The FASEB Journal article fj.201700327R. Published online August 10, 2017.

THE

JOURNAL • RESEARCH • www.fasebj.org

IFN-g aggravates neointimal hyperplasia by inducing

endoplasmic reticulum stress and apoptosis in
macrophages by promoting ubiquitin-dependent liver
X receptor-a degradation
Qiang Zhao,* Dong Zhou,* Hongjun You,* Bowen Lou,* Yan Zhang,* Yuling Tian,* Ning Guo,* Xiaoli Chen,*
Yan Liu,* Yue Wu,* Zuyi Yuan,*,†,1 and Juan Zhou*,‡,2
*Department of Cardiovascular Medicine, First Affiliated Hospital of Medical School and †Key Laboratory of Environment and Genes Related
to Diseases, Ministry of Education, Xi’an Jiaotong University, Xi’an, China; and ‡Key Laboratory of Molecular Cardiology of Shanxi Province,
Xi’an, China

ABSTRACT: Neointimal hyperplasia is the main cause of restenosis after percutaneous coronary interventions (PCIs).
Both IFN-g and macrophages play nonredundant roles in the pathogenesis of vascular intimal hyperplasia;
however, the underlying mechanisms remain elusive and must be further investigated. In mouse peritoneal
macrophages, IFN-g significantly accelerated degradation and up-regulated polyubiquitination of liver X
receptor (LXR)-a. Signal transducer and activator of transcription 1 (STAT1) inhibitor, fludarabine, and
PIAS1 knockdown reduced ubiquitination and increased the expression of LXR-a in IFN-g–treated macro-
phages. IFN-g also increased the expression of endoplasmic reticulum (ER) stress–related proteins, including
p-PERK, p-eIIF2a, and CCAAT-enhancer-binding protein homologous protein (CHOP), as well as apoptosis
of macrophages. Treatment with ER stress inhibitor, 4-phenylbutyric acid (4-PBA), and LXR agonist, T0901317
(T0), alleviated IFN-g–induced apoptosis in macrophages. Neointimal hyperplasia was significant after ca-
rotid ligation for 4 wk in ApoE2/2 mice. IFN-g mAb, T0, and 4-PBA treatment not only significantly attenuated
neointimal hyperplasia but also decreased CD68+TUNEL+ double-positive macrophages in the hyperplastic
neointima. Moreover, after 4-PBA or T0 administration, the number of CD68+p-eIIF2a+ and CD68+CHOP+
double-positive cells in neointimal was also apparently decreased. Taken together, these results defined an
unexpected role of IFN-g and LXR-a in the development of neointimal hyperplasia. The PIAS1/STAT1-
dependent LXR-a degradation induced by IFN-g promoted ER stress and apoptosis in macrophages, which
leads to aggravated neointimal hyperplasia. LXR agonist efficiently improved neointimal hyperplasia, which
may be a promising new strategy to ameliorate restenosis and vascular remodeling after PCI.—Zhao, Q.,
Zhou, D., You, H., Lou, B., Zhang, Y., Tian, Y., Guo, N., Chen, X., Liu, Y., Wu, Y., Yuan, Z., Zhou, J. IFN-g
aggravates neointimal hyperplasia by inducing endoplasmic reticulum stress and apoptosis in macro-
phages by promoting ubiquitin-dependent liver X receptor-a degradation. FASEB J. 31, 000–000 (2017).
www.fasebj.org
KEY WORDS: inflammation • ER stress • restenosis

ABBREVIATIONS: 4-PBA, 4-phenylbutyrate; CHOP, CCAAT-enhancer-

Restenosis is the most important long-term adverse
binding protein homologous protein; ER, endoplasmic reticulum; Flu, outcome of stent implantation for coronary artery disease,
fludarabine; GAPDH, glyceraldehyde-3-phosphate dehydrogenase; occurring in 12% of drug-eluting stents and 24% of
HDAC4, histone deacetylase 4; IP, Immunoprecipitation; LXR, liver X
receptor; PCI, percutaneous coronary intervention; PI, propidium iodide;
bare-metal stents (1). As the main cause of restenosis after
STAT1, signal transducer and activator of transcription 1 percutaneous coronary interventions (PCIs), neointimal
1
Correspondence: Department of Cardiovascular Medicine, First Affili-
hyperplasia is characterized by arterial wall thickening
ated Hospital of Xi’an Jiaotong University, 277 West Yanta Rd., Xi’an and decreased arterial lumen space. When vessel walls are
2
710061, Shaanxi, China. E-mail: zuyiyuan@mail.xjtu.edu.cn damaged, smooth muscle cells and myofibroblasts pro-
Correspondence: Department of Cardiovascular Medicine, First Affili- liferate and migrate into the neointimal space, which re-
ated Hospital of Xi’an Jiaotong University, 277 West Yanta Rd., Xi’an
710061, Shaanxi, China. E-mail: 1306899042@qq.com sults in simultaneous secretion of a large amount of matrix
proteins (2). Recently, data have suggested that other cell
doi: 10.1096/fj.201700327R
This article includes supplemental data. Please visit http://www.fasebj.org to lines may also play significant roles in lesion formation.
obtain this information. Macrophages are the predominant type of myeloid cells

0892-6638/17/0031-0001 © FASEB 1
Downloaded from www.fasebj.org to IP 128.59.222.107. The FASEB Journal Vol., No. , pp:, August, 2017
that are recruited to injured artery (3). They exhibit a dis- ubiquitin were obtained from Abcam (Cambridge, MA, USA).
tinct profile via the release of cytokines and growth factors, Abs against CCAAT-enhancer-binding protein homologous
thereby facilitating the accumulation of vascular smooth protein (CHOP), phospho-eIF2a (Ser51), and phospho-signal
transducer and activator of transcription 1 (STAT1) were sup-
muscle cells and myofibroblasts within the intima (4, 5). plied from Cell Signaling Technology (Danvers, MA, USA). Abs
Few studies have proven that the thickness of neointima is against phospho-PERK, PIAS1, histone deacetylase 4 (HDAC4),
increased with the aggravation of inflammation (3). b-actin, and STAT1 inhibitor, fludarabine (Flu), were provided
Several studies have indicated the pathogenic role of by Santa Cruz Biotechnology (Santa Cruz, CA, USA). FITC-
IFN-g in the development of neointimal hyperplasia (6, 7). labeled goat anti-mouse IgG Abs were purchased from Abbkine
Expression of IFN-g is increased in areas of arterial injury, (Wuhan, China). CY3-labeled goat anti-rabbit IgG Abs were
and deficiency of IFN-g reduces neointimal formation (8). obtained from Abcam. LXR agonists T0901317 and GW3965 as
well as protein synthase inhibitors, cycloheximide and MG-132,
Preliminary data show that IFN-g induces human vascu- were supplied by Sigma-Aldrich (St. Louis, MO, USA). HDAC4
lar smooth muscle cell proliferation and intimal expansion inhibitors, trichostatin A and LMK 235, were provided by Tocris
via the mammalian target of rapamycin raptor complex 1 Bioscience (Bristol, United Kingdom). Anti-mouse IFN-g mAb
pathway (9), but its underlying mechanisms are still far was purchased from R&D Systems (Minneapolis, MN, USA).
from clear. Normal rat IgG was obtained from Sigma-Aldrich.
Recently, endoplasmic reticulum (ER) has been emerged
as an important pathophysiologic paradigm that un-
Cell culture and treatment
derlies chronic metabolic disease. ER homeostasis can be
disturbed by many physiologic and pathologic factors,
Resident peritoneal macrophages were collected and cultured in
such as unfolded protein response, viral infections, tox- RPMI 1640 medium (Thermo Fisher Scientific, Waltham, MA,
ins, and inflammatory cytokines (10). Severe or pro- USA) that was supplemented with 10% fetal bovine serum as
longed perturbed ER homeostasis or ER stress will previously described (16). Cells were seeded in a 6-well plate and
ultimately lead to the deterioration of cellular functions incubated in serum-free medium overnight before treatment.
and cell death, mainly apoptosis. It has been proven that RAW 264.7 macrophage cell line was purchased from the
ER stress promotes neointimal formation, and allevia- American Type Culture Collection (Manassas, VA, USA) and
maintained in DMEM (Thermo Fisher Scientific) that was sup-
tion of ER stress inhibits neointimal hyperplasia in a
plemented with 10% fetal bovine serum (Thermo Fisher Scien-
wire-induced vascular injury model (11). tific) at 37°C in a humidified atmosphere of 5% CO2. To explore
As a member of the nuclear receptor family of tran- the effects of several signaling molecules, cells were pretreated
scription factors, liver X receptor (LXR) is famous as a lipid with specific inhibitors for desired periods of time before cytokine
and cholesterol metabolism sensor. LXR protects cells exposure.
from cholesterol overload by inducing the transcription of
numerous cholesterol-related genes, including ABCA1
(ATP-binding cassette transporter A1), ABCG1 (ATP- RNA isolation and real-time quantitative PCR
binding cassette transporter G1), ARL7 [ADP-ribosylation
Total RNA was isolated from cultured cells by using TRIzol re-
factor (ARF)-like 7], and lipoprotein lipase (12, 13). Re-
agent (Thermo Fisher Scientific) according to manufacturer in-
cently, data have indicated that LXR also regulates ER structions. Purified RNA was reverse-transcribed into cDNA by
stress via dynamic modulation of membrane phospholipid using a RevertAid First Strand cDNA Synthesis Kit (Thermo Fisher
composition. Toxic lipids regulate activities of fatty acid Scientific), and newly synthesized cDNA was analyzed by real-
synthase and stearoyl CoA desaturase by inhibiting LXR-a, time quantitative PCR on iQ5 Multicolor Real-Time PCR detection
which results in a reduction of polyunsaturated fatty acid system (Bio-Rad, Hercules, CA, USA) with a SYBR Premix Ex Taq
and a higher ratio of membrane phospholipid to choles- II RT-PCR Kit (Takara, Kyoto, Japan). Glyceraldehyde-3-phos-
phate dehydrogenase (GAPDH) was used as a housekeeping gene.
terol, which eventually triggers ER stress (14). The following primers were used in this study: mouse LXR-a:
Current therapies have not targeted fundamental forward: 59-GCGTCCATTCAGAGCAAGTGT-39; reverse: 59-
disease-modifying mechanisms, and have resulted in only TCACTCGTGGACATCCCAGAT-39; mouse GAPDH; forward:
modest improvement in morbidity and mortality of reste- 59-AAGGTCATCCCAGAGCTGAA-39; reverse: 59-CTGCTTCAC-
nosis (15). Mechanisms of obliterative vascular remodeling CACCTTCTTGA-39.
and neointimal hyperplasia remain largely unexplored. In
the present study, we report that IFN-g aggravates the
neointimal hyperplasia by inducing ER stress–mediated Immunoblotting
apoptosis in macrophages by promoting ubiquitin-
Cultured cells were lysed in RIPA buffer (Cybrdi, Gaithersburg,
dependent LXR-a degradation. Taken together, we MD, USA) that contained protease inhibitor (Roche Diag-
propose that LXR-a may be a new target for the treat- nostics, Indianapolis, IN, USA). Protein concentration was
ment of neointimal hyperplasia and vascular restenosis. determined by using BCA protein assay reagent kit (Thermo
Fisher Scientific). Equal amounts of protein were loaded onto
4–12% Bis-Tris precast gels for electrophoresis and were
electrotransferred onto a PVDF membrane (Roche Diagnos-
MATERIALS AND METHODS tics). After blocking for 1 h at room temperature, membranes
were sequentially incubated with primary Abs at 4°C over-
Reagents night and secondary Abs at room temperature for 1 h. The
protein signal was detected by chemiluminescence. All den-
Recombinant murine IFN-g was purchased from PeproTech sitometric data for target protein were normalized with
(Rock Hill, NJ, USA). Abs against LXR-a, LXR-b, CD68, and b-actin, which was used as loading control.

2 Vol. 31 December 2017 The FASEB Journal x www.fasebj.org ZHAO ET AL.

Downloaded from www.fasebj.org to IP 128.59.222.107. The FASEB Journal Vol., No. , pp:, August, 2017
Immunoprecipitation selected visual fields, and the average number of double-positive
cells in each specimen was calculated.
Cell extracts were prepared by using modified RIPA buffer (Cell
Signaling Technology). Cell lysates (300 mg) were incubated with
1 mg of Ab at 4°C overnight and precipitated with protein G TUNEL labeling and CD68 costaining
agarose beads (Santa Cruz Biotechnology) at 4°C for 2 h.
Immunoprecipitated proteins were separated by SDS-PAGE and Mouse carotid artery was first stained with mAb against CD68 at
transferred onto a PVDF membrane. Membranes were then se- room temperature for 1 h, followed by incubation with a second
quentially incubated with primary and secondary Abs. Bands Ab as shown above. Cell death in situ of carotid artery was then
were visualized by using an ECL system. examined by using the In Situ TUNEL Kit (Merck Millipore,
Darmstadt, Germany) as previously described (19). In brief, tis-
sue sections were sequentially incubated with TUNEL reaction
Annexin V/propidium iodide staining mixture and anti–FITC-AP at 37°C in a humidified chamber for
60 min. Images were obtained using identical exposure settings
Annexin V/propidium iodide (PI) assay was performed fol- and were quantified by using ImagePro Plus 6.0 software.
lowing manufacturer instructions (Annexin V-FITC kit; Roche
Life Sciences, Basel, Switzerland). In brief, cells were washed with
PBS, then incubated with the mixture of Annexin V and PI at Statistical analysis
room temperature for 10 min in the dark. The fluorescence signal
was examined by using a fluorescence microscope (Olympus, Data between 2 groups were analyzed with SPSS 18.0 (SPSS,
Tokyo, Japan). Photomicrographs were obtained using identical Chicago, IL, USA) using Student’s t test (2-tailed, unpaired). One-
exposure settings, and all images were quantified using Image- way ANOVA was used for multiple groups. All values were
Pro Plus 6.0 software (Media Cybernetics, Rockville, MD, USA). expressed as means 6 SD unless otherwise stated. P , 0.05 was
considered statistically significant.
Animals and treatment protocol

Experimental protocols in our study were approved by the In- RESULTS

stitutional Ethics Committee for Animal Experiments of Xi’an
Jiaotong University. All procedures conformed to the U.S. IFN-g increases ER stress–dependent apoptosis
National Institutes of Health Guide for the Care and Use of in macrophages, whereas LXR activation
Laboratory Animals. All mice were housed in a controlled reverses it
environment (20 6 2°C, 12-h light/dark cycle) and maintained on
a standard chow diet with free access to water. The carotid liga- Although IFN-g–induced apoptosis has been confirmed in
tion model was carried out by using 8-wk-old male ApoE2/2
mice according to a previously described protocol (17). In brief, macrophages (20), its underlying mechanisms remain
mice were anesthetized with ketamine (75 mg/kg) and xylazine largely unexplored. Previous studies have indicated that
(5 mg/kg) by i.p. injection. Anesthesia was monitored by pinch- ER stress is a mechanism that is responsible for cell apo-
ing the tail and toe. The right common carotid arteries were dis- ptosis via prolonged activation of IRE1 or CHOP (21, 22).
sected and ligated near to the bifurcation by using 6.0 silk suture. To explore whether IFN-g induced apoptosis via the ER
Left common carotid arteries were dissected without ligation and stress–mediated pathway and whether LXR played a role
used as sham controls. After surgery, mice were administered
in such processes, we treated macrophages with different
analgesia (buprenorphine 0.05 mg/kg per 12 h, i.p.) 48 h for
postoperative analgesia. Mice were then fed a high-fat diet and concentrations of IFN-g for 24 h and examined apoptosis
treated with PBS, LXR agonist T0 (50 mg/kg per time, twice per using Annexin-V and TUNEL analyses. Figure 1A, B show
wk), 4-phenylbutyric acid (4-PBA; 100 mg/kg per time, twice per that treatment of IFN-g (20 ng/ml) significantly increased
wk), anti-mouse IFN-g mAb (100 mg, once per wk), or normal rat Annexin-V and TUNEL-positive macrophages com-
IgG (100 mg, once per wk) for 4 more weeks. Mice were eutha- pared with the control group; however, pretreatment
nized with CO2 7 d after ligation or 24 d after the last challenge. of ER stress inhibitor, 4-PBA, notably reduced such an in-
All animals were fixed by perfusion with 4% paraformaldehyde crease, which suggests that IFN-g caused apoptosis, in part,
for 3 min. After excision of the left and right carotid arteries, vessels
were fixed in 70% ethanol overnight, then cut in half and frozen in
in macrophages via the ER stress–dependent pathway.
optimal cutting temperature compound embedding medium. Recently, it has been reported that the toxic lipid-induced
inhibition of nuclear receptor LXR-a triggers ER stress by
impacting polyunsaturated fatty acid content of the ER
Histopathologic staining and membrane (14); therefore, we were curious about the role
immunofluorescence analysis
of LXR in IFN-g–induced apoptosis in macrophages.
The carotid artery was dissected and frozen in optimal cutting
Combination treatment of IFN-g and LXR synthetic ligand,
temperature compound embedding medium in 8-mm serial sec- T0 (T0901317), also sharply decreased Annexin-V and
tions. Carotid artery tissue was stained with hematoxylin and TUNEL-positive macrophages compared with IFN-g
eosin to evaluate pathologic lesions. The lesion area was quantified treatment alone (Fig. 1A, B). In addition, IFN-g stimulation
as previously described (18). CD68+CHOP+ or CD68+p-eIF2a+ significantly increased the levels of p-PERK, p-eIF2a, and
double-positive macrophages were analyzed by immunofluores- CHOP—3 ER stress markers—in macrophages, whereas
cence staining with mAbs against mouse CD68 or rabbit CHOP treatment of LXR agonists, T0 or GW (GW3965), reduced,
and p-eIF2a at room temperature for 1 h, followed by incubation
with FITC-labeled goat anti-mouse IgG or CY3-labeled goat anti- in part, the expression of the above-mentioned 3 markers at
rabbit IgG. All sections were photographed with identical expo- the protein level (Fig. 1C, D). These data illustrated that
sure settings, and all images were quantified by using ImagePro IFN-g could induce ER stress–dependent apoptosis in
Plus 6.0 software. Each specimen was examined in 5 randomly macrophages, in part, via the LXR-related pathway.

IFN-g AGGRAVATES NEOINTIMAL HYPERPLASIA BY INDUCING ER STRESS AND APOPTOSIS IN MACROPHAGES 3

Downloaded from www.fasebj.org to IP 128.59.222.107. The FASEB Journal Vol., No. , pp:, August, 2017
Figure 1. IFN-g increases ER stress–dependent apoptosis in macrophages, whereas LXR activation reverses it. A) Annexin
V-FITC/PI double staining for apoptotic macrophages treated with IFN-g, IFN-g + PBA, or IFN-g + T0 for 24 h. Early apoptotic
cells were marked only in green, whereas late apoptotic cells were presented in red with green (n = 5 in each group). B) TUNEL
assay for apoptotic macrophages that were treated with IFN-g, IFN-g + PBA, or IFN-g + T0 for 24 h. All apoptotic cells were
presented in red (n = 5 in each group). C, D) Representative Western blot analysis of p-PERK, p-eIF2a, and CHOP in
macrophages that were treated with IFN-g with or without T0 or GW, 2 LXR agonists, for 24 h. b-Actin was used as loading control
(n = 5 in each group for Western blot). *P , 0.05 vs. control.

IFN-g decreases the expression of LXR-a, but that LXR-a expression at the protein level was reduced
not LXR-b, at the protein level in macrophages after cycloheximide (50 mg/ml) treatment for 3 or 6 h,
which reveals that LXR-a protein could be degraded nat-
To explore the exact role of LXR in IFN-g–induced apo- urally. Pretreatment of 20 ng/ml IFN-g for 24 h accelerated
ptosis in macrophages, we first assessed the effects of LXR-a degradation (Fig. 3A, B), whereas proteasome in-
IFN-g on the expression of LXR. Mouse peritoneal mac- hibitor MG-132 almost totally reversed such degradation.
rophages were treated with 10, 20, and 50 ng/ml IFN-g for We confirmed the role of the ubiquitin-proteasome system
24 h or with 20 ng/ml IFN-g for 6, 12, or 24 h. Figure 2A in effects of IFN-g on LXR-a expression by using coim-
shows that IFN-g significantly decreased LXR-a expres- munoprecipitation (co-IP) assay. Figure 3C indicates that
sion at the protein level, but not LXR-b, in a dose- and time- IFN-g treatment significantly triggered the up-regulation
dependent manner. We then investigated the influence of of polyubiquitin and LXR-a. These results indicate that
IFN-g on the expression of LXR-a at the mRNA level. IFN-g decreased the expression of LXR-a via the ubiquitin-
Figure 2B shows that the expression of LXR-a at the proteasome system in macrophages.
mRNA level was not notably altered after IFN-g (20 ng/ml)
treatment for 3, 6, 9, or 12 h, which indicates that IFN-g
post-transcriptionally decreased the expression of LXR-a IFN-g induces degradation of LXR-a via the
at the protein level in macrophages. ubiquitin-proteasome system depending on
LXR-a/PIAS1/STAT1 tripolymer formation
IFN-g decreases the expression of LXR-a via
the ubiquitin-proteasome system It has been reported that LXR-a can be post-transcriptionally
in macrophages modified by histone deacetylase 4 (HDAC4) via SUMOylation
in IFN-g–stimulated primary astrocytes (23). To detect
To explore the underlying mechanisms of LXR-a down- whether post-transcriptional modification played a role
regulation induced by IFN-g treatment in macrophages, in the accelerated degradation of LXR-a in macrophages
we treated primary peritoneal macrophages with cyclo- that were treated by IFN-g, we detected interacting
heximide, a protein synthesis inhibitor. Figure 3A shows proteins with LXR-a by using co-IP assay. Figure 4A, B

4 Vol. 31 December 2017 The FASEB Journal x www.fasebj.org ZHAO ET AL.

Downloaded from www.fasebj.org to IP 128.59.222.107. The FASEB Journal Vol., No. , pp:, August, 2017
Figure 2. IFN-g decreases the protein expression of LXR-a, but not LXR-b, in macrophages. A) Representative Western blot
analysis of LXR-a and LXR-b in mouse peritoneal macrophages that were treated with different concentrations of IFN-g or with
20 ng/ml IFN-g for 6, 12, and 24 h. b-Actin was used as loading control. B) mRNA expression of LXR-a in mouse peritoneal
macrophages that were treated with IFN-g (20 ng/ml) for 3, 6, 9, and 12 h. GAPDH was used as a housekeeping gene. Data are
presented as means 6 SD from 3 independent experiments (n = 9 in each group for PCR and n = 5 in each group for Western
blot). *P , 0.05 vs. control.

exhibits the failure of coprecipitation of HDAC4 and treatment significantly increased the expression of poly-
LXR-a in macrophages that were treated by IFN-g. In ubibiquitin of LXR-a, whereas Flu and PIAS1 shRNA
addition, both TSA and LMK235—2 HDAC4 inhibitors— notably alleviated ubiquitination of LXR-a. These data
did not affect LXR-a expression induced by IFN-g in indicate the important roles of STAT1 and PIAS1 in the
macrophages (Supplemental Fig. 2). These data indicate process of LXR-a degradation induced by IFN-g.
that HDAC4 did not take part in LXR-a degradation
induced by IFN-g in macrophages. Surprisingly, copre- IFN-g mAb attenuates neointimal hyperplasia
cipitation of LXR-a, and not only p-STAT1, but also and inhibits apoptosis in experimental
protein inhibitor of activated STAT-1 (PIAS1), was de- arterial restenosis
tected under the action of IFN-g (Fig. 4A, C). STAT1 is an
important upper signaling molecule in IFN-g activation. It has been indicated that IFN-g and inflammation play
Similar to HDAC4, PIAS1 is not only an important reg- vital roles in the development of neointimal hyperplasia,
ulatory molecule in the IFN-g pathway but also a SUMO but its underlying mechanisms must be further in-
E3 ligase (24); therefore, we speculated that IFN-g vestigated. To investigate the role of IFN-g–induced apo-
treatment promoted the phosphorylation of STAT1 in ptosis of macrophages in the pathologic process of
macrophages. Under the action of p-STAT1, both p- neointimal hyperplasia, we reproduced the carotid liga-
STAT1 and PIAS1 synergistically modified LXR-a pro- tion model by using 8-wk-old ApoE2/2 mice. Mice were
tein, which made it easier to be degraded via the given high-fat diet when they were 7 wk old, and the
ubiquitin-proteasome system. We next treated macro- treatment lasted until the end point of the experiment.
phages with Flu, a STAT1-specific inhibitor, and PIAS1 After 1 wk of high-fat diet, the left common carotids of all
shRNA in our study. Figure 4D, E shows that IFN-g experimental mice were ligated, and the right common

IFN-g AGGRAVATES NEOINTIMAL HYPERPLASIA BY INDUCING ER STRESS AND APOPTOSIS IN MACROPHAGES 5

Downloaded from www.fasebj.org to IP 128.59.222.107. The FASEB Journal Vol., No. , pp:, August, 2017
Figure 3. IFN-g decreases the expression of LXR-a via the ubiquitin-proteasome system in macrophages. A) Representative
Western blot analysis of LXR-a in mouse peritoneal macrophages that were treated with cycloheximide (CHX; 50 mg/ml), a
protein synthesis inhibitor, for 3 and 6 h after being administered 20 ng/ml IFN-g for 24 h. b-Actin was used as loading control.
B) Representative Western blot analysis of LXR-a in mouse peritoneal macrophages that were treated with CHX with or without
MG-132, a proteasome inhibitor, for 3 and 6 h after being administered 20 ng/ml IFN-g for 24 h. b-Actin was used as loading
control. C ). IP and immunostaining analysis of the relationship between LXR-a and polyubiquitin in mouse peritoneal
macrophages that were treated by 20 ng/ml IFN-g for 24 h (n = 5 in each group for Western blot). *P , 0.05 vs. control.

carotid underwent sham surgery as a control. Mice were LXR agonist and ER stress inhibitor attenuate
i.p. administered 100 mg anti-mouse IFN-g mAb or normal neointimal hyperplasia
rat IgG once per wk. All mice were euthanized at the
fourth wk after surgery. Hematoxylin and eosin staining in ER stress is the main cause of apoptosis (21). We assessed
Fig. 5A shows that neointimal hyperplasia, which is de- whether LXR activation could attenuate restenosis and
cided by medial index, became significant after carotid neointimal hyperplasia to explore the role of LXR-a in ER
ligation for 4 wk compared with sham surgery, whereas stress–dependent apoptosis of macrophages induced by
IFN-g mAb significantly attenuated neointimal hyper- IFN-g. ApoE2/2 mice with carotid ligation were i.p. ad-
plasia compared with IgG treatment. In addition, Fig. 5B ministered T0 (10 mg/kg) and 4-PBA (1 g/kg) twice per
reveals that IFN-g mAb also significantly reduced the in- wk after surgery. 4-PBA has been confirmed to decrease
filtration of CD68+ cells—a specific cell surface marker of the intima-to-media ratio by 50% after wire injury (10). In
macrophages—in the hyperplastic neointima compared our experiment, hematoxylin and eosin staining showed
with IgG treatment. We further evaluated apoptosis in that both LXR agonist T0 and ER stress inhibitor 4-PBA
hyperplastic neointima by using TUNEL and CD68 cos- treatment relieved neointimal hyperplasia sharply in 4-wk
taining. Figure 5B indicates that a great number of apo- carotid ligation compared with the injured group (Fig. 6A).
ptotic cells were found in injured arterial tissue. Moreover, In addition, there were no significant changes in body
most apoptotic cells were CD68+ macrophages in hyper- weight, plasma cholesterol, and triglyceride levels among
plastic neointima. IFN-g mAb treatment significantly re- these groups in our experiments (Supplemental Fig. 1).
duced macrophage apoptosis compared with IgG-treated Moreover, we also determined the apoptosis of CD68+
mice. These data confirm the important role of IFN-g and macrophages in carotid artery tissue to investigate the
inflammation in the development of neointimal hyper- molecular mechanisms that underlies the effects of LXR
plasia. Taken together, IFN-g promoted neointimal hy- agonist on neointimal hyperplasia. As shown in Fig. 6B,
perplasia, in part, by inducing macrophage infiltration and apoptotic macrophages—indicated as TUNEL+CD68+
facilitating cellular apoptosis. double-positive cells—were significantly increased in the

6 Vol. 31 December 2017 The FASEB Journal x www.fasebj.org ZHAO ET AL.

Downloaded from www.fasebj.org to IP 128.59.222.107. The FASEB Journal Vol., No. , pp:, August, 2017
Figure 4. IFN-g induces the degradation of LXR-a via the ubiquitin-proteasome system depending on LXR-a/PIAS1/STAT1 tripolymer
formation. A–C) IP and immunostaining analysis of the relationship between LXR-a, p-STAT1, PIAS1, and HDAC4 in macrophages that
were treated with 20 ng/ml IFN-g for 24 h. b-Actin was used as loading control. D) IP and immunostaining analysis of the relationship
between LXR-a, STAT1 and polyubiquitin in macrophages that were treated with or without Flu, a STAT-1 activation inhibitor, before
being administered 20 ng/ml IFN-g for 24 h. b-Actin was used as loading control. E) IP and immunostaining analysis of the relationship
between LXR-a, PIAS1, and polyubiquitin in macrophages that were treated with or without PIAS1 shRNA before being administered
20 ng/ml IFN-g for 24 h. b-Actin was used as loading control (n = 5 in each group for Western blot).

hyperplastic neointima after carotid ligation for 4 wk (2, 15, 25). Recent studies have indicated that IFN-g and ER
compared with the sham control group. After T0 or 4-PBA stress can critically affect neointimal hyperplasia after ar-
administration, both infiltration and apoptosis of CD68+ tery injury (6, 10). Our studies not only affirm the vital
macrophages were apparently decreased (P , 0.05). To effect of IFN-g and inflammation, but also define an un-
further determine the role of ER stress in macrophages and expected role for LXR-a degradation-related ER stress and
the effects of LXR agonist on neointimal hyperplasia, we apoptosis of macrophages in the development of neo-
labeled macrophages with P-eIF2a or CHO. Figure 6C, D intimal hyperplasia. Therefore, LXR-a may be a promising
reveals that carotid ligation for 4 wk significantly increased target to reverse vascular remodeling.
CD68+ macrophages in hyperplastic neointima, which in- Although drug-eluting stents have been applied for
dicates infiltration of inflammatory cells, whereas such an more than a dozen years in patients with PCI to prevent
infiltration was notably lessened after 4-PBA or T0 treat- restenosis, acute and chronic vessel occlusion rates are still
ment. In addition, the number of CD68+P-eIF2a+ and as high as 10–30% (5). As a result of artery wall injury and
CD68+CHOP+ double-positive macrophages was obvi- healing response, restenosis consists of 2 main processes:
ously increased in neointima after 4 wk of carotid ligation neointimal hyperplasia and vessel remodeling (26). Char-
compared with the sham control group; however, after acterized by accelerated accumulation of smooth muscle
4-PBA or T0 treatment, the number of CD68+P-eIF2a+ and cells, myofibroblasts, and macrophages, neointimal hy-
CD68+CHOP+ double-positive macrophages was signifi- perplasia is considered to be the main cause of chronic
cantly decreased. Taken together, these results indicate that restenosis after PCI (27). Inflammatory reaction triggered
4-PBA and T0 relieved neointimal hyperplasia by reducing by artery injury leads to the recruitment and accumulation
ER stress and apoptosis of macrophages in ApoE2/2 mice. of macrophages into the injured lesion, which maintains
inflammatory cascade reactions and further accelerates the
development of neointimal hyperplasia (2).
DISCUSSION As a well-known inflammatory cytokine, IFN-g is in-
volved in the development of several cardiovascular
Characterized by neointimal hyperplasia, restenosis is diseases. It has been confirmed that IFN-g promotes
the most frequent adverse outcome in patients after PCI neointimal hyperplasia by inducing expression of

IFN-g AGGRAVATES NEOINTIMAL HYPERPLASIA BY INDUCING ER STRESS AND APOPTOSIS IN MACROPHAGES 7

Downloaded from www.fasebj.org to IP 128.59.222.107. The FASEB Journal Vol., No. , pp:, August, 2017
Figure 5. IFN-g mAb attenuates injury-induced neointimal hyperplasia and inhibits apoptosis in injury-induced neointimal lesion.
A) Representative hematoxylin and eosin staining (HE) of sham surgery or injured common carotid artery tissues of ApoE2/2
mice that were treated with IFN-g mAb or normal rat IgG. B) Representative immunofluorescence staining of the intracuff
carotid artery with a CD68 (green) Ab and TUNEL (red) in ApoE2/2 mice that were treated with IFN-g mAb or normal rat IgG.
CD68+ TUNEL+ double-positive macrophages are shown by the yellow area (arrow; n = 10 in each group). *P , 0.05 vs. control.

inflammatory genes and cell signaling molecules, such as increased in hyperplastic neointima tissue after carotid liga-
ICAM-1 and VCAM-1, in macrophages and dysregulating tion, and treatment of ER stress inhibitor 4-PBA significantly
eNOS activity in endothelial cells, which, in turn, accelerates decreased the number of these apoptotic macrophages
the accumulation of smooth muscle cells and neointimal that demonstrate the vital role of ER stress–induced ap-
hyperplasia (9, 28). In the present study, we observed that optosis of macrophages in neointimal hyperplasia (Fig. 6B).
IFN-g was also involved in neointimal hyperplasia by in- In addition, CD68+CHOP+ and CD68+p-eIF2a+ double-
ducing apoptosis of macrophages. Blockage of IFN-g by Ab positive cells were largely accumulated in neointimal hy-
neutralization efficiently alleviated the infiltration and apo- perplastic lesions (Fig. 6C, D), which further indicated that
ptosis of macrophages in injured arterial tissues and im- ER stress was present in macrophages in the development
proved neointimal hyperplasia in ApoE2/2 mice. of restenosis. Treatment with 4-PBA significantly de-
It has been reported that macrophages participate in the creased the number of CD68+CHOP+ or CD68+p-eIF2a+
restenosis process via the release of cytokines, metal- double-positive macrophages and significantly improved
loproteinases, and growth factors (4). In the present study, neointimal hyperplasia simultaneously. Taken together,
CD68+TUNEL+ double-positive cells were significantly these findings suggest that ER stress–dependent apoptosis

8 Vol. 31 December 2017 The FASEB Journal x www.fasebj.org ZHAO ET AL.

Downloaded from www.fasebj.org to IP 128.59.222.107. The FASEB Journal Vol., No. , pp:, August, 2017
Figure 6. LXR agonist and ER stress inhibitor attenuate neointimal hyperplasia. A) Hematoxylin and eosin staining (HE) of sham
surgery or injured common carotid artery tissues in ApoE2/2 knockout mice with or without T0 or PBA treatment for 4 wk after carotid
(continued on next page)

IFN-g AGGRAVATES NEOINTIMAL HYPERPLASIA BY INDUCING ER STRESS AND APOPTOSIS IN MACROPHAGES 9

Downloaded from www.fasebj.org to IP 128.59.222.107. The FASEB Journal Vol., No. , pp:, August, 2017
Figure 7. Schematic diagram of the molecular mechanisms by which IFN-g aggravates neointimal hyperplasia.

of macrophages plays an important role in the develop- regulator of ER homeostasis in macrophages, which is
ment of neointimal hyperplasia (Fig. 7). consistent with the study by Erbay et al. (31). Still, the
It has been proved that the activation of nuclear receptor underlying mechanisms of LXR-a–mediated ER ho-
signaling molecules, including peroxisome proliferator- meostasis in macrophages remain poorly understood.
activated receptors and LXR, ameliorates neointimal Ligand-free LXR-a can be rapidly degraded via the
hyperplasia via anti-inflammation and antiproliferation ubiquitin-proteasome system (24). In IFN-g–stimulated
(29, 30). In 2010, Erbay et al. (31) first demonstrated the brain astrocytes, LXR-a and LXR-b can be post-
central roles for LXR-a in the regulation of ER homeostasis transcriptionally modified via SUMOylation by HDAC4
of macrophages in the development of cardiovascular or PIAS1, 2 SUMO E3 ligases (23). To explore the un-
diseases. In the present study, we observed that the deg- derlying mechanism of accelerated degradation induced by
radation of LXR-a, but not LXR-b, was accelerated by IFN- IFN-g in macrophages, we detected the interaction protein
g, which induced ER tress and subsequent apoptosis of with LXR-a by using co-IP assay. Figure 3 shows that we
macrophages. LXR agonist T0 ameliorated ER stress and detected the combination of PIAS1, but not HDAC4, and
apoptosis of macrophages induced by IFN-g in vitro. LXR-a in IFN-g–treated macrophages. STAT1-specific in-
Furthermore, T0 relieved ER stress in macrophages and hibitor, Flu, or PIAS1 shRNA notably decreased the ex-
apoptosis, in part, in injured arterial tissues and improved pression of polyubiquitin and LXR-a at the protein level
experimental neointimal hyperplasia in vivo. LXR usually induced by IFN-g in macrophages. Detailed mechanisms
functions as a lipid metabolism regulator, and its activa- of how the combination of LXR-a and PIAS1 promoted its
tion also has obvious anti-inflammatory effects in ubiquitination remain unclear. Our preliminary data
the process of cardiovascular diseases (12). Of interest, show that knockdown of ubiquitin-conjugating enzyme 9
there was no significant difference in cholesterol and (UBC9), the unique E2 SUMOylation ligase, not only
triglyceride levels of plasma (Supplemental Fig. 1) in the completely abolished the ubiquitination of LXR-a induced
T0, PBA, and control groups, which indicates that LXR by IFN-g, but also increased its expression in macro-
agonist relieved neointimal hyperplasia independent of phages (data not shown). These data indicate that the
changes in the lipid profile. Here, our data show that SUMOylation of LXR-a induced by IFN-g via PIAS1
LXR-a, but not LXR-b, is an important sensor and promoted the instability and degradation of LXR-a by

arteries ligation (CAL; n = 10 in each group). B) Representative immunofluorescence staining of the carotid artery with a CD68
(green) Ab and TUNEL (red) in sham surgery or injured common carotid artery tissues in ApoE2/2 knockout mice with or
without T0 or PBA treatment for 4 wk after CAL. CD68+ TUNEL+ double-positive macrophages are shown by the yellow area (arrow;
n = 10 in each group). C) Immunofluorescent double staining of vascular tissue with Abs against CD68 and P-eIF2a, green in panel C
indicates CD68, red indicates P-eIF2a, and blue indicates DAPI-stained cellular nuclei. CD68+P-eIF2a+ double-positive macrophages
are shown by the yellow area (arrow; n = 10 in each group). D) Immunofluorescent double staining of vascular tissue with Abs against
CD68 and CHOP, green in panel D indicates CD68, red indicates CHOP, and blue indicates DAPI-stained cellular nuclei. CD68
+
CHOP+ double-positive macrophages are shown by the yellow area (arrow; n = 10 in each group). *P , 0.05 vs. control.

10 Vol. 31 December 2017 The FASEB Journal x www.fasebj.org ZHAO ET AL.

Downloaded from www.fasebj.org to IP 128.59.222.107. The FASEB Journal Vol., No. , pp:, August, 2017
proteasome. Taken together, the degradation LXR-a in 11. Noda, T., Maeda, K., Hayano, S., Asai, N., Enomoto, A., Takahashi, M.,
macrophages induced by well-known inflammatory cyto- and Murohara, T. (2015) New endoplasmic reticulum stress
regulator, gipie, regulates the survival of vascular smooth muscle
kine IFN-g via the ubiquitin-proteasome system depended cells and the neointima formation after vascular injury. Arterioscler.
on LXR-a/PIAS1/STAT1 tripolymer formation. Thromb. Vasc. Biol. 35, 1246–1253
In conclusion, IFN-g accelerated the degradation 12. Hong, C., and Tontonoz, P. (2014) Liver X receptors in lipid
metabolism: opportunities for drug discovery. Nat. Rev. Drug Discov.
of LXR-a in macrophages via LXR-a/PIAS1/STAT1 13, 433–444
tripolymer-dependent pathway. Degradation of LXR-a 13. Sun, X., Haas, M. E., Miao, J., Mehta, A., Graham, M. J., Crooke, R. M.,
induced by IFN-g triggered ER stress and apoptosis in Pais de Barros, J. P., Wang, J. G., Aikawa, M., Masson, D., and
macrophages, which leads to aggravated neointimal hy- Biddinger, S. B. (2016) Insulin dissociates the effects of liver X
receptor on lipogenesis, endoplasmic reticulum stress, and
perplasia. LXR agonist efficiently improved neointimal inflammation. J. Biol. Chem. 291, 1115–1122
hyperplasia and restenosis. To explore potential new 14. Rong, X., Albert, C. J., Hong, C., Duerr, M. A., Chamberlain, B. T.,
therapies to ameliorate inflammatory disorders, it is im- Tarling, E. J., Ito, A., Gao, J., Wang, B., Edwards, P. A., Jung, M. E.,
Ford, D. A., and Tontonoz, P. (2013) LXRs regulate ER stress and
portant to understand how LXR activation exerts its ef- inflammation through dynamic modulation of membrane
fects on inflammatory diseases. phospholipid composition. Cell Metab. 18, 685–697
15. Alfonso, F., Byrne, R. A., Rivero, F., and Kastrati, A. (2014) Current
treatment of in-stent restenosis. J. Am. Coll. Cardiol. 63, 2659–2673
ACKNOWLEDGMENTS 16. Zhang, X., Gonçalves R., and Mosser, D. M. (2008) The isolation and
characterization of murine macrophages. Curr. Protoc. Immunol.
The author would like to express heartfelt gratitude to her Chapter 14, Unit 14.1
husband for valuable encouragement and spiritual support during 17. Kumar, A., and Lindner, V. (1997) Remodeling with neointima
formation in the mouse carotid artery after cessation of blood flow.
the study. This work was supported by the National Natural
Arterioscler. Thromb. Vasc. Biol. 17, 2238–2244
Science Foundation of China (Grants 91639301, 91339116, and 18. Dai, X., Ding, Y., Liu, Z., Zhang, W., and Zou, M. H. (2016)
81300226). The authors declare no conflicts of interest. Phosphorylation of CHOP (C/EBP homologous protein) by the
AMP-activated protein kinase alpha 1 in macrophages promotes
CHOP degradation and reduces injury-induced neointimal disrup-
AUTHOR CONTRIBUTIONS tion in vivo. Circ. Res. 119, 1089–1100
19. Wang, Q., Zhang, M., Ding, Y., Wang, Q., Zhang, W., Song, P., and
Q. Zhao, Z. Yuan, and J. Zhou designed experiments; Zou, M.H. (2014) Activation of NAD(P)H oxidase by tryptophan-
Q. Zhao, D. Zhou, H. You, Y. Zhang, and Y. Tian performed derived 3-hydroxykynurenine accelerates endothelial apoptosis and
dysfunction in vivo. Circ. Res. 114, 480–492
experiments; Q. Zhao, B. Lou, N. Guo, X. Chen, Y. Liu, and 20. Schroder, K., Hertzog, P. J., Ravasi, T., and Hume, D. A. (2004)
Y. Wu analyzed data; and Q. Zhao and J. Zhou wrote the paper. Interferon-gamma: an overview of signals, mechanisms and functions.
J. Leukoc. Biol. 75, 163–189
21. Szegezdi, E., Logue, S. E., Gorman, A. M., and Samali, A. (2006)
REFERENCES Mediators of endoplasmic reticulum stress-induced apoptosis. EMBO
Rep. 7, 880–885
1. Lee, J. Y., Park, D. W., Kim, Y. H., Yun, S. C., Kim, W. J., Kang, S. J., Lee, 22. Tabas, I., and Ron, D. (2011) Integrating the mechanisms of apoptosis
S. W., Lee, C. W., Park, S. W., and Park, S. J. (2011) Incidence, induced by endoplasmic reticulum stress. Nat. Cell Biol. 13, 184–190
predictors, treatment, and long-term prognosis of patients with 23. Lee, J. H., Park, S. M., Kim, O. S., Lee, C. S., Woo, J. H., Park, S. J., Joe,
restenosis after drug-eluting stent implantation for unprotected left E. H., and Jou, I. (2009) Differential SUMOylation of LXRalpha and
main coronary artery disease. J. Am. Coll. Cardiol. 57, 1349–1358 LXRbeta mediates transrepression of STAT1 inflammatory signaling
2. Park, S. J., Kang, S. J., Virmani, R., Nakano, M., and Ueda, Y. (2012) In- in IFN-gamma-stimulated brain astrocytes. Mol. Cell 35, 806–817
stent neoatherosclerosis: a final common pathway of late stent failure. 24. Kim, K. H., Yoon, J. M., Choi, A. H., Kim, W. S., Lee, G. Y., and Kim, J. B.
J. Am. Coll. Cardiol. 59, 2051–2057 (2009) Liver X receptor ligands suppress ubiquitination and degradation
3. Shah, P. K. (2003) Inflammation, neointimal hyperplasia, and of LXRalpha by displacing BARD1/BRCA1. Mol. Endocrinol. 23, 466–474
restenosis: as the leukocytes roll, the arteries thicken. Circulation 25. Cassese, S., Byrne, R. A., Tada, T., Pinieck, S., Joner, M., Ibrahim, T.,
107, 2175–2177 King, L. A., Fusaro, M., Laugwitz, K. L., and Kastrati, A. (2014)
4. Bayes-Genis, A., Campbell, J. H., Carlson, P. J., Holmes, D. R., Jr., and Incidence and predictors of restenosis after coronary stenting in 10
Schwartz, R. S. (2002) Macrophages, myofibroblasts and neointimal 004 patients with surveillance angiography. Heart 100, 153–159
hyperplasia after coronary artery injury and repair. Atherosclerosis 163, 26. Costa, M. A., and Simon, D. I. (2005) Molecular basis of restenosis and
89–98 drug-eluting stents. Circulation 111, 2257–2273
5. Danenberg, H. D., Fishbein, I., Gao, J., Mönkkönen, J., Reich, R., Gati, 27. Toutouzas, K., Colombo, A., and Stefanadis, C. (2004) Inflammation
I., Moerman, E., and Golomb, G. (2002) Macrophage depletion by and restenosis after percutaneous coronary interventions. Eur. Heart J.
clodronate-containing liposomes reduces neointimal formation after 25, 1679–1687
balloon injury in rats and rabbits. Circulation 106, 599–605 28. Hui, D. Y. (2008) Intimal hyperplasia in murine models. Curr. Drug
6. Rose, M. L. (2007) Interferon-gamma and intimal hyperplasia. Circ. Targets 9, 251–260
Res. 101, 542–544 29. Blaschke, F., Leppanen, O., Takata, Y., Caglayan, E., Liu, J., Fishbein,
7. Tellides, G., and Pober, J. S. (2007) Interferon-gamma axis in graft M. C., Kappert, K., Nakayama, K. I., Collins, A. R., Fleck, E., Hsueh, W. A.,
arteriosclerosis. Circ. Res. 100, 622–632 Law, R. E., and Bruemmer, D. (2004) Liver X receptor agonists suppress
8. Tavakoli, N. N., Harris, A. K., Sullivan, D. R., Hambly, B. D., and Bao, S. vascular smooth muscle cell proliferation and inhibit neointima
(2012) Interferon-g deficiency reduces neointimal formation in a formation in balloon-injured rat carotid arteries. Circ. Res. 95, e110–e123
model of endoluminal endothelial injury combined with atherogenic 30. Guzeloglu, M., Reel, B., Atmaca, S., Bagrıyanık, A., and Hazan, E. (2012)
diet. Int. J. Mol. Med. 30, 545–552 The effects of PPARg agonist rosiglitazone on neointimal hyperplasia in
9. Wang, Y., Bai, Y., Qin, L., Zhang, P., Yi, T., Teesdale, S. A., Zhao, L., rabbit carotid anastomosis model. J. Cardiothorac. Surg. 7, 57
Pober, J. S., and Tellides, G. (2007) Interferon-gamma induces hu- 31. Erbay, E., Babaev, V. R., Mayers, J. R., Makowski, L., Charles, K. N.,
man vascular smooth muscle cell proliferation and intimal expansion Snitow, M. E., Fazio, S., Wiest, M. M., Watkins, S. M., Linton, M. F., and
by phosphatidylinositol 3-kinase dependent mammalian target of Hotamisligil, G. S. (2009) Reducing endoplasmic reticulum stress
rapamycin raptor complex 1 activation. Circ. Res. 101, 560–569 through a macrophage lipid chaperone alleviates atherosclerosis.
10. Ishimura, S., Furuhashi, M., Mita, T., Fuseya, T., Watanabe, Y., Nat. Med. 15, 1383–1391
Hoshina, K., Kokubu, N., Inoue, K., Yoshida, H., and Miura, T. (2014)
Reduction of endoplasmic reticulum stress inhibits neointima Received for publication April 12, 2017.
formation after vascular injury. Sci. Rep. 4, 6943 Accepted for publication July 25, 2017.

IFN-g AGGRAVATES NEOINTIMAL HYPERPLASIA BY INDUCING ER STRESS AND APOPTOSIS IN MACROPHAGES 11

Downloaded from www.fasebj.org to IP 128.59.222.107. The FASEB Journal Vol., No. , pp:, August, 2017
IFN-γ aggravates neointimal hyperplasia by inducing
endoplasmic reticulum stress and apoptosis in macrophages by
promoting ubiquitin-dependent liver X receptor-α degradation
Qiang Zhao, Dong Zhou, Hongjun You, et al.

FASEB J published online August 10, 2017

Access the most recent version at doi:10.1096/fj.201700327R

Supplemental http://www.fasebj.org/content/suppl/2017/08/10/fj.201700327R.DC1
Material

Subscriptions Information about subscribing to The FASEB Journal is online at

http://www.faseb.org/The-FASEB-Journal/Librarian-s-Resources.aspx

Permissions Submit copyright permission requests at:

http://www.fasebj.org/site/misc/copyright.xhtml

Email Alerts Receive free email alerts when new an article cites this article - sign up at
http://www.fasebj.org/cgi/alerts

© FASEB

Downloaded from www.fasebj.org to IP 128.59.222.107. The FASEB Journal Vol., No. , pp:, August, 2017
 A

B C

Suppl
ementalFi
gur
e1.Bodywei
ghtandl
ipi
dpar
amet
ersofpl
asmai
nApoE-
/-knockoutmi
cewi
thorwi
thoutT0orPBAt
reat
mentf
or
4weeksaf
terCAL.
Af
ter1week’
shighf
atdi
et,commoncar
oti
dar
ter
ieswer
eli
gat
edasdescr
ibedi
nmet
hods.Bodywei
ght(
A),pl
asmachol
est
erol(
B)andt
ri
glycer
ide(
C)l
evel
s
weredet
ect
ed.
 A HDAC4
B HDAC4

LXR-
α LXR-
α

β-
act
in β-
act
in

I
FN-
γ - + - + I
FN-
γ - + - +
TSA - - + + LMK235 - - + +

Suppl
ement
alFi
gur
e2.HDAC4doesnott
akepar
tinLXR-
αdegr
adat
ioni
nducedbyI
FN-
γinmacr
ophages.
Repr
esentat
ivewest
ernblotanal
ysi
sofLXR-αandHDAC4inmouseper
itonealmacr
ophagest
reat
edby20ng/
mLI
FN-
γfor24hwi
thorwi
thoutHDAC4i
nhi
bit
ors
TSA(A)orLMK135(B).β-
actinwasusedasaloadi
ngcont
rol
.