Final Report Summary - NEWGENES (The role of de novo evolution in the emergence of new genes)
The canonical view of how new genes arise was that they can not come from scratch, i.e. not out of random non-coding sequences. Instead, it was thought that they are only generated by duplicating or re-shuffling of existing genes. This had left the question open how the very first genes would have arisen during evolution. A general paradigm assumes that the first building blocks of genes may have arisen during the origin of life phase, where different chemical conditions may have existed. The idea that de novo evolution of genes out of random sequences could be an alternative mechanism was initially fuelled by some isolated examples, one of them described from our laboratory in the mouse. These discoveries justified a much larger project to search systematically for de novo evolved genes, since the proof of a frequent de novo evolution would change the text book knowledge on the evolution of genes.
Our project focused on three general topics: 1) generation of a large comparative sequence data set to unequivocally trace the emergence of new transcripts and genes from non-coding DNA, 2) an experimental evolutionary analysis of synthetically generated random nucleotide sequences and 3) genetic studies on a subset of identified de novo evolved genes.
In the first project part we found that a much larger fraction of the genome is transcribed in RNA than had been suspected so far. Instead of less than 10% of the genome that is annotated as being transcribed, we found around 50% yielded transcripts, many of which only in a subset of the respective taxa. We could show that new transcripts arise at a very high rate from non-coding parts of the genome, but most get also quickly lost again. Our findings imply that basically the whole genome is "scanned" for being transcribed within an evolutionary time span of 10 Million years.
Hence, this project part showed that the emergence of new transcripts is far more efficient than had been assumed. The next question, namely whether any of these transcripts could carry a function was addressed in project part 2). Here we tested whether sequences with a random nucleotide or amino-acid composition could influence the growth rate of bacterial cells, i.e. would show a bioactivity. Again, we found that this happens much more often than it would possibly have been considered before. In some experiments we found that up to 70% of random sequences had a bioactivity, making the emergence of new genes out of non-coding sequences a very credible possibility. An important practical insight from this project is that our approach yields basically unlimited numbers of candidate sequences for possible pharmacological use. The results from project part 3) confirm these general findings. Using classical genetic approaches to inactivate newly evolved genes, we can show that they have a clear genetic function, including a key innovation in an evolutionary lineage, such as the bony skull in higher vertebrates.
Thanks to the funding of this project, we have made major steps forward in understanding the role of de novo evolution of genes. The results are at a level where they will change the text books and where they will inspire much subsequent work including possible economically relevant applications.
Our project focused on three general topics: 1) generation of a large comparative sequence data set to unequivocally trace the emergence of new transcripts and genes from non-coding DNA, 2) an experimental evolutionary analysis of synthetically generated random nucleotide sequences and 3) genetic studies on a subset of identified de novo evolved genes.
In the first project part we found that a much larger fraction of the genome is transcribed in RNA than had been suspected so far. Instead of less than 10% of the genome that is annotated as being transcribed, we found around 50% yielded transcripts, many of which only in a subset of the respective taxa. We could show that new transcripts arise at a very high rate from non-coding parts of the genome, but most get also quickly lost again. Our findings imply that basically the whole genome is "scanned" for being transcribed within an evolutionary time span of 10 Million years.
Hence, this project part showed that the emergence of new transcripts is far more efficient than had been assumed. The next question, namely whether any of these transcripts could carry a function was addressed in project part 2). Here we tested whether sequences with a random nucleotide or amino-acid composition could influence the growth rate of bacterial cells, i.e. would show a bioactivity. Again, we found that this happens much more often than it would possibly have been considered before. In some experiments we found that up to 70% of random sequences had a bioactivity, making the emergence of new genes out of non-coding sequences a very credible possibility. An important practical insight from this project is that our approach yields basically unlimited numbers of candidate sequences for possible pharmacological use. The results from project part 3) confirm these general findings. Using classical genetic approaches to inactivate newly evolved genes, we can show that they have a clear genetic function, including a key innovation in an evolutionary lineage, such as the bony skull in higher vertebrates.
Thanks to the funding of this project, we have made major steps forward in understanding the role of de novo evolution of genes. The results are at a level where they will change the text books and where they will inspire much subsequent work including possible economically relevant applications.