Binding experiment KD values and physiologically active GABA concentrations: an only apparent contradiction?