In the presence of acetylated low-density lipoprotein, macrophages infected with an adeno-virus expressing human adipophilin showed a 31% increase in triglyceride content and a greater n
Trang 1Adipophilin increases triglyceride storage in human
macrophages by stimulation of biosynthesis and
inhibition of b-oxidation
Guilhem Larigauderie1,2,3, Clarisse Cuaz-Pe´rolin1,2,3, Amena B Younes4, Christophe Furman1,2,3, Catherine Lasselin1,2,3, Corinne Copin1,2,3, Michael Jaye5, Jean-Charles Fruchart1,2,3and
Mustapha Rouis1,2,3
1 Inserm, U545, Lille, F-59019 France
2 Institut Pasteur de Lille, De´partement d’Athe´roscle´rose, Lille, F-59019 France
3 Universite´ de Lille 2, Faculte´ de Pharmacie, Lille, F-59019 France
4 Inserm IFR-17, Laboratoire de Microscopie Electronique, Lille, France
5 GlaxoSmithKline, King of Prussia, PA, USA
Lipid-enriched macrophage-derived foam cells are an
early and characteristic feature of atherosclerotic
lesions Lipid loading of macrophages in vitro can be
achieved by chemical modification of the
apolipopro-tein B component of low-density lipoproapolipopro-tein (LDL), aggregation of LDL induced by either vortexing or treatment with lipases, or complexing of LDL with gly-cosaminoglycans or antibodies which bind macrophages
Keywords
adipophilin; macrophage; atherosclerosis;
lipid droplet; triglycerides
Correspondence
M Rouis, INSERM UR545, Institut Pasteur
de Lille, 1 rue du Professeur Calmette,
59019 Lille, France
Fax: +33 3 20 87 73 60
Tel: +33 3 20 87 73 79
E-mail: mustapha.rouis@pasteur-lille.fr
(Received 19 April 2006, revised 30 May
2006, accepted 5 June 2006)
doi:10.1111/j.1742-4658.2006.05357.x
Lipid accumulation alters macrophage biology and contributes to lipid retention within the vessel wall In this study, we investigated the role of adipophilin on triglyceride accumulation and lipid-droplet formation in THP-1-derived macrophages (THP-1 macrophages) In the presence of acetylated low-density lipoprotein, macrophages infected with an adeno-virus expressing human adipophilin showed a 31% increase in triglyceride content and a greater number of lipid droplets compared with control cells Incubation of macrophages with very low-density lipoprotein (VLDL) dra-matically increased cellular triglyceride content similarly in control and adipophilin-overexpressing cells By itself, VLDL increased adipophilin expression, which explains the lack of effect of adipophilin overexpression
on cellular triglyceride content in macrophages loaded with VLDL The lipid-droplet content of macrophages was increased by overexpression of adipophilin and⁄ or loading with VLDL In contrast, inhibition of adipo-philin expression using siRNA prevented lipid-droplet formation and signi-ficantly reduced intracellular triglyceride content Using inhibitors of b-oxidation and acyl-coenzyme A synthetase, results were obtained which suggest that adipophilin elevates cellular lipids by inhibition of b-oxidation and stimulation of long-chain fatty acid incorporation into triglycerides Adipophilin expression in THP-1 macrophages altered the cellular content
of different lipids and enhanced the size of lipid droplets, consistent with a role for adipophilin in human foam cell formation
Abbreviations
ACAT-1, acetyl-coenzyme A acetyltransferase 1; AcLDL, acetylated LDL; ADRP, murine adipose differentiation-related protein; AICAR, 5’-phosphoribosyl-5-aminoimidazole-4-carboxamide; CE, cholesteryl ester; FC, free cholesterol; HSL, hormone-sensitive lipase; LDL, low-density lipoprotein; m.o.i., multiplicity of infection; oxLDL, oxidized LDL; PPAR, peroxisome proliferator-activated receptors; TG, triglycerides; THP-1 macrophages, THP-1-derived macrophages.
Trang 2and promote LDL uptake by endocytosis [1] In
addi-tion, several studies have reported that macrophages
can accumulate large amounts of cholesteryl ester (CE)
through the uptake of oxidized LDL (oxLDL) by a
variety of mechanisms, including the scavenger
path-way [2], and that VLDL are capable of inducing CE
and triglyceride (TG) accumulation in macrophages
[3,4] The mechanism of TG accumulation in human
monocyte–macrophages primarily involves the direct
uptake of free fatty acids generated by the extracellular
lipoprotein lipase-mediated hydrolysis of VLDL–TG
followed by intracellular reesterification into lipids,
however, receptor-mediated uptake of intact VLDL
particles is also implicated [4–6]
Lipid accumulation in macrophages not only
contri-butes to cholesterol and TG retention within the vessel
wall, but also alters macrophage biology Indeed,
sev-eral studies have indicated that a diversity of effects on
macrophage function can be attributed to lipid loading
These include the upregulation of the genes for
apo-lipoprotein E [7], elastase [8] and tissue factor [9], as
well as altered expression of several other genes [10]
Within cells, lipid is stored in spherical organelles
called lipid droplets [11] which have been reported to
play active and diverse roles in the cellular life cycle
Indeed, lipid droplets are involved in the maintenance
of intracellular cholesterol balance in fibroblasts [12]
and appear to be the principal source of fatty acids in
adipose and liver [13] Moreover, correlations between
lipid droplets and certain human diseases such as
athe-roma plaque, steatosis, obesity and cancers have been
reported [11]
Lipid droplets are composed of a CE and TG core
surrounded by a phospholipid monolayer and coated
with specific proteins [11] Adipophilin, or adipose
dif-ferentiation-related protein (ADRP), a 50 kDa protein
initially described in adipocytes [14], is a marker of
lipid accumulation and is among the lipid
droplet-associated proteins present in a variety of cells such as
hepatocytes, adipocytes, muscle cells, mammary
epithe-lial cells, fibroblasts, endotheepithe-lial cells and macrophages
[15,16]
Macrophage expression of adipophilin is upregulated
by oxLDL [17], acetylated LDL (AcLDL) [18],
enzy-matically modified LDL [19] and by synthetic agonists
of the peroxisome proliferator-activated nuclear
recep-tors d (PPARd) [20,21] and c (PPARc) [22,23]
How-ever, the precise role of adipophilin in macrophage
foam cell formation and, in turn, in the development
of atherosclerotic lesions remains unclear In this
study, we investigated the impact of adipophilin
over-expression or downregulation on lipid accumulation
and droplet formation in human THP-1 macrophages
Results
We have previously shown that adipophilin expression was greater in human atherosclerotic lesions than in healthy areas of the same artery and that the majority
of adipophilin mRNA in atheromatous tissue was attributed to lipid-rich macrophages (CD68+ cells) [18] We have also reported that THP-1 cells differenti-ated into macrophages with phorbol esters were able to rapidly take up AcLDL and to subsequently develop a foam cell-like morphology Under these conditions, adi-pophilin expression was enhanced dramatically [18] To further study the function of adipophilin in human macrophages, we generated an adenovirus vec-tor-expressing human adipophilin (Ad.CMV.adipo-philin) Using the control Ad.CMV.GFP vector, we demonstrated nearly 100% infection of THP-1 macro-phages (data not shown) We assessed adipophilin expression using both quantitative PCR and immuno-blotting in cells infected with two different amounts of Ad.CMV.adipophilin At multiplicity of infection values (m.o.i.) of 100 and 500, adipophilin mRNA was increased 14 ± 0.8- and 39 ± 7.8-fold, respectively, and adipophilin protein was increased
6.5 ± 1.7- and 38 ± 15-fold, respectively, com-pared with control cells (Fig 1)
When cells were loaded with 100 lgÆmL)1 AcLDL, adipophilin overexpression resulted in a modest but significant increase in TG (1.3-fold, P < 0.05) (Fig 2) and altered cellular CE and free cholesterol
Fig 1 Expression of adipophilin protein and mRNA in adenovirus-infected THP-1 macrophages THP-1 macrophages were adenovirus-infected with Ad.CMV.GFP or Ad.CMV.adipophilin at 100 and 500 m.o.i Three days later, total proteins and total RNA were isolated Total protein was analysed by western blotting (upper) and RNA was quantified using real-time qPCR (lower) *The difference between Ad.CMV.GFP-infected cells and cells infected with Ad.CMV adipophilin was significant at P < 0.01.
Trang 3(FC) content (1.4-fold increase, P < 0.01 and more
than twofold decrease, P < 0.01, respectively, data
not shown) However, when cells were loaded with 10,
50 and 100 lgÆmL)1VLDL or 10 lgÆmL)1AcLDL, no
significant difference in TG content was seen between
control and adipophilin-overexpressing cells (Fig 2),
whereas cellular CE and FC contents were altered
similarly to AcLDL-loaded cells in the presence of
100 lgÆmL)1 VLDL (1.6-fold increase, P < 0.05 and
60% decrease P < 0.05, respectively, data not shown)
The fact that we did not observe increased TG content
in adipophilin-overexpressing vs control cells
follow-ing incubation with VLDL is due to the dramatic
increase in the cellular TG content in all cells under these conditions To examine whether VLDL increased adipophilin in human macrophages, we incubated THP-1 macrophages with increasing concentrations of VLDL and examined adipophilin levels by immuno-blotting (Fig 3) In the presence of 10 and
100 lgÆmL)1VLDL, adipophilin increased 12 ± 2.8-and 28 ± 5.2-fold relative to control macrophages cultured in lipid-free medium Thus, elevation of cellu-lar adipophilin by VLDL renders it impossible to observe an effect of Ad.CMV.adipophilin-mediated adipophilin overexpression on TG content (Fig 2) To confirm this hypothesis, we incubated
Ad.CMV.adipo-0 25 50 75 100 125 150 175 200
No added
non-infected cells Ad.CMV.GFP Ad.CMV.adipophilin
VLDL (µg/ml)
AcLDL (µg/ml)
*
Fig 2 Effect of adipophilin overexpression on lipid mass in THP-1 macrophages incubated with VLDL Cells were infected with 500 m.o.i.
of Ad.CMV.GFP (control) or Ad.CMV.adipophilin and incubated with 0, 10, 50 or 100 lgÆmL)1VLDL or AcLDL (10, 100 lgÆmL)1) in medium containing 1% fetal bovine serum for 48 h The results are the means ± SD of three independent experiments performed in quadruplicate.
*The difference between control and cells infected with Ad.CMV.adipophilin in the presence of 100 lgÆmL)1AcLDL was significant at
P < 0.05.
Fig 3 Expression of adipophilin protein levels in human THP-1 macrophages loaded with VLDL THP-1 macrophages were incu-bated with 0, 10 or 100 lgÆmL)1VLDL in RPMI-1640 containing 0.4% BSA for 48 h Total proteins were isolated and samples of
20 lg were separated by SDS ⁄ PAGE (10%) and blotted onto a nitrocellulose membrane The results are mean ± SD of three inde-pendent experiments *The difference bet-ween control and cells incubated with VLDL was significant at P < 0.01.
Trang 4philin-infected macrophages with 10 and 100 lgÆmL)1
VLDL for 48 h and measured adipophilin levels by
immunoblotting On top of the already elevated level
of adipophilin expression in
Ad.CMV.adipophilin-infected cells, 10 and 100 lgÆmL)1 VLDL increased
adipophilin 1.8 ± 0.71- and 5.2 ± 1.91-fold,
respectively, relative to Ad.CMV.adipophilin-infected
macrophages cultured in lipid-free medium In
com-parison, no significant difference in adipophilin
expres-sion was observed in Ad.CMV.adipophilin-infected
macrophages loaded or not with 10 and 100 lgÆmL)1
of AcLDL (data not shown)
The induction of adipophilin expression and the
excessive lipid loading of THP-1 macrophages in
response to VLDL treatment was confirmed using
immunolocalization experiments which revealed the
presence of numerous large lipid droplets surrounded
by adipophilin (Fig 4B) and by Oil Red O staining
(Fig 5E,F) In cells cultured in the absence of VLDL
(control cells), nominal diffuse cytoplasmic adipophilin
staining was observed (Fig 4A); in the presence of
VLDL, adipophilin staining was pronounced in both
Ad.CMV.adipophilin- and Ad.CMV.GFP-infected cells
(Fig 4G,H) Adenoviral-mediated overexpression of
adipophilin followed by an incubation with 100
lgÆmL)1AcLDL for 24 h showed a significant increase
in lipid-droplet formation (Fig 4F) compared with
AcLDL-loaded control cells (Fig 4E) or
adipophilin-overexpressing cells incubated without VLDL or
AcLDL (Fig 4D)
To further assess the impact of adipophilin levels on
lipid-droplet formation we manipulated adipophilin
levels in THP-1 macrophages by infection with
Ad.CMV.adipophilin or by transfection of cells
with adipophilin siRNA Noninfected and control
Ad.CMV.GFP-infected THP-1 macrophages grown in
serum-free RPMI-1640 or in 10% fetal bovine serum
showed a quasi absence of lipid droplets in the
cyto-plasm following Oil Red O staining (Fig 5A,C,I)
However, incubation of adipophilin-infected
macro-phages in RPMI-1640 supplemented with 10% fetal
bovine serum (Fig 5D) followed by staining with Oil
Red O showed a significant increase in lipid droplets
in comparison with Ad.CMV.GFP-infected cells
(Fig 5C) In agreement with our intracellular TG
measurements, cells incubated with 100 lgÆmL)1
VLDL (Fig 5B,E,F,I) contained a greater number of
lipid droplets than cells incubated with or without
10% fetal bovine serum (Fig 5A,C,D,I) When THP-1
macrophages were transfected with siRNA–adipophilin
or siRNA–GAPDH (control) followed by incubation
with 100 lgÆmL)1 VLDL for 24 h, there was a
sub-stantial reduction in the number and size of lipid
drop-lets in siRNA–adipophilin-transfected cells (Fig 5H), whereas control siRNA–GAPDH-transfected cells accumulated a large number of lipid droplets (Fig 5G) To verify the implication of adipophilin in lipid-droplet formation, we measured the intracellular accumulation of TG in siRNA–adipophilin-transfected macrophages following 48 h incubation with either
10 or 100 lgÆmL)1 VLDL Inhibition of adipophilin expression decreased cellular TG content in both cases
by 30% compared with control cells (Fig 6) Potential mechanisms by which adipophilin increased lipid content in THP-1 macrophages include protection of lipid droplets against the activity of intracellular lipases such as hormone-sensitive lipase (HSL), inhibition of fatty acid oxidation (which would favour the recycling of fatty acids), enhancement of acetyl-coenzyme A acetyltransferase 1 (ACAT-1) este-rification activity or stimulation of lipid synthesis The effect of adipophilin on intracellular lipase activity was determined by adding the acyl-coenzyme A synthetase inhibitor triacsin C to the medium of macrophages preloaded with oleate, to inhibit the reutilization of fatty acids released from hydrolysed TG [24,25] No significant differences could be observed between triascin C-treated Ad.CMV.adipophilin, Ad.CMV GFP and noninfected cells, which all contained
58 ± 9.9% of the initial TG mass at 24 h post infec-tion (data not shown), suggesting that adipophilin does not elevate cellular TG by protecting it from cellular lipases However, because these results were obtained using an indirect method (triacsin C inhibition) affecting total cellular lipase, we subsequently measured more specifically HSL activity in lysate-infected macrophages loaded with AcLDL as an exogenous source of lipids (because VLDL strongly induced adipophilin expression even in Ad.CMV.GFP infected macrophages) Our data showed a significant twofold decrease (P < 0.001) in HSL activity in Ad.CMV.adipophilin-infected cells compared with control cells (Fig 7) This inhibitory effect on HSL activity, seen in the presence of elevated amounts of adipophilin, may explain to some extent the increased storage of lipids
To examine the effect of adipophilin on TG synthe-sis, Ad.CMV.adipophilin- and Ad.CMV.GFP-infected macrophages were loaded with 400 lm palmitate, either alone or with 2.5 lm triacsin C, and the cellular
TG content was quantified In the absence of triac-sin C, adipophilin-overexprestriac-sing cells produced more
TG from palmitate than control cells (133.6 ± 8.4 vs
95 ± 9.7 lgÆmg)1 cell protein, respectively) The addi-tion of triacsin C together with palmitate reduced the
TG mass in both Ad.CMV.GFP- and Ad.CMV adipophilin-infected cells to approximately the same
Trang 5levels (66.5 ± 11.2 and 61.5 ± 10.7 lgÆmg)1 cell
pro-tein, respectively) (Fig 8) These results suggest that
increased TG in cells overexpressing adipophilin is at
least partly due to acyl-coenzyme A synthetase
eleva-ted activity or to the downstream incorporation of
acyl-CoAs into TG Lipid esterification was quantified
in either THP-1 macrophages overexpressing
adipophi-lin (following infection with Ad.CMV.adipophiadipophi-lin)
or after downregulation of adipophilin expression
by transfection with siRNA–adipophilin Neither
enhanced nor reduced adipophilin expression had an
effect on ACAT-1 activity, given that similar amounts
of [14C]oleate incorporation into cholesteryl oleate were measured in both cases (data not shown) To fur-ther probe this, we assessed the impact on TG accu-mulation of pharmacological inhibition of ACAT-1
in Ad.CMV.adipophilin-infected and control THP-1 macrophages Enhanced TG accumulation in adipophi-lin-overexpressing cells was dependent on the addition
of palmitate, whereupon adipophilin-overexpressing cells accumulated 1.6 times more TG than control cells In the absence of palmitate, TG accumulation
Fig 4 Immunocytochemical analysis of adi-pophilin in THP-1 macrophages (magnifica-tion, 63·) Cells were cultured in RPMI-1640 supplemented with 10% fetal bovine serum Adipophilin was immunolocalized using a specific polyclonal antibody as described in Experimental procedures (A) Control cells grown without added lipids for 24 h (B) Cells incubated with VLDL (100 lgÆmL)1) for
24 h (C) Cells infected with Ad.CMV.GFP (D) Cells infected with Ad.CMV.adipophilin (E) Cells infected with Ad.CMV.GFP, then treated for 24 h with AcLDL (100 lgÆmL)1) (F) Cells infected with Ad.CMV.adipophilin, then incubated for 24 h with AcLDL (100 lgÆmL)1) (G) Cells infected with Ad.CMV.GFP, then treated for 24 h with VLDL (100 lgÆmL)1) (H) Cells infected with Ad.CMV.adipophilin, then incubated for 24 h with VLDL (100 lgÆmL)1).
Trang 6in adipophilin-overexpressing and control cells was
similar and unaffected by addition of the ACAT-1
inhibitor CAY10486 The absolute amounts and fold
stimulation of TG accumulation in palmitate-loaded,
adipophilin-overexpressing cells (133.6 ± 6.7 lgÆmg)1 cell protein, 1.5-fold) were similar when cells were co-incubated with palmitate plus CAY10486 (150.2 ± 7.6 lgÆmg)1 cell protein, 1.6-fold) These results
I
Fig 5 Lipid-droplet staining with Oil Red O
in THP-1 macrophages (magnification, 63·).
(A) Cells cultured in serum-free RPMI-1640.
(B) Cells cultured in RPMI-1640
supplemen-ted with 10% fetal bovine serum and
100 lgÆmL)1VLDL (C) Cells infected with
500 m.o.i of Ad.CMV.GFP then maintained
in culture in RPMI-1640 supplemented with
10% fetal bovine serum (D) Cells infected
with 500 m.o.i of Ad.CMV.adipophilin then
incubated in RPMI-1640 supplemented with
10% fetal bovine serum (E) Cells infected
with 500 m.o.i of Ad.CMV.GFP then
incuba-ted in RPMI-1640 supplemenincuba-ted with 100
lgÆmL)1VLDL (F) Cells infected with 500
m.o.i of Ad.CMV.adipophilin then incubated
in RPMI-1640 supplemented with 100 lgÆ
mL)1VLDL (G) siRNA–GAPDH-transfected
cells incubated in RPMI-1640 containing
100 lgÆmL)1 (H)
siRNA–adipophilin-transfected cells incubated in RPMI-1640
containing 100 lgÆmL)1VLDL (I) Average
number of lipid droplets in THP-1
macrophages cultured and treated as
described in the preceding legend (A–H)
expressed as fold change from control cells
(described in A) * The difference was
significant at P < 0.05 #, non significant.
Trang 7cate that enhanced accumulation of TG in
adipophilin-overexpressing cells does not involve ACAT-1, because
specific inhibition of ACAT-1 was without effect
Next, we examined whether inhibition of fatty acid
oxidation may also contribute to TG elevation by
adi-pophilin in THP-1 macrophages TG accumulation in
Ad.CMV.GFP- and Ad.CMV.adipophilin-infected cells
incubated in the presence of palmitate (positive
con-trol) was compared with TG accumulation in cells incubated with palmitate plus bromopalmitate, a non-metabolized inhibitor of fatty acid oxidation [26] For these experiments, the concentrations of palmitate and bromopalmitate used were 100 lm, because higher bromopalmitate concentrations were toxic for THP-1 macrophages Cells were infected or not with Ad.CMV.GFP or Ad.CMV.adipophilin and then loa-ded with 100 lm fatty acids for 48 h Adipophilin-infected cells accumulated 1.4 times more TG than control cells (Fig 9) No differences in TG accumula-tion were observed between any of the cell-treatment groups when cells were incubated only with bromo-palmitate, which is poorly incorporated into TG The addition of both palmitate and bromopalmitate to noninfected cells, as well as to cells infected with the control vector (Ad.CMV.GFP) showed an increase in
TG mass in comparison with cells incubated with palmitate only This indicates that fatty acid oxidation
is an ongoing process in THP-1 macrophages; inhibi-tion of fatty acid oxidainhibi-tion by bromopalmitate results
in elevated cellular TG content In contrast to control cells, no significant differences were observed between adipophilin-overexpressing cells incubated with palmi-tate alone or with bromopalmipalmi-tate plus palmipalmi-tate This
Fig 6 Effect of adipophilin downregulation on lipid mass in THP-1
macrophages incubated with VLDL Cells were transfected with
siRNA–GAPDH (control) or siRNA–adipophilin and 24 h later,
incuba-ted with 10 or 100 lgÆmL)1 VLDL or AcLDL (10, 100 lgÆmL)1) in
medium containing 1% fetal bovine serum for 48 h The results are
the means ± SD of three independent experiments performed in
triplicates *The difference between control and cells transfected
with siRNA-adipophilin was significant at P < 0.05 # The difference
between macrophages incubated in the presence of VLDL and
con-trol, Ad.CMV.GFP or Ad.CMV.Adipophilin-infected macrophages
(without VLDL) was significant at P < 0.05.
Fig 7 Effect of adipophilin overexpression on HSL activity in
THP-1 macrophages THP-1 cells infected with either Ad.CMV.
adipophilin or Ad.CMV.GFP and incubated for 48 h with AcLDL.
HSL activity was assayed as neutral CE by following the release of
[1- 14 C] oleic acid from cholesteryl [1- 14 C]oleate as described in
Experimental procedures *The difference between control and
Ad.CMV.adipophilin cells was significant at P < 0.001.
0 20 40 60 80 100 120 140 160
No added lipids
Palmitate Palmitate +
triacsin C
CAY10486 Palmitate +
CAY10486
non-infected cells Ad.CMV.GFP Ad.CMV.adipophilin
*
*
NS
Fig 8 Triglycerides synthesis is inhibited by triacsin C but not by
an ACAT-1 inhibitor in THP-1 macrophages overexpressing adipo-philin THP-1 macrophages were infected or not with 500 m.o.i of Ad.CMV.GFP and Ad.CMV.adipophilin and then incubated either for
16 h at 37 C with 400 l M palmitate complexed to BSA in the pres-ence or abspres-ence of 2.5 l M triacsin C or for 48 h at 37 C with
400 l M palmitate complexed to BSA in the presence or absence of
60 l M CAY10486 TG content was quantified on lipid extracts.
*The difference between Ad.CMV.adipophilin cells in the presence
of palmitate or palmitate + CAY10486 and cells in the absence of lipids or in the presence of palmitate (only control THP-1 macroph-ages), palmitate + triacsin C, CAY10486 or palmitate + CAY10486 (only control THP-1 macrophages) was significant at P < 0.01 NS, non significant.
Trang 8suggests that the enhanced TG content in
adipophilin-overexpressing cells may be partly due to inhibition of
fatty acid oxidation, because it cannot be further
inhibited by bromopalmitate
To confirm this hypothesis, we subsequently treated
adipophilin-overexpressing or control THP-1
macro-phages with
5’-phosphoribosyl-5-aminoimidazole-4-carboxamide (AICAR), which stimulates fatty acid
oxidation by activation of AMP-activated protein
kin-ase Cells were incubated with or without palmitate
and TG levels were assessed Addition of AICAR
alone did not significantly change the intracellular TG
content, however, there was a trend for increased TG
in cells treated with both AICAR and palmitate
(Fig 9) The rather modest effects of AICAR in
palmi-tate-loaded cells is not surprising, because AICAR
treatment did not affect adipophilin expression (data
not shown) and AICAR induces fatty acid oxidation
and therefore the degradation of palmitate In the
presence of AICAR, TG levels in
adipophilin-overex-pressing cells were similar to control cells (57.0 ± 6.2
and 63.0 ± 6.0 lgÆmg)1 cell protein, respectively)
Thus, stimulation of fatty acid oxidation by AICAR
appears to be dominant over the inhibition of fatty acid oxidation by adipophilin
Discussion
We studied the impact of adenoviral-mediated overex-pression of adipophilin in THP-1-derived macrophages
on the accumulation of TG when the cells were incu-bated in the presence of VLDL, AcLDL or palmitate Adipophilin is a lipid droplet-associated protein which
is expressed in a wide range of lipid-accumulating cells including macrophages [10,16,27] However, little is known about the function of adipophilin in macro-phages By analogy with adipocytes, which share certain common features [28,29], and preadipocytes, which may convert to macrophages [30], stimulation of human adipophilin expression might induce lipid-droplet formation in macrophages The function of ADRP, the murine equivalent of human adipophilin, has been analysed in murine fibroblasts and the results showed that ADRP stimulated lipid accumulation and lipid-droplet formation without induction of other adi-pocyte-specific genes or other lipogenic genes [31] More recently, ADRP-deficient mice were created which showed reduced hepatic TG content as well as protection from diet-induced fatty liver compared with wild-type mice [32]
lgÆmL)1), adipophilin overexpression resulted in eleva-ted cellular TG content Incubation of macrophages with VLDL dramatically elevated the TG content of both adipophilin-overexpressing and control cells To strengthen these results, we investigated the presence
of adipophilin around these lipid droplets using immu-nofluorescence microscopy of THP-1 macrophages Our results confirmed the presence of adipophilin sur-rounding all sizes of lipid droplet in THP-1 macro-phages (Fig 4) Moreover, droplet formation was stimulated in cells overexpressing adipophilin and con-versely, adipophilin expression was strongly increased
in macrophages loaded with VLDL (Fig 3)
To determine whether exogenous lipid or cellular adipophilin content was rate-limiting for cellular TG accumulation, we quantitated TG in Ad.CMV.adipo-philin-infected macrophages The intracellular TG content accumulation was dependent on the amount
of VLDL in the culture media, and no significant dif-ferences were observed between Ad.CMV.adipophilin and control macrophages incubated with 0–100 lgÆmL)1 VLDL (Fig 2) Because endogenous adipo-philin expression is induced by VLDL loading of cells (Fig 3), these data do not indicate whether the level
of lipids or adipophilin is rate-limiting for TG
accu-0
20
40
60
80
100
120
140
160
Palmitate
Bromopalmitate
AICAR
non-infected cells
Ad.CMV.GFP
Ad.CMV.adipophilin
-+ +
-+
*
*
# #
NS
NS
NS
##
##
NS
Fig 9 Effect of adipophilin on TG content in THP-1 macrophages
grown in culture medium without or with 100 l M palmitate, 100 l M
bromopalmitate and 500 l M AICAR THP-1 macrophages were
infected with 500 m.o.i of Ad.CMV.GFP and Ad.CMV.adipophilin.
Infected and noninfected cells were then incubated for 48 h at
37 C with or without fatty acids and AICAR as indicated on the
fig-ure *The difference between Ad.CMV.adipophilin cells and
con-trols in the absence of lipids or in the presence of palmitate or
palmitate + bromopalmitate was significant at P < 0.05 # The
dif-ference between controls in the presence of bromopalmitate and
controls in the presence of palmitate alone or bromopalmitate +
palmitate was significant at P < 0.05 ## The difference between
controls in the presence of bromopalmitate + palmitate and
con-trols in the absence of lipids or in the presence of palmitate alone
was significant at P < 0.05 NS, non significant.
Trang 9mulation The mechanism of adipophilin stimulation
by VLDL has been described in murine macrophages
and shown to be dependent on activation of the
nuc-lear receptor PPARd [21] To further investigate
whe-ther adipophilin or lipids were rate-limiting for lipid
accumulation, we incubated adipophilin-infected
macrophages with lipids and showed an increase in
the size of lipid droplets (Fig 5D,I) In contrast,
si-RNA–adipophilin-transfected cells accumulated
sub-stantially less lipid in the presence of VLDL (Figs 5H
and 6) These results suggested that
adipophilin-deple-ted cells might take up less VLDL and clearly
indica-ted that adipophilin was rate-limiting for lipid
accumulation in human macrophages This conclusion
is strengthened by our previous data showing that
siRNA–adipophilin-transfected macrophages
accumu-lated 50% less TG compared with control cells
[18] An additional hypothesis to consider would be
that because lipids could not be stored in lipid
drop-lets in adipophilin-deficient macrophages; their
distri-bution may also be different under these conditions
This latter hypothesis is strengthened by the fact that
although livers from ADRP-deficient and wild-type
mice showed similar total lipid abundance,
ADRP-deficient mice contained significantly less TG In these
mice, subcellular distribution analyses revealed that
TG was reduced in the cytosolic fraction but
increased twofold in the microsomal fraction [32]
As shown in Fig 3, human THP-1 macrophages
incubated for 48 h with 100 lgÆmL)1 VLDL contained
30 times more adipophilin protein level than control
cells Adipophilin levels in VLDL-loaded control cells
were similar to those measured in VLDL-loaded
Ad.CMV.adipophilin-infected macrophages (data not
shown), and no differences in adipophilin staining was
observed between VLDL-loaded
Ad.CMV.adipophilin-and Ad.CMV.GFP-infected macrophages (Fig 4G,H)
The data suggest that adipophilin in excess of the level
induced by lipid loading may be degraded Consistent
with this, when adipophilin was overexpressed simply
with the adenovirus without added lipids, only a
lim-ited amount of adipophilin was retained by the
nom-inal amount of intracellular lipids, as observed by Oil
Red O staining (Fig 5D) It appears that macrophages
adjust their adipophilin content depending on the
presence of cellular lipids and the ectopic expression of
adipophilin following by adenoviral infection is
degra-ded Thus, adipophilin is rate-limiting for lipid
accumulation in human macrophages, and both its
expression and stability appear to be regulated by
lipids This hypothesis is supported by the fact that in
the absence of ADRP, mice were resistant to
diet-induced fatty liver [32]
Because Ad.CMV.adipophilin-infected macrophages contained significantly more lipid droplets than control macrophages, we investigated the possible impact of adipophilin on fatty acid oxidation by using the non-metabolizable fatty acid bromopalmitate, an inhibitor
of fatty acid oxidation In palmitate-loaded control cells, bromopalmitate elevated cellular TG content, which indicates that fatty acid oxidation is an ongoing process in THP-1 macrophages In contrast, in adipo-philin-overexpressing cells loaded with palmitate, bromopalmitate failed to increase the already elevated level of TG These results suggest that the presence of
an elevated pool of adipophilin is sufficient to protect fatty acids from b-oxidation To examine whether adi-pophilin may increase cellular TGs by inhibition of fatty acid oxidation, experiments were performed with AICAR, which stimulates fatty acid oxidation by acti-vation of AMP-activated protein kinase Incubation of cells with AICAR resulted in loss of enhanced TG accumulation in adipophilin-overexpressing cells The data suggest that stimulation of fatty acid oxidation by AICAR is dominant over the inhibition of fatty acid oxidation by adipophilin The mechanism by which this occurs is unknown, but may be complex, because adipophilin is not known to be phosphorylated by AMP-activated kinase
Another mechanism by which adipophilin might ele-vate cellular TG is by stimulation of acyl-coenzyme A synthetase and⁄ or the incorporation of acyl-CoA into
TG To address this, triacsin C was used to inhibit acyl-coenzyme A synthetase, a key enzyme whose fatty acyl-CoA products may be incorporated into TG or become substrates for fatty acid oxidation Inhibition
of acyl-coenzyme A synthetase abrogated the elevated level of TG in adipophilin-overexpressing cells, sug-gesting that increased TG in adipophilin-overexpress-ing cells is due, at least in part, to elevated activity of acyl-coenzyme A synthetase or to the downstream incorporation of acyl-CoAs into TG
However, neither enhanced adipophilin expression nor its inhibition had an effect on whole-cell esterifica-tion activity (data not shown) The elevated TG pool in Ad.CMV.adipophilin-infected cells remained increased after specific ACAT-1 inhibition, suggesting that ACAT-1 was not implicated in the TG increase and that the fatty acid pool utilized by ACAT-1 was either very small compared with or not the same as that used to generate intracellular TG We also assessed whether adipophilin protected TG from hydrolysis For this, cells were preloaded with oleate followed by treatment with triacsin C to block acyl-coenzyme A synthetase and hence fatty acid incorpor-ation into TG Under these conditions, no difference
Trang 10in TG content was observed between control and
pophilin-overexpressing cells, which showed that
adi-pophilin does not protect TG from lipolysis However,
we did observe a significant decrease in HSL activity
on exogenous substrate in lysates of
Ad.CMV.adipo-philin-infected cells compared with controls This result
can be reconciled with the conclusion from the
triacsin C experiment that adipophilin does not protect
TG from lipolysis by the proposition that TG in lipid
droplets in adipophilin-overexpressing cells is relatively
inaccessible to HSL and other lipases This
interpret-ation is in agreement with results obtained in murine
adipocytes and in vivo in ADRP-deficient mice
show-ing that there was no significant effect of adipophilin
on basal and isoproterenol-stimulated lipolysis [32]
Because HSL also hydrolyses CE, the reduced activity
of HSL may explain the elevated levels of CE
meas-ured in adipophilin-overexpressing cells [18]
Macro-phages also contain CE hydrolase [33] and in future
studies it will be of interest to compare the effects of
adipophilin overexpression on macrophage expression
of these two lipases
In summary, our results suggest that adipophilin
increases TG in macrophages by stimulating
incorpor-ation of acyl-CoA into TG as well as by inhibition of
fatty acid oxidation This contrasts with findings in
ADRP-deficient mice, in which no difference in the
rates of fatty acid oxidation were observed between
ADRP-deficient and wild-type mice [32] This
discrep-ancy may reflect differences between the roles of
hep-atic vs nonhephep-atic adipophilin, species differences
(mouse vs human) or methodological differences
Concerning the latter, our conclusions regarding the
effects of adipophilin on fatty acid oxidation and TG
biosynthesis are largely based on TG mass
measure-ments in cells treated with different pharmacological
agents, whereas the disparate conclusions from
adipo-philin-deficient mice are based on radioisotopic
meas-ures in primary hepatocytes We note that in
adipophilin knockout mice, TG and nonesterified fatty
acids accumulated in the microsomal compartment
where TG is synthesized [32] These findings are in
agreement with our suggestion that adipophilin might
associate with intracellular fatty acids, which then
escape from b-oxidation pathways and are redirected
for esterification and storage Thus, when lipids
accu-mulate inside the macrophage, such as occurs in
conse-quence to VLDL loading, adipophilin expression is
stimulated, and lipid transport or incorporation into
nascent or ongoing lipid droplets ensues
In conclusion, we provided clear evidence that
adenoviral-mediated overexpression of human
adipophilin enhanced lipid-droplet formation in human
macrophages Our results indicated that adipophilin contributes to TG accumulation by stimulating the generation and⁄ or incorporation of fatty acyl CoAs into TG and⁄ or by inhibiting fatty acid oxidation Additional experiments are required to more precisely define the mode of action of adipophilin in human macrophages and its relevance in atherosclerosis
Experimental procedures
Cell culture and siRNA transfection assays
Human monocytic THP-1 cells (ATTC TIB-202, LGC Promochem, Molsheim, France) were maintained in
RPMI-1640 (BioWhittaker-Cambrex, Emerainville, France) con-taining 25 mmolÆL)1 Hepes buffer and 10% fetal bovine serum (Eurobio, Courtaboeuf, France) Three days before transfection, cells were seeded in six-well culture dishes (Falcon, Becton-Dickinson Labware, Franklin Lakes, NJ)
at a density of 1· 106cells⁄ well Differentiation of THP-1 monocytes to macrophages occurred in the presence of
160 nm phorbol 12-myristate 13-acetate (Sigma, Saint Quentin, France) for 72 h [34] Transfections of siRNA were carried out as described previously [18] About 80% inhibition of adipophilin expression was obtained
Recombinant adenovirus expression
Recombinant vectors were constructed using standard tech-niques [35] The full-length adipophilin cDNA was gener-ated by RT-PCR from total RNA of THP-1 cells using oligonucleotides designed to create XhoI (5¢) and MluI (3¢) cutting sites The digested fragment was cloned under the control of the CMV promoter in the pShuttle-CMV vector (Stratagene, La Jolla, CA) The recombinant adenovirus was constructed in 293 cells by in vivo homologous recom-bination between shuttle plasmids and pAdEASY-1 [36] and plaque purified High titre stocks of Ad.CMV.adipo-philin and Ad.CMV.GFP (2.7· 1012
and 8.5· 1012
viral particlesÆmL)1, respectively) were produced in 293 cells and purified on CsCl gradients THP-1 macrophages (1· 106cells⁄ well) were infected by highly purified adeno-virus vectors at a m.o.i of either 100 or 500 plaque-forming units⁄ cell in RPMI-1640, and 24 h later the infected macro-phages were ready for further studies
RNA analysis
Total RNA from THP-1 macrophages was extracted using the RNeasy kit (Qiagen) For RT-PCR analyses, 5 lg of total RNA was treated by DNAseI (Invitrogen Life Tech-nologies, Cergy-Pontoise, France) and reverse transcribed using random hexamer primers (Clontech Laboratories, Mountainville, NJ) and M-MLV reverse transcriptase