ANALYTICAL

BIOCHEMISTRY

196,122-128

(1991)

The Labeling of Lipoproteins for Studies of Cellular Binding with a Fluorescent Lipophilic Dye’ James

P. Corsetti,2

Charles

H. Weidner,*

Joanne

Cianci,

and Charles

E. Sparks

Department of Pathology and Laboratory Medicine, University of RochesterMedical Center, 601 Elmwood Avenue, Box 608, Rochester,New York 14642; and *Department of Chemistry, University of Rochester,Rochester,New York 14627

Received

October

18,199O

N,N-dipentadecylaminostyrylpyridinium iodide is a dye that is approximately loo-fold more intensely fluorescent in a lipid than aqueous environment. This observation suggests its potential as a fluorescence stain for lipoproteins. This work reports the staining of LDL with this dye for use in studies of cellular binding. The staining procedure is simple, resulting in stable attachment of the dye as determined by transfer experiments, physical properties essentially identical to native LDL as demonstrated by virtually identical electrophoretic mobility, and consistent results in studies of cellular binding using flow cytometry. Increased signal to noise ratio over other dyes used for lipoprotein staining including the widely used Dil(3,3’-dioctadecylindocarbocyanine iodide) allows determinations of greater sensitivity and precision to be made. This is demonstrated by the flow cytometric determination of the 4°C binding curve of LDL with freshly isolated human peripheral blood lymphocytes (i.e., cells not LDL receptor upregulated). Mediation of binding by the LDL receptor is demonstrated by correspondence between the LDL receptor dissociation constant derived from this work and literature values; increased specific binding in lymphocytes cultured in lipoprotein-deficient media to up-regulate the LDL receptor; and decreased specific binding in lymphocytes cultured in the presence of 25-hydroxy cholesterol for 48 h to suppress the LDL receptor. 0 1991

Academic

Prese,

Inc.

Studies with fluorescently labeled lipoproteins have provided important information on the physical properties of lipoproteins and the cellular binding and uptake

i This work Grant-In-Aid ‘To whom dressed.

was supported in part by American Heart Association 889793 and PHSHL29837. correspondence and reprint requests should be ad-

of lipoproteins especially in studies related to atherogenesis. Lipoprotein fluorescence labels include sterol, steryl ester, and phospholipid fluorescent analogues; covalent fluorochrome-apolipoprotein conjugates; and lipophilic dyes (1). One of the most useful of all has proven to be the lipophilic, carbocyanine dye; 3,3’dioctadecylindocarbocyanine iodide (Di1)3 (2,3). Carbocyanine dyes have been extensively used as probes of membrane potential (4). Lipoproteins are easily labeled with Dil with preservation of essentially unalteredphysical properties and firm binding of probe to lipoprotein (2). Indeed, several recent reports (5-10) demonstrate studies with Dil-lipoproteins using flow cytometry for the quantitation of cellular binding and uptake providing results previously attainable only with radiolabeling techniques. An additional significant advantage with flow cytometry is that it provides information on a cellby-cell basis allowing investigations of heterogeneity of cellular lipoprotein metabolism. The use of Dil-lipoproteins, however, is somewhat limited to cells with large numbers of receptors especially in binding studies because of limitations in signal to noise (6,7). The situation is especially significant in flow cytometry, where for any particular cell, signal sampling occurs just once and only for an instant. In view of the demonstrated utility of Dil-lipoproteins, extension of the technique to cells with relatively low levels of receptors, and increasing signal to noise ratios in general, is desirable. Staining with higher levels of Dil is unacceptable as lipoprotein physical properties begin to change significantly (2). A search was undertaken to identify a more intensely fluorescent dye for lipoproteins with the simultaneous preservation of ease of staining, minimal alteration in

3 Abbreviations used: Dil, 3,3’-dioctadecylindocarbocyanine di-15-ASP, N,N-dipentadecylaminostyrylpyridinium physiologic saline solution; PBS, phosphate-buffered

122

Copyright All

rights

of

iodide; saline.

iodide; PSS,

ooO3-2697/91$3.66 0 1991 by Academic Press, Inc. reproduction in any form reserved.

FLUORESCENT

LABELING

physical properties, and firm binding. This work reports the synthesis of N,N-dipentadecylaminostyrylpyridinium iodide (di-15ASP) and its demonstration as a lipoprotein stain. It is an aminostyrylpyridinium salt, a class of fluorescent dyes also originally developed as probes of membrane potential (11,12). The dye is especially well suited for flow cytometry because of conveniently available excitation wavelength (488 nm), and the intense fluorescence of the labeled lipoprotein. MATERIALS

AND

METHODS

Synthesis of N,N-Dipentadecylaminopyridinium

Iodide

OF

LIPOPROTEINS

123

treated with piperidine (0.25 ml) at room temperature. The resulting mixture was heated to 110°C in a sealed tube for 35 min, and then cooled to 0°C. The precipitate was filtered and recrystallized from methanol to yield 0.264 g (70%) of pure dye as a red solid, mp 204-206°C. ‘H NMR (CD&, 300 MHz) 6 8.78-8.70 (m, 2H), 7.857.73 (d, J = 6.0 Hz, 2H), 7.60 (d, J = 15.8 Hz, lH), 7.50 (d, J = 8.8 Hz, 2H), 6.81 (d, J = 15.8 Hz, lH), 6.64 (d, J = 8.8 Hz, 2H), 4.41 (s, 3H), 3.34 (t, J = 7.4 Hz, 4H), 1.70-1.20 (m, 52H), 0.89 (t, J = 6.8 Hz, 6H); uv-vis (ethanol) X,, 495 nm, E= 48700. Lipoprotein Isolation and Staining

The required substituted p-N,N-dipentadecylaminoHuman LDL (density 1.019-1.063 g/ml) and HDL benzaldehyde was prepared by the alkylation of aniline (density 1.085-1.215 g/ml) were isolated from the serum with 1-bromopentadecane according to the procedure of of normal healthy donors by sequential density gradient Foster and Hammick (13) followed by Vilsmeier formyultracentrifugation (17). The LDL was filter sterilized lation (14). Condensation of the aldehyde with through a 0.2-pm filter and dialyzed extensively against 4,Ndimethylpyridinium iodide, following the general 0.9% saline and 0.01% EDTA. After dialysis, the LDL procedure of Brooker et al. (15), then afforded the de- was again filter sterilized through a 0.2-pm filter and sired styryl dye. stored in sterile polystyrene tubes for periods less than 5 days at 4°C until ready for use. N,N-Dipentadecylaminobenzene. Freshly distilled For the preparation of di-15-ASP-LDL, a solution of aniline (0.722 g, 7.76 mmol) and 1-bromopentadecane (4.52 g, 15.50 mmol) were treated with potassium hy- LDL (0.5 mg LDL total protein/ml) and BSA (0.12%) in droxide (0.870 g, 15.50 mmol) and the mixture was PSS (137 mM NaCl, 2.7 mM KCL, 1.0 mM MgCl,, 3.3 heated to 135°C for 13 h. After cooling, the mixture was mM NaH,PO,, 5.6 mM glucose, 20 mM Hepes, pH 7.4 poured into water (25 ml) and extracted with ethyl ether with 1 N NaOH) was prepared to which was added di(2 X 25 ml). The combined ethereal extracts were dried 15-ASP to a final concentration of 75 pM from a 5 mM over magnesium sulfate, concentrated in vacua, and pu- stock solution in DMSO. After mixing well by gentle rified by flash column chromatography (16) (SiO,; hex- inversion, the mixture was incubated for 3 h at room anes elution, followed by 1% ethyl acetate in hexanes) to temperature in the dark. The density of the mixture was then adjusted to 1.063 with NaBr and ultracentrifuged afford 1.44g (36%) of N,N-dipentadecylaminobenzene as a waxy solid, mp 42-43°C. ‘H NMR (CDCl,, 300 for 20 h at 14°C and 50,000 rpm on a Beckman 80Ti rotor. The stained LDL was recovered by aspiration of MHz) 6 7.21 (dd, J= 8.4,7.3 Hz, 2H), 6.65-6.60 (m, 3H), layer which was then passed through a 0.23.25 (t, J = 7.6 Hz, 4H), 1.65-1.55 (m, 4H), 1.40-1.20 (m, the surface pm filter and dialyzed extensively against 0.9% saline 48H), 0.89 (t, J = 6.9 Hz, 6H); m+/e 514; TLC (Merck F254-60 SiO,, 0.25 mm, 1.25% ethyl acetate in hexanes) and 0.01% EDTA. To further separate the stained LDL from free dye, the preparation was then sequentially Rf = 0.42. passed through two heparin-agarose affinity columns 4-N,N-Dipentadecylaminobenzaldehyde. Phospho(LDL-Direct Cholesterol Audit System, Isolabs Inc., rous oxychloride (0.209 g, 1.36 mmol) was added to 0.5 Akron, OH) with retention of the second 500 ~1 of the p ml of DMF at 0°C. N,N-Dipentadecylaminobenzene eluate in each case (the free dye is retained in the col(0.700 g, 1.36 mmol) was added to this all at once and the mixture was heated to 100°C for 1 h. Ice (1 g) and 1 N umn). The stained LDL was stored in sterile polystyrene tubes at 4°C and again filter sterilized through a sodium acetate (1 ml) were added to the mixture, which was allowed to stir for an additional hour at room tem- 0.2~pm filter just before use. Electrophoresis of stained and unstained LDL preparations was performed on agaperature. The precipitated solid was filtered and recrystallized from ethanol to afford 0.715 g (97%) of the alde- rose (Corning Universal Agarose Plates, Palo Alto, CA) hyde as a tan solid, mp 52-52.5”C. ‘H NMR (CDCl,, 300 in barbital buffer at 90 V at room temperature for 35 MHz) 6 9.71 (s, lH), 7.70 (d, J = 8.9 Hz, 2H), 6.64 (d, J min and stained with Fat Red 7B. Lipoprotein total pro= 8.9 Hz, 2H), 3.34 (t, J = 7.6 Hz, 4H), 1.70-1.20 (m, tein determinations were by the method of Markwell et al. (18). 52H), 0.89 (t, J = 6.9 Hz, 6H); TLC (2:l hexanes:ethyl acetate) R, = 0.59. Lymphocyte Preparation and Lipoprotein Binding 1-(p-N,N-Dipentadecylaminophenyl)-2-(4-(N-methylpyridinium) ethylene iodide. A suspension of 4,NdiMononuclear cells were separated from heparinized methylpyridinium iodide (0.118 g, 0.50 mmol) and the blood by differential sedimentation centrifugation with aldehyde (0.271 g, 0.5 mmol) in 2.5 ml of methanol was Lymphoprep (Accurate Chemical and Scientific Corpo-

124

CORSETTI

ration, Westbury, NY). The interface layer was collected and washed two times, the first with PBS (NaCl 137 mM, KH,PO, 1.47 mM, Na,HPO, 8.09 mM, KC1 2.68 InM; pH 7.4), and the second with RPMI/0.2% BSA (Sigma, St. Louis, MO) with centrifugations at 400g for 10 min at room temperature. The cells were then resuspended in RPMI/0.5% BSA (Miles Inc., Kankakee, IL). Monocytes were eliminated by adhesion to plastic as follows: the mononuclear cell preparation was incubated in a loo-mm polystyrene tissue culture dish (Corning Glass Works, Corning, NY) for 1 h at 37°C in a 95% 0,, 5% CO,. The cells were kept in culture in RPMI/0.5% BSA at 37°C in 95% O,, 5% CO, for the indicated periods of time. For experiments with 25-hydroxy cholesterol, the cells were kept in culture for 48 h with 25-hydroxy cholesterol (2 pg/ml) (Steraloids Inc., Wilton, NH) plus cholesterol (10 pg/ml) (Steraloids Inc., Wilton, NH). For binding experiments, lymphocytes were diluted to a density of 2 X lo5 cells/ml with RPMI/0.2% BSA and then di-15-ASP-LDL was added. For binding experiments with blocking, unstained LDL was added to RPMI/0.2% BSA at a final concentration of 400 pg/ml total protein. Cells were added to this to a final density of 2 X lo5 cells/ml, and then di-15-ASP-LDL was added. The cells were then incubated for 2 h at 4°C. After incubation, the cells were washed twice with 1 ml RPMI/ 0.2% BSA and resuspended in RPMIIO.B% BSA to a final density of 1 X lo6 cells/ml. Fluorescence

Determinations

Fluorescence spectra and determinations were all performed on a Perkin-Elmer 650-10s spectrofluorometer (Perkin-Elmer, Norwalk, CT). Excitation spectra were monitored at 620 nm and emission spectra excited at 488 nm with all determinations at room temperature. Spectra were uncorrected for instrumental response. Analytical determinations were all monitored at 560 nm.

ET

AL.

‘-q&qy;;y;; 2 14

FIG. 1. The structure iodide (di-X-ASP).

Dye Transfer

of NJ-dipentadecylaminostyrylpyridinium

Experiments

Dye transfer from stained LDL to native HDL was assessed by mixing experiments in which stained LDL (50 pg/ml total protein) was incubated with native HDL in final concentrations ranging from 0 to 400 gg/ml total protein at 4°C for 2 h. This gave a maximal molar HDL/ LDL ratio of 40/l. After incubation, the HDL and LDL were separated by affinity chromatography on heparinagarose columns and the fluorescence of the eluates was determined and corrected for dilutional effects. Flow Cytometry The cells were analyzed with an Epics C flow cytometer (Coulter Electronics, Hialeah, FL) equipped with an argon-ion laser. Excitation was 500 mW at 488 nm. Optimum signal to noise for collection of the di-15-ASP fluorescence was obtained using two 550-nm long-pass absorption filters (3802052, Coulter Electronics). Intact single lymphocytes were gated using forward and rightangle light scatter. All fluorescence data collection was in linear mode. Photomultiplier tube voltage and gain were set to give a mean fluorescence channel number of approximately 120 for the most intensely fluorescent specimen of the day which resulted in greater than 90% of cells on scale. When necessary, calibration was achieved by the use of microbead standards. Where significant, autofluorescence was handled by subtracting the mean fluorescence channel number of an autofluorescence control of unstained cells from the stained cells. RESULTS

Microscopy Electron microscopy of negatively stained di-15ASP-LDL with phosphotungstate (19) was performed on an Hitachi HS-8 electron microscope (Hitachi, Tokyo, Japan). Fluorescence microscopy of primary rat hepatocytes on cover slips was performed with a Nikon Labophot microscope (Nippon Kogaku, Garden City, NY) with episcopic fluorescence attachment using a 495-nm excitation filter and a 515-nm emission filter. Primary rat hepatocytes were prepared as previously described (20) and kept in culture overnight. They were then incubated with 50 pglml of di-15-ASP-LDL for 2 h at 4”C, washed, and kept at 4°C until microscopic examination (cover slip inverted on a slide with focusing at the cell-cover slip interface).

The amphiphilic nature of di-15-ASP is clearly demonstrated by its chemical structure shown in Fig. 1. Figure 2 demonstrates the room temperature excitation (monitored at 620 nm) and emission (excited at 488 nm) spectra of 5 nM di-15-ASP-LDL in PSS at room temperature. All spectra were uncorrected for instrumental response. The excitation spectrum demonstrates significant absorption at 488 nm corresponding to an intense line of the argon-ion laser commonly used as an excitation source in flow cytometry. Lipoprotein electrophoresis revealed virtually identical mobility of di-15-ASP-LDL and native LDL. Figure 3 is a negatively stained electron photomicrograph of di-15-ASP-LDL demonstrating discrete LDL particles similar in size and shape to native LDL, and Fig. 4 is a

FLUORESCENT

WAVELENGTH

LABELING

(nm)

FIG.

2. The fluorescence excitation spectrum (dashed line) and emission spectrum (solid line) of 5 nM di-15-ASP-LDL in PSS at room temperature. The excitation was monitored at 620 nm, and the emission excited at 488 nm. The spectra were uncorrected for instrumental response.

fluorescence photomicrograph of rat hepatocytes demonstrating grainy, circumferential plasma-membrane staining of di-15-ASP-LDL. Figure 5 demonstrates the total and nonspecific binding of di-15-ASP-LDL by freshly isolated human peripheral blood lymphocytes after incubation for 2 h at 4°C. Nonspecific binding was determined from blocking experiments with unstained LDL in which cells and then di-15-ASP-LDL were added to media containing the unstained LDL (400 pg/ml total protein). The incubations were performed immediately after lymphocyte isolation providing minimal opportunity for up-regulation of the LDL receptor. Binding was determined as the mean fluorescence channel number from flow cytometric analysis in linear mode. This is necessarily a relative determination as calibration to some measure of absolute mass has not yet been effected for the stained LDL. Nonspecific binding was approximately 15% of total at di-15-ASP-LDL concentration of 50 pg/ml. Also in Fig. 5 is the specific binding curve which was obtained by subtracting the nonspecific from the total. The dissociation constant, Kdof the LDL receptor from the data of Fig. 5 is approximately 19 bg/ml which was estimated as the concentration at half-maximal binding. Similar experiments were also performed on human erythrocytes which are not expected to express the LDL receptor, and thus served as a negative control for specific cellular LDL binding. The results are shown in Fig. 6, which demonstrates erythrocyte total and nonspecific binding; also reproduced are the lymphocyte total binding data of Fig. 5 for comparison. As expected, erythrocyte total binding and nonspecific binding were virtually identical and a small fraction of lymphocyte total binding. Figure 7 demonstrates the firm binding of di-15-ASP to LDL as determined by mixing experiments in which di-15-ASP-LDL (50 pg/ml total protein) was incubated in 0.9% saline at 4°C for 2 h with a range of concentrations of unlabeled HDL. The lipoproteins were sepa-

OF LIPOPROTEINS

125

rated using heparin-agarose affinity chromatography and fluorescences of LDL and HDL fractions determined. Figure 7 is a plot of the fluorescence of the LDL and HDL fractions as a function of HDL concentration and clearly demonstrates firm binding of the dye to LDL. Figure 8 demonstrates suppression of up-regulation of the LDL receptor by 25-hydroxy cholesterol and upregulation by culture in media free of lipoproteins. Figure 8A demonstrates the results of binding experiments with di-15-ASP-LDL (25 pg/ml) on lymphocytes in culture for 48 h with and without (control) 2 pg/ml of 25hydroxy cholesterol and 10 pg/ml of cholesterol. The cells demonstrated specific binding approximately 60% of that of control. Figure 8B demonstrates the results of binding experiments with di-15-ASP-LDL (100 *g/ml) on freshly isolated lymphocytes (control) and lymphocytes in culture for 22 h in media free of lipoproteins (0.5% BSA). The cells demonstrated a twofold increase in specific binding over that of control. It should be noted that the results of Figs. 8A and 8B are not directly comparable as different conditions were used for each of the two experiments. DISCUSSION The structure of di-15-ASP as shown in Fig. 1 clearly demonstrates the amphiphilic nature of this molecule. As such, the binding of the dye to lipoprotein is most likely via dissolution into the phospholipid monolayer surrounding the lipoprotein particle. Phosphatidylcholine is the predominant phospholipid of lipoproteins; and hence, the dye probably acts as a phosphatidylcholine analogue in the phospholipid monolayer. The amphiphilic nature of the di-l5-ASP also has implications for the purification of stained lipoproteins as the dye probably forms micelles in an aqueous environment.

FIG.

3. Phosphotungstate negatively stained electron photomicrograph (~104,500) of di-15ASP-LDL demonstrating discrete LDL particles similar to native LDL in size and shape.

126

4. Fluorescence FIG. kdm d of di-15ASP-LDL cell-cover slip interface).

CORSETTI

ET

photomicrograph (Xl,000 with 495 nm excitation at 4°C for 2 h and kept at 4°C until microscopic

Preliminary experiments in our laboratory (data not shown) show a discontinuity in light scatter properties as a function of dye concentration which is presumptive evidence for a change in the intermolecular organization of the dye. This is further substantiated by the fact that stained lipoprotein preparations cannot be separated from the free dye by dialysis alone. This is unexpected in view of the small size of the dye molecule relative to a lipoprotein particle. Further separation is effected by affinity chromatography on heparin-agarose. Separation of free and dissociable dye from stained lipoprotein is critical as free dye would directly stain cells and other lipoproteins present in the system. The significantly higher fluorescence intensity of diX-ASP in a lipid than aqueous environment plays a major role in the utility of the dye as a lipoprotein stain.

AL.

and 515 nm emission examination (cover

filters) of rat hepatocytes slip inverted on a slide

incubated wit1 150 with focusing at the

This enhancement most likely is a result of constraints on the geometrical conformations available to the molecule in the phospholipid monolayer favoring those with high emission quantum yields. The superiority of di-15 ASP for LDL labeling over the currently, widely used Dil is demonstrated by the ability to obtain the specific binding curve of Fig. 5. This is the binding curve at 4°C of di-15ASP-LDL by freshly isolated human peripheral blood lymphocytes. Since these cells are not cultured in lipoprotein deficient media there is little chance for upregulation of the LDL receptor. The curve was determined using flow cytometry and is a result that has not been possible with Dil-labeled LDL (6). The experiment with di-15-ASP-LDL results in flow cytometric histograms with clearly delineated peaks even at low stained LDL concentrations. This is in contrast to Dil-LDL

FLUORESCENT

LABELING

OF

127

LIPOPROTEINS

-00

I 100

I

I 200

HDL LDL

o-0-0 8

I 300

1

I 400

1 ! iCIO

(p/ml)

@g/ml)

FIG. 5. The binding of di-15-ASP-LDL by freshly isolated human peripheral blood lymphocytes as a function of LDL concentration. Binding was at 4°C for 2 h and determined as the mean fluorescence channel number in linear mode from flow cytometric analysis. There is total binding (circles) and nonspecific binding (squares) as determined in the presence of unlabeled LDL (400 rg/ml total protein). Specific binding (triangles) was determined by subtraction of nonspecific binding from total binding. The results are from a typical experiment on a single individual. Precision studies for such cases result in standard deviations of binding (mean fluorescence channel number) on the order of 2 channel numbers.

where even amplification of the lymphocyte fluorescence intensity by monitoring uptake at 37°C instead of binding at 4°C still results in inability to clearly determine peak position, so that a 75th percentile technique (6) is used for data analysis. Reported values of the Kd of the LDL receptor range between 2.4 and 15 /*g/ml for human fibroblasts, lymphocytes, and rat hepatocytes (7,21-25) with the higher values more representative of determinations at 37°C

FIG. 7. Transfer of d&ASP from labeled LDL to unlabeled HDL as a function of HDL concentration. Aliquots of native HDL at the indicated concentrations were incubated with stained LDL (50 pg/ml total protein) for 2 h at 4°C in 0.9% NaCl. The LDL and HDL were then separated using heparin-agarose affinity chromatography and the fluorescences determined: LDL fraction (circles) and HDL fraction (squares). The fluorescences were corrected for dilutional factors.

and the lower values more representative of determinations at 4°C. The value of the Kd at 4’C of the LDL receptor, estimated as the concentration at half-maximal binding from the data of Fig. 5, is 19 pg/ml. Although this value is somewhat high, it must be viewed as semiquantitative in that more definitive Scatchard analysis cannot be performed since the flow cytometric experiment does not provide the requisite unbound stained LDL concentration. With this level of precision, our value agrees well with the previously reported values. Alternatively, if the value is accurate, it more closely resembles previous results reported at 37°C. This may be a result of dye-related attenuation of LDL phase-related changes in apolipoprotein B conforma-

6 [1 LDL @g/ml) FIG. 6. The binding of di-15-ASP-LDL by freshly isolated human erythrocytes as a function of LDL concentration. Binding was at 4°C for 2 h and determined as the mean fluorescence channel number in linear mode from flow cytometric analysis. There is total binding (open circles), nonspecific binding (filled circles) as determined in the presence of unlabeled LDL (400 pg/ml total protein), and typical total binding (squares) of freshly isolated lymphocytes for comparison. The results are from a typical experiment on a single individual. Precision studies for such cases result in standard deviations of binding (mean fluorescence channel number) on the order of 2 channel numbers.

I 2

FIG. 8. Regulation of the LDL receptor on human lymphocytes. (A) Inhibition of up-regulation of the LDL receptor by 25 hydroxy cholesterol as demonstrated by binding with di-15-ASP-LDL (25 pg/ ml) at 4°C for 2 h. Column 1, control of lymphocytes in culture for 48 h. Column 2, lymphocytes in culture for 48 h in the presence of 25 hydroxy cholesterol (2 pg/ml) and cholesterol (10 rig/ml). (B) Upregulation of the LDL receptor by culture in media free of lipoproteins for 22 h as demonstrated by binding with di-15-ASP-LDL (100 pglml) at 4°C for 2 h. Column 1, control of freshly isolated lymphocytes. Column 2, lymphocytes cultured in media free of lipoproteins (0.5% BSA) for 22 h. The data were individually normalized to the respective controls and represent the means of three separate experiments with cells from different individuals.

128

CORSETTI

tion which may underlie the decrease of the Kd with decrease in temperature for native LDL. The minimal alteration in physical properties of LDL upon staining with di-15-ASP is demonstrated by (i) preparations with electrophoretic mobility virtually identical to native LDL, (ii) the preservation of discrete LDL particles similar to native LDL in size and shape as shown by the electron photomicrograph of the stained LDL of Fig. 3, and (iii) the expected grainy, circumferential plasma membrane staining with di-15-ASP-LDL shown in Fig. 4 consistent with LDL receptor binding. The firm binding of di-15-ASP to LDL at the conditions of the binding experiments is demonstrated by the data of Fig. 7. The HDL curve shows that even at the maximum HDL concentration, there is at most only a 9% transfer of dye. The LDL curve is consistent with this although absolute quantitation is not possible because of experimental error fluctuations in the data. It should be noted that for the binding experiments in this work, dye transfer is probably negligible. This is because the experiments were not done in the presence of other lipoproteins. Hence, transfer of dye could only occur to cell membranes which was insignificant as demonstrated by the human erythrocyte binding experiments which showed total binding virtually identical to nonspecific binding and a small fraction of typical lymphocyte total binding. The use of the dye for lipoprotein staining for quantitative studies of cellular binding mandates the evaluation of spectral characteristics and dye transfer in the individual system. Four lines of evidence support mediation of binding of di-15-ASP-LDL via the LDL receptor. First, the value of the Kd (19 pg/ml) of the LDL receptor is in good agreement with previously reported values. Second, the control binding experiments with human erythrocytes of Fig. 6 demonstrate total binding virtually identical to nonspecific binding. Third, the binding experiments of Fig. 8A demonstrate less specific binding of stained LDL in the cells exposed to 25-hydroxy cholesterol for 48 h than in the 48-h control cells. This is consistent with the well-known inhibitory effects of oxysterols on the biosynthesis of cholesterol as well as the expression of the LDL receptor (26-27). Fourth, the binding experiments of Fig. 8B demonstrate increased specific binding of stained LDL in lymphocytes cultured for 24 h in lipoprotein-deficient media than in the control of freshly isolated lymphocytes. This is strongly suggestive of the expected up-regulation of the LDL receptor in cells cultured in the absence of lipoproteins (28). In summary, data have been presented demonstrating the lipophilic fluorescent dye, di-15-ASP, to be an excellent stain for LDL for cellular binding studies. The stained LDL has physical properties close to native LDL; there is insignificant loss of staining; binding is via the LDL receptor; and the preparations allow deter-

ET

AL.

minations of greater made than with Dil.

sensitivity

and precision

to be

REFERENCES 1. Via, D. P., and Smith, L. C. (1986) in Methods in Enzymology (Albers, J. J., and Segrest, J. P., Eds.), Vol. 129, pp. 848-857, Academic Press, San Diego. 2. Pitas, R. E., Innerarity, T. L., Weinstein, J. N., and Mahley, R. W. (1981) Arteriosclerosis 1,177-185. 3. Barak, L. S., and Webb, W. W. (1981) J. Cell. Biol. 90,595-604. 4. Sims, P. J., Waggoner, A. S., Wang, C. H., and Hoffman, J. F. (1974) Biochemistry 13,3315-3330. 5. Berg, K. A., Berry, M. L., Sapareto, S. A., and Petty, H. R. (1986) Biochim. Biophys. Acta 887,304-314. 6. Traill, K. N., Bock, G., Winter, U., Hilchenbach, M., Jurgens, G., and Wick, G. (1986) J. Hi&o&em. Cytochem. 34,1217-1221. 7. Traill, K. N., Jurgens, G., Bock, G., Huber, L., Schonitzer, D., Widhalm, K., Winter, U., and Wick, G. (1987) Mech. Ageing Deu. 40,261-288. 8. Traill, K. N., Jurgens, G., Bock, G., and Wick, G. (1987) Zmmunobiology 176,447-454. 9. Jurgens, G., Xu, Q., Huber, L. A., Bock, G., Howanietz, H., Wick, G., and Traill, K. N. (1989) J. Biol. Chem. 264,8549-8556. 10. Suzuki, K., Sakata, N., Kitani, A., Hara, M., Hirose, T., Hirose, W., Norioka, K., Harigai, M., Kawagoe, M., and Nakamura, H. (1990) Biochim. Biophys. Acta 1042, 210-216. 11. Loew, L. M., Bonneville, G. W., and Surow, J. (1978) Biochemistry 17,4065-4071. 12. Loew, L. M., Simpson, L., Hassner, A., and Alexanian, V. (1979) J Am. Chem.Soc. 101,5439-5440. 13. Foster, R., and Hammick, D. L. (1954) J. Chem. Sot. 2685-2692. 14. Campaigne, E., and Archer, W. L. (1953) J. Am. Chem. Sot. 75, 989-991. 15. Brooker, L. G. S., Keyes, G. H., Sprague, R. H., Van Dyke, R. H., Van Lare, E., Van Zandt, G., White, F. L., Cressman, H. W. J., and Dent, S. G., Jr. (1951) J. Am. Chem. Sot. 73,5332-5350. 16. Still, W. C., Kahn, M., and Mitra, A. J. (1978) J. Org. Chem. 43, 2923-2925. 17. Havel, R. J., Eder, H. A., and Bragdon, J. H. (1955) J. Clin. Znuest. 34,1345-1353. 18. Markwell, M. K., Haas, S. M., Bieber, L. L., and Tolbert, N. E. (1978) Anal. Biochem. 87, 206-210. 19. Forte, T. M., and Nordhausen, R. W. (1986) in Methods in Enzymology (Segrest, J. P., and Albers, J. J., Eds), Vol. 128, pp. 442457, Academic Press, Orlando, FL. 20. Sparks, C. E., Sparks, J. D., Bolognino, M., Salhanick, A., Strumph, P. S., and Amatruda, J. M. (1986) Metabolism 35, 1128-1136. 21. Brown, M. S., and Goldstein, J. L. (1974) Proc. Nat. Acad. Sci. USA 71,788-792. M. S., and Goldstein, J. L. (1974) J. Biol. Chem. 249, 22. Brown, 5153-5162. 23. Pitas, R. E., Innerarity, T. L., Arnold, K. S., and Mahley, R. W. (1979) Proc. Nat. Acad. Sci. USA 76.2311-2315. 24. Harkes, L., and VanBerkel, T. J. C. (1982) Biochim. Biophys. Actu

712,677-683. 25. Sanghvi, 26. Brown,

A., and Warty, V. (1985) Biochem. J. 227,397-404. M. S., and Goldstein, J. L. (1975) Cell 6,307-316.

27. Goldstein, J. L., and Brown, M. S. (1990) Nature 343,425-430. M. S., Bilheimer, D. W., and Goldstein, J. L. 28. Ho, Y. K., Brown, (1976) J. Clin. Znuest. 58, 1465-1474.