THE CONE SYNAPSES OF CONE BIPOLAR CELLS OF PRIMATE RETINA

One each of bipolar cell types DB2 and DB4, together with a flat and a n invaginating midget bipolar cell, were taken from a Golgi-stained rh esus macaque retina; then serially sectioned for EM examination of the ir synapses with cone pedicles. The cone input to the dendrites of the DB2 cell was exclusively at basal junctions; it had a characteristic distribution. Fifty per cent of the basal synapses were with cone pedi cle membrane immediately adjacent to the dendrite of a bipolar cell in vaginating to end opposite the ribbon of a cone triad (this, therefore , is called triad-associated). The remainder were one or more synapses distant from the triad-associated position (and, therefore, non-triad associated). The DB4 cell had both basal (predominantly in the triad- associated position) and ribbon-related synaptic input. But the basal to invaginating ratio differed from that of our previously published c ell; 56% basal, 43% invaginating, as compared with 31% basal and 69% i nvaginating. Like foveal IMB cells the synapses of the mid-peripheral invaginating midget bipolar cell were exclusively invaginating; but we re about 25% more numerous. The flat midget bipolar cell made exclusiv ely basal synapses. These were 2.5 times more numerous than those of f oveal flat midget bipolar cells, and 3.5 times the number of invaginat ing midget bipolar synapses at equivalent eccentricity. The synapses b etween cones and diffuse and midget bipolar cells are characteristic f or each particular bipolar cell type, but the details depend on a cell 's distance from the fovea (eccentricity). A rather constant number of cone pedicle synaptic ribbons 38.6+/-2.5 (n=60) was found across mid- peripheral macaque and vervet monkey retinae. The smaller mean number for vervet monkey, 27.4+/-3.5 (n=23), suggests there can also be gener ic differences in synaptic detail at cone bipolar cell synapses.