Errors introduced in radioligand binding studies due to displaceable, cation dependent, [3H]prazosin binding to glass-fibre filters and glass surfaces.

[3H]prazosin not only specifically and homogeneously labels alpha 1-adrenoceptors, but also binds to glass surfaces and non-linearly to the glass-fibre filters, commonly used in radioligand binding experiments. Binding to filters can be modulated by unlabeled alpha-adrenergic compounds and cations. If no correction is applied for displaceable filter binding, analysis of [3H]prazosin binding experiments leads to erroneous results. Analysis of [3H]prazosin saturation experiments on guinea-pig cerebral cortex membranes with correction for filter binding before the non-linear fit procedure indicated that [3H]prazosin labels a homogeneous population of alpha 1-adrenoceptors (Rtot: 8.33 fmol.mg-1 wet tissue) with a dissociation constant of 1.28 x 10(-10) M. However, analysis of the same data after correction for non-specific binding, (determined in parallel experiments by adding 10 microM phentolamine to the incubation medium) resulted in a best fit to a model in which [3H]prazosin labels two alpha 1-adrenoceptor subpopulations (R1: 15.0 fmol.mg-1 and R2: 14.6 fmol.mg-1 wet tissue) with dissociation constants of respectively 1.78 x 10(-10) and 5.63 x 10(-9) M. The discrepancy between the two methods of analysis is due to displacement of the radioligand from the filters by phentolamine. Prazosin and oxymetazoline are also able to displace filter-bound [3H]prazosin. The extent to which displaceable filter binding distorts the proper results depends on the actual magnitude of the error and also on the method of analysis.