Cellular cholesterol flux studies: methodological considerations.

Reverse cholesterol transport (RCT) is the process in which peripheral cells release cholesterol to an extracellular acceptor such as high-density lipoprotein (HDL) which then mediates cholesterol delivery to the liver for excretion. RCT represents a physiological mechanism by which peripheral tissues are protected against excessive accumulation of cholesterol. The first step in RCT is the interaction of the cell with lipoprotein particles, a process that results in both the cellular uptake and release of cholesterol. The various components of this cholesterol flux can be viewed as efflux, influx and net flux. Experimental protocols for measuring each of these components of cholesterol flux are very different, and a number of considerations are required to design experimental approaches for the quantitation of flux parameters. Although many flux studies have been conducted in the past, the recent discoveries of the scavenger receptor B1 (SR-B1) and ATP binding cassette 1 (ABCA1), which mediate the movement of cholesterol between cells and extracellular acceptors, has led to increased interest in studies of cellular cholesterol flux. The aim of this review is to present a discussion of the methodological considerations that should be evaluated during the design and analysis of cellular cholesterol flux experiments.

Evidence that newly synthesized esterified cholesterol is deposited in existing cytoplasmic lipid inclusions.

Esterified cholesterol (EC) and triglyceride (TG) can be stored in cells as cytoplasmic inclusions. The physical state of the EC in these lipid droplets varies from liquid to liquid crystalline, depending on a number of factors, including the amount of TG co-deposited in the inclusion. The lipid in these droplets undergoes turnover via hydrolysis and resynthesis. We determined whether newly synthesized lipid is incorporated into existing cytoplasmic droplets, forms a discrete cytoplasmic droplet, or forms a small inclusion that fuses with an existing droplet. This was accomplished by monitoring the physical state of the lipid within the cytoplasmic inclusions following sequential deposition of TG and EC. Fu5AH cells were initially grown in media containing oleic acid to produce TG-rich, isotropic inclusions. The cells were then incubated with medium containing free cholesterol-phospholipid dispersions to promote synthesis and deposition of EC. To inhibit cytoplasmic TG hydrolysis, the lipase inhibitor, diethylumbelliferyl phosphate (UBP), was added at the time of cholesterol enrichment. The phase behavior of lipid droplets isolated from the lipid-rich cells was determined using polarizing light flow cytometry and microscopy. An anisotropic droplet population (EC-rich inclusions) was not detected, although there was an increase in cellular EC mass and no change in cellular TG mass. Therefore, under conditions where there is no turnover of cytoplasmic TG, newly synthesized EC is incorporated into existing TG inclusions.

Expression of scavenger receptor BI in COS-7 cells alters cholesterol content and distribution.

Previous studies have shown that scavenger receptor BI (SR-BI) stimulates the bidirectional flux of free cholesterol (FC) between HDL and SR-BI-expressing cells. A major component of the enhanced FC flux appears to occur independently of HDL binding to SR-BI and may be due to changes in membrane lipid domains resulting from SR-BI expression (1). In the present study, the impact of SR-BI on cellular cholesterol metabolism was determined by examining SR-BI-mediated changes in cellular cholesterol mass, the esterification of HDL-derived FC, and changes in membrane lipid pools. Growth of SR-BI-expressing cells in medium containing HDL led to increased cellular cholesterol mass, most of which accumulated as ester. The esterification of HDL-derived FC was enhanced by SR-BI-expression to a far greater extent than the SR-BI mediated increase in FC uptake, suggesting an SR-BI-mediated effect on cholesterol utilization in the cell. This observation was tested by comparing FC esterification rates in SR-BI positive and negative cells when equivalent amounts of extracellular FC were taken up via cyclodextrins or apolipoprotein AI/phospholipid disks, neither of which contained cholesteryl ester. Under these conditions, SR-BI did not preferentially stimulate cholesterol esterification. These results indicate that the enhanced esterification of HDL-derived FC in SR-BI-expressing cells is due to the expanded pool of cellular FC and not to a specific effect of SR-BI on cholesterol utilization. Two approaches were used to test the effects of SR-BI expression on membrane lipid organization. In the first, the sensitivity of cellular FC to exogenous cholesterol oxidase was tested under conditions in which there is a preferential oxidation of caveolar cholesterol. SR-BI-expression was found to greatly increase the fraction of cellular cholesterol available to the oxidase as compared to either vector-transfected cells or cells expressing the related class B scavenger receptor CD36. These results suggest that SR-BI expression alters the distribution of membrane-free cholesterol to a caveolar fraction or alters the accessibility of this membrane fraction to exogenous cholesterol oxidase. In the second approach, the efflux of cellular FC to high concentrations of cyclodextrins was monitored under conditions where desorption of FC from the plasma membrane is rate limiting for efflux. SR-BI-expressing cells showed a shift in the distribution of FC between two kinetic pools with more FC in the fast pool and less in the slow pool. These data support a model in which SR-BI expression leads to a redistribution of cholesterol to membrane domains that serve to facilitate the flux of FC between cells and lipoproteins.

Cytotoxic cholesterol is generated by the hydrolysis of cytoplasmic cholesteryl ester and transported to the plasma membrane.

The present study examines the fate and effects of free cholesterol (FC) generated by the hydrolysis of cytoplasmic cholesteryl esters (CE) in model macrophage foam cells. J774 or elicited mouse peritoneal macrophages (MPM) were enriched with CE by incubating with acetylated low density lipoprotein (acLDL) and FC/phospholipid dispersions, thus creating model foam cells. Treatment of the foam cells with the acyl coenzyme-A:cholesterol acyltransferase (ACAT) inhibitor, CP-113,818, in the absence of any extracellular cholesterol acceptors, resulted in cellular toxicity. This was accompanied by an increase in the amount of FC available for oxidation by an exogenous cholesterol oxidase. Furthermore, cellular toxicity was proportional to the size of the oxidase susceptible pool of FC over time. Morphological analysis and in situ DNA fragmentation assay demonstrated the occurrence of apoptosis in the ACAT inhibited cells. Co-treatment with the hydrophobic amine U18666A, an intracellular cholesterol transport inhibitor, led to a dose dependent reduction in cytotoxicity and apoptosis, and blocked the movement of FC into the oxidase susceptible pool. In addition, treating model foam cells with CP-113,818 plus chloroquine, a compound that inhibits the function of acidic vesicles, also diminished cellular toxicity. Staining with the cholesterol binding dye filipin revealed that the macrophages treated with CP-113,818 contained a cholesterol oxidase accessible pool of FC in the plasma membrane. These results suggest that FC generated by the hydrolysis of cytoplasmic CE is transported through acidic vesicles to the plasma membrane, and accumulation of FC in this pool triggers cell death by necrosis and apoptosis.

Crystallization of free cholesterol in model macrophage foam cells.

-The present study examined free cholesterol (FC) crystallization in macrophage foam cells. Model foam cells (J774 or mouse peritoneal macrophages [MPMs]) were incubated with acetylated low density lipoprotein and FC/phospholipid dispersions for 48 hours, resulting in the deposition of large stores of cytoplasmic cholesteryl esters (CEs). The model foam cells were then incubated for up to 5 days with an acyl-coenzyme A:cholesterol acyltransferase (ACAT) inhibitor (CP-113,818) in the absence of an extracellular FC acceptor to allow intracellular accumulation of FC. FC crystals of various shapes and sizes formed in the MPMs but not in the J774 macrophages. Examination of the MPM monolayers by microscopy indicated that the crystals were externalized rapidly after formation and thereafter continued to increase in size. Incubating J774 macrophages with 8-(4-chlorophenylthio)adenosine 3':5'-cyclic monophosphate (CPT-cAMP) in addition to CP-113,818 caused FC crystal formation as a consequence of CPT-cAMP stimulation of CE hydrolysis and inhibition of cell growth. In addition, 2 separate cholesterol phases (liquid-crystalline and cholesterol monohydrate) in the plane of the membrane bilayer were detected after 31 hours of ACAT inhibition by the use of small-angle x-ray diffraction of J774 macrophage foam cells treated with CPT-cAMP. Other compounds reported to inhibit ACAT, namely progesterone (20 microgram/mL) and N-acetyl-D-sphingosine (c(2)-ceramide, 10 microgram/mL), induced cellular toxicity in J774 macrophage foam cells and FC crystallization when coincubated with CPT-cAMP. Addition of the extracellular FC acceptors apolipoproteins (apo) E and A-I (50 microgram/mL) reduced FC crystal formation. In MPMs, lower cell density and frequent changes of medium were conducive to crystal formation. This may be due to "dilution" of apoE secreted by the MPMs and is consistent with our observation that the addition of exogenous apoE or apoA-I inhibits FC crystal formation in J774 macrophage foam cells cotreated with CP-113,818 plus CPT-cAMP. These data demonstrate that FC crystals can form from the hydrolysis of cytoplasmic stores of CEs in model foam cells. FC crystal formation can be modulated by the addition of extracellular FC acceptors or by affecting the cellular rate of CE hydrolysis. This process may contribute to the formation of FC crystals in atherosclerotic plaques.

Cell cholesterol efflux: integration of old and new observations provides new insights.

Numerous studies using a variety of cell/acceptor combinations have demonstrated differences in cholesterol efflux among cells. These studies also show that different acceptors, ranging from simple molecules like cyclodextrins to serum, stimulate efflux through a variety of mechanisms. By combining early observations with data derived from recent studies, it is now possible to formulate a model for cell cholesterol efflux which proposes that an array of different mechanisms, including aqueous diffusion, lipid-free apolipoprotein membrane microsolubilization, and SR-BI-mediated cholesterol exchange contribute to cholesterol flux. In this model the relative importance of each mechanism would be determined both by the cell type and the nature of the extracellular cholesterol acceptor.

Comparative disposition of the nephrotoxicant N-(3, 5-dichlorophenyl)succinimide and the non-nephrotoxicant N-(3, 5-difluorophenyl)succinimide in Fischer 344 rats.

Disposition of the nephrotoxicant N-(3,5-dichlorophenyl)succinimide (NDPS) was compared with that of a nontoxic analog, N-(3, 5-difluorophenyl)succinimide (DFPS). Male Fischer 344 rats were administered 0.2 or 0.6 mmol/kg [14C]NDPS or [14C]DFPS (i.p. in corn oil). Plasma concentrations were determined from blood samples obtained through the carotid artery. Urine samples were analyzed for metabolite content by HPLC. Rats were sacrificed at 3 h (DFPS) or 6 h (NDPS) and tissue radiolabel content and covalent binding were determined. [14C]NDPS-derived plasma radioactivity levels were 6- to 21-fold higher and peaked later than those from [14C]DFPS. Six hours after dosing, NDPS was 40% eliminated in the urine compared with approximately 90% for DFPS. By 48 h, only 67% of the NDPS dose was eliminated in urine. In contrast, DFPS excretion was virtually complete within 24 h. NDPS underwent oxidative metabolism to a slightly greater extent than DFPS. Distribution of [14C]NDPS-derived radioactivity into the kidneys was 3- to 6-fold higher than that into the liver or heart, and was more extensive than with [14C]DFPS. NDPS also covalently bound to plasma, renal, and hepatic proteins to a greater extent than DFPS. In summary, NDPS achieves higher tissue and plasma concentrations, covalently binds to a greater extent, and is eliminated more slowly than DFPS. Differences in the lipid solubility of NDPS metabolites and DFPS metabolites may help explain these results. The overall greater tissue exposure of NDPS and its metabolites may contribute to differential toxicity of these analogs.

Scavenger receptor BI (SR-BI) mediates free cholesterol flux independently of HDL tethering to the cell surface.

In addition to its effect on high density lipoprotein (HDL) cholesteryl ester (CE) uptake, scavenger receptor BI (SR-BI) was recently reported to stimulate free cholesterol (FC) flux from Chinese hamster ovary (CHO) cells stably expressing mouse SR-BI, a novel function of SR-BI that may play a role in cholesterol removal from the vessel wall where the receptor can be found. It is possible that SR-BI stimulates flux simply by tethering acceptor HDL particles in close apposition to the cell surface thereby facilitating the movement of cholesterol between the plasma membrane and HDL. To test this, we used transiently transfected cells and compared the closely related class B scavenger receptors mouse SR-BI and rat CD36 for their ability to stimulate cholesterol efflux as both receptors bind HDL with high affinity. The results showed that, although acceptor binding to SR-BI may contribute to efflux to a modest extent, the major stimulation of FC efflux occurs independently of acceptor binding to cell surface receptors. Instead our data indicate that SR-BI mediates alterations to membrane FC domains which provoke enhanced bidirectional FC flux between cells and extracellular acceptors.

Effects of intracellular free cholesterol accumulation on macrophage viability: a model for foam cell death.

This study was designed to identify cellular responses associated with free cholesterol (FC) accumulation in model macrophage foam cells. Mouse peritoneal macrophages (MPMs) or J774 macrophages were loaded with cholesteryl esters using acetylated LDL and FC/phospholipid dispersions and were subsequently exposed to an acyl coenzyme A:cholesterol acyltransferase (ACAT) inhibitor. This treatment produced a rapid accumulation of cellular FC. The FC that accumulated due to ACAT inhibition was more readily available for efflux to 2-hydroxypropyl-beta-cyclodextrin (which removes cholesterol from the plasma membrane) than FC in untreated control cells. After a 3-hour exposure to an ACAT inhibitor, a significant increase in phospholipid synthesis was seen, followed by the leakage of LDH after 12 hours of treatment. We also observed, by electron and fluorescence microscopy, morphological indications of both apoptosis and necrosis in cells treated with an ACAT inhibitor. In addition, inhibition of ACAT for 48 hours resulted in the formation of FC crystals in MPMs but not in J774 cells. If compound 3beta-[2-(diethylamino)ethoxy]androst-5-en-17-one (U18666A), which modulates intracellular trafficking of cholesterol, was added together with the ACAT inhibitor, each of the metabolic changes elicited by the accumulation of excess FC was either diminished or eliminated. The protective affect of U18666A was not due to a decrease in cellular FC concentrations, because cells treated with an ACAT inhibitor accumulated similar amounts of FC in the presence or absence of U18666A. Thus, treatment with U18666A results in the sequestering of FC in a pool that prevents it from causing various responses to FC deposition in macrophages. The metabolic changes that were produced when these model foam cells were treated with the ACAT inhibitor parallel the pathological events that have been shown to occur in the developing atherosclerotic plaque.

Cytochrome P450-mediated metabolism and nephrotoxicity of N-(3,5-dichlorophenyl)succinimide in Fischer 344 rats.

The agricultural fungicide N-(3,5-dichlorophenyl)succinimide (NDPS) is nephrotoxic in rats. Previous studies have suggested that oxidative hepatic biotransformation is required for the induction of kidney damage. The experiments described in this paper were designed to further investigate the relationship between NDPS metabolism and nephrotoxicity using various modulators of cytochrome P450 activity. Male Fischer 344 rats were pretreated with the P450 inducers Aroclor 1254 (ARO), isoniazid (INH), 3-methylcholanthrene (3-MC), and phenobarbital (PB), or the P450 inhibitor 1-aminobenzotriazole (ABT). Control animals received vehicle only. NDPS metabolism was investigated using hepatocytes isolated from the various treatment groups. Separate experiments were also conducted to evaluate the effects of these pretreatments on NDPS-induced nephrotoxicity in rats. PB and ARO enhanced formation of the known nephrotoxic NDPS metabolites, N-(3,5-dichlorophenyl)-2-hydroxysuccinimide, N-(3,5-dichlorophenyl)-2-hydroxysuccinamic acid, and N-(3,5-dichlorophenyl)-3-hydroxysuccinamic acid, by the hepatocytes. In contrast, ABT inhibited formation of the nephrotoxic metabolites, whereas INH and 3-MC did not alter NDPS biotransformation. NDPS-induced renal damage was potentiated by pretreating the rats with PB or ARO and was attenuated by ABT. Compared with control animals, toxicity was unaffected by INH or 3-MC pretreatments. Thus, there was a correlation between pretreatments that induce P450-mediated NDPS metabolism and the effects that these compounds have on NDPS-induced nephrotoxicity. The data indicate that specific P450 isozymes metabolize NDPS to its hydroxylated products and suggest that these metabolites mediate the nephrotoxicity induced by NDPS.

The effect of aromatic fluorine substitution on the nephrotoxicity and metabolism of N-(3,5-dichlorophenyl)succinimide in Fischer 344 rats.

N-(3,5-Difluorophenyl)succinimide (DFPS) is a non-toxic analogue of the nephrotoxic fungicide N-(3,5-dichlorophenyl)succinimide (NDPS). Although NDPS must be metabolized to produce renal damage, the metabolic fate of DFPS is unknown. These studies were therefore designed to examine the nephrotoxic potential of putative DFPS metabolites and to determine if DFPS is metabolized differently from NDPS. Male Fischer-344 rats were administered (1.0 mmol/kg. i.p. in corn oil) DFPS, N-(3,5-difluorophenyl)succinamic acid (DFPSA), N-(3,5-difluorophenyl)-2-hydroxysuccinimide (DFHS), N-(3,5-difluorophenyl)-2- or -3-hydroxysuccinamic acids (2- and 3-DFHSA, respectively), N-(3,5-difluoro-4-hydroxyphenyl)succinimide (DFHPS). N-(3,5-difluoro-4-hydroxyphenyl) succinamic acid (DFHPSA) or corn oil only (1.2 ml/kg). Although some of the compounds produced changes in renal function and histology, these alterations were not indicative of irreversible kidney damage. DFPSA, 2-DFHSA, 3-DFHSA and DFHPSA were detected in the urine of rats 3 h after administration of 0.2 mmol/kg [14C]DFPS. The same metabolites were produced by isolated rat hepatocytes, but not by renal proximal tubule cells. Formation of the oxidative metabolites in vitro was prevented by the cytochrome P450 inhibitor 1-aminobenzotriazole. It appears that DFPS undergoes hepatic biotransformation similar to NDPS and that some of its metabolites have reversible effects on renal proximal tubules.

Nephrotoxic potential of N-(3,5-dichlorophenyl)glutarimide and N-(3,5-dichlorophenyl)glutaramic acid in Fischer 344 rats.

The agricultural fungicide N-(3,5-dichlorophenyl)succinimide (NDPS) produces kidney damage in rats. Although many NDPS analogues have been screened as possible nephrotoxicants, the one-carbon homologue, N-(3,5-dichlorophenyl)glutarimide (NDPG), has not been evaluated. This study examined the nephrotoxic potential of NDPG and a putative metabolite, N-(3,5-dichlorophenyl)glutaramic acid (NDPGA). Male Fischer 344 rats (N = 3-4 per group) were administered a single i.p. injection in corn oil of NDPG or NDPGA (0.4 or 1.0 mmol/kg), NDPS (0.4 mmol/kg), or corn oil alone. Renal function was monitored for 48 h. In contrast to NDPS, NDPG and NDPGA did not significantly alter renal function or kidney morphology when compared to corn oil-treated controls. These experiments show that replacement of the succinimide ring in NDPS with a glutarimide ring abolishes toxicity.