Protein precipitation with 96-well plates

The study concluded that Once wash steps are optimized, samples prepared by solid phase extraction are cleaner than those prepared by protein precipitation. Samples prepared by extraction with a Multi-SPE plate resulted in lower LOQs than samples prepared by solvent precipitation. Drug recoveries were acceptable (>80%) for both the SPE and the solvent precipitation methods. Well-to-well reproducibility of samples was slightly better with extraction with a Multi-SPE plate. Evaporation and reconstitution, while more time-consuming, yield better chromatographic performance, allow analysis of lower concentration samples, and require optimization for good analyte recovery. 

Typical protein precipitation procedures use one volume of plasma plus three to six volumes of acetonitrile or methanol (or a mixture) with the internal standard at an appropriate concentration for the assay. Poison et al.102 reported that protein precipitation using acetonitrile eliminates at least 95% of the proteins after filtration or centrifugation, the supernatant can often be directly injected into the HPLC/MS/MS system. Usually this step is performed using 96-well plates that are ideal for semi-automation of sample preparation. Briem et al.103 reported on a robotic sample preparation system for plasma based on a protein precipitation step and a robotic liquid handling system that increased throughput by a factor of four compared to a manual system. 

Hopfgartner et al. (2002) compared ternary column online SPE LC/MS and TFC with offline 96-well plate SPE LC/MS to quantitate three drug candidates in human plasma. A protein precipitation step was performed before the SPE LC/MS. Dual trapping columns (YMS AQ, 10 x 2.0 mm, 5 /tm) were used with an analytical column (Intertsil Phenyl, 50 x 2.1 mm, 5 /tm). The run cycle was 3 min calibration range was 0.2 to 250 ng/mL. The run cycle was 2 min with a calibration range of 5 to 1000 ng/mL for TFC. Offline SPE LC/MS achieved the same calibration range with a run time of 2 min. 

An example for an automated stability test in plasma is described by Linget and du Vignaud (1999). Incubations are performed on a 215 Gilson liquid handler. Incubation was done at substrate concentrations of 50 pM on 96 deep well plates. Each incubation tube contained 375 pL of a 200 pM test compound solution (in 0.1 M Tris buffer with 3% BSA, added to assist dissolution of compounds with poor solubility) and 1125 pL of plasma. Samples are taken after incubation times of 0, 1, 2, 3, 4 and 5 min. At each of these time points an aliquot of the incubation mixture was transferred from the incubation tube into a well in a 96 deep well plate containing an equal volume of acetonitrile for quenching by protein precipitation followed by centrifugation of the plates. Supernatants were analyzed by HPLC for metabolic screening. 

A potential problem with the analysis of biological fluids, especially plasma samples from in vivo studies, is the risk of clogging of SPE cartridges and/or analytical columns. Therefore, filtration of such samples prior to LC or SPE is recommended. Combined filtering and protein precipitation in 96-well plate format was described [97-98]. The samples are collected and stored frozen in sealed 96-well polypropylene filter plates. Prior to SPE and LC-MS analysis, the seals are removed and the plate is placed on top of a 96-well SPE manifold As the plasma thaws, it passes through the filter and into the SPE device. 

Protein precipitation was automated into a 96-well plate format by means a robotic liquid handler by Watt et al. [101]. The procedure is described in more detail in Ch. 1.5.1. It enabled a 4-fold improvement in sample throughput on the LC-MS instrument, compared to previous manual protein precipitation procedures. The method was applied as a generic sample pretreatment method in combination with a generic LC-MS method to a variety of drug candidates [102]. 

Biological matrices are not directly compatible with LC-MS analysis, since these samples tend to block LC columns and contaminate the ion source. Extraction of compounds of interest from biological fluids is required prior to LC-MS/MS analysis [20]. Sample extraction can be achieved off-line with protein precipitation (PP), liquid-liquid extraction (LLE), or solid-phase extraction (SPE) [21]. With the ease of use and sophistication of automated liquid-handling systems, sample extraction procedures in a 96-well format can handle microliter volumes with multiple sorbents per plate and can simplify and expedite SPE method development [22,23]. The technique can be used to routinely develop methods for multiple analytes and examine a set of eluent compositions for each analyte [16]. 

Variations on the vapour diffusion method have met with considerable success. A solution of the protein containing a salt concentration approximately 10% below that needed for precipitation is equilibrated by vapour diffusion with a larger volume of more concentrated salt solution which is only slightly below the concentration needed for precipitation. With non-volatile precipitants water distils out from the protein solution to the reservoir. With volatile solvents, distillation and equilibration will proceed in the opposite direction. The hanging drop version of this method allows numerous trials of different conditions with very little protein material. Plastic tissue culture plates (for example, with 24 cylindrical wells of 2 ml volume) may be used. The precipitant solution (1 ml volume) is placed in the wells. These are then sealed with a coverslip onto which a drop of protein solution (5-20 pi) has been placed and then inverted. A drop of light oil on the rim of the well makes for an air-tight seal. The method allows ready inspection of the drops without disturbing them. 

Many different sample preparation techniques are available to the drug discovery scientist. Off-line sample preparation procedures include protein precipitation, filtration, dilution followed by injection, liquid-liquid extraction (LLE), and solid-phase extraction (SPE). Typically, these procedures are performed in an automated, high-throughput mode that features a 96-well plate format. Online sample preparation procedures include SPE and turbulent flow chromatography (TFC) with conventional chromatographic media or restricted access media (RAM). These online approaches are often simple and easy to automate. 

Experimentally, high-throughput microsomal stability assays are conducted by incubating compounds (0.5-5 iM) at 37°C with resuspended liver microsomes in buffer along with NADPH, usually in a %-well plate format. Aliquots of the reaction mixture are taken at predetermined intervals and quenched with organic solvent to stop the reaction as well as precipitate microsomal proteins. The quenched mixture is then centrifuged and the supernatant is removed for LC-MS analysis. SRM is typically used to monitor the disappearance of the parent compounds, and the results at each time point are expressed in the form of percent of compound remaining as compared with time zero (TO). 

An LC-MS-MS method measures directly the concentration of amantadine (1-adamantylamine, used for treatment of influenza, hepatitis C, parkinsonism, and multiple sclerosis) without protein precipitation, centrifugation, extraction, and derivatization steps. Only 50 p,l sample is needed. Internal standard is l-(l-adamantyl)pyridinium bromide. The serum sample is diluted by water in a 96-well plate. The chromatographic separation is performed using an eluent of isocratic water/acetonitrile (60/40, v/v) with 5 g/1 formic acid on a C8 column. Run time is 3 min. Electrospray atmospheric pressure ionization, positive ion, and selective reaction monitoring mode were used. Detection hmit 20 mg/1, linearity 20-5000 mg/1, intraassay/interassay coefficient of variation <6%/<8% recovery 99-101%. 

---