Associative memory cells of encoding fear signals and anxiety are recruited by neuroligin-3-mediated synapse formation

Chen, Bingchen; Zhang, Yun; Xiao, Huajuan; Wang, Lei; Li, Jiayi; Xu, Yang; Wang, Jin-Hui

doi:10.1038/s42003-024-07170-w

Download PDF

Article
Open access
Published: 08 November 2024

Associative memory cells of encoding fear signals and anxiety are recruited by neuroligin-3-mediated synapse formation

Bingchen Chen¹,
Yun Zhang¹,
Huajuan Xiao¹,
Lei Wang¹,
Jiayi Li¹,
Yang Xu¹ &
…
Jin-Hui Wang ORCID: orcid.org/0000-0001-8340-717X¹

Communications Biology volume 7, Article number: 1464 (2024) Cite this article

2514 Accesses
3 Citations
Metrics details

Subjects

Abstract

Acute severe stress may induce fear memory and anxiety. Their mechanisms are expectedly revealed to explore therapeutic strategies. We have investigated the recruitment of associative memory cells that encode stress signals to cause fear memory and anxiety by multidisciplinary approaches. In addition to fear memory and anxiety, the social stress by the resident/intruder paradigm leads to synapse interconnections between somatosensory S1-Tr and auditory cortical neurons in intruder mice. These S1-Tr cortical neurons become to receive convergent synapse innervations newly from the auditory cortex and innately from the thalamus as well as encode the stress signals including battle sound and somatic pain, i.e., associative memory neurons. Neuroligin-3 mRNA knockdown in the S1-Tr cortex precludes the recruitment of associative memory neurons and the onset of fear memory and anxiety. The stress-induced recruitment of associative memory cells in sensory cortices for stress-relevant fear memory and anxiety is based on neuroligin-3-mediated new synapse formation.

Auditory cortical neurons are recruited to encode fear signals and anxiety by neuroligin-3-mediated synapse formation

Article Open access 11 April 2025

Neuronal types in the mouse amygdala and their transcriptional response to fear conditioning

Article Open access 26 October 2023

Synaptic potentiation of engram cells is necessary and sufficient for context fear memory

Article Open access 04 June 2025

Introduction

Physical and/or psychological stresses often induce fear memory and anxiety in individuals, e.g., posttraumatic stress disorder and phobia to specific events or objects^{1,2,3,4,5,6,7,8,9,10,11}. The persistence of these affective disorders also leads to the suppressions of the immune system and the cardiovascular system¹². It is critical to reveal cellular and molecular mechanisms underlying the correlations among stresses, fear memory and affective disorders. In general, the acute severe stress induces fear memory and anxiety, and the chronic mild stress initiates memories to negative events and defeats to induce the depression^{4,13,14,15,16,17}. The erasing of stress-induced memories to negative outcomes is the essential step to relieve these pathological moods and to treat secondary diseases¹⁸. Efforts to achieve this goal remain unsuccessful^{18,19,20,21,22}. Less success may result from the possibility that many types of stresses, including acute severe stress versus chronic mild stress, psychological stress versus physical stress and social stress versus natural hazard stress, may induce different cellular and molecular changes in the brain correlated to the suffering of anxiety, depression or their mixtures^{5,18,23,24,25,26}. The comprehensive molecular profiles and neuronal circuits relevant to different stresses and consequences remain elucidated to understand the pathogenesis of such affective disorders as well as to develop their therapeutic strategies.

Many brain regions have been presumably involved in fear memory, anxiety and depression^{1,11,27,28,29}, such as the amygdala, the prefrontal cortex and the nucleus accumbens^{30,31,32,33,34,35,36,37,38,39,40,41,42}. Neuronal circuits in these structures have been thought of as relevance to the balance between reward memory and fear memory^{43,44,45,46,47,48,49}. The abnormality of these structures in the brain often leads to memories to negative events and affective disorders, e.g., anxious and depressive moods^{10,17,39,50,51,52,53,54,55,56,57,58,59,60,61}. Whether other brain areas that primarily receive the stressful signals from sensory systems, such as auditory, visual and somatic sensory cortices, play the role in these affective disorders remain to be examined^11,62. Whether the neurons in such cortical areas are recruited to encode multiple signals associated in the social stress and to cause fear memory and anxiety, similar to the working principle of associative memory cells in other cortices¹¹, remains to be identified.

Many molecules relevant to the fear memory and anxiety induced by the social stress have been identified at levels of mRNA and miRNA in the brain^{4,6,63,64,65,66,67,68,69,70,71,72,73,74}. Studies have less been done to address the causal relationship between these molecules and psychological disorders. How these molecules initiate the pathological changes at the synapses, neurons, glia cells and neural circuits in different brain areas remains to be addressed. As memories to various signals are presumably caused by synapse formation¹⁸ and synapse plasticity^{75,76,77,78,79,80}, the proteins for synapse linkages, e.g., neuroligin-3 and neurexin^{81,82,83,84,85,86,87}, may be required for fear memory and anxiety evoked by the social stresses. To all of these points about the neural circuits and molecular profiles of mood disorders induced by the specific type of stresses, we intend to examine how associative memory neurons are recruited in the sensory cortices to encode the stress signals as well as to induce fear memory and anxiety by neuroligin-3-mediated new synapse formation.

Our strategies to address this issue are given below. C57 mice as intruders were subjected to the physical/psychological stress by a resident/intruder paradigm, in which they experienced the attacks from a resident CD1 mouse^{4,88,89,90,91,92,93}. Mice in social stress and control groups were placed in a social interaction cage to measure fear memory and on an elevated-plus maze to assess anxious state. Intruder C57 mice that showed stress-induced fear memory and anxiety as well as control C57 mice were studied by the neural tracing to track new synapse formation and interconnections and by the recording the responses of somatosensory cortical neurons to the stress signals including the battle sound and the pain signal from somatic injury regions. Synapse interconnections of associative memory neurons among cross-modal cortices were examined by microinjecting adeno-associated viruses that carried genes of encoding fluorescent proteins in one of sensory cortices and by detecting their expression in those connected cortical areas, or the other way around. The recruitment of associative memory neurons was ensured when they morphologically received the new synapse contacts between fluorescent-labeled presynaptic axon boutons and postsynaptic spines along with innate synapse contacts over neuronal dendrites in the convergent manner^94,95,96 as well as when they demonstrated the increased spike-encoding in response to those stressful signals^94,97. The roles of neuroligin-3 in the formation of fear memory and the recruitment of associative memory neurons were studied by the short-hairpin RNA (shRNA) that was specific to silence neuroligin-3 mRNA.

Results

In this section, we present our data about the stress-induced recruitment of the associative memory neurons in the somatosensory cortex that encode the stressful signals as well as result in fear memory and anxiety by neuroligin-3-mediated synapse formation. The fear memory to CD1 resident mouse in C57 intruder mice was induced by CD1 resident attacks in a resident/intruder paradigm, in which the social stress signals included the somatic pain from body-injury areas and the sound signal from their battles. These stress signals were inputted to the cerebral cortices by the somatosensory system and the auditory system. The fear memory to this CD1 resident mouse in C57 intruder mice was examined by the social interaction test (Methods). The anxiety state in C57 intruder mice was tested by the elevated-plus maze. The functional presence of associative memory neurons recruited in S1-Tr cortices of C57 intruder mice was identified by recording their responses to the battle sound and the pain stimuli to the somatic injury areas. The morphological identification of associative memory neurons in S1-Tr cortices of C57 intruder mice was done by detecting the convergent synapse innervations on S1-Tr cortical neurons from the auditory cortex and the thalamus as well as the interconnections between such two cortical areas. The influences of neuroligin-3 in the formation of new synapses and the recruitment of associative memory cells were examined by the knockdown of neuroligin-3 mRNA with shRNA carried by adeno-associated virus (AAV).

The social stress by resident/intruder paradigm induces fear memory and anxiety

C57BL/6JThy1-YFP mice were initially screened by the social interaction test in an open field including an interaction zone and a small box as well as by an elevated-plus maze to have their self-control data. Subsequently, the associative learning by resident/intruder paradigm^{4,88,89,90,91,92,93} was done for three weeks in a group of C57 mice to induce their fear memory and anxiety, compared with a group of control C57 mice (Fig. 1A). The formation of fear memory was accepted when the C57 intruder mice showed the avoidance specific to the CD1 resident mouse placed in the small box within the interaction zone, which had attacked C57 intruder mice in resident/intruder paradigm^4,6,68,73,74. The anxiety state in these C57 intruder mice was featured by less access to the open arms on an elevated-plus maze^26,98,99,100.

**Fig. 1: The social stress leads to fear memory and anxiety-like behavior in mice.**

The motions of C57 intruder mice in the social interaction test with the presence of the CD1 resident mouse are presented by heat-maps in Fig. 1B. Fear memory was presumably formed in C57 intruder mice when they avoided the box containing this CD1 mouse and less accessed an interaction zone (white dash-line). The stay time in the interaction zone before (self-control) and after the social stress was measured. The significant reduction in the ratio of the stay time in the interaction zone after and before the social stress indicated the avoidance of C57 intruder mice to the CD1 resident mouse, i.e., the formation of fear memory to this CD1 resident mouse in C57 intruder mice. In the comparison of their behaviors before and after the social stress, the C57 intruder mice appeared to stay away from the interaction zone, or the avoidance of C57 intruder mice to this CD1 resident mouse. C57 control mice appeared not to avoid the interaction zone. The values about the ratio of the stay time in the interaction zone after the social stress to that before the social stress are 1.25 ± 0.17 in C57 control mice (n = 15) and 0.31 ± 0.07 in C57 intruder mice (n = 15, p < 0.0001, Fig. 1C). This result indicates the formation of fear memory in C57 intruder mice to the CD1 resident mouse. It is noteworthy that these C57 intruder mice did not show the avoidance to the empty box in the interaction zone and the ambient sound in their living environment (Supplementary Fig. S1) as well as to other C57 mice^4,73, such that fear memory in the intruder mice is specific to the CD1 resident mouse.

The behavioral tasks of C57 mice on the elevated-plus maze are presented in the heat-maps of Fig. 1D. The anxiety state was measured based on their less access to the open arms of the elevated-plus maze (white line), compared to accesses to close arms (yellow line). The stay time in open arms before and after the social stress was measured. The significant reduction in the ratio of the stay time in the open arms after the social stress to the stay time before this social stress indicates their avoidance to open arms, or the formation of anxiety in C57 intruder mice. Compared with their behaviors in open arms before and after this resident/intruder paradigm, C57 mice in the intruder group appeared to stay in closed arms and away from open arms, or the avoidance to open arms to be anxiety state. C57 mice in the control group appeared not to avoid open arms. The values about the ratio of the stay time in open arms after to before this social stress paradigm are 1.40 ± 0.20 in C57 control mice (n = 15) and 0.48 ± 0.08 in C57 intruder mice (n = 15, p < 0.001, Fig. 1E). Thus, the social stress induces anxiety in intruder mice, which supports experimental data in the open field (Fig. 1B,C).

The resident/intruder paradigm includes a few of fear signals, such as the painful signal from body-injury area and the sound signal from their battle. The association of such fear signals in the resident/intruder paradigm appears associative learning. Fear memory and fear-relevant anxiety induced by these fear signals fell into the category of associative memory^{4,6,11,68,73,74}. We aimed to study whether associative memory neurons were recruited to encode these fear signals in intruder mice with fear memory and anxiety by morphological and functional approaches.

It is noteworthy that associative memory neurons are thought of as the basic unit of engram circuits¹¹. Engram cells have been identified by detecting activity-dependent genes, e.g., c-Fos^101,102,103, such that we initially tested the recruitment of engram cells relevant to fear memory by immunofluorescence. As shown in Fig. 2A, the number of c-Fos-labeled cells appear to be higher in the S1-Tr cortex, hippocampus and amygdala from C57 intruder mice than in these regions from C57 control mice. The absolute number of engram cells per mm2 are 682.33 ± 44.16 in the S1-Tr cortex, 175.22 ± 20.4 in hippocampal CA3 area, 188.60 ± 20 in hippocampal CA1 area and 109.4 ± 11.44 in the amygdala of C57 intruder mice (n = 15–40 slices from three mice), and the absolute number of engram cells per mm2 are 293.93 ± 30.58 in the S1-Tr cortex, 83.44 ± 13.85 in hippocampal CA3 area, 100.13 ± 17.16 in hippocampal CA1 area and 80.10 ± 10.00 in the amygdala of C57 control mice (n = 15–40 slices from three mice, four asterisks, p < 0.0001; three asterisks, p < 0.001 and one asterisk, p < 0.05 in Fig. 2B). The percentages of engram cells in total cells are 27.99 ± 1.37 in the S1-Tr cortex, 7.96 ± 0.74 in hippocampal CA3 area, 8.47 ± 0.88 in hippocampal CA1 area and 7.42 ± 0.76 in the amygdala from C57 intruder mice (n = 15–40 slices from three mice), whereas the percentages of engram cells in total cells are 12.17 ± 1.01 in the S1-Tr cortex, 3.84 ± 0.63 in hippocampal CA3 area, 3.93 ± 0.61 in hippocampal CA1 area and 4.89 ± 0.66 in the amygdala from C57 control mice (n = 15–40 slices from three mice, four asterisks, p < 0.0001; three asterisks, p < 0.001 and an asterisk, p < 0.05 in Fig. 2C). This result indicates that associative memory cells are recruited in the S1-Tr cortex, the hippocampus and amygdala in the fear memory induced by the social stress. As c-Fos-labeled engram cells are dominantly seen in the S1-Tr cortex, we focused on studying the recruitment of associative memory cells in the S1-Tr cortex by morphological and functional approaches.

**Fig. 2: The resident/intruder paradigm enhances c-Fos expression in the S1-Tr cortex of mice.**

The S1-Tr cortex and the auditory cortex interconnect after the social stress

When C57 intruder mice were attacked by CD1 resident mouse, the stress signals including the battle sound and the painful signal from body-injury areas were inputted to the S1-Tr cortex and the auditory cortex, respectively^4,73. Based on the rule of coactivity together and interconnection together¹¹, the pain signal activated S1-Tr cortical neurons and the battle sound activated auditory cortical neurons in C57 intruder mice. Their co-activities may induce the synapse interconnections between the S1-Tr cortex and the auditory cortex as well as make the S1-Tr cortical neurons receive new synapse innervations from the auditory cortex alongside innate synapses from the thalamus. Synapse interconnections were morphologically examined by the injections of AAVs that carried the genes of encoding the fluorescent proteins into local neurons and by the detection of fluorescent proteins in their axon target areas with the anterograde neural tracing. Synapse contacts were the close attachments between fluorescent-labeled presynaptic axon boutons and YFP-labeled postsynaptic dendritic spines.

A low magnification image in Fig. 3A shows the coronal sections of cerebral brains under the confocal microscopy with the injection of AAV-CMV-GFP into the S1-Tr cortex of C57 intruder mice. The number of GFP-labeled axon boutons in the auditory cortex from an intruder mouse (right panel in Fig. 3B) looks higher than that from a control mouse (left). In statistical analyzes of Fig. 3C, the densities of axon boutons in the auditory cortices are 2.88 ± 0.33 × 10⁴ per mm³ in controls (n = 7 cubes from 3 mice) and 2.25 ± 0.75 × 10⁵ per mm³ in intruder group (n = 7 cubes from 3 mice; p < 0.0001). This result indicates that the connection from the S1-Tr cortex to the auditory cortex emerges and strengthens in stress-induced fear memory and anxiety. Moreover, a low magnification image in Fig. 3D shows the coronal sections of cerebral brains with the injection of AAV-CMV-tdTomato in the auditory cortex of C57 intruder mouse. The projection of RFP-labeled axonal boutons in the S1-Tr cortex from an intruder mouse (right panel in Fig. 3E) appears higher than that from a control mouse (left panel). In statistical analysis of Fig. 3F, the densities of axonal boutons in S1-Tr cortices are 1.86 ± 0.19 × 10⁵ per mm3 in controls (n = 7 cubes from 3 mice) and 5.12 ± 0.55 × 10⁵ per mm³ in intruders (n = 7 cubes from 3 mice; p < 0.001). This result indicates that the connection from the auditory cortex to the S1-Tr cortex is emerged and strengthened in stress-induced fear memory and anxiety. This indication has also been tested and confirmed by using the retrograde neural tracing (Supplementary Fig. S2).

**Fig. 3: The social stress induces interconnections between the S1-Tr cortex and the auditory cortex.**

We have also examined whether the interconnections between S1-Tr and auditory cortices are functional or not. Experiments were conducted in the in vivo electrophysiology by stimulating the auditory cortex and by recording the responsiveness of S1-Tr cortical neurons (Fig. 4A). By using the same intensity of electrical stimulations, S1Tr cortical neurons in C57 intruder mice with fear memory appear to respond to electrical stimulations, in comparison with C57 control mice (Fig. 4B). The amplitudes of field excitatory postsynaptic potentials in response to 100 μA are 0.024 ± 0.003 mV in intruder mice (red symbols, n = 6) and 0.003 ± 0.003 mV in control mice (blue symbols, n = 6, p < 0.001, Fig. 4C). The amplitudes of excitatory postsynaptic potentials in response to 500 μA are 0.031 ± 0.004 mV in intruder mice (red symbols, n = 6) and 0.017 ± 0.004 mV in controls (blue symbols, n = 6, p < 0.001, Fig. 4C). Thus, synapse interconnections between the S1-Tr cortex and the auditory cortex are functional.

**Fig. 4: The functional connection between the S1-T and auditory cortices is established after the resident/intruder paradigm.**

S1-Tr cortical neurons receive convergent synapse innervations after the social stress

New interconnections between the S1-Tr cortex and the auditory cortex in fear memory and anxiety mice may cause the following cellular alternations. The S1-Tr cortical neurons receive new synapse innervations from the auditory cortex and innate synapses from the thalamus. The S1-Tr neurons encode those fear signals associatively acquired in the social stress including the battle sound from the auditory cortex and the pain signal from body-injury areas, or other way around. The S1-Tr cortical neurons are recruited to encode associative fear signals for fear memory and anxiety, whose features are similar to associative memory cells^11,104.

To test this implication, AAV-CMV-tdTomato and AAV-CMV-BFP were microinjected into the auditory cortex (left panel in Fig. 5A) and the thalamus (right panel) in C57 mice, respectively. After the social stress, red and blue axon boutons were tracked on the spines protruded from the apical dendrites of S1-Tr cortical neurons labeled by the YFP to examine convergent innervations. In comparison with S1-Tr dendrites in control mouse, synapse contacts from the auditory cortex (red boutons) and the thalamus (blue boutons) can be detected on the spines of an individual dendrite of S1-Tr cortical neuron (white arrows in Fig. 5B). The spines per 100 μm dendrite on S1-Tr cortical neurons are 64.50 ± 3.85 in control group (n = 30 dendrites from 3 mice) and 86.87 ± 3.76 in intruder group (n = 34 dendrites from 4 mice; p < 0.001, Fig. 5C). The head sizes of dendritic spines (μm) on S1-Tr cortical neurons are 0.52 ± 0.01 in control group (n = 131 spines from 3 mice) and 0.66 ± 0.02 in intruder group (n = 132 spines from 3 mice; p < 0.0001, Fig. 5D). These data indicate that the stress-induced fear memory and anxiety are associated with the increases in the number of dendritic spines for receiving more axon inputs and the size of dendritic spines for more efficient synapse transmission.

**Fig. 5: The high density of dendritic spines and the emergence of associative memory cells in intruder mice with fear memory and anxiety.**

In addition, synapse contacts per 100 μm dendrite on S1-Tr cortical neurons made by axons from auditory cortical neurons (red boutons on yellow spines) are 1.46 ± 0.27 in control group (n = 30 dendrites from 3 mice) and 6.99 ± 0.41 in intruder group (n = 34 dendrites from 4 mice; p < 0.0001, Fig. 5E). Synapse contacts per 100 μm dendrite on S1-Tr cortical neurons made by axons from thalamic neurons (blue boutons on yellow spines) are 0.88 ± 0.25 in control group (n = 30 dendrites from 3 mice) and 5.69 ± 0.56 in intruder group (n = 34 dendrites from 4 mice; p < 0.0001, Fig. 5F). The stress-induced fear memory and anxiety are associated with the increase in the number of convergent synapses on somatosensory cortical neurons from the auditory cortex and the thalamus.

Interestingly, the ratios of postsynaptic spines on S1-Tr cortical neurons in intruder mice to those in control mice are 1.32 ± 0.05 per 10 μm dendrite (n = 30), the ratio of red presynaptic axon boutons of auditory cortical neurons on the S1-Tr cortex in intruder mice to those in control mice is 2.70 ± 0.1 mm³ (n = 5 cubes), as well as the ratio of blue presynaptic axon boutons of thalamic neurons onto the S1-Tr cortex in intruder mice to those in control mice is 2.53 ± 0.16 mm³ (n = 5 cubes; p < 0.0001, ANOVA in Fig. 5G). Furthermore, the ratios of presynaptic boutons from the auditory cortex onto the postsynaptic spines of S1-Tr cortical neurons are 8.71 ± 0.6 in control mice (n = 30) and 18.15 ± 0.78 in intruder mice (n = 30; p < 0.001, Fig. 5H). These data indicate that the social stress induces the projection of new axons dominantly than the rise of dendritic spines, such that the presynaptic factor may constitute a major driving force for the formation of new synapse innervations. This indication is supported by our observation that a presynaptic kinetic protein, kif1a, is required for new axon projection and synapse formation induced by the social stress.

This stress-induced convergent synapse innervations on S1-Tr cortical neurons may endorse these S1-Tr neurons to encode associative stressful fear signals including the battle sound from the auditory cortex and the pain signal from somatic injury areas. The functional features of the associative memory neurons¹¹ were investigated by the electrophysiological recording in S1-Tr cortices.

Associative memory neurons recruited in S1-Tr cortex encode the battle sound and the pain

The rehearsals of stress and fear signals acquired from the resident/intruder paradigm may induce C57 intruder mice to retrieve their fear memory and express anxiety. Associative memory neurons as the basic units in engrams or memory traces¹¹ may encode these stressful signals, based on their reception of new synapse innervations from the auditory cortex after the social stress along with innate synapses from the thalamus.

Electrophysiological recordings in vivo were conducted in the S1-Tr cortices of C57 mice in the meantime to give the battle sound signal and the pain stimulus to body-injury areas (Fig. 6A). Compared with S1-Tr cortical neurons in control mice that respond to the pain stimulus to body-injury areas only (blue trace in Fig. 6B), S1-Tr cortical neurons in intruder mice become to encode the battle sound alongside the pain stimulus (red trace in Fig. 6B). Percentages of S1-Tr cortical neurons in response to pain and sound signals are 9.1% in controls (n = 66 neurons from 5 mice) and 21.3% in intruders (n = 75 neurons from 6 mice; p < 0.05, X²-test in Fig. 6C,D). The fact that more S1-Tr cortical neurons in the intruder mice showing fear memory and anxiety become to encode the new battle sound alongside the innate somatosensory signal indicates the recruitment of associative memory neurons in the S1-Tr cortex during the social stress.

**Fig. 6: S1-Tr cortical neurons in stress-induced fear memory and anxiety mice encode the somatic pain stimulus and the battle sound.**

Moreover, the activity levels of those S1-Tr cortical neurons in response to the battle sound and the somatic pain were analyzed. A left panel in Fig. 6E shows that the response strengths (normalized spike frequencies) of S1-Tr cortical neurons to the pain signal are 1.6 ± 0.16 in control mice (n = 66 neurons) and 2.05 ± 0.17 in intruder mice (n = 75 neurons; p < 0.05). Right panel in Fig. 6E illustrates normalized spike frequencies at S1-Tr cortical neurons in response to the battle sound signal are 1.08 ± 0.10 in control mice (n = 66 neurons) and 1.57 ± 0.18 in intruder mice (n = 75 neurons; p < 0.05). This result indicates that associative memory neurons recruited in the S1-Tr cortex are functionally upregulated.

In noteworthy, the S1-Tr cortical neurons from control mice that encode the signals of the somatic pain and the battle sound also encode the ambient sound. The S1-Tr cortical neurons in intruder mice that encode the somatic pain and the battle sound do not respond to the ambient sound (Supplementary Fig. S3). This observation indicates that the S1-Tr cortical neurons in the mice with stress-induced fear memory specifically store the associated stress signals, i.e., the associative memory neurons, compared to a few S1-Tr cortical neurons non-specific to the stress signals in control mice.

Neuroligin-3 is required for stress-induced fear memory, anxiety and associative memory cells

Stress-induced fear memory and anxiety are associated with the recruitment of associative memory neurons by the reception of new synapse innervation in S1-Tr cortical neurons. In terms of its molecular mechanism, neuroligin-3 may be required, since it is one synapse linkage protein^{81,83,84,85,86} and is involved in memory-relevant synapse formation⁸⁷. Neuroligin-3 was knocked down by a shRNA approach in S1-Tr cortices before C57 mice experienced the social stress by the resident/intruder paradigm. If neuroligin-3 knockdown downregulates new synapse innervations, associative memory neurons and fear-relevant behaviors, the neuroligin-3 is one of molecules for these processes as well as a causal relationship between cellular changes and fear memory/anxiety is ensured.

The shRNA specific for neuroligin-3 knockdown was carried by AAV-D/J8-U6-mNlgn3-EGFP. Neuroligin-3 shRNA and its scramble control were microinjected to S1-Tr cortices of those mice in shRNA-NLGN3 group and shRNA-scramble control group, respectively. After experienced the social stress, these mice were examined in their fear memory and anxiety by behavioral tests as well as the formation of new synapse innervations and the recruitment of associative memory neurons by functional and morphological approaches.

Behavioral tasks to examine the effectiveness of neuroligin-3 knockdown on fear memory and anxiety are presented in Fig. 7. The intruder plus neuroligin-3 knockdown mice appear to spend more time in the interaction zone (Fig. 7A) and open arms (Fig. 7C), compared to the intruder plus scramble control mice. The ratios of the activity time in the interaction zone after the social stress to the activity time before the social stress are 0.92 ± 0.14 in the intruder plus neuroligin-3 knockdown mice (green symbols/bar in Fig. 7B, n = 12) and 0.47 ± 0.09 in intruder plus scramble control mice (purple symbols, n = 12; p < 0.05). The ratios of the stay time in open arms after the social stress to the stay time before the social stress are 1.15 ± 0.26 in the intruder plus neuroligin-3 knockdown mice (green symbols in Fig. 7D, n = 12) and 0.36 ± 0.14 in intruder plus scramble control mice (purple symbols, n = 12; p < 0.05). These data indicate that neuroligin-3 knockdown in S1-Tr cortices relieves stress-induced fear memory and anxiety.

**Fig. 7: Neuroligin-3 knockdown in the S1-Tr cortex prevents stress-induced fear memory and anxiety.**

The influence of neuroligin-3 knockdown on the recruitment of associative memory neurons was examined by neural tracing. AAV-CMV-tdTomato and AAV-CMV-BFP were microinjected into the auditory cortex (left panel in Fig. 8A) and the thalamus (middle panel in Fig. 8A) in C57 intruder mice, respectively, in addition to the microinjections of neuroligin-3 shRNA or shRNA- scramble control into S1-Tr cortices (right panel in Fig. 8A). Compared to the shRNA-scramble mice (left panel in Fig. 8B), intruder plus neuroligin-3 knockdown mice appears an attenuation of convergent synapse innervations on S1-Tr cortical neurons, i.e., the prevention of associative memory neuron recruitment (right panel in Fig. 8B). Spines per 100 μm dendrite on S1-Tr cortical neurons are 80.86 ± 4.93 in intruder plus scramble control group (purple symbols; n = 30 dendrites from 3 mice) and 42.91 ± 3.19 in intruder plus neuroligin-3 knockdown (green symbols; n = 34 dendrites from 4 mice; p < 0.0001, Fig. 8C). Synapse contacts per 100 μm dendrite on S1-Tr cortical neurons made by the axons of auditory cortical neurons (red boutons on yellow spines) are 2.37 ± 0.37 in intruder plus scramble group (n = 23 dendrites from 3 mice) and 1.17 ± 0.31 in intruder plus neuroligin-3 knockdown (n = 23 dendrites from 4 mice; p < 0.05, Fig. 8D). Synapse contacts per 100 μm dendrite on S1-Tr cortical neurons from the axons from thalamic neurons (blue boutons on yellow spines) are 2.17 ± 0.45 in intruder plus scramble group (n = 23 dendrites from 3 mice) and 0.60 ± 0.21 in intruder plus neuroligin-3 knockdown group (n = 23 dendrites from 4 mice; p < 0.01, Fig. 8E). These data indicate that neuroligin-3 works for the formation of convergent synapse innervations on S1-Tr cortical neurons induced by the social stress.

**Fig. 8: Neuroligin-3 knockdown in the S1-Tr cortex prevents for the formation of new synapses and the recruitment of associative memory neurons.**

The influence of neuroligin-3 knockdown on the recruitment of associative memory neurons was also studied by electrophysiology, in which neuroligin-3 shRNA or its scramble control was microinjected in S1-Tr cortices in the mice before their social stress. In vivo spikes were recorded in S1-Tr cortices of C57 mice in two groups, in which the battle sound and the somatic stimulus to body-injury regions were given. S1-Tr cortical neurons in intruder plus scramble control mice (purple trace in Fig. 9A) are able to respond to the painful stimulus and the battle sound, but S1-Tr cortical neurons in intruder plus neuroligin-3 knockdown mice mainly respond to the pain stimulus (green trace in Fig. 9A). Percentages of S1-Tr cortical neurons in response to both pain and sound signals are 16.2% in intruder plus scramble control (n = 68 neurons from 5 mice) and 4.4% in intruder plus neuroligin-3 knockdown (n = 91 neurons from 6 mice; p < 0.05, X²-test in Fig. 9B,C). The social stress may result in an increased proportion of the activated neurons in intruder mice (p < 0.01, X² = 9.186, X²-test in Fig. 9D). The knockdown of neuroligin-3 does not affect the quantity of activated neurons but disrupts the transition from tactile-encoding cells to associative memory cells. In addition, the responses of S1-Tr cortical neurons to the battle sound and somatic stimulus in intruder plus neuroligin-3 knockdown and intruder plus scramble mice are presented in Supplementary Fig. S4). It is noteworthy that the changes in the behaviors in relevance to fear memory and anxiety as well as in the spines and new synapse contacts among the mice in the control, the intruder, the intruder plus neuroligin-3 knockdown and the intruder plus scramble control groups are compared to show the influence of neuroligin-3 knockdown on the stress-induced fear memory, anxiety and their encoders (Supplementary Fig. S5). These data strengthen the indication that neuroligin-3 works for the recruitment of associative memory neurons in S1-Tr cortices from the mice experiencing a resident/intruder paradigm. Neuroligin-3-mediated synapse formation is essential to the recruitment of associative memory neurons that encode fear signals to cause fear memory and anxiety.

**Fig. 9: Neuroligin-3 knockdown in the S1-1Tr cortex prevents associative memory neurons to encode the somatic pain stimulus and the battle sound.**

Discussion

The resident/intruder paradigm induces fear memory and anxiety in mice by the associative learning of stress signals including the pain from body-injury areas and the battle sound (Fig. 1). The interconnections between the auditory cortex and the S1-Tr cortex are associated to the stress-induced fear memory and anxiety (Figs. 3–4). S1-Tr cortical neurons receive new synapse innervations from the auditory cortex alongside innate synapse innervations from the thalamus (Fig. 5). Certain S1-Tr cortical neurons become able to specifically encode the pain signal from body-injury areas and the battle sound (Fig. 6). Thus, associative memory cells are recruited to encode fear memory and anxiety based on the coactivity and interconnection between S1-Tr and auditory cortical neurons in the social stress. Importantly, the recruitment of associative memory neurons and the formation of fear memory and anxiety are downregulated by the knockdown of neuroligin-3-mediated synapse linkage (Figs. 7–9). These results strengthen an indication about the causal relationship for associative memory neurons to encode fear memory and anxiety.

The amygdala, prefrontal cortex and nucleus accumbens have been found to be involved in fear memory and anxiety^{1,27,28,30,31,32,33,34,35,36,37,38,39,40,41}. The imbalance among these units in neural circuits may strengthen the memory to negative events and cause affective disorders^{10,39,50,51,52,53,54,55,56,57,58,59,60}. These brain areas reside in the downstream of sensory cortices, and do not directly receive the sensory signals and stress signals. These brain areas are likely the second orders or the secondary places to encode fear memory and affective disorders. How the neurons in sensory cortices become to encode fear memory and anxiety remain examined¹¹. Our studies here have identified the social stress-induced recruitment of associative memory cells in the somatosensory cortex that receive new synapse innervations from the auditory cortex along with innate synapse innervations from the thalamus and that encode the stress signals including the battle sound and the somatic painful signal for the fear memory and anxiety (Figs. 3–6). In addition, the fact that neuroligin-3 knockdown downregulates the somatosensory neurons to encode the stress signals as well as the fear memory and anxiety (Figs. 7–9) endorses a conclusion that these associative memory neurons are correlated to fear memory and anxiety. This result provides a clue that the molecules relevant to new synapse formation and associative memory cell recruitment are ready to being developed for the relief of fear memory and anxiety.

The emerged and enhanced synapse interconnections between the auditory cortex and the somatosensory cortex after the stress-induced fear memory and anxiety indicate the recruitment of associative memory neurons in the auditory cortex. These auditory cortical neurons recruited for the associative memory receive the new synapse innervations from the somatosensory cortex and the innate synapse innervations from the medial geniculate body as well as encode the stress signals including the battle sound and the somatic pain signal. This proposal as the supplement of our conclusion above is granted by our preliminary results currently. With the interconnections between somatosensory and auditory cortices for the recruitment of associative memory cells, we suggest that the stress signals are strengthened by the interactions among the interconnected associative memory neurons and that any one of such stress signals is able to induce the retrieval of those associated stress signals, or the other way around, which further triggers the expression of the fear memory and anxiety by the secondary associative memory neurons in the prefrontal cortex, amygdala, nucleus accumbens and so on¹¹. This hypothesis is granted by our observations that the social stress induces the recruitment of the secondary associative memory neurons in the medial prefrontal cortex, which interconnect those primary associative memory neurons in the sensory cortices including visual, auditory and somatosensory cortices for their interactions. It is noteworthy that the stress-induced recruitment of associative memory neurons is correlated to other types of affective disorders, such as depression and schizophrenia.

The identification of molecules underlying the stress-induced fear memory and anxiety has been a hotspot to develop the therapeutic strategies for the relief of these pathological courses. Numerous molecules at levels of mRNA and microRNA in the amygdala, the nucleus accumbens and the prefrontal cortex were associated with fear memory and anxiety^{4,6,63,64,65,67,68,70,71,72,73,74}. The causal relationship between these molecules and negative mood by lowering expressions of genes and proteins in vivo has not been systematically studied. Whether these molecules in sensory cortices play the role in fear memory and anxiety remains to be studied. As memories to various signals presumably result from the synapse formation¹⁸, the proteins for synapse linkages^{81,83,84,85,86} may be required for the stress-induced fear memory and anxiety. Our studies show the correlations of neuroligin-3 to the recruitment of associative memory cells in somatosensory cortices as well as the formation of fear memory and anxiety in relevance to the social stress. Based on the rule of coactivity together and connections together¹¹, the cascades triggered by the stress signals to recruit associative memory neurons for fear memory and anxiety may include major steps below, the coactivity of somatosensory and auditory neurons by their intense action potentials, the change of epigenetic processes by microRNA-324/133, the expression of neuroligin-3 and the synapse linkage for the new synapse formation in these cortical neurons. The similar chain reaction has been indicated in another form of associative learning/memory by pairing whisker tactile and olfactory signals^{11,95,104,105,106}.

Associative memory neurons as basic units of memory trace or engrams are recruited in the formation of associative memory¹¹. This supposition is based on the studies in associative learning by pairing whisker tactile, odor and tail temperature signals^11,87,94,104. Associative memory cells are featured by the recruitment from the coactivity of neurons, the synapse interconnections among the coactive neurons, the convergent synapse innervations on these coactive neurons from their new and innate inputs as well as the encoding of multiple signals inputted from these synapse innervations on the coactive neurons¹¹. In current study, a resident/intruder paradigm by the association of the battle sound and the somatic pain leads to fear memory and anxiety in the intruder mice. Somatosensory cortical neurons possess the features of associative memory neurons. Certain S1-Tr cortical neurons interconnect auditory cortical neurons, receive the synapse innervations newly from the auditory cortex and innately from the thalamus as well as encode those stressful signals from auditory and somatosensory systems (Figs. 1–6). This is another line of evidences related to the recruit of associative memory neurons by the associative learning of social stress signals. In two animal models of associative learning, we have identified associative memory neurons for the joint storage and the reciprocal retrieval of multiple signals associated in the learning. These studies strengthen the concept of associative memory cells as basic units in memory traces or engrams for the associative learning and memory¹¹. This strengthened conclusion is encouraging neuroscientists’ investigations to examine whether associative memory neurons are also recruited in other types of associative learning and memory. It is noteworthy that the resident/intruder paradigm includes a few stressful signals, e.g., the battle sound that are inputted by the auditory system, the somatic pain that are inputted by the somatosensory system and the battle sight that are inputted by the visual system, to cerebral cortices. The signals have not been separated clearly in previous studies. The definite separation of different signals and cues from various studies of learning and memory is suggested to detect associative memory neurons and their working principles to encode those associated signals and to guide those cognitive and emotional behaviors¹¹.

Another line of studies in learning and memory has been focused onto identifying memory traces or engrams based on activity-dependent genes, e.g., c-Fos, Arc and others^{101,102,103,107,108,109,110}. There may be the link between the expression of activity-dependent genes and the active strength of memory cells. Their parallel changes lead to a thought that those cells labeled by activity-dependent genes may be engram cells^{111,112,113,114,115}. However, these studies have not examined the structural and functional features of the engrams. We have identified the stress-induced recruitments of the associative memory neurons to encode the fear memory, which are featured by the reception of multiple synapse innervations and the encoding of the stress signals inputted via these pathways (Figs. 3–6). The associative memory neurons in the S1-Tr cortex are labeled by c-Fos (Fig. 2), and the sensory cortical neurons with synapse interconnections are labeled by c-Fos after the social stress. With our data to demonstrate the associative memory neurons by their fear signal encoding, synapse interconnections and molecular labeling, other studies in relevance to the memory engrams are better to demonstrate their structural and functional properties in addition to their labeling by activity-dependent genes. This encourage is especially important under the situation that these activity-dependent genes are not specific for memory engrams. For instance, the upregulation of activity-dependent genes has been seen in those hyperactive neurons with the seizure discharge^{116,117,118,119} as well as the neuron toxicity in brain ischemia^120,121,122. That is, the activity-dependent genes may be suitable to identify all of those active neurons.

The formation of new synapse interconnection among cross-modal sensory cortices and the recruitment of associative memory neurons to encode stress signals for fear memory and anxiety are functionally and morphologically identified in a mouse model of resident/intruder paradigm. Stress-induced psychological behaviors and cellular changes are based on neuroligin-3-mediated new synapse linkage. A diagram in Fig. 10 illustrates that the social stress induces S1-Tr cortical neurons to receive new synapse innervations from the auditory cortex alongside innate synapses from the thalamus as well as the interconnections between the S1-Tr cortex and auditory cortex, such that S1-Tr cortical neurons become able to encode all of those stress signals inputted from auditory and somatosensory systems. Our study reveals associative memory neurons of encoding multiple stress signals and anxiety induced by the social stress, which strengthens the concept of associative memory cells recruited in other types of associative learning¹¹. Our study also reveals that the cellular working principle for fear memory and anxiety is based on associative memory neurons to encode stress signals and psychological behaviors, which has not been indicated in previous studies.

**Fig. 10: Neuroligin-3 knockdown mitigates fear memory, anxiety, and alters recruitment of associative memory neurons in the S1-Tr Cortex of Intruder Mice.**

Methods

The use of animals

Experiments were done in accordance with guidelines and regulations by the Administration Office of Laboratory Animal in Beijing China. All experiment protocols were approved by Institutional Animal Care and Use Committee in Administration Office of Laboratory Animal at Beijing China (B10831).

C57BL/6JThy1-YFP mice (Jackson Laboratory, USA) have been used in all of our experiments, whose glutamatergic neurons in the cerebral brain were genetically labeled by yellow fluorescent protein (YFP)^123,124. Experimental mice were accommodated in the sterile barrier facility under the condition of twelve hours for day and night with sufficient food and water. The ambient temperature was 22 ± 2 °C. The relative humidity was 55 ± 5%. These conditions were included in the specific pathogen free (SPF). The timeline for our experiments was illustrated in Fig. 1A. C57 mice in postnatal days eighteen and well-developed body were randomly placed in the groups of the control and the social stress by resident/intruder paradigm. The mice were taken into the laboratory for them to be familiar with the experimenters and the training apparatus for one week. These mice were then either the control or the social stress for three weeks. Subsequently, the mice were studied in the formation of fear memory and anxiety and the recruitment of associative memory neurons by approaches of behavioral tasks, neural tracing, electrophysiology and molecular biology.

Behavioral study by the social stress

The induction of fear memory and anxiety in C57 mice was done by the social stress with a resident/intruder paradigm^{88,89,90,91,92}. The social stress by physical and psychological stresses during mouse social interactions is more realistic in lifespan, compared with the electrical stimulation for the stressful situation used in other studies^91,92,93. The choice of CD1 male mice as aggressive residents (aggressors) used in the social stress paradigm was based on a criterion that the latency of their attacking to those unfamiliar C57 mice was within two minutes when they met. Strain C57 mice as intruders were used for our experiments starting at postnatal weeks three. After the adaptation period to be familiar with experimental operators and training apparatus for a week, the qualified C57 male mice were divided to four groups including control, intruder, intruder plus neuroligin-3 knockdown that was fulfilled by the microinjection of AAV-CMV-GFP-U6-mNGln3 into the S1-Tr cortex to produce short-hairpin RNA specifically for silencing neuroligin-3 mRNA, as well as intruder plus shRNA-scramble. The control and intruder mice were selected based on their higher activity and lower anxious state. Their anxious state was tested by the elevated-plus maze. Their fear memory to CD1 attack was examined by the social interaction test in a cage including the open field, an interaction zone and a small box for holding CD1 mouse. In the adaptation period, these mice were allowed to be familiar with those cages for their living and social interactions. The CD1 resident mouse, the CD1 attacks and the battle sound were new stress signals to C57 mice in this resident/intruder paradigm, in which these signals plus the pain signal from somatic injury areas are associatively learnt and memorized by C57 intruder mice. In this resident/intruder paradigm (i.e., social stress), C57 intruder mice received physical attacks from a male CD1 resident. During this period, these C57 intruder mice experienced the stressful environment, where the stress signals inputted to multiple sensory modalities include the battle sound to the auditory cortex, the image of CD1 mouse presence to the visual cortex, the painful signal from body injuries to the somatosensory cortices (S1-Tr), and so on. These physically and psychologically stress signals were presumably inducing the fear memory of C57 intruder mice to CD1 resident mouse^4,6,73,74.

A pair of CD1 male and female mice lived in a room space (29×17.5×12.5 cm) of normal cage for their “marriage” relationship. In the adaptation period, C57 mice lived in a normal cage were placed in the social interaction cage to collect control data about the stay time in the interaction zone. These C57 mice were also placed in the neighboring room of a CD1 living room for them to be familiar with it. To each of C57 intruder mice before experiencing the social stress paradigm, self-control values about the stay time in the interaction zone with the presence of CD1 mouse in a small box (target presence) and the stay time in the interaction zone without CD1 in this box (no target) were also collected in this period. Those C57 mice, which had this ratio above 0.5 as well as the stay time in open arms of elevated-plus maze above 10% and consistent values within mean± 2 SD, were selected for our study.

In subsequent three weeks for the associative learnings by the social stress, those C57 mice in the groups of intruders, intruder plus shRNA scramble control and intruders plus neuroligin-3 knockdown experienced the physical and psychological stress. In the first two weeks, each of the C57 intruder mice was placed in the living house of CD1 resident mice in the afternoon everyday, when the male CD1 aggressive mouse was present and female CD1 mouse was taken out of this living house. The duration for C57 intruder mice to stay in the CD1-living house was based on the attack times when CD1 male resident mouse had bitten C57 intruder mice five times on the back of their bodies. In the third week, each of these C57 intruder mice was placed in the living house of CD1 resident mice once two days. In this model, these C57 intruder mice experienced physical and psychological stress by the attack from CD1 mice. The stress signals in this social stress could be dissected into the battle sound during CD1 attacks, the image of CD1 aggressive appearance and the pain stimulus from their body injury areas bitten by a CD1 mouse. The C57 intruder mice have associatively learnt these stress signals inputted from auditory, vision and somatosensory systems. The stimulations based on these signals were used to detect their behavioral responses to these associated stress signals in order to examine the formation of associative fear memory as well as the responses of S1-Tr cortical neurons to these associated signals in order to confirm the recruitment of associative memory neurons.

The test of fear memory formation

After this social stress period, C57 mice in the groups of control, intruder, intruder plus neuroligin-3 knockdown by its specific shRNA microinjection into the S1-Tr cortex and intruder plus shRNA-scramble control were studied in the formation of fear memory. The object to test their fear memory was the CD1 resident male mice that had attacked C57 mice. The test of fear memory formation was conducted in an interaction cage that included a small box of holding a CD1 resident mouse and an interaction zone in the open field (Fig. 1). The identification of fear memory formation in the C57 mice was based on the avoidance of C57 intruder mice to the box of holding a CD1 resident male mouse as well as less access toward the interaction zone, but not avoidance to the box of holding a familiar C57 mouse. The values about the ratio of the stay time in interaction zone with CD1 mouse presence in a small box to the stay time in the interaction zone without CD1 resident presence in this box were measured again for the comparison to their self-control before the social stress paradigm. The significant reductions of the stay time in interaction zones with the presence of CD1 aggressor in the small box before and after the social stress as well as of the stay time in interaction zone with the presences of CD1 aggressor versus C57 mouse in the small box indicate the formation of fear memory in C57 intruder mice.

It is noteworthy that intruder and control mice were separately housed in their own cages during the adaptation period as well as the intervals of the social stress paradigm and of the tests of fear memory and anxiety. In other words, there were no chances for these mice among inter- groups to the direct interaction for the establishment of the social behaviors and empathy. In addition, these C57 intruder mice had no loss of social interaction capability since they did not avoid the empty box (Fig. S6) as well as other C57 mice^4,73, except for CD1 resident mouse. Furthermore, these C57 intruder mice had no loss of auditory ability since they responded to the battle sound but not ambient sound (Figs. S1 and S6).

The test of anxiety state

Anxiety-like behaviors in C57 mice treated by stressful condition or control were evaluated by an elevated-plus maze (EPM), which was thought to be a validated and classic method to assess the level of anxiety in rodents^99,100. The EPM consists of two open arms that are 30 cm in the length, 5 cm in the width and 0 cm in the wall height opposite to two closed arms that are 30 cm in the length, 5 cm in the width and 15.25 cm in the wall height. The arms extended from a central platform (5 cm × 5 cm). The EPM was located 40 cm above the floor. All of the experiments were performed between 8:00 to 14:00. Mice naturally avoid the open field. On the other hand, they intend to explore new environments for the food. In this regard, the merit for the mice to avoid the open field was measured by the duration when the mice stayed in the closed arms, or the duration in the closed arms versus total experimental time, whereas the merit for mice to explore the new environment was measured by the exploration times toward open arms. Therefore, the exploration times and the duration in the closed arms were used to evaluate the level of anxiety, which were recorded by the automatic video-tracking system for five minutes. C57 mice were placed at the open field of elevated-plus maze with facing to a closed arm at the beginning of experiments. The higher level of anxiety-like behaviors was the situation that the mice spent more time in the closed arms and having lower exploration times toward the open arms^26,98.

The identification of associative memory cells

Associative memory neurons are recruited based on the rule of coactivity together and interconnections together among the neurons by a chain reaction including intensive action potentials, epigenetic events as well as gene and protein expression in relevance to their new axon projection and synapse formation. Associative memory neurons among cross-modal cortices and within intramodal cortex are featured by their synapse interconnections, such that each of them receives new synapse innervations from active neurons alongside innate synapse innervations as well as encodes those signals inputted by these axons and synapse inputs^{11,94,95,96,125}. Because the resident/intruder paradigm included various signals that were transmitted into the central nervous system of C57 intruder mice from the auditory system that encodes the battle sound generated in the attack of CD1 mouse to C57 mice, the somatosensory system that encodes the pain signal from body-injury areas bitten by a CD1 resident mouse, and the visual system that encodes CD1 resident and fighting images, these stressful signals to C57 intruder mice would cause their associative learning about auditory, somatosensory and visual signals. These stressful signals presumably for the joint storages and the reciprocal retrievals of the associative memory included the somatosensory signal (the painful signal from body-injury areas), the auditory signal (the battle sound generated during their fighting) as well as the visual signal from CD1 resident image and fight scene. In this regard, the fear memory to CD1 resident mouse in C57 intruder mice might be induced by seeing CD1 mouse that had attacked, listening the battle sound or receiving the pain stimuli in body-injury areas. New synapse interconnections might be detected through neural tracing among S1-Tr, auditory and visual cortices. Associative memory neurons, which encoded the pain signal from stimulating body-injury areas, the battle sound during their fighting and the visual image from the fighting scene, would be identified by electrophysiological approach in these cortical areas.

Neural tracing to localize associative memory neurons

The morphological identification of the interconnections between S1-Tr and auditory cortices was conducted by using AAV-carried genes of fluorescent proteins. AAV was an abbreviation of adeno-associated viruses^{94,95,96,97,126}. A few AAVs with CMV-promoter were used in our experiments, such as AAV-CMV-GFP, -BFP, and -tdTomato (OBiO Inc., Shanghai China). In the study of synapse interconnections between the S1-Tr cortex and auditory cortex, AAV-CMV-EGFP was microinjected into the S1-Tr cortex (-1.34 mm posterior to the bregma, 1.5 mm lateral to the middle line and 0.3 mm depth away from the bregma; the brain map¹²⁷) and AAV-CMV-tdTomato was microinjected into the auditory cortex (−2.7 mm posterior to the bregma, 4.0 mm lateral to the middle line and 1.2 mm depth away from the bregma) before the resident/intruder paradigm. The microinjections were done by using glass electrodes, and were controlled by the syringe held with the three-dimensional stereotaxic apparatus (RWD Life science, Shenzhen, China). The injected AAVs were about 0.2 μl in volume and 30 min for injection period. AAVs-CMV-GFP were uptaken and then expressed in S1-Tr cortical neurons, where the green fluorescent protein (GFP) was produced. The GFP was transported toward entire axons at the target areas in the anterograde manner, such that axon boutons and terminals were labeled by the GFP. AAVs-CMV-tdTomato were uptaken and then expressed in auditory cortical neurons, where the red fluorescent protein (RFP) was produced. The RFP was transported to entire axons at the target areas in the anterograde manner, so that axon boutons and terminals were labeled by the RFP^{94,95,96,97,126,128}.

In the study of convergent synapse innervations on S1-Tr cortical neurons from the auditory cortex and the thalamus, or associative memory neuron in the S1-Tr cortex, AAV-CMV-tdTomato (0.2 μl) was microinjected into the auditory cortex, and AAV-CMV-BFP (0.2 μl) was microinjected into the thalamus. AAVs were uptaken and then expressed at those neurons in the injected areas, where fluorescent proteins were produced. These fluorescent proteins were transported toward the entire axons at the target areas in the anterograde manner, so that certain axon boutons and terminals labeled by these fluorescent proteins were detected in the target areas and even on the dendritic spines of S1-Tr cortical neurons to be synapse contracts, called as a morphological identification of associative memory neurons^{94,95,96,97,126}. The microinjections of AAVs two days before the resident/intruder paradigm allowed the transportation of the expressed-fluorescent proteins to axonal terminals along with the projection of learning-induced axonal growth.

After the injections of AAVs in intruder and control mice were about three weeks, the mice were anesthetized by the intraperitoneal injections of urethane (1.5 g/kg). 50 ml of 0.9% saline and then 50 ml of 4% paraformaldehyde were perfused into their left ventricle until their bodies were rigid. The brains were quickly isolated and post-fixed in 4% paraformaldehyde in additional 24 h. Cerebral brains were sliced by a vibratome in a series of coronal sections with 100 μm in thickness. In order to clearly show three-dimensional images about new synapses in the S1-Tr cortex, these brain slices were placed in Sca/eA2 solution for 10 min to make them transparent^95,129. These brain slices were rinsed by the phosphate buffer solution (PBS) for 3 times, air-dried and cover-slipped. The images of cortical neurons, dendrites, dendritic spines, axonal boutons and synapse contacts were taken and collected under a 60X lens for the high magnification on the confocal microscope (Nikon A1R plus). The anatomic images of cerebral brains were taken by a 4× lens for low magnification in this confocal microscope. In the C57BL/6JThy1-YFP mice, the postsynaptic neuron dendrites and spines were labeled by the YFP. Presynaptic axon boutons were labeled by the GFP, BFP or RFP produced from AAV-CMV-FPs being microinjected. Those contacts between yellow dendritic spines and blue or red presynaptic axonal boutons with <0.1 μm space cleft were presumably the chemical synapses^95,97. The wavelength of excitation laser-beam 488 nm was used to activate the GFP and YFP. The wavelength of excitation laser-beam 561 nm was used to activate the RFP. The wavelength of excitation laser-beam 405 nm was used to activate the BFP. The wavelengths of the emission spectra of the BFP, GFP, YFP and RFP were 412–482 nm, 492–512 nm, 522–552 nm and 572–652 nm, respectively. The images of those spines, boutons and synapse contacts were quantitatively analyzed by the ImageJ and Imaris⁹⁵. The associative memory neurons were accepted by detecting two sources of buttons convergently to those spines on each of dendrites from YFP-labeled S1-Tr cortical neurons^11,94.

Electrophysiological neuronal recordings to identify associative memory neurons

Before the electrophysiological recording of S1-Tr cortical neurons, the mice in intruder group with fear memory and anxiety or control group were anesthetized by intraperitoneal injections of urethane (1.5 g/kg) for surgical operations after the social stress paradigm was done. The body temperature was kept at 37 °C by computer-controlled heating blanket. The craniotomy (2 mm in diameter) was done on the mouse skull above the left side of the S1-Tr cortex (−1.34 mm posterior to the bregma and 1.5 mm lateral to the midline)⁹⁷. Electrophysiological recordings at S1-Tr cortical neurons in vivo were conducted in these mice under light anesthetic condition with the withdrawal reflex by pinching, the eyelid blinking reflex by air-puffing and the muscle relax. The unitary discharge of cortical neurons in a category of local field potential (LFP) was recorded in layers II-III of the S1-Tr cortices by using glass pipettes filled with a standard solution (150 mM NaCl, 3.5 mM KCl and 5 mM HEPES). The resistance of those recording pipettes was 30 MΩ. The electrical signals of S1-Tr cortical neurons in their spontaneous spikes and evoked-spikes by the fighting-noisy sound or the stimulus to injury areas were recorded and acquired by AxoClamp-2B amplifier and Digidata 1322 A and were analyzed by pClamp 10 system (Axon Instrument Inc. CA, USA). Spiking signals were digitized at 20 kHz and filtered by the low-pass at 5 kHz. The 100–3000 Hz band-pass filter and the second-order Savitzky-Golay filter were used to isolate spike signals. Spike frequencies were quantitatively measured. Relative spike frequencies in response to the noise sound and somatic stimulations were the ratio that the spike frequencies in response to the stimulations were divided by spontaneous spike frequencies 20 s before the stimulations. If the ratio of those evoked-spike frequencies to spontaneous spike frequencies was 1.5 or above, S1-Tr cortical neurons were deemed as the response to the stimulations^94,95,97,126. Associative memory neurons were accepted by detecting the situation that S1-Tr cortical neurons respond to those two sources of the paired stressful signals^11,94.

Electrophysiological experiments were also fulfilled by stimulating auditory cortical neurons and recording S1-Tr cortical neurons from those mice in control group and intruder group, which was used to examine the functional connection from the auditory cortex to the S1-Tr cortex, or the other way around. LFP peaks were the most common of consistent parameters used to be quantitated. The differences between peaks and baselines in individual LFPs were measured and averaged to show stimulus-evoked LFP amplitudes^130,131. In electrophysiology recording about the activities of S1-Tr cortical neurons (−1.34 mm posterior to the bregma and 1.5 mm lateral to the midline)⁹⁷, electrical stimulations were given to auditory cortices (−2.7 mm posterior to the bregma, 4.0 mm lateral to the middle line and 0.5 mm depth away from the bregma) in those mice of control and intruder groups, The frequency of electrical stimulations was 0.1 Hz.

Immunofluorescence for detecting c-Fos expression in engram cells

In the meantime of identifying associative memory neurons by neural tracing and electrophysiological recording, the recruitment of associative memory cells as basic units in engram circuits in the resident/intruder paradigm were also examined by detecting the expression of activity-dependent genes, such as c-Fos¹¹⁵, in S1-Tr cortical neurons with the approach of western-blot to quantify this protein. The mice in control and intruder groups were anaesthetized with urethane (1.5 g/kg). They were perfused transcardially with 40 ml of the PBS and then 50 ml of a cold PBS plus 4% paraformaldehyde. Their brains were isolated and post-fixed in 4% paraformaldehyde at 4 °C for additional 2 h. Brain tissues were embedded in Optimum Cutting Temperature (O.C.T., Tissue-Tek). By using a Thermo Scientific cryostat HM525, the brain slices with free-floating 20 μm were sectioned. These sections were washed by PBS for 5 min about three times and then treated with 0.5% Tween20 in TBS at room temperature for 1 hour. Subsequently, these sections were washed with PBS for 5 min about three times, blocked with 10% normal goat serum, 1% BSA, 0.3 M glycine, and 0.1% Tween20 in PBS at room temperature for 2 h. These sections were incubated with mouse anti-c-Fos antibody (1:1000 CST) at 4 °C overnight. After washed with PBS three times (10 min for each time), the tissue sections were incubated in rabbit anti-mouse antibody at room temperature for 1.5 hour (1:1,000; Alexa Fluor 647 Conjugate, CST). The tissue sections were washed with PBS again three times (10 min for each time). The tissue sections were then stained by DAPI to label cellular nuclei for twenty minutes. Subsequently, these brain slices were mounted on glass slides with Vectashield mounting medium (Vector Laboratories). These brain slices mounted on glass slides were viewed and scanned under the confocal microscope (Nikon A1R plus, Japan)¹³².

The study of molecules for the recruitment of associative memory neurons

In the study of roles of neuroligin-3 in the formation of new synapse inputs and the recruitment of associative memory cells in the S1-Tr cortex, the approach of neuroligin-3 mRNA knockdown was used by its specific shRNA that was carried by AAVs and microinjected into this cortical area^{133,134,135,136,137}. In the meantime, the control was conducted by microinjecting neuroligin-3 shRNA-scramble. By the injection of AAV-CMV-GFP-U6-mNlgn3 (pAAV[shRNA]-GFP-U6-mNlgn3) into the S1-Tr cortex three days before the social stress paradigm, this approach was expected to lower neuroligin-3 expression of in S1-Tr cortical neurons, to prevent the formation of new synapse innervations from the auditory cortex to the S1-Tr cortex and to lower the recruitment of associative memory neurons in the S1-Tr cortex. Experiments in neuroligin-3 knockdown were jointly conducted with AAV-mediated neural tracing and electrophysiological recording to test its effectiveness on the morphology and function of associative memory neuron recruitment. After the resident/intruder paradigm, those mice from the groups of intruder plus shRNA and intruder plus shRNA-scramble were studied in their stress-induced fear memory and anxiety by behavioral tasks, the synapse convergent innervations onto S1-Tr cortical neurons by neural tracing, and the responses of the associative memory neurons to the battle sound and the somatic pain by the electrophysiological recording. The quantities of S1-Tr cortical neurons in response to the two signals were analyzed in these groups. The influences of shRNA specific for neuroligin-3 on new synapse innervations and associative memory neuron recruits would be confirmed when the number of new synapse contacts and associative memory neurons in the group of intruder plus neuroligin-3 knockdown was significantly lowered in comparison with the scramble control group.

It is noteworthy that shRNA designed to knockdown neuroligin-3 has been proved to be their effectiveness on the expression levels of neuroligin-3 mRNA by RT-PCR and neuroligin-3 protein by western-blot (Fig. S7). In the meantime, we have made sure that neuroligin-3 and PSD-95 are co-localized, which is associated with their upregulation of both neuroligin-3 and PSD-95 in the mice with associative memory (Fig. S8).

Statistical analyzes

The students’ t test and one-way ANOVA was used for the comparisons of experiment data including behavior tasks, axon boutons, synapse contacts as well as neuronal responses to the battle sound and the pain stimulus to body-injury areas in mice. X²-test was applied for the statistical comparison of the alternations in the percentage of associative memory neurons identified by electrophysiological study. The sample size was set by no less than ten samples for behavioral tasks and no less than fifty cells from three mice in cellular morphological and functional studies.

Reporting summary

Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.

Data availability

The data supporting the findings of this study are available within the article and/or its supplementary materials (Supplementary Data 1). Requests for additional data can be fulfilled via the corresponding author on reasonable request.

Code availability

The electrophysiological data reported herein utilzied the pCLAMP version 10.5.2. Imaris version 9.0.1 and ImageJ were employed for image processing. Matlab version R2020a were used for animal behavioral data analyzes. Any requests for code or batch functions used in these analyzes can be fulfilled via the corresponding author upon reasonable request.

References

Baldi, E. & Bucherelli, C. Brain sites involved in fear memory reconsolidation and extinction of rodents. Neurosci. Biobehav Rev. 53, 160–190 (2015).
Article PubMed Google Scholar
Coutellier, L. & Usdin, T. B. Enhanced long-term fear memory and increased anxiety and depression-like behavior after exposure to an aversive event in mice lacking TIP39 signaling. Behav. Brain Res. 222, 265–269 (2011).
Article PubMed PubMed Central Google Scholar
Desmedt, A., Marighetto, A. & Piazza, P. V. Abnormal fear memory as a model for posttraumatic stress disorder. Biol. Psychiatry 78, 290–297 (2015).
Article PubMed Google Scholar
Du, K., Lu, W., Sun, Y., Feng, J. & Wang, J. H. mRNA and miRNA profiles in the nucleus accumbens are related to fear memory and anxiety induced by physical or psychological stress. J. Psychiatr. Res. 118, 44–65 (2019).
Article PubMed Google Scholar
Izquierdo, I., Furini, C. R. & Myskiw, J. C. Fear Memory. Physiol. Rev. 96, 695–750 (2016).
Article PubMed Google Scholar
Sun, J. et al. mRNA and microRNA profiles in the amygdala are relevant to susceptibility and resilience to psychological stress induced in mice. J. Mol. Neurosci. 70, 1771–1796 (2020).
Article CAS PubMed Google Scholar
Makkar, S. R., Zhang, S. Q. & Cranney, J. Behavioral and neural analysis of GABA in the acquisition, consolidation, reconsolidation, and extinction of fear memory. Neuropsychopharmacology 35, 1625–1652 (2010).
Article CAS PubMed PubMed Central Google Scholar
Orsini, C. A. & Maren, S. Neural and cellular mechanisms of fear and extinction memory formation. Neurosci. Biobehav Rev. 36, 1773–1802 (2012).
Article PubMed PubMed Central Google Scholar
Parsons, R. G. & Ressler, K. J. Implications of memory modulation for post-traumatic stress and fear disorders. Nat. Neurosci. 16, 146–153 (2013).
Article CAS PubMed PubMed Central Google Scholar
Si, Y., Song, Z., Sun, X. & Wang, J. H. microRNA and mRNA profiles in nucleus accumbens underlying depression versus resilience in response to chronic stress. Am. J. Med Genet B Neuropsychiatr. Genet 177, 563–579 (2018).
Article CAS PubMed PubMed Central Google Scholar
Wang, J. H. Associative memory cells: basic units of memory trace. the first edn, 1-275 (Springer Nature, 2019).
Thrall, G., Lane, D., Carroll, D. & Lip, G. Y. A systematic review of the effects of acute psychological stress and physical activity on haemorheology, coagulation, fibrinolysis and platelet reactivity: Implications for the pathogenesis of acute coronary syndromes. Thromb. Res. 120, 819–847 (2007).
Article CAS PubMed Google Scholar
Sun, X., Song, Z., Si, Y. & Wang, J. H. microRNA and mRNA profiles in ventral tegmental area relevant to stress-induced depression and resilience. Prog. Neuropsychopharmacol. Biol. Psychiatry 86, 150–165 (2018).
Article CAS PubMed Google Scholar
Overstreet, D. H. Modeling depression in animal models. Methods Mol. Biol. 829, 125–144 (2012).
Article CAS PubMed Google Scholar
Southwick, S. M. & Charney, D. S. The science of resilience: implications for the prevention and treatment of depression. Science 338, 79–82 (2012).
Article CAS PubMed Google Scholar
Willner, P., Towell, A., Sampson, D., Sophokleous, S. & Muscat, R. Reduction of sucrose preference by chronic unpredictable mild stress, and its restoration by a tricyclic antidepressant. Psychopharmacol. (Berl.) 93, 358–364 (1987).
Article CAS Google Scholar
Zhu, Z., Wang, G., Ma, K., Cui, S. & Wang, J. H. GABAergic neurons in nucleus accumbens are correlated to resilience and vulnerability to chronic stress for major depression. Oncotarget 8, 35933–35945 (2017).
Article PubMed PubMed Central Google Scholar
Wang, J. H. Searching basic units in memory traces: associative memory cells. F1000Res 8, 457 (2019).
Article PubMed PubMed Central Google Scholar
de Quervain, D., Schwabe, L. & Roozendaal, B. Stress, glucocorticoids and memory: implications for treating fear-related disorders. Nat. Rev. Neurosci. 18, 7–19 (2017).
Article PubMed Google Scholar
Flores, A., Fullana, M. A., Soriano-Mas, C. & Andero, R. Lost in translation: how to upgrade fear memory research. Mol. Psychiatry 23, 2122–2132 (2018).
Article PubMed Google Scholar
Maren, S. Seeking a spotless mind: extinction, deconsolidation, and erasure of fear memory. Neuron 70, 830–845 (2011).
Article CAS PubMed PubMed Central Google Scholar
Sandkuhler, J. & Lee, J. How to erase memory traces of pain and fear. Trends Neurosci. 36, 343–352 (2013).
Article PubMed PubMed Central Google Scholar
Ma, K. et al. Impaired GABA synthesis, uptake and release are associated with depression-like behaviors induced by chronic mild stress. Transl. Psychiatry 6, 1–10 (2016).
Article CAS Google Scholar
Xu, A., Cui, S. & Wang, J. Incoordination among subcellular compartments is associated to depression-like behavior induced by chronic mild stress. Int. J. Neuropsychopharmacol. 19, pyv122 (2015).
Article PubMed Central Google Scholar
Sutoo, D. & Akiyama, K. Neurochemical changes in mice following physical or psychological stress exposures. Behav. Brain Res 134, 347–354 (2002).
Article CAS PubMed Google Scholar
Zhang, F., Liu, B., Lei, Z. & Wang, J. mGluR1,5 activation improves network asynchrony and GABAergic synapse attenuation in the amygdala: implication for anxiety-like behavior in DBA/2 mice. Mol. Brain 5, 20 (2012). 1756-6606-5-20.
Article CAS PubMed PubMed Central Google Scholar
Herry, C. & Johansen, J. P. Encoding of fear learning and memory in distributed neuronal circuits. Nat. Neurosci. 17, 1644–1654 (2014).
Article CAS PubMed Google Scholar
Wang, J. H. & Cui, S. Associative memory cells: formation, function and perspective. F1000Res 6, 283 (2017).
Article PubMed PubMed Central Google Scholar
Wang, J. H. & Cui, S. Associative memory cells and their working principle in the brain. F1000Res 7, 108 (2018).
Article PubMed PubMed Central Google Scholar
Ehrlich, I. et al. Amygdala inhibitory circuits and the control of fear memory. Neuron 62, 757–771 (2009).
Article CAS PubMed Google Scholar
Fanselow, M. S. & Gale, G. D. The amygdala, fear, and memory. Ann. N. Y Acad. Sci. 985, 125–134 (2003).
Article PubMed Google Scholar
Floresco, S. B. The nucleus accumbens: an interface between cognition, emotion, and action. Annu Rev. Psychol. 66, 25–52 (2015).
Article PubMed Google Scholar
Garrett, A. & Chang, K. The role of the amygdala in bipolar disorder development. Developmental Psychopathol. 20, 1285–1296 (2008).
Article Google Scholar
Haruno, M., Kimura, M. & Frith, C. D. Activity in the nucleus accumbens and amygdala underlies individual differences in prosocial and individualistic economic choices. J. Cogn. Neurosci. 26, 1861–1870 (2014).
Article PubMed Google Scholar
Hastings, R. S., Parsey, R. V., Oquendo, M. A., Arango, V. & Mann, J. J. Volumetric analysis of the prefrontal cortex, amygdala, and hippocampus in major depression. Neuropsychopharmacology 29, 952–959 (2004). 1300371 [pii].
Article PubMed Google Scholar
Keele, N. B. The role of serotonin in impulsive and aggressive behaviors associated with epilepsy-like neuronal hyperexcitability in the amygdala. Epilepsy Behav. 7, 325–335 (2005).
Article PubMed Google Scholar
Lebow, M. A. & Chen, A. Overshadowed by the amygdala: the bed nucleus of the stria terminalis emerges as key to psychiatric disorders. Mol. Psychiatry 21, 450–463 (2016).
Article CAS PubMed PubMed Central Google Scholar
Lee, J. H., Lee, S. & Kim, J. H. Amygdala Circuits for Fear Memory: A Key Role for Dopamine Regulation. Neuroscientist 23, 542–553 (2016).
Article PubMed Google Scholar
Monk, C. S. et al. Amygdala and nucleus accumbens activation to emotional facial expressions in children and adolescents at risk for major depression. Am. J. Psychiatry 165, 90–98 (2008).
Article PubMed Google Scholar
Price, J. L. Comparative aspects of amygdala connectivity. Ann. N. Y Acad. Sci. 985, 50–58 (2003).
Article PubMed Google Scholar
Tzschentke, T. M. & Schmidt, W. J. Functional relationship among medial prefrontal cortex, nucleus accumbens, and ventral tegmental area in locomotion and reward. Crit. Rev. Neurobiol. 14, 131–142 (2000).
Article CAS PubMed Google Scholar
Vialou, V. et al. Prefrontal cortical circuit for depression- and anxiety-related behaviors mediated by cholecystokinin: role of DeltaFosB. J. Neurosci. 34, 3878–3887 (2014).
Article PubMed PubMed Central Google Scholar
Fadok, J. P., Darvas, M., Dickerson, T. M. K. & Palmiter, R. D. Long-term memory for pavlovian fear conditioning requires dopamine in the nucleus accumbens and basolateral amygdala. Plos One 5, e12751 (2010). ARTN.
Article PubMed PubMed Central Google Scholar
Hikida, T., Morita, M. & Macpherson, T. Neural mechanisms of the nucleus accumbens circuit in reward and aversive learning. Neurosci. Res. 108, 1–5 (2016).
Article PubMed Google Scholar
Ikemoto, S. Dopamine reward circuitry: two projection systems from the ventral midbrain to the nucleus accumbens-olfactory tubercle complex. Brain Res Rev. 56, 27–78 (2007).
Article CAS PubMed PubMed Central Google Scholar
Kochenborger, L. et al. Modulation of fear/anxiety responses, but not food intake, following alpha-adrenoceptor agonist microinjections in the nucleus accumbens shell of free-feeding rats. Neuropharmacology 62, 427–435 (2012).
Article CAS PubMed Google Scholar
Russo, S. J. et al. The addicted synapse: mechanisms of synaptic and structural plasticity in nucleus accumbens. Trends Neurosci. 33, 267–276 (2010).
Article CAS PubMed PubMed Central Google Scholar
Thomas, K. L., Hall, J. & Everitt, B. J. Cellular imaging with zif268 expression in the rat nucleus accumbens and frontal cortex further dissociates the neural pathways activated following the retrieval of contextual and cued fear memory. Eur. J. Neurosci. 16, 1789–1796 (2002).
Article PubMed Google Scholar
Yang, F. C. & Liang, K. C. Interactions of the dorsal hippocampus, medial prefrontal cortex and nucleus accumbens in formation of fear memory: Difference in inhibitory avoidance learning and contextual fear conditioning. Neurobiol. Learn. Mem. 112, 186–194 (2014).
Article CAS PubMed Google Scholar
Anderson, E. M. et al. Overexpression of the histone dimethyltransferase G9a in nucleus accumbens shell increases cocaine self-administration, stress-induced reinstatement, and anxiety. J. Neurosci. 38, 803–813 (2018).
Article CAS PubMed PubMed Central Google Scholar
Barrot, M. et al. Regulation of anxiety and initiation of sexual behavior by CREB in the nucleus accumbens. P Natl Acad. Sci. USA 102, 8357–8362 (2005).
Article CAS Google Scholar
Bosch-Bouju, M., Larrieu, T., Linders, L., Manzoni, O. J. & Laye, S. Endocannabinoid-mediated plasticity in nucleus accumbens controls vulnerability to anxiety after social defeat stress. Cell Rep. 16, 1237–1242 (2016).
Article CAS PubMed Google Scholar
Feng, J. et al. Tet1 in nucleus accumbens opposes depression- and anxiety-like behaviors. Neuropsychopharmacology 42, 1657–1669 (2017).
Article CAS PubMed PubMed Central Google Scholar
Francis, T. C. & Lobo, M. K. Emerging role for nucleus accumbens medium spiny neuron subtypes in depression. Biol. Psychiatry 81, 645–653 (2017).
Article CAS PubMed Google Scholar
Fu, K. Q. et al. Overexpression of transmembrane protein 168 in the mouse nucleus accumbens induces anxiety and sensorimotor gating deficit. Plos One 12, e0280117 (2017).
Article Google Scholar
Heshmati, M. et al. Mefloquine in the nucleus accumbens promotes social avoidance and anxiety-like behavior in mice. Neuropharmacology 101, 351–357 (2016).
Article CAS PubMed Google Scholar
Kim, K. S. et al. Adenylyl cyclase-5 activity in the nucleus accumbens regulates anxiety-related behavior. J. Neurochemistry 107, 105–115 (2008).
Article CAS Google Scholar
Levita, L., Hoskin, R. & Champi, S. Avoidance of harm and anxiety: A role for the nucleus accumbens. Neuroimage 62, 189–198 (2012).
Article PubMed Google Scholar
Lim, B. K., Huang, K. W., Grueter, B. A., Rothwell, P. E. & Malenka, R. C. Anhedonia requires MC4R-mediated synaptic adaptations in nucleus accumbens. Nature 487, 183–189 (2012).
Article CAS PubMed PubMed Central Google Scholar
Salamone, J. D., Correa, M., Mingote, S. M. & Weber, S. M. Beyond the reward hypothesis: alternative functions of nucleus accumbens dopamine. Curr. Opin. Pharm. 5, 34–41 (2005).
Article CAS Google Scholar
Wang, H. et al. Social defeat drives hyperexcitation of the piriform cortex to induce learning and memory impairment but not mood-related disorders in mice. Transl. Psychiatry 12, 380 (2022).
Article PubMed PubMed Central Google Scholar
Nomura, H. et al. Memory formation and retrieval of neuronal silencing in the auditory cortex. Proc. Natl Acad. Sci. USA 112, 9740–9744 (2015).
Article CAS PubMed PubMed Central Google Scholar
Bagot, R. C. et al. Circuit-wide transcriptional profiling reveals brain region-specific gene networks regulating depression susceptibility. Neuron 90, 969–983 (2016).
Article CAS PubMed PubMed Central Google Scholar
Blanchard, R. J., McKittrick, C. R. & Blanchard, D. C. Animal models of social stress: effects on behavior and brain neurochemical systems. Physiol. Behav. 73, 261–271 (2001).
Article CAS PubMed Google Scholar
Cestari, V., Rossi-Arnaud, C., Saraulli, D. & Costanzi, M. The MAP(K) of fear: from memory consolidation to memory extinction. Brain Res. Bull. 105, 8–16 (2014).
Article CAS PubMed Google Scholar
Ganzel, B. L., Morris, P. A. & Wethington, E. Allostasis and the human brain: Integrating models of stress from the social and life sciences. Psychol. Rev. 117, 134–174 (2010).
Article PubMed PubMed Central Google Scholar
Johansen, J. P., Cain, C. K., Ostroff, L. E. & LeDoux, J. E. Molecular mechanisms of fear learning and memory. Cell 147, 509–524 (2011).
Article CAS PubMed PubMed Central Google Scholar
Lu, Y., Yang, J., Sun, J., Lu, W. & Wang, J. H. mRNA and miRNA profiles in the nucleus accumbens are associated with psychological stress-induced susceptible and resilient mice. Pharm. Biochem Behav. 199, 173062 (2020).
Article CAS Google Scholar
Liang, J. et al. Conditional neuroligin-2 knockout in adult medial prefrontal cortex links chronic changes in synaptic inhibition to cognitive impairments. Mol. Psychiatry 20, 850–859 (2015).
Article CAS PubMed Google Scholar
Kyrke-Smith, M. & Williams, J. M. Bridging Synaptic and Epigenetic Maintenance Mechanisms of the Engram. Front Mol. Neurosci. 11, 369 (2018).
Article CAS PubMed PubMed Central Google Scholar
Otis, J. M., Werner, C. T. & Mueller, D. Noradrenergic regulation of fear and drug-associated memory reconsolidation. Neuropsychopharmacology 40, 793–803 (2015).
Article PubMed Google Scholar
Sandi, C. & Haller, J. Stress and the social brain: behavioural effects and neurobiological mechanisms. Nat. Rev. Neurosci. 16, 290–304 (2015).
Article CAS PubMed Google Scholar
Sun, Y., Lu, W., Du, K. & Wang, J. H. microRNA and mRNA profiles in the amygdala are relevant to fear memory induced by physical or psychological stress. J. Neurophysiol. 122, 1002–1022 (2019).
Article CAS PubMed Google Scholar
Yang, J. et al. Correction to: mRNA and microRNA profiles are associated with stress susceptibility and resilience induced by psychological stress in the prefrontal cortex. Psychopharmacol. (Berl.) 237, 3095 (2020).
Article CAS Google Scholar
Bliss, T. & Lomo, T. Long-lasting potentiation of synaptic transmission in the dentate area of the anaesthetized rabbit following stimulation of the perforant path. J. Physiol. (Lond.) 232, 331–356 (1973).
Article CAS PubMed Google Scholar
Bliss, T. V. P. & Lynch, M. A. in Long-term Potentiation: From Biophysics to Behavior 35 Neurology and Neurobiology (edsP. W. Landfield & S. A. Deadwyler) 3-72 (Alan R. Liss, Inc., 1988).
Poo, M. M. et al. What is memory? The present state of the engram. BMC Biol. 14, 40 (2016).
Article PubMed PubMed Central Google Scholar
Lisman, J., Cooper, K., Sehgal, M. & Silva, A. J. Memory formation depends on both synapse-specific modifications of synaptic strength and cell-specific increases in excitability. Nat. Neurosci. 21, 309–314 (2018).
Article CAS PubMed PubMed Central Google Scholar
Wang, J. H., Ko, G. & Kelly, P. T. Cellular and molecular bases of memory: synaptic and neuronal plasticity. J. Clin. Neurophysiol. 14, 264–293 (1997).
Article CAS PubMed Google Scholar
Zhang, M., Hung, F., Zhu, Y., Xie, Z. & Wang, J. Calcium signal-dependent plasticity of neuronal excitability developed postnatally. J. Neurobiol. 61, 277–287 (2004).
Article CAS PubMed Google Scholar
Craig, A. M. & Kang, Y. Neurexin-neuroligin signaling in synapse development. Curr. Opin. Neurobiol. 17, 43–52 (2007).
Article CAS PubMed PubMed Central Google Scholar
Hamilton, S. M. et al. Fmr1 and Nlgn3 knockout rats: novel tools for investigating autism spectrum disorders. Behav. Neurosci. 128, 103–109 (2014).
Article CAS PubMed Google Scholar
Li, H., Guo, R., Guan, Y., Li, J. & Wang, Y. Modulation of Trans-Synaptic Neurexin-Neuroligin Interaction in Pathological Pain. Cells 11, https://doi.org/10.3390/cells11121940 (2022).
Lise, M. F. & El-Husseini, A. The neuroligin and neurexin families: from structure to function at the synapse. Cell Mol. Life Sci. 63, 1833–1849 (2006).
Article CAS PubMed PubMed Central Google Scholar
Sudhof, T. C. Synaptic neurexin complexes: a molecular code for the logic of neural circuits. Cell 171, 745–769 (2017).
Article CAS PubMed PubMed Central Google Scholar
Uchigashima, M., Cheung, A. & Futai, K. Neuroligin-3: a circuit-specific synapse organizer that shapes normal function and autism spectrum disorder-associated dysfunction. Front Mol. Neurosci. 14, 749164 (2021).
Article CAS PubMed PubMed Central Google Scholar
Xu, Y., Cui, T. L., Li, J. Y., Chen, B. & Wang, J. H. Associative memory neurons of encoding multi-modal signals are recruited by neuroligin-3-mediated new synapse formation. eLife 12, https://doi.org/10.7554/eLife.87969 (2023).
Berton, O. et al. Essential role of BDNF in the mesolimbic dopamine pathway in social defeat stress. Science 311, 864–868 (2006).
Article CAS PubMed Google Scholar
Bjorkqvist, K. Social defeat as a stressor in humans. Physiol. Behav. 73, 435–442 (2001).
Article CAS PubMed Google Scholar
Hammels, C. et al. Defeat stress in rodents: From behavior to molecules. Neurosci. Biobehav Rev. 59, 111–140 (2015).
Article CAS PubMed Google Scholar
Martinez, M., Calvo-Torrent, A. & Pico-Alfonso, M. A. Social defeat and subordination as models of social stress in laboratory rodents: a review. Aggressive Behav. 24, 241–256 (1998).
Article Google Scholar
Tsankova, N. M. et al. Sustained hippocampal chromatin regulation in a mouse model of depression and antidepressant action. Nat. Neurosci. 9, 519–525 (2006).
Article CAS PubMed Google Scholar
Vasconcelos, M., Stein, D. J. & de Almeida, R. M. Social defeat protocol and relevant biomarkers, implications for stress response physiology, drug abuse, mood disorders and individual stress vulnerability: a systematic review of the last decade. Trends Psychiatry Psychother. 37, 51–66 (2015).
Article PubMed Google Scholar
Feng, J. et al. Barrel cortical neuron integrates triple associated signals for their memory through receiving epigenetic-mediated new synapse innervations. Cereb. Cortex 27, 5858–5871 (2017).
Article PubMed Google Scholar
Gao, Z. et al. Coactivations of barrel and piriform cortices induce their mutual synapse innervations and recruit associative memory cells. Brain Res. 1721, 146333 (2019).
Article CAS PubMed Google Scholar
Lei, Z. et al. Synapse innervation and associative memory cell are recruited for integrative storage of whisker and odor signals in the barrel cortex through miRNA-mediated processes. Front. Cell. Neurosci. 11, 1–11 (2017).
Article Google Scholar
Wu, R., Cui, S. & Wang, J. H. miRNA-324/-133a essential for recruiting new synapse innervations and associative memory cells in coactivated sensory cortices. Neurobiol. Learn Mem. 172, 107246 (2020).
Article CAS PubMed Google Scholar
Liu, B., Feng, J. & Wang, J.-H. Protein kinase C is essential for kainate-induced anxiety-related behavior and glutamatergic synapse upregulation in prelimbic cortex. CNS Neuroscience and Therapeutics 20, in press, https://doi.org/10.1111/cns.12313 (2014).
Pellow, S. & Chopin, P. Validation of open:closed arm entries in an elevated plus-maze as a measure of anxiety in the rats. J. Neurosci. Methods 14, 149–167 (1985).
Article CAS PubMed Google Scholar
Walf, A. A. & Frye, C. A. The use of elevated plus maze as an assay of anxiety-related behavior in rodents. Nat. Protoc. 2, 322–328 (2007).
Article CAS PubMed PubMed Central Google Scholar
Josselyn, S. A., Kohler, S. & Frankland, P. W. Finding the engram. Nat. Rev. Neurosci. 16, 521–534 (2015).
Article CAS PubMed Google Scholar
Josselyn, S. A. & Tonegawa, S. Memory engrams: Recalling the past and imagining the future. Science 367, https://doi.org/10.1126/science.aaw4325 (2020).
Kim, J. I., Cho, H. Y., Han, J. H. & Kaang, B. K. Which neurons will be the engram - activated neurons and/or more excitable neurons? Exp. Neurobiol. 25, 55–63 (2016).
Article PubMed PubMed Central Google Scholar
Wang, D. et al. Neurons in the barrel cortex turn into processing whisker and odor signals: a cellular mechanism for the storage and retrieval of associative signals. Front Cell Neurosci. 9, 320 (2015).
Article PubMed PubMed Central Google Scholar
Liu, Y. et al. Piriform cortical glutamatergic and GABAergic neurons express coordinated plasticity for whisker-induced odor recall. Oncotarget 8, 95719–95740 (2017).
Article PubMed PubMed Central Google Scholar
Yan, F. et al. Coordinated plasticity between barrel cortical glutamatergic and GABAergic neurons during associative memory. Neural Plasticity 2016, 1–20 (2016).
CAS Google Scholar
Reijmers, L. G., Perkins, B. L., Matsuo, N. & Mayford, M. Localization of a stable neural correlate of associative memory. Science 317, 1230–1233 (2007).
Article CAS PubMed Google Scholar
Tayler, K. K., Tanaka, K. Z., Reijmers, L. G. & Wiltgen, B. J. Reactivation of neural ensembles during the retrieval of recent and remote memory. Curr. Biol. 23, 99–106 (2013).
Article CAS PubMed Google Scholar
Pignatelli, M. et al. Engram Cell Excitability State Determines the Efficacy of Memory Retrieval. Neuron, (2018).
Tonegawa, S., Pignatelli, M., Roy, D. S. & Ryan, T. J. Memory engram storage and retrieval. Curr. Opin. Neurobiol. 35, 101–109 (2015).
Article CAS PubMed Google Scholar
Bissiere, S. et al. Electrical synapses control hippocampal contributions to fear learning and memory. Science 331, 87–91 (2011). 331/6013/87 [pii].
Article CAS PubMed PubMed Central Google Scholar
Grundemann, J. & Luthi, A. Ensemble coding in amygdala circuits for associative learning. Curr. Opin. Neurobiol. 35, 200–206 (2015).
Article CAS PubMed Google Scholar
Tanaka, K. Z. & McHugh, T. J. The hippocampal engram as a memory index. J. Exp. Neurosci. 12, 1179069518815942 (2018).
Article PubMed PubMed Central Google Scholar
Wang, K. H. et al. In vivo two-photon imaging reveals a role of arc in enhancing orientation specificity in visual cortex. Cell 126, 389–402 (2006).
Article CAS PubMed Google Scholar
Tonegawa, S., Liu, X., Ramirez, S. & Redondo, R. Memory engram cells have come of age. Neuron 87, 918–931 (2015).
Article CAS PubMed Google Scholar
Kiessling, M. & Gass, P. Immediate early gene expression in experimental epilepsy. Brain Pathol. 3, 381–393 (1993).
Article CAS PubMed Google Scholar
Meldrum, B. S. Concept of activity-induced cell death in epilepsy: historical and contemporary perspectives. Prog. Brain Res. 135, 3–11 (2002).
Article CAS PubMed Google Scholar
Wang, X., Song, X., Wu, L., Nadler, J. V. & Zhan, R. Z. Persistent hyperactivity of hippocampal dentate interneurons after a silent period in the rat pilocarpine model of epilepsy. Front Cell Neurosci. 10, 94 (2016).
Article PubMed PubMed Central Google Scholar
Simonato, M. et al. Differential expression of immediate early genes in the hippocampus in the kindling model of epilepsy. Mol. Brain Res 11, 115–124 (1991).
Article CAS PubMed Google Scholar
Abe, H. & Nowak, T. S. Jr Induced hippocampal neuron protection in an optimized gerbil ischemia model: insult thresholds for tolerance induction and altered gene expression defined by ischemic depolarization. J. Cereb. Blood Flow. Metab. 24, 84–97 (2004).
Article CAS PubMed Google Scholar
Bokesch, P. M., Marchand, J. E., Connelly, C. S., Wurm, W. H. & Kream, R. M. Dextromethorphan inhibits ischemia-induced c-fos expression and delayed neuronal death in hippocampal neurons. Anesthesiology 81, 470–477 (1994).
Article CAS PubMed Google Scholar
Kiessling, M. et al. Differential transcription and translation of immediate early genes in the gerbil hippocampus after transient global ischemia. J. Cereb. Blood Flow. Metab. 13, 914–924 (1993).
Article CAS PubMed Google Scholar
Feng, G. et al. Imaging neuronal subsets in transgenic mice expressing multiple spectral variants of GFP. Neuron 28, 41–51 (2000).
Article CAS PubMed Google Scholar
Zhang, G., Gao, Z., Guan, S., Zhu, Y. & Wang, J. H. Upregulation of excitatory neurons and downregulation of inhibitory neurons in barrel cortex are associated with loss of whisker inputs. Mol. Brain 6, 2 (2013).
Article CAS PubMed PubMed Central Google Scholar
Li, J. Y., Xu, Y., Wang, D. G. & Wang, J. H. The interconnection and function of associative memory neurons are upregulated for memory strengthening. Front Neural Circuits 17, 1189907 (2023).
Article PubMed PubMed Central Google Scholar
Gao, Z. et al. Associations of unilateral whisker and olfactory signals induce synapse formation and memory cell recruitment in bilateral barrel cortices: cellular mechanism for unilateral training toward bilateral memory. Front. Cell. Neurosci. 10, 1–16 (2016).
Article Google Scholar
Paxinos, G. & Watson, C. The Mouse Brain: in stereotaxic coordinates. 1-350 (Elsevier Academic Press, 2005).
Tervo, D. G. et al. A designer AAV variant permits efficient retrograde access to projection neurons. Neuron 92, 372–382 (2016).
Article CAS PubMed PubMed Central Google Scholar
Hama, H. et al. Scale: a chemical approach for fluorescence imaging and reconstruction of transparent mouse brain. Nat. Neurosci. 14, 1481–1488 (2011). nn.2928 [pii].
Article CAS PubMed Google Scholar
Wang, D. et al. Neurons in the barrel cortex turn into processing whisker and odor signals: a cellular mechanism for the storage and retrieval of associative signals. Frontier. Cell. Neurosci. 9, 1–12 (2015).
Google Scholar
Zhao, J., Wang, D. & Wang, J. H. Barrel cortical neurons and astrocytes coordinately respond to an increased whisker stimulus frequency. Mol. Brain 5, 12 (2012).
Article PubMed PubMed Central Google Scholar
Guo, L. et al. microRNA-15b contributes to depression-like behavior in mice by affecting synaptic protein levels and function in the nucleus accumbens. J. Biol. Chem. 295, 6831–6848 (2020).
Article CAS PubMed PubMed Central Google Scholar
Chang, K., Elledge, S. J. & Hannon, G. J. Lessons from Nature: microRNA-based shRNA libraries. Nat. Methods 3, 707–714 (2006).
Article CAS PubMed Google Scholar
Khatri, N., Rathi, M., Baradia, D., Trehan, S. & Misra, A. In vivo delivery aspects of miRNA, shRNA and siRNA. Crit. Rev. Ther. Drug Carr. Syst. 29, 487–527 (2012).
Article CAS Google Scholar
Pardridge, W. M. shRNA and siRNA delivery to the brain. Adv. Drug Deliv. Rev. 59, 141–152 (2007).
Article CAS PubMed PubMed Central Google Scholar
Pushparaj, P. N., Aarthi, J. J., Manikandan, J. & Kumar, S. D. siRNA, miRNA, and shRNA: in vivo applications. J. Dent. Res 87, 992–1003 (2008).
Article CAS PubMed Google Scholar
Rao, D. D., Vorhies, J. S., Senzer, N. & Nemunaitis, J. siRNA vs. shRNA: similarities and differences. Adv. Drug Deliv. Rev. 61, 746–759 (2009).
Article CAS PubMed Google Scholar

Download references

Acknowledgements

This study is funded by Natural Science Foundation of China (81971027, U2241209 and 81930033) to Jin-Hui Wang.

Author information

Authors and Affiliations

College of Life Science, University of Chinese Academy of Sciences, Beijing, China
Bingchen Chen, Yun Zhang, Huajuan Xiao, Lei Wang, Jiayi Li, Yang Xu & Jin-Hui Wang

Authors

Bingchen Chen
View author publications
Search author on:PubMed Google Scholar
Yun Zhang
View author publications
Search author on:PubMed Google Scholar
Huajuan Xiao
View author publications
Search author on:PubMed Google Scholar
Lei Wang
View author publications
Search author on:PubMed Google Scholar
Jiayi Li
View author publications
Search author on:PubMed Google Scholar
Yang Xu
View author publications
Search author on:PubMed Google Scholar
Jin-Hui Wang
View author publications
Search author on:PubMed Google Scholar

Contributions

Bingchen Chen, Yun Zhang, Huajuan Xiao, Lei Wang, Jiayi Li, Yang Xu contributed to experiments and data analyzes. Jin-Hui Wang contributed to the concept, project design and paper writing.

Corresponding author

Correspondence to Jin-Hui Wang.

Ethics declarations

Competing interests

All authors declare no competing interest. All authors have read and approved the final version of the manuscript.

Ethics approval and consent to participate

Experiments were done in accordance with guidelines and regulations by the Administration Office of Laboratory Animal in Beijing China. All experiment protocols were approved by Institutional Animal Care and Use Committee in Administration Office of Laboratory Animal at Beijing China (B10831).

Consent for publication

For the publication of this manuscript, the consent for the use of any 803 individual person’s data, including individual details, images, or videos, has been obtained where 804 applicable. All necessary consent for publication has been secured.

Peer review

Peer review information

Communication Biology thanks Gissella Vetere and the other, anonymous, reviewer for their contribution to the peer review of this work. Primary Handling Editor: Benjamin Bessieres. A peer review file is available.

Additional information

Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Consent for publication

For the publication of this manuscript, the consent for the use of any individual person’s data, including individual details, images, or videos, has been obtained where applicable. All necessary consent for publication has been secured.

Supplementary information

Transparent Peer Review file

Supplementary Information

Description of Additional Supplementary File

Supplementary Data 1

Reporting Summary

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by-nc-nd/4.0/.

Reprints and permissions

About this article

Cite this article

Chen, B., Zhang, Y., Xiao, H. et al. Associative memory cells of encoding fear signals and anxiety are recruited by neuroligin-3-mediated synapse formation. Commun Biol 7, 1464 (2024). https://doi.org/10.1038/s42003-024-07170-w

Download citation

Received: 14 April 2024
Accepted: 29 October 2024
Published: 08 November 2024
DOI: https://doi.org/10.1038/s42003-024-07170-w

This article is cited by

Auditory cortical neurons are recruited to encode fear signals and anxiety by neuroligin-3-mediated synapse formation
- Yun Zhang
- Bingchen Chen
- Jin-Hui Wang
Translational Psychiatry (2025)

Subjects

Abstract

Similar content being viewed by others

Introduction

Results

The social stress by resident/intruder paradigm induces fear memory and anxiety

The S1-Tr cortex and the auditory cortex interconnect after the social stress

S1-Tr cortical neurons receive convergent synapse innervations after the social stress

Associative memory neurons recruited in S1-Tr cortex encode the battle sound and the pain

Neuroligin-3 is required for stress-induced fear memory, anxiety and associative memory cells

Discussion

Methods

The use of animals

Behavioral study by the social stress

The test of fear memory formation

The test of anxiety state

The identification of associative memory cells

Neural tracing to localize associative memory neurons

Electrophysiological neuronal recordings to identify associative memory neurons

Immunofluorescence for detecting c-Fos expression in engram cells

The study of molecules for the recruitment of associative memory neurons

Statistical analyzes

Reporting summary

Data availability

Code availability

References

Acknowledgements

Author information

Authors and Affiliations

Contributions

Corresponding author

Ethics declarations

Competing interests

Ethics approval and consent to participate

Consent for publication

Peer review

Peer review information

Additional information

Consent for publication

Supplementary information

Rights and permissions

About this article

Cite this article

Share this article

This article is cited by

Search

Quick links