Fig. 6 . Blocking e-ATP signaling increases the frequency of cholinergic cells not obeying the minimal spacing characteristic of these cells. (A) Autocorrelation plot of a normal cholinergic array at P2. The autocorrelation plots all the distances between the cells in the sample analyzed. In the normal cholinergic arrays the autocorrelation is a uniform distribution with a central hole. This means that the only spatial constraint these cells obey is to avoid getting closer to one another than a minimal distance, as already described (Galli-Resta et al., 1997). (B) Autocorrelation plot of an apyrase-treated cholinergic array at P2. The autocorrelation is still a uniform distribution with a central hole, but this is smaller than normal. This indicates the reduced efficacy of a mechanism normally ensuring the minimal spacing typical of the cholinergic arrays. (C,D) Examples of density recovery profiles (DRPs) histograms, plotting the density of counts in the autocorrelation as a function of the distance from the center of the coordinates, are shown for control (C) and treated (D) cells. In both cases the DRP rises from zero to a constant density, but does so in much shorter distances in the treated (D) than in the normal case (C). Count density is normalized to the average plateau density; an arrow indicates the exclusion radius (ER) value. (E) The size of the central exclusion region in the autocorrelation is quantified using the ER. The average ER is plotted with its standard deviation for the GCL cholinergic mosaics in control (vehicle injected) and apyrase- or oATP-injected retinas. e-ATP signaling blockade induces a significant ER reduction with respect to control (t-test, apyrase P<10^-7; oATP P<10^-6). Data were derived from the autocorrelation of all sampled fields (for each treatment n=8 retinas, 4 samples per retina).