Translational Psychiatry www.nature.

com/tp

ARTICLE OPEN

Beneficial effects of running exercise on hippocampal

microglia and neuroinflammation in chronic unpredictable
stress-induced depression model rats
Kai Xiao1,2, Yanmin Luo2,3, Xin Liang2,4, Jing Tang1,2, Jin Wang1,2, Qian Xiao5, Yingqiang Qi6, Yue Li1,2, Peilin Zhu1,2, Hao Yang2,7,
✉ ✉
Yuhan Xie1,2, Hong Wu 1,2 and Yong Tang 1,2

© The Author(s) 2021

Running exercise has been shown to relieve symptoms of depression, but the mechanisms underlying the antidepressant effects
are unclear. Microglia and concomitant dysregulated neuroinflammation play a pivotal role in the pathogenesis of depression.
However, the effects of running exercise on hippocampal neuroinflammation and the number and activation of microglia in
depression have not been studied. In this study, rats were subjected to chronic unpredictable stress (CUS) for 5 weeks followed by
treadmill running for 6 weeks. The depressive-like symptoms of the rats were assessed with a sucrose preference test (SPT).
Immunohistochemistry and stereology were performed to quantify the total number of ionized calcium-binding adapter molecule
1 (Iba1)+ microglia, and immunofluorescence was used to quantify the density of Iba1+/cluster of differentiation 68 (CD68)+ in
subregions of the hippocampus. The levels of proinflammatory cytokines in the hippocampus were measured by qRT-PCR and
ELISA. The results showed that running exercise reversed the decreased sucrose preference of rats with CUS-induced depression. In
addition, CUS increased the number of hippocampal microglia and microglial activation in rats, but running exercise attenuated the
CUS-induced increases in the number of microglia in the hippocampus and microglial activation in the dentate gyrus (DG) of the
hippocampus. Furthermore, CUS significantly increased the hippocampal levels of inflammatory factors, and the increases in
inflammatory factors in the hippocampus were suppressed by running exercise. These results suggest that the antidepressant
effects of exercise may be mediated by reducing the number of microglia and inhibiting microglial activation and
neuroinflammation in the hippocampus.

Translational Psychiatry (2021)11:461 ; https://doi.org/10.1038/s41398-021-01571-9

INTRODUCTION regulation and control of emotion [10]. Postmortem studies have

Major depressive disorder (MDD) is a common mental disorder provided evidence that several inflammatory genes are upregu-
characterized primarily by symptoms of pervasive depressed lated in the hippocampi of MDD patients [11], and the expression
mood along with anhedonia and feelings of guilt [1]. It is of some inflammation-related genes has been shown to be
estimated that the aggregate prevalence and lifetime risk of correlated with a decreased hippocampal volume [12]. Moreover,
depression are approximately 12.9 and 10.8%, respectively [2]. As several studies also demonstrated that the levels of inflammatory
the leading cause of disability and suicide among psychiatric mediators are increased in the hippocampi of depression model
disorders, MDD severely affects patients’ quality of life and rats subjected to chronic stress [13, 14]. Microglia, the resident
increases the global disease burden [3, 4]. immune cells of the central nervous system (CNS) [15], plays a
Although the exact etiology and pathogenesis of depression central role in mediating the inflammatory response in MDD [16].
remain unclear, increasing evidence has shown that the inflam- Mounting evidence indicates that microglia in several brain
matory response is a main factor involved in the development of regions, especially the hippocampus, might be linked to the
the disorder [5, 6]. Numerous clinical studies have verified that the pathogenesis of depression in both depressed suicide cases and
circulating levels of proinflammatory cytokines in depressed animal models [17–19]. The hippocampus comprises several
individuals are higher than those in nondepressed subjects subregions, including the cornu ammonis (CA) 1, CA2/3, and
[7, 8]. In addition, chronic mild stress substantially increases the dentate gyrus (DG) subregions, and each subregion plays specific
risk of elevated inflammatory cytokine levels in the brains of rats functional roles [20, 21]. Several studies have reported that the
[9]. The hippocampus is the key structure responsible for the number of microglia in the CA2/3 and DG is increased in

1
Department of Histology and Embryology, Chongqing Medical University, 400016 Chongqing, P. R. China. 2Laboratory of Stem Cells and Tissue Engineering, Chongqing Medical
University, 400016 Chongqing, P. R. China. 3Department of Physiology, Chongqing Medical University, 400016 Chongqing, P. R. China. 4Department of Pathophysiology,
Chongqing Medical University, 400016 Chongqing, P. R. China. 5Department of Radioactive Medicine, Chongqing Medical University, 400016 Chongqing, P. R. China. 6Institute of
Life Science, Chongqing Medical University, 400016 Chongqing, P. R. China. 7Department of Biomedical Engineering, Chongqing Medical University, 400016 Chongqing, P. R.
China. ✉email: [email protected]; [email protected]

Received: 23 March 2021 Revised: 5 August 2021 Accepted: 19 August 2021

 K. Xiao et al.
2
depression animal models exposed to chronic stress [17, 19, 22]. In Finally, we investigated the effects of treadmill exercise on
contrast, Kreisel et al. reported that chronic unpredictable stress microglial activation and the levels of inflammatory cytokines
(CUS)-exposed mice exhibit decreased numbers of microglia in the using immunofluorescence assays, polymerase chain reaction
DG of the hippocampus [23]. Thus, the types of changes in the (PCR), and enzyme-linked immunosorbent assay (ELISA).
number of microglia that occur in the hippocampal subregions in
depression models are still controversial. To further explore this
issue, the current study used an unbiased stereological technique MATERIALS AND METHODS
to precisely quantify microglia in specific subregions of the Animals
hippocampus (the CA1, CA2/3, and DG) in depression model rats. Sixty male Sprague-Dawley rats (6–8 weeks old and weighing 200 ± 20 g)
Microglia are characterized by different states, including a were provided by the Experimental Animal Center of Chongqing Medical
quiescent state in which the cells are ramified or resting and an University (Chongqing, P. R. China). The rats were housed under a 12-h
activated state in which the cells are hyper-ramified or phagocytic light/dark cycle at an appropriate temperature (22 ± 2 °C) and provided
access to food and water. Before the interventions began, all rats were
[24]. In their quiescent state, ramified microglia have been found allowed to adapt to the housing conditions for 2 weeks. During the animal
to monitor synaptic function and actively respond to neuronal experiment, the animals in each group were treated by the investigators
activity [25]. Once brain homeostasis is disturbed, microglia without blinding. All procedures (Fig. 1) were approved by the Animal Care
integrate inflammatory signals in the periphery and the CNS and Committee of Chongqing Medical University and performed in accordance
enter an activated state through structural and functional with the National Institutes of Health Guide for the Care and Use of
changes [26]. The limitations of the morphological classification Laboratory Animals.
standards of microglia in different states have promoted more
functional research. A neuroimaging study demonstrated sig- CUS paradigm
nificant elevations in the density of translocator protein, a marker The CUS procedure was performed as described in our previous reports
of microglial activation, in the hippocampi of patients currently [32, 41]. After a 2-week baseline adjustment of sucrose preference, rats
experiencing a major depressive episode [27]. Furthermore, with sucrose preference above baseline were randomly divided into the
recent preclinical studies have demonstrated that the activation control group (n = 23) and the CUS group (n = 37). The control rats were
of microglia and the release of proinflammatory cytokines can be housed at a density of four to five rats per cage, and the CUS rats
1234567890();,:

induced by chronic stress and can increase vulnerability to were housed in individual cages. The two groups of rats were housed in
separate rooms and had no contact with each other. The model rats were
depressive-like behaviors [28]. Therefore, activated microglia and subjected to different types of stressors, including water or food
concomitant dysregulated neuroinflammation in the hippocam- deprivation, empty bottle exposure, intermittent illumination, cage tilting,
pus might play a vital role in the development of depression. restraint, damp bedding exposure, no bedding exposure, cold/hot stress,
However, the change in the number of activated microglia in the strobe light exposure, noise exposure, electric foot shock exposure, tail
hippocampus that occur in the context of depression remain pinching, and cage shaking. Two types of stressors were applied daily, and
unclear. To further investigate the functional changes in microglia no stimulus was repeated within 5 days (Table 1). The entire CUS
associated with depression, we analyzed microglial activation and procedure lasted for 5 weeks. Then the CUS rats were randomly divided
concomitant inflammatory cytokine levels in the hippocampi of into the CUS + standard group (CUS + STD group, n = 17) and the CUS +
CUS-exposed rats. running group (CUS + RUN group, n = 20).
Physical exercise is an effective nonpharmaceutical treatment
for MDD. Various clinical surveys have demonstrated that exercise Treadmill running protocol
is effective in improving depressive symptoms in individuals with In the CUS + RUN group, the rats ran on a six-lane motorized treadmill for
mild or moderate depression [29, 30]. Our previous rodent studies 20 min each day, 5 days per week for 6 weeks (Fig. 1). During the first week,
the treadmill speed was gradually increased from 10 to 20 m/min. During
further revealed that moderate-intensity treadmill running exer-
the remaining 5 weeks, the speed was maintained at 20 m/min. This
cise can improve depressive-like behaviors in rats exposed to CUS moderate-intensity treadmill running pattern was used as described in our
[31–33]. However, the mechanisms underlying the antidepressant previous study [32].
effects of exercise in MDD are unclear. As mentioned above, During the running exercise period, the rats in the CUS + STD group and
hippocampal microglia and the inflammatory response are the CUS + RUN group were housed at a density of one rat per cage, while
associated with the development of depression. An inhibitor of the rats in the control group remained housed under normal circum-
activated microglia, minocycline, has been shown to attenuate stances with four to five rats per cage.
neuroinflammation induced by lipopolysaccharide or interferon-
alpha and ameliorate depressive-like behavior [34, 35], implying Behavior tests
that inhibiting microglial activation might be a potential Body weight test. The body weights of the rats were measured during a
therapeutic strategy for depression. Numerous clinical and fixed time frame each week.
preclinical studies have suggested that exercise can prevent or
attenuate depression and that this effect is correlated with Sucrose preference test. The SPT was carried out as previously described
decreases in neuroimmune factor levels [36, 37]. In addition, [42]. Before the experiment, the rats were acclimated to the sucrose
running exercise has been shown to decrease the numbers of solution (1%, w/v). Briefly, two bottles of sucrose solution were placed in
microglia in the hippocampi of high-fat diet (HFD)-fed model rats
and aged mice [38, 39]. Kohman et al. found that exercise
contributed to reducing the activation of microglia isolated from
the hippocampi of aged mice [40]. However, the effects of running
exercise on the numbers and activation of hippocampal microglia
and the accompanying inflammatory response in depression still
need to be explored. Fig. 1 Timeline of the experimental procedures. The experiment
Therefore, we hypothesized that running exercise attenuates lasted a total of 13 weeks. The animals were allowed to habituate to
depressive symptoms by exerting protective effects against the housing conditions for 2 weeks before any interventions were
initiated. The CUS model rats were exposed to two stressors per day
microglial activity and inflammatory cytokines in CUS models. To for 5 weeks. Then the CUS + RUN group rats were subjected to
test this hypothesis, we verified the antidepressant effects of treadmill running exercise for 6 weeks. Sucrose preference and body
running exercise with a sucrose preference test (SPT). Then we weight were assessed during a fixed time frame per week. CUS
investigated whether running exercise affects these changes using chronic unpredictable stress, CUS + RUN CUS + running, SPT
immunohistochemistry and unbiased stereological methods. sucrose preference test, BWT body weight test.

Translational Psychiatry (2021)11:461

 K. Xiao et al.
3
Sunday each cage for 24 h and then replaced with two bottles filled with sucrose
solution and water. After the acclimation period, the rats were allowed

BWT

BWT

BWT

BWT

BWT
SPT

SPT

SPT
SPT

SPT
access to one bottle of sucrose solution and one bottle of water.
After 24 h, the weights of the consumed sucrose solution and water were
recorded. The percentage of sucrose preference was calculated as sucrose

Continuous darkness during

consumption/(sucrose consumption + water consumption) × 100% [43].
The SPT was assessed during a fixed time frame each week.
Intermittent illumination

Intermittent illumination

Intermittent illumination
Intermittent illumination
Food deprivation

Perfusion and tissue processing

In the following procedures, all experiments and data analyses were

Cage shaking
carried out by investigators blinded to the treatment conditions. After

Cold stress
the behavioral testing, five rats were randomly selected from each group.
Saturday

Restraint

Restraint
daytime

The rats were anesthetized with 1% pentobarbital sodium (4 mL/kg,

intraperitoneal injection) and transcardially perfused with heparinized
saline followed by 4% paraformaldehyde in phosphate-buffered saline
(PBS, 0.6 M, pH 7.4). Then the cerebrum of each rat was removed and cut
into two hemispheres along the sagittal suture. The brain tissue was
postfixed in 4% paraformaldehyde for 24 h and subsequently immersed in
Food deprivation

Damp bedding a series of sucrose solutions of increasing concentration. The right or left
hemisphere was chosen randomly and cut into 50 μm serial sections using
Tail pinching

a cryostat (CM1860, Leica, Germany). Every sixth section containing the

Intermittent
illumination
Strobe light

Strobe light
Cage tilting
Cold stress

hippocampus was sampled in a systematic random manner, and 12 sets of

sampled sections were acquired for the following analyses.
Friday
Noise

Noise

Immunohistochemistry and stereological analyses

Two separate sets of serial sections containing the hippocampus were
Continuous darkness during

Continuous darkness during

chosen from each group of rats. The free-floating sections were rinsed in
PBS with 0.3% Triton X-100 (Sigma, USA) and 0.1% Tween (PBS + T) 6 times
Empty bottle exposure

for 10 min each. Then, the sections were blocked with 1% fetal bovine
Continuous lighting

Continuous lighting

serum (FBS), 10% SP 9001-A, 0.1% cold water fish gelatin, and PBS + T for
Bedding removed

2 h at 37 °C. The sections were then incubated with a rabbit anti-Iba1

primary antibody (1:2000, ab178847, Abcam, Cambridge, UK) in PBS + T for
Cage shaking
Cage tilting

2 days at 4 °C. After being rewarmed and rinsed, the sections were
Thursday

Restraint

Restraint
daytime

daytime

incubated with SP 9001-B (biotinylated goat anti-rabbit immunoglobulin,

1:20) for 3 h at 37 °C and followed by SP 9001-C (biotin-HRP-streptavidin)
for 2 h at 37 °C. Subsequently, the sections were immersed in a
diaminobenzidine solution (ZLL-9032, ZSGB, Beijing, P. R. China) for
approximately 1 min. Finally, the sections were dehydrated by sequential
Electric foot shock

Electric foot shock

Bedding removed
Water deprivation

Food deprivation

immersion in a graded ethanol series (75, 95, and 100%; 5 min each) and
xylene (3 × 10 min).
Tail pinching

Tail pinching
Wednesday
Strobe light

Strobe light

All quantitative analyses were performed as previously described

Cage tilting

[41, 44]. The hippocampal fields were delineated by cytoarchitectonic

criteria. According to the morphological features of the neurons in the
hippocampus, the contours of the CA1, CA2/3 and DG were delineated at a
×2.5 magnification (Supplementary Fig. 1A). The numbers of Iba1+ cells in
the three subfields of the hippocampus were estimated using the optical
Continuous darkness during

fractionator method in the stereological system (Glostrup, Denmark). The

total number of Iba1+ cells in each hippocampal subregion was calculated
Empty bottle exposure

using the following formula:

X
Continuous lighting

1 1 1
Q ´
Electric foot shock

N¼ ´ ´
Water deprivation

Water deprivation

Water deprivation

ssf asf hsf

Cage shaking

where the variable N is the total number of Iba1+ cells in the CA1, CA2/3,
and DG in each animal, ∑Q− is the total number of Iba1+ cells actually
Hot stress

Hot stress
Tuesday

daytime

counted in the specimens, ssf is the section sampling fraction, asf is the
SPT sucrose preference test, BWT body weight test.

area sampling fraction, and hsf is the section height sampling fraction
(for the stereological sampling scheme, see Supplementary Fig. 1B, C and
Table 2). In the present study, the ssf was 1/6 according to the above
Schedule of the CUS paradigm.

Continuous darkness during

section sampling method. The optical disector counting frames were

placed in the delineated area of the sections in a systematically random
manner. The asf, the ratio between the area of the unbiased counting
frame and the rectangular area, was set to 6%. The guard zone was set at a
Continuous lighting

thickness of 3 μm, and microglia were counted at a depth of 15 μm below

Water deprivation
Food deprivation
Damp bedding

Damp bedding

the guard zone. Hsf represents the ratio between the counted height and
the mean thickness of each section.
Cage tilting

Cage tilting

Restraint
Monday

daytime

Coimmunostaining and cell counting

Noise

One set of serial sections containing the hippocampus was randomly

chosen from each group of rats. Free-floating sections were rinsed with
PBS + T 6 times for 10 min each. Then the sections were blocked with 1%
Week 1

Week 2

Week 3

Week 4

Week 5
Table 1.

FBS, 10% SP 9001-A, and PBS + T for 2 h at 37 °C. The sections were then
Time

incubated with a rabbit anti-Iba1 primary antibody (1:500, ab178847,

Abcam, Cambridge, UK) and a mouse anti-CD68 primary antibody (1:100,

Translational Psychiatry (2021)11:461

 K. Xiao et al.
4
Table 2. Sampling scheme for the estimation of the number of Iba1+ cell.
Control group CUS + STD group CUS + RUN group
Number of sections sampled
CA1 13–17 13–18 13–18
CA2/3 12–17 12–17 13–18
DG 14–20 14–19 14–19
Number of counting frames sampled
CA1 260 (244–298) 343 (310–378) 288 (230–334)
CA2/3 214 (176–273) 269 (237–293) 238 (193–309)
DG 262 (204–345) 306 (271–352) 301 (235–350)
Number of Iba1+ cells counted
CA1 391 (274–462) 650 (459–804) 436 (253–548)
CA2/3 317 (197–517) 520 (368–747) 359 (197–514)
DG 439 (263–598) 637 (471–818) 502 (301–702)
Note: The number of sampled sections is presented as the range, whereas the numbers of counting frames and the numbers of sampled Iba1+ cells are
presented as the mean with the range in parentheses.
CUS + STD CUS + standard, CUS + RUN CUS + running.

Table 3. Primer sequences for qRT-PCR.

Gene Forward primer (5′– 3′) Reverse primer (5′– 3′)
IL-1β TGAGGCTGACAGACCCCAAAAGAT GCTCCACGGGCAAGACATAGGTAG
iNOS GAGACGCACAGGCAGAGGTTG AGCAGGCACACGCAATGATGG
IL-6 ACTTCCAGCCAGTTGCCTTCTTG TGGTCTGTTGTGGGTGGTATCCTC
TNF-α TCGTAGCAAACCACCAAGCG AGAGAACGGATGAACACGCCA
β-Actin TGTCACCAACTGGGACGATA GGGGTGTTGAAGGTCTCAAA
qRT-PCR quantitative real-time polymerase chain reaction.

ab955, Abcam, Cambridge, UK) in PBS + T for 2 days at 4 °C. The sections analysis was used to verify the primer specificity. Finally, the relative gene
were washed three times for 10 min at room temperature in PBS + T. For expression was analyzed by the 2−ΔΔCt method.
the immunofluorescence staining, the sections were incubated with
DyLight 549- and DyLight 488-labeled secondary antibodies (1:200;
Abbkine, USA) for 2 h at 37 °C. 4’,6-Diamidino-2-phenylindole (AR1176, Enzyme-linked immunosorbent assay
Boster, Wuhan, P. R. China) was used for the nuclear staining. Fluorescent The protein expression levels of IL-1β, iNOS, IL-6, and TNF-α in the
images were captured at a ×200 magnification under a Zeiss fluorescence hippocampus were examined by ELISA. The hippocampi of the rats were
microscope (Zeiss, Germany). At least ten representative images of each isolated on ice and washed with normal saline. Then the tissue samples
hippocampal subregion were acquired from each rat. The Iba1+/CD68+ were homogenized and centrifuged at 12,000 rpm for 15 min. The
cells in each image were manually quantified, and the cell density and supernatant was collected, and the total protein concentration was
percentage were calculated and analyzed. The percentage of activated determined using a BCA Protein Assay Kit (Beyotime, P. R. China). Samples
microglia is expressed as the ratio of Iba1+/CD68+ cells to Iba1+ cells. containing equivalent amounts of protein were calculated and subpack-
aged. Subsequently, the protein concentrations of inflammatory cytokines
in the hippocampus were determined using ELISA kits (EIAab, P. R. China)
Quantitative real-time PCR (qRT-PCR) according to the manufacturer’s instructions. The optical density at 450 nm
The levels of the proinflammatory cytokines IL-1β, inducible nitric oxide was determined using a microplate reader (Bio-Rad, Hercules, CA, USA),
synthase (iNOS), IL-6, and tumor necrosis factor (TNF)-α are associated with and the concentration of these cytokines was calculated according to
the severity of depression [7, 45]. To evaluate the mRNA levels of these standard curves.
proinflammatory cytokines in the hippocampi of depression model rats, we
randomly selected five rats from each group for the qRT-PCR analysis. The
brain tissues of the rats were quickly removed on ice after anesthesia and
decapitation. The hippocampal tissues were dissected, frozen in liquid STATISTICS
nitrogen, and subsequently stored at −80 °C. The statistical analyses were performed using SPSS version 23.0
The total RNA was extracted from the hippocampi of the rats using a (IBM, Armonk, NY, USA). All data are expressed as the mean ±
Total RNA Extraction Kit (LS1040, Promega, Shanghai, P. R. China) standard deviation (SD). All data were assessed for normality and
according to the manufacturer’s protocol. Complementary DNA (cDNA) homogeneity of variance. The body weight data were analyzed
synthesis was performed using a PrimeScript RT Reagent Kit with gDNA using a repeated-measures analysis of variance (ANOVA). Regard-
Eraser (RR047A, Takara, Japan) according to the manufacturer’s instruc- ing the remaining data, one-way ANOVA followed by a least
tions. The primer sequences and sequence specificity of the primers were significant difference post hoc test was used to compare the
designed and tested, respectively, using the National Center for
Biotechnology Information (NCBI) Primer-BLAST. The primer sequences results among multiple groups and Student’s t tests were used for
are listed in Table 3. PCR amplification of cDNA was performed using 2× comparisons between two groups. The sample size of each
SYBR Green qPCR Master Mix (B21202, Bimake, Houston, USA). The experiment was chosen based on previous experience with the
formation of PCR products was detected in real time using a CFX96 Real- aim of detecting at least a p value <0.05 in the different tests
Time PCR Detection System (Bio-Rad, Hercules, CA, USA). A melting curve applied. No animals were excluded from the present study.

Translational Psychiatry (2021)11:461

 K. Xiao et al.
5

Fig. 2 The effects of running exercise on the depressive-like behaviors of CUS-exposed rats. A Body weights of the rats in the control
group (n = 23) and the CUS group (n = 37) during the first 7 weeks. B Body weights of the rats in the control group (n = 23), the CUS + STD
group (n = 17) and the CUS + RUN group (n = 20) during the last 7 weeks. C Sucrose preferences of the rats in the control group (n = 23) and
the CUS group (n = 37) at different stages. D Sucrose preferences of the rats in the control group (n = 23), the CUS + STD group (n = 17), and
the CUS + RUN group (n = 20) at different stages. **p < 0.01. CUS + STD CUS + standard, CUS + RUN CUS + running.

RESULTS group had decreased numbers of Iba1+ microglia in the CA1, CA2/
Running exercise reversed depressive-like behavior in CUS- 3, and DG compared with the CUS + STD group (p = 0.000, p =
exposed rats 0.010, p = 0.047, respectively; Fig. 3B and Table 4). In the present
The effect of running exercise on CUS-induced depressive-like study, sampling met the criteria because the observed variance of
behavior was analyzed by behavioral tests. The body weight of the individual estimate (OCE2) was much <50% of the observed
the CUS group was similar to that of the control group during the interindividual variance (OCV2) (Table 4).
2-week baseline adjustment phase (p > 0.05, Fig. 2A). The body
weight of the CUS-exposed rats was found to be significantly
lower than that of the normal rats from the first week of the CUS Running exercise decreased the density and ratio of activated
intervention (p < 0.01, Fig. 2A). During the running exercise period, microglia in subfields of the hippocampus in the CUS-exposed
the rats in the CUS + STD group and CUS + RUN group rats
persistently showed distinctly lower body weights than the rats The activation status of microglia was observed using Iba1 and
in the control group (p < 0.01, p < 0.01, respectively; Fig. 2B). In CD68 immunofluorescence staining. CD68 is a transmembrane
addition, the sucrose consumption of the normal rats was similar glycoprotein that is highly expressed in activated and phagocytic
to that of the CUS-exposed rats during the baseline period (t = microglia. Red staining indicated Iba1+ microglia, and green
−0.170, p = 0.866, Fig. 2C). However, after the 5-week CUS staining indicated CD68 (Fig. 4A, B). There were significant
intervention, the sucrose preference in the CUS group was differences in the densities of Iba1+/CD68 + microglia in the CA1,
significantly lower than that in the control group (t = −4.990, CA2/3, and DG (F(2,12) = 6.482, p = 0.012; F(2,12) = 6.349, p =
p = 0.000, Fig. 2C). After 6 weeks of running exercise, there were 0.013; F(2,12) = 18.676, p = 0.000) and the percentages of
significant differences in sucrose preference among the three Iba1+/CD68+ microglia in Iba1+ cells in the CA1, CA2/3, and
groups (F(2,57) = 28.003, p = 0.000). The rats in the CUS + RUN DG (F(2,12) = 5.851, p = 0.017; F(2,12) = 5.935, p = 0.016; F(2,12) =
group showed a significantly higher sucrose preference than the 15.876, p = 0.000) among the three groups. Specifically, relative
rats in the CUS + STD group (p = 0.000, Fig. 2D). to the rats in the control group, the densities of Iba1+/CD68+
microglia and percentages of Iba1+/CD68+ microglia in Iba1+
cells in the CA1, CA2/3, and DG areas were elevated in the CUS +
Running exercise decreased the numbers of microglia in STD group rats (p = 0.008, p = 0.005, and p = 0.000 for the
subfields of the hippocampus in the CUS-exposed rats densities of Iba1+/CD68 + cells; p = 0.014, p = 0.006, and p =
The number of Iba1+ microglia in the hippocampus was observed 0.000 for the percentages of Iba1+/CD68+ cells), indicating that
and quantified using a stereological method combined with CUS promoted microglial activation. In addition, the density of
immunohistochemistry. Representative pictures of the immuno- Iba1+/CD68+ microglia and percentage of Iba1+/CD68+ micro-
histochemical staining using an anti-Iba1 antibody are shown in glia in Iba1+ cells in the DG area of the hippocampus in the
Fig. 3A. Iba1 is a protein specifically expressed in microglia in the CUS + RUN group were found to be decreased (p = 0.003 and
CNS. There were significant differences in Iba1+ microglia in the p = 0.007, respectively). However, the densities of Iba1+/CD68+
CA1, CA2/3, and DG among the three groups (F(2,12) = 24.182, p = microglia and percentages of Iba1+/CD68+ microglia in Iba1+
0.000; F(2,12) = 10.510, p = 0.002; F(2,12) = 7.899, p = 0.006). Speci- cells in CA1 and CA2/3 of the hippocampus in the CUS + RUN
fically, there were markedly more Iba1+ microglia in the CA1, CA2/ group did not significantly differ from those in the CUS + STD
3, and DG in the CUS + STD group of rats than in the control group (p = 0.857 and p = 0.36 for the density of Iba1+/CD68+
group of rats (p = 0.000, p = 0.001, p = 0.002, respectively). cells; p = 0.863 and p = 0.463 for the percentage of Iba1+/CD68+
Moreover, after 6 weeks of running exercise, the CUS + RUN cells) (Fig. 4C, D and Table 5).

Translational Psychiatry (2021)11:461

 K. Xiao et al.
6

Fig. 3 The effects of running exercise on the numbers of microglia in subfields of the hippocampus in CUS-exposed rats. A Representative
images of immunohistochemical staining of Iba1 in each subregion of the hippocampus (×100 oil). Scale bar = 30 μm. B Quantitative analyses
comparing the numbers of Iba1+ cells in the CA1, CA2/3, and DG areas among the control group, the CUS + STD group, and the CUS + RUN
group. The data are expressed as the mean ± SD (n = 5 per group). *p < 0.05 and **p < 0.01. CUS + STD CUS + standard, CUS + RUN CUS +
running.

Running exercise reversed the increases in the mRNA levels of in the hippocampi of the rats in the CUS + STD group (p = 0.048;
inflammatory cytokines in the hippocampus of the CUS- Fig. 5A). There were no significant differences in the mRNA levels
exposed rats of IL-6 and TNF-α in the hippocampus among the three groups
The changes in inflammatory cytokines (IL-1β, iNOS, IL-6, and (F(2,12) = 0.810, p = 0.468; F(2,12) = 0.497, p = 0.621; Fig. 5C, D).
TNF-α) at the genetic level were detected by qRT-PCR. As shown in
Fig. 5, there were significant differences in the mRNA levels of Running exercise reversed the increases in the protein levels
IL-1β and iNOS in the hippocampus among the three groups of inflammatory cytokines in the hippocampus of the CUS-
(F(2,12) = 7.940, p = 0.006; F(2,12) = 4.248, p = 0.040). Specifically, exposed rats
compared to the control group, the CUS exposure significantly The protein levels of inflammatory cytokines (IL-1β, iNOS, IL-6, and
increased the mRNA levels of IL-1β and iNOS in the hippocampus TNF-α) were measured by ELISA. There were significant differences
(p = 0.002 and p = 0.016, respectively; Fig. 5A, B). In addition, in the protein levels of IL-1β and IL-6 in the hippocampus among
running exercise significantly decreased the mRNA levels of IL-1β the three groups (F(2,12) = 6.725, p = 0.011; F(2,12) = 10.003,

Translational Psychiatry (2021)11:461

 K. Xiao et al.
7
+
Table 4. Stereological results of the total numbers of Iba1 cells in the hippocampus.
Control group CUS + STD group CUS + RUN group
CA1
Number of Iba1+ cells (×105) 1.22 ± 0.18 2.04 ± 0.14** 1.38 ± 0.26##
−3
OCV (×10 ) 144.97 68.14 185.10
OCE (×10−3) 2.89 2.22 2.73
CA2/3
Number of Iba1+ cells (×105) 0.97 ± 0.22 1.63 ± 0.26** 1.18 ± 0.22##
−3
OCV (×10 ) 231.96 157.34 189.71
OCE (×10−3) 3.28 2.50 2.97
DG
Number of Iba1+ cells (×105) 1.32 ± 0.22 1.95 ± 0.23** 1.60 ± 0.29#
−3
OCV (×10 ) 167.96 115.85 184.56
OCE (×10−3) 2.79 2.28 2.54
Note: The numbers of Iba1+ cells are represented as the mean ± SD.
CUS + STD CUS + standard, CUS + RUN CUS + running, OCE observed coefficient of error, OCV observed coefficient of variation.
**p < 0.01 when the CUS + STD group was compared with the control group. #p < 0.05 when the CUS + RUN group was compared with the CUS + STD group.
##
p < 0.01 when the CUS + RUN group was compared with the CUS + STD group.

p = 0.003; Fig. 5E, G). Specifically, the protein levels of IL-1β and IL- methods used. In the studies by Cai et al. and Tong et al., the
6 were enhanced in the CUS + STD group compared with those in density of Iba1+ microglia in the DG of the hippocampus was
the control group (p = 0.043 and p = 0.035, respectively). Com- evaluated via a semiquantitative measurement of Iba1+ material;
pared with the CUS + STD group, the protein levels of IL-1β and IL- however, the microglial density may not directly reflect the total
6 were significantly diminished in the hippocampi of the rats in number of microglia. In our study, an unbiased stereological
the CUS + RUN group (p = 0.003 and p = 0.001, respectively). In technique, a three-dimensional quantitative method, was used to
addition, the protein levels of iNOS in the CUS + STD group were accurately quantify microglia in the hippocampus; thus, we
significantly increased compared to those in the control group provided more precise results. The second possible reason for
(p = 0.040; Fig. 5F). However, the protein levels of TNF-α in the the different observed effects of CUS on the number of microglia
hippocampus did not significantly differ among the three groups might be the types and duration of stressors. Cai et al. and Tong
(F(2,12) = 0.136, p = 0.875; Fig. 5H). et al. established the CUS model by adopting ten different types of
stressors. In our study, more types of stressors, including food
deprivation, empty bottle exposure, no bedding exposure, hot
DISCUSSION stress, electric foot shock exposure, and tail pinching, were added
In the present study, CUS was used to induce depressive-like to the CUS paradigm. The duration of each stress was generally
behavior in rats. The CUS model can simulate the various stresses different. We hypothesize that the heterogeneity of these stressors
that humans encounter in life, causing rodents to exhibit the core may be another important reason for the different changes in the
symptoms of depression, such as anhedonia [46]. The SPT was number of microglia in animal models of depression. The third
performed to evaluate the degree of anhedonia of the CUS model possible reason for the different observed effects of CUS on the
rats. Consistent with a previous study [47], we found that 5 weeks number of microglia might be the different animal species
of the CUS intervention reduced body weight and sucrose studied. In the present study, which used a rat model, we found
preference in the rats, indicating that the rat depression model that the number of Iba1+ microglia was increased in the
was successfully established. hippocampi of the depression model rats. This result is consistent
Increasing evidence suggests that alterations in the number of with the findings of previous research in which a rat model was
microglia are associated with the onset of depression, and the adopted [17]. Cai et al. and Tong et al. evaluated the changes in
change in the number of total microglia helps determine whether microglia using mouse models and found a decreased microglial
microglia are in the activated or senescent phase [26, 48]. Iba1 is density in the hippocampus. We hypothesize that hippocampal
constitutively expressed in all microglia, including resting ramified microglia in rats are more susceptible to CUS exposure than those
microglia and activated microglia [49]. Thus, an analysis of Iba1+ in mice. In addition, we systemically measured the number of
microglia along with an assessment of depressive symptoms microglia in the hippocampal CA1, CA2/3, and DG. We found that
might aid in the prediction of the progression and treatment CUS significantly increased the total number of Iba1+ microglia in
effects of depression. The hippocampus is a stress-sensitive brain each subregion of the hippocampus in the depression model rats
area that has been implicated in the pathophysiology and and that the trends in the variation in the number of microglia
progression of depression [50–53]. In the current study, our across the three regions of the hippocampus after the CUS
results show that the total number of Iba1+ microglia in the intervention were consistent. Therefore, we speculate that the
hippocampus in the depressed model rats was significantly higher changes in the total number of microglia (including activated and
than that in the control rats, suggesting that the total number of resting microglia) induced by CUS are consistent across the three
microglia increases in response to CUS exposure. In contrast, Cai subregions of the hippocampus. Collectively, our present results
et al. and Tong et al. examined the change in the number of suggest that microglial number changes in the hippocampus may
microglia in a defined area of the hippocampus in CUS-exposed be involved in the pathogenesis of the depression model, and we
mice by an immunofluorescence assay and found that the density are the first to provide evidence that CUS affects the number of
of microglia was decreased in the hippocampi of depressed mice microglia in each subregion of the hippocampus. Although most
[54, 55]. The first possible reason for the different effects of CUS on previous studies and the present study found microglial changes
the number of microglia might be the differential quantitative in the hippocampi of depression models [19, 22, 56, 57], changes

Translational Psychiatry (2021)11:461

 K. Xiao et al.
8

Fig. 4 The effects of running exercise on the density and ratio of activated microglia in subfields of the hippocampus in CUS-exposed
rats. A Representative confocal images of Iba1+/CD68+ microglia in the hippocampus. Scale bar = 30 μm. B Representative confocal images of
immunofluorescence staining using antibodies against Iba1 and CD68 in the hippocampal subregions in the control group, the CUS + STD
group, and the CUS + RUN group. The arrowheads indicate representative Iba1+/CD68+ cells. Scale bar = 50 μm. C, D Quantitative analyses
comparing the densities and percentages of Iba1+/CD68+ cells in the CA1, CA2/3, and DG areas among the control group, the CUS + STD
group, and the CUS + RUN group. The data are expressed as mean ± SD (n = 5 per group). *p < 0.05 and **p < 0.01. CUS + STD CUS + standard,
CUS + RUN CUS + running.

in microglia in other brain regions cannot be ruled out, which demonstrated that chronic stress increases the density of
merits further study. activated/phagocytic microglia in the DG of the hippocampus
In addition to the changes in the total number of microglia, [64]. However, a systematic study on the change in activated
microglial functional alterations, including microglial activation microglia in each subregion of the hippocampus has not been
and concomitant elevations in pro-inflammatory biomarker levels, reported. In the present study, we assessed microglial activation
play crucial roles in the deterioration of depression [58, 59]. mainly by analyzing the number of microglia labeled by Iba1 and
Changes in microglial activation state have been shown to be CD68 per unit area and found that CUS increased the density of
sensitive markers in the brains of depressed suicides [18]. Iba1+/CD68+ cells and the percentage of Iba1+/CD68+ cells
Furthermore, numerous studies have reported that the levels of among Iba1+ cells in the CA1, CA2/3, and DG of the hippocampus,
CD68, an ideal marker of activated and phagocytic microglia [60], revealing that chronic stress induced microglial activation in the
are elevated in depression models [61–63]. Another study hippocampus. Microglia can be activated during inflammatory

Translational Psychiatry (2021)11:461

 K. Xiao et al.
9
+ +
Table 5. Density and percentage of Iba1 /CD68 cells.
Control group CUS + STD group CUS + RUN group
Density of Iba1+/CD68 + cells (/mm2)
CA1 3.30 ± 1.12 5.70 ± 1.21** 5.56 ± 1.23
CA2/3 3.71 ± 1.28 6.89 ± 1.90** 6.01 ± 1.05
DG 4.83 ± 1.37 9.69 ± 1.14** 6.69 ± 1.29##
+ +
Percentage of Iba1 /CD68 cells (%)
CA1 3.37 ± 1.18 5.49 ± 1.08* 5.62 ± 1.23
CA2/3 3.75 ± 1.26 6.55 ± 1.74** 5.90 ± 0.90
DG 4.75 ± 1.27 8.80 ± 0.91** 6.47 ± 1.21##
+
Note: The density and percentage of Iba1 cells are presented as mean ± SD.
CUS + STD CUS + standard, CUS + RUN CUS + running.
*p < 0.05 when the CUS + STD group was compared with the control group. **p < 0.01 when the CUS + STD group was compared with the control group.
##
p < 0.01 when the CUS + RUN group was compared with the CUS + STD group.

conditions that cause deviations from microglial homeostasis, Parkinson’s disease [70, 71]. However, whether running exercise
such as stress, stroke, infection, and neurodegenerative diseases has positive effects on hippocampal microglia in depression
[26]. In the current study, CUS exposure might be the reason for model rats has remained unclear. Using an unbiased stereological
the increased microglial activation. However, many stress-related method, we found that running exercise could reverse the CUS-
factors, such as stress hormones, neurotransmitters, and pattern induced increases in Iba1+ microglial populations in the
recognition receptor agonists, can greatly affect the activation of hippocampal CA1, CA2/3, and DG subregions in depression model
microglia [65]. Thus, the specific mechanisms underlying stress- rats. The current study is the first to provide unbiased and
associated microglial activation need to be further studied. The accurate quantitative evidence suggesting that the total number
current study investigated microglial activation in each subfield of of microglia in the hippocampus is altered in CUS model rats after
the hippocampus in depression model rats from a functional running exercise. More importantly, our results indicate that
perspective, and the results indicate that activated microglia in the hippocampal microglia might be structural targets for the
hippocampus are involved in the pathogenesis of depression. ameliorative effects of running exercise on depression.
Furthermore, we found that CUS induced elevated mRNA and As mentioned above, microglial activation and the levels of
protein levels of the proinflammatory cytokines IL-1β and iNOS in inflammatory factors can better reflect the effects of microglia on
the hippocampus, which is consistent with a report by Liu et al. the neuroinflammatory response than the total number of
[45]. As IL-1β and iNOS released by microglia help amplify microglia. In the present study, we found that running exercise
neuroinflammation in depression and subsequently lead to attenuated the increases in the number of Iba1+ microglia in
neurotoxicity and pathological changes [66, 67], this finding the CA1, CA2/3, and DG subregions of the hippocampus in the
suggests that IL-1β and iNOS in the hippocampus may play an depression model rats but attenuated the increase in the density
important role in the processes of microglial activation. However, of Iba1+/CD68+ microglia only in the DG region of the
no significant changes in the hippocampal gene expression of IL-6 hippocampus. Impaired functional activity in the hippocampal
were found in the depressed model rats. In contrast, the IL-6 CA1, CA2/3, and DG subregions has been found to be associated
protein level in the hippocampi of the CUS + STD group rats was with depressive symptoms [72], and the subregions of the
significantly higher than that in the hippocampi of the control hippocampus interact to modulate depressive emotions. The DG
group rats. Gene expression is a complex process involving region acts as a gateway that processes sensory inputs from the
transcription, RNA splicing, translation, and posttranslational environment and integrates these inputs into the hippocampus
protein modification. The different results of the gene expression to moderate neuroplasticity [73]. A study by Wohleb et al.
and protein level of IL-6 could be due to improved gene demonstrated that microglia mediate neuronal remodeling and
translation efficiency, enhanced posttranslational modifications, synaptic deficits in depression model mice, indicating that
or mRNA instability, which ultimately affects the protein level, but neuronal plasticity is closely related to microglia during the
the specific mechanism remains unclear and requires further development of depression [74]. In addition, treadmill exercise
study. Collectively, the evidence indicates that microglial activa- has been shown to have neuroprotective effects in the DG area of
tion and elevations in inflammatory cytokines may contribute to the hippocampus by inhibiting HFD-induced microglial activation
the development of depression. and the expression of proinflammatory cytokines [75]. Intrigu-
Running exercise is recognized as an effective therapeutic ingly, running exercise can regulate hippocampal synaptic
strategy for depression [68]. Consistent with the results of a plasticity and neurogenesis by selectively enhancing long-term
previous study [69], our behavioral results demonstrate that potentiation in the DG region instead of the CA1 [76]. Thus, we
6 weeks of running exercise increased the sucrose preference of speculate that activated microglia in the DG area are more likely
the depression model rats, indicating that running exercise can to respond to running exercise treatment than those in other
mitigate the depressive-like behaviors induced by CUS. Therefore, hippocampal subregions in the context of depression. Moreover,
these results confirm that running exercise has a therapeutic as the sensitivity of the microglial state might vary across
effect against depression. Zheng et al. showed that the inhibition different hippocampal regions, we speculate that CUS-induced
of microglial activation helps attenuate IFN-α-induced neuroin- activated/phagocytic microglia in the CA1 and CA2/3 of the
flammation and depressive-like behaviors in mice [35], which may hippocampus might not transition to the quiescent state or might
be associated with important antidepressant mechanisms. Several not transition as quickly as microglia in the DG region after
studies in related fields clearly demonstrated that treadmill running exercise. The regional specificity of running exercise in
exercise attenuates the increase in microglia in the hippocampal alleviating microglial activation and neuroinflammation needs to
CA1 region and cerebellar vermis in subjects with neurodegen- be further investigated in future studies. In addition, it can be
erative disorders, such as sporadic Alzheimer’s disease and concluded that physical exercise reduces the detrimental effects

Translational Psychiatry (2021)11:461

 K. Xiao et al.
10

Fig. 5 The effects of running exercise on the mRNA and protein levels of proinflammatory cytokines in the hippocampus in CUS-exposed
rats. A–D Quantitation of the relative mRNA expression of the proinflammatory cytokines IL-1β, iNOS, IL-6, and TNF-α in the hippocampi of the
control group, the CUS + STD group, and the CUS + RUN group. The data are presented as mean ± SD (n = 5 per group). E–H Quantitation of
the protein levels of the proinflammatory cytokines IL-1β, iNOS, IL-6, and TNF-α in the hippocampi of the three groups. The data are presented
as mean ± SD (n = 5 per group). *p < 0.05 and **p < 0.01. CUS + STD CUS + standard, CUS + RUN CUS + running.

of neuroinflammation in MDD patients and depression model we speculate that running exercise has antidepressant effects by
rodents [77, 78]. Our results show that running exercise reduced suppressing microglial activation and the release of inflammatory
the gene expression and protein level of IL-1β in the hippocam- cytokines.
pus of rats subjected to CUS. Furthermore, running exercise In conclusion, our study shows that the number and activation
reduced the protein level of IL-6 in the hippocampus of the of hippocampal microglia along with accompanying neuroin-
depression model rats. There were no differences in the gene flammation in the hippocampus may be involved in the
expression and protein levels of iNOS and TNF-α after running pathogenesis of depression in model rats. In addition, running
wheel exercise. Finally, our results prove that running exercise has exercise reverses these changes in the context of depression,
protective effects against the elevation of some inflammatory indicating that changes in hippocampal microglia and neuroin-
factors in the hippocampus of CUS model rats. Taken together, flammation might be the structural bases by which running

Translational Psychiatry (2021)11:461

 K. Xiao et al.
11
exercise exerts antidepressant effects. Our findings may shed new 24. Streit WJ, Walter SA, Pennell NA. Reactive microgliosis. Prog Neurobiol.
light on immunomodulatory targets related to the antidepressant 1999;57:563–81.
effects of exercise and provide a scientific basis for the clinical 25. Salter MW, Beggs S. Sublime microglia: expanding roles for the guardians of the
treatment strategy for depression. CNS. Cell. 2014;158:15–24.
26. Yirmiya R, Rimmerman N, Reshef R. Depression as a microglial disease. Trends
Neurosci. 2015;38:637–58.
27. Setiawan E, Wilson AA, Mizrahi R, Rusjan PM, Miler L, Rajkowska G, et al. Role of
REFERENCES
translocator protein density, a marker of neuroinflammation, in the brain during
1. Barnett R. Depression. Lancet. 2019;393:2113. major depressive episodes. JAMA Psychiatry. 2015;72:268–75.
2. Lim GY, Tam WW, Lu Y, Ho CS, Zhang MW, Ho RC. Prevalence of depression in the 28. Dai J, Ding Z, Zhang J, Xu W, Guo Q, Zou W, et al. Minocycline relieves depressive-
community from 30 countries between 1994 and 2014. Sci Rep. 2018;8:2861. like behaviors in rats with bone cancer pain by inhibiting microglia activation in
3. GBD 2017 Disease and Injury Incidence and Prevalence Collaborators. Global, hippocampus. Anesth Analg. 2019;129:1733–41.
regional, and national incidence, prevalence, and years lived with disability for 354 29. Helgadóttir B, Forsell Y, Hallgren M, Möller J, Ekblom Ö. Long-term effects of
diseases and injuries for 195 countries and territories, 1990-2017: a systematic exercise at different intensity levels on depression: a randomized controlled trial.
analysis for the Global Burden of Disease Study 2017. Lancet. 2018;392:1789–858. Prev Med. 2017;105:37–46.
4. Walker ER, McGee RE, Druss BG. Mortality in mental disorders and global disease 30. Bailey AP, Hetrick SE, Rosenbaum S, Purcell R, Parker AG. Treating depression with
burden implications: a systematic review and meta-analysis. JAMA Psychiatry. physical activity in adolescents and young adults: a systematic review and meta-
2015;72:334–41. analysis of randomised controlled trials. Psychol Med. 2018;48:1068–83.
5. Miller AH, Raison CL. The role of inflammation in depression: from evolutionary 31. Xiao Q, Wang F, Luo Y, Chen L, Chao F, Tan C, et al. Exercise protects myelinated
imperative to modern treatment target. Nat Rev Immunol. 2016;16:22–34. fibers of white matter in a rat model of depression. J Comp Neurol.
6. Loftis JM, Huckans M, Morasco BJ. Neuroimmune mechanisms of cytokine- 2018;526:537–49.
induced depression: current theories and novel treatment strategies. Neurobiol 32. Luo Y, Xiao Q, Wang J, Jiang L, Hu M, Jiang Y, et al. Running exercise protects
Dis. 2010;37:519–33. oligodendrocytes in the medial prefrontal cortex in chronic unpredictable stress
7. Miller AH, Maletic V, Raison CL. Inflammation and its discontents: the role of cyto- rat model. Transl Psychiatry. 2019;9:322.
kines in the pathophysiology of major depression. Biol Psychiatry. 2009;65:732–41. 33. Chen L-M, Zhang AP, Wang FF, Tan CX, Gao Y, Huang CX, et al. Running exercise
8. Hiles SA, Baker AL, de Malmanche T, Attia J. A meta-analysis of differences in IL-6 protects the capillaries in white matter in a rat model of depression. J Comp
and IL-10 between people with and without depression: exploring the causes of Neurol. 2016;524:3577–86.
heterogeneity. Brain Behav Immun. 2012;26:1180–8. 34. Henry CJ, Huang Y, Wynne A, Hanke M, Himler J, Bailey MT, et al. Minocycline
9. You Z, Luo C, Zhang W, Chen Y, He J, Zhao Q, et al. Pro- and anti-inflammatory attenuates lipopolysaccharide (LPS)-induced neuroinflammation, sickness beha-
cytokines expression in rat’s brain and spleen exposed to chronic mild stress: vior, and anhedonia. J Neuroinflammation. 2008;5:15.
involvement in depression. Behav Brain Res. 2011;225:135–41. 35. Zheng L-S, Kaneko N, Sawamoto K. Minocycline treatment ameliorates interferon-
10. Cameron HA, Glover LR. Adult neurogenesis: beyond learning and memory. Annu alpha- induced neurogenic defects and depression-like behaviors in mice. Front
Rev Psychol. 2015;66:53–81. Cell Neurosci. 2015;9:5.
11. Mahajan GJ, Vallender EJ, Garrett MR, Challagundla L, Overholser JC, Jurjus G, et al. 36. Paolucci EM, Loukov D, Bowdish DME, Heisz JJ. Exercise reduces depression and
Altered neuro-inflammatory gene expression in hippocampus in major depressive inflammation but intensity matters. Biol Psychol. 2018;133:79–84.
disorder. Prog Neuropsychopharmacol Biol Psychiatry. 2018;82:177–86. 37. Liu W, Sheng H, Xu Y, Liu Y, Lu J, Ni X. Swimming exercise ameliorates depression-
12. Savitz J, Frank MB, Victor T, Bebak M, Marino JH, Bellgowan PS, et al. Inflammation like behavior in chronically stressed rats: relevant to proinflammatory cytokines
and neurological disease-related genes are differentially expressed in depressed and IDO activation. Behav Brain Res. 2013;242:110–6.
patients with mood disorders and correlate with morphometric and functional 38. Kang E-B, Koo J-H, Jang Y-C, Yang C-H, Y Lee Y, Cosio-Lima LM, et al. Neuropro-
imaging abnormalities. Brain Behav Immun. 2013;31:161–71. tective effects of endurance exercise against high-fat diet-induced hippocampal
13. Goshen I, Kreisel T, Ben-Menachem-Zidon O, Licht T, Weidenfeld J, Ben-Hur T, neuroinflammation. J Neuroendocrinol. 2016. https://doi.org/10.1111/jne.12385.
et al. Brain interleukin-1 mediates chronic stress-induced depression in mice via 39. He XF, Liu DX, Zhang Q, Liang FY, Dai GY, Zeng JS, et al. Voluntary exercise
adrenocortical activation and hippocampal neurogenesis suppression. Mol Psy- promotes glymphatic clearance of amyloid beta and reduces the activation of
chiatry. 2008;13:717–28. astrocytes and microglia in aged mice. Front Mol Neurosci. 2017;10:144.
14. Tagliari B, Tagliari AP, Schmitz F, da Cunha AA, Dalmaz C, Wyse AT. Chronic 40. Kohman RA, Bhattacharya TK, Wojcik E, Rhodes JS. Exercise reduces activation of
variable stress alters inflammatory and cholinergic parameters in hippocampus of microglia isolated from hippocampus and brain of aged mice. J Neuroin-
rats. Neurochem Res. 2011;36:487–93. flammation. 2013;10:114.
15. Wake H, Moorhouse AJ, Miyamoto A, Nabekura J. Microglia: actively surveying 41. Tang J, Liang X, Zhang Y, Chen L, Wang F, Tan C, et al. The effects of running
and shaping neuronal circuit structure and function. Trends Neurosci. exercise on oligodendrocytes in the hippocampus of rats with depression
2013;36:209–17. induced by chronic unpredictable stress. Brain Res Bull. 2019. https://doi.org/
16. Calcia MA, Bonsall DR, Bloomfield PS, Selvaraj S, Barichello T, Howes OD. Stress 10.1016/j.brainresbull.2019.04.001.
and neuroinflammation: a systematic review of the effects of stress on microglia 42. Borbély É, Hajna Z, Nabi L, Scheich B, Tékus V, László K, et al. Hemokinin-1
and the implications for mental illness. Psychopharmacology. 2016;233:1637–50. mediates anxiolytic and anti-depressant-like actions in mice. Brain Behav Immun.
17. Tynan RJ, Naicker S, Hinwood M, Nalivaiko E, Buller KM, Pow DV, et al. Chronic 2017;59:219–32.
stress alters the density and morphology of microglia in a subset of stress- 43. Liu MY, Yin CY, Zhu LJ, Zhu XH, Xu C, Luo CX, et al. Sucrose preference test for
responsive brain regions. Brain Behav Immun. 2010;24:1058–68. measurement of stress-induced anhedonia in mice. Nat Protoc. 2018;13:1686–98.
18. Torres-Platas SG, Cruceanu C, Chen GG, Turecki G, Mechawar N. Evidence for 44. Liang X, Tang J, Chao FL, Zhang Y, Chen LM, Wang FF, et al. Exercise improves
increased microglial priming and macrophage recruitment in the dorsal anterior depressive symptoms by increasing the number of excitatory synapses in the
cingulate white matter of depressed suicides. Brain Behav Immun. 2014;42:50–9. hippocampus of CUS-induced depression model rats. Behav Brain Res.
19. Zhao Q, Wu X, Yan S, Xie X, Fan Y, Zhang J, et al. The antidepressant-like effects of 2019;374:112115.
pioglitazone in a chronic mild stress mouse model are associated with PPARγ- 45. Liu LL, Li JM, Su WJ, Wang B, Jiang CL. Sex differences in depressive-like beha-
mediated alteration of microglial activation phenotypes. J Neuroinflammation. viour may relate to imbalance of microglia activation in the hippocampus. Brain
2016;13:259. Behav Immun. 2019;81:188–97.
20. Wang X, Zhang D, Lu XY. Dentate gyrus-CA3 glutamate release/NMDA trans- 46. Hammen C. Stress and depression. Ann Rev Clin Psychol. 2005;1:293–319.
mission mediates behavioral despair and antidepressant-like responses to leptin. 47. Yan W, Liu JF, Han Y, Zhang W, Luo YX, Xue YX, et al. Protein kinase Mζ in medial
Mol Psychiatry. 2015;20:509–19. prefrontal cortex mediates depressive-like behavior and antidepressant response.
21. Ballmaier M, Narr KL, Toga AW, Elderkin-Thompson V, Thompson PM, Hamilton L, Mol Psychiatry. 2018;23:1878–91.
et al. Hippocampal morphology and distinguishing late-onset from early-onset 48. Singhal G, Baune BT. Microglia: an interface between the loss of neuroplasticity
elderly depression. Am J Psychiatry. 2008;165:229–37. and depression. Front Cell Neurosci. 2017;11:270.
22. Dong SQ, Zhang QP, Zhu JX, Chen M, Li CF, Liu Q, et al. Gypenosides reverses 49. Ito D, Tanaka K, Suzuki S, Dembo T, Fukuuchi Y. Enhanced expression of Iba1,
depressive behavior via inhibiting hippocampal neuroinflammation. Biomed ionized calcium-binding adapter molecule 1, after transient focal cerebral
Pharmacother. 2018;106:1153–60. ischemia in rat brain. Stroke. 2001;32:1208–15.
23. Kreisel T, Frank MG, Licht T, Reshef R, Ben-Menachem-Zidon O, Baratta MV, et al. 50. Thomas RM, Hotsenpiller G, Peterson DA. Acute psychosocial stress reduces cell
Dynamic microglial alterations underlie stress-induced depressive-like behavior survival in adult hippocampal neurogenesis without altering proliferation.
and suppressed neurogenesis. Mol Psychiatry. 2014;19:699–709. J Neurosci. 2007;27:2734–43.

Translational Psychiatry (2021)11:461

 K. Xiao et al.
12
51. Treadway MT, Waskom ML, Dillon DG, Holmes AJ, Park M, Chakravarty MM, et al. 74. Wohleb ES, Terwilliger R, Duman CH, Duman RS. Stress-induced neuronal colony
Illness progression, recent stress, and morphometry of hippocampal subfields and stimulating factor 1 provokes microglia-mediated neuronal remodeling and
medial prefrontal cortex in major depression. Biol Psychiatry. 2015;77:285–94. depressive-like behavior. Biol Psychiatry. 2018;83:38–49.
52. McKinnon MC, Yucel K, Nazarov A, MacQueen GM. A meta-analysis examining 75. Han T-K, Leem Y-H, Kim H-S. Treadmill exercise restores high fat diet-induced
clinical predictors of hippocampal volume in patients with major depressive disturbance of hippocampal neurogenesis through β2-adrenergic receptor-
disorder. J Psychiatry Neurosci. 2009;34:41–54. dependent induction of thioredoxin-1 and brain-derived neurotrophic factor.
53. Koolschijn PC, van Haren NE, Lensvelt-Mulders GJ, Hulshoff Pol HE, Kahn RS. Brain Brain Res. 2019;1707:154–63.
volume abnormalities in major depressive disorder: a meta-analysis of magnetic 76. van Praag H, Christie BR, Sejnowski TJ, Gage FH. Running enhances neurogenesis,
resonance imaging studies. Hum Brain Mapp. 2009;30:3719–35. learning, and long-term potentiation in mice. Proc Natl Acad Sci USA.
54. Cai, Z., Ye T, Xu X, Gao M, Zhang Y, Wang D, et al. Antidepressive properties of 1999;96:13427–31.
microglial stimulation in a mouse model of depression induced by chronic 77. Schuch FB, Deslandes AC, Stubbs B, Gosmann NP, Silva CT, Fleck MP. Neuro-
unpredictable stress. Prog Neuropsychopharmacol Biol Psychiatry. 2020;101:109931. biological effects of exercise on major depressive disorder: a systematic review.
55. Tong L, Gong Y, Wang P, Hu W, Wang J, Chen Z, et al. Microglia loss contributes Neurosci Biobehav Rev. 2016;61:1–11.
to the development of major depression induced by different types of chronic 78. Ignácio ZM, da Silva RS, Plissari ME, Quevedo J, Réus GZ. Physical exercise and
stresses. Neurochem Res. 2017;42:2698–711. neuroinflammation in major depressive disorder. Mol Neurobiol. 2019;56:
56. Hellwig S, Brioschi S, Dieni S, Frings L, Masuch A, Blank T, et al. Altered microglia 8323–35.
morphology and higher resilience to stress-induced depression-like behavior in
CX3CR1-deficient mice. Brain Behav Immun. 2016;55:126–37.
57. Wang Y-L, Han QQ, Gong WQ, Pan DH, Wang LZ, Hu W, et al. Microglial activation AUTHOR CONTRIBUTIONS
mediates chronic mild stress-induced depressive- and anxiety-like behavior in K.X. designed the study, conducted the experiments, analyzed the data, and wrote
adult rats. J Neuroinflammation. 2018;15:21. the article. Y.L. conducted the animal experiments and guided the experiments. X.L.
58. Wohleb ES, Hanke ML, Corona AW, Powell ND, Stiner LM, Bailey MT, et al. and J.T. guided the experiments and provided technical assistance. J.W. and Q.X.
β-Adrenergic receptor antagonism prevents anxiety-like behavior and microglial conducted the animal experiments and provided technical assistance. H.Y., Y.Q., Y.L.,
reactivity induced by repeated social defeat. J Neurosci. 2011;31:6277–88. P.Z., and Y.X. provided technical assistance. H.W. and Y.T. designed the study and
59. Wachholz S, Eßlinger M, Plümper J, Manitz MP, Juckel G, Friebe A. Microglia provided comments concerning the article.
activation is associated with IFN-alpha induced depressive-like behavior. Brain
Behav Immun. 2016;55:105–13.
60. Murphy JE, Tedbury PR, Homer-Vanniasinkam S, Walker JH, Ponnambalam S.
Biochemistry and cell biology of mammalian scavenger receptors. Athero- FUNDING INFORMATION
sclerosis. 2005;182:1–15. This study was supported by the National Natural Science Foundation of China (NSFC:
61. Nie X, Kitaoka S, Tanaka K, Segi-Nishida E, Imoto Y, Ogawa A, et al. The innate 81871073, 82001436, 82001435).
immune receptors TLR2/4 mediate repeated social defeat stress-induced social
avoidance through prefrontal microglial activation. Neuron. 2018;99:464.e7–79.e7.
62. Xu N, Tang XH, Pan W, Xie ZM, Zhang GF, Ji MH, et al. Spared nerve injury COMPETING INTERESTS
increases the expression of microglia M1 markers in the prefrontal cortex of rats The authors declare no competing interests.
and provokes depression-like behaviors. Front Neurosci. 2017;11:209.
63. Lehmann ML, Cooper HA, Maric D, Herkenham M. Social defeat induces
depressive-like states and microglial activation without involvement of peripheral ADDITIONAL INFORMATION
macrophages. J Neuroinflammation. 2016;13:224. Supplementary information The online version contains supplementary material
64. Feng X, Fan Y, Chung CY. Mefenamic acid can attenuate depressive symptoms by available at https://doi.org/10.1038/s41398-021-01571-9.
suppressing microglia activation induced upon chronic stress. Brain Res.
2020;1740:146846. Correspondence and requests for materials should be addressed to Hong Wu or
65. Wohleb ES, Franklin T, Iwata M, Duman RS. Integrating neuroimmune systems in Yong Tang.
the neurobiology of depression. Nat Rev Neurosci. 2016;17:497–511.
66. He MC, Shi Z, Sha NN, Chen N, Peng SY, Liao DF, et al. Paricalcitol alleviates Reprints and permission information is available at http://www.nature.com/
lipopolysaccharide-induced depressive-like behavior by suppressing hypothalamic reprints
microglia activation and neuroinflammation. Biochem Pharmacol. 2019;163:1–8.
67. Brown GC, Neher JJ. Microglial phagocytosis of live neurons. Nat Rev Neurosci. Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims
2014;15:209–16. in published maps and institutional affiliations.
68. Kandola A, Ashdown-Franks G, Hendrikse J, Sabiston CM, Stubbs B. Physical
activity and depression: towards understanding the antidepressant mechanisms
of physical activity. Neurosci Biobehav Rev. 2019;107:525–39.
69. Kim T-K, Kim JE, Park JY, Lee JE, Choi J, Kim H, et al. Antidepressant effects of
exercise are produced via suppression of hypocretin/orexin and melanin- Open Access This article is licensed under a Creative Commons
concentrating hormone in the basolateral amygdala. Neurobiol Dis. 2015;79:59–69. Attribution 4.0 International License, which permits use, sharing,
70. Lu Y, Dong Y, Tucker D, Wang R, Ahmed ME, Brann D, et al. Treadmill exercise adaptation, distribution and reproduction in any medium or format, as long as you give
exerts neuroprotection and regulates microglial polarization and oxidative stress appropriate credit to the original author(s) and the source, provide a link to the Creative
in a streptozotocin-induced rat model of sporadic Alzheimer’s disease. J Alzhei- Commons license, and indicate if changes were made. The images or other third party
mers Dis. 2017;56:1469–84. material in this article are included in the article’s Creative Commons license, unless
71. Lee J-M, Kim TW, Park SS, Han JH, Shin MS, Lim BV, et al. Treadmill exercise indicated otherwise in a credit line to the material. If material is not included in the
improves motor function by suppressing Purkinje cell loss in Parkinson disease article’s Creative Commons license and your intended use is not permitted by statutory
rats. Int Neurourol J. 2018;22:S147–55. regulation or exceeds the permitted use, you will need to obtain permission directly
72. Fujii T, Saito DN, Yanaka HT, Kosaka H, Okazawa H. Depressive mood modulates from the copyright holder. To view a copy of this license, visit http://creativecommons.
the anterior lateral CA1 and DG/CA3 during a pattern separation task in cogni- org/licenses/by/4.0/.
tively intact individuals: a functional MRI study. Hippocampus. 2014;24:214–24.
73. Wright BJ, Jackson MB. Long-term potentiation in hilar circuitry modulates gating
by the dentate gyrus. J Neurosci. 2014;34:9743–53. © The Author(s) 2021

Translational Psychiatry (2021)11:461