Mitochondrial Respiratory Enzyme Complexes in Rostral Ventrolateral Medulla as Cellular Targets of Nitric Oxide and Superoxide Interaction in the Antagonism of Antihypertensive Action of eNOS Transgene

Nitric oxide (NO) is a gaseous molecule that plays an active role in central cardiovascular regulation (Zanzinger, 1999). One brain region in which NO is of pivotal importance in this regulatory machinery is the rostral ventrolateral medulla (RVLM) (Chan et al., 2001; Mayorov, 2005), which provides tonic sympathetic outflow for the maintenance of vasomotor tone (Dampney, 1994). Accumulating evidence suggests that NO deficiency in the RVLM may underlie sympathetic over-excitation (Kagiyama et al., 1998; Chan et al., 2003; Tandai-Hiruma et al., 2005) or impaired baroreflex regulation (Kishi et al., 2003) during hypertension. Accordingly, overproduction of NO by gene transduction of endothelial nitric-oxide synthase (eNOS) in the RVLM promotes sympathoinhibition and depressor response (Kishi et al., 2002; Tai et al., 2005) and improves the impaired baroreflex control of heart rate (Kishi et al., 2003) in hypertensive rats. The antihypertensive effects of eNOS gene transduction, however, fail to provide a sustained normalization of arterial pressure (Kishi et al., 2002; Tai et al., 2005) and are followed by rebound hypertension (Tai et al., 2005). The cellular mechanisms for those intriguing observations are essentially unknown.

NO is also a free radical messenger that at physiological concentrations binds to mitochondrial respiratory enzyme complex IV (MRC-IV) to donate single electrons to the oxygen molecule and promote the release of superoxide anion () (Cleeter et al., 1994). NO may additionally be converted to a number of reactive nitrogen species, including S-nitrosothiols, to inhibit MRC-I (Brown and Borutaite, 2004), leading to production of (Borutaite and Brown, 2006). Moreover, NO increases levels by down-regulating mitochondrial uncoupling protein-2 (UCP-2) (Merial et a l., 2000), a member of the mitochondrial anion carrier family that dissipates the proton gradient in the inner mitochondrial membrane, leading to a decrease in ROS production (Korshunov et al., 1997; Andrews et al., 2005). Under normal conditions, the mitochondrion-derived is rapidly degraded by superoxide dismutases (SODs) to hydrogen peroxide, which is subsequently converted to water by catalase and glutathione peroxidase. However, under pathological conditions such as hypertension, when the antioxidant defense machinery is defective (Kishi et al., 2004; Tai et al., 2005), NO may cause an accumulation of by inhibiting mitochondrial respiratory enzyme activity and down-regulating UCP-2. in the RVLM in turn evokes pressor response via activation of sympathetic vasomotor activity (Kimura et al., 2005a; Chan et al., 2006).

We demonstrated previously (Chan et al., 2006) that the molecular synthesis and enzyme activity of mitochondrial manganese SOD (SOD2) are reduced in the RVLM of spontaneously hypertensive rats (SHR). It follows that an interaction between NO and in the RVLM, using MRC as the cellular target, may contribute to the unsustained depressor and rebound hypertensive effects after eNOS gene transduction in SHR. This hypothesis was validated in the present study using adenoviral vectors to overexpress eNOS (AdeNOS) and/or SOD2 (AdSOD2) in the RVLM of SHR or normotensive Wistar-Kyoto (WKY) rats.

Materials and Methods

All experimental procedures were carried out in compliance with the guidelines of our institutional animal care committee and were in accordance with the Guide for the Care and Use of Laboratory Animals as adopted and promulgated by the U.S. National Institutes of Health.

Animals. Male adult (10-12-week-old) SHR (215-240 g, n = 336) or WKY rats (210-245 g, n = 286), purchased from the Experimental Animal Center of the National Applied Research Laboratories, Taiwan, were used. They were housed in an animal room under temperature control (24 ± 0.5°C) and 12-h light-dark (8:00 AM-8:00 PM) cycle. Standard laboratory rat chow (PMI Nutrition International, Brentwood, MO) and tap water were available ad libitum. All animals were allowed to acclimatize for at least 7 days before experimental manipulations.

Construction and Purification of Adenovirus Vectors. Adenoviral vectors encoding eNOS or SOD2 gene were constructed and purified as detailed previously (Tai et al., 2005; Chan et al., 2006). Based on 4,6-diamidino-2-phenylindole staining of adenoviral construct carrying green fluorescence protein (GFP), 80 to 90% of cultured GH₃ neuronal cells were infected with minimal toxicity at a concentration of 100 multiplicity of infection (Supplemental Fig. S1). In addition, the efficacy of AdeNOS to enhance NO production or AdSOD2 to inhibit accumulation has been demonstrated (Tai et al., 2005).

In Vivo Gene Transduction of eNOS and/or SOD2 in the RVLM. For in vivo gene delivery, bilateral microinjection of AdeNOS or AdSOD2, adenoviral vectors encoding βgal (Adβgal), or green fluorescence protein (AdGFP) (Tai et al., 2005; Chan et al., 2006) was carried out stereotaxically and sequentially into the RVLM of SHR or WKY rats that were anesthetized with sodium pentobarbital (50 mg/kg i.p.). An adenoviral suspension containing 0.5 × 10⁹ plaqueforming units (pfu)/50 nl was delivered to each injection site over 10 to 15 min using a glass micropipette. For cotransduction, a mixture of AdeNOS and AdSOD2 containing 1 × 10⁹ pfu/50 nl was administered. A total of eight injections (four on each side) were made at two rostrocaudal levels at stereotaxic coordinates of 4.5 or 5 mm posterior to lambda, 1.8 to 2.1 mm lateral to the midline, and 8.0 to 8.5 mm below the dorsal surface of cerebellum. These coordinates cover the confines of RVLM within which sympathetic premotor neurons reside (Ross et al., 1984). The wound was closed in layers, and animals were allowed to recover in individual cages with free access to food and water.

Microinjection of Test Agents into the RVLM. Various pharmacological agents were used to establish a causal role for MRC, NO, and in the cellular events and cardiovascular phenotypes induced by gene transduction with AdeNOS and/or AdSOD2 in the RVLM. Bilateral microinjection of test agents was carried out stereotaxically and sequentially into the RVLM, using a glass micropipette and at an injection volume of 50 nl (Tai et al., 2005; Chan et al., 2006, 2007a,b). The stereotaxic coordinates of RVLM used were 4.6 to 4.8 mm posterior to lambda, 1.8 to 2.1 mm lateral to the midline, and 8.0 to 8.5 mm below the dorsal surface of cerebellum. Test agents employed included the MRC-I inhibitor rotenone (Sigma-Aldrich, St. Louis, MO); a mobile mitochondrial electron carrier, coenzyme Q₁₀ (CoQ₁₀; Sigma-Aldrich); a nonselective NOS inhibitor, N^ω-nitro-l-arginine-methyl ester (l-NAME, Sigma-Aldrich); a potent eNOS inhibitor, N⁵-(1-iminoethyl)-l-ornithine (l-NIO; Tocris Cookson, Bristol, UK); an active peroxynitrite decomposition catalyst, 5,10,15,20-tetrakis-(N-methyl-4′-pyridyl)-porphyrinato iron (III) (FeTMPyP; Calbiochem, San Diego, CA); a NADPH oxidase inhibitor, apocynin (Calbiochem); or a peroxisome proliferator-activated receptor γ activator, rosiglitazone (Cayman Chemical, Ann Arbor, MI). Microinjection of artificial cerebrospinal fluid (aCSF) or dimethyl sulfoxide served as the vehicle and volume control. The composition of aCSF was 117 mM NaCl, 25 mM NaHCO₃, 11 mM glucose, 4.7 mM KCl, 2.5 mM CaCl₂, 1.2 mM MgCl₂, and 1.2 mM NaH₂PO₄. The doses used were established in pilot studies or were the same as in our previous studies (Tai et al., 2005; Chan et al., 2001, 2005a, 2006), in which individual test agent was used for the same purpose as in the present study.

Measurement of Systemic Arterial Pressure and Heart Rate. Changes in cardiovascular parameters were evaluated after gene transduction of eNOS and/or SOD2 into the RVLM. Mean systemic arterial pressure (MSAP) and heart rate (HR) were recorded in conscious rats using a radiotelemetry system (UA-10; Data Sciences International, Minneapolis, MN) (Chan et al., 2007b). Under sodium pentobarbital (50 mg/kg i.p.) anesthesia, a flexible catheter attached to a telemetry transmitter was inserted into the rat abdominal aorta just below the renal arteries. The catheter was secured in place with tissue glue, and the transmitter was attached to the abdominal muscle and remained in the abdominal cavity for the duration of the experiment. The skin was closed using nonabsorbable suture (3-0). Rats were allowed to recover from surgery for 3 days in individual cages positioned over an RLA-3000 radiotelemetry receiver before gene transduction. The MSAP and HR were recorded continuously for 30 min every day between 13:00 and 15:00, and telemetry data were collected for a maximum of 49 days. In a separate series of experiments, the short-term effect of rotenone, CoQ₁₀, or apocynin on MSAP or HR was determined on day 14 or 28 after gene transduction in animals that were maintained under propofol anesthesia (30 mg/kg/h i.v.).

Collection of Tissue Samples from Ventrolateral Medulla. To correlate biochemical changes in the RVLM with cardiovascular phenotypes after gene transduction, we routinely collected ventrolateral medullary tissues at various intervals after microinjection of AdeNOS and/or AdSOD2 into the RVLM. Rats were killed with an overdose of sodium pentobarbital (100 mg/kg i.p.) and perfused intracardially with 150 ml of warm (37°C) saline containing heparin (100 U/ml). The brain was rapidly removed and placed on dry ice, blocked in the coronal plane, and sectioned at 300-μm thickness in a cryostat. Both sides of the ventrolateral medulla covering the RVLM were collected by micropunches made with a stainless steel bore (1 mm i.d.) (Chan et al., 2005a, 2006, 2007a; Tai et al., 2005). Medullary tissues collected from animals under anesthesia but without treatment served as the basal control.

Isolation of Mitochondrial Fractions. To investigate whether MRC is the cellular target for NO and interaction after gene transduction of eNOS and/or SOD2 in the RVLM or to measure UCP-2 expression, mitochondria were isolated from the ventrolateral medulla by discontinuous Percoll gradient centrifugation according to procedures described previously (Chan et al., 2005b, 2007a). This procedure, which was performed at 4°C and completed within 2 h after tissue collection, yields 10 to 15% of the total mitochondria and enriches the mitochondrial fraction by at least 10-fold compared with tissue homogenates (Kantrow et al., 1997). The purity of the mitochondrial-rich fraction was verified by the expression of MRC-IV (Naithani et al., 2003). Total protein in the mitochondrial or cytosolic extracts was estimated by the method of Bradford with a protein assay kit (Bio-Rad Laboratories, Hercules, CA).

Assays For Activity of Mitochondrial Respiratory Enzymes. Activities of MRC were measured immediately after mitochondrial isolation, according to procedures reported previously (Wu et al., 2007) and using a thermostatically regulated ThermoSpectronic spectrophotometer (Fisher Scientific UK, Loughborough, UK). At least duplicate determination was carried out for each tissue sample in all enzyme activity assays, and the activity was expressed in nmol/mg protein/min. For NADH cytochrome c reductase (NCCR; marker for coupling capacity between complexes I and III) activity, the mitochondrial fraction (20 μg of protein) was incubated in a mixture containing 50 mM K₂HPO₄ buffer, pH 7.4, 1.5 mM KCN, 1 mM β-NADH, and 20 μM rotenone at 37°C for 2 min.

After the addition of 0.1 mM cytochrome c, the reduction of oxidized cytochrome c was measured as the difference in the presence or absence of rotenone at 550 nm for 3 min at 37°C. The molar extinction coefficient of cytochrome c at 550 nm is 18,500 M/cm.

Determination of succinate cytochrome c reductase (SCCR; marker for coupling capacity between complexes II and III) activity in the mitochondrial fraction (30 μg) was performed in 40 mM K₂HPO₄ buffer, pH 7.4, and 1.5 mM KCN, supplemented with 20 mM succinate. After a 5-min equilibration at 37°C, 50 μM cytochrome c was added, and the reaction was monitored at 550 nm for 3 min at 37°C.

The cytochrome c oxidase (CCO, marker enzyme for MRC-IV) activity was measured by oxidation of the reduced cytochrome c. The activity is defined as the first-order rate constant and is calculated from known concentration of ferrocytochrome c and the enzyme amount in the assay mixture. Mitochondrial fraction (30 μg) and 500 mM K₂HPO₄, pH 7.4, was preincubated at 30°C for 5 min. A 45 μM ferrocytochrome c was added to start the reaction monitored at 550 nm for 3min at 30°C. The background rate was measured after the addition of 1.0 μMK₃Fe(CN)₆.

Measurement of ATP Concentration. We measured tissue levels of ATP to decipher whether AdeNOS- and/or AdSOD2-induced changes in levels and cardiovascular responses are consequential to bioenergetic failure. Samples collected from ventrolateral medulla of rats that received the same experimental manipulations were pooled and homogenized in a protein extraction solution (Pierce, Rockford, IL). The supernatant after centrifugation at 10,000g for 10 min was subject to determination of ATP concentration, using an ATP bioluminescence assay (Roche Diagnostics, Mannheim, Germany) (Wu et al., 2007). Light emitted from a luciferase-mediated reaction and measured by a tube luminometer (Berthold Detection Systems, Pforzheim, Germany) was used to calculate the measured values.

Measurement of by Lucigenin-Enhanced Chemiluminescence. production was determined by lucigenin-enhanced chemiluminescence according to previously described and validated methods (Chan et al., 2005a, 2006; Tai et al., 2005). production was calculated and expressed in micromoles per minute per milligram of protein. Specificity for was determined by adding SOD (350 U/ml) to the incubation medium.

Western Blot Analysis. Western blot analysis of total protein extracted from ventrolateral medulla was carried out as detailed previously (Tai et al., 2005; Chan et al., 2006) using a mouse monoclonal antiserum against nitrotyrosine (1:1000; Calbiochem), a rabbit polyclonal antiserum against UCP-2 (1:2000; Calbiochem), α-tubulin (1:5000; Sigma), or a mitochondrial inner membrane marker, prohibitin (1:1000; NeoMarkers, Fremont, CA). This was followed by incubation with horseradish peroxidase-conjugated goat anti-rabbit IgG (Jackson Immunoresearch Laboratories, West Grove, PA) or sheep anti-mouse IgG (GE Healthcare, Chalfont St. Giles, Buckinghamshire, UK). Specific antibody-antigen complex was detected using an enhanced chemiluminescence Western Blot detection system (PerkinElmer Life Sciences, Waltham, MA). The amount of detected nitrotyrosine was quantified by Photo-Print Plus software (ETS Vilber-Lourmat, Marne La Vallee, France) and was expressed as the ratio to α-tubulin protein.

Isolation of RNA and Reverse Transcription Real-Time Polymerase Chain Reaction. A fundamental premise in the present study is that delivery of the transgenes to the RVLM effectively induced the targeted gene products. To detect copper/zinc SOD1, SOD2, extracellular SOD3, nNOS, iNOS, or eNOS mRNA expression, total RNA from samples of ventrolateral medulla was isolated with TRIzol reagent (Invitrogen, Carlsbad, CA) according to the manufacturer's protocol. All RNA isolated was quantified by spectrophotometry, and the optical density 260/280-nm ratio was determined. Reverse transcription (RT) reaction was performed using a SuperScript Preamplification System (Invitrogen) for the first-strand cDNA synthesis. Real-time polymerase chain reaction (PCR) analysis was performed by amplification of cDNA using a LightCycler instrument (Roche Diagnostics) (Chan et al., 2005a, 2007b). The primers used in real-time PCR amplification (Table 1) were designed using Roche LightCycler probe design software 2.0 based on sequence information from the NCBI database, and were synthesized by Genemed Biotechnologies (San Francisco, CA). PCR reaction for each sample was carried out in duplicate for all the cDNA and for the glyceraldehyde-3-phosphate dehydrogenase control. Fluorescence signals from the amplified products were quantitatively assessed using the LightCycler software program (version 3.5). Second derivative maximum mode was chosen with baseline adjustment set in the arithmetic mode (Chan et al., 2005a, 2007b).

Double Immunofluorescence Staining and Laser Confocal Microscopy. We assessed the distribution of eNOS and SOD2 protein in the RVLM after gene transduction by double immunofluorescence staining. In brief, free-floating 25-μm sections of the medulla oblongata were incubated with a mouse monoclonal antiserum against eNOS (1:1000; BD Biosciences, San Jose, CA), together with a rabbit polyclonal antiserum against SOD2 (1:1000; StressGene, Victoria, BC, Canada). The sections were subsequently incubated concurrently with a goat anti-mouse IgG conjugated with Alexa Fluor 564 (Invitrogen) for eNOS and a goat anti-rabbit IgG conjugated with Alexa Fluor 488 (Invitrogen) for SOD2. Viewed under a Fluorview FV300 laser scanning confocal microscope (Olympus, Tokyo, Japan), immunoreactivity for eNOS exhibited red fluorescence, and that for SOD2 exhibited green fluorescence. The occurrence of yellow fluorescence on merged images indicated coexpression of eNOS and SOD transgenes.

Histology. With the exception of animals used for biochemical or molecular analyses, the brain stem was removed from animals after they were killed by an overdose of sodium pentobarbital (100 mg/kg i.v.), and fixed in 30% sucrose in 10% formaldehyde-saline solution for ≥72 h. Frozen 25-μm sections of the medulla oblongata were stained with 1% Neural red for histological verification of the location of microinjection sites. The expression of GFP in the medulla oblongata was detected by confocal microscopy.

Statistical Analysis. All values are expressed as mean ± S.E. One- or two-way analysis of variance with repeated measures was used to assess group means, as appropriate, to be followed by the Scheffé multiple-range test for post hoc assessment of individual means. A value of p < 0.05 was taken to indicate statistical significance.

Results

Effects of Gene Transduction of eNOS and/or SOD2 in RVLM on Mitochondrial Respiratory Enzyme Activity and ATP Content in Ventrolateral Medulla. Our first series of experiments determined whether WKY rats or SHR exhibited differential bioenergetic responses to gene transduction of eNOS or SOD2 in the RVLM. Microinjection of AdeNOS to the RVLM of WKY rats, alone or together with AdSOD2, inhibited CCO activity but not NCCR or SCCR activity in ventrolateral medulla when determined on day 7 or 14 after injection (Fig. 1A). On the other hand, NCCR activity was depressed 7, 14, or 21 days after transduction of eNOS in the RVLM of SHR, and CCO activity was reduced on days 7 or 14 (Fig. 1B). The suppression of NCCR, but not CCO, activity by the eNOS transgene was significantly reversed on cotransduction with AdSOD2. Microinjection of AdSOD2 alone, on the other hand, did not exert discernible effect on the activity of those mitochondrial respiratory enzymes in SHR or WKY rats. On day 14 after eNOS gene transduction, the decrease of NCCR or CCO activity in SHR was significantly reversed in a dose-related manner by bilateral microinjection into the RVLM of an eNOS inhibitor, l-NIO (10, 50, or 100 pmol) (Fig. 2A). Given at an equimolar concentration, the nonselective NOS inhibitor l-NAME (100 pmol) promoted comparable antagonism on the reduced NCCR (80 ± 5 versus 74 ± 6%, n = 5) or CCO (74 ± 4 versus 70 ± 6%, n = 5) activity after eNOS gene transduction. Microinjection of the peroxynitrite decomposition catalyst FeTMPyP (10 or 50 pmol), on the other hand, attenuated only the reduction in NCCR, but not CCO, activity (Fig. 2B). The reduction in CCO activity on day 14 after AdeNOS transfection in WKY rats was also prevented by l-NAME or l-NIO but not by FeTMPyP (data not shown).

Compared with aCSF control, bilateral microinjection of AdeNOS and/or AdSOD2 into the RVLM of SHR or WKY rats induced minimal alterations of ATP content in ventrolateral medulla, detected on day 14 or 21 after gene transduction (Table 2). Control injection of Adβgal was also minimally effective.

Temporal Effects of Gene Transduction of eNOS and/or SOD2 in RVLM on Superoxide Levels in Ventrolateral Medulla. Our second series of experiments examined whether gene transduction of eNOS or SOD2 also elicited disparate effects on the differential levels of in the RVLM of the two strains of rats. As observed in our previous studies (Tai et al., 2005; Chan et al., 2006), the basal level of in ventrolateral medulla was higher in SHR than WKY rats. Measured on day 14 or 21 after gene delivery of AdeNOS, there were further increases of this level in SHR, whereas AdSOD2 elicited a decrease (Fig. 3). On the other hand, neither transgene affected the level in ventrolateral medulla of WKY rats. The AdeNOS-induced increase in level in the SHR was significantly attenuated by cotransduction of AdSOD2 (Fig. 3). Similar response was also observed in the SHR after bilateral microinjection into the RVLM of the nonselective NOS inhibitor l-NAME (100 pmol) or a selective eNOS inhibitor, l-NIO (100 pmol) (Fig. 4A). The tissue level of in the RVLM of SHR, but not WKY rats (data not shown), was also elevated on day 28 or 35 after eNOS transduction (Fig. 4B). Microinjected into the bilateral RVLM of a mobile mitochondrial electron carrier on day 28 or 35 after the gene transduction of eNOS, CoQ₁₀ (2.5 nmol), but not the NADPH oxidase inhibitor apocynin (2.5 nmol), significantly attenuated the induced accumulation of level in the RVLM of SHR. Treatment with l-NIO (0.43 ± 0.12 μmol/min/mg protein, n = 5), l-NAME (0.35 ± 0.09 μmol/min/mg protein, n = 5), CoQ₁₀ (0.35 ± 0.11 μmol/min/mg protein, n = 5), or apocynin (0.41 ± 0.09 μmol/min/mg protein, n = 5) exerted no discernible effect on level in the RVLM of WKY (0.36 ± 0.13 μmol/min/mg protein, n = 6), which was not altered on day 28 after gene delivery of AdeNOS (compare Figure 3A).

Temporal Effects of Gene Transduction of eNOS in RVLM on Peroxynitrite Production in Ventrolateral Medulla. The observation that the peroxynitrite decomposition catalyst FeTMPyP blunted the AdeNOS-induced reduction in mitochondrial NCCR activity and the augmentation in level effected by the transgene suggest that the eNOS-derived NO may interact with to form peroxynitrite in the RVLM of SHR. Our third series of experiments evaluated this possibility. Compared with WKY rats, the expression of 3-nitrotyrosine, an experimental index of peroxynitrite (Reiter et al., 2000), in ventrolateral medulla was significantly increased 7, 14, or 21 days after gene transduction of eNOS in the RVLM of SHR (Fig. 5A). Measured on day 14 or 21, the AdeNOS-induced increase in 3-nitrotyrosine in SHR was significantly attenuated by cotransduction of AdSOD2 (Fig. 5B).

Temporal Effects of Gene Transduction of eNOS or SOD2 in RVLM on MSAP and HR. The cardiovascular effects of manipulations of NO and level in the RVLM by gene transduction were investigated in our fourth series of experiments. Consistent with our previous report (Tai et al., 2005), application of AdeNOS (0.5 × 10⁹ pfu) to the RVLM promoted a significant decrease in MSAP or HR in SHR or WKY rats that lasted 14 days after gene transduction (Fig. 6). The AdeNOS-induced hypotension in SHR was followed by a rebound hypertension that became significant between days 28 and 35 after injection. Transduction of AdSOD2 (0.5 × 10⁹ pfu) resulted in significant hypotension and bradycardia in SHR, but not WKY rats, that lasted 11 days. Comicroinjection of AdeNOS and AdSOD2 (final titer, 0.5 × 10⁹ pfu) into the bilateral RVLM significantly augmented the amplitude and prolonged the duration (21 days) of the induced hypotension and bradycardia in SHR but not WKY rats. In particular, the maximal decrease in MSAP after cotransduction of AdeNOS and AdSOD2 (59.6 ± 2.9 mm Hg, n = 10) was significantly greater than the mathematical sum of that promoted individually by AdeNOS (31.5 ± 2.3 mm Hg, n = 10) or AdSOD2 (15.1 ± 2.4 mm Hg, n = 10). Furthermore, the rebound hypertension observed in AdeNOS-treated SHR was absent on delivery of AdeNOS and AdSOD2. Transduction of Adβgal, on the other hand, had no discernible effect on MSAP or HR in either strain of rat. Cotransduction of AdeNOS and AdSOD2 to regions adjacent to the confine of RVLM (e.g., lateral reticular nucleus, ventromedial medulla, areas lateral to the nucleus ambiguus) also did not affect the baseline MSAP or HR in SHR or WKY rats (data not shown).

Effect of Mitochondrial Respiratory Complex I on Depressor Response to Gene Transduction of eNOS in RVLM of WKY Rats. Results from our first four series of experiments suggest that depression of MRC-I activity and increase in production in the RVLM are associated with an antagonism of the vasodepressor effect of the eNOS transgene in the RVLM of SHR. Our fifth series of experiments was performed to ascertain that such a relationship is causal with the use of a loss-of-function approach by depressing the activity of MRC-I in the RVLM of normotensive WKY rats, which was not affected by eNOS transgene (Fig. 1A). On day 14 after AdeNOS transduction in the RVLM of WKY rats, bilateral microinjection of the MRC-I inhibitor rotenone (500 pmol) into the RVLM significantly reversed the depressor and bradycardic responses induced by the eNOS transgene during the 120-min observation period (Fig. 7). Rotenone (500 pmol) also evoked a pressor response in WKY rats that received Adβgal gene transduction. The rotenone-evoked pressor response was significantly reversed by bilateral comicroinjection into the RVLM of the SOD mimetic tempol (50 nmol; +9.6 ± 1.7 versus +2.1 ± 1.1 mm Hg, n = 5), in the Adβgal-transduced WKY rats. In a separate series of experiments, bilateral microinjection into the RVLM of rotenone resulted in a significant increase in tissue level of (rotenone, +76 ± 6%, n = 5) in ventrolateral medulla of WKY rats on day 14 after eNOS gene transduction.

Effect of NADPH Oxidase Inhibitor or Protection of the Mitochondrial Respiratory Enzyme Activity on Rebound Hypertension after eNOS Gene Transduction in RVLM of SHR. Our sixth series of experiments further deciphered the contribution of generated by the MRC or NADPH oxidase to the rebound hypertension after eNOS gene transduction in the SHR. On day 28 (Fig. 8) or 35 (data not shown) after AdeNOS gene transduction in SHR, microinjection of a mobile mitochondrial electron carrier CoQ₁₀, (5 nmol), but not the NADPH oxidase inhibitor apocynin (5 nmol), into the bilateral RVLM significantly alleviated the rebound hypertension during the 120-min observation period. No rebound hypertension in WKY (123.6 ± 3.8 mm Hg, n = 5) rats was detected on day 28 after eNOS gene transduction; CoQ₁₀ (118.3 ± 4.5 mm Hg, n = 5) or apocynin (121.6 ± 3.9 mm Hg, n = 5) treatment also did not significantly affect MSAP. At the end of a 120-min observation period, the tissue level of in ventrolateral medulla of AdeNOS-transduced SHR was significantly lower in CoQ₁₀-treated rats (0.66 ± 0.12 μmol/min/mg protein, n = 5) than apocynin-treated (0.92 ± 0.15 μmol/min/mg protein, n = 5).

Temporal Effects of eNOS Gene Transduction in RVLM on Mitochondrial UCP-2 Expression in Ventrolateral Medulla. The observation that rebound hypertension after gene transduction in the RVLM of SHR occurred after waning of eNOS expression prompted us to search for the mechanism that underlies this lag period. One possibility is that NO may suppress mitochondrial UCP-2 (Merial et al., 2000), leading to elevated tissue levels of that elicited rebound hypertension. Our seventh series of experiments evaluated this possibility. Bilateral microinjection into the RVLM of AdeNOS significantly decreased UCP-2 protein expression in the mitochondrial fraction of samples from ventrolateral medulla of SHR or WKY rats (Fig. 9). Such a down-regulation of mitochondrial UCP-2 expression became significantly different from sham control on day 7, 14, or 21 after gene transduction of eNOS in the RVLM and was not affected by cotransduction of AdSOD2 into the RVLM of SHR or WKY rats.

Effect of Transcriptional Up-Regulation of UCP in the RVLM on the Elevated Superoxide Levels in Ventrolateral Medulla after Gene Transduction of eNOS in SHR. Our eighth series of experiments further established a causal role for UCP down-regulation and the induced increase in level in the ventrolateral medulla. On day 21 or 28 after eNOS gene transduction in the RVLM of SHR, the increase in tissue level of in ventrolateral medulla was reversed in a dose-related manner by bilateral microinjection into the RVLM of rosiglitazone (1 or 5 nmol), an activator of the peroxisome proliferator-activated receptor that acts as a transcriptional factor for UCP-2 induction (Nolte et al., 1998), delivered 24 h before measurement of (Fig. 10). The level in ventrolateral medulla of WKY rats, which was not altered after eNOS gene transduction (0.40 ± 0.12 μmol/min/mg protein, n = 6), was not affected by rosiglitazone (5 nmol).

Temporal Expression of eNOS or SOD2 mRNA in Ventrolateral Medulla after Microinjection of AdeNOS and/or AdSOD2 into RVLM. A fundamental premise in the present study is that delivery of the transgenes to the RVLM effectively induced the targeted gene products. Our ninth series of experiments ascertained this premise. Consistent with our previous findings on protein expression (Tai et al., 2005; Chan et al., 2006), real-time RT-PCR analysis revealed that, given individually (Fig. 11) or together (data not shown), bilateral microinjection into the RVLM of AdeNOS and/or AdSOD2 resulted in significant and comparable upregulation of eNOS and SOD2 mRNA expression in ventrolateral medulla of SHR or WKY rats that lasted, respectively, 14 or 11 days. In contrast, there were no parallel changes in nNOS, iNOS, SOD1, or SOD3 mRNA (data not shown). Microinjection of Adβgal also did not affect basal eNOS or SOD2 mRNA expression in the RVLM (Fig. 11).

Cellular Expression of eNOS and SOD2 after Microinjection of AdeNOS and AdSOD2 into RVLM. Another critical issue is that gene delivery is effective at the cellular level and is restricted to the RVLM. Our final series of experiments addressed this issue. Figure 12, A and B, showed the presence of eNOS or SOD2 immunoreactivity in RVLM cells of SHR 11 days after cotransduction of AdeNOS and AdSOD2 in this medullary site. It is noteworthy that against the clearly defined nucleus, eNOS or SOD2 expression was confined to the cytoplasm. Furthermore, merged images (Fig. 12C) demonstrated that all eNOS-positive cells in the RVLM were immunoreactive to SOD2. Based on the manifestation of green fluorescence from GFP, Fig. 12D illustrates that microinjection of AdGFP resulted in a distribution that was restricted to the anatomical confines of the RVLM. Similar results were observed in WKY rats that received cotransduction of AdeNOS and AdSOD2 in the RVLM (data not shown).

Discussion

Overexpression of NO in the RVLM by eNOS gene transduction was proposed for treatment of neurogenic hypertension (Zanzinger, 1999) because NO deficiency in the RVLM contributes to the neural mechanisms of hypertension in SHR (Kagiyama et al., 1998; Chan et al., 2003; Tandai-Hiruma et al., 2005). The eNOS transgene, however, fails to provide sustained depressor effect (Kishi et al., 2002; Tai et al., 2005) and is followed by rebound hypertension (Tai et al., 2005). The present study presented novel observations to suggest that an interaction between NO and in the RVLM, using MRC-I in the mitochondrial respiratory chain as the cellular target via formation of peroxynitrite, contributes to those cardiovascular outcomes after eNOS gene transduction in SHR.

Our results indicate that MRC-I is the main target for the interaction between NO and in the RVLM during neurogenic hypertension. We found that gene transduction of AdeNOS in SHR, but not WKY rats, significantly inhibited MRC-I activity, as indicated by a suppression of NCCR (marker for coupling capacity between complexes I and III), but not SCCR (marker for coupling capacity between complexes II and III) activity. The observations that l-NIO or l-NAME reversed the inhibition of NCCR activity by AdeNOS indicated that those cellular processes are dependent on eNOS-derived NO production. The manifestation of comparable degree of antagonism by equimolar concentration of l-NIO or l-NAME further imply minimal contributions from nNOS- and iNOS-derived NO to the inhibition of NCCR or CCO activity after eNOS gene transduction. We also found that microinjection of AdeNOS into the RVLM selectively up-regulated eNOS, but not nNOS or iNOS mRNA. These results are consistent with previous reports on protein expression of the NOS isoforms after eNOS gene transduction (Kishi et al., 2002, 2003; Tai et al., 2005).

It is noteworthy that the suppression of MRC-I activity in the RVLM of SHR after eNOS gene transduction was also significantly prevented in the short term by a peroxynitrite decomposition catalyst or in the long term by cotransduction of AdSOD2. We reported previously (Chan et al., 2006) that a reduction in protein expression and enzyme activity of SOD2 underlies the augmented basal level of in the RVLM of SHR. Furthermore, MRC-I is a main source of mitochondrial (Kudin et al., 2004) and is selectively sensitive to peroxynitrite (Murray et al., 2003). It is thus conceivable that a reaction between the augmented NO production and the enhanced tissue level of after gene transduction of eNOS leads to the formation of peroxynitrite in the RVLM, which when coupled with SOD2 deficiency (Chan et al., 2006) results in suppression of MRC-I activity in ventrolateral medulla of SHR. The temporal increase in expression of nitrotyrosine, the product of peroxynitrite-meditated nitration of tyrosine residue (Reiter et al., 2000), in the ventrolateral medulla of AdeNOS-treated SHR that paralleled those of enhanced tissue level of or depression of MRC-1 activity, but not WKY rats, provides further credence to this notion. Moreover, it is also likely that an action of peroxynitrite on MRC-1 in the RVLM of SHR may also contribute to the sustained augmentation of level in the RVLM of SHR after AdeNOS transduction.

Our results also indicate that MRC-IV is another target for eNOS transgene in the RVLM. Transfection of eNOS depressed CCO (marker enzyme for MRC-IV) activity in the RVLM of both SHR and WKY rats, an action that was antagonized by l-NAME or l-NIO, but not FeTMPyP or cotransfected AdSOD2. This suggests that eNOS-derived NO exerts a direct depressant effect on MRC-IV activity. In this regard, NO binds to the Fe²⁺ center of cytochrome α₃ and the Cu²⁺ B center of CCO to elicit a reversible inhibition on MRC-IV activity (Stevens et al., 1979; Mason et al., 2006). It is intriguing that despite the induced MRC-IV inhibition by the eNOS transgene there was only minimal alteration in level in ventrolateral medulla of WKY rats. This observation implies the presence of an effective mitochondrial SOD2 system in the RVLM of WKY rats for degradation of the accumulated after gene transduction. Alternatively, it suggests that MRC-IV plays a minor role in mitochondrial formation and that MRC-I is the major -producing sites in brain mitochondria (Kudin et al., 2004).

Mitochondrial dysfunction is associated with the pathogenesis of several cardiovascular disorders, including hypertension (Kimura et al., 2005b; Callera et al., 2006). Together with our previous findings (Chan et al., 2006), we propose that there are at least two mechanisms via which mitochondrial dysfunction in the RVLM of SHR may contribute to neurogenic hypertension. The first mechanism involves a reduction in protein expression and enzyme activity of mitochondrial SOD2 in the RVLM (Chan et al., 2006), resulting in accumulation of that induces sympathetic overexcitation and pressor effect (Kishi et al., 2002; Kimura et al., 2005a; Chan et al., 2006). Mitochondria isolated from sod2-null mice exhibit higher levels of mitochondrial than the wild type, of which is attenuated by a SOD mimetic (Morten et al., 2006). The present results further showed that gene transduction of SOD2 to the RVLM resulted in depressor responses in SHR. The second mechanism engages counteraction of the sympathoinhibitory and depressor effect of NO (Zanzinger et al., 1995; Zanzinger, 1999) by mitochondria-derived in the RVLM. We observed that cotransduction of AdSOD2 in the RVLM of SHR augmented both the magnitude and duration of hypotensive response induced by overexpression of eNOS, together with a significant reversal of AdeNOS-induced production to levels that are comparable with those in WKY rats. Moreover, the MRC-I inhibitor attenuated the depressor response to eNOS overexpression in the RVLM of normotensive WKY rats. It is noteworthy that all these cellular events and hemodynamic changes occurred with minimal alteration in cellular ATP contents. As such, it is unlikely that the AdeNOS- and/or AdSOD2-induced changes in levels and cardiovascular responses are consequential to bioenergetic failure.

Enhancement of NO production by gene transduction of eNOS to the RVLM (Kishi et al., 2004) has been demonstrated to promote sympathoinhibitory and depressor responses in hypertensive rats (Kishi et al., 2002; Tai et al., 2005), although the normalized arterial pressure is followed by rebound hypertension. We reported previously (Tai et al., 2005) that this rebound phenomenon is attributed to generation. The present study extends this finding to reveal that CoQ₁₀, a mobile mitochondrial electron carrier, but not apocynin, a NADPH oxidase inhibitor, significantly reversed the augmented level in the RVLM, alongside a suppression of rebound hypertension seen after eNOS gene transduction. These observations suggest that cellular generated by the mitochondrial MRC, but not NADPH oxidase, is responsible for the AdeNOS-promoted rebound hypertension. We noted a trend of reduction in tissue level in the RVLM by CoQ₁₀ in animals treated with Adβgal. These findings pointed to a crucial engagement of mitochondrial electron transport chain in tonic regulation of cellular ROS homeostasis.

We recognize that the up-regulated eNOS mRNA expression in ventrolateral medulla has returned to baseline 21 days after AdeNOS gene transduction, the time point at which rebound hypertension began. One candidate that we identified to link transduction of AdeNOS gene with this rebound hypertension is mitochondrial UCP-2, which diminishes mitochondrial production (Nègre-Salvayre et al., 1997). We found a closely correlated temporal sequence of up-regulation of eNOS mRNA or protein expression (Tai et al., 2005), down-regulation of mitochondrial UCP-2 or elevated production in ventrolateral medulla, and manifestation of rebound hypertension after AdeNOS transduction in the RVLM of SHR. A causal role for UCP-2 down-regulation in the augmentation of production was established when transcriptional up-regulation of UCP-2 with the peroxisome proliferator-activated receptor γ activator rosiglitazone abrogated the elevated during which rebound hypertension took place. It is intriguing to note that whereas mitochondrial UCP-2 was down-regulated in the ventrolateral medulla of SHR and WKY rats after the eNOS gene transduction, elevated tissue level of and rebound hypertension were only observed in SHR. Given the abundance of SOD2 in the mitochondria of WKY rats, we reasoned that the increased production of because of the AdeNOS-induced UCP-2 down-regulation may be rapidly dismutated. This antioxidant defense mechanism, on the other hand, was significantly reduced in the mitochondria of SHR (Chan et al., 2006), allowing oxidative injury to go unchecked and persisted over time after AdeNOS gene transduction. Our results showed that comparable inhibition of UCP-2 expression was exhibited in ventrolateral medulla of SHR or WKY rats that received AdeNOS alone or in combination with AdSOD2. These results suggest the possibility of a direction inhibitory action of NO on UCP-2 expression that is independent of . The underlying molecular pathway remains to be established. UCP-2 belongs to the mitochondrial anion carrier family located in the inner mitochondrial membrane, which dissipates the mitochondrial proton gradient and causes a decrease in ROS production (Korshunov et al., 1997; Andrews et al., 2005). Our results therefore suggest that NO generated by the eNOS transgene while it was still prevalent down-regulated UCP-2 expression, leading to the subsequent increase in mitochondrial level in the RVLM that underlies the rebound hypertension in SHR.

The reason for NO-dependent cardiovascular effects to outlast the expression of individual mRNA after gene transduction of eNOS and SOD2 are not immediately clear. Oxidative stress has been shown to cause a deficiency in tetrahydrobiopterin (BH₄), an essential cofactor for eNOS, in blood vessels (Landmesser et al., 2003). Conversely, reducing generation enhances BH₄ levels and prompts endogenous eNOS to regain its normal function to produce NO (Antoniades et al., 2006; Li et al., 2006). We also recognize that apart from the mitochondrial MRC, eNOS may produce at the expense of NO through a process referred to as “eNOS uncoupling” (Xia et al., 1998). Whether this mechanism is involved in the generation of and its contribution to cardiovascular effects after eNOS gene transduction, however, awaits further investigation. One hint arises from eNOS transgenic mice (eNOS-Tg) and double transgenic mice (eNOS/GCH-Tg) that included both eNOS and cyclohydrolase 1 (GCH), the rate-limiting enzyme in BH₄ biosynthesis (Bendall et al., 2005). Whereas NO production is markedly increased in the aorta of both strains of transgenic mice compared with wild-type littermates (Adlam et al., 2007), the magnitude of augmented NO level is comparable in eNOS/GCH-Tg and eNOS-Tg mice. This suggests that eNOS uncoupling after overexpression of eNOS is minimal.

We recognize that a fundamental premise in the present study is that the delivered transgenes effectively transduced the targeted gene products in the RVLM. Our results from real-time RT-PCR confirmed that coadministration of AdeNOS or AdSOD2 into the RVLM elicited an up-regulation of eNOS or SOD2 mRNA in ventrolateral medulla that exhibited a temporal profile that was comparable with the induced eNOS or SOD2 protein expression (Kishi et al., 2002, 2003; Kimura et al., 2005; Tai et al., 2005; Chan et al., 2006). Double immunofluorescence staining coupled with confocal microscopy further established that application of AdeNOS and AdSOD2 leads to cotransduction of eNOS and SOD2 within the cytoplasm of RVLM cells. Our results also refuted the possibility that the unsustained depressor response after overexpression eNOS in RVLM of SHR was a consequence of reduced transduction efficiency. Both SHR and WKY rats exhibited comparable time course and amplitude of up-regulation of eNOS or SOD2 in ventrolateral medulla on cotransduction of AdeNOS and AdSOD2.

Three factors determined our choice of eNOS gene transduction to increase NO production in the RVLM. First, transduction of adenoviral vectors encoding eNOS in brain tissue has been demonstrated to exert minimal inflammatory and neurotoxic effects while eliciting a 2-fold increase in NO levels (Kishi et al., 2002, 2003). Second, eNOS is reported to be more useful in discriminating de novo production from endogenous production than nNOS in the RVLM (Kishi et al., 2002, 2003). Third, contrary to an augmented nNOS (Edwards et al., 2004) or a reduced iNOS (Chan et al., 2001) mRNA level in SHR, eNOS transcription in the RVLM is comparable between SHR and WKY rats (Kishi et al., 2002; Tai et al., 2005). Transduction of eNOS gene would thus minimize the potential confounding influence of difference in transcriptional regulation of the transduced gene between SHR and WKY rats. We are aware that overexpression of eNOS increases systemic blood flow (Brevetti et al., 2003). The contribution of differential blood flow induced in the RVLM to our observed differential hypotensive effects of eNOS transgene in SHR and WKY awaits further elucidation. The possibility that the differential cardiovascular responses to eNOS gene transduction between SHR and WKY rats is due to differential membrane injury because of oxidative stress as a result of enhanced and in the RVLM of SHR is deemed unlikely. The amount of thiobarbituric acid-reactive substances, an experimental index for lipid peroxidation by free radicals, in stroke-prone SHR rats that received eNOS gene transduction in the RVLM is not different from those that received Adβgal (Kishi et al., 2003).

In conclusion, based on differential results obtained from transduction of AdeNOS and/or AdSOD2 in the RVLM of SHR or WKY rats, the present study revealed that an interaction between NO and that targets MRC-I in the RVLM via the formation of peroxynitrite contributes to the unsustained depressor response after overexpression eNOS in SHR. Prophylactic or therapeutic potential against neurogenic hypertension is implicated for an enhancement of NO production by up-regulating eNOS expression in the RVLM. Nevertheless, our results from gene transduction also prompted a precautionary note; there is a potential detrimental consequence for eNOS-generated NO in the RVLM in an environment of pre-existing oxidative stress. A recent study (Waki et al., 2006) demonstrated the contribution of eNOS-generated NO in the nucleus tractus solitarii to hypertension in SHR. This demonstration, together with results from our present study, raises the issues of whether simple up-regulation of eNOS expression by gene transduction is a reasonable strategic target for the prevention or therapy of neurogenic hypertension. On the other hand, the possibility that regulation of mitochondrial functions may be amenable to manipulation opens a new vista for the development of remedial strategies against neurogenic hypertension.