Department of Pharmacology, University of Virginia,
Charlottesville, Virginia 22908
We examined the effects of
destroying bulbospinal catecholaminergic neurons with the immunotoxin
anti-dopamine 
-hydroxylase-saporin (anti-DH-Sap) on splanchnic
nerve activity (SNA) and selected sympathetic reflexes in rats.
Anti-DH-Sap was administered into the thoracic spinal cord with the
retrograde tracer fast blue. After 3-5 wk, anti-DH-Sap
eliminated most bulbospinal C1 (>74%), C3 (~84%), A5 (~98%),
and A6 cells. Noncatecholaminergic bulbospinal neurons of the rostral
ventrolateral medulla and serotonergic neurons were spared. Under
chloralose anesthesia, mean arterial pressure and heart rate of
anti-DH-Sap-treated rats (3-5 wk) were normal. Resting SNA was
not detectably altered, but the baroreflex range and gain were reduced
~40% (P < 0.05). Phenyl biguanide-induced decreases
in mean arterial pressure, heart rate, and SNA were unchanged by
anti-DH-Sap, but the sympathoexcitatory response to intravenous
cyanide was virtually abolished (P < 0.05). Rats that
received spinal injections of saporin conjugated to an anti-mouse IgG
had intact bulbospinal C1 and A5 cells and normal physiological responses. These data suggest that C1 and A5 neurons contribute modestly to resting SNA and cardiopulmonary reflexes. However, bulbospinal catecholaminergic neurons appear to play a prominent sympathoexcitatory role during stimulation of chemoreceptors.
 |
INTRODUCTION |
THE ROSTRAL
VENTROLATERAL MEDULLA (RVLM) is an essential structure for
generation of sympathetic vasomotor tone and a necessary central
component of many sympathetic reflexes (5, 6,
13, 16, 23, 25,
30). The RVLM controls sympathetic tone via bulbospinal
neurons (henceforth called presympathetic neurons) that project to
spinal sympathetic preganglionic neurons (20). Presympathetic neurons display spontaneous activity that is tightly correlated to vasomotor sympathetic nerve discharge (18)
and generally display a pattern of discharge highly reminiscent of that
of individual vasoconstrictor or cardiac sympathetic efferents (2, 5, 27). Two-thirds of the
presympathetic RVLM neurons have been identified as C1 cells, whereas
others are not catecholaminergic (14, 21).
The relative roles of these two populations of presympathetic neurons
in the generation of sympathetic tone and the production of sympathetic
reflexes through the RVLM have been difficult to determine. In many
cases, such as the arterial baroreceptor reflex and the Bezold-Jarisch
reflex, changes in the activities of C1 and non-C1 RVLM neurons
resemble the responses seen in sympathetic nerves (18,
30). Selective removal of either of these classes of
presympathetic neurons has not been possible, because the C1 cells are
insensitive to the classic catecholaminergic neurotoxin 6-hydroxydopamine (10), and a marker specific for the
noncatecholaminergic neurons has not been identified.
The recent development of an immunotoxin, saporin, conjugated to an
antibody for dopamine--hydroxylase (DH, DH-Sap), promises to
provide an effective tool for the selective elimination of C1 neurons
in the RVLM. Indeed, microinjection of anti-DH-Sap into the cerebral
ventricles or directly into the RVLM appears to effectively lesion
neurons of the C1 cell group (15, 32). However, with either of these approaches, the elimination of C1 cells
is not selective for the presympathetic portion of the C1 cell group.
In addition, injection of the toxin into the RVLM produces an
unavoidable measure of nonselective local damage (e.g., gliosis) to the
area of interest (15).
In the present study we sought to determine whether
microinjection of anti-DH-Sap into the thoracic spinal cord, a
primary terminal projection site for C1 cells, would produce an
effective and selective lesion of the bulbospinal C1 cells in the RVLM. We then used this model to determine the importance of presympathetic C1 RVLM neurons for the generation of sympathetic tone and the production of several sympathetic reflex responses that are mediated through the RVLM.
| |
MATERIALS AND METHODS |
Animals.
Male Sprague-Dawley rats (250-275 g; Hilltop Laboratories,
Scottdale, PA) were housed in groups (4 per cage) with a 12:12-h light-dark cycle and food and water available ad libitum. All procedures were performed in accordance with National Institutes of
Health and University of Virginia animal care and use guidelines.
Microinjections of saporin conjugates and fast blue into the
thoracic spinal cord.
Anesthesia was induced with 5% halothane (in 100% O2).
During surgery, halothane was applied through a nose cone (1.9% in 100% O2). The rat was placed in a stereotaxic apparatus
with the mouthpiece set at 11 mm below the interaural line. After a
dorsal laminectomy, the dorsal process of a lower thoracic spinal
vertebra was clamped to stabilize the upper thoracic spinal cord. The
spinal cord was exposed between vertebrae by incising the overlying
disks and meninges. Microinjections of anti-DH-Sap (42 ng/200 nl per injection; Chemicon, Temecula, CA) were made bilaterally at two levels
of the thoracic spinal cord: T2 and T4
(n = 4) or T4 and T6
(n = 4). In another four rats, anti-DH-Sap was
administered at 21 ng/100 nl per injection bilaterally at
T2 and T4. In addition, a subset of these rats
(n = 6) received bilateral microinjections of fast blue (FB,
200 nl of a 2% solution in sterile isotonic saline; Sigma Chemical,
St. Louis, MO) into alternate thoracic spinal levels (T3
and T5 or T5 and T7). An additional
group of rats (n = 5) received microinjections of FB
only (at T3 and T5 or at T5 and
T7). To control for potential nonspecific damage caused by
anti-DH-Sap, another group of rats (n = 6) received bilateral microinjections of saporin conjugated to a goat anti-mouse IgG (IgG-Sap, 42 ng/200 nl per injection at T2 and
T4; Chemicon). All microinjections were directed toward the
interomediolateral cell column (0.8 mm lateral to the midline and 1.0 mm ventral to the dorsal surface). After completion of the
microinjections, the wound was closed and the rats were placed on a
warm pad to maintain body temperature during recovery. All rats were
treated postoperatively with an analgesic (buprenorphine, 4 µg/kg im; Buprenex, Reckitt and Colman, Richmond, VA) and an antibiotic (penicillin G procaine, 7,500 U/kg im; G. C. Hanford, Syracuse, NY).
Physiological experiments.
Rats were studied 3-5 wk after spinal microinjections. Anesthesia
was induced by 5% halothane (in 100% O2). Rats were
intubated and artificially ventilated with 1.5-1.8% halothane in
100% O2 for surgical procedures. A brachial artery was
cannulated to record mean arterial pressure (MAP) and heart rate (HR),
and a brachial vein was cannulated to administer anesthetic and
paralytic agents. A femoral vein was cannulated for administration of
drugs used to elicit reflex responses. An inflatable snare was placed
around the abdominal aorta just below the diaphragm to permit rapid
control of upper body MAP (21). The left splanchnic nerve
was isolated via a retroperitoneal approach, and the segment distal to
the suprarenal ganglion was placed on two Teflon-coated silver wires that had been bared at the tip (250 µm bare diameter; A-M Systems, Everett, WA). The nerve and wires were embedded in a dental impression material (polyvinylsiloxane; Darby Dental Supply, Westbury, NY), and
the wound was closed around the exiting recording wires.
On completion of surgery, the halothane anesthesia was terminated and
was replaced by 
-chloralose (30 mg/kg solution in 3% sodium borate;
70 mg/kg initial bolus followed by hourly supplements of 20 mg/kg iv;
Fisher Scientific, Pittsburgh, PA). Rats were allowed to stabilize for
45 min before reflex tests began. End-tidal CO2 was
monitored by infrared spectroscopy and was maintained between 3.5 and
4.0% (11). Body temperature (measured rectally) was
maintained at 37°C. Ten minutes before reflex tests began, rats were
paralyzed with pancuronium bromide (1 mg/kg iv; Elkins-Sinn, Cherry
Hill, NJ).
All physiological variables were monitored on a chart recorder (model
RS 3600, Gould, Valley View, OH) and simultaneously stored on a
videocassette recorder via a digitizer interface (model 3000A,
frequency range: DC-22 kHz; Vetter Digital, Rebersburg, VA) for
off-line computer analysis. The MAP was calculated from the pulse
pressure measured by a transducer (Statham P10 EZ, Gould) connected to
the brachial arterial catheter. The HR was determined by triggering
from the pulse pressure (Biotach, Gould). Splanchnic nerve activity
(SNA) was filtered (10 Hz-3 kHz band pass with a 60-Hz notch
filter), full-wave rectified, and averaged in 1-s bins
(9). The baseline SNA (100%) was arbitrarily defined as activity during the resting state immediately preceding each
physiological test, and the minimum SNA (0%) was determined after
injection of clonidine (10 µg/kg iv; Sigma Chemical) at the end of
the study (9).
All animals underwent a series of reflex tests that were performed in
the same order separated by 5 min with drug doses established in pilot
experiments. The drugs were prepared in sterile isotonic saline for
injection in 50-µl volumes. The femoral venous catheter (dead space
100 µl) was loaded with each test solution and was flushed with 200 µl of saline to expel the drug. The Bezold-Jarisch reflex was
elicited with phenyl biguanide (5 and 20 µg/kg iv; Aldrich Chemical,
Milwaukee, WI). A chemoreflex was produced with sodium cyanide (100 and
200 µg/kg; Mallinckrodt, Paris, KY). The arterial baroreceptor reflex
was examined by raising arterial pressure (AP) with the abdominal
aortic snare and lowering AP with sodium nitroprusside (10 µg/kg iv;
Sigma Chemical). At the end of the experiment the animal was deeply
anesthetized with halothane and transcardially perfused with PBS (250 ml, pH 7.4) followed by 500 ml of 4% phosphate-buffered formaldehyde.
Histology.
The brain stem and thoracic spinal cord were removed and stored in
fixative overnight at 4°C. Brain stems and spinal cords were cut
using a Vibratome (30-µm coronal sections and 50-µm horizontal sections, respectively) and stored in a cryoprotectant solution at
20°C (21). The brain stem sections were later
processed for visualization of one of the following cellular markers:
tyrosine hydroxylase (TH),
phenylethanolamine-N-methyltransferase (PNMT), tryptophan
hydroxylase (TrypH), or Nissl substance. In each case, one of every six
sections was used. Solutions were prepared in Tris-buffered saline
(TBS; 0.1 M Tris, pH 7.4) and used at room temperature unless indicated
otherwise. On removal from cryoprotectant solution, all sections were
rinsed in TBS and blocked for 30 min in 10% normal goat serum (NGS;
GIBCO BRL, Grand Island, NY). Immunohistochemical detection of TH was
performed by incubation with a mouse monoclonal antibody (1:2,000 with
10% NGS and 0.1% Triton X-100 overnight at 4°C; Chemicon) followed
by a biotinylated goat anti-mouse IgG (1:200 for 45 min; Vector
Laboratories, Burlingame, CA) and then streptavidin indocarbocyanine
(Cy3, 1:1,000 for 1 h; Jackson Immunoresearch Laboratories, West
Grove, PA). Immunohistochemical detection of PNMT was performed by
incubation with a rabbit polyclonal antibody (1:2,000 with 10% NGS and
0.1% Triton X-100 overnight at 4°C; DiaSorin, Stillwater, MN)
followed by a biotinylated goat anti-rabbit IgG (1:200 for 45 min;
Vector Laboratories) and then streptavidin Cy3 (1:1,000 for 1 h;
Jackson Immunoresearch Laboratories). Immunohistochemical detection of
TrypH was performed by incubation with a mouse monoclonal antibody
(1:2,000 with 10% NGS and 0.1% Triton X-100 overnight at 4°C; Sigma
Chemical) followed by a biotinylated rabbit anti-mouse IgG3
(1:200 for 45 min; Zymed Laboratories, South San Francisco, CA) and
then streptavidin Cy3 (1:1,000 for 1 h; Jackson Immunoresearch Laboratories). All sections were rinsed in TBS and then in phosphate buffer (0.1 M, pH 7.4) and mounted onto microscope slides. Coverslips were applied with Krystalon mounting medium (EM Diagnostic Systems, Gibbstown, NJ).
To stain for Nissl substance, sections were mounted onto gelatin-coated
slides and immersed in a series of alcohols and xylenes. Nissl
substance was revealed by 0.25% thionin. Coverslips were affixed with
DPX mounting medium (Aldrich Chemical).
Spinal cord sections were mounted onto slides, and coverslips were
applied with Krystalon mounting medium. The location of FB injection
sites was examined by fluorescence microscopy, and injection sites from
anti-DH-Sap and IgG-Sap were visualized by dark-field illumination.
Brain mapping, cell counting, and imaging.
Sections were examined by using a fluorescence microscope (Leitz,
Heidelberg, Germany) with an N2 filter to visualize the red Cy3
fluorescence and an A filter to visualize the FB. The locations of
Cy3-positive and/or FB-positive neuronal profiles were plotted along
with the outline of each section and several anatomic landmarks with
Neurolucida software (Microbrightfield, Colchester, VT) and a Ludl
motor-driven microscope stage (21). Neuroanatomic
nomenclature and planes of sections are according to Paxinos and Watson
(19). Cell counts included all neuronal perikaryal
profiles, regardless of whether a nucleus was detectable. The effects
of spinal microinjections were determined by generating simple areal
density ratios among groups (4). Counts were performed in
all sections from each series of one in six sections.
The locations of PNMT-immunoreactive (PNMT-ir) C1, C2, and C3 cells
were plotted in sections from 13.8 to 11.2 mm caudal to bregma (12 sections/animal). In the case of C1 cells, the ventral quadrant of both
sides of the section was plotted, and profile counts reflect an average
of the two sides of the medulla. Profile counts for C2 and C3 cells
reflect the total number of profiles in the dorsal half of each
section. In the subset of animals that received FB microinjections,
PNMT-ir profiles were also examined for the presence of FB. In the four
sections immediately caudal to the facial nucleus (12.8 to 11.8 mm
caudal to bregma), the locations of FB-positive profiles without PNMT
immunoreactivity in RVLM were plotted. The locations of
TrypH-immunoreactive (TrypH-ir) profiles were plotted in the ventral
half of sections from 11.6 to 11.2 mm caudal to bregma (3 sections/animal), and counts reflect the total number of profiles in
each section. The presence of FB in TrypH-ir profiles was also
examined. The locations of TH-immunoreactive (TH-ir) profiles in the A5
cell group were plotted in sections from 10.4 to 9.0 mm caudal to
bregma (8 sections/animal), and counts reflect an average from the two
sides of the pons. The presence of FB in A5 cell profiles was also examined.
Examples of bulbospinal PNMT-ir, TH-ir, and TrypH-ir cells with FB were
photographed with 35-mm film (1,600 ASA push process color slide film)
at ×250 magnification. Color 35-mm slides were scanned on a flatbed
scanner (1,000 dots/in.; Ultra Saphir, Young Phillips, Richmond, VA)
and acquired as Adobe Photoshop documents (Adobe Systems, Mountain
View, CA). Multiple photomicrograph figures were assembled in Photoshop
software, with the resolution adjusted so that the image sizes fit the
page. In Photoshop, the output levels were limited to the range of
levels containing pixels. Hue, contrast, and brightness were adjusted
to reflect the original image as much as possible. Lettering, scale
bars, and arrows also were added in Photoshop.
Data analysis.
The effects of treatment with anti-DH-Sap or IgG-Sap on the number
of catecholaminergic medullary neurons and various physiological measurements were determined by ANOVA followed by Tukey-Kramer post hoc
tests when significant F values were obtained. Counts for
TrypH-ir profiles and FB-positive/non-PNMT-ir profiles in RVLM in
control rats and rats treated with anti-DH-Sap were each compared by
one-way ANOVA with repeated measures. Because the counts for TrypH-ir
profiles were comparable at the three medullary levels examined within
each group, the data from these levels were pooled and are presented as
one mean value for each group.
An arterial baroreflex curve relating MAP and SNA was constructed for
each rat. The SNA at resting MAP was set at 100%, and SNA after
clonidine (10 µg/kg iv) was set as the minimum (i.e., 0) value.
Boltzman sigmoidal curves, with use of the equation SNA = (A1 A2)/{1 + exp [A3 · (MAP A4)]} + A2,
were fitted to the experimental data points by use of the software
program Origin (Microcal Software, Northampton, MA), where A1 defines the upper plateau of the curve, A2 defines the lower plateau of the
curve, A3 describes the distribution of the gain along the curve, and
A4 (MAP50) is the midpoint of the curve. Maximum gain (Gmax) was calculated using the formula
Gmax = A3 · (A1 A2)/4 (22).
| |
RESULTS |
Depletion of bulbospinal adrenergic cells by anti-DH-Sap.
In the control rats, PNMT-ir profiles (C1 cells) were found within the
ventral quadrants of all 12 medullary levels examined, as expected
(Fig. 1A; examples in Figs.
2A and
3A). The presence of FB in
PNMT-ir profiles showed that the spinally projecting C1 cells were
located at the rostral end of the C1 cell group, with 69% of the
PNMT-ir/FB-positive profiles located between 12.0 and 11.2 mm
caudal to bregma (Fig. 1B; examples in Fig. 2, A and B, and Fig. 3A), as previously shown
(8, 17, 24, 29). Rats treated with anti-DH-Sap had a massive depletion of
the rostral, bulbospinal C1 cell group (Fig. 1, A and
B; examples in Fig. 2, C and D, and
Fig. 3B). The depletion of C1 cells did not differ by
injection volume of anti-DH-Sap or thoracic spinal levels
injected, so the lesioned animals were combined for group histological
analyses. The percent depletion of PNMT-ir profiles between 12.0 and
11.2 mm produced the most reliable and closest estimate of the
depletion of bulbospinal PNMT-ir profiles (r = 0.97),
but with a consistent underestimation. This underestimation was
predicted, given that even the very rostral end of the C1 cell group
contains bulbospinal and nonspinally projecting C1 cells
(17, 24, 31). In the 12 rats
treated with anti-DH-Sap, counts of PNMT-ir profiles
between 12.0 and 11.2 mm revealed an average depletion of 71 ± 4% (43-89%). In the six lesioned rats also injected with FB,
counts at the same medullary levels revealed an average depletion of
61 ± 5% of PNMT-ir profiles but an average depletion of 73 ± 7% (48-94%) of bulbospinal PNMT-ir profiles. Therefore,
the average depletion of bulbospinal PNMT-ir cells in the 12 rats
treated with anti-DH-Sap is likely to be slightly higher
than 74%.
View larger version (28K):
[in this window]
[in a new window]
|
Fig. 1.
Effects of spinal microinjection of
anti-dopamine--hydroxylase-saporin (anti-DH-Sap) or anti-mouse
IgG-saporin (IgG-Sap) on the number of
phenylethanolamine-N-methyltransferase-immunoreactive
(PNMT-ir) neuronal profiles in ventral (A and B;
C1 cells) and dorsal (C and D; C2 and C3 cells)
rostral medulla. A: number of PNMT-ir profiles per ventral
quadrant (C1 cell profiles) at 12 medullary levels in rats that
received no spinal injection (control rats, n = 6), in
rats treated with anti-DH-Sap (n = 12), and in rats
treated with IgG-Sap (n = 5). Note the substantial
decrease in the number of rostral PNMT-ir profiles in the
anti-DH-Sap-treated rats. B: subset of rats in
A that also received spinal microinjections of the
retrograde tracer fast blue (FB). The FB-positive (FB+), PNMT-ir
profiles are restricted to the rostral C1 cells. The number of
PNMT-ir/FB-positive profiles is markedly reduced in rats treated with
anti-DH-Sap (n = 6) compared with
combined groups of control rats (n = 5) and rats
treated with IgG-Sap (n = 2). C: number of
PNMT-ir profiles per dorsal half (C2 and C3 cells) of the sections in
A. IgG-Sap had no effect on the number of PNMT-ir profiles,
but treatment with anti-DH-Sap produced a marked decrease
in the number of profiles of C3 neurons at the rostral medullary
levels. D: subset of the rats in C that also
received spinal microinjections of FB (same rats in B). The
number of FB-positive/PNMT-ir profiles, which are restricted to the C3
cell group, is substantially diminished by treatment with
anti-DH-Sap. Values are means ± SE.
|
|
View larger version (50K):
[in this window]
[in a new window]
|
Fig. 2.
Representative photomicrographs of 2 control rats
(A, B, E, and F) and 2 rats treated with
anti-DH-Sap (C, D, G, and H) showing PNMT-ir
C1 cells (A and C) and tryptophan
hydroxylase-immunoreactive (TrypH-ir) serotonergic neurons
(E and G). Retrograde label of bulbospinal
neurons produced by spinal FB injections is also shown (B,
D, F, and H). A: PNMT-ir
cells visualized with indocarbocyanine (Cy3) in a control rat at 11.8
mm caudal to bregma. B: the same area of the section in
A viewed to reveal FB. Arrows in A and
B are directed toward PNMT-ir profiles with FB, indicating
bulbospinal C1 cells. *, Bulbospinal, non-C1 neuronal profiles (FB
positive without PNMT). C: PNMT-ir cells visualized with Cy3
at 11.8 mm caudal to bregma in a rat treated with anti-DH-Sap.
D: the same area of the section in C viewed to
reveal FB. Arrows in C and D are directed toward
PNMT-ir profiles with FB, indicating bulbospinal C1 cells. Arrowheads
are directed toward FB particles remaining after depletion of
bulbospinal C1 cells by anti-DH-Sap. *, Bulbospinal, non-C1
neuronal profile (FB positive without PNMT) E: TrypH-ir
cells visualized with Cy3 to reveal serotonergic cells at 11.8 mm
caudal to bregma in a control rat. F: the same area of the
section in E viewed to reveal FB. Arrows in E and
F are directed toward TrypH-ir profiles with FB, indicating
bulbospinal serotonergic neurons. G: TrypH-ir cells
visualized with Cy3 in a rat treated with anti-DH-Sap. This section
is from the same medullary level as the section in E, and
the number of TrypH-ir neurons is comparable. H: the same
area of the section in G viewed to reveal FB. Arrows in
G and H are directed toward TrypH-ir profiles
with FB, indicating bulbospinal serotonergic neurons. Scale bar, 500 µm. py, Pyramidal tract. In A-D, dorsal is toward the
top of the photomicrographs, medial is toward the
right, and the ventrolateral edge of the medulla is at the
bottom left. In E-H, dorsal is toward the
top of the photomicrographs, and the ventral surface of the
medulla at the midline is centered at the bottom.
|
|
View larger version (18K):
[in this window]
[in a new window]
|
Fig. 3.
Distribution of PNMT-ir and FB-positive neuronal profiles
in a control rat (A) and a rat treated with anti-DH-Sap
(B) plotted using Neurolucida. Dashed lines, boundaries used
to define C1 cells in Fig. 1, A and B; solid
lines, triangular area used to define rostral ventrolateral medulla
(RVLM) for the purpose of counting non-PNMT-ir/FB-positive neuronal
profiles. For RVLM, the dorsal boundary is the center of the nucleus
ambiguus, the medial boundary is the lateral edge of the pyramidal
tract, the lateral boundary is the ventral edge of the spinal
trigeminal tract, and the ventral boundary is the ventral edge of the
section.  , PNMT-ir profiles without FB;  ,
PNMT-ir profiles with FB; ×, FB-positive profiles without PNMT. Small
dots represent the FB particles in rats treated with anti-DH-Sap,
presumably the result of elimination of bulbospinal C1 cells in the
absence of a means for degrading the retrograde tracer. Scale bar, 500 µm. Amb, nucleus ambiguus; ION, inferior olivary nuclei; py,
pyramidal tract; icp, internal capsule; sp5, spinal trigeminal tract.
|
|
At the caudal medullary levels (13.8 to 13.0 mm caudal to bregma)
the PNMT-ir profiles did not contain FB (Fig. 1, A and B), as expected (17, 24,
29), and the number of PNMT-ir profiles at these levels
was unaffected by treatment with anti-DH-Sap (Fig.
1A). Rats treated with IgG-Sap showed no depletion of C1 cell profiles at any medullary level examined (Fig. 1A).
Although most C1 neurons were physically eliminated from the rostral
tip of the RVLM by microinjection of anti-DH-Sap into the
spinal cord (cf. Fig. 2, A and B, with Fig. 2,
C and D), these cells constitute such a small
portion of the total neurons within the region that their disappearance
was not apparent by examination of Nissl substance (Fig.
4). In addition, because the
microinjections of toxin were made into the spinal cord, no gliosis or
any other readily observable cytological difference was detected in the RVLM (Fig. 4).
View larger version (148K):
[in this window]
[in a new window]
|
Fig. 4.
Representative photomicrographs of Nissl-stained sections
(revealed with thionin) containing the RVLM from a control rat
(A) and a rat treated with anti-DH-Sap (B).
Despite the depletion of bulbospinal C1 cells from the RVLM in the
anti-DH-Sap-treated rat, the Nissl stain reveals no obvious
differences between the 2 rats. Remaining cells appear normal, with no
evidence of gliosis. The cluster of large cells at the center of the
top of the photomicrographs is part of the nucleus ambiguus.
The ventral surface of the medulla can be seen at the bottom
of the photomicrographs. Scale bar, 100 µm.
|
|
Counts of PNMT-ir profiles in the dorsal adrenergic cell groups
revealed a pattern similar to that seen with the ventral PNMT-ir profiles. The more caudally located C2 neurons did not contain FB (Fig.
1, C and D), as expected (17,
24), and the number of these profiles was not altered by
treatment with anti-DH-Sap (Fig. 1C). In
contrast, the more rostrally located C3 neurons were markedly depleted
by anti-DH-Sap (58 ± 3% of PNMT-ir profiles between 12.0 and 11.2 mm; Fig. 1C). The presence of FB
in the PNMT-ir profiles indicated that 97% of the dorsal bulbospinal PNMT-ir neurons are located between 12.0 and 11.2 mm caudal to
bregma, as shown previously (17, 24). In the
six rats also injected with FB, anti-DH-Sap produced a
53 ± 2% depletion of PNMT-ir profiles and an 84 ± 3%
depletion of the bulbospinal PNMT-ir profiles in the C3 cell group
(Fig. 1D). These data indicate that many C3 neurons do not
project to the cord, and counts of PNMT-ir profiles alone greatly
underestimate the depletion of the bulbospinal portion of the C3 cell
group. Treatment with IgG-Sap did not affect the number of PNMT-ir
profiles at any dorsal medullary level examined (Fig. 1C).
Depletion of bulbospinal noradrenergic cells by anti-DH-Sap.
Although the aim of the present study was to examine the effect of
depletion of the bulbospinal C1 cells, the large intraspinal microinjections of anti-DH-Sap were expected to also
eliminate the bulbospinal noradrenergic cell groups within the pons
(3). Examination of the A5 cell group within the
ventrolateral pons revealed a near-total elimination of these
noradrenergic cell profiles (Fig. 5 and
Fig. 6, A-D). Counts
of TH-ir profiles from 8 pontine levels in 12 rats treated with
anti-DH-Sap revealed an 88 ± 2% depletion
(78-97%) compared with control rats (Fig. 5A). Counts
from the five rats also injected with FB revealed an 84 ± 3%
depletion of TH-ir profiles and a 98 ± 1% depletion (96-99%) of bulbospinal TH-ir profiles compared with control rats (Fig. 5B). Similar to the results obtained by counting
PNMT-ir cells in the C1 cell group, counts of TH-ir cells alone may
slightly underestimate the depletion of bulbospinal A5 neurons. No
significant correlation between the magnitude of depletion of the
bulbospinal TH-ir profiles of the A5 cell group and the bulbospinal
PNMT-ir profiles of the C1 cell group was observed.
View larger version (20K):
[in this window]
[in a new window]
|
Fig. 5.
Effects of spinal microinjection of anti-DH-Sap on the
number of tyrosine hydroxylase-immunoreactive (TH-ir) neuronal profiles
in the ventrolateral pons. A: number of TH-ir profiles per
ventral quadrant (A5 cells) at 8 pontine levels in control rats
(n = 5) and rats treated with anti-DH-Sap
(n = 11). The number of TH-ir profiles in the rats
treated with anti-DH-Sap was markedly reduced at all
pontine levels examined. B: a subset of rats in A
that also received spinal microinjections of FB. TH-ir profiles with FB
were found at all pontine levels examined. At all 8 medullary levels
examined, TH-ir profiles with FB were virtually eliminated in rats
treated with anti-DH-Sap (n = 6) compared
with control rats (n = 5).
|
|
View larger version (82K):
[in this window]
[in a new window]
|
Fig. 6.
Representative photomicrographs of 2 control rats
(A, B, E, and F) and 2 rats treated with
anti-DH-Sap (C, D, G, and H) showing TH-ir A5
cells (A and C) and TH-ir neurons in the locus
ceruleus (E and G). The retrograde label of
bulbospinal neurons produced by spinal FB injections is also shown
(B, D, F, and H).
A: TH-ir A5 cells visualized with Cy3 at 10.0 mm caudal to
bregma in a control rat. B: the same area of the section in
A viewed to reveal FB. Arrows in A and
B are directed toward TH-ir profiles with FB, indicating
bulbospinal A5 cells. C: TH-ir cells visualized with Cy3 at
10.0 mm caudal to bregma in a rat treated with anti-DH-Sap. No
TH-ir neuronal profiles are present in the A5 region. D: the
same area of the section in C viewed to reveal FB.
Arrowheads are directed toward FB particles remaining after depletion
of bulbospinal A5 cells by anti-DH-Sap. E: TH-ir cells in
the locus ceruleus visualized with Cy3 at 9.4 mm caudal to bregma in
a control rat. F: the same area of the section in
E viewed to reveal FB. Arrows in E and
F are directed toward TH-ir profiles with FB, indicating
bulbospinal noradrenergic neurons of the subceruleus. *, FB-positive
neuronal profile that is not noradrenergic. G: TH-ir cells
in the locus ceruleus at 9.4 mm caudal to bregma visualized with Cy3
in a rat treated with anti-DH-Sap. The majority of noradrenergic
neurons of the locus ceruleus remain intact after treatment with
anti-DH-Sap. H: the same area of the section in
G viewed to reveal FB. No FB-positive/TH-ir neuronal
profiles are present in the subceruleus. *, FB-positive neuronal
profile that is not noradrenergic. Arrowheads indicate FB particles
from bulbospinal noradrenergic neurons that have been depleted by
anti-DH-Sap. Scale bar, 500 µm. 7n, Facial nerve; LSO, lateral
superior olive; 4V, 4th ventricle.
|
|
Examination of TH-ir profiles in the locus ceruleus indicated that the
vast majority of these cells remained after treatment with
anti-DH-Sap (Fig. 6, E and G). Although the
depletion of bulbospinal TH-ir neurons within the subceruleus was not
quantified, treatment with anti-DH-Sap eliminated most of these
cells (cf. Fig. 6, E and F, with Fig. 6,
G and H), comparable to the depletion of A5
neurons. Bulbospinal cells within the subceruleus that were not TH-ir
were spared (Fig. 6, E and H).
Selectivity of the depletion of bulbospinal catecholaminergic
neurons by anti-DH-Sap.
We examined two populations of noncatecholaminergic bulbospinal neurons
to determine whether the toxicity of anti-DH-Sap was
selective for cells that make DH. First, we previously
showed that the population of putative presympathetic cells within the RVLM is comprised of C1 cells and noncatecholaminergic cells
(21). To determine whether anti-DH-Sap
selectively depleted the catecholaminergic portion of this population,
we counted the number of profiles within the RVLM that contained FB but
no PNMT immunoreactivity. Although, these profiles were not from
functionally characterized neurons, strict anatomic borders for RVLM
were used to provide an estimate of this population of putative
presympathetic neurons (Fig. 3). In control rats and rats treated with
anti-DH-Sap, FB-positive/non-PNMT-ir profiles were
counted bilaterally from four levels of the medulla (12.8, 12.3,
12.0, and 11.8 mm caudal to bregma; Fig. 3). The counts from the
two sides of each section were averaged for comparison with counts of
C1 cell profiles. The number of FB-positive/non-PNMT-ir profiles in
RVLM did not differ between control rats (n = 5, 14.9 ± 3.0 profiles/side) and rats treated with
anti-DH-Sap (n = 6, 15.5 ± 1.4 profiles/side) at any of the medullary levels examined.
Second, we estimated the density of bulbospinal serotonergic neurons in
the raphe pallidus of control rats and rats treated with
anti-DH-Sap. Serotonergic cells were revealed by
immunohistochemical detection of TrypH in animals that had been
injected with FB into the thoracic spinal cord (5 control rats and 5 rats treated with anti-DH-Sap). The medullary sections
from control rats (Fig. 2, E and F) were
indistinguishable from sections from rats treated with
anti-DH-Sap (Fig. 2, G and H).
Counts of profiles obtained from the three levels of the medulla
examined in each animal were averaged. The number of TrypH-ir profiles
in control rats (142.7 ± 14 profiles/section) was not different
from the number counted in rats treated with anti-DH-Sap
(141.3 ± 13 profiles/section). Similarly, the number of
bulbospinal TrypH-ir profiles in control rats (67 ± 9.6 profiles/section) and the number in rats treated with
anti-DH-Sap (52.3 ± 6.0 profiles/section) were comparable.
Effects of saporin conjugates and FB at the spinal cord.
Although the lesion produced by microinjection of
anti-DH-Sap appears to be specific for spinally
projecting adrenergic and noradrenergic neurons at the levels of the
medulla and pons, nonspecific damage at the sites of the injections did
occur. Examination of the thoracic spinal cords revealed that the
microinjections were centered approximately in the region of the
intermediolateral cell column in the upper thoracic spinal cord where
expected, but the injectate had clearly spread beyond this region into
the levels of the dorsal and ventral horns. The injectates apparently diffused predominantly along the length of the cord but did not reach
the midline or lateral edges of the cord. Injections of anti-DH-Sap, IgG-Sap, and FB produced patches of
nonspecific damage surrounded by areas of gliosis. The nature of the
damage at the center of the injections was often impossible to assess because the tissue did not cut well. However, an outline of the area of
detectable gliosis was plotted from several sections in representative
rats from each group to compare the size of the area of the gliosis
produced by the microinjections. The 100-nl injections of
anti-DH-Sap produced areas of gliosis (3.73 ± 0.18 mm rostrocaudally and 0.77 ± 0.4 mm mediolaterally) comparable to
those seen with the 200-nl injections of anti-DH-Sap
(4.91 ± 0.69 mm rostrocaudally and 0.92 ± 0.04 mm
mediolaterally). However, with 200-nl injection volumes, the tissue was
more fragile and difficult to cut. The 200-nl injections of IgG-Sap
produced areas of gliosis (4.95 ± 0.62 mm rostrocaudally and
0.88 ± 0.15 mm mediolaterally) comparable to those seen with
200-nl injections of anti-DH-Sap. Microinjection of FB
(200 nl) produced smaller areas of damage and gliosis (2.49 ± 0.22 mm rostrocaudally and 0.39 ± 0.02 mm mediolaterally) than
those seen with the 200-nl microinjections of anti-DH-Sap
(P < 0.001, t-tests). The midline and
lateral fibers of passage appeared intact in all spinal cords examined, and these were readily observable in rats injected with FB. Rats showed
no obvious deficits in locomotion after microinjections into the spinal cord.
Arterial baroreceptor reflex.
Microinjection of anti-DH-Sap or IgG-Sap into the
thoracic spinal cord produced no detectable change in baseline MAP
(Tables 1 and
2) or HR (Table 2) in rats anesthetized
with chloralose, artificially ventilated, and paralyzed. Baseline SNA
was not noticeably altered by treatment with anti-DH-Sap; the signal
was observed with the same degree of amplification in all groups of
animals. However, analysis of the relationship between MAP and SNA
(baroreflex curves) revealed an effect of the treatment with
anti-DH-Sap. Although the baroreflex operated around a
comparable MAP (MAP50 in Table 1) in all groups, in rats
treated with anti-DH-Sap, a greater percentage of SNA was
insensitive to increased MAP (lower plateau in Table 1, examples in
Fig. 7, A and B).
Thus the operating range and the gain of the baroreflex (range and
Gmax in Table 1) were decreased in rats treated with
anti-DH-Sap. These differences in the arterial
baroreceptor reflex were not seen in rats treated with IgG-Sap (Table
1, Fig. 7C).
View this table:
[in this window]
[in a new window]
|
Table 1.
Average parameters of sigmoidal baroreflex curves for control rats and
rats treated with anti-DH-Sap or IgG-Sap
|
|
View this table:
[in this window]
[in a new window]
|
Table 2.
Effects of phenyl biguanide on MAP, HR, and SNA in control rats and
rats treated with anti-DH-Sap or IgG-Sap
|
|
View larger version (18K):
[in this window]
[in a new window]
|
Fig. 7.
Baroreflex curves in a control rat (A), a rat
treated with IgG-Sap (B), and a rat treated with
anti-DH-Sap (C). Curves relating splanchnic nerve
activity (SNA) and mean arterial pressure (MAP) were generated by
lowering MAP with nitroprusside and increasing MAP by constricting an
abdominal aortic snare. Baseline SNA was set to 100%, and minimum SNA
was determined after intravenous injection of clonidine (10 µg/kg).
Details for generating the sigmoidal curve that best fit the data are
stated in MATERIALS AND METHODS. In rats treated with
anti-DH-Sap, the slope of the baroreflex curve is reduced as a
result of a reduced range of the SNA that is sensitive to changes in
MAP. These changes were not seen in rats treated with IgG-Sap. See
Table 1 for group data.
|
|
Phenyl biguanide-induced Bezold-Jarisch reflex.
Phenyl biguanide produced a marked decrease in MAP, HR, and SNA in all
control rats in a dose-dependent manner (Fig.
8, A-C, Table 2). These
responses were not diminished by treatment with IgG-Sap or
anti-DH-Sap (Table 2). In rats treated with
anti-DH-Sap, the lower dose of phenyl biguanide produced
larger decreases in HR and MAP with a longer time to recover MAP to
baseline levels (Fig. 8, D-F, Table 2).
View larger version (27K):
[in this window]
[in a new window]
|
Fig. 8.
Phenyl biguanide-induced changes in MAP, SNA, and HR in a
control rat (A-C) and a rat treated with anti-DH-Sap
(D-F). Intravenous phenyl biguanide produced a decrease
in MAP, SNA, and HR in the control rat. In the rat treated with
anti-DH-Sap, phenyl biguanide also produced a marked decrease in
MAP, SNA, and HR. Arrows, onset of phenyl biguanide injections. See
Table 2 for group data.
|
|
Cyanide-induced changes in SNA and AP.
In control rats, intravenous injection of sodium cyanide produced a
burst in SNA, which was accompanied by a small rise in MAP (Fig.
9, A-D, and Fig.
10). Although the period of
increased SNA was brief compared with previously observed responses in
urethan-anesthetized, vagotomized rats (11,
12), this SNA response was observed in all control rats.
The increases in SNA and MAP were always followed by a decrease in SNA
and MAP (Fig. 9, A-D, and Fig. 10). The two doses of
cyanide produced effects of comparable magnitude (Fig. 10), indicating
that the lower dose of cyanide (100 µg/kg) was maximally effective.
In rats treated with IgG-Sap, cyanide produced SNA and MAP responses
comparable to those seen in control rats (Fig. 9, E-H,
and Fig. 10). In contrast, in rats treated with anti-DH-Sap, the lower dose of cyanide no longer produced
a burst in SNA or a rise in MAP in any of the rats examined (Fig. 9,
I and K, and Fig. 10). With the higher dose of
cyanide, the rise in MAP was absent in all rats and the burst in SNA
was absent in all but two rats (Fig. 9, J and L,
and Fig. 10). The decreases in SNA and MAP in response to cyanide
persisted in rats treated with anti-DH-Sap (Fig. 9,
I-L, and Fig. 10).
View larger version (28K):
[in this window]
[in a new window]
|
Fig. 9.
Examples of cyanide-induced changes in SNA and MAP in a
control rat, a rat treated with IgG-Sap, and a rat treated with
anti-DH-Sap. A, C, E, G, I, and K: effects of
100 µg/kg iv of cyanide; B, D, F, H, J, and L:
effects of 200 µg/kg iv of cyanide. Both doses of cyanide produced an
increase in MAP and SNA that was followed by a decrease in both
parameters in the control rat (A-D) and the rat treated
with IgG-Sap (E-H). In the rat treated with
anti-DH-Sap (I-L), cyanide did not increase SNA or
MAP at either dose. However, the decreases in MAP and SNA produced by
cyanide were present in the rat treated with anti-DH-Sap. Arrows,
onset of cyanide injections.
|
|
View larger version (24K):
[in this window]
[in a new window]
|
Fig. 10.
Effects of cyanide on SNA and MAP in control rats, rats
treated with IgG-Sap, and rats treated with anti-DH-Sap.
Left: effects of the lower dose of cyanide (100 µg/kg);
right: effects of the higher dose of cyanide (200 µg/kg).
The 2 bars for each condition illustrate the peak of the initial
increase and the trough of the decrease in SNA and MAP in all control
rats (solid bars, n = 8 for the low dose and
n = 5 for the high dose), rats treated with IgG-Sap
(hatched bars, n = 5 for both doses), and rats treated
with anti-DH-Sap (open bars, n = 10 for the low dose
and n = 7 for the high dose). * Significantly
different from control rats at the same dose (P < 0.05).
|
|
| |
DISCUSSION |
The present study demonstrates that microinjection of the
immunotoxin anti-DH-Sap into the upper thoracic spinal cord produces an effective and highly selective depletion of bulbospinal
catecholaminergic neurons in rats. Treatment with anti-DH-Sap
produced a virtually complete disappearance of the noradrenergic
bulbospinal A5 and A6 neurons and destroyed the majority of the
bulbospinal C1 and C3 adrenergic cells. Noncatecholaminergic
bulbospinal neurons in the RVLM and serotonergic neurons of the raphe
nuclei were spared. Under chloralose anesthesia, the most prominent
deficit displayed by rats treated with anti-DH-Sap was a reduction
in their sympathoexcitatory response to intravenous cyanide, suggesting that bulbospinal catecholaminergic neurons play a key role in the
carotid chemoreflex. These rats also exhibited some reduction in the
gain and operating range of their sympathetic baroreflex. Given the
extent of the lesion of bulbospinal A5 and C1 cells, the latter changes
were not as large as anticipated. These results suggest that C1 and A5
cells may play a relatively minor role in generating the resting level
of sympathetic tone that maintains MAP, but bulbospinal
catecholaminergic neurons may be recruited when sympathetic vasomotor
tone is vigorously activated.
Selective depletion of noradrenergic and adrenergic neurons by
anti-DH-Sap.
The anti-DH-Sap conjugate was developed as a tool to selectively
destroy neurons that express DH on their plasma membrane (noradrenergic and adrenergic, but not dopaminergic, neurons). This
immunotoxin provides the first tool for selectively destroying adrenergic neurons (e.g., C1 cells), which are insensitive to the
classic catecholaminergic neurotoxin 6-hydroxydopamine
(10). Its specificity relies on the conjugation of the
ribosomal toxin Sap to an anti-DH antibody (32).
Because DH is vesicular and membrane bound, it is exteriorized
during exocytosis and acts as an extracellular membrane receptor for
the internalization of anti-DH-Sap. Once inside the cell, Sap
irreversibly inactivates the 60S subunit of ribosomes, which prevents
protein synthesis and eventually kills the cell (26).
Previous reports have shown that anti-DH-Sap effectively eliminates
cells when deposited in the region of the axon terminals (noradrenergic
neurons of the locus ceruleus) (1) or even the cell bodies
(C1 cells in the RVLM) (15).
Microinjection of anti-DH-Sap into the thoracic spinal cord produced
the disappearance of most bulbospinal C1 neurons (
74% on average and
94%) in the present study. We believe that these percentages reflect
a depletion of the cells and not a reduction in the detectability of
PNMT immunoreactivity for two reasons. First, the quality of the PNMT
stain was evaluated in each rat by counting the number of PNMT-ir
profiles in the caudal aspects of the C1 or C2/C3 groups that do not
project to the thoracic spinal cord (Fig. 1, A and
C) (24, 29). Because cell counts in these areas were unaffected by anti-DH-Sap, we could obtain objective evidence that the visibility of PNMT immunoreactivity was the
same in lesioned and in control rats. Second, our estimate of the
depletion of cells was also based on counts of retrogradely labeled
neurons. If anti-DH-Sap had merely reduced the level of expression
of PNMT without killing the bulbospinal C1 cells, the number of healthy
FB-positive neurons of the RVLM should have remained unchanged. This
was clearly not the case (Figs. 1 and 2). For similar reasons, we
conclude that anti-DH-Sap also eliminated bulbospinal A5 and A6
neurons rather than reduced the detectability of TH. The percent
depletion of the bulbospinal noradrenergic neurons was ~100%, i.e.,
slightly better than that of the adrenergic cells. The particular
sensitivity of A5 neurons to anti-DH-Sap may be the result of a
higher concentration of exteriorized DH on the terminals or axons of
noradrenergic neurons. Alternately, the spinal projection pattern of A5
cells may be less specific than that of C1 neurons, such that a larger
proportion of bulbospinal A5 neurons may converge onto the spinal
thoracic segments injected. Although the A7 noradrenergic group of the
pons was not examined in the present study, it is probable that this
noradrenergic cell group that projects to the spinal cord was also
destroyed by treatment with anti-DH-Sap.
The lesion of bulbospinal A5 and C1 neurons by anti-DH-Sap was
clearly due to a process that required the specific binding of the
anti-DH antibody, because microinjection of the same dose of saporin
conjugated to an anti-mouse IgG had no effect on the number of these
neurons (Fig. 1). Furthermore, noncatecholaminergic neurons were
insensitive to anti-DH-Sap. The numbers of serotonergic neurons and
the nonadrenergic component of the RVLM bulbospinal projection
(FB-positive/non-PNMT-ir neurons) in rats treated with anti-DH-Sap
were comparable to those counted in unoperated control rats (Figs. 2
and 3). These results are in agreement with a recent study by Madden et
al. (15) in which microinjection of anti-DH-Sap into
the RVLM depleted ~90% of PNMT-ir neurons within the RVLM without
altering the number of bulbospinal, barosensitive non-C1 RVLM neurons.
In addition, in the present study, rats treated with anti-DH-Sap
displayed no obvious alteration in locomotor activity, hindlimb tone,
or behavior, which suggested that other descending spinal projections
were intact.
Despite the clear selectivity of anti-DH-Sap relative to descending
axons and axons of passage, substantial local damage occurred at the
site of injection in the form of gliosis and tissue necrosis that
persisted up to 5 wk after injection. When minimally effective doses of
anti-DH-Sap are employed, local lesions and gliosis can reportedly
be reduced (15). However, the window for this effective
and minimally damaging dose appears to be narrow, and injury is
probably never totally absent (15). In the present study
we chose to inject the toxin at sites distal to the cell bodies of
interest and to use a higher dose of anti-DH-Sap to increase the
likelihood of an effective depletion of the C1 cells. Because the toxin
was injected into a terminal area, we found that the RVLM was
cytologically intact in all animals, despite the selective
disappearance of the bulbospinal C1 cells (Fig. 4). The local damage
produced by microinjection of anti-DH-Sap may be due to the
nonselective binding of the saporin conjugate to neurons, glial cells,
or blood vessels, because treatment with IgG-Sap produced comparable
damage at the sites of injection in the present study. However, because
IgG-Sap produced equivalent local damage without depleting bulbospinal
catecholaminergic neurons, this neurotoxin provided a useful operated
control group for comparison with rats treated with anti-DH-Sap in
the physiological experiments.
Effect of anti-DH-Sap on the arterial baroreflex and
Bezold-Jarisch reflex.
A compelling finding of the present study is that depletion of the vast
majority of bulbospinal C1, A5, C3, and A6 cells had no detectable
effect on resting MAP and HR. Although SNA cannot be accurately
quantified between animals, rats treated with anti-DH-Sap clearly
had sympathetic tone, which was inhibited by increasing MAP and
eliminated by clonidine. The effectiveness of clonidine indicated that
afferent nerve traffic from the abdominal region made an equally
negligible contribution to the splanchnic nerve activity recorded in
control and lesioned rats. The relationship between SNA and MAP,
measured by classic logistic curve fitting (22), was
altered by treatment with anti-DH-Sap. Specifically, the proportion
of resting SNA remaining at saturation of the baroreflex (baroinsensitive SNA) was greater in rats treated with anti-DH-Sap, and the gain of the baroreflex was consequently attenuated. However, because this analysis requires SNA to be measured as a percentage of
the baseline value, there are several equally plausible interpretations of the changes induced by anti-DH-Sap. First, the increase in the
baroinsensitive component of SNA could be the result of a reduction in
the ability of baroreceptors to inhibit splanchnic vasoconstrictor
sympathetic efferents (decreased baroreflex inhibition). This
interpretation is somewhat unlikely, because the MAP50 of the reflex was unchanged (Table 1). Alternately, the data could indicate that at all levels of AP the activity of splanchnic
vasoconstrictor sympathetic efferents may be somewhat lower after
treatment with anti-DH-Sap. This second interpretation is compatible
with the widely held notion that the discharges of C1 cells contribute to resting sympathetic tone (14, 18,
20, 21). Finally, the activity of
baroinsensitive (nonvasomotor) splanchnic efferents might have been
enhanced by anti-DH-Sap, causing the increase in the percentage of
resting SNA remaining at saturation of the baroreflex.
In contrast, rats treated with IgG-Sap displayed no detectable change
in any of the measured physiological variables compared with unoperated
controls. Accordingly, the nonselective damage produced by
anti-DH-Sap in the upper thoracic spinal cord could not have caused
the changes observed when this saporin conjugate was injected. Thus the
selective depletion of bulbospinal catecholaminergic neurons is the
most likely explanation for the altered baroreflex responses observed
in rats treated with anti-DH-Sap. These results suggest that
bulbospinal catecholaminergic neurons may contribute to the efficacy of
baroreceptor-mediated changes in SNA but do not provide the essential
signal for these changes.
Intravenous administration of phenyl biguanide, a selective agonist of
5-HT3 receptors, activates cardiopulmonary chemosensitive vagal afferents to produce marked reductions in HR, SNA, and MAP. This
pattern of responses, known as the Bezold-Jarisch reflex, is mediated
in large part by the inhibition of RVLM presympathetic neurons
(30), including C1 and non-C1 cells (31). In
the present study, the magnitudes of phenyl biguanide-induced decreases
in SNA, AP, and HR were not altered by treatment with anti-DH-Sap or
IgG-Sap. These data suggest that, in the absence of the majority of
bulbospinal C1 cells, inhibition of the remaining presympathetic non-C1
neurons by phenyl biguanide decreases SNA to produce a comparable
decrease in MAP. Furthermore, preservation of this reflex suggests that
the RVLM likely continues to be a source of tonic excitatory drive to
sympathetic vasomotor neurons.
Effect of anti-DH-Sap on cyanide-induced sympathoexcitation.
In vagotomized rats anesthetized with urethan, intravenously
administered cyanide or brief inhalation of 100% nitrogen elicits a
marked increase in SNA that is dependent on intact carotid
chemoreceptors (7, 12). We assume that under
-chloralose anesthesia the initial excitatory component of the
sympathetic nerve response to cyanide is also due, at least in part, to
activation of carotid chemoreceptors. Although -chloralose produces
an anesthetized preparation that is excellent for examining sympathetic
responses to phenyl biguanide and changes in AP, this anesthetic was
found to be less than optimal for producing a sympathetic response to cyanide. In any case, the virtual abolition of the cyanide-induced increase in SNA in rats treated with anti-DH-Sap suggests that the
carotid chemoreflex is especially sensitive to the lesion of
bulbospinal catecholaminergic cell groups. This observation is
consistent with prior evidence that A5 and C1 cells make a significant
contribution to the chemoreflex. Microinjection of muscimol into the A5
region reduces the carotid chemoreflex-induced increase in SNA by
54-82% (11), and this sympathetic reflex also
clearly relies on the excitation of a subset of the RVLM presympathetic
neurons (13, 28). Although the RVLM neurons were not phenotypically identified in these two studies, it is likely
that some of the excitatory responses were elicited from presympathetic
C1 cells. Briefly, prior evidence suggests that bulbospinal A5 and C1
cells play a role in the sympathoexcitation produced by stimulation of
peripheral chemoreceptors. These observations are consistent with the
marked attenuation of cyanide-induced sympathoexcitation seen in the
present study in rats treated with anti-DH-Sap.
Perspectives
In the absence of the vast majority of bulbospinal
catecholaminergic neurons, rats continued to have a normal AP and HR.
Although changes in resting sympathetic vasomotor tone are difficult to accurately determine, rats treated with anti-DH-Sap clearly
maintained a measurable sympathetic nerve activity under anesthesia.
Their SNA was modulated by baroreceptor inputs, with an
MAP50 comparable to that of intact rats, and it was
inhibited normally by the stimulation of cardiopulmonary chemosensitive
afferents. The discrepancy between the magnitude of the destruction of
the C1 and A5 cells and the relatively modest consequences of these
lesions on basal sympathetic tone and cardiopulmonary reflexes suggests
at least three alternative explanations. First, anti-DH-Sap may have
caused only minor alterations, because terminal sprouting from the few
remaining bulbospinal C1 cells combined with other adaptive changes in
the spinal cord to compensate for the loss of a majority of the C1
cells. However, this explanation is not likely, because the rats with
near-total elimination of C1 cells displayed a physiology that was
indistinguishable from those with slightly less effective lesions.
Furthermore, compensation by remaining C1 cells would not be expected
to maintain most sympathetic reflexes with a specific reduction in the
sympathetic response to cyanide. A second interpretation is that
seemingly normal sympathetic function is observed after lesions of the
C1 neurons, because a brain stem area other than the RVLM now maintains sympathetic outflow. However, this explanation would require that inhibition of sympathetic outflow by increased AP or phenyl biguanide is mediated through central sites beyond the RVLM in rats treated with
anti-DH-Sap. A third interpretation is that basal sympathetic tone
generation and inhibitory cardiopulmonary reflexes (baroreflex and
Bezold-Jarisch reflex) rely in large part on the activity of the
noncatecholaminergic presympathetic neurons of the RVLM (14, 21) rather than on the discharges of the
bulbospinal catecholaminergic cells. Indeed, there is substantial
evidence that the non-C1 presympathetic neurons in the RVLM are
unaffected by anti-DH-Sap (Figs. 2 and 3) (15). The
lesion of up to 94% of spinally projecting C1 cells may thus leave a
sufficient number of non-C1 presympathetic neurons intact to maintain
an apparently normal sympathetic tone and AP at rest. In contrast, when
the sympathetic system is vigorously activated (e.g., by the carotid chemoreflex), deficits become readily apparent. Thus the presympathetic C1 neurons may play a subordinate role in the generation of sympathetic vasomotor tone by the RVLM under resting conditions but provide essential enhancement of this drive during activation of the
sympathetic nervous system.
This work was supported by National Heart, Lung, and Blood
Institute Grant HL-28785.
Address for reprint requests and other correspondence: P. G. Guyenet, Dept. of Pharmacology, University of Virginia Health System, PO Box 800735, 1300 Jefferson Park Ave., Charlottesville, VA
22908 (E-mail: pgg{at}virginia.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.
H-saporin
TOP
ABSTRACT
INTRODUCTION
