Epilepsy is a neurological disorder that affects approximately 65 million people worldwide. While numerous treatment options are available to control seizures associated with epilepsy, approximately 20–30% of epileptic patients are resistant to treatment. Pharmaceutical and biomedical device companies continue to develop new treatments that are more efficacious, while minimizing undesirable side-effects. One such drug, levetiracetam, alone as a monotherapy, or combined with another treatment as an adjunct, is widely used to control partial onset and generalized seizures (1). Despite its use as a relatively safe and effective anti-convulsant treatment option, the precise mechanism(s) of action are not fully understood.

The anti-epileptic effects of levetiracetam may occur, at least in part, by acting directly on neurons. For example, levetiracetam is known to bind to the synaptic vesicle protein SV2A (2) and to inhibit presynaptic calcium channels (3). Studies have shown that levetiracetam may also exert its anti-epileptic effects by reducing calcium currents in CA3 pyramidal neurons of the hippocampus (4), or by rescuing neurons in the hippocampus and dentate gyrus from death (5). Another possibility is that levetiracetam alters mitochondrial membrane potential (mΔΨ), although the reports using the in vivo perforant pathway stimulation paradigm are conflicting. In one study, levetiracetam effectively mitigated mitochondrial dysfunction in the hippocampus following established status epilepticus (6), but not at acute time points after the onset of status epilepticus (7). Thus, it is unclear if, or to what extent, the therapeutic effects of levetiracetam can be attributed to its neuronal interactions.

While the purpose of the current study is to determine if there are direct effects of levetiracetam on neuronal and immune cells, other studies have suggested glial cell mechanisms of action. For example, Ueda et al. (8) suggested that levetiracetam exerts its neuroprotective effects through its actions on glial cells. Similarly, Haghikia et al. (9) showed that levetiracetam has anti-inflammatory effects on astrocytes. Consistent with this finding, Kim et al. (10) showed that levetiracetam reduced gliosis in epileptic brains and inhibited IL-1B, and Stienen et al. (11) showed that the anti-inflammatory effects of levetiracetam on astrocytes may be mediated by TGFβ1.

In addition to effects in the CNS, levetiracetam could also be exerting its effects in the periphery. Supporting this notion are the results of a previous study demonstrating that levetiracetam inhibits the function of some CD8⁺ T Lymphocytes (12). Such interactions with the peripheral immune system might explain the increased incidence of pharyngitis and rhinitis in levetiracetam-treated patients (13–18). However, studies are lacking that provide a thorough analysis of the effects of levetiracetam on peripheral immune cells.

Understanding the mechanism(s) of action of levetiracetam is important because this knowledge could lead to more efficacious treatments and better understanding of the epileptic condition. Due to the lack of a unified theory for the mechanism(s) of action, the current study was designed to determine if levetiracetam affects the mΔΨ of peripheral immune cells and neuronal cells. Moreover, this study sought to address possible immune-mediated mechanisms by determining if levetiracetam alters the expression of immune receptor–ligand pairs.

Previous studies have indicated mitochondrial differences in the presence of levetiracetam (6, 7). Therefore, we determined if these differences were specific for neuronal or immune cells. Analysis of mΔΨ in C17.2 cells revealed no significant differences in the absence of NGF (Figure 1A). In the presence of NGF, levetiracetam resulted in a significant increase (Figure 1A) in mΔΨ at all concentrations tested (1.5 μm. p < 0.03; 15 μm, p < 0.05; 0.15 mM, p < 0.04; 1.5 mM, NS). It is pertinent to note that treatment with levetiracetam did not cause any observable alterations to the morphology of the C17.2 cells, either with or without NGF (data not shown). In contrast to the increased mΔΨ in the presence of NGF and levetiracetam, the impact of levetiracetam on spleen cells (Figure 1B) was a significant reduction in mΔΨ (p < 0.007). This reduction appeared to be invariant chain dependent, as splenocytes from mice deficient in invariant chain showed no significant changes in mΔΨ in response to levetiracetam (Figures 1C,D).

Elevated mΔΨ can be associated with elevated CD95 (21, 22). Therefore, to address the possibility that levetiracetam alters receptor–ligand pairs on neurons, we used the mouse neuronal stem cell line, C17.2, which can be differentiated in the presence of NGF. We assessed CD95, a member of the BGF superfamily and its ligand, FasL (CD178) to determine if levetiracetam can influence cell proliferation, differentiation, and survival. We also examined alterations in the co-stimulatory molecules B7.1 (CD80), or B7.2 (CD86) to assess the potential of levetiracetam to alter co-stimulation of T cell activation. The results from analysis of C17.2 cells revealed that levetiracetam treatment alone had no significant effects on CD95 (Figure 2A), CD178 (Figure 2B), CD80 (Figure 2C), or CD86 (Figure 2D). In the presence of NGF, no significant differences were observed for CD95, CD178, CD80, or CD86 at the 1.5 or 15 μm concentrations. However, at 0.15 and 1.5 mM, a significant increase in CD95 (p < 0.02 and p < 0.001, respectively) was observed. At these latter two concentrations, no significant differences were observed for CD80 or CD86. For CD178, no significant differences were observed for the three lowest concentrations of levetiracetam, but at the 1.5 mM concentration, a significant increase was observed for CD178 (p < 0.05).

In addition to examining neuronal cells, we also examined peripheral immune cells from the spleen. We examined numbers of T cells and numbers of MHCII⁺ cells (which includes macrophages and B cells), as well as CD95 expression on these cells. The results showed that levetiracetam treatment resulted in no significant effect on the number of CD3⁺ T cells (Figure 3A), MHCII⁺ (Figure 3B) cells, nor on the levels of CD95 expression by T cells (Figure 3C), and non-T cells (Figure 3D). In addition, we examined overall levels of MHCII and CLIP on non-T cells (Figures 4A,B) to address the possibility that levetiracetam alters immunogenicity of peripheral immune cells. No changes were observed for either of these variables (Figures 4A,B). To further detect levetiracetam-induced changes in processing or presentation by immune cells, we assessed lysosomal acidity and found no significant changes (Figure 4C).

Levetiracetam is well established as a beneficial anti-seizure medication and as an adjunct to other anti-seizure medications. The molecular mechanisms accounting for the efficacy of levetiracetam for seizure activity are largely unknown. The results from the present study suggest that levetiracetam induces expression of CD95 and CD178 on NGF-treated C17.2 neuronal cells. The results also demonstrate that the increased mΔΨ in response to levetiracetam on C17.2 neuronal stem cells requires the presence of NGF. This is likely due to the differentiating effect of NGF on neural stem cells. In contrast, the study shows that levetiracetam lowers the mΔΨ of splenocytes and this effect is dependent on intact invariant chain. These results suggest that both neuronal and non-neuronal anti-epileptic activities of levetiracetam involve control over energy metabolism, more specifically, mΔΨ.

Epilepsy has traditionally been considered primarily a neuronal disease. Growing evidence also implicates astrocytes, microglia, peripheral leukocytes, and blood–brain barrier breakdown in the pathogenesis of epilepsy. Here, we show two novel observations that potentially link peripheral leukocytes to neurons. Our results suggest that levetiracetam affects mitochondrial energy metabolism as reflected by changes in mΔΨ. Interestingly, these changes are inversely related when comparing splenocytes to NGF-treated neuronal stem cells. That is, levetiracetam causes a statistically significant increase in the membrane potential of NGF-treated C17.2 cells and a significant decrease in mΔΨ of levetiracetam-treated spleen cells. It is pertinent to note that in mice deficient for CD74, the levetiracetam-induced change to splenocyte mΔΨ was ameliorated. Thus, it is possible that levetiracetam-induced changes in mitochondrial activity result from cell–cell contact because CD74 can be expressed on the cell surface and mediate interactions with other cells through its cognate ligand, CD44. Alternatively, the requirement for CD74 to see the effects.

Previous studies have suggested that mitochondrial dysfunction contributes to the epileptic condition (23). The putative functional significance of levetiracetam-induced alterations to mΔΨ in the epileptic brain is its known effects on proton transport. Previous studies have demonstrated that alterations to the mΔΨ in neurons (24) and non-neuronal cells (25), directly influences ion concentrations in the cytosol, thereby influencing plasma membrane potential. Alterations to mΔΨ have also been shown to influence oxidative stress (26), which may be another anti-epileptic mechanism. A third potential mechanism through which altered mΔΨ could influence seizures is by altering the cytosolic pH. An acidification of cytosol as a result of protonation is known to hyperpolarize the plasma membrane (27), which may raise the seizure threshold. Support for this latter suggestion is observed in epileptic hippocampal slices where the pattern of epileptic activity corresponds to mΔΨ and ion concentration (23).

Previous work from our lab demonstrated that Fas/FasL interactions can facilitate neurite outgrowth subsequent to nerve crush injury (28). CD95 and its ligand CD178, a member of the NGF/NGF receptor superfamily of death-inducing receptor–ligand pairs. Many members of this superfamily are involved in cell fate decisions including cell death, cell proliferation, and differentiation. We addressed the possibility that if levetiracetam altered mitochondrial activity, it might also affect CD95 expression because elevated mΔΨ can be associated with elevated CD95. The results from the current study are consistent with this idea because we found that in neuronal cells, in the presence of NGF and >0.15 mM levetiracetam, mΔΨ is increased as is Fas expression. Therefore, levetiracetam may be involved in stabilizing mΔΨ in the presence of elevated levels of NGF.

Another potential effect of levetiracetam on leukocytes could be related to some of the side-effects associated with levetiracetam. In particular, an increased incidence of pharyngitis and rhinitis has been observed in levetiracetam-treated patients (13–18). These findings are consistent with a role for alterations to an effective immune response that may involve alterations in CD74 expression and function. A second possibility is related to the finding of reduced mΔΨ in splenocytes, which may reflect altered levels of immune function, including increased inflammation accounting for pharyngitis and rhinitis.

Overall, the data from the current study indicate that levetiracetam differentially affects the mΔΨ of neuronal C17.2 and non-neuronal splenocytes. The results also show that in the presence of elevated NGF, neuronal C17.2 cells express CD95 and CD178. The results from this study could help to explain some of the mechanisms of action of levetiracetam, including some of its side-effects. More studies are needed to better understand the implications of these findings so that more efficacious treatments with minimal side-effects can be developed.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.