1521-0111/92/6/718–730$25.00 https://doi.org/10.1124/mol.117.109645MOLECULAR PHARMACOLOGY Mol Pharmacol 92:718–730, December 2017Copyright ª 2017 by The American Society for Pharmacology and Experimental Therapeutics

Hydrogen Sulfide Preserves Endothelial Nitric Oxide SynthaseFunction by Inhibiting Proline-Rich Kinase 2: Implications forCardiomyocyte Survival and Cardioprotection s

Sofia-Iris Bibli, Csaba Szabo, Athanasia Chatzianastasiou, Bert Luck, Sven Zukunft,Ingrid Fleming, and Andreas PapapetropoulosLaboratory of Pharmacology, Faculty of Pharmacy, National and Kapodistrian University of Athens, Athens, Greece (S.-I.B., A.P.);Institute for Vascular Signalling, Centre for Molecular Medicine, Goethe University, Frankfurt am Main, Germany (S.-I.B., B.L.,S.Z., I.F.); Department of Anesthesiology, University of Texas Medical Branch, Galveston, Texas (C.S.); “George P. Livanos andMarianthi Simou” Laboratories, First Department of Pulmonary and Critical Care Medicine, Evangelismos Hospital, Faculty ofMedicine, National and Kapodistrian University of Athens, Athens, Greece (A.C.); and Clinical, Experimental Surgery andTranslational Research Center, Biomedical Research Foundation of the Academy of Athens, Athens, Greece (A.P.)

Received June 13, 2017; accepted October 11, 2017

ABSTRACTHydrogen sulfide (H2S) exhibits beneficial effects in the car-diovascular system, many of which depend on nitric oxide (NO).Proline-rich tyrosine kinase 2 (PYK2), a redox-sensitive tyrosinekinase, directly phosphorylates and inhibits endothelial NOsynthase (eNOS). We investigated the ability of H2S to relievePYK2-mediated eNOS inhibition and evaluated the importanceof the H2S/PYK2/eNOS axis on cardiomyocyte injury in vitroand in vivo. Exposure of H9c2 cardiomyocytes to H2O2 orpharmacologic inhibition of H2S production increased PYK2(Y402) and eNOS (Y656) phosphorylation. These effects wereblocked by treatment with Na2S or by overexpression ofcystathionine g-lyase (CSE). In addition, PYK2 overexpressionreduced eNOS activity in a H2S-reversible manner. Theviability of cardiomyocytes exposed to Η2Ο2 was reducedand declined further after the inhibition of H2S production.

PYK2 downregulation, L-cysteine supplementation, or CSEoverexpression alleviated the effects of H2O2 on H9c2 cardio-myocyte survival. Moreover, H2S promoted PYK2 sulfhydrationand inhibited its activity. In vivo, H2S administration reducedreactive oxygen species levels, as well as PYK2 (Y402) andeNOS (Y656) phosphorylation. Pharmacologic blockade ofPYK2 or inhibition of PYK2 activation by Na2S reduced myo-cardial infarct size in mice. Coadministration of a PYK2 inhibitorand Na2S did not result in additive effects on infarct size. Weconclude that H2S relieves the inhibitory effect of PYK2 oneNOS, allowing the latter to produce greater amounts of NO,thereby affording cardioprotection. Our results unravel theexistence of a novel H2S-NO interaction and identify PYK2 asa crucial target for the protective effects of H2S underconditions of oxidative stress.

IntroductionHydrogen sulfide (H2S) has emerged as an important gaseous

signaling molecule in mammalian cells regulating a multitudeof basic biologic processes, including bioenergetics, prolifera-tion, apoptosis, and necrosis (Mustafa et al., 2009; Li et al.,2011; Szabó and Papapetropoulos, 2011; Wang, 2012; Módiset al., 2014). Endogenous H2S is produced by three enzymes—namely, cystathionine g-lyase (CSE), cystathionine b-synthase,and 3-mercaptopyruvate sulfur transferase (Kimura, 2011;

Kabil and Banerjee, 2014; Papapetropoulos et al., 2015).Although all three enzymes are expressed in the cardiovascularsystem, existing data suggest that CSE plays a major role incardiovascular physiology (Wang, 2012; Polhemus and Lefer,2014; Katsouda et al., 2016). CSE exerts angiogenic(Papapetropoulos et al., 2009), hypotensive (Yang et al.,2008), cardioprotective (Elrod et al., 2007; Bibli et al., 2015a),as well as antioxidant and anti-inflammatory effects in themyocardium and the vessel wall (Calvert et al., 2009; Kimura,2011; Szabó et al., 2011; Shibuya et al., 2013; Salloum, 2015;Kanagy et al., 2017). Reduced generation or increased break-down of H2S leads to lower levels of this gasotransmitter and isassociated with several cardiovascular pathologies and condi-tions, such as endothelial dysfunction, atherosclerosis, hyper-tension, heart failure, and preeclampsia (Polhemus and Lefer,2014;Wang et al., 2015; Greaney et al., 2017; Kanagy et al., 2017).

This work was supported by European Union FP7 REGPOT CT-2011-285950 [SEE-DRUG], by the Cooperation in Science and Technology COSTAction BM1005 [ENOG: European network on gasotransmitters], and by theDeutsche Forschungsgemeinschaft [SFB 834/A9].

https://doi.org/10.1124/mol.117.109645.s This article has supplemental material available at molpharm.

aspetjournals.org.

ABBREVIATIONS: ANOVA, analysis of variance; AOAA, aminoxyacetic acid; CSE, cystathionine-g lyase; DHE, dihydroethidium; DMEM, Dulbecco’smodified Eagle’s medium; eNOS, endothelial nitric oxide synthase; FBS, fetal bovine serum; GFP, green fluorescent protein; HEK, human embryonickidney cell line; H2S, hydrogen sulfide; KO, knockout; LAD, left anterior descending coronary artery; MDA, malondialdehyde; MOI, multiplicity ofinfection; MTT, 3-(4, 5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide; NO, nitric oxide; PBS, phosphate-buffered saline; PC, protein carbonyls;PYK2, proline-rich kinase 2; ROS, reactive oxygen species; Scrsi, scrambled small interfering RNA; Sol, solvent; TCA, trichloroacetic acid.

718

http://molpharm.aspetjournals.org/content/suppl/2017/10/13/mol.117.109645.DC1Supplemental material to this article can be found at:

at ASPE

T Journals on N

ovember 16, 2021

molpharm

.aspetjournals.orgD

ownloaded from

Studies from several laboratories have proven that endoge-nously produced and exogenously administered H2S limitischemia-reperfusion injury and reduces infarct size inisolated hearts and in vivo (Johansen et al., 2006; Pan et al.,2006; Elrod et al., 2007; Calvert et al., 2009; Szabó et al., 2011;King et al., 2014; Polhemus et al., 2014; Polhemus and Lefer,2014; Bibli et al., 2015a; Das et al., 2015).To exert its biologic responses, H2S uses a variety of

signaling pathways by regulating the activity of kinases,phosphatases, transcription factors, and ion channels(Szabó, 2007; Paul and Snyder, 2012; Wang, 2012;Polhemus and Lefer, 2014; Kanagy et al., 2017). Many ofthese actions are attributed to a post-translational modifi-cation of cysteine residues in a modification referred to assulfhydration or persulfidation (Paul and Snyder, 2012). Inaddition, some biologic effects exerted by H2S require nitricoxide (NO) production. Angiogenesis, vasodilation, andcardioprotection are reduced or blunted when endothelialNO synthase (eNOS) is inhibited (Coletta et al., 2012; Kinget al., 2014; Bibli et al., 2015a). At the molecular level, theH2S-NO interaction involves increased eNOS phosphoryla-tion at the activator site S1177 (Minamishima et al., 2009;Papapetropoulos et al., 2009; Coletta et al., 2012; Altaanyet al., 2013; Kondo et al., 2013; King et al., 2014; Bibli et al.,2015a; Chatzianastasiou et al., 2016; Karwi et al., 2017) andreduced phosphorylation at the T495 inhibitory site(Coletta et al., 2012; Polhemus et al., 2013; Bibli et al.,2015b). Moreover, H2S promotes eNOS dimerization andcoupling through the sulfhydration of C443 (Altaany et al.,2014). The aforementioned post-translational modifica-tions of eNOS enhance NO production and/or bioavailabil-ity after exposure to H2S, as evidenced by increases incGMP accumulation or NO metabolite levels (Coletta et al.,2012; Predmore et al., 2012; Kondo et al., 2013; King et al.,2014; Szabo, 2017). With respect to cardioprotection, theimportance of S1176 phosphorylation in vivo was demon-strated using S1176A knock-in mice in which H2S donoradministration was ineffective in limiting infarct size(King et al., 2014).The proline-rich tyrosine kinase 2 (PYK2) is a redox-

sensitive kinase (Lev et al., 1995; Tokiwa et al., 1996; Taiet al., 2002; Chappell et al., 2008; Loot et al., 2009) that hasbeen linked to cardiac remodeling (Takeishi, 2014), hyper-trophic responses (Hirotani et al., 2004), dilated cardiomy-opathy (Koshman et al., 2014), and ischemia reperfusioninjury (Fisslthaler et al., 2008). Recently, we demonstratedthat PYK2 directly phosphorylates eNOS on Y657 (humaneNOS sequence; corresponds to murine Y656), rendering itinactive (Fisslthaler et al., 2008; Loot et al., 2009). PYK2 isactivated in the early minutes of myocardial reperfusionafter ischemia, resulting in increased phosphorylation ofeNOS on the Y656 inhibitory residue and reduced NOoutput; this mechanism defines myocardial infarct size,and PYK2 is proposed to serve as a novel therapeutic targetfor cardioprotection (Bibli et al., 2017). Since H2S is known topossess antioxidant properties, we hypothesized that it couldblock oxidative stress-induced PYK2 activation in earlyreperfusion, limiting eNOS inhibition and reducing myocar-dial infarct size. Data from the current study indicate thatH2S restrains the activation of PYK2, providing a novelmechanism of positive interaction between H2S and NO incardiomyocytes.

Materials and MethodsChemicals and Reagents

All chemicals and reagents— including aminoxyacetic acid (AOAA),L-cysteine, Na2S, PF-431396, Triton �100, NaCl, NaF, EDTA, EGTA,phenylmethyl sulfonyl fluoride, protease and phosphatase inhibitors,glycerol phosphatase, and MTT were purchased from Sigma-Aldrich(Taufkirchen, Germany); DMSO, H2O2, Tris, and SDSwere purchasedfrom AppliChem (Bioline Scientific, Athens, Greece). LipofectAMINERNAiMAX was obtained from Invitrogen (Antisel, Athens, Greece);Dulbecco’s modified Eagle’s medium (DMEM), sodium pyruvate,antibiotics, Opti-MEM, and fetal bovine serum (FBS) wereobtained from Gibco (Antisel); the cGMP kit was obtained from Enzo(Zafeiropoulos SA, Athens, Greece); Electrogenerated chemilumines-cence was obtained from Thermoscientific (Bioanalytica, Athens,Greece). The pPYK2, PYK2, eNOS, b-tubulin, nitrotyrosine, andsecondary antibodies were purchased from Cell Signaling Technol-ogies (Danvers, MA); anti-CSE was from Proteintech Europe(Manchester, UK); the peNOS(Y657) was generated by Eurogentec(Köln, Germany). SSP4 was purchased from Dojindo EU (Munich,Germany). The Amplex red assay kit was purchased from Invitrogen(Thermo Fischer Scientific, Darmstadt, Germany). The ADP-Glo kinaseassay kit was obtained from Promega (Mannheim, Germany).

Cell Culture

The rat embryonic heart–derived H9c2 cell line was obtained fromAmerican Type Culture Collection (CRL-1446) (ATCC, LGC Stan-dards, Middlesex, UK). H9c2 cells were cultured in DMEM containing25 mM D-glucose and 1 mM sodium pyruvate, supplemented with10% FBS, 2 mM L-glutamine, 1% streptomycin (100 mg/ml), and1% penicillin (100U/ml) at pH 7.4 in a 5%CO2 incubator at 37°C. Cellswere maintained in a subconfluent condition of a maximum of 70%before passaging to avoid differentiation. For differentiation, H9c2cells were seeded and allowed to grow to confluence. The medium wasthen replaced to DMEM containing 1% FBS with 10 nM all-trans-retinoic acid for 7 days. After 7 days, cells were elongated, connectingat irregular angles, and cardiac differentiation markers such ascardiac troponin and MLC2v transcripts were elevated, reminiscentof cells with a cardiac phenotype, as described before (Bibli et al.,2017). Human embryonic kidney cells (HEK) cells were cultured inMEM supplemented with 1mM sodium pyruvate and 10%FBS, 2 mML-glutamine, 1% streptomycin (100 mg/ml), and 1% penicillin(100 U/ml) at pH 7.4 in a 5% CO2 incubator at 37°C.

In Vitro Treatments

For in vitro experiments, differentiated H9c2 cells were pretreatedwith the appropriate drug as follows. Single treatments were asfollows: AOAA was used at a concentration of 1 mM for 45 minutes,L-cysteine at 500 mM for 45 minutes, PF-431396 5 mM for 45 minutes,and Na2S 100 mM for 30 minutes. For double treatments, AOAA andL-cysteine in the previously mentioned concentrations were addedsimultaneously; PF-431396 was added 15 minutes before Na2S. Aftertreatment, cells were lysed for biochemical analysis or subsequentcell-viability studies. To induce in vitro oxidative stress injury, H9c2cells (1.5 � 104/well) were treated with 500 mM H2O2 in serum-freeDMEM for 12 hours in a 5% CO2 incubator at 37°C.

Small Interfering RNA-Mediated Downregulation of PYK2

H9c2 cells were seeded until 80% confluence. Transient transfectionof small interfering RNA (siRNA) (100 nM) was performed usingLipofectAMINE RNAiMAX according to the manufacturer’s instruc-tions. The transfection complex was diluted into Opti-MEM mediumand added directly to the cells. After 24 hours, the Opti-MEM wasreplaced with complete DMEM medium with 10% FBS for cellviability assays or with serum-free DMEM for biochemical studies.

PYK2, H2S, eNOS, and Cardioprotection 719

at ASPE

T Journals on N

ovember 16, 2021

molpharm

.aspetjournals.orgD

ownloaded from

The efficacy of siRNA PYK2 gene knockdown, 48 hours post-transfection, has been previously confirmed (Bibli et al., 2017).

Adenoviral Infections

H9c2 cells were seeded until confluence and differentiated for7 days. On day 6 of differentiation, cells were infected with green

fluorescent protein (GFP) or CSE adenoviruses (Bucci et al., 2010) at10 MOI for 36 hours.

MTT Measurements

After oxidative stress injury, cell survival was assessed in differen-tiated H9c2 cells by using the conversion of MTT (3-(4, 5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide) to formazan.

Fig. 1. Endogenous production of H2S regulates PYK2/eNOS phosphorylation in cardiomyocytes. Differentiated H9c2 were treated with 50 mMH2O2 for10 minutes and lysed, and the proteins were subjected to SDS-PAGE. Representative Western blots, along with densitometric for (A) pPYK2(Y402) and(B) peNOS(Y656) of cells treated with AOAA 1 mM for 45 minutes or L-cysteine 500 mM for 45 minutes. Cells infected with GFP Ad or CSE Ad (10 MOI,36 hours before H2O2) were evaluated for (C) pPYK2(Y402) and (D) peNOS(Y656). Phosphorylated protein levels were normalized to total protein levels;n = 5 independent experiments; *P , 0.05; **P , 0.001; ***P , 0.0001 (two-way ANOVA, Bonferroni).

720 Bibli et al.

at ASPE

T Journals on N

ovember 16, 2021

molpharm

.aspetjournals.orgD

ownloaded from

PYK2, H2S, eNOS, and Cardioprotection 721

at ASPE

T Journals on N

ovember 16, 2021

molpharm

.aspetjournals.orgD

ownloaded from

Cells were incubated withMTT at a final concentration of 0.5 mg/ml for2 hours at 37°C. The formazan formed was dissolved in solubilizationsolution (10% Triton-X 100 in acidic 0.1 N HCl in isopropanol);subsequently, absorbance was measured at 595 nm with a backgroundcorrection at 750 nm using a microplate reader.

Western Blot Analysis

H9c2 cells seeded in six-well plates until confluence, treatedas described already, were washed twice with phosphate-bufferedsaline (PBS) and further lysed with lysis solution (1% Triton �100,20 mM Tris pH 7.4–7.6, 150 mM NaCl, 50 mM NaF, 1 mM EDTA,1 mM EGTA, 1 mM glycerolphosphatase, 1% SDS, and 100 mMphenylmethyl sulfonyl fluoride, supplemented with protease andphosphatase inhibitor cocktail. Frozen ischemic samples were pulver-ized and homogenized with the lysis buffer. The lysates werecentrifuged at 11,000g for 15 minutes at 4°C. The supernatants werecollected, and the protein concentration was determined based on theLowry assay. The supernatant was mixed with a buffer containing4% SDS, 10% 2-mercaptoethanol, 20% glycerol, 0.004% bromophenylblue, 0.125 M Tris/HCl. The samples were then heated at 100°C for10 minutes and stored at 280°C. An equal amount of protein wasloaded in each well and then separated by SDS-PAGE electrophoresisand transferred onto a polyvinylidene difluoride membrane. Afterblocking with 5% nonfat dry milk, membranes were incubated over-night at 4°C with primary antibody. The following primary antibodieswere used: phospho PYK2 (Y402), phosphor-eNOS (Y656), total PYK2,total eNOS, nitrotyrosine, and b-tubulin. Membranes were thenincubated with secondary goat anti-rabbit HRP antibody for 2 hoursat room temperature and developed using the Supersignal electro-generated chemiluminesence Western blotting detection reagents.Relative densitometry was determined using a computerized softwarepackage (National Institutes of Health Image), and the phosphory-lated values were normalized to the values for total proteins re-spectively. All the presented total proteinswere derived from the samegel/experiment as the phosphoprotein presented after stripping of themembrane.

eNOS Activity Measurements

The following plasmids were used: myc-tagged human eNOS cDNA(GenBank accession no. NM_000603) cloned in pcDNA3, 1myc/His,PYK2 cDNA (NC_000008.11), and a dominant negative PYK2mutant(Fisslthaler et al., 2008). For the transfection experiments, HEK293cells were plated in 12-well plates, grown overnight, and transfectedwith the indicated plasmids using a total of 2 mg DNA and 4 ml ofLipofectAMINE transfection reagent per well in Opti-MEM medium.After 24 hours, cells were used for liquid chromatography-massspectrometry (LC-MS) and Western blot measurements. For Westernblot studies, cells were starved for 4 hours in serum-free mediumwiththe addition of 0.1% of bovine serum albumin. For NOmeasurements,HEKcells were treatedwith sepiapterin, 10mmol for 2 hours. Argininedepletion was performed by adding stable isotope labeling with aminoacids in cell culture medium for 12 hours of pretreatment. Heavylabeled Arg (13C6, 15N4) was added for 2 hours before collection foractivity measurements. Cells were immediately emerged in liquidnitrogen, and 85% high-performance liquid chromatography–gradeMeOH was added for cell disruption. The lysates were centrifuged in12,000g for 15 minutes in 4°C. Fifty microliters of supernatant wasmixed with 50 ml of MeOH containing 1 mg/ml glutamine as internal

standard. Conversion of heavy labeled Arg (13C6, 15N4) to heavycitrulline (13C6, 15N3) was analyzed by hydrophilic interactionchromatography coupled to MS. LC separation was performed on anAgilent 1290 Infinity pump system (Agilent, Waldbronn, Germany)using a Phenomenex (Aschaffenburg, Germany) Kinetex HILICcolumn (100 mm � 2.1 mm, 2.6 mm) at a column oven temperatureof 35°C. The gradient between solvent A (10 mM ammonium formate)and solvent B (acetonitrile 1 0.15% formic acid) was as follows:0.0–0.5 minute 10% A, then increased to 85% A for 0.5–6 minutes,increased to 98% A for 6.1 minutes, and held until 13.6 minutes.Subsequently, the column was reconditioned with 90% A for4.4 minutes. The flow rate was set to 350 ml/min, and the injectionvolume was 2.5 ml. MS was performed using a QTrap 5500 massspectrometer (Sciex, Darmstadt, Germany) with electrospray ioniza-tion at 300°Cwith 3500 V in positivemode.MS parameters were set toCUR 30 psi, GS1 50 psi, and GS2 40 psi. Data acquisition andinstrument control were managed through the software Analyst 1.6.2.Peak integration, data processing, and analyte quantification wereperformed usingMultiQuant 3.0 (Sciex). The area under the peak wasused as the quantitative measurement. The specific MRM transitionfor every compound was normalized to appropriated isotope labeledinternal standards.

H2S Measurements

Intracellular levels of H2S were measured by monitoring thereaction of SSP4 with H2S. In brief, cells were seeded in 12- or48-well plates and allowed to reach confluence. The culture mediumwas replaced with phenol red-free DMEM supplemented with 0.1%bovine serum albumin. For inhibition of endogenous H2S production,cells were pretreated with AOAA (1 mM) for 45 minutes. To enhanceH2S production, cells were incubated with L-cysteine (500 mM) for45 minutes. Subsequently, medium was replaced and SSP4(10 mmol/liter) was added for 60 minutes. Thereafter, the cell super-natant was collected, and floating cells were removed by centrifuga-tion (16,000g, 10minutes, 4°C). The specific products of the reaction ofH2S with SPP4 were quantified by LC-MS/MS.

Oxidatives Stress Detection in Cardiomyocytes. Hydrogenperoxide levels weremeasured in cardiomyocytes by using the AmplexRed Assay Kit according to the manufacturer’s instructions. In brief,differentiatedH9c2 cells were infectedwith aGFP- or CSE-expressingadenovirus or treated with the Na2S salt as described. In some wells,50 mM H2O2 was added to the cells for 10 minutes. The reaction wasstopped on ice. Cells were collected and washed three times with ice-cold PBS to wash out the exogenous H2O2; 10

6 cells per condition wereused for H2O2 determination. A standard curve for H2O2 was used toquantify the endogenously produced H2O2.

PYK2 Activity Assay

The effects of H2S on PYK2 activity were determined using theADP-Glo kinase assay kit. The assay was performed in the presence ofsolvent or different concentrations of Na2S on purified PYK2 proteinaccording to the manufacturer’s instructions.

S-Sulfhydration Detection

Sulfhydration was detected using a modified biotin switch assay. Inbrief, H9c2- differentiated cells were treated with Na2S (100 mM) for30 minutes. Reactions were stopped on ice, and cells were washedwith ice-cold PBS. Subsequently, samples were precipitated with

Fig. 2. Exogenous Na2S reduces PYK2 activation and eNOS phosphorylation on Y656/7 during oxidative stress injury. Differentiated H9c2 cells weretreated with Na2S 100 mM for 20minutes before H2O2. Subsequently, H2O2 (50 mM for 10minutes) was applied. RepresentativeWestern blots along withdensitometric analysis of (A) pPYK2(Y402) and (B) peNOS(Y656). Phosphorylated protein levels were normalized to total proteins; n = 5 independentexperiments; **P , 0.001; ***P , 0.0001. (C) HEK cells were transfected with WT human eNOS, with or without WT PYK2 or a kinase-dead PYK2mutant. Representative Western blots for peNOS (Y657), eNOS, and PYK2. Phosphorylated eNOS levels were normalized to total eNOS levels; n =4 independent experiments; **P , 0.001; ***P , 0.0001. (two-way ANOVA, Bonferroni).

722 Bibli et al.

at ASPE

T Journals on N

ovember 16, 2021

molpharm

.aspetjournals.orgD

ownloaded from

20% trichloroacetic acid (TCA) and stored at280°C. TCA precipitateswere washed with 10% and 5% TCA and then centrifuged (16,000g,30 minutes, 4°C) before being suspended in HENs buffer (250 mmol/literHEPES-NaOH, 1 mmol/liter EDTA, 0.1 mmol/liter neocuproine,100 mmol/liter deferoxamine, 2,5% SDS) containing 20 mmol/litermethanethiosulfonate to block free thiols and protease and phospha-tase inhibitors. Acetone precipitation was performed, and pellets wereresuspended in 300 ml qPerS-SID lysis buffer (6 mol/liter urea,100 mmol/liter NaCl, 2% SDS, 5 mmol/liter EDTA, 200 mmol/literTris pH 8.2; 50 mmol/liter iodoacetyl-PEG2-biotin, 2.5 mmol/literdimedone), sonified, and incubated for 2 hours at room temperature inthe dark. Lysates (500mg) were precipitated with acetone, and proteinpellets were resuspended in 50 ml Tris/HCl (50 mmol/liter, pH 8.5)containing guanidinium chloride (GdmCl 6mmol/liter), and incubatedat 95°C for 5 minutes. A negative control was generated for eachsample by adding dithiothritol (1 mmol/liter) during biotin cross-linking. Biotin was then immunoprecipitated using a high-capacitystreptavidin resin (Thermo Scientific, Heidelberg, Germany) over-night at 4°C. Elution was performed by the addition of 3% SDS, 1%b-mercaptoethanol, 8 mol/liter urea, and 0.005% bromophenol blue inPBS for 15 minutes at room temperature, followed by 15 minutes at95°C. Sulfhydrated proteins were detected after SDS-PAGE byWestern blotting.

cGMP Enzyme Immunoassay

Cyclic nucleotides were extracted by HCl and measured using acommercially available enzyme immunoassay kit (Enzo Life Sciences)according to the manufacturer’s instructions. For tissue samples,frozen ischemic tissue was pulverized. Powdered samples from myo-cardial ischemic tissue were lysed with 0.1 N HCl (1:5 v/w) to extractcGMP, the content of which was measured using enzyme immunoas-say according to the manufacturer’s instructions. Protein concentra-tionwas determined by the Lowrymethod, and results were expressedas picomoles cGMP/mg protein.

Malondialdehyde and Protein Carbonylation Assessment

Tissue sampleshomogenateswere used tomeasuremalondialdehyde(MDA) and protein carbonyls (PC). MDA was determined spectro-photometrically as previously described (Andreadou et al., 2014). Aspectrophotometric measurement of 2,4-dinitrophenylhydrazine de-rivatives of PC was used to quantify PC content as previouslydescribed (Andreadou et al., 2014).

Dihydroethidium Staining of Cardiac Reactive OxygenSpecies Formation

Cardiac reactive oxygen species (ROS) production was qualitativelydetected by dihydroethidium (DHE) (1 mM)-derived fluorescence inheart tissue cryosections of 8 mm as described previously (Andreadouet al., 2014).

Animals

All animal procedures complied with the European Communityguidelines for the use of experimental animals; experimental protocolswere approved by the Ethical Committee of the Prefecture of Athens(790/2014). Animals received standard rodent laboratory diet. In thepresent study, we used male mice C57BL/6J mice.

Surgical Procedures

Murine In VivoModel of Ischemia-Reperfusion Injury. Malemice 10–12 weeks old were anesthetized by intraperitoneal injectionwith a combination of ketamine, xylazine, and atropine (0.01 ml/g,final concentrations of ketamine, xylazine, and atropine 10 mg/ml,2 mg/ml, 0.06 mg/kg, respectively). A tracheotomy was performed forartificial respiration at 120–150 breaths/min and PEEP 2.0 (0.2 mltidal volume) (Flexivent rodent ventilator; Scireq, Montreal, QC).

Electrocardiogram recordings were performed by a lead I ECGrecording with PowerLab 4.0 (ADInstruments, Sydney, Australia).Recordings were analyzed by LabChart 7.0 software. A thoracotomywas then performed between the fourth and fifth ribs, and thepericardium was carefully retracted to visualize the left anteriordescending coronary artery (LAD), which was ligated using a 8-0Prolene (Ethicon, Somerville, NJ)monofilament polypropylene sutureplaced 1 mm below the tip of the left ventricle. The heart wasstabilized for 15 minutes before ligation to induce ischemia. Afterthe ischemic period, the ligaturewas released and allowed reperfusionof the myocardium. Throughout the experiments, body temperaturewas maintained at 37 6 0.5°C by way of a heating pad. Afterreperfusion, hearts were rapidly excised from mice and directlycannulated and washed with 2.5 ml of saline-heparin 1% for bloodremoval. Five milliliters of 1% TTC phosphate buffer 37°C wereinfused via the cannula into the coronary circulation, followed byincubation of the myocardium for 5 minutes in the same buffer; 2.5 mlof 1% Evans blue, diluted in distilled water, was then infused into theheart. Hearts were kept in220°C for 24 hours, sliced in 1-mm sectionsparallel to the atrioventricular groove, and then fixed in 4% formal-dehyde overnight. Slices were then placed between glass plates 1 mmapart and photographed with a Cannon Powershot A620 DigitalCamera through Zeiss 459300 microscope and measured with theScion Image program. Infarct and risk area volumeswere expressed incubic centimeters, and the percentage of infarct-to-risk area ratio(percentage of ischemia/reperfusion) was calculated.

Experimental Protocol

In animals treated with the pharmacologic inhibitor of PYK2,PF-431396 was administrated at 5 mg/g in 2% DMSO (100 ml) i.v.,and Na2S was given as an i.v. bolus at 100 mg/kg as describedpreviously (Bibli et al., 2015a). In the first experimental series,animals received the indicated drugs either 10 minutes beforesacrifice for tissue collection (sham-operated groups) or 10 minutesbefore reperfusion (ischemia-reperfusion injury groups). The leftventricle was isolated and submerged in liquid nitrogen for preserva-tion before analysis.

Fig. 3. Na2S increases eNOS activity. Heavy citrulline to heavy arginineratio of supernatants from cells treated as described in (Fig. 2C). eNOSactivity was assessed by its ability to produce heavy citrulline on depletionof arginine for 12 hours and the addition of heavy arginine for 2 hours.Results were analyzed by LC-MS measurements; n = 4 independentexperiments; **P , 0.001 (two-way ANOVA, Bonferroni).

PYK2, H2S, eNOS, and Cardioprotection 723

at ASPE

T Journals on N

ovember 16, 2021

molpharm

.aspetjournals.orgD

ownloaded from

In a second series of experiments,micewere subjected to 30minutesof regional ischemia of the myocardium, followed by 2 hours ofreperfusion, and randomized into four groups as follows: 1) Sol group(n 5 8): administration of solvent [water for injection containing2% DMSO (100 ml) i.v. 10 minutes before the ischemic insult];2) PF-431396 group (n 5 8): administration of 5 mg/g PF-431396[dissolved in water for injection containing 2% DMSO; (100 ml) iv10 minutes before reperfusion; 3) Na2S group (n 5 8): administrationof 100 mg/kg Na2S (100 ml i.v. 10 minutes before reperfusion), and4) PF-4313961Na2S group (n5 6): administration of PF-431396 andNa2S as in groups 2 and 3.

Statistical Analysis

One- or two-way analysis of variance (ANOVA) was used to detectthe differences between multiple groups or unpaired two-tailed

Student’s t test to compare two groups. A value of P , 0.05 wasconsidered statistically significant. All statistical calculations wereperformed using Prism 4 analysis software (GraphPad Software, Inc.,La Jolla, CA). Data are shown as mean 6 S.E.M. values.

ResultsEndogenously Generated H2S Inhibits PYK2 in

Cultured Cardiomyocytes. To evaluate the role of endog-enous H2S on PYK2 activation, we evaluated PYK2 phosphor-ylation in differentiated H9c2 cardiomyocytes in the presenceof a pharmacologic inhibitor of H2S synthesis (AOAA)(Asimakopoulou et al., 2013) (Fig. 1, A and B) or the substratefor H2S synthesis (L-cysteine) (Fig. 1, A and B). In analternative approach, endogenous H2S production was en-hanced via the adenoviral-mediated overexpression of CSE(Fig. 1, C and D). In agreement with our recently publishedobservations (Bibli et al., 2017), H2O2 treatment resulted inthe increased phosphorylation of PYK2 on Y402 (Fig. 1A).Moreover, H2O2 treatment increased eNOS phosphorylationon the eNOSY656 inhibitory site (Fig. 1B). The effects of H2O2

treatment were mimicked by inhibiting endogenous H2Sproduction with AOAA (Fig. 1, A and B). PYK2 and eNOSphosphorylation were not enhanced by the combined treat-ment of H2O2 and AOAA. In addition, supplementation withthe H2S substrate L-cysteine (Fig. 1, A and B), or CSEoverexpression (Fig. 1, C and D) abrogated the biochemicalchanges on PYK2 (Fig. 1, A and C) and eNOS (Fig. 1, B and D)triggered by H2O2.Pharmacologic Administration of H2S Results in

PYK2 Inhibition and Activation of eNOS. Having estab-lished that endogenously produced H2S regulates PYK2activation, we sought to determine whether H2S supplemen-tation could mitigate the effects of H2O2 on the PYK2/eNOSpathway. We pretreated H9c2 cells with a sulfide salt, Na2S.In these experiments, we observed that Na2S eliminated boththe H2O2-induced PYK2 activation (Fig. 2A) and eNOSphosphorylation on Y656 (Fig. 2B). To prove that the effectsobserved on eNOS Y656 phosphorylation were mediated byPYK2, we used a heterologous expression system. HEK cellswere cotransfected with wild-type eNOS and an emptypcDNA3 vector, a wild-type PYK2, or a dominant negativePYK2 plasmid as in our previous studies (Fisslthaler et al.,2008). Cotransfection with wild-type eNOS and wild-typePYK2 resulted in an increase in peNOS on Y657 from the

Fig. 5. H2S inhibits PYK2 activity. (A). Effects ofincreasing concentrations of Na2S on purified PYK2enzymatic activity; n = 3 independent experiments(B). Effects of Na2S (100 mM, 30 minutes) on PYK2sulfhydration in H9c2 cells; n = 5 independent exper-iments; **P, 0.001; ***P, 0.0001 from Sol (two-wayANOVA, Bonferroni).

Fig. 4. Correlation of H2S levels with oxidative stress in cardiomyocytes.H2S produced byH9c2 cells overexpressing CSE (A) or after pharmacologicmanipulation of endogenous CSE activity (B) was assessed by using SSP4.(C and D) Hydrogen peroxide levels in cells with increased or decreasedH2S production; n = 5 independent experiments; **P , 0.001; ***P ,0.0001. t test (A); one-way ANOVA; Newman-Kleus (B); two-way ANOVA,Bonferroni (C and D).

724 Bibli et al.

at ASPE

T Journals on N

ovember 16, 2021

molpharm

.aspetjournals.orgD

ownloaded from

basal activity of overexpressed PYK2 (Fig. 2C); Na2S admin-istration inhibited eNOS Y656 phosphorylation. Exposure ofHEK cells to H2O2 further increased eNOS tyrosine phosphor-ylation; the effect of was reversed by incubation with Na2S. Incontrast to what was observed with wild-type PYK2, HEKcells transfected with the dominant negative PYK2 showed noincrease in eNOS phosphorylation under the conditionsstudied, confirming that eNOSY657 phosphorylation dependson PYK2 activity. To evaluate the effect of the pharmacologictreatments on eNOS activity, we measured the conversion ofL-arginine to L-citrulline (Fig. 3). Although exposure to Na2Sincreased eNOS activity, incubation of cells with H2O2 did notalter L-citrulline formation in cells that did not express PYK2.PYK2 overexpression reduced the L-citrulline/L-arginine ratioin line with our biochemical data (enhanced eNOS phosphor-ylation on Y657); this effect was reversed by Na2S. Incubationof PYK2-transfected cells with H2O2 potentiated the inhibi-tory effect on eNOS activity. Exogenous application of H2S toPYK2-transfected cells treated with H2O2 restored eNOSactivity, confirming that H2S inhibits PYK2 activation andalleviates its inhibitory effect on eNOS.Mechanisms of PYK2 Inhibition by H2S. To study the

mechanisms through which H2S inhibits PYK2 preventingeNOS inhibition, we determined the effect of H2S levels on thelevels of ROS, a known trigger for PYK2 activation. Over-expression of CSE or L-cysteine supplementation increasedH2S levels, whereas AOAA reduced these levels (Fig. 4, A andB). When H2S production was enhanced, H2O2 levels werereduced and vice versa (Fig. 4, C andD). IncreasedH2O2 levelsled to PYK2 Y402 and eNOS Y657 phosphorylation (Fig. 1),and H2S reversed this effect. To determine whether H2S alsohas direct effects on PYK2, we determined its ability to inhibitPYK2 activity. H2S elicited a robust inhibitory effect onrecombinant PYK2 with an IC50 in the sub-micromolar range(Fig. 5A). The inhibition of PYK2 by H2S was associated withenhanced sulfhydration of the kinase (Fig. 5B)H2S Salvages PYK2-Induced Cardiomyocyte Death

In Vitro. To study the effects of PYK2 on cardiomyocytesurvival, we used an in vitro model of H2O2-triggered oxida-tive stress injury and cell death. Inhibition of CSE/CBS–derived H2S by AOAA resulted in augmented cardiomyocytedeath under baseline conditions (Supplemental Fig. 1A), aswell as after H2O2 (Fig. 6A) administration. Providing addi-tional L-cysteine in the culturemedia increased cardiomyocytesurvival in the presence of H2O2 did not but did not reverse theeffect of AOAA (Fig. 6A; Supplemental Fig. 1A). To evaluatethe ability of endogenous H2S production to regulate PYK2activity in the context of cell survival, PYK2 was silencedusing an siRNA approach (Bibli et al., 2017). In line with ourprevious findings (Bibli et al., 2017), PYK2 silencing increasedcardiomyocyte survival in cells treated with H2O2. Moreover,the deleterious effect of AOAA treatment was not observedafter PYK2 knockdown (Fig. 6A). L-cysteine supplementationpartially reversed the effect of H2O2 in nontransfected andscrambled RNA-transfected cells, but not in PYK2 silencedcells (Fig. 6A). We next overexpressed CSE and exposed cellsto H2O2 to determine the contribution of PYK2 to the pro-tective effect of endogenously produced H2S. When PYK2 wasexpressed (nontransfected and scrambled ScrsiRNA trans-fected conditions), CSE restricted the deleterious effect ofH2O2 on cardiomyocyte survival (Fig. 6B); however, noadditional effect of CSE overexpression was observed in cells

in which PYK2 was silenced, indicating that H2S protectscardiomyocytes by inhibiting PYK2 under oxidative stressconditions. Finally, experiments were conducted in the

Fig. 6. H2S reduces oxidative stress-induced cardiomyocyte death in aPYK2-dependent manner. Differentiated H9c2 cells were subjected to12 hours of H2O2 500 mM exposure. To inhibit endogenously H2S pro-duction, the cells were pretreated with AOAA 1 mM (A). To stimulateendogenous H2S production, cells were pretreated with L-cysteine 500 mM(A). To assess the effect of inhibition of H2S production during H2O2exposure, cells were pretreatedwith a combination of AOAAand L-cysteineas described already (A). The effects of endogenous H2S production wereevaluated on both naïve cells and in cells inwhich PYK2was inhibited withPYK2 siRNA for 48 hours. In (B) H9c2 cells, survival was determined afterinfection with a GFP or CSE Ad (10 MOI, 36 hour) before the addition ofH2O2; control cells or cells in which PYK2 was silenced were used. Toincrease H2S availability, cells were treated with Na2S 100 mM for30 minutes and then exposed to H2O2. In the same experimental series,PYK2 was pharmacologically inhibited with 5 mM PF-431396 for 45 min-utes before H2O2 (C). At the end of the incubation, cell viability wasassessed by the MTT assay. n = 6 independent experiments; *P , 0.05 vs.No injury (white cirles); #P , 0.05 vs. H2O2;

xP , 0.05 vs. AOAA (A); (two-way ANOVA, Bonferroni A–C).

PYK2, H2S, eNOS, and Cardioprotection 725

at ASPE

T Journals on N

ovember 16, 2021

molpharm

.aspetjournals.orgD

ownloaded from

presence of Na2S as an exogenous source of H2S andPF-431396 as a pharmacologic inhibitor of PYK2. In thisseries of experiments, we observed that both the PYK2inhibitor and Na2S improved cell survival after H2O2 expo-sure; however, combining the sulfide salt with PF-431396 didnot exert an additional effect (Fig. 6C).H2S Alleviates eNOS Inhibition During Reperfu-

sion In Vivo. To test whether the observed in vitrofindings could be extrapolated in vivo, we used a LADligation model. In these experiments, administration ofNa2S in sham-operated animals did not affect the basallevels of PYK2 and eNOS Y656 phosphorylation (Supple-mental Fig. 2, A and B) or cGMP levels (Supplemental Fig. 2C).When Na2S was administrated intravenously 10 minutesbefore reperfusion, however, a 50% reduction in the phos-phorylation of PYK2 was observed in the early minutes ofreperfusion (Fig. 7A). At the same time, we also noted areduction in Y656 phosphorylation of eNOS (Fig. 7B) alongwith an increase in the levels of the surrogate NO marker,cGMP (Fig. 7C). In addition, administration of Na2S duringischemia resulted in a reduction in the oxidative andnitrosative stress biomarkers malondialdehyde (MDA;Fig. 8A) and PCs (Fig. 8B) in the early minutes of reperfu-sion. Similarly, nitrotyrosine levels (Fig. 8C) and DHE-reactive products (Fig. 8D) were attenuated in Na2S-treatedanimals compared with the solvent-treated animals. These

findings taken together demonstrate that Na2S inhibitsoxidative stress, limits PYK2 activity, and de-represseseNOS activity.Cardioprotective Effects of H2S Are Dependent on

the PYK2/eNOS Pathway. Next, we tested whether theincrease in eNOS activity brought about by the H2S-mediated inhibition of PYK2 yields functionally relevantoutcomes in vivo. To do so, we assessed the infarct size inmice subjected to ischemia/reperfusion injury in the pres-ence of a pharmacologic PYK2 inhibitor or/and an H2Ssource in a dose previously reported by our group not toaffect hemodynamic parameters (Chatzianastasiou et al.,2016). In agreement with our previously published obser-vations (Bibli et al., 2017), the pharmacologic inhibition ofPYK2 reduced myocardial infarct size (36.8%6 2.0% for theSol group, 18.0% 6 0.9% for the PF-431396 group). Simi-larly, we observed that the administration of Na2S wasprotective (36.8% 6 2.0% for the Sol group and 17.8% 61.6% for the Na2S group, *P , 0.05). However, the simul-taneous administration of PF-431396 and Na2S exerted noadditional beneficial effects with respect to infarct size(20.2% 6 2.5%), implying that these two agents rely on thesame downstream molecular targets (Fig. 9, A and C). Nostatistically significant differences were observed in thearea at risk to whole myocardial area among the studiedgroups (Fig. 9B).

Fig. 7. Exogenous Na2S regulates PYK2/eNOS activa-tion in early reperfusion in vivo. Mice were subjected toischemia (30minutes) and reperfusion for 3minutes. Onegroup of mice received Na2S as an i.v. bolus at 100 mg/kg10 minutes before myocardial reperfusion. Ischemictissues from the left ventricle were then collected andPYK2 (A), eNOS phosphorylation (B), and cGMP levels(C) were determined. Representative Western blots anddensitometric analysis of (A) pPYK2 (Y402) and (B)peNOS (Y656). Phosphorylated protein levels were nor-malized to total protein levels. (C) cGMP levels in theischemic myocardium in mice receiving solvent or Na2Streatment; n = 6 animals per group; **P , 0.001 (t test).

726 Bibli et al.

at ASPE

T Journals on N

ovember 16, 2021

molpharm

.aspetjournals.orgD

ownloaded from

DiscussionMyocardial ischemia induces cellular damage via malad-

aptive biochemical responses in the ischemic organ (Yellonand Hausenloy, 2007). Subsequent reperfusion, althoughbeneficial, leads to further paradoxical intracellular injuryand increased myocardial death. No pharmacologic strategieshave been introduced so far in to routine clinical practice toreduce infarct size in patients undergoing acute myocardialinfarction (Hausenloy and Yellon, 2016). Better understand-ing of the intracellular signaling of ischemia/reperfusioninjury is expected to lead to novel therapeutic strategies foracute myocardial infarction patients. Herein, we investigatedthe impact of H2S on the PYK2/eNOS axis and its relevance tocardioprotection.Initially, we set out to determine whether endogenously

produced H2S regulates PYK2 phosphorylation. When H9c2cells were treated acutely with the CSE/CBS inhibitor AOAA(Asimakopoulou et al., 2013), we observed an increase inPYK2 tyrosine phosphorylation, indicative of enhanced PYK2activation. Since PYK2 is activated by ROS (Jones and Bolli,2006) and H2S exhibits both direct and indirect antioxidanteffects, (Ju et al., 2013; Xie et al., 2016), AOAA-triggered PYK2activation might be the result of increased oxidative stress onlowering H2S levels. In line with this hypothesis, exogenouslyadded H2O2 mimicked the effects of AOAA on PYK2. More-over, coincubation of cells with AOAA and H2O2 exerted noadditional effects on PYK2 phosphorylation, indicating acommon mechanism of action for AOAA and H2O2. Finally,

supplementation with theH2S synthesis substrate L-cysteine,CSE overexpression or exogenously added H2S reversed theeffects of H2O2 on PYK2, providing further evidence that H2Slimits PYK2 activation by counteracting the action of oxidantmolecules. In addition to preventing PYK2 activation byreducing oxidative stress, we observed that H2S can directlyinhibit PYK2 activity, suggesting a dual mechanism of actionfor H2S.In agreementwith the fact that PYK2 phosphorylates eNOS on

Y656, in all the experiments performed, changes in eNOS tyrosinephosphorylationparalleled those inPYK2phosphorylation.Directevidence for the involvement of PYK2 on H2O2-induced eNOSY656 phosphorylation was provided by a dominant negativeapproach. Whereas most researchers study S1177 phosphoryla-tion (human eNOS sequence, corresponds to murine S1176) as asurrogate marker of eNOS activity (Dimmeler et al., 1999; Fultonet al., 1999), we believe it is more appropriate to study Y657. Wehave previously shown that phosphorylation of Y657 exerts adominant effect compared with S1177; whereas S1177 phosphor-ylation leads to eNOS activation, dual phosphorylation of Y657/S1177 abolishes eNOS activity (Bibli et al., 2017). The observedchanges on eNOS phosphorylation triggered by H2S were accom-panied by changes in activity. H2O2 reduced eNOSactivity only incells expressing PYK2; this effect was reversible by H2S.To determine the ability of H2S to inhibit PYK2-mediated

cell toxicity, we exposed H9c2 cells to H2O2. Treatment ofcardiomyocytes with H2O2 significantly reduced cell survival;the effect of H2O2 could be ameliorated by increasing H2S

Fig. 8. Na2S reduces ROS production in earlyreperfusion. Mice were subjected to ischemia(30 minutes) and reperfusion for 3 minutes.One group ofmice receivedNa2S as an i.v. bolusat 100 mg/kg 10 minutes before myocardialreperfusion. Tissues were then collected and(A) MDA or (B) PC determined by colorimetricassays. (C) Nitrotyrosine semiquantitative lev-els were determined with a nitrotyrosine Ab,and (D) myocardial ROS were determined byDHE staining in 8 mM cryosections of theischemic myocardial area; n = 6 animals pergroup. *P , 0.05; ** P , 0.001 (t test).

PYK2, H2S, eNOS, and Cardioprotection 727

at ASPE

T Journals on N

ovember 16, 2021

molpharm

.aspetjournals.orgD

ownloaded from

production via L-cysteine, CSE overexpression, or the exoge-nous addition of Na2S. Our findings agree with previouslypublished findings that H2S donors protect H9c2 cells fromH2O2 toxicity (Szabó et al., 2011; Zhao et al., 2015;Chatzianastasiou et al., 2016; Bibli et al., 2017). We recentlyreported that the toxicity of H2O2 in cultured cardiomyocytescould be inhibited by pharmacological PYK2 inhibition orPYK2 silencing and that the effect of PYK2 was eNOS-dependent (Bibli et al., 2017). In the present series ofexperiments, we found that H2S was unable to improve cellsurvival in H9c2 when PYK2 was silenced or inhibited.Similarly, reducing endogenous H2S production did not leadto greater toxicity when PYK2 was silenced. Taken together,the aforementioned observations suggest that H2S signals viaPYK2 to protect cardiomyocytes against oxidative stress-induced toxicity. Based on our findings that: 1) the protectiveeffect of PYK2 inhibition in H9c2 is linked to de-repression ofeNOS (Bibli et al., 2017), and 2) the herein reported observa-tion that H2S blocks PYK2 activation, we propose that theprosurvival effect of H2S is mediated by an increase in NObioavailability that results from PYK2 inhibition.Nitric oxide is considered a key player in myocardial

ischemia/reperfusion injury in vivo (Jones and Bolli, 2006;Andreadou et al., 2015). We and others have shown that eNOSis required for the cardioprotective actions of H2S in vivo(Kondo et al., 2013; King et al., 2014; Bibli et al., 2015a;Chatzianastasiou et al., 2016; Karwi et al., 2017). Our group

has recently demonstrated that the capability of eNOS toproduce NO during ischemia/reperfusion injury is regulatedby the redox-sensitive kinase PYK2 (Bibli et al., 2017). PYK2phosphorylation peaks roughly 3 minutes after reperfusion,returning to baseline within 10 minutes. The time course ofeNOS phosphorylation on Y656 parallels that of PYK2activation, resulting in reduced NO production, contributingto myocardial death. We found that administration of H2S caninhibit PYK2 phosphorylation in the early minutes of reper-fusion in the infarcted left ventricle, which resulted inalleviation of the inhibitory eNOS tyrosine phosphorylationand a subsequent increase of the cardiac levels of the NOsurrogate marker cGMP. The observed biochemical changestranslated to functional outcomes, as inhibition of PYK2kinase either via Na2S or via its pharmacologic inhibitorPF-431396 resulted in a reduction of myocardial infract size inthe murine hearts. In compliance with our in vitro cell-survival studies, simultaneous administration of Na2S andPF-431396 did not exert additional beneficial effects in myo-cardial survival, further suggesting that Na2S exerts itseffects thought PYK2 inhibition.As ROSs are a likely trigger for PYK2 activation after

reperfusion, we assessed their levels 3 minutes afterischemia/reperfusion injury when PYK2 activity is maximal(Bibli et al., 2017) by determining four different indexes ofoxidative stress, namely, MDA, PC, nitrotyrosine levels, andDHE-reactive species. All the indexes measured were lower in

Fig. 9. Na2S limits myocardial infarctsize in a PYK2-dependent manner.Mice were subjected to LAD ligation;infracted area, area at risk, and totalarea were determined. (A). Infarcted toarea at risk ratio as percentage; n =8 for Sol group, n = 8 for PF-431396group, n = 8 for Na2S, n = 6 forPF-43139+ Na2S group; **P , 0.001.(B). Ratio of area at risk to whole myo-cardial area; p = NS among groups. (C)Representative pictures from differenttreatment groups (two-way ANOVA,Bonferroni).

728 Bibli et al.

at ASPE

T Journals on N

ovember 16, 2021

molpharm

.aspetjournals.orgD

ownloaded from

mice receiving Na2S compared with vehicle-treated mice at aspecific time. Moreover, the reduced oxidative stress at earlytime points of reperfusion correlated with lower levels of PYK2and eNOS tyrosine phosphorylation. Several reports havedemonstrated that the cardioprotective effects of H2S dependon its antioxidant properties. Upregulation of antioxidantprotein expression via Nrf-2 activation has been suggested asa major protective pathway used by H2S (Calvert et al., 2009;Peake et al., 2013; Shimizu et al., 2016); however, the acuteprotective effect observed in our studies was likely independentof the transcriptional activation of antioxidant defense genes.In addition, although direct scavenging of several ROShas beenshown to occur in vitro (Li and Lancaster, 2013; Kabil et al.,2014), the biologic relevance of these reactions has not beendemonstrated in vivo and is likely of minor, if any, importanceowing to their slow rates and the limited amount of free H2Scomparedwith other reducing compounds in living cells. Amoreplausible mechanism for the action for H2S is persulfidationand modification of the activity of proteins involved in regulat-ing ROS levels. One such example was provided by Sun et al.(2012), who demonstrated that H2S decreases the levels of ROSby inhibitingmitochondrial complex IV and increasingMn- andCuZn-SOD activities in cardiomyocytes; NaHS also activatedCuZn-SOD in a cell-free system. ThemechanismbywhichNa2Slimits ROS production during early reperfusion and preventsPYK2 activation is a matter of future investigations.So far, numerous synergistic interactions between NO and

H2S have been reported (Szabo, 2017). For example, eNOSexpression increases after treatment with H2S, (Meng et al.,2013), and exposure to H2S promotes eNOS coupling, dimer-ization and phosphorylation on the S1177 residue (Colettaet al., 2012; King et al., 2014). H2S can also release NO fromintracellular storage pools (Bir et al., 2012; Olson, 2013),inhibit phosphodiesterase activity to boost cGMP signaling(Bucci et al., 2010; Bucci et al., 2012), and preserve theNO receptor, soluble guanylate cyclase, in its reduced,NO-responsive state (Zhou et al., 2016). In the present study,we report an additional level of H2S-NO crosstalk and furtherunderscore the importance of NO in the biologic activity ofH2S. In conclusion, our study confirms the importance ofPYK2 inhibition for cardioprotection and provides a novelmolecular pathway through which H2S donors exert theirbeneficial cardiovascular effects.

Authorship Contributions

Participated in research design: Bibli, Szabo, Fleming,Papapetropoulos.

Conducted experiments: Bibli, Chatzianastasiou, Luck, Zukunft.Performed data analysis: Bibli, Papapetropoulos.Wrote or contributed to the writing of the manuscript: Bibli, Szabo,

Fleming, Papapetropoulos.

References

Altaany Z, Ju Y, Yang G, and Wang R (2014) The coordination of S-sulfhydration,S-nitrosylation, and phosphorylation of endothelial nitric oxide synthase by hy-drogen sulfide. Sci Signal 7:ra87.

Altaany Z, Yang G, and Wang R (2013) Crosstalk between hydrogen sulfide and nitricoxide in endothelial cells. J Cell Mol Med 17:879–888.

Andreadou I, Iliodromitis EK, Rassaf T, Schulz R, Papapetropoulos A,and Ferdinandy P (2015) The role of gasotransmitters NO, H2S and CO in myo-cardial ischaemia/reperfusion injury and cardioprotection by preconditioning,postconditioning and remote conditioning. Br J Pharmacol 172:1587–1606.

Andreadou I, Mikros E, Ioannidis K, Sigala F, Naka K, Kostidis S, Farmakis D, Tenta R,Kavantzas N, Bibli SI, et al. (2014) Oleuropein prevents doxorubicin-induced car-diomyopathy interfering with signaling molecules and cardiomyocyte metabolism.J Mol Cell Cardiol 69:4–16.

Asimakopoulou A, Panopoulos P, Chasapis CT, Coletta C, Zhou Z, Cirino G, GiannisA, Szabo C, Spyroulias GA, and Papapetropoulos A (2013) Selectivity of com-monly used pharmacological inhibitors for cystathionine b synthase (CBS) andcystathionine g lyase (CSE). Br J Pharmacol 169:922–932.

Bibli SI, Andreadou I, Chatzianastasiou A, Tzimas C, Sanoudou D, Kranias E,Brouckaert P, Coletta C, Szabo C, Kremastinos DT, et al. (2015a) Cardioprotectionby H2S engages a cGMP-dependent protein kinase G/phospholamban pathway.Cardiovasc Res 106:432–442.

Bibli SI, Yang G, Zhou Z, Wang R, Topouzis S, and Papapetropoulos A (2015b) Role ofcGMP in hydrogen sulfide signaling. Nitric Oxide 46:7–13.

Bibli SI, Zhou Z, Zukunft S, Fisslthaler B, Andreadou I, Szabo C, Brouckaert P,Fleming I, and Papapetropoulos A (2017) Tyrosine phosphorylation of ENOS reg-ulates myocardial survival after an ischaemic insult: role of PYK2. Cardiovasc Res113:926–937.

Bir SC, Kolluru GK, McCarthy P, Shen X, Pardue S, Pattillo CB, and Kevil CG (2012)Hydrogen sulfide stimulates ischemic vascular remodeling through nitric oxidesynthase and nitrite reduction activity regulating hypoxia-inducible factor-1a andvascular endothelial growth factor-dependent angiogenesis. J Am Heart Assoc 1:e004093.

Bucci M, Papapetropoulos A, Vellecco V, Zhou Z, Pyriochou A, Roussos C, Roviezzo F,Brancaleone V, and Cirino G (2010) Hydrogen sulfide is an endogenous inhibitor ofphosphodiesterase activity. Arterioscler Thromb Vasc Biol 30:1998–2004.

Bucci M, Papapetropoulos A, Vellecco V, Zhou Z, Zaid A, Giannogonas P, CantalupoA, Dhayade S, Karalis KP, Wang R, et al. (2012) cGMP-dependent protein kinasecontributes to hydrogen sulfide-stimulated vasorelaxation. PLoS One 7:e53319.

Calvert JW, Jha S, Gundewar S, Elrod JW, Ramachandran A, Pattillo CB, Kevil CG,and Lefer DJ (2009) Hydrogen sulfide mediates cardioprotection through Nrf2signaling. Circ Res 105:365–374.

Chappell AE, Bunz M, Smoll E, Dong H, Lytle C, Barrett KE, and McCole DF (2008)Hydrogen peroxide inhibits Ca21-dependent chloride secretion across colonic ep-ithelial cells via distinct kinase signaling pathways and ion transport proteins.FASEB J 22:2023–2036.

Chatzianastasiou A, Bibli SI, Andreadou I, Efentakis P, Kaludercic N, Wood ME,Whiteman M, Di Lisa F, Daiber A, Manolopoulos VG, et al. (2016) Cardioprotectionby H2S donors: nitric oxide-dependent and ‑independent mechanisms.J Pharmacol Exp Ther 358:431–440.

Coletta C, Papapetropoulos A, Erdelyi K, Olah G, Módis K, Panopoulos P,Asimakopoulou A, Gerö D, Sharina I, Martin E, et al. (2012) Hydrogen sulfideand nitric oxide are mutually dependent in the regulation of angiogenesisand endothelium-dependent vasorelaxation. Proc Natl Acad Sci USA 109:9161–9166.

Das A, Samidurai A, Hoke NN, Kukreja RC, and Salloum FN (2015) Hydrogen sulfidemediates the cardioprotective effects of gene therapy with PKG-Ia. Basic ResCardiol 110:42.

Dimmeler S, Fleming I, Fisslthaler B, Hermann C, Busse R, and Zeiher AM (1999)Activation of nitric oxide synthase in endothelial cells by Akt-dependent phos-phorylation. Nature 399:601–605.

Elrod JW, Calvert JW, Morrison J, Doeller JE, Kraus DW, Tao L, Jiao X, Scalia R,Kiss L, Szabo C, et al. (2007) Hydrogen sulfide attenuates myocardial ischemia-reperfusion injury by preservation of mitochondrial function. Proc Natl Acad SciUSA 104:15560–15565.

Fisslthaler B, Loot AE, Mohamed A, Busse R, and Fleming I (2008) Inhibition ofendothelial nitric oxide synthase activity by proline-rich tyrosine kinase 2 in re-sponse to fluid shear stress and insulin. Circ Res 102:1520–1528.

Fulton D, Gratton JP, McCabe TJ, Fontana J, Fujio Y, Walsh K, Franke TF, Papa-petropoulos A, and Sessa WC (1999) Regulation of endothelium-derived nitric oxideproduction by the protein kinase Akt. Nature 399:597–601.

Greaney JL, Kutz JL, Shank SW, Jandu S, Santhanam L, and Alexander LM (2017)Impaired hydrogen sulfide-mediated vasodilation contributes to microvascularendothelial dysfunction in hypertensive adults. Hypertension 69:902–909.

Hausenloy DJ and Yellon DM (2016) Ischaemic conditioning and reperfusion injury.Nat Rev Cardiol 13:193–209.

Hirotani S, Higuchi Y, Nishida K, Nakayama H, Yamaguchi O, Hikoso S, Takeda T,Kashiwase K, Watanabe T, Asahi M, et al. (2004) Ca(21)-sensitive tyrosine kinasePyk2/CAK beta-dependent signaling is essential for G-protein-coupled receptoragonist-induced hypertrophy. J Mol Cell Cardiol 36:799–807.

Johansen D, Ytrehus K, and Baxter GF (2006) Exogenous hydrogen sulfide (H2S)protects against regional myocardial ischemia-reperfusion injury–Evidence for arole of K ATP channels. Basic Res Cardiol 101:53–60.

Jones SP and Bolli R (2006) The ubiquitous role of nitric oxide in cardioprotection.J Mol Cell Cardiol 40:16–23.

Ju Y, Zhang W, Pei Y, and Yang G (2013) H(2)S signaling in redox regulation ofcellular functions. Can J Physiol Pharmacol 91:8–14.

Kabil O and Banerjee R (2014) Enzymology of H2S biogenesis, decay and signaling.Antioxid Redox Signal 20:770–782.

Kabil O, Motl N, and Banerjee R (2014) H2S and its role in redox signaling. BiochimBiophys Acta 1844:1355–1366.

Kanagy NL, Szabo C, and Papapetropoulos A (2017) Vascular biology of hydrogensulfide. Am J Physiol Cell Physiol 312:C537–C549.

Karwi QG, Bornbaum J, Boengler K, Torregrossa R, Whiteman M, Wood ME, SchulzR, and Baxter GF (2017) AP39, a mitochondria-targeting hydrogen sulfide (H2 S)donor, protects against myocardial reperfusion injury independently of salvagekinase signalling. Br J Pharmacol 174:287–301.

Katsouda A, Bibli SI, Pyriochou A, Szabo C, and Papapetropoulos A (2016) Regula-tion and role of endogenously produced hydrogen sulfide in angiogenesis.Pharmacol Res 113 (Pt A):175–185.

Kimura H (2011) Hydrogen sulfide: its production, release and functions. AminoAcids 41:113–121.

King AL, Polhemus DJ, Bhushan S, Otsuka H, Kondo K, Nicholson CK, Bradley JM,Islam KN, Calvert JW, Tao YX, et al. (2014) Hydrogen sulfide cytoprotective

PYK2, H2S, eNOS, and Cardioprotection 729

at ASPE

T Journals on N

ovember 16, 2021

molpharm

.aspetjournals.orgD

ownloaded from

signaling is endothelial nitric oxide synthase-nitric oxide dependent. Proc NatlAcad Sci USA 111:3182–3187.

Kondo K, Bhushan S, King AL, Prabhu SD, Hamid T, Koenig S, Murohara T,Predmore BL, Gojon G, Sr, Gojon G, Jr, et al. (2013) H2S protects againstpressure overload-induced heart failure via upregulation of endothelial nitricoxide synthase. Circulation 127:1116–1127.

Koshman YE, Chu M, Kim T, Kalmanson O, Farjah M, Kumar M, Lewis W, GeenenDL, de Tombe P, Goldspink PH, et al. (2014) Cardiomyocyte-specific expression ofCRNK, the C-terminal domain of PYK2, maintains ventricular function and slowsventricular remodeling in a mouse model of dilated cardiomyopathy. J Mol CellCardiol 72:281–291.

Lev S, Moreno H, Martinez R, Canoll P, Peles E, Musacchio JM, Plowman GD, Rudy B,and Schlessinger J (1995) Protein tyrosine kinase PYK2 involved in Ca(21)-inducedregulation of ion channel and MAP kinase functions. Nature 376:737–745.

Li L, Rose P, and Moore PK (2011) Hydrogen sulfide and cell signaling. Annu RevPharmacol Toxicol 51:169–187.

Li Q and Lancaster JR, Jr (2013) Chemical foundations of hydrogen sulfide biology.Nitric Oxide 35:21–34.

Loot AE, Schreiber JG, Fisslthaler B, and Fleming I (2009) Angiotensin II impairsendothelial function via tyrosine phosphorylation of the endothelial nitric oxidesynthase. J Exp Med 206:2889–2896.

Meng J, Ganesan Adaikan P, and Srilatha B (2013) Hydrogen sulfide promotes nitricoxide production in corpus cavernosum by enhancing expression of endothelialnitric oxide synthase. Int J Impot Res 25:86–90.

Minamishima S, Bougaki M, Sips PY, Yu JD, Minamishima YA, Elrod JW, Lefer DJ,Bloch KD, and Ichinose F (2009) Hydrogen sulfide improves survival after cardiacarrest and cardiopulmonary resuscitation via a nitric oxide synthase 3-dependentmechanism in mice. Circulation 120:888–896.

Módis K, Bos EM, Calzia E, van Goor H, Coletta C, Papapetropoulos A, Hellmich MR,Radermacher P, Bouillaud F, and Szabo C (2014) Regulation of mitochondrialbioenergetic function by hydrogen sulfide. Part II. Pathophysiological and thera-peutic aspects. Br J Pharmacol 171:2123–2146.

Mustafa AK, Gadalla MM, Sen N, Kim S, Mu W, Gazi SK, Barrow RK, Yang G, WangR, and Snyder SH (2009) H2S signals through protein S-sulfhydration. Sci Signal2:ra72.

Olson SD (2013) Fluid dynamic model of invertebrate sperm chemotactic motilitywith varying calcium inputs. J Biomech 46:329–337.

Pan TT, Feng ZN, Lee SW, Moore PK, and Bian JS (2006) Endogenous hydrogensulfide contributes to the cardioprotection by metabolic inhibition preconditioningin the rat ventricular myocytes. J Mol Cell Cardiol 40:119–130.

Papapetropoulos A, Foresti R, and Ferdinandy P (2015) Pharmacology of the‘gasotransmitters’ NO, CO and H2S: translational opportunities. Br J Pharmacol172:1395–1396.

Papapetropoulos A, Pyriochou A, Altaany Z, Yang G, Marazioti A, Zhou Z, JeschkeMG, Branski LK, Herndon DN, Wang R, et al. (2009) Hydrogen sulfide is anendogenous stimulator of angiogenesis. Proc Natl Acad Sci USA 106:21972–21977.

Paul BD and Snyder SH (2012) H2S signalling through protein sulfhydration andbeyond. Nat Rev Mol Cell Biol 13:499–507.

Peake BF, Nicholson CK, Lambert JP, Hood RL, Amin H, Amin S, and Calvert JW(2013) Hydrogen sulfide preconditions the db/db diabetic mouse heart againstischemia-reperfusion injury by activating Nrf2 signaling in an Erk-dependentmanner. Am J Physiol Heart Circ Physiol 304:H1215–H1224.

Polhemus D, Kondo K, Bhushan S, Bir SC, Kevil CG, Murohara T, Lefer DJ,and Calvert JW (2013) Hydrogen sulfide attenuates cardiac dysfunction after heartfailure via induction of angiogenesis. Circ Heart Fail 6:1077–1086.

Polhemus DJ, Calvert JW, Butler J, and Lefer DJ (2014) The cardioprotective actionsof hydrogen sulfide in acute myocardial infarction and heart failure. Scientifica(Cairo) 2014:768607.

Polhemus DJ and Lefer DJ (2014) Emergence of hydrogen sulfide as an endogenousgaseous signaling molecule in cardiovascular disease. Circ Res 114:730–737.

Predmore BL, Kondo K, Bhushan S, Zlatopolsky MA, King AL, Aragon JP,Grinsfelder DB, Condit ME, and Lefer DJ (2012) The polysulfide diallyl tri-sulfide protects the ischemic myocardium by preservation of endogenous hy-drogen sulfide and increasing nitric oxide bioavailability. Am J Physiol HeartCirc Physiol 302:H2410–H2418.

Salloum FN (2015) Hydrogen sulfide and cardioprotection–Mechanistic insights andclinical translatability. Pharmacol Ther 152:11–17.

Shibuya N, Koike S, Tanaka M, Ishigami-Yuasa M, Kimura Y, Ogasawara Y, Fukui K,Nagahara N, and Kimura H (2013) A novel pathway for the production of hydrogensulfide from D-cysteine in mammalian cells. Nat Commun 4:1366.

Shimizu Y, Nicholson CK, Lambert JP, Barr LA, Kuek N, Herszenhaut D, Tan L,Murohara T, Hansen JM, Husain A, et al. (2016) Sodium sulfide attenuatesischemic-induced heart failure by enhancing proteasomal function in an Nrf2-dependent manner. Circ Heart Fail 9:e002368.

Sun WH, Liu F, Chen Y, and Zhu YC (2012) Hydrogen sulfide decreases the levels of ROSby inhibiting mitochondrial complex IV and increasing SOD activities in cardiomyocytesunder ischemia/reperfusion. Biochem Biophys Res Commun 421:164–169.

Szabó C (2007) Hydrogen sulphide and its therapeutic potential.Nat Rev Drug Discov6:917–935.

Szabo C (2017) Hydrogen sulfide, an enhancer of vascular nitric oxide signaling:mechanisms and implications. Am J Physiol Cell Physiol 312:C3–C15.

Szabó C and Papapetropoulos A (2011) Hydrogen sulphide and angiogenesis: mech-anisms and applications. Br J Pharmacol 164:853–865.

Szabó G, Veres G, Radovits T, Gero D, Módis K, Miesel-Gröschel C, Horkay F, KarckM, and Szabó C (2011) Cardioprotective effects of hydrogen sulfide.Nitric Oxide 25:201–210.

Tai LK, Okuda M, Abe J, Yan C, and Berk BC (2002) Fluid shear stress activatesproline-rich tyrosine kinase via reactive oxygen species-dependent pathway.Arterioscler Thromb Vasc Biol 22:1790–1796.

Takeishi Y (2014) Pivotal roles of regulating the proline-rich tyrosine kinase 2 (PYK2)signaling in cardiac function and remodeling. J Mol Cell Cardiol 74:295–296.

Tokiwa G, Dikic I, Lev S, and Schlessinger J (1996) Activation of Pyk2 by stresssignals and coupling with JNK signaling pathway. Science 273:792–794.

Wang R (2012) Physiological implications of hydrogen sulfide: a whiff explorationthat blossomed. Physiol Rev 92:791–896.

Wang R, Szabo C, Ichinose F, Ahmed A, Whiteman M, and Papapetropoulos A (2015)The role of H2S bioavailability in endothelial dysfunction. Trends Pharmacol Sci36:568–578.

Xie ZZ, Liu Y, and Bian JS (2016) Hydrogen sulfide and cellular redox homeostasis.Oxid Med Cell Longev 2016:6043038.

Yang G, Wu L, Jiang B, Yang W, Qi J, Cao K, Meng Q, Mustafa AK, Mu W, Zhang S,et al. (2008) H2S as a physiologic vasorelaxant: hypertension in mice with deletionof cystathionine gamma-lyase. Science 322:587–590.

Yellon DM and Hausenloy DJ (2007) Myocardial reperfusion injury. N Engl J Med357:1121–1135.

Zhao Y, Yang C, Organ C, Li Z, Bhushan S, Otsuka H, Pacheco A, Kang J, AguilarHC, Lefer DJ, et al. (2015) Design, synthesis, and cardioprotective effects ofN-mercapto-based hydrogen sulfide donors. J Med Chem 58:7501–7511.

Zhou Z, Martin E, Sharina I, Esposito I, Szabo C, Bucci M, Cirino G,and Papapetropoulos A (2016) Regulation of soluble guanylyl cyclase redox state byhydrogen sulfide. Pharmacol Res 111:556–562.

Address correspondence to: Dr. Andreas Papapetropoulos, Faculty ofPharmacy, Panepistimiopolis, Zografou, Athens 15771, Greece. E-mail:apapapet@pharm.uoa.gr

730 Bibli et al.

at ASPE

T Journals on N

ovember 16, 2021

molpharm

.aspetjournals.orgD

ownloaded from