November 15, 2013 (Vol. 33, No. 20)
Josh P. Roberts
There are two major design subfields in microfluidics—chip-in-a-lab and lab-on-a-chip. Chip-in-a-lab aims to shrink laboratory operations down to the microliter (or smaller) scale, mainly for the cost, time, and reagent savings.
Lab-on-a-chip tries to fit a laboratory benchtop into a handheld device. Both approaches were in evidence at the recent Gordon Research Conference, “Microfluidics: Challenges, Advances, and New Technologies for Diagnostics.”
Although microfluidics, and in particular paper microfluidics, is a hot, sexy new topic and seems like the latest, greatest thing to hit the practical sciences, “almost everything in the field is old,” according to Barry Lutz, Ph.D., research assistant professor in the University of Washington’s department of bioengineering. The theories have been around, in some cases, for a few hundred years. The materials themselves may be advanced, but even they have been exploited at least since the early 1980s in the form of home pregnancy tests.
In these tests, fluid moves on its own without the aid of a pump. There is no need to pipette. All the reagents are stored on the device. It’s very fast, very easy to use. “The simple lateral flow test is just an ingenious invention,” Dr. Lutz exclaimed.
The paper may look uniform, “but it’s really a bunch of tortuous pathways,” Dr. Lutz said. Yet by understanding a few basic physical principles of how fluids move through paper, which in many ways is very similar to how pollutants in groundwater move through sand or rock, a researcher should be able to understand and predict behavior in a microfluidics device.
The Reynolds number (a relationship between a fluid’s inertia and its viscosity) helps to determine whether flow is fast or slow, turbulent or not, for example. The boundary conditions are different for flow through paper than they are for flow through a tube. Fluid moves through the entire width of a paper conduit, while fluid near the wall of a tube is essentially pinned and does not move. Wicking power, or capillary pressure, propels the fluid into the paper. The flow due to wicking does not happen instantaneously because of a countering force—flow resistance. “We have to drag the fluid through these small pores,” Dr. Lutz said.
There is also dispersion. “We have soluble components moving with the fluid—proteins, small molecules, salts, maybe nucleic acids,” said Dr. Lutz. “An important question is: Do they move with the fluid or does the paper change the way they are distributed in the fluid?” It turns out that both effects matter—classical chromatographic effects, where the molecule sticks to the paper for a few seconds and then unsticks, as well as hydrodynamic effects, where the fluid wends its way through the paper.
Dr. Lutz discussed his lab’s work using dissolvable sugar as a programmable flow delay in a multistep microfluidic system. By applying different concentrations of sucrose to the paper and allowing it to dry, the group was able exploit the varying time delays created as the fluid reached the sugar and had to dissolve it. “We were running an ELISA-like test in which the reagents have to be run sequentially,” Dr. Lutz said. In this way, fluid could be added to all the legs of the device at the same time, yet the first leg would turn itself on while the second waited its turn, and so forth.
Compared with a traditional 96-well-plate-scale ELISA, in which incubation times are on the order of minutes to hours, one of the brilliant things about flow tests is that mixing takes place in seconds by diffusion, noted Dr. Lutz.
Diagnostics on Demand
The University of Washington has been working with GE Global Research and other collaborators under a DARPA “Diagnostics on Demand” program to develop technology that will enable low-cost, high-performance tests to be used in a point-of-care or even a limited-resource setting.
The paper- or membrane-based handheld device will be user-friendly, instrument-free, and fully disposable, with power contained within (so there is no need to plug it into a wall). “We want to be able to do a test from sample to readout in less than an hour, ideally 30 minutes,” explained David Moore, Ph.D., the project’s co-principal investigator and manager of GE Global Research’s Membrane and Separation Technologies Lab.
According to Dr. Moore, “membranes and papers are what make up the porous channels that allow for the transport of the biospecimens that you’re collecting, processing, amplifying, and then detecting.” The steps will all be carried out within modules of the microfluidic device with no further intervention necessary, producing a visual result.
The first test will be for detection of MRSA DNA from a nasal swab. MRSA is problematic in institutional settings such as military bases (thus the interest of DARPA), prisons, and hospitals where many physicians who will later practice in remote settings do their training.
The groups are also looking at using blood as a sample fluid, and at identifying DNA or RNA of other pathogens such as STDs and viruses. Ultimately, they hope to develop a device that could be used in the field to detect a panel of infectious diseases, for example, so that a soldier’s health could be monitored in a war setting.
The readout might be a series of lines or a pattern of dots “that could then be imaged or captured using a cell phone,” Dr. Moore explained. “The phone would have an algorithm associated with it to capture, process, store, and transmit that information to the healthcare provider of interest.”
For now, they’re in the process of doing multiple iterative loops of developing what a prototype will look like, with a variety of different pack-of-cards-sized designs being considered. But although the form factor has not yet been decided upon, “I can say we’ve made very nice progress integrating the modules of the technology into something that works from sample all the way to result,” Dr. Moore noted.
DARPA is not the only agency looking to develop rapid diagnostics for use in remote locations. Wave 80 Biosciences is working under NIH contract to develop rapid quantification of HIV from a single drop of blood. The company is designing an enclosed microfluidic cartridge to directly target the HIV virion RNA molecule.
Patients sometimes present at clinics in places like rural Africa with flu-like symptoms and very large levels of virus in their blood before anti-HIV antibodies can be detected. The infrastructure doesn’t exist to send them home, run a sophisticated diagnostic overnight, and expect them to come back the following day for a result. They may be highly infectious, so it’s important to have a test to determine the presence of HIV RNA on the spot.
The device’s 10-cm-long cartridge is designed to fit inside a handheld processing unit containing electronics for operating fluidic actuators and incubating the sample at a specified temperature. Once the incubation is complete, “you stick it in the analyzer unit, which has the user interface as well as the optics for reading out the luminescent signal,” explained Wave 80 scientist William Behnke-Parks, Ph.D. Up to 10 samples can be processed in parallel.
“A lot of steps have to occur in a small amount of real estate,” Dr. Behnke-Parks said. The process begins by placing a drop of blood onto one side of the disposable cartridge, and then the company’s proprietary slit-capillary array fluidic actuators (SCAFAs) act like pumps to move the blood through. The blood is lysed to release the HIV RNA. It’s passed through a solid-phase extraction to clean up the sample, and the RNA is amplified isothermally. HIV RNA can then be optically detected following an “ultrasensitive bipartite signal amplification assay.”
The process is even more complex than it sounds because “you’re working with a limited number of molecules and have to push through very small channels to control it—the physics is very different at that length scale,” pointed out Dr. Behnke-Parks. For example, when mixing on scales that we’re used to—centimeters or meters—there is a large component of the flow that is inertial, and this causes turbulence. But microliter-size samples and submillimeter dimensions operate under a very low Reynolds number regime where inertia, and therefore turbulence, is almost negligible.
The team at Wave 80 used a pulsatile flow to rapidly and successively interleave two fluids such that each pulse produces a plug so small that they can mix by diffusion to produce a homogenous distribution within seconds.
Some issues in microfluidics—mixing, for example—are easier to deal with in a laboratory setting with “big, powerful pumps to pump fluids through small mixing geometries that then have very high backpressures,” Dr. Behnke-Parks said.
Yet using syringe pumps or peristaltic pumps to control flow has its own drawbacks, noted Anne Le Nel, Ph.D., R&D manager at Fluigent. Such solutions in the nanoliter range can lead to poor temporal response times and, because of the rotating motors, pulsing and instabilities in flow delivery.
In 2005, the Paris-based company introduced the Microfluidic Flow Control System (MFCS), a pressure-driven flow controller with no mechanical parts that can be used to independently drive liquid through multiple microfluidic channels. To do so, it applies pressure from a source such as bottled gas or a compressor onto liquid reservoirs. “The use of pressure to move liquid can have the consequence of having very stable flows and also short settling and response times,” said Dr. Le Nel.
Yet customers began asking to know the flow rate. So Fluigent integrated precise flow sensors to monitor the velocity of the liquids inside the microfluidic systems.
Still not content, some customers wanted to be able to control the flow rate as well. “So we developed a specific algorithm in which the pressure pump adjusts its pressure to be able to control the flow rate,” Dr. Le Nel explained.
It is possible to order flow rates in a syringe pump system, for example. “But then you don’t know if there is some delay in the response, and you also don’t know if there is stability or instability in the flow,” Dr. Le Nel said. “Because you don’t have a sensor, you are only pushing liquids.”