Memory Flashcards
• Whitlock et al (2006):
Measures fEPSP of CA1 (good measure of synaptic strength and hence induction of LTP). Inhibitory avoidance task (an associative learning paradigm). Control either didn’t receive the shock or given the shock and immediately remove so they couldn’t form an association. Monitor synaptic transmission in CA1 by electrically stimulating Schaffer collaterals and evoking fEPSP (multielectrode array used)
o All animals showed strong avoidance of going back into the dark side compared to the controls (no footshock) → memory is very Changes in electrically evoked neural responses after learning
o Short lasting increase in AMPA receptor subunits (GluA1 and GluA2) in post-synaptic membranes after learning
o Found that EPSP response doesn’t change much at all after IA training compared to baseline and control group but between channel variance increases substantially (256%) →Most channels showed a slight decrease in fEPSP after behaviour but some showed a substantial increase which appears rapidly and persists throughout the session, but only in IA trained group
o found an inverse relationship between magnitude of behavioural LTP and electrical LTP suggesting a common mechanism → Concluded that hippocampal LTP supported associative spatial memory but only in specific hippocampal regions
o Next measured levels in GluA1 in post synaptic membrane by isolating synaptoneurosomes from the dorsal hippocampus and used immunohistochemistry with selective antibodies to look at delivery of AMPAR subunits into the post-synaptic membrane→ found increase in both GluA1 and GluA2 in post-synaptic membrane of IA trained group vs controls (rapid and transient). No change in NR1 (no change in NMDAR numbers) consistent with the LTP mode
o No change in cerebellum, no change in overall AMPAR subunit levels in crude hippocampal homogenates (i.e. reflecting trafficking)
o Correlation not causality→ timings is a problem here
• Tononi and Cirelli (2006):
plastic process occurring during wakefulness result in net increase in synaptic strength in many brain circuits. Role or sleep is to downscale synaptic strength so to a baseline level that is energetically sustainable→ beneficial for memory and learning. Sleep is the price we have to play → goal is homeostatic regulation of the total synaptic weight impinging on neurons
• Vyazovskiy et al (2008):
Rat cortex and hippocampus. Collecting SNS and comparing levels of GluA1 in post-synaptic density in animals that had been awake for 6 hours compared to animals asleep for 6 hours. Immunohistochemistry
o Showed that GluA1 containing AMPAR increase across 6 hr wake and reduce over 6hrs of sleep.
o Also changes in phosphorylation states of AMPARs , CAMPKII and GSK3β are consistent with synaptic potentiation during wakefulness and depression during sleep
o Amplitude od cortical evoked responses increase after wakefulness and decrease after sleep and correlate with change in SWA
o -Observed molecular changes cannot prove conclusively that synaptic efficiency in vivo has changes as molecular assays do not reflect the functional state of synapse necessarily → e.g. synaptoneurosomes can not distinguish between sureface and internal receptors or pre and postsynaptic pools
o Observed molecular and electrophysiological changes
• O’Keefe and Dostrovsky (1971):
Microelectrodes as rats roamed→ Place cells
• Morris et al (1986):
Water maze task. Chronic intracerebroventricular infusion using osmotic minipumps of AP5.
• (first major piece of evidence to support the LTP memory hypothesis)
o D-AP5 (NMDAR antagonist) impairs acquisition of spatial reference memory watermaze task → learned more slowly and took more indirect route than the controls but no significant difference. Rats did not show any preference in transfer test
o Impairs performance of spatial reference memory WM but they are not useless
o Suggest that NMDA receptor dependant learning is involved in spatial learning.
o -Learning vs performance?
o Sensorimotor side effects (changes in motivation or emotion)
o -Drug is going into the ventricles so will go to all regions of the brain, Hippocampus or extrahippocampus ?
o Certainly the case there were sensorimotor side effects. Have a problem climbing to platform and particularly when they try to shake themselves dry they often fell back into the water and had to find the platform again (are these effecting their performance in the water maze) – Keith and Rudy 1991, Saucier and Cain 1995
• Bannerman et al (1995);
Upstairs/downstairs water maze tas. Spatial pre-training. 2 different water mazes. Compare rats that are trained for the first time in the water maze while receiving d-AP5 (experimentally niave rats) vs WM experienced rats now receiving DAP5 and learning new maze
o Spatial pretraining removes the spatial learning deficit with AP5 however experimentally niave mice impaired in upstairs WM
o So may not need hippocampal NMDA receptors for spatial navigation to form an association between a spatial location and a platform
o (Despite the fact that if take the animals and try to induce LTP in the hippocampus at the end of the study, can see a massive reduction in LTP)
• Tsien et al (1996):
CA1-specific GluN1 KO mice. MWM. First demonstration of a mouse where not only could a key protein required for the induction of LTP be KO but was done in a region and cell type specific manner. KO was restricted to CA1 subfield of the hippocampus Used standard Cre technology
o If take a hippocampal slice and try to induce LTP there is absence of LTP (functional verification)
o Impaired in the water maze → Slower to learn where the platform was on the MWM, clear impairment on task
o Suggest that NMDA receptor dependant LTP is required for acquisition of spatial learning and memory
o - But turned out that the KO was not selective for the hippocampus at all (expression outside in the cortex) (Fukaya et al 2003 showed clear reduction of GluN1 levels in cortex within 2 months of the mice being born)
o Indication of this in paper. As did a control task where they had to swim to a visible platform (beacon) – hippocampus independent. But KO mice were impaired on the visible platform task
• Bannerman et al (2012):
(CA1 and DG)-specific NMDAR KO. GluN1hpc KO mice (Grin 1). From granule cells in DG and pyramidal cells in the CA1 Cre expression.
o Impaired on radial maze but not on the watermaze (both require spatial reference memory)→ is this due to ambiguity of the radial maze?
o But impaired on spatial discrimination (Identical Beacon task – one is escape platform). (e.g. how many times did they swim to the decoy than the escape beacon)→ they made a lot more errors) All mistakes were nearly all where trials starting close to the decoy. So they are much slower at stopping themselves from going to first beacon they see. So they’re much less able to behaviourally inhibit the very strong approach response they will exhibit when they see a black ball
o Not impaired on non-spatial, visual discrimination beacon task (move the location of platform every trial)
o Important here is that there isn’t a deficit because of a general spatial learning and memory deficit instead the deficit is more likely to reflect ambiguity associated with the spatial choice test.
o Hypothesise that the role of the hippocampus in disambiguating between overlapping memories during memory retrieval
o Suggests hippocampal NMDA receptors are not required to encode associative spatial memories
• Tang et al (1999):
Effect of NR2 overexpression in cortex and hippocampus of mice Overexpression in the forebrains of transgenic mice
o Enhanced and prolonged opening of NMDAR in individual synapses. fEPSPs were significantly greater→ suggested enhanced NMDAR response and following tetanic stimulation there was larger potentiation, both inhibited AP5
o In retention test, transgenic mice showed stronger preference than wildtype to the novel object than one they had already explored → implies better long term memory
o Contextual (hippocampal dependant) and cue (not) fear conditioning, mice showed elevated freezing in both
o Fear extinction paradigm – they were quicker to learn to dissociate preciously associated event MWM faster spatial learning
o Implies that MNDA receptors are important not only just spatial memory but other forms of long term memory
o Overexpression was throughout the forebrain so was it hippocampal dependant?
• Zamanillo et al (1999):
KO GluA1 mice→ brain slices. CA1 pyramidal cells, showed a functional AMPAR reduction
o Lack LTP in hippocampus, consistent with the trafficking idea
o CA1 soma patch current were strongly reduced, Glu synaptic currents were unaltered
o LTP was absent in CA3-CA1 synapses but spatial learning in the water maze not impaired
o Suggests that CA1 hippocampal LTP is controlled by the number of subunit composition of AMPAR and show a dichotomy between LTP and CA1 ad acquisition of spatial memory
• Reisel et al (2002):
GluA1 KO mice. GluA1 KO mice display normal spatial RM acquisition but have selective spatial WM deficit
o No impairment on associative long-term spatial memory tasks
o Normal spatial reference learning and memory, both on h MWM and on appetitively motivated Y-maze task
o Showed specific spatial working memory impairment during tests on non-matching to place on both the Y maze and elevated T make
o Suggest different forms of hippocampus-dependant spatial memory involve different aspects of neural processing within the hippocampus
• Bygrave et al (2019)
Recorded hippocampal-prefrontal coherence while assessing spatial working memory and short term habituation when KOdGria1 (GluA1). LFPs recorded
o Gria1–/– knockout mice exhibited a pronounced deficit in short-term habituation, which was reflected by a deficit on the spatial novelty preference test, which was largely rescued by the re-introduction of GluA1 into CA2/CA3
o Found that β and low-γ frequency coherence could predict working memory performance , theta coherence was unrelated to performance
o Novel environments , theta coherence specifically tracked exploration-related attention in wt
o Reintroduction of GluA selectively into CA2 and CA3 restored abnormal short term habituation, theta coherence and hippocampal and prefrontal theta oscillation
o Sustained elevation of hippocampal prefrontal theta coherence may underlie a failure in regulating novelty -related selective attention leading to aberrant salience → mechanistic link between GRIA1 and schizophrenia ? Excessive and prolonged levels of selective attention
• Scmidtt et al (2003):
GluA1 deletion→ radial maze
o Spatial reference memory (which arms are baited and which ones aren’t)
o To keep tract of which arms have already visited on a particular trial (Spatial working memory)
o GluA1 doesn’t impair spatial reference memory but impairs spatial working memory as they make many more errors
o So conclude that GluA1 is not required for spatial reference memory but required for spatial working memory
o Selective spatial WM deficit
o Suggest that different neuronal mechanisms within the hippocampus may support different kinds of information processing
• Sanderson et al (2011):
GluA1 KO mic→ looked at discrete non-spatial visual stimuli on operant chamber. Light then another same or different → short term memory deficit
o Stimulus-specific sensitisation of GluA1 KO mice
o Aberrant salience and schizophrenia
o Overall GuA1 KO mice have impaired short term memory which results in deficits in habituation can display stimulus specific sensitisation to a recent stimulus, preserved long-term associative memory
o Suggests GluA1 is not necessary for encoding but affects the way that short term memory is expressed
