Frequency potentiation in granule cells in vivo at θ frequency perforant path stimulation

EXPERIMENTAL

113,74-78

NEUROLOGY

(1991)

Frequency Potentiation in Granule Cells in Viva at 0 Frequency Perforant Path Stimulation MAR~ADOLORES Neurologicr

Mmoz,

Experimental, Departamento de Morfologia, Facultad

ANGELNWEZ,*ANDELIOGARC~A-AUS?T de Investigacio’n, Hospital “Ram& de Medicina, Universidad Auto’rwma,

MATERIALS

AND METHODS

Transmembrane potentials were recorded in 10 urethane-anesthetized (0.7 g/kg, ip) and -curarized (1-2 mg/kg, ip) Sprague-Dawley rats, weighing 200-300 g. The experimental methodology followed has been previously described (20). Briefly, trephine holes were drilled in the skull at preselected stereotaxic coordinates (13), a macroelectrode was lowered into the fascia dentata (A: 4.6, L: 2.6, H: 3) to record the field activity, and an indifferent electrode was placed in the frontal lobe. Bipolar stimulating electrodes were aimed at an angle of 30” and placed in the angular bundle to stimulate the perforant pathway (A: -1.9, L: 4.1, H: 3.5) and in the mossy fibers for antidromic stimulation (A: 1.4, L: 2.4, H: 3.5). The stimulation parameters consisted in square pulses of 0.3 ms duration and 50-100 PA intensity delivered at 0.4 Hz. This stimulation was followed by a ~-HZ or lo-Hz stimulation train during lo-60 s (test stimulation). In other cases the stimulation frequency was increased in steps from 0.4 to 4 Hz. A postcontrol stimulation of 0.4 Hz was always applied after the stimulation test. Intracellular recordings were performed with micropipets filled with 3 M potassium acetate or 3 A4 KC1 (40-80 MQ resistance). The micropipets were placed on the cerebral cortex and then were lowered in micrometer steps. When they were above the hippocampal fissure according to stereotaxic coordinates, stimulation started. The micropipet recorded the field potential generated by perforant path stimulation. The hippocampal anatomical organization determines a well-known laminar field potential profile which reflects the position of the micropipet within the hippocampus. When the granule neuron was impaled, a hyperpolarizing current was injected through the micropipet to compensate for the depolarization evoked by the penetration; the current injection was slowly removed.

INTRODUCTION Repetitive stimulation of hippocampal afferents has been found to increase synaptic efficacy in eliciting orthodromic spikes in pyramidal cells (2,3,8,9, 18). This result has often been termed frequency potentiation and has also been reported in the peripheral (12,22) and central nervous systems (7,14). Another case of potentiation is long-term potentiation (LTP) which is manifested as a long-lasting increase in the magnitude of the postsynaptic response due to tetanic stimulation ((5,6), for review see (24)). Stimulation frequencies used in LTP are generally above 50 Hz and are applied during a short period. However, stimulation trains at theta (0) frequency elicit a more stable LTP (11,15). This stimulation pattern resembles the rhythmic inputs generated by the entorhinal cortical neurons (1) which reach the hippocampal region through the perforant path, and suggests that fl rhythm may play a modulatory role in synaptic efficacy (11). The present series of in viva recordings was undertaken in order to determine changes in the excitatory or inhibitory postsynaptic potentials (EPSPs and IPSPs, 74 Inc. reserved.

and *Departamento Spain

respectively) evoked in the granule cells while stimulating the perforant path at fl frequency. The results indicate that synaptic efficacy increases fundamentally when the stimulation frequency is in the fl frequency band (between 4.0 and 8.0 Hz).

The effect of frequency potentiation on the postsynaptic potential in granule cells was studied stimulating the perforant path in curarized and urethanized rats. At stimulation frequencies between 2.0-6.0 Hz, synaptic efficacy in eliciting an orthodromic action potential increased despite the hyperpolarization of the transmembrane potential. The excitatory postsynaptic potential (EPSP) slope and duration also increased while the inhibitory postsynaptic potential (IPSP) was reduced. Stimulation frequencies greater than 6.0 Hz produced similar changes in the transmembrane potential and EPSP-IPSP sequence, but they did not increase synaptic efficacy. The frequency potentiation at frequencies into the 0 band suggest that this potentiation participates in 0 rhythm genesis in this structure which, in turn, suggests that the fascia dentata could work as a band-pass filter. In 12.6% of cases postpotenktiOn WRS ak30 observed. 0 1991 Academic PAM, IN.

0014-4336191 $3.00 Copyright 0 1991 by Academic Press, All rights of reproduction in any form

y Cajal;” Madrid,

FREQUENCY

POTENTIATION

When the recording neuron was controlled, the experiment was begun. Data were stored in FM tape and analyzed off-line. Input resistance was determined according to the voltage drop during injections of hyperpolarizing current pulses. Neurons were included for analysis only if they exhibited resting potentials greater than -50 mV with overshooting spikes. Postsynaptic potentials were analyzed off-line in a digital computer in order to measure the IPSP area and to calculate the EPSP slope (see below). At least 30 s of continuously recorded data were fed into the computer and stored at an 8-kHz sampling frequency; postsynaptic potential averages were also calculated1 The EPSP slope was calculated by linear interpolation. The EPSP duration was measured, assigning zero voltage as the resting potential level and zero time as the EPSP latency. The EPSP ended when the voltage crossed zero voltage again. To establish the statistical significance between the two averaged records and classification of neurons the Wilcoxon test and Cluster test were used, respectively. RESULTS

Sixteen of the 35 neurons recorded in the upper and lower blades of the fascia dentata were identified as granule cells (GC) by antidromic driving from mossy fiber stimulation. The remaining neurons presented the same functional characteristics as those identified as GCs and were therefore included in the same population. Neurons showed a resting potential between -50 and -68 mV (mean -58.3 + SEM 2.8 mV) and an input resistance of 30.3 & 2.7 MO. Spontaneous spikes had steep rising and falling slopes, with a mean duration of 1.1 ~fr0.1 ms at the halfway point between the firing and peak levels. Peak amplitude was 64.2 -t 2.3 mV, and the spontaneous firing rate was 7.1 f 0.6 Hz. According to membrane properties the neurons belonged to the same class (Cluster test was used); according to the discharge patterns, three types of neurons were found and classified in the presence of continuous B rhythm in the fascia dentate field activity (19). The existence of stable membrane potentials and overshooting spikes indicates that in vivo intracellular recordings were made in healthy neurons. When the perforant path was stimulated at a frequency of 0.4 Hz the GCs always displayed an EPSPIPSP sequence (Fig. IA, arrows 1 and 4; Fig. lB, 1 and4) and an orthodromic spike with a mean latency of 3.8 Z&0.5 ms was evoked in 40% of the neurons (see also (19)). When control stimulation (0.4 Hz) was followed by frequency stimulation at 4.0 Hz the transmembrane potential immediately became hyperpolarized (Fig. lA, between 2 and 3; Fig. lB, 2 and 3). Nevertheless, the probability of evoking an action potential through perforant path stimulation increased to 100%. At 4-O-Hz

IN

IN VZVO GRANULE

CELLS

75

stimulation the EPSP slope increased significatively (P < 0.05) in 46.7% of neurons. The neurons that increased their EPSP slope simultaneously augmented the EPSP duration (P < 0.01) in 57% of cases. Of all neurons recorded, 73% showed a significant increase of EPSP duration (P < 0.01) at 4.0 Hz stimulation. The membrane potential returned to the resting level when the 4.0-Hz frequency stimulation ceased or when it was restimulated at 0.4 Hz, and the rate of orthodromic spike generation was similar to that seen with control stimuli (Fig. lA, arrow 4; Fig. lB, 4). However, after 30 s of stimulation at 4.0 Hz, synaptic postpotentiation was observed in some cases (12.5%) during stimulation frequencies of 0.4 Hz. This postpotentiation consisted in the appearance, some 60-80 s after (ending) the 4.0 Hz stimulation, of two or three action potentials over the EPSP (Fig. XC). This effect lasted approximately 60 s. In other experiments the stimulation frequency was gradually increased from 0.4 to 10.0 Hz. As can be observed in Fig. 2, the intensity of electric stimulation at 0.4 Hz was adjusted to just below the orthodromic spike threshold and remained unaltered throughout the experiment. Perforant path stimulation with increasing frequencies evoked orthodromic spikes in 60% of the neurons with a stimulation frequency of 1.0 Hz and in 100% of the neurons when stimulation reached 2.0 or 3.0 Hz. Sometimes a second spike was elicited when stimulation reached 4.0 Hz. Repetitive stimulation at frequencies greater than or equal to 2.0 Hz produced membrane hyperpolarization (7 mV at 2.0 Hz and 15 mV at 4.0 Hz). This change in the transmembrane potential did not affect the increase in synaptic efficacy. Synaptic efficacy remained unchanged when the membrane potential was carried to the resting level or to more depolarized levels through the injection of depolarizing currents (not shown). Figure 2 also shows that the slope and duration of the EPSP increased and it was difficult to measure the EPSP amplitude, because the spikes overlapped. A decrease in IPSP was also observed (see below). When stimulating the perforant path at frequencies equal to or above 5.0 Hz synaptic efficacy did not increase and the membrane potential was more hyperpolarized (30 mV) than with earlier stimulations at lower frequencies as may be observed in the recording of the stimulation at 10.0 Hz (Fig. 2). The resting level was recovered when stimulation returned to 0.4 Hz. The evolution of two IPSP periods during repetitive stimulation was analyzed by calculating the area corresponding to the first 40 ms after the end of the EPSP (termed initial IPSP period) and the area corresponding to the following 90 ms (late IPSP period). This classification was made according to the different ionic mechanism described previously (19). Figure 3 shows that the area of the initial (A) and late (B) IPSP periods decreased significantly with stimulation frequencies greater than or equal to 2.0 Hz (P < 0.001). The IPSP

76

MUfiOZ,

NtifiEZ,

AND

GARCIA-AUSTT

c

FIG. 1. Effect on the GCs of perforant path stimulation at 4.0 Hz. (A) An example of intracellular recordings during perforant path stimulation {arrows). (B) Two superimposed traces of the synaptic potential evoked by perforant path stimulation under different conditions: (i) at a stimulation frequency of 0.4 Hz (only the last stimulation is shown); (ii) at the beginning of the 4.0-Hz stimulation; (iii) after 20 s of the 4.0-Hz stimulation; and (iv) 5 s after the end of the ~-HZ stimulation. During the ~-HZ stimulation the transmembrane potential is hyperpolarized and synaptic efficacy increases. After stimulation, the synaptic and transmembrane potential returns to the previous level. (C) The effect of the perforant path stimulation at 0.4 Hz after a train of stimuli at 4.0 Hz; a potentiation on the response is observed. Dotted lines indicate the resting membrane potential level.

area remained reduced when stimulus frequency returned to 0.4 Hz during the total time the neuron was impaled. In order to study the influence that decreasing the IPSP amplitude had on enhancing synaptic efficacy, during repetitive stimulation of the perforant path, the initial or late IPSP periods were abolished with intracellular diffusions of Cl- or Cs, respectively, as has previously been described (19). Synaptic efficacy increased and IPSP was reversed after 10 s of Cl- diffusion. Cs+ diffusion did not produce any change in the orthodromic firing of GCs (not shown). Hyperpolarizing current pulses were applied after the orthodromic spikes so that the changes in postsynaptic potential conductance during stimulation at frequencies between 0.4 and 4.0 Hz (Fig. 4) could be determined. The control voltage drop values measured at the resting membrane potential before stimulation were considered as 100%. The input resistance decreased when stimulation was at 0.4 Hz. Probably due to EPSP, maximum values were seen approximately 10 ms after stimulus, and they returned to control levels lo-20 ms later. When the frequency increased to 2.0,3.0, or 4.0 Hz, in-

put resistance remained decreased up to 40, 60, or 90 ms, respectively. The minimum value of input resistance also lasted longer when the stimulation frequency was increased (about 20 ms). DISCUSSION

As described in Results, repetitive electric stimulation in the perforant pathway evoked changes in the membrane potential and in the postsynaptic potential efficacy of GCs. The membrane potential hyperpolarization that we observed during repetitive stimulation has also been described in in vitro rat (21) and in in uiuo rabbit (3) recordings of pyramidal cells at a lO.O-Hz stimulation frequency. This membrane potential hyperpolarization was not related to the increase in synaptic efiicacy, because synaptic transmission remained increased when the membrane potential was depolarized. This effect may be due to the higher excitation of the inhibitory feedback, to a direct effect on the inhibitory interneurons, and/or to long-lasting changes in K+ conductances. The increase in synaptic efficacy was observed at

FREQUENCY

POTENTIATION

IA VZVO GRANULE

77

CELLS

Two different factors may contribute to the synaptic efficacy increase at stimulation frequencies above 1.0 Hz: (i) The slope and duration of EPSP increase and (ii) the amplitude of both IPSP periods decreases. The changes in the EPSP may be due to a larger excitatory input and/or to the reverse of the initial IPSP component. A resistance decrease of approximately 100% during EPSP and around 30% during IPSP has been described (19). We found that increasing the stimulating frequency prolonged the decrease in input resistance by roughly lOO%, suggesting that the EPSP lasts longer (see Fig. 4). Periods in which the input resistance decrements were about 30% also lasted longer, indicating that the IPSP, although reduced in amplitude, was more sustained. The increase in the EPSP with a maximum enhancement of conductance could be explained by the opening of new channels due to the repetitive stimulation. Many authors studying LTP generation have proposed this as a possible mechanism for increasing EPSP (17). The decrease in IPSP amplitude was not due to sustained hyperpolarization because it did not change after membrane depolarization. The reduction in IPSP may contribute to the increase in orthodromic firing, as has been demonstrated during Cl- intracellular diffusion

‘..... ............. IC

10Hz

t

IN

5%

FIG. 2. Effect of stimulation frequency increase on the GCs. The perforant path is stimulated at frequencies of 0.4,1.0,2.0,3.0,4.0, and 10.0 Hz, as indicated in each recording. Three traces were superimposed. The intensity of the electric stimulation stimulus was adjusted to just below the orthodromic spike threshold at 0.4 Hz. Stimulation frequencies equal to or above 1.0 Hz elicit orthodromic firing and the discharge probability increases with the stimulation frequency as does the transmembrane potential hyperpolarization. A stimulation frequency of 10.0 Hz abolishes orthodromic spikes and the membrane potential hyperpolarization reverses the IPSP. Dotted lines indicate the resting membrane potential level.

stimulating frequencies that fundamentally include the 13band. Similar results have been described with extracellular recordings (see (4) for review). This GC activation pattern also occurs during periods of natural fl frequency stimulation due to the rhythmic discharges of the entorhinal neurons (1) that elicit a rhythmic input through the perforant path. Synaptic efficacy must therefore increase during 0 periods favoring GC firing. Many authors have reported that LTP is better elicited when using 8 frequency stimulation (11, 15, 16,23) suggesting that the 6 rhythm plays a modulatory role in LTP generation (11). This mechanism is surely the same one we are describing. These results suggest that the fascia dentata could work as a band filter that is fundamentally sensitive to determinate frequency inputs, or in other words, it specifically processes information which arrives modulated at this frequency. Perhaps the size of the sustained hyperpolarization at frequencies above 5.0 Hz prevents spike firing.

0.4

1

2

3

4

0.4 Hz

2

3

4

0.4 Hz

120 1 B

l-

0.4

FIG. 3.

1

Changes in the area of the initial (A) and late (B) periods of IPSP during the frequency increase in perforant path stimulation. The 100% corresponded to the area of the initial or late period of IPSP at a stimulation frequency of 0.4 Hz. The IPSP area decreases at frequencies above 1.0 Hz. When the stimulation frequency returns 0.4 Hz, the area of the initial (A) and the late (B) period of IPSP remain diminished (*P < 0.001).

78

MUfiOZ,

NfiTr;lEZ,

AND

GARCIA-AUSTT

100

6.

1

7. l

0.4

Hz

.2

Hz q 3Hz

8.

r4Hz

9. 10

30

50

70

90ms

FIG. 4. Changes in the input resistance after perforant path stimulation. Input resistance was tested by the injection of constant short hyperpolarizing pulses. Abscissae, time after stimulation. Ordinate, percentage voltage drop. The voltage drop in the hyperpolarizing pulses before stimulation is considered as 100%. The synaptic stimulation evoked an input resistance decrease which lasted longer when the stimulation frequency was increased. Only the last percentage voltage values at 3 and 4 Hz are shown.

10. 11. 12. 13.

which reverts the initial component of the IPSP to depolarizing. levels and increases synaptic efficacy (19). Postpotentiation responses after stimulation frequencies of 4.0 Hz were only observed in 12.5% of the neurons. The fact that only a few neurons were potentiated could be due to the low frequency single pulse stimulation since higher stimulation frequencies or stimuli trains are normally used to elicit potentiation (for review see (24)). On the other hand, some of the literature states that not all neurons are capable of developing postpotentiation (10, 25). ACKNOWLEDGMENT We thank Ms. C. F. Warren for her linguistic assistance.

14. 15. 16.

17. 18. 19.

REFERENCES 1. ALONSO, A., AND E. GARCIA-Aus~. 1987. Neuronal sources of theta rhythm in the entorhinal cortex of the rat. I. Laminar distribution of theta field potential. Exp. Brain Res. 67: 493501. 2. ANDERSEN, P. 1960. Interhippocampal impulses. II. Apical dendritic activation of CA1 neurons. Actu Physiol. &and. 48: 178208. 3. ANDERSEN, P., AND T. LOMO. 1967. Controlof hippocampal output by afferent volley frequency. In Structure rind Function of the Limbic System (W. R. Adey and T. Tokizane, Eds.), pp. 400412. Elsevier, Amsterdam/New York. 4. BLAND, B. H. 1986. The physiology and pharmacology of hippocampal formation theta rhythms. Prog. Neurobiol. 26: l-54. 5. BLISS, T. V. P., AND T. LIMO. 1973. Long-lasting potentiation of svnantic transmission in the dentate area of the anaesthetized

20. 21. 22. 23. 24. 25.

rabbit following stimulation of the perforant path. J. Physiol. 232: 331-356. BLISS, T. V. P., AND A. R. GARDNER-MED~IN. 1973. Long-lasting potentiation of synaptic transmission in the dentate area of the unanaesthetized rabbit following stimulation of the perforant path. J. Physiol. 232: 357-374. COLLINS, W. P. III, M. G. HONIG, AND L. M. MENDELL. 1984. Heterogeneity of group Ia synapses on homonymous cu-motoneurons as revealed by high-frequency stimulation of Ia afferent fibers. J. Neurophysiol. 52: 980-993. &EAGER, R., T. DUNWIDDIE, AND G. LYNCH. 1980. Paired-pulse and frequency facilitation in the CA1 region of the in vitro rat hippocampus. J. Physiol. 299: 409-424. DUDEK, F. E., S. A. DEADWYLER, C. W. COTMAN, AND G. LYNCH. 1976. Intracellular responses from granule cell layer in slices of rat hippocampus: Perforant path synapse. J. Neurophysiol. 39: 384-393. DUFFY, C. J., AND T. J. TEYLER. 1978. Development of potentiation in the dentate gurus of rat: Physiology and anatomy. Brain Res. Bull. 3: 425-430. GREENSTEIN, Y., C. PAVLIDES, AND J. WINSON. 1988. Long-term potentiation in the dentate gyrus is preferentially induced at theta rhythm periodicity. Brain Res. 438: 331-334. KATZ, B., AND R. MILEDI. 1968. The role of calcium in neuromuscular facilitation. J. Physiol. 195: 481-492. KBNIG, J. F. R., AND R. A. KLIPPEL. 1963. The Rat Bruin: A Stereotaxic Atlas of the Forebrain and Lower Parts of the Brain Stem, Williams & Wilkins (Eds.), Baltimore. KUNO, M. 1964. Mechanism of facilitation and depression of the excitatory synaptic potential in spinal motoneurons. J. Physiol. 175: 100-112. LARSON, J., D. WONG, AND G. LYNCH. 1986. Patterned stimulation at the theta frequency is optimal for the induction of hippocampal long-term potentiation. Bruin Res. 368: 347-350. LARSON, J., AND G. LYNCH. 1989. Theta patterns stimulation and the induction of LTP: The sequence in which synapses are stimulated determines the degree to which they potentiate. Brain Res. 489: 49-58. LYNCH, G., AND M. BAUDRY. 1984. The biochemistry of memory: A new and specific hypothesis. Science 224: 1057-1063. MACVICAR, B., AND F. E. DUDEK. 1979. Intracellular recordings from hippocampal CA3 pyramidal cells during repetitive activation of mossy fibers in vitro. Bruin Res. 168: 377-381. Mtioz, M. D., A. N&JEZ, AND E. GARCIA-AUSTT. 1990. In vivo intracellular analysis of rat dentate granule cells. Bruin Res. 509: 91-98. N~~Ez, A., E. GARCIA-Aus=, AND W. BUIQO,JR. 1987. Intracellular &rhythm generation in identified hippocampal pyramids. BrainRes. 416:289-300. PITLER, T. A., AND P. W. LANDFIELD. 1987. Postsynaptic membrane shifts during frequency potentiation of the hippocampal EPSP. J. Neurophysiol. 56: 866-882. RAHAMIMOFF, R. 1968. A dual effect of calcium ions on neuromuscular facilitation. J. Physiol. 196: 471-480. STAUBLI, U., AND G. LYNCH. 1987. Stable hippocampal longterm potentiation elicited by “theta” pattern stimulation. Brain Res. ?35: 227-234. TEYLER, T. J., AND P. DISCENNA. 1987. Long-term potentiation. Annu. Reu. Neurosci. 10: 131-161. YAMAMOTO, C., AND S. SAWADA. 1981. Important factors in induction of long-term potentiation in thin hippocampal sections. Exp. Neural. 74: 122-130.