Posted comment on ´Engram size varies with learning and reflects memory content and precision` written by J. Leake, R. Zinn, L.H. Corbit, M.S. Fanselow and B. Vissel and published in Journal of Neuroscience 2021 41 (18) p. 4120 doi 10.1523/JNEUROSCI.2786-20.2021
SUMMARY
Leake and colleagues investigated neuronal assemblies in the dentate gyrus that represented memories formed and recalled with contextual fear conditioning. They looked at how learning time could affect memory content and object discrimination. For their experiments they used 8-12 weeks old transgenic mice that were heterozygous for the Fos2AiCREER/+ gene and homozygous for the reporter gene. Contextual fear conditioning experiments were carried out using standard fear conditioning chambers (Context A – test, aniseed essence scented, Context B – control, ethanol) and an unconditional stimulus of a 2sec 1mA footshock. Conditioning was carried out by placing the animals in the chamber for different periods of time before the footshock was given and for the behavioural testing the mice were placed in either conditioning context A or B for up to 30 minutes without the shock. For the targeted labelling experiments, the mice were habituated to the intraparietal injections for 7 days before the beginning of the experiments. They then received a single intraparietal injection of 4-hydroxytamoxifen (100mg/kg) immediately after removal from the conditioning chamber. The mice were then tested 7 days after conditioning to allow for transgene expression.
Behavioural change was assessed by measuring every 4 seconds incidences of freezing (immobility) or not. The data was analysed by taking the number of freezings divided by the number of samples to give the percentage. Immunohistochemistry was also carried out and this involved Leake and team performing c-fos immunohistochemistry on brain coronal slices of the hippocampus. Fluorescent confocal images of the dorsal hippocampus were acquired using an inverted laser scanning confocal microscope and cell counts performed using the semiautomated custom designed macro, ImageJ. Statistical analyses (ANOVA, Student t tests) were performed on all results.
The first set of experiments examined the relationship between PSI (Placement shock interval) and accuracy of memory when the mice were subject to in the conditioning chamber footshock paired with Context A or Context B with no footshock. The test period was 3 minutes since this was found to give the optimum PSI for both conditions. Leake and team found that conditional freezing increased with PSI as did discrimination with the shock context preferentially expressed. Fear in Context A was initially low but increased rapidly to its 3minute limit whereas PSI did not change for Context B, remaining close to zero. For other time intervals, fear freezing did not differ significantly for both 0secs and 30secs intervals for both Context A/shock or Context B/no shock whereas mice demonstrated more fear at 720secs interval for Context A/shock than for Context B whose freezing value was close to baseline still. This indicated to the authors that the PSI mediated conditioned fear memory precision and that longer PSI produced better discrimination.
The second set of experiments examined the number of cells activated during the memory task with the premise that the greater level of contextual information acquired with the longer PSI, the greater the number of dorsal hippocampal cells activated (greater the number of c-fos+ cells). This was seen to be the case with PSI increase increasing the number of c-fos+ cells in the hippocampal regions studied, ie. dentate gyrus (DG – 5.187, CA3 – 11.45, CA1 – 8.28). The activity increase was at its maximum at 180 secs in the DG area where it reached a plateau whereas the plateau was reached later at 720secs PSI in the CA3 and 1800secs for the CA1. These results supported Leake`s hypothesis that increased activation of the hippocampal cells was linked to extended time in the context before the footshock and that different areas reached plateaus of activity at different times (DG first, then CA3, then CA1). This allowed more contextual information to be processed and learnt leading to more precise discrimination in future trials.
The third set of experiments looked at whether the c-fos+ expression activation after recall was also affected by the length of the initial exposure to context. Leake and team used conditioned mice with a training PSI of 30 seconds(elicits imprecise memories) or 720 seconds (elicits precise memories). These mice were then exposed to the context for up to 1800secs. The authors found that the number of c-fos+ cells increased with re-exposure duration in the DG, CA3 and CA1 regions and at both PSI intervals. They also found that the maximum number of activated cells was reached at the same re-exposure duration at both PSIs with DG reaching a plateau at 180secs (about 90 c-fos+ cells) and CA3 and CA1 at 720secs. At 1800 secs time point a slight decrease in number of c-fos+ cells was observed although this was not significantly different to that at 720secs. Therefore, it was concluded that the duration of the PSI did not affect the rate of cellular activation at recall. The observation of the slight decrease in c-fos+ cells at longer time intervals was suggested as possibly due to degradation of some of the c-fos protein that may have been generated at the start of the trial session.
The fourth set of experiments was set-up to confirm that the numbers of cells activated after conditioning at different PSIs (30, 180, 720 secs) were the same as those activated after recall sessions of equivalent durations (30, 180, 720 secs). It was found that the number of c-fos+ cells and c-fos activity increased as a function of re-exposure duration in all three hippocampal regions (DG, CA3, CA1), but there was no significant difference to those found for the learning history indicated by the results for the previous PSIs. The authors explained these observations as the total number of c-fos+ cells being influenced by the duration of the session preceding the c-fos assessment itself and not by the extent of the PSI and learning of the previous contextual experience.
The next experiments in this set involved repeating the contextual fear conditioning and PSI duration variations, but with a TRAP2 mouse line. This was carried out because c-fos expression in the TRAP2 mice drives the integration of tamoxifen-inducible Cre recombinase (CreERT2). CreERT2 translocates to the nucleus to initiate recombination and permanent expression of the effector gene when the neuron is active in the presence of tamoxifen or its metabolite 4-OHT. The actual mice used were double transgenic since the TRAP2 mouse line was crossed with a tdTomato reporter line (Ai14) to form TRAP2:Ai14 mice. These mice underwent context fear conditioning (30 or 720 sec PSI) and were then injected with either 4-OHT or the vehicle. The tdTomato expression (tdTomato+ cells) was nearly absent in the latter mice but was present in the 4-OHT mice. Since the tagging was low in both CA3 and CA1 regions, the authors performed experiments only on the DG region.
The TRAP2:Ai14 mice then underwent the context fear conditioning test. The mice were placed in either Context A or Context B for 720 seconds without footshock and then tissue was tested for tdTomato expression or c-fos expression. It was found that the mice were able to differentiate between similar contexts at the 720 secs PSI (incidence of freezing), but not at the quicker 30 secs PSI, but there was no difference in the number of tdTomato or c-fos expressing cells in any condition. (The lack of effect of tdTomato expression was against previous findings of the authors, but this was explained by smaller effect size and longer 4-OHT labelling time.) Again at 30 secs PSI there was only a low number of cells that were active during learning reactivated and there was no significant difference between contexts. However at 720 secs PSI, cells that were active during learning were preferentially reactivated in Context A (there were more cells reactivated) compared to Context B. Leake and colleagues explained these results as being not completely due to the level of conditioning since there was no significant correlation between reactivation rate and freezing levels in the conditioning context. This implied to the authors that the 720 secs PSI produces a larger neural representation encoding more information and that the degree of discrimination across different PSIs tracks the proportion rather than the absolute number of reactivated cells of the neuronal assembly. A lower proportion of cells reactivated in different contexts correspond to the discrimination.
Leake and colleagues then went on to discuss their results. They found that the total number of c-fos expressing cells increased to a plateau with the duration of the current session independent of what happened in the past. However, there was a subset of cells that were active during initial learning and reactivated at test and these corresponded to the degree of initial learning and subsequent discrimination. These cells were said to constitute the memory. Therefore, poorer learning was said to lead to smaller memory populations resulting in impaired memory performance and behavioural discrimination. The amount of learning was found to be dictated by the duration of the PSI, eg. the longer the PSI (in Leake`s case 720 secs) the greater the time the subject has to sample and encode the features of the context leading to an increased number of memory-encoding cells as well as an overall increase in number of active cells in general. This leads to a better discrimination in Leake`s experiments between Context A and Context B at 720 secs PSI than at 30 secs PSI.
Leake and colleagues concluded that since there was no correlation between reactivated cell numbers and freezing for Context A independent of PSI length, this ruled out the level of fear being responsible for the higher number of reactivated cells at 720 secs PSI for Context A. They proposed that only time determined how much information could be encoded. The authors then went on to say that that their idea appeared to contradict the research of others which showed that neural assemblies representing memory is stereotyped in size and does not differ with experience (Rao-Ruiz; Liu, Taylor – 2-8% of DG cells active regardless of task valence or conditioning) because of likely intrinsic excitability and inhibitory circuits limiting size. They explained that their short 30 secs PSI was sufficient to reveal the variability in neural assembly size with time before the plateau is reached. Therefore, once the plateau size is reached then the neural assembly remains stable, but before this time differences in learning condition can lead to differences in memory content and later performance.
The discussion continued with a look at content. Recruitment of cells in the DG leads to memory specificity with the more cells activated encoding richer information and improving discriminatory behaviour. However, Leake and team found that there was a difference between total number of cells reactivated, time and context. In Context B, freezing differed with the different PSIs but the total numbers of reactivated cells remained the same. However, in Context A freezing was independent of PSI but the number of reactivated cells was different. This indicated to the authors that discrimination was not totally dependent on the absolute number of reactivated cells, but was dependent on the proportion of the assembly relating to Context A that was reactivated in Context B. The authors found that the neuronal assembly at 720 sec PSI was larger, but less engaged in Context B whereas at 30 secs PSI it was smaller but engaged to the same extent in both contexts. This mirrored behavioural generalisation and supported the view that the larger neuronal assemblies support discrimination and that this discrimination is associated with greater cellular reactivation in the conditioning context relative to a different context. However, the authors did state that memory precision at its peak actually only requires around 2% of the total population of cells being activated in this DG area known to be involved in memory.
This led to a discussion about the association between memory precision and cellular reactivation in the subset of c-fos expressing cells. The authors found that the total number of c-fos expressing cells was unrelated to previous learning history, but did reflect the duration of the current session. The increase was not linear and a plateau was reached after about 12minutes. This suggested that the majority of c-fos expressing cells was not involved in the memory storage per se and that the total c-fos expression was indicative of the degree to which the hippocampus was stimulated during the session. This was supported by the observation that the stimulation would decrease as the context became more familiar and would re-occur on re-entry into the context. Only a small proportion of c-fos expressing cells are reactivated at test. These cells were suggested as being involved in new learning or information updating and this supports work by others, eg. c-fos expressing cells would develop place fields characteristics that correspond with contextual identity.
Leake and colleagues concluded their article by saying that assembly size varied with learning condition and larger assemblies support improved neuronal and behavioural discrimination. They also state that fear generalisation depends on the similarity of environment and on the extent of initial learning. Therefore, generalisation can be induced by insufficient contextual learning and can be reduced by further context exposure to allow updating of memories with additional information. This could be of a therapeutic benefit in alleviating overgeneralisation of fear, for example in individuals suffering from post-traumatic stress disorder.
COMMENT
What makes this topic interesting is that it explores the role of the dentate gyrus (DG) in fear conditioning. The work of Leake and colleagues reaffirms the observation that the active neuronal assemblies seen in the DG during contextual fear conditioning tasks are representative of the task being undertaken at the time. The neuronal assembly is assumed to be part of the global neural representation of the information and obtains information from real-time input, working memory processing including memory recall to varying degrees (by that is meant that the level ranges from none when no learning has yet taken place to a full blown recall of event characteristics of previous encounters when full learning has taken place) as well as other information relating directly and indirectly to the task and physiological status at the time. The size of the neuronal assembly firing at any one time is therefore flexible and can adapt rapidly to respond to changing conditions. Leake`s experiments looked at one set of environmental conditions and what happens to firing in the DG when one feature of that condition is changed (smell in the environment). They took mice, used a standard fear conditioning task (place preference) and investigated the size of the neuronal assembly active at the time in the DG when footshock was given (ie. PSI – designated ´end` of test) or not. What they found under the test conditions for their conditioning task (Context A – aniseed smell + pain) was as expected, ie. the longer the time the animal spent in the environment before the footshock, the better the learning of the characteristics of the environment and the better the precision of the memory in terms of discrimination when faced with other contexts. What they also found was that the total level of firing was not associated with learning and past experience, but to the preceding real-time events. This indicates that the role of DG in contextual fear conditioning tasks is likely to be primarily a relay centre of information rather than an actual memory storage site.
The DG is structurally one part of the hippocampus, a brain region in general known to be linked to object recognition, memory formation and long-term memory recall. The DG shows firing characteristics of plasticity associated with long-term changes and reliant on DA1 and DA5 receptors activity and protein synthesis (Deng) plus it contains a large percentage of multi-synaptic boutons (40%) indicating wide connectivity to other brain areas. Input to the DG area comes from the entorhinal cortex (EC) layer 2 and output goes to the other hippocampal areas CA3 via mossy fibres (although there is a direct route EC2 to CA3) and then onto the CA1. The information coming into the EC medial region comes from the parahippocampal cortex which receives input from the WHERE pathway and associative cortices including the representation of space. Input also comes from the lateral EC which receives input from the perirhinal cortex from the WHAT pathway which is associated with input from all sensory regions. Therefore, one part of the firing in the DG is real-time information from the sensory pathways and this would be instrumental in object recognition tasks. In relation to Leake`s contextual fear conditioning task experiments, sensory information would not only be visual, but also olfactory since the presence or absence of pain from the footshock was linked to a particular smell (Context A – aniseed smell + pain, Context B – control, ethanol + no pain). Therefore, Leake`s conditioning task is essentially an ´object recognition` type task rather than the more commonly studied spatial memory-based tasks. The ´object` consists of both visual information (ie. detailing the visual aspects of the environment) and olfactory information (ie. the smell the mouse is subjected to within its real-time environment). The visual information remains relatively constant since the mouse is placed in the same environment for all test trials and the olfactory information is limited to one of two stimuli contributing to the control trial or the test trial.
Therefore, in the case of Leake`s contextual fear conditioning experiment, subjecting the mice to the relevant environments leads to appropriate neuronal firing and neuronal assembly (NA) formation. In their experiments this neural activity was assessed using the c-fos marker whose mRNA is upregulated when neurons fire and action potentials occur. As given above, the NA formed represents firing instigated from various sources. Event characteristics such as visual and in this case olfactory are integrated (Fanselow) and bound together along with other information such as stimulus categories, attentional control demands, concrete feature selective information (primarily from the visual cortex), category selective information (posterior frontal cortex) and control demand selective information from the insula, caudate, anterior cingulate and parietal cortex (Jang). This forms the global neural representation of which activity in the DG forms a part.
As expected, the extent of the NA and the quality (eg. contributing areas, level and type of firing) determine the real-time level and quality of active information and processing possible. In the case of the contextual fear conditioning experiments of Leake and team, the number of c-fos activated cells within the DG increased with time during the learning phase. This suggested as expected that the amount of informational input increased with time spent experiencing the environment. This is supported by work from others that show that place field activity appears within seconds of entering a novel context. This activity continues for many minutes with progressive increases in hippocampal (IEG) gene expression until full stabilisation of the hippocampal place fields occurs after several minutes. This is what Leake and team saw with the NA firing of the DG since the maximal number of cells activated at recall was lower at shorter PSIs (duration of time in learning mode) and that secondly, there was a finite time for learning with a threshold of firing which event characteristic input does not go above. In the DG for Leake`s contextual fear conditioning experiments, the maximum number of c-fos cells was achieved with 180secs PSI whereas in rats (Burman) fear conditioning with footshock reached a maximum much quicker at 30secs PSI. The other areas of the hippocampus also showed different plateau times with the CA3 reaching a maximum at 720 secs PSI and the CA1 even longer at 1800 secs PSI and with a maximum number of cells active at 180secs even higher than the DG. Leake concluded that the results demonstrated that extended time in the context before the footshock resulted in activation of more hippocampal cells dependent on area.
This neuronal firing and NA formation in the DG within the constructs of the contextual fear conditioning experiment with relation to PSI time has behavioural consequences, eg. better ´object` recognition; improved storage and recall; and finally, as shown by Leake and colleagues, better discrimination between similar contexts. With regards to better object recognition, this is a real-time effect and is due to sensory input (visual and olfactory) plus optimal cognitive processing and recall conditions, eg. attention, working memory. (Even though the contextual fear condition task involves footshock which is a ´fear` situation as it induces pain, it appears that activity in the DG is not linked to negative emotional value since Leake found that there were no significant differences between freezing rate and PSI durations.) This fits in with the general role of the hippocampus in object recognition with the neural firing patterns changing with tasks (Lee), feature selectivity given in the CA1 (Buttner) and location and visual information associated with different locations in the hippocampus dependent on what is required from the task at hand (Opris). (The hippocampus is also linked to spatial information and memory, but Leake`s experiments relate to smell and space and therefore, as said before the information is more ´place plus feature` rather than location change.)
Reports of time scale differences with firing (Wang) such as sensory input (second time scale, supports formation of spatial firing fields) and memory related firing activity (second long + millisecond long dependent on septum input- supports sequential firing and episodic memory) leads on to the second behavioural consequence of NA firing in the DG and that is of memory storage and recall relevant to learning and remembering conditioning situations and required behaviour. In general, there is an association between hippocampal functioning and conditioning particularly with regards to memory, eg. oscillation and phase locking in the hippocampus is required (Nokia), conditioned fear leads to increased phosphorylation of CREB (Monti), brief exposure in fear conditioning leads to reconsolidation requiring NK-kbeta1 (de la Finente) and with particular reference to contextual fear conditioning, heightened activity of hippocampal extrasynaptic GABA A receptors impair fear and memory encoding and retrieval (Jovasevic). Memory recall required for the identification of the stimulus (in Leake`s experiments, the context smell) involves activation in the hippocampal areas, eg. rapid item recognition signal in the perirhinal cortex triggers source retrieval in the hippocampus as seen by increased connectivity between the two areas (Staresinal). The retrieval of episodic memory is said to involve the re-enactment of the encoding processes to result in a reinstatement of the original encoded pattern as well as the reinstatement of a transformed representation of the encoded information (Xiao). This implies firing of the same cells in encoding as in retrieval.
Activity in the DG during memory retrieval is also reported, but not all activity of the area is related to this function. Leake`s experiments on DG involvement in the recall of fear memory during the conditioning test showed that the maximum number of activated cells as shown by c-fos expression was reached at the same re-exposure duration for the two PSIs tested with a maximum number of active cells achieved with 180secs PSI. This implied that the same cells were involved in both encoding and recall. This implies that the cells have not been physiologically altered by previous experience and therefore, is indicative that the cells of the DG do not store event characteristics per se in the form of long-term memories since long-term memory formation leads to physiological changes to the cell that increases efficiency of firing. If this was the case, then firing activity during recall would be quicker as evidenced in human memory retrieval which occurs at a faster timescale than during encoding (Yaffe). Therefore, the total number of DG cells activated in retrieval corresponds not just to memory, but to the real-time session, ie. the NA consists of firing cells responsible for incoming sensory information, working memory processing (including memory recall) plus other required cognitive capabilities. This observation matches that seen for the hippocampus in general, eg. the degree of overlap between IEG+ cells activated after conditioning and testing was reported as low suggesting that only a fraction of the original hippocampal cell population was incorporated into the memory (Liu).
The observation by Leake and team that the same cells involved in encoding are active in retrieval supports the view that the DG acts primarily as a relay and that activity in the DG relating to memory event characteristics retrieved is likely to be only a small proportion of the total active cells and Leake and team confirmed this using their c-fos tagging method of TRAP2 cells. These cells were active as expected in both learning and recall situations and probably represent the event characteristics relating to environment location plus olfactory information. Leake found that learning time associated particularly with the smell feature of the environment correlated to the size of this subset of cells. Short PSIs led to same number of event characteristics with both smell contexts and found that longer PSIs led to a greater number of double labelled cells and hence, were linked to greater details of the event characteristics represented in the NA. This observation was also seen for the CA3 and CA1 areas except the time of the PSI was longer (720secs) before maximum was reached.
The idea that the hippocampus plays an informational relay role in long-term memory supersedes the idea that the hippocampus is a site for long-term information storage. Long-term storage of details themselves is likely to be in the higher brain regions with the hippocampal cells and others at this level being the ´neuronal conduit` to these regions, eg. memory formation is part of the route to the prefrontal cortex, visual cortex and not responsible for event characteristic memory in object recognition (Broadbent). Therefore, the hippocampal cells will respond to repetitive firing by undergoing physiological changes to make the firing and information transfer more efficient (eg. LTP and neurogenesis with a stabilisation of connectivity between the cortex and hippocampus connections for recall – Frankland), but these changes are likely to be less stable and hence, firing capability is more adaptable. This is evidenced with for example: new information requiring protein synthesis and neurogenesis with lesions of the hippocampus leading to loss of new information, but retention of familiar (Winocur); a shift from hippocampus dependent object recognition memory with time (Broadbent); and mice lacking in NMDARs in the DG and CA1 areas acquire spatial reference memory from water maze tasks the same as controls (Bannerman). The advantage of having the hippocampus and with relevance to this blog, the DG, as part of the ´relay` system and not the main storage site is that it allows the hippocampus to be part of the ´information/task processing` system rather than purely an information ´storage` site. With regards to Leake`s contextual fear conditioning experiments then, the DG would be part of the ´working memory` system which would take part in the decision-making of whether the real-time context would result in footshock or not. Therefore, the quality and extent of the NA representing the context will determine the level of discrimination.
Discrimination requires decision-making whether conscious or non-conscious. In general, the hippocampus is known to play a role in decision-making and working memory (Duncan, Fixmacher) in processes that require glutamate excitatory and inhibitory firing, eg. NMDAR antagonists reduce working memory (Takedi) and task irrelevant information is affected by affected by NMDA R antagonists (Gage). The hippocampus in this case can support at least two active NAs so that comparisons can be made as supported by the report that the hippocampus plays a role as comparator capable of individual representations of overlapping inputs (Zeamer). Within this concept, the DG appears to be linked to have a more specific functioning role where similar situations require comparison in order that actions can be taken. This observation is supported for example by reports that the DG plays a role in discrimination with maximal similarity between objects place pairs, but with no overlap then there is no functioning (Lee); and in mice undergoing a spatial memory water maze task functioning is important for the selection of alternative response that arise from competing or overlapping memories (Bannerman). With reference to Leake`s article, the DG is reported as being required for fear expression when the mice subjects have to distinguish between similar contexts with one neutral and the other feared (Bernier). Inhibition of the DG is said to lead to increased generalisations of fear to an unfamiliar context that is similar to a feared one and impaired fear expression in the conditioned context when it is similar to a neutral one (Bernier). This type of functioning is also seen in Leake and colleagues experiments where DG activity determines discrimination between the two olfactory contexts. In this situation, the shorter PSIs would be equivalent to Bernier`s inhibition and are linked to poorer learning and generalisation whereas better discrimination would be observed with longer PSIs, ie. longer periods of learning before the footshock event. In Leake`s experiments activity of the maximum number of c-fos cells and maximum level of discrimination was achieved with 180secs PSI.
In Leake`s experiments the different contexts produced different outcomes that the mouse could do nothing about and therefore, discrimination of the contexts was only linked to expectation, ie. aniseed smell leads to footshock and pain, ethanol smell, nothing. Therefore, neuronal firing and NA formations with the exception of olfactory information for example, remain similar during the course of the experiment. In other scenarios, however, discrimination may result in a new course of action following information processing and decision-making, eg. movement to take the animal out of the reach of the footshock source. In these cases, new information would have to inputted and stored as learning of the sequence of events takes place because of trial-and-error feedback, eg. no footshock if move to the left. This new learning would supersede the information represented by the original NA. Therefore, new memories and new patterns of efficient firing in the DG would be formed although not necessarily all features of the existing NA would be overwritten, eg. visual information about the location`s walls. This process of updating is shown by reports for example that newly generated axons juxtapose or displace previously established synapses which leads to the pre-existing NAs being degraded (Martrig) and lesions of the hippocampus leads to loss of new information, but retention of familiar (Winocur). One aspect of this appears to be linked to DG functioning since it has been seen that inhibition of the DG impairs fear extinction (Bernier). This inability of fear memories to be overwritten and over-generalisation of event characteristics leading to inaccurate object recognition and risk assessment are seen to be problems occurring in some psychological conditions. Leake and team correlated their findings using mice to human post-traumatic stress disorder (PTSD) saying that the formation of impoverished memories may be associated with the generalisation of events with relation to fear as seen in PTSD. The generalisation is linked to lack of details due to short PSI durations, ie. time for learning when the subject is in the fear environment. However, it should be said that PTSD is not necessarily linked to duration of experience or lack of details, but may be more associated with repetition of events and inability to process information beyond the event itself. Therefore, although PTSD is not necessarily linked to lack of event information, it does give an indication that supplementation to quality of recall may lead to a shift away from generalisation of situations resulting in fear to discrimination of actual situation and fear-independence. This type of shift would be achieved by reactivation of the memory via re-exposure allowing modification of the memory engram (NA) during the reactivation in a process already said to be a possible therapeutic mechanism for PTSD (Radiske). However, two factors should be considered. The first is that if Leake`s hypothesis is valid then the PTSD sufferer needs to spend more time in the ´fear context` in order to counteract the impoverished memory formation of the original exposure and secondly, that the original memory may not be fully over-written and therefore, recurrence of fear can occur. This second factor comes from the observation that fear extinction forms new memories, but does not fully erase the originals (Liu). It is shown in contextual fear conditioning that some hippocampal cells re-map in both fear conditioning and extinction (Wang) whereas others respond predominantly during extinction. Other active cells are stable throughout learning and these, just like in Leake`s experiments and in the explanation given above, provide the encoding of static unchanging aspects of the environment (Wang). Therefore, it is unlikely that counteracting the impoverishment of the quality of the memory by exposing to more information would only be successful in the treatment of PTSD if it was accompanied by modified informational processing on NA reactivation and event recall.
Therefore, Leake`s experiments show that the DG is part of the hippocampal relay system of information input, transfer and processing and it plays a role in contextual fear conditioning. In this particular scenario of Leake`s, the DG appears to be involved in the comparison of similar contexts and hence, level of detail represented by the active neuronal assembly will determine success at event discrimination as well as aiding decision-making. Duration of learning of event characteristics can therefore, affect neurochemically the size and content of the neuronal assemblies formed and reactivated, as well as affecting the corresponding behavioural performances. Impoverishment of detail from short learning times leads to fear generalisation whereas longer durations in the context can lead to more successful discrimination. It appears that a maximum number of cells are involved in the DG and a plateau reached independent to past experience demonstrating that neuronal assembly activity in the DG area relates to the real-time situation rather than past experiences.
Since we`re talking about the topic……………………….
…..object recognition is said to be impaired in older female mice and in ovariectomised mice with co-administered diphenyl diselenide and 17beta estradiol (Da Rocha). If Leake`s experiments are repeated using older female mice, would the timing of PSI for the maximum learning and behavioural discrimination have to be increased in order to counteract the impaired working memory performance?
…….would the administration of rolipram (a phosphodiesterase inhibitor) which is said to lead to increased memory consolidation of conditioned fear (Monti) result in reduced PSI durations being required to reach maximums for memory retrieval and discrimination?
…….can we assume that activity in the DG would also show the same discrimination activity as in Leake`s experiments when pleasant and unpleasant smells are paired during sleep with specific auditory tones (Arzi) and the subjects are re-exposed to the tones during waking-time? Would the introduction of a third context (smell) and paired sound produce learning at shortened PSIs due to the improved encoding and recall capabilities which are said to be associated with the novel context being associated with the existing knowledge (Brod)?
neurochitchat 