Send to printer »

Feature Articles : Nov 15, 2012 (Vol. 32, No. 20)

As Sample Prep Improves, So Does Mass Spec

  • John Russell

Though unglamorous, sample prep is widely regarded as the most critical step in performing robust mass spectrometry.

Any number of problems—contamination, incomplete digestion, unwanted side reactions, technician error, etc.—may arise during sample prep to undermine a mass spec run. Recently, advances in reagents, protocols, and automation are helping mitigate sample-prep problems, speeding throughput, and improving reproducibility.

Detergent removal is a case-in-point. Used to isolate proteins from various sample types, detergents (e.g., SDS, CHAPS, Triton X-100, NP-40) are notoriously incompatible with mass spec (MS) analysis. Successful removal of detergents with good protein or peptide recovery, especially for low-abundance proteins, is critical for accurate MS analysis.

Earlier this year, at the ASMS conference in Vancouver, Thermo Fisher Scientific presented a new approach for detergent removal.

“Our goal was to identify a resin or workflow that would allow removal of detergents that would be detergent independent so it wouldn’t just work with anionic detergent, for instance, but would also be compatible with any sample type,” said John Rogers, Ph.D., manager R&D, mass spec reagents.

Thermo identified a resin with a strongly hydrophobic pocket; detergents bind within the pocket but the pocket is small enough that it excludes proteins or peptides.

“What’s unique about this product is it works with either protein or peptide samples. The advantage is you can perform a digestion with a complex sample in the presence of detergents and then easily remove the detergent,” added Babu Antharavally, Ph.D., senior scientist.

Thermo reports the method successfully removed >95% of nonionic, ionic, and zwitterionic detergents from 0.5−1% solutions, with high recovery of proteins or peptides. Tandem mass spectrometric analysis of 2.5−10 micrograms of BSA enzymatic digests at 25−100 microgram/mL—prepared in the presence of detergents and processed to remove detergent—revealed sequence coverage and MASCOT scores as good as or better than control BSA samples processed without detergent.

How Good Is Your Trypsin?

Trypsin digestion is a common step in protein sample prep, but there are few standardized methods for determining trypsin efficiency. JPT Peptide Technologies has developed a kit to monitor trypsin efficiency and showed supporting work at ASMS.

“When you do a trypsin cleavage there’s always an efficiency question. How much do you add? At what temperature should the reaction run and for how long? What are the buffers,” said Ulf Reimer, Ph.D., head, R&D. “Some people do it overnight in the cold room. Others do it at 37º. But there’s no easy way to measure trypsin cleavage efficiency.”

JPT is one of few peptide providers using a paper-based spot synthesis instead of beads, noted Dr. Reimer. Part of JPT’s manufacturing process involves putting a trypsin cleavage tag on peptides enabling JPT to accurately quantify production. The same principle is used in the kit.

“To quantify the completion of the trypsin, you would put the peptides in this kit —several different target peptides (different lengths and sequences) with different tags on the N-terminus—in with your sample. On the basis of recovery of the cleaved controlled peptides you can determine the efficiency of the trypsin in your prep process,” said Dr. Reimer.

He cites the kit’s advantages as accurate in situ determination of trypsin cleavage efficiency and reproducibility; compatibility with routine sample digestion workflows; analysis using standard LC-MS protocols; and absolute quantification of cleavage products in the samples using heavily labeled tags. Kit availability is planned before the end of the year.

Miniature HPLC Columns

Automating sample prep in modest scale settings has proved difficult because of cost (equipment) and skill (automation expertise) requirements. Agilent Technologies presented an approach it calls “Rapid, Automated Sample Preparation for Peptide Mapping,” based on the company’s AssayMAP Bravo platform. The result, said Agilent, is improved reproducibility, robustness, and increased throughput.

“We have developed a disposable cartridge that has a 5-microliter packed bed in it and along with technology for packing any resin in there with high chromatograph efficiency. These are like miniature disposable HPLC columns,” said Scott Fulton, MS bioengineering, and head, AssayMAP operations and workflow development.

Built into the robotic head is a probe syringe and when a cartridge is connected, it provides positive displacement flow control so there’s no air gap. “We are able to control the flow at an appropriate rate for running through this tiny packed bed,” said Fulton.

Typical flow rates are on the order of 2-to-10 microliters per minute, “so it’s really slow and we’re able to do that using 96 little syringe pumps operating in parallel.”

A technician is still required for some steps. The platform is “not hugely expensive but it has some limited deck space, and in between the digestion and purification steps you might need to go in and spend two minutes putting different plates on the deck and moving a couple of things around,” according to Fulton.

Binding is quantitative, he continued, producing a high product yield from a small volume. He said the approach is aimed at applications such as biomarker validation efforts “where you need to run something on the order a thousand patient samples to do a study and you really want to do that campaign in a week or two.”

Accelerating Throughput

Shimadzu demonstrated the ability of its Perfinity workstation, which it describes as an integrated online protein sample-preparation platform combining the selectivity of immunoaffinity, the speed of online digestion, and the resolving power of LC/MS.

“Performing these steps off-line requires as little as 8 hours, but frequently greater than 24 hours. With the Perfinity workstation the full process can be completed in as little as 30 minutes,” said Rachel Lieberman, Ph.D., applications scientist.

Key to performance is use of the Perfinity immobilized trypsin column where the equivalent of a 24-hour digestion can be obtained in 60 seconds.

“The application of mass spectrometry to clinical samples depends upon mass spectrometry outperforming ELISA tests,” noted Dr. Lieberman. “The detection of isoforms is one such example. Hemoglobin is an often-sited example of where a single amino acid variation is of significant clinical relevance. As such the hemoglobin variant study was performed as a proof of concept for the demonstration of the system.”

While cycle times can be as little as seven minutes for certain applications, in cases where extremely large numbers of samples (≥ 200) must be run in less than 24 hours the end user is better off using a plate-based technique, said Dr. Lieberman. Furthermore, the current configuration of the Perfinity workstation operates at analytical flow rates and it is not suited for lower abundant biomarkers.

Having cut digestion time down to as little as one minute the pretreatment (reduction and alkylation) steps comprise the new bottleneck.

“Our current goal is to dramatically reduce the impact of these steps. Improvements that enable increased sensitivity and higher throughput are also in progress,” pointed out Dr. Lieberman.

Improving Optimization

Used properly, statistical DOE (design of experiments) techniques can be more efficient and produce better results than OFAT (change one process factor at a time) in many applications. Tecan discussed a study in which DOE, automation of sample preparation, and multiparameter LC-MS/MS analysis enabled efficient and reproducible optimization of sample prep and yield for immunosuppressant drugs simultaneously.

In the study, a half factorial design was employed with five parameters at two levels, where a total of 15 parameters were screened in three sessions (five parameters each). The critical factors identified proved to work well (e.g., increase of signal in this study).

Although powerful, manual execution of DOE series is error-prone and tedious. To mitigate this problem, Tecan used an automated liquid handler (Freedom EVO®) driven by flexible software (Freedom EVOware®) that can pipette according to dedicated scripts using input for variables from Excel tables.

“In such a table, the concentrations defined in the DOE software can be converted to liquid volumes to be pipetted, if appropriate stock solutions representing all factors are defined,” explained Roland Geyer, Ph.D., innovation manager, Tecan.

Another key piece of technology used is Tecan’s AC extraction plate, still in development, which facilitates sample cleanup by a modified type of Liquid-Liquid extraction.

“It consists of a deep-well plate with wells coated with an immobilized liquid used as the extraction substrate. The coated-well surface as such enables simplification of the entire sample-preparation process. The procedure works without any protein precipitation, filtration, centrifugation or solvent evaporation, hence accelerating and streamlining the complete sample-preparation workflow,” said Dr. Geyer.

New Interface for LC-MALDI

Development of MALDI-TOF/TOF instruments with acquisition rates up to 1,000 Hz, has made LC-MALDI-MS a viable alternative to LC-ESI-MS for many proteomic and peptidomic applications. Despite this, LC-MALDI-MS is not nearly as widely used, said Johan Gobom, Ph.D., associate professor, University of Gothenburg, Institute of Neuroscience and Physiology.

“We believe an important reason for this is the robustness and ease of use of the LC-MALDI interface,” he said.

Dr. Gobom and colleagues have developed a LC-MALDI interface based on solenoid-valve microdispensing jets for contact-free deposition of analyte and matrix on the MALDI sample plate. This overcomes some of the problems associated with the currently available LC-MALDI interfaces.

Example: use of syringe pumps to tee in the matrix solution, and fractionating by contact deposition are prone to interruptions caused by matrix crystallization in the interface and low tolerance to deviations in the x-y-z calibration of the fraction collector.

Contact-free sample fractionation for LC-MALDI is likely to be adopted quickly and widely among LC-MALDI users, said Dr. Gobom. The solenoid-valve based technology used in this work is inexpensive, and the cost of the dispensing robot is similar to that of an ordinary LC-MALDI spotter.

Plus, the new interface increases the walk-away time of LC-MALDI; a technician can load a sample tray on the LC and leave the system to fractionate the samples unattended for several days.

The described LC-MALDI interface was developed in a collaborative project between University of Gothenburg, Sweden and M2 Automation. A LC-MALDI fractionation robot based on the developed technology has been commercialized by M2 Automation.