J Ph~siohgy
OElsevier,
( 1996) 90, 343.-347
(Ptrrisl
Paris
Modulation of synaptic plasticity in the hippocampus and piriform cortex by physiologically meaningful olfactory cues in an olfactory association task FA Chaillan, FS Roman*, B Soumireu-Mourat
Summary - Animals were trained to discriminate two natural odors while another group was trained to discriminate between a patterned electrical stimulation distributed on the lateral olfactory tract (LOT), labelled olfaco-mimetic stimulation (OMS). used as an olfactory cue ver.su.s a natural odor. No statistically significant difference was observed in behavioral data between these two groups. The animals trained to learn the meaning of the OMS exhibited a gradual long-term potentiation (LTP) phenomenon in the piriform cortex. When a group of naive animals was pseudo-conditioned, giving the OMS for the same number of sessions but without any olfactory training, no LTP was recorded. These results indicate that the process of learning olfactory association gradually potentiates cortical synapses in a defined cortical terminal field. and may explain why LTP in the piriform cortex is not elicited by the patterned stimulation itself, but only in an associative context. As olfactory and hippocampus regions are connected tlicl the lateral entorhinal cortex. the olfactomimetic model was used to study the dynamic of involvement of the dentate gyrus (DG) in learning and memory of this associative olfactory task. Polysynaptic field potentials, evoked by the LOT stimulation, were recorded in the molecular layer of the ipsilateral DG. An early and rapid (2nd session) potentiation was observed when a significant discrimination of the two cues began to be observed. The onset latency of the potentiated response was 304 ms. When a group of naive animals was pseudoconditioned, no change was observed. Taken together. these results support the hypothesis that early activation of the DC during the learning of olfactory cue allows the progressive storage of olfactory information in a defined set of potentiated cortical synapses. The onset latency of the polysynaptic potentiated responses suggests the existence of a reactivating hippocampal loops during the processing of olfactory information. odor discrimination
/ learning I long-term potentiation
(LTP) / piriform cortex I dentate gyros I behavior
Introduction Learning and memory are considered to be the result of distributed neural network reorganizations following modifications in synaptic strength (Bloch and Laroche, 1984). The phenomenon of long-term potentiation (LTP) is a likely candidate for a cellular modification postulated by different theories to conserve a stable memory trace. LTP is a long-term enhancement of a monosynaptic excitatory transmission elicited by short, high frequency stimulation of the afferent pathway (Bliss and Gardner-Medwin, 1973; Bliss and Lomo, 1973). Although numerous in vitro and in vivo studies have reported LTP in a variety of central nervous structures, data relating a direct link between LTP and mnesic processes are scarce. In our laboratory, we are attempting to find data to link learning and memory processes with neural plasticity, like the LTP phenomenon in the brain (Soumireu-Mourat and Roman, 1991; Roman et al, 1993a, 1994). To do so, we chose to train rats using an olfactory associative task which has several interesting behavioral, anatomical and electrophysiological features. Briefly, rats can perform an olfactory task as well
*Correspondence
and reprints
as primates trained with visual cues (Slotnick and Katz, 1974; Slotnick and Kaneko, 1981) giving us a tool to dissociate different kinds of memory (Eichenbaum et ul, 1986; Roman et al, 1993b). The anatomy of the olfactory system is relatively simple. There is only one synapse between the olfactory bulb (OB) and the piriform (Pir Cx) and between the OB and the entorhinal (Ent Cx) cortices twitter et al, 1991). These structures are connected to the hippocampal system known to participate in the learning and memory processes (Scoville and Milner, 1957). Olfactory information reached the dentate gyms (DG) specifically via the lateral entorhinal cortex (Lat Ent Cx) and the lateral perforant path (LPP) (Witter ef al, 1991). Moreover, the Pir Cx is much simpler than the neocortex, and the projections from the OB generate a dense but restricted terminal field in a cell-sparse dendritic layer, a situation that is ideal for extracellular recording of monosynaptic responses. Finally, the electrical stimulation of the lateral olfactory tract (LOT) can evoke a monosynaptic and a polysynaptic field potential, in the Pir Cx and the DG respectively (Haberly and Price. 1977; Wilson and Steward, 1978). The goal of the following combined electrophysiological and behavioral experiments was two-fold. Firstly, it was to specify the kinetics of the involvement of the Pir Cx and the hippocampus, studying the DC during the learning and the memorization of an olfactory associative task. Secondly, it was to determine what
344
kind of synaptic plasticity vation of these structures.
is involved
during
the acti-
Materials and methods All rats were implanted with two bipolar stainless steel stimulating electrodes in the LOT. One electrode was used to apply the patterned stimulation during the training sessions and was labelled the active electrode (AE). The other was used as a control electrode (CE) to apply only single test pulses, before and after each session. One group of animals was implanted with a monopolar platinum recording electrode in the ipsilateral Layer I of the Pir Cx to record the monosynaptic field potential elicited by the LOT stimulation. Another group was implanted with the same kind of recording electrode in the ipsilateral granular cell layer of the DG to record the polysynaptic field potential elicited by the LOT stimulation. The discrimination training was an associative Go-No Go learning task in which animals learned to associate a correct odor with a reward (water). Water-deprived rats were trained to discriminate two natural odors, one associated with water and the other with a non-aversive light, respectively called the positive (0+) and negative (0-) odors. The rat had to respond for 0+ and not for 0-. For the electrophysiological study, animals were trained to discriminate between a patterned electrical stimulation and a natural odor. The paradigm was the same as above except that the positive odor was replaced by a patterned electrical stimulation on the LOT (36 ms burst of four pulses (at 100 S-l ) delivered at an interburst interval of 160 ms). The patterned stimulation induced strong sniffing in rats and was called olfactomimetic stimulations (OMS). The animals were trained once a day for 5 days in a 60 random trial session. One group of animals was trained to discriminate between two natural odors (control group). Two groups were trained to discriminate between the patterned electrical stimulation and the natural odor, one with the recording electrode positioned in the Pir Cx and the other in the DG. To test the possible effects of the OMS by itself (ie without behavioral training), two groups were pseudoconditioned. One group of naive animals with implants in the Pir Cx and the other with implants in the DG were given the same amount of experience (five sessions) with the patterned electrical stimulation of the LOT. natural odor, water and light rewards with no association.
Results The behavioral data obtained for the two-odor discrimination (control group) showed that the rats improved their performance across sessions (fig IA) reaching a correct response level of more than 80% on session 5. Significant discrimination was observed starting on the third session (66.73 f 4.92%; P < 0.05 compared to
A
-_
5
80
8 ww
70
$3
60
y
so
4
40 1
” ‘:T lO.O-
B
s
7.5-
f ; 0
5.0-
w' !i
2s-
0.0-j 0
1
I
I
8
I
1
2
3
4
5
SESSION (day) Fig 1. Behavioral performance of rats trained to discriminate between two natural odors. Each bar represents the mean and the SEM of the Hkial daily session. Animals were trained for 5 days. A. Mean correct response rate for positive (O+) and negative (&) odors. B. Response latency curve across the five training sessions. The latency gradually decreased for 0+ and gradually increased for &.
sessions 1 (chance level) and 3: Mann-Whitney U-test, two-tailed). Response latencies for O+ and 0- (fig 1B) were significantly different across the five sessions (MANOVA, F(4, 48) = 20.75; P < 0.001). Across sessions, response latencies decreased for 0+ and increased for 0- because animals gradually learned to respond for 0+ and not to respond for 0-. Comparison between the control group and the Pir Cx group showed no statistical differences (MANOVA, F(4,48) = 1.56; ns). The changes in the field potentials induced by the patterned electrical stimulation used as a discriminative cue are reported in figure 2A for the Pir Cx group. A statistically significant change in field potential slope evoked by AE was present prior to the 4th session (P
345
A w n
s fn +
30-
2O-
ICE -AE
T T
T
lo-
T_
kY fi
O
5
-lO-
s
-2oj 1
2
3
4
5
4
5
2B) exhibited a long-term depression that was statistically significant across the five sessions (P < 0.01; Mann-Whitney U-test, two-tailed). Recordings in the DG (fig 3A) showed no statistically significant changes in the polysynaptic field potentials of the DG after the 1st and before the 2nd training sessions compared to the baseline (before the 1st session). Just after the 2nd session, electrophysiological recordings indicated a statistically significant (P < 0.01; Mann-Whitney U-test, two-tailed) increase in the slope of the polysynaptic field potential evoked by the AE. Although not as large, this increase was still statistically significant before and after all of the fol-
B
i!
0
cj
-10
v) g
-20
g
-30
2 5
-40
$
-50
-60 ’
I
2
3
SESSION (day) Fig 2. Change in field potentials
recorded in the pirifotm cortex before and after each session. Each bar represents the average value of the slopes of 20 evoked field potentials. A. Slope for the trained animals. The field potentials evoked by the active electrode (AE) increased across sessions. The field potentials of the control electrode (CE) exhibited a slight decrease. B. Slope for the pseudo-conditioned animals. Field potentials evoked by both AE and CE exhibited a decrease across the five training sessions. The pseudo-conditioned animals received patterned electrical stimulation, natural odor, water. and light without any association. No statistically significant increase or decrease in the slope of either AE or CE was observed.
B
s g
2 0 I
0.10
1
T
T
*
T
T T
o.os-
o.oo-
< 0.05; compare the responses collected prior to the 4th session with the baseline collected prior to the 1st session; Mann-Whitney U-test, two-tailed).This increase was maintained after the end of the session (P < 0.05) and still remained 24 h later, prior to and after the 5th session (P < 0.001). No statistically significant change was observed in the slope of the field potential evoked by CE. A highly significant correlation was found between the change of the slope of AE and the response latency differences across sessions (n = 91; r = 0.59; P < 0.001). The field potentials evoked by AE and CE of the Pir Cx pseudo-conditioned group (fig
0
I
2
3
4
SESSION
Fig 3. Slope of the polysynaptic
5
(day)
field potential evoked by the active electrode (AE) and the control electrode (CE) during the daily training sessions. Each point is the average slope of 20 single polysynaptic responses, before and after each session. The value of the slope before the 1st session was used as the baseline. A. Polysynaptic field potential curve for animals trained to discriminate the OMS versus natural odor. B. Polysynaptic field potential curve for pseudo-conditioned animals. These animals received patterned electrical stimulation, natural odor, water, and light without any association. No statistically significant increase or decrease in the slope of either AE or CE was observed.
346
lowing training sessions (P < 0.05). The onset latency of the polysynaptic field potential which exhibits a potentiation was between 3040 ms. The peak amplitude latency was around 59 ms. The slope of the polysynaptic field potential evoked by CE decreased slightly during the successive learning sessions. The only two statistically significant decreases were after session I and before session 5 compared to the baseline before the 1st session (P < 0.05). For AE and CE in the pseudo-conditioned group, there was a non-statistically significant trend toward a decrease in the slope of the polysynaptic field potential across the five sessions (fig 3B). The only significant correlation between the change of the slope of the AE and the response latency differences was during session 2 (r = 0.57; ?z = 20; P < 0.01).
Discussion and conclusion We used an olfactory discrimination task with OMS in order to find out what kind of neural plasticity was induced in different brain structures, what kinetics were involved, and how these events might give rise to longterm memory. Previous results showed that when rats were trained on a series of two-natural odor discriminations, an immediate and extremely stable LTP was induced as soon as one of the natural odors was replaced by OMS (Roman et al, 1987). Here the results showed that a gradual long-term potentiation phenomenon was induced in the Pir Cx while rats were learning the meaning of the OMS without previous training. When a group of naive animals was pseudo-conditioned, subjected to OMS for the same number of sessions without any training, no LTP was recorded. These results indicate that the process of learning an olfactory association gradually potentiates cortical synapses in a defined cortical field (fig 4A), and may explain why LTP in the Pir Cx is not elicited by the patterned stimulation by itself, but only in an associative context. The DG recordings showed that trained rats exhibited a potentiation of a polysynaptic field potential elicited by the active electrode used to apply the OMS. This phenomenon was correlated with the improvement in the animal’s performance only during session 2, suggesting early activation of the DG during the association learning process. Moreover, the 30-40 ms onset latency of the potentiated polysynaptic field suggests the existence of a reactivating hippocampal loop. No electrophysiological change for the active and control electrodes was observed in the DG of the pseudo-conditioned animals, suggesting that the polysynaptic potentiation was specifically linked to the learning con-
Fig 4. Examples of a change in the field potentials evoked in the piriform cortex and dentate gyrus induced by patterned electrical stimulation of the lateral olfactory tract. A. Monosynaptic field potential evoked by the active electrode recorded in the piriform cortex before (b) and after (a) sessions 1, 3, and 5 61. S3. S5). The solid line is the baseline and the dotted line represents the daily recording. B. Polysynaptic field potential evoked by the active electrode recorded in the dentate gyms before and after sessions I, 2, and 5 61, S2. S5).
text (fig 4B). De Curtis et al (1991) demonstrated the existence of ‘reverberant activation of the entorhinal (cortex)-hippocampal-entorhinal (cortex) circuit following a single electrical stimulation of the LOT’, in vitro. Their data indicated that LOT stimulation ‘can produce a sequential activation that arises in the entorhinal cortex and reenters into this cortex after having traveled throughout the extent of the hippocampal circuit’. According to this hypothesis, a similar kind of sequential activation to that reentering the entorhinal cortex could also reenter the DG. Taken together these data provide evidence of the differential temporal activation of the limbic structures and cortex during learning and memory processes. These results support the hypothesis that early dynamic activation of the DG during the learning of an olfactory cue, which is reflected by the observed polysynaptic
347
potentiation in the DG, allows for the gradual storage of olfactory information in a defined set of potentiated cortical synapses, as seen on the gradual LTP recordings in the Pir Cx. An explanation which is consistent with the literature on learning and memory (Squire, 1987) would be that the limbic circuits related to the olfactory system play an important role in the active suppression of target cells in the Pir Cx (Hasselmo and Bower, 1991; Barkai and Hasselmo, 1994), which normally allow specific modifications of the synapses supporting the odor-reward association. The hippocampus may influence the synaptic transmission and storage of olfactory information in different structures via the diagonal band of Broca (Kunze et al, 1991; Roman et al, 199313). In conclusion, these data suggest an association between hippocampal functions (the association between discrimination cues and reward) and the transfer and storage of relevant information in a specific cortical neural network located at least in the Pir Cx.
References Barkai E, Hasselmo ME (1994) Modulation of input/output function of rat piriform cortex pyramidal cells. J Neumphysiol 72, 644 658 Bliss TVP, Gardner-Medwin AR (1973) Long-lasting potentiation of synaptic transmission in the dentate area of unaesthetized rabbit following stimulation of the perforant path. J Ph,vsio[ 232, 357-374 Bliss TVP, Lomo T (1973) Long-lasting potentiation of synaptic transmission in the dentate area of the unaesthetized rabbit following stimulation of the perforant path. J Physiol232, 331-356 Bloch V, Laroche S (1984) Facts and hypotheses related to search for engram. In:Neurobiology of learning and memoq Guilford, New York, 249-260 De Curtis M, Pare D, Llinas RR (1991) The elctrophysiology of the olfactory-hippocampal circuit in the isolated and perfused adult mammalian brain in vitro. Hippocampus 1, 341-354
Eichenbaum H, Fagan A. Cohen NJ (1986) Normal olfactory discrimination learning set and facilitation of reversal learning after medial-temporal damage in rats: implications for an account of preserved learning abilities in amnesia. J Neurosci 6, 1876-1884 Haberly LB, Price JL (1977) The axonal projection patterns of the mitral and tufted cells of the olfactory bulb in the rat. Brain Res 129, 152-157 Hasselmo ME, Bower JM (1991) Selective suppression of afferent but not intrinsic fiber synaptictransmission by 2-amino-4-phosphonobutyric acid (AP4) in piriform cortex. Brain Res 548,248255 Kunze WWA, Shafton AD. Kemm RE, McKenzie JS (1991) Effects of stimulating the nucleus of the horizontal limb of the diagonal band on single unit activity in the olfactory bulb. Neuroscience 40, 21-27 Roman FS, Staubli U, Lynch G (1987) Evidence for synaptic potentiation in a cortical network during learning. Brain Res 418, 22 1-226 Roman FS, Chaillan FA, Soumireu-Mourat B (1993a) Long-term potentiation in rat piriform cortex following discrimination leaming. Brain Res 601. 265-272 Roman FS, Simonetto I. Soumireu-Mourat B (1993b) Learning and memory of odor-reward association: selective impairment following horizontal diagonal band lesions. Behav Neurosci 107, 72-8 1 Roman FS, Chaillan FA, Soumireu-Mourat B (1994) Long-term potentiation and long-term depression in piriform cortex in viva. In: Long-term potentiation: a debate of current issues (Baudry M, Davis JL, eds), vol 2, 291-304 Scoville WB, Milner B (1957) Loss of recent memory after bilateral hippocampal lesions. .I Neural Neurosurg fsychiatr 20. 1 l-21 Slotnick BM, Katz HM (1974) Olfactory learning set formation in rats. Science 185, 796798 Slotnick BM, Kaneko N (1981) Role of dorsomedial thalamic nucleus in olfactory discrimination in rats. Science 214, 91-92 Soumireu-Mourat B. Roman FS (1991) Long-term potentiation and learning and memory. In: Long-term potentiation: a debate of current issues (Baudry M, Davis JL, eds) vol 1. 317-323 Squire LR ( 1987) Memon; and Brain. Oxford University Press, New York. 315 p Wilson RC. Steward 0 (1978) Polysynaptic activation of the dentate gyms of the hippocampal formation: an olfactory input via the entorhinal cortex. Exp Brain Res 33, 523-534 Witter MP, Groenewegen HJ, Lopes Da Silva FH, Lohman AHM (1991) Functional organization of the extrinsic an intrinsic circuitry of the parahippocampal region. Prog Neurobioi 33, 161253