Atherosclerosis, 96 (1992) 203-214 0 1992 Elsevier Scientific Publishers Ireland, Ltd. All rights reserved. 0021-9150/92/$05.00

203

Printed and Published in Ireland

ATHERO 04907

Cytokine regulation of macrophage apo E secretion: opposing effects of GM-CSF and TGF-p Steven H. Zuckerman, Glenn F. Evans and Laura O’Neal Lilly Research Labs, Indianapolis, IN 46285, 317-276-4385 (USA)

(Received 26 February, 1992) (Revised, received 11June, 1992) (Accepted 22 July, 1992)

Summary Biosynthesis of apolipoprotein (apo) E has been previously demonstrated to be regulated in macrophages by intracellular free cholesterol levels as well as by macrophage activating factors. In this report, the regulation of apo E secretion by cytokines detected within atherosclerotic lesions has been investigated. Granulocyte macrophage-colony stimulating factor (GM-CSF) stimulated macrophages had a 3-j-fold reduction in apo E secretion, comparable to that observed for gamma interferon (IFNy), while tumor necrosis factor alpha (TNFa) and interleukin 1 beta (IL-IO) resulted in a 2-fold decrease. In contrast to the reduction in apo E secretion by these cytokines, transforming growth factor beta (TGF-/3) stimulated macrophages secreted 3-fold greater amounts of apo E than controls. The reduced secretion of apo E by GM-CSF was reversible, heat labile, dose dependent, maximal 48 h after cytokine exposure and was coincident with an increase in Iibronectin secretion. The opposing effects of GM-CSF and TGF-/3 on apo E secretion were consistent with similar changes detected in apo E mRNA levels. Cytokine effects on apo E secretion in cholesterol loaded macrophages were also investigated and found to be similar to the nonloaded cells with GM-CSF decreasing and TGF-fl increasing apo E secretion. The observed differences in apo E secretion did not correlate with any significant changes in either cellular cholesterol distribution in the non-cholesterol loaded macrophages or in basal ACAT activity. In addition to changes in apo E secretion, cytokine treated macrophages pulsed with [ t4C]oleate and acetylated LDL for 2-6 h had a 2-fold increase (GM-CSF) or decrease (TGF-/3) in cholesterol esterilication. Therefore, GM-CSF and TGF-0 mediated changes in apo E secretion may occur through a mechanism independent of changes in cellular free cholesterol levels. These results suggest that cytokines expressed within an atheroma may play an important role in the modulation of macrophage mediated reverse cholesterol transport.

Key words: Apolipoprotein

E; Reverse cholesterol transport; Macrophage; Cytokines; ACAT

Correspondence to: Steven H. Zuckeman, Lilly Research Labs, Indianapolis, IN 46285, 317-276-4385, USA.

Introduction

Apolipoprotein E, a 34-kDa glycoprotein, is associated with multiple serum lipoprotein fractions including very low density lipoproteins (VLDL), chylomicrons, remnants and a subpopulation of (high density lipoproteins (HDLs) [l-3]. Approximately 20-40% of the apo E synthesized is from extra hepatic sites including brain, heart, adrenals, smooth muscle, adipocytes, macrophages and other mesenchymal and nonmesenchymal sources [4-61. The tissue distribution for apo E synthesis is unique among the serum apolipoproteins and may reflect the diverse functions associated with apo E. Apo E is involved in reverse cholesterol transport by shunting apo Econtaining particles back to the liver with uptake by the apo B, E receptor (LDL receptor) and a distinct hepatic apo E receptor (remnant receptor) [l-3]. In addition to facilitating lipid transport and metabolism, apo E has been reported to have immunosuppressive effects on lymphocytes [7,8], to interact with matrix components [9,10] and has been detected in high quantities within atherosclerotic plaques and with neurofibrillary tangles and amyloid plaques in brain biopsies from Alzheimer patients [ 11,121. The broad tissue distribution of apo E has necessitated a greater understanding of the mechanisms involved in the regulation of apo E synthesis. Macrophages synthesize significant quantities of apo E and have served as a model for understanding apo E synthesis and secretion [ 13- 151. Results from numerous laboratories have demonstrated increased synthesis of apo E by macrophages following cholesterol loading. Experiments with acetylated LDL, 25-hydroxycholesterol and with the Sandoz acyl CoA:cholesterol acyltransferase, ACAT, (EC 2.3.1.26) inhibitor SA58035 as well as HDL coculture experiments, suggested that increased apo E synthesis was related to increased levels of intracellular unesteritied cholesterol [ 16-221. In distinction to these stimulatory effects, macrophages activated in vivo with pyran copolymer or Bacillus Calmette Guerin (BCG), or in vitro with lipopolysaccharide (LPS) or IFNy have decreased apo E secretion [23-251. The effect of other cytokines which modulate macrophage ef-

fector functions on apo E biosynthesis has not been determined. The demonstration of TGF-fl within atherosclerotic lesions [26] as well as the effects of colony stimulating factors on macrophage cholesterol metabolism [27] necessitated an analysis of their role in reverse cholesterol transport by regulating apo E synthesis. The present study demonstrates that GM-CSF and TGF-/3 have opposing effects on apo E, with the former decreasing and the latter increasing apo E secretion. In both instances, these effects were not due to changes in cellular free cholesterol and were apparent in both non-loaded and cholesterolloaded macrophages. Cytokine mediated changes in apo E secretion were dose dependent and maximal by 48 h after cytokine exposure. These results suggest that the relative concentrations of GMCSF and TGF-/3 within the atheroma microenvironment can significantly effect macrophage apo E secretion and reverse cholesterol transport. Materials and Methods Ceil culture

Peritoneal macrophages were obtained from female Balb/C mice 3-5 days after an intraperitoneal injection of thioglycollate broth and were cultured in 24 well plates (Costar, Cambridge, MA) at 2 x lo5 cells per well in RPM1 1640 media supplemented with 2% fetal calf serum. Following overnight incubation macrophages were stimulated with lipopolysaccharide (LPS) (055:B5, Difco, Detroit, MI), phorbol myristate acetate (PMA) human interleukin 1 (IL-l@, or murine tumor necrosis factor alpha (TNFar) (provided by Dr. L. Bobbitt, Lilly Research Labs, Indianapolis, IN), murine granulocyte macrophagecolony stimulating factor, GM-CSF (Amgen, Thousand Oaks, CA), murine gamma interferon (IFNy) (Amgen), or porcine transforming growth factor beta (TGF-&) (R & D systems, Minneapolis, MN) for the designated intervals at the specified concentrations. Biosynthetic labeling, immunoprecipitation and electrophoresis

Macrophages incubated with cytokines were labeled for the final 5 h of culture in methionine free media with 1% fetal calf sera and 60 &X/ml of

205

[ 35S]methionine (Amersham, Arlington Heights, IL). Supernatants were collected and an aliquot was quantitated for TCA precipitable counts. All [ 35S]methionine labeled supematants were normalized to approximately lo6 counts/min of TCA precipitable protein and diluted to a final volume of 0.75 ml. These normalized supematants were then incubated with both goat anti human apo E (Chemicon, Temecula, CA) and rabbit anti mouse fibronectin (Chemicon) at final dilutions of l/100 or normal goat and rabbit sera at similar dilutions overnight at 4°C. Immunoprecipitates were obtained using the appropriate secondary antibody coupled to agarose (Sigma, St. Louis, MO), washed with PBS containing 1% Triton X-100, 0.1% SDS, 0.5% sodium deoxycholate and electrophoresed on lo-20% gradient polyacrylamide gels under denaturing conditions. Gels were fluorographed, dried onto nitrocellulose filters and autoradiographed. Band intensities were evaluated by quantifying immunospecific counts directly from the nitrocellulose with a Betagen model 603 blot analyzer (Waltham, MA) or by relative integrated optical density units from autoradiograms using a Bio Image scanner and Visage software (Bio Image Products, Ann Arbor, MI). RNA analysis

Control or cytokine treated cells were washed at intervals from O-7 days, lysed in 5 M guanidine thiocyanate and centrifuged through 5.7 M CsCl. RNA pellets were phenol-chloroform extracted, ethanol precipitated, dissolved in water, applied to nitrocellulose and hybridized with a 5 ’ end labeled antisense probe specific for murine apo E (CAGGCGTATTTGCTGGGTCTGTTCCTCCAT) [28] or a similarly labeled oligo d(T) probe (Promega Biotech, Madison, WI) using procedures previously described [29]. The extent of oligonucleotide hybridization to each slot was quantitated by Betagen prior to autoradiography. Macrophage A CA T assay

A macrophage based ACAT assay was developed using modifications of the ACAT assay described by Goldstein et al. [30]. Briefly, macrophage cultures were incubated for 48 h with GM-CSF or TGF-@ prior to addition of 50 pg/ml

of acetylated LDL (Biomedical Technologies, Inc., Stoughton, MA) and 0.4 &i/ml of [l-‘4C]oleic acid (Amersham). At 2, 4 or 6 h after [14C]oleate addition triplicate wells were washed, scraped in PBS and cells were chloroform-methanol extracted. Non-cholesterol loaded macrophages were incubated with cytokines for 48 h and pulsed with [14C]oleate for 2 and 4 h prior to lysis. The organic phase from macrophage cultures incubated in the presence or absence of acetylated LDL was concentrated and chromatographed on Whatman (Whatman Ltd., Maidstone, UK) silica coated flexible TLC plates using a petroleum ether/ethyl ether/glacial acetic acid (90: 10: 1) solvent system. Cholesterol ester and triglyceride spots were identified by iodine vapor and plates were subsequently autoradiographed and cholesterol ester and triglyceride spots were cut and quantitated by liquid scintillation. Cellular cholesterol distribution

Control or macrophages incubated with GMCSF or TGF-fi for 24 h were labeled with 1 pCi/ml of [4-‘4C]cholesterol (Amersham) for an additional 24 h in the presence of cytokines. Macrophage wells in triplicate were then washed and incubated in RPM1 1640 containing 2 mg/ml of bovine serum albumin for 24 h to permit isotopic equilibration. Wells were then washed and processed as described for the ACAT assay. Esterified and free cholesterol were resolved in the cell lysates by thin layer chromatography identified on autoradiograms and quantitated by liquid scintillation counting. Results The effect of TNF, IL-l, IFN?, GM-CSF and TGF-fl on macrophage secretion of apo E was evaluated after a 48-h preincubation with cytokines followed by a 5-h [35S]methionine pulse. As evident in Fig. 1, incubation of macrophages with 10 rig/ml of the above cytokines or with 100 @ml of LPS or PMA, resulted in the modulation of apo E secretion. Both GM-CSF and IFNy reduced macrophage secretion of apo E approximately 5-fold while TNF and IL-10 had a more modest effect. TGF-/.I treated macrophages

206

in contrast secreted 2.5fold greater amounts of apo E. The down regulation of apo E secretion by GMCSF was comparable to that observed for IFNy and LPS, both of which have been demonstrated to reduce apo E secretion [23,25] and served in the present study as a positive control. Further studies, therefore, focused on GM-CSF and TGF/3 effects on apo E secretion as examples of negative and positive cytokine regulators. The optimal concentrations and kinetics for apo E regulation by GM-CSF and TGF-/3 were then established and compared with the regulation of tibronectin secretion in the same cultures. The suppressive effect of GM-CSF after a 48-h incubation was maximal at 1- 10 @ml (Fig. 2). At concentrations below 1 ng/ml, the suppressive effect of GMCSF was still apparent, although the inhibition observed below 0.01 @ml was variable between experiments (Fig. 2B). The down regulation of apo E secretion by GM-CSF was specific and did not reflect a more generalized suppressive effect on macrophage secretion. Concentrations of GMCSF which resulted in maximal suppression of apo E induced a 3-fold increase in tibronectin secretion (Fig. 2C). Furthermore, the modulatory effect of GM-CSF on apo E secretion was heat labile with a greater than 80% loss of activity observed after boiling (data not shown) suggesting that the reduction in apo E secretion by GM-CSF was not due to contaminating endotoxin.

Fig. 1. Cytokine mediated regulation of apo E synthesis by murine peritoneal macrophages. Control or macrophages incubated with 10 @ml of GM-CSF, IFNy, TGF-8, TNF, IL-l, or 100 @ml of LPS or PMA for 48 h were [ 35S]methionine labelled for 5 h prior to immunoprecipitation. The apo E precipitable band was quantitated by densitometry and expressed as integrated optical density units (A). The 25-40-kDa molecular weight region of the autoradiogram from which densitometry was performed is presented (B). Representative experiment of 3.

C 22 -

Fibmnectin

g

10

Fibronectin

B 5

14

t b

10

t S 0 .oool 001 14.3-

.Ol

.I

1

GU-CSF(n@nl)

IO

loo

6

6 L;i;*. 16 0 .3001.001 01

1

1

IO loo

GMCSF(~S'~I)

Fig. 2. GM-CSF decreases apo E and stimulates tibronectin synthesis in a concentration dependent manner. Control macrophages immunoprecipitated with a mixture of normal goat and rabbit sera (lane 1) or with both anti apo E and anti tibronectin sera (lane 2) were compared to anti apo E and libronectin coprccipitates from macrophages stimulated with 100, 10, 1,O. 1, 0.01, or 0.001 ng/ml of GM-CSF (lanes 3-8) for 48 h. Supematants in triplicate following a 5-h [35S]methionine pulse were normalized for TCA precipitable counts prior to immunoprecipitation. Bands identified from the autoradiogram (A) as apo E and tibronectin were quantitated by densitometry and expressed as integrated optical density units for apo E (B) and tibronectin (C). Brackets indicate S.E.M. from a representative experiment of 4.

207 16000 m

Apo E 0 Fibronectin

12000

0

Control

A GM-CSF

-

0 TGF-/j m E

0

0

I

I

.OOOl

,001

I

I

I

1

.Ol

.l

1

10

TGF$(ng/ml) 4 Fig. 3. TGF-/3 increases synthesis of both apo E and fibronectin in a concentration dependent manner. Control or macrophages stimulated with TGF-fl for 48 h were [35S]methionine labeled and supernatants, in triplicate, were processed as described in Fig. 2. Bands corresponding to apo E and tibronectin were quantitated by Betagen blot analyzer from the original gels dried onto nitrocellulose. Representative experiment of 3, brackets indicate S.E.M.

The increased secretion of ape E by TGF-/3 necessitated further studies to determine whether this effect was concentration dependent and specific for apo E. Increased secretion of apo E by TGF-/3 was maximal when macrophages were stimulated with l- 10 @ml of cytokine (Fig. 3). At TGF-0 concentrations below 0.1 ng/ml, apo E secretion approached control values. In distinction to the opposite effects of TGF-/3 and GM-CSF on apo E, both cytokines stimulated libronectin release although the increase mediated by TGF-6 was consistently less than with GM-CSF. The modulatory effects of both TGF-/3 and GM-CSF were also dependent on the duration of cytokine exposure with maximal changes observed by 48 h for both apo E (Fig. 4A) and fibronectin (Fig. 4B). Increased secretion of apo E by TGF-/3 occurred by 24 h and returned towards control levels by 6 days. The depressed secretion of apo E in GM-CSF stimulated macrophages was apparent by 24 h and persisted through the entire time course. The enhanced apo E secretion observed for control macrophages with culture suggests that GM-CSF acts in part by preventing this increase from occurring. However, if fluid phase GM-CSF is removed by washing the monolayers, then the cytokine treated macrophages approach control levels of apo E secretion (75% of control) by 96 h.

6

B _

0

2

b Time (days)

Fig. 4. Kinetics of GM-CSF and TGF-8 regulation of apo E synthesis. Control or macrophage cultures stimulated with 10 @ml of GM-CSF or TGF-fl for O-7 days were labeled with [3SS]methionine for the last 5 h of each day of cytokine exposure. Supernatants were normalized for TCA precipitable counts and immunoprecipitated for apo E (A) or fibronectin (B). Immunoprecipitable bands were quantitated from autoradiograms by densitometry and expressed as integrated optical density units. Total cell protein (n = 2) at 48 h after cytokine treatment was not significantly different between the cultures and ranged between 90 and 116 pg cell protein per IO6 cells. Representative experiment of 3, brackets indicate S.E.M.

The effects of GM-CSF and TGF-0 on apo E secretion were also evaluated on cholesterol loaded macrophages following ingestion of acetylated LDL. As demonstrated in Fig. 5, macrophages preloaded with acetylated LDL for 24 h, followed by cytokines for an additional 24 h had reduced apo E secretion when exposed to GM-CSF and enhanced apo E when stimulated with TGF-0.

208

A

*

TABLE 1 EFFECI OF GM-CSF AND TGFQ CHOLESTEROL DISTRIBUTION

8 holesterol loaded

.z

ON CELLULAR

6

Conditionsa

Cholesterol esterb

Free cholesterol b

1 0 8‘Z

4

Control GM-CSF TGF-6

Cl00 Cl00 Cl00

8330 (330) 9100 (205) 8186 (1333)

P zP

2

1

z 0

Cholesterol Loaded

Control

1

2

3

4

5

Wntreated or macrophages incubated with 10 ng/rnl of GMCSF or TGF-6 for 24 h were labeled for an additional 24 h with 1 &i/ml of [ t4C]cholesterol. Macrophages were shifted to RPM1 1640 containing 2 mg/ml of bovine serum albumin for a final 24 h in the continued presence of cytokines and then lysed and chloroform-methanol extracted. Esterified and free cholesterol were resolved by thin layer chromatography and quantitated by liquid scintillation counting. bNumbers represent the counts/mm for esteritied and free cholesterol per 2.5 x 10’ cells per well. The mean counts/mm (n = 3) is presented with the value from each experiment determined from triplicate wells. Parentheses indicates S.E.M.

6

Fig. 5. Cytokine modulation of apo E secretion in cholesterol loaded macrophages. Macrophages were incubated in the presence (cholesterol loaded) or absence of 50 pglml of acetylated LDL for 24 h prior to exposure to 10 ng/ml of GMCSF or TGF-6. Both control and cholesterol loaded macrophages were pulsed for 5 h with [35S]methionine 24 h after cytokine treatment and supematants were immunoprecipitated. Immunoprecipitates were electrophoresed, fluorographed and apo E precipitated band intensitities were quantitated by densitometry and expressed as integrated optical density units (A). The 2%40-kDa region of the gel which was scanned is presented(B). Lanes 1,4 represent the untreated controls while lanes 2,5 represent GM-CSF and lanes 3,6 TGF/3 stimulated macrophages, respectively.

While total apo E secretion is enhanced in the acetylated LDL treated macrophages when compared with the parallel non-lipid loaded cultures, the changes in apo E secretion with cytokine treatment were of a similar magnitude. Since intracellular cholesterol pools are known to regulate apo E synthesis [17-211, it was necessary to consider that the modulation of apo E by GM-CSF and TGF-/3 was occurring through changes in cellular free cholesterol levels. Control and cytokine treated macrophages were labeled with 1 $X/ml of [14C]cholesterol for 24 h followed by an additional 24 h incubation in RPM1 + 2 mg/ml BSA to permit isotopic equilibration.

Macrophage lysates were chloroform-methanol extracted and the relative amounts of free and esterified cholesterol were quantitated by liquid scintillation counting following thin layer

TABLE 2 ACAT ACTIVITY CELLS

IN NON-CHOLESTEROL

LOADED

Conditionss

[ t4C]oleate (h)

Cholesterol esterb

Triglyceridesb

Control

2 4 2 4 2 4

249 (59) 392 (164) 183 (25) 193 (30) 211 (74) 318 (116)

3718 (678) 5457 (1079) 2590 (66) 4425 (1091) 7072 (2120) 12824 (2495)’

GM-CSF TGF-fl

“Control or macrophages incubated with 10 @ml of GM-CSF or TGF-fl for 48 h were pulsed with 0.2 &i of [ t4C)oleate in 0.5 ml of media for 2 or 4 h. Following the [t4C]oleate pulse triplicate wells were lysed, chloroform-methanol extracted and resolved by thin layer chromatography. bCholesterol ester and triglycerides were quantitated by liquid scintillation counting and expressed as the mean counts/mm from 3 separate experiments. *Significant at P < 0.1 by 2-tailed Student’s r-test. Parentheses indicate S.E.M.

209

chromatography. Under these conditions, greater than 98% of the [i4C]cholesterol in both the control and cytokine treated non-cholesterol loaded macrophages migrated as free cholesterol (Table 1). The relative amounts of free cholesterol between the control and cytokine treated macrophages were indistinguishable by this technique suggesting that the cytokine mediated changes in apo E secretion were not related to significant changes in the cellular free cholesterol pool. In additional experiments, intracellular free cholesterol was quantitated by fluorodensitometry and also found to be unchanged between the control and cytokine treated macrophages with values between 60 and 63 pg free cholesterol per mg cell protein obtained. Finally, in additional experiments with acetylated LDL preloaded cytokine

treated macrophages, there were no significant differences between cytokine and control cells in free cholesterol (data not shown) although apo E was down regulated 3-fold by GM-CSF and up regulated 1.7-fold by TGF-8. Although there was no evidence for differences in cellular free cholesterol in the cytokine treated macrophages, the possibility that alterations in basal ACAT activity may have occurred, resulting in changes in free cholesterol was examined (Table 2). The basal levels of ACAT activity as determined by quantitation of [ i4C]cholesterol ester following a 2- and 4-h [14C]oleate pulse were not significantly different between the cytokine stimulated and control macrophages. TGF-8 stimulated macrophages did however demonstrate an increase in triglyceride synthesis over the con-

C 15000 I

80000

-o-Control --t TGFB

-13 Control + GM-CSF

60000

1 oooo-

40000 500020000 0

0

I

123456

I

I

I

I

I

2

3

4

5

6

D

B

-I

65000 -

I

20000 4 g

45000 -

$ ‘E E 25000= .’ c

10000

yy

0

5000 123456

1

Hours

2

3

4

5

6

Hours

Fig. 6. Effect of GM-CSF and TGF-j3 on macrophage ACAT activity. Control, GM-CSF (A,B), or TGF-fl (CD) stimulated macrophage cultures in triplicate were pulsed after 48 h with 50 &ml of acetylated LDL and [14C]oleic acid for 2,4, or 6 h. Lysates were extracted and the organic phase chromatographed on flexible TLC plates. Cholesterol ester (A,C) and triglycerides (B,D) were identified on autoradiograms and quantitated by liquid scintillation counting. Approximately 1250 counts/mm corresponds to 1 nmol/mg cell protein of cholesterol ester. Brackets indicates standard deviations, each cytokine treatment and its respective control were evaluated in three separate experiments and a representative experiment is presented.

210

trol or GM-CSF stimulated macrophages. Therefore, changes in cellular free or esterified cholesterol were not observed in GM-CSF or TGF-8 stimulated macrophages, suggesting a distinct mechanism for apo E regulation. Although GM-CSF and TGF-/3 had no effect on basal ACAT activity, similar experiments were performed to determine cytokine effects on macrophage ACAT activity and cholesterol ester accumulation in the presence of acetylated LDL as a source of modified LDL. Control or cytokine treated macrophages were incubated for 48 h and pulsed the final 2,4, or 6 h with [ 14C]oleate and acetylated LDL. As demonstrated in Fig. 6A, GMCSF stimulated macrophages showed a significant increase in cholesterol esterification when compared with controls. This effect did not merely reflect differences in oleate uptake as the corresponding increase in triglycerides in the GMCSF treated cultures was not statistically signiticant (Fig. 6B). Macrophages incubated with TGF/3 in contrast, had a 2-fold reduction in cholesterol ester formation (Fig. 6C) while differences with the parallel controls for triglycerides (Fig. 6D) were not apparent. These results suggest that GM-CSF and TGF-/3 can affect the rate of cholesterol esterification following ACAT induction. However, these effects on ACAT activity and cholesterol ester accumulation were less significant if cytokine treated macrophages were incubated with acetylated LDL for 24 h prior to the [14C]oleate pulse (data not shown). The effects of GM-CSF and TGF-/3 on apo E mRNA levels were then investigated to determine whether apo E regulation was translational or pre translational. As demonstrated in Fig. 7, the effects of these cytokines on apo E secretion were consistent with changes in the steady state mRNA levels. GM-CSF resulted in a 2-fold reduction in apo E mRNA 24-72 h after cytokine exposure. No significant changes were detected at 8 h, and differences with the control cultures were less apparent by 5-7 days. TGF-fi in contrast resulted in a 2-3-fold increase in apo E mRNA 8 h after cytokine exposure and this effect persisted at 7 days. In distinction to the quantitative differences in apo E mRNA, hybridization with the oligo d(T) probe (Fig. 7B) indicated comparable amounts of total RNA were loaded per slot. These results were

Cl Control

Oi

4

Time (days)

Apo E

6

1

2

dT

3

4

5

6

Fig. 7. Effect of GM-CSF and TGF-/3 on apo E mRNA levels. Control or macrophages stimulated with TGF-fl and GM-CSF at 2 and 10 ne/ml, respectively, from 8 h to 7 days, were lysed and the resulting RNA was blotted onto nitrocellulose and hybridized with oligonucleotide probes for apo E and oiigo d(T) to correct for unequal RNA loading. (A) Hybridizable counts were quantitated by Betagen blot analyzer and each point with the exception of the day 7 control and TGF-,!I represents the mean of 3-5 separate experiments. Bars indicate S.E.M. (B) Representative autoradiogram is presented for control (lanes 1, 4), GM-CSF (lanes 2, 5) and TGF-~3 (lanes 3, 6) stimulated macrophage cultures. Slots (top to bottom) represent 8 h and days 1, 3, 5 and 7 cultures in the presence or absence of cytokines.

consistent with regulation of apo E synthesis by GM-CSF and TGF-fi occurring at a pretranslational level. Discussion Regulation of apo E has been demonstrated in vitro and in vivo with elevated levels reported in animal models associated with hypercholesterolemia [36-381. In vitro regulation of macrophage apo E synthesis in response to cholesterol loading with VLDL or acetylated LDL was initially

211

reported in the studies of Basu et al. [ 13,141 and has since been confirmed and extended in other primary and transformed macrophage cell lines [ 17-221. These studies demonstrated that elevations in the intracellular pool of free cholesterol regulated transcriptional activation and secretion of apo E. The earlier studies by Werb and coworkers suggested that macrophage activating factors including LPS, zymosan ingestion, IFN-/, pyran and BCG elicitation resulted in reduced apo E secretion [l&23-25]. While several macrophage activating agents suppress apo E secretion, increased apo E secretion by macrophages, with the exception of cholesterol loading, has not been reported for defined mediators. Takagi et al. have reported that human monocyte derived macrophages when incubated with platelets secrete greater amounts of apo E than cells comparably loaded with cholesterol following incubation with acetylated LDL [39]. While the platelet mediator(s) involved in this up regulation of apo E secretion were not defined, the role of TGF-0 was considered in the present study due to its effects on macrophages and inflammation [40]. GM-CSF and TGF-/3 have been reported to affect both cholesterol esterification and scavenger receptor activity in macrophages or macrophage like cell lines [27,31] as well as to modulate other effector functions associated with inflammation [32-351. GM-CSF stimulated macrophages in the present study exhibited reduced apo E secretion which was concomitant with lower mRNA levels. In distinction to the suppressive effect of GM-CSF on apo E synthesis, fibronectin secretion was enhanced 3-5-fold over controls. Therefore, the down regulation of apo E synthesis by GM-CSF was not the result of a generalized suppression of protein synthesis or secretion. The effects of GMCSF were also reversible if the cytokine was removed after 24 h. TGF-/3, in contrast, increased apo E secretion and this increase was consistent with elevations in apo E mRNA, suggesting pretranslational regulation. The effects of TGF-P on apo E secretion appear to be less significant after 4 days exposure (Fig. 4) although apo E mRNA remains elevated at 7 days (Fig. 7). Whether this difference in the kinetics is related to a general effect on protein secretion, catabolism or reuptake is not clear at present. The distinction be-

tween apo E mRNA and protein secretion was also apparent in the control macrophages where the increase in apo E secretion with culture occurred without any significant change in apo E mRNA. Therefore, in agreement with previous studies [17-211, regulation of apo E secretion involves both transcriptional and post-transcriptional mechanisms. The cytokine mediated effects on apo E secretion were also apparent on macrophages preloaded with acetylated LDL. These cells by oil red 0 staining resemble the lipid laden foam cells present within atherosclerotic lesions. Whereas the ability of classic macrophage activating factors to down regulate apo E secretion has been investigated [ 15,23-251, the cytokine mediated changes in apo E secretion in macrophage derived foam cells demonstrated in the present study has not been previously reported. In both lipid loaded and control macrophage cultures, cytokine modulation of apo E secretion appears to be independent of changes in cellular free cholesterol. These results suggest distinct mechanisms for regulation of apo E synthesis which are independent of effects on cellular free cholesterol pools. Cytokine effects on cholesterol esterification and ACAT activity were also investigated in the . present study for both control and acetylated LDL preloaded macrophages. Neither GM-CSF nor TGF-/3 had any significant effect on basal ACAT activity, although TGF-P treated macrophages had greater levels of [14C]oleate derived triglycerides. This increase in triglycerides was only detected in the non-loaded macrophages; acetylated LDL pulsed macrophages did not show a similar increase in triglycerides over the control cultures. In distinction to the lack of cytokine mediated effects on basal ACAT activity, GMCSF increased and TGF-0 decreased ACAT activity when cholesterol esterification was induced by including acetylated LDL during the 2-6-h [‘4C]oleate pulse. A similar effect of TGF-P on ACAT activity has been reported in the human monocytic line THP-1 [31]. The detection of TGF-0 mRNA in atherosclerotic lesions from cholesterol fed rabbits [26] suggests its involvement in the developing atheroma. Clearly, several effects of TGF-/3 could be involved in mitigating this lesion. In addition to

212

increasing apo E secretion as demonstrated in the present study, TGF-/3 has been reported to inhibit the generation of hydrogen peroxide and nitrogen oxide in IFNy stimulated macrophages [41,42]. This suppression of the oxidative burst may depress macrophage oxidation of LDL. The decreased scavenger receptor activity on TGF-@ treated THP-1 cells [31] provides another mechanism for lesion regression by reducing foam cell formation. The ability of TGF-/3 to down regulate IL-1 receptors [43,44] suggests that it may also down regulate cytokine receptors on smooth muscle cells involved in proliferation. However, its involvement in lesion progression also has to be considered since TGF-fl stimulates monocyte IL- 1 synthesis, is a monocyte chemoattractant and induces myeloid GM-CSF receptors [32,33,45]. The relationship between GM-CSF and macrophage activation suggests a proinflammatory role for this cytokine within the atherosclerotic lesion. GM-CSF stimulates myeloid differentiation and effector functions and can synergize with other factors to further increase antibody dependent cytotoxicity, phagocytosis, chemotaxis, secretion, cholesterol esterification, as well as the expression of tissue factor and other cell surface determinants [27,34,35,46-481. However, while these inflammatory properties may contribute to the progression of the atherosclerotic lesion, patients in phase I/II clinical trials receiving GM-CSF demonstrated decreases in total serum cholesterol during therapy [49]. Similar effects with M-CSF have been reported in normal and hypercholesterolemic rabbits as well as in nonhuman primates [50]. Neither serum apo E levels nor changes in the apo E containing HDL fraction in the animal models or patients on CSF were reported in these studies. While cytokines have been demonstrated in vitro to regulate apo E expression and in vivo to reduce total plasma cholesterol, other studies have reported the converse, namely, the effects of LDL or modified LDL particles on cytokine expression. Uptake of oxidized LDL has been demonstrated to inhibit TNF transcription [51] while injection of minimally modified LDL into mice increases MCSF expression [52]. The ability of modified LDL to stimulate the expression of endothelial cell derived GM-CSF [53] and of lipoprotein(a) to in-

hibit TGF-fl activation [54] suggests that these particles may also contribute to atherogenesis by regulating the synthesis of cytokines which may be involved in lesion progression. The macrophage secretion of apo E is presumably regulated by both negative (GM-CSF, IFN$ and positive (TGF-fl) cytokines with the relative concentrations of these mediators influencing the ability of the macrophage to promote reverse cholesterol transport. These cytokines by modulating apo E secretion, macrophage scavenger receptor activity, cholesterol esterification, LDL modification and cytokine release determine whether the macrophage plays a beneficial or detrimental role within an atheroma. Clearly, a greater understanding of the complex interrelationships between macrophages, cytokines and modified lipoproteins within the microenvironment of the atheroma is necessary for defining the critical regulatory events involved in lesion regression or progression. Acknowledgements The authors thank Bruce Glover for providing the apo E antisense oligonucleotide probe, Dr. Chandrasekhar for his critical review of this manuscript and Dr. Patrick Eacho for his supportive fluorodensitometric studies to quantitate free cholesterol in cytokine treated macrophages. References Mahley, R.W., Apolipoprotein E: Cholesterol transport protein with expanding role in cell biology, Science, 240 (1988) 622. Getz, G.S., Mazzone, T., Soltys, P. and Bates, S.R., Atherosclerosis and apoprotein E. An enigmatic relationship, Arch. Pathol. Lab. Med., 112 (1988) 1048. Wilson, C., WardelI, M.R., Weisgraber, K.H., Mahley, R.W., and Agard, D.A., Three dimensional structure of the LDL receptor-binding domain of human apolipoprotein E, Science, 252 (1991) 1817. Chao, Y.-S., Yamin, T.-T., Thompson, G.M. and Kroon, P.A., Tissue-specific expression of genes encoding apolipoprotein E and apolipoprotein A-I in rabbits, J. Biol. Chem., 259 (1984) 5306. Reue, K.L., Quon, D.H., O’Donnell, K.A., Dizikes, G.J., Fareed, G.C. and Lusis, A.J., Cloning and regulation of messenger RNA for mouse apolipoprotein E, J. Biol. Chem., 259 (1984) 2100. Simonet, W.S., Bucay, N., Lauer, S.J., Wirak, D.O..

213

1

8

9

10

11

12

13

14

15

16

17

18

19

Stevens, ME., Weisgraber, K.H., Pitas, R.E. and Taylor, J.M., In the absence of a downstream element, the apolipoprotein E gene is expressed at high levels in kidneys of transgenic mice, J. Biol. Chem., 265 (1990) 10809. Curtiss, L.K. and Edgington, TX, Immunoregulatory plasma low density lipoprotein: the biologic activity and receptor-binding specificity is independent of neutral lipid, J. Immunol., 126 (1981) 1008. Hui, D.Y., Harmony, J.A.K., Innerarity, T.L. and Mahley, R.W., Immunoregulatory plasma lipoproteins. Role of apoprotein E and apoprotein B, J. Biol. Chem., 255 (1980) 11775. Boyles, J.K. Zoellner, CD., Anderson, L.T., Kosik, L.M., Pitas, R.E., Weisgraber, K.H., Hui, D.Y., Mahley, R.W. et al., A role for apolipoprotein E, apolipoprotein A-l and low density lipoprotein receptor in cholesterol transport during regeneration and remyelination of the rat sciatic nerve, J. Clin. Invest., 83 (1989) 1015. Weisgraber, K.H., Rall, SC., Mahley, R.W., Milne, R.W., Marcel, Y.L. and Sparrow, J.T., Human apolipoprotein E. Determination of the heparin binding sites of apolipoprotein Es, J Biol. Chem., 261 (1986) 2068. Vollmer, E., Roessner, A., Bosse, A., Backer, W., Kaesberg, B., Robenek, H., Sorg, C. and Winde, G., Immunohistochemical double labeling of macrophages, smooth muscle cells and apolipoprotein E in the atherosclerotic plaque, Path. Res. Pratt., 187 (1991) 184. Namba, Y., Tomonaga, M., Kawasaki, H., Otomo, E. and Ikeda, K., Apolipoprotein E immunoreactivity in cerebral amyloid deposits and neurotibrillary tangles in Alzheimer’s disease and kuru plaque amyloid in CreutzfeldtJakob disease, Brain Res., 541 (1991) 163. Basu, S.P., Brown, M.S., Ho, Y.K., Have], R.J. and Goldstein, J.L., Mouse macrophages synthesize and secrete a protein resembling apolipoprotein E, Proc. Natl. Acad. Sci. USA, 78 (1981) 7545. Basu, S.P., Ho, Y.K., Brown, MS., Bilheimer, D.W., Anderson, R.G.W. and Goldstein, J.L., Biochemical and genetic studies of the apoprotein E secreted by mouse macrophages and human monocytes, J. Biol. Chem., 257 (1982) 9788. Werb, Z. and Chin, J.R., Apoprotein E is synthesized and secreted by resident and thioglycollate elicited macrophages but not by pyran copolymer or Bacillus Calmette Guerin activated macrophages, J. Exp. Med., 158 (1983) 1272. Basu, S.P., Goldstein, J.L. and Brown, M.S., Independent pathways for secretion of cholesterol and apolipoprotein E by macrophages, Science, 219 (1983) 871. Dory, L., Synthesis and secretion of apo E in thioglycollate-elicited mouse peritoneal macrophages: effect of cholesterol efilux, J. Lipid Res., 30 (1989) 809. Dory, L., Regulation of apolipoprotein E secretion by high density lipoproteins in mouse macrophages, J. Lipid Res., 32 (1991) 783. Mazzone, T., Gump, H., Diller, P. and Getz, G.S.,

20

21

22

23

24

25

26

27

28

29

30

31

32

Macrophage free cholesterol content regulates apolipoprotein E synthesis, J. Biol. Chem., 262 (1987) 11657. Mazzone, T., Basheeruddin, K. and Poulos, C., Regulation of macrophage apolipoprotein E gene expression by cholesterol, J. Lipid Res., 30 (1989) 1055. Mazzone, T. and Basheeruddin, K., Dissociated regulation of macrophage LDL receptor and apolipoprotein E gene expression by sterol: J. Lipid Res., 32 (1991) 507. Rouis, M., Nigon, F., Eggerman, T.L., Brewer, H.B. and Chapman, M.J., Apolipoprotein E expression by humanmonocyte-derived macrophages. Modulation by opsonized zymosan and cholesterol, Eur. J. Biochem., 189 (1990) 447. Werb, Z. and Chin, J.R., Endotoxin suppresses expression of apoprotein E by mouse macrophages in vivo and in culture. A biochemical and genetic study, J. Biol. Chem., 258 (1983) 10642. Takemura, R. and Werb, Z., Modulation of apoprotein E secretion in response to receptor-mediated endocytosis in resident and inflammatory macrophages, J. Exp. Med., 159 (1984) 167. Oropeza, R.L., Schreiber, R. and Werb, Z., Regulation of apolipoprotein E expression in macrophages by yinterferon. In: Sorg, C. and Schimpl, S. (Eds.), Cellular and Molecular Biology of Lymphokines, Academic Press, NY, 1985, pp. 303-307. Ross, R., Masuda, J., Raines, E.W., Gown, A.M., Katsuda, S., Sasahara, M., Malden, L.T., Masuko, H. and Sato, H., Localization of PDGF-B protein in macrophages in all phases of atherogenesis, Science, 248 (1990) 1009. Ishibashi, S., Inaba, T., Shimano, H., Harada, K., Inoue, I., Mokuno, H., Mori, N., Gotoda, T., Takaku, F. and colony-stimulating factor Yamada, N., Monocyte enhances uptake and degradation of acetylated low density lipoproteins and cholesterol esteritication in human monocyte-derived macrophages, J. Biol. Chem., 265 (1990) 14109. Horiuchi, K., Tajima, S., Menju, M. and Yamamoto, A., Structure and expression of mouse apolipoprotein E gene, J. B&hem., 106 (1989) 98. Zuckerman, S.H., Evans, G.F. and Guthrie, L., Transcriptional and post transcriptional mechanisms involved in the differential expression of LPS induced IL- 1 and TNF mRNA, Immunology, 73 (1991) 460. Goldstein, J.L., Basu, SK. and Brown, M., Receptormediated endocytosis of low density lipoprotein in cultured cells, Methods Enzymol., 98 (1983) 241. Bottalico, L.A., Wager, R.E., Agellon, L.B., Assoian, R.K. and Tabas, I., Transforming growth facto@ inhibits scavenger receptor activity in THP-1 human macrophages, J. Biol. Chem., 266 (1991) 22866. Wahl, SM., Hunt, D.A., Wakefield, L.M., McCartneyFrancis, N., Wahl, L.M., Roberts, A.B. and Spom, M.B., Transforming growth factor type beta induces monocyte chemotaxis and growth factor production, Proc. Natl. Acad. Sci. USA. 84 (1987) 5788.

214 33

34

35

36

31

38

39

40 41

42

43

44

Wahl, SM., McCartney-Francis, N., Allen, J.B., Dougherty, E.B. and Dougherty, S.F., Macrophage production of TGF-fi and regulation by TGF-0, Ann. NY Acad. Sci., 593 (1990) 188. Lang, R.A., Metcalf, D., Cuthbertson, R.A., Lyons, I., Stanley, E., Kelso, A., Kannourakis, G., Williamson, D.J., Klintworth, G.K., Gonja, T.J. et al., Transgenic mice expressing a hemopoietic growth factor gene (GMCSF) develop accumulations of macrophages, blindness and a fatal syndrome of tissue damage, Cell, 51 (1987) 615. Gregory, S.H., Magee, D.M. and Wing, E.J., The role of colony-stimulating factors in host defenses, Proc. Sot. Exp. Biol. Med., 197 (1991) 349. Lin-Lee, Y.-C., Tanaka, Y., Lin, C.-T. and Chan, L., Effects of an atherogenic diet on apolipoprotein E biosynthesis in the rat, Biochemistry, 20 (1981) 6474. Apostolopoulos, J.J., Howlett, G.J. and Fidge, N., Effects of dietary cholesterol and hypothyroidism on rat apolipoprotein mRNA metabolism, J. Lipid Res., 28 (1987) 642. Kim, M.H., Nakayama, R., Manos, P., Tomhnson, J.E., Choi, E., Ng, J.D. and Holten, D., Regulation of apohpoprotein E synthesis and mRNA by diet and hormones, J. Lipid Res., 30 (1989) 663. Takagi, Y., Dyer, C.A. and Curtiss, L.K., Plateletenhanced apolipoprotein E production by human macrophages: a possible role in atherosclerosis, J. Lipid Res., 29 (1988) 859. Nathan, C. and Spom, M., Cytokines in context, J. Cell. Biol., 113 (1991) 981. Tsunawaki, S., Spom, M., Ding, A. and Nathan, C., Deactivation of macrophages by transforming growth factor 0, Nature, 334 (1988) 260. Ding, A., Nathan, C.F., Graycar, J., Derynck, R., Stuehr, D.J. and Srimal, S., Macrophage deactivating factor and transforming growth factors &, & and & inhibit induction of macrophage nitrogen oxide synthesis by IFNy, J. Immunol., 145 (1990) 940. Dubois, C.M., Ruscetti, F.W., Palaszynski, E.W., Falk, L.A., Oppenheim, J.J. and Keller, J.R., Transforming growth factor fl is a potent inhibitor of interleukin 1(IL-I) receptor expression: Proposed mechanism of inhibition of IL-1 action, J. Exp. Med., 172 (1990) 737. Harvey, A.K., Hrubey, P.S. and Chandrasekhar, S., Transforming growth factor-8 inhibition of interleukin 1

45

46

41

48

49

50

51

52

53

54

activity involves down-regulation of interleukin 1 receptors on chondrocytes, Exp. Cell Res., 195 (1991) 376. Keller, J.R., Jacobsen, S.E.W., Sill, K.T., Ellingsworth, L.R. and Ruscetti, F.W., Stimulation of granulopoiesis by transforming growth factor 0: Synergy with granulocytelmacrophage-colony stimulating factor, Proc. Natl. Acad. Sci. USA, 88 (1991) 7190. Zuckerman, S.H. and Surprenant, Y.M., Induction of endothelial cell/macrophage procoagulant activity: Synergistic stimulation by gamma interferon and granulocyte-macrophage colony stimulating factor, Thromb. Haemostasis, 61 (1989) 178. Zuckerman, S.H., Surprenant, Y.M. and Tang, J., Synergistic effect of granulocyte-macrophage colonystimulating factor and 1,25 dihydroxyvitamin D3 on the differentiation of the human monocytic cell line U937, Blood, 71 (1988) 619. Metcalf, D., Begley, C.G., Johnson, G.R., Nicola, N.A., Vadas, M.A., Lopez, A.F., Williamson, D.J., Wong, G.G., Clark, SC. and Wang E,A., Biologic properties in vitro of a recombinant human gramdocyte-macrophage colony stimulating factor, Blood, 67 (1986) 37. Nimer, S.D., Champlin, R.E. and Golde, D.W., Serum cholesterol-lowering activity of granulocyte-macrophage colony-stimulating factor, J. Am. Med. Assoc., 260 (1988) 3291. Stoudemire, J.B. and Gamick, M.B., Effects of recombinant human macrophage colony-stimulating factor on plasma cholesterol levels, Blood, 77 (1991) 750. Hamilton, T.A., Ma, G.P. and Chisolm, G.M., Oxidized low density lipoprotein suppresses the expression of tumor necrosis factor-alpha mRNA in stimulated murine peritoneal macrophages, J. Immunol., 144 (1990) 2343. Liao, F., Berliner, J.A., Mehrabian, M., Navab, M., De.mer, L.L., Lusis, A.J. and Fogelman, A.M., Minimally modified low density lipoprotein is biologically active in vivo in mice, J. Clin. Invest., 87 (1991) 2253. Rajavashisth, T.B., Andalibi, A., Territo, MC., Berliner, J.A., Navab, M., Fogelman, A.M. and Lusis, A.J., Induction of endothelial cell expression of granulocyte and macrophage colony stimulating factors by modified lowdensity lipoproteins, Nature, 344 (1990) 254. Kojima, S., Harpel, P.C. and Rifkin, D.B., Lipoprotein(a) inhibits the generation of transforming growth factor 13: An endogenous inhibitor of smooth muscle cell migration., J. Cell Biol., 113 (1991) 1439.