High-Throughput Screening Applications for Enantiomeric Excess Determination Using ESI-MS

In the past decades, high-throughput screening (HTS) applications have been established in many fields of drug development. Thus, suitable qualitative and quantitative analytical methods have become an important research topic. Since more than 80% of drugs and active ingredients are chiral substances,1 the development of fast and reliable analytical methods for the determination of the enantiomeric excess (ee%) is required for improving drug efficacy and reducing drug development costs. Common analytical methods for the ee% determination of chiral compounds include HPLC, GC, and capillary electrophoresis (CE). Due to their long analysis times and the use of chiral columns and materials, these analytical techniques cannot be used in high-throughput experimentation.

Traverse et al.2 reviewed several techniques for the determination of ee% in HTS, including mass spectrometry. Lindner et al.3 gave an overview on the use of MS for ee% determination. A multitude of different methods, such as host–guest complex formation,4,5 ion–molecule reactions,6 collision-induced decay of diastereomeric complexes,7 and parallel kinetic resolution with mass-tagged auxiliaries, have been established.8–13 In order to meet the requirements of high-throughput systems, it is necessary to combine sample preparation, analysis, and data evaluation in a versatile and easily adaptive processing system with a minimum of manual work. In this study, various substrates and auxiliaries from different compound classes were investigated and used in HTS systems.

Parallel kinetic resolution for ee% determination

Figure 1 - General principle of parallel kinetic resolution.

Reetz et al.8 and Finn et al.9,10 described a method for the ee% determination of chiral catalysts using parallel kinetic resolution and mass spectrometry. A chiral substrate was derivatized with two pseudoenantiomeric mass-tagged auxiliaries. Due to the mass differences of the auxiliaries used, usually four products with two characteristic masses are formed in this derivatization process (see Figure 1 for the general principle of parallel kinetic resolution). Because of the different reaction kinetics of the enantiomers and auxiliaries, various quantities of the products result. The ratios of the integrated peak areas of both masses are characteristic for each ee% value of the substrate. Figure 2 shows an example with two characteristic peaks for the determination of (–)-carbobenzyloxy-L-proline (100ee%). Finn et al.9,10 reported the analysis of secondary alcohols and primary and secondary amines as chiral substrates. (S)-N-benzoyl-2-pyrrolidinecarboxylic acid, (R)-N-(p-toluoyl)-2-pyrrolidinecarboxylic acid, and their correlative “switched” versions were used as mass-tagged pseudoenantiomeric auxiliaries. In this study, the method was extended and optimized for further compound classes.

Figure 2 - Mass spectrum of derivated (–)-carbobenzyloxy-L-proline (100ee%).

Figure 3 - Investigated chiral substrates: a) (–)-carbobenzyloxy-L-proline, b) (+)-carbobenzyloxy-D-proline, c) (S)-(–)-1-carbobenzyloxy-2-piperidinecarboxylic acid, d) (R)-(+)-1-carbobenzyloxy-2- piperidinecarboxylic acid, e) (S)-(+)-2-heptanol, f) (R)-(–)-2-heptanol, g) (S)-(+)-2-octanol, h) (R)-(−)-2-octanol, i) N-boc-cis-4-hydroxy-L-proline methyl ester, j) N-boc-trans-4-hydroxy-L-proline methyl ester, k)
(S)-(–)-3-chloro-1-phenyl-1-propanol, l) (R)-(+)-3-chloro-1-phenyl-1-propanol, m) (R)-(+)-2-amino-3-phenyl-1-propanol, n) (S)-(–)-2-amino-3-phenyl-1-propanol, o) (1R,2S,5R)-(–)-menthol, p) (1S,2R,5S)-(+)-menthol.

Experimental

Chemicals and reagents

Different carboxylic acids, alcohols, and amino alcohols were analyzed (Figure 3). (–)-Carbobenzyloxy-L-proline (a), (+)-carbobenzyloxy-D-proline (b), (S)-(–)-1-carbobenzyloxy-2-piperidinecarboxylic acid (c), (R)-(+)-1-carbobenzyloxy-2-piperidinecarboxylic acid (d), (S)-(+)-2-heptanol (e), (R)-(–)-2-octanol (h), (1R,2S,5R)-(–)-menthol (o), (1S,2R,5S)-(+)-menthol (p), (S)-(–)-3-chloro-1-phenyl-1-propanol (k), (R)-(+)-3-chloro-1-phenyl-1-propanol (l), N-boccis-4-hydroxy-L-proline methyl ester (i), N-boc-trans-4-hydroxy-L-proline methyl ester (j), and N,N′-dicyclohexylcarbodiimide (DCC) were purchased from Sigma Aldrich (Steinheim, Germany). (R)-(–)-2-heptanol (f), (S)-(+)-2-octanol (g), (R)-(+)-2- amino-3-phenyl-1-propanol (m), (S)-(–)-2-amino-3-phenyl-1-propanol (n), and 4-(dimethylamino)-pyridine (DMAP) were obtained from Fluka (Buchs, Switzerland); methanol (HPLC gradient grade) was from Roth (Karlsruhe, Germany), and dichloromethane (HPLC grade) and toluene were from AppliChem (Darmstadt, Germany).

Sample preparation

All tested substrates, with the exception of (R)-(+)-2-amino-3-phenyl-1-propanol and its corresponding enantiomer, were solved in dichloromethane with a concentration of 20 μmol/L. To create the calibration curve, five stock solutions were prepared with 100, 50, 0, –50, and –100ee%. For the analysis of the carboxylic acids, an auxiliary solution with (S)-(+)-2-heptanol and (R)-(−)-2-octanol, each in a concentration of 100 μmol/L, was prepared with toluene. For the analysis of alcohols and amino alcohols, the auxiliary solution contains (–)-carbobenzyloxy-L-proline as well as (R)-(+)-1-carbobenzyloxy-2-piperidinecarboxylic acid in the same concentration. Finally, a toluene solution, which contains the reagents DCC (200 μmol/L) and DMAP (2 μmol/L), was prepared.

(R)-(+)-2-amino-3-phenyl-1-propanol and its corresponding enantiomer were solved with a concentration of 0.2 μmol/mL in toluene. The corresponding auxiliary solution contains (–)-carbobenzyloxy-L-proline and (R)-(+)-1-carbobenzyloxy-2-piperidinecarboxylic acid, each in a concentration of 2 μmol/mL. The reagent solution contains 5 μmol/mL DCC and 0.5 μmol/mL DMAP.

The derivatization was performed in 1-mL vials from Agilent (Waldbronn, Germany) and in 96-well multiple-well plates with a well volume of 500 μL from Greiner-BioOne (Essen, Germany); 50 μL of the solution with chiral substrates, 100 μL of the auxiliary solution, and 50 μL of the solution with the reagents were added and mixed in a thermo shaker (Thermomixer comfort, Eppendorf, Hamburg, Germany) for 1 hr at 20 °C followed by the addition of 300 μL methanol. For the MS analysis, the samples were diluted with methanol (1:100, v/v).

Instrumentation

Table 1 - Characteristic masses (m/z) of the [M+Na]+ ions of the derivative pairs

The mass spectrometric analyses were carried out on an Agilent LC-MS system consisting of a G1379B vacuum degasser, G1312B binary pump, G1367C high- performance automated liquid sampler, and G1969A time-of-flight mass spectrometer (TOF-MS) with an electrospray ion source. Five microliters of the sample solution were injected with a prior needle wash. The mobile phase had a flow of 1 mL/min and consisted of methanol and water with 0.1% formic acid (90:10, v/v). The TOF-MS was operated in positive ion mode with the following parameters: nitrogen as nebulizer and drying gas, nebulizer pressure of 35 psig, drying gas flow of 10 L/min, drying gas temperature of 350 °C, capillary voltage of 4000 V, fragmentor voltage of 215 V, skimmer voltage of 60 V, and octopol voltage set at 250 V. The measurement was operated using the SCAN mode in the range of 50–1000 m/z. Data acquisition was done using Mass Hunter Data Acquisition software from Agilent. Table 1 shows the masses (m/z) of the pairs of derivatives analyzed as protonated ions [M+Na]+.