Loss of Mouse P2Y6 Nucleotide Receptor Is Associated with Physiological Macrocardia and Amplified Pathological Cardiac Hypertrophy*

The study of the mechanisms leading to cardiac hypertrophy is essential to better understand cardiac development and regeneration. Pathological conditions such as ischemia or pressure overload can induce a release of extracellular nucleotides within the heart. We recently investigated the potential role of nucleotide P2Y receptors in cardiac development. We showed that adult P2Y4-null mice displayed microcardia resulting from defective cardiac angiogenesis. Here we show that loss of another P2Y subtype called P2Y6, a UDP receptor, was associated with a macrocardia phenotype and amplified pathological cardiac hypertrophy. Cardiomyocyte proliferation and size were increased in vivo in hearts of P2Y6-null neonates, resulting in enhanced postnatal heart growth. We then observed that loss of P2Y6 receptor enhanced pathological cardiac hypertrophy induced after isoproterenol injection. We identified an inhibitory effect of UDP on in vitro isoproterenol-induced cardiomyocyte hyperplasia and hypertrophy. The present study identifies mouse P2Y6 receptor as a regulator of cardiac development and cardiomyocyte function. P2Y6 receptor could constitute a therapeutic target to regulate cardiac hypertrophy.

The study of the mechanisms leading to cardiac hypertrophy is essential to better understand cardiac development and regeneration. Pathological conditions such as ischemia or pressure overload can induce a release of extracellular nucleotides within the heart. We recently investigated the potential role of nucleotide P2Y receptors in cardiac development. We showed that adult P2Y 4 -null mice displayed microcardia resulting from defective cardiac angiogenesis. Here we show that loss of another P2Y subtype called P2Y 6 , a UDP receptor, was associated with a macrocardia phenotype and amplified pathological cardiac hypertrophy. Cardiomyocyte proliferation and size were increased in vivo in hearts of P2Y 6 -null neonates, resulting in enhanced postnatal heart growth. We then observed that loss of P2Y 6 receptor enhanced pathological cardiac hypertrophy induced after isoproterenol injection. We identified an inhibitory effect of UDP on in vitro isoproterenol-induced cardiomyocyte hyperplasia and hypertrophy. The present study identifies mouse P2Y 6 receptor as a regulator of cardiac development and cardiomyocyte function. P2Y 6 receptor could constitute a therapeutic target to regulate cardiac hypertrophy.
Cardiovascular pathologies are the main cause of mortality in developed countries. Heart failure is characterized by cardiac hypertrophy and fibrosis. Pathological cardiac hypertrophy can be caused inter alia by hypertension, myocardial infarct, or diabetes and is associated with cardiac cell death, cardiomyocyte hypertrophy, and increased proliferation of cardiac fibroblasts.
The role of P2Y nucleotide receptors in the heart has not been extensively studied. Cardioprotective effect of UTP has been previously described: treatment with UTP prior to myocardial infarction leads to reduced ischemic damage through P2Y 2 receptor activation (1,2). Another UTP receptor, P2Y 4 , plays an important role in the heart: P2Y 4 knock-out (KO) mice display reduced cardiac angiogenesis and cardiac postnatal development (3), as well as lower exercise capacity and reduced exercise-induced cardiac hypertrophy (4). More recently, it has been shown that P2Y 4 KO mice display lower infarct size and reduced cardiac inflammation and fibrosis in a model of ligation of the left anterior descending artery (5). The study of the role of P2Y 6 UDP receptor in the heart has also been initiated. UDP can have an inotropic effect on cardiomyocytes, mediated by P2Y 6 receptor (6). Nishida et al. (7) showed that inhibition of P2Y 6 receptor using the specific antagonist MRS2578 decreased collagen deposition after transverse aortic constriction, without affecting cardiac hypertrophy induced in this model. Also, P2Y 6 receptor seems to have a deleterious role in atherosclerosis, being enriched in the sclerotic lesions and favoring inflammation (8,9). These various effects of nucleotides are depending on cell-specific expression of P2Y subtypes involved: mouse P2Y 4 receptor is not expressed in cardiomyocytes and is mainly expressed in cardiac endothelial cells, whereas mouse P2Y 2 and P2Y 6 receptors were found in all tested cardiac cells (2,3,6,7).
We decided here to investigate P2Y 6 involvement in cardiac hypertrophy by analyzing postnatal heart development of P2Y 6 -null mice and by using P2Y 6 -null mice in a pathological cardiac hypertrophy-induced model. Cardiac hypertrophy is a hallmark of cardiac remodeling. There are two types of cardiac hypertrophy: physiological and pathological hypertrophy. Physiological hypertrophy arises during embryonic and postnatal development, pregnancy, or chronic physical exercise. This hypertrophy is reversible and characterized by normal or increased cardiac function, associated with normal sarcomeric organization and gene expression (10). Pathological hypertrophy is associated with re-expression of fetal genes considered as hypertrophy markers: Acta1 (skeletal muscle actin ␣1), Nppa (atrial natriuretic peptide), and Myh7 (myosin heavy chain ␤), and develops in response to pressure overload, myocardial infarct, or diabetes and as a consequence of genetic mutations (11), and is linked to cell death leading to fibrosis. Fibrosis decreases cardiac contractility and capillary density, which decreases oxygen availability and leads to cardiac ischemia and heart failure. This kind of hypertrophy is irreversible.
The presence of the P2Y 6 receptor in mouse cardiomyocytes and its involvement in cardiac fibrosis (6, 7) led us to study its potential role in cardiac development. We analyzed the role of this UDP receptor in physiological and pathological cardiac hypertrophy using P2Y 6 -null mice. We used the model of intraperitoneal injection of isoproterenol (ISO), 5 a non-selective agonist of ␤-adrenergic receptors, which is a model frequently used to induce pathological cardiac hypertrophy (12)(13)(14).

Experimental Procedures
Ethics Statement-C57Bl6/J P2Y 6 knock-out (P2Y 6 KO) male mice were obtained as previously described (15). Animal works were assessed according to the national and international guidelines and approved by ethic committee CEBEA (Ethic Committee of Animal Well-Being).
Cardiac Echography Experiments-Cardiac echography experiments were assessed as previously described (5). Briefly, transthoracic echography was performed on anesthetized mice (1% isoflurane), using a Vevo 2100 system equipped with a 40 MHz transducer (Visualsonics, Toronto, Canada). Cardiac parameters were measured in B mode long axis, M mode, or Doppler mode.
Immunohistochemistry Experiments-5-Day-old wild type (WT) and P2Y 6 KO mice were injected with 1 mM EdU overnight to evaluate proliferation of cardiomyocytes in neonate mice. Then neonates were decapitated and hearts were harvested, rinsed, and frozen in Tissue-Tek OCT TM (Sakura Finetek, Alphen aan den Rijn, The Netherlands). Hearts were sliced to 5 m thickness with cryostat Leica 3050S. Then EdU was revealed with the Click-iT EdU Imaging Kit (Invitrogen, Life Technologies) and cardiomyocytes were stained with troponin T antibody (Sigma) and nuclei with Hoechst (Thermo Scientific, Waltham, MA). Images were taken randomly with a Zeiss Axioimager Z1 and analyzed with ImageJ software. Data were obtained by counting 5 images per heart.
For histological analysis, hearts from adult mice were harvested and frozen in Tissue-Tek OCT TM , cut at 10 m thickness and stained with hematoxylin-eosin or Masson's trichrome for collagen staining. Cross-sectional areas and fibrosis quantification were evaluated with ImageJ software.
Flow Cytometry Analysis-Hearts were harvested from 5-day-old WT and P2Y 6 KO mice and total ventricular cells were extracted. These cells were stained with Ki67 (BD Biosciences) and troponin T (antibodies-online, Atlanta, GA) antibodies. Number of double-stained cells was analyzed with BD LSRFortessa cell analyzer (BD Biosciences).
Isoproterenol Injection in Vivo-WT and P2Y 6 KO mice aged between 8 and 12 weeks were intraperitoneally injected once with 50 mg/kg of ISO (Sigma) or saline solution 2 min before echographic measurements (acute injection), or with 50 mg/kg/day of ISO or saline solution, daily during 7 days. Then, mice were sacrificed, weighted, tibias were measured, and hearts were harvested, rinsed, and weighted. Ventricles were then stored in TRIzol reagent (Life Technologies) for RNA extraction or in Tissue-Tek OCT TM for histological examination.
RNA Sequencing Analysis-Indexed cDNA libraries were obtained using the TruSeq Stranded mRNA sample preparation kit (Illumina) following the manufacturer's recommendation. The multiplexed libraries (9.5 pM) were loaded on a high throughput HS flow cell. Sequences were produced using a TruSeq PE Cluster kit version 4 and SBS-kit version 4 (250 cycles) on a HiSeq 1500 (Illumina). 29.2 and 26.1 million pairedend reads (2 ϫ 125 bases) corresponding to P2Y 6 KO mice and control mice, respectively, were imported into the web-based Galaxy platform. Read quality reports of the sequenced libraries were obtained using the FastQC software. Reads were then trimmed to 90 bp to obtain sequences with a per base quality value Ͼ30 and remove the Illumina universal adapter sequences. Reads were mapped against the mouse reference genome (NCBI build 37.2/mm9) using TopHat2 to generate read alignmensts for each sample. The mean coverage obtained was ϫ45.2 and ϫ44.9 for the library corresponding, respectively, to P2Y 6 KO and control mice. Annotations were obtained from iGenomes (Illumina). Cuffdiff were then used to calculate the level of differential gene expression. RNAseq was assessed on a mixture of RNA from three WT hearts and three P2Y 6 KO hearts harvested after 7 days of saline or ISO injection. Genes with counts per million (cpm) Ն1 in KO RNA and with cpm KO/cpm WT ratio Ն2 were selected and analyzed with Enrichnet software, which sorted genes regulated in KO mice in function of KEGG pathways in its database. Regulated genes involved in interesting pathways were then analyzed with STRING software to obtain interaction networks. RNA sequencing data have been deposited and assigned GEO accession number GSE76215.
Quantitative PCR Analysis-RNA was extracted from ventricles in TRIzol with an RNeasy extraction kit (Qiagen, Venlo, The Netherlands), then reverse transcription was assessed using random hexamers and Superscript II Reverse Transcriptase (Invitrogen, Life Technologies). Gene expression was then analyzed by quantitative PCR on 5 ng of cDNA and with specific primers and reactions were run on 7500 Fast Real Time PCR system (Applied Biosystems, Life Technologies). C T genes were normalized to housekeeping gene (Rpl32) C T and results were expressed on 2 Ϫ⌬CT .
cAMP Assays-Hearts were harvested, boiled, and crushed in 2 ml of water. After centrifugation, the supernatants were evaporated with a concentrator centrifuge SpeedVac TM RC1010 (Thermo Scientific). Pellets were then diluted in water, cAMP was acetylated with KOH and anhydride acetic solutions, and cAMP was then measured by radioimmunoassay. cAMP was then normalized with protein content dosed with a Pierce BCA Protein Assay Kit, following the manufacturer's recommendations (Thermo Scientific).
Culture of Neonate Cardiomyocytes-1-Day-old neonates were decapitated and ventricles were harvested and rinsed in PBS. Total cells were obtained after digestion in 2 mg/ml of collagenase B, 1 mg/ml of pancreatin (Sigma), and 50 g/ml of DNase I (Sigma). Cardiomyocytes were selected by Percoll gradient (GE Healthcare, Little Chalfont, United Kingdom) and plated in wells coated with 0.1% gelatin in medium containing half DMEM, half M199/F-12, 10% horse serum, 5% fetal bovine serum, 2% penicillin/streptomycin, and 1% L-glutamine (Gibco, Life Technologies).
To induce hypertrophy in vitro, cells were serum deprived during 24 h and treated with 1 M ISO and/or 100 M UDP (Sigma) during 48 h in serum-deprived medium. After treatment, cells were fixed with 4% paraformaldehyde (Sigma) and stained with ACTA1 antibody (Proteintech group, Chicago, IL). Secondary antibody was from Jackson ImmunoResearch (Suffolk, United Kingdom). ϩ/ϩ ) and P2Y 6 KO (P2Y 6 Ϫ/Ϫ ) male mice aged between 8 and 12 weeks with hematoxylin-eosin (n ϭ 5 mice, scale bar ϭ 1 mm). Middle and right panels, respectively, heart weight and normalization of heart weight to body weight of WT and P2Y 6 KO male mice (n ϭ 20 mice). B, quantification of expression of cardiac hypertrophy markers Acta1, Serca2a, Nppa, and Nppb by quantitative PCR in WT and P2Y 6 KO hearts (n ϭ 6 mice). C and D, echocardiographic analysis of WT and P2Y 6 KO mice cardiac function. Left ventricle mass, EDV, and EF were calculated in B and M modes (panel C: n WT ϭ 16, n KO ϭ 19). A wave and E/A ratio were obtained by Doppler analysis for P2Y 6 KO and WT mice (panel D, n ϭ 9). Results are mean Ϯ S.E., analyzed by t tests, ***, p Ͻ 0.001; **, p Ͻ 0.01; *, p Ͻ 0.05.
To analyze cardiomyocyte proliferation in vitro, cells were serum deprived overnight and treated with 100 M UDP during 48 h in serum containing medium. After treatment, cells were fixed with 4% paraformaldehyde and stained with Alexa F488conjugated WGA (Molecular Probes, Life Technologies) and eF570-conjugated Ki67 (eBiosciences, San Diego, CA) antibodies. Immunofluorescent images (20 images per well) were taken randomly with Zeiss Axioimager Z1 and analyzed with Axiovision software.
Statistics-Results are expressed as mean Ϯ S.E. Comparison between groups were realized using one-way analysis of variance and/or t tests (Mann-Whitney test for non-parametric data), or two-way analysis of variance test.
We then performed echocardiography experiments using adult WT and P2Y 6 KO mice to analyze cardiac function (Fig. 1, C and D, and Table 1). These experiments revealed a higher left ventricle mass in P2Y 6 KO mice than in WT mice (98.8 Ϯ 4.0 versus 78.5 Ϯ 4.6 mg, respectively, mean Ϯ S.E., n WT ϭ 16, n KO ϭ 19 mice, ***, p Ͻ 0.001, Fig. 1C, left panel). Left ventricle hypertrophy in P2Y 6 KO mice was also significant after normalization to body weight: 0.39 Ϯ 0.02 versus 0.32 Ϯ 0.02%, for P2Y 6 KO and WT mice, respectively (*, p Ͻ 0.05, data not shown). Echocardiographic measurements of cardiac parameters in B and M modes like end-diastolic or end-systolic volumes (EDV, ESV), ejection fraction (EF) or fractional shortening (FS), the thickness of ventricular walls (LVPW), and septum (IVS) were realized for P2Y 6 KO and WT mice (Fig. 1C, Table 1). No significant difference for these major cardiac parameters was observed (Fig. 1C, Table  1). Tissue Doppler measurements were also performed to investigate a potential diastolic dysfunction in P2Y 6 KO mice. We observed a significant increase of the A wave value in P2Y 6 KO compared with WT hearts (490.21 Ϯ 18.40 versus 415.11 Ϯ 20.16 mm/s, respectively, n ϭ 9 mice; **, p Ͻ 0.01, Fig. 1D, left panel). E/A ratio was lowered in P2Y 6 KO hearts but not significantly (Fig. 1D, right panel). We also observed no difference in IVRT and IVCT parameters between WT and P2Y 6 KO mice ( Table 1).

P2Y 6 Loss Amplifies Pathological Cardiac Hypertrophy
Induced by Isoproterenol-We investigated the potential role of P2Y 6 receptor in pathological cardiac hypertrophy induced by isoproterenol injection. We injected 50 mg of ISO/kg/day during 7 days and analyzed cardiac hypertrophy in WT and P2Y 6 KO mice. Interestingly, Fig. 3A shows that ISO-induced cardiac hypertrophy was greater in P2Y 6 KO mice than in WT mice (Fig. 3A, Table 2 We then analyzed fibrosis deposition in P2Y 6 KO mice (Fig.  3B). Only a restricted fibrosis was observed in ISO-treated hearts with no significant difference: 1.61 Ϯ 0.32 versus 1.94 Ϯ 0.55% in WT ISO versus KO ISO mice, respectively; mean Ϯ S.E., n ϭ 4 mice, Fig. 3B). Then we investigated the expression of cardiac hypertrophy markers Nppa, Nppb, Serca2a, Acta1, Myh6, and Myh7 in P2Y 6 KO and WT mice. Expressions of Nppa and Acta1 were higher in P2Y 6 KO than WT hearts after isoproterenol injection (Fig. 4A), which is characteristic of pathological hypertrophy, whereas the expressions of Nppb, Myh6, and Myh7 were not significantly modified between CTL and ISO-injected mice. The expression of Serca2a, which is known to be decreased in some cardiac pathologies (16 -18), was lower after ISO injection, but in a comparable way in WT and P2Y 6 KO mice.
␤-Adrenoreceptor expression was evaluated in control and ISO-treated P2Y 6 KO and WT hearts (Fig. 4B). Comparable expression was observed for ␤1and ␤2-adrenoreceptor expression (Fig. 4B) but ␤3-adrenoreceptor expression was decreased in the untreated P2Y 6 KO heart (Fig. 4B). We also quantified the cardiac cAMP level in P2Y 6 KO and WT mice after acute ISO treatment (50 mg/kg during 10 min, Fig. 4C). We observed that acute ISO induced a more significant cAMP increase in WT than in P2Y 6 KO hearts (Fig. 4C).
To understand the effects of ISO on cardiac function, we assessed echocardiography experiments on acute ISO-treated mice (Table 3). We can observe that ISO injection induced modifications of cardiac contractility and heart rate, but with no difference between WT and P2Y 6 KO mice (Table 3).
UDP Inhibits in Vitro Isoproterenol-induced Cardiomyocyte Hypertrophy-To investigate if the positive effect of P2Y 6 receptor loss on ISO-induced cardiac hypertrophy was linked to a negative effect of the ligand UDP on cardiomyocyte hypertrophy, we isolated and treated 1-day-old neonate P2Y 6 WT and KO cardiomyocytes with 1 M ISO and/or 100 M UDP. After 2 days of cardiomyocyte stimulation, we observed that UDP alone had no effect on cardiomyocyte size, whereas ISO was able to induce hypertrophy of cardiomyocytes (47.8 Ϯ 20.5%, mean Ϯ S.E. of 4 independent experiments; *, p Ͻ 0.05, Fig. 5). Interestingly, UDP was able to significantly inhibit the positive effect of isoproterenol on cardiomyocyte size (Fig. 5). As previously observed in Fig. 2C, the increase in basal cardiomyocyte size in P2Y 6 KO cultures was confirmed (Fig. 5). ISO and UDP effects on P2Y 6 KO myocyte size were not significant (Fig. 5).
Identification of Genes Regulated in Control and Isoproterenol-treated P2Y 6 KO Hearts-RNAseq experiments were assessed on hearts isolated from WT and P2Y 6 KO mice 7 days after saline or ISO injection (supplemental Tables S1 and S2). Comparison of gene expression between P2Y 6 KO and WT basal mice showed that only about 120 genes were differentially regulated. After analysis with Enrichnet (KEGG pathway) and   ϩ/ϩ ) mice injected with 50 mg/kg/day of ISO during 7 days (n ϭ 10 mice). B, ␤-adrenergic receptors expression was quantified by quantitative PCR in mice injected with 50 mg/kg/day of ISO during 7 days (n ϭ 10 mice). C, quantification of cardiac cAMP level 10 min after ISO injection (50 mg/kg) in P2Y 6 KO and WT mice (n ϭ 7 mice). Results are mean Ϯ S.E., analyzed by two-way analysis of variance. ***, p Ͻ 0.001; **, p Ͻ 0.01; *, p Ͻ 0.05.

TABLE 3 Echocardiographic data of WT and P2Y 6 KO mice after acute ISO injection
Mice were injected intraperitoneally with ISO (50 mg/kg) or saline (NaCl 0.9%) and echography was realized 2 min after ISO injection in B and M modes. Results are mean Ϯ S.E.

Loss of P2Y 6 Receptor Induces Cardiac Hypertrophy
STRING softwares, supplemental Table S1 revealed genes involved in MAPK, peroxisome proliferator-activated receptor, and Toll-like receptor signaling pathways, metabolic pathways, and cell cycle. Then, we compared the regulation of genes between ISO-injected P2Y 6 KO versus ISO-injected WT mice (supplemental Table S2 and Fig. 6). Among the 1130 regulated genes in ISO-treated P2Y 6 KO compared with WT mice, analysis with Enrichnet (KEGG pathway) and STRING softwares revealed genes involved in processes like hypertrophic or dilated cardiomyopathies (Acta1, Des, Itga11, and Itgb5), regulation of actin cytoskeleton (Vav1, Rac2, Pak 1 and 3), or extracellular matrix-receptor interactions (Fig. 6, supplemental Table S2). We also identified P2Y 6 target genes in RNAseq experiments that are involved in signaling pathways such as the JAK/STAT or MAPK pathways (supplemental Table S2). Additionally, quantitative PCR experiments have been performed to confirm up-regulation of some target genes involved inter alia in hypertrophic and dilated cardiomyopathy ( Table 4). The shading in Tables 4 and 5 indicate the gene regulations corresponding to a ratio superior or equal to 2 (Ն2). Table 5 is focused on genes significantly regulated by ISO in P2Y 6 KO hearts. We identified inter alia genes with an important KO ISO/WT ISO ratio involved in cardiovascular function (Lox, Ctgf, and Fn1) or JAK/STAT pathway (Lif, Il7r, Il21r, and Il11).

Discussion
During the first weeks after birth, cardiomyocyte proliferation accompanied by massive growth of capillaries and coronary vessels leads to an important development of the heart (19,20). Cardiomyocyte hypertrophy plays a major role in further growth of the heart with a determinant role of ␣ 1A/C and ␣ 1B receptors, indicating the role of sympathetic innervation (21). The observation of increased heart weight in P2Y 6 KO mice was very intriguing. The expression of P2Y 6 in cardiomyocytes was previously demonstrated (22). We observed increased cardiomyocyte size in 1-day-old P2Y 6 KO neonates compared with WT neonates. Additionally in vivo proliferation of cardiomyocytes was significantly increased in the hearts of 5-day-old P2Y 6 -null mice compared with wild type mice. Increased heart weight at adult age in P2Y 6 -null mice appears as a direct consequence of increased proliferation and size of neonate cardiomyocytes. P2Y 6 receptor appeared thus as a negative regulator of both early hypertrophy and hyperplasia of cardiac myocytes.
We have observed that Nppb expression is increased in P2Y 6 KO adult mice, compared with WT. BNP is a circulating hormone playing an important role in vascular tone regulation and compensation in heart failure (23). Dysregulation of heart

TABLE 4 Confirmation of gene regulations in ISO-treated P2Y 6 KO vs ISO-treated WT mice by RNAseq and quantitative PCR
Mice were injected daily with ISO (50 mg/kg daily) during 7 days. Then, hearts were harvested and RNA was extracted to perform RNAseq experiments using a RNA pool from 3 hearts for each condition. Some genes up-regulated in ISO-injected P2Y 6 KO mice compared to ISO-injected WT mice (P2Y 6 KO ISO/WT ISO) were confirmed by quantitative qPCR analysis using 10 mice per condition. Regulated genes were analyzed with Enrichnet (KEGG pathway) and STRING softwares and have been classified in different pathways as represented by a black dot (•). JULY 22, 2016 • VOLUME 291 • NUMBER 30 development was reported for mice deficient for many target genes and mainly related to loss of their expression in cardiomyocytes and cardiomyocyte dysfunction. Neuropilin-1 or creatine kinase knock-out mice display hearts with an enhanced size due to cardiomyocyte hypertrophy (24,25). The cardiac phenotype of P2Y 6 -null mice could also be compared with that of p27 KIP1 knock-out mice (26) and transgenic mice expressing c-myc constitutively in cardiomyocytes (27), which is characterized by increased heart weight with enhanced proliferation of cardiomyocytes. Until now, the study of UDP effects in the heart was very restricted and P2Y 6 receptor only reported to be involved in overload-induced cardiac fibrosis (7). The use of P2Y 6 antagonist MRS2578 had no effect on transverse aortic constriction-induced cardiac hypertrophy (7). Even if our model of ISO injection in P2Y 6 -null mice also implies adrenoreceptor activation, it differs from transverse aortic constriction-induced pathological cardiac hypertrophy involving endogenous catecholamines with the use of a P2Y 6 antagonist having possible secondary effects. We demonstrated previously that mice deficient for another P2Y receptor, the P2Y 4 receptor, display smaller hearts, a microcardia phenotype related to a cardiac angiogenic defect without any change in cardiac function (3) but a reduced exercise capacity (4). P2Y 6 KO and WT mice display comparable values for major cardiac parameters in B and M modes. Interestingly Doppler measurements revealed a significant increase of A wave value in P2Y 6 KO hearts. E/A ratio was lowered in P2Y 6 KO hearts but not significantly. Tissue Doppler imaging suggests a potential diastolic dysfunction related to cardiac hypertrophy in P2Y 6 KO mice.

Loss of P2Y 6 Receptor Induces Cardiac Hypertrophy
Cardiac and urine levels of norepinephrine and dopamine were decreased in P2Y 4 -null mice (4), whereas isoproterenol action on cardiac hypertrophy was enhanced in P2Y 6 -null mice. The potentiation or inhibition of adrenergic responses could be a determinant in the action of extracellular nucleotides on cardiac development and hypertrophy.
The present study supports P2Y 6 involvement in both physiological and pathological conditions. Besides the macrocardia phenotype of P2Y 6 -null mice, we observed a more pronounced cardiac-induced hypertrophy in these mice. Effectively, isoproterenol injection induced a more robust increase in heart weight and expression of hypertrophy markers such as Nppa and Acta1 in P2Y 6 -null mice than in wild type mice. These in vivo data were correlated with inhibitory effects of UDP in vitro on hyperplasia and isoproterenol-induced hypertrophy of cardiomyocytes. Echocardiography experiments showed comparable cardiac parameters after acute ISO treatment of P2Y 6 KO and WT mice.
Interestingly, ␤3-adrenoceptors expression was lower in P2Y 6 KO compared with WT basal hearts, whereas P2Y 6 KO and WT mice displayed comparable ␤3-adrenoreceptor cardiac expression 7 days after ISO injection. Overexpression TABLE 5 Genes more strongly upregulated after ISO injection in P2Y 6

KO than in WT mice
Mice were injected daily with ISO (50 mg/kg daily) during 7 days. Then, hearts were harvested and RNA was extracted to perform RNAseq experiments using a RNA pool from 3 hearts for each condition. Ratios were reported for ISO-injected P2Y 6 KO mice compared to ISO-injected WT mice (KO ISO/WT ISO), P2Y 6 KO compared to WT saline-injected mice (KO control (CTL)/WT CTL), ISO-injected compared to saline-injected P2Y 6 KO mice (KO ISO/KO CTL), ISO-injected compared to saline-injected WT mice (WT ISO/WT CTL). Regulated genes were analyzed with Enrichnet (KEGG pathway) and STRING softwares and have been classified in different pathways as represented by a black dot (•).
of ␤3-adrenoceptors was reported to inhibit hypertrophic response to ISO in vitro and in vivo through endothelial nitricoxide synthase stimulation (28). Third generation ␤-adrenoreceptor blockers such as Nebivolol have further properties including stimulation of NOS and/or ␤3-adrenoceptors (29).
Signaling pathways involved in the development of cardiac hypertrophy are diverse and complex. Generally, physiological cardiac hypertrophy is mediated by the IGF-1/PI3K/AKT pathway, effects mediated by a receptor-tyrosine kinase (7,11). The causes of pathological cardiac hypertrophy are also complex and multiple: mechanical stress or extracellular stimuli like catecholamines, endothelin-1, and angiotensin-II that activate G-protein coupled receptors. ISO-induced cardiac hypertrophy has been shown to be mediated first by G s proteins then in a PKA-dependent manner by G i proteins, leading to Src and Ras activation (12). Development of myocardial hypertrophy starts with neurohumoral factors, such as angiotensin II and norepinephrine, which activate receptors coupled to members of the G q , G 12 , and G i families (30 -32) and hypertrophic gene expression in cardiomyocytes through the Ca 2ϩ -dependent pathway (33). Most of P2Y receptors including P2Y 6 receptor are coupled to G␣ q . The major role of the G q family in the pathogenesis of hypertrophy was demonstrated by the use of transgenic overexpression or conditional knock-out mice (34,35).
Our data support that the P2Y 6 receptor acts as a negative regulator of hyperplasia during postnatal heart development as well as an inhibitor of isoproterenol-induced hypertrophy in adult mice. Interestingly RNAseq analysis revealed different sets of genes regulated by loss of P2Y 6 in basal and ISO-treated hearts (supplemental Table S1 and S2), suggesting differential involvement of this receptor in physiological and pathological cardiac hypertrophy. RNAseq analysis of the hearts of isoproterenol-treated mice showed inter alia the up-regulation in P2Y 6 -null mice of genes directly involved in hypertrophic and dilated cardiomyopathies (Nppa, Acta1, Igf-1, Tgf-␤2, Tgf-␤3, and integrins Itga11 and Itgb5), regulation of actin cytoskeleton (Vav1, Rac2, and integrins Itgb2 and Itgam) and also up-regulation of genes implicated in the JAK/STAT pathway (Il-10ra, Il-7r, Il-9r, Socs1, and Socs3), which is involved in the development of pathological hypertrophy. We confirmed the regulation of many genes in ISO-treated control mice that were previously described as ISO target genes in the heart (36). Interestingly the ISO-mediated regulation of some of these genes such as lectin-like oxidized LDL receptor-1 (Lox-1) and periostin was strongly amplified in the P2Y 6 -null mice ( Table  5). Lox-1 abrogation is known to reduce cardiac hypertrophy by reducing oxidative stress (37), whereas periostin is described as a factor in cardiac remodeling after clinical unloading of the failing heart (38).
UTP release was observed in human heart during myocardial infarction (6) and mechanical stretch releases nucleotides from cardiomyocytes through pannexin hemichannels (7). Uracil nucleotides can protect cardiomyocytes from hypoxic stress (39) as well as reduce infarct size and improve heart function after myocardial infarct (2), reflecting the multiple effects of UTP on P2Y 2 receptors expressed on cardiomyocytes and endothelial P2Y 2 and P2Y 4 receptors. We recently demonstrated P2Y 4 involvement in cardiac fibrosis and protection against cardiac ischemia (5). It appears thus that P2Y 2 , P2Y 4 , and P2Y 6 receptors could play various roles at different levels in cardiac function, development, and remodeling.
The present study constitutes an important advance in the field of extracellular nucleotide receptors and cardiac development. UDP appears as an inhibitor of physiological and pathological cardiac hypertrophy acting through P2Y 6 receptor activation. P2Y 6 agonists could reduce cardiac hypertrophy and associated damages and constitute a therapeutic target as a regulator of cardiac remodeling.