Rapid Miniaturized Shake Flask Determination of LogP, LogD, Lipophilicity and pKa using ILE 96-Well Plates

This study is currently being conducted, and the final results will be presented at AAPS (San Diego, Nov, 2007), and published here upon its completion.

Monday, November 12
Session Time: 01:00 pm - 05:00 pm
Location: Exhibit Hall, San Diego Convention Center

Purpose. To develop a high throughput method to determine LogP, LogD, lipophilicity and pKa of compounds over a wide pH range using polymeric coated wells of a 96-well plate as a miniature “shake flasks.”

Methods. This method uses the premise that the octanol/water partition ratio of a compound is approximately equal to its PDMS/water partition ratio. Each well of an ILE 96 well plate coated with 0.1uL (phase ratio = 1000) of polydimethylsiloxane was used as a miniature “shake flask.” 100uL of buffer ranging in pH from 2 to 11 was added to each well (12) of a plate row. Each well was spiked with 0.1ug of of cocaine, flurazepam, imipramine, loratidine and nicotine. The pates were sealed with alumaseal and extracted for 60 minutes by vortexing with a Scientific Industries Microplate Genie. The Plates were then emptied. The wells were back extracted into 150uL acetonitrile:methanol (2:1) and agitating for twenty minutes on a Gilson 205H Orbital Shaker for 20 minutes @ 400 rpm. 1uL of the supernant from each well was analyzed by GC/MS using an HP-5971 GC-MS in SIM mode, splitless injection on a 30m X 0.25mm i.d. X 0.25 VB-5 column.

Partition coefficients were determined for each well. LogD was then plotted against pH. Curve fitting algorithms were then used to determine LogD at pH 7.4, Log P and pKa. LogP data was determined by taking the maximum LogD for each compound.

Results. Phase ratio was found to affect the accuracy of the determinations depending upon extraction efficiency. Cocaine gave the closest correlation to literature values with determined LogP of 2.35, LogD7.4 of 1.65 and pKa of 8.52/5.38, as compared to literature values of LogP 2.3 and pKa of 8.61/5.59. Nicotine extraction efficiency on PDMS was insufficient to supply data. Flurazepam showed a pKa of 7.9 and LogP of 3.02 versus the literature values of 8.16 and 3.81 respectively. Loratadine showed a LogP of 3.32 versus the literature value of 3.64.

Conclusions. It is possible to rapidly determine LogP, LogD, lipophilicity and pKa of compounds in a 96-well format where individual wells act as miniature "shake flasks." When the phase ratios of extractions are adjusted to give extraction efficiencies between 20 and 50%, it is possible to determine a compound's LogD and LogP to two significant digits while simultaneously determining pKa.