Brain Research, 586 (1992) 247-255 © 1992 Elsevier Science Publishers B.V. All rights reserved 0006-8993/92/$05.00
247
BRES 17929
In vivo intracellular correlates of hippocampal formation theta-on and theta-off cells Jan Konopacki t, Brian H. Bland, Luis V. Colom 2 and Scott D. eddie Departnwnt of Psychology, Behat'ioral Neuroscience Research Group, The Unit'ersity of Calgary, Calgary, Alta. (Canada) (Accepted 3 March 1992)
Key words: Hippocampal intracellular correlate; Spike characteristics membrane oscillation; Theta-on theta-off
Using urethane-anesthetized rats, intracellular recordings were made in hippocampal formation cells classified according to previously established criteria as either theta-on or theta-off, in order to further define the electrophysiological characteristics of these cells. Four cells classified as phasic theta-off cells had short duration spikes ( < I ms), high input resistances (54-61 M.O) and large fast afterhyperpolarizations (6-10 mV), thus sharing some of the properties of identified hippocampal interneurons. Phasic theta-off cells also exhibited rhythmic membrane potential oscillations (MPOs) ranging from 4 to 10 mV in amplitude during the simultaneous occurrence of extracellular theta field activity, but noi during the occurrence of large amplitude irregular field activity (LIA). The MPOs of phasic theta-off cells were the same frequency as and were highly coherent with the extracellular theta field activity, in all four phasic theta-off cells the positive peak of the MPO was in phase with the positive peak of the local theta field activity. At the onset of extraceUular theta field activity above 4-5 Hz, the membrane potentials of phasic theta-off cells showed a 5-10-mV hyperpolarizing shift, accompanied by MPOs without spike discharges. As theta frequency slowed down there was a return to baseline membrane potential levels and spike discharges occurred near the positive peak of the MPOs. The seven cells classified as phasic theta-on had longer duration spikes (>1 ms), lower input resistances (22-36 M e ) and small (approx. 1.0 mV) fast aflerhyperpolarizations, thus sharing some of the properties of hippocampal projection cells. Phasic theta-on cells also exhibited MPOs (4-10 mV) during theta but not during LIA and the MPOs were also the same frequency as and were highly coherent with the extracellular theta field activity. The positive peak of the MPOs were in phase with the negative peak of the local ~xtracellular theta. At the: onset of extracellular theta the membrane potentials of phasic th~ta-on cells showed a 5-10 mV depolarizing shift accompanied by MPOs and spike discharges occurring n~ar the positiw peak of the MPOs. The five tonic them-on cells and two non-related cells had similar electrophysiological properties as the phasic theta-on cells. Tonic theta-on and non-related cells fililed to exhibit MPOs during the occurrence of either extracellular theta field activity or UA.
INTRODUCTION The hippocampal formation exhibits one of the most well-documented rhythmicities of the central nervous system of mammals, termed theta activity. The other two most commonly occurring hippocampal field activities are non-rhythmic patterns, termed large amplitude irregular activity (LIA) 4s and fast (beta) waves 7,2H,45,4~,. We carried out a number of extracellular studies of the spike train dynamics of hippocampai formation neurons in relation to the field activities of theta and LIA. These studies led us to develop a set of criteria for classifying cells in relation to theta field activity and we subsequently used these criteria for classifying cells in
the medial septum.vertical limb of the diagonal band of Broca and the posterior cingulate cortex. Theta-related cells in the hippocampal formation, posterior cingulate cortex and the medial septum-vertical limb of the diagonal band of Broca, comprised two distinct populations we termed theta-on and theta-off. Theta-on cells increased their activity during theta and theta-off cells decreased their activity during theta. Each of these populations had two subtypes based on their pattern of cell discharges. The first subtype we termed phasic, so-named since they discharged in rhythmic bursts related to a particular phase of the extracellular theta field activity. The second subtype we termed tonic since they discharged in a constant regular or
Correspondence: B.H. Bland, Department of Psychology, Behavioral Neuroscience Research Group, The University of Calgary, Calgary, Alta., Canada T2N IN4. Fax: (1) (403) 282-8249. t Present address: Department of Animal Physiology, Laboratory of Limbic System Physiology, University of L6d~, 90-222 L6d~., Poland. 2 Present address: Division of Neuroscience, Baylor College of Medicine, One Baylor Plaza, Houston Medical Center, Houston, TX 77030, USA.
248 irregular non-bursting manner with no consistent relation to the phase of each cycle of the extracellular field activity, in addition to the above, we have investigated the pharmacology, afferent inputs and behavioral correlates of theta-on and theta-off cells 3-5.8-13.16.44 The unequivocal morphological identity of theta-related cells remains as a yet unattained but important objective. There have been suggestions made for the identity of theta-related cells in the hippocampal formation. Fox and Ranck ~s,~,~concluded that their data were consistent with most theta cells (our theta-on cells) being interneurons, although they conceded in the latter paper that some theta cells may be projection cells. Bland et al. z presented evidence for both CA1 and dentate granule cells being theta cells. Rose et al..1, concluded that dentate granule cells were theta cells and Buzsfiki et al. 7 argued that their physiologically identified interneurons and granule cells met the criteria tbr being theta cells. Nffiez et al. ~' and Mufioz et al..~s have shown that CA I pyramidal cells and dentate granule cells, respectively, show all the characteristics of theta-on cells. Bland and Colom 4 advanced the hypothesis that hippocampal theta-off cells were inhibitory interneurons receiving inhibitory GABAergic projections from the medial septum-vertical limb of the diagonal band of Broca. This hypothesis was supported by a recent study in which these septal nuclei were manipulated using the technique of reversible procaine suppression 44. The purpose of the present study was to investigate the intracellular correlates of theta-related cells in the in rive, urethane-anesthetized rat, in order to provide more infi)rmation concerning the ¢lectrophysiological characteristics of theta-on and theta.off cells, Particular attention was given to the presence or absence of fast spike afterhyperpolarizations, spike durations and the presence or absence of rhythmic membrane potential oscillations (MPOs).
to the midline and 2.4-2.8 mm ventral to the dural surface). Intracellular recordings in the left hippocampal formation were made at the same posterior and lateral coordinates, starting at the alvear surface and continuing ventral to the lower (endal) blade of the dentate granule cells. The area for recording was prepared in the following manner: a small (approximately 4 x 4 ram) square of bone was drilled out after a small hole was previously drilled in the center of the square. This square of bone was then replaced in the skull and pushed down to provide pressure on the cortex. The bone was cemented in place with cyanolite. The animal's body rested on a tilted platform mounted to the top of a scissor jack. While viewing the hole in the skull through a dissecting microscope, the animal's body was raised until the movement of the cerebrospinal fluid in the hole was maximally reduced. Cells were recorded with glass microelectrodes (80-120 M,f~) filled with 2 M potassium acetate, lntracellular signals were amplified by a conventional high-impedance amplifier (Neurodata), displayed on an oscilloscope, and these and all other signals were stored on an FM tape recorder for off-line analysis. Once an intracellular recording was stabilized, the protocol consisted of collecting I-3 rain of cell activity during large amplitude irregular hippocampal field activity (LIA) and !-3 rain of cell activity during synchronized (theta) hippocampal field activity. Following this a series of depolarizing and hyperpolarizing currents (20 ms duration) were injected to determine discharge properties and input resistances. Spike height and duration was determined from the first action potential evoked at threshold depolarization. Fast afterhyperpolarizations were measured from the base of the action potential to the peak of the fast afterhyperpolarization. No attempt was made to measure time constants, Cells selected for subsequent off-line analysis had to have resting membrane potentials of > 50 mV and overshooting action potentials. Data analysis included: (i) autocorrclation histograms of the spike trains, during LIA and theta conditions; (2) cross-correlations between the spike trains and the simultaneously-occurring field activities of either LIA or theta; (3) t-tests for determining the difference in mean firing rates between LIA and thcta conditions, The first 3 analyses allowed us to classify the cells as theta-on, theta-off or unrelated, according to our previous criteria a. Ouc to time constraints lind sensitivity of the preparation, we were not able to produce a wide enough range of theta frequenci~:s to carry out linear regression allalysesl (4) determination of the presence or absence of hyperpolarizing and depolarizing after potentials and of spike duration; (5) frequency spectrographs of b,~lh the extr~lc~:llular I'[eld activity ,nd :nemb~an~ potentials: (6) measurements of coherence between the ¢xtracellular field activity and the membrane potentials; (7) frozen sections (40 gin) were taken serially and mounted on glass slides for subsequent thionine staining and determination of electrode tract locations, Judgements of the layer the cells were recorded in were made on the basis of depth profiles carried out with the field electrode, the depth of the cell recording and histology,
RESULTS MATERIALS AND METHODS
The data were obtained from 50 male black-hooded rats (0,1500.2(10 kg) supplied by the Animal Care Services at the University of Calgary. The rats were initially anesthetized with halothane while tracheal and jugular cannulae were inserted, Halothane was then discontinued and urethane was administered via the jugular cannula to maintain an appropriate level of anesthesia during the remaining surgical and experimental procedures, The rats were placed in the stereotaxic instrument with the plane between hregma and lambda leveled to horizontal, Body temperature was maintained at 3"/°C and heart rate was monitored constantly throughout the experiment, An uninsulated tungsten wire placed in the cortex, anterior to bregma, served as an indifferent electrode and the stereotaxic frame was connected to ground, A tungsten microelectrode (0,2-0,5 Mf~) for recording hippocampal field activity was placed in the right dorsal hippocampal formation in the dentate molecular layer (3,0-3,3 mm posterior to hregma, 2,0-2,5 mm lateral
A total of 43 cells were recorded and of these, 18 were held long enough to complete the entire experimental protocol for classification as theta-on, theta-off or non-related cells. The remaining 25 cells were not held long enough to be classified according to our criteria for theta-related cells, Out of the 43 cells, 17 were located in the region of the CA1 layer and 26 were located in the region of the upper blade of the dentate.
General electrophysiological findings Resting membrane potentials ranged from - 5 4 to - 6 7 mV (.~ = 60.53 + 5.28 mV) and action potentials
249 TABLE !
Electrophysiologicai characteristics of cells classified as theta.related or non-related and of cells not hem long enough to be classified
Number (in parentheses) given when less than the total in group. R M P , resting membrane potential, MPO, membrane potential oscillation, f A H P , fast afterhyperl~darization measured from the base of the action potential to the peak of the f A H P . Cell
RMP (mV)
Spike height (mV)
h~put resistance(M)
MPO (mV) Theta
Lia
(ms)
62.3-1- 4.2
60.7+5.5
28.7+7.1 (4)
7.9+2.1
-
58.8+4.7
60.24.2.6
29.34.5.4 (2)
-
59.54.5.9
50.54.2.6
5 7 . 8 + 3 . 2 (2)
61.04-1.0
54.04- 2.0
32.44-6.9
60.5+5.5
58.34.6.3
classification Phasic thetaon (n = 7) Tonic thetaon (n = 5) Phasic thetaoff(n = 4) Non-related (n = 2)
D,ration
fAHP (mV)
Dischargerates "(Hz) Theta LIA
1.8+0.2
1.4+0.2
8.84-3.98
3 . 8 + 1.5
-
2.14.0.7
1.24-0.3
7.84-2.1
3.7+0.9
7.84.2.6
-
0.9+0.3
8.74.2.3
2.94.0.4
9.14-2.2
-
-
1.74-0.6
1.64-0.1
2.64-0.1
2.5 4-0.2
-
-
2.2+0.8(24) 0.8(1)
1.34.0.4(24) 9.4(1)
-
-
Non-classified (n = 25)
-
ranged from 48 mV to 66 mV ( . ~ - 61.27 + 6.01 mV). Five of the 43 cells (12%) in this study had large fast afterhyperpolarizations (AHP) (6-10 mV) and short spike durations ( < 1 ms). The remaining 38 cells (88%) had a small fast afterhyperpolarization and longer duration action potentials (1.8-2.2 ms). Table I presents a summary of the electrophysiological characteristics of the 43 cells recorded in the study, grouped according to their classification in relation to theta field activity for the cases where this was successfully achieved and into a non.classified group where it was not.
cells and 2 were non-related. The main results of analyses of these cells were as follows: (1) only the 4 cells classified as phasic theta-off had large AHP's and short duration spikes; (2) both phasic theta-on and phasic theta-off cells had rhythmic membrane potential oscillations (MPOs) during the simultaneous occurrence of extracellular theta field activity but not during the occurrence of LIA; (3) tonic theta-on cells and non-related cells never showed MPOs; (4) the MPOs of both phasic theta-on and phasic theta-off cells were the same frequency as the simultaneously occurring extracellular theta activity and there was a high degree of coherence between the intra- and extracellular activities; (5) a~ the onset of extracellular theta field activity the membrane-potentials of phasic theta-on cells showed a depolarizing shift accompanied by the appearance of MPOs and spike discharges occurring of-
CX,lls classified in relation to theta field activity As stated above, 18 cells were held long enough to successfully classify them in relation to theta field activity. Seven cells were classified as phasic theta-on cells, 4 as phasic theta-off cells, 5 as tonic theta-on 8pontaneouo There
8pontineoue LIA
B,
AB
,8 mV Extrecelluler Dentate Layer
vlmV~,r-w
w-
*
w
-
,,,*w
- v
Intr-,oellulmr Dentate Layer
w ~ . ~ ,
. . . . .
10 mV
,,v..'-v---.
1No
lO mv
~ ~ v r 10 mmm
110 mV
llllO mmec
Fig. I. lntracellular correlates of a phasic theta-on ceil. A: fast sweep of a spontaneous discharge. Spike discharged on a slow depolarizing
potential. Note the wide spike width and the absence of a large fast afterhyperpolarization. B: upper panel shows the relationship between theta field activity and the simultaneously occurring spike discharges (spikes chopped). In this and all other phasic theta-on cells, the positive peak of the MPOs were in phase with the negative peak of the extracellular theta field activity. The lower panel shows the spike discharges at a faster sweep. Note the membrane potential oscillations. C: upper panel shows the relationship between LIA and the simultaneously occurring spike discharges. The lower panel shows the discharges at a faster sweep. Note the absence of membrane potential oscillations.
25O ten near the positive peak of the MPOs; (6) at the onset of extracellular theta field activity above 4-5 Hz, there was a marked hyperpolarizing shift in the membrane potentials of phasic theta-off cells, accompanied by the appearance of MPOs without spike discharges. As the theta frequency slowed down there was a return to baseline membrane potential levels and spike discharges occurred near the positive peak of the MPOs. Phasic them-on cells (n -- 7 ) Five of these cells were located in the dentate layer and the remaining two were in the CAI layer. A representative example of a phasic theta-on celt is shown in Fig. 1. This cell was located in the dentate granule cell region but no measures were taken to positively identify it as a granule cell. Fig. IA shows a fast sweep of the spike discharge taken during spontaneous activity. The spike discharged on a slow depolarizing potential, had a spike duration of slightly over 2 ms and very little fast a~terhyperpolarization. The upper panel in Fig. I B shows the simultaneous recording of theta field activity from the dentate and the accompanying cell discharges, in this figure, extracellular negativity is up, indicating that the cell tended to discharge near extracellular peak negativity (verified by cross-correlation). The same phase relation was observed for the other 6 phasic theta-on cells. Also evi.
dent was the presence of rhythmic MPOs with the spike discharges occurring near the positivity of the MPOs. it was occasionally observed that MPOs could occur without spike discharges, in the total sample of phasic theta.on cells. The bottom panel in Fig. I B shows it faster time base of the cell discharge and the occurrence of MPOs slightly over 6 mV. in this study, MPOs of phasic theta.on cells ranged from 4 mV to almost 10 inV. At the onset of theta field activity, the
A.
B.
membrane potentials of all phasic theta-on cells typically showed a depolarizing shift in the range of 5-10 mV, compared to the average membrane potential levels measured during LIA. The top panel in Fig. 1C shows the cell discharging in a non-rhythmic pattern during the occurrence of LIA. The bottom panel shows the faster time base, taken during LIA. Note the absence of MPOs. The mean discharge of this cell during theta was 9.9 + 0.991 spikes/s and 5.7 + 0.9 spikes/s during LIA (significantly different at P < 0.001). The cell discharges and field activity were cross-correlated during theta but not during LIA and the auto-correlation histograms again showed rhythmicity during theta but not during LIA. Phasic theta-off cells (n = 4) Three of these cells were located in the CAI layer and the remaining cell was in the dentate layer. A representative example of a phasic theta-off cell is shown in Fig. 2. This cell was located in the CAI cell body region. Fig. 2A shows a fast sweep of the spike discharge taken during spontaneous activity. The resting membrane potential is indicated by the dotted line. The spike discharged on a slow depolarizing potential, had a spike duration of less than l ms and showed a large (8.0 mV) afterhyperpolarization. Fig. 2B is a continuous record of the spike train, showing the transition fi'om LIA to theta produced by a tail pinch (marked by arrow). The resting membrane potential is
indicated by the dotted line. At the onset of thcta, all 4 cells showed a hyperpolarizing shift of the membrane potential (approx. 5-10 mV) accompanied by the appearance of MPOs. The membrane potential then returned to baseline approximately 2 s later. As theta frequency declined, spikes occurred near the positive peak of the MPOs. With the transition to LIA, spike
r.
LIA
That.
Extro©ellular Dentate Layer
Int..llula,
cA1 Lay°r
I,S mV
F mY t Tall plncheh~j lo rnv
__] 10 rose©
~ Tall Pinch
J10 my 1 sac
Fig. 2. Intracellular cor;elates of a phasic theta-off cell, A: fast sweep of a spontaneous discharge, The resting membrane potential is indicated by the dotted line here and in B, The cell discharged on a slow depolarizing potential, Note the narrow spike width and presence of a large fast afterhyperpolarization. B: a continuous record of the spike discharges showing the transition from L I A to theta induced by a tail pinch (arrow). Note the cessation of discharges, the hypcrpolarizing shift and presence of membrane potential oscillations during theta. ]n this and all other phasic theta-off cells, the positive pe~,k of the MPOs were in phase with the positive peak of the extracellular field activity. C: details of the relationship between the extracellular field activity and spike discharges for the sequence shown in B.
251 discharges increased in frequency in a non-rhythmic pattern. F~g. 2C shows the simultaneous occurrence of field activity recorded from dentate region and spike discharges from the CA1 region. In this case, Fig, 2C corresponds exactly to Fig, 2B. Fig. 2C also illustrates that the positive peak of the MPOs was occurring on the positive phase of the extracellular theta field activity, since the dentate field was phase-reversed 180" to the local CA1 theta field activity. Indeed, all the MPOs of all 4 phasic theta-off cells exhibited the same phase relation, which was opposite to the phase relations demonstrated between the MPOs of phasic theta-on cells and the extracellular theta field activity.
Tonic theta-on cells (n
=
cantly different at P < 0.001). The cell discharges were not cross-correlated during either theta or LIA, nor did the auto-correlation histograms show any degree of rhythmicity for either condition.
Frequency and coherence relationships between field actit,ity and membrane potentials Phasic them-on cells All phasic theta-on cells (n = 7) exhibited MPOs during the occurrence of theta field activity but not during LIA. Fig. 4 shows a representative analysis of one of these cells. The top two panels show the frequency spectographs of the extracellular field activity and intracellular membrane potentials during theta (left side) and LIA (right side) expressed as amplitude in decibels (dB). The bottom panel gives the results of the coherence testing between the two signals, expressed as a correlation coefficient (rho). In this example both the extracellular theta and the membrane potential had a peak frequency at 3.5 Hz with a very high coherence. As indicated by the panels on the right, no peak frequency was observed for either the extracellular field activity or the membrane potential during LIA and the coherence was correspondingly low.
5)
Four of these cells were located in the dentate layer and one was located in the CAI layer. Fig. 3 shows a representative example of a tonic theta-on cell located in the dentate layer. Fig. 3A shows a fast sweep of the spike discharge taken during spontaneous activity. The spike discharged on a slow depolarizing potential, had a spike duration of slightly over 2 ms and did not have a large fast afterhyperpolarization. The upper panel in Fig. 3B shows the simultaneous recording of theta field activity from the dentate and the accompanying cell discharges. The cell discharged in an irregular pattern during theta field activity. As can be seen in the lower panel of Fig. 3B, the membrane potential did not show the occurrence of MPOs during theta activity. The top panel of Fig, 3C shows tl~.e reduced number of cell discharges occurring during LIA, again in a nonrhythmic pattern, The bottom panel shows the spike discharges during LIA, with a faster time base. Note again the complete absence of MPOs during LIA. The mean discharge of this cell during theta was 11.1 :t: 2.04
Phasic theta-off cells The frequency spectral and coherence analyses of the extracellular field activity accompanying the membrane potential of a representative phasic theta-off cell (n = 4) is given in Fig. 5. The upper panels on the left clearly indicate an identical peak frequency (4.25 Hz) for the extraceilular theta and the membrane potential. The high degree of coherence between the two signals is indicated in the lower left panel. The corresponding
spikes/s and 5.3 + 1.0 spikes/s during LIA (signifi-
''NN Spontaneous Thets
All
Spontaneous LIA Extraoellular Dentate Layer
I .5 mV
Intrscellular 1IF ~
v
-
•
w
-~-..,w-,r-,,
Ir
Dentate L a y e r r - ~ , , . -
v . . . . . . , ~ - ' - - ' . . .
-
i lOmV
1 el©
10 mV 10 mV
.eo,.,,-J . . . . . . . . . .
10 mse¢
I
250 meec
Fig. 3. lntracellular correlates of a tonic theta-on cell. A: fast sweep of a spontaneous discharge. Spike dischargeson a slow depolarizing potential. Note the wide spike and the absence of a large fast afterhyperpolarization.B: upper panel shows the relationshipbetween theta field activity and the simultaneouslyoccurringspike discharges(spikes chopped). The lower panel shows the spike dischargesat a faster sweep. Note the absence of membrane potential oscillations. C: upper panel shows the relationship between LIA and the simultaneouslyoccurring spike discharges.The lowerpanel shows the dischargesat a faster sweep. Note absenceof membrane potential oscillations.
252
Theta
Theta
LIA
| oIlill[IT~'n~
Cell
Phasic-Off
Phasic-On Cell
LIA
g d! Ill lit I : I ~
Illl'lllll~
247
BRES 17929
In vivo intracellular correlates of hippocampal formation theta-on and theta-off cells Jan Konopacki t, Brian H. Bland, Luis V. Colom 2 and Scott D. eddie Departnwnt of Psychology, Behat'ioral Neuroscience Research Group, The Unit'ersity of Calgary, Calgary, Alta. (Canada) (Accepted 3 March 1992)
Key words: Hippocampal intracellular correlate; Spike characteristics membrane oscillation; Theta-on theta-off
Using urethane-anesthetized rats, intracellular recordings were made in hippocampal formation cells classified according to previously established criteria as either theta-on or theta-off, in order to further define the electrophysiological characteristics of these cells. Four cells classified as phasic theta-off cells had short duration spikes ( < I ms), high input resistances (54-61 M.O) and large fast afterhyperpolarizations (6-10 mV), thus sharing some of the properties of identified hippocampal interneurons. Phasic theta-off cells also exhibited rhythmic membrane potential oscillations (MPOs) ranging from 4 to 10 mV in amplitude during the simultaneous occurrence of extracellular theta field activity, but noi during the occurrence of large amplitude irregular field activity (LIA). The MPOs of phasic theta-off cells were the same frequency as and were highly coherent with the extracellular theta field activity, in all four phasic theta-off cells the positive peak of the MPO was in phase with the positive peak of the local theta field activity. At the onset of extraceUular theta field activity above 4-5 Hz, the membrane potentials of phasic theta-off cells showed a 5-10-mV hyperpolarizing shift, accompanied by MPOs without spike discharges. As theta frequency slowed down there was a return to baseline membrane potential levels and spike discharges occurred near the positive peak of the MPOs. The seven cells classified as phasic theta-on had longer duration spikes (>1 ms), lower input resistances (22-36 M e ) and small (approx. 1.0 mV) fast aflerhyperpolarizations, thus sharing some of the properties of hippocampal projection cells. Phasic theta-on cells also exhibited MPOs (4-10 mV) during theta but not during LIA and the MPOs were also the same frequency as and were highly coherent with the extracellular theta field activity. The positive peak of the MPOs were in phase with the negative peak of the local ~xtracellular theta. At the: onset of extracellular theta the membrane potentials of phasic th~ta-on cells showed a 5-10 mV depolarizing shift accompanied by MPOs and spike discharges occurring n~ar the positiw peak of the MPOs. The five tonic them-on cells and two non-related cells had similar electrophysiological properties as the phasic theta-on cells. Tonic theta-on and non-related cells fililed to exhibit MPOs during the occurrence of either extracellular theta field activity or UA.
INTRODUCTION The hippocampal formation exhibits one of the most well-documented rhythmicities of the central nervous system of mammals, termed theta activity. The other two most commonly occurring hippocampal field activities are non-rhythmic patterns, termed large amplitude irregular activity (LIA) 4s and fast (beta) waves 7,2H,45,4~,. We carried out a number of extracellular studies of the spike train dynamics of hippocampai formation neurons in relation to the field activities of theta and LIA. These studies led us to develop a set of criteria for classifying cells in relation to theta field activity and we subsequently used these criteria for classifying cells in
the medial septum.vertical limb of the diagonal band of Broca and the posterior cingulate cortex. Theta-related cells in the hippocampal formation, posterior cingulate cortex and the medial septum-vertical limb of the diagonal band of Broca, comprised two distinct populations we termed theta-on and theta-off. Theta-on cells increased their activity during theta and theta-off cells decreased their activity during theta. Each of these populations had two subtypes based on their pattern of cell discharges. The first subtype we termed phasic, so-named since they discharged in rhythmic bursts related to a particular phase of the extracellular theta field activity. The second subtype we termed tonic since they discharged in a constant regular or
Correspondence: B.H. Bland, Department of Psychology, Behavioral Neuroscience Research Group, The University of Calgary, Calgary, Alta., Canada T2N IN4. Fax: (1) (403) 282-8249. t Present address: Department of Animal Physiology, Laboratory of Limbic System Physiology, University of L6d~, 90-222 L6d~., Poland. 2 Present address: Division of Neuroscience, Baylor College of Medicine, One Baylor Plaza, Houston Medical Center, Houston, TX 77030, USA.
248 irregular non-bursting manner with no consistent relation to the phase of each cycle of the extracellular field activity, in addition to the above, we have investigated the pharmacology, afferent inputs and behavioral correlates of theta-on and theta-off cells 3-5.8-13.16.44 The unequivocal morphological identity of theta-related cells remains as a yet unattained but important objective. There have been suggestions made for the identity of theta-related cells in the hippocampal formation. Fox and Ranck ~s,~,~concluded that their data were consistent with most theta cells (our theta-on cells) being interneurons, although they conceded in the latter paper that some theta cells may be projection cells. Bland et al. z presented evidence for both CA1 and dentate granule cells being theta cells. Rose et al..1, concluded that dentate granule cells were theta cells and Buzsfiki et al. 7 argued that their physiologically identified interneurons and granule cells met the criteria tbr being theta cells. Nffiez et al. ~' and Mufioz et al..~s have shown that CA I pyramidal cells and dentate granule cells, respectively, show all the characteristics of theta-on cells. Bland and Colom 4 advanced the hypothesis that hippocampal theta-off cells were inhibitory interneurons receiving inhibitory GABAergic projections from the medial septum-vertical limb of the diagonal band of Broca. This hypothesis was supported by a recent study in which these septal nuclei were manipulated using the technique of reversible procaine suppression 44. The purpose of the present study was to investigate the intracellular correlates of theta-related cells in the in rive, urethane-anesthetized rat, in order to provide more infi)rmation concerning the ¢lectrophysiological characteristics of theta-on and theta.off cells, Particular attention was given to the presence or absence of fast spike afterhyperpolarizations, spike durations and the presence or absence of rhythmic membrane potential oscillations (MPOs).
to the midline and 2.4-2.8 mm ventral to the dural surface). Intracellular recordings in the left hippocampal formation were made at the same posterior and lateral coordinates, starting at the alvear surface and continuing ventral to the lower (endal) blade of the dentate granule cells. The area for recording was prepared in the following manner: a small (approximately 4 x 4 ram) square of bone was drilled out after a small hole was previously drilled in the center of the square. This square of bone was then replaced in the skull and pushed down to provide pressure on the cortex. The bone was cemented in place with cyanolite. The animal's body rested on a tilted platform mounted to the top of a scissor jack. While viewing the hole in the skull through a dissecting microscope, the animal's body was raised until the movement of the cerebrospinal fluid in the hole was maximally reduced. Cells were recorded with glass microelectrodes (80-120 M,f~) filled with 2 M potassium acetate, lntracellular signals were amplified by a conventional high-impedance amplifier (Neurodata), displayed on an oscilloscope, and these and all other signals were stored on an FM tape recorder for off-line analysis. Once an intracellular recording was stabilized, the protocol consisted of collecting I-3 rain of cell activity during large amplitude irregular hippocampal field activity (LIA) and !-3 rain of cell activity during synchronized (theta) hippocampal field activity. Following this a series of depolarizing and hyperpolarizing currents (20 ms duration) were injected to determine discharge properties and input resistances. Spike height and duration was determined from the first action potential evoked at threshold depolarization. Fast afterhyperpolarizations were measured from the base of the action potential to the peak of the fast afterhyperpolarization. No attempt was made to measure time constants, Cells selected for subsequent off-line analysis had to have resting membrane potentials of > 50 mV and overshooting action potentials. Data analysis included: (i) autocorrclation histograms of the spike trains, during LIA and theta conditions; (2) cross-correlations between the spike trains and the simultaneously-occurring field activities of either LIA or theta; (3) t-tests for determining the difference in mean firing rates between LIA and thcta conditions, The first 3 analyses allowed us to classify the cells as theta-on, theta-off or unrelated, according to our previous criteria a. Ouc to time constraints lind sensitivity of the preparation, we were not able to produce a wide enough range of theta frequenci~:s to carry out linear regression allalysesl (4) determination of the presence or absence of hyperpolarizing and depolarizing after potentials and of spike duration; (5) frequency spectrographs of b,~lh the extr~lc~:llular I'[eld activity ,nd :nemb~an~ potentials: (6) measurements of coherence between the ¢xtracellular field activity and the membrane potentials; (7) frozen sections (40 gin) were taken serially and mounted on glass slides for subsequent thionine staining and determination of electrode tract locations, Judgements of the layer the cells were recorded in were made on the basis of depth profiles carried out with the field electrode, the depth of the cell recording and histology,
RESULTS MATERIALS AND METHODS
The data were obtained from 50 male black-hooded rats (0,1500.2(10 kg) supplied by the Animal Care Services at the University of Calgary. The rats were initially anesthetized with halothane while tracheal and jugular cannulae were inserted, Halothane was then discontinued and urethane was administered via the jugular cannula to maintain an appropriate level of anesthesia during the remaining surgical and experimental procedures, The rats were placed in the stereotaxic instrument with the plane between hregma and lambda leveled to horizontal, Body temperature was maintained at 3"/°C and heart rate was monitored constantly throughout the experiment, An uninsulated tungsten wire placed in the cortex, anterior to bregma, served as an indifferent electrode and the stereotaxic frame was connected to ground, A tungsten microelectrode (0,2-0,5 Mf~) for recording hippocampal field activity was placed in the right dorsal hippocampal formation in the dentate molecular layer (3,0-3,3 mm posterior to hregma, 2,0-2,5 mm lateral
A total of 43 cells were recorded and of these, 18 were held long enough to complete the entire experimental protocol for classification as theta-on, theta-off or non-related cells. The remaining 25 cells were not held long enough to be classified according to our criteria for theta-related cells, Out of the 43 cells, 17 were located in the region of the CA1 layer and 26 were located in the region of the upper blade of the dentate.
General electrophysiological findings Resting membrane potentials ranged from - 5 4 to - 6 7 mV (.~ = 60.53 + 5.28 mV) and action potentials
249 TABLE !
Electrophysiologicai characteristics of cells classified as theta.related or non-related and of cells not hem long enough to be classified
Number (in parentheses) given when less than the total in group. R M P , resting membrane potential, MPO, membrane potential oscillation, f A H P , fast afterhyperl~darization measured from the base of the action potential to the peak of the f A H P . Cell
RMP (mV)
Spike height (mV)
h~put resistance(M)
MPO (mV) Theta
Lia
(ms)
62.3-1- 4.2
60.7+5.5
28.7+7.1 (4)
7.9+2.1
-
58.8+4.7
60.24.2.6
29.34.5.4 (2)
-
59.54.5.9
50.54.2.6
5 7 . 8 + 3 . 2 (2)
61.04-1.0
54.04- 2.0
32.44-6.9
60.5+5.5
58.34.6.3
classification Phasic thetaon (n = 7) Tonic thetaon (n = 5) Phasic thetaoff(n = 4) Non-related (n = 2)
D,ration
fAHP (mV)
Dischargerates "(Hz) Theta LIA
1.8+0.2
1.4+0.2
8.84-3.98
3 . 8 + 1.5
-
2.14.0.7
1.24-0.3
7.84-2.1
3.7+0.9
7.84.2.6
-
0.9+0.3
8.74.2.3
2.94.0.4
9.14-2.2
-
-
1.74-0.6
1.64-0.1
2.64-0.1
2.5 4-0.2
-
-
2.2+0.8(24) 0.8(1)
1.34.0.4(24) 9.4(1)
-
-
Non-classified (n = 25)
-
ranged from 48 mV to 66 mV ( . ~ - 61.27 + 6.01 mV). Five of the 43 cells (12%) in this study had large fast afterhyperpolarizations (AHP) (6-10 mV) and short spike durations ( < 1 ms). The remaining 38 cells (88%) had a small fast afterhyperpolarization and longer duration action potentials (1.8-2.2 ms). Table I presents a summary of the electrophysiological characteristics of the 43 cells recorded in the study, grouped according to their classification in relation to theta field activity for the cases where this was successfully achieved and into a non.classified group where it was not.
cells and 2 were non-related. The main results of analyses of these cells were as follows: (1) only the 4 cells classified as phasic theta-off had large AHP's and short duration spikes; (2) both phasic theta-on and phasic theta-off cells had rhythmic membrane potential oscillations (MPOs) during the simultaneous occurrence of extracellular theta field activity but not during the occurrence of LIA; (3) tonic theta-on cells and non-related cells never showed MPOs; (4) the MPOs of both phasic theta-on and phasic theta-off cells were the same frequency as the simultaneously occurring extracellular theta activity and there was a high degree of coherence between the intra- and extracellular activities; (5) a~ the onset of extracellular theta field activity the membrane-potentials of phasic theta-on cells showed a depolarizing shift accompanied by the appearance of MPOs and spike discharges occurring of-
CX,lls classified in relation to theta field activity As stated above, 18 cells were held long enough to successfully classify them in relation to theta field activity. Seven cells were classified as phasic theta-on cells, 4 as phasic theta-off cells, 5 as tonic theta-on 8pontaneouo There
8pontineoue LIA
B,
AB
,8 mV Extrecelluler Dentate Layer
vlmV~,r-w
w-
*
w
-
,,,*w
- v
Intr-,oellulmr Dentate Layer
w ~ . ~ ,
. . . . .
10 mV
,,v..'-v---.
1No
lO mv
~ ~ v r 10 mmm
110 mV
llllO mmec
Fig. I. lntracellular correlates of a phasic theta-on ceil. A: fast sweep of a spontaneous discharge. Spike discharged on a slow depolarizing
potential. Note the wide spike width and the absence of a large fast afterhyperpolarization. B: upper panel shows the relationship between theta field activity and the simultaneously occurring spike discharges (spikes chopped). In this and all other phasic theta-on cells, the positive peak of the MPOs were in phase with the negative peak of the extracellular theta field activity. The lower panel shows the spike discharges at a faster sweep. Note the membrane potential oscillations. C: upper panel shows the relationship between LIA and the simultaneously occurring spike discharges. The lower panel shows the discharges at a faster sweep. Note the absence of membrane potential oscillations.
25O ten near the positive peak of the MPOs; (6) at the onset of extracellular theta field activity above 4-5 Hz, there was a marked hyperpolarizing shift in the membrane potentials of phasic theta-off cells, accompanied by the appearance of MPOs without spike discharges. As the theta frequency slowed down there was a return to baseline membrane potential levels and spike discharges occurred near the positive peak of the MPOs. Phasic them-on cells (n -- 7 ) Five of these cells were located in the dentate layer and the remaining two were in the CAI layer. A representative example of a phasic theta-on celt is shown in Fig. 1. This cell was located in the dentate granule cell region but no measures were taken to positively identify it as a granule cell. Fig. IA shows a fast sweep of the spike discharge taken during spontaneous activity. The spike discharged on a slow depolarizing potential, had a spike duration of slightly over 2 ms and very little fast a~terhyperpolarization. The upper panel in Fig. I B shows the simultaneous recording of theta field activity from the dentate and the accompanying cell discharges, in this figure, extracellular negativity is up, indicating that the cell tended to discharge near extracellular peak negativity (verified by cross-correlation). The same phase relation was observed for the other 6 phasic theta-on cells. Also evi.
dent was the presence of rhythmic MPOs with the spike discharges occurring near the positivity of the MPOs. it was occasionally observed that MPOs could occur without spike discharges, in the total sample of phasic theta.on cells. The bottom panel in Fig. I B shows it faster time base of the cell discharge and the occurrence of MPOs slightly over 6 mV. in this study, MPOs of phasic theta.on cells ranged from 4 mV to almost 10 inV. At the onset of theta field activity, the
A.
B.
membrane potentials of all phasic theta-on cells typically showed a depolarizing shift in the range of 5-10 mV, compared to the average membrane potential levels measured during LIA. The top panel in Fig. 1C shows the cell discharging in a non-rhythmic pattern during the occurrence of LIA. The bottom panel shows the faster time base, taken during LIA. Note the absence of MPOs. The mean discharge of this cell during theta was 9.9 + 0.991 spikes/s and 5.7 + 0.9 spikes/s during LIA (significantly different at P < 0.001). The cell discharges and field activity were cross-correlated during theta but not during LIA and the auto-correlation histograms again showed rhythmicity during theta but not during LIA. Phasic theta-off cells (n = 4) Three of these cells were located in the CAI layer and the remaining cell was in the dentate layer. A representative example of a phasic theta-off cell is shown in Fig. 2. This cell was located in the CAI cell body region. Fig. 2A shows a fast sweep of the spike discharge taken during spontaneous activity. The resting membrane potential is indicated by the dotted line. The spike discharged on a slow depolarizing potential, had a spike duration of less than l ms and showed a large (8.0 mV) afterhyperpolarization. Fig. 2B is a continuous record of the spike train, showing the transition fi'om LIA to theta produced by a tail pinch (marked by arrow). The resting membrane potential is
indicated by the dotted line. At the onset of thcta, all 4 cells showed a hyperpolarizing shift of the membrane potential (approx. 5-10 mV) accompanied by the appearance of MPOs. The membrane potential then returned to baseline approximately 2 s later. As theta frequency declined, spikes occurred near the positive peak of the MPOs. With the transition to LIA, spike
r.
LIA
That.
Extro©ellular Dentate Layer
Int..llula,
cA1 Lay°r
I,S mV
F mY t Tall plncheh~j lo rnv
__] 10 rose©
~ Tall Pinch
J10 my 1 sac
Fig. 2. Intracellular cor;elates of a phasic theta-off cell, A: fast sweep of a spontaneous discharge, The resting membrane potential is indicated by the dotted line here and in B, The cell discharged on a slow depolarizing potential, Note the narrow spike width and presence of a large fast afterhyperpolarization. B: a continuous record of the spike discharges showing the transition from L I A to theta induced by a tail pinch (arrow). Note the cessation of discharges, the hypcrpolarizing shift and presence of membrane potential oscillations during theta. ]n this and all other phasic theta-off cells, the positive pe~,k of the MPOs were in phase with the positive peak of the extracellular field activity. C: details of the relationship between the extracellular field activity and spike discharges for the sequence shown in B.
251 discharges increased in frequency in a non-rhythmic pattern. F~g. 2C shows the simultaneous occurrence of field activity recorded from dentate region and spike discharges from the CA1 region. In this case, Fig, 2C corresponds exactly to Fig, 2B. Fig. 2C also illustrates that the positive peak of the MPOs was occurring on the positive phase of the extracellular theta field activity, since the dentate field was phase-reversed 180" to the local CA1 theta field activity. Indeed, all the MPOs of all 4 phasic theta-off cells exhibited the same phase relation, which was opposite to the phase relations demonstrated between the MPOs of phasic theta-on cells and the extracellular theta field activity.
Tonic theta-on cells (n
=
cantly different at P < 0.001). The cell discharges were not cross-correlated during either theta or LIA, nor did the auto-correlation histograms show any degree of rhythmicity for either condition.
Frequency and coherence relationships between field actit,ity and membrane potentials Phasic them-on cells All phasic theta-on cells (n = 7) exhibited MPOs during the occurrence of theta field activity but not during LIA. Fig. 4 shows a representative analysis of one of these cells. The top two panels show the frequency spectographs of the extracellular field activity and intracellular membrane potentials during theta (left side) and LIA (right side) expressed as amplitude in decibels (dB). The bottom panel gives the results of the coherence testing between the two signals, expressed as a correlation coefficient (rho). In this example both the extracellular theta and the membrane potential had a peak frequency at 3.5 Hz with a very high coherence. As indicated by the panels on the right, no peak frequency was observed for either the extracellular field activity or the membrane potential during LIA and the coherence was correspondingly low.
5)
Four of these cells were located in the dentate layer and one was located in the CAI layer. Fig. 3 shows a representative example of a tonic theta-on cell located in the dentate layer. Fig. 3A shows a fast sweep of the spike discharge taken during spontaneous activity. The spike discharged on a slow depolarizing potential, had a spike duration of slightly over 2 ms and did not have a large fast afterhyperpolarization. The upper panel in Fig. 3B shows the simultaneous recording of theta field activity from the dentate and the accompanying cell discharges. The cell discharged in an irregular pattern during theta field activity. As can be seen in the lower panel of Fig. 3B, the membrane potential did not show the occurrence of MPOs during theta activity. The top panel of Fig, 3C shows tl~.e reduced number of cell discharges occurring during LIA, again in a nonrhythmic pattern, The bottom panel shows the spike discharges during LIA, with a faster time base. Note again the complete absence of MPOs during LIA. The mean discharge of this cell during theta was 11.1 :t: 2.04
Phasic theta-off cells The frequency spectral and coherence analyses of the extracellular field activity accompanying the membrane potential of a representative phasic theta-off cell (n = 4) is given in Fig. 5. The upper panels on the left clearly indicate an identical peak frequency (4.25 Hz) for the extraceilular theta and the membrane potential. The high degree of coherence between the two signals is indicated in the lower left panel. The corresponding
spikes/s and 5.3 + 1.0 spikes/s during LIA (signifi-
''NN Spontaneous Thets
All
Spontaneous LIA Extraoellular Dentate Layer
I .5 mV
Intrscellular 1IF ~
v
-
•
w
-~-..,w-,r-,,
Ir
Dentate L a y e r r - ~ , , . -
v . . . . . . , ~ - ' - - ' . . .
-
i lOmV
1 el©
10 mV 10 mV
.eo,.,,-J . . . . . . . . . .
10 mse¢
I
250 meec
Fig. 3. lntracellular correlates of a tonic theta-on cell. A: fast sweep of a spontaneous discharge. Spike dischargeson a slow depolarizing potential. Note the wide spike and the absence of a large fast afterhyperpolarization.B: upper panel shows the relationshipbetween theta field activity and the simultaneouslyoccurringspike discharges(spikes chopped). The lower panel shows the spike dischargesat a faster sweep. Note the absence of membrane potential oscillations. C: upper panel shows the relationship between LIA and the simultaneouslyoccurring spike discharges.The lowerpanel shows the dischargesat a faster sweep. Note absenceof membrane potential oscillations.
252
Theta
Theta
LIA
| oIlill[IT~'n~
Cell
Phasic-Off
Phasic-On Cell
LIA
g d! Ill lit I : I ~
Illl'lllll~
