Primary and solvent kinetic isotope effects (KIEs) for the proton transfer between a butanone molecule and the microsolvated ion OH-(H2O)(n=0,1,2) to give the (H2O)(n+1)... butenolate complex have been theoretically calculated in order to study solvation changes that take place along the proton transfer process. To take into account in the transition state calculations the two dynamical bottlenecks involved in those cluster reactions (one corresponding to the ion-dipole approach and the other one to the actual proton transfer), the canonical unified statistical theory has been used. Only for the n = 2 case, the local bottleneck in the proton transfer region controls the global process at the whole range of temperatures studied. The different values of the KIEs as a function of temperature and depending on the relative weight of the two dynamical bottlenecks are analyzed. The degree of coupling between the motion of solvent molecules and the motion of the proton that is being transferred is studied by testing the rule of the geometrical mean. This empirical rule holds very well when only one bottleneck governs the global process, but an important breakdown of this rule is found when the two bottlenecks are comparable. Assuming a monotonical variation of solvent isotope effect from reactants to product, the solvent KIE, by comparison with the equilibrium solvent isotope effect, indicates that the hydroxide ion is 44% desolvated at the transition state of the reaction with n = 2 at 298 K. All these results have been obtained without including tunneling effects. A first estimation of the tunneling contribution to kinetic isotope effects has been evaluated by interpolated variational transition state theory.