Vol. 44, No. 2, February 1998
BIOCHEMISTRYand MOLECULAR BIOLOGY INTERNATIONAL Pages 347-362
TRANSFER OF CHOLESTEROL FROM MACROPHAGES
TO LYMPHOCYTES IN CULTURE Paulo I. H o m e m de Bittencourt Jr. 1 and Rui Curi 2
1Department of Physiology,Institute of Basic Health Sciences, Federal University of Rio Grande do Sul, 90050-170 Porto Alegre, RS, Brazil and Department of Physiologyand Btophystcs,Institute of Biomedical Sciences, University ofSio Paulo, 05508-900 Silo Paulo, SP, Brazil 2
9
9
9
Received July 29, 1997 Received afterrevision September 18, 1997 S u m m a r y . A major feature of macrophage metabolism is its capacity to produce and export cholesterol. Several reports have shown that the manipulation of lymphocyte cholesterol content elicits important changes in lymphocyte proliferation. These findings lead to an inquiry as to whether macrophage-derived cholesterol released into the lymphocyte surroundings may be transferred to ttie latter thus affecting lymphocyte function. In this study, cholesterol transfer from macrophages to lymphocytes was examined in vitro using rat cells in culture. The findings indicate that there may be a significant transfer of cholesterol from [4-14C]cholesterol labeled resident peritoneal macrophages to mesenteric lymph node resting lymphocytes (up to 173.9 + 2.7 omol/107 lvmphocytes/107 macrophages when co-cultivated for 48 h), in a lipoprotein-dependent'manner. This represents the mass transfer of ca. 17 nmoles of cholesterol molecules per 107 lymphocytes from 1 0 7 macrophages (calculated on the basis of specific radioactivity incorporated mto macrophages after the pre-labelling period), which suggests that macrophages are capable of replacing the whole lymphocyte cholesterol pool every 21 h. Moreover, an 11 l%-increase in the total cholesterol content oflymphocytes was found after co-cultivation with macrophages for 48 h. When compared to peritoneal cells, monocytes/macrophages obtained from circulating blood leukocytes presented a much higher cholesterol transfer capacity to lymphocytes (3.06 + 0.10 nmol/107 lymphocytes/107 macrophages co-cultivated for ~4 h). Interestingly, inflammatory macrophages dramatically reduced their cholesterol transfer ability (by up to 91%, as compared to resident macrophages). Cholesterol transfer may involve a humoral influence, since it is not only observed when cells are co-cultivated in a single-well chamber system (cells in direct contact), but also in a two-compartment system (where cells can communicate but not by direct contact). Co-cultivation with macrophages decreased the basal incorporation of [2J4C]thymidine into lymphocyte DNA and the additmn of cholesterol to lymphocyte culture media also impaired the lymphoeyteproliferative response to the mitogens concanavalin A (Con A) and bacterial lipopolysaccharide (LPS). The above results suggest that macrophages may transfer cholesterol to lymphocytes (fromboth lymph nodes and blood), thus regulating lymphocyte function by raising the lntracellular cholesterol content and suppressing lymphocyte proliferative activity. If this is so, a words: cholesterol transfer, lipid transfer, cell-to-cell interaction, lipid metabolism, lymphocyte proliferation, lymphocytes, monocytes/macrophages. Key
Abbreviations: apo-E etc., apolipoprotein-Eetc.; BSA, bovine serum albumin; Con A, concanavalinA; DTNB,
5,5'-dithio-bis-(2-nitrobenzoicacid); FCS, fetalcalf serum; HI)L, high-densitylipoprotein;HMG-CoA,3-hydroxy3-methylghitaryl-CoA;IDL, intermediate-densitylipoprotein; LCAT, lecithin-cholesterolacyltransfcrasr LDL, low-density lipoprotein; LTP, lipid transfer protein; LPS, bacterial lipopolysaccharide;PBS, phosphate-buffered saline; TLC, thin-layerchromatography;VLDL,verylow-densitylipoprotein. Address for reprint requests: Paulo lye Homcm do Bittcncourt Jr., Dept. of Physiology, Instituteof Basic Health
Sciences,FederalUniversityof Rio Grandedo Sal, Rua SarmentoLeite 500, 90050-170 Porto Alegre,RS, Brazil.
Phone: +55-51-3163621; Fax: +55-51-3163166; e-mall:
[email protected]
1039-9712/98/020347-16505.00/0 347
Copyright 9 1998 by Academic Press Australia. All rights of reproduction in any form reserved.
Vol. 44, No. 2, 1998
BIOCHEMISTRY and MOLECULAR BIOLOGY INTERNATIONAL
modulatory role for the transfer of cholesterol in both physiological (e.g. immune response) and pathological conditions (e.g. atherosclerosis) may be postulated. This hypothesis is currently under investigation in our laboratory. INTRODUCTION Macrophages have an intense cholesterol metabolism, being capable of recycling their entire plasma membrane once every 30 minutes (1) and having a high ability to produce and export cholesterol, which may be modulated by prostaglandins (2,3). On the other hand, lymphocytes treated with cholesterol exhibited an enriched sterol content in their plasma membrane and a markedly suppressed mitogenic-induced differentiation to blast cells (4). Although cholesterogenesis ( d e n o v o synthesis of cholesterol from acetyl-CoA) is essential for cell proliferation (5), cholesterol p e r s e may be regarded as a potential inhibitor of cell proliferation, since it is a feedback inhibitor of cholesterogenesis by blocking 3-hydroxy-3-methylglutaryl-CoA (HMG-CoA) reductase, a key-enzyme of the pathway (5-8). It has also been shown that lowdensity lipoproteins (LDL), the major source of cholesterol in the plasma, potently inhibit HMG-CoA reductase (5). Moreover, LDL effectively impairs both phytohemagglutinin (PHA)and concanavalin A (Con A)-induced lymphocyte proliferation (9-13), although at subphysiological concentrations LDL may stimulate Con A-induced lymphocyte activation (14-15). The above information led us to argue that macrophage-derived cholesterol released into the lymphocyte surroundings may be transferred to the lymphocyte free cholesterol pool, thus affecting its function. In this study, this possibility is tested using rat peritoneal resident or inflammatory macrophages, rat monocytes/macrophages from the blood, and rat lymphocytes from the mesenteric lymph nodes and blood. MATERIALS AND METHODS C e l l s - Mesenteric lymph node lymphocytes and peritoneal resident macrophages were obtained from the animals (male Wistar rats, 4-6-months old, 7 per experiment) sacrificed by cervical dislocation, following the University of S$o Paulo Animal Committee guidelines, as previously described (2). Inflammatory macrophages were obtained from thioglycollate-injected (intraperitoneally) animals as described elsewhere (16). Mesenteric lymph node lymphocytes obtained from these animals are called here thioglycollate-stimulated lymphocytes, since intraperitoneal thioglycollate treatment has been shown to affect lymphocyte function (16). Cells were then pooled and cultured at 37~ in a 5% (v/v) CO2 atmosphere in air in Eagle's minimum essential medium (G-ibco, BRL, UK) supplemented with antibiotics and 10% fetal calf serum (FCS, Cultilab, Campinas, Brazil) except when stated. Prior to labeling procedures, macrophages were left to attach on polystyrene dishes (Coming, USA) and, after 2 h, non-attached cells were removed. After preparation, lymphocytes (> 98% pure) were seeded and cultivated in polystyrene culture flasks (Coming, USA) for 2 h in order to eliminate all attachable cells. C i r c u l a t i n g b l o o d l e u k o c y t e s - In order to investigate the significance of the cholesterol transfer process in cells that are constantly in contact with one another, as under physiological conditions, monocytes/macrophages and lymphocytes were prepared from circulating blood. For this purpose, rats were heparinized (200 1U/kg, i.p.) 10 min before the experiments and blood was collected by decapitation. Blood samples were then pooled (10 animals per experiment) and circulating leukocytes were separated by layering cell suspension over 1.077 g/ml Lymphoprep | solution (Nycomed Pharma AS, Oslo, Norway), following the manufacturer's instructions. Afterwards, leukocytes were suspended in the medium and incubated for 2 h, at 37~ in a 5% (v/v) COE atmosphere in air. After this period, non-attachable cells (essentially lymphocytes) were harvested from the plates containing monocytes/macrophages. During the experiments, original blood cell ratios were preserved. 348
Vol. 44, No. 2, 1998
BIOCHEMISTRY and MOLECULAR BIOLOGY INTERNATIONAL
Labeling procedures - In order to label the macrophage intracellular cholesterol pool, resident
and thioglycollate-stimulated cells (~107/well in 1-ml final volume, except when stated) were cultivated in 24-well polystyrene dishes (Coming, USA) with pre-purified (thin-layer chromatography, TLC) 0.1 ~tCi/ml [4Jac]cholesterol (57.1 mCi/mmol, NEN, Du Pont Co., USA) dissolved in ethanol (0.05% final concentration, by vol.) in Eagle's essential medium supplemented with antibiotics and 10 mg/ml dialyzed (against PBS, pH 7.4, for 24 h at 4~ lipid-free bovine serum albumin (BSA) prepared as described (17). Preliminary experiments established that, under the above conditions, the most appropriate labeling period for macrophages was found to be 18 h, when about 50-75% of radioactivity from [4-14C]cholesterol (depending on the cell number) was incorporated into macrophages essentially into free cholesterol (94-95%) and cholesteryl ester (5-6%) pools (Table 1); longer incubation periods, although ensuring stronger labeling, have shown less effective cholesterol transfer. To examine the possibility of cholesterol transfer from lym~hocytes to macrophages, mesenteric lymph node resting lymphocytes were also labeled with [4-1 C]cholesterol and co-cultivated as when assessing cholesterol transfer from macrophages to lymphocytes. In order to assess the transfer of intracellularly synthesized 14C-cholesterol from macrophages to lymphocytes, resident and thioglycollate-stimulated peritoneal macrophages (~107/well in 1-ml final volume) were also labeled with [1J4C]acetate (sodium salt, 52.8 mCi/mmol, Amersham, UK) at a final concentration of 2 mM (0.4 p~Ci/ml) for different time periods. The best labeling period with [1J4C]acetate for transfer studies was found to be 12 h, although the incorporation of acetyl units into macrophage lipids could reach higher values with longer incubation periods (Table 2). Co-cultivation and transfer o f radioactivity between cells - After labeling periods, supernatant fluids were discarded, macrophage monolayers were washed three times with Eagle's medium (remaining radioactivity equaling background counts) and fresh medium was added. For cocultivation, a lymphocyte suspension (1-2 x 107 cells) was then prepared in 10-mm diameter polycarbonate-membrane inserts (Nunc, Roskilde, Denmark) which were mounted onto macrophage-attached 24-well dishes. This two-compartment chamber system allows full humoral exchange and communication while hindering direct contact between the two cell types. Afterwards, the cells were co-cultured for 12, 24 or 48 h at 37~ in a 5% (v/v) CO2 atmosphere in air in Eagle's essential medium supplemented with antibiotics and 10% FCS (v/v). Lymphocytes were obtained from the same macrophage-donor animals, except when indicated in the figure legends. At the end of culture periods, lymphocytes were harvested from the inserts (upper compartment) by using a micropipette and macrophages were collected with teflon cell scrapers (Sigma, USA). Alternatively, when specifically stated, cells were co-cultivated in the conventional one-chamber model (both cell types in the same well, thus permitting direct cell contact). In this case, after co-culturing periods, lymphocytes were harvested by using a micropipette and, in order to eliminate macrophages eventually detached during co-cultivation, the cells were incubated for 30 min, at 37~ under agitation with iron powder (mesh 400, -10 mg/107 lymphocytes in 2 ml medium). This procedure allows prompt phagocytosis of iron by macrophages, which become heavier, thus precipitating under Lymphoprep | gradient centrifugation. The suspension was then layered over 1.077 g/ml Lymphoprep| centrifuged at 200 x g, for 5 min, at 4~ The turbid interface between the medium and Lymphoprep | solution (>99% lymphocytes) was collected and the infranatant fluid (containing macrophages with phagocytosed iron particles) saved for further analysis or, eventually, discarded. By using either one- or two-compartment systems, lymphocytes were transferred to 2-ml microcentrifuge tubes, washed twice with cold PBS, pelleted and disrupted by 0.5 ml of methanol, on ice, under agitation (by using a pipette). One milliliter of chloroform and 0.4 ml PBS were added and samples were agitated with the use of a vortex and centrifuged (15,000 xg) for 2 min. The chloroform phase was saved and the aqueous-methanolic phase (upper layer) was re-extracted with 0.5 ml chloroform. Chloroform fractions were then pooled and washed once with the mixture chloroform/methanol/water (3:48:47, by vol.). After centrifugation (15,000 x g for 2 min), the upper phase was discarded and the chloroform fraction dried in a SpeedVac concentrator (Savant Instruments, UK) and kept at -20~ until use. To be analyzed, samples were dissolved in 50 ~tl chloroform/methanol (2:1 v/v) containing 2.5 ~ butylated hydroxytoluene (Sigma, USA), chromatographed by TLC and the lipid samples of interest were then gently scraped from the silica plates and counted in 2 ml scintillation cocktail (2). Recoveries for cholesterol, cholesteryl
349
Vol. 44, No. 2, 1998
BIOCHEMISTRY and MOLECULAR BIOLOGY INTERNATIONAL
Table 1. Profile o f [4-~4C]cholesterol incorporation into cholesterol, cholesterol esters and oxysterol derivatives of resident and thioglycollate-elicited (inflammatory) rat peritoneal macrophages after 18 h. Cell type T L C band
Resident
Inflammatory
Cholesterol Cholesteryl esters Oxysterols
801.2 + 34.0 45.4 + 5.0 6.4 + 3.5
1252.0 + 96.3 65.6 + 7.6 9.0 + 2.4
Total
853.0 + 35.0
1326.6 + 96.6
Macrophages were pre-labeled for 18 h, cellular lipids were extracted and chromatographed by TLC and then radioactivity incorporated into the bands of interest were counted. Data are expressed in pmoles of cholesterol units incorporated into 107 macrophages as the means + S.E.M. of four individual experimental preparations, each one performed in triplicate. Total values represents the sum of the radioactivity found in all the three TLC bands. For details, see Materials and Methods.
Table2. Time-course of [1-~4C]acetate incorporation into lipid extracts (from cells and supernatant fluids) of resident and thioglycollate-elicited (inflammatory) rat peritoneal macrophages cultured for up to 48 h. Cell type
Culture time (h) Resident
Inflammatory
Cell fraction
12 48
1.80 + 0.28 3.24 + 0.26
2.94 + 0.10 5.00 + 0.40
Supernatant Fluid
12 48
2.99 + 0.25 10.39 + 0.76
5.61 + 0.47 16.94 + 1.24
Macrophages were pre-labeled for the indicated time periods. Afterwards, total lipids from cells and supernatants (synthesized 14C lipids exported from labeled cells) were extracted, chromatographed by TLC and radioactivity was counted. Data are expressed in nmoles of acetyl units incorporated into total lipids per 107 macrophages as the means + S.E.M. of five individual experimental preparations, each one performed in triplicate. For details, see Materials and Methods.
350
Vol. 44, No. 2, 1998
BIOCHEMISTRY and MOLECULAR BIOLOGY INTERNATIONAL
esters, triacylglycerols, free fatty acids and phospholipids by using the above procedure were always higher than 85%, as inferred with radiolabeled internal standards of the respective compounds. All data derived from radiolabeling procedures in this work were obtained by counting TLC lipid bands after the steps described above. In some experiments, total lipids from cells and supernatant solutions were extracted as described above and dried samples were dissolved in 100 Ixl of ethanol for the determination of total cholesterol content (cholesterol and cholesteryl esters) in 20-1xl samples by using the cholesteryl-ester hydrolase/cholesterol oxidase method (18). ~ Isolation oflipoproteins - Lipoproteins were prepared from FCS (pooled from the same lots used in transfer experiments - frozen/thawed once) by inverted rate-zonal ultracentrifugation and sequential flotation in a vertical rotor (VTi 80, Beckman, USA) at 510,000 xg, for 45 min, at 12~ in the presence of 2.5 I.U./ml penicillin, 2.5 rtg/ml streptomycin, 0.01% (w/v) thiomersal, 50 laM phenylmethylsulfonyl fluoride, 0.04% (w/v) EDTA, 0.05% (w/v) NAN3, 50 ixM DTNB, as adapted from reference (19). The following fractions were obtained: very-low-density lipoproteins (VLDL), 0.940-1.006g/ml; intermediate-density lipoproteins (IDL), 1.006-1.019 g/ml; lowdensity lipoproteins (LDL), 1.019-1.063 g/ml; high-density lipoproteins (HDL), 1.063-1.210 g/ml; lipid-transfer protein inhibitor activity, 1.210-1.250 g/ml; lipid-transfer protein activity present in lipoprotein deficient serum (>1.250 g/ml). In some experiments, when stated, FCS-derived lipoprotein deficient serum was purchased from Sigma (USA). After preparation, all the above fractions were dialyzed twice against 100 volumes PBS, pH 7.4, at 4~ for 12 h, sterilized by filtration (0.22 Ixm membranes, Millipore, USA) and then protein content was assessed (20). Total cholesterol present in each fraction was determined by the cholesterol oxidase method (18). The FCS used throughout the experiments, including that for the preparation of lipoprotein fractions, had been previously frozen/thawed once. Nevertheless, the transfer phenomenon was always observed with its addition to co-culture medium whereas its absence markedly impaired the referred process, in spite of the possible alterations that freezing/thawing could bring to lipoprotein integrity. Cell integrity and incorporation o f labeled thymidine into lymphocyte D N A - After culturing and co-culturing periods, cell viability was assessed in macrophages and lymphocytes by trypan blue exclusion. In order to evaluate lymphocyte proliferative activity, the rates of [2J4C]thymidine incorporation into cell DNA were measured in resting (unstimulated), 5 ~tg/ml Con-A and 10 rtg/ml LPS-stimulated lymphocytes, as described elsewhere (3). To study lymphocyte proliferation in co-culturing systems, both one- and two-compartment chamber models were employed under the same conditions used for labeling procedures. In the first model, macrophages (10 ~ cells) and lymphocytes (10 ~ cells) were seeded together in 96-well dishes, whereas, in the two-compartment model, macrophages (105 cells in 150 Ixl medium I were plaqued in the lower chamber (bottom of 96-well plates) and lymphocyte suspension (10 cells in 60 rtl medium) was added to the interior of upper chamber (8-well strip Anopore TM inserts, Nunc, Denmark). In the experiments where unlabeled free cholesterol was added to culture media, the sterol (usually 5 rtg/ml) was dissolved in ethanol (0.05% final concentration, v/v) and controls received equal amounts of the vehicle. Statistical analysis - As required in each case, differences were compared using the unpaired bitailed Student's t-test or by analysis of variance (ANOVA) with the Scheff~'s t-test; the significance level was set for p5000 dpm, background radioactivity
