[1] 20w The influence of electric fields on the epileptiform bursts induced by high potassium in CA3 region of rat hippocampal slice

[1] 200w The e ffe cts of pulsed direct current (de) electric fields on the frequency of spontaneous bursting in a model epileptic focus were studied. The high potassium hippocampal slice model was used to generate spontaneous burst firing activity similar to interictalspikes in the pyramidal eel/layer of CA3. Electric fields were generated from platinum subdural electrodes placed in the perfusion bath. Three hundred and seventy-eight experimental trials were performed on 7 0 hippocampal slices from 7 0 rats and the effects of field polarity, field strength and duration of stimuli on firing frequency was examined. Hippocampal slices were oriented horizontally with the CA3 layer towards the positive electrode, the average interburst interval did not correlate significantly with polarity of the delivering pulses (one-way ANOVA, p = 0.96). Average interburst interval showed a significant correlation with pulse duration of 200 and 400msec (p = 0.030 and p = 0.004, respectively). As a fun ction of field strength, there were significant average interval changes for fields o f 33, 46, and 73 mV/mm (p = 0.024, p = 0 .007 and p = 0 .007, respectively). In conclusion, CA3 burst firing activity in high potassium concentration can therefore be altered by electric fields.

[1] 232w The fundamental characteristic of epileptiform activity is an in crease in synchrony of large populations of neurons. Numerous studies have demonstrated evidence that both endogenous and externally applied electrical fields can modify the excitability and synchronicity of neu ronal population. In 1962, Creutzfeld 1 successfully altered the spontaneous and evoked activity of neocortical neurons with transcortical extracellular de currents; two years later, Bindman and colleagues 2 were able to increase the firing rate of pyramidal neurons in rat cerebral cortex with brief polarizing currents. Using in vitro hippocampal slice preparations; Jefferys 3 was ab le to modify the synchrony of discharge of dentate granule cells of guinea-pig hippocampus with extracellular de pulses generated from nonpolarizab le electrodes oriented perpendicular to the slice. Richardson 4 demonstrated also that extracellular field potentials can ephaptically discharge CAl neurons and may play a role in recruitment and synchronization of neuronal activity in the hippocampus. Durand 5 successfully showed how focal electrical currents applied to penici llin induced epilepti form activity of rat hippocampa l slices decreased the amplitude of evoked population spikes. Nevertheless, whether electric fields can reliably modify the activity of an autonomously firing neuronal ensemble has to our knowledge never been explored. In order to examine this, we investigate the effect of pulsed de electric fi!=lds on spontaneous burst-firing freq uency in the CA3 pyramidal ce lls from high potass ium hippocampal brain slices.

[1] 12w Female Sprague-Dawley rats weighing 125-140 g were anesthetized with diethyl-ether and decapitated.

[2] 52w Transverse slices 400 ,um thick were prepared from the hippocampus with a tissue chopper, and placed horizontally on a nylon net midway between and 1 mm above the two platinum electrodes in the interface chamber. Following 90 min of incubation in humidified oxygenated (9 5% 0 2, 5% C02) ACSF containing normal

[3] 143w [KCI] (K+ = 3.5 mM, CJ-= 136 mM), the perfusate was switched to ACSF containing high [KCI] (K+ = 8.5 mM, CJ-= 141 mM). After 15-20 min of high [KCI] perfusion, spontaneous burst firing was recorded from the pyramidal cell layer of CA3. Systematically, electric pul ses with different polarity, current intensity and duration were delivered to the platinum electrodes; the interval between stimulation was 30-60 sec. Recordings were digitized across 12 bits at 5kHz with a Digidata 1200 analog to digital converter (Axon Instruments, Foster City, CA, USA), and stored on a personal computer using Axotape 2.0 (Axon Instruments). The intervals between the burst firing were measured over a period of 4.5 sec immediately before and after stimulation and these intervals were summed and averaged. The anatomy of the transverse hippocampal sl ice and arrangement of electrodes are illustrated in Figure 7.

[4] 138w Figure 2 is a graph of electric potential measurements taken at the peak of de pulses as a function of distance from the electrodes at different stimulus currents. The graph abscissa is in mm and zero is the center of the chamber with the ( +) and (-) electrodes at ± 5 mm respectively . There is a linear relationship between electric potential and the distance along the line joining the electrode centers indicating a constant electric field through the slice under study (electric field is the gradient of the electric potential). These results confirm that the slice rests within the center of the 'near field' of the dipole we have constructed, with the expected zero potential midway between the negative and positive electrodes. In the current ranges used 2-8 mA, field strength of 15-73 mV/mm was generated.

[5] 75w Ten hippocampal slices from 10 rats were studied. We examined the effects of the electric field on interburst interval by alternately changing the polarity, stimulus duration and field strength. Three hundred and seventyeight trials w ere analyzed. As illustrated in Figure 7 slices were always oriented w ithin the chamber w ith the CA3 towards the left and dentate gyrus dow n. We indicate the polarity of the left/right electrode pair as +Ior -I+ (left/right).

[6] 197w Figure 3 presents selected examples of the change of interburst interval immediately before and after the stimulus as a function of polarity (Figure 3A), stimulus duration (Figure 38) and field strength (Figure 3C) in three different experiments on the same hippocampal slice. Arrow s indicate the stimulus artifacts associated with the applied de pulses. In Figure 3A, the upper trace shows a decrease in average interval length after changing polarity from positive near CA3 (+1-) to negative (-I + ) while the lower trace shows an increase A 8 c -•-.sma -e-1 ma -..&.-2ma -+--4ma -X-6ma -*-8ma 1 mvl 1 sec Figure 3: Selected examples of burst-firin g frequency changes as fun ction of A: Pola rity: to p, negative polarity (-/+); bottom, positive polarity (+/-). B: Stimulus duration: top, 10 msec; botto m, 400 msec. C: Fie ld strength: to p, 2 ma (15 mBV/mm); botto m, 8 ma (73 mV/mm). Arrows ind icated the onset of stimulus pulse A significant lengthening of the interval between burst-firing was seen w hen the CA3 layer oriented toward the positive electrode and the stimulation duration was set at 200-400 msec, p = 0.03 and p = 0.004 respectively.

[7] 144w in average interval after reversing from negative(-/+ ) to positive(+/-) polarity. Likewise, in Figure 3B the upper trace shows a decrease in average bursting interval after a 10 msec stimulus duration while the lower trace is an example of an increase in average bursting interval after a400 msec pulse. In Figure 3C, the top trace shows the average bursting interval decreases when a field strength of 15 mV/mm was applied to the hippocampal slice, and increases in the bottom trace for a field strength of 73 mV/mm. Polarity for both 3b and 3c is+/ -. Table 1 summarizes the results of the average change in interburst interval length as a function of po larity regardless of field strength and duration. There were more increases than decreases in interval, but this did not correlate with a particular choice of polarity (oneway ANOVA, p = 0.96).

[8] 61w Table 2 summarizes the average change in interburst interval versus stimulus duration. With negative polarity (-/+) near CA3, no correlation is seen between interval and stimulus duration. However, with positive polarity (+/ -) a significant increase in average interval length was seen with stimulus duration of 200 msec and 400 msec (one-way ANOVA, p = 0.030 and p = 0.004, respectively).

[9] 83w Table 3 summarizes the resu lts of the average change in interburst interval in response to varying levels of field strength. No significant correlation is seen for negative polarity (-/ +) near CA3. For positive polarity (+ / -), average interval significantly increased for sti mulus currents of 4, 6, and 8 ma which are equivalent to delivered potential fields of 33, 46, and 73 mV/ mm (oneway ANOVA, p = 0.024, p < 0.001 , and p < 0. 001 , respectively).

[10] 55w Figure 4 is a 3D-scatter graph demonstrat ing the interplay between field strength and duration on average interburst interval. Higher field strength of 33-73 mv/ mm and longer duration of 10-400 mV/mm lengthen the interburst intervals. On the other hand smaller field of 15-33 mV/mm and shorter duration 1 0-50 msec shorten the interburst intervals.

[11] 53w In order to develop regression equation for these findings, a multifactor ANOVA of polarity, sti mulus duration and field strength adjusted for po larity and duration is shown in Table 4. This analysis confirms that polarity show ed no significant effect on average interburst interval. On the other hand, the average interburst (/)

[12] 72w .s:::. interval is signfificantly dependent on both stimulus duration and field strength (p < 0 .0001 and p = 0 .0076, respectively). Using this model, multiple regression yields the following relationship y = -52.68 + 12.68a + Q.41 b where y is the mean change of interburst interval, -52.68 is the constant (p = 0.04t a is the field strength (p = 0.003) and b is the stimulusduration (p < 0.001 ).

[1] 214w Potassium-induced spontaneous population events in rat hippocampal slices were used in this study to investigate the effects of electric fields on the frequency of burst firing in CA3. This high potassium model has been extensively studied and it has been suggested that these bursts of neuronal activity share ph~siological characteristics with human interictal spikes -9 . The hippocampal slices in our experimental setup were oriented horizontally between two electrodes in order to allow the CA3 layer to be consistently oriented with respect to the electric field. This placement arrangement has bee n shown previously as an effective way of modifying the electrical excitability of dentate granule cells 3 . 546 Neurological Research, 1998, Volume 20, September Electric fields generated by de pulses altered the frequency of spontaneous burst firing in CA3. Polarity, stimulus duration and field strength could change the bursting behavior of this system. There was a trend of increasing the interburst interval when the CA3 layer was placed either towards the positive or negative electrode and on the same hippocampal slice the opposite effect, which is decreasing the interburst interval, was seen when these polarities were reversed (Figure 3A). Nevertheless, polarity alone was a rather inconsistent modulator of interburst interval, illustrated in the lack of statistical significance in Tables 1 and 4.

[2] 157w Previous authors have investigated the effect of polarity in the in vitro evoked potentials from hippocampal slices 3 ' 10 and in anesthetized rat and cat cerebral neocortex 1 ' 2 . These studies suggested that both positive polarity and negative polarity applied to a cell body layer could increase the excitability of the underlying cortex. Unfortunately, some of these studies share little similarity to our present experiments. In a more similar study to ours, Jefferys 3 successfully demonstrated that orienting dentate granule cells parall el to an electric field was effective in modifying their electrical excit" ability. In order to settle the issue whether polarity has any affect on the bursting activity, surgically isolating a Downloaded by [Australian Catholic University] at 04:53 08 August 2017 portion of the CA3 layer would increase the uniformity of the orientation of CA3 pyramidal cells within the field, and could serve to increase the consistency of experiments such as these.

[3] 230w The relationship between pulsed electric field duration and changes in mean interburst interval was approximately linear. With a small duration (i.e. 1 0 msec or 50 msec), we were able to either increase or decrease the frequency of burst firing. On the other hand, longer duration consistently lengthened the interval (1 00, 200 and 400 msec). We also used 600, 800 and 1 000 msec in some of our preparations but did not include these results in this report because with long sti mulation duration, the electric field artifacts interfered with our recordings. Bindman (1964) studied the effects of stimulus duration on burst-firing frequency in anesthetized rat motor cortex; prolonged stimulation of 5-1 0 min with different polarities could either lengthen or shorten the spontaneous electrical activity of the neurons. A current of 2-8 ma delivered to stimulus electrodes submerged in high [KCI] ACSF in our set up could generate a potential field of 15-73 mV/mm. With these fields, we found a current between 4-8 ma (corresponding to an electric field of 33-73 mV/mm) could significantly alter the frequency of burst firing. The mean interspikes interval between burst-firing could be lengthened with increasing the field strength. This finding again shared the similarity to the finding by Creutzfeld et a/. 1 in which a transcortical potential change of 50-75 mV/mm could modify the spontaneous firing of motor neurons in anesthetized cat.

[4] 129w To explain the results of electric fields imposed on our preparation, we consider the potential gradient across various parts of the CA3 pyramidal neurons. In horizontally oriented neurons, where the apical or basal dendrites would be oriented towards the positive electric field, and inward current flow could hyperpolarize the dendrites and the cell soma thus resulting in decreased neu ronal excitability and increased interburst interval. Since these neurons are actually aligned along a horseshoe shaped region from CA4 to CA3, the orientation of the electric field and the synchronizing neurons is complex. This explains some of our difficulty in establishing consistent effects from polarity changes. Again, we speculate whether surgically isolating a region of CA3 with a small cut in the slice would increase the consistency of such experiments.

[5] 65w We recently were able to control the burst-firing frequency of CA3 in hippocampal slices perfusing with high [KCI] ACSF using direct antidromic stimulation of the Schaffer collateral fibers and similar control can be performed with a chaos control algorithmfoot_1 • If such control of burst-firing frequency could be performed with electric field, the prospect of in vivo control without invasive electrodes would be most attractive.

[1] 119w Figure 1: Schematic illustration of arrangement of hippocampal slice and electrodes, viewed from above Potential measurements were then sampled at 1-mm intervals along these imaginary lines. A total of 50 potential points were measured. These electrical potential maps were made from averages of two separate measurements. The measurements were made at the depth of 1 mm from the fluid surface. Using the ellipse tool from Harvard Graphics (Harvard Graphics for Windows Version 1.0, Harvard Graphics, Ocean, NJ, USA), the potential curves were then constructed by connecting the measurements of similar value. There were an average of ten measurements for each complete potential curve that fit inside the grid and a minimum of three measurements for each incomplete potential curve.

[1] 226w Two platinum disc electrodes (5 mm in diameter) embedded in a silastic strip with their center 1 em apart (PMT Corp.) were placed 2 mm beneath the fluid surface and 1 mm beneath the nylon support netting of an interface type perfusion chamber. These electrodes are identical to the subdural strip electrodes used to localize seizure foci in human patients undergoing epi lepsy surgery. The chamber was perfused with high [KCI] arti fic ial ce rebrospinal fluid (ACSF) flowin g at 2 ml min-1 and composed of 155 mM Na+, 14 1 mM c[-, 8.5 mM K+, 1.2 mM Ca 2 +, 1.2 mM Mg 2 +, 1.25 mM Po~-, 24 mM HCO), 1.2 mM so~-, and 10 mM dextrose. The temperature of the chamber was kept at 32-35°C. The two platinum subdural electrodes were connected in parall el w ith a 10 KQ resistor and a photoelectric stimulus isolating unit (Model SIU 15, Grass Corp, West Warwick, Rl, USA) driven by a d igital stimulator (Model S8800, Grass Corp.). Recordings were made with 2-4 MQ glass needle electrodes filled with 150 mM NaCI. To construct maps ofthe electric fields, a square grid of 5 x 5 mm was drawn around each electrode. Nine imaginary lines perpendicular to the horizontal axis connecting the two centers of the electrodes was outlined. These lines were 2.5 mm apart.

[2] 62w To quantify these results, both one-way and multifactor analysis of variance (ANOVA) were applied to test the effect of polarity, field strength, and stimulus duration on interburst interval length; multiple regression analysis was used to determine the correlation of the significant factors seen in the ANOV A. Statistical methods were implemented with Statgraphics Plus 1.0 for Windows (Manugistics, Inc., Princeton, NJ, USA).