INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

IN THE MATURE NERVOUS system, transsynaptic signals play a major role in the regulation of postsynaptic receptors and their associated signaling cascades (14, 30, 39, 61). Typically for many neurotransmitters, increased presynaptic activity elicits uncoupling of postsynaptic receptors from their response elements (desensitization) (14, 39), and with continued stimulation, cell membrane receptors are removed from the surface and internalized (downregulation) (30, 61). Conversely, reductions in presynaptic activity are accompanied by supersensitivity and receptor upregulation (14, 30, 39, 61). However, in the developing organism, a number of studies suggest that these relationships are absent or even reversed (8, 12, 13, 16, 59). For cardiac -adrenoceptors (-ARs), repeated neonatal stimulation with the receptor agonist isoproterenol fails to elicit desensitization, and instead, signaling elements downstream from the receptor are induced, resulting in agonist-induced sensitization instead of the expected desensitization (13). Similarly, denervation at birth fails to elicit the expected -AR supersensitivity (59). Indeed, in many systems, removal of transsynaptic input during development permanently obtunds the subsequent receptor-mediated responses (8, 12, 16); as these effects are often elicited before the formation of the majority of synaptic connections, the first few "pioneer" synapses may be of key importance in programming the subsequent development of receptor expression and function.

In the developing heart, -AR input appears to be important in establishing the competence of a variety of nonadrenergic, heterologous signals that operate through adenylyl cyclase (AC), by influencing the expression and function of signaling proteins other than the -AR itself (13, 15, 16, 66-68). The current study addresses receptor "cross talk" in myocardial cells, by assessing whether -AR activity regulates the ontogeny of m₂ACh receptors (m₂AChRs), receptors whose actions directly oppose those of the -ARs. The ontogenetic profiles of both the receptors themselves and their neuronal inputs are well established. -ARs emerge earlier in embryonic life than do mAChRs (41) and regulate fetal heart function initially (7) through circulating catecholamines (25) or intrinsic cardiac adrenergic cells (17). However, in terms of neuronal input, vagal cholinergic function appears first, in the early neonatal period, followed by the onset of sympathetic function during the second to third postnatal week (32, 53). To delineate the role played by -AR stimulation, we reversed the normal postnatal sequence by administering isoproterenol to neonates; in addition, we evaluated the effects of neonatal treatment with 6-hydroxydopamine (6-OHDA), which destroys catecholaminergic projections and thus maintains cholinergic dominance (13, 59). By using terbutaline, a -AR agonist that penetrates the placenta and the blood-brain barrier, we determined the potential for cross talk between -ARs and m₂AChRs, and we compared effects on cardiac receptors with those in the brain. Whereas cardiac -ARs and m₂AChRs are colocalized on myocytes, the receptors are present on disparate cell populations within the fetal brain (21, 43, 45, 52). Presumably, control of the expression and/or function of m₂AChRs by -ARs would be likely to require cross talk within the same cell. Finally, in light of the widespread use of -AR agonists to arrest preterm labor, this study provides an animal model of -AR/m₂AChR cross talk to evaluate the potential for tocolytics to disrupt cardiac autonomic function.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Animal treatments. All studies were carried out in accordance with the declaration of Helsinki and with the Guide for the Care and Use of Laboratory Animals as adopted and promulgated by the National Institutes of Health. Timed pregnant Sprague-Dawley rats were shipped by a climate-controlled truck (transit time 12 h) and housed with free access to food and water. For studies of the effects of postnatal -AR stimulation, pups were randomized at birth and redistributed to the nursing dams with a litter size of 10. Additionally, dams were reassigned randomly to different litters each day so as to distribute any dam-related differences equally among all litters. Within each litter, equal numbers of animals were assigned to the different treatment groups in a sex-matched design using approximately equal proportions of males and females. Beginning on postnatal day (PN) 2, animals were given a 4-day regimen of once-daily subcutaneous injections of either 1.25 mg/kg of L-isoproterenol HCl, 10 mg/kg of terbutaline hemisulfate, or an equivalent volume (1 ml/kg) of vehicle, consisting of 0.9% saline plus 0.1% ascorbic acid. These treatments have been shown to elicit maximal cardiac -receptor stimulation in the neonate (1, 2, 13, 50). Twenty-four hours after the last injection, animals were killed, hearts and brains were dissected, and the brain stem was isolated from the rest of the brain by removing the cerebellum and performing a cut rostral to the thalamus. All tissues were frozen with liquid nitrogen and stored at 45°C; preliminary studies verified that freezing and storage did not change any of the measured variables. In keeping with the vast literature on neonatal rat development, the cross fostering required for randomization did not affect nursing behavior nor did the dams reject or cannibalize any of the pups. Growth rates and receptor expression in the control group were within normative limits for unmanipulated animals (36, 54, 56, 58). We also performed preliminary experiments comparing unmanipulated animals to cross-fostered controls and found no differences in receptor expression (data not shown).

Receptor binding. Receptor binding capabilities were assessed by methods described in earlier publications (31, 65). The overall strategy was to examine binding at a single subsaturating ligand concentration in preparations from every animal. The selection of a single concentration of radioligand for the receptor analysis enables the detection of changes in either K_d or B_max but does not permit distinction between the two possible mechanisms. Accordingly, we used additional membrane preparations pooled from several animals in each treatment group for Scatchard analysis to identify which of the binding parameters was affected. This strategy was necessitated by the requirement to measure binding of two different ligands as well as six different measures of AC activity for each of the hundreds of membrane preparations involved in the study.

AC activity. Membrane fractions were prepared as already described for -AR binding. The membrane preparations were diluted 20-fold with 250 mM sucrose, 1 mM EGTA, and 10 mM Tris (pH 7.4) before the assay. Aliquots of cardiac membrane preparation containing 30 µg of protein were incubated for 30 min at 30°C with final concentrations of 100 mM Tris · HCl (pH 7.4), 10 mM theophylline, 1 mM adenosine 5'-triphosphate, 10 mM MgCl₂, 1 mg bovine serum albumin, and a creatine phosphokinase-ATP-regenerating system consisting of 10 mM sodium phosphocreatine and 8 IU phosphocreatine kinase, in a total volume of 250 µl. The enzymatic reaction was stopped by placing the samples in a 90-100°C water bath for 5 min, followed by sedimentation at 3,000 g for 15 min, and the supernatant solution was assayed for cAMP using radioimmunoassay kits. Preliminary experiments showed that the enzymatic reaction was linear well beyond the assay time period and was linear with membrane protein concentration; concentrations of cofactors were optimal.

Data analysis. Data are presented as means ± SE. For convenience, some data are presented as the percent change from control values. However, statistical tests were performed on the original data. Effects of -AR agonist treatment were evaluated by multivariate ANOVA incorporating all relevant variables (treatment, age, tissue, receptor subtype, and in vitro conditions for the AC determinations), and with data log transformed whenever variance was heterogeneous. Where appropriate, interaction terms were obtained for treatment × other variables, and individual differences between treatment groups were then established using Fisher's protected least significant difference. However, where there was only a main treatment effect in a multivariate design, only main effects were reported. Because there were no effects of any of the treatments on the concentration of membrane proteins (data not shown), results were the same regardless of whether receptor binding or AC activity was expressed per unit tissue weight or per milligram of membrane protein, and we restricted presentation to the latter parameter. Significance for main treatment effects was assumed at P < 0.05. However, for interactions at P < 0.1, we also examined whether lower-order main effects were detectable after subdivision of the interactive variables (60).

Materials. Animals were purchased from Zivic Laboratories (Pittsburgh, PA). cAMP radioimmunoassay kits were purchased from Amersham Pharmacia Biotech (Piscataway, NJ), and [³H]AFDX384 (specific activity 133 Ci/mmol) and [¹²⁵I]iodopindolol (specific activity 2,200 Ci/mmol) were obtained from PerkinElmer Life Sciences (Boston, MA). All other chemicals were obtained from Sigma Chemical (St. Louis, MO).

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Administration of isoproterenol to neonatal rats daily on PN 2-5 produced a significant 15% decrease in cardiac m₂AChR binding on PN 6, as evaluated at a single subsaturating ligand concentration (1 nM): control 189 ± 7 fmol/mg protein (n = 20); isoproterenol 159 ± 5 fmol/mg (n = 19), P < 0.003 vs. control. Scatchard determinations (Fig. 1) indicated a larger decrease of 25% in the number of binding sites (B_max); the smaller effect seen at a single ligand concentration represented partial offsetting of the decrease in receptor number by a significant decrease in receptor affinity (K_d). The change in K_d was not an artifact of residual isoproterenol in the membrane preparation, as that would have increased not decreased the K_d.

View larger version (21K): [in this window] [in a new window]   Fig. 1.   Scatchard plot for effects of neonatal isoproterenol treatment [postnatal days (PN) 2-5] on cardiac m2ACh receptor (m2AChR) binding assessed on PN 6. Data were obtained from 3 separate membrane preparations for each treatment group. ANCOVA indicates a significant treatment difference between the 2 plots (P < 0.05), and the inset table indicates the effects on the individual plot parameters, evaluated by Fisher's protected least significant difference. Bmax, maximum binding capacity or density of binding sites; B, bound ligand; B/F, ratio of bound to free ligand.

To determine whether the decrement in receptor binding elicited a comparable reduction in cell signaling capabilities, we evaluated the ability of m₂AChR agonists to inhibit AC activity (Fig. 2). In control rats, addition of carbachol or oxotremorine to cardiac membrane preparations produced significant decrements in AC activity (main effect of agonist P < 0.0001), indicating effective linkage of the receptors to AC via the inhibitory G protein (G_i). The effect of muscarinic agonists was greatest in the presence of stimulation by forskolin (interaction of m₂AChR agonist × stimulant P < 0.0001). In vivo treatment of neonatal rats with isoproterenol on PN 2-5 produced a significant reduction of the effectiveness of m₂AChR agonists by about the same proportion (15-20%) as the decline seen for receptor binding.

View larger version (15K): [in this window] [in a new window]   Fig. 2.   Adenylyl cyclase (AC) response to an m2AChR agonist in membrane preparations from control rats and rats that received isoproterenol on PN 2-5. Data represent means ± SE obtained from the number of animals shown in parentheses, evaluated as the percent change in AC activity elicited by a maximally effective concentration (100 µM) of carbachol or oxotremorine. ANOVA across all stimulants appears at the top; separate tests for each stimulant were not conducted because of the absence of a treatment × stimulant interaction. Activities (in pmol · min1 · mg protein1) in the absence of the m2AChR stimulants were as follows: control group, basal activity 18.1 ± 0.9; isoproterenol-stimulated activity 27.0 ± 0.7; forskolin-stimulated activity 362 ± 23; isoproterenol-treated group 20.0 ± 0.7, 29.3 ± 1.0, and 404 ± 15.

We next determined if the period for -AR regulation of m₂AChR expression and function extends back into fetal stages. To ensure penetration of the -AR agonist to the fetus, we used terbutaline, which is known to cross the placenta (4, 18, 33). In addition to assessing m₂AChR binding, we assessed -AR downregulation so as to verify that fetal tissues were exposed to terbutaline levels sufficient to alter expression of its primary receptor target (1). After 4 days of terbutaline treatment on GD 17-20, fetal cardiac m₂AChR and -AR binding sites were downregulated to a small but significant degree, with approximately the same effect seen for the two receptor types (Fig. 3). Terbutaline also penetrates into the brain (33), so we next compared the effects across the two tissues. Fetal terbutaline treatment failed to affect m₂AChR expression in the brain, despite the fact that it elicited much greater -AR downregulation than was seen in the heart. We also determined whether comparable effects could be elicited by postnatal terbutaline administration, given on the same schedule as in the isoproterenol studies (treatment on PN 2-5 and measurements on PN 6). Postnatal terbutaline treatment produced the same small (10%) decrement in cardiac m₂AChRs that had been seen with fetal treatment: control 177 ± 3 fmol/mg protein (n = 16); terbutaline 160 ± 4 fmol/mg (n = 16), P < 0.002 vs. control.

View larger version (21K): [in this window] [in a new window]   Fig. 3.   Effects of gestational terbutaline (Ter) treatment [gestational days (GD) 17-20] on cardiac and brain m2AChR and -adrenoceptor (AR) binding assessed on GD 21. Data represent means ± SE obtained from the number of animals shown in parentheses, presented as the percent change from the corresponding control (Con) values, which were: heart m2AChR 257 ± 6 fmol/mg protein; heart -AR 12.5 ± 0.2; brain m2AChR 136 ± 2; and brain -AR 9.5 ± 0.7. ANOVA across both receptor subtypes and both tissues appears at the top, with subdivision by tissue at the bottom. In the heart, the significant treatment difference was not distinguishable between the 2 receptor types (no interaction of treatment × subtype), so only the main effect is reported without testing of individual subtypes. In the brain, the differences were subtype selective (interaction of treatment × subtype), and the * denotes the selective treatment difference for -AR binding.

We next determined if the small degree of fetal cardiac m₂AChR downregulation seen after fetal terbutaline treatment was paralleled by a decrease in the ability of the receptors to inhibit AC activity (Fig. 4). Although m₂AChR agonists caused significant inhibition of AC on GD 20 (P < 0.0001), the effect was much larger by GD 21 (main effect of age P < 0.0001); terbutaline treatment had no significant effect on this pattern.

View larger version (14K): [in this window] [in a new window]   Fig. 4.   AC response to m2AChR stimulation in membrane preparations from control rats and rats that received terbutaline on GD 17-20. Data represent means ± SE obtained from the number of animals shown in parentheses, evaluated as the percent change in isoproterenol-stimulated AC activity elicited by a maximally effective concentration (100 µM) of carbachol or oxotremorine. ANOVA across both ages appears at the top; separate tests for each age were not conducted because of the absence of an effect of treatment or interaction of treatment × age. Activities (pmol · min1 · mg protein1) in the absence of the m2AChR stimulants were as follows: control group, GD 20 17.4 ± 0.8, GD 21 24.8 ± 0.7; terbutaline-treated group 19.3 ± 0.9 and 25.3 ± 0.7.

Finally, we determined whether tonic sympathetic activity played a role in the establishment of m₂AChR expression. Newborn rats (PN 1) were given 6-OHDA to destroy the developing sympathetic nerve terminals so as to interrupt endogenous catecholaminergic input (57). Three days later, we observed a significant increase in cardiac m₂AChRs of nearly 25%: control 238 ± 6 fmol/mg protein (n = 13); 6-OHDA 294 ± 19 fmol/mg (n = 13), P < 0.01 vs. control. The effect on m₂AChRs was bigger than the effect on -ARs, which actually showed a small (5%) nonsignificant decrease: control 9.3 ± 0.8 fmol/mg protein (n = 6); 6-OHDA 8.8 ± 0.6 fmol/mg (n = 6). The effect on m₂AChRs was statistically distinguishable from the lack of effect on -ARs (treatment × subtype interaction P < 0.05). The promotional effect of 6-OHDA on m₂AChRs persisted through the period over which sympathetic innervation normally develops [first 3 wk postnatally (53)] but at a lower magnitude than that seen in the immediate neonatal period, with a 6% increase on PN 7 and a 15% increase on PN 17 (data not shown). Although these smaller effects were not in themselves statistically significant, they were also not distinguishable from the larger effect obtained on PN 4 (P < 0.05 for the main effect of 6-OHDA, but no significant interaction for 6-OHDA × age), so the actual degree of persistence would need to be verified in future work. We also contrasted the effects of neonatal 6-OHDA treatment on cardiac m₂AChR expression with results obtained in the brain stem. 6-OHDA penetrates the neonatal blood-brain barrier to elicit massive norepinephrine depletion (64). Nevertheless, we did not observe upregulation of m₂AChRs in the brain on PN 4, the point at which peak effects were seen in the heart: control 212 ± 3 fmol/mg protein (n = 4); 6-OHDA 208 ± 13 fmol/mg (n = 6). The lack of effect in the brain stem was distinguishable from the upregulation seen in the heart (significant 6-OHDA × tissue interaction).

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Results obtained in this study indicate that fetal or neonatal -AR stimulation modulates the concentration and function of m₂AChRs in the developing heart. Repeated administration of isoproterenol elicited up to a 25% decrease in the number of m₂AChRs, an effect that was paralleled by a loss of the ability of m₂AChR stimulation to inhibit AC activity. In addition, isoproterenol administration also shifted the affinity of m₂AChRs, as evidenced by a decreased K_d. Ordinarily, ligand binding affinity for m₂AChRs tends to decrease with development, i.e., the K_d increases with age (22), so that our findings suggest that, in addition to altering the number and function of m₂AChRs, excess -AR stimulation promotes the maintenance of the immature conformation. It is unlikely that the alterations in binding affinity represent a difference in primary sequence of the receptor protein and, indeed, similar increases in K_d are seen for cardiac -ARs during neonatal development (27). Taken together, these results suggest that the alterations in K_d reside in factors that are common to multiple types of receptors, such as the membrane lipid milieu in which the receptors are embedded. Both age and -AR stimulation are known to alter membrane lipids, leading to changes in receptor affinities (3), and -AR stimulation has a profound effect on fetal/neonatal lipid metabolism (25). Repetitive or excessive stimulation by exogenous -AR agonists would thus be expected to influence the composition of cell membranes and hence the properties of membrane-associated receptors.

Our findings indicate that the reduction in m₂AChRs elicited by -AR stimulation likely represents specific, developmentally specified cross talk between the two receptor types. First, no such modulation of m₂AChR levels is seen with -AR agonist treatment in adults (29). Second, the reduction in m₂AChRs elicited by isoproterenol administration to neonates was observed, despite the fact that there was no downregulation of -ARs; the inability of isoproterenol to elicit homologous receptor downregulation in neonatal heart agrees with our earlier findings (59). Third, the downregulation was specific to the heart, where the -ARs and m₂AChRs are on the same cells, whereas no such effect was seen in the brain, where the two receptors reside on different cell populations (21, 43, 45, 52). Fourth, our results indicated receptor-subtype selectivity, which will be discussed below. Although our results do not address the cellular mechanisms for -AR/m₂AChR cross talk, these are likely to require additional interactions between cardiac cells and their systemic environment. Earlier work with cultured neonatal myocytes indicates that -AR stimulation evokes only a transient decrease in m₂AChR expression, followed by more prolonged increases in m₂AChRs (34). However, culturing of myocytes alters their differentiation state and specifically affects the expression and function of G protein-coupled receptors (19, 63). Given the prominent role played by endocrine factors in the expression and development of -ARs and mAChRs as identified in earlier work from our and other laboratories (5, 26, 27, 38, 40, 42, 48, 55, 62), maintenance of the in vivo hormonal milieu may provide key elements necessary to -AR/m₂AChR cross talk.

The ability of -AR stimulation to reduce cardiac m₂AChRs was also demonstrable in the fetus, indicating that the period for receptor cross talk extends to earlier stages of development. The effects of terbutaline, a ₂-selective agonist, on fetal m₂AChRs were smaller than those seen after postnatal administration of isoproterenol, but they were indistinguishable from the effect of terbutaline administered postnatally. It is unlikely that the smaller fetal effect represents failure of terbutaline to penetrate the placenta: we found significant downregulation of -ARs in both heart and brain, and our earlier work confirmed equivalent downregulation of hepatic ₂-ARs in the fetus and neonate after these terbutaline regimens (1, 2). Instead, the current results point to preferential effects for the ₁-AR subtype. Terbutaline elicited a smaller downregulation of cardiac m₂AChRs than did isoproterenol, a mixed ₁/₂-agonist, despite the fact that terbutaline, but not isoproterenol, evoked significant -AR downregulation. The fetal heart has a higher proportion of ₂-ARs than does the neonatal or adult heart (1), so that, if this subtype were highly effective in modulating m₂AChR expression, terbutaline would be more effective, not less effective, in the fetus than in the neonate. That does not mean that ₂-ARs are totally devoid of activity toward m₂AChRs. Indeed, cross talk via the ₂-AR is likely to be more important in tissues in which this is the predominant subtype, such as the developing lung, where -AR stimulation has been shown to elicit a robust decrease in m₂AChRs (46). However, as those studies were performed in cell cultures, it remains to be demonstrated whether prevailing ₂-AR/m₂AChR cross talk can be elicited to the same extent in vivo.

Although both neonatal isoproterenol and fetal terbutaline administration evoked a decrease in the number of cardiac m₂AChRs, only the postnatal treatment elicited a significant decrement in the AC response to receptor stimulation. The linkage of m₂AChRs to AC undergoes rapid development during the fetal treatment period, evidenced by the dramatic increase in muscarinic agonist effect between GD 20 and GD 21. Accordingly, there may be a significant proportion of spare (i.e., uncoupled) receptors at this time, so that the small decline elicited by terbutaline may be relatively less important. Alternatively, m₂AChR downregulation may be offset by the much higher rate of protein synthesis in the fetus; the heart grows fivefold between GD 17 and GD 21, so most of the receptors and downstream signaling proteins are formed after the initiation of treatment. As yet a third alternative, the smaller degree of downregulation of m₂AChRs seen with fetal terbutaline treatment (<10% decrease) may simply be too small to enable detection of an impact on cell signaling with the optimal conditions under which AC is measured in vitro. Studies of m₂AChR function in vivo may be required to reveal the physiological significance of the loss of fetal m₂AChRs.

The results obtained with isoproterenol or terbutaline treatment illustrate the effects of -AR overstimulation on m₂AChRs and their linked cellular responses. However, they do not address the issue of whether m₂AChR expression is under tonic control by the development of adrenergic input. Accordingly, we performed additional studies in which sympathetic nerves were destroyed at birth by the administration of 6-OHDA. Neonatal sympathectomy elicited significant increases in cardiac m₂AChRs of about the same magnitude seen for the decreases in receptors elicited by -AR agonists. Indeed, the effect of 6-OHDA on m₂AChRs was actually bigger than that on -ARs, which showed no induction at the same age; failure of 6-OHDA to change -AR expression in the neonate has been reported (59). As with -AR agonist treatment, there was no cross talk in the brain, again emphasizing that cross talk requires that the -ARs and m₂AChRs be located on the same cell. The promotional effect of 6-OHDA on cardiac m₂AChRs persisted through the stage in which sympathetic innervation shows its most dramatic increases (53), but perhaps surprisingly, at a lower magnitude of effect. This suggests that tonic adrenergic input does play a role in the programming of m₂AChR development, but primarily in early stages, before the onset of reflex control of sympathetic activity. Thus only a few "pioneer" synapses may be required for receptor cross talk. The subsequent decline in the effect of denervation is likely to represent increasing contributions of nonneuronal factors to receptor expression [e.g., hormones (53)]. Ultimately, then, the role of sympathetic neurons to the control of cardiac m₂AChRs is a modulatory one, rather than representing an essential input.

-AR agonists are widely used as tocolytics in preterm labor, with treatment often extending for 10-20% of the total gestational period (9, 47), just as used in the current study. Accordingly, it is essential to establish whether the current findings are likely to have corresponding functional effects on fetal/neonatal cardiac function. Although vagal control of heart rate and contractility are demonstrable at birth, parasympathetic input is relatively weak (32, 51), so that even a minor effect on mAChRs may have an adverse functional effect. Indeed, imbalances of muscarinic/adrenergic receptor expression in the heart and in brain stem areas involved in cardiorespiratory regulation are implicated in perinatal morbidity and mortality, including SIDS (11, 20, 37, 54). It is therefore notable that the changes seen here are comparable in magnitude to decreases in brain stem muscarinic receptor binding seen in infants that died of SIDS (20) or to shifts in brain stem and cardiac -ARs and m₂AChRs in animal models that recapitulate the cardiovascular changes thought to underlie perinatal hypoxia-induced brain damage (54). Future work should concentrate on whether perinatal exposure to -AR agonists compromises neonatal cardiac function, both under basal conditions and in situations like hypoxia that may trigger cardiovascular collapse.