Neuronal nitric oxide synthase gene transfer decreases [Ca2+]i in cardiac sympathetic neurons

Available online at www.sciencedirect.com

Journal of Molecular and Cellular Cardiology 43 (2007) 717 – 725 www.elsevier.com/locate/yjmcc

Original article

Neuronal nitric oxide synthase gene transfer decreases [Ca 2+ ]i in cardiac sympathetic neurons Lijun Wang, Michael Henrich, Keith J. Buckler, Mary McMenamin, Christopher J. Mee, David B. Sattelle, David J. Paterson ⁎ Burdon Sanderson Cardiac Science Centre, Department of Physiology, Anatomy and Genetics, Sherrington Building, Parks Road, Oxford, OX1 3PT, UK MRC Functional Genetics Unit, Le Gros Clark Building, South Parks Road, Oxford, OX1 3QX, UK Received 13 July 2007; received in revised form 30 August 2007; accepted 6 September 2007 Available online 19 September 2007

Abstract Gene transfer of neuronal nitric oxide synthase (nNOS) can decrease cardiac sympathetic outflow and facilitate parasympathetic neurotransmission. The precise pathway responsible for nitric oxide (NO) mediated inhibition of sympathetic neurotransmission is not known, but may be related to NO–cGMP activation of cGMP-stimulated phosphodiesterase (PDE2) that enhances the breakdown of cAMP to deactivate protein kinase A (PKA), resulting in a decrease in Ca2+ influx mediated exocytosis of the neurotransmitter. We investigated depolarization evoked Ca2+ influx in nNOS gene transduced sympathetic neurons from stellate ganglia with a noradrenergic cell specific vector (Ad.PRS-nNOS or empty vector), and examined how nNOS gene transfer affected cAMP and cGMP levels in these neurons. We found that targeting nNOS into these sympathetic neurons reduced amplitudes of voltage activated Ca2+ transients by 44%. nNOS specific inhibition by N-[(4S)-4-Amino-5-[(2aminoetyl](amino] pentyl]-N′-nitroguanidine (AAAN) reversed this response. nNOS gene transfer also increased intracellular cGMP (47%) and decreased cAMP (29%). A PDE2 specific inhibitor Bay60-7557 reversed the reduction in cAMP caused by Ad.PRS-nNOS. These results suggest that neuronal NO modulates cGMP and PDE2 to regulate voltage gated intracellular Ca2+ transients in sympathetic neurons. Therefore, we propose this as a possible key step involved in NO decreasing cardiac sympathetic neurotransmission. © 2007 Elsevier Inc. All rights reserved. Keywords: Nitric oxide; Sympathetic neurons; Norepinephrine; cGMP; cAMP; PDE2; [Ca2+]i

1. Introduction Excessive sympathetic activation is well documented in patients with hypertension and heart failure with direct evidence of increased levels of norepinephrine (NE) [1,2] that may be related to oxidative stress impairing autonomic function [3]. Emerging evidence indicates the release of NE from cardiac sympathetic nerve terminals can be directly inhibited by endogenous or exogenous nitric oxide (NO) acting postganglionically, but pre-synaptically to the neuroeffector site of pacemaking [4]. We have recently demonstrated that noradrenergic neuronspecific gene transfer of neuronal nitric oxide synthase (nNOS) into the cardiac sympathetic nerve decreases NE release in nor⁎ Corresponding author. Department of Physiology, Anatomy and Genetics, Sherrington Building, University of Oxford, Parks Road, Oxford, OX1 3PT, UK. Tel.: +44 1865 272518; fax: +44 1865 282170. E-mail address: [email protected] (D.J. Paterson). 0022-2828/$ - see front matter © 2007 Elsevier Inc. All rights reserved. doi:10.1016/j.yjmcc.2007.09.005

motensive rats [5] and to a greater extent in hypertensive animals [6]. Spatial localization of nNOS in the cardiac neuronal axis appears to be important in conferring specificity of action of NO in the cardiovascular system [7]. nNOS is positioned in discrete cellular domains in the cardiac myocytes to regulate excitation contraction coupling [8] and Ca2+ handling [9]. Calcium signalling governs important processes in neurons, including membrane excitability and neurotransmitter release [10]. Calcium influx through N-type Ca2+ channels in sympathetic neurons initiates norepinephrine (NE) release from varicosities [11]. The signalling pathway responsible for the inhibitory effect of NO on sympathetic neurotransmission is not completely understood. This is partially due to the difficulties in differentiating the effects between the two branches of the autonomic nervous system due to their close proximity in effector regions and subsequent interactions. Given the multiple actions of NO on cardiac control, interpretation of a specific response after a

718

L. Wang et al. / Journal of Molecular and Cellular Cardiology 43 (2007) 717–725

global intervention on the NO system (systemic nNOS over expression or nNOS knockout) adds to the difficulties of making accurate conclusions. By using our recently developed noradrenergic neuron specific gene transfer technique [5], we are able to target nNOS into the cardiac sympathetic neuron only. This approach allows us to study the role of NO in sympathetic neurotransmission with minimum interference from other cell types. In this study, we used primary cultures of dissociated sympathetic neurons from stellate ganglia transduced with a noradrenergic neuron-specific adenoviral vector encoding nNOS. This method was employed to test the hypothesis that nNOS overexpression decreases NE release through elevated NO–cGMP signalling, thereby stimulating PDE2 mediated degradation of cellular cAMP leading to a subsequent inhibition of intracellular Ca2+ mediated exocytotic release of NE. 2. Materials and methods 2.1. Sympathetic neuron isolation and adenoviral vector transduction Cardiac sympathetic neurons were isolated from neonatal Sprague–Dawley (SD) rats as previously described [5]. Briefly, 8–12-day-old neonatal SD rats were humanely killed by an approved Home Office schedule 1 method: overdose of pentobarbital followed by exsanguination. Sympathetic stellate ganglia were dissected and digested using a combination of collagenase and trypsin. Dissociated neurons were purified by seeding on a collagen coated dish for 1 h to minimize the number of fibroblasts and Schwann cells in the culture. The supernatant containing mostly neurons was plated onto poly-Llysine/laminin substrate before transduction with an adenoviral vector encoding nNOS driven by a noradrenergic promoter [12]. An empty adenoviral vector was used as control for comparing the effect of viral transduction on the neurons. The cells were infected with 1 × 108 pfu of adenoviral vector per well (1.9 cm2/well, Nunc, Denmark). Replication-deficient adenoviral vectors encoding nNOS under control of the noradrenergic neuron-specific promoter (PRS×8) were generated as described previously [5,12]. 2.2. Measurement of free intracellular calcium concentration Intracellular free calcium concentration ([Ca2+]i) was determined in single neurons using Fura-2 fluorescence ratio imaging. Subconfluent sympathetic neurons grown on 6 mm cover slips were loaded with 5 μM Fura-2-AM at 37 °C for 30 min. The neurons were then transferred to a temperaturecontrolled (37 °C), gravity fed, perfusion chamber (volume: 100 μl); a flow rate of 2 ml/min permitted rapid solution exchange. Fura-2 was excited alternately at 340 nm and 380 nm at 0.5 Hz and the emitted fluorescence measured at 510 nm using a Nikon Diaphot microscope equipped with a Cairn Optoscan monochromator and photomultiplier tube. Data acquisition and analysis was performed using a CED 1401, and a PC running Spike 2 software (Cambridge Electronic Design, UK). Fluores-

cence excitation ratios were transformed into [Ca2+]i concentrations using the equation derived by Grynkiewicz et al. [13]. The Grynkiewicz equation for calibration of the Fura-2 ratio signal reads as follows: ½Ca2þ i ¼ Kd  ðSf 2 =Sb2 Þ  ðRRmin Þ=ðRmaxRÞ Kd = dissociation constant of Fura-2 (224 nM under standard conditions), Sf2 = fluorescence signal at 380 nm of Ca2+ unbound (free) form, Sb2 = fluorescence signal at 380 nm in Ca2+ bound form, R = actual measured Fura-2 ratio, Rmin = minimal Fura-2 ratio in zero [Ca2+], Rmax = maximum Fura-2 ratio at saturating [Ca2+]. The constants Rmin and Rmax were determined in situ using Ionomycin [14]. The calcium intensity image profile was obtained using an UltraView imaging system (Perkin Elmer, UK) housed on an inverted Olympus IX70 microscope [15,16]. Emitted fluorescence was collected through a dichroic mirror set, passed to an image intensifier and then to a Hamamatsu 1394 ORCA-ERA camera (Hamamatsu, Japan). Measurements of [Ca2+]i were obtained from cells bathed in HEPES buffered Tyrode solution. Cell depolarization was induced by 30 s exposure to a high K+ HEPES Tyrode containing 50 mM KCl (with equimolar reduction in NaCl). To examine the effect of nNOS inhibition, sympathetic neurons were treated with 5 μM of the nNOS specific inhibitor N-[(4S)-4-Amino-5-[(2aminoetyl](amino] pentyl]-N′-nitroguanidine (AAAN) (Sigma: A5727) [17], for 2–5 min before high K+ stimulation. 2.3. Immunohistochemistry Cultured primary neurons were fixed with 4% Paraformaldehyde and permeabilized with 0.1% Triton X100 and 1% BSA. Cells were then processed for immuno-reactivity with mouse anti tyrosine hydroxylase (TH), 1:200 (Sigma) and rabbit anti nNOS, 1:200 (Invitrogen) sequentially. Fixed cells were incubated with 10% normal horse serum, primary antibody and biotinylated secondary antibody. The immunofluorescent signals were detected by Texas Red Streptavitin (SA-5006, Vector Labs) for TH or Fluorescein Streptavitin (SA-5001, Vector Labs) for nNOS. Streptavidin/Biotin Blocking Kit (SP2002, Vector Labs) was used before applying TH staining and between TH and nNOS staining. This procedure ensured all endogenous biotin, biotin receptors, or non-specific streptavidin binding sites present in tissues were blocked prior to the addition of the labelled streptavidin reagent. 2.4. Measurement of nNOS activity nNOS activity in sympathetic neurons was analyzed by measuring biochemical conversion of L-arginine to L-citrulline by NOS. 3H L-arginine (Amersham) was added to cell homogenates with or without a nNOS specific inhibitor AAAN at a concentration of 10 μM. After 30 min incubation, the reactions were stopped with a HEPES buffer (pH 5.5) containing EDTA to inactivate the NOS. Converted L-citrulline was separated from positively charged unreacted arginine using a Dowex AG 50WX8 resin (8% crosslinking and mesh size 200–400) filled spin

L. Wang et al. / Journal of Molecular and Cellular Cardiology 43 (2007) 717–725

719

FLUOstar Fluorescence Microplate Reader (BMG Labtech). The standard curve of the assay was obtained by plotting relative light units (RLU) versus log cGMP concentration using 9-point serially diluted cGMP standards. RLU obtained from each assay allowed the determination of the amount of cGMP produced in neurons. 2.6. Measurement of intracellular cAMP levels cAMP concentration in neurons was measured using TRFRET technology (LANCE cAMP Assay, Perkin Elmer) [19,20]. About 10000 cultured sympathetic neurons were used for each assay according to manufacture's protocol. The fluorescent signal was read on a TRF detection instrument, Victor 3V (Perkin Elmer). The standard curve of the assay was obtained by plotting LANCE signal versus log cAMP concentration using 10-point serially diluted cAMP standards. Counts at 665 nM obtained from each assay allowed the determination of the amount of cAMP produced in neurons. 2.7. Statistical analysis Analysis was performed using the paired or un-paired Student's t-test as appropriate. For all experiments, data represent mean ± S.E.M, statistical significance was accepted at p b 0.05. 3. Results 3.1. Sympathetic neuron specific gene expression mediated by Ad.PRS-nNOS Sympathetic neuron specific nNOS gene expression was detected as early as 2 days post-transduction with Ad.PRS-

Fig. 1. Representative immunofluorescent images of extra-cardiac sympathetic neurons after gene transfer with a noradrenergic neuron-specific adenoviral vector Ad.PRS-nNOS. A, Anti-nNOS stain. B, Same cell preparation stained with sympathetic neuron marker anti-TH. C, Overlay of anti-nNOS (Fluorescein Streptavitin), anti-TH (Texas-red Streptavitin).and DAPI. Note that all nNOS expressing neurons were colocalized with TH positive neurons. Scale bar: 50 μm.

cup (Stratagene, 204500). The NOS activity was then quantified by counting the radioactivity in the outflow. The results are expressed in fmol citrulline/mg protein/min. 2.5. Measurement of intracellular cGMP levels For measurement of cGMP concentration in neurons, the enzyme fragment complementation (EFC) method (HitNunter cGMP Assay, DiscoverRx) was used [18]. About 10,000 cultured sympathetic neurons were used for each assay according to the manufacture's protocol. The luminescence was read in a

Fig. 2. nNOS activity in nNOS gene transferred sympathetic neurons as determined by measuring the conversion of [3H]-L-arginine to [3H]-L-citrulline. Six independent experiments with 50,000 neurons in each one were performed. Neurons treated with Ad.PRS-nNOS show a 64% increase in nNOS activity compared to the empty vector control group (⁎⁎p b 0.01, n = 6, unpaired t-test). The nNOS inhibitor AAAN reduced nNOS activity caused by nNOS over expression to a level similar to that seen in the empty vector group (⁎⁎⁎p b 0.001, n = 6, paired t-test) (NS: no significance, n = 6, unpaired t-test).

720

L. Wang et al. / Journal of Molecular and Cellular Cardiology 43 (2007) 717–725

nNOS, an adenoviral vector with a noradrenergic promoter encoding nNOS [5]. nNOS was clearly detected by an nNOS antibody in over 90% of sympathetic neurons. Immunohistochemistry in Ad.PRS-nNOS transduced cells showed that nNOS co-localized with the sympathetic neuron marker TH (Fig. 1). There was no detectable leakage of nNOS expression in other cell types since the other cells in the background stained with DAPI did not express nNOS. DAPI stains nuclei of all cell types including fibroblast and Schwann cells co-cultured with sympathetic neurons [6]. 3.2. nNOS gene transfer increased nNOS activity in sympathetic neurons Sympathetic neurons from stellate ganglia were transduced with a noradrenergic specific adenoviral vector Ad.PRS-nNOS or Ad.PRS-empty, a non-coding empty vector as a control. nNOS activity was measured in 6 independent experiments 2 days post-transduction (Fig. 2). nNOS activity in nNOS gene transferred group increased 64% compared to the empty vector (n = 6 in each group, p b 0.01, unpaired t-test). The nNOS specific inhibitor AAAN reduced nNOS activity caused by nNOS over expression to a level similar to that seen in the empty vector group (n = 6 in each group, p b 0.001, paired t-test).

Fig. 4. A, An example recording from Ad.PRS-nNOS or Ad.PRS-empty transduced single sympathetic neuron loaded with Fura-2-AM. K+ indicates brief (30 s) exposure to a 50 mM K+ containing Tyrode to depolarize the neuron and evoke voltage-gated calcium entry. B, The average amplitudes of voltagegated Ca2+ transients in sympathetic neurons treated with a nNOS vector or a control empty vector. Neurons transduced with nNOS had average 44% lower amplitudes of voltage-gated Ca2+ transients compared to neurons transduced with an empty control vector (⁎⁎⁎p b 0.001, n = 15, unpaired t-test).

There was no significant change in nNOS activity before and after AAAN treatment in the empty vector control group indicating that nNOS activity is generally low in sympathetic neurons without nNOS gene transfer. 3.3. nNOS gene transfer into sympathetic neurons increased intracellular cGMP level We observed that intracellular cGMP concentration was significantly increased in the sympathetic neuron culture 3 days after nNOS gene transfer (Fig. 3A). Ad.PRS-nNOS increased cGMP level by 47% compared to an empty control vector. Six independent gene transduction experiments followed by cGMP assays in sympathetic neurons were performed (n = 6, ⁎⁎⁎p b 0.001, unpaired t-test).

Fig. 3. A, Cellular cGMP levels in sympathetic neurons 3 days after transduction with Ad.PRS-nNOS was determined using EFC technology. Cellular cGMP levels in sympathetic neurons transduced with Ad.PRS-nNOS were significantly higher (47%) than that detected in neurons transduced with an empty control vector (⁎⁎⁎p b 0.001, n = 6, unpaired t-test). B, Cellular cAMP level in sympathetic neurons 3 days after transduction with Ad.PRS-nNOS was determined using TRFRET technology. Neurons transduced with Ad.PRS-nNOS detected a significant decrease in cAMP levels (29%) compared with the empty vector treatment (⁎⁎p b 0.01, n = 7, unpaired t-test). This attenuation was reversed by PDE2 specific inhibitor Bay 60-7550 (NS: no significance, n = 6, unpaired t-test).

3.4. nNOS gene transfer into sympathetic neurons decreased intracellular cAMP level Seven independent gene transduction experiments followed by cAMP assays in sympathetic neurons were performed. Neurons transduced with Ad.PRS-nNOS detected a significant decrease in cAMP levels (29%) compared with the empty vector treatment (Fig. 3B) (⁎⁎p b 0.01, n = 7, unpaired t-test). This attenuation was reversed by a PDE2 specific inhibitor Bay 607550 (NS: no significance, n = 6, unpaired t-test). Treatment

L. Wang et al. / Journal of Molecular and Cellular Cardiology 43 (2007) 717–725

721

intracellular calcium in single transduced neurons. Calcium imaging was performed 48 h post-transduction of either Ad. PRS-nNOS or Ad.PRS-empty vector transduced neurons. Brief exposure to a Tyrode solution containing 50 mM K+ evoked a rapid and robust increase in [Ca2+]i that quickly recovered upon return to normal Tyrode in both control vector and Ad.PRSnNOS transduced neurons. These [Ca2+]i transients were highly reproducible (see Fig. 4A). Quantitative analysis of [Ca2+]i responses to high K+ induced membrane depolarization revealed that the average [Ca2+]i attained in neurons treated with the vector encoding nNOS was 44% lower (Fig. 4B) than in neurons transduced with a control empty vector (⁎⁎⁎p b 0.001, unpaired t-test, 15 neurons from five independent transduction experiments in each group). A typical calcium intensity profile of a sympathetic neuron from stellate ganglia responding to 50 mM KCl challenge is shown in Fig. 5. These profiles display the changes in calcium distribution from neurons measured in the resting state (immediate prior to KCl challenge), depolarization state (upon KCl challenge) and repolarization state (recovering from KCl challenge). Round cells are sympathetic neurons. Cells marked blue indicate baseline Fura-2-AM loading. Cells marked red indicate Ca2+ influx in responding to 50 mM KCl induced

Fig. 5. A, Typical calcium intensity profile of a stellate ganglion sympathetic neuron after a challenge with 50 mM KCl. These profiles display the changes in Ca2+ distribution from that measured in A, resting state, (color blue indicating baseline Fura-2-AM loading) B, depolarization (color red indicating Ca2+ influx) and C, repolarization state. Round cells are sympathetic neurons.

with PDE2 specific inhibitor greatly increased cAMP levels in both empty and nNOS vector transduced neurons indicating that PDE2 constantly broke down cAMP in these neurons [21]. 3.5. Sympathetic neuron nNOS gene transfer reduced high K+ evoked elevation of [Ca2+]i To evaluate the effects of nNOS gene transfer on voltagegated Ca-signalling, we performed ratiometric recordings of

Fig. 6. A, An example recording from Ad.PRS-nNOS transduced sympathetic neurons responding to nNOS specific inhibitor AAAN (5 μM). K+ indicates 50 mM KCl stimulation for 30 s. B, Statistical data based on experiments on 11 neurons from 5 independent transduction experiments. The average increase of voltage-activated Ca2+ transients during application of nNOS inhibitor AAAN was 42% (⁎⁎⁎p b 0.001, n = 11, paired t-test).

722

L. Wang et al. / Journal of Molecular and Cellular Cardiology 43 (2007) 717–725

depolarization. During repolarization, sympathetic neurons change color back to blue. This suggests that cells are healthy enough to respond to stimuli and exert a normal depolarization/ repolarization pattern and Ca2+ homeostasis after 3 days of adenovirus mediated gene transfer. We observed no obvious difference in the response kinetics between neurons transduced with adenoviral vectors encoding nNOS and those transduced with a control empty vector.

functions at the cellular level by demonstrating that in cardiac sympathetic neurons, NO can induce cGMP synthesis, leading to activation of the cAMP-hydrolyzing activity of PDE2 and a consequent degradation of intracellular cAMP; this, in turn, is responsible for a reduction in [Ca2+]i and exocytotic release of NE.

3.6. nNOS specific inhibitor (AAAN) reversed reduction of [Ca2+]i caused by nNOS gene transfer

Evidence from central administration of NOS inhibitors into the paraventricular nucleus supports the concept of an inhibitory influence of NO on central sympathetic outflow [22,23]. In the peripheral nervous system, NOS inhibition also facilitates cardiac NE release [24] and the heart rate response to sympathetic nerve activation [4]. Bath application of NE in the presence of NOS inhibitors does not change the heart rate response suggesting the action of NO is working presynaptically to modulate neurotransmitter release [4,25]. In the spontaneous hypertensive rat (SHR), the NO–cGMP pathway in the peripheral nervous system is impaired at two sites; post gangionically but pre-synaptically to the pacemaking cells [6] and within the pacemaker itself [26]. Here hyperresponsiveness to β-adrenergic stimulation was associated with elevated cAMP levels and reduced cGMP levels and accompanied by an increased basal and NE-stimulated L-type calcium current (ICaL) in sinoatrial node cells [26]. nNOS gene transfer into the pacemaker region normalized the heart rate responsiveness to NE, increased cGMP, decreased cAMP and NE stimulated ICaL [26] suggesting a strong coupling between

To confirm that the effect of nNOS gene transfer on reduction of voltage-gated Ca2+ transients is through NO production, we tested whether a nNOS specific inhibitor can reverse this effect. AAAN, is a potent inhibitor of nNOS with the highest selectivity for nNOS over eNOS (greater than 2500-fold) and a 320-fold selectivity over iNOS [17]. We observed that application of 5 μM AAAN to the perfusion buffer during high K+ stimulation greatly enhanced the amplitude of voltage-gated Ca2+ transients in neurons transduced with Ad.PRS-nNOS (Fig. 6). The average increase of [Ca2+]i after applying AAAN is 42% (n = 11, ⁎⁎⁎p b 0.001, paired t-test). We did not find the effect of AAAN on control neurons treated with an empty vector, data not shown. 4. Discussion Our present study advances current understanding on the mechanism of NO–cGMP in the regulation of sympathetic

4.1. NO–cGMP signalling in modulating sympathetic activity

Fig. 7. Roles of neuronal NO in cardiac sympathetic neurotransmission. NO generated from nNOS signals through activation of sGC and subsequent elevation of cGMP production, which in turn activates cGMP stimulated PDE2 that decreases cAMP–PKA dependent phosphorylation of neuronal voltage-gated Ca2+ channels resulting in decreased exocytotic release of NE, which activates β1 adrenoceptors and stimulatory G proteins (Gs) of cardiac myocytes.

L. Wang et al. / Journal of Molecular and Cellular Cardiology 43 (2007) 717–725

NO–cGMP and intracellular Ca2+ handling in cardiac cells. Emerging evidence has also reported enhanced sympathetic activity with increased evoked NE release and impaired sGCcGMP signalling in the SHR [6]. Targeted gene transfer using a noradrenergic neuron specific adenoviral vector increased the bioavailability of nNOS and cGMP levels resulting in decreased evoked NE release in the SHR to levels seen in the normotensive control [6]. These results demonstrated that NO generated by nNOS can act post-ganglionically to inhibit sympathetic hyperactivity by decreasing neurotransmitter release. 4.2. Role of neuronal NO in cardiac sympathetic neurotransmission It is well established that NO activates sGC and stimulate cGMP production [27]. Upon NO activation, cGMP accumulates and interacts with PDE2, which attenuates the βadrenergic regulation of calcium currents [28]. The distinguishing feature of PDE2 is that it is allosterically stimulated by cGMP binding [29]. This binding stimulates PDE2 leading to increased hydrolysis of cAMP. Thus, elevation of cGMP activates PDE2 that in turn lowers cAMP [21]. cAMP activates cAMP-dependent protein kinase (PKA) which is directly involved in phosphorylation of several proteins critically involved in excitation–contraction and secretion coupling [30,31]. PKA dependent phosphorylation regulates voltage gated Ca2+ channels where these channels are responsible for Ca2+ influx mediated NE release from varicosities of sympathetic neurons [31]. Therefore interventions that target cGMP signalling cascades work in part via modulation of cyclic nucleotide dependent phosphodiesterase. A schematic overview illustrating the role of neuronal NO in cardiac sympathetic neurotransmission is shown in Fig. 7. The amplitude of any [Ca2+]i response to membrane depolarization typically reflects the concerted action of a number of processes, principally Ca2+-entry through voltage-gated channels, calcium induced calcium release (CICR) from the endoplasmic reticulum (ER) and calcium buffering by mitochondria. Theoretically, NO could modulate any or all of these processes. In sympathetic neurons however CICR only makes a relatively minor contribution to the rise in [Ca2+]i in response to a strong depolarization [32]. Moreover, if NO were to activate ryanodine receptors in these cells as occurs in skeletal and cardiac cells [33–35], we would expect to observe an augmented Ca2+ transient in nNOS transduced cells, but we did not. Mitochondria usually take up Ca2+ when [Ca2+]i exceeds approximately 500 nM and then release this Ca2+ back into the cytosol when [Ca2+]i falls below this threshold [36]. This process normally results in a second elevation, or plateau, in [Ca2+]i during recovery from depolarization such that the Ca2+ transient appears biphasic [36]. We did not observe such “tell tale” biphasic Ca-responses to depolarization in the adenovirus transduced sympathetic neurons used in this study (see Fig. 4). Moreover the elevation of [Ca2+]i in response to depolarization was relatively modest (i.e. 600 nM in control neurons) such as to barely exceed the threshold for mitochondrial Ca-uptake. We

723

therefore conclude that mitochondrial Ca-buffering also contributes little to the control of [Ca2+]i under the experimental conditions employed in this study. In consequence the only likely cause of the much reduced Ca-transient seen in neurons transduced with Ad.PRS-nNOS is NO and cGMP dependent inhibition of voltage-gated Ca-entry. We cannot however exclude the possibility that enhanced NO production may also influence other aspects of cellular Ca-homeostasis. 4.3. Yin-Yang hypothesis of cyclic nucleotides in regulation of cardiac neurotransmission Calcium and the cyclic nucleotides cGMP and cAMP are interrelated second messengers and essential for coupling excitation to secretion [31,37]. Cyclic nucleotide phosphodiesterases (PDEs) are not only important in keeping the homeostasis of cyclic nucleotides but also act to limit the movement of cAMP and cGMP, thereby establishing intracellular signalling microdomains [38]. The Yin-Yang hypothesis of cyclic nucleotides is based on opposite effects of cAMP and cGMP on cardiac function [27,39]. Combining the results from the present study with previous work suggests that NO–cGMP pathway can not only activate the cAMP pathway through cGMP-inhibited PDE (PDE3) in a cholinergic setting to increase ACh release [40], but can also antagonize the cAMP pathway through cGMP-stimulated PDE (PDE2) in sympathetic neurons to negatively modulate exocytotic release of NE. Recent evidence indicates that both cyclic nucleotides and PDEs are compartmentalized [38,41]. It is clear that cAMP and cGMP signaling in cardiac cells is regulated by different PDEs organized in a network, and that factors relating to the spatial organization and regulation of the network are of key importance in shaping the cyclic nucleotide response [21]. 4.4. Perspectives NO can have diverse effects within the nervous system where the spatial localization of the nNOS is thought to be important for conferring specificity of action. Our noradrenergic neuronspecific gene transfer approach has its great advantage on selectivity, thus adding to our knowledge of NO's site specific actions in relation to its target [5], although the technique does not allow us to target nNOS in the precise cellular microdomain. Nevertheless we observed that the nNOS immuno-reactivity in nNOS gene transferred sympathetic neuron is mostly in the cytoplasm but not in the nucleus, suggesting the expression pattern is in line with its native cellular location. In contrast, we found Ad.PRS-eGFP transduced cells expressed eGFP throughout the cell [5]. A primary culture of dissociated sympathetic neurons is an ideal model to investigate NO modulated transmitter release, since other sources of endogenous modulators are absent and autoinhibitory feedback is negligible [31]. We did not perform direct measurements of ICaN since the N type Ca channel responsible for NE release in sympathetic neurons has been extensively characterized [11,42,43]. We also did not directly

724

L. Wang et al. / Journal of Molecular and Cellular Cardiology 43 (2007) 717–725

measure NE release in this model. However, when all observations are taken together with the current data and recent findings, there is compelling evidence to suggest that increased signalling through the nNOS–NO–cGMP pathway can decrease NE release from cardiac sympathetic neurons due to modulation of [Ca2+]i. Acknowledgments We thank Professor Christopher Schofield for using his TRF detection instrument, and Dr. Jeremy Taylor for his technical advice on sympathetic neuron isolation. This work was funded by a grant from the Medical Research Council, UK. M. Henrich was funded by DFG (HE3678/1-1). References [1] Viquerat C, Daly P, Swedberg K, Evers C, Curran D. Endogenous catecholamine levels in chronic heart failure. Relation to the severity of hemodynamic abnormalities. Am J Med 1985;78(3):455–60. [2] Goldstein DS. Plasma catecholamines and essential hypertension. An analytical review; 1983. p. 86–99. [3] Danson EJ, Paterson DJ. Reactive oxygen species and autonomic regulation of cardiac excitability. J Cardiovasc Electrophysiol May 2006;17(Suppl 1):S104–12. [4] Choate JK, Paterson DJ. Nitric oxide inhibits the positive chronotropic and inotropic responses to sympathetic nerve stimulation in the isolated guinea-pig atria. J Auton Nerv Syst Feb 15 1999;75(2–3):100–8. [5] Wang L, Li D, Plested CP, Dawson T, Teschemacher AG, Paterson DJ. Noradrenergic neuron-specific overexpression of nNOS in cardiac sympathetic nerves decreases neurotransmission. J Mol Cell Cardiol Aug 2006;41(2):364–70. [6] Li D, Wang L, Lee CW, Dawson TA, Paterson DJ. Noradrenergic cell specific gene transfer with neuronal nitric oxide synthase reduces cardiac sympathetic neurotransmission in hypertensive rats. Hypertension Jul 2007;50(1):69–74. [7] Paton JF, Kasparov S, Paterson DJ. Nitric oxide and autonomic control of heart rate: a question of specificity. Trends Neurosci Dec 2002;25(12): 626–31. [8] Hare JM. Nitric oxide and excitation–contraction coupling. J Mol Cell Cardiol Jul 2003;35(7):719–29. [9] Sears CE, Bryant SM, Ashley EA, Lygate CA, Rakovic S, Wallis HL, et al. Cardiac neuronal nitric oxide synthase isoform regulates myocardial contraction and calcium handling. Circ Res Mar 21 2003;92(5):e52–9. [10] Clapham DE. Calcium signaling. 0092-8674, 80, 2; Jan 27 1995. p. 259–68. [11] Plummer MR, Rittenhouse A, Kanevsky M, Hess P. Neurotransmitter modulation of calcium channels in rat sympathetic neurons. J Neurosci Aug 1991;11(8):2339–48. [12] Hwang DY, Carlezon Jr WA, Isacson O, Kim KS. A high-efficiency synthetic promoter that drives transgene expression selectively in noradrenergic neurons. Hum Gene Ther Sep 20 2001;12(14):1731–40. [13] Grynkiewicz G, Poenie M, Tsien RY. A new generation of Ca2+ indicators with greatly improved fluorescence properties. J Biol Chem Mar 25 1985;260(6):3440–50. [14] Scanlon M, Williams DA, Fay FS. A Ca2+-insensitive form of fura-2 associated with polymorphonuclear leukocytes. Assessment and accurate Ca2+ measurement. J Biol Chem May 5 1987;262(13):6308–12. [15] Watson R, Jepson JE, Bermudez I, Alexander S, Hart Y, McKnight K, et al. Alpha7-acetylcholine receptor antibodies in two patients with Rasmussen encephalitis. Neurology Dec 13 2005;65(11):1802–4. [16] Raymond-Delpech V, Towers PR, Sattelle DB. Gene silencing of selected calcium-signalling molecules in a Drosophila cell line using doublestranded RNA interference. Cell Calcium Feb 2004;35(2):131–9. [17] Hah JM, Roman LJ, Martasek P, Silverman RB. Reduced amide bond peptidomimetics. (4S)-N-(4-amino-5-[aminoakyl]aminopentyl)-N′-nitro-

[18]

[19] [20] [21]

[22]

[23] [24]

[25]

[26]

[27]

[28] [29] [30] [31]

[32]

[33]

[34]

[35]

[36]

[37] [38]

[39]

guanidines, potent and highly selective inhibitors of neuronal nitric oxide synthase. J Med Chem Aug 2 2001;44(16):2667–70. Eglen RM. Enzyme fragment complementation: a flexible high throughput screening assay technology. ASSAY Drug Dev Technol 2002;1(1): 97–104. Williams C. cAMP detection methods in HTS: Selecting the best from the rest. Nat Rev Drug Discov 2004;3(2):125–35. Hemmila I. LANCE: homogeneous assay platform for HTS. J Biomol Screen 1999;4(6):303–8. Zaccolo M, Movsesian MA. cAMP and cGMP signaling cross-talk: role of phosphodiesterases and implications for cardiac pathophysiology; 2007. p. 1569–78. Li YF, Mayhan WG, Patel KP. Role of the paraventricular nucleus in renal excretory responses to acute volume expansion: role of nitric oxide. Am J Physiol Heart Circ Physiol Oct 2003;285(4):H1738–46. Patel KP, Li Y-F, Hirooka Y. Role of nitric oxide in central sympathetic outflow; 2001. p. 814–24. Schwarz P, Diem R, Dun NJ, Forstermann U. Endogenous and exogenous nitric oxide inhibits norepinephrine release from rat heart sympathetic nerves. Circ Res Oct 1 1995;77(4):841–8. Sears CE, Choate JK, Paterson DJ. Effect of nitric oxide synthase inhibition on the sympatho-vagal control of heart rate. J Auton Nerv Syst Aug 27 1998;73(1):63–73. Heaton DA, Lei M, Li D, Golding S, Dawson TA, Mohan RM, et al. Remodeling of the cardiac pacemaker L-type calcium current and its {beta}-adrenergic responsiveness in hypertension after neuronal NO synthase gene transfer; 2006. p. 443–52. Fischmeister R, Castro L, Abi-Gerges A, Rochais F, Vandecasteele G. Species- and tissue-dependent effects of NO and cyclic GMP on cardiac ion channels. Comp Biochem Physiol A Mol Integr Physiol 2005;142(2): 136–43. Lohmann SM, Fischmeister R, Walter U. Signal transduction by cGMP in heart. Basic Res Cardiol Nov–Dec 1991;86(6):503–14. Bender AT, Beavo JA. Cyclic nucleotide phosphodiesterases: molecular regulation to clinical use; 2006. p. 488–520. Bers DM. Cardiac excitation–contraction coupling. Nature Jan 10 2002;415(6868):198–205. Kubista H, Boehm S. Molecular mechanisms underlying the modulation of exocytotic noradrenaline release via presynaptic receptors. Pharmacol Ther Oct 2006;112(1):213–42. Thayer SA, Hirning LD, Miller RJ. The role of caffeine-sensitive calcium stores in the regulation of the intracellular free calcium concentration in rat sympathetic neurons in vitro. Mol Pharmacol Nov 1988;34(5): 664–73. Fischmeister R, Castro L, Abi-Gerges A, Rochais F, Vandecasteele G. Species- and tissue-dependent effects of NO and cyclic GMP on cardiac ion channels. Comp Biochem Physiol A Mol Integr Physiol Oct 2005;142(2): 136–43. Tocchetti CG, Wang W, Froehlich JP, Huke S, Aon MA, Wilson GM, et al. Nitroxyl improves cellular heart function by directly enhancing cardiac sarcoplasmic reticulum Ca2+ cycling. Circ Res Jan 5 2007;100(1):96–104. Sun J, Xu L, Eu JP, Stamler JS, Meissner G. Nitric oxide, NOC-12, and S-nitrosoglutathione modulate the skeletal muscle calcium release channel/ryanodine receptor by different mechanisms. An allosteric function for O2 in S-nitrosylation of the channel. J Biol Chem Mar 7 2003;278(10):8184–9. Werth JL, Thayer SA. Mitochondria buffer physiological calcium loads in cultured rat dorsal root ganglion neurons. J Neurosci Jan 1994;14(1): 348–56. Rasmussen H, Goodman DB. Relationships between calcium and cyclic nucleotides in cell activation. Physiol Rev Jul 1977;57(3):421–509. Fischmeister R, Castro LRV, Abi-Gerges A, Rochais F, Jurevicius J, Leroy J, et al. Compartmentation of cyclic nucleotide signaling in the heart: the role of cyclic nucleotide phosphodiesterases; 2006. p. 816–28. Goldberg N, Haddox M, Nicol S, Glass D, Sanford C, Kuehl F, et al. Biologic regulation through opposing influences of cyclic GMP and cyclic AMP: the Yin Yang hypothesis. Adv Cyclic Nucleotide Res 1975;5: 307–30.

L. Wang et al. / Journal of Molecular and Cellular Cardiology 43 (2007) 717–725 [40] Herring N, Paterson DJ. Nitric oxide–cGMP pathway facilitates acetylcholine release and bradycardia during vagal nerve stimulation in the guinea-pig in vitro. J Physiol Sep 1 2001;535(Pt 2):507–18. [41] Mongillo M, Tocchetti CG, Terrin A, Lissandron V, Cheung YF, Dostmann WR, et al. Compartmentalized phosphodiesterase-2 activity blunts betaadrenergic cardiac inotropy via an NO/cGMP-dependent pathway. Circ Res Feb 3 2006;98(2):226–34.

725

[42] Lin Z, Harris C, Lipscombe D. The molecular identity of Ca channel alpha 1-subunits expressed in rat sympathetic neurons. J Mol Neurosci Winter 1996;7(4):257–67. [43] Rittenhouse AR, Zigmond RE. Role of N- and L-type calcium channels in depolarization-induced activation of tyrosine hydroxylase and release of norepinephrine by sympathetic cell bodies and nerve terminals. J Neurobiol Aug 1999;40(2):137–48.