yang D, Cheng H, Crow MT, Xiao RP. Activation of CaMKII deltaC is a common intermediate of diverse death stimuliinduced heart muscle cell apoptosis. J Biol Chem 282: 10833–10839, 2007. 41. Zhu WZ, Wang SQ, Chakir K, Yang D, Zhang T, Brown JH, Devic E, Kobilka BK, Cheng H, Xiao RP. Linkage of beta1-adrenergic stimulation to apoptotic heart cell death through protein kinase a-independent activation of Ca2⫹/calmodulin kinase II. J Clin Invest 111: 617–625, 2003.

AJP-Cell Physiol • doi:10.1152/ajpcell.00145.2013 • www.ajpcell.org