C-X-C Motif Chemokine 3 Promotes the Inflammatory Response of Microglia after Escherichia coli-Induced Meningitis

Meningitis is a major clinical manifestation of Escherichia coli (E. coli) infection characterized by inflammation of the meninges and subarachnoid space. Many chemokines are secreted during meningitic E. coli infection, of which C-X-C motif chemokine 3 (CXCL3) is the most highly expressed. However, it is unclear how CXCL3 plays a role in meningitic E. coli infection. Therefore, this study used in vitro and in vivo assays to clarify these contributions and to identify novel therapeutic targets for central nervous system inflammation. We found a significantly upregulated expression of CXCL3 in human brain microvascular endothelial cells and U251 cells after meningitic E. coli infection, and the CXCL3 receptor, C-X-C motif chemokine receptor 2 (CXCR2), was expressed in microglia. Furthermore, CXCL3 induced M1 microglia by selectively activating mitogen-activated protein kinases signaling and significantly upregulating tumor necrosis factor-α (TNF-α), interleukin (IL)-1β, IL-6, nitric oxide synthase 2 (NOS2), and cluster of differentiation 86 (CD86) expression levels, promoting an inflammatory response. Our findings clarify the role of CXCL3 in meningitic E. coli-induced neuroinflammation and demonstrate that CXCL3 may be a potential therapeutic target for future investigation and prevention of E. coli-induced neuroinflammation.


Introduction
A common infectious disease that affects the central nervous system (CNS) is bacterial meningitis. Organisms that cause meningitis include Escherichia coli, Haemophilus influenzae, Streptococcus pneumoniae, and Neisseria meningitidis. It is possible for E. coli to infect the CNS and cause neuroinflammation [1]. The blood-brain barrier (BBB) is a structural and functional barrier formed by endothelial cells interacting with pericytes, astrocytes, neurons, and microglia [2,3]. Various molecules pass through it to and from the brain, maintaining the neural microenvironment and protecting the brain from microbes and toxins in the blood [4]. The brain microvascular endothelial cells (BMECs) regulate molecular transport between the bloodstream and the brain in order to maintain a highly controlled neurovascular environment for the proper functioning of neuronal circuits [5]. Astrocytes play an important role in neurodegenerative diseases by contributing to the dynamic regulation of the neural system [6]. Microglia play a critical role in supporting key functions in the CNS [7]. The majority of CNS diseases involve microglia, which convert from a resting/surveillance state in the normal brain to a fully active state in the diseased brain [8]. When the brain parenchyma is infected or stimulated, microglia become active and are involved in a variety of CNS disorders, such as brain trauma, stroke, and Parkinson's disease; thus, they have contributed significantly to neurological diseases. Studies have shown that, similar to macrophages, microglia have two distinct activation phenotypes: classical (M1) and alternative (M2). M1 microglia are more likely to induce neuronal death than M2 microglia. Collectively, these two microglia activation phenotypes, neurotoxicity, and neuroprotection after activation play important roles in disease pathogenesis.
A chemokine is a chemotactic cytokine that influences the migration patterns and positioning of immune cells [9]. Chemokines play an important role in pathophysiological processes, such as in inflammation, angiogenesis, and asthma [10]. Many cytokines are secreted during meningitic E. coli infection, including the chemokine C-X-C motif chemokine 3 (CXCL3), also called growth-related oncogene γ (GRO-γ) or macrophage inflammatory protein 2 (MIP-2) [11,12]. CXCL3 is a neutrophil-activating chemokine that belongs to the growth-related oncogene subfamily of CXC chemokines, which exert their biological roles through the chemokine receptors, C-X-C motif chemokine receptors (CXCRs) 1 and 2 [13,14]. In 2000, Addison et al. demonstrated that CXCR2 mediates the biological functions of CXCL3 [15].
Macrophages, osteoblasts, airway epithelial cells, and dendritic cells secrete CXCL3. In addition, CXCL3 promotes blood vessel formation, tumor cell growth, cancer cell migration, cluster of differentiation (CD) 31 vascular cell infiltration, and smooth muscle cell migration [16]. It is still unclear how CXCL3 contributes to meningitis. Therefore, this study investigated the function of CXCL3 in meningitic E. coli infection to identify novel therapeutic targets for CNS inflammation.

Meningitic E. coli Significantly Stimulates CXCL3 Expression In Vivo and In Vitro
Previous RNA sequencing data demonstrated upregulated CXCL3 expression in hB-MECs and U251 cells [17,18]. As a result of bacterial challenge, hBMEC transcription of CXCL3 increased significantly over time and was sustained ( Figure 1A). Furthermore, meningitic E. coli infection promoted CXCL3 protein expression ( Figure 1B). CXCL3 expression was also upregulated in U251 cells after the bacterial challenge ( Figure 1C,D). CXCL3 transcription increased rapidly 1 h post infection, especially in the U251 cells; transcription in these cells, was considerably higher than that in hBMECs. immunoblot results from whole cell extracts after infecting (C) hBMECs or (D) U251 cells after meningitic E. coli infection at an MOI of 10 for 1, 2, and 3 h. (E) CXCL3 transcription levels (brain lysates; real-time PCR) and (F) serum CXCL3 concentrations after meningitic E. coli infection (intravenous injection) (n = 5) at 1 × 10 7 colony forming units for 2, 4, 6, and 8 h. (G) CXCL3 protein expression from challenged mice (brain lysates; western blot). Data were presented as mean ± SD from three independent assays. p < 0.05 (*) was considered statistically significant; p < 0.01 (**) and p < 0.001 (***) indicated extremely significant differences.
The brains of mice with meningitic E. coli infection also showed significant and timedependent increases in CXCL3 transcription ( Figure 1E), as did the serum CXCL3 protein levels ( Figure 1F). Furthermore, serum CXCL3 expression increased 2 h post infection and continued increasing time-dependently ( Figure 1G). In summary, treatment with meningitic E. coli promoted CXCL3 expression in vitro and in vivo.

Microglia Express CXCR2 during Meningitic E. coli Infection
High CXCL3 levels have been detected in mouse brain tissue after infection, and CXCL3 interacts with CXCR2 (the receptor) to produce a range of biological effects. Therefore, we examined the expression of CXCR2 in the mouse brain by immunofluorescence. Complement component receptor-3 alpha (CD11b), a classical microglia marker, was used to detect the location of CXCR2 in the brain; it was expressed in microglia ( Figure 2A). Moreover, we observed elevated CXCR2 expression levels in BV2 cells after 3 h of stimulation with recombinant mouse CXCL3 ( Figure 2B). Consistently, CXCR2 messenger RNA transcription and protein levels were time-dependently upregulated in response to CXCL3 ( Figure 2C). Therefore, we propose that meningitic E. coli infection facilitates CXCL3 expression and acts on microglia. immunoblot results from whole cell extracts after infecting (C) hBMECs or (D) U251 cells after meningitic E. coli infection at an MOI of 10 for 1, 2, and 3 h. (E) CXCL3 transcription levels (brain lysates; real-time PCR) and (F) serum CXCL3 concentrations after meningitic E. coli infection (intravenous injection) (n = 5) at 1 × 10 7 colony forming units for 2, 4, 6, and 8 h. (G) CXCL3 protein expression from challenged mice (brain lysates; western blot). Data were presented as mean ± SD from three independent assays. p < 0.05 (*) was considered statistically significant; p < 0.01 (**) and p < 0.001 (***) indicated extremely significant differences.
The brains of mice with meningitic E. coli infection also showed significant and timedependent increases in CXCL3 transcription ( Figure 1E), as did the serum CXCL3 protein levels ( Figure 1F). Furthermore, serum CXCL3 expression increased 2 h post infection and continued increasing time-dependently ( Figure 1G). In summary, treatment with meningitic E. coli promoted CXCL3 expression in vitro and in vivo.

Microglia Express CXCR2 during Meningitic E. coli Infection
High CXCL3 levels have been detected in mouse brain tissue after infection, and CXCL3 interacts with CXCR2 (the receptor) to produce a range of biological effects. Therefore, we examined the expression of CXCR2 in the mouse brain by immunofluorescence. Complement component receptor-3 alpha (CD11b), a classical microglia marker, was used to detect the location of CXCR2 in the brain; it was expressed in microglia ( Figure 2A). Moreover, we observed elevated CXCR2 expression levels in BV2 cells after 3 h of stimulation with recombinant mouse CXCL3 ( Figure 2B). Consistently, CXCR2 messenger RNA transcription and protein levels were time-dependently upregulated in response to CXCL3 ( Figure 2C). Therefore, we propose that meningitic E. coli infection facilitates CXCL3 expression and acts on microglia. Immunofluorescence results of CXCR2 expression (red) and CD11b (green, microglia) from the brain of mice after meningitic E. coli infection. Scale bar, 50 µm. (B) CXCR2 protein expression in BV2 cells in response to 3 or 6 h of C-X-C motif chemokine 3 (CXCL3) treatment (10 ng/mL) (western blot). (C) CXCR2 transcription levels in BV2 cells in response to 3 or 6 h of CXCL3 treatment (10 ng/mL) (n = 3; real-time polymerase chain reaction). GAPDH was used as an internal control for normalization. Data were presented as mean ± SD from three independent assays. p < 0.01 (**) indicated extremely significant differences; ns, not significant.

CXCL3 Facilitated Microglia Polarization towards a Pro-Inflammatory Profile
Microglia become active after stimulation. We therefore investigated whether CXCL3 affects the polarization of microglia toward a pro-inflammatory phenotype in meningitic E. coli infection. Compared to control mice, meningitic E. coli infection mice had a significant increase in CD16/32 and CD11b double-positive microglia, but a significant decrease in CD206 and CD11b double-positive microglia ( Figure 3A). To determine the effect of CXCL3 on the polarization of BV2 cells, we used flow cytometry to detect CD16/32 and CD206 expression and found that CXCL3 (10 ng/mL) treatment for 3 h significantly increased the percentage of CD16/32 positive microglia compared to that in the PBS group, but the percentages of CD206 positive microglia were consistent between the treatment and PBS groups ( Figure 3B). These results indicate that CXCL3 induces M1 microglia polarization. Immunofluorescence results of CXCR2 expression (red) and CD11b (green, microglia) from the brain of mice after meningitic E. coli infection. Scale bar, 50 µm. (B) CXCR2 protein expression in BV2 cells in response to 3 or 6 h of C-X-C motif chemokine 3 (CXCL3) treatment (10 ng/mL) (western blot). (C) CXCR2 transcription levels in BV2 cells in response to 3 or 6 h of CXCL3 treatment (10 ng/mL) (n = 3; real-time polymerase chain reaction). GAPDH was used as an internal control for normalization. Data were presented as mean ± SD from three independent assays. p < 0.01 (**) indicated extremely significant differences; ns, not significant.

CXCL3 Facilitated Microglia Polarization towards a Pro-Inflammatory Profile
Microglia become active after stimulation. We therefore investigated whether CXCL3 affects the polarization of microglia toward a pro-inflammatory phenotype in meningitic E. coli infection. Compared to control mice, meningitic E. coli infection mice had a significant increase in CD16/32 and CD11b double-positive microglia, but a significant decrease in CD206 and CD11b double-positive microglia ( Figure 3A). To determine the effect of CXCL3 on the polarization of BV2 cells, we used flow cytometry to detect CD16/32 and CD206 expression and found that CXCL3 (10 ng/mL) treatment for 3 h significantly increased the percentage of CD16/32 positive microglia compared to that in the PBS group, but the percentages of CD206 positive microglia were consistent between the treatment and PBS groups ( Figure 3B). These results indicate that CXCL3 induces M1 microglia polarization. ) in BV2 cells in response to CXCL3 (10 ng/mL) stimulation for 6 h (n = 3; quantitative real-time polymerase chain reaction). GAPDH was used as an internal control for normalization. Data were presented as mean ± SD from three independent assays. p < 0.05 (*) was considered statistically significant; p < 0.01 (**) and p < 0.001 (***) indicated extremely significant differences.
Next, we investigated whether the ERK1/2 pathway was involved in pro-inflammatory factor production. Following pre-treatment with U0126 (a specific ERK1/2 inhibitor), iNOS, and CD16 expression significantly decreased ( Figure 4B). In addition, IL-1β, IL-6, TNF-α, and iNOS, as well as CD86 expression substantially decreased to different extents than those in their respective control groups ( Figure 4C).
Finally, we performed in vivo experiments, administering U0126 intraperitoneally (10 mg/kg) for 12 h to block ERK signaling before the intravenous injection of meningitic E. coli. The subsequent real-time PCR analysis indicated that U0126 treatment decreased IL-1β, IL-6, iNOS, and CD86 expression ( Figure 4D). According to these results, ERK1/2 is active and participates in the CXCL3-induced inflammatory responses in BV2 cells.

CXCR2 Blockade Inhibits the Expression of Pro-Inflammatory Factors
CXC chemokines exert their biological roles through the chemokine receptors CXCR1 and CXCR2. To confirm the critical role of CXCL3 in microglia activation, we intraperitoneally injected 1 mg/kg of the CXCR2 antagonist, SB225002 (a specific CXCR2 antagonist), to inhibit CXCL3 activity 12 h before the intravenous injection of meningitic E. coli. SB225002 treatment decreased the IL-1β, IL-6, TNF-α, iNOS, and CD86 expression levels (real-time PCR; Figure 5A). We also pre-treated BV2 cells with SB225002, finding similar results to the in vitro experiment ( Figure 5B). Next, we investigated whether the ERK1/2 pathway was involved in pro-inflammatory factor production. Following pre-treatment with U0126 (a specific ERK1/2 inhibitor), iNOS, and CD16 expression significantly decreased ( Figure 4B). In addition, IL-1β, IL-6, TNF-α, and iNOS, as well as CD86 expression substantially decreased to different extents than those in their respective control groups ( Figure 4C).
Finally, we performed in vivo experiments, administering U0126 intraperitoneally (10 mg/kg) for 12 h to block ERK signaling before the intravenous injection of meningitic E. coli. The subsequent real-time PCR analysis indicated that U0126 treatment decreased IL-1β, IL-6, iNOS, and CD86 expression ( Figure 4D). According to these results, ERK1/2 is active and participates in the CXCL3-induced inflammatory responses in BV2 cells.

CXCR2 Blockade Inhibits the Expression of Pro-Inflammatory Factors
CXC chemokines exert their biological roles through the chemokine receptors CXCR1 and CXCR2. To confirm the critical role of CXCL3 in microglia activation, we intraperitoneally injected 1 mg/kg of the CXCR2 antagonist, SB225002 (a specific CXCR2 antagonist), to inhibit CXCL3 activity 12 h before the intravenous injection of meningitic E. coli. SB225002 treatment decreased the IL-1β, IL-6, TNF-α, iNOS, and CD86 expression levels (real-time PCR; Figure 5A). We also pre-treated BV2 cells with SB225002, finding similar results to the in vitro experiment ( Figure 5B). , interleukin (IL)-1β, IL-6, nitric oxide synthase 2 (NOS2), and cluster of differentiation (CD) 86 in the brains of mice that were pretreated with the CXCR2 antagonist SB225002 (1 mg/kg; intraperitoneal injection) for 12 h then infected with meningitic E. coli for 5 h (n = 5 per group; was determined by real-time polymerase chain reaction [PCR]). (B) mRNA expression of pro-inflammatory markers in BV2 cells were pretreated with SB225002 (50 nM) for 2 h then treated with C-X-C motif chemokine 3 (CXCL3) (10 ng/mL) for 6 h (n = 3; quantitative real-time PCR). β-actin was used as an internal control for normalization. Data were presented as mean ± SD from five independent assays. p < 0.05 (*) was considered statistically significant; p < 0.01 (**) and p < 0.001 (***) indicated extremely significant differences. , interleukin (IL)-1β, IL-6, nitric oxide synthase 2 (NOS2), and cluster of differentiation (CD) 86 in the brains of mice that were pretreated with the CXCR2 antagonist SB225002 (1 mg/kg; intraperitoneal injection) for 12 h then infected with meningitic E. coli for 5 h (n = 5 per group; was determined by real-time polymerase chain reaction [PCR]). (B) mRNA expression of pro-inflammatory markers in BV2 cells were pretreated with SB225002 (50 nM) for 2 h then treated with C-X-C motif chemokine 3 (CXCL3) (10 ng/mL) for 6 h (n = 3; quantitative real-time PCR). β-actin was used as an internal control for normalization. Data were presented as mean ± SD from five independent assays. p < 0.05 (*) was considered statistically significant; p < 0.01 (**) and p < 0.001 (***) indicated extremely significant differences.
Finally, immunofluorescence indicated that SB225002 or U0126 treatment significantly decreased the percentage of CD16/32 and CD11b double-positive microglia after meningitic E. coli infection compared to those without SB225002 or U0126 treatment ( Figure 6). These results further indicated that CXCL3 plays a critical role in M1 microglia activation. Finally, immunofluorescence indicated that SB225002 or U0126 treatment significantly decreased the percentage of CD16/32 and CD11b double-positive microglia after meningitic E. coli infection compared to those without SB225002 or U0126 treatment (Figure 6). These results further indicated that CXCL3 plays a critical role in M1 microglia activation. Figure 6. The effect of C-X-C motif chemokine receptor 2 (CXCR2) or extracellular signal-regulated protein kinases 1 and 2 (ERK1/2) antagonists on microglia polarization in vivo. Immunofluorescence results for CD11b (green), CD16/32 (red), and CD206 (red) in the brains of mice were pretreated with SB225002 (1 mg/kg; intraperitoneal injection) or U0126 (10 mg/kg; intraperitoneal injection) for 12 h then infected with meningitic E. coli for 5 h. Scale bar, 50 µm. Quantitative analysis of CD11b and CD16/32 double-positive and CD11b and CD206 double-positive cells. Data were presented as mean ± SD from three independent assays. p < 0.05 (*) was considered statistically significant; p < 0.01 (**) indicated extremely significant differences; ns, not significant.

Discussion
E. coli is an important Gram-negative bacterium and an important contributor to meningitis [5,17]. E. coli induces inflammation in the CNS by breaking through the BBB. However, the mechanism by which it breaks through the BBB remains unclear. Previous studies have mainly reported the important role of E. coli virulence factors in bacterial BBB penetration. In contrast, few studies have examined how host-intrinsic factors may influence the development of meningitis. Our previous study found that infection with meningitic E. coli can significantly increase the expression of chemokines and cytokines, such as IL-6, IL-1β, TNF-α, CXCL3, CXCL2, and C-C motif chemokine ligand (CCL) 2, and some Figure 6. The effect of C-X-C motif chemokine receptor 2 (CXCR2) or extracellular signal-regulated protein kinases 1 and 2 (ERK1/2) antagonists on microglia polarization in vivo. Immunofluorescence results for CD11b (green), CD16/32 (red), and CD206 (red) in the brains of mice were pretreated with SB225002 (1 mg/kg; intraperitoneal injection) or U0126 (10 mg/kg; intraperitoneal injection) for 12 h then infected with meningitic E. coli for 5 h. Scale bar, 50 µm. Quantitative analysis of CD11b and CD16/32 double-positive and CD11b and CD206 double-positive cells. Data were presented as mean ± SD from three independent assays. p < 0.05 (*) was considered statistically significant; p < 0.01 (**) indicated extremely significant differences; ns, not significant.

Discussion
E. coli is an important Gram-negative bacterium and an important contributor to meningitis [5,17]. E. coli induces inflammation in the CNS by breaking through the BBB. However, the mechanism by which it breaks through the BBB remains unclear. Previous studies have mainly reported the important role of E. coli virulence factors in bacterial BBB penetration. In contrast, few studies have examined how host-intrinsic factors may influence the development of meningitis. Our previous study found that infection with meningitic E. coli can significantly increase the expression of chemokines and cytokines, such as IL-6, IL-1β, TNF-α, CXCL3, CXCL2, and C-C motif chemokine ligand (CCL) 2, and some chemokines remain at high levels in the blood and brain for a long time [17,19]. In this study, we demonstrated that CXCL3 promotes microglia M1 polarization by binding to CXCR2, which ultimately induces the expression of pro-inflammatory cytokines (Figure 7). chemokines remain at high levels in the blood and brain for a long time [17,19]. In this study, we demonstrated that CXCL3 promotes microglia M1 polarization by binding to CXCR2, which ultimately induces the expression of pro-inflammatory cytokines ( Figure  7). CXCL3, an individual that belongs to the CXC chemokine subfamily, is widely expressed and associated with organ injury and inflammatory responses [20]. CXCL3 is highly upregulated during esophageal carcinogenesis and contributes to vascular invasion in gastric cancer and human melanoma. As well as promoting tumor cell migration, CXCL3 is associated with the migration of airway smooth muscle cells and neurons. Furthermore, in preeclampsia, activation of CXCL3 by endogenous factors promotes trophoblast invasion, migration, and proliferation in human trophoblasts and is critical to the pathogenesis of preeclampsia [21]. CXCL3 appears to play a role in several pathophysiological processes. However, the role of CXCL3 in CNS inflammation remains to be investigated. Compared to BMECs, astrocytes release significantly higher levels of CXCL3 in response to stimulation with the meningitis E. coli infection, suggesting that astrocytes are the main source of CXCL3, which acts on other cells and causes a series of biological effects. CXCL3, an individual that belongs to the CXC chemokine subfamily, is widely expressed and associated with organ injury and inflammatory responses [20]. CXCL3 is highly upregulated during esophageal carcinogenesis and contributes to vascular invasion in gastric cancer and human melanoma. As well as promoting tumor cell migration, CXCL3 is associated with the migration of airway smooth muscle cells and neurons. Furthermore, in preeclampsia, activation of CXCL3 by endogenous factors promotes trophoblast invasion, migration, and proliferation in human trophoblasts and is critical to the pathogenesis of preeclampsia [21]. CXCL3 appears to play a role in several pathophysiological processes. However, the role of CXCL3 in CNS inflammation remains to be investigated. Compared to BMECs, astrocytes release significantly higher levels of CXCL3 in response to stimulation with the meningitis E. coli infection, suggesting that astrocytes are the main source of CXCL3, which acts on other cells and causes a series of biological effects.
The biological functions of the chemokines in the CXC family depend on binding to their receptors CXCR1 or CXCR2. CXCR2 is expressed on endothelial cells, oligodendrocytes, and various immune cells; multiple sclerosis, traumatic brain injury, and Alzheimer's disease are associated with CXCR2 and its ligands [22][23][24]. CXCR2 antagonists have been proposed as a therapeutic strategy for treating inflammatory diseases [25]. Microglia do not express CXCR2 under homeostatic conditions; however, CXCL3 robustly induces high levels of CXCR2 mRNA and protein expression in BV2 microglia, and its expression is upregulated when activated during CNS pathologies. In contrast, in this study, we identified CXCR2 expression in the control group, perhaps because of differences between the cells in vitro and primary cells; the culture environment in vitro also differs from that in vivo. Therefore, we speculate that CXCL3 could either promote or inhibit the inflammatory response in microglia.
Some studies have shown that CXCL3 promotes prostate cancer cell proliferation and migration by upregulating p-ERK1/2, p-Akt, and Bcl2 and downregulating Bax [33]. Simultaneously, CXCL3 promotes the formation of fat cells by activating the ERK/MAPK and JNK/MAPK pathways [11], which suggests that CXCL3 induces M1 microglia by selectively activating ERK1/2 MAPK signaling, promoting an inflammatory response.

Bacterial Strains and Cell Culture
In this study, we used a meningitic E. coli strain (PCN033) maintained in our laboratory. Luria-Bertani medium was used to grow the bacterial cells.

Meningitic E. coli Infection of hBMECs and U251 Cells
E. coli was cultured overnight and diluted in serum-free medium before being added separately to the confluent hBMECs and U251 cell monolayer cultures. Samples were then harvested and processed using TRIzol reagent or cell lysis buffer after three washes with pre-chilled phosphate-buffered saline (PBS).

Western Blotting
After dilution of brain homogenates or cell lysates in loading buffer, they were boiled at 100 • C for 10 min. An acrylamide-sodium dodecyl sulfate gel containing 12% acrylamide was used for loading. Polyvinylidene difluoride membranes were used to transfer the gels. TBST (10 mM Tris-buffered saline with 0.05% Tween 20) was used to block the membranes for 2 h, followed by overnight incubation with either antibody. After washing, a species-specific horseradish peroxidase-conjugated antibody was added, and the blots were visualized with ECL reagent after incubation.

Isolation and Quantitative Real-Time Polymerase Chain Reaction (qRT-PCR) Analysis of RNA
Total RNA was isolated from brain lysates or cells using Trizol reagent (Aidlab Biotech, Beijing, China). HiScript II Q RT SuperMix for qPCR gDNA wiper (Vazyme, Nanjing, China) was used to synthesize complementary DNA from RNA aliquots from each sample. Using MonAmp SYBR Green qPCR Mix (RN04005M, Monad Biotech Co., Ltd. Wuhan, China), real-time PCR was performed on a qTOWER3/G quantitative real-time PCR thermal cycler (Analytikjena, Jena, Germany). The target gene expression was normalized using glyceraldehyde-3-phosphate dehydrogenase (GAPDH) or β-actin as a control.

Secretory CXCL3 Determination by Enzyme-Linked Immunosorbent Assay (ELISA)
Mice were challenged with meningitic E. coli as described above. Mice were euthanized at the indicated time points, and the serum was collected and stored at −80 • C. ELISA kits (Elabscience, Houston, TX, USA) were used to measure serum CXCL3 concentrations according to the manufacturer's instructions.

Flow Cytometry Analysis
Flow cytometry was applied to compare the expression of CD16/32 and CD206 in BV2 cells stimulated by CXCL3. Cells were seeded at a density of 1 × 10 6 in 6-well plates and cultured for 24 h. The cells were incubated with CXCL3 for 3 h, treated with trypsin without EDTA, and washed with cell staining buffer. Cells were incubated with FITC-conjugated monoclonal rat CD16/32 antibodies at 4 • C for 30 min, washed with a cell staining buffer, and then incubated with a fixation buffer at room temperature for 20 min. Next, the cells were treated with a permeabilization wash buffer then incubated with APC-conjugated monoclonal rat CD206 antibodies at 4 • C for 30 min and washed with a cell staining buffer. Finally, the cells were washed with a permeabilization wash buffer and suspended in a cell staining buffer in the dark at 4 • C. CD16/32 and CD206 expression levels were analyzed using a FACS Calibur Flow cytometer (Becton Dickinson, Franklin Lakes, NJ, USA). The FITC anti-CD16/32 (FITC-65080) and APC anti-CD206 (APC-65155) were obtained from Proteintech (Chicago, IL, USA).

Immunofluorescence Analysis
Mice with the typical CNS disorder were given ketamine-xylazine (0.1 mL/10 g) and perfused with PBS. Brain samples were then removed, fixed in 4% formaldehyde solution, and embedded in paraffin.
Sections were incubated with the primary antibody, conjugated with either Cy3 or FITC, and finally incubated with the appropriate secondary antibody, DAPI, to stain the nucleus.

Animal Infection Assay
Four-week-old Kunming (female) mice were obtained from the Laboratory Animal Center of China at Three Gorges University (Wuhan, China) for animal infection studies. These mice were then injected with meningitic E. coli through the tail vein with 1 × 10 7 colony-forming units (CFUs) and then sacrificed. At the designated times, the mice were anesthetized, their peripheral blood was collected for serum extraction, and it was then processed for further testing.
SB225002 (HY-16711, MedChemExpress; Summit, NJ, USA) is a selective non-peptide CXCR2 (the CXCL3 receptor) inhibitor and was used to inhibit CXCL3 activity. To block CXCR2, SB225002 (1 mg/kg) was injected intraperitoneally into mice 12 h before the intravenous injection of meningitic E. coli; a similar amount of DMSO was injected into the control group.

Statistical Analyses
Data were expressed as means ± standard deviations unless otherwise specified. Betweengroup differences were analyzed using one-way analysis of variance (ANOVA) in GraphPad Prism version 7.0 (GraphPad Software, La Jolla, CA, USA). p-values of < 0.05 indicated significant differences, and those <0.01 and <0.001 indicated extremely significant differences.

Conclusions
In conclusion, our results indicate that meningitic E. coli infection promotes CXCL3 expression, and CXCL3 binds to CXCR2 to promote neuroinflammation by regulating downstream signaling pathways and expression of pro-inflammatory factors in microglia. These findings suggest that CXCL3 may be a novel therapeutic target for E. coli-induced CNS inflammation.