On the basis of neurophysiological and biochemical studies, it has been suggested that the amino acid, glycine, is a major inhibitory transmitter in the mammalian spinal cord (I, 23). The evidence in support of a role for glycine as a neurotransmitter includes the demonstration that glycine (a) mimics the action of naturally occurring inhibitory transmitter when it is iontophoretically applied to motor neurons (5, 6, 24), (b) is released after dorsal root stimulation (12), (c) is reduced in concentration when inhibitory interneurons are selectively destroyed (7), and (d) is taken up into brain slices and a distinct population of synaptosomes by a glycine-specific high-affinity transport system (2,16,17,22). This uptake into specific synaptosomes can be demonstrated only in those regions of the central nervous system in which glycine is iontophoretically active and present in high concentration.Electron microscope autoradiography has been used in attempts to identify synaptic terminals possessing high affinity uptake systems for glycine (10,11). These autoradiographic studies of [SH] glycine uptake in vitro and in vivo have shown silver grains over nerve terminals with flattened vesicles (9, 14, 15). However, previous reports have not (a) provided quantitative evidence for selective labeling of synapses, (b) quantitatively analyzed the topographical distribution of labeled synapses, or (c) demonstrated that the label was in glycine and not in protein or metabolites.In the present investigation, synapses on rat ventral horn cells were labeled by [3H]glycine in vivo' and were identified by electron microscope autoradiography. Our findings indicate that (a) the grain density is greatest in synaptic terminals, (b) a high proportion of axosomatic and proximal axodendritic synapses take up [aH]glycine, (c) many of the labeled terminals have elliptical and pleomorphic vesicles, and (d) [SH]glycine is not metabolized significantly under the conditions of our experiment.
MATERIALS AND METHODS
Microinjection Technique and FixationSprague-Dawley rats weighing 150-300 g were anesthetized with chloral hydrate (40 mg/kg body weight), the lumbar spinal cord was exposed, and a small dural puncture 1.2 mm from midline was made over the L5 cord level with a fine shaq~ned needle. By means of a micromanipulator, a micropipette (tip diameter 15 ,m) was advanced perpendicularly 1.6 mm into the ventral horn. 0.4 ,l [aH]glycine (10 Ci/mmol, 4.5 • 10 5 dpm/,l, N~w England Nuclear, Boston, Mass.) in 0.1 M phosphate buffer (pH 7.3) was injected over 5 rain, using a screw-driven apparatus. The calculated concentration of [aH]glycine at the site of injection was 2.1 • 10 -2 raM. All animals were sacrificed 10 min after the beginning of the injection. Groups of animals were studied in three ways. Autoradiography was done on four female rats perfused with 2.5% glutaraldehyde in 0.1 M cacodylate buffer with 3% sucrose (Group I). By using a dissecting