by Jeff Perkel
For every repetitive laboratory process, there comes a point when you must ask whether the cost savings associated with doing the thing manually exceeds the value of the time spent actually doing it.
Consider MALDI sample prep. There are two primary proteomics workflows that lead to MALDI mass spectrometry. In one, proteins are separated electrophoretically (on either a one- or two-dimensional gel), protein spots are picked from the gel, digested into peptides, eluted from the gel plug, mixed with matrix, and finally spotted onto a MALDI target plate for mass spec analysis. The other workflow is simpler, substituting a liquid chromatographic separation for the gel.
Both processes may be run manually. But manual sample processing is tedious, error-prone, and can lead to user-to-user variability. For small labs, those caveats are perhaps a mere nuisance. But for larger labs, or for core facilities and industrial labs with substantially larger workloads, such limitations can have significant consequences.
Fortunately, robots exist to automate every step along the path from gel to mass spectrometer, as well as to spot LC fractions directly onto a MALDI target plate (which is used to store the column profile for off-line analysis instead of analyzing it on-the-fly). Such automation doesn't come cheap – expect to spend upwards of $150,000 total for an automated spot picker, digestion and sample preparation robot, and MALDI spotter. LC sample spotters are less expensive, but of course, you will also need the LC system itself.
Though a variety of options exist, robots for automating the gel workflow all perform more or less the same tasks.
A spot-picker
images a gel, and then either automatically (based on preset criteria) or manually, produces a "hit list" of spots to core out of the gel and place into an on-board microtiter plate. A processing robot digests the proteins in each plug with protease, elutes the peptides from the gel, and transfers them (either by vacuum or pipetting) to a secondary plate. Spotting robots (whether LC-coupled or not) then combine these samples with MALDI matrix
and transfer them to target plates.
Often, but not always, these instruments can interface with other manufacturers' software and consumables. Many spotting robots, for instance, are "manufacturer agnostic," meaning they can spot to any type of target plate (though they may require a special adaptor to do so). Some gel spot-pickers will allow users to import hit lists generated by other imagers, as well.
The Proteomics/Mass Spectrometry Facility at thePenn State College of Medicine has robots to automate both gel- and LC-based MALDI workflows, according to Bruce Stanley, the facility's director. These include a Tempo LC MALDI spotting system
from Applied Biosystems/MDS Sciex for spotting fractions coming out of a nanoLC column, as well as spot-picking robots from Leap Technologies
and GE Healthcare.
What the facility does not automate, however, are the in-gel digestion and elution steps, because in Stanley's experience, manual processing yields more, and higher confidence, protein identification: 70% to 95% when done manually, as compared to 60% to 70% when done using robots.
"It's a philosophical question of how much time you want to put in on the front end, versus how much information you want out on the back end," he says.
If you decide to use robotics, you can automate as much or as little as you want. Some robots, for instance, do it all. Leap Technologies' 2D IDx is an all-in-one system that automates the entire gel-based process, including gel imaging, spot picking, proteolytic digestion, peptide elution, and spotting. So does Shimadzu Biotech's Xcise.
Other systems have more discrete functionalities. Dedicated spot-picking robots include the PROTEINEER spII
from Bruker Daltonics, the Ettan Spot Picker from GE Healthcare, the ProPic II
from Genomic Solutions, and the EXQuest
from Bio-Rad Laboratories.
Downstream gel-processing systems include the Bruker's PROTEINEER dp (in-gel digestion, elution, and plate spotting), the Ettan Digester and Ettan Spotter from GE Healthcare, the standalone ProGest
digester and ProMS spotter from Genomic Solutions, the multifunctional ProPrep
(a higher throughput digestion, elution, and spotting instrument), also from Genomic Solutions, and the TOFPrep
plate spotter from PerkinElmer Life & Analytical Sciences.
And for those interested in automating the LC-MALDI workflow, options include Bruker Daltonics' PROTEINEER fc, Eksigent Technologies' nanoLC-MALDI spotting system, and Leap Technologies' PAL (prep-and-load) system.
Rather than buying off-the-shelf instruments dedicated to a particular application, some researchers prefer more versatile liquid handling robots, which can be reconfigured for other applications should the need arise. PerkinElmer Life & Analytical Sciences' 8-tip JANUS Varispan, for instance, is a multifunctional liquid handler preconfigured for in-gel digestion, elution, and spotting.
But, says Gary Reznik, Applications Science Group Leader for Automation and Liquid Handling at PerkinElmer Life and Analytical Sciences, the robot can also be used for any liquid-handling task, as well.
"Instead of being limited to a single application, the JANUS is versatile and can be reconfigured," he says, for instance, to handle new plate formats, additional robotic arms, on-board options like vacuum manifolds and shakers, as well as integration with other systems. "JANUS can have 28 different configurations, not counting options and integration," Reznik says.
With all the available options, the real question potential users must ask is, do they really need automation?
Obviously throughput is a big part of the answer. Stanley says that from his perspective, cutting and digesting 10 to 20 spots per day is a reasonable manual workload. "If I were pushing the gels, doing 50 to 100 spots per gel, I'd definitely use robotics. The area in between is hard to decide."
According to Gordon Nye, Proteomics Product Manager at Leap Technologies, for many labs throughput alone does not justify automation– even if they run a dozen two-dimensional gels at a time, they typically pick spots from only one or two of them.
But there are other reasons to automate. Robotics also ensure user-to-user reproducibility, streamline lab processes and data analysis, and cut down on contamination (most robots are enclosed to keep out keratin, for instance).
"What gel-cutting robotic systems can do is to eliminate errors and the tedium of manually cutting gels," Nye says. "It makes life in the lab more enjoyable."
