February 28, 2011
Make a psychotropic drug by a new route. Asenapine (5) is an antipsychotic drug for treating schizophrenia, acute mania associated with bipolar disorder, and depression. The drug is the racemate of the trans isomer, and both enantiomers of the racemate are equally effective. Inventors J. Zhu, R. Keltjens, and J. Firet summarize existing routes used to make 5; the main problems are associated with achieving high trans/cis isomer ratios of intermediates during the synthesis.
The patent is unusual because it describes a theoretical route to achieve a trans configuration of the bridging C–C bond between the two rings in the synthesis of 5. The philosophy is to leave the five- or seven-membered ring open until the desired configuration is achieved and then close the ring. The authors decided to use a strategy starting from the seven-membered ring diester 1. The key aspect of the patent is the selective reduction of 1 to form trans-2.
The inventors found, surprisingly, that the preferred reduction method uses magnesium metal and an alcohol. The patent claims also cover the use of metal hydrides and catalytic reduction, but no examples of these are given.
In the reduction of 1 with Mg–MeOH, 2 is isolated in almost 100% yield with a trans/cis ratio of 97:3. In the second step, 2 is reduced with LiAlH4 to form diol 3, which is recovered in crude form and converted to its bis(methanesulfonate) 4. The final step is the reaction of 4 with aq MeNH2 to produce desired molecule 5 in 97% yield with a cis content of <2%. The reactions are carried out on gram or milligram scales, so whether they can be scaled up is unknown.
The patent also describes the synthesis of starting material 1, which is outlined in the second sequence. The first step is the base-catalyzed condensation of ester 6 with dimethyl oxalate to form diester 7. After workup, the crude product is isolated in 83% yield and treated with polyphosphoric acid to cyclize 7 to anhydride 8, isolated in only 15% yield.
In the final step, diester 1 is produced from 8 by heating it with KF in MeOH, followed by adding MeI. The product is isolated in 78.7% yield. 1H NMR data are given for 1 and 5, and 13C NMR data are given for 1. An alternative route is suggested for making 1 by way of 8 that starts from a regioisomer of 6 in which the ester group is in the other ring. However, no details are given.
The process gives the desired compound with high selectivity, but the synthesis of the starting material is very inefficient. (Synthon BV [Nijmegen, The Netherlands]. US Patent 7,875,729, Jan. 25, 2011; Keith Turner)