Mixing and Matching Microbes for Food and Fuel—What do cows and termites have in common? “Nada, zip, zilch,” most people would say. But Bryan White, a professor of animal sciences and a member of the Molecular Bioengineering of Biomass Conversion and Host-Microbe Systems themes at IGB, would say, “not so fast.” Cows and termites both digest plant cell walls—cows specialize in grasses, of course, and termites in wood. That makes them both “ligno-cellulosic deconstruction systems.

Why should we care? Because, among other reasons, identifying the most efficient mechanisms for breaking down plant cell walls would solve a major bottleneck in the development of biofuels. Cows and termites, or more precisely, enzymes from microbes in the cows’ and termites’ digestive systems, can provide some insight.

Although biofuels are much in the news these days, there are several obstacles blocking their viability. The first involves identifying a plant substrate that does not compete with a food source (like corn does) and delivers enough energy per unit to be economically produced.

The next step is to quickly, inexpensively, and efficiently break that plant material into something that can be fermented. That’s where White comes in. For many years, White has studied the cow rumen’s treasure trove of bacteria that have evolved to break down plant cell walls.

“If we were going to pick a system to find enzymes that break down plant cell walls the rumen seems very logical, because cows live on plant cell walls from all different types of plants,” says White. “It (the rumen microbiome) sees everything except for wood—and sometimes it even sees wood. And these microbial enzymes (in the rumen) have been breaking down plant cells walls since the dawn of time so these organisms have seen everything and have probably evolved their system to be able handle everything.”

White first became interested in the rumen and its enzymes for a different reason. His original project was to find the feedstock that cows digested most efficiently so that farmers, particularly in developing countries, could raise less expensive meat, milk, and leather goods. Conveniently, the science behind that project and the biofuels one is essentially the same.

“In the back of our heads we always thought we could use this same approach (of studying enzymes from rumen microbes) for biofuels,” says White.

Not much is known about this microbiome, however. There are, says White, about 1011 organisms per milliliter in the rumen. Researchers’ best guess is that those organisms represent at least 3,500 species.

“We have the genomes for fewer than 10,” says White.

Not long ago, White had a paradigm shift that made learning about the biome more efficient. He began thinking of the microbes in the rumen not as individual species, but rather as a single organism working together. That means that, instead of trying to tease out a single organism, researchers can essentially reach in, grab a handful, and work with them all together.

“You take all DNA as if it were a single chromosome even though it’s 3,000 chromosomes or 7,000—I don’t even know the number,” says White. “But I’m taking everything that’s there in theory and I’m querying it all,” says White.

This “metagenomic” approach yielded results, but with some limits.

“I’m only querying what I can clone, so if it’s uncloneable DNA, I’m not querying it. If it’s there in low abundance I’m not querying it, I’m only querying things in high abundance.”

This approach was successful. The first three plant-cell-wall-degrading enzymes another lab found, using this approach, had never been seen before.

“That says ‘hey this really is a resource,’” says White.

But then, next generation sequencing, or pyrosequencing, was developed. Previously scientists identified an organism, cloned it, and then sequenced it, whereas pyrosequencing enables them to skip the cloning step. This technique meant White’s team could examine in “unprecedented depth” the microbe populations of the rumen because it became possible to find DNA that was not clonable and/or occurred in low numbers. This technique also was much faster than previous ones.

By comparing the enzyme components of the rumen with that of another system (cue termite) that also breaks down cell walls, White confirmed that the set of enzymes present depends on what plant matter the system is digesting, such as grass (cow) or wood (termite). His findings were published in the Proceedings of the National Academy of Sciences (PNAS) in February. This result supports White’s hypothesis that enzymes are “substrate specific.” His research now is focusing on a designer enzyme cocktail for each given substrate, from Miscanthus to mahogany.

White is most excited about using pyrosequencing to ask bigger biological questions: what is, precisely, a species, and how do enzymes evolve, and what are the driving forces? He can use the architecture of the cellulosome, the system that breaks down cellulose, to look at evolution and diversity.

This ties in very neatly with White’s overarching interest, which has always been to understand how hosts and microbes interact.

“Fundamentally the science that drives this is really cool,” says White. “It’s just a nice bonus that these fundamental answers very often have outcomes that are very useful to society.