Differential effects of electrical stimulation patterns, motivational-behavioral stimuli and their order of application on functional plasticity processes within one input in the dentate gyrus of freely moving rats in vivo

Hippocampal long-term potentiation (LTP) is a long-lasting increase in synaptic efficacy considered to be the cellular basis of memory. LTP consists of an early, protein synthesis-independent phase (E-LTP) and a late phase that depends on protein synthesis (L-LTP). In water-deprived rats E-LTP in the dentate gyrus (DG) can be reinforced into L-LTP, if the rats were allowed to drink within 15 min after E-LTP induction (behavioral LTP-reinforcement, BR). LTP can be depotentiated by low-frequency stimulation (LFS) to the same synaptic input if applied shortly after tetanization (<10 min). Here, we addressed the question of whether a BR protocol is able to recover LTP at depotentiated synaptic inputs. We show that LTP, depotentiation, LFS and BR specifically interact within one afferent input, which could be explained by the "synaptic tagging" hypothesis outlined by [Frey and Morris (1997) Nature 385:533-536]. E-LTP induced by a weak tetanus (WTET) sets tags in the activated inputs which are able to capture and to process plasticity-related proteins (PRPs) required for L-LTP, the synthesis of which was induced by BR. Synaptic tags could be reset by LFS. BR alone was unable to rescue depotentiated LTP, but the combination of BR and subsequent WTET transformed E-LTP into L-LTP. We show that LTP, LTD and behavioral stimuli alternatively and reversibly affect a single afferent input for long periods of time by LTP as well as LTD mechanisms, competing with each other under the influence of different concurrent stimuli. Affective modulation can shift the balance to one or the other. We show that the result will depend not only on the last stimulus, but on the history of previous stimuli applied to the specific input. Afferent stimuli activate alternative, but partially overlapping cascades with long-lasting consequences for the input including spaced-associative processes of "synaptic tagging" as well as "cross-tagging" which could be demonstrated in single synaptic afferents to one neuronal population in freely behaving animals.