NMDA RECEPTOR BLOCKADE ALTERS THE TOPOGRAPHY OF NATURALLY-OCCURRING GANGLION-CELL DEATH IN THE RAT RETINA

During the first 10 days after birth around half the rat’s retinal ganglion cells die. Previous work has shown that ganglion cells whose axons have made large topographic targeting errors are preferentially eliminated during this period and that the selection of such cells for preferential elimination is dependent on an activity-driven mechanism: this process is one way in which the postnatal refinement of the topography of the retinocollicular projection is achieved. We have given systemic injections of the NMDA channel blocker MK801 during the first 14 days of life to see whether this activity-dependent process works via the NMDA channel. We assessed the topographic pattern of retinal ganglion cell death by making localized injections of fast blue into the superior colliculus at birth and measuring the distribution of labeled ganglion cells either at Postnatal Day 2 or at Day 14. We find that overall retinal ganglion cell death, measured by optic nerve axon counts, is not prevented by MK801 treatment. However, whereas in untreated animals ganglion cells whose axons have reached the most topographically inappropriate target area are preferentially eliminated, in MK801-treated animals ganglion cell death appears to be random, in that we see no evidence of the preferential elimination of retinal ganglion cells whose axons have made large topographic targeting errors in MK801-treated animals. NMDA receptor blockade therefore has the same effect on the pattern of retinal ganglion cell death as tetrodotoxin blockade of the retina. © 1993 by Academic Press, Inc.