| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
ARTICLES |
Laboratory of Neuroendocrinology, Department of Biomedical Sciences, University Medical School (A.J.D., S.S., J.A.R., G.L.), Edinburgh, United Kingdom EH8 9XD; and Department of Neurobiology, The Babraham Institute (D.B.), Cambridge, United Kingdom CB2 4AT
Address all correspondence and requests for reprints to: Dr. A. J. Douglas, Laboratory of Neuroendocrinology, Department of Biomedical Sciences, University Medical School, Edinburgh, United Kingdom EH8 9XD. E-mail: alison.j.douglas{at}ed.ac.uk
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
1-noradrenergic receptor antagonist,
benoxathian, into the supraoptic nucleus before and during iv oxytocin
administration, Fos expression in supraoptic neurons was significantly
less than that in vehicle controls. Thus, at term pregnancy, uterine
contractions activate both oxytocin and vasopressin neurons in the SON,
and this activation involves a noradrenergic pathway. |
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
During parturition in the rat, the firing rate of magnocellular oxytocin neurons in the rat supraoptic nucleus (SON) is increased, and superimposed upon this elevated electrical activity are intermittent high frequency bursts of action potentials, which occur at about the time of delivery of each pup (4) and lead to pulsatile oxytocin secretion from the posterior pituitary (5, 6). The response of the rat uterus to systemic oxytocin administration increases on the day of delivery, in the last few hours before labor starts (7), and the pulsatile pattern of oxytocin appears to be particularly efficacious, as the onset of labor and optimal progress of parturition (8, 9) can be promoted by pulsatile administration of oxytocin at doses that are less effective when given continuously (9). Uterine contractions and the passage of fetuses through the birth canal further enhance oxytocin secretion; thus, oxytocin release from the pituitary during parturition is driven by a positive feedback loop (the Ferguson reflex) (10).
This positive feedback is relayed via a neuronal pathway through the brainstem (11) including the A2 neurons of the nucleus of the tractus solitarius (NTS). Neurons in both the SON and the NTS express Fos during parturition, reflecting their activation at this time, and pulsatile administration of oxytocin to term pregnant rats results in the induction of Fos expression at both sites even before birth is induced (8, 9). Triple labeling has shown that some of these NTS neurons are both retrogradely labeled from the SON and contain tyrosine hydroxylase, the rate-limiting enzyme in noradrenaline synthesis (12).
The above pattern of Fos expression in the SON and NTS during birth superficially resembles that after iv cholecystokinin (CCK). CCK increases the firing rate of oxytocin neurons, accompanied by increased Fos expression in the SON and increased oxytocin secretion (13, 14); CCK also induces Fos expression (8, 14) in noradrenergic neurons in the NTS and increases noradrenaline release within the SON (15), and local destruction of the noradrenergic innervation of the SON by selective neurotoxins prevents CCK-induced Fos expression in the SON, but not expression in the NTS or in the intact SON contralateral to the lesion (14). Thus, noradrenaline appears to mediate the activation of oxytocin neurons in the SON after systemic CCK. As noradrenaline release within the SON increases immediately before and during parturition (16), we hypothesized that this noradrenergic pathway may also mediate positive feedback from the contracting uterus to the SON.
In these experiments we tested the hypotheses that contractions of the
uterus before the delivery of pups activate oxytocin neurons, and that
they do so via a noradrenergic projection to the SON. Here, we studied
intrauterine pressure changes, the electrical activity of
identified SON neurons, and Fos expression in the SON in term pregnant,
anesthetized rats during pulsatile oxytocin administration. We also
measured noradrenaline release in the SON and studied the effects of
blocking the actions of endogenous noradrenaline by infusion of an
1-noradrenergic antagonist (benoxathian)
into the SON during pulsatile oxytocin administration.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Effect of oxytocin pulse administration on uterine contractile
activity and SON neuron firing rate in term pregnant rats
On the day of expected parturition (day 22 of pregnancy), 16
Sprague Dawley (Edinburgh) or Wistar (Babraham) rats were anesthetized
with urethane (ethyl carbamate, 1.25 g/kg) or pentobarbitone (18 mg in
0.3 ml; Sagatal, Rhone Merieux Ltd., Harlow, UK) given ip on the
morning of experiment. The two anesthetics were used in both Edinburgh
and Babraham, and this allowed comparison of the traditionally used
urethane with pentobarbitone, which is known not to induce SON neuron
activation (17). Rats were laparotomized by a lateral
longitudinal incision, and one horn of the pregnant uterus was removed
from the abdominal cavity and kept moist with warm 0.9% saline. A
small purse-string suture was made between the first and second pups
from the ovarian end using a curved needle and fine silk thread. A
small incision was made in the middle of the purse-string suture, and a
deflated latex balloon (condom tip) connected to saline-filled
polyethylene tubing was inserted between two pups. The uterine incision
was closed and secured by drawing the purse-string suture together. The
uterine horn was returned to the abdominal cavity, and the abdominal
wall and skin incisions were sutured closed. In some rats, a balloon
was inserted in this way into both uterine horns to compare their
contractile activities. In other rats a second balloon was inserted in
the cervix (per vaginum). The balloon(s) was filled with 1–1.5 ml
saline and was connected to a voltmeter via a pressure transducer;
pressure was sampled at 1 Hz and recorded using Spike 2 computer
software on a personal computer.
For uterine inflation/deflation experiments, in eight rats the pup at the cervical end in the horn of the uterus was removed and replaced with a balloon filled with approximately 4 ml saline, the volume of which could be decreased or increased by withdrawal or injection of saline to effect deflation or inflation, respectively.
The jugular vein and the trachea were then cannulated. The SON and the neural stalk of the pituitary were exposed using a ventral surgical approach as described previously (18). Briefly, a bipolar stimulating electrode was placed on the neural stalk, and the electrical activity of single SON neurons was recorded extracellularly using a glass micropipette filled with 0.9% saline, introduced into the SON under visual control. Neurons identified as projecting to the neurohypophysis by antidromic activation were classified as oxytocin or vasopressin cells on the basis of firing pattern and response to CCK (20 µg/kg, iv; Sigma, Dorset, UK). Oxytocin cells fired continuously and were excited by iv administration of CCK. Vasopressin cells fired phasically or continuously and were either inhibited or showed no response to iv CCK (19).
After a control period of 30–60 min, iv bolus injections of oxytocin
were administered (10 mU for 2 h and 20 mU for 2 h;
Syntocinon, Sandoz Pharmaceuticals Corp., Basel, Switzerland;
each pulse was in 20 or 30 µl given every 10 min, diluted in 0.9%
NaCl, and was administered in less than 2 s; maximum volume
injected over 4 h was 760 µl/pregnant rat, which has a total
blood volume of 
20 ml). Four control rats received no iv oxytocin
injections while the SON neuron firing rate and/or uterine pressure
were continuously recorded for up to 3 h. To analyze the effects
of oxytocin administration on uterine and cervical pressure and cell
activity, each experiment was divided into a control period, before any
drug administration, and four 1-h periods of oxytocin
administration.
Uterine and cervical pressure were measured, and the baseline pressure, frequency, and amplitude of contractions were calculated in the last 10 min of each experimental period; the mean pressure was calculated as the arithmetic average of the recorded values, the frequency of contractions by counting the number of peaks in each 10-min period (defined as a rise and fall in pressure of at least 3 mmHg), and the amplitude of contractions by taking the pressure at each peak and subtracting the lowest pressure before the next peak.
The firing rate of cells (mean ± SEM spikes/sec, in 60-sec bins) was used as the measure of cell activity, except when the activity pattern was phasic (putative vasopressin neurons), at which point the activity quotient (the proportion of time that a cell is active) and the intraburst firing rate were also calculated. The effect of the oxytocin pulses was measured by comparing firing rate in the control period with the last 10 min of each hour of oxytocin administration.
The correlation coefficient between intrauterine pressure in two uterine horns in the same rat, between intrauterine pressure and intracervical pressure in the same rat, or between intrauterine pressure and SON neuron activity in the same rat were calculated over each separate 1-h period.
Uterine pressure was not recorded in the following experiments to avoid induction of SON neuron activity and Fos expression during the additional surgical procedures.
Effect of oxytocin pulse administration on Fos expression in the
SON and NTS
Virgin (287 ± 20 g BW) and late pregnant (21 and 22
days; 418 ± 10 and 442 ± 18, respectively; each with 8–16
fetuses in utero) Sprague Dawley rats were anesthetized with
Sagatal as described above [Sagatal does not induce Fos in the SON
(17)], and a jugular vein was cannulated. At least 1
h after surgery, iv injections (pulses) were given every 10 min for
4 h; the rats received pulses of either isotonic saline (30 µl)
or oxytocin [Syntocinon; 30-µl pulses of 10 mU for 2 h,
followed by 20 mU for 2 h, (8)]. Ten minutes after
the last pulse, the rats were killed by decapitation, and their brains
and brainstems were rapidly removed and frozen in crushed dry ice.
Virgin rats injected ip with hypertonic saline (1.5
M NaCl, 4 ml/kg) were killed after 90 min, and
their brains were collected as described above as positive technical
controls; hyperosmotic stimulation activates Fos expression in both
oxytocin and vasopressin neurons in the SON (20). The
tissues were stored at -70 C until sectioning and
immunocytochemistry.
To assess whether Fos was expressed in oxytocin and/or vasopressin neurons, additional 22-day pregnant rats were treated similarly (vehicle pulses, n = 4; oxytocin pulses, n = 4) to generate tissue for double immunocytochemistry. Virgin rats were not investigated due to their known lack of basal Fos protein expression [in frozen brains (21) and in perfused-fixed brains (22)]. After pulse treatment the rats were perfused-fixed transcardially with heparinized isotonic saline and 4% paraformaldehyde in 0.1 M PBS; the brain and brainstem were removed and postfixed in 15% sucrose in 4% paraformaldehyde followed by 30% sucrose in PBS for 2 days and then frozen on dry ice before storage at -70 C and processing.
Effect of oxytocin pulse administration on noradrenaline release in
the SON
Seventeen 22-day pregnant Sprague Dawley rats were anesthetized
with Sagatal on the morning of experiment (as described above), and a
jugular vein was cannulated. A microdialysis probe (concentric
semipermeable membrane, 1 mm long and 0.5 mm wide; CMA12, Carnegie
Medicin, Biotech Instruments Ltd. (Kimpton, UK); prepared according to
the manufacturer’s instructions), flushed with degassed filtered
artificial cerebrospinal fluid (aCSF) at 1.3 µl/min, was
stereotaxically positioned in the SON (1.1 mm caudal and 1.8 mm lateral
to bregma, and 8.9 mm below the surface of the dura with the skull
level). After flushing for more than 1 h, microdialysis samples
were collected at intervals of 10–15 min. Basal samples were collected
for 60 min, then rats were given iv pulses of either vehicle (n =
9) or oxytocin [n = 8; every 10 min at 10 mU for 60 min (first
period), then 20 mU for 60 min (second period)]. Samples were
collected into tubes containing 2% perchloric acid (1 µl/7 µl,
vol/vol, dialysate) at 4 C to prevent oxidation of noradrenaline.
Dialysate samples were rapidly frozen on dry ice within 10 min of
collection. After the experiment, the brains were collected from the
rats, fixed in 10% formol saline, histologically stained, and examined
for probe location; only data from rats with confirmed probe location
within the SON were analyzed. The in vitro recovery rate of
the microdialysis probes, measured by HPLC followed by electrochemical
detection or by the percent recovery of
[3H]noradrenaline in dialysates, was 9.8
± 0.4%. Noradrenaline was quantified at The Babraham Institute using
reverse phase HPLC with electrochemical detection (M460 ECD,
Waters Corp., Milford, MA; in conjunction with a BAS 6-mm
Unijet detector cell) according to the method of Mefford
(23). Aliquots of 8 µl were injected onto a 2 x
150-mm Sphereclone ODS2 column (Phase Separations, Deeside, UK)
via a CMA200 autoinjector (Carnegie Medicin). Data capture and
integration were performed using the Gynkotek Software system (Severn
Analytical, Macclesfield, UK) run on a personal computer. Samples from
vehicle- and oxytocin-treated rats were analyzed randomly; standards
and blank controls were analyzed before each batch of experimental
samples. The limit of detection for the noradrenaline was 1.3 pg/8
µl. All rats had at least eight fetuses in utero; mean
maternal body weight was 414 ± 13 g.
Effect of an intra-SON infusion of a noradrenergic antagonist on
Fos expression in the SON induced by oxytocin pulse administration in
pregnant rats
Under halothane anesthesia (2–3% in
oxygen-N2O at 1200 ml/min for 25 min) pregnant
Sprague Dawley rats were fitted via a dorsal stereotaxic approach with
a 8.6-mm long, 22-gauge guide cannula (Plastics One, Inc., Semat, St.
Albans, UK) unilaterally in the SON (1.1 mm caudal and 1.8 mm lateral
to bregma with the skull level between 
and bregma) 2 days before
the experiment; the cannula was held in position with dental cement
fixed to the skull with two screws. On day 22 of pregnancy (384 ±
17 g BW), the rats were anesthetized with Sagatal (18 mg, ip), and
a jugular vein cannula filled with heparinized saline was implanted. At
least 1 h later an intra-SON infusion [0.8 µl/h for 2 h
from a gas-tight 10-µl Hamilton syringe (Reno, NV; Supplier,
Sigma, driven by a Braun perfusor pump] of either aCSF or
the noradrenergic 1-antagonist, benoxathian
(Sigma; 1 or 10 µg/ml, dissolved in aCSF), was begun
using a 28-gauge injection cannula (Plastics One, Inc.), which
projected 0.5 mm beyond the tip of the guide cannula into the SON.
After 30 min of infusion, iv pulses of isotonic saline or oxytocin
(Syntocinon, 20 mU/30 µl, at 30 µl every 10 min) were started and
continued for 90 min. Thus, there were four groups: rats receiving an
aCSF infusion in the SON given iv pulses of vehicle or oxytocin, and
rats receiving 1 µg/µl benoxathian or 10 µg/µl benoxathian
infusion in the SON and given iv oxytocin pulses. Ten minutes after the
last pulse, the rats were given a lethal dose of Sagatal iv and
perfused-fixed transcardially with heparinized isotonic saline and then
paraformaldehyde in PBS; the brain and brainstem were removed and
postfixed as described above. The location of the cannula tip near the
SON (<0.5 mm laterally and dorsally) was confirmed in the processed
sections.
Effect of intra-SON infusion of a noradrenergic antagonist on the
expression of Fos in the SON induced by osmotic stimulation
Under halothane anesthesia, virgin Sprague Dawley rats were
fitted with a guide cannula unilaterally in a SON, as described above,
2 days before experiment. On the day of experiment, the rats (253
± 9 g BW) were anesthetized with Sagatal, and a jugular vein
cannula was implanted. One hour later an intra-SON infusion (flow rate,
0.8 µl/h for 2 h) of either aCSF or benoxathian (1 µg/µl)
was begun. After 30 min of infusion the rats received an ip injection
of hypertonic saline (1.5 M NaCl, 4 ml/kg). Ninety minutes
later, the rats were given an overdose of Sagatal and perfused-fixed
transcardially, and the brain was removed and prepared as described
above.
Fos immunocytochemistry
From the nonfixed frozen brains 15-µm coronal cryostat
sections were cut through the SON and NTS and mounted on gelatinized
slides. Immunocytochemistry for Fos was performed as previously
described (24). Briefly, sections containing SON
(1000–1300 µm caudal from bregma) (25) or NTS
(500 µm rostral to obex, also containing area postrema) were then
fixed, and endogenous peroxidase was deactivated. The sections were
preincubated in normal sheep serum and incubated for 36 h at 4 C
with a polyclonal antiserum against Fos (Oncogene Science, Inc.; Supplier, Cambridge Bioscience, Cambridge, UK;
1:1000), then in the second antibody (antirabbit IgG complex,
peroxidase labeled, Vector Laboratories, Inc.
Peterborough, UK; 1:500) for 24 h at 4 C. Nuclear staining was
enhanced using the nickel-intensified (glucose-amino-oxidase)
3',3'-diaminobenzidine (DAB) method (21). The sections
were dehydrated, cleared in xylene, and coverslipped. Fos-positive
nuclei in the SON (16 SON profiles/rat) and NTS (4 NTS bilateral
profiles/rat) were counted, with the identity of the sections coded,
using a microscope with a x10 objective and a brightfield condenser.
For each rat the number of Fos-positive neurons per unit area SON or
NTS (i.e. per mm2) was calculated by
dividing the number of Fos-positive neurons by the area of the SON,
which was measured using Image (NIH image analysis computer program).
Sections from all pregnant rats were processed together, but additional
virgin rats given vehicle or oxytocin pulses were processed separately
and counted by the same person who processed the sections from the
pregnant rats. Slides from positive control rats (given hypertonic
saline) were included in each immunocytochemistry run as technical
controls.
Coronal sections (50 µm) cut from perfused-fixed brains on a freezing microtome were processed by Fos immunocytochemistry using a free-floating method as previously described (8). Briefly, endogenous peroxidase was deactivated, and sections were preincubated in normal sheep serum. Sections were incubated in a polyclonal antiserum against Fos (Cambridge Bioscience; 1:1000 diluted in phosphate buffer with 0.2% Triton X-100 and 1% normal sheep serum) for 36 h at 4 C, then in second antibody using an ABC kit (Vectastain Elite, Vector Laboratories, Inc.) and visualized using the nickel-enhanced DAB reaction. The sections were then mounted on gelatinized slides, dehydrated, cleared, and coverslipped. Fos-positive neurons were counted in both infused (ipsilateral) and contralateral sides of the SON and in the NTS on coded slides as described above. Fos-positive nuclei in the SON (20–24 SON profiles/rat) and NTS (4 NTS bilateral profiles/rat, at the rostro-caudal level of the area postrema) were counted, and for each rat the number of Fos-positive neurons per unit area SON or NTS was calculated as described above. Sections from all rats from each experiment were counted by the same person. Slides from positive control rats were included in each immunocytochemistry run as technical controls.
Double immunolabeling
Sections containing SON were double labeled for Fos and oxytocin
or Fos and vasopressin; sections containing NTS were double labeled for
Fos and tyrosine hydroxylase. Coronal sections (40 µm) were cut on a
freezing microtome and processed first by Fos immunocytochemistry as
described above, resulting in a black color visible in the nucleus of
positive neurons. The sections were then preincubated again with 1%
normal sheep serum and incubated with either a polyclonal antiserum
against oxytocin (1:2000; a gift from Dr. T. Higuchi, Fukui, Japan), a
monoclonal antiserum against vasopressin (1:2000; Chemicon
International, Inc., Temecula, CA), or a monoclonal antiserum against
tyrosine hydroxylase (1:400; Chemicon International, Inc.).
Visualization was achieved using
H2O2 and DAB and resulted
in brown cytoplasmic color visible in positive neurons. The sections
were then mounted on gelatinized slides, dehydrated, cleared, and
coverslipped. The number of Fos-positive neurons was counted in 10 SON
or NTS profiles per rat; as dense cell packing made accurate counting
of the number of oxytocin or vasopressin cells in the SON difficult,
the data are expressed as the percentage of Fos-positive cell nuclei
associated with cytoplasmic oxytocin or vasopressin labeling. Sections
from all pregnant rats were processed together for the double
immunolabeling.
Statistical analysis
Data are presented as the group mean ± SEM;
P < 0.05 was considered statistically significant. For
experiments on uterine activity, a 
2 test was
used to evaluate the difference in effect of oxytocin pulses between
the experimental group and the control group. For the cell firing rate,
all multiple comparisons were made with one-way ANOVA for repeated
measures, and single comparisons were by Student’s t test.
A Genstat software program was used to analyze correlations between
uterine/cervical pressure and electrical activity of SON neurons. The
z-test was used to compare proportions in the microdialysis experiment.
For other experiments, statistical analysis was performed using
SigmaStat for Windows (Jandel Scientific, San Rafael, CA) statistical
software. ANOVA (one or two way) for repeated measures was used to
compare paired data and consecutive samples.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
2 test). There
was no consistent change in the frequency of contractions, which
occurred at intervals of 40–90 sec in each rat (for examples, see Fig. 1
).
|
). The contractile activity of the two horns
within each rat was not tightly synchronized in the control period;
there was no consistent phase relationship between contractions in one
horn and those in the other, but the activity in the two horns was
positively correlated by the end of the oxytocin administration in each
experiment (r2 = 0.74, 0.85, and 0.9
respectively; each P < 0.05), and in one case it
switched from a negative correlation (r2 =
-0.78) in the control period to a positive correlation
(r2 = 0.74) in the fourth hour of oxytocin
administration.
Effect of oxytocin pulse administration on cervical contractile
activity in term pregnant rats
In five rats intracervical pressure was recorded
concomitantly with intrauterine pressure (e.g. Fig. 1D), and
in another two rats intracervical pressure alone was recorded. The
baseline pressure in the cervix during oxytocin administration declined
in all but one rat; it decreased most rapidly in the first hour of
oxytocin administration. However, in six of the seven rats, the
amplitude of cervical pressure changes increased with oxytocin
administration (to 1.6–31 mmHg) and remained greater at the end of
oxytocin administration than in the control period (P
< 0.05, by one-way ANOVA for repeated measures). Intracervical
pressure was not significantly correlated with intrauterine pressure
during either the control period or the final period of oxytocin
administration, and there were no periods of consistent synchrony
between cervical and uterine activities with oxytocin administration.
There was no consistent change in the frequency of contractions.
The type of anesthetic used did not influence the uterine or cervical responses to oxytocin pulse administration.
Effect of oxytocin pulse administration on SON neuron firing rate
in term pregnant rats
Oxytocin cells. Ten oxytocin cells were identified in the SON
of term pregnant rats with simultaneous intrauterine pressure
recording. Five of these were recorded throughout the initial 1-h
control period and the subsequent 4 h of treatment; the other
cells were each recorded for at least 2 h during part of the
experimental protocol. All cells showed an increase in continuous
activity during oxytocin administration (by one-way ANOVA for repeated
measures, P < 0.01) compared with the control period,
with no evidence of pulsatile activation (burst firing, see Fig. 1, A
and B). In eight oxytocin neurons that were recorded through a control
period and for the whole of the first hour of oxytocin administration,
the firing rate was significantly higher at the end of the first hour
than in the control period (mean change, +1.0 ± 0.3 spikes/s;
P < 0.05, by paired t test). Seven cells
were recorded throughout the second hour, and five of these showed a
further increase in activity, whereas the firing rate of the other two
cells remained above the initial control rate (mean change from first
to second hour, +0.2 ± 0.2 spikes/sec).
Vasopressin cells. Nine vasopressin cells were recorded during
oxytocin administration. Six cells were recorded through a control
period and for the whole of the first hour of treatment; their firing
rate was significantly higher at the end of the first hour than in the
control period (see Fig. 1, C and D; mean change, +0.6 ± 0.2
spikes/s; P < 0.05, by paired t test).
Three of six cells showed a further significant increase in activity
from the first to the second hour. Seven cells were recorded for 3
h during pulse administration. Six of the seven cells showed a
significant increase from the second to the third hour and displayed
significantly higher firing rates than first recorded (by one-way ANOVA
for repeated measures, P = NS; by paired t
test, P < 0.05, first period compared with control,
and third period compared with first). The activity quotient increased
in five of the vasopressin cells, and the intraburst firing rate
increased in all of the vasopressin cells (both P <
0.05; although in one cell this was not sustained until the end of
oxytocin administration).
The activity of one oxytocin and one vasopressin cell was recorded over periods of 2–3 h without intervention. In these experiments no significant changes were observed in cell firing rate (control period to end, 2.5 ± 0.1 to 2.2 ± 0.1 and 4.5 ± 0.2 to 4.2 ± 0.2 spikes/sec in oxytocin and vasopressin cells, respectively).
Correlation between intrauterine pressure and neuronal
activity
The correlation between SON neuron electrical activity and
intrauterine pressure was calculated for both oxytocin and vasopressin
cells. Thirteen cells (seven oxytocin and six vasopressin) were
examined, in which recordings of both neuron firing rate and
intrauterine pressure were made. During the control period the
correlation coefficients were not significantly different from zero
(oxytocin cells: P = 0.76; range of
r2 = -0.98 to +0.62; vasopressin cells:
P = 0.25; range of r2 = -0.4 to
+0.33). The correlation coefficient significantly increased between the
control period and the last treatment period for both oxytocin
(P = 0.046, by paired t test vs.
control period) and vasopressin cells (P = 0.045, by
paired t test vs. control period). Thus, the
correlation coefficients for the vasopressin cells were significantly
different from zero at the end of stimulation (P =
0.044; range of r2 = -0.02 to +0.69), and
statistical significance was almost reached for oxytocin cells
(P = 0.059; range of r2 = +0.43
to +0.83).
Whereas oxytocin cells fire continuously, vasopressin cells generally
fire phasically, with bursts of 20- to 60-sec duration separated by
silent periods of around 20-sec duration (see Fig. 2C) (62). Despite the
overall correlation of vasopressin cell firing rate with intrauterine
pressure, phasic bursts of activity were not consistently synchronized
with uterine contractions in any cell. However, for one vasopressin
cell that showed a clear phasic activity and was recorded throughout
4 h of oxytocin pulse treatment, a strong association between
burst periodicity and individual uterine contractions was apparent. For
this cell, the correlation of firing rate to mean intrauterine pressure
over the entire 4-h oxytocin pulse treatment was highly
significant (Fig. 2A; r2 = 0.46;
P < 0.01). Additionally, each 200-sec interval
throughout this period of recording was analyzed statistically and
revealed a positive correlation between uterine pressure and cell
activity for almost every interval (r2 =
0.35–0.6; overall P < 0.01), and this was apparent as
a clear overall relationship between spike probability and the phase of
uterine contraction (Fig. 2, B and C, firing rate and intrauterine
pressure recording for the same cell). No significant correlations were
found for the intrauterine pressure and cell firing rate in the
control, untreated rats.
|
In eight rats the activities of four oxytocin and five vasopressin
cells were recorded while a balloon placed in the lower uterus
(i.e. replacing a pup) was inflated or deflated by injection
or removal of 1.5–4 ml warm saline from an attached syringe. Oxytocin
cells showed a small, but consistent, decrease in electrical activity
during uterine distension induced by balloon inflation and a
corresponding increase in activity after deflation (Fig. 3; P < 0.05, by
t test, comparison of pre- and poststimulus value; for four
cells mean decrease after inflation, 0.6 ± 0.08 spike/sec).
Vasopressin cells showed a less consistent response to inflation and
deflation, with no overall significance detected.
|
). Oxytocin pulses induced a further
significant increase in the number of neurons expressing Fos compared
with the vehicle-treated rats on the same day (by two-way ANOVA,
P < 0.0001; by Newman-Keuls post-hoc test,
P < 0.05) and compared with oxytocin-treated day 21
pregnant rats (P < 0.05; Fig. 4A). The number of
neurons expressing Fos in the SON on day 22 of pregnancy was
approximately one third of that in the hypertonic saline-positive
controls (585 ± 141 Fos-positive nuclei/mm2
SON; n = 4).
|
Fos expression in the NTS was low in vehicle-treated day 21 pregnant
rats and did not significantly increase after oxytocin pulses (Fig. 4B). Rats on day 22 of pregnancy given oxytocin pulses showed
significantly more Fos-positive nuclei in the NTS than rats given
vehicle pulses (by ANOVA, P < 0.0001; Fig. 4B) or
rats given oxytocin pulses on day 21. Fos expression in tyrosine
hydroxylase-positive neurons comprised 29.0 ± 12.3% of all
Fos-expressing neurons in vehicle-treated rats and 37.4 ± 11.0%
of all Fos-expressing neurons in rats treated with oxytocin pulses
(P = NS).
In this experiment all of the oxytocin-pulse treated rats on day 22 of pregnancy went into labor or gave birth during the experiment compared with two of the vehicle pulse-treated rats on day 22 of pregnancy even though they remained under anesthesia; none of the rats given either vehicle or oxytocin pulses on day 21 of pregnancy showed any signs of labor or birth.
Effect of oxytocin pulse administration on noradrenaline release in
the SON
Of the nine vehicle-treated rats, one gave birth to a pup during
the experiment. Two of the vehicle-treated rats showed strong signs of
labor before the experiment (presence of blood/mucus at vaginal opening
and abdominal contractions); thus, three rats (33%) in the vehicle
group were in labor or gave birth. In oxytocin-treated rats, four of
eight (50%) gave birth to one or more pups during the second injection
period. In the rats that showed clear signs of labor or later gave
birth to a pup (four oxytocin-treated and three in vehicle-treated
rats), the mean basal noradrenaline content in dialysates from the SON
was 49.1 ± 11.9 pg/ml (mean of the four samples before
pulses started), significantly greater than that in rats with no signs
of labor (19.6 ± 7.1 pg/ml; n = 10; P <
0.05, by t test).
The noradrenaline content of the dialysates did not change
significantly in rats given vehicle pulses and that were not in labor
or did not give birth to a pup (Fig. 5A).
In both the oxytocin-treated group and the labor/pup group, some rats
exhibited peaks of noradrenaline content (Fig. 5, B–D). Peaks were
defined as noradrenaline content reaching a value of greater than 5
SD from the mean of the basal samples for that rat. The
proportion of rats showing 1 or more noradrenaline peaks was
significantly greater for oxytocin-treated and labor/pup groups (8 of
11 rats) than for the vehicle-pulsed group (0 of 6 rats; by z-test to
compare proportions, P < 0.05). Of oxytocin-treated
rats, 5 of 8 exhibited peaks of noradrenaline, all of which were
observed during the second postbasal sampling period when the high
concentration pulses of oxytocin were administered. Of vehicle-treated
rats, 2 of 9 exhibited peaks of noradrenaline during the basal control
period; these were the 2 rats in labor, and the peaks were not related
to the vehicle pulses. Of vehicle-treated rats, 1 of 9 exhibited a peak
during the second postbasal sampling period, and this was in the rat
that gave birth to a pup.
|
, A and B; by Student’s Newman
Keuls post-hoc test, P < 0.05 compared with
vehicle-treated rats). Infusion of benoxathian into one SON before and
during the oxytocin pulses at concentrations of 1 and 10 µg/µl
strongly attenuated the oxytocin-induced Fos expression in the infused
ipsilateral side compared with the effect of aCSF (P <
0.05, by post hoc test; Fig. 6A); thus, numbers of
Fos-containing neurons in the oxytocin-treated groups treated with
benoxathian in the SON were not significantly different from those in
the vehicle-treated controls (Fig. 6A). Fos expression was also
significantly attenuated in the contralateral SON (at a benoxathian
concentration of 10 µg/µl; P < 0.05, by
post-hoc test; but not the 1 µg/µl dose; Fig. 6B). The
number of Fos-positive neurons was counted separately in the dorsal SON
region (defined as dorsal of the line between the lateral ventral edge
of the SON and the top of the dorsal, lateral edge of the optic chiasm,
previously described as the predominant location of oxytocin neurons),
ipsilateral to the infusion side. The ratio of labeled neurons in the
dorsal SON/whole SON was not significantly different after benoxathian
compared with vehicle infusion in the oxytocin-treated rats
(benoxathian, 0.49 ± 0.03; vehicle, 0.49 ± 0.02),
indicating that Fos expression was attenuated similarly by benoxathian
in both the oxytocin-rich SON area and the whole SON, which also
contains vasopressin neurons.
|
). Benoxathian
infusion in the SON did not alter the number of oxytocin-induced
Fos-positive nuclei in the NTS (Fig. 6C), and there was no significant
difference between the ipsi- and contralateral sides (individual data
not shown). In these rats, 6 of 19 oxytocin-pulse treated rats went into labor or gave birth during the experiment compared with none of the vehicle pulse-treated rats; benoxathian infusion in the SON had no significant effect on whether labor started at either dose (4 of 6 rats that were in labor were given benoxathian).
Effect of intra-SON infusion of a noradrenergic antagonist on Fos
expression in the SON induced by osmotic stimulation
Fos expression in SON neurons was high in osmotically stimulated
virgin rats, as expected, with 2273 ± 400 positive
nuclei/mm2 in controls (mean of both SON in
vehicle-treated rats). Infusion of benoxathian into the SON did not
significantly attenuate the Fos expression induced by osmotic
stimulation in either the ipsilateral or contralateral sides (Fig. 7) compared with that in vehicle-treated
rats.
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
It appears that the contractile activity of the uterus is itself the source of activation of neurons in the SON. As uterine activity, at least in the two horns, was not tightly synchronized in basal control recordings and was poorly correlated with cervical activity, a tight correspondence between uterine contractions in any one horn and neural activity could not be anticipated. Nonetheless, the activity of the two horns of the uterus was positively correlated after oxytocin treatment in the experiments in which both were measured simultaneously, and the electrical activity of SON neurons and intrauterine pressure were significantly correlated after oxytocin treatment. The phase relationship in these cases between cell activity and uterine activity suggests activation during contraction rather than during relaxation. This together with the inhibitory effect of artificial uterine distension on cell activity suggest that SON neurons are activated as a result of uterine contraction rather than as a result of uterine distension accompanying the movement of fetuses along the uterus.
Supporting these findings, the same oxytocin pulse regimen also rapidly induced Fos expression in SON neurons in term pregnant rats. As the SON is within the blood-brain barrier, which is relatively impermeable to oxytocin (27), the half-life of circulating oxytocin is only 1–2 min, and oxytocin administration does not induce Fos expression in the SON of virgin rats, the injected oxytocin is unlikely to have had a direct effect on SON neurons. Furthermore, it seems that the presence of the oxytocin-receptive uterus is required for SON neuron activation. During parturition there is an increase in the discharge frequency of uterine afferent nerves, including the pelvic nerve (28), which when lesioned prevents the fetus expulsion reflex and reduces oxytocin secretion (29, 30). Therefore, our data are consistent with the hypothesis that uterine activity can modulate supraoptic neuronal activity before and during delivery.
Unlike the inhibitory effects of uterine distension described here, vaginal distension increases the SON neuron firing rate ( (31, 32) in pregnant and lactating rats. In addition, cervical stimulation can specifically initiate bursting discharges in term pregnant rats. Thus, we suggest that contractile activity in the uterus preceding the first delivery stimulates further oxytocin release that enhances that delivery, facilitating the movement of fetuses along the birth canal. At the point of delivery, stimulation of the cervix and distension of the vagina initiate a burst discharge of oxytocin, which may either facilitate the expulsive phase of delivery or initiate separation of the placenta from the uterus of the next pup for delivery.
Double immunolabeling and firing rate responses showed that enhanced uterine contraction excited vasopressin neurons of the SON as well as oxytocin neurons. It has previously been reported that both vasopressin neuron Fos expression and vasopressin secretion increase during parturition (33, 34). The presence of vasopressin V1 receptors in the uterus that can mediate contractions (35, 36, 37) suggests that vasopressin could augment oxytocin-induced uterine contractions. However, vasopressin V1 receptor antagonist administration does not delay birth (38), and it seems likely that the major role of vasopressin in the uterus during parturition is to promote vasoconstriction, which would reduce the risk of hemorrhage after delivery.
Efferents from the uterus to the SON are relayed via the brainstem; increased uterine contractile activity, when artificially or naturally activated by pulsatile oxytocin infusion (8) or during parturition (9), respectively, induces Fos expression in brainstem neurons, including the NTS and ventrolateral medulla, and in the dorsal vagal complex. There is strong neuroanatomical and pharmacological evidence that the noradrenaline released in the SON is from axon terminals of neurons with cell bodies in these brainstem areas. Catecholaminergic neurons in the A1 and A2 regions project to magnocellular SON and PVN neurons (39, 40, 41), and electrical stimulation of the NTS activates both oxytocin and vasopressin neurons in the SON (11). Directly projecting A2 neurons of the NTS have been implicated in the brainstem drive to oxytocin neurons (42), whereas A1 neurons seem to selectively regulate the vasopressin neurons (40). We now show that giving oxytocin pulses to term pregnant rats (enhancing uterine contractile activity) induces noradrenaline release in the SON, observed mainly as sudden peaks, and that rats in which parturition is imminent have higher basal extracellular noradrenaline content in the SON, confirming our previous reports of increased noradrenaline content in the SON just before and during spontaneous parturition (16). Furthermore, we have shown, using retrograde tracing techniques, that NTS neurons projecting to the SON express tyrosine hydroxylase (a rate-limiting enzyme in noradrenaline synthesis) and are activated (express Fos) during parturition (12).
We show here for the first time that an
1-receptor noradrenergic antagonist attenuated
Fos expression in the SON induced by driving uterine contractile
activity. In conjunction with the data showing that increased uterine
contractile activity correlates with increased SON neuron firing rate
and Fos expression and the increase in peaks of noradrenaline release
in the SON, these data indicate a physiological role for endogenous
noradrenaline in the activation of oxytocin neurons at parturition.
Noradrenaline acts on both 1- and
2-adrenoceptors in the SON [respectively,
excitatory (43, 44) and presynaptic inhibitory (45, 46)], and here we used a selective
1-adrenoceptor antagonist, benoxathian
(47), previously shown to be effective in inhibiting the
oxytocin neuron firing rate (48). In fact, benoxathian has
previously been shown to inhibit the CCK-induced increase in the
oxytocin neuron firing rate (48), and in preliminary
experiments we found that it also restrains CCK-induced Fos expression
in the SON compared with vehicle (307 ± 149 vs.
805 ± 199 positive neurons/mm2,
respectively; P < 0.05). Therefore, the effects on Fos
expression that we report here probably reflect prevention of an
excitatory action of noradrenaline on oxytocin and vasopressin neurons.
Thus, these data support previous reports of intracerebral
noradrenaline increasing oxytocin secretion (49) and SON
neuron Fos expression (50), and -adrenergic antagonist
administration inhibiting basal and stimulated oxytocin neuron firing
rates (48) and delaying the milk ejection reflex
(51).
An 1-noradrenergic agonist has previously been
shown to modulate the firing pattern in SON neurons in vitro
(52) to produce clustered spike discharge, suggesting that
it may facilitate the high frequency bursting activity of oxytocin
cells observed during parturition and the milk ejection reflex.
However, in the experiments reported here, uterine contractions alone
did not induce burst-like activation; it is possible that during
parturition bursts may be specifically triggered by vagino/cervical
activation during passage of the fetus or placenta, as bursts can be
evoked in oxytocin cells in late pregnant rats by vaginal probing
(53). Thus, the peaks of noradrenaline release observed in
the SON in parturient rats, some of which were in labor and delivered
pups, albeit slowly, under anesthesia, may reflect activation not
present in the rats anesthetized with urethane for electrophysiological
experiments, in which movement of the fetuses was also obstructed by
insertion of balloons. However, the background elevation of the
noradrenaline concentration in parturient rats, presumably reflecting
increased tonic activity of the noradrenergic innervation of the SON
during labor, probably contributes to the elevated oxytocin cell
activity.
However, a large proportion of the NTS neurons that express Fos in parturition or after oxytocin-induced uterine contractions and that project to the SON do not contain tyrosine hydroxylase (8, 12), so another neurotransmitter(s)/factor(s) appears to be additionally required for (or is capable of) induction of oxytocin neuron activity and secretion. Whether noradrenaline causes the enhanced basal Fos expression in the SON on day 22 of pregnancy in control rats cannot easily be shown by seeking changes in Fos expression. In other preliminary experiments, benoxathian infusion into the SON for up to 5 h before and during birth in conscious rats was unable to reduce Fos expression during birth compared with that in vehicle-infused controls. Therefore, the signal(s) that induced basal Fos expression before birth was not predominantly noradrenergic, and/or once the peptide product of the Fos gene was generated, it could not be switched off by simply removing the source of the excitatory signal, and the protein remained in the cell for a time depending upon its stability. Local factors within the SON may also be required to induce neuronal burst firing. Noradrenaline has been implicated in regulating oxytocin release from the dendrites of oxytocin neurons in lactation (54) and inducing changes in morphology of SON neurons (55); both phenomena are exhibited during parturition and contribute to the generation of burst firing (56). Release of oxytocin from dendrites within the SON increases during parturition and prevention of its actions via the oxytocin receptor (57, 58) with a local infusion of oxytocin antagonist inhibits oxytocin release in the SON and delays pup births (59); while oxytocin infusion into the SON can induce SON neuron Fos expression (preliminary findings). So it is possible that noradrenaline contributes to the enhanced firing rate and Fos expression in SON neurons during parturition by regulating local dendritic oxytocin release. Either the anesthetized rat model used here or uterine contraction alone is insufficient to permit burst firing of SON neurons to occur, although burst firing can be observed under urethane anesthesia during the milk ejection reflex in lactating rats (60). As local vasopressin release does not increase in the SON during parturition (57), it is unlikely to contribute to the induction of Fos in our model, even though vasopressin is released within the SON under other physiological circumstances, and vasopressin infusion in the SON induces SON neuron Fos expression (61).
Although both oxytocin and vasopressin neurons can be activated by cardiovascular, osmotic, and other abdominal stimuli as well as by parturition and suckling, it is unlikely that changes in blood pressure or volume could have accounted for their activation. Blood pressure was not measured in these experiments, but previous measurements of blood pressure did not show any difference between oxytocin- and saline-treated pregnant female rats (8). We have also reported that the same oxytocin pulse treatment does not activate neurons in the subfornical organ or area postrema (8), areas that express Fos after osmotic stimulation (62) or after changes in blood pressure (63, 64) and that project to the SON. However decreased, and not increased, blood volume increases Fos expression in SON neurons (63), and oxytocin has variously been reported to have no effect (65), to decrease (66), or to increase (67) arterial blood pressure. Certainly, the changes in blood volume imposed in our experiments amount to less than 5% of the blood volume of a pregnant rat, and saline injections do not in themselves increase neuron activation. In any case, the induced changes in plasma oxytocin in our experiments mimic normal changes during parturition, and presumably any uterine contraction-mediated blood pressure changes associated with that.
In our experiments, benoxathian attenuated Fos expression in the SON, both ipsi- and contralateral to the infusion site. This may reflect diffusion of benoxathian across the hypothalamus to the contralateral side, cross-talk between the two SON (68, 69), or actions in regions of the hypothalamus, for example the anterior periventricular structures (70), that have bilateral projections to the SON. The extent of diffusion of the antagonist is unknown for this experiment, but lesion of the anterior periventricular region does not disrupt the parturition process (71). An alternative explanation, however, is that action in even one SON reduces oxytocin release and so reduces uterine contractile activity, resulting in reduced feedback drive to both nuclei. Benoxathian did not produce a general reduction in SON neuron responsiveness, as it did not prevent SON neuron Fos expression in response to a hyperosmotic stimulus, which involves both a direct increase in activation by the osmoresponsive SON neurons and trans-synaptic activation through osmoresponsive neurons in the lamina terminalis. Although the lack of effect against a hyperosmotic stimulus could be because it induced a large activation of the neurons, benoxathian has previously been reported to inhibit hyperexcited SON neurons under other circumstances (48). Therefore, overall, our data indicate that increasing uterine contractile activity at term activates oxytocin and vasopressin neurons in the SON at least in part via noradrenergic action in the SON.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
2 Current address: Max Planck Institute of Psychiatry,
Kraepelinstrasse, 80804 Munich, Germany.
Received July 25, 2000.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
)]. Typical recordings of oxytocin neurons
from rats with simultaneous intrauterine pressure recording (A) and
simultaneous intrauterine pressure recording in both uterine horns of
the same rat (B); the excitatory response to CCK is also shown (
).
Typical recordings of vasopressin neurons from rats with simultaneous
intrauterine pressure recording (C) and with simultaneous intrauterine
and intracervical pressure recording from the same rat (D).
Horizontal bars, 10 min.
; 30 µl every 10 min for 4 h) or oxytocin pulses
(
; 10 mU/30 µl·pulse every 10 min for 2 h, followed by 20
mU/30 µl·pulse every 10 min for 2 h). A, Fos expression in the
SON in virgin (n = 4 and 9 for vehicle and oxytocin groups,
respectively), 21-day pregnant (n = 5 and 5, respectively), and
22-day pregnant (n = 6 and 7, respectively) rats. By two-way
ANOVA: P < 0.0001. by Newman-Keuls
post-hoc tests: *, P < 0.05
vs. day 21; #, P < 0.05 between
oxytocin-treated and vehicle control on same day. B, Fos expression in
the NTS in 21-day pregnant (n = 5 and 4) and 22-day pregnant
(n = 6 and 7) rats given vehicle or oxytocin pulses, respectively.
By one-way ANOVA: P < 0.05; #,
P < 0.05 compared with all other groups.
; 1 µg/ml; n = 6) or
aCSF (