172 Brain Research, 132 (1977) 172 175 ~:) Elsevier/North-Holland Biomedical Press Membrane linked proteins at CNS synapses SHU-HUI YEN, RONALD K. H. LIEM, PAUL T. KELLY, CARL W. COTMAN and MICHAEL L. SHELANSK1 Department o/ Neuroscience, Mental Retardation Research Center, Children's Hospital Medical Center, and Department of Neuropathology, Harvard Medical School, Boston, Mass. 02115, and De- partment of Psychobiology, University of California, Irvine, Calif. (U.S.A.) (Accepted April 27th, 1977) Considerable progress has been made in recent years toward understanding the molecular architecture of synaptic junctions in the central nervous system. The postsynaptic density (PSD), the undercoating of the postsynaptic membrane at most excitatory synapses, consists predominantly of proteinsl,L Two of the proteins associated with the PSD are the fibrous proteins, actin and tubulin6,1°,13,14. We now report an association of brain filament protein (BF) with the PSD. BF is the presumptive subunit protein of intermediate brain filaments: fibers 10 nm in diameter which course through the neuronal cytoplasm15 with no obvious membrane association. These filaments have been proposed to maintain the shape of these cells, but have not previously been related to synapses. Since the reported BF subunit molecular weight of about 51,0005,11,12,15 approximates that of the major polypeptide constituent of the PSD 1,7, we evaluated the possibility that synaptic j unc- tions contain BF. Highly specific antisera to calf BF were raised in rabbits, using the 51,000 mole- cular weight band cut from sodium dodecyl sulfate (SDS) polyacrylamide gels15. The F(ab')2 fragment of anti-BF antibody was conjugated to horseradish peroxidase, and this complex was applied to sections of mouse brain 16. Electron microscopic examina- tion showed heavy staining of PSDs (Fig. 1), as well as neuronal and glial filaments1~. Synaptic boutons did not stain. Absorption of the antiserum with BF caused a com- plete loss of PSD staining as well as filament staining. In view of this immuno- histochemical evidence and previous evidence for the presence of tubulin in the PSD, we undertook further studies on isolated synaptic junctions and PSDs to determine whether their major polypeptide constituents react with anti-BF or anti-tubulin anti- sera. These antibodies have been extensively tested for possible cross-reaction against each other. No anti-tubulin activity was found in the anti-BF antiserum, and there was only an insignificant amount of anti-BF activity in the anti-tubulin antiserum when tested against bovine antigensL Anti-BF was also found to react with mouse s and rat brain filaments, but not with rat brain tubulin (Liem, Yen and Shelanski, unpublished observations).