Blankenship and colleagues propose that the ancestral nitrogenase enzyme must have evolved from a simpler enzyme that performed a simpler metabolic pathway and the researchers suggest that some of the uncharacterized nitrogenase homologue genes that they have discovered - genes that code for enzymes that neither perform nitrogen fixation or photosynthesis - may hold clues to what that pathway

We've discovered a new group of enzymes that were not previously known and because of their position in the phylogenetic tree, we think of these as more ancient, he said. We think that these as yet uncharacterized enzymes are probably also doing some kind of similar chemical reduction - reducing a multiple bond to a single bond. The structure of most nitrogenase homologues is such that they take the input of a large amount of energy and use it to break multiple bonds. We're pretty sure that these enzymes are doing a similar sort of activity, but we don't yet know what.

Our idea is that the very early activity was something that was not nitrogen fixation per se, but some other, probably simpler, chemical transformation, probably also involving nitrogen compounds. Then, at a later time, the ability to do nitrogen fixation evolved from that.

Since a number of the organisms that have the unknown enzymes live in exotic environments with toxic and/or hydrothermic conditions, the researchers hypothesize that the uncharacterized enzymes may be relics of an ancestral enzyme that derived nutrients from toxic compounds in the ancient environment.

We think it may be something that may have been useful in an ancient biosphere where certain compounds were more abundant - perhaps an enzyme for cyanide (HN 3 ) or azide (N 3 - ) reduction but we're trying to keep an open mind, Blankenship said.

According to Blankenship, the team is currently testing the metabolic activity of some of the uncharacterized genes by splicing them into E. coli and exposing the modified bacteria to various proposed toxic compounds.

Regardless of what the uncharacterized enzymes are found to metabolize, Raymond, Blankenship and colleagues feel that the process of nitrogenase evolution that they have uncovered also exposes some basic truths concerning how evolution happens in biochemical processes.

Evolution is a great recycler, a junkman who takes some piece that was invented for some other purpose and reuses it in a new way, Blankenship said. Once you do that, you can end up with some amazing things, like the nitrogenase complex - a stunningly complicated molecular machine that does extremely difficult chemistry.