As amino acids go, glutamine is much more than a simple protein building block. Functioning either in the brain during recycling of the neurotransmitter glutamate, or as a key step in the metabolic pathway that produces urea to detoxify ammonia, the enzyme that produces glutamine is essential for most creatures' well being. Consequently glutamine synthetase has `an extraordinarily long evolutionary history' explains Patricia Wright. Most mammals only have a single glutamine synthetase gene, but as there are two copies of each chromosome in mammals' cells, there are always two almost indistinguishable copies of the gene, one on each half of a chromosome pair,called alleles. However, sometime during the last 100 million years, some fish species broke away from the rest of the fish evolutionary tree, and duplicated their diploid chromosomes, resulting in an animal with four sets of chromosomes. So when Wright and Tom Mommsen set out independently in their laboratories to investigate the tetraploid rainbow trout's glutamine synthetase gene, they fully expected to find that the fish had doubled up a single glutamine synthetase gene when they duplicated their chromosomes,resulting in two different glutamine synthetase genes. But after months of patient cloning and sequencing, both scientists began to realise that the sequences just didn't match up! Instead of having two alleles of each gene,the `alleles' were too different to code for the same protein, so the rainbow trout must have four glutamine synthetase genes(p. 1511)! Somewhere along the line, the rainbow trout's diploid ancestor must have already duplicated its glutamine synthetase gene before it went tetraploid.But if the rainbow trout had four glutamine synthetase genes, other fish that had stuck firmly to the diploid branch of the evolutionary branch might have a duplicated gene too, just like the trout's ancestor. Brent Murray dived into the zebra fish and fugu genomes, and after searching through thousands of unnamed genes in the enormous databases, he discovered that both fish had two glutamine synthetase genes. And when Wright and Mommsen reconstructed the enzyme's phylogentic tree, they realised that the fugu and zebra fish's ancestors had both duplicated the gene, but at different times. All of which made Pat Walsh wonder whether his favourite, the toadfish, might also have multiple copies of the gene.He began scrutinising mRNA throughout the fish's body, to see if the glutamine synthetase from different tissues had been produced by a single gene. But when he analysed the nucleic acid that produced the gill's glutamine synthetase, the gene was completely different from the glutamine synthetase produced in the fish's brain and liver. Instead of having a single glutamine synthetase gene, the toadfish had also duplicated the enzyme, but only used the second gene to produce glutamine synthetase in its gills(p. 1523).Walsh was intrigued to find out when the toadfish had duplicated the gene,and constructed another phylogenetic tree, which he adds agrees well with Wright's and Mommsen's phylogeny. He looked for the new gene's closest relative. But instead of resembling other fish glutamine synthetases, the new gene was right out on a limb, and looked more like the Xenopusenzyme!Why would a fish enzyme look so much like an amphibian's when the kinetics of urea production are so different on land and in the water? Walsh explains that for some unknown reason, toadfish suddenly stop excreting ammonia. This should cause the toxic waste product to accumulate in the gill, unless the fish could suddenly switch on urea production. By ramping up glutamine synthetase production, the enzyme could mop up the stray ammonia and convert it into non-toxic glutamine, but he explains that the enzyme in the toadfish's gill would have to a high affinity for ammonia, just like a terrestrial animal's glutamine synthetase. Walsh thinks that this `ammonia trapping'mechanism could explain the similarity between the fish and the frog genes,but he knows that there are more questions to be answered before this theory becomes more than speculation.