Physiology & Behavior 291 (2025) 114788

Contents lists available at ScienceDirect

Physiology & Behavior

journal homepage: www.elsevier.com/locate/physbeh

Intermittent social isolation enhances social investigation but impairs social

memory in adult male mice
Shuyan Geng a,b,1, Zixu Zhang a , Xing Liu a , Haoyu Sun a,b , Tianxiang Xu a,b, Chuanyao Sun b,
Shengru Hu a,b , An Liu b , Zhiyuan Yang c,* , Wei Xie b,*, Mingdao Mu a,b,*,1
a
School of Medicine, Southeast University, 87 Dingjiaqiao Road, Nanjing, PR China
b
The Key Laboratory of Developmental Genes and Human Disease, Ministry of Education, The School of Life Science and Technology, Southeast University, 2 Sipailou
Road, Nanjing, PR China
c
School of Artificial Intelligence, Hangzhou Dianzi University, Hangzhou, PR China

A R T I C L E I N F O A B S T R A C T

Keywords: Social isolation profoundly impacts motivated behavior and neural plasticity. While the effects of acute and
Intermittent social isolation chronic social isolation have been extensively studied, the consequences of intermittent isolation during adult­
Social investigation hood, particularly relevant to modern lifestyles, remain poorly understood. This study investigated the impact of
Social interaction
intermittent social isolation (ISI) on social behavior and brain activation in adult male mice. Compared to group-
Social memory
Stress
housed controls, ISI males exhibited heightened social investigation and increased social interaction, reminiscent
Neuronal activity of craving-like behaviors. Intriguingly, this enhanced social investigation was accompanied by impaired social
recognition memory in a three-chamber sociability test. Furthermore, ISI induced distinct patterns of neural
activation in brain regions governing social processing, including the paraventricular nucleus of the hypothal­
amus, the intermediate part of lateral septum, the paraventricular nucleus of the thalamus, and the thalamic
periventricular gray. Notably, ISI did not affect anxiety-like behaviors or spatial memory, emphasizing its specific
impact on social domains. These findings demonstrate that ISI during adulthood selectively enhances social
investigation while disrupting social memory in male mice, possibly mediated by distinct neural circuits. Un­
derstanding the neurobiological mechanisms underlying these effects may inform interventions for individuals
experiencing social isolation, an increasingly prevalent phenomenon in modern society.

1. Introduction social isolation during critical developmental stages, such as adoles­

cence, can lead to a multitude of behavioral alterations in rodents,
The significance of social interactions for the well-being of social including increased anxiety-like behaviors, aggression, and impairments
species, such as humans and mice, has been widely recognized [1,2]. In in social recognition [6,7]. These behavioral changes have been asso­
today’s fast-paced and dynamic society, individuals often experience ciated with neurobiological alterations, such as decreased hippocampal
intermittent periods of social isolation due to various factors, including neurogenesis, altered synaptic plasticity, and neurotransmitter imbal­
work demands, travel, or personal circumstances. This form of social ances [5]. However, the impact of intermittent social isolation during
stress has become increasingly prevalent, yet its impact on behavior and adulthood, when the brain has matured and social behaviors are more
neural function remains relatively unexplored compared to the stable, remains poorly characterized.
well-studied effects of acute and chronic social isolation [2–5]. Investi­ Stress influences multiple neural pathways through varied molecular
gating the consequences of intermittent social isolation during adult­ signals, leading to different behavioral responses based on the stressor’s
hood is crucial for understanding its potential role in the development of nature, intensity, and duration [2,8]. The impact of stress on motivated
neuropsychiatric disorders and informing interventions for those expe­ behavior is particularly intricate; it can both enhance and diminish
riencing this unique form of social stress [2,4]. reward-seeking actions [8,9]. Viewing socializing as a stimulus akin to
Previous studies have consistently demonstrated that prolonged food provides a compelling framework to explore how social

* Corresponding authors.
E-mail addresses: [email protected] (Z. Yang), [email protected] (W. Xie), [email protected] (M. Mu).
1
These authors contribute equally.

https://doi.org/10.1016/j.physbeh.2024.114788
Received 1 May 2024; Received in revised form 10 December 2024; Accepted 20 December 2024
Available online 20 December 2024
0031-9384/© 2024 Elsevier Inc. All rights are reserved, including those for text and data mining, AI training, and similar technologies.
S. Geng et al. Physiology & Behavior 291 (2025) 114788

deprivation, particularly intermittent isolation, may heighten the drive intensifying the pursuit of social rewards, possibly through neuroplastic
for social interaction [2,10,11]. This offers valuable insights into the changes in brain areas related to social processing.
complex interplay between stress, social motivation, and behavioral The present study aims to address this knowledge gap by investi­
adaptations. The motivational dynamics of reward, deprivation, and gating the behavioral and neurobiological consequences of intermittent
craving, extensively studied in substance abuse, might similarly apply to social isolation, a form of social stress, in adult male mice. By examining
social interactions [3,12]. For instance, intermittent exposure to the effects of periodic social deprivation on social investigation/inter­
addictive substances increases craving [13,14], just as an intermittent action, social memory, and neural activity in different brain regions, we
high-sugar diet heightens the desire for sugar [15]. This suggests that seek to elucidate the precise impact of intermittent social contact on
periodic social contact deprivation might trigger craving-like states, mature mammals. We anticipate that the findings will not only provide

Fig. 1. Overview of the experimental design. (A) The timeline illustrates the housing conditions for both Group and ISI mice. Group mice were maintained in their
respective groups throughout the study period, while ISI mice were subjected to alternating cycles of 24-hour grouped housing and 24-hour isolation over a 14-day
period. (B) On Day 15, a series of behavioral assessments were conducted, including the social investigation test, free social interaction test, object investigation test,
spatial memory test, and the three-chamber social preference and social novelty test (consisting of habituation, sociability, and social novelty preference trials).
Additionally, the open field test, light/dark box test, and elevated plus maze test were performed to assess anxiety-related behavior and general locomotion. Sixty
minutes after the completion of 10-minute social stimuli, c-Fos immunofluorescence staining was performed to evaluate neural activation patterns.

2
S. Geng et al. Physiology & Behavior 291 (2025) 114788

greater clarity on nuances of mouse social behaviors, but also offer isolation, two unfamiliar mice of similar age and weight were placed in a
valuable parallels to human social modern dynamics. Particularly, it will 50 cm (L) x 50 cm (W) x 50 cm (H) open field arena for a 5-minute
enhance understanding of the impacts of social disruptions on mental interaction test. A condition-blind observer manually recorded the
health and social well-being. duration of active social behaviors such as sniffing, following, and
allogrooming. BORIS and MATLAB software were used to create dy­
2. Materials and methods namic graphs of social interaction times.

2.1. Animal handling and housing 2.2.4. Three-Chambered social preference and social novelty test
Adapted from a previous method [18], this test used a
Adult male C57BL/6 J mice (10–12 weeks old), sourced from Gem­ three-chamber, non-reflective white acrylic plexiglas apparatus to
Pharmatech Co. Ltd, Jiangsu Province, China, were housed in standard evaluate sociability and social novelty preference in mice over three
cages (290 mm x 178 mm x 160 mm) with 3–5 mice per cage on 10-minute phases. Initially, during the habituation phase, mice accli­
ventilated racks. After a 7-day acclimatization period, they either un­ mated to the chamber for 10 min. In the sociability phase, an unfamiliar
derwent intermittent social isolation or remained group-housed ac­ mouse (Stranger 1) was enclosed in square container in one side
cording to the experimental design. The facility maintained a chamber, while another container with an object (O1) was placed in the
temperature of 21–23 ◦ C, a humidity of 55 ± 10 %, and a 12-hour light/ opposite chamber. Following a 10-minute interval, the social novelty
dark cycle with lights on from 8:00 to 20:00. Mice had free access to phase commenced, wherein a new unfamiliar mouse (Stranger 2) was
autoclaved water and irradiated rodent chow (Xietong Organism, introduced into the previously container that contained O1. The time
Jiangsu Province, China). All procedures complied with the Guide to the spent investigating with Stranger 1 and Stranger 2 was analyzed using
Care and Use of Experimental Animals and received approval from the ANY-MAZE software. A preference for Stranger 2 indicated a novelty
Institutional Animal Care and Use Committee at Southeast University. bias and the ability to discriminate between novel and familiar social
stimuli.
2.2. Behavioral assays
2.2.5. Spatial memory test (Object location recognition)
Behavioral assessments were performed in daylight conditions. Spatial memory of mice was assessed using the Object Location
Equipment was sterilized with 70 % ethanol between trials. To minimize Recognition Test, which consisted of three 5-minute phases conducted
interference while optimizing animal usage, separate cohorts of mice in an open arena. Initially, during the habituation phase, mice were
were employed for different sets of tests: Cohort 1 underwent social allowed to acclimate to an empty arena for 5 min. This was immediately
investigation test, free social interaction test, object investigation test, followed by the familiarization phase, where mice explored two iden­
and spatial memory test; Cohort 2 was subjected to three-chamber social tical objects (O1 and O2) for 5 min, enabling them to form spatial
preference and social novelty test; Cohort 3 was utilized for open field memories of the objects’ locations. After a 1-hour inter-trial interval, the
test, light/dark box test, elevated plus maze test, and c-Fos immuno­ recognition phase commenced, in which one of the objects (O2) was
fluorescence staining. All tests within a cohort were completed on the relocated to a novel position, and the mice were reintroduced into the
same day with appropriate time intervals, as detailed in Fig. 1. Behav­ arena. Their behavior was recorded and analyzed for 5 min using the
ioral tracking and quantification were performed using ANY-MAZE behavioral analysis software ANY-MAZE. Investigation times with both
software (version 7.4, Stoelting Co., Wood Dale, IL, USA), while the novel and familiar locations were quantified. Preferential investi­
BORIS (Behavioral Observation Research Interactive Software) [16] gation of the relocated object was indicative of intact spatial memory,
facilitated detailed categorization of free social behaviors. whereas a lack of preference suggested spatial memory impairment.

2.2.1. Open field test (OFT) 2.2.6. Light/Dark box test

The Open Field Test assessed anxiety-like behaviors and locomotor This test evaluated anxiety-like behaviors in socially isolated mice
activity in mice following intermittent social isolation. The arena, con­ using a two-chamber apparatus (28 cm (L) x 25 cm (W) x 25 cm (H) for
structed from non-reflective white acrylic plexiglas, measured 50 cm (L) light chamber), featuring a darkened chamber (<10 lux) and a brightly
× 50 cm (W) × 50 cm (H). During each 5-minute session, the mice were illuminated one (about 600 lux), connected by a 3 cm × 3 cm opening.
placed in the open field apparatus and allowed to explore freely. ANY- During each 5-minute trial, mice were initially placed in the illuminated
MAZE software (version 7.4, Stoelting Co., USA) recorded the time chamber and video-tracked. The time spent in the lighted side and
spent in the central 25 cm x 25 cm zone and the total distance traveled. transitions between chambers were recorded and analyzed using ANY-
Less time in the central zone indicates anxiety-like behavior, while MAZE software. A preference for the dark chamber suggested higher
greater distance traveled measures locomotor activity [17]. anxiety levels [17].

2.2.2. Open field test with enclosed target 2.2.7. Elevated plus maze test
This test was conducted in a 50 cm (L) × 50 cm (W) × 50 cm (H) non- This test assessed anxiety-like behaviors in mice subjected to inter­
reflective white acrylic Plexiglas arena, with a transparent, perforated mittent social isolation using an elevated plus-shaped plexiglas maze.
acrylic square container measuring 5 cm (L) × 5 cm (W) × 20 cm (H) The maze consisted of a central platform (5.5 cm × 5.5 cm), two open
placed at the center. The container housed either a mouse or a non-social arms (30 cm (L) × 5.5 cm (W)), and two enclosed arms (35 cm (L) × 5.5
object. Subject mice were placed in the arena and given 5 min to freely cm (W) × 15 cm (H)), elevated 60 cm above the ground. Mice were
investigate their surroundings. The duration of each subject mouse’s video-tracked for 5 min after being placed in the center platform facing
head being within a 5-cm zone surrounding the container was quantified an open arm. Time spent in and entries into open versus closed arms
using ANY-MAZE software, which also generated activity heatmaps. The were quantified. More open arm time and entries indicated lower anx­
time spent in close proximity to the container was considered a measure iety levels [17].
of social investigation, with increased time spent near the container
housing another mouse being interpreted as an indicator of heightened 2.2.8. c-Fos immunofluorescence staining
social investigation. To identify brain regions potentially involved in ISI-induced social
investigation, we performed c-Fos immunofluorescence staining, a
2.2.3. Free social interaction test technique previously described [17]. Briefly, male mice from either the
To assess changes in social investigation due to intermittent Group or ISI conditions were exposed to a novel same-sex conspecific for

3
S. Geng et al. Physiology & Behavior 291 (2025) 114788

10 min. One hour post-stimulation, mice were deeply anesthetized and investigation (Fig. 3F; unpaired t-test, t(26) = 2.114, p = 0.0443).
transcardially perfused with 4 % paraformaldehyde. Brains were har­ However, in a subsequent phase, ISI mice, unlike Group mice, did not
vested, post-fixed, dehydrated in 30 % sucrose, embedded in OCT show a preference for a novel mouse over a familiar one, suggesting
compound, and sectioned into 40 μm slices using a cryostat. impaired social recognition memory (Fig. 3G-H; two-way ANOVA with
Free-floating sections were processed for immunofluorescence using a sidak multiple comparison test, t(52) = 2.540, p = 0.0281 for Group
primary antibody against c-Fos (1:1000, CST, #2250s) followed by a mice; t(52) = 1.718, p = 0.3015 for ISI mice in Fig. 3H). Quantitative
Cy3-conjugated goat anti-rabbit IgG (H + L) (1:1000, Proteintech, Cat analysis confirmed this significant deficit in social memory due to ISI
No: SA00009–2). Sections were counterstained with DAPI to label cell (Fig. 3I; unpaired t-test, t(26) = 2.176, p = 0.0397). These results indi­
nuclei and mounted on slides for confocal microscopy. The c-Fos cate that while ISI enhances social investigation, it concurrently impairs
expression patterns were compared between the Group and ISI condi­ social memory in adult male mice, highlighting the nuanced effects of
tions to identify brain regions exhibiting differential activation in social environments on cognitive functions.
response to social stimuli. Quantification of c-Fos positive cells was
performed using ImageJ software. For each condition, 3–4 mice were 3.3. Evaluating locomotion, anxiety, and spatial memory after
analyzed to ensure reproducibility of the findings. intermittent isolation in adult male mice

2.3. Data analysis To determine if the observed social behavior and memory changes in
ISI adult male mice relate to differences in locomotion, anxiety, or
Data are presented as mean ± SEM. GraphPad Prism software was spatial cognition, we conducted a series of behavioral tests. The open
used for statistical analyses. The unpaired t-test was used to compare field test revealed no notable differences in locomotor investigation or
differences between two independent groups. The paired t-test was used anxiety behaviors between ISI and group-housed males, as evidenced by
to compare differences between two related groups. Two-way ANOVA similar total distances traveled and time spent in center zones (Fig. 4A-C;
with sidak multiple comparison test was conducted to compare differ­ unpaired t-test, t(12) = 0.1160, p = 0.9096 in Fig. 4B; unpaired t-test, t
ences between multiple independent groups. p values <0.05 were (12) = 1.767, p = 0.1026 in Fig. 4C). In the light-dark box test, both ISI
considered statistically significant. and control groups displayed comparable anxiety levels, shown by their
similar distribution of time and transitions between light and dark areas
3. Results (Fig. 4D-F; unpaired t-test, t(12) = 0.3192, p = 0.7551 in Fig. 4E; un­
paired t-test, t(12) = 0.6799, p = 0.5095 in Fig. 4F). Similarly, the
3.1. Enhanced social investigation in adult male mice after intermittent elevated plus maze test results, assessing anxiety-like behavior, were
isolation alike for both ISI and group-housed mice in terms of time and entries
into open and closed arms (Fig. 4G-I; unpaired t-test, t(12) = 0.4985,
Upon establishing an intermittent social isolation (ISI) model p = 0.6272 in Fig. 4H; unpaired t-test, t(12) =0.1155, p = 0.9100 in
(Fig. 1), we assessed its impact on adult male mice’s social investigation Fig. 4I).
behavior. In an open field featuring a restrained, age-matched, same-sex For the novel object location recognition test, evaluating spatial
unfamiliar mouse (Fig. 2A), ISI mice spent significantly more time in the memory, indicated comparable cognitive performance in both groups,
central social interaction zone (the subject mouse’s head within a 5-cm as seen in their investigation preferences (Fig. 5; two-way ANOVA with
region surrounding the container) compared to group-housed controls, sidak multiple comparison test, t(28) = 0.2059, p= 0.9739 for Group
indicating heightened social investigation. Heatmap analysis revealed mice in Fig. 5C; t(28) =0.2435, p = 0.9637 for ISI mice in Fig. 5C; two-
ISI mice’s extended stays in the central social zone (Fig. 2B), and time way ANOVA with sidak multiple comparison test, t(28) = 2.905,
course data indicated more social investigation with the stranger (S1) by p = 0.0141 for Group mice in Fig. 5F; t(28) =3.153, p = 0.0076 for
ISI mice (Fig. 2C; unpaired t-test, t(14) = 4.625, p = 0.0004). During ISI mice in Fig. 5F; two-way ANOVA with sidak multiple comparison test
free social interaction tests, ISI mice engaged in longer social in­ indicated no significant difference in O2 investigation time between
teractions with an unfamiliar mouse compared to group-housed controls Group and ISI mice t(28) =0.0043, p = 0.9999 in Fig. 5F).
(Fig. 2D-G; unpaired t-test, t(14) = 4.203, p = 0.0009 in Fig. 2F). These collective results from diverse, established behavioral tests
However, both groups spent similar amounts of time investigating a suggest that 14-day ISI does not significantly impact locomotion, anxiety
novel object (Fig. 2H-J; unpaired t-test, t(16) = 0.2118, p = 0.8349 in behaviors, or spatial memory in adult male mice.
Fig. 2J), highlighting ISI mice’s enhanced preference for social over
object interactions. 3.4. Neurobiological effects of intermittent social isolation on brain
To further elucidate the differential effects of intermittent and pro­ regions
longed isolation, we conducted additional experiments comparing mice
subjected to 14 days of continuous isolation with those housed in To investigate the neurobiological mechanisms underlying the
groups. As illustrated in Supplementary Figure 1, prolonged isolation behavioral changes observed in intermittently socially isolated adult
did not significantly influence social investigation, which aligns with male mice, we compared neural activity in ISI and group-housed mice
previous reports [19–21]. Taken together, these findings suggest that following exposure to an unfamiliar mouse. Brains were perfused one
intermittent isolation, rather than prolonged isolation, notably enhances hour post-exposure, and immunofluorescence staining for c-Fos, a
social investigation in adult male mice. neuronal activation marker, was performed (Fig. 6A). Through a
comprehensive c-Fos screening of various brain regions, including the
3.2. Intermittent isolation impairs social memory in adult male mice forebrain, midbrain, and hindbrain, we discovered that certain areas
exhibited significantly higher activation in ISI mice compared to group-
To further explore the impact of intermittent isolation on social housed controls when exposed to social stimuli (Fig. 6B). Notably, the
memory, we used a three-chamber social preference and social novelty paraventricular nucleus (PVN) (Fig. 6B,C; unpaired t-test, t(6) = 6.864, p
test in adult male mice (Fig. 3A-C). Initially, both control and ISI mice = 0.0005), the intermediate part of the lateral septum (LSI) (Fig. 6B, D;
showed a strong preference for interacting with an unfamiliar mouse unpaired t-test, t(6) = 3.052, p = 0.0224), the paraventricular nucleus of
over an object (Fig. 3D-E; two-way ANOVA with sidak multiple com­ the thalamus (PVT) (Fig. 6B; unpaired t-test, t(6) = 2.715, p = 0.0349)
parison test, t(52) = 3.002, p < 0.0082 for Group mice; t(52) = 7.161, and the thalamic periventricular gray (PVG) (Fig. 6B; unpaired t-test, t
p < 0.0001 for ISI mice in Fig. 3E), with ISI mice demonstrating even (6) = 3.447, p = 0.0137) displayed enhanced c-Fos expression in ISI
greater social preference index, indicating heightened social mice relative to controls. This indicates heightened neural response to

4
S. Geng et al. Physiology & Behavior 291 (2025) 114788

Fig. 2. Enhanced Social Investigation in Adult Male Mice after Intermittent Isolation. (A) Experimental setup for the 5-minute social investigation test. (B)
Average heatmaps illustrating the investigation frequencies of Group and ISI mice with a stimulus mouse (S1). (C) Time spent by the subject mouse’s head within a 5-
cm zone surrounding S1 in the open field. (D) Experimental setup for the 5-minute free social interaction test. (E) Cumulative duration of free social interaction. (F)
Social interaction time. (G) Representative raster plots depicting social interaction patterns in Group and ISI mice. (H) Experimental setup for the 5-minute object
investigation test. (I) Average heatmaps showing the interaction frequencies of Group and ISI mice with a stimulus object (O1). (J) Time spent by the subject mouse’s
head within a 5-cm region around O1 in the open field. (***, p< 0.001; ns, no significant difference, p > 0.05).

5
S. Geng et al. Physiology & Behavior 291 (2025) 114788

Fig. 3. Intermittent Isolation Impairs Social Memory in Adult Male Mice. (A-C) Experimental design for assessing sociability and social novelty preference, (A)
with initial 10-minute habituation, (B) followed by a 10-minute interaction with an unfamiliar mouse (S1) versus an object (O1) in Trial 2 for sociability, and (C) after
a 10-minute interval, a 10-minute choice between a novel mouse (S2) and the familiar mouse (S1) in Trial 3 for social novelty preference. (D, G) Heatmaps illus­
trating the interaction frequencies of group-housed (Group) and intermittently socially isolated (ISI) mice in Trials 2 and 3, respectively. (E) Investigation time during
Trial 2. (F) Social preference index for Trial 2, calculated as (S1 investigation time - O1 investigation time) / (S1 investigation time + O1 investigation time). (H)
Investigation time during Trial 3. (I) Social novelty preference index for Trial 3, calculated as (S2 investigation time - S1 investigation time) / (S2 investigation time +
S1 investigation time). (*, p < 0.05; **, p < 0.01; ***, p < 0.001; ns, no significant difference, p > 0.05).

social stimulus in these regions in ISI mice, suggesting that ISI may Intermittent social isolation, a form of social stress, profoundly
trigger neural plasticity changes in specific brain areas, potentially impacted motivated behavior in adult male mice. In this study,
influencing the altered social behaviors observed in ISI mice. following periods of isolation, males exhibited heightened social
investigation and increased social interaction, resembling craving-like
4. Discussion behaviors seen in addiction models [14,15]. This suggests that inter­
mittent isolation might create a social deficit, triggering a compensatory
This study provides novel insights into the behavioral and neurobi­ need for social contact [2,3,11]. The similarity between this social drive
ological consequences of intermittent social isolation during adulthood, and reward-seeking behaviors in substance abuse highlights the potent
an area that has received relatively less research attention compared to reinforcing nature of social interactions and the potential for social
long-term isolation or early developmental deprivation. Our findings in deprivation to evoke craving-like states.
adult male mice reveal a complex interplay between stress, social Notably, while boosting social investigation, intermittent isolation
investigation, and social memory, highlighting the importance of weakened social memory in adult male mice. Their difficulty in recog­
consistent social interactions for maintaining optimal social functioning. nizing familiar from novel peers after isolation periods aligns with

6
S. Geng et al. Physiology & Behavior 291 (2025) 114788

Fig. 4. Locomotor Activity and Anxiety-related Behavioral Assessments in Adult Male Mice Following Intermittent Social Isolation. (A) Heatmaps from the
Open Field Test (OFT) compare the activity patterns of group-housed (Group) and intermittently socially isolated (ISI) mice during a 5-minute test. (B) Quantification
of the total distance travelled and (C) the time spent in the center of the open field arena. (D) Heatmaps from the Light/Dark Test (LDT) compare the behavior of
Group and ISI mice during a 5-minute test. (E) Quantification of the time spent in the dark compartment and (F) the number of entries into the light compartment. (G)
Heatmaps from the Elevated Plus Maze Test (EPM) compare the investigation patterns of Group and ISI mice during a 5-minute test. (H) Quantification of the time
spent in the open arms and (I) the number of entries into the open arm. (ns, no significant difference, p > 0.05).

findings from long-term or adolescent social isolation studies [22,23]. memory.

This similarity indicates that both short-term and prolonged social dis­ The dissociation between heightened social investigation and
ruptions can similarly impact social memory processing, including its impaired social memory in intermittently isolated adult male mice
consolidation and retrieval. This emphasizes the delicate nature of social suggests a complex interplay between distinct neural circuits governing
cognition, highly susceptible to environmental alterations, and the these aspects of social behavior. The differential neural activation
critical importance of regular social interaction in maintaining social observed in ISI males following social stimuli exposure indicates that

7
S. Geng et al. Physiology & Behavior 291 (2025) 114788

Fig. 5. Intermittent Isolation Does Not Significantly Affect Spatial Recognition in Adult Male Mice. (A, D) Setup for the two-trial novel object recognition task:
Familiarization Stage (Stage 1) and Recognition Stage (Stage 2). (B, E) Heatmaps depict mouse activity around objects O1 and O2 for group-housed (Group) and
intermittent social isolation (ISI) conditions during both trials. (C, F) Investigation time of the mice to different object (*, p 〈 0.05; ns, no significant difference,
p〉 0.05).

specific brain regions may play crucial roles in mediating the effects of environmental characteristics, combined with the fact that all behav­
intermittent isolation on social investigation and memory. Brain re­ ioral tests were conducted during the daytime under bright lighting
gions, such as the PVN, LSI, PVT, and PVG, have been implicated in the conditions [35,36], may have introduced an additional level of stress to
regulation of social behavior and motivation. The LSI, a region inter­ the mice, thereby reducing their baseline investigation behavior.
connected with the hippocampus and hypothalamus, has been associ­ Importantly, these testing conditions were applied uniformly across all
ated with social behavior and reward [24,25]. While the PVN, a key groups, ensuring valid internal comparisons and reliable conclusions
component of the hypothalamic-pituitary-adrenal (HPA) axis, is within our study. However, differences in key experimental variables,
involved in stress responses and social behavior regulation [26,27]. In such as arena size, wall and floor color, lighting conditions, or other
addition to the PVT, with its neural connectivity to brain regions methodological factors, may account for the observed discrepancies
involved in social processing and reward, has been implicated in the with published data. Future studies are warranted to systematically
regulation of social motivation and behavior [28–30]. The PVG, adja­ evaluate the impact of such environmental variables on baseline inves­
cent to the third ventricle, integrates emotional and autonomic re­ tigation behavior, which could further clarify the interplay between
sponses and coordinates behavioral responses to stressful stimuli [31, experimental conditions and behavioral outcomes in similar paradigms.
32], although its specific role in social behavior remains less explored. While our study provides valuable insights into the effects of inter­
The enhanced activation of these regions in isolated males suggests that mittent social isolation on adult male mice, it is important to acknowl­
intermittent isolation may induce neuroplastic changes in the neural edge some limitations and potential avenues for future research. First,
circuits underlying social investigation and memory, potentially our study focused exclusively on male mice, and therefore, the findings
contributing to the observed behavioral alterations. Further research is may not be generalizable to females. Given the well-documented sex
needed to elucidate the precise mechanisms by which these brain re­ differences in social behavior and stress responses [37], future studies
gions interact and contribute to the complex effects of intermittent social should investigate the impact of intermittent isolation on female mice to
isolation on social behavior and cognition. determine whether they exhibit similar or distinct behavioral and
Regarding the relatively low investigation times observed in the neurobiological changes compared to males. This would provide a more
group-housed control mice compared to published data, this outcome comprehensive understanding of the sex-specific effects of intermittent
likely reflects differences in experimental conditions. Behavioral studies isolation and inform the development of targeted interventions. Second,
are inherently sensitive to various factors, including mouse strain, batch our study examined the effects of intermittent isolation in adult mice,
variability, and other environmental influences [33,34]. In our study, all but it would be valuable to explore how this form of social stress in­
groups, including the control (group-housed) and experimental (isola­ fluences mice at different developmental stages. The impact of inter­
tion or intermittent isolation) groups, were assessed in a relatively large mittent isolation may vary depending on the age at which it occurs, with
open field with white walls and a white floor. These specific potentially distinct consequences for juvenile, adolescent, and aged

8
S. Geng et al. Physiology & Behavior 291 (2025) 114788

Fig. 6. Neurobiological response to intermittent social isolation in mice. (A) Experimental design for c-Fos staining in group-housed and intermittently socially
isolated (ISI) mice exposed to a stimulus mouse (S1). (B) Quantification of c-Fos-positive cells in various brain regions of group-housed and ISI mice, including the
anterior olfactory nucleus (AO), the paraventricular nucleus of the thalamus (PVT), the lateral periaqueductal gray (LPAG), the intermediate part of the lateral
septum (LSI), the paraventricular nucleus of the hypothalamus (PVN), the rostral periolivary region (RPO), the ventral tegmental area (VTA), the medial amygdala
(MeA), the ventral lateral geniculate nucleus (VLG), the thalamic periventricular gray (PVG), and Zona Incerta (ZI). (C-F) Representative images of c-Fos immu­
noreactivity in different brain areas, including (C) PVN, (D) LSI, (E) AO, and (F) LPAG. Statistical significance is denoted as follows: *, p〈 0.05; ***, p < 0.001; ns, no
significant difference, p〉 0.05. Scale bar, 150 μm.

mice [38]. Investigating the age-dependent effects of intermittent identified several specific brain regions (including PVN, LSI, PVT, and
isolation would provide insights into the developmental trajectories of PVG) that exhibited increased neural activation in response to social
social motivation and memory and help identify critical periods of stimuli following intermittent isolation, further research is needed to
vulnerability or resilience to social stress. Third, while our study elucidate the precise neurobiological mechanisms underlying the

9
S. Geng et al. Physiology & Behavior 291 (2025) 114788

observed behavioral changes. Future studies could employ techniques Supplementary materials
such as optogenetics, chemogenetics, or targeted pharmacological in­
terventions to manipulate the activity of specific neural circuits or Supplementary material associated with this article can be found, in
neurotransmitter systems involved in social motivation and memory, the online version, at doi:10.1016/j.physbeh.2024.114788.
providing a more detailed understanding of the causal relationships
between neural activity patterns and behavioral outcomes. Data availability
From a translational perspective, our findings in adult male mice
model provide a novel framework for investigating the neurobiological Data will be made available on request.
basis of social interaction and social motivation in the context of inter­
mittent isolation, with significant implications for understanding human References
social behavior in situations of periodic social deprivation. The craving-
like state induced by intermittent isolation in male mice parallels the [1] A.K. Beery, D. Kaufer, Stress, social behavior, and resilience: insights from rodents,
increased desire for social interaction often reported by individuals Neurobiol Stress. 1 (2015) 116–127.
[2] C.R. Lee, A. Chen, K.M. Tye, The neural circuitry of social homeostasis:
experiencing loneliness or social disconnection. Moreover, the impair­ consequences of acute versus chronic social isolation, Cell 184 (2021) 1500–1516.
ment in social memory observed in isolated males may have relevance [3] G.A. Matthews, K.M. Tye, Neural mechanisms of social homeostasis, Ann. N. Y.
for understanding the social cognitive deficits associated with certain Acad. Sci. 1457 (2019) 5–25.
[4] A. Steptoe, A. Shankar, P. Demakakos, J. Wardle, Social isolation, loneliness, and
psychiatric disorders characterized by social withdrawal or isolation. all-cause mortality in older men and women, Proc. Natl. Acad. Sci. 110 (2013)
5797–5801.
5. Conclusion [5] A. Orben, L. Tomova, S.-J. Blakemore, The effects of social deprivation on
adolescent development and mental health, The Lancet Child Adolesc. Health 4
(2020) 634–640.
This study underlines the significant effects of intermittent social [6] D.C. Li, E.A. Hinton, S.L. Gourley, Persistent behavioral and neurobiological
isolation on social investigation and memory in adult male mice, of­ consequences of social isolation during adolescence. Seminars in Cell &
Developmental Biology, Elsevier, 2021, pp. 73–82.
fering new insights into how stress intersects with motivated behavior. It
[7] L.F. Parise, C.J. Burnett, S.J. Russo, Early life stress and altered social behaviors: a
highlights the crucial role of regular social interaction in optimal social perspective across species, Neurosci. Res. (2023).
functioning and suggests that periodic social deprivation can trigger [8] N.G. Hollon, L.M. Burgeno, P.E. Phillips, Stress effects on the neural substrates of
motivated behavior, Nat. Neurosci. 18 (2015) 1405–1412.
craving-like states and hinder social cognition. However, future research
[9] M. Ironside, P. Kumar, M.-S. Kang, D.A. Pizzagalli, Brain mechanisms mediating
should explore sex and age differences and the specific neurobiological effects of stress on reward sensitivity, Curr. Opin. Behav. Sci. 22 (2018) 106–113.
mechanisms behind these behavioral changes. Expanding our knowl­ [10] V. Trezza, P. Campolongo, L.J. Vanderschuren, Evaluating the rewarding nature of
edge of how social experiences, motivation, and cognition interact social interactions in laboratory animals, Dev. Cogn. Neurosci. 1 (2011) 444–458.
[11] L. Tomova, K.L. Wang, T. Thompson, G.A. Matthews, A. Takahashi, K.M. Tye, et al.,
across various demographics and life stages can inform interventions for Acute social isolation evokes midbrain craving responses similar to hunger, Nat.
those facing social isolation or challenges in social functioning. Neurosci. 23 (2020) 1597–1605.
[12] D. Sazhin, A.M. Frazier, C.R. Haynes, C.R. Johnston, I.K.-Y. Chat, J.B. Dennison, et
al., The role of social reward and corticostriatal connectivity in substance use,
Author notes J. Psychiatry Brain Sci. 5 (2020).
[13] S. Shiffman, M.S. Dunbar, X. Li, S.M. Scholl, H.A. Tindle, S.J. Anderson, et al.,
These authors contributed equally: Mingdao Mu, Shuyan Geng Craving in intermittent and daily smokers during ad libitum smoking, Nicotine
Tobacco Res. 16 (2014) 1063–1069.
[14] A. Garcia, I. Webb, L. Yager, M. Seo, S. Ferguson, Intermittent but not continuous
CRediT authorship contribution statement access to cocaine produces individual variability in addiction susceptibility in rats,
Psychopharmacol. (Berl.) 237 (2020) 2929–2941.
[15] J.J. DiNicolantonio, J.H. O’Keefe, W.L. Wilson, Sugar addiction: is it real? A
Shuyan Geng: Methodology, Investigation, Formal analysis. Zixu narrative review, Br. J. Sports Med. 52 (2018) 910–913.
Zhang: Software, Methodology, Investigation. Xing Liu: Software, [16] O. Friard, M. Gamba, BORIS: a free, versatile open-source event-logging software
Methodology, Investigation. Haoyu Sun: Methodology, Data curation. for video/audio coding and live observations, Methods Ecol. Evol. 7 (2016)
1325–1330.
Tianxiang Xu: Methodology, Data curation. Chuanyao Sun: Valida­
[17] M.-D. Mu, H.-Y. Geng, K.-L. Rong, R.-C. Peng, S.-T. Wang, L.-T. Geng, et al.,
tion, Methodology. Shengru Hu: Methodology. An Liu: Writing – re­ A limbic circuitry involved in emotional stress-induced grooming, Nat. Commun.
view & editing, Validation, Resources. Zhiyuan Yang: Writing – review 11 (2020) 2261.
& editing, Validation, Supervision. Wei Xie: Validation, Supervision, [18] R. Dang, Y. Zhou, Y. Zhang, D. Liu, M. Wu, A. Liu, et al., Regulation of social
memory by lateral entorhinal cortical projection to dorsal hippocampal CA2,
Resources, Project administration, Conceptualization. Mingdao Mu: Neurosci. Bull. 38 (2022) 318–322.
Writing – review & editing, Writing – original draft, Validation, Super­ [19] S. Musardo, A. Contestabile, M. Knoop, O. Baud, C. Bellone, Oxytocin neurons
vision, Project administration, Methodology, Investigation, Funding mediate the effect of social isolation via the VTA circuits, Elife 11 (2022).
[20] J.K. Rivera-Irizarry, M.J. Skelly, K.E. Pleil, Social Isolation Stress in Adolescence,
acquisition, Formal analysis, Data curation, Conceptualization. but not Adulthood, Produces Hypersocial Behavior in Adult Male and Female
C57BL/6J Mice, Front. Behav. Neurosci. 14 (2020) 129.
Declaration of competing interest [21] Y.K. Ma, P.Y. Zeng, Y.H. Chu, C.L. Lee, C.C. Cheng, C.H. Chen, et al., Lack of social
touch alters anxiety-like and social behaviors in male mice, Stress. 25 (2022)
134–144.
The authors declare no competing interests. [22] N. Leser, S. Wagner, The effects of acute social isolation on long-term social
recognition memory, Neurobiol. Learn. Mem. 124 (2015) 97–103.
[23] Y. Xiong, H. Hong, C. Liu, Y.Q. Zhang, Social isolation and the brain: effects and
Acknowledgments mechanisms, Mol. Psychiatry 28 (2023) 191–201.
[24] R. Menon, T. Süß, V.E. de Moura Oliveira, I.D. Neumann, A. Bludau, Neurobiology
This work was supported by Natural Science Foundation of China of the lateral septum: regulation of social behavior, Trends Neurosci. 45 (2022)
27–40.
(NSFC 82201705, M.M.), Natural Science Foundation of Jiangsu Prov­ [25] L. Li, R. Durand-de Cuttoli, A.V. Aubry, C.J. Burnett, F. Cathomas, L.F. Parise, et al.,
ince (BK2022040821, M.M.), Medical Specialty Ascent Program of Social trauma engages lateral septum circuitry to occlude social reward, Nature
Southeast University (4060692202/005, M.M.), Fundamental Research 613 (2023) 696–703.
[26] A.L. Heck, J.A. Sheng, A.M. Miller, S.A. Stover, N.J. Bales, S.M. Tan, et al., Social
Funds for the Central Universities (1124007126, to M.M.), and the
isolation alters hypothalamic pituitary adrenal axis activity after chronic variable
Scientific and Technological Innovation 2030 Major Projects (STI2030- stress in male C57BL/6 mice, Stress. 23 (2020) 457–465.
Major Projects 2021ZD0204001 to W.X. and 2021ZD0204002 to M.M.). [27] J.P. Herman, N. Nawreen, M.A. Smail, E.M. Cotella, Brain mechanisms of HPA axis
We appreciate Dr. Moyi Li, Dr. Junhua Geng, and Dr. Yanan Wang for regulation: neurocircuitry and feedback in context Richard Kvetnansky lecture,
Stress. 23 (2020) 617–632.
their valuable discussions and suggestions on experimental design. We [28] M.A. Penzo, C. Gao, The paraventricular nucleus of the thalamus: an integrative
also thank Dr. Yue Gao and Dr. Xun Lu for their valuable help. node underlying homeostatic behavior, Trends. Neurosci. 44 (2021) 538–549.

10
S. Geng et al. Physiology & Behavior 291 (2025) 114788

[29] K. Yamamuro, L.K. Bicks, M.B. Leventhal, D. Kato, S. Im, M.E. Flanigan, et al., [34] F.C. Hankenson, E.M. Prager, B.R. Berridge, Advocating for Generalizability:
A prefrontal-paraventricular thalamus circuit requires juvenile social experience to accepting Inherent Variability in Translation of Animal Research Outcomes, Annu
regulate adult sociability in mice, Nat. Neurosci. 23 (2020) 1240–1252. Rev. Anim. Biosci. 12 (2024) 391–410.
[30] L.D. Hill-Bowen, J.S. Flannery, R. Poudel, Paraventricular Thalamus Activity [35] A. Kapogiannatou, E. Paronis, K. Paschidis, A. Polissidis, N.G. Kostomitsopoulos,
during Motivational Conflict Highlights the Nucleus as a Potential Constituent in Effect of light colour temperature and intensity on the behaviour of male C57CL/6J
the Neurocircuitry of Addiction, J. Neurosci. 40 (2020) 726–728. mice, Appl. Anim. Behav. Sci. 184 (2016) 135–140.
[31] H. Luo, Y. Huang, A.L. Green, T.Z. Aziz, X. Xiao, S. Wang, Neurophysiological [36] L.K. Fonken, M.S. Finy, J.C. Walton, Z.M. Weil, J.L. Workman, J. Ross, et al.,
characteristics in the periventricular/periaqueductal gray correlate with pain Influence of light at night on murine anxiety- and depressive-like responses, Behav.
perception, sensation, and affect in neuropathic pain patients, Neuroimage Clin. 32 Brain Res. 205 (2009) 349–354.
(2021) 102876. [37] S. Yan, R. Xu, T.D. Stratton, V. Kavcic, D. Luo, F. Hou, et al., Sex differences and
[32] D. Nandi, X. Liu, C. Joint, J. Stein, T. Aziz, Thalamic field potentials during deep psychological stress: responses to the COVID-19 pandemic in China, BMC. Public
brain stimulation of periventricular gray in chronic pain, Pain. 97 (2002) 47–51. Health 21 (2021) 1–8.
[33] B. Voelkl, N.S. Altman, A. Forsman, W. Forstmeier, J. Gurevitch, I. Jaric, et al., [38] D.M. Magalhães, M. Mampay, A.M. Sebastião, G.K. Sheridan, C.A. Valente, Age-
Reproducibility of animal research in light of biological variation, Nat. Rev. related impact of social isolation in mice: young vs middle-aged, Neurochem. Int.
Neurosci. 21 (2020) 384–393. 174 (2024) 105678.

11