LTP consolidation: Substrates, explanatory power, and functional significance Gary Lynch a, * , Christopher S. Rex b , Christine M. Gall b,c a Department of Psychiatry and Human Behavior, Gillespie Neuroscience Research Facility, University of California, Irvine, CA 92697-4292, USA b Department of Neurobiology and Behavior, University of California, Irvine, CA 92697, USA c Department of Anatomy and Neurobiology, University of California, Irvine, CA 92697, USA Received 11 May 2006; received in revised form 5 July 2006; accepted 17 July 2006 Abstract Long-term potentiation (LTP) resembles memory in that it is initially unstable and then, over about 30 min, becomes increasingly resistant to disruption. Here we present an hypothesis to account for this initial consolidation effect and consider implications that follow from it. Anatom- ical studies indicate that LTP is accompanied by changes in spine morphology and therefore likely involves cytoskeletal changes. Accordingly, theta bursts initiate calpain-mediated proteolysis of the actin cross-linking protein spectrin and trigger actin polymerization in spine heads, two effects indicative of cytoskeletal reorganization. Polymerization occurs within 2min, has the same threshold as LTP, is dependent on integrins, and becomes resistant to disruption over 30 min. We propose that the stabilization of the new cytoskeletal organization, and thus of a new spine morphology, underlies the initial phase of LTP consolidation. This hypothesis helps explain the diverse array of proteins and signaling cascades implicated in LTP, as well as the often-contradictory results about contributions of particular molecules. It also provides a novel explanation for why LTP is potently modulated by factors likely to be released during theta trains (e.g., BDNF). Finally, building on evidence that normal patterns of activity reverse LTP, we suggest that consolidation provides a delay that allows brain networks to sculpt newly formed memories. Ó 2006 Elsevier Ltd. All rights reserved. Keywords: Long-term potentiation; Integrin; Actin; Cytoskeleton; Brain-derived neurotrophic factor; Hippocampus 1. Introduction The idea that memories are fragile when new but become more stable with time is so commonplace that it is probably pointless to ask when it was first recorded. Ribot in 1882 writes as though it was an accepted principle, citing the fol- lowing anecdote from a still earlier researcher: ‘‘[Mr H.] was driving his wife and child in a phaeton, when the horse took fright and ran away; and, all attempts to pull him in being unsuccessful, the phaeton was at last violently dashed against a wall, and Mr H. was thrown out, sustain- ing a severe concussion of the brain. On recovering, he found that he had forgotten the immediate [original italics] antecedents of the accident, the last thing he remembered being that he had met an acquaintance on the road about two miles from the scene of it. Of the efforts that he made, and the terror of his wife and child, he has not, to this day, any recollection whatsoever.’’ Why Mr H. remembered meeting the acquaintance perhaps 20 min before the accident, but lost all memory of events, some quite striking (‘the terror of his wife and child’), that transpired afterwards up to the accident, is a central question in the neurobiological study of memory. Ribot took the intui- tive position that newly learned material progressively stabi- lizes (‘consolidates’) with time, and this idea has, with occasional challenges (Misanin et al., 1968), dominated think- ing in the field ever since (McGaugh, 2000). The discovery of consolidation-like effects in laboratory animals (Duncan, * Corresponding author. Tel.: þ1 949 824 1517; fax: þ1 949 824 1255. E-mail address: glynch@uci.edu (G. Lynch). 0028-3908/$ - see front matter Ó 2006 Elsevier Ltd. All rights reserved. doi:10.1016/j.neuropharm.2006.07.027 Neuropharmacology 52 (2007) 12e23 www.elsevier.com/locate/neuropharm