Abstract

In a subset of patients with epilepsy, patterned visual stimuli can trigger clinical seizures. The etiology of this phenomenon, and the complex interaction between functional architecture and epilepsy, were investigated in ferret visual cortex. Optical imaging of intrinsic signals was used to visualize maps of orientation, ocular dominance and spatial frequency. Acute interictal spike foci were then induced within V1 using focal iontophoresis of bicuculline methiodide and optically mapped during presentation of patterned visual stimuli. We found that specific orientations and spatial frequencies could preferentially trigger epileptiform events, depending on the location of the epicenter of the epileptic focus within the columnar architecture of visual cortex. These data support a cortical etiology of the clinical phenomenon of pattern-sensitive epilepsy. We were not able to demonstrate a spatial correlation between the functional architecture maps and the topography of the epileptic focus. These findings implicate short-range rather than long-range horizontal excitatory connections in the lateral spread of interictal spikes, which may be specific to the epilepsy model of acute focal disinhibition. Orientation and spatial frequency maps were severely disturbed in the region of the focus but were unaltered in the surrounding cortex. Thus, optical imaging of intrinsic signals can be used to simultaneously map epilepsy and normal functional anatomy with high spatial resolution.

Intoduction

One of the best clinical examples of the interaction between functional brain architecture and epilepsy is photosensitive epilepsy. This phenomenon occurs in up to 15% of patients with epilepsy, of which as many as 70% patients will demonstrate a phenomenon called ‘pattern-sensitive epilepsy’, in which epileptic events are triggered by patterned visual stimuli (Binnie and Wilkins, 1998; Dorothée and Trenité, 1998). Clinical studies have demonstrated that the most effective stimuli consist of alternating black and white linear gratings oscillating in a direction orthogonal to their orientation, and the most important parameters are spatial frequency, binocular presentation and luminance (Binnie and Wilkins, 1998). In rare cases of astigmatism, one particular orientation is more likely to trigger an event than all other orientations (Chartrian et al., 1970; Wilkins et al., 1979). The pathophysiology of pattern-sensitive epilepsy is unknown, but the etiology is thought to arise from dysfunction in an inhibitory network surrounding a trigger zone comprising a population of orientation selective neurons in visual cortex (Wilkins et al., 1979). Laboratory studies of experimentally induced interictal spike (IIS) foci in mammalian visual cortex using focal disinhibition, however, have only succeeded in triggering IISs following changes in luminance but not orientation-specific stimulation (Ebersole and Levine, 1975; Gabor and Scobey, 1975, 1980; Gabor et al., 1979; Ebersole and Chatt, 1981). Although the etiology of this phenomenon remains unclear, the ability of physiological stimulation to trigger paroxysmal discharges implies a direct functional interaction between the normal cortical architecture and the genesis of epileptiform events.

One of the fundamental organizing principals of the neocortex is that neurons with similar response properties are organized into columnar modules. These functional columns have anatomic correlates found in the arborization of thalamocortical afferents, as well as the interconnectedness of iso-orientation domains via long-range horizontal excitatory connections (Mountcastle, 1957, 1982; Hubel and Wiesel, 1962, 1963; Gilbert and Wiesel, 1989; Malach et al., 1993; Weliky et al., 1995; Bosking et al., 1997; Dalva et al., 1997). For many years, investigators have postulated that the cortical architecture, in particular the functional column, may play an important role in the initiation and propagation of epileptic events (Ebersole and Levine, 1975; Gabor et al., 1979; Reichenthal and Hocherman, 1979; Gabor and Scobey, 1980; Ebersole and Chatt, 1981). By focally applying small doses of epileptogenic agents and performing cortical transections to isolate slabs of cortex in vivo, attempts have been made to define the minimal volume of cortex necessary to generate an IIS (Morrell and Hambery, 1969; Reichenthal and Hocherman, 1977; Lueders et al., 1980; Bashir and Holmes, 1993). Since the cortical column might possess sufficient recurrent excitatory inter-connections to facilitate epileptogenesis (Ayala et al., 1973), several investigators have theorized that the cortical column may be the basic epileptic unit (Goldensohn et al., 1977; Reichenthal and Hocherman, 1977, 1979; Gabor et al., 1979). Data from rodent neocortical slices, however, have shown that the minimal volume of cortex necessary to elicit an IIS is actually much smaller than a cortical column, and there appears to be no relationship between the two structures (Gutnick et al., 1982; Silva et al., 1991; Chesi and Stone, 1998). These findings have not been confirmed in vivo.

In addition to influencing the initiation of epileptiform events, the columnar organization of the cortex has also been suggested as an important factor in the lateral spread of neocortical epilepsy. Simultaneous electrophysiological recordings from variable distances around the IIS trigger-zone, in both in vitro and in vivo preparations, have demonstrated a non-uniform horizontal propagation velocity (Tharp, 1971; Goldensohn et al., 1977; Chervin et al., 1988; Wong and Prince, 1990; Wadman and Gutnick, 1993). This spatial periodicity is thought to reflect variations in the density in long-range horizontal excitatory connections, which have been shown to link populations of neurons with similar receptive field properties (Gilbert and Wiesel, 1989; Malach et al., 1993; Weliky et al., 1995; Bosking et al., 1997; Dalva et al., 1997).

Our understanding of the cortical column has evolved in recent years with the introduction of optical recording techniques that permit the simultaneous sampling of large areas of cortex (Grinvald et al., 1988; Cohen, 1989; Ebner and Chen, 1995; Bonhoeffer and Grinvald, 1996). These studies have demonstrated that the distribution of orientation preference across visual cortex does not occur in a box-like, discrete fashion, as proposed by earlier models based on electrophysiological data (Hubel and Wiesel, 1963). Rather, the range of orientations is smoothly distributed around a central point in a circular, pinwheel-like pattern (Bonhoeffer and Grinvald, 1991, 1993). Rapid transitions, or fractures, only occur at pinwheel centers, or ‘singularities’ (Blasdel and Salama, 1986; Bonhoeffer and Grinvald, 1991). This finding calls into question the relationship between the epileptic generator and the cortical column. In addition, the existence of additional overlying maps in visual cortex, which have been demonstrated with optical recording techniques, such as spatio-temporal frequency and directionality maps, further confuses the spatial limits of the alleged epileptic generator (Shmuel and Grinvald, 1996; Weliky et al., 1996; Hübener et al., 1997; Shoham et al., 1997).

The purpose of this study was to examine the interaction between an acute pharmacologically-induced IIS focus and the surrounding functional architecture in ferret visual cortex with optical imaging of intrinsic signals. In vivo optical mapping of an IIS focus has been recently demonstrated by Schwartz and Bonhoeffer (2001). In this study, we generated an interictal focus in ferret visual cortex and mapped both the columnar organization of V1 and V2 as well as the epileptic focus in order to visualize directly how the two phenomena interact. Although we were able to preferentially trigger an IIS with oriented visual stimuli, we found little evidence that either the minimal volume of cortical tissue required for an IIS or the lateral margins of horizontal spread has any relationship with the functional architecture. We cannot exclude that these findings may be specific to this experimental model based on the acute injection of a disinhibiting epileptogenic agent.

Materials and Methods

Animal care, surgery and optical imaging were done as previously described (Bonhoeffer and Grinvald, 1996; Chapman and Bonhoeffer, 1998; Schwartz and Bonhoeffer, 2001). All procedures were approved by regional government authorities.

Surgery

Nine adult (6 months to 5 years) ferrets (Marshall Farms, New Rose, NY) were used in this study. Anesthesia was induced with a mixture of ketamine (15–30 mg/kg i.m.) and xylazine (1.5–2.0 mg/kg i.m.), supplemented with atropine (0.15 mg/kg i.m.). Animals received a tracheotomy and were ventilated with 60–70% N2O, 30–40% O2 and 1.4–1.6% halothane (halothane was reduced to 0.8–0.9% for the imaging). End-tidal CO2 was maintained at 3.2–3.8% and animals were hydrated with 2 cc/kg/h dextrose/Ringer’s solution and paralyzed with intravenous gallamine triethiodide (10 mg/kg/h). EKG, EEG and rectal temperature were continually monitored. The scalp was incised, then a craniotomy was performed over the dorsal occipital lobe and the dura was retracted. Withdrawal of cerebrospinal fluid from the cisterna magna was performed when necessary to reduce cerebral pulsations. Animals were fitted with contact lenses to focus the eyes onto the screen, keep the corneas moist and atropine (0.5%) was used to dilate the pupils. Agar (2%) and a glass coverslip were placed over the cortex for imaging.

Visual Stimulus

For visual cortex mapping, animals were presented with a 20-condition random sequence of moving sine-wave gratings projected onto a frosted glass screen. The sequence consisted of four different orientations at two spatial frequencies (0.08 and 0.2 cycles per degree) interleaved with four blanks, iso-luminant with the oriented stimuli. Each non-blank stimulus was presented using computer controlled eye shutters to randomly alternate monocular stimulation, for a total of 20 behavioral conditions. Stimuli appeared on the screen and then began moving in a direction perpendicular to their orientation after 1800 ms. Visual stimuli were generated with a VSG Series Three stimulator (Cambridge Research System, Rochester, UK). During epilepsy mapping, animals were presented with six different visual stimuli, each for 1 min, in a nonrandom order: blank, 0°, 45°, 90°, 135°, blank.

Optical Imaging

The brain was illuminated (707 ± 10 nm) with two fiber-optic light guides connected to a halogen lamp. The signal was recorded with a cooled CCD camera (Theta system — ORA 2001, Germantown, NY) equipped with a tandem lens (Ratzlaff and Grinvald, 1991) focused ∼500 µm beneath the cortical surface. Orientation, spatial frequency and ocular dominance maps were first generated in normal cortex prior to epileptogenesis. Five frames of 600 ms duration were recorded at the start of stimulus movement, exactly 1800 ms after stimulus onset, for each of 20 visual stimuli (conditions). These were then averaged over 24 repetitions. To map epileptic foci, 120 frames of 500 ms duration were recorded for each of six visual stimuli (conditions). Each of these 6 min blocks was separated by a 95 s inter-block interval, during which images were digitized at 12-bit resolution and stored on the hard drive.

Electrophysiology and Iontophoresis

Epidural electrocorticography (ECoG) was monitored with two AgCl electrodes on either side of the craniotomy, ∼5 mm from the epileptic focus. Field potential recording (f.p.) and iontophoresis were performed through two glass micropipettes (o.d. = 1.0 mm, i.d. = 0.58 mm) pulled and broken to a tip diameter of 5 µm and separated by 500 µm. F.p. micropipettes were filled with 3 M NaCl. ECoG and f.p. signals were amplified, band-pass filtered between 1 and 100 Hz, and digitized at 200 Hz. Single-units were recorded during visual stimulation and quantitative orientation and direction tuning curves were calculated after spike-sorting based on spike shapes (Brainware, Oxford, UK). Interictal foci were induced with iontophoresis of 5 mM bicuculline methiodide [BMI (Sigma, St Louis, MO)] diluted in 165 mM NaCl, pH 3.0. The micropipette was advanced 800 µm with an approach angle of 30–40° in order to place the tip in cortical layers II–III of area V1. The currents used were –15 to –20 nA for retention and + 50 to 500 nA for epileptogenesis. Positive currents were maintained until stereotypical biphasic IIS appeared (∼5 min) and then titrated to the minimal dose to maintain a constant amplitude and frequency with no afterdischarges.

Single-unit Analysis

Responses to drifting gratings of 16 different directions from the optically determined ‘focus’ and ‘surround’ regions were bandpass filtered between 0.3 and 3 kHz and digitized using Brainware. Orientation tuning averaged over five trials was quantified by fitting smooth curves to the data with a fast Fourier transform using the zero-, first- and second-order components (Wörgotter and Eysel, 1987). Spontaneous activity was assessed by the response to a blank stimulus. Bursting was defined by the number of inter-spike intervals < 5 ms using the Friedman test (Wyler and Ward, 1980). In addition, inter-spike intervals from 4 to 8 ms were analyzed.

Imaging Analysis

Visual Cortex Maps

Orientation, ocular dominance and spatial frequency maps were computed in a standard fashion (Bonhoeffer and Grinvald, 1996) using commercially available analysis programs (MAPS, Germantown, NY). In brief, differential orientation maps were generated by dividing images obtained while animals viewed one orientation by images obtained during presentation of the perpendicular orientation. Single condition maps are created by dividing images obtained during presentation of an orientation by images obtained during presentation of the blank stimulus. Vectorial summation of the single-condition maps yields an angle map of orientation preference that can then be color-coded. Fracture maps, which demonstrate rapid transitions orientation preference, are produced from the two-dimensional derivative of the angle map. Spatial frequency maps result from dividing the sum of images obtained while the animal views all orientations presented at one spatial frequency by the sum of all orientations presented at a second spatial frequency; ocular dominance maps were computed accordingly. All images were clipped to set the upper and lower 1.5% of pixel values to gray levels of 255 and 0 respectively, and high-pass filtered to cut-off components larger than the field of view of the camera. Orientation tuning was quantified for each pixel within a region of interest by fitting smooth curves to the data with a fast Fourier transform in a fashion similar to that used for the single units (Wörgotter and Eysel, 1987). The area of the column was determined by thresholding the top 15% pixel values in the differential maps. The determination of this value was based on prior reports (Chen et al., 2001) and our own data, which, when varied, did not significantly change the results.

Epilepsy Maps

Blank-divided (BD) epilepsy maps were produced by dividing each individual frame acquired during interictal spiking from an average of 30 control images obtained with a negative holding current when no interictal spiking occurred. Spike-triggered (ST) epilepsy maps were obtained by dividing a single frame following the IIS by a single frame preceding the IIS (Schwartz and Bonhoeffer, 2001). The optical area of the IIS was derived from the BD maps by thresholding to a pixel value 1 SD above the pixel values from the same location of the focus during control conditions. This threshold was chosen since it most reliably reproduced the appearance of the raw data. This area was manually outlined in NIH image (NIH) and calculated from the known area of the field of view of the camera. Control images were generated by triggering divisions from frames in which spikes did not occur. Both BD maps and ST maps could be visualized as single frame divisions or averages of multiple frames to increase the signal to noise ratio. BD maps were spatially-filtered with a high-pass to cut off components larger than the field of view of the camera. ST maps were unfiltered. As a control, maps were generated in a similar fashion with a large negative iontophoretic current.

Correlating the Epilepsy and Visual Cortex Maps

Contour plots of a region of interest surrounding the focus were superimposed over the angle, spatial frequency and ocular dominance maps using IDL (Research Systems, Inc.). Four evenly spaced contours were chosen to represent the range of the optical amplitude of the focus starting at a level equal to one standard deviation above the mean pixel value in the image. A warping algorithm, based on alignment of the blood vessel pattern, was used in cases in which the camera was moved between mapping trials. Two-dimensional cross correlations were performed between a region of interest (ROI) surrounding the IIS focus and the same region in the functional map with bin size equal to the size of a pixel. To obtain statistical significance, the ROI containing the focus was randomly shifted to 1000 locations over the functional maps. The correlation coefficient (r) of the actual data was then compared with the distribution of randomly generated r-values. Since multiple statistical tests were performed, which increase the risk of false positive finding, we chose P < 0.01 as the cut-off for significance. For these statistical comparisons, maps were smoothed with a 120 µm low-pass filter to eliminate both false negative correlations from single pixel deviations and false positive correlations from noise. Images were not thresholded for this comparison to avoid the loss of any subtle structure to the signal, which might be of significance.

Results

Functional Architecture of Ferret Visual Cortex

Orientation, ocular dominance and spatial frequency maps were produced from ferret visual cortex (n = 6) using optical recording of intrinsic signals (Bonhoeffer and Grinvald, 1996). Orientation preference in the ferret is organized in a similar manner as in the cat, tree-shrew or primate (Fig. 1D) (Blasdel and Salama, 1986; Grinvald et al., 1986; Bonhoeffer and Grinvald, 1991; Chapman et al., 1996; Fitzpatrick, 1996; Weliky et al., 1996; Rao et al., 1997). Vertical columns consisting of populations of neurons preferring the same orientation are distributed in a smooth circular pattern into ‘pinwheel’ shapes surrounding a central point (Bonhoeffer and Grinvald, 1991). These ‘singularites’, or ‘fractures’, are characterized by rapid changes in orientation preference between neighboring neurons (Maldonado et al., 1997). By calculating the two-dimensional derivative of the angle map, we generated fracture maps, which revealed that, similar to the findings of Rao et al. (1997, pinwheel centers had a density of 4.8 ± 0.8/mm2 and more discontinuities than in the cat (Fig. 1E). By thresholding the images and quantifying the pixels preferentially responding to each orientation, we determined that the area of cortex devoted to horizontal and vertical orientations was on average 26 ± 14% larger than the area devoted to oblique orientations (Fig. 1B,C). This anisotropy was found in all animals. A similar, although less dramatic, contrast was reported in the visual cortex of the developing and adult ferret in earlier reports (Chapman and Bonhoeffer, 1998; Coppola et al., 1998). The average width of the orientation columns derived from differential maps was 340 ± 60 µm (Fig. 1B,C), which is equivalent to an average area of 0.12 mm2. However, columns were variable in length, some being circular and others oblong.

Ocular dominance maps have a bizarre, irregular structure in the ferret with a large degree of inter-animal variation (Law et al., 1988; Redies et al., 1990; Issa et al., 1999; White et al., 1999). Although three separate regions with differing ocular dominance patterns have been identified, our maps were limited to the most rostral map on the surface of the occipital cortex which is most easily imaged in the in vivo anesthetized preparation (Law et al., 1988; White et al., 1999). In this region, a caudal contralateral eye band sits adjacent to a more rostral ipsilateral eye band (Law et al., 1988; White et al., 1999). Single-unit recordings have shown that these ocular dominance areas are almost completely monocular and their interface represents the V1/V2 border (White et al., 1999) [but see Issa et al. (1999) for contrasting data on binocularity of neurons in V1]. As in previous reports, we found a complex inter-digitation of V1 into V2, consisting of large irregularly shaped blobs which fragments V2 into isolated cortical territories (Fig. 1F) (Issa et al., 1999; White et al., 1999). Although widely variable, the average area of the OD columns was 0.94 ± 0.78 mm2.

Optical imaging and 2-DG data have recently demonstrated that spatial frequency is also organized in a columnar fashion in visual cortex of both the cat and primate (Tootell et al., 1981; Bonhoeffer et al., 1995; Hübener et al., 1997; Shoham et al., 1997; Issa et al., 2000). Spatial frequency columns have not been examined in ferret visual cortex. Although a wide range of spatial frequencies have been shown in the cat to be represented in a continuous fashion (Issa et al., 2000), for the purposes of these epilepsy experiments we only mapped high and low spatial frequency. We found a clear preference for high spatial frequency stimuli caudally and low spatial frequency rostrally in the region of occipital cortex available for optical imaging (Fig. 1G). Within these larger areas, smaller regions of spatial frequency preference were identified that had a mean area of 0.61 ± 0.57 mm2. It is recognized in the cat that neurons in area V1 are tuned preferentially for high spatial frequency while those in area V2 are tuned for low spatial frequency (Movshon et al., 1978; Bonhoeffer et al., 1995; Issa et al., 2000). Although the border between regions with high and low spatial frequency preference in the ferret was almost parallel to the V1/V2 border as defined by the ocular dominance bands, they were clearly separated by as much as 1 mm.

Interictal Spikes are Triggered by Specific Orientations and Spatial Frequencies

After visualizing the functional architecture, an acute IIS focus was created using iontophoresis of BMI into cortical layers II–III. Within 3–5 min of starting a +500 nA ejection current, IIS could be recorded from the local f.p. As demonstrated previously (Schwartz and Bonhoeffer, 2001), these events began as small negative deflections (0.2–0.4 mV) in the local field potential (f.p.) and rapidly developed into well-formed biphasic spikes (2–6 mV), occurring at regular intervals (0.26–0.76 Hz) and lasting for several hours (Fig. 2A,B). Optical recording of intrinsic signals was used to generate spike-triggered epilepsy maps (Schwartz and Bonhoeffer, 2001). These maps provide an image of the location and topography of the population of neurons participating in each IIS with a spatial resolution of <100 µm. During the early small amplitude events, the optical epilepsy maps demonstrated a focal change in reflectance in light that increased in size in parallel with the amplitude of the IIS. After several minutes, well-developed epileptic foci were generally circular, centered on the location of the micropipette tip with an area of 2.84 ± 1.59 mm2 and a sharp border (Fig. 2C). Control injections using a negative current did not produce any changes in reflectance. In order to evaluate whether visual stimulation with a particular orientation was sufficient to trigger an IIS, blank and grating stimuli were shown in sequence and the number of IIS occurring in the first 500 ms after initiating stimulus movement were quantified. As is demonstrated in Figure 3A, the probability of an IIS was greater during a window from 40 to 300 ms following presentation of oriented stimuli than in a similar window after presentation of the blank stimuli. This was seen in all cases (n = 6) (Fig. 3A). In many cases, one or two particular orientations were more likely to trigger an IIS (Fig. 3B). In one animal, presentation of 0°-oriented stimuli consistently precipitated an afterdischarge, a small, self-limited, seizure-like event, which was documented in both the electrical and optical recordings (Fig. 3C). Afterdischarges were not seen following presentation of the other orientations.

The preferred orientation for triggering the IIS was determined by the quantity of IISs that appeared within a 500 ms window after the onset of presentation of each orientation. The most common orientations that triggered an IIS were 0° and 135° (Table 1). In order to determine the explanation for this phenomenon, we hypothesized that the preferred orientation for triggering the IIS would correlate with the columnar location of the iontophoresis pipette and the epicenter of the IIS optical map. By superimposing the epilepsy maps derived from the earliest small amplitude spikes over the angle maps of the orientation columns derived from the same area of cortex at an earlier time, prior to iontophoresis of BMI (Fig. 4), we could show that the location of the epileptic focus was strongly correlated with the orientation of the visual stimulus that preferentially triggered the IIS (Table 1). The columnar location was derived from the number of pixels in the angle map preferring each orientation underneath the area of the superimposed focus.

Another way to determine the preferred orientation for triggering an IIS is to examine the angle map, resulting from visual stimulation with an IIS focus in visual cortex. Angle maps are created by color-coding the vectorial summation of responses to each orientation. The length of each vector is the magnitude of the response to that orientation divided by a genuine blank stimulus (single-condition map). If one orientation were more likely to elicit an IIS than the blank, the epileptic focus would appear in the single-condition map, otherwise it would be divided out. Thus, in the resulting angle map, the dominant color in the focus reveals the orientation most likely to trigger the IIS. Figure 5A (middle) reveals the resulting angle maps from an IIS focus triggered preferentially by 0° > 45° stimuli.

We also found that spatial frequency, in addition to orientation, was a powerful trigger for epileptiform events. In Figure 5B (middle), a sum of the optical responses to high spatial frequency stimuli is divided by the optical response to low spatial frequency stimuli. If IISs were equally likely to be triggered by stimuli of both spatial frequencies, the optical signal from the focus would be divided out of the resulting image. In the example shown in Figure 5B, IISs are clearly more likely to occur in response to low spatial frequency stimuli since the optical signal from the focus appears to have a uniform shading in the divided image. Similar results were found in all six animals.

It is not clear from examining the angle maps in Figure 5A, however, if the orientation preference of the neurons within the focus has also been shifted, since the signal from the IIS is so large. In order to distinguish between these two possibilities, we performed extracellular single-unit recordings from the same neuron before, during and after induction of an IIS focus. Single-units (n = 7 units in four animals) within the focus demonstrated a significant increase in overall firing frequency and bursting simultaneous with the IIS (P < 0.05). Concomitantly, there was a marked widening of the tuning curve from control values. The orientation selectivity index (Wörgotter and Eysel, 1987) decreased from a mean of 1.15 ± 0.48 before epileptogenesis to 0.85 ± 0.42 during the period of IIS, then returned to 0.95 ± 0.64 after reversal of the iontophoresis. There was also a mean shift in orientation preference by 23.4 ± 35° (Fig. 6A,B). Such a shift in orientation preference has also been reported following iontophoresis of sub-epileptic doses of BMI (Toth et al., 1997). Hence, although the number of single units examined is limited, the data support the idea that the optical signal recorded from the focus represents both the preferred orientation for eliciting IIS and a shift in orientation preference of the involved neurons.

The Functional Architecture in the ‘Surround’ is not Affected by the Focus

Epilepsy maps generated with optical imaging of intrinsic signals demonstrate a sharp border between the focus and surrounding cortex (Schwartz and Bonhoeffer, 2001). Neurons in the cortex surrounding an interictal focus are known to exhibit inhibitory post-synaptic potentials (IPSPs) concurrently with each IIS, presumably triggered by recurrent long-range inhibitory connections from within the focus (Prince and Wilder, 1967; Dichter and Spencer, 1969a,b). In order to assess how this decrease in neuronal activity in the ‘inhibitory surround’ affects the functional architecture, we imaged orientation preference in the cortex surrounding an IIS focus and compared it with the functional architecture maps obtained before and after epileptogenesis. It was not clear if the alterations in cerebral hemodynamics induced by the metabolic demands of the IIS focus would affect the optical signals from adjacent cortex.

As can be seen in Figure 5, the layout of the orientation columns outside the area of the epileptic focus is not significantly affected by the adjacent IIS. Extracellular single-unit recordings (n = 7 units in four animals) from neurons in the’ surround’ confirm the optical data (Fig. 6C,D). Although these neurons represent only a small fraction of the entire neuronal pool, unlike those recorded from the focus, they did not manifest bursting simultaneous with the IIS and showed a mean decrease in their firing rate by 36% (P < 0.05). Mean change in orientation preference was –1.01 ± 6.83°.

The Shape of the Interictal Focus Does not Correlate with the Functional Architecture

In order to determine if the spatial distribution of the IIS conforms to the underlying functional architecture, we superimposed the epilepsy maps over the cortical architecture maps. Epilepsy maps generated with spike-triggered image division were used for this analysis since blank-divided (BD) maps have larger vascular artifacts that could induce false positive correlations. Contour plots of the IIS superimposed over the functional maps did not reveal any obvious relationship between the focus and the underlying architecture (Fig. 7). For example, the margins of the IIS did not spread preferentially into any one orientation or avoid crossing either the borders between ocular dominance and spatial frequency columns or the fractures in the orientation maps. Neither was there a region outside the border of the focus, within a particular functional ‘column’, in which the intrinsic signal rose >1 SD from the mean of the entire image. We investigated any subtle relationship between the topography of the focus and the underlying functional architecture using two-dimensional cross correlations. A region of interest surrounding the interictal focus was correlated with the same region in the functional architecture maps and the resulting r-value was compared with the distribution of r-values obtained by randomly shifting the focus to 1000 different locations within the functional map. There were no correlations at P < 0.01 for either the orientation, ocular dominance or spatial frequency maps. To explore the possibility that a rapid change in contour might occur in the region of a fracture, a similar comparison was performed between the two-dimensional derivative map of the focus and the fracture map. No significant correlations were found.

The size and shape of the minimal epileptic aggregate has no relationship to the functional architecture

To determine the minimal epileptic aggregate capable of sustaining an IIS, we used BD maps acquired during the first, low-amplitude spikes, which appear early in the development of the interictal focus. BD maps, in contrast to ST maps, represent the full optical signal since the residual signal from the prior spike is not subtracted out of the final image (Schwartz and Bonhoeffer, 2001). Although BD maps have more artifacts, they are also more sensitive to small signal changes. Since we could not reliably produce optical maps from a single spike at the earliest stages of the evolution of the focus, we averaged the maps from the first 10 spikes, which were of similar amplitude.

The first ten spikes in seven different foci had a mean amplitude and width of 0.43 ± 0.07 mV and 25.9 ± 6.3 ms respectively. Simultaneously acquired BD maps had a mean area of 0.12 ± 0.02 mm2 with a range of 0.08–0.14 mm2 (Fig. 4). Hence, the size of the minimal epileptic aggregate is similar to the size of the iso-orientation domains but much smaller than the ocular dominance and spatial frequency columns. In addition, there is much less variance in the size of the minimal epileptogenic aggregate than in the functional columns, which vary widely from column to column and animal to animal. Since BD maps are more affected by surface vasculature, their contour is not symmetrical. To test whether any anisotropy was caused by correlations with the underlying functional architecture, these first-spike BD maps were cross-correlated with the functional architecture maps and compared with the distribution of r-values obtained by randomly shifting the position of the focus to 1000 different locations. No significant (P < 0.01) correlations were found.

Discussion

The IIS is the electrophysiological hallmark of epileptic cortex, which consists of the synchronous paroxysmal depolarization of a population of pathologically interconnected neurons, typically lasting less than 100 milliseconds (Dichter and Ayala, 1987). These neurons are also capable of producing high frequency oscillations termed ‘fast ripples’ (Ylanen et al., 1995; Bragin et al., 2002a). Surrounding the epileptic focus, neurons exhibit varying degrees of inhibition, felt to be important in controlling the spread of hypersynchronous excitatory activity (Prince and Wilder, 1967; Dichter and Spencer, 1969a,b; Bragin et al., 2002a). The acute application of disinhibiting pharmacological agents such as penicillin (PCN) or bicuculline (BMI), which act via blockade of GABAA receptors, is one of the oldest and most frequently utilized models of the IIS focus (Fisher, 1989). Although this model is not a perfect replication of the chronic epileptic focus, intracellular recordings from neurons within the disinhibited focus reveal paroxysmal depolarizing shifts similar to recordings from neurons in more chronic foci, as well as facilitation of ‘fast ripple’ activity (Matsumoto and Ajimone-Marsan, 1964; Prince, 1968; Bragin et al., 2002a).

How Does the Functional Architecture Influence the Initiation of the Epileptic Event?

By presenting an oriented stimulus that was iso-luminant with the blank stimulus, we were able to preferentially trigger IISs with oriented stumuli. These results provide direct evidence that the neurophysiologic basis of photosensitive and pattern-sensitive epilepsy resides in primary visual cortex. Earlier attempts to trigger IIS foci in visual cortex with orientation-specific stimuli were not successful (Ebersole and Levine, 1975; Scobey and Gabor, 1977). We attribute our success to the fact that the initial presentation of the oriented stimulus did not cause a change in luminance and spikes were triggered with stimulus movement rather than stimulus presentation. Since change in luminance is a known trigger of visually evoked epileptic events (Scobey and Gabor, 1977; Wilkins et al., 1979), it is probable that the orientation-specific response in prior studies was masked by the large luminance change at stimulus onset.

We also show that certain orientations are more likely to trigger epileptiform events than other orientations and we can predict this preference based on the orientation column into which the BMI is iontophoresed. Hence, the initiation of epileptiform events is directly related to the underlying columnar architecture and can be explained based solely on focal alterations in neocortical physiology.

The ‘Minimal Epileptogenic Unit’

The minimal number of neurons required to support paroxysmal population activity has been investigated both in vivo and in vitro. In vivo cortical transections can isolate small areas of cortex to define the minimal amount of tissue required to support an IIS, generally reported between 0.5 and 0.8 mm2 (Morrell and Hambery, 1969; Reichenthal and Hocherman, 1977; Lueders et al., 1980; Bashir and Holmes, 1993). In vitro techniques, on the other hand, have shown that subdivided neocortical slices consisting of only one or two lamina are capable of sustaining paroxysmal events (Gutnick et al., 1982; Silva et al., 1991; Chesi and Stone, 1998).

The area of cortex from which intrinsic signals were recorded during the earliest IISs in our study was on the order of 0.1 mm2. One reason the size of the minimal epileptogenic aggregate we report is smaller than reported in previous in vivo studies involving cortical transections is that we are recording the amount of tissue participating in each event, rather than the minimal amount required to sustain the event. Surgical transections undoubtedly damage a considerable number of neurons adjacent to the incision, which decreases the number of functional neurons in the remaining slab of tissue. Our technique probably overestimates the size of the epileptogenic aggregate for two additional reasons. First, we summed the maps from the first 10 spikes in an evolving focus to increase our signal to noise and the tenth spike was larger than the first. Also, the intrinsic signal represents subthreshold activity in addition to action potentials and thus includes spatial information about EPSPs in surrounding neurons outside of the bursting focus (Das and Gilbert, 1995). Our findings are further supported by optical recordings of spontaneous events from slices stained with voltage-sensitive dyes (VSDs), which report that events arise from a population of neurons with a volume of <0.04 mm3 or an area of 0.09 mm2 (the spatial resolution of the technique) (Tsau et al., 1998, 1999). In addition, Chesi and Stone (1998) recently compared the minimal epileptogenic unit in vitro to the rodent barrel fields and found that the epileptic generator was clearly smaller than the volume of an individual barrel and that there was no apparent relationship between the two structures.

Both primary and secondary visual cortices are organized into multiple overlying maps, including direction, spatial frequency and ocular dominance in addition to orientation, which are distributed relatively smoothly across the cortical surface (Bonhoeffer and Grinvald, 1991, 1993; Shmuel and Grinvald, 1996; Weliky et al., 1996; Hübener et al., 1997; Shoham et al., 1997; Issa et al., 2000). The size of these columns is extremely variable, even within a single animal. Hence, it is not clear how there could be a single epileptogenic unit that could conform to, or be determined by, such a structurally diverse and complex cortical organization.

How Does the Functional Architecture Influence the Propagation of the Epileptic Event?

In vivo and in vitro field recordings from disinhibited epileptic foci have reported that horizontal propagation occurs in ‘fits and starts’ over preferred pathways across the cortex, possibly related to columnar architecture (Tharp, 1971; Petsche et al., 1974; Chervin et al., 1988; Wong and Prince, 1990; Wadman and Gutnick, 1993). The rate of horizontal propagation in these studies (0.04–0.08 m/s) implicates synaptic transmission mediated by axon collaterals of excitatory pyramidal cells (Chervin et al., 1988; Wong and Prince, 1990; Wadman and Gutnick, 1993;). Both excitatory and inhibitory short-range horizontal excitatory connections (<500 µm) are spatially symmetric, targeting neurons of all orientations in visual cortex, whereas long-range excitatory connections (>1 mm) are patchy and anisoptropic, connecting domains with similar receptive fields (Weliky et al., 1995; Bosking et al., 1997; Dalva et al., 1997; Toth et al., 1997). Hence, the reports of inhomogeneous horizontal epileptic spread implicate long-range rather than short-range connections.

In vitro optical recordings of epileptic spread with voltage sensitive dyes (VSD), however, have not shown a similar non-uniform horizontal spread but rather smooth propagation (Sutor et al., 1994; Tanifuji et al., 1994; Albowitz and Kuhnt, 1995), perhaps caused by higher doses of pharmacological disinhibition used in the VSD experiments (Telfian and Connors, 1998). Experiments in ferret neocortical slices have shown that high doses of BMI can alter horizontal propagation velocities and eliminate the influence of long-range horizontal projections (Nelson and Katz, 1995). Hence, a model of epilepsy based on disinhibition may underestimate the role of long-range projections.

Our in vivo optical recordings were made with the minimal dose of BMI that could sustain constant amplitude IISs and we found a homogeneous, symmetrical spike topography that did not appear to correlate with the underlying columnar architecture. Our negative results may be attributable to the temporal resolution of the intrinsic signal which is on the order of several hundred milliseconds (Frostig et al., 1990; Bonhoeffer and Grinvald, 1996). In support of our data, London et al. (1989) optically recorded BMI-induced IIS in vivo using VSDs and a photodetector with a temporal resolution of <5 ms and reported no spatial inhomogeneities in the lateral spread of the optical signal.

How Does the IIS Focus Influence the Functional Architecture?

Although a small percentage of patients with chronic epilepsy show an overall decline in cognitive abilities, it is thought to result more from long-term anticonvulsant therapy or an underlying structural abnormality rather than from the epileptic events themselves (Cassidy and Corbett, 1997). Nevertheless, there is a convincing evidence that the constant barrage of electrical events can induce changes in the nervous system that may underlie some of the chronic cognitive/behavioral decline found in epileptic patients (Engel et al., 1991; Lothman, 1997).

Our results indicate that focal epileptogenic disinhibition not only induced a large decrease in orientation selectivity but also a small, but significant shift in the orientation preference of neurons in the focus, compared with neurons recorded from the inhibitory ‘surround’. Studies of orientation preference following focal application of subepileptic disinhibiting as well as inhibiting pharmacological agents also show significant changes in orientation preference several hundred microns away from the site of iontophoresis (Crook and Eysel, 1992; Sillito et al., 1980; Toth et al., 1997). Toth et al. (1997) demonstrated that these shifts occur in the direction of the orientation preference of the cortical column into which the disinhibiting agent was applied. Hence, it is likely that the changes in orientation preference found in our study may be more attributable to pharmacological disinhibition than the epileptic events.

Optical and electrophysiological data from the inhibitory ‘surround’, however, is a more reliable reflection of the influence of the IIS on the surrounding architecture since the recordings are performed 4–6 mm from the application site, which is beyond most estimates of diffusion distance following iontophoretic application (Fox et al., 1989). We found that the architecture of the iso-orientation domains did not change in the ‘surround’ despite the overall decrease in neuronal activity and the proximity to the epileptic focus. However, an acute model of epileptogenesis may not be sufficient to induce alterations in synaptic plasticity found in chronic epilepsy.

Our ability to optically map functional architecture in close proximity to the epileptic focus parallels the fMRI literature. Cerebral hemodynamics appear tightly regulated around pathological lesions such as epileptic foci and vascular malformations which might interfere with blood flow and subsequently influence the optical signals (Morris et al., 1994; Latchaw et al., 1995; Maldjian et al., 1996; Schwartz et al., 1998). The ability to optically map functional architecture adjacent to pathological lesions is critical for intraoperative applications of intrinsic signal imaging to guide neurosurgical resections (Schwartz and Bonhoeffer, 2001).

Special thanks to Tobias Bonhoeffer, in whose laboratory these experiments were performed. Also thanks to Sven Schütt, who contributed substantially to the data analysis, Frank Sengpiel and Mark Hübener, for expert advice throughout the study and finally, Volker Staiger, Iris Kehrer and Frank Brinkmann, for providing outstanding technical assistance. This work was supported by the Max-Planck Gesellschaft, as well as by grants to T.H.S. from the Epilepsy Foundation of America, the van Wagenan Fellowship of the American Association of Neurological Surgeons and the Alexander von Humboldt Stiftung.

Figure 1. Optical imaging of intrinsic signals reveals the functional architecture in ferret visual cortex. (A) Blood vessel pattern of the surface of the visual cortex. (B) Differential map produced by dividing images obtained with 0° and 90° stimulation. (C) Differential map produced by dividing images obtained with 45° and 135° stimulation. The size of the 45° and 135° iso-orientation domains is smaller than that of the 0° and 90° domains. (D) The angle map is generated by color-coded vectorial summation of each single condition map on a pixel-by-pixel basis. (E) Fracture map generated from the two-dimensional derivative of the angle map demonstrates rapid changes in orientation preference at pinwheel centers. (F) Ocular dominance map in the ferret has a seemingly disorganized structure and column diameters vary widely both between and within animals. The V1/V2 border correlates with the margin between the caudal contralateral and rostral ipsilateral eye bands. (G) Spatial frequency maps demonstrate high spatial frequency preference caudally and low-spatial frequency preference rostrally. The spatial frequency border is roughly parallel to the ocular dominance border but shifted caudally. R, rostral; C, caudal; L, lateral; M, medial. Scale bar = 1 mm.

Figure 1. Optical imaging of intrinsic signals reveals the functional architecture in ferret visual cortex. (A) Blood vessel pattern of the surface of the visual cortex. (B) Differential map produced by dividing images obtained with 0° and 90° stimulation. (C) Differential map produced by dividing images obtained with 45° and 135° stimulation. The size of the 45° and 135° iso-orientation domains is smaller than that of the 0° and 90° domains. (D) The angle map is generated by color-coded vectorial summation of each single condition map on a pixel-by-pixel basis. (E) Fracture map generated from the two-dimensional derivative of the angle map demonstrates rapid changes in orientation preference at pinwheel centers. (F) Ocular dominance map in the ferret has a seemingly disorganized structure and column diameters vary widely both between and within animals. The V1/V2 border correlates with the margin between the caudal contralateral and rostral ipsilateral eye bands. (G) Spatial frequency maps demonstrate high spatial frequency preference caudally and low-spatial frequency preference rostrally. The spatial frequency border is roughly parallel to the ocular dominance border but shifted caudally. R, rostral; C, caudal; L, lateral; M, medial. Scale bar = 1 mm.

Figure 2. Optical epilepsy maps reveal the area of cortex involved in each IIS. (A) Cortical surface map with the location of the BMI iontophoresis and f.p. recording electrode. (B) Periodic biphasic interictal spike recorded from the field potential electrode reach a stable amplitude within a few minutes after starting iontophoresis. (C) Epilepsy maps from four different animals generated with spike-triggered image division. The first map in the upper left-hand corner corresponds with the cortex and spikes presented in (A) and (B). Each map is the an average of the spike-triggered image division generated from all spikes occurring in a 6 min period, which varied from 70 to 160 spikes, depending on the frequency. Scale bar = 1 mm.

Figure 2. Optical epilepsy maps reveal the area of cortex involved in each IIS. (A) Cortical surface map with the location of the BMI iontophoresis and f.p. recording electrode. (B) Periodic biphasic interictal spike recorded from the field potential electrode reach a stable amplitude within a few minutes after starting iontophoresis. (C) Epilepsy maps from four different animals generated with spike-triggered image division. The first map in the upper left-hand corner corresponds with the cortex and spikes presented in (A) and (B). Each map is the an average of the spike-triggered image division generated from all spikes occurring in a 6 min period, which varied from 70 to 160 spikes, depending on the frequency. Scale bar = 1 mm.

Figure 3. Preferential triggering of IIS and afterdischarges by oriented stimuli. (A) The timing of each IIS following initiation of movement of the grating stimuli pooled over all animals. Each diamond represents the occurrence of an IIS in response to the stimuli displayed on the X-axis. Time 0 corresponds to the onset of the movement of the orientated stimulus and the corresponding time point following the blank stimulus. The Y-axis demonstrates the latency of the spike. Grating stimuli are more likely to elicit an IIS in a window from 40 to 300 ms following stimulus movement than are blank stimuli. (B) Examples of the number of spikes that occur in a window from 40 to 300 ms following stimulus movement in four different animals averaged over all trials. In these examples, the orientations most likely to trigger an IIS were 135° > 0° > 90° > 45° (upper left), 135° = 0° > 90° > 45° (upper right), 0° > 45° > 135° (lower left) and 0° (lower right). Error bar = SEM. Stimuli which always triggered a spike have no error bars. (C) Optical (top trace) and f.p. (bottom trace) recordings from one animal show that 0° stimuli trigger an afterdischarge in two separate trials. Timescale of f.p. recording is expanded to facilitate visual inspection.

Figure 3. Preferential triggering of IIS and afterdischarges by oriented stimuli. (A) The timing of each IIS following initiation of movement of the grating stimuli pooled over all animals. Each diamond represents the occurrence of an IIS in response to the stimuli displayed on the X-axis. Time 0 corresponds to the onset of the movement of the orientated stimulus and the corresponding time point following the blank stimulus. The Y-axis demonstrates the latency of the spike. Grating stimuli are more likely to elicit an IIS in a window from 40 to 300 ms following stimulus movement than are blank stimuli. (B) Examples of the number of spikes that occur in a window from 40 to 300 ms following stimulus movement in four different animals averaged over all trials. In these examples, the orientations most likely to trigger an IIS were 135° > 0° > 90° > 45° (upper left), 135° = 0° > 90° > 45° (upper right), 0° > 45° > 135° (lower left) and 0° (lower right). Error bar = SEM. Stimuli which always triggered a spike have no error bars. (C) Optical (top trace) and f.p. (bottom trace) recordings from one animal show that 0° stimuli trigger an afterdischarge in two separate trials. Timescale of f.p. recording is expanded to facilitate visual inspection.

Figure 4. Optical maps of the earliest, small amplitude IIS, superimposed over the orientation maps. (A) Example of the f.p. recording from the first low-amplitude spikes at the beginning of the development of the IIS focus. (B) Blank-divided maps obtained by thresholding to the pixel value 1 SD above the mean prior to the appearance of the first spikes (seen in A) superimposed over the angle maps from that same area of cortex. In these examples, epileptic foci are located over 0° > 45° (left) and 135° > 0° (right). The area of the focus is indicated below the map. The irregular contour of the spike foci is likely caused by vascular artifacts since the blank divided maps highlight the vasodilation that accompanies the IIS focus.

Figure 4. Optical maps of the earliest, small amplitude IIS, superimposed over the orientation maps. (A) Example of the f.p. recording from the first low-amplitude spikes at the beginning of the development of the IIS focus. (B) Blank-divided maps obtained by thresholding to the pixel value 1 SD above the mean prior to the appearance of the first spikes (seen in A) superimposed over the angle maps from that same area of cortex. In these examples, epileptic foci are located over 0° > 45° (left) and 135° > 0° (right). The area of the focus is indicated below the map. The irregular contour of the spike foci is likely caused by vascular artifacts since the blank divided maps highlight the vasodilation that accompanies the IIS focus.

Figure 5. Orientation- and spatial-frequency triggering of IISs can be seen in the angle and differential maps. Optical recording from the surrounding cortex reveals preserved cortical architecture and intact cerebral autoregulation. (A) Angle map generated in the presence of an IIS focus. The intrinsic signal within the focus is distorted by the occurrence of the spikes. The intrinsic signal from the surrounding cortex is unaffected. The dominant color in the focus indicates that 0° stimuli were more likely to trigger spikes than other orientations. Sample of the f.p. recording simultaneous with the imaging. Scale bar = 1 mm. (B) Spatial frequency maps from the same animal demonstrate that low-spatial frequency stimuli are more likely to trigger IIS than high-spatial frequency stimuli.

Figure 5. Orientation- and spatial-frequency triggering of IISs can be seen in the angle and differential maps. Optical recording from the surrounding cortex reveals preserved cortical architecture and intact cerebral autoregulation. (A) Angle map generated in the presence of an IIS focus. The intrinsic signal within the focus is distorted by the occurrence of the spikes. The intrinsic signal from the surrounding cortex is unaffected. The dominant color in the focus indicates that 0° stimuli were more likely to trigger spikes than other orientations. Sample of the f.p. recording simultaneous with the imaging. Scale bar = 1 mm. (B) Spatial frequency maps from the same animal demonstrate that low-spatial frequency stimuli are more likely to trigger IIS than high-spatial frequency stimuli.

Figure 6. Radial plots of orientation preference from extracellular single unit recordings from the focus and surround. (A) Two examples of units in the focus from two separate animals demonstrate a shift in orientation preference, a decrease in selectivity and an increase in spike frequency in the presence of the IISs. (B) Two units from the surround recorded from two separate animals show minimal change in orientation preference and selectivity although there is a clear decrease in firing frequency. This electrophysiological data confirm the optical data in Figure 5 and indicate that the orientation preference of the neurons in the focus does shift.

Figure 6. Radial plots of orientation preference from extracellular single unit recordings from the focus and surround. (A) Two examples of units in the focus from two separate animals demonstrate a shift in orientation preference, a decrease in selectivity and an increase in spike frequency in the presence of the IISs. (B) Two units from the surround recorded from two separate animals show minimal change in orientation preference and selectivity although there is a clear decrease in firing frequency. This electrophysiological data confirm the optical data in Figure 5 and indicate that the orientation preference of the neurons in the focus does shift.

Figure 7. Lack of correlation between the IIS focus and the functional architecture. (A) Epilepsy map generated with spike-triggered image division. Contour plot of the focus superimposed over the (B) 0/90° differential map, (C) 45/135° differential map, (D) spatial frequency map and (E) ocular dominance map. (F) The two-dimensional derivative of the epilepsy map highlights the rapid change in optical signal at the edge of the focus. (G) Superimposition of the contour plot on the two-dimensional epilepsy map over the fracture map indicates no correlation between rapid changes in orientation preference and the edge of the focus. Scale bar = 1 mm.

Figure 7. Lack of correlation between the IIS focus and the functional architecture. (A) Epilepsy map generated with spike-triggered image division. Contour plot of the focus superimposed over the (B) 0/90° differential map, (C) 45/135° differential map, (D) spatial frequency map and (E) ocular dominance map. (F) The two-dimensional derivative of the epilepsy map highlights the rapid change in optical signal at the edge of the focus. (G) Superimposition of the contour plot on the two-dimensional epilepsy map over the fracture map indicates no correlation between rapid changes in orientation preference and the edge of the focus. Scale bar = 1 mm.

Table 1


 Relationship between the location of the imaged focus with respect to the underlying orientation columns and the preferred orientation of the visual stimulus for triggering the interictal spikes (IIS)

 Preferred orientation triggering IIS  Location of imaged focus 
135° = 0° > 90° > 45° 135° > 0°=90° 
0° 0° > 45° 
135° > 0° 135° > 0° 
135° > 0° > 90° > 45° 135° > 0° 
0° > 45° > 135° 0° > 45° 
0° = 45° = 135° > 90° 0° = 45° = 90° = 135° 
 Preferred orientation triggering IIS  Location of imaged focus 
135° = 0° > 90° > 45° 135° > 0°=90° 
0° 0° > 45° 
135° > 0° 135° > 0° 
135° > 0° > 90° > 45° 135° > 0° 
0° > 45° > 135° 0° > 45° 
0° = 45° = 135° > 90° 0° = 45° = 90° = 135° 

References

Albowitz B, Kuhnt U (
1995
) Epileptiform activity in the guinea-pig neocortical slice spreads preferentially along supragranular layers  — recordings with voltage-sensitive dyes.
Eur J Neurosci
 
7
:
1273–
1284.
Ayala GF, Dichter M, Gumnit RJ, Matsumoto H, Spencer WA (
1973
) Genesis of epileptic interictal spikes. New knowledge of cortical feedback systems suggests a neurophysiological explanation of brief paroxysms.
Brain Res
 
52
:
1–
17.
Bashir ZI, Holmes O (
1993
) Phases in the development of a penicillin epileptiform focus in rat neocortex.
Exp Brain Res
 
96
:
319–
327.
Binnie CD, Wilkins AJ (
1998
) Visually induced seizures not caused by flicker (intermittent light stimulation). In: Advances in neurology (Zifkin BG, Andermann F, Beaumanoir A, Rowan AJ, eds), pp. 
123–
138. Philadelphia, PA: Lippincott-Raven.
Blasdel GG, Salama G (
1986
) Voltage-sensitive dyes reveal a modular organization in monkey striate cortex.
Nature
 
321
:
579–
585.
Bonhoeffer T, Grinvald A (
1991
) Iso-orientation domains in cat visual cortex are arranged in pinwheel-like patterns.
Nature
 
353
:
429–
431.
Bonhoeffer T, Grinvald A (
1993
) The layout of iso-orientation domains in area 18 of cat visual cortex: optical imaging reveals a pinwheel-like organization.
J Neurosci
 
13
:
4157–
4180.
Bonhoeffer T, Grinvald A (
1996
) Optical imaging based on intrinsic signals. The methodology. In: Brain mapping. The methods (Toga AW, Mazziota JC, eds), pp.
55–
99. San Diego, CA: Academic Press.
Bonhoeffer T, Kim D-S, Maloniek, Shoham D, Grinvald A (
1995
) Optical imaging of the layout of functional domains in area 17 and across the area 17/18 border in cat visual cortex.
Eur J Neurosci
 
7
:
1973–
1988.
Bosking WH, Zhang Y, Schofield B, Fitzpatrick D (
1997
) Orientation selectivity and the arrangement of horizontal connections in tree shrew striate cortex.
J Neurosci
 
17
:
2112–
2127.
Bragin A, Mody I, Wilson CL, Engel JJ (
2002
) Local generation of fast ripples in epileptic brain.
J Neurosci
 
22
:
2012–
2021.
Cassidy G, Corbett J (
1997
) Learning disorders. In: Epilepsy. A comprehensive textbook (Engel JJ, Pedley TA, eds), pp.
2053–
2063. Philadelphia, PA: Lippincott-Raven.
Chapman B, Bonhoeffer T (
1998
) Overrepresentation of horizontal and vertical orientation preferences in developing ferret area 17.
Proc Natl Acad Sci USA
 
95
:
2609–
2614.
Chapman B, Stryker MP, Bonhoeffer T (
1996
) Development of orientation preference maps in ferret primary visual cortex.
J Neuroscience
 
16
:
6443–
6453.
Chartrian GE, Lettich E, Miller LH, Green JR (
1970
) Pattern-sensitive epilepsy. 1. An electrographic study of its mechanisms.
Epilepsia
 
11
:
125–
149.
Chen LM, Friedman RM, Ramsden BM, LaMotte RH, Roe AW (
2001
) Fine-scale oraganization of S1 (Area 3b) in the squirrel monkey revealed with intrinsic optical imaging.
J Neurophysiol
 
86
:
3011–
3029.
Chervin RD, Pierce PA, Connors BW (
1988
) Periodicity and directionality in the propagation of epileptifom discharges across neocortex.
J Neurophysiol
 
60
:
1695–
1713.
Chesi AJR, Stone TW (
1998
) Epileptiform activity in supragranular and infragranular blocks of mouse neocortex.
Epilepsy Res
 
31
:
29–
38.
Cohen L (
1989
) Optical approaches to neuronal function. In: Annual review of physiology (Hoffman JF, De Weer P, eds), pp.
487–
582. Palo Alto, CA: Annual Review Inc.
Coppola DM, White LE, Fitzpatrick D, Purves D (
1998
) Unequal representation of cardinal and oblique contours in ferret visual cortex.
Proc Natl Acad Sci USA
 
95
:
2621–
2623.
Crook JM, Eysel UT (
1992
) GABA-induced inactivation of functionally characterized sites in cat visual cortex (area 18): effects on orientation tuning.
J Neurosci
 
12
:
1816–
1825.
Dalva MB, Weliky M, Katz LC (
1997
) Relationships between local synaptic connections and orientation domains in primary visual cortex.
Neuron
 
19
:
871–
880.
Das A, Gilbert CD (
1995
) Long-range horizontal connections and their role in cortical reorganization revealed by optical recording of cat primary visual cortex.
Nature
 
375
:
780–
784.
Dichter M, Spencer WA (
1969
) Penicillin-induced interictal discharges from cat hippocampus. I. Characteristics and topographical features.
J Neurophysiol
 
32
:
649–
662.
Dichter M, Spencer WA (
1969
) Penicillin-induced interictal discharges from cat hippocampus. II. Mechanisms underlying origin and restriction.
J Neurophysiol
 
32
:
663–
687.
Dichter MA, Ayala GF (
1987
) Cellular mechanisms of epilepsy: a status report.
Science
 
237
:
157–
164.
Dorothée GA, Trenité K-N (
1998
) Reflex seizures induced by intermittent light stimulation. In: Advances in neurology (Zifkin BG, Andermann F, Beaumanoir A, Rowan AJ, eds), pp.
99–
121. Philadelphia, PA: Lippincott-Raven.
Ebersole JS, Chatt AB (
1981
) Toward a unified theory of focal penicillin epileptogenesis: an intracortical evoked potential investigation.
Epilepsia
 
22
:
347–
363.
Ebersole JS, Levine RA (
1975
) Abnormal neuronal responses during evolution of a penicillin epileptic focus in cat visual cortex.
J Neurophysiol
 
38
:
250–
266.
Ebner TJ, Chen G (
1995
) Use of voltage-sensitive dyes and optical recordings in the central nervous system.
Prog Neurobiol
 
46
:
463–
506.
Engel JJ, Bandler R, Griffith NC, Caldecott-Hazard S (
1991
) Neurobiological evidence for epilepsy-induced interictal disturbances. In: Advances in neurology (Smith D, Treiman D, Trimble M, eds), pp. 
97–
111. New York: Raven Press.
Fisher RS (
1989
) Animal models of the epilepsies.
Brain Res Rev
 
14
:
245–
278.
Fitzpatrick D (
1996
) The functional organization of local circuits in visual cortex: insights from the study of tree shrew cortex.
Cereb Cortex
 
6
:
329–
341.
Fox K, Sato H, Daw N (
1989
) The location and function of NMDA receptors in cat and kitten visual cortex.
J Neurosci
 
9
:
2443–
2454.
Frostig RD, Lieke EE, Ts’o DY, Grinvald A (
1990
) Cortical functional architecture and local coupling between neuronal activity and the microcirculation revealed by in vivo high-resolution optical imaging of intrinsic signals.
Proc Natl Acad Sci USA
 
87
:
6082–
6086.
Gabor AJ, Scobey RP (
1975
) Spatial limits of epileptogenic cortex: its relationship to ectopic spike generation.
J Neurophysiol
 
38
:
395–
404.
Gabor AJ, Scobey RP (
1980
) The physiological basis of epileptogenic neural aggregates.
Trends Neurosci
 
00
:
000–
000.
Gabor AJ, Scobey RP, Wehrli CJ (
1979
) Relationship of epileptogenicity to cortical organization.
J Neurophysiol
 
42
:
1609–
1625.
Gilbert CD, Wiesel TN (
1989
) Columnar specificity of intrinsic horizontal connections in cat visual cortex.
J Neurosci
 
3
:
1116–
1133.
Goldensohn ES, Zablow L, Salazar A (
1977
) The penicillin focus. I. Distribution of potential at the cortical surface.
Electroencephalogr Clin Neurophysiol
 
42
:
480–
492.
Grinvald A, Lieke EE, Frostig RD, Gilbert CD, Wiesel TN (
1986
) Functional architecture of cortex revealed by optical imaging of intrinsic signals.
Nature
 
324
:
361–
364.
Grinvald A, Frostig RD, Lieke E, Hildesheim R (
1988
) Optical imaging of neuronal activity.
Physiol Rev
 
68
:
1285–
1365.
Gutnick MJ, Grossman Y, Carlen P (
1982
) Epileptogenesis in subdivided neocortical slices.
Neurosci Lett Supp
 
10
:S226.
Hubel DH, Wiesel TN (
1962
) Receptive fields, binocular interaction and functional architecture in the cat’s visual cortex.
J Physiol
 
160
:
106–
154.
Hubel DH, Wiesel TN (
1963
) Shape and arrangement of columns in cat’s striate cortex.
J Physiol
 
165
:
559–
568.
Hübener M, Shoham D, Grinvald A, Bonhoeffer T (
1997
) Spatial relationships among three columnar systems in cat area 17.
J Neurosci
 
17
:
9270–
9284.
Issa NP, Trachtenberg JT, Chapman B, Zahs KR, Stryker MP (
1999
) The critical period for ocular domiance plasticity in the ferret’s visual cortex.
J Neurosci
 
19
:
6965–
6978.
Issa NP, Trapel C, Stryker MP (
2000
) Spatial frequency maps in cat visual cortex.
J Neurosci
 
20
:
8504–
8514.
Latchaw RE, Xioping H, Ugurbil K et al. (
1995
) Functional magnetic resonance imaging as a management tool for cerebral arteriovenous malformations.
Neurosurgery
 
37
:
619–
626.
Law MI, Zahs KR, Stryker MP (
1988
) Organization of primary visual cortex (area 17) in the ferret.
J Comp Neurol
 
278
:
157–
180.
London JA, Cohen LB, Wu J-Y (
1989
) Optical recordings of the cortical response to whisker stimulation before and after the addition of an epileptic agent.
J Neurosci
 
9
:
2182–
2190.
Lothman EW (
1997
) Biological consequences of repeated seizures. In: Epilepsy. A comprehensive textbook (Engel JJ, Peddley TA, eds), pp. 
481–
497. Philadelphia, PA: Lippincott-Raven.
Lueders H, Bustamente L, Zablow L (
1980
) Quantitative studies of spike foci induced by minimal concentrations of penicillin.
Electroencephalogr Clin Neurophysiol
 
48
:
80–
89.
Malach R, Amir Y, Harel M, Grinvald A (
1993
) Relationship between intrinsic connections and functional architecture revelaed by optical imaging and in vivo targeted biocytin injectios in primate striate cortex.
Proc Natl Acad Sci USA
 
90
:
10469–
10473.
Maldjian J, Atlas SW, Howard RS et al. (
1996
) Functional magnetic resonance imaging of regional brain activity in patients with intracerebral arteriovenous malformations before surgical or endovascular therapy.
J Neurosurg
 
84
:
477–
483.
Maldonado PE, Gödecke I, Gray CM, Bonhoeffer T (
1997
) Orientation selectivity in pinwheel centers in cat striate cortex.
Science
 
276
:
1551–
1555.
Matsumoto H, Ajmone-Marsan C (
1964
) Cortical cellular phenomena in experimental epilepsy: interictal manifestations.
Exp Neurol
 
9
:
286–
304.
Morrell F, Hambery J (
1969
) A new surgical technique for the treatment of focal cortical epilepsy [abstract].
Electroencephalogr Clin Neurophysiol
 
26
:120.
Morris GL, Mueller WM, Yetkin FZ et al. (
1994
) Functional magnetic resonance imaging in partial epilepsy.
Epilepsia
 
35
:
1194–
1198.
Mountcastle VB (
1957
) Modality and topgraphic properties of single neurons of cat’s somatic sensory cortex.
J Neurophysiol
 
20
:
408–
434.
Mountcastle VB (
1982
) An organizing principle of cerebral function: the unit module and the distributed system. In: The mindful brain (Schmitt HO, ed.), pp.
1–
50. Cambridge, MA: MIT Press.
Movshon JA, Thompson ID, Tolhurst DJ (
1978
) Spatial and temporal contrast sensitivity of neurons in areas 17 and 18 of the cat’s visual cortex.
J Physiol
 
283
:
101–
120.
Nelson DA, Katz LC (
1995
) Emergence of functional circuits in ferret visual cortex visualized by optical imaging.
Neuron
 
15
:
23–
34.
Petsche H, Prohaska O, Rappelsburger P, Vollmer R, Kaiser A (
1974
) Cortical seizure patterns in a multidimensional view: the information content of equipotential maps.
Epilepsia
 
15
:
439–
463.
Prince DA (
1968
) The depolarizing shift in ‘epileptic’ neurons.
Exp Neurol
 
21
:
467–
485.
Prince DA, Wilder J (
1967
) Control mechanisms in cortical epileptogenic foci. ‘Surround’ inhibition.
Arch Neurol
 
16
:
194–
202.
Rao SC, Toth LJ, Sur M (
1997
) Optically imaged maps of orientation preference in primary visual cortex of cats and ferrets.
J Comp Neurol
 
387
:
358–
370.
Ratzlaff EH, Grinvald A (
1991
) A tandem-lens epifluourescence macroscope: hundred-fold brightness advantage for wide field imaging.
J Neurosci Methods
 
36
:
127–
137.
Redies C, Diksic M, Riml H (
1990
) Functional organization in the ferret visual cortex: a double-label 2-deoxyglucose study.
J Neurosci
 
10
:
2791–
2803.
Reichenthal E, Hocherman S (
1977
) The critical cortical area for development of penicillin-induced epilepsy.
Electroencephalogr Clin Neurophysiol
 
42
:
248–
251.
Reichenthal E, Hocherman S (
1979
) A critical epileptic area in the cat’s cortex and its relation to the cortical columns.
Electroencephalogr Clin Neurophysiol
 
47
:
147–
132.
Schwartz TH, Bonhoeffer T (
2001
) In vivo optical mapping of epileptic foci and surround inhibition in ferret cerebral cortex.
Nature Med
 
7
:
1063–
1067.
Schwartz TH, Resor SR Jr, De La Paz R, Goodman RR (
1998
) Functional magnetic resonance imaging localization of ictal onset to a dysplastic cleft with simultaneous sensorimotor mapping: intraoperative electrophysiological confirmation and postoperative follow-up: technical note.
Neurosurgery
 
43
:
639–
644.
Scobey RP, Gabor AJ (
1977
) Properties of epileptogenic focus: activation field.
J Neurophysiol
 
40
:
1199–
1213.
Shmuel A, Grinvald A (
1996
) Functional organization for direction af motion and its relationship to orientation maps in cat area 18.
J Neurosci
 
16
:
6945–
6964.
Shoham D, Hübener M, Schulze S, Grinvald A, Bonhoeffer T (
1997
) Spatio-temporal frequency domains and their relation to cytochrome oxidase staining in cat visual cortex.
Nature
 
385
:
529–
533.
Sillito AM, Kemp JA, Milson JA, Berardi N (
1980
) A re-evaluation of the mechanisms underlying simple cell organization selectivity.
Brain Res
 
194
:
517–
520.
Silva LR, Amitai Y, Connors B (
1991
) Intrinsic oscillation of neocortex generated by layer 5 pyramidal neurons.
Science
 
251
:
432–
435.
Sutor B, Hablitz JJ, Rucker F, Bruggencate G (
1994
) Spread of epileptiform activity in the immature rat neocortex studied with voltage-sensitive dyes and laser scanning microscopy.
J Neurophysiol
 
4
:
1756–
1768.
Tanifuji M, Sugiyama T, Murase K (
1994
) Horizontal propagation of excitation in rat visual cortical slices revealed by optical imaging.
Science
 
266
:
1057–
1059.
Telfian AE, Connors BW (
1998
) Layer-specific pathways for the horizontal propagation of epileptiorm discharges in neocortex.
Epilepsia
 
39
:
700–
708.
Tharp BR (
1971
) The penicillin focus: a study of field characteristics using cross-correlation analysis.
Electroencephalogr Clin Neurophysiol
 
31
:
45–
55.
Tootell RBH, Silverman MS, De Valois RL (
1981
) Spatial frequency columns in primary visual cortex.
Science
 
214
:
813–
815.
Toth LJ, Kim D-S, Rao SC, Sur M (
1997
) Interaction of local inputs in visual cortex.
Cereb Cortex
  7.
Tsau Y, Guan L, Wu J-Y (
1998
) Initiation of spontaneous epileptiform activity in the neocortical slice.
J Neurophysiol
 
80
:
978–
982.
Tsau Y, Guan L, Wu J-Y (
1999
) Epileptiform activity can be initiated in various neocortical layers: an optical imaging study.
J Neurophysiol
 
82
:
1965–
1973.
Wadman WJ, Gutnick MJ (
1993
) Non-uniform propagation of epileptiform discharge in brain slices of rat neocortex.
Neuroscience
 
52
:
255–
262.
Weliky M, Kandler K, Fitzpatrick D, Katz LC (
1995
) Patterns of excitation and inhibition evoked by horizontal connections in visual cortex share a common relationship to orientation columns.
Neuron
 
15
:
541–
552.
Weliky M, Bosking WH, Fitzpatrick D (
1996
) A systematic map or direction preference in primary visual cortex.
Nature
 
379
:
725–
728.
White LE, Bosking WH, Williams SM, Fitzpatrick D (
1999
) Maps of central visual space in Ferret V1 and V2 lack matching inputs from the two eyes.
J Neurosci
 
19
:
7089–
7099.
Wilkins AJ, Binnie CD, Darby CE (
1979
) Visually-induced seizures.
Prog Neurobiol
 
15
:
85–
117.
Wong BY, Prince DA (
1990
) The lateral spread of ictal discharges in neocortical brain slices.
Epilepsy Res
 
7
:
29–
39.
Wörgotter F, Eysel UT (
1987
) Quantitative determination of orientational and directional components in the response of visual cortical cells to moving stimuli.
Biol Cybern
 
57
:
349–
355.
Wyler AR, Ward AAJ (
1980
) Epileptic neurons. In: Epilepsy: a window to brain mechanisms (Lockard JSS, Ward AAJ, eds), pp.
51–
68. New York: Raven.
Ylanen A, Bragin A, Nádasdy Z et al. (
1995
) Sharp wave associated high frequency oscillation (200 Hz) in the intact hippocampus: network and intracellular mechanisms.
J Neurosci
 
14
:
30–
46.