Ecotoxicology and Environmental Safety 228 (2021) 112979

Contents lists available at ScienceDirect

Ecotoxicology and Environmental Safety

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

Melatonin alleviates hippocampal GR inhibition and depression-like

behavior induced by constant light exposure in mice
Yang Yang a, b, Wenduo Jiang a, b, Yue Feng a, b, Jie Liu a, b, Hongwu Chen c, Deyun Wang a, d,
Ruqian Zhao a, b, *
a
MOE Joint International Research Laboratory of Animal Health & Food Safety, Institute of Immunology, Nanjing Agricultural University, Nanjing 210095, PR China
b
Key Laboratory of Animal Physiology & Biochemistry, College of Veterinary Medicine, Nanjing Agricultural University, Nanjing 210095, PR China
c
Department of Biochemistry and Molecular Medicine, University of California Davis, Sacramento, CA, USA
d
Institute of Traditional Chinese Veterinary Medicine, College of Veterinary Medicine, Nanjing Agricultural University, Nanjing 210095, PR China

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

Edited by Professor Bing Yan Light pollution has become a potential health risk factor worldwide. Chronic exposure to constant light (CCL)
leads to depressive-like behavior, yet the mechanism remains unclear. In this study, mice exposed to CCL for 3
Keywords: weeks exhibited depression-like behaviors, with decreased melatonin in plasma and increased oxidative stress in
Chronic constant light hippocampus. Meanwhile, CCL-exposed mice showed elevated plasma corticosterone (CORT) levels and
Melatonin
diminished glucocorticoid receptor (GR) phosphorylation in hippocampus. Concurrently, glycogen synthase ki­
Oxidative stress
nase 3 beta (GSK3β) was inactivated with increased phosphorylation at Ser9. The interrelationship of GSK3β and
GSK3β
GR GR was clarified in mouse hippocampal neuron (HT-22) cells. GSK3β inhibitor CHIR-99021 induced GR inhi­
Hippocampus bition with diminished phosphorylation, while GR inhibitor RU486 did not affect GSK3β expression or phos­
phorylation. Furthermore, GSK3β-mediated GR inhibition was reproduced in vitro in HT-22 cells treated with
melatonin receptor antagonist luzindole and H2O2 in combination. Finally, melatonin reversed GSK3β-mediated
GR inhibition in hippocampus and improved CCL-induced depression-like behavior in mice. These results
indicate that CCL induces melatonin deficiency and oxidative stress in hippocampus, which in turn leads to
GSK3β-mediated GR inhibition and depression-like behavior in mice.

1. Introduction behavior in weanling mice (Cisse et al., 2016) and adult rats (Borniger
et al., 2014) exposed to dim light at night. Therefore, CORT is regarded
Artificial light pollution has become a potential health risk factor as an important indicator of depression-like behavior in rodents (Lee
worldwide (Navara and Nelson, 2007). Aberrant light conditions in­ et al., 2015). Glucocorticoid receptors (GR) expressed in hippocampus
crease susceptibility to heart disease (Ha and Park, 2005), cancer (Davis serves as a brake to inhibit HPA activity through negative feedback of
and Mirick, 2006), sleep disturbances (Deboer et al., 2007), as well as elevated CORT during stress response (Jacobson and Sapolsky, 1991).
major depressive disorder and mood disorders (Dumont and Beaulieu, Suppression of GR in hippocampus usually leads to chronically higher
2007). Our previous studies found that chronic constant light (CCL) plasma CORT and depression-like behavior (Jacobson and Sapolsky,
exposure for 3 weeks from weaning to 6 weeks of age suppresses hip­ 1991; De Kloet and Reul, 1987). However, whether and how hippo­
pocampal neurogenesis and impairs cognitive behaviors (Yang et al., campal GR inhibition is involved in CCL-induced depression-like
2021). Numerous studies have shown that CCL exposure induces a behavior remain unclear.
depressive-behavioral phenotype in neonatal (Coleman et al., 2016; The serine/threonine kinase glycogen synthase kinase 3 (GSK3) is a
Coleman and Canal, 2017) and adult (Zhou et al., 2018) mice. Increased multifunctional kinase with a critical role in the regulation of oxidative
plasma corticosterone (CORT) levels are linked with depression-like stress. The kinase activity of GSK3 is inhibited through phosphorylation

Abbreviations: CCL, chronic constant light; CORT, corticosterone; DHE, Dihydroethidium; GR, glucocorticoid receptor; GSH-Px, glutathione peroxidase; GSK3β,
glycogen synthase kinase 3 beta; HPA, hypothalamic-pituitary-adrenal; MDA, malondialdehyde; MT, melatonin; ROS, reactive oxygen species; SOD, superoxide
dismutase.
* Correspondence to: Key Laboratory of Animal Physiology & Biochemistry, Nanjing Agricultural University, Nanjing 210095, PR China.
E-mail address: zhaoruqian@njau.edu.cn (R. Zhao).

https://doi.org/10.1016/j.ecoenv.2021.112979
Received 25 July 2021; Received in revised form 27 October 2021; Accepted 7 November 2021
Available online 15 November 2021
0147-6513/© 2021 The Authors. Published by Elsevier Inc. This is an open access article under the CC BY-NC-ND license
(http://creativecommons.org/licenses/by-nc-nd/4.0/).
Y. Yang et al. Ecotoxicology and Environmental Safety 228 (2021) 112979

of a serine residue located at the N-terminus of the proteins, S21 and S9 2.3. Behavioral tests
of GSK3α and GSK3β, respectively (Godemann et al., 1999). It was re­
ported that oxidative stress induces increased GSK3β phosphorylation, The open-field test, elevated plus maze test, forced swimming test,
thus reduced GSK3β activity, in mice liver (Fan et al., 2018). Also, and the tail suspension test were performed as previously described
oxidative stress causes GSK3β downregulation in primary human (Pietropaolo, 2010).
amnion epithelial cells and amnion mesenchymal cells (Lavu et al.,
2019), as well as SH-SY5Y cells (Lingappa et al., 2021) and lung 2.3.1. Open field test
epithelial cells (Li et al., 2017). GSK3β has been reported to phosphor­ To assess the possible effects of chronic constant light on locomotor
ylate GR (Galliher-Beckley et al., 2008; Yang et al., 2019). However, activity, exploratory drive and anxiety, mice were evaluated in the open-
whether GSK3β mediates CCL-induced GR inhibition in hippocampus field paradigm. Mice were individually placed in a 40 × 60 × 50 cm
remains unknown. wooden box. The floor was divided into 24 rectangles. We counted the
Melatonin plays a key role in controlling circadian behavioral re­ number of ambulating (square crossings), rearing (vertical activity),
sponses (Reiter, 1993). It also acts as a potent antioxidant by scavenging grooming, excreting and the time spent in center area in a 6 min session.
reactive oxygen species (ROS) (Galano et al., 2011). CCL-induced To hide animal clues, the apparatus was cleaned with 10% ethanol be­
depressive-like behavior in rodents is usually associated with low tween tests.
melatonin and high ROS levels (Tchekalarova et al., 2018; Wideman and
Murphy, 2009). Melatonin could rescue CCL-induced ROS production in 2.3.2. Elevated plus maze test
rat hippocampus, cortex and cerebellum (Baydas et al., 2002), as well as To evaluate the possible effects of chronic constant light on anxiety
kidney and liver (Baydas et al., 2001). Nevertheless, it is to be explored responses, mice were subjected to the elevated plus maze task. The
whether melatonin can rescue CCL-induced hippocampal GR inhibition experimental apparatus, shaped like a “plus” sign, consisted with two
and depression-like behavior. open arms (30 × 5 cm), two equal-sized closed (30 × 5 × 15 cm) arms
Therefore, the objectives of the present study were, firstly, to elab­ and a central platform (5 × 5 cm). The elevated plus maze placed at a
orate the effects of chronic constant light on anxiety and depressant-like height of 1 m from the ground. The test consisted of placing a mouse in
activity; secondly, to delineate whether and how hippocampal GR in­ the central platform and allowing it to freely explore the maze for 5 min.
hibition is involved in CCL-induced depression-like behavior; and The animal crossing all four paws over the line dividing the central
thirdly, to explore whether melatonin can rescue CCL-induced hippo­ square from the open arm is defined as entering one arm. The time of
campal GR inhibition and depression-like behavior. entries into the open and closed arms were recorded as an indicator of
anti-anxiety behavior of animals in this test.
2. Materials and methods
2.3.3. Forced swimming test
2.1. Ethics statement To evaluate the depressant-like activity in mice, mice were subjected
to the forced swimming test. Mice were individually forced to swim in an
The experimental protocol was approved by the Animal Ethics open cylindrical container. The container height 25 cm and diameter 10
Committee of Nanjing Agricultural University. The project number is cm, containing 19 cm of water at 25 ± 1 ◦ C. And the total duration of
2016YFD0500502. The sampling procedures according to the “Guide­ immobility was measured in 6 min. Mouse was judged to be immobile
lines on Ethical Treatment of Experimental Animals’’ (2006) No. 398 set when it stopped struggling and remained floating motionless in the
by the Ministry of Science and Technology, China. water, making only keep its head above water. The duration of immo­
bility in the last 5 min of the (total) 6 min of swimming time was
2.2. Animals and experimental design recorded with a video camera and record immobility time.

C57BL/6 mice weighing 18–22 g (8 weeks of age) were purchased 2.3.4. Tail suspension test
from Model Animal Research Center of Nanjing University (Nanjing, The tail suspension test was performed for assessing the depressant-
China). A female-to-male ratio of 2:1 was applied for mating. The like activity in mice. Mice were suspended by the tail from a ledge with
offspring pups were used in the following experiments after weaning at 3 adhesive tape (about 1 cm from the tip of the tail). The distance between
weeks of age. the tip of the tail of the mouse and the floor was about 50 cm. Immobility
was defined as the absence of movement for 6 min. The duration of
2.2.1. Experiment 1 immobility in the last 5 min of the (total) 6 min was recorded with a
Twenty male mice weighing 9 ± 0.5 g were randomly divided into video camera and record immobility time.
control (CON, n = 10) and chronic constant light (CCL, n = 10) groups.
Mice in CON group were housed in 12:12 h light: dark cycle with lights 2.4. Total protein extraction and western blotting
on at 8:00 a.m. for 3 weeks, whereas mice in CCL group were housed in
24 h light at 300 lx for 3 weeks. The behavioral tests were performed at Total protein extraction and western blot analysis for glucocorticoid
the end of the light treatment before sacrificed for sampling. receptor (GR; ab2768, diluted 1:1000, Abcam, USA), phospho-
glucocorticoid receptor (p-GR; 4161s, diluted 1:1000, Cell Signaling
2.2.2. Experiment 2 Technology, USA), GSK3α (9338, diluted 1:1000,Cell Signaling Tech­
Twenty-four male mice (3 weeks of age) were randomly divided into nology, USA), phospho-GSK3α (9316, diluted 1:1000, Cell Signaling
CON, CCL and CCL supplemented with melatonin (CCL + Mel) groups (8 Technology, USA), GSK3β (9315, diluted 1:500, Cell Signaling Tech­
mice per group). The light treatment was the same as in the Experiment nology, USA), phospho-GSK3β (9336, diluted 1:500, Cell Signaling
1, except that mice in CCL + Mel group received intraperitoneal injec­ Technology, USA) was carried out according to the recommended pro­
tion of melatonin (10 mg/kg/day) at 6:00 pm for 3 weeks. The dosage of tocols provided by the manufacturers, and β-actin (AP0060, Bioworld,
melatonin is determined according to previous studies (Long et al., USA, diluted 1:10,000) was used as loading control. Images were
2018; Xu et al., 2020). All the mice were subjected to the behavioral tests captured by VersaDoc 4000MP system (Bio-Rad, USA) and the band
at 6 weeks of age before sacrificed for sampling. density was analyzed with Quantity One software (Bio-Rad, USA).

2
Y. Yang et al. Ecotoxicology and Environmental Safety 228 (2021) 112979

2.5. Immunofluorescence of phospho-glucocorticoid receptor 2.10.2. Fluorescence microscope analysis

HT22 cells were incubated with 5 μM MitoSOX Red for 30 min at
Phospho-GR immunopositive cells in hippocampus were detected 37 ◦ C. Then, cells were visualized and imaged with a fluorescence mi­
with immunofluorescent staining as previously described (Yang et al., croscope after washing with PBS.
2019). Hippocampal frozen sections (20 µm) were treated with 0.3%
Triton X-100 for 30 min on coverslips. After washing with PBS for three 2.11. Statistical analysis
times and blocking with 10% FBS for 1 h. And then the frozen sections
were incubated with rabbit polyclonal antibody against phospho-GR All data are presented as means ± SEM. The differences between
(4161, Cell Signaling Technology, USA, diluted 1:400) overnight at groups were analyzed using Student’s t-test or 2-way analysis of vari­
4 ◦ C. Signal of primary antibodies was amplified by ance followed by Tukey’s test for multiple comparisons with SPSS 20.0
Fluorescein-conjugated AffiniPure goat anti-rabbit IgG (H+L) (129302, (IBM, Armonk, NY). The differences were considered statistically sig­
Jackson ImmunoResearch, USA, diluted 1:200) and then visualized and nificant when P < 0.05.
imaged with a fluorescence microscope.
3. Results
2.6. Measurement of corticosterone and melatonin
3.1. Chronic constant light (CCL) exposure increases average daily feed
Corticosterone concentration was determined by Enzyme Immuno­ intake and induces depressive-like behavior
assay (EIA) kit (No. ADI-900-097, Enzo, Farmingdale, NY, USA)
following the manufacturer’s instructions. Serum melatonin levels were CCL exposure did not influence body weight (P > 0.05), but signif­
measured using Mouse MT (Melatonin) ELISA Kit (KE1427, Immuno­ icantly increased average daily feed intake (P < 0.01) (Fig. S1). Open
Way Biotechnology, USA). field test showed that the ambulating entries (Fig. 1A) was significantly
increased (P < 0.01), however, the center area entries and duration
2.7. Measurement of superoxide anion (O2-) in hippocampus (Fig. 1B and C), as well as the grooming and rearing (Fig. 1D and E) were
significantly decreased (P < 0.05) in CCL group, while the excreting was
The superoxide anion was determined using Dihydroethidium (DHE) not significantly changed (Fig. 1F) (P > 0.05). In elevated plus maze test,
kit (S0063, Beyotime, China), according to the manufacturer’s in­ the frequency and duration of entering open arms were significantly (P
structions. Hippocampal frozen sections (20 µm) were incubated with 5 < 0.01) decreased (Fig. 1G and H), while the frequency and duration of
μM DHE for 30 min at 37 ◦ C, then were visualized and imaged with a entering closed arms were significantly (P < 0.01) increased (Fig. 1I and
fluorescence microscope after washing with PBS. J). In forced swimming test (Fig. 1K) and tail suspension test (Fig. 1L),
the immobility time were significantly (P < 0.05) increased.
2.8. Measurement of MDA, SOD and GSH-Px activity
3.2. CCL exposure increases plasma CORT concentration and decreases
Malondialdehyde (MDA), superoxide dismutase (SOD) and gluta­ melatonin concentration
thione peroxidase (GSH-Px) activities in mice hippocampus were
analyzed using MDA, SOD and GSH-Px assay kits (Nanjing Jiancheng CCL exposure significantly (P < 0.01) increased plasma CORT level
Bioengineering Institute, Nanjing, China), respectively, according to the (Fig. 2A) and decreased plasma melatonin level (Fig. 2B).
manufacturer’s instructions.
3.3. CCL exposure induces oxidative stress in hippocampus
2.9. Cell culture and cell viability
CCL exposure significantly increased superoxide anion (O2-)
Mouse hippocampal neuron cells (HT-22) obtained from Shanghai (P < 0.01) (Fig. 3A and B) and MDA (P < 0.05) (Fig. 3C) levels, while
HuiYing Biological Technology Co. Ltd (China) were cultured in Dul­ significantly decreased SOD (P < 0.01) (Fig. 3D) and GSH-Px (P < 0.05)
becco’s Modified Eagle’s Medium (DMEM) (Lot AAH204831, (Fig. 3E) activities in hippocampus.
SH30243.01, Hyclone, USA) containing 10% (v/v) FBS at 37 ◦ C under
5% CO2. Cell viability was tested by a colorimetric assay with CCK-8 3.4. Chronic constant light exposure decreases hippocampus GR and
(G021, Nanjing Jiancheng Bioengineering Institute, China). Cells were GSK3β activity
cultured on 96-well cell culture plate with or without H2O2, melatonin
and luzindole (a competitive antagonist of melatonin receptors). CCL exposure significantly (P < 0.05) decreased phospho-GR protein
content (Fig. 4A and B), and significantly (P < 0.05) increased phospho-
2.10. Measurement of reactive oxygen species (ROS) in HT-22 cells GSK3β protein content (Fig. 4C), indicating inhibit GSK3β activity in
hippocampus. To confirm that GSK3 could regulate the phosphorylation
Mitochondrial ROS was tested using MitoSOX Red (M36008, Invi­ levels of GR in HT-22 cells, the GSK3β activity was inhibit by 3 μM CHIR-
trogen Co., Ltd., USA) according to our previous publication (Yang et al., 99021 (a specific inhibitor of GSK3). Decrease of GSK3α and GSK3β
2017). (Fig. S2A) significantly (P < 0.01) decreased the phosphorylation levels
of GR (Fig. S2B). To further test whether GR could regulate the levels of
2.10.1. Flow cytometer analysis GSK3β, GR activity was blocked by 5 μM RU486 (a specific antagonist of
HT22 cells were digested by trypsin/EDTA and collected into a 1.5 GR). RU486 significantly (P < 0.01) decreased the protein expression of
mL eppendorf tube followed by washing twice with PBS, MitoSOX red phospho-GR (Fig. S2C), yet GSK3β (Fig. S2D) protein expression were
was diluted in PBS and added to cell at 1 × 106 cells/mL. After incu­ not affected by RU486. These results indicate that GSK3β could regulate
bation for 30 min at 37 ◦ C, the cells were centrifuged and collected into a the phosphorylation levels of GR, while GR could not regulate the levels
1.5 mL eppendorf tube, then resuspended in 1 mL PBS and subsequently of GSK3β in HT-22 cells.
transferred to 5-mL FACS tubes. The MitoSOX red fluorescence was then
measured on a flow cytometer (FACS Verse™, BD Biosciences, USA), the 3.5. H2O2 and luzindole in combination simulates the cellular and
argon laser excitation at 488 nm was coupled with emission measure­ molecular events induced by CCL exposure in mouse hippocampus
ments using 585/42 bandpass filters and at least 1 × 104 events were
recorded per sample. Based on the in vivo observation that CCL-exposed mice show

3
Y. Yang et al. Ecotoxicology and Environmental Safety 228 (2021) 112979

Fig. 1. Chronic constant light (CCL) exposure increases average daily feed intake and induces depressive-like behavior. In order to elaborate the effect of
CCL on locomotor activity, anxiety and depressive-like activity, the open-field test, elevated plus maze test, forced swimming test, and the tail suspension test were
performed. (A–F) The counts of ambulating, center area entry, rearing, grooming, excreting, and time spent in center area in the open field test; (G, I) The number of
entries in open and closed arms in the elevated plus maze test; (H, J) The time spent in open and closed arms of the elevated plus maze; (K) The immobility time in the
forced swimming test; (L) The immobility time in the tail suspension test. Values are mean ± SEM (n = 10); *p < 0.05; **p < 0.01, compared with control.

decreased melatonin together with increased ROS, we treated HT-22 phospho-GR/GR (Fig. 5F and L).
cells with H2O2 (100 μM) and melatonin receptor antagonist luzindole
(5 μM) to establish an in vitro model for further mechanistic investiga­
3.6. Melatonin rescues CCL-induced depressive-like behavior
tion. The concentrations of luzindole and H2O2 were determined in
preliminary trials (Fig. S3). H2O2 alone or in combination with luzin­
To further confirm the effect of melatonin on CCL-induced depres­
dole, significantly (P < 0.01) decreased cell number (Fig. 5A), increased
sive-like behavior in vivo, we injected CCL exposure mice with mela­
ROS production (Fig. 5B–C and E) and reduced cell viability (Fig. 5D).
tonin (10 mg/kg/day) for 3 weeks. Melatonin significantly (P < 0.05)
Nevertheless, neither H2O2 nor luzindole could reproduce the intracel­
increased the body weight (Fig. S4A) and decreased the average daily
lular molecular events observed in vivo. Only when HT-22 cells treated
feed intake affected by CCL exposure (Fig. S4B). In open field test,
with H2O2 and luzindole in combination showed the same responses
melatonin significantly rescued the ambulating entries (Fig. 6A)
observed in vivo, with significantly (P < 0.05) increased phospho-
(P < 0.05), the center area entries (Fig. 6B) (P < 0.05) and duration
GSK3β/GSK3β (Fig. 5F and I), and significantly (P < 0.05) decreased
(Fig. 6C) (P < 0.05), as well as the occurrence of grooming (Fig. 6D)

4
Y. Yang et al. Ecotoxicology and Environmental Safety 228 (2021) 112979

Fig. 2. CCL exposure increases plasma CORT concentration and decreases melatonin concentration. (A) Plasma corticosterone level; (B) Plasma melatonin
level. Values are mean ± SEM (n = 12); **p < 0.01, compared with control.

Fig. 3. CCL exposure induces oxidative stress in hippocampus. The superoxide anion was determined using Dihydroethidium (DHE). Hippocampal frozen
sections (20 µm) were incubated with 5 μM DHE for 30 min at 37 ◦ C. (A–B) Superoxide anion (O2-) levels in CA1, CA3, and DG regions (n = 4); (C) Malondialdehyde
(MDA) level (n = 6); (D) Superoxide dismutase (SOD) activity (n = 6); (E) Glutathione peroxidase (GSH-Px) activity (n = 6). Values are means ± SEM. *P < 0.05,
**P < 0.01, compared with control.

5
Y. Yang et al. Ecotoxicology and Environmental Safety 228 (2021) 112979

Fig. 4. Chronic constant light exposure decreases hippocampus GR and GSK3β activity. (A) Phospho-GR-immunopositive neurons in CA1, CA3, and DG regions
(n = 4); (B) Contents of GR and phospho-GR protein (n = 6); (C) Contents of GSK3β and phospho-GSK3β protein (n = 6). Values are means ± SEM. *P < 0.05,
**P < 0.01, compared with control.

(P < 0.05), without affecting counts of rearing (Fig. 6E) (P < 0.05) or 4. Discussion
excreting (Fig. 6F). In elevated plus maze test, the frequency (Fig. 6G)
and duration (Fig. 6H) of entering open arms, as well as the frequency The present study provides the first evidence that chronic constant
(Fig. 6J) of entering closed arms were significantly (P < 0.05) rescued, light exposure for 3 weeks from weaning to 6 weeks of age induces
while the duration (Fig. 6I) of entering closed arms was not significantly depression and anxiety-like behaviors in C57BL/6 mice, which is asso­
(P > 0.05) rescued. Meanwhile, the immobility time tended to be ciated with higher CORT and lower melatonin levels in the blood. In this
reversed in forced swimming test (Fig. 6K) and tail suspension test study, the overall behavioral and endocrine consequences of constant
(Fig. 6L). These results indicate that melatonin rescues CCL-induced light exposure (CCL) were observed on a mouse model in vivo. This
anxiety and depressive-like behavior. model provides a holistic view on the general responses of the body and
helps to pinpoint the specific target and to propose a hypothesis for
3.7. Melatonin rescues CCL-induced elevation of plasma CORT level and further mechanistic investigation. Hippocampus was chosen as the focus
suppression of hippocampal GSK3β/GR activity of the study because it is reported as a “setpoint” for HPA axis activity
(Jacobson and Sapolsky, 1991), a target of melatonin action (Valdes-­
Melatonin significantly rescued plasma CORT level (Fig. 7A), as well Tovar et al., 2018), and a key player in depressive behavior (Otte et al.,
as the ratio of phospho-GR/GR (Fig. 7B), phospho-GSK3α/GSK3α 2016). Further molecular analysis on mouse hippocampus determined
(Fig. 7C) and phospho-GSK3β/GSK3β (Fig. 7D) in hippocampus, affected inhibition of GSK3/GR pathway and increased oxidative stress. Based on
by CCL exposure. the results obtained in vivo, we come up with a hypothesis that CCL
induces melatonin deficiency and oxidative stress in hippocampus,

6
Y. Yang et al. Ecotoxicology and Environmental Safety 228 (2021) 112979

Fig. 5. H2O2 and luzindole in combination simulates the molecular events induced by CCL exposure in mouse hippocampus. HT-22 cells were pretreated
with melatonin receptor antagonist 5 μM luzindole for 3 h, then treated with 100 μM H2O2 for 24 h, and then treated with 50 μM melatonin for 6 h. (A) Microscopic
pictures of cells cultured with or without H2O2, melatonin and luzindole; Mitochondrial ROS was tested using 5 μM MitoSOX Red, incubation for 30 min at 37 ◦ C. (B)
ROS production assessed by immunofluorescence (n = 4); (C, E) ROS production assessed by flow cytometry (n = 4); (D) Cell viability measured by CCK-8 assay
(n = 10). (F) Images of bands detected in Western blot analyses; (G) Protein content of GSK3β; (H) Protein content of phospho-GSK3β; (I) Protein content of phospho-
GSK3β/GSK3β; (J) Protein content of GR; (K) Protein content of phospho-GR; (L) Protein content of phospho-GR/GR. Values with different superscripts are
significantly different from each other (P < 0.05) (n = 3). Note that the loading order of the samples in Western blot differs from the order of the bar graph.

7
Y. Yang et al. Ecotoxicology and Environmental Safety 228 (2021) 112979

Fig. 6. Melatonin supplementation recuses CCL-induced depressive-like behavior. To further confirm the effect of melatonin on CCL-induced depressive-like
behavior in vivo, CCL exposure mice were injection with 10 mg/kg/day melatonin for 3 weeks. (A–F) The counts of ambulating, center area entry, rearing, grooming,
excreting, and time spent in center area in the open field test; (G, I) The number of entries in open and closed arms in the elevated plus maze test; (H, J) The time
spent in open and closed arms of the elevated plus maze test; (K) The immobility time in the forced swimming test; (L) The immobility time in the tail suspension test.
Values with different superscripts are significantly different from each other (P < 0.05) (n = 8).

which in turn leads to GSK3β-mediated GR inhibition and the depressive behavior in CCL-exposed mice.
depression-like behavior in mice. However, it is not possible to verify Constant light exposure is known to lower pineal melatonin synthesis
this hypothesis in the in vivo mouse model. So, hippocampal cells were and to decrease mean melatonin concentrations in the blood (Pablos
used to delineate the proposed signaling pathway with respective in­ et al., 1998). A lot of studies have reported that patients with the
hibitors in vitro. The in vitro model provides convenience in deductive symptomatic nature of affective disorders, such as depression and anx­
cause-consequence analysis through loss-of-function approaches, iety, are characterized with disrupted rhythms of both melatonin and
although it may not reflect the real situation in the animal body. cortisol (Germain and Kupfer, 2008; Monteleone et al., 2011; Stetler
Therefore, we returned to in vivo mouse model to prove that melatonin et al., 2004). Nevertheless, the link between melatonin and cortisol is
can alleviate the inhibition of GSK3/GR pathway in hippocampus and unclear.

8
Y. Yang et al. Ecotoxicology and Environmental Safety 228 (2021) 112979

Fig. 7. Melatonin supplementation rescues CCL-induced high plasma CORT level and low hippocampus GSK3β/GR activity. CCL exposure mice were in­
jection with 10 mg/kg/day melatonin for 3 weeks. (A) Plasma corticosterone level (n = 8); (B) Contents of GR and phospho-GR protein (n = 6); (C) Contents of
GSK3α and phospho-GSK3α protein (n = 6); (D) Contents of GSK3β and phospho-GSK3β protein (n = 6). Values with different superscripts are significantly different
from each other (P < 0.05).

Free radicals are important mediators of tissue injury in several and depression-like behaviors in CCL-exposed mice, but the signaling
neurodegenerative diseases (Almli et al., 2001), as brain is especially pathway by which CCL exposure causes hippocampal GR inhibition re­
susceptible to oxidative insults because of its high oxygen utilization and mains elusive.
low levels of antioxidative protective enzymes and radical scavengers GSK3 is a multifunctional kinase that can phosphorylate GR (Galli­
(Reiter, 1998). Previous studies have shown that constant light exposure her-Beckley et al., 2008). In this study, we found that CCL-induced
leads to an increase in MDA and a reduction in SOD and GSH-Px in the hippocampal GR inhibition was accompanied by GSK3β inactivation
brain (Baydas et al., 2001; Albarran et al., 2001). Our previous study by its phosphorylation at Ser9. We conducted in vitro study with HT-22
also found that white LED light exposure induces higher mitochondrial cells to have further confirmed that role of GSK3 in GR phosphorylation.
ROS level in HT-22 cells (Yang et al., 2017). Melatonin exerts neuro­ Both melatonin and H2O2 are reported to affect GSK3 function. For
protective effects due to its direct free-radical scavenging properties instance, melatonin prevents LPS-induced oxidative stress in human
(Reiter et al., 2001; Baydas et al., 2003) and indirect antioxidant ac­ alveolar epithelial cells via the GSK3β/Nrf2 pathway (Park et al., 2013;
tivities by stimulating major antioxidant enzymes in the central nervous Ding et al., 2020). H2O2 induces cell apoptosis by triggering activation of
system (Rodriguez et al., 2004). Indeed, we detected high ROS and MDA Akt and GSK3β in H9c2 cells (Park et al., 2013). Therefore, we used
production, accompanied by lower SOD and GSH-Px activities, in the HT-22 cells to test our hypothesis that CCL affects GSK3-mediated GR
hippocampus of CCL-exposed mice. Therefore, we speculate that CCL pathway in hippocampus via melatonin deficiency and oxidative stress.
induced depression and anxiety-like behaviors in mice directly or indi­ Indeed, we found that, neither luzindole nor H2O2, can simulate the
rectly through melatonin deficiency causes oxidative stress. results of in vivo experiments when treated alone. Only when HT-22
GR is abundantly expressed in hippocampus which serves as a brake cells treated with H2O2 and luzindole in combination showed the
to inhibit HPA activity through negative feedback of elevated CORT same responses observed in vivo. This observation, combined with other
during stress response (Jacobson and Sapolsky, 1991). Previous studies publications, indicate a divergent signaling pathway of melatonin, other
reported that chronic stress induces depressive-like behaviors in asso­ than H2O2-induced oxidative stress, is involved in the cascade of GR
ciation with higher plasma CORT level and hippocampal GR inhibition regulation.
(Wang et al., 2019). In this study, we also found that CCL decreased the Moreover, melatonin has been shown to rescue the depressive-like
phosphorylation levels of GR. Therefore, we speculate that CCL leads to behavior induced by different types of chronic stress (Ramirez-Ro­
GR inhibition in the hippocampus and thereby impairs the feedback of driguez et al., 2014; Ali et al., 2020; Crupi et al., 2010). In this study, we
HPA activity. This hypothesis can explain the excessive CORT secretion show that melatonin can protect mice from CCL-induced depressive-like

9
Y. Yang et al. Ecotoxicology and Environmental Safety 228 (2021) 112979

behavior. All the CCL-induced alterations, including higher plasma Baydas, G., Reiter, R.J., Nedzvetskii, V.S., Nerush, P.A., Kirichenko, S.V., 2002. Altered
glial fibrillary acidic protein content and its degradation in the hippocampus, cortex
CORT level, hippocampal GR and GSK3β, are reversed by melatonin
and cerebellum of rats exposed to constant light: reversal by melatonin. J. Pineal
treatment. Res. 33, 134–139.
Baydas, G., Kutlu, S., Naziroglu, M., Canpolat, S., Sandal, S., Ozcan, M., Kelestimur, H.,
2003. Inhibitory effects of melatonin on neural lipid peroxidation induced by
5. Conclusions intracerebroventricularly administered homocysteine. J. Pineal Res. 34, 36–39.
Borniger, J.C., McHenry, Z.D., Abi Salloum, B.A., Nelson, R.J., 2014. Exposure to dim
The present study provides the first evidence that chronic constant light at night during early development increases adult anxiety-like responses.
Physiol. Behav. 133, 99–106.
light exposure for 3 weeks induces depressive-like behaviors in adoles­ Cisse, Y.M., Peng, J., Nelson, R.J., 2016. Dim light at night prior to adolescence increases
cent mice, in which diminished plasma melatonin level coincided with adult anxiety-like behaviors. Chronobiol. Int. 33, 1473–1480.
increased oxidative stress in hippocampus, and increased plasma CORT Coleman, G., Canal, M.M., 2017. Postnatal light effects on pup stress axis development
are independent of maternal behavior. Front. Neurosci. 11, 46.
level was associated with GR inhibition in hippocampus. We identified
Coleman, G., Gigg, J., Canal, M.M., 2016. Postnatal light alters hypothalamic-pituitary-
GSK3β as a mediator for GR inhibition and further elucidated the adrenal axis function and induces a depressive-like phenotype in adult mice. Eur. J.
signaling pathway linking CCL-induced melatonin deficiency and Neurosci. 44, 2807–2817.
oxidative stress with GSK3β-mediated GR inhibition in HT-22 hippo­ Crupi, R., Mazzon, E., Marino, A., La Spada, G., Bramanti, P., Cuzzocrea, S., Spina, E.,
2010. Melatonin treatment mimics the antidepressant action in chronic
campal neuronal cells in vitro. Finally, we demonstrated the protective corticosterone-treated mice. J. Pineal Res. 49, 123–129.
effect of melatonin on CCL-induced depressive-like behavior in mice Davis, S., Mirick, D.K., 2006. Circadian disruption, shift work and the risk of cancer: a
through the mitigation of GSK3β-mediated hippocampal GR inhibition. summary of the evidence and studies in Seattle. Cancer Cause Control 17, 539–545.
De Kloet, E.R., Reul, J.M., 1987. Feedback action and tonic influence of corticosteroids
Our findings implicate possible adverse consequences of excessive light on brain function: a concept arising from the heterogeneity of brain receptor
exposure on brain function and behavior, either in daily life or in systems. Psychoneuroendocrinology 12, 83–105.
therapy. Deboer, T., Detari, L., Meijer, J.H., 2007. Long term effects of sleep deprivation on the
mammalian circadian pacemaker. Sleep 30, 257–262.
Ding, Z., Wu, X., Wang, Y., Ji, S., Zhang, W., Kang, J., Li, J., Fei, G., 2020. Melatonin
CRediT authorship contribution statement prevents LPS-induced epithelial-mesenchymal transition in human alveolar
epithelial cells via the GSK-3beta/Nrf2 pathway. Biomed. Pharmacother. = Biomed.
Pharmacother. 132, 110827.
YY contributed to behavior tests, data analysis, and drafting of the Dumont, M., Beaulieu, C., 2007. Light exposure in the natural environment: relevance to
manuscript. YY and WJ were responsible for animal care, breeding and mood and sleep disorders. Sleep Med. 8, 557–565.
sampling. YF and JL provided technical support. RZ and DW contrib­ Fan, X., Wang, L., Huang, J., Lv, H., Deng, X., Ci, X., 2018. Pterostilbene reduces
acetaminophen-induced liver injury by activating the Nrf2 antioxidative defense
uted to conception, experimental design and data interpretation. RZ and system via the AMPK/Akt/GSK3beta pathway. Cell. Physiol. Biochem.: Int. J. Exp.
HC contributed to critical revision of the manuscript. All authors read Cell. Physiol. Biochem. Pharmacol. 49, 1943–1958.
and approved the final manuscript. Galano, A., Tan, D.X., Reiter, R.J., 2011. Melatonin as a natural ally against oxidative
stress: a physicochemical examination. J. Pineal Res. 51, 1–16.
Galliher-Beckley, A.J., Williams, J.G., Collins, J.B., Cidlowski, J.A., 2008. Glycogen
synthase kinase 3beta-mediated serine phosphorylation of the human glucocorticoid
Declaration of Competing Interest receptor redirects gene expression profiles. Mol. Cell. Biol. 28, 7309–7322.
Germain, A., Kupfer, D.J., 2008. Circadian rhythm disturbances in depression. Hum.
The authors declare that they have no known competing financial Psychopharm. Clin. 23, 571–585.
Godemann, R., Biernat, J., Mandelkow, E., Mandelkow, E.M., 1999. Phosphorylation of
interests or personal relationships that could have appeared to influence tau protein by recombinant GSK-3beta: pronounced phosphorylation at select Ser/
the work reported in this paper. Thr-Pro motifs but no phosphorylation at Ser262 in the repeat domain. FEBS Lett.
454, 157–164.
Ha, M., Park, J., 2005. Shiftwork and metabolic risk factors of cardiovascular disease.
Acknowledgment J. Occup. Health 47, 89–95.
Jacobson, L., Sapolsky, R., 1991. The role of the hippocampus in feedback regulation of
the hypothalamic-pituitary-adrenocortical axis. Endocr. Rev. 12, 118–134.
This work was supported by the National Key Research and Devel­
Lavu, N., Richardson, L., Radnaa, E., Kechichian, T., Urrabaz-Garza, R., Sheller-Miller, S.,
opment Program of China (2016YFD0500502), the Priority Academic Bonney, E., Menon, R., 2019. Oxidative stress-induced downregulation of glycogen
Program Development of Jiangsu Higher Education Institutions (PAPD), synthase kinase 3 beta in fetal membranes promotes cellular senescencedagger. Biol.
the Postgraduate Research & Practice Innovation Program of Jiangsu Reprod. 101, 1018–1030.
Lee, D.Y., Kim, E., Choi, M.H., 2015. Technical and clinical aspects of cortisol as a
Province (KYCX18_0716), and Jiangsu Collaborative Innovation Center biochemical marker of chronic stress. Bmb Rep. 48, 209–216.
of Meat Production and Processing, Quality, and Safety Control. The Li, X., Lai, Y., Li, J., Zou, M., Zou, C., 2017. Oxidative stress destabilizes protein arginine
funders had no role in study design, data collection and analysis, deci­ methyltransferase 4 via glycogen synthase kinase 3beta to impede lung epithelial cell
migration. Am. J. Physiol. Cell Physiol. 313, C285–C294.
sion to publish, or preparation of the manuscript. Lingappa, S., Shivakumar, M.S., Manivasagam, T., Somasundaram, S.T., Seedevi, P.,
2021. Neuroprotective effect of epalrestat on hydrogen peroxide induced
neurodegeneration in SH-SY5Y cellular model. J. Microbiol. Biotechnol. 31,
Appendix A. Supplementary material
867–874.
Long, T., Yang, Y., Peng, L., Li, Z., 2018. Neuroprotective effects of melatonin on
Supplementary data associated with this article can be found in the experimental allergic encephalomyelitis mice via anti-oxidative stress activity.
online version at doi:10.1016/j.ecoenv.2021.112979. J. Mol. Neurosci.: MN 64, 233–241.
Monteleone, P., Martiadis, V., Maj, M., 2011. Circadian rhythms and treatment
implications in depression. Prog. Neuropsychopharmacol. Biol. Psychiatry 35,
References 1569–1574.
Navara, K.J., Nelson, R.J., 2007. The dark side of light at night: physiological,
epidemiological, and ecological consequences. J. Pineal Res. 43, 215–224.
Albarran, M.T., Lopez-Burillo, S., Pablos, M.I., Reiter, R.J., Agapito, M.T., 2001.
Otte, C., Gold, S.M., Penninx, B.W., Pariante, C.M., Etkin, A., Fava, M., Mohr, D.C.,
Endogenous rhythms of melatonin, total antioxidant status and superoxide
Schatzberg, A.F., 2016. Major depressive disorder. Nat. Rev. Dis. Prim. 2, 16065.
dismutase activity in several tissues of chick and their inhibition by light. J. Pineal
Pablos, M.I., Reiter, R.J., Ortiz, G.G., Guerrero, J.M., Agapito, M.T., Chuang, J.I.,
Res. 30, 227–233.
Sewerynek, E., 1998. Rhythms of glutathione peroxidase and glutathione reductase
Ali, T., Rahman, S.U., Hao, Q., Li, W., Liu, Z., Ali Shah, F., Murtaza, I., Zhang, Z.,
in brain of chick and their inhibition by light. Neurochem. Int. 32, 69–75.
Yang, X., Liu, G., Li, S., 2020. Melatonin prevents neuroinflammation and relieves
Park, E.S., Kang, J.C., Kang, D.H., Jang, Y.C., Yi, K.Y., Chung, H.J., Park, J.S., Kim, B.,
depression by attenuating autophagy impairment through FOXO3a regulation.
Feng, Z.P., Shin, H.S., 2013. 5-AIQ inhibits H2O2-induced apoptosis through reactive
J. Pineal Res.
oxygen species scavenging and Akt/GSK-3beta signaling pathway in H9c2
Almli, L.M., Hamrick, S.E.G., Koshy, A.A., Tauber, M.G., Ferriero, D.M., 2001. Multiple
cardiomyocytes. Toxicol. Appl. Pharmacol. 268 (1), 90–98.
pathways of neuroprotection against oxidative stress and excitotoxic injury in
Pietropaolo, S., 2010. Mood and anxiety-related phenotypes in mice: characterization
immature primary hippocampal neurons. Dev. Brain Res. 132, 121–129.
using behavioral tests. Genes Brain Behav. 9.
Baydas, G., Ercel, E., Canatan, H., Donder, E., Akyol, A., 2001. Effect of melatonin on
Ramirez-Rodriguez, G., Vega-Rivera, N.M., Oikawa-Sala, J., Gomez-Sanchez, A., Ortiz-
oxidative status of rat brain, liver and kidney tissues under constant light exposure.
Lopez, L., Estrada-Camarena, E., 2014. Melatonin synergizes with citalopram to
Cell Biochem. Funct. 19, 37–41.

10
Y. Yang et al. Ecotoxicology and Environmental Safety 228 (2021) 112979

induce antidepressant-like behavior and to promote hippocampal neurogenesis in Wang, W., Liu, L., Yang, X., Gao, H., Tang, Q.K., Yin, L.Y., Yin, X.Y., Hao, J.R., Geng, D.
adult mice. J. Pineal Res. 56, 450–461. Q., Gao, C., 2019. Ketamine improved depressive-like behaviors via hippocampal
Reiter, R.J., 1993. The melatonin rhythm – both a clock and a calendar. Experientia 49, glucocorticoid receptor in chronic stress induced-susceptible mice. Behav. Brain Res.
654–664. 364, 75–84.
Reiter, R.J., 1998. Oxidative damage in the central nervous system: protection by Wideman, C.H., Murphy, H.M., 2009. Constant light induces alterations in melatonin
melatonin. Prog. Neurobiol. 56, 359–384. levels, food intake, feed efficiency, visceral adiposity, and circadian rhythms in rats.
Reiter, R.J., Acuna-Castroviejo, D., Tan, D.X., Burkhardt, S., 2001. Free radical-mediated Nutr. Neurosci. 12, 233–240.
molecular damage – mechanisms for the protective actions of melatonin in the Xu, C.N., Kong, L.H., Ding, P., Liu, Y., Fan, Z.G., Gao, E.H., Yang, J., Yang, L.F., 2020.
central nervous system. Ann. N. Y. Acad. Sci. 939, 200–215. Melatonin ameliorates pressure overload-induced cardiac hypertrophy by
Rodriguez, C., Mayo, J.C., Sainz, R.M., Antolin, I., Herrera, F., Martin, V., Reiter, R.J., attenuating Atg5-dependent autophagy and activating the Akt/mTOR pathway.
2004. Regulation of antioxidant enzymes: a significant role for melatonin. J. Pineal Biochim. Biophys. Acta Mol. Basis Dis. 1866, 165848.
Res. 36, 1–9. Yang, Y., Zong, Y., Sun, Q., Jia, Y., Zhao, R., 2017. White light emitting diode suppresses
Stetler, C., Dickerson, S.S., Miller, G.E., 2004. Uncoupling of social zeitgebers and diurnal proliferation and induces apoptosis in hippocampal neuron cells through
cortisol secretion in clinical depression. Psychoneuroendocrinology 29, 1250–1259. mitochondrial cytochrome c oxydase-mediated IGF-1 and TNF-alpha pathways. Free
Tchekalarova, J., Stoynova, T., Ilieva, K., Mitreva, R., Atanasova, M., 2018. Agomelatine Radic. Biol. Med. 113, 413–423.
treatment corrects symptoms of depression and anxiety by restoring the disrupted Yang, Y., Jia, Y., Sun, Q., Dong, H., Zhao, R., 2019. White light emitting diode induces
melatonin circadian rhythms of rats exposed to chronic constant light. Pharmacol. autophagy in hippocampal neuron cells through GSK-3-mediated GR and RORalpha
Biochem. Behav. 171, 1–9. pathways. Aging 11, 1832–1849.
Valdes-Tovar, M., Estrada-Reyes, R., Solis-Chagoyan, H., Argueta, J., Dorantes-Barron, A. Yang, Y., Feng, Y., Hu, Y., Liu, J., Shi, H., Zhao, R., 2021. Exposure to constant light
M., Quero-Chavez, D., Cruz-Garduno, R., Cercos, M.G., Trueta, C., Oikawa-Sala, J., impairs cognition with FTO inhibition and m(6)A-dependent TrkappaB repression in
Dubocovich, M.L., Benitez-King, G., 2018. Circadian modulation of neuroplasticity mouse hippocampus. Environ. Pollut. 283, 117037.
by melatonin: a target in the treatment of depression. Br. J. Pharmacol. 175, Zhou, Y., Zhang, H.K., Liu, F., Lei, G., Liu, P., Jiao, T., Dang, Y.H., 2018. Altered light
3200–3208. conditions contribute to abnormalities in emotion and cognition through HINT1
dysfunction in C57BL/6 mice. Front. Behav. Neurosci. 12, 110.

11