Perivascular sympathetic nerves are important determinants of vascular function that are likely to contribute to vascular complications associated with hyperglycemia and diabetes. The present study tested the hypothesis that glucose modulates perivascular sympathetic nerves by studying the effects of 7 days of hyperglycemia on norepinephrine (NE) synthesis [tyrosine hydroxylase (TH)], release, and uptake. Direct and vascular-dependent effects were studied in vitro in neuronal and neurovascular cultures. Effects were also studied in vivo in rats made hyperglycemic (blood glucose >296 mg/dl) with streptozotocin (50 mg/kg). In neuronal cultures, TH and NE uptake measured in neurons grown in high glucose (HG; 25 mM) were less than that in neurons grown in low glucose (LG; 5 mM) (P < 0.05; n = 4 and 6, respectively). In neurovascular cultures, elevated glucose did not affect TH or NE uptake, but it increased NE release. Release from neurovascular cultures grown in HG (1.8 ± 0.2%; n = 5) was greater than that from cultures grown in LG (0.37 ± 0.28%; n = 5; P < 0.05; unpaired t-test). In vivo, elevated glucose did not affect TH or NE uptake, but it increased NE release. Release in hyperglycemic animals (9.4 + 1.1%; n = 6) was greater than that in control animals (5.39 + 1.1%; n = 6; P < 0.05; unpaired t-test). These data identify a novel vascular-dependent effect of elevated glucose on postganglionic sympathetic neurons that is likely to affect the function of perivascular sympathetic nerves and thereby affect vascular function.
- sympathetic nervous system
- vascular smooth muscle
the sympathetic nervous system is a major determinant of cardiovascular function. The sympathetic nervous system acts in part via release of neurotransmitters from postganglionic sympathetic neurons innervating blood vessels. These neurons are critical for maintenance of blood pressure and distribution of blood flow (19, 21). Cardiovascular complications are a leading cause of morbidity and mortality in patients with diabetes (1), and evidence suggests that vascular sympathetic neurons contribute to these complications (8, 17). Despite the importance of vascular sympathetic neurons for cardiovascular function and the clinical relevance of cardiovascular complications associated with diabetes, little is known about how diabetes affects the function of these neurons.
Elevated glucose causes many of the complications associated with diabetes (3, 36). The effects of elevated glucose on vascular sympathetic neurons are not well understood. Glucose may act directly on the neurons. In addition, elevated glucose affects vascular cells (3), and vascular cells modulate the function of postganglionic sympathetic neurons (10–12, 23, 24). Thus elevated glucose may also affect vascular sympathetic neurons indirectly via vascular-dependent mechanisms.
In the present study, in vitro and in vivo models were used to assess direct and vascular-dependent effects of elevated glucose on vascular sympathetic neurons. In vitro studies indicate that in neuronal cultures, elevated glucose decreased tyrosine hydroxylase (TH) expression, decreased norepinephrine (NE) uptake, and had no effect on NE release. In contrast, in neurovascular cultures, elevated glucose did not affect TH or NE uptake but increased NE release. In vivo, TH and NE uptake were unchanged, and NE release from vascular sympathetic neurons was increased in hyperglycemic streptozotocin (STZ)-treated rats. These data suggest that elevated glucose increases NE release from vascular sympathetic neurons via vascular-dependent mechanisms. This increase is likely to contribute to vascular complications of hyperglycemia and diabetes.
MATERIALS AND METHODS
Sprague-Dawley rats were used in this study. Adult (250–300 g) male and postpartum females were used to prepare cultures of tail artery vascular smooth muscle. Male and female 3- to 4-day-old rats were used to prepare cultures of postganglionic sympathetic neurons. Femoral arteries from adult male rats were used for measurement of NE release and uptake. Hyperglycemia was induced in a subset of adult male rats by injecting STZ [50 mg/kg in sodium acetate buffer (pH 4.5), ip]. Corresponding control rats were injected with sodium acetate buffer. Two days after injection, blood glucose was assessed in the STZ animals to confirm that these animals were hyperglycemic. Superior ganglia and femoral arteries were harvested from these animals for Western analyses and measurement of NE release and uptake. The use of animals in this study was approved by the Animal Care and Use Committee at the University of Vermont and was in accordance with the National Institutes of Health guidelines for the use of animals in research.
Postganglionic sympathetic neurons were isolated from superior cervical ganglia of male and female Sprague-Dawley neonatal rats. Ganglia were collected and dissociated for 10 min at 37°C in a collagenase-hyaluronidase solution (10 mg/ml bovine serum albumin, 4 mg/ml collagenase, and 1 mg/ml hyaluronidase) and then for 10 min in trypsin (3 mg/ml). Dissociated cells were resuspended in neuronal medium [DMEM-F12 supplemented with 5% fetal bovine serum (FBS), 10% NuSerum (Collaborative), 50 ng/ml NGF, and penicillin-streptomycin] and applied to collagen-coated tissue culture dishes. The cells were allowed to attach overnight in a humidified 5% CO2 environment maintained at 37°C. Nonneuronal cells were then growth arrested with mitomycin C (10 μg/ml for 1 h).
Vascular smooth muscle cells (VSM) were obtained from tail arteries of adult male and female Sprague-Dawley rats as described by Ross (29). The cells were grown in low-glucose DMEM supplemented with 10% FBS, 100 units of penicillin, and 100 units of streptomycin. Cells were maintained at 37°C in a humidified 5% CO2 environment. VSM were used for experiments after two passages with trypsin. Neurovascular cultures were prepared by adding VSM to mitomycin-treated postganglionic sympathetic neurons. These cultures were then grown in neuronal medium.
Cells were pelleted in PBS and then lysed in enhanced RIPA buffer. Tissues were excised from adult rats and homogenized in enhanced RIPA buffer (50 mM Tris base, 150 mM NaCl, 10 mM EDTA, 0.25% deoxycholate, 1% Nonidet P-40 substitute, 10% glycerol, 1% protease inhibitor cocktail, 1 mM DTT, and 0.1% sodium dodecyl sulfate). Cell and tissue samples were diluted with an equal volume of 2× electrophoresis loading buffer, boiled for 5 min, electrophoresed on 4–20% gradient Tris-glycine polyacrylamide gels, and transferred to nitrocellulose membranes. The membranes were blocked with 3% nonfat dry milk in PBS containing 0.05% Tween (PBST) for 20 min at room temperature and then incubated overnight at 4°C in blocking solution containing the appropriate primary antibody. Unbound primary antibodies were then removed with three 5-min washes (PBST), and the membranes were incubated for 1 h at room temperature in PBST containing 3% nonfat dry milk and a 1:3,000 dilution of horseradish peroxidase-conjugated secondary antibody. Unbound secondary antibodies were removed with three 5-min washes (PBST). Horseradish peroxidase was then detected with enhanced chemiluminescence (Pierce) and documented on autoradiographic film. Signals were quantified densitometrically.
NE release and uptake.
NE release was assessed using tritiated NE (Amersham). These assays were performed using low (5 mM d-glucose, 20 mM l-glucose)- or high (25 mM d-glucose)-glucose HEPES-buffered Krebs solution [122 mM NaCl, 3 mM KCl, 0.4 mM MgSO4·H2O, 1.2 mM KH2PO4, 20 mM HEPES, 1.3 mM CaCl2·2H2O, 1 mM ascorbic acid, and 10 μM pargyline (pH 7.4)]. Cell cultures or freshly isolated femoral arteries were preincubated at 37°C with 100 nM tritiated NE for 30 (cells) or 60 min (arteries). Cells and arteries were then washed (6 × 5 min) and pharmacologically [cells; 3 μM 1,1-dimethyl-4-phenylpiperazinium iodide (DMPP)] or electrically stimulated (arteries; 1 min; 4 Hz, 60 mV, 0.5-ms pulse duration). Cells and arteries were then solubilized. Tritiated NE in all samples was collected and analyzed using a Beckman LS6000IC liquid scintillation counter. Basal release was calculated as (cpm in last wash)/(total cpm available for release). Stimulated release was calculated as (stimulated cpm − basal cpm)/(total cpm available for release). NE uptake [total cpm taken up by cells or by the artery (cpm/wet weight)] was also determined.
Data are means ± SE. One-sample (Western analyses and NE uptake) or unpaired (NE release) t-tests were performed to compare data. Differences were considered significant if P values were <0.05.
Neurotransmitters from postganglionic sympathetic neurons innervating target organs produce the effects of the sympathetic nervous system. The primary neurotransmitter released by postganglionic sympathetic neurons is NE. The present study considered the direct and vascular-dependent effects of elevated glucose on TH, the rate-limiting enzyme for the synthesis of NE, NE release, and NE uptake in postganglionic sympathetic neurons.
Direct effects of elevated glucose were studied in cultures of postganglionic sympathetic neurons. Neurons were grown for 7 days in low (LG; 5 mM) or high glucose (HG; 25 mM). These concentrations of glucose correspond to plasma glucose concentrations in normoglycemic and hyperglycemic animals and humans. Figure 1 shows the effects of elevated glucose on TH and GAP43 expression. TH is the rate-limiting enzyme in catecholamine synthesis and is thus a determinant of NE synthesis. GAP43 is a neuronal marker that was not reproducibly affected by the experimental conditions used in the present studies and is a marker of the amount of neuronal protein present in the samples. Representative (n = 4) Western analyses and quantitative analysis are shown. These data indicate that elevated glucose decreased TH. TH expression in neurons grown in HG was less than that in neurons grown in LG. The direct effects of 7 days of elevated glucose on NE release and uptake are shown in Fig. 2. Elevated glucose did not affect basal (Fig. 2A) or stimulated NE release (Fig. 2B), but it decreased NE uptake (Fig. 2C).
Many effects of elevated glucose are due to increased levels of reactive oxygen species (ROS) (3). I determined whether ROS contributed to the observed effect of HG on NE uptake by assessing the effects of HG in the presence of polyethylene glycol (PEG)-catalase. Catalase converts hydrogen peroxide to water and thus decreases levels of ROS (16). PEG-catalase reduced the effect HG on NE uptake (Fig. 2C).
Elevated glucose markedly affects vascular cells (3), and vascular cells markedly affect postganglionic sympathetic neurons (10–12, 23, 24). Thus elevated glucose may affect postganglionic sympathetic neurons indirectly by affecting vascular cells. Vascular-dependent effects of elevated glucose were assessed in sympathetic neurovascular cultures. In blood vessels, the primary targets of postganglionic sympathetic neurons are VSM. In the present study, these vascular cells were considered. In neurovascular cultures, postganglionic sympathetic neurons were grown in the presence of VSM derived from adult rat tail arteries. These VSM were chosen because they are representative of VSM from sympathetically innervated arteries (12). The sympathetic neurons were in direct contact or in close proximity to the VSM, and thus there was reciprocal contact (or proximity)-dependent communication as well as reciprocal exchange of soluble mediators.
The effects of elevated glucose on TH and GAP43 expression in sympathetic neurovascular cocultures were assessed. As noted above, in these cocultures, the neurons and VSM were in direct contact, and thus the samples analyzed would contain both neurons and VSM. Figure 3A shows representative Western analyses of TH and GAP43 expression in postganglionic sympathetic neuron cultures and tail artery VSM. These data indicate that VSM do not express detectable levels of TH or GAP43 and that TH and GAP43 in the neurovascular cultures would be attributable to the postganglionic sympathetic neurons and not the VSM.
Elevated glucose did not affect TH or GAP43 expression in sympathetic neurovascular cultures (Fig. 3, B and C; n = 4; P > 0.05; 1-sample t-test). These data indicate that the effect of elevated glucose on neurons grown in the presence of VSM differs from that on neurons grown in the absence of VSM (Fig. 1), suggesting that the VSM altered the response of the neurons to glucose. Was this effect specific to VSM? Figure 3D shows representative Western analyses (n = 4) of TH and GAP43 in sympathetic neuron/HEK-293 cultures grown for 7 days in LG and HG. Elevated glucose markedly reduced TH and GAP43 in these cultures, suggesting that both VSM and HEK-293 cells altered the neurons response to glucose, but in markedly different ways.
The effects of HG on NE release and uptake in sympathetic neurovascular cultures are shown in Fig. 4. Elevated glucose did not affect basal NE release or NE uptake but increased DMPP-stimulated NE release. PEG-catalase did not inhibit the HG-induced increase in stimulated NE release (n = 2), which suggests that ROS did not contribute to this effect.
To assess the effects of elevated glucose on postganglionic sympathetic neurons in vivo, rats were made hyperglycemic with STZ. The effects of 7 days of hyperglycemia were then studied. Seven days after injection, STZ rats were hyperglycemic (blood glucose > 296 mg/dl) compared with control rats (blood glucose = 108 ± 5.4 mg/dl). Western analyses indicate that in vivo, 7 days of elevated glucose did not affect TH in cell bodies of postganglionic sympathetic neurons in superior cervical ganglia (Fig. 5A) or in sympathetic nerve fibers innervating femoral arteries (Fig. 5B). Representative analyses and corresponding quantitative analyses are shown. TH in the ganglia was normalized to GAP43, which is an index of the amount of neuronal protein in each sample. TH in the arteries was normalized to smooth muscle α-actin, which is an index of the amount of vascular smooth muscle protein in the sample.
In vivo effects of elevated glucose on femoral artery sympathetic nerves were also studied. Electrically stimulated NE release (Fig. 6B) from perivascular nerves of isolated femoral arteries from rats that had been hyperglycemic for 7 days (STZ) was greater than that from arteries of control rats. Basal release (Fig. 6A) and uptake (Fig. 6C) were not significantly different.
In some blood vessels, NE release from perivascular sympathetic nerves is inhibited by presynaptic α2-adrenergic receptors, and the function of these receptors can be inhibited by ROS (14). Thus HG-induced increases in ROS could increase NE release by inhibiting the function of presynaptic α2-adrenergic receptors. This does not appear to be the case in the present study, since the α2-adrenergic antagonist yohimbine (1 μM) did not affect stimulated NE release from femoral arteries of normoglycemic animals. Release in the presence of yohimbine (11.1 ± 0.3%) was not different from that in the absence of yohimbine (14.8 ± 2.6%; P > 0.05; unpaired t-test; n = 3).
The sympathetic nervous system is a major determinant of cardiovascular function. Sympathetic control of vascular and cardiovascular function is mediated in via perivascular sympathetic nerve fibers (4–6). Vascular complications are a major cause of morbidity and mortality in patients with diabetes (1), and studies suggest that perivascular nerves contribute to vascular complications of diabetes (8, 17). Elevated glucose causes many of the complications of diabetes (3, 15, 36). The present study considers how elevated glucose affects postganglionic sympathetic neurons and perivascular sympathetic nerves.
Postganglionic sympathetic neurons are modulated by the targets they innervate (10–12, 23, 24). In blood vessels, vascular cells affect survival (10, 23), growth (12, 24), and neurotransmitter expression (11) of these neurons. Hyperglycemia markedly affects vascular cells (3), which suggests that hyperglycemia would affect vascular modulation of perivascular nerves. For example, work in this laboratory has shown that vascular-derived VEGF promotes the growth of perivascular nerves (24), and Natarajan et al. (27) have shown that hyperglycemia increases VSM production of VEGF. These studies suggest that hyperglycemia is likely to indirectly affect perivascular nerves by increasing perivascular VEGF. The present studies tested the hypothesis that VSM modulate the effects of glucose on postganglionic sympathetic neurons.
To test the proposed hypothesis, the effects of glucose on cultured postganglionic sympathetic neurons grown in the absence and presence of VSM were compared. In the absence and presence of VSM, 7 days of elevated glucose did not affect GAP43 expression (Figs. 1 and 3). Since GAP43 is an index of neuronal protein, these data suggest that in these cultures, this duration of elevated glucose did not affect survival or neurite outgrowth. Elevated glucose decreased TH expression and NE uptake in neurons grown in the absence but not in the presence of VSM (Figs. 1–4). Elevated glucose increased stimulated NE release from neurons grown in the presence but not in the absence of VSM (Figs. 2 and 4). These data indicate that VSM altered the effects of glucose on postganglionic sympathetic neurons and suggest that under hyperglycemic conditions, the VSM maintain or enhance sympathetic neurotransmitter release.
The effects of hyperglycemia on postganglionic sympathetic neurons were also studied in vivo. STZ was used to induce hyperglycemia in adult male rats. This model is routinely used as a model of hyperglycemia and type 1 diabetes (2, 9, 20, 25, 30, 35). Western analyses indicate that 7 days of hyperglycemia did not affect TH or GAP43 expression in cell bodies of postganglionic sympathetic neurons in superior cervical ganglia (Fig. 5A). These data are consistent with previous studies (31, 33) and with data obtained in neuronal cultures grown in the presence of VSM (Fig. 3). These data are not consistent with those obtained in neuronal cultures grown in the absence of VSM (Fig. 1), suggesting that in vivo, direct effects of glucose on the neurons are modulated by other mechanisms that are likely to include vascular-dependent mechanisms. Western analyses also indicate that 7 days of hyperglycemia did not affect TH in perivascular nerve fibers innervating rat femoral arteries. Since TH is the rate-limiting enzyme for NE synthesis, these data suggest that glucose modulation of TH was not a mechanism for affecting NE synthesis in perivascular nerve fibers. Previous studies indicated that hyperglycemia and diabetes did not affect perivascular nerve density in the rat STZ model (35). The present findings that glucose did not modulate TH or GAP43 are consistent with these previous findings. As noted, the superior cervical ganglia data are consistent with the data obtained in neuronal cultures grown in the presence of VSM (Fig. 3) and are not consistent with those obtained in neuronal cultures grown in the absence of VSM (Fig. 1), suggesting that in vivo, direct effects of glucose on the neurons are modulated by other mechanisms that are likely to include vascular-dependent mechanisms. Postganglionic sympathetic neurons in superior cervical ganglia innervate many targets, including blood vessels, but they do not innervate the femoral artery (13). Thus the similar lack of an effect of glucose on TH and GAP43 in neurons in the superior cervical ganglia and in nerve fibers innervating femoral arteries suggests that this is not unique to these tissues. Seven days of hyperglycemia in vivo increased electrically stimulated NE release measured in vitro (Fig. 6B), which is consistent with the increase in DMPP-stimulated NE release observed in neuronal cultures grown in the presence (Fig. 4B) but not in the absence of VSM (Fig. 2B). These data further support the hypothesis that vascular-dependent mechanisms contribute to the effects of glucose on perivascular sympathetic nerves.
Elevated glucose is known to induce oxidative stress in neurons (34) and in vascular cells (3), and oxidative stress affects many cellular processes. The present studies suggest that ROS contribute to direct effects of HG on NE uptake in the neuronal cultures (Fig. 2C) but do not contribute to the effect of HG in the sympathetic neurovascular cultures (Fig. 4B). In addition, the present study found that inhibition of α2-adrenergic receptors did not affect NE release from isolated rat femoral arteries, which suggests that ROS-dependent inhibition of presynaptic α2-adrenergic receptors did not mediate the HG-induced increase in NE release from these arteries (Fig. 6B). Potential mechanisms underlying the effects of HG in the sympathetic neurovascular cultures and in the isolated femoral arteries are currently under investigation.
Elevated glucose also inhibits ATP-sensitive potassium (KATP) channels (26, 28), which if present could depolarize the nerves and increase NE release (7). Studies in rat tail artery suggest that this is not a primary mechanism underlying the glucose-induced increase in NE release shown in Fig. 6B. Acute (1 h) elevations of glucose, which would inhibit KATP channels, did not affect stimulated NE release from isolated rat tail arteries (P > 0.05; n = 4; data not shown).
It is well established that hyperglycemia has detrimental effects on many neurons and that neuropathy is a debilitating complication of diabetes (15, 34, 36, 38). Many studies have indicated that vascular sympathetic nerves remain functional in patients and animals with hyperglycemia and diabetes. Inhibition of NE binding to α-adrenergic receptors decreases blood pressure (8) in patients with diabetes. This indicates that in these patients, NE released by functioning postganglionic neurons, including vascular neurons, affects blood pressure and blood flow. Morphological analyses in humans and rats indicate that some, but not all, postganglionic sympathetic neurons are susceptible to hyperglycemia (30, 31, 33, 34). Vascular sympathetic neurons were not specifically identified in these studies, but additional evidence indicates that the function of this subset of postganglionic sympathetic neurons is maintained in hyperglycemic animals. Frisbee (18) reported enhanced activity of vascular sympathetic nerves in skeletal muscle of hyperglycemic obese Zucker rats, and Hart et al. (20) reported that arteries from hyperglycemic STZ-treated rats contained less but took up and released more NE than arteries from normoglycemic rats. Speirs et al. (35) reported that 3 mo of hyperglycemia did not affect the density or function of sympathetic nerves innervating rat tail arteries. Martinez-Nieves and Dunbar (25) found that acute inhibition of NE binding to α-adrenergic receptors increased blood flow in femoral arteries of rats that had been hyperglycemic for 5–6 wk, suggesting that NE was being released from postganglionic sympathetic neurons innervating these arteries. The persistence of sympathetic neurovascular function in diabetes suggests that perivascular sympathetic nerves are resistant to the detrimental effects of hyperglycemia. Little is known about the effects of hyperglycemia on perivascular sympathetic nerves and the mechanisms underlying the potential resistance of these nerves to hyperglycemia. The present studies demonstrate that glucose is likely to have direct and vascular-dependent effects on perivascular sympathetic nerves and suggest that vascular-dependent mechanisms maintain or enhance sympathetic neurovascular transmission. A recent study suggested that diabetes depresses synaptic transmission in sympathetic ganglia (9). Vascular-dependent increases in NE release may act to oppose decreased ganglionic transmission and thereby maintain sympathetic neurovascular transmission. Additional in vivo studies are warranted to further investigate the effects of hyperglycemia on sympathetic neurovascular transmission and how these contribute to the vascular and cardiovascular complications associated with diabetes.
This work was supported by National Heart, Lung, and Blood Institute Grant HL076774.
No conflicts of interest, financial or otherwise, are declared by the author(s).
I gratefully acknowledge the expert technical assistance of Rachel Poole.
- Copyright © 2011 the American Physiological Society