Shikonin inhibits oxidized LDL-induced monocyte adhesion by suppressing NFkB activation via up-regulation of PI3K/Akt/Nrf2-dependent antioxidation in EA.hy926 endothelial cells

Shikonin inhibits oxidized LDL-induced monocyte adhesion by suppressing NFκB activation

via up-regulation of PI3K/Akt/Nrf2-dependent antioxidation in EA.hy926 endothelial cells

Chin-Shiu Huang

, Ai-Hsuan Lin

, Ting-Chun Yang

, Kai-Li Liu

, Haw-Wen Chen

,

Chong-Kuei Lii

a

Department of Health and Nutrition Biotechnology, Asia University, Taichung 413, Taiwan

b

Department of Medical Research, China Medical University Hospital, China Medical

University, Taichung 404, Taiwan

c

Department of Nutrition, China Medical University, Taichung 404, Taiwan

d

School of Nutrition, Chung Shan Medical University, Taichung 402, Taiwan

*

Corresponding authors at: Department of Nutrition, China Medical University, 91 Hsueh-

Shih Road, Taichung 404, Taiwan. Tel.: +886 4 2205 3366 ext. 7519; fax: +886 4 2206 2891.

1

Authors contributed equally to this work.

E-mail addresses: [email protected] (C.S. Huang), [email protected] (A.H. Lin),

[email protected] (T.C. Yang), [email protected] (K.L. Liu),

[email protected] (H.W. Chen), [email protected] (C.K. Lii).

ABSTRACT

Oxidized low-density lipoprotein (oxLDL) is a key contributor to atherogenesis through

multiple mechanisms, including the reactive oxygen species (ROS)-mediated nuclear factor-

kappaB (NFκB) signaling pathway. Although shikonin, one of the main active components

isolated from the Chinese herb Lithospermum erythrorhizon, has been shown to possess

cardioprotective, antioxidative, and anti-inflammatory effects, the mechanisms underlying

these actions are not well understood. In this study, we used EA.hy926 endothelial-like cells

to examine the anti-atherogenic activity of shikonin. Shikonin (0-1 µM) concentration-

dependently induced heme oxygenase-1, glutamate cysteine ligase modifier subunit, catalase,

superoxide dismutase 1, glutathione peroxidase 1, and glutathione reductase protein and mRNA

expression and glutathione content via activation of the phosphatidylinositol 3-kinase

(PI3K)/Akt/Nrf2 signaling pathway. In the presence of oxLDL (40 µg/ml), shikonin

pretreatment reversed oxLDL-induced ROS production, antioxidant response element

reporter activity, NFκB nuclear translocation, and intercellular adhesion molecule (ICAM)-1

and E-selectin expression and suppressed the increase of monocyte adhesion to endothelial

cells. Nrf2 knockdown by using RNA interference attenuated the ability of shikonin to inhibit

oxLDL-induced NFκB DNA binding activity, adhesion molecule expression, and monocyte

adhesion. Taken together, these results suggest that shikonin protects against oxLDL-induced

endothelial damage by suppressing ROS/NFκB-mediated ICAM-1 and E-selectin expression

via up-regulation of PI3K/Akt/Nrf2-dependent antioxidant enzyme expression.

Keywords: Adhesion molecules; Antioxidant defense; Nuclear factor-kappaB (NFκB);

Nuclear factor erythroid 2-related factor 2 (Nrf2); oxidized LDL; Shikonin

1. Introduction

Atherosclerosis, a major cause of death worldwide, is a chronic inflammatory disease

of the arterial wall that is initiated by the accumulation of oxidized low-density lipoprotein

(oxLDL) and inflammatory cells such as monocytes [1]. oxLDL plays multiple

proatherogenic roles in vascular endothelial disorders by promoting monocyte and leukocyte

recruitment; increasing endothelium apoptosis, thrombosis, and vascular smooth muscle cell

proliferation; and stimulating the release of inflammatory mediators, such as reactive oxygen

species (ROS) and cytokines [2]. Some evidence has provided insights into the mechanisms

by which oxLDL-induced cellular ROS mediate monocyte adhesion to endothelial cells [3,4].

For example, an excess of ROS activates intracellular signaling, including tyrosine kinase,

phosphatidylinositide 3-kinase (PI3K)/Akt, and mitogen-activated protein kinases (MAPK),

and subsequently activates nuclear factor-kappaB (NFκB) [5,6]. Activated NFκB is then

released from inhibitory kappa B (IκB), translocates into the nucleus, binds to the κB binding

site, and up-regulates the expression of monocyte chemotactic protein-1 and several

endothelial cell-specific adhesion genes, such as intercellular adhesion molecule (ICAM)-1,

vascular cell adhesion molecule 1, and E-selectin, and finally triggers the progression of

atherosclerosis [7]. A promising strategy for atherosclerosis prevention is therefore inhibition

of oxLDL-mediated ROS production.

To counteract oxidative stress, an adaptive response has evolved in aerobic organisms.

Nuclear factor E2-related factor 2 (Nrf2) is a redox-sensitive transcription factor known for

its pivotal role not only in antioxidant defense but also in anti-inflammation and drug

metabolism [8]. In the resting state, Nrf2 is associated to Kelch like-ECH-associated protein

(Keap) 1 and is sequestered in the cytoplasm. In response to oxidative or electrophilic stress,

Nrf2 is dissociated from the Nrf2-Keap1 complex through oxidative or covalent modification

of cysteine residues of Keap1 [9] or through direct phosphorylation of Nrf2 by PI3K/Akt,

MEK/MAPK, protein kinase C, and cyclic-AMP-activated protein kinase [10-13]. Recently,

accumulating evidence has indicated that Nrf2-Keap1 is regulated in a more complex way.

For instance, the interaction of proteins such as p21 and p62 with Nrf2 or Keap1 promotes

their disassociation [14,15]. Moreover, several microRNAs such as miR-144, miR28, and

miR200a may change Nrf2-mediated signaling by targeting Nrf2 or Keap1 mRNA [16-18].

Free Nrf2 then quickly translocates into the nucleus, forms a heterodimer with small Maf, and

binds to a conserved sequence named the antioxidant-response element (ARE). Nrf2-ARE

signaling regulates the expression of an array of antioxidant defense and drug detoxification

genes, such as glutamate-cysteine ligase (GCL), superoxide dismutase (SOD) 1, glutathione

(GSH) peroxidase (GPx) 1, GSH reductase (GSR), heme oxygenase (HO) 1, GSH S-

transferases, NADPH quinone oxidoreductase, UDP-glucuronyl transferase, and p-

glycoprotein [19]. GCL is a heterodimer composed of a catalytic subunit (GCLC) and a

modifier subunit (GCLM) that catalyzes the first and rate-limiting step of de novo GSH

synthesis [20].

Lithospermum erythrorhizon (Siebbold et Zuccarini), a Chinese medicinal herb, has

been widely used for the treatment of burns, anal ulcers, hemorrhoids, dermatitis, and

infectious diseases for thousands of years [21]. Shikonin, a naphthoquinone pigment, has

been reported to be the most active component responsible for such pharmacological

functions of L. erythrorhizon [22]. Shikonin possesses diverse biological functions such as

antioxidative, anti-inflammatory, antithrombotic, antimicrobial, wound healing, and

antitumorigenic properties. In an autoimmune-prone strain of mice, oral administration of 20

mg/kg shikonin reduces mononuclear cell infiltration in the perivascular and interstitial

regions and down-regulates the mRNA expression of ICAM-1 and VCAM-1 in the kidney

[23]. In LPS-activated macrophages, shikonin (0.3-3 µM) effectively inhibits nitric oxide

production by suppressing the mRNA and protein expression of inducible nitric oxide

synthase [24]. Recent studies have reported that shikonin exerts anti-inflammatory and

anticancer effects via modulation of the Nrf2 signaling pathway [25,26]. These potent anti-

inflammatory and antioxidative activities also account for the role of shikonin in

cardiovascular protection [22]. Although several cellular proteins and active targets of

shikonin have been proposed, the precise molecular protection mechanism of shikonin has

not been elucidated. Therefore, we were interested in understanding whether the anti-

atherogenic protective function of shikonin works through the Nrf2/ARE pathway and

whether this subsequently inhibits NFκB -regulated inflammatory events.

In this study, we used an in vitro model of oxLDL-induced human endothelial

EA.hy926 cell dysfunction to examine the protective effect of shikonin on oxLDL-induced

expression of ICAM-1 and E-selectin and the subsequent adherence of human monocytic

THP-1 cells to endothelial cells. We also investigated the possible mechanisms underlying

these actions.

2. Materials and methods

2.1. Chemicals

Shikonin (>97%) was obtained from Merck (Darmstadt, Germany). Dulbecco’s modified Eagle medium (DMEM), RPMI 1640, and penicillin-streptomycin solution were from Gibco/BRL (Grand Island, NY, USA). Fetal bovine serum was purchased from HyClone (Logan, UT, USA). PD98059 (ERK inhibitor) and SB203580 (MAPK p38 inhibitor) were from TOCRIS (Ellisville, MO). LY294002 (PI3K inhibitor), GSH, HEPES, heparin, CuSO

, 2’,7’-dichlorodihydrofluorescein diacetate (DCFDA), and all other chemicals were obtained from Sigma-Aldrich (St. Louis, MO, USA). Antibodies against PI3K p85α, Akt, phospho-Akt (S473), phospho-Akt (T308), p38, phospho-p38, ERK, phospho-ERK, PARP, and ICAM-1 were obtained from Cell Signaling Technology (Danvers, MA, USA); E-selectin was from BioVision ( Mountain View, CA, USA); HO-1 was from Merck (Darmstadt, Germany);

GCLM, GCLC, IκBα, JNK, phospho-JNK, and Nrf2 were from Santa Cruz Biotechnology (Santa Cruz, CA, USA); NFκB p65 was from BD Biosciences (San Jose, CA, USA); catalase, SOD-1, SOD-2, GPx-1, and GSR were from Gene Tex (San Antonio, TX, USA).

2.2. LDL isolation and oxLDL preparation

LDL was isolated from plasma of healthy volunteers, and oxLDL was prepared as

described previously [27]. Written informed consent as approved by the Review Board for

Human Research of the Chung Shan Medical University was signed by all participants. The

generation of thiobarbituric acid-reactive substances was monitored by the fluorometric

method [28], and the values of malondialdehyde equivalents were 3.12±0.58 and 63.5±5.16

nmol/mg protein of native LDL and CuSO

-treated LDL, respectively.

2.3. Cell cultures and treatments

The human endothelial cell line EA.hy926 (ATCC CRL-2922), which is derived by

fusing human umbilical vein endothelial cells (HUVECs) with A549 human lung carcinoma

cells, was grown in DMEM supplemented with 1.5 g/l NaHCO

, 10% fetal bovine serum, 100

units/ml penicillin, and 100 µg/ml streptomycin; the human monocytic cell line THP-1

(ATCC TIB-202) was grown in a RPMI 1640 medium supplemented with 1.5 g/l NaHCO

,

4.5 g/l glucose, 10% fetal bovine serum, 100 units/ml penicillin, and 100 µg/ml streptomycin.

Cells were incubated at 37

C in a 5% CO

humidified incubator. For fluorescence labeling,

THP-1 cells were treated with 1 µM 2',7'-bis(2-carboxyethyl)-5(6)-carboxyfluorescein

acetoxymethyl ester (BCECF-AM, Molecular Probes, Eugene, OR, USA) for 30 min and

were then changed to the phenol red-free RPMI1640 medium supplemented with 10% FBS.

The cytotoxicity of shikonin was determined by (3-(4,5)-dimethylthiazol(2y1)-2,5-di-

phenyltetrazolium bromide (MTT) assay. Incubation of shikonin at 0-1 µM for 24 h in the

absence or presence of 40 µg/ml oxLDL for an additional 16 h did not change cell viability

(>90%) in EA.hy926 cells. Therefore, treatment with shikonin at 0-1 µM was chosen for the

following experiments. For each experiment, EA.hy926 cells were grown to 80% confluence

and were then treated with different concentrations of shikonin (0, 0.25, 0.5, and 1 µM) for

16 h followed by stimulation with 40 µg/ml oxLDL for an additional 24 h. Cells were

harvested in a lysis buffer containing 10 mM Tris-HCl, 320 mM sucrose, 5 mM EDTA, 0.1%

Triton X-100, 1 mM PMSF, and 1 mM dithiothreitol, pH 8.0. Cell extracts were then

centrifuged at 10,000g at 4

C for 15 min. The supernatants were recovered and the total

protein was analyzed by use of the Coomassie Plus protein assay kit (Pierce, Rockford, IL,

USA).

2.4. Western blotting analysis

Proteins were first electrophoresed in an SDS-polyacrylamide gel and were then

transferred to polyvinylidene fluoride membranes. After blocking with 5% nonfat milk,

membranes were probed with primary HO-1, GCLC, GCLM, GPx-1, GSR, SOD-1, SOD-2,

catalase, phospho-Akt (S473), phospho-Akt (T308), Akt, phospho-p38, p38, phospho-ERK,

ERK, phospho-JNK, JNK, ICAM-1, E-selectin, IκBα, NFκB p65, and Nrf2 antibodies. The

membranes were then probed with the horseradish peroxidase-labeled secondary antibody.

The bands were visualized by using an enhanced chemiluminescence kit (PerkinElmer Life

Science, Boston, MA, USA) and were quantitated with an AlphaImager 2000 (Alpha

Innotech).

2.5. Real-time quantitative reverse transcriptase-PCR

Total RNA of EA.hy926 cells was isolated by using Trizol reagent (Life Technologies,

Carlsbad, CA, USA) as described by the manufacturer. RNA extracts were reverse

transcribed with Moloney murine leukemia virus reverse transcriptase (Promega) for

synthesis of cDNA. cDNA was amplified with TaqMan® Universal PCR Master Mix primers

and probes and the reactions were performed in an ABI 7000 Real Time PCR (Applied

Biosystems). The primers for quantitative analyses of HO-1, GCLM, catalase, SOD-1, GPx-

1, GSR, and GAPDH mRNA were designed by using ProbeFinder version 2.46 for human

DNA from Roche Applied Science (Basel, Switzerland). Primers sequences are HO-1 (5'-

GGGTGATAGAAGAGGCCAAGA-3'; 5'-AGCTCCTGCAACTCCTCAAA-3'), GCLM (5'-

GAAGAAGATATTTTTCCTGTCATTGAT-3'; 5'-

CCATTCATGTATTGAAGAGTGAATTT-3'), catalase (5'-CCATCGCAGTTCGGTTCT-3';

5'-GGGTCCCGAACTGTGTCA-3'), SOD-1 (5’-TCATCAATTTCGAGCAGAAGG-3’; 5’-

CAGGCCTTCAGTCAGTCCTTT-3’), GPx-1 (5’-CAACCAGTTTGGGCATCAG-3’; 5’-

GTTCACCTCGCACTTCTCG-3’), GSR (5’-AACAACATCCCAACTGTGGTC-3’; 5’-

CCATATTTATGAATGGCTTCATCTT-3’), and GAPDH (5’-

AGCCACATCGCTCAGACAC-3’; 5’-GCCCAATACGACCAAATCC-3’). GAPDH was

used as an internal standard gene and the threshold cycle (Ct) of a test sample to a control

sample (2

) was used for relative quantification of target gene expression.

2.6. Measurement of intracellular reduced and oxidized GSH content

Cellular reduced and oxidized GSH content were determined as described with some

modifications [29]. After treatment with shikonin for 16 h, cells were harvested in 100 mM

potassium phosphate buffer, pH 7.4. After sonication, cell lysates were centrifuged at

10,000xg at 4

C for 20 min. One hundred μl of the cytosolic fraction was then reacted with

200 μl of a buffer containing 100 mM potassium phosphate and 10 mM 5,5’-dithiobis(2-

nitrobenzoic acid), pH 7.4, by gentle mixing, followed by the addition of 60 μl of 20% 5-

sulfosalicylic acid to cause acid precipitation. After centrifugation at 10,000xg and 4°C for 10

min, 100 μl of supernatant was used to analyze reduced and oxidized GSH content by use of

HPLC/MS. The detection limit of GSH and GSSG is 1.1 and 0.48 nmol/mL, respectively.

2.7. Plasmids, transfection, and luciferase assay

A 2×ARE/Luc fragment (pGL3-2×ARE/Luc) containing tandem repeats of double-

stranded oligonucleotides spanning the Nrf2 binding site, 5’-TGACTCAGCA-3’, was a kind

gift from Dr. BS Wung (Department of Applied Microbiology, National Chiayi University,

Chiayi, Taiwan). The subsequent transfection experiment was performed with Nanofectin

reagent (PAA, Pasching, Austria) with some modifications as described previously [30].

Briefly, cells were transiently transfected with 0.1 μg of pGL3-2×ARE/Luc plasmid and 0.1

μg of pCMV-β-galactosidase plasmid by using 0.4 μl of Nanofectin in OPTI-MEM medium

for 16 h. After transfection, cells were changed to DMEM medium and treated with shikonin

for an additional 16 h. The luciferase activity was measured by using a luciferase substrate kit

(Promega, Madison, WI) in a microplate luminometer (Tropix TR-717, Applied Biosystems).

For each experiment, luciferase activity was determined and was normalized to β-

galactosidase activity.

2.8. Reactive oxygen species measurement

The probe H

DCFDA was used to detect intracellular peroxide production as described

[30]. At the end of oxLDL treatment, EA.hy926 cells were incubated with 5 μM H

DCFDA

for an additional 10 min. The DCF fluorescence was detected by using a confocal microscope

detection system (Leica TCS SP2).

2.9. Monocyte adhesion assay

EA.hy926 cells in 12-well plates were allowed to grow to 80% confluence and were

then pretreated with shikonin for 16 h followed by incubation with 40 μg/ml oxLDL for

another 24 h. A total of 1 × 10

THP-1 cells labeled with BCECF-AM were added to each

well, and the cells were co-incubated with EA.hy926 for 30 min [31]. The cells were washed

and filled with cell culture medium, and the plates were sealed, inverted, and centrifuged at

100xg for 5 min to remove nonadherent THP-1. Bound THP-1 cells were lysed in a 1% SDS

solution, and the fluorescence intensity was determined in a fluoroscan ELISA reader

(FLX800, Bio-Tek, Winooski, VT, USA) with excitation and emission wavelengths of 480

nm and 520 nm, respectively.

2.10. RNA interference by siRNA of PI3K p85α and Nrf2

EA.hy926 cells were transfected with PI3K p85α (Santa Cruz Biotechnology, Santa

Cruz, CA, USA), Nrf2 siRNA (GE Dharmacon, Lafayette, CO, USA) or corresponding

nontargeting siRNA control by using DharmaFECT1 transfection reagents according to the

manufacturer’s protocols. After 16 h of transfection, cells were treated with shikonin

followed by incubation with 40 μg/ml oxLDL for the periods specified in the figures.

Knockdown efficiency was ascertained by immunoblotting.

2.11. Electrophoretic mobility shift assay

Nuclear extract proteins were prepared as described previously [27]. Electrophoretic

mobility shift assay (EMSA) was performed with the LightShift Chemiluminescent EMSA

kit (Pierce) to measure the effect of shikonin on oxLDL-induced nuclear NFκB binding

activity to the κB binding site. The biotin-labeled double-stranded oligonucleotides used were

as follows: 5’-GCCTGGGAAAGTCCCCTCAACT-3’; mut, 5’-

GCCTGGGAAAGTCgCCTCAACT-3’. After electrophoresis on a 6% Tris-boric acid-EDTA

polyacrylamide gel, the separated nuclear protein-DNA complex was electrotransferred to a

Hybond-N

membrane. The membranes were then incubated with streptavidin horseradish

peroxidase, and the nuclear protein-DNA bands were developed by using an enhanced

chemiluminescence kit.

2.12. Statistical analysis

The statistical analysis was performed by using SAS statistical software (Cary, NC, USA).

Data are expressed as means ± SD from at least three independent experiments. The

significance of the difference among mean values was determined by one-way analysis of

variance followed by Tukey’s test or by a two-tailed Student’s t-test. p<0.05 was taken to be

statistically significant.

3. Results

3.1. Shikonin increases antioxidant defense

Incubation of EA.hy926 cells with shikonin for 24 h increased the protein expression

of GCLM, catalase, SOD-1, GPx-1, and GSR in a concentration-dependent manner and in a

time-dependent manner up to 24 h at 1 μM of shikonin (Fig. 1A). Protein levels of GCLC and

SOD-2, however, were not changed by shikonin. Similarly, the mRNA levels of GCLM,

catalase, SOD-1, GPx-1, and GSR were time-dependently increased up to 24 h by shikonin

(Fig. 1B). A concentration-dependent up-regulation of HO-1 protein expression was also

noted in cells treated with shikonin. The time course changes of HO-1 mRNA and protein,

however, were different from those noted for the other antioxidant enzymes. The maximum

induction of HO-1 protein (Fig. 1A) and mRNA (Fig. 1B) by shikonin was observed at 8 and

4 h, respectively, after which expression gradually decreased. Consistent with the induction

of GCLM expression, shikonin increased cellular GSH contents in a concentration-dependent

manner (Fig. 1C). Compared with that in the control cells, GSH content was 68% higher in

cells treated with 1 μM shikonin (p<0.05).

3.2. Nrf2/ARE activation by shikonin via the PI3K/Akt pathway

Next, we examined the effect of shikonin on Nrf2 activation, which plays an important

role in modulating antioxidant and phase II detoxification enzyme gene expression. With

shikonin treatment, Nrf2 nuclear translocation was time-dependently increased up to 120 min

(Fig. 2A). By transient transfection with reporter constructs harboring the ARE, shikonin at

0.5 and 1 μM significantly increased luciferase activity (p<0.05), with a 160% increase with 1

μM shikonin treatment (Fig. 2B).

Nrf2 activation is known to be under the regulation of several kinases. Thus, we

investigated whether Akt, p38, ERK, and JNK were involved in the shikonin-induced

Nrf2/ARE activation and HO-1 protein expression. Compared with controls, phosphorylation

of Akt (S473 and T308), p38, and ERK1/2 increased in response to shikonin, and maximal

activation was noted at 30 min (Fig. 2C). No change in JNK phosphorylation was found.

Pretreatment with the PI3K/Akt inhibitor LY294002 and the p38 inhibitor SB203580

significantly reduced shikonin-induced Nrf2 nuclear translocation, whereas no effect of the

ERK inhibitor PD98059 was found (Fig. 2D). Attenuation of shikonin-induced ARE-

luciferase activity (Fig. 2E) as well as HO-1 expression (Fig. 2F) was found by PI3K/Akt

inhibitor but not by ERK and p38 inhibitors. By using siRNA to create a PI3K p85α

knockdown model, shikonin induction of Nrf2 nuclear translocation and HO-1 expression

(Fig. 2G) were abolished. Taken together, these results indicated that shikonin induction of

Nrf2-dependent gene transcription is likely associated with the activation of the

PI3K/Akt/Nrf2/ARE signaling pathway.

3.3. Shikonin attenuates oxLDL-induced ROS production, adhesion molecule expression,

THP-1 adhesion, and NFκB activation

In the presence of 40 μg/ml oxLDL, intracellular DCF fluorescence, which represents

the intracellular levels of ROS, time-dependently increased and reached a maximum at 30

min. Therefore, EA.hy926 cells treatment with oxLDL for 30 min was chosen for the

following experiments. Cells pretreated with shikonin had lower intracellular ROS formation

in a concentration-dependent manner (p<0.05, Fig. 3).

As shown by measuring ICAM-1 and E-selectin, adhesion molecules were upregulated

by 1.9-fold and 2.0-fold, respectively, after 24 h exposure to oxLDL treatment (Fig. 4A).

Accompanying this induction of adhesion molecules, the number of THP-1 cells adhering to

the EA.hy926 cells also increased with oxLDL treatment (Fig. 4B). Pretreatment with 1 μM

shikonin attenuated the oxLDL-induced ICAM-1 and E-selectin protein expression and

concentration-dependently suppressed THP-1 adhesion as well.

The expression of adhesion molecules and other inflammation-associated genes is

highly dependent on the activation of NFκB. To elucidate the upstream signaling pathways

involved in the oxLDL-mediated induction of adhesion molecule expression, we examined

the changes in oxLDL-induced NFκB activation by shikonin in EA.hy926 cells. As shown in

Fig 4C, oxLDL decreased cytosolic IκBα content up to 6 h in a time-dependent manner.

Accompanied by the reduction of IκBα, nuclear translocation of p65 was increased by

oxLDL, and this increase of nuclear p65 was attenuated by shikonin pretreatment (Fig. 4D).

3.4. Shikonin inhibition of adhesion molecule expression is Nrf2 dependent

We further confirmed the role of Nrf2 in shikonin’s inhibition of oxLDL-induced

adhesion molecule expression by using siRNA to create an Nrf2 knockdown model. The

effectiveness of Nrf2 knockdown was ascertained by Western blot assay (Fig. 5A). The effect

of shikonin on oxLDL-induced NFκB activation as measured by EMSA (Fig. 5B) and ICAM-

1 and E-selectin expression (Fig. 5C) were apparently abolished by knocking down of Nrf2 in

EA.hy926 cells. Shikonin suppression of THP-1 adhesion to EA.hy926 cells was also

abolished by knocking down Nrf2 (Fig. 5D).

4. Discussion

Oxidative damage and inflammation are two major risk factors of atherosclerosis.

Accordingly, enhancement of intracellular antioxidant defense and suppression of

inflammation are effective approaches to protecting endothelial cells against oxidant insult

such as oxLDL. In this study, we demonstrated for the first time that shikonin effectively

induces the expression of PI3K/Akt/Nrf2-dependent antioxidant genes, including HO-1,

SOD-1, catalase, GCLM, GPx-1, and GSR, and increases intracellular GSH content.

Furthermore, these events subsequently lead to inhibition of oxLDL-induced intracellular

ROS accumulation, NFκB activation, adhesion molecule expression, and monocyte adhesion.

The present study results also support the concept that Nrf2 activation prevents NFB-

mediated vascular inflammatory dysfunction [8,27,32].

The increase in ROS generation in endothelial cells by oxLDL is largely through

activation of NADPH oxidase, uncoupling of endothelial nitric oxide synthase signaling, and

damage to mitochondrial function [33-35]. For instance, oxLDL causes endothelial

dysfunction by inducing NADPH oxidase expression and activity, which lead to superoxide

anion and H

O

production [34,36]. oxLDL causes endothelial nitric oxide synthase

uncoupling, whereby endothelial nitric oxide synthase generates superoxide anion rather than

nitric oxide in the endothelium [35,37]. To maintain endothelial function, an effective

antioxidant defense is necessary to eliminate oxLDL-initiated oxidative damage. SOD is one

antioxidant that protects against superoxide-mediated cytotoxicity by catalyzing superoxide

anion to form H

O

, which is subsequently metabolized into water and molecular oxygen by

both GSH redox cycling and catalase [38]. Also, up-regulation of HO-1 is demonstrated to

inhibit H

O

-induced ROS generation and vascular endothelial dysfunction [39]. Here, we

showed that shikonin increases the expression of HO-1, SOD-1, GCLM, GPx-1, GSR, and

catalase at both the protein level and the mRNA level and also increases GSH contents in

EA.hy926 cells (Fig. 1). Furthermore, we showed that these actions of shikonin lead to a

reduction in oxLDL-induced ROS production (Fig. 3). These results suggest that the

protection against ROS-mediated oxidative injury in endothelial cells by shikonin can partly

be explained by its up-regulation of antioxidant enzymes and GSH synthesis.

The induction of HO-1, SOD-1, catalase, and GSH redox cycle gene transcription is

highly dependent on the presence of a conserved oligonucleotide sequence called the ARE.

Although several transcription factors bind to the ARE, Nrf2 is the key regulator of the

antioxidant defense against vascular disorders [40]. Increasing evidence indicates that

numerous phytochemicals protect cardiovascular function by up-regulating Nrf2/ARE

activation [41]. For instance, protection by isothiocyanates and andrographolide against

oxLDL- and TNFα-induced endothelial damage is mediated by the Nrf2-ARE pathway [27,

42]. Recently, Nrf2 was also reported to be involved in shikonin-induced NAD(P)H-quinone

oxidoreductase 1 transcription in human breast cancer cells [26] and heat shock protein 70

expression in human lymphoma U937 cells [43]. Herein, we found that shikonin increases

Nrf2 nuclear translocation (Fig. 2A) and ARE-driven luciferase activity (Fig. 2B), which

suggests that this naphthoquinone pigment is an effective Nrf2 activator.

Several protein kinases including PI3K/Akt, MAPK, protein kinase C, and cyclic-

AMP-activated protein kinase are involved in the activation of Nrf2 [10-13]. Evidence

indicates that shikonin modulates the activity of several protein kinases in various cells. For

instance, shikonin induction of apoptosis in Bcr/Abl chronic myelogenous leukemia cells

works through ROS-mediated JNK activation [44]. In the human gastric cancer cell line

AGS, shikonin induces apoptosis through the inhibition of PI3K/Akt and ERK activities and

the augmentation of p38 activity [45]. In the cervical cancer cell line HeLa, the colon cancer

cell line Hct116, the lung cancer cell line A549, and the hepatocellular carcinoma cell lines

Hep3B, BEL7402, and Huh7, shikonin increases ROS production, inactivates PI3K/Akt, and

subsequently promotes apoptosis [46,47]. Shikonin promotes glucose transporter 4

translocation into the plasma membrane via the PI3K/Akt pathway [48]. In this study, when

EA.hy926 endothelial cells were exposed to shikonin, Akt (S473 and T308), p38, and

ERK1/2 phosphorylation was increased (Fig. 2C). The increases in Nrf2 nuclear translocation

(Fig. 2D), ARE-reporter activity (Fig. 2E), and HO-1 expression (Fig. 2F) as well as the

decrease in oxLDL-induced THP-1 adhesion to endothelial cells (data not shown) by

shikonin were apparently reversed by PI3K/Akt inhibitor. Moreover, the importance of

PI3K/Akt on shikonin-induced Nrf2 activation and HO-1 expression was supported by using

a PI3K p85α knockdown model (Fig. 2G). ERK inhibitor exhibited minor effect. Although

p38 inhibitor SB203580 inhibited shikonin-induced Nrf2 nuclear translocation, SB203580

did not reverse HO-1 induction by shikonin. It needs to address that, in addition to the Nrf2,

transcription factor such as the AP-1 regulates HO-1 gene transcription as well [42]. Small-

molecule kinase inhibitors have been widely used in numerous researches, whereas the

specificity of these inhibitors has been questioned. For instance, p38 is not the only target of

SB203580. Cyclin G-associated kinase, casein kinase1δ, receptor-interacting protein 2, and

glycogen synthase kinase 3β are inhibited by SB203580 as well [49]. To clarify the actual

role of p38 in shikonin-induced Nrf2 activation, using a knockdown approach is warranted.

Taken together, these results support that the vascular protection by shikonin is associated

with activation of Nrf2 and its downstream antioxidant defense genes, mainly through the

PI3K/Akt signaling pathway.

Similar to Nrf2, NFκB is a redox-sensitive transcription factor involved in immune

response, inflammation, development, and apoptosis [50]. Evidence indicates that cross-talk

exists between the Nrf2 and the NFκB pathways. In vascular endothelium, improvement of

cellular antioxidation prevents tumor necrosis factor α-induced vascular inflammation by

inhibiting ROS activation of NFκB upstream modulators or by inducing NFκB

glutathionylation. These events in turn lead to suppression of NFκB-driven gene transcription

including ICAM, vascular cell adhesion molecule, and E-selectin, as well as the subsequent

monocyte adhesion to endothelial cells [8,30,51]. Tumor necrosis factor α-induced leukocyte

adhesion to EA.hy926 cells is enhanced by silencing Nrf2-driven HO-1 expression [52].

Furthermore, dextran sulfate sodium-induced colonic colitis is more severe in Nrf2 knockout

mice than in wild-type mice [53]. These findings strongly support that the anti-inflammatory

property of various of phytochemicals, e.g., sulforaphane, phenethyl isothiocyanate,

curcumin, hydroxytyrosol, and andrographolide, is positively related to their potency on Nrf2

activation [27,39,54,55]. In this study, oxLDL time-dependently decreased cytosolic

IκBα(Fig. 4C) and increased the nuclear p65 level (Fig. 4D). With shikonin pretreatment,

oxLDL-induced p65 nuclear translocation was abrogated (Fig. 4D), which led to the

attenuation of oxLDL-induced increase in ICAM-1 and E-selectin expression (Fig. 4A) and

THP-1 adhesion to EA.hy926 (Fig. 4B). With Nrf2 knockdown, inhibition by shikonin of

oxLDL-induced NFκB activation (Fig. 5B) and both ICAM-1 and E-selectin protein

expression (Fig. 5C) were reversed. Shikonin inhibition of THP-1 adhesion to EA.hy926 was

abrogated by Nrf2 silencing as well (Fig. 5D). These findings clearly indicate that shikonin

activation of Nrf2 plays an important role in preventing NFκB-mediated inflammatory events

in the endothelium.

In conclusion, this study demonstrates that shikonin effectively protects against

oxLDL-induced endothelial dysfunction including ROS release, adhesion molecule

expression, and monocyte adhesion, and that these effects are likely mediated by up-

regulation of PI3K/Akt/Nrf2-dependent antioxidant defense, which inhibits oxLDL-induced

NFκB activation (Fig. 6). These findings provide a potential mechanism for the anti-

atherogenic action of shikonin.

Acknowledgements

This work was supported by CMU101-ASIA-02 from the China Medical University

and Asia University, Taiwan.

Conflict of interest

The authors declare that they have no conflict of interest.

References

[1] Ross R. Atherosclerosis--an inflammatory disease. N Engl J Med 1999;340:115-26.

[2] Lusis AJ. Atherosclerosis. Nature 2000;407:233-41.

[3] Botham KM, Wheeler-Jones CP. Postprandial lipoproteins and the molecular regulation

of vascular homeostasis. Prog Lipid Res 2013;52:446-64.

[4] Gao P, Wang XM, Qian DH, Qin ZX, Jin J, Xu Q, et al. Induction of oxidative stress by

oxidized LDL via meprinα-activated epidermal growth factor receptor in macrophages.

Cardiovasc Res 2013;97:533-43.

[5] Mazière C, Salle V, Gomila C, Mazière JC. Oxidized low density lipoprotein enhanced

RANKL expression in human osteoblast-like cells. Involvement of ERK, NFkappaB and

NFAT. Biochim Biophys Acta 2013;1832:1756-64.

[6] Dandapat A, Hu C, Sun L, Mehta JL. Small concentrations of oxLDL induce capillary

tube formation from endothelial cells via LOX-1-dependent redox-sensitive pathway.

Arterioscler Thromb Vasc Biol 2007;27:2435-42.

[7] Wang YK, Hong YJ, Yao YH, Huang XM, Liu XB, Zhang CY, et al. 6-Shogaol protects

against oxidized LDL-induced endothelial injuries by inhibiting oxidized LDL-evoked

LOX-1 signaling. Evid Based Complement Alternat Med 2013;2013:503521.

[8] Kim JH, Choi YK, Lee KS, Cho DH, Baek YY, Lee DK, et al. Functional dissection

of Nrf2-dependent phase II genes in vascular inflammation and endotoxic injury using

Keap1 siRNA. Free Radic Biol Med 2012;53:629-40.

[9] Dinkova-Kostova AT, Holtzclaw WD, Cole RN, Itoh K, Wakabayashi N, Katoh Y, et al.

2002. Direct evidence that sulfhydryl groups of Keap1 are the sensors regulating

induction of phase 2 enzymes that protect against carcinogens and oxidants. Proc Natl

Acad Sci USA 2002;99:11908-13.

[10] Huang HC, Nguyen T, Pickett CB. Regulation of the antioxidant response element by

protein kinase C-mediated phosphorylation of NF-E2-related factor 2. Proc Natl Acad

Sci USA 2000;97:12475-80.

[11] Mo C, Wang L, Zhang J, Numazawa S, Tang H, Xiaoqiang T, et al. The crosstalk

between Nrf2 and AMPK pathways is important for the anti-inflammatory effect of

berberine in LPS-stimulated macrophages and endotoxin-shocked mice. Antioxid

Redox Signal 2013;20:574-88.

[12] Min KJ, Lee JT, Joe EH, Kwon TK. An IκBα phosphorylation inhibitor induces heme

oxygenase-1(HO-1) expression through the activation of reactive oxygen species

(ROS)-Nrf2-ARE signaling and ROS-PI3K/Akt signaling in an NF-κB-independent

mechanism. Cell Signal 2011;23:1505-13.

[13] Al-Huseini LM, Aw Yeang HX, Sethu S, Alhumeed N, Hamdam JM, Tingle Y, et al.

Nuclear factor-erythroid 2 (NF-E2) p45-related factor-2 (Nrf2) modulates dendritic

cell immune function through regulation of p38 MAPK-cAMP-responsive element

binding protein/activating transcription factor 1 signaling. J Biol Chem

2013;288:22281-8.

[14] Chen W, Sun Z, Wang XJ, Jiang T, Huang Z, Fang D, et al. Direct interaction between

Nrf2 and p21(Cip1/WAF1) upregulates the Nrf2-mediated antioxidant response. Mol

Cell 2009;34:663-73.

[15] Lau A, Wang XJ, Zhao F, Villeneuve NF, Wu T, Jiang T, et al. A noncanonical

mechanism of Nrf2 activation by autophagy deficiency: direct interaction between

Keap1 and p62. Mol Cell Biol 2010;30:3275-85.

[16] Eades G, Yang M, Yao Y, Zhang Y, Zhou Q. miR-200a regulates Nrf2 activation 854 by

targeting Keap1 mRNA in breast cancer cells. J Biol Chem 2011;286:40725-33.

[17] Sangokoya C, Telen MJ, Chi JT. MicroRNA miR-144 modulates oxidative stress

tolerance and associates with anemia severity in sickle cell disease. Blood

2010;116:4338-48.

[18] Yang M, Yao Y, Eades G, Zhang Y, Zhou Q. MiR-28 regulates Nrf2 expression through

a Keap1-independent mechanism. Breast Cancer Res Treat 2011;129:983-91.

[19] Chun KS, Kundu J, Kundu JK, Surh YJ. Targeting Nrf2-Keap1 signaling for

chemoprevention of skin carcinogenesis with bioactive phytochemicals. Toxicol Lett

2014;229:73-84.

[20] Lu SC. Glutathione synthesis. Biochim Biophys Acta 2013;1830:3143-53.

[21] Andújar I, Recio MC, Giner RM, Ríos JL. Traditional chinese medicine remedy to jury:

the pharmacological basis for the use of shikonin as an anticancer therapy. Curr Med

Chem 2013;20:2892-8.

[22] Andújar I, Ríos JL, Giner RM, Recio MC. Pharmacological properties of shikonin- a

review of literature since 2002. Planta Med 2013;79:1685-97.

[23] Wang XC, Feng J, Huang F, Fan YS, Wang YY, Cao LY, et al. Effects

of shikonin isolated from zicao on lupus nephritis in NZB/W F1 mice. Biol Pharm Bull

2009;32:1565-70.

[24] Cheng YW, Chang CY, Lin KL, Hu CM, Lin CH, Kang JJ. Shikonin derivatives

inhibited LPS-induced NOS in RAW 264.7 cells via downregulation of MAPK/NF-

kappaB signaling. J Ethnopharmacol 2008;120:264-71.

[25] Zhang B, Chen N, Chen H, Wang Z, Zheng Q. The critical role of redox homeostasis

in shikonin-induced HL-60 cell differentiation via unique modulation of the Nrf2/ARE

pathway. Oxid Med Cell Longev 2012;2012:781516.

[26] Yao Y, Brodie AM, Davidson NE, Kensler TW, Zhou Q. Inhibition of estrogen

signaling activates the NRF2 pathway in breast cancer. Breast Cancer Res Treat

2010;124:585-91.

[27] Huang CS, Lin AH, Liu CT, Tsai CW, Chang IS, Chen HW, et al. Isothiocyanates

protect against oxidized LDL-induced endothelial dysfunction by upregulating Nrf2-

dependent antioxidation and suppressing NFκB activation. Mol Nutr Food Res

2013;57:1918-30.

[28] Fraga CG, Leibovitz BE, Tappel AL. Lipid peroxidation measured as thiobarbituric

acid-reactive substances in tissue slices: characterization and comparison with

homogenates and microsomes. Free Radic Biol Med 1998;4:155-61.

[29] Guan X, Hoffman B, Dwivedi C, Matthees DP. A simultaneous liquid

chromatography/mass spectrometric assay of glutathione, cysteine, homocysteine and

their disulfides in biological samples. J Pharm Biomed Anal 2003;31:251-61.

[30] Yang YC, Lii CK, Wei YL, Li CC, Lu CY, Liu KL, et al. Docosahexaenoic acid

inhibition of inflammation is partially via cross-talk between Nrf2/heme oxygenase 1

and IKK/NF-κB pathways. J Nutr Biochem 2013;24:204-12.

[31] Lei YP, Chen HW, Sheen LY, Lii CK. Diallyl disulfide and diallyl trisulfide suppress

oxidized LDL-induced vascular cell adhesion molecule and E-selectin expression

through protein kinase A- and protein kinase B-dependent signaling pathways. J Nutr

2008;138:996-1003.

[32] Li W, Khor TO, Xu C, Shen G, Jeong WS, Yu S, et al. Activation of Nrf2-antioxidant

signaling attenuates NFkappaB-inflammatory response and elicits apoptosis. Biochem

Pharmacol 2008;76:1485-9.

[33] Galle J, Hansen-Hagge T, Wanner C, Seibold S. Impact of oxidized low density

lipoprotein on vascular cells. Atherosclerosis 2006;185:219-26.

[34] Heinloth A, Heermeier K, Raff U, Wanner C, Galle J. Stimulation of NADPH oxidase

by oxidized low-density lipoprotein induces proliferation of human vascular

endothelial cells. J Am Soc Nephrol 2000;11:1819-25.

[35] Förstermann U, Münzel T. Endothelial nitric oxide synthase in vascular

disease: from marvel to menace. Circulation 2006;113:1708-14.

[36] Zhao R, Ma X, Xie X, Shen GX. Involvement of NADPH oxidase in oxidized LDL-

induced upregulation of heat shock factor-1 and plasminogen activator inhibitor-1 in

vascular endothelial cells. Am J Physiol Endocrinol Metab 2009;297:E104-11.

[37] Xie L, Liu Z, Lu H, Zhang W, Mi Q, Li X, et al. Pyridoxine inhibits endothelial NOS

uncoupling induced by oxidized low-density lipoprotein via the PKCα signalling

pathway in human umbilical vein endothelial cells. Br J Pharmacol 2012;165:754-64.

[38] Kalyanaraman B. Teaching the basics of redox biology to medical and graduate students:

Oxidants, antioxidants and disease mechanisms. Redox Biol 2013;1:244-57.

[39] Zrelli H, Matsuoka M, Kitazaki S, Araki M, Kusunoki M, Zarrouk M, et al.

Hydroxytyrosol induces proliferation and cytoprotection against oxidative injury in

vascular endothelial cells: role of Nrf2 activation and HO-1 induction. J Agric Food

Chem 2011;59:4473-82.

[40] Gupte AA, Lyon CJ, Hsueh WA. Nuclear factor (erythroid-derived 2)-like-2 factor

(Nrf2), a key regulator of the antioxidant response to protect

against atherosclerosis and nonalcoholic steatohepatitis. Curr Diab Rep 2013;13:362-

71.

[41] Kumar H, Kim IS, More SV, Kim BW, Choi DK. Natural product-derived

pharmacological modulators of Nrf2/ARE pathway for chronic diseases. Nat Prod Rep

2014;31:109-39.

[42] Lu CY, Yang YC, Li CC, Liu KL, Lii CK, Chen HW. Andrographolide inhibits TNFα-

induced ICAM-1 expression via suppression of NADPH oxidase activation and

induction of HO-1 and GCLM expression through the PI3K/Akt/Nrf2 and

PI3K/Akt/AP-1 pathways in human endothelial cells. Biochem Pharmacol 2014;91:40-

50.

[43] Ahmed K, Furusawa Y, Tabuchi Y, Emam HF, Piao JL, Hassan MA, et al. Chemical

inducers of heat shock proteins derived from medicinal plants and cytoprotective genes

response. Int J Hyperthermia 2012;28:1-8.

[44] Mao X, Yu CR, Li WH, Li WX. Induction of apoptosis by shikonin through a

ROS/JNK-mediated process in Bcr/Abl-positive chronic myelogenous leukemia

(CML) cells. Cell Res 2008;18:879-88.

[45] Lee MJ, Kao SH, Hunag JE, Sheu GT, Yeh CW, Hseu YC, et al. Shikonin time-

dependently induced necrosis or apoptosis in gastric cancer cells via generation of

reactive oxygen species. Chem Biol Interact 2014;211:44-53.

[46] Gong K, Li W. Shikonin, a Chinese plant-derived naphthoquinone, induces apoptosis in

hepatocellular carcinoma cells through reactive oxygen species: A potential new

treatment for hepatocellular carcinoma. Free Radic Biol Med 2011;51:2259-71.

[47] Ahn J, Won M, Choi JH, Kim YS, Jung CR, Im DS, et al. Reactive oxygen species-

mediated activation of the Akt/ASK1/p38 signaling cascade and p21(Cip1)

downregulation are required for shikonin-induced apoptosis. Apoptosis 2013;18:870-

81.

[48] Kamei R, Kitagawa Y, Kadokura M, Hattori F, Hazeki O, Ebina Y, et al.

Shikonin stimulates glucose uptake in 3T3-L1 adipocytes via an insulin-independent

tyrosine kinase pathway. Biochem Biophys Res Commun 2002;292:642-51.

[49] Bain J, Plater L, Elliott M, Shpiro N, Hastie CJ, McLauchlan H, et al. The selectivity of

protein kinase inhibitors: a further update. Biochem J 2007;408:297-315.

[50] Gloire G, Piette J. Redox regulation of nuclear post-translational modifications during

NF-kappaB activation. Antioxid Redox Signal 2009;11:2209-22.

[51] Yeh PY, Li CY, Hsieh CW, Yang YC, Yang PM, Wung BS. CO-releasing molecules

and increased heme oxygenase-1 induce protein S-glutathionylation to modulate NF-

κB activity in endothelial cells. Free Radic Biol Med 2014;70C:1-13.

[52] Yang HL, Chang HC, Lin SW, Senthil Kumar KJ, Liao CH, Wang HM, et al. Antrodia

salmonea inhibits TNF-α-induced angiogenesis and atherogenesis in human

endothelial cells through the down-regulation of NF-κB and up-regulation

ofNrf2 signaling pathways. J Ethnopharmacol 2014;151:394-406.

[53] Khor TO, Huang MT, Kwon KH, Chan JY, Reddy BS, Kong AN. Nrf2-deficient mice

have an increased susceptibility to dextran sulfate sodium-induced colitis. Cancer Res

2006;66:11580-4.

[54] Yu AL, Lu CY, Wang TS, Tsai CW, Liu KL, Cheng YP, et al. Induction of heme

oxygenase 1 and inhibition of tumor necrosis factor alpha-induced intercellular

adhesion molecule expression by andrographolide in EA.hy926 cells. J Agric Food

Chem 2010;58:7641-8.

[55] Jeong WS, Kim IW, Hu R, Kong AN. Modulatory properties of various natural

chemopreventive agents on the activation of NF-kappaB signaling pathway. Pharm

Res 2004;21:661-70.

Figure legends

Fig. 1. Shikonin induces antioxidant enzyme expression and GSH contents in endothelial

cells. EA.hy926 cells were treated with 0-1 µM shikonin for 24 h or with 1 µM shikonin for

the indicated time periods. The protein expression of GCLM, GCLC, catalase (CAT), SOD-1,

SOD-2, GPx-1, GSR, and HO-1 was determined by Western blot (A). Levels of GCLM,

GCLC, catalase (CAT), SOD-1, SOD-2, GPx-1, GSR, and HO-1 mRNA were determined by

real-time PCR (B). Concentration-dependent changes in cellular GSH contents (reduced GSH

+ 2X GSSG) were measured after 16 h of shikonin treatment (C). GSH contents (reduced

GSH + 2 x GSSG) of the control cells were 270 ± 30.3 nmol/mg protein. Values are mean ±

SD, n=3. Values not sharing the same letter are significantly different (p<0.05).

Fig. 2. PI3K/Akt and MAPK activation and Nrf2 nuclear translocation by shikonin. EA.hy926

endothelial cells at 80% confluence were treated with 1 µM shikonin for up to 120 min. Time-

dependent changes of nuclear Nrf2 protein level were determined by Western blot (A). Cells

were transiently transfected with ARE-luciferase construct for 16 h followed by being

maintained in the fresh medium for another 2 h, and were then treated with 1 µM of shikonin

for an additional 16 h (B). PI3K/Akt, p38, ERK1/2, and JNK phosphorylation induced by

shikonin were determined over a 90-min period. Cells were pretreated with 20 µM PD98059,

LY294002, or SB2023580 for 1 h before incubation with 1 µM shikonin (C). The nuclear

Nrf2 level (D) was measured after 120 min of shikonin treatment and ARE luciferase activity

(E) and HO-1 expression (F) were determined after 16 h of shikonin treatment. Cells were

transfected with nontargeting control siRNA (NTC) or sip85α for 16 h, followed by treatment

with 1 µM shikonin for another 6 h (p85α and nuclear Nrf2 levels) or 16 h (HO-1 expression)

(G). Values are mean ± SD, n=3. Values not sharing the same letter are significantly different

(p<0.05). PARP, poly(ADP-ribose) polymerase.

Fig. 3. Shikonin inhibits oxLDL-induced ROS generation. EA.hy926 cells were pretreated

with 0 (A, B), 0.25 (C), 0.5 (D), or 1 µM (E) shikonin for 16 h followed by loading with

ROS-sensitive fluorophore H

DCFDA for another 15 min and then were stimulated without

(A) or with 40 µg/ml oxLDL (B, C, D, and E) for an additional 30 min. The DCF

fluorescence was detected by using a confocal microscope. Values are mean ± SD, n=3.

Values not sharing the same letter are significantly different (p<0.05).

Fig. 4. Shikonin suppresses oxLDL-induced NFκB activation and THP-1 monocyte adhesion

to endothelial cells. EA.hy926 cells were pretreated with 0.5 and 1 µM shikonin for 16 h

followed by stimulation with 40 µg/ml oxLDL for an additional 30 min. ICAM-1 and E-

selectin levels were analyzed by Western blot (A). Human monocytic THP-1 cell adhesion to

EA.hy926 cells was analyzed by BCECF-AM fluorescent assay (B). Changes in cytosolic

IκBα over 12 h of oxLDL treatment were determined (C). Nuclear NFκB p65 and cytosolic

IκBα contents were determined after 6 h of oxLDL treatment with or without a 16-h shikonin

pretreatment (D). Values are mean ± SD, n=3. Values not sharing the same letter are

significantly different (p<0.05).

Fig. 5. siNrf2 attenuates shikonin inhibition of adhesion molecule expression and THP-1

adhesion. EA.hy926 cells were transfected with nontargeting control siRNA (NTC) or siNrf2

for 16 h, followed by treatment with 0-1 µM shikonin for an additional 16 h. oxLDL (40

µg/ml) was then added. Knockdown efficiency of Nrf2 was confirmed by Western blot (A).

DNA binding activity of nuclear NFκB was determined after 6 h of oxLDL treatment by

EMSA (B). Unlabeled double-stranded NFκB oligonucleotide (cold) and a mutant double-

stranded oligonucleotide (mut) were added for the specificity assay. Effects of siNrf2 on

shikonin inhibition of ICAM-1 and E-selectin expression (C) and THP-1 adhesion (D) were

measured after 24 h of oxLDL exposure. Values are mean ± SD, n=3. Values of groups

treated with NTC or siRNA not sharing the same letter are significantly different (p<0.05). *

vs. NTC, p<0.05.

Fig. 6. Scheme summarizing the mechanisms of shikonin inhibition of monocyte adhesion to

vascular endothelial cells stimulated by oxidized low-density lipoprotein (oxLDL). CAT,

catalase; GCL, glutamate-cysteine ligase; SOD-1, superoxide dismutase 1; GPx-1, glutathione

peroxidase 1; GSR, glutathione reductase; HO-1, heme oxygenase 1; ICAM-1, intercellular

adhesion molecule 1; ROS, reactive oxygen species.