131
Biochimica et Biophysics Acta, 529 (1978) 13X-137 0 Elsevier/Nor~h-HoIl~d Biomedical Press
BBA 57158
INDUCTION OF LOW DENSITY LIPOPROTEIN RECEPTOR SYNTHESIS BY HIGH DENS~Y L~OPROTE~ IN C~TURES OF ~~ SKIN FIBROBLASTS
NORMAN E. MILLER
Unit of Ca~io~c~~r Meta~ol~m and Na~i~on Research, Baker medical Research fnstitute, Melbourne (Australia) (Received September 23rd, 1977) (Revised manuscript received December
22nd, 1977)
Further studies have been made of the effects of high density lipoprotein (HDL) on the surface binding, internalization and degradation of 12sI-labeled low density lipoprotein (12’1-labeled LDL) by cultured normal human fibroblasts. In agreement with earlier studies, during short incubations HDL inhibited the surface binding of ‘2sI-labeled LDL. In contrast, following prolonged ~cubations 1251-labeledLDL binding was consistently greater in the presence of HDL. The increment in 12sI-labeled LDL binding induced by HDL was: (a) associated with a decrease in cell cholesterol content; (b) inhibited by the addition of cholesterol or cycloheximide to the incubation medium; and (c) accompanied by similar increments in 12sI-labeled LDL internalization and degradation. It is concluded that HDL induces the synthesis of high affinity LDL receptors in human fibrobl~s by promoting the efflux of cholesterol from the cells. Introduction The surface membrane of human fibrobl~ts contains specific receptors for plasma low density lipoprotein (LDL) [1,2]. After the lipoprotein is bound to the receptors, it is taken into the cell and catabolised. The protein component is degraded to trichloroacteic acid-soluble fragments, which are released from the cell, and the cholesteryl ester is hydrolysed [1,4,5]. The free cholesterol so released inhibits cellular cholesterol synthesis [5,6] and the synthesis of LDL receptors [ 7] _ Fib&blasts also bind plasma high density lipoprotein (HDL). The binding of HDL is quantitatively similar to that of LDL, but HDL is much more slowly * Abbreviations: tein (1.063-1.21
LDL, low density g/ml) [31.
lipoprotein
(1.006-1.063
g/ml) [31; HDL,
hish
density lipopro-
132
internalized [8]. Accordingly, HDL does not increase cell cholesterol content [8]. On the contrary, when fibroblasts have been grown in medium containing LDL, subsequent exposure to HDL reduces cell cholesterol content and enhances cholesterol synthesis from acetate [9,10]. The binding of HDL to the cell membrane inhibits that of LDL, producing corresponding reductions of LDL internalization and degradation [ 111. The effects of HDL on cell cholesterol efflux and on LDL binding have been suggested as alternative explanations for the negative correlation which exists in vivo between body cholesterol pool size and the plasma HDL cholesterol concentration [ 12,131. The effects of HDL on LDL binding by fibroblasts have to date been examined during short incubations only. This was essential for the purpose of characterizing the interaction during binding in the absence of any changes in receptor function which might result from internalization and degradation of lipoprotein or from lipid transfer between lipoproteins and the cell membrane. The present study was undertaken to examine the possibility that such changes might significantly modify the LDL-HDL interaction. Methods Materials. Carrier-free Na12’I was obtained from The Radiochemical Centre, Amersham, U.K. Sodium heparin (grade II; porcine intestinal mucosa), bovine y-globulin, 5a-cholestane and cycloheximide were purchased from Sigma Chemical Co., St. Louis, U.S.A. Cholesterol (>99% pure) was obtained from B.D.H. Chemicals Ltd., Melbourne, Australia. Polystyrene tissue culture ware was purchased from Falcon Plastics, Oxnard, U.S.A. Fetal calf serum (cholesterol, 34 mg/dl; protein, 40 mg/ml), bovine serum albumin, Eagle’s basal medium [14], Dulbecco’s phosphate-buffered saline [15] and normal human skin fibroblasts at the 9th doubling number were obtained from the Commonwealth Serum Laboratories, Melbourne, Australia. Cell cultures. Fibroblast monolayers were maintained in 75 cm2 flasks in a humidified incubator (95% sir/5% C02) at 37°C in Eagle’s medium containing 10% (v/v) fetal calf serum/l0 mM NaHCOJ2 mM glutamine/lOO units/ml penicillin G/100 pg/ml streptomycin/2 pg/ml amphotericin B. Experiments were performed with cells grown to 60-90% confluency in 60 X 15 mm dishes (4-6 days after seeding with 1 + lo’--2 - 10’ cells). At confluency there were 1.5 * lo6 cells and 250 pg of cell protein per dish. Cells were studied between the 1st and 10th passages. Lipoproteins and lipoprotein-deficient serum. Lipoproteindeficient serum (density >1.21 g/ml; 50 mg protein/ml; 13 pg cholesterol/ml), LDL and HDL were prepared as previously described [8,11] from fasting healthy normoof lipidemic human subjects. 1251-labeled LDL was prepared by a modification the ICI method of McFarlane [ 161 as previously described [8]. Specific activities were 62-239 cpm/ng protein. Less than 2% of ‘251-labeled LDL radioactivity was soluble in trichloroacetic acid and less than 0.5% was extractable into CHClJCH3OH (2/l, v/v). Cellular uptake and degradation of lipoprotein. At the end of an incubation with ‘251-labeled LDL the medium was removed and the monolayers were washed six times at 0-4°C as previously described [8,11]. Surface binding of
133
LDL was quantified from the amount of lzsI which was released from the cells during incubation (20 min, 4°C) with 2 ml 0.6% (w/v) heparin/O.lfi M NaCl [17]. The cells were dissolved in 2 ml 0.1 M KOH (60 min, 37°C) and aliquots taken for measurement of residual cell-associated 1251, representing internalized lipoprotein [2,4,8], and of cell protein using albumin as standard [ 181. Degradation of 1251-labeled LDL by fibroblasts was calculated from the accumulation of non-iodide trichloroacetic acid-soluble “‘1 in the medium in excess of that occurring in the absence of cells [1,8,19]. All data on the concentration, specific activity, and metabolism of “‘Ilabeled LDL are expressed in terms of protein content. Cell cholesterol content. Cell monolayers were washed and dissolved in KOH as described above. After the removal of aliquots for protein assay [18], 10 I.rg 5cY-cholestane were added as internal standard and the mixture saponified (1 M KOH/70% ethanol for 3 h at 60°C). The non-saponifiable lipids were extracted into hexane and their cholesterol content determined by gas-liquid chromatography of the trimethylsilyl ether according to Miettinen et al. [ 201. Results The effect of HDL on LDL binding by fibroblasts as a function of time (3-72 h) was first examined using cells which had been preincubated for 24 h in medium containing 5% (v/v) lipoprotein-deficient serum. The results of a representative experiment appear in Fig. 1. In accordance with previous studies [11], after a 3 h incubation LDL binding was reduced by 24% in the presence of HDL at a protein mass ratio to LDL of 12 : 1. With continued incubation, progressive reductions in LDL binding were observed in both the presence and the absence of HDL. This change was much less marked in monolayers exposed
~150 E
-
S LL =: tllW P c uJ50 C 1%
f ?? !z
0 0
24 43 TIME (HOURS)
72
Fig. 1. Effect of HDL on the surface binding of 1251-labeled LDL by cultured human fibroblasts as a function of the time of incubation (cells prelncubated with lipoprotein-deficient serum). Fibroblast monolayers were incubated for 24 h ln 2 ml of medium contalning 5% (v/v) lipoproteindeficient serum. The medium was then replaced with 2 ml of fresh medium contalning 5% lipoprotein-deficient serum and either 21 pg/ml of 1251-labeled LDL alone (a) or 21 rg/mI of 1251.labeled LDL plus 250 fig/nil of HDL (01. After further incubation for the indicated periods of time. duplicate dishes were harvested for measurement of LDL binding and cell cholesterol content as described under Methods. At each time interval after 3 h the medium in those dlahes which were not harvested was replaced with 2 ml of fresh medium of identical composition. Values presented are the means of duplicate determinations. Coefficient of variation: 3%. 12gI-labeled LDL specific acitivity. 239 cpm/ng. Mean cell protein per dish, 166 pg.
134
to both lipoproteins, however, to the extent that after 72 h LDL binding was 78% greater in the presence of HDL. The smaller decrement in LDL binding with time which occurred in the presence of HDL was associated with a smaller increment in cell cholesterol content (8 vs. 27 pg/mg cell protein). Fibroblasts that have been preincubated for 24 h with lipoprotein-deficient serum have a low cholesterol content due to continued loss of cholesterol in the absence of LDL uptake [ 91, and a maximum number of LDL receptors due to a lack of suppression of receptor synthesis by cellular cholesterol [7]. The increase in cell cholesterol content and decrease in LDL binding with time which were observed in the above experiments can be attributed to internalization and degradation of ‘251-labeled LDL. Similar experiments were subsequently performed using fibroblasts directly after growth in medium containing 10% fetal calf serum and unlabeled LDL (Table I). As would be anticipated, such cells showed only small changes in cholesterol content and LDL binding during incubation with ‘251-labeled LDL alone. Nevertheless, the time course of the effect of HDL on LDL binding under these circumstances resembled that observed in the first experiments: an initial inhibitory effect, demonstrable after 3 h, being reversed during continued incubation for 48 h. The enhancement of LDL binding by HDL now represented an absolute increase, was associated with an absolute decrease in cell cholesterol content, and was found to be prevented by the addition of cholesterol to the medium. These observations suggested that the increase in LDL binding induced by HDL reflected a rise in LDL receptor synthesis occurring in response to an increase in cell cholesterol efflux. This conclusion was supported by the following results: (1) Binding of LDL by fibroblasts was not increased following preincubation (48 h, 37°C) of the ‘251-labeled LDL preparation (4 pg/ml) with TABLE I LDL BY CULTURED HUMAN EFFECTS OF HDL ON THE SURFACE BINDING OF 1251-LABELED FIBROBLASTS, AND ON CELL CHOLESTEROL CONTENT, AS A FUNCTION OF THE TIME OF INCUBATION Cells used directly after growth in medium containing fetal calf serum and LDL. Fibroblast monolayers were grown in medium containing 10% (v/v) fetal calf serum and 26 pg/ml of unlabeled human LDL. The cells were then washed with 2 ml of Dulbecco’s phosphate-buffered saline and 2 ml of fresh medium was added containing 2% lipoprotein-deficient serum and 26 /.ig/ml of 12s~labeled LDL plus: nil, 255 pg/ml of HDL. or 255 !&ml of HDL and 93 &ml of cholesterol (added in 15 ~1 of ethanol). After further incubation for 3 or 48 h, duplicate dishes were harvested for measurement of LDL binding and cell cholesterol content as described under Methods. Dishes incubated for 48 h were given a change of 2 ml of fresh medium of identical composition after 24 h. Values presented are the means of duplicates. Figures in parentheses are the percentage changes relative to values obtained with 1 251-labeled LDL alone. Coefficients of variation: LDL binding. 6%; cell cholesterol content, 9%. 1251-labeled LDL specific activity, 62 cpm/ng. Mean cell protein per dish, 232 pg. Incubation time
Addition to incubation medium
(h)
Nil
HDL
LDL bound (ng/mg cell protein)
3 48
60 48
41 (-32%) 123 (+156%)
CelI cholesterol content (&mg cell protein)
3 48
41 64
45 26
HDL plus cholesterol
(-4%) (-54%)
25 (-48%) 88 (+63%)
135
HDL (80 Erg/ml) in medium containing 2% lipoproteindeficient serum. In contrast, preincubation of fibroblasts (after growth in medium containing unlabeled LDL) under identical conditions increased their capacity to bind 12’1labeled LDL (18 pg/ml) during a subsequent incubation (3 h, 37°C) by 80112% relative to cells preincubated without HDL. Addition of albumin (5 mg/ ml) or of y-globulin (5 mg/ml) to the preincubation medium had no effect on subsequent LDL binding. (2) The increase in LDL binding which was induced by preincubating fibroblasts with HDL could be completely or partially prevented by the addition of cholesterol or LDL to the preincubation medium (Table II). Cycloheximide, an inhibitor of protein synthesis [21], had a similar effect (Table II). (3) Preincubation (48 h, 37°C) of fibroblasts with HDL (513 E.cg/ml) did not increase their capacity to bind LDL (12 ,ug/ml) when the cells had already been depleted of cholesterol by prior exposure for 48 h to medium containing 10% lipoprotein-deficient serum. On the contrary, the ability of fibroblasts to bind LDL during a 3 h incubation at 37°C was now reduced (P < 0.01) from 250 + 8 to 90 _+12 ng/mg-cell protein (mean f S.E., n = 3) by preincubation with HDL. Separate studies with 1251-labeled HDL showed that the reduction of LDL binding in these experiments could be attributed to the presence of residual surface-bound HDL, measured as trypsinreleasable “‘1 [8]. Thus, the ability of HDL to increase LDL binding was dependent upon its ability to reduce cell cholesterol content, and required normal cellular protein synthesis. In subsequent experiments LDL binding was examined as a function of LDL concentration following preincubation of the fibroblasts for 48 h with differing
TABLE II SURFACE BINDING OF 1251-LABELED LDL BY CULTURED HUMAN FIBROBLASTS FOLLOWING PREINCUBATION OF THE CELLS WITH OR WITHOUT HDL, AND THE EFFECT OF ADDING LDL, CHOLESTEROL OR CYCLOHEXIMIDE TO THE PREINCUBATION MEDIUM Cells used directly after growth in medium containing fetal calf serum and LDL. After growth in medhun containing 16% (v/v) fetal calf serum and unlabeled human LDL (20 &ml), fibroblsst monolayers were prcincubated for 48 h in 2 ml of medium containing 2% lipoprotein-deficient serum with or without HDL (215 pg/ml). to which had been added: nil. LDL, cholesterol or cycloheximide as indicated. The cells were then washed with 2 ml of Dulbecco’s phosphate-buffered saline and placed in 2 ml of fresh medium containing 2% lipoprotein-deficient serum and 1251.labeled LDL (16 Irg/ml). After further incubation for 3 h. the cells were harvested for measurement of LDL bindlng as described under Methods. Results are given as mean f S.E. of values obtained from three dishes. 12sI-labeled LDL specific activity. 177 cpm/ng. Mean cell protein Per dish. 196 fig. P values were obtained by Student’s t-test. n.s.. P > 0.06. Addition to preincubation medium
LDL bound (ng/mg cell protein)
P
Difference (%)
Biochimica et Biophysics Acta, 529 (1978) 13X-137 0 Elsevier/Nor~h-HoIl~d Biomedical Press
BBA 57158
INDUCTION OF LOW DENSITY LIPOPROTEIN RECEPTOR SYNTHESIS BY HIGH DENS~Y L~OPROTE~ IN C~TURES OF ~~ SKIN FIBROBLASTS
NORMAN E. MILLER
Unit of Ca~io~c~~r Meta~ol~m and Na~i~on Research, Baker medical Research fnstitute, Melbourne (Australia) (Received September 23rd, 1977) (Revised manuscript received December
22nd, 1977)
Further studies have been made of the effects of high density lipoprotein (HDL) on the surface binding, internalization and degradation of 12sI-labeled low density lipoprotein (12’1-labeled LDL) by cultured normal human fibroblasts. In agreement with earlier studies, during short incubations HDL inhibited the surface binding of ‘2sI-labeled LDL. In contrast, following prolonged ~cubations 1251-labeledLDL binding was consistently greater in the presence of HDL. The increment in 12sI-labeled LDL binding induced by HDL was: (a) associated with a decrease in cell cholesterol content; (b) inhibited by the addition of cholesterol or cycloheximide to the incubation medium; and (c) accompanied by similar increments in 12sI-labeled LDL internalization and degradation. It is concluded that HDL induces the synthesis of high affinity LDL receptors in human fibrobl~s by promoting the efflux of cholesterol from the cells. Introduction The surface membrane of human fibrobl~ts contains specific receptors for plasma low density lipoprotein (LDL) [1,2]. After the lipoprotein is bound to the receptors, it is taken into the cell and catabolised. The protein component is degraded to trichloroacteic acid-soluble fragments, which are released from the cell, and the cholesteryl ester is hydrolysed [1,4,5]. The free cholesterol so released inhibits cellular cholesterol synthesis [5,6] and the synthesis of LDL receptors [ 7] _ Fib&blasts also bind plasma high density lipoprotein (HDL). The binding of HDL is quantitatively similar to that of LDL, but HDL is much more slowly * Abbreviations: tein (1.063-1.21
LDL, low density g/ml) [31.
lipoprotein
(1.006-1.063
g/ml) [31; HDL,
hish
density lipopro-
132
internalized [8]. Accordingly, HDL does not increase cell cholesterol content [8]. On the contrary, when fibroblasts have been grown in medium containing LDL, subsequent exposure to HDL reduces cell cholesterol content and enhances cholesterol synthesis from acetate [9,10]. The binding of HDL to the cell membrane inhibits that of LDL, producing corresponding reductions of LDL internalization and degradation [ 111. The effects of HDL on cell cholesterol efflux and on LDL binding have been suggested as alternative explanations for the negative correlation which exists in vivo between body cholesterol pool size and the plasma HDL cholesterol concentration [ 12,131. The effects of HDL on LDL binding by fibroblasts have to date been examined during short incubations only. This was essential for the purpose of characterizing the interaction during binding in the absence of any changes in receptor function which might result from internalization and degradation of lipoprotein or from lipid transfer between lipoproteins and the cell membrane. The present study was undertaken to examine the possibility that such changes might significantly modify the LDL-HDL interaction. Methods Materials. Carrier-free Na12’I was obtained from The Radiochemical Centre, Amersham, U.K. Sodium heparin (grade II; porcine intestinal mucosa), bovine y-globulin, 5a-cholestane and cycloheximide were purchased from Sigma Chemical Co., St. Louis, U.S.A. Cholesterol (>99% pure) was obtained from B.D.H. Chemicals Ltd., Melbourne, Australia. Polystyrene tissue culture ware was purchased from Falcon Plastics, Oxnard, U.S.A. Fetal calf serum (cholesterol, 34 mg/dl; protein, 40 mg/ml), bovine serum albumin, Eagle’s basal medium [14], Dulbecco’s phosphate-buffered saline [15] and normal human skin fibroblasts at the 9th doubling number were obtained from the Commonwealth Serum Laboratories, Melbourne, Australia. Cell cultures. Fibroblast monolayers were maintained in 75 cm2 flasks in a humidified incubator (95% sir/5% C02) at 37°C in Eagle’s medium containing 10% (v/v) fetal calf serum/l0 mM NaHCOJ2 mM glutamine/lOO units/ml penicillin G/100 pg/ml streptomycin/2 pg/ml amphotericin B. Experiments were performed with cells grown to 60-90% confluency in 60 X 15 mm dishes (4-6 days after seeding with 1 + lo’--2 - 10’ cells). At confluency there were 1.5 * lo6 cells and 250 pg of cell protein per dish. Cells were studied between the 1st and 10th passages. Lipoproteins and lipoprotein-deficient serum. Lipoproteindeficient serum (density >1.21 g/ml; 50 mg protein/ml; 13 pg cholesterol/ml), LDL and HDL were prepared as previously described [8,11] from fasting healthy normoof lipidemic human subjects. 1251-labeled LDL was prepared by a modification the ICI method of McFarlane [ 161 as previously described [8]. Specific activities were 62-239 cpm/ng protein. Less than 2% of ‘251-labeled LDL radioactivity was soluble in trichloroacetic acid and less than 0.5% was extractable into CHClJCH3OH (2/l, v/v). Cellular uptake and degradation of lipoprotein. At the end of an incubation with ‘251-labeled LDL the medium was removed and the monolayers were washed six times at 0-4°C as previously described [8,11]. Surface binding of
133
LDL was quantified from the amount of lzsI which was released from the cells during incubation (20 min, 4°C) with 2 ml 0.6% (w/v) heparin/O.lfi M NaCl [17]. The cells were dissolved in 2 ml 0.1 M KOH (60 min, 37°C) and aliquots taken for measurement of residual cell-associated 1251, representing internalized lipoprotein [2,4,8], and of cell protein using albumin as standard [ 181. Degradation of 1251-labeled LDL by fibroblasts was calculated from the accumulation of non-iodide trichloroacetic acid-soluble “‘1 in the medium in excess of that occurring in the absence of cells [1,8,19]. All data on the concentration, specific activity, and metabolism of “‘Ilabeled LDL are expressed in terms of protein content. Cell cholesterol content. Cell monolayers were washed and dissolved in KOH as described above. After the removal of aliquots for protein assay [18], 10 I.rg 5cY-cholestane were added as internal standard and the mixture saponified (1 M KOH/70% ethanol for 3 h at 60°C). The non-saponifiable lipids were extracted into hexane and their cholesterol content determined by gas-liquid chromatography of the trimethylsilyl ether according to Miettinen et al. [ 201. Results The effect of HDL on LDL binding by fibroblasts as a function of time (3-72 h) was first examined using cells which had been preincubated for 24 h in medium containing 5% (v/v) lipoprotein-deficient serum. The results of a representative experiment appear in Fig. 1. In accordance with previous studies [11], after a 3 h incubation LDL binding was reduced by 24% in the presence of HDL at a protein mass ratio to LDL of 12 : 1. With continued incubation, progressive reductions in LDL binding were observed in both the presence and the absence of HDL. This change was much less marked in monolayers exposed
~150 E
-
S LL =: tllW P c uJ50 C 1%
f ?? !z
0 0
24 43 TIME (HOURS)
72
Fig. 1. Effect of HDL on the surface binding of 1251-labeled LDL by cultured human fibroblasts as a function of the time of incubation (cells prelncubated with lipoprotein-deficient serum). Fibroblast monolayers were incubated for 24 h ln 2 ml of medium contalning 5% (v/v) lipoproteindeficient serum. The medium was then replaced with 2 ml of fresh medium contalning 5% lipoprotein-deficient serum and either 21 pg/ml of 1251-labeled LDL alone (a) or 21 rg/mI of 1251.labeled LDL plus 250 fig/nil of HDL (01. After further incubation for the indicated periods of time. duplicate dishes were harvested for measurement of LDL binding and cell cholesterol content as described under Methods. At each time interval after 3 h the medium in those dlahes which were not harvested was replaced with 2 ml of fresh medium of identical composition. Values presented are the means of duplicate determinations. Coefficient of variation: 3%. 12gI-labeled LDL specific acitivity. 239 cpm/ng. Mean cell protein per dish, 166 pg.
134
to both lipoproteins, however, to the extent that after 72 h LDL binding was 78% greater in the presence of HDL. The smaller decrement in LDL binding with time which occurred in the presence of HDL was associated with a smaller increment in cell cholesterol content (8 vs. 27 pg/mg cell protein). Fibroblasts that have been preincubated for 24 h with lipoprotein-deficient serum have a low cholesterol content due to continued loss of cholesterol in the absence of LDL uptake [ 91, and a maximum number of LDL receptors due to a lack of suppression of receptor synthesis by cellular cholesterol [7]. The increase in cell cholesterol content and decrease in LDL binding with time which were observed in the above experiments can be attributed to internalization and degradation of ‘251-labeled LDL. Similar experiments were subsequently performed using fibroblasts directly after growth in medium containing 10% fetal calf serum and unlabeled LDL (Table I). As would be anticipated, such cells showed only small changes in cholesterol content and LDL binding during incubation with ‘251-labeled LDL alone. Nevertheless, the time course of the effect of HDL on LDL binding under these circumstances resembled that observed in the first experiments: an initial inhibitory effect, demonstrable after 3 h, being reversed during continued incubation for 48 h. The enhancement of LDL binding by HDL now represented an absolute increase, was associated with an absolute decrease in cell cholesterol content, and was found to be prevented by the addition of cholesterol to the medium. These observations suggested that the increase in LDL binding induced by HDL reflected a rise in LDL receptor synthesis occurring in response to an increase in cell cholesterol efflux. This conclusion was supported by the following results: (1) Binding of LDL by fibroblasts was not increased following preincubation (48 h, 37°C) of the ‘251-labeled LDL preparation (4 pg/ml) with TABLE I LDL BY CULTURED HUMAN EFFECTS OF HDL ON THE SURFACE BINDING OF 1251-LABELED FIBROBLASTS, AND ON CELL CHOLESTEROL CONTENT, AS A FUNCTION OF THE TIME OF INCUBATION Cells used directly after growth in medium containing fetal calf serum and LDL. Fibroblast monolayers were grown in medium containing 10% (v/v) fetal calf serum and 26 pg/ml of unlabeled human LDL. The cells were then washed with 2 ml of Dulbecco’s phosphate-buffered saline and 2 ml of fresh medium was added containing 2% lipoprotein-deficient serum and 26 /.ig/ml of 12s~labeled LDL plus: nil, 255 pg/ml of HDL. or 255 !&ml of HDL and 93 &ml of cholesterol (added in 15 ~1 of ethanol). After further incubation for 3 or 48 h, duplicate dishes were harvested for measurement of LDL binding and cell cholesterol content as described under Methods. Dishes incubated for 48 h were given a change of 2 ml of fresh medium of identical composition after 24 h. Values presented are the means of duplicates. Figures in parentheses are the percentage changes relative to values obtained with 1 251-labeled LDL alone. Coefficients of variation: LDL binding. 6%; cell cholesterol content, 9%. 1251-labeled LDL specific activity, 62 cpm/ng. Mean cell protein per dish, 232 pg. Incubation time
Addition to incubation medium
(h)
Nil
HDL
LDL bound (ng/mg cell protein)
3 48
60 48
41 (-32%) 123 (+156%)
CelI cholesterol content (&mg cell protein)
3 48
41 64
45 26
HDL plus cholesterol
(-4%) (-54%)
25 (-48%) 88 (+63%)
135
HDL (80 Erg/ml) in medium containing 2% lipoproteindeficient serum. In contrast, preincubation of fibroblasts (after growth in medium containing unlabeled LDL) under identical conditions increased their capacity to bind 12’1labeled LDL (18 pg/ml) during a subsequent incubation (3 h, 37°C) by 80112% relative to cells preincubated without HDL. Addition of albumin (5 mg/ ml) or of y-globulin (5 mg/ml) to the preincubation medium had no effect on subsequent LDL binding. (2) The increase in LDL binding which was induced by preincubating fibroblasts with HDL could be completely or partially prevented by the addition of cholesterol or LDL to the preincubation medium (Table II). Cycloheximide, an inhibitor of protein synthesis [21], had a similar effect (Table II). (3) Preincubation (48 h, 37°C) of fibroblasts with HDL (513 E.cg/ml) did not increase their capacity to bind LDL (12 ,ug/ml) when the cells had already been depleted of cholesterol by prior exposure for 48 h to medium containing 10% lipoprotein-deficient serum. On the contrary, the ability of fibroblasts to bind LDL during a 3 h incubation at 37°C was now reduced (P < 0.01) from 250 + 8 to 90 _+12 ng/mg-cell protein (mean f S.E., n = 3) by preincubation with HDL. Separate studies with 1251-labeled HDL showed that the reduction of LDL binding in these experiments could be attributed to the presence of residual surface-bound HDL, measured as trypsinreleasable “‘1 [8]. Thus, the ability of HDL to increase LDL binding was dependent upon its ability to reduce cell cholesterol content, and required normal cellular protein synthesis. In subsequent experiments LDL binding was examined as a function of LDL concentration following preincubation of the fibroblasts for 48 h with differing
TABLE II SURFACE BINDING OF 1251-LABELED LDL BY CULTURED HUMAN FIBROBLASTS FOLLOWING PREINCUBATION OF THE CELLS WITH OR WITHOUT HDL, AND THE EFFECT OF ADDING LDL, CHOLESTEROL OR CYCLOHEXIMIDE TO THE PREINCUBATION MEDIUM Cells used directly after growth in medium containing fetal calf serum and LDL. After growth in medhun containing 16% (v/v) fetal calf serum and unlabeled human LDL (20 &ml), fibroblsst monolayers were prcincubated for 48 h in 2 ml of medium containing 2% lipoprotein-deficient serum with or without HDL (215 pg/ml). to which had been added: nil. LDL, cholesterol or cycloheximide as indicated. The cells were then washed with 2 ml of Dulbecco’s phosphate-buffered saline and placed in 2 ml of fresh medium containing 2% lipoprotein-deficient serum and 1251.labeled LDL (16 Irg/ml). After further incubation for 3 h. the cells were harvested for measurement of LDL bindlng as described under Methods. Results are given as mean f S.E. of values obtained from three dishes. 12sI-labeled LDL specific activity. 177 cpm/ng. Mean cell protein Per dish. 196 fig. P values were obtained by Student’s t-test. n.s.. P > 0.06. Addition to preincubation medium
LDL bound (ng/mg cell protein)
P
Difference (%)
