Today's paper revisits a topic that seems to have become a recurring theme for this blog – the evolution of gene regulation. (It also involves primates, which is always a bonus). Ryuichi Sakate, from the Japanese Biological Informatics Consortium, and colleagues from across Japan recently published what they call an “initial analysis” of potential differences in gene expression between humans and chimpanzees. “Initial” and “potential” are important words here – this paper really is just one small step towards cataloguing all regulatory differences that evolved since humans and chimps went their separate ways around 5 million years ago.
But small steps still take us forwards.
The researchers took brain tissue from a chimpanzee who had died of natural causes, and a skin sample from a healthy chimpanzee. They then worked backwards from the RNA gene transcripts present in each tissue to determine exactly which pieces of chimp genomic DNA were actively expressed in brain and skin. These copy DNA sequences (cDNAs) were then compared to the human genome, and to databases of human cDNAs, to look for any inter-species differences in the transcription of the 87 genes under study.
The bulk of the paper is made up of statistics detailing the occurrence, location and type of differences between human and chimpanzee gene transcripts. The differences include single nucleotide changes, small insertions or deletions, and splicing of different pieces of genomic DNA (exons) into the transcript. In general, the protein-coding segments of the transcripts contained the fewest changes. This finding is in line with many other studies; most changes to the protein sequence will be harmful, and therefore subject to negative selective pressure, while relatively more mutations are tolerated in the non-coding regions of the transcript. Human and chimpanzee transcripts were more likely to differ in their site of initiation, rather than termination, and may therefore be regulated by different initiating signals.
One possible explanation for this finding is that many different sequences can act as transcription initiation sites, whereas the transcript termination signal is more precisely defined, i.e. it is “easier” to evolve a new initiation signal. However, it is also true that many, if not most, genes use a range of different transcription initiation sites, and the researchers seem to realise that they may not be comparing the full spectrum of chimpanzee transcripts to their human equivalents. This was, to me, the major weakness in the study, and one that the authors acknowledge. As they say in relation to one particular difference they found, “...it is difficult to distinguish the difference between inter-species and intra-species types of [variation]; further data collection is needed”.
Primate researchers are definitely at a disadvantage here. The human cDNAs in the public databases have been accumulated over many years, usually involving multiple independent studies. They therefore represent a reasonably accurate picture of the many different ways in which the same gene can be regulated in different individuals, different tissues, and in response to different environmental cues. The human databases contain examples of single nucleotide polymorphisms (SNPs, single base changes that are present in some, but not all, humans), alternative exon splicing events, alternative gene promoters, and alternative transcription start sites within the same promoter. The full picture is still far from complete, but we've made an excellent start.
The chimpanzee cDNA databases are much less comprehensive. Chimpanzees are rightly protected, and I know from experience that obtaining tissue samples is no easy task. Therefore when a new study finds differences between a chimp cDNA and the equivalent human transcript, it is often impossible to say that this is a genuine species-specific difference rather than a rare chimpanzee SNP or alternative splicing event. We're going to need to sequence many more chimpanzee cDNAs before we can start to make truly meaningful comparisons to human gene transcription.
The paper by Sakate et al did not quite live up to the promise of its title. Don't get me wrong, this is an important piece of work, and I'm glad that someone did it. But I think that the addition of more chimpanzee cDNAs to the public databases is a more valuable contribution than the paper's somewhat premature comparative analysis.