Characteristics of Plateau Activity During the Latent Period Prior to
Epileptiform Discharges in Slices From Rat Piriform Cortex
REZAN DEMIR, LEWIS B. HABERLY, AND MEYER B. JACKSON
Departments of Physiology and Anatomy and Center for Neuroscience, University of Wisconsin Medical School,
Madison, Wisconsin 53706
INTRODUCTION
Studies both in vivo (Croucher et al. 1988; Piredda and Gale
1985, 1986; Racine et al. 1988; Stevens et al. 1988) and in vitro
(Hoffman and Haberly 1991, 1996) have identified the piriform
cortex (PC) as a brain region with an unusually high seizure
susceptibility. Microelectrode recordings in brain slices have
identified the endopiriform nucleus (En), situated beneath deep
layer III of the PC, as a focus for the initiation of epileptiform
discharges (Hoffman and Haberly 1991, 1993). Subsequent
voltage imaging experiments showed that under some condiThe costs of publication of this article were defrayed in part by the payment
of page charges. The article must therefore be hereby marked “advertisement”
in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
1088
tions layer VI of the neighboring neocortex can also contribute
(Demir et al. 1998).
When slices of PC are stimulated electrically with currents
just above the threshold for generation of epileptiform discharges, a pronounced latency of up to 100 –150 ms elapses
before a discharge appears (Demir et al. 1998, 1999; Hoffman
and Haberly 1989, 1991). However, a slice is not completely
quiescent during this “latent period.” An accelerating buildup
of electrical activity can be seen at the site of discharge onset
(onset activity) (Demir et al. 1999; Hoffman and Haberly
1993). Similar activity in the hippocampus has provided support for models of discharge initiation by regenerative buildup
through recurrent excitatory synapses (Traub and Miles 1991;
Traub et al. 1989). Multisite voltage imaging has revealed an
additional form of latent period activity with very different
temporal qualities. A low-amplitude, sustained depolarization
(plateau activity) appeared at the beginning of the latent period
and continued until discharge onset (Demir et al. 1999). Plateau activity generally occurred at a site distinct from the site
of discharge onset, at the boundary region of En and deep layer
III of PC; onset activity generally occurred deeper within the
En. The finding that CoCl2 and kynurenic acid blocked discharges when applied to the sites of discharge onset and
plateau activity indicated that both types of latent period activity are required for the generation of interictal-like discharges (Demir et al. 1999).
In a recent study from this laboratory, the site of discharge
onset was systematically mapped in PC slices under various
conditions, including different rostrocaudal levels, different in
vitro models of epilepsy, and different sites of stimulation in
overlying PC (Demir et al. 1998). This study revealed variations in the site of onset, which imply that different structures
could contribute to discharge generation under different conditions. Therefore we conducted a parallel study to map sites of
plateau activity under the same range of conditions. Here we
report variations in the site of plateau activity that occur in
parallel with those seen in the site of onset. Under conditions
in which the adjacent neocortex contributed to discharge onset,
a parallel neocortical participation in plateau activity was also
seen. The coordination between the sites of plateau and onset
activity support the hypothesis that two distinct circuits feedback on one another to initiate epileptiform discharges. Electrical stimulation at the site of onset evoked low-amplitude
plateau activity over a broader region, indicating that there is
flexibility in this component of the discharge generating circuitry.
0022-3077/00 $5.00 Copyright © 2000 The American Physiological Society
Downloaded from http://jn.physiology.org/ by 10.220.33.4 on June 17, 2017
Demir, Rezan, Lewis B. Haberly, and Meyer B. Jackson. Characteristics of plateau activity during the latent period prior to epileptiform discharges in slices from rat piriform cortex. J. Neurophysiol.
83: 1088 –1098, 2000. The deep piriform region has an unusually high
seizure susceptibility. Voltage imaging previously located the sites of
epileptiform discharge onset in slices of rat piriform cortex and
revealed the spatiotemporal pattern of development of two types of
electrical activity during the latent period prior to discharge onset. A
ramplike depolarization (onset activity) appears at the site of discharge onset. Onset activity is preceded by a sustained low-amplitude
depolarization (plateau activity) at another site, which shows little if
any overlap with the site of onset. Because synaptic blockade at either
of these two sites blocks discharges, it was proposed that both forms
of latent period activity are necessary for the generation of epileptiform discharges and that the onset and plateau sites work together in
the amplification of electrical activity. The capacity for amplification
was examined here by studying subthreshold responses in slices of
piriform cortex using two different in vitro models of epilepsy. Under
some conditions electrically evoked responses showed a nonlinear
dependence on stimulus current, suggesting amplification by strong
polysynaptic excitatory responses. The sites of plateau and onset
activity were mapped for different in vitro models of epilepsy and
different sites of stimulation. These experiments showed that the site
of plateau activity expanded into deep layers of neighboring neocortex
in parallel with expansions of the onset site into neocortex. These
results provide further evidence that interactions between the sites of
onset and plateau activity play an important role in the initiation of
epileptiform discharges. The site of plateau activity showed little
variation with different stimulation sites in the piriform cortex, but
when stimulation was applied in the endopiriform nucleus (in the sites
of onset of plateau activity), plateau activity had a lower amplitude
and became distributed over a much wider area. These results indicate
that in the initiation of epileptiform discharges, the location of the
circuit that generates plateau activity is not rigidly defined but can
exhibit flexibility.
CHARACTERISTICS OF PLATEAU ACTIVITY IN PIRIFORM CORTEX
METHODS
Piriform cortex slices
In vitro models of epilepsy
We used the same two in vitro models as in our previous studies to
make slices generate interictal-like activity (Demir et al. 1998, 1999).
The first model, termed induction, consisted of an N-methyl-D-aspartate (NMDA) receptor– dependent transformation process (Hoffman
and Haberly 1989; Stasheff et al. 1989). Slices were subjected to a
transient period of spontaneous bursting in low-Cl⫺ ACSF (in which
93% of the Cl⫺ was replaced by isethionate) for 45–90 min at 34°C,
and then returned to normal ACSF for optical recordings. Epileptiform discharges could be evoked in these induced slices for up to 7 h
after return to normal saline. The second model, termed disinhibition,
involved blockade of GABAA receptors by bath application of bicuculline methiodide (5–10 ␮M; Sigma, St. Louis, MO).
Recordings were made at 32 ⫾ 2°C. Epileptiform events were
evoked by electrical stimulation of slices with a saline-filled glass
pipette with a tip diameter of 20 –50 ␮m. Current pulses (200 ␮s)
were delivered with a stimulus isolator. The threshold for discharge generation was determined by careful variation of stimulus
intensity.
Voltage imaging
The voltage-sensitive fluorescent dye RH414 (Molecular Probes,
Eugene, OR) was used to image epileptiform discharges. Before
imaging, slices were stained for 30 – 45 min in 200 ␮M RH414 in
ACSF bubbled with carbogen. A 464-element, hexagonally arranged,
photodiode fiber optic device (Chien and Pine 1991) served to measure fluorescence. The optical imaging setup used in this laboratory
follows that of Wu and Cohen (1993) and has been described previously (Demir et al. 1998). It employs an upright epifluorescent microscope (Reichert-Jung Diastar), equipped with a 100-W tungstenhalogen light source, a 475- to 565-nm band-pass excitation filter, a
590-nm dichroic mirror, and a 610-nm longpass emission filter. Imaging data were collected with a Zeiss ⫻5 Fluar objective (NA ⫽
0.25) that gave a distance between neighboring photodetector fields of
144 ␮m. The output of each photodetector was individually amplified
to a final level of 0.2 V/pA of photocurrent, digitized, and read into a
Pentium computer. Fluorescent signals were high-pass filtered with a
500-ms time constant and low-pass filtered with a corner frequency of
300 or 500 Hz. Trans-illuminated pictures of slices were taken with a
charge-coupled device camera, and read into a PC with a framegrabber (Data Translation, Marlboro, MA). Most video pictures were
recorded with a lower magnification ⫻2.5 objective to facilitate
visualization of anatomic landmarks.
Data acquisition and analysis
Optical signals were acquired and analyzed with the computer
program Neuroplex (OptImaging, Fairfield, CT). Traces displayed
individually represent averages of signals from two to seven neighboring photodiodes from a single trial. Fluorescence traces were
overlaid on video images with computer programs written in IDL
(Research Systems, Boulder, CO) (Demir et al. 1998; Jackson and
Scharfman 1996). The contours for the site of discharge onset were
drawn around the first photodetectors where the fluorescence
change surpassed 50 –70% of its maximum amplitude. This cutoff
excludes plateau activity, which has an amplitude ⬃25% of the
maximum amplitude of an interictal-like discharge. The contours
for plateau activity were prepared manually by identifying detectors where a sustained depolarization was clearly visible above
baseline noise. We have shown previously that plateau activity
falls off sharply at a discreet boundary (see Fig. 7 of Demir et al.
1999), so that in spite of the subjectivity of this manual method
there is ordinarily little error in demarcating the region of plateau
activity. In the present study, some of the fluorescence traces
showed overlap between local responses and plateau activity when
experiments were performed with stimulation in En at or near the
site of plateau activity (see RESULTS). In these cases there was more
uncertainty in mapping activity. This may have led to overestimation of the size of the region showing plateau activity within ⬃300
␮m of the stimulus electrode (see Fig. 6B).
RESULTS
Onset and plateau activity during the latent period
Previous work from this laboratory identified plateau and
onset activity as two specific forms of abnormal electrical
activity occurring during the latent period between electrical
stimulation and discharge onset in PC slices (Demir et al.
1999). These are illustrated with fluorescence traces from
induced (Fig. 1A) and disinhibited (Fig. 1B) slices. These
traces all show the interictal-like epileptiform discharges
typically seen with these two in vitro models of epilepsy. As
pointed out previously (Demir et al. 1998), discharges in the
disinhibited model are longer in duration as compared with
the induced model. Other differences were observed in the
sites of onset and plateau activity, as will be discussed in
detail below.
Onset activity was seen in both models [top traces (1) of
Fig. 1, A and B]. Onset activity was a ramplike depolarization at the site of onset, beginning ⬃20 – 60 ms after stimulation and culminating in an epileptiform discharge. The
middle traces of this figure (2) show plateau activity, with a
roughly constant, low-amplitude depolarization beginning
within ⬃10 ms of stimulation and continuing during the
entire latent period. The bottom traces (3) show local cortical responses to stimulation followed by discharges. Mapping of the plateau and onset sites will be presented below,
but in general, plateau activity was seen at a site distinct
from the site of discharge onset. Outside the sites of discharge onset and plateau activity, little electrical activity
Downloaded from http://jn.physiology.org/ by 10.220.33.4 on June 17, 2017
PC slices with a thickness of 350 ␮m were prepared from the brains
of male Sprague Dawley rats (175–240 g) as described previously
(Demir et al. 1998; Hoffman and Haberly 1991). Slices were cut with
a Vibratome in a near coronal plane perpendicular to the brain surface.
During slicing, subsequent storage, and recording, slices were maintained in artificial cerebrospinal fluid (ACSF) consisting of (in mM)
124 NaCl, 5 KCl, 26 NaHCO3, 1.2 KH2PO4, 2.4 CaCl2, 1.3 MgSO4,
and 10 glucose bubbled with 95% O2-5% CO2 (carbogen). Slices were
defined as anterior, intermediate, and posterior PC according to stereotaxic level (Demir et al. 1998). Anterior PC slices taken from
stereotaxic levels ⫺0.70 to 1.60 mm contained lateral olfactory tract,
claustrum, and agranular insular cortex (AI) and were further characterized by a thick layer Ia. Posterior slices were taken from stereotaxic
levels ⫺2.80 to ⫺1.80 mm and were distinguished by the presence of
anterior perirhinal cortex (PRha), and the absence of claustrum. Intermediate slices were taken from stereotaxic levels ⫺1.80 to ⫺0.70
mm and were characterized by an adjoining neocortex at the transition
between the AI and PRha (see Figs. 4, 5, and 7 for labeled video
pictures and sketches). Most of our previous work on plateau activity
was conducted in intermediate PC slices (Demir et al. 1999), in which
the claustrum was either small or absent.
After imaging experiments, all slices were fixed with 4% paraformaldehyde, sectioned at 60 ␮m, and Nissl stained, as described
previously (Demir et al. 1998). Identification of sites of plateau
activity were based on examination of these stained sections.
1089
1090
R. DEMIR, L. B. HABERLY, AND M. B. JACKSON
Downloaded from http://jn.physiology.org/ by 10.220.33.4 on June 17, 2017
FIG. 1. Epileptiform discharges and different forms of latent period activity prior to discharge onset. A: results from an induced
slice. B: results from a disinhibited slice. The video image on the left shows the piriform cortex (PC) slices (intermediate in both
cases) from which the optical recordings were taken. Traces on the right, from the indicated sites, show onset activity (1), plateau
activity (2), and rapidly decaying local responses in layer II of the overlying PC (3). All traces show epileptiform discharges,
although in B3 the amplitude is very small. Brackets indicate onset and plateau activity in the relevant traces. Dashed vertical line
indicates time of stimulus (94 ␮A in A and 74 ␮A in B, which was applied in layer Ib at sites indicated by the jagged line symbol).
Dotted vertical lines indicate the time of discharge onset at the site of onset. Each trace represents an average of 6 neighboring
detectors.
was seen during the latent period. At these sites, an epileptiform discharge either erupted from a flat baseline or immediately followed the rapidly decaying local response seen
near the site of stimulus. With shorter duration latent periods, the local response occasionally showed temporal overlap with the epileptiform discharge (compare traces A3 and
B3 of Fig. 1). In general, stimulus currents slightly above
threshold were used to make the latent period as long as
possible and thus provide optimal resolution of latent period
activity.
Properties of local responses
Responses from the site of plateau activity increased in
amplitude in a graded manner in response to increasing stimulus currents (Demir et al. 1999). In contrast, epileptiform
CHARACTERISTICS OF PLATEAU ACTIVITY IN PIRIFORM CORTEX
1091
discharges were all-or-none. Here we analyzed stimulus-response relationships during the latent period both at the site of
plateau activity and at other sites in the PC slice. Figure 2
shows optical signals from different sites in a disinhibited slice.
These sites include the site of discharge onset in the dorsalmost part of En (Fig. 2Ba), the site of plateau activity at the
border of En and deep layer III (Fig. 2Bb), a site in superficial
layer III (Fig. 2Bc), and a site in layer Ib near the site of
electrical stimulation (Fig. 2Bd). Responses evoked by three
subthreshold and one suprathreshold stimuli are displayed. The
amplitudes of subthreshold responses at these sites increased in
a nonlinear fashion with increasing stimulus intensity (Fig. 2:
columns 1 and 2). This behavior was observed not only at the
site of plateau activity (Fig. 2Bb), but also at other sites in the
slice wherever local responses could be seen (Fig. 2Bc; Fig. 3,
B1–B5). The only exception to this was responses very near the
stimulus site, where responses were predominantly monosynaptic excitatory synaptic potentials (and possibly axonal
spikes) evoked directly by the stimulus current (Fig. 2Bd).
Because the nonlinear increase in amplitude is steepest at
stimulus intensities subthreshold for epileptiform discharge
generation, it will be referred to here as “subthreshold nonlinear behavior.”
Stimulus-response plots illustrate these different behaviors
for control (Fig. 3A) and disinhibited (Fig. 3B) slices. The
control plots were fitted by linear functions (5/5 slices), and the
disinhibited plots were fitted by sigmoidal functions (5/5
slices). The linear stimulus-response behavior seen in control
slices stands in striking contrast to the nonlinear behavior of
disinhibited slices. In induced slices, stimulus-response plots
from the site of plateau activity were linear (Demir et al. 1999).
Thus in this particular respect, the induced model was closer to
control than to the disinhibited model.
Mapping of plateau activity: stimulation in PC
To determine the precise location of plateau activity, contours were drawn around the region where it was observed
(Fig. 4, A2–D2, blue shaded regions). In addition, contours
showing the sites of onset were included (Fig. 4, A2–D2, red
shaded regions) to emphasize the different site of plateau
activity. The slices were Nissl stained after imaging experiments, and the photographs are shown above each site map
(Fig. 4, A1–D1) to assist in relating the contours to anatomic
regions. In these experiments, epileptiform activity was evoked
by stimulus currents less 10% above the threshold for discharge generation. This provided the longest latency so that the
activity preceding discharge onset could be most reliably assessed. Stronger stimulus currents increase the amplitude of
plateau activity. The size of both the plateau and onset sites are
also increased, but remain in the same basic location (Demir et
al. 1999).
Previous work has shown that in induced slices, the site of
discharge onset consists of the dorsal-most part of the En and
the adjacent parts of layer VI of neocortex (AI in anterior slices
and PRha in posterior slices) (Demir et al. 1998). In both
induced anterior and posterior slices, the site of plateau activity
included the boundary between En and PC deep layer III (Fig.
Downloaded from http://jn.physiology.org/ by 10.220.33.4 on June 17, 2017
FIG. 2. Responses to subthreshold stimuli. A: a video image of a disinhibited intermediate slice with arrows pointing to the
locations from which fluorescence traces in B were taken. B: traces show responses from the site of discharge onset (a), the site
of plateau activity (b), superficial layer III of the overlying PC (c), and the site of stimulus in layer Ib of piriform cortex (d).
Stimulus currents are indicated above. Note the disproportionate increase in responses when the stimulus intensity was increased
from 55 to 65 ␮A (b and c). Traces represent averages of 2 neighboring detectors from the indicated sites. The site of stimulus is
indicated by the jagged line symbol.
1092
R. DEMIR, L. B. HABERLY, AND M. B. JACKSON
4, A2 and C2, blue shaded regions). In anterior slices plateau
activity extended into layers V and VI of the adjoining AI (n ⫽
8; Fig. 4A2), and in posterior slices plateau activity included
small portions of layers V and VI of adjoining PRha (n ⫽ 18;
Fig. 4C2). The contribution of deep layers of PRha to plateau
activity in posterior slices was somewhat less than that of deep
layers of AI in anterior slices (compare Figs. 4, A2 with C2).
The extension of the site of plateau activity into AI or PRha in
induced slices parallels the extension of the site of onset into
the same neocortical structures.
In contrast to induced anterior slices, in disinhibited anterior
slices the site of plateau activity was smaller and confined to a
portion of En bordering deep layer III of PC (n ⫽ 6; Fig. 4B2).
In disinhibited posterior slices (n ⫽ 5; Fig. 4D2) plateau
activity was not confined to En, but was also seen in a substantial part of neighboring deep layer III of PC. As reported
previously (Demir et al. 1998), the sites of onset in disinhibited
anterior and posterior slices consisted of a site in En, without
any involvement of deep layers of neighboring AI or PRha (red
regions in Fig. 4, B2 and D2). The site of plateau activity
showed a similar trend of not extending into adjoining parts of
neocortex. Thus in disinhibited slices neither onset nor plateau
activity appear in neighboring neocortex, and this was the case
for both anterior and posterior slices. Comparison of the results
from induced and disinhibited slices suggests that even as these
sites vary, plateau activity and onset activity maintain a characteristic spatial relationship during the initiation of an epileptiform discharge.
In contrast to anterior and posterior slices, in intermediate
slices the sites of plateau activity and onset activity were not
influenced by choice of model. In induced (n ⫽ 20; Fig. 5) and
disinhibited (n ⫽ 6) intermediate slices, plateau activity was
confined to a portion of En and a part of adjacent PC layer III.
In these slices, discharge onset was seen at essentially the same
site for both models of epilepsy used in this study (Demir et al.
1998). Likewise, the site of plateau activity did not differ
significantly between induced versus disinhibited intermediate
slices. A summary of the site of plateau activity in different
planes of slicing along the rostrocaudal axis is presented in
Table 1.
For a given slice, the site of plateau activity was relatively
constant as long as the site of stimulation remained in PC. In
experiments where two (n ⫽ 6) or three (n ⫽ 6) different
stimulation sites within PC were compared, the site of plateau
Downloaded from http://jn.physiology.org/ by 10.220.33.4 on June 17, 2017
FIG. 3. Linear and nonlinear stimulus-response plots. Peak amplitude of electrically evoked fluorescence change was plotted
against stimulus current in control (A) and disinhibited (B) slices. Measurements were taken from a site in layer III. Both A and
B show data from the same slice, before and after disinhibition with bicuculline. Stimulation was applied in layer Ib. The control
plot could be fitted by a line, but the disinhibition plot was nonlinear (B). Linear or sigmoidal functions were drawn through the
data to emphasize the different behaviors. Note the different scales for the 2 plots; bicuculline produced a large increase in
amplitude for the same stimulus current. The threshold for the epileptiform discharge was 55 ␮A in this experiment. Traces on the
right side represent responses used in B. Note the disproportionate increase in peak amplitude from B2 to B3.
CHARACTERISTICS OF PLATEAU ACTIVITY IN PIRIFORM CORTEX
1093
Downloaded from http://jn.physiology.org/ by 10.220.33.4 on June 17, 2017
FIG. 4. Mapping sites of onset and plateau activity. Slices were Nissl stained (A1–
D1) after site mapping (A2–D2). Contours in
A2–D2 indicate the sites of discharge onset
(red) and plateau activity (blue) for induced
anterior (A2), disinhibited anterior (B2), induced posterior (C2), and disinhibited posterior (D2) slices. In induced slices plateau
activity extends into layers V and VI of the
neighboring AI (A2) and PRha (C2) in anterior and posterior slices, respectively. In contrast, in disinhibited slices (B2 and D2) plateau activity did not extend into neocortex.
Sites of stimulus are marked by the white
jagged lines. Threshold stimulus currents
were 70 ␮A (A2), 117 ␮A (B2), 600 ␮A
(C2), and 120 ␮A (D2). CPu, caudate-putamen; ec, external capsule; Cl, claustrum; En,
endopiriform nucleus; LOT, lateral olfactory
tract; AI, agranular insular cortex; RF, rhinal
fissure; PC, piriform cortex; PRha, anterior
perirhinal cortex. Arrowheads mark the border between PC and AI in A1 and B1 and
between PC and PRha in C1 and D1.
1094
R. DEMIR, L. B. HABERLY, AND M. B. JACKSON
activity showed a variability of only one to three detector fields
(150 – 450 ␮m; Fig. 5, B1–B3). In general, with stimulation in
overlying PC, this small variability in the site of plateau
activity was comparable to that described previously for the
site of onset (Demir et al. 1998). Further, the variations in
plateau site and onset site tended to occur in parallel, again
supporting the idea that the sites maintain a characteristic
spatial relationship.
Although most experiments showed plateau activity in the
regions indicated in Table 1, the results of some experiments
deviated from this pattern. In one of eight induced anterior
slices, the site of plateau activity was located primarily in a
portion of the En at the boundary with deep layer III, with little
contribution from AI and deep layer III of PC. In this slice, the
site of onset was localized in the central portion of En rather
than the dorsal-most part and, unlike most induced anterior
slices, did not include layer VI of AI. In 14 of 18 induced
posterior slices, the site of plateau activity was as indicated in
Table 1, but in two slices deep layer III of PC was excluded. In
these slices the site of plateau activity was confined to a deep
part of PRha and a portion of En just below the boundary with
deep layer III of PC. In the two remaining slices the site of
plateau activity included a portion of En and deep layer III of
the overlying PC without any involvement of deep layers of
PRha. Four of 20 induced intermediate slices failed to show
plateau activity at the usual site consisting of adjoining parts of
Downloaded from http://jn.physiology.org/ by 10.220.33.4 on June 17, 2017
FIG. 5. Sites of plateau activity with different stimulation sites in PC. A: a photograph of a Nissl-stained intermediate PC
slice (see Fig. 4 for abbreviations). An arrowhead marks the border of PC and AI/
PRha. B1–B3: sites of onset (red) and plateau
(blue) activity in an induced slice are indicated by the shaded contours. Epileptiform
activity was evoked by stimulation in layer
Ib (B2), superficial layer III (B2), and deep
layer III (B3). Sites of stimulus are marked
by the bold jagged lines. The threshold stimulus currents were 55 ␮A (B1), 75 ␮A (B2),
and 40 ␮A (B3). Note that the site of plateau
activity did not change significantly with different stimulus sites in the overlying PC.
CHARACTERISTICS OF PLATEAU ACTIVITY IN PIRIFORM CORTEX
TABLE
1.
Sites of plateau activity
Anterior PC slices
Intermediate PC slices
Posterior PC slices
1095
DISCUSSION
Induction
Disinhibition
Adjoining parts of En
and layer V and VI
of AI; variable
involvement of
deep layer III of
PC (n ⫽ 7)
Adjoining parts of En
and deep layer III
of PC (n ⫽ 16)
Adjoining parts of En
and deep layer III
of PC with or
without a small
part of layer V of
PRha (n ⫽ 16)
Portion of En; variable
involvement of deep
layer III of PC
(n ⫽ 6)
Adjoining parts of En
and deep layer III of
PC (n ⫽ 5)
Adjoining parts of En
and deep layer III of
PC (n ⫽ 5)
En and deep layer III of PC. In these slices, plateau activity was
mainly concentrated in a part of En just below the border with
deep layer III of PC.
In disinhibited slices, there was much less variability. Out of
17 disinhibited slices from different planes, only 1 (an intermediate slice) showed a small deviation from the typical pattern. In this slice, plateau activity was contained within the En
near its boundary with deep layer III of the overlying PC.
Mapping of plateau activity: stimulation in En
When stimulation was applied in En, plateau activity was
observed, but it was much lower in amplitude. Furthermore,
the region where it was observed was larger, far more variable,
and shifted relative to the location seen with stimulation in the
overlying PC (Fig. 6). Because the threshold for generating a
discharge was generally about 10-fold lower for stimulation at
the sites of onset and plateau activity compared with stimulation in PC, the graded local responses were small. Plateau
activity was correspondingly also small but was clearly visible
in seven of nine slices (8 intermediate and 1 posterior). In these
slices, plateau activity showed an expansion toward the site of
onset. With either stimulation at the site of discharge onset
(Fig. 6B) or the site of plateau activity (Fig. 6C), plateau
activity extended from the original site seen with stimulation in
PC (Fig. 6A) toward the site of discharge onset. Although the
presence of local responses in the region created some uncertainty in mapping plateau activity (for example, Fig. 6B, trace
1), the extension of plateau activity in the ventral direction was
at a considerable distance from the stimulus electrode so that
mapping was unambiguous (Fig. 6B, trace 2). With these deep
stimulation sites there was substantial overlap between the sites
of onset and plateau activity. By contrast, stimulation in overlying PC produced plateau activity and onset at sites showing
little or no overlap (Demir et al. 1999).
The site of onset was also examined for discharges evoked
by stimulation in En and was found to be very similar to the
site of onset seen with stimulation in overlying PC, for all
conditions tested (Demir et al. 1998). Thus the location of
onset activity was more consistent and well defined than the
location of plateau activity.
Downloaded from http://jn.physiology.org/ by 10.220.33.4 on June 17, 2017
Summary of the results shown in Figs. 4 and 5 on the site of plateau activity.
Some variations in these sites were seen, as discussed in the text.
This study investigated subthreshold responses in the induced and disinhibition models of epilepsy. The stimulusresponse plots for subthreshold stimulus currents were highly
nonlinear in disinhibited slices, both at the site of plateau
activity as well as most other sites. The amplitude of the
fluorescence change increased steeply over a narrow range of
stimulus currents below the threshold for epileptiform discharge generation (Fig. 3B). Epileptiform behavior is intrinsically nonlinear, in that a discharge is evoked when the stimulus
current exceeds a characteristic threshold. The present observation suggests that there is an additional form of nonlinearity
that can be associated with epileptiform behavior. This type of
behavior was never observed in control slices (Fig. 3A), where
stimulus-response relations were linear, but it was also not seen
in subthreshold responses in induced slices (Demir et al. 1999,
Fig. 4C). Thus subthreshold nonlinearity is not necessary for
epileptiform behavior, and its presence in the disinhibited
model may reflect a difference in the underlying mechanisms
of discharge generation. Both models clearly increase neuronal
excitability, but induction increases excitability without introducing nonlinear responses to stimuli below the discharge
threshold.
Previous work has shown that the induction and disinhibition models also differ in discharge onset site, and in the
duration of the interictal-like discharge (Demir et al. 1998). It
is likely that in the induced model the changes brought on by
spontaneous bursting in low-Cl⫺ are more restricted spatially.
Because GABAA receptors are ubiquitous, and are blocked
throughout the slice by bath application of bicuculline, disinhibition will lead to a more global change in excitability, not
only in En but also in overlying PC. In induced slices, linear
behavior was seen even at the plateau site, but subthreshold
responses at the site of onset deeper in En were too small to
assess. Thus we cannot rule out the possibility that localized
changes in excitability in the induced model lead to subthreshold nonlinearity at a highly restricted location.
One possible explanation for the subthreshold nonlinear
behavior seen in disinhibited slices is positive feedback
through recurrent excitatory synaptic pathways. Such fiber
systems have been described in the PC, both between pyramidal cells and between pyramidal and deep multipolar cells
(Behan and Haberly 1999; Luskin and Price 1983). Disinhibition would presumably unmask di- and/or polysynaptic potentials, which can be evoked in PC slices (Tseng and Haberly
1989). From this perspective, the more linear stimulus-response behavior seen in layer I (Fig. 2, bottom row of traces)
could reflect differences in the strength of synaptic inputs to
apical versus basal dendrites of pyramidal cells. Over short
distances, excitatory inputs from association fibers in the PC
project predominantly to pyramidal cell basal dendrites located
in layer III (Haberly and Presto 1986; Rodriguez and Haberly
1989). Subthreshold nonlinear behavior was clearly seen in this
region (Figs. 2B and 3B). Pyramidal cell axons typically traverse longer distances before ascending to layer I to innervate
apical dendrites. These projections are less likely to be preserved in slices, so direct activation of association axons by
stimulation in layer Ib would evoke mostly monosynaptic
events, and these show a linear dependence on stimulus current.
1096
R. DEMIR, L. B. HABERLY, AND M. B. JACKSON
In this work a detailed mapping study was conducted in
slices of PC to locate the sites where plateau activity can be
seen before the onset of epileptiform discharges. These sites
are summarized in Table 1 and sketched on drawings of
slices in Fig. 7 to make their spatial relationships clear.
Although the site of plateau activity varied with the different
conditions tested, it always included a portion of En at the
boundary with deep layer III of the overlying PC. Additional
sites where plateau activity was observed depended on the
rostrocaudal level of the slice, as well as the in vitro model
used to generate epileptiform behavior. A portion of deep
layer III at the boundary with En and deep layers of neocortex also showed plateau activity under some conditions.
These regions thus possess properties that allow them to
produce a weak and roughly constant depolarization. It is
not known whether this depolarization depends on intrinsic
regenerating potentials or reciprocal excitatory synaptic interactions, but the neurons in these regions must have the
cellular properties or synaptic circuitry capable of sustaining plateau activity for the entire latent period, which can
last ⬎100 ms.
Participation of neocortical structures adjoining PC had
been reported previously in the onset of epileptiform discharges (Demir et al. 1998). In induced anterior and posterior PC slices the sites of onset not only consisted of the
dorsal portion of En, but also a portion of deep layer VI of
adjoining neocortex (either AI or PRha, for anterior and
posterior slices, respectively). Here we found that as the site
of onset expanded to include deep layers of the neighboring
cortex, there was a parallel expansion of plateau activity
Downloaded from http://jn.physiology.org/ by 10.220.33.4 on June 17, 2017
FIG. 6. Plateau activity evoked by stimulation at the sites of discharge onset and
plateau activity. Video images (A–C) taken
during recording show contours for the sites
of onset (red) and plateau activity (blue),
with electrical stimulation in layer II of PC
(glass electrode; A), in the site of onset (B),
and in the site of plateau activity (C). Sites of
onset and plateau activity were 1st determined for stimulation in layer II (A), and
epileptiform events were then evoked by
electrical stimulation at those sites (B and C).
The site of onset remained roughly the same
(red contours), but the site of plateau activity
(blue contours) changed. Overlap between
local responses and plateau activity (trace 1
of B) created some uncertainty in mapping
plateau activity, so the blue contour close to
the stimulus site is not as accurate as elsewhere in the figure (see METHODS). Threshold
stimulus currents were 250 ␮A (A), 14 ␮A
(B), and 16 ␮A (C) (note the lower threshold
at the sites of onset and plateau activity).
Although signals during the latent period
were small due to the lower thresholds at
these sites, onset and plateau activity could
still be distinguished in the fluorescence
traces numbered 1 and 2 in B and C. Traces
were taken from the same sites in A–C, as
indicated by the numbers 1 and 2 in the video
images. Each trace is an average of 6 neighboring detectors from a single trial.
CHARACTERISTICS OF PLATEAU ACTIVITY IN PIRIFORM CORTEX
1097
toward deep layers of neocortex (in Fig. 7, compare anterior-induced with anterior-disinhibited, and posterior-induced with posterior-disinhibited). It is noteworthy that
participation of adjacent neocortex in both onset and plateau
activity is seen only in the induced model of epilepsy, and
not in the disinhibited model. The finding of different degrees of neocortical participation in discharge onset led us to
hypothesize previously that there are regional variations in
synaptic circuitry (Demir et al. 1998). Thus if GABAA
receptors play a greater role in limiting synchronous firing
in En, this could account for preferential onset in En in the
disinhibited model. The present finding of a similar exclusion of plateau activity from adjoining neocortex in the
disinhibited model argues for a similar role for GABAA
receptors at the boundary region between En and deep layer
III of PC (the principle site of plateau activity). The properties of the adjoining neocortex that allow it to contribute
to discharge onset in the induced model may be related to
those properties that allow it to contribute to plateau activity.
The sites of both onset and plateau activity varied, and it is
significant that the variation of these two sites occurred in
parallel. This is consistent with the idea that functional interactions between the two circuits generate these forms of electrical activity. The parallel spread into neocortex preserves a
characteristic spatial relation, and this spatial relation may be
necessary for the communication between these two circuits
that allows them to amplify electrical activity and initiate an
Downloaded from http://jn.physiology.org/ by 10.220.33.4 on June 17, 2017
FIG. 7. Sites of onset and plateau activity. Solid blue contours indicate the site of
plateau activity on sketches of PC slices
from different stereotaxic levels for the 2
models used in this study. Results from 4
typical experiments are shown for each
condition. The dashed red ellipses show
the site of onset reported by Demir et al.
(1998). Many of the imaging experiments
used in that previous study were also used
in the present study. Note that although the
contours demarcating sites of onset and
plateau activity overlap when results from
multiple experiments are combined, in individual experiments there was little if any
overlap (Figs. 4 and 5). All of the data
summarized here are for stimulation in PC.
Sites were more variable and overlapped
more when stimulation was applied in En
(Fig. 6). Dashed line indicates the boundary between PC and neocortex. For label
definitions, see Fig. 4.
1098
R. DEMIR, L. B. HABERLY, AND M. B. JACKSON
Address for reprint requests: M. Jackson, Dept. of Physiology, SMI 127,
University of Wisconsin Medical School, 1300 University Ave., Madison, WI
53706.
Received 30 July 1999; accepted in final form 18 October 1999.
REFERENCES
BEHAN, M. AND HABERLY, L. B. Intrinsic and efferent connections of the
endopiriform nucleus in rat. J. Comp. Neurol. 408: 532–548, 1999.
CHIEN, C.-B. AND PINE, J. Voltage-sensitive dye recording of action potentials
and synaptic potentials from sympathetic microcultures. Biophys. J. 60:
697–711, 1991.
CROUCHER, M. J., BRADFORD, H. F., SUNTER, D. C., AND WATKINS, J. C.
Inhibition of the development of electrical kindling of the prepiriform cortex
by daily focal injections of excitatory amino acid antagonists. Eur. J. Pharmacol. 152: 29 –38, 1988.
DEMIR, R., HABERLY, L. B., AND JACKSON, M. B. Voltage imaging of epileptiform activity in slices from rat piriform cortex: onset and propagation.
J. Neurophysiol. 80: 2727–2742, 1998.
DEMIR, R., HABERLY, L. B., AND JACKSON, M. B. Sustained and accelerating
activity at two discrete sites generate epileptiform discharges in slices of
piriform cortex. J. Neurosci. 19: 1294 –1306, 1999.
HABERLY, L. B. AND PRESTO, S. Ultrastructural analysis of synaptic relationships of intracellular stained pyramidal cell axons in piriform cortex.
J. Comp. Neurol. 248: 464 – 474, 1986.
HOFFMAN, W. H. AND HABERLY, L. B. Bursting induces persistent all-or-none
EPSPs by an NMDA-dependent process in piriform cortex. J. Neurosci. 9:
206 –215, 1989.
HOFFMAN, W. H. AND HABERLY, L. B. Bursting-induced epileptiform EPSPs in
slices of piriform cortex are generated by deep cells. J. Neurosci. 11:
2021–2031, 1991.
HOFFMAN, W. H. AND HABERLY, L. B. Role of synaptic excitation in the
generation of bursting-induced epileptiform potentials in the endopiriform
nucleus and piriform cortex. J. Neurophysiol. 70: 2550 –2561, 1993.
HOFFMAN, W. H. AND HABERLY, L. B. Kindling-induced epileptiform potentials
in piriform cortex slices originate in the underlying endopiriform nucleus.
J. Neurophysiol. 76: 1430 –1438, 1996.
JACKSON, M. B. AND SCHARFMAN, H. E. Positive feedback from hilar mossy
cells to granule cells in the dentate gyrus revealed by voltage-sensitive dye
and microelectrode recording. J. Neurophysiol. 76: 601– 616, 1996.
LUSKIN, M. B. AND PRICE, J. L. The topographic organization of associational
fibers of the olfactory system in the rat, including centrifugal fibers to the
olfactory bulb. J. Comp. Neurol. 216: 264 –291, 1983.
PIREDDA, S. AND GALE, K. A crucial epileptogenic site in the deep prepiriform
cortex. Nature 317: 623– 625, 1985.
PIREDDA, S. AND GALE, K. Role of excitatory amino acid transmission in the
genesis of seizures elicited from the deep prepiriform cortex. Brain Res.
377: 205–210, 1986.
RACINE, R. J., MOSHER, M., AND KAIRISS, E. W. The role of the pyriform cortex
in the generation of interictal spikes in the kindled preparation. Brain Res.
454: 251–263, 1988.
RODRIGUEZ, R. AND HABERLY, L. B. Analysis of synaptic events in the opossum
piriform cortex with improved current source density analysis techniques.
J. Neurophysiol. 61: 702–718, 1989.
STASHEFF, S. F., ANDERSON, W. W., CLARK, S., AND WILSON, W. A. NMDA
antagonists differentiate epileptogenesis from seizure expression in an in
vitro model. Science 245: 648 – 651, 1989.
STEVENS, J. R., PHILIPS, I., AND BEAUREPAIRE, R. Gamma-vinyl GABA in
endopiriform area suppresses kindled amygdala seizures. Epilepsia 29:
404 – 411, 1988.
TRAUB, R. D. AND MILES, R. Neuronal Networks of the Hippocampus. New
York: Cambridge Univ. Press, 1991, 281 pp.
TRAUB, R. D., MILES, R., AND WONG, R.K.S. Model of the origin of rhythmic
population oscillations in the hippocampal slice. Science 243: 1319–1325, 1989.
TSENG, G. F. AND HABERLY, L. B. Deep neurons in piriform cortex. I. Morphology and synaptically evoked responses including a unique high-amplitude paired shock facilitation. J. Neurophysiol. 62: 369 –385, 1989.
WU, J. Y. AND COHEN, L. B. Fast multisite optical measurements of membrane
potential. In: Fluorescent and Luminescent Probes for Biological Activity,
edited by W. T. Mason. London: Academic, 1993, p. 389 – 404.
Downloaded from http://jn.physiology.org/ by 10.220.33.4 on June 17, 2017
epileptiform discharge. This supports the hypothesis that the
sites of onset and plateau activity work together to generate an
epileptiform discharge, presumably through reciprocal excitatory synaptic pathways (Demir et al. 1999).
The variation in the site of plateau activity was far more
pronounced when the stimulus electrode was moved to the
site of discharge onset in En. The threshold was very low in
these experiments, so that plateau activity had a much lower
amplitude and became distributed over much of the En. This
suggests that the properties that are necessary for plateau
activity are distributed widely through the En and neighboring areas of PC and neocortex. By contrast, the site of onset
was generally smaller and remained at essentially the same
location (Fig. 6). What may make the site of onset unique is
its strong reciprocal excitatory synaptic connections with a
large surrounding area. Plateau activity can then appear at
variable locations within this larger region, according to
where the stimulus is applied. But the discharge onset will
occur at the one location receiving strong excitatory input
from all these areas.