Even if we just think about the protein-coding genes, we know that thousands are typically switched on in a given human cell type. These genes are spread out across our 46 chromosomes, so we might expect that if we analysed a cell to visualise the geographical locations of the genes that are switched on we would see thousands of tiny dots spread throughout the nucleus. Instead, as shown schematically in Figure 12.4, there are about 300 to 500 larger spots.{249} Gene expression in our cells isn’t a cottage industry. Instead it takes place in discrete locations in the nucleus known as factories.{250}
Figure 12.4 The dots represent the positions of protein-coding genes in the nucleus. If genes were positioned in the nucleus solely as a function of their position on chromosomes, we would see a diffuse pattern such as the one on the left. Instead, genes cluster together in three-dimensional space, creating a punctate pattern of gene localisation represented by the situation on the right.
Each factory contains between four and 30 copies of the enzyme that makes a messenger RNA molecule from the DNA template, plus a large number of other molecules required to do the work.{251},{252} The enzymes stay in one place and the relevant gene is reeled through to be copied.{253} In order for the gene to reach the factory, the DNA has to loop out to reach the right part of the cell nucleus. But the really ingenious bit is that more than one gene can be copied into messenger RNA at a time in the same factory. The combination of genes found in a single factory isn’t random. The genes tend to be ones that code for proteins that are used for related functions in the cell. This is equivalent to having multiple parallel assembly lines in one physical factory. Once all the lines have completed their individual tasks, the factory can assemble the final product from the components. One factory produces boats, another builds food mixers. In our cells, the factories ensure that genes are expressed in a coordinated fashion. This means lots of loops unfurling from chromosomes and localising to the same regions of the nucleus simultaneously.
One example of this is a factory for the genes that code for the proteins required to create the complex haemoglobin molecule, which carries oxygen around in the blood.{254} Another factory is used to generate the proteins required in order to mount a strong immune response.{255} One important component of an effective immune response is the production of proteins called antibodies. Antibodies circulate in the blood and other body fluids, binding to any foreign matter that they detect. Scientists activated the cells that produce antibodies and then studied how certain key genes looped out. The genes they analysed were the ones required to create antibody molecules. They found that these key genes moved to the same factory as each other. Remarkably, some of these genes were completely physically separate from each other normally, as they are located on different chromosomes.
Although this is a remarkable way of coordinating gene expression, it may also carry risks. Burkitt’s lymphoma is the aggressive cancer we met earlier in this chapter. The cell type that becomes abnormal in this disease is the cell type that produces antibodies. In this condition, a strong promoter from one chromosome gets abnormally positioned next to a gene from another chromosome. Until recently we didn’t understand why these regions were susceptible to joining up, because we thought of them as being physically distant from each other, as they are on different chromosomes. But now we know that the regions that ‘swap’ to create the dangerous abnormal hybrid chromosome are both regions that move to the factory described in the previous paragraph. This might be how the two different chromosomes get close enough together to swap their material, perhaps if both break simultaneously and are wrongly repaired when in the factory.
While it might seem that evolution would have selected against this dangerous situation, we need to remember yet again that natural selection is about compromise, not perfection. The advantages of producing antibodies to fight off infections and thereby keep us alive long enough to reproduce clearly outweigh the potential disadvantages of an increased cancer risk.
13. No Man’s Land
When we think about the First World War, the prevailing image many of us have is probably of men in trenches. Opposing armies dug into the muddy landscapes for the ultimate exposition of war as months of boredom, punctuated by moments of acute terror.{256} The trenches occupied by the armies were separated by stretches of terrain that were not under the control of either combatant. These stretches were ‘No Man’s Land’, and could be as narrow as a couple of hundred metres, or over a kilometre wide. At night, the soldiers would creep out of the trenches for reconnaissance, to lay barbed wire and to retrieve injured or dead compatriots.
The human genome contains multiple regions of No Man’s Land, keeping different elements apart from each other. Just like the quagmires of the First World War, these genomic barriers vary in size and are fairly fluid, depending on where they lie in relation to their troop movements. And just like the No Man’s Land of Europe in those awful few years of slaughter, these regions are anything but devoid of activity. The No Man’s Land of the human genome binds proteins, garners epigenetic modifications and regulates the interactions of different genetic elements in a highly active way.
This is important to our cells, because most of our genes are all over the place.[36]{257} By this we mean that genes are scattered around on our 23 pairs of chromosomes in a fairly nonsensical way. As we have already seen, the genes that code for the proteins required to make haemoglobin are brought together by changes in the three-dimensional arrangements of chromosomes. This compensates for the fact that they aren’t arranged next to each other in a nice, neat way. If we look at how most of our genes are distributed, they are like the donations to a jumble sale or charity shop before they’ve been organised sensibly.
This can mean that our cells contain a gene that codes for a protein required in the foetal liver next to a gene for a protein expressed in the adult skin. There’s a huge number of such situations and this creates potential difficulties. It means that our cells require barriers between different elements to maintain different patterns of gene expression. The control needs to be relevant to a specific cell type, and to the particular developmental stage. We don’t want teeth genes expressed in our eyes or heart genes expressed in our bladders.
We know that epigenetic modifications influence gene expression. If we take the brain as an example, there are some genes that are never expressed in neuronal cells. For instance, the protein keratin is used in hair and nails, but isn’t used by our adult grey matter. In brain cells, the keratin gene is switched off and it’s kept in an inactive state by a particular pattern of epigenetic modifications. But as we’ve already seen, epigenetic modifications are blind to DNA sequence. What’s to stop these repressive modifications from creeping along from the keratin gene and starting to switch off other genes as well?
36
There are a couple of exceptions to this where genes are clustered in a way that reflects their expression patterns. The main ones are the HOX genes that control body patterning and the Ig genes that code for antibodies.