Activation of Akt is essential for acetylcholine to trigger generation of oxygen free radicals

Abstract

Objectives: Acetylcholine (ACh) receptor activation in the heart causes mitochondrial production of reactive oxygen species (ROS) that is dependent on mitochondrial K_ATP channel opening. Recent data show that Akt (also known as protein kinase B) is phosphorylated at its activation site following exposure to ACh. However, since no reliable Akt inhibitor is available, it has not been possible to determine whether Akt activation is an actual step in the protective pathway. Methods: Cultured rat vascular smooth muscle cells (A7r5) were transiently transfected with a dominant negative Akt (Akt-AAA), thus inhibiting the ability of ACh in these cells to phosphorylate Akt. Transfected cells were identified by co-transfection of enhanced green fluorescent protein (EGFP). ROS production was determined by incubating the cells for 15 min with 1 mM reduced MitoTracker Red which becomes fluorescent only after reacting with ROS. Cells were then triple-washed to remove any voltage-dependent pool of dye and single cell fluorescence was measured. Results: ACh exposure (2 mM) led to a 1.64±0.15-fold increase in the average fluorescence over that seen in untreated cells (P = 0.002). A similar increase in ROS production occurred after treatment with either the K_ATP channel opener diazoxide (DIAZ) or the potassium ionophore valinomycin (VAL). Akt-AAA transfection abolished ACh-induced ROS production, but not increased ROS production after treatment with either DIAZ or VAL. Conclusion: Thus, at least in the smooth muscle cell model, Akt phosphorylation is an important step in the signal transduction pathway leading from ACh receptor activation to the generation of ROS. The experiments reveal that Akt is positioned between the receptor and the K_ATP channel in this model.

Acetylcholine, Free radicals, K-ATP channel, Protein phosphorylation, Signal transduction, Smooth muscle

Time for primary review 19 days.

1 Introduction

The signal transduction mechanisms mediating receptor-induced cardioprotective effects of ischemic preconditioning have been the subject of many recent investigations. It is known that occupation of several G_i-coupled receptors is a triggering event for this protection [1]. We [2] as well as others [3] have demonstrated that the neurotransmitter acetylcholine (ACh) can mimic preconditioning by occupying cardiac muscarinic receptors which are G_i-coupled. We have found that in the rabbit heart cardiac receptors can be divided into two classes [2]. Those in the ‘muscarinic’ class trigger protection through a pathway dependent on opening of mitochondrial ATP-dependent potassium channels (mitoK_ATP) and the generation of reactive oxygen species (ROS). These include the bradykinin, α₁-adrenergic and opioid receptors. Adenosine receptors, which are also G_i-coupled, trigger the preconditioned state in a manner that requires neither mitoK_ATP opening nor ROS generation [2]. We recently developed a smooth muscle cell model to study the mitoK_ATP-ROS pathway [4]. We found that in vascular smooth muscle cells ACh indeed leads to the generation of ROS, and that the latter could be blocked by 5-hydroxydecanoate, a potent inhibitor of mitoK_ATP channels [5]. Interestingly we found that wortmannin, an inhibitor of phosphatidylinositol 3-kinase (PI3-kinase) could abolish the ROS signal following ACh treatment, but not that seen after administration of diazoxide, a direct opener of mitoK_ATP[5]. That led us to propose that PI3-kinase is positioned between the muscarinic receptor and the mitoK_ATP. We explored this pathway in the rabbit heart and found that the serine/threonine kinase Akt (also known as protein kinase B, PKB) was activated after infusion of ACh [6]. Akt was studied because it is known that Akt is a direct downstream target of PI3-kinase. PI3-kinase's 3-phosphorylated lipid products, phosphatidylinositol 3,4,5-trisphosphate [PtdIns(3,4,5)P₃] and PtdIns(3,4)P₂, activate 3′-phosphoinositide-dependent kinase-1 (PDK1) and PDK2 which in turn phosphorylate Akt on its serine and threonine residues [7,8,9]. Indeed, wortmannin was able to block the ACh-induced activation of Akt in the rabbit heart [6]. The mechanism by which PI3-kinase is activated in response to G_i-protein-coupled receptor activation is not fully known, but transactivation of one or more receptor tyrosine kinases seems likely [6,10,11].

So far, 3 isoforms of Akt have been identified, Akt1, -2, and -3 (PKBα, β and γ), and it is still not fully known which isoforms are activated upon stimulation with the various ligands [12]. In the present study, we concentrated on Akt1 which is known to be expressed at high levels in the heart and is activated through the hierarchical phosphorylation of both Thr308 and Ser473.

Although PI3-kinase activation is required for preconditioning's protection [13,14], PI3-kinase has multiple downstream targets including Src-family tyrosine kinases as well as Akt. Because there is no known pharmacological inhibitor of Akt, it has not been possible to test whether Akt phosphorylation is a step in the signal transduction pathway leading to protection or whether its activation is merely an epiphenomenon. To answer this question we turned to the A7r5 cell model in which mitochondrial ROS generation can be directly measured [5]. We transfected these cells with a dominant negative form of Akt and tested whether the resulting suppression of Akt would affect mitochondrial ROS generation in response to ACh.

2 Methods

2.1 A7r5 cells

Aortic vascular smooth muscle cells (A7r5) from Rattus norvegicus were purchased from American Type Culture Collection (ATCC, Manassas, VA). Cells were grown in 75 cm² tissue culture flasks (Corning Inc., Corning, NY), and replated before confluence. Experiments were performed on cells (60–80% confluent) which were grown for 24 to 48 h in 35 mm cell culture dishes (Corning Inc., Corning, NY). Cells of passages 3 to 9 were used for the experiments. Dulbecco's Modified Eagle's Medium (DMEM) containing 10% fetal bovine serum (FBS), 4 mM l-glutamine, 4.5 g/l glucose, 1.5 g/l sodium bicarbonate, 0.11 g/l pyruvate, penicillin, gentamicin, and fungizone was used for cell growth as recommended by ATCC. Cells were grown and incubated with experimental drugs in the dark at 37°C in air enriched with 5% CO₂.

2.2 Experimental design

For each experiment the medium was changed to FBS- and pH-indicator-free DMEM. Cells were incubated for a total of 40 min in the new medium and then reduced MitroTracker Red (MTR) (Molecular Probes, Eugene, OR) was added to the medium (1 mM final concentration) for an additional 15 min. When the dye is irreversibly oxidized by ROS, it becomes fluorescent and is concentrated in mitochondria. If an activator [acetylcholine (ACh), diazoxide (DIAZ) or valinomycin (VAL)] were used, it was added 10 min after changing the medium. After 15 min incubation with MTR, cells were washed three times with fresh DMEM containing the drugs but no MTR to remove extra-mitochondrial and unbound MTR. Fluorescence was determined at the end of the protocol. Four to 6 repetitions were performed for each group on different days, with cells of different passages, and with different MTR lots.

2.3 Transfection

The dominant negative Akt-AAA in a plasmid using a CMV promoter was kindly provided by Dr. Woodgett (Ontario Cancer Institute, Toronto, Canada). In this N-terminal haemagglutinin (HA)-tagged mutant of Akt1 (PKBα) the two major regulatory phosphorylation sites (Thr308 and Ser473) and the phosphate transfer residue in the catalytic site (Lys179) were replaced by alanine residues rendering the protein product inactive (see [15] for details). Transient transfections were performed in 35 mm cell culture dishes via liposome-mediated transfer according to the Effectene product manual (Qiagen Inc., Valencia, CA). Transfection was performed with a mixture of 1.3 μg Akt-AAA and 0.25 μg enhanced green fluorescent protein (EGFP) with a constant ratio of 8:1 for the enhancer/DNA complexes. The EGFP-plasmid [pEF1α(1a)EGFP] was kindly provided by Dr. David Engler (Texas Heart Institute, Houston, TX). DNA complexes were introduced to the cells 24 h before the start of the experiments and then removed. To rule out an independent effect of EGFP, transfection with identical enhancer/DNA ratio was performed with EGFP alone.

2.4 Flow cytometric cell sorting

Cells were transfected with either EGFP alone (tf0) or with both EGFP and Akt-AAA (tf). Non-transfected cells acted as controls. All cells were treated with ACh (2 mM). Transfected cells were sorted with a Becton/Dickinson Vantage SE fluorescence activated cell sorter (FACS) using an excitation of 488 nm and emission at 515–545 nm to detect cells expressing EGFP. Cells were immediately lysed and used for immunoblotting with the phospho-specific Akt antibody.

2.5 Immunoblotting

Cells were lysed in ice-cold lysis buffer containing Tris–HCl (pH 7.5) 20 mM, NaCl 150 mM, Na₂EDTA 1 mM, EGTA 1 mM, 1% Triton, sodium pyrophosphate 2.5 mM, β-glycerophosphate 1 mM, Na₃VO₄ 1 mM and leupeptin 1 μg/ml. Protein content of the supernatant was measured using the Bradford technique. SDS–Page was performed with equal amounts of protein on a 10% separation gel. After transfer to a nitrocellulose membrane the blot was incubated with a primary antibody for 18 h at 4°C. For analysis of cells after FACS cell sorting a primary phospho-specific antibody to Ser473 of Akt was used, whereas a primary antibody against haemagglutinin (HA) was used to test the expression of the Akt-AAA plasmid. After incubation with the appropriate horseradish peroxidase-conjugated secondary antibody, the signal was detected with LumiGLO reagent (Cell Signaling Technology, Beverly, MA).

2.6 Measurement of ROS

Production of mitochondrial ROS was analyzed by measuring single cell fluorescence using a Leica TCS SP2 (Leica Microsystems, Exton, PA) confocal laser scanning microscope. To measure ROS, excitation at 543 nm and emission at 575–650 nm were used, whereas identification of EGFP was performed at an excitation of 488 nm and emission of 515–535 nm. In non-transfected control cells fluorescence was quantified after manually tracing the perimeters of 4–10 cells per field in 10 random fields per dish, leading to analysis of 60 to 100 cells per experiment. The average single-cell fluorescence for each dish was computed. This fluorescence value was then compared to that measured in dishes of treated cells studied on the same day. In transfection experiments transfected cells were identified by expression of EGFP. ROS was measured as described above by tracing 1–5 of these fluorescent green cells per field in 20 random fields, leading as well to analysis of 60–100 cells per experiment.

2.7 Acetylcholine, diazoxide and valinomycin

In the first group of studies the impact of 3 activators, DIAZ (200 μM), VAL (100 nM) and ACh (2 mM), on ROS production in non-transfected cells was investigated by comparing fluorescence in cells treated with each of these agents to that of a parallel, untreated control sample. Fluorescence of each group of treated cells was compared to that of the corresponding control experiment at the same time, with cells from the same passage, and with the same MTR lot.

In the second group the 3 activators were tested in the same manner in cells previously transfected with Akt-AAA. The success of the transfection was confirmed by identification of the results of co-transfection with EGFP genetic material. Only cells fluorescing green, and, therefore, expressing EGFP, were studied. To exclude a possible effect of EGFP itself on ROS production, transfection with EGFP alone was performed in additional cells and the effect of activation with ACh on ROS generation was measured.

2.8 Chemicals

All experimental drugs and the monoclonal anti-HA antibody were purchased from Sigma Chemical (St. Louis, MO). The phospho-specific antibody against Ser473 of Akt (#9271) and the HRP-conjugated goat anti-rabbit IgG antibody were purchased from Cell Signaling Technology. The Effectene Transfection Reagent Kit was purchased from Qiagen (Valencia, CA), FBS was purchased from Gibco BRL, Life Technologies (Rockville, MD) and MitroTracker Red from Molecular Probes (Eugene, OR). The HRP-conjugated goat anti-mouse IgG antibody was obtained from Upstate Biotechnology (Waltham, MA). Either distilled water (ACh) or DMSO (MTR, DIAZ, VAL) was used to dissolve the drugs and prepare stock solutions. The final DMSO concentration was kept below 1%. All stock solutions including MTR were made freshly every day.

2.9 Data analysis

For untreated cells we have noted that the fluorescence measurements vary widely between individual cultures and MTR lots. To account for that variability fluorescence of cells depicted in the figures is expressed as a percentage of that in an untreated control dish that was studied in parallel fashion during each experiment. The text reports the relative fluorescence of the experimental cells normalized to that of the respective control cells. These were then averaged for the different experiments. Confidence limits are standard errors of the mean. A t-test was used to test for differences in the relative fluorescence intensity of the groups. Differences were considered significant if the P value was <0.05.

3 Results

3.1 Transfection with Akt-AAA

Co-transfection of the A7r5 cells with Akt-AAA and EGFP using the Effectene transfection reagent resulted in a transfection rate of approximately 10–20% as evaluated by counting the fluorescent green cells expressing EGFP. To further investigate whether the transfection of the HA-tagged Akt-AAA plasmid was effective, a Western blot was made of non-transfected and transfected cells using an anti-HA primary antibody. Fig. 1A presents a blot representing one of 4 similar experiments documenting a single band at ≈60 kD representing the expression of HA-tagged Akt-AAA in the transfected cells. As expected, the blot shows the presence of the HA tag only in the transfected cells.

Fig. 1

A. Representative immunoblot for anti-HA antibody in untreated (control) and dominant negative HA-tagged Akt-AAA transfected A7r5 cells. tf=transfected cells. B. Representative immunoblot for phospho-Akt after treatment with ACh (2 mM) in non-transfected A7r5 cells (control), cells transfected with EGFP alone (tf0) and cells transfected with EGFP and Akt-AAA (tf).

3.2 Akt-AAA acts as a dominant-negative inhibitor of Akt1

To test whether the transfection with Akt-AAA can suppress the phosphorylation of Akt, cells were transfected with either EGFP alone or EGFP and Akt-AAA. After extracting only the cells expressing EGFP with FACS, western blot analysis with a phospho-specific antibody against Akt was performed to test whether Akt-AAA transfection suppresses Akt phosphorylation. As seen in Fig. 1B the non-transfected cells (control) and the cells transfected with only the EGFP (tf0) had a significant increase in Akt phosphorylation after treatment with ACh, whereas this response was abolished in the cells transfected with EGFP and Akt-AAA (tf).

3.3 ROS production following ACh treatment

Whereas control cells with no intervention create a basal level of ROS measured by single cell fluorescence, incubation with 2 mM ACh caused a 1.64±0.15-fold increase compared to that in untreated cells (n = 6 dishes, P = 0.002 vs. untreated) (Fig. 2). We next looked at ROS production in cells transfected with EGFP alone. Fluorescence measurements were made only in the green fluorescent cells expressing EGFP. ACh caused a significant increase in ROS production over that seen in the untreated EGFP-transfected cells (1.26±0.17-fold vs. untreated, n = 4 dishes, P = 0.02). While the ACh-induced ROS production averaged less in the EGFP-transfected cells than that seen in the non-transfected cells, the difference was not significant. ROS production in response to ACh was totally abolished in cells co-transfected with Akt-AAA and EGFP (0.92±0.08 fold vs. control, n = 6 dishes, P = ns) (ACh-tf in Fig. 2). As noted above, MTR fluorescence was measured only in cells expressing EGFP.

Fig. 2

ROS production as measured by cell fluorescence. Treatment of A7r5 cells with 2 mM acetylcholine (ACh) increases ROS generation in non-transfected cells compared to the baseline fluorescence seen in untreated cells. Transfection with the dominant negative Akt-AAA totally abolished the ACh-induced ROS generation (ACh-tf, P = ns vs. control). Transfected cells were identified by co-transfection of EGFP. The bars depict the group average±S.E.M. All fluorescence is normalized to that seen in untreated (control) cells from a paired culture. *P<0.05 vs. control.

3.4 Diazoxide and valinomycin increase ROS independently of Akt

Exposure of non-transfected A7r5 cells to 200 μM DIAZ caused the production of ROS to increase to 1.35±0.26-fold of the basal single cell fluorescence measured in parallel in control cells (n = 4 dishes, P = 0.037 vs. control) (Fig. 3). DIAZ caused a similar increase in ROS production in Akt-AAA transfected cells with a 1.46±0.27-fold increase over transfected cells not exposed to diazoxide (n = 5 dishes, P = 0.005) (DIAZ-tf in Fig. 3). Again only cells expressing EGFP were studied.

Fig. 3

ROS production as measured by cell fluorescence. Treatment of A7r5 cells with the mitoK_ATP opener diazoxide (DIAZ) (200 μM) increases ROS generation by a similar amount in both non-transfected (DIAZ) and transfected (DIAZ-tf) cells as compared to the baseline fluorescence seen in untreated cells. Transfected cells were identified by co-transfection of EGFP. *P<0.05 vs. control.

Similar results were seen after treating the cells with the mitochondrial potassium ionophore VAL (100 nM). Exposure of non-transfected A7r5 cells led to a 1.42±0.37-fold increase in cell fluorescence (n = 6 dishes, P = 0.018 vs. control) (Fig. 4). A similar increase with VAL was seen in cells transfected with Akt-AAA (1.45±0.43-fold vs. control, n = 5 dishes, P = 0.047) (VAL-tf in Fig. 4). The experiments with DIAZ and VAL prove that the transfected cells were still capable of producing ROS when potassium entered the mitochondria, either from the opening of K_ATP channels or from the action of the ionophore.

Fig. 4

ROS production as measured by cell fluorescence. Treatment of A7r5 cells with the potassium ionophore valinomycin (VAL) (100 nM) increases ROS generation by a similar amount in both non-transfected (VAL) and transfected (VAL-tf) cells as compared to the baseline fluorescence seen in untreated (control) cells. Transfected cells were identified by co-transfection of EGFP. *P<0.05 vs. control.

4 Discussion

There is compelling evidence that activation of PI3-kinase is an important step in the signaling cascade leading to preconditioning [13,14,16]. There are also several reports that Akt, a downstream target of PI3-kinase, is activated during preconditioning [6,13]. However, because a suitable inhibitor of Akt has not been available, it has been difficult to determine whether activation of Akt is an actual step in transducing the protective signal or just an epiphenomenon. We [2,17] and others [18–20] have proposed that G_i-coupled receptors trigger protection by causing opening of mitoK_ATP which causes the mitochondria to produce ROS. Indeed mitoK_ATP-dependent ROS production has now been seen in a variety of cell models [4,5,18,20]. In the present study we present strong evidence using an Akt-deficient model that Akt is an essential link in the muscarinic receptor's ability to generate ROS in A7r5 cells. Furthermore we can show that Akt is positioned in the signaling pathway somewhere between the surface muscarinic receptor and mitoK_ATP.

For these studies we transfected A7r5 cells with a mutant Akt gene having alanine substitutions at the 3 key phosphorylation sites in the Akt protein making it impossible for the latter to be activated. The dominant-negative Akt abolished ACh-induced ROS production. Yet the transfection had no effect on ROS generation induced by the mitoK_ATP opener diazoxide or the potassium ionophore valinomycin indicating that the transfected cells still had the ability to generate ROS upon K⁺ entry. These functional data suggest that the activation of Akt is a key step in ACh signaling that leads to mitochondrial ROS generation which ultimately triggers the preconditioned state.

Ischemic preconditioning is a well known endogenous protective mechanism first described by Murry et al. [21] in which brief periods of ischemia/reperfusion make the heart very resistant to infarction from a subsequent period of prolonged ischemia. Much research has been devoted to determine the mechanism of this phenomenon. During the preconditioning ischemia ligands to several G_i-coupled receptors are produced by the cell including adenosine, bradykinin and opioids. Past studies have established that the neurotransmitter ACh also has the ability to trigger the preconditioning response [2,22], although ACh is not released by the heart during ischemic preconditioning. ACh acts through the G_i-coupled muscarinic receptor and has been a useful tool for pharmacological activation of preconditioning's signal transduction pathway. In a previous study using rabbit hearts we found that ACh, opioids and bradykinin all trigger preconditioning's anti-infarct effect through the mitoK_ATP/ROS pathway [2]. ACh is the drug of choice for the A7r5 model because these cells express the muscarinic receptor [23]. Indeed, 2 mM ACh induces ROS production in these cells that is entirely receptor-mediated since atropine abolishes the signal [5].

Using the PI3-kinase inhibitor wortmannin it has been easy to demonstrate that PI3-kinase is a necessary step in the protection from either ischemic [13] or ACh [14] preconditioning. Furthermore wortmannin blocks ROS production from ACh in A7r5 cells [5]. Akt is a downstream target for PI3-kinase in many cell types [24]. Three isoforms of Akt have been identified (Akt1, -2, and -3). The most prevalent isoform, Akt1, needs to be phosphorylated by both 3′-phosphoinositide-dependent kinase-1 (PDK1) at the Thr308 domain and PDK2 at Ser473 to achieve maximal activation [12]. The Akt1 isoform is ubiquitous and most of the studies, as well as the present one, have either focused on Akt1 or have not discriminated among the different isoforms.

In a previous study Matsui and colleagues [25] transfected cardiomyocytes with either a constitutively active Akt or PI3-kinase. In either transfection the cells became very resistant to apoptosis. More recently they transfected rat hearts with a constitutively active Akt and found the hearts to be very resistant to infarction [26]. Several other reports suggest that Akt activation is protective to the heart [13,16,27]. But, to our knowledge, there are no reports of the use of a selective Akt inhibitor and all of these studies have used gene transfer of either constitutively active or dominant negative forms of Akt. In the present study, we found that expression of the dominant negative Akt-AAA mutant blocked the pathway thought to be required for triggering entrance into the preconditioned state.

An obvious limitation of the present study is the use of rat vascular smooth muscle A7r5 cells to investigate intracellular signaling. To achieve a sufficient transfection rate and expression of Akt-AAA we needed a fast growing cell line in which a G_i-coupled receptor could elicit a robust K_ATP-dependent ROS signal. Because adult rabbit cardiomyocytes do not divide in culture the transfection efficiency would be expected to be low. However, the recently reported results in A7r5 cells pertinent to intracellular signal transduction [5] have now been reproduced in adult rabbit cardiac myocytes [28]. Based on these studies we believe that the A7r5 cell is an adequate model for investigation of the mitoK_ATP/ROS pathway.

In these studies we used reduced MitoTracker as our fluorochrome for measuring ROS production. Reduced MitoTracker is not fluorescent until it is oxidized by ROS. The oxidized form is then sequestered in the mitochondria as a result of both the voltage gradient and chemical binding to proteins in the mitochondria. Because opening of mitoK_ATP channels can change the mitochondrial potential it is important to assure that the fluorescence that we measure reflects only ROS production. To accomplish this we incubate the cells for 15 min in reduced MitoTracker and then triple-wash the cells to remove any unbound chemical. Washing apparently removes all of the unbound, and thus voltage-dependent, pool since we have found that the fluorescence in the present model is unaffected by depolarizing the mitochondria with dinitrophenol [4]. Additionally, either of the free radical scavengers N-(2-mercaptopropionyl) glycine or Tiron could completely block the fluorescence signal proving that ROS was causing it [4]. Therefore we do believe that MitoTracker as used here is a reliable indicator of ROS production in A7r5 cells.

We co-transfected the cells with EGFP which was used as a reporter. Measurements of cell fluorescence in the transfected dishes revealed that EGFP had been expressed in 10–20% of the cells. Unfortunately, the EGFP was not on the same plasmid as the HA-tagged Akt-AAA. Rather we mixed the two plasmids, but the mixture was such that there was a predominance of the Akt-AAA plasmid. This is an accepted technique and the reasoning is that any cell that took up the EGFP plasmid would then most certainly have taken up the HA-tagged Akt-AAA plasmid as well. We were able to test for the expression of the HA-tagged Akt-AAA plasmid by immunoblotting with anti-HA antibody (Fig. 1A) and clearly some of the cells were expressing the HA-tagged Akt-AAA as well. The ability of this method to sufficiently abolish Akt activity has been shown by others using the same Akt-AAA plasmid and an identical transfection protocol [15]. We indeed found that ACh-induced phosphorylation of the Akt-AAA transfected cells that had been sorted according to their green fluorescence was virtually abolished proving that Akt activity had been suppressed (Fig. 1B).

In summary, this study provides compelling evidence that Akt plays a central and crucial role in the signaling cascade by which ACh leads to mitochondrial ROS production. We believe that the signal transduction system that we have measured in the A7r5 cells is identical to the one present in the heart that is responsible for receptor ligand triggering of the ischemic preconditioning phenomenon. Furthermore because the transfection with a dominant negative Akt could not block protection from a direct opener of mitoK_ATP channels, we conclude that Akt is positioned in the pathway between the receptor and the mitoK_ATP channel.

Acknowledgements

This study was supported in part by grants HL-20648 (J.M.D.) and HL-50688 (M.V.C.) from the National Heart, Lung, and Blood Institute of NIH. T.K. was supported by a grant from Aventis Pharmaceuticals Inc.