Title: Altered Arc transcription in the hippocampus and cortex of mice lacking
the GluA1 AMPAR-subunit
Name: Mik Schutte
Student ID: 11651067
Assessors: Dr. Hui Xiong & Dr. Jan A. Gorter
Submission date: 24-01-2020
Altered Arc transcription in the hippocampus and
cortex of mice lacking the GluA1 AMPAR-subunit
Mik Schutte
Abstract
α-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid receptors (AMPARs) are a primary mediator of excitatory synaptic transmission in the brain. Their synaptic exocytosis contributes to long-term synaptic potentiation (LTP) which strengthens synapses resulting in more connected neurons. LTP is believed to regulate memory formation, by connecting neurons into an engram; the neural
representation of a memory. AMPAR-subunit GluA1 knockout (GluA1 KO) mice show impaired memory formation and lowered synaptic transmission. However, it’s unknown if these GluA1 KO mice show the same number of memory-engram cells compared to their wildtype (WT) littermates. Here we use Arc::dVenus mice, expressing fluorescent dVenus on the activity-regulated cytoskeleton-associated protein (Arc) promotor to label engram cells in the dentate gyrus (DG), Cornu Ammonis 1 (CA1) and the primary visual cortex (V1) of WT and GluA1 KO mice, without and after
fear-conditioning. GluA1 KO mice show more Arc+ cells within the DG and CA1, but not within V1,
compared to their WT littermates. These results suggest a role for AMPAR-subunit composition in Arc-related mechanisms.
Introduction
α-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid receptors (AMPARs) are one of several receptors that mediate excitatory synaptic transmission within the brain (Greger et al., 2017; Hammond, 2008). The trafficking of AMPARs is a mechanism through which synaptic plasticity is conveyed. More specifically, the exo- and endocytosis of AMPARs alters the strength of a synapse which regulates the plasticity’s direction either towards long-term potentiation (LTP) or long-term depression (LTD) (Kessels & Malinow, 2009). Both LTP and LTD are commonly revered to as forms of Hebbian plasticity. AMPARs are assembled as homo- or heterotetramers from a merger of four subunits, GluA1, GluA2, GluA3 and GluA4. Additionally, AMPAR-trafficking and function are modified by subunit-specific protein binding, post-translational modification and auxiliary proteins (Greger et al., 2017; Diering & Huganir, 2018). However, this paper will mainly focus on the effect of AMPAR-subunit composition. Throughout the mammalian brain, the AMPAR composition shows regional as well as temporal diversity (Schwenk et al., 2014). Yet within the mature hippocampus, most excitatory neurons contain GluA1/2 or GluA2/3 heteromers (Wenthold et al., 1996). Through the exocytosis of AMPARs into the synapse, LTP connects neurons, contributing to the formation of neuronal engrams. An engram is believed to be the neural representation of a memory. Research implies that an engram consists of a sparse population of neurons linking various brain regions like the cortex and
hippocampus (Tonegawa et al., 2015; Rao-Ruiz et al., 2019b). Moreover, it’s representational capacity is influenced by its sparsity or number of contributing cells. These engrams are generally identified by labelling immediate early genes (IEGs) (Tonegawa et al., 2015; Park et al., 2016; Rao-Ruiz et al., 2019a, Rao-Ruiz et al., 2019b). Furthermore, optogenetic stimulation of neurons labelled during fear-conditioning results in a freezing-response, further emphasizing the engram’s representational capacity (Lacagnina et al., 2019). The hippocampal sub-structure, the dentate gyrus (DG), is especially involved in the formation of contextual fear memory engrams (Hernández-Rabaza et al., 2008; Lacagnina et al., 2019). Additionally, the GluA1 subunit plays a role in the induction of LTP via AMPAR incorporation (Boehm et al., 2006; Diering & Huganir, 2018). This process of incorporation occurs both during learning and LTP induction. (Kessels & Malinow, 2009). Even though LTP classically occurs though the incorporation of GluA1-containing AMPARs, the synapse does demonstrate remarkable flexibility to potentiate via different glutamate receptor configurations (Granger et al., 2012; Renner et al., 2017).
Albeit that AMPAR exocytosis facilitates the formation of memories, ever-advancing incorporation may cause the synapse to become saturated, losing its capacity to encode information and form new engrams (Moser et al., 1998). In order to prevent the saturation of synapses, the synaptic strength has to be submissive to homeostatic regulation. Homeostatic regulation is believed to adjust the strength of every excitatory synapse within a neuron. Seeing as Hebbian plasticity alters the strength of individual synapses and homeostatic regulation alters the excitability of a neuron it’s conceivable that they influence each other. Moreover, homeostatic regulation, like Hebbian plasticity, functions through the trafficking and modification of AMPARs. However, previous episodes of Hebbian plasticity should not be nullified by homeostatic regulation, for this might impair the integrity of existing engrams (Shepherd et al., 2006; Rial Verde et al., 2006; Turrigiano, 2008; Diering & Huganir, 2018). The direction of homeostatic regulation is based upon a neuron’s activity. For example, neuronal
hyperactivity reduces the AMPAR-mediated current and synaptic AMPAR expression (Turrigiano et al., 1998; O’Brien et al., 1998). Diering and Huganir have proposed that there exists an AMPAR code. Conform this code, AMPAR-subunit composition is one of several factors determining a hierarchy of AMPAR-trafficking. This signifies that different AMPAR variants are more or less likely to be trafficked during Hebbian plasticity and homeostatic regulation (Diering & Huganir, 2018).
This is in line with the aforementioned literature on the importance of GluA1-containing AMPAR trafficking within LTP. Seeing as LTP underlies engram formation and the GluA1-subunit is heavily involved in LTP, it is conceivable that this subunit influences engram formation itself. Moreover, GluA1 knockout (GluA1 KO) mice show a lowered synaptic transmission and a decreased memory
(Supplementary Fig. 1a,b & 2a,b). However, the effect of a GluA1 KO on engram formation remains unknown. Therefore, we set out to answer whether GluA1 KO mice contain the same amount of engram cells within the hippocampal sub-regions DG and Cornu Ammonis 1 (CA1), compared to their wildtype (WT) littermates. Additionally, engram cells within the primary visual cortex (V1) were also analysed. Since the presence of GluA1 is necessary for LTP (Zamanillo et al., 1999; Zhou et al., 2018); LTP occurs during memory formation (Kelleher et al., 2004; Whitlock et al., 2006) and engrams are the neural representation of a memory (Tonegawa et al., 2015; Park et al., 2016; Rao-Ruiz et al., 2019b); it’s hypothesized that GluA1-KO mice contain a lower number of engram cells.
To verify this, fear-conditioning-induced memory engram cells were labelled using the IEG of activity-regulated cytoskeleton-associated protein (Arc) (Gouty-Colomer et al., 2016; Gruene et al., 2016; Lacagnina et al., 2019). Brain regions were chosen based on the DG’s and CA1’s role in memory and engram formation (Yassa et al., 2011; Park et al., 2016; Rao-Ruiz et al., 2019a, Rao-Ruiz et al., 2019b) as well as CA1’s thoroughly researched electrophysiological properties. V1 was selected based on the well-studied effect of homeostatic regulation in the visual system (Turrigiano, 2008). Based on the aforementioned connection with LTP and engram formation, it’s hypothesized that the lack of GluA1 results in a debilitated engram, which shows through a decrease in the number of Arc+ cells.
Materials & Methods
MiceArc::dVenus mice expressing the destabilized fluorescent reporter of dVenus, coupled to the Arc promotor (gifted by prof. dr. Steven Kushner, Erasmus University Rotterdam) and WT mice (Harlan, The Netherlands) were backcrossed for over 10 generations to establish C57BL/6J mice.
Subsequently, Arc::dVenus mice were crossed with GluA1 KO mice (obtained from dr. R. Huganir) to establish the Arc::dVenus-GluA1KO mice used for the experiment. The animals had ad libitum access to food and water and were kept on a 12h light/dark cycle. Housing conditions for the group-housed mice were typical standard housing. As for the solitary-housed mice, these were solitarily
accomodated for one week prior to experiments. All mice were 5 to 10 weeks of age when sacrificed. The animal welfare committee of the University of Amsterdam approved the experimental protocol.
Genotyping
The genotype of the mice was confirmed prior to analysis. Raw genetic material was provided as an ear snip and genomic DNA was extracted by adding the pre-mixed (4:1) extraction solution (N3910 Sigma) and tissue preparation solution (T3073 Sigma) to each sample. Samples with mix were then
incubated at room temperature for 1 hour and consecutively incubated at 95˚C for 3 minutes. Following, incubation the neutralization solution B (N3910 Sigma) was added. PCR was then used to amplify the Arc and the GluA1 gene (Supplementary Fig. 3 & 4). The product, containing the amplified genetic material was then run through a gel electrophoresis with 1% agarose at 120V for 45 minutes. Bromouridine was used to visualize the DNA-strands on the gel. Genetic information was used to evenly distribute mice into conditions (Supplementary Fig. 5).
Contextual fear-conditioning
The contextual fear-conditioning paradigm was conducted in a standard fear-conditioning chamber provided with a stainless steel, shock generating grid. Upon placement in the chamber mice could accustom to it for 2 minutes. Following the acclimatization period, 3 consecutive footshocks (0.8 mA, 1 s) each spaced by 1 minute were delivered through the electrical grid. Succeeding the final shock was a 2-minute rest period prior to the ending of the trial (Fig. 1a).
Tissue preparation
2 hours post fear-conditioning, the animals were deeply anaesthetized with pentobarbital (50 mg/kg) and perfused with phosphate-buffered saline (0.1 M PB, pH: 7.4) and paraformaldehyde (4%).
Perfused mice were sacrificed by decapitation. Extracted brains were kept in PFA (4%) and were incubated in 15% sucrose for 4 hours and 30% sucrose overnight, prior to slicing. Brains were sliced in a coronal fashion into 40 µm for the group-housed and 30 µm thick sections for the solitary-housed mice using a R. Jung AG Heidelberg microtome. Slice thickness was decreased in solitary-housed mice to facilitate a more precise count. Slices were stored in PBS with 10% sodium azide (1000:1) at 4 ˚C. 8 slices (Bregma -1.34 mm to -3.16 mm) for DG and CA1 and 6 slices (Bregma -2.70 mm to -3.64 mm) for V1 were mounted on slides and vectashield with DAPI (Vectashield Mounting Medium with DAPI, H-1200, Vector Laboratories Inc.) was used for coverslipping.
Quantitating analysis
Fluorescent dVenus and DAPI positive cells were imaged using a Nikon DS-Ri2 microscope at X10 magnification. Innate dVenus fluorescence was imaged using a 510 µm excitation wavelength and 460 µm for DAPI. Images were z-stacked (±10 µm) around a manually set focus. Images were imported to ImageJ where grayscale images were used for analysis. DG dVenus+ density was acquired by
determining the surface area of the DG granular layer and counting the dVenus+ cells within this area
manually. CA1 dVenus+ and DAPI+ cells were counted at a self-defined region of interest. As for V1
dVenus+ density, these were counted using a modified ImageJ Cell Counter plugin.
Statistical analysis
To statistically test the differences between groups a Student’s T-test was used. Student’s T-test was preferred over ANOVA, due to selective comparison (between genotype or between condition). Prior to statistical testing, all data was checked on normality of distribution via a Shapiro test and on equal variance through a Levene’s test. If these checks were not met, log-transformation was tried to counter this. When checks were still not met, the assumptions for using the Student’s T-test would have been violated and thus resulted in the usage of the non-parametrical alternative; the Mann-Whitney U test. Data was determined to be significant at α ≤ 0.05. No correction for multiple comparisons was conducted because of the exploratory nature of this study.
Results
Increased number of dVenus+ cells in the dentate gyrus of GluA1 KO mice.
In order to determine the role of GluA1 in engram formation, the number of dVenus+ cells in the DG
was compared between GluA1 KO mice and WT littermates. GluA1 KO mice showed a significant increase in their DG dVenus+ cell density in both the home-caged (Whitney-U: W=0, p=0.016) as well
as the fear-conditioned (Whitney-U: W=0, p=0.008) mice, compared to their WT littermates (Fig. 1b,c). These results lead to the rejection of the previously stated hypothesis that GluA1 KO mice would, in fact, contain less Arc+ cells and thus show a lower dVenus+ density. However,
fear-conditioning the mice did not seem to affect the dVenus+ density; neither in the WT nor in the GluA1
KO mice. This finding questions the representative power of dVenus+ cells as engram cells within this
experiment. Which is why we can only conclude an increase in dVenus+ cell density of GluA1 KO mice
and not an increase in engram cells compared to their WT littermates. Taken together, this data shows that GluA1 KO mice show an increased number of Arc+ cells within the DG, regardless of
fear-conditioning.
Increased dVenus/DAPI percentage in CA1 of GluA1 KO mice.
The CA1, an additional substructure of the hippocampus has been extensively studied for its
electrophysiological properties and it’s alterations of these properties upon the induction of Hebbian plasticity. To examine the effect of a GluA1 KO on engram formation, the same pictures, as used for DG analysis, were inspected. Both the number of DAPI+ cells as the number of dVenus+ cells were
counted. Subsequently, a percentage of dVenus+ cells per DAPI+ cells was deducted to systematically
determine engram formation. Percentage was used instead of density to make comparison between literature easier. The CA1 data showed that there was an increase in the percentage of dVenus+ cells
in the CA1 of fear-conditioned mice once GluA1 was knocked out (Whitney-U: W=23, p=0.032). This raise in dVenus+ percentage did not reach significance in home-caged WT versus home-caged GluA1
KO mice (Whitney-U: W=18, p=0.063) (Fig. 1d,e). Thus, within the CA1, the percentage of dVenus+
cells per DAPI+ cells is elevated in fear-conditioned GluA1 KO mice compared to their WT littermates
and a similar trend is present in their home-caged counterparts. Thus, that GluA1 KO mice have more Arc expression within the CA1 compared to their WT littermates when fear-conditioned. Moreover, a similar effect is implied between home-caged WT and GluA1 KO mice.
Implied dVenus+ density increase in fear-conditioned WT and GluA1 KO mice.
In order to study the effect of a GluA1 KO in the cortex, cells within layers II & III of V1 were analyzed for Arc expression through dVenus-positivity. Counting was conducted by a modified Cell Counter plugin in ImageJ. Data showed a non-significant trend in which dVenus+ cell density in layer II/III of V1
increased upon fear-conditioning for both WT (Whitney-U: W=3, p=0.110) as well as GluA1 KO (Whitney-U: W=4, p=0.095) mice (Fig. 1f,g). Interestingly, a genotype effect was absent in layers II & III of V1. This shows that fear-conditioning does affect dVenus+ within V1 irrespective of genotype and
that a GluA1 KO doesn’t alter expression levels. All in all, it appears that knocking out GluA1 doesn’t affect Arc expression within layers II/III of V1, nor does it impede an effect of fear-conditioning within this brain region.
Fig. 1) dVenus+ cells in WT and GluA1 KO hippocampus and V1. a) Fear-conditioning paradigm. Mice received 3 consecutive 0.8mV
footshocks of 1 second in duration with 2 minutes prior and after fear conditioning. Sacrifice was conducted 2 hours post conditioning. b) Number of dVenus+
cells per 1.0 mm2
of DG granular layer for home-caged wildtype (HC-WT: n=4; µ±SEM=322.5±92.1), fear-conditioned wildtype WT: n=5; µ±SEM=414.0±65.9), home-caged GluA1 KO (HC-KO: n=5; µ±SEM=1546.3±493.2) and fear-conditioned GluA1 KO (FC-KO: n=5; µ±SEM=1717.7±243.3). Mann-Whitney U test; effect of knockout: HC-WT vs HC-(FC-KO: W=0, p=0.016, FC-WT vs FC-(FC-KO: W=0, p=0.008.
c) Representative images of the DG from each condition. d) Percentage of dVenus+ cells against DAPI+ cells in CA1 for all conditions (WT-HC:
µ±SEM=28.5±7.96, WT-FC: µ±SEM=40.0±3.25, KO-HC: µ±SEM=45.2±3.00 & KO-FC: µ±SEM=57.5±4.94). Mann-Whitney U test; effect of knockout: HC-WT vs HC-KO: W=18, p=0.063, FC-WT vs FC-KO: W=23, p=0.032. e) Representative images of the CA1 from each condition. f) Number of dVenus+ cells per 1.0 mm2 of layer II/II V1 cortex for WT-HC (µ±SEM=134.1±84.6), WT-FC (µ±SEM=325.4±63.4), KO-HC
(µ±SEM=115.8±15.7) & KO-FC (µ±SEM=407.8±60.1). Mann-Whitney U test; effect of fear-conditioning: WT vs FC: W=3, p=0.110, HC-KO vs FC-HC-KO: W=4, p=0.095. g) Representative images of V1 layer II/II. Data are presented as mean ± SEM. Scale bar: 200µm.
Social impoverishing suggests a decrease in Arc+ cells in home-caged GluA1 KO mice.
Previous research has shown that a social environment affects Arc expression (Pinaud, 2009). Since our study utilizes group-housed mice it is possible that these housing-conditions contribute to the results described above. To control for this effect the same experiment was carried out using solitary-housed mice. However, only home-caged mice of either genotype were used in the solitary-solitary-housed condition. Additionally, no statistical tests were performed due to the low sample size (n=2) of the solitary-housed condition. Data is controlled for different slice thickness. Analysis of the DG for the solitary-housed mice showed a decreasing trend in dVenus+ density for GluA1 KO mice, compared to
their group-housed opposites. These findings imply that an impoverished social environment through solitary-housing lowers the number of Arc+ cells within the DG for home-caged GluA1 KO mice
compared to their group-housed counterparts (Fig. 2b). As for CA1, results also allude towards a tendency to decrease dVenus+ percentage in home-caged GluA1 KO mice upon solitary-housing (Fig.
2d). This solitary-housing induced decrease in dVenus positivity appears to constrain itself to GluA1 KO mice and is not observed in WT mice (Fig. 2a,c). The presence of a social impoverishing induced decrease in Arc positivity in the hippocampus of GluA1 KO mice shows that the social environment might indeed affect Arc expression.
Fig. 2) The effect of solitary-housing on dVenus positivity in home-caged WT and GluA1 KO mice. a) dVenus+ density of group- (n=4) and
solitary- housed (n=2) HC WT mice within the DG. b) dVenus+ density of group- (n=5) and solitary-housed (n=2) HC GluA1 KO mice within the
DG. Data suggests solitary-housing might reduce dVenus+
dVenus+/DAPI+ percentage of group- (n=4) and solitary-housed (n=2) HC-WT mice in CA1. d) dVenus+/DAPI+ percentage of group- (n=5) and
solitary-housed (n=2) HC GluA1 KO mice in CA1. Data implies a decrease in dVenus+ percentage.
Discussion
This study set out to identify the role of the GluA1-subunit on memory engram sparsity within the hippocampus and cortex. This was established by determining whether GluA1 KO mice contained the same amount of engram cells within the hippocampal sub-regions of the DG and CA1 as well as V1 of the cortex, compared to their WT littermates. First, it was shown that a GluA1 deficiency through knockout increases the number of Arc+ cells within the granular layer of the DG. However, since our
experiment failed to induce an increase in Arc+ cells upon fear-conditioning within WT mice it is
unsure whether Arc truly marks a memory engram in this setup. Therefore, it cannot be concluded that, since knocking out GluA1 increases the number of Arc+ cells, a GluA1 KO also results in an
increase of engram cells within the DG. Secondly, this study shows a similar increase in Arc+ cells
within the CA1 of fear-conditioned mice and implies a comparable effect in home-caged mice. Taking these findings together, we conclude that GluA1 KO mice do not contain the same, but an increased amount of Arc+ cells within the hippocampal sub-regions like the DG and CA1, compared to their WT
littermates. This contradicts the earlier stated hypothesis that a GluA1 KO would decrease the number of Arc+ cells. Additionally, Arc+ cells within layer II/III of V1 were assessed for both genetic
conditions. Results showed no difference between WT and GluA1 KO mice in the number of Arc+ cells.
Interestingly, fear-conditioning did increase the amount of Arc+ cells in V1 and this was the case for
both genotypes. However, it’s debatable whether these cells can be classified as memory engram cells for they don’t necessarily represent a memory. All in all, we conclude that GluA1 KO mice contain the same number of memory engram cells within layer II/III of V1 compared to their WT littermates and imply that fear-conditioning increases the engram cell count in both WT and GluA1 KO mice. This shows a difference in engram associated mechanisms, such as Arc positivity, within the cortex compared to the sub-cortical structure of the hippocampus.
It is important to emphasize that this study was unable to replicate an increase in Arc+ cells upon
fear-conditioning in WT mice. One possible explanation for this is that all mice were highly stressed at baseline, which was also implied by their behaviour (biting, plucking etc.). These high stress levels can interfere with the formation of a stress-induced memory via fear-conditioning. Additionally, stress increases Arc expression which could also impair the identification of an increase in Arc expression upon fear-conditioning (Ons et al., 2004). It’s possible that the social environment of group-housing contributed to the increased stress levels (Pinaud, 2009). Several articles researching Arc use solitary-housed mice to control for this effect. Nevertheless, we were unable to decrease Arc expression by solitarily housing the WT mice, yet it did seem to work on GluA1 KO. However, this might be the case due to the low sample size of only two mice. Moreover, we have yet to determine whether solitary-housing does enable a fear-conditioning effect in WT mice. Additionally, solitary-housed mice brains were sliced to 30 µm thick slices whereas group-housed brains were sliced to 40 µm thick slices. Even though we controlled for this in the data processing, follow-up research studying the effect of social impoverishing should indisputably use same slice thickness. Also, for group-housed CA1 analysis, the images were focused upon the DG. This means that it was difficult to identify individual cells. The lack of image sharpness might make these results less accurate.
The conclusion that mice lacking the GluA1-subunit show more Arc+ cells within the hippocampus
implies a role for AMPAR-subunit composition to influence Arc+ mechanisms. This possibly places
subunit-composition upstream of Arc. Interestingly, Arc has been linked to the endocytosis of GluA2/3 AMPARs (Rial Verde et al., 2006; Shepherd et al., 2006). One possible link between Arc-mediated GluA2/3 endocytosis and increased Arc positivity in a GluA1 KO lies within the neuron’s coping capabilities. It is possible that neurons lacking the GluA1-subunit cope with the GluA1-deficit by incorporating more GluA2/3 AMPARs into their synapse. An increased synaptic presence of GluA2/3 AMPARs could then cause an overexpression of Arc to enable sufficient Arc to mediate GluA2/3 endocytosis. This further contributes to the theory of the AMPAR code and the upstream contribution of AMPAR-subunit composition in Arc mechanisms. However, this does make the assumption that GluA1 KOs contain more GluA2/3 AMPARs which, as of yet, hasn’t been proven scientifically. A study assessing a possible increase in synaptic GluA2/3 AMPARs in a GluA1 KO would back this implied mechanism up. If AMPAR-subunit composition is indeed linked in an upstream fashion to Arc expression through GluA2/3 endocytosis, an absence of the GluA3-subunit should decrease the number of Arc+ cells. Preliminary data from GluA3 KO mice does imply a decrease in Arc+
cells compared to a GluA1 KO, yet it’s impossible to draw any conclusions as of yet (Supplementary Fig. 6a-b). Research further developing the effect of AMPAR-subunit composition on Arc expression could also make use of organotypic slices from GluA1- and GluA3- KOs, for In vitro work prevents any effect of housing to contribute to the results and is worthwhile to use.
Additionally, GluA1 containing AMPARs might control a neurons intrinsic excitability, which is one of the factors determining a neurons contribution to a memory engram (Rao-Ruiz et al., 2019b). GluA1 KO mice show a lowered synaptic transmission which suggests a decreased excitability
(Supplementary Fig. 1a-b). The absence of GluA1 might disrupt the neuron’s intrinsic excitability, which causes a decrease in engram sparsity through an increase in cells contributing to the engram. Yet, further research is necessary to determine GluA1’s role in resolving a neuron’s recruitment to an engram. Another determining factor in engram recruitment is network architecture, such as local microcircuitry. Interneurons provide inhibitory currents to the network which constrain engram sparsity (Rao-Ruiz et al., 2019b). Within some interneurons, the direction of homeostatic plasticity is opposite to that of principal neurons. For instance, parvalbumin-expressing interneurons increase their excitatory post-synaptic strength upon network hyperactivity, which increases their inhibitory output (Chang et al., 2010). Due to GluA1 containing AMPARs incorporation being mostly responsible for homeostatic scaling up, it is likely that this process is disrupted in a GluA1 KO (Diering & Huganir, 2018). Thus, due to GluA1 absence, interneurons within the hippocampal microcircuitry fail to homeostatically scale up their post-synaptic strength, which causes a decreased inhibitory output to principal neurons and a failure to constrain engram sparsity. Further research should take the local microcircuitry of the hippocampus in to account instead of solely focusing on the principle neurons. Possibly, a selective GluA1 KO for interneurons might clarify the role of AMPAR-subunit composition in engram constrainment by local microcircuitry.
References
1. Boehm, J., Myoung-Goo, K., Johnson, R.C., Esteban, J., Huganir, R.L. & Malinow, R. (2006) Synaptic Incorporation of AMPA Receptors during LTP Is Controlled by a PKC Phosphorylation Site on GluR1. Neuron 51, 213-225.
2. Chang, M.C., Park, J.M., Pelkey, K.A., Grabenstatter, H.L., Xu, D., Linden, D.J., Sutula, T.P., McBain, C.J. & Worley, P.F. (2010). Narp regulates homeostatic scaling of excitatory synapses on parvalbumin expressing interneurons. Nat Neurosci 13, 1090-1097.
3. Diering, G.H. & Huganir, R.L. (2018) The AMPA Receptor Code of Synaptic Plasticity. Neuron
100, 314-329.
4. Gouty-Colomer, L.A., Hosseini, B., Marcelo, I.M., Schreiber, J., Slump, D.E., Yamaguchi, S., Houweling, A.R., Jaarsma, D., Elgersma, Y. & Kushner, S.A. (2015). Arc expression identifies the lateral amygdala fear memory trace. Molecular Psychiatry 21, 364-375.
5. Granger, A.J., Shi, Y., Lu, W., Cerpas, M. & Nicoll, R.A. (2012). LTP requires a reserve pool of glutamate receptors independent of subunit type. Nature 493, 495-500.
6. Greger, I.H., Watson, J.F. & Cull-Candy, S.G. (2017). Structural and Functional Architecture of AMPA-type Glutamate Receptors and Their Auxiliary Proteins. Neuron 94, 713-730.
7. Gruene, T., Flick, K., Rendall, S., Cho, J.H., Gray, J. & Shansky, R. (2016). Activity-Dependent Structural Plasticity After Aversive Experiences in Amygdala and Auditory Cortex Pyramidal Neurons. Neuroscience 328, 157-164.
8. Hammond, C. (2008). Cellular and Molecular Neurophysiology (3rd ed.). Elsevier Ltd.
9. Hernández-Rabaza V., Hontecillas-Prieto L., Velázquez-Sánchez C., Ferragud A., Pérez-Villaba A., Arcusa A., Barcia, J.A., Trejo, J.L. & Canales, J.J. (2008). The hippocampal dentate gyrus is essential for generating contextual memories of fear and drug-induced reward. Neurobiol Learn Mem 90(3), 553–559.
10. Kelleher, R.J, Govindarajan, A. & Tonegawa, S. (2004). Translational Regulatory Mechanisms in Persisten Forms of Synaptic Plasticity. Neuron 44, 59-73.
11. Kessels, H.W. & Malinow, R. (2009). Synaptic AMPA Receptor Plasticity and Behavior. Neuron
61, 340-350.
12. Lacagnina, A.F., Brockway, E.T., Crovetti, C.R., Shue, F., McCarty, M.J., Sattler, K.P., Lim, S.C., Leal Santos, S., Denny, C.A. & Drew, M.R. (2019). Distinct hippocampal engrams control extinction and relapse of fear memory. Nat Neurosci 22, 753–761.
13. Moser, E.I., Krobert, K.A., Moser, M.B. & Morris, R.G. (1998). Impaired spatial learning after saturation of long-term potentiation. Science 281(5385), 2038-2042.
14. O’Brien R.J., Kamboj, S., Ehlers, M.D., Rosen, KR., Fischbach, G.D. & Huganir, R.L. (1998) Activity-dependent modulation of synaptic AMPA receptor accumulation. Neuron 21(5), 1067-1078.
15. Ons, S., Marti, O. & Armario, A. (2004). Stress-induced activation of the immediate early gene Arc (activity-regulated cytoskeleton-associated protein) is restricted to telencephalic areas in the rat brain: relationship to c-fos mRNA. Journal of Neurochemistry 89(5), 1111-1118.
16. Park, S., Kramer E.E., Mercaldo, V., Rashid, A.J. Insel, N., Frankland, P.W & Josselyn, S.A. (2016). Neuronal Allocation to a Hippocampal Engram. Neuropsychopharmacology 41, 2987-2993.
17. Rao-Ruiz, P., Couey, J.J., Marcelo, I.M., Bouwkamp, C.G., Slump, D.E., Matos, M.R., Van der Loo, R.J., Martins, G.J., Van den Hout, M., Van IJken, W.F., Costa, M.R., Van den Oever, M.C. & Kushner, S.A. (2019a). Engram-specific transcriptome profiling of contextual memory
consolidation. Nat Commun 10:2232.
18. Rao-Ruiz, P., Yu, J., Kushner, S.A. & Josselyn, S.A. (2019b). Neuronal competition: microcircuit mechanisms define the sparsity of the engram. Current Opinion in Neurobiology 54, 163-170. 19. Renner, M.C., Albers, E.H., Gutierrez-Castellanos, N., Reinders, N.R., Van Huijstee, A.N., Xiong,
H., Lodder, T.R. & KEssels, H.W. (2017). Synaptic plasticity through activation of GluA3-containing AMPA-receptors. eLife 6, 1-22.
20. Rial Verde, E.M., Lee-Osbourne, J., Worley, P.F., Malinow, R. & Cline, H.T. (2006). Increased Expression of the Immidiate-Early Gene Arc/Arg3.1 Reduces AMPA Receptor-Mediated Synaptic Transmission. Neuron 52, 461-474.
21. Schwenk, J., Baehrens, D., Haupt, A., Bildl, W., Boudkkazi, S., Roeper, J., Fakler, B. & Schulte, U. (2014). Regional Diversity and Developmental Dynamics of the AMPA-Receptor Proteome in the Mammalian Brain. Neuron 84, 41-54.
22. Shepherd, J.D., Rumbaugh, G., Wu, J., Chowdhury, S., Plath, N., Kuhl, D., Huganir, R.L. & Worley, P.F. (2006). Arc/Arg3.1 Mediates Homeostatic Synaptic Scaling of AMPA Receptors. Neuron 52, 475-484.
23. Tonegawa, S., Liu, X., Ramirez, S. & Redondo, R. (2015). Memory Engram Cells Have Come of Age. Neuron 87(5), 918-931.
24. Turrigiano, G.G. (2008). The Self-Tuning Neuron: Synaptic Scaling of Excitatory Synapses. Cell
135(3), 422-435.
25. Turrigiano, G.G., Leslie, K.R., Desai, N.S., Rutherford, L.C. & Nelson, S.B. (1998). Activity-dependent scaling of quantal amplitude in neocortical neurons. Nature 391, 892-896. 26. Wenthold, R.J., Petralia, R.S., Blahos, J.I.I. & Niedzielski, A.S. (1996). Evidence for multiple
AMPA receptor complexes in hippocampal CA1/CA2 neurons. J Neurosci. 16(6), 1982-1989. 27. Whitlock, J.R., Heynen, A.J., Shuler, M.G. & Bear, M.F. (2006). Learning induces long-term
potentiation in the hippocampus. Science 313, 1093–1097.
28. Yassa, M.A. & Stark, C.E. (2011). Pattern separation in the hippocampus. Trends in Neuroscience 34(10), 515-525.
29. Zamanillo, D., Sprengel, R., Hvalby, O., Jensen, V., Burnashev, N., Rozov, A., Kaiser, K.M., Koster, H.J., Borchardt, T., Worley, P., et al. (1999). Importance of AMPA receptors for hippocampal synaptic plasticity but not for spatial learning. Science 284, 1805–1811. 30. Zhou, Z., Liu, A., Xia, S., Leung, C., Qi, J., Meng, Y., Xie, W., Park, P., Collingridge, G.L. & Jia, Z.
(2018). The C-terminal tails of endogenous GluA1 and GluA2 differentially contribute to hippocampal synaptic plasticity and learning. Nat Neurosci 21, 50–62.
Supplementary figures
Supplementary Fig. 1) mEPSC amplitude and frequency of WT and GluA1 KO mice. a) mEPSC amplitude of home-caged wildtype (HC-WT:
n=8; µ±SEM=8.22±0.329), fear-conditioned wildtype (FC-WT: n=8; µ±SEM=9.82±0.630), home-caged GluA1 KO (HC-KO: n=9;
µ±SEM=6.62±0.413) and fear-conditioned GluA1 KO (FC-KO: n=7; µ±SEM=6.71±0.245). b) mEPSC frequency of home-caged wildtype (HC-WT: n=9; µ±SEM=0.590±0.147), fear-conditioned wildtype (FC-(HC-WT: n=8; µ±SEM=0.945±0.120), home-caged GluA1 KO (HC-KO: n=7; µ±SEM=0.060±0.011) and fear-conditioned GluA1 KO (FC-KO: n=9; µ±SEM=0.101±0.018). Data provided by Dr. Hui Xiong.
Supplementary Fig. 2) Freezing levels directly and 2 hours after fear-conditioning of WT and GluA1 KO mice. a) Freezing levels directly after
fear-conditioning in WT (n=5) and GluA1 KO (n=5) mice. GluA1 KO mice show low freezing levels (µ±SEM=28.2±2.09) compared to their WT littermates (µ±SEM=0.413±0.110). b) Freezing levels 2 hours after fear-conditioning in WT (n=5) and GluA1 KO (n=5) mice. Decreased freezing levels in GluA1 KO mice (µ±SEM=19.2±9.28) compared to their WT littermates (µ±SEM=69.8±11.7). Data provided by Dr. Hui Xiong.
Arc Primers Sequence
Arc Forward GCG-ACG-TAA-ACG-GCC-ACA-AGT-TCA-GCG-TGT
Arc Reverse ACC-TCC-AGC-AGG-ACC-ATG-TGA-TCG-CGC-TTC
Supplementary Fig. 3) Genetic sequence of Arc primers used for PCR to establish Arc mice genotypes. Primers supplied by Eurogenetics.
GluA1 Primers Sequence
KT330 TTC-CTG-GTC-AGC-CGT-TTC-AGT-CCT-TA
KT331 TCC-TCC-ATC-TCT-GTG-TCC-CAA-GTC-CT
KT488 TTG-ATA-TCG-AAT-TCC-TGC-AGC-CCA-TTG
Supplementary Fig. 4) Genetic sequence of GluA1 primers used for PCR to establish GluA1 mice genotype. Primers supplied by
Genotype Condition Housing n=
Wildtype Home-caged Group-housed 4
Wildtype Fear-conditioned Group-housed 5
GluA1 knockout Home-caged Group-housed 5
GluA1 knockout Fear-conditioned Group-housed 5
Wildtype Home-caged Solitary-housed 2
GluA1 knockout Home-caged Solitary-housed 2
Supplementary Fig. 5) Group size of all conditions of all experiments mentioned within the main text.
Supplementary Fig. 6) dVenus+ cells in DG and CA1 of WT and GluA3 KO mice. a) dVenus+ density per 1.0 mm2 DG of home-caged WT
(HC-WT: n=1; µ±SEM=135.3±NA), home-caged GluA3 KO (HC-KO: n=2, µ±SEM=178.73±0.385) and fear-conditioned GluA3 KO (FC-KO: n=3; µ±SEM=237.68±7.05). b) dVenus+/DAPI+ percentage in CA1 of home-caged WT WT: n=1; µ±SEM=6.59±NA), home-caged GluA3 KO
