Devouring data

Hundreds of brutal, efficient killers crowd Carter Butts’ home, bringing grim death to any unwise enough to venture into range. But fear not. These are magnificent bug-eating plants that populate nearly every continent.

Drosera capensis, commonly known as the cape sundew, also replicates human digestive processes under extreme circumstances. So Butts, a University of California, Irvine sociology and statistics professor, and his research colleague Rachel Martin – a fellow carnivorous plant enthusiast – wondered: Why not install one in a lab and see what it could teach humans?

The result: an 18-month project to sequence and study the genome of D. capensis, the first of its family and the third carnivorous plant ever to be sequenced.

They chose the cape sundew, in part, because it drew Charles Darwin’s attention in his 1875 book on insect-devouring plants. The 194 species of sundew use a sweet, sticky secretion to snag insects on long tentacles. Other tentacles trap the bugs, and sessile glands digest them.

Butts and Martin overcame hurdles in extracting DNA, analyzed genomic sequences with tools typically used in sociology, and explored novel enzymes that break down proteins – and which could provide significant medical benefits to humans.

“Carnivorous plants are neat,” says Martin, a UCI associate professor of chemistry and molecular biology & biochemistry. “Nobody knows exactly what proteins and enzymes they possess. They started out trying to kill insects to protect themselves from being eaten. This has evolved six separate times into mechanisms for trapping insects and using them for nutrients.”

The duo’s discoveries, reported recently in two peer-reviewed journals, show that these awe-inspiring killers may be equally extraordinary in what they can do for humans – erasing biofilms from medical instruments and safely battling stubborn fungal infections, for example.

Environmental challenges faced by the plants suggest they have creative ways of surviving, according to Butts. “It’s a lot harder for a carnivorous plant to eat things than it is for us,” he says. “You can’t heat up the environment to make the enzymes go faster, and you have to compete with bacteria and fungi and other creatures that are also trying to eat your food. For them to survive, they need to have really effective and interesting properties.”

Butts and Martin, already collaborators on a National Science Foundation project focusing on proteins, conducted the plant study under the aegis of the California Institute for Telecommunications & Information Technology, where Butts has his lab.

The institute is devoted to multidisciplinary research – in this case, pairing Butts’ methodology training and Calit2’s computational resources with Martin’s keen understanding of biochemistry.

“My work has mostly been on social networks, simulation methods and statistical methods,” Butts says. “This whole arena was an entirely new direction for me. We adapted a technique that I had previously used to study individual life histories, believe it or not, to look at proteins.”

The team soon ran into difficulties. First, they needed to extract usable DNA – a challenge because plants excrete substances that mask their DNA when chopped up.

“A lot of the genome sequencing techniques and the techniques for collecting DNA are optimized for human studies because the driving force for a lot of this is medicine,” Martin says. “I mixed and matched techniques until I came up with one that actually worked. UCI’s Genomics High-Throughput Facility was very helpful in checking whether the genome was good enough to sequence.”

The effort took roughly three months, followed quickly by creation of the sequencing data – a process shortened to about a week by technical developments.

“We took advantage of next-generation sequencing technology, which has reduced, quite a lot, the cost and time of getting usable reads,” Butts says. “That takes the extracted DNA, breaks it into tiny pieces and then uses a machine to read out the DNA sequences of these little pieces. You get a puzzle that has hundreds of millions of pieces, and you have to put it together, which is a computational challenge.”

Calit2 provided key resources during this step.

“It requires access to high-performance computing, which is something I have in my lab and was able to bring to bear on the problem,” Butts says.

The team reviewed gigabytes of data to figure out how to organize the puzzle pieces into readable sections of the genome.

“There’s no universal strategy that always works. You experiment with modifying the techniques and try different strategies. We went through quite a number of assembly algorithms and strategies for processing and cleaning the data until we found something that finally worked,” Butts says.

“We believe we have about 90 percent of what’s there. That’s pretty good for an initial assembly, and that means what we have is good enough that we can start to find proteins and do other kinds of studies of the organism with fairly high confidence,” he adds.

The data yielded a rich trove of information.

“It really is exciting to have this huge basket of stuff and realize there are treasures in there. You just have to learn how to get them out,” Butts says.

Martin quickly identified 44 protein sequences, far too many to economically produce. Using molecular modeling, she and Butts helped target the most novel and exciting sequences. She focused on proteins that may work as antifungal agents, producing three in her lab. Two are chitinases, meaning they generate enzymes that digest the chitin in bug exoskeletons.

“This is a very valuable agent to use as an antifungal because it can poison fungi but not humans, since humans don’t have chitin,” Martin explains.

She also has created a protein-specific insert, a subsequence of a protein that cleaves off after the protein is made. It is itself an enzyme whose functions may include defending against pathogens and cleaning cell walls.

“This is something the plant seems to be using as an antifungal or antibacterial agent. We’re thrilled about catching that in the lab,” Martin says.

The duo has more genomes in the pipeline too. And they’re eager to see what they may discover from other carnivorous plants.

“We found all this great stuff from only the first carnivorous plant that we’ve looked at, and there are numerous kinds of plants and lineages. So we’re really excited about comparing what we found here to what’s going on in other carnivorous plants,” Martin says. “We expect to see a lot of similarities, because things evolve in similar directions to get the same function, but there’s no reason it has to be exactly the same.”