Solid-phase synthesis and biological evaluation of a combinatorial library of philanthotoxin analogues

The modular structure of philanthotoxins was exploited for construction of the first combinatorial library of these compounds using solid-phase parallel synthesis. (S)-Tyrosine and (S)-3-hydroxyphenylalanine were used as amino acid components, spermine, 1,12-dodecanediamine, and 4,9-dioxa-1,12-dodecanediamine as amine components, and butanoyl, phenylacetyl, and cyclohexylacetyl as N-acyl groups. Following automated preparative HPLC, the resulting 18 compounds were isolated as the S-forms in 40-70% yields. The purity of the products was determined by HPLC with evaporative light scattering detection and by (1)H and (13)C NMR. The thus obtained philanthotoxins were tested electrophysiologically for their antagonist properties on human muscle-type nicotinic acetylcholine receptors (nAChR) expressed in TE671 cells and on rat brain non-NMDA glutamate receptors (non-NMDAR) expressed in Xenopus oocytes. 4-Hydroxy analogues lacking the secondary amino groups (PhTX-12 and 4,9-dioxa-PhTX-12 and their analogues) were inactive on non-NMDAR, whereas the potency of the spermine derivatives (PhTX-343 and its analogues) increased with steric bulk of the N-acyl group. The analogue of PhTX-343 in which the N-butanoyl group was replaced by phenylacetyl group had IC(50) of 15 +/- 4 nM on non-NMDAR. Increasing the steric bulk of the N-acyl group was not advantageous for activity at nAChR, and a sharp decrease in potency with increased steric bulk was observed with the derivatives of PhTX-12. 3-Hydroxy analogues generally exhibited lower activity and different response to alterations of the N-acyl groups as compared to the 4-hydroxy analogues. Since the acyl group alterations in PhTX-343 and 4,9-dioxa-PhTX-12 have a similar effect on potency, which is distinctly different from that observed for PhTX-12, the two former compounds may bind to nAChR in a similar fashion but differently from that of PhTX-12. The combinatorial library approach described in this work represents a prototype methodology for future exploration of structure-activity relationships of philanthotoxins.