One of the chief limitations to LC-MS/MS is that it is a serial technique. While typical gradients are on the order of a few minutes, the dwell time for an individual SRM channel is typically only 50 ms. Consequently, several reported methods have attempted to take advantage of this temporal disparity. In one example, Korfmacher et al. combined the postcolumn effluent from two LC systems to a single MS, which performed alternating detection of the analytes and internal standards injected in the two respective samples . While stacked injections have been used for a number of years with isocratic systems, Van Pelt and co-workers published results from a parallel gradient system . The method involves simultaneous injection of four samples with a multiprobe autosampler. Although injection occurs simultaneously, the arrival at the MS is staggered by plumbing a different gradient delay for each sample.
Another approach that has received great attention is the use of parallel chromatography coupled to a single MS equipped with multiple indexed ESI sprayers. A commercially available system, referred to as multiplexed ESI or MUXTM, allows up to eight ESI sprayers to be consecutively sampled, each connected to an individual LC column. Sampling occurs by mechanical means, with a rotating device that occludes all but a single sprayer at any given instant.
Yang and co-workers demonstrated the viability of MUX for bioanalysis, and validation for loratadine and its metabolite descarboethoxyloratadine was performed in four preclinical species with a four-sprayer MUX interface . Other examples that have been reported have been more closely linked to drug discovery. For instance, Bayliss et al. used a four-sprayer MUX in conjunction with TFC to achieve an overall throughput of 120 samples per hour . With a somewhat similar design, Deng and co-workers used a parallel four-extraction column TFC system in combination with a four-sprayer MUX interface . In this system, CS was used to couple an extraction column with an analytical column for each channel. Methotrexate was used as a test compound, and this technology achieved successful cross-validation with results obtained with a more conventional LC-MS/MS method. In previous work, this group demonstrated improved chromatography with a dual-column approach that was applied to cassette dosing .
The combination of MUX with monolithic LC has also been investigated as a means to increase bioanalytical throughput. Deng and co-workers used four monolithic C-18 columns (4.6 x 100 mm) with a four-sprayer MUX interface to analyze twelve 96-well plates in 10 hours . In this example, the analyte, oxazepam, was chromatographically separated with a run time of 2 minutes to yield an overall throughput of 30 s/sample. Note that this rate did not include sample preparation, which was performed with automated SPE with human plasma as the matrix.
Despite the throughput provided by the MUX interface, several practical considerations should be noted. First, the MUX system cannot be used with flow rates that exceed 0.1 mL/min. In addition, a small amount of carryover exists between adjacent sprayers, typically on the order of 0.1%. While each of these issues can be tolerated, the overhead associated with sampling multiple effluent streams limits the number of points that can be acquired across a chromatographic peak. In a four-channel MUX system, each sprayer is accessed about every 1.2 seconds . Thus, to permit the acquisition of 10 data points, the LC peak must be at least 12-seconds wide. Because typical drug-discovery runs include more than one analyte, as well as an internal standard, the MUX system is not truly compatible with fast LC methods that typically produce peak widths less than 6 seconds wide.
Because of the serial nature of LC-MS, much of the current discussion has centered on ways to reduce LC-MS/MS cycle time. Often the rate-limiting step for drug-discovery bioanalysis lies in the speed with which methods (sample preparation and chromatography) can be prepared for NCEs. One of the frequently overlooked steps is the need to tune and optimize MS/MS transitions for the various analytes studied. Fortunately, most MS vendors now offer semi- or fully automated procedures to perform this task. The origin for these procedures can be traced to the seminal work of Whalen et al., who published an automated procedure known as AUTOSCAN . It is possible to establish experimental conditions with this procedure in the flow-injection analysis (FIA) mode for 96 analytes in less than one hour.
Was this article helpful?