Angiotensin II type 2 receptor-coupled nitric oxide production modulates free radical availability and voltage-gated Ca²⁺ currents in NTS neurons

Abstract

The medial region of the nucleus tractus solitarius (mNTS) is a key brain stem site controlling cardiovascular function, wherein ANG II modulates neuronal L-type Ca²⁺ currents via activation of ANG II type 1 receptors (AT₁R) and production of reactive oxygen species (ROS). ANG II type 2 receptors (AT₂R) induce production of nitric oxide (NO), which may interact with ROS and modulate AT₁R signaling. We sought to determine whether AT₂R-mediated NO production occurs in mNTS neurons and, if so, to elucidate the NO source and the functional interaction with AT₁R-induced ROS or Ca²⁺ influx. Electron microscopic (EM) immunolabeling showed that AT₂R and neuronal NO synthase (nNOS) are coexpressed in neuronal somata and dendrites receiving synapses in the mNTS. In the presence of the AT₁R antagonist losartan, ANG II increased NO production in isolated mNTS neurons, an effect blocked by the AT₂R antagonist PD123319, but not the angiotensin (1–7) antagonist D-Ala. Studies in mNTS neurons of nNOS-null or endothelial NOS (eNOS)-null mice established nNOS as the source of NO. ANG II-induced ROS production was enhanced by PD123319, the NOS inhibitor N^G-nitro-l-arginine (LNNA), or in nNOS-null mice. Moreover, in the presence of losartan, ANG II reduced voltage-gated L-type Ca²⁺ current, an effect blocked by PD123319 or LNNA. We conclude that AT₂R are closely associated and functionally coupled with nNOS in mNTS neurons. The resulting NO production antagonizes AT₁R-mediated ROS and dampens L-type Ca²⁺ currents. The ensuing signaling changes in the NTS may counteract the deleterious effects of AT₁R on cardiovascular function.

the nucleus of the solitary tract (NTS), the major site of termination of vagal afferents arising from aortic baroreceptors and cardiorespiratory chemoreceptors, plays a critical role in normal cardiovascular regulation and in the neurohumoral dysfunction associated with hypertension (7, 13, 21, 31). The octapeptide ANG II is heavily expressed in the medial NTS (mNTS) and contributes to cardiovascular regulation and dysregulation, in large part, by activation of ANG II type 1 receptors (AT₁R) and type 2 receptors (AT₂R) (30, 33, 47). One of the signaling mechanisms by which AT₁R influence the mNTS neurons involves production of reactive oxygen species (ROS) by the enzyme NADPH oxidase (Nox) (18, 49, 50, 55). AT₁R-mediated ROS, in turn, enhance voltage-gated L-type Ca²⁺ currents in the mNTS neurons, which may contribute to the deleterious cardiovascular effects of ANG II (45, 49, 50). Much less is known, however, about the role of AT₂R in the mNTS. AT₂R are expressed in brain stem not only early in development, but also in adulthood (15, 17). In other autonomic neurons, AT₂R activation has been linked to production of nitric oxide (NO), which is thought to exert beneficial effects by counterbalancing some of the deleterious actions of AT₁R activation (10, 15). However, it is not known whether AT₂R activation in the mNTS neurons triggers NO production. Furthermore, the spatial relationships between AT₂R and NO synthase (NOS) have not been established.

ROS react avidly with NO to generate a wide variety of nitrated species (19, 32), and ROS levels can have a marked impact on NO availability and biological activity (22, 32). Therefore, it is likely that ROS produced by AT₁R activation modulate the bioavailability of NO produced by activation of AT₂R, and vice versa. However, the functional interactions between AT₁R-mediated ROS and AT₂R-mediated NO in the mNTS neurons remain unclear, particularly in reference to L-type Ca²⁺ currents, which have an important role in the function of the mNTS neurons (45, 49, 50) and are modulated by NO and ROS (5, 50).

In this study, we sought to address some of these outstanding questions concerning the role of AT₂R in the mNTS neurons. Using electron microscopic (EM) immunolabeling, we found that AT₂R have a close spatial relationship with neuronal nitric oxide synthase (nNOS) in the mNTS neurons. Consistent with this finding, NO imaging studies in isolated mNTS neurons identified nNOS as the predominant source of NO. Finally, patch-clamping studies demonstrated that NO produced by AT₂R activation dampens voltage-gated L-type Ca²⁺ currents induced by ANG II. The findings suggest a powerful modulatory role of AT₂R on Ca²⁺ channel activity in the mNTS neurons, which is likely to have a major influence on the NTS function.

MATERIALS AND METHODS

All experiments were performed in compliance with the guidelines of the Institutional Animal Care and Use Committee at Weill Cornell Medical College.

Materials.

Male Sprague-Dawley rats (2–3 wk old) were purchased from Charles River Laboratories. Mice were also used for EM study or to take advantage of knockout models to study the role of selected NOS isoforms and Nox2. C57BL/6J mice (wild-type controls), Nox2-null mice and nNOS-null mice (4–6 wk old) were obtained from an in-house colony (20, 50). Endothelial nitric oxide synthase (eNOS)-null mice were purchased from Jackson Laboratory. All null mice were on a C57BL6/J background. ANG II, PD123319, N^G-nitro-l-arginine (LNNA), and 2-phenyl-4,4,5,5-tetramethylimidazoline-1-oxyl-3-oxide (PTIO) were purchased from Sigma-Aldrich Chemical. D-Ala [A-779, (D-Ala7)-ANG I/II (1–7)] and ANG II-(1–7) were purchased from Bachem Bioscience Americas. Fluorescent dyes 4-amino-5-methylamino-2′,7′-difluorofluorescein diacetate (DAF-FM) and 6-carboxy-2′,7′-dichlorodihydro-fluorescein di(acetoxymethyl ester) (C-DCDHF-DA) were purchased from Invitrogen. Losartan was a gift from the DuPont Merck Pharmaceutical. The nNOS, glial fibrillary acidic protein (GFAP), and cluster of differentiation 3 (CD3) antibodies were purchased from BD Biosciences. The AT₂R Neun and Iba1 antibodies were purchased from Abcam.

Electron microscopy.

EM dual-immunolabeling has been described previously (50). Male wild-type mice were anesthetized with pentobarbital sodium, and their brains were fixed by perfusion with heparin-saline (1,000 U/ml) followed consecutively by 3.8% acrolein (Polysciences) in 2% paraformaldehyde in 0.1 mol/l phosphate buffer. Coronal sections of 40 μm were cut through the mNTS starting at 7.32 caudal to bregma and ending at 7.76 caudal to bregma using a vibratome (Leica Microsystems), and all of the sections included were at the level of the area postrema (34). The tissue sections were incubated for 48 h at 4°C in 0.1 mol/l Tris-buffered saline/0.1% BSA solution containing a 1:200 dilution of a mouse monoclonal nNOS antibody for immunoperoxidase labeling and a 1:1,000 dilution of a rabbit polyclonal AT₂R antiserum for immunogold-silver (9, 11). The nNOS antibody (cat. no. 610309) was against a sequence at the C-terminal (amino acids 1095–1289) of human nNOS. Western blot analysis showed that the purified antiserum recognizes a single band of ∼155 kDa, corresponding to the molecular weight of nNOS (44). The AT₂R antiserum was against a synthetic peptide sequence from the COOH-terminal domain (amino acids 349–363) of the rat AT₂R (Abcam, cat. no. ab19134). This antiserum specifically recognizes AT₂R in human cervical carcinoma (HeLa) cells transfected with AT₂R, but not AT₁R (11). After the primary antisera incubation, the sections were placed for 30 min each in biotinylated secondary horse anti-mouse IgG (1:400; Incstar) and avidin-biotin peroxidase complex (ABC Elite Kit, Vector Laboratories). The bound peroxidase was visualized by reaction of the sections for 5 min in 3,3′-diaminobenzidine (Sigma-Aldrich) and hydrogen peroxide. For immunogold labeling, the sections were placed for 2 h in a 1:50 dilution of donkey anti-rabbit IgG conjugated with 1-nm colloidal gold, and a silver intense EM kit was used for 6 min to enhance bound gold particles (Amersham Biosciences). The sections were postfixed in 2% osmium tetroxide and embedded in EM bed-812. Single ultrathin sections (70 nm) containing the mNTS at the level of the area postrema were cut using an EM UC6 ultratome (Leica Microsystems) and collected and dried on 400-mesh copper grids. Ultrathin sections were then counter-stained for ultrastructural analysis using a Philips CM10 transmission electron microscope.

Dissociation of the mNTS neurons.

As described in detail elsewhere (49, 50), rats were killed by CO₂, the brain stem was quickly removed and transferred to a chamber containing ice-cold sucrose-artificial cerebrospinal fluid (aCSF) oxygenated with a mixed gas containing 95% O₂ and 5% CO₂ (39, 40). Although hypercapnia and hypoxia at death could influence NTS chemoreceptive neurons, these changes are likely to be reversed by equilibration of the tissue in 95% O₂ and 5% CO₂. Coronal slices (300 μm in thickness) were obtained and incubated with 0.02% pronase and 0.02% thermolysin at 36°C in the oxygenated lactic acid (l)-aCSF. With the area postrema as a landmark, the mNTS region was punched, and its cells were mechanically dissociated. The mNTS neurons were identified based on their typical cellular morphology, e.g., small round or oval bipolar cells with long thin processes (49, 50).

ROS or NO detection.

Methods for ROS and NO detection have been described in detail elsewhere (20, 24–26, 39, 49) and are briefly summarized. NO and ROS were detected using DAF-FM (20, 24–26) and C-DCDHF-DA (39, 49), respectively. The mNTS neurons were incubated with C-DCDHF-DA (5 μmol/l) or DAF-FM (5 μmol/l) in oxygenated l-aCSF for 30 min, and then rinsed in control buffer. Time-resolved fluorescence (FITC filter) was measured at 30-s intervals with a Nikon diaphot 300 inverted microscope equipped with CCD digital camera (Princeton Instruments) and using IPLab software (Scanalytics) (49). Recordings were started after a stable baseline measurement was achieved, usually 10 min. For all experiments, concurrent vehicle recordings were performed. In some experiments, the cell type(s) producing NO was identified by immunofluorescence labeling with antibodies for the neuronal marker NeuN (1:500), the astrocytic marker GFAP (1:500), the microglial marker Iba1 (1:500), or the endothelial marker CD31 (1:500).

Whole-cell patch recording.

Voltage-gated Ca²⁺ currents were elicited from whole-cell patched NTS neurons. An Axopatch-200A patch-clamp amplifier was used with Cs⁺ electrode solution (49, 50). Using 2 mM Ca²⁺ as a charge carrier in oxygenated l-aCSF, we observed that inward Ca²⁺ currents were elicited by a 500 ms-depolarizing pulse from a holding potential of −60 mV to −10 mV using pClamp 8 (Axon Instruments). The amplitude of inward currents measured at the end of depolarizing pulses is considered as that of L-type Ca²⁺ currents (49, 50).

Data analysis.

The electron microscopy images were obtained exclusively at the surface (1–2 μm) of the flat-embedded tissue, where penetration of the immunoreagents is optimal. Profiles containing two or more gold particles were considered to be immunogold-labeled. Peroxidase-immunoreactive profiles had a granular, electron-dense precipitate greater than that seen in any other morphologically similar profiles in the surrounding neuropil. Illustrations were prepared by importing digital images directly into Adobe Photoshop. ROS or NO fluorescence data are expressed as the ratio Ft/Fo, where Ft is fluorescence following the application of ANG II in a given cell, and Fo is the baseline fluorescence of the same cell immediately before application of ANG II. There was no relationship between baseline fluorescence (Fo) and the fluorescence increase induced by ANG II (Ft) under the experimental conditions studied (20, 49). Data (mean ± SE) were statistically evaluated by the paired Student's t-test, or ANOVA, as appropriate. P values < 0.05 were considered significant.

RESULTS

AT₂R are coexpressed with nNOS in somata and dendrites of the mNTS.

Immunogold labeling for AT₂R was seen mainly within somatodendritic profiles, some of which also contained diffuse immunoperoxidase reaction product for nNOS in the mNTS of mouse (Fig. 1, A–C) or rat (data not shown). In the soma, AT₂R immunogold was detected away from the plasma membrane and near cytoplasmic organelles, including endomembrane of smooth endoplasmic reticulum and Golgi lamellae (Fig. 1A). In dendrites, AT₂R were localized to intracellular, as well as surface membranes. The surface labeling was evident on nonsynaptic, as well as synaptic portions of the plasmalemmal surface (Fig. 1, B and C). Occasionally, AT₂R gold particles were observed on the plasma membrane beneath a postsynaptic membrane specialization (Fig. 1C). Dendrites containing AT₂R and nNOS were among those that received synaptic contacts from axon terminals. Some of these terminals contained sparse AT₂R immunolabeling (Fig. 1B) and were typical of glutamatergic visceral vagal afferents, forming multiple dendritic contacts (3). These results provide ultrastructural evidence that AT₂R and nNOS are present in the same neurons.

Fig. 1.

Electron micrographs illustrating angiotensin type 2 receptor (AT₂R) and neuronal nitric oxide synthase (nNOS) coexpression in somato-dendritic profiles in the mouse mNTS. A: immunogold labeling for the AT₂R (red arrows) is seen in the cytoplasm near endomembranes (em) and the Golgi lamellae (Go) in a soma containing nNOS immunoreactivity (AT₂R/nNOS-s). The peroxidase nNOS labeling (green arrows) is diffusely distributed throughout the cytoplasm, with a slightly denser distribution near endomembranes (em) and mitochondria (m). B: large vagal-like axon terminal containing one immunogold particle (red arrows, AT₂R-t) on the plasma membrane is forming two synapses (yellow curved arrows) with a dually labeled dendrite (dd). The postsynaptic labeling for AT₂R (red arrows) is seen in the cytoplasm, away from the synaptic inputs. C: unlabeled terminals form synapses (yellow curved arrow) with dual-labeled dendrites (dd-1 and dd-2). One of the dendrites (dd-2) shows a single AT₂R gold particle (red arrows) located in the synaptic portion (yellow curved arrow) of the plasma membrane contacted by an unlabeled axon. An immunogold-labeled axon terminal (AT₂-t) is seen in the neuropil. Scale bars = 0.5 μm.

Activation of AT₂R increases NO production from nNOS.

We used DAF-FM imaging to determine whether the AT₂R activation results in NO production in the mNTS neurons. In the presence of control buffer (l-aCSF), ANG II (10–4,000 nmol/l) had little effect on the NO signal (Fig. 2B). However, in the presence of the AT₁R antagonist losartan (3 μmol/l), ANG II increased the NO signal markedly (Fig. 2B). To determine the cellular source of NO, we examined NO-dependent fluorescence in neurons, microglia, astrocytes, and endothelial cells identified using cell-specific markers. We found that ANG II-induced NO fluorescence (P < 0.01 vs. control) was colocalized only with neurons (n = 26) (Fig. 2C). The enhancement of NO-dependent fluorescence by coapplication of ANG II with losartan was blocked by the AT₂R antagonist PD123319 (30 μmol/l, P > 0.05 vs. ANG II alone, n = 9). NO-dependent fluorescence was also blocked by the cell-permeable NO scavenger PTIO (51) (50 μmol/l, P > 0.05 vs. ANG II, n = 14) or the NOS inhibitor LNNA (1 mmol/l, P > 0.05 vs. ANG II, N = 6), attesting to the specificity of the NO signal. The ANG-(1–7) receptor antagonist D-Ala (2 μmol/l) did not attenuate the increase in NO-dependent fluorescence induced by ANG II (P > 0.05 vs. losartan, n = 8) (Fig. 3A). Moreover, ANG-(1–7) (100 nM) did not increase NO-dependent fluorescence (P > 0.05 vs. losartan, n = 9) (Fig. 3B). We then used eNOS or nNOS-null mNTS neurons to examine the enzymatic source of NO. In the presence of losartan, ANG II (100 nmol/l) increased NO production in the mNTS neurons of wild type mice (P < 0.05 vs. control, n = 6) or eNOS-null mice (P < 0.05 vs. control, n = 9), but not in nNOS-null mice (P > 0.05 vs. control, n = 9) (Fig. 4). Therefore, nNOS is required for the NO production evoked by activation of AT₂R.

Fig. 2.

AT₂R activation potentiates nitric oxide (NO) production in rat medial nucleus tractus solitarius (mNTS) neurons. A: representative images showing that ANG II (100 nmol/l) increased NO-dependent fluorescence in mNTS neurons. Scale bars = 10 μm. B: with application of control buffer ANG II increases the NO signals (control: n = 11; ANG II: 10 nmol/l, n = 11; 30 nmol/l, n = 11; 100 nmol/l, n = 11; 4,000 nmol/l, n = 11). Pretreatment of the AT₁R antagonist losartan (Los, 3 μmol/l) potentiated the NO signals induced by ANG II [10 nmol/l, 30 nmol/l, 100 nmol/l, and 4,000 nmol/l for control (n = 6) and ANG II (n = 6)]. C: application of losartan (3 μM) and ANG II (100 nM) increases the NO signal in neurons (NeuN; n = 26) but not in microglia (lba1; n = 6), astrocytes (GFAP; n = 4), or endothelial cells (CD31; n = 5). *P < 0.05, **P < 0.01; paired Student's t-test or ANOVA was used.

Fig. 3.

AT₂R- and NOS-dependent NO signals in rat mNTS neurons. A: pretreatments with the AT₂R antagonist PD123312 (PD; 30 μmol/l, n = 9), the NO scavenger 2-phenyl-4,4,5,5-tetramethylimidazoline-1-oxyl-3-oxide (PTIO) (50 μmol/l; n = 14), or the NOS inhibitor N^G-nitro-l-arginine (LNNA) (1 mmol/l, n = 10) inhibited ANG II (100 nmol/l)-mediated potentiation of the NO signals (P > 0.05 vs. control). However, pretreatment with the angiotensin (1–7) receptor antagonist D-Ala (2 μmol/l, n = 8) did not affect the potentiation (P < 0.01 vs. control). B: ANG-(1–7) (100 nM), in the presence of losartan, does not affect the NO-dependent fluorescence of isolated mNTS neurons (n = 9). *P < 0.05, **P < 0.01; paired Student's t-test or ANOVA.

Fig. 4.

AT₂R-induced NO production is dependent on nNOS in mouse mNTS neurons. In the presence of losartan (3 μmol/l), ANG II (100 nmol/l), potentiated NO production in wild-type (WT) (n = 6) and eNOS-null mNTS neurons (n = 9). However, the ANG II (100 nmol/l)-induced increase in NO was not observed in nNOS-null neurons (n = 9). *P < 0.05; paired Student's t-test or ANOVA.

AT₂R-dependent NO production modulates AT₁R-dependent ROS production.

NO rapidly interacts with the ROS superoxide to produce peroxynitrite and other nitrated species (19, 32). Therefore, we sought to determine whether the NO produced by activation of AT₂R counteracts the ROS produced by activation of AT₁R. In agreement with previous results (49, 50), ANG II, at a concentration producing maximal effects (1 μmol/l), increased the ROS signal in rat mNTS neurons (P < 0.01 vs. control, n = 16). The increase was blocked by the ROS scavenger Mn(III) tetrakis(4-benzoic acid)porphyrin chloride (30 μmol/l, P > 0.01 vs. control, n = 10), attesting to the specificity of the ROS signal. The ROS increase was also suppressed by losartan (3 μmol/l, P > 0.05 vs. control, n = 11), but it was facilitated by PD123319 (30 μmol/l, P < 0.05 vs. ANG II alone, n = 18), or LNNA (1 mmol/l, P < 0.05 vs. ANG II, n = 7) (Fig. 5A). ANG II (1 μmol/l) also increased ROS production in wild-type mouse mNTS neurons (P < 0.01 vs. control, n = 7). The ROS increase was blocked by losartan (3 μmol/l, P > 0.05 vs. control, n = 8) and was not observed in Nox2-null mice (P < 0.05 vs. control, n = 9), confirming Nox2 as the source of ROS. In contrast, the ROS production was enhanced in nNOS-null mNTS neurons (P < 0.05 vs. WT, n = 9) (Fig. 5B). These results indicate that the NO produced by AT₂R activation suppresses the Nox2-derived ROS evoked by AT₁R receptor activation.

Fig. 5.

AT₂R-induced NO production antagonizes AT₁R-induced reactive oxygen species (ROS) in the mNTS neurons. A: in rat mNTS neurons, ANG II (1 μmol/l) increases the ROS signal (n = 16), an effect blocked by the ROS scavenger Mn(III) tetrakis(4-benzoic acid)porphyrin chloride (30 μmol/l; n = 10) or the AT₁R antagonist losartan (Los, 3 μmol/l; n = 11), but potentiated by the AT₂R antagonist PD123319 (PD, 30 μmol/l; n = 18) or the NOS inhibitor LNNA (1 mmol/l; n = 7). B: in mouse wild-type mNTS neurons, ANG II (1 μmol/l) potentiated the ROS signal (n = 7), an effect blocked by losartan (Los, 3 μmol/l; n = 8). However, ANG II (1 μmol/l) reduced the ROS signal in Nox2-/- mNTS neurons (n = 10). ANG II (1 μmol/l) enhanced the ROS signal in nNOS-/- mNTS neurons (n = 9), an effect completely blocked by pretreatment with losartan (Los; 3 μmol/l; n = 6). *P < 0.05; **P < 0.01; paired Student's t-test or ANOVA.

AT₂R-mediated NO production down-regulates L-type Ca²⁺ currents.

ANG II increases voltage-gated L-type Ca²⁺ currents in the mNTS neurons via AT₁R activation and Nox2-derived ROS (49, 50). Ca²⁺ current enhancement in the mNTS plays a role in the mechanisms by which ANG II induces the neurohumoral dysfunction underlying hypertension (45). To gain insight into the functional implications of the interplay between NO and ROS in the mNTS neurons, we examined the effect of the AT₂R-mediated NO production on the Ca²⁺ currents evoked by ANG II. Without losartan, ANG II (100 nmol/l) increased the amplitude of L-type Ca²⁺ current (P < 0.01 vs. control, n = 5) (Fig. 6, A and F). However, in the presence of losartan, ANG II decreased the amplitude of L-type Ca²⁺ current (P < 0.01 vs. control; n = 7) (Fig. 6, B and F), an effect blocked by PD123319 (30 μmol/l) (Fig. 6, C and F) or LNNA (1 mmol/l) (Fig. 6, D and F). We then used the NO donor SNAP to determine whether NO itself was able to downregulate ANG II-induced Ca²⁺ currents. Consistent with this hypothesis, SNAP (2 μmol/l) decreased the Ca²⁺ current (P < 0.01 vs. control; n = 6) (Fig. 6, E and F). These findings indicate that NO produced by AT₂R activation downregulates the ANG II-induced Ca²⁺ current.

Fig. 6.

Activation of AT₂R inhibits L-type Ca²⁺ currents via NO in rat mNTS neurons. A: with control buffer, ANG II (100 nmol/l) increased the L-type Ca²⁺ current (n = 5). B: however, ANG II (100 nmol/l) inhibited the L-type Ca²⁺ current in the presence of losartan (Los; 3 μmol/l; n = 7). This inhibitory effect was blocked by coapplication of PD (30 μmol/l; n = 5) (C) or LNNA (1 mmol/l; n = 6) (D). E: NO donor SNAP (2 μmol/l; n = 6) mimics the inhibitory effect of ANG II plus losartan on L-type Ca²⁺ currents. F: histogram shows percentage changes in the amplitude of L-type Ca²⁺ currents affected by ANG II (100 nmol/l) with or without losartan, PD123319, LNNA, and SNAP. **P < 0.01; paired Student's t-test or ANOVA.

DISCUSSION

We have shown that AT₂R are coexpressed with nNOS in somata and dendrites of mNTS neurons, and NO production by these neurons is markedly enhanced by AT₁R inhibition. Such NO increase is prevented by an AT₂R antagonist, suggesting the involvement of AT₂R in the signaling pathway. Results in rats were generally comparable to those obtained in mice. Using NOS-null mice, we demonstrated that nNOS, not eNOS, is the source of NO. Interestingly, experiments with ANG-(1–7) and D-Ala suggested that, at variance with findings in other systems (54), ANG-(1–7) or its receptor does not affect NO production in mNTS. Because of the close interaction between NO and ROS, we examined the effect of AT₂R activation on ROS production. We found that AT₂R inhibition or NOS inhibition enhances the ROS production evoked by ANG II. Similarly, ROS production was enhanced in nNOS-null mNTS neurons, an effect attenuated by AT₁R inhibition. To determine the functional implications of the AT₂R-dependent NO production, we tested its effect on voltage-gated L-type Ca²⁺ currents, an important signaling mechanism in the mNTS neurons (45, 49, 50). AT₁R inhibition suppressed the Ca²⁺ currents evoked by ANG II, an effect dependent on AT₂R and NO. Taken together, these findings provide evidence of a reciprocal interaction between AT₂R-dependent NO production and AT₁R-dependent ROS production. The resulting balance between NO and ROS has an impact on voltage-gated Ca²⁺ channels, which are critically involved in the functional output of the mNTS neurons (22).

The present study sought to compare the findings from rats and mice, including genetically modified mice, to integrate all information derived from these two different models to generate a synergistic platform for understanding the roles of AT₂R in nNOS- vs. eNOS-derived NO. Although it is not unlikely that the mechanisms underlying the AT₂R-mediated NO release in the rat NTS is different from that in the mouse NTS, the similarities between rat and mice, from physiology to genomics, allows us to explain present findings obtained from the rat NTS neurons.

Somatodendritic profiles in the mNTS contain both AT₂R and nNOS.

The primary location of AT₂R in the mNTS was in postsynaptic somata and dendrites, many of which also contained nNOS. The colocalization of AT₂R and nNOS provides ultrastructural support for the functional interaction between ANG II and NO. In dual-labeled somatodendritic profiles, AT₂R were present in the cytoplasmic compartment, associated with membranous structures involved in synthesis and intracellular trafficking of proteins to and from the surface (35, 43). Plasmalemmal labeling of AT₂R was occasionally seen in small- and medium-sized dendrites receiving synaptic input from one or more axon terminals. The presence of AT₂R at the synapses in the nNOS-containing neuron suggests that activation of postsynaptic AT₂R may result in generation of NO that presynaptically modulates the release of inhibitory and/or excitatory transmitters.

Age-dependent expression of AT₂R.

In contrast with early observations of mRNA of AT₂R expression at very high levels in the fetus and its rapid regression to low levels in adults (37), recent publications using Western blot analysis shows that the brain stem from rat and mouse fetuses and neonates exhibit a significantly lower AT₂R protein expression compared with adult rats (15, 17, 53). These papers also show that 2–3-wk-old brain stem expresses considerable amounts of AT₂R protein. Therefore, AT₂R are also present in the adult brain stem, highlighting their physiological relevance (15, 17).

AT₂R induce nNOS-dependent NO production.

NO is an important signaling molecule in the NTS, but its enzymatic source(s) and relationships to AT₂R activation have not been clearly defined (17). Some studies have provided evidence that NTS neurons are endowed with AT₂R (27, 29, 36). In addition, other reports show that NTS neurons express nNOS but not eNOS (14, 28), whereas others have suggested that eNOS in the NTS plays a role in the regulation of blood pressure in spontaneously hypertensive rats (33, 48, 52). The identification of the enzymatic sources of NO is complicated by the difficulties in selectively inhibiting specific NOS isoforms, especially eNOS (4). To overcome this problem, in the present study, we used nNOS and eNOS-null mice and identified nNOS as the sole source of NO in the mNTS. However, on the basis of our observations in dissociated NTS cells in which AT₁R were blocked, we cannot support or contradict the findings in vivo demonstrating the involvement of eNOS-derived NO in the dampening of baroreceptor sensitivity induced by exogenous ANG II (33, 48, 52). Our immunoelectron microscopy studies demonstrated a colocalization of nNOS and AT₂R, providing ultrastructural evidence supporting a functional interaction between these proteins. However, the signaling pathways by which AT₂R activate nNOS remain to be established. Activation of bradykinin receptors has been implicated in AT₂R-induced NO production (42, 46), a finding not confirmed by other studies (1). Therefore, the mechanisms linking AT₂R activation to NO production remain to be elucidated.

Interaction between AT₁R-derived ROS and AT₂R-derived NO.

NO reacts avidly with ROS to generate a nitrated species, some of which contribute to its physiological and pathological effects (8, 22, 23). Our findings suggest that when neurons are exposed to ANG II, AT₂R-induced NO is scavenged by AT₁R-derived ROS. Thus, robust increases in NO are observed only if ROS production is suppressed by AT₁R inhibition. Thus, robust increases in NO are observed only if ROS production is suppressed by AT₁R inhibition. Conversely, the ROS production induced by AT₁R activation is dampened by the NO produced by AT₂R activation. In the normal state, ROS production predominates over NO production, and ROS can be detected, even in the absence of AT₂R inhibition. However, as demonstrated here, AT₂R inhibition reduces NO and potentiates AT₁R-dependent ROS generation, suggesting that loss of NO enhances ROS availability. This effect is unlikely to be related to NO-induced Nox2 inhibition because NO does not suppress, but promotes, Nox2-dependent ROS formation (20). Therefore, ROS and NO can reciprocally influence their bioavalability and biological activities in the mNTS neurons. Previous studies have demonstrated that AT₂R facilitate vasodilation through NOS/NO-mediated signal after chronic AT₁R antagonism and that AT₁R antagonists increase both AT₂R expression and cGMP (6, 40). These effects may contribute, in part, to the beneficial actions of AT₁R blockers in the treatment of hypertension. Thus, our finding that blockade of AT₁R in NTS neurons may significantly enhance the AT₂R-mediated NO release in vivo may have clinically relevant implications.

Effects on voltage-gated Ca²⁺ channels.

Voltage-gated L-type Ca²⁺ channels play an important role in neurotransmission and long-term potentiation, and, as such, are critical to neuronal function in the NTS. AT₂R have been proposed to modulate neuronal Ca²⁺ channel activity by NO through two different mechanisms, cGMP and S-nitrosylation (2). The present results suggest that AT₂R activation suppresses Ca²⁺ currents through NO production, whereas AT₁R receptor activation enhances Ca²⁺ currents via Nox2-derived ROS (49, 50). The reduction in Ca²⁺ currents by NO shown here is in agreement with similar findings in other neuronal types (5). The mechanisms of the effects of NO on L-type Ca²⁺ channels have not been elucidated in full, but they may include activation of soluble guanylyl cyclase or S-nitrosylation of Ca²⁺-handling proteins (2).

Perspectives and Significance

We investigated the role of AT₂R in the regulation of NO and ROS homeostasis in mNTS neurons. We found that AT₂R and nNOS are colocalized at postsynaptic sites and that AT₂R activation enhances NO production via activation of nNOS, while reducing the availability of ROS generated by AT₁R activation via Nox2. These effects on NO and ROS have an impact on L-type Ca²⁺ currents since AT₂R-mediated NO attenuates the Ca²⁺ currents triggered by AT₁R activation. These results suggest that AT₂R-mediated NO production may offset the deleterious effects of AT₁R-mediated ROS in the mNTS, possibly, by suppressing L-type Ca²⁺ currents. Our findings also unveil a previously unrecognized interaction between AT₁R and AT₂R in the NTS. Such interaction is relevant to clinical conditions, such as hypertension and cardiac failure, in which the balance between central AT₁R and AT₂R is altered (16). Furthermore, AT₂R-mediated NO may help explain the beneficial effects of AT₁R inhibitors for the treatment of hypertension (12, 41). Finally, the functional significance of the interaction between mNTS AT₁R, and AT₂R in the regulation of sympathetic outflow and blood pressure remains to be defined (12). Further studies using in vitro and in vivo approaches will be required to address these outstanding questions.

GRANTS

This work was supported by the National Institutes of Health Grants HL-096571 and MH-40342.

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the authors.