A ROLE FOR A CHICKEN HOMOLOG OF THE NEURAL CELL-ADHESION MOLECULE L1 IN CONSOLIDATION OF MEMORY FOR A PASSIVE-AVOIDANCE TASK IN THE CHICK

Intracranial injection of antibodies directed against the neural cell adhesion molecule L1 resulted in amnesia for passive avoidance trainin g in day-old chicks tested 24 hr subsequently. L1 antibodies were amne sic when administered at one of two time windows: 30 min pretraining a nd 5.5-8 hr post-training. No amnesia was apparent if injections were made at times before, between, or after these time windows (-2, +1, +3 , +4, or +12 hr relative to training). A fragment of the L1 molecule d erived from the external fibronectin domains FN1-5 produced amnesia on ly when injected at the 5.5-hr timepoint, whereas a fragment of the im munoglubin-like domains Ig I-VI produced amnesia only when injected 30 min prior to training. We have shown previously that long-term memory for the passive avoidance task requires two waves of glycoprotein syn thesis, the first occurring immediately after training, and the second some 6 hr thereafter. The glycoprotein synthesis inhibitor 2-deoxygal actose results in amnesia if injected at either time, whereas the neur al cell adhesion molecule (N-CAM) is specifically involved only in the second wave. The coincidence of the time course of memory disruption resulting from injection of L1 antibodies with that occurring with 2-d eoxygalactose supports the hypothesis that establishment of an endurin g memory for the experience of passive avoidance training requires two waves of glycoprotein synthesis, each wave being biochemically and fu nctionally discrete. The differential effects of the two L1 fragments suggests that separate mechanisms of synaptic stabilization are involv ed at the two time points.