The kinetic of S. Aureus alpha-toxin channels were examined by """"""""noise"""""""" analysis at several different pH's to measure the kinetics of proton binding to ionizable sites within the channel. Recognizing that this binding follows first-order kinetics, and noting that proton binding creates a step change in channel conductance one sees that the responsible ionizable sites have a pk of 5.5 and that the association and dissociation rate constants are 8x10-9 and 10-5 respectively. These values imply that the ionizable residues are freely accessible to the aqueous phase and that the active groups are either histidines, glutamic acids or aspartic acids. By exposing ionic channels to neutral polyethyleneglycols of different molecular weight, we have been able to generalize the use of water- soluble polymers to probe channel structure. Large excluded solutes in membrane- bathing solutions create an osmotic work of transition between different conductance states, a work proportional to the difference in polymer-inaccessible aqueous spaces of these states. Progressively smaller polymers begin to penetrate the channel space, to cause a decrease in channel conductance and to exert a correspondingly reduced effective osmotic pressure. We find that the channel, a kind of ultimate molecular sieve, appears to exclude polymers on the basis of their rotational volume calculated from radius of gyration (rather than the polymer cross-section or diameter typically used to gauge channel size). In a particular case of alamethicin, polymer probing reveals a fixed (about 3,000 A-3) aqueous volume change with each successive step between the five different conductance states. These states, then, are collections of conducting units of nearly the same conductance rather than pores of successively larger radius.
