Disease modelling of a human neurodevelopmental disorder using mouse embryonic stem cells

As cultured embryonic stem (ES) cells can be differentiated in neurons under well-defined conditions, they provide a unique opportunity to model and study diseases of the nervous system such as Rett syndrome. Most cases of this tragic neurodevelopmental disorder affecting about 1 young girl in 10’000 are caused by mutations in MECP2, a gene encoding the methyl-CpG-binding protein 2. Despite the generation of very useful mouse models recapitulating some important features of Rett syndrome, like the lack of normal brain growth, breathing arrhythmia and movement deficits, the function of MeCP2 in the brain remains unclear at this point. A few years ago, our laboratory established a differentiation procedure allowing the generation of virtually pure progenitors defined as Pax6-positive radial glial cells. As they do in vivo, these cells go on to generate glutamatergic neurons. One key characteristic of this system is that essentially all cells differentiate synchronously and I used it during the course of my PhD thesis to compare wild-type ES cells with ES cells lacking MeCP2. During the transition from progenitors to neurons, the size of nuclei increases by about 40% within a period of about 10 days, while nuclei of neurons lacking MeCP2 fail to grow at the same rate. Both the acute re-expression of MeCP2 in MeCP2 Stop-Floxed neurons following the excision of a stop cassette using Cre, as well as the viral delivery of MeCP2 to null neurons were found to rescue the small nuclear size phenotype. In an effort to correlate the size of the nuclei with their transcriptional activity, I compared the rate of total in vitro RNA transcription in pure nuclei isolated from wild-type vs. Mecp2 -/y (null) and Mecp2 Stop-Floxed ES cell-derived neurons. Regardless of the type of MeCP2 mutation, the neuronal nuclei lacking MeCP2 were found to be significantly less transcriptionally active compared to the wild-type nuclei. These results suggest that MeCP2 may have a general, genome-wide role in regulating the rate of RNA transcription, including ribosomal RNA that represents the major species investigated in such experiments. As BDNF levels have been reported to be reduced in the brain of Mecp2 -/y mice, I also examined if ES cell-derived neuron lacking MeCP2 would reflect this potentially important aspect of the disease, since decreased levels of BDNF could be hoped to be corrected and to improve the patients’ condition. At about 15 days in vitro, the levels of BDNF were found to be reduced by about 30% and the acute expression of MeCP2 was found to restore the expression of BDNF to levels similar to those seen in wild-type neurons.
In the second part of my work, I found that the sphingosine-1 phosphate analogue fingolimod (FTY720), recently introduced as the first oral treatment of multiple sclerosis, increases BDNF levels both in wild-type and Mecp2 -/y ES cell-derived neurons by mechanisms involving MAPK signaling and neuronal activity. In MeCP2 mutant mice, FTY720 increases BDNF levels in the striatum as well its volume, ameliorates locomotor activity and extends the lifespan by 50%.
As microcephaly is one of major symptoms of Rett syndrome, I also investigated in the third part of my PhD work the effects of the lack of MeCP2 on cell proliferation. While MeCP2-deficient ES cells do not have a proliferation defect, neuronal progenitors lacking MeCP2 generate fewer neurons compared with wild-type cells. This difference between ES cells and neuronal progenitors may result from the fact that the levels of MeCP2 expression are much higher in neurons and neuronal progenitors compared to ES cells.
Taken together, my results show that an ES cell based system represent a useful tool to understand molecular mechanisms underlying human neurological disorders.