By Anthony Strattner
May 7, 2002 | Imagine a cell programmed to record events over time, a kind of flight recorder for living systems.
It soon may be possible by commercially engineering biological systems such as molecular memory, which enables stable recording of internal and external cellular events. An engineered cell might be designed to perform specific functions, such as recording cellular activity during a specified time period and then retrieving that information, making analysis of drug interaction much faster and more precise.
This is part of the biomolecular engineering vision of Joseph Jacobson, head of the molecular machines group at MIT's Media Lab, and his colleagues Eric Lander, director of the Whitehead Institute Center for Genome Research, and George Church, both leaders in the Human Genome Project. Together with Noubar Afeyan, chairman and CEO of venture capital firm NewcoGen Group, these scientists set out to create a company that could design and build novel biological systems with science-fiction sounding capabilities.
The company was launched in late 2000 with the name engeneOS Inc. (pronounced "ingenious"). Its stated mission: harness "nature's operating system" by leveraging recent advances in the physical sciences and
|Creating CAD Systems Biomolecular Engineering
|A computer-based biomolecular design system would be very similar to those used to develop integrated circuits and electronic systems.
IT to automate biomolecular engineering. Just as electronic-design automation (EDA) technology revolutionized electronics manufacturing in the 1980s — enabling companies like Intel to make microprocessors progressively smaller, faster, and cheaper — engeneOS wants to create a similar automated system for making biomolecular devices.
A glimpse of engeneOS' technology can be gleaned from a recent paper in Nature, "Remote electronic control of DNA hybridization through inductive coupling to an attached metal nanocrystal antenna," of which Jacobson was the senior author and Kimberly Hamad-Schifferli the lead author (Nature, January 10, 2002). The group used radio frequency radiation to induce denaturing of bound DNA while leaving surrounding molecules relatively unaffected. The switching was fully reversible. EngeneOS is a sponsor of the Media Lab, and has licensed the technology from MIT.
"We were interested in creating a new method to electronically and remotely control individual biomolecules. These directions may be useful in interfacing the molecular scale machinery of biology to the infrastructure of digital electronics," says Jacobson.
To that end, researchers at engeneOS are charged with taking molecular biology beyond identifying and characterizing novel drug targets. They treat genomic information gathered from the Human Genome Project and other research as "building blocks," or engineerable parts, to be catalogued in a database and licensed to companies who want to build biomolecular machines.
To make biological molecules engineerable — that is, capable of being mixed and matched in combinations not found in nature — engeneOS researchers must map out the rules of each molecule's behavior. They will then be able to design molecular components based on these rules, build a library of components, and ultimately piece together certain components into molecular machines for sale.
In addition to selling its own biomolecular devices, which it plans to patent, engeneOS intends to make its component library available to drug companies, and chemical and materials concerns such as DuPont or BASF — any company that wants or needs to manipulate its products at a molecular level. EngeneOS executives envision their component library, along with patented design rules and automated tools, helping to foster the rise of a new class of engineer: a multidisciplinary professional cross-trained in molecular biology, physics, and perhaps chemistry. These engineers would be designing products, such as drug-discovery assays or molecular electronic switches, on a nanometer (one billionth of a meter) scale .
Tall Automation Order Using a combination of genomic mining software, such as Hidden Markov Model applications from the University of California at Santa Cruz, and Computer Aided Software Engineering-based reasoning software, a team of research scientists at engeneOS' office in Waltham, Mass., has begun the Herculean job of mining a wide range of databases for DNA sequence data. "There are currently about 800-plus completed genomes," says John Chan, director of bioinformatics at engeneOS. "All those are fair game for us."
In addition to DNA sequences, other sources of data include X-ray crystallography and nuclear magnetic resonance (NMR), as well as literature and lab work.
"We're pulling in twice as much data as a typical genomic company needs for determining gene function," he says. "We must deal with a lot of global structure and sequence comparisons."
The company uses a cluster of Dell symmetric multiprocessor (SMP) servers running Linux as its main computational platform. The cluster combines multiple SMP servers using high-speed interconnects to form a parallel, virtual supercomputer — one of the most economical ways to get mainframe horsepower, says Chan. EngeneOS is currently evaluating the suitability of an Oracle database to house the critical mapping and component data, although no decision has been reached as yet.
Tiny Antennas, Big Bucks Chan plans to have an automated bioCAD system up and running by early next year, using another 12 to 18 months for troubleshooting and debugging before licensing the technology. Meanwhile, a library of proprietary modular components will have been built, including engineered cells and proteins (which engeneOS officials declined to identify), as well as hybrid devices composed of biological and nonbiological materials.
One type of hybrid machine would use a "nano antenna" similar to the molecular machine developed at MIT's Media Lab and reported in Nature. The antenna, attached to a biological molecule, is a crystal of single digit, nano-scale size (about one-hundred-thousandth the width of human hair). When a magnetic field is applied from a radio frequency, the crystal absorbs the energy and transmits it to the attached molecule. The heat of the magnetic field could, for instance, alter the shape of antenna-equipped DNA. After removing the heat, the strands would return to their original form.
Different nano antenna devices could be customized for drug discovery, diagnostics, and gene profiling. For example, a molecular "cage" may be designed that assumes a certain shape when a certain frequency is applied. An inert component, such as a drug, could be placed inside the cage and released at a very precise time and location.
EngeneOS has also targeted applications in addition to the life sciences. "We envision multiple opportunities to engineer biological systems to address major needs in important industries," says Frank Lee, president and CEO. In IT, for instance, chip manufacturers are attracted to the prospect of using biomolecular components to build processors 1,000 times denser and significantly cheaper, bit for bit.
"They see this technology as dramatically increasing the requirements for computation and storage capabilities," says Anjan Mehta, engeneOS chief strategy and business officer. Indeed, figures from Merrill Lynch and Salomon Smith Barney estimate the market opportunity made possible by biomolecular engineering to be worth upward of several billion dollars by 2005, and Mehta wants his company at the leading edge.
"The whole idea of entering this business," he says, "is to create something useful in two or three years, not 10 years."
Anthony Strattner is a writer based in Framingham, Mass., and can be reached at email@example.com.