Abstracts

Rationale: Stroke in the immature brain presents with seizures which are treated with anticonvulsants. Post-stroke neurogenesis in the SGZ of the immature brain is decreased and may contribute to long-term cognitive impairments. Histone deacetylase (HDAC) inhibitors prevent HDAC from blocking DNA transcription and replication. Valproic acid (VPA) is a HDAC inhibitor, but also enhances GABA neurotransmission and blocks sodium channels. We aimed to determine the impact of chronic HDAC inhibition on post-stroke recovery in the immature brain. Methods: P12 CD1 mice received right common carotid artery ligation or sham surgery and acute seizures were scored. Mice received either VPA (200mg/kg) or vehicle twice daily from postnatal day (P)16 to P28. 5-bromo-2'-deoxyuridine (BrdU; 50mg/kg) was injected five times between P24 and P26 prior to sacrifice on P42. Percent of brain atrophy was quantified and immunohistochemistry was performed for Brdu/NeuN colabeling in the SGZ. In a subset of mice, serum was collected at P26 and analyzed for concentrations of VPA. Results: Mortality was increased, but not significantly in VPA-treated ligated mice (6/21 mice; 28.5% compared to 2/20; 10.0% in saline-treated mice). VPA-treated mice exhibited significantly lower body weights than saline-treated mice (Repeated measures ANOVA, p=0.03 injured, p=0.04 uninjured, and p=0.003 non-ligated controls). This difference was apparent from P17, the day after VPA was first administered, to P28 but disappeared by P42, 2 weeks after drug administration ceased. Hemispheric and hippocampal atrophy was not significantly different in the VPA- compared to vehicle- treated stroke injured mice. Among injured VPA-treated mice, there was a 208% increase in BrdU/NeuN-positive cells in the ipsilateral dentate gyrus (DG) when compared to counts in saline treated mice (p= 0.024); there was a 55.4% increase in contralateral DG (p= 0.023). In uninjured VPA-treated mice, there was a 49.4 and 61.5 % increase in cell counts in contralateral and ipsilateral DG versus vehicle treated (p= 0.007 and 0.004 respectively). Peak drug levels of VPA in serum one hour after dose were: 251.7 ± 13.3 mcg/ml with a maximum half-life of 1.24 ± 0.26 hours (4/8 trough samples were below detectable levels; the average trough level of the remaining 4 was 2.1 ± 0.23 mcg/ml). Conclusions: Chronic VPA treatment increased SGZ neurogenesis. This effect may be related to HDAC inhibition given that previous studies with Trichostatin A (TSA), a strong selective HDAC-inhibitor, produced similar results; the impact of this increased SGZ neurogenesis requires functional assessment at about P50. The transient increased lag in weight gain in VPA- treated mice does not appear to be related to HDAC inhibition however, since this effect was not seen in TSA- compared to vehicle- treated mice; it may be related to valproate toxicity from oxidative stress or impact on mitochondrial function. Chronic administration of valproate results in a very short half-life in these juvenile mice.