Functional cardiac orexin receptors: role of orexin-B/orexin 2 receptor in myocardial protection

Cardiovascular disease (CVD), including myocardial infarction, is the leading cause of morbidity and mortality in the Western world [1,2]. Orexins, also referred to as hypocretins, exist in two functional forms, i.e. orexin A (OR-A) and orexin B (OR-B), which are derived from a common 130 amino acid precursor peptide [3,4] and exert their actions by activating the orexin-1 (OX1R) and orexin-2 (OX2R) receptors. OXRs belong to the G protein-coupled receptor (GPCR) superfamily [3] and are distributed throughout both central and peripheral sites where they regulate endocrine, metabolic and cardiovascular functions [5–7]. Of note, several established cardiovascular drugs exert their cardiac actions via GPCRs, such as the muscarinic, angiotensin, adrenergic, and endothelin receptors [8].

Growing evidence indicates that central orexin neurons are implicated in cardiovascular regulation [5]. For example, in the central nervous system (CNS), both orexin and OX2R-containing nerve fibres have been identified in the paraventricular nucleus (PVN) [9], an area which constitutes a vital central site for integration of sympathetic outflow and cardiovascular function [10–17]. Thus, although orexins can influence the autonomic control of the cardiovascular system, little is known about any direct effects of orexins in the heart.

We and others have provided evidence for a wide distribution of orexins and their receptors peripherally [18–20]. We have also demonstrated the presence of OXRs in human adipose tissue, where they modulate adipogenesis and adipose tissue metabolism [7]. Herein, we hypothesised that orexins may also exert direct cardiac effects and we sought to investigate the expression and function of cardiac OXRs using in vitro, ex vivo and in vivo models, corroborating this with clinical data.

Study treatment agents were: OR-A, OR-B (Phoenix Peptides), OX2R agonist ([Ala11, d-Leu15]-OR-B; Tocris), OX1R antagonist (SB-408124; Tocris), wortmannin (PI3Kinase (PI3K) inhibitor; Sigma–-Aldrich) and mitogen-activated protein kinase (MAPK) inhibitor U0126 (Tocris). According to the manufacturer’s specifications on biological activity, the used OX2R agonist ([Ala11, d-Leu15]-OR-B; Tocris) is a ‘highly potent and selective OX2 receptor agonist; displays 400-fold selectivity over OX1 receptors. EC₅₀ values are 0.13 and 52 nM for human OX2 and OX1 receptors respectively’. Similarly, the used OX1R antagonist (SB-408124) is a ‘selective non-peptide orexin OX1 receptor antagonist (Kb values are 21.7 and 1405 nM for human OX1 and OX2 receptors respectively)’. Triphenyl-tetrazolium chloride (TTC, Sigma–Aldrich) was used to determine the infarct size following ischaemia and reperfusion. All other reagents used were of the highest purity commercially available.

All animal experiments were performed in accordance with U.K. legal requirements and the guidelines from Directive 2010/63/EU of the European Parliament on the protection of animals used for scientific purposes. This investigation also conforms to the guidelines for the welfare of animals in experimental neoplasia as they have been developed by the United Kingdom Coordinating Committee of Cancer Research (UKCCCR). Adult male Wistar rats (250–300 g) were housed in environmentally controlled conditions (22 ± 2°C, humidity 40–60%) under a 12:12-h light–dark schedule (lights on 06:00). After a week of habituation to these conditions, rats were killed using cervical dislocation. Hearts were immediately removed for the following studies: (i) dissection and immediate snap-freezing for later tissue analysis; (ii) isolation of ventricular cardiomyocytes by enzymatic digestion; (iii) immersion in ice-cold oxygenated Tyrode’s solution and then perfusion via the aortic cannula in the modified Langendorff model with Tyrode’s solution at 37°C, as previously described [21]. In brief, contractile parameters were measured by insertion of a fluid-filled latex balloon through the left atrium (LA) into the left ventricle (LV), this was connected to a pressure transducer (MTL0380, ADInstruments Ltd, U.K.) and the balloon volume was adjusted to give end-diastolic pressure of <10 mmHg. Aortic perfusion pressure was monitored with a second pressure transducer in series with the aortic cannula. Data were continuously recorded using a Power Lab 8 preamplifier/digitiser (ADInstruments Ltd, U.K.). Hypothalamic dissection took place as previously described [22]. All procedures were approved by the ethics committee and complied with the University’s care and welfare guidelines, in strict accordance with the Home Office Guidance (PPL 70/7175) on Research and Testing using animals and the Animal (Scientific Procedures) Act, 1986 and the Guide for the Care and Use of Laboratory Animals published by the U.S. National Institutes of Health (NIH Publication No. 85-23, revised 1996).

Commercially available human cardiovascular multiple tissue cDNA (MTC) panel (Clontech) were used for the present study. All tissues were normal (i.e. non-diseased) and the amount of pooled patients varied from 11 up to 32. The sets of primers used for reverse transcription-PCR (RT-PCR) amplification were: (i) OX1R sense 5′-ccttcctggctgaagtgaag-3′ and antisense 5′-agtgggagaaggtgaagcag-3′ and (ii) OX2R sense 5′-gtcgcaactggtcatctgct-3′ and antisense 5′-cgtcctcatgtggtggttct-3′.

Under the microscope, single rat cardiomyocytes (n=6) were harvested via the patch clamp micro-electrode pipette, without losing the gigaseal to prevent contamination with extracellular fluid [23]. Subsequently, the pipette contents were expelled into a sterile 0.5 ml tube.

Total RNA from different rat heart chambers (n=3) was extracted and cDNA synthesis was performed. PCRs were carried out as previously described [7].

Adult rat cardiomyocytes, seeded on glass cover-slips were fixed with 4% paraformaldehyde in PBS. The primary rabbit antiserum OR-A and OR-B (Phoenix Europe GmbH, Karlsruhe, Germany) and the primary goat polyclonal OX1R (Santa Cruz Biotechnology Inc., U.S.A.) and a monoclonal OX2R antibody (Abcam, U.K.) were used at a 1:500 dilution. All dilutions were made in 3% BSA in PBS-0.01% Triton X-100. Specimens were incubated with primary antibody for 60 min, then washed three times with PBS (5 min each time) before incubation with anti-rabbit IgG-FITC and Texas Red anti-goat secondary antibody (Vector Laboratories, Peterborough, U.K.) for 45 min. The slides were then thoroughly rinsed with PBS, and the cell nuclei were visualised by applying the DNA-specific dye DAPI at a final concentration of 1.5 µg/ml.

Whole heart tissue sections from adult rats (n=3) were cut at 3 μm. Endogenous peroxidase was blocked by incubating sections in a 1.5% H₂O₂ solution for 20 min. The detection system used was an immunoperoxidase-based system (Vector Universal Elite ABC kit) with a diaminobenzidine tetrahydrachloride visualisation agent. Sections were incubated with primary antibody (OX1R, OX2R, OR-A and OR-B) for 60 min at room temperature. Sections were then washed for 2 × 5 min in buffer. Secondary biotinylated antibody was applied for 30 min at room temperature. The avidin–biotin complex solution was then applied to the sections and incubated for 30 min at room temperature (Vector Elite ABC reagent). DAB solution was prepared according to the manufacturer’s instructions and applied to the sections for 5 min. Sections were then dehydrated, cleared and mounted.

For controls, rat spleen, testes (positive controls) and thymus (negative control) were stained using paraffin-embedded tissues from US Biomax (#RA162, MD, U.S.A.). Furthermore, we have repeated the OR-A staining on rat heart array (#TR031; US Biomax MD, U.S.A.).

Protein lysates from rat brain, rat heart and human myometrium were separated using SDS/PAGE (15% polyacrylamide) under reducing conditions and then transferred on to immunoblot nitrocellulose membrane. Blocking of non-specific binding was achieved by placing the membrane in a blocking buffer (1% BSA in 1× TBST) at 4°C overnight. The following day, membrane was rinsed with TBST for 5–10 min. Membranes were then incubated with primary anti-rabbit OR-A and OR-B antibodies (Phoenix Peptides), diluted 1:500 and further incubated at 4°C overnight under gentle agitation. The next day, after rinsing the membrane three times in TBST for 10 min to remove unbound primary antibodies, the membranes were exposed to anti-rabbit IgG–HRP conjugate (Sigma–Aldrich) (diluted 1:2000 in blocking solution) for 1 h at room temperature. Detection of immunocomplexes took place using the ECL method, exposing X-ray film (Fuji) at different time points (30 s to 1 min) following by developing. Molecular weight approximations are taken by comparing the stained bands to that of the marker proteins.

Following the isolation of calcium tolerant cardiomyocytes (n=6), the cells were treated in a concentration (0.01–100 nM) and time (0–60 min) dependent manner with both OR-A and OR-B. Following these optimisation experiments, orexin-induced phosphorylation of various signalling cascades, i.e. Akt, extracellular signal-regulated kinase (ERK) 1/2 (ERK1/2), Troponin-I (TnI), and Myosin Light Chain were determined by Western blot analyses, as described previously [24].

Contraction and intracellular calcium [Ca²⁺]i were determined in isolated ventricular cardiomyocytes (n=6), as previously described [21]. Cells were classified as either contractile or non-contractile in response to electrical stimulation. To measure [Ca²⁺]i, myocytes were loaded with fura-2 and excited alternately at 340 and 380 nm with a monochromator and emitted light collected at >520 nm and measured using a photomultiplier tube (Photon Technology International). Contraction strength is shown as percentage cell shortening and [Ca²⁺]_i as Fura2 ratio.

All experiments lasted for 195 min. Hearts were allowed to stabilise for 30 min with the standard Tyrode’s solution perfusion. Contractile parameters from the data files were calculated by averaging a minimum of six cycles per data point to obtain the mean value for left ventricular systolic pressure (LVSP mmHg), and left ventricular end diastolic pressure (LVEDP mmHg) (measured at t0, t30, t40, t75, t85, t100, t130, t160 and t195 min). The difference between these two pressure values represented the left ventricular developed pressure (LVDP, mmHg). Heart rate (HR, beats.min⁻¹) was determined at the same time points together with the rate pressure product (RPP, mmHg), which was calculated by multiplying the LVDP and HR (average value of LVDP is obtained from averaging both systolic and diastolic values, then the mean is multiplied by HR to obtain RPP).

Global ischaemia was produced for 30 min by stopping flow of physiological solution to the heart. At the end of reperfusion, the heart was frozen at −20°C before being sliced into 2 mm thick transverse sections and stained with 2% TTC, which reacts with intracellular dehydrogenases present in viable cells producing a red pigment and staining viable tissue red (whereas infarcted tissue remains pale) and then fixed in 10% formalin overnight to enhance the contrast. Each section was then photographed with a digital camera and then traced on to a clear acetate sheet to determine infarct size (% infarct area = infarct area on each slice/total area of the heart slice × 100) [25]; the area was calculated using computer-assisted planimetry [NIH image 1.57] [26–28].

For the inhibitor studies, the rats were randomly assigned to one of the following groups: (i) control group, where they were subjected to 30 min of global ischaemia followed by 120 min of reperfusion, and the treated heart groups, which received inhibitor (wortmannin; 1 μmol/l or U0126; 10 μmol/l) together with OR-B perfusion which was added to Tyrode’s buffer after stabilisation and perfused for 10 min prior to 30-min global ischaemia and switched back to normal Tyrode’s during 120 min reperfusion. In addition, some hearts only received the inhibitors (wortmannin or U0126) to rule out any influence they may have directly upon infarct size.

Rats were anaesthetised with 3% isoflurane inhalation and then intubated via a tracheostomy and ventilated with a mixture of O₂ and 1.5–3% isoflurane, using a rodent ventilator. A left thoracotomy was performed through the fourth intercostal space, pericardium opened and the left coronary artery (LCA) located and ligated with a 4-0 silk suture. The ligation was confirmed successfully when the anterior wall of the LV turned pale around the area that was tied off using a snare. Following 30 min of regional ischaemia, the snare was released and blood allowed to flow freely during 120 min of reperfusion. At the end of the experimental protocol, the hearts were removed and perfused on to the Langendorff system for staining. The ligation was tied off and hearts perfused with 1% Evans Blue to demarcate the area at risk (AAR), followed by TTC staining to identify the viable tissue staining deep red. The infarct area within the risk area was then calculated (I/R %). The same surgical procedure was followed for sham group without occlusion of the LCA (data not shown).

At the end of reperfusion, the rat was overdosed with pentobarbital and the heart was rapidly removed and mounted on to the Langendorff apparatus as above. The left anterior descending artery was ligated, 1 ml of 0.25% Evans Blue (in PBS) was then injected into the heart and allowed to perfuse followed by normal Tyrode’s for a further 2 min to delineate the area not at risk (ANAR) staining it blue [29]. The heart was stored at −20°C before being sectioned into 2-mm sections and stained with TTC to determine the infarct size as stated above. The percentage infarct tissue (I) within the AAR was calculated as I/AAR %. Furthermore, the ischaemic area (AAR) was distinguished from ANAR.

The clinical study was approved by the Research Ethics Committee South-Central Oxford C (11/SC/0140). All participants provided written informed consent, in accordance with the Declaration of Helsinki.

To explore whether our findings from the cell cultures and animal model could be of relevance to CVD in humans, we investigated OX2R gene expression levels in human myocardium samples. A cohort of 54 patients undergoing coronary artery bypass grafting with available myocardial cDNA library was used for this purpose (Table 1); relative expression was calculated using the Pfaffl method, with phosphoglycerate kinase 1 (PGK1) as housekeeping gene [30].

To examine the direct effects of OR-A and OR-B on intracellular signalling in the human heart, human myocardial tissue samples were collected from the right atrial appendage of eight patients undergoing cardiac surgery. These myocardial tissue samples were washed in ice-cold Krebs HEPES buffer and then cut into thin strips containing all myocardial layers. The tissue was first equilibrated for 10 min in Krebs HEPES buffer at 37°C and then incubated for 20 min in the presence or absence of human recombinant OR-A (100 nM) or OR-B (100 nM) (Phoenix Pharmaceuticals Inc., Burlingame, CA, U.S.A.). At the end of the experiment, incubated tissue samples were collected and stored at −80°C until assayed. Western immunoblotting was used to examine the direct effects of OR-A and OR-B on p-Akt (Ser⁴⁷³), total-Akt, (Cell Signaling, Danvers, MA), and phospho/total ERK1/2 (Abcam, Cambridge, England, United Kingdom) in human myocardial tissue. Briefly, myocardial tissue samples were homogenised for 30 s using a pre-cooled electric Polytron homogenizer in 220 μl of lysis buffer (Invitrogen, U.K.) containing a protease plus phosphatase inhibitor cocktail (Roche Applied Science). Homogenates were spun at 13000 rpm for 10 min at 4°C. The protein concentration of the supernatants was then measured using the BCATM Protein Assay kit (Pierce, U.K.). Protein lysates were separated on 12% gradient SDS-NuPAGE gel (Invitrogen, U.K.), and proteins transferred to PVDF membranes (Amersham UK Ltd.), followed by blocking with 5% powdered skimmed milk. The membranes were incubated with the respective primary antibodies overnight and immunodetection of the primary antibodies was performed with horseradish-peroxidase–conjugated secondary antibodies (Promega), and enhanced chemifluorescence (Amersham Bioscience UK Ltd.), and quantified in relation to the house-keeping protein GAPDH (Santa Cruz Biotechnology, Santa Cruz, CA).

All results are expressed as mean ± S.E.M., unless indicated otherwise. For Western blotting experiments, the densities of the immunoreactive bands were measured using a scanning densitometer coupled to scanning software (ImageQuant; Molecular Dynamics, Amersham Pharmacia, U.K.). Data were evaluated as in the study by Maulik et al. [31]. Statistical analyses among the Langendorff model perfused groups were performed using one-way ANOVA followed by Bonferroni’s post hoc test with significance determined at the level of P<0.05.

Using qRT-PCR, we demonstrate that both OXRs are expressed in the rat heart. Using rat hypothalamus as a positive control, the expression of OX1R was significantly less in the heart (P<0.001), whereas OX2R expression levels were similar between the heart and the hypothalamus (Figure 1A). Both OXRs were uniformly expressed throughout the rat heart including the LV, left auricle (also known as left atrial appendage), interventricular septum, right ventricle (RV) and right auricle (Figure 1D). Single-cell PCR from cardiomyocytes revealed expression of both OXRs in single cells investigated (Figure 1E).

Using Western blotting, we also confirmed the expression of OXRs at the protein level in total rat heart and cardiomyocyte cell lysates (data not shown). Immunohistochemical analysis of ventricular sections using specific antibodies for OX1R (Figure 1B(I)) and OX2R (Figure 1B(II)) revealed a widespread cytoplasmic and membrane distribution of both OXRs across the rat ventricle. Cellular distribution of OXRs was further investigated in rat cardiomyocytes in vitro. Immunofluorescence revealed a similar distribution of OX1R (Figure 1C(I)) and OX2R (Figure 1C(II)). This is consistent with the expression of GPCRs in the cytoplasm (due to trafficking or internalisation), as well as on the cell membrane [32]. We then proceeded to study the expression of OXRs in the human heart using a cardiovascular MTC panel demonstrating expression of both OXRs in multiple human cardiac compartments (Figure 1F).

We also demonstrated that prepro-orexin is expressed as a 302-bp PCR product in the LV, LA, interventricular septum, RV and right atrium (RA) of the rat heart (Figure 1G). We have used Western blotting to demonstrate that the prepropeptide can be cleaved into full-length OR-A (Figure 1H) and OR-B (Figure 1I), with molecular weights of approximately 3.5 and 2.9 kDa, respectively. We then assessed the protein expression using immunohistochemistry. Antibodies were validated using rat thymus as a negative control and rat spleen and rat testes as positive control (Figure 1J and Supplementary Figure S2). Subsequent immunohistochemical analysis of rat ventricular sections revealed protein expression of both cleaved peptides across the ventricle (Figure 1K and Supplementary Figure S2). Similar expression was evident in individual cardiomyocytes (Figure 1L).

Superfusion of cardiomyocytes with Tyrode’s solution containing 10 nM OR-B increased the shortening of twitch contraction (Figure 2A), from 15.2 ± 1.2 to 20.6 ± 1.7% (n=6, P<0.01; Figure 2B). This was not associated with any change on either the amplitude (Figure 2A) or the rate of the Ca²⁺ transient decline (n=6, P=NS (no supplement); Figure 2C). OR-A had no effect on the rat cardiomyocyte cell length even with concentrations as high as 100 µM (data not shown; P=0.07). The OR-B effect was replicated using the specific OX2R agonist ([Ala11, d-Leu15]-OR-B) (Figure 2D), while it was not blocked by an OX1R-specific antagonist (SB-408124) (Figure 2E), suggesting that these effects are predominantly mediated via an OR-B/OX2R pathway.

Furthermore, OR-B, but not OR-A (data not shown), significantly induced phosphorylation of TnI at Ser²³/Ser²⁴ in cardiomyocytes. This effect was dose-dependent, with a maximum activation at 100 nM of OR-B (P<0.001; Figure 2F). Similarly, OR-B, but not OR-A (data not shown), induced significant phosphorylation of the myosin light chain (MLC) at Ser¹⁹ with maximal effects at 100 nM and 15 min (P<0.001; Figure 2G).

In both the isolated rat heart and the in vivo rat model, we measured LVDP, LV pressure decay and HR. The RPP was then calculated; this was used as a clear and direct indication of the energy demand of the heart and, thus, a good measure of the energy consumption of the heart pre- and post-ischaemia.

Initially, pre-treatment with either OR-A, OR-B or OX2R agonist, prior to the ischaemic insult, did not alter the haemodynamic parameters or HR between any of the groups (data not shown). Contractile parameters were recorded at various time points throughout the experimental protocol (Supplementary Figure S1). Following 30 min of global ischaemia in the isolated Langendorff perfused rat heart model, the expected decline in cardiac function during reperfusion, as observed in the control hearts, was not affected by pre-treatment with OR-A (Figure 3A). In contrast, the pre- and post-treatment groups for both OR-B and the OX2R agonist resulted in significantly improved recovery compared with control during reperfusion (P<0.05; Figure 3A,B). This was further replicated in the in vivo rat model where the control group (saline treated) demonstrated poor cardiac functional recovery and cardiac function following 30 min of regional ischaemia and 120 min reperfusion, whereas the OR-B pre-treated group [bolus injection (i.v.) of OR-B (100 nM) 10 min prior to regional ischaemia] showed significant improvement in functional recovery during reperfusion (P<0.05; Figure 3C). Concentration-dependent effects were noted with OR-B (20–100 nM), with a maximal response at 100 nM (data not shown).

In isolated rat heart model pre-treated with OR-B for 10 min prior to global ischaemia and reperfusion, the LV infarct size was reduced from 61.5 ± 3.0% in the control to 22.6 ± 1.9% in the treated group (P<0.001; Figure 3D). In addition, the OR-B post-ischaemic treatment was just as effective at providing myocardial protection as pre-treatment, since it also significantly reduced the LV infarct size to 21.8 ± 3.9% (P<0.001; Figure 3E). The OX2R-specific agonist exhibited a similar protective effect to OR-B treatment, inducing a reduction in LV infarct size both during pre-treatment (25.7 ± 3.6%) and at the start of reperfusion (34.7 ± 3.1%, both P-values <0.001 compared with control; Figure 3D,E). However, pre-treatment with OR-A alone showed no significant reduction compared with control (58.3 ± 3.0 compared with 61.5 ± 3.0%, respectively; P=NS) in the LV infarct size (Figure 3D). Moreover, in the in vivo model pre-treatment with a bolus injection (i.v.) of OR-B (and of ORX2Ra, but not of OR-A) 10 min prior to regional ischaemia and reperfusion significantly reduced the %LV infarct size compared with the control (saline) group (66.7 ± 3.0 compared with 26.8 ± 14.9% in the control and OR-B treated group, respectively; P<0.001; Figure 3F).

The isolated rat heart model was pre-treated either with an OX1R-specific antagonist or a PI3 kinase (PI3K) inhibitor (wortmannin) in the presence or absence of OR-B prior to ischaemia/reperfusion. Pre-treatment with either the OX1R-specific antagonist or wortmannin alone showed no change in functional recovery compared with the control group (Figure 4A) and no effect on the %LV infarct size (61.5 ± 3.0% in the control group compared with 61.8 ± 1.9% with OX1R antagonist alone, and 56.5 ± 1.5% with wortmannin alone; P=NS, Figure 4B). Pre-treatment with OR-B together with wortmannin showed no myocardial protection on cardiac functional recovery (P=NS; Figure 4A); and, abolished the OR-B protective effect on LV infarct size (P=NS, Figure 4B), suggesting the importance of the PI3K/Akt pathway in OR-B induced cardioprotection. Improved functional recovery with OR-B pre-treatment was not inhibited by the simultaneous administration of an OX1R-specific antagonist (P<0.05; Figure 4A). Similarly, the OR-B protective effect on %LV infarct size was not blocked by the OX1R-specific antagonist (P<0.001; Figure 4B). Both these findings suggest that the noted effect is OX2R specific.

Given that ischaemic preconditioning protects the heart by phosphorylating the pro-survival kinases Akt and ERK1/2 at reperfusion, we investigated further whether orexins can affect their phosphorylation [33]. Treatment with OR-B increased the phosphorylation of both Akt and ERK1/2 compared with basal levels (P-values <0.001; Figure 4C–E); in contrast with OR-A which had no significant effect on ERK1/2 (Figure 4E) and Akt (data not shown). The OR-B effect was concentration dependent (Figure 4D). The time course for the ERK1/2 activation was for 3, 5, 10 and 15 min. The involvement of an MAPK pathway was further corroborated using the selective MAPK inhibitor U0126, demonstrating that inhibition of the MAPK pathway using U0126 abrogates the protective effect of OR-B on LV infarct size (Figure 4F).

To further evaluate the potential implication of orexins and their receptors in cardiac disease in humans, we studied OXR downstream signalling and expression in cardiac tissue samples from adult patients undergoing cardiac surgery and coronary artery bypass grafting. In the context of these studies, all human myocardial tissue samples were collected from the right atrial appendage during cardiac surgery.

Using Western blotting, we examined the direct effects of OR-A and OR-B on the phosphorylation of Akt and ERK1/2 in heart tissue samples from eight patients. Treatment with OR-A or OR-B did not change Akt phosphorylation compared with basal levels (Figure 5A), while only OR-B significantly increased ERK1/2 phosphorylation compared with control (non-treated heart tissue; P<0.05; Figure 5B).

In addition, we measured OXR expression levels in heart tissue samples from heart failure patients undergoing coronary artery bypass grafting (n=54; Table 1). Our results demonstrated a negative correlation between the severity of heart failure symptoms and the expression of OX2R. As such, when patients were categorised using the New York Heart Association (NYHA) functional classification of heart failure (classes I–IV based on increasing severity of clinical symptoms) there was a significant decrease in OX2R in NYHA-II compared with NYHA-I and in NYHA-III/IV compared with NYHA-II and NYHA-I (P-values <0.05; Figure 5C).

In the present study, we have shown that the heart constitutes a source of orexins and expresses functional OXRs. Moreover, our experiments revealed that OR-B can exert direct cardioprotective effects in both ex vivo and in vivo rat heart models.

Indeed, using the Langendorff perfused rat heart model, pre- and post-myocardial ischaemia treatments with both OR-B and a selective OX2R agonist resulted in significantly increased RPP which is a direct indication of the energy demand of the heart and, thus, a good measure of the energy consumption of the heart pre- and post-ischaemia.

In the case of the ex vivo studies that replicate the in vivo results, additional interactions between sympathetic activity, which is increased in ischaemia [34], and OXR signalling through cAMP [35] cannot be ruled out. Additionally, treatment of ventricular cardiomyocytes with OR-B, but not OR-A, lead to significantly increased Tn-I and MLC phosphorylation, followed by a concomitant increase in the strength of their twitch contraction. Of note, these findings agree with previous studies demonstrating that changes in the phosphorylation of key cardiac regulatory proteins can affect cardiac function [36,37].

This is the first observation of a differential specificity of OR-A and OR-B in the heart. OR-A activates both OX1R and OX2R receptors with similar potencies (∼36 nM), whereas OR-B selectively activates OX2R compared with OX1R (∼420 nM) [38]. Both receptors belong to the GPCR superfamily and are highly promiscuous in terms of their G-protein coupling characteristics [39]. These tend to be cell- and tissue-specific and influence subsequent activation of downstream signalling pathways. For example, we have previously shown that expression of human OX2R is regulated by a complex involving a proximal PKA/PKC-regulated promoter and a distal promoter regulating tissue-specific expression of alternative transcripts which in turn post-transcriptionally regulate receptor levels [40]. Therefore, future studies are required to provide further insight into the regulation of the promoter of OX2R in the myocardium.

Moreover, a dichotomous role of OXRs is evident in the rat brain, where OX1Rs mediate the neurobiological effects that drive drug seeking, whereas OX2Rs play a crucial role in sleep/wake cycle regulation and arousal [41]. Interestingly, OR-A, but not OR-B, can activate electrophysiologically the subfornical organ of rat neurons [42]. In a more recent study, OR-A (100  nM) and OR-B (100  nM) increased leptin gene expression in differentiated preadipocytes by 49.9 and 71.3% respectively [43], indicative of differential modulation of transcriptional responses by these two peptides.

This phenomenon has been previously described for other GPCRs. In the case of smooth muscle cell CRH-receptors, urocortin, but not CRH, activated the MAPK signal transduction pathway despite similar binding affinities [44]. Thus, it is possible that differences on OR-A and OR-B responses could also be influenced by the cardiac microenvironment, leading to alternative coupling of G proteins and activation of distinct signalling pathways. In previous studies, we have shown that OX2Rs can differentially couple to Gq, Gi/o and Gs proteins in different tissues [19,22,39]. Emerging studies have also shown that OXRs can form functional dimers. It is possible that dimerisation can also take place at the cardiac level and therefore influence pharmacological properties, such as the affinity of the receptors for orexin ligands, as well as cellular distribution and trafficking [45]. Finally, polymorphisms of the OX2R gene have also been described previously [46]. To this date, dimerisation or polymorphisms have not been explored in the myocardium. Future studies should determine whether dimerisation or polymorphisms of cardiac OXRs may affect ligand binding, or indirectly affect ligand binding by altering receptor structure.

In the present study, using an OX1R antagonist and an OX2R agonist, we conclude that the primary signalling at the rat heart is mediated via an OR-B/OX2R pathway. OXRs are able to activate certain key plasticity-regulating cascades, such as PI3K and ERK, and p38 MAPK. Here, pre-treatment with wortmannin abolished the OR-B protective effect on LV infarct size; suggesting the involvement of the PI3K/Akt pathway in OR-B induced cardioprotection. This signalling pathway has been implicated in myocardial protection; and is referred to as the reperfusion injury signalling kinase (RISK) pathway [33]. Mediators that target the activation of the RISK (PI3K/Akt) pathway have been also shown to ameliorate recovery following ischaemic stress [25,47]. In our study, OR-B, but not OR-A, was able to induce the phosphorylation status of both Akt and ERK1/2. Our data contradict previous studies showing that both orexin peptides can increase Akt phosphorylation. For example, in a study of neuronal cell cultures OR-B was more potent than OR-A, but the maximal effect of both peptides on the Akt activation was very similar [48]. This apparent contradiction could be attributed to the fundamental differences between cardiomyocytes and neuronal cells including morphological polarity, shape and function of action potentials, synaptic and homoeostatic plasticity; not to mention the spatial expression of OXRs in the brain [49]. As such, it is plausible that orexins exert their effect in an organ- and cell-specific manner.

The OR-B selectivity in post-translational modification of key mediators is corroborated by previous studies in other tissues. Indeed, in the human testis, when IP3 turnover was measured, OR-B exerted a more potent effect compared with OR-A [18]. Similarly, in human H295R adrenocortical cells’ diverse roles have been noted for OR-A and OR-B in terms of activation of MAPKs and cortisol release; i.e. only OR-A (at 1 nM) activated ERK1/2 and p38 [50].

Based on our findings, herein we propose the presence of a signalling pathway involving OX2R-ERK1/2-MLC in the heart with a net effect of increased Ca²⁺ sensitivity and cell length/contractility following OR-B superfusion in cardiomyocytes [21,51–54]. Activation of ERK1/2 is involved in the activation of contractile responses through mechanisms involving direct phosphorylation of the Ca²⁺/calmodulin-dependent MLC kinase (MLCK) [33,55–57].

In a seminal review by Spinazzi et al. [58], a case is made for mediation of Ca²⁺ by OXRs. However, this effect of OX1R and OX2R on Ca²⁺-release is through IP3 receptor activation, which, although evident in the hypothalamus, does not impact on the contraction of rat ventricular cardiac muscle under basal conditions, where the ryanodine receptors (RyR) dominate excitation–contraction (E–C) coupling [59]. As such, this may have more of a role in pathophysiology of heart failure and arrhythmias [60]. Interestingly, the dominant expression of IP3 receptors is on the nuclear membrane where they are thought to be involved in cardiac hypertrophy [61].

Conversely, inhibition of ERK attenuates force development by lowering MLC phosphorylation in cardiac tissue [62]. Because of its central role in the regulation of contraction and relaxation of the heart, cardiac TnI (cTnI) is also phosphorylated by OR-B to ensure appropriate function. Changes in the phosphorylation status of cTnI are well documented during acute cardiac events and in patients with heart failure [63].

We have expanded on these observations using human heart tissue samples from heart failure patients. So far, a number of studies indicated that there is a dysregulation of the orexin system in the CNS in models of heart failure [64–66]. Here we further report that in adults with heart failure (NYHA classes I–IV) there was a significant negative correlation between the severity of clinical symptoms (NYHA class) and OX2R expression. The fact that OX2R expression was significantly lower in patients with NYHA classes III/IV compared with less symptomatic patients with NYHA class I/II, indicates that OX2R signalling is compromised in more severe heart failure. These observations are further supported by the findings of Perez et al. [67], where in a heart failure model, hypocretin receptor 2 (HCRTR2)/OX2R-deficient mice exhibited poorer cardiac function, and greater myocardial scarring. Moreover, in the same study, differential OX2R expression data from microarray experiments were recorded in patients with dilated cardiomyopathy and ischaemic cardiomyopathy [67].

Based on the established role that orexins have in maintaining the awake state, orexin antagonism has also been identified as a novel therapeutic approach for the treatment of insomnia [68]. As such, suvorexant, a non-selective dual orexin (OX1R/OX2R) receptor antagonist (DORA), has been approved by the FDA, in 2014, for treating insomnia patients with difficulties in sleep onset and/or sleep maintenance (http://www.acc.org/latest-in-cardiology/articles/2015/06/16/08/40/new-insomnia-drugs-in-the-context-of-cardiovascular-disease - BELSOMRA® (suvorexant) package insert: http://www.merck.com/product/usa/pi_circulars/b/belsomra/belsomra_pi.pdf). To date, suvorexant use has not been associated with any reported adverse cardiovascular safety outcomes; however, it must be highlighted that it is unknown whether suvorexant has adverse effects in CVD patients, since patients with significant and/or recent CVD (e.g. with acute coronary syndrome or congestive heart failure) were excluded from the relevant trials (http://www.acc.org/latest-in-cardiology/articles/2015/06/16/08/40/new-insomnia-drugs-in-the-context-of-cardiovascular-disease). Therefore, and based on the findings from our study and from others [67], the use of orexin antagonists may warrant further assessment and careful monitoring for CVD adverse effects, particularly in heart failure patients [69].

Finally, our human ex vivo study results were also consistent with the in vitro rat data, since orexin treatment of human myocardial tissue demonstrated that OR-B, but not OR-A, was able to induce ERK1/2 phosphorylation. However, there was no change in Akt phosphorylation upon treatment with OR-A or OR-B. The latter could be due to the fact that chronic heart failure reduces Akt phosphorylation, so signalling is already compromised at basal levels. Indeed, compared with subjects with normal cardiac function, human skeletal muscle of heart failure patients demonstrates reduced Akt phosphorylation [70]. It should be noted that the cardiovascular and musculoskeletal system are interrelated, whereby a functional deterioration in one is reflected similarly in the other [71]. Differences in the phosphorylation of Akt can also be due to species differences, since certain proteins, like ETS-related gene (ERG), are differentially expressed in the rat and human heart [72].

We also acknowledge certain limitations with our experimental system using healthy young adult rats compared with patients used in the present study that have certain comorbidities. To date, it has proven difficult to find the ‘perfect’ translational in vivo model for CVD. This is primarily due to the high level of complexity, as well as heterogeneity of these diseases, in addition to potential impact/influence from environmental or genetic factors. Future studies can make use of different animal models for examples such as; heart failure, aged rats or rats fed on a high-fat diet to generate further data on the involvement of OR-B/OX2R at the cardiac level. However, to the best of our knowledge, there is no perfect in vivo model to represent the entire repertoire of defects/comorbidities at the cardiac level.

In summary, our present findings document that OR-B exhibits a protective role on cardiac function pre- and post-ischaemia, leading to reduction in the infarct size. We also provide novel conclusive evidence that the rat heart expresses functional OXRs and is a source of orexins. Cardiac OXRs can also influence contractile tone via mechanisms involving phosphorylation of TnI and MLC. Thus, in addition to the known CNS effects of orexins/OX2R [5,73], our data illustrate the importance of the OX2R as a potential therapeutic target directly at the heart level, suggesting that targeted use of OX2R-specific agonists may have a role in the prevention and treatment of CVD.

• Several cardiovascular drugs exert their cardiac actions via GPCRs, and orexins acting via their GPCRs can affect the heart centrally. In the present study, we tested whether orexins and their receptors can exert a direct effect at cardiac level.

• The human and rat heart constitute a source of orexins and express functional OXRs. OR-B can exert direct cardioprotective effects in both ex vivo and in vivo rat heart models. In adults with heart failure there was a significant negative correlation between the severity of clinical symptoms (NYHA class) and OX2R expression in the human heart.

• Based on the findings from our study, the use of orexin antagonists warrants further assessment and careful monitoring for CVD adverse effects, particularly in heart failure patients.

The authors declare that there are no competing interests associated with the manuscript.

This work was supported by the British Heart Foundation [grant number 03/131/16192]; and the Heart Research U.K. [grant number RG2564/08/09].

H.S.R. conceived and planned the experiments. V.H.P., J.C., E.K., C.A., A.A. and B.K.T. carried out the experiments. V.H.P., E.K., I.K., H.S.M., G.K.D., G.R., C.A., E.W.H., A.N. and H.S.R. contributed to the interpretation of the results. H.S.R. and E.K. took the lead in writing the manuscript. All authors provided critical feedback and helped shape the research, analysis and manuscript.

AAR

area at risk

ANAR

area not at risk

[Ca²⁺]i

intracellular calcium

CNS

central nervous system

cTnI

cardiac troponin I

CVD

cardiovascular disease

ERK

extracellular signal-regulated kinase

GPCR

G protein-coupled receptor

HR

heart rate

LA

left atrium

LCA

left coronary artery

LV

left ventricle

LVDP

left ventricular developed pressure

MAPK

mitogen-activated protein kinase

MTC

multiple tissue cDNA

MLC

myosin light chain

NS

no supplement

NYHA

New York Heart Association

OR-A

orexin A

OR-B

orexin B

OX1R

orexin receptor-1

OX2R

orexin receptor-2

OXR

orexin receptor

PI3K

PI3 kinase

RISK

reperfusion injury signalling kinase

RPP

rate pressure product

RT-PCR

reverse transcription-PCR

RV

right ventricle

Tn

troponin

TTC

triphenyl-tetrazolium chloride