Figure 17.9.4.

Sampling size and the resolving power of antibody-mediated FRET. (A) Diagram of a model IgG molecule bound to an antigen on the membrane surface (left). The Fc and each Fab domain are modeled as a cylinder of diameter 3 nm, height 7 nm, connected by flexible hinges. Arrows indicate directions of allowed rotational flexibility. The range of potential positions that can be occupied by dyes conjugated to the antibody surface is indicated on the right. Dyes are allowed to be on the surface of the stalk of the mushroom-shaped space, and anywhere in the volume of the head. (B) A simulated donor and acceptor labeled antibody (4 dyes/IgG, randomly distributed as described in panel A) were bound to antigens separated by distances of between 8 and 16 nm, and the % energy transfer between the dyes calculated and plotted. Each datum represents the average of between 1 and 1000 such simulations as indicated (n) on each graph. Custom macros (available upon request from the authors) were written for NIH Image 1.62 to perform the antibody-mediated FRET simulations. In essence, an algorithm was designed to simulate the stochastic binding of a mixture of donor- and acceptor-labeled antibodies to a set of antigens on a membrane surface, followed by a calculation of the FRET between the randomly distributed dyes on all of the bound antibodies. The algorithm encompassed the following steps: (1) the x-y positions of the appropriate number of antigens were distributed on a hypothetical surface of defined area (usually 0.5 × 0.5 μm) at the indicated density and configuration (either randomly distributed, or in clusters of three). Clusters were not allowed to overlap, and the minimal distance separating adjacent antigens was limited to 8 nm, as determined by the steric hindrance of bound IgG molecules. (2) Each antigen was randomly assigned to either be unoccupied, bound by a donor antibody, or bound by an acceptor antibody. The relative probabilities of each assignment were determined by the desired occupancy and donor/acceptor ratio. (3) The x-y-z positions of dyes were randomly chosen relative to each antigen by the criteria outlined in the text. (4) Once the x-y-z positions for all of the donor and acceptor dyes were set, the summed FRET efficiency that would be expected for this distribution of dyes was calculated according to previously established equations (Förster, 1948; Dewey and Hammes, 1980).