SEQUENCING genomes may be all the rage, but the real prize is the proteome,
all the proteins in a particular plant or animal. It鈥檚 the proteins that do all
the vital work in any organism. Now European and Canadian researchers have
developed a computer program that will let researchers get their hands on the
proteome even if the organism鈥檚 genome hasn鈥檛 been sequenced.
鈥淚t鈥檚 a really insightful piece of work. It opens up proteomics to any
organism that people want to study,鈥� says Timothy Veenstra, a proteomics
researcher at the Pacific Northwest National Laboratory in Richland,
Washington.
To identify a protein, biologists usually turn to mass spectrometry, which
they use to measure the mass of the protein molecule. Then they break the
protein into smaller and smaller fragments. From the mass of these fragments
it鈥檚 possible to deduce a number of possible amino-acid sequences for the
protein. If one of the sequences matches that of a known protein, this must be
the one.
Advertisement
But this approach requires an exact match between one of the predicted
sequences and part of a known protein or gene. It fails if you don鈥檛 know an
organism鈥檚 genome, says Andrej Shevchenko of the European Molecular Biology
Laboratory in Heidelberg, Germany.
So Shevchenko and his colleagues wrote a program that looks for proteins that
are similar, not just exact matches. It works like this. Suppose an unknown
protein is broken into five main pieces, and the spectrometer data suggests 200
possible sequences for each piece. If a large number of these are similar to the
sequence of a known protein, then the protein being analysed should also be
similar to that known protein. In this way, you can identify proteins even
though you don鈥檛 have exact sequence data.
His team has already used the method to identify 15 out of 19 proteins
purified from the yeast Pichia pastoris. They were able to do this by comparing
these proteins with those of a distantly related fungus, Saccharomyces
cerevisiae, whose genome has already been sequenced.
The technique will allow biologists to study proteins in animals that are
easy to work with, but whose genomes are unknown. This will help them understand
what the related proteins do in people. 鈥淭he characterisation of proteomes of
organisms where the genome is not known can be done with the same sensitivity
and throughput as if the genome was known,鈥� says Shevchenko, whose work will
appear in Analytical Chemistry.
For example, cancer researchers study chromosome segregation using the frog
Xenopus laevis. 鈥淏ut its genome is not known,鈥� says Shevchenko.
The software can also help identify novel proteins, Veenstra says. He and his
colleagues are studying the bacterium Deinococcus radiodurans, which has a
genome similar to that of E. coli, but which can survive nearly 500 times as
much radiation. Shevchenko鈥檚 software could first identify the common proteins,
says Veenstra. Then you can look at the proteins unique to Deinococcus that may
be responsible for radiation resistance.