Abstract
Depression is one of the most common psychiatric diseases worldwide. With the increase in the number of depressive episodes, cognitive dysfunction may be accelerated. Although significant findings related to the pathogenesis of depression have been reported, the precise molecular mechanisms of depression-related cognitive disorders have not yet been fully clarified. In this study, we collected serum copper levels and evaluated cognitive functions in patients with major depressive disorder (MDD) and healthy controls. Furthermore, we adopted a chronic restraint stress paradigm to induce depressive-like behaviors in mice, namely stress mice, and C57BL/6J mice were regarded as naive mice. We further measured the copper levels in hippocampus and dendritic spines of hippocampal neurons in stress mice and naive mice. Besides, we evaluated the changes of N-methyl-D-aspartic acid receptor subunit 2B (GluN2B) and postsynaptic density protein 95 (PSD95) levels in hippocampus, and dendritic spines of hippocampal neurons in stress mice with a copper inhibitor. The results revealed that high levels of copper and decreased memory scores exhibited a significant correlation in MDD patients. We further found that the copper inhibitor increased GluN2B and PSD95 levels in hippocampus, which could be involved in the regulation of dendritic spines of hippocampal neurons in stress mice. These results suggested that high levels of copper suppressed GluN2B and PSD95 levels in hippocampus, damaged synaptic function, and caused memory disorders in depression. Our findings provided a promising perspective for high levels of copper in patients with depression-related cognitive disorders, and copper may even be targeted for therapeutic manipulation.
Major depressive disorder (MDD) is one of the most common psychiatric disorders, which is characterized by increased incidence due to continuous stress on life and work (1). MDD patients have impaired cognitive functions, including memory, processing, language, and vision domains (2). With the increase of depressive episodes, cognitive dysfunction may be accelerated (3). Although pharmacological treatments have achieved a great progress, the precise molecular mechanisms of depression-related cognitive disorders have not yet been fully elucidated. Consequently, it is imperative and essential to investigate the underlying mechanisms and facilitate the identification of new therapeutic targets for depression.
At present, accumulating evidence has been put into the functions of glutamic acid, which is the most excitatory neurotransmitter in the brain (4,5). Glutamic acid is transmitted to postsynaptic neurons under the mediation of glutamate receptors (6). The postsynaptic membrane receives neurotransmitters transmitted by glutamic receptors to the dendritic spines, causing excitation of the next neurons and maintaining the integrity of cognitive function (7). It was reported that glutamic acid metabolic disturbance and abnormal N-methyl-D-aspartic acid receptor subunit 2B (GluN2B) levels were involved in MDD (8,9). Postsynaptic density protein 95 (PSD95) can induce localization and aggregation of GluN2B on postsynaptic membrane, and contribute to glutamic acid transmission to dendritic spines (10). However, the potential mechanism of glutamic acid transmission in depression-related cognitive dysfunction remains elusive.
With the occurrence of depression, people’s feeding behaviors have changed. The material circulation and dynamic balance of metal ions are destroyed, leading to the abnormal glutamic metabolism (11). It was reported that there are several abnormal metal ions in aging brains (12). Studies have shown that the disorder of metal ion metabolism is an important pathophysiological mechanism of Alzheimer’s disease (13–15). Copper, as an important metal ion, can be found in the brain, and it is actively involved in oxidation-reduction reactions in neurons (16). A previous study has shown that the oxidation of high levels of copper caused damage to neurotransmitter transmission and cellular functions of neurons (17). In addition, high levels of copper could produce excessive reactive oxygen species and accelerate the aging process of neurons (18). To study the effects of copper levels on MDD, we measured serum copper levels and evaluated the cognitive functions of MDD patients and healthy controls (HCs). In addition, we measured copper levels in serum and hippocampus tissues, and dendritic spines of hippocampal neurons from mice with depressive-like behaviors using the chronic restraint stress paradigm. We supposed that copper might provide new insights into the pathogenesis of depression-related cognitive dysfunction. Therefore, targeting copper could be a valuable strategy to early identify and treat depression-related memory disorders.
Method
Human Subjects
In this study, 20 MDD patients (age, 33.90 [7.10] years old) and 20 age- and gender-matched HCs (age, 33.20 [6.01] years old) were recruited. All subjects were right-handed and belonged to the Han nationality. The demographic and clinical characteristics of all subjects were assessed by a psychiatrist. The symbols of 33.20 (6.01) and 33.90 (7.10) represent the mean (standard deviation).
MDD Group
A total of 20 cases with MDD were included
The inclusion criteria for the MDD group were as follows: (a) MDD diagnosis by a psychiatrist based on the fifth edition of the Diagnostic and Statistical Manual of Mental Disorders (DSM-5), (b) the 24-item Hamilton Depression Scale (HAMD-24) score ≥ 20 points, (c) the Young Manic Rating Scale score < 6 points, (d) psychiatric drug-naive and treatment-naive, and (e) voluntarily participation in the study and signing informed consent form.
The exclusion criteria for the MDD group were as follows: (a) cases with any other mental or neurological disorders, (b) cases with organic mental disorders, (c) no manic or hypomanic episodes, pregnant or lactating women, (d) cases with an abnormal brain structure or contraindications for magnetic resonance imaging (MRI), and (e) receiving antipsychotic drugs.
HC Group
A total of 20 HCs were recruited
The inclusion criteria for the HC group were as follows: (a) the absence of any underlying diseases or conditions, (b) HAMD-24 score < 8 points, and (c) voluntarily participation in the study and signing informed consent form.
The exclusion criteria for the HC group were as follows: (a) psychiatric diseases or family history of mental disorders, (b) pregnant or lactating women, and (c) cases with an abnormal brain structure or contraindications for MRI.
Collection and Analysis of Blood Samples
Participants’ blood samples were obtained within 48 hours. In this study, 2 mL of blood samples were obtained from each participant in the morning. Then, 40 μL blood was added into 1 mL of protein precipitant and was fully shaken up. After that, the blood samples were centrifuged for 5 minutes at 2 000 rpm to obtain the serum samples. Serum samples were stored at −20°C until being analyzed. The analysis of blood samples was performed in a biochemical laboratory.
Ethics Statement
Animal experimental protocols were approved by the Ethics Committee of Experimental Animals of South China University of Technology (Guangzhou, China). We conducted an exploration on mice in accordance with the National Institutes of Health Guide for the Care and Committee Guidelines of Laboratory Animals. The present study was approved by the Ethics Committee of Guangzhou First People’s Hospital (Guangzhou, China).
Chronic Restraint Stress Paradigm
Four-week-old male C57BL/6J mice were purchased from Guangdong Medical Laboratory Animal Center (Guangzhou, China). Some C57BL/6J mice, as the naive group, were placed in a mouse cage, and some C57BL/6J mice with chronic restraint stress paradigm, as the stress group, were placed in another mouse cage. Mice of 2 groups were placed in a cyclic rhythm environment of 12 hours light/12 hours dark with food and water availability. Apart from that, every mouse of the stress group was placed in a tight-fitting and closed tube with only a small hole for breathing within 3 hours every day, and mice could never move. The earlier-mentioned operation was repeated on mice of the stress group for twenty-one consecutive days to induce depressive-like behaviors. Finally, behavioral tests were performed to validate depressive-like behaviors and to evaluate the spatial memory ability of mice.
Experimental Design With Copper Inhibitor (Tetrathiomolybdate [TM])
The TM is used to reduce copper concentration. Some C57BL/6J mice, as the stress + TM group (with copper inhibitor and chronic restraint stress paradigm), were placed in a mouse cage, and some C57BL/6J mice (with sterilized saline solution and chronic restraint stress paradigm), as the stress group, were placed in another mouse cage. Two groups of mice were placed with the earlier-mentioned chronic restraint stress paradigm at 7 pm within 3 hours every day. Apart from that, every mouse of stress + TM group was injected with the copper inhibitor, TM (Sigma), by gavage at a dosage of 12 mg/kg/d at 9 am every day. Every mouse of the stress group was injected with sterilized saline solution (1 mL/kg/d) as control at 9 am every day. The earlier-mentioned operation was repeated on 2 groups of mice for twenty-one consecutive days. Finally, behavioral tests were performed to validate depressive-like behaviors and to evaluate spatial memory ability in the 2 groups of mice.
Behavioral Tests
Weight test
The weight of mice before and after twenty-one days of experiment was measured by an electronic weighing instrument.
Open field test
Mice were placed in an open box made of light-proof materials with length, width, and height of 100, 100, and 50 cm, respectively, and the bottom of the box was black, which was divided into 25 squares of 20 cm × 20 cm with white lines. The experiment was conducted in a quiet environment from 08:00 am to 12:00 pm. After placing mice into the central compartment, behaviors of mice were recorded for 5 minutes, including the time spent in the central compartment and the distance moved. Each mouse was subjected to 1 behavioral measurement, and after the measurement, a wet cloth was used to remove feces left in the compartment, and then, the open box was cleaned with 75% alcohol cotton balls around the perimeter and bottom of the box, and the next mouse was tested after the box was thoroughly cleaned.
Forced swimming test
On the first day, an appropriate amount of 25°C water was poured into the plastic bucket used for testing, where the water was highly enough to ensure that tails of mice could not touch the bottom of the bucket, and then, mice were put in for 15 minutes of acclimatization training. On the second day, an operation performed on the first day was repeated, and mice were acclimatized for 1 minute before entering the water. The resting time (no movement time), struggling time, and swimming time of each mouse in the second 5 minute were recorded with a timer. Resting time included the time when mice kept floating in the water with their forelimbs curled up and only their heads were exposed. Struggling time included the time when mice were scratched and climbed along the wall of the barrel. Swimming time included the time when mice kept swimming along the wall of the barrel in the water. Before each behavioral test, the plastic bucket was cleaned, and the used water was replaced with fresh water to avoid the odor left by the last animal to interfere with the test results, and to avoid the influences of the noise in the environment on the experimental results. Mice were placed on the heater after the test to dry their fur and to reduce unnecessary mortality of mice.
The measurement index included the resting time of mice in the plastic bucket, that was, the time when mice were floating on the water surface with only a slight physical activity or the time when the body weight of mice was perpendicular to the horizontal surface, and only the nose of mice was exposed to the water surface to breathe.
Sucrose preference test
First, 1% sucrose solution was prepared with ddH20, followed by the following operations: (a) on the first day (24 hours), 2 bottles of 1% concentration of 150 mL sucrose solution were placed at the same height in the same position in each cage; (b) on the second day (24 hours), 1 bottle of 1% concentration of 150 mL sucrose solution and 1 bottle of 150 mL capacity of ddH20 were placed at the same height in each cage at the same time. After 12 hours, the sucrose solution and ddH20 were swapped; (c) on the third day (24 hours), no water and no food were allowed; (d) on the fourth day (2 hours), a bottle of 1% sucrose solution and a bottle of 150 mL ddH20 were placed simultaneously at the same height in each cage, and positions were swapped after 1 hour, and the consumption of sucrose solution and ddH20 in each cage was measured using a measuring cylinder. The sugar-water preference of mice was calculated as follows: sugar-water preference of mice = (sugar-water consumption/(sugar-water consumption + ddH20 consumption) × 100%).
Morris water maze (MWM) test
The experimental principle of the MWM test is based on the characteristic that mice are strong swimmers, while are averse to being in water. Mice instinctively force themselves to swim to find a platform above the horizontal surface to rest after entering the water. During the training phase of the experiment, mice are artificially placed into the water maze, and by observing objects in the maze, they force themselves to find an appropriate place to stay, to collect information about their own spatial location, to organize information around them and store it in their memory, and to help them find their destination quickly. Through 5 days of repetitive training, mice repeatedly learn the environment around the maze and deepen their localization awareness of spatial locations in the maze, thereby assessing their memory on the sixth day.
(a) Localization experiment: mice were put into the pool from the entry points of the 4 quadrants, and the time for mice to find the platform within 1 minute was recorded, which was the evasion latency period, and the earlier-mentioned experiment was repeated in 5 days. (b) Probe test: On the sixth day of the experiment, the platform in the water was removed, and mice were put into the pool from the opposite quadrant of the target platform, and the percentage of the distance moved across the platform quadrant in 1 minute was recorded, as well as the number of times that the platform was crossed, in order to respond to the memory of mice.
Observational index: The MWM test was performed to determine the memory of mice, including the evasion latency during localization, the number of times crossing the platform during the probe test, and the percentage of the distance moved in the target quadrant to the whole moving distance.
Measurement of Copper Levels
Serum sample: We took a glass capillary tube with 2–2.5 cm long capillary segments, soaked it in 1% heparin solution for 1 minute, and dried it for use. An operator used the right hand holding the capillary tube from the inner canthus of the mouse to quickly insert the conjunctiva, and then, gently pushed toward the bottom of the eye, and gently rotated the capillary tube to cut through the venous plexus, so that the venous blood flew along the capillary wall to the EP tube. The EP tube was placed at room temperature for 3 hours, centrifugation was performed at 2 000 rpm for 15 min, and the supernatant was taken in a new EP tube.
Tissues samples: After mice were neck-broken and executed, their hippocampal tissues were peeled off, were placed in the EP tube, 50 µL ddH2O was added, the hippocampal tissues were grounded on ice, centrifuged at 2 000 rpm for 15 min, and the supernatant was taken in a new EP tube.
Examination: First, 50 µL of the sample (supernatant of tissues or serum) was added to 50 µL of the complexing agent 3,5 dibromo-PAESA; the reaction was mixed, and incubated at 37°C for 5 minutes to produce a blue complex. The absorbance of the blue complex was measured using a spectrophotometer at a measurement wavelength of 600 nm, and the reading was taken to obtain the copper levels.
Quantitative Reverse Transcription-Polymerase Chain Reaction (RT-qPCR)
In this study, TRIzol reagent was used to isolate total RNA from hippocampal tissue. An RNA kit was used to synthesize cDNA based on the protocol, and a SYBP Rremix Ex Taq II system was utilized to make PCR reactions.
GluN2B: Forward 5′-CAGCAAAGCTCGTTCCCAAAA-3′, and
Reverse 5′-GTCAGTCTCGTTCATGGCTAC-3′.
PSD95: Forward 5′-TCAGACGGTCACGATCATCGCT-3′, and
Reverse 5′-GTTGCTTCGCAGAGATGCAGTC-3′.
Western Blot Assay
Radio-immunoprecipitation assay lysis buffer was used to isolate total protein, and a bicinchoninic acid assay kit was utilized to qualify protein. Proteins samples were transferred onto polyvinylidene difluoride membranes according to the protocol. The membranes were incubated with primary antibodies (GluN2B: ab28373, PSD95: ab18258, β-Tublin: ab15568, GAPDH: ab9485; Abcam) at 4°C overnight, and then, incubated with secondary antibodies for 1 hour in the following day at room temperature. Protein bands were visualized using an enhanced chemiluminescence detection system.
Immunohistochemistry (IHC)
We put the paraffin-embedded sections into xylene, anhydrous ethanol, alcohol in turn, then added an appropriate amount of 0.01 M EDTA buffer to repair, added 3% hydrogen peroxide dropwise, and dried the slides with an absorbent paper. We added diluted normal goat serum dropwise for 30 minutes at room temperature, then added diluted primary antibodies (GluN2B: ab28373, PSD95: ab18258; Abcam) dropwise and incubated overnight at 4°C in a wet box after adding the primary antibodies. The tissue sections were thrice washed with phosphate-buffered saline (PBS), dried with an absorbent paper, and incubated for 30 minutes at 37°C. The tissue sections were rinsed 4 times with PBS, freshly prepared 3, 3′-diaminobenzidine color development solution was added dropwise, and the tissue sections were rinsed with tap water after color development. The tissue sections were restained with Harris hematoxylin solution, washed with water, fractionated with 1% hydrochloric acid alcohol, and then returned to blue with PBS dip wash. After rinsing the tissue sections with water, they were sequentially placed in alcohol, anhydrous ethanol, and xylene, they were air-dried in a fume hood, with drops of neutral gum placed next to the tissue and covered with coverslips, and the sealed sections were placed flat in a fume hood to dry.
Golgi Staining
The brain tissues were placed in a fixed solution for more than 48 hours. We cut the brain tissues into 2–3 mm thick pieces, gently rinsed the brain tissues with saline water for several times, placed them in a 45 mL round-bottomed EP tube, and added Gorky’s dye to completely submerge the brain tissues. Afterward, we put them into a cool and ventilated place to avoid light treatment. After 48 hours of immersion, we replaced the old dye with a new dye, and then replaced with a new dye every 3 days for a total of 14 days. On the 15th day, the brain tissues were thrice washed with distilled water, and then 80% glacial acetic acid was poured to submerge the tissues overnight. We cut the tissues into 100 μm using an oscillating tissue slicer, affixed them to the gelatin slides, and air-dried them overnight in the dark. The dried tissue slides were treated with concentrated ammonia for 15 minutes, washed with distilled water for 1 minute, dried, and sealed with glycerin gelatin.
Statistical Analysis
Statistical analysis was conducted by SPSS 22.0 software (IBM, Armonk, NY). The differences in gender between the 2 groups (MDD and HC groups) were compared using the Chi-square test. Other demographic and clinical characteristics were compared using the independent-samples t test. Data from RT-qPCR, Western blotting, and IHC were compared using the independent-samples t test between the 2 groups. Data from the open field test, forced swimming test, sucrose preference test, weight test, and MWM test were compared using the independent-samples t test between the 2 groups. p-Values were adjusted using false discovery rate correction via the Benjamini-Hochberg procedure. p-Values < .05 were considered statistically significant.
Results
Demographics and Clinical Characteristics
Patients’ demographic and clinical characteristics were collected from 20 MDD patients and 20 HCs, and it was found that HAMD-24 scores were higher in MDD patients than those in HCs (Table 1). These results showed that the depression degree in MDD patients was higher than that in HCs.
Demographics and Clinical Characteristics
| Variables | MDD | HC | X2/t | Adjusted p |
|---|---|---|---|---|
| Gender (male/female) | 6/14 | 7/13 | 0.114 | .736 |
| Age (years) | 33.90 ± 7.10 | 33.20 ± 6.01 | 0.336 | .739 |
| Age of onset (years) | 32.10 ± 7.19 | NA | ||
| Duration of disease (months) | 18.40 ± 8.33 | NA | ||
| Number of episodes | 2.05 ± 0.68 | NA | ||
| Educational level (years) | 16.00 ± 2.51 | 16.35 ± 2.53 | −0.438 | .664 |
| HAMD-24 (scores) | 26.10 ± 3.65 | 3.25 ± 1.20 | 5.437 | <.001* |
| MoCA (scores) | 24.95 ± 1.43 | 29.05 ± 0.68 | −11.548 | <.001* |
| Variables | MDD | HC | X2/t | Adjusted p |
|---|---|---|---|---|
| Gender (male/female) | 6/14 | 7/13 | 0.114 | .736 |
| Age (years) | 33.90 ± 7.10 | 33.20 ± 6.01 | 0.336 | .739 |
| Age of onset (years) | 32.10 ± 7.19 | NA | ||
| Duration of disease (months) | 18.40 ± 8.33 | NA | ||
| Number of episodes | 2.05 ± 0.68 | NA | ||
| Educational level (years) | 16.00 ± 2.51 | 16.35 ± 2.53 | −0.438 | .664 |
| HAMD-24 (scores) | 26.10 ± 3.65 | 3.25 ± 1.20 | 5.437 | <.001* |
| MoCA (scores) | 24.95 ± 1.43 | 29.05 ± 0.68 | −11.548 | <.001* |
Notes: p-Values were adjusted using false discovery rate correction. Adjusted p-value for gender was obtained by the Chi-square test. Adjusted p-values for age, educational level, HAMD-24, and MoCA were obtained by the independent-samples t test. Values for MDD and HC were presented as mean ± SD. N = 20 in each group. MDD = major depressive disorder; HC = healthy control; HAMD = Hamilton Depression Scale; MoCA = Montreal cognitive assessment; SD = standard deviation.
*Adjusted p-value < .05.
Demographics and Clinical Characteristics
| Variables | MDD | HC | X2/t | Adjusted p |
|---|---|---|---|---|
| Gender (male/female) | 6/14 | 7/13 | 0.114 | .736 |
| Age (years) | 33.90 ± 7.10 | 33.20 ± 6.01 | 0.336 | .739 |
| Age of onset (years) | 32.10 ± 7.19 | NA | ||
| Duration of disease (months) | 18.40 ± 8.33 | NA | ||
| Number of episodes | 2.05 ± 0.68 | NA | ||
| Educational level (years) | 16.00 ± 2.51 | 16.35 ± 2.53 | −0.438 | .664 |
| HAMD-24 (scores) | 26.10 ± 3.65 | 3.25 ± 1.20 | 5.437 | <.001* |
| MoCA (scores) | 24.95 ± 1.43 | 29.05 ± 0.68 | −11.548 | <.001* |
| Variables | MDD | HC | X2/t | Adjusted p |
|---|---|---|---|---|
| Gender (male/female) | 6/14 | 7/13 | 0.114 | .736 |
| Age (years) | 33.90 ± 7.10 | 33.20 ± 6.01 | 0.336 | .739 |
| Age of onset (years) | 32.10 ± 7.19 | NA | ||
| Duration of disease (months) | 18.40 ± 8.33 | NA | ||
| Number of episodes | 2.05 ± 0.68 | NA | ||
| Educational level (years) | 16.00 ± 2.51 | 16.35 ± 2.53 | −0.438 | .664 |
| HAMD-24 (scores) | 26.10 ± 3.65 | 3.25 ± 1.20 | 5.437 | <.001* |
| MoCA (scores) | 24.95 ± 1.43 | 29.05 ± 0.68 | −11.548 | <.001* |
Notes: p-Values were adjusted using false discovery rate correction. Adjusted p-value for gender was obtained by the Chi-square test. Adjusted p-values for age, educational level, HAMD-24, and MoCA were obtained by the independent-samples t test. Values for MDD and HC were presented as mean ± SD. N = 20 in each group. MDD = major depressive disorder; HC = healthy control; HAMD = Hamilton Depression Scale; MoCA = Montreal cognitive assessment; SD = standard deviation.
*Adjusted p-value < .05.
Comparison of Cognitive Function Scores and Levels of Metal Ions Between MDD Patients and HCs
The Montreal cognitive assessment (MoCA) scale was used to assess the cognitive function of 20 MDD patients and 20 HCs, and it was found that the total MoCA scale scores (Table 1), memory scores, and attention scores (Figure 1A) in MDD patients were lower than those in HCs. In addition, MDD patients’ cognitive function was reduced, mainly in memory and attention domains.
Comparison of cognitive function scores and levels of metal ions between MDD group and HC group. (A) Comparison of cognitive function scores between the 2 groups. (B) Comparison of levels of metal ions between the 2 groups. (C) Correlations among copper levels and HAMD-24 scores in MDD patients. (D) Correlations among copper levels and memory scores in MDD patients. N = 20 in each group. All data were presented as mean ± SD. *p < .05. MDD = major depressive disorder; HC = healthy control; HAMD = Hamilton Depression Scale; SD = standard deviation.
Serum levels of metal ions were compared between MDD patients and HCs, and it was revealed that serum copper levels were elevated in MDD patients compared with those in HCs, and serum levels of magnesium and calcium were reduced in MDD patients compared with those in HCs (Figure 1B). The abnormal serum levels of copper, magnesium, and calcium might have effects on the MDD development.
Correlations Among Levels of Metal Ions, Cognitive Function Scores, and HAMD-24 Scores in MDD Patients
Pearson correlation analysis of serum levels of metal ions and HAMD-24 scores in 20 MDD patients indicated a significant positive correlation between serum copper levels and HAMD-24 scores (Figure 1C). Pearson correlation analysis of serum levels of metal ions with memory scores and attention scores in 20 MDD patients revealed a significant negative correlation between serum copper levels and memory scores (Figure 1D). These data suggested that with rising copper levels, the degree of depression was aggravated, and memory declined in MDD patients. However, serum levels of magnesium and calcium had no significant correlations with depression and cognitive function in MDD patients.
Comparison of Depressive-Like Behaviors Between Stress Mice and Naive Mice
We adopted the chronic restraint stress paradigm on C57BL/6J mice to induce depressive-like behaviors, as well as C57BL/6J wild-type mice as the naive group. We used the open field test, forced swimming test, sucrose preference test, and weight test to detect depressive-like behaviors, and the MWM test was employed to evaluate the spatial memory ability of mice.
In the open field test, the center distance, center duration, and the percentage of center distance to total distance were significantly reduced in stress mice than those in naive mice (Figure 2A–C). In the forced swimming test, it was found that stress mice stayed significantly longer in water than naive mice (Figure 2D). In the sucrose water preference test, the sucrose preference rate of stress mice was significantly lower than that of naive mice (Figure 2E). In the weight test, it was indicated that the weight of stress mice was significantly reduced than that of naive mice (Figure 2F). In the MWM test, it was found that with the prolongation of training days during the first 5 days of navigation training, the escape latency of mice finding the platform was shortened, while the escape latency of naive mice was significantly shorter than that of stress mice. On the sixth day of the probe test, naive mice crossed the platform more than stress mice, and the percentage of moving distance in the platform quadrant of naive mice was significantly higher than that of stress mice (Figure 2G–I).
Comparison of ethological indicators between stress mice and naive mice. (A) Center distance in the open field test. (B) Center duration in the open field test. (C) Percentage of center distance to total distance in the open field test. (D) Stationary time in the forced swimming test. (E) Sucrose preference rate in the sucrose preference test. (F) Weight test. (G) Escape latency in the Morris water maze test. (H) The time of mice staying in the target quadrant in the Morris water maze test. (I) The number of mice crossed the target platform in the Morris water maze test. N = 6 in each group. All data were presented as mean ± SD. *p < .05. NW = Northwest; SD = standard deviation.
It was found that stress mice had depressive-like behaviors with the chronic restraint stress paradigm, and their spatial memory ability more significantly decreased compared with naive mice.
Stress Mice Had High Levels of Copper, Decreased Levels of GluN2B and PSD95, and Reduced Dendritic Spines of Hippocampal Neurons Versus Naive Mice
The results of complexation colorimetry demonstrated that copper levels in the serum and hippocampus of stress mice were higher than those in naive mice (Figure 3A). Levels of GluN2B and PSD95 in the hippocampus were evaluated based on RT-qPCR, Western blotting, and IHC between stress mice and naive mice. Levels of GluN2B (Figure 3B) and PSD95 (Figure 3C) were reduced in hippocampus of stress mice compared with those in naive mice.
Stress mice showed high levels of copper, decreased GluN2B and PSD95 levels, and reduced dendritic spines in hippocampus versus in naive mice. (A) Stress mice showed high levels of copper in serum and hippocampus versus in naive mice. (B) Stress mice exhibited the decreased GluN2B levels in hippocampus than those in naive mice. (C) Stress mice showed decreased PSD95 levels in hippocampus than those in naive mice. (D) Stress mice had decreased dendritic spines in hippocampus that in naive mice. N = 6 in each group for copper levels and RT-qPCR. N = 3 in each group for Western blotting, immunohistochemistry, and dendritic spines. All data were presented as mean ± SD. *p < .05. HP = hippocampus; GluN2B = N-methyl-D-aspartic acid receptor subunit 2B; PSD95 = postsynaptic density protein 95; RT-qPCR = quantitative reverse transcription-polymerase chain reaction; SD = standard deviation.
Golgi staining was performed to analyze the dendritic spines of hippocampal neurons. It was found that dendritic spines of hippocampal neurons decreased in stress mice compared with naive mice (Figure 3D).
It was revealed that high levels of copper could play an important role in the decreased GluN2B levels, PSD95 levels, and dendritic spines of hippocampal neurons in stress mice.
Copper Inhibitor Elevated Levels of GluN2B and PSD95 and Increased Dendritic Spines of Hippocampal Neurons of Stress Mice
In order to understand the copper interactions, copper inhibitor, TM, was injected into stress mice by gavage to reduce copper levels in hippocampus. First, it was confirmed that copper levels in serum and hippocampus were reduced with the administration of copper inhibitor (TM; Figure 4A). Next, levels of GluN2B and PSD95 in hippocampus were measured using RT-qPCR, Western blotting, and IHC in the 2 groups. The results showed that copper inhibitor (TM) increased the levels of GluN2B (Figure 4B) and PSD95 (Figure 4C) in hippocampus of stress mice.
Copper inhibitor elevated GluN2B and PSD95 levels and increased dendritic spines in hippocampus of stress mice. (A) Copper levels in serum and hippocampus were reduced in stress mice with the use of TM. (B) GluN2B levels were risen in stress mice with the use of TM. (C) PSD95 levels were elevated in stress mice with the use of TM. (D) Increased dendritic spines in hippocampus of stress mice with the use of TM. N = 6 in each group for copper levels and RT-qPCR. N = 3 in each group for Western blotting, immunohistochemistry, and dendritic spines. All data were presented as mean ± SD. *p < .05. HP = hippocampus; TM = tetrathiomolybdate; GluN2B = N-methyl-D-aspartic acid receptor subunit 2B; PSD95 = postsynaptic density protein 95; RT-qPCR = quantitative reverse transcription-polymerase chain reaction; SD = standard deviation.
Regarding the copper interactions, the changes of dendritic spines of hippocampal neurons were assessed. It was found that copper inhibitors increased the dendritic spines of hippocampal neurons in stress mice (Figure 4D).
It was obvious that high levels of copper inhibited levels of GluN2B and PSD95, and caused damage to dendritic spines of hippocampal neurons in stress mice.
Copper Inhibitor Relieved Depressive-Like Behaviors and Improved Spatial Memory Ability of Stress Mice
In order to verify the effects of copper inhibitor on behaviors of mice, open field test, forced swimming test, sucrose preference test, and MWM test were performed. It was revealed that copper inhibitor relieved depressive-like behaviors and improved spatial memory ability of stress mice.
In the open field test, the center distance, center duration, and the percentage of center distance to total distance were significantly risen in stress mice with TM than those in stress mice without TM (Supplementary Figure A–C). In the forced swimming test, it was found that stress mice with TM stayed significantly shorter in water than stress mice without TM (Supplementary Figure D). In the sucrose water preference test, the sucrose preference rate in stress mice with TM was significantly higher than that in stress mice without TM (Supplementary Figure E). In the weight test, it was revealed that the weight of stress mice with TM significantly increased than that of stress mice without TM (Supplementary Figure F). In the MWM test, it was indicated that with the prolongation of training days during the first 5 days of navigation training, the escape latency of mice finding the platform was shortened, while the escape latency of stress mice with TM was significantly shorter than that of stress mice without TM. On the sixth day of probe test, stress mice with TM crossed the platform more than stress mice without TM, and the percentage of moving distance in platform quadrant of stress mice with TM was significantly higher than that of stress mice without TM (Supplementary Figure G–I).
The earlier-mentioned results showed that high levels of copper aggravated the depressive-like behaviors and the decline of spatial memory ability in stress mice.
Discussion
To the best of our knowledge, cognitive dysfunction is a common clinical manifestation in MDD patients. In this study, it was found that the memory scores of MDD patients were significantly lower than those of HCs. Moreover, high levels of copper and decreased memory scores exhibited a significant correlation in MDD patients. Copper may provide promising insights into the mechanism of depression-related memory disorders.
Memory is a multicomponent psychological process, including memorization, recognition, recollection, etc. The frontal lobe, hippocampus, and thalamus are involved in memory activities, and the damaged neurons in hippocampus mainly indicate memory disorders (19). MDD patients suffer from word learning and memory loss, especially short-term memory ability and instantaneous memory ability (20). A study of 66 MDD patients showed that memory ability was significantly reduced compared with HCs (21). A previous study showed that memory is the susceptible domain to decline among cognitive domains (22). However, there is still the lack of evidence to explore the molecular mechanisms of depression-related memory disorders.
The evidence showed that the levels of metal ions were abnormal in neurological diseases (23). Metal ions play an important role in glutamic transmitter transmission between neurons, affecting the synaptic connection of excitatory glutamic neurons and causing changes in cognitive functions (24). In this study, levels of metal ions were compared between 20 MDD patients and 20 HCs. There were significant differences in levels of copper, magnesium, and calcium between the 2 groups (Figure 1B). In conclusion, it was revealed that abnormal levels of metal ions could play a role in the potential pathogenesis of depression-related memory disorders.
To further explore the functions of metal ions in depression-related memory disorders, correlation analysis was performed between levels of metal ions and cognitive scores. Importantly, it was found that a high level of copper was significantly correlated with the decreased memory score in MDD patients (Figure 1D). The copper level in neuronal gray matter is twice that in the fibrous white matter of the brain (25). High levels of copper could cause toxic effects and damage to brain functions (26). No significant correlation was found among serum magnesium levels, calcium levels, memory scores, and attention scores in MDD patients in the present study. It was revealed that the damage of high-level copper was mainly concentrated in the memory domain of a cognitive function. These results may be valuable to deeply investigate the effects of copper levels on the pathogenesis of depression-related memory disorders.
In this study, the chronic restraint stress paradigm was utilized to induce depressive-like behaviors in mice, and the effects of copper levels on depression were explored. High levels of copper in hippocampus were also detected in stress mice (Figure 3A). Moreover, decreased dendritic spines of hippocampal neurons in stress mice were correlated with decreased memory ability (Figure 3D). To reconfirm the effects of high levels of copper on memory ability, a copper inhibitor was used to explore the copper interactions in the changes of dendritic spines in hippocampal neurons. The results showed that the copper inhibitor could effectively reduce copper levels (Figure 4A), and reversed dendritic spines of hippocampal neurons in stress mice (Figure 4D). Therefore, high levels of copper could play an important role in the decreased dendritic spines of hippocampal neurons.
In order to explore the mechanism of copper levels on dendritic spines in depression, we concentrated on neurotransmitter disorders in the hippocampus. It is broadly accepted that disorders of glutamate metabolism occur in response to infrequent intracerebral transmitters and may involve a range of pathophysiological processes of depression (27). GluN2B mediates glutamic transmission across the postsynaptic membrane (28). PSD95 is a structural protein of glutamatergic synapses and a key regulator of dendritic spines in vivo (29). Therefore, GluN2B and PSD95 are requisite for the transmission of glutamic acid from synaptic gap to the postsynaptic membrane. It was reported that the changed levels of glutamic receptors and PSD95 might affect the synaptic signaling, which could be involved in the depression process (30). As illustrated in Figure 3B and C, GluN2B and PSD95 levels in hippocampus were reduced in stress mice compared with those in naive mice. With the administration of copper inhibitor, the levels of GluN2B and PSD95 in hippocampus were significantly elevated in stress mice (Figure 4B and C). These results indicated that high levels of copper suppressed GluN2B and PSD95 levels in hippocampus, and could be involved in the regulation of dendritic spines of hippocampal neurons in depression. PSD95 is a scaffolding protein that binds to GluN2B in glutamatergic postsynaptic densities and stabilizes receptor regulation of synaptic function (31). Our findings could provide interactions of high levels of copper for exploration of the potential mechanisms of depression-related memory disorders, and contribute to find potential therapeutic targets for depression.
This is the valuable study to evaluate the effects of high levels of copper on the development of depression-related memory disorders. Nevertheless, some potential limitations of the present study should be pointed out. First, the sample size of MDD patients and HCs was relatively small, and this might restrict the generalizability of the results, indicating the necessity of further study with a larger sample size. Second, 1 sample source in a specific period might limit the results. Additional sample sources from diverse periods using the same molding technique are required to reconfirm outcomes. Finally, the interactions of high levels of copper have still remained complex, and our research might not comprehensively reveal the detailed mechanism of copper in glutamate metabolism in MDD. Further exploration is essential based on excluding aforementioned factors.
Conclusion
In summary, it was found that the memory scores of MDD patients were significantly lower than those of HCs. Moreover, high levels of copper and decreased memory scores exhibited a significant correlation in MDD patients. Further exploration of the mechanisms of high levels of copper in depression-related memory disorders is necessary. We used chronic restraint stress paradigm to induce depressive-like behaviors in mice, and found that high levels of copper suppressed GluN2B and PSD95 levels in hippocampus, which could be involved in the regulation of dendritic spines of hippocampal neurons in depression. These preliminary data suggested that high levels of copper could provide a promising direction for the potential pathogenesis of depression-related memory disorders, and even be targeted for therapeutic manipulation. However, further large-scale study is required to better understand the effects of copper levels on the pathogenesis of depression-related memory disorders.
Funding
This study was supported by the National Natural Science Foundation of China (No: 81860214), Guangzhou Key Discipline of Medical (Geriatric Medicine, No: ZDXK202103), and Basic Research Program and Applied Basic Research Project of Guangzhou City, China (No: 202301010449 and 202201010091).
Conflict of Interest
None declared.
References
Author notes
These authors contributed equally to this work.
