Remodeling and shuttling. Mechanisms for the synergistic effects between different acceptor particles in the mobilization of cellular cholesterol

Abstract

In normal physiology, cells are exposed to cholesterol acceptors of different sizes simultaneously. The current study examined the possible interactions between two different classes of acceptors, one large (large unilamellar phospholipid vesicles, LUVs) and one small (HDL or other small acceptors), added separately or in combination to Fu5AH rat hepatoma cells. During a 24-hour incubation, LUVs of palmitoyl-oleoyl phosphatidylcholine at 1 mg phospholipid (PL) per milliliter extracted approximately 20% of cellular unesterified cholesterol (UC) label and mass in a slow, continuous fashion (half-time [t1/2] for UC efflux was approximately 50 hours) and human HDL3 at 25 micrograms PL per milliliter extracted approximately 15% cellular UC label with no change in cellular cholesterol mass (t1/2 of approximately 8 hours). In contrast, the combination of LUVs and HDL3 extracted over 90% of UC label (t1/2 of approximately 4 hours) and approximately 50% of the UC mass, indicating synergy. To explain this synergy, specific particle interactions were examined, namely, remodeling, in which the two acceptors alter each other's composition and thus the ability to mobilize cellular cholesterol, and shuttling, in which the small acceptor ferries cholesterol from cells to the large acceptor. To examine remodeling, LUVs and HDL were coincubated and reisolated before application to cells. This HDL became UC depleted, PL enriched, and lost a small amount of apolipoprotein A-I. Compared with equivalent numbers of control HDL particles; remodeled HDL caused faster efflux (t1/2 approximately 4 hours) and exhibited a greater capacity to sequester cellular cholesterol over 24 hours (approximately 38% versus approximately 15% for control HDL), consistent with their enrichment in PL. Remodeled LUVs still extracted approximately 20% of cellular UC. Thus, remodeling accounted for some but not all of the synergy between LUVs and HDL. To examine shuttling, several approaches were used. First, reisolation of particles after an 8-hour exposure to cells revealed that HDL contained very little of the cellular UC label. The label was found almost entirely with the LUVs, suggesting that LUVs continuously stripped the HDL of cellular UC. Second, bidirectional flux studies demonstrated that LUVs blocked the influx of HDL UC label into cells, while the rate of efflux of cellular UC was maintained. These kinetic effects explained the massive net loss of cellular UC to LUVs with HDL. Third, cyclodextrin, an artificial small acceptor that does not acquire PL and hence does not become remodeled, exhibited substantial synergy with LUVs, supporting shuttling. Thus, the presence of large and small acceptors together can overcome intrinsic deficiencies in each. Small acceptors are efficient at extracting cellular cholesterol because they approach cell surfaces easily but have a low capacity, whereas large acceptors are inefficient but have a high capacity. When present simultaneously, where the small acceptor can transfer cholesterol quickly to the large acceptor, high efficiency and high capacity are achieved. The processes responsible for this synergy, namely, remodeling and shuttling, may be general phenomena allowing cooperation both during normal physiology and after therapeutic administration of acceptors to accelerate tissue cholesterol efflux in vivo.

FIG 1

Time courses of [³H]UC efflux from rat Fu5AH hepatoma cells to LUVs, serum, HDL, and combinations. Fu5AH cells trace-labeled with [³H]UC in 22-mm tissue-culture wells were incubated for 24 hours at 37°C in a humidified incubator (5% C0₂) with 1.0 mL of test medium (MEM/HEPES containing 1% BSA, 1.0 μg Sandoz compound 58035, and the indicated cholesterol acceptors). A, Time course for the release of UC label from cells exposed to POPC LUVs 1 mg PL per milliliter (▼), 1 % serum (■), and the combination of these two acceptors at these concentrations (▲). The calculated sum of cholesterol release to LUVs and to serum is shown by ◆, dashed line and is marked “additive.” B, Time course for the release of cellular [³H]UC by POPC LUVs 1 mg PL per milliliter POPC LUVs (▼), HDL₃ at 25 μg PL per milliliter (■), and the combination of these two acceptors at these concentrations (▲). The calculated sum of cholesterol release to LUVs and to HDL added separately is shown by ◆, dashed line and is marked “additive.” Data points represent the mean±SD of the value derived from three cell wells after correction for efflux to control media (MEM/HEPES containing 1% BSA). All curves were fit to a single-exponential decay equation as described in “Methods.”

FIG 2

The depletion of cellular UC and EC mass in cholesterol-loaded Fu5AH cells exposed to LUVs, HDL, and the combination. A, Net loss of cellular UC mass expressed as micrograms UC per milligram cell protein. B, Net loss of cellular EC mass in micrograms EC per milligram cell protein after a 24-hour exposure to the indicated acceptors. Data shown represent the mean±SD of the net loss of sterol measured from three cell wells after correcting for sterol mass present in cells exposed to control media (MEM/HEPES/1% BSA). UC levels in control cells after a 24-hour exposure to control medium were 22.5±1.2 μg/mg protein and EC levels were 53.4±1.3 μg/mg protein.

FIG 3

Isolation and characterization of LUVs and HDL after coincubation. LUVs and HDL used to create remodeled particles were coincubated overnight at 37°C (ratio of LUV PL to HDL PL was 10:1) in the absence of cells. After coincubation, LUVs and HDL were separated on a Sepharose CL-4B column (90×1.5 cm) equilibrated with TBS and run at 0.75 mL/min at 25°C (A). Fractions corresponding to each repurified particle were pooled and concentrated before determination of particle diameter by native PAGGE. Electrophoretic scans obtained from Phast Gel native PAGGE of HDL coincubated with saline (B), HDL remodeled by LUVs (C), and molecular-weight standards (D) are shown. The peaks marked 1 through 4 of panel D represent the migration of the molecular-weight standards (1) thyroglobulin, (2) ferritin, (3) catalase, and (4) lactase dehydrogenase.

FIG 4

Cholesterol efflux from cells to control and remodeled HDL particles. The cell efflux conditions were the same as described for Fig 1. A, Time courses for the release of labeled UC from cells exposed to HDL (25 μg PL per mL) that had been preincubated with saline (■ “original” HDL) and HDL remodeled (see “Methods” and Fig 3) by LUVs (○) placed on cells at the same particle numbers (ie, same EC concentrations) as the original HDL. B, Time courses for the release of cellular [³H]UC by remodeled LUVs at 1 mg PL per milliliter (∇), remodeled HDL placed on cells at the same particle numbers present in original HDL at 25 μg PL per milliliter (○), and the combination of these remodeled particles at these concentrations (▲) versus time. The calculated sum of cholesterol release to remodeled LUVs and HDL is shown by ◊, line-dash-line and is marked “additive.” C, Relative contributions of remodeling and shuttling to cellular cholesterol efflux. The time courses for the release of cellular cholesterol to (1) the sum of release to original LUVs and to HDL added separately (◆, dashed line, taken from Fig 1B), (2) the sum of release to remodeled LUVs and remodeled HDL added separately after reisolation (◊, line-dash-line, taken from Fig 4B), and (3) remodeled LUVs and remodeled HDL added in combination after reisolation (▲, taken from Fig 4B) are shown. The vertical arrow A (between “original additive” and “particle combination”) represents total synergy, and the distance marked by arrow B (between “remodeled additive” and “particle combination”) represents residual synergy. Data points represent the mean±SD of [³H]UC remaining in cells determined from three cell wells after correcting for efflux to control media. All curves were fit to a single-exponential decay equation as described in “Methods.”

FIG 5

Bidirectional flux of [³H]UC from labeled cells to media containing [¹⁴C]UC-labeled HDL with or without LUVs. Fu5AH cells were prepared as described in Fig 1. [¹⁴C]UC-labeled HDL was generated as described in “Methods” and then coincubated with LUVs or saline before addition to cells. Data shown represent the fraction of [³H]UC remaining in cells and the fraction of [¹⁴C]UC that had accumulated in cells from media HDL at the indicated times. Solid symbols represent the efflux of [³H]UC (■) from cells to HDL₃ at 200 μg PL per milliliter and the influx of [¹⁴C]UC (●) from HDL₃. Open symbols represent the efflux of [³H]UC (□) and the influx of [¹⁴C]UC (○) in cells exposed to 200 μg PL per milliliter HDL₃ and 1 mg/mL LUV PL. Each data point represents the mean±SD obtained from three cell wells.

FIG 6

Mechanisms for the synergy between LUVs and HDL in the mobilization of cellular cholesterol. A, Cholesterol efflux from a cell monolayer to LUVs and HDL in the absence of any interactions. B, Synergy between LUVs and HDL due to remodeling. HDL remodeled by LUVs becomes UC depleted and PL enriched, making it a more efficient acceptor for cellular cholesterol. C, Synergy between LUVs and HDL due to a continuous shuttling of UC from cells to LUVs by HDL, which acts as a carrier.