| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
The Edward Mallinckrodt Department of Pediatrics, Washington University School of Medicine, St. Louis, Missouri 63110
 |
ABSTRACT |
|---|
 TOP
 ABSTRACT
 INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
|
|---|
This study combined single
and transneuronal labeling to define the origin of midline-crossing
vagal fibers projecting to the rat's lungs. Injections of the
-subunit of cholera toxin (CT-) into the lungs labeled similar
numbers of neuronal somata in the nucleus ambiguus and dorsal motor
nucleus of the vagus on both sides of the medulla, even though vagal
stimulation increased lung resistance 50% less in the contralateral
than in the ipsilateral lung. Unilateral cervical vagotomy prevented
CT- labeling of ipsilateral neuronal somata and sensory fibers,
indicating that lung-bound vagal fibers undergo decussation only inside
the thorax. Injections of CT- and FluoroGold into opposite main stem
bronchi double labeled 30% and 11% of all neuronal somata
immunoreactive for CT- and FluoroGold, respectively, showing that
one single vagal motoneuron can innervate airways on both sides.
Injections of pseudorabies virus into the right lung revealed a
bilateral network of infected neurons, even after unilateral vagotomy.
The latter did not prevent infection of the ipsilateral vagal nuclei. These findings demonstrate that vagal motoneurons that project to the
lungs receive contralateral inputs from the airway premotor network and
vagal bronchomotor centers.
parasympathetic system; nucleus ambiguus; dorsal motor nucleus of the vagus; cholera toxin; pseudorabies virus
| |
INTRODUCTION |
|---|
|
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
|
|---|
LUNGS AND AIRWAYS RECEIVE the preponderance of their motor nerve supply from medullary neurons, the somata of which are located in the nucleus ambiguus complex or in the dorsal motor nucleus of the vagus (11, 12, 15, 25). These neurons are regulated by an extensive brain stem neuronal network (10, 11, 25) that integrates sensory information from respiratory mechano- and chemoreceptors, autonomic centers, and premotor inspiratory neurons into a respiratory-modulated outflow (34). Vagal motor fibers carry this outflow to the airways, where they are thought to establish obligatory synapses with an intrinsic network of parasympathetic ganglia, which in turn provide direct innervation to airway smooth muscle, glands, and blood vessels.
For such a system to be mechanically efficient, however, preganglionic inputs must reach distant regions of both lungs at approximately the same time and with similar intensity. Vagal stimulation and neuroanatomic tracing studies suggest that these requisites are met in part by a dual-innervation arrangement in which each lung receives fibers from both vagus nerves. The existence of such an arrangement has been suspected since the early neuroanatomic studies of Larsell and Mason in rabbits (16) and Honjin in mice (13) revealed that a substantial number of lung vagal preganglionic and sensory fibers do not undergo degeneration after ipsilateral cervical vagotomy. Olsen et al. (23) provided functional evidence for the presence of midline-crossing vagal fibers by showing that unilateral electrical stimulation of the vagus nerve increases the airflow resistance of both lungs in dogs and cats. More recently, using horseradish peroxidase as a retrograde neuronal tracer, Kalia and Mesulam (15) demonstrated that the right main stem bronchus of the cat receives a bilateral supply of motor fibers that originate from the nucleus ambiguus and dorsal motor nucleus of the vagus. Interestingly, the same investigators reported that the apical lobe of the right lung receives bilateral innervation only from the dorsal motor nucleus of the vagus; innervation from the nucleus ambiguus was unilateral.
The present study was designed to uncover some unexplored features of
this dual-innervation arrangement in the rat. First, we assessed the
size and trajectory of the contingent of vagal motor and sensory fibers
that cross the midline to innervate the contralateral lung by analyzing
the effects of cervical and thoracic vagotomies on the topography of
neuronal labeling by the -subunit of cholera toxin (CT-) injected
into the lungs. Second, we investigated whether individual vagal motor
fibers are committed to the innervation of unilateral targets or divide
to reach bilateral targets by determining the rate of double labeling
of medullary neurons after injections of CT- and FluoroGold (FG)
into opposite main stem bronchi. Finally, we elucidated the existence
of intramedullary connections between vagal preganglionic centers by
examining the effects of unilateral vagotomy on the retrograde
transsynaptic labeling of brain stem neurons by pulmonary injections of
pseudorabies virus.
| |
METHODS |
|---|
|
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
|
|---|
All experiments were performed in male Sprague-Dawley rats (300-400 g body wt, 10-12 wk old; Charles River, Wilmington, MA) following protocols approved by the Washington University Animal Studies Committee. The rats were housed at 23°C in a climate-controlled room with access to standard rat chow and water.
Midline Crossing by Vagal Motor and Sensory Fibers
The experiments described here ascertained the trajectory and size of the contingent of vagal fibers that cross the midline to innervate the contralateral lung. Our experimental approach included two complementary strategies. The first, aimed at defining whether vagal motor and sensory fibers undergo decussation in the brain stem (Fig. 1A), compared the topographical patterns of neuronal labeling produced in the medulla oblongata by the injection of CT- into the lungs of rats that had
intact vagi (n = 16, divided into 2 groups of 8 rats,
depending on the side of the injection) or had undergone unilateral
cervical vagotomies (n = 32, divided into 4 subgroups
of 8 rats, each representing a different combination of lung injection
and vagotomy sides). The second strategy was designed to elucidate
whether midline-crossing vagal fibers exit the vagal trunk above or at
the origin of the recurrent laryngeal nerve or follow the path of
smaller parahilar branches that emerge from the nerve near the tracheal
bifurcation. To this end, we analyzed the effects of a low-thoracic
vagotomy on the response of each individual lung to unilateral vagal
stimulation (Fig. 1D; n = 18) and the
retrograde labeling of vagal medullary neurons by CT- injected into
the ipsilateral lung (n = 6).
|
Pulmonary injections of CT-.
The rats were anesthetized with halothane (0.5-3%), which was
piped initially into an induction box or, after the institution of
mechanical ventilation, blended into the inspiratory limb of the
ventilator's circuit. The trachea was intubated through the mouth with
the help of a pediatric otoscope, and the endotracheal cannula was
connected to a rodent ventilator (Harvard Apparatus, South Natick, MA).
The lung, right or left, was exposed under a dissecting microscope
through a lateral thoracotomy at the fourth or fifth intercostal space.
A 50-µl volume of 0.1% CT- suspension (List Biological
Laboratories, Campbell, CA) was then injected into the lung parenchyma
with use of a Hamilton microsyringe (Reno, NV). The injectate volume
was divided into 5-10 injections, each performed with the needle
tip ~1-3 mm below the pleural surface while lung volume was held
constant to minimize disruption of the lung tissue. After every
injection, the visceral pleura was blotted with a cotton-tipped probe
to minimize nonspecific neuronal labeling by leaked CT- suspension.
After completion of the injections, the lung surface was washed
thoroughly with normal saline and allowed to dry for 3-5 min
before the thoracotomy was closed by layers. A small silicone rubber
tube was left inside the chest for removal of any air remaining in the
pleural space at the end of the surgery. On recovery from anesthesia,
the rat was returned to the holding facility, where it remained for
9-11 days.
Cervical and thoracic vagotomies. Cervical vagotomies were carried out through a paramedial neck incision. The vagus nerve was isolated from the carotid artery, and a 0.5-cm segment of the nerve distal to the origin of the superior laryngeal nerve was removed. The incision was closed by layers. Thoracic vagotomies were performed only on the right side to minimize animal usage. The vagus nerve was visualized through a lateral thoracotomy by retracting the lung anteriorly. The entire trajectory of the nerve caudal to the origin of the recurrent laryngeal nerve was dissected from the surrounding tissues and removed. The thoracotomy was closed as described above.
Effect of unilateral vagal stimulation on lung resistance. Rats were anesthetized with pentobarbital sodium (50 mg/kg ip). The trachea and the vagus nerves were exposed through a midline neck incision, and a tracheostomy was performed at the second intercartilaginous space. A cannula made from a 6-cm section of PE-190 tubing (0.12 cm ID) tapered at the end was inserted into the trachea and secured with a tie placed around the trachea and cannula. The tracheal cannula was connected to the rodent ventilator through an ensemble consisting of a small tube with a lateral port for measurement of airway pressure (model MAP45, ±56 cmH2O, Validyne Engineering, Northridge, CA) and a Fleisch no. 000 pneumotachograph attached to a differential pressure transducer (model MAP45, ±2.3 cmH2O) for measurement of airway gas flow. The length of the tubing connections was adjusted to ensure that the transducer outputs were well matched dynamically and had an amplitude-frequency response appropriate for the conditions of the experiments. The transducer outputs were amplified and recorded digitally at an acquisition rate of 500 Hz.
To determine the effects of vagal stimulation on lung resistance, we first exposed the lungs through a widened median sternotomy. After cutting the vagus nerves caudal to the origin of the superior laryngeal nerves, we denuded the distal stump of each nerve and placed it on a bipolar electrode attached to a nerve stimulator (model 50-5008, Harvard Apparatus), with the nerve and electrode covered with mineral oil to prevent dehydration. Next, we defined the supramaximal stimulus for each rat by varying stimulation frequency between 5 and 40 Hz and potential between 2 and 30 V while monitoring airway pressure as described previously (24). Pulse width was maintained constant at 0.5 ms. We then loosened the tie around the trachea and advanced the tracheal cannula into a main stem bronchus, right or left in random order, until the cannula's tapered end was wedged inside the bronchial lumen without obstructing segmental bronchial branches, as demonstrated by the selective and complete inflation of the corresponding lung. Before performing any measurements, we inflated the lung to an airway pressure of 30 cmH2O for 1-2 s to provide a consistent volume history. We then applied the supramaximal stimulus to the distal stump of the vagus nerve while the lung was ventilated with a peak pressure of 15 cmH2O at a rate of 80 breaths/min. The stimulus was discontinued as soon as its effect on airway pressure reached a maximum, usually within 30 s of stimulation. Finally, we pulled the cannula tip back into the trachea and allowed the rat to recover for 3-5 min before repeating the measurements in the contralateral lung. Total lung resistance was calculated as the ratio of the airway pressures and flows measured at midvolume points during inspiration and expiration (20). The increase in lung resistance produced by vagal stimulation was computed as the difference between the maximum resistance and the resistance determined before the initiation of the stimulus.Bilateral Innervation of Airways by Vagal Motoneurons
These experiments tested the hypothesis that a single vagal motoneuron can provide innervation to airways in both lungs. Our experimental approach was based on a simple double-labeling scheme in which CT- and FG were used as retrograde neuronal markers (Fig.
1B). To ensure consistency between injections and reduce the
potential for contamination of other organs within the chest, we
injected both markers into the main stem bronchi at the lung hilum. At
this location, the bronchi are large enough to allow visual inspection
of the site as the injection progresses. They are also separated from
each other by the mediastinum and distant enough from the esophagus and
the large vessels and heart to minimize the potential for nonspecific
labeling of vagal neurons innervating these structures when small
volumes of tracer are injected.
The rats (n = 24) were prepared as described above for
the pulmonary injections. The main stem bronchus, right or left, was visualized through a lateral thoracotomy by retracting the lung anteriorly. CT- (0.1%) or FG (15% solution, Fluorochrome, Denver, CO), dissolved in a 1- to 1.5-µl volume, was injected through a glass
micropipette mounted on an X-Y-Z micromanipulator
(Stoelting, Wood Dale, IL) in two to five individual injections
distributed over the external surface of the bronchus. The bronchial
serosa was lifted slightly with the tip of the pipette before the
injections to facilitate diffusion of the injectate within the
subserosal space and minimize backpressure-induced leakage. Once again,
the surface of the bronchus was blotted with a cotton-tipped probe after each injection, and the pleural space was washed thoroughly with
normal saline before the thoracotomy was closed. The contralateral bronchus was injected with the alternative neuronal marker 24 h
later after a similar procedure. The survival period before removal of
the brain and spinal cord was 11-12 days.
Interconnections Between Vagal Medullary Centers: Pseudorabies Virus Injections
This final group of experiments was designed to determine whether vagal motoneurons projecting to the lungs receive inputs from contralateral vagal premotor or motor neurons. We took advantage of the retrograde transneuronal labeling properties of the Bartha strain of pseudorabies virus (29). We also relied on the demonstration by the experiments described above that no pulmonary-bound vagal motor fibers cross the midline above the neck. We reasoned that a unilateral cervical vagotomy would preclude labeling of ipsilateral medullary neurons unless this labeling occurred transsynaptically through neurons located in the contralateral side of the medulla (Fig. 1C).Pseudorabies virus was injected into the right lung of 28 rats, which
were divided into 2 groups of 14 rats each, depending on whether the
cervical vagotomy was on the right or left side. (Once again, no
injections were performed into the left lung to minimize animal usage.)
The procedure was analogous to that described above for the CT-
injections. The viral suspension containing 3-4 × 107 plaque-forming units of the Bartha strain of
pseudorabies virus was injected in 0.3-µl aliquots to a total
injection volume of 1-1.5 µl. The survival period after the
injections was 5 days.
Selectivity of Neuronal Tracer Injections
The selectivity of neuronal labeling by the pulmonary injections of the various neuronal tracers used in the study relied on the restriction of the injectate to the injected lung. In designing the experiments, we were especially concerned that leakage from the lung CT- injection sites into the pleural cavity and diffusion of CT-
across the visceral pleura into thoracic organs such as the esophagus
or the heart, which receive a rich supply of motor and sensory nerve
fibers from the vagus (1, 5, 7,
9, 28, 30, 31),
could result in spurious labeling of medullary neurons
(8). We performed four types of control experiments to
address this concern.
The first experiment studied the distribution over time of a sample of
125I-CT- injected into the lungs of nine rats.
Radioiodination of CT- was performed as described previously
(17). A standard 0.25% suspension of CT- (List
Biological) was diluted 1:1 to a final volume of 200 µl with 0.1 M
NaPO4-buffered saline and incubated for 10 min with 20 µl
of a solution containing 2 mCi of Na125I (New England
Nuclear, Boston, MA) in the presence of two Iodo-Beads (Pierce
Chemical, Rockford, IL). After addition of 30 µl of cytochrome c (20 mg/ml; Sigma-Aldrich, St. Louis, MO) as a color label,
the reaction volume was fractionated in a Sephadex G-10 column
(Sigma-Aldrich). A 20- to 40-µl volume of the radiolabeled CT-
fraction (specific activity ~4,600 cpm/ng, concentration 0.08%) was
then injected in 10-µl aliquots into the right or the left lung
following the procedure described above. After survival periods of
8 h (n = 2), 24 h (n = 3),
and 7 days (n = 4), the thoracic viscera were removed.
The lungs were cut in coronal slices, which were then mounted flat on a
film plate along with the trachea and esophagus to obtain an
autoradiographic image. Lastly, the lung tissue was homogenized, and
the number of radioactive counts in representative samples from each
lung was determined in a scintillation counter.
The second control experiment was aimed at establishing whether CT-
suspension leaked from the injection sites may cause unintended
labeling of medullary neurons innervating other organs in the thoracic
cavity. We performed a small right lateral thoracotomy in six rats and
instilled a 30-µl volume of a 0.1% suspension of CT- on the lung
surface. The thoracotomy was then closed by layers. The survival period
after the injections was 11 days.
The third control experiment compared the topographic distribution of
medullary neurons labeled by concomitant injections of CT- into the
lung and FG into the esophagus. This comparison was intended to
identify double-labeled neurons, which would alert us of the presence
of CT- leakage from lung injection sites. The injections were
performed through a right thoracotomy in 13 rats. CT- (50 µl of
0.1% suspension) was injected into the right rostral and medial lobes
as described above. FG (0.8 µl of a 4% solution) was injected into
the right aspect of the wall of the midthoracic portion of the
esophagus with a glass micropipette. The survival period was 11-13 days.
Finally, the last control experiment tested our ability to discern
between neurons innervating adjacent airway targets by comparing the
distribution of medullary neurons labeled by CT- and FG, each
injected alternately into the extra- or intrathoracic trachea of six
rats. Because innervation of the trachea appears to follow a segmental
pattern (15), we predicted that these injections would
label two distinct populations of parasympathetic preganglionic
neurons. CT- (1 µl of 0.1% suspension) and FG (1 µl of 15%
solution) were injected into the anterior or lateral wall of the
trachea, which was exposed through a midline neck incision for the
extrathoracic injections and through a lateral thoracotomy for the
intrathoracic injections. The survival period was 11-13 days.
Preparation and Staining of Brain Tissue
At the end of the preestablished survival period after the neuronal tracer injections, anesthesia was induced again, this time with pentobarbital sodium (50 mg/kg ip). The heart and lungs were exposed through an extended median sternotomy, and heparin (1,000 U/kg) was injected into the right ventricle. The rat's systemic circulation was perfused with 0.1 M sodium phosphate-buffered saline and then with 4% paraformaldehyde in 0.1 M sodium phosphate buffer (pH 7.40) through a cannula inserted into the aortic root through the left ventricular wall. The spinal cord and brain were removed, submerged in 4% paraformaldehyde solution for 2 days, and stored in buffered 30% sucrose until they were sectioned.The brain stems were cut into 50-µm-thick coronal sections on a
freezing microtome. A 1-in-5 series of the sections was immersed in a
blocking solution containing 5% donkey serum (Sigma-Aldrich) in 0.3%
Triton X-100 and 0.02 M potassium phosphate buffer at room temperature
for 30 min and incubated overnight, also at room temperature, with the
appropriate solution of primary antiserum (Table
1). After they were rinsed thoroughly in
buffered saline, the sections were placed for 3-4 h in a 1:100
dilution of biotinylated anti-IgG antiserum (single-labeling studies)
or 1:50 dilutions of fluorescent-labeled (FITC or FITC and
tetramethylrhodamine isothiocyanate) anti-IgG antisera (double-labeling
studies) at room temperature. All the species-specific anti-IgG
antibodies were raised in donkey and purchased from Jackson
ImmunoResearch Laboratories (West Grove, PA). The sections treated with
biotinylated antisera were washed, treated for 1 h with an
avidin-horseradish peroxidase complex (Vectastain ABC kit, PK-400,
Vector Laboratories, Burlingame, CA), and stained with 0.05%
diaminobenzidine tetrahydrochloride in buffered saline containing
0.003% hydrogen peroxide. Staining was intensified by sequential
incubation with 1.5% silver nitrate for 1 h at 56°C, 0.2% gold
chloride for 15 min, and 5% sodium thiosulfate for 5 min, both at room
temperature, then the sections were counterstained with 0.6% thionin
in 0.2 M acetic acid buffer. The processed sections were mounted on
gelatinized glass slides, covered with a buffered glycerol solution
(containing 0.1% p-phenylenediamine for fluorescent-stained
sections to prevent fading), and protected with coverslips.
|
Data Analysis
Diaminobenzidine- and fluorescent-stained sections were examined with bright-field and fluorescence microscopy, respectively. Labeled neuronal somata and fibers were placed on a computerized map of the rat's brain stem (23a); a color code was used to differentiate single- and double-labeled neuronal somata. Labeled neurons were counted in medullary sections obtained at 0.5-mm intervals starting at the medullary-spinal junction, and the results were analyzed for the effects of injection site, vagotomy, and section level with an ANOVA with replication. Fisher's protected least significant difference was used for a posteriori comparisons. Increases in total lung resistance during vagal stimulation were compared for differences between lungs or the side of the stimulated vagus also with an ANOVA. All the photomicrographs used in the illustrations were scanned digitally into Adobe Photoshop format for mounting and labeling. Only color and contrast were adjusted to reproduce the conditions viewed under the microscope.| |
RESULTS |
|---|
|
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
|
|---|
Midline Crossing by Vagal Motor and Sensory Fibers
Characteristics of neuronal and sensory fiber labeling by lung
injections of CT-.
Neuronal somata and sensory fibers were labeled in the medulla of all
but 3 of the 32 rats that survived the 9- to 11-day period allowed for
transport of CT- from the lung injection site to the central nervous
system. Deaths usually occurred within 12 h of the surgery,
bearing no apparent relationship in their frequency to the side of the
injection or to whether the rat had undergone a cervical vagotomy.
into one lung
caused labeling of surprisingly similar numbers of neuronal somata on
both sides of the medulla (Fig. 2).
Although labeling density varied substantially from rat to rat, the
total number of CT--labeled neurons was consistently greater after
left than after right lung injections. Labeled neurons were grouped in
two distinctive nuclear divisions. The main or ventral division
consisted of large, often multipolar, neurons, which were clustered in
the rostralmost portion of the nucleus ambiguus proper, corresponding to the compact formation of the nucleus ambiguus (5), or
distributed more loosely in the region of the lateral tegmental field
immediately ventral to the nucleus (Fig.
3), in the subdivision known as the external formation in the same nomenclature. Although isolated neurons
were labeled caudally in the area often referred to as nucleus
retroambigualis, the majority of the neurons labeled in the nucleus
ambiguus complex were aligned rostrocaudally in a 1.5-mm-long column
rostral to the obex (from 13 to 11.5 mm caudal to the bregma). The
second or dorsal division of retrogradely labeled neurons was
considerably sparser and included smaller, frequently bipolar neuronal
somata located in the dorsal region of the dorsal motor nucleus of the
vagus.
|
|
injection, but in some sections, differences between
sides were not apparent. Labeled sensory terminals were more abundant
below the obex (Fig. 4), concentrating in
the commissural subnucleus, in the caudalmost sections, and in the
medial and ventrolateral subnuclei of the nucleus of the tractus
solitarius, more rostrally.
|
Effect of unilateral vagotomy on neuronal and sensory fiber labeling. Cervical vagotomies prevented labeling of neuronal somata and sensory fibers on the ipsilateral medulla but did not affect labeling in the contralateral medulla (Figs. 2-4). The vagotomies had no apparent effect on the proportions of neuronal somata labeled in the nucleus ambiguus complex and dorsal motor nucleus of the vagus or, within the nucleus ambiguus, on the partition of labeled neurons between the compact and external formations.
Effect of thoracic vagotomy on response to vagal stimulation and
CT- labeling after lung injections of CT-.
Electrical stimulation of the distal stump of the cervical vagus
increased the resistances opposed by both lungs to airflow (Fig.
5). The increases were maximal at
stimulation frequencies of 30-40 Hz and were more pronounced in
the ipsilateral than in the contralateral lung (2.1 ± 0.2 and
2.9 ± 0.7 times greater, depending on whether the left or the
right vagus was stimulated) and during left than during right vagal
stimulation. Interruption of the right vagus nerve caudal to the origin
of the recurrent laryngeal nerve reduced the responses of both lungs to
stimulation of the right cervical vagus (left lung by 75% and right
lung by 59%), without changing their responses to stimulation of the
left cervical vagus. The right thoracic vagotomy reduced substantially, but did not eliminate, neuronal labeling in the right nucleus ambiguus
(data not shown) after right lung injections of CT-. We therefore
concluded that the residual increases in airflow resistance produced by
cervical vagal stimulation after thoracic vagotomy were elicited by
bronchomotor fibers exiting the vagus with the recurrent laryngeal
nerve or rostral to it.
|
Bilateral Innervation of Airways by Vagal Motoneurons
Labeling with CT- and FG was found to coexist in 13 of the 18 rats that survived the appointed period of 11 days after the injections
(6 injected with FG into the left bronchus and CT- into the right
bronchus and 7 injected with the alternative scheme). Although there
were no differences in the anatomic organization of the neurons labeled
by the two retrograde neuronal markers (Fig.
6), considerably fewer neurons were
labeled by CT- than by FG. The topographic distribution of the
neuronal somata labeled by the bronchial injections was
undistinguishable from that produced by the injection of CT- into
the lungs.
|
Double labeling (Fig. 7), indicating
uptake of both markers by the same neuron, was observed in one or more
sections in all the rats. Of all the neurons labeled by CT- in the
nucleus ambiguus, 30% were also labeled by FG; conversely, of all the
nucleus ambiguus neurons labeled by FG, 11% were also labeled by
CT-.
|
Interconnections Between Vagal Medullary Centers
We found extensive neuronal labeling indicating transsynaptic infection with pseudorabies virus on both sides of the brain stem in 13 of the 24 rats that survived 5 days after injection of the virus (Fig. 8). In addition to areas known to contain vagal preganglionic neurons (nucleus ambiguus and dorsal motor nucleus of the vagus), virus-infected neurons were present in the nucleus of the tractus solitarius (including the subnuclei found to contain labeled sensory fiber terminals after lung injections of CT-), midline reticular nuclei (raphe pallidus, raphe obscurus,
parapyramidal, and raphe magnus nuclei), ventrolateral medulla (lateral
paragigantocellular, gigantocellular, and retrotrapezoid nuclei), and
noradrenergic nuclei of the medulla and pons (A5, locus ceruleus, and
locus subceruleus). There were no labeled neurons in the gracilis or cuneate nucleus, confirming previous reports that, at the viral doses
used in the experiments, the virus does not infect sensory pathways
(26).
|
Unilateral cervical vagotomies reduced but did not eliminate the
presence of virus-infected neurons in the nucleus ambiguus and dorsal
motor nucleus of the vagus ipsilateral to the cut vagus nerve (Fig.
9). Simultaneous staining with antibodies
against choline acetyltransferase confirmed the cholinergic identity of these neurons (Fig. 10).
|
|
Controls for Selectivity of Tracer Injections
Distribution of 125I-CT- injected into the lung.
Autoradiographic analysis confirmed that, although the
injected 125I-CT- was almost entirely removed
from the lung after 7 days, within the lungs, the tracer did not
migrate beyond the immediate vicinity of the injection sites (Fig.
11). No accumulation of radioactivity was found in the contralateral lung, the trachea, or the esophagus. The
ratio of total lung count measured in homogenates of injected and
contralateral lungs removed at 8 h to that removed at 24 h was <0.02.
|
Labeling of vagal medullary neurons by intrapleural CT-.
None of the six rats injected with CT- suspension into the pleural
space exhibited labeling of neuronal somata or sensory fiber terminals
in the medulla (data not shown).
Distinction between pulmonary and esophageal motoneurons.
Injections of CT- into the rostral and medial lobes of the right
lung and FG into the wall of the midthoracic esophagus produced concurrent labeling in the nucleus ambiguus and dorsal nucleus of the
vagus in 7 of the 12 rats that survived the procedure. No double
labeling was detected in any of these rats (Fig.
12). The somata of
the majority of the neurons labeled by the injections of CT- into
the lungs were once again distributed bilaterally between the compact
and the external formations of the nucleus ambiguus in the sections
located between 13.5 and 11.5 mm caudal from the bregma. The somata of
the neurons labeled by the injections of FG into the esophageal wall
were also found bilaterally, but they were circumscribed to the
compact formation, where they were interspersed with the
CT--labeled neurons. Esophageal neurons adopted a more rostral
distribution than the pulmonary neurons; they were found rarely in
medullary sections >12.5 mm caudal to the bregma.
|
Distinction between motoneurons innervating extra- and
intrathoracic trachea.
Labeling by CT- and FG injected into extra- and intrathoracic
segments of the trachea coexisted in all six rats included in this
control experiment. We found no double-labeled neuronal somata in the
nucleus ambiguus or dorsal motor nucleus of the vagus in any of these
rats (Fig. 13).
|
| |
DISCUSSION |
|---|
|
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
|
|---|
The autonomic innervation of bilateral organs that are functionally interdependent must contain mechanisms to prevent asymmetries in neural input from interfering with mechanical efficiency. In the case of the lungs, the mechanism appears to be based in part on a system of double innervation, whereby each lung receives supplies of motor and sensory nerve fibers from both vagus nerves. In the present study we demonstrate that 1) as many as one-half of the vagal motor fibers and a smaller, but still substantial, proportion of the sensory fibers destined for the rat's lungs cross the midline, and they do so exclusively inside the thorax; 2) at least for the main stem bronchi, a single medullary motoneuron can innervate airways on both sides of the midline; and 3) medullary vagal motoneurons that project to the lungs receive inputs from an extensive bilateral neuronal network that includes sensory relay areas in the nucleus tractus solitarius and the contralateral nucleus ambiguus and dorsal motor nucleus of the vagus.
Midline Crossing by Vagal Motor Fibers
The counts of neurons labeled by lung injections of CT- suggest
that each lung receives a unexpectedly symmetrical supply of vagal
motor fibers from both sides of the medulla. This symmetry contrasts
with the ipsilateral preponderance of lung innervation reported by
Kalia and Mesulam (15), who used injections of horseradish peroxidase to analyze the topography of the medullary neurons that
innervate the cat's lungs and airways. These investigators found that,
when the tracer was injected into the right main stem bronchus, over
one-third of the neuronal somata labeled in the nucleus ambiguus and
dorsal motor nucleus of the vagus and a substantial portion of the
central sensory fiber terminals were on the left side of the medulla.
When the injections were made into the apical lobe of the right lung,
however, only dorsal motor nucleus neurons and sensory terminals were
labeled in the left side of the medulla. No labeled neurons were
present in the left nucleus ambiguus. Differences in the distributions
of lung nerves between cats and rats provide the most obvious
explanation for the discrepancy between these findings and ours. There
are, however, other potential explanations worthy of consideration.
First, labeling of the contralateral nucleus ambiguus in our rats could
have resulted from transbronchial migration of the tracer to the
opposite lung. The distribution of the radiolabeled CT- injected
into the lung, however, makes this explanation unlikely. Bilateral
labeling of vagal neurons in the rats' medulla could have been also
the unsuspected consequence of contamination of the esophagus or the
heart by leaked CT-. The results of the control studies also render
this explanation improbable. Finally, it is possible that the
injections of Kalia and Mesulam were inefficient at labeling
contralateral neurons. This interpretation derives some support from
the results of the vagal stimulation experiments. Particularly when
considered in contrast to the symmetry with which both lungs are
represented in the vagal nuclei, the weak response of the contralateral
lung to unilateral vagal stimulation suggests that the midline-crossing motor fibers have a more limited arborization than their ipsilateral counterparts. If this is so, horseradish peroxidase, a less-sensitive retrograde tracer than CT- (33), may not have been
taken up by these fibers in amounts sufficient for the identification
of neuronal somata in the contralateral medulla. The consequences of
the tracer's reduced sensitivity may have been compounded by the
relatively large size of the cat's lungs. A wider separation between
the pleural surface and the lung hilum in this species may have placed
the lung injections at a greater distance from central airways (where
intrapulmonary parasympathetic ganglia concentrate) than in the rat.
Two additional observations of our study deserve mention in connection
with the lateral distribution of vagal fibers in the lungs. First, left
lung injections labeled a larger number of neurons in the nucleus
ambiguus (right and left) than right lung injections. Although we did
not compare the diffusion of CT- in the lungs quantitatively, we
believe that this disparity is likely to have resulted from a greater
spread of the neuronal tracer in the unilobar left lung than in the
multilobar right lung. Second, the increases in airflow resistance
elicited by unilateral vagal stimulation were greater in the
ipsilateral than in the contralateral lung. To interpret this
observation (or any other quantitative results obtained from vagal
stimulation), it is important to consider that lung resistance
increases during vagal stimulation are influenced by factors such as
the geometry of the airways (24) and the effectiveness
with which the electrical stimulus mimics physiological conduction in
the nerve. The latter, in particular, is influenced by the proportion
of myelinated and unmyelinated fibers present in each vagus. There is
convincing evidence that, in the guinea pig, capsaicin-responsive
fibers remain committed to the ipsilateral vagus nerve
(32). Thus, at least in this species, unilateral vagal
stimulation at the frequencies and pulse duration used in our
experiments (32) is likely to elicit greater contraction
of ipsilateral than contralateral airways. Tachykinins released
antidromically by ipsilateral unmyelinated sensory fibers in response
to the stimulus can be expected to sensitize local ganglia to the
effects of other depolarizing stimuli (4, 21)
and may cause bronchoconstriction directly (18). A similar
effect would be observed if vagal motor fibers were segregated to
provide preferential innervation to ipsilateral airway smooth muscle
targets (as opposed to glands or blood vessels) or established a larger
number of ganglionic synapses in the ipsilateral than in the
contralateral airways. Such arrangements would have the advantage of
allowing the vagal centers to retain individual control of the
cholinergic outflow to each lung without losing the benefits of a
bilateral distribution in the prevention of large asymmetries in airway tone.
Location of Lung Vagal Motoneurons and Central Afferents in the Medulla Oblongata
The organization of the lower airway motoneurons and central afferents has received surprisingly little attention in the neuroanatomic literature compared with other thoracic viscera, such as the esophagus (1-3, 5) or the heart (1, 5, 7, 9, 28, 30, 31). The precise origin of the airways' motor nerve supply was not examined with modern neuroanatomic methods until Kalia and Mesulam (15) demonstrated that the cat's trachea, bronchi, and lung parenchyma receive variable proportions of preganglionic fibers from neurons located in the compact and external formations of the nucleus ambiguus and in the dorsal motor nucleus of the vagus. Subsequent studies by Haxhiu et al. (11), Haxhiu and Loewy (12), Hadziefendic and Haxhiu (10), and our own laboratory (25) have confirmed the participation of the same two subdivisions of the nucleus ambiguus and, to a lesser extent than reported originally by Kalia and Mesulam, of the dorsal motor nucleus of the vagus in the innervation of the trachea and lungs in the rat, dog, ferret, and sheep.The participation of the compact formation of the nucleus ambiguus in
the innervation of the airway and lung tissues challenges some of the
views expressed by Bieger and Hopkins (5) in their authoritative analysis of the representation of the upper alimentary tract in the rat's medulla oblongata. These investigators asserted that this subdivision of the nucleus ambiguus is dedicated exclusively to the efferent innervation of striated esophageal muscle. The results
of our control experiments, especially the absence of double labeling
of neuronal somata by the concurrent injections of CT- into the lung
and FG into the esophagus, argue against any concerns that the
visualization of compact formation neurons after lung tracing
experiments was an artifact caused by spread of the tracer to the
adventitial surface of the esophagus. Far from negating that the
nucleus ambiguus is organized in a viscerotopic fashion, the
coexistence of esophageal and lung neurons in the same nuclear
subdivision is, in our opinion, consistent with the common embryologic
origin of the two organs and with the observation that a subpopulation
of esophageal parasympathetic ganglia participates directly in the
control of airway tone (6).
Kalia and Mesulam (15) were also the first to report that
afferent nerve fibers from the cat's right lung end bilaterally in the
nucleus of the tractus solitarius, concentrating in the ventrolateral,
dorsolateral, and commissural subnuclei of the nucleus of the tractus
solitarius. More recently, Haxhiu and Loewy (12), using
CT- as an anterograde tracer, described a slightly different
topographical termination pattern for the trachea's sensory nerves of
rats, ferrets, and dogs, whereby the majority of the sensory fibers
ended in the medial, ventrolateral, and commissural subnuclei of the
same nucleus. These findings, which are similar to those presented here
for the rat's lungs, verify that the airways and the esophagus have
anatomically separate sensory relays in the nucleus of the tractus
solitarius, with most of the esophageal fibers ending in the central
subnucleus (1). Evidence of this separation, which had
been noticed by Altschuler et al. (1) in their detailed
description of the viscerotopic sensory representation of the upper
alimentary tract in the rat, adds further support to the contention
that CT- injected into the lung did not spread to the esophagus in
our experiments.
Bilateral Innervation of Airways by Vagal Motoneurons
The presence of double-labeled neurons after the injections of CT- and FG into the main stem bronchi demonstrates that a single
vagal motoneuron can innervate airway structures located on both sides
of the midline. Although this finding reveals a previously unsuspected
characteristic of the vagal innervation of the airways, its validity is
restricted to the main stem bronchi. Our present results offer no proof
that a similar arrangement is in place for more distal airways.
Clearly, a considerable number of vagal motoneurons project to both
main stem bronchi. However, the observed discrepancy in the degree of
neuronal labeling by CT- and FG (most likely the result of the
greater tissue spread of the latter) limits our ability to draw a more
quantitative estimate of the number of cells with bilateral
projections. This discrepancy may reflect differences in the uptake,
transport, and immunoreactive characteristics of the two markers rather
than to nonspecific labeling by either of them. Except for the number
of labeled neurons, the distributions of the CT-- and FG-labeled
neurons were undistinguishable. Furthermore, the labeling disparity
occurred, despite the use of similar injectate volumes of CT- and FG
and the alternation of injection side. Finally, previous studies have
shown that neither CT- nor FG is incorporated by undamaged fibers of
passage (19, 27), which reduces the
possibility that FG simply labeled a greater number of fibers destined
for distal airways than CT-.
Interconnections Between Vagal Medullary Centers
Injection of pseudorabies virus into the right lung revealed an extensive bilateral network of infected brain stem neurons, even when all the efferent fibers from one side of the medulla were severed by unilateral cervical vagotomies. This network comprised first-order neurons (presumably preganglionic parasympathetic neurons) contralateral to the vagotomy and second- or higher-order premotor neurons in a variety of locations throughout the brain stem, including the nucleus of the tractus solitarius, the nucleus ambiguus, and the dorsal motor nucleus of the vagus ipsilateral to the vagotomy. The topographic distribution of the infected neurons was akin to those described after similar injections into rat and sheep tracheae and lungs (10, 11, 25). The pattern of medullary infection was also similar to that produced by injection of pseudorabies virus into the esophagus (3), denoting a considerable overlap in the regulatory networks of the respiratory and alimentary derivatives of the primitive foregut.A note of caution is pertinent in interpreting these results. Unlike some of the earlier experiments (10, 11, 25), our design did not include the interruption of the sympathetic pathways from the lung (11, 25). The obvious conclusion that preganglionic neurons receive regulatory inputs from both sides of the medulla must therefore be tempered by the possibility that some of the extensive bilateral labeling by the virus occurred through sympathetic nerves.
Nevertheless, retrograde passage via the sympathetic system is unlikely to explain the presence of the virus in the vagal nuclei. Injections of pseudorabies virus into structures such as the stellate ganglion or the adrenal gland, which are innervated by medullary sympathetic neurons, do not infect neurons in the ambigual complex or in the dorsal motor nucleus of the vagus (14). This leaves no plausible route for the infection of the vagal nuclei ipsilateral to the vagotomy other than the contralateral lung-projecting vagal motoneurons. The specific occurrence of interconnections between the nuclei ambigua is well documented by direct injections of retrograde markers into various portions of the ambiguus complex in the rat (22). Here we show that at least some of these interconnections may be involved in the coordination of vagal output to the lungs from both sides of the medulla. Our present data do not allow us to establish, however, whether the coordinating inputs are channeled through medullary interneurons, respiratory vagal motoneurons, or other vagal premotor neurons.
Perspectives
Our results demonstrate that each lung receives a bilateral supply of vagal motor and sensory fibers. This system of double innervation may serve to coordinate parasympathetic outflow to the lungs by 1) relaying unilateral lung stimuli to sensory neurons in both nuclei of the tractus solitarius, 2) coordinating preganglionic responses to these stimuli via interconnections between bilateral vagal motor centers, and 3) ensuring that each lung receives a balanced apportionment of the outflow from both vagal motor centers.| |
ACKNOWLEDGEMENTS |
|---|
The authors thank Prof. Arthur D. Loewy for advice and technical assistance.
| |
FOOTNOTES |
|---|
This work was supported by National Heart, Lung, and Blood Institute Grant HL-57998.
Address for reprint requests and other correspondence: J. J. Pérez Fontán, Dept. of Pediatrics, Washington University School of Medicine, St. Louis Children's Hospital, One Children's Place, St. Louis, MO 63110 (E-mail:FONTAN{at}KIDS.WUSTL.EDU).
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. §1734 solely to indicate this fact.
Received 15 January 2000; accepted in final form 14 March 2000.
| |
REFERENCES |
|---|
|
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
|
|---|
1.
Altschuler, SM,
Bao X,
Bieger D,
Hopkins DA,
and
Miselis RR.
Viscerotopic representation of the upper alimentary tract in the rat: sensory ganglia and nuclei of the solitary and the spinal trigeminal tracts.
J Comp Neurol
283:
248-268,
1989[ISI][Medline].
2.
Altschuler, SM,
Bao X,
and
Miselis RR.
Dendritic architecture of nucleus ambiguus motoneurons projecting to the upper alimentary tract in the rat.
J Comp Neurol
309:
402-414,
1991[ISI][Medline].
3.
Barrett, RT,
Bao X,
Miselis RR,
and
Altschuler SM.
Brain stem localization of rodent esophageal premotor neurons revealed by transneuronal passage of pseudorabies virus.
Gastroenterology
107:
728-737,
1994[ISI][Medline].
4.
Belvisi, MG,
Patacchini R,
Barnes PJ,
and
Maggi CA.
Facilitatory effects of selective agonist for tachykinin receptors on cholinergic neurotransmission: evidence for species differences.
Br J Pharmacol
111:
103-110,
1994[ISI][Medline].
5.
Bieger, D,
and
Hopkins DA.
Viscerotopic representation of the upper alimentary tract in the medulla oblongata in the rat: the nucleus ambiguus.
J Comp Neurol
262:
546-562,
1987[ISI][Medline].
6.
Canning, BJ,
and
Undem BJ.
Evidence that antidromically stimulated vagal afferents activate inhibitory neurones innervating guinea-pig trachealis.
J Physiol (Lond)
480:
613-625,
1994[ISI].
7.
Ciriello, J,
and
Calaresu FR.
Medullary origin of vagal preganglionic axons to the heart of the cat.
J Auton Nerv Syst
5:
9-22,
1982[ISI][Medline].
8.
Fox, EA,
and
Powley TL.
False-positive artifacts of tracer strategies distort autonomic connectivity maps.
Brain Res Rev
14:
53-77,
1989[Medline].
9.
Geis, GS,
and
Wurster RD.
Horseradish peroxidase localization of cardiac vagal preganglionic somata.
Brain Res
182:
19-30,
1980[ISI][Medline].
10.
Hadziefendic, S,
and
Haxhiu MA.
CNS innervation of vagal preganglionic neurons controlling peripheral airways: a transneuronal labeling study using pseudorabies virus.
J Auton Nerv Syst
76:
135-145,
1999[ISI][Medline].
11.
Haxhiu, MA,
Jansen ASP,
Cherniack NS,
and
Loewy AD.
CNS innervation of airway-related parasympathetic preganglionic neurons: a transneuronal labeling study using pseudorabies virus.
Brain Res
618:
115-134,
1993[ISI][Medline].
12.
Haxhiu, MA,
and
Loewy AD.
Central connections of the motor and sensory vagal systems innervating the trachea.
J Auton Nerv Syst
57:
49-56,
1996[ISI][Medline].
13.
Honjin, R.
Experimental degeneration of the vagus and its relation to the nerve supply of the lung of the mouse, with special reference to the crossing innervation of the lung by the vagi.
J Comp Neurol
106:
1-19,
1956.
14.
Jansen, ASP,
Nguyen XV,
Karpitskiy V,
Mettenleiter TC,
and
Loewy AD.
Central command neurons of the sympathetic nervous system: basis of the fight-or-flight response.
Science
270:
644-646,
1995
15.
Kalia, M,
and
Mesulam MM.
Brain stem projections of sensory and motor components of the vagus complex in the cat. II. Laryngeal, tracheobronchial, pulmonary, cardiac, and gastrointestinal branches.
J Comp Neurol
193:
467-508,
1980[ISI][Medline].
16.
Larsell, O,
and
Mason ML.
Experimental degeneration of the vagus nerve and its relation to the nerve terminations in the lung of the rabbit.
J Comp Neurol
33:
509-516,
1921.
17.
Lencer, WI,
Chu S-HW,
and
Walker WA.
Differential binding kinetics of cholera toxin to intestinal microvillus membrane during development.
Infect Immun
55:
3126-3130,
1987
18.
Lundberg, JM,
Brodin E,
and
Saria A.
Effects and distribution of vagal capsaicin-sensitive substance P neurons with special reference to the trachea and lungs.
Acta Physiol Scand
119:
243-252,
1983[ISI][Medline].
19.
Luppi, P-H,
Fort P,
and
Jouvet M.
Iontophoretic application of unconjugated cholera toxin B subunit (CTb) combined with immunohistochemistry of neurochemical substances: a method for transmitter identification of retrogradely labeled neurons.
Brain Res
534:
209-224,
1990[ISI][Medline].
20.
Mead, J,
and
Whitteberger JL.
Physical properties of human lungs measured during spontaneous respiration.
J Appl Physiol
5:
779-796,
1953.
21.
Myers, AC,
and
Undem BJ.
Electrophysiological effects of tachykinins and capsaicin on guinea-pig bronchial parasympathetic ganglion neurones.
J Physiol (Lond)
470:
665-679,
1993
22.
Nuñez Abades, PA,
Portillo F,
and
Pasaro R.
Characterisation of afferent projections to the nucleus ambiguus of the rat by means of fluorescent double labeling.
J Anat
172:
1-15,
1990[ISI][Medline].
23.
Olsen, CR,
Colebatch HJH,
Mebel PE,
Nadel JA,
and
Staub NC.
Motor control of pulmonary airways by nerve stimulation.
J Appl Physiol
20:
202-208,
1965
23a.
Paxinos, G,
and
Watson C.
The Rat Brain in Stereotaxic Coordinates. San Diego, CA: Academic, 1997.
24.
Pérez Fontán, JJ,
and
Kinloch LP.
Control of bronchomotor tone during perinatal development in sheep.
J Appl Physiol
75:
1486-1496,
1993
14.5 and B 
