This morning, Tom Murray at The Hastings Center pointed me to a new paper from James Liao's lab at UCLA demonstrating the first engineered bug that produces isobutanol from cellulose.  Wendy Higashide, et al, ported the artificial butanol synthesis pathway from the group's earlier work in E. coli (see this previous post) into Clostridium cellulolyticum.  Here is the article.

Recall that butanol is a much better biofuel than is ethanol.  Butanol is also not hygoscopic (doesn't suck up water), which means it can be blended at any point in the distribution chain, whereas ethanol must be trucked/barged/piped in dedicated infrastructure until just upstream of a gas station in order to avoid pulling contaminating water into the fuel stream.  Butanol has a long history of use as a transportation fuel, and has been demonstrated to be a drop in replacement for gasoline in existing engines.  See, for example, the work of the 2007 iGEM team from Alberta, and my earlier post "A Step Toward Distributed Biofuel Production?"  One advantage of making butanol instead of ethanol is that butanol spontaneously phase separates from water (i.e., it floats to the top of the tank) at concentrations above about 7.5% by volume, which substantially reduces the energy required to separate the molecule for use as a fuel.

The press release accompanying the Higashide paper describes the work as a "proof of concept".  The team attempted to insert the butanol synthesis pathway into a Clostridum strain isolated from decaying grass -- a strain that naturally consumes cellulose.  Unfortunately, this Clostridium strain is not as well characterized as your average lab strain of coli, nor does it have anywhere near the same number of bells, knobs, and whistles for controlling the inserted metabolic genes.  The short summary of the paper is that the team managed to produce 660 mg of butanol per liter of culture.  This is only about 0.07% by volume, or ~100 times below the concentration at which butanol phase separates from water.  The team lays out a number of potential routes to improving this yield, including better characterization of the host organism, or simply moving to a better characterized organism.

So, a nice proof of principle.  This is exactly the sort of technological transformation I discuss in my book.  But this proof of concept is not anywhere near being economically useful or viable.  Nonetheless, this progress demonstrates the opportunities ahead in relying on biology for more of our industrial production.