There are also far more enhancers than we might expect. A recent comprehensive study looked at the patterns of histone modifications in nearly 150 human cell lines. When they assessed these lines for patterns that looked like enhancers, they found nearly 400,000 candidate enhancer regions.{235} This is far more than required if there was a one-to-one relationship between enhancers and protein-coding genes. It’s even too many if we assume that long non-coding RNAs have enhancers.
The enhancers weren’t all found in every cell type. This is consistent with a model where the same stretch of DNA can have different functions in different cells, depending on how it is epi-genetically modified.
For many years, we had no clear models of how enhancers really work. We now suspect that in many cases they may be critically dependent on another type of junk: the long non-coding RNAs. In fact, specific classes of long non-coding RNAs may be expressed from the enhancers themselves.{236} Many of the long non-coding RNAs we met in Chapter 8 are involved in repressing expression of other genes. But it is now believed that there is also a large class of long non-coding RNAs that enhance gene expression. This was first suggested to be the case for long non-coding RNAs that regulate neighbouring genes. If expression of the long non-coding RNA was increased experimentally, the expression of the neighbouring protein-coding gene also increased. Conversely, if the expression of the long non-coding RNA was knocked down experimentally, the protein-coding gene also showed lower expression.{237}
Further evidence came from analysing the timing patterns for specific long non-coding RNAs and the messenger RNAs they were believed to regulate. Researchers treated cells with a stimulus that they knew caused expression of a specific gene. They found that the enhancing long non-coding RNA was switched on before the messenger RNA from the neighbouring protein-coding gene.{238},{239} This is consistent with a model where the long non-coding RNA located in the enhancer is switched on in response to a stimulus, and then in turn helps to switch on expression of the protein-coding gene.
The long non-coding RNA doesn’t drive this increase on its own. The process is reliant on the presence of a large complex of proteins. The complex is known as Mediator. The long non-coding RNA binds to the Mediator complex, directing its activity to the neighbouring gene. One of the proteins in the Mediator complex is able to deposit epigenetic modifications on the adjacent protein-coding gene.[32] This helps to recruit the enzyme that creates the messenger RNA copies which are used as the templates for protein production. There is a consistent relationship between the Mediator complex and the long non-coding RNA. Experimentally generated decreases in expression of either the long non-coding RNA or a member of the complex each lead to decreased expression of the neighbouring gene.{240}
The importance of a physical interaction between long non-coding RNAs and the Mediator complex has been shown by a human genetic condition. This disorder is called Opitz-Kaveggia syndrome. Children born with this condition have learning disabilities, poor muscle tone and disproportionally large heads.{241} The affected children have inherited a mutation in a single gene. This codes for the protein in the Mediator complex that interacts with long non-coding RNA molecules.[33]
The more that scientists analysed the activity of the Mediator complex, the more interested they became. One of the reasons was that the Mediator complex is responsible for the actions of a group of enhancers with special powers. These are the super-enhancers, and they are particularly important in embryonic stem (ES) cells, the pluripotent cells that have the potential to become any cell type in the human body (see page 105).{242}
The super-enhancers are clusters of enhancers all acting together. They are about ten times the size of normal enhancers. Because of this, proteins can bind to the super-enhancers at very high levels, much higher than are found on normal enhancers. This allows the super-enhancers to really ramp up expression of the gene they are regulating. But it’s not just the numbers of proteins that bind that interested the researchers. It’s the identities of these proteins.
As we saw in Chapter 8, ES cells don’t stay pluripotent by chance or passively. In order for ES cells to maintain their potential, they regulate their genes very carefully. Even relatively mild perturbations in gene expression can start to push an ES cell down a pathway that converts it into a specialised cell type. One way of visualising this is to think of a Slinky at the top of a tall flight of stairs. Just the slightest nudge to push it over the edge of the top step is enough to send that Slinky on a very long journey. Perhaps an even better analogy might be a Slinky that is held back from falling down the stairs by a small weight on its trailing end. Remove the weight, and off the Slinky will go.
There are a set of proteins that are absolutely vital for maintaining the pluripotency of ES cells. These are known as master regulators, and they are like the small weight on the trailing end of the Slinky. The master regulators are expressed very highly in ES cells, but at much, much lower levels in specialised cells.
The importance of these proteins was unequivocally demonstrated in 2006. Researchers in Japan expressed a combination of four of these master regulators at very high levels in differentiated cells. Astonishingly, this set in motion a chain of molecular events which culminated in the creation of cells that were almost identical in action to ES cells.{243} This is analogous to a Slinky at the bottom of a flight of stairs moving all the way back up to the top step. The cells created by this route have the potential to be converted into any cell type in the body.[34] This remarkable work, and the research that followed on from it, has generated enormous excitement because potentially we can create replacement cells to treat a large number of disorders. These range from blindness to type 1 diabetes, and from Parkinson’s disease to cardiac failure.
Until this new technology was developed, it was extremely difficult to create appropriate cells to treat human conditions. This is because cells from a different individual usually can’t be implanted into another person. The immune system will recognise the donated cells as foreign and kill them, as if they were an invading organism. But, as shown in Figure 12.1, we now have the potential to make cells that are a perfect match for the patient.
The 2006 work has spawned an industry potentially worth billions of dollars, and also resulted in one of the fastest awards of a Nobel Prize in Medicine or Physiology ever, just six years after the original publication.{244}
In normal ES cells, some of these master regulator proteins bind at very high densities to the super-enhancers. The super-enhancers themselves are regulating some key genes that maintain the pluripotent state of the cells. The Mediator complex is also present at very high levels in the same locations. Knocking down the expression of a master regulator, or of Mediator, has very similar effects on the expression of these key genes. The expression levels drop, making the ES cells more likely to start differentiating into specialised cell types.
Figure 12.1 The theory behind using patient-derived cells to create therapies tailored for a specific individual.
32
The modification is addition of a phosphate group (one phosphorus atom and four oxygen atoms) to a specific position on histone H3. This modification is usually associated with active genes.