Oncostatin M Up-regulates Low Density Lipoprotein Receptors in HepG2 Cells by a Novel Mechanism*

Oncostatin M is a growth regulatory protein secreted by macrophages and activated T lymphocytes. In a hepatoma cell line (HepG2) the polypeptide very po-tently increased low density lipoprotein (LDL) uptake with an ECBo of 0.1-0.2 nM. The stimulation of LDL uptake was detectable by 2 h, was maximal by 8 h, and remained elevated through 20 h of oncostatin M incubation. In a similar fashion, oncostatin M also in- creased the number of cell surface LDL receptors by a mechanism that was inhibited by cycloheximide or the protein kinase C inhibitor H-7. Oncostatin M stimulation of LDL uptake and receptor protein occurred re- gardless of the state of cholesterol-dependent regulation of HepG2 LDL receptor (Le. cells incubated in medium containing lipoproteins responded to the same extent as did cells incubated in the absence of lipopro- teins). No significant effects were observed on sterol synthesis over 8 h or on DNA synthesis over 24 h. Oncostatin M induced


Oncostatin M Up-regulates Low Density Lipoprotein Receptors in HepG2 Cells by a Novel Mechanism*
(Received for publication, May 2, 1991) Robert I. Grovef, Charles E. Mazzucco, Susan F. Radka Oncostatin M is a growth regulatory protein secreted by macrophages and activated T lymphocytes. In a hepatoma cell line (HepG2) the polypeptide very potently increased low density lipoprotein (LDL) uptake with an ECBo of 0.1-0.2 nM. The stimulation of LDL uptake was detectable by 2 h, was maximal by 8 h, and remained elevated through 20 h of oncostatin M incubation. In a similar fashion, oncostatin M also increased the number of cell surface LDL receptors by a mechanism that was inhibited by cycloheximide or the protein kinase C inhibitor H-7. Oncostatin M stimulation of LDL uptake and receptor protein occurred regardless of the state of cholesterol-dependent regulation of HepG2 LDL receptor (Le. cells incubated in medium containing lipoproteins responded to the same extent as did cells incubated in the absence of lipoproteins). No significant effects were observed on sterol synthesis over 8 h or on DNA synthesis over 24 h.
Oncostatin M induced rapid alterations in HepG2 phospholipid metabolism. Within 5-15 min there was a 20-50% increase in incorporation of 32P into several classes of phospholipids, including the phosphoinositides. Radiolabeled diacylglycerol levels were elevated 20% by 2 min and nearly 50% by 15 min. In addition, the polypeptide induced rapid increases (within 1 min) in phosphorylation of HepG proteins on tyrosine residues. Stimulation of both phosphotyrosine and LDL receptor up-regulation by oncostatin M was decreased by the tyrosine kinase inhibitor genistein. We propose that oncostatin M up-regulates HepG2 LDL receptor expression by a mechanism that includes stimulation of a tyrosine kinase followed by generation of phospholipid-related second messengers.
The hepatic low density lipoprotein (LDL)' receptor plays a major role in cholesterol homeostasis (1-4). Circulating cholesterol in the form of LDL is removed from plasma by the highly specific LDL receptor and is internalized via receptor-mediated endocytosis. Upon degradation of the LDL particle in the lysosomal compartment, the LDL-derived cholesterol elevates the intracellular free cholesterol concentration. The elevated free cholesterol (or an oxysterol derivative) * The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. signals the hepatocyte to decrease transcription of the message of some of the key enzymes in the cholesterol biosynthetic pathway (5), resulting in a decrease in de nouo cholesterol synthesis. LDL receptor message and protein are also down-regulated by elevated intracellular cholesterol (5,6), resulting in decreased ability of the liver to remove additional LDL cholesterol from the plasma (3,7,8). A mechanism to up-regulate the LDL receptor independently would therefore be expected to yield an additional decrease in plasma cholesterol levels (4).
Mononuclear leukocytes secrete a number of polypeptides that modulate a wide variety of different cellular functions, including proliferation and differentiation (9). Recently, evidence that macrophages may affect cholesterol homeostasis has been emerging. Colony stimulating factors that activate macrophages cause a dramatic decrease in total serum cholesterol in humans (10) and in primates (11). A study in dietinduced hypercholesterolemic rats indicated that injection of a macrophage activator (zymosan) caused a significant decrease in total serum cholesterol (12). More recent evidence has suggested that endotoxin-stimulated macrophage-conditioned media were able to induce a significant increase in uptake of LDL and in LDL receptor number in a human liver cell line.2 The conditioned media also caused a decrease in cholesterol synthesis which did not appear to be related to the LDL receptor up-regulation. Further analysis of the conditioned media indicated the majority of the LDL receptor up-regulatory activity was caused by oncostatin M (13).
Originally discovered in conditioned media from a human macrophage cell line, oncostatin M is a novel glycoprotein that has growth regulatory properties. Depending on the cell type, the polypeptide may or may not alter proliferation (14). Oncostatin M has a molecular mass of 28 kDa and binds to high affinity cell surface receptors of approximately 150 kDa as determined by cross-linking studies (15). In the current study, we have characterized the interaction of oncostatin M with HepGZ cells and report on the mechanism by which it up-regulates LDL receptors.

MATERIALS AND METHODS
Cells and Reagents-The hepatoma cell line HepG2 was obtained from ATCC (Bethesda, MD) and cultured in RPMI medium supplemented with 10% fetal bovine serum. Cell culture reagents were obtained from GIBCO except for lipoprotein-deficient serum and the fluorescent DiI-LDL (Biotechnologies Inc., Boston). Genistein was from ICN. Oncostatin M and monoclonal antibodies were from Oncogen (Seattle). The OM1 and OM2 monoclonal antibodies are directed against epitopes on native oncostatin M; OM1 did not neutralize the effects of oncostatin M in a growth inhibition assay whereas OM2 completely abrogated oncostatin "induced activity:' Rabbit anti-phosphotyrosine antisera were raised and purified essentially as described by Kamps and Sefton (16) and were a gift from Dr. J. Ledbetter (Oncogen). On occasion, a commercial anti-phosphotyrosine antibody raised against tyrosine phosphorylated v-ab1 peptide (Pharmingen) was also used. Radioisotopes were purchaed from Du Pont-New England Nuclear. All other reagents were from Sigma.
LDL Uptake Assay-LDL uptake in HepG2 cells cultured in 12well plates (Costar) at 2 X 105/well was assayed by the fluorescent LDL method (17)(18)(19) using a fluorescence microscope and video camera (SIT 66) attached to a computer (Laser 25) with JAVA image analysis software (Jandel Scientific, CA) as reported previously.' This method has been shown to be comparable to "'I-LDL methods and therefore to reflect LDL receptor activity and regulation reliably (17)(18)(19). Specifically, 2 pg/ml DiI-LDL in serum-free media was added to HepG2 monolayers for 2 h. The monolayers were washed three times to remove free DiI-LDL and fixed with 4% formalin in phosphatebuffered saline. Accumulated fluorescence was measured by quantitating the average intensity of 25-50 cells. Three different fields in each of duplicate or triplicate monolayers were quantitated and averaged. The amount of LDL protein taken up was determined as follows. Cell monolayers (unfixed) were solubilized with 0.1% sodium dodecyl sulfate, transferred into glass tubes, and the lipids (and DiI) were extracted into chloroform using the Bligh-Dyer extraction procedure (20). Fluorescence in the chloroform layer was quantitated with a fluorescence spectrophotometer (Perkin-Elmer) and compared with a standard curve generated by extracting DiI from known quantities of DiI-LDL. Protein was estimated using Coomassie Blue kits (Pierce Chemical Co.).
LDL Receptor ELISA-The amount of surface LDL receptor was estimated with an ELISA that employed the LDL receptor antibody C-7 (21). The HepG2 monolayers were fixed with 4% formalin and blocked for 2 h with phosphate-buffered saline (PBS) containing 3% (w/v) bovine serum albumin. The blocking solution was replaced with PBS containing bovine serum albumin (0.5%) and 5 pg/ml C-7 and incubated at 25 "C for 2 h. The monolayers were washed extensively to remove excess antibody with PBS, 0.5% bovine serum albumin, 0.05% Triton X-100 and then incubated for 1 h in PBS containing 20 pg/ml peroxidase-conjugated goat anti-mouse IgG (Cooper Biochemicals). Excess antibody was removed, and peroxidase substrate (0-diphenylenediamine (4 mg/ml) and hydrogen peroxide (0.00012%) in 0.1 M citrate phosphate, pH 5.0) was added to the monolayers. Color development was stopped after approximately 10 min by the addition of HC1 (2.5 N, final concentration). Optical density was measured at 490 nm in a microtiter plate reader (Molecular Devices).
Sterol Synthesis Assay-Cholesterol synthesis was estimated from the sterol synthesis method of Mosley et al. (22). Briefly, cell cultures were incubated with 2 pCi/ml [14C)acetate for 1 h, washed once with PBS, and then incubated with 1.5 N NaOH for 10 min before transfer into glass test tubes. The lysate was saponified at 70 "C for 1.5 h. The unsaponified lipids were extracted with petroleum ether, and the upper phase was dried under a stream of NS at 45 "C. The lipids were first resuspended with ethano1:acetone (1:l) and then the sterols precipitated by the addition of 1% digitonin. The precipitates were washed twice with acetone and counted in a scintillation counter.
DNA Synthesis Assay-HepGP monolayers were incubated with [,"H]thymidine at 1 pCi/ml for 1 h. The monolayers were washed two times with phosphate-buffered saline and treated with ice-cold 5% trichloroacetic acid. After three washes with PBS the precipitates were solubilized with 0.1% sodium dodecyl sulfate and counted in a scintillation counter.
Lipid Analyses-Lipids from cells radiolabeled with ["P]orthophosphate (1 Ci/mmol) or with ['H]glycerol (200 mCi/mmol) were extracted by the method of Bligh and Dyer (20) as modified previously (23). Extracts, dried under a stream of nitrogen, were solubilized with 100 p1 of chloroform and spotted on thin-layer chromatography (TLC) plates (Silica Gel 60, Merck). The phospholipids were resolved either with a solvent system designed to separate the inositides (propanol, 4.3 M ammonium hydroxide, 65:35, v/v) or with a system designed to separate phosphatidylcholine and phosphatidylethanolamine (chloroform:methanol:acetic acidwater 50:35:8:1, v/v). Neutral lipids were resolved with benzene:ethyl acetate (8020, v/v). The radiolabeled phospholipids were identified as spots on autoradiographs that comigrated with authentic standards (detected with iodine vapors) and were quantitated by scraping the spots into vials and counting in a scintillation counter. Neutral lipid standards (diacylglycerol and triacylglycerol) were added to the [3H]glycerol-labeled lipid extracts before chromatography and were identified with iodine vapors. The spots were scraped and analyzed as outlined above.
Phosphotyrosine Analysis-HepG2 cells were plated into six-well plates at 4 X lo5 cells/well. After 48 h (about 70% confluence), the cells were incubated with various agents for the indicated time, rapidly washed once with ice-cold PBS, and lysed with 100 p1 of SDS sample buffer containing 50 mM 2-mercaptoethanol and 500 p~ sodium vanadate. The buffer was mixed rapidly over the entire well before being removed and heated to 100 "C for 15 min. The samples (25 pl) were separated by SDS-polyacrylamide gel electrophoresis on 10-20% gradient gels and transferred to Hybond C Extra membranes (Amersham Corp.). The membranes were blocked with PBS containing 5% bovine serum albumin, 1% ovalbumin, and 1 mM orthovanadate for 16 h at 4 "C. Proteins containing phosphotyrosine were detected using an affinity-purified rabbit anti-phosphotyrosine antibody followed by '2'II-labeled protein A (ICN).
Oncostatin M Receptor Cross-linking-Either H2981 cells (lung adenocarcinoma) or HepG2 cells were plated in 24-well tissue culture plates (Costar) and grown to confluence (approximately 2 X lo5 cells/ well). After two washes with Dulbecco's modified Eagle's medium supplementedwith 50 mM BES (pH 6.8), 0.1% bovine serum albumin, and 10% fetal bovine serum (binding buffer), the cells were incubated for 3 h at 25 "C with 200 pl of binding buffer containing 4 nM I2'Ioncostatin M either with or without 400 nM unlabeled oncostatin M.
To cross-link, disuccinyl suberate (0.4 mM) was added to the cells for 20 min at 25 "C. The monolayers were washed and the cells lysed with 2% SDS-electrophoresis sample buffer containing 5% 2-mercaptoethanol, boiled for 5 min, and then separated by polyacrylamide gel electrophoresis on 7.5% gels. Gels were dried and autoradiographed on Kodak X-AR5 film.

RESULTS
The effects of oncostatin M on LDL uptake in HepG2 cells cultured in media that up-regulated (LPDS) or down-regulated (fetal bovine serum) LDL receptors were investigated. Oncostatin M, at 100 ng/ml (4 nM) stimulated LDL uptake approximately 60% over the fetal bovine serum controls (Fig.  1). This stimulation was similar in magnitude to the upregulation induced in HepG2 cells by the LPDS control medium. However an increase of about 55% in LDL uptake still occurred when oncostatin M was added to cells cultured in LPDS (Fig. 1). The time and concentration dependences of the oncostatin M effect are shown in Fig. 2. The earliest detectable response occurred within 2 h, and the peak response at 8 h was maintained over 20 h of incubation ( Fig.   2 A ) . The concentration of oncostatin M required to give a half-maximal increase was 3-5 ng/ml(O.l-O.2 nM); at concentrations of 50 ng/ml or higher a maximum stimulation of 70% was achieved (Fig. 2B).
To determine whether the LDL uptake stimulation was caused by increased receptor number, we employed an LDL receptor ELISA using the C-7 antibody (21).   (24). To determine whether the oncostatin M effects on LDL receptor were linked to cholesterol synthesis inhibition, we studied the incorporation of radiolabeled acetate in HepG2 cell sterols. For the first 8 h, when LDL uptake had already peaked, no effect on cholesterol synthesis was detected. After 20 h, oncostatin M inhibited cholesterol synthesis by 60% (Table I).
It has been reported that certain peptide growth factors can increase LDL receptor by stimulating cell proliferation (25).
In contrast, oncostatin M, a t concentrations that gave maximal up-regulation of the LDL receptor, slightly inhibited ["HI thymidine incorporation a t 24 h (Table I).
The effects of oncostatin M on HepG2 cells suggested that these cells possess a specific oncostatin M receptor. When

TARLE I Effect of oncostatin M on HepGP sterol and DNA s.vnthesis
For sterol synthesis studies, cells incubated in LPDS media were treated with 100 ng/ml oncostatin M for the indicated times. One h before the end of the experiments, ["Clacetate (2 pCi/ml) was added, and incorporation of radiolabel into nonsaponifiable lipids was determined as described under "Materials and Methods." DNA synthesis was estimated by incorporation of radiolabeled thymidine in cells Since HepG2 cells appeared to possess a specific oncostatin M receptor, we examined the effects of the protein on second messenger pathways. Within 2 min after the addition of oncostatin M, a 20% increase in diacylglycerol was detected (Fig. 4A). The stimulation reached 50% by 30 min and remained elevated for a t least 60 min. T o investigate whether the elevated diacylglycerol arose from induced degradation of phospholipids, we studied the effects of oncostatin M on phospholipid metabolism. Within 15 min the polypeptide increased 'tzP incorporation into phosphatidylcholine by 50% and into the phosphoinositides by 25% (Fig. 4R).
The oncostatin "induced increase in diacylglycerol suggested that the signal transduction pathway involved activation of protein kinase C. Treatment of HepG2 cells with phorbol myristate acetate (50 nM) for 4 h produced a 40% increase in LDL uptake, and this increase was inhibited by pretreatment of the cells with the protein kinase C inhibitor H-7. Furthermore, H-7 completely blocked the stimulation of HepG2 LDL uptake induced by 4 h of incubation with oncostatin M (see Table 111). Recently, it has been reported that diacylglycerol levels can be elevated by a mechanism that includes tyrosine kinasemediated activation of phospholipase C-71 (26). The possibility that a tyrosine kinase could be activated by oncostatin M was investigated. Oncostatin M induced increases in tyrosine phosphorylation of several proteins, including a 165/180-kDa doublet and two other heavier bands a t 110 and 125 kDa ( Fig. 5C and Fig. 6). Phosphorylation could be detected a t low concentrations of 1-5 ng/ml. At optimal concentrations (25-50 ng/ml), phosphorylation of the doublet a t 165/180 kDa increased 5-10-fold over unstimulated cells whereas phosphotyrosine content in the bands a t 110 and 125 kDa increased 2-3-fold (Fig. 5, A and C). Phosphorylation of the 165/180 kDa doublet occurred very rapidly. It was detectable after 1 min (the earliest time point taken), peaked between 3-5 min, and then decreased over the next 25-30 min (Fig. 5, B and  C ) . Phosphorylation of other bands (80 and 95 kDa) remained elevated for a t least 30 min (Fig. 5C). As indicated under "Materials and Methods," phosphorylation was measured by Western blotting techniques using anti-phosphotyrosine antibodies. Phenylphosphate or phosphotyrosine (10 mM) completely blocked binding of the antibody to nitrocellulose; phosphoserine had no effect (data not shown).
The effects of other growth factors known to stimulate tyrosine kinase activity on both tyrosine phosphorylation and LDL receptors in HepG2 cells were assessed ( Fig. 6 and Table  11). Although insulin (10 pg/ml) induced the phosphorylation of a new band a t about 160 kDa and enhanced phosphorylation of the bands at 110 and 125 kDa, it had only minor effects on LDL uptake. EGF (100 ng/ml) stimulated the added to HepG2 monolayers for the indicated times before the cells were analyzed. The data were derived from the experiment shown in C. C, time dependence. This is a picture of the autoradiograph from which the data in R were derived. 10 pg/ml) was added to HepG2 monolayers for 10 min hefore the cells were lysed. Phosphotyrosine levels on HepG2 proteins were analyzed with Western blot techniques as described under "Materials and Methods." phosphorylation of a band at 175 kDa and enhanced phosphorylation of the 110-and 125-kDa bands. EGF also stimulated LDL uptake by approximately 40%. Treatment with PDGF (10 ng/ml) gave no observable increase in tyrosine phosphorylation or LDL uptake. EGF or insulin, but not PDGF, gave rise to small increases in DNA synthesis (Table  11).

FIG
Since the results suggested an early role for tyrosine kinase activity in oncostatin "induced changes, the effects of the tyrosine kinase inhibitor genistein (27) were investigated. Pretreatment of HepG2 cells with genistein inhibited in-  creases in tyrosine phosphorylation, in LDL uptake, and diacylglycerol stimulated by oncostatin M (Table 111). The effects of genistein were not caused by general cytotoxicity since the inhibition was reversed by washing genistein from the cells prior to stimulation and also because the inhibitor had no effect on HepG2 DNA synthesis over 24 h. In addition, genistein only slightly inhibited LDL uptake stimulated by treatment of the cells with phorbol myristate acetate.
In contrast to these results, the phosphotyrosine phosphatase inhibitor sodium orthovanadate enhanced the effect of oncostatin M on LDL uptake (Table 111) although this required 2-4 h of preincubation with vanadate. In control HepG2 cells treated with sodium vanadate, tyrosine phosphorylation was enhanced greatly. The vanadate effect was so marked, however, that it was difficult to identify oncostatin "induced tyrosine phosphorylation in the presence of vanadate. Alone, the phosphatase inhibitor only partially increased LDL uptake (Table 111).

DISCUSSION
Agents that increase LDL receptors by the sterol regulatory pathway (i.e. competitive inhibitors of cholesterol synthetic enzymes) first act to decrease the concentration of intrahepatocyte cholesterol. This reduction in sterols causes derepression of the LDL receptor gene, and transcription occurs (24). In the present study we show that the polypeptide oncostatin M is recognized by specific receptors on the surface of a liver cell line and rapidly up-regulates LDL receptors. Oncostatin M does not appear to act via the sterol-dependent pathway since up-regulations of similar magnitude were induced by the factor both in the presence and absence of exogenous cholesterol. In addition, both LDL uptake and the number of LDL receptors were maximally up-regulated (4-8 h) whereas effects on cholesterol synthesis were not yet detected. Thus oncostatin M apparently up-regulates the expression of the LDL receptors by a novel mechanism.
We have shown that within 1-2 min after treatment of HepG2 cells with oncostatin M there is a large increase in the levels of phosphotyrosine of several proteins. Pretreatment of the cells with the phosphotyrosine kinase inhibitor genistein partially blocked both tyrosine phosphorylation and LDL receptor up-regulation induced by the peptide. Additionally the phosphotyrosine phosphatase inhibitor vanadate increased both the levels of tyrosine phosphorylation and oncostatin "induced LDL uptake. These results suggest a role for oncostatin "induced tyrosine kinase activation in the LDL receptor up-regulation mechanism.
In addition to tyrosine kinase, protein kinase C may be involved in the pathway in which oncostatin M induces LDL receptor up-regulation. It has been reported that tyrosine kinase activation by other peptide factors leads to the activation of phospholipase c-yl by phosphorylation of a tyrosine residue (26). Our data support a similar mechanism for oncostatin M. The rapid rise in diacylglycerol levels and the phospholipid metabolism alterations are consistent with phospholipase C activation. Protein kinase C involvement is likely, based on the increase in diacylglycerol, the finding that the protein kinase C inhibitor (H-7) prevented the up-regulation, and the ability of the phorbol ester activator of protein kinase C to increase LDL uptake. This last result extends the observation by others that protein kinase C activators increase the message levels for the LDL receptor in HepG2 cells (29). From these findings it is tempting to speculate that oncostatin M stimulates a tyrosine kinase which then phosphorylates and activates a phospholipase. The inhibition of production of diacylglycerol by pretreatment with genistein, but the lack of effect of genistein on phorbol myristate acetate induction of LDL receptors, indicates that tyrosine kinase activation precedes the generation of diacylglycerol. Diacylglycerol activates protein kinase C , which in turn induces LDL receptor expression, presumably by stimulation of transcription. Protein kinase C activation may also explain the delayed onset of cholesterol synthesis inhibition which occurs by 20 h of incubation. It has been reported that phosphorylation of the rate-limiting enzyme for cholesterol synthesis (hydroxymethylglutaryl-coenzyme A reductase; EC 1.1.1.88) by protein kinase C inhibits its activity (28).
The effects of several factors that are known to stimulate tyrosine kinase activation were compared in HepG2 cells. Epidermal growth factor (EGF) stimulated the phosphorylation of peptides of 110,125, and 175 kDa and caused increases in LDL uptake. Insulin increased tyrosine phosphorylation but had no effect on LDL uptake. Since not all peptides that stimulate tyrosine kinase activity up-regulated the LDL receptor, oncostatin M must stimulate the activity of a specific kinase or the phosphorylation of a specific substrate that is involved in regulating the levels of the LDL receptor. The identities of the oncostatin "stimulated tyrosine kinase and its substrates are under investigation. Possible candidates are the oncostatin M receptor itself and phospholipase (2-71. The discovery that a polypeptide secreted by macrophages (14) has effects on liver cell lipoprotein and cholesterol metabolism is intriguing and suggests a possible connection between the immune system and cholesterol homeostasis. Reports in the literature indicate that activators of macrophages can dramatically decrease LDL levels in both animals (11, 12) and humans (10). The novel effects of oncostatin M on HepG2 cells indicate a possible way that hepatocyte LDL receptor expression can be regulated by macrophages. It is interesting to note that the liver has a large number of macrophage-derived Kupffer cells residing in close proximity to the parenchymal cells where secretory factors would be especially accessible to the hepatocyte.
Although regulation of oncostatin M secretion is not understood, the presence of oncostatin M in media from macrophages stimulated with endotoxin (13) indicates that endotoxin can increase oncostatin M production. Identification of other stimulators of oncostatin M secretion by macrophages is currently in progress.