Anat Embryol (1989) 180:1-15 Anatomy and Embryology 9 Springer-Verlag 1989 Review article Quantitative electron microscopic immunocytochemistry of neuroactive amino acids Ole P. Ottersen Anatomical Institute, University of Oslo, Karl Johansgate 47, N-0162 Oslo 1, Norway Summary. Amino acids are of crucial importance in brain function, not only as metabolic intermediates and building blocks of proteins, but also as mediators of interneuronal communication. This dual role of the amino acids distingu- ishes them from other neurotransmitter candidates, and im- plies that they are unlikely to be restricted to neurons using them as transmitters. This calls for a quantitative approach when attempts are made to analyse the distribution of trans- mitter amino acids by means of immunocytochemistry. The present review deals with recent methodological develop- ments that have made it possible to utilize specific antisera to explore the cellular and subcellular distribution of neu- roactive amino acids in a quantitative manner. Key words: Glutamate Taurine - Glycine Immunocyto- chemistry Cerebellum - Hippocampus - Medulla spinalis Introduction Amino acids act as building blocks of proteins and as meta- bolic intermediates in all living cells. In brain tissue, several amino acids in addition serve important roles as neurotrans- mitters (Fonnum 1984; Iversen and Goodman 1986; Cot- man et al. 1987). Amino acids seem ideally suited as media- tors of synaptic transmission: they are small molecules that diffuse rapidly and can be effectively handled by cellular uptake and release mechanisms; they can be stored in small volumes and can be quickly replenished. The same properties that seem attractive in a biological sense have hampered attempts to unravel the precise local- ization of the amino acids in nerve tissue. Redistribution and loss of amino acids during subcellular fractionation procedures preclude accurate measurements of amino acid concentrations in synaptosomes and organelles, and their rapid turnover may introduce inaccuracies in biochemical analyses even at the regional level (review: Ottersen and Storm-Mathisen 1984a). With regard to immunocytochem- istry, the small size of the amino acid molecules long de- terred attempts to develop antibodies against them. In 1983, however, Storm-Mathisen et al. showed that polyclonal antisera raised against glutaraldehyde-bovine serum albu- min conjugates of GABA and glutamate could be used to demonstrate the respective amino acids in sections of glutar- aldehyde-fixed brains. The success of this strategy reflects the ability of glutaraldehyde to couple the free amino acids in the tissue to brain macromolecules, thus forming com- plexes that because of their structural similarity with the immunogens are recognized by the specific antibodies (for reviews on methodological aspects, see Ottersen et al. 1986; Storm-Mathisen and Ottersen 1986; Ottersen and Storm- Mathisen 1987a). The same or similar strategies have since been used by several groups to prepare antibodies against a series of dif- ferent amino acids thought to act as neurotransmitters or neurotransmitter precursors (GABA: Seguela et al. 1984; Hodgson et al. 1985; Maley and Newton 1985; Mature and Streit 1986; Wenthold et al. 1986; and others, glutamate: Mature et al. 1987; Hepler et al. 1988; Liu et al. 1989; and others, aspartate: Ottersen and Storm-Mathisen 1985; Campistron et al. 1986c; Saito et al. 1986; Aoki et al. 1987; Hepler et al. 1988; and others, giycine : Pourcho and Goebel 1985; Campistron et al. 1986a; Dale et al. 1986; Ottersen etal. 1987; Wenthold et al. 1987; Aoki etal. 1988; van den Pol and Gores 1988; and others, taurine: Madsen et al. 1985; Ottersen etal. 1985; Campistron etal. 1986b; Yoshida et al. 1986; Ida et al. 1987; and others, glutamine: Laake et al. 1986). Several of the antibodies referred to above are monoclonals (e.g., Matute and Streit 1986; Matute et al. 1987). Alternative immunocytochemieal ap- proaches for the detection of amino acids include immuni- zations with conjugates of dimers of amino acids (Madl et al. 1987) or with gold particles instead of proteins as amino acid carriers (Shiosaka et al. 1986). The different procedures for raising and applying amino acid antisera will be evaluated in a forthcoming review (Ottersen and Storm-Mathisen 1989b). Since all of the amino acids that are supposed to act as transmitters (with the possible exception of GABA) also play a diversity of metabolic roles, they would be expected to occur in varying amounts in most or all cell compart- ments. In the case of glutamate, for example, biochemical studies have pointed to the existence of several pools: a transmitter pool in glutamatergic terminals, a neuronal met- abolic pool in non-glutamatergic neurons (including a pre- cursor pool in GABAergic cells), and a glial pool (Fonnum 1984; Fonnum et al. 1986). The distribution of neuroactive amino acids should therefore ideally be analysed by immu- nocytochemical procedures that permit a quantitative as- sessment of the immunoreactivity. The immunogold proce- dure (reviews: De Mey 1983; Roth 1983; van den Pol 1984) is particularly suitable for such studies; in addition to pro- viding possibilities for quantification, this procedure offers a high resolution and is compatible with good preservation of ultrastructural detail. In the present review an attempt