1 Anatomy
2 Physiology, Northeastern Ohio Universities College of Medicine, Rootstown, Ohio 44272-0095
3 Department of Physiology, Wayne State University School of Medicine, Detroit, Michigan 48201
4 Department of Neurology, Boston University School of Medicine, Boston Massachusetts 02118
![]() |
ABSTRACT |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
sympathetic nervous system; norepinephrine; cardiac function; knockout mice
![]() |
INTRODUCTION |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Studies that utilized targeted inactivation of neurotrophin genes to investigate the roles of neurotrophins in vivo focused primarily on the impact of a complete absence of these factors on neuron populations in -/- mice (4, 8, 9). A major limitation of this type of study is that -/- mice die at birth or shortly thereafter due to pleiotropic effects on different organs or a failure to thrive (5, 8). Neurotrophins are also required by mature, fully differentiated neurons (24, 29, 30). Therefore, the premature mortality of -/- mice limits their usefulness for investigating the function of the neurotrophins in postnatal mice.
Heterozygous mutant (+/-) NT3 mice, in which a single copy of the NT3 gene has been inactivated, are visually indistinguishable from wild-type (+/+) mice and reproduce well (8). Few studies have closely examined the phenotype of NT3 mice. However, while conducting a detailed analysis of -/- NT3 mice, Ernfors et al. (8) noted that +/- NT3 mice exhibit a significant loss of sensory neurons. Similarly, deficits in sensory (4) and sympathetic (2) neurons occur in NGF +/- mice.
This study addressed whether two copies of the NT3 gene are necessary for proper pre- and postnatal development of the sympathetic innervation of the heart by systematic analysis of anatomical, biochemical, and physiological parameters of sympathetic innervation in NT3 +/- mice at birth, weaning, and adulthood. We show that deletion of a single copy of the NT3 gene results in quantifiable structural and functional deficits in the sympathetic innervation of the heart of postnatal mice, thereby indicating a gene-dosage effect for the NT3 gene.
![]() |
METHODS |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
|
Neuron counts.
The number of neurons in the left stellate ganglion (LSG) of +/+, +/-, and -/- newborn (postnatal day 0 or P0) mice and +/+ and +/- weanling (P21) and adult (P60) mice were counted to estimate the number of cardiac sympathetic neurons. The thoraxes of P0 and P21 mice were immersed overnight in Bouin's fixative, transferred to 70% ethanol, and processed for paraffin embedding. Serial sections were cut at 6 µm, stained with hematoxylin and eosin, and mounted on glass slides. Only nucleated cells with a prominent nucleolus (a feature of neurons in the LSG) were counted at 30-µm intervals throughout the ganglion. The LSG of adult mice were dissected and fixed in cold 2% paraformaldehyde-2.5% glutaraldehyde in 0.1 M sodium cacodylate buffer. After postfixation in 1% OsO4, ganglia were dehydrated and embedded in Embed 812. Serial 1-µm sections were collected at 50-µm intervals throughout the ganglion, and neurons were counted using the same criteria as for P0 and P21 mice. The average nucleolar diameter was estimated from 20 randomly sampled nuclei in a series of serial 1-µm sections. The total number of neurons in the LSG of P0, P21, and P60 mice was estimated by a classic technique. The counts of neurons in sections were summed, multiplied by the interval length, and divided by the mean nucleolar length (27).
TUNEL assay and tyrosine hydroxylase immunocytochemistry.
A terminal transferase-mediated dUTP-biotin nick end labeling (TUNEL) assay was used to identify apoptotic cells in the LSG from two mice of each genotype on P0, and data were correlated with anti-tyrosine hydroxylase (anti-TH) immunocytochemistry on adjacent sections to demonstrate that apoptotic cells were sympathetic neurons. Tissues were fixed in 4% paraformaldehyde for 24 h at 4°C and 24 h at 25°C, embedded in paraffin, sectioned sagittally at 6 µm, and mounted on glass slides. Sections were transferred through xylene and descending concentrations of ethanol, then rehydrated in PBS. TUNEL sections were treated for 20 min with 25 µg/ml proteinase K in PBS. Positive controls were then exposed to 1 µg/ml DNase I for 5 min. Endogenous peroxidase activity was blocked by incubating the sections in 3% H2O2 for 5 min. Sections were rinsed in 0.5 M Tris, pH 7.5, 50 mM MgCl2, 0.6 mM mercaptoethanesulfonic acid, and 0.5 mg/ml BSA for 10 min, then incubated at 37°C for 2.5 h in the same buffer containing 2 µl Klenow enzyme (300 U/ml) and 20 µl biotinylated nucleotide mix. Negative controls were incubated in the same mixture without the Klenow enzyme. Rinsing in 3 M NaCl and 300 mM sodium citrate solution for 5 min terminated the reaction. After a rinse in distilled H2O, sections were incubated with a 1:500 mixture of streptavidin-peroxidase conjugate and strept-avidin-peroxidase diluent for 10 min. Subsequently, sections were stained with blue label dye (Trevigen, Gaithersburg, MD) for 10 min, dehydrated with an ethanol series, cleared with xylene, and coverslipped with mounting medium.
Adjacent sections of the thoracic regions of +/+, +/-, and -/- P0 mice and the LSG of +/+ and +/- P60 mice were processed for anti-TH immunocytochemistry. Endogenous peroxidase activity was blocked as described previously. Incubating sections in goat serum blocked nonspecific binding. Sections were incubated overnight at 4°C with rabbit monoclonal anti-TH (Protos Biotech) diluted 1:100. Antibody binding was detected by an indirect peroxidase (Vectastain ABC) method followed by detection using diaminobenzidine tetrahydrochloride (DAB) as a substrate.
To trace the growth of sympathetic axons into the heart, hearts of +/+, +/-, and -/- mice on P0 and +/+ and +/- mice on P21 and P60 were fixed, embedded in paraffin, sectioned, and processed for anti-TH immunocytochemistry according the protocol used for identifying sympathetic neurons in the LSG.
Norepinephrine measurement.
The extent of cardiac sympathetic innervation was estimated by measuring norepinephrine (NE) concentration in the heart. At P0, P21, and P60, mice were weighed and then euthanized by decapitation (P0) or cervical dislocation (P21, P60). Hearts were excised and weighed for calculation of heart/body weight ratios. Cardiac NE concentrations were measured in +/+, +/-, and -/- mice on P0 and in +/+ and +/- mice on P21 and P60. Entire hearts of P0 mice or ~1020 mg of the free wall of the left ventricle of P21 or P60 mice were immersed in cold 0.1 M perchloric acid. Tissue samples were homogenized in a glass-walled tissue grinder and centrifuged at 3,000 rpm for 5 min at 4°C to pellet cell debris. The supernatant was aspirated through a 0.45-µm filter, and the concentration of NE in the filtrate of each sample was analyzed by HPLC.
Measurement of sympathetic tonus.
All instrumentation of adult mice was performed using aseptic surgical procedures. Mice were anesthetized by an intramuscular injection of a mixture containing ketamine (40 mg/kg) and xylazine (8 mg/kg) and were instrumented for measurement of arterial pressure (AP), mean arterial pressure (MAP), and heart rate (HR) by inserting a polytetrafluoroethylene catheter into the descending aorta via the left common carotid artery. The catheter was also used for infusion of cardiac autonomic antagonists. The arterial catheter was flushed daily, filled with heparin (1,000 U/ml), and plugged with a stainless steel obturator. Mice were monitored for signs of infection and a change in body weight during recovery from surgery. During this recovery period, mice were familiarized with the experimental apparatus. All mice were fully recovered and healthy at the time of data collection.
Instrumented, conscious, unrestrained mice were placed in a large Plexiglas box (30.5 cm3) and allowed to adapt to the laboratory environment for 1 h. After baseline hemodynamic measurements were obtained, HR, AP, and MAP responses to cardiac autonomic (muscarinic-cholinergic and ß1-adrenergic) blockade were measured. Cardiac muscarinic-cholinergic receptor blockade was achieved by infusion of the nonselective receptor antagonist, methylatropine (MA, 3 mg/kg) through the carotid arterial catheter. HR response to MA peaked 1015 min after infusion; therefore, this interval was selected as a standard before HR was measured. Cardiac ß1-adrenergic receptor blockade was achieved by infusion of the selective ß1-adrenergic antagonist, metoprolol (MT, 10 mg/kg) into the catheter. MT was infused 15 min after MA, and HR was measured again after 15 min. Intrinsic heart rate (HRI) was considered to be the HR after complete cardiac autonomic (muscarinic-cholinergic and ß1-adrenergic) receptor blockade. Sympathetic tonus was calculated as HRM - HRI, where HRM is the heart rate after muscarinic-cholinergic receptor blockade.
Gender differences.
Initially, data were analyzed separately for male and female mice within each group. No statistically significant gender differences in numbers of LSG neurons, cardiac NE concentrations, MAP, resting HR, or sympathetic tonus were noted for age/genotype-matched mice (P > 0.05). Hence, data for males and females were pooled for analyses.
Statistical analysis.
Mice were grouped according to age and genotype. Data were analyzed using a SigmaStat statistical program. Heart/body weight ratios were transformed by an arcsine square root transformation prior to ANOVA. A square root transformation was performed on neuron counts in the LSG prior to analyses. Differences indicated by ANOVA were elucidated by post hoc tests. Although transformed data were analyzed, data are reported as untransformed values. NE concentrations and sympathetic tonus were compared between corresponding data cells grouped by age and genotype using an unpaired Student's t-test. Alpha levels of 0.05 were used to determine statistical significance in all tests.
![]() |
RESULTS |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Neuron deficits in the stellate ganglion of NT3 +/- and -/- mice.
The majority of postganglionic cardiac sympathetic neurons originate in the stellate ganglion in the rat (19). Therefore, we examined the LSG (Fig. 2A), to determine whether inactivation of one copy of the NT3 gene resulted in a reduction in the number of cardiac sympathetic neurons. At birth, -/- mice had significantly fewer (P < 0.05) neurons than did +/- mice, which in turn had fewer (P = 0.06) neurons than +/+ mice (Fig. 3A). In addition, the neuron deficit in +/- mice was 22% or approximately one-half that of -/- mice (46%) at birth. Neuron deficits in +/- mice persisted postnatally. The reduction in neuron number in +/- mice, relative to +/+ mice, at P21 and P60 approximated that at birth (25% and 20%, respectively). Neuron number decreased by relatively the same degree in both +/+ and +/- mice between P0 and P21. No significant decreases in neuron number occurred between P21 and P60 in the LSG of either +/+ or +/- mice.
|
|
Decreased anti-TH reactivity in NT3 +/- mice.
Deletion of one or more NT3 genes qualitatively affected the intensity of anti-TH reactivity in the LSG. Sections of the LSG from +/- mice bound anti-TH antibodies more intensely than those of -/- mice and less intensely than those of +/+ mice on P0 (Fig. 2, DF). In addition, the LSG of +/- mice bound anti-TH less intensely than the LSG of +/+ mice on P60 (Fig. 2, G and H).
Lowered cardiac NE concentrations in NT3 +/- mice.
The principal neurotransmitter stored in nerve terminals of cardiac sympathetic neurons is NE (15). Thus NE concentrations were measured to provide an index of cardiac sympathetic innervation. Mean cardiac NE concentrations were equivalent among +/+, +/-, and -/- mice at birth, ranging from 100 to 160 pg/mg tissue (Fig. 3B). Ventricular NE concentrations increased gradually with age. NE concentrations differed significantly (P < 0.05) between +/+ and +/- mice at weaning and this difference was more pronounced (P < 0.05) in adults (Fig. 3B). Ventricular NE concentrations of adult +/- mice were 37% less than those of +/+ mice (Fig. 3B). Cardiac NE concentrations were higher than those reported for adult mice (15), but this difference reflects a strain difference between Balb/c 129 and C57/Bl6 mice (Walro, unpublished observations).
The hearts of +/+, +/-, and -/- mice on P0 and +/+ and +/- mice on P21 and P60 were stained with anti-TH antibodies to determine whether low cardiac NE concentrations on P0 and gradually increasing cardiac NE concentrations through P60 were due to the proliferation and growth of cardiac sympathetic nerve endings within the heart. Sympathetic nerve endings were restricted to the epicardium on P0 (Fig. 4, A and B). On P21, sympathetic nerve endings were detected to the greatest extent in the epicardium but were observed coursing with blood vessels in the myocardium in both +/+ and +/- mice (Fig. 4, C and D). By P60, numerous sympathetic nerve endings were visible among myocytes in the myocardium of both +/+ and +/- mice (Fig. 4, E and F). However, no differences in density of these fibers were discernible among mice all three genotypes of mice on P0 and between +/+ and +/- mice on P21 and P60.
|
|
![]() |
DISCUSSION |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Role of NT3 in the survival of cardiac sympathetic neurons.
The importance of NT3 for the survival of sympathetic neurons is well documented. The superior cervical ganglion (SCG) of newborn mice lacking the NT3 gene contains approximately one-half the number of neurons as the SCG from wild-type mice (8, 9). We observed a 46% deficit in the LSG of newborn mice lacking both copies of the NT3 gene and a 22% deficit in mice lacking one copy of the NT3 gene; thus paravertebral sympathetic ganglia other than the SCG are dependent on NT3 for survival prior to birth. These data differ from that of Brennan et al. (2), who found no difference in numbers of SG neurons in +/- and +/+ NT3 mutant mice.
Excessive cell death due to apoptosis occurs prior to birth in the SCG of NT3 -/- mice, although data regarding the specific days during gestation when cell death occurs are conflicting (6, 28). The rates of apoptosis in the SCG of NT3 -/- mice and +/- mice are equivalent at birth (9). Consequently, the decrease in LSG cell number observed in +/- and -/- mice on P0 likely results from excessive cell death prior to birth. We observed few TUNEL+/TH+ cells in the LSG of +/+, +/-, and -/- mice on P0, which indicates that the rate of apoptosis may be declining postnatally. No qualitative difference in the number of TUNEL+/TH+ LSG cells among +/+, +/-, and -/- mice on P0 and stable cell numbers in +/- mice relative to +/+ mice through P60 indicate that lowered levels of NT3 do not lead to a significant increase in postnatal apoptosis of LSG neurons.
Sympathetic neurons of postnatal animals require both NGF and NT3 for survival (24, 28). The number of neurons in the LSG of both NT3 +/+ and +/- mice decreased slightly between P0 and P21, but the decreases were proportional for both genotypes. Similarly, deficits in numbers of neurons originate prior to birth in the trigeminal ganglion and SCG, but no further loss in +/- mice relative to +/+ mice occurs postnatally (6, 7, 28). Thus the decrease we observed in +/- mice from P0P21 may reflect naturally occurring cell death, a process that occurs through postnatal day 7 (P7) in the SCG of the rat (26), rather than a reduction in NT3 concentrations.
Administration of antibodies specific for NT3 to 1- to 2-wk-old rats results in an 80% deficit of neurons in the SCG; thus NT3 has been reported to be essential for maintenance of postnatal sympathetic neurons (29). The stable sizes of the sympathetic populations in the stellate ganglia of +/+ and +/- mice from P21-P60 suggest that reduced concentrations of NT3 still provide a maintenance level of postnatal support for the smaller population of neurons present in the LSG of +/- mice.
Effect of NT3 on cardiac NE concentrations.
Cardiac NE concentrations of all three genotypes of mice were low at birth (1013% of adults), and those of NT3 +/- and +/+ mice increased gradually with age. TH+ sympathetic endings were present in the epicardium of +/+, +/-, and -/- mice on P0 but were not observed in the myocardium of +/+ or +/- mice until P60. Collectively, these data reflect the immaturity of sympathetic innervation of the heart at birth and are consistent with data obtained previously in mammals (15, 23, 25). Sympathetic neurons first contact target organs ~9 days before birth in rodents (16, 22), but the sympathetic innervation of the heart may not mature until 60 days after birth (25). Ursell et al. (25) noted that the sympathetic innervation of the dog heart was restricted to the epicardium in mid-late gestational pups and that sympathetic neurons penetrated the myo- and endocardia and matured 2 mo after birth. Thus, although sympathetic neurons reach the mouse heart early in development, the low cardiac NE concentrations in all genotypes of NT3 mice at birth suggest that the penetration and ramification of sympathetic neurons into the myo- and endocardia are postnatal events.
The qualitative density of TH+ nerve fibers was equivalent in the epicardium of all three genotypes at birth. Similarly, El Shamy et al. (6) observed no qualitative difference in innervation in the hearts of NT3 -/- and +/+ mice after TH immunocytochemical staining on P7 (6). In the present study, location and density of TH+ nerve fibers within the heart were similar in NT3 +/+ and +/- mice through P60. However, +/- mice exhibited lower cardiac NE concentrations than did +/+ mice at P21, and this deficit became more severe by P60. The increasing disparity in cardiac NE concentrations was not paralleled by a proportionate loss of neurons in the LSG or by a loss of nerve terminals in the heart, thus cell death could not have accounted for the lower cardiac NE concentrations. In the absence of a proportional loss of neurons, reduced synthesis of NE is the most plausible source for the increasing deficits in cardiac NE concentrations in +/- mice relative to +/+ mice.
Gene dosage effect of NT3.
Deletion of the NT3 gene may affect the size of sympathetic neuron populations in a dose-dependent fashion. The 22% deficit in sympathetic neurons present in NT3 +/- mice at birth was approximately one-half the deficit (46%) observed in NT3 -/- mice. Indirect evidence suggests that the number of cellular copies of NT3 and NGF genes regulates the size of sensory or sympathetic neuron populations. Ernfors et al. (8) noted a 50% reduction in muscle spindles, which exist in a 1:1 ratio with the group Ia afferent and fusimotor neurons that innervate spindles, of NT3 +/- mice. Likewise, deletion of a single copy of the NGF gene results in a 50% reduction in sympathetic neurons (2), and 1324% fewer calcitonin gene-related peptide (CGRP)-reactive neurons are present in dorsal root ganglion (DRG) L45 of NGF +/- mice (4). Although haplo-insufficiency, or deficits in neurotrophin levels of target organs resulting from inactivation of one copy of a gene, has not been documented in NT3 +/- mice, brain-derived neurotrophic factor (BDNF) +/- mice have been shown to have reduced tissue levels of BDNF by ELISA (1). The deletion of one or more BDNF genes affects the survival of neurons in vestibular and nodose ganglia in a dose-dependent manner. The number of neurons in nodose/petrosal and vestibular ganglia of BDNF +/- mice is intermediate between those of BDNF +/+ and -/- mice (1).
Haplo-insufficiency may affect the phenotype of sympathetic neurons in addition to affecting their survival. A reduced synthesis of NE in NT3 +/- mice relative to +/+ mice may be a secondary source of lower ventricular NE concentrations observed in weanling and adult NT3 +/- mice. Sensory neurons in NGF +/- mutant mice bound anti-CGRP less intensely than did neurons from wild-type littermates, thereby suggesting a reduced content of CGRP (4). Similarly, we observed a lower intensity of anti-TH staining in the LSG of -/- and +/- mice compared with +/+ mice on P0 and lower intensity for +/- mice relative to +/+ mice on P60.
Functional consequences of inactivating an NT3 gene.
The functional consequence of inactivating an NT3 gene was determined in chronically instrumented conscious, unrestrained mice. This approach eliminated many of the confounding influences of anesthesia on the autonomic nervous system. Moreover, the use of intact conscious animals compliments and extends the morphological and biochemical components of this study.
The sympathetic nervous system is critical for maintaining both resting and stressed cardiac function based on its dominant role in regulating cardiac inotropism and chronotropism. Several investigators have recently documented the influence of the sympathetic nervous system on cardiac function by determining sympathetic tonus (3, 11, 13, 18). These studies have demonstrated that sympathetic tonus reflects changes in HR and is an appropriate method for indirectly determining the influence of the sympathetic nervous system on cardiac regulation. In this study, we have demonstrated that deletion of a single copy of the NT3 gene results in a lower resting HR as well as a lower sympathetic nerve activity in conscious, unrestrained mice.
Lower resting HRs in the adult NT3 +/- mouse are consistent with the reduction in sympathetic tonus in these animals. Ganglion blockade, which blocks both parasympathetic and sympathetic outflow, results in small but nonsignificant decreases in resting HR in mice (21). These data demonstrate that sympathetic drive predominates in maintaining resting HR in the mouse (14, 21). Furthermore, sympathetic drive mediates increases in HR, cardiac output, and AP during exercise or other stressors. The HR of a mouse can increase to ~800 beats/min (21). The mechanisms mediating the increase in HR are vagal withdrawal and an increased activity of the sympathetic nervous system (21). Given the important role that the sympathetic nervous plays in the regulation of cardiac inotropy and chronotropy, NT3 +/- mice may have a reduced ability to respond to exercise or other stressors due to the reduced chronotropic and inotropic reserve. These important questions merit further investigation.
Sympathetic neurons regulate HR mainly through ß1-adrenergic receptors (21). However, other receptors also mediate sympathetically induced chronotropic effects. Specifically, 1-adrenergic receptors (20) and the receptors for neuropeptide Y (10) have direct chronotropic coupling in isolated atria. These data may explain the failure to find lower resting HRs in ß1-adrenergic receptor knockout mice (21). Importantly, the HRs of ß1-adrenergic receptor knockout mice were significantly lower than wild-type mice after blocking cardiac muscarinic receptors. Blockade of parasympathetic outflow reduced HR variability to the point where differences in resting HR could be observed. These data support the notion that parasympathetic outflow has a profound influence on resting HR in the mouse (21).
NT3 +/-mice as a model for cardiac dysautonomias.
NT3 -/- mice die within a few weeks of birth due to adverse pleiotropic effects on the outflow vessels of the heart or a general failure to thrive (5, 8, 9). Consequently, researchers have been unable to study how neuron deficits in these mice affect organ function in postnatal mice. Although inactivation of a single copy of the NT3 gene results in significant deficits in the sympathetic nervous system, NT3 +/- mice survive into adulthood. Individuals afflicted with dysautonomias, a broad category of disorders affecting the autonomic nervous system of humans, also survive to adulthood. When the sympathetic nervous system is affected, the hallmarks of the disease include catecholamine deficiencies and loss of myocardial nerve terminals (12). Hence, the phenotype of NT3 +/- mice mimics the phenotype of individuals affected by sympathetic dysautonomias in many respects.
![]() |
ACKNOWLEDGMENTS |
---|
![]() |
FOOTNOTES |
---|
![]() |
REFERENCES |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|