Centrifugal microfluidic platforms based on the compact disc format offer advantages for tackling the typical steps in a micro total analysis system for nucleic acid IVDs.
Share [4]
Microfluidics research has become extensive, particularly regarding the development of sample-to-answer platforms focusing on protein and nucleic acid (NA) IVDs. The holy grail in this field is the development of a single system that can accept a wide range of biological samples (e.g., blood or saliva), perform the required sample preparation and analysis, and quickly produce test results with little or no user input required (besides the initial sample introduction). But few examples of such sample-to-answer systems for NA IVDs have gone beyond the research and development stage, and even fewer are commercially available. In contrast, there are notable sample-to-answer systems for measuring simpler analytes in biological samples, such as fingerstick blood glucometers, blood gas and electrolyte panels by Abbott Laboratories (Abbott Park, IL) and Epocal Inc. (Ottawa, Canada), and urine-based pregnancy tests.1
A myriad of terminology in the microfluidics field describes integrated systems that can accept samples and perform a complete analysis: micro total analysis system (microTAS), lab-on-a-chip (LOC), and lab-on-a-CD, when referring to centrifugal platforms. Other terms include point-of-care (POC) and sample-to-answer. Sample-to-answer and microTAS are approximate synonyms for the methods and systems discussed in this three-part article. While LOC and lab-on-a-CD refer to particular steps or processes adapted for microscale devices and centrifugal microscale devices, respectively, they do not necessarily indicate a complete system. Point-of-care refers most commonly to IVD tests performed at or near a patient's location, as opposed to sending samples to a lab. For the purposes of this article, the terms microTAS, POC, and sample-to-answer are used interchangeably when discussing complete microfluidic IVD systems.
The previous article (IVD Technology, November/December 2009) discussed the typical steps in an NA sample-to-answer system by considering biological and engineering factors. This article discusses centrifugal microfluidic platforms based on the compact disc (CD) format, along with the advantages they offer for tackling the steps in a microTAS for nucleic acid IVDs. In the third article, the biological and microfluidic IVD groundwork will be used to present a forward-looking system capable of performing NA diagnostics.
Centrifugal Microfluidics
Microfluidic platforms have the potential to tackle and integrate the NA diagnostics steps discussed in the previous article. There are numerous demonstrations of the benefits gained by moving from a typical wet-bench set-up to a microfluidic device. Such benefits include reduced reagent use, significantly decreased total processing time, increased abilities for parallel processing, and reduced process variability via automation.2,3
As with any field, microfluidics has its particular set of challenges. Foremost is the issue of laminar flow typical for the low-Reynolds-number conditions in many microfluidic conduits. Mixing, which is required for fast reactions, in low-Reynolds-number conditions remains challenging. Other problems include the poor downscaling behavior of mechanical pumping systems, stiction, which can cause many valves to fail, and channel-clogging air bubbles. Despite such challenges, centrifugal microfluidic platforms based on the CD format provide many solutions to overcome some of these limitations inherent to microfluidic disposables.
CD Advantages
When developing any fluidic devices, the main concern is how to get liquids to and from the areas of interest in a controlled manner. This general problem can be encapsulated by the need for two related technologies: pumps and valves. The centrifugal platform based on the CD format provides simple and effective modes for pumping and valving.
 [5] |
| Figure 1. (click to enlarge [5]) Schematic of microchannels on a CD: a) Two reservoirs connected by a single channel; b) Hydrophobic valve made by a channel restriction in a hydrophobic material; c) Hydrophobic valve made by functionalization of the channel surface with hydrophobic material; and d) Capillary valve made by a channel widening in a hydrophilic material. |
Centrifugally induced pressure on the fluid as the CD spins causes fluid propulsion on the CD. Researchers have characterized this type of flow extensively in the literature.4,5 The volumetric flow rate depends on the speed at which the disc spins, the distance the liquid is from the center of the disc, the geometry of the fluidic channels, and the fluidic properties (i.e., density, viscosity, and surface energy) (see Figure 1a). By using combinations of different channel geometries and spin speeds, flow rates ranging from 5 nl/s to 0.1 ml/s can be achieved with a high degree of accuracy and precision. The typical fluid-pumping rotation speeds range from 300 to 2000 revolutions per minute.
Centrifugal pumping forces on the CD provide many advantages compared with other alternative pumping methods, such as syringe, peristaltic, and electroosmotic pumping. While pressure-driven syringe and peristaltic pumps provide good control over large flow rates, they can be unwieldy when trying to miniaturize and/or conduct parallel processing.6 In addition, the pressures needed to move fluids through the microchannels do not scale well at 1/r4 so they become very large in the micro domain, which makes implementing the pumps into small, high-throughput platforms difficult.
Electroosmotic pumping methods can overcome such problems since they scale better in the micro domain and can be easily adapted into microchannels using microfabrication. However, these methods highly depend on the pH and ionic strength of the fluid being pumped.6 Moreover, the high-voltage power supplies (more than 1 kVolt) required in these systems make them expensive and impractical.
Centrifugal pumping forces do not need large power supplies (only a low power motor), do not depend on the fluid's pH or ionic strength, and do not need fluidic interconnects or tubing for force application.
Consequently, different assay steps requiring different sample and buffer properties (e.g., surface energy, pH) can be combined into a single CD device. Centrifugal pumping also provides forces across the entire length of a fluid element, allowing smooth, controlled, and reproducible flow. This feature offers additional benefits: bubbles that develop have less chance of disrupting downstream fluidic processes, which also results in low residual hold-up of fluids during transfer. Finally, many individual systems can be placed on a single CD, making parallel processing possible without further complexities in the pumping hardware.
Valving on the CD is performed using two main valve types: hydrophobic and capillary (see Figure 1b-d). Hydrophobic valves can take on two different forms: one using changes in channel geometries and the other using surface modifications (see Figure 1b-c). In both cases, a fluid can be forced past the hydrophobic valve by increasing the spin frequency past a critical value. Capillary valves are commonly used in CD platforms and are the result of the balance between centrifugal and surface tension forces in a hydrophilic material (see Figure 1d).
When fluid being pumped through a narrow channel by centrifugal forces reaches an abrupt widening, a large surface tension force develops at that widening. If the surface tension force is greater than the centrifugal force, the fluid flow will stop even though the CD continues to spin. At a certain spin speed, known as the burst frequency, the centrifugal forces will overcome the surface tension forces, and the fluid will continue to flow down the channel. By designing microfluidic structures with channels of varying capillary valve sizes, controlling when a valve opens can be achieved by increasing the CD's rotational speed. The passive valving available on CD platforms facilitates implementation of complex microfluidic structures without the need for complex mechanics beyond the motor required for pumping.
The pumping and valving mechanisms unique to centrifugal platforms make modular design and implementation of complex fluidics achievable, providing significant advantages compared with standard microfluidic platforms for POC applications. A more comprehensive overview of CD platforms can be found in the literature.7 The advantages of the CD platforms discussed above have led to the development of several platforms geared toward NA diagnostics. When combined with the fabrication methods described below, centrifugal microfluidic platforms can be prototyped with relative ease.
CD Fabrication
CDs for microfluidic applications can be prototyped using multilayer structures made of inexpensive polycarbonate plastic and double-sided, pressure-sensitive adhesives (PSA). Using simple computer numerical controlled (CNC) machining, channel widths as small as 1 mm are machined into the polycarbonate plastic. A cutter-plotter cuts channel widths as narrow as 200 μm in thinner materials such as 100 μm-thick PSA or thin plastic films. Once the appropriate pieces have been designed and machined, they are aligned centrally and radially, and laminated together using the PSA layers.
 [6] |
|
Figure 2. (click to enlarge [6]) Schematic showing the assembly of a typical five-layer, microfluidic prototype CD. Hard plastic fluidic CDs (mm-scale thickness) are bonded together using thin microfluidic adhesive layers (μm-scale thickness), providing a large range of dimensions on a single device.
|
Even the simplest microfluidic CD consists of no less than the following five layers: top polycarbonate CD with CNC-machined sample loading, sample removal, and air venting holes sealed using a thin adhesive film during operation; pressure-sensitive adhesive with microchannel features cut using a plotter; middle polycarbonate CD with CNC-machined channel features; another pressure-sensitive adhesive; and solid bottom polycarbonate CD to seal off the channels (see Figure 2). Microfluidic CD platforms can involve more layers to accommodate more complex fluidics. Moreover, different devices and substances can be placed inside the CD during fabrication (e.g., beads, lyophilized reagents, or filters). The CDs can also be exposed to O2 plasma treatment or functionalized with bovine serum albumin to create hydrophilic and hydrophobic surfaces, respectively.
Using such fabrication methods allows for a manageable transition from prototyping to mass production of injection molded plastic disposables. The ability to produce injection-molded parts with microscale features is important for successful manufacturing and is discussed further in part three of this article.
Current CD Platforms
Several CD platforms have been developed that are geared toward NA diagnostics and the analysis steps described in part one of this article. While this section will highlight certain platforms, it is by no means an exhaustive review of all microfluidic CD systems.
Part one of this article discussed a blood sample preparation system that involves almost all of the necessary preparation steps. Along these lines, researchers have developed a platform designed to extract pathogens from μL-sized samples of whole blood.8 By using the aforementioned CD fabrication methods, a system was developed that operates as follows: whole blood is separated via centrifugation, specific cells/viruses are captured from plasma using antibody-labeled beads, the cells/viruses are eluted to perform concentration and purification, and finally lysis is performed. The entire procedure takes approximately 12 minutes to run. This system is an excellent example of the feasibility of combining several NA analysis steps on a single CD platform.
However, as expected, this platform is complex, and requires several different components and technologies. Since ferrowax (a mixture of wax and magnetic nanoparticles) is used as a valve, the system requires placement of the ferrowax inside the CD during assembly, a laser for heating, and a moving magnetic platform to actuate the wax once it is melted. The system reportedly could not prepare polymerase chain reaction (PCR)-ready materials directly from whole blood, which may be due to cell lysis being performed as the last step, thereby possibly releasing inhibitors. Consequently, another sample preparation step would be required in a product-ready NA diagnostics system, which again underscores the complexity of sample preparation and the non-triviality of integrating sample preparation steps.
The most crucial sample preparation step is lysis, and researchers have developed a standalone microfluidic lysis CD that relies on a bead-beating method powered by magnetic forces.9 This CD uses small magnetic disks placed inside each lysing chamber that oscillate by interacting with stationary magnets on a static CD platform. The oscillation causes shear forces that result in lysis. Once lysed, solids are removed by centrifugation, and the supernatant (nucleic acid) is extracted using a unique siphon valve.
Validation of the CD platform showed that it can accept a raw sample of up to 60 μL and deliver a clarified sample from E. coli, B. globigii, and yeast cells in approximately five minutes. Moreover, the platform has also proven to be efficient for viral lysis and sample homogenization. This system requires a stationary magnetic platform, and placement of the materials inside the CD during assembly. This CD delivers only a clarified sample and does not perform NA concentration or purification.
The next common step in NA analysis is amplification via PCR. Researchers have developed a microfluidic PCR card system that is designed for future integration on a CD capable of rapid and efficient NA amplification.10 Peltier thermoelectric devices were selected to perform active heating and cooling of the PCR cards. Keeping future CD integration in mind, the researchers realized that not only liquid valves but also vapor valves will be needed during thermocycling. Thermocycling causes vapors to accumulate as the liquids approach boiling temperatures (95º C). Large pressures build up as a result, causing the sample to expand, so the PCR chamber needs to be sealed and isolated to avoid loss of sample solution. With Peltier devices already integrated onto the PCR platform, a novel ice-valve scheme developed by Phasiks Inc.was chosen in which Peltier devices are used to freeze a small plug of liquid at each end of the reaction chamber, preventing the movement of both liquid and vapor during thermocycling.
The microfluidic PCR card platform achieved rapid heating and cooling ramping speeds of up to 10° C/s. Because the Peltier devices allowed active heating and cooling based on polarity of the current used, shorter thermocycling times were achieved compared to standard PCR systems that rely on active heating but feature passive cooling only. Validation of the microfluidic system's PCR abilities was performed by amplifying an E. coli gene in a 25-μL reaction volume. A 40-cycle amplification was conducted by using a temperature profile of 7s, 15s, and 15s (ramping times included) at 95° C, 60° C, and 72° C, respectively. Using gel electrophoresis, the observed limit of detection was 10 copies for the particular assay and conditions chosen, and was completed in less than half the time of a standard benchtop thermocycler.
While this microfluidic PCR card system provides solutions to several PCR system problems (e.g., thermocycling and valving), it requires a large amount of power both for thermocycling and ice-valve actuation. However, all hardware is left out of the PCR card, making it a good example of partitioning of functions between the disposable and the instrument.
The final step in NA analysis is detection. A rapid, flow-through DNA hybridization disposable has been developed that is designed for detection via an optical DNA microarray.11,12 The rate-limiting step when performing analysis with DNA microarrays is the time for hybridization to occur, which can take upwards of 18-24 hours. Enough time must be given for the DNA of interest to diffuse passively through the entire solution and be exposed to every capture probe on the surface. By using thin and narrow microfluidic chambers, hybridization time can be reduced by making the diffusion distances much smaller. Moreover, by incorporating active flow, mass transport of sample DNA to the capture probes can be enhanced.
Using microfabrication, a silicone microfluidic disposable was developed that provides equivalent or better performance (i.e., signal intensities and signal-to-noise ratios) compared with standard array kits, and has a total processing time of less than 15 minutes. Glass slides were spotted with capture probes, and the microfluidic part with embedded fluidics passively adhered to the slides, which were then mounted in a centrifugal holder. Serial capillary valves were used to release sequentially the fluorescently labeled sample, a wash solution, and a rinse solution based on increasing spin speeds. The slides were then spun dry, the silicone microfluidic device was removed, and the fluorescence was scanned using a standard glass slide scanner.
The system could detect the differences between several Staphylococcus strains that differed only by as little as a single nucleotide. The signal intensities were 10-fold higher compared with passive systems, which verified that efficient hybridization did occur due to the gains in mass transport from the rapid, microfluidic flow-through system.
 [7] |
|
Figure 3. (click to enlarge [7]) Sampling of current microfluidic platforms designed to tackle the steps of nucleic acid diagnostics: a) Integrated benchtop system developed by Samsung for nucleic acid extraction from whole blood; b) Close-up of a single device on the integrated cell lysis and clarification CD; c) Schematic of the rapid, microfluidic PCR amplification card; and d) Schematic of the rapid microfluidic system for DNA microarray hybridization assays.
|
The microfluidic systems described above and the steps they accomplish are shown in Figure 3. Robustly integrating such platforms into a single microTAS for NA diagnostics is no small task. As mentioned above, these systems are only a sampling of many of the CD platforms that have been developed. Notably, many successful CD technologies have been produced by both industry and academia, such as the Institute for Microsystem Technology (Freiburg, Germany), Tecan (Mannedorf, Switzerland), Abaxis (Union City, CA), and Gyros AB (Uppsala, Sweden). However, many hurdles have yet to be overcome, including challenges inherent to centrifugal platforms.13
CD Challenges
While the CD indeed provides effective pumping and valving options to facilitate the integration of NA diagnostic steps, the method is not without limitations. For example, few of the CD valving schemes presented above act as vapor valves: they perform liquid valving only. Thus, liquid reagents could not be stored for long periods of time since evaporation and cross-chamber contamination could occur. In addition, high-temperature heating steps (e.g., PCR) would cause the sample to escape and evaporate, and would result in a failed step and possible channel clogging once vapors condense downstream. Such conditions require a valve that not only holds back liquids but also prevents the movement of vapors, similar to the ice and ferrowax valves in the systems discussed above.
Also, pumping on the CD still requires energy to create the centrifugal forces. Other pumping methods (e.g., manually-driven pumps or lateral-flow) do not require onboard power sources. Perhaps the largest obstacle of centrifugal pumping is that it is unidirectional: fluids flow only from the center of the CD radially outward. Properties of the fluids being pumped can affect fluidic behavior, especially when implementing capillary valves. Thus, designing and testing a CD from the outset using similar biological fluids and samples that will be used in the final disposable is essential. As mentioned in part one of this article, the test fluids used must represent both the average and extreme cases of relevant fluidic properties.
The unique advantages of the CD platform make it well suited for POC applications, such as rapid screening in physician offices, clinics, or the field. However, the CD platform is not well suited for high-throughput screening (HTS) applications. While HTS systems will retain their utility, the CD platform is a likely candidate to address the POC IVD market.
Conclusion
While the problems discussed so far are limited to CD platforms in particular, many broader issues will affect any microfluidic device designed for NA diagnostics. As covered in parts one and two of this article, an in-depth understanding of some of the biological and microfluidic issues facing microTAS systems for NA diagnostics has offered a better appreciation of such problems. The third article of this three-part series will discuss such challenges, along with solutions and recommendations for success. In addition, a nucleic acid IVD system for POC applications that relies on CD microfluidic technologies will be presented.
References
1. P St-Louis, “Status of Point-of-Care Testing: Promise, Realities, and Possibilities,” Clinical Biochemistry 33, no. 6 (2000): 427-40.
2. A Manz, N Graber, and HM Widmer, “Miniaturized Total Chemical Analysis Systems: A Novel Concept for Chemical Sensing,” Sensors and Actuators B: Chemical 1, no. 1–6 (1990): 244–248.
3. DJ Beebe, GA Mensing, and GM Walker, “Physics and Applications of Microfluidics in Biology,” Annual Review of Biomedical Engineering 4 (2002): 261-86.
4. DC Duffy, et al., “Microfabicated Centrifugal Microfluidic Systems: Characterization and Multiple Enzymatic Assays,” Analytical. Chemistry 71, no. 20 (1999): 4669-4678.
5. MJ Madou and GJ Kellogg, “The LabCD: A Centrifuge-Based Microfluidic Platform for Diagnostics, Systems, and Technologies,” in Clinical Diagnostics and Drug Discovery, ed. GE Cohn and A Katzir (San Jose, CA: SPIE, 1998), 80-93.
6. MJ Madou, Fundamentals of Microfabrication (Boca Raton, FL: CRC Press, 2002).
7. MJ Madou, et al., “Lab on a CD,” Annual Review of Biomedical Engineering 8 (2006): 601-628.
8. C Ko, “One-Step Pathogen Specific DNA Extraction from Whole Blood on a Centrifugal Microfluidic Device,” Lab on a Chip 7 no. 5 (2007): 565-573.
9. H Kido, et al., “A Novel, Compact Disk-Like Centrifugal Microfluidics System for Cell Lysis and Sample Homogenization,” Colloids and Surfaces B: Biointerfaces 58, no. 1 (2007): 44-51.
10. G Jia, et al., “A Low-Cost, Disposable Card for Rapid Polymerase Chain Reaction,” Colloids and Surfaces B: Biointerfaces 58, no. 1 (2007): 52-60.
11. G Jia, “Dynamic Automated DNA Hybridization on a CD (Compact Disc) Fluidic Platform,” Sensors and Actuators B: Chemical 114, no. 1 (2006): 173-181.
12. R Peytavi, et al., (2005). “Microfluidic Device for Rapid (<15 Min) Automated Microarray Hybridization,” Clinical Chemistry 51, no. 10 (2005): 1836-44.
13. J Ducrée, “The Centrifugal Microfluidic Bio-Disk Platform,” Journal of Micromechanics and Microengineering, 17, no. 7 (2007): 103-115. 5
Jonathan Siegrist is a PhD candidate near completion in the biomedical engineering department at the University of California, Irvine, and a co-founder of iGlyko Inc., a medical device company focused on in-hospital glucose monitoring. jsiegris@uci.edu [8]
Regis Péytavi, PhD, is a project leader at the Centre de Recherche en Infectiologie of Université Laval in Québec City, and the technological coordinator of a multidisciplinary team project focused on integration of nucleic acid assays with centrifugal microfluidics devices. regis.peytavi@crchul.ulaval [9]
Michel Bergeron, MD, is professor and chairman of the Centre de Recherche en Infectiologie of Université Laval in Québec City, and the founder of Infectio Diagnostic Inc., which is now BD Diagnostic-GeneOhm. michel.g.bergeron@crchul.ulaval [10]
Marc Madou, PhD, is a chancellor's professor in the mechanical and aerospace engineering department at the University of California, Irvine, and a pioneer in the fields of centrifugal microfluidics and microfabrication both in industry and academia. mmadou@uci.edu [11]