Phosphoproteins Regulated by the Interaction of High-Density Lipoprotein With Human Skin Fibroblasts
Abstract Interaction of HDL with cells activates protein kinase C (PKC), a process that may be important in stimulating efflux of excess cellular cholesterol. Here we report that HDL treatment of cholesterol-loaded fibroblasts increases 32P labeling of three acidic phosphoproteins. These phosphoproteins, called pp80, pp27, and pp18 based on apparent Mr in kD, were also phosphorylated by acute treatment of cells with phorbol myristate acetate, suggesting that they are regulated in response to PKC activation. The HDL-stimulated phosphorylation of pp80 and pp18 was significant after only 30 seconds and was sustained for at least 30 and 120 minutes, respectively, while increased phosphorylation of pp27 was transient, reaching a maximum at 10 minutes. Both pp27 and pp18 were phosphorylated on serine/threonine residues, whereas pp80 was phosphorylated on serine/threonine and tyrosine residues. Immunoprecipitation studies suggested that pp80 is the myristoylated alanine-rich C kinase substrate protein, but the identities of pp27 and pp18 are unknown. HDL and trypsin-digested HDL stimulated phosphorylation of pp80 and pp27, while purified apoA-I, apoA-II, or apoE had no stimulatory effects, indicating that the active component in HDL was trypsin resistant and unlikely to be an apolipoprotein. Conversely, HDL, apoA-I, apoA-II, and apoE all stimulated pp18 phosphorylation, while trypsin-digested HDL had less effect, consistent with pp18’s being responsive to HDL apolipoproteins. Treatment of cholesterol-depleted cells with apoA-I also stimulated phosphorylation of pp18, but only transiently. These results suggest that HDL interaction with cells activates diverse PKC-mediated pathways that target different phosphoproteins. Of these three phosphoproteins, only pp18 has a phosphorylation response consistent with its being involved in apolipoprotein-mediated lipid transport.
The concentration of HDL in plasma has an inverse correlation with the incidence of coronary heart disease,1 2 suggesting that HDL is antiatherogenic. HDL is hypothesized to carry cholesterol from peripheral cells to the liver for eventual elimination from the body, a process referred to as “reverse cholesterol transport,”3 High-affinity binding sites for HDL have been described using cultured cells and plasma membrane preparations. Cell-surface HDL binding proteins have been identified that may be physiological receptors for HDL.4 5 6 7 8 The interaction of HDL with these cell-surface binding sites has been found to be mediated by HDL apolipoproteins, notably apoA-I, apoA-II, apoE, and apoA-IV.4 9 This interaction appears to stimulate excretion of excess cholesterol from cells.10 11
Several laboratories have investigated the cellular signaling responses generated as a result of HDL interacting with cells. HDL or apoA-I–containing phospholipid vesicles have been shown to induce diacylglycerol production in adipocytes,12 platelets,13 and fibroblasts.14 15 16 17 Both the lipid and protein components in HDL have been implicated in activating different classes of cellular phospholipases15 16 17 18 and mobilizing intracellular calcium.19 The biological consequences of these diverse signaling responses is unclear. It has been reported that activation of PKC by HDL components stimulates the excretion of a variety of cellular molecules, including cholesterol from fibroblasts,14 15 16 17 18 19 20 smooth muscle cells,21 and adipocytes12 ; surfactant from lung alveolar type II cells22 ; and endothelin-1 from endothelial cells.23 24 HDL-induced signals may also be involved in stimulating mitogenesis.25
Phosphorylation of key intracellular proteins is an expected response to the interaction of HDL with the cell surface and protein kinase activation. In the current study, we identified by high-resolution 2D PAGE three phosphoproteins that were further phosphorylated as a result of HDL interacting with cholesterol-loaded human skin fibroblasts. Two of these phosphoproteins with Mr of 80 kD (pp80) and 27 kD (pp27) were phosphorylated in response to the cellular interaction of protease-resistant HDL components and were unaffected by purified apolipoproteins, suggesting that HDL lipids may be involved in regulating these phosphoproteins. The third phosphoprotein, an 18-kD protein (pp18), was rapidly phosphorylated by the interaction of lipid-free apoA-I, apoA-II, and apoE with cells. These findings support previously published results suggesting the involvement of multiple cellular signaling pathways stimulated by different HDL components. Only one of these phosphoproteins, pp18, is regulated in a manner consistent with its being involved in modulating apolipoprotein-mediated cholesterol efflux from cells.
Preparation of HDL, Apolipoproteins, and Trypsinized HDL
HDL3 (herein referred to as HDL) was isolated from the plasma of healthy male volunteers by standard sequential ultracentrifugation techniques at d=1.125 to 1.210 g/mL. HDL was subjected to heparin-agarose affinity chromatography to remove apoB- and apoE-containing particles.26 Proteolytically digested HDL particles were generated by treating HDL with trypsin for 10 minutes at 37°C at an HDL:trypsin protein ration of 40:1.10 ApoA-I, apoA-II, and apoE were purified from HDL as described.20 27
Human skin fibroblasts obtained from skin biopsies were grown to confluence in 60-mm dishes, using DMEM supplemented with 10% fetal bovine serum. Cells were washed twice with PBS containing 2 mg/mL BSA and incubated for 24 hours in serum-free DMEM containing 2 mg/mL BSA plus 30 μg/mL cholesterol to load cells with sterol.10 20 27 Cells were then washed twice with DMEM containing 1 mg/mL BSA and maintained overnight in the same medium to allow cellular pools of cholesterol to equilibrate. For experiments using fibroblasts not loaded with cholesterol, cells were incubated for 48 hours with medium containing 10% LPDS.
Radiolabeling of Cells
Cells were washed twice with PF-DMEM containing 1 mg/mL BSA and incubated with PF-DMEM/BSA containing 32PO4 (0.5 mCi per plate) for 1 hour to label the endogenous adenosine triphosphate pool. Cells were then washed four times with PF-DMEM/BSA.
Incubation of Cells With HDL, Apolipoproteins, and Trypsinized HDL
Cells were incubated with PF-DMEM/BSA (control) or PF-DMEM/BSA plus HDL (50 μg protein/mL), trypsinized HDL (normalized to the same phospholipid content as HDL), or apoA-I, apoA-II, or apoE (10 μg/mL). Cells were then chilled on ice, washed three times in ice-cold buffer A (10 mmol/L Tris-HCl, pH 7.4, 2.0 mmol/L EDTA, 150 mmol/L NaCl, 1 mmol/L PMSF, 1 mmol/L leupeptin, 1 mmol/L pepstatin, 1 mmol/L aprotinin, 50 mmol/L NaF, 0.1 mmol/L NaVO3), dislodged from the dishes with a policemen, and sedimented by microcentrifugation.
Extraction of Cellular Protein
To extract total cellular proteins, the pelleted cells were solubilized in 200 μL of buffer B (50 mmol/L Tris-HCl, pH 7.4, 35 mmol/L SDS, 1.0 mmol/L EDTA, 50 mmol/L NaF, 0.1 mmol/L NaVO3, 0.1 mol/L 2-mercaptoethanol, and buffer A protease inhibitors). Samples were immediately boiled for 3 minutes and then centrifuged at 100 000g for 30 minutes at 4°C. The solubilized protein in the supernatant was precipitated by adding 700 μL cold acetone and incubating on ice for 1 hour, pelleted by microcentrifugation, dried in a speed vacuum concentrator, and solubilized in 50 μL of buffer C (8.0 mol/L urea, 10 mmol/L Tris-HCl, pH 7.4, 65 mmol/L CHAPS, 4.4% 3/10 ampholytes, 2.0 mmol/L SDS, and 120 mmol/L 2-mercaptoethanol) with agitation at room temperature for 1 hour. Protein was quantitated using the Bradford protein assay.28
To determine membrane association of the phosphoproteins, cells were collected and sonicated in 200 μL of buffer A followed by centrifugation at 100 000g for 30 minutes. The soluble proteins in the supernatant were precipitated with 700 μL cold acetone and prepared for IEF as described above. Insoluble proteins in the pellet were solubilized in 200 μL of buffer B and boiled for 3 minutes. The sample was then recentrifuged at 100 000g for 30 minutes, and the supernatant fraction was precipitated with 700 μL cold acetone and prepared for IEF.
High-Resolution 2D PAGE
Proteins resuspended in buffer C (30 μg) were focused for 15 hours at 400 V in polyacrylamide tube gels (11 cm×2 mm) as described by O’Farrell.29 Extruded gels were equilibrated for 15 minutes before separation, using a 7% to 17% gradient SDS-PAGE, run at 40 V for 14 hours. Slab gels were fixed and shrunk in 50% ethanol:10% acetic acid (2 hours) and then dried between two cellophane sheets. Radiolabeled proteins were identified by autoradiography.
Quantitation of Phosphoproteins
Phosphoproteins were quantitated by phosphorimaging at the PhosphorImager Facility, Markey Molecular Medicine Center, University of Washington, using a PhosphorImager model 400 S (Molecular Dynamics). Each spot, corresponding to a particular phosphoprotein, was quantitated using an equal ellipsoid volume.
Specific precautions were required to detect and quantitate phosphoproteins regulated as a result of HDL stimulation. For instance, if one dish of cells was exposed to 50 μg/mL HDL in PF-DMEM/BSA, a control dish was exposed to only PF-DMEM/BSA for the same amount of time. Both plates were scraped, cells were pelleted, and the proteins were separated by 2D gels. Criteria for control phosphoproteins used to quantify the HDL effect included: (1) separation from other phosphoproteins; (2) unresponsiveness to HDL; and (3) a phosphorylation ratio of a particular control protein to other control proteins of 1.0±.1 after various treatments (see formula below and Fig 1A⇓). The control proteins served two purposes: (1) to standardize the gels for variations in protein loading and [γ−32P]ATP specific activity, and (2) to serve as a standard for calculating the degree of protein 32P labeling. Changes in protein 32P labeling was determined using the formula (RPt/CPt)/(RPc/CPc), where RP and CP are individual responsive and control 32P-labeled proteins, respectively, isolated on gels containing proteins from either treated (t) or control (c) fibroblasts. Between three and five control phosphoproteins were used to calculate an average degree of phosphorylation for each responsive phosphoprotein.
Phosphorylated proteins were separated using 2D gel electrophoresis followed by transfer to a PVDF membrane (Millipore Corp). The immobilized phosphoproteins were detected after autoradiography and then acid hydrolyzed as described by Kamps and Sefton.30 Separation of the resulting phosphoamines was then performed using the method outlined by Munoz and Marshall,31 and detected by autoradiography.
Immunoprecipitation With MARCKS Antibody
Proteins from cells radiolabeled and incubated with HDL were obtained as described above, and the non–membrane-associated protein samples were precleared using agarose-protein A (Bio-Rad) for 1 hour at 4°C. Bovine brain MARCKS antibody (gift of Angus C. Nairn, The Rockefeller University) was then added, and samples were incubated overnight at 4°C, followed by addition of agarose-protein A for 1 hour to immunoprecipitate MARCKS protein. The agarose beads were washed five times with 0.5 mL 50 mmol/L Tris-HCl, (pH 7.4), 0.15 mol/L NaCl at 4°C and stored at −70°C.
Immunoprecipitated proteins were boiled for 5 minutes using SDS-PAGE sample buffer and then separated using 8% SDS-PAGE. Proteins were transferred to nitrocellulose (16 hours at 30 V), and 32P-labeled proteins were visualized by autoradiography. Tyrosine phosphorylation of immunoprecipitated protein was determined by incubating the nitrocellulose in blocking buffer (PBS, pH 7.4, 0.05% Tween 20, and 2% (wt/vol) BSA) for 1 hour, and then incubating with 1:500 dilution of anti-phosphotyrosine antiserum (Sigma) for 1 hour. After washes (3×10 minutes), secondary antibody (conjugated with horseradish peroxidase) was incubated with the nitrocellulose membrane for an additional hour. Tyrosine-phosphorylated proteins were visualized using enhanced chemiluminescence (ECL, Amersham).
Average values ±SEM were determined by combining values from separate experiments performed under identical conditions. Student’s t test was performed where indicated to determine significance between the groups of data.
Identification of HDL-Regulated Phosphoproteins
To identify cellular phosphoproteins that are phosphorylated in the presence of HDL, 32P-labeled human skin fibroblasts were incubated with HDL (50 μg/mL) for 10 minutes, and total cellular proteins were separated using high-resolution 2D PAGE. For most experiments, cells were first incubated with serum-free medium containing cholesterol, since previous studies showed that HDL had maximum effects on cholesterol excretion when cells were both growth arrested by serum deprivation and loaded with cholesterol.10 At 50 μg/mL, HDL promotes cholesterol efflux from cholesterol-loaded fibroblasts by mostly high-affinity processes, which saturate at 20 to 30 μg/mL HDL.10 20 26 Both IEF and nonequilibrium pH gradient electrophoresis were initially performed to identify 32P-labeled proteins along the entire pH spectrum (data not shown). Using a specific combination of ampholytes, 2D PAGE revealed three phosphoproteins that exhibited increased 32P labeling as a result of HDL treatment (Fig 2A⇓ and 2B⇓). The largest phosphoprotein (pp80) had an apparent Mr of 80 kD and a pI of 4.5. The second largest phosphoprotein (pp27), had an apparent Mr of 27 kD and a pI of approximately 6.0. The third phosphoprotein (pp18) had an apparent Mr of 18 kD and a pI of 4.7 to 4.8. No other 32P-labeled proteins detected by this procedure consistently showed a change in phosphorylation in response to HDL treatment.
Acute treatment of fibroblast cells with PMA (160 μmol/L for 10 minutes) also increased the phosphorylation of these three phosphoproteins (Fig 2C⇑). In contrast, the membrane-permeable cAMP analogue, 8-bromo-cAMP (1 mmol/L), did not affect their phosphorylation (Fig 2D⇑).
By narrowing the pH gradient in the first dimension (IEF), it was possible to resolve pp18 into a doublet on the basis of net charge. These two spots, referred to as pp18A (pI=4.7) and pp18B (pI=4.8), became evident after HDL treatment (Fig 3⇓). It remains to be determined whether these two spots represent distinct proteins or isoforms of the same protein differing in pI or phosphorylation state. In any event, both phosphoproteins were phosphorylated by HDL treatment and thus were quantified together for the remainder of the study.
The phosphorylation of these three phosphoproteins by HDL was examined as a function of time (Fig 1⇑). The effect of HDL on protein phosphorylation was quantified using a formula that normalizes 32P labeling of a responsive phosphoprotein after HDL treatment to the labeling of the same protein after control incubations (no HDL) and to the labeling of three to five control proteins that are unaffected by HDL treatment (see “Methods”). This corrects for both intragel and intergel variability between experiments and permits statistical analysis. When values for each control protein were calculated using this formula, the relative 32P labeling was unchanged by treatment of cells with HDL during 30-minute incubations (Fig 1A⇑). In contrast, pp80 was rapidly phosphorylated (30 seconds, the earliest time point in Fig 1⇑) to 200% to 250% of its basal phosphorylated state, which was sustained throughout the 30-minute treatment period. Increased phosphorylation of pp27 was transient, being significantly phosphorylated only at 10 minutes to approximately 150% of its basal phosphorylated state. Phosphorylation of pp18 increased significantly at all time points to 130% to 180% of its basal phosphorylated state.
Membrane Association of pp80, pp27, and pp18
To test whether these cellular phosphoproteins were membrane bound or soluble, intact cells were treated with HDL for 10 minutes and sonicated immediately after dislodgment from the plate. A soluble (supernatant) and membrane (pellet) fraction were isolated after centrifugation at 100 000g. Analysis by 2D PAGE indicated that pp80 was present in both the soluble and membrane fractions in roughly equal amounts (Fig 4⇓). In contrast, pp27 was detected only in the membrane fraction and pp18 was detected only in the soluble fraction. The distribution of these three phosphoproteins did not change as a result of HDL treatment (data not shown).
Phosphoamine Analysis of pp80, pp27, and pp18
To further examine the classes of protein kinases involved in HDL-stimulated phosphorylation of pp80, pp27, and pp18, we analyzed their 32P-labeled phosphoamine content after acute HDL treatment of fibroblasts. Tyrosine and threonine/serine were heavily phosphorylated in pp80 (Fig 5⇓). Phosphoproteins pp27 and pp18 were phosphorylated exclusively on threonine/serine residues. Although threonine and serine were not resolved in Fig 5⇓, other studies indicated that all three proteins were phosphorylated on both threonine and serine. The same pattern of 32P-labeled amino acids was observed in the basal state (data not shown), but phosphorylation of tyrosine (pp80) and threonine/serine (pp80, pp27, pp18) residues was increased by treatment of cells with HDL. These results imply that phosphorylation of pp80, but not pp27 and pp18, is mediated by tyrosine kinase(s), as well as PKC. Further support for this conclusion stems from experiments showing that chronic (24 hours) treatment of fibroblasts with PMA only partially (50% to 80%) reduced the ability of HDL to stimulate phosphorylation of pp80, whereas it completely abolished the stimulatory effects of HDL on pp27 and pp18 phosphorylation (data not shown). Immunoblot analysis revealed that 24-hour PMA treatment completely suppressed expression of PKCα, and PKCβ, the major PMA-sensitive PKC isotypes in cultured fibroblasts.32
HDL-Stimulated Phosphorylation of the MARCKS Protein
The apparent Mr and pI of pp80 are similar to those previously reported for the MARCKS protein,33 a ubiquitous substrate for PKC. Polyclonal antibody generated against bovine brain MARCKS (provided by Angus C. Nairn, The Rockefeller University) immunoprecipitated a 32P-labeled 80-kD protein that was phosphorylated to a greater extent by treatment of cells with HDL (Fig 6A⇓), consistent with the possibility that pp80 is the MARCKS protein. Immunoblot analysis with anti-phosphotyrosine IgG indicated that the immunoprecipitated protein was phosphorylated on tyrosine residues (Fig 6B⇓), consistent with the phosphoamino acid analyses.
Effects of Trypsinized HDL and Purified ApoA-I on Phosphorylation of Cellular Proteins
Previous studies10 20 34 showed that HDL and purified apoA-I, but not trypsin-digested HDL, stimulate efflux of intracellular sterol. This process appears to involve activation of PKC, suggesting that sterol efflux is initiated by signals elicited by the interaction of HDL apolipoproteins with cells. To determine whether protein phosphorylation has the same response pattern, we compared the effects of HDL, trypsinized HDL, and purified apoA-I on phosphorylation of the three HDL-stimulated phosphoproteins.
Trypsinized HDL consistently and significantly stimulated the phosphorylation of pp80 at both 0.5 and 10 minutes (Fig 7A⇓). Also, incubation of cells with trypsinized HDL increased phosphorylation of pp27 at 10 minutes, similar to what was observed with unmodified HDL (Fig 7B⇓). This suggested that trypsin-insensitive components of HDL, possibly lipid, were responsible for stimulating phosphorylation of these two phosphoproteins. Purified apoA-I had no significant effect on phosphorylation of either of these two proteins when added at a concentration (10 μg/mL) twice that needed to maximally stimulate lipid efflux from cells.27 34 35
In contrast to its effects on pp80 and pp27, lipid-free apoA-I consistently and significantly (Fig 7C⇑) increased phosphorylation of pp18 at 0.5 and 10 minutes, and trypsinized HDL had only a slight, insignificant ability to stimulate the phosphorylation of pp18. These results indicate that the phosphorylation of pp18 is stimulated by apoA-I and trypsin-sensitive proteins associated with HDL.
Effects of Purified ApoA-I, ApoA-II, and ApoE on Phosphorylation of Cellular Proteins
Apolipoprotein-mediated cholesterol efflux has broad specificity for multiple apolipoproteins, including apoA-I, apoA-II, and apoE.27 35 36 We examined the effects of these different apolipoproteins on phosphorylation on pp80, pp27, and pp18. Treatment of cholesterol-loaded fibroblasts for 10 minutes with 10 μg/mL of any of these three apolipoproteins had no significant effect on phosphorylation of pp80 and pp27, except that apoE caused a significant decrease in phosphorylation of pp27 compared with controls (P<.05) (Fig 8⇓). In contrast, treatment with apoA-I, apoA-II, and apoE all significantly increase phosphorylation of pp18 (P<.002). Thus, phosphorylation of pp18 is stimulated by multiple exchangeable apolipoproteins.
Effects of ApoA-I on Phosphorylation of Proteins in Cholesterol-Depleted Cells
Previous studies showed that cellular binding of HDL37 and apolipoprotein-mediated lipid efflux36 37 are enhanced by cholesterol loading of cells. To determine whether apoA-I also stimulates phosphorylation of pp18 in cholesterol-depleted cells, 32P labeling of this and the two other phosphoproteins was measured as a function of time of treatment with apoA-I after cells were grown for 48 hours in the presence of LPDS. HDL apolipoprotein-mediated sterol efflux is almost completely suppressed when fibroblasts are first grown in the presence of serum lacking a source of cholesterol.10 Exposure of LPDS-treated cells to apoA-I caused a decrease in phosphorylation of pp80 and pp27 during 10-, 20-, and 30-minute incubations (Fig 9A⇓), but only the 10- and 30-minute time points for pp80 reached significance (P<.05). In contrast, apoA-I significantly increased phosphorylation of pp18 (P<.005) after 10- and 20-minute incubations. When cells were exposed to apoA-I for 30 minutes, however, there was only a modest increase in phosphorylation of pp18 that was not significantly different from baseline (P=.1). These results suggest that apoA-I also stimulates phosphorylation of pp18 in cholesterol-depleted cells, but only transiently.
To further test the possibility that apoA-I–stimulated phosphorylation of pp18 was transient in cholesterol-depleted cells but sustained in serum-deprived, cholesterol-loaded cells, we measured the ability of apoA-I to stimulate phosphorylation of pp18 during 2-hour chase incubations after cells were incubated with either LPDS or cholesterol-rich, serum-free medium. Exposure of LPDS-treated cells to apoA-I for 2 hours had no significant effect on the phosphorylation of pp18 (Fig 9B⇑). In contrast, increased phosphorylation of pp18 was still evident after serum-deprived, cholesterol-loaded cells were incubated with apoA-I for 2 hours. These results provide additional evidence that apoA-I elicits a sustained increase in phosphorylation of pp18 in growth-arrested, cholesterol-loaded fibroblasts but not in cells grown in the presence of LPDS.
We also compared the effects of the different preincubation conditions on the relative 32P labeling of pp80, pp27, and pp18 as normalized to five control phosphoproteins. With serum-deprived and cholesterol-loaded cells, 32P labeling of pp80 relative to control proteins was over twofold greater than that observed with cells grown in the presence of LPDS (Fig 9C⇑). Under the same conditions, 32P labeling of pp27 was increased slightly but insignificantly, and labeling of pp18 was significantly decreased (31%, P<.001). These results indicate that serum deprivation and cholesterol loading of fibroblasts leads to increased phosphorylation of pp80 and decreased phosphorylation of pp18 relative to control proteins but has no significant effect on phosphorylation of pp27.
The current study shows that acute treatment of growth-arrested, cholesterol-loaded fibroblasts with HDL stimulates phosphorylation of at least three acidic cellular phosphoproteins. The largest phosphoprotein (pp80) has an apparent Mr of 80 kD and a pI of 4.5, is partially associated with cell membranes, and was phosphorylated over twofold within 30 seconds of HDL treatment, which was sustained for at least 30 minutes. The second largest (pp27) is a membrane-associated phosphoprotein with an apparent Mr of 27 kD and a pI of 6.0 that underwent a transient 50% increase in phosphorylation after 10 minutes of HDL treatment. The smallest of the three (pp18) is a soluble 18-kD phosphoprotein that showed a 50% increase in phosphorylation after only 30 seconds of HDL treatment, which was sustained for at least 2 hours with cholesterol-loaded cells. This phosphoprotein was frequently resolved into two spots on 2D gels with pIs of 4.7 and 4.8, suggesting isoforms of the same protein. Increased phosphorylation of both spots was observed after HDL treatment. In addition, phosphorylation of all three phosphoproteins was increased in response to acute treatment of cells with PMA, but not with a cAMP analogue, suggesting they may be specific targets of a PKC signaling pathway. No other acidic or basic 32P-labeled proteins detectable on 2D gels showed consistent changes in phosphorylation after acute treatment of cholesterol-loaded fibroblasts with HDL.
Phosphoprotein 80 was phosphorylated on tyrosine and serine/threonine residues, whereas pp27 and pp18 were phosphorylated exclusively on serine/threonine. These results indicate that pp80 is capable of being phosphorylated by protein tyrosine kinases as well as PKC. Chronic treatment of cells with PMA, conditions that completely downregulate PKCα and PKCβ isotypes in fibroblasts, only partially suppressed the ability of HDL to stimulate pp80 phosphorylation, further indicating that protein kinases other than these two PKC isotypes are involved. In contrast, HDL-stimulated phosphorylation of both pp27 and pp18 was completely suppressed by chronic PMA treatment. This raises the possibility that PKCα and/or PKCβ mediate these phosphorylation events, although other PKC isotypes cannot be excluded.
The three phosphoproteins appeared to be regulated by different components associated with HDL. Treatment of fibroblasts with either HDL or trypsin-digested HDL stimulated phosphorylation of pp80 and pp27, whereas treatment with purified apoA-I, apoA-II, or apoE had either no effect or caused a slight decrease. This implies that the active component in HDL for stimulating phosphorylation of these two proteins is trypsin resistant and unlikely to be an apolipoprotein. It is possible that the stimulating agent is an HDL lipid, since lipids such as phosphatidic acid and lysophosphatidylcholine, which are present in HDL,19 have been shown to directly or indirectly activate PKC.38 39 In contrast to what was observed for pp80 and pp27, both HDL and purified apolipoproteins stimulated phosphorylation of pp18, while trypsinized HDL had no significant effect. Thus, it is likely that pp18 is being phosphorylated in response to the interaction of HDL apolipoproteins with cells. Because trypsinized HDL appeared to modestly stimulate pp18 phosphorylation in some but not all experiments, we cannot exclude the possibility that trypsin-resistant components in HDL may also have some effects on this protein.
The degree of phosphorylation of these three phosphoproteins differed in response to the preincubation conditions used in this study. When normalized to control phosphoproteins, 32P labeling of pp80 was greater in growth-arrested, cholesterol-loaded fibroblasts than in cholesterol-depleted cells grown in the presence of lipoprotein-deficient serum. In contrast, pp18 was 32P labeled to a lesser extent in cholesterol-loaded cells than in cholesterol-depleted cells, while labeling of pp27 appeared to be unaffected by these preincubation conditions. These results suggest that cholesterol loading of cells and/or depriving cells of serum components increase phosphorylation of pp80 but decrease phosphorylation of pp18.
These preincubation conditions also led to a different temporal pattern of stimulation of pp18 phosphorylation by apoA-I. As with growth-arrested and cholesterol-loaded cells, apoA-I stimulated phosphorylation of pp18 in cells preincubated with LPDS. In this case, however, the stimulation was significant only during the first 20 minutes. Between 30 and 120 minutes of exposure to apoA-I, phosphorylation of pp18 was not significantly above basal levels. Thus, in contrast to its effects on cholesterol-loaded cells, apoA-I caused only a transient increase in phosphorylation of pp18 in cholesterol-depleted cells. These results suggest that apoA-I–stimulated phosphorylation of pp18 in cholesterol-depleted fibroblasts is desensitized after 20 minutes, perhaps by the action of a phosphatase. Alternatively, stimulation of pp18 phosphorylation by apoA-I may have both an early (transient) and a late (sustained) phase, with the late phase being absent in cholesterol-depleted cells.
Several lines of evidence suggest that pp80 is the MARCKS protein, a well-characterized and ubiquitous substrate for PKC. First, the apparent Mr on SDS-PAGE and the pI correspond to those reported previously for MARCKS.33 Second, an antibody specific for MARCKS immunoprecipitated an 80-kD protein that was phosphorylated by HDL treatment of fibroblasts. Third, both pp80 and the immunoprecipitated protein contained phosphotyrosine residues. These findings also indicate that phosphorylation of the MARCKS protein is mediated by protein tyrosine kinases as well as PKC. Wu and Handwerger40 identified an acidic cytosolic 80-kD protein in human trophoblasts that was phosphorylated by treatment of cells with PMA, HDL, and purified apoA-I. The effect of apoA-I, however, was observed at concentrations 10-fold higher than that used in the current study (>100 μg/mL) and 20-fold higher than that needed to maximally stimulate cholesterol efflux from fibroblasts and macrophages,27 35 suggesting that it was related to lower-affinity responses to apoA-I, such as promoting lactogen release. Although they have similar properties, it is unknown whether the 80-kD phosphoproteins in trophoblasts and fibroblasts are the same molecule.
The identities of pp27 and pp18 are unknown. The Mr of pp27 corresponds to a protein previously reported by Darbin et al25 to be phosphorylated in response to treatment of vascular endothelial cells with HDL. Tazi et al41 has described three HDL-responsive phosphoproteins in human adenocarcinoma cells of 28, 24, and 20 kD. It is possible that at least one of these may be the same protein as pp27 or pp18. The 20-kD protein, however, appears to be phosphorylated in adenocarcinoma cells by a PMA-insensitive kinase. We investigated the possibility that pp18 is Op18, an 18-kD oncoprotein involved in cell-cycle control,42 and found that an Op18 antibody did not cross-react with pp18.
The physiological relevance of these HDL-mediated phosphorylation events remains to be determined. The different patterns of phosphorylation suggest that they may reflect multiple and divergent signaling responses to the interaction of HDL with cells. We showed previously that modification of HDL by treatment with tetranitromethane abolished the ability of HDL to stimulate translocation of PKC from the cytosol to the plasma membrane,20 a process frequently but not exclusively associated with PKC activation.43 More recently, we found that treatment of fibroblasts with HDL, trypsinized HDL, and tetranitromethane-modified HDL, but not with purified apolipoproteins, stimulates mitogen-activated protein kinase, which was partially dependent on PKC activation.44 These and the current findings suggest that modified HDL particles activate PKC by mechanisms different from those elicited by HDL apolipoproteins. This may reflect involvement of different isoforms of PKC or different subcellular sites of enzyme translocation, as has been described for other agonists.45 Divergent signaling response to HDL is further supported by results showing that HDL components activate different phospholipases in cholesterol-loaded fibroblasts.16 17 Moreover, HDL and its apolipoproteins have been shown to stimulate a variety of cellular processes associated with activation of PKC, including efflux of cholesterol and other lipids,12 14 15 16 17 18 19 20 21 22 excretion of proteins,23 24 and mitogenesis.25
Since both phospholipids and apolipoproteins in HDL can promote cholesterol efflux from cells by independent mechanisms,46 it is possible that any one or more of these phosphoproteins is involved in cellular lipid transport pathways. The interaction of HDL apolipoproteins with cholesterol-loaded fibroblasts stimulates translocation of intracellular sterol to the plasma membrane by a PKC-mediated signaling pathway.20 Since both HDL and purified apolipoproteins, but not trypsinized HDL, stimulate sterol translocation,10 20 increased phosphorylation of pp80 and pp27 can be excluded as being involved in this stimulatory process. Three lines of evidence, however, are consistent with the possibility that pp18 may be a component of the signaling or transport pathways that promote efflux of excess cholesterol from cells. First, with cholesterol-loaded cells, HDL apolipoproteins stimulate sustained phosphorylation of pp18, translocation of intracellular sterols to the plasma membrane,10 15 16 20 and cholesterol and phospholipid efflux.27 34 35 36 Second, with cholesterol-depleted fibroblasts, apoA-I stimulates pp18 phosphorylation only transiently and has much less ability to remove cellular lipids compared with its effects on cholesterol-loaded cells.36 37 Third, cholesterol loading of fibroblasts selectively decreases phosphorylation of pp18 and inhibits translocation of intracellular sterols to the plasma membrane.10 Additional structural and functional characterization of pp18 will be required to establish its role in lipid transport. Regardless of their functions, the three phosphoproteins described here may be useful markers for the divergent signaling processes induced by the interaction of HDL with cells.
Selected Abbreviations and Acronyms
|DMEM||=||Dulbecco’s modified Eagle’s medium|
|MARCKS||=||myristoylated alanine-rich C kinase substrate|
|PAGE||=||polyacrylamide gel electrophoresis|
|PKC||=||protein kinase C|
|PMA||=||phorbol 12-myristate 13-acetate|
This work was supported by National Institutes of Health grants HL-18645 and DK02456. Dr Deeg was a recipient of a Pfizer postdoctoral fellowship. Bovine brain MARCKS antibody was a gift of Angus C. Nairn, of The Rockefeller University.
Gordon DJ, Knoke J, Probstfield JL, Superko R, Tyroler HA. High-density lipoprotein cholesterol and coronary artery disease in hypercholesterolemic men: the Lipid Research Clinics Coronary Primary Prevention Trial. Circulation. 1986;73:1217-1225.
Glomset JA. The plasma lecithins: cholesterol acyltransferase reaction. J Lipid Res. 1968;9:155-167.
Graham DL, Oram JF. Identification and characterization of a high density lipoprotein-binding protein in cell membrane by ligand blotting. J Biol Chem. 1987;262:7439-7442.
McKnight GL, Reasoner J, Gilbert T, Sundquist KO, Hokland BM, McKernan PA, Champagne J, Johnson CJ, Bailey MC, Holly R, O’Hara PJ, Oram JF. Cloning and expression of a cellular high density lipoprotein-binding protein that is up-regulated by cholesterol loading of cells. J Biol Chem. 1992;267:12131-12141.
Tozuka M, Fidge NH. Purification and characterization of two high-density-lipoprotein binding proteins from rat and human liver. Biochem J. 1989;261:239-244.
Bond HM, Morrone G, Venuta S, Howell KE. Characterization and purification of proteins which bind high-density lipoprotein. Biochem J. 1991;279:633-641.
Barbaras R, Puchois P, Fruchart J-C, Pradines-Figuëres A, Ailhaud G. Purification of an apolipoprotein A binding protein from mouse adipose cells. Biochem J. 1990;269:767-773.
Fidge NH, Nestel PJ. Identification of apolipoproteins involved in the interaction of human high density lipoproteins with receptors on cultured cells. J Biol Chem. 1985;260:3570-3575.
Oram JF, Mendez AJ, Slotte JP, Johnson TF. High-density lipoprotein apolipoproteins mediate removal of sterol from intracellular pools but not from plasma membranes of cholesterol-loaded fibroblasts. Arterioscler Thromb. 1991;11:403-414.
Rogler G, Trumbach B, Klima B, Lackner KJ, Schmitz G. HDL-mediated efflux of intracellular cholesterol is impaired in fibroblasts from Tangier disease patients. Arterioscler Thromb Vasc Biol. 1995;15:683-690.
Walter M, Reinecke H, Nofer J-R, Seedorf U, Assmann G. HDL3 stimulates multiple signaling pathways in human skin fibroblasts. Arterioscler Thromb Vasc Biol. 1995;15:1975-1986.
Pörn MI, Ackerman KEO, Slotte JP. High-density lipoprotein induces a rapid and transient release of Ca2+ in cultured fibroblasts. Biochem J. 1991;279:29-33.
Mendez AJ, Oram JF, Bierman EL. Protein kinase C as a mediator of high density lipoprotein receptor-dependent efflux of intracellular cholesterol. J Biol Chem. 1991;266:10104-10111.
Voyno-Yasenetskaya TA, Dobbs LG, Erickson SK, Hamilton RL. Low density lipoprotein and high density lipoprotein mediated signal transduction and exocytosis in alveolar type II cells. Proc Natl Acad Sci U S A. 1993;90:4256-4260.
Hu RM, Chuang MY, Prins B, Kashyap ML, Frank HJ, Pedram A, Levin ER. High density lipoproteins stimulate production and secretion of endothelin-1 from cultured bovine aortic endothelial cells. J Clin Invest. 1994;93:1056-1062.
Darbon JM, Tournier JF, Tauber JP, Bayard F. Possible role of protein phosphorylation in the mitogenic effect of high density lipoproteins on cultured vascular endothelial cells. J Biol Chem. 1986;261:8002-8008.
Biesbroeck RC, Oram JF, Albers JJ, Bierman EL. Specific high affinity binding of HDL to cultured human skin fibroblasts and arterial smooth muscle cells. J Clin Invest. 1983;61:525-539.
Mendez AJ, Anantharamaiah GM, Segrest JP, Oram JF. Synthetic amphipathic helical peptides that mimic apolipoprotein A-I in clearing cellular cholesterol. J Clin Invest. 1994;94:1698-1705.
O’Farrell PH. High resolution two-dimensional electrophoresis of proteins. J Biol Chem. 1975;250:4007-4021.
Halsey DL, Girard PR, Kuo JF, Blackshear PJ. Protein kinase C in fibroblasts. J Biol Chem. 1987;262:2234-2243.
Blackshear PJ. The MARCKS family of cellular protein kinase C substrates. J Biol Chem. 1993;268:1501-1504.
Francis GA, Knopp RH, Oram JF. Defective removal of cellular cholesterol and phospholipids by apolipoprotein A-I in Tangier disease. J Cell Invest. 1995;96:78-87.
Hara H, Yokoyama S. Interaction of free apolipoproteins with macrophages. J Biol Chem. 1991;266:3080-3086.
Bielicki JK, Johnson WJ, Weinberg RB, Glick JM, Rothblat GH. Efflux of lipid from fibroblasts to apolipoproteins: dependence on elevated levels of cellular unesterified cholesterol. J Lipid Res. 1992;33:1699-1709.
Oram JF, Brinton EA, Bierman EL. Regulation of HDL receptor activity in cultured human skin fibroblasts and human arterial smooth muscle cells. J Clin Invest. 1983;72:1611-1621.
Bell RM, Burns DJ. Lipid activation of protein kinase C. J Biol Chem. 1991;266:4661-4664.
Kugiyama K, Ohgushi M, Sugiyama S, Murohara T, Fukunaga K, Miyamoto E, Yasue H. Lysophatidylcholine inhibits surface receptor-mediated intracellular signals in endothelial cells by a pathway involving protein kinase C activation. Circ Res. 1992;71:1422-1428.
Tazi KA, Bonnafous M, Faure G, Soula G, LeGaillard F. Involvement of the Ca(2+)-dependent phosphorylation of a 20 kDa protein in the proliferative effect of high density lipoproteins (subclass 3) on the adenocarcinoma cell line A549. Biochem J. 1995;307:557-561.
Deeg MA, Bowen RF, Oram JF, Bierman EL. High density lipoproteins stimulate mitogen-activated protein kinases in human skin fibroblasts. Arterioscler Thromb Vasc Biol. 1997;17:1667-1674.
Haller H, Ziegler W, Lindschau C, Luft FC. Endothelial cell tyrosine kinase receptor and G protein–coupled receptor activation involves distinct protein kinase C isoforms. Arterioscler Thromb Vasc Biol. 1996;16:678-686.
Oram JF, Yokoyama S. Apolipoprotein-mediated removal of cellular cholesterol and phospholipids. J Lipid Res. 1996;37:2473-2491.