Abstracts

Rationale: Delayed neuronal death is a well-established phenomenon in the hippocampus and other brain regions. The process is of interest as a means to explain clinical deterioration after acute brain injury. However, mechanisms underlying delayed neuronal death and its relationship to apoptosis remain poorly understood. Recently, new fluorophores and microscopy techniques have made possible more detailed explorations of this process. We describe our initial findings in an in vitro preparation in which this process can be studied. Methods: We evaluated the death of neurons in a chronically epileptic in vitro preparation in which multiphoton microscopy could be performed over a period of several days. Organotypic hippocampal slice cultures were made from wild-type C57BL/6J mice, and imaged with transgenic fluorophores as well as the Na+ dye SBFI-AM. Organotypic slice cultures were prepared on P6 and incubated in vitro until use, with SBFI added 24 hours prior to imaging. Two-photon imaging was used to excite SBFI at both Na+-sensitive and -insensitive wavelengths, allowing for the ratiometric determination of the [Na+]i. Propidium iodide (PI), an indicator of cell death, was used as a costain in most experiments to exclude moribund neurons from the analysis. Results: The earliest detectable events in the neurons post-trauma were an increase in caspase activity (as indicated by FLICA positivity), a reduction in the emission of virus-induced and transgenically-expressed fluorescent proteins (TurboRFP and Clomeleon, respectively), and an apparent retraction of all dendrites and axons. Next, neuronal membrane permeability progressively increased over several days and esterifed dyes such as SBFI-AM and Fura-AM permeated the cytosol. Cell membrane permeability to Na+ increased, which was associated with decreased membrane potential and increased cytoplasmic Na+ concentration with minimal change in neuronal volume. Permeability to PI was a late event, and neurons with cytoplasmic Na+ concentrations greater than 100 mM were most likely to stain for propidium iodide. Cytoplasmic Ca2+ also increased a few hours before cells became unrecognizable. The terminal event was a sudden reduction in neuronal volume that appeared to be associated with engulfment by microglia. Mitochondrial potentials and Na+/K+ ATPase activity were sustained throughout the process. Membrane permeability was reduced by cyclooxygenase (COX-2) inhibitors and Bax antagonists, but no intervention completely reversed the process. Conclusions: We describe an in vitro model of delayed neuronal cell death in the developing hippocampus.  Dying cells were rapidly removed from neuronal circuits, and thus are unlikely to contribute directly to seizures after brain injury. However, they continued to consume oxygen and glucose, which might be in short supply in the penumbra of injury. Delayed neuronal death was characterized by a prolonged phase of progressively increasing membrane permeability that could be blocked, without altering the course of neuronal death. Funding: NIH/NINDS 5R01NS034700-22