Looking for Novel Cellulose Degrading Enzymes in Cow Guts

Scientific American has an interesting article out about a research team at the Joint Genome Institute which is looking for novel enzymes that may help to degrade cellulose.  Almost all ethanol currently produced comes from the fermentation of simple sugars.  In the U.S., the simple sugars often come from corn.  The problem is that this is not terribly efficient.  First, corn for fuel comes from the same land as corn for crops, meaning that the more fuel produced from corn, the less available for food, and we see rising food prices.  Second, the simple sugars we extract from corn come almost entirely from the kernels.   So we grow these great big corn stalks, take the kernels and through away the stalk and cob, which account for the vast majority of the mass in the plant.  The reason we do this is because the stalk and cob are primarily composed of cellulose, a complex sugar which is not easily fermentable.  What would be nice is if we could break the cellulose down into simple sugars, and then we could use existing technology to ferment these  into ethanol for fuel.  If we were able to break down cellulose, we could even skip corn, and concentrate on plants that don't compete with our food supply.  The most commonly referred to possibility is switch grass, a very hardy plant which can grown in all sorts of soil and environments that are typically hostile to traditional agriculture.

So back to the SciAm article.  The JGI researchers figured that cows eat all sort of grass that is chalk full of cellulose.  So, they took a cow which essentially has a door inserted on its side, allowing for easy access to the rumen, the cow's first stomach.  They placed a bag of switch grass inside and waited 3 days.  They then removed the bag and sequenced the DNA of all microbes found feeding on the switchgrass.  Of course there were thousands if not millions of different species of microbes there, but in their sequencing process they were able to divy out the contributions from the various microbes.  (This is a technique called metagenomic analysis.)  They then searched the sequenced genes looking for DNA sequences which are similar to known DNA sequences of genes that work on carbohydrates.  They identified over 27,000 genes.  They actually took 90 of these genes, expressed the genes, took the resulting enzymes and looked to see if they do in fact help breakdown cellulose.  Over half of them did.

It would be interesting to screen a larger set of these enzymes in a high-throughput fashion, identify a set of the most effective enzymes, and then perform multiple rounds of directed evolution, selecting for the most efficient enzyme during each round.  This might give an indication of how much low-hanging fruit there is in improving the efficiency of such enzymes...