Redefining the Function of the Rett Syndrome Protein
Just before the holidays I had an opportunity to discuss with Adrian Bird the new data reported in his latest paper, published today in Molecular Cell. Most readers of this blog will know that Prof. Bird discovered the MeCP2 protein in the early 1990s while working at the Research Institute for Molecular Pathology in Vienna. Almost a decade later, Huda Zoghbi’s finding that mutations in MeCP2 cause Rett Syndrome propelled Prof. Bird into the realm of neuroscience. He found himself working, for the first time, on scientific issues with great relevance to human disease. In 2007 he published the dramatic reversal experiments.
We’ve come to expect novel and significant insights from the Bird lab; this new paper redefines our concept of both the scope and function of MeCP2. In the words of co-author Peter Skene, it may be “the watchdog of the neuronal genome.”
[SPANISH] Redefiniendo la Función de la Proteína del Síndrome de Rett
[ITALIAN] Ridefinire le Funzioni della Proteina Chiave della Sindrome di Rett
[GERMAN SUMMARY] MeCP2 – Neudefinition der Funktion des Rett-Syndrom Proteins
MeCP2 Goes Global
Monica Coenraads: I found the data in your latest paper regarding the high levels and broad distribution of MeCP2 to be quite striking.
Adrian Bird: Yes, MeCP2 is exceptionally abundant. Most transcription factors, proteins that turn genes on or off, exist in 10,000 to at most 100,000 molecules per cell. We are seeing 100 to 1,000 times more than that of MeCP2. In fact, there is almost as much MeCP2 in the nucleus as there are nucleosomes, which are the fundamental repeating structural units of chromatin. That means that there is enough MeCP2 to potentially cover nearly all of the genome.
MC: I was intrigued by the fact that MeCP2 binds to non-genes as well as genes.
AB: As far as MeCP2 is concerned it doesn’t seem to care whether it binds to genes or not. It simply binds everywhere there are methyl groups.
MC: So MeCP2 follows methylation across the genome.
AB: Indeed, and this tracking of DNA methylation could explain the reversibility of severe Rett symptoms that we see in mice. The important developmental step is to establish the correct pattern of methylation, and that appears to happen normally in Rett patients. Once you have that pattern set down, and you put MeCP2 back in, as we did in our reversal experiment, the protein simply goes where it’s told by methylation and resumes its function.
The Genome – It’s Not All About Genes
MC: This is probably a good time to remind our readers that only 5% of the genome is made up of genes. The rest comprises what is still sometimes referred to as “junk DNA” because scientists have not been able to ascribe any function to it. I’ve always found the term “junk DNA” a bit arrogant – I doubt that 95% of our genome is junk and in fact recent work has suggested that the junk might in fact have important regulatory functions.
AB: You are absolutely right; we shouldn’t dismiss any of the genome as junk. Much of this so- called “junk DNA” has actually been conserved over many millions of years and that fact alone suggests that there is a good reason for that “junk” to be there.
MC: In recent years the idea that MeCP2 binds to methylated DNA has been questioned a bit. This paper reaffirms and expands on that. Where is this leading us?
AB: I think this confirmation, combined with an abundance of MeCP2 sufficient to cover all the methyl groups in the genome, is telling us something about the function of MeCP2.
MC: So can we still say that Rett symptoms are caused by faulty repression of downstream genes by MeCP2?
AB: That remains a hypothesis that needs proving. We are still waiting for evidence that particular genes, when misexpressed due to mutated MeCP2, are causing Rett. We have a lot of work yet to do to figure out the connection between the absence of repression by MeCP2 and the symptoms of Rett.
MC: So what about the papers that claim particular genes are targets of MeCP2?
AB: Indeed, there have been quite a lot of papers – some written by our lab- which say that certain genes appear to be changed when MeCP2 is missing. The finding is followed up with biochemistry experiments which show that MeCP2 binds to these genes, so the data seems to make sense. However, once you find that MeCP2 binds absolutely everywhere, the concept of target genes becomes a bit less interesting and perhaps less relevant.
MC: If MeCP2 is not a transcription factor, as previously thought, what would you call it?
AB: I would call it an alternative linker histone 1. Ages ago we showed that MeCP2 and the linker histone, HI, compete with each other to assemble chromatin on methylated DNA. In this paper we show that when MeCP2 is absent, the amounts of HI, which are normally very low in the brain, go up dramatically. In that sense MeCP2 clearly resembles a histone.
MC: Let’s give a bit of background for our readers. Histones are proteins which act as spools around which DNA is wound. This winding, or compaction, allows the 1.8 meters of DNA material to fit inside each of our cells. There are two classes of histones – core histones and linker histones. Core histones form the spool around which DNA winds – resembling beads on a string. And linker histones are the DNA separating the beads. HI is one of two linker histones. So, in effect, linker histone is the string between the beads of a necklace.
Might It Be Simpler?
MC: Yet another observation of your paper is that MeCP2 is likely performing the same function throughout the brain. Please elaborate.
AB: Some think that MeCP2 does different things in different neurons. Our data suggests that the pattern of MeCP2 binding is similar regardless of the brain region. My emphasis has turned to the idea that, in the absence of MeCP2, there is a generic problem with neurons and that the regional effects have something to do with what those neurons do in the brain and not so much that MeCP2 does different things in different places. In other words, MeCP2 does the same thing everywhere but its consequences are different.
Currently there is a lot of data from many labs pulling us in multiple directions. I would like to see if we can slice through all that complexity and say, in all these neurons this is what is wrong. I’m excited about the possibility that perhaps it’s not that complicated after all.
MC: That would be an elegant and welcome scenario. Thank you, Prof. Bird, for discussing your latest paper. I look forward to bringing our readers an update soon regarding your work.
Share this article: