Resolution for 125I-labeled specimens under electron microscope (EM) autoradiographic conditions was assessed experimentally. With this isotope the size of the silver halide crystal was the most important resolution-limiting factor. Heavy metal staining such as is routinely used in preparing animal tissues for EM autoradiography produced an improvement in resolution of -15-20%. For a 500-1,000-/~ biological tissue section fixed with OsO4 and stained with uranyl acetate, we obtained resolution (half distance, HD) values of -800 -120/~ using Ilford L4 emulsion and 500 -70 A using a Kodak NTE-type emulsion. General aspects of resolution-limiting factors and comparison with 3H and 14C values are discussed.On theoretical grounds, 125I is a highly favorable isotope (see Appendix A) for electron microscope (EM) autoradiography. This was early recognized by Kayes et al. (11) in 1962. However, although some calibration studies for both sensitivity and resolution have been performed using this isotope (6, 10), the resolution attainable with this isotope under varying experimental conditions has not yet been systematically assessed. In the present study we have adapted previous calibration procedures used for tritium and ~4C (2,7,20,22) to obtain resolution values for mzsI as a function of section thickness, photographic emulsions, developing procedures, and heayy metal staining as employed in Em autoradiography. The results allowed us to reassess the critical parameters affecting resolution for isotopes of different energy.
MATERIALS AND METHODS
Resolution SpecimensTwo resolution specimens were used. One was to test resolution in plastic specimens (density 1.1) as a function of section thickness and different emulsion-developer combinations; the second was to test the effect of increasing specimen density by the incorporation of heavy metals such as osmium and uranium, which are used in the fixation and staining of biological material for electron microscopy.RESOLUTION USING PLASTIC TEST SPECI-MENS: The calibration specimen was modeled after that described previously for tritium and 14C (2,20,22) in which a thin film of radioactive polystyrene was sandwiched between a polymerized Epon block and a layer of nonradioactive methacrylate. (The radioactive styrene was thus never exposed to the solvents normally involved in embedding for EM autoradiography, which would