NF‐κB‐mediated inflammatory damage is differentially affected in SH‐SY5Y and C6 cells treated with 27‐hydroxycholesterol

Abstract Previous studies have demonstrated that 27‐hydroxycholesterol (27‐OHC), a cholesterol metabolite, was involved in the inflammatory process of Alzheimer's disease (AD). The present study aimed to investigate the 27‐OHC‐induced inflammatory damage to neurons and astrocytes and the underlying mechanism(s) accounting for this damage. Human neuroblastoma cells (SH‐SY5Y cells) and rat glioma cells (C6 cells) were treated with vehicle or 27‐OHC (5, 10, or 20 μM) for 24 hr. The levels of secreted interleukin‐1β (IL‐1β), interleukin‐10 (IL‐10), tumor necrosis factor alpha (TNF‐α), and inducible nitric oxide synthase (iNOS) were determined by using an enzyme‐linked immunosorbent assay (ELISA). Immunofluorescence staining was used to determine the cellular expression of toll‐like receptor 4 (TLR4) and transforming growth factor‐β (TGF‐β). The mRNA and protein expression levels of nuclear factor‐κB p65 (NF‐κB p65), nuclear factor‐κB p50 (NF‐κB p50) and cyclooxygenase‐2 (COX‐2) in both SH‐SY5Y and C6 cells were also detected by real‐time PCR and Western blot, respectively. The results of this study showed that 27‐OHC treatment increased secretion of TNF‐α and iNOS and decreased secretion of IL‐10, upregulated expression of TGF‐β, NF‐κB p65 and p50, and downregulated expression of COX‐2 in SH‐SY5Y cells. In C6 cells, treatment with 27‐OHC resulted in decreased secretion of IL‐1β, IL‐10, TNF‐α, and iNOS, and increased expression of TLR4 and TGF‐β. These results suggest that 27‐OHC may cause inflammatory damage to neurons by activating the TGF‐β/NF‐κB signaling pathway and to astrocytes by activating the TLR4/TGF‐β signaling, which results in the subsequent release of inflammatory cytokines.

Increasing numbers of studies have suggested that 27-hydroxycholesterol (27-OHC), an oxidative metabolite of cholesterol, may cause inflammation in the brain and is associated with an increased risk of dementia including AD (Dias, Brown, Shabir, Polidori, & Griffiths, 2018;Testa et al., 2014Testa et al., ,2016. Our recent studies showed the toxic effect of 27-OHC in C6 glioma cells and in the brains of Sprague-Dawley rats fed a high-cholesterol diet. Our results suggest that 27-OHC may be able to inhibit cholesterol synthesis in the brain Zhang et al., 2018). In addition to being involved in cholesterol metabolism, 27-OHC is also involved in the inflammatory process. Gargiulo et al. (2015) demonstrated that 27-OHC could enhance the release of IL-1β and TNF-α through a nuclear factor-κB (NF-κB)/toll-like receptor 4 (TLR4)-dependent pathway in human promonocytic U937 cells. The expression of NF-κB and TNF-α was upregulated in retinal pigment epithelial cells treated with 27-OHC at concentrations between 10 and 20 μM (Prasanthi et al., 2009). An additional study reported that 27-OHC significantly increased the levels of NF-κB and transcription factor binding to βsite amyloid precursor protein cleaving enzyme 1 (BACE1), which in turn increased the formation of Aβ and the transcription of BACE1 in SH-SY5Y cells (Marwarha, Raza, Prasanthi, & Ghribi, 2013).
Accumulating evidence suggests that 27-OHC may be involved in the inflammatory processes that contribute to the development of AD (Testa et al., 2014(Testa et al., ,2016. However, few studies have explored the role(s) of 27-OHC in neurons and astrocytes. In the present study, to investigate the effects of 27-OHC on inflammation in neurons and astrocytes, we tested the secretion of IL-1β, IL-10, TNF-α, and iNOS, and measured the expression of toll-like receptor 4 (TLR4) and transforming growth factor β (TGF-β) and their downstream factors NF-κB and COX-2 in SH-SY5Y and C6 cells treated with 27-OHC. containing 10% fetal bovine serum and streptomycin (10 U/L) and incubated in 5% CO 2 at 37°C. All experiments were conducted under treatment with various treatments (5, 10, or 20 μM) of 27-OHC or vehicle (DMEM) for 24 hr, and cells or culture media were then collected for the indicated assay measurements. SH-SY5H and C6 cells were treated with DMEM or 27-OHC (5, 10 or 20 μM) for 24 hr and harvested for measurement. The dose range of 27-OHC was based on our previous study Ma, Li et al., 2015).

| Enzyme-linked immunosorbent assay (ELISA)
SH-SY5H and C6 cells (approximately 1 × 10 5 cells/well) were treated with DMEM or 27-OHC (5, 10 or 20 μM) for 24 hr. Supernatants from the cultures were collected and centrifuged at 800 g for 10 min. The levels of IL-1β, TNF-α, and iNOS were determined by using corresponding ELISA assay kit (Ray Biotech, USA) according to the manufacturer's instructions. The level of IL-10 was also detected with an ELISA assay kit (United States Biological, USA). In both assays, absorbance at 450 nm was measured, and the levels of IL-1β, TNF-α, iNOS, and IL-10 were calculated according to the standard curves.
The cells were then washed three times with PBS and then fixed in 4% phosphate-buffered paraformaldehyde for 30 min. Next, the cells were permeabilized with phosphate-buffered 0.1% Triton X-100 for 20 min. The permeabilized cells were then incubated with fetal bovine serum (FBS) for 1 hr and subsequently incubated in primary rabbit anti-TLR-4 (Wuhan Servicebio Technology Company, China; 1:400 dilution) or rabbit anti-TGF-β (Wuhan Servicebio Technology Company, China; 1:400 dilution) overnight at 4°C, followed by incubation in goat anti-rabbit CY3 secondary antibody (Wuhan Servicebio Technology Company, China; 1:300 dilution) at room temperature for 50 min. Immunolabeled cells were washed and mounted on gelatin-coated slides using DAPI containing antifade mounting medium for 10 min. Then, DAPI was added to stain the cell nucleus for 10 min. Finally, the slips were examined for blue and red fluorescence by using a fluorescence microscope (Nikon, Japan). Image-Pro Plus 6.0 was used to analyze the average fluorescence intensity.

| Real-time PCR analysis
The mRNA levels of NF-κB p65, NF-κB p50, and COX-2 were determined by using quantitative real-time PCR as previously described (Starr et al., 2011). Briefly, 1.0 μg of total RNA was reverse transcribed into cDNA using an M-MLV reverse transcriptase kit (Thermo Scientific, USA). The expression of NF-κB p65, NF-κB p50, and COX-2 mRNAs was quantified by real-time PCR using an ABI real-time PCR system (Applied Biosystems, USA) under standard qRT-PCR conditions. Primer sequences are provided in Table 1. The calculation of mRNA was based on the 2 −ΔΔCT method with normalization to GAPDH.

| Western blot analysis
SH-SY5Y and C6 cells were treated with either vehicle or 27-OHC (5, 10 or 20 μM) for 24 hr. Protein expression of NF-κB p50, NF-κB p65, and COX-2 was measured by Western blot as previously described (Sui et al., 2009). Cell lysates were collected using RIPA buffer, and protein concentrations were determined using a BCA protein assay kit (Pierce Biotechnology, USA). Protein samples (50 μg/well) were loaded and separated with 12% SDS-polyacrylamide gel electrophoresis and wet transferred to a polyvinylidene fluoride (PVDF) membrane at a voltage of 60 V for 2 hr. Membranes were blocked with fresh blocking buffer (Tris-buffered saline containing 5% skim milk powder) at room temperature for 1 hr. The membranes were then incubated with primary antibodies for NF-κB p50 (1:1,000, Santa Cruz Biotechnology, USA), NF-κB p65 (1:1,000, Abcam, USA), COX-2 (1:1,000, Abcam, USA), and β-actin (1:1,000, Cell Signaling, USA) overnight at 4°C, followed by incubation with corresponding secondary antibodies for 1 hr at room temperature. The blots were washed three times with TBST buffer, and protein bands were visualized by using an alkaline phosphatase reaction kit according to the manufacturer's instructions. The FluorChem FC 2 software (Alpha Innotech, USA) was used to acquire images and quantify the grayscale value of each protein. The protein expression level for each gene was normalized to the level of β-actin.

| Statistical analysis
All data were analyzed by using the SPSS 18.0 software (IBM, USA).
All data are presented as the mean ± standard deviation (SD) of at least three independent experiments. The means of different groups were compared by one-way ANOVA followed by LSD or Tamhane's post hoc analysis. A two-tailed p < 0.05 was considered statistically significant.

| The effects of 27-OHC on the expression of TLR-4 and TGF-β
Immunofluorescence staining showed that exposure to 27-OHC at 5, 10, and 20 μM did not lead to significant alterations in TLR4 expres-

| 27-OHC did not alter NF-κB p65, NF-κB p50, or COX-2 protein expression
The protein expression levels of NF-κB p65, NF-κB p50, and COX-2 were not significantly altered in both SH-SY5Y and C6  however, IL-1β was downregulated in a dose-dependent manner in astrocytes treated with 27-OHC for 24 hr. This downregulation may indicate that IL-1β was increased as an acute inflammatory factor in the brain at early time points in response to 27-OHC stimulation, but cell damage was aggravated after 24 hr when the levels of IL-1β were decreased. Rosklint, Ohlsson, Wiklund, Noren, and Hulten (2002) reported that 27-OHC dose-dependently increased the expression of IL-1β in macrophages stimulated with lipopolysaccharide (LPS), but upregulation of IL-1β was not stable without this 27-OHC-LPS costimulation. In the present study, IL-1β was increased in SH-SY5Y cells in response to 27-OHC during the first 24 hr; however, these changes were not detectable beyond 24 hr, suggesting that the release of inflammatory factors may occur in the early stages of the inflammatory reaction.
As an anti-inflammatory factor, IL-10 may inhibit the activation of macrophages and microglia and reduce the synthesis of cytokines. Lee, McGeer, and McGeer (2015) reported that IL-10 could reduce the expression of tau protein in microglia that was stimulated by LPS. However, IL-10 did not seem to play the same  Tumor necrosis factor alpha is widely expressed in microglia, astrocytes, macrophages, endothelial cells, and neurons. An unnecessary increase in TNF-α is reported to be toxic to neurons. Reinsfelt, Westerlind, Blennow, Zetterberg, and Ricksten (2013) reported that increased levels of TNF-α in serum and cerebrospinal fluid were related to the levels of Aβ. The amount of TNF-α in the brain was increased in AD rats, and the mRNA expression of TNF-α in the hippocampus was also significantly increased as well (Solmaz et al., 2015). These results indicate that TNF-α likely participates in the pathogenesis of AD. The metabolite 27-OHC has been shown to activate the phosphorylation of tau protein and upregulate the expression of TNF-α in the hippocampus of rabbits (Prasanthi, Larson, Schommer, & Ghribi, 2011). Our study shows that 27-OHC increased TNF-α secretion in neurons but decreased its secretion in astrocytes. This result suggests that increases in TNF-α expression may occur earlier in astrocytes than in neurons. Therefore, severe inflammatory reactions have already occurred in C6 cells after treatment with 27-OHC for 24 hr, and these subsequent inflammatory reactions may result in apoptosis.
Under pathological conditions, iNOS is found at increased levels in the central nervous system. Upregulated iNOS leads to the release of ROS, which can cause serious damage to neurons. Expression of iNOS is assumed to activate glial cells and reflects the extent of lipid peroxidation and generation of ROS in the body (Chen et al., 2017).
For example, the expression of iNOS was significantly higher in the rat pheochromocytoma PC12 cells than in normal cells (Kim et al., 2015). Our study indicates that 27-OHC can increase the expression of iNOS in both neurons and astrocytes, thus suggesting that 27-OHC may cause inflammation in neurons and astrocytes through iNOS-mediated mechanisms.
Toll like receptor 4 is recognized as a component of the primary innate immune receptor-mediated inflammatory signaling pathway (Li et al., 2018). Researchers demonstrated that the functional activation of TLR4 is essential for increased TGF-β mRNA expression in response to 27-OHC in epithelial cells (Pei, Lin, Song, Li, & Yao, 2008  The mRNA expression of COX-2 was measured with qRT-PCR. Data represented mean ± SD of three independent experiments. Mean value was significantly different from that of the control group *p < 0.05; **p < 0.01. COX-2: cyclooxygenase-2; NF-κB p65: nuclear factor-κB p65; NF-κB p50: nuclear factor-κB p50 signaling pathways, ultimately leading to the death of the neurons.

ACK N OWLED G M ENTS
This work was supported by grants from the State Key Program of National Natural Science of China (No. 81330065).

CO N FLI C T O F I NTE R E S T
The authors declare that they do not have any conflict of interest.

E TH I C A L S TATEM ENTS
This study does not involve any human or mammal testing. Quantification of panel c. Data represented mean ± SD of three independent experiments. COX-2: cyclooxygenase-2; NF-κB p65: nuclear factor-κB p65; NF-κB p50: nuclear factor-κB p50