Researchers look to nature for answers about efficient biofuel production
By Caitlin Elizabeth Rockett
Biofuel seems like a buzzword these days. The term surfaces in news about global warming, energy security, environmental awareness, and economic stability. While the topic seems new, the use of biofuel in cars is nearly as old as the automotive industry itself—the first diesel engine and Henry Ford’s Model T both initially used plant-derived fuels.
But crude oil became the cheaper alternative to what we now call “alternative fuels.” Interest in biofuels has waxed and waned through the decades but growing concerns over the consequences of petroleum fuel use has resurrected the biofuel debate: Can we produce plant-derived fuels that are cost effective, efficient, and environmentally sustainable?
The National Institute for Computational Sciences (NICS) provides the large-scale systems that help scientists like Gregg Beckham and Mike Crowley study how to resourcefully convert biomass—plant matter like waste wood chips or wheat straw—into liquid fuels. Their group is particularly interested in cellulose, a polymer of the sugar glucose, which forms much of the structure of plants on Earth, but is very resistant to deconstruction.
“Plant matter such as wood and straw contain significant amounts of cellulose, but breaking the chemical bonds linking the glucose building blocks is quite difficult,” explained Beckham, a senior engineer at the National Renewable Energy Laboratory (NREL) in Colorado. “But some organisms breakdown cellulose using cocktails of enzymes. We’re trying to figure out why cellulose is so hard to breakdown and how these groups of enzymes do it at the molecular level.”
The team used two NICS resources—a Cray XT5 called Kraken and an XT4 called Athena—for their research on the enzymatic deconstruction of cellulose, particularly to measure how much energy a particular set of enzymes use to break down cellulose.
With the world’s population recently situated at 7 billion and counting, securing food supplies to support mankind becomes more important with each passing year. This stark reality makes using food products like corn or soybeans as biofuel sources utterly impracticable, while the inedible cellulose becomes a serious contender.
“It’s the most abundant biological material on the planet” said Michael Crowley, a principal scientist at NREL. “Every plant on Earth makes cellulose, from trees to algae. So we’re focused on using non-food crops, and plants that can grow just about anywhere.”
Primarily using Athena and another large computing system at the Texas Advanced Computing Center called “Ranger,” the NREL team looked at how two enzymes—Cel7A and Cel6A, found together in natural enzyme mixtures—breakdown cellulose.
Looking to natureIn the late 1940s the US Army sent scientists to investigate why canvas and other fibrous materials used by soldiers in the South Pacific were rotting so quickly. The scientists isolated the fungus Trichoderma reesei that used a particularly effective set of enzymes to break down cellulose. Researchers around the world subsequently discovered that Cel7A and Cel6A were the two primary enzymes the fungus used to degrade cellulose. Since then, scientists have devoted much time to studying T. reesei and Cel7A and Cel6A in particular, but the NREL computational group has explored uncharted territory with their simulations of these cellulose-chomping enzymes.
The Trichoderma reesei Family 7 cellulase (Cel7A, shown in light purple) engaged on a cellulose crystal (green). From left to right, the enzyme consists of a small carbohydrate-binding module, a long flexible linker, and a large catalytic domain that can thread a single chain of cellulose into the catalytic tunnel. The enzyme is decorated on the carbohydrate-binding module and linker with O-glycosylation (shown in yellow) and the catalytic domain is decorated with N-glycosylation (shown in dark blue).
Cel7A breaks down biomass much like a human would unravel rope strand by strand. Cel6A is like a partner in crime, so to speak. Cellulose is directional, so these enzymes specialize in breaking cellulose down from opposite directions. The enzymes thread a single chain of cellulose into their catalytic machinery, eventually clipping off a disaccharide of glucose called cellobiose. Cellobiose can easily be converted to glucose, which can then be converted various fuels, but getting to it is difficult.
The logical first step for the NREL team involved taking an in-depth look at how these enzymes interact with cellulose to produce cellobiose.
Breaking down the problem
The team knew from experimental work that as cellobiose builds up it begins to inhibit Cel7A’s and Cel6A’s ability to degrade cellulose. NREL scientist Lintao Bu studied which amino acids in Cel7A were binding to these sugars, ultimately contributing to inhibition of the enzymes. The goal was to see if converting the native amino acids to other kinds of amino acids would reduce the adhesion between the inhibiting sugars and the enzyme. It was indeed found that several amino acid sites exhibit high binding affinity, which could be reduced through experimental mutations. This information was passed on to experimentalists at NREL and published in the Journal of Biological Chemistry so that other groups could examine the viability of these mutations.
Working with Vanderbilt University professor Clare McCabe and graduate student Courtney Taylor, the NREL team was also able to study the binding module of Cel7A—the component of the enzyme that affixes the catalytic machinery of the enzyme to cellulose to rip it apart. This work emphasized the glycosylation of this binding module, which had not been considered explicitly before. Previous experimental research showed that one to two sugars are attached to the binding module, and Taylor calculated the difference in the strength of binding if the sugar was removed versus if the sugar was present. Her research showed that addition of sugar on the glycosylation site on the binding module can increase the binding affinity dramatically, meaning that the engineering of sugars at these sites might be another strategy to enhance cellulosic breakdown.
|The carbohydrate-binding module from Cel7A contains two sites that exhibit O-glycosylation (shown in yellow). These glycosylation sites have not been considered explicitly in previous experimental studies, but computational work on Athena and Kraken predicted that the impact of these glycans sites on the binding affinity is significant.|
Glycosylation is generally difficult to study experimentally, and many studies remove glycosylation sites because of this inconvenience. The NREL computational group has put marked effort into realistically simulating Cel7A by including glycosylation sites. This work was published in the December 6 edition of the Journal of Biological Chemistry.
In a different study, NREL scientist Christina Payne examined the impact of large aromatic residues (specifically tryptophans) in the catalytic tunnel of Cel6A. These large hydrophobic residues (shown in yellow above) line the catalytic domain tunnels of most cellulose-degrading enzymes and bind to cellulose chains (shown in blue and red) when a single chain is threaded into the enzyme tunnel for cleavage into cellobiose. Payne computationally mutated the tryptophan residues one-by-one to a much smaller hydrophobic residue (alanine) to study the impact on the binding to the cellulose chain along the tunnel. Her results demonstrated that the binding affinity varies dramatically according to their position in the tunnel, suggesting that nature employs a wide range of binding affinities for various functions. This work was recently published in the Journal of Biological Chemistry.
The catalytic domain tunnel of Cel6A is lined with aromatic residues that bind to cellulose chains. The NREL group used Athena to understand the impact of mutating these large hydrophobic residues to smaller amino acids. (Figure by Christina M. Payne)
Beckham, Crowley, and colleagues plan to study other enzymes and cellulose in depth, but their current work has given experimentalists at NREL and in the biomass conversion community much to examine. Their most recent research conducted at NICS is on molecular simulations of the entire Cel7A enzyme on cellulose. Using Athena, the team constructed a molecular level picture of Cel7A using simulations that ran 2 orders of magnitude longer than any previous models of the enzyme to date. These simulations revealed several new features of Cel7A behavior that are experimentally inaccessible, but for which large-scale computing resources can offer significant insights. These results are being experimentally tested currently.
While the journey to sustainable and economically viable cellulosic biofuels isn’t complete, the NREL team has cleared a path for future research —both computational and experimental—into how enzymes break down cellulose. But according to Beckham, they couldn’t have done it without world-class resources.
“NSF and NICS are providing resources that enable not only us but science of all domains to develop a more secure and sustainable future for the US and the global community at large.”
About NICS: The National Institute for Computational Sciences (NICS) is a joint effort of the University of Tennessee and Oak Ridge National Laboratory that is funded by the National Science Foundation (NSF). Located on the campus of Oak Ridge National Laboratory, NICS is a major partner in NSF’s Extreme Science and Engineering Discovery Environment (XSEDE).