A cynic might suggest that there's an `in' game among the apostles of post-genomic technology to invent the most obscure `ome'. By now, we're all aware of the transcriptome, the proteome and, perhaps, the metabolome. But how about the peptidome? At present, there are only six papers on this topic in the whole of PubMed, yet the latest delivers information of real utility to comparative physiology.

The peptidome is defined as all the peptides in a cell or tissue, together with their post-translational modifications. It's thus a cut-price proteome, a snapshot of all peptides small enough to extract and load onto a mass spectrometer, without prior tryptic digestion. This sounds dry, until you see it in the context in which it will invariably be unleashed; the tissue is the central nervous system, and the peptidome is all the neuropeptides. So in a single experiment, involving (in this case) surprisingly small batches of 50 Drosophila larval brains, we get a view of all the neuropeptides that interest the animal, rather than the peptides that we (as experimenters) deduce by staring at the genome.

How well did it work? Surprisingly, considering that it's something of a sport to pick off Drosophila brain-related genes, only seven neuropeptides have been purified traditionally, and 18 have been identified from other routes. Here, the authors identified 28 neuropeptides, including eight that had not been identified or predicted in any way previously. That's a pretty impressive hit rate. Of course, the approach was not perfect, as low abundance peptides, or those with restricted temporal patterns of expression, might not be picked up in a single assay. The authors flag this limitation with a list of known or plausible peptides that do not show up on their assay. Among them, for example, is a CRF-like diuretic peptide that our group has just shown to be expressed in only six cells, so this may provide a lower estimate for the sensitivity of the system.

Could this approach be extended to other species? Very quickly and easily: Drosophila is famous for having a sequenced genome and a whole battery of genetic resources. However, only the former quality is significant here. It is easiest to deduce peptide identities by comparing mass fragmentation patterns against all possible peptides encoded by the genome (there are computer programs like MASCOT to do exactly this). There are several other insect genomes coming onstream in the next few months (Anopheles gambiae Aedes aegypti and, perhaps, Apis). So relatively few experiments could provide us with a database of most insect neuropeptides of species of interest, and modest scaling up could do the same for fish, rodents...

The learning curve and capital cost for some of the new `-omics' technologies can baffle or frustrate the physiologist: but here is one that delivers real results now. Given that J. Exp. Biol. has published 53 papers on neuropeptides in the last few years, peptidomics is likely to become important to many of us.