Abstract
Network Hyperexcitability in Hippocampal Slices From Mecp2 Mutant Mice Revealed by Voltage-Sensitive Dye Imaging.
Calfa G, Hablitz JJ, Pozzo-Miller L. J Neurophysiol 2011;105:1768–1784.
Dysfunctions of neuronal and network excitability have emerged as common features in disorders associated with intellectual disabilities, autism, and seizure activity, all common clinical manifestations of Rett syndrome (RTT), a neurodevelopmental disorder caused by loss-of function mutations in the transcriptional regulator methyl-CpG-binding protein 2 (MeCP2). Here, we evaluated the consequences of Mecp2 mutation on hippocampal network excitability, as well as synapse structure and function using a combination of imaging and electrophysiological approaches in acute slices. Imaging the amplitude and spatiotemporal spread of neuronal depolarizations with voltage-sensitive dyes (VSD) revealed that the CA1 and CA3 regions of hippocampal slices from symptomatic male Mecp2 mutant mice are highly hyperexcitable. However, only the density of docked synaptic vesicles and the rate of release from the readily releasable pool are impaired in Mecp2 mutant mice, while synapse density and morphology are unaffected. The differences in network excitability were not observed in surgically isolated CA1 minislices, and blockade of GABAergic inhibition enhanced VSD signals to the same extent in Mecp2 mutant and wild-type mice, suggesting that network excitability originates in area CA3. Indeed, extracellular multiunit recordings revealed a higher level of spontaneous firing of CA3 pyramidal neurons in slices from symptomatic Mecp2 mutant mice. The neuromodulator adenosine reduced the amplitude and spatiotemporal spread of VSD signals evoked in CA1 of Mecp2 mutant slices to wild-type levels, suggesting its potential use as an anticonvulsant in RTT individuals. The present results suggest that hyperactive CA3 pyramidal neurons contribute to hippocampal dysfunction and possibly to limbic seizures observed in Mecp2 mutant mice and RTT individuals.
Commentary
Rett syndrome (RTT) is an X-chromosome-linked autism spectrum disorder that almost exclusively affects females. Although the initial postnatal development appears to be normal, progressive deterioration of normal development begins at 6 to 18 months of age. Symptoms of RTT include severe problems with language and communication, learning, motor coordination, and seizures (recurrent partial and generalized), which develop in 80 percent of patients (1, 2). The genetic defect responsible for the development of the majority of RTT cases has been linked to loss of function mutations of an X-chromosomal gene (located on Xq28) encoding the DNA-binding protein, methyl-CpG-binding protein 2 (MeCP2) (3). MeCP2 binds to methylated CpG islands within promoter regions of the DNA and thereby plays an important role in the regulation of gene transcription. Thus, a loss of function of the upstream regulator MeCP2 is expected to affect the expression of a variety of genes, which play crucial roles in brain development and network excitability, simultaneously. RTT has successfully been replicated in mutant mouse models that either lack MeCP2 completely or that express a nonfunctional truncated version thereof (4, 5). Global re-expression of MeCP2 within a Mecp2-null background reverses neurologic deficits (6).
Scientifically, RTT is a fascinating condition since it combines comorbid behavioral alterations (e.g., communication and learning deficits) with seizures that can most likely be attributed to complex changes in network homeostasis within the hippocampal formation. This is important because seemingly unrelated comorbid conditions might be linked to the excitability state of the hippocampal network. To study this possibility, it is imperative to investigate hippocampal excitability on the network level. In an elegant study by Calfa and colleagues, acute hippocampal slice preparations from hemizygous male MeCP2-deficient mutant mice (knockouts) were used to study hippocampal network excitability employing a combination of imaging and electrophysiologic tools. Key for the assessment of hippocampal network activity was the use of a hexagonal 464-element photodiode array to monitor the spatiotemporal excitation of a voltage-sensitive dye (VSD) following electrical stimulation of Schaffer collaterals (to image membrane depolarization in area CA1) or of mossy fibers (to image membrane depolarization in area CA3). Vesicular release from presynaptic terminals was visualized by multiphoton imaging, and the ultrastructure of synapses was investigated by electron microscopy.
Using VSD imaging (7), the authors first demonstrate pronounced hyperexcitability in area CA1 of acute hippocampal slices from symptomatic Mecp2 mutant animals, reflected by a spatiotemporal spread of the fluorescent signal. These effects were specific for slices taken from symptomatic animals and were not found in slices taken from presymptomatic animals. Quantitative ultrastructural analysis of synapses in area CA1 of symptomatic Mecp2 mutants revealed a deficit in synaptic vesicles docked at their active zone. Presynaptic release properties were further investigated by multiphoton excitation microscopy of activity-dependent destaining of a fluorescent dye. Data from those experiments revealed a selective deficit in the activity-dependent release from the readily releasable pool of synaptic vesicles. These morphologic and kinetic data on synaptic release properties are in apparent contrast to the increased CA1 hyperexcitability. The results suggest that differences in synapse structure may not be a major contributor to the profound differences in VSD signals from the Mecp2 mutant animals and that the origin of the CA1 hyperexcitability might be located elsewhere (i.e., upstream in the hippocampal circuitry). To address this question, area CA1 was surgically separated from CA3. This procedure prevented hyperexcitability in CA1 under conditions where GABAergic inhibition was maintained, indicating that input from area CA3 was needed to trigger the hyperexcitability of area CA1. As expected, the frequency of spontaneous multiunit activity in area CA3 was higher in slices derived from Mecp2 mutant mice compared with wild-type littermates. Likewise, the spatiotemoral spread of neuronal depolarizations evoked in area CA3 by mossy fiber stimulation was increased in slices from the mutants. Further, in intact slices the stimulation of afferent mossy fibers resulted in larger neuronal depolarizations in area CA1 of slices from the mutant animals. These results demonstrate hyperexcitability of the entire hippocampal network and resulting increased seizure susceptibility in Mecp2 knockout mice. This condition appears to originate from a high level of spontaneous activity in the pyramidal cell layer of area CA3. Finally, Calfa and colleagues demonstrate that hippocampal excitability in slices of Mecp2 mutant mice could be reduced to levels comparable with those seen in wild-type slices after the application of 50 μM of the brain's endogenous neuromodulator and anticonvulsant adenosine (8), providing support for its potential therapeutic use as an anticonvulsant in RTT patients.
The study from Calfa and colleagues sheds important new light on the synaptic bases of network excitability in the hippocampus of modeled RTT. Importantly, the hyperexcitability of the hippocampal network of Mecp2 mutant mice appears to reside in the hyperactive CA3 region. These findings are consistent with the characteristic hippocampal dysfunction and the seizure phenotype in RTT patients and might provide an explanation for comorbid behavioral and seizure-related phenotypes. A limitation of Calfa's study is the focus on neuronal networks in an in vitro preparation. Network dysfunction in situ in RTT might be more complex. Recent findings demonstrate that the loss of MeCP2 affects not only neurons but also glial cells of RTT brains, and functional studies indicate that astrocytes with MeCP2 mutations have non-cell-autonomous effects on neuronal properties (9). Intriguingly, therapeutic re-expression of Mecp2 preferentially in astrocytes within an Mecp2-deficient background significantly improved behavioral abnormalities associated with RTT such as aberrant respiration, hypoactivity, and decreased dendritic complexity, as well as greatly prolonged the life span of Mecp2 knockout mice (10). These studies demonstrate that functional glia, by providing an optimal homeostatic environment for neurons to function properly, might play an important role in RTT. Unfortunately, the seizure phenotype was not investigated in the latter study. It would be interesting to know whether glial expression of MeCP2 might also affect hippocampal network excitability and seizure susceptibility within a Mecp2-null background.
As the work from Calfa and colleagues shows, RTT presents as a condition of complex network dysfunction, resulting in a sophisticated comorbid phenotype. Therefore, it is highly unlikely that RTT can be treated with conventional drugs that usually have highly specific (neuronal) downstream targets. Rather, reconstruction of homeostatic network functions within the brain might be needed in order to find a treatment for RTT. The condition of RTT is a prime example of how an upstream regulator (MeCP2) leads to complex alteration of network function. Mecp2 mutants thereby provide excellent model systems to study novel homoeostatic therapies, which might be based on modulating glial function or on using neuromodulation (e.g., by adenosine as tested by the authors) as a novel therapeutic modality to reconstruct network excitability.