Genes To Gasoline

THE DEPARTMENT of Energy estimates that about 1 billion dry tons of cellulosic biomass per year in the U.S. could be converted into biofuels. This estimate is a limit of how much plant material fuel makers could use without upsetting the normal supply of food, feed, and fiber and without requiring a major change in energy infrastructure and agricultural practices.

If every scrap of that material were converted into transportation fuels and combined with the biofuels that can be made from available plant starches, sugars, and oils, the total could replace about one-third of U.S. petroleum consumption by 2030 and about half by 2050, according to several estimates.

The starches, sugars, and oils are ready to go into this new fuel stream: Chemical and fermentation processes are already commercialized or are being launched to convert them into ethanol or other fuels such as butanol, gasoline, jet fuel, and diesel fuel (C&EN, Nov. 17, page 57). On the raw biomass side, most of the biofuels are expected to come from breaking down lignocellulose in bioenergy crops or agricultural residues by deconstructing cell-wall biopolymers into their sugar building blocks.

But therein lies a problem. The "recalcitrance" of the plant material—that is, the inability to quickly and cost-effectively break down lignocellulose—stands in the way.

Picture yourself inside the cell wall of a poplar tree or a blade of switchgrass, and you will see why. Lignocellulose is an impenetrable maze of cellulose chains bundled together into cablelike fibrils. These fibrils are shrouded by a matrix of hemicellulose and other polysaccharides, all of which is glued together by lignin, a randomly structured biopolymer. Lignocellulose is built tough to not come apart.

Few details are known about how cellulose and hemicellulose are constructed in plants, their distribution within cell walls, and how they attach to one another, lignin, or cell-wall proteins, notes Martin Keller, director of DOE's BioEnergy Science Center (BESC), which has its headquarters at Oak Ridge National Laboratory. For that reason, BESC's 300 scientists located at federal labs, academic institutions, companies, and nonprofit organizations across the country have been given the challenge of designing a path through the recalcitrance problem.

By applying the kind of automated high-throughput screening technology and genomics tools perfected by scientists working on decoding the human genome and by drug discovery scientists, they have gotten off to a fast start.

"Our goals are to identify and characterize genes involved in cell-wall biosynthesis and structure and to establish which genes can have an impact on altering biomass recalcitrance," Keller explains. "That includes discovering microbes that efficiently degrade biomass and reengineering them to produce enzymes that can quickly deconstruct cell walls."

BESC is one of three DOE bioenergy centers established in June 2007. The others are the Great Lakes Bioenergy Research Center, led by the University of Wisconsin, Madison, in partnership with Michigan State University; and the Joint BioEnergy Institute, located at Lawrence Berkeley National Laboratory. Each center began with a $125 million treasure chest in funding and five years to see what it can accomplish in reengineering biological processes to develop new, more efficient methods for utilizing biomass.

"The mix of all three centers is good because they are approaching the biomass-to-biofuels challenge from different angles to help facilitate the research," Keller says. While BESC is homing in on the biomass recalcitrance problem, the Great Lakes center is focusing on increasing production of starches, sugars, and oils that can be converted into biofuels, and the Joint BioEnergy Institute is concentrating on microbe-based conversion of sugars into fuels besides ethanol.

The three bioenergy centers are part of DOE's Genomics:GTL program, formerly called Genomes to Life, which focuses on using plants and microbes for energy-related technologies such as biofuel production, carbon sequestration, and environmental cleanup of toxic waste from nuclear weapons development. Genomics:GTL is a sister program to DOE's Joint Genome Institute, which was a contributor to the Human Genome Project and continues to lead efforts to sequence the genomes of plants and microorganisms.

Among plant genomes that have already been sequenced is that of Populus trichocarpa, a fast-growing poplar tree that has promise as a nonfood energy crop in cooler climates; it is the first tree to have its complete genome sequenced. Among microbial genomes that have been sequenced is that of Trichoderma reesei, a fungus originally discovered because it was degrading cotton military uniforms and canvas tents in the South Pacific during World War II. Scientists quickly learned that T. reesei is a prolific producer of cellulose- and hemicellulose-degrading enzymes (C&EN, May 12, page 11).

Other sequenced genomes include those of Clostridium thermocellum, a well-studied soil bacterium that degrades cellulose and then ferments the glucose into ethanol and other products, and the white rot fungus Phanerochaete chrysosporium, which produces enzymes that degrade lignin, helping to clear a pathway to cellulose. Another species on this list is Pichia stipitis, a yeast that ferments the five-carbon sugar xylose from hemicellulose to make ethanol; its activity complements that of Saccharomyces cerevisiae (baker's yeast), which is the current standard for making ethanol from six-carbon sugars such as glucose but can't ferment five-carbon sugars without genetic modification.

Other plants whose genomes are still being sequenced include switchgrass and miscanthus grass. Among the other microbes scientists are sequencing are ones that live in the gut of ruminants such as cows and in insects such as termites, which unlike people have the ability to digest cellulose.

"The Apollo moon shot and the Human Genome Project rallied support for massive R&D efforts that created the capabilities to overcome obstacles that were not contemplated at the outset of those initiatives," says Joint Genome Institute Director Edward M. Rubin. "Similarly, today's barriers to improving biofuels are significant, but genetics and genomics can catalyze progress toward delivering—in the not-too-distant future—economically viable and more socially acceptable biofuels based on lignocellulose."

SEVERAL APPROACHES to breaking down biomass into accessible sugars already exist, BESC's Keller explains. A leading method is to grind the plant material, pretreat it with dilute sulfuric acid, and then let enzymes loose on it to release the sugars. Pretreatment serves to remove or alter hemicellulose or lignin, remove acetyl groups from hemicellulose, decrystallize cellulose, and open the lignocellulose structure to give the enzymes room to work, he says. But lingering problems include the high cost of the enzymes and the fact that the subsequent fermentation by native and modified microorganisms is still relatively slow for handling large volumes of biomass.

Other established ways to process biomass could be more practical for some types of plants and in some locations. For example, gasification is a traditional catalytic method that uses extreme heat to convert coal or natural gas into synthesis gas (carbon monoxide and hydrogen), which can be converted to hydrocarbon fuels and chemicals—processes already employed in refineries. Lignocellulosic biomass can be treated the same way. Pyrolysis is a milder process that uses medium heat and low-oxygen conditions to partially break down plant biomass into "biocrude," an oil that can be upgraded to hydrocarbon fuels.

These processes bypass converting biomass to sugars, but they have trade-offs in energy and capital costs and in control of the types of fuels and chemicals that can be made. Deconstructing lignocellulose to accessible sugars followed by chemical or fermentation processes is expected to be the most practical pathway to biofuels, if scientists and engineers can rein in costs.

"Overall, pretreatment is a very expensive step in biofuel production," Keller says. "What we have in mind is to eliminate or significantly reduce the amount of pretreatment needed. For example, we plan to ultimately come up with plants that are more easily digested and might need just a hot-water pretreatment without chemicals. We can optimize hydrolyzing this biomass to sugars by incorporating our efforts to develop more efficient microbes and enzymes. Once we do that, we will look to improve other plant characteristics, such as yield; the more biomass you produce at lower cost, the cheaper your fuel will be."

A second goal that Keller and the BESC scientists have their eyes on is engineering a multitalented microbe that can disassemble the plant cell wall and ferment the resulting sugars into biofuel in one go, a strategy known as consolidated bioprocessing. This type of one-organism, one-pot process "could be a major breakthrough for low-cost production of ethanol or other fuels and chemicals," Keller says.

BESC's approach to conquering recalcitrance has started with a major study on a fast-growing poplar tree from the Pacific Northwest and switchgrass native to the prairies of North America. Because no single parameter characterizes recalcitrance, Keller says, the standard operating procedure is to screen multiple plant samples under a variety of conditions and measure the amount of sugar produced at the end of each test run.

In traditional biomass testing, the analyst places a sample in a metal test tube, adds water or dilute sulfuric acid, and then heats the tube in a sand or oil bath, he explains. After a prescribed amount of time, the tube is cooled, the resulting slurry filtered, and then enzymes and buffer solution are added to the solids for the cellulose hydrolysis step. Different runs are needed to test various parameters, such as acid strength, temperature, length of reaction time, single enzymes or combinations of enzymes, and the amount of enzymes used. And each run needs to be replicated. The manual process with multiple wet chemistry steps is labor-intensive and time-consuming, and the number of samples adds up quickly, Keller points out.

BESC scientists plan to screen some 4,000 plant samples per year, Keller says. By the time the various pretreatment and enzyme permutations are completed, that could add up to more than 100,000 tests per year.

"TO DO THAT the traditional way would be a staggering task and incredibly expensive," says chemical and environmental engineering professor Charles E. Wyman of the University of California, Riverside. Wyman, who is an expert on biomass pretreatment, has led the BESC effort to develop a high-throughput biomass pretreatment process to handle the workload.

In less than a year, Wyman, postdoc Michael H. Studer, and graduate students Jaclyn DeMartini and Heather McKenzie, working in consultation with scientists at the National Renewable Energy Laboratory (NREL), in Golden, Colo., developed a unique 96-well plate system that permits combining pretreatment and enzyme hydrolysis screens in one process. They construct the well plates from Hastelloy, a commercial corrosion-resistant alloy, so that the well plates can handle the temperature, pressure, and acidic conditions of the pretreatment step.

For pretreatment, a robot places 3–6 mg of sample into each 0.3-mL well and then adds water or dilute acid to nearly fill the wells. Once the well plate is loaded and sealed, it's placed in a chamber and blasted with steam. Using steam to heat up the plate is another innovation in pretreatment studies that achieves even heating of all 96 wells, Wyman says. After a prescribed reaction time, from a few minutes to 20 minutes or longer, the chamber is quickly cooled by flooding it with cold water; the heating and cooling cycles each take less than a minute.

With no filtering required, the robot adds enzymes and a buffer solution to the pretreated slurry to finish the biomass digestion. In the initial screening studies, a standard cocktail of commercial cellulase and xylanase enzymes is used, but in later rounds of study, a variety of commercial, precommercial, and new enzymes will be tested, Wyman says. His group uses high-performance liquid chromatography to analyze the conversion rate and yield of lignocellulose to glucose, xylose, and other sugars. But in practice, scientists will use a simpler colorimetric analysis, he says.

"The new process can identify promising combinations of plants, pretreatment, and biological conversion systems, comparable to the conventional manual approach, but it significantly increases sample throughput at low cost," Wyman says.

"This system should revolutionize high-throughput screening for modified plants as well as for novel enzymes," Keller adds. "Its success demonstrates the power of combining engineering with next-generation biofuels research and shows how engineering can solve the current processing bottlenecks."

Wyman and colleagues continue to optimize the screening process, but they have handed off the technology, which BESC is in the process of patenting, to NREL scientists who are using it to evaluate poplar and switchgrass samples. The new process fits seamlessly into a high-throughput robotic system that NREL was already using for testing enzymes.

"THE PHENOTYPE we are looking for in these plants—recalcitrance—is quite complex," says Brian H. Davison, chief scientist of systems biology and biotechnology at Oak Ridge National Laboratory and leader of BESC's Biomass Characterization & Modeling team. "Finding a way to get a reproducible measure of the resistance to cell-wall deconstruction for accessing sugars is not an easy phenotype to study," Davison points out. "It's not like eye color."

That has required Wyman, Davison, NREL scientists Mark Davis and Stephen R. Decker, and colleagues to set up a multistep process to evaluate plants for recalcitrance. The first step is a compositional screen using "analytical pyrolysis" to look at the natural range of diversity in a pool of plants grown in different locations to quickly determine the least recalcitrant and most recalcitrant examples for further study, Davison says.

The procedure uses molecular-beam mass spectrometry to analyze a few milligrams of sample ground from each plant, Davison explains. It takes a minute to complete. The purpose of this step is to map the distribution of cellulose, hemicellulose, and lignin content of the samples, he says, which can help determine the subsequent pretreatment conditions. BESC scientists have looked at about 1,000 poplar trees and a few hundred switchgrass plants so far, he notes.

In the second step, selected plants move on to an initial high-throughput pretreatment and enzyme hydrolysis screening, where a few different conditions are tested. Some of the more interesting-looking plants from this group will be retested under a broader range of conditions.

"After this initial screening, the question to ask is why a certain tree or grass plant is more easily digested," Davison says. "Is it because of some environmental factor? Or is it really based on genetic factors?"

To make that determination, the BESC scientists follow up with a third step that starts the process of genetic modifications. The team is testing groups of plants from an "activation tagging" study in which snippets of known DNA were randomly inserted into different locations on a plant chromosome.

Separately, BESC scientists are using literature reports, screening evidence, and intuition based on their experience to select groups of candidate genes that they think could significantly influence recalcitrance in the plants, Davison says. The team is manipulating these targeted genes using deletion, knockout, and silencing techniques.

The plants tested in these random and targeted studies are being rescreened for composition, pretreatment, and digestibility using the high-throughput method. For these genetic studies, the researchers monitor changes in gene expression to see which genes have been modified in the random study or what the plant response is in the targeted study.

The genetic modifications are carried out on single cells or groups of cells, Davison points out, so for each round of study, the scientists have to regrow the modified plants using tissue culture to generate a shoot, which then needs to grow for six to nine months before rescreening. The scientists carefully plan out the timing of the various steps so they don't end up with idle time, he notes.

"With this approach, we can start to link the digestibility of the plant sample to some of the genes," Davison says. "These association studies are similar to ones used in medicine to associate diseases with certain places on chromosomes and then ultimately to specific genes."

Scientists don't yet know how many genes are involved in building and maintaining plant cell walls, Davison notes. Researchers estimate that about 4,000 such genes are present in poplar trees and that even more are active in switchgrass, he says.

As soon as the BESC researchers develop a database of "impact genes"—ones that if modified will suddenly make the plant much easier to digest—they will move forward with the project. In the next phase, which is set to begin, the scientists will take the top 50 and bottom 50 impact genes and start validating which changes can be made to overcome recalcitrance without compromising traits such as plant size, productivity, drought tolerance, resistance to pests, fertilizer requirements, and cell-wall stability. "The goal at this point is to create improved plants that express reduced recalcitrance and enhanced biomass production," Davison says.

Understanding recalcitrance takes more than just chemical and genomic analyses, however. Important helping hands in BESC's recalcitrance studies are imaging and spectroscopy techniques that permit analysis of specific plants from the cellular to the molecular level, Davison notes. These studies help the researchers visualize how the lignocellulose structure affects enzyme activity and vice versa. As the scientists start to correlate low recalcitrance with certain genes, he says, the ability to get a closeup look at what is happening inside the cell wall can help explain why.

During BESC's first year of operation, NREL's Shi-You Ding and Michael E. Himmel and coworkers in BESC's Advanced Cell Wall Characterization team have been using a range of techniques to study plant cells and lignocellulose. For example, they use total internal reflectance fluorescence microscopy to look at the macroscale level of plant cells and track the distribution and movement of labeled cells and enzymes.

They also use high-resolution atomic force microscopy to more precisely image the molecular architecture of plant cell walls, including viewing the surface topography and binding of microbial cells and enzymes to cell walls. Other scientists are using one-dimensional and two-dimensional nuclear magnetic resonance spectroscopy techniques and Fourier transform infrared spectroscopy to study cellulose crystallinity at different stages of pretreatment and enzyme hydrolysis.

Keller is confident that BESC will make significant headway toward circumventing recalcitrance in poplar and switchgrass within its five-year appropriation window. But that is just part of what the center is trying to accomplish, he says. The other major thrust is coming up with a model organism for consolidated bioprocessing.

One of the leading candidates is C. thermocellum. Keller points out seminal work by BESC team member Lee R. Lynd of Dartmouth College and colleagues to determine how C. thermocellum utilizes cellulose. Whereas most microbes produce only a few cellulose-degrading enzymes, C. thermocellum produces more than 25 degrading and fermentation enzymes.

The bacterium's strategy for combined biomass deconstruction and conversion to ethanol is to employ a "cellulosome," Keller explains. Cellulosomes are large, multifunctional complexes made by certain bacteria that assemble several enzymes into a single protein structure. These complexes protrude from bacterial surfaces, latch on to plant cell walls, tear apart lignocellulose, and are thought to feed on cellulose polymer chains like slurping a spaghetti noodle.

"We need to better understand and learn how to engineer cellulosomes for improving deconstruction of the complex plant cell wall," Keller says. To that end, Lynd, Keller, and other BESC scientists are studying gene expression of C. thermocellum during cellulose digestion and quantifying changes in expression in response to fermenting different types of sugars.

BESC researchers are also studying microorganisms from hot springs. Teams of scientists have ventured to geothermal areas such as Yellowstone National Park to poke around looking for microbes that grow on trees or plants surrounding hot springs or on trees that have fallen into the pools, Keller says. Learning about the genetic machinery of these microbes could help reengineer species like C. thermocellum to be more heat-tolerant, for example.

These are sensitive organisms that have to be isolated and handled carefully, notes Keller, who traces his scientific roots to the study of "extremophile" microorganisms as a student at the University of Regensburg, in Germany, and during a 10-year stint at enzyme-producer Diversa (now Verenium) before he joined Oak Ridge National Laboratory.

"We are isolating these organisms and growing them on solid substrates, such as poplar or switchgrass," Keller says. "The idea is to keep the microbes growing on a natural material instead of an artificial substrate such as a lab dish."

This method has been successful and has helped the BESC scientists discover several new biomass-degrading organisms that are "very exciting," including one that grows at 85 ºC, Keller says.

BESC IS NOT working in a vacuum when it comes to solving the biomass recalcitrance problem, Keller points out. The other bioenergy centers, other research institutions, and large and small companies are tackling various aspects of recalcitrance, from pretreatment to enzyme development to microbial fuel production. But Keller believes BESC has the largest integrated and focused effort. Research at most institutions tends to focus on one or two aspects of recalcitrance, Keller says, whereas larger commercial efforts are well integrated but target only first-generation technologies for rapid deployment.

"As we reach our one-year mark at BESC, we are looking at how to best collaborate and share information with the rest of the scientific community," Keller adds. "DOE has required that we protect intellectual property, and we will be jointly licensing technology after negotiating with companies that want to commercialize the various technologies." Private-sector involvement is needed to meet the ambitious goals of bioenergy development, Keller says.

A lot of "What if?" scenarios will need to play themselves out as cellulosic biofuels continue to develop. Agricultural economists are starting to weigh in with more authority to urge scientists and policymakers to begin thinking about how to deal with the unintended consequences of cellulosic biofuels to ensure that they can be produced sustainably. Impacts on water and land use, availability, and quality—especially nutrient runoff—as well as greenhouse gas emissions must be considered, they say.

Those warnings underscore the impetus for solving the biomass recalcitrance problem, which is "one of the great scientific challenges of our lifetime," Keller believes.

"The current generation of schoolchildren will be the first to rely on cellulosic biofuels," observes Keller, who has three young children. "It is important to educate the public on what is required to make biofuels. It is an investment in our future. We have to show them how to change our lifestyles to a more sustainable pathway."