C57 BL/6J adult male mice, aged 8–10 weeks, were provided by the Laboratory Animal Center of Southern Medical University (Guangzhou, Guangdong). All animals were raised in a single cage in a standard specific pathogen-free experimental environment (12 h light/dark cycle, with ad libitum access to dry food and clean water). The indoor temperature was maintained at 23 ± 1°C, and the humidity was maintained at 50–60%. Animal treatments, including anesthesia induction and euthanasia, were conducted in accordance with the Principle of Laboratory Animal Care (NIH Publication no. 85–23, revised 1985). All experimental procedures were conducted in accordance with the requirements of the China Animal Ethics Committee and were approved by the Animal Ethics Committee of The Third Affiliated Hospital of Sun Yat-sen University.

In addition to those included in the control group, the remaining mice were subjected to chronic unpredictable stress. The procedures for CMS model development were adapted from a previous study (Qin et al., 2019). The stress paradigm included fasting, day/night reversal, forced swimming, noise, stroboscopy, restraint stress, water deprivation, and a wet cage; these stressors were randomly subjected 2–3 times every day. The unpredictable stimuli were administered in different ways for 7 weeks. After chronic stress stimulation for 3 weeks, the CMS mice were then randomly assigned to the SSD treatment group (CMS+SSD) or nontreated CMS group (CMS). The SSD dosage was adopted from previous validated research (Xu et al., 2019; Chao et al., 2020) (M3936, Abmole, Houston, TX, United States, >98%, 1 mg/kg; dissolved in saline and administered via gavage) and the treatment lasted three weeks; in contrast, the mice in the CMS group were treated with saline. In the preliminary experiment, we had used 1 mg/kg dosage in age-matched mice to verify the potential adverse effects of SSD. The SSD treatment had little effect on depressive and sexual behaviors of the mice, after which we performed the following experiments. All experiments were strictly implemented in accordance with the 3R principles; each group had 20 mice. At the end of the CMS modeling and SSD treatment, the animals were used for behavioral, morphological, electrophysiological, and biochemical analyses. All animals were assigned a numerical code and the investigators were always blinded to the treatment groups until the completion of data analysis.

The sucrose preference test is the main method of testing the core symptom of CMS depression model-anhedonia (Cao et al., 2013). After 23 h of fasting and water abstinence, each mouse was provided with two bottles of water (weights measured in advance): one bottle contained 1% sucrose water and the other had pure water. Pure water consumption and sugar water consumption of the mice were calculated after 1 h. The percentage of sugar water consumption was calculated in terms of the total weights of liquid consumed.

The mice were placed in a swimming bucket filled with water at 24 ± 2°C, with a depth of 20 cm, for 6 min. The time the mice spent in immobility within the last 4 min was recorded. The duration of immobility was recorded when mice ceased struggling.

This experiment was conducted in a closed Y-shaped maze as previously described (Petrulis and Johnston, 1999). The maze was made of PVC plastic and consisted of three long arms and perforated plexiglass boxes were placed at the ends of two top arms. The animals were allowed to explore the animals behind the transparent perforated plexiglass boxes. The mice were allowed to move freely in the maze for 5 min, and the time spent exploring each end of the Y-maze was recorded.

In reference to research on the sexual motivation of rats (Chu and Ågmo, 2016), we established a mouse model for the sexual motivation test. The experimental device used is similar to that used in rats. The equipment consisted of a rectangular open field having two chambers with openings on the diagonally opposite walls. The front of the chamber is made of a wire mesh, which allows the animals to see, smell and touch the animals in the chamber. Male and female animals are respectively placed in the two chambers. The virtual area in front of the cage is the motivation exploration area. A camera was suspended to track the movements of the animals, and the preference score was defined by female motivation area time/(female motivation area time + male motivation area time).

The sexual interaction test was adapted from a previous study (Benelli et al., 2002). Briefly, the experiment was conducted in a quiet, dimly lit environment at nightphase. After three days of preadaptation to the experimental arena (20 min per day; arena area, 40 cm*40 cm*40 cm), the tested male mice were allowed to interact with aphrodisiac female mice (Grønli et al., 2005) for 30 min. The recorded masculine sexual behavior include the latency to first mount, mount frequency, intromission frequency, and the ratio of intromission and total mount.

Following paraformaldehyde perfusion and fixation, the mouse brains were dehydrated with 30% sucrose. After cryofixation and obtaining cryosections of the whole brain, the brain slices were incubated in 5% bovine serum albumin containing 0.5% Triton X-100 for 1 h, after which they were treated with the following primary antibodies: rabbit polyclonal anti-Iba1 (1:300; ab153696; Abcam) and rabbit polyclonal anti-glial fibrillary acid protein (GFAP) (1:300; ab7260; Abcam); the primary antibodies were incubated at overnight at 4°C. On the next day, following three washes with 0.1M phosphate-buffered saline (PBS), the slices were incubated with secondary antibody at room temperature for 1.5 h, mounted with DAPI-containing mounting medium, and sealed after washing again with PBS. Fluorescent images were acquired by confocal microscopy (Leica, Wetzlar, Germany) and analyzed using the Image J software.

The brain protein samples were extracted using RIPA buffer containing a protease inhibitor cocktail (Roche, Basel, Switzerland), and the extracted proteins were electrophoresed in a 10–12% sodium dodecyl sulfate-polyacrylamide gel (Willget, China). Thereafter, the separated proteins were transferred onto polyvinylidene difluoride membranes; the membranes were then probed overnight with the following primary antibodies: mouse polyclonal anti-GFAP (1:800; cat# 3670; Cell Signaling Technology, MA, United States), rabbit polyclonal anti-MAP2 (1:800; cat# 4542; Cell Signaling Technology) and mouse polyclonal anti-GAPDH (1:2,000, sc-365062, Santa Cruz Biotechnology, TX, USA). On the next day, after washing, the membranes were incubated with horseradish peroxidase (HRP)-conjugated secondary antibody for 1 h at room temperature. The proteins were then visualized using an enhanced chemiluminescent substrate (ECL; Thermo Fisher Scientific, IL, United States).

Oxidative stress in the MePD was evaluated by determining the generation of ROS. Quantitative measurements of ROS were performed using a double-sandwich ELISA method (ROS assay kit; QYE2656; Qualityard, Beijing, China) in accordance with the manufacturer’s instructions. The OD value of each hole was measured sequentially with a microplate reader (PerkinElmer) with the wavelength of 450 nm and the values were expressed relative to the signal of controls.

Stereotaxic surgeries were conducted as previously described (Wang et al., 2019b). The mice were anesthetized with intraperitoneal injection of 4% pentobarbital sodium (40 ml/kg) and placed on a stereotaxic apparatus (RWD, Shenzhen, China). The scalp was shaved, and the skin was disinfected with 75% alcohol. The skin was cut along the median line of the exposed bone, and the subcutaneous tissue was separated to fully expose the skull. After drilling through the skull, above the MePD (AP -1.7; ML 2.0; DV -5.0) nucleus, adeno-associated viruses (AAV) expressing the green fluorescent protein (GFP; BrainVTA, Wuhan, China) was injected into the target site.

The GFP-stained brain sections were observed under a confocal microscope (LSM510, Meta, Zeiss). For dendritic spine density analysis, z-series stacks were obtained with 10 consisted scans at a high zoom at 1 mm intervals in the Z-axis. The dendritic spines were sampled from 3–5 distal dendrites of all GFP-positive neurons. The density of dendritic spines in the MePD was measured using Image J software.

After the mice were placed on a stereotaxic apparatus and the skull was exposed, a glass electrode was slowly lowered toward the MePD to record neuronal activity. An electrical stimulation was delivered (STG4008 stimulator, Multichannel Systems, Germany) through the electrode—intensity, 0.1–0.9 mA; duration, 0.2 ms; frequency, 0.2–40 Hz. At the end of the experiments, only the data collected from the right stimulation sites were used for analyses.

The GraphPad Prism 7.0 software was used to analyze the data and create the figures. Data are presented as mean ± standard error of mean (SEM) and analyzed using a Student’s t-test or an analysis of variance, according to the different experiments. The significance level for all tests was set to p/q < 0.05.

Adult male mice underwent a seven-week CMS paradigm andthe SSD administration started from the end of the third week of CMS induction. In the sucrose preference test to evaluate anhedonia, the percentage of sucrose consumption was significantly reduced in the CMS group than in the non-stressed control group; however, SSD treatment for three weeks significantly improved sucrose consumption (Figure 1A). The results of FST revealed that the CMS group showed significantly extended duration of immobility than the non-stressed control group. In contrast, SSD administration remarkably rescued this immobility in the FST (Figure 1B). Therefore, these findings indicate that SSD demonstrated antidepressant effects in a mouse model of CMS.

In a previous study, the CMS-related rodent model had shown abnormally low sexual desire (Shen et al., 2020). In a mice Y-maze sexual preference test (Figure 2A), healthy male mice preferred to explore an estrous female mouse rather than a sexually active male mouse (Figure 2B); however, the CMS mice spent nearly the same amount of time around the male/female target arm. SSD treatment remarkably alleviated the adverse effects of CMS, with a statistically significant increase in the time spent exploring an estrous female mouse than a sexually active male mouse. In another sexual motivational test adapted from a rat study (Figure 2C), when the mice were allowed to freely explore the arena, similar to our abovementioned results, the CMS-induced male mice spent less time in the female zone than those in the control group. In addition, the SSD treatment restored female exploration behavior (Figure 2D). SSD administration improved the sexual motivation of the male CMS mice.

After the sexual motivation test, we performed the sexual performance test assay to further assess the sexual performance of male CMS mice. In the sexual interaction test (Figure 3A), the latency to mount, number of mounts, number of intromissions, and the ratio between intromission and mount were recorded. SSD-treated CMS mice showed significantly reduced latency for first mouth with female mouse than vehicle-treated mice (Figure 3B). This further validated our previous results of the sexual motivation test. Moreover, SSD treatment significantly increased the number of mounts and intromission (Figures 3C,D), as well as increased the intromission ratio of male CMS mice to mount a female mouse (Figure 3E). SSD administration improved the overall sexual performance of male CMS mice.

Similar to other amygdala subareas, the MePD mainly consists of two types of cells: neurons and glial cells. As a main glial cell type in the MePD, astrocytes are sensitive to environmental stress and may be affected in terms of their morphology and functions (Tynan et al., 2013), which can impact normal neural functions. In the MePD, we observed that the GFAP + astrocytic soma and primary processes responded sensitively to stress stimuli (Figure 4A). The animals in the CMS group had a lower GFAP + astrocyte density and shrunken astrocytic morphologies than those in the control group. The results of WB further confirmed the reduced GFAP expression in the MePD of CMS mice (Figure 4B); however, these deficits were reversed by SSD administration. Furthermore, chronic stress increased the number of Iba1-positive microglia and promote their activation in the MePD of CMS mice (Figure 4A).

Nextly, we tested whether SSD could attenuated neuroinflammation and inhibiting the oxidative stress. We performed ELISA studies to measure the IL-1β and IL-6 in the MePD immediately after the behavioral tests. CMS increased the IL-1β and IL-6 levels as compared to control group, the treatment of SSD significantly decreased the IL-1β and IL-6 levels in MePD of the mice as compared to CMS group (Figures 4C,D). CMS induces significantly increased ROS production compared with the control, however, SSD treatment effectively mitigate the elevation of the ROS levels in the MePD of the mice (Figure 4E).

Two weeks before the end of the CMS modeling, we infected MePD neurons with AAV-CMV-GFP. We measured the spine density in these local pyramidal neuronal cells that were tagged with GFP to reveal the morphological characteristics (Figure 5A). Spines protruding from second order dendrites were separately assessed for distal dendrites. The CMS mice showed significantly decreased spine density arising from the distal second order dendrites than that in the normal control mice (Figure 5B). The spines of dendrites in the SSD-treated group showed a significant reversion of the above-stated deficit. Consistent with the abovementioned observation, the expression of the neuronal marker MAP2 was decreased in CMS mice (Figure 5C), whereas MAP2 expression was significantly increased in the MePD of CMS+SSD mice.

The firing rate of MePD neurons was measured by inserting an electrode into the brain (Figures 6A,B). CMS modeling significantly decreased the firing rate of MePD neurons (Figure 6C). This decreased neuronal firing was reversed in CMS+SSD group, which indicated that the restoration of the firing properties of the MePD may underpin the effects of SSD.