- Email: [email protected]
Serotonergic modulation of the respiratory rhythm generator at birth: An in vitro study in the rat
Neuroscience Letters, 143 (1992) 91 95
91
© 1992 Elsevier Scientific Publishers Ireland Ltd. All rights reserved 0304-3940/92/$ 05.00 NSL 08864
Serotonergic modulation of the respiratory rhythm generator at birth: an in vitro study in the rat E. Di P a s q u a l e , D. M o r i n , R. M o n t e a u a n d G . H i l a i r e Biologie des Rvthmes et du D~;veloppement, URA C N R S 0205. Faculty; St. d~;rdme. Marseil/us ( k)'am'~' ;
(Received 26 March 1992: Revised version received 20 May 1992: Accepted 20 Ma5 1992) Kqv wor&v
Respiratory center; Serotonin, Neuromodulation, Newborn rat: In vitro
In order to investigate the mechanisms through which serotonin (5-HT) modulates the activity of the respiratory rhythm generator, respiratory activity was recorded from cervical ventral roots of the superfused isolated brainstem-spinal cord preparation of the newborn rat. Replacing the normal bathing medium by a medium containing 5-HT (30/.tM) increased the respiratory frequency by 70% of the control value. Intact pontomedullary structures are necessary for this effect to take place, however, since the 5-HT-induced increases in respiratory frequency were no longer observed after elimination (section and electrolytic lesion) of the caudal ventro-lateral pons containing the A5 areas. Local applications of 5-HT (dual bath, microdialysis and microinjection experiments) revealed, however, that 5-HT acts at the medullary level and that its effects are not due to a diffuse action on all the neurons of the medullary respiratory centers but to a specitic action focusing on structures located in the rostral ventro-latcral medulla.
Studies on the n e u r o m o d u l a t i o n o f the central respiratory activity exerted by endogenous substances such as serotonin (5-HT) have often yielded conflicting results since, depending on the m o d e of administration (peripherally or centrally), various respiratory effects [7, 8] can be elicited by 5-HT. The reduced preparation consisting of isolated n e w b o r n rat brainstem-spinal cord [19] is still able to produce s p o n t a n e o u s and rhythmic respiratory discharges in vitro in the phrenic nerves which are assumed to be generated by the same system that operate in vivo [17] or at least a part o f this system: this preparation has proved to be a useful tool for analyzing respiratory n e u r o m o d u l a t o r y processes since the elimination o f all the peripheral effects means that the results can be attributed with certainty to central mechanisms only. Previous pharmacological experiments performed with this preparation d e m o n s t r a t e d that the activity o f the medullary respiratory r h y t h m generator is increased by 5-HT. A d d i n g exogenous 5-HT to the bathing medium [9, 1 1], activating the 5-HT biosynthesis chain [12] and stimulating raphe structures [10] increased the respiratory fi'eq u e n c y . ( R F ) . This effect appears to be mediated by 5HTt receptors [12], since (i) it was blocked by 5-HT~ 2 Correspondencu." E. Di Pasquale, Biologie des Rythmes et du Ddveloppement, URA CNRS 0205, Facult6 St. Jdr6me, BP 332, Marseilles. France.
(methysergide) but not by specific 5-HT 2 antagonists (ketanserine), and (ii) it could be elicited by 5-HT t (8O H - D P A T ) but not 5-HT_~ agonists (~-CH3-5-HT). The aim o f the present report was to shed some light on the mechanisms whereby 5-HT may modulate the R F by limiting the access o f 5-HT to certain parts o f the brainstem. Results reveal that intact ponto-medullary structures are necessary for 5-HT modulation of the R F to occur, but the structures on which 5-HT acts are likely located in the rostral ventro-lateral medulla, an area which has been reported to contain the "pacemaker-like" neurons of the respiratory centers [4, 14, 15, 17, 18]. The brainstem and cervical spinal cord oF newborn rats (0--3 days old) were dissected and placed i~ ~i 2 ml in vitro c h a m b e r superfused with artifii:ial cerebl'ospinal fluid, using procedures identical to those previously described [2, 3]. Briefly, electrical activity from a phrenic nerve root was recorded using suction electrodes, filtered (5-3,000 Hz), amplified (Neurolog System, Digitimdr), integrated (leaky integrator, time constant 50 ms) and displayed on an oscilloscope and a chart recorder. Lesions were performed by passing D C currents (1 mA, 5 s) via tungsten microelectrodes. Serotonin (Sigma) was added to the normal medium (30 y M ) and superfused (2 ml/min) t\~r 5 rain. To apply 5-HT to the whole preparation, the pons alone, or the medulla alone, a barrier (improved with vaseline) was
92
:'()(}
• .::
,- .:
-- ~
.~ 15{) >
.,, i
C
I ?5 15(}
175
100
125 100
125 75
75
50
5{)
25
25
-) ~ ' ~ ; ; - --~ -- ''q . . . . .
) -'l~i
0
5 HT
~" 20O
zoo
D
5 HT
Time (min)
E
20o
175
175
175
'~
150
150
150
~
125
125
c
0
F
t25
pr"-~
100
100 75
I00 75
50
50
50
25
25
25
I
75
0
5 HT
Time (min)
Time (min)
0
5HT
Time (min)
5HT
Time (rain)
Fig. 1. Effects of general application of 5-HT on respiratory rhythm. A: schematic arrangement showing the type of inspiratory discharge recorded on the brainstem-spinal cord preparation of newborn rat. B F: in each graph, mean respiratory frequency (expressed as a percentage of the mean control respiratory frequency) and S.E.M. plotted every minute vs. time (expressed in rain). First 4 rain: control under normal medium; following 5 min: test under 5-HT medium (30/JM, black area); last 5 min: recovery under normal medium. B-D: 5-HT bathing under normal conditions increased the respiratory frequency (B, n=14) but this effect was not observed in C (n= 15) after transection performed at the level of the ponto-medullary junction (see double-headed arrow in A) or in D (n=6) after bilateral coagulation of A5 (large black dots in A). E-F: double bath experiment with a barrier placed at the level of the double-headed arrow; 5-HT elicited an increase in the respiratory frequency when applied to the medulla (F) but not when applied to the pons (E).
installed at the ponto-medullary border (double-headed arrow in Fig. 1).In 30 experiments, multibarelled micropipettes were used to apply 5-HT in specific sites of the brainstem (raphe nuclei, nucleus ambiguus, rostral ventro-lateral medulla) by pressure pulse (100 mM, 1-2 s). Prior to all the experiments, control ejections were performed under direct visual observation using a microscope: the ejected volumes never exceeded 5 nl. In 17 microdialysis experiments, a microinjection pump (Carnegie Med. CMA/100, flow rate 2 pl/min) connected to a syringe selector (CMA/111, to select either normal medium or medium containing 5-HT) and a microdialysis probe (CMA/11, outer diameter 240/.tm, length 1 mm) was used to slowly apply 5-HT by diffusion in specific parts of the brainstem. The probe, fixed to a micromanipulator (LPC) was inserted vertically 1 mm below the ventral surface into either the raphe nuclei or the rostral ventro-lateral medulla. The probe was perfused for 30 min with a normal medium and 5-HT medium was then applied for 5 min (500 pM; since the recovery of the probe was in the range of 10% in vitro, the actual 5-HT concentration in the tissue was assumed to be about 50
pM). Normal medium and 5-HT medium (applied either by superfusion or microdialysis) had identical pH (7.3). In order to be able to make comparisons with previously published results [9, 12], identical experimental procedures and data compilation methods were adopted. A control period of at least 5 rain was used to define the mean RF every minute prior to any experimental manoeuvre. The resulting frequency changes were expressed as percentages of the mean control values (100%) in the figures, but the absolute values in c.min -~ are given in the text. The experiments were repeated on several brainstem preparations (at least 5) with a standardized procedure to evaluate the mean effects on the RF which were expressed as the mean _+ S.E.M. Any differences were tested using Student's t-test and were taken to be significant at P values lower than 0.05. The activity recorded in vitro from the cervical ventral roots of newborn rat brainstem-spinal cord preparations is illustrated in Fig. 1A. This rhythmic activity, which was taken to constitute the respiratory frequency (RF), was stable under control conditions (Fig. 1B, mean rate 4.4 + 0.8 c-min ~, n=14). When the normal medium was
93
A
B
175
100 75 5O 25 0
5HT
a:'l
..j
D i t,l,i,lil Ill 1;:ii lp,rI t I[
!l!lll!
1!1
7
'
i, lj 1 Ilil |
,i.l-j-,3_.3- I
.~ 1'25
C
o
rdm
5 HT
T i m e (min)
l lllialll[nlJl,ltt,,lllltl
l til:l~Jl[.l_l,tll
T i m e (rain)
~l..it
[rl I lYVIIlrrrlVITTI I I1I I [Tll FIT r v'J r r T"r'f l l T rll7 •
5HT
E
8
2 min
Fig. 2. Effects of local application of 5-HT on respiratory rhythm. A: schematic illustration showing the site of local 5-HT applications (black dot) in the rostral ventro-lateral medulla. B,C: same ordering as in Fig. 1B F showing the mean increase in respiratory frequency elicited when 5-HT was applied to the rostral ventro-lateral medulla either by microdialysis (B, 500/aM. n 9) or by pressure pulse via multibarrcllcd micropipettes (('. 100 raM, n 25L D,E: the inspiratory discharge recorded from cervical root in two different preparations: note the increases in respiratory l'rcqucncy elicited by 5-HT microdialysis (D, between arrows} and 5-HT pressure microinjection (E, at arrm~) in individual cxpcriments. Time as indicatcd.
replaced by a medium containing 5-HT ( 3 0 / a M ) for 5 min, the R F increased during the first minute o f perfusion, and reached a peak within 2 3 min (mean: 7.5 + 1.4 c-min ~; n=14: range 5- 20 c.min L). Resuming the control conditions .'estored the control R F within 5 10 min. Eliminating the pons by performing a transection at the level o f the p o n t o - m e d u l l a r y junction (double-headed arrow in Fig. IA) increased the R F to 11.5 +_ 0.7 c.min ] (see also ret;. 2 and 3). 5-HT medium no longer significantly increased the R F (n=15, Fig. IC). To localize the pontine structures necessary for 5-HT excitatory effects to be exerted on the RF, microdissections were performed in the pons under binocular control. 5-HT-induced increases in the R F were not significantly altered by eliminating the medial region o f the pons (hatched area in Fig. 1: n=9) but were abolished by eliminating the lateral regions (dotted area in Fig. 1: n=5). To define more precisely what structures in the lateral pons were responsible for these respiratory e f fects, bilateral electrolytic lesions were placed in the caudal ventro-lateral pons to eliminate A5 areas by passing D C currents via a lungsten microelectrode inserted just medial to the entry o f the V roots (large filled circles in
Fig. IA, see refs. 2, 3 and 16) to 300/am below the venlral surface, The R F was increased by both electrolytic lesions and ponto-medullary sections (no statistical difference) and the 5-HT-induced increases in the R F were suppressed (n=6, Fig. 1D), To determine whether 5-HT acts on pontine o1" medullaw targets, a barrier was phtced at the ponto-medullary junction (double-headed arrow in Fig. I A) with a view to applying various media to either the pons or tile medulla. When the pons was bathed with 5-HT medium (30/aM) and the medulla with normal medium, no significant changes in thc R F were elicited (n-10: Fig. IE). When the ports was bathed with nornial medium and the medulla with 5-HT medium, the R F increased signiticantly (9.1 _+ 1.7 c.min ',/7=11, Fig. IF): the mean effect was not significantly different from that evoked by applying 5-HT to the whole brainstem. Intact ponto-medullary structures are therefore necessary for 5-HT modulation o f the respiratory r h y t h m generator to occur, but the structures on which 5-HT acts are probably in the medulla. These medullary structures were identified by performing microinjections o f 5-HT via multibarelled mi-
94 cropipettes inserted into the brainstem (see ref. 16). No significant changes in the RF resulted from injecting 5HT in the vicinity of the ponto-medutlary raphe nuclei (n=8) and the ambiguus nuclei (n--26). Increases in the RF were frequently elicited by injecting 5-HT into the rostral ventro-lateral medulla, 200-300/lm below the ventral surface. The increase in the RF occurred during the first minute after the injection (8.3 + 1.9 c.min ~, n=15, Fig. 2C,E); the RF returned to control values within 2-3 min. At these sites, saline injections had no effects. Microdialysis experiments were performed to apply 5HT locally to the rostral ventro-lateral medulla during longer periods. The microdialysis probe perfused with normal bathing medium was first slowly inserted in the rostral ventro-lateral medulla to a depth of 1 mm. Depending on the experiments, the RF was either unaffected (n=2), decreased (n--2), increased (n=3) or was suppressed for a while (5-10 min, n=2) when the probe was lowered in the rostral ventro-lateral medulla. After a 30-rain period, the control R F was restored and the microdialysis probe was then perfused with a 5-HT medium (500/~m) for 5 min instead of the normal one. The RF increased within 2 min (8.7 _+ 1.6 c.min -1) and remained high during the 5 min dialysis (n=8, Fig. 2B,D). Replacing the dialysate with a normal medium resulted in the RF returning to control values with a period of 15--20 min. When the microdialysis probe was placed into medial raphe sites at either the pontine (n=4) or the medullary (n--5) levels, 5-HT perfusion of the probe did not significantly affect the RF. The present results fully agree with previous studies [9-13] describing the excitatory modulation of the respiratory rhythm generator exerted by 5-HT and suggest furthermore that 5-HT acts specifically on structures in the rostral ventro-lateral medulla but not diffusely on the neurons of the ventral respiratory group [6]. This does not rule out the possible existence of other sites of action for 5-HT, such as the hypoglossal nucleus where it reduces the hypoglossal inspiratory activity [13], but in this case it is the amplitude, and not the frequency, of the motor output which is affected. The results of our experiments involving (i) perfusing the medulla and pons separately, (ii) performing microdialysis and (iii) using microejections, presented herein, strongly suggest that 5-HT modulation of RF is mediated by structures located in the rostral ventro-lateral medulla. This region seems to play a critical role in the genesis and modulation of respiratory rhythmicity since in this area, (i) electrical stimulation initiates premature inspirations [3, 14, 15], (ii) electrolytic lesions suppress respiratory rhythmicity [3, 14, 15], (iii) unitary recordings reveal the existence of 'pacemaker-like' respiratory
neurons [4, 14, 15, 17], and (iv) local application ofcatecholamine reduces the RF [3]. Our 5-HT data theretbre indicate the functional importance of this area in defining the RF. The exact structures of the rostral ventro-lateral medulla on which 5-HT acts are difficult to determine but the following are some possible candidates: (i) the "pacemaker-like' respiratory neurons [4, 14, 15, 17, 18], (ii) the B3 serotonergic neurons located in this area close to respiratory neurons [1] and (iii) other 5-HT sensitive neurons. The mechanisms whereby 5-HT elicits an increase in the RF appear to be complex, since 5-HT at the medullary level increases the RF only if the ponto-medullary structures are intact. In the caudal ventro-lateral pons, A5 areas have been found to exert an inhibitory modulation on the activity of the medullary respiratory rhythm generator via a continuous release of endogenous noradrenaline on structures of the rostral ventro-lateral medulla [2, 3]. Section and lesion experiments which eliminated this inhibition (see above) confirmed these previous results since the resting RF was doubled and the effects of 5-HT on the RF were abolished. At least two hypotheses can be put forward to explain why 5-HT had no effect when the noradrenergic modulation originating from the A5 areas was eliminated. First, this may mean that the level of activity of the respiratory rhythm generator is defined by interactions between 5HT and noradrenaline. Secondly, the frequency of the respiratory rhythm generator may already be maximal after elimination of the A5 areas (mean rate around 10 c.min ~) and cannot be further increased by 5-HT. This hypothesis seems unlikely, however, since in some individual experiments, a faster respiratory rate (15-20 c.min -~) could be elicited by applying 5-HT medium to the intact brainstem. In conclusion, the respiratory rhythm generator, which has been assumed to be located in the rostral ventro-lateral medulla [4, 14, 15, 17, t 8], seems to be continuously modulated by a dual control system with an inhibitory noradrenergic [2, 3] and excitatory serotonergic influences. Since these mechanisms are functional at birth but not yet mature [5], any dysfunction in the bioaminergic maturation processes during the perinatal period, until a relatively stable configuration is achieved, might give rise to respiratory disorders such as obstructive [12] or central apneas. The authors would like to acknowledge the highly useful assistance of Mrs. A.M. Lajard, Mrs. L Pio and Mr. M. Manneville. They are most grateful to Dr. S. Iscoe for his valuable review of the manuscript. This research was supported by the CNRS, the INSERM and the Conseil R6gional PACA.
95 1 Connelly, C., Ellenberger, H. and Feldman, J., Are there serotonergic projections from raphe and retrotrapezoid nuclei to the ventral respiratory group in the rat? Neurosci. Lett., 105 (1989) 34~,0. 2 Errchidi, S., Hilaire, G. and Monteau, R., Permanent release of noradrenaline modulates respiratory frequency in the newborn rat: an in vitro study, J. Physiol., 429 (1990) 497-510. 3 Errchidi, S., Monteau, R. and Hilaire, G., Noradrenergic modulation of the medullary respiratory rhythm generator in the newborn rat, J. Physiol., 443 (1991)477498. 4 Feldman, J., Smith, J.C. and Liu, G., Respiratory pattern generation in mammals: in vitro en bloc analyses, Curr. Op. Neurobiol., 1 (1991) 590 594. 5 Hamon, M. and Bourgoin, S., Characteristic of 5-HT metabolism and function in developing brain, in Osborne (Ed.), Biology of Serotonergic Transmission. J. Wiley, New York, 1982, pp. 197 220. 6 Hilaire, G., Monteau, R., Gauthier, P., Rega, P. and Morin, D., Functional significance of the dorsal and ventral respiratory group in adult and newborn rats: in vivo and in vitro studies, Neurosc'i. Lett.,lll(1990) 133 138. 7 Millhorn, D., Eldridge, F., Waldrop, T. and Klingler, L., Centrally and peripherally administered 5-HTP have opposite effects on respiration, Brain Res., 264 (1983) 349 354. 8 Millhorn, D.E., Physiological significance of pharmacological studies of respiratory regulation. In C. Von Euler and H. Langercrantz (Eds.), Neurobiology of the Control of Breathing, Raven, New York, 1986, pp. 89 95. 9 Monteau, R., Morin, D., Hennequin, S. and Hilaire, G., Differential effects of serotonin on respiratory activity of hypoglossal and cervical motoneurons: an in vitro study on the newborn rat, Neurosci. Lett., 111 (1990) 127 132. 10 Morin, D., Hennequin, S., Monteau, R. and Hilaire, G., Depressant effect of raphe stimulation on inspiratory activity of the hypoglossal
11
12
13
14
15
16 17
18
19
nerve: in vitro study in the newborn rat, Neurosci. Lett., 116 (1990) 299-- 303. Morin, D., Hennequin, S., Monteau, R. and Hilaire, G., Serotonergic influences on central respiratory activity: an in vitro study in the newborn rat, Brain Res., 535 (1990) 281 287. Morin, D., Monteau, R. and Hilaire, G., 5-Hydroxytryptamine modulates central respiratory activity in the newborn rat: an in vitro study, Eur. J. Pharmacol., 192 (1991) 89 95. Morin, D., Monteau, R., and Hilaire, G., Compared effect of 5-HT on the inspiratory activity of hypoglossal and cervical motoneurons: an in vitro study in the newborn rat, J. Physiol., 451 (1992) 605 629. Onimaru, H. and Homma, I., Respiratory rhythm generator neurons in medulla ofbrainstem-spinal cord preparation from newborn rat, Brain Res., 403 (1987) 380 384. Onimaru, H., Arata, A. and Homma, I., Localization of respiratory rhythm generating neurons in the medulla of brainstem-spinal cord preparations of newborn rats, Neurosci. Lett., 78 (1987) 151 155. Paxinos, G., TOrk, I., Tecott, L. and Valentino, K., Atlas of the Developing Rat Brain, Academic Press, New York, 1991. Smith, J.C., Greer, J.J., Liu, G. and Feldman, J.L., Neural mechanisms generating respiratory pattern in mammalian brain stem-spinal cord in vitro: spatiotemporal patterns of motor and medullary neuron activity, J. Neurophysiol., 64 (1990) 1149 1169. Smith, J.C., Ellenberger, H.H., Ballanyi, K., Richter, D.W. and Feldman, J.L., Pre-B6tzinger complex: a brainstem region that may generate respiratory rhythm in mammals, Science. 254 (1991) 726 729. Suzue, T., Respiratory rhythm generation in the in vitro brain stemspinal cord preparation of neonatal rat, J. Physiol., 354 (1984) 173 183.
91
© 1992 Elsevier Scientific Publishers Ireland Ltd. All rights reserved 0304-3940/92/$ 05.00 NSL 08864
Serotonergic modulation of the respiratory rhythm generator at birth: an in vitro study in the rat E. Di P a s q u a l e , D. M o r i n , R. M o n t e a u a n d G . H i l a i r e Biologie des Rvthmes et du D~;veloppement, URA C N R S 0205. Faculty; St. d~;rdme. Marseil/us ( k)'am'~' ;
(Received 26 March 1992: Revised version received 20 May 1992: Accepted 20 Ma5 1992) Kqv wor&v
Respiratory center; Serotonin, Neuromodulation, Newborn rat: In vitro
In order to investigate the mechanisms through which serotonin (5-HT) modulates the activity of the respiratory rhythm generator, respiratory activity was recorded from cervical ventral roots of the superfused isolated brainstem-spinal cord preparation of the newborn rat. Replacing the normal bathing medium by a medium containing 5-HT (30/.tM) increased the respiratory frequency by 70% of the control value. Intact pontomedullary structures are necessary for this effect to take place, however, since the 5-HT-induced increases in respiratory frequency were no longer observed after elimination (section and electrolytic lesion) of the caudal ventro-lateral pons containing the A5 areas. Local applications of 5-HT (dual bath, microdialysis and microinjection experiments) revealed, however, that 5-HT acts at the medullary level and that its effects are not due to a diffuse action on all the neurons of the medullary respiratory centers but to a specitic action focusing on structures located in the rostral ventro-latcral medulla.
Studies on the n e u r o m o d u l a t i o n o f the central respiratory activity exerted by endogenous substances such as serotonin (5-HT) have often yielded conflicting results since, depending on the m o d e of administration (peripherally or centrally), various respiratory effects [7, 8] can be elicited by 5-HT. The reduced preparation consisting of isolated n e w b o r n rat brainstem-spinal cord [19] is still able to produce s p o n t a n e o u s and rhythmic respiratory discharges in vitro in the phrenic nerves which are assumed to be generated by the same system that operate in vivo [17] or at least a part o f this system: this preparation has proved to be a useful tool for analyzing respiratory n e u r o m o d u l a t o r y processes since the elimination o f all the peripheral effects means that the results can be attributed with certainty to central mechanisms only. Previous pharmacological experiments performed with this preparation d e m o n s t r a t e d that the activity o f the medullary respiratory r h y t h m generator is increased by 5-HT. A d d i n g exogenous 5-HT to the bathing medium [9, 1 1], activating the 5-HT biosynthesis chain [12] and stimulating raphe structures [10] increased the respiratory fi'eq u e n c y . ( R F ) . This effect appears to be mediated by 5HTt receptors [12], since (i) it was blocked by 5-HT~ 2 Correspondencu." E. Di Pasquale, Biologie des Rythmes et du Ddveloppement, URA CNRS 0205, Facult6 St. Jdr6me, BP 332, Marseilles. France.
(methysergide) but not by specific 5-HT 2 antagonists (ketanserine), and (ii) it could be elicited by 5-HT t (8O H - D P A T ) but not 5-HT_~ agonists (~-CH3-5-HT). The aim o f the present report was to shed some light on the mechanisms whereby 5-HT may modulate the R F by limiting the access o f 5-HT to certain parts o f the brainstem. Results reveal that intact ponto-medullary structures are necessary for 5-HT modulation of the R F to occur, but the structures on which 5-HT acts are likely located in the rostral ventro-lateral medulla, an area which has been reported to contain the "pacemaker-like" neurons of the respiratory centers [4, 14, 15, 17, 18]. The brainstem and cervical spinal cord oF newborn rats (0--3 days old) were dissected and placed i~ ~i 2 ml in vitro c h a m b e r superfused with artifii:ial cerebl'ospinal fluid, using procedures identical to those previously described [2, 3]. Briefly, electrical activity from a phrenic nerve root was recorded using suction electrodes, filtered (5-3,000 Hz), amplified (Neurolog System, Digitimdr), integrated (leaky integrator, time constant 50 ms) and displayed on an oscilloscope and a chart recorder. Lesions were performed by passing D C currents (1 mA, 5 s) via tungsten microelectrodes. Serotonin (Sigma) was added to the normal medium (30 y M ) and superfused (2 ml/min) t\~r 5 rain. To apply 5-HT to the whole preparation, the pons alone, or the medulla alone, a barrier (improved with vaseline) was
92
:'()(}
• .::
,- .:
-- ~
.~ 15{) >
.,, i
C
I ?5 15(}
175
100
125 100
125 75
75
50
5{)
25
25
-) ~ ' ~ ; ; - --~ -- ''q . . . . .
) -'l~i
0
5 HT
~" 20O
zoo
D
5 HT
Time (min)
E
20o
175
175
175
'~
150
150
150
~
125
125
c
0
F
t25
pr"-~
100
100 75
I00 75
50
50
50
25
25
25
I
75
0
5 HT
Time (min)
Time (min)
0
5HT
Time (min)
5HT
Time (rain)
Fig. 1. Effects of general application of 5-HT on respiratory rhythm. A: schematic arrangement showing the type of inspiratory discharge recorded on the brainstem-spinal cord preparation of newborn rat. B F: in each graph, mean respiratory frequency (expressed as a percentage of the mean control respiratory frequency) and S.E.M. plotted every minute vs. time (expressed in rain). First 4 rain: control under normal medium; following 5 min: test under 5-HT medium (30/JM, black area); last 5 min: recovery under normal medium. B-D: 5-HT bathing under normal conditions increased the respiratory frequency (B, n=14) but this effect was not observed in C (n= 15) after transection performed at the level of the ponto-medullary junction (see double-headed arrow in A) or in D (n=6) after bilateral coagulation of A5 (large black dots in A). E-F: double bath experiment with a barrier placed at the level of the double-headed arrow; 5-HT elicited an increase in the respiratory frequency when applied to the medulla (F) but not when applied to the pons (E).
installed at the ponto-medullary border (double-headed arrow in Fig. 1).In 30 experiments, multibarelled micropipettes were used to apply 5-HT in specific sites of the brainstem (raphe nuclei, nucleus ambiguus, rostral ventro-lateral medulla) by pressure pulse (100 mM, 1-2 s). Prior to all the experiments, control ejections were performed under direct visual observation using a microscope: the ejected volumes never exceeded 5 nl. In 17 microdialysis experiments, a microinjection pump (Carnegie Med. CMA/100, flow rate 2 pl/min) connected to a syringe selector (CMA/111, to select either normal medium or medium containing 5-HT) and a microdialysis probe (CMA/11, outer diameter 240/.tm, length 1 mm) was used to slowly apply 5-HT by diffusion in specific parts of the brainstem. The probe, fixed to a micromanipulator (LPC) was inserted vertically 1 mm below the ventral surface into either the raphe nuclei or the rostral ventro-lateral medulla. The probe was perfused for 30 min with a normal medium and 5-HT medium was then applied for 5 min (500 pM; since the recovery of the probe was in the range of 10% in vitro, the actual 5-HT concentration in the tissue was assumed to be about 50
pM). Normal medium and 5-HT medium (applied either by superfusion or microdialysis) had identical pH (7.3). In order to be able to make comparisons with previously published results [9, 12], identical experimental procedures and data compilation methods were adopted. A control period of at least 5 rain was used to define the mean RF every minute prior to any experimental manoeuvre. The resulting frequency changes were expressed as percentages of the mean control values (100%) in the figures, but the absolute values in c.min -~ are given in the text. The experiments were repeated on several brainstem preparations (at least 5) with a standardized procedure to evaluate the mean effects on the RF which were expressed as the mean _+ S.E.M. Any differences were tested using Student's t-test and were taken to be significant at P values lower than 0.05. The activity recorded in vitro from the cervical ventral roots of newborn rat brainstem-spinal cord preparations is illustrated in Fig. 1A. This rhythmic activity, which was taken to constitute the respiratory frequency (RF), was stable under control conditions (Fig. 1B, mean rate 4.4 + 0.8 c-min ~, n=14). When the normal medium was
93
A
B
175
100 75 5O 25 0
5HT
a:'l
..j
D i t,l,i,lil Ill 1;:ii lp,rI t I[
!l!lll!
1!1
7
'
i, lj 1 Ilil |
,i.l-j-,3_.3- I
.~ 1'25
C
o
rdm
5 HT
T i m e (min)
l lllialll[nlJl,ltt,,lllltl
l til:l~Jl[.l_l,tll
T i m e (rain)
~l..it
[rl I lYVIIlrrrlVITTI I I1I I [Tll FIT r v'J r r T"r'f l l T rll7 •
5HT
E
8
2 min
Fig. 2. Effects of local application of 5-HT on respiratory rhythm. A: schematic illustration showing the site of local 5-HT applications (black dot) in the rostral ventro-lateral medulla. B,C: same ordering as in Fig. 1B F showing the mean increase in respiratory frequency elicited when 5-HT was applied to the rostral ventro-lateral medulla either by microdialysis (B, 500/aM. n 9) or by pressure pulse via multibarrcllcd micropipettes (('. 100 raM, n 25L D,E: the inspiratory discharge recorded from cervical root in two different preparations: note the increases in respiratory l'rcqucncy elicited by 5-HT microdialysis (D, between arrows} and 5-HT pressure microinjection (E, at arrm~) in individual cxpcriments. Time as indicatcd.
replaced by a medium containing 5-HT ( 3 0 / a M ) for 5 min, the R F increased during the first minute o f perfusion, and reached a peak within 2 3 min (mean: 7.5 + 1.4 c-min ~; n=14: range 5- 20 c.min L). Resuming the control conditions .'estored the control R F within 5 10 min. Eliminating the pons by performing a transection at the level o f the p o n t o - m e d u l l a r y junction (double-headed arrow in Fig. IA) increased the R F to 11.5 +_ 0.7 c.min ] (see also ret;. 2 and 3). 5-HT medium no longer significantly increased the R F (n=15, Fig. IC). To localize the pontine structures necessary for 5-HT excitatory effects to be exerted on the RF, microdissections were performed in the pons under binocular control. 5-HT-induced increases in the R F were not significantly altered by eliminating the medial region o f the pons (hatched area in Fig. 1: n=9) but were abolished by eliminating the lateral regions (dotted area in Fig. 1: n=5). To define more precisely what structures in the lateral pons were responsible for these respiratory e f fects, bilateral electrolytic lesions were placed in the caudal ventro-lateral pons to eliminate A5 areas by passing D C currents via a lungsten microelectrode inserted just medial to the entry o f the V roots (large filled circles in
Fig. IA, see refs. 2, 3 and 16) to 300/am below the venlral surface, The R F was increased by both electrolytic lesions and ponto-medullary sections (no statistical difference) and the 5-HT-induced increases in the R F were suppressed (n=6, Fig. 1D), To determine whether 5-HT acts on pontine o1" medullaw targets, a barrier was phtced at the ponto-medullary junction (double-headed arrow in Fig. I A) with a view to applying various media to either the pons or tile medulla. When the pons was bathed with 5-HT medium (30/aM) and the medulla with normal medium, no significant changes in thc R F were elicited (n-10: Fig. IE). When the ports was bathed with nornial medium and the medulla with 5-HT medium, the R F increased signiticantly (9.1 _+ 1.7 c.min ',/7=11, Fig. IF): the mean effect was not significantly different from that evoked by applying 5-HT to the whole brainstem. Intact ponto-medullary structures are therefore necessary for 5-HT modulation o f the respiratory r h y t h m generator to occur, but the structures on which 5-HT acts are probably in the medulla. These medullary structures were identified by performing microinjections o f 5-HT via multibarelled mi-
94 cropipettes inserted into the brainstem (see ref. 16). No significant changes in the RF resulted from injecting 5HT in the vicinity of the ponto-medutlary raphe nuclei (n=8) and the ambiguus nuclei (n--26). Increases in the RF were frequently elicited by injecting 5-HT into the rostral ventro-lateral medulla, 200-300/lm below the ventral surface. The increase in the RF occurred during the first minute after the injection (8.3 + 1.9 c.min ~, n=15, Fig. 2C,E); the RF returned to control values within 2-3 min. At these sites, saline injections had no effects. Microdialysis experiments were performed to apply 5HT locally to the rostral ventro-lateral medulla during longer periods. The microdialysis probe perfused with normal bathing medium was first slowly inserted in the rostral ventro-lateral medulla to a depth of 1 mm. Depending on the experiments, the RF was either unaffected (n=2), decreased (n--2), increased (n=3) or was suppressed for a while (5-10 min, n=2) when the probe was lowered in the rostral ventro-lateral medulla. After a 30-rain period, the control R F was restored and the microdialysis probe was then perfused with a 5-HT medium (500/~m) for 5 min instead of the normal one. The RF increased within 2 min (8.7 _+ 1.6 c.min -1) and remained high during the 5 min dialysis (n=8, Fig. 2B,D). Replacing the dialysate with a normal medium resulted in the RF returning to control values with a period of 15--20 min. When the microdialysis probe was placed into medial raphe sites at either the pontine (n=4) or the medullary (n--5) levels, 5-HT perfusion of the probe did not significantly affect the RF. The present results fully agree with previous studies [9-13] describing the excitatory modulation of the respiratory rhythm generator exerted by 5-HT and suggest furthermore that 5-HT acts specifically on structures in the rostral ventro-lateral medulla but not diffusely on the neurons of the ventral respiratory group [6]. This does not rule out the possible existence of other sites of action for 5-HT, such as the hypoglossal nucleus where it reduces the hypoglossal inspiratory activity [13], but in this case it is the amplitude, and not the frequency, of the motor output which is affected. The results of our experiments involving (i) perfusing the medulla and pons separately, (ii) performing microdialysis and (iii) using microejections, presented herein, strongly suggest that 5-HT modulation of RF is mediated by structures located in the rostral ventro-lateral medulla. This region seems to play a critical role in the genesis and modulation of respiratory rhythmicity since in this area, (i) electrical stimulation initiates premature inspirations [3, 14, 15], (ii) electrolytic lesions suppress respiratory rhythmicity [3, 14, 15], (iii) unitary recordings reveal the existence of 'pacemaker-like' respiratory
neurons [4, 14, 15, 17], and (iv) local application ofcatecholamine reduces the RF [3]. Our 5-HT data theretbre indicate the functional importance of this area in defining the RF. The exact structures of the rostral ventro-lateral medulla on which 5-HT acts are difficult to determine but the following are some possible candidates: (i) the "pacemaker-like' respiratory neurons [4, 14, 15, 17, 18], (ii) the B3 serotonergic neurons located in this area close to respiratory neurons [1] and (iii) other 5-HT sensitive neurons. The mechanisms whereby 5-HT elicits an increase in the RF appear to be complex, since 5-HT at the medullary level increases the RF only if the ponto-medullary structures are intact. In the caudal ventro-lateral pons, A5 areas have been found to exert an inhibitory modulation on the activity of the medullary respiratory rhythm generator via a continuous release of endogenous noradrenaline on structures of the rostral ventro-lateral medulla [2, 3]. Section and lesion experiments which eliminated this inhibition (see above) confirmed these previous results since the resting RF was doubled and the effects of 5-HT on the RF were abolished. At least two hypotheses can be put forward to explain why 5-HT had no effect when the noradrenergic modulation originating from the A5 areas was eliminated. First, this may mean that the level of activity of the respiratory rhythm generator is defined by interactions between 5HT and noradrenaline. Secondly, the frequency of the respiratory rhythm generator may already be maximal after elimination of the A5 areas (mean rate around 10 c.min ~) and cannot be further increased by 5-HT. This hypothesis seems unlikely, however, since in some individual experiments, a faster respiratory rate (15-20 c.min -~) could be elicited by applying 5-HT medium to the intact brainstem. In conclusion, the respiratory rhythm generator, which has been assumed to be located in the rostral ventro-lateral medulla [4, 14, 15, 17, t 8], seems to be continuously modulated by a dual control system with an inhibitory noradrenergic [2, 3] and excitatory serotonergic influences. Since these mechanisms are functional at birth but not yet mature [5], any dysfunction in the bioaminergic maturation processes during the perinatal period, until a relatively stable configuration is achieved, might give rise to respiratory disorders such as obstructive [12] or central apneas. The authors would like to acknowledge the highly useful assistance of Mrs. A.M. Lajard, Mrs. L Pio and Mr. M. Manneville. They are most grateful to Dr. S. Iscoe for his valuable review of the manuscript. This research was supported by the CNRS, the INSERM and the Conseil R6gional PACA.
95 1 Connelly, C., Ellenberger, H. and Feldman, J., Are there serotonergic projections from raphe and retrotrapezoid nuclei to the ventral respiratory group in the rat? Neurosci. Lett., 105 (1989) 34~,0. 2 Errchidi, S., Hilaire, G. and Monteau, R., Permanent release of noradrenaline modulates respiratory frequency in the newborn rat: an in vitro study, J. Physiol., 429 (1990) 497-510. 3 Errchidi, S., Monteau, R. and Hilaire, G., Noradrenergic modulation of the medullary respiratory rhythm generator in the newborn rat, J. Physiol., 443 (1991)477498. 4 Feldman, J., Smith, J.C. and Liu, G., Respiratory pattern generation in mammals: in vitro en bloc analyses, Curr. Op. Neurobiol., 1 (1991) 590 594. 5 Hamon, M. and Bourgoin, S., Characteristic of 5-HT metabolism and function in developing brain, in Osborne (Ed.), Biology of Serotonergic Transmission. J. Wiley, New York, 1982, pp. 197 220. 6 Hilaire, G., Monteau, R., Gauthier, P., Rega, P. and Morin, D., Functional significance of the dorsal and ventral respiratory group in adult and newborn rats: in vivo and in vitro studies, Neurosc'i. Lett.,lll(1990) 133 138. 7 Millhorn, D., Eldridge, F., Waldrop, T. and Klingler, L., Centrally and peripherally administered 5-HTP have opposite effects on respiration, Brain Res., 264 (1983) 349 354. 8 Millhorn, D.E., Physiological significance of pharmacological studies of respiratory regulation. In C. Von Euler and H. Langercrantz (Eds.), Neurobiology of the Control of Breathing, Raven, New York, 1986, pp. 89 95. 9 Monteau, R., Morin, D., Hennequin, S. and Hilaire, G., Differential effects of serotonin on respiratory activity of hypoglossal and cervical motoneurons: an in vitro study on the newborn rat, Neurosci. Lett., 111 (1990) 127 132. 10 Morin, D., Hennequin, S., Monteau, R. and Hilaire, G., Depressant effect of raphe stimulation on inspiratory activity of the hypoglossal
11
12
13
14
15
16 17
18
19
nerve: in vitro study in the newborn rat, Neurosci. Lett., 116 (1990) 299-- 303. Morin, D., Hennequin, S., Monteau, R. and Hilaire, G., Serotonergic influences on central respiratory activity: an in vitro study in the newborn rat, Brain Res., 535 (1990) 281 287. Morin, D., Monteau, R. and Hilaire, G., 5-Hydroxytryptamine modulates central respiratory activity in the newborn rat: an in vitro study, Eur. J. Pharmacol., 192 (1991) 89 95. Morin, D., Monteau, R., and Hilaire, G., Compared effect of 5-HT on the inspiratory activity of hypoglossal and cervical motoneurons: an in vitro study in the newborn rat, J. Physiol., 451 (1992) 605 629. Onimaru, H. and Homma, I., Respiratory rhythm generator neurons in medulla ofbrainstem-spinal cord preparation from newborn rat, Brain Res., 403 (1987) 380 384. Onimaru, H., Arata, A. and Homma, I., Localization of respiratory rhythm generating neurons in the medulla of brainstem-spinal cord preparations of newborn rats, Neurosci. Lett., 78 (1987) 151 155. Paxinos, G., TOrk, I., Tecott, L. and Valentino, K., Atlas of the Developing Rat Brain, Academic Press, New York, 1991. Smith, J.C., Greer, J.J., Liu, G. and Feldman, J.L., Neural mechanisms generating respiratory pattern in mammalian brain stem-spinal cord in vitro: spatiotemporal patterns of motor and medullary neuron activity, J. Neurophysiol., 64 (1990) 1149 1169. Smith, J.C., Ellenberger, H.H., Ballanyi, K., Richter, D.W. and Feldman, J.L., Pre-B6tzinger complex: a brainstem region that may generate respiratory rhythm in mammals, Science. 254 (1991) 726 729. Suzue, T., Respiratory rhythm generation in the in vitro brain stemspinal cord preparation of neonatal rat, J. Physiol., 354 (1984) 173 183.