Biochips take a lesson from 'Proteomics'
By Chappell Brown, EE Times
September 23, 2003 (1:36 p.m. EST)
URL: eetimes.com
Though it is far from evident in the day-to-day business of semiconductors, the integrated circuits business has deep roots in biology. Digital architecture became an unstoppable technological force in the 20th century, analogous to the clanking steam-driven machinery that transformed 19th century society. But it is all but forgotten that digital circuits, designed in the 1950s and 1960s, were inspired by that era's understanding of how the nervous system operates.
That the pioneers basically got it wrong has not dampened the enthusiasm generated by machines originally called "electronic brains" that have evolved to populate the developed world. Despite the seeming impossibility of creating realistic artificial intelligence, artificial life and autonomous life-like robots, inventors and researchers never tire of trying. And in trying, they have generated ubiquitous electronic systems that are increasingly linked by communications networks, giving the inorganic industrial world a nervous system.
The biological metaphor is simply too compelling to ignore, and that is perhaps the best argument for a bright future for the emerging technology of BioMEMS. A combination of silicon chip technology and the basic molecular processes of living organisms, BioMEMS promise to go beyond mere metaphor by actually merging organic processes with electronic circuitry.
Visionaries are talking about blending molecular biology with computational systems at the atomic scale to create a true hybrid called bio-nanomechanical systems (BioNEMS), which could become a major force in nanotechnology. But in the near term, many small biotech concerns and some major semiconductor makers are already fielding products based on BioMEMS.
The potential market, and the commitment of the semiconductor industry to develop it, is somewhat hazy. Most major semiconductor companies have small BioMEMS research projects mainly targeted at DNA analysis, but actual product initiatives seem to be sparse.
Motorola got into the BioMEMS game early on in 1998 and was marketing a DNA analysis system called Codelink by 2000. Then, in 2002, as a result of a company reorganization, it scaled back its program and sold the Codelink team and product rights to a British company. At the time, Motorola's CEO, Christopher B Galvin, said that he still viewed the field as important but felt the company needed to focus on its core technologies. Indeed, despite a very tough semiconductor market, Motorola has stayed in the game, marketing a second DNA analysis system called the eSensor DNA Detection System. Its mode of operation is also significant: the bioarray uses DNA hybridization to drive the assembly of circuits on the array. A completed circuit signals the presence of a specific DNA strand, perhaps a forerunner of truly biological circuits.
Intel recently disclosed a 12-person lab devoted to BioMEMS [visit www.eetimes.com/at/im/news/OEG20030320S0028], and IBM has at least one researcher at its T.J. Watson research center working in that area. IBM has made a bigger commitment to developing software for DNA analysis, which turns out to be a formidable problem in itself. In fact, software has stimulated the fastest advances in DNA research by allowing research teams to coordinate the massive databases they generate in sequencing projects.
But DNA analysis is only the first rung on the biological ladder. Without an understanding of how DNA information is transformed into proteins, it has limited technological use. There are a lot of promising applications for DNA detection: linking gene deficiencies to disease, drug design, patient monitoring-many high-growth applications are out there. But there is a much larger field beyond DNA.
The next rung on the biological ladder is proteomics, the study of how proteins are formed and how they operate. DNA has some mechanical abilities based on its lock-and-key method of bonding, but proteins are the physical building blocks and basic mechanical units of living organisms.
While DNA is essentially linear strand, proteins exist in three-dimensional configurations that are so complex that supercomputers are needed to analyze their physical and chemical functions.
A Nucleic Acid-Programmable Protein Array (NAPPA) mimics the DNA process within cells that create the proteins.
If BioMEMS research can reach the same level in proteomics that it has now achieved for DNA, a genuinely disruptive technology will emerge. But getting to that level will require a much more varied group of biotechnologies than is found in the typical BioMEMS lab.
One company that appears to be mounting a serious effort in that direction is ST Microelectronics. Its lab in Catania, Italy, is assembling a broad set of expertise that goes beyond current BioMEMS technology.
During the past six months, the lab has announced a record-breaking silicon LED, a DNA analysis chip and a protein film that enhances the performance of solar cells.
Expertise wanted
STMicro is trying to assemble a range of expertise in areas such as silicon optoelectronics, bioelectronics and nano-organics as an essential step in realizing next-generation technology. And proteomics is on its agenda, according to the lab director, Salvo Coffa. He said he thinks proteomics is clearly a high-growth area.
The big problem facing anyone trying to make progress in that direction is the "fragmentation of expertise" in the field, according to Coffa. Not only does his lab need wide ranging expertise in electronics, materials science, biotechnology and organic chemistry, the target application areas are highly specialized.
"It is similar to the beginnings of the semiconductor industry, where you would have to have all your expertise in-house to design and build a circuit," he explained.
Today, sophisticated computer systems do most of the work in laying out a working circuit and it is taken for granted that the design can be shipped to any number of fab lines that will crank out working chips. But in the beginning, circuit designers, chemists and silicon crystal growth specialists were all working on the same project to get anything that worked.
STMicro is entering product development projects with a number of biotech firms in order to tap an even broader range of expertise. There is an explosion of research into the path-from a DNA-coded gene to the complex of proteins it produces.
Mostly conventional macro scale biotechnology equipment is used in this work, providing an opportunity for anyone who can reduce it to the chip level.
The proteins produced by a genome are not only large in number and complex in their behavior and structure, but the control system that produces them can mutate during the process, producing a wide variety of related proteins from a single DNA code.
The inverse problem faced in clinical applications is how to find the gene that produced a given protein, or looking at the protein profile of a healthy patient compared with a diseased person. Solving those problems efficiently will require the mass chemical analysis only available with BioMEMS technology coupled with sophisticated bioinformatic software.
Many startups are looking at ways to get into the game by designing simple-to-use systems that can perform the entire analysis. But success will require semiconductor MEMS expertise along with a host of other obscure specialties.
That type of clinical diagnosis system is only a starting point. The capability that will open up with the development of automated proteomic systems could be turned around to create protein complexes with predefined properties. One example is research into how to attach metal and semiconducting nanoclusters to proteins. One could conceive of a bioarray that creates 3-D protein clusters with intricate semiconducting devices and circuits-the BioMEMS analog of a silicon wafer full of chips.
Judging from history, it is unlikely that the major semiconductor companies of today will spawn the BioMEMS revolution. Fairchild Camera and Instrument Corp. didn't appreciate the efforts to integrate diverse electronic functions on a substrate by an electrochemist and an electrical engineer, Gordon Moore and Robert Noyce. They struck out on their own, becoming prime movers in the semiconductor revolution. One wonders what hindsight in the year 2020 will reveal about the BioMEMS revolution. |
