Activation of oxytocin (OT)ergic projections from the hypothalamic paraventricular nucleus (PVN) to the nucleus tractus solitarii contributes to cardiovascular adjustments during exercise training (EXT). Moreover, a deficit in this central OTergic pathway is associated with altered cardiovascular function in hypertension. Since PVN catecholaminergic inputs, known to be activated during EXT, modulate PVN cardiovascular-related functions, we aimed here to determine whether remodeling of PVN (nor)adrenergic innervation occurs during EXT and whether this phenomenon is affected by hypertension. Confocal immunofluorescence microscopy and tract tracing were used to quantify changes in (nor)adrenergic innervation density in PVN subnuclei and in identified dorsal vagal complex (DVC) projecting neurons (PVN-DVC) in EXT normotensive [Wistar-Kyoto rat (WKY)] and hypertensive [spontaneously hypertensive rat (SHR)] rats. In WKY, EXT increased the density of PVN dopamine β-hydroxylase immunoreactivity (DBHir) (160%). Furthermore, the number and density of DBHir boutons overlapping PVN-DVC OTergic neurons were also increased during EXT (130%), effects that were blunted in SHR. Conversely, while DBHir in the medial parvocellular subnucleus (an area enriched in corticotropin-releasing hormone neurons) was not changed by EXT in WKY, a diminished DBHir was observed in trained SHR. Overall, these data support the concept that the PVN (nor)adrenergic innervation undergoes plastic remodeling during EXT, an effect that is differentially affected during hypertension. The functional implications of PVN (nor)adrenergic remodeling in relation to the central peptidergic control of cardiovascular function during EXT are discussed.
- synaptic remodeling
- sympathetic nervous system
- synaptic plasticity
there is growing evidence that exercise activity reduces the risk of numerous pathological conditions, including obesity, cardiovascular disease, and diabetes (23). For example, exercise training (EXT) promotes several beneficial cardiovascular adjustments, including remodeling of the heart and skeletal muscle circulation (5, 46), improvement of autonomic control of the heart, and the appearance of resting bradycardia, a characteristic marker of EXT (13, 51). Moreover, EXT has been shown to reduce cardiovascular risk in hypertension, by diminishing both sympathetic activity and pressure in hypertensive individuals (5, 6, 14, 46).
Accumulating evidence in recent years has started to elucidate the mechanisms underlying EXT-induced cardiovascular adjustments. In this sense, recent studies from our laboratories showed that within the nucleus tractus solitarii (NTS), vasopressin (VP) and oxytocin (OT) projections originating in the hypothalamic paraventricular nucleus (PVN) participate in the reflex control of the heart, as well as cardiovascular adjustments during EXT (9, 25, 35, 48, 49). For example, our studies support that activation of OT PVN-NTS projections results in improved reflex bradycardia (35) and restrained exercise-induced tachycardia (9).
While activation of the OT-PVN-NTS pathway seems critical for proper cardiovascular adaptations during EXT, the underlying mechanisms leading to its enhanced activation during this condition remain at present unknown. In principle, this could occur in response to increased activity and/or efficacy of functionally relevant afferent inputs.
In this sense, (nor)adrenergic fibers originating mostly from brain stem cell groups, including the A1, A2, and A6 regions (the caudal ventral medulla, the NTS, and the locus coeruleus, respectively), constitute major visceroceptive pathways providing the PVN with critical peripheral cardiovascular information needed for the proper generation of descending homeostatic responses (60). Within the PVN, norepinephrine, via activation of both α- and β-adrenoreceptors (18–20, 28, 59, 63), is a key neurotransmitter modulating both neuroendocrine and autonomic functions (41, 54, 58). For example, norepinephrine within the PVN elicits independent bradycardic and pressor responses, effects mediated by activation of α1- and α2-receptors, respectively (29).
The functional importance of PVN (nor)adrenergic transmission in the regulation of autonomic reflexes during EXT is supported by several lines of evidence. For example, numerous studies support increased activation of (nor)adrenergic hypothalamic inputs (including those innervating the PVN) during EXT (11, 24, 40, 52). Moreover, blockade of PVN adrenoreceptors prevented exercise-mediated increments in norepinephrine and corticosterone plasma levels (61, 62), supporting the view that activation of (nor)adrenergic PVN afferent inputs contributes to autonomic and neuroendocrine adjustments during EXT (61, 62). Finally, given that (nor)epinephrine within the PVN stimulates the OT system (72) and that EXT activates both PVN (nor)adrenergic inputs as well as the OT-PVN-NTS pathway (which leads to improved reflex bradycardia), it is reasonable to speculate that noradrenergic modulatory actions on autonomic responses within the PVN during EXT are mediated, at least in part, via the activation of peptidergic PVN-NTS descending pathways.
In general, increased activity/efficacy of an afferent input during conditions of sustained or repetitive activation, such as during EXT, has been linked to activity-dependent remodeling of the activated input, inducing as a result changes in synaptic density, number and/or molecular configuration of postsynaptic receptors, and intrinsic neuronal excitability, among others (7, 10, 15, 31, 70). For example, activity-dependent changes in (nor)adrenergic synaptic boutons contacting magnocellular neurosecretory neurons have been shown to occur in response to sustained stimuli, such as lactation and dehydration. Given this phenomenon, we hypothesized in the present study that EXT results in structural remodeling (i.e., increased density of synaptic contacts) of PVN (nor)adrenergic inputs contacting OT PVN-dorsal vagal complex (DVC) neurons, contributing, in turn, to activation of this peptidergic pathway and to the autonomic cardiovascular adjustments associated with EXT. Furthermore, and on the basis of previous observations demonstrating that the central OTergic system is differentially affected by EXT and hypertension (45), we also evaluated in this study whether EXT-induced afferent remodeling in the PVN was also modulated during hypertension. To this end, we combined confocal immunofluorescence microscopy and tract tracing to quantify changes in (nor)adrenergic innervation density in PVN subnuclei and in identified OT-PVN-DVC projecting neurons in EXT normotensive [Wistar-Kyoto rat (WKY)] and hypertensive [spontaneously hypertensive rat (SHR)] rats.
MATERIALS AND METHODS
Animals, Training Protocol, and Functional Measurements
Male WKY and SHR (both 2 mo old) were housed on a 12:12-h light-dark cycle and allowed free access to food and water. Rats were preselected for their ability to walk on a treadmill (Inbramed, KT-300, ∼10 sessions, 10–15 min/day, 0.3–0.6 km/h, 0% grade), and only active rats (20 WKY, 20 SHR) were used in this study. Half of the animals of each group (n = 10) were submitted to a low-intensity training protocol (T; = 50–60% of maximal exercise capacity as determined by exercise tests on treadmill, performed 5 days/wk, 1 h/day over 3 mo), while the other half (n = 10), with equivalent capability, were kept sedentary (S) for a similar period of time (9, 25, 45). Maximal exercise tests (graded exercise on treadmill, starting at 0.3 km/h with increments of 0.3 km/h every 3 min up to exhaustion) were performed at weeks 0, 6, and 12 to determine maximal individual exercise capacities, define the training intensity, and compare the efficacy of the training protocol among groups. Rats assigned to the S protocol were handled every day. In a separate set of experiments, as a control for a possible direct effect of the DVC microinjection on (nor)adrenergic remodeling, a subgroup of control WKY lacking a DVC tracer microinjection (n = 8, aged 2 mo) were subjected to the same T (n = 4) and S (n = 4) protocols and measurements of dopamine β-hydroxylase immunoreactivity (DBHir) as detailed below. Finally, to further confirm the efficacy of training, age-matched rats (n = 3/group), submitted to similar T and S protocols, were decapitated for removal of the soleus muscle for measurement of citrate synthase activity (3) at the end of protocols. Briefly, 100 mg of tissue was homogenized in 1 ml of Tris·HCl (20 mM)-EDTA (1 mM) solution at pH 7.4 and centrifuged to isolate the supernatant. Citrate synthase activity was determined in 10 μl of the enzyme extract added to the assay buffer [100 nM tris(hydroxymethyl)aminomethane, 0.2 mM DTNB, 0.1 mM acetyl-CoA, and 0.05% (vol/vol) Triton X-100 at pH 8.1, in a final volume of 950 μl] and to 50 μl of oxaloacetate (final concentration of 10 mM). Absorbance was measured over 10 min at 412 nm, and the results were expressed in nanomoles per minute per milligram of protein. Protein content was measured according to Ref. 8.
At the end of the T and S protocols, and after the third maximal exercise test, WKY and SHR were subjected to microinjection of a retrograde tracer in the NTS (see below) and allowed to recover for 4–5 days. The T protocol was reestablished 2 days after the surgery and continued during this period of time. One day before the functional experiments were conducted, rats were lightly anesthetized [ketamine-xylazine-acepromazine, 0.7:0.2:0.1 (vol/vol/vol), 0.4 ml/kg im] for catheterization of the femoral artery (45). Mean arterial pressure (MAP) and heart rate (HR) were measured 24 h after the catheter implantation and 26–30 h after the last training session in conscious, freely moving rats resting in their home cages. The arterial catheter was connected to the recording system (P23Db transducer, 3400 Recorder; Gould), and a variable period of time (20–45 min) was allowed for stabilization of cardiovascular parameters before starting the simultaneous measurement of AP and HR for ∼30 min (basal values).
Retrograde Tracer Injection
The four groups of WKY and SHR (WKYS, WKYT, SHRS, and SHRT) were anesthetized (0.8 ml/kg im) and placed in a stereotaxic apparatus (David Kopf). The dorsal brain stem was exposed, and Fluorogold was administered into the medial NTS (200 nl), according to the following coordinates: 0.6 mm lateral (right or left) and 0.9 mm deep, considering the calamus scriptorius as the reference zero. The site and extension of the microinjections were verified histologically as previously described (65). While the injections were in most cases centered at the NTS, the dye was found to “leak” into the nearby dorsal motor nucleus of the vagus. Thus injections were considered to comprise the DVC, and retrogradely labeled neurons were consequently named as PVN-DVC. To control for a possible direct effect of the DVC microinjection on (nor)adrenergic remodeling, a subset of control WKY used in these studies did not receive a retrograde tracer injection.
All protocols and surgical procedures used were in accordance with the “Ethical Principles in Animal Research” adopted by the Brazilian College of Animal Experimentation and were approved by the Institutional Animal Care and Use Committees at the University of São Paulo and the University of Cincinnati.
After functional measurements, rats were deeply anesthetized (pentobarbital sodium, 60 mg/kg ip) and submitted to transcardiac perfusion with 0.01 M PBS (150 ml) followed by fixative (4% paraformaldehyde in 0.1 M PBS, pH 7.2, ∼500 ml). Rats were decapitated, and brains were postfixed (4% paraformaldehyde for 4 h at 4°C), cryoprotected (0.1 M PBS containing 30% sucrose at 4°C) for a minimum of 72 h, blocked, and stored at −80°C until processing. Hypothalamic coronal sections (Leica cryostat CM3050, 25 μm) were collected in tissue culture wells with 0.01 M PBS at 4°C. Sections were incubated with 0.01% Triton X-100 and 10% normal horse serum for 1 h. For triple immunofluorescence reactions, sections were incubated for 24 h in a mixture of primary antibodies that included a polyclonal guinea pig anti-oxytocin (Bachem, 1:100,000 dilution), a mouse monoclonal anti-DBH (1:20,000 dilution), and a polyclonal rabbit anti-Fluorogold (Chemicon, 1:10,000 dilution). Reactions with primary antibodies were followed by a 2-h incubation with secondary antibodies [donkey anti-guinea pig-Cy3 labeled (1:400 dilution), donkey anti-mouse Cy5 labeled (1:50 dilution), and donkey anti-rabbit Cy2 labeled (1:400 dilution), diluted in PBS containing 0.1% Triton X-100 and 0.4% NaN3 and purchased from Jackson ImmunoResearch Laboratories]. Control experiments were performed by omitting primary or secondary antibodies.
Fluorescence Confocal Imaging Acquisition and Analysis
Because of a variety of factors, including misplacement of retrograde tracers and loss of tissue during histological procedures, the number of brains available for these studies was unevenly diminished among the WKY and SHR groups (WKYS = 6; WKYT = 8, SHRS = 10, SHRT = 5).
Histological sections were examined with a Leica TCS SL confocal microscope (Leica Microsystems, Mannheim, Germany). Argon-krypton and HeNe lasers were used to excite the FITC and Cy3 fluorochromes at 488 and 543 nm and the Cy5 fluorochrome at 633 nm, respectively. Fluorescent signal cross talk among the different channels was avoided by setting image acquisition parameters with individually labeled sections. Each optical section was averaged three times. All images from the different experimental groups were digitized with identical microscope settings. In this case, omission of primary or secondary antibodies resulted in an almost complete negative (i.e., dark) image (results not shown).
Quantification of DBHir within specific PVN subnuclei.
Stacks of six consecutive optical focal planes (z step: 1 μm, image size 1,024 × 1,024 pixels, × 20 lens) were obtained from two rostrocaudal levels of the PVN [bregma level (B): −1.80 mm and B: −2.10 mm], and a projection image was generated. Imaging analysis was performed with Image Pro software (Media Cybernetics, Silver Spring, MD). An automated tracing procedure incorporating a threshold paradigm (4) was applied to the channel containing the DBHir information. Background intensity was calculated from random adjacent areas in the neuropil, and the threshold was set to pass intensities 1.5× above background immunofluorescence. Regions of interest (ROIs, n = 3) of fixed sizes were drawn within the following PVN subnuclei, according to Ref. 66: lateral magnocellular (LM), parvocellular ventromedial (VM), dorsal cap (DC), parvocellular posterior (PaPo), and medial parvocellular (MP). The distributions of retrogradely labeled and OT immunoreactive neurons were also used as landmarks to delineate the various subnuclei (see ⇓Fig. 2). The density of DBHir thresholded signal within each ROI (expressed as % threshold area = area occupied by thresholded signal/total ROI area*100) was calculated, and a mean value for each PVN subnuclei per animal was obtained (38). Mean DBHir density values were then obtained for each experimental condition/group.
Quantification of DBHir in identified OTergic, PVN-DVC projecting neurons.
Double-labeled OT-Fluorogold neurons in the PaPo subnucleus were identified by merging confocal images obtained from the Cy2 and Cy3 fluorophores. A positive colocalization was considered by the appearance of yellow (red + green) profiles in the merged image. Stacks of confocal images containing the three channels (Cy2, Cy3, and Cy5) and spanning the whole extension of double-labeled neurons (∼50 sections, z step: 0.5 μm, image size 1,024 × 1,024 pixels, ×63 oil immersion lens, ×4 digital zoom) were then obtained. Only double-labeled neurons whose somata were fully contained within the z stack were selected for quantification. All neurons contained within the sampled sections fulfilling these criteria were used for quantification purposes. These image stacks were imported into Volocity (Improvision, Lexington, MA), and neurons, along with associated DBHir boutons, were reconstructed in three dimensions. Since Fluorogold results in a punctate staining, the OT signal (red channel), which provides a diffuse, homogeneous staining of the neuron, along with the blue (DBHir) channel were classified to establish an intensity threshold for neuronal and bouton signals. Threshold was set to pass intensities 5× above background immunofluorescence, as determined from the frequency distribution of voxel intensities calculated for each respective channel. The surface area and volume of positive elements in the neuronal and bouton channels were measured independently. Volocity software algorithms enabled colocalization analysis, discriminating between boutons that overlap with the neuron from those that do not. The density of overlapping boutons was calculated by dividing the number of overlapping boutons by the respective neuronal surface area, also provided by Volocity. In addition, the mean intensity of overlapping boutons was calculated and expressed as fluorescent arbitrary units (AUs), ranging from 0 (absolute black) to 255 (absolute blue). These approaches have been recently validated to study changes in synaptic innervation density (50), presynaptic varicosity density (68), and synaptic shape (73).
Since DBH is responsible for the conversion of dopamine into norepinephrine, the DBHir structures are considered to represent (nor)adrenergic fibers and are mentioned throughout the text as such.
Results are expressed as means ± SE. Differences in MAP, HR, and running distance between groups (WKY and SHR) and conditions (S and T) were analyzed with a two-way ANOVA. A nonparametric ANOVA (Kruskal-Wallis) was used to compare DBHir data (data were not normally distributed; results not shown). Dunn's multiple comparison tests were used as post hoc analyses. Differences were considered significant at P < 0.05.
Efficacy of Training Protocol and Functional Measurements
As shown in Fig. 1, the gain in performance was similar in the trained WKY and SHR groups at the end of protocols: running distance attained during maximal exercise tests on treadmill was increased by 451 ± 16 and 434 ± 29 m from week 0 to week 12 in WKYT and SHRT groups, respectively (P < 0.05; n = 10 /group), while in the sedentary controls effective running distance was unchanged or slightly decreased during the 12-wk period (−7 ± 5 and −114 ± 34 m in WKYS and SHRS, respectively). In addition, citrate synthase activity, measured in the soleus muscle in a subgroup of sedentary and trained rats that underwent the same training protocol, revealed that exercise was effective to increase muscle oxidative metabolism by 22% (from 232 ± 20 to 282 ± 19 nmol·min−1·mg protein−1). In these trained rats, maximal exercise tests also showed a significant increase of 406 ± 45 m in treadmill running distance. Hemodynamic measurements (Fig. 1) showed that SHRS exhibited higher resting MAP and HR than respective normotensive controls (168 ± 3 mmHg and 362 ± 11 beats/min vs. 112 ± 2 mmHg and 307 ± 5 beats/min in WKYS; P < 0.05). EXT was effective in decreasing baseline HR (9% reduction, P < 0.05; Fig. 1) in WKY and SHR but did not change resting MAP in either group.
(Nor)adrenergic Afferent Density in PVN: Dependence on Exercise Training, Hypertension, and Subnuclei Distribution
To study changes in the density and/or distribution of (nor)adrenergic inputs in various PVN subnuclei, retrograde tract tracing (Fluorogold injected into the DVC) along with triple immunofluorescence for DBH, OT, and Fluorogold was used. Representative examples of DBHir in LM, DC, MP, VM, and PaPo subnuclei at two different PVN rostrocaudal levels (B: −1.80 mm and B: −2.10 mm) are shown in Fig. 2.
Initially, DBHir was measured within the whole PVN and values were compared between conditions (S × T) and groups (WKY and SHR) [n = 6 (WKYS), 8 (WKYT), 10 (SHRS), and 5 (SHRT)]. Results are summarized in Fig. 3, A and B.
Our results indicate that DBHir varied significantly among groups (P < 0.001, Kruskal-Wallis test). Post hoc analyses showed that while DBHir increased in WKYT compared with WKYS (160%; P < 0.005), a significantly decreased DBHir was observed in SHRT compared with SHRS (44%; P < 0.05). Furthermore, while DBHir in S rats was not different between WKY and SHR (P > 0.05), DBHir was significantly higher in WKYT compared with SHRT (∼260%; P < 0.05). A similar increase in PVN relative DBHir density was observed in control EXT WKY that did not receive a DVC microinjection of the retrograde tracer (S: 6.2 ± 0.7 relative % density; T: 15.4 ± 4.12 relative % density; P < 0.02, Mann Whitney U-test, n = 4/group) (see discussion).
To determine whether the influence of experimental condition (S × T) or experimental group (WKY × SHR) on DBHir was dependent on its topographical distribution within the PVN, data were further analyzed according to identified neuroendocrine and autonomic-related PVN subnuclei. Results are summarized in Fig. 3C. Our results indicate that DBHir varied significantly among groups (P < 0.0001, Kruskal-Wallis test). In WKY, a statistically significant increase in DBHir induced by EXT was observed in the PaPo subnucleus (∼320%; P < 0.05), a region known to contain autonomic-related PVN neurons. Despite a tendency for increased DBHir levels in other autonomic-related subnuclei (i.e., DC and VM), differences did not reach a statistically significant level (P > 0.1). Conversely, in SHR DBHir was significantly reduced during EXT in the MP subnuclei (∼63%; P < 0.05), a region known to predominantly contain corticotropin-releasing factor neurons (67). No significant differences in DBHir between SHRS and SHRT were observed in the other PVN subnuclei (P > 0.1 for all subnuclei).
While the relative density of DBHir varied among groups, no differences in the total PVN surface area were observed among groups (P > 0.2 for B: −1.80 mm and B: −2.10 mm; see Table 1).
Changes in Density of DBHir Boutons Overlapping Identified PVN-DVC OTergic Projecting Neurons
To determine whether changes in PVN DBHir among the different experimental conditions and groups were associated with OT-immunoreactive PVN-DVC projecting neurons (OT-PVN-DVC), the degree of DBHir boutons in close apposition with retrogradely labeled OT-PVN-DVC was calculated. OT-PVN-DVC neurons from WKYS (n = 19), WKYT (n = 20), SHRS (n = 17), and SHRT (n = 18), located in the PaPo subnucleus, were reconstructed in three dimensions, and the number, density, and intensity of DBHir boutons overlapping with the membrane surface of somatic and proximal dendritic compartments of OT-PVN-DVC neurons were calculated in each experimental group (see materials and methods). A representative example of a three-dimensional (3D) reconstructed OT-PVN-DVC neuron and associated DBHir boutons is shown in Fig. 4.
DBHir boutons overlapping OT-PVN-DVC neurons had a mean surface area of 7.0 ± 0.3 μm2 and a mean volume of 0.85 ± 0.06 μm3. No differences in the size of DBHir boutons were observed among groups (see Table 2).
Our results indicate that the number of DBHir boutons overlapping OT-PVN-DVC neurons varied significantly among groups (P < 0.02, Kruskal-Wallis test). As summarized in Fig. 5A, the number of overlapping DBHir boutons significantly increased in WKYT compared with WKYS (130%; P < 0.05). On the other hand, no differences were observed between SHRT and SHRS (P > 0.5). Similarly, no differences were observed between WKYS and SHRS or WKYT and SHRT (P > 0.5 in all cases).
To determine whether the density of overlapping DBHir boutons in OT-PVN-DVC neurons was also affected, the surface area of 3D reconstructed neurons was calculated. The density of DBHir boutons was determined by calculating the number of overlapping DBHir boutons/unit of neuronal membrane area (see materials and methods). Results are summarized in Fig. 5B.
The density of DBHir boutons associated with OT-PVN-DVC projecting neurons still varied among experimental conditions/groups (P < 0.05, Kruskal-Wallis; Fig. 5B), and results from post hoc tests indicated that the density of overlapping DBHir boutons was significantly enhanced in WKYT compared with WKYS (170%; P < 0.05). On the other hand, no significant differences were observed between SHRS and SHRT (P > 0.5). Differences in DBHir density were not due to changes in neuronal surface area, because the latter was found to be similar among experimental groups (P > 0.5, Kruskal-Wallis test; Fig. 5C). Furthermore, highly weak correlations were observed between the number or the density of overlapping DBHir boutons and neuronal surface area (R2 values for DBHir counts: 0.2, 0.3, 0.1, and 0.3 for WKYS, WKYT, SHRS, and SHRT, respectively; R2 values for DBHir density: 0.2, 0.3, 0.1, and 0.1 for WKYS, WKYT, SHRS, and SHRT, respectively). Finally, no differences in the mean intensity signal of DBHir overlapping boutons were observed among groups (WKYS: 86.6 ± 1.3 AU, WKYT: 87.6 ± 1.5 AU, SHRS: 85.9 ± 1.7 AU, SHRT: 88.5 ± 1.3 AU; P > 0.5, Kruskal-Wallis).
Interestingly, while no differences in neuronal surface area were observed among experimental conditions/groups, we did observe differences in OT-PVN-DVC cell volume, which varied significantly among experimental conditions/groups (P < 0.05, Kruskal-Wallis test). As summarized in Fig. 5D, results from post hoc tests indicated that cell volume was significantly larger in SHRS compared with WKYS (117%; P < 0.01). Furthermore, EXT resulted in a decreased cell volume in SHR (28%; P < 0.05) but not in WKY (P > 0.5).
Our results support plastic remodeling of PVN (nor)adrenergic innervation during EXT. The higher EXT-induced DBHir in the PVN of control animals, as well as the increased density of DBHir boutons overlapping OT-PVN-DVC neurons, support an enhanced synaptic action of (nor)adrenergic transmitters in this autonomic-related neuronal population. Thus the present studies indicate that, along with the previously reported plasticity in the intrinsic membrane properties of OT-PVN-DVC neurons during EXT (36), structural remodeling of (nor)adrenergic synaptic inputs constitutes an important underlying mechanism contributing to enhanced activation of the OT-PVN-DVC pathway and associated autonomic cardiovascular adjustments during EXT. Finally, while no differences were observed between sedentary, normotensive, and sedentary hypertensive rats, our results indicate that EXT-induced plasticity was differentially affected in hypertensive rats.
Important methodological aspects of these studies deserve to be considered. First, since the efficacy of EXT was similar in WKY and SHR rats, our data suggest that differences in EXT-induced changes in DBHir between the groups were not dependent on different degrees of training. However, we cannot rule out at present a contribution of differential adaptive mechanisms (e.g., differences in oxidative metabolism, muscle weight, etc) among groups. Second, since these studies were restricted to the “established, chronic” phase of the hypertension, we are unable to determine at present whether differences observed between WKY and SHR are a result of the hypertension per se or due to inherent differences between WKY and SHR. Future studies using animals at a prehypertensive stage will be needed to address this point. Third, using confocal microscopy with 3D reconstructions, we assessed changes in DBHir with two complementary approaches: measurements of overall immunoreactive density in discrete PVN regions and quantification of immunoreactive boutons overlapping target-specific and neurochemically identified autonomic-related PVN neurons. The latter approach was recently validated as an efficient tool to assess plasticity of neurotransmitter innervation in identified neurons after prolonged stimulation (50). Furthermore, measurement of PVN DBH innervation density was recently shown to be highly reproducible, with low interexperimental variability (38). In fact, the PVN regions and single-cell DBHir density values reported in this study are in general agreement with those reported by others (38, 50).
Individual DBHir boutons were contained in two or three 0.5-μm optical sections (within the resolving power of the confocal system used), enabling us to count them individually. However, because labeled dendrites could only be followed within the thickness of a given section, and in many instances were obscured by labeled neighboring neurons or dendrites, we were limited to quantifying boutons in identified somata and proximal dendrites.
In agreement with previous unpublished observations from our laboratories, we found that microinjection of tracers in the DVC (as well as cannulations) interfered with the ability of EXT to reduce blood pressure in hypertensive rats. This is in contrast with previous studies from our own laboratories in non-DVC instrumented animals, in which we reported a clear pressure fall in SHR with a training protocol similar to that used here (5, 6, 46), as expected from previous results both in experimental animals and hypertensive humans (12–14, 45). Thus the lack of EXT-induced fall in blood pressure in the SHR in the present study does not reflect an intrinsic limitation of our EXT model.
Finally, we observed a similar increase in PVN DBHir density in EXT rats that did not receive a tracer injection in the DVC, results that argue against a contribution of this manipulation to EXT-induced increases in (nor)adrenergic innervation, However, a caveat to be taken into consideration is that since these studies were restricted to normotensive animals, an effect of the tracer injection manipulation on DBHir in the hypertensive group cannot be conclusively ruled out at present.
EXT-Induced Increment of DBHir Density in Normotensive Rats
The relative density of PVN DBHir innervation in WKY increased during EXT, and despite the fact that differences reached statistical significance only within the PaPo, a region that mostly contains autonomic-related neurons, including those innervating the dorsal brain stem subnucleus (66, 67), a similar tendency was observed in the other subnuclei analyzed. Thus it would be premature at this stage to conclude that remodeling of (nor)adrenergic inputs to the PVN during EXT occurred in a subnucleus-specific manner. Quantification at the single-cell level demonstrated an increased DBHir innervation density in identified OT-PVN-DVC neurons. It is worth noting that the degree of changes observed within the PaPo subnucleus as a whole was larger than that observed at the single-cell level. This could be due in part to the fact that the quantification in the latter was limited to somatic and proximal dendritic compartments (see materials and methods). Since (nor)adrenergic inputs to the PVN synapse both on somatic and dendritic segments (21, 42, 43), it is likely that exclusion of distal dendrites from our single-cell analysis resulted in an underestimation of the degree of changes in DBH innervation per neuron during EXT. However, we cannot rule out that other autonomic-related neurons contained within the PaPo [either neurochemically different PVN-DVC projecting (e.g., VP) or neurons projecting to other targets] may be affected as well, and to a different extent than OT-PVN-DVC neurons. In light of previous studies from our laboratory supporting changes in the PVN-NTS VPergic system during EXT (49), it would be important in future studies to determine whether EXT-induced plasticity in DBHir also affected this particular PVN neuronal population.
The precise mechanisms underlying the increased DBHir density are at present unknown. In principle, this could result from formation of new immunoreactive terminals and/or detection of preexisting, previously unresolved terminals, the latter due to increased expression of DBH above the detectability threshold of the immunofluorescence approach. In general, increments in DBHir within preexisting terminals would be expected to result in an increased mean immunoreactive intensity signal (unless these changes were restricted to previously undetected terminals). Thus our results showing increased DBH density without concomitant changes in immunoreactivity intensity argue against this latter mechanism. It is important, however, to emphasize that the approaches used here cannot accurately differentiate between these alternative, though not mutually exclusive, mechanisms. Finally, while changes in DBHir density could be secondary to changes in cell size, our results showing no differences in neuronal surface area among groups, as well as a lack of correlation between number of overlapping DBHir boutons and neuronal surface area argue against this possibility.
While numerous studies support an important role of PVN catecholaminergic inputs in neuroendocrine and autonomic control (41, 54, 58), as well as increased activation of these inputs during EXT (11, 24, 40, 52), the potential contribution of catecholaminergic remodeling to enhanced PVN-DVC activity during this condition remains to be elucidated. Interestingly, recent work indicates that OT mRNA expression in the PVN is modulated by changes in local norepinephrine levels (72). Thus, based on our results showing concomitant increments in DBHir density (this work) and PVN/NTS OT mRNA levels (45) during EXT, one could speculate that the PVN (nor)adrenergic remodeling during EXT reported here may contribute to the increased efficacy of the OTergic PVN-DVC pathway, constituting, in turn, an important mechanism leading to OT-mediated cardiovascular adjustments during training. This is also supported by the concomitant lack of EXT-induced (nor)adrenergic remodeling and blunted OT mRNA in SHR (see below).
Interactions Between Exercise Training and Hypertension
Abundant evidence supports an involvement of the PVN in the pathophysiology of hypertensive disorders (1, 2, 32, 45, 69), including changes in PVN noradrenergic levels during hypertension (56, 57). Moreover, our recent findings showing depressed OT content and OT receptor mRNA levels in the PVN and NTS, respectively, during hypertension (45), along with a blunted tachycardic response in hypertensive rats following OT receptor antagonist microinjection in the NTS (34), support a general deficit in the central OT system during hypertension. Importantly, we found this deficit to be only partially improved by EXT (45). On the basis of these data, and the well-established relationship between central (nor)adrenergic and the OTergic hypothalamic system (see Refs. 53 and 64 for a review), we hypothesized interactions between EXT and hypertension. A recent report by DiCarlo et al. (22) showed that exercise normalized the decreased number of diaphorase (nitric oxide synthase)-positive neurons found in the PVN of hypertensive rats. Our present data suggest a different pattern of interaction between exercise and hypertension in terms of the PVN catecholaminergic innervation: while no differences in (nor)adrenergic innervation density were observed between WKYS and SHRS, the EXT-induced increase in PVN (nor)adrenergic innervation was absent, or even decreased, during hypertension, depending on the PVN subnuclei analyzed. The blunted DBHir remodeling in the PaPo subnuclei, and specifically in OT-PVN-DC neurons of hypertensive rats, is in agreement with our recent results showing that the robust EXT-induced increment in PVN and NTS OT mRNA levels was partially blunted in hypertensive rats (45). Moreover, previous studies also support opposing effects of EXT on brain norepinephrine function in normotensive and hypertensive animals (24, 26, 40). Thus the absent EXT-induced changes in (nor)adrenergic innervation during hypertension in OT-PVN-NTS neurons may contribute to the diminished efficacy of EXT to improve central OT function during hypertension (45).
As a caveat, it is important to consider that changes in the density, distribution, and/or efficacy of presynaptic (nor)adrenergic inputs may be modulated or offset by concomitant changes in postsynaptic receptor function. Thus future studies will be needed to confirm a mechanistic and functional link between (nor)adrenergic afferent remodeling and plasticity in the OT and VP central peptidergic systems during EXT.
While EXT-induced increase in DBHir in the PaPo subnucleus was blunted in SHRT, an opposite effect was observed in the MP subnucleus, where a diminished DBHir was observed only in SHRT. The significance of this differential effect observed in the MP subnucleus, an area enriched in corticotropin-releasing hormone (CRH) neurons (67), is at present unknown. Direct catecholaminergic inputs from the brain stem to this region have been described (16, 17), activation of which is usually associated with a facilitatory effect on the hypothalamic-pituitary-adrenal (HPA) axis (see Ref. 33 for a review). Interestingly, numerous studies support that both EXT (55, 71) and hypertension increase central activation of the HPA axis (27, 30, 39), effects that, in light of the present results, may not require an increased degree of catecholaminergic innervation in a PVN region enriched in CRH neurons. Studies evaluating the interactions between training and hypertension in the central regulation of the HPA axis are, unfortunately, scarce. Worth mentioning is an early study reporting that the increased plasma corticosterone levels following a session of acute exercise were diminished in trained compared with untrained SHR (44). While highly speculative at this point, a diminished catecholaminergic innervation of CRH-containing PVN regions in SHRT (as found in this study) may contribute to a diminished HPA axis activation in this group.
While exercise activity is associated with a variety of compensatory cardiovascular adjustments, the mechanisms and pathways underlying such responses remain obscure. A growing body of evidence indicates that the central (nor)adrenergic system as well as OTergic and VPergic PVN-DVC projections are important components of the neuronal circuitry underlying adaptive cardiovascular responses during exercise. Within this context, recent data from our laboratory showing plasticity in the intrinsic membrane properties of PVN-DVC neurons (37), along with the EXT-induced remodeling in their (nor)adrenergic innervation (present study), support the concept that functional and structural plasticity of intrinsic and extrinsic factors controlling neuronal function within the PVN may constitute key mechanisms contributing to the ability of these circuits to adapt to the demanding conditions associated with exercise training.
Interestingly, numerous studies support a remarkable neuroplastic capacity of the hypothalamic (nor)adrenergic system. For example, hypothalamic (nor)adrenergic innervation decreased during osmotic stimuli (50) and in response to stimuli provoking CRH release (38). Conversely, an increased (nor)adrenergic innervation density was observed during lactation (47). Thus these studies together support the general concept that (nor)adrenergic synaptic rearrangement in the hypothalamus may be differentially regulated, depending on the stimulus modality and the neuronal types affected. Whether and how these adaptive physiological plastic mechanisms are affected by various pathological conditions will be a challenge for future investigations.
This work was supported by National Heart, Lung, and Blood Institute Grant RO1-HL-68725 (J. E. Stern) and Fundação de Amparo a Pesquisa do Estado de São Paolo Grants 02/00621-0 and 05/60244-4.
We thank Dr. Wenfeng Zhang for excellent technical support.
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
- Copyright © 2007 the American Physiological Society