Project Summary
In this revised proposal, we continue to develop an approach to repair base mutations at the level of
RNA, for attenuating symptoms in mouse models of human neurological disease. The experiments
are an outgrowth of a pilot NIH Director’s Transformative Research Award that supported both
recently published and preliminary results attesting to feasibility of this approach. Our work is
currently focused on Rett syndrome, a devastating neurological disease due to mutations in the gene
encoding the transcription factor, MECP2. We focused initially on a human patient guanosine (G) to
adenosine (A) mutation in MECP2, MECP2317G>A, which interferes severely with its ability to bind to
chromatin and results in Rett syndrome. We showed, for the first time, that endogenous Mecp2317G>A
RNA can be recoded to the wild type amino acid efficiently in non-dividing neurons cultured from a
Mecp2317G>A mouse line that exhibits severe Rett-like symptoms. The recoding occurred through site-
directed deamination by a hijacked catalytic domain of Adenosine Acting on RNA 2 (ADAR2)
(Editase), fused to a bacteriophage RNA binding peptide, which we targeted to the Mecp2 mutation
by an RNA guide. In this revised application, we present new data indicating that recoding also
occurs in vivo, in 3 different hippocampal neuronal populations, after direct hippocampal injection of
AAV encoding the hybrid ADAR2 protein and Mecp2 RNA guides. Moreover, recoding resulted in
amount of MeCP2 localization to chromatin consistent with amount of editing at the RNA level. We
have developed the tools and reagents that now place us in an ideal position to address critical
unanswered questions for reversing neurological phenotypes of Rett syndrome, and for testing
hypotheses related to site-directed repair. In Aim 1, we test the hypotheses that brain-wide repair of
Mecp2317G>A RNA, by site-directed editing, can be tuned to high efficiency and specificity in mice, and
restores proper chromatin interaction. For this purpose, we perform whole transcriptomic RNA seq
analysis across the brain after peripheral injections of an efficient brain AAV serotype virus encoding
optimized editing or control components. In Aim 2, we test new guides for the ability to recruit
endogenous ADAR2 to mutant Mecp2 RNA in vivo, circumventing potential immune responses to the
bacteriophage moiety in the hybrid ADAR2 protein and potentially minimizing off-target editing. Our
initial model is Mecp2311G>A that has the ideal nonsense codon for this approach. In Aim 3, we inject
peripherally the virus encoding our current and optimized editing components, or controls, to test our
hypothesis that site-directed RNA editing can stabilize/reverse Rett-like symptoms in both Mecp2G>A
mouse lines. In addition to Rett syndrome, our approach has the potential to cure thousands of
additional pathogenic G>A mutations.