Science and sensibility

The pufferfish Fugu rubripes has long been infamous as a dangerous way to eat your lunch but it is also garnering increasing fame amongst biologists as a model organism. Fugu may seem an odd creature to study as a model of human biology, being separated from us in evolution by around 400 million years and not as amendable to experimentation as our closer relatives the mouse and rat. Nevertheless this little self inflating fish, along with a number of its piscine brethren, are teaching us a lot about how we work and how we got here in the first place.

By the late 1990s the human genome project was pouring out data on the A's T's C's and G's that make people, well… people. The problem was that the A's T's C's and G's really don't mean anything by themselves. In order to work out what a newly sequenced gene might do one usually compares it against a database of other genes whose functions are well understood. If you have a stretch of DNA that has a sequence very similar to a gene for turning DNA into RNA then it's a safe bet that your gene is involved with copying DNA into RNA. If we are to understand the code that the human genome project spits out then we need to decipher and understand a whole bunch of other organism's genomes. How such data is "understood" or annotated in the first place is a fascinating topic that I intend to cover here at some future point - for the time being suffice it to say that it is and that we have databases full of it.

In the early days of genome biology "simple organisms" like the bacteria Escheria coli provided scientists with a picture of the minimal set of genes required for basic cellular life . An E. coli needs to make a copy itself (including its DNA) convert its DNA into RNA into functional proteins and metabolise chemicals from its environment. All organisms on earth are faced with these same problems so it is of no surprise that genomes of prokaryotes like E. coli have helped assign functions to sequences derived from the human genome project. In fact many E. coli genes are not merely similar to human ones; they are the result of the same ancestral gene taking very different paths through evolutionary history.

Moving up in complexity the genome of the baker's yeast (Saccharomyces cerevisiae) shows what genes are required for the more intricate eukaryotic cell structure to develop. These organisms are of course of great interest in themselves but are limited in their use as model for human biology because they are single celled organisms. If you are looking for genes that allow the development of organs, nervous systems and intricately balanced multi-cellular processes then obviously you will need to study the genomes of organisms that contain such structures. In 1998 the 1mm long nematode worm, Caenorhabditis elegans became the first animal to have its genome deciphered. This sequence contained some 20 000 genes and when the fruit fly Drosophila melanogaster was added to the growing number of organisms to be the subject of a genome project scientists felt they were developing a pretty good idea of the set of genes it took to make an animal. But even with Drosophila and C. elegans undergoing sequencing there were significant evolutionary, developmental and morphological gaps to be bridged to get to something like a human being. It was decided a model vertebrate genome should be sequenced and there was one stand out candidate. Fugu.

So what makes Fugu's genome such a great model of our own? First, it is approximately seven and a half times smaller than our own and it was thought that, being a vertebrate, it would hold a total number of genes similar to our own genome. Perhaps more importantly Fugu's genome contains very little so called repetitive DNA. Over half of your genome is made from small sequences of DNA repeated thousands or even millions of times, Fugu by comparison has 15% of its genome made from repetitive sequences.. The quickest way to sequence a genome is to sequence enough small stretches of DNA that you get a lot of overlapping sequences and all the fragments can be put together like a jigsaw puzzle. To extend the jigsaw analogy repetitive sequences are like sky, hundreds of identically cut pieces of sky. Since they are all exactly the same it makes joining up overlapping sequences nearly impossible and since Fugu has less repetitive DNA it was easier to piece together the overlapping sequences.

The Fugu genome is now more or less completely sequenced and it has proved to be remarkably valuable tool. Fugu is effectively a "readers digest" version of the human genome - all the essential information with very little of the excessive baroque passages that litter our own DNA. As well as the repetitive DNA mentioned earlier eukaryotic genes tend to contain large sections of DNA called introns that are never translated into proteins. In mammals these introns often account for more DNA that the actual coding sections of a gene. The position of introns in genes is highly conserved between Fugu and human but the Fugu ones are routinely twenty times smaller. Additionally the space between genes is greatly reduced. Also 75% of the genes predicted from the human genome project are shared with Fugu.

So what use is this reduced genome? Well for a start the lack of large introns and much junk DNA makes it easier for computer algorithms to predict which sequences from the human genome project are actually genes. The initial sequence of the Fugu genome revealed almost 1000 new genes in the human equivalent.

A number of Fugu genes are being used as models for more complex human ones. A prime example is a gene called parkin - mutated parkin genes are the main cause for inherited Parkinson's disease. The human parkin gene is massive but the protein it codes for is actually very small. Of the 1 400 000 bases that make up your parkin gene only about 4 500 are ever translated into protein. Fugu also has a parkin gene but it is 350 times smaller than yours, having skipped out on the massive intron expansion that has occurred in us mammals. More excitingly for Parkinson's researchers Fugu parkin does exactly the same job in Fugu cells as it does in ours and the actual protein coding sequence seems to have changed little in the last 400 millions years. Researchers hope to use the smaller Fugu version of the gene to make minor changes in the regulatory information (on/off switches in and around a gene) that would probably be lost in the mess of the much greater human counterpart. Doing so may lead to a better understanding of what changes in human parkin lead to Parkinson's disease. Information on 25 other papers made possible by Fugu genome project can be read here, and many hundreds of papers featuring Fugu can be accessed here.

The bit I find most exhilarating about the pufferfish's contribution to our understanding of the world is they way it can tell us how evolution has happened in the so called higher organisms. Firstly by looking directly at the molecular evidence and retracing some of the events that have occurred to us and the fishes since our ancestors took to the terra firma and theirs continued their watery ways. The venerable PZ Myers describes exactly how this has been done using a pufferfish related to Fugu; even to the extent that we can guess what the ancestral vertebrate genome looked like.

Secondly we can reflect on the remarkable fact that you, a pufferfish and I have the same basic set of genes. If a fish can live and breathe underwater and even make a toxin capable of turning people into zombies with basically the same genes as you then how on earth can anyone still claim that the human genome is a 'blueprint' for humanity. Fugu and the other genome projects have taught as that a lot of evolution has had nothing to do with what genes one has, but how one uses them. Sure gene and genome duplications have created more genes for selection to work on but just as importantly evolution has lead to novel expression of existing genes. To use a better analogy than the blueprint one (but one just about as well worn) the genome of an organism is like a recipe, describing when each gene should be turned on when to induce the developmental programs that make us people and fish fish. The cues for these instructions are also encoded in the genome in "regulatory sequences" which were the topic of a recent paper freely available from the public library of science. Since so much of vertebrate evolution has been retooling and retuning existing proteins and pathways with subtle changes in regulation the Fugu genome represents a great opportunity to investigate the processes that lead to us air breathing, land dwelling divergent fish's emergence.