| COMMENTARIES |
A New Concept for Cellular Cholesterol Homeostasis
Yvonne Lange1 and Theodore L. Steck2, 1Department of Pathology, Rush University Medical Center, Chicago, IL 60612, 2Department of Biochemistry & Molecular Biology, University of Chicago, 920 E. 58th Street, Chicago, IL 60637, Tel: 312-942-5256, Email: ylange@rush.edu
Most of the body burden of cholesterol is of endogenous origin. Consequently, the way cells manage their cholesterol impacts atherogenesis. Cells control their cholesterol levels quite closely; we have determined, for example, that the level of cholesterol in fibroblasts remains constant to within ± 10% over weeks of cultivation. But how does a cell know how much cholesterol it has and how much cholesterol it needs; and how does it adjust the difference? The last of these questions has been thoroughly explicated: multiple feedback pathways respond in parallel to the level of cellular cholesterol or the oxysterols derived therefrom [1, 2]. We now present a new hypothesis concerning the first two questions: how cells set their cholesterol requirement and sense deviations from it [3,4].
We begin with the premise that cholesterol enters into stoichiometric complexes with membrane bilayer phospholipids [5]. The strongest among these weak and dynamic sterol associations is with sphingolipids and phosphatidylcholine [6,7]. These complexes are stabilized by the hydrogen bonding of polar head groups and van der Waals contacts among nonpolar surfaces; furthermore, the hydrophobic exclusion of sterols from water can drive them to settle under the protective umbrella of the large polar head groups of their lipid partners [5,8]. The complexes so formed can segregate laterally into condensed and ordered liquid phases that serve as the basis for rafts [9,10]. It may be that most of the phospholipids in the plasma membrane complex with cholesterol, given that the mole ratio of cholesterol to phospholipids in these membranes is almost unity [11,12] and that each cholesterol molecule typically associates with at least one phospholipid molecule [5]. This picture is consistent with the fact that cholesterol increases the impermeability and mechanical strength of the entire bilayer and not just patches of it. It follows that cholesterol complexes may comprise a large fraction of the plasma membrane.
What concerns us here is the novel state of the cholesterol that accumulates in excess of the complexing capacity of the bilayer phospholipids. These surplus cholesterol molecules, being dispersed in the bilayer in a relatively unrestrained state, have a high escape tendency, chemical activity or fugacity [5,13]. We shall refer to this attribute as their activity [4]. We imagine that active cholesterol has an increased frequency and/or extent of projection out of the bilayer into the surrounding water space [4]. One reflection of this active state is that plasma membrane cholesterol is normally a very poor substrate for the enzyme cholesterol oxidase but the addition of small amounts of cholesterol to cells greatly stimulates cholesterol oxidation. This behavior suggests that the active excess but not the complexed sterol is the substrate for the enzyme. Similarly, incrementing the cholesterol in synthetic lipid membranes or in plasma membranes beyond the normal mole ratio with phospholipids promotes cholesterol transfer to the extracellular acceptor, cyclodextrin [3,14]. We argue below that the increased tendency of uncomplexed sterol molecules to leave the membrane (i.e. the abundance of the active fraction) is a central factor in the homeostatic regulation of cell cholesterol. How?
We assume that the physiological level of plasma membrane cholesterol is determined by the saturation point of the participating phospholipids—the stoichiometric balance point at which neither is in excess. Then, any rise in plasma membrane cholesterol would be countered thermodynamically by the high activity (leaving tendency) of the uncomplexed sterol. Conversely, depletion of the plasma membrane would be reversed by equilibration with cholesterol pools in the plasma and cytoplasm. The lipid transfer protein, scavenger receptor class B type I (SR-BI) facilitates such passive exchange with the circulating high-density lipoproteins, HDL [15]. In addition, the ATP-binding cassette transport protein, ABCA1, actively exports cholesterol to nascent apo-HDL acceptors [16]. We suggest that SR-BI and ABCA1 preferentially draw upon the active fraction of plasma membrane cholesterol as their substrate, thereby titrating plasma membrane cholesterol to its physiological set-point. This makes sense a priori: limiting export to the active excess would prevent the potentially injurious depletion of plasma membrane cholesterol.
Circumstantial evidence supports this hypothesis. For one thing, SR-BI promotes the susceptibility of plasma membrane cholesterol to cholesterol oxidase attack [17] as if activating it; that is, facilitating its partial projection into the aqueous compartment. In addition, the biphasic kinetics with which plasma membrane cholesterol is transferred to HDL by SR-BI suggests that there are two plasma membrane compartments [18]; the small, fast-moving pool could be comprised of active cholesterol, the preferred substrate for SR-BI, while the large, slow pool would be the cholesterol held in complexes that gradually replenishes the fast pool as it is depleted.
Just as the case of SR-BI, ABCA1 promotes the attack of cholesterol oxidase on the plasma membrane [19]. That ABCA1 apparently resides in cholesterol-poor regions of the plasma membrane rather than in cholesterol-rich rafts is also consistent with the suggestion that its substrate is uncomplexed, high activity cholesterol [20]. Furthermore, our working hypothesis explains why digesting the cell surface with sphingomyelinase C, which liberates cholesterol from sphingomyelin complexes [21-24], stimulates cholesterol export by ABCA1 [25]. Finally, synthetic ceramides that displace cholesterol from phospholipid complexes [26] also activate ABCA1-mediated transport [27,28]. A strong inference is that active cholesterol is the true substrate for the lipid transfer proteins. These processes, of themselves, would serve to restore perturbed plasma membrane cholesterol levels to the physiological set-point of the resting cell.
Several mechanisms for cholesterol homeostasis respond to the active fraction of cholesterol in the plasma membrane. First, note that the major fraction of cell cholesterol resides in the plasma membranes and in the endomembrane compartments with which it is in intermittent continuity. The membranes of mitochondria, nuclei, and the endoplasmic reticulum or ER have much smaller pools [11]. For example, it has been estimated that > 80% of the cholesterol is in the plasma membranes of cultured human fibroblasts and ~ 0.5% is in their ER [12,29]. This distribution runs roughly parallel to the disparate affinities of the phospholipid species characteristic of each organelle; this is in keeping with the premise that cholesterol is held in bilayers as phospholipid complexes [11,30,31]. However, there may also be metabolic regulation of the intracellular allocation of cholesterol, e.g. through the phosphorylation of proteins [32]. Cholesterol is rapidly conveyed among cytoplasmic compartments; soluble cytosolic proteins and/or bridges at the contact sites between organelles seem more likely pathways than do vesicular shuttles [31].
The ER is clearly the nexus of the intracellular homeostatic activities that maintain plasma membrane cholesterol at its physiological set-point. This battery of ER functions includes cholesterol esterification by acyl-cholesterol acyltransferase (ACAT); cholesterol biosynthesis, governed by HMG-CoA reductase; and regulation of the expression of multiple proteins mediating cholesterol accretion by the sterol regulatory element binding protein, SREBP [1,2]. The proteins associated with each of these homeostatic processes have regulatory binding sites for cholesterol or oxysterols derived therefrom [33]. Small increments in plasma membrane cholesterol rapidly evoke large increases in ER cholesterol and in the activity of the regulatory proteins [3,29]. The threshold for the steep dose-response curve rests at the physiological set-point of the plasma membrane and ER pools. This behavior suggests, once again, that it is the active fraction of plasma membrane cholesterol – that in excess of its basal level – that drives the homeostatic responses.
The biosynthesis of oxysterols from cholesterol in the ER and mitochondria runs parallel to this regulatory system. Oxysterols serve as potent signals of cholesterol excess that counter its further accumulation [34]. For one thing, oxysterols activate nuclear receptors, thereby stimulating the expression of ABCA1 and promoting cholesterol export from the plasma membrane. In addition, oxysterols reduce the activity of homeostatic systems in the ER. They help to sequester SREBP and, in that way, limit the biosynthesis of proteins driving cholesterol accretion. Furthermore, 27-hydroxycholesterol reaches the ER within minutes of its biosynthesis in mitochondria and, there, stimulates the proteolytic inactivation of HMG-CoA reductase, curtailing cholesterol production [2,34,35]. The enhanced water solubility of oxysterols confers yet another homeostatic function: exit from overloaded peripheral cells (e.g. macrophages) for disposal by the liver [36]. The prediction of our hypothesis that oxysterol biosynthesis should ramp up acutely when cell cholesterol exceeds its physiologic quotient (i.e. the complexing capacity of plasma membrane phospholipids) has recently been confirmed for mitochondrial 27-hydroxycholesterol production (Lange, Ye, Lanier, Ory, Steck, in preparation). It would be worth looking for a similar threshold in the conversion of cholesterol to bile salts and steroid hormones.
References