Serotonin modulates retinotectal and corticotectal convergence in the superior colliculus

A dense serotonin (5-HT)-containing projection to the superficial layers of the superior colliculus (SC) has been demonstrated in diverse mammalian species, but how 5-HT may affect visual signals within these laminae is largely unknown. This study undertook to investigate the distribution of 2 types of 5-HT receptors in the SC and to ascertain their physiological effects on transmission of visual signals to the SC from the retinotectal and corticotectual pathways. Autoradiography of tissue sections exposed to [3H]-8-OH-DPAT (8-hydroxy-dipropylaminotetraline) or to [125I]cyanopindolol plus isoproterenol showed that 5-HT1A and 5-HT1B receptors, respectively, were present in the superficial SC layers. In unilaterally enucleated animals, binding of ligand to 5-HT1B receptors was greatly reduced on the deafferented (contralateral) side, which is consistent with the possibility that these receptors are located on preterminal axons. Binding to 5-HT1A receptors was unaltered by enucleation. In recordings of superficial layer neurons from SC slices, application of 5-HT during blockade of 5-HT1A receptors with spiperone reduced the amplitude of EPSPs evoked by stimulation of the optic tract. The 5-HT concentration for a 50% reduction in EPSP amplitude was 6 microM. Under these conditions, there were no significant alterations in either membrane potential or input resistance concurrent with 5-HT mediated reduction in EPSPs. During extracellular in vivo recordings, 5-HT, applied by iontophoresis or micropressure or by endogenous release produced by electrical stimulation of the dorsal raphé nucleus, strongly suppressed visual activity in SC neurons. The effectiveness of 5-HT application was significantly stronger on responses evoked by electrical stimulation of the optic chiasm (an average response decrement of 92.2%) than on these evoked in the same neurons by stimulation of visual cortex (an average response reduction of 32.3%). These results support the following conclusions. The 5-HT1B receptors are located preferentially on optic axon terminals and exert presynaptic inhibition of retinotectal inputs. Secondly, 5-HT1A receptors probably have a postsynaptic localization and may affect activity of SC neurons irrespective of the source of input. The combined effect of 5-HT at both subtypes would bias SC visual activity toward information received from the corticotectal pathway.