Separation of the optical antipodes of the novel σ₁ receptor ligand [¹⁸F]fluspidine using chiral HPLC

Objectives: (±)-[¹⁸F]Fluspidine represents a novel ligand with high binding affinity and selectivity toward σ₁ receptors [1,2]. With the aim to identify the individual potential of the two enantiomers of (±)-[¹⁸F]fluspidine for particular disease-related imaging studies, this racemate as well as its tosylate precursor needed to be enantioseparated. Therefore, we performed an extensive chiral HPLC study with different chiral stationary phases (CSPs) and eluent modes.

Methods: The use of the immobilized polysaccharide type column Chiralpak IA (Chiral Technologies Europe, Daicel) offered the possibility to study various solvents, which are not common for HPLC. We used mixtures of several alkenes with ethers, esters, alcohols and chlorinated solvents as NP mode eluents. ACN and MeOH as organic modifier mixed with aqueous TEA/AcOH or NH₄OAc as buffers were used in RP mode. Working with this column type in the polar ionic mode by using 100% ACN or MeOH with TEA/AcOH was possible as well. Furthermore, we studied the enantioselectivity on different coated polysaccharide phases (cellulose and amylose based), a Pirkle-type CSP (with pi-acceptor-donor character) and a glycopeptide CSP (teicoplanin derived; Dr. Maisch, Germany). The identification of the enantiomers was carried out with a chiral detector (OR 2090, JASCO) and is based on the measurement of optical rotation.

Results: The tosylate precursor could be best enantioseparated with the Chiralpak IA column using 90% n-hexane/THF/0.1% DEA. At a flow rate of 1mL/min the (–)-enantiomer eluted with t_R=13.9 min in front of the (+)-enantiomer with t_R=32.4 min. This remarkable separation is appropriate for a semi-preparative separation of the precursor compound to provide enantiomerically pure precursor compounds for the radiosynthesis of (S)-(–)-[¹⁸F]fluspidine and (R)-(+)-[¹⁸F]fluspidine. In contrast to the results obtained with the tosylate, the corresponding fluoro compound fluspidine could not be enantioseparated with the Chiralpak IA column. Also most of the other columns investigated were not successful. Exclusively, the Reprosil-Chiral OM column (comparable to Chiralcel OD) was appropriate to separate the fluspidine enantiomers sufficiently in NP mode using 97% n-hexane/isopropanol/0.1% DEA (Figure). This method was transferred to the automated radiosynthetic procedure to analyze the enantiomeric purity of (S)-(–)-[¹⁸F]fluspidine and (R)-(+)-[¹⁸F]fluspidine, which was higher than 97%.

Figure. Analytical enantioseparation of (±)-fluspidine (left) and verification of (R)-(+)-[¹⁸F]fluspidine after radiosynthesis (right) on Reprosil-Chiral OM with 97% n-hexane/isopropanol/0.1% DEA

Conclusion: We successfully developed two chiral HPLC methods for enantioseparation of fluspidine and its tosylate precursor. Furthermore, we have repeatedly shown, that it is not predictable which chiral phase will be suitable to separate the enantiomers of a compound even if the derivatives are structurally related.

Acknowledgment: Supported by DFG (STE 601/10-2, WU 176/7-2).
References: [1] Fischer S, et al (2011) Eur J Nucl Med Mol Imaging, 38, 540-551. [2] Maisonial A, et al (2012) Bioorg Med Chem, 20, 257-269.