Controlled initiation of enzymatic reactions in micrometer-sized biomimetic compartments

We present a technique to initiate chemical reactions involving few reactants inside micrometer-scale biomimetic vesicles (10(-12) to 10(-15) L) integral to three-dimensional surfactant networks. The shape of these networks is under dynamic control, allowing for transfer and mixing of two or several reactants at will. Specifically, two nanotube-connected vesicles were filled with reactants (substrate and enzyme, respectively) by microinjection. Initially, the vesicles are far apart and any diffusive mixing (on relevant experimental time scales) between the contents of the separated vesicles is hindered because of the narrow diameter and long axial extension of the nanotube. To initiate a reaction, the vesicles were brought close together, the nanotube was consumed by the vesicles and at a critical distance, the nanotube-vesicle junctions were dilated leading to formation of one spherical reactor, and hence mixing of the contents. We demonstrate the concept using a model enzymatic reaction, which yields a fluorescent product (two-step hydrolysis of fluorescein diphosphate by alkaline phosphatase), where product formation was measured as a function of time using a FRAP fluorescence microscopy protocol. By comparing the enzymatic activity with bulk measurements, the enzyme concentration inside the vesicle could be determined. Reactions could be followed for systems having as few as approximately 15 enzyme molecules confined to a reactor vesicle. To describe the experiments we use a simple diffusion-controlled reaction model and solve it using a survival probability approach. The agreement with experiment is qualitative, but the model describes the trends well. It is shown that the model correctly predicts (i) single-exponential decay after a few seconds, and (ii) that the substrate decay constant depends on the number of enzymes and geometry of reaction container. The numerical correction factor Lambda is introduced in order to ensure semiquantitative agreement between experiment and theory. It was shown that this numerical factor depends weakly on vesicle radius and number of enzymes, thus it is sufficient to determine this factor only once in a single calibration measurement.