The Berkeley Science Review: Articles

Lab on a Chip
Tiny technologies offer big possibilities (view PDF)
by Paul Hauser

What do the next Mars rover mission, an infectious disease testing unit, and a mobile forensics van have in common? In the very near future, a group of UC Berkeley scientists hope to equip all of these with diagnostic biological "chips" that can run batteries of tests on microscopic samples—from Martian ice crystals to minute drops of blood. The chips, some as small as dimes and some as big as compact discs, harbor integrated technologies to manipulate and characterize biological samples, allowing the work of an entire clinical lab to be done on a single, portable biochip. With their size, accuracy and time-saving advantages, these mobile testing units are fast becoming the weapon of choice for diagnosing and fighting diseases ranging from equatorial dengue fever to rare genetic disorders.

These "labs-on-a-chip" are so named because they integrate laboratory procedures for sample preparation, reagent mixing and biochemical analysis on a single chip, essentially replacing much of the bulky instrumentation or manual operations found in a traditional laboratory setting. Because the elements contained within these chips are highly variable, they can be designed to monitor everything from food and beverage contamination to the spread of HIV. And much like the microprocessor chip powered the computing revolution, researchers hope to seed the next wave of biological discovery and diagnostics with their surprisingly small and elegantly powerful devices.

Expanded scope, shrinking size
Spurred on by the potential of combining the physics of fluid dynamics with the power of materials fabrication and design, as well as the hope of using modern biological tools to solve complex problems, UC Berkeley researchers are working to create integrated solutions in the form of biological microchips. "These small chips give significant performance advantages compared to larger devices, making analysis a lot faster while still retaining the capacity to achieve very high resolution data," says Professor Amy Herr of the Department of Bioengineering. Much of the improvement in data quality is a direct result of the chips' size, since the engineering and physics principles governing microchips are markedly different from those that limit analogous large-scale laboratory instruments. Couple the increased resolution with the extremely small amount of liquid necessary to run such chip analyses and it becomes clear why these devices are making a big impact in diagnostic applications where samples, man-hours, and time are all in limited supply.

As the lab-on-a-chip field races to innovate ever more integrated, more automated, and higher precision chips, biochip engineers are faced with a host of engineering challenges and opportunities. This is where creativity can flourish, because creating microchips that combine multiple laboratory manipulations presents nearly limitless possibilities for design, implementation, and testing. To accomplish these specialized tasks, scientists engineer chip components that filter, purify, sort, distribute, catalogue, and even make copies of biological elements. The fundamental physical and chemical properties of the chip system simply provide the jumping off point for implementing the appropriate combination of intricate design strategies that solve the problem at hand.

Putting cells in their place
So what do these miniaturized on-chip devices look like? In Professor Luke Lee's bioengineering lab, the chips commonly take the form of an array of small moldable polymer "traps" that capture cells from a microfluidic flow. Imagine a glass-encased microchip that looks strikingly similar to the Pac-Man game of yesteryear, with cells maneuvering through a maze of paths and barriers. Pressurized flows push the cells along while single or multi-cell mechanical barriers trap or gather cells, preventing them from flowing out the other end of the chip and housing the cells for downstream manipulations. The trapping wells can be manufactured to vary in size ranging from large wells designed to collect clusters of cells to microscopic traps that capture single cells.

In addition to the maze of capture wells, the lab has developed an on-chip procedure to break trapped cells apart using a targeted electrical current, releasing the cells' internal contents for further analysis. Under an alternative scheme, an electric pulse briefly compresses the cells' outer membranes to allow external genetic material to enter the cell and become incorporated into the host genome. This on-chip genetic trick can be combined with on-chip cell culture in which a temperature-controlled chamber is used as an incubator to proliferate single cells after applying the genetic modifications. Lee's lab is now using these basic cell manipulations to create chips that test and validate current models of cancer cell adhesion, tumor formation, and nerve cell function.

From cells to genes
While Lee's group manipulates whole cells, chemistry professor Richard Mathies focuses on manipulating the biological molecules that compose the genetic basis of cells, namely the cellular nucleic acids, DNA and RNA. The idea to focus on nucleic acids started rather unexpectedly from his group's early explorations of a simple valve system that precisely gates and controls the on-chip flow. "What started as a very fundamental exploration of the power of a simple valve to control fluidics expanded to meet the many needs of a wide variety of practical problems," explains Mathies. The valves, which can control fluids with very high precision, have allowed Mathies's group to think creatively about the management of biological molecules with sensitivities and scales unheard of just a decade ago—an approach that has proved particularly powerful for manipulating molecules like DNA and RNA.

The microvalves are manufactured by sandwiching a thin piece of rubberized polymer between two etched glass plates or wafers. On the glass plate above the membrane, a thin glass obstruction blocks the flow of liquid while the plate below contains an etched depression that is connected to a vacuum pump. Any time a vacuum is administered from beneath the depression, the rubber is displaced down and away from the glass barrier to allow fluid to flow through the valve as long as suction is applied. This elegant control mechanism has given biochips the capability to precisely measure and control nanoliters of fluid (one billionth of a liter), allowing for more efficient fluid control and thus more sensitivity for on-chip measurements. As a result, chips can contain capillaries where, as Mathies describes, "one microliter of fluid [one millionth of a liter or close to the smallest visibly discernable drop of liquid] would now be nearly two meters long on-chip." With this kind of accuracy and control, standard manipulations are transformed into entirely new dimensions of distance, time, and efficiency.

One traditional laboratory process that has benefitted tremendously from the reduced scale and integrated fluid management is the polymerase chain reaction (PCR), a process that uses repetitive synthetic reactions to make millions of copies of identical DNA sequences. The on-chip equivalent of PCR now uses only hundreds of nanoliters of liquid instead of the classical instruments that use a thousand times more volume. Because PCR requires successive heating, cooling, and mixing cycles, the chip can automate this process and the reaction time is reduced to less than 30 minutes, compared to the two or more hours needed for the standard, lab-scale procedure.

In more advanced experiments, the lab has expanded technology developed for the Human Genome Project to form rapid sequencing and genotyping chips that read each "letter" of the genomic material within a minute biological sample. In collaboration with Matthew Francis and Carolyn Bertozzi, also in the Department of Chemistry, the Mathies group developed a single chip that combines single cell capture, PCR amplification of unicellular DNA or RNA, and a genetic screen to compare that cell's DNA or RNA with other cells in a mixed population. Each of these steps involved significant testing and validation in isolation. More such collaborations are in the future, as single-cell genotyping experiments are seen as a necessary tool for stem cell biologists, tumor and cancer researchers and geneticists alike.

Prototyping protein chips
Efforts such as the Human Genome Project and large-scale genetic screening for heritable diseases have provided compelling reasons to pursue DNA-based microchip analysis, but Herr and other researchers are keenly aware that protein analyses are equally powerful for disease diagnostics. It is proteins that really do the work of keeping cells healthy and fighting off disease. Accurate and real-time measures of such processes through protein profiling affords great power for improving treatment regimes and individualized drug development. Herr and her students are primarily interested in harnessing the advantages of the reduced-scale chips to arrive at new and more effective protein manipulations with the hope of improving both laboratory and clinically relevant measurements. As a first step in this direction, they have introduced protein separation, labeling, and detection systems into their glass-etched chips.

On-chip protein separation gels are one of the many customizable tools in Herr's on-chip arsenal. These homogenous polymer gels start as a liquid that is introduced into an empty chip and solidified into place with the aid of laser light. Once the gel is in place, an electrical field is applied to the chip to drive charged proteins through the gel in accordance with their molecular charge and size. Additional membrane-like elements that filter or concentrate proteins from a microfluidic solution also add separation power. These complementary elements are used in concert through a combination of fluid flow and electric current to move and enrich proteins in a prescribed way.

This technology is only useful if the proteins can be visualized, which is usually done by some sort of labeling procedure. In the lab and on-chip, researchers use purified antibodies that have light-emitting molecules attached to them to bind to and visualize proteins that they are interested in following. While long available on the lab scale, the reduced scale environment of the chip allows Herr to concentrate and then label proteins to measure samples of very low abundance. These measures are then used to distinguish normal and disease-associated protein profiles in just a single drop of liquid, such as from a tear, a drop of saliva or spinal fluid extractions, creating "a cleaner format through which standard clinical measurements can be done in an automated or hands-free way," as Herr says.

Beyond the halls of science
While careful not to overstate the impacts of their work that is, at the moment, still largely performed within the confines of the lab, UC Berkeley-based lab-on-a-chip researchers all have visions of developing and validating chip systems with broad applicability to real world problems. As Herr notes, "most of my lab right now is working on the basic material science, systems engineering, and electrophysics of what's happening inside the integrated chips, but all of them have the goal of making quantitative measurements of clinical relevance." And why not, since the microchip platform can address problems that in some cases other methods cannot. In addition, many of the on-chip assays are faster, more portable, sparing in their sample usage, and often more sensitive than their traditional counterparts. While cost can be a barrier for the one-of-a-kind chips, options for large-scale manufacturing are being pursued that may soon create chips that can be produced cheaply enough to be disposable. With this as a future paradigm, UC Berkeley groups are eager to show that biochip assays can provide immediate solutions for diagnostic and industrial needs.

In a recent proof-of-principle study, Herr and collaborators from the UC Berkeley-affiliated Joint BioEnergy Institute (JBEI) and the University of Michigan School of Dentistry published a paper showing that on-chip microfluidics can be used to accurately test saliva for proteins that have clinical importance for oral health. This so-called point-of-care diagnostic study shows that fluorescence-based measurements of a saliva protein known to be associated with progressing gum disease can be used to accurately screen patients for their relative risk of advancing oral disease. And while the results are highly significant themselves, the added benefit is that Herr had only to "ask for small samples and could get out meaningful results, making their clinical collaborators very happy." Additionally, she says, "if we can do it fast and in an automated way they are even happier because they don't have to dedicate their time and resources examining these samples."

Another project currently in the early stages of testing and development in her lab is a diagnostic measurement of traumatic brain injury using protein assays of cerebrospinal fluid and serum. While the details of which proteins will be measured in the study are still under consideration, she and her collaborators plan to measure the levels of a few select proteins that will indicate the degree of trauma and may also quantify the degree of natural and post-operative healing and repair that is taking place following a head injury. From a diagnostic perspective this holds great potential because many proteins can be measured from a single small sample, which eliminates the need for repeated, extremely painful extractions of cerebrospinal fluid. Novel discovery is also possible by augmenting the cerebrospinal fluid studies with highly sensitive blood serum measurements to see if leakage from the central nervous system to the plasma can be detected, and whether this correlates with the degree of trauma.

The diagnostic power of biochips is certainly not limited to diseases of the Western and developed world. Herr is just beginning to become involved in infectious disease diagnosis through affiliations with organizations like the UC Berkeley Center for Emerging and Neglected Diseases and an international group, the Program for Appropriate Technologies for Healthcare (PATH), that seeks to bring healthcare solutions to international rural communities.

With a similar mission, and eye towards solving international problems cost-efficiently, Bernhard Boser of the electrical engineering department has been working for some time on chip-based measurements of dengue fever infection. Boser has developed a small silicon chip that can directly measure blood samples for the presence of dengue fever antigens as a result of a viral infection. Inside his chip, the patient's blood is mixed with tiny magnetic beads attached to antibodies that bind the dengue fever antigens. There is also an immobile plate coated with dengue antibodies that is anchored to the chip. When the virus is present, an antigen sandwich is formed that tethers the magnetic beads to the chip's antibody coated plate via the viral antigen. Any unbound magnetic beads are washed out of the chip and then a device is activated within the chip to temporarily magnetize the bound beads. This magnetic signal strength is recorded by the chip, which corresponds to the number of bound beads and thus the patient's viral load. Because his chips are cheap (less than $1 each), thanks to economies of scale brought about by the consumer electronics industry and the use of older (but still very reliable) chip technology, he has long envisioned modifying the chips to measure a whole host of disease antigens to assess infection rates among isolated, at-risk populations. This type of on-the-ground rapid testing also interests international health organizations who wish to track disease epidemics as they spread by making population-wide comparisons of strain variation and environmental response to the propagating epidemic.

Taking chips far and wide
It is easy to recognize how biochips are immediately applicable to health care and disease diagnostics, yet this is a narrow subset of the practical applications for microchip technology. Lab-on-a-chip groups like Mathies's have repeatedly shown that chip-based chemical detection systems that measure a plethora of compounds have immense commercial and industrial utility for analysis and discovery projects. Although the launch timeline has recently been delayed to 2016, his lab has already manufactured a chip and reader that will travel aboard the next Mars Rover to look for biochemical traces of life on the Red Planet. This chip uses small samples collected from the planet's surface to measure the levels and compositions of molecules that can provide strong evidence as to the past or current existence of life on Mars. The group took great pains to demonstrate that the instrumentation was sufficiently sensitive to make accurate measurements from a variety of samples found on the planet's surface. "We really feel that we developed the most sensitive measure of its kind that is currently available, all because of the technological advances that can be housed within our chips," Mathies says.

Mathies and his group have also been shaking up detection system development with more earthbound innovations by combining their DNA analysis technology with their chip manufacturing expertise. A group of his scientists recently designed, tested, and validated a fully self-contained human DNA forensics chip. The chip intakes a miniscule amount of human blood from a crime scene and outputs a signature DNA profile that can be compared to a national databank of previously characterized genomes. Within the chip, the extracted blood-borne DNA is subjected to a PCR reaction to copy certain regions of human chromosomes that are so highly variable from person to person that they are considered a unique DNA fingerprint for every individual, even members of an immediate family. This is the same technology that is routinely used as evidence in criminal court, but the Mathies lab has developed a chip that can analyze blood samples at the site of the crime in only a few hours by a single technician. With the advent of this chip, the multiple days of work by several forensic scientists are reduced to a chip, a chip-reading instrument, a laptop, and a few man-hours of work.

Positioned on the brink of upward expansion, the lab-on-a-chip researchers at UC Berkeley all envision powerful ways that biological microchips can expand opportunity both inside and outside the lab. If chip technologies successfully reach the marketplace, they will be able to provide up-to-date information to help individuals and organizations to monitor and improve personal and public health, food and drug safety, and environmental metrics. In what has commonly been called our current "age of information," these biological microchips are poised to be a key instrument for furthering our comprehension of the biological information that will be ubiquitously desired for its diagnostic and instructive power in the coming century.

Paul Hauser is a graduate student in nutrition science.

From the sidebar:
There are all kinds of different cells, and often the number and type of cells in a sample can provide a lot of information about the health of a patient or the success of an experiment. Wouldn't it be great if there were a quick and easy way to count and characterize cells? Mechanical engineering professor Lydia Sohn's lab designs microscopic pores, hardly wider than a single cell, to do just that.

The pores are millimeter-long channels, through which liquid flows and conducts a current. Foreign objects (including cells) entering the pore block the flow and thus cause a drop in conductivity. To actually characterize specific cells that pass through, the pores are lined with antibodies that interact with markers on the cell surface, slowing cells with the marker of interest. By measuring the duration of the conductivity drop, it is possible to distinguish between cells with the marker (long drop) and those without (short drop). Sohn published a paper in the August 2008 issue of Lab on a Chip demonstrating proof-of-principle for this system.

The potential medical applications of these pores were the driving force in their creation. Professor Lucy Godley of University of Chicago, a coauthor on the Lab on a Chip paper, is interested in developing a means to quickly diagnose acute promyelocytic leukemia (APL), a blood cancer that usually afflicts adults and can cause fever, fatigue, weight loss, and other unpleasant symptoms. APL is readily curable, but treatment must be administered quickly, often within 24 hours of admittance, and is dangerous when given to non-APL patients. APL patients admitted on nights and weekends are sometimes not diagnosed quickly and may die as a result. Godley believes a device based on the pore system could provide a quick and simple method of detecting APL-associated cell types, allowing a faster response and hopefully saving lives. The portability and low cost of the system should make its incorporation into the medical field relatively easy. The devices themselves cost less than a penny each, and the handheld box to monitor pore conductivity costs only $1000.

Direct cell characterization can be useful for a number of clinical and research applications, but Swomitra Mohanty, a postdoc in the Sohn lab, is interested in a different use for the pores: antibody detection. The body's immune system creates specific antibodies in response to different infections, so identifying antibodies can be a means for diagnosis. The assay uses glass beads that only bind a given antibody of interest. The beads are exposed to a patient's blood serum and then passed though the pores. If the antibodies of interest are in the serum, they will coat the beads, effectively increasing their size and leading to a greater drop in conductivity when traveling through the pore. Mohanty hopes to use this system to study neglected diseases, especially because the pores' portability and low price make them ideal for use in the developing world. Currently, he is investigating the capability of the pores to distinguish between leptospirosis and dengue fever, two tropical diseases with similar symptoms but very different treatments. With any luck, the pores will be out in the field making diagnoses within a few years.

In addition to directly saving lives, the pores can aid researchers working to better understand diseases. Bioengineering professor Irina Conboy studies stem cell aging, a process strongly tied to conditions such as Parkinson's and Alzheimer's. She needs to separate specific stem cells from mixed populations but finds existing sorting techniques such as fluorescent-activated cell sorting (FACS) to be too difficult and destructive. FACS requires special preparation to label the cells before use, and during this time, cell function can be affected. By contrast, cells can be put through the pores without any modification. In addition, while FACS requires many thousands of cells, the pores are effective with just a few hundred. Conboy is finding the pores to be a useful addition to her toolkit, and many more researchers could similarly benefit from this new technology.

Sidebar by Michael J. Brown

Comments on this article? Drop us a line at with 'letter to the editor' in the subject!