Therefore, the SPRY2 domain detection was also carried out in native conditions using dot blot. As a control antibody, we used a commercial Wortmannin antibody against RyR1, antibody 34C. In Western blot, the signal for the detection of the SPRY2 domain in purified, denatured RyR1 was weak as compared to the control RyR1 antibody. In contrast, native dot blot detection of purified RyR1 was very intense in all cases, suggesting strong nativestructure dependence for the immunodetection. The native dot blot detection levels in rabbit skeletal muscle vesicles were similar amongst all three anti-SPRY2 antibodies, but lower than the 34C antibody. At least for anti-SPRY B, all this is in full agreement with the report that this antibody recognizes native RyR1 in SR vesicles and that it can immunoprecipitate purified RyR1 with a third of the efficacy measured for the antibody 34C. Since for a symmetric shape there are several redundant configurations of the Euler angles that yield the same 2D projection, for every decorated raw image where the antibody was visible and did not fall in a specified corner of the RyR1, we replaced its Euler angles for the symmetrically related Euler angles that would place the antibody on the specified corner. If the raw image had more than one antibody bound, this was repeated by providing the symmetrically related Euler angles that would place this second bound antibody in the specified location, and so on. Finally, the 3D reconstruction was performed with all antibodies back-projected to the specified repeat. The above method works as long as the antibody is visible and always binds to the same site. To establish this we first made a rough identification of the 3D position of the antibody relative to the RyR1 by creating a simulated 3D of RyR1 with four antibodysized spheres located at four equivalent positions and projecting this simulated model in all orientations. Then once the RyR1 in the raw image was matched to the corresponding projection, the position of the projected spheres was compared to the position of the extra mass. The four spheres were then placed in other locations and the process was repeated until the match between the set of simulated projections and the set of raw images was optimal, which indicated the approximate 3D location of the antibody. With this new method it has been possible not only to identify the antibody in the individual particles but it has also been possible to recover the signal in the context of the 3D structure. SPRY domains have been shown to mediate protein-protein interaction processes involved in diverse cellular functions. The SPRY domains of several proteins have been recently crystallized, revealing a high structural conservation for this domain. A distinctive feature of SPRY domains for the protein-protein interaction specificity is the presence of an unstructured and flexible loop between b-strands. The flexible loop is thought to mediate specific protein-protein interactions. This loop shows conformational exchange in intermediate time scales, which appears to be important for protein-protein interactions. The 3D location of SPRY2 within domain 6 of RyR1 found in this study.