Dlg1 Knockout Inhibits Microglial Activation and Alleviates Lipopolysaccharide-Induced Depression-Like Behavior in Mice

Microglia-mediated neuroinflammation is widely perceived as a contributor to numerous neurological diseases and mental disorders including depression. Discs large homolog 1 (Dlg1), an adaptor protein, regulates cell polarization and the function of K+ channels, which are reported to regulate the activation of microglia. However, little is known about the role of Dlg1 in microglia and the maintenance of central nervous system homeostasis. In this study, we found that Dlg1 knockdown suppressed lipopolysaccharide (LPS)-induced inflammation by down-regulating the activation of nuclear factor-κB signaling and the mitogen-activated protein kinase pathway in microglia. Moreover, using an inducible Dlg1 microglia-specific knockout (Dlg1flox/flox; CX3CR1CreER) mouse line, we found that microglial Dlg1 knockout reduced the activation of microglia and alleviated the LPS-induced depression-like behavior. In summary, our results demonstrated that Dlg1 plays a critical role in microglial activation and thus provides a potential therapeutic target for the clinical treatment of depression.


Introduction
Major depressive disorder (MDD) is a severe psychiatric condition that affects nearly 15% of people worldwide [1], and it significantly increases the risk of suicide [2]. However, the pathophysiology of MDD is yet to be completely elucidated. Neuroinflammation is increasingly being accepted as a risk factor for emotional disorders, especially depression [3,4]. Several clinical and preclinical studies have revealed that signatures of neuroinflammation, including activated microglia and inflammatory mediators, are commonly found in patients with depression [4,5].
Microglia, the resident immune cells of the central nervous system (CNS), play indispensable roles in both health and disease [6]. Upon neuroinflammation, microglia are the first activated cells in the brain parenchyma [7]. They not only participate in the processes of neurodegenerative diseases [8][9][10] but are also involved in psychiatric disorders [11,12]. Evidence has shown that microglial status is strongly associated with depression. Endotoxin stimulation can activate microglia and thus cause negative affective disorders [13], and microglia-derived pro-inflammatory cytokines are correlated with depression-like behaviors [14]. Furthermore, by targeting serotonin (5-HT), antidepressants inhibit depression-induced microglial activation and cytokine production, and alleviate depression-like symptoms [15][16][17]. Given that microgliainduced neuroinflammation is an important contributor to MDD, a major goal is to understand the precise underlying mechanism of microglial function in MDD. Discs large homolog 1 (Dlg1, also known as SAP-97) is a member of the membrane-associated guanylate kinase family [18]. It is an adaptor protein that regulates cell polarity, neuronal Kv1 channel localization, and Kv1.3 channel function [19,20]. Moreover, the Kv1.3 potassium channel has been reported to regulate the function of microglia, and selectively blocking Kv1.3 channels inhibits microglial activation [21]. In the CNS, Dlg1 interacts with AMPA (a-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid) and NMDA (N-methyl-D-aspartic acid) receptors when they are sorted and transported from the soma to synapses [22]. In the peripheral immune system, Dlg1 is involved in the activation of dendritic cells and T cells, and knockout of Dlg1 inhibits the release of inflammatory cytokines [23,24]. In addition, Dlg1 contributes to neuropsychiatric disorders, and is regarded as a risk factor for schizophrenia [25]. Despite the multiple roles of Dlg1, its roles in microglial activation and in MDD remain to be clarified.
In this study, we demonstrated that Dlg1 regulates microglial activation by targeting nuclear factor-jB (NF-jB) signaling and the mitogen-activated protein kinase (MAPK) pathway. In vivo, conditional knockout of microglial Dlg1 inhibited microglial activation, decreased inflammatory cytokine levels, and alleviated depressionlike behaviors in mice, providing a potential therapeutic strategy to treat or slow down the progression of depression.

Materials and Methods
Mice CX3CR1 CreER mice were purchased from the Jackson Laboratory (Bar Harbor, USA), and Dlg1 flox/flox mice were gifted by Dr. Wanli Liu (Tsinghua University, China). The generation of the Dlg1 flox/flox mouse has been described [26]. To generate microglia-specific knockout mice, homozygous Dlg1 flox/flox mice were crossed with mice expressing tamoxifen (TAM)-inducible Cre-recombinase under the control of the CX3CR1 promoter (CX3CR1 CreER mice [27,28]). Mice were given TAM by intragastric administration at the age of 6 weeks to induce microgliaspecific knockout. All mice were housed in a specific pathogen-free environment at the Animal Care Facility in our institute. All animal experiments were approved by the Institutional Animal Care and Use Committee of the Beijing Institute of Basic Medical Sciences.

Cell Culture and Transfection
All procedures were performed as previously described [29]. Briefly, BV2 and HEK293T cell lines were obtained from the American Type Culture Collection (Manassas, USA) and maintained in Dulbecco's modified Eagle's medium (Invitrogen, Waltham, USA) supplemented with 10% fetal bovine serum (Gibco, Grand Island, USA) and 1% penicillin-streptomycin (Invitrogen). All cells were maintained in a 5% CO 2 atmosphere at 37°C.

Plasmid Constructs and Selection of Cells with Stable Expression
Full-length Dlg1 was amplified from a mouse cDNA library by PCR and inserted into the pCDH-CMV-MCS-EF1-Puro expression vector (Sigma-Aldrich, St Louis, USA) using the EcoRI and NotI restriction sites. To generate cells that stably overexpressed Dlg1, BV2 cells were infected with the retroviral vector pCDH encoding Dlg1 and selected with puromycin.

Drugs and Treatment
Lipopolysaccharide (LPS; Sigma-Aldrich) was dissolved in saline (0.9% NaCl), and injected intraperitoneally at 1 mg/ kg. Tamoxifen (#S1238, Selleckchem, Houston, USA) was dissolved in corn oil. Mice at the age of 6 weeks were given a total dose of 20 mg tamoxifen intragastrically for three consecutive days to induce microglial Dlg1-specific knockout.

Dual-Luciferase Reporter System
The NF-jB reporter was generated in our laboratory. Briefly, the NF-jB promoter was cloned into a pGL3luciferase reporter vector (Promega, Madison, USA). The pCMV-Renilla plasmid and NF-jB reporter were cotransfected into HEK293T cells using Lipofectamine 2000 transfection reagent (#11668019, Invitrogen). Sixteen hours after transfection, the cells were lysed and luciferase activity was measured using a dual-luciferase reporter detection system (Promega).

Open Field Test (OFT)
The mice were kept in the test room for 2 h in advance for preconditioning. During the experiment, each mouse was placed in a 32 9 32 cm 2 test box. The test program was started to record the spontaneous activity of the mouse during 5 min and to count the total distance travelled and the time spent in the central area.

Elevated Plus Maze (EPM)
The mice were kept in the test room for 2 h in advance for preconditioning. During the test, each mouse was gently placed in the center area facing the open arm. The recording software was then started to record the activity time and the number of entries into the open arm during 5 min.

Tail Suspension Test (TST)
The mice were kept in the test room for 2 h in advance for preconditioning. We stuck medical tape to the tail tip, and the mouse was hung on the instrument with a clip. The experimental time was 6 min and the immobility time of each mouse was recorded in the last 4 min of the test.

Forced Swimming Test (FST)
The day before the experiment, the mice were allowed to swim in water for 5 min. On the day of the experiment, the mice were kept in the test room for preconditioning. During the test, each mouse was placed in a beaker (volume 3 L) filled with water at 18-21°C. The total test time was 6 min. The immobility time was recorded during the last 4 min of the test.

Statistical Analysis
Data are expressed as the mean ± SEM from at least two independent experiments. Statistical differences between the test and control values were analyzed using Student's ttest. For multiple comparisons, statistical differences were analyzed by applying ordinary one-way ANOVA (Tukey's multiple comparison test). Data were considered significant as follows: *P \ 0.05, **P \ 0.01, and ***P \ 0.001. Statistical analysis was performed using GraphPad Prism (version 6.07, GraphPad, San Diego, USA).

Knockdown of Dlg1 in Microglia Inhibits LPSinduced Inflammation
LPS is a commonly-used agonist that induces the activation of microglia. This process is characterized by the overwhelming release of pro-inflammatory cytokines [30,31]. To investigate the role of Dlg1 in microglial activation, we exposed BV2 cells with Dlg1 knockdown to LPS (1 lg/ mL) at different time points (Fig. 1A). Western blot analysis showed that the iNOS protein level was significantly lower in Dlg1-knockdown cells than in control cells (Fig. 1B, C). In addition, knockdown of Dlg1 inhibited the up-regulation of pro-inflammatory cytokines, including tumor necrosis factor (TNF-a), interleukin-1 beta (IL-1b), and IL-6 ( Fig. 1D-H). These results suggest that Dlg1 is involved in LPS-induced inflammation, and knockdown of Dlg1 in microglia significantly reduces inflammatory cytokine levels.
To support our findings, we overexpressed Dlg1 in BV2 cells, and these cells were also exposed to LPS at different time points. Consistently, both the iNOS protein level and the cytokine levels were significantly increased when Dlg1 was overexpressed (Fig. 1I-N). Taken together, these results indicate that Dlg1 plays an important role in microglia-mediated inflammation.

Dlg1 Participates in Microglia-mediated Inflammation by Targeting NF-jB Signaling and the MAPK Pathway
It has been well corroborated in previous studies that LPS is an activator of NF-jB signaling and the MAPK pathway [30,32]. To study the mechanism underlying Dlg1 and microglia-mediated inflammation, we used a short-term LPS challenge, and found that the phosphorylation levels of various pathway proteins declined markedly in Dlg1knockdown BV2 cells, including P-IKKa/b, P-IjBa, P-JNK, and P-P38 ( Fig. 2A). Correspondingly, overexpression of Dlg1 significantly increased the levels of these proteins (Fig. 2B). These results suggest that Dlg1 participates in microglia-mediated inflammation by regulating NF-jB signaling and the MAPK pathway. Since alterations in Dlg1 affect the protein levels of NF-jB signaling, we then asked whether Dlg1 affects its activation. Using a dual-luciferase reporter system, we found that the NF-jB promoter luciferase showed enhanced activity upon overexpression of Dlg1 (Fig. 2C). Collectively, these results indicate that Dlg1 regulates NF-jB signaling and the MAPK pathway and thus plays an important role in microglia-mediated inflammation.
Previous studies have shown that Dlg1 is an adaptor protein of the Kv1.3 potassium channel by ensuring K ? outflow and subsequent Ca 2? inflow events [22,23]. To determine whether Dlg1 regulates microglia-mediated inflammation depending on the activity of K ? channels, we applied a serial concentration gradient of K ? culture medium to inhibit K ? channel activity [33] and found that the decreased levels of p-IKKa/b, P-JNK, and P-P38 regulated by Dlg1 knockdown diminished as the concentration of K ? increased (Fig. 2D). Furthermore, the levels of the inflammatory cytokines TNF-a and IL-6 were significantly inhibited when the K ? channel activity was blocked by high concentrations of K ? (Fig. 2E, F). Together, these results indicate that Dlg1 regulates microglia-mediated inflammation in a K ? -dependent manner.

Microglial Dlg1 Knockout Alleviates Depression-like Behavior
Given that Dlg1 knockdown significantly reduced the inflammatory response in microglia, we next sought to explore the role of Dlg1 in depression. We generated mice with inducible microglia-specific knockout of Dlg1 (Dl-g1 flox/flox ; CX3CR1 CreER , abbreviated as cKO mice), and used tamoxifen to induce Cre-recombinase activity in microglia (Fig. 3A). Compared to Dlg1 flox/flox mice, the Dlg1 level in cKO mice decreased significantly (Fig. 3B). The LPS-induced depression model has been commonly used in studies based on the causal link between neuroinflammation and depression [34][35][36]. In our experiments, we applied an LPS challenge after tamoxifen administration to induce depression-like behavior in Dlg1 flox/flox and cKO mice as assessed by behavioral tests (Fig. 3A). As shown in  suggesting that microglial Dlg1 knockout does not impair LPS-induced anxiety-like behaviors.
However, in the TST, cKO mice had a shorter immobility time than Dlg1 flox/flox mice, suggesting that microglial Dlg1 knockout alleviates the depression-like behavior (Fig. 3I, J). To further characterize the role of Dlg1, we used the FST. Consistent with the results in the TST, microglial Dlg1 knockdown significantly decreased the immobility time, suggesting an improvement in depression-like behavior (Fig. 3K, L). Together, these results demonstrate that deletion of microglial Dlg1 alleviates inflammation-induced depression-like behaviors in mice.

Dlg1 Knockout Inhibits Microglial Activation In Vivo
We next investigated why Dlg1 knockout alleviates depression-like behavior and how microglia change in cKO mice. To address this, we performed IBA1 immunofluorescence staining and found that microglia were robustly activated in the brain of mice injected with LPS, including changes in density and morphology (Fig. 4A, B). Microglial Dlg1 knockout significantly rescued the LPS-induced reduction in branch number and length, as well as increasing the soma area (Fig. 4C-F). Furthermore, LPS injection increased microglial numbers in the hippocampus of Dlg1 flox/flox mice, and Dlg1 microglial conditional knockout significantly reduced this increase (Fig. 4G, H), with no effect on astrocyte number (Fig. 4G, I). These results suggest that Dlg1 knockout in microglia markedly inhibits microglial activation in the mouse brain.
To determine whether inflammatory cytokines also declined, we measured the protein levels of iNOS and IBA1 in the hippocampus and found that upon LPS challenge they were significantly reduced in cKO mice, with a decreasing trend of GFAP levels compared with Dlg1 flox/flox mice (Fig. 5A-D). Moreover, we found that the inflammatory cytokines TNF-a and IL-6 were significantly reduced in cKO mice (Fig. 5E, F), with a decreasing trend of IL-1b levels (Fig. 5G).
These results collectively indicate that microglial conditional knockout of Dlg1 inhibits microglial activation, and thus alleviates the depression-like behavior induced by neuroinflammation (Fig. 5H).  J Immobility times in the TST. K Cartoon of forced-swimming test (FST). L Immobility times in the FST. Data are presented as the mean ± SEM; *P \ 0.05; **P \ 0.01, ***P \ 0.001, two-tailed unpaired Student's t-test.

Discussion
As a representative of bacterial infection, systemic LPS injection (1 mg/kg) causes time-dependent behavioral changes in mice. Pro-inflammatory cytokines in the blood peak at *2 h, and depression-like behaviors occur 24 h later [14,[37][38][39]. In this study, we used LPS injection to create a depression-like mouse model. Microglial Dlg1 deletion ameliorates the symptoms of depression in mice, and the hippocampus is relatively vulnerable to acute inflammatory attack during depression [34]. Our results demonstrated that Dlg1 knockout reduced microglial activation and inflammatory cytokine levels in the hippocampus, suggesting that Dlg1 functions as an important regulator of inflammation-induced depression.
Neuroinflammation is involved in the regulation of motivational and emotional states in various pathological processes. The increased expression and release of inflammatory mediators, generally manifested as pro-inflammatory cytokines, acute phase proteins, chemokines, and adhesion molecules, are related to human depressive symptoms and negative affective states in mice [40][41][42][43][44][45]. Elevated levels of IL-6, IL-1b and TNF-a are found in both peripheral blood and cerebrospinal fluid [40,[46][47][48][49].
Moreover, the concentrations of these cytokines in plasma are positively correlated with the severity of depressive symptoms [41,50]. Furthermore, previous studies have revealed that functional allelic variants of the IL-1b and TNF-a genes increase the risk of depression and are associated with decreased responsiveness to antidepressant therapy [51,52]. Mechanistically, the increased proinflammatory cytokines mediated by NF-jB signaling [53] cause long-term alterations in neuropeptide and neurotransmitter synthesis, leading to metabolic changes in serotonin, dopamine, and other related neurotransmitters Data are presented as the mean ± SEM; *P \ 0.05, **P \ 0.01, ***P \ 0.001, two-tailed unpaired Student's t-test. [54,55]. Moreover, the MAPK pathway is associated with dysregulation of the hypothalamic-pituitary-adrenal axis, one of the strongest biological correlates of MDD [56,57]. Here, we found that the NF-jB signaling and MAPK pathways were involved in microglia-mediated inflammation. Knockout of Dlg1 in microglia inhibited the activation of the NF-jB signaling and MAPK pathways, and markedly alleviated depression-like behaviors. Numerous studies have revealed that voltage-gated K ? channels are involved in microglial activation [58,59]. In this study, we found that the effects of Dlg1 were partially rescued by incubation with high concentrations of K ? , suggesting that K ? flux is involved in the effect of Dlg1 in microglia. However, the target and the regulatory mechanism of Dlg1 in microglial activation need to be further investigated.
Microglia are resident immune cells of the CNS. Our recent studies showed that inhibition of microglial activation-induced neuroinflammation alleviates brain injury in multiple neurodegenerative diseases and psychological disorders [9,10,60,61]. Moreover, a recent study showed that microglial activation regulates negative emotions through prostaglandin-dependent striatal neurons, and thus plays a key role in the development of major depression [62]. In this study, we found that knockout of Dlg1 in microglia alleviated neuroinflammation-induced depression-like behaviors in mice, providing a novel way to interfere with microglial activation and an alternative treatment strategy for depression.

Conclusions
In summary, our study demonstrated that knockdown of Dlg1 inhibits the activation of NF-jB signaling and the MAPK pathway in microglia, and consequently reduces microglial activation and the release of pro-inflammatory cytokines. Furthermore, deletion of microglial Dlg1 alleviates inflammation-induced depression-like symptoms in mice. Our findings demonstrate that Dlg1 plays a vital role in microglial activation and provides a potential therapeutic target for the treatment of depression.

Conflict of interest The authors declare no conflicts of interest.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons. org/licenses/by/4.0/.