European Jalurna! of Pharmacology,
195 (1991)
241-244
241
0 1991 Elsevier Science Publishers B.V. 0014.2999/91/$03.50 ADONIS
0014299991002442
WP 51787
Potentiation of synaptic reflexes by II-serine in the rat spinal cord in vitro Richard DUPHA
J. Siarey, Stephen
R B. V.. P.O. Box 900. 1380 DA
K. Long and Richard
Weesp, The Netherlands
and
’
Department
H. Evans 1
of Pharmacolom,
School of Medical
Sciences,
Bristol BS8 1 TD, U.K.
Received 25 October 1990. revibed MS received 28 December 1990. accepted 8 January 1991
6-Cyano-7-nitroquinoxaline-2,3-dione (CNQX) (IO PM) depressed dorsal root-evoked ventral and dorsal root potentials of the in vitro immature rat spinal cord to 26.3 f 5.2 S.E.M. and 40.8 f 2.7% of control values respectively. These depressant effects of CNQX were partially reversed by D-serine (EC,, values 39.7 PM f 8.7 S.E.M. N = 6 and 34.9 + 12.5 PM, N = 5 for ventral root potential and dorsal root potential respectively). Under our experimental conditions, which included the presence of Mg2+ (0.75 mM) in the bathing medium, no measurable potentiation of these synaptic reflexes by D-serine was recorded in the absence of
CNQX. These data indicate that CNQX. in addition to its depressant effect at non-NMDA receptors, depresses an NMDA receptor-mediated component of segmental transmission through its action at the glycine site of the NMDA receptor complex. D-Serine; Spinal cord; CNQX (6.cyano-7-nitroquinoxaline-2.3.dione);
1. introduction The NMDA receptor/channel is proving to be a very complex structure with many ligand binding sites. One of these is the strychnine-insensitive glycine binding site at which D-serine is an agonist (Johnson and Ascher, 1987; reviewed by Thomson, 1990). 6-Cyano-7nitroquinoxaline-2,3-dione (CNQX) has been recently introduced as a potent and selective antagonist of nonNMDA receptors (Honor6 et al., 1988). Subsequent studies have shown that CNQX can antagonise NMDA-induced responses and thai this antagonism can be reversed by D-serine. Thus the relatively weak antagonism at NMDA receptors produced by CNQX is probably due to an action at the strychnine-insensitive glycine site. This interaction is suggested from binding site displacement studies (Kessler et al., 1989; Lester et al., 1989), NMDA-evoked release studies (Harris and Miller, 1989) and from electrophysiological responses to NMDA of in vitro spinal cord preparations (Birch et al., 1988; Long et al., 1990) and isolated neurones (Lester et al., 1989). However, there are relatively few studies which show operation of the glycine site during synaptic transmission. In the absence of applied antagonists glycine (Thomson et al., 1989) and D-serine (Salt, 1989) have been reported to potentiate NMDA receptor-mediated excitatory synaptic responses in
Correspondence to: R.H. Evans, Department of Pharmacology. School of Medical Sciences, Bristol BS8 1TD. U.K.
(Electrophysiology)
cerebrocortical slice preparations and in the thalamus in vivo respectively. In the presence of CNQX D-serine has been shown to reverse the synaptic depressant action of a glycine site antagonist in cerebrocortical (Fletcher et al., 1989) and hippocampal (Bashir et al., 1990; Gaiarsa et al., 1990) slice preparations. The present paper describes an investigation of this action of D-serine on synaptic transmission in the neonate rat spinal cord.
2. Materials and methods Two to five day old rats were anaesthetised with urethane (1.8 g per kg i.p.). Following decapitation spinal cords were removed, hemisected and mounted for electrical stimulation of dorsal roots and recording from an adjacent dorsal root (dorsal root potential) and/or corresponding ventral root (ventral root potential) as previously d,escribed (Long et al., 1990). Synaptic potentials were recorded with an SMR laboratory interface connected to an IBM microcomputer and controlled by a software package developed in house (J.M. Van der Zalm, Duphar BV R & D Automation Department). The programme was used to measure amplitudes and integrate areas (over 1.65 s) of synaptic potentials. Mean values &S.E. (from N preparations) are presented in Results. The significance of differences between mean values was assessed from a rank order test (Wilcoxon-Mann-Whitney).
3 Resurts
CONTROL
A
prqarations were superfused with continuously 8% ~ium at a rate of 1 ml per min maintains at S”C and gassed with a mixture of 95%-S% of O&t& to pH 7.4. The bathing medium had the following ~~rn~~on (mM): NaCl 118. NaHCC& 24, KC! 3, CaCl z 1.5. MgsO, 0.75, dextrose 12. All drugs were diluted in the bathing medium before superfusion on to the preparation. CNQX and ( + )2-amino-Sphosphono pentanoate (AP5) were supplied by Tocris Chemicals.
n
DR-*RL./ L J-L_._ L B
A&J
I
In the present seven preparations the mean latency and the mean time to the initial peak (monosynaptic component) of the ventral root potential (8.9 + 0.6 and It.8 k 0.6 ms respectively) corresponded with the initial report on this preparation (Otsuka and Konishi, 1974). Componding values for the dorsal root potential in the present experiments were 17.4 & 0.8 and 110 Z!Z 6 ms. The latter time to peak is similar to that reported by Preston and Wallis (1980) in the absence of added Mg’+. The dorsal root reflex was not prominent in the present experiments probably due to a combination of low stimulus intensity (Preston and Wallis. 1980) and presence of 0.75 mM Mg” (Preston and Wallis, 1980). The mean Peak amplitude, area and time to half decay of the ventral root potential were 3.06 f 0.62 mV, 805 + 228 @J - s and 11.I zt:0.5 ms (N = 7). Corresponding values for the dorsal root potential were 1.37 + 0.5 mV, 599 +_241 PV - s and 42.5 + 2.7 ms (N = 6). As expected from previous experiments with mature spinal preparations (Long et al., 1990) CNQX (10 PM) reversibly depressed both the ventral root potential and dorsal root potential (fig. 1A). The amplitude of the initial monos~ap~c component of the ventral root potential was depressed from control (100%) values to 6.7 + 1.7% (N = 7) and the maximum amplitude of the dorsal root potential was depressed from control values to 44.9 + 3.8%. Areas of ventral root potential and dorsal root potential, which reveal polysynaptic components, were depressed from control values to 26.3 + 5.2 and 40.8 + 2.7% (N = 6) respectively. The difference in CNQX-sensitivity between the two synaptic responses was highly significant (P < 0.002). The selective antagonist AP5 was used to show the involvement of NMDA receptors in generation of both synaptic potentials. As can be seen in fig. 1B a brief (2 min) application of AP5 (25 PM), in the presence of CNQX, produced a further depression of both potentials. As well as depressing the ~p~tudes the CNQX produced a significant increase in latency of the synaptic potentials. Thus in the presence of CNQX (10 PM) latencies of ventral and dorsal root potentials were increased to 14.3 -t 1.3 and 25.2 + 33 ms respectively (N = 7). Tilese values were not significantly altered fol-
CNQX IO
D-SERINE200
25
I
l-----i
I
Fig. 1. (A) Depression of dorsal root-evoked dorsal (DR-DRP) and ventral (DR-VRP) root potentials by CNQX (10 FM). The left band traces are controls 5 min before intr~ucti~ of CNQX and the right hand traces were recorded 20 min Wowing introductZo%of CEQX. The upper and lower traces are synchronous and show depression of DR-VRP and DR-DRP respectively. Vertical calibration I mV, horizontal 10 ms (DR-VRP) and 200 ms (DR-DRP). (B) Same preparation as (A) following intr~uction of CNQX. APS and D-serine were introduced as indicated above the traces. AP5 was present for 2 min and D-serine for 5 min. APS depressed and D-serine potentiated both potentials. Recovery is shown 20 min following introduction of AP5and D-se&e-free medium. Vertical calibration 0.5 mV (upper) 1 mV (lower), horizontal 20 ms (upper) and 500 ms (lower).
lowing introduction of APS (25 FM) or D-serine (200 FM). The potentiating effect of D-serine is illustrated in fig. 1B following recovery from the depressant effect of APS. These effects on several preparations are summarised in table 1 where it can be seen that in the presence of D-serine potentials were enhanced above the control levels in the presence of CNQX alone. The effects of AP5 and D-serine on amplitude of the ventral root potential shown in the first column of table 1 are on the longer latency peak and not the monosynaptic component since the latter peak was absent in the presence of CNQX. The potentiating effect of D-serine (200 PM) was significant at the P < 0.014 level. In order to determine EC, values D-serine was added to the superfusate in cumulative concentrations as illustrated in fig. 2A. EC,, values for the potentiating effect of D-serine, determined from dose-response plots as illustrated in fig. 2B, were 39.7 f 8.7 CM and 34.9 f 12.5 FM on the ventral root potential and dorsal root potential respec-
243
4. Discussion
TABLE 1 Effect of APS (25 FM) and R-serine (200 CM) on amplitude and areas of dorsal root-evoked ventral and dorsal root potentials in the presence of CNQX (10 PM). N = 7 for the ventral and 6 for the dorsal root potential. The synaptic responses were depressed by AP5 and potentiated by D-serine. Percentages are of control areas before introduction of CNQX. Mean absolute control values are given in the text. The drugs were applied as illustrated in fig. 2. Treatment
Ventral
CNQX CNQX + AP5
Dorsal
Peak
Area
Peak
Area
6.7kl.7 4.7 f 0.8
26.3k5.2 6.8kO.9 47.7+7.3
44.9k3.8 l&4&4.6 54.7+5.4
40.8*2.7 13.2f3.9 54.4&4.6
CNQX+D-serine 8.Ok2.2
tively. These values are not significantly different (P = 0.33). In the absence of CNQX no significant potentiating effect of D-serine was observed on three preparations when applied at ~ncentrations up to 1 mM.
240
300
360
420
480
540
so -
60-
40 -
x-
/ i
d
oL 10
’
“‘11 too
c
Fig. 2. (A) Plot of areas (pV+s) of dorsal root-evoked ventral root potential against time (mm). Introduction of drugs at micromolar concentration is indicated by the bars above the plot. The plot of percentage potentia~ion against concentration of D-serine (PM) shown in (B) was produced from the cumulative effect of D-set-me shown in (A).
The present study shows that CNQX is a Potent depressant of synaptic potentials in the neonate rat spinal cord in agreement with previous obserations on the ~P~b~~ (Fletcher et al., 1988) and the adult rat (Long et al., 1990) spinal cord. The depressant effects of both CNQX and APS on the dorsal root potential (fig. 1 and table 1) shows that a combination of excitatory amino acid receptors are involved in this synaptic potential of neonate preparations as has been previously shown to occur in preparations from adult rats (Evans and Long, 1989). In the present experiments CNQX increased the latencies of the two synaptic potenti&. This CNQX-induced increase in latency was not reversed by D-set-me which is consistent with the role of non-NMDA receptors in mediation of the initial cornponents of these spinal reflexes (Evans and Long, 1989). A significant difference between neonate (present results) and adult (Long et al., 1990) reflexes appears to be the presence of a relatively short latency APS-sensitive component in the ventral root potential of the neonate, in the presence of Mg*+, which was revealed following application of CNQX (fig. 1). This may be a general feature of immature synaptic circuitry because a similar difference has been observed in the hippocampus (Gaiarsa et al., 1990). Despite the presence of a significant cont~bution of NMDA receptors to the population excitatory post-synaptic potential no potentiating effect of D-set-me was seen in the absence of CNQX. It is possible that the difference between spinal and cerebral (Thomson et al., 1989) preparations in the lat:er respect could be explained by heterogeneity of the NMDA-linked glycine receptor between spinal cord and forebrain (Danysz et al., 1990). However, in the hip~~pal slice preparation the effect of D-serine was similar to the present report in that no potentiation of synaptic responses was observed in the absence of a glycine site antagonist (Bashir et al., 1990). The reversal by D-serine of the synaptic depressant effects of CNQX observed in the present study could be apparent rather than a reflection of direct competition between CNQX and D-serine. However, the concentrations of D-serine required to reverse the synaptic depressant effect of CNQX in the present study were similar to those previously reported to reverse the antagonism of NMDA-induced depolarizations in these (Long et al., 1990) and cerebrocor?ical (Fletcher et al., 1989) preparations. These concentrations of D-serine or glycine are at least an order of mag~tude higher than concentrations which are effective on NMDA activated ionic currents of isolated cells or on displacement of radioligand from washed isolated membranes (Lester et af., 1989). This difference in effective con~ntrations has been attributed to the presence of a SatUKttiUg TeVeTof
~~te~t~t~~~glycine in whole true preparations (Fletcher et a&.. ~~~~~. The isolated ceil and men~brane studies qsrc:Thomson. 1990) show the dissociaGon constants For &cinr or D-se&c at the tglycine site ( < 0.3 pM) to be ccmsistcnt with more than 95% saturation of the site al a~~ro~~rnate~y6 $4 agonist. In the present e~per~rnents it can be assumed that intrinsic glycine was close to a saturating concentra\ion because NMDA recceptor~~~iat~ excitation was not enhancedin the absence of sntsgonist.CFXjX (10 pM) would produce a dose-ratio of approsimatcly six (calculated from the K, of 2.2 PM reported by Kesskr et al.. 1989) at the glycine site. Thus the present EC,, of 30-40 PM for reversal of the depressant effect of CNQX by D-serine would be consistent with restoration of a saturating level of glycine site agonist by a 6-fold increase in concentration. The uptake of exogenous D-serine or glycine could also be involved in reducing their apparent potency on whole tkssue preparations. Thus the present data suggest that CNQX, in addition to a depressant effect at non-NMDA receptors. depresses an NMDA receptor-mediated component of segmental transmission through an action at the strychnine-insensitive glycine site of the NMDA receptor compfex.
R.J.S. is a DUPHAR research scholar and R.H.E. was supportrd Research Council. The Taberner Trust and The Wellc~me Trust.
h?: the Mrdkd
References Bashir. 2.1.. 3. Tam and G.L. Colling~dge. 199% Activation of the glycins site in the NMDA receptor is necessary for the induction of LTP. Neurosci. Lett. 108, 261. Birch. P.J.. CJ. Grossman and A.G. Hayes, 1988, 6,7-Dinitroquinoxalin~~3-dione and 6-nitr~7~yan~quinoxaline-2,3dione antagonise responses to NMDA in the rat spinal cord via an action at the strychnine-insensitive glycine receptor, European J. Pharmacol. 156, 177. Danysz. W.. E. Fadda. J.T. Wrohlewski and E. Costa. 1990. (‘HI-DSerine Iabets st~chnine-insensitive glycine recognition sites of rat central nervous system. Life Sci. 46. 155.
Fvanb. R.H. and S.K. Long, 1989. primary afferent depolarization in the mt spinal cord is mediated by pathways utilising NMDA and n.m-NMDA receptors. Neurosci.Lett. 100. 231. Ftetrner. E.J.. D. Martin. J.A. Aram. D. Lodge and T. Honor& 1988. Quinoxalinediones selectively block quisqualate and kainate receptors and synaptic events in rat ncocortex and hippocampus and frog spinal cord in vitro, bi. J. Ph~rmacol. 95, 585. Fletcher. E.J.. J.D. Milfar, S. Zeman and D.Lodge, 1989. Non-competitive antagonism of N-methyl-D-aspartate by displacement of an endogenous glycine-like substance, European J. Neurosci. 1,196. Gaiarsa. J.L.. R. Corradetti. E. Cherubini and Y. Ben Ari. 1990. The allosteric glycine site of the N-methyl-D-aspartate receptor modulates GABAergic-mediated synaptic events in neonatal rat CA3 hippocampal neurons, Proc. Natl. Acad. Sci. U.S.A. 87, 343. Harris. K.M. and R.J. Mifler, 1989. CNQX (6-cyano-%nitroquinoxatine-2,3-dione) antagonizes NMDA-evoked J-‘H]GABA release from cultured cortical neurons via an inhibitory actisn at the strychnine-insensitive glycine site, Brain Res. 489, 185. Honor&, T., S.N. Davies. J. Drejer. E.J. Fletcher. P. Jacobsen, D. Lodge and F.E. Nielsen, 1988. Quinoxalin~iones: Potent competitive non-NMDA glutamate receptor antagonists, Science 241, 701. Johnson. J.W. and P. Ascher. 1987. Glycine potentiates the NMDA response in cultured mouse brain neurons. Nature [London) 325, 529. Kessler. M.. T. Terramani. G. Lynch and M. Baudrey. 1989. A glycine site associated with N-methyl-D-aspartic acid receptors: characterization and identification of a new class of antagonists. J. Neurechem. 52. 131s. Lester. R.A.J.. M.L. Quarum. J.D. Parker. E. Weber and C.E. Jahr, 1989, Interaction of 6-cyano-7.nitroquinoxaline-2.3-dione with the N-methyl-D-aspartate receptor-associated glycine binding site, Mol. Pharmacol. 35. 565. Long. SK., D.A.S. Smith. R.J. Siarey and R.H. Evans, 1990, Effect of 6-cyano-2,3-dihydroxy-%nitro-quinoxalone (CNQX) on darsai root-, NMDA-, kainate-. and quisqualate-mediated depolarization of rat motoneurones in vitro. Br.J.Pharmacol. 100.850. Otsuka, M. and S. Konishi. 1974, Ei~troph~sioIogy of mammalian spinal cord in vitro, Nature (London) 252.733. Preston, P.R. and D.l.Wallis. 1980. Characteristics of dorsal root potentials recorded from the isolated spinal cord of the neonate rat. J. Neural Trans. 48, 271. Salt. T.E., 1989. M~ulation of NMDA r~eptor-mediate responses hy glycine and D-se&e in the rat thalarnus in vivo, Brain Res. 481,403.
Thomson. A.M.. 1990, Glycine is a coagonist at the NMDA receptor/ channel complex. Prog. Neurobiol. 35. 53. Thomson, A.M.. V.E. Walker and D.M. Flynn, 1989, GlycZne enhances NMDA-receptor-mediated synaptic potentials in neocortical slices. Nature (London} 338. 422.
195 (1991)
241-244
241
0 1991 Elsevier Science Publishers B.V. 0014.2999/91/$03.50 ADONIS
0014299991002442
WP 51787
Potentiation of synaptic reflexes by II-serine in the rat spinal cord in vitro Richard DUPHA
J. Siarey, Stephen
R B. V.. P.O. Box 900. 1380 DA
K. Long and Richard
Weesp, The Netherlands
and
’
Department
H. Evans 1
of Pharmacolom,
School of Medical
Sciences,
Bristol BS8 1 TD, U.K.
Received 25 October 1990. revibed MS received 28 December 1990. accepted 8 January 1991
6-Cyano-7-nitroquinoxaline-2,3-dione (CNQX) (IO PM) depressed dorsal root-evoked ventral and dorsal root potentials of the in vitro immature rat spinal cord to 26.3 f 5.2 S.E.M. and 40.8 f 2.7% of control values respectively. These depressant effects of CNQX were partially reversed by D-serine (EC,, values 39.7 PM f 8.7 S.E.M. N = 6 and 34.9 + 12.5 PM, N = 5 for ventral root potential and dorsal root potential respectively). Under our experimental conditions, which included the presence of Mg2+ (0.75 mM) in the bathing medium, no measurable potentiation of these synaptic reflexes by D-serine was recorded in the absence of
CNQX. These data indicate that CNQX. in addition to its depressant effect at non-NMDA receptors, depresses an NMDA receptor-mediated component of segmental transmission through its action at the glycine site of the NMDA receptor complex. D-Serine; Spinal cord; CNQX (6.cyano-7-nitroquinoxaline-2.3.dione);
1. introduction The NMDA receptor/channel is proving to be a very complex structure with many ligand binding sites. One of these is the strychnine-insensitive glycine binding site at which D-serine is an agonist (Johnson and Ascher, 1987; reviewed by Thomson, 1990). 6-Cyano-7nitroquinoxaline-2,3-dione (CNQX) has been recently introduced as a potent and selective antagonist of nonNMDA receptors (Honor6 et al., 1988). Subsequent studies have shown that CNQX can antagonise NMDA-induced responses and thai this antagonism can be reversed by D-serine. Thus the relatively weak antagonism at NMDA receptors produced by CNQX is probably due to an action at the strychnine-insensitive glycine site. This interaction is suggested from binding site displacement studies (Kessler et al., 1989; Lester et al., 1989), NMDA-evoked release studies (Harris and Miller, 1989) and from electrophysiological responses to NMDA of in vitro spinal cord preparations (Birch et al., 1988; Long et al., 1990) and isolated neurones (Lester et al., 1989). However, there are relatively few studies which show operation of the glycine site during synaptic transmission. In the absence of applied antagonists glycine (Thomson et al., 1989) and D-serine (Salt, 1989) have been reported to potentiate NMDA receptor-mediated excitatory synaptic responses in
Correspondence to: R.H. Evans, Department of Pharmacology. School of Medical Sciences, Bristol BS8 1TD. U.K.
(Electrophysiology)
cerebrocortical slice preparations and in the thalamus in vivo respectively. In the presence of CNQX D-serine has been shown to reverse the synaptic depressant action of a glycine site antagonist in cerebrocortical (Fletcher et al., 1989) and hippocampal (Bashir et al., 1990; Gaiarsa et al., 1990) slice preparations. The present paper describes an investigation of this action of D-serine on synaptic transmission in the neonate rat spinal cord.
2. Materials and methods Two to five day old rats were anaesthetised with urethane (1.8 g per kg i.p.). Following decapitation spinal cords were removed, hemisected and mounted for electrical stimulation of dorsal roots and recording from an adjacent dorsal root (dorsal root potential) and/or corresponding ventral root (ventral root potential) as previously d,escribed (Long et al., 1990). Synaptic potentials were recorded with an SMR laboratory interface connected to an IBM microcomputer and controlled by a software package developed in house (J.M. Van der Zalm, Duphar BV R & D Automation Department). The programme was used to measure amplitudes and integrate areas (over 1.65 s) of synaptic potentials. Mean values &S.E. (from N preparations) are presented in Results. The significance of differences between mean values was assessed from a rank order test (Wilcoxon-Mann-Whitney).
3 Resurts
CONTROL
A
prqarations were superfused with continuously 8% ~ium at a rate of 1 ml per min maintains at S”C and gassed with a mixture of 95%-S% of O&t& to pH 7.4. The bathing medium had the following ~~rn~~on (mM): NaCl 118. NaHCC& 24, KC! 3, CaCl z 1.5. MgsO, 0.75, dextrose 12. All drugs were diluted in the bathing medium before superfusion on to the preparation. CNQX and ( + )2-amino-Sphosphono pentanoate (AP5) were supplied by Tocris Chemicals.
n
DR-*RL./ L J-L_._ L B
A&J
I
In the present seven preparations the mean latency and the mean time to the initial peak (monosynaptic component) of the ventral root potential (8.9 + 0.6 and It.8 k 0.6 ms respectively) corresponded with the initial report on this preparation (Otsuka and Konishi, 1974). Componding values for the dorsal root potential in the present experiments were 17.4 & 0.8 and 110 Z!Z 6 ms. The latter time to peak is similar to that reported by Preston and Wallis (1980) in the absence of added Mg’+. The dorsal root reflex was not prominent in the present experiments probably due to a combination of low stimulus intensity (Preston and Wallis. 1980) and presence of 0.75 mM Mg” (Preston and Wallis, 1980). The mean Peak amplitude, area and time to half decay of the ventral root potential were 3.06 f 0.62 mV, 805 + 228 @J - s and 11.I zt:0.5 ms (N = 7). Corresponding values for the dorsal root potential were 1.37 + 0.5 mV, 599 +_241 PV - s and 42.5 + 2.7 ms (N = 6). As expected from previous experiments with mature spinal preparations (Long et al., 1990) CNQX (10 PM) reversibly depressed both the ventral root potential and dorsal root potential (fig. 1A). The amplitude of the initial monos~ap~c component of the ventral root potential was depressed from control (100%) values to 6.7 + 1.7% (N = 7) and the maximum amplitude of the dorsal root potential was depressed from control values to 44.9 + 3.8%. Areas of ventral root potential and dorsal root potential, which reveal polysynaptic components, were depressed from control values to 26.3 + 5.2 and 40.8 + 2.7% (N = 6) respectively. The difference in CNQX-sensitivity between the two synaptic responses was highly significant (P < 0.002). The selective antagonist AP5 was used to show the involvement of NMDA receptors in generation of both synaptic potentials. As can be seen in fig. 1B a brief (2 min) application of AP5 (25 PM), in the presence of CNQX, produced a further depression of both potentials. As well as depressing the ~p~tudes the CNQX produced a significant increase in latency of the synaptic potentials. Thus in the presence of CNQX (10 PM) latencies of ventral and dorsal root potentials were increased to 14.3 -t 1.3 and 25.2 + 33 ms respectively (N = 7). Tilese values were not significantly altered fol-
CNQX IO
D-SERINE200
25
I
l-----i
I
Fig. 1. (A) Depression of dorsal root-evoked dorsal (DR-DRP) and ventral (DR-VRP) root potentials by CNQX (10 FM). The left band traces are controls 5 min before intr~ucti~ of CNQX and the right hand traces were recorded 20 min Wowing introductZo%of CEQX. The upper and lower traces are synchronous and show depression of DR-VRP and DR-DRP respectively. Vertical calibration I mV, horizontal 10 ms (DR-VRP) and 200 ms (DR-DRP). (B) Same preparation as (A) following intr~uction of CNQX. APS and D-serine were introduced as indicated above the traces. AP5 was present for 2 min and D-serine for 5 min. APS depressed and D-serine potentiated both potentials. Recovery is shown 20 min following introduction of AP5and D-se&e-free medium. Vertical calibration 0.5 mV (upper) 1 mV (lower), horizontal 20 ms (upper) and 500 ms (lower).
lowing introduction of APS (25 FM) or D-serine (200 FM). The potentiating effect of D-serine is illustrated in fig. 1B following recovery from the depressant effect of APS. These effects on several preparations are summarised in table 1 where it can be seen that in the presence of D-serine potentials were enhanced above the control levels in the presence of CNQX alone. The effects of AP5 and D-serine on amplitude of the ventral root potential shown in the first column of table 1 are on the longer latency peak and not the monosynaptic component since the latter peak was absent in the presence of CNQX. The potentiating effect of D-serine (200 PM) was significant at the P < 0.014 level. In order to determine EC, values D-serine was added to the superfusate in cumulative concentrations as illustrated in fig. 2A. EC,, values for the potentiating effect of D-serine, determined from dose-response plots as illustrated in fig. 2B, were 39.7 f 8.7 CM and 34.9 f 12.5 FM on the ventral root potential and dorsal root potential respec-
243
4. Discussion
TABLE 1 Effect of APS (25 FM) and R-serine (200 CM) on amplitude and areas of dorsal root-evoked ventral and dorsal root potentials in the presence of CNQX (10 PM). N = 7 for the ventral and 6 for the dorsal root potential. The synaptic responses were depressed by AP5 and potentiated by D-serine. Percentages are of control areas before introduction of CNQX. Mean absolute control values are given in the text. The drugs were applied as illustrated in fig. 2. Treatment
Ventral
CNQX CNQX + AP5
Dorsal
Peak
Area
Peak
Area
6.7kl.7 4.7 f 0.8
26.3k5.2 6.8kO.9 47.7+7.3
44.9k3.8 l&4&4.6 54.7+5.4
40.8*2.7 13.2f3.9 54.4&4.6
CNQX+D-serine 8.Ok2.2
tively. These values are not significantly different (P = 0.33). In the absence of CNQX no significant potentiating effect of D-serine was observed on three preparations when applied at ~ncentrations up to 1 mM.
240
300
360
420
480
540
so -
60-
40 -
x-
/ i
d
oL 10
’
“‘11 too
c
Fig. 2. (A) Plot of areas (pV+s) of dorsal root-evoked ventral root potential against time (mm). Introduction of drugs at micromolar concentration is indicated by the bars above the plot. The plot of percentage potentia~ion against concentration of D-serine (PM) shown in (B) was produced from the cumulative effect of D-set-me shown in (A).
The present study shows that CNQX is a Potent depressant of synaptic potentials in the neonate rat spinal cord in agreement with previous obserations on the ~P~b~~ (Fletcher et al., 1988) and the adult rat (Long et al., 1990) spinal cord. The depressant effects of both CNQX and APS on the dorsal root potential (fig. 1 and table 1) shows that a combination of excitatory amino acid receptors are involved in this synaptic potential of neonate preparations as has been previously shown to occur in preparations from adult rats (Evans and Long, 1989). In the present experiments CNQX increased the latencies of the two synaptic potenti&. This CNQX-induced increase in latency was not reversed by D-set-me which is consistent with the role of non-NMDA receptors in mediation of the initial cornponents of these spinal reflexes (Evans and Long, 1989). A significant difference between neonate (present results) and adult (Long et al., 1990) reflexes appears to be the presence of a relatively short latency APS-sensitive component in the ventral root potential of the neonate, in the presence of Mg*+, which was revealed following application of CNQX (fig. 1). This may be a general feature of immature synaptic circuitry because a similar difference has been observed in the hippocampus (Gaiarsa et al., 1990). Despite the presence of a significant cont~bution of NMDA receptors to the population excitatory post-synaptic potential no potentiating effect of D-set-me was seen in the absence of CNQX. It is possible that the difference between spinal and cerebral (Thomson et al., 1989) preparations in the lat:er respect could be explained by heterogeneity of the NMDA-linked glycine receptor between spinal cord and forebrain (Danysz et al., 1990). However, in the hip~~pal slice preparation the effect of D-serine was similar to the present report in that no potentiation of synaptic responses was observed in the absence of a glycine site antagonist (Bashir et al., 1990). The reversal by D-serine of the synaptic depressant effects of CNQX observed in the present study could be apparent rather than a reflection of direct competition between CNQX and D-serine. However, the concentrations of D-serine required to reverse the synaptic depressant effect of CNQX in the present study were similar to those previously reported to reverse the antagonism of NMDA-induced depolarizations in these (Long et al., 1990) and cerebrocor?ical (Fletcher et al., 1989) preparations. These concentrations of D-serine or glycine are at least an order of mag~tude higher than concentrations which are effective on NMDA activated ionic currents of isolated cells or on displacement of radioligand from washed isolated membranes (Lester et af., 1989). This difference in effective con~ntrations has been attributed to the presence of a SatUKttiUg TeVeTof
~~te~t~t~~~glycine in whole true preparations (Fletcher et a&.. ~~~~~. The isolated ceil and men~brane studies qsrc:Thomson. 1990) show the dissociaGon constants For &cinr or D-se&c at the tglycine site ( < 0.3 pM) to be ccmsistcnt with more than 95% saturation of the site al a~~ro~~rnate~y6 $4 agonist. In the present e~per~rnents it can be assumed that intrinsic glycine was close to a saturating concentra\ion because NMDA recceptor~~~iat~ excitation was not enhancedin the absence of sntsgonist.CFXjX (10 pM) would produce a dose-ratio of approsimatcly six (calculated from the K, of 2.2 PM reported by Kesskr et al.. 1989) at the glycine site. Thus the present EC,, of 30-40 PM for reversal of the depressant effect of CNQX by D-serine would be consistent with restoration of a saturating level of glycine site agonist by a 6-fold increase in concentration. The uptake of exogenous D-serine or glycine could also be involved in reducing their apparent potency on whole tkssue preparations. Thus the present data suggest that CNQX, in addition to a depressant effect at non-NMDA receptors. depresses an NMDA receptor-mediated component of segmental transmission through an action at the strychnine-insensitive glycine site of the NMDA receptor compfex.
R.J.S. is a DUPHAR research scholar and R.H.E. was supportrd Research Council. The Taberner Trust and The Wellc~me Trust.
h?: the Mrdkd
References Bashir. 2.1.. 3. Tam and G.L. Colling~dge. 199% Activation of the glycins site in the NMDA receptor is necessary for the induction of LTP. Neurosci. Lett. 108, 261. Birch. P.J.. CJ. Grossman and A.G. Hayes, 1988, 6,7-Dinitroquinoxalin~~3-dione and 6-nitr~7~yan~quinoxaline-2,3dione antagonise responses to NMDA in the rat spinal cord via an action at the strychnine-insensitive glycine receptor, European J. Pharmacol. 156, 177. Danysz. W.. E. Fadda. J.T. Wrohlewski and E. Costa. 1990. (‘HI-DSerine Iabets st~chnine-insensitive glycine recognition sites of rat central nervous system. Life Sci. 46. 155.
Fvanb. R.H. and S.K. Long, 1989. primary afferent depolarization in the mt spinal cord is mediated by pathways utilising NMDA and n.m-NMDA receptors. Neurosci.Lett. 100. 231. Ftetrner. E.J.. D. Martin. J.A. Aram. D. Lodge and T. Honor& 1988. Quinoxalinediones selectively block quisqualate and kainate receptors and synaptic events in rat ncocortex and hippocampus and frog spinal cord in vitro, bi. J. Ph~rmacol. 95, 585. Fletcher. E.J.. J.D. Milfar, S. Zeman and D.Lodge, 1989. Non-competitive antagonism of N-methyl-D-aspartate by displacement of an endogenous glycine-like substance, European J. Neurosci. 1,196. Gaiarsa. J.L.. R. Corradetti. E. Cherubini and Y. Ben Ari. 1990. The allosteric glycine site of the N-methyl-D-aspartate receptor modulates GABAergic-mediated synaptic events in neonatal rat CA3 hippocampal neurons, Proc. Natl. Acad. Sci. U.S.A. 87, 343. Harris. K.M. and R.J. Mifler, 1989. CNQX (6-cyano-%nitroquinoxatine-2,3-dione) antagonizes NMDA-evoked J-‘H]GABA release from cultured cortical neurons via an inhibitory actisn at the strychnine-insensitive glycine site, Brain Res. 489, 185. Honor&, T., S.N. Davies. J. Drejer. E.J. Fletcher. P. Jacobsen, D. Lodge and F.E. Nielsen, 1988. Quinoxalin~iones: Potent competitive non-NMDA glutamate receptor antagonists, Science 241, 701. Johnson. J.W. and P. Ascher. 1987. Glycine potentiates the NMDA response in cultured mouse brain neurons. Nature [London) 325, 529. Kessler. M.. T. Terramani. G. Lynch and M. Baudrey. 1989. A glycine site associated with N-methyl-D-aspartic acid receptors: characterization and identification of a new class of antagonists. J. Neurechem. 52. 131s. Lester. R.A.J.. M.L. Quarum. J.D. Parker. E. Weber and C.E. Jahr, 1989, Interaction of 6-cyano-7.nitroquinoxaline-2.3-dione with the N-methyl-D-aspartate receptor-associated glycine binding site, Mol. Pharmacol. 35. 565. Long. SK., D.A.S. Smith. R.J. Siarey and R.H. Evans, 1990, Effect of 6-cyano-2,3-dihydroxy-%nitro-quinoxalone (CNQX) on darsai root-, NMDA-, kainate-. and quisqualate-mediated depolarization of rat motoneurones in vitro. Br.J.Pharmacol. 100.850. Otsuka, M. and S. Konishi. 1974, Ei~troph~sioIogy of mammalian spinal cord in vitro, Nature (London) 252.733. Preston, P.R. and D.l.Wallis. 1980. Characteristics of dorsal root potentials recorded from the isolated spinal cord of the neonate rat. J. Neural Trans. 48, 271. Salt. T.E., 1989. M~ulation of NMDA r~eptor-mediate responses hy glycine and D-se&e in the rat thalarnus in vivo, Brain Res. 481,403.
Thomson. A.M.. 1990, Glycine is a coagonist at the NMDA receptor/ channel complex. Prog. Neurobiol. 35. 53. Thomson, A.M.. V.E. Walker and D.M. Flynn, 1989, GlycZne enhances NMDA-receptor-mediated synaptic potentials in neocortical slices. Nature (London} 338. 422.
