INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

AT LATE STAGES of embryonic development, acute changes in arterial oxygen tension alter heart rate and the distribution of cardiac output (13). This chemoreflex produces an increase in vagal nerve activity to the heart and an increase in sympathetic activity toward the periphery (13, 18, 21). Through humoral and neurogenic catecholamines this can lead to an elevation of peripheral arterial tone (12, 13). The strength of the peripheral arterial constriction in the embryo depends on the degree of maturation of the arterial smooth muscle as well as the supply of vasoconstrictor stimuli. During acute embryonic hypoxemia, the flow to vital organs such as the brain and heart is preserved at the expense of the peripheral circulation (17, 19, 25, 26). Increases in arterial tone may occur predominantly in peripheral tissues (12, 19, 26). Alternatively, hypoxemia may increase the conductance of the central vascular compartments through local vasodilator substances such as adenosine (22) and endothelium-derived nitric oxide (NO) (1). In the adult, the pharmacological control of the coronary and cerebral circulations differs extensively from that of peripheral vascular beds. It is, however, not clear at which developmental stage this regionality emerges and which mechanisms promote it. The chemoreflex may be subject to developmental changes (17). In fetal sheep, peripheral vasomotor control starts to develop at 0.7 gestation and becomes more prominent as development proceeds (10, 15, 19). Control of the cerebral and coronary circulations may develop earlier and may rely on local myogenic and metabolic mechanisms rather than on neurohumoral mediators (15).

The chicken embryo is an attractive model to study cardiovascular responses during fetal development (25, 30, 31). Unlike in mammalian species, cardiovascular responses in avian embryos are not influenced by maternal or placental vasoactive hormones. We previously demonstrated in 0.47- to 0.9-gestation chicken embryos that acute hypoxemia leads to bradycardia (30) and redistribution of cardiac output in favor of the heart and brain (26). We noted that, similar to results in fetal sheep, the fraction directed to the heart and brain increased with gestational age (26), indicating a maturation of vascular control mechanisms. This may be due to changes in the efficacy of the chemoreceptors and changes in the ability of peripheral arteries to regulate resistance, as well as developmental changes in sympathetic nervous and endothelium-dependent control of these peripheral arteries.

The goal of the present study was to describe development and regionality of arterial vasomotor control in the chicken embryo. A central artery, the carotid artery, and a peripheral artery, the femoral artery, were isolated from 0.7-, 0.8-, and 0.9-incubation chicken embryos (i.e., after 15, 17, and 19 days of the 21-day incubation period). Their responses to adrenergic stimuli, sympathetic nerve stimulation, and ACh were compared. The latter compound is generally considered to be an endothelium-dependent vasodilator (5, 7, 8, 10, 14), but this has not yet been demonstrated in the chicken embryo. The choice of carotid and femoral arteries was justified by their anatomic location, their size, and by the relative ease with which these vessels could be isolated from even the smallest embryos investigated.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Vessel isolation. Experimental procedures followed Dutch laws for animal experiments. Fertilized Lohman-selected White Legghorn eggs were incubated at 38°C and 60% relative air humidity and were rotated once every hour, until day 15, 17, or 19 (0.7, 0.8, or 0.9 incubation time, respectively). The eggs were opened at their blunt side. The eggshell and the outer eggshell membrane were carefully removed using forceps to expose the inner eggshell membrane, which was then superfused with Krebs-Ringer bicarbonate (KRB) solution. An incision of 2 cm was then made in the chorioallantoic membrane. The egg was turned upside down, and the embryo was collected on a petri dish. The extraembryonic membranes were removed, and the embryo was transferred to a petri dish that had been coated with Sylgard (Dow Corning) and filled with KRB. The right carotid and right femoral artery were carefully dissected from the embryo.

Recording of arterial reactivity. The isolated arteries were mounted as ring segments (length 1.7 to 2.0 mm) between an isometric force transducer (Kistler Morce DSC 6, Seattle, WA) and a displacement device in a myograph (model 610M, J.P. Trading, Aarhus, Denmark) using two stainless steel wires (diameter 40 µm). Force was divided by twice the vessel segment length to calculate wall tension. During mounting and experimentation, the myograph organ bath (5 ml vol) was filled with KRB maintained at 37°C and aerated with 95% O₂-5% CO₂. Each artery was stretched to its individual optimal lumen diameter, i.e., the diameter at which it developed the strongest contractile response to 125 mM K⁺-KRB (K-KRB), using a diameter-tension protocol as previously described for mammalian small arteries (Ref. 29; Fig. 1).

View larger version (13K): [in this window] [in a new window]   Fig. 1.   Typical tracing of isometric force vs. time illustrating contractile responses in a carotid artery of 0.9-incubation chicken embryo as a function of imposed distension. Isolated arterial segment was mounted in organ chamber, and its lumen diameter was increased in steps of 30 µm from 200 to 410 µm (arrows). At each level of distension, vessel was intermittently stimulated by exposure to 125 mM K+ [between K+ and w (wash)]. Note that maximal contractile response (Emax) was obtained at a diameter of 380 µm.

Morphometry. Fixed vessels were transferred to 70% ethanol and embedded in paraffin. To determine medial cross-sectional area, we performed immunohistochemistry on thin cross sections (4 µm) with antibodies against smooth muscle -actin (Sigma Chemical, St. Louis, MO). Endogenous peroxidase was blocked by 0.1% H₂O₂ in methanol for 20 min at room temperature, and the sections were incubated for 45 min with primary antibody (1:4,000). Horseradish peroxidase-conjugated rabbit anti-mouse antibodies (Daka; dilution 1:200) and diaminobenzioline-H₂O₂ substrate were subsequently used to visualize the immunoreactivity. The area occupied by smooth muscle -actin was determined using a Zeiss Axioscope (Zeiss, Germany), a standard CCD camera (Stemmer, Germany), and commercially available software (JAVA 1.21, Jandel Scientific, Corte Madera, CA).

Staining of perivascular sympathetic nerves. To demonstrate the presence of l-norepinephrine (NE)-containing nerves, we stained whole mount vessel preparations with glyoxylic acid (24, 29). Vessel segments were opened longitudinally and were incubated for 30 min at 20°C in 2% glyoxylic acid in 10% sucrose containing phosphate buffer (pH 7.2). Subsequently, the vessel segments were transferred to a mounting glass, air-dried for 30 min, stretched at 100°C for 4 min, and enclosed with Etellan and a coverslip. The presence of glyoxylic acid-induced fluorescence, representing catecholamine-containing nerves, was visualized using fluorescent microscopy (microscope Nikon Diaphot, BA 470-DM 455 filter; Nikon FE2 camera). Photographs of the vessel segments were taken through the microscope (objective Fluo ×10, Nikon) using Kodak 320 ASA film.

Solutions and drugs. The composition of KRB solution (in mmol/l) was as follows: 118.5 NaCl, 4.7 KCl, 1.2 MgSO₄ · 7H₂O, 1.2 KH₂PO₄, 25.0 NaHCO₃, 2.5 CaCl₂, and 5.5 glucose. In K-KRB (125 mmol/l K⁺) all NaCl was replaced by an equimolar amount of KCl. K⁺-containing solution (35 mmol/l) was prepared by mixing appropriate volumes of K-KRB and KRB. Phosphate-buffered solution consisted of 0.1 mol/l NaH₂PO₄ · H₂O and 0.1 mol/l Na₂HPO₄ · 2H₂O. ACh, glyoxylic acid, L-NAME, l-NE, l-phenylephrine (PE), and prazosin were all obtained from Sigma Chemical. BHT-933 (azepexol) was a generous gift from Boehringer. For all agents 1,000× concentrated stock solutions were prepared in double-distilled water.

Data analysis. Concentration-response curves were analyzed in terms of sensitivity and maximal response by fitting the individual experimental data with a nonlinear sigmoid regression curve (GraphPad Software, San Diego, CA). Maximal contractile responses (E_max) were expressed in terms of active wall tension (force divided by twice the segment length; N/m); sensitivities are shown as pD₂, where pD₂=  log₁₀ EC₅₀ (EC₅₀ represents the concentration at which 50% of the maximal responses were observed). Changes and differences in E_max and pD₂ with time and between types of vessels were evaluated by ANOVA with Bonferroni's correction for multiple comparisons. P < 0.05 was accepted to represent statistical significance. Data are shown as means ± SE.

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Femoral and carotid arteries obtained from 0.7- to 0.9-incubation chicken embryos responded to depolarizing high-K⁺ solution with a contraction (Fig. 1). In both types of vessels, the diameter at which maximal responses were obtained (D_opt) and the amplitude of the maximal response (K_max) increased significantly with increasing incubation (Figs. 2 and 3). In femoral arteries, D_opt averaged 357 ± 8, 409 ± 13, and 449 ± 15 µm and K_max averaged 0.26 ± 0.03, 0.73 ± 0.15, and 1.78 ± 0.14 N/m at 0.7, 0.8, and 0.9 incubation, respectively (n = 10-18). In carotid arteries, D_opt averaged 368 ± 8, 406 ± 8, and 468 ± 15 µm; and K_max averaged 0.19 ± 0.02, 0.21 ± 0.04, and 1.01 ± 0.15 N/m at 0.7, 0.8, and 0.9 incubation, respectively (n = 8-13). Despite comparable diameters (Fig. 2), contractile responses to K⁺ were significantly larger in femoral than in carotid arteries at 0.8 and 0.9 incubation (Fig. 3).

View larger version (19K): [in this window] [in a new window]   Fig. 2.   Lumen diameter and medial cross-sectional area (CSA) of chicken embryonic arteries as a function of gestational age. A: diameter at which Emax to 125 mM K+ could be obtained is shown for isolated femoral arteries (Fem) and carotid arteries (Car). B: medial CSA was determined on anti-smooth muscle -actin-stained cross sections. Values are means ± SE (number of arteries: n = 8-18). *Differences between both types of arteries were not statistically significant (P < 0.05).

View larger version (15K): [in this window] [in a new window]   Fig. 3.   Contractile reactivity of chicken embryonic arteries as function of incubation. Emaxto 125 mM K+ (A), norepinephrine (NE, B), and phenylephrine (C) are shown for isolated Fem and Car. Values are means ± SE (n = 8-18). *Difference from Fem is statistically significant (P < 0.05). Data analysis was restricted to arterial segments that exhibited K+-induced contractions exceeding the detection limit (0.1 N/m).

Part of the reason for these results seems to be the significantly larger medial mass in femoral than in carotid arteries (Fig. 2), but statistically significant differences between both vessels persist when contractility was normalized to medial thickness. Active wall stress averaged 69 ± 7 and 41 ± 5 N/m² at 0.8 incubation and 142 ± 11 and 85 ± 9 N/m² at 0.9 incubation in femoral and carotid arteries, respectively.

In femoral and carotid arteries at 0.7, 0.8, and 0.9 incubation, high concentrations of relaxing agonists, such as ACh (10 µmol/l), adenosine (1 mmol/l), and papaverine (1 mmol/l) failed to affect basal tension, indicating the absence of spontaneous tone under the experimental conditions.

In all femoral arteries from 0.7 incubation onwards, and in all carotid arteries at 0.9 incubation, NE and the selective ₁-adrenergic agonist PE induced concentration-dependent tonic contractile responses (Figs. 3 and 4). At 0.7 and 0.8 incubation, only 30-40% of the carotid arteries investigated responded to the agonists; maximal responses were close to the detection limit of 0.01 N/m. At all time points, E_max to the adrenergic stimuli were significantly larger in femoral than in carotid arteries (Fig. 3). At 0.9 incubation, the sensitivity (pD₂) for NE was significantly smaller in femoral (6.03 ± 0.8) than in carotid arteries (6.53 ± 0.11), whereas the sensitivity for PE was comparable in femoral (5.58 ± 0.09) and carotid arteries (5.27 ± 0.17). Both femoral and carotid arteries failed to contract in response to BHT-933, a selective ₂-adrenergic agonist (data not shown), indicating a lack of functional ₂-adrenergic receptors on arterial smooth muscle at the developmental stages investigated.

View larger version (15K): [in this window] [in a new window]   Fig. 4.   Contractile responses of 0.9-incubation chicken embryonic Fem to electrical field stimulation of perivascular nerves (A) and exogenously supplied NE (B). Values are means ± SE (n = 10).

Electrical field stimulation caused frequency-dependent contractions in femoral arteries at 0.9 incubation (Fig. 4) but failed to induce contraction in femoral arteries at earlier developmental stages and in carotid arteries at all stages investigated. Glyoxylic acid staining confirmed the presence of catecholamine-containing nerves in femoral arteries at 0.9 incubation (Fig. 5). In the presence of the ₁-adrenergic receptor antagonist prazosin (0.1 µmol/l), contractile responses of 0.9-incubation femoral arteries to electrical field stimulation were reduced by >85% (data not shown).

View larger version (128K): [in this window] [in a new window]   Fig. 5.   Glyoxylic acid-induced fluorescence in a whole mount preparation of 0.9-incubation chicken embryonic Fem highlighting the presence of NE-containing perivascular nerves. Calibration bar: 280 µm.

Embryonic arteries that had been constricted with 35 mmol/l K⁺ responded to ACh with concentration-dependent relaxations (Fig. 6). In 0.9-incubation femoral arteries, the relaxing responses to ACh were abolished by mechanical or chemical removal of the endothelium (Fig. 6). Sensitivity to the cholinergic agonist and its maximal effect did not change significantly between 0.7 and 0.9 incubation and did not differ significantly between femoral and carotid arteries (Fig. 7). L-NAME (0.1 mmol/l) increased the contractile responses to 35 mmol/l K⁺ in both femoral and carotid arteries (Fig. 6) and reduced the sensitivity and the maximal responsiveness to the endothelium-dependent relaxing effects of ACh in both types of vessels (Fig. 7). Although K⁺-induced contraction was both on an absolute and a relative basis stronger in femoral than in carotid arteries, the L-NAME-resistant component of the relaxing effects of ACh was significantly more pronounced in carotid than in femoral arteries at 0.7 and at 0.9 incubation (Fig. 7). It is noteworthy that the contractile effect of L-NAME in embryonic arteries was not significantly altered by endothelium removal (Fig. 6).

View larger version (16K): [in this window] [in a new window]   Fig. 6.   Typical tracings of isometric tension vs. time illustrating effects of 35 mM K+, ACh, and NG-nitro-L-arginine methyl ester (L-NAME) in intact [+endothelium (+E); top] and denuded [endothelium (E); bottom] segments of 0.9-incubation chicken embryonic Fem. Removal of endothelium was obtained by perfusion with 0.1% Triton X-100. Concentrations of pharmacological agents are shown as log M. Horizontal bar: 5 min; vertical bars: 1 N/m.

View larger version (29K): [in this window] [in a new window]   Fig. 7.   Concentration-response curves illustrating relaxing effects of ACh in Fem (left) and Car (right) obtained from 0.7- (A), 0.8- (B), and 0.9-incubation (C) chicken embryos. Experiments were performed in absence and in presence of 0.1 mmol/l L-NAME. Data shown as means ± SE (n = 6-8). Note that L-NAME reduced responses to ACh to a larger extent in Fem than in Car.

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Between 0.7 and 0.9 incubation, the contractile reactivity to ₁-adrenergic and receptor-independent stimulation increased in the femoral and carotid arteries of chicken embryos. Late developmental increases in contractile reactivity, in neurogenic vasoconstrictor responses, and in the responsiveness to endothelium-derived NO were more pronounced in the peripheral femoral artery than in the central carotid artery. Such regional heterogeneity may contribute to the redistribution of cardiac output during hypoxemia in the embryo and may find its origin in neurogenic and endothelial influences on the functional differentiation of arterial smooth muscle.

Previous studies of cardiovascular responses in chicken embryos revealed similarities to those in fetal sheep (17, 26). Most notably, acute hypoxemia results not only in bradycardia but also in a redistribution of cardiac output from the peripheral circulation to the heart and brain in both species. Catecholamines participate in this embryonic chemoreflex in both systems (12, 15, 17, 18, 20, 21, 25, 26, 30). In the present study we analyzed isolated arteries of the chicken embryo. The use of in vitro approaches, which are well established for adult arteries (e.g., 7, 8, 29), allow quantification of responses and sensitivities of individual blood vessels to various vasoactive agents and allow us to quantify neurogenic and endothelium-dependent responses in the absence of modulatory circulating hormones. We compared femoral and carotid arteries as model systems for a peripheral and a central vascular bed. The size and reactivity of the youngest arteries that we studied were only moderately above the limits of the in vitro techniques that are currently available. Vessels further downstream were therefore not included in this study.

In femoral and carotid arteries obtained from 0.7-incubation embryos, direct depolarization induced significant contraction. This indicates that at this developmental stage the arterial smooth muscle cells are already equipped with contractile proteins and an excitation-contraction coupling that most likely involves voltage-operated calcium channels. In both types of vessels, the receptor-independent contractile responses to depolarization increased five- to sevenfold between 0.7 and 0.9 incubation, whereas medial mass, as judged from measurements of medial cross-sectional area, increased only 65-75%. The contractile responses were larger in femoral than in carotid arteries despite comparable lumen diameter, and significant differences persisted after correction for differences in medial thickness. It is most likely that smooth muscle cell differentiation contributes to the developmental increase in arterial contractile reactivity in general and to the difference between femoral and carotid arteries as regards contractile strength in particular.

At all time points investigated, NE and PE induced contraction in the embryonic femoral artery. Adrenergic responsiveness and pharmacomechanical coupling thus develop rather early in the chicken embryo. Previous research (11) established that cardiac adrenergic responses can be observed as early as day 4 of incubation. In sharp contrast, consistent ₁-adrenergic contractile responses could not be obtained in the carotid arteries before 0.9 incubation. With the use of 0.3 nM [³H]prazosin and previously described ligand-binding techniques in intact arterial segments (28), we observed that the density of ₁-adrenergic receptors was comparable in 0.9-incubation femoral arteries (10.8 ± 1.3 fmol/mg total protein, n = 8) and carotid arteries (11.1 ± 1.8 fmol/mg total protein, n = 4). Consequently, aspects beyond the receptors, most likely involving the coupling of these sarcolemmal structures to the contractile apparatus, account for the regional difference. Little is known about the mechanisms that control ₁-adrenergic mechanisms in general and during development in particular. The observed difference between the two types of arteries excludes a major role for circulating factors such as glucocorticoids and catecholamines. Recent findings of our group suggest, on the other hand, that local aspects of smooth muscle cell differentiation and of perivascular sympathetic innervation play a pivotal role in this respect. We demonstrated in adult arteries that dedifferentiation of arterial smooth muscle resulting from balloon injury is accompanied by a marked reduction of ₁-adrenergic receptor density (3) and that the perivascular sympathetic innervation promotes the presence of _1A-adrenergic receptors while reducing the density of _1B- and _1D-adrenergic receptors (28).

In this study we approached perivascular sympathetic nerves histochemically and from a functional point of view. Glyoxylic acid-induced fluorescence of perivascular nerve fibers was prominent in late-incubation chicken femoral arteries. Prazosin-sensitive contractile responses to perivascular nerve stimulation were obtained in late-gestation femoral arteries but not in carotid arteries. In the femoral arteries, constrictor responses to exogenous NE could be obtained before neurogenic responses. In adult arteries, the density of perivascular sympathetic innervation varies considerably between anatomic locations (28). Regionally selective vascular sympathetic innervation during development is brought about by timely secretion by vascular smooth muscle of a mixture of nerve-attracting mediators such as nerve growth factor and nerve-repelling substances. This secretion is restricted to a rather limited period of time and may be limited to an intermediate vascular smooth muscle cell phenotype (9, 16). It may furthermore be more prominent in vascular smooth muscle cells of mesodermal origin than in those that are derived from the neural crest and which primarily populate the blood vessels in the cranial region. In addition to altering the pharmacological properties of the innervated blood vessel (e.g., see 29), the perivascular sympathetic innervation can initially stimulate growth and proliferation of the smooth muscle cells and subsequently promote the development and maintenance of a contractile phenotype (for review, see Ref. 6). We thus propose that the stronger contractility of femoral arteries during late development and their larger responsiveness to ₁-adrenergic stimulation are related to their sympathetic innervation. The causal interrelationship between these aspects clearly remains to be established.

Another vessel wall component that plays important roles in the morphogenesis and development of the vascular system and in the control of vasomotor tone is the endothelium. Endothelial cells give rise to the earliest vascular channels and later attract mesenchymal cells to the vessel wall (4). Subsequently, endothelium-derived mediators promote the differentiation of these vessel wall cells into contractile smooth muscle cells (4). Endothelium-derived NO that stimulates cGMP production and protein kinase G activity has been proposed to participate in this differentiating action (2, 23). We approached this mediator with the use of ACh, an agent that induces endothelium-dependent relaxation in adult arteries of various species (5, 7, 8, 10), including chicken aorta (14). Our experiments with mechanical and chemical endothelium removal demonstrate the endothelium dependency of cholinergic vasodilatation in the chicken embryo. Partial blockade of these relaxations by the NOS inhibitor L-NAME indicates an involvement of endothelium-derived NO. Relaxing effects of endothelium-derived NO could be demonstrated from the earliest developmental stages investigated and were at all stages more prominent in femoral arteries than in carotid arteries. Our current findings do not allow us to attribute the regional difference to a larger endothelial release of NO or to a more elaborated guanylate cyclase-protein kinase G system in femoral than in carotid arteries. Yet, the findings are consistent with a contribution of endothelium-derived NO to the differentiation of arterial smooth muscle (2, 23) and to the developmental increase in arterial contractile reactivity. Future pharmacological intervention studies are, however, needed to strengthen this proposal.

It is noteworthy that in chicken embryonic arteries, as in some adult mammalian systems (5) and in the aorta of mature chickens (14), the endothelium-dependent relaxing responses to ACh could only partly be blocked by L-NAME. A role for an endothelium-derived hyperpolarizing factor (5) seems unlikely because the relaxing effects of ACh were studied during contraction induced by depolarizing high-K⁺ solution. The exact nature of the L-NAME-resistant component of endothelium-dependent relaxation remains to be established in chicken embryonic arteries. This also applies for the observed endothelium-independency of the L-NAME-induced contraction.

In summary, we observed differences between carotid and femoral arteries of chicken embryos as regards the developmental increase in contractile strength, ₁-adrenergic vasoconstriction, perivascular sympathetic innervation, and endothelium-dependent vasodilatation involving NO. From approximately "mid-incubation," the cardiovascular system of the chicken embryo can respond to hypoxemia with a redistribution of cardiac output from peripheral vascular beds to the heart and brain (25, 26). As development progressed, this chemoreflex became more prominent. Based on our observations, the development of arterial contractile reactivity, i.e., its pharmacological control and regional heterogeneity, may participate herein and seems to involve neurogenic and endothelial influences on arterial smooth muscle cell differentiation.