You are here

Adrian Bird


Beatrice Alexander-Howden, Kashyap Chhatbar, Justyna Cholewa-Waclaw, John Connelly, Dina De Sousa, Jacky Guy, Martha Koerner, Matthew Lyst, Timo Quante, Jim Selfridge, Ruth Shah, Konstantina Skourti-Stathaki, Christine Struthers, Rebekah Tillotson
Bird Lab Homepage

Modelling the variable severity of Rett syndrome mutations

Adrian gives a brief overview of his research.

Rett syndrome is caused by mutations in the X-linked MECP2 gene, which encodes a chromosomal protein that binds to methylated DNA. We earlier generated a mouse model by deletion of the gene for MeCP2 and found that this closely mirrored the human disorder. The availability of a convincing model of Rett allows us to investigation the origin of the phenotype at a molecular level. Although the gene knockout is useful in this respect, there is a need for genetic models that mimick the actual mutations seen in Rett syndrome patients. We are particularly interested in missense mutations, which substitute the wrong amino acid at a single position, but leave the rest of the protein unaffected, as these allow us to determine the importance of specific protein domains for MeCP2 function. To this end we created a Mecp2 “allelic series” representing the three most common missense Rett syndrome (RTT) mutations. Together these three mutations (R133C, T158M and R306C) make up about one quarter of all RTT mutations in humans. Interestingly the mutations vary significantly in average severity in humans and this spectrum is closely mimicked in the mouse models; R133C being least severe, T158M most severe and R306C being intermediate.

We found that R133C and T158M mutations each affect two aspects of MeCP2 function, as they destabilize the protein and they also compromise DNA binding. Destabilisation is particularly apparent in the case of T158M, which presumably accounts for the high clinical impact of this mutation. This work also allowed us to evaluate a provocative hypothesis suggesting that the MeCP2 binds to hydroxymethylated cytosine (generated by oxidation of 5-methylcytosine) in the brain and that the R133C mutant had specifically lost this ability. Our studies showed that MeCP2 does not in fact bind to the major form of hydroxymethylated DNA in the brain and we did not find any evidence that R133C selectively lost binding to this modification. Overall, our allelic series recapitulates human Rett syndrome severity, revealing compound molecular aetiologies and providing a valuable resource in the search for personalized therapeutic interventions against this distressing condition. Our current work is designed to test hypotheses regarding the mechanism of action of MeCP2. We have proposed that the protein acts as a bridge between methylated DNA and a large multi-protein gene-silencing machine. Our goal now is to challenge this model by altering the genes encoding Mecp2 and its interacting proteins. In this way we hope to establish how MeCP2 works in the healthy brain and how mutations lead to brain dysfunction.

Selected publications:

Cook PC, Owen H, Deaton AM, Borger JG, Brown SL, Clouaire T, Jones GR, Jones LH, Lundie RJ, Marley AK, et al. A dominant role for the methyl-CpGbinding protein Mbd2 in controlling Th2 induction by dendritic cells. Nature Communications. 2015;6(6920).

Brown K, Selfridge J, Lagger S, Connelly J, De Sousa D, Kerr A, Webb S, Guy J, Merusi C, Koerner MV, et al. The molecular basis of variable phenotypic severity among common missense mutations causing Rett syndrome. Hum Mol Genet. 2015.

Lyst MJ, and Bird A. Rett syndrome: a complex disorder with simple roots. Nature Reviews Genetics. 2015.

 Modelling the most common Rett syndrome mutations in mice.
A. MeCP2 protein represented as a linear bar, showing the MBD (Methylated DNA Binding Domain), NID (NCoR corepressor-Interaction Domain) and the most common missense mutations that cause Rett syndrome. B. Survival of MeCP2 male mice carrying the R133C (green), R306C (blue) and T158M (red) mutations in comparison with males lacking MeCP2 altogether (grey). Severity of phenotypes recapitulates the human severity spectrum for these same mutations. C. A section of the hippocampal region of the mouse brain showing nuclei stained for DNA (“DAPI”) and green fluorescent protein (EGFP), which tracks MeCP2. Wildtype protein (WT-GFP) localizes to heterochromatic dots in the nuclei as expected, as does R306C as the mutation does not strongly affect DNA binding. Both R133C and T158M are less abundant and poorly localized due to instability and compromised binding to DNA respectively.