Guest Commentary: Digital sensor technologies are transforming biomedicine
Having traveled the journey of one medical transformation enabled by digital sensor technology—the management of diabetes—I believe we are poised on the brink of another, much larger revolution in digital sensing.
I have devoted my career to marrying electronics and biologyin the development of bioelectronic systems. Since my doctoral studies inbiosensing at the University of Oxford, I have been captivated by the eleganceof using bioelectronic coupling to analyze the microscopic world, and the massivetechnology leaps that these types of technologies may induce. Having traveledthe journey of one medical transformation enabled by digital sensortechnology—the management of diabetes—I believe we are poised on the brink ofanother, much larger revolution in digital sensing.
Perhaps no better example exists of personalizedmedicine—following the right treatment for the right person at the righttime—than in the current management of diabetes. Diabetes, whether type 1 ortype 2, is ultimately caused by disruption of insulin signaling, resulting inpoor regulation of blood glucose levels that can be controlled through acombination of drugs, diet and exercise.
Whereas the results of most diagnostic assays are acted onby physicians on a timescale of hours to days or longer, a diabetic patient mayneed to know his blood glucose concentration in less than 20 minutes to avoidlife-threatening hypoglycemia. Critically, tight control of high blood sugarlevels is essential in avoiding common complications of the disease includingkidney, retinal and cardiovascular damage. As a result, approximately 6 billionglucose assays are performed by this patient population every year, a numberthat exceeds all other diagnostic assays combined.
Despite obvious utility and demand, this widespread adoptionof personal glucose testing by diabetic patients did not happen until decadesafter the development of the first home-testing systems in the 1970s. Homeglucose monitoring took off, not through any fundamental advancement in theglucose chemistry underlying the technology, but through a shift from opticalto electrical detection that simplified the workflow and leveraged themicroelectronics revolution.
The basic chemistry behind glucose monitoring involves an enzymaticreaction that occurs in the presence of glucose to produce free electrons. Thefirst blood glucose tests relied on additional chemistry to couple glucoseoxidation to photobleaching of a dye, which could then be read by an opticaldevice. Translating that chemistry into a system for home use created amulti-step process that involved a noticeable skin prick to extract blood (0.1ml), blotting the blood on a strip, waving the strip about to air-dry it,timing the reaction and measuring the result in an optical meter the size of asmall house brick.
A fundamental shift occurred when Professor H. Allen Hilland his colleagues at the University of Oxford devised a novel method ofdirectly measuring the oxidative states of the glucose enzymes involved as anelectric signal. This enabled them to produce a device the size of a pen andrequiring only a tenth of the volume of blood (0.01 ml) that was used in theoptics-based system. Interfacing the basic glucose chemistry directly into adigital system without the need for secondary reporter chemistry meant a veryquick, painless, single-step process that could leverage the rapidly increasingminiaturization of microelectronics to become a barrier-free and invaluablepersonal tool for medical management. Current systems use a 0.001ml bloodsample, resulting in virtually pain-free testing.
Like glucose detection in the 1970s, most current systems toidentify single molecules rely on surrogate identification of a molecularlabel. Particularly common are optical technologies, usually to detectfluorescent labels. These methods can be laborious, time consuming andexpensive. As in the case of glucose measurement, I believe a similarlytransformative label-free bioelectronic coupling technology is at hand that is broadlyapplicable to single molecule detection—nanopores.
Nanopores are nanoscale holes in membranes. Experimentally,this hole is most commonly formed by the pore-forming protein alpha hemolysin.When chambers on either side of the membrane are bathed in electrolyte and acharge is set across the membrane, the current through the nanopore channel canbe measured. When a target analyte interacts with the nanopore, acharacteristic disruption in current can be measured to identify that target.Through protein engineering and combination with sophisticated hardware andinstrument fluidics, nanopores may be turned into high-throughput, directelectronic sensors for a variety of analytes.
Nanopores have been explored for more than 20 years in theresearch environment, pioneered by academics including professor Hagan Bayleyat the University of Oxford and professor Dan Branton of Harvard University.However, analysis of one nanopore at a time as has been traditional in academicsetting, is impractical for the industrial environment. To create a highthroughput technology, a method of arraying the measurement of multiplenanopores must be developed. It is only recently that this technology beenpoised for industrial application and is currently under development.
One of the most prominent applications for nanoporetechnology is DNA sequencing; the analysis of DNA to ascertain the order inwhich the four standard bases appear and to gain additional information such asthe description of epigenetic modifications. While the technology has advanceddramatically in recent years, today's sequencing systems may still be comparedwith the mainframe computers of the 1970s. Today, many systems are located in anumber of large centers (genome centers and core labs) around the world thatperform cost-effective DNA sequence-based experiments through efficiencies ofscale. DNA sequencing is startingto be democratized but the scale and complexity of existing technologies stillmakes it a relatively centralized experiment.
As the National Human Genome Research Institute noted in its$1000 genome challenge in 2004, "the ability to sequence an individual genomecost-effectively could enable health care professionals to tailor diagnosis,treatment and prevention to each person's unique genetic profile." This "race for the $1,000 genome" hasbeen the headline of many articles in recent years. While a useful metric, whenmeasuring the cost of a 3 billion base-pair haploid human genome sequence it isimportant to factor in everything; sample preparation, instrument amortisation,reagents, labor, informatics, instrument downtime and even energy. The fullyfactored cost today is still tens of thousands of dollars.
If the insatiable demand for DNA information by researchers,and soon clinicians, is to be satisfied at a realistic cost, a step-change inprice and complexity of workflow is needed. That is most likely to be met whenoptical methods are replaced by direct, electronic ones that are sensitive atthe single molecule level. Nanopore sequencing fits this profile; by removingthe need for amplification, fluorescent labelling, optical detection andtranslation of an optical signal into DNA information. Limited by silicon rather than optics,the development of nanopore-based electronic sequencers would be free to followMoore's law of scaling-up and fulfill the vision of a mainframe-to-PC stylechange in DNA analysis technology.
DNA sequence data is still largely a tool for scientificresearchers, but it is taking steps towards the clinic, particularly inoncology. Reports emerged in 2009 of early personal cancer genomics, withcomparison of the genomes of patients' healthy and cancer cells leading toclues about treatment strategies. Companion diagnostics exist and more are expected;a classic example is that of the breast cancer drug Herceptin, which in theUnited Kingdom is given purely to people who have a particular gene variantthat identifies them as treatment-responsive. With the advent of lower-costtechnology that allows routine large-scale analysis, we await further news ofpersonalized medical developments such as efficacy of drugs against particularviral strains or the pharmacodynamics of drugs according to variants in humanmetabolic enzymes.
Beyond DNA sequencing, there are many further areas in whichdirect electronic analysis of single molecules might contribute towardspersonalized medicine and more. For example, a protein nanopore might betailored for the identification of specific proteins, small organic moleculesand ionic species.
Just as having a simpler, cheaper methodology for DNAsequencing would catalyze the use of DNA sequence information in drug discoveryprocesses, the same potential applies to the analysis of proteins. There isstill no gold-standard platform for protein analysis during the validation anddiscovery phases of drug development. The platform of choice—massspectrometry—resembles the DNA sequencer of today, requiring large start-upcosts and overhead. Immunoassays can also be used to identify specificproteins, but require knowledge of which proteins are likely to be of interestfrom the outset. In the drug discovery phase, immunoassays are simply tooexpensive and cumbersome to be used for proteomic analysis.
Nanopore-based protein analysis couldoffer a simple, electronic and real-time tool for protein discovery, validationand potentially the same technology could be applied in the clinic. Earlyindustrial work is underway to explore the performance of nanopore-ligandcomplexes for the specific and sensitive detection of target proteins.
Another related application of this technology lies in theanalysis of ion channels, the naturally occurring pore-forming proteins thatregulate electrochemical balance across biological membranes. Ion channels areboth drug targets (for example one of the druggable targets on the influenzavirus) and sources of toxicity (for example, the hERG channel).
The same instrumentation that reads the currents acrossnanopores may be used to analyze the action of drugs and potential drugs on ionchannels. In the future, these methods may be incorporated into ahigh-throughput screening tool for routine use in research laboratories.
Just as the development of an electrical readout for bloodglucose monitoring has changed the face of diabetes management, the use ofdigital sensing technology in molecular analysis could have a transformativeimpact in many scientific disciplines.
Gordon Sanghera is co-founder of Oxford NanoporeTechnologies Ltd. He was appointed CEO in June 2005. Sanghera has more than 20years of experience in the design, development and global launch of novel,point-of-care bioelectronic systems. At Abbott Laboratories, he held both U.K.and U.S. director level positions, including research and manufacturing processdevelopment. Before its acquisition by Abbott, Sanghera led the R&D ofMedisense Inc., where he was instrumental in the launch of several generationsof blood glucose systems for the consumer and medical markets. He has also developedand validated market production processes to meet with the regulatoryrequirements for the United States and Europe. Sanghera has a Ph.D inbioelectronic technology and a degree in chemistry.