Human neuroblastoma SH-SY5Y cells were treated with 5, 10 and 25 μM of 24-OHC, 27-OHC, or a mixture of 24-OHC and 27-OHC, and Aβ42 levels were determined with ELISA. ELISA measurements showed that treatment with 5, 10 or 25 μM 24-OHC did not induce significant changes in secreted Aβ42 levels compared to levels in medium of untreated cells (Fig. 1a). Conversely to 24-OHC, treatment with 5, 10 or 25 μM 27-OHC led to a substantial increase in Aβ42 levels (Fig. 1b). Treatment with a mixture of 24-OHC + 27-OHC did not induce significant changes in the levels of Aβ42 compared to levels from untreated cells or cells treated with 27-OHC (Fig. 1c). These results suggest that, although it doesn't reduce Aβ42 levels per se, 24-OHC, when added to 27-OHC, prevents the 27-OHC-induced significant increase in Aβ42 levels.

In order to determine potential mechanisms involved in 27-OHC-induced increase in Aβ42 levels, we examined first the effects of 24-OHC and 27-OHC on levels of APP and BACE1. We chose to carry out our experiments with a concentration of 10 μM 24-OHC or 27-OHC. At 10 μM/mL concentration, 24-hydroxycholesterol has been shown to increase APP gene expression in human neuronal cells 13 and 27-hydroxycholesterol has been shown to inhibit neutral sphingomyelinase in human endothelial cells 14. Western blot (Fig. 2a) and densitometric (Fig. 2b) analyses show the effect of 24-OHC, 27-OHC, or a mixture of 24-OHC + 27-OHC on levels of APP and BACE1. There was a significant increase in APP levels following treatment with 27-OHC. Treatment with 24-OHC or with a mixture of 24-OHC + 27-OHC did not statistically increase APP levels. Increase in APP levels with 27-OHC was associated with an increase in BACE1 levels. Treatment with 24-OHC did not significantly alter BACE1 levels. The mixture of 24-OHC + 27-OHC increased BACE1 levels significantly, indicating that 24-OHC does not reverse the effects of 27-OHC on BACE1 levels. These results suggest that 27-OHC-induced increase in levels of APP and BACE1 favors the amyloidogenic pathway that leads to increased Aβ42 production.

The immunofluorescence imaging (Fig. 3) showed a reduced immunoreactivity to 6E10, an antibody that detects full length APP as well as Aβ, in cells treated with 24-OHC in comparison to control cells. A substantial increase in 6E10 staining was observed in cells with 27-OHC compared to control cells or cells treated with 24-OHC. In cells treated with a mixture of 24-OHC+27-OHC, the intensity of the immunoreactivity to 6E10 antibody appears similar to that in control cells. These results are in accordance with the Western blot data showing increased APP and Aβ42 levels with 27-OHC compared to treatment with 24-OHC.

Extracellular sAPPα levels were determined with Western blot analysis in cell medium of control and of cells treated with 24-OHC, 27-OHC, and 24-OHC+27-OHC. Treatment with 24-OHC led to a substantial increase in sAPPα levels (Fig. 4a and 4b). No significant changes in sAPPα levels were observed in cells treated with 27-OHC or with a mixture of 24-OHC+27-OHC when compared to control cells (Fig. 4a and 4b). These results suggest that 24-OHC favors the processing of APP via the non-amyloidogenic pathway that precludes Aβ42 production.

To determine whether increased Aβ42 levels following treatment with 27-OHC is associated with reduced levels of Aβ40, the most abundant Aβ species in normal brain, Aβ40 levels in the medium were measured using ELISA. None of the 24-OHC, 27-OHC and 24-OHC + 27-OHC treatments altered Aβ40 levels (Fig. 4c). These results suggest that the increase in Aβ42 we found with 27-OHC originates from increased processing of APP by BACE1 and not from shift of APP processing from Aβ40 to Aβ42. Collectively, these results demonstrate that 24-OHC favors the non-amyloidogenic pathway while 27-OHC enhances the amyloidogenic pathway.

We also determined the effect of treatment with 24-OHC and 27-OHC on ABCG1 and ABCA1, two cholesterol transporters that may be involved in APP processing and Aβ production. Western blot (Fig. 5a) and densitometric (Fig. 5b) analysis show the effect of 24-OHC, 27-OHC, or a mixture of 24-OHC + 27-OHC on levels of ABCG1 and ABCA1. Treatment with 24-OHC significantly increases levels of ABCG1 and ABCA1. Neither 27-OHC nor a mixture of 24-OHC + 27-OHC significantly altered ABCG1 and ABCA1 levels. These results suggest that regulation of ABCG1 and ABCA1 levels with 24-OHC may be involved in the processing of APP to the non-amyloidogenic pathway.

Purified 24-OHC was obtained from Biomol International (Plymouth Meeting, PA) and 27-OHC from Medical Isotopes, Inc (Pelham, NH). Stock solutions of 24-OHC and 27-OHC were prepared in ethanol and stored at -70°C. All cell culture reagents were obtained from Invitrogen.

Human neuroblastoma SH-SY5Y cells were cultured in 25-cm² cell culture flasks using Dulbecco's modified Eagle's medium: Ham's F12 with Glutamax (DMEM:F12; 1:1; v/v) and 10% FBS. When the cells reached 80% confluence, they were incubated for 24 h at 37°C in DMEM: F12 with vehicle (ethanol), 24-OHC, 27-OHC, or a mixture (1:1) of 24-OHC and 27-OHC. Cells incubated with vehicles were used as a control.

Following treatments, the culture medium was collected, supplemented with protease and phosphatase inhibitors cocktail, and centrifuged at 16,000 × g for 5 min at 4°C. 100 μl of supernatant was used for Aβ40 and Aβ42 quantification by colorimetric sandwich ELISA (Covance, Denver, PA) according to the manufacturer's protocol. Treatments were performed in triplicate, and the quantity of Aβ in each sample was measured in duplicate and expressed as mean ± standard error for the samples. Aβ40 and Aβ42 levels are expressed in pg/ml.

Control, 24-OHC, 27-OHC, and 24-OHC + 27-OHC treated cells were lysed with a protein extraction reagent (M-PER; Thermo Scientific, Rockford, IL). For sAPPα immunoprecipitation procedure prior to Western blotting, media samples were incubated at 4°C overnight with sAPPα antibody (2B3 clone, IBL-America, Minneapolis, MN) using the Catch and Release Reversible Immunoprecipitation System (Millipore, Billerica, MA). Protein concentrations were determined with the BCA protein assay reagent by standard protocol. Proteins were separated in SDS-PAGE gels, transferred to a polyvinylidene difluoride membrane (Millipore, Bedford, MD) and incubated with antibodies to APP (1:1000, Chemicon International, Temecula, CA), sAPPα (1:100), BACE1 (1:100, Chemicon International, Temecula, CA), ABCG1 (1:100, Novus Biologicals, Littleton, CO), and ABCA1 (1:100, Neuromics, Edina, MN). β-actin was used as gel-loading control. The blots were developed with enhanced chemiluminiscence (Immmun-star HRP chemiluminiscent kit, Biorad, Hercules, CA). The results are quantified by densitometry and represented as total integrated densitometric values.

The immunostaining with 6E10 for detection of Aβ in control, 24-OHC, 27-OHC, and 24-OHC +27-OHC treated cells was carried out using confocal microscopy. The cells were fixed with paraformaldehyde, blocked with 5% normal goat serum, and reacted overnight at 4°C with 6E10 antibody (1:250, Signet laboratories Inc., Dedham, MA). Cells were then washed and incubated with secondary antibodies conjugated to Alexa fluor-488 (Molecular Probes, Inc., Eugene, OR) for one hour at room temperature in the dark and washed with PBS. The cells were mounted with Vectasheild containing DAPI (Vector laboratories, Inc., Burlingame, CA). The cells were visualized with a Zeiss LSM 510 META confocal system coupled to a Zeiss Axiophot 200 inverted epifluorescence microscope.

Data was analyzed for statistical significance using analysis of variance (ANOVA) followed by Dunnett's multiple comparison test with GraphPad Prism software 4.01. All values obtained from three different experiments were expressed as mean value ± SEM