A recent study has lent more weight to the view that 'Junk DNA' may be anything but junk. A joint effort by the European Molecular Biology Laboratory (EMBL) in Heidelberg, Germany and Stanford University, California, US, has uncovered large differences between the non-coding DNA of different individuals, which may be associated with differing levels of disease risk and other traits too.
The findings were published in two papers in the journals Science and Nature. A third study, carried out at the European Bioinformatics Institute (EBI) in Cambridge, UK, drew similar conclusions, and was published in Science.
Only two per cent of the human genome is made up of protein-coding genes that shape attributes like our hair or eye colour. These 23,000 genes vary by only about 0.025 per cent across all humans, and decades of research have been spent trying to tease out what effects these minute differences have on us.
Until recently, scientists could find little explanation or function for the remaining 98 per cent of our DNA that didn't produce proteins, hence the unflattering name 'junk DNA'. But, more recently, researchers have shown that this non-coding DNA, which can vary by as much as one to four per cent between individuals, plays a crucial role in determining which genes are active and how much of a particular protein those genes produce. This has led many to theorise that the key to our individuality may not be our genes, but in these 'junk' sequences that surround and control the genes.
'We largely have the same sets of genes. It's just how they're regulated that makes them different', says Professor Michael Snyder, a geneticist from Stanford University. Snyder's research involved transcription factors - proteins that attach to areas of non-coding DNA and dictate how neighbouring genes make proteins. They focused on two transcription factors in particular, NF-kappa-B (a protein involved in immune responses) and Pol-II (a protein that helps convert DNA to RNA) and noticed that, in different individuals, these transcription factors act in very different locations on the genome.
Snyder's team began by identifying some 15,000 genome-wide protein binding sites for NF-kappa-B, and some 19,000 for Pol-II, in a chimpanzee and 10 humans (five of European descent, three of west African descent, and two of east Asian descent). They then investigated if there was variation between the 11 individuals or if the proteins bound equally strongly.
They found that some 7.5 per cent of NF-kappa-B sites, and a massive 25 per cent of Pol-II sites, exhibited significant binding differences between the humans, in some cases by as much as two orders of magnitude. Interestingly, there was only a 32 per cent difference in Pol-II binding regions between the chimp and humans, not that much larger than the 25 per cent found between two humans. Perhaps most significantly, the researchers found that many of the variable binding regions were near genes involved in diseases including type-1 diabetes, lupus, leukaemia and schizophrenia.
The EBI team performed a similar analysis in their study except they compared cells from two families consisting of a mother, father and two children. They examined two processes involved in the regulation of genes: the transcription factor binding, and whether Chromatin (a complex combination of DNA and proteins that makes up chromosomes) was 'open' (essentially whether genes were 'allowed' to be transcribed). They found that Chromatin sites that tended to be open in the parents, tended to be open in the children too, suggesting that Chromatin structure is heritable.
If a single nucleotide polymorphism (SNP) in non-coding DNA produces an inheritable difference in transcription factor binding, or Chromatin structure, the consequences could be very exciting. It may offer a potential mechanism for transmission of characteristics such as disease susceptibility between generations.
The hope of these scientists and others like them is that, if they can identify the regulatory changes that underlie different diseases, they can treat them in the future with drugs that make a particular gene more or less active. Jan Korbel, the team leader from EMBL, is optimistic: 'The picture that emerges is very complex, but in order to start treating diseases, you don't have to understand the full picture,' he says.