INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

A MAIN EFFECT OF cardiovascular maturation during the immediate postnatal period is a reduction in the dependence on circulating amines typical of fetal life, counterbalanced by increased reliance on neuronally released amines as is characteristic of a mature circulation (10). Although the pace and character of these changes are species specific, depend on the neurotransmitter involved, and vary among different vascular beds (6, 11, 21), all mammals undergo dramatic postnatal alterations in vascular structure and composition that lead to improvements in contractile capacity (29), stiffness (36), ion balance (20), and autonomic innervation (14, 15). In turn these changes enable enhanced reactivities to autonomic neurotransmission (15, 29) and reflex baroreceptor control (32), which are essential for postnatal cardiovascular homeostasis.

Whereas it is clear that cardiovascular autonomic function develops rapidly after birth in most species (1, 15, 40), this maturation clearly involves much more than proliferation of nerve fibers. In many vascular beds, reactivity to autonomic agonists appears long before the perivascular innervation becomes functional (3, 10, 14). Indeed, in many cases, reactivity to autonomic agonists such as norepinephrine (NE) is greater in immature than in mature arteries from the same vascular bed (19, 38, 44), suggesting that immature adrenergic receptors, by virtue of their lack of direct innervation, may exhibit a perinatal version of denervation supersensitivity (31). In some vascular beds, the ability of nonneuronal tissues to inactivate neuronally released transmitters also appears before the arrival of a functional innervation (3, 8). Autonomic perivascular nerve fibers generally first appear as a longitudinal network with few varicosities (12), then later develop into the circular meshwork pattern with higher node densities typical of adult arteries (12, 24). Commensurate with the proliferation of the sympathetic perivascular innervation, neuronal reuptake of NE becomes significant (14, 15), neuronal NE content rises (3, 10, 15), and, not long thereafter, nerve stimulation is able to release neuronal transmitter stores (3, 10, 14). A final step in perivascular autonomic maturation involves increased neuronal activation under both basal conditions (21, 33) and during reflex responses (27).

Although the general features of perivascular autonomic maturation appear well established, the precise timing and importance of the events involved are highly variable. In contrast to the general picture given, for example, in some vascular beds the NE content (14), monoamine oxidase activity (14), overall sympathetic nerve density (22), sympathetic nerve activity (33), contractile responses to transmural nerve stimulation (35, 37), and/or hypotensive effects of ganglionic blockade (10) may be greater in immature than in mature animals of the same species. Given this variability, any clear picture of the pattern of autonomic maturation involved in a given vascular bed requires multiple separate assessments of the mechanisms potentially involved.

The present studies were designed to assess the pattern and mechanisms involved in maturation of the cerebrovascular sympathetic innervation. In adults, this innervation has important vasomotor influences (9, 35, 44) that can contribute significantly to cerebral autoregulation at elevated arterial pressures (9). The potential contribution of cerebrovascular sympathetic nerves to autoregulation in the fetus and neonate is unknown, although it is clear that autoregulatory capacity is less efficient in the immature than in mature cerebral circulation (42) and may be influenced by sympathetic denervation (28). To better characterize the effects of maturation on the capacity for sympathetic cerebral vasoconstriction, the present studies addressed the general hypothesis that cerebrovascular maturation alters adrenergic neurovascular coupling through changes in both the presynaptic release of and the postsynaptic reactivity to NE. To directly test this hypothesis we examined neuronal NE content, NE uptake capacity, stimulation-induced NE release, and postsynaptic reactivity to both exogenous and neuronally released NE in fetal and adult sheep cerebral arteries. This parallel assessment of these effects of maturation was intended to provide a unique and unprecedented view of the adrenergic processes involved in ovine cerebrovascular maturation.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

All protocols and procedures used in these studies were reviewed and approved by the Animal Research Committee of Loma Linda University. Pregnant and nonpregnant ewes of mixed breed were obtained from a single supplier (Nebeker Ranch, Lancaster, CA). Pregnant animals were killed after 138-142 days gestation by administration of 100 mg/kg intravenous pentobarbital sodium into the jugular vein, and fetuses were immediately delivered by cesarean section. Nonpregnant adults were similarly killed. After death, the brains were immediately removed both from fetuses and adults and placed in ice-cold Krebs solution with subsequent dissection of the middle cerebral arteries. The Krebs solution contained (in mM) 118 NaCl, 4.8 KCl, 1.5 CaCl₂, 1.2 KH₂PO₄, 25 NaHCO₃, 1.2 MgSO₄, 0.3 ascorbic acid, and 11.5 glucose.

Protocol 1: responses to transmural stimulation and NE. After removal of excess adipose and connective tissue, we cut the arteries into individual 1- to 3-mm-long ring segments. To avoid endothelial influences, we removed the endothelium via mechanical abrasion by carefully passing a small needle through the lumen several times. We then mounted the segments on paired Tungsten wires (OD 250 µm) between low-compliance force transducers (Kulite BG-10) and posts attached to micrometers used to vary resting tension. We equilibrated the arteries at normal ovine core temperature for at least 30 min in baths containing a bicarbonate Krebs solution with (in mM) 120 NaCl, 5.56 dextrose, 25.6 NaHCO₃, 5.17 KCl, 2.49 MgSO₄, and 1.60 CaCl₂ with 114 µM ascorbic acid and 27 µM disodium EDTA that was continuously bubbled with 95% O₂-5% CO₂. All segments were equilibrated at internal diameters of 1.1 (fetal) and 1.3 (adult) mm, which were previously shown to be optimal for our preparation (30). Contractile tensions from all artery segments were continuously recorded using online computers that also digitized and normalized the data.

After 30 min of equilibration at optimum stretch, we contracted all arteries with a potassium-Krebs solution similar to that described with the exception of NaCl having been eliminated and exchanged for KCl on an equimolar basis yielding a final potassium ion concentration of 120 mM. The arteries were then contracted with 10 µM serotonin (5-HT). Once agonist-induced tone stabilized, the arteries were exposed to 10 µM acetylcholine to verify successful removal of the endothelium. Artery segments exhibiting relaxations to acetylcholine greater than 10% of control contractions were discarded.

After verification of endothelial denudation, the arteries were equilibrated for another 30 min in normal Krebs to which we added 1 µM atropine to block involvement of muscarinic receptors, 10 µM capsaicin to deplete sensory peptidergic neurons, and 10 µM nitro-L-arginine methyl ester (L-NAME) to block possible influences from nitric oxidergic innervation. During this incubation, the bath solution was changed at 10-min intervals to eliminate the possible accumulation of any released products in the bath. At the end of the incubation period, the arteries were stimulated at 8 Hz in 30-s bursts using square wave pulses of 0.3-ms duration delivered by a Grass S-44 stimulator (Grass Instruments, Quincy, MA). The output of the stimulator was connected in parallel to eight individual constant-current amplifiers, one for each vessel bath, which in turn delivered current to platinum wire electrodes placed in parallel on both sides of each mounted arterial segment. The current delivered to each artery segment was measured as a voltage drop across a fixed-series resistance displayed on an oscilloscope and was titrated individually for each artery segment to determine both the threshold and optimal (lowest current that produced a maximal contractile response) currents for stimulation. Once optimal currents were determined and set for each artery segment, all arteries were stimulated simultaneously at 1, 2, 4, 8, and 16 Hz in 45-s bursts of 0.3-ms duration. A rest period of 5 min was allowed between each 45-s burst of stimulation. Once the final 16-Hz stimulation had been delivered, the arteries were incubated for 30 min in normal Krebs containing (in µM) 2 guanethidine, 0.1 propranolol, 0.1 cocaine, 10 capsaicin, 1 atropine, and 10 L-NAME. Then, the arteries were again stimulated at 8 Hz for 45 s at a duration of 0.3 ms. Next, 0.3 µM TTX was added to the baths and 15 min later the arteries were again stimulated at 8 Hz for 45 s with a pulse duration of 0.3 ms. After this final stimulation, the arteries were washed, rested for 10 min in normal Krebs containing 0.1 µM propranolol and 0.1 µM cocaine, then exposed to cumulative increasing concentrations of NE added in one-half-log increments from 10⁹ to 10^3.5 M.

Protocol 2: NE content and release measurements. Segments of middle cerebral arteries, 3-4 cm in length, were dissected and cannulated at both ends with polyethylene tubing and placed in a low-volume in vitro perfusion system as previously described (7). The entire perfusion assembly was immersed in a circulating water bath and kept at 37°C. At the beginning of the experiment tissues were perfused for 1 h with aerated (95% O₂-5% CO₂) Krebs solution containing 10⁵ M cocaine and 10⁵ M deoxycorticosterone to inhibit neuronal and extraneuronal NE uptake, respectively.

Electrical field stimulation was delivered by a Grass S-48 stimulator through two platinum electrodes placed at either end of the tissue chamber. The parameters for excitation were 8 Hz, 60 V, 1-ms duration, and 480 pulses (1-min stimulation). Perivascular nerves were activated two times in succession for 1 min each with a 30-min equilibration period between stimulations. After the second activation, the tissues were exposed to TTX (1 µM) for 30 min, and once again the nerves were activated for 1 min. The perfusate was collected continuously starting at the beginning of each stimulation period until 5 ml were collected. Basal NE release was monitored by a collection of 5 ml of perfusate after each treatment. At the end of each experiment, arteries were homogenized in 3 ml of 0.1 N perchloric acid. After centrifugation, the supernatant was saved for determination of NE content as previously described (17).

Perfusates and tissue homogenate supernatants were extracted and quantified with dihydroxybenzylamine (DHBA) as an internal standard (300 pg), following a previously described protocol (5). A 100-µl sample of extracted amines was injected into an ESA coulochem II high-performance liquid chromatograph (ESA, Bedford, MA) and separated on an ESA reverse-phase C-18 column with an aqueous mobile phase from ESA (MD-TM) that contained (in mM) 75 Na₂H₂PO₄, 0.5 sodium dodecyl sulfate, and 0.025 EDTA with 20% acetonitrile and 5% methanol. The following formula was used to calculate the amount of NE in the injected sample: pg NE = (NE peak height sample/peak height NE standard) × 100 pg DHBA × (peak height DHBA standard/DHBA peak height sample). Recoveries varied from 85-95%. Fractional release was calculated as follows: fractional NE release = pg NE released/(pg NE tissue content × number of stimulation pulses).

Protocol 3: NE uptake. Middle cerebral artery segments cleaned of excess adipose and connective tissue were cut in half, and each half was placed in a separate bath containing 50 ml of Krebs bubbled and equilibrated as described in Protocol 1: responses to transmural stimulation and NE. In one bath, 0.1 µM cocaine was present to inhibit neuronal NE uptake, and in the other bath cocaine was absent. After 15 min of incubation, tritiated NE of known specific activity was added to both baths in a final concentration of 0.2 µM. One hour later, the arteries were removed, quickly rinsed in a solution of cold Krebs containing 0.1 µM unlabeled NE, weighed, then digested overnight in 1.5 ml ammonium hydroxide in toluene (Beckman BTS-450). The following day, the samples were added to 10 ml of toluene-based scintillation cocktail and counted until a minimum of 10,000 counts/min had accumulated. Neuronal NE uptake was calculated as the difference in disintegrations per minute between the untreated and cocaine-treated samples divided by the specific activity of the labeled NE.

Data analysis and statistics. All contractile data were normalized relative to the maximal response to potassium. Frequency-response data were fitted to a rectangular hyperbola using nonlinear regression to obtain estimates of S_max (the maximal response to stimulation) and F₅₀ (the frequency at which 50% of the maximal response to stimulation was observed). Concentration-response data were fitted to the logistic equation using nonlinear regression to obtain pD₂ (log EC₅₀) and E_max (maximal contractile response) values. Age-related differences in response to transmural stimulation and exogenous NE were determined using a repeated-measures ANOVA followed by post hoc comparisons performed with a Duncan's analysis. All other age-related differences were determined using an unpaired Student's t-test. In cases in which no significant difference was observed, a power analysis was performed.

Drugs used. Atropine, capsaicin, cocaine hydrochloride, deoxycorticosterone, dihydroxybenzylamine, guanethidine monosulfate, L-NAME, NE, propranolol, and TTX were all obtained from Sigma Chemical Company (St. Louis, MO).

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

A total of 57 fetal and 45 adult middle cerebral artery segments was taken from 37 term lambs and 29 adult sheep, respectively. When multiple segments were taken from the same animal, the resultant values were averaged into a single value such that each animal contributed equally to the means and standard errors reported. All reported values of n refer to the number of animals used, and statistical significance implies P < 0.05 unless otherwise stated.

Responses to transmural stimulation. In the adult middle cerebral artery segments mounted for study of responses to electrical stimulation, maximum responses to 120 mM potassium averaged 3.43 ± 0.13 g (n = 6; Fig. 1). In response to transmural electrical stimulation, the adult segments contracted significantly at all frequencies applied, and the magnitudes of these contractions were related to stimulation frequency. At 16 Hz, the maximum response averaged 9.5 ± 3.7% of the maximum response to 120 mM potassium. When the adult frequency-response relationship was fitted to a rectangular hyperbola, the curve of best fit had an S_max of 23.1% of the maximum response to 120 mM potassium and an F₅₀ of 24.0 Hz.

View larger version (17K): [in this window] [in a new window]   Fig. 1.   Effects of age on responses to transmural stimulation. Transmural stimulation of middle cerebral artery segments at supramaximal current and 0.3-ms duration in presence of 10 µM capsaicin, 1 µM atropine, and 10 µM nitro-L-arginine methyl ester produced frequency-dependent increases in contractile tension that are expressed relative to maximum responses to 120 mM potassium Krebs. Across all frequencies (ANOVA), responses to stimulation were significantly greater in adult than in fetus. * Significant age-related differences at each frequency, as indicated by post hoc Duncan's analyses. All responses to stimulation were blocked by pretreatment with either 2 µM guanethidine or 0.3 µM TTX. Vertical error bars indicate SE for artery segments from 10 fetuses and 6 adults.

In fetal middle cerebral artery segments, maximum responses to 120 mM potassium averaged 1.41 ± 0.35% (n = 10), which was significantly less than those observed in adult segments. Transmural stimulation produced significant contractions at all frequencies applied, but the magnitudes of these contractions were quite small and, at frequencies of 4 Hz, were significantly less than those observed in the adult segments. At 16 Hz, the maximum response averaged 1.1 ± 0.6% of the maximum response to 120 mM potassium. When the fetal frequency-response relationship was fitted to a rectangular hyperbola, the curve of best fit had an S_max of only 1.25% of the maximum response to 120 mM potassium and an F₅₀ of 3.0 Hz. In both the fetal and adult segments, pretreatment with either 2 µM guanethidine or 0.3 µM TTX completely eliminated all responses to transmural stimulation.

NE concentration- response relationship. In the presence of 0.1 µM propranolol and 0.1 µM cocaine, cumulative additions of NE produced concentration-related contractions of which maximum magnitudes averaged 106 ± 20% and 67 ± 13% of the maximum responses to 120 mM potassium in fetal (n = 6) and adult (n = 9) middle cerebral artery segments, respectively (Fig. 2). The magnitudes of the responses to NE were significantly greater in the fetus than in the adult at concentrations of 10 µM. In contrast, pD₂ values for NE did not vary significantly with age and averaged 6.11 ± 0.12 in the fetus (n = 6) and 6.33 ± 0.09 in the adult (n = 9).

View larger version (18K): [in this window] [in a new window]   Fig. 2.   Effects of age on norepinephrine (NE) concentration-response relationship. Cumulative additions of NE in presence of 0.1 µM propranolol and 0.1 µM cocaine produced concentration-related contractions that had a greater maximum magnitude in fetal (106 ± 20%) than in adult (67 ± 13%) middle cerebral artery segments. * Significant age-related differences in contractile response as detected by post hoc Duncan's analyses. Corresponding pD2 values averaged 6.11 ± 0.12 M in fetus and 6.33 ± 0.09 M (n = 9) in adult. These values were not significantly different. Vertical error bars indicate SE for 6 fetuses and 9 adults.

NE content and release. Maturation had no significant effect on total NE content, which averaged 32.4 ± 5.0 and 32.5 ± 3.9 ng/mg wet wt in fetal (n = 17) and adult (n = 13) middle cerebral arteries, respectively (Fig. 3). In contrast, maturation significantly depressed basal release as well as stimulation-evoked mass and fractional NE release. Basal release averaged 0.0045 ± 0.0009 and 0.0024 ± 0.0005 pg/mg wet wt in fetus (n = 17) and adult (n = 13), respectively. Stimulation-evoked mass NE release averaged 883 ± 184 and 416 ± 106 pg/mg wet wt in the fetus (n = 17) and the adult (n = 13), respectively. Corresponding values for stimulation-evoked fractional release were 51.1 ± 5.3 and 22.8 ± 3.8 pg/pg NE content per pulse × 10⁶ for fetus (n = 17) and adult (n = 13), respectively.

View larger version (16K): [in this window] [in a new window]   Fig. 3.   Effects of age on NE content and release. Maturation had no significant (ns) effect total NE content but significantly reduced both total mass of NE released during stimulation and fraction of NE released per pulse. Data indicated represent means ± SE for 10 fetuses and 6 adults. * P < 0.05 as indicated by Student's t-test.

NE uptake. As observed for NE content, maturation also had no significant effect on cocaine-sensitive NE uptake in ovine middle cerebral artery segments (Fig. 4). Uptake averaged 1.55 ± 0.40 and 1.84 ± 0.51 pmol/mg wet wt. in the fetus (n = 10) and the adult (n = 7), respectively.

View larger version (12K): [in this window] [in a new window]   Fig. 4.   Effects of age on NE uptake. Maturation had no significant (ns) effect on cocaine-sensitive NE uptake. Data shown indicate means ± SE of mass of NE accumulated in 30 min measured in 10 fetal and 7 adult arteries.

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

The mature cerebral circulation receives a rich sympathetic innervation in numerous species including the mouse (24), rat (1, 40), and human (12). Stimulation of this innervation can modestly vasoconstrict adult cerebral arteries (35, 44), and in the present study the magnitude of this response averaged ~10% of the maximum tension produced by complete potassium depolarization. The finding that both 0.3 µM TTX and 2 µM guanethidine blocked these responses indicates that they were adrenergic in nature. Because stimulation currents were individually optimized for each segment studied, the magnitudes of the contractions observed most probably represent maximum responses. Finally, because these responses were obtained in endothelium-denuded arteries in the presence of 1 µM atropine, 10 µM capsaicin, and 10 µM L-NAME, it is most unlikely that the release of acetylcholine, sensory neuronal peptides, or nitric oxide contributed to these responses.

In contrast to the adult arteries, the fetal arteries responded poorly to transmural stimulation and produced significantly less tension than adult arteries at all stimulation frequencies employed (Fig. 1). This result is consistent with a wide variety of studies in which vascular responses to nerve stimulation generally improve with maturation (10) but contradicts other studies in dog coronary (37) and mesenteric (39) arteries, rat tail arteries (34), and ovine renal arteries (16, 31) in which responses to nerve stimulation declined with maturation. Although this inconsistency may reflect differences in the rates of adrenergic maturation in different vascular beds, nerve stimulation-induced vasoconstriction has also been reported to be of greater magnitude in immature than in mature ovine pial (44) and monkey basilar and middle cerebral (35) arteries. Why these two latter studies yielded results opposite to those of the present study is uncertain, but most probably derives from the fact that only the present experiments were conducted in endothelium-denuded arteries depleted of neuronal peptides and treated with L-NAME to inhibit nitric oxide production. The experiments by Toda (35) in monkey cerebral arteries were performed on helical strips precontracted with 0.2-0.8 µM PGF_2a, which may have potentiated subsequent responses to neuronally released NE (2). The experiments by Wagerle et al. (44) were carried out in vivo using measurements of pial window diameter in response to superior cervical ganglion stimulation, a perturbation that can have multiple cerebrovascular effects.

One possible explanation for the age-related differences in responses to stimulation observed in the present studies could be that postsynaptic adrenergic receptor sensitivity to NE was less in fetal than in adult arteries. However, consistent with numerous previous studies of cerebral arteries from the dog (38), monkey (35), baboon (19), and sheep (13, 44), in the present study fetal arteries were no less sensitive to NE than were adult arteries (Fig. 2). This fact, combined with the observation that alpha adrenergic receptor affinity for NE may be greater in immature than in mature sheep cerebral arteries (13), strongly suggests that depressed postsynaptic reactivity to NE cannot explain why responses to transmural nerve stimulation were weaker in the fetal than in the adult arteries. Alternatively, these observations suggest that synaptic concentrations of NE during stimulation must have been greater in adult than in fetal arteries.

A key determinant of synaptic neurotransmitter concentration is neuronal release, particularly in preparations such as were used in the present study in which neuronal reuptake of transmitter was inhibited. In these experiments, the stimulation-induced release of endogenous NE, whether expressed as fractional release per pulse or as total mass released per stimulation train, was significantly greater in fetal than in adult arteries (Fig. 3). Interestingly, this significant age-related difference was observed even though total NE content was virtually identical in fetal and adult arteries. Together, these findings suggest that nerve stimulation in some way is more efficient at releasing NE in fetal than in adult arteries; either the number of vesicles released per pulse or the mass of NE per vesicle must be greater in fetal than in adult cerebral arteries. Consistent with either of these possibilities, a variety of studies suggest that NE content per nerve may actually be greater in immature than in mature sympathetic neurons due to the fact that immature neurons can synthesize the transmitter but cannot export it to axons and varicosities that are not yet fully developed (10, 12, 14, 15).

Aside from the neuronal release and reuptake of NE, the other main determinant of synaptic NE concentration is synaptic volume (4). If both total release and postsynaptic sensitivity to NE were greater, but the average synaptic NE concentration was less in fetal than in adult arteries, then clearly total synaptic volume must have been greater in fetal than in adult arteries. Because synaptic cleft width is a main determinant of synaptic volume, the data therefore suggest that synaptic cleft width may have been greater in immature than in mature ovine cerebral arteries. Alternatively, it remains possible that the total number of synapses per unit weight of tissue may have been greater in fetal than in adult arteries. In this case, because NE content per unit weight did not differ with age, a larger number of synapses could yield a greater total tissue synaptic volume and thus a lower average synaptic concentration of NE, even though total tissue release was greater in fetal compared with adult. To test this latter possibility, we measured NE uptake capacity as a measure of synaptic density in the arteries employed.

As shown in Fig. 4, neuronal NE uptake capacity and presumably synaptic density were not significantly different in fetal and adult middle cerebral arteries. Certainly, the use of uptake as an index of synaptic density assumes that the specific activity of each uptake carrier pump is the same and that the average concentration of the carrier is the same in each nerve. However, the effects of maturation on NE transport proteins, their biochemistry, and their distribution have not been studied in any detail. Nonetheless, the uptake findings are consistent with the general view that neuronal NE reuptake capacity is under hormonal control (8) and increases significantly (15) during the immediate postnatal period. Other studies further suggest that the formation of synapses is a relatively late event in adrenergic maturation (12, 26), particularly in the cerebral circulation (41). More importantly, the uptake results suggest that the total number of adrenergic synapses was probably similar in fetal and adult arteries, and, therefore, synaptic cleft width was most probably greater in fetal than in adult arteries.

Whereas the above interpretation assumes NE to be the sole transmitter released in response to nerve stimulation, it remains possible that corelease of another transmitter could have been involved. A variety of previous studies have suggested that other transmitters such as neuropeptide Y (NPY) (43), ATP (25), or nitric oxide (25, 45) may also be released during transmural nerve stimulation. Although the release of nitric oxide probably did not contribute to stimulation-induced responses owing to the presence of 100 µM L-NAME in the baths, possible contributions by NPY and ATP cannot be excluded. Although responses to stimulation were blocked by pretreatment with 2 µM guanethidine, indicating an adrenergic localization of the transmitter(s) involved, it is possible that ATP and/or NPY were colocalized within adrenergic preterminal vesicles. If this were the case, then the adrenergic release of one or both of these transmitters, or some other as yet unidentified vasoactive transmitter, must have been greater in adult than in fetal arteries to be consistent with the present results.

Perspectives