each experiment, an input-output curve (afferent stimulusintensity versus EPSP amplitude) was plotted at the testfrequency. For all experiments, the amplitude of the testEPSP was adjusted to one-third of maximum, usually about1±1.2 mV. LTP/LTD was measured at 25±30 min postNMDA application. HFS-induced LTP was induced byeighttrainseachofeightstimuliat200Hz,intertraininterval0.2 s. NMDA was obtained from Sigma. Recordings wereanalyzed using p-Clamp. Values are the mean
^
SEM for
n
slices and two-tailed Student's
t
-test was used for statisticalcomparison.In previous experiments in CA1, perfusion of a relativelyhigh concentration (150
m
M) of NMDA for a brief timeinduced a large transient depression followed by a transientpotentiation [14,15]. Similar ®ndings were observed in thepresent study in the dentate gyrus using a high concentrationofNMDAforabrieftime.More recent workinCA1showedthat longer bath application of NMDA (20
m
M for 3 min)inducedaSTD,followedbyaLTD[10].Wethereforeinves-tigated the effect of a lower concentration of NMDA (10±20
m
M)perfusedforalongerperiod(5min)inthedentate.Intheinitial set of experiments, recordings were made from themedial perforant path. Bath perfusion of NMDA (10
m
M, 5min)causedasmall(1±5%)butnon-signi®cantdepressionof the EPSP during the actual perfusion, followed, upon wash-out, by an increasing enhancement of the EPSP over thesubsequent 10 min and then induction of a stable LTPmeasuring 21
^
5% (
n
6) (Fig. 1A). Repeat applicationsofNMDAinducedadditionalLTP,withtheNMDA-inducedLTPreachingamaximumof31
^
11%;
n
6afterthethirdapplication of NMDA (Fig. 1B). Such NMDA-induced LTPdid not occlude HFS induced LTP. Thus control HFS-inducedLTPmeasured50
^
6%;
n
7(Fig.1C).Followingmaximal induction of LTP by NMDA application, themagnitude of HFS-induced LTP was not altered, measuring58
^
7%;
n
6(Fig. 1B,C).Stimulationduring theapplica-tion of NMDA was not found to be necessary for the induc-tion of LTP. Thus cessation of test stimulation during and 15min after the perfusion of NMDA did not inhibit the induc-tion of LTP, which measured 19
^
3;
n
5 (Fig. 1D). In themedial perforant path, paired stimuli applied at 40 ms inter-pulse interval results in paired pulse depression of the EPSPof16
^
3%;
n
6.ThepairedpulsedepressionwasreducedinthepresenceofNMDA,to5
^
4%,butfollowingwashoutof NMDA, returned to the control level, i.e. no change inpaired pulse depression accompanied NMDA induced LTP(Fig. 1E).Recordings from the lateral perforant path showed thatapplication of NMDA had a very different action to that inthe medial perforant path. Thus perfusion of NMDA (10
m
M, 5 min) induced LTD of the test EPSP. The LTDinduced by a single application of NMDA was small andnot signi®cant. However, additional applications of NMDAproduced a summation of the LTD, and LTD became promi-nent, measuring 25
^
5%;
n
4, after the 2nd to 4th appli-cation (Fig. 2).The present study demonstrates that the plasticity evokedby NMDA has a regional speci®city. In the medial perforantpath of the dentate gyrus, NMDA induces LTP, whereas inthe lateral perforant path, NMDA induces LTD. The ®ndingthatNMDAapplicationcaninduceLTPinthemedialperfor-antpathisthe®rstsuchobservationofNMDA-inducedLTP.PreviousstudieshaveshownthatNMDAapplicationinduceseither only a short-term potentiation when relatively highconcentrations ofNMDAwereappliedforabrieftimeeither[6,9,14,15]orLTDwhenlowconcentrationsofNMDAwereapplied for a longer time [10]. The induction of LTP byNMDA in the medial perforant path in the present study,rather than short-term potentiation, is likely to be due to theapplication of NMDA at a low concentration and for a rela-tively prolonged period. Interestingly, the LTP induced byexogenous NMDA application did not occlude with HFS-inducedLTP.ThiscouldbeduetodifferentsitesofinductionoftheLTPinducedbyexogenousNMDAapplication andbyHFS. For example, perhaps exogenous NMDA applicationresults in a presynaptic site for LTP induction via activationof presynaptic NMDAR [5] or NMDA receptors located onastrocytes, as suggested by Araque et al. [1], in contrast toHFS inducing LTP at a post-synaptic site. As no measurablechanges in paired-pulse depression accompanied the LTPinduced by exogenous NMDA application, such LTP, if presynaptic, would involve a change in the number of sitesof release rather than a change in release probability. Thedetailed mechanisms of induction of the LTP induced byexogenous NMDA are presently under investigation. TheinductionofLTDbyNMDAapplicationinthelateralperfor-ant path is identical to that found in the study [10] in CA1, inwhich a similar concentration and duration of application of NMDA to the present study was used.It is well established that Ca
2
1
in¯ux via NMDA receptor
A.M. Rush et al. / Neuroscience Letters 298 (2001) 175±178
177
Fig. 2. NMDA induces LTD in the lateral perforant path of thedentate gyrus in vitro. An initial application of NMDA by bathperfusion did not induce LTD, but further applications inducedsigni®cantLTD measuring25
^
5,
n
4, after the 4th applicationof NMDA.
Leave a Comment