Implications of decreased hippocampal neurogenesis in chronic temporal lobe epilepsy

Dentate neurogenesis in animal models of chronic TLE

The dentate neurogenesis has been examined rigorously in two different kainic acid (KA) models of chronic TLE in rats (Hattiangady et al., 2004, 2005a). The first model is a unilateral intracerebroventricular (ICV) KA model in which administration of 0.5 µg of KA directly into the posterior lateral ventricle under anesthesia induces extensive degeneration of CA3 pyramidal neurons and loss of sizable fraction of dentate hilar neurons in the hippocampus ipsilateral to the KA administration (Shetty & Turner, 1995, 1996, 1997a, 1997b, 1999a, 1999b). Following this injury, multiple epileptogenic changes occur in both DG and CA1 subfield, resulting in hippocampal hyperexcitability (Turner & Wheal, 1991; Shetty & Turner, 1997a, 1999a, 2000, 2001; Shetty et al., 2003, 2005a). However, because of the unilateral nature of the injury, the frequency of SRMS has been found to be minimal in this model (Hattiangady et al., 2004). The other model is an intraperitoneal (IP) KA model in which graded IP injections of KA at a dose of 3 mg/kg body weight (BW) (once every hour for 3–5 h) induces continuous stages III–V seizures for over 3 h, resulting in bilateral loss of fractions of neurons in the dentate hilus and CA1 and CA3 subfields of the hippocampus (Hattiangady et al., 2004, 2005a; Rao et al., 2006a). This injury also induces multiple epileptogenic changes in the hippocampus during the silent phase that eventually culminate in chronic epilepsy characterized by robust SRMS (Rao et al., 2006a, 2007), learning and memory deficits, and depression (Stafstrom et al., 1993; Mikati et al., 2001). Interestingly, DG neurogenesis declines substantially when animals exhibit chronic epilepsy in both models.

Dentate neurogenesis in human TLE

Decreased DG neurogenesis observed in animal models of chronic TLE is also supported by the findings in hippocampal tissues resected from human TLE patients (Mathern et al., 2002; Crespel et al., 2005; Pirttila et al., 2005a, 2005b; and see review by Siebzehnrubl and Blumcke, 2008). It also appeared that the extent of decline in neurogenesis is directly proportional to the severity of lesion in chronic TLE (Mikkonen et al., 1998). A recent study confirms decreased DG neurogenesis in TLE patients through observation of decreased synthesis of mRNA for DCX (Fahrner et al., 2007) and absence of cells positive for Ki-67 (a marker of proliferating cells). Thus, studies in hippocampi of both TLE patients and animals exhibiting chronic TLE clearly suggest that DG neurogenesis substantially declines in chronic epilepsy.

Alterations in hippocampal stem cell number

A recent study suggests that the numbers of putative NSCs that are immunopositive for the transcription factor sex determining region of Y-chromosome-2 (Sox2) decline by ~30% in the SGZ of rats exhibiting chronic TLE (Shetty, personal communication). This finding also corroborated with the counts of cells positive for vimentin (i.e., radial glial cells or putative stem cells) in the dentate SGZ-GCL (Shetty, personal communication). These results are, however, in contrast to the human study, where an increased number of cells positive for Musashi-1 (another marker of putative NSCs) was seen in the epileptic hippocampal tissues obtained from TLE patients (Crespel et al., 2005). Most of the Musashi-1-positive cells in this study were found in the SGZ and the hippocampal fissure. Many Musashi-1-positive cells also coexpressed vimentin or nestin, suggesting that these cells are likely NSCs. Musashi-1 is known to potentiate notch signaling that maintains NSCs in an undifferentiated state in pathological conditions (Morrison, 2001) or instructs them to proliferate and differentiate into glia (Imai et al., 2001; Okano et al., 2002) in control conditions. Based on these characteristics, Crespel and colleagues (2005) suggest that increased expression of Musashi-1 during chronic epilepsy instructs NSCs to remain in an undifferentiated state and/or to switch into glial fate. Either or both of these possibilities can reduce the overall DG neurogenesis during chronic epilepsy. From the above, it is clear that currently there is no consensus regarding reductions in the number of NSCs during chronic epilepsy. A rat study suggests ~30% depletion in the number, whereas a human study implies no such reductions (Shetty, personal communication; Crespel et al., 2005). Yet, from both studies, it emerges that substantial numbers of NSCs persist in the SGZ of the chronically epileptic hippocampus (Fig. 2). Persistence of NSCs during chronic epilepsy is encouraging for employing strategies that have the ability to induce neuronal differentiation in newly born cells derived from NSCs.

Changes in the hippocampal microenvironment

It is well known from multiple studies that higher concentrations of neurotrophic factors such as epidermal growth factor (EGF), fibroblast growth factor-2 (FGF-2), insulin-like growth factor-1 (IGF-1), brain-derived neurotrophic factor (BDNF), and vascular endothelial growth factor (VEGF) can enhance DG neurogenesis (Cheng et al., 2001; Lichtenwalner et al., 2001; Jin et al., 2002a, 2002b; Yoshimura et al., 2003; Rai et al., 2007). Consistent with this idea, increased neurogenesis observed shortly after SE or brain injury has been found to be associated with the upregulation of these factors. However, chronic epilepsy has been found to be associated with reduced concentration of many of these factors (Fig. 2). These include FGF-2, IGF-1, and BDNF (Shetty et al., 2003; Hattiangady et al., 2004). The chronically epileptic hippocampus exhibited 75% less FGF-2, 53% less IGF-1, and 64% less BDNF than intact hippocampus (Shetty et al., 2003; Hattiangady et al., 2004). In addition to these, several other factors that are important for DG neurogenesis may decrease during chronic epilepsy. Indeed, a preliminary study shows decreased expression of GDNF during chronic epilepsy (Shetty, personal communication). Because GDNF exhibits both neurogenesis-promoting effects (Chen et al., 2005; Kobayashi et al., 2006) and anticonvulsant properties (Kanter-Schlifke et al., 2007), decreased GDNF may contribute to both diminished neurogenesis and increased frequency of SRMS during chronic epilepsy. Furthermore, an inhibitor of the wingless-type murine mammary tumor virus (MMTV) integration site family protein (Wnt) Dickkopf-1 is elevated in mesial TLE (Busceti et al., 2007). This may also contribute to decreased DG neurogenesis during chronic epilepsy because overexpression of Wnt3 increases neurogenesis in hippocampal NSCs and blockade of Wnt signaling abolishes DG neurogenesis (Lie et al., 2005).

However, VEGF expression in a variety of hippocampal cells was found to increase during chronic epilepsy in both humans and animal models (Rigau et al., 2007). Increased VEGF expression was also associated with increased angiogenic processes and disruption of the blood– brain barrier (Rigau et al., 2007). It is possible that increased VEGF levels maintain the proliferation of NSCs during chronic epilepsy at levels observed in age-matched intact hippocampus (Hattiangady et al., 2005a). This is because VEGF is considered as one of the important mitogenic factors for NSCs. On the other hand, decreased levels of BDNF, IGF-1, FGF-2, GDNF, and Wnt proteins likely dampen the neuronal differentiation of newly born cells. Thus, considerable changes occur in the microenvironment of the hippocampus in chronic epilepsy that likely influence the overall decline in DG neurogenesis.

Chronic inflammation

Hippocampal inflammation commonly observed after injury can considerably suppress DG neurogenesis (Crespel et al., 2002; Ekdahl et al., 2003; Monje et al., 2003). Inflammation persists during the chronic phase of epilepsy, though it is considerably less than that observed shortly after SE (Crespel et al., 2002; Hattiangady et al., 2004; Bonde et al., 2006). A study examining the hippocampi from patients with mesial TLE and hippocampal sclerosis reports association of neuronal loss with considerable gliosis in both CA1 and CA3 subfields (Crespel et al., 2002). Dual immunolabeling analyses revealed overexpression of nuclear factor (NF) kappaB-p65 in reactive astrocytes occupying lesioned areas, suggesting that inflammatory processes are chronically active or transiently reinduced by recurrent seizures in the epileptic hippocampus with sclerosis (Crespel et al., 2002). However, evaluation of activated microglial cells in the chronically epileptic hippocampus at 5 months after the initial KA-induced injury in a rat model of TLE demonstrated considerably reduced density of activated microglial cells, in comparison to their density at 2 weeks after KA-induced injury (Hattiangady et al., 2004). Thus, in the KA model, increased inflammation was seen when neurogenesis was dramatically upregulated, and reduced inflammation was observed when neurogenesis was radically decreased. These results underscore that chronic inflammation is not the major reason underlying decreased neurogenesis in chronic epilepsy. However, some contribution from the persisting activated microglial cells toward reduced neurogenesis cannot be ruled out.

Granule cell dispersion

Granule cell dispersion (GCD) is a characteristic feature observed in ~40% cases of TLE with hippocampal sclerosis and in some animal models of TLE (Houser, 1990; Harding & Thom, 2001; Kralic et al., 2005; Fahrner et al., 2007). A recent study implies that GCD is caused by the displacement of mature granule cells due to a deficiency in reelin (Heinrich et al., 2006; see review by Parent and Murphy this issue). It has also been suggested that seizure-induced increase in DG neurogenesis (during the early phase after an IPI) does not contribute to GCD as the occurrence of GCD was found to be associated with a progressive decline and absence of DG neurogenesis (Heinrich et al., 2006; Fahrner et al., 2007). However, it is also unlikely that GCD causes the decline in DG neurogenesis as decreased DG neurogenesis in animals exhibiting chronic epilepsy has been consistently observed in the absence of GCD (Hattiangady et al., 2004, 2005a). Thus, GCD is another pathology observed in a fraction of TLE cases with hippocampal sclerosis. Although GCD is associated with decreased DG neurogenesis, currently there is no evidence to support that GCD induces decreased neurogenesis in animals exhibiting chronic epilepsy or in human TLE.

Changes in reelin and cystatin C

Reelin is an extracellular matrix protein essential for the proper migration of neurons. Recent studies have suggested a link between DG neurogenesis and reelin expression. First, a study demonstrates reduced neurogenesis with preferential differentiation of newly born cells into astrocytes in mice lacking reelin, also called the reeler mice (Zhao et al., 2007). Second, reduced reelin expression has been observed after the pilocarpine-induced SE in rats, which appeared to induce ectopic chain migration and aberrant integration of newborn dentate granule cells (Gong et al., 2007). Thus, it is possible that reelin deficiency observed during chronic epilepsy (Heinrich et al., 2006) contributes to decreased neurogenesis and aberrant migration of newly born neurons into the dentate hilus. Cystatin C, a protease inhibitor, is another factor that is upregulated during chronic epilepsy (Pirttila et al., 2005a, 2005b). In patients with TLE, higher level of cystatin C expression was associated with decreased neurogenesis. However, the extent of aberrant migration of newly born neurons was enhanced, implying that increased cystatin C levels might be promoting the abnormal migration of newly born neurons. Thus, simultaneous downregulation of reelin and upregulation of cystatin C may contribute to both decreased DG neurogenesis and abnormal migration of newly born neurons during chronic epilepsy.

Dentate hyperexcitability and spontaneous seizures

There is no direct evidence to support that diminished neurogenesis plays a role in the maintenance of SRMS during chronic epilepsy. However, a study suggests that ~14% of newly born neurons in the DG of young rats differentiate into inhibitory GABAergic basket cells (Liu et al., 2003). If this scenario were also true in diseased conditions, dramatically decreased DG neurogenesis during chronic epilepsy would result in minimal addition of new GABAergic basket cells into the DG circuitry. This may contribute to the persistence of reduced inhibitory neurotransmission and exacerbation of dentate hyperexcitability with time. Studies in animal models of TLE suggest both reduced inhibition of granule cells and reduced number of GABAergic basket cells in the DG during chronic epilepsy (Shetty & Turner, 2001; Kobayashi & Buckmaster, 2003; Morimoto et al., 2004). Nevertheless, it remains to be established whether these changes in inhibitory transmission occur because of decreased addition of new GABAergic cells through neurogenesis or simply through the loss of existing GABAergic interneurons over time. Another study, using an electrical stimulation model of SE, suggests that granule cells that are born and integrated into the GCL at extended time points after SE integrate in such a way that they receive reduced excitatory synaptic input and exhibit an enhanced inhibitory synaptic drive (Jakubs et al., 2006). These results imply that functional integration of neurons born in pathological conditions is adjusted to the prevailing functional state in the pathological hippocampal network (Jakubs et al., 2006; Kempermann, 2006). Considering this, it is possible that decreased addition of new neurons into the GCL interferes with the possible spontaneous repair of hyperexcitability in the DG. Thus, in conditions such as chronic epilepsy, where the neural circuitry is already proexcitatory, substantially declined neurogenesis may contribute to the maintenance and/or exacerbation of DG hyperexcitability. However, it should be noted that the pattern of integration of new granule cells reported byJakubs et al. (2006) does not seem to happen in the chemoconvulsant (KA or pilocarpine) models of TLE. This is because, in these models, decreased DG neurogenesis during chronic epilepsy is associated with aberrant migration of considerable fractions of newly born neurons into the dentate hilus (Hattiangady et al., 2004). Additionally, new granule cells that integrate into the SGZ-GCL in these models demonstrate an increased occurrence of basal dendrites (Shapiro & Ribak, 2006; Jessberger et al., 2007; Walter et al., 2007), a feature known to promote aberrant synaptic reorganization in the epileptic hippocampus.

Learning and memory function

It is hypothesized that the maintenance of hippocampal-dependent learning and formation of the temporal clusters of long-term episodic memories require continuous addition of newly functional granule cells into the GCL circuitry (Shors et al., 2002; van Praag et al., 2002; Jessberger & Kempermann, 2003; Kempermann et al., 2004; Aimone et al., 2006; Kee et al., 2007). From this perspective, decreased addition of new functional granule cells into the GCL during chronic epilepsy would likely impair the hippocampal-dependent learning and memory function and culminate in cognitive deficits associated with chronic TLE (Fig. 2). Indeed, clinical studies report that patients with a longer duration of refractory TLE exhibit more severe cognitive impairments (Jokeit & Ebner, 1999; Brown-Croyts et al., 2000; Alessio et al., 2004). Thus, it is possible that cognitive deterioration observed in patients with refractory TLE is at least partially linked to diminished DG neurogenesis. However, a cause–effect relationship between decreased neurogenesis and cognitive impairments during chronic TLE remains to be established in future studies. If the potential linkage turns out be true, strategies that are efficacious for enhancing DG neurogenesis during chronic epileptic conditions will be therapeutic. Nevertheless, it should be noted that KA administration or other brain insults lead to persistent loss of significant fractions of several hippocampal cell types in addition to decreased neurogenesis, which may also affect the learning ability.

Depression

Many studies have suggested a role for hippocampal neurogenesis in mediating some of the behavioral effects of antidepressants (Sahay et al., 2007). Animal studies show that disruption of the antidepressant-induced neurogenesis blocks behavioral improvements mediated by antidepressants (Santarelli et al., 2003; Drew & Hen, 2007; Sahay & Hen, 2007). It has been suggested that reversal of the decreased neurogenesis is one way by which antidepressant drugs exert their positive effects (Malberg, 2004). Thus, reduced DG neurogenesis might contribute to the pathophysiology of depression observed in patients with chronic epilepsy. In this context, development of strategies that enhance the production of new neurons in the chronically epileptic hippocampus may be useful for alleviating depression in epileptic patients.

Grafting of stem cells

Grafting of stem cells into the hippocampus may be beneficial for improving neurogenesis during chronic epilepsy. A recent study demonstrates that grafting of NSCs or glial progenitors is efficacious for improving neurogenesis in the aging rat hippocampus (Hattiangady et al., 2007). The precise mechanisms by which this approach increased neurogenesis from endogenous NSCs were unclear. However, it appeared to be mediated through increased concentration of neurotrophic factors such as FGF-2, BDNF, and IGF-1 as both NSCs and glial progenitors exhibited robust expression of these factors (Hattiangady et al., 2007). As both decreased neurogenesis and reduced concentration of multiple neurotrophic factors that support hippocampal neurogenesis observed in chronic epilepsy also occur during aging (Kuhn et al., 1996; Hattiangady et al., 2005b; Rao et al., 2005; Shetty et al., 2005b; Rao et al., 2006b), grafting of NSCs or glial progenitors appears to be a promising approach for improving neurogenesis in the chronically epileptic hippocampus. Grafting of stem cells into the chronically epileptic hippocampus may also induce other beneficial effects such as seizure control and improved learning and memory (Shetty & Hattiangady, 2007b). For example, if grafted stem cells generate significant numbers of GABAergic interneurons (as observed when they were grafted early after SE, Chu et al., 2004), this may lead to decreased frequency and intensity of SRMS through connectivity of these newly generated interneurons with the host neurons exhibiting increased excitability. Alternatively, grafted stem cells may induce antiseizure effects through secretion of anticonvulsant factors. Thus, stem cell grafting approach for chronic TLE has considerable promise. However, rigorous studies using animal models of chronic TLE are needed in future to comprehend the behavior, differentiation, integration, and neurotrophic activity of grafted NSCs. These studies also need to ascertain the possible long-term efficacy of this approach for restraining seizures, improving DG neurogenesis, reversing learning and memory dysfunction, and alleviating depression.

Administration of neurotrophic factors

Chronic epilepsy is associated with reduced levels of several neurotrophic factors that are considered as positive regulators of DG neurogenesis in the hippocampus that include FGF-2, IGF-1, and BDNF (Shetty et al., 2003; Hattiangady et al., 2004). Therefore, it is possible that dramatically declined neurogenesis during chronic epilepsy is a result of decreased levels of these neurotrophic factors. From this perspective, administration of these factors (either individually or in combination) may stimulate increased neurogenesis from endogenous NSCs that persist in the chronically epileptic hippocampus. Although no studies are available on the effects of administration of these factors into the chronically epileptic hippocampus, studies in other models strongly support their use for improving neurogenesis. These include increased neurogenesis observed in the aging hippocampus following intracerebroventricular administration of IGF-1 (Lichtenwalner et al., 2001) and FGF-2 (Jin et al., 2003; Rai et al., 2007) using osmotic minipumps. Furthermore, as the blood–brain barrier is leaky during chronic epilepsy, likely because of increased VEGF expression and angiogenic processes, these neurotrophic factors may be delivered through either subcutaneous or systemic injections. If successful, this will avoid the invasive approach of delivering factors through cannula implantation into the brain and prevent the possible brain infections linked to the procedure. Although the effects of subcutaneous administration of neurotrophic factors during chronic epilepsy have not been examined so far, subcutaneous administration of FGF-2 and/or BDNF was found to be efficacious for greatly improving neurogenesis in the injured aged hippocampus (Shetty, personal communication). Thus, administration of neurotrophic factors during chronic epilepsy has promise for improving DG neurogenesis.

Physical exercise

Physical exercise may be useful for improving neurogenesis during chronic epilepsy. It is well known that physical exercise induces beneficial effects on overall health and cognitive functions (Cotman et al., 2007). In the hippocampus, physical activity considerably increases neurogenesis through an enhanced proliferation of NSCs and increased neuronal differentiation of newly born cells (van Praag et al., 1999; Kempermann et al., 2000; van Praag et al., 2005). Voluntary running exercise from 3 to 9months of age in mice can maintain DG neurogenesis to levels typically observed in younger animals (Kronenberg et al., 2006). A moderate exercise has also been shown to be useful for enhancing DG neurogenesis and learning and memory function in rats with hippocampal injury (Chen et al., 2006). A recent study shows that exposure to voluntary wheel running after stroke enhances survival of newborn cells, upregulates the phosphorylated cyclic adenosine monophosphate (cAMP) response element-binding protein (pCREB) in the DG, and reverses the ischemiainduced spatial memory impairments (Luo et al., 2007). It also emerges that voluntary exercise increases the levels of BDNF, NGF, VEGF, and other growth factors that stimulate neurogenesis and improves cognitive function (Neeper et al., 1995; Neeper et al., 1996; Fabel et al., 2003; Soya et al., 2007). Considering these, moderate physical exercise during chronic epilepsy appears to be beneficial for improving neurogenesis. This may also have some positive effects on the frequency and intensity of SRMS. Indeed, clinical studies suggest that physical exercise decreases the frequency and severity of seizures (Denio et al., 1989; Eriksen et al., 1994). Studies in animal models are also supportive of the beneficial effects of physical exercise on epilepsy. Physical exercise retards the development of kindling (Arida et al., 1998). Furthermore, in a pilocarpine model of epilepsy, physical exercise diminishes the frequency of seizures (Arida et al., 1999), reduces the hyperresponsiveness of CA1 pyramidal neurons and increases the magnitude of hippocampal long-term potentiation (Arida et al., 2004), and preserves a substantial fraction of hippocampal parvalbumin immunopositive interneurons (Arida et al., 2007). Another study suggests that physical exercise improves learning after KA-induced hippocampal neurodegeneration (Gobbo & O’Mara, 2005). Thus, physical exercise has promise for increasing neurogenesis, reducing seizures, and improving cognitive function during chronic epilepsy. Long-term studies are, however, required to assess the extent of improvements with this approach.

Exposure to enriched environment

Exposure to enriched environment is an attractive approach for inducing robust neuronal plasticity in both intact and injured brain (Nithianantharajah & Hannan, 2006), and may be useful for enhancing neurogenesis in the chronically epileptic hippocampus. Environmental enrichment typically consists of many components, such as expanded learning opportunities, increased social interaction and physical activity, and larger housing (Nithianantharajah & Hannan, 2006). The plasticity mediated by environmental enrichment includes multiple morphological, physiological, neurochemical, and cognitive changes. The morphological changes include increases in the brain weight Bennett et al., 1969; Henderson, 1970), cerebral cortex thickness (Greer et al., 1981), dendritic growth and spine number (Johansson & Belichenko, 2002; Leggio et al., 2005), and synapses (Foster et al., 1996; Mora et al., 2007). The physiological changes comprise enhanced synaptic strength (Foster & Dumas, 2001) and increased excitatory glutamatergic synaptic transmission (Nichols et al., 2007). The neurochemical changes involve increased concentration of neurotrophic factors such as GDNF (Young et al., 1999), neurotrophins BDNF, NGF, and NT-3 (Ickes et al., 2000; Zhu et al., 2006), and VEGF (During & Cao, 2006). Alterations in cognitive function include enhanced long-term potentiation (Duffy et al., 2001; Tang & Zou, 2002) and improved spatial learning and memory performance (Kempermann et al., 1997; Moser et al., 1997; Pham et al., 2002; Huang et al., 2007). Moreover, several studies have shown that exposure to enriched environment increases DG neurogenesis (Kempermann et al., 1997; Olson et al., 2006). Furthermore, a link between increased DG neurogenesis and improved functional outcome in animals exposed to an enriched environment has been suggested in models of aging (Kempermann et al., 2002), traumatic brain injury (Gaulke et al., 2005), Huntington’s and Alzheimer’s diseases (Lazic et al., 2006; Levi & Michaelson, 2007), and cranial irradiation (Fan et al., 2007).

Pertaining to epilepsy and enrichment, a study has shown that preenrichment in adult rats increases the seizure threshold and decreases hippocampal neurodegeneration when challenged with KA (Young et al., 1999; Auvergne et al., 2002). Furthermore, housing animals in environmentally enriched conditions after the induction of SE improves learning and memory abilities and minimizes the probability of developing behavioral abnormalities and depression in immature rats (Faverjon et al., 2002; Rutten et al., 2002; Koh et al., 2005, 2007). Interestingly, immature rats that underwent SE and were housed in enriched environment exhibited increases in DG neurogenesis and phosphorylation of CREB (Faverjon et al., 2002). Thus, several beneficial effects of enrichment have been demonstrated in epilepsy models. However, hitherto, no studies that examine the effects of environmental enrichment on DG neurogenesis or the frequency and intensity of SRMS during chronic epilepsy are available.

Antidepressant therapy

It is now well known that virtually all antidepressant treatments enhance DG neurogenesis (Santarelli et al., 2003; Malberg, 2004; Drew & Hen, 2007; Perera et al., 2007). A greatly enhanced DG neurogenesis following chronic antidepressant treatment is likely mediated by increased: (1) concentrations of serotonin and/or norepinephrine (Duman et al., 1997; Carlezon et al., 2005); (2) cAMP–CREB cascade (Nestler et al., 1989; Nibuya et al., 1996; Thome et al., 2000); and (3) levels of BDNF, VEGF, FGF-2, and IGF-1 (Fabel et al., 2003; Khawaja et al., 2004). Pertaining to the link between DG neurogenesis and depression, a study demonstrates that the disruption of the antidepressant-induced neurogenesis blocks the positive behavioral responses mediated by antidepressants (Santarelli et al., 2003). However, it is not clear whether decreased DG neurogenesis is one of the characteristic pathophysiological features of depression. Thus, from the studies conducted so far, it appears that though depression is not caused by decreased DG neurogenesis, clinical improvements through antidepressant therapy are mediated through increased DG neurogenesis (Drew & Hen, 2007; Sahay & Hen, 2007). As decreased DG neurogenesis, learning and memory impairments, and depression coexist during chronic epilepsy, prolonged antidepressant therapy may be useful. Rigorous studies in animal models of chronic TLE examining all three issues are needed in future studies.