We used ECS to study the effect of repeated brief seizures on the levels of trophic factors in the developing brain. Repeated ECS exposure over 7 day periods had no significant impact on the health or growth of the rat pups, as indicated by measures of brain and body weight. This indicates that our seizure model is distinct from that of supramaximal ECS which negatively impacts both brain and body weight of immature rats (Wasterlain and Plum, 1973
; Vernadakis et al., 1967
). As the rat pups matured, the behavioral seizure manifestations became more organized, with the seizures resembling the adult phenotype by the third postnatal week. However, enhanced seizure severity is not likely to be responsible for the induction of neurotrophic factors at different developmental stages, since seizure intensity started to increase in the second postnatal week without significant upregulation of neurotrophic factors. As reported previously, acute brief non-injurious seizure activity lasting 5–10 sec is generally sufficient to induce significant increase of neurotrophic factors in adults (Conti et al., 2009
; Gwinn et al., 2002
The gradual increase in baseline expression of FGF-2 in the rhinal cortex over the first 3 postnatal weeks is consistent with the developmental profile documented in a previous report (Kuzis et al., 1995
). Most of the postnatal increase derived from changes in the high molecular weight isoform. Similarly, baseline BDNF expression increased with age, but the most pronounced change was somewhat earlier than of FGF-2.
During the first postnatal week, no ECS-induced changes in FGF-2 or BDNF proteins were detected, but during the second postnatal week, repeated ECS evoked a small but significant increase in hippocampal BDNF. By the third postnatal week, ECS treatment caused a net increase in BDNF that was more than five times that observed at P13, and a significant increase in FGF-2 protein in rhinal cortex was seen. It is interesting to note that in the case of both FGF-2 and BDNF, the age of onset of responsiveness to seizure-induced upregulation corresponds to the period when baseline levels have reached peak values. This raises the possibility that during the postnatal maturational increase in BDNF and FGF-2 protein levels, their synthesis is largely regulated by activity-independent mechanisms, and that once a critical level of protein is achieved, activity-dependent regulation plays a greater role.
The most likely explanation for the relative lack of induction of FGF-2 and BDNF in response to seizures during the first two postnatal weeks is either the lack of activity-dependent regulation of these trophic factors or a highly constrained range for activity-dependent regulation, with seizure activity exceeding the upper limit of this window. It is also possible that the seizures in the early neonatal period propagate through networks other than those engaged by the seizures in the older pups. The different seizure patterns observed across different developmental stages would support this possibility. However, given the high susceptibility of the hippocampus to all types of seizures, it is unlikely that this area would not be engaged by the seizure activity in the younger animal (Stafstrom at al., 1992
Baseline levels of NGF in the hippocampus and frontal cortex remained relatively constant over the developmental period examined in our study. This is in agreement with the observations of Das et al (2001)
who found little change in NGF protein in hippocampus between P7 and P21. Repeated ECS seizures did not alter NGF expression in the hippocampus and frontal cortex in the rat pups during any of the postnatal ages examined. The absence of an effect in the immature frontal cortex contrasts with the ECS-induced increase, albeit modest, in NGF protein in adult frontal cortex (Conti et al., 2009
; Angelucci et al., 2002
The model of seizures that we employed allows for 24 h intervals between seizure sessions. This treatment regimen is sufficient to cause maximal induction of FGF-2 and BDNF in adults and, as we have seen here, in 3 week-old pups. To address the question of whether more frequent seizures may trigger an upregulation in less mature animals, we conducted pilot studies (unpublished data) in which ECS sessions were given at 6 h intervals between P6 and P8 (i.e., 10 sessions of 3 ECS/session over 54 h). Under these conditions, we saw a 40% increase in hippocampal BDNF, but because this was associated with a significant decrease in the growth rate of the pups, the findings are inconclusive. Nevertheless, this increase was still several fold lower as compared with the induction following seizures during the third postnatal week.
The relative absence of upregulation of trophic factors by recurrent seizures in early postnatal development suggests that seizure activity in this period may not promote neuroplasticity and synaptic remodeling to the extent that it does in the adult brain. This may explain the fact that long-term outcomes following seizures in the immature brain are different from those in adults. For example, an increase in BDNF and TrkB activation is thought to contribute to epileptogenesis (Binder et al., 1999
), so the relative lack of seizure-induced BDNF induction may explain the decreased epileptogenicity (i.e., induction of spontaneous seizures) following exposure to status epilepticus during the first two postnatal weeks (Danzer et al., 2004
). Spontaneous seizure activity developed after status epilepticus only in pups older than P18 (Danzer et al., 2004
; Priel et al., 1996
; Stafstrom et al., 1992
). In addition, repeated pentylenetetrazol-induced recurrent seizures during the second postnatal week did not induce later spontaneous seizures (Huang et al., 2002
Our data has implications for the potential therapeutic impact of ECT in children (Stein et al., 2006
). If induction of neurotrophic factors is an important component of the therapeutic action of ECT, the relative lack of this component in early childhood may predict reduced efficacy of this treatment approach in this population. On the other hand, we have found a substantial ECS-induction of at least two neurotrophic factors in pre-adolescent animals, suggesting that by late childhood, these mechanisms are operative. However, it should be noted that the lack of increase in NGF protein in frontal cortex and hippocampus and the lack of increase in FGF-2 in frontal cortex are features that distinguish the effect of ECS on the immature brain from that in adults. The extent to which these differences may affect the therapeutic efficacy of ECT are unknown, but they underscore the need for further characterization of the molecular and behavioral consequences of ECS in immature animals.
In conclusion, while the upregulation of neurotrophic factors may mediate many of the short and long-term effects of seizures in the mature brain, whether therapeutic (as in the case of ECT for affective disorders) or deleterious (as in the case of epileptogenesis), a similar extent of seizure-related upregulation does not occur in the immature brain, corresponding to the period of early childhood. This may have implications for the application of ECT to children, as well as for understanding the neurodevelopmental impact of childhood seizures.