Reperfusion of ischaemic rat or mouse hearts causes NE [noradrenaline (‘norepinephrine’)] release, stimulation of α1-ARs (α1-adrenergic receptors), PLC (phospholipase C) activation, Ins(1,4,5)P3 generation and the development of arrhythmias. In the present study, we examined the effect of increased α1A-AR drive on these responses. In hearts from non-transgenic mice (α1A-WT), Ins(1,4,5)P3 generation was observed after 2 min of reperfusion following 30 min of zero-flow ischaemia. No Ins(1,4,5)P3 response was observed in hearts from transgenic mice with 66-fold overexpression of α1A-AR (α1A-TG). This was despite the fact that α1A-TG hearts had 8–10-fold higher PLC responses to NE than α1A-WT under normoxic conditions. The immediate phospholipid precursor of Ins(1,4,5)P3, PtdIns(4,5)P2, responded to ischaemia and reperfusion similarly in α1A-WT and α1A-TG mice. Thus the lack of Ins(1,4,5)P3 generation in α1A-TG mice is not caused by limited availability of PtdIns(4,5)P2. Overall, α1-AR-mediated PLC activity was markedly enhanced in α1A-WT mice under reperfusion conditions, but responses in α1A-TG mice were not significantly different in normoxia and post-ischaemic reperfusion. Ischaemic preconditioning prevented Ins(1,4,5)P3 generation after 30 min of ischaemic insult in α1A-WT mice. However, the precursor lipid PtdIns(4,5)P2 was also reduced by preconditioning, whereas heightened α1A-AR activity did not influence PtdIns(4,5)P2 responses in reperfusion. Thus preconditioning and α1A-AR overexpression have different effects on early signalling responses, even though both prevented Ins(1,4,5)P3 generation. These studies demonstrate a selective inhibitory action of heightened α1A-AR activity on immediate post-receptor signalling responses in early post-ischaemic reperfusion.
A brief period of ischaemia and reperfusion in rat or mouse hearts causes NE [noradrenaline (‘norepinephrine’)] release from the cardiac sympathetic nerves , activation of α1-ARs (α1-adrenergic receptors), generation of substantial amounts of Ins(1,4,5)P3 and the initiation of arrhythmias [2–4]. In the heart, Ins(1,4,5)P3 is not produced in large amounts in response to α1-AR activation under normoxic conditions [5,6]. Generation of Ins(1,4,5)P3 after periods of ischaemia involves an overall increase in PLC (phospholipase C) activity, as well as increased cleavage of PtdIns(4,5)P2 (Figure 1). The changes in Ins(1,4,5)P3 production and arrhythmogenesis both require the release of NE and activation of α1-ARs, but it is not currently known which subtype of the α1-AR is involved in the response. Hearts of most species express two subtypes of α1-AR, α1A and α1B, at the protein level [7,8]. Expression of α1A-ARs predominates in human heart, but the α1B subtype is more plentiful in rat and mouse myocardium [9,10]. Both of these subtypes couple via Gq to PLCβ isoforms and, thus, would be expected to initiate similar responses in the myocardium . However, studies using genetically modified mouse lines have shown that the two receptor subtypes respond differently to pathological stimuli. Hearts that overexpress α1B-AR are predisposed to hypertrophy with no contractile phenotype [11,12], whereas α1A-AR-overexpressing hearts have no hypertrophy but have markedly heightened contractility . It is possible, therefore, that ischaemia and post-ischaemic reperfusion have different effects on α1A- and α1B-AR-mediated responses.
Generation of Ins(1,4,5)
P3 and its metabolites
In previous studies, we have shown that expression of a constitutively active α1B-AR in the mouse heart prevented PLC activation during ischaemia/reperfusion and, thus, no Ins(1,4,5)P3 was generated, and the hearts were protected against arrhythmias . This might represent a preconditioning action of the overexpressed α1B-AR. On the other hand, hearts from α1B-AR-overexpressing mice had reduced expression of the α1A-AR subtype. Therefore it is also possible that the protection afforded in α1B-AR-transgenic mice was indirect and related to the suppressed α1A-AR expression. This hypothesis proposes that the Ins(1,4,5)P3 responses in early reperfusion are mediated exclusively by α1A-ARs. If this were the case, hearts overexpressing α1A-ARs should have heightened Ins(1,4,5)P3 generation in early post-ischaemic reperfusion, compared with WT (wild-type) controls, and the present study was undertaken to address this question. Previous studies from our laboratory  and others  have suggested that the α1A-AR subtype mediated pathological growth responses in the heart. Stimulation of α1A-AR was shown to be sufficient to cause cardiomyocyte hypertrophy. Furthermore, hypertrophy in vivo, as well as in cell models, was associated with an increased expression of α1A-AR mRNA [14,15]. However, recent studies have suggested, on the contrary, that α1A-ARs are cardioprotective [16,17]. In the present study, we investigated the possible role of the α1A-AR subtype in mediating the Ins(1,4,5)P3 response in early reperfusion, or alternatively its role in protecting hearts from this response. We found that heightened activity of α1A-ARs increased PLC responses in normoxia, but, despite this, completely suppressed Ins(1,4,5)P3 responses in reperfusion.
MATERIALS AND METHODS
[3H]Inositol was purchased from Auspep, and scintillation fluid (Flo-Scint IV) was from PerkinElmer. All other chemicals were AR grade, and solutions were prepared in highly purified Milli-Q water.
All procedures were approved by the Alfred Medical Research and Education Precinct Animal Ethics Committee and followed the Guidelines of the Australian Code of Practice for the Care and Use of Animals for Scientific Purposes. Transgenic (TG) mice (A1A2 line) (α1A-TG) with cardiac-restricted overexpression (66-fold) of the α1A-AR under the control of the α-myosin heavy chain promoter were generated on a FVB/N genetic background at the Victor Chang Cardiac Research Institute, Sydney, Australia . WT (α1A-WT) littermates were used as controls. Both male and female mice were used, and the animals ranged in age from 3–6 months. No significant differences were observed in body or heart weights between α1A-TG and α1A-WT mice bred in our animal facility.
3H-Labelled inositol phosphates in isolated perfused mouse hearts
Adult mice were injected with heparin (5 international units/g of body weight) 30 min before killing by cervical dislocation. Hearts were removed and transferred immediately into ice-cold saline containing 30 mmol/l 2,3-butanedione monoxime. Hearts were retrogradely perfused using the Langendorf method with Hepes-buffered Krebs medium [20 mmol/l Hepes buffer (pH 7.4), 11 mmol/l glucose, 138 mmol/l Na+, 4.5 mmol/l K+, 1.2 mmol/l Mg2+, 25 mmol/l HCO3−, 1.2 mmol/l PO4 and 2 mmol/l Ca2+] (37 °C) equilibrated with 5% CO2/95% O2 at a rate of 2 ml/min, as described in detail previously . After a 15-min equilibration period, hearts were labelled with [3H]inositol (40 μCi/ml) for 2 h. Labelled medium was removed and replaced with medium containing propranolol (1 μmol/l) and LiCl (10 mmol/l) for 10 min to block β-ARs and to inhibit inositol phosphate metabolism respectively. Hearts were then subjected to the different protocols, as described below. At the end of each experiment, inositol phosphate accumulation was terminated by freezing the hearts in liquid nitrogen, following excision at the atrioventricular junction. The frozen ventricles were weighed and inositol phosphates were extracted and quantified as described below.
The protocols followed are shown in Figure 2. For experiments under normoxic conditions, NE (100 μmol/l) was added to the perfusate for 2 min. For experiments investigating ischaemia and reperfusion, [3H]inositol-labelled hearts were subjected to ischaemia by ceasing perfusion and oxygen flow for 30 min. Reperfusion was achieved by restarting perfusate flow and oxygen for 2 min. Ischaemic preconditioning was achieved by subjecting [3H]inositol-labelled hearts to three cycles of 5 min of ischaemia and 5 min of reperfusion, followed by 30 min ischaemia without or with 2 min of reperfusion, as described previously . Timed controls were also prepared.
Protocols used for the perfused mouse heart experiments
Extraction and quantification of 3H-labelled inositol phosphates
Inositol phosphates were extracted from frozen hearts in 2 ml of 5% (v/v) TCA (trichloroacetic acid) solution containing 2.5 mmol/l EDTA and 5 mmol/l phytic acid using a Polytron homogenizer followed by sonication, as described previously . Samples were centrifuged to remove TCA-insoluble material, the supernatant was collected and the pellet was re-extracted in 0.5 ml of 5% (v/v) TCA. Supernatants were pooled and extracted with 1,1,2-trichloroethane/tri-n-octylamine (1:1, v/v) (Sigma). The final aqueous phase was treated at 50 °C with proteinase K (50 μg/ml). Treated samples were then passed through a Dowex-50 column (4% cross-linked; mesh, 4–400) and eluted with 1 ml of water. 3H-Labelled inositol phosphates were separated using anion-exchange HPLC and were quantified using an on-line β-counter (Packard Instruments Model CR), as described previously .
Extraction and quantification of 3H-labelled inositol phospholipids
TCA pellets remaining after inositol phosphate extraction were dissolved with 1 ml of chloroform/methanol/HCl (200:100:1, by vol.) using sonication and vigorous vortex-mixing. EDTA (0.5 ml; 5 mmol/l) was added, and the phases were separated by centrifugation at low speed. The upper phase of each sample was discarded, and the final organic phase was evaporated under vacuum. The dried lipids were deacylated with methylamine/methanol/butanol (42:47:9, by vol.) for 45 min at 50 °C, followed by evaporation under vacuum. The samples were dissolved in 1 ml of water and extracted with butanol/light petroleum/ethyl formate (20:40:1, by vol.). The aqueous phase of each sample (1 ml) was applied to 1 ml Dowex-1 columns (formate form). Columns were washed with 20 ml of water. Glycerophosphoinositol (deacylated PtdIns) was eluted with 20 ml of 180 mmol/l ammonium formate and 5 mmol/l sodium tetraborate. After washing the columns with 20 ml of the above solution, glyceroPtdIns4P (deacylated PtdIns4P) was eluted with 20 ml of 400 mmol/l ammonium formate and 100 mmol/l formic acid. GlyceroPtdIns(4,5)P2 [deacylated PtdIns(4,5)P2] was eluted with 7 ml of 1 mol/l ammonium formate and 100 mmol/l formic acid. Sample radioactivity was counted using a β-spectrometer.
Measurement of NE
Samples of perfusate were collected before and after 2 min of reperfusion and were extracted with alumina. NE in these samples was separated by HPLC and then quantified by electrochemical detection, as described previously .
Values are means±S.E.M. Statistical analysis of 3H-labelled inositol phosphates and 3H-labelled inositol phospholipids [PtdIns4P and PtdIns(4,5)P2] were carried out using a two-way ANOVA, followed by Tukey's test. Significance was determined at P<0.05.
α1A-TG mice have heightened PLC responses under physiological conditions
To evaluate the activity of the overexpressed α1A-AR, PLC responses to maximally effective concentrations of NE were assessed in [3H]inositol-labelled hearts from α1A-WT and α1A-TG mice. In the absence of NE, there were no differences in the content of any of the 3H-labelled inositol phosphates in hearts from α1A-WT and α1A-TG mice. Maximally effective concentrations of NE caused only a small 3H-labelled inositol phosphate response in α1A-WT mice, and the only 3H-labelled inositol phosphate to show an increase was [3H]Ins4P, the ultimate metabolite of Ins(1,4,5)P3 (Figures 1 and 3). In marked contrast, in hearts from α1A-TG mice, large increases were detected in [3H]Ins(1,4)P2 as well as in [3H]Ins4P. There was no increase in [3H]Ins(1,4,5)P3 in either strain (Figure 3). Increases in 3H-labelled inositol phosphates were larger in both strains after 20 min of NE stimulation, and the response in α1A-TG mice was 8–10-fold higher than that in α1A-WT mice (results not shown).
Overexpression of α1A-AR enhances 3H-labelled inositol phosphate responses to NE in mouse hearts
Post-ischaemic reperfusion causes [3H]Ins(1,4,5)
P3 responses in hearts from α1A-WT, but not α1A-TG, mice
Our previous studies in rat and mouse hearts have shown that activation of α1-AR in early post-ischaemic reperfusion either by endogenously released NE, or NE added to the perfusate, causes substantial Ins(1,4,5)P3 generation [3,4]. Ischaemia for 30 min did not alter the content of [3H]Ins(1,4,5)P3 but, as shown in Figure 4, reperfusion for 2 min, after 30 min of ischaemia, caused a robust [3H]Ins(1,4,5)P3 response in hearts from α1A-WT mice. We next investigated whether heightened α1A-AR activity increased this response. In contrast with hearts from α1A-WT mice, there was no increase in [3H]Ins(1,4,5)P3 in hearts from α1A-TG mice after 2 min of reperfusion following 30 min of ischaemia (Figure 4).
Hearts from α1A-WT, but not α1A-TG, mice generate [3H]Ins(1,4,5)
P3 in early post-ischaemic reperfusion
The NE content of the perfusate increased substantially with 2 min of reperfusion in hearts from both α1A-WT and α1A-TG mice, and the amount of release was not different between the two groups. Thus the failure of α1A-TG to generate Ins(1,4,5)P3 during reperfusion cannot be explained by a lack of NE release activating the overexpressed receptors (Table 1).
|.||NE (nmol/l) .|
|Treatment .||α1A-WT mice .||α1A-TG mice .|
|Ischaemia (20 min)||0||0|
|Ischaemia (20 min)+reperfusion (2 min)||2.36±0.58**||1.54±0.31**|
|Ischaemia (30 min)||0||0|
|Ischaemia (30 min)+reperfusion (2 min)||1.92±0.32**||1.76±0.34**|
|.||NE (nmol/l) .|
|Treatment .||α1A-WT mice .||α1A-TG mice .|
|Ischaemia (20 min)||0||0|
|Ischaemia (20 min)+reperfusion (2 min)||2.36±0.58**||1.54±0.31**|
|Ischaemia (30 min)||0||0|
|Ischaemia (30 min)+reperfusion (2 min)||1.92±0.32**||1.76±0.34**|
In previous studies, we have shown that ischaemia/reperfusion causes a substantial increase in overall PLC activity in addition to Ins(1,4,5)P3 generation [4,21]. Thus not only is Ins(1,4,5)P3 increased above the level seen in normoxia, but so are all of its metabolites. To make a direct comparison of the overall 3H-labelled inositol phosphate response (which reflects PLC activity) in normoxia and reperfusion, we added saturating amounts of NE to the perfusate during the 2-min reperfusion period and compared the total 3H-labelled inositol phosphate (PLC) responses with responses to 2 min of NE in normoxia. As shown in Figure 5 (upper-left-hand panel), the total 3H-labelled inositol phosphate (PLC) response in hearts from α1A-WT mice was substantially heightened under reperfusion conditions. In marked comparison, there was no significant increase in overall 3H-labelled inositol phosphate accumulation in hearts from α1A-TG mice under reperfusion conditions (Figure 5, upper-right-hand panel). Thus α1-AR-mediated PLC activity is increased by ischaemia/reperfusion in α1A-WT mice, but not in α1A-TG mice.
Hearts from α1A-WT, but not α1A-TG, mice have enhanced PLC responses in early reperfusion
As shown in the lower panels in Figure 5, α1A-WT mice had an increase in [3H]Ins(1,4,5)P3 with added NE during 2 min of post-ischaemic reperfusion. However, even with NE added, there was no [3H]Ins(1,4,5)P3 response in α1A-TG mice. This shows that the lack of Ins(1,4,5)P3 response seen in α1A-TG mice is not due to a lack of available endogenous NE.
Post-ischaemic reperfusion increases PtdIns(4,5)
P2 in hearts from α1A-WT and α1A-TG mice
Only hearts from α1A-WT mice released Ins(1,4,5)P3 upon reperfusion after ischaemic insult and this was not due to a lack of NE. Another possible explanation for this was that there were differences in substrate availability between the two strains, for instance that [3H]PtdIns(4,5)P2 was limiting in hearts from α1A-TG but not in α1A-WT mice. To investigate this possibility, 3H-labelled inositol phospholipids were extracted and quantified in the same hearts used for the 3H-labelled inositol phosphate studies. Ischaemia caused significant decreases in [3H]PtdIns(4,5)P2, and this response was similar in hearts from α1A-WT and α1A-TG mice. In both α1A-WT and α1A-TG hearts, [3H]PtdIns(4,5)P2 increased substantially with 2 min of reperfusion (Figure 6). Thus differences in PtdIns(4,5)P2 availability do not appear to explain the different Ins(1,4,5)P3 responses in the two strains.
Changes in [3H]PtdIns(4,5)
P2 with ischaemia and reperfusion are similar in hearts from α1A-WT and α1A-TG mice
Preconditioning prevents Ins(1,4,5)
P3 generation in hearts from α1A-WT mice after 30 min of ischaemia
Because preconditioning by α1A-ARs has been suggested previously , we considered the possibility that the inhibitory effect of heightened α1A-AR activity was caused by a preconditioning mechanism. To investigate this possibility, we examined the effect of ischaemic preconditioning on Ins(1,4,5)P3 responses in α1A-WT mice after 30 min of ischaemia. Preconditioning did not cause any change in 3H-labelled inositol phosphates after 30 min of ischaemia, as we have reported in rat hearts previously . However, preconditioning completely prevented the Ins(1,4,5)P3 response in early reperfusion in α1A-WT mice (Figure 4). Similar experiments were performed using hearts from α1A-TG mice, where preconditioning had no detectable effect. Thus, in terms of Ins(1,4,5)P3 generation, the effects of heightened α1A-AR activity appeared to be similar to the effects of preconditioning.
To examine further the possibility that heightened α1A-AR activity mimicked ischaemic preconditioning, we next examined increases in [3H]PtdIns(4,5)P2 during 2 min of reperfusion after 30 min of ischaemia. Ischaemic preconditioning prevented the increase in [3H]PtdIns(4,5)P2 in early reperfusion, unlike the scenario seen with α1A-AR overexpression (Figure 6). It is likely, therefore, that preconditioning limits the Ins(1,4,5)P3 responses in reperfusion by inhibiting the generation of the precursor PtdIns(4,5)P2. Thus, although both increased α1A-AR activity and ischaemic preconditioning protected the hearts from Ins(1,4,5)P3 increases in early reperfusion, these may involve different mechanisms.
In both rat and mouse hearts, a brief ischaemic insult causes substantial enhancement of early signalling events initiated via α1-ARs, resulting in heightened PLC activity and quantifiable production of Ins(1,4,5)P3 in the post-ischaemic reperfusion period [4,21]. The results of the present study show that overexpression of α1A-ARs prevents the Ins(1,4,5)P3 generation that occurs when hearts are subjected to reperfusion after brief ischaemic periods (Figure 4). Hearts from α1A-WT mice, when subjected to ischaemia and reperfusion, had substantially heightened PLC responses and quantifiable generation of Ins(1,4,5)P3. In contrast, in hearts from α1A-TG mice, there was no Ins(1,4,5)P3 response in reperfusion, and the overall PLC activity was not heightened further above that observed in normoxia. The lack of response in early reperfusion in α1A-TG mice was not related to insufficient NE release, as endogenously released NE in the perfusate was similar in α1A-WT and α1A-TG mice (Table 1). Furthermore, addition of excess NE to the perfusate did not uncover a heightened response in α1A-TG mice (Figure 5) and, even in the presence of excess NE, α1A-TG mice did not have Ins(1,4,5)P3 production in reperfusion (Figure 5, lower panels). Both of these observations argue that the lack of reperfusion response in α1A-TG mice is not related to availability of NE.
In addition to Ins(1,4,5)P3 generation, reperfusion following a brief ischaemic period increased [3H]PtdIns(4,5)P2, the immediate precursor of Ins(1,4,5)P3. PtdIns(4,5)P2, in addition to its role as the immediate precursor of Ins(1,4,5)P3, is also a major regulator of a number of ion channels and exchangers critical for the maintenance of cardiac rhythm [23,24]. Thus changes in PtdIns(4,5)P2 could be arrhythmogenic in their own right, independently of any effect on Ins(1,4,5)P3 generation. In the present study, we have shown that PtdIns(4,5)P2 increased in reperfusion similarly in hearts from α1A-WT and α1A-TG mice and, thus, PtdIns(4,5)P2 availability does not explain the lack of Ins(1,4,5)P3 response in α1A-TG mice.
Traditionally the α1A-AR subtype has been implicated as a mediator of pathological responses in heart. Stimulation of α1A-ARs is sufficient to cause cardiomyocyte hypertrophy and, furthermore, hypertrophy in vivo and in neonatal rat cardiomyocytes selectively increased the expression of α1A-AR mRNA. This implies a self-perpetuating action of α1A-ARs [14,15]; however, more recent studies have suggested, on the contrary, that α1A-AR are cardioprotective [16,17]. Evidence was presented that the α1A-AR subtype, when overexpressed, improved functional recovery after ischaemic injury [8,16]. Furthermore, even physiological levels of α1A-AR activity appear to be effective in protecting the myocardium from apoptosis caused by insults that mimic ischaemia . Thus it is possible that the observed effect of α1A-AR activity in reperfusion represents a cardioprotective action. If this is so, then the mechanisms involved appear to be different from those involved in ischaemic preconditioning. Ischaemic preconditioning prevented Ins(1,4,5)P3 responses in early reperfusion, but, in this case, the loss of the Ins(1,4,5)P3 response appeared to be associated with a failure to increase the precursor PtdIns(4,5)P2. Thus the well-established effectiveness of preconditioning in reducing reperfusion arrhythmias [25,26] might be related to the loss of either Ins(1,4,5)P3 or PtdIns(4,5)P2 responses, either of which have the potential to trigger arrhythmias.
Ins(1,4,5)P3 causes Ca2+ release from IP3-Rs [Ins(1,4,5)P3 receptors] localized on the sarcoplasmic reticulum and the perinuclear membrane [27,28]. Even though the expression of IP3-Rs is low in the heart, there is evidence that IP3-R activation can perturb the Ca2+-induced Ca2+-release programme orchestrated by the more prevalent ryanodine receptors and leads to increased Na+/Ca2+ exchange , a possible mediator of the observed arrhythmias. Furthermore, removal of the type 2 IP3-R, the subtype expressed in working cardiomyocytes , prevented potentially arrhythmogenic responses to PLC activation in mouse atrial myocytes . Importantly, increased IP3-R expression is seen in failed human ventricular tissue  and in atrial samples from patients predisposed to atrial fibrillation , suggesting a contribution of Ins(1,4,5)P3 to human pathology. In previous studies, we have shown that the generation of Ins(1,4,5)P3 in early reperfusion is associated with the initiation of arrhythmias. Although rat hearts develop ventricular fibrillation under such conditions [2,3], we have shown previously that mice of the C57 strain only develop ventricular tachycardia and ectopic beats . Moreover, both ventricular tachycardia and ectopic beats are eliminated by overexpression of α1B-ARs, and this effect was associated with the loss of an Ins(1,4,5)P3 response. Unfortunately, the FVB strain of mouse used in the present study was resistant to ventricular tachycardia development after 30 min of ischaemia. In fact, we observed only the occasional ectopic beat in hearts from either α1A-WT or α1A-TG mice. Therefore it was not possible to determine whether the lack of Ins(1,4,5)P3 responses in hearts from α1A-TG mice protects against arrhythmias. Heightened Ins(1,4,5)P3 causes ectopic Ca2+ spikes that can initiate early after-depolarizations and interfere with the functioning of sarcolemmal ion channels and exchangers [29,31]. It is likely, therefore, that limiting Ins(1,4,5)P3 would protect from arrhythmogenesis and may be an important component of ischaemic preconditioning . In previous studies, we  and others  have shown that overexpressed α1A-AR protect from ischaemic injury, whereas α1B-AR do not. Activation of the α1-AR has been reported to mediate a preconditioning action apparently similar to that afforded by brief ischaemic periods [35,36]. However, in the present study, heightened α1A-AR activity probably prevented Ins(1,4,5)P3 responses by mechanisms that differ from ischaemic preconditioning.
The experiments in the present study show that ischaemia and subsequent reperfusion in mouse hearts causes a complex array of changes in the generation of the inositol phosphates and their precursor inositol phospholipids. Of these, only the generation of Ins(1,4,5)P3 during early reperfusion is altered by heightened α1A-AR activity, which prevented the response. This points to a possible protective action of α1A-ARs in heart.
The work was supported by grants from the National Health and Medical Research Council of Australia, 826921 and 317802, as well as a Principal Research Fellowship to E.A.W. (317803). R.M.G. is supported by Programme Grant #354400 from the National Health and Medical Research Council of Australia. NE measurements were performed by Dr Jacqueline Hastings (Human Neurotransmitter Laboratory, Baker Heart Research Institute).