|Home | About | Journals | Submit | Contact Us | Français|
Cholesterol homeostasis is regulated not only by cholesterol, but also by oxygenated cholesterol species, referred to as oxysterols. Side-chain oxysterols, such as 25-hydroxycholesterol (25-HC), regulate cholesterol homeostasis through feedback inhibition and feed-forward activation of transcriptional pathways that govern cholesterol synthesis, uptake, and elimination, as well as through direct nongenomic actions that modulate cholesterol accessibility in membranes. Elucidating the cellular distribution of 25-HC is required to understand its biological activity at the molecular level. However, studying oxysterol distribution and behavior within cells has proven difficult due to the lack of fluorescent analogs of 25-HC that retain its chemical and physical properties. To address this, we synthesized a novel intrinsically fluorescent 25-HC mimetic, 25-hydroxycholestatrienol (25-HCTL). We show that 25-HCTL modulates sterol homeostatic responses in a similar manner as 25-HC. 25-HCTL associates with lipoproteins in media and is taken up by cells through LDL-mediated endocytosis. In cultured cells, 25-HCTL redistributes among cellular membranes and, at steady state, has a similar distribution as cholesterol, being enriched in both the endocytic recycling compartment as well as the plasma membrane. Our findings indicate that 25-HCTL is a faithful fluorescent 25-HC mimetic that can be used to investigate the mechanisms through which 25-HC regulates sterol homeostatic pathways.
Mammalian cells obtain cholesterol by de novo synthesis in the endoplasmic reticulum (ER) and by endocytosis of lipoproteins. Cellular cholesterol levels are tightly regulated by coordinated homeostatic mechanisms, primarily involving the regulated proteolysis of the sterol regulatory element-binding protein (SREBP)-2 transcription factor in the Golgi (1). When ER sterols are abundant, SREBP-2 is retained in the ER in a complex with the cholesterol-sensing protein, SREBP cleavage-activating protein (SCAP), and the ER retention proteins, Insig-1 or -2 (2, 3). When cholesterol content in the ER is depleted, SCAP undergoes a conformational change and is released from Insig, allowing the SREBP-SCAP complex to translocate to the Golgi apparatus (4, 5) where it undergoes proteolytic maturation with release of the SREBP transcription factor (6).
Cholesterol homeostasis is regulated not only by cholesterol, but also by oxygenated cholesterol species, referred to as oxysterols (1, 7–9) Side-chain oxysterols, such as 25-hydroxycholesterol (25-HC), are generated enzymatically and act as important regulators of cholesterol homeostasis, despite their presence in cells being only at 0.1% the concentration of cholesterol (9). At the transcriptional level, 25-HC can bind the liver X receptors (LXRs) to active LXR-mediated transcription that results in increased cholesterol efflux and elimination (10, 11). 25-HC also inhibits SREBP maturation and subsequent transcription of genes involved in cholesterol biosynthesis and uptake (1). In contrast to cholesterol, which directly interacts with SCAP, 25-HC enhances the interaction between SCAP and Insig proteins, resulting in retention of the SREBP-2-SCAP complex in the ER (12, 13).
In addition to transcriptional regulation, 25-HC also regulates cholesterol homeostasis by modulating cholesterol accessibility in membranes (14, 15). While the exact mechanism is not known, 25-HC has been shown to have membrane-disordering effects and thins the bilayer, which increases cholesterol accessibility (16, 17) and results in enhanced flux of plasma membrane cholesterol to the ER (18–20). Studies with the enantiomer of 25-HC have implicated direct interactions with membranes for inhibition of SREBP-2 processing as well as degradation of HMG-CoA reductase (HMGR) (14, 15), sterol homeostatic responses that likely are modulated by ER cholesterol accessibility (21). How endogenous oxysterols, which are present at 0.1% the level of cellular cholesterol, participate in regulation of these physiological responses is unclear, though local enrichment in specific compartments offers a possible explanation.
Elucidating the cellular distribution and trafficking of 25-HC is required to understand its biological activity at the molecular level. However, this has proven difficult because of the lack of minimally perturbed fluorescent 25-HC mimetics (22). For this study, we developed a novel fluorescent 25-HC mimetic, 25-hydroxycholestatrienol (25-HCTL), and characterized its fidelity as a 25-HC mimic, as well as its activity and distribution in cells. Our study provides a powerful new tool for investigating the mechanisms through which 25-HC regulates sterol homeostatic pathways.
All tissue culture supplies, Lipofectamine transfection reagents, Trizol reagent, SuperScript III synthesis kits, Alexa labeling kits, and dyes, geneticin, and LipidTox were from Life Technologies. 25-HC was from Steraloids. SYBR Green was from Applied Biosystems. The Dual-Glo Luciferase assay system was purchased from Promega. All other chemicals, including human transferrin (Tf) and HRP, were from Sigma. Iron-loaded Tf was purified on a Sephacryl S-300 gel filtration system (23).125I-Tf was prepared as described previously (23). Alexa488 or -546 was conjugated to iron-loaded Tf following the manufacturer’s instructions. HRP was conjugated to iron-loaded Tf, as described previously (24). [1,2-3H]25-HC (48 Ci/mmol) and [3H]cholesterol were purchased from PerkinElmer Life Sciences.
The scheme and methods for synthesis of 25-HCTL are described in the supplementary data.
Chinese hamster ovary (CHO)-K1 cells (CRL-9618) were obtained from ATCC, and CHO ldlA7 cells were a gift from M. Krieger (25). CHO-K1, CHO ldlA7, and CHO-HMGR (15) cell lines were maintained in normal medium consisting of 1:1 DMEM:Ham’s F-12, 5% (v/v) FBS (Sigma), 2 mM glutamine, 50 units/ml penicillin, and 50 μg/ml streptomycin. For esterification assays, cells were incubated in media containing DMEM/5% lipoprotein-deficient serum (LPDS). Cholesterol starvation medium consisted of LDPS medium plus 20 μM lovastatin and 50 μM mevalonate. For luciferase, flow cytometry, and quantitative (q)PCR, cells were incubated in starvation media, DMEM/5% LPDS with 20 μM lovastatin and 50 μM mevalonate. TRVb1 cells are modified CHO cells that lack the endogenous Tf receptor and express the human Tf receptor (26). TRVb1 cells were grown in bicarbonate-buffered Ham’s F-12 medium supplemented with 200 μg/ml geneticin, 100 units/ml penicillin, 100 μg/ml streptomycin, 5% FBS, and 2 mg/ml glucose. Human osteosarcoma U2OS cells stably expressing scavenger receptor A (SRA) were grown in McCoy’s 5A medium with 1 mg/ml geneticin, 100 units/ml penicillin, 100 μg/ml streptomycin, and 10% FBS (27). U2OS-SRA Niemann-Pick C (NPC)1 knockdown cells were a gift from B. Balch (Scripps Research Institute). U2OS-SRA NPC1 knockdown cells were grown in McCoy’s 5A medium with 5 μg/ml puromycin, 1 mg/ml geneticin, 100 units/ml penicillin, 100 μg/ml streptomycin, and 10% FBS. All cells were grown at 37°C in a 5% CO2 humidified incubator.
Cholesterol esterification was performed as previously described (28). Briefly, cells were pulsed with 1 μCi/ml [3H]cholesterol and 1 μM 25-HC, 1 μM 25-HCTL, or ethanol vehicle in LPDS media at 37°C for 2 and 4 h. Total lipids were extracted and percent cholesterol esterification was calculated by scintillation counting of [3H]cholesteryl ester (CE) and [3H]cholesterol species.
Oxysterol-stimulated HMGR degradation was performed as described (15).
CHO cells were seeded in normal media and then treated with 0–3 μM 25-HC or 25-HCTL in starvation media for 24 h. RNA was extracted using Trizol. cDNAs were detected and amplified using SYBR Green Master Mix (Applied Biosciences) (see primer list in supplementary Table 1).
For quantification of LXR-dependent gene expression, transfection assays with CHO cells were performed as described (15). Briefly, cells were transfected with a reporter containing 990 bp of the human ABCA1 promoter linked to a luciferase reporter and pTK-renilla. Wells were treated for 24 h in starvation media with 0–3 μM 25-HC or 25-HCTL, or 10 μM T0901317, which served as a control for LXR activation. Luciferase activity was measured and normalized to renilla.
Fluorescent images were collected on a DMIRB inverted microscope (Leica Microsystems, Deerfield, IL) equipped with an Andor iXonEM Blue EMCCD camera driven by MetaMorph imaging system software (Universal Imaging/Molecular Devices, Sunnyvale, CA) using 40× (1.25 NA) and 63× (1.36 NA) objectives with 2 × 2 pixel binning (29). 25-HCTL was imaged using a filter cube obtained from Chroma Technology [Bellows Falls, VT; 335 nm (20 nm band pass) excitation filter, 365 nm-long pass dichromatic filter, and 405 nm (40 nm band pass) emission filter]. Standard TRITC, FITC, and Cy5 cubes were obtained from Chroma. Filipin staining was performed as previously described (30), and was imaged using an A4 filter cube (Leica, Wetzlar, Germany). Fluorescence quantification was carried out using MetaMorph image processing software (Molecular Devices), as described previously (30, 31).
CHO cells were seeded at 1 × 105 cells/well in a 6-well dish. At 24 h, cells were treated with 5 μM 25-HC, 5 μM 25-HCTL, or ethanol vehicle in normal media containing [d7]linoleate-BSA (28). Cells were harvested from 0 to 18 h, lipids extracted, and incorporation of oxysterols into oxysterol linoleate esters quantified by LC-MS/MS. For measurement of 25-HCTL incorporation into lipid droplets, U2OS-SRA cells were labeled overnight with 5 μM 25-HCTL in complete medium. To accelerate lipid droplet formation, cells were also incubated with 50 μg/ml acetylated LDL (AcLDL). Following overnight incubation, cells were washed and chased for 4 h to ensure delivery of AcLDL and redistribution of cholesterol from the lysosome. Cells were incubated with Alexa546 Tf for 20 min to identify the endocytic recycling compartment (ERC). Cells were then fixed and lipid droplets were stained using the neutral lipid dye, LipidTox.
Subcellular fractionation was performed as described by Hao et al. (32), with minor alterations. Briefly, TRVb1 cells were grown to confluency in 150 × 25 mm tissue culture dishes. Cells were incubated with M2G medium [150 mM NaCl, 5 mM KCl, 2 mg/ml glucose, 1 mM CaCl2, 1 mM MgCl2, and 20 mM HEPES (pH 7.4)] containing 1 μCi/ml [1,2-3H]25-HC for 60 min and chased for 60 min at 37°C to allow for equilibration. Cells were then incubated with 1.2 μg/ml 125I-Tf and 20 μg/ml HRP-Tf for 20 min at 37°C. After labeling with radioactive probes, cells were chilled on ice. 3′3-Diaminobenzidine (DAB) (0.4 mg/ml), 50 mM ascorbic acid, and 0.025% hydrogen peroxide (H2O2) were added to half of the cells (referred to as “density shifted”), and only DAB and ascorbic acid were added to the other half (“non-shifted”) (33, 34). The dishes were incubated at 4°C on ice in the dark for 60 min. Cells were harvested by incubation with EDTA on ice for 20 min. Cells were washed and resuspended in 250 mM sucrose and 50 mM HEPES (pH 7.2). Cells were lysed by 21 passages through a 27 gauge needle and centrifuged to remove unbroken cells and nuclei. The postnuclear supernatant was applied to a 10–50% continuous sucrose gradient (35). Gradients were subjected to centrifugation at 137,000 g at 4°C for 18 h. The fractionated gradients were taken for radioactivity measurement.
Wild-type U2OS-SRA and U2OS-SRA NPC1 knockdown cells were pulse labeled with 0.5 mg/ml rhodamine dextran 70 kDa and 5 μM 25-HCTL overnight in serum-containing medium. Following incubation, cells were washed and chased in serum medium for 3 h. At the end of the 3 h chase, cells were pulse labeled with 20 μg/ml Alexa488-Tf for 20 min. Cells were imaged using identical parameters using a 63× (1.36 NA) oil objective. 25-HCTL and dextran images were analyzed using the MetaMorph colocalization program, as previously described (36).
All results are expressed as mean ± SE. The statistical significance of differences in mean values was determined by Student’s t-test. Data shown are representative of at least two similar experiments.
To investigate the cellular distribution of side-chain oxysterols, we synthesized an oxysterol analog of 25-HC that contained a C25 -OH group and a conjugated triene system in the steroid B and C rings (Fig. 1A). The synthesis of the probe, 25-HCTL, is detailed in the supplementary data. The steroid ring system in 25-HCTL is identical to that of cholestatrienol (CTL), a cholesterol analog that possesses intrinsic fluorescence due to the presence of the two additional double bonds (37). Top edge views of the two oxysterols reveal the similarity between the 25-HC and 25-HCTL structures (Fig. 1B). Because CTL has previously been shown to mimic the membrane behavior of cholesterol (38, 39), we reasoned that the 25-HCTL probe similarly would be a faithful mimic of 25-HC in model membrane studies.
To test whether 25-HCTL could serve as an authentic oxysterol probe, we compared the biochemical properties of 25-HCTL with 25-HC with respect to a broad range of cellular sterol homeostatic responses. We first examined the effect of the oxysterols on internalization of plasma membrane cholesterol. CHO cells were pulsed with [3H]cholesterol, and arrival of cholesterol at the ER was monitored by esterification by ACAT. We found that the rate and extent of cholesterol esterification resulting from treatment with the two oxysterols were indistinguishable (Fig. 2A). We next examined the effects of the oxysterols on the stability of the HMGR, the degradation of which is mediated by the oxysterol-dependent interaction of HMGR with the ER retention protein, Insig (14, 15). For quantification of HMGR degradation, we treated CHO cells stably expressing an HMGR-green fluorescent protein (GFP) fusion protein with 25-HCTL or 25-HC (15). We found that 25-HCTL and 25-HC were equally effective in accelerating HMGR degradation (Fig. 2B). Next, we examined SREBP-2-dependent expression and found that the oxysterols were also equivalent with respect to suppression of HMGR, HMG-CoA synthase, and the LDL receptor (LDLR) expression (Fig. 3). Finally, we assessed the ability of 25-HCTL to bind to LXR, a nuclear hormone receptor that is activated by side-chain oxysterols and promotes cholesterol elimination and efflux (10). We used an LXR-luciferase reporter assay and also quantified mRNA expression of ABCA1, a canonical LXR target (15). Treatment with the oxysterols resulted in similar dose-response increases in both assays, although the response to 25-HCTL was somewhat blunted in comparison to 25-HC (Fig. 4).
Because 25-HCTL closely mimics the membrane behavior of 25-HC in both biophysical and biochemical assays, we hypothesized that this synthetic fluorescent analog would provide a suitable probe to examine the uptake and intracellular distribution of 25-HC. TRVb1 cells were labeled overnight with various concentrations of 25-HCTL in growth medium (Fig. 5). Cells were then fixed and stained with the cholesterol-binding dye, filipin, in order to compare the distribution of the oxysterol probe relative to cholesterol. Previous work has shown that cholesterol is enriched in the plasma membrane and ERC and can be visualized using filipin (29, 32). At high concentrations, 10 μM and 20 μM, 25-HCTL distributed among the plasma membrane and ERC, but it was also enriched in vesicular structures (Fig. 5A). Cholesterol was not enriched in these vesicular structures, as indicated by the low filipin intensity. We identified these vesicular structures as lysosomes because they colocalized with the acidic compartment dye, LysoTracker, and endocytosed high molecular mass dextrans (Fig. 5B). At lower concentrations, 2.5 and 5 μM, 25-HCTL exhibited a similar distribution as cholesterol and became enriched in the plasma membrane and ERC. Additionally, cholesterol loading of cells using AcLDL resulted in redistribution of 25-HCTL to the ER, where it was esterified and deposited into lipid droplets (Fig. 6A). Redistribution to ER membranes was confirmed biochemically by quantifying esterification of 25-HC and the 25-HCTL analog (Fig. 6B). Thus, 25-HCTL traffics to the ER and is esterified in a similar manner as 25-HC, and demonstrates a similar distribution as cholesterol in living cells.
We examined whether the lysosomal enrichment at high concentrations of the fluorescent analog might have resulted from internalization via endocytosis of the oxysterol probe associated with lipoproteins. To determine whether 25-HCTL is internalized and targeted to lysosomes by initially associating with lipoproteins, we incubated 20 μM 25-HCTL in PBS, 1 mg/ml LDL, or growth medium for 30 min at 37°C, followed by fractionation through a Sephadex G-10 column (Fig. 7). As controls, we fractionated LDL and growth medium that were not incubated with 25-HCTL. We measured the absorbance at 280 nm to detect protein (Fig. 7A) and at 325 nm for 25-HCTL (Fig. 7B) in the individual fractions. 25-HCTL that was incubated in PBS did not have absorbance at 280 nm and was recovered in fractions 9 and 10, as detected by absorbance at 325 nm. This is consistent with elution of a small molecule. Strikingly, 25-HCTL that was incubated with growth medium cofractionated with serum proteins and lipoproteins (fractions 4–6), while growth medium alone had low absorbance at 325 nm. Not surprisingly, 25-HCTL that was incubated with 1 mg/ml LDL also cofractionated with LDL (fractions 4–6), while LDL alone did not have absorbance at 325 nm. To investigate whether 25-HCTL primarily enters cells via association with lipoproteins, we incubated wild-type and LDLR-deficient (ldlA7) CHO cells, which lack a functional LDLR, with 25-HCTL. Levels of 25-HCTL were lower in LDLR-deficient cells, both at steady-state following an overnight incubation (Fig. 7C) and after a 4 h pulse under lipoprotein starvation conditions (Fig. 7D). Taken together, these data show that 25-HCTL associates with lipoproteins, including LDL, and is internalized via the LDLR, suggesting that oxysterols may be internalized with lipoproteins and traffic first to endosomes and lysosomes before redistribution to the plasma membrane and ERC.
To investigate the distribution of 25-HCTL following internalization, we performed pulse chase experiments in live cells using the fluorescent analog (Fig. 8). In these experiments lower 25-HCTL concentrations (5 μM) were used to more closely approximate the trafficking of oxysterols under physiological conditions. To identify the ERC, TRVB1 cells were pulse labeled with 20 μg/ml fluorescent Tf for 15 min prior to fixation. At t = 0, 25-HCTL was enriched in punctate vesicular structures and colocalized poorly with Tf. After 30 min, the 25-HCTL fluorescence in the vesicular structures was reduced, and there was partial colocalization of 25-HCTL with Tf. After 60 min, the majority of the 25-HCTL fluorescence had trafficked from the vesicular structures to the ERC, colocalizing with Tf, as well as the plasma membrane. In addition to enrichment in the ERC and plasma membrane, 25-HCTL fluorescence was observed in a perinuclear compartment that could be the Golgi. These data show that following association with lipoproteins and endocytosis, the oxysterol analog is trafficked from the lysosome and then distributed throughout the cell, becoming enriched in the plasma membrane and ERC, as well as to some extent in other perinuclear organelles.
The fluorescence microscopy data shows that at low concentrations 25-HCTL, like cholesterol, is enriched in the plasma membrane and ERC. To examine whether the results obtained using 25-HCTL reflect the cellular distribution of natural 25-HC, we determined the distribution of 25-HCTL and 25-[3H]hydroxycholesterol following subcellular fractionation (Fig. 9). Cells were labeled with 25-[3H]hydroxycholesterol for 60 min, chased for 60 min to allow for equilibration through cellular compartments, and incubated for 20 min with 125I-Tf prior to disruption and separation of postnuclear supernatants on sucrose density gradients. Both 125I-Tf and 25-[3H]hydroxycholesterol distribute in two peaks (Fig. 9A). To identify the plasma membrane in the gradient, cells were labeled with 125I-Tf on ice, which retains Tf exclusively at the plasma membrane (Fig. 9B). Under these conditions, the peak at the top of the gradient (fractions 1–4) is maintained, while the heavier peak (fractions 10–15) is less apparent. This suggests that the fractions 1–4 contain plasma membrane, whereas fractions 10–15 contain various cellular organelles, including ERC, trans-Golgi, and Golgi. To isolate the ERC from other cellular organelles of similar density, we took advantage of a density shifting reaction using HRP. Cells were incubated with 25-[3H]hydroxycholesterol followed by incubation with 125I-Tf and HRP-Tf for 20 min. Under these conditions, 125I-Tf and HRP-Tf become enriched in the ERC. The cells are then reacted with ascorbic acid, DAB, and H2O2 to generate a dense product in the ERC. The accumulation of the DAB polymer in the ERC results in a significant increase in the compartment’s density, whereas organelles that do not have HRP-Tf are unaffected. Comparing the nonshifted (Fig. 9A) and shifted (Fig. 9C) profiles, we see that the denser peak in the nonshifted profile moves to the bottom of the sucrose gradient (50%), carrying both 125I-Tf and 25-[3H]hydroxycholesterol as a result of the density shifting reaction. Taken together, the shifted and nonshifted profiles show that a significant portion of 25-[3H]hydroxycholesterol accumulates in the ERC and plasma membrane.
As 25-HCTL entered cells via absorption on lipoproteins and trafficked to the lysosome before redistribution throughout the cell, we investigated the role of the NPC1 protein in efflux of 25-HCTL from lysosomes. We compared the 25-HCTL localization in wild-type U2OS and NPC1 knockdown cells (U2OS-NPC1) (Fig. 10A). At steady-state in wild-type U2OS cells, 25-HCTL redistributes throughout the plasma membrane and ERC with some accumulation in vesicular structures. However, in U2OS-NPC1 cells, 25-HCTL colocalizes with the rhodamine dextran (Fig. 10B), indicating that it is largely sequestered in the vesicular structures with some distribution throughout the cell. Thus, NPC1 is required for efficient trafficking of 25-HCTL from lysosomes, similar to what was previously reported for an alkyne oxysterol analog (40).
Although side-chain oxysterols are important physiological regulators of cholesterol homeostasis, their cellular distribution remains poorly understood (20). In this study, we synthesized and characterized a novel intrinsically fluorescent 25-HC analog, 25-HCTL, to investigate the cellular activity and distribution of 25-HC. We show that 25-HCTL exhibits similar biochemical behavior as 25-HC with respect to regulation of cholesterol homeostatic responses, indicating that it is a faithful mimetic of 25-HC. Through its association with lipoproteins, 25-HCTL gains entry into the cell via the endocytic pathway, initially trafficking to lysosomes and eventually redistributing to the ERC and plasma membrane, similar to cholesterol. Thus, 25-HCTL provides a minimally perturbed fluorescent oxysterol probe for investigation of the itinerary and distribution of 25-HC in cells, as well as for understanding the mechanisms through which oxysterols exert their homeostatic effects.
Fluorescence microscopy is a powerful tool for studying intracellular distribution and transport processes. Fluorescent lipid molecules have been widely used to study behavior of lipids in model membranes and cells. The utility of these lipid analogs, however, is often limited because addition of a bulky fluorophore can alter the physical properties of the molecule (41–43). For cholesterol trafficking, two approaches have been successfully used. One approach is the introduction of an alkyne moiety to lipids that can be conjugated to a fluorophore using bio-orthognal chemistry (40, 44, 45). Although the alkyne moiety does not greatly alter the biophysical properties of the sterols, it precludes live cell imaging due to the conjugation requirements (40, 44). An alternative approach takes advantage of intrinsically fluorescent sterols, including the naturally occurring cholesterol analog, dehydroergosterol (DHE) (32), or a synthetic sterol, CTL (46). These fluorescent sterols are similar to cholesterol in biophysical behavior and structure (18, 39), but include two additional double bonds in the steroid ring system that are responsible for their fluorescent properties.
For this study, we designed and synthesized a fluorescent mimetic of 25-HC. Like CTL (37), 25-HCTL possesses a conjugated steroid ring, making it fluorescent without greatly altering the structure of 25-HC (Fig. 1). As a structural mimetic of 25-HC, it was anticipated that 25-HCTL would have comparable biochemical activity in regulating cholesterol homeostasis. Similar to 25-HC, 25-HCTL treatment increased cholesterol flux to the ER for esterification (15), suppressed SREBP-dependent gene expression, and decreased HMGR protein levels (47). These findings indicate that 25-HCTL, like 25-HC, partitions into membranes, resulting in increased accessibility of cholesterol and enhanced sterol flux to the ER (14, 17, 19). The reduction of HMGR protein and SREBP target gene expression suggests that 25-HCTL has similar effects as 25-HC on modulating the activity of the Insig proteins (13). Likewise, 25-HCTL is able to modulate LXR gene expression, albeit to a lesser extent than 25-HC (10, 48). The difference in activation of LXR may be attributable to the addition of the two double bonds in 25-HCTL that results in flattening of the chair conformations of the steroid B and C rings. Taken together, 25-HCTL is an excellent structural and biochemical mimetic of 25-HC.
We exploited the fluorescent properties of 25-HCTL to examine the itinerary and distribution of 25-HC in cells. At high concentrations (10–20 μM), the majority of 25-HCTL was retained in lysosomes, while a portion redistributed throughout cell membranes. The enrichment of 25-HCTL in lysosomes did not alter the cholesterol distribution of the cells, suggesting that the retention was not due to attenuated activity of the NPC proteins and/or reduced sterol efflux. On the other hand, redistribution of 25-HCTL from lysosomes does require the action of NPC1 protein, as the probe was retained in lysosomes of NPC1-deficient cells. A potential mechanism to explain the initial enrichment of 25-HCTL in lysosomes is that the probe associates with lipoproteins in media and is subsequently internalized via the LDLR. This is supported by experiments demonstrating that 25-HCTL partitions into serum lipoproteins or purified LDL and enters the cell via LDLR, rather than through passive diffusion. This mechanism may be shared with other oxysterol probes that also initially transit through the lysosome and then redistribute to other organelles (40, 44).
At lower concentrations (2.5–5 μM), 25-HCTL has a similar membrane distribution as cholesterol. This distribution is similar to that of the fluorescent cholesterol probes, DHE (32) and CTL (46), yet markedly different from the previously reported 20(S)-hydroxycholesterol alkyne probe (40). This finding suggests that 25-HCTL, like other sterol probes (40, 44), initially transits through the lysosome and then redistributes among membranes (e.g., ERC and plasma membrane) to achieve a steady-state distribution similar to that of cholesterol. 25-HCTL is also esterified and incorporated into lipid droplets, indicating that, like 25-HC, the analog traffics to the ER. Radiolabeled 25-HC similarly localized to the ERC and plasma membrane at steady-state, strongly suggesting that the fluorescent analog can serve as a faithful probe to study 25-HC trafficking. The rapid redistribution of 25-HC/25-HCTL among cell membranes is likely mediated by both vesicular and nonvesicular transport mechanisms (18, 20). As cholesterol distribution is maintained largely by a nonvesicular mechanism (32), it seems plausible that 25-HC and 25-HCTL trafficking between cellular membranes might be mediated, at least in part, by sterol transport proteins such as the OSBP/ORP protein family members or STARD4 (49–51). Future studies will be needed to investigate the role of the sterol transport proteins in oxysterol trafficking between organelles.
In summary, we present the design, synthesis, and characterization of a novel fluorescent analog of 25-HC. We provide evidence that the addition of two double bonds in the steroid ring system does not greatly alter its structure or biochemical activity, while making it a suitable probe for fluorescence studies. Through use of this fluorescent analog, we can study, for the first time, the trafficking kinetics of 25-HC in living cells. Additionally, this probe reveals the time-resolved heterogeneous distribution of side-chain oxysterols in cells, and provides insight into how the local environment of a membrane organelle may potentially be altered by submicromolar oxysterol levels. Taken together, these findings validate the use of 25-HCTL as a fluorescent mimetic of 25-HC for study of biological activity of oxysterols.
This work was supported by grants to D.B.I. (F31 DK104631), D.S.O. (HL067773), F.R.M. (R37 DK27083), and A.A.B. (F30 HL97563), and by the Washington University Diabetes Research Center (P30 DK020579) and the Taylor Family Institute for Innovative Psychiatric Research (D.F.C.).
[S]The online version of this article (available at http://www.jlr.org) contains a supplement.