If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Contact information for corresponding author: Anil G Cashikar, PhD, Department of Psychiatry, Washington University School of Medicine, 425 S Euclid Ave, BJCIH 9205, St Louis, MO 63110 USA. Phone: 314-273-1863.
Department of Psychiatry, Washington University School of Medicine, St Louis, MO 63110Hope Center for Neurological Disorders, Washington University School of Medicine, St Louis, MO 63110Taylor Family Institute for Innovative Psychiatric Research, Washington University School of Medicine, St Louis, MO 63110Department of Neurology, Washington University School of Medicine, St Louis, MO 63110
Hope Center for Neurological Disorders, Washington University School of Medicine, St Louis, MO 63110Department of Neurology, Washington University School of Medicine, St Louis, MO 63110Knight Alzheimer Disease Research Center, Washington University School of Medicine, St Louis, MO 63110
Hope Center for Neurological Disorders, Washington University School of Medicine, St Louis, MO 63110Department of Neurology, Washington University School of Medicine, St Louis, MO 63110Knight Alzheimer Disease Research Center, Washington University School of Medicine, St Louis, MO 63110
Department of Psychiatry, Washington University School of Medicine, St Louis, MO 63110Hope Center for Neurological Disorders, Washington University School of Medicine, St Louis, MO 63110Taylor Family Institute for Innovative Psychiatric Research, Washington University School of Medicine, St Louis, MO 63110Department of Neurology, Washington University School of Medicine, St Louis, MO 63110
Neuroinflammation, a major hallmark of Alzheimer’s disease and several other neurological and psychiatric disorders, is often associated with dysregulated cholesterol metabolism. Relative to homeostatic microglia, activated microglia express higher levels of Ch25h, an enzyme that hydroxylates cholesterol to produce 25-hydroxycholesterol (25HC). 25HC is an oxysterol with interesting immune roles stemming from its ability to regulate cholesterol metabolism. Since astrocytes synthesize cholesterol in the brain and transport it to other cells via apolipoprotein E (ApoE)-containing lipoproteins, we hypothesized that secreted 25HC from microglia may influence lipid metabolism as well as extracellular ApoE derived from astrocytes. Here we show that astrocytes take up externally added 25HC and respond with altered lipid metabolism. Extracellular levels of ApoE lipoprotein particles increased after treatment of astrocytes with 25HC without an increase in Apoe mRNA expression. In mouse astrocytes expressing human ApoE3 or ApoE4, 25HC promoted extracellular ApoE3 better than ApoE4. Increased extracellular ApoE was due to elevated efflux from increased Abca1 expression via LXRs as well as decreased lipoprotein reuptake from suppressed Ldlr expression via inhibition of SREBP. 25HC also suppressed expression of Srebf2, but not Srebf1, leading to reduced cholesterol synthesis in astrocytes without affecting fatty acid levels. We further show that 25HC promoted the activity of sterol-o-acyl transferase that led to a doubling of the amount of cholesteryl esters and their concomitant storage in lipid droplets. Our results demonstrate an important role for 25HC in regulating astrocyte lipid metabolism.
). Commonly recognized by the increased numbers of activated microglia and astrocytes in specific regions of the brain, neuroinflammation also entails dramatic changes in gene expression profiles of microglia and astrocytes. In AD, and in mouse models of AD, a specific activated state of microglia is commonly referred to as disease-associated microglia (DAM) (
). The gene expression signature of DAM shows down-regulation of homeostatic genes and up-regulation of several activation markers. One such activation marker is Ch25h, encoding cholesterol-25-hydroxylase. Single cell and single nuclei transcriptomics studies show myeloid cell specific overexpression of Ch25h in various mouse models of neurological disorders ((
We have recently demonstrated that Ch25h is overexpressed in AD brain as well as in mouse models of amyloid deposition and tau-mediated neurodegeneration (
cDNA cloning of mouse and human cholesterol 25-hydroxylases, polytopic membrane proteins that synthesize a potent oxysterol regulator of lipid metabolism.
). Activation of primary microglia with the TLR-4 agonist, lipopolysaccharide (LPS) also upregulates Ch25h and results in increased synthesis and release of 25HC into the extracellular milieu (
). 25HC has been shown to have anti-viral properties and exhibit context-dependent immune-modulatory effects which can be either pro- or anti-inflammatory (
) suggesting that 25HC may also exhibit neuron-specific activities. These findings illustrate that 25HC not only exhibits autocrine effects on Ch25h-expressing myeloid cells but also supports paracrine effects on neighboring cells.
However, the paracrine effects of 25HC on astrocytes are not well understood. As astrocytes are important for cholesterol synthesis in brain and are a major source of ApoE-containing lipoprotein particles central to cholesterol transport between astrocytes, neurons, and other cells (
), we posited that 25HC released from activated microglia may directly impact lipid metabolism in astrocytes and the secretion of lipoproteins containing ApoE. APOE is the most important risk factor for AD and contributes significantly to the risk of developing several other neurological and non-neurological disorders (
). A single copy of the APOE-E4 allele confers a ∼ 3.7 times greater risk than the common variant, APOE-E3, whereas homozygous APOE-E4/E4 carriers have a ∼12-15 times greater risk for developing AD. In a mouse model of amyloid deposition, ApoE has been reported to promote the localization of astrocytes to plaques (
). Astrocytes expressing APOE4 have impaired lipid metabolism and increased accumulation of lipid droplets (LDs) relative to APOE3-expressing astrocytes (
). Despite its pathophysiological roles in AD and other CNS diseases, the mechanisms underlying ApoE production and secretion in the brain are not clear. Here, we report that 25HC directly modulates astrocyte lipid metabolism via activation of LXR-mediated and inhibition of SREBP-mediated gene expression as well as by increasing cholesterol esterification. As a result, 25HC promotes extracellular ApoE lipoprotein levels via a post-translational mechanism. We also show that 25HC exerts a differential effect on extracellular levels of the human ApoE isoforms expressed in mouse astrocytes from human APOE-knockin mice. In addition, 25HC enhances esterification of cholesterol and the accumulation of lipid droplets in astrocytes.
MATERIALS AND METHODS
Chemicals
Oxysterols – 25HC (Cayman Chemical Company, 11097), 25-(C4 Top fluor®)-25HC (Avanti Polar Lipids, 810299P), 7α25diHC (Cayman Chemical Company, 11032), cholesterol (Avanti Polar Lipids, 700000), LPS (Sigma Aldrich, L2630), GSK2033 (Sigma Aldrich, SML1617), T0901317 (Cayman Chemical Company 71810). LY295427 was a kind gift from Dr. Douglas Covey at Washington University School of Medicine, St Louis, MO (
Murine versus human apolipoprotein E4: differential facilitation of and co-localization in cerebral amyloid angiopathy and amyloid plaques in APP transgenic mouse models.
). Wildtype C57BL/6J (referred to as ‘wildtype’ or WT) and B6.129S6-Ch25htm1Rus/J (referred to as ‘ch25h-/-’ or KO) mice were from Jackson Laboratories (Strain #016263). ApoE KI mice with the targeted replacement of the endogenous murine Apoe gene with human APOE-ε3 or APOE-ε4 were described earlier (
). All mice were maintained on C57BL/6J background and housed in Association for Assessment and Accreditation of Laboratory Animal Care (AAALAC) accredited facilities with ad libitum access to food and water on a 12-hour light/dark cycle. All animal procedures were approved by the Institutional Animal Care and Use Committee (IACUC) at Washington University and were in agreement with the AAALAC and WUSM guidelines.
Cells
Primary astrocyte cultures were obtained from the cortex of neonatal mice (2-3 days old). Meninges were removed and cortices dissected in cold Hanks’ Balanced Salt solution without calcium and magnesium. The tissue was digested with 0.25% trypsin at 37°C for 5 min, and the cell suspension was filtered using a 100μm sterile filter. Cells were centrifugated at 300g for 5 minutes at 23 ºC and resuspended in astrocyte medium (DMEM + 10% heat-inactivated fetal bovine serum + 1X Glutamax + 1X Sodium pyruvate and 1X Penicillin/Streptomycin – all from GIBCO). Cells were seeded in a T-75 flask previously coated with 10μg/ml poly-D-lysine (Sigma Aldrich, P7280) overnight at 37ºC. Cells were cultured for 7-10 days with media changes every 3-4 days to reach a confluent layer of astrocytes. For astrocyte cultures, loosely attached cells microglia were shaken off, followed by trypsinization and replating into desired tissue culture plates at a density of 2 X 105 cells/ml. All treatments were conducted on astrocyte serum-free media (DMEM/Neurobasal (1:1) + 1X Glutamax + 1X Sodium pyruvate + 0.1% of fatty acid-free bovine serum albumin and 1X penicillin/streptomycin – all were from GIBCO). For 25HC treatment, an identical volume of ethanol was added as Vehicle.
For microglia cultures, the procedure was similar except the culture media also contained 5ng/ml GM-CSF (GoldBio # 1320-03-5) to stimulate microglial proliferation. When large numbers of microglia were observed floating, the flasks were shaken at 200rpm for 30 minutes at 37ºC. The floating cells were collected by centrifugation and the cells were resuspended in microglia media (DMEM/F12 + 1X Glutamax + 1X Sodium pyruvate + 0.1% of fatty acid-free bovine serum albumin and 1X penicillin/streptomycin + 1X ITS supplement (R&D Systems AR013) + 25 ng/mL M-CSF (GoldBio 1320-09-10). All treatments were done in the same serum-free medium.
Fluorescent 25HC uptake
Cells were seeded at a density of 4 X 104 cells/ well in 8-Well chambered #1.5 coverglass (CellVis, C8-1.5P). A stock of 2mM 25-(C4 Top fluor®)-25HC was prepared in ethanol. Astrocyte treatment media containing 1.8μM unlabeled 25HC (90%) and 0.2μM topfluor-25HC (10%) was prepared to result in a total 25HC concentration of 2μM. Astrocyte culture media was replaced with treatment media containing 25HC. A kinetic study of 25HC uptake was conducted by fixing cells at 30 min, 1h, 2h, 4h, 1 day with paraformaldehyde (Electron microscopy Sciences, Cat. No. 15710). Cells were counterstained with DAPI and imaged on Zeiss LSM880 using identical settings at the Washington University Center for Cellular Imaging. Images were processed similarly using ImageJ (Fiji).
25HC quantitation
The amount of 25HC released into conditioned culture media by astrocytes and microglia was quantified using previously described methods (
). Briefly, conditioned media was subjected to derivatization with dimethyl glycine followed by liquid chromatography-mass spectrometry. The amount of 25HC was estimated based on an internal standard of deuterated-25HC.
Drug treatments
Mouse primary astrocytes were plated at a density of 2 X 105 cells/ml in a poly-D-lysine coated multi-well plate as needed. Cells were allowed to attach for 2 days, media were removed, and the treatments were diluted in astrocyte serum-free media. Stock solutions of 25HC, cholesterol, T0901317, and LY295427 were made in ethanol and for GSK2033 in dimethyl sulfoxide (DMSO) at 10mM final concentration. 25HC treatments were at 2μM final concentration (5000-fold dilution of the stock solution for all experiments. For vehicle control samples, an identical volume of ethanol was added resulting in a final ethanol concentration of 0.02% (∼3mM). For the 25HC concentration series, the highest concentration tested was 4μM and corresponding amounts of ethanol as vehicle was used as control. In experiments to test how 25HC influences LXR- and SREBP-pathways, LY295427 and GSK2033 were added to astrocyte serum-free media at a final concentration of 5μM. For GSK2033, DMSO was the vehicle control. Treatments were terminated after 1, 2, or 4 days by collecting media and cells for downstream processing as needed.
ApoE ELISA
ELISA for mouse and human ApoE was performed as described earlier (
). The capture antibody was HJ6.2 and HJ15.7 the detection antibody was biotinylated-HJ6.8 and biotinylated HJ15.4 for quantifying mouse and human ApoE, respectively. The ELISA plates were coated with 10μg/ml HJ15.7 (human ApoE) or 5μg/ml HJ6.2 (mouse ApoE) for 2h at room temperature. The plates were washed with PBS and blocked with 2% BSA in PBS. Samples were diluted appropriately in 1% BSA in PBS with protease inhibitors, loaded into the plates in duplicates and incubated overnight at 4ºC. The next day, the plates were washed with PBS, followed by addition of the appropriate detection antibody at 50ng/ml - biotinylated-HJ6.8 (mouse ApoE) or biotinylated HJ15.4 (human ApoE) and incubated for 1h at room temperature. The plates were washed and incubated with streptavidin-poly-HRP40 (Fitzgerald #65R-S104PHRP) for 1h at room temperature. The plates were then developed with Super Slow TMB (Sigma-Aldrich, Cat# T5569). The development was stopped using 2N sulfuric acid and the plate was read at 450nm.
Immunoblotting
For secreted proteins, media were collected after treatment of cells as required. For cellular proteins, cells (treated as appropriate) were washed with PBS and lysed with RIPA buffer (Millipore, Cat. No. 20-188) containing ProBlock Gold™ Mammalian protease inhibitor cocktail (Goldbio, GB-331-1) and Benzonase (Sigma, E1014-5KU). Samples were run on 4-15% Mini protean TGX stain free gels (Biorad, Cat. No.4568085) at 100V for 1h. Gels were transferred to PVDF membranes at 100V by 1 h. Membranes were blocked with a 5% non-fat dry milk in tris-buffered saline with 0.1% Tween-20 (TBST). Primary antibody was diluted in blocking buffer and incubated at 4ºC overnight. Blots were washed and incubated with the secondary antibody conjugated to HRP (Leinco Technologies, M114). Blots were developed with chemiluminescent development using Immobilon ECL Ultra Western HRP substrate (Millipore, CS222617) and imaged using the ChemiDoc System (Bio-Rad Labs). Band intensities were quantified using the Image Lab software (Bio-Rad Labs).
ApoE non-denaturing gradient gel electrophoresis
Cells were plated in a 6-well plate with a total density of 4 X 105 cells/well. Supernatants were collected 4 days after treatment. Protein was concentrated from the media using VivaSpin Turbo 4 (10000 MWCO) concentrator tubes (Sartorius, VS04T01). Protein samples were run on 4-15% Mini protean TGX stain-free gels (Bio-Rad) at 100V for 14h at 4ºC. Gels were transferred to PVDF membrane at 100V for 1h at 4ºC. Immunoblotting was carried out as described above.
qPCR
RNA was isolated using the Quick-RNA Miniprep Plus Kit (Zymo Research, Cat. No. R1058). cDNA was synthesized from total RNA using High-Capacity cDNA Reverse Transcription kit with RNase inhibitor (Applied Biosystems, Cat No.4374966 Foster City, CA, USA). For the qPCR reaction mix, PrimeTime probe-based qPCR assays (see Table I) and PrimeTime Gene Expression Mastermix (Cat. No.1055771) were obtained from IDT (Integrated DNA Technologies, Inc). qPCR reactions were run using the Fast mode on a QuantStudio™ 3 Real-Time PCR Instrument (Applied Biosystems by ThermoFisher, A28131). For gene expression experiments, data were normalized against actin (Actb) and the 2−ΔΔCt method was used to calculate relative gene expression value (Relative Quantity, RQ).
TABLE IProbe-based qPCR assays from IDT (Integrated DNA Technologies)
For absolute quantification of Ch25h expression in microglia and astrocytes, PrimeTime probe-based qPCR assay for Ch25h (Cat. No. Mm.PT.58.42792394.g) from IDT was used to determine a standard curve for threshold cycle (CT) values against the number of copies of Ch25h cDNA. Since Ch25h is an intronless gene, the plasmid construct with the mouse Ch25h gene in pcDNA3.1 (Genscript, OMu18523D) was used to define the standard curve. The CT values obtained for cDNA prepared from 200ng RNA isolated from equal number (2 X 105) of microglia or astrocytes were interpolated against the standard curve to determine the absolute number of copies of Ch25h in microglia and astrocytes.
Cholesterol efflux assay
To determine how 25HC influences cholesterol efflux, we used the protocol as described earlier (
). Briefly, we first loaded astrocytes for 24 h in the presence of 100 μg/ml of acetylated low-density lipoprotein from human plasma, (AcLDL, Invitrogen, L35354) previously incubated with 5 μM of Bodipy-cholesterol (Avanti Polar lipids, 810255) at 37°C for 30 minutes. Cells were equilibrated in serum free media for 2 h. To induce cholesterol efflux, cells were incubated with or without 10% mouse serum for 24h in the presence of 25HC (2μM) or corresponding vehicle control. After the efflux period media were discarded, and cells were washed with 1X PBS and detached using flow cytometry (FACS) buffer (PBS/2%FBS/1mM EDTA) for 10 min at 37°C. Cholesterol efflux was assessed directly at a single-cell level by flow cytometry using the Canto II analyzer. Fluorescence in cells without addition of mouse serum indicates “Uptake” and cells with mouse serum stimulation of efflux indicates “Efflux”. Lower levels of cellular fluorescence indicate higher efflux rates.
Lipoprotein reuptake assay
To test how 25HC influences lipoprotein reuptake, we used the LDL Uptake Cell-Based Assay kit (Cayman Chemicals Cat# 10011125) and followed manufacturer instructions. Briefly, we first treated mouse primary astrocytes with vehicle or 2μM 25HC for 24 hours and followed by addition of low-density lipoprotein (LDL) labeled with DyLight 550 to the culture media for 3 hours. At the end of the treatment, cells were washed with cold PBS, fixed with 4% paraformaldehyde, stained with DAPI and imaged using a microscope. LDL uptake was quantified using ImageJ (Fiji) by normalizing for number of cells (DAPI-positive nuclei) and expressed as LDL sum area per cell.
Lipid droplet analysis
Mouse primary astrocytes were seeded at a density of 3 X 104 cells in a 8-well chamber slide (CellVis, C8-1.5P). Oleic acid was added as Oleic acid-bovine serum albumin (Sigma O-3008) and used with cell culture BSA (Sigma A-1595) as vehicle control. To address the ability of 25HC to induce lipid droplets, the following treatments were added separately or in combination in astrocyte serum-free media: Vehicle (ethanol + BSA), 25HC (2μM), oleic acid (30μM), and the acyl-coenzyme A: cholesterol acyltransferase (SOAT) inhibitor, avasimibe (1 μM; APExBIO cat#A4318). After 24 hours, the cells were washed, fixed in 4% paraformaldehyde and stained with BODIPY 493/503 (Thermo Fisher Scientific, D3922) and DAPI. Images from three independent experiments were acquired on the Nikon Spinning Disk Confocal microscope at the Washington University Center for Cellular Imaging. Lipid droplets and cells counts were analyzed using ImageJ (Fiji).
Lipidomic analysis
1 X 106 astrocytes were grown in six 6-cm dishes and treated with Vehicle or 25HC (2μM). After 2 days, the media were collected and centrifugated at 4000 rpm x 5 min to remove cell debris. The cells were washed and collected in DPBS 1X. Cells were transferred into a microtube and the supernatant was removed after centrifugation at 10000 g X 5 min at 4ºC. Samples were frozen on dry ice and stored at -80C. Lipidomic analysis was carried out at the Functional Lipidomics Core at Barshop Institute for Longevity and Aging Studies of the University of Texas Health Science Center, San Antonio, TX. Briefly, lipid extracts of cultured cells were prepared by the procedure of Bligh-Dyer extraction (
) and a premixed solution of internal standards was added based on the protein content of cell homogenates. Individual molecular species of free fatty acids were identified and quantified after derivatization as previously described (
). The mass levels of individual cholesterol ester species were calculated from total cholesteryl ester mass and their composition was profiled using precursor-ion scanning of m/z 369. Data processing based on the principles of shotgun lipidomics such as selective ionization, low concentration of lipid solution, and correction for differential isotopologue patterns and kinetics of fragmentation was conducted as previously described (
Selection of internal standards for accurate quantification of complex lipid species in biological extracts by electrospray ionization mass spectrometry-What, how and why.
Means of 2-3 technical replicates for each biological replicate were plotted. To determine statistical significance (p value of <0.05 was considered significant; * 0.05-0.01; **0.01-0.001; ***<0.001). Student’s t test or two-way ANOVA with appropriate post hoc multiple comparisons tests (as indicated) were carried out with the help of GraphPad Prism version 9.0.
RESULTS
Inflammatory stimuli induce secretion of 25HC in microglia, but not astrocytes
Multiple studies reporting single cell-RNAseq data have shown that Ch25h is predominantly expressed in microglia in the brain. However, expression of Ch25h has also been reported in primary cultured astrocytes (
) and may indicate an altered astrocyte gene expression profile upon isolation and culture. To validate microglia-specific expression of Ch25h, we first tested for purity of primary mouse microglia and astrocytes using immunofluorescence with antibodies specific to Iba1 and GFAP (Sup Fig 1A & B). While all cells in the microglia culture were positive for Iba1 with no detectable GFAP-positive cells, nearly all cells in the astrocyte cultures were GFAP-positive with only about 1% of Iba1-positive microglia. Next, we measured the expression of Ch25h in cultured primary mouse microglia and astrocytes using a qPCR method for absolute quantitation. A standard curve was constructed with cycle threshold (CT) value against gene copy number using a known concentration of purified plasmid DNA (pCDNA3.1 with mouse Ch25h gene) as shown in Sup Fig 1C. Microglia and astrocytes were challenged with 100ng/ml LPS to stimulate transcription of inflammatory genes including Ch25h. Equal amounts of RNA (120ng) from microglia and astrocytes were used to prepare cDNA for qPCR. The CT values for the samples were compared against the standard curve to determine the expression of Ch25h between the two cell types. In microglia, Ch25h was low in the absence of LPS and increased about 100-200-fold after treatment with LPS (Fig 1A). In astrocytes, Ch25h was low in the absence of LPS and increased only marginally upon treatment with LPS, presumably resulting from the low levels of contaminating microglia (Fig 1A).
Figure 1Microglia express Ch25h and secrete 25HC in response to LPS treatment. (A) Comparison of Ch25h expression between mouse microglia and astrocytes by qPCR in response treatment with 100ng/ml LPS for 24h. Data are shown absolute copy numbers of Ch25h mRNA per 200,000 cells. A plot of CT values versus copy number is provided in Supplementary Figure 1. (p-values, * < 0.05; two-way ANOVA) (B) Measurement of 25HC in conditioned culture media from microglia or astrocytes with or without treatment with 100ng/ml LPS, as indicated. (p-values, *** <0.0002, Two-way ANOVA with Tukey test for multiple comparisons). (C) Comparison of amount of 25HC in cells versus media at 6 or 24h after treatment of wildtype microglia with vehicle (control), 10ng/ml LPS or 100ng/ml LPS. For direct comparison of the distribution of 25HC, cell extracts were prepared in a volume proportional to the volume of media used for cell culture (200,000 cells in 1ml media).
We next examined the synthesis of 25HC by microglia and astrocytes with or without LPS. Cells were treated with LPS for 24h as before. Conditioned culture media was collected, debris precleared by centrifugation and the media supernatants were subjected to derivatization by dimethyl glycine followed by liquid chromatography and mass spectrometry-based quantitation as described previously (
). While microglia produced high levels of 25HC after LPS treatment, astrocytes produced no appreciable amounts of 25HC with or without LPS (Fig 1B). High levels of 25HC produced by LPS-treated microglia reflected the high levels of Ch25h expression, demonstrating a causal relationship between Ch25h expression and 25HC production in microglia. Therefore, we conclude that inflammatory stimuli triggered overexpression of Ch25h and production of 25HC by microglia, but not astrocytes. A very low level of Ch25h expression could be observed in our primary cultures of isolated neonatal mouse astrocytes most likely due to contaminating microglia. However, no 25HC production and secretion was detected from astrocytes with or without endotoxin stimulus.
It has been suggested that 25HC may be synthesized by enzymes other than Ch25h (
). To test this, we examined 25HC synthesis after LPS challenge in microglia from wild-type and ch25h-/- mice (Fig 1C). Increasing amounts of 25HC as a function of time were observed exclusively in media from wild-type microglia, whereas no 25HC was detectable in media from ch25h-/- microglia. Thus, microglia from ch25h-/- mice produce no 25HC as previously demonstrated (
We next examined the proportion of 25HC that was secreted into the culture media relative to that retained within microglial cells. To do this, we prepared cell extracts after collecting the conditioned media. The volume of extraction buffer used was equal to the volume of the conditioned culture media to facilitate a direct comparison of 25HC levels. About 30-40% of 25HC produced by microglia after exposure to LPS was released extracellularly into the media while retaining 60-70% of it within cells (Fig 1D). This distribution is in agreement with that observed earlier for ApoE2 and ApoE4-expressing microglia (
), suggesting that microglial 25HC not only elicit autocrine effects within microglia but may also exert paracrine effects on surrounding cells such as neurons, astrocytes, and other cells to facilitate inter-cellular crosstalk.
Astrocytes internalize 25HC
To test whether cultured astrocytes could take up and internalize 25HC, astrocytes were treated with 2μM 25HC mixed with 10% (0.2μM) BODIPY-labeled fluorescent 25HC (TopFluor-25HC) in culture media, and the uptake of fluorescent 25HC was measured over a period of 24 hours (Fig 2A). Very little fluorescence, if any, could be detected at 2h. By 4h, a general increase in fluorescence intensity outlining the cells was visible suggesting that the 25HC was present in the plasma membrane. By 24h, 25HC was clearly present in intracellular vesicles in virtually every cell. To confirm whether astrocytes were able to take up unlabeled 25HC added externally into the culture media in the same timeframe, we quantified 25HC remaining in the culture media after 24 hours of contact with astrocytes. Interestingly, when astrocytes were treated with 1μM or 2μM 25HC, only 44.9nM (± 1.1nM) and 84.0nM (± 8.6nM) 25HC remained in the conditioned media after 24 hours (Fig 2B). These results suggest that nearly 95% of the externally added 25HC was readily taken up by astrocytes within 24h.
Figure 2Uptake of 25HC by astrocytes. (A) Fluorescence microscopy of astrocytes treated with 2μM 25HC (containing 25HC (1.8μM) and TopFluor-25HC (0.2μM)) for various times. Time-dependent increase in fluorescence intensity of cells within intracellular vesicles was observed within 1 day. (B) On the left is a schematic of the experiment showing the addition of 25HC to a final concentration of 1μM or 2μM to astrocytes. Uptake of 25HC by astrocytes was monitored using LCMS to measure 25HC remaining (detected) in the culture media after 1 day of incubation on the right.
), we tested whether 25HC impacts extracellular ApoE from astrocytes. Astrocytes were incubated with 25HC, 7α,25-dihydroxycholesterol (7α25diHC), cholesterol, T0901317 (a synthetic LXR agonist) or vehicle (ethanol) and the ApoE in the culture media accumulating over 48 hours was measured by ELISA (Fig 3A). While cholesterol or ethanol (vehicle control) had no effect on extracellular ApoE by astrocytes, 25HC increased extracellular ApoE levels by about 200%. The stimulatory effect of 25HC was comparable to an established LXR agonist such as T0901317. Interestingly, 7α25diHC, an oxysterol synthetized from 25HC by the enzyme Cyp7b1 also showed no stimulatory effect on extracellular levels of ApoE indicating that the effect of 25HC was not due to its metabolism.
Figure 3Figure 3: 25HC increases extracellular ApoE in astrocytes without altering Apoe mRNA expression. (A) Astrocytes were incubated with 2μM final concentration of 25HC, 7α25diHC, cholesterol, T0901317 or with an appropriate amount of vehicle (ethanol). ApoE was measured by ELISA in the conditioned media after 2 days. (p-values ****<0.0001 – Two-way ANOVA with Sidak’s multiple comparisons test) (B) Effect of 25HC concentration and time on the accumulation of extracellular ApoE from astrocytes. (C) Expression of Apoe mRNA was measured by qPCR in astrocytes treated with 2μM each of cholesterol, 25HC or T0901317 or an appropriate amount of vehicle (ethanol) or without any treatment (none). Data was normalized to mouse Actb (actin B) as the endogenous control gene and shown as relative expression compared to ‘None’. (*0.05-0.01; One-way ANOVA with Dunnett’s multiple comparisons test). (D) Immunoblots of cell lysates from astrocytes treated with 0, 1, 2 or 5μM 25HC for 48 hours. Antibodies for ApoE, ApoJ or actin are shown on the left of corresponding blots. Bar graphs below the blots show band quantitation after baseline correction normalized for actin band intensity. (E) Immunoblots of conditioned media from astrocytes treated with 0, 1, 2 or 5μM 25HC for 48 hours. Antibodies for ApoE or ApoJ are shown on the left of corresponding blots. Bar graphs below the blots show band quantitation after baseline correction. (F) Immunoblot for ApoE of conditioned media from astrocytes treated with 0, 1, 2 or 5μM 25HC for 48 hours, and samples were run on native polyacrylamide gel electrophoresis. (G) Extracellular cholesterol in serum-free conditioned media from astrocytes treated with vehicle (black bars) or 2μM 25HC (grey bars). (H) Extracellular ApoE in mouse astrocytes from knockin mice expressing ApoE3 or ApoE4. (I) Expression of APOE was measured by qPCR in astrocytes treated with vehicle (ethanol) or 2μM 25HC and normalized to mouse Actb (actin B) (p-values * 0.05-0.01; **0.01-0.001; ***<0.001; ****<0.0001 – Two-way ANOVA with Dunnett’s multiple comparisons test).
We next tested the effect of 25HC at different concentrations over time. 25HC increased ApoE levels derived from astrocytes in a concentration- and time-dependent manner. At earlier time points (1-day), basal secreted levels of ApoE in conditioned media were low under control conditions and increased by 4-5 fold over 2 days (Fig 3B). The amount of extracellular ApoE increased 2-3-fold following treatment with 25HC (2μM) relative to basal levels at both time points.
To better understand the cellular mechanisms responsible for the increased extracellular ApoE in 25HC-treated astrocytes, we isolated RNA from astrocytes treated for 1 day and examined Apoe mRNA levels by qPCR. 25HC failed to alter the expression of Apoe mRNA at any time. Treatment of astrocytes with vehicle (ethanol) or cholesterol did not alter the expression of Apoe mRNA (Fig 3C). A synthetic non-steroidal LXR agonist, T0901317, increased Apoe mRNA expression by 1.5-fold.
We further examined the amount of ApoE in cell lysates by immunoblotting and compared this with a control protein such as actin B or another apolipoprotein, ApoJ. None of these other astrocytic proteins were altered by 25HC when measured in cell lysates (Fig 3D). We next tested conditioned media for secreted ApoE and ApoJ. 25HC increased extracellular ApoE but not ApoJ suggesting that the effect was specific for ApoE (Fig 3E). These observations show that the observed increase in the levels of extracellular ApoE following 25HC treatment occurs without substantially altering intracellular levels of ApoE.
Next, we examined whether the increase in extracellular ApoE by 25HC was associated with an alteration in ApoE quaternary structure. To this end, we concentrated the variously treated conditioned media and separated the secreted ApoE by non-denaturing (native) gel electrophoresis followed by immunoblotting as described previously (
). While each band on the native blot intensified with increasing 25HC concentration, the relative proportions of each of the bands were not different between 25HC-treated and vehicle-treated samples (Fig 3F). This suggests that the extracellular ApoE following 25HC treatment of astrocytes was not associated with a change in quaternary structure. Free cholesterol levels in the conditioned media containing ApoE also increased (Fig 3G) suggesting that the increased extracellular ApoE was lipidated.
To test whether 25HC also increases extracellular levels of human ApoE isoforms, ApoE3 and ApoE4, we treated astrocytes from human APOE knock-in mice (
) with various concentrations of 25HC for 48 hours. Extracellular human ApoE quantified by ELISA (Fig 3H) indicated that 25HC strongly promoted an increase of extracellular ApoE3 by 2-fold and 3-fold at 2μM and 4μM 25HC, respectively. On the other hand, 25HC had a much weaker influence on extracellular ApoE4 – Extracellular ApoE4 levels increased marginally at 2μM 25HC and by about 1.4-fold by 4μM 25HC. Negative stain electron microscopy analyses of ApoE immuno-purified using an antibody specific for human ApoE (HJ15.4) from conditioned culture media of mouse astrocytes expressing human ApoE3 or ApoE4 showed similar lipoprotein particles of about 20nm diameter for ApoE3 and ApoE4 with or without 25HC (Sup Fig 2 and Sup Table I) suggesting similar lipidation levels. While the electron micrographs are also suggestive of 25HC-mediated increase in ApoE3 and ApoE4 containing lipoproteins, quantitation of this increase would be inappropriate due to the immuno-isolation step. As in the case of mouse Apoe (Fig 2C), 25HC failed to influence expression of human APOE3 or APOE4 in the mouse astrocytes from the human APOE-knockin mice (Fig 3I) suggesting post-transcriptional control of extracellular ApoE3 and ApoE4.
25HC upregulates Abca1 and downregulates Ldlr expression
We next assessed the key genes responsible for ApoE efflux and reuptake by astrocytes by focusing on Abca1, which is central to the efflux of ApoE lipoproteins and Ldlr, which facilitates its reuptake (
), 25HC may increase the expression of Abca1, a gene controlled by LXRs. 25HC is also a strong suppressor of the sterol regulatory element binding protein (SREBP)-mediated transcription of genes including Ldlr (
). By qPCR, we observed that 25HC increased Abca1 expression after 1 day of treatment by about 2.5-fold (p< 0.01), similar to a 3-fold increase observed with the synthetic LXR agonist, T0901317 (Fig 4A). Vehicle (ethanol) had no significant effect on Abca1 expression. On the other hand, 25HC strongly suppressed the expression of Ldlr to about 10% of control (p < 0.001), while treatments with vehicle or T0901317 showed no effect on Ldlr expression (Fig 4B). Similar changes in Abca1 and Ldlr mRNA was observed in mouse astrocytes expressing ApoE3 and ApoE4 (Sup Fig 3), suggesting that differences in the expression of efflux and reuptake genes did not account for the differences in extracellular ApoE3 and ApoE4 (ApoE3>ApoE4). We confirmed the effects of 25HC on Abca1 and Ldlr mRNA expression by immunoblotting for the respective proteins in mouse astrocytes. In accordance with the mRNA expression levels Abca1 protein levels increased, while Ldlr protein levels decreased following 25HC treatment (Fig 4C).
Figure 425HC increases cholesterol efflux and decreases lipoprotein reuptake. (A & B) Expression of Abca1 (A) and Ldlr (B) mRNA in response to treatment of astrocytes with vehicle, 25HC (2μM) or T0901317 (2μM) was assessed by qPCR. Data is normalized to Actb (actin B) as the endogenous control gene and shown as relative expression compared to ‘vehicle’. (*0.05-0.01; **0.01-0.001; ***<0.001; ****<0.0001; Two-way ANOVA with Sidak’s multiple comparisons test). (C) Immunoblots of cell lysates from astrocytes treated with 0, 1, 2 or 5μM 25HC for 48 hours. Antibodies for Abca1, Ldlr or actin are shown on the left of corresponding blots. Bar graphs below the blots show band quantitation after baseline correction and normalized for actin band intensity. (D) Schematic diagram showing the action of inhibitors to block the effect of 25HC on LXR and SREBP pathways. (E, F & G) Expression of Abca1 (E), Ldlr (F) and Apoe (G) mRNA in response to treatment of astrocytes with vehicle or 25HC (2μM) in the presence of vehicle (DMSO), 5μM LY295427 or 5μM GSK2033 was assessed by qPCR. Data is normalized to Actb (actin B) as the endogenous control gene and shown as relative expression compared to ‘DMSO’. (*0.05-0.01; **0.01-0.001; ***<0.001; ****<0.0001; Two-way ANOVA with Sidak’s multiple comparisons test). (H) ApoE in the conditioned media of astrocytes treated with vehicle or 25HC (2μM) in the presence of vehicle (DMSO), 5μM LY295427 or 5μM GSK2033 was measured by ELISA in the conditioned media after 2 days. (I) Fluorescence of BODIPY-cholesterol within astrocytes quantified by FACS at the uptake step and the efflux steps after treatment with vehicle (ethanol) or 2μM 25HC for 24h. Lower fluorescence within cells indicate higher efflux. (J) The amount of TopFluor-LDL uptaken by astrocytes treated with vehicle (ethanol) or 2μM 25HC for 24h was measured by microscopy as the fluorescence sum area in the field and normalized to the total number of cells in the field. (*0.05-0.01; **0.01-0.001; ***<0.001; ****<0.0001; Two-way ANOVA with Sidak’s multiple comparisons test).
25HC-mediated upregulation of LXR-mediated and suppression of SREBP-mediated gene expression are important for ApoE secretion
To determine whether the effects of 25HC on LXR- and SREBP-mediated gene expression are responsible for the increase in extracellular ApoE, we tested the effect of drugs that interfere with each pathway (Fig 4D). To inhibit the effect of 25HC on the LXR pathway, we treated astrocytes with 25HC in the presence of the LXR antagonist, GSK2033 (
). In the presence of GSK2033, very little expression of Abca1 was observed with or without 25HC treatment (Fig 4E) supporting earlier observations that Abca1 expressed in astrocytes is LXR-dependent. Interestingly, LY295427 reversed 25HC-dependent upregulation of Abca1 expression (Fig 4E). Expression of Ldlr was suppressed by 25HC as expected and this suppression was unaffected by the presence of GSK2033 (Fig 4F). To interfere with 25HC’s ability to suppress the SREBP pathway, we tested LY295427, a hypocholesterolemic agent that reduces plasma cholesterol levels by increasing Ldlr expression in the liver (
). LY295427 also reversed 25HC-dependent suppression of Ldlr expression (Fig 4F). These effects of LY295427 may be related to increased expression of Insig1 as previously observed in cells treated with LY295427 (
Neither GSK2033 nor LY295427 had any effect on Apoe gene expression (Fig 4G). However, when the level of ApoE measured in the conditioned media was monitored by ELISA (Fig 4H), both drugs reduced ApoE levels to about 50% of that of control (vehicle-treated) astrocytes. On the other hand, in the presence of 25HC, both drugs eliminated 25HC-stimulated ApoE levels. These results suggest that interfering with either the LXR pathway or the SREBP pathway eliminates the stimulatory effect of 25HC on extracellular ApoE levels in astrocytes.
25HC modulates efflux and reuptake of ApoE lipoprotein
To determine whether the increased extracellular ApoE resulted from increased efflux (secretion) or from decreased reuptake of lipoproteins, we examined the effect of 25HC on efflux and uptake of lipoproteins. Towards this goal we first fed astrocytes with BODIPY-labeled cholesterol for 24h. The cells were treated with vehicle or 25HC together with or without addition of mouse serum as an acceptor of effluxed BODIPY-cholesterol. Fluorescence of BODIPY-cholesterol retained in cells was quantified by flow-cytometry. In the absence of mouse serum (termed as “Uptake”), no differences were observed between cells treated with vehicle or 25HC. In the presence of mouse serum as acceptor (termed as “Efflux"), fluorescence within cells decreased significantly after treatment with 25HC compared to vehicle control (Fig 4I). This suggests that 25HC treatment of cells increased efflux of cholesterol.
Next, to examine how 25HC treatment influences reuptake of lipoproteins as a result of decreased Ldlr, we treated astrocytes with vehicle or 25HC for 24 hour and allowed them to uptake fluorescently labeled LDL for 4 hours. The amount of fluorescence within cells was quantified using microscopy. Treatment of astrocytes with 25HC markedly reduced uptake of labeled LDL relative to vehicle control (Fig 4J). Together, we conclude that 25HC not only increased ApoE-lipoprotein efflux from astrocytes but also reduced lipoprotein reuptake.
25HC treatment reduces cholesterol biosynthesis in astrocytes
Astrocytes are major producers of cholesterol and other lipids in the CNS. 25HC is a well-characterized oxysterol regulator of the SREBP pathway in peripheral tissues. However, its importance in the CNS has not been studied. We examined how 25HC influences lipid and cholesterol biosynthetic genes in astrocytes compared to a synthetic LXR agonist, T0901317 (Fig 5A-D). Expression of Srebf2 (encoding SREBP2, the transcription factor regulating sterol synthesis genes) was reduced to 50% following 25HC treatment of astrocytes for 1 day, whereas T0901317 did not affect Srebf2 expression (Fig 5A). Similarly, while 25HC also reduced the expression of Insig1 (a major regulator of the SREBP pathway) to about 20%, T0901317 marginally increased Insig1 (Fig 5B). Interestingly, while 25HC had no effect on the expression of Srebf1 (encoding SREBP1, the transcription factor regulating fatty acid synthesis genes), T0901317 increased Srebf1 by about 50% (Fig 5C). Likewise, while 25HC did not affect the expression of Fasn (fatty acid synthase), T0901317 increased Fasn by 250% (Fig 5D). 25HC did not affect the expression of Srebf1 and Insig2.
Figure 525HC suppresses sterol biosynthesis genes in astrocytes. Expression of Srebf2 (A), Insig1 (B), Srebf1 (C) and Fasn (D) mRNA in response to treatment of astrocytes with vehicle, 25HC (2μM) or T0901317 (2μM) for 1 day was assessed by qPCR. Data was normalized to Actb (actin B) as the endogenous control gene and shown as relative expression compared to ‘vehicle’. (E) Schematic showing selected steps of cholesterol biosynthesis and the enzyme involved. (F) Expression of Acat2, Hmgcr, and Dhcr24 in response to treatment of astrocytes with vehicle (black bars) or 2μM 25HC (grey bars) for 1 day was quantified by qPCR. Data was normalized to Actb (actin B) as the endogenous control gene and shown as relative expression compared to ‘vehicle’. Quantitation of total free fatty acids (G) and free cholesterol (H) in mouse astrocytes treated with vehicle (black bars) or with 2μM 25HC (grey bars).(*0.05-0.01; **0.01-0.001; ***<0.001; ****<0.0001; Two-way ANOVA with Sidak’s multiple comparisons test).
We next examined the expression of three key genes - namely Acat2, Hmgcr and Dhcr24 -coding for enzymes required for cholesterol biosynthesis as outlined in Fig 5E. Acat2 codes for acetyl-coA acyltransferase-2, which catalyzes the first step of cholesterol synthesis. Hmgcr codes for 3-hydroxy-3-methylglutaryl-coA reductase, an enzyme catalyzing a key step in the synthesis of mevalonate. Dhcr24 codes for 24-dehydrocholesterol reductase, which catalyzes the last step in cholesterol biosynthesis. The sterol synthesis genes Acat2, Hmgcr and Dhcr24 were markedly decreased after 1 day of treatment with 25HC (Fig 5F).
Considering the dramatic changes induced by 25HC in cholesterol biosynthesis genes, we next examined whether the changes in gene expression also resulted in corresponding changes in lipids. Astrocytes treated with 25HC or vehicle were tested for changes in free fatty acids (Fig 5G) and free cholesterol (Fig 5H). No difference was observed in free fatty acid levels (Fig 5G) in agreement with a lack of effect of 25HC on Srebf1 and Fasn (Fig 5C &D). On the other hand, free cholesterol levels were reduced by nearly 10% in 25HC-treated astrocytes relative to vehicle-treated astrocytes (Fig 5H) reflecting the strong reductions in the expression of Srebf2 and cholesterol biosynthetic enzymes (Fig 5E & F).
Augmentation of cholesterol esterification and lipid droplet accumulation by 25HC
Cholesterol ester formation is greatly enhanced by 25HC by the activation of fatty acyl coA:cholesterol acyl transferase enzyme in human fibroblasts (
) and other cells. To determine whether 25HC also enhanced cholesterol esterification in astrocytes, we examined the intracellular levels of cholesteryl esters (CE) in mouse astrocytes treated with vehicle or 25HC for 24 hours. Interestingly, the amount of CE dramatically increased in astrocytes treated with 25HC relative to vehicle (Fig 6A). Deeper analysis of CE revealed that 25HC significantly increased the amounts of CE14:0, CE16:1, CE16:0, CS18:2, CE18:1, CE20:4 and CE22:5, suggesting that 25HC promoted cholesterol esterification with both saturated as well as mono- or poly-unsaturated fatty acids (Fig 6B). Previous studies have shown that 25HC activates cholesterol esterification by sterol-O-acyl transferase (SOAT; also known as acyl-coA cholesterol acyltransferase, ACAT) (
). We tested whether 25HC also upregulated expression of the two SOATs - Soat1 and Soat2. While 25HC did not affect expression levels (relative to vehicle control) of Soat1 (Fig 6C), it increased expression of Soat2 (Fig 6D). However, the average CT values for Soat2 were very high (>35) suggesting the expression levels were very low. The low Soat2 expression levels and higher Soat1 expression levels agree with publicly available databases for expression of these genes in astrocytes in the brain (
Figure 625HC increases cholesteryl esters and lipid droplet accumulation in astrocytes. Quantitation of total cholesterol esters (A) in mouse astrocytes treated with vehicle (black bars) or with 2μM 25HC (grey bars). Detailed analyses of cholesteryl esters (B) are shown. Expression of Soat1 (C) and Soat2 (D) in astrocytes treated with vehicle (black bars) or 2μM 25HC (grey bars) was measured by qPCR and normalized to Actb as before. (E) Mouse astrocytes treated with vehicle (ethanol), 2μM 25HC, 30μM OA or OA+25HC were stained for lipid droplets with BODIPY (green) and DAPI (blue). Top row of images is without the SOAT/ACAT inhibitor, avasimibe and bottom row had 1μM avasimibe. Bar = 100μm. (F & G) Quantification of the microscopy data in E and expressed as total LD counts/total nuclei. (H) Extracellular ApoE levels quantified by ELISA. Light grey bars are without avasimibe and dark grey bars are with avasimibe. (*0.05-0.01; **0.01-0.001; ***<0.001; ****<0.0001; Data compared with unpaired t-tests in A. For all others two-way ANOVA with Sidak’s multiple comparisons test).
Since cholesterol esters are mostly stored in lipid droplets, we hypothesized that 25HC might promote lipid droplet (LD) biogenesis in astrocytes, which has been thought to play an important role in astrocyte function and neuroinflammation (
). To test this, we treated astrocytes with 25HC in the absence or presence of oleic acid (OA), a well-studied fatty acid enhancer of LD biogenesis as well as with avasimibe, an inhibitor of SOAT/ACAT. LDs were detected using BODIPY 493/503 dye which stains neutral lipids (Fig 6E). In the absence of OA, 25HC increased LD counts in astrocytes and this increase was reversed by avasimibe (Fig 6E & F) suggesting that the increased cholesteryl esters in 25HC-treated astrocytes were stored in lipid droplets. OA provides the acyl groups required for the formation of triacylglycerols (TAGs) from diacyl glycerides (DAGs) as well as for the formation of cholesteryl esters, which are key neutral lipid components of LDs. Incubation of astrocytes with OA, increased LDs by nearly 10-fold (compare vehicle control in Fig 6F with OA in Fig 6G). Interestingly, coincubation of astrocytes with OA and 25HC, nearly doubled the LD content from that resulting from OA alone (Fig 6E & G). This suggests that in 25HC-treated astrocytes OA further enhanced cholesteryl esters. While avasimibe had no effect on LDs induced by OA alone, it eliminated the LD-stimulatory effect of 25HC and reverted the levels to that of OA alone. Neither additional of OA nor inhibition of cholesterol esterification by avasimibe altered basal or 25HC-stimulated levels of extracellular ApoE (Fig 6H). These results suggest that cholesteryl esters produced in response to treatment of astrocytes with 25HC were stored in LDs but not in ApoE-containing lipoproteins at basal levels as well as when co-treated with an acyl-donor such as OA.
To investigate how 25HC affects LD accumulation in astrocytes expressing human ApoE3 and ApoE4 isoforms, we treated primary astrocytes from knock-in mice expressing human APOE3 or APOE4 with vehicle or 25HC and examined intracellular LD (Sup Fig 4). Contrary to the differential secretion of ApoE lipoproteins between ApoE3- and ApoE4-expressing astrocytes (Fig 3H), no difference was obvious in LD accumulation between ApoE3- and ApoE4-expressing astrocytes at any concentration of 25HC tested (Sup Fig 4).
DISCUSSION
In this study, we have provided evidence that 25HC, a cholesterol metabolite produced almost exclusively by activated microglia alters lipid metabolism in astrocytes (Fig 7). We have shown that 25HC alters astrocyte lipid homeostasis in various ways including altering cholesterol metabolism, lipoprotein efflux and reuptake as well as lipid droplet accumulation. Following treatment of mouse astrocytes with physiologically relevant concentrations of 25HC (
), increased secretion as well as decreased reuptake of ApoE lipoproteins resulted in higher levels of extracellular ApoE. Genes required for cholesterol biosynthesis were also strongly downregulated in astrocytes by 25HC. These effects resulted from upregulation of the LXR-mediated gene expression and downregulation of SREBP-mediated gene expression by 25HC. 25HC also led to a marked increase in cholesteryl ester levels and its storage in lipid droplets within astrocytes.
Figure 7Lipid metabolism and ApoE secretion in astrocytes is modulated by 25HC, a cholesterol metabolite secreted by activated microglia. Top: Overexpression of cholesterol 25-hydroxylase (Ch25h) in microglia activated by inflammatory stimuli converts cholesterol to 25-hydroxycholesterol (25HC). Paracrine effects of secreted 25HC affects astrocyte lipid metabolism and ApoE secretion. Bottom: Schematic showing different ways 25HC alters cholesterol metabolism in astrocytes. 25HC inhibits the SREBP2 pathway to suppress cholesterol biosynthesis by reducing the expression of cholesterol biosynthetic enzymes such as Hmgcr. By suppressing LDL-receptor (Ldlr), 25HC also reduces cholesterol reuptake. Activation of LXRs by 25HC promotes expression of Abca1 that increase cholesterol efflux. Decreased reuptake of lipoproteins as well as increased secretion of lipoproteins result in higher levels of extracellular ApoE. Finally, 25HC enhances the activity of the cholesterol acyl transferase enzyme, Acat1 to promote cholesterol esterification. Thus, 25HC effectively reduces free cholesterol in astrocytes. This figure was made using BioRender.
). However, we have shown that astrocytes readily take up and accumulate 25HC suggesting that the intracellular concentrations of 25HC may be relatively high. Expression of APOE is controlled by LXRs (
24(S)-hydroxycholesterol participates in a liver X receptor-controlled pathway in astrocytes that regulates apolipoprotein E-mediated cholesterol efflux.
). However, our studies with primary mouse astrocytes show that Apoe gene expression is only weakly stimulated by the synthetic LXR agonist, T0901317. Further, Apoe expression is neither elevated by 25HC treatment nor inhibited by treatment with the LXR antagonist, GSK2033. Expression of human APOE are also unaffected by 25HC in mouse astrocytes expressing APOE3 or APOE4 isoforms. This suggests that expression of mouse Apoe mRNA or human APOE mRNA does not appear to be strongly dependent on LXRs in mouse astrocytes and in contrary human-derived cells. However, we observed that 25HC significantly increases the expression of Abca1, a well-known LXR target gene (
) and that blocking LXR activity with GSK2033 reduced Abca1 expression in the presence or absence of 25HC. Interestingly, we find that 25HC produced primarily by activated microglia differentially affects extracellular levels of human ApoE3 and ApoE4 without differentially affecting the lipidated form of ApoE or the expression of APOE, Abca1 or Ldlr. This suggests that the lower extracellular levels of ApoE4-containing lipoproteins relative to ApoE3-containing lipoproteins in response to 25HC treatment may involve other unknown mechanisms.
Astrocytes are a major cell type in the brain for cholesterol biosynthesis, which is controlled by Srebf2 (
) and may play an important role in the pathogenesis of AD. Cholesterol biosynthesis controlled by the SREBP-pathway encompasses the following key steps: (i) when cholesterol levels are low Scap-Srebp2 complex migrates from the ER to the Golgi via COP-II-dependent trafficking; (ii) in the Golgi Srebp2 is proteolytically cleaved by the S1P and S2P proteases to generate the cytosolic Srebp2 fragment; (iii) the transcriptionally active Srebp2 fragment translocates to the nucleus to stimulate transcription of cholesterol synthesis genes. Polymorphisms in the SREBF2 gene (rs2269657) have been reported to be associated with the risk of developing late onset AD and SREBF2 mRNA was reported to be significantly higher in AD temporal cortex and cerebellum (
). In mice, astrocyte-specific loss of Srebf2 resulted in reduced cholesterol synthesis and defects in neurite outgrowth, behavior, and energy metabolism (
). Treatment of primary astrocytes with 25HC resulted in markedly reduced expression of Insig1 as well as Srebf2, which encodes the Srebp2 protein itself. When 25HC binds to Insig1, it also stabilizes the protein by reducing proteasomal degradation of Insig1 thus promoting its interaction with Scap-Srebp2 (
), expression of Insig1 is also reduced in the presence of 25HC. The intricate control of Insig1 gene expression, protein stability and function establishes a feedback mechanism by which cholesterol biosynthesis is regulated by 25HC.
Following treatment of mouse astrocytes with 25HC, we observed a dramatic decrease in the expression of Ldlr, a gene expressed under the control of the SREBP pathway. Ldlr is a major receptor for the reuptake of ApoE lipoproteins (
The low density lipoprotein receptor regulates the level of central nervous system human and murine apolipoprotein E but does not modify amyloid plaque pathology in PDAPP mice.
). Ldlr is normally localized to the plasma membrane where it binds extracellular ApoE lipoproteins followed by endocytosis and subsequent release of bound lipids in the lysosome (
). On the other hand, by reducing Ldlr levels, 25HC markedly limits reuptake of ApoE resulting in greater amounts of extracellular ApoE. A previous study showed that LY295427 reverses the effect of 25HC on the SREBP-pathway by elevating the expression of Ldlr (
) presumably by creating a sink for 25HC. In agreement, we observed that 25HC failed to suppress Ldlr expression in astrocytes treated with LY295427. However, it is unclear whether LY295427 also reverses the suppression of Srebf2 expression by 25HC. Taken together, microglial 25HC overproduced under neuroinflammatory conditions may be a potentially important feedback regulator of the SREBP machinery in astrocytes leading to reduced cholesterol biosynthesis as well as lipoprotein reuptake.
25HC also nearly doubled levels of cholesterol esters and their storage in LDs in astrocytes. Storage and utilization of lipids in LDs in astrocytes are important for astrocyte lipid homeostasis and function (
). Our data suggest a role for activated microglia in promoting lipid dysregulation in astrocytes. The increase in 25HC-induced LDs could be blocked by an inhibitor of SOAT/ACAT suggesting the accumulation in CE-rich lipid droplets in 25HC-treated astrocytes. Unlike triglyceride-enriched LDs, CE-enriched LDs are thought to play an important role in the biosynthesis of steroid hormones (
). It will be interesting to examine whether mice lacking Ch25h are defective in the production of neurosteroids.
With increased cholesterol efflux, increased cholesterol esterification and decreased cholesterol biosynthesis, the net effect of 25HC on astrocytes is a reduction in overall free cholesterol levels by about 10% in astrocytes as we empirically established. Over long periods, these cholesterol-reducing effects are likely to be detrimental to cell survival.
The results presented herein highlight a potential role for 25HC, the immune-oxysterol produced by microglia to regulate extracellular levels of ApoE as well as cholesterol metabolism in astrocytes. We postulate that 25HC may function as a potential mediator of microglia-astrocyte crosstalk leading to dysregulation of astrocyte lipid metabolism under conditions of neuroinflammation. It will be interesting to examine the effects 25HC in in vivo models of neuroinflammation and neurodegeneration.
DATA AVAILABILITY
All data are contained within the manuscript.
Funding Sources
This research is supported by funding from Centene Corporation contract (P19-00559) for the Washington University-Centene ARCH Personalized Medicine Initiative to AGC and SMP and by a grant from the National Institute on Aging of the National Institutes of Health under Award Number U19AG069701 (Project 2: AGC, SMP, and DMH and Core F: XH). Support to XH for the Functional Lipidomics Core (P30 AG013319 and P30 AG044271) is acknowledged. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.
SUPPORTING INFORMATION
This article contains supporting information.
AUTHOR CONTRIBUTIONS
AGC and SMP, conceived the study; AGC, SMP and DMH, obtained funding; AGC, DTR, DT, JR, MS and XH, performed experiments and analyzed data; DTR, JML, DMH and SMP, critically reviewed the manuscript; AGC, wrote the manuscript
CONFLICT OF INTEREST
SMP is a co-founder, board member and shareholder of Sage Therapeutics, Voyager Therapeutics and Alnylam Pharmaceuticals. He is also a board member and shareholder of Karuna Pharmaceuticals and a Venture Partner at Third Rock Ventures. D.M.H. co-founded and is on the scientific advisory board of C2N Diagnostics. D.M.H. is on the scientific advisory board of Denali and Cajal Neuroscience and consults for Genentech and Alector.
ACKNOWLEDGEMENTS
We wish to acknowledge the gift of LY295427 from Dr. Doug Covey, Department of Developmental Biology, Washington University School of Medicine, St Louis, MO.
cDNA cloning of mouse and human cholesterol 25-hydroxylases, polytopic membrane proteins that synthesize a potent oxysterol regulator of lipid metabolism.
Murine versus human apolipoprotein E4: differential facilitation of and co-localization in cerebral amyloid angiopathy and amyloid plaques in APP transgenic mouse models.
Selection of internal standards for accurate quantification of complex lipid species in biological extracts by electrospray ionization mass spectrometry-What, how and why.
24(S)-hydroxycholesterol participates in a liver X receptor-controlled pathway in astrocytes that regulates apolipoprotein E-mediated cholesterol efflux.
The low density lipoprotein receptor regulates the level of central nervous system human and murine apolipoprotein E but does not modify amyloid plaque pathology in PDAPP mice.
00023-8&id=gr1.jpg){kind=link}
00023-8&id=gr2.jpg){kind=link}
00023-8&id=gr3.jpg){kind=link}
00023-8&id=gr4.jpg){kind=link}
00023-8&id=gr5.jpg){kind=link}
00023-8&id=gr6.jpg){kind=link}
00023-8&id=gr7.jpg){kind=link}