By Tom Ulrich
While the genome’s As, Ts, Cs, and Gs hold the instructions for making proteins, how does a cell know when to read a gene? And could it relate to developmental disorders?
These gene-reading instructions are encoded in our epigenome, a set of factors that give our cells exquisite control over when and where to turn individual genes on and off. This control involves a delicate and complex dance between DNA and proteins called histones – DNA wraps itself around histones to create a complex called chromatin – as well as the many different types of epigenetic tags.
Yang Shi, of the Division of Newborn Medicine at Children’s Hospital Boston, wants to understand what happens when the genome doesn’t read the epigenome’s instructions correctly, which in the developing brain can cause intellectual disabilities.
"Epigenetics can be broken down into three steps," says Shi. "First, writing and/or erasing tags on DNA and histones by specific enzymes, marking when to use a gene. Second, reading the tags, using proteins containing specialized recognition modules. Third, recruitment of additional enzymes and proteins that cause the cell to start or stop some activity."
The list of proteins in that second step – the tag readers – is huge. Shi explains why, using histone tagging as an example: "There are four kinds of histones found in nature that, together with DNA, form a nucleosome, the basic unit of chromatin. Each histone can be tagged at multiple different points with multiple different tags at the same time. And the tags at each location on each histone each have their own readers.
"If you calculate the nearly astronomical number of possible theoretical combinations of histone states," Shi continues, "you come up with a very large number of potential readers."
Any errors in a system that complex are bound to have dramatic effects. For boys, errors that involve the X chromosome can wreak particular havoc. Every cell in a girl has two copies of the X chromosome, so if a gene on one copy gets mutated, the cell can compensate as long as the other copy is undamaged. Boys have only a single copy of the X (paired up with a Y chromosome), so their X chromosome genes have no backup should something go wrong. (In girls, damage to both copies of an X chromosome gene puts them in similar straits.)
Shi’s laboratory is particularly interested in the relationships between epigenetic gene errors on the X chromosome and neurodevelopmental, cognitive and craniofacial disorders. "We currently know of 82 X-chromosome genes that are mutated in intellectual disability disorders," says Shigeki Iwase, a postdoctoral fellow in Shi’s laboratory. "Of these, 19 have features suggesting that they help edit and read epigenetic tags. But for the most part we don’t yet know which tags they read."
For one of these 19 genes – the ATRX (alpha-thalassemia/mental retardation, X-linked) gene – Shi and Iwase may have found the answer. Mutations in ATRX result in a disorder (namely, ATRX syndrome) exclusively seen in boys; there are only about 150 recognized cases worldwide. The mutations that cause the disorder tend to cluster 50/50 in two sections of the ATRX gene.
The pair of researchers, along with collaborators from Tsinghua University and Memorial Sloan Kettering Cancer Center, recently announced in Nature Structural and Molecular Biology that one of these sections, called the ADD domain, is actually a hitherto unknown epigenetic reader.
"We knew that mutations in the ATRX gene’s other domain prevent it from depositing histones where needed, which keeps DNA from wrapping up when it should," says Iwase. "But no one previously understood what mutations in the ADD region did. Now we know that this section of ATRX gives it a particular dual-tag reading capability that’s not been seen before."
Numbers game
One of the challenges in trying to understand relationship between epigenetics and developmental disorders, especially X-linked ones, is the small numbers of patients. "Intellectual disabilities are found in only about 2 or 3 percent of the population," Iwase says. "Each syndrome only represents a small fraction of the cases in that percentage. And each of the 82 X-linked genes represent a fraction of that fraction."
Shi also notes that while linking disorders and readers singly is both useful and necessary, it doesn’t reflect how different epigenetic mechanisms work in concert within the cell. "We’re still at the early stage of cataloging which reader works with which epigenetic modification, one at a time," he says. "But eventually we want to understand how constellations of modifications get read, which is much closer to what happens in vivo."
"We want to understand the forest," he concludes, "but we have to start with the trees."
