Paul W. Frankland

The organization of recent and remote memories

A fundamental question in memory research is how our brains can form enduring memories. In humans, memories of everyday life depend initially on the medial temporal lobe system, including the hippocampus. As these memories mature, they are thought to become increasingly dependent on other brain regions such as the cortex. Little is understood about how new memories in the hippocampus are transformed into remote memories in cortical networks. However, recent studies have begun to shed light on how remote memories are organized in the cortex, and the molecular and cellular events that underlie their consolidation.

Memory Reconsolidation and Extinction Have Distinct Temporal and Biochemical Signatures

Memory retrieval is not a passive phenomenon. Instead, it triggers a number of processes that either reinforce or alter stored information. Retrieval is thought to activate a second memory consolidation cascade (reconsolidation) that requires protein synthesis. Here, we show that the temporal dynamics of memory reconsolidation are dependent on the strength and age of the memory, such that younger and weaker memories are more easily reconsolidated than older and stronger memories. We also report that reconsolidation and extinction, two opposing processes triggered by memory retrieval, have distinct biochemical signatures: pharmacological antagonism of either cannabinoid receptor 1 or L-type voltage-gated calcium channels blocks extinction but not reconsolidation. These studies demonstrate the dynamic nature of memory processing after retrieval and represent a first step toward a molecular dissection of underlying mechanisms.

Preferential incorporation of adult-generated granule cells into spatial memory networks in the dentate gyrus

Throughout adulthood, new neurons are continuously added to the dentate gyrus, a hippocampal subregion that is important in spatial learning. Whether these adult-generated granule cells become functionally integrated into memory networks is not known. We used immunohistochemical approaches to visualize the recruitment of new neurons into circuits supporting water maze memory in intact mice. We show that as new granule cells mature, they are increasingly likely to be incorporated into circuits supporting spatial memory. By the time the cells are 4 or more weeks of age, they are more likely than existing granule cells to be recruited into circuits supporting spatial memory. This preferential recruitment supports the idea that new neurons make a unique contribution to memory processing in the dentate gyrus.

The Involvement of the Anterior Cingulate Cortex in Remote Contextual Fear Memory

Although the molecular, cellular, and systems mechanisms required for initial memory processing have been intensively investigated, those underlying permanent memory storage remain elusive. We present neuroanatomical, pharmacological, and genetic results demonstrating that the anterior cingulate cortex plays a critical role in remote memory for contextual fear conditioning. Imaging of activity-dependent genes shows that the anterior cingulate is activated by remote memory and that this activation is impaired by a null α-CaMKII mutation that blocks remote memory. Accordingly, reversible inactivation of this structure in normal mice disrupts remote memory without affecting recent memory.

The dorsal hippocampus is essential for context discrimination but not for contextual conditioning

The authors describe how (a) the timing of hippocampal lesions and (b) the behavioral-representational demands of the task affect the requirement for the hippocampus in contextual fear conditioning. Post- but not pretraining lesions of the hippocampus greatly reduced contextual fear conditioning. In contrast, pretraining lesions of the hippocampus abolished context discrimination, a procedure in which mice are trained to discriminate between 2 similar chambers (shock context vs. no-shock context). Whereas either contextual- or cue-based strategies can be used to recognize an aversive context, discrimination between similar contexts is optimally acquired by contextual (hippocampal)-based strategies. In keeping with the lesion results, Nf1(+/-)/Nmdar1(+/-) mutant mice, which have spatial learning deficits, are impaired in context discrimination but not in contextual conditioning. Together, these data dissociate hippocampal and nonhippocampal contributions to contextual conditioning, and they provide direct evidence that the hippocampus plays an essential role in the processing of contextual stimuli.

α-CaMKII-dependent plasticity in the cortex is required for permanent memory

Cortical plasticity seems to be critical for the establishment of permanent memory traces. Little is known, however, about the molecular and cellular processes that support consolidation of memories in cortical networks. Here we show that mice heterozygous for a null mutation of α-calcium-calmodulin kinase II (α-CaMKII+/-) show normal learning and memory 1–3 days after training in two hippocampus-dependent tasks. However, their memory is severely impaired at longer retention delays (10–50 days). Consistent with this, we found that α-CaMKII+/- mice have impaired cortical, but not hippocampal, long-term potentiation. Our results represent a first step in unveiling the molecular and cellular mechanisms underlying the establishment of permanent memories, and they indicate that α-CaMKII may modulate the synaptic events required for the consolidation of memory traces in cortical networks.

The acoustic startle reflex: neurons and connections.

The startle reflex protects animals from blows or predatory attacks by quickly stiffening the limbs, body wall and dorsal neck in the brief time period before directed evasive or defensive action can be performed. The acoustic startle reflex in rats and cats is mediated primarily by a small cluster of giant neurons in the ventrocaudal part of the nucleus reticularis pontis caudalis (RPC) of the reticular formation. Activation of these RPC neurons occurs 3-8 ms after the acoustic stimulus reaches the ear. Undetermined neurons of the cochlear nuclei activate RPC via weak monosynaptic and strong disynaptic connections. The strong disynaptic input occurs via neurons of the contralateral ventrolateral pons, including large neurons of the ventrolateral tegmental nucleus that integrate auditory, tactile and vestibular information. RPC giant neurons, in turn, activate hundreds of motoneurons in the brain stem and the length of the spinal cord via large reticulospinal axons near the medial longitudinal fasciculus. To hindlimb motoneurons, monosynaptic connections from the reticulospinal tract are weak, but disynaptic connections via spinal cord interneurons are stronger and show temporal facilitation, like the startle response itself.

Selective Erasure of a Fear Memory

Memories are thought to be encoded by sparsely distributed groups of neurons. However, identifying the precise neurons supporting a given memory (the memory trace) has been a long-standing challenge. We have shown previously that lateral amygdala (LA) neurons with increased cyclic adenosine monophosphate response element–binding protein (CREB) are preferentially activated by fear memory expression, which suggests that they are selectively recruited into the memory trace. We used an inducible diphtheria-toxin strategy to specifically ablate these neurons. Selectively deleting neurons overexpressing CREB (but not a similar portion of random LA neurons) after learning blocked expression of that fear memory. The resulting memory loss was robust and persistent, which suggests that the memory was permanently erased. These results establish a causal link between a specific neuronal subpopulation and memory expression, thereby identifying critical neurons within the memory trace.

Spaced training induces normal long-term memory in CREB mutant mice

BACKGROUND The cAMP responsive element binding protein (CREB) is a transcription factor the activity of which is modulated by increases in the intracellular levels of cAMP and calcium. Results from studies with Aplysia, Drosophila and mice indicate that CREB-activated transcription is required for long-term memory. Furthermore, a recent study found that long-term memory for olfactory conditioning can be induced with a single trial in transgenic Drosophila expressing a CREB activator, whereas in normal flies, with presumably lower CREB-mediated transcription levels, conditioning requires multiple spaced trials. This suggests that CREB-mediated transcription is important in determining the type of training required for long-term memory of olfactory conditioning in Drosophila. Interestingly, studies with cultured Aplysia neurons indicated that removing a CREB repressor promoted the formation of long-term facilitation, a cellular model of non-associative memory. RESULTS Here, we have confirmed that mice lacking the alpha and Delta CREB proteins (CREBalphaDelta-) have abnormal long-term, but not short-term, memory, as tested in an ethologically meaningful task. Importantly, additional spaced training can overcome the profound memory deficits of CREBalphaDelta- mutants. Increasing the intertrial interval from 1 to 60 minutes overcame the memory deficits of the CREBalphaDelta- mice in three distinct behavioral tasks: contextual fear conditioning, spatial learning and socially transmitted food preferences. CONCLUSIONS Previous findings and results presented here demonstrate that CREB mutant mice have profound long-term memory deficits. Importantly, our findings indicate that manipulations of CREB function can affect the number of trials and the intertrial interval required for committing information to long-term memory. Remarkably, this effect of CREB function is not restricted to simple conditioning tasks, but also affects complex behaviours such as spatial memory and memory for socially transmitted food preferences.

The HMG-CoA reductase inhibitor lovastatin reverses the learning and attention deficits in a mouse model of neurofibromatosis type 1.

Neurofibromatosis Type 1 (NF1) is a common neurological disorder caused by mutations in the gene encoding Neurofibromin, a p21Ras GTPase Activating Protein (GAP). Importantly, NF1 causes learning disabilities and attention deficits. A previous study showed that the learning and memory deficits of a mouse model of NF1 (nf1+/-) appear to be caused by excessive p21Ras activity leading to impairments in long-term potentiation (LTP), a cellular mechanism of learning and memory. Here, we identify lovastatin as a potent inhibitor of p21Ras/Mitogen Activated Protein Kinase (MAPK) activity in the brain. Lovastatin is a specific inhibitor of three-hydroxy-3-methylglutaryl coenzyme A (HMG-CoA) reductase, used commonly for the treatment of hypercholesterolemia. We report that lovastatin decreased the enhanced brain p21Ras-MAPK activity of the nf1+/- mice, rescued their LTP deficits, and reversed their spatial learning and attention impairments. Therefore, these results demonstrate that lovastatin may prove useful in the treatment of Neurofibromatosis Type 1.