Ottersen O P
Anatomical Institute, University of Oslo, Norway.
J Chem Neuroanat. 1989 Jan-Feb;2(1):57-66.
A series of glutamate-glutaraldehyde-brain protein conjugates, prepared with different glutamate concentrations, was embedded in resin, cut, and incorporated in a multilayered sandwich which was embedded anew. Ultrathin cross-sections through the sandwich were collected on mesh grids and incubated together with tissue sections in an antiserum recognizing fixed glutamate, followed by a second layer antibody coupled to colloidal gold particles. This model system revealed a roughly linear relationship between the concentration of fixed glutamate (up to 154 mmol/l) and gold particle density. Extrapolated to the accompanying tissue sections this finding implies that a tissue compartment displaying 2n gold particles/microns2 contains twice as many fixed glutamate residues per volume as a compartment with n gold particles/microns2. By use of the model system the concentration of fixed glutamate in presumed glutamatergic nerve terminals in the cerebellum was estimated to be two to three times the average tissue level and about five times higher than the concentration of fixed glutamate in terminals thought to use gamma-aminobutyric acid as transmitter. This study represents a step towards the use of electron microscopic immunocytochemistry as a tool to assess the absolute concentrations of neuroactive amino acids in nerve terminals and other tissue compartments.