Send to printer »

Feature Articles : Sep 1, 2010 (Vol. 30, No. 15)

Less Is More for Mass Spec Sample Prep

Goals Set to Increase the Throughput, Cost-Effectiveness, and Overall Analytical Efficiency
  • Vicki Glaser

A common thread wove through the diverse scientific presentations and new product launches at the recent American Society for Mass Spectrometry (ASMS) annual conference in Salt Lake City.

Researchers want to be able to acquire qualitative and quantitative information in one experiment performed on a single instrument platform with high resolution and high mass accuracy. Mass spec manufacturers are meeting that demand with an array of new systems that offer increased throughputs, innovative ion transfer technologies, and high resolving power.

An emphasis on time of flight (ToF) methodologies highlighted the growing interest in applying advanced ToF systems that offer high sensitivity and mass accuracy to do quantitative MS analysis. Another factor driving development in the MS arena is the new human metabolites in safety testing (MIST) guidelines that call for characterization and quantification of a parent compound and its metabolites. To save time and resources, drug developers want to be able to obtain this information in one MS run.

Several presentations at the conference focused on sample-prep strategies that can increase the throughput, cost-effectiveness, and overall efficiency of MS analysis. Panos Hatsis, an investigator at Novartis Institutes for Biomedical Research, described a typical 34.5-hour workflow for ADME sample analysis, beginning with the receipt of samples, batch creation, use of triple quadrupole MS for analyte tuning, 24 hours for sample analysis, followed by data analysis and data reporting.

“We need to make a positive impact on this time line,” said Hatsis. He identified the 24-hour sample-analysis time frame as the biggest bottleneck in the process and described the use of Biocius’ RapidFire® Mass Spectrometry System to perform solid-phase extraction (SPE), with “an ultrafast autosampler” that allows for completion of a method cycle in about seven seconds. Hatsis and colleagues evaluated about 200 compounds, comparing SPE and triple quadrupole MS/MS to conventional LC-MS/MS for generating data on the total ion concentration of an analyte and an internal standard.

This approach reduced the 24-hour sample-analysis time to 2 hours and the total 34.5-hour workflow to 12.5 hours. As a result, the process bottleneck shifted to the analyte-tuning process. To determine the need for compound tuning, the researchers replaced MS analysis on a triple quadrupole system with ToF MS, which provided higher resolution, increased sensitivity, and greater mass accuracy. Hatsis proposed that switching to a quadrupole ToF or orbital-trap MS system could eliminate the need for analyte tuning, further shortening the workflow.

Jon Williams, investigator at GlaxoSmithKline, spoke on the “Development of a Novel Sample Preparation Strategy for LC/MS/MS Quantitation of Serum Binding Domain Antibodies for PK Studies” at the meeting.

The use of LC-MS/MS to quantify biotherapeutic agents provides increased sensitivity and specificity and a wide linear range, with slightly less sensitivity compared to traditional antibody-based methods.

In contrast to pharmacokinetic (PK) studies to evaluate small molecule drugs, quantification of therapeutic proteins in plasma samples requires more demanding method development and multistep sample preparation, typically including an enzymatic proteolysis step to break down large proteins, sample clean-up and protein depletion using either SPE or electrophoretic methods, followed by a detection step.

Williams emphasized that method development for peptide multiple reaction monitoring experiments should focus on achieving high peptide recovery from the digestion and extraction process, selection of a peptide that is unique and not present endogenously at high levels, the formation of an intense fragment ion with a mass-to-charge ration (m/z) greater than the precursor ion, and it should not contain easily oxidized amino acids. Process development should also aim to identify putative cleavage sites in the peptide sequence.

He described pharmacokinetic (PK) studies performed on Exendin-4 (a peptide found in the saliva of the Gila monster that is homologous to mammalian glucagon-like peptide-1) chemically coupled to AlbudAb™, an antibody engineered to bind to albumin for the purpose of extending the half-life of the therapeutic agent.

Williams’ group devised a process stream in which proteolysis is performed before sample clean-up. One of the challenges in protein digestion is not to produce too many interfering peptides that can mask the signal of the target peptide.

Proteolysis can be controlled through enzyme selection, amount of enzyme, and reaction temperature and duration. The group used SPE for sample clean-up, optimizing the stationary phase, bed volume, and loading/elution buffers to enhance recovery. Their initial method yielded only 9% recovery efficiency.

By adjusting the digestion protocol they were able to reduce the proteolysis step from a 24-hour to a 70-minute digest that produced no background ion, and by increasing the SPE bed volume from 10 mg to 30 mg, they were able to trap more peptide, resulting in improved recovery efficiency to about 70%, Williams said.

“LC-MS/MS is a viable technology for PK analysis of biotherapeutics,” concluded Williams, but improved overall throughput is needed to compete with ELISA-based methods. In the future, ultrahigh pressure LC and multiplexed LC systems will help to achieve this.

Whereas ELISA is a better alternative for determining the concentration of free protein in biological samples, LC-MS/MS is preferable for measuring total biotherapeutic concentration. However, LC-MS/MS based methods may be used to differentiate between PK results obtained on the intact biotherapeutic and the metabolized portions of a biotherapeutic, which may not contain an epitope recognized by an antibody.

Erasmus Cudjoe, Ph.D. candidate at the University of Waterloo, Canada, spoke about “Automated Solid-Phase Microextraction in 96-well Plate Format: High Throughput Analysis and Ligand-Receptor Binding Studies.”

Solid-phase microextraction (SPME) yields “good sample clean-up that is compatible with MS,” said Cudjoe, and “is applicable to sampling of biofluids” such as plasma or whole blood. Additionally, it yields “information on free and total concentration in one experiment.” This technique also enables simultaneous analysis of multiple samples in parallel in a 96-well plate format, with a reuseable extraction phase.

Cudjoe described an open-bed, multi-fiber SPME sampling system that simultaneously performs multiple extractions on an automated platform that carries out all of the steps necessary for sample extraction, desorption, washing, and fiber preconditioning. Based on experiments using a benzodiazepine drug as a model to study system optimization and performance, Cudjoe reported “excellent reproducibility” and a sample prep time of about 75 minutes for 96 samples, without the need for extraction phase preconditioning.

An underlying principle of SPME is that the amount of analyte that can be extracted is directly proportional to the volume of the extraction phase. To minimize the process volume, Cudjoe and colleagues explored the use of thin-film technology in which the surface area to volume ratio of the extraction phase is increased, leading to an acceleration of the extraction rate (by about twofold) and shortening of the extraction time. The result has been greater sensitivity and higher absolute recovery.

Analyzing Dried Blood Spots

Dried blood spots (DBS) preserved on filter paper represent a complex matrix used in drug development and newborn blood screening. DBS are an easy way to collect, transport, and store blood samples for testing without the need for refrigeration.

Analyzing compounds in the blood sample can be challenging due to matrix effects caused by the complex makeup of blood and by the components of the paper. Joseph Stankovich, an intern in Gary Van Berkel’s lab at Oak Ridge National Laboratory, described an automated, high-throughput technique called liquid extraction surface analysis (LESA) using Advion BioSystems’ TriVersa Nanomate® nano-electrospray ionization (nano-ESI) MS system for analyzing drugs in DBS.

The success of MS analysis following direct liquid sample extraction from the DBS will depend on the solubility of the analyte and the ionization efficiency, as well as the characteristics of the surface, including the size, shape, topography, and hydrophobicity of the blood spot.

Several techniques have been used to facilitate liquid extraction of DBS, including coating the surface of the filter paper with silicone, creating wells on the surface (for example, by printing a wax pattern on the paper, followed by heating, causing the wax to form shallow wells), or punching out intact spots and putting the filter paper disks directly in the wells of a microtiter plate where the extraction then takes place.

Stankovich showed that all three of these methods yielded comparable results in terms of analyte quantification (in this instance of the drug propranolol), down to 10 ng/mL with an internal standard spiked directly into the liquid blood before spotting. Propranolol was quantified down to a level of 20 ng/mL with addition of an internal standard in methanol on top of the spots after the blood had dried—simulating a practical laboratory situation.

The third technique, punching out the DBS, was the only one that allowed the researchers to apply an internal standard after the initial application of dosed blood and to achieve acceptable accuracy and reproducibility.

Stankovich and colleagues concluded that LESA is a viable method for automated analysis of DBS, and that several techniques can be used to facilitate sample extraction. LESA enabled quantitative sample analysis in less than two minutes per sample.

Emerging from MIST

Joanna Pols, associate principal scientist at Merck Research Laboratories, presented a study on the development and optimization of an LC-MS method to separate the major metabolites of a drug candidate, as drug developers face the challenges presented by the MIST guidelines.

In her talk, Pols described the goal of a method-development study as generating a metabolite-rich matrix of 14C-labeled compounds that can be used for LC-MS method development and optimization.

Pols outlined several objectives of sample processing: removal of co-eluting/interfering endogenous components; quantitative extraction and recovery of the parent drug and its metabolites; spiking of 14C-labeled parent drug into blank matrix to assess the extraction recovery of the parent drug; and concentration of the sample to prepare it for LC-MS.

A variety of extraction methods can be used, the most common being SPE, solvent extraction with protein precipitation, and liquid-liquid extraction. Direct injection may also be possible depending on the body fluid being analyzed.

“Sample reconstitution after extraction is an often overlooked, critical step,” said Pols. Following extraction, the supernatants are combined and evaporated to near dryness under vacuum to achieve a final volume of 100–500 µL. Based on the results of stability experiments, Pols and colleagues determined that both the pH and concentration of biological matrices are important factors affecting degradation and that drying extracts under vacuum conditions in some cases is preferable to drying under nitrogen.

Once the samples are ready for metabolite profiling they can be analyzed using standard MS instruments such as triple quadrupole or Orbitrap (Thermo Fisher Scientific) systems, optimizing the processing conditions for the parent compound. Pols emphasized the value of starting with a robust, well-understood method that can then support all future metabolite-profiling studies throughout the development life cycle of a particular compound.