Microglia activation in the mPFC mediates anxiety‐like behaviors caused by Staphylococcus aureus strain USA300

Abstract Introduction: Staphylococcus aureus (S. aureus) is considered as one of the major causative agents of serious hospital‐ and community‐acquired infections. Recent studies have reported that S. aureus infection induced neuroinflammation and was linked with some mental disorders. To evaluate the effects of S. aureus infection on abnormal behaviors, we conducted the present study. Methods: A S. aureus USA300‐infected mouse model was established using bacterial suspension injection into tail vein. A series of behavioral tests were performed after USA300 infection. The expression of cytokines was detected in serum and mPFC. The number and some morphological parameters of microglia were also evaluated by immunofluorescence staining. Results: Anxiety‐like behaviors, instead of locomotor activity impairment or depression‐like behaviors, were observed in mice infected with S. aureus USA300 compared with control. S. aureus USA300 infection caused overexpression of IL‐6, TNF‐α, and IL‐1β in serum, resulted in microglial over‐activation and excessive release of proinflammatory cytokines in the mPFC. In addition, overexpression of TLR2 accompanied by increased GLS1 and p‐STAT3 was observed in the mPFC of mice infected with S. aureus USA300. Conclusion: This study provides evidence that S. aureus USA300 infection can lead to neuroinflammation in the mPFC of mice, which may contribute to the development of anxiety.

2019). The exact etiology of anxiety remains poorly understood. Accumulating evidence has suggested microbial infection at various life stages may confer a significant risk of anxiety (Dinan & Cryan, 2017).

Staphylococcus aureus (S. aureus) is a Gram-positive bacterium that causes infection ranging from skin and soft tissue infection to invasive
infection, such as sepsis, endocarditis, osteomyelitis, and pneumonia (Tam & Torres, 2019;Turner et al., 2019). Remarkably, the spread of methicillin-resistant S. aureus (MRSA) strains has seriously reduced antibiotic treatment options and brought a global medical concern.
In a clinical trial, increased anxiety was noted in children with atopic dermatitis, which is usually associated with S. aureus infection (Huang et al., 2021). A recent study has reported that S. aureus could cause a series of anxiety-like behaviors in brain abscess mice, and these abnormal behaviors were improved by the treatment of bactericidal agent ciprofloxacin, sensitive to the S. aureus strain (Dey & Bishayi, 2021). However, the potential role of S. aureus infection in the development of anxiety-like behaviors is still unclear.
A growing stream of studies have suggested that S. aureus-induced peripheral inflammatory or neuroinflammation is involved in the pathogenesis of mental disorders (Balczon et al., 2019;Mahanti et al., 2015). S. aureus could engage TLR2 to recognize bacterial lipoproteins and induce inflammatory cytokine production in the central nervous system (CNS) to trigger and maintain the neuroinflammatory process (de Morais et al., 2021). A recent study has found that the activation of microglia and upregulation of TLR2 triggered anxiety-like behaviors in mice . In addition, clinical and basic studies have proven that microglial activation as well as proinflammatory cytokines, such as TNF-α and IL-6, arising from both peripheral and CNS induced by infection or stress response was involved in the development of anxiety-like behaviors (Ishikawa & Furuyashiki, 2022). Therefore, S. aureus-induced neuroinflammation may play a critical role in anxiety.
The medial prefrontal cortex (mPFC) has been posited to serve a variety of social, affective, and cognitive functions (Lieberman et al., 2019) and is also extremely vulnerable to inflammatory injury. It was reported that the modulation of mPFC was closely related to psychiatric disorders and abnormal behaviors (Adams et al., 2022;Isserles et al., 2021). An animal study has shown that low-intensity focused ultrasound stimulation could inhibit social avoidance behavior by inhibiting activation of the inflammatory response, increasing neuronal excitation, and protecting the integrity of the neuronal structure in the mPFC . In a chronic social defeat stress depression model, obvious microglial activation and excess release of proinflammatory cytokines were found in the mPFC of mice with depressive and anxiety-like behaviors (Dang et al., 2022). The mPFC may be an important brain area mediating anxiety.
In our study, we evaluated the causal relationship between S. aureus USA300-induced neuroinflammation in the mPFC and anxiety-like behaviors. The results indicated that the upregulation of GLS1 as well as the NF-κB/STAT3 signaling pathway in the mPFC drive mount neuroinflammation, which is involved in the development of anxiety-like behaviors caused by USA300 infection. This study provides a new idea for the mechanism and treatment of anxiety.

Mice
Male BALB/C mice (6-8 weeks old and weighing 20-24 g) were purchased from Hunan Saike Jingda Experimental Animal Co. Ltd. All mice were housed on a 12/12 light/dark cycle and fed a standard diet with tap water. All testing procedures were performed in a room to minimize any stress response potentially induced by novel environmental cues. All experimental procedures were approved by the Third Military Medical University and performed according to the Guidelines of Laboratory Animal Care and Use.

Body weight and survival rate monitoring
Male BALB/c mice were randomly divided into the following two groups: control and USA300. Mice were injected with 100 μl saline or USA300 suspension (1 × 10 5 CFU and 1 × 10 7 CFU) via tail vein.
The body weights and survival rates of mice in different groups were recorded for 8 days.

Behavior tests
All behavioral experiments were performed during the light phase.

Light-dark transitions
The light-dark transitions test was used to evaluate the anxiety levels of mice according to previous procedures (Flannery et al., 2015).
Mice were placed into the "light" side (∼400 lx) and allowed to move freely for 10 min. Time spent in the dark side and the total number of transitions were automatically recorded by Ethovision XT 11.0 (Noldus).

Elevated plus-maze
The elevated plus-maze (EPM) was structured with two open arms and two closed arms (30 cm × 6 cm ×15 cm) extending from a central area The maze was thoroughly cleaned using 70% ethanol between each test session.

Sucrose preference test (SPT)
Mice were given two bottles in the home cage from which they could drink freely. During the first day, the two bottles contained water and allowed habituation to the bottles. On the following day, one bottle was replaced with a 1% sucrose solution. Mice were then allowed to drink from the bottles over two days, and consumption of each bottle was recorded daily. The bottles were switched daily to ameliorate side bias.
Fluid intake was measured afterwards by weighing the drinking bottles.
The sucrose preference was calculated from the amount of sucrose solution consumed, expressed as a percentage of the total amount of liquid drunk: sucrose preference = (sucrose intake (g)) / (sucrose intake (g) + water intake (g)) 100% (Williams et al., 2020).

Tail suspension test (TST)
Mice were suspended by a hook 50 cm above the floor using adhesive tape. The hook was placed ∼1 cm from the tip of the tail. The activities of mice were recorded for 6 min. The immobility duration was analyzed during the last 4 min.

Forced swimming test (FST)
A glass cylinder (height: 20 cm, diameter: 10 cm) was filled with water up to a level of 10 cm. The temperature was maintained at 23 ± 2 • C.
Mice were placed in the middle of the given cylinder and allowed to swim for 6 min. The immobility durations of mice during the last 4 min of the total of 6 min were evaluated by two independent operators.

New object recognition (NOR)
The new object recognition task was conducted in a square gray Plexiglas arena (40 cm × 40 cm × 30 cm) located in a sound attenuated room.
Briefly, after a 10 min habituation period, two identical cylinders (A) were introduced in the diagonal corners of the arena. First, mice were placed in the box and allowed to explore freely the objects for 10 min.
Second, one of the cylinders was replaced with a new cuboid object (B).
The time for exploring freely was 10 min, which was the same as the first trial. Only the second trial was evaluated. An exploratory preference index defined as the ratio of the time spent exploring the cuboid object (B) to the total time spent exploring both objects (A and B) was calculated. Individual movement tracks were analyzed with Ethovision XT 11.0 (Noldus).

Nest building task
Mice were individually housed in a clean cage for 24 h. The nesting material consisting of nearly 2.5 g/5 cm 2 of square compressed cotton (Ancare, USA) were placed in the middle of the cage. The percentage of the shredded nestlet was recorded after 24 h.

Real-time PCR
The mPFC tissues were collected, and the mRNA levels of TLR2, NF-κB, IL-6, TNF-α, IL-4, and IL-10 in the mPFC were determined by real-time PCR. Total RNA was extracted from the mPFC using Trizol (Invitrogen, Carlsbad, CA, USA) according to the manufacturer's instructions, and this step was followed by reverse transcription. Quantitative PCR amplification was performed in triplicate using the SYBR Green kit (Takara Company, Japan) with the following program: 1 cycle of 95 • C for 30 s, 40 cycles of 95 • C for 5 s and 60 • C for 30 s. The primer sequences used are listed in Table S1. All data were normalized against the housekeeping control GAPDH expression using the Thermal Cycler Dice Real Time system (TaKaRa Company, Japan).

Cytokine ELISA
Serum IL-6, TNF-α, and IL-1β concentrations were detected by sandwich enzyme-linked immunosorbent assay (ELISA) and calculated based on the standard curve. Serum cytokine levels were expressed in pg/ml, following manufacturer's protocol. The minimum detectable dose for IL-6, TNF-α, and IL-1β was typically 1.6 pg/ml, 1.88 pg/ml, and 2.31 pg/ml, respectively.

Immunofluorescence
According to our previous method , mice were anesthetized and transcardially perfused with saline followed by 4% paraformaldehyde (PFA). Subsequently, the brains were postfixed in 30% sucrose solution with 4% PFA. Serial coronal brain sections (30-μm-thick) were collected on a cryostat and preserved in a cryoprotectant at −20 • C. The sections were incubated with the following primary antibodies (rabbit anti-ionized calcium binding adapter molecule Microglia morphology was quantified according to a previous study (He et al., 2022). Eight-bit 30 μm z-stack images of Iba1 + cells were obtained with no more than a 2 μm interval between planes. Images were converted to binary by using Image J; Soma size, branch numbers, and branch length of microglia were measured using Image J with the plugin AnalyzeSkeleton.

Western blot
The protein of mouse mPFC tissues were extracted, and protein concentration was measured using a bicinchoninic acid kit (Beyotime (1:1000, CST). The next day, the membranes were washed and incubated with the antirabbit or antimouse secondary antibodies for 1 h at room temperature. The specific protein bands were visualized in membranes by chemiluminescence method, and the optical density of protein bands were analyzed using Bio-Rad Image-Lab 6.0 software.

Statistical analyses
All the data were presented as the mean ± SEM and analyzed by SPSS 24.0 software (IBM Corporation, Armonk, NY, USA). Repeatedmeasures ANOVA were employed to analyze the subjecting data from the total distance and center time in the OFT and body weight monitoring. Two groups comparison were performed with independent t-tests.
The Welch's correction was employed when the variance between the two groups was significantly different, Mann-Whitney U test were used when normality of samples was failed according to our previous study . When the p value was less than .05, it was considered statistically significantly different in our study.

Ethics approval and informed consent
All experimental procedures were approved by Third Military Medical University and were performed according to the guidelines of laboratory animal care and use.

Effects of USA300 infection on the body weight and survival rate of mice
The body weights of mice were monitored for 8 days. There was no apparent loss of body weight infected with USA300 (1 × 10 5 CFU), instead of USA300 (1 × 10 7 CFU), compared with control ( Figure S1a).
The survival rate of mice was also monitored for 8 days. All the mice infected with USA300 (1 × 10 5 CFU) or saline survived. However, the survival rate of mice infected with USA300 (1 × 10 7 CFU) was obviously decreased ( Figure S1b). These results suggested that 1 × 10 5 CFU, instead of 1 × 10 7 CFU, USA300 infection could not result in death in mice. Therefore, the dose of 1×10 5 CFU USA300 was selected to perform the next behavior tests.

Anxiety-like behaviors were induced in mice infected with S. aureus USA300
To explore whether acute S. aureus USA300 infection could induce anxiety-like behaviors, male mice aged 6-8 weeks underwent the open field test, elevated plus-maze, and light-dark transition test after tail vein injection. No remarkable difference in the total distance traveled over the total 30-min period among the USA300 infection and control groups, which indicated that locomotor activity was not affected by S. aureus USA300 infection (Figure 1a Total transitions of mice infected with USA300 were significantly higher compared to the control (Figure 2d, p < .05). Collectively, these results suggested S. aureus USA300 infection could induce anxiety-like behaviors in BALB/C mice.

3.2.1
Effects of S. aureus USA300 infection on short-term memory and depression-like behaviors The novel object recognition test was conducted to determine whether S. aureus USA300 infection resulted in apparent impairments in the

S. aureus USA300 infection induced cytokine overproduction in serum
To determine the host cytokine responses, the production of IL-1β, IL-6, and TNF-α in serum were respectively examined 6 h after S. aureus  In line with the immunohistochemical data, Iba1 protein levels in the mPFC were markedly increased in the mice infected with USA300 compared with the control (Figure 5d and e, p < .01). As shown in Figure 5f, microglia exhibited ramified morphology with smaller soma and more microglial processes in the mPFC of control, while microglia displayed an amoeboid morphology, with retracted, thickened processes and enlarged soma in the mPFC of mice infected with USA300.

S. aureus USA300 infection induced microglia activation and altered proinflammatory cytokine production in the mouse mPFC
Compared to the control group, microglial soma size was significantly larger (Figure 5h Although there was no significant difference in the mRNA expressions of TNF-α in the mPFC of mice in the two groups, the protein levels of TNF-α were significantly increased in the mice infected with S. aureus USA300 compared to the control (Figure 6d, p > .05; Figure 6h and j, p < .05). In addition, the mRNA levels of anti-inflammatory cytokines IL-4 and IL-10 were detected and not statistically changed in the mPFC of mice infected with S. aureus USA300 compared to the control ( Figure   S3a and b, p > .05). These results suggested that S. aureus USA300 infection induced neuroinflammation in mPFC through TLR2/NF-κB signaling pathway.

S. aureus USA300 induced excessive expression of GLS1 and STAT3 activation in the mPFC
Increasing evidence showed that GLS1 plays an important role in neuroinflammation (Ji et al., 2021). In the CNS, GLS is an enzyme that catalyzes the hydrolytic deamidation of glutamine to glutamate and plays an important role in excitotoxic glutamate generation. The expression of GLS1 in the mPFC was detected using western blot.
As shown in Figure 7a and b, the levels of GLS1 in the mPFC of S. aureus USA300-infected mice were significantly higher than those of control mice (p < .05). Importantly, glutamate derived from glutaminolysis seems to play an essential role in STAT3 activation, which has been increasingly considered to play a critical role in anxiety (Shentu et al., 2021). Western blot was used to further assess the expression of p-STAT3. Consistently, we found that the p-STAT3 level was significantly upregulated in S. aureus USA300-infected mice compared to the control group (Figure 7c and d, p < .05). These results indicated that the overexpression of GLS1 after S. aureus USA300 infection was associated with the activation of the STAT3-signaling pathway.

DISCUSSION
In this work, we found that acute S. aureus USA300 (1 × 10 5 CFU) infection resulted in anxiety-like behaviors. Increased inflammatory cytokines in serum and mPFC were observed in the S. aureus USA300infected mice accompanied by excessive activation of microglia in the mPFC and upregulation of TLR2. S. aureus USA300 infection also upregulated the GLS1 and STAT3-signaling pathway, implicated with anxiety-like behaviors.
S. aureus is an important Gram-positive human pathogen that causes serious hospital-and community-acquired infections, including bacterial infections, bacteremia, and sepsis (Vautor et al., 2009). At present, S. aureus is one of the most common organisms found in abscess related to traumatic brain injury and neurosurgical procedure (Lu et al., 2002;Patel & Clifford, 2014). Patients with S. aureus infection are accompanied by impairments in cognition, memory, sensory processing, and emotional functioning (Iwashyna et al., 2010;Lazosky et al., 2010;Semmler et al., 2013). Mounting clinical evidence has revealed that S. aureus infection was closely related to anxiety (Andersson et al., 2016;Mursaloglu et al., 2021). Dey and Bishayi (2021) have found that 2.0×10 8 CFU of S. aureus (AG-789) could cause a series of anxietyand depression-like behaviors in a brain abscess model of Swiss albino mice. In addition, male albino mice were given 5 × 10 6 CFU of S. aureus Representative the densitometric analysis of TNF-α (n = 3 for each group). Data are presented as mean ± SEM. *p < .05 Bacterial invasion always leads to host inflammation, an adaptive response triggered by noxious stimuli and conditions (Medzhitov, 2008). There is strong evidence that excessive inflammatory response plays a crucial role in the development of anxiety (Duivis et al., 2013;Pitsavos et al., 2006). A recent meta-analysis has shown that peripheral IL-6 and TNF-α were significantly increased in patients with generalized anxiety disorder (Costello et al., 2019). Consistent with this, we demonstrated that the serum levels of IL-6 and TNF-α were significantly higher in mice infected with S. aureus USA300 compared to the control group. In addition, activation of TLR2 could mediate intracellular signaling in defense against S. aureus infection (Barton & Medzhitov, 2003;Esen et al., 2004), trigger TLR2-MyD88 signaling, and then result in activating NF-κB signaling via a cascade of intermediary steps. Subsequently, a series of proinflammatory cytokines, such as IL-6 and TNF-α, were released (Nguyen & Gotz, 2016). Downregulation of TLR2 could prevent the overexpression of TNF-α, IL-1β, and IL-6 caused by S. aureus in mice . Some previous studies have pointed out that the imbalances of pro-and anti-inflammatory cytokines were relation with abnormal behaviors (You et al., 2011). IL-4 and IL-10 were reported to play important roles in abnormal behaviors Zhang et al., 2020). The mRNA levels of IL-4 and IL-10 in the mPFC were detected by real-time PCR. We found that the levels of IL-4 and IL-10 in the mPFC of mice infected with USA300 were not significantly changed compared to the control group. It meant that USA300 infection may not affect the expression of IL-4 and IL-10 in the mPFC.
The mPFC is a primary target of stress and connected with many complex behaviors in human and animal studies. The stimulation of extraneous factors could give rise to microglia activation and excessive inflammation in mPFC, which were associated with many psychiatric disorders, such as anxiety and depression Yang et al., 2020). Our previous study showed that loss of LXRβ in astrocytes in mPFC leaded to anxiety-like behaviors in mice . In line with these studies, microglia activation was found in the mPFC after S. aureus infection, accompanied by the increase of TLR2/NF-κB and overexpression of proinflammatory cytokines IL-6 and TNF-α in the mPFC. These results indicated that neuroinflammation following S. aureus infection in the mPFC was associated with anxiety-like behaviors. Hippocampus is involved in learning and memory and is also extremely vulnerable to environmental factors. Several studies have suggested that hippocampus may be the other targeted brain region of S. aureus infection. Peripheral S. aureus infection could promote histone H3 hypoacetylation and decreases tyrosine hydroxylase protein level in the hippocampus and prefrontal cortex of rat (Choudhury et al., 2019). Mice challenged intraperitoneally with S. aureus showed increase of anxiety-like behaviors, which may be mediated by the overexpression of IL-6 and TNF-α in prefrontal cortex and hippocampus (Shal et al., 2019). We should pay more attention to the hippocampus after S. aureus infection in the following study.
Glutaminase (GLS) mediates the conversion of glutamine to glutamate, supporting most excitatory neurotransmission in the CNS (Kosten et al., 2019). GLS have two isoforms GLS1 and GLS2. GLS1 is the main glutamine metabolic enzyme in the brain rather than GLS2.
High levels of GLSl protein were identified in many psychiatric and neurological diseases (Hamed et al., 2018;Huang et al., 2011). Microglial GLS1 deficiency could mitigate neuroinflammation in the lipopolysaccharide (LPS)-induced depression model (Ji et al., 2021). JHU083, a glutaminase inhibitor, was reported to attenuate behavioral deficits and improve disease outcomes in some neuroinflammation models, such as chronic social defeat stress and Alzheimer's disease (Hollinger et al., 2020;Zhu et al., 2019). In addition, Chen et al. (2012) found that TNF-α contributed to GLS upregulation and increased glutamate release in Japanese encephalitis virus-infected microglia. Consistent with these results, we confirmed an increased level of GLS1 in the mPFC following S. aureus USA300 infection, suggesting that abnormal GLS1 induced by excessive inflammatory response contributed to anxiety-like behaviors. Glutaminolysis can regulate the activation of STAT3, and GLS1 inhibition was reported to reduce phosphorylated STAT3 expression (Guo et al., 2016;Xia et al., 2018). Moreover, we found that excessive inflammatory cytokines induced an elevation of STAT3 via enhancing the phosphorylation of the latter in mPFC after S. aureus USA300 infection. These results reveal that S. aureus USA300 induced excessive expression of GLS1 and STAT3 activation in mPFC.

CONCLUSIONS
In conclusion, we provided the evidence that S. aureus USA300 infection could result in anxiety-like behaviors, which was involved with neuroinflammation in mPFC. These findings extend our understanding of S. aureus infection in the nervous system and may provide novel insights to the model and therapy of anxiety.

CONFLICT OF INTEREST
The authors confirm that there are no conflicts of interest.

AUTHOR CONTRIBUTIONS
JZ and WS designed and conducted the experiments, collated and analyzed the data, and wrote the original manuscript. LY and LW contributed to conception and design of the study. XL, TL, and JZ monitored the status of the mice and collected the samples. LY and LW contributed to the collection of data for quantification and statistical analysis. XF, JG, and XR reviewed the experimental plan, provided technical and financial support, and revised the manuscript. All authors contributed to the article and approved the submitted version.

DATA AVAILABILITY STATEMENT
The data in this study are available from the corresponding author based on the reasonable request.

PEER REVIEW
The peer review history for this article is available at https://publons. com/publon/10.1002/brb3.2715