Brain Research, 578 (1992) 289-296 © 1992 Elsevier Science Publishers B.V. All rights reserved. 0006-8993/92/$05.00
289
BRES 17651
Interaction between thyrotropin-releasing hormone (TRH) and NMDA-receptor-mediated responses in hypoglossal motoneurones Jens C. Rekling Institute of Neurophysiology, University of Copenhagen, Copenhagen (Denmark) (Accepted 10 December 1991)
Key words: Thyrotropin-releasing hormone; Hypoglossal motoneuron; Transmitter interaction; Neuromodulation, N-Methyl-o-aspartate; Excitatory amino acid; Brain slice
The effect of thyrotropin-releasing hormone (TRH) on the responses to excitatory amino acids was investigated in hypoglossal motoneurones in an in vitro preparation of the brainstem from guinea pigs using current clamp and discontinuous single electrode voltage clamp (dSEVC). Bath application of 20-50/~M TRH markedly potentiated the response to iontophoretically applied NMDA, whereas no potentiation of the response to glutamate, aspartate or quisqualic acid was seen. Voltage clamp experiments showed that TRH did not increase the current flowing through NMDA channels, thus a direct modulatory role of TRH on NMDA channels was not a likely explanation of the potentiation. Voltage clamp studies of the current-voltage relationship showed that the potentiation of the response to NMDA and lack of potentiation of the response to quisqualic acid was a result of an interaction between the actions of TRH and the amino acids on the electroresponsive profile of the membrane. Endogenous NMDA receptor activation was produced by tetanic stimulation of the reticular formation dorsolaterally to the hypoglossal nucleus, evoking large APV sensitive EPSPs in the presence of CNQX, a non-NMDA blocker. The amplitude and duration of these potentials were increased at more positive membrane potentials in response to TRH. It is concluded that TRH can act as a neuromodulator-potentiating the response to NMDA receptor activation-simply by changing the electroresponsive properties of the membrane. INTRODUCTION I m p r o v e m e n t s in immunohistochemical and in situ hybridization techniques have revealed that the T R H systems in the m a m m a l i a n brain are considerably m o r e extensive than previously assumed 7'22. The initial interest in this tripeptide was centered around its role in the hypothalamic control of secretion of thyroid-stimulating h o r m o n e (TSH). The transmitter action of T R H in extrahypothalamic tissues has, however, attracted increasing attention recently. T R H has been shown to have potent actions on the m o t o r system at several anatomical levels. Injection of T R H into the nucleus accumbens induced increased l o c o m o t o r activity and behaviors such as licking, chewing and sniffing 4. Intrathecal administration of T R H induced wet-dog shakes and forepaw-licking behavior, possibly resulting from an action at the level of the spinal cord or brainstem 9. Injection of T R H in the fourth ventricle increased respiratory rate suggesting an action on the medullary respiratory centers 5. A t the level of the spinal cord T R H has been shown to increase the excitability of spinal m o t o n e u r o n e s as m e a s u r e d by an increase in the antidromic field potentials, ventral root depolarizations and increase in the monosynaptic reflex 2'
3,19. T h e effects of T R H in the spinal cord has been explained by a direct action of T R H on spinal motoneurons 17'18'24 and facilitation of presynaptic release of excitatory transmitters 12. The anatomical distribution of extrahypothalamic T R H also involves several brainstem m o t o r nuclei, including the hypoglossal nucleus, where the native horm o n e , immunohistochemically characterized T R H - c o n taining nerve terminals and T R H binding sites have been found 6'1t'13'14. m study from this l a b o r a t o r y provided electrophysiological indications for a transmitter action of T R H in the hypoglossal nucleus, showing that T R H induced depolarization and increase in the input resistance of hypoglossal m o t o n e u r o n e s 2°. Like many of the o t h e r putative neuropeptidergic transmitters T R H is believed to have a n e u r o m o d u l a t o r y function-interacting with o t h e r transmitter systems. In cortical neurons T R H has been found to potentiate the excitatory action of acetylcholine 1 and in rat spinal motoneurones in vivo T R H enhanced both glutamate- and a s p a r t a t e - e v o k e d activity25. This study investigates the actions of T R H on the response to excitatory amino acids in hypoglossal m o t o n e u r o n e s and demonstrates how a n e u r o m o d u l a t o r y role of T R H can be explained by the
Correspondence: J.C. Rekling, Institute of Neurophysiology, Panum Institute, Blegdamsvej 3C, 2200 Copenhagen N, Denmark.
290 action of T R H itself o n t h e e l e c t r o r e s p o n s i v e profile o f the n e u r o n a l m e m b r a n e .
MATERIALS AND METHODS The experiments were performed on transversely cut brainstem slices from albino guinea pigs according to a previously described method 16. In brief, 400/~m thick slices from the brainstem were immediately transferred to the recording chamber. It had a volume of 2 ml, a temperature of 36.5°C and was constantly superfused at a rate of 2 ml/min with preheated, oxygenated Ringer's solution composed of 134 mM NaC1, 2 mM KCI, 20 mM NaHCO 3, 1.25 mM KH2PO 4, 1.5 mM CaC12, 2 mM MgC12, 9 mM glucose. The solution was bubbled with 95% 02/5% CO 2 and this gas mixture also filled the space above the medium in the recording chamber. After a period of 30 rain to 1 h the slices were transilluminated with a point source from below and the recording electrode placed in the hypoglossal nucleus under visual guidance through a stereomicroscope. Recording system. Electrodes were pulled from filamented glass tubes (outer diameter of 1.5 mm, inner diameter of 0.8 mm) and filled with a 2-3 M solution of potassium acetate. The electrodes had resistances of between 15 and 60 MfL A bridge-balance amplifier (built in the laboratory) with a cut-off frequency of 8 kHz was used in the current clamp experiments. Discontinuous single electrode voltage clamp (dSEVC) was performed using the homebuilt amplifier or an Axoclamp 2A amplifier at switching frequencies of 2.5 to 3 kHz. The recorded signals were stored on a 4-track tape recorder (Racal) with a band width of DC to 2.5 kHz. Offline analysis was done on a personal computer digitizing the signals at 5 to 82 kHz. lontophoresis. Microelectrodes with broken tips were filled with NMDA (N-methyl-D-aspartic acid; 0.05 M; Sigma), L-glutamic acid (0.5 M; Sigma), L-aspartic acid (0.225 M; Sigma) or quisqualic acid (0.05 M; Sigma) dissolved in distilled water, pH was adjusted to 8 with NaOH. The iontophoresis electrode was positioned in the slice independently of the recording electrode and a bridge-balance amplifier was used as current source (50-100 nA ejection current and 10-30 nA backing current). Extracellular stimulation. Electrical stimulation was done between two isolated, 5/~m thick, twisted platinum wires. The unisolated
A
,~...~
ends of the platinum wires, spaced 100 pm, were placed in the slice, dorso-lateral to the hypoglossal nucleus in the reticular formation. The rectangular pulses lasted 100 ~s at voltages of 25 V to 40 V. Materials. Thyrotropin-releasing hormone (Cambridge Research Biochemicals Ltd.) was applied to the superfusion solution reaching a final concentration of 20, 40 or 50/~M, TTX (tetrodotoxin; Sigma) was used in a concentration of 1/~M. CNQX (6-cyano-2,3dihydroxy-7-nitro-quinoxaline; Tocris Neuramin) was used at 50 pM, APV (DL-2-amino-5-phosphono-valeric acid; Sigma) at 100 /~M. QX-314 (A generous gift from Astra) was dissolved in 3 M potassium acetate pipette solution reaching a final concentration of 100-200 mM.
RESULTS T h e findings p r e s e n t e d h e r e are b a s e d o n stable intracellular r e c o r d i n g s f r o m 45 hypoglossal m o t o n e u r o n s . T h e i d e n t i t y o f the cells was b a s e d o n the characteristic electrophysiological
properties
previously
d e s c r i b e d 16
and p o s i t i o n i n g of the r e c o r d i n g e l e c t r o d e inside the b o u n d a r i e s o f the hypoglossal nucleus, which was visible in the u n s t a i n e d slice.
Potentiation of the response to iontophoretically applied NMDA In o r d e r to test for a n e u r o m o d u l a t o r y action o f T R H in the hypoglossal m o t o n e u r o n e s i o n t o p h o r e s i s of exci° t a t o r y a m i n o acids was p e r f o r m e d . T h e i o n t o p h o r e s i s e l e c t r o d e was p l a c e d in the slice i n d e p e n d e n t l y of the r e c o r d i n g e l e c t r o d e , but in close p r o x i m i t y to the tip o f it, j u d g e d visually. W h e n p e r f o r m i n g i o n t o p h o r e s i s c a r e was t a k e n to o b t a i n a r e g u l a r cycle o f c u r r e n t e j e c t i o n and as small a h o l d i n g c u r r e n t as possible. C u r r e n t was i n j e c t e d into the e l e c t r o d e for 4 s at fixed intervals (30
C
B
2O mV
~
NMDA
D
NMDA
E
F
"l"rx
l"rX, TRH
NMD'~
TTX, W~mh
NMD'~'A
Fig. 1. Potentiation of the response to NMDA iontophoresis by TRH. A, B and C: iontophoresis of NMDA before, during TRH (20/~M) and after 10 min of wash. D, E and F: iontophoresis of NMDA performed with TI'X (1/~M) in the superfusion solution. Membrane potential was kept constant using bias current.
291
A
.-.
B
,.--.-
Control
C Wash
TRH
~u
D
E
F
G
...-
H
I
Control
QuJs
Wash
TRH
eamrol
I
TFIH
I
~
Fig. 2. Action of TRH on iontophoresis of glutamate, aspartate and quisqualic acid. TRH did not potentiate the response to iontophoresis of glutamate (A, B, C), aspartate (D, E, F) or quisquahc acid (G, H, I) at a concentration of 50/~M. Wash (10 min). Membrane potential was kept constant using bias current.
or 40 s), since this regime resulted in the smallest change in the response to the compounds over time. The first compound tested was NMDA and as shown in Fig. 1 TRH had a striking effect on the iontophoretic NMDA response. In Fig. 1A,B,C a pulse lasting 4 s driving NMDA out of the iontophoresis pipette was applied before (Control), during T R H (20/~M; 3 min after start of superfusion) and 10 min after re-entering a normal Ringer's solution (Wash). The membrane potential was kept constant, counteracting the depolarizing action of TRH, using hyperpolarizing bias current and the extracellular iontophoresis artefact has been subtracted. During TRH, NMDA iontophoresis evoked a much longer spike train commencing earlier after the onset of the pulse. When the test was performed with TI'X (1 /~M) in the superfusion solution (Fig. 1D,E,F) the depolarizing action of NMDA was markedly potentiated during TRH. The depolarizing potential rose faster after the onset of the pulse and 18 mV positive to resting potential a 'high noise' plateau potential emerged. Eleven cells were tested with NMDA iontophoresis and all cells displayed potentiation of the response by TRH. During continued superfusion with T R H the action on the NMDA response desensitized, pursuing the same time course as the desensitization of the depolarizing action of T R H previously described 2°.
No potentiation of the response to glutamate, aspartate or quisqualic acid A similar test was done iontophoresing glutamate (5
cells), aspartate (3 cells) and quisqualic acid (4 cells) (Fig. 2). Attempts to iontophorese kainic acid were unsuccessful, because of the manifest toxic effects of this compound on the viability of the hypoglossal motoneurons. None of the neurons tested with these excitatory amino acids showed potentiation of the response when T R H (20-50 #M) was added to the superfusion solution.
Voltage dependence of the NMDA potentiation Fig. 3 shows that the size of the potentiation of the NMDA response by TRH is dependent on the membrane potential. Iontophoresis of NMDA was performed in a ~ (1/~M) containing medium and hyperpolarizing pulses of constant strength were given to monitor the apparent input resistance (2 cells). Using bias current the NMDA iontophoresis was given from three different membrane potentials (-66, -70 and -74 mV) before and during TRH (20/~M). Note that the control response to NMDA is voltage dependent, increasing at more positive potentials due to a decrease in the Mg 2+ block. During T R H the potentiation of the depolarizing action of NMDA was very pronounced at the two most depolarized holding potentials (-66 mV and -70 mV), measured as the area under the curve (80% increase at -66 mV, 113% increase at -70 mV). At -74 mV the potentiation was much smaller (22% increase). In terms of amplitude and input resistance the relative difference is not so clear probably because the amplitude of the response during TRH reaches a plateau at the two most depolarized potentials, but both parameters are substantially smaller at
292
A
",-rx
NMDA
C
TTX,TRH
NMDA
~
D
NMDA
E
Voltage clamp studies
B
NMDA
~-x
F
Tr~,
J
2OmV
NMDA
NMDA
Fig. 3. Voltage dependence of the TRH induced potentiation of the NMDA response. A, C and E: iontophoresis of NMDA performed in a TTX (1 ~M) containing solution rising from three different membrane potentials (-66 mV, -70 mV, -74 mV). Hyperpolarizing current pulses of constant strength were given to monitor input resistance. B, D and F: NMDA iontophoresis rising from the same membrane potentials as A, C, E, but 3 min into a TRH (50/~M) containing solution.
-74 mV than at -70 mV (amplitude: 58% increase vs. 133%, input resistance: 50% increase vs. 93%). The initial conclusion from these experiments is that T R H in a voltage-dependent manner potentiates the response to N M D A , but not the response to glutamate, aspartate or quisqualic acid. This raises two questions. Does T R H have a direct modulatory action on the current carried through N M D A channels? Or, does the change in the electroresponsive properties of the membrane induced by T R H somehow cause the response to N M D A to be augmented, but not the response to the other amino acid? To address these questions a series of voltage clamp experiments were performed.
A
co~
To investigate the possibility that the potentiation of the N M D A response was the result of a direct modulation of the current carried through N M D A channels, d S E V C (see Materials and Methods) of the iontophoretic response was performed (2 cells). In Fig. 4A a hypoglossal motoneurone in a TTX containing solution was voltage clamped at -63 mV, a membrane potential at which T R H induced a marked potentiation of the N M D A response in current clamp. Iontophoresis of N M D A (4 s pulse) induced a transient inward current. Three minutes after entering a T R H (50/~M) containing solution the test was repeated (Fig. 4B). As seen in Fig. 4B T R H induces a steady inward current (0.4 nA, dotted line), but the amplitude of the N M D A response is unaltered (-0.8 n A vs. - 0 . 8 nA). Thus T R H does not increase the current flowing through N M D A channels. In order to construct I - V curves before and during superfusion with T R H , N M D A or quisqualic acid a command voltage having the shape of a rising ramp was applied in d S E V C (10 cells). The rate of the ramp was relatively slow (6-11 mV/s) allowing transient currents to pass to steady state. The recordings shown in Fig. 5 were lowpass filtered at 0.3 kHz and smoothed digitally (running average of 20 ms windows). Fig. 5A (trace a) shows an I - V curve obtained from a neurone in a T I ' X (1/~M) containing solution. Three minutes after entering a solution containing 50 /~M T R H the I - V curve shifted downward due to the development of an inward current (trace b). The configuration of the curve was changed with a decrease in membrane slope conductance developing positive to -50 mV. These observations are in good agreement with the depolarization and increase in input resistance seen in current clamp in response to T R H 2°. A n intersection between the two curves at hyperpolarized potentials was not seen, probably because the current passing characteristics of the electrode did not permit c o m m a n d voltages negative to approximately -80 mV without losing control of the voltage. The interaction between N M D A and T R H is illustrated in Fig. 5B, showing a set of I - V curves obtained
B
TRH 20 mV
V
NMDA
NMDA
4s
Fig. 4. Voltage clamp of the iontophoretic NMDA response. A: NMDA iontophoresis performed in dSEVC mode inducing a transient inward current. TI'X (1/~M) was added to the medium. B: NMDA iontophoresis 3 min after entering a TRH (50/~M) containing solution. Note that TRH induces a steady inward current, but the response to NMDA is unchanged.
293 A
/
+.s
"~. . . . . . . i . . . . . . . . . . . . . . . .
~
-
~
"" ......
b. . . . . . .
45 -80
B
-60
..411
4
-20
~ a ~ . - - - ~ . . . . . - ....
2 ....... . - . . . . . . ~
-6 ...............................i..................................... I
-90
C
-75
-60
-45
-30
2
go 0
-3
-4 -90
:
:
-75
-60
i
-45
-30
Membrane potential (mV)
Fig. 5. I-V curves based on dSEVC during bath application of TRH, NMDA or quisqualic acid. A: plot of the current-voltage relationship in a solution containing TI'X (1 /~M; trace a) and 3 min after adding TRH (50 ~M; trace b). B: I-V curves in TI"X (1 gM; trace a), 10 min into superfusion with NMDA (100 gM; trace b) and 3 min after adding TRH (50 ~M; trace c). C: I-V curves in T r x (1 #M; trace a), 10 min into superfusion with quisqualic acid (150/~M; trace b) and 3 rain after adding TRH (50 #M; trace c).
in q-"l'X (1/~M; trace a), N M D A (100/~M; trace b) and N M D A + T R H (50/~M; trace c). A p p l i c a t i o n of N M D A to the supeffusion solution resulted in the d e v e l o p m e n t
A
B Normal ringer
E
C CNQX
F
/~TRH
of a p r o m i n e n t v o l t a g e - d e p e n d e n t inward current making the I - V curve highly non-linear. A s in the case of bath applied T R H alone, the m e m b r a n e slope conductance was decreased positive to a p p r o x i m a t e l y - 4 8 m V in response to N M D A . A p p l i c a t i o n of T R H (50/~M) in addition to N M D A in the solution resulted in developm e n t of an additional steady inward current, displacing the I - V curve in a parallel m a n n e r negative to - 5 0 mV. Positive to - 5 0 m V T R H induced a further decrease in the m e m b r a n e slope conductance, now resulting in a negative slope conductance, which causes flow of inward current throughout the c o m m a n d voltage range. A relative increase in inward current is thus induced by T R H acting together with N M D A at potentials positive to approximately - 5 0 mV. In Fig. 5C a set of I - V curves o b t a i n e d in T T X (1/~M; trace a), quisqualic acid (150 g M ; trace b) and quisqualic a c i d + T R H (50/~M; trace c) shows that activation of the A M P A type of excitatory amino acid r e c e p t o r does not lead to the same v o l t a g e - d e p e n d e n t action on the I - V relationship as N M D A r e c e p t o r activation. Quisqualic acid induced an inward current (trace b), but as o p p o s e d to N M D A the m e m b r a n e slope conductance was increased in the most depolarized region (positive to - 5 2 mV). W h e n T R H (50/~M) was a d d e d to the solution (trace c) additional inward current was induced, but the m e m b r a n e slope conductance of the cell now shifted back to the control value throughout the voltage range. Thus T R H acting together with quisqualic acid actually induced a relative change in the current flowing at m o r e positive m e m b r a n e potentials, but importantly this effect was a decrease in outward current (Fig. 5C; trace c). These voltage clamp experiments strongly point to the hypothesis that the potentiation of the N M D A response
D CNQX
G
/~TRH
CNOX, APV
10mV
H
f~/TRH
TRH
10mV
Control
Fig. 6. Effect of TRH on NMDA receptor mediated EPSPs. A, B, C, D: pharmacological isolation of a NMDA receptor mediated EPSP. A: EPSP elicited by a single stimuli, performed with QX-314 in the recording electrode. B: blockade of the EPSP by 50 gM CNQX, leaving behind only a very small prolonged depolarization. C: tetanic stimulation of the same cell as in A, B at 333 Hz lasting 40 ms, giving rise to a depolarizing potential. D: blockade of the depolarizing potential by APV (100/~M). All records are average of 8 to 10 sweeps. E, F, G and H: tetanic stimuli evoking NMDA receptor mediated EPSPs, elicited from four different membrane potentials, before and during TRH (50 /~M). Note the relative increase in amplitude at the two most positive potentials. All records are average of 8 to 10 sweeps.
294 and lack of potentiation of the other amino acids seen in current clamp should be understood in terms of the effects these compounds and T R H have on the I - V relationship.
negative potentials (-70 and -75 mV), however, were not potentiated but displayed a minor decrease in amplitude and a small prolongation of the falling phase. DISCUSSION
Voltage-dependent increase in the amplitude of NMDA receptor mediated EPSPs The potentiation of the iontophoretic NMDA response seen in current clamp could be an experimental curiosity with no apparent 'meaning' in terms of the synaptic integrative functioning of the motoneurons. Fig. 6, however, shows that TRH probably plays an important role in regulating the response to endogenous NMDA receptor activation. These experiments were based on pharmacological distinction between NMDA receptor and non-NMDA receptor activation using selective antagonists. Motoneurones were penetrated by electrodes containing QX-314 (100-200 raM) dissolved in 2 M potassium acetate, blocking sodium spikes, without affecting synaptic activation. Single electrical stimuli with a bipolar electrode placed dorsolaterally to the hypoglossal nucleus in the reticular formation evoked an EPSP (Fig. 6A). When the non-NMDA blocker CNQX (50/~M) was added to the superfusion solution the stimulus evoked EPSP disappeared, leaving behind a very small prolonged depolarization (Fig. 6B). A train of stimuli at 333 Hz lasting 40 ms was then applied. This induced a depolarizing potential with a longer rise- and decay time than an EPSP elicited by a single stimuli (Fig. 6C). As shown in Fig. 6D the depolarizing potential was blocked by APV (100/~M), a selective NMDA blocker (3 cells). Thus it can be concluded that the tetanic stimulation in the presence of CNQX elicits an EPSP caused by activation of NMDA receptors. A small depolarization, however, persisted after application of APV (Fig. 6D). This was probably a rebound depolarization, caused by a small summating train of IPSPs uncovered by the blockade of excitatory synaptic transmission (CNQX, APV), since the depolarization could be mimicked by a hyperpolarizing pulse of 40 ms driving the membrane potential to approximately the same hyperpolarized level. The action of T R H (50/~M) on the NMDA receptor mediated EPSP was investigated in 4 other cells one of which is shown in Fig. 6E,F,G,H. The average of 10 sweeps of EPSPs elicited from four different membrane potentials (-60, -65, -70 and -75 mV) is shown with the control record and the T R H record superimposed. When TRH (50/~M) was applied to the superfusion solution, the EPSPs starting from the two most positive membrane potentials (-60 and -65 mV) were greatly augmented in amplitude and duration (27% and 30% increase in amplitude). EPSPs starting from the two most
The purpose of this study was to test for a possible neuromodulatory role of T R H in the hypoglossal nucleus, inspired by several extracellular studies showing facilitatory or inhibitory effects of TRH on the response to other transmitters. The principal finding of the study was that T R H potentiated the response to NMDA receptor activation. It is suggested that the potentiation can be explained by an interaction between the effects TRH and NMDA have on the electroresponsive properties of the membrane and not a direct modulation of the current carried through NMDA channels. The modulatory actions of TRH previously reported mainly describes facilitatory or inhibitory effects of T R H on the response to excitatory amino acids or acetylcholine. TRH was found to enhance both glutamate- and aspartate-induced excitation of spinal motoneurons 25 and facilitate glutamate-induced firing in spinal cord dorsal cells8. In cortical neurons T R H potentiated the excitatory action of acetylcholine 1, but depressed the glutamate excitation 21. Another study, however, could not replicate the depressant action of T R H on amino acidinduced excitation 23, to the contrary the study showed an enhancement of the response to quisqualic acid and no effect on NMDA induced excitations. In hypothalamic neurons T R H was found to modulate the response to glutamate but not the response to acetylcholine ~°. All of the above-mentioned studies were performed using extracellular recordings and it was therefore of interest to see whether TRH had similar neuromodulatory actions in the hypoglossal motoneurones using intracellular recordings, which potentially can give a more precise characterization of the action of the different compounds on the neuronal membrane. Application of TRH on hypoglossal motoneurones resulted in a marked potentiation of the response to NMDA iontophoresis. TRH was previously shown to increase the input resistance of the hypoglossal motoneutons 2° and this in itself would tend to increase the response to excitatory input, thus providing a satisfactory explanation to the NMDA potentiation. It was therefore a puzzling finding that T R H did not affect the response to glutamate, aspartate or quisqualic acid. This selectivity then could be the result of a direct modulatory action of T R H on NMDA channels, but this is not consistent with the fact that glutamate as well as aspartate have NMDA receptor agonist properties. The second possibility was that the change in the electroresponsive
295 properties of the membrane induced by TRH somehow caused the response to N M D A to be augmented, but not the response to the other amino acids. The major electrophysiological difference between iontophoretic NMDA responses and responses to the other excitatory amino acids is the prominent voltage dependence of the NMDA response, brought about by the Mg 2÷ block 15. In a series of current clamp experiments not shown here this phenomenon was also noted in the hypoglossal motoneurons, reflected in the change in the input resistance during iontophoresis of the different amino acids. During iontophoresis of glutamate, aspartate and quisqualic acid the input resistance of hypoglossal motoneurones was decreased, in contrast to the increase during iontophoresis of NMDA. A further indication of an interaction between NMDA and T R H based on the effect these compounds have on the electroresponsive profile of the membrane was the observation that the facilitatory action of TRH on the NMDA response was sensitive to the holding potential from which the iontophoresis was given. Iontophoresis of NMDA from a relative negative membrane potential resulted in a much smaller potentiation than iontophoresis from more positive potentials. In order to determine whether TRH interacted with the currents induced by the amino acids a series of voltage clamp experiments were performed. In voltage clamp the inward current induced by NMDA iontophoresis was not affected by TRH, thus a direct modulatory effect on the NMDA channels was not a likely explanation of the potentiation seen in current clamp. A molecular interaction, however, depending on an unconstrained change in membrane potential in response to NMDA and T R H cannot be ruled out. Current-voltage relationship experiments showed that both NMDA and T R H applied to the bath alone induced an inward current with a decrease in membrane slope conductance positive to approximately -50 mV and when applied together the flow of current was inward throughout the command voltage range, with a negative membrane slope conductance at potentials positive to approximate -50 mV. This combined action on the membrane then explains the potentiation of the iontophoretic NMDA response seen in current clamp. During NMDA iontophoresis the membrane depolarized due to the inward current flowing through NMDA channels. When TRH was added to the bathing solution the iontophoresis was performed from the same membrane potential giving hyperpolarizing bias current to counterbalance the depolarizing effect of TRH. Because of the combined action of TRH and NMDA, the membrane slope conductance positive to -50 mV was now negative. This means that during NMDA iontophoresis, with TRH in the bath, an increased amount of inward current flowed at potentials positive to -50 mV,
potentiating the response to NMDA. The development of a negative slope conductance was probably the cause for generation of a plateau potential when the iontophoresis was done with TI'X in the medium (Fig. 1E), which will further augment the potentiation. Bath application of quisqualic acid in voltage clamp led to the development of a inward current with an increase in membrane slope conductance at more positive potentials. Importantly, however, the overall current flowing through the membrane at the more positive command potentials was outward. Adding TRH resulted in development of an additional inward current and as expected reduced the membrane slope conductance in the depolarized region of the I - V plot. The opposing action by T R H on the membrane slope conductance was, however, not great enough to change the overall current from outward to inward at the more positive potentials. Thus the following picture of the action of TRH on quisqualic acid iontophoresis in current clamp emerges. During quisqualic acid the membrane depolarized due to the inward current flowing through A M P A channels. Adding TRH to the bath induced a depolarization that was counterbalanced by bias current. The depolarizing action of quisqualic acid, then, was not augmented because the membrane potential never reached the point where TRH had changed the membrane slope conductance, since the overall current flowing through the membrane in this region was outward. This mechanism probably also holds true for glutamate and aspartate since these compounds like quisqualic acid produces a decrease in input resistance during iontophoresis in current clamp. This mechanism of action would also explain some of the conflicting findings of modulatory effects of T R H in other neuronal populations. A possible error in the conclusions drawn from extracellular studies of the modulatory effects of TRH seen in this context is the claim, that the modulation of a transmitter response by T R H cannot be due to a non-specific change in neuronal excitability induced by TRH, because the response to one transmitter is facilitated, but the response to another is not. This is incorrect, because the extracellular studies do not disclose whether one of the compounds have a voltage-dependent action of its own. The experiments on the pharmacologically isolated NMDA receptor mediated EPSPs show that the mechanism outlined above might have an important physiological function with respect to the integration of synaptic potentials in the hypoglossal motoneurons. When T R H had acted on the neurons, NMDA receptor mediated EPSPs were increased in amplitude and duration at more positive potentials. The increase in input resistance of the neurons induced by TRH should, however, also be taken into consideration in this phenomenon. An in-
296 crease in the input resistance of the cells will make the
with the combined effect on the m e m b r a n e slope con-
cells electrotonically more 'compact', facilitating the space clamp efficacy of the intracellular electrode. Hold-
ductance outlined above will never the less cause T R H to acts as a n e u r o m o d u l a t o r , boosting synaptic input card e d by the endogenous N M D A receptor ligand.
ing the m e m b r a n e potential at a value positive to the resting m e m b r a n e potential would then result in a dendritic m e m b r a n e potential closer to the somatic holding potential, when T R H is present. As a result of the voltage dependency of the N M D A receptor mediated EPSPs this would then m e a n a larger amplitude 'seen' from the somatic recording. This mechanism together
REFERENCES 1 Braitman, D.J., Auker, C.R. and Carpentar, D.O., Thyrotropin-releasing hormone has multiple actions in cortex, Brain Res., 194 (1980) 244-248. 2 Clarke, K.A. and Stirk, G., Motoneurone excitability after administration of a thyrotropin releasing hormone analogue, Br. J. Pharmacol., 80 (1983) 561-565. 3 Cooper, B.R. and Boyer, C.E., Stimulant action of thyrotropin releasing hormone on cat spinal cord, Neuropharmacology, 17 (1978) 153-156. 4 Heal, D.J. and Green, A.R., Administration of thyrotropin releasing hormone (TRH) to rats releases dopamine in n. accumbens but not in n. caudatus, Neuropharmacology, 18 (1979) 2331. 5 Homma, I., Oouchi, M. and Ichikawa, S., Facilitation of inspiration by intracerebroventricular injection of thyrotropin-releasing hormone in rabbits, Neurosci. Len., 44 (1984) 265-269. 6 Hrkfelt, T., Fuxe, K., Johansson, O,, Jeffcoate, S. and White, N., Distribution of thyrotropin-releasing hormone (TRH) in the central nervous system as revealed with immunohistochemistry, Eur. J. Pharmacol., 344 (1975) 389-392. 7 H6kfelt, T., Tsuruo, Y., Ulfhake, B., Cullheim, S., Arvidsson, U., Foster, G.A., Schultzberg, M., Schalling, M., Arborelius, L., Freedman, J., Post, C. and Visser, T., Distribution of TRHlike immunoreactivity with special reference to coexistence with other neuroactive compounds, Ann. N Y Acad. Sci., 553 (1989) 76-105. 8 Jackson, D.A. and White, S.R., Thyrotropin releasing hormone (TRH) modifies excitability of spinal cord dorsal horn cells, Neurosci. Lett., 92 (1988) 171-176. 9 Johnson, J.V., Fone, K.C.E, Havler, M.E., Tullock, I.E, Bennett, G.W. and Marsden, C.A., A comparison of the motor behaviors produced by the intrathecal administration of thyrotropin-releasing hormone and thyrotropin-releasing hormone analogues in the conscious rat, Neuroscience, 29 (1989) 463470. 10 Kow, L.M and Pfaff, D.W., Neuropeptides TRH and cyclo(HisPro) share neuromodulatory, but not stimulatory, action on hypothalamic neurons in vitro: implication for the regulation of feeding, Exp. Brain Res., 67 (1987) 93-99. 11 Kubek, M.J., Rea, M.A., Hodes, Z.I. and Aprison, M.H., Quantitation and characterization of thyrotropin-releasing hormone in vagal nuclei and other regions of the medulla oblongata of the rat, J. Neurochem., 40 (1983) 1307-1313. 12 Lacey, G., Nistri, A. and Rhys-Maitland, E.R., Large enhance-
Acknowledgements. This work was supported by Dir. Ib Henriksens Fond, Foundation for Experimental Research in Neurology, the Lundbeck Foundation and The Danish Medical Research council (SLF-12-8279, 12-8715).
ment of excitatory postsynaptic potentials and currents by thyrotropin-releasing hormone (TRH) in frog spinal motoneurones, Brain Res., 488 (1989) 80-89. 13 Manaker, S., Winokur, A., Rostene, W.H. and Rainbow, T,C., Autoradiographic localization of thyrotropin-releasing hormone receptors in the rat central nervous system, J. Neurosci., 5 (1985) 167-174. 14 Mantyh, P.W. and Hunt S.P., Thyrotropin-releasing hormone (TRH) receptors. Localization by light microscopic autoradiography in rat brain using [3H][3-Me-His2]TRH as the radioligand, J. Neurosci., 5 (1985) 551-561. 15 Mayer, M.L. and Westbrook, G.L. The action of N-methyl-Daspartic acid on mouse spinal neurones in culture, J. Physiol., 361 (1985) 65-90. 16 Mosfeldt Laursen, A. and Rekling, J.C., Electrophysiological properties of hypoglossal motoneurones of guinea pigs studied in vitro, Neuroscience, 30 (1989) 619-637. 17 Nicoll, R.A., Excitatory action of TRH on spinal motoneurons, Nature, 65 (1977) 242-243. 18 Nistri, A., Fisher, N.D. and Gurnell, M., Block by the neuropeptide TRH of an apparently novel K÷ conductance of rat motoneurones, Neurosci. Lett., 120 (1990) 25-30. 19 Ono, H. and Fukuda, H., Ventral root depolarization and spinal reflex augmentation by a TRH analog in rat spinal cord, Neuropharrnacology, 21 (1982) 739-744. 20 Rekling, J.C., Excitatory effects of thyrotropin-releasing hormone (TRH) in hypoglossal motoneurones, Brain Res., 510 (1990) 175-179. 21 Renaud, L.P., Blume, H.W., Pittman, Q J., Lamour, Y. and Tan, A.T., Thyrotropin-releasing hormone selectively depresses glutamate excitation of cerebral cortical neurons, Science, 205 (1979) 1275-1276. 22 Segerson, T.P., Hoefler, H., Childers, H., Wolfe, H.J., Wu, P., Jackson, I.M.D. and Lechan, R.M., Localization of thyrotropin-releasing hormone prohormone messenger ribonucleic acid in rat brain by in situ hybridization, Endocrinology, 121 (1987) 98-107. 23 Stone, T.W,, Actions of TRH and cyclo-(His-Pro) on spontaneous and evoked activity of cortical neurones, Eur. J. Pharrnacol., 92 (1983) 113-118. 24 Takahashi, T., Thyrotropin-releasing hormone mimics descending slow synaptic potentials in rat spinal motoneurons, Proc. R. Soc. Lond. B, 225 (1985) 391-398. 25 White, S.R., A comparison of the effects of serotonin, substance P and thyrotropin-releasing hormone of rat spinal motoneurons in vivo, Brain Res., 335 (1985) 63-70.
289
BRES 17651
Interaction between thyrotropin-releasing hormone (TRH) and NMDA-receptor-mediated responses in hypoglossal motoneurones Jens C. Rekling Institute of Neurophysiology, University of Copenhagen, Copenhagen (Denmark) (Accepted 10 December 1991)
Key words: Thyrotropin-releasing hormone; Hypoglossal motoneuron; Transmitter interaction; Neuromodulation, N-Methyl-o-aspartate; Excitatory amino acid; Brain slice
The effect of thyrotropin-releasing hormone (TRH) on the responses to excitatory amino acids was investigated in hypoglossal motoneurones in an in vitro preparation of the brainstem from guinea pigs using current clamp and discontinuous single electrode voltage clamp (dSEVC). Bath application of 20-50/~M TRH markedly potentiated the response to iontophoretically applied NMDA, whereas no potentiation of the response to glutamate, aspartate or quisqualic acid was seen. Voltage clamp experiments showed that TRH did not increase the current flowing through NMDA channels, thus a direct modulatory role of TRH on NMDA channels was not a likely explanation of the potentiation. Voltage clamp studies of the current-voltage relationship showed that the potentiation of the response to NMDA and lack of potentiation of the response to quisqualic acid was a result of an interaction between the actions of TRH and the amino acids on the electroresponsive profile of the membrane. Endogenous NMDA receptor activation was produced by tetanic stimulation of the reticular formation dorsolaterally to the hypoglossal nucleus, evoking large APV sensitive EPSPs in the presence of CNQX, a non-NMDA blocker. The amplitude and duration of these potentials were increased at more positive membrane potentials in response to TRH. It is concluded that TRH can act as a neuromodulator-potentiating the response to NMDA receptor activation-simply by changing the electroresponsive properties of the membrane. INTRODUCTION I m p r o v e m e n t s in immunohistochemical and in situ hybridization techniques have revealed that the T R H systems in the m a m m a l i a n brain are considerably m o r e extensive than previously assumed 7'22. The initial interest in this tripeptide was centered around its role in the hypothalamic control of secretion of thyroid-stimulating h o r m o n e (TSH). The transmitter action of T R H in extrahypothalamic tissues has, however, attracted increasing attention recently. T R H has been shown to have potent actions on the m o t o r system at several anatomical levels. Injection of T R H into the nucleus accumbens induced increased l o c o m o t o r activity and behaviors such as licking, chewing and sniffing 4. Intrathecal administration of T R H induced wet-dog shakes and forepaw-licking behavior, possibly resulting from an action at the level of the spinal cord or brainstem 9. Injection of T R H in the fourth ventricle increased respiratory rate suggesting an action on the medullary respiratory centers 5. A t the level of the spinal cord T R H has been shown to increase the excitability of spinal m o t o n e u r o n e s as m e a s u r e d by an increase in the antidromic field potentials, ventral root depolarizations and increase in the monosynaptic reflex 2'
3,19. T h e effects of T R H in the spinal cord has been explained by a direct action of T R H on spinal motoneurons 17'18'24 and facilitation of presynaptic release of excitatory transmitters 12. The anatomical distribution of extrahypothalamic T R H also involves several brainstem m o t o r nuclei, including the hypoglossal nucleus, where the native horm o n e , immunohistochemically characterized T R H - c o n taining nerve terminals and T R H binding sites have been found 6'1t'13'14. m study from this l a b o r a t o r y provided electrophysiological indications for a transmitter action of T R H in the hypoglossal nucleus, showing that T R H induced depolarization and increase in the input resistance of hypoglossal m o t o n e u r o n e s 2°. Like many of the o t h e r putative neuropeptidergic transmitters T R H is believed to have a n e u r o m o d u l a t o r y function-interacting with o t h e r transmitter systems. In cortical neurons T R H has been found to potentiate the excitatory action of acetylcholine 1 and in rat spinal motoneurones in vivo T R H enhanced both glutamate- and a s p a r t a t e - e v o k e d activity25. This study investigates the actions of T R H on the response to excitatory amino acids in hypoglossal m o t o n e u r o n e s and demonstrates how a n e u r o m o d u l a t o r y role of T R H can be explained by the
Correspondence: J.C. Rekling, Institute of Neurophysiology, Panum Institute, Blegdamsvej 3C, 2200 Copenhagen N, Denmark.
290 action of T R H itself o n t h e e l e c t r o r e s p o n s i v e profile o f the n e u r o n a l m e m b r a n e .
MATERIALS AND METHODS The experiments were performed on transversely cut brainstem slices from albino guinea pigs according to a previously described method 16. In brief, 400/~m thick slices from the brainstem were immediately transferred to the recording chamber. It had a volume of 2 ml, a temperature of 36.5°C and was constantly superfused at a rate of 2 ml/min with preheated, oxygenated Ringer's solution composed of 134 mM NaC1, 2 mM KCI, 20 mM NaHCO 3, 1.25 mM KH2PO 4, 1.5 mM CaC12, 2 mM MgC12, 9 mM glucose. The solution was bubbled with 95% 02/5% CO 2 and this gas mixture also filled the space above the medium in the recording chamber. After a period of 30 rain to 1 h the slices were transilluminated with a point source from below and the recording electrode placed in the hypoglossal nucleus under visual guidance through a stereomicroscope. Recording system. Electrodes were pulled from filamented glass tubes (outer diameter of 1.5 mm, inner diameter of 0.8 mm) and filled with a 2-3 M solution of potassium acetate. The electrodes had resistances of between 15 and 60 MfL A bridge-balance amplifier (built in the laboratory) with a cut-off frequency of 8 kHz was used in the current clamp experiments. Discontinuous single electrode voltage clamp (dSEVC) was performed using the homebuilt amplifier or an Axoclamp 2A amplifier at switching frequencies of 2.5 to 3 kHz. The recorded signals were stored on a 4-track tape recorder (Racal) with a band width of DC to 2.5 kHz. Offline analysis was done on a personal computer digitizing the signals at 5 to 82 kHz. lontophoresis. Microelectrodes with broken tips were filled with NMDA (N-methyl-D-aspartic acid; 0.05 M; Sigma), L-glutamic acid (0.5 M; Sigma), L-aspartic acid (0.225 M; Sigma) or quisqualic acid (0.05 M; Sigma) dissolved in distilled water, pH was adjusted to 8 with NaOH. The iontophoresis electrode was positioned in the slice independently of the recording electrode and a bridge-balance amplifier was used as current source (50-100 nA ejection current and 10-30 nA backing current). Extracellular stimulation. Electrical stimulation was done between two isolated, 5/~m thick, twisted platinum wires. The unisolated
A
,~...~
ends of the platinum wires, spaced 100 pm, were placed in the slice, dorso-lateral to the hypoglossal nucleus in the reticular formation. The rectangular pulses lasted 100 ~s at voltages of 25 V to 40 V. Materials. Thyrotropin-releasing hormone (Cambridge Research Biochemicals Ltd.) was applied to the superfusion solution reaching a final concentration of 20, 40 or 50/~M, TTX (tetrodotoxin; Sigma) was used in a concentration of 1/~M. CNQX (6-cyano-2,3dihydroxy-7-nitro-quinoxaline; Tocris Neuramin) was used at 50 pM, APV (DL-2-amino-5-phosphono-valeric acid; Sigma) at 100 /~M. QX-314 (A generous gift from Astra) was dissolved in 3 M potassium acetate pipette solution reaching a final concentration of 100-200 mM.
RESULTS T h e findings p r e s e n t e d h e r e are b a s e d o n stable intracellular r e c o r d i n g s f r o m 45 hypoglossal m o t o n e u r o n s . T h e i d e n t i t y o f the cells was b a s e d o n the characteristic electrophysiological
properties
previously
d e s c r i b e d 16
and p o s i t i o n i n g of the r e c o r d i n g e l e c t r o d e inside the b o u n d a r i e s o f the hypoglossal nucleus, which was visible in the u n s t a i n e d slice.
Potentiation of the response to iontophoretically applied NMDA In o r d e r to test for a n e u r o m o d u l a t o r y action o f T R H in the hypoglossal m o t o n e u r o n e s i o n t o p h o r e s i s of exci° t a t o r y a m i n o acids was p e r f o r m e d . T h e i o n t o p h o r e s i s e l e c t r o d e was p l a c e d in the slice i n d e p e n d e n t l y of the r e c o r d i n g e l e c t r o d e , but in close p r o x i m i t y to the tip o f it, j u d g e d visually. W h e n p e r f o r m i n g i o n t o p h o r e s i s c a r e was t a k e n to o b t a i n a r e g u l a r cycle o f c u r r e n t e j e c t i o n and as small a h o l d i n g c u r r e n t as possible. C u r r e n t was i n j e c t e d into the e l e c t r o d e for 4 s at fixed intervals (30
C
B
2O mV
~
NMDA
D
NMDA
E
F
"l"rx
l"rX, TRH
NMD'~
TTX, W~mh
NMD'~'A
Fig. 1. Potentiation of the response to NMDA iontophoresis by TRH. A, B and C: iontophoresis of NMDA before, during TRH (20/~M) and after 10 min of wash. D, E and F: iontophoresis of NMDA performed with TI'X (1/~M) in the superfusion solution. Membrane potential was kept constant using bias current.
291
A
.-.
B
,.--.-
Control
C Wash
TRH
~u
D
E
F
G
...-
H
I
Control
QuJs
Wash
TRH
eamrol
I
TFIH
I
~
Fig. 2. Action of TRH on iontophoresis of glutamate, aspartate and quisqualic acid. TRH did not potentiate the response to iontophoresis of glutamate (A, B, C), aspartate (D, E, F) or quisquahc acid (G, H, I) at a concentration of 50/~M. Wash (10 min). Membrane potential was kept constant using bias current.
or 40 s), since this regime resulted in the smallest change in the response to the compounds over time. The first compound tested was NMDA and as shown in Fig. 1 TRH had a striking effect on the iontophoretic NMDA response. In Fig. 1A,B,C a pulse lasting 4 s driving NMDA out of the iontophoresis pipette was applied before (Control), during T R H (20/~M; 3 min after start of superfusion) and 10 min after re-entering a normal Ringer's solution (Wash). The membrane potential was kept constant, counteracting the depolarizing action of TRH, using hyperpolarizing bias current and the extracellular iontophoresis artefact has been subtracted. During TRH, NMDA iontophoresis evoked a much longer spike train commencing earlier after the onset of the pulse. When the test was performed with TI'X (1 /~M) in the superfusion solution (Fig. 1D,E,F) the depolarizing action of NMDA was markedly potentiated during TRH. The depolarizing potential rose faster after the onset of the pulse and 18 mV positive to resting potential a 'high noise' plateau potential emerged. Eleven cells were tested with NMDA iontophoresis and all cells displayed potentiation of the response by TRH. During continued superfusion with T R H the action on the NMDA response desensitized, pursuing the same time course as the desensitization of the depolarizing action of T R H previously described 2°.
No potentiation of the response to glutamate, aspartate or quisqualic acid A similar test was done iontophoresing glutamate (5
cells), aspartate (3 cells) and quisqualic acid (4 cells) (Fig. 2). Attempts to iontophorese kainic acid were unsuccessful, because of the manifest toxic effects of this compound on the viability of the hypoglossal motoneurons. None of the neurons tested with these excitatory amino acids showed potentiation of the response when T R H (20-50 #M) was added to the superfusion solution.
Voltage dependence of the NMDA potentiation Fig. 3 shows that the size of the potentiation of the NMDA response by TRH is dependent on the membrane potential. Iontophoresis of NMDA was performed in a ~ (1/~M) containing medium and hyperpolarizing pulses of constant strength were given to monitor the apparent input resistance (2 cells). Using bias current the NMDA iontophoresis was given from three different membrane potentials (-66, -70 and -74 mV) before and during TRH (20/~M). Note that the control response to NMDA is voltage dependent, increasing at more positive potentials due to a decrease in the Mg 2+ block. During T R H the potentiation of the depolarizing action of NMDA was very pronounced at the two most depolarized holding potentials (-66 mV and -70 mV), measured as the area under the curve (80% increase at -66 mV, 113% increase at -70 mV). At -74 mV the potentiation was much smaller (22% increase). In terms of amplitude and input resistance the relative difference is not so clear probably because the amplitude of the response during TRH reaches a plateau at the two most depolarized potentials, but both parameters are substantially smaller at
292
A
",-rx
NMDA
C
TTX,TRH
NMDA
~
D
NMDA
E
Voltage clamp studies
B
NMDA
~-x
F
Tr~,
J
2OmV
NMDA
NMDA
Fig. 3. Voltage dependence of the TRH induced potentiation of the NMDA response. A, C and E: iontophoresis of NMDA performed in a TTX (1 ~M) containing solution rising from three different membrane potentials (-66 mV, -70 mV, -74 mV). Hyperpolarizing current pulses of constant strength were given to monitor input resistance. B, D and F: NMDA iontophoresis rising from the same membrane potentials as A, C, E, but 3 min into a TRH (50/~M) containing solution.
-74 mV than at -70 mV (amplitude: 58% increase vs. 133%, input resistance: 50% increase vs. 93%). The initial conclusion from these experiments is that T R H in a voltage-dependent manner potentiates the response to N M D A , but not the response to glutamate, aspartate or quisqualic acid. This raises two questions. Does T R H have a direct modulatory action on the current carried through N M D A channels? Or, does the change in the electroresponsive properties of the membrane induced by T R H somehow cause the response to N M D A to be augmented, but not the response to the other amino acid? To address these questions a series of voltage clamp experiments were performed.
A
co~
To investigate the possibility that the potentiation of the N M D A response was the result of a direct modulation of the current carried through N M D A channels, d S E V C (see Materials and Methods) of the iontophoretic response was performed (2 cells). In Fig. 4A a hypoglossal motoneurone in a TTX containing solution was voltage clamped at -63 mV, a membrane potential at which T R H induced a marked potentiation of the N M D A response in current clamp. Iontophoresis of N M D A (4 s pulse) induced a transient inward current. Three minutes after entering a T R H (50/~M) containing solution the test was repeated (Fig. 4B). As seen in Fig. 4B T R H induces a steady inward current (0.4 nA, dotted line), but the amplitude of the N M D A response is unaltered (-0.8 n A vs. - 0 . 8 nA). Thus T R H does not increase the current flowing through N M D A channels. In order to construct I - V curves before and during superfusion with T R H , N M D A or quisqualic acid a command voltage having the shape of a rising ramp was applied in d S E V C (10 cells). The rate of the ramp was relatively slow (6-11 mV/s) allowing transient currents to pass to steady state. The recordings shown in Fig. 5 were lowpass filtered at 0.3 kHz and smoothed digitally (running average of 20 ms windows). Fig. 5A (trace a) shows an I - V curve obtained from a neurone in a T I ' X (1/~M) containing solution. Three minutes after entering a solution containing 50 /~M T R H the I - V curve shifted downward due to the development of an inward current (trace b). The configuration of the curve was changed with a decrease in membrane slope conductance developing positive to -50 mV. These observations are in good agreement with the depolarization and increase in input resistance seen in current clamp in response to T R H 2°. A n intersection between the two curves at hyperpolarized potentials was not seen, probably because the current passing characteristics of the electrode did not permit c o m m a n d voltages negative to approximately -80 mV without losing control of the voltage. The interaction between N M D A and T R H is illustrated in Fig. 5B, showing a set of I - V curves obtained
B
TRH 20 mV
V
NMDA
NMDA
4s
Fig. 4. Voltage clamp of the iontophoretic NMDA response. A: NMDA iontophoresis performed in dSEVC mode inducing a transient inward current. TI'X (1/~M) was added to the medium. B: NMDA iontophoresis 3 min after entering a TRH (50/~M) containing solution. Note that TRH induces a steady inward current, but the response to NMDA is unchanged.
293 A
/
+.s
"~. . . . . . . i . . . . . . . . . . . . . . . .
~
-
~
"" ......
b. . . . . . .
45 -80
B
-60
..411
4
-20
~ a ~ . - - - ~ . . . . . - ....
2 ....... . - . . . . . . ~
-6 ...............................i..................................... I
-90
C
-75
-60
-45
-30
2
go 0
-3
-4 -90
:
:
-75
-60
i
-45
-30
Membrane potential (mV)
Fig. 5. I-V curves based on dSEVC during bath application of TRH, NMDA or quisqualic acid. A: plot of the current-voltage relationship in a solution containing TI'X (1 /~M; trace a) and 3 min after adding TRH (50 ~M; trace b). B: I-V curves in TI"X (1 gM; trace a), 10 min into superfusion with NMDA (100 gM; trace b) and 3 min after adding TRH (50 ~M; trace c). C: I-V curves in T r x (1 #M; trace a), 10 min into superfusion with quisqualic acid (150/~M; trace b) and 3 rain after adding TRH (50 #M; trace c).
in q-"l'X (1/~M; trace a), N M D A (100/~M; trace b) and N M D A + T R H (50/~M; trace c). A p p l i c a t i o n of N M D A to the supeffusion solution resulted in the d e v e l o p m e n t
A
B Normal ringer
E
C CNQX
F
/~TRH
of a p r o m i n e n t v o l t a g e - d e p e n d e n t inward current making the I - V curve highly non-linear. A s in the case of bath applied T R H alone, the m e m b r a n e slope conductance was decreased positive to a p p r o x i m a t e l y - 4 8 m V in response to N M D A . A p p l i c a t i o n of T R H (50/~M) in addition to N M D A in the solution resulted in developm e n t of an additional steady inward current, displacing the I - V curve in a parallel m a n n e r negative to - 5 0 mV. Positive to - 5 0 m V T R H induced a further decrease in the m e m b r a n e slope conductance, now resulting in a negative slope conductance, which causes flow of inward current throughout the c o m m a n d voltage range. A relative increase in inward current is thus induced by T R H acting together with N M D A at potentials positive to approximately - 5 0 mV. In Fig. 5C a set of I - V curves o b t a i n e d in T T X (1/~M; trace a), quisqualic acid (150 g M ; trace b) and quisqualic a c i d + T R H (50/~M; trace c) shows that activation of the A M P A type of excitatory amino acid r e c e p t o r does not lead to the same v o l t a g e - d e p e n d e n t action on the I - V relationship as N M D A r e c e p t o r activation. Quisqualic acid induced an inward current (trace b), but as o p p o s e d to N M D A the m e m b r a n e slope conductance was increased in the most depolarized region (positive to - 5 2 mV). W h e n T R H (50/~M) was a d d e d to the solution (trace c) additional inward current was induced, but the m e m b r a n e slope conductance of the cell now shifted back to the control value throughout the voltage range. Thus T R H acting together with quisqualic acid actually induced a relative change in the current flowing at m o r e positive m e m b r a n e potentials, but importantly this effect was a decrease in outward current (Fig. 5C; trace c). These voltage clamp experiments strongly point to the hypothesis that the potentiation of the N M D A response
D CNQX
G
/~TRH
CNOX, APV
10mV
H
f~/TRH
TRH
10mV
Control
Fig. 6. Effect of TRH on NMDA receptor mediated EPSPs. A, B, C, D: pharmacological isolation of a NMDA receptor mediated EPSP. A: EPSP elicited by a single stimuli, performed with QX-314 in the recording electrode. B: blockade of the EPSP by 50 gM CNQX, leaving behind only a very small prolonged depolarization. C: tetanic stimulation of the same cell as in A, B at 333 Hz lasting 40 ms, giving rise to a depolarizing potential. D: blockade of the depolarizing potential by APV (100/~M). All records are average of 8 to 10 sweeps. E, F, G and H: tetanic stimuli evoking NMDA receptor mediated EPSPs, elicited from four different membrane potentials, before and during TRH (50 /~M). Note the relative increase in amplitude at the two most positive potentials. All records are average of 8 to 10 sweeps.
294 and lack of potentiation of the other amino acids seen in current clamp should be understood in terms of the effects these compounds and T R H have on the I - V relationship.
negative potentials (-70 and -75 mV), however, were not potentiated but displayed a minor decrease in amplitude and a small prolongation of the falling phase. DISCUSSION
Voltage-dependent increase in the amplitude of NMDA receptor mediated EPSPs The potentiation of the iontophoretic NMDA response seen in current clamp could be an experimental curiosity with no apparent 'meaning' in terms of the synaptic integrative functioning of the motoneurons. Fig. 6, however, shows that TRH probably plays an important role in regulating the response to endogenous NMDA receptor activation. These experiments were based on pharmacological distinction between NMDA receptor and non-NMDA receptor activation using selective antagonists. Motoneurones were penetrated by electrodes containing QX-314 (100-200 raM) dissolved in 2 M potassium acetate, blocking sodium spikes, without affecting synaptic activation. Single electrical stimuli with a bipolar electrode placed dorsolaterally to the hypoglossal nucleus in the reticular formation evoked an EPSP (Fig. 6A). When the non-NMDA blocker CNQX (50/~M) was added to the superfusion solution the stimulus evoked EPSP disappeared, leaving behind a very small prolonged depolarization (Fig. 6B). A train of stimuli at 333 Hz lasting 40 ms was then applied. This induced a depolarizing potential with a longer rise- and decay time than an EPSP elicited by a single stimuli (Fig. 6C). As shown in Fig. 6D the depolarizing potential was blocked by APV (100/~M), a selective NMDA blocker (3 cells). Thus it can be concluded that the tetanic stimulation in the presence of CNQX elicits an EPSP caused by activation of NMDA receptors. A small depolarization, however, persisted after application of APV (Fig. 6D). This was probably a rebound depolarization, caused by a small summating train of IPSPs uncovered by the blockade of excitatory synaptic transmission (CNQX, APV), since the depolarization could be mimicked by a hyperpolarizing pulse of 40 ms driving the membrane potential to approximately the same hyperpolarized level. The action of T R H (50/~M) on the NMDA receptor mediated EPSP was investigated in 4 other cells one of which is shown in Fig. 6E,F,G,H. The average of 10 sweeps of EPSPs elicited from four different membrane potentials (-60, -65, -70 and -75 mV) is shown with the control record and the T R H record superimposed. When TRH (50/~M) was applied to the superfusion solution, the EPSPs starting from the two most positive membrane potentials (-60 and -65 mV) were greatly augmented in amplitude and duration (27% and 30% increase in amplitude). EPSPs starting from the two most
The purpose of this study was to test for a possible neuromodulatory role of T R H in the hypoglossal nucleus, inspired by several extracellular studies showing facilitatory or inhibitory effects of TRH on the response to other transmitters. The principal finding of the study was that T R H potentiated the response to NMDA receptor activation. It is suggested that the potentiation can be explained by an interaction between the effects TRH and NMDA have on the electroresponsive properties of the membrane and not a direct modulation of the current carried through NMDA channels. The modulatory actions of TRH previously reported mainly describes facilitatory or inhibitory effects of T R H on the response to excitatory amino acids or acetylcholine. TRH was found to enhance both glutamate- and aspartate-induced excitation of spinal motoneurons 25 and facilitate glutamate-induced firing in spinal cord dorsal cells8. In cortical neurons T R H potentiated the excitatory action of acetylcholine 1, but depressed the glutamate excitation 21. Another study, however, could not replicate the depressant action of T R H on amino acidinduced excitation 23, to the contrary the study showed an enhancement of the response to quisqualic acid and no effect on NMDA induced excitations. In hypothalamic neurons T R H was found to modulate the response to glutamate but not the response to acetylcholine ~°. All of the above-mentioned studies were performed using extracellular recordings and it was therefore of interest to see whether TRH had similar neuromodulatory actions in the hypoglossal motoneurones using intracellular recordings, which potentially can give a more precise characterization of the action of the different compounds on the neuronal membrane. Application of TRH on hypoglossal motoneurones resulted in a marked potentiation of the response to NMDA iontophoresis. TRH was previously shown to increase the input resistance of the hypoglossal motoneutons 2° and this in itself would tend to increase the response to excitatory input, thus providing a satisfactory explanation to the NMDA potentiation. It was therefore a puzzling finding that T R H did not affect the response to glutamate, aspartate or quisqualic acid. This selectivity then could be the result of a direct modulatory action of T R H on NMDA channels, but this is not consistent with the fact that glutamate as well as aspartate have NMDA receptor agonist properties. The second possibility was that the change in the electroresponsive
295 properties of the membrane induced by TRH somehow caused the response to N M D A to be augmented, but not the response to the other amino acids. The major electrophysiological difference between iontophoretic NMDA responses and responses to the other excitatory amino acids is the prominent voltage dependence of the NMDA response, brought about by the Mg 2÷ block 15. In a series of current clamp experiments not shown here this phenomenon was also noted in the hypoglossal motoneurons, reflected in the change in the input resistance during iontophoresis of the different amino acids. During iontophoresis of glutamate, aspartate and quisqualic acid the input resistance of hypoglossal motoneurones was decreased, in contrast to the increase during iontophoresis of NMDA. A further indication of an interaction between NMDA and T R H based on the effect these compounds have on the electroresponsive profile of the membrane was the observation that the facilitatory action of TRH on the NMDA response was sensitive to the holding potential from which the iontophoresis was given. Iontophoresis of NMDA from a relative negative membrane potential resulted in a much smaller potentiation than iontophoresis from more positive potentials. In order to determine whether TRH interacted with the currents induced by the amino acids a series of voltage clamp experiments were performed. In voltage clamp the inward current induced by NMDA iontophoresis was not affected by TRH, thus a direct modulatory effect on the NMDA channels was not a likely explanation of the potentiation seen in current clamp. A molecular interaction, however, depending on an unconstrained change in membrane potential in response to NMDA and T R H cannot be ruled out. Current-voltage relationship experiments showed that both NMDA and T R H applied to the bath alone induced an inward current with a decrease in membrane slope conductance positive to approximately -50 mV and when applied together the flow of current was inward throughout the command voltage range, with a negative membrane slope conductance at potentials positive to approximate -50 mV. This combined action on the membrane then explains the potentiation of the iontophoretic NMDA response seen in current clamp. During NMDA iontophoresis the membrane depolarized due to the inward current flowing through NMDA channels. When TRH was added to the bathing solution the iontophoresis was performed from the same membrane potential giving hyperpolarizing bias current to counterbalance the depolarizing effect of TRH. Because of the combined action of TRH and NMDA, the membrane slope conductance positive to -50 mV was now negative. This means that during NMDA iontophoresis, with TRH in the bath, an increased amount of inward current flowed at potentials positive to -50 mV,
potentiating the response to NMDA. The development of a negative slope conductance was probably the cause for generation of a plateau potential when the iontophoresis was done with TI'X in the medium (Fig. 1E), which will further augment the potentiation. Bath application of quisqualic acid in voltage clamp led to the development of a inward current with an increase in membrane slope conductance at more positive potentials. Importantly, however, the overall current flowing through the membrane at the more positive command potentials was outward. Adding TRH resulted in development of an additional inward current and as expected reduced the membrane slope conductance in the depolarized region of the I - V plot. The opposing action by T R H on the membrane slope conductance was, however, not great enough to change the overall current from outward to inward at the more positive potentials. Thus the following picture of the action of TRH on quisqualic acid iontophoresis in current clamp emerges. During quisqualic acid the membrane depolarized due to the inward current flowing through A M P A channels. Adding TRH to the bath induced a depolarization that was counterbalanced by bias current. The depolarizing action of quisqualic acid, then, was not augmented because the membrane potential never reached the point where TRH had changed the membrane slope conductance, since the overall current flowing through the membrane in this region was outward. This mechanism probably also holds true for glutamate and aspartate since these compounds like quisqualic acid produces a decrease in input resistance during iontophoresis in current clamp. This mechanism of action would also explain some of the conflicting findings of modulatory effects of T R H in other neuronal populations. A possible error in the conclusions drawn from extracellular studies of the modulatory effects of TRH seen in this context is the claim, that the modulation of a transmitter response by T R H cannot be due to a non-specific change in neuronal excitability induced by TRH, because the response to one transmitter is facilitated, but the response to another is not. This is incorrect, because the extracellular studies do not disclose whether one of the compounds have a voltage-dependent action of its own. The experiments on the pharmacologically isolated NMDA receptor mediated EPSPs show that the mechanism outlined above might have an important physiological function with respect to the integration of synaptic potentials in the hypoglossal motoneurons. When T R H had acted on the neurons, NMDA receptor mediated EPSPs were increased in amplitude and duration at more positive potentials. The increase in input resistance of the neurons induced by TRH should, however, also be taken into consideration in this phenomenon. An in-
296 crease in the input resistance of the cells will make the
with the combined effect on the m e m b r a n e slope con-
cells electrotonically more 'compact', facilitating the space clamp efficacy of the intracellular electrode. Hold-
ductance outlined above will never the less cause T R H to acts as a n e u r o m o d u l a t o r , boosting synaptic input card e d by the endogenous N M D A receptor ligand.
ing the m e m b r a n e potential at a value positive to the resting m e m b r a n e potential would then result in a dendritic m e m b r a n e potential closer to the somatic holding potential, when T R H is present. As a result of the voltage dependency of the N M D A receptor mediated EPSPs this would then m e a n a larger amplitude 'seen' from the somatic recording. This mechanism together
REFERENCES 1 Braitman, D.J., Auker, C.R. and Carpentar, D.O., Thyrotropin-releasing hormone has multiple actions in cortex, Brain Res., 194 (1980) 244-248. 2 Clarke, K.A. and Stirk, G., Motoneurone excitability after administration of a thyrotropin releasing hormone analogue, Br. J. Pharmacol., 80 (1983) 561-565. 3 Cooper, B.R. and Boyer, C.E., Stimulant action of thyrotropin releasing hormone on cat spinal cord, Neuropharmacology, 17 (1978) 153-156. 4 Heal, D.J. and Green, A.R., Administration of thyrotropin releasing hormone (TRH) to rats releases dopamine in n. accumbens but not in n. caudatus, Neuropharmacology, 18 (1979) 2331. 5 Homma, I., Oouchi, M. and Ichikawa, S., Facilitation of inspiration by intracerebroventricular injection of thyrotropin-releasing hormone in rabbits, Neurosci. Len., 44 (1984) 265-269. 6 Hrkfelt, T., Fuxe, K., Johansson, O,, Jeffcoate, S. and White, N., Distribution of thyrotropin-releasing hormone (TRH) in the central nervous system as revealed with immunohistochemistry, Eur. J. Pharmacol., 344 (1975) 389-392. 7 H6kfelt, T., Tsuruo, Y., Ulfhake, B., Cullheim, S., Arvidsson, U., Foster, G.A., Schultzberg, M., Schalling, M., Arborelius, L., Freedman, J., Post, C. and Visser, T., Distribution of TRHlike immunoreactivity with special reference to coexistence with other neuroactive compounds, Ann. N Y Acad. Sci., 553 (1989) 76-105. 8 Jackson, D.A. and White, S.R., Thyrotropin releasing hormone (TRH) modifies excitability of spinal cord dorsal horn cells, Neurosci. Lett., 92 (1988) 171-176. 9 Johnson, J.V., Fone, K.C.E, Havler, M.E., Tullock, I.E, Bennett, G.W. and Marsden, C.A., A comparison of the motor behaviors produced by the intrathecal administration of thyrotropin-releasing hormone and thyrotropin-releasing hormone analogues in the conscious rat, Neuroscience, 29 (1989) 463470. 10 Kow, L.M and Pfaff, D.W., Neuropeptides TRH and cyclo(HisPro) share neuromodulatory, but not stimulatory, action on hypothalamic neurons in vitro: implication for the regulation of feeding, Exp. Brain Res., 67 (1987) 93-99. 11 Kubek, M.J., Rea, M.A., Hodes, Z.I. and Aprison, M.H., Quantitation and characterization of thyrotropin-releasing hormone in vagal nuclei and other regions of the medulla oblongata of the rat, J. Neurochem., 40 (1983) 1307-1313. 12 Lacey, G., Nistri, A. and Rhys-Maitland, E.R., Large enhance-
Acknowledgements. This work was supported by Dir. Ib Henriksens Fond, Foundation for Experimental Research in Neurology, the Lundbeck Foundation and The Danish Medical Research council (SLF-12-8279, 12-8715).
ment of excitatory postsynaptic potentials and currents by thyrotropin-releasing hormone (TRH) in frog spinal motoneurones, Brain Res., 488 (1989) 80-89. 13 Manaker, S., Winokur, A., Rostene, W.H. and Rainbow, T,C., Autoradiographic localization of thyrotropin-releasing hormone receptors in the rat central nervous system, J. Neurosci., 5 (1985) 167-174. 14 Mantyh, P.W. and Hunt S.P., Thyrotropin-releasing hormone (TRH) receptors. Localization by light microscopic autoradiography in rat brain using [3H][3-Me-His2]TRH as the radioligand, J. Neurosci., 5 (1985) 551-561. 15 Mayer, M.L. and Westbrook, G.L. The action of N-methyl-Daspartic acid on mouse spinal neurones in culture, J. Physiol., 361 (1985) 65-90. 16 Mosfeldt Laursen, A. and Rekling, J.C., Electrophysiological properties of hypoglossal motoneurones of guinea pigs studied in vitro, Neuroscience, 30 (1989) 619-637. 17 Nicoll, R.A., Excitatory action of TRH on spinal motoneurons, Nature, 65 (1977) 242-243. 18 Nistri, A., Fisher, N.D. and Gurnell, M., Block by the neuropeptide TRH of an apparently novel K÷ conductance of rat motoneurones, Neurosci. Lett., 120 (1990) 25-30. 19 Ono, H. and Fukuda, H., Ventral root depolarization and spinal reflex augmentation by a TRH analog in rat spinal cord, Neuropharrnacology, 21 (1982) 739-744. 20 Rekling, J.C., Excitatory effects of thyrotropin-releasing hormone (TRH) in hypoglossal motoneurones, Brain Res., 510 (1990) 175-179. 21 Renaud, L.P., Blume, H.W., Pittman, Q J., Lamour, Y. and Tan, A.T., Thyrotropin-releasing hormone selectively depresses glutamate excitation of cerebral cortical neurons, Science, 205 (1979) 1275-1276. 22 Segerson, T.P., Hoefler, H., Childers, H., Wolfe, H.J., Wu, P., Jackson, I.M.D. and Lechan, R.M., Localization of thyrotropin-releasing hormone prohormone messenger ribonucleic acid in rat brain by in situ hybridization, Endocrinology, 121 (1987) 98-107. 23 Stone, T.W,, Actions of TRH and cyclo-(His-Pro) on spontaneous and evoked activity of cortical neurones, Eur. J. Pharrnacol., 92 (1983) 113-118. 24 Takahashi, T., Thyrotropin-releasing hormone mimics descending slow synaptic potentials in rat spinal motoneurons, Proc. R. Soc. Lond. B, 225 (1985) 391-398. 25 White, S.R., A comparison of the effects of serotonin, substance P and thyrotropin-releasing hormone of rat spinal motoneurons in vivo, Brain Res., 335 (1985) 63-70.
