Introduction

In 1970, it was proposed that adenosine 5′-triphosphate (ATP) is an important signaling molecule, which could act like conventional neurotransmitters to initiate potent biological effects (Burnstock et al. 1970; reviewed Burnstock 2007). Implicit to this idea was the presence of purinoreceptors, which led Burnstock to propose that receptors for adenosine and adenine nucleotides be classified into two types: P1 purinoreceptors, which are most sensitive to adenosine and are blocked by methylxanthines, and P2 purinoreceptors, which are most sensitive to ATP (Burnstock 2008).

There is a growing interest in the effects of activation of P2 receptors by ATP within the cardiovascular system. It is now well established that ATP is produced by a variety of sources within the heart, in both normal physiological and pathophysiological states. Indeed, it is a cotransmitter alongside adrenaline in sympathetic nerve terminals in the amphibian heart (Hoyle and Burnstock 1986). In pathophysiological states such as hypoxia and ischemia, cardiac myocytes release ATP into the extracellular milieu (Forrester and Williams 1977; Wee et al. 2007). Molecular cloning studies have revealed two subfamilies of P2 receptors: P2X receptors, which are ligand-gated non-specific cation channels, and P2Y receptors, which are heptahelical receptors coupled to G proteins. Seven P2X receptors (P2X1–7) and eight P2Y receptors (P2Y1,2,4,6,11,12,13,14—the missing P2Y receptors are non-mammalian orthologs) have so far been cloned (Vassort 2001; Köles et al. 2008). Members of both the P2X and P2Y family are expressed in the heart (Vassort 2001). Activation of P2X receptors leads to the opening of a non-specific cation channel permeable to Na+, Ca2+, and K+ (Vassort 2001). At negative potentials, these channels carry an inward current, which reverses around 0 mV (Vassort 2001). Members of the P2Y family are coupled to G proteins. They can be further subdivided into two groups: the first group (i.e., P2Y1, P2Y2, P2Y4, P2Y6, and P2Y11) couples to Gq proteins to stimulate phospholipase C followed by increases in intracellular inositol phosphates and the mobilization of intracellular Ca2+ stores; in addition, the activation of P2Y11 induces an increase in adenylate cyclase activity. The second group (i.e., P2Y12, P2Y13 and P2Y14) couples to Gi proteins and leads to the inhibition of adenylate cyclase followed by a decrease in intracellular cAMP levels (von Kügelgen 2006; Köles et al. 2008). Interference with intracellular Ca2+ is expected to have effects throughout the heart. For example, in the ventricles, an increase in systolic Ca2+ will have a positive inotropic effect (Danziger et al. 1988), and recently, the role played by intracellular Ca2+ in pacemaking has been the subject of intense scrutiny (Ju et al. 2003; Lyashkov et al. 2007).

To date, ATP has been demonstrated to both increase and decrease the L-type Ca2+ current in ventricular cells (Alvarez et al. 1990; Qu et al. 1993), modulate K+ currents in atrial cells (Matsuura and Ehara 1997), activate an ATP-activated cationic current in rabbit sinoatrial nodal cells (Shoda et al. 1997), and modulate intracellular Ca2+ in amphibian nodal cells (Ju et al. 2003). ATP activation of P2 receptors also plays a role in inotropy. Activation of P2X receptors increases intracellular Ca2+ transients and contractility in single ventricular myocytes from murine and rat hearts (Vassort 2001; Shen et al. 2006). Elucidating the roles played by members of the P2X and P2Y family has been complicated by the lack of subtype specific agonists and antagonists. Furthermore, an additional complication exists with the discovery that P2X receptors can coassemble to form heteromeric complexes with distinct pharmacological and operational profiles when compared to their homomeric subunits (Torres et al. 1999; Brown et al. 2002; Köles et al. 2008). Similarly, P2Y receptor subtypes can form homo- or hetero-oligomers with each other or with adenosine receptors (Köles et al. 2008). A detailed description of the distribution of P2 receptors in discrete regions of the heart should facilitate a further understanding of P2 receptor heart function. Therefore, the aim of this study was to examine the expression profile of the P2 receptors in the sinoatrial node and cardiac muscle of rat and human heart.

This study demonstrates a widespread expression pattern of P2 receptors in different regions of the rat and human heart. Important differences were observed in the distribution of P2 receptors between rat and human hearts.

Materials and methods

Preparation of tissues

Wistar rats (male; 250–300 g) were humanely killed according to the United Kingdom Animals (Scientific Procedures) Act, 1986. Intact sinoatrial node (SAN) was isolated from male rat hearts (~300 g) for RNA isolation (n = 7) and in situ hybridization (ISH; n = 3) as previously described (Dobrzynski et al. 2005; Tellez et al. 2006). Central SAN samples, for RNA isolation, were taken from the intercaval region approximately at the level of the leading pacemaker site in the SAN (adjacent to the main branch from the crista terminalis). Samples of ventricular and atrial muscle were taken from the left ventricle (LV; free wall) and right atrium (RA). Samples were snap frozen (for ISH) in OCT before being immersed in isopentane cooled by liquid N2 (−162°C) and stored at −80°C. The samples were immersed in drop freezing medium (OCT, BDH) and frozen in liquid N2. Tissue sections (12 µm) were subsequently cut on a cryostat and were stored for up to several months at −80°C. The rat heart failure model used has been described previously (Maczewski and Maczewska 2006). Briefly, heart failure was induced by generating an extensive myocardial infarction caused by ligation of the proximal left coronary artery. Rats were humanely killed 3 months postmyocardial infarction. Heart failure (HF, n = 8) and sham-operated (SO, n = 8) rats were used in this study for quantitative PCR (qPCR). Separate samples of heart from these rats were also used for another study where hemodynamic and remodeling data are described in detail (Tellez et al., manuscript in preparation). Briefly, in that study, rats showed characteristic signs of heart failure (corrected lung weight (g/kg), HF 9.48 ± 0.83, SO 5.81 ± 0.23; corrected heart weight (g/kg), HF 5.16 ± 0.34, SO 3.33 ± 0.14; n = 10). In addition, the mean infarct size was 49.3 ± 2.5%; n = 10. The infarct size was estimated using echocardiography, as described previously (Maczewski and Maczewska 2006). Furthermore, HF rats showed an increase in left ventricular diastolic pressure (317%) and decreases in left ventricular systolic pressure (19%), left ventricular developed pressure (32%), maximum rate of rise of left ventricular pressure (37%), maximum rate of fall of left ventricular pressure (41%), and left ventricular ejection fraction (60%, all from ten observations, Tellez et al., manuscript in preparation).

Human SAN preparations (n = 4) were obtained from The Prince Charles Hospital, Chermside, Australia (Human Research Ethics committee approval EC2565). Samples were from brain dead patients, whose hearts were not used for organ donation, but were available for research (Table 1). SAN preparations were snap frozen in isopentane precooled in liquid nitrogen. Tissue sections (30 µm) were cut from the preparations and stained using hematoxylin and eosin in order to identify the SAN; SAN and right atrium were then micro-dissected for RNA isolation.

Table 1 Summary of patient profile
Full size table

RNA isolation and cDNA production

Total RNA, from rat samples (ventricular muscle, atrial muscle, and SAN), was isolated as previously described (Dobrzynski et al. 2005; Tellez et al. 2006). RNA integrity was confirmed on a formaldehyde-agarose gel and quantified using a ND-1000 nanodrop (Labtech, UK). Total RNA (150 ng) was reverse transcribed with Superscript II reverse transcriptase (Invitrogen, UK) and random hexamer primers, according to the manufacturer's protocol. cDNA was diluted tenfold in water for subsequent qPCR analysis. Total RNA was isolated from human samples in a similar manner; 530 ng of total human RNA was reverse-transcribed.

Quantitative PCR and primers

qPCR was carried out as previously described (Tellez et al. 2006). Briefly, the relative abundance of selected cDNA fragments was determined with qPCR on an ABI Prism 7700 (Applied Biosystems) and detection with a SYBR green probe. Assay primers were obtained from Qiagen (Qiagen, UK). All runs were 40 cycles in duration. For all transcripts and samples, at least three separate measurements were made with 1 μl aliquots of each cDNA sample. The abundance of a cDNA fragment is expressed relative to the abundance of 28S (housekeeper). It was assumed that the relative abundance of a cDNA fragment reflects the relative abundance of the corresponding mRNA.

Assessment of purity of samples

The samples used in this study were used in two other studies in which the purity of the samples was assessed and reported. In study 1, Chandler et al. (2009) investigated the expression of over 100 transcripts in the human SAN, whereas in study 2, Tellez et al. (manuscript in preparation) investigated the expression of over 100 transcripts in the rat SAN. In both studies, the purity of the samples was estimated using HCN4 (the major isoform of the pacemaker channel). The data from these two studies show that in both rat and human, there is more expression of HCN4 in the SAN compared to the surrounding atrial muscle (data not shown). The data from these two studies therefore confirm that there is a minimal contamination in the different samples. In addition, our data also show that there is more expression of ANP mRNA in the human and rat RA as compared to the SAN (Chandler et al. 2009; Tellez et al. manuscript in preparation).

In situ hybridization

ISH was carried out as previously described (Dobrzynski et al. 2006; Tellez et al. 2006). Briefly, we used digoxigenin (DIG)-labeled cRNA probes. cDNA fragments were isolated from rat heart or brain cDNA by PCR and linked to SP6 or T7 promoters by subcloning into pGEM-T Easy (Promega, UK). cRNA probes were synthesized by in vitro transcription from SP6 or T7 promoters using the Mega Script kit (Ambion, UK) at a ratio of ~4 unlabeled UTPs (Ambion) to one DIG-UTP (Roche, UK).

Statistics

All results are given as means ± SEM. Differences were evaluated by one-way ANOVA to compare receptor mRNA levels for rat left ventricle, right atrium, and SAN and Student's t test to compare receptor mRNA levels for human right atrium and SAN with SigmaStat software. Differences were considered significant at the level of P < 0.05.

Results

Determination of P2X receptor mRNA in rat ventricular and atrial muscle and the sinoatrial node by quantitative PCR

cDNA for all P2X receptors, except P2X6, was successfully amplified in rat left ventricle, right atrium, and SAN (Fig. 1). The most abundant mRNA was P2X5 > P2X7 > P2X1–4 in all heart regions (Fig. 1). Further analysis showed heterogeneity between heart regions for some receptor subtypes (Fig. 1), and statistically significant differences are shown in Table 2. Of note, P2X1, P2X2, P2X3, and P2X7 mRNAs were significantly more abundant in SAN compared with left ventricle (for P2X1, P2X2, P2X3) and right atrium (for P2X4, P2X7). P2X4 mRNA had a different pattern of expression with left ventricle > SAN > right atrium. In rats with coronary artery ligation induced heart failure, P2X4 mRNA was upregulated in the SAN (average upregulation of 93%; Fig. 1 inset). A trend towards upregulation was also shown in rat left ventricle (43% increase, P > 0.05) and right atrium (38% increase, P > 0.05).

Fig. 1
figure 1

Relative abundance, as measured by qPCR, of P2X transcripts in the rat left ventricle, right atrium, and SAN. Means ± SEM (n = 7) shown

Full size image
Table 2 Summary of significant differences in expression levels between tissue types
Full size table

Determination of P2Y receptor mRNA in rat ventricle, atrium, and the sinoatrial node by quantitative PCR

cDNA for all P2Y receptors studied was successfully amplified in all areas of the rat heart (Fig. 2). P2Y11 has not been cloned from the rat and therefore was not tested. In left ventricle, mRNA for P2Y1, P2Y2, and P2Y14 were the most abundant transcripts (Fig. 2). In the right atrium, P2Y2 mRNA was the most abundant (Fig. 2), and in the SAN, P2Y2, P2Y6, P2Y12, and P2Y14 all showed an equal level of mRNA expression (Fig. 2). Table 2 provides a summary of differences between regions.

Fig. 2
figure 2

Relative abundance, as measured by qPCR, of P2Y transcripts in the rat left ventricle, right atrium, and SAN. Means ± SEM (n = 7) shown

Full size image

Abundance of P2 receptor mRNAs compared to HCN4 and the A1 receptor

We sought to gain a perspective of the relative levels of the most abundant SAN purine receptor mRNAs (P2X4, P2X5, P2X7, P2Y14) by making a comparison with HCN4 (the primary α-subunit for the pacemaker current If) and the A1 adenosine receptor (Fig. 3a). P2X5 had a similar mRNA abundance to that seen for HCN4, which was several fold higher than the next three most abundant P2 transcripts and adenosine A1 receptor mRNA.

Fig. 3
figure 3

a Comparative abundance of HCN4, P2X4, P2X5, P2X7, P2Y14, and adenosine A1 receptor mRNAs in the rat SAN. Means ± SEM (n = 7) shown. b Comparative abundance of HCN4, P2X4, P2X7, P2Y1, P2Y2, P2Y14, and adenosine A1 receptor mRNAs in the human SAN. Means ± SEM (n = 4) shown

Full size image

In situ hybridization on rat ventricular muscle, atrial muscle, and sinoatrial node

Since the heart is a heterogeneous organ, comprising not only myocytes but also vascular, neuronal, connective, and adipose tissue, in situ hybridization was used to determine the localization of purine receptor mRNA in left ventricle, right atrium, and SAN. The most abundant purine receptor mRNAs, P2X4, P2X5, P2X7, and P2Y14, were investigated. For all P2 receptors studied with in situ hybridization, the labeling of mRNA appeared as dark spots at low power and intracellular rings at high power (Fig. 4), which presumably correspond to the rough endoplasmic reticulum (in which mRNA is translated) in the perinuclear region. Figure 4 shows high levels of mRNA in all cardiac regions not for only P2X5 but also P2X4 and P2Y14 (Fig. 4a–f, j–l). In the ventricles, expression of P2X4 and P2Y14 mRNA (Fig. 4a and j) appears to be confined to the cardiac myocytes. However, P2X5 mRNA labeling was not restricted to ventricular myocytes (Fig. 4d) but also occurred to the lining of the extensive vascular network within ventricular tissue, most likely representing P2X5 expression to the endothelial lining of the ventricular vasculature (data not shown). This is not surprising because P2X5 mRNA has been previously reported to be expressed in bovine endothelial aortic cells (Ramirez and Kunze 2002). As with the ventricle, labeling in the atrial muscle of the crista terminalis was clearly localized to the perinuclear region in the atrial myocytes (Fig. 4b, e, and k). Of most interest was the strong presence of mRNA for P2X4, P2X5, and P2Y14 in SAN myocytes, in the familiar perinuclear pattern (Fig. 4c, f, and l). We are confident that these cells are nodal for several reasons. The images were taken from adjacent tissue sections to sections that had been labeled for Cx43 and HCN4 as negative (Musa et al. 2002) and positive (Tellez et al. 2006) markers of the SAN, respectively. Furthermore, the small size of the cells (~40 µm) suggests that these cells are nodal. Within the nodal tissue sections, labeling of P2X4, P2X5, and P2Y14 appeared to be confined to the nodal myocytes. The data for P2X7 are different to that seen for P2X4 and P2X5. Interestingly, in situ hybridization failed to detect P2X7 mRNA in the ventricles (Fig. 4g). This is in contrast to the qPCR data, which suggested that P2X7 is the second most abundant P2X transcript within the ventricles. Labeling within the atrium and SAN was present for the P2X7 transcript, but was less abundant in comparison to that seen for P2X4, P2X5, and P2Y14 (Fig. 4h and i). The data suggest that P2X7 mRNA is expressed within cells other than cardiac myocytes in the ventricles.

Fig. 4
figure 4

Distribution of P2X4, P2X5, P2X7, and P2Y14 mRNA in the left ventricle (a, d, g, and j), the right atrium (b, e, h, and k), and the SAN (c, f, i, and l), as determined by in situ hybridization. Labeling of mRNA appears as black perinuclear rings. Scale bar, 100 µm

Full size image

Quantitative PCR of purine receptor mRNA to human atrial muscle and sinoatrial node

Quantitative PCR for purine receptor mRNA was determined in human right atrium and SAN. Unlike in rat, not all P2 receptors could be amplified from human SAN and right atrial muscle. In atrial muscle (Fig. 5), specific primers for P2X1, P2X2, P2X3, and P2Y11 failed to amplify any transcript and therefore were assumed to be absent. Of the transcripts that were successfully amplified, the predominant P2X receptor mRNAs present in atrial muscle were P2X4 and P2X7. This is in contrast to the data obtained for the rat (Fig. 1). The most abundant P2Y receptor mRNAs were P2Y1, P2Y2, and P2Y14 (Fig. 1). In human SAN, unlike rat SAN, P2X4 tended to be the most abundant P2X isoform (Fig. 5). P2X2, P2X3, and P2Y11 primers failed to detect mRNA within human SAN samples and therefore was assumed to be absent. SAN had greater levels of P2X1 and P2Y14 mRNA than right atrium (Table 2). In order to determine the relative abundance of the P2 receptor mRNAs, a comparison was made between the most abundant P2 receptor transcripts with mRNA for HCN4 and the A1 receptor (Fig. 3b). Unlike the rat, the P2 receptor transcripts showed a lower abundance than HCN4 or A1 receptor mRNAs.

Fig. 5
figure 5

Relative abundance, as measured by qPCR, of P2X and P2Y transcripts in the human right atrium and SAN Means ± SEM (n = 4) shown

Full size image

Discussion

This study demonstrates a widespread expression pattern of P2 receptors at mRNA level in different regions of rat heart and a somewhat more restricted distribution in human heart. It also demonstrates, for the first time, the P2 receptor expression pattern in the human and rat SAN. Information about receptor distribution was limited to mRNA because, unfortunately, it was not possible to investigate the expression of P2 receptors at the protein level since currently available antibodies proved unreliable for both immunohistochemistry and western blotting.

Atrium and ventricle

We observed P2X1–5 and 7, P2Y1,2,4,6,12–14 mRNA in rat atrium and ventricle, P2X4,7, P2Y1,2,4,6,12–14 in human right atrium, which could predict cardiac effects. Functional correlates for all these have not been determined yet. Nevertheless, several studies have investigated the activation of P2 receptors by ATP and the role they play in inotropy. At the forefront of these are a series of studies carried out by Liang and coworkers (Shen et al. 2006; He et al. 2001; Mei and Liang 2001; Shen et al. 2007a, b; Yang et al. 2004). Utilizing a transgenic mouse over-expressing P2X4, they demonstrated that activation of P2X receptors and specifically P2X4 is capable of enhancing the contractile state of myocytes and the intact heart. This increase in contractility is associated with an increase in sarcoplasmic reticulum Ca2+ loading and is independent of any change in cAMP or the activity of the L-type Ca2+ channel in mouse ventricular cells (Hu et al. 2001; Mei and Liang 2001; Shen et al. 2007b; Yang et al. 2004). In a coronary artery ligation model of rat heart failure, overexpression of P2X4 receptors caused enhanced cardiac contractile performance at 7 days and 1 and 2 months postinfarction compared to wild-type controls (Sonin et al. 2008). Furthermore, transgenic P2X4 mice showed a survival benefit at all time points (Sonin et al. 2008). In the present study, in situ hybridization demonstrated that P2X4 mRNA was highly expressed in the rat heart and predicted to be localized over myocytes (Fig. 4). qPCR data demonstrated that the most abundant of the P2X receptor mRNAs was P2X5 mRNA in the rat (Fig. 1). Therefore, P2X5 may play a significant role within the heart. In human atrial muscle, P2X4 mRNA was the most abundant of the P2X transcripts. In electrically stimulated isolated human right atrial trabeculae from patients with coronary artery disease with symptomatic heart failure (NYHA II–III) undergoing coronary artery bypass, ATP caused a transient reduction in contractile force, followed by a sustained increase (Gergs et al. 2008). The receptor/mechanism for the transient cardiodepressant effect is not known. The positive inotropic component was mimicked by 2-methyl-thio-ATP and was resistant to blockade by the P2 receptor antagonists, suramin, PPADS, or RB2 and was tentatively suggested to be mediated through activation of the P2X4 receptor (Gergs et al. 2008). It is interesting that P2X4 mRNA showed a trend toward an increase in both right atrium (38% increase) and left ventricle (43% increase) in our coronary artery ligation model of heart failure in the rat. It is possible that the P2X4 receptor was upregulated in patients with heart failure (NHYA II–III; Gergs et al. 2008), which may have facilitated the appearance of P2X4-receptor like positive inotropic effects (Gergs et al. 2008). Myocardial ischemia causes an increase in levels of ATP, ADP, and UTP in mouse heart using a langendorf model (Wee et al. 2007). UTP was shown to attenuate myocardial dysfunction following ischemia–reperfusion possibly through activation of P2Y2 receptors (Wee et al. 2007), one of the most abundant P2Y receptors.

Studies on patients suffering from dilated cardiomyopathy (DCM) have demonstrated an increase in the expression of P2X1 (Berry et al. 1999) in left atria. The authors postulated that the increase in P2X1 expression may provide a route for “cytotoxic Ca2+” and plays a role in the increase in cell death seen in patients with DCM. The mechanism for increased P2X1 receptors in left atrium from patients with dilated cardiomyopathy has not been determined. qPCR failed to detect the presence of P2Y11 in the human right atrium or SAN, which is in agreement with the human heart Northern blot data of Communi et al. (1997). However, our data contrast with Hou et al. (1999), who demonstrated strong bands for P2Y11 in all four chambers of the heart from patients with DCM. Combining our data with the data of Hou et al. (1999) suggests either a compensatory upregulation of P2Y11 or a pathophysiological role for these receptors in the development of heart failure. The work of Liang and coworkers (Shen et al. 2006; He et al. 2001; Mei and Liang 2001; Shen et al. 2007a, b; Yang et al. 2004) suggests that activation of the P2 receptors maybe a potential route to augment myocardial contractility when the heart is in a compromised state.

Sinoatrial node

This is the first report of the expression of P2 receptors in the rat and human SAN. ATP-dependent modulation of heart rate could be the net sum of effects mediated by multiple receptor subytpes. Ju et al. (2003) reported that activation of P2Y1 receptors modulates intracellular Ca2+, which in turn modulates the firing rate in cane toad pacemaker cells. Application of 2-MeSADP induced a biphasic effect on [Ca2+]i, which was associated with a transient tachycardia followed by a slowing of heart rate. A similar result was seen by Takikawa et al. (1990), who reported in the rabbit heart that the application of >300 µM ATP induced a sinus tachycardia. They postulated that this response was due to the activation of P2 receptors linked to phospholipase C and production of inositol phosphate and diacylglycerol. Interestingly, in the present study, in the rat SAN, P2X5 was the most abundant of the P2 receptors at the mRNA level (Fig. 1). Its expression level was several fold more abundant than A1 receptor mRNA (Fig. 3a), suggesting that it has the potential to play a significant role in modulating pacemaking. Activation of P2X5 by ATP leads to an inward current at negative membrane potentials with a reversal potential of −4 mV (Garcia-Guzman et al. 1996) and an outward current at positive potentials (Garcia-Guzman et al. 1996). Therefore, activation may lead to a tachycardia with activation of members of the P2Y family potentially leading to a bradycardia. A balance between activation of P2X and P2Y receptors may lead to the biphasic response seen by Ju et al. (2003). In the human SAN, P2X4 was the most abundant of the P2X receptors at the mRNA level (Fig. 5). Although the abundance of P2 receptor mRNAs was lower compared with HCN4 and A1 receptor mRNAs (Fig. 3b), this does not rule out the possibility that they play a role in pacemaking. A pathophysiological state or states of myocardial stress may trigger the activation of these receptors. Consistent with this, an average upregulation of P2X4 by 93% was seen in the SAN in the rat heart failure tissue. It would be interesting to investigate whether there is an upregulation of any of the P2 receptors in the SAN from patients with heart failure.