Send to printer »

Feature Articles : Jan 15, 2011 (Vol. 31, No. 2)

Overcoming Lab Automation Challenges

Flexibility Is Essential as There Is No Such Thing as One-Size-Fits-All Instrumentation
  • Ellyn Kerr

The bioindustry has, in its essence, been about translational medicine, and it is strictly the rise of high throughput (and, hence, automation) that has permitted its development. As several presenters at the upcoming “LabAutomation” meeting acknowledge, effecting the subtle refinements for successful automation of cell culture, experimentation, data collection, and analysis in translational medicine R&D requires some agility of approach and an understanding of complexity.

Gary Smejkal, research assistant at the Harvard Catalyst Laboratory for Innovative Translational Technologies and an affiliate assistant professor at University of New Hampshire’s Hubbard Center for Genome Studies, likens the demands for automation to aspects of a particular myth.

“There is no ‘bed of Procrustes’,” Smejkal asserts. “In Greek mythology, Procrustes was a host who offered guests a magical bed that would adapt to one’s body size. But the bed wasn’t magical: he would either stretch guests out on a rack or cut their legs, to make them fit. The moral in automation is that there is no one-size-fits-all, whether for reagents or platforms.

“A DNA hexamer translates to about 4,096 possible sequences. By comparison, if you take a hexapeptide that is six amino acids long, you have 64 million possible sequences. It’s less straightforward than DNA analysis,” he notes. Furthermore, proteins of potential pharmaceutical interest are often expressed in low abundance, confounded further by the broad concentration range over which proteins are expressed.

“In human plasma, the mass of albumin is nearly 10 billion times greater than that of cell-signaling proteins like the interleukins,” Smejkal observes. All this creates a challenge for effective isolation and collection. Much about any sample-prep strategy will depend on the particular downstream applications and whether the proteins of interest are membrane or cytoplasmic proteins, for instance.

Smejkal cites advantages of a Pressure Cycling Technology (PCT) sample-preparation system by Pressure BioSciences that applies an oscillating pressure to release nucleic acids, small molecules, and proteins “rapidly and reproducibly,” in a way designed to minimize the denaturing effects of thermal and chemical methods. The system thus offers reported benefits over bead beaters, sonicators, homogenizers, and mortar-and-pestle grinding.

The PCT platform comprises two main components: a Pressure BioSciences Barocycler® instrument, which alternates pressures from ambient to up to 35,000 psi, and single-use processing containers called Pulse tubes.

“Proteomic sample preparation is one of the areas for which automation has the furthest to go; strides are being made but we still need better automation for proteomics. We can’t afford 80% or 70% efficiency,” Smejkal insists. “We need to access all the protein constituents of a sample, even the low-abundance proteins, and do this without bias for some proteins—for example, higher-abundance ones—over others.”

Microchromatography

Control of flow rate and the capacity for quantification are among the key requirements for meaningful automation of sample preparation.

Scott Fulton, founder and CEO of BioSystem Development, highlights details of a platform that will be co-launched with Agilent Technologies at “LabAuto”.

The collaborative effort centers on BioSystem Development’s AssayMap® technology for high-throughput microchromatography, which is built around a cartridge device with a 5 µL volume bed of chromatographic resin (or other solid-phase support), designed to function either as a spin column or with an automated liquid-handling system. The cartridges function as miniaturized chromatography columns to allow a range of affinity-based or physical chromatographic separations and enzymatic reactions, says Fulton.

“We’ve combined Agilent’s Bravo automated 96-channel robotic liquid handlers with our AssayMAP microchromatography probe syringe technology, where the cartridge is coupled directly to small syringes mounted in the head of the liquid handler,” Fulton explains.

“These allow very precise control of flow rate, down to the microliters-per-minute range or even slower, in up to an entire microplate of 96 cartridges at once.” The result is an automated platform for “precise separations and enzymatic digestions” on disposable, 5 µL packed-bed columns in a 96-channel SBS standard microplate format.

“With the technology, we’re able to do applications like microplate ELISAs that normally might take 4 to 24 hours. With the packed-bed format and enhanced mass transport our technology affords, the binding reactions occur much faster. We can do the same ELISA on an entire plate in about 30 minutes, fully automated, without changing the reagents at all. This is critical for companies that have already invested into successful assay validation,” Fulton acknowledges.

Going forward, Agilent is to market protein-purification offerings built on the joint platform and expand the AssayMAP content portfolio. BioSystem Development will develop new chemistries and bring new “content partners” into the AssayMAP alliance, says Fulton.

The Medium Is the Message

Veit Bergendahl, Ph.D., head of R&D for cell culture technology at Miltenyi Biotec, will speak about the challenges of culturing live cells in the face of heterogeneity of living systems and the high variability inherent in assays, cell growth, and differentiation.

“The tall challenge for automation is to integrate tools for process control and software capable of feedback loops and scheduling. Most laboratory automation equipment was originally designed to fit the needs of chemical screening,” he notes.

“But the demands of biological material and cells often challenge current formats. Limitations become apparent in tasks like maintaining sterility, flexibility (optimizing for different cell lines or applications), accommodating biological buffers (which can often be abrasive to equipment), or delivering volumes larger than 5 mL.”

This past summer, Tecan and Miltenyi entered into an agreement based on Miltenyi’s MACS® magnetic-bead technology and Tecan’s Freedom EVO® liquid-handling platforms. Tecan has integrated Miltenyi Biotec’s MultiMACS™ Separator biomolecular and cellular separation devices onto the deck of the EVO instrument and has developed software to accommodate Miltenyi Biotec’s systems, including the autoMACS® Pro Separator, gentleMACS™ Dissociator, and MACSQuant® Analyzer.

Miltenyi Biotec’s philosophy is to explore bidirectional benefits of automation: to apply automation in support of primary cell culture (improving the quality of cells by bringing in standardization and reproducibility to automated processes), and to systematically develop new specialty media by using screens, “taking advantage of what we’ve learned with HTS and small molecule screening in order to develop optimal cell culture media, rather than the time-consuming task of discretely testing different compounds with known effects,” Dr. Bergendahl says.

“The big challenge is to come up with good assays that monitor quality or can verify that a particular component improves culture. To bring in automation to this development is not yet common.”

The importance of this is obvious when acknowledging the goal to avoid the uncharacterized nature of undefined serum components and preemptively realizing that “a lot of these techniques will eventually go toward clinical applications, so we try to develop media that have the potential to then be produced under GMP guidelines with the qualities necessary for clinical application,” says Dr. Bergendahl.

Separations and Detection

Yolanda Fintschenko, Ph.D., director of sales, marketing, and new technologies at LabSmith discusses her work on an automated modular interface for microfluidic separations and fluorescent detection. She conducted this work with the team of Holger Becker, Ph.D., co-founder and CSO at microfluidic ChipShop. The two concur with the problems of variability and range as mentioned by Harvard’s Smejkal.

“Current challenges in automation are very application-dependent. For R&D, one of the greatest challenges is having both automation and flexibility. Particularly in the field of microfluidics, researchers and product developers alike go through a fairly long stage of breadboarding their systems,” Dr. Fintschenko says. “However, while transient, this phase is extremely important, with reproducible and accurate data required to move on to the next step. Automation of fluid and sample manipulation, particularly outside the chip where necessary, can be helpful, but it also needs to be flexible.”

Dr. Becker concurs. “One of the main challenges lies in the varying requirements (volumes, flow rates, etc.) from the application side that demand a high degree of flexibility from any automation solution. Furthermore, a lack of standards in the microfluidics world continues to aggravate this problem.”

In November 2010, the companies announced an agreement under which LabSmith will distribute select microfluidic ChipShop chips that are compatible with LabSmith’s CapTite™ microfluidic interconnection products. The idea is to facilitate researchers’ building and rebuilding of microfluidic circuits.

The microfluidic ChipShop chips are also being sold as part of LabSmith’s LabPackage microfluidics laboratory kit that includes LabSmith’s HVS448 high-voltage sequencer, SVM340 synchronized video microscope for real-time visual monitoring and video capture, and SPS01 programmable syringe pumps for volumetrically controlled flow rates, in addition to the CapTite components.

“Our approach is to simplify and standardize tools by focusing on modular components and a breadboard platform designed for ease of use, as well as control of fluid connections, pressure-driven flow, and ease of high-quality visualization,” Dr. Fintschenko says. “Our partnership with microfluidic-ChipShop highlights this approach by allowing us to provide off-the shelf chips that are easily connected to and controlled by these modules.”

Dr. Fintschenko believes that the field of “microfluidics, in particular, has become quite mature. This means we have to continue to understand how the automated control of small volumes, pressure, and electric fields are being taken advantage of by scientists of diverse backgrounds who may not be interested in microfluidics, per se. These scientists want to apply off-the-shelf tools to automate measurements that are very difficult to make in any other way.”

“The message should be that using microfluidics to solve an application doesn’t require a degree in rocket science,” Dr. Becker adds. “We see a significant development in the commercialization of microfluidics as a true enabling technology for analytical sciences, drug discovery, biology, and diagnostics. Some of the early promises, like speeding up the analytical process or integrating a complete analysis on a chip, are becoming reality, and this is very exciting.”

Sample Prep and Target Analysis

As with R&D applications, the challenges in point-of-care molecular diagnostics lie in preparation of biosamples, target analysis, and “providing users with comprehensible readouts,” explains Jeff Tza-Huei Wang, Ph.D., assistant professor of mechanical and biomedical engineering at Johns Hopkins University’s Whitaker Biomedical Engineering Institute. “All of this involves multiple, separated processes done with a number of instruments that, to date, despite the high degree of automation for individual tasks, still require operator intervention to transfer samples between different equipment at various stages.”

Dr. Wang’s research team has aimed to develop an “entire sample-to-answer process in a fully automated fashion, with the high degree of parallelization desired for the great number of samples that often need to be tested in a single study,” he says.

They have developed a droplet-based micro total analysis system (μTAS) that, according to Dr. Wang, is most beneficial in its simplicity.

“No complicated microfluidic component is required, since the droplet itself functions as the fluid-containment compartment as well as the transportation unit. With magnetic particles, the droplet is actuated with an external magnet. The functionalized magnetic particles then serve as the substrate for biomolecule adsorption, allowing solid-phase-based sample preparation in the droplets.”

This simplification affords integration of multiple processes within a single µTAS platform that, when combined with a miniaturized sample-handling stage and fluorescence-detection module, allows, as Dr. Wang describes, “true sample-in/answer-out capability.”

“Droplet-based microfluidic systems have been emerging at a fast pace in recent years. Numerous studies that focus on various aspects of droplet microfluidics have been reported,” he says. “Nonetheless, only a limited number of groups have looked deep into the potential of droplets as a great platform for automated biosample analysis.”

Automation 101

  • Translational medicine R&D and diagnostics require flexibility of automation solutions that target a range of specific challenges for culturing cells and isolating and characterizing proteins.
  • There is a need for, and movement in automation toward, integration of platform capability, from cell culture and sample preparation through to analysis.
  • Industry collaborations continue to bridge innovations in handling biological materials, with developments in instrumentation.