Life Sciences 93 (2013) 359–366

Contents lists available at ScienceDirect

Life Sciences

j ourna l homepage: www.e lsev ie r .com/ locate / l i fesc ie

Inhibition of sphingosine-1-phosphate lyase rescues sphingosinekinase-1-knockout phenotype following murine cardiac arrest

Irina A. Gorshkova a,f,1, Huashan Wang b,1, Gerasim A. Orbelyan c, Jonathan Goya d,Viswanathan Natarajan a,e,f, David G. Beiser c, Terry L. Vanden Hoek b, Evgeny V. Berdyshev a,f,⁎a Department of Medicine, University of Illinois at Chicago, Chicago, IL, United Statesb Deparment of Emergency Medicine, University of Illinois at Chicago, Chicago, IL, United Statesc Section of Emergency Medicine, Department of Emergency Medicine, University of Chicago, Chicago, IL, United Statesd Department of Quantitative and Computational Biology, Princeton University, Princeton, NJ, United Statese Department of Pharmacology, University of Illinois at Chicago, Chicago, IL, United Statesf Institute for Personalized Respiratory Medicine, University of Illinois at Chicago, Chicago, IL, United States

⁎ Corresponding author at: University of Illinois at Chica3111, M/C 719, Chicago, IL 60612, United States. Tel.: +1

E-mail address: eberdysh@uic.edu (E.V. Berdyshev).1 Both authors contributed equally to this manuscript.

0024-3205/$ – see front matter © 2013 Elsevier Inc. All rihttp://dx.doi.org/10.1016/j.lfs.2013.07.017

a b s t r a c t

a r t i c l e i n f o

Article history:

Received 14 March 2013Accepted 12 July 2013

Keywords:Cardiac arrestDihydrosphingosine-1-phosphateS1P lyaseS1P signaling system

Aims: To test the role of sphingosine-1-phosphate (S1P) signaling system in the in vivo setting of resuscitationand survival after cardiac arrest.Mainmethods:Amousemodel of potassium-induced cardiac arrest and resuscitationwas used to test the impor-tance of S1P homeostasis in resuscitation and survival. C57BL/6 and sphingosine kinase-1 knockout (SphK1-KO)female mice were arrested for 8 min then subjected to 5 minute CPR with epinephrine bolus given at 90 s afterthe beginning of CPR. Animal survival wasmonitored for 4 h post-resuscitation. Upregulation of tissue and circu-latory S1P levels were achieved via inhibition of S1P lyase by 2-acetyl-5-tetrahydroxybutyl imidazole (THI).Plasma and heart tissue S1P and ceramide levels were quantified by targeted ESI-LC/MS/MS.

Key findings: Lack of SphK1 and low tissue/circulatory S1P levels in SphK1-KOmice led to poor animal resuscita-tion after cardiac arrest and to impaired survival post-resuscitation. Inhibition of S1P lyase in SphK1-KO micedrastically improved animal resuscitation and survival. Improved resuscitation and survival of THI-treatedSphK1-KO mice were better correlated with cardiac dihydro-S1P (DHS1P) than S1P levels. The lack of SphK1and the inhibition of S1P lyase by THI were accompanied by modulation in cardiac S1PR1 and S1PR2 expressionand by selective changes in plasma N-palmitoyl- and N-behenoyl-ceramide levels.Significance: Our data provide evidence for the crucial role for SphK1 and S1P signaling system in resuscitationand survival after cardiac arrest, which may form the basis for development of novel therapeutic strategy tosupport resuscitation and long-term survival of cardiac arrest patients.

© 2013 Elsevier Inc. All rights reserved.

Introduction

The sphingosine-1-phosphate (S1P) signaling system is well docu-mented to be critical for the resistance to different stress conditions.From yeast to mammalian cells, the enforcement of S1P-mediated sig-naling provides protection from heat (Mao et al., 1999; Cowart et al.,2010), metabolic (Oskouian and Saba, 2010; Ponnusamy et al., 2010),or oxidative (Nikolova-Karakashian and Reid, 2011; Van Brocklyn andWilliams, 2012; Karliner, 2013) stress-induced cell damage. In cancercells, the S1P signaling system plays an important role in promotingcancer cell proliferation and in protecting them from chemotherapeuticdrug-induced apoptosis (Oskouian and Saba, 2010; Ponnusamy et al.,2010; Pyne and Pyne, 2010; Watters et al., 2011). S1P exerts its action

go, 909 S.Wolcott Ave., COMRB312 355 5918.

ghts reserved.

by ligating to five known S1P receptors (S1PR1–5), of which S1PR1 isthe most reported to mediate pro-survival properties of extracellularlyinitiated S1P signaling (Uhlig and Gulbins, 2008; Skoura and Hla,2009;Means and Brown, 2009; Diab et al., 2010). In addition, intracellu-lar S1P is able to directly bind to its protein targets and thus to affectsignaling responses (Suomalainen et al., 2005; Alvarez et al., 2010);however, only a limited number of such targets are currently known.The final decision between cellular death and survival largely dependson the balance between intracellular formation of S1P and ceramides.Ceramides are metabolic precursors as well as signaling counterpartsof S1P, but their signaling is fully dependent on binding to intracellulartargets such as the inhibitor of PP2A (Mukhopadhyay et al., 2009; Kimet al., 2010) that leads to the activation of apoptosis as an overall out-come of excessive ceramide formation.

Tissue ischemia and the reoxygenation, which happens duringevents related to temporal arrest of blood flow, lead to a severe cellinjury with resulting tissue necrosis and apoptosis. Within a plethora ofmediators involved in this process, S1P occupies an important position

360 I.A. Gorshkova et al. / Life Sciences 93 (2013) 359–366

by providing tissue protection from ischemia/reperfusion-induced inju-ry. S1P-mediated signaling directly protects cells from apoptosis butalso tightens endothelium junctions, thus increasing overall integrity ofvasculature and decreasing vascular leak (McVerry and Garcia, 2005;Limaye, 2008; Skoura and Hla, 2009; Zhao et al., 2011). In the contextof cardiac ischemia/reperfusion-induced injury, a pioneering work byKarliner's group (Jin et al., 2002, 2004, 2007; Vessey et al., 2006, 2009;Zhang et al., 2007; Karliner, 2013) and by others (Knapp, 2011; VanBrocklyn and Williams, 2012) identified the S1P and S1PR1/3 signalingsystem to be critical for cardiac survival in ex vivo as well as in vivoexperimental settings. Furthermore, in humans, circulatory S1P levelsare known to inversely correlatewith the occurrence of ischemic disease(Argraves et al., 2011) and to decrease after the onset of myocardialinfarction (Knapp et al., 2009). This suggests the significance of a propermetabolic and circulatory “S1P tone” to maintain normal physiologicalresponses and tissue functioning. Of two enzymes directly responsi-ble for S1P biosynthesis, sphingosine kinase 1 (SphK1) rather thansphingosine kinase 2 (SphK2) is considered to be most critical forsignaling mechanisms involved in the protection of the heart fromischemia/reperfusion-induced injury (Jin et al., 2007, 2008). However,Sphk2 was also recently suggested to play a part in cardioprotectionby regulating mitochondrial permeability transition (Gomez et al.,2011). This last finding directly links S1P and sphingosine kinases toceramide/S1P homeostasis as mitochondrial permeability transition isknown to be a major target for ceramide-induced and PP2A-mediatedapoptotic signaling responses (Gudz et al., 1997; Ghafourifar et al.,1999; Siskind et al., 2006; Mukhopadhyay et al., 2009).

While the in vitro (Tao et al., 2007; Vessey et al., 2009) and ex vivo(Jin et al., 2004, 2007; Bandhuvula et al., 2011) data clearly suggesta protective role of SphK1/S1P signaling system for cardiomyocytesurvival and heart functionality during I/R, there is only a limited infor-mation (Duan et al., 2007) supporting their beneficial role in the in vivoischemia/reperfusion models targeting cardiovascular system. Wedecided to explore the impact of the global modulation of S1P meta-bolic and signaling homeostasis on animal resuscitation and short-term survival using the most drastic in vivo model of cardiac andglobal ischemia–reperfusion injury — the mouse model of cardiacarrest and resuscitation we have developed to model and to studysudden death (Abella et al., 2004; Zhao et al., 2008; Beiser et al.,2010, 2011). We applied this model to SphK1-KO mice in conjunc-tion with a manipulation in the endogenous S1P levels through thein vivo inhibition of S1P lyase, an enzyme which irreversiblydegrades S1P. Our study has confirmed the importance of SphK1and S1P/dihydroS1P homeostasis for animal resuscitation aftercardiac arrest (CA) and for the support of cardiovascular systemand survival after resuscitation.

Materials and methods

Standards and reagents

S1P, DHS1P, a 17-carbon analog of S1P (C17-S1P), individual N-14:0–24:1 ceramides and N-17:0-ceramide were obtained from AvantiPolar Lipids (Alabaster, AL, USA). 2-acetyl-5-tetrahydroxybutyl imidazole(THI) was purchased from Cayman Chemicals (Ann Arbor, MI).

Animal preparation and cardiac arrest protocol

All animal procedures were approved by the Institutional AnimalCare and Use Committee of the University of Illinois at Chicago andwere performed as previously described (Beiser et al., 2010, 2011).Briefly, five to eight months old (25–30 g) female C57BL/6 orSphK1-KO mice (global knockout) generated on C57BL/6 back-ground (Allende et al., 2004) were anesthetized with 80 μg/g ofketamine (Phoenix Scientific, St. Joseph, MO) and 12 μg/g xylazine(Ben Venue Laboratories, Bedford, OH) with periodic redosing of

20–30% of the initial ketamine dose as required to maintain surgicalanesthesia. Rectal temperature was monitored and maintained at37 ± 0.5 °C throughout the surgical preparation. Mice were orallyintubated and mechanically ventilated with a tidal volume of 12.5 μl/g,a respiratory rate of 110 breaths/min, and a positive end-expiratorypressure of 2 cm H2O. Microcatheters (BioTime, Berkeley, CA) wereinserted into the left jugular vein and right carotid artery for fluidadministration and blood pressure monitoring, respectively, and needleprobes were placed to provide three-lead ECG.

Animals were monitored for 50–60 min from the time of anesthesiainduction, then asystolic cardiac arrest was induced by the intravenousadministration of 0.08 mg/g potassium chloride solution (Sigma, St.Louis, MO). After 8 min of arrest, resuscitation was attempted withchest compressions and mechanical ventilation with an epinephrinebolus administered through the jugular vein catheter at 90 s after thebeginning of CPR (1.5 μg/mouse in 150 μl saline followed by a catheterflush with 250 μl saline). ROSC was defined as the return of sinusrhythm with a MAP of N40 mm Hg lasting at least 5 min. CPR wasterminated after 5 min or upon hemodynamic evidence of initial ROSC.Successfully resuscitated animals were maintained on the mechanicalventilator with heating support to maintain body temperature at37 ± 0.5 °C. Blood pressure and end-tidal CO2 (ETCO2) parameterswere monitored for up to 240 min during which time animals receivedscheduled intravenous injections of 0.9% saline at a rate of 100 μl/h.Protocol was terminated at T = 240 min (survived animals) or prema-turely and animals were considered dead when blood pressuredecreased to MAP of b30 mm Hg and stayed below 30 mm Hg for atleast 5 min.

Animal treatment with 2-acetyl-5-tetrahydroxybutyl imidazole (THI)

THI was dissolved in 10 g/L glucose in water (pH = 3.5). FemaleC57BL/6 mice (5–8 months old, 25–30 g) received THI at 10 μg/mousetwice a day for 3 days by gavage in a volume of 0.1 ml. Control femalemice received only solvent. The efficiency of the chosen regimen ofTHI application was confirmed in preliminary experiments whichshowed a substantial upregulation of circulatory and cardiac S1P levels.

Lipid extraction and sample preparation for ESI-LC/MS/MS

Tissue lipids were extracted by a modified Bligh and Dyer proce-dure (Bligh and Dyer, 1959) with the use of 0.1 N HCl for phase sep-aration as described (Berdyshev et al., 2006). C17-S1P (40 pmol)and C17-ceramide (30 pmol) were employed as internal standards,and were added during the initial step of lipid extraction. Theextracted lipids were dissolved in methanol/chloroform (4:1, v/v),and aliquots were taken to determine total phospholipid contentas described (Vaskovsky et al., 1975). Samples were concentratedunder a stream of nitrogen, re-dissolved in methanol, transferredto autosampler vials, and subjected to sphingolipid LC/MS/MS analysis.Plasma samples were processed similarly except that lipid phosphoruswas not determined and data were expressed per sample volume.

Analysis of sphingoid base-1-phosphates and ceramides

Analyses of sphingoid base-1-phosphates and ceramides wereperformed by electrospray ionization tandem mass spectrometry(ESI-LC/MS/MS). The instrumentation employed was an AB Sciex5500 QTRAP hybrid triple quadrupole linear ion-trap mass spec-trometers (Applied Biosystems, Foster City, CA, USA) equippedwith a turbo ion spray ionization source interfaced with an automatedAgilent 1200 series liquid chromatograph and autosampler (AgilentTechnologies, Wilmington, DE, USA). S1P and DHS1P were analyzed asbis-acetylated derivatives with C17-S1P as the internal standardemploying reverse-phase HPLC separation using Ascentis Express C8column (75 × 2.1 mm, 2.7 micron particle size) with a gradient elution

Table 1Baseline and resuscitation characteristics of control and THI‐treated mice.

C57BI/6 control SphK1-KO SphK1-KO, THItreated

n = 10 n = 10 n = 10

Baseline Weight (g) 28.6 ± 1.9 28.3 ± 4.4 26.7 ± 2.6Heart rate (bpm) 260 ± 41 282 ± 48 274 ± 35ETCO2 (mm Hg) 39.9 ± 3.8 35.6 ± 2.3 40.0 ± 3.9MAP (mm Hg) 81.5 ± 12.8 70.9 ± 17.2 82.1 ± 7.2

Resuscitation ROSC, n (%) 9 (90) 5 (50) 9 (90)ROSC time (s) 132 ± 22 153 ± 27 136 ± 21

150 min Survival, n (%) 7 (70) 5 (50) 9 (90)Heart rate (bpm) 442 ± 54 476 ± 56 488 ± 60ETCO2 (mm Hg) 38.1 ± 5.0 31.7 ± 3.7 35.3 ± 5.4

240 min Survival, n (%) 5 (50) 1 (10) 8 (80)Heart rate (bpm) 462 ± 79 572 483 ± 100ETCO2 (mm Hg) 33.8 ± 4.2 24.3 34.0 ± 6.6MAP (mm Hg) 46.4 ± 7.7 38.1 57.9 ± 18.2

Fig. 1. (A) Kaplan–Meier plot of wild-type (C57BL/6), SphK1-KO, and THI-treated SphK1-KOmice survival following cardiac arrest. * — p b 0.05 versus C57BL/6 mice; ## — p b 0.01versus non-treated SphK1-KO mice, N.S. — non-significant versus C57BL/6 mice. THI treat-ment of SphK1 animals conferred statistically significant survival benefit at T = 240 minpost-ROSC relative to non-treated SphK1-KO group (Mantel–Cox log-rank). (B) Meanarterial pressure (MAP) at T = 150 min post-ROSC in control C57BL/6, SphK1-KO, andTHI-treated SphK1-KO mice. THI treatment statistically significantly increased MAP inSphK1-KO animals in comparison with non-treated SphK1-KO mice.

361I.A. Gorshkova et al. / Life Sciences 93 (2013) 359–366

frommethanol:water:formic acid (60:40:0.5, v/v,with 5 mMammoniumformate) system to acetonitrile:water:formic acid (99:0.5:0.5, v/v, with5 mM ammonium formate) system, negative ion ESI, and MRM analysisessentially as described in (Berdyshev et al., 2005). Ceramides andsphingoid bases were analyzed with N-17:0-ceramide as internalstandards employing reverse-phase HPLC separation using AscentisExpress C8 column (75 × 2.1 mm, 2.7 micron particle size) with agradient elution from methanol:water:formic acid (60:40:0.5, v/v,with 5 mM ammonium formate) system to acetonitrile:chloroform:water:formic acid (80:20:0.5:0.5, v/v, with 5 mM ammonium formate)system, positive ion ESI, and MRM analysis essentially as described in(Berdyshev et al., 2006).

RNA isolation and real time RT-PCR

Total RNA was isolated from heart tissues using TRIzol® reagentaccording to the manufacturer's instruction. RNA (1 μg) was reverses-transcribed using cDNA synthesis kit (Bio-Rad), and real-time PCR andquantitative PCR were performed to assess expression of the S1PR1,2,3using primers designed for mouse mRNA sequences. iQ SYBR GreenSupermix was used to perform the real time measurements usingiCycler by BioRad. Amplicon expression in each sample was normalizedto 18S RNA content. Analysis of results and fold differences were deter-mined using the comparative CT method. Fold change was calculatedusing a comparative quantification algorithm from the ΔΔCt valueswith the formula (2−ΔΔCt), and data are presented as relative to theendogenous normalizer 18S mRNA expression.

Immunoblotting

Cardiac tissue was pulverized in liquid nitrogen then tissue sampleswere lysed with 100–300 μl cell lysis buffer (Cell Signaling Technology,Danvers, MA, USA) containing Halt Thermo-Fisher protease inhibitorcocktail (Fisher, Scientific, Pittsburgh, PA, USA). Tissue lysates werecleared by centrifugation at 5000 ×g for 10 min, and boiled with theLaemmli sample buffer for 5 min. Tissue lysates (20–30 μg protein)were separated on 10% or 4–20%NuPage precast gels (Life Technologies,Grand Island, NY, USA), transferred to PVDF membranes, blocked inTBST containing 5% BSA prior to incubation with primary antibodies(1:1000 dilution) overnight. After blocking, washing and incubationwith appropriate secondary antibody, blots were developed using anECL chemiluminescence kit. The bands of interest from immunoblotswere scanned by densitometry and integrated density of pixels in iden-tified areas was quantified using ImageJ (NIH, Bethesda, DC, USA).

Statistical analysis

A GraphPad Prizm 5.02 statistical package was used for statisticalanalyses. One way ANOVA with Bonferroni post-hoc test was usedto determine statistical difference between experimental groups.Differences between groups were considered statistically significantat p b 0.05. Results are expressed as means ± S.E.M. The log-rank(Mantel–Cox) test was applied to compare survival curves.

Results

SphK1-KO mice have limited resuscitation and survival after cardiac arrestthat is rescued by the inhibition of S1P lyase

When subjected to our standard cardiac arrest/return of sponta-neous circulation (CA/ROSC) protocol, SphK1-KO mice have shownpoor resuscitation and four hour survival rates. At baseline, SphK1-KOmice did not differ much from control mice in all measured phys-iological parameters with only a tendency for having a slightly lowermean arterial pressure (MAP) and higher heart rate parameters(Table 1). It should be noted that ketamine–xylazine-provided

anesthesia results in generally low baseline heart rate that is charac-teristic to employed model of cardiac arrest and post-resuscitationmonitoring (Beiser et al., 2010, 2011).

When subjected to CA/ROSC protocol, SphK1-KOmice demonstratedpoor ROSC and survival rates, with only one animal reaching our targetend-point criteria (MAP N 30 mm Hg at T = 240 min) (Table 1,Fig. 1A). Interestingly, most of resuscitated SphK1-KO mice did notdiffer from C57Bl/6 mice in terms of their MAP until T = 120 min(data not shown) but quickly dropped their blood pressure afterthat (Fig. 1B) and died around 3 h post-ROSC (Fig. 1A). Treatmentof SphK1-KO mice with 2-acetyl-5-tetrahydroxybutyl imidazole(THI), the in vivo inhibitor of S1P lyase that irreversibly degrades S1P

362 I.A. Gorshkova et al. / Life Sciences 93 (2013) 359–366

to Δ2-hexadecenal and ethanolamine phosphate, dramatically im-proved resuscitation and survival rates of SphK1-KO mice (Table 1,Fig. 1A). Moreover, both MAP at T = 150 min and survival rate param-eters in THI-treated SphK1-KO mice were the same or slightly betterthan corresponding parameters in control C57BL/6 mice (Table 1,Fig. 1A, B). These data point out at the importance of a proper S1P ho-meostasis for a proper cardiovascular recovery and functioning aftercardiac arrest.

Rescue of SphK1-KO phenotype through the inhibition of S1P lyase is linkedto normalization of disbalanced S1P and ceramide homeostasis

THI inhibits S1P lyase in vivo and up-regulates circulatory and tissueS1P levels (Schwab et al., 2005). To test if the chosen regimen of THIapplication (10 μg/mouse p.o. 2×/day × 3 days), instead of the usuallyemployed 50 mg/L THI in drinking water, ad libitum, provides measur-able changes in tissue and circulatory S1P levels, we treated naïve con-trol C57BL/6 and SphK1-KO mice with THI as noted above or with itscarrier (5 g/L glucose in water, pH = 3.5) and analyzed cardiac tis-sue and plasma levels of S1P, its endogenous analog DHS1P, and

Fig. 2. The effect of THI treatment on plasma and heart levels of sphingosine-1-phosphate and dihwith THI (10 μg/mouse p.o. 2×/day × 3 days) and the content of sphingoid base-1-phosphateswatissue S1P levels (A, C) in naïve animals but onlyminimally affected cardiac tissue and circulatoryversus control C57BL/6 group. N.S.— non significant.

S1P metabolic precursors and signaling counterparts, ceramides, bythe LC/MS/MS. We found that circulatory and cardiac levels of S1Pand DHS1P in naïve animals were substantially reduced in SphK1-KO mice in comparison with control C57BL/6 mice (Fig. 2A–D).Animal treatment with THI significantly increased S1P levels (Fig. 2A,C) but only minimally upregulated DHS1P levels in both cardiac tissueand in plasma in control as well as in SphK1-KO mice (Fig. 2B, D).Importantly, circulatory and cardiac levels of S1P-DHS1P in THI-treatednaïve SphK1-KO mice approached or even exceeded their levels innon-treated C57BL/6mice (Fig. 2). Thus, SphK1-KO phenotype is charac-terized by lower cardiac and circulatory levels of S1P and DHS1P, whichcan be upregulated through the inhibition of S1P lyase.

Next, we looked at the molecular species of ceramides in heart andplasma of naïve control and THI-treated animals. Naïve SphK1-KOmice had triple amount of ceramides in plasma compared to C57BL/6micewhile the heart tissue of naïve SphK1-KOmice contained substan-tially less ceramides than the heart tissue of C57BL/6 mice (Fig. 3A, B).This increase in plasma ceramides in SphK1-KO was clearly associatedwith a preferred upregulation of the N-behenoyl ceramide (22:0-ceramide) molecular species (Fig. 3C) with the exception for 16:0-

ydrosphingosine-1-phosphate in naïve C57BL/6 and SphK1-KOmice. Animals were treateds analyzed by the LC/MS/MS. THI treatment substantially upregulated circulatory and cardiaclevels of DHS1P (B,D). Data are presented asmean ± S.E.M. n = 3–4 per group. *— p b 0.05

Fig. 3. The effect of THI treatment on plasma and heart levels of ceramides in naïve C57BL/6 and SphK1-KOmice. Animals were treatedwith THI (10 μg/mouse p.o. 2×/day × 3 days) andthe content of ceramideswas analyzed by the LC/MS/MS. (A) The effect of THI treatment on heart tissue level of total ceramides. Note that cardiac tissue of non-treated SphK1-KOmice hassubstantially lower content of ceramides in comparison to C57BL/6 mice which is fully restored by THI treatment. (B) The effect of THI treatment on plasma level of total ceramide and(C) individual ceramide molecular species. Note that ceramide plasma content in naïve SphK1 mice is tripled in comparison to their level in plasma of C57BL/6 mice mostly due to aselective upregulation of N-22:0-ceramide content. Also note that naïve SphK1-KO mice are characterized by a very low circulatory level of N-16:0-ceramide and that THI treatmentnormalized plasma level of N-22:0-ceramide but not that of N-16:0-ceramide. Data are presented as mean ± S.E.M. n = 3–4 per group; N.S. — non significant.

363I.A. Gorshkova et al. / Life Sciences 93 (2013) 359–366

ceramide, which level was drastically reduced. The inhibition of S1Plyase by THI completely abolished this ceramide SphK1-KO-specificphenotype and decreased plasma ceramide levels, and specificallythat of 22:0-ceramide, to the level observed in control C57BL/6 mice.However, the inhibition of S1P lyase did not affect 16:0-ceramidelevel in plasma of SphK1-KO mice (Fig. 3C) thus pointing out at thelink between 16:0-ceramide formation and the functional SphK1 ratherthan the S1P levels. In contrast, heart tissue ceramide levels in naïveSphK1-KO mice were lower than that in the heart of C57BL/6 mice butincreased as a result of S1P lyase inhibition without demonstratingany preference toward any individual ceramide molecular species(data not shown). These results show that in SphK1-KO mice, plasmaand heart show an inverse pattern of ceramide dynamics in responseto S1P lyase inhibition by THI (Fig. 3A).

Survival of THI-treated SphK1-KO mice is linked with better stability oftissue DHS1P-S1P levels post-resuscitation

To test if the stability of cardiac tissue levels of S1P-DHS1P is linkedto animal survival post-resuscitation, we compared cardiac S1P-DHS1Plevels in sham-operated, non-surviving, and in surviving non-treatedand THI-treated SphK1-KO animals. The LC/MS/MS analysis of cardiacS1P-DHS1P levels revealed a potential link between animal survivaland the stability of cardiac DHS1P rather than S1P levels. We foundthat there were some minor differences in the effect of THI on cardiacS1P-DHS1P levels in sham-operated (Fig. 4A, B) versus naïve (Fig. 2)SphK1-KO mice most probably due to the fact that sham-operatedanimals were kept on ventilator for about an hour from the beginningof the surgery to stabilize blood pressure and heart rate parametersbefore tissue collection while naïve animals were not subjected tosurgery and were sacrificed immediately after attaining anesthesia.Regardless of these minor differences, survival of THI-treated SphK1-KO mice correlated better with stably elevated tissue DHS1P (Fig. 4B)rather than with S1P (Fig. 4A) level even at the end of 4 h post-ROSCperiod. Interestingly, and in full concordance with the critical role ofS1P-DHS1P in protection from ischemia/reperfusion-induced injury,the DHS1P-S1P level in cardiac tissue in each of either non-survived ornon-resuscitated THI-treated animal was below that in any animalfrom sham-operated control or survived until 4 h post-ROSC groups(Fig. 4). Further, the background S1P-DHS1P tissue levels were practi-cally the same for sham-operated, non-survived, or non-resuscitated

SphK1-KO mice that correlates with observed difficulties to resuscitatethese animals and suggests the requirement for a minimum (DH)S1Ptone to support cardiac function after cardiac arrest.

S1P lyase inhibition results in upregulation of S1PR2 expression

S1P receptors 1, 2 and 3 have been implicated in cardioprotection(Means et al., 2007; Zhang et al., 2007; Tao et al., 2009). Therefore, wetested if SphK1-KO phenotype and S1P lyase inhibition affect theexpression of S1PR1–3. Real-time qPCR analysis revealed negative effectof SphK1deletion on the expression of S1PR1 and S1PR2 but not S1PR3 incomparison to their mRNA levels in C57BL/6 mice (Fig. 5). Importantly,the inhibition of S1P lyase with THI normalized the expression level ofS1PR2 and partially restored the expression of S1PR1 in SphK1-KOmice; however it did not affect mRNA level for S1PR3. These datasuggest a potential link between S1PR2 and to a lesser degree S1PR1,

but not S1PR3, expression in the heart with tissue and/or circulatorylevels of S1P.

Discussion

The mouse model of cardiac arrest and resuscitation has been suc-cessfully employed by our group in recent years to study the effect ofhypothermia, Akt, and NOS3 signaling in the most drastic setting ofischemia/reperfusion injury (Abella et al., 2004; Zhao et al., 2008;Beiser et al., 2010, 2011). This model includes potassium chloride-induced eight minute cardiac arrest and five minute cardio-pulmonaryresuscitation (CPR) with epinephrine bolus applied at 90 s after thebeginning of CPR (Beiser et al., 2010, 2011). In this model, 25–30 gfemale C57BL/6 mice have a historically stable ROSC rate (about 80%)and four hour survival rate (around 50%) (Beiser et al., 2010, 2011).Importantly, the four hour survival rate of female C57BL/6 mice atabout 40–50% ensures good comparison with animals experimentallyaffected to either deteriorate or to support pro-surviving signaling andsurvival when even relatively small animal numbers are employed.

The S1P signaling system is well documented to be vital for resis-tance to cellular stress challenges and in particular for protection fromischemia/reperfusion injury (Hofmann et al., 2009; Knapp, 2011; VanBrocklyn and Williams, 2012; Karliner, 2013). Yet, there is only oneknown reportwhere a beneficial role for SphK1/S1P-mediated signalingwas shown in the in vivo model where SphK1 was adenovirally

Fig. 4. Four hour survival of THI-treated SphK1-KOmice after cardiac arrest is associatedwiththe stability of upregulated cardiac DHS1P level. The content of S1P (A) and DHS1P (B) wasanalyzed in the heart tissue at respected time point or at four hours post-resuscitation. Non-survivors are defined as animals with MAP b 30 mm Hg. Sham-operated animals weremaintained on ventilator for the same time as experimental animals until the initiation ofcardiac arrest. Note that THI-treated SphK1-KO mice that survived until 4 h post-ROSChave statistically significantly upregulated tissueDHS1P levelwhile S1P content did not dem-onstrate strong relationship with survival. Also, note that each of non-ROSC/non-survivedTHI-treated SphK1-KO mice had the lowest measured levels of S1P and DHS1P. Data arepresented as individual measurements with the mean ± S.E.M.

364 I.A. Gorshkova et al. / Life Sciences 93 (2013) 359–366

transfected to animals at the moment of acute left anterior descendingcoronary artery ligation that resulted in decreased infarct size, collagencontent, increased neovascularization, and improved left ventricularfunction two weeks post-intervention (Duan et al., 2007). Therefore,we further addressed the vital role of SphK1/S1P homeostasis forcardioprotection and survival using genetically engineeredmice lackingSphK1 and the in vivo model of sudden death. Of the two sphingosine

kinase isoforms, SphK1 is most often implicated in pro-survivalsignaling (Limaye, 2008; Shida et al., 2008; Karliner, 2013); therefore,when designing the study, we expected to observe a poor survival phe-notype in SphK1-KOmice. To overcome decreased tissue and circulato-ry S1P levels due to the lack of SphK1, we also introduced an in vivoinhibition of S1P lyase with THI. S1P lyase irreversibly degrades S1P/DHS1P to ethanolamine phosphate and Δ2-hexadecenal/hexadecanal(Saba and Hla, 2004) and is considered to be the most critical catabolicenzyme controlling intracellular S1P level. Therefore, in the presence offunctional SphK2, S1P lyase inhibition should increase S1P/DHS1P sta-bility and their intracellular and extracellular levels and signaling.

In full agreementwith our prediction, SphK1-KOmice demonstrateda very poor rate of resuscitation and minimal survival until four hourspost-ROSC (Table 1, Fig. 1). The poor resistance of SphK1-KO mice toischemia/reperfusion injury was clearly associated with substantiallydecreased cardiac and circulatory levels of S1P and DHS1P (Fig. 2), andnot with the lack of functional SphK1, as the upregulation of S1P-DHS1Plevels in the heart and in circulation through the inhibition of S1P lyasemade SphK1-KO mice extremely protected against global I/R-inducedinjury (Table 1, Figs. 1, 2, 4). These findings are fully supported by earlierobservations showing the ability of S1P to protect SphK1-KO mousehearts ex vivo (Jin et al., 2007) and the beneficial effect of S1P lyase inhi-bition by THI on functional recovery and infarct size in mouse heartssubjected to ischemia and reperfusion ex vivo (Bandhuvula et al., 2011).Our experiments also revealed a surprisingly better association of survivalof THI-treated SphK1-KO mice with cardiac DHS1P rather than S1P level(Fig. 4) as half of survived THI-treated animals had cardiac S1P level aslow as the non-survived and even non-resuscitated SphK1-KO mice. Onthe contrary, all survived THI-treated SphK1-KO animals maintainedcardiac DHS1P level above its level in any of the non-treated SphK1-KOmice suggesting the requirement for a particular “DHS1P cellular tone”to support cardiac function after ischemia and reperfusion.

It is known that external S1P-initiated signaling persists to providepro-survival signaling in the presence of functionally impaired SphK1(Jin et al., 2007). S1P exerts its cardioprotection through S1PR1–3, ofwhich S1PR1 is the most abundantly expressed in cardiomyocytes(Nakajima et al., 2000; Mazurais et al., 2002; Zhang et al., 2007).S1PR1 (Zhang et al., 2007) as well as S1PR2 and S1PR3 (Means et al.,2007) were shown to mediate S1P-induced Akt phosphorylation inmouse cardiomyocytes and to confer S1P-induced cardioprotection. Inour hands, we could not clearly relate Akt phosphorylation status withthe survival of THI-treated SphK1-KO animals due to most probablysignificant differences in time passed after cardiac arrest until tissuecollectionbetween surviving andnon-surviving animals (SupplementaryFig. 1). However, our quantitative analysis of S1PR1–3 expression in theheart of SphK1-KOmice by qRT-PCR revealed that the lack of functioningSphK1 and most probably the decrease in tissue/circulatory S1P-DHS1P(Fig. 2) results in diminished expression of S1PR1 and S1PR2 but notS1PR3 in the heart of SphK1-KO mice in comparison to their expressionlevel in the heart of wild-type C57BL/6 mice (Fig. 5). The fact that theinhibition of S1P lyase restored the level of S1PR1–2 expression to thatin wild-type mice clearly indicates that the expression of S1PR1–2 iscontrolled in part by intracellular/extracellular S1P-DHS1P levels.Interestingly, S1PR3 expression seems to be affected by neither thelack of SphK1 nor the upregulation of S1P-DHS1P levels (Fig. 5). Fur-ther studies are necessary regarding “S1P sensing” and regulation ofS1PR1–2 expression as the decrease in S1P receptor expression aftermyocardial infarction might be part of the global mechanism leadingto poor survival and recovery. Several research groups have showndecreased plasma levels for S1P-DHS1P after myocardial infarctionin rats (Knapp et al., 2012) as well as in humans (Knapp et al.,2009; Sattler et al., 2010; Argraves et al., 2011), which persisted forup to 5 days (Knapp et al., 2009). This may provoke a decrease inS1PR1,2 expression as a result of ischemia/reperfusion injury andfurther decrease of already diminished pro-survival tone of S1P-induced and S1PR-mediated signaling. In this regard, a supportive

Fig. 5. Cardiac expression of S1PR2 and S1PR1 but not S1PR3 is affected by the lack of functional SphK1 and by THI treatment. The expression of major cardiac S1P receptors was measuredby qRT-PCR. Strong association was found between the presence of functional SphK1 and the expression of S1PR2 and S1PR1. THI treatment of SphK1-KO mice restored mRNA level ofcardiac S1PR2 and to a lesser degree that of S1PR1 but did not affect S1PR3 expression. Data are presented as mean ± S.E.M. n = 3–5 animals per group.

365I.A. Gorshkova et al. / Life Sciences 93 (2013) 359–366

therapy with S1P lyase inhibitors may provide a solution for sustainedupregulation of S1PR expression and S1P-induced cardiomyocytesurvival. At the global level, increased (DH)S1P tone and S1PR1 expres-sion will be especially beneficial for maintaining vascular integrity(McVerry and Garcia, 2005; Limaye, 2008; Zhao et al., 2011) and atten-uating post-infarction inflammatory responses through the decrease inlymphocyte egress from lymph nodes (Saba and Hla, 2004; Schwabet al., 2005).

Our study also revealed a previously unknown link between SphK1expression and circulatory levels of two particular ceramide molecularspecies, N-16:0- and N-22:0-ceramides. Ceramides are S1P signalingcounterparts (Pyne and Pyne, 2010; Van Brocklyn and Williams, 2012)and play a critical role in the initiation of apoptosis through the activa-tion of PP2A-mediated pathway (Mukhopadhyay et al., 2009) and bydirectly affecting mitochondria permeability transition (Gudz et al.,1997; Ghafourifar et al., 1999; Siskind et al., 2006). While the cardiacceramide levels decreased in SphK1-KO mice in comparison withC57BL/6 mice (Fig. 3A) without any specificity toward any particularceramidemolecular species (data not shown), ceramide levels in plasmaincreased about three-fold, withN-22:0-ceramide andN-16:0-ceramideclearly demonstrating differential dynamics (Fig. 3B, C). In particular,22:0-ceramide, being one the four major ceramide molecular speciesin mouse plasma, demonstrated the most striking eight-foldupregulation in the absolute amounts and three-fold increase ifexpressed as relative percent of all ceramide molecular species inSphK1 mice. On the contrary, the level of 16:0-ceramide was sub-stantially decreased in plasma of SphK1 mice. Surprisingly, S1Plyase inhibition with THI restored both cardiac and plasmaceramides in SphK1-KO mice to “wild-type” levels in all but plasmaN-16:0-ceramide levels (Fig. 3C). This observation suggests a defi-nite link between SphK1 protein expression and the ability torelease N-16:0-ceramide into circulation. On the contrary, plasma N-22:0-ceramide level seems to be under the control of circulating/tissueS1P-DHS1P levels and not of SphK1 protein asN-22:0-ceramide level inplasmawas themost upregulated by the lack of SphK1 and themost de-creased by THI inhibition. Currently, the role of circulating N-22:0-ceramide in resuscitation and survival after cardiac arrest isunclear and further studies are necessary. Argraves et al. (2011)

found negative association between serum N-24:1-ceramide levelsand ischemic heart disease in humans. On the contrary, Knapp et al.(2012) found an increase in plasma ceramides in a ratmodel ofmyocar-dial infarction. Separate studies are required to understand reasons andconsequences of the observed relationship between SphK1/S1P andN-22:0-ceramide and if this metabolic ceramide imbalance contributesto poor survival of SphK1-KO mice after cardiac arrest.

In conclusion, our study suggests a potential strategy for noveltherapeutic interventions applicable to cardiac arrest patients fromthe moment of CPR through post-resuscitation support. The associa-tion found between tissue and circulatory S1P-DHS1P levels and thesurvival rate after resuscitation suggests a potential benefit for animmediate i.v. infusion therapy by S1P/DHS1P or S1PR1,2 ligandsthat would help resuscitation and support heart function earlypost-ROSC. Thereafter, patients can be treated with the inhibitor(s)of S1P lyase that will provide global long-term upregulation of S1P-DHS1P metabolic and signaling tone to improve vascular integrity,cardiac contractility, and to counteract post-resuscitation inflamma-tory response associated with poor survival of resuscitated patients.While further studies are required to test this proposed strategy, ourresults show that (DH)S1P signaling system is critical for resuscita-tion and survival after cardiac arrest.

Supplementary data to this article can be found online at http://dx.doi.org/10.1016/j.lfs.2013.07.017.

Conflict of interest statement

No conflict of interests.

Acknowledgments

This work was supported by American Heart Association SDGgrant0930028N to EVB.

References

Abella BS, Zhao D, Alvarado J, Hamann K, Vanden Hoek TL, Becker LB. Intra-arrest coolingimproves outcomes in a murine cardiac arrest model. Circulation 2004;109:2786–91.

366 I.A. Gorshkova et al. / Life Sciences 93 (2013) 359–366

Allende ML, Sasaki T, Kawai H, Olivera A, Mi Y, van Echten-Deckert G, et al. Mice deficientin sphingosine kinase 1 are rendered lymphopenic by FTY720. J Biol Chem 2004;279:52487–92.

Alvarez SE, Harikumar KB, Hait NC, Allegood J, Strub GM, Kim EY, et al. Sphingosine-1-phosphate is a missing cofactor for the E3 ubiquitin ligase TRAF2. Nature2010;465:1084–8.

Argraves KM, Sethi AA, Gazzolo PJ, Wilkerson BA, Remaley AT, Tybjaerg-Hansen A, et al.S1P, dihydro-S1P and C24:1-ceramide levels in the HDL-containing fraction ofserum inversely correlate with occurrence of ischemic heart disease. Lipids HealthDis 2011;10:70.

Bandhuvula P, Honbo N, Wang GY, Jin ZQ, Fyrst H, Zhang M, et al. S1P lyase: a noveltherapeutic target for ischemia–reperfusion injury of the heart. Am J PhysiolHeart Circ Physiol 2011;300:H1753–61.

Beiser DG, Wojcik KR, Zhao D, Orbelyan GA, Hamann KJ, Vanden Hoek TL. Akt1 geneticdeficiency limits hypothermia cardioprotection following murine cardiac arrest. AmJ Physiol Heart Circ Physiol 2010;298:H1761–8.

Beiser DG, Orbelyan GA, Inouye BT, Costakis JG, Hamann KJ, McNally EM, et al. Geneticdeletion of NOS3 increases lethal cardiac dysfunction following mouse cardiac arrest.Resuscitation 2011;82:115–21.

Berdyshev EV, Gorshkova IA, Garcia JG, Natarajan V, HubbardWC. Quantitative analysis ofsphingoid base-1-phosphates as bisacetylated derivatives by liquid chromatography-tandem mass spectrometry. Anal Biochem 2005;339:129–36.

Berdyshev EV, Gorshkova IA, Usatyuk P, Zhao Y, Saatian B, Hubbard W, et al. De novobiosynthesis of dihydrosphingosine-1-phosphate by sphingosinekinase 1 inmammaliancells. Cell Signal 2006;18:1779–92.

Bligh EG, DyerWJ. A rapid method of total lipid extraction and purification. Can J BiochemPhysiol 1959;37:911–7.

Cowart LA, Shotwell M, Worley ML, Richards AJ, Montefusco DJ, Hannun YA, et al. Reveal-ing a signaling role of phytosphingosine-1-phosphate in yeast. Mol Syst Biol 2010;6:349. http://dx.doi.org/10.1038/msb.2010.3.

Diab KJ, Adamowicz JJ, Kamocki K, Rush NI, Garrison J, Gu Y, et al. Stimulation of sphingo-sine 1-phosphate signaling as an alveolar cell survival strategy in emphysema. Am JRespir Crit Care Med 2010;181:344–52.

Duan HF, Wang H, Yi J, Liu HJ, Zhang QW, Li LB, et al. Adenoviral gene transfer of sphin-gosine kinase 1 protects heart against ischemia/reperfusion-induced injury andattenuates its postischemic failure. Hum Gene Ther 2007;18:1119–28.

Ghafourifar P, Klein SD, Schucht O, Schenk U, Pruschy M, Rocha S, et al. Ceramide inducescytochrome c release from isolatedmitochondria. Importance of mitochondrial redoxstate. J Biol Chem 1999;274:6080–4.

Gomez L, PaillardM, PriceM, Chen Q, Teixeira G, Spiegel S, et al. A novel role for mitochon-drial sphingosine-1-phosphate produced by sphingosine kinase-2 in PTP-mediatedcell survival during cardioprotection. Basic Res Cardiol 2011;106:1341–53.

Gudz TI, Tserng KY, Hoppel CL. Direct inhibition of mitochondrial respiratory chain com-plex III by cell-permeable ceramide. J Biol Chem 1997;272:24154–8.

Hofmann U, Burkard N, Vogt C, Thoma A, Frantz S, Ertl G, et al. Protective effectsof sphingosine-1-phosphate receptor agonist treatment after myocardialischaemia–reperfusion. Cardiovasc Res 2009;83:285–93.

Jin ZQ, Zhou HZ, Zhu P, Honbo N, Mochly-Rosen D, Messing RO, et al. Cardio-protection mediated by sphingosine-1-phosphate and ganglioside GM-1 inwild-type and PKC epsilon knockout mouse hearts. Am J Physiol Heart CircPhysiol 2002;282:H1970–7.

Jin ZQ, Goetzl EJ, Karliner JS. Sphingosine kinase activation mediates ischemicpreconditioning in murine heart. Circulation 2004;110:1980–9.

Jin ZQ, Zhang J, Huang Y, Hoover HE, Vessey DA, Karliner JS. A sphingosine kinase 1mutation sensitizes the myocardium to ischemia/reperfusion injury. Cardiovasc Res2007;76:41–50.

Jin ZQ, Karliner JS, Vessey DA. Ischaemic postconditioning protects isolated mouse heartsagainst ischaemia/reperfusion injury via sphingosine kinase isoform-1 activation.Cardiovasc Res 2008;79:134–40.

Karliner JS. Sphingosine kinase and sphingosine 1-phosphate in the heart: a decade ofprogress. Biochim Biophys Acta 2013;1831:203–12.

Kim SW, Kim HJ, Chun YJ, Kim MY. Ceramide produces apoptosis through induction ofp27(kip1) by protein phosphatase 2A-dependent Akt dephosphorylation in PC-3prostate cancer cells. J Toxicol Environ Health A 2010;73:1465–76.

Knapp M, Baranowski M, Czarnowski D, Lisowska A, Zabielski P, Górski J, et al. Plasmasphingosine-1-phosphate concentration is reduced in patients with myocardialinfarction. Med Sci Monit 2009;15:CR490–3.

Knapp M. Cardioprotective role of sphingosine-1-phosphate. J Physiol Pharmacol2011;62:601–7.

Knapp M, Zendzian-Piotrowska M, Błachnio-Zabielska A, Zabielski P, Kurek K, GórskiJ. Myocardial infarction differentially alters sphingolipid levels in plasma, eryth-rocytes and platelets of the rat. Basic Res Cardiol 2012;107:294.

Limaye V. The role of sphingosine kinase and sphingosine-1-phosphate in the regulationof endothelial cell biology. Endothelium 2008;15:101–12.

Mao C, Saba JD, Obeid LM. The dihydrosphingosine-1-phosphate phosphatases ofSaccharomyces cerevisiae are important regulators of cell proliferation andheat stress responses. Biochem J 1999;342:667–75.

Mazurais D, Robert P, Gout B, Berrebi-Bertrand I, Laville MP, Calmels T. Cell type-specificlocalization of human cardiac S1P receptors. J Histochem Cytochem 2002;50:661–70.

McVerry BJ, Garcia JG. In vitro and in vivo modulation of vascular barrier integrity bysphingosine 1-phosphate: mechanistic insights. Cell Signal 2005;17:131–9.

Means CK, Xiao CY, Li Z, Zhang T, Omens JH, Ishii I, et al. Sphingosine 1-phosphate S1P2and S1P3 receptor-mediated Akt activation protects against in vivo myocardialischemia–reperfusion injury. Am J Physiol Heart Circ Physiol 2007;292:H2944–51.

Means CK, Brown JH. Sphingosine-1-phosphate receptor signalling in the heart.Cardiovasc Res 2009;82:193–200.

Mukhopadhyay A, Saddoughi SA, Song P, Sultan I, Ponnusamy S, Senkal CE, et al. Directinteraction between the inhibitor 2 and ceramide via sphingolipid-protein bindingis involved in the regulation of protein phosphatase 2A activity and signaling.FASEB J 2009;23:751–63.

Nakajima N, Cavalli AL, Biral D, Glembotski CC, McDonough PM, Ho PD, et al. Expressionand characterization of Edg-1 receptors in rat cardiomyocytes: calcium deregulationin response to sphingosine 1-phosphate. Eur J Biochem 2000;267:5679–86.

Nikolova-Karakashian MN, Reid MB. Sphingolipid metabolism, oxidant signaling, andcontractile function of skeletal muscle. Antioxid Redox Signal 2011;15:2501–17.

Oskouian B, Saba JD. Cancer treatment strategies targeting sphingolipid metabolism. AdvExp Med Biol 2010;688:185–205.

Ponnusamy S, Meyers-NeedhamM, Senkal CE, Saddoughi SA, Sentelle D, Selvam SP, et al.Sphingolipids and cancer: ceramide and sphingosine-1-phosphate in the regulationof cell death and drug resistance. Future Oncol 2010;6:1603–24.

Pyne NJ, Pyne S. Sphingosine 1-phosphate and cancer. Nat Rev Cancer 2010;10:489–503.Saba JD, Hla T. Point-counterpoint of sphingosine 1-phosphate metabolism. Circ Res

2004;94:724–34.Sattler KJ, Elbasan S, Keul P, Elter-Schulz M, Bode C, Gräler MH, et al. Sphingosine

1-phosphate levels in plasma and HDL are altered in coronary artery disease. BasicRes Cardiol 2010;105:821–32.

Schwab SR, Pereira JP, Matloubian M, Xu Y, Huang Y, Cyster JG. Lymphocyte sequestrationthrough S1P lyase inhibition and disruption of S1P gradients. Science 2005;309:1735–9.

Shida D, Takabe K, Kapitonov D, Milstien S, Spiegel S. Targeting SphK1 as a new strategyagainst cancer. Curr Drug Targets 2008;9:662–73.

Siskind LJ, Kolesnick RN, Colombini M. Ceramide forms channels in mitochondrial outermembranes at physiologically relevant concentrations. Mitochondrion 2006;6:118–25.

Skoura A, Hla T. Lysophospholipid receptors in vertebrate development, physiology, andpathology. J Lipid Res 2009;50:S293–8.

Suomalainen L, Pentikäinen V, Dunkel L. Sphingosine-1-phosphate inhibits nuclear factorkappaB activation and germ cell apoptosis in the human testis independently of itsreceptors. Am J Pathol 2005;166:773–81.

Tao R, Zhang J, Vessey DA, Honbo N, Karliner JS. Deletion of the sphingosine kinase-1 geneinfluences cell fate during hypoxia and glucose deprivation in adult mousecardiomyocytes. Cardiovasc Res 2007;74:56–63.

Tao R, Hoover HE, Zhang J, Honbo N, Alano CC, Karliner JS. Cardiomyocyte S1P1receptor-mediated extracellular signal-related kinase signaling and desensitization.J Cardiovasc Pharmacol 2009;53:486–94.

Uhlig S, Gulbins E. Sphingolipids in the lungs. Am J Respir Crit Care Med 2008;178:1100–14.Van Brocklyn JR, Williams JB. The control of the balance between ceramide and

sphingosine-1-phosphate by sphingosine kinase: oxidative stress and the see-saw of cell survival and death. Comp Biochem Physiol B Biochem Mol Biol2012;163:26–36.

Vaskovsky VE, Kostetsky EY, Vasendin IM. A universal reagent for phospholipidanalysis. J Chromatogr 1975;114:129–41.

Vessey DA, Kelley M, Li L, Huang Y, Zhou HZ, Zhu BQ, et al. Role of sphingosine kinaseactivity in protection of heart against ischemia reperfusion injury. Med Sci Monit2006;12:BR318–24.

Vessey DA, Li L, Honbo N, Karliner JS. Sphingosine 1-phosphate is an important endoge-nous cardioprotectant released by ischemic pre- and postconditioning. Am J PhysiolHeart Circ Physiol 2009;297:H1429–35.

Watters RJ, Wang HG, Sung SS, Loughran TP, Liu X. Targeting sphingosine-1-phosphatereceptors in cancer. Anticancer Agents Med Chem 2011;11:810–7.

Zhang J, Honbo N, Goetzl EJ, Chatterjee K, Karliner JS, Gray MO. Signals from type 1sphingosine 1-phosphate receptors enhance adult mouse cardiac myocytesurvival during hypoxia. Am J Physiol Heart Circ Physiol 2007;293:H3150–8.

Zhao D, Abella BS, Beiser DG, Alvarado JP, Wang H, Hamann KJ, et al. Intra-arrest coolingwith delayed reperfusion yields higher survival than earlier normothermic resuscita-tion in a mouse model of cardiac arrest. Resuscitation 2008;77:242–9.

Zhao Y, Gorshkova IA, Berdyshev E, He D, Fu P, Ma W, et al. Protection of LPS-inducedmurine acute lung injury by sphingosine-1-phosphate lyase suppression. Am J RespirCell Mol Biol 2011;45:426–35.