Synaptic structure and function in area CA1 does not account for hippocampal hyperexcitability in slices from symptomatic Mecp2 null mice

CALFA G, AMARAL MD, HABLITZ JJ, POZZO-MILLER L

Leesburg, Virginia

Congreso; 11th Annual Rett Syndrome Symposium, International Rett Syndrome Foundation (IRSF); 2010

IRSF

Partial and generalized, convulsive or silent (i.e. absence) seizures are common clinical manifestation in Rett syndrome patients. Generalized myoclonic jerks coupled with recurrent abnormal EEG discharges were also observed in Mecp2 null mice. Here, we evaluated hippocampal network excitability by voltage-dye imaging in slices from symptomatic Mecp2 null mice and age-matched wildtype littermates (P35-P42). Acute hippocampal slices (300μm-thick) were stained with RH-414 dye (30μM), and its voltage-sensitive signals evoked by single afferent stimulation of Shaffer collaterals (100μsec, 30μA) were imaged in CA1 with an array of 464 photodiodes. Voltage-sensitive dye (VSD) signals had larger peak amplitude, were longer lasting and spread more in area CA1 of Mecp2 null slices than in wildtype slices. Consistently, the input/output relationship of CA3-CA1 extracellularly recorded excitatory postsynaptic potentials (field EPSPs) had a steeper slope in Mecp2 null slices than in wildtype slices. Interestingly, these differences in the amplitude, duration and spatial spread of VSD signals were not observed between slices of presymptomatic Mecp2 null and age-matched wildtype mice (P16-23). However, all these parameters of VSD signals were significantly larger in the younger slices than in older slices of both genotypes, suggesting that hippocampal network excitability undergoes a developmental decrease that seems to be impaired in Mecp2 null slices. Quantitative electron microscopy of area CA1 revealed no significant differences in the density of asymmetric synapses on dendritic spines (presumptive excitatory) or of symmetric synapses on dendritic shafts (presumptive inhibitory) between symptomatic Mecp2 null mice and age-matched wildtype littermates. However, Mecp2 null mice had a lower density of docked synaptic vesicles at the active zone of both asymmetric spine synapses and symmetric shaft synapses, but had comparable densities of the total pool of synaptic vesicles within presynaptic terminals. Multiphoton imaging of synaptic vesicle recycling within glutamatergic (distal) and GABAergic (proximal) presynaptic terminals labeled with FM1-43 revealed that only the sucrose-sensitive readily releasable pool (RRP) of vesicles had slower rates of activity-dependent destaining in Mecp2 null slices than in wildtype slices. On the other hand, the rate of destaining from the total recycling pool (TRP) of vesicles labeled with hyperkalemic solutions was comparable between symptomatic Mecp2 null and wildtype slices. Finally, the state of GABAergic inhibition in CA1 was evaluated in slices exposed to the GABAA receptor antagonist bicuculline after surgical isolation of CA1 from CA3 (minislices). Surprisingly, VSD signals in CA1 mini-slices from symptomatic Mecp2 null had comparable amplitude, duration and spatial spread to those evoked in wildtype slices. In addition, bicuculline enhanced the duration and spatial spread of VSD signals to a comparable extent in Mecp2 null and wildtype mini-slices , indicating that Mecp2 deletion does not affect local GABAergic inhibition in area CA1. In conclusion, quantitative electron microscopy and multiphoton imaging of synaptic vesicle recycling demonstrated that synaptic structure and function in area CA1 could not account for the hyperexcitability revealed by voltage-sensitive dye imagingin slices from symptomatic Mecp2 null mice. Since the surgical isolation of CA1 from CA3 abolished the differences in VSD signals between Mecp2 null and wildtype slices, and increased them in CA1 of wildtype slices to reach levels of Mecp2 null slices, we reasoned that an excitation/inhibition imbalance must be occurring upstream of CA1 to account for the hyperexcitability observed in CA1 are of mutant mice. Our electrophysiological observations in area CA3 confirm this model (see poster by Amaral et al.).