| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Department of Psychology and Trinity College Institute of Neuroscience, University of Dublin, Trinity College, Dublin 2, Ireland
Submitted 22 August 2002; accepted in final form 3 March 2003
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMETHODSRESULTSDISCUSSIONACKNOWLEDGMENTSREFERENCES |
|---|
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONACKNOWLEDGMENTSREFERENCES |
|---|
,
1995). The anatomy, physiology,
and functions of the dentate gyrus and hippocampus have been extensively
investigated (for reviews, Eichenbaum
1999; O'Keefe
1979,
1999;
O'Mara 1995). The subiculum,
by contrast, has received little attention
(O'Mara et al. 2001), although
it is the major output structure of the hippocampus proper, receiving a
massive input from hippocampal area CA1
(Amaral et al. 1991;
Witter et al. 1989) and
returning a lesser oligosynaptic projection to CA1
(Commins et al. 2002). The
subiculum plays a pivotal role in epilepsy
(Dreier and Heinemann 1991),
Alzheimer's disease (Braak and Braak
1991), as well as in spatial and mnemonic processing, as indicated
by lesion studies (Schenk and Morris
1985) and neuronal recordings
(Sharp and Green 1994). The
CA1-subiculum projection sustains long-term potentiation (LTP) (Commins et al.
1998a,b)
and is very sensitive to both behavioral
(Commins and O'Mara 2000) and
endotoxic (Commins et al.
2001) stress.
Several studies have attempted to classify cell types within the subiculum,
using intracellular (Behr et al.
1996; Greene and Totterdell
1997; Mason 2000;
Stewart 1997;
Stewart and Wong 1993;
Taube 1993) and patch-clamping
recording in vitro (Staff et al.
2000), and unit recording in vivo
(Gigg et al. 2000;
Sharp and Green 1994). There
is broad agreement regarding a distinction between bursting cells (which fire
2–6 fast action potentials with approximately 5-ms interspike intervals
riding on a slow potential) and regular spiking cells (which fire single
action potentials with interspike intervals in the 60- to 160-ms range).
Bursting in the in vitro preparation is not abolished by blocking excitatory
synaptic transmission and has thus been described as an intrinsic membrane
property of subicular pyramidal cells (Mason 1993); the situation in the in
vivo preparation may, however, be somewhat different. Bursting cells can fire
single action potentials by shifting the cell membrane potential to more
depolarized levels (Mason 1993; Stewart
and Wong 1993). Staff et al.
(2000) suggest it may be most
appropriate to describe the firing properties of subicular principal neurons
as lying on a "propensity to burst" continuum; this propensity may
be regulated by Ca2+ tail currents since the amplitude
of these currents is correlated with the strength of bursting across subicular
neurons (Jung et al. 2001).
Thus subicular pyramidal neurons might form a single neuronal class, sharing a
burst-generating mechanism that is stronger in some cells than others.
Hippocampal theta consists of oscillations in EEG of an approximately
sinusoidal waveform of regular amplitude with a frequency range of 6–12
Hz, whereas irregular activity is typically slower than theta (1–4 Hz)
and has been subdivided into large amplitude irregular activity (LIA) and
small amplitude irregular activity (SIA)
(Buzsaki and Vanderwolf 1985;
Vanderwolf et al. 1975). Much
of LIA appears to be sharp wave activity (large amplitude EEG irregular waves
reflecting increased excitability of intrahippocampal circuitry, resulting
from temporary disinhibition from afferent control;
Buzsaki 1986). There is little
information available regarding the differing types of unit activity in the
subiculum and on the relationship between subicular unit activity and
subicular population activity. We here attempt to clarify the in vivo
classification of subicular cells and to describe subicular EEG by recording
unit activity and EEG in unrestrained freely moving rats in a simple setting,
as well as correlating both types of activity with the simultaneous behavior
of the animal.
|
METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONACKNOWLEDGMENTSREFERENCES |
|---|
Five adult male Wistar rats (BioResources Unit, University of Dublin), weighing 300–400 g on arrival in the laboratory were used as subjects. Rats were housed singly in a temperature-controlled laminar airflow cupboard and maintained on a 12:12 h light-dark cycle (light: 0800–2000 h). All testing was carried out during the light phase. All subjects were well handled prior to the experiment. Local, national, and European Union regulations governing animal welfare were obeyed, and numbers used were minimized consistent with statistical power.
Apparatus
All recordings for this experiment were conducted while the rats occupied a small, familiar square chamber in the experimental room (side 25 cm, height 30 cm). The chamber rested on the floor of the experimental room, close to the experimenter so that the ongoing behavior of the rat could be easily observed; it occupied the same position for all recording sessions. Two 12-V bulbs attached to the ceiling illuminated the experimental room.
Microdrive and electrodes
The microdrive assembly was based on that of John O'Keefe (e.g.,
O'Keefe and Speakman 1987;
Speakman and O'Keefe 1990) and
consisted of a vertical post (17 gauge) carrying a hollow screw, which when
turned, allows the vertical movement of an acrylic "nut." The nut
itself is prevented from turning by a post mounted on, and parallel to, the
screw post. The nut holds a 14-pin connecting plug behind and a hollow guide
cannula (25 gauge) in front that contains the electrode wires. A small metal
turner contacting flanges on the top of the screw turns the screw. To ensure
accuracy in the movement, a spring maintains the contact of the lower
screw-bearing surface against a fixed bush with sufficient pressure to prevent
vertical displacement. The electrodes can be lowered in steps of approximately
50 µm by making one-eighth turns. Recording electrodes consisted of eight
Nichrome wires (80:20 nickel/chromium alloy; formvar-insulated; 25 µm bare
wire diam; Advent Research Materials) twisted together to form two bundles of
four electrodes. The bundles were cut flat to expose the tips at approximately
the same level. Both bundles were inserted into one stainless steel guide
cannula (25 gauge) fixed to the microdrive. All units were recorded using
single electrode techniques (operating in differential mode to reduce noise
common to both channels, i.e., a quiet electrode was used as the differential
electrode); single electrodes were used because of previous work suggesting
that no advantage in unit separation is conferred by using multiple electrode
configurations (Sharp 1997;
unpublished data).
Surgery
Rats were anesthetized with intraperitoneal injections of sodium
pentobarbital (60 mg/kg Sagatal, Rhone-Merieux), followed by atropine 0.5
ml/50 kg, (Bimeda), with supplementary doses of sodium pentobarbital
administered if necessary. They were placed in a stereotaxic apparatus (Kopf),
the skull was exposed, and a small burr hole was drilled over the electrode
implantation site. Seven bone screws were inserted around the exposed skull to
anchor the microdrive. One screw acted as the ground connection. The
microdrive electrode bundles were aimed at the dorsal subiculum: AP,
–6.8 mm (relative to bregma); L, 4.0 mm; DV, 2.4 mm
(Paxinos and Watson 1996).
Once the electrodes had been lowered to the required depth, the small exposed
area between the cannula and the surrounding skull was packed with sterile gel
(Granugel, Convatec), and the entire apparatus was cemented to the skull with
dental acrylic (Associated Dental). After surgery, the wound area was dusted
with antibiotic powder (Cicatrin, Wellcome Ireland), and the rat was kept
under observation for several hours. The animal was allowed to recover for at
least 7 days before experimentation began. Electrodes were implanted in the
left hemisphere in four animals and the right hemisphere in one animal; no
hemispheric differences in unit characteristics were noted.
Recording techniques
A headstage (NB Labs) containing field effect transistors (FETs), one for each channel, was plugged into the microdrive connector for recording purposes. The output from the headstage was fed to a filter-amplifier (Model 1700, A-M Systems): each spike channel signal was amplified (x10,000) and filtered (band-pass 300 Hz–5 kHz); EEG channels were amplified (x1,000) and filtered (band-pass 1–500 Hz). The amplifier output was fed into a loudspeaker, an oscilloscope, and a computerized AD conversion system (DataWave Technologies). The waveforms of each event (spikes) were digitized, typically at 33 kHz or 32 data points per spike, and produced 1 ms of activity for each spike. The DataWave system also received the EEG output from the amplifier and stored this information as a continuous record at a typical sampling rate of 150 Hz. "Event flags" were used to record the time and type of behavioral or other event by pressing a key on the keyboard when the event occurred.
Protocol
Each session typically lasted approximately 20 min. During this time, the rat was not required to perform any task. Usually rats were very active for the first few minutes, exploring the chamber rapidly, sniffing and rearing frequently. By the end of the session, this very active behavior usually subsided to "alert, still" or "quiet" behaviors, punctuated regularly by brief "alert, moving" periods. The experimenter recorded the ongoing behavior of the rat by pressing a key on the keyboard that was saved to file as an event-flag together with a time-stamp, which were then used to construct peri-event time histograms (PETHs) of unit firing with respect to the event flag-coded behaviors. PETHs were constructed by binning the unit activity with respect to the event flag time (as a 0 reference), within a user-defined time range; all like event flags were collected to produce a cumulative histogram.
Data analysis
Output ASCII files were exported from DataWave into custom-written analysis software (Matlab-5, MathWorks). Spike sorting was conducted using a template-matching algorithm with conservative criteria for acceptance. EEG was visualized using a data plotter that also showed simultaneous spike and event flag data; as such, portions of the EEG trace could be cropped for further analysis. After spike sorting, individual unit data were further processed to generate inter-spike interval histograms (ISIHs), autocorrelation histograms (ACHs), PETHs based on the event flags, and raster plots. ISIHs, ACHs, and raster plots were used, together with the electrophysiological characteristics of the units (spike duration and spike amplitude), to classify each unit according to known subicular cell types. The firing characteristics of the putative bursting cells were subjected to further analysis. EEG was analyzed using discrete Fourier transforms.
Unit separation
Satisfactory subicular unit separation is difficult to achieve possibly
owing to the number of simultaneously active neurons in the subiculum
(Barnes et al. 1990). We
conducted unit separation using a custom-written template-matching program
(Matlab), which uses a template-matching algorithm measuring the so-called
"city block distance" (CBD) (Wheeler 1998) between each spike and
a template constructed from a sample of spikes. This is expressed as
 |
). The algorithm
was quite conservative, which meant that we rejected just over one-half of all
potential units. Most units were rejected for having too few spikes to be
classifiable, often because the algorithm separated the firing of putative
single cells into more than one unit. The algorithm did cope quite well with
the well-known phenomenon in putative bursting cells of decreasing successive
spike amplitudes as evidenced by the significant difference between successive
spike amplitudes reported in RESULTS. Electrophysiological analysis
ISIHs were constructed by summing (or "binning," in 1-ms bins) the time intervals between consecutive spikes (with a maximum interval of 500 ms); ACHs were constructed by binning the time intervals between consecutive spikes where spiken, spiken+1, spiken+2, etc. occur at 0 ms (with a maximum interval of 500 ms). The amplitude (µV) and duration (ms) of each unit were also calculated. Simple mean rate (in Hz) was calculated for each unit by dividing the total number of spikes in a session generated by that unit by the session time (s).
Bursting unit analysis
The putative bursting units were subjected to further analysis to
characterize this particular mode of firing. We defined a burst as a series of
spikes in which each ISI was 
10 ms. Bursts consisted, therefore, of a
minimum of two spikes. ISIs of 10 ms are at the top-end of the range of ISIs
expected of bursting units (based on previously defined bursting
characteristics, e.g., Sharp and Green
1994) and well below the ISIs typically displayed by regular
spiking units.
The following characteristics were analyzed: 1) the number of spikes per burst; 2) the bursting ISI (ISIb: i.e., the ISIs between those spikes that, under our formal definition, comprised a burst); 3) the burst duration; and 4) the inter-burst interval (IBI). We also examined the relationship between successive spikes in each burst, specifically any differences in spike amplitude or ISIb. Finally, we calculated a "propensity to burst" measure as an index of the predominance of bursting units to fire in a bursting mode.
Statistical analysis
Where appropriate, differences between firing rates were analyzed
parametrically with one-way ANOVA or t-test for independent samples.
Paired t-test were used for comparisons of unit firing rates before
and after the onset of a particular behavior (as recorded with the event flags
and displayed in the PETHs), as the pre- and postflag differences were
normally distributed. All statistics were calculated using Statistical Package
for the Social Sciences (SPSS) software (
= 0.05).
EEG analysis
EEG was visualized using a data replay plotter that also allowed the simultaneous viewing of spike times and event flags; as such, portions of the EEG trace could be cropped for further analysis. Continuous records of unit and EEG activity were also visualized. EEG was analyzed using discrete Fourier transforms to reveal constituent sinusoids of different frequencies, from which power spectra were plotted to reveal dominant frequencies present in the EEG.
EEG/unit relationship
The theta phase angle at which each spike fired was calculated in two stages: first, the entire EEG record for each session was scanned for periods of theta oscillation; second, the phase angle at which spikes fired during these theta oscillations was calculated. The entire algorithm was as follows.
Once the EEG phase angle of each spike was found, the methods of
directional statistics were used to calculate the mean phase and mean
resultant length for each unit as described by Mardia and Jupp
(2000) (see
King et al. 1998 for a
previous application of these methods to spike and EEG data). The mean
resultant length is used here as a measure of dispersion of the phase angles
of each spike around the mean for each unit. To test for uniformity in the
spread of phase angles (and hence to look for units whose spikes showed
"nonuniform" phase-related firing), Rayleigh's test was used for
each unit (as described by Mardia and Jupp
2000).
Histological analysis
Following the experiment, rats were deeply anesthetized using sodium
pentobarbitol and then perfused. Brains were removed, stored in 4%
formaldehyde for several days, then frozen to –20°C, sectioned on a
cryostat, and stained with cresyl violet. The sections were subsequently
analyzed to verify electrode placements in dorsal subiculum by reference to
the atlas of Paxinos and Watson
(1996).
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONACKNOWLEDGMENTSREFERENCES |
|---|
|
Unit classification
Of 130 subicular units, only 61 (47%) could unequivocally be assigned to a class (2.35 ± 0.21 SE units per session; range, 1–4; all recorded on a single electrode). Eighty-six percent of the remaining units had too few spikes to be classified confidently (ISIHs and ACHs become impossible to interpret with fewer than approximately 200 spikes), and these were discarded from any subsequent analyses. The remaining 14% appeared to be the result of poor separations rather than different unit classes and were also discarded. Four unit classes were defined using electrophysiological and firing characteristics: bursting units, regular spiking units, theta-modulated units, and fast spiking units (see Table 1). There was no obvious relationship between recording site and unit classification. The units were classed as follows:
|
): a
comparison of the ISIHs and ACHs for each unit failed to clearly distinguish
bursting units (Sharp and Green
1994) from depolarized bursters
(Sharp and Green 1994;
Stewart and Wong 1993). Hence
all units showing evidence of burst firing were assigned to a single bursting
unit class. The results of the burst analysis are presented later.
|
Unit electrophysiological characteristics and spike train analysis
Table 1 displays the electrophysiological measures of the accepted units separated by unit type. The mean firing rate for all accepted units was 1.36 Hz, reflecting either our relatively conservative unit separation techniques or the behavioral conditions in this study. One-way ANOVAs with Tukey's honestly significant difference (HSD) post hoc tests were conducted to test for any significant differences in firing characteristics between the unit classes. The fast spiking unit class fired at a significantly higher rate and had a significantly smaller spike width than all the other unit classes (both P < 0.0001). The theta-modulated unit class had a significantly smaller spike height than the bursting unit and regular spiking unit classes (P < 0.001 and P < 0.01, respectively). There were no other significant differences.
Bursting unit analysis
The results of the bursting unit analysis (showing are the mean of all the individual unit minimum (min) and maximum (max) values for each measure; all values are mean ± SE) were as follows.
The examination of successive spikes within each burst showed that successive spike amplitudes decreased in general, with a mean change of –3.66 ± 0.71% (i.e., between spiken and spiken+1); a one-sample t-test (test value = 0) of the change in successive spike amplitudes was highly significant (P < 10–6), indicating that, although small, the decrease in spike amplitudes over successive spikes in a burst was a robust finding for the bursting unit class. The successive ISIbs did not decrease in general, with a mean of –0.22 ± 0.12 ms (i.e., between spiken and spiken+1); a one-sample t-test (test value = 0) of the change in successive ISIbs was not significant.
Since the bursting unit class showed much variation in firing
characteristics, we calculated a measure of propensity to burst for each
bursting unit (see also Staff et al.
2000), defined as the percentage of intervals that occurred in the
burst ISI range (1–9 ms). For example, a bursting unit containing 2,000
spikes of which the number occurring in the burst ISI range is 250 would give
a propensity to burst value (PtB) of (250/1999) x 100, or approximately
12.5%. We found that PtBs ranged between 1.5 and 33% (mean of 9.3 ±
1.8%), where PtBs close to 0% indicate relatively little bursting and PtBs
above 10–20% indicate relatively larger degrees of bursting.
Figure 3 displays a histogram
of PtB values for the 25 bursting units. Although the majority of PtBs are
small, there is a wide spread of values across the range. Interestingly, there
was no correlation between a bursting unit's PtB value and its overall firing
rate (Spearman's rho: 0.247, P > 0.2), suggesting that the
differences in the appearance of ACHs for bursting units (e.g.,
Fig. 2) are not owing to
differences in cell firing rates.
|
EEG analysis
EEG from 26 of the 30 sessions (EEG was not recorded or was unstable during 4 sessions) were analyzed using fast Fourier transforms, and a periodogram of power against frequency was then plotted to reveal dominant frequencies (see Fig. 4, A–E, for examples of periodograms). Examples of subicular EEG can be seen in Figs. 5, 6, 7. Two common patterns of EEG activity emerged, represented by unimodal and multimodal periodograms. In 15/26 cases (57.6%), a multimodal distribution of frequencies occurred as shown in Fig. 4, A–C. Two or more peaks occur, with dominant peaks occurring in the 1- to 3-Hz range (presumably reflecting LIA) and 6- to 10-Hz range (presumably reflecting theta); other smaller peaks at higher frequencies were sometimes apparent (e.g., Fig. 4C has a small peak around 11 Hz). The ratio of the size of the LIA peak and the size of the theta peak varied. In the remaining 11/26 cases (42.3%), a unimodal distribution was evident, with the dominant peak in either the theta range (Fig. 4D) or the LIA range (Fig. 4E). Most theta oscillations had a frequency of about 6–7 Hz.
|
|
|
|
EEG/unit relationship
Figures 5 and
6 show unit firing for each
class superimposed on simultaneous EEG (theta oscillations in
Fig. 5; LIA including sharp
waves in Fig. 6). Whereas
bursting, regular spiking, theta-modulated, and fast spiking units all showed
at least some evidence of firing in phase with theta, it was not a strong
relationship (and significant for only some units: see Phase-related
firing). This latter point probably explains why most ACHs do not show
much evidence of theta modulation. As with hippocampal neurons, all unit
classes fired during LIA, probably related to sharp wave activity
(Buzsaki 1986). Interestingly,
units fire during some but not all sharp waves; for example, the
theta-modulated unit in Fig.
6C fires strongly in the first half of the trace during
sharp waves, but does not fire at all in the second.
Phase-related firing
A total of 18/56 units showed phase-locking with theta, i.e., showed a
preferential phase as assessed using Rayleigh's test (P < 0.05; 5
units, 1 theta-modulated unit, and 4 fast spiking units could not be assessed
because EEG was not recorded or was unstable during their acquisition). Of
these, 9/25 (36%) of the bursting units, 3/12 (25%) of the regular spiking
units, 6/18 (33%) of the theta-modulated units, and 0/1 of the fast spiking
units were phase-locked to theta. Of the units that showed phase-locking to
theta, the mean phase of EEG for bursting units was 314° (n = 9),
for regular spiking units was 245° (n = 3), and for
theta-modulated units was 8° (n = 6). While each of these
selected units showed a preferential phase of firing, no unit class,
however, showed a preferential phase (see
Fig. 8). The grand mean phase
angle for all phase-locked units combined was 318° (n = 18;
Fig. 9). Using a two-sample
Watson-Williams test (Mardia and Jupp
2000), a significant difference was found between the mean phase
angles of the bursting unit class and those of the theta-modulated unit class
(P < 0.05).
|
|
Of the units that showed phase-locking to theta, the mean ± SE of each unit class's mean resultant length was as follows: bursting units, 0.46 ± 0.05 (n = 9); regular spiking units, 0.31 ± 0.1 (n = 3); and theta-modulated units, 0.24 ± 0.04 (n = 18). The bursting units showed significantly greater mean resultant lengths than the theta-modulated class (Student's t-test, P < 0.05), meaning that bursting units showed a greater degree of phase modulation than theta-modulated units. All other comparisons were nonsignificant.
Behavioral analysis
UNITS. PETHs revealed few significant differences between pre- and post-flag firing rates (15 significant differences across 14 units). PETHs may not be the best way to investigate firing rate differences between behavioral states because it is difficult to determine precisely when those states shift. All significant differences (P < 0.001–0.05) appeared to be related to changes in arousal levels or related to movement since they were associated with either the alert, still, alert, moving, or rearing flags and were not specific to a unit class.
EEG. Subicular EEG showed a similar relationship with behavior to hippocampal EEG. Theta rhythms were invariably evident during alert, active, rearing, and grooming behaviors. In relation to rearing, theta rhythms would often continue after the initial rearing movement, presumably related to sniffing the air. Both theta and irregular activity appeared intermittently during alert, still and quiet behaviors, although it was the irregular activity and sharp waves that dominated these periods (see Fig. 6).
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONACKNOWLEDGMENTSREFERENCES |
|---|
;
Muller et al. 1991;
Sharp and Green 1994).
Furthermore, EEG has not been previously recorded simultaneously with
subicular unit firing. As well as classifying subicular units according to
their firing patterns, we have studied the broad relationship between
subicular unit firing, EEG, and behavioral state. We conclude that subicular
units can be separated into at least four classes (bursting, regular
spiking, theta-modulated, and fast spiking) on the basis of the
electrophysiological characteristics of their firing rate, spike duration,
relationship with simultaneously recorded EEG, and spiketrain time
characteristics. We have also found that subicular bursting units show large
variation in their propensity to burst (see also
Staff et al. 2000). The
analysis of unit firing against behavioral state revealed few significant
differences between preand post-event flag firing rates, and these appeared to
be related to arousal levels or movement.
The ACHs for bursting, regular spiking, and the fast spiking unit classes
are similar to those of Sharp and Green
(1994); although the bursting
units described here show more variation than Sharp and Green
(1994), it is possible that
their "depolarized bursters" are classified here as bursters.
Sharp and Green (1994) do not
report theta-modulated units, but did not record EEG in their recordings, so
these units may have been assigned to their nonbursting class. It is difficult
to say whether or not we encountered what Sharp and Green call depolarized
bursters: some of the bursting units recorded here show many ISIs longer than
the typical burst ISIs, but whether this means that these units are shifting
from a bursting to a regular spiking mode is another matter, especially as a
recent patch-clamp study did not find any cells showing this shift
(Staff et al. 2000). Sharp and
Green do state that their classification is tentative, and only that it
"is possible that [the depolarized bursters] pattern [of firing]
resulted from the oscillation of bursting cells between a bursting and
nonbursting mode, as described by Stewart and Wong
(1993)."
Staff et al. (2000) suggest
that subicular bursting units may lie on a propensity to burst continuum,
given that they find a range of cells, from weakly to strongly bursting cells,
which exhibit different levels of bursting activity, and that despite this
variation in burst propensity, these neurons do not differ in any other
measured parameter. The variation in firing characteristics of our bursting
units is an interesting finding in itself and the propensity to burst
hypothesis may explain the variation in bursting unit ACH shape that we have
found. Bursting is hypothesized to increase the probability of synaptic
vesicle release and as such may increase the reliability of synaptic
transmission (Commins et al. 1998; Lisman
1997). This mechanistic advantage of bursting may also convey an
informational advantage—place cells appear to have smaller place fields
when single spikes are ignored and only bursts are taken into account
(Otto et al. 1991). If the
subiculum does indeed mediate hippocampal-cortical interaction, reliability in
transferring information to downstream cortical circuits would be expected.
While this suggests why the subiculum contains a large proportion of bursting
units to nonbursting units, it does not explain why bursting neurons within
the subiculum should vary according to their propensity to burst.
The analysis of the bursting units revealed properties in common with
bursting cells recorded in other subicular studies, as well as in other areas
in the hippocampal formation. The mean number of spikes and the mean ISIs in
each burst are both very similar to values reported in other subicular single
unit studies (e.g., Sharp and Green
1994, ISIb range: 2–4 ms;
Staff et al. 2000, spikes per
burst: 2–6 and ISIb range: 4–5 ms;
Taube 1993, spikes per burst:
3–5). Successive spike amplitudes within each burst decreased in
amplitude, another property in common with bursting units throughout the
hippocampal formation, and reported also for units in the subiculum (Mason
1993; Staff et al. 2000;
Stewart 1993). Successive within-burst ISIbs, however, did not, in
general, change, unlike one report of increases in ISIbs during
later spikes in bursts (Staff et al.
2000), perhaps reflecting differences between in vivo and in vitro
recordings.
Fast spiking units are very probably interneurons: they fire at very high
rates and have narrow spikes, known features of hippocampal interneurons
(e.g., O'Keefe 1979). We also
find them in a similar proportion to previous reported proportions (e.g.,
Greene and Totterdell 1997,
their fast spiking units; Sharp and Green
1994, their theta units). The theta-modulated unit class is only
evident in the in vivo preparation with simultaneous EEG recordings (the
latter because theta-modulated units were distinguished from regular spiking
units by increased firing when a theta rhythm was present in the EEG). Of
course, in vitro studies cannot help to explain this finding because many
inputs to the subiculum are severed in the process of making a slice,
including a major source of thetarhythm activity in the hippocampus, the
septum (Butcher and Woolf
1986). No other in vivo subicular study
(Barnes et al. 1990; Sharp
1997,
1999a,
b,
c;
Sharp and Green 1994) reports
simultaneously recorded EEGs, and hence no other study would be able to
discriminate theta-modulated units from regular spiking units. Theta-modulated
units increase their firing rates when a theta rhythm appears in the EEG, as
do putative interneurons (e.g., O'Keefe
1979); however, the large spike width of these units, which was
not significantly different from the bursting unit and regular spiking unit
spike widths (presumed pyramidal cells) but was significantly different from
the fast spiking unit spike width, led us to assign these units to their own
class. It will be interesting to see if other in vivo studies also report this
unit class.
The analysis of spike phase angles showed that, as with other structures in
the hippocampal formation, a proportion of units from each unit class were
phase-locked to theta. A theta-rhythmic signal resonates in hippocampal
circuits (probably spreading further afield too, to the rest of the
hippocampal formation and beyond; Vertes
et al. 2001) and is believed to entrain unit firing in separate
structures to a common rhythm for information processing purposes (cf.
Gray and Singer 1989),
possibly involving mnemonic functions of the circuit. The fact that many units
in the hippocampal formation are phase-locked to theta rhythms suggests that
information processing in these structures is shared between the individual
structures of the hippocampal formation and reveals something of the dynamics
of the hippocampal network. This phase-locking is not uniform, however. King
et al. (1998) found that 79.7%
of medial septal/diagonal band of broca (MS/DBB) neurons (59/74) recorded in
the freely moving rat were phase-locked to theta; Dragoi et al.
(1999) report even greater
phase locking (but with a much smaller sample: 23/25 units, 92%). Similarly
high proportions of cells show theta phase-locking in the median raphe (80%,
Viana Di Prisco et al. 2002).
The proportion of cells showing phase-locking in area CA1 appears to be
slightly less; Csicsvari et al.
(1999) report that 65%
(241/369) neurons in the urethane-anesthetized rat show phase-locking. By
contrast, in the subicular recordings reported here, 18/56 (approximately 32%
of classifiable cells) were phase-locked to theta, a much smaller proportion
than in prior areas, even those such as CA1 or the MS/DBB, from which the
subiculum takes major inputs (O'Mara et
al. 2001). The relative lack of entrainment of subicular neurons
by this important intrinsic rhythm is suggestive of a limit to which theta
might be capable of affecting both subicular and hippocampal information
processing more generally.
|
ACKNOWLEDGMENTS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONACKNOWLEDGMENTSREFERENCES |
|---|
The Wellcome Trust supported this work.
|
FOOTNOTES |
|---|
Address for reprint requests: S. M. O'Mara, Dept. of Psychology, and Trinity College Institute of Neuroscience, University of Dublin, Trinity College, Dublin 2, Ireland (E-mail: smomara{at}tcd.ie).
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONACKNOWLEDGMENTSREFERENCES |
|---|
Amaral DG and Witter MP. The three-dimensional organisation of the hippocampal formation: a review of anatomical data. Neuroscience 31: 571–591, 1989.[Web of Science][Medline]
Amaral DG and Witter MP. Hippocampal formation. In: The Rat Nervous System, 2nd ed, edited by Paxinos G. New York: Academic Press, 1995, p. 247–291.
Barnes CA, McNaughton B, Mizumori SJ, Leonard BW, and Lin LH. Comparison of spatial and temporal characteristics of neuronal activity in sequential stages of hippocampal processing. Prog Brain Res 83: 287–300, 1990.[Web of Science][Medline]
Behr J, Empson D, Schmitz T, and Heinemann U. Electrophysiological properties of rat subicular neurons in vitro. Neurosci Lett 220: 41–44, 1996.[Web of Science][Medline]
Braak H and Braak E. Neuropathological staging of Alzheimer-related changes. Acta Neuropathol (Berl) 82: 239–259, 1991.[Medline]
Butcher LL and Woolf NJ. Central cholinergic systems: synopsis of anatomy and overview of physiology and pathology. In: The Biological Substrates of Alzheimer's Disease, edited by Scheibel AB and Wechsler AF. New York: Academic Press, 1986, p. 182–201.
Buzsaki G. Hippocampal sharp waves: their origin and significance. Brain Res 398: 242–252, 1986.[Web of Science][Medline]
Buzsaki G, Leung LW, and Vanderwolf CH. Cellular bases of hippocampal EEG in the behaving rat. Brain Res 287: 139–171, 1983.[Medline]
Buzsaki G and Vanderwolf CH. Electrical Activity of the Archicortex. Budapest: Akademia Kiado, 1985.
Commins S, Aggleton JP, and O'Mara SM. Physiological evidence for a possible projection from dorsal subiculum to hippocampal area CA1. Exp Brain Res 146: 155–160, 2002.[Web of Science][Medline]
Commins S, Gigg J, Anderson M, and O'Mara SM. The projection from hippocampal area CA1 to the subiculum sustains long-term potentiation. Neuroreport 9: 847–850, 1998a.[Web of Science][Medline]
Commins S, Gigg J, Anderson M, and O'Mara SM. Interactions between paired-pulse facilitation and long-term potentiation in the projection from hippocampal area CA1 to the subiculum. Neuroreport 9: 4109–4113, 1998b.[Web of Science][Medline]
Commins S and O'Mara SM. Interactions between paired-pulse facilitation, low-frequency stimulation and behavioural stress in the pathway from hippocampal area CA1 to the subiculum: dissociation of baseline synaptic transmission from paired-pulse facilitation and depression of the same pathway. Psychobiology 28: 1–11, 2000.
Commins S, O'Neill LAJ, and O'Mara SM. The effects of the bacterial endotoxin lipopolysaccharide on synaptic transmission and synaptic plasticity in the hippocampal area CA1-subiculum pathway in vivo. Neuroscience 102: 273–280, 2001.[Web of Science][Medline]
Csicsvari J,
Hirase H, Czurko A, Mamiya A, and Buzsaki G. Oscillatory coupling of
hippocampal pyramidal cells and interneurons in the behaving rat. J
Neurosci 19:
274–287, 1999.
Dragoi G, Carpi
D, Recce M, Csicsvari J, and Buzsaki G. Interactions between hippocampus
and medial septum during sharp waves and theta oscillation in the behaving
rat. J Neurosci 19:
6191–6199, 1999.
Dreier JP and Heinemann U. Regional and time dependent variations of low Mg2+ induced epileptiform activity in rat temporal cortex slices. Exp Brain Res 87: 581–596, 1991.[Web of Science][Medline]
Eichenbaum H. The hippocampus and mechanisms of declarative memory. Behav Brain Res 103: 123–133, 1999.[Web of Science][Medline]
Gigg J, Finch DM, and O'Mara SM. Responses of morphologically characterized rat subicular neurons to stimulation of CA1 and lateral entorhinal cortex in vivo. Brain Res 884: 35–50, 2000.[Web of Science][Medline]
Gray CM and
Singer W. Stimulus-specific neuronal oscillations in orientation columns
of cat visual cortex. Proc Natl Acad Sci USA
86: 1698–1702,
1989.
Greene JRT and Totterdell S. Morphology and distribution of electrophysiologically defined classes of pyramidal and nonpyramidal neurons in the rat ventral subiculum in vitro. J Comp Neurol 380: 395–408, 1997.[Web of Science][Medline]
Jung HY, Staff
NP, and Spruston N. Action potential bursting in subicular pyramidal
neurons is driven by a calcium tail current. J
Neurosci 21:
3312–3321, 2001.
King C, Recce M, and O'Keefe J. The rhythmicity of cells in the medial septum/diagonal band of Broca in the awake freely moving rat: relationships with behaviour and hippocampal theta. Eur J Neurol 10: 464–477, 1998.
Lisman JE. Bursts as a unit of neural information: making unreliable synapses reliable. Trends Neurosci 20: 38–43, 1997.[Web of Science][Medline]
Mardia KV and Jupp PE. Directional Statistics. Chichester, England: John Wiley, 2000.
Mason A. Electrophysiology and burst-firing of rat subicular pyramidal neurons in vitro: a comparison with CA1. Brain Res 600: 174–178, 2000.
Muller RU, Kubie JL, Bostock EM, Taube JS, and Quirk GJ. Spatial firing correlates of neurons in the hippocampal formation of freely moving rats. In: Brain and Space, edited by Paillard J. Oxford: Oxford, 1991, p. 177–191.
O'Keefe J. A review of the hippocampal place cells. Prog Neurobiol 13: 419–439, 1979.[Web of Science][Medline]
O'Keefe J. Do hippocampal pyramidal cells signal non-spatial as well as spatial information? Hippocampus 9: 352–364, 1999.[Web of Science][Medline]
O'Keefe J and Speakman A. Single unit activity in the rat hippocampus during a spatial memory task. Exp Brain Res 68: 1–27, 1987.[Web of Science][Medline]
O'Mara SM. Spatially selective firing properties of hippocampal formation neurons in rodents and primates. Prog Neurobiol 45: 253–274, 1995.[Web of Science][Medline]
O'Mara SM, Commins S, Anderson M, and Gigg J. The subiculum: a review of form, physiology and function. Prog Neurobiol 64: 129–155, 2001.[Web of Science][Medline]
Otto T, Eichenbaum H, Wiener SI, and Wible CG. Learning-related patterns of CA1 spike trains parallel stimulation parameters optimal for inducing hippocampal long-term potentiation. Hippocampus 1: 181–192, 1991.[Medline]
Paxinos G and Watson C. The Rat Brain in Stereotaxic Coordinates, 3rd ed. San Diego, CA: Academic Press, 1996.
Schenk F and Morris RGM. Dissociation between components of spatial memory in rats after recovery from the effects of retrohippocampal lesions. Exp Brain Res 58: 11–28, 1985.[Web of Science][Medline]
Sharp PE. Subicular cells generate similar spatial firing patterns in two geometrically and visually distinctive environments: comparison with hippocampal place cells. Behav Brain Res 85: 71–92, 1997.[Web of Science][Medline]
Sharp PE. Subicular place cells expand or contract their spatial firing pattern to fit the size of the environment in an open field but not in the presence of barriers: comparison with hippocampal place cells. Behav Neurosci 113: 643–652, 1999a.[Web of Science][Medline]
Sharp PE. Complimentary roles for hippocampal versus subicular/entorhinal place cells in coding place, context, and events. Hippocampus 9: 432–443, 1999b.[Web of Science][Medline]
Sharp PE. Comparison of the timing of hippocampal and subicular spatial signals: implications for path integration. Hippocampus 9: 158–172, 1999c.[Web of Science][Medline]
Sharp PE and Green C. Spatial correlates of firing patterns of single cells in the subiculum of the freely moving rat. J Neurosci 14: 2339–2356, 1994.[Abstract]
Speakman A and O'Keefe J. Hippocampal complex spike cells do not change their place fields if the goal is moved within a cue controlled environment. Eur J Neurosci 2: 544–555, 1990.[Web of Science][Medline]
Staff NP, Jung
H-Y, Thiagarajan T, Yao M, and Spruston N. Resting and active properties
of pyramidal neurons in subiculum and CA1 of rat hippocampus. J
Neurophysiol 84:
2398–2408, 2000.
Stewart M. Antidromic and orthodromic responses by subicular neurons in rat brain slices. Brain Res 769: 71–85, 1997.[Web of Science][Medline]
Stewart M and
Wong RKS. Intrinsic properties and evoked responses of guinea pig
subicular neurons in vitro. J Neurophys
70: 232–245,
1993.
Taube JS. Electrophysiological properties of neurons in the rat subiculum in vitro. Exp Brain Res 96: 304–318, 1993.[Web of Science][Medline]
Vanderwolf CH, Kramis R, Gillespie LA, and Bland B. Hippocampal rhythmic slow activity and neo-cortical low-voltage fast activity: relations to behaviour. In: The Hippocampus, edited by Isaacson RL and Pribram KH. New York: Plenum Press, vol. 2, 1975.
Vertes RP, Albo Z, and Viani Di Prisco G. Theta-rhythmically firing neurons in the anterior thalamus: implications for mnemonic functions of Papez's circuit. Neurosci 104: 619–625, 2001.[Web of Science][Medline]
Viana Di Prisco G, Albo Z, Vertes RP, and Kocsis B. Discharge properties of neurons of the median raphe nucleus during hippocampal theta rhythm in the rat. Exp Brain Res 145: 383–394, 2002.[Web of Science][Medline]
Witter MP, Groenewegen HJ, Lopes da Silva FH, and Lohman AHM. Functional organisation of the extrinsic and intrinsic circuitry of the parahippocampal region. Prog Neurobiol 33: 161–253, 1989.[Web of Science][Medline]
This article has been cited by other articles:
|
|
 |
|
  W. B. Wilent and D. A. Nitz Discrete Place Fields of Hippocampal Formation Interneurons J Neurophysiol, June 1, 2007; 97(6): 4152 - 4161. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. P. Wu, H. L. Huang, M. N. Asl, J. W. He, J. Gillis, F. K. Skinner, and L. Zhang Spontaneous rhythmic field potentials of isolated mouse hippocampal-subicular-entorhinal cortices in vitro J. Physiol., October 15, 2006; 576(2): 457 - 476. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Visit Other APS Journals Online |
