Michael Roukes isn't a biologist, but he sure sounds like one. Roukes is a physicist at the California Institute of Technology whose initial claim to fame -- at least in the esoteric world of nanotechnology -- was his work building molecule-sized machines. He began to recognize the tantalizing promise that these ultratiny devices have as biological sensors. Someday, he believes that these sensors will be able to track in real time the complex symphony of electrochemical signaling that goes on inside the human cell. Today, the methods scientists use to understand these processes -- everything from DNA transcription to the way cells respond to invaders -- can only be done by destroying the cell and examining the chemical traces of what remains.
Roukes' vision is to use these tools to help build a new era of personalized health care. Someday nanoscale electromechanical devices linked to powerful computers will be able to monitor human health by constantly sampling molecules flowing through the body. Early signs of a virus coming on might generate a warning to get rest. Cancer markers would send the patient to a doctor before the serious damage begins. And the treatment would be customized to each patient's unique pathologies.
So compelled is the physicist by this biological vision that he has put aside his work on further developing the machines themselves in favor of exploring how they can be used to change biology. In April, 2004, Roukes formed Nanokinetics, a startup that will explore and develop these technologies. He recently spoke with BusinessWeek's Industries Editor Adam Aston about the future of nanomachines and human health. Edited excerpts of their conversation follow:
Q: What are the key innovations in nanoscale devices in biology?
A: In terms of their size and sensitivity, nanoscale devices offer the potential for interacting with single molecules. Nanoscale mechanical devices enable us to measure the forces that bind biomolecules -- for example, an antibody to an antigen -- on the scale of the individual hydrogen bonds involved.
Q: What does this mean for biology and medicine?
A: This has enormous implications. Think about the genomics revolution. We think of the polymerase chain reaction [PCR] as one of the keys to genomics: how genes replicate and turn on and off. PCR enables us to fill test tubes with vast volumes of replicated DNA to do the testing necessary for this.
Contrast that with proteomics, the study of structure and function of proteins. Nanomachines could help us unlock a tremendous amount of knowledge in proteomics. This is the next step in understanding molecular biology, yet we don't have the ability to replicate proteins in the same way, in such volumes, so they're much harder to study. Plus we're learning that there are many times the number of proteins [in the human proteome] than there are genes in our genome.
Q: How will nanomachines make this possible?
A: The nanosensors and nanomachines that are now emerging suggest that we'll ultimately be able to do individually "personalized" medicine. Consider the example of cancer. Through all the innovations and tools that have emerged in the genomics era, we can now do a molecular profiling of large populations of patients suffering from cancer on a patient-by-patient basis. But today, detailed tissue samples would typically come from autopsies. Real-time genomic analysis could help reveal that each patient's version of a similar cancer is distinct at the molecular level.
Q: What does this mean for the business of medicine?
A: Today, big drug makers target these diseases very broadly, not individually. But at the molecular level, each incidence of cancer is different, so the blockbuster drug approach is simply not applicable. [In the future] each disease will be treated uniquely, as a molecularly distinct manifestation. This new paradigm is disruptive to big pharma. There won't be any Lipitor, no big billion-dollar drugs if we're worrying about individuals instead of general classes of disease.
Q: How soon will this happen?
A: Laboratories and clinicians will be making the first use of these devices in the next couple years, and commercial production of very simple versions of these devices are likely within the next five years.
Q: So describe what a patient of the future might experience given these advances?
A: Medicine will be personal, quantitative, and will be executed in real-time. At first, we'll have little devices that will take a bit of saliva or a pin prick of blood a few times a day and provide a complete readout of our current physiological state. This data will be crunched by powerful bioinformatics software that will output what we should do at the moment-- whether it's to help recover more quickly from the previous night's indulgence, fend off a cold, or something more worrisome.
More serious conditions will probably necessitate intervention from mainstream medicine. But with these devices, one's clinical team will from the outset be armed with quantitative information upon which they'll base their initial steps of intervention.
Later, these devices will likely be employed in vivo -- as medical implants, providing continuous real-time diagnosis. These will probably be linked wirelessly to external processors and huge medical databases. Yes, this leads to significant issues concerning personal privacy and medical ethics. But I am absolutely certain that this sea change in medicine is inevitable.
Q: How does technology make this possible?
A: There are many enablers. They're taking shape in labs now: ultrasensitive sensors that work at the molecular level, microfluidic biochemical processors on chip that can manipulate and read out signatures from individual cells, molecular biology, and so on. The biological data that will stream from these new protocols is truly massive, and will only increase. So immense amounts of computational power will be necessary to process the huge amounts of data produced by this new generation of tools.
Q: How is this sort of sensor able to do real-time biological sensing?
A: Technology in progress today will enable entire gene and protein "arrays" to be shrunk down to the size of a single cell, from centimeter-sized today, and their signals will be read out as electrical signals. These will be embedded within microfluidic biochemical processors that allow each cell in a chip-scale cell culture to be monitored continuously.
Q: What are the key hurdles to implementing nanoscale biosensing devices?
A: There are three. For now, most of nanoscience is focused upon -- and perhaps too enamored with -- single devices and interesting fundamental effects. Productive, full deployment of these sensors will only become possible with their large-scale integration to produce systems of great complexity. And, of course, techniques for the mass production of these systems, so they aren't simply one-of-a-kind curiosities, are essential.
Second, in many ways the theory of nanoscience is still in its infancy. It makes relatively simple assumptions based upon fundamental physical science, but the parallel operation of hundreds or thousands of nanoscale processes will introduce huge oceans of unknown interactions and consequences that we haven't yet faced.
Third, there's the materials issue. In nature, biology works by the fusion of chemistry, biology, ionics, microfluidics, and so on. When we really enter the realm of nanoscale real-time biological sensing, it will certainly be through a similar fusion of systems and methodologies from different disciplines.