The role of nitric oxide in hippocampal long-term potentiation

neuron,

Vol. 8. 211-216,

February,

1992, Copyright

0 1992 by Cell Press

The Role of Nitric Oxide in Hippocampal Long-Term Potentiation Jane E. Haley,*+ George 1. Wilcox,+ and Paul F. Chapman* *Department of Psychology +Department of Pharmacology University of Minnesota Minneapolis, Minnesota 55455

Summary long-term potentiation is a long-lasting, use-dependent increase in the strength of synaptic connections. We investigated the role of nitric oxide (NO) in determining the duration of potentiation induced by high frequency stimulation of afferents in the CA1 region of the rat hippocampus. The calciumkalmodulindependent production of NO can be initiated by activation of excitatory amino acid receptors and results in increased levels of cCMP in target cells. Here we report that only a relatively short-term potentiation can be induced in the presence of nitro-L-arginine methyl ester (L-NAME), an NO synthase inhibitor. The effects of L-NAME on the duration of potentiation are partially reversed by coadministration of L-arginine, a precursor of neuronal NO, and by dibutyryl cGMP. Hemoglobin, which binds extracellular NO, also shortens the duration of stimulus-induced potentiation. The results suggest a role for NO in the maintenance of activity-dependent synaptic enhancements, possibly via the generation of cCMP. Introduction Long-term potentiation (LTP) is considered a cellular analog of learning and memory in the mammalian nervous system (Bliss and Lynch, 1988; Brown et al., 1988; Madison et al., 1991). The mechanisms bywhich this increase in synaptic efficacy is maintained over long periods of time has been an area of increasing interest (Malenka, 1991). While the role of calcium in LTP induction has been well established (Bliss and Lynch, 1988; Malenkaet al., 1988), the means bywhich postsynaptic calcium concentrations affect synaptic strength are still unknown. Activation of the N-methyl+aspartate (NMDA) subtype of excitatory amino acid receptor, which permits the influx of calcium, is necessary for the induction of LTP in the CA1 cell field of hippocampus. Interestingly, NMDA receptor activation has also been shown to stimulate an increase in cyclic guanosine monophosphate (cGMP) in the hippocampus. This increase in cGMP was blocked by inhibitors of nitric oxide synthase (NOS) (East and Garthwaite, 1991). Nitric oxide (NO), which is synthesized by the calcium/ calmodulin-dependent NOS, is known to activate soluble guanylate cyclase, suggesting that NO may mediate excitatory amino acid-induced cGMP elevation. The fact that cGMP levels are elevated by NMDA

receptor activation in the hippocampus and that this elevation depends on a calcium/calmodulin-dependent enzyme suggests that NO could be involved in the induction or maintained expression of synaptic enhancements (Garthwaite, 1991; Snyder and Bredt, 1991). Indeed, recent evidence suggests that inhibition of NOS affects synaptic plasticity (B6hme et al., 1991). We examined the role of NO in the maintained expression of LTP by assessing the effects of inhibiting NOS with nitro-L-arginine methyl ester (L-NAME) and by administering extracellular hemoglobin, which serves as a false substrate for NO. Our results indicate that either blocking the synthesis of NO or sequestering NO extracellularly reduces the duration of synaptic plasticity induced by high frequency stimulation. While L-NAME effectively limited the duration of tetanus-induced potentiation, the stereoisomer, D-NAME, was ineffective. The results of NOS inhibition by L-NAME could be overcome by coadministration of the enzyme precursor, L-arginine, or the cGMP analog dibutyryl cGMP (DiBcGMP). In contrast, L-arginine was unable to reverse the effects of hemoglobin at concentrations that were effective at counteracting L-NAME. Results LTP was induced in hippocampal slices maintained in vitro by stimulation of the Schaffer collateral/commissural inputs to CA1 pyramidal neurons. Figures IAand IB illustrate representative extracellular field excitatory postsynaptic potentials (EPSPs) recorded from the stratum radiatum before and after delivery of tetanic stimulation in the presence of L-NAME or D-NAME. Six of seven slices that received high frequency stimulation in the presenceof 100 nM L-NAME showed an increase of greater than 20% 15 min after tetanus. In these same slices, the slope returned to pretetanus baseline 50 min after tetanus (mean slope at 50 min = 101% of baseline, + 3%, SEM). By contrast, when tetanic stimulation was delivered in the presence of 100 nM D-NAME (Figure ID), seven of eight slices were still potentiated at 60 min after tetanus (mean slopeat60min = 137% of baseline, + 5%, SEM). Thus L-NAME significantly reduces the duration of the synaptic enhancement. L-NAME (10 PM) had no effect on synaptic responses to low frequency stimuli, either before or after LTP induction, even at 100 times the lowest concentration affecting the duration of synaptic enhancement. Thirty minutes after washing in IO PM L-NAME, the slope of the field EPSP in response to low frequency stimulation was 92% of baseline (+ 5%, SEM). This was not significantly different from the response before wash-in (p >0.3). In four slices in which IO PM L-NAME was washed in after tetanus, LTP expression was unaf-

C

250,

501

5OO-

0

IO

20

Figure

1. The

Effect

of Pretreatment

30

40

50

60

Time (minutes)

Time (minutes) with

L-NAME

and the Inactive

Stereoisomer

D-NAME

on LTP in Rat Hippocampal

Slices

(A and C) Pretreatment with 100 nM L-NAME; (B and D) pretreatment with the same concentration of o-NAME. (A and B) Representative recordings of extracellular EPSPs prior to tetanic stimulation (trace 1) and 15 min (trace 2) and 60 min (trace 3) following delivery of the tetanus. Each waveform is the average of ten consecutive responses, evoked at 1 per 15 s. The mean negative slope of the rising EPSP is plotted against time in (C) and (D). Data were normalized to the pretetanus baseline and calculated as the mean (* SEM) per minute over several slices. Tetanic stimulation was delivered after a 5 min stable baseline had been established, and this resulted in an increase in the EPSP slope. The increase that was induced while L-NAME was in the perfusate (C) was transient and declined steadily. Mean EPSP slope returned to the pretetanic level about 40-50 min after tetanus (at 60 min; p = 0.5, n = 7, compared with pretetanus baseline). By contrast, high frequency stimulation in the presence of the inactive stereoisomer o-NAME (D) resulted in an immediate increase in the slope that declined to a significantly higher level 60 min after tetanus, compared with the pretetanus baseline (p < 0.01, n = 7).

posttetanus response beforewash-in (p>O.l). Despite its lack of effects on both synaptic responses to low frequency stimulation and previously induced LTP, we often observed that the slope of the field EPSP was

fected (Figure 2). The mean EPSP slope 30 min after L-NAME wash-in (approximately 60 min after tetanic stimulation) was 125% (+ 5%, SEM) of the pretetanus response and was not significantly different from the

Figure

IA -p+/--f-os-a,;-pf-B

rso-

1

lO,,ML-NAME

10

20

30

40

Time (minutes)

so

60

70

in After

Tetanus

HadNoEffecton LTp

200 -

5o I 0

Washed

Representative averaged EPSPs at pretetanus (trace 1) and 15 min (trace2) and 60 min (trace 3) posttetanus are shown in (A). The normaliied EPSP slope data from a single experiment are shown in (B). L-NAME was washed in at a dose 100 times greater than that requited to block maintenance as a pretreatment. There was no change in the potentiated slope of the EPSP in all four slices.

2so-

100 r-

2. L-NAME

80

Kale of Nitric 213

Oxide

in LTP

n

D-NAME

Figure

@

L-NAME

tially

q q

i- NAME + L- arginine

ginine

L- arginlne

* T

“”

D-NAME

IOOnM

1 VM

10UM

L-NAME $0 PM L-arglnw

smaller than baseline 60 min after tetanus delivered in the presence of 10 WM L-NAME (see Figure 3). NOS catalyzes the production of NO and citrulline from L-arginine (Garthwaite, 1991). The addition of L-arginine has been shown to reduce or abolish the effects of NOS inhibitors in both the cardiovascular and the central nervous systems (Palmer et al., 1989; East and Garthwaite, 1991; Shibuki and Okada, 1991). L-NAME was applied at concentrations ranging from 10 nM to IO PM (Figure 3). When L-arginine (3-10 times the L-NAME concentration) was added to the perfusate along with L-NAME (100 nM to 10 PM), the effects of thesynthaseinhibitorwere markedlyattenuated. L-ATginine alone (30 PM) had no statistically significant effect on the size of the synaptic enhancement following tetanic stimulation. One of the principal effects of NOS in both the cardiovascular and the central nervous systems is the production of cGMP via the activation of guanylate cyclase (Ignarro et al., 1986; Garthwaite et al., 1988; Bredt and Snyder, 1989). It is possible, therefore, that the conversion of short-term potentiation (STP) to LTP is the result of elevated cGMP levels following high frequency stimulation. The actions of L-NAME on the duration of synaptic enhancement may result from the prevention of NO-generated increases in cCMP production (East and Garthwaite, 1991). In an attempt to counter the effects of L-NAME, we added the phosphodiesterase-resistant cGMP analog DiBcGMP to the perfusate along with L-NAME (Figure 4). Slices that were given high frequency stimulation in medium containing 100 nM L-NAME and 100 PM DiBcGMP still demonstrated significant increases in EPSP slope 60 min after tetanus (mean percentage of pretetanus baseline = 112% + 4%, SEM; p < 0.02). DiBcGMP applied in the absence of L-NAME had no effect on synaptic responses to low frequency stimulation 60 min after wash-in (mean percentage of baseline be-

3. The Effects of L-NAME Were ParReversed by the NO Precursor L-Ar-

The data represent the mean slope of the EPSP (normalized to the pretetanus baseline) 50-60 min afterthedeliveryof thetetanus, for several slices in each group. Pretreatment with t-NAME (IO nM, n = 5; 100 nM, n = 7; 1 PM, n = 8; 10 PM, n = 6) resulted in EPSP slopes that were lower than those seen with 100 nM D-NAME (n = 7). t-Arginine administered with t-NAME at a dose 3-10 times higher (100 nM + 1 kM L-arginine, n = 6; 1 PM + 3 fiM L-arginine, n = 8; IO PM + 30 PM L-arginine, n = 5) resulted in mean slopes greater than those seen with L-NAME alone. Tetanus delivered in the presence of L-arginine alone (30 PM, n = 4) resulted in LTP that was not significantly different from that elicited in D-NAME. A single asterisk indicates p < 0.05; double asterisks, o < 0.01, compared with D-NAME.

fore wash-in = 95% f 2%, SEM; p > 0.2). Moreover, DiBcGMP alone did not affect LTP (mean percentage of pretetanus baseline 60 min after tetanus = 117% + 5%, SEM). NO is known to bind with high affinity to the heme moiety in hemoglobin. To examine the effect of removing endogenously generated NO, hemoglobin was bath applied at a concentration of 100 PM (Figure

Figure 4. DiBcCMP Partially Synaptic Enhancement

Reverses

the Effects

of L-NAME

on

(A) The size of the mean EPSP slope measured 60 min after tetanus in 100 PM DiBcGMP and 100 nM L-NAME (n = 8) is significantly greater than that of slices bathed in 100 nM L-NAME alone (n = 7). Alone, 100 FM DiBcCMP has no effect on LTP (A, n = 3), or on the synaptic responses to low frequency stimulation (B, n = 4). A single asterisk indicates p < 0.03, compared with 100 nM L-NAME alone.

NWKXl 214

maintained expression of long-term synaptic enhancements produced by tetanic stimulation in the hippocampus. Administration of L-arginine, the precursor of NO, is capable of reversing the effects of L-NAME. Our results here are consistent with other reports indicating that blockage of NOS can be partially overcome by supplying additional precursor (Palmer et al., 1989; Bohmeet al., 1991; Shibuki and Okada, 1991). Both the ineffectiveness of the stereoisomer D-NAME and the reversibility of the actions of L-NAME by L-arginine attest to the specificity of the action of L-NAME on NOS. Equally important is the result with extracellular hemoglobin application, since it is a large protein that is unable to access the intracellular environment. The ability of hemoglobin to produce the same effect on potentiation as L-NAME strongly implies that thediffusion of NO through the extracellular space plays a significant role in synaptic plasticity.This suggeststhe possibility that NO might be a retrograde messenger produced in postsynaptic neurons by elevated calcium, but acting on presynaptic neurons, on glia, or at neighboring postsynaptic sites. Hemoglobin was still able to shorten the duration of tetanus-induced potentiation even when coadministered with high doses of L-arginine. Thus, while L-arginine was able to overcome NO synthesis inhibition, it was not sufficient to ameliorate the effects of sequestering released NO. This may result from the inability of NOS to produce enough NO to saturate the hemoglobin. One important consequence of NO production is an increase in levels of cGMP. Indeed, application of NMDA to hippocampal as well as cerebellar slices

5). High frequency stimulation of the Schaffer collaterals resulted in STP, which closely resembled the responseobservedwith L-NAME.Theslopeofthefield EPSP 60 min after tetanus was not significantly different from the pretetanus value (mean percentage of pretetanus baseline = 102% k 6%, SEM). The effect of hemoglobin could not be reversed by application of 30 t.rM L-arginine (Figure 5), a dose that prevented even the highest concentration of L-NAME from affecting the duration of LTP (mean percentage of pretetanus baseline at 60 min = 102% + 4%, SEM). Administration of 100 PM hemoglobin had no effect on synaptic responses to low frequency stimulation (mean percentage of baseline before wash-in, at 30 min = 83% k 8%, SEM; p = 0.19, df = 2), as was the case with L-NAME. Discussion The conversion of STP to LTP is likely to involve the actions of a calcium-dependent second messenger system (Malenkaetal.,1988;Malinowetal.,1988; Madison et al., 1991; Malenka, 1991). A particularly attractive possiblity is that calcium and/or calmodulin are activating an enzyme in the postsynaptic neuron as a result of high frequency stimulation. The calcium/ calmodulin dependence of NOS raises the possibility that this enzyme could be activated under these conditions and therefore play a critical role in LTP (Garthwaite, 1991; Snyder and Bredt, 1991). Moreover, NOS activation can be induced by excitatory amino acid agonists in a calcium-dependent manner (CarthWaite et al., 1988). Our results indicate that using L-NAME to block the synthesis of NO prevents the

Figure 5. Hemoglobin tion of Tetanus-Induced

250,

+ -

50-1

0

v

10

1 20

30 Time

40 (minutes)

100 pM hemoglobin ~30 pM L-arglnine 100 pM hemoglobin

50

60

Shortens the DuraPotentiation

High frequency stimuli given in the presence of 100 PM hemoglobin (n = IO) produce potentiation that returns to pretetanus baseline values within 60 min (p Z 0.6, df = 9). The addition of 30 PM L-arginine (n = 11) does not reverse the effects of hemoglobin (p > 0.4, df = IO).

Role of Nttric 215

Oxide

in LTP

results in an increase in the levels of cGMP, probably via the direct activation of guanylate cyclase by NO (East and Garthwaite, 1991; Southam and Garthwaite, 1991). It is likely, therefore, that the effects of NO are mediated by cGMP or a cGMP-dependent process. Thus, an increase in the intracellular concentration of cGMP may contribute to the maintained expression of LTP. When cCMP levels were increased by exogenous application of the phosphodiesterase-resistant DiBcGMP, L-NAME no longer reduced the duration of synaptic enhancement. We did not observe an effect of DiBcGMP alone on synaptic efficacy or on the response to high frequency stimulation at concentrationscapableof reversing the effects of L-NAME. This evidence suggests a role for cGMP in determining the duration of synaptic enhancement produced by high frequency afferent stimulation. One popular hypothesis states that the duration of LTP depends on the activation of protein kinase C and/or calcium/calmodulin-dependent kinase II. Inhibitors of these two kinases are capable of preventing the maintenance of synaptic enhancements (Lovinger et al., 1987; Malinow et al., 1988; Reymann et al., 1988; Malenka et al., 1989; Tsien et al., 1990), as is L-NAME. There is no reason to believe, however, that these two second messenger systems are mutually exclusive. Indeed, it is possible that protein kinase C and/or calcium/calmodulin-dependent kinase II and an NO-dependent process are all critical for establishing long-duration synaptic enhancements, perhaps by acting at different points on a multistep molecular cascade. While it has recently been suggested that NO may be released from neurons and may participate in synaptic transmission (Bredt et al., 1990; Garthwaite, 1991; Shibuki and Okada, IVVI), our results indicate that blocking either the synthesis or the extracellular diffusion of GO does not’affect synaptic responses to low frequency stimulation at Schaffer collateral/commissural synapses in CAI. However, NO appears to play an important role in the maintenance of synaptic plasticity in the hippocampus. NO is known to diffuse out of the cells in which it is generated (Garthwaite, 1991; Shibuki and Okada, 1991). Thus, it could serve as the retrograde signal that would permit the participation of presynaptic neurons and/or glia in the expression of LTP induced by the activation of postsynaptic NMDA receptors. Indeed, evidence from three other laboratories corroborates our findings and suggests a similar mechanism (Btihme et al., 1991; O’Dell et al., 1991; Schuman and Madison, 1991).

mately2 ml/min.Therecordingchamberwas maintainedat room temperature (22°C) throughout the experiments. L-NAME, o-NAME, L-arginine, hemoglobin, and DiBcGMP were dissolved from stock solutions (in distilled water) into ACSF duringexperimental treatment.Alldrugswereobtainedfrom Sigma, except D-NAME, which was purchased from Bachem. Picrotoxin (20 vM) was routinely added to the perfusing medium in order to block inhibition mediated by y-aminobutyric acid type A receptors. When pretreatment experiments were performed (Fig ures 1, 3, 4A, and 5), slices were incubated for 15 min in drugcontaining solutions before they were transferred to the recording chamber. Extracellularfield potentials were recorded in thestratum radiatum of CA1 with 2 M NaCI-filled electrodes, in response to stimulation of the stratum radiatum at the CAZ/CAl border. Stimuli were delivered once per 15 s via a bipolar, stainless steel stimulating electrode. Stimulus strength was adjusted sothatthe response was just subthreshold for evoking a population spike. LTP was induced by delivering a train of ten, 40 ms bursts at 100 Hz; the interval between bursts was 200 ms (Larson et al., 1986). Population EPSPs were digitized and stored on a computer. The peak amplitude and the slope of the initial portion of the EPSPs were monitored on-line. Experiments were performed only on slices that demonstrated stable responses to low frequency stimulation for at least 5 min. When appropriate, statistical comparisons were made using Student’s t test for either paired or unpaired groups.

Experimental

East, S. J., and Garthwaite, J. (1991). NMDA receptor activation in rat hippocampus induces cyclic CMP formation through the L-arginine nitric oxide pathway. Neurosci. Lett. 723, 17-19.

Procedures

H ippocampal slices (450 pm) were prepared from male rats using standard methods (Kelso and Brown, 1986). Slices were incubated in artificial cerebrospinal fluid (ACSF) containing the following: 125 mM NaCI, 2 mM KCI, 1.25 mM NaPO+ 26 mM NaHC03, IO mM o-glucose, 4 mM MgSO+ 4 mM CaC12, and 1 WM glycine. Slices were transferred oneat a time to an interface-type recording chamber through which ACSF flowed at approxi-

Acknowledgments This work was supported by USPHS grant #ROl-DA-04274 to C. L. W. and by grants from the University of Minnesota Graduate School to P. F. C. We would like to thank Nancy Peterson for technical advice and Michael Mauk, David Wyllie, Steven Traynelis, and Edward Kairiss for critical comments on the manuscript. We aregrateful to Sigma Chemical Company for their gift of DiBcGMP. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “advertisement” in accordance with 18 USC Section 1734 solely to indicate this fact. Received

October

14, 1991; revised

December

4, 1991.

References Bliss, T. V. P., and Lynch, M. A. (1988). Long-term potentiation of synaptic transmission in the hippocampus: properties and mechanisms. In Long-Term Potentiation: From Biophysics to Behavior, P. W. Landfield and S. A. Deadwyler, ed. (New York: Alan R. Liss, Inc.), pp. 3-72. B6hme, G. A., Bon, C., Stutzmann, 1. M., Doble, A., and Blanchard, J. C. (1991). Possible involvement of nitric oxide in long-term potentiation. Eur. J. Pharmacol. 799, 379-381. Bredt, D. S., and Snyder, S. H. (1989). Nitric oxide mediates glutamate-linked enhancement of cGMP levels in the cerebellum. Proc. Natl. Acad. Sci. USA 86, 9030-9033. Bredt, D. S., Hwang, P. H., and Snyder, S. H. (1990). Localization of nitric oxide synthase indicating a neural role for nitric oxide. Nature 347, 768-770. arown, (1988).

T. H., Chapman, P. F., Kairiss, Long-term synaptic potentiation.

E. W., and Keenan, C. L. Science 242, 724-728.

Garthwaite, J. (1991). Glutamate, nitric oxide and cell-cell ling in the nervous system. Trends Neurosci. 74, 60-67.

signal-

Garthwaite, J., Charles, S. L., and Chess-Williams, R. (1988). Endothelium-derived relaxing factor release on activation of NMDA receptors suggests role as intracellular messenger in the brain. Nature 336, 385-388.

Neuron 216

Ignarro, L. J., Harbison, R. G., Wood, K. S., and Kadowitz, P. J. (1986). Activation of purified soluble guanylate cyclase by endothelium-derived relaxingfactorfrom intrapulmonaryartery and vein: stimulation by acetylcholine, bradykinin and arachidonic acid. J. Pharmacol. Exp. Ther. 237, 893-900. Kelso, S. R., and Brown, T. H. (1986). Differential conditioning of associative synaptic enhancement in hippocampal brain slices. Science 232, 85-87. Larson, J., Won& D., and Lynch, C. (1986). Patterned stimulation at the theta frequency is optimal for the induction of hippocampal long-term potentiation. Brain Res. 368, 347-350. Lovinger, D. M., Wong, K. L., Murakami, K., and Routtenberg, A. (1987). Protein kinase C inhibitors eliminate hippocampal long term potentiation. Brain Res. 436, 177-183. Madison, D. V., Malenka, R. C., and Nicoll, R. A. (1991). Mechanisms underlying long-term potentiation of synaptic transmission. Annu. Rev. Neurosci. 74, 379-397. Malenka, R. C. (1991). Postsynaptic factors control the duration of synaptic enhancement in area CA1 of the hippocampus. Neuron 6, 53-60. Malenka, R. C., Kauer, J. A., Zucker, R. S., and Nicoll, R. A. (1988). Postsynaptic calcium is sufficient for potentiation of hippocampal synaptic transmission. Science 242, 81-84. Malenka, R. C., Kauer, J. A., Perkel, D. J., Mau k, M. D., Kelly, P. T., Nicoll, R. A., and Waxham, M. N. (1989).An essential role for postsynapticcalmodulin and protein kinaseactivityin long-term potentiation. Nature 340, 554-557. Malinow, R., Madison, D. V., and Tsien, R. W. (1988). Persistent H-7-sensi;ive, sphingosine-insensitive protein kinaseactivity underlies long-term potentiation. Nature 335, 820-822. O’Dell, T. J., Hawkins, R. D., Kandel, E. R., and Arancio, 0. (1991). Tests on the roles of two diffusable substances in LTP: evidence for nitric oxide as a possible early retrograde messenger. Proc. Natl. Acad. Sci. USA 88, 11285-11289. Palmer, R. M. J., Ashton, D. S., and Moncada, endothelial cells synthesize nitric oxide from 333, 664-666.

S. (1989). Vascular r-arginine. Nature

Reymann, K. G., Frey, U., Jerk, R., and Matthies, H. (1988). Polymyxin B, an inhibitor of protein kinase C, prevents the maintenance of synaptic long-term potentiation in hippocampal CA1 neurons. Brain Res. 440, 305-314. Schuman, E. M., and Madison, intercellular messenger nitric Science 254, 1503-1506.

D. V. (1991). A requirement for the oxide in long-term potentiation.

Shibuki, K., and Okada, D. (1991). Endogenous lease required for long-term synaptic depression lum. Nature 349, 326-328. Snyder, S. H., and Bredt, D. S. (1991). Nitric oxide messenger. Trends Pharmacol. Sci. 72, 125-128. Southam,E.,and of nitric oxide

Garthwaite,J.(1991).Climbingfibresasasource in the cerebellum. Eur. J. Neurosci.

nitric oxide rein the cerebelas a neuronal

3, 379-382.

Tsien, R. W., Schulman, H., and Malinow, R. (1990). Peptide inhibitors of PKC and CaMK block induction but not expression of long-term potentiation. Biol. Med. Signal Trans. 24, 101-107.