Endocytosis Is Enhanced in Tangier Fibroblasts

A human genetic disorder, Tangier disease, has been linked recently to mutations in ATP-binding cassette protein A1 (ABCA1). In addition to its function in apoprotein A-I-mediated lipid removal, ABCA1 was also shown to be a phosphatidylserine (PS) translocase that facilitates PS exofacial flipping. This PS translocation is crucial for the plasma membrane to produce protrusions enabling the engulfment of apoptotic cells. In this report, we show that ABCA1 also plays a role in endocytosis. Receptor-mediated endocytosis, probed by both transferrin and low density lipoprotein, is up-regulated by more than 50% in homozygous Tangier fibroblasts in comparison with controls. Fluid-phase uptake is increased similarly. We also demonstrate that bulk membrane flow, including lipid endocytosis and exocytosis, is accelerated greatly in Tangier cells. Moreover, endocytosis is similarly enhanced in normal fibroblasts when ABCA1 function is inhibited by glyburide, whereas glyburide has no effect on endocytosis in Tangier cells. In addition, we demonstrate a decreased annexin V binding in Tangier fibroblasts as compared with controls, supporting the notion that PS transmembrane distribution is indeed defective in the presence of ABCA1 mutations. Furthermore, adding a PS analog to the exofacial leaflet of the plasma membrane normalizes endocytosis in Tangier cells. Taken together, these data demonstrate that ABCA1 plays an important role in endocytosis. We speculate that this is related to the PS translocase function of ABCA1. A loss of functional ABCA1, as in the case of Tangier cells, enhances membrane inward bending and facilitates endocytosis.

ATP-binding cassette (ABC) 1 protein A1 belongs to the ABC transporter superfamily, one of the largest and most highly conserved gene superfamilies (1). This family of transporters consists, minimally, of a highly conserved ABC-ATPase and a much less conserved multimembrane-spanning domain. By hydrolyzing ATP, ABC transporters are capable of transporting a wide variety of substrates including lipids across the membrane. The substrate specificity of ABC transporters is thought to be determined by the nature of the membrane-spanning domains. In humans, several clinical disorders are linked to defects in ABC transporters (2).
Tangier disease is a rare genetic disorder in humans characterized by extremely low plasma concentrations of high density lipoprotein cholesterol and apoprotein A-I (3). The absence of high density lipoprotein in Tangier patients was attributed to the impairment of apoprotein A-I-mediated cholesterol efflux (4). Recently, several mutations in ABCA1 have been linked genetically to the disease (5)(6)(7)(8). This was confirmed further by studies using a knockout mouse model (9,10). These animals also demonstrate a nearly complete absence of high density lipoprotein. The mechanisms by which ABCA1 facilitates apoprotein A-I-mediated lipid efflux, however, remain largely unknown. ABCA1 may serve as a receptor on the cell surface for apo-AI as indicated by chemical crosslinking studies (11,12). Apo-AI and ABCA1, however, were shown to have rather distinct diffusional coefficients on the plasma membrane, indicating that direct interaction between apo-AI and ABCA1 is limited (13). ABCA1 is known to specifically transport phosphatidylserine (PS) from the internal leaflet of the plasma membrane to the exofacial leaflet (PS translocase) (14). This could influence the lipid microenvironment on the cell surface and may sequentially facilitate apoprotein A-I binding (13). In addition, the plasma membrane, together with endosomal membranes, is the main reservoir of cellular free cholesterol (15). Endosomal membrane trafficking may be linked directly to apo-AI-mediated cholesterol and phospholipid efflux (16).
We have hypothesized that impairment of PS translocase caused by functional mutations in ABCA1 or inhibition of ABCA1 would result in the alteration in endosomal membrane trafficking. In the present report, we provide evidence demonstrating that a loss of functional ABCA1 results in an increase in endocytosis. Both receptor-mediated endocytosis and fluidphase uptake are enhanced in Tangier fibroblasts. Membrane recycling in the endosomal system is also accelerated. Importantly, this enhanced endocytosis can be duplicated in normal fibroblasts by pharmacologically inhibiting ABCA1 function. Furthermore, endocytosis in Tangier cells can be attenuated by lyso-PS, a phospholipid analog specifically inserted into the exofacial leaflet of the plasma membrane.

EXPERIMENTAL PROCEDURES
Cell Culture-Tangier fibroblasts (TD1 and TD2) and two normal control fibroblasts (N1 and N2) were obtained from Dr. J. Oram (Washington University, Seattle, WA). These cells were immortalized by transfection with human papillomaviruses E6 and E7 (17). Another independent Tangier primary fibroblast cell line (TD3) and two normal * This work was supported by a Canadian Institutes of Health Research (CIHR) Group Grant in Atherosclerosis (44360) (to R. M.) and CIHR/Wyeth-Ayerst Chair in Cardiovascular Disease (to R. M.). The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
Endocytosis Measurements-Cells were grown 2-3 days on coverslip dishes before the experiments. For LDL uptake, cells were changed into Dulbecco's modified Eagle's medium with 10% lipoprotein-deficient serum overnight to up-regulate LDL receptors. For all uptake experiments, the cells were preincubated in Dulbecco's modified Eagle's medium/HEPES plus 2 mg/ml BSA for 10 min at 37°C followed by either Cy3-transferrin (Tf) (10 g/ml) for 10 min or DiI-LDL (10 g/ml) for 30 min in the same medium. The cells were then rinsed three times by PBS and fixed with 4% paraformaldehyde. Surface receptor binding was performed by incubating cells on ice with Cy3-Tf or DiI-LDL for 30 min. This was followed by PBS rinses and fixation on ice with paraformaldehyde. Fluid-phase endocytosis was measured by incubating cells with fluorescein-dextran (5 mg/ml) for 30 min at 37°C. The cells then were rinsed with PBS three times and fixed with 4% paraformaldehyde.
Membrane Recycling-Liposomes containing BODIPY-C 5 -SM were made by mixing BODIPY-C 5 -SM and dioleylphosphatidylcholine, both in ethanol, by a molar ratio of 2:3. The mixture was then dried under N 2 and redissolved in ethanol. Liposome stock (20 mM total lipid concentration) was produced by rapidly injecting lipid mixture into medium 1 while vortexing followed by dialyzing against PBS overnight at 4°C to remove the ethanol. Cell surface membrane labeling was achieved by incubating cells with liposomes in medium 1 (50 M total lipid concentration) on ice for 30 min. The cells were then washed three times on ice before chasing at 37°C for 10 or 30 min to allow endocytosis to occur. Cells were cooled down to 4°C immediately and washed with ice-cold medium 1 containing 5% fatty acid-free BSA (six times and five min each) to remove remaining surface label. This was followed by a light fix with paraformaldehyde on ice, and the cells were examined immediately by microscopy.
For membrane recycling, cells were incubated with FM 1-43 (10 M) in medium 1 plus 2 mg/ml glucose for 15 min at 37°C to label the endosomal system. Cells were rinsed three times with room temperature dye-free medium/glucose to remove surface-associated dye. Cells were then chased on the microscope stage maintained at 32-34°C, and a time series of fluorescence images was acquired at 0, 0.5, 1, 2, 3, 5, 7, 10, 15, and 20 min.
Annexin V Surface Binding-Both control and Tangier cells were preincubated on ice for more than 30 min. Cells were then incubated with Alexa Fluor 488 conjugated annexin (1:5 dilution from manufacturer stock) in medium 1 for 1 h at 4°C. After three washes with cold PBS, cells were lysed with 2% Triton X-100. The lysates were measured for fluorescence using a spectrofluorometer, RTC-2000 (Photon Technology International, Inc., Princeton, NJ). The protein contents in the lysates were determined by the Lowry assay.
Fluorescence Microscopy Measurement-Most experiments reported in this paper involved fluorescence intensity measurement following a procedure described earlier (20,21). The measurement was performed on an Olympus IX50 inverted fluorescence microscope equipped with a cooled 12-bits CCD camera (microMax, Princeton Instruments). Images were acquired with a 40ϫ (0.75 numerical aperture) objective to collect fluorescence from the entire cell thickness. Total fluorescence intensity of each image was measured using Winview, the program that also drives the CCD camera. After background correction, the total fluorescence intensity of each image was then divided by the number of cells (ϳ20 -30 cells for fibroblasts) in the image to give a mean fluorescence intensity/cell (FI/cell). Each data point represented an average of 5-8 such images (ϳ200 cells). For membrane recycling experiments, a time series of images was taken from each dish of cells, and 4 -5 dishes were used for each cell line. For each dish, fluorescence intensity at time 0 was counted as 100% or normalized to 1. Fluorescence intensities at each time point were normalized according to initial fluorescence intensity (FI t /FI initial ) to give relative FI. The data are the average relative FIs at each time point from 3 or 4 dishes. Curve fitting was a single exponential decay using SigmaPlot.

ABCA1 Inhibitor, Glyburide, Increases Endocytosis in Hu-
man Fibroblasts-Because ABCA1 has been shown to influence lipid distribution in the plasma membrane, we first asked if the inhibition of ABCA1 would affect endocytosis. Normal human fibroblasts were preincubated with glyburide for 30 min and then incubated with Tf in the presence of glyburide for an additional 10 min at 37°C. Tf uptake was measured and compared with correspondent untreated cells. As shown in Fig. 1a, glyburide treatment resulted in an increase in Tf endocytosis in three independent control fibroblast cell lines, ranging from 30 to 50%. This strongly suggested that ABCA1 is indeed involved in membrane trafficking. We then tested glyburide on three Tangier fibroblast cell lines. The inhibition of ABCA1 by glyburide, as expected, had no detectable effect on all the Tangier fibroblasts tested (Fig. 1b). Because Tangier fibroblasts express only nonfunctional ABCA1, we then used these Tangier cells to study in detail the effect of impaired ABCA1 function on endocytosis.
Receptor-mediated Endocytosis Is Increased in Tangier Fibroblasts-We then compared receptor-mediated endocytosis in control and Tangier fibroblasts. Tf and LDL are known to bind to surface receptors and to be internalized through clathrin-coated pits (22). Fibroblasts were incubated with Cy3-Tf for 10 min at 37°C before microscopy observation. A typical Tf intracellular distribution in control fibroblasts is shown in Fig.  2a (left). Tf is seen in punctate endosome structures throughout the cells. There were also bright perinuclear clusters in most of the cells, presumably the recycling endosomes. Tf uptake in two Tangier fibroblasts (TD1 and TD2) evidently is greater compared with the two normal (N1 and N2) cell lines. This was verified by quantitative measurements in three Tangier fibroblasts and three normal controls as shown in Fig. 2b. There is an ϳ50% increase of Tf endocytosis during a 10-min incubation in Tangier cells in comparison with controls. Increased endo-cytosis of Tf is not caused by an altered surface receptor expression, because Tf surface binding was similar in all six cell lines (Fig. 2b, light gray bar).
We next measured LDL uptake in these cells. Both control and Tangier fibroblasts were incubated with DiI-LDL for 30 min at 37°C and then fixed for microscopy measurements. Typical DiI-LDL distribution is shown in Fig. 3a. To correct for possible differences in surface receptor expression, LDL surface binding was first measured by incubating cells with DiI-LDL on ice for 30 min (Fig. 3b, black bars). This was used to normalize final DiI-LDL uptake in each cell line (Fig. 3b). Similar to Tf, DiI-LDL uptake in all three Tangier cell lines is more than double that of control cells after a 30-min incubation. Together with Tf uptake results, we conclude that the receptor-mediated endocytosis is up-regulated in Tangier cells including both immortalized and primary Tangier fibroblasts.
Fluid-phase Uptake Is also Increased in Tangier Fibroblasts-In addition to receptor-mediated endocytosis, extracellular nutrients can also be taken up by cells through other mechanisms such as pinocytosis (23). A fluid-phase marker, fluorescein-dextran, was used to measure overall endocytosis. The cells were incubated with dextran for 30 min at 37°C, and the amount of uptake was quantified. The two Tangier cell lines accumulated more than double the amount of dextran during a 30-min period as compared with normal control cells (Fig. 4).
Both Membrane Lipid Endocytosis and Recycling Is Accelerated in Tangier Fibroblasts-Another method to examine endocytosis is to monitor membrane lipid flow. A short chain sphingomyelin fluorescence analog (BODIPY-C 5 -SM) was used to assay membrane endocytosis (24). Cells were incubated on ice with liposomes containing BODIPY-C 5 -SM to allow the sphingomyelin analog to insert into the cell surface. The cells then were rinsed with dye-free medium and chased at 37°C for 10 or 30 min to allow endocytosis to occur. By this time, a population of endosomes containing BODIPY-C 5 -SM became visible (Fig. 5a). The majority of the labeling was still on the cell surface, and the overall degree of labeling was similar in Tangier and control cells. The cells were then cooled down on ice once again, and the remaining surface BODIPY-C 5 -SM was removed by back exchange with fatty acid-free BSA. The only BODIPY-C 5 -SM remaining with cells at this point was within the endosomes. Intracellular BODIPY-C 5 -SM was then quantified as shown in Fig. 5b. BODIPY-C 5 -SM endocytosis was more than doubled in Tangier fibroblasts in comparison with normal controls at both 10 and 30 min. This is again in agreement with results described earlier.
Membrane recycling was measured with a fluorescence lipophilic dye, FM 1-43. This dye partitions between membrane and aqueous phase and is fluorescent only in the membrane (a 10 5 increase in quantum yield) (25). Cells were incubated with FM 1-43 for 15 min at 37°C to label endosomal compartments. The cells then were rinsed several times with medium to wash off the dye on the cell surface. The only dye left associated with cells at this time was in the endosomal compartments. The cells were then chased in a dye-free medium. During the chase, lipids in the endosomal compartments move back to the cell surface (membrane recycling), and FM 1-43 loses its fluorescence upon reaching the surface by rapidly dissociating from membrane. The rate of cell-associated fluorescence decay, therefore, is an indicator of membrane exocytosis (26). When this experiment was performed with control cells, membrane lipids recycle back to the cell surface with t1 ⁄2 ϳ 9 min (Fig. 6), similar to the rate observed by others (26). In Tangier cells, however, the fluorescence decay is faster with t1 ⁄2 ϳ 5 min. This indicates that the membrane flow from endosomal compart- ments back to the cell surface is accelerated in Tangier fibroblasts. Together with our observations with BODIPY-C 5 -SM described above, we conclude that the rate of membrane recycling in Tangier cells is about twice that of normal control fibroblasts.
Exogenous Lyso-PS Attenuates Endocytosis in Tangier Fibroblasts But Not in Normal Cells-ABCA1 has shown to be a PS translocase that transports PS from the internal leaflet to the exofacial leaflet of the membrane. The loss of functional ABCA1, such as in Tangier fibroblasts, could lead to an alteration of PS distribution across the bilayers of the plasma membrane. PS distribution between bilayers is thought to be main-tained minimally by two flippases (27); an aminophospholipid flippase flops PS inward, and ABCA1 flips outward. At steady state, most PS is on the internal leaflet of the plasma membrane with a small fraction in the exofacial leaflet. With impaired ABCA1 function and a normal aminophospholipid flippase, however, Tangier fibroblasts may have less PS in the exofacial leaflet. This might be expected to have an impact on membrane bending, which consequently could influence endocytosis. To verify that Tangier cells indeed have an altered PS distribution, Alexa Fluor 488 conjugated annexin V was used to quantitate the amount of PS in the outer leaflets. Annexin V is known to have a high affinity for PS and is widely used to detect apoptotic cells where the membrane asymmetry is lost (50). Alexa 488 annexin V surface binding (1 h at 0°C) was extremely low in fibroblasts, and we failed to detect any significant signal by fluorescent microscopy, indicating a predominant inner leaflet PS distribution in nonapoptotic cells. When we, however, measured annexin V binding by fluorospectrometer using lysates from a large number of cells, we consistently could detect decreased annexin V binding in Tangier fibroblasts (Fig. 7). This supports the notion that PS transmembrane distribution is indeed defective in the presence of ABCA1 mutations.
To determine whether enhanced endocytosis in Tangier cells is caused by an alteration in PS distribution, we used a lyso-PS that shares the same head group as PS but lacks one of the fatty acid chains. Lyso-PS, unlike PS, is not the substrate of aminophospholipid flippase (28). This enabled us to specifically supplement PS to the exofacial leaflet and possibly to "restore" Tangier cells to "normal" to some extent. Both immortalized and primary Tangier as well as normal fibroblasts were preincubated with lyso-PS (10 M) for 20 min and then incubated with Tf in the presence of lyso-PS for 10 min at 37°C. As shown in Fig. 8, Tf endocytosis was decreased by 30% in the two Tangier fibroblast cell lines treated with lyso-PS, whereas lyso-PS at this concentration had no significant effect on normal cells. The number of Tf receptors on the cell surface was not altered significantly by lyso-PS treatment (data not shown). This suggests that enhanced endocytosis in Tangier cells could be the direct consequence of an alteration in PS distribution caused by the loss of functional ABCA1. DISCUSSION The studies presented here demonstrate that the loss of ABCA1 function in Tangier fibroblasts is associated with enhanced endocytosis. Although the three Tangier cell lines employed here have different mutations in ABCA1, all share a common phenotype: the total absence of apo-AI-mediated lipid efflux (17,29). This was also observed in cells lacking ABCA1 or its equivalent such as in the ABCA1Ϫ/Ϫ mouse (9,14). In addition, overexpression of ABCA1 correlates with an increase in apo-AI-mediated lipid efflux (12,30). All these data strongly support an absence of functional ABCA1 in Tangier cell lines. We therefore used Tangier cells to probe the role of this transporter in endocytosis.
Using three primary or immortalized Tangier cell lines, we obtained several lines of evidence indicating that ABCA1 plays a role in membrane trafficking. First, we demonstrated that receptor-mediated endocytosis, specifically the uptake of Tf and LDL, was increased significantly in Tangier fibroblasts. We observed similar increases of fluid-phase uptake and endosomal membrane recycling in Tangier cells. Importantly, we demonstrated that the Tangier phenotype of enhanced endocytosis could be reproduced in normal control fibroblasts by a specific ABCA1 inhibitor, glyburide (31). As anticipated, this inhibitor had no effect on endocytosis in the three Tangier fibroblasts. Interestingly, glyburide has also been shown to have a potent inhibitory effect on cholesterol and lipid efflux in normal cells, reproducing what is observed in Tangier cells (32). These data support our conclusion that the loss of functional ABCA1 in Tangier cells is indeed responsible for the observed alteration in endocytosis.
It is not entirely clear at present how ABCA1 might regulate endocytosis. Signaling processes such as activation of phospholipases D and C were shown to be partially impaired in Tangier cells (33). This could have an impact on membrane trafficking. Activation of phospholipase D or C, however, was shown to increase endocytosis (34,35). It is therefore unlikely that the defects in the phospholipase D or C signaling pathway could explain our observations of enhanced endocytosis in Tangier cells.
Alternatively, as suggested by the "rescue effect" of lyso-PS in Tangier fibroblasts, ABCA1 is necessary for the maintenance of cross-leaflet PS distribution of the plasma membrane. Lipids in the plasma membrane are distributed asymmetrically (36), with phosphatidylcholine and SM mainly in the exofacial leaflet and phosphatidylethanolamine and PS in the internal leaflet of the membrane. This asymmetry is maintained actively by flippases and translocases that utilize ATP (37). Such a system provides cells with a stable yet highly responsive and dynamic membrane. Within a lipid bilayer, for example, any unidirectional lipid flipping or flopping between two leaflets would cause an alteration of relative surface area of the leaflets and therefore a change in membrane curvature (27). This would facilitate either inward (invagination) or outward membrane bending (evagination) (38). Membrane bending then initiates vesiculation leading to either endocytosis or blebbing.
Although some invaginations on the cell surface are assisted by protein coatings such as clathrin-coated pits or caveolae, a large number of vesiculation events occur without apparent coating. This is evident by the fact that fluid-phase uptake, a measure of overall vesiculation events, is rather insensitive to the inhibition of clathrin-coated pit endocytosis (39). This implies that vesiculation events could occur simply by the modifications of membrane leaflets, possibly through a dynamic adjustment of relative surface areas in a lipid bilayer. Hydrolysis of sphingomyelin (an exofacial leaflet lipid on the plasma membrane), for example, results in extensive vesiculations in the absence of visible coatings and is independent of ATP (40). This has been attributed to a decrease of surface area in the exofacial leaflet of the membrane, which in turn forces inward membrane bending (40). Similarly, enrichment of the internal leaflet with exogenous PS is known to produce an increase in endocytosis (41) that was dose-dependent and relying on PS being translocated into the internal leaflet of the plasma membrane by an aminophospholipid flippase (42). A lyso-PS, similar to the one employed in our study, had an inhibitory effect on endocytosis by remaining on the exofacial leaflet of the membrane (28). Interestingly, both receptor-mediated and fluidphase endocytosis was affected in this case, similar to our observations. When lyso-PS was added in a concentration that did not have any significant impact on the endocytosis of cells that have functional ABCA1, we found that endocytosis in Tangier cells was attenuated significantly. We cannot rule out the possibility that Tangier cells may have other metabolic If indeed ABCA1 functions as PS flippase, it would be responsible for supplying extra PS to the exofacial leaflet of the membrane while depleting PS from internal leaflet. Both these movements favor an outward membrane bending (43). This in fact is consistent with the observation that ABCA1 or its homologue is required for the engulfment of apoptotic cells (31,44), a process characterized by membrane protrusion initiated by an outward membrane bending (45). A loss of functional ABCA1 would lead to relatively more PS in the inner leaflet of the membrane. This relative excess inner leaflet area not only suppresses membrane protrusion (engulfment), but also favors membrane inward bending, which in turn facilitates endocytosis. Our interpretation is also in line with the recent finding that another ABC transporter, Drs2p, influences the formation of clathrin-coated vesicles from the trans-Golgi network in yeast (46). This ABC transporter was shown to be an aminophospholipid flippase (47) that translocates PS or phosphatidylethanolamine from luminal leaflet to cytoplasmic leaflet on the trans-Golgi network. This may imply that although the budding of clathrin-coated vesicles depends on protein coats, changes in membrane bending energy caused by defects in lipid Both normal and Tangier fibroblasts (TD1) were labeled on ice with liposomes containing BODIPY-C 5 -SM for 30 min, rinsed with ice-cold medium free of lipid analog, and then warmed to 37°C for a 10-or 30-min chase. The live cells were then imaged with fluorescence microscopy. BODIPY-C 5 -SM is seen on the plasma membrane and in the punctate dots, the endosomal compartments (a). The cells then were cooled on ice and washed with 5% BSA to remove surfaceremaining BODIPY-C 5 -SM. This resulted in BODIPY-C 5 -SM only in the intracellular compartments, representing the amount of membrane lipid endocytosed during the chase at 37°C. The amount of cell-associated BODIPY-C 5 -SM was then measured, and the results are shown in b. flippase (Drs2P) or translocase (ABCA1) could add an additional driving force to inhibit or promote initial invagination.
The most striking phenotype of Tangier disease is the total absence of lipid efflux from these cells. Recently, a retroendocytosis model has been proposed as a possible mechanism for lipid efflux (16). Lipid acceptors, either apo-AI or high density lipoprotein, may constantly traffic through endosomes and then recycle back to the cell surface. Possibly within endosomal organelles, the acceptors acquire lipids. Tangier fibroblasts have an enhanced endocytosis yet totally lack lipid efflux. This apparent paradox may be resolved by the fact that the binding of apo-AI is correlated positively with the expression of a functional ABCA1 (12,13). Without functional ABCA1, Tangier cells would not be able to bind to apo-AI efficiently or to shed lipid. Apo-AI cell association and binding in these fibroblasts is extremely low, and in these studies we failed to detect any significant binding or association (data not shown).
An outstanding question regarding ABCA1 is how a PS translocase activity possibly facilitates lipid efflux. Using cell lines with or without ABCA1 expression, Fielding et al. (32) recently demonstrated that to acquire cholesterol apo-AI must first be lipidated with phospholipids, especially phosphatidylcholine. Lipidated apo-AI can then efficiently acquire cholesterol independent of ABCA1 expression. This is in line with recent evidence that ABCA1-mediated cholesterol efflux does not depend on "rafts," a membrane microdomain rich in cholesterol and sphingomyelin but depleted in phosphatidylcholine (48). It seems unlikely that PS can interact with cholesterol directly. In contrast, the relationship between PS and phosphatidylcholine may be important and remains to be explored.
In summary, we report that ABCA1 plays an important role in endosomal membrane trafficking. These data support recent studies indicating that ABCA1 functions as a PS translocase (14) and are in line with reports that ABCA1-green fluorescent protein is mainly localized to the plasma membrane (14,49). Further work is required to understand whether ABCA1 also functions in other vesicular transport processes.
FIG. 6. Lipid exit from endosomal compartments is accelerated in Tangier cells. Both control and Tangier (TD1) cells were labeled with 10 M FM 1-43 for 15 min at 37°C and rinsed several times with PBS/glucose at 37°C. The cells then were moved to a microscope stage maintained at 32-34°C. Fluorescence images were taken at 0, 0.5, 1, 2, 3, 5, 7, 10, 15, and 20 min. FI at each time point (FI t ) was divided by initial FI (FI initial ) to normalize cell-associated fluorescence: FI t /FI initial . The data are averages from 3-4 dishes at each time point. Curve fitting was by a single exponential decay.
FIG. 7. Annexin V surface binding is reduced in Tangier fibroblasts. Both primary control (N3) and primary Tangier (TD3) fibroblasts were incubated with Alexa 488 conjugated annexin V for 1 h at 0°C and then lysed with 2% Triton X-100. Cell-associated fluorescence (FI) was measured and then ratioed to cell protein (mg/ml). A representative binding from three measurements is shown here, and error bars are the standard deviations (p Ͻ 0.001).