Curare alkaloids from Matis Dart Poison: Comparison with d-tubocurarine in interactions with nicotinic, 5-HT(3) serotonin and GABA(A) receptors

Several novel bisbenzylisoquinoline alkaloids (BBIQAs) have recently been isolated from a Matis tribe arrow poison and shown by two-electrode voltage-clamp to inhibit mouse muscle nicotinic acetylcholine receptors (nAChR). Here, using radioligand assay with Aplysia californica AChBP and radioiodinated α-bungarotoxin ([(125)I]-αBgt), we show that BBIQA1, BBIQA2, and d-tubocurarine (d-TC) have similar affinities to nAChR orthosteric site. However, a competition with [(125)I]-αBgt for binding to the Torpedo californica muscle-type nAChR revealed that BBIQAs1, 2, and 3 are less potent (IC(50)s = 26.3, 8.75, and 17.0 μM) than d-TC (IC(50) = 0.39 μM), while with α7 nAChR in GH(4)C(1) cells, BBIQA1 was less potent that d-TC (IC(50)s = 162 μM and 7.77 μM, respectively), but BBIQA2 was similar (IC(50) = 5.52 μM). In inhibiting the Ca(2+) responses induced by acetylcholine in Neuro2a cells expressing the mouse adult α1β1εδ nAChR or human α7 nAChR, BBIQAs1 and 2 had similar potencies to d-TC (IC(50)s in the range 0.75–3.08 μM). Our data suggest that BBIQA1 and BBIQA2 can inhibit adult muscle α1β1εδ nAChR by both competitive and noncompetitive mechanisms. Further experiments on neuronal α3β2, α4β2, and α9α10 nAChRs, expressed in Xenopus laevis oocytes, showed that similar potencies for BBIQAs1, 2, and d-TC. With α3β2γ2 GABA(A)R currents were almost completely inhibited by d-TC at a high (100 μM) concentration, but BBIQAs1 and 2 were less potent (only 40–50% inhibition), whereas in competition with Alexa Fluor 546-α-cobratoxin for binding to α1β3γ2 GABA(A)R in Neuro2a cells, d-TC and these analogs had comparable affinities. Especially interesting effects of BBIQAs1 and 2 in comparison with d-TC were observed for 5-HT(3A)R: BBIQA1 and BBIQA2 were 5- and 87-fold less potent than d-TC (IC(50) = 22.63 nM). Thus, our results reveal that these BBIQAs differ from d-TC in their potencies towards certain Cys-loop receptors, and we suggest that understanding the reasons behind this might be useful for future drug design.