Human Leukemia Inhibitory Factor Upregulates LDL Receptors on Liver Cells and Decreases Serum Cholesterol in the Cholesterol-Fed Rabbit
Abstract In a previous study, we found that the cytokine (human) leukemia inhibitory factor (hLIF) significantly reduced plasma cholesterol levels and the accumulation of lipid in aortic tissues of cholesterol-fed rabbits after 4 weeks of treatment. The mechanisms by which this occurs were investigated in the present study. This involved examining the effect of hLIF on (1) the level of plasma cholesterol at different times throughout the 4-week treatment and diet period; (2) smooth muscle cell (SMC) and macrophage-derived foam cell formation in vitro; and (3) LDL receptor expression and uptake in the human hepatoma cell line HepG2. At time zero, an osmotic minipump (2-mL capacity; infusion rate, 2.5 μL/h; 28 days) containing either hLIF (30 μg · kg−1 · d−1) or saline was inserted into the peritoneal cavity of New Zealand White rabbits (N=24). Rabbits were divided into four groups of six animals each. Group 1 received a normal diet/saline; group 2, a normal diet/hLIF; group 3, a 1% cholesterol diet/saline; and group 4, a 1% cholesterol diet/hLIF. hLIF had no effect on the plasma lipids or artery wall of group 2 rabbits (normal diet). However, in group 4 rabbits, plasma cholesterol levels and the percent surface area of thoracic aorta covered by fatty streaks was decreased by ≈30% and 80%, respectively, throughout all stages of the 4-week treatment period. In vitro, hLIF failed to prevent lipoprotein uptake by either SMCs or macrophages (foam cell formation) when the cells were exposed to β-VLDL for 24 hours. In contrast, hLIF (100 ng/mL) added to cultured human hepatoma HepG2 cells induced a twofold or threefold increase in intracellular lipid accumulation in the medium containing 10% lipoprotein-deficient serum or 10% fetal calf serum, respectively. This was accompanied by a significant non–dose-dependent increase in LDL receptor expression in hLIF-treated HepG2 cells incubated with LDL (20 μg/mL) when compared with controls (P<.05) incubated in control medium alone (P<.05). We suggest that the hLIF-induced lowering of plasma cholesterol and tissue cholesterol levels (inhibition of fatty streak formation) in the hyperlipidemic rabbit is due in part to upregulation of hepatic LDL receptors, with resultant increased clearance of lipoprotein-associated cholesterol from the circulation. There is an additional and as-yet-unknown mechanism acting at the level of the vessel wall that appears to be affecting the process of arterial cholesterol accumulation.
- Received December 19, 1995.
- Accepted August 6, 1996.
Leukemia inhibitory factor is a cytokine that was originally described in the early 1980s on the basis of its ability to induce terminal differentiation in some hemopoietic cell lines.1 2 Other properties that have been attributed to this cytokine include stimulation of proliferation of hemopoietic cells and myoblasts, inhibition of totipotent murine embryonic stem cell differentiation, stimulation of bone remodeling, promotion of neuronal survival and differentiation, hepatocyte acute-phase protein synthesis, and promotion of cachexia and lipoprotein metabolism (for a review, see References 3 and 43 4 ).
The clinical, epidemiological, and experimental evidence linking hyperlipidemia to the development of atherosclerosis is now well established.5 LDL is the major carrier of cholesterol in the blood and is most significantly associated with atherosclerotic plaque formation.6 However, β-VLDL, which appears in high concentrations in the plasma after ingestion of a cholesterol-rich meal, may also play an important role. This lipoprotein occurs in high concentrations in the blood of rabbits fed a cholesterol-enriched diet.7
Lipoprotein metabolism at the endothelium involves LPL-mediated degradation of VLDL by loss of triglycerols and apoproteins, thereby producing free fatty acids and VLDL remnants. These remnants are further processed by LPL to the more atherogenic form, LDL.8 Mori and coworkers9 have shown that the cytokine LIF can act as an adipocytic LPL inhibitor. In a previous study, we showed that plasma cholesterol concentration and aortic tissue cholesterol accumulation were significantly reduced by administration of hLIF to hyperlipidemic rabbits.7 This and other factors, such as the specific decrease in LDL after hLIF administration,7 led us to suggest that it may be hLIF’s inhibition of LPL activity, ie, prevention of VLDL hydrolysis at the endothelium, that plays a major role in the inhibition of atherosclerosis development in the rabbit. However, it was also suggested that other mechanisms may be involved.
In the present study, we examined the action of hLIF on lipoprotein uptake by SMCs and macrophages and the potential effect of hLIF on hepatic lipoprotein metabolism by examining LDL uptake and LDL receptor expression in vitro using the human hepatoma cell line HepG2.
Recombinant hLIF was a gift of Dr Nic Gough from the Australian Medical Research and Development (AMRAD) Corp (Victoria, Australia) and was prepared as previously described.10 11 12 hLIF was administered at a dose of 30 μg · kg−1 · d−1.7
Twenty-four male 18-week-old New Zealand White rabbits weighing ≈2.30 kg were obtained from commercial colonies. They were caged individually in a room with a 12-hour day/night cycle and fed ad libitum with water, fruit/vegetables, and standard rabbit chow. Hyperlipidemia was induced in 12 of the rabbits via a 1% (wt/wt) cholesterol-supplemented diet as described.7
All major surgical procedures were carried out under general anesthesia as described previously.7 An abdominal midline incision allowed intraperitoneal placement of an Alzet Osmotic Minipump (model 4ML2, ALZA Corp) containing either PBS (control) or hLIF.7 These pumps are specified to deliver 2.6 μL/h, with a mean volume of 2.17 mL and therefore are potentially active for 36 days.
Analysis of Hyperlipidemic Plasma
Blood collected weekly (2.5 mL) and by cardiac puncture at sacrifice was spun at 90g for 20 minutes at room temperature (25°C) to separate plasma from blood cells. Plasma cholesterol levels were determined by using an automated version of an enzymatic colorimetric method (CHOD-PAP)13 on the COBAS Bio-analyzer (Roche) and a commercially available kit (Monotest Cholesterol, Boehringer Mannheim).
Fatty Streak Formation
The thoracic aorta was excised at termination, fixed in 10% buffered formal saline (Conform, Medos Corp), longitudinally cut, flattened, and stained en face with oil red O as previously described.7
Aortic SMCs were obtained from thoracic and abdominal aortas of 9- to 12-week-old New Zealand White rabbits via enzymatic dispersion as previously described.14 Primary cells were cultured through four passages in medium 199 and 5% FCS (CSL). Macrophages were the J774A.1 cell line derived from the BALB/c mouse strain and were cultured in DMEM and 10% FCS. The human hepatoma cell line HepG2 was obtained from the American Type Culture Collection and cultured in RPMI and 10% FCS.
Foam Cell Formation
SMCs and macrophages were cultured in medium 199 and 5% LPDS or DMEM and 10% LPDS in the presence (preexposure) of different hLIF concentrations (50, 100, 200, or 500 ng/mL) and β-VLDL (100 μg/mL) in a final volume of 500 μL per well. The cells were kept in a CO2-gassed incubator at 37°C for 24 hours, after which the medium was removed and the cells trypsinized and resuspended in 50 μL of 5 mmol/L NaOH in 10-mL centrifuge tubes. The tubes were then sonicated for 1 hour, after which 5 μL of 20% sodium deoxycholate (BDH Chemicals) was added to each tube and the tubes incubated at room temperature for a minimum of 48 hours. The amount of β-VLDL–associated cholesterol accumulation was assessed by using a Monotest cholesterol kit and the COBAS Bio-analyzer.
HepG2 Lipoprotein (LDL) Uptake
HepG2 cells were cultured in experimental medium of either RPMI and 10% FCS or RPMI and 10% LPDS and then incubated with hLIF (1, 10, 50, 100, or 500 ng/mL) for 8 hours at 37°C in a 5% CO2 atmosphere. The cells were trypsinized and resuspended in 50 μL of 5 mmol/L NaOH in 10-mL centrifuge tubes and sonicated for 1 hour, after which 5 μL of 20% sodium deoxycholate (BDH Chemicals) was added to each tube; the tubes were then incubated at room temperature for a minimum of 48 hours. LDL-associated cholesterol accumulation was assessed by using a Monotest cholesterol kit and the COBAS Bio-analyzer as described above. Some cells were set aside for staining of intracellular lipid with oil red O.
HepG2 LDL Receptor Expression
Cells were plated on 10-mm-diameter glass coverslips in 24-well plates at a density of 4×104 cells per well and incubated at 37°C in an atmosphere of 5% CO2 for 24 hours to allow cell adhesion to the coverslips. Experimental medium (1 mL, RPMI and 10% FCS) containing various hLIF concentrations (1, 10, 50, 100, or 500 ng/mL) and/or LDL (20 μg/mL)15 was used, and the cells were incubated for another 8 hours. The medium was removed and the cells were washed in PBS (×3) before being fixed in chilled methanol (BDH Chemicals) for 10 minutes. Nonspecific binding was blocked with 5% sheep plasma (diluted in PBS, 100 μL per coverslip) for 20 minutes. Primary antibody C7 (100 μL, 1:50 dilution; mouse monoclonal, anti–LDL receptor antibody; Amersham) was added to each coverslip and incubated at room temperature for 2 hours in a humidifying chamber, after which 100 μL (1:100 dilution) of the FITC-conjugated secondary antibody (sheep anti-mouse immunoglobulin, Silenus Laboratories) was added and the cells incubated for 1 hour in the dark at room temperature in a humidifying chamber. The coverslips were then washed, removed from the wells, and mounted on microscope slides. Cells were examined under a Bio-Rad MRC 600 confocal photomicroscope for LDL receptor expression, with images displayed concurrently on a Sony monitor. Receptor expression was determined semiquantitatively by use of a color image analyzer, which detects fluorescent staining by measurement of pixels (number and intensity), thus determining the percentage fluorescence in a defined area.
Preparation of Lipoprotein and LPDS
β-VLDL and LDL were obtained from the plasma of rabbits that received the cholesterol-enriched diet by a differential flotation technique.16 Prior to use, the lipoprotein was sterilized by filtration through a series of acrodisc filters (0.8, 0.4, and 0.2 μm, Gelman) and then dialyzed in four changes of PBS with 1 mmol/L EDTA over a period of 48 hours. LPDS was prepared from normolipidemic rabbit plasma via differential flotation,16 dialyzed, filter sterilized as described above, and stored at −20°C. LPDS standardizes the basal number of lipoprotein receptors on the cells and also adds no extra lipoprotein to the medium in which the cells are incubated under experimental conditions.
For all analyses, paired t tests and Mann-Whitney U tests were performed on an IBM-compatible DX PC using SigmaStat software (Jandel Scientific). Significance level was set at P<.05.
Osmotic Pump Function
As shown in Fig 1⇓, the Alzet osmotic minipumps delivering hLIF operated efficiently over the 4-week experimental period exhibiting a linear delivery rate of 0.49±0.02 mL/wk. This rate is equivalent to ≈2.90±0.11 μL/h and is comparable to the factory specification of 2.6 μL/h.
Effect of hLIF on Plasma Cholesterol Levels
To examine the effect of hLIF (30 μg · kg−1 · d−1) on the development of hypercholesterolemia, group 3 and 4 rabbits were placed on a 1% cholesterol–enriched diet for 28 days without or with hLIF administration and blood samples collected every 7 days. As expected, group 3 rabbits displayed a significant increase in plasma cholesterol concentration compared with normolipidemic group 1 rabbits (P<.001), even after only 7 days (Fig 2⇓). Plasma cholesterol concentrations rose sharply, reaching maximum levels during the first 2-week period before plateauing over the remainder of the experimental period to ≈10.80 mmol cholesterol per liter of plasma (from 0.98±0.01 mmol cholesterol per liter of plasma). Cholesterol-fed rabbits receiving hLIF (group 4) displayed a significant 30% reduction in plasma cholesterol levels after the first week of treatment (P<.05; Fig 2⇓). Maximum levels of plasma cholesterol were reached by the second week of cholesterol feeding (7.44±0.87 mmol cholesterol per liter of plasma) and remained relatively constant over the following 2 weeks (7.53±0.73 and 8.09±0.69 mmol cholesterol per liter of plasma after weeks 3 and 4, respectively). Although cholesterol levels were significantly lowered with hLIF treatment than the levels observed in the control cholesterol-fed rabbits, they were still significantly greater than cholesterol levels seen in group 1 (normal diet) animals (P<.05). There was no difference in cholesterol levels of normolipidemic rabbits with or without hLIF (P<.05).
Fatty Streak Formation
Fatty lesions appeared in the thoracic aortas of group 3 rabbits within 7 days of cholesterol feeding and covered ≈18% of the luminal surface (Fig 3⇓). With continuous cholesterol feeding during the next 3 weeks, the area of luminal surface covered with fatty lesions increased to 39%, 55%, and 82%, respectively (Fig 3⇓). hLIF treatment in group 4 cholesterol-fed rabbits retarded the formation of aortic fatty streaks (Fig 4⇓), so that the extent of lesion coverage reached only ≈12% of the luminal surface by weeks 3 and 4.
Effect of hLIF on Lipoprotein Uptake by SMCs and Macrophages
The effect of hLIF on foam cell formation was examined in vitro using SMCs and macrophages, both of which accumulate lipid in the atherosclerotic lesion. Both SMCs and macrophages exposed to 100 μg/mL β-VLDL alone for 24 hours displayed significant increases in lipoprotein accumulation of 250% and 450%, respectively, compared with control cells (100%; P<.05). Both SMCs and macrophages exposed to β-VLDL after preexposure to hLIF concentrations of 50, 100, 200, or 500 ng/mL accumulated lipid to relatively the same degree as did cells exposed to β-VLDL alone (P>.05; control SMCs, 206.4±37.8 ng/mg protein; treated SMCs, 213.6±42.4 ng/mg protein; control macrophages, 413.2±67.8 ng/mg protein; treated macrophages, 422.8±34.4 ng/mg protein).
Effect of hLIF on LDL Uptake in HepG2 Cells
The effect of hLIF on LDL uptake in HepG2 cells after culture in media that upregulated (LPDS) or downregulated (FCS) the LDL receptors was examined. Cholesterol accumulation in HepG2 cells treated with 100 ng/mL hLIF was significantly higher than in control cells exposed to 100 μg/mL LDL only (control, P<.05). LDL uptake was increased threefold, raising intracellular cholesterol levels from 163±8.56 (control cells) to 505±9.34 ng/mg protein (Fig 5⇓). A significant twofold increase in LDL uptake also occurred in cells incubated in LPDS in the presence of 100 ng/mL hLIF (P<.05; from 280±11.34 [control] to 616±8.74 ng/mg protein). In both FCS and LPDS cultures, all other doses of hLIF (1, 10, 50, and 500 ng/mL) failed to induce a significant change in cholesterol accumulation compared with controls (P>.05).
Intracellular oil red O staining (Fig 6a⇓) illustrates the extent of lipid accumulation by cells incubated in RPMI and 10% FCS in the presence of 100 ng/mL hLIF compared with non–hLIF-treated cells incubated in the same medium (Fig 6b⇓).
Effect of hLIF on HepG2 LDL Receptor Expression
To determine whether the stimulation of LDL uptake was caused by an increased number of receptors, LDL receptors were visualized after an 8-hour incubation with the LDL receptor–specific antibody C7 and evaluated semiquantitatively using a color image analyzer. Cells incubated with LDL displayed a decrease in LDL receptor expression compared with cells cultured in control medium; however, the trend was not statistically significant (P>.05). Treatment of cells with 100 ng/mL hLIF in the presence of LDL (20 μg/mL) resulted in an increase in LDL receptor expression by ≈35% (P<.05) compared with receptor expression in control cells and cells exposed to LDL only. In accordance with the LDL accumulation data, 100 ng/mL hLIF was the only effective concentration to significantly increase LDL receptor expression (P<.05; Figs 7⇓ and 8⇓).
The effect of hLIF on plasma cholesterol levels and fatty streak development in the cholesterol-fed rabbit described in our earlier publication7 was reexamined in the present study to determine whether the effect occurred at all stages of hLIF treatment. Because parameters of LIF activity, eg, platelet number, peaked at 2 weeks and then decreased (leaving open the possibility that all LIF within the osmotic pumps had been administered in the early stages of the experiment), we wanted to verify that the effects of hLIF were long-lasting. Specifically, both plasma cholesterol level and fatty streak development were examined on a weekly basis over the 4-week experimental period. The present study showed that the osmotic minipumps delivered the prescribed volume throughout the entire 4-week treatment. It also showed that hLIF lowers serum cholesterol levels in cholesterol-fed rabbit within 1 week of administration and keeps these levels depressed by ≈30% for the entire 4 weeks. This result is in accord with results from our previous study7 that showed a significant decrease in plasma cholesterol level as a result of both lowered plasma LDL- and β-VLDL–associated cholesterol. Also consistent with this previous work was the ability of hLIF to inhibit fatty streak formation on the luminal surface of the thoracic aorta of hyperlipidemic rabbits, with ≈80% inhibition occurring throughout the entire treatment period.
Fatty streaks are commonly found in the intimal layer of arterial walls at anatomic sites that are prone to plaque development; however, most fatty streaks remain clinically insignificant or disappear.17 They are characterized as ill-defined, slightly raised regions consisting primarily of lipid-laden macrophages and SMCs (foam cells).18 In the present study, we have shown that hLIF has no direct inhibitory effect on lipoprotein (β-VLDL) uptake by these cells in vitro. Thus, the 80% inhibition of fatty streak formation in the thoracic aorta of cholesterol-fed rabbits by hLIF in vivo and the previously described 60% reduction in aortic arch cholesterol,7 while undoubtedly due in part to the 30% decrease in plasma cholesterol levels, must involve some mechanism other than a direct effect on macrophage or SMC lipoprotein receptors or metabolism. For example, circulating lipoproteins are converted to more atherogenic species after enzymatic interaction with LPL on the capillary endothelium of various tissues.19 LIF is a known LPL inhibitor9 and would therefore prevent hydrolysis of VLDL and thereby reduce the quantity of circulating LDL. LPL bound to the extracellular matrix can retain lipoproteins in the extravascular space, thereby making them more susceptible to oxidative modifications, which in turn facilitates their uptake by SMCs and macrophages via the scavenger receptor.20 21 Also, interaction between proteoglycans and lipoproteins in the vessel wall increases the susceptibility for lipid ingestion by macrophages. Recent studies by Bell and Carroll22 and Carroll et al23 show that LIF is a potent inhibitor of proteoglycan synthesis in articular cartilage. By inhibiting the activity of LPL and/or interfering with the biological properties of the extracellular matrix, hLIF may prevent the entrapment of lipoprotein in the vessel wall, thus preventing its modification and ingestion by SMCs and macrophages.
As stated above, the reduction in fatty streak formation in the presence of hLIF is undoubtedly related in part to the lower plasma cholesterol levels, which in turn suggests an increase in systemic (hepatic) regulation. In the present study, it was shown that hLIF possesses the ability to increase, in a non–dose-dependent manner, both LDL receptor expression and LDL uptake in vitro by the human hepatoma cell line HepG2. This indicates that lipoprotein clearance in vivo may be increased by hLIF through upregulation of hepatic LDL receptors (B/E receptors), which are naturally downregulated under hyperlipidemic conditions. Grove et al15 24 examined the effect of the LIF-related cytokine oncostatin M on lipoprotein metabolism by hepatic cells in vitro. They found that upregulation of LDL receptors by oncostatin M was of similar magnitude in the presence and absence of exogenous cholesterol, with an associated 80% increase in stimulation of LDL uptake.25 It was suggested, therefore, that oncostatin M upregulates these receptors and induces increased uptake of lipoprotein by HepG2 cells via a novel mechanism acting outside the sterol-dependent pathway. The same appears to be true for hLIF, because stimulation of receptor upregulation occurred in cultures exposed to hLIF both with and without LDL (20 μg/mL). Grove et al15 speculated that the upregulation mechanism might be oncostatin M–induced tyrosine kinase activation followed by generation of phospholipid-related second messengers. The exact role of tyrosine kinase activity in the stimulation of LDL receptor expression is unclear. However, examination of the 5′ flanking region of the LDL receptor gene upstream from the initiator Met codon revealed a putative Egr-1–binding motif, GAGGGGGCG, at position −328.26 Separate studies have shown that Egr-1 is a nuclear signal transducer that bridges early biochemical events generated by the cell membrane, with long-term changes in cell proliferation or differentiation resulting from regulation of transcription.27 Activation of protein kinase C can induce Egr-1 message.28 Consequently, it has been demonstrated that upregulation of lipoprotein receptors in HepG2 cells follows induction of Egr-1 by oncostatin M. The correlation between tyrosine kinase phosphorylation, Egr-1 induction, and LDL receptor regulation has implicated Egr-1 as a nuclear signal transducer that is utilized by oncostatin M to induce transcription of the LDL receptor gene.29 It is possible that hLIF operates the same way, as recent work by Schiemann and Nathanson30 has shown that protein kinase C–dependent activation of second-messenger protein kinases in 3T3-L1 cells is stimulated with LIF treatment in vitro. However, when comparing the effect of oncostatin M on HepG2 cell lipoprotein metabolism with other related cytokines, Liu et al29 found that neither interleukin-6 nor LIF (50 to 100 ng/mL) induced the Egr-1–binding motif and that each was “ineffective” with respect to stimulation of LDL uptake. These findings contrast with those of the present study, which demonstrate that hLIF treatment of HepG2 cells induces a 35% increase in LDL receptor expression and an increase in the uptake of lipoprotein-associated cholesterol of as much as 67%. It should be stressed, however, that the present study did not examine HepG2 LDL receptor expression at a molecular-genetic level. Nevertheless, it is important to note that Liu et al29 did observe a statistically significant 20% increase in lipoprotein uptake by HepG2 cells in the presence of LIF regardless of the fact that they did not observe LIF induction of the LDL receptor transcription signal transducer Egr-1. Further investigation at the molecular level is necessary to explain this phenomenon.
In conclusion, we suggest that the mechanism of hLIF-induced lowering of plasma cholesterol and inhibition of fatty streak formation in the hyperlipidemic rabbit incorporates upregulation of hepatic LDL receptors, with resultant increased clearance of lipoprotein-associated cholesterol from the circulation. There is an additional and as-yet-unknown mechanism acting at the level of the vessel wall that affects the process of arterial cholesterol accumulation.
Selected Abbreviations and Acronyms
|DMEM||=||Dulbecco’s modified Eagle’s medium|
|FCS||=||fetal calf serum|
|LIF||=||leukemia inhibitory factor|
|SMC||=||smooth muscle cell|
This work was supported by a grant from Australian Medical Research and Development (AMRAD Corp) and the Australian Research Council.
Tomida M, Yamamoto-Yamaguchi Y, Hozumi M. Purification of a factor inducing differentiation of mouse myeloid leukemic M1 cells from conditioned medium of mouse fibroblasts l929 cells. J Biol Chem.. 1984;259:10978-10986.
Tomida M. Induction of differentiation of WEHI-3B D+ leukemic cells tranfected with differentiation-stimulating factor/leukemia inhibitory factor receptor cDNA. Blood.. 1995;85:217-221.
Steinberg D. Current theories of the pathogenesis of atherosclerosis. In: Steinberg D, Olefsky JM, eds. Hypercholesterolemia and Atherosclerosis—Pathogenesis and Prevention. New York, NY: Churchill/Livingstone; 1987:5.
Auwerx J, Marzetta C, Hokanson J. Relationship between hepatic triglyceride lipase and low density lipoprotein characteristics. Arteriosclerosis.. 1989;9:319-325.
Moran CS, Campbell JH, Simmons DL, Campbell GR. Human leukemia inhibitory factor inhibits development of experimental atherosclerosis. Arterioscler Thromb.. 1994;14:1356-1363.
Gearing DP, Nicola NA, Metcalf D, Foote S, Wilson TA, Gough NM, Williams RL. Production of leukemia inhibitory factor (LIF) in Escherichia coli by a novel procedure and its use in maintaining embryonic stem (ES) cells in culture. Biotechnology.. 1989;7:1157-1166.
Katterman R, Jaworek D, Möller G, et al. Multicentre study of a new enzymatic method of cholesterol determination. J Clin Chem Clin Biochem.. 1984;22:254-261.
Grove RI, Mazzucco C, Radka SF, et al. Oncostatin M up-regulates low density lipoprotein receptors in HepG2 cells by a novel mechanism. J Biol Chem.. 1991;266:18194-18199.
Havel RJ, Eder HA, Bragdon JH. The distribution and chemical composition of ultracentrifugally separated lipoproteins in human serum. J Clin Invest.. 1955;34:1345-1353.
Stary HC, Blankenhorn DH, Chandler AB, Glagor S, Insull W Jr, Richardson M, Rosenfeld ME, Schaffer SA, Schwartz CJ, Wagner WD, Wissler RW. A definition of the intima of human arteries and its atherosclerosis-prone regions. Arteriosclerosis.. 1992;12:120-127.
Edwards IJ, Goldberg IJ, Parks JS, Xu H. Lipoprotein lipase enhances the interaction of low density lipoprotein with artery derived extracellular matrix proteoglycans. J Lipid Res.. 1993;34:1155-1161.
Grove RI, Mazzucco C, Allegretto N, et al. Macrophage-derived factors increase low density lipoprotein uptake and receptor number in cultured human liver cells. J Lipid Res.. 1991;32:1889-1897.
Liu J, Grove RI, Vestal RE. Oncostatin M activates low density lipoprotein receptor gene transcription in sterol-repressed liver cells. Cell Growth Diff.. 1994;5:1333-1338.
Sudhoff TC, Goldstein LJ, Brown MS, Russell DW. The LDL receptor gene: a mosaic of exons shared with different proteins. Science.. 1985;228:815-822.
Lemaire P, Vesque C, Schmitt J, et al. The serum-inducible mouse gene Krox-24 (Erg-1) encodes a sequence-specific transcriptional activator. Mol Cell Biol.. 1990;10:3456-3467.
Hallahan DE, Sukhatme VP, Sherman ML, et al. Protein kinase C mediates x-ray inducibility of nuclear signal transducers Egr-1 and c-jun. Proc Natl Acad Sci U S A.. 1991;266:10802-10806.
Liu J, Shoyab M, Grove RI. Induction of Egr-1 by oncostatin M precedes up-regulation of low density lipoprotein receptors in HepG2 cells. Cell Growth Diff.. 1993;4:611-616.
Schiemann WP, Nathanson NM. Involvement of protein kinase C during activation of the mitogen-activated protein kinase cascade by leukaemia inhibitory factor: evidence for participation of multiple signalling pathways. J Biol Chem.. 1994;269:6376-6382.