1. Behavioral Tests.
Epileptic animals showed increased locomotor activity, but 2-DG treatment was not able to reverse kainic acid-induced enhancement in locomotor activity in the epileptic group.
The epileptic and epileptic+2DG animals travelled a longer distance than the control group (2749 ± 224, 2276 ± 125.8, 4207 ± 379.7, 4140 ± 426.8 cm for control (N=8), control+2-DG (N=6), epileptic (N=8), and epileptic+2-DG (N=8) groups, respectively for all the behavioral tests unless noted, P<0.05 for both; Fig. 2 A). Furthermore, the velocity of the movement had a similar pattern (4.53 ± 0.38, 3.79 ± 0.20, 7.01 ± 0.63, 6.90 ± 0.71 cm/s, P<0.05 for both; Fig. 2B) indicating more locomotor activity in epileptic and epileptic+2-DG animals, and 2-DG treatment could not reverse the changes.
Epileptic animals expressed reduced anxiety-like behavior, 2-DG only slightly reversing the changes.
Less anxiety-like behavior in the epileptic animals compared to control group was extrapolated from more time spent in the open arms (43.1 ± 6.34, 23.68 ± 8.93, 115 ± 11.47, 109.8 ± 23.56 (N=7), P<0.01; Fig. 2C), more open arm entry (7± 1.52, 2.83 ± 1.01, 18 ± 2.68, 11.43 ± 3.21 (N=7), P<0.01; Fig. 2D), smaller latency of first entry to open arm (122.7 ± 31.55, 181.9 ± 39.45, 20.95 ± 5.91, 56.49 ± 20.02, P<0.01; Fig. 2E), and more head dipping frequency (23.56 ± 2.10, 20.17 ± 2.96, 40.56 ± 4.41, 49.25 ± 7.02, P<0.05; Fig. 2F). Nevertheless, body stretching frequency, which is deemed to reflect anxiety level the most [22] is not significantly different between the groups (14.58 ± 0.84, 16.5 ± 0.67, 18.9 ± 1.69, 20.25 ± 3.16; Fig. 2G). Even though 2-DG seems to have slightly reversed the alterations (number of open arm entry and latency of first open arm entry), the drug group cannot resemble the control group. 2-DG had a similar effect on control+2DG animals as it had on the epileptic animals (Fig. 2).
Increased locomotor activity seems to be the main player in the zero-maze test.
ANCOVA (analysis of covariance) revealed that the locomotion state of the animals affects the zero maze parameters as a covariate. Pearson’s test (or Spearman’s test) demonstrated a significant correlation between moved distance in the open field and all the zero maze parameters except for body stretching frequency; there was no interaction between the group and locomotor variables hence, a necessary prerequisite of ANCOVA test was met. Masking locomotion’s effect by the ANCOVA test, animals in the control, epileptic, and epileptic+2-DG showed no significant difference in the parameters of anxiety-like behavior but head dipping frequency (Table1).
Table 1
Correlation between locomotion status and zero maze parameters and the effect of masking locomotion’s interference in zero maze results.
|
Open-arm time
|
Number of open-arm entry
|
Latency of first open-arm entry
|
Head dipping frequency
|
Body stretching frequency
|
Correlation
|
r=0.588**
|
r=0.511*
|
r=0.39*
|
r=0.635**
|
r=0.224
|
ANCOVA test p-value
|
0.16
|
0.15
|
0.08
|
0.04
|
-
|
ANOVA test
p-value
|
0.019
|
0.03
|
0.033
|
0.006
|
-
|
There was a significant correlation between locomotion status and the zero maze parameters except for body stretching frequency (the first row). ANCOVA test results revealed that there was no significant difference in the zero maze parameters between control, epileptic, and epileptic+2-DG animals after masking locomotion’s interference (second row) while ANOVA test results had shown a significant decrease in anxiety status of the animals (last row; and figure 2). Head dipping frequency, however, was significantly different between the groups even after masking locomotion’s effect. *P < 0.05, **P < 0.01. |
Epileptic animals appeared to be depressed-like, 2-DG partially leading to more sweet water consumption.
The ratio of sweet water to tap water consumption was higher in the epileptic animals, being partially reversed in the epileptic group receiving 2-DG (9.57 ± 0.96, 8.68 ± 2.10, 3.82 ± 1.54, 6.25 ± 1.31, P<0.05; Fig. 2H). It is worth noting that depression-like behavior deduced from the result of this test suggests an all-or-none pattern. It is to say, each epileptic animal was either depressed (tap water consumption was much higher than sweet water consumption) or not-depressed (sweet water consumption was much higher than tap water consumption). 2-DG had no considerable effect in the control+2-DG group (Fig. 2H).
2. Electrophysiology
2-DG failed to suppress interictal sharp-wave complexes.
A preferential enhancement of glycolysis in the activated brain during diseases, like epilepsy, has been reported [4], and based on this, the antiepileptic effect of glycolysis inhibition has been proposed [5, 7, 9, 23]. Therefore, we next examined the electrophysiological consequences of 2-DG treatment, which inhibits competitively glycolysis and prevents ATP production, in a kainic acid model of temporal lobe epilepsy.
13 hours of continuous LFP recording was conducted from the dorsal hippocampus of the epileptic animals. Sharp-waves were frequently seen in epileptic group while were never observed in control animals (Fig. 3A&B). analysis of these sharp-wave complexes revealed no significant difference in the ratio of total sharp-wave complex time (1.87 ± 0.71, 1.25 ± 0.32 in epileptic+2DG and epileptic+saline groups respectively; Fig. 3C), the ratio of mean sharp wave duration (1.15 ± 0.09, 1.03 ± 0.08; Fig. 3D), and sharp wave frequency (1.02 ± 0.07, 1.01 ± 0.04; Fig. 3E) between the Epileptic-2-DG (N=5) and Epileptic-Saline (N=5) groups; The ratios were obtained by division of the quantities taken in 21st and 28th days in each group (28th/21st). Although the parameters are reported for the whole 13 hours, the sharp-wave complexes were scrutinized in one-hour epochs during the whole 13 hours and no notable suppression was noted (data not shown). LFP results suggest that 2-DG is not able to suppress epileptiform activity at circuit-level.
Glycolysis inhibition by 2-DG cannot completely reverse epilepsy-induced cellular electrophysiological changes in the CA1 pyramidal neurons.
By applying hyperpolarizing currents (Fig. 4A) Patch-clamp Findings showed that resting membrane potential remained unchanged either following induction of epilepsy (-60.43 ± 2.424 in control (N=8); -59.12 ± 3.527, in the epileptic group (N=6)) or inhibition of glycolysis in the epileptic group (-60.37 ± 1.4 mV, epileptic+2-DG group (N=7); Fig. 4B). The membrane resistance was also not affected by intrahippocampal kainic acid injection (122.3 ± 18.59 MΩ in control vs epileptic group 119.2 ± 21.11 MΩ) and 2-DG treatment (105.1 ± 22.90 MΩ; Fig. 4C). However, the time constant was shorter in both epileptic and epileptic+2-DG groups (5.76± 1.26, P<0.01 and 5.12±0.80 ms, P<0.001, respectively; Fig. 4D) compared to control group (13.11 ± 1.47ms). Furthermore, the membrane capacitance was decreased in these groups (0.05±0.01 pF, P<0.01 and 0.06±0.01 pF, P<0.05, respectively; Fig. 4E) when compared to the control neurons (0.17±0.03 pF). Changes in membrane capacitance were associated with a significantly smaller cell size of survived pyramidal neurons (Figs. 4F, G&H) both in epileptic ((control group: 13.89 ± 1.24 µm (N=4) vs epileptic group: 8.45 ± 0.09 µm(N=4), P<0.05) and in the epileptic plus 2-DG group (8.7 ± 0.15 µm, P<0.05; Fig. 4I). As it can be seen in Nissl-stained sections, induction of epilepsy also led to a remarkable cell loss in the epileptic group and inhibition of glycolysis did not stop the loss of neurons in the epileptic group receiving 2-DG (Fig. 1D).
Whole-cell current-clamp recordings revealed that neurons from epileptic mice showed enhanced neuronal excitability (Fig. 5A) and exhibited a significant higher firing frequency (1.28 ± 0.42 Hz (N=5) in control neurons vs 4.93 ± 0.63 Hz (N=3), P<0.05; Fig. 5C), but treatment with 2-DG following induction of epilepsy reduced the neuronal excitability and firing frequency (2.18 ± 0.74 Hz (N=5), Figs. 5A&C).
Furthermore, although the amplitude of action potential was not affected by epilepsy induction when compared with the control group (80.74 ± 4.81 mV vs 85.16 ± 7.52 mV) inhibition of glycolysis in the epileptic mice who received 2-DG resulted in a significant increase in the AP amplitude (105.4 ± 7.06 mV (N=4), P= 0.033; Fig. 5D). Induction of epilepsy following intrahippocampal injection of kainic acid led to a slightly slower depolarization phase of action potential (Control:1.09 ± 0.25 ms vs epileptic:1.28 ± 0.07 ms), as evidenced by an increased rising tau compared to control cells, but 2-DG treatment resulted in a significantly faster time constant of the rising phase of the action potential, as compared both to the control and epileptic groups (0.41 ± 0.05 ms, P<0.05; Fig. 5E). We also found that either induction of epilepsy or 2-DG treatment had no significant effect on the AP half-width (Fig. 5F). Nevertheless, due to afterdepolarization (ADP), which is only noted in the epileptic+2-DG group (amplitude=49.891±8.312 mV, decay tau=2.584±0.672 ms, area under curve=143.6±50.87 ms. mV; Fig. 5b), the duration of AP (measured at threshold voltage) was significantly increased both in epileptic and epileptic+2-DG groups compared to control group (4.19 ± 0.55, 9.81 ± 1.63, 16.15 ± 1.40 ms (N=4), in control, epileptic and epileptic+2-Dg groups, respectively, P<0.05; Fig. 5G). The amplitude of AHP did not differ significantly between the groups (Fig. 5H). Inhibition of glycolysis, but not induction of epilepsy alone, was accompanied by a significant increase in the coefficient of variation of interspike interval (ISI) when compared to either control (0.64 ± 0.04 vs 0.97 ± 0.11, P<0.05) or epileptic (0.60 ± 0.12, P<0.05; Fig. 5I) groups. This may indicate the irregularity of firing pattern of the CA1 pyramidal neurons in epileptic mice that received 2-DG.
The latency of the first post-inhibitory rebound spike, following a hyperpolarizing ramp current, was significantly shorter in the epileptic neurons than in control ones (34.19 ± 7.85, 5.36 ± 0.75 ms, P<0.01); likewise, it was significantly lower in epileptic+2-DG group compared to epileptic group (7.15 ± 0.99 ms, P<0.05, Fig. 5J&K).
To further assess the impact of 2-DG treatment on the firing pattern, we analyzed quantitative burst activity. Burst activity in hippocampal pyramidal neurons has already been demonstrated in epileptic cells [24, 25]. Here, neurons obtained from all the epileptic animals showed severe burst activity (Fig. 6A); in the epileptic+2-DG group, however, neurons from 2 of the 5 animals showed no burst activity; the other three animals showed attenuated burst activity compared with the epileptic animals (Fig. 6B), Even though the difference between the parameters did not reach to significance level; the number of bursts decreased by 61% (19 ± 3.60, 11.67 ± 2.33 for epileptic and epileptic+2-DG groups respectively; Fig. 6C), the mean duration of each burst diminished by 207% (4.46 ± 1.27, 1.45 ± 0.62; Fig. 6D), the mean AP number in each burst decreased by 71% (11.78 ± 4.98, 6.88 ± 2.13; Fig. 6E), and the mean pause between the bursts dwindled by 41% (1.99 ± 0.88, 2.812 ± 0.45; Fig. 6F). Patch-clamp results, hence, suggest that 2-DG could notably reverse the alterations that ensued from epilepsy induction at the cellular level. Moreover, it led to alterations in AP properties that did not assemble the control animals.
3. Histology
Kainic acid led to severe pyramidal and NADPH-d positive cell loss in the dorsal CA1 pyramidal layer.
Nissl staining showed that the integrity of the dorsal CA1 pyramidal cell layer was disrupted and remarkable cell loss, especially in CA1, was evident following intrahippocampal kainic acid administration (Figs. 1C&D). In parallel, NADPH histochemical staining for nitrergic neurons revealed that the number of NADPH-d+ cells in the pyramidal layer of dorsal CA1was significantly decreased in the intrahippocampal kainic acid injected animals when compared with the control group (11.27 ± 0.76, 4.93 ± 0.84 in control (N=4 mice, 28 sections) and epileptic (N=3 mice, 21 sections) groups respectively; P<0.05. Figs. 7A, a, B, b, C, c, D, d & E).
2-DG treatment increased NADPH-d+ neurons in the dorsal hippocampus of control animals while failing to alter their number in the kainic acid-treated animals
In control+2-DG group, elevated NADPH-d+ was observed compared with control group (20.71 ± 1.01 in the control+2DG group (N=4 mice, 28 sections); p-value < 0.001; Fig. 7E). In the epileptic group receiving 2-DG, however, no significant alteration in the number of NADPH-d-stained neurons (6.61 ± 1.50 (N=4 mice, 28 sections); p-value = 0.73; Fig. 7E) was noted compared with the epileptic group. In contrast, in the contralateral dorsal hippocampus of epileptic and the epileptic+2-DG group, the NADPH-d+ cell number was approximately 3-fold more in comparison with the ipsilateral side (15 ± 1.375 in the epileptic group, P<0.05, and 13.07 ± 1.64 in the epileptic+2-DG group, P<0.05; Figs. 7F&G); Note that the number of NADPH-d+ neurons in the contralateral hippocampus of these groups was even mildly more compared to control group. There was no notable difference between ipsilateral and contralateral hippocampus in control group (data not shown).