The desire by drug discovery researchers to conduct ADME assays earlier in the development process has led to a demand for the analysis of larger numbers of samples. Moreover, unlike the majority of assays commonly used in lead discovery and high-throughput applications, most ADME assays are not amenable to homogenous assay formats without significant compromises in data quality or biological relevance. The analytical tool of choice in ADME assays is mass spectrometry due to its ability to perform direct quantitative detection of a wide range of analytes.
The ability of mass spectrometry to perform direct and highly quantitative analysis is offset by several drawbacks. Mass Spectrometers (MS) are serial detection systems capable of analyzing only one sample at a time. Mass spectrometry is generally incompatible with solutions containing high ionic strength such as salts and buffers and requiring sample preparation, and it is a high capital- and operational-cost platform.
Traditionally, mass spectrometers have been coupled to liquid chromatography systems (such as HPLC) as a means of on-line sample preparation prior to analysis. While HPLC is an effective method of sample preparation, it is time-consuming, with cycle times ranging from several minutes to tens of minutes. Even with a fast, two-minute HPLC method, less than two 384-well plates can be analyzed in 24 hours. The high cost and low speed of LC-MS has led to much research into ways of increasing MS throughput.
Increasing MS Throughput
Improving the HPLC process itself is an obvious solution and several vendors have been actively involved in this area. Others have attempted to solve the HPLC bottleneck by designing parallel HPLC systems coupled to a single MS detector that allow for injections to be staggered, facilitating overlapping analyses. Another attempt to overcome the sample preparation bottleneck is an MS source that facilitates multiple HPLC systems being simultaneously coupled to a single detector.
While these approaches do have shortcomings such as complicated method development and decreased sensitivity, in the case of parallel MS, they enable faster throughput. Reports in the literature claim that throughputs under one minute per sample have been achieved in some applications, facilitating analysis of four 384-well plates in 24 hours.
Yet another solution to the MS throughput bottleneck is to decouple the LC from the analysis. A parallel sample preparation is performed, typically with a 96-well SPE plate. The time associated with the sample preparation is amortized over the 96-samples.
Several methods are available for the MS analysis of such samples, including flow-injection analysis. These approaches have unique strengths and weaknesses. Parallel LC or MS systems integrate sample preparation and analysis into a single automated step but provide only incremental increases in throughput. Decoupling the sample preparation facilitates much higher throughputs but puts the throughput burden on the sample-preparation step. Even with advanced robotics, running large numbers of 96-well SPE-plates per day is an expensive, labor-intensive, and error-prone process.
Ideally, high-throughput MS systems are highly automated and address all steps of analysis, including sample preparation, data acquisition, data analysis, and generation of reports for exportation to databases. Any method that doesn’t solve all steps pushes the throughput bottleneck on other steps.
RapidFire™ Mass Spectrometry
A novel solution to the challenge of achieving high-throughput MS analysis is the RapidFire™ system from BioTrove (www.biotrove.com). RapidFire Mass Spectrometry (RF-MS) couples micro-scale solid-phase extraction to a triple-quadrupole mass spectrometer. RF-MS is fully automated and performs integrated sample purification and MS analysis. Sample cycle times are as fast as five seconds per sample, translating to over 4,500 data points in eight hours (Figure 1).
The core of the RF-MS system is a fast yet highly effective SPE device capable of rapidly desalting and purifying samples in preparation for MS analysis. The proprietary SPE cartridge used in the RF-MS system has a small bed volume with high surface area and is optimized for high-throughput applications. The system is fully compatible with microsomal enzyme preparations and has been used successfully in applications involving cell culture experiments and tissue extracts.
The RapidFire system was developed as a high-throughput screening tool (HTS) to enable the identification of lead compounds against targets that are difficult or economically infeasible to screen with conventional technologies. Such targets include many enzymes in metabolic disease, anti-inflammatory, cardiovascular, and anti-infective/anti-fungal therapeutic areas. In typical HTS applications the system directly quantifies the substrate and product of the assay. While test compounds may be present, there is no need for their quantification. The activity of a test compound can be inferred from the relative amount of substrate remaining and product formed (Figure 2). In applications where it is not possible or desirable to monitor substrate and product simultaneously such as reactions where the substrate and product have different ionization or there is low conversion, the formation of product is normalized to an internal standard.
An example of RapidFire in HTS is a kinase assay. Quenched kinase assay plates are loaded onto the RapidFire system (capacity is 18 microtiter plates) without any sample preparation. The automated system serially analyzes each well and provides a quantitative output of the peptide and phosphopeptide concentrations in each well. Wells containing active test compounds can be identified as containing lower concentrations of phosphopeptide product and higher concentrations of peptide substrate.
While there are many commercially available solutions for screening kinases that have greater throughput, including many homogeneous assays, the advantage of RapidFire is that the approach to running a kinase assay is identical to any other protein-modifying enzyme such as a methylase, deacetylase, hydroxylase, phosphatase, ubquitinase, farnesyl transferase, or protease. Because simple homogeneous assays are not readily available for many enzymes in the latter classes, assay development and validation is a major bottleneck. With RapidFire, HTS assays can be developed and run quickly and efficiently in a label-free format using native substrates without requiring the use of radioactivity or antibodies.
Another application for the RapidFire system is P450 inhibition assays for performing drug-drug interaction (DDI) assays. The workflow for DDI assays is similar to an HTS application in that a drug probe that is specific for a CYP isoform is incubated with the enzyme (typically liver microsomes) in the presence of the test compound. Test compounds that are substrates or inhibitors of that CYP isoform attenuate the conversion of the drug probe to its metabolite. As in HTS applications, there is no need to monitor the test compound itself—the activity of the test compound can be inferred through the amount of probe metabolite formed (Figure 3). To date, 12 different drug probes have been validated against six CYP isoforms (CYP1A2, 2C8, 2C9, 2C19, 2D6, and 3A4) for running DDI assays with RapidFire.