Effects of CSF-1 on Cholesterol Accumulation and Efflux by Macrophages
To assess whether human monocyte-specific colony-stimulating factor (CSF-1) might influence atherogenesis, CSF-1–induced macrophage responses that might contribute to enhanced clearance of low-density lipoprotein (LDL) or modified LDL were investigated. Careful account was made of cell preservation and increases in cell volume and protein (representing increased cell surface area, and thus endocytically active membrane) during culture with CSF-1. This permitted distinction between selective and nonspecific effects of CSF-1, the latter paralleling increases in cellular mass and volume. CSF-1 enhanced mouse peritoneal macrophage survival in vitro during exposure to lipoprotein-deficient serum with or without native LDL or acetylated LDL (Ac-LDL), as judged by maintenance of cellular DNA and cell numbers. In the presence of copper-oxidized LDL (Ox-LDL), such effects were very slight. In all conditions, CSF-1 increased cellular protein content. CSF-1 increased the uptake of both Ac-LDL and Ox-LDL calculated per culture, but this was entirely explicable by the increased cell protein, indicating that there was no selective enhancement of scavenger receptor or other routes for uptake of the modified LDLs. Similarly, CSF-1 also increased the accumulation of cholesterol and its esters nonspecifically. CSF-1 did have a marked and specific effect on the composition of cholesterol esters, decreasing the proportion of polyunsaturated esters relative to monounsaturated and saturated esters. Finally, cholesterol efflux induced by apolipoprotein A1 from Ac-LDL–loaded macrophages was not influenced by CSF-1. Thus, the enhanced macrophage catabolism of modified LDLs by CSF-1 is part of a nonspecific action on the cells but could contribute to a reduction in circulating cholesterol, observed in some situations of CSF-1 presentation in humans.
- Received November 23, 1995.
- Revision received April 24, 1996.
Human monocyte-specific colony stimulating factor controls proliferation and differentiation of monocytic precursor cells.1 In addition, it has a range of other effects on the functional activity and survival of developing and matured mononuclear phagocytes.2 3 During a phase I/II trial of its activity against refractory aplastic anemia, it was observed that CSF-1 lowered serum cholesterol by an average of 37%.4 Various authors have considered mechanisms that might contribute to this effect, and enhanced clearances of LDL or modified LDL have been the favored candidates. Studies in rabbits5 indicate accelerated clearance of both nLDL and methylated LDL, implying roles for both mechanisms. Similarly, several authors6 7 have claimed that the cellular uptake of Ac-LDL may be enhanced by CSF-1. A key question, as yet poorly resolved, is whether any physiologically modified LDL may be subject to accelerated clearance or uptake as a result of CSF-1 treatment.
We have addressed this issue in some detail. In interpreting data concerning uptake and metabolism of LDLs by mononuclear phagocytes, it is crucial to take account of the large increment in cell volume and cell protein (representing increased cell surface area, and therefore endocytically active membrane), per cell or per unit DNA, which is consequent on CSF-1 exposure.2 3 6 The action of CSF-1 on the survival of mononuclear phagocytes2 3 impacts equally on the interpretation of uptake experiments in vitro. Enhanced uptake could be attributed to enhanced cell protein and enhanced survival or to direct stimulation of the uptake mechanism, and these possibilities need to be dissociated. We find that CSF-1 enhanced the uptake per cell of both Ac-LDL and Ox-LDL, but this was directly proportional to the increment in cell protein. Thus, the actions of CSF-1 observed were primarily due to enhanced cell protein and survival.
Recombinant murine CSF-1 was donated by Cetus Corp (Emeryville, Calif). Pure, lyophilized human apo A-1 was provided by Dr K-A. Rye (Adelaide, Australia) and reconstituted as described.8
Isolation, Modification, and Radiolabeling of LDL
Blood was collected from healthy, normolipidemic, fasting volunteers into tubes containing EDTA (3 mmol/L), aprotinin (90 kallikrein inhibitory units/mL; Sigma Chemical Co) and soybean trypsin inhibitor (20 μg/mL; Sigma). LDL was isolated by a single ultracentrifugation step in a vertical rotor (Beckman VTi50) by use of a discontinuous gradient for 2.5 hours at 242 000g.9 The LDL was then washed to remove contaminating albumin by a second overnight centrifugation at a density of 1.063 in an angle rotor (Beckman Ti70) and then dialyzed for 24 hours against several changes of 50 to 100 volumes of PBS containing 1.0 mg/mL EDTA and 0.1 mg/mL chloramphenicol. All dialysis solutions were deoxygenated by bubbling with nitrogen, and all dialyses were performed under near-anoxic conditions at 4°C in filled, stoppered bottles to minimize LDL oxidation during these preparative steps. The isolated LDL was sterilized by membrane filtration (0.45 μm), stored in the dark in PBS containing 1 mg/mL EDTA and 0.1 mg/mL chloramphenicol at 4°C under nitrogen gas, and used within 2 weeks of preparation.
Acetylation was achieved with the use of acetic anhydride as previously described.8 Excess reagents were removed by dialysis for 24 hours against three 1-L changes of PBS containing 1 mg/mL EDTA and 0.1 mg/mL chloramphenicol.
Before oxidation, freshly prepared LDL was dialyzed for 24 hours against two 1-L changes of PBS containing ≈5 mg/mL Chelex-100 resin (Biorad) and 0.1 mg/mL chloramphenicol and an additional two 1-L changes of PBS with 0.1 mg/mL chloramphenicol alone. Copper-mediated oxidation was subsequently achieved by incubating LDL (400 μg protein/mL) in PBS with a sterile solution of CuCl2 (10 μmol/L final concentration) at 37°C for 24 hours.
We checked the extent of LDL modification before use by assessing the relative electrophoretic mobility of both Ac-LDL and Ox-LDL against that of unmodified nLDL after electrophoresis on 1% agarose gels (Ciba-Corning) at 90 V for 45 minutes in barbital buffer (pH 8.6). The lipoproteins were visualized by staining with fat red 7B.
Radiolabeling of LDL
LDL was labeled with Na125I (5 to 10 mCi/mg; Amersham) by the iodine monochloride method.10 Iodinated LDL (specific activity, 50 to 300 dpm/ng LDL protein) was separated from unbound iodine by use of a Sephadex G-25 column (PD-10; Pharmacia), then either acetylated or oxidized as described above.
Isolation and Culture of Macrophages
Resident macrophages were isolated from QS mice by peritoneal lavage with ice-cold DMEM (Cytosystems ) containing 0.38% (wt/vol) sodium citrate, penicillin G (50 U/mL), and streptomycin (50 μg/mL). The cells were plated in 35-mm-diameter tissue-culture wells (Costar) at 5×106 cells/well, incubated for 1 to 2 hours, washed three times with PBS to remove nonadherent cells, then incubated in DMEM containing 10% (vol/vol) heat-inactivated fetal calf serum, 50 U/mL penicillin, and 50 μg/mL streptomycin at 37°C for 24 hours. These procedures were in accordance with the guidelines of the National Health and Medical Research Council of Australia and were approved by the Central Sydney Area Health Service Animal Welfare Committee.
Measurement of Cell Protein and DNA
Cells were washed twice with PBS and incubated in DMEM containing 10% (vol/vol) LPDS (15 mg protein/mL; ρ>l.24), 50 U/mL penicillin, and 50 μg/mL streptomycin. When present, CSF-1 was normally added to a final concentration of 2×104 U/mL, unless otherwise stated. After 24 or 48 hours, cells were lysed by the addition of 1 mL 0.2 mol/L NaOH and a subsequent 30-minute period of gentle rotation on a flat-bed orbital shaker. Two 200-μL aliquots from the cell homogenate were analyzed for their protein and DNA content. Protein concentrations were determined by use of the bicinchoninic acid method, with BSA (fraction V; Sigma) as standard. Samples were incubated for 60 minutes at 60°C and absorbance measured at 562 nm. DNA concentration was assessed via the fluorescent method described by Labarca and Paigen.11 Briefly, 200 μL of samples and standards (salmon sperm DNA; Sigma) were added to 2 mL of a working solution containing 100 ng/mL fluorescent Hoechst 33258 dye (Sigma) in a solution of PBS (0.05 mol/L sodium phosphate, 2 mol/L NaCl, pH 7.4). The fluorescence of the samples was subsequently measured in a fluorimeter (Hitachi) with excitation at 360 nm, emission at 460 nm, and both the excitation and emission bandpasses set to 5 nm.
Measurement of Cell Number, Size, and Thymidine Incorporation
To measure cell number and size, cultures were washed twice with PBS (prewarmed to 37°C) and then stained with Coomassie brilliant blue R250 as described by Klomp et al.12 Cell number and area were then determined in five randomly selected fields per culture with the use of a ×200 magnification in a Nikon TMS inverted microscope and chromatic color image analysis system (version 2.2).
Incorporation of (6-[3H])-thymidine (Amersham TRK 306; 2 Ci/mmol), added at 1 μCi/mL, was measured in the last 18 hours before harvest. Cells were harvested by washing twice in PBS and then lysing in 1 mL 0.1% (vol/vol) Triton X-100. Two hundred microliters was mixed with 100 μL BSA and 1 mL 3 mol/L TCA and then centrifuged at 3000g for 10 minutes. The pellets were redissolved in 1 mol/L NaOH and counted for tritium with external standards correction.
Determination of Uptake of Modified LDL by Macrophages
125I-LDL was either acetylated or oxidized as described above. Peritoneal macrophages were washed twice with PBS then incubated for 24 or 48 hours in 2 mL DMEM containing 10% (vol/vol) LPDS, 50 U/mL penicillin, and 50 μg/mL streptomycin, plus 50 μg/mL 125I-LDL modified LDL (acetylated or oxidized) and CSF-1 (2×104 U/mL ) where indicated. After 24 or 48 hours, LDL uptake was measured by determination of iodide-free degradation (TCA-soluble) products in the culture medium and accumulation of cell-associated radioactivity, as described previously.13 14 The total uptake of LDL by the cells was defined as the sum of these two measurements.
Measurement of Accumulation of Cholesterol Esters and Oxidation Products by Macrophages
Cells were incubated for 24 or 48 hours in DMEM containing 10% (vol/vol) LPDS, 50 U/mL penicillin, 50 μg/mL streptomycin, and 50 μg/mL of either Ac-LDL or Ox-LDL plus CSF-1 (2×104 U/mL ) where indicated. Medium was removed and cultures were gently washed with PBS and then dissolved in 1 mL 0.2 mol/L NaOH, as described above. Lipid accumulation was quantitatively determined by HPLC analysis; 400-μL aliquots of cell lysate were extracted with 2.5 mL ice-cold methanol and 10 mL hexane in sequence, with vigorous vortexing after each extraction, then centrifuged (Beckman GS-6R) at 1700 rpm at 4°C for 2 minutes; 8 mL of the hexane layer was removed, evaporated under vacuum, and redissolved in 200 μL of eluent (this varied according to the analysis performed), from which 150 μL was injected onto a reverse-phase C-18 column (Supelco; 25-cm length, 0.46-cm diameter, 5-cm guard column, and 5-μm particle size). For cells loaded with Ac-LDL, analysis of cholesterol and cholesterol esters was performed by detecting absorbance at 210 nm after elution with acetonitrile/isopropanol (30/70, vol/vol) as previously described.15 For Ox-LDL–loaded macrophages, oxidized derivatives of cholesterol and cholesteryl esters were analyzed at 234 nm by use of an acetonitrile/isopropanol/water (44/54/2, vol/vol/vol) solvent.15 Separate cultures were grown on coverslips and stained with oil red O after lipoprotein uptake as described above.
Measurement of Cholesterol Efflux From Ac-LDL–Loaded Macrophages
Macrophages harvested as described above were incubated for 24 hours in DMEM containing 10% (vol/vol) LPDS, 50 μg/mL Ac-LDL, 50 U/mL penicillin, 50 μg/mL streptomycin, and CSF-1 (2×104 U/mL). In control cultures, CSF-1 was omitted. After this loading period, cells were washed twice with PBS then incubated for an additional 24 hours in 2 mL/well DMEM containing 100 μg/mL albumin, 50 U/mL penicillin, and 50 μg/mL streptomycin in the presence of apo A-1 (25 μg/mL) where indicated. To measure cholesterol efflux, 1.5 mL of medium was collected from each well, then centrifuged at 4°C for 5 minutes at 10 000 rpm to pellet any detached cells. One milliliter of this medium was placed in a borosilicate tube, 10 μL trifluoroacetic acid was added, and the mixture was extracted with 2.5 mL methanol and 10 mL hexane and analyzed for its cholesterol content by HPLC at 210 nm as described above. After removal of excess medium, the adherent macrophages were washed twice with cold PBS, dissolved in 0.2 mol/L NaOH, extracted into methanol/hexane, and assayed for cholesterol and cholesteryl ester content by HPLC analysis.
Macrophages were loaded with cholesteryl esters by incubation for 24 hours in DMEM containing 10% (vol/vol) LPDS, 50 μg/mL Ac-LDL, 50 U/mL penicillin, and 50 μg/mL streptomycin. The cells were washed with PBS, then incubated for 24 hours in DMEM containing 100 μg/mL albumin, 50 U/mL penicillin, and 50 μg/mL streptomycin to which apo A-1 and CSF-1 (25 μg/mL and 2×104 U/mL, respectively ) were added as appropriate. Both medium and cell homogenate were then analyzed for cholesterol content by HPLC (210 nm) as described above.
Where appropriate, we assessed the significance of differences between groups using the t test with pooled variances, with the use of the MYSTAT package. Probability values <.05 have been taken as significant and are quoted.
Effect of CSF-1 on Macrophage Viability
The protein and DNA content of macrophage cultures maintained in the absence or presence of CSF-1 were compared. Data in Tables 1⇓ and 2 and Fig 1⇓ are for cells that were simultaneously exposed to both CSF-1 and modified LDLs (50 μg/mL), to assist comparison with subsequent data. The addition of CSF-1 to cells for 24 or 48 hours in the presence of 50 μg/mL Ac-LDL produced a dose-dependent change in the recovered cell protein content of the cultures (Fig 1⇓). In the presence of CSF-1, cell protein was largely maintained during 48-hour culture, whereas without this addition, a large and progressive loss in cell protein was observed. The data in Fig 1⇓, which are taken from a single experiment, illustrate a trend that was consistently observed in eight independent experiments. Thus, macrophages cultured in the absence of CSF-1 contained 63±7% (n=8) of the cell protein measured in CSF-1–supplemented cells after 24 hours and only 32±11% (n=8) of equivalent but CSF-1–supplemented (2×104 U/mL) cells after 48 hours.
Measurement of the cell size, cell number, and DNA content of macrophage cultures exposed to CSF-1 indicated that the changes described for cell protein per culture were also reflected in the cell number of the cultures (Table⇑s 1 and 2). CSF-1 inclusion in the culture medium led to a sustained or slight increase in DNA content during a 24-hour period, compared with a substantial loss of DNA in its absence. The enhancing effects of CSF-1 on cell number and DNA, and also on cell size (Table 2⇓), were pronounced in the absence of added lipoproteins or in the presence of nLDL or Ac-LDL but were absent or slight in the case of Ox-LDL. Indeed, when CSF-1 was omitted, the DNA, cell number, and protein content of Ox-LDL–treated cultures were higher than equivalent Ac-LDL, nLDL, or LDL-free cells. To determine whether this effect, which was seen in four independent experiments, is a consequence of improved cell survival in the presence of Ox-LDL or is due to limited stimulation of macrophage proliferation, which has recently been reported to occur in the presence of certain forms of oxidized LDL,16 we also studied 3H-thymidine incorporation. Thymidine incorporation was negligible in the absence of CSF-1 (<500 dpm/culture), although it was slightly greater in Ox-LDL than in the other cultures at 48 hours (413±16 versus 156±19 dpm/culture for nLDL). Thus, the higher protein content of these cultures reflects improved cell survival rather than a stimulatory effect of Ox-LDL on cell proliferation. CSF-1 could stimulate thymidine incorporation into nLDL and Ac-LDL cultures ≈5-fold at 24 hours and 40- to 50-fold at 48 hours (nLDL, 156±19 and 10 828±1079 dpm/culture; Ac-LDL, 298±27 and 11 527±235; without and with CSF-1 at 48 hours, respectively). CSF-1 did not stimulate incorporation into Ox-LDL cultures at 24 or 48 hours, consistent with its very limited effect on cell number or cell DNA in these conditions. Because protein per cell was significantly enhanced by CSF-1 in all conditions and especially at 48 hours (Table 2⇓), we conclude that enhanced cell survival and protein accumulation are the predominant factors in the actions of CSF-1 in our culture conditions and that the slight changes in cell proliferation are of limited importance.
Effect of CSF-1 on Uptake of Modified LDLs by Macrophages
Mouse peritoneal macrophages, cultured without added CSF-1, endocytose Ac-LDL and Ox-LDL at similar rates13 14 and to a large extent by the same scavenger receptors. CSF-1 has previously been reported to lead to an increased expression of the scavenger receptor in human monocyte-derived macrophages during their in vitro differentiation6 and in elicited mouse peritoneal macrophages.7 It is not clear whether this effect is the consequence of accelerated differentiation caused by CSF-1 or a more specific effect on the scavenger receptor per se. We therefore studied the effects of CSF-1 on scavenger receptor activity in the more fully differentiated resident mouse peritoneal macrophages by measuring the endocytosis of modified LDL during 24 hours' incubation of the cells with 125I-labeled lipoproteins. Uptake was measured as the sum of cell-associated radioactivity and iodide-free, TCA-soluble degradation products in the medium at the end of the incubation, and these parameters are plotted separately in Figs 2⇓ and 3. We have previously shown13 14 that endocytosis of both Ac-LDL and Ox-LDL proceeds at a linear rate during this period. Ac-LDL and Ox-LDL were endocytosed at similar rates, as shown previously (Figs 2 and 3⇓⇓). Again, consistent with other studies, Ac-LDL apo B was readily degraded, although a substantial proportion of the internalized Ox-LDL remained largely intact (precipitated by 10% [vol/vol] TCA; data not shown) within the cells. The addition of 2×104 U/mL CSF-1 during these incubations led to an increased uptake of both Ac-LDL (Fig 2A⇓) and Ox-LDL (Fig 3A⇓) per culture. However, because CSF-1 also had a substantial effect on total protein per culture as indicated in Fig 1⇑, when modified LDL endocytosis was expressed per cell protein (Figs 2B and 3B⇓⇓) or per DNA (data not shown), the stimulatory effect was lost or slightly reversed. From these data, it was concluded that these differentiated murine resident peritoneal macrophages do not specifically increase their expression of scavenger receptors in response to CSF-1, at least during the period (48 hours) studied here.
Effect of CSF-1 on Lipid Accumulation by Macrophages
The degree of lipid accumulation in macrophages after incubation with either Ac-LDL or Ox-LDL was assessed both microscopically (after oil red O staining) and quantitatively (by use of HPLC analysis). In all instances, cells loaded with modified LDLs in the presence of CSF-1 appeared to be larger and more heavily loaded with lipid droplets (giving them an exaggerated “foamy” appearance) than the control cells (data not shown). In Ac-LDL–loaded macrophages, the lipid droplets tended to be most dense toward the perimeter of the cell, whereas Ox-LDL–loaded cells were more spherical and had a more even distribution of lipid droplets within the cell body.8 This rounded appearance of the cells was not associated with toxicity and was partially reversible on removal of Ox-LDL from the culture medium.
Subsequent HPLC analyses confirmed the qualitative impression that CSF-1 enhanced cellular lipid accumulation in macrophages loaded with modified LDL. Table 3⇓ presents data from a typical experiment illustrating this point. It shows the concentration of cholesterol and its associated esters for cells incubated with and without CSF-1 in the presence of Ac-LDL for 24 or 48 hours. There is a clear enhancement in overall lipid accumulation after 24 hours, and this is particularly marked after 48-hour incubation. Table 2⇑ also reveals that CSF-1 treatment decreased the proportion of cholesterol esters containing polyunsaturated fatty acids (expressed as the ratio of cholesteryl arachidonate and cholesteryl linoleate to total cholesterol esters) from 85% without CSF-1 to 55% in its presence, at 48 hours. In Ox-LDL–loaded cells, there was also a CSF-1–induced enhancement in the net accumulation of cholesterol and cholesterol ester oxidation products within macrophages (data not shown).
Although CSF-1 caused an increase in overall lipid accumulation, it is evident from both the protein measures and oil red O staining that there was a concomitant increase in cell size and number. As a result, when lipid accumulation is corrected for cell protein (see Table 2⇑), there is no remaining positive effect of the cytokine.
Influence of CSF-1 on Cholesterol Efflux Induced by Apo A-1
The influence of CSF-1 on apo A-1–induced efflux from Ac-LDL–loaded macrophages was investigated by means of two different procedures, detailed in “Methods.” In procedure A (Fig 4⇓), CSF-1 was present only during loading of cells with lipid from Ac-LDL, whereas in procedure B (Fig 5⇓), CSF-1 was present only during the efflux phase. As we have previously shown,8 efflux from Ac-LDL–loaded macrophages in the absence of apo A-1 is normally <10% of the total cellular cholesterol pool (cholesterol plus cholesterol esters), whereas 40% to 50% of the total pool is exported in the presence of apo A-1. In the present study, substantial efflux was again induced by apo A-1, but CSF-1 had no effect on efflux in either condition, whether present during lipid loading or during efflux.
Our evidence indicates that the actions of CSF-1 on the uptake of lipoproteins by macrophages can be explained largely on the basis of the enhanced survival and cell size it induces. As Tables 1 and 2⇑⇑ show, in the absence of CSF-1, cell protein and DNA and cell numbers of cultured resident peritoneal macrophages declined, whereas in its presence they were preserved or enhanced. This phenomenon of CSF-1–mediated stimulation of cell size is consistent with previous studies using rat alveolar macrophages3 and human monocyte-derived macrophages.6 Increased cell number, size, and surface area were also evident in peritoneal macrophages isolated from mice treated with recombinant CSF-1, as well as Kupffer cells in the livers of such animals.2 The effects of CSF-1 on Ox-LDL cultures were quite limited in comparison with the marked actions in nLDL, Ac-LDL, and no-LDL conditions. Studies of thymidine incorporation again showed a lack of response to CSF-1 in Ox-LDL cultures and relatively low incorporation in all conditions. Thus, cell survival and protein accretion was the critical site of action of CSF-1 in our experiments, and cell proliferation was of little importance.
CSF-1 treatment elevates uptake of both Ac-LDL and Ox-LDL per culture but not per cell protein. Enhanced modified lipoprotein uptake per culture may therefore be explained by increased cell survival. Although others suggest that CSF-1 selectively increases scavenger receptor–dependent modified LDL uptake, we believe that some such reports can be largely explained either as above (by increased cell survival) or on the basis of increased protein per cell. For example, although de Villiers et al7 showed substantially increased scavenger receptor protein expression in elicited murine macrophages in response to CSF-1, the increment in functional cell surface Ac-LDL binding (judged by flow cytometry on a per cell basis) was modest and quite possibly within the range explicable by increased cell protein. Similarly, in studies of human monocyte-derived macrophages,6 cholesterol and cholesterol ester accumulation were enhanced on a per DNA basis (with a variety of cholesterol donors). However, per cell protein, the differences in total cholesterol (cholesterol plus cholesterol ester) are not statistically significant. On the other hand, it is clear that in human monocyte-derived macrophages, CSF-1 did enhance Ac-LDL uptake (as opposed to the consequent lipid deposition) and that this exceeds the increments in cell protein and cell survival.6 This is a somewhat different system, because CSF-1 promotes terminal differentiation of monocytes to cells resembling tissue macrophages, a process that would occur in vivo as cells leave the circulation. The resident peritoneal macrophages used in the present study are already differentiated “tissue” macrophages, and as such, their responses to CSF-1 might also be expected to differ.
To appreciate the potential impact of CSF-1 on either cellular lipid levels or plasma cholesterol levels, it is necessary to consider not only lipid uptake into cells but also its intracellular metabolism and efflux from cells. The influence of CSF-1 on intracellular enzymatic handling of endocytosed cholesterol was not previously directly studied. Our data (Table 3⇑) show marked decreases in the ratio of polyunsaturated to total cholesteryl esters and corresponding increases in the ratio of saturated and monounsaturated to total esters. We have not investigated the mechanism of this occurrence, although it presumably involves changes in the activities of the enzymes responsible for fatty acid synthesis and metabolism within the cells.
Several authors17 18 19 have presented conflicting data on changes in the enzymes relevant to the overall catabolism of the endocytosed cholesterol esters, the lysosomal acidic cholesterol esterase, and the extralysosomal acetyl CoA: cholesterol acyltransferase and neutral esterase. Some authors suggested that the overall enzymatic balance shifts in favor of hydrolysis and argued that this might enhance cholesterol efflux. However, in the present study, this was not the case; there was no effect of CSF-1 on the proportional cellular cholesterol efflux to apo A-1 (Figs 4 and 5⇑⇑). Ishibashi et al,6 however, report that addition of CSF-1 during 24 hours of Ac-LDL uptake by monocyte-derived macrophages led to an enhanced efflux to LPDS in a subsequent 8-hour period in the absence of CSF-1. This was argued solely on the basis of measured decreases in cholesterol esters, without assessment of free sterol or sterol recoveries. Furthermore, other workers,17 in agreement with our data, have also failed to detect any differences in efflux caused by CSF-1.
Our observations of nonselective enhancement of lipoprotein uptake, effective on both Ac-LDL and Ox-LDL and explicable on the basis of enhanced cell quantities and size, provide one explanation of the CSF-1–induced reduction in circulating cholesterol originally described in humans.4 Both increased cell survival and increased cell protein per cell may contribute to the enhanced uptake we observed. Furthermore, it is well documented that CSF-1 causes large increases in the numbers of circulating leukocytes and tissue macrophages in vivo, which could contribute similarly to the removal of serum cholesterol.2 4 In animals, CSF-1–induced reductions of circulating cholesterol have been inconsistent, with both positive20 21 and negative17 18 results obtained in studies with cholesterol-fed or Watanabe rabbits. At present, it is impossible to interpret these in vivo data in a consistent manner. Recent studies on the Watanabe rabbit have indicated that the cholesterol-lowering effects of CSF-1 involve increased uptake of cholesterol by hepatic parenchymal cells and secretion in bile, without any evidence of accumulation in hepatic macrophages.23 Those authors speculated on the role of increased macrophage production of apo E and lipoprotein lipase. Preliminary favorable results of CSF-1 administration in individuals with familial hypercholesterolemia were reported.23
The question therefore remains whether the actions of CSF-1 are antiatherogenic or proatherogenic. At best, the sometimes observed decline in serum cholesterol may be beneficial, provided its consequence is excretion or degradation of cholesterol rather than its tissue deposition, as recent reports suggest.18 23 At worst, they might exacerbate conditions that are themselves intrinsically proatherogenic, rather than being independent initiators of atherogenesis.22 A study24 of op0/E0 mice, in which the CSF-1 and apo E genes are inactivated, associated a lack of CSF-1 with reduced atherogenesis and elevated plasma cholesterol and defined a gene dosage effect on these parameters. Although the mechanisms were not elucidated further and tissue cholesterol was not measured, those results taken together with our own suggest that in atherosclerotic lesions, high local concentrations of CSF-125 26 may contribute to formation of large, lipid-laden foam cells by the mechanisms we have observed. In the presence of local CSF-1, systemically administered CSF-1 may not necessarily provide an additional stimulus. These mechanisms will need to be resolved if CSF-1 is to fulfill its potential in therapy for hypercholesterolemia and atherosclerosis.
Selected Abbreviations and Acronyms
|Ac-LDL||=||acetylated low-density lipoprotein|
|CSF-1||=||human monocyte-specific colony-stimulating factor|
|DMEM||=||Dulbecco's modified Eagle medium|
|HPLC||=||high-performance liquid chromatography|
|nLDL||=||native low-density lipoprotein|
|Ox-LDL||=||oxidized low-density lipoprotein|
The generous gifts of recombinant human CSF-1 (Cetus Corp) and purified apo A-1 (Dr K.-A. Rye, University of Adelaide) and the excellent assistance of Merron Shutter during preliminary studies are gratefully acknowledged. We also thank Coralie Cornish (Immunology Unit, Heart Research Institute) for advice and assistance with cell imaging.
Metcalf D. Control of granulocytes and macrophages: molecular, cellular and clinical aspects. Science.. 1991;254:529-533.
Hume DA, Pavli P, Donahue RE, Fidler IJ. The effect of human recombinant macrophage colony stimulating factor (CSF-1) on the murine mononuclear phagocyte system in vivo. J Immunol.. 1988;141:3405-3409.
Mori N, Gotoda T, Ishibashi S, Shimano H, Harada K, Inaba T, Takaku F, Yazaki Y, Yamada N. Effects of human recombinant macrophage colony-stimulating factor on the secretion of lipoprotein lipase from macrophages. Arterioscler Thromb.. 1991;11:1315-1321.
Shimano H, Yamada N, Ishibashi S, Harada K, Matsumoto A, Mori N, Inaba T, Motoyoshi K, Itakura H, Takaku F. Human monocyte colony stimulating factor enhances the clearance of lipoproteins containing apo-B 100 via both low density lipoprotein receptor-dependent and -independent pathways in rabbits. J Biol Chem.. 1990;265:12869-12875.
Ishibashi S, Inaba T, Shimano H, Harada K, Inoue I, Mokuno H, Mori N, Gotoda T, Takaku F, Yamada N. Monocyte colony stimulating factor enhances uptake and degradation of acetylated low density lipoproteins and cholesterol esterification in human monocyte-derived macrophages. J Biol Chem.. 1990;265:14109-14117.
de Villiers WJS, Fraser IP, Hughes DA, Doyle AG, Gordon S. M-CSF selectively enhances macrophage scavenger receptor expression and function. J Exp Med.. 1994;180:705-709.
Kritharides L, Jessup W, Dean RT. Apolipoprotein A-I-mediated efflux of sterols from oxidized LDL-loaded macrophages. Arterioscler Thromb Vasc Biol.. 1995;15:276-289.
Labarca C, Paigen K. A simple, rapid and sensitive DNA assay procedure. Anal Biochem.. 1979;94:344-352.
Yui S, Sasaka T, Miyazaki A. Induction of murine macrophage growth by modified LDLs. Arterioscler Thromb.. 1993;13:331-337.
Ishibashi T, Yokoyama K, Shindo J, Hamazaki Y, Endo Y, Sato T, Takahashi S, Kawarabayasi Y, Shiomi M, Yamamoto T, Maruyama Y. Potent cholesterol-lowering effect by human granulocyte-macrophage colony-stimulating factor in rabbits: possible implications of enhancement of macrophage functions and an increase in mRNA for VLDL receptor. Arterioscler Thromb.. 1994;14:1534-1541.
Schaub RG, Bree MP, Hayes LL, Rudd MA, Rabbani L, Loscalzo J, Clinton SK. Recombinant human macrophage colony-stimulating factor reduces plasma cholesterol and carrageenan granuloma foam cell formation in Watanabe heritable hyperlipidemic rabbits. Arterioscler Thromb.. 1994;14:70-76.
Schaub RG, Donelly LH, Parker TS, Clinton SK, Garnick MB. CSF-1 and cholesterol clearance. In: Proceedings of the 10th MCDB/ISU Symposium: Colony Stimulating Factor 1—Molecular Mechanisms in Development and Disease. Ames, Iowa: In press.
Smith JD, Trogan E, Ginsberg M, Grigaux C, Tian J, Miyata M. Decreased atherosclerosis in mice deficient in both macrophage colony-stimulating factor (op) and apolipoprotein E. Proc Natl Acad Sci U S A.. 1995;92:8264-8268.
Shimada M, Inaba T, Shimano H, Gotoda T, Watanabe Y, Yamamoto K, Motoyushi K, Yazaki Y, Yamada N. Platelet-derived growth factor BB-dimer suppresses the expression of macrophage colony-stimulating factor in human vascular smooth muscle cells. J Biol Chem.. 1992;267:15455-15458.