The transcription factor CREB enhances interleukin-17A production and inflammation in a mouse model of atherosclerosis

Page 1

(293), ra83. [DOI: 10.1126/scisignal.2004214] 6Science SignalingGadiparthi N. Rao (17 September 2013) Sivareddy Kotla, Nikhlesh K. Singh, Mark R. Heckle, Gabor J. Tigyi andInflammation in a Mouse Model of Atherosclerosis

The Transcription Factor CREB Enhances Interleukin-17A Production and`

This information is current as of 17 September 2013. The following resources related to this article are available online at http://stke.sciencemag.org.

Article Tools http://stke.sciencemag.org/cgi/content/full/sigtrans;6/293/ra83

Visit the online version of this article to access the personalization and article tools:

References http://stke.sciencemag.org/cgi/content/full/sigtrans;6/293/ra83#otherarticles

This article cites 55 articles, 32 of which can be accessed for free:

Glossary http://stke.sciencemag.org/glossary/

Look up definitions for abbreviations and terms found in this article:

Permissions http://www.sciencemag.org/about/permissions.dtl

Obtain information about reproducing this article:

the American Association for the Advancement of Science; all rights reserved. byAssociation for the Advancement of Science, 1200 New York Avenue, NW, Washington, DC 20005. Copyright 2008

(ISSN 1937-9145) is published weekly, except the last week in December, by the AmericanScience Signaling

on Septem

ber 17, 2013 stke.sciencem

ag.orgD

ownloaded from

Page 2

R E S E A R C H A R T I C L E

P H Y S I O L O G Y

The Transcription Factor CREB EnhancesInterleukin-17A Production and Inflammationin a Mouse Model of AtherosclerosisSivareddy Kotla, Nikhlesh K. Singh, Mark R. Heckle, Gabor J. Tigyi, Gadiparthi N. Rao*

stke.sD

ownloaded from

The enzyme 15-lipoxygenase (15-LO) plays a role in atherogenesis (also known as atherosclerosis),but the underlying mechanisms are unclear. We found that 15(S)-hydroxyeicosatetraenoic acid [15(S)-HETE],the major 15-LO–dependent metabolite of arachidonic acid, stimulated the production of reactive oxygenspecies (ROS) by monocytes through the xanthine oxidase–mediated activation of nicotinamide adeninedinucleotide phosphate (NADPH) oxidase. ROS production led to the Syk-, Pyk2-, and mitogen-activatedprotein kinase (MAPK)–dependent production of the proinflammatory cytokine interleukin-17A (IL-17A) in amanner that required the transcription factor CREB (cyclic adenosine monophosphate response element–binding protein). In addition, this pathway was required for the 15(S)-HETE–dependent migration and adhesionof monocytes to endothelial cells. Consistent with these observations, we found that peritoneal macrophagesfrom apolipoprotein E–deficient (ApoE−/−) mice fed a high-fat diet (a mouse model of atherosclerosis) ex-hibited increased xanthine oxidase and NADPH oxidase activities; ROS production; phosphorylation of Syk,Pyk2, MAPK, and CREB; and IL-17A production compared to those from similarly fed ApoE−/−:12/15-LO−/−

mice. These events correlated with increased lipid deposits and numbers of monocytes andmacrophages inthe aortic arches of ApoE−/− mice, which resulted in atherosclerotic plaque formation. Together, these ob-servations suggest that 15(S)-HETE exacerbates atherogenesis by enhancing CREB-dependent IL-17A pro-duction and inflammation.

cienc

on S

eptember 17, 2013

emag.org

INTRODUCTION

Cardiovascular diseases are the leading cause of death and disability inthe world (1). Atherosclerosis, which is a chronic inflammatory disease,is the major cause of heart failure and stroke, and it accounts for most ofthe deaths caused by cardiovascular diseases (1, 2). Atherosclerosis is char-acterized by thickening of the arterial wall as a result of the accumulation oflipid-laden foam cells of macrophage and smooth muscle cell origin, aswell as calcification (3). A large body of data shows that lipoxygenases(LOs), particularly 5-LO and 15-LO, which are enzymes that catalyze di-oxygenation of polyunsaturated fatty acids, play a role in the pathogenesisof this disease (4–7). One of the most explored and best appreciatedmechanisms of the involvement of LOs in atherogenesis is the capacityof 15-LO to oxidize low-density lipoprotein (LDL) (8–11). LOs catalyzethe stereospecific insertion of molecular oxygen into cis-polyunsaturatedfatty acids, such as arachidonic acid (AA) and linoleic acid, which re-sults in the formation of hydroperoxyeicosatetraenoic acids (HpETEs)and hydroperoxyoctadecadienoic acids (HpODEs), respectively (12, 13).HpETEs and HpODEs are then nonenzymatically converted to hydroxy-eicosatetraenoic acids (HETEs) and hydroxyoctadecadienoic acids (HODEs),respectively. 15-LO1 and 15-LO2 catalyze the conversion of AA mainlyto 15(S)-HpETE (14, 15), whereas the murine ortholog of 15-LO1, whichis known as 12/15-LO, converts AA to 12(S)-HpETE and 15(S)-HpETE(13). Some studies have demonstrated that when exposed to AA, athero-sclerotic arteries produce more 15-HETE than do normal arteries (16, 17).Furthermore, many of the risk factors for cardiovascular disease, such ashypercholesterolemia, diabetes, obesity, and smoking, are associated withthe increased abundance or activity of 12/15-LO (18–22). In addition,

Department of Physiology, University of Tennessee Health Science Center,894 Union Avenue, Memphis, TN 38163, USA.*Corresponding author. E-mail: [email protected]

www.S

HETEs are pro-oxidants (that is, they induce oxidative stress) (23), andmany reports showed that oxidative stress is linked to the pathogenesisof various disorders, including cardiovascular diseases, cancer, and rheu-matoid arthritis (24, 25). Despite this, little is known regarding the role ofHETEs in atherosclerosis other than the fact that 15-LO mediates theoxidation of LDL (9–11).

Because the migration of monocytes and macrophages to sites ofvascular injury and their adhesion to endothelium are considered thehallmarks of the initiation of atherogenesis (1), we hypothesized that15(S)-HETE, the major 15-LO–dependent product of AA, enhancedthe migration and adhesion of monocytes and macrophages to endothe-lium, thereby exacerbating inflammation. Here, we found that 15(S)-HETEenhanced the migration and adhesion of macrophages to monolayers ofendothelial cells in vitro. In addition, we showed that 15(S)-HETE stimu-lated the production of reactive oxygen species (ROS) by activating thexanthine oxidase–dependent nicotinamide adenine dinucleotide phosphate(NADPH) oxidase, which led to stimulation of the transcription factorCREB (cyclic adenosine monophosphate response element–binding pro-tein) and production of the proinflammatory cytokine interleukin-17A(IL-17A), which facilitated monocyte and macrophage migration and ad-hesion. Stimulation of CREB depended on the activities of the nonrecep-tor tyrosine kinases Syk and Pyk2, as well as the mitogen-activated proteinkinases (MAPKs). Furthermore, peritoneal macrophages isolated frommice deficient in apolipoprotein E (ApoE−/− mice) that were fed a high-fat diet (a mouse model of atherosclerosis) exhibited substantially increasedactivities of xanthine oxidase and NADPH oxidase, production of ROS,phosphorylation of Syk, Pyk2, MAPK, and CREB, and production ofIL-17A compared to those isolated from ApoE−/−:12/15-LO−/− mice feda high-fat diet. These observations also correlated with increased amountsof ROS, lipid deposits, numbers of monocytes and macrophages, IL-17Aproduction, and plaque formation in the aortic arches of ApoE−/− mice

CIENCESIGNALING.org 17 September 2013 Vol 6 Issue 293 ra83 1

Page 3

R E S E A R C H A R T I C L E

compared to those of ApoE−/−:12/15-LO−/− mice fed a high-fat diet, whichis suggestive of a crucial role for 12/15-LO in the development of high-fatdiet–induced oxidant stress and inflammation. These findings indicate that12/15-LO plays a role in atherogenesis by enhancing CREB-dependentIL-17A production.

on Septem

ber 17, 2013 stke.sciencem

ag.orgD

ownloaded from

RESULTS

15(S)-HETE stimulates the migration of THP1 cellsand their adhesion to endothelial cells in aROS-dependent mannerInflammation is considered an underlying factor in atherogenesis (26, 27).In addition, a role for LOs in atherogenesis and inflammation has beendemonstrated (4–7, 21). However, the mechanisms by which LOs influenceatherogenesis are not clear other than that 15-LO oxidizes LDL (9–11).To understand the role of LOs in atherogenesis, we studied the effects of5(S)-HETE, 12(S)-HETE, and 15(S)-HETE, which are derived by cleavageof AA by 5-LO, 12-LO, and 15-LO, respectively, on the migration and ad-hesion of monocytes. Of all the HETEs tested, 15(S)-HETE was the mostpotent in stimulating both the migration of THP1 cells (a human monocyticcell line) and their adhesion to a monolayer of human umbilical vein en-dothelial cells (HUVECs) (Fig. 1, A and B). 15(S)-HETE was more potentthan 15(R)-HETE in stimulating both the migration and adhesion of THP1cells (Fig. 1, C and D), which suggests that these effects are specific to15-LO metabolites of AA. Analysis of the dose-response effect of 15(S)-HETE showed that the maximal effects on THP1 cell migration and adhe-sion occurred at 0.1 mM 15(S)-HETE (Fig. 1, E and F).

To understand the mechanisms underlying the 15(S)-HETE–dependentmigration and adhesion of THP1 cells, we tested its effects on ROS pro-duction. 15(S)-HETE stimulated ROS production in a time-dependent man-ner, with a maximal fourfold increase in ROS generation 1 hour aftertreatment (Fig. 2A). A convincing body of evidence suggests that NADPHoxidase and xanthine oxidase are the major producers of ROS in most celltypes (25, 28, 29). Therefore, to find the source of 15(S)-HETE–inducedROS, we tested the roles of NADPH oxidase and xanthine oxidase inTHP1 cells. 15(S)-HETE stimulated the activities of both NADPH oxidaseand xanthine oxidase in a time-dependent manner, with a maximal three-to fivefold increase 30 min after treatment (Fig. 2, B and C). In addition,we found that the NADPH oxidase inhibitors apocynin and diphenyl-eneiodonium (30), as well as the xanthine oxidase inhibitor allopurinol(30), substantially reduced the extent of 15(S)-HETE–dependent ROSproduction (Fig. 2D).

To confirm these findings, we used an antisense oligonucleotide (ASO)approach. Depletion of either p47Phox, a component of NADPH oxidase(28), or xanthine oxidase substantially attenuated 15(S)-HETE–dependentROS production (Fig. 2E) in THP1 cells. Inhibition or depletion of NADPHoxidase or xanthine oxidase also blocked the 15(S)-HETE–dependentmigrationand adhesion of THP1 cells (Fig. 2, F and G). Because inhibition or deple-tion of NADPH oxidase and xanthine oxidase similarly attenuated 15(S)-HETE–dependent ROS production, we next asked whether there was anyinteraction between these oxidases. Inhibition of NADPH oxidase activity ordepletion of NADPH oxidase protein by pharmacological or ASO-based ap-proaches, respectively, had minor effects on 15(S)-HETE–dependent xan-thine oxidase activity (Fig. 2H). On the other hand, inhibition of xanthineoxidase activity or depletion of xanthine oxidase protein by pharmacologicalor ASO-based approaches, respectively, substantially attenuated 15(S)-HETE–dependent NADPH oxidase activity (Fig. 2I), similarly to that of specificNADPHoxidase inhibitors, suggesting that 15(S)-HETE–stimulatedNADPHoxidase activity and ROS production depend on xanthine oxidase activity.

www.S

We also tested the role of xanthine oxidase in the phosphorylation ofthe NADPH oxidase component p47Phox (25, 28). We found that 15(S)-HETE stimulated the phosphorylation of p47Phox in a time-dependentmanner (Fig. 2J), and knockdown of xanthine oxidase by ASO substantial-ly inhibited this effect (Fig. 2J). These findings suggest that 15(S)-HETEstimulated NADPH oxidase activity in a xanthine oxidase–dependentmanner, leading to the production of ROS. To obtain additional evidencefor these effects of 15(S)-HETE, we also tested its effects on the activitiesof NADPH oxidase and xanthine oxidase and ROS production in mouseprimary macrophages. 15(S)-HETE stimulated NADPH oxidase andxanthine oxidase activities and ROS production in a time-dependent man-ner, with maximal increases after 1 hour of treatment (Fig. 3, A to C).

0

10

20

30

5

15

25

3540

Mig

ratio

n (c

ells

/fiel

d)

45

Mig

ratio

n (c

ells

/fiel

d)

Con

trol 0.01 0.05 0.1 0.5 1.0

15(S)-HETE ( M)

0

10

20

30

40

50

60

A B

C D

E

0

100

200

400

500

600

300

700 Control5(S)-HETE12(S)-HETE15(S)-HETE

Control5(S)-HETE12(S)-HETE15(S)-HETE

Mig

ratio

n (c

ells

/fiel

d)

0

10

20

30

5

15

25

354045 Control

15(R)-HETE15(S)-HETE

0

200

400

600

100

300

500

700800900 Control

15(R)-HETE15(S)-HETE

*

*

*

*

*

**

*

**

*

Adh

esio

n (R

FU

) A

dhes

ion

(RF

U)

Con

trol 0.01 0.05 0.1 0.5 1.0

15(S)-HETE ( M)

0

10

20

30

40

50

60

Adh

esio

n (R

FU

)

*

* **

F

Fig. 1. 15(S)-HETE stimulates the migration and adhesion of THP1 cells.(A to D) Effects of various HETEs on the migration and adhesion of THP1

cells. (A and C) The migration of THP1 cells in response to vehicle (control)or the indicated HETE (0.1 mM) was measured by the modified Boydenchamber method. (B and D) Quiescent THP1 cells were treated withvehicle (control) or the indicated HETE (0.1 mM) for 1 hour, labeled with10 mM BCECF-AM [2′,7′-bis-(2-carboxyethyl)-5(and-6)-carboxyfluorescein-acetoxymethyl] for 30 min, overlaid onto a monolayer of quiescent HUVECs,and incubated for 2 hours. Adherent cells were determined by measur-ing fluorescence intensity. RFU, relative fluorescence units. (E and F) Themigration and adhesion of THP1 cells were measured as described for (A)to (D) in response to the indicated doses of 15(S)-HETE. Data are means ±SD from three independent experiments. *P < 0.01 versus control.

CIENCESIGNALING.org 17 September 2013 Vol 6 Issue 293 ra83 2

Page 4

R E S E A R C H A R T I C L E

on Septem

ber 17, 2013 stke.sciencem

ag.orgD

ownloaded from

0

40

80

120

160

p47Phox ASO2XO ASO2

Control ASO

– –

– ––+ +– –

– –+ +–

+ +

––

– –

– ––+ +– –

– –+ +–

+ +

––

-Tubulin

p47Phox

-Tubulin

XO

0

100

200

300

15(S)-HETE

** **

– ––

– –– –

– –– + + ––+ + – –

– – – – + +

– –

– ––+ +– –

– –+ +–

+ +

––– –

–

– –– –

– –– + + ––+ + – –

– – – – + +

– –

– ––+ +– –

– –+ +–

+ +

––– –

–

– –– –

– –– + + ––+ + – –

– – – – + +

**

*

*

*

*

RO

S p

rodu

ctio

n (R

FU

)

RO

S p

rodu

ctio

n (R

FU

)

NA

DP

H o

xida

se a

ctiv

ity (

M)

0

1

5

6

7

2

3

4

5 60 12010

Con

trol 30

15(S)-HETE (min)

5 60 12010

Con

trol 30

15(S)-HETE (min)

**

**

XO

act

ivity

(R

FU

)

0

500

100

200

300

400

600

5 60 12010

Con

trol 30

15(S)-HETE (min)

*

*

*

0

100

200

300

ApocyninDPI

Allopurinol

15(S)-HETE

– ––

– –– –

– –– + + ––+ + – –

– – – – + +

**

**

**

*

RO

S p

rodu

ctio

n (R

FU

)

p47Phox ASO1– –p47Phox ASO2

Control ASO –––+

–+

+

XO ASO1– –XO ASO2

Control ASO –––+

–+

+

pp47Phox

XO

p47Phox

– –

XO ASO2

+ + – –p47Phox ASO2 ––– – + +

15(S)-HETE – + + – +–

Control ASO

– – ++––

pp47Phox

p47Phox

Con

trol

6030 120 240 360

15(S)-HETE (min)

0

1

5

6

7

2

3

4

8

ApocyninDPI

Allopurinol

15(S)-HETE*

**

****

NA

DP

H o

xida

se a

ctiv

ity (

M)

0

5

10

15

20

25

30*

**

**

15(S)-HETE

XO ASO2p47Phox ASO2

Control ASO

NA

DP

H o

xida

se a

ctiv

ity (

M)

-Tubulin

p47Phox ASO2XO ASO2

Control ASO0

10

20

30

40

50

60

70

Mig

ratio

n (c

ells

/fiel

d)

15(S)-HETE

Mig

ratio

n (c

ells

/fiel

d)

0

20

40

60

80

120

140

ApocyninDPI

Allopurinol

100

15(S)-HETE

** ******

**

* *

0

500

1000

2000

2500

1500

ApocyninDPI

Allopurinol

15(S)-HETE

p47Phox ASO2XO ASO2

Control ASO0

1000

2000

3000

4000

5000 15(S)-HETE

** **

******

**

Adh

esio

n (R

FU

)

Adh

esio

n (R

FU

)

– –

– ––+ +– –

– –+ +–

+ +

––– –

–

– –– –

– –– + + ––+ + – –

– – – – + +

0

2000

4000

6000

8000

10,000

12,000

14,000

16,00015(S)-HETE

**

*

ApocyninDPI

Allopurinol

XO

act

ivity

(R

FU

)

*

**

0

2000

4000

6000

8000

10,000

12,000

XO ASO2p47Phox ASO2

Control ASO

15(S)-HETE*

**

XO

act

ivity

(R

FU

)

*

A B C D

E

JI

G H

F

Fig. 2. The 15(S)-HETE–stimulated migration and adhesion of THP1 cellsrequires ROS production. (A to D) THP1 cells were treated with (A to C)

absence of the indicated inhibitors before being analyzed for (F) migration,(G) adhesion, (H) xanthine oxidase activity, and (I) NADPH oxidase activity.

vehicle (control) or 0.1 mM 15(S)-HETE for the indicated times or with (D)vehicle (control) or 0.1 mM 15(S)-HETE in the absence or presence of100 mM apocynin, 10 mM diphenyleneiodonium (DPI), or 100 mM allopurinolfor 1 hour. (A and D) ROS production, (B) NADPH oxidase activity, and (C)xanthine oxidase (XO) activity were measured. (E) Left: THP1 cells weretransfected with control, p47Phox-specific, or xanthine oxidase–specificASOs (100 nM). Forty-eight hours later, the abundances of p47Phox andxanthine oxidase were analyzed by Western blotting and were normalizedto that of b-tubulin. Right: Forty-eight hours after transfection, quiescentcells were treated with vehicle or 0.1 mM 15(S)-HETE for 1 hour before beinganalyzed for ROS production. (F to I) Left: THP1 cells were treated withvehicle (empty bars) or 0.1 mM 15(S)-HETE (filled bars) in the presence or

www.S

Right: THP1 cells transfected with control ASO or the indicated ASOs weretreated with vehicle (empty bars) or 0.1 mM 15(S)-HETE (filled bars) beforebeing analyzed for (F) migration, (G) adhesion, (H) xanthine oxidase activity,and (I) NADPH oxidase activity. (J) Top: Whole-cell lysates of vehicle-treatedTHP1 cells or THP1 cells that were treated with 0.1 mM 15(S)-HETE for the in-dicated times were analyzed byWestern blotting with antibodies against theindicated proteins. Bottom: THP1 cells were transfected with the indicatedASOs and allowed to rest before being treated with 15(S)-HETE for 1 hour andthen were analyzed by Western blotting with antibodies against the indicatedproteins. Data in (A) to (I) aremeans ± SD from three independent experiments.*P< 0.01 versus control; **P < 0.01 versus 15(S)-HETE. Western blots in (J) arefrom one experiment and are representative of three independent experiments.

CIENCESIGNALING.org 17 September 2013 Vol 6 Issue 293 ra83 3

Page 5

R E S E A R C H A R T I C L E

Consistent with these effects, 15(S)-HETE also induced the migrationand adhesion of these cells to a monolayer of mouse endothelial cells(Fig. 3, D and E).

15(S)-HETE–stimulated migration and adhesionof THP1 cells requires ROS-dependent activationof Syk and Pyk2ROS act as intracellular signals primarily through their capacity to in-hibit protein tyrosine phosphatases (PTPs), thereby tilting the balanceof the system toward the activation of protein tyrosine kinases (PTKs)(31, 32). Having found that 15(S)-HETE stimulated ROS production, wenext asked whether ROS led to the activation of any nonreceptor tyrosinekinases. We tested the effect of 15(S)-HETE on the activation of three non-receptor tyrosine kinases: Src, Syk, and Pyk2. Although 15(S)-HETEcaused only a marginal increase in the tyrosine phosphorylation of Src, itstimulated the tyrosine phosphorylation of both Syk and Pyk2 in a time-

www.S

on Septem

ber 17, 2013 stke.sciencem

ag.orgD

ownloaded from

dependent manner, with near-maximal increases in phosphorylation 1 hourafter treatment (Fig. 4A). To test the roles of Syk and Pyk2 in the 15(S)-HETE–dependent migration and adhesion of THP1 cells, we used a com-bination of pharmacological and genetic approaches. We found thatBAY61-3606 and PF431396, specific inhibitors of Syk (33) and Pyk2(34), respectively, inhibited 15(S)-HETE–stimulated migration and adhe-sion of THP1 cells (Fig. 4B). Similarly, ASO-mediated reduction in theabundance of Syk or Pyk2 also reduced the 15(S)-HETE–dependent mi-gration and adhesion of THP1 cells (Fig. 4C).

To understand the mechanisms by which Syk and Pyk2 were ac-tivated by 15(S)-HETE, we tested the role of ROS. Inhibition or de-pletion of either NADPH oxidase or xanthine oxidase attenuated the15(S)-HETE–dependent phosphorylation of Syk and Pyk2 (Fig. 4D).Because 15(S)-HETE activated both Syk and Pyk2 with similar timecourses, we asked whether there was any interaction between thesetwo kinases. Inhibition or depletion of Syk blocked the 15(S)-HETE–dependent phosphorylation of Pyk2 (Fig. 4E). Conversely, inhibitionor depletion of Pyk2 substantially inhibited the phosphorylation ofSyk (Fig. 4E), which suggested that each kinase depended on the otherfor activation.

On the basis of these observations, we performed coimmunoprecipita-tion experiments to determine whether Syk and Pyk2 interacted with eachother. We found that both Syk and Pyk2 physically associated with eachother in THP1 cells, and that this interaction was enhanced in response to15(S)-HETE (Fig. 4F). To explore potential signaling events downstreamof the 15(S)-HETE–dependent activation of Syk and Pyk2, we tested thestimulation of MAPKs. 15(S)-HETE induced the phosphorylation of ex-tracellular signal–regulated kinases 1 and 2 (ERK1/2), c-Jun N-terminalkinases 1 and 3 (JNK1/3), and p38 in a time-dependent manner, with maxi-mal responses 1 hour after treatment (Fig. 5A). In addition, PD098059,SP600125, and SB203580, which specifically inhibit MAPK-ERK kinase1 and 2 (MEK1/2; kinases that activate ERK1/2), JNK1/2, and p38, re-spectively (35), inhibited 15(S)-HETE–stimulated THP1 cell migrationand adhesion (Fig. 5B). Furthermore, inhibition or depletion of NADPHoxidase, xanthine oxidase, Syk, or Pyk2 blocked 15(S)-HETE–dependentphosphorylation of ERK1/2, JNK1/3, and p38 (Fig. 5, C and D). Thesefindings revealed that 15(S)-HETE stimulated the activation of ERK1/2,JNK1/3, and p38 in a manner dependent on xanthine oxidase, NADPHoxidase, Syk, and Pyk2.

15(S)-HETE–stimulated migration and adhesionof THP1 cells requires ROS-, Syk-, Pyk2-, andMAPK-dependent activation of CREBHaving observed that 15(S)-HETE stimulated MAPK activation, wenext asked whether these responses led to the activation of transcriptionfactors. We showed that MAPKs mediate the activation of ATF2 (acti-vating transcription factor 2) and CREB during 15(S)-HETE–stimulatedmigration of endothelial cells and vascular smooth muscle cells (15, 36).Therefore, we tested the effect of 15(S)-HETE on ATF2 and CREB ac-tivation in THP1 cells. We found that whereas 15(S)-HETE had no ef-fect on ATF2 phosphorylation, it stimulated CREB phosphorylationin a time-dependent manner with a maximal response after 1 hour oftreatment (Fig. 6A). ASO-mediated knockdown of CREB inhibitedthe 15(S)-HETE–stimulated migration and adhesion of THP1 cells(Fig. 6B). To elucidate the mechanisms underlying 15(S)-HETE–dependent CREB activation, we tested the roles of NADPH oxidase,xanthine oxidase, Syk, Pyk2, and MAPKs. Depletion of NADPH oxi-dase, xanthine oxidase, Syk, or Pyk2 or pharmacological inhibition ofMAPKs attenuated the 15(S)-HETE–dependent phosphorylation ofCREB (Fig. 6, C to E).

Adh

esio

n (1

x 1

02 R

FU

)

0

2

10

12

4

6

8

Control15(S)-HETE

*

5 60 12010

Con

trol

15(S)-HETE (min)

30

XO

act

ivity

(1

x 10

3 R

FU

)

0

2

10

4

6

8* **

*

*

5 60 12010

Con

trol

15(S)-HETE (min)

30

NA

DP

H o

xida

se a

ctiv

ity (

M)

0

1

5

6

2

3

4

*

*

**

*

02

1012

468

1618

14

RO

S p

rodu

ctio

n (1

x 1

03 R

FU

)

5 60 12010

Con

trol

15(S)-HETE (min)

30

* **

*

*

Mig

ratio

n (c

ells

/fiel

d)

05

2530

101520

4045

35

Control15(S)-HETE*

A B

C

E

D

Fig. 3. 15(S)-HETEstimulatesNADPHoxidaseand xanthine oxidase activities and ROS pro-duction in mouse peritoneal macrophagesand enhances their migration and adhesion.(A to C) Quiescent mouse peritoneal macro-phages were treated with vehicle (control) or0.1 mM 15(S)-HETE for the indicated times be-fore being analyzed for (A) NADPH oxidaseactivity, (B) xanthine oxidase activity, and (C)ROS production. (D) The migration of quies-

cent macrophages in response to vehicle (control) or 0.1 mM 15(S)-HETEwas measured by the modified Boyden chamber method. (E) Quiescentmacrophages were treated with vehicle (control) or 0.1 mM 15(S)-HETEfor 1 hour, labeled with 10 mM BCECF-AM for 30 min, overlaid onto amonolayer of quiescentmouse endothelial cells, and incubated for 2hours.Adherent cells were determined by measuring fluorescence intensities.Data are means ± SD from three independent experiments. *P < 0.01 ver-sus control.

CIENCESIGNALING.org 17 September 2013 Vol 6 Issue 293 ra83 4

Page 6

R E S E A R C H A R T I C L E

www.SCIENCESIGNALING.org 17 S

on Septem

ber 17, 2013 stke.sciencem

ag.orgD

ownloaded from

15(S)-HETE–stimulatedmigration and adhesion of THP1cells requires ROS-, Syk-, Pyk2-,MAPK-, and CREB-dependentproduction of IL-17AWe next studied the effects of 15(S)-HETEon the production of the proinflammatorycytokines IL-11, IL-17A, and tumor necrosisfactor–a (TNF-a) by THP1 cells. Whereas15(S)-HETE reduced the abundances ofboth IL-11 and TNF-a mRNAs in THP1cells, it increased IL-17A mRNA abun-dance in a time-dependent manner, witha maximal effect after 2 hours of treatment(Fig. 7A). Because 15(S)-HETE increasedthe abundance of IL-17A mRNA, we nextexamined its effects on IL-17A protein abun-dance, and found that it stimulated maximalIL-17A protein production after 2 hours oftreatment (Fig. 7B). IL-17A stimulated themigration and adhesion of THP1 cells in adose-dependent manner, with a maximumresponse at a concentration of 20 ng/ml(Fig. 7C). Furthermore, neutralizing anti–IL-17AantibodiesblockedtheIL-17A–dependent,aswell as the15(S)-HETE–dependent,migra-tion and adhesion of THP1 cells (Fig. 7D).ASO-mediated depletion of NADPH oxi-dase, xanthine oxidase, Syk, Pyk2, or CREBand pharmacological inhibition of MAPKsattenuated the 15(S)-HETE–dependent pro-duction of IL-17A (Fig. 7, E to H). Theseobservations suggest that 15(S)-HETEstim-ulates the migration and adhesion of THP1cells through the xanthine oxidase–dependentNADPH oxidase–mediated production ofROS, which leads to the activation of a sig-naling axis consisting of Syk, Pyk2, and vari-ous MAPKs, which culminates in activationof the transcription factor CREB and theproduction of IL-17A.

12/15-LO mediates theproduction of IL-17A inmice fed a high-fat dietApolipoprotein E–deficient (ApoE−/−) micedevelop atherogenesis spontaneously, andthus, these mice are used as a model to studythe pathogenesis of atherosclerosis (2). Pre-vious studies have shown that deletion of12/15-LO in ApoE−/− mice attenuates high-fat diet–induced atherogenesis (5, 10).Therefore, to explore the pathophysiolog-ical importance of these findings, we usedbothApoE−/− andApoE−/−:12/15-LO−/−mice.After feedingApoE−/−miceandApoE−/−:12/15-LO−/−micewith a high-fat diet for 8 weeks,we studied the signaling events thatwe char-acterized earlier in monocytes and macro-phages. We found that macrophages from

pSyk

Syk

pPyk2

Pyk2

Con

trol

6030 120 240 360

15(S)-HETE (min)

Pho

spho

prot

ein

abun

danc

e(f

old

stim

ulat

ion)

pSrc

Src

PF431396

0

1000

2000

3000

4000

BAY61-3606

*

15(S)-HETE

0

2

4

6

8

30 240 36060

Con

trol 120

15(S)-HETE (min)

pSyK

pPyk2pSrc

*

*

*

*

*

*

*

*

*

*

Mig

ratio

n (c

ells

/fiel

d)

0

20

40

60

80

Adh

esio

n (R

FU

)

– ––

– –– + + ––

+ +

– ––

– –– + + ––

–+ – – –+

+ + – ––

– –– + + ––

–+ – – –+

+ +

– ––

– –– + + ––

+ +BAY61-3606

PF431396

* 15(S)-HETE

†

††

†

0

5

10

15

20

30

25

Mig

ratio

n (c

ells

/fiel

d)

Syk ASO3Pyk2 ASO3

Control ASO

*

0

1000

2000

3000

4000

5000

Syk ASO3 Pyk2 ASO3

Control ASO

* 15(S)-HETE15(S)-HETE

Adh

esio

n (R

FU

)

Syk

-Tubulin

Syk ASO1– –Syk ASO2

Control ASO –––

–––+

– – –Syk ASO3 +

–+

+

Pyk2

-Tubulin

Pyk2 ASO1– –Pyk2 ASO2

Control ASO –––

–––+

– – –Pyk2 ASO3 +

–+

+

††

† †

ApocyninDPI

Allopurinol15(S)-HETE

pPyk2

Pyk2

pSyk

Syk

– –

XO ASO2

+ + – –p47Phox ASO2 ––– – + +

15(S)-HETE – + + – +–

Control ASO

– – ++––

p47Phox

XO

-Tubulin

0

1

3

2

4

– –

XO ASO2

+ + – –p47Phox ASO2 ––– – + +

15(S)-HETE – + + – +–

Control ASO

– – ++––

pSykpPyk2* *

pPyk2

Pyk2

pSyk

Syk

– ––

– + +– –– –

– –

– + +

– + + ––+ + – –

– – – – + +– –

0.0

1.0

2.0

3.0

4.0

ApocyninDPI – –

–

Allopurinol15(S)-HETE – + +

– –– –

– –

– + +

– + + ––+ + – –

– – – – + +– –

* * pSykpPyk2

Pho

spho

prot

ein

abun

danc

e(f

old

stim

ulat

ion)

††

†

††

†

†

†† †

Syk

IP:Pyk2

Con

trol

60 120 240 360

15(S)-HETE (min)

30

Pyk2

IB:

pPyk2

Pyk2

pSyk

Syk

PF431396– –

15(S)-HETE – + +– –

– +

+ + – –– – + +

–

BAY61-3606

– –Pyk2 ASO3+ +– – –

Syk ASO3 – –– –

15(S)-HETE – + + – +–

Control ASO

–

+ +

–

–

pSyk

Syk

pPyk2

Pyk2

-Tubulin

+ +

Con

trol

60 120 240 360

15(S)-HETE (min)

30 IB:

Syk

Pyk2

IP:Syk

Pho

spho

prot

ein

abun

danc

e(f

old

stim

ulat

ion)

A B

C

D

E

F

Fig. 4. Syk and Pyk2 mediate the 15(S)-HETE–dependent migration and adhe-sion of THP1 cells. (A) Whole-cell lysatesofvehicle-treatedTHP1cellsorcells treatedwith 0.1 mM 15(S)-HETE for the indicatedtimes were analyzed by Western blottingwith antibodies against the indicatedproteins. Bar graph showsdensitometric

analysis of the fold increase in the abundances of pSyk, pSrc, and pPyk2 at the indicated times in 15(S)-HETE–treatedcells compared to that in vehicle-treatedcells. (B) THP1cellswerepretreatedwith vehicle, 10mMBAY61-3606, or 5 mMPF431396 for 30 min before being treated with 0.1 mM 15(S)-HETE and then analyzed formigration (left) and adhesion (right) as described earlier. (C) Top: Western blotting analysis of the effects of Syk-and Pyk2-specific ASOs on their targets. Bottom: THP1 cells transfectedwith the indicated ASOs and allowed torest were treated with vehicle or 0.1 mM15(S)-HETE before being analyzed for migration and adhesion. (D) Left:THP1 cells were treatedwith vehicle or 0.1 mM15(S)-HETE in the absence or presence of the indicated inhibitorsfor 1 hour and then were analyzed by Western blotting with antibodies specific for the indicated proteins. Right:THP1cells transfectedwith the indicatedASOswereallowedtorestbeforebeingtreatedwithvehicleor0.1mM15(S)-HETE for 1 hour and then were analyzed by Western blotting with antibodies against the indicated proteins. Bargraphshowsdensitometricanalysisof the fold increase in theabundancesofpSykandpPyk2 foreachconditioncompared to that in vehicle-treatedcells. (E) Inhibition or knockdownof SykorPyk2 reciprocally affects their 15(S)-HETE–dependent phosphorylation. THP1 cells treated with the indicated inhibitors (top) or transfected with theindicated ASOs (bottom) were treated with vehicle or 0.1 mM 15(S)-HETE before being analyzed by Westernblotting with antibodies against the indicated proteins. (F) THP1 cells treated with vehicle (control) or 0.1 mM15(S)-HETEfor the indicated timesweresubjected to immunoprecipitation (IP)withanti-Pyk2oranti-Sykantibodies,and the immunoprecipitateswereanalyzedbyWesternblotting (IB)withantibodiesspecific forSykorPyk2.Data in(A) to (D) aremeans ± SD from three independent experiments. *P< 0.01 versus control; †P< 0.01 versus 15(S)-HETE. Western blots are from one experiment and are representative of three independent experiments.

eptember 2013 Vol 6 Issue 293 ra83 5

Page 7

R E S E A R C H A R T I C L E

on Septem

ber 17, 2013 stke.sciencem

ag.orgD

ownloaded from

ApoE−/−:12/15-LO−/− mice exhibited substantially reduced migration andadhesion capacities compared to those of ApoE−/−mice, as well as demon-strated comparatively reduced NADPH oxidase and xanthine oxidase activ-ities, ROSproduction, Syk, Pyk2,MAPK, andCREBphosphorylation, andIL-17A production (Fig. 8, A to C). Furthermore, the plasma concentration

www.SCIENCESIGNALIN

of IL-17Awas substantially lower in ApoE−/−:12/15-LO−/−

mice than inApoE−/−mice (Fig. 8D). In contrast, the con-centration of the anti-inflammatory cytokine IL-10 wassubstantially greater in ApoE−/−:12/15-LO−/− mice thanin ApoE−/− mice (Fig. 8D). Thus, these in vivo findingssuggest that 12/15-LO plays a role in enhancing the in-flammation elicited by the xanthine oxidase– andNADPHoxidase–dependent ROS production that leads to Syk-Pyk2-MAPK–mediated CREB activation and IL-17A pro-duction. These observations also suggest that 12/15-LOincreases the cellular oxidative stress of inflammatorycells, particularly monocytes and macrophages, in re-sponse to the cardiovascular risk factors of high-fat dietand increased caloric intake.

To correlate these observations with plaque forma-tion, we isolated the aortic arches from ApoE−/− miceand ApoE−/−:12/15-LO−/− mice that were fed a high-fatdiet for 8 weeks, and examined them for ROS produc-tion, lipid deposits, and numbers of monocytes and mac-rophages, as well as for IL-17A production. We observedincreased amounts of ROS (both superoxide anion andH2O2), lipid deposits (by Oil Red O staining), and num-bers of monocytes and macrophages in the aortic archesof ApoE−/− mice compared to those of ApoE−/−:12/15-LO−/− mice (Figs. 8, E and F, and 9, A and B). In addi-tion, immunofluorescence staining showed that the aorticarches of ApoE−/− mice had increased abundance ofIL-17A compared to those of ApoE−/−:12/15-LO−/−

mice (Fig. 9C). Similarly, double immunofluorescencestaining for the macrophage marker Mac3 and IL-17A re-vealed that there was an increased number of IL-17A–producing macrophages in the aortic arches of ApoE−/−

mice compared to those of ApoE−/−:12/15-LO−/− micefed a high-fat diet (Fig. 9C).

DISCUSSION

The presence of abnormal amounts of LDL represents arisk factor for cardiovascular diseases (8, 37). The majormechanism underlying the role of the oxidative modi-fication of LDL in atherosclerosis is its uptake and re-tention by macrophages, which then transform intofoam cells and accumulate in the arterial wall, resultingin increased thickness and hardening (1, 26, 27). Thegradual accumulation of lipid-laden foam cells leadsto plaque formation. Whereas the progression of theplaque eventually leads to occlusion of the artery, un-stable plaques can also rupture and expose their throm-bogenic matrix to blood, which leads to the formation ofthrombo-embolic occlusions (37). Many reports showedthat 15-LO1 and its murine ortholog 12/15-LO play arole in atherogenesis by mediating the oxidation of LDL(4–7, 9–11). Although the mechanisms by which 15-LO1modifies LDL are still unclear, some reports showed thatupon exposure to AA, atherosclerotic arteries, liver, or

macrophages in vivo, ex vivo, or in vitro generate 15-HETE at submicro-molar concentrations (16, 17, 38, 39), which raises the possibility thatthis lipid mediator might have some influence in atherogenesis. Indeed,15-HpETE, the peroxide form of 15-HETE, is capable of enhancing theoxidation of LDL in the presence of catalytic amounts of copper (23).

0

40

80

120

160

PD098059

Mig

ratio

n (c

ells

/fiel

d)

15(S)-HETE

– – –– + + ––

*

– –– –– – + + – –

– – – – + +

– – –– + + ––– –

– –– – + + – –

– – – – + +SB203580SP600125

0

1000

2000

3000

4000

*

PD098059

SB203580SP600125

15(S)-HETE

Adh

esio

n (R

FU

)

pERK1/2

ERK2

pJNK1/3

pp38 MAPK

p38 MAPK

JNK1

Con

trol

6030 120 240 360

15(S)-HETE (min)

†† †

†

††

pJNK1/3pERK1/2

pp38 MAPK5

6

4

3

2

1

0

15(S)-HETE (min)

Con

trol 30 60 120 240 360

*

*

**

*

*

*

*

*

*

pERK1/2

pp38 MAPK

p38 MAPK

ERK2

JNK1

pJNK1/3

ApocyninDPI – –

–

Allopurinol15(S)-HETE – + +

– –– –

– –

– + +

– + + ––+ + – –

– – – – + +– –

0

1

2

3

4

5

6

pp38 MAPK

pERK1/2pJNK1/3

ApocyninDPI – –

–

Allopurinol15(S)-HETE – + +

– –– –

– –

– + +

– + + ––+ + – –

– – – – + +– –

– –

XO ASO2

+ + – –p47Phox ASO2 ––– – + +

15(S)-HETE – + + – +–

Control ASO

– – ++––

pERK1/2

pJNK1/3

pp38 MAPK

ERK2

JNK1

p38 MAPK

XO

p47Phox

-Tubulin

0

1

2

3

4

5

– –

XO ASO2

+ + – –p47Phox ASO2 ––– – + +

15(S)-HETE – + + – +–

Control ASO

– – ++––

pp38 MAPK

pERK1/2pJNK1/3*

†

*

*

†

†

†

†††

†

†

*

†

*

*

†

†

†††

pERK1/2

pp38 MAPK

p38 MAPK

ERK2

JNK1

pJNK1/3

PF431396– –

15(S)-HETE – + +– –

– +

+ + – –– – + +

–

BAY61-3606

0

1

2

3

PF431396– – BAY61-3606

15(S)-HETE – + +– –

– +

+ + – –– – + +

–

pp38 MAPK

pERK1/2pJNK1/3

– –Pyk2 ASO3+ +– – –

Syk ASO3 – –– –

15(S)-HETE – + + – +–

Control ASO

–

+ +

–

–

pERK1/2

ERK2

pJNK1/3

JNK1

pp38 MAPK

p38 MAPK

0

1

2

4

5

6

3

pp38 MAPK

pERK1/2pJNK1/3

Syk

Pyk2

-Tubulin

+ +

– –Pyk2 ASO3+ +– – –

Syk ASO3 – –– –

15(S)-HETE – + + – +–

Control ASO

–

+ +

–

–

+ +

**

*

* *

††

†

††

†

†

††

†

††

Pho

spho

prot

ein

abun

danc

e(f

old

stim

ulat

ion)

Pho

spho

prot

ein

abun

danc

e(f

old

stim

ulat

ion)

Pho

spho

prot

ein

abun

danc

e(f

old

stim

ulat

ion)

Pho

spho

prot

ein

abun

danc

e(f

old

stim

ulat

ion)

Pho

spho

prot

ein

abun

danc

e(f

old

stim

ulat

ion)

A B

C

D

Fig. 5. MAPKs mediatethe 15(S)-HETE–dependentmigration and adhesion ofTHP1 cells. (A) Whole-celllysates of vehicle-treatedTHP1 cells (control) or THP1cells treated with 0.1 mM15(S)-HETE for the indi-cated times were analyzedbyWestern blotting with anti-bodies specific for the indi-cated proteins. Bar graphshows densitometric analy-sis of the fold increase in theabundances of pERK1/2,pJNK1/3, and pp38 MAPKat the indicated times in 15(S)-HETE–treated cells com-paredto that invehicle-treatedcells. (B) THP1cellswerepre-treated for 30 min with vehi-cle, 30mMPD098059, 10mMSP600125,or10mMSB203580;treatedwith vehicleor0.1mM15(S)-HETE; and then ana-lyzed for migration (left) andadhesion (right). (C and D)THP1 cells were treated withthe indicated inhibitorsorwere transfectedwith the indicatedASOsbefore being treated with vehicle or 0.1 mM15(S)-HETE for 1 hour.

Whole-cell lysates were then analyzed byWestern blottingwith antibodies specific for the total andphosphorylated forms of the indicatedMAPKproteins, p47Phox, xanthine oxidase, Syk, and Pyk2.b-Tubulin was used as a loading control. Bar graph shows densitometric analysis of the fold in-crease in the abundances of pERK1/2, pJNK1/3, and pp38 MAPK for each condition comparedto that invehicle-treatedcells.Westernblotsare fromoneexperimentandare representativeof threeindependent experiments. Data in bar graphs are means ± SD from three independentexperiments. *P < 0.01 versus control; †P < 0.01 versus 15(S)-HETE.

G.org 17 September 2013 Vol 6 Issue 293 ra83 6

Page 8

pCREB

CREB

-Tubulin

0

10

20

30

40

50

60

– –

XO ASO2

+ + – –p47Phox ASO2 ––– – + +

15(S)-HETE – + + – +–

Control ASO

– – ++––

pCREB

CREB

p47Phox

XO

0

1

2

3

4

– –

XO ASO2

+ + – –p47Phox ASO2 ––– – + +

Control ASO

– – ++––

0

10

20

30

40

50

60

70

Mig

ratio

n (c

ells

/fiel

d)

CREB ASO3 + Control ASO

– ––+ + –

+

CREB ASO3 + Control ASO

– ––+ + –

+

15(S)-HETE

15(S)-HETE*

**

** **

**

**

CREB

Adh

esio

n (R

FU

)

Con

6030 120 240 360

pATF2

ATF2

– –CREB ASO2–

–+– –CREB ASO3 +

– –

Pyk2 ASO3

+ + – –Syk ASO3 ––– – + +

15(S)-HETE – + + – +–

Control ASO

– – ++––

pCREB

CREB

Syk

Pyk2

-Tubulin

0

4

6

8

2

– –

Pyk2 ASO3

+ + – –Syk ASO3 ––– – + +

Control ASO

– – ++––

*

**

**0

1

3

4

5

2

SB203580

PD098059– –

–

– –– –

– –– + + ––+ + – –

– – – – + +SP600125

*

**

** **

15(S)-HETE

15(S)-HETE

15(S)-HETESB203580

PD098059– –

–

– + +– –– –

– –

– + +

– + + ––+ + – –

– – – – + +– –

SP600125

pCREB

CREB

pCR

EB

abu

ndan

ce(f

old

stim

ulat

ion)

pCR

EB

abu

ndan

ce(f

old

stim

ulat

ion)

pCR

EB

abu

ndan

ce(f

old

stim

ulat

ion)

C

D E

R E S E A R C H A R T I C L E

on Septem

ber 17, 2013 stke.sciencem

ag.orgD

ownloaded from

Because HpETEs are mostly unstable, we asked whether their rela-tively stable form, 15(S)-HETE, influences the atherogenic process. Wepreviously showed that 15(S)-HETE enhances vascular wall remodelingin response to injury (14) and increases angiogenesis in response to ische-mia (35). Having observed that 15(S)-HETE enhanced the migration ofendothelial cells and smooth muscle cells, we reasoned that it might alsopromote the migration of monocytes and macrophages as well as their ad-hesion to endothelium, thereby influencing atherogenesis. Here, we foundthat 15(S)-HETE stimulated monocyte migration and adhesion to endothe-lium. Oxidative stress plays a crucial role in vascular diseases (24, 28).Because some reports showed that 15-HpETE promotes oxidative stressand enhances the oxidation of LDL in the presence of copper (23), weasked whether 15(S)-HETE had any role in producing ROS. We foundthat 15(S)-HETE enhanced ROS production in a time-dependent manner,and that this effect depended on the activities of both xanthine oxidase andNADPH oxidase.

We also showed that 15(S)-HETE–stimulated NADPH oxidase activa-tion depended on xanthine oxidase activity. A large body of evidence sug-gests that NADPH oxidase is the major producer of ROS in vascular cellsand in other cell types (25, 28). Early studies showed that xanthine oxidaseplays a role in tissue reperfusion injury (29); however, to our knowledge,the possibility of there being an interaction between these two unrelatedoxidases has not been examined before. Because allopurinol, a competi-tive inhibitor of xanthine oxidase (33), or knockdown of xanthine oxidasewith a specific ASO blocked the 15(S)-HETE–dependent phosphorylationof p47Phox, a regulatory component of NADPH oxidase (29), it is likelythat xanthine oxidase plays a role in the activation of NADPH oxidase.Furthermore, phosphorylation of p47Phox is required for its membranetranslocation where it forms a complex with Nox to assemble the activeNADPH oxidase (25, 29). Thus, our findings uncover a potentially impor-tant mechanism by which xanthine oxidase and NADPH oxidase interactto stimulate the production of ROS in response to 15(S)-HETE.

www.S

Previous studies showed that oxidants, such as H2O2, inhibit PTPsand thereby enhance the phosphorylation of receptor and nonreceptortyrosine kinases, causing their activation (31, 32). Here, we provideevidence that 15(S)-HETE stimulated the tyrosine phosphorylation ofSyk and Pyk2 through the xanthine oxidase– and NADPH oxidase–dependent production of ROS, which, in turn, was required for themigration of THP1 cells and their adhesion to a HUVEC monolayer.Furthermore, Syk and Pyk2 mutually depended on each other for theirphosphorylation because inhibition of either kinase blocked the phos-phorylation of the other. In support of this mechanism, we observed thata physical interaction between Syk and Pyk2 was enhanced in responseto 15(S)-HETE. One possible explanation for the mutually dependentphosphorylation of Syk and Pyk2 might be that it is a result of theirassociation with a PTP. Inhibition of the associated PTP may lead to thephosphorylation and activation of one of these kinases acutely, which, in

Fig. 6. CREB mediates the 15(S)-HETE–dependent migration and adhe-sion of THP1 cells. (A) Whole-cell lysates of vehicle-treated THP1 cellsor THP1 cells treated with 0.1 mM 15(S)-HETE for the indicated timeswere analyzed by Western blotting with antibodies against the indicatedproteins. (B) THP1 cells transfected with control ASO or CREB-specificASOs were treated with vehicle or 0.1 mM 15(S)-HETE for 1 hour andthen analyzed by Western blotting with antibodies against the indicatedproteins (top) or analyzed for migration (middle) or adhesion (bottom).(C and D) THP1 cells transfected with control ASO or with ASOs specificfor (C) the indicated oxidases or (D) the indicated kinases were allowedto rest after transfection before they were treated with vehicle or 0.1 mM15(S)-HETE for 1 hour. Whole-cell lysates were analyzed by Western blot-ting with antibodies specific for the indicated proteins. Bar graph showsdensitometric analysis of the fold increase in the abundance of pCREB nor-malized to that of total CREB for each condition compared to that in vehicle-treated cells transfected with control ASO. (E) THP1 cells were treated withvehicle or 0.1 mM 15(S)-HETE in the absence or presence of the indicatedinhibitors for 1 hour before being analyzed by Western blotting with anti-bodies against the indicated proteins. Bar graph shows densitometric anal-ysis of the fold increase in the abundance of pCREB normalized to that oftotal CREB for each condition compared to that in vehicle-treated cells.Western blots are from one experiment and are representative of threeindependent experiments. Data in bar graphs are means ± SD fromthree independent experiments. *P < 0.01 versus control; **P < 0.01 versus15(S)-HETE.

C

trol 15(S)-HETE (min) CREB ASO1

Control ASO –––

–––

++A B

IENCESIGNALING.org 17 September 2013 Vol 6 Issue 293 ra83 7

Page 9

Con

t

60 120 240

15(S)-HETE (min)

30 360

TNF-

IL-11

IL-17

-Actin

Co

60 120 24030

IL-17

-Tubulin

360

0

468

2

1012141618

30 240 36060

Con

trol

15(S)-HETE (min)

120

*

*

*

*

IL-1

7 ab

unda

nce

(fol

d st

imul

atio

n)

0

20

40

60

10

30

50

70

Mig

ratio

n (c

ells

/fiel

d

**

*

*

Con

trol 5 10 20 50 100

IL-17 (ng/ml)

0

50

100

150

200

250

300

*

*

**

Adh

esio

n (R

FU

)

0

20

40

60

80

Mig

ratio

n (c

ells

/fiel

d)

IL-17 NabIgG IL-17 NabIgG

*

**

IL-17*

**

15(S)-HETE

0

10

20

40

30

**

*

Adh

esio

n (R

FU

)

**

*

IL-17

-Tubulin

– –XO ASO2+ +– – –

p47Phox ASO2 – –– –

15(S)-HETE – + + – +–

Control ASO

–

+ +

–

–

++

0

4

6

7

2

1

3

5

– –

XO ASO2

+ + – –p47Phox ASO2 ––– – + +

Control ASO

– – ++––

**

**

*

15(S)-HETESB203580

PD098059– –

–

– + +– –– –

– –

– + +

– + + ––+ + – –

– – – – + +– –

SP600125

IL-17

-Tubulin

CREB ASO3 + Control ASO

– –15(S)-HETE + +–

–+ + –+

–IL-17

CREB

-Tubulin

0

1

3

4

5

2

SB203580

PD098059– –

–

– –– –

– –– + + ––+ + – –

– – – – + +SP600125

*

**

** **

15(S)-HETE

0

0.5

1.0

3.0

2.0

1.5

2.5

3.5

CREB ASO3 + Control ASO

– ––+ + –

+

*

**

15(S)-HETE

– –Pyk2 ASO3+ +– – –

Syk ASO3 – –– –

15(S)-HETE – + + – +–

Control ASO

–

+ +

–

–

IL-17

-Tubulin

++

0

1

3

4

5

2

– –Pyk2 ASO3+ +– – –

Syk ASO3 – –– –Control ASO

–

+ +

–

–

**

**

* 15(S)-HETE

++

15(S)-HETE

IL-1

7 ab

unda

nce

(fol

d st

imul

atio

n)

IL-1

7 ab

unda

nce

(fol

d st

imul

atio

n)

IL-1

7 ab

unda

nce

(fol

d st

imul

atio

n)

IL-1

7 ab

unda

nce

(fol

d st

imul

atio

n)

E

D

F G H

R E S E A R C H A R T I C L E

on Septem

ber 17, 2013 stke.sciencem

ag.orgD

ownloaded from

turn, might have a priming effect on the phosphorylation of the other witha feed-forward mechanism.

Although a role for Syk in inflammation is not well established, theinvolvement of Pyk2 in inflammation and atherogenesis has been reported(40, 41). In addition, both Syk and Pyk2 play a role in vascular wall re-modeling in response to injury (42, 43). Here, we showed that both Syk

www.S

and Pyk2 interacted with each other and mediated monocyte migration.Previously, we showed that 15(S)-HETE activates MAPKs in endothelialcells and vascular smooth muscle cells to mediate their migration (15, 35).Consistent with these observations, we found that 15(S)-HETE also stimu-lated the activation of ERK1/2, JNK1/3, and p38 in THP1 cells to mediatetheir migration and adhesion. In addition, the 15(S)-HETE–dependent

Fig. 7. IL-17A mediates the 15(S)-HETE–stimulated migration and adhesion ofTHP1 cells. (A) THP1 cells were treatedwith vehicle or 0.1 mM 15(S)-HETE for theindicated times and then were analyzedby reverse transcription polymerase chainreaction (RT-PCR) to determine the abun-dances of mRNAs for IL-11, IL-17, TNF-a,and b-actin. (B) Whole-cell lysates of vehicle-treated THP1 cells or THP1 cells treatedwith 0.1 mM 15(S)-HETE for the indicatedtimes were analyzed by Western blottingto determine the abundance of IL-17 pro-tein. Bar graph shows densitometric analy-sis of the fold increase in the abundance ofIL-17 protein normalized to that of b-tubulinfor the indicated times in 15(S)-HETE–treatedcells compared to that in vehicle-treatedcells. (C) THP1 cells were treated with vehi-cle (control) or the indicated concentrationsof IL-17A before being analyzed for migra-tion (top) and adhesion (bottom). (D) THP1cells were pretreated with normal immuno-globulin G (IgG) or neutralizing antibody(Nab) against IL-17A, treated with IL-17A(20 ng/ml; left) or 0.1 mM 15(S)-HETE (right),and then analyzed for migration (top) or ad-hesion (bottom). (E and F) THP1 cells weretransfected with control ASO or with ASOsspecific for (E) the indicated oxidases or(F) the indicated kinases, rested, and thentreated with vehicle or 15(S)-HETE for 2 hours.Whole-cell lysates were then analyzed byWestern blotting with antibodies against theindicated proteins. Bar graphs show densi-tometric analysis of the fold increase in theabundance of IL-17 protein normalized tothat of b-tubulin for each condition com-pared to that in vehicle-treated cells trans-fected with control ASO. (G) THP1 cells weretreated with vehicle or 0.1 mM 15(S)-HETE inthe absence or presence of the indicated in-hibitors for 2 hours before being analyzed byWestern blotting with antibodies against theindicated proteins. Bar graph shows densito-

metric analysis of the fold increase in the abundance of IL-17 protein nor-malized to that of b-tubulin for each condition compared to that in vehicle-treated cells. (H) THP1 cells were transfected with control or CREB-specificASOs, rested, and then treated with vehicle or 0.1 mM 15(S)-HETE for 2 hoursbefore being analyzed by Western blotting with antibodies against the in-dicated proteins. Bar graph shows densitometric analysis of the fold in-

C

crease in the abundance of IL-17 protein normalized to that of b-tubulinfor each condition compared to that in vehicle-treated cells transfected withcontrol ASO. Western blots are from one experiment and are representativeof three independent experiments. Data in bar graphs are means ± SD fromthree independent experiments. *P < 0.01 versus control; **P < 0.01 versus15(S)-HETE.

rol ntro

l

15(S)-HETE (min)8090

) *

15(S)-HETEA B C

IENCESIGNALING.org 17 September 2013 Vol 6 Issue 293 ra83 8

Page 10

R E S E A R C H A R T I C L E

on Septem

ber 17, 2013 stke.sciencem

ag.orgD

ownloaded from

activation ofMAPK signaling required the activation of xanthine oxidase,NADPH oxidase, Syk, and Pyk2.

Many reports have shown that Src and Pyk2 mediate MAPK activationin response to external cues (31, 32, 44). We showed that by phosphoryl-ating and activating CREB, MAPKs modulate the migration of vascularsmooth muscle cells in response to 15(S)-HETE (15). Here, we also found

www.SCIENCESIGNALING.org 17 S

that 15(S)-HETE activated CREB, and thatstimulation of CREB activity was requiredfor the migration of THP1 cells and fortheir adhesion to endothelial cells. Further-more, the 15(S)-HETE–dependent activa-tion of CREB required the stimulation ofMAPK signaling downstream of xanthineoxidase, NADPH oxidase, Syk, and Pyk2.Studies indicate that CREB plays a role inthe hormonal regulation of cell metabolism,neuronal plasticity, memory, and cell sur-vival (45, 46). Our findings showed thatin monocytes, CREB enhanced the produc-tion of proinflammatory cytokines, such asIL-17A, revealing a possible role for CREBin mediating inflammation. Furthermore,the blockade of 15(S)-HETE–dependentmigration and adhesion of THP1 cells byneutralizing anti–IL-17A antibodies sug-gests that IL-17A mediates the inflamma-tory effects of CREB.

The adhesion of THP1 cells to a mono-layer of HUVECs may require the availabil-ity of adhesion molecules. Studies showedthat the LO-dependent products derivedfrom linoleic acid, such as 13-HpODE,stimulate the production of vascular cell ad-hesion molecule (VCAM) by human aorticendothelial cells (47). On the basis of thisinformation, we hypothesize that 15(S)-HETEmight stimulate the production of celladhesion molecules, such as VCAM, byHUVECs, which could facilitate the adhe-sion of monocytes and macrophages, andthat this effect may also depend on IL-17A.

Many studies have shown that deletionof the gene encoding 12/15-LO results indecreased atherosclerosis in ApoE−/− micefed a high-fat (Western) diet, which sug-gests a role for 12/15-LO in atherogenesis(5, 10). In addition, it was reported that mac-rophages expressing 12/15-LO contributeto atherogenesis (48). However, the pro-tective role of loss of 12/15-LO in athero-sclerosis was attributed to the involvementof 12/15-LO in LDL oxidation. Here, weobserved that peritoneal macrophages fromApoE−/− mice fed a high-fat diet exhibitedsubstantially increased xanthine oxidaseand NADPH oxidase activities, producedmore ROS, had increased phosphorylationand activation of Syk, Pyk2, MAPK, andCREB, and produced more IL-17A than didApoE−/−:12/15-LO−/−mice fedahigh-fat diet.

These observations suggest that 12/15-LO plays a critical role in ROS pro-duction, and that this response depends on the activities of xanthine oxidaseand NADPH oxidase.

A number of studies have shown that IL-17A plays a role in vascularinflammation and atherogenesis (49, 50); however, some reports indicatethat IL-17A alone may not be sufficient to account for aortic plaque burden

0

10

20

30

40

*

Mig

ratio

n (c

ells

/fiel

d)

0

200

400

600

800

1000

1200

Adh

esio

n (R

FU

)

*

ApoE–/–:12/15-LO–/–ApoE–/–

ApoE–/–:12/15-LO–/–ApoE–/–

ApoE–/–:12/15-LO–/–ApoE–/–

0

200

400

600

800

RO

S p

rodu

ctio

n (R

FU

)

*

0NA

DP

H o

xida

se a

ctiv

ity (

M)

20

40

60

80

100

120

140

*

0

50

200

100

150

250

XO

act

ivity

(R

FU

)

*

pSyk

pPyk2

Syk

Pyk2

pERK1/2

pJNK1/3

pp38 MAPK

p38 MAPK

JNK1

pCREB

IL-17

CREB

-Tubulin

ERK2

Apo

E–/

–

Apo

E–/

– :12

/15-

LO–/

–

0

2000

4000

6000

8000

10,000

12,000

IL-1

7 ab

unda

nce

(pg/

ml)

*

0

2000

4000

6000

8000

10,000

12,000

14,000

IL-1

0 ab

unda

nce

(pg/

ml) *

Ap

oE

–/–

DHE (red) H2DCFDA (green) Merged

H2O

2 ab

unda

nce

(1 ×

10

6 RF

U)

ApoE–/–:12/15-LO–/–ApoE–/–

0

1

2

3

4

*

Sup

erox

ide

anio

n ab

unda

nce

(1 ×

10

6 RF

U)

0

5

10

15

20

25

*

Ap

oE

–/–

Ap

oE

–/–:1

2/15

-LO

–/–

4× 10×

ApoE–/–:12/15-LO–/–ApoE–/–

Lesi

on a

rea

(1 ×

10

2 m

2 )

0

2

4

6

8

10

*

Apo

E:1

2/15

-LO

–/–

–/–

A B

D

C

E

FFig. 8. Loss of 12/15-LO inhibitsdiet-induced, CREB-dependentIL-17A production in macro-phages. (AandB)ApoE−/−miceand ApoE−/−:12/15-LO−/− micewere fed a high-fat diet for8weeks.Peritonealmacrophageswere then isolatedand (A)weresubjected to migration and ad-hesionassaysor (B)were lysed

and assayed for ROS production, NADPH oxidase activity,and xanthine oxidase activity. (C) Whole-cell lysates of peri-tonealmacrophages isolated from themicedescribed in (A)

wereanalyzedbyWesternblottingwithantibodies specific for the indicatedproteins.b-Tubulinwasusedasaloading control. (D) IL-17A and IL-10 concentrations in the plasma of the mice described in (A) weremeasured by enzyme-linked immunosorbent assay (ELISA). (E) ROS abundances in atherosclerotic lesionswere determined by staining cryosections of aortic roots isolated from the mice described in (A) withdihydroethidium (DHE; to detect the superoxide anion) and 2′,7′-dichlorodihydrofluorescein diacetate(H2DCFDA; to detect H2O2) andmeasuring fluorescence intensities. Left: images of the sections; right: pooleddata based on the fluorescence intensities. (F) Atherosclerotic lesions were assessed by Oil Red O stainingof cryosectionsof theaortic roots isolated from themicedescribed in (A) andmeasuring the lesionareas. Left:images of the sections; right: pooled data based on the staining intensities.

eptember 2013 Vol 6 Issue 293 ra83 9

Page 11

R E S E A R C H A R T I C L E

www.SC

stke.scD

ownloaded from

(51). Because loss of 12/15-LO in mice fed a high-fat diet reduced theproduction of IL-17A, we propose that 12/15-LO plays a role in high-fat diet–induced IL-17A production in ApoE−/−mice. In addition, we foundthat the increased abundance of IL-17A correlated with the number ofmacrophages in the lesion area, which was substantially higher in ApoE−/−

mice than in ApoE−/−:12/15-LO−/− mice. This suggests that the 12/15-LO–12/15(S)-HETE axis plays a role in stimulating IL-17A productionin macrophages, causing their migration in an autocrine manner. Further-more, given the known role for IL-17A in promoting inflammation (49–51),it is tempting to suggest that 12/15-LO, through its regulatory effect onIL-17A production, plays a major role in atherogenic inflammation. Inthis context, we should also point out that a report showed that 12/15-LOmediates inflammation in adipose tissue (21). Other reports showed thatthe cytokine IL-8 stimulates the production of 15-HETE by neutrophils(52). Here, we showed that 12/15-LO and 15(S)-HETE both played arole in increasing the production of IL-17A. On the basis of these obser-vations, we suggest that there is a feed-forward circuitry between 12/15-LOand cytokine regulation in the modulation of atherogenic inflammation.Because 15(S)-HETE stimulates the xanthine oxidase– and NADPHoxidase–dependent generation of ROS, it is possible that this mechanism,activated by 12/15-LO,mediates the oxidation of LDL in atherogenesis. Onthe basis of these findings, we suggest that 12/15-LO plays an importantrole in the generation of oxidative stress and inflammation in response toa high-fat or Western diet, and thereby might promote the development ofatherosclerotic lesions. Together, these findings suggest that 12/15-LOappears to be an upstream modulator of atherogenesis and that it mightbe a target for drug development.

on Septem

ber 17, 2013 iencem

ag.org

MATERIALS AND METHODS

ReagentsAnti-pCREB (9198), anti-pERK1/2 (9101), anti-pJNK1/3 (9251), anti–pp38 MAPK (9215), anti-pPyk2 (3291), and anti-pSyk (2715) antibodieswere purchased from Cell Signaling Technology. Anti-CD68 (SC-9139),anti-CREB (SC-58), anti-ERK2 (SC-154), anti–IL-17 (SC-374218), anti-JNK3 (SC-752), anti-Mac3 (SC-19991), anti–p38 MAPK (SC-538), anti-p47Phox (SC-14015), anti-Syk (SC-573), anti–b-tubulin (SC-9104), andanti–xanthine oxidase (SC-20991) antibodies were obtained from SantaCruz Biotechnology Inc. Anti-Pyk2 (ab32571) and anti-CD11b (ab8878)antibodies, as well as the IL-17–specific ELISA kit (ab100702), were pur-chased fromAbcam. 5(S)-HETE (34230), 12(S)-HETE (34570), 15(R)-HETE(34710), 15(S)-HETE (34720), and the xanthine oxidase kit (10010895) wereobtained from Cayman Chemicals. The NADPH oxidase kit (KA1663) wasobtained from Abnova. Thioglycolate medium, brewer-modified (21176),was purchased from BD Biosciences. Apocynin (A10809) and allopurinol(A8003) were obtained from Sigma-Aldrich Chemicals. Diphenyleneiodoniumchloride (BML-CN240) and PD098059 (BML-EI360) were purchasedfrom Enzo Life Science. SP328007 and SB559389 were obtained fromCalbiochem. Lipofectin transfection reagent (15596018), CM-H2DCFDA[5-(6)-chloromethyl-2′,7′-dichlorodihydrofluorescein diacetate] (C6827),DHE (D11347), BCECF-AM (B1170), and TRIzol reagent were obtainedfrom Invitrogen. Human recombinant IL-17 (317-ILB-050), IL-17–specificneutralizing antibody (AF-317-NA), and the IL-10 ELISA kit (M10000B) wereobtained fromR&DSystems. The enhanced chemiluminescence (ECL)West-ern blotting detection reagents (RPN2106) were obtained fromGEHealthcare.The following phosphorothioate-modified ASOs were synthesized byIDT: hControl ASO, 5′-GGGGGUTCTCTGCGTACGGTGCUAGU-3′;hp47Phox (NM_000265) ASO1, 5′-GGGCUCAGGGTCTTCCGTCUCGUC-3′;hp47Phox ASO2, 5′-GUUGGGCTCAGGGTCTTCCGUCUC-3′; human

ApoE–/–

ApoE–/– :12/15-LO–/–

4× 10×

ApoE–/–

ApoE–/– :12/15-LO

–/–

CD11b (red) CD68 (green) Merged

IL-17 (red) Mac3 (green) Merged (40×)

ApoE–/–

ApoE–/–:12/15-LO–/–

01234

67

5

89

*IL-1

7 ab

unda

nce

(1 ×

10

6 RF

U)

IL-1

7-po

sitiv

e ar

ea (

1 ×

103

m2 ) 10

12

14

8

6

4

2

0

**

Mac

3-po

sitiv

e ar

ea

(1 ×

103

m2 )

1012141618

86420 IL

-17/

Mac

3-po

sitiv

e ar

ea

1.0

0.8

0.6

0.4

0.2

0

ApoE–/–:12/15-LO–/–ApoE–/–

A

B

C

Fig. 9. Loss of 12/15-LO attenuates the migration and adhesion of macro-phages. (A) The extent of recruitment of monocytes in atherosclerotic le-

sions was examined by immunofluorescence staining of the cryosectionsdescribed in Fig. 8E for CD11b and CD68. (B) The recruitment of macro-phages in atherosclerotic lesions was examined by immunostaining of thecryosections described in Fig. 8E for Mac3. (C) Double immunofluores-cence staining of the cryosections described in Fig. 8E for IL-17A (red)and Mac3 (green). The bar graphs represent IL-17A abundance in thelesion, the IL-17A–positive area in the lesion, the Mac3-positive area inthe lesion, and the lesion area positive for both IL-17A and Mac3 in cryo-sections from ApoE−/− mice and ApoE−/−:12/15-LO−/− mice fed a high-fatdiet. *P < 0.01 versus ApoE−/− mice (n = 6 mice per group).

IENCESIGNALING.org 17 September 2013 Vol 6 Issue 293 ra83 10

Page 12

R E S E A R C H A R T I C L E

on Septem

ber 17, 2013 stke.sciencem

ag.orgD

ownloaded from

xanthine oxidase (NM_000379) ASO1, 5′-CCCAUCTCCTCCACAGCAU-CCACC-3′; human xanthine oxidase ASO2, 5′-GCCUCCTCCCATTCTCT-TCACUCG-3′; hSyk (NM_001174168)ASO1, 5′-CCCAUCCGCTCTCCTU-UCUC-3′; hSyk ASO2, 5′-CUUCCCTGTCTTGTCUUUGU-3′; hSykASO3, 5′-UUCCCTGTCTTGTCTUUGUC-3′; hPyk2 (NM_173175)ASO1, 5′-GGCGGUTCTGCGTACGGTGCUAGU-3′; hPyk2 ASO2, 5′-GGUCUGTACTTAGGTCGGCUGGGC-3′; hPyk2 ASO3, 5′-CCU-GUGTCCATAGCCCAGAGUACC-3′; hCREB (NM_004379) ASO1,5′-CUCUGCTGGTTCTCGGCUCC-3′; hCREB ASO2, 5′-GUGGCAA-TCTGTGGCUGGGC-3′; and hCREB ASO3, 5′-GCUGCTTCCCTGTT-CUUCAU-3′. h, human.

Cell cultureTHP1 cells, a human leukemicmonocyte cell line, were purchased from theAmericanTypeCultureCollection (ATCC) and cultured inRPMI1640 sup-plemented with 10% heat-inactivated fetal bovine serum (FBS), penicillin(50 U/ml), streptomycin (50 mg/ml), and 50 mM b-mercaptoethanol.HUVECs were obtained from Invitrogen and were cultured in Medium 200with low-serum growth supplement, containing gentamicin (10 mg/ml) andamphotericin B (0.25 mg/ml). Mouse pancreatic endothelial cells were ob-tained from ATCC and were maintained in Dulbecco’s modified Eagle’smedium–F12 supplemented with 10% FBS, penicillin (50 U/ml), andstreptomycin (50 mg/ml). All cell culturesweremaintained in a humidified95% air and 5% CO2 atmosphere at 37°C.

Cell migration assaysA modified Boyden chamber method was used to measure cell migration(15). THP1 cells were serum-starved overnight in serum-free RPMI 1640.Cells were resuspended in serum-free medium and plated on Matrigel-coated cell culture inserts at 5 × 104 cells per insert. 15(S)-HETE was addedto a final concentration of 0.1 mM to both the upper and lower chambers,and the cells were incubated for 8 hours at 37°C. The inserts were thenlifted, nonmigrated cells on the upper surface of the membrane were re-moved with a cotton swab, and the membrane was then fixed in methanoland stained with 4′,6-diamidino-2-phenylindole (DAPI) to visualize thenuclei. Whenever ASOs were used, cells were transfected first with the in-dicated ASOs for 6 hours in serum-free medium, maintained in completemedium for 36 hours, and then subjected to growth arrest and analyzed formigration. Whenever inhibitors were used, cells were incubated first withthe indicated inhibitor for 30 min. The DAPI-positive cells were countedunder an inverted microscope (Carl Zeiss AxioVision AX10), and the extentof migration was expressed as the number of cells per field of view.

Cell adhesion assaysThe adhesion of THP1 cells to HUVECs was measured with a fluorometricmethod (53). Briefly, quiescent THP1 cells were left untreated or were treatedwith 0.1 mM15(S)-HETE for 1 hour at 37°C and thenwere labeledwith 10 µMBCECF-AMin serum-freemediumfor 30min.The labeled cellswereplatedona quiescentmonolayer ofHUVECs at 8 × 104 cells per well andwere incubatedfor a further 2 hours, after which, any nonadherent THP1 cells werewashed outwith phosphate-buffered saline (PBS). The adherent BCECF-AM–labeledTHP1 cells were then lysed in 0.2 ml of 0.1M tris-HCl containing 0.1% TritonX-100, and the fluorescence intensity was measured in a SpectraMax GeminiXSspectrofluorometer (MolecularDevices)with excitationat 485nmandemis-sion at 535 nm. Cell adhesion was expressed as relative fluorescence units.

Enzyme-linked immunosorbent assayThe concentrations of IL-10 and IL-17 in plasma from mice were mea-sured with specific ELISA kits according to the manufacturers’ instructions(R&D Systems and Abcam, respectively).

www.SC

ROS detectionIntracellular ROS generation was measured with membrane-permeableCM-H2DCFDA, as described previously (54). After the treatments indicatedin the figure legends, THP1 cells were incubated with 10 mMCM-H2DCFDAfor 30 min, washed with PBS, and resuspended in serum-free medium. The flu-orescence intensities of the resuspended cells were measured in a SpectraMaxGemini XS spectrofluorometer (Molecular Devices) with excitation at 485 nmand emission at 535 nm. To measure ROS concentrations in mouse ar-teries, aortic arch sections of ApoE−/− mice and ApoE−/−:12/15-LO−/− micefed a high-fat diet were stained with the dyes DHE and CM-H2DCFDA todetect superoxide anions and H2O2, respectively (55), and fluorescenceintensities were measured with Nikon NIS-Elements software version AR3.1. ROS production was expressed as relative fluorescence units.

Reverse transcription polymerase chain reactionTotal cellular RNA was extracted from THP1 cells with TRIzol reagentaccording to the manufacturer’s instructions. Reverse transcription was per-formed with a High Capacity cDNA Reverse Transcription Kit (AppliedBiosystems). Complementary DNA (cDNA) was then used as a templatefor amplification by PCR with the following primers: human IL-11(NM_000641), 5′-CTGAGCCTGTGGCCAGATACA-3′ (forward) and5′-CTCCAGGGTCTTCAGGGAAGA-3′ (reverse); human TNF-a(NM_000594), 5′-CAGAGGGCCTGTACCTCATC-3′ (forward) and 5′-GGAAGACCCCTCCCAGATAG-3′ (reverse); human IL-17A(NM_002190), 5′-GCAATGAGGACCCTGAGAGA-3′ (forward) and 5′-CCCACGGACACCAGTATCTT-3′ (reverse); human b-actin (NM_001101),5′-AGCCATGTACGTTGCTAT-3′ (forward) and 5′-GATGTCCACGTCA-CACTTCA-3′ (reverse). The amplification was performed with GeneAmpPCR System 2400 (Applied Biosystems). The amplified PCR products wereseparated on 1.5% (w/v) agarose gels and stained with ethidium bromide,and images were captured with the Kodak In Vivo Imaging System.

Western blottingWhole-cell lysates containing equal amounts of protein were resolved byelectrophoresis on 0.1% (w/v) SDS and 10% (w/v) polyacrylamide gels.The proteins were transferred electrophoretically onto a nitrocellulose mem-brane. After being blocked in either 5% (w/v) nonfat dry milk or 5% (w/v)bovine serum albumin (BSA), the membrane was incubated with the ap-propriate primary antibodies, followed by incubation with horseradishperoxidase–conjugated secondary antibodies. The antigen-antibody complexeswere detected with the ECL detection reagent kit (GE Healthcare).

Measurement of NADPH oxidase andxanthine oxidase activitiesNADPH oxidase and xanthine oxidase activities were measured with kitsaccording to the manufacturers’ instructions (Abnova and Cayman Chemicals,respectively).

TransfectionsCells were transfected with the indicated ASOs (100 nM) with Lipofectintransfection reagent for 6 hours according to the manufacturer’s instruc-tions. After transfection, cells were maintained in complete RPMI 1640for 36 hours and then were incubated with serum-free medium overnightbefore being used for experiments.

Isolation of peritoneal macrophages and collection ofplasma and aortas from miceApoE−/−mice (stock no. 002052) were obtained from Jackson Laboratories.ApoE−/−mice and ApoE−/−:12/15-LO−/−mice (5) were bred and maintainedaccording to the guidelines of the Institutional Animal Care and Use Facility.

IENCESIGNALING.org 17 September 2013 Vol 6 Issue 293 ra83 11

Page 13

R E S E A R C H A R T I C L E

on Septem

ber 17, 2013 stke.sciencem

ag.orgD

ownloaded from

All of the experiments involving mice were approved by the Animal Careand Use Committee of the University of Tennessee Health Science Center.Six-week-old male ApoE−/− mice and ApoE−/−:12/15-LO−/− mice werefed a high-fat diet (protein, 15.2%; carbohydrate, 42.7%; fat, 42%; Harlancat. no. TD.88137) for 8 weeks. To collect peritoneal macrophages, micewere injected intraperitoneally with 2 ml of autoclaved 4% thioglycolate.Four days later, the animals were anesthetized with ketamine and xylazine,and the peritoneal lavage was collected in RPMI 1640. Cells were incu-bated at 3 × 105 cells/cm2 in RPMI medium containing penicillin (50 µg/ml)and streptomycin (50 µg/ml). After 3 hours, floating cells (mostly red bloodcells) were removed by washing with cold PBS. Adherent cells (macro-phages) were used either for the extraction of total cellular protein forWestern blotting analysis or to perform assays to measure migration, ad-hesion, ROS production, NADPH oxidase activity, or xanthine oxidaseactivity. Blood was drawn into 1.5-ml tubes containing anticoagulant(20 ml of 0.5 M EDTA per tube) by cardiac puncture and centrifugedat 3500g for 5 min at 4°C to collect the plasma. The plasma was used tomeasure the concentrations of cytokines. The aortas and aortic arches wereperfused with PBS, dissected, and fixed in OCT (optimal cutting tempera-ture) compound.

ImmunohistochemistrySerial aortic arch sections (10 mm) were made with a Leica CM3050 Scryostat machine, fixed in an equal volume of cold acetone and methanolfor 10 min, permeabilized in 0.2% Triton X-100 for 10 min, and blockedwith 3% BSA for 30 min. The sections were then incubated with anti-Mac3 (clone M3/84) primary antibody overnight at 4°C, which was fol-lowed by incubation with biotin-conjugated mouse anti-rat secondaryantibody for 1 hour at room temperature. The sections were incubatedwith the ABC (avidin-biotin complex) reagent for 30 min, developedwith DAB (3,3′-diaminobenzidine) reagent (Vector Laboratories), andcounterstained with hematoxylin. The sections were observed with aNikon Eclipse 50i microscope with 4×/0.10 or 10×/0.25 magnification,and images were captured with a Nikon Digital Sight DS-L1 camera.

Oil Red O stainingAfter being fixed with 4% (v/v) paraformaldehyde, the aortic arch sectionswere stained with Oil Red O and counterstained with hematoxylin. Imageswere captured as described for immunohistochemistry.

Immunofluorescence stainingThe aortic arch sections were fixed with acetone/methanol (1:1) for 10 min,permeabilized in 0.2% Triton X-100 for 10 min, blocked with 5% goatserum in 3% BSA for 1 hour, and incubated with rat anti-mouse CD11band rabbit anti-mouse CD68 or with rat anti-mouse Mac3 and mouse anti–human IL-17 antibodies (all at a 1:100 dilution), followed by incubationwith Alexa Fluor 568–conjugated goat anti-rat and Alexa Fluor 488–conjugated goat anti-rabbit or with Alexa Fluor 488–conjugated goat anti-rat and Alexa Fluor 568–conjugated goat anti-mouse secondary antibodies,respectively. The sections were observed with a Zeiss inverted microscope[Zeiss AxioObserver.Z1; type, LD Plan-Neofluar; magnification at 10×/NA(numerical aperture) 0.4 or 40×/NA 0.6], and the fluorescence images werecaptured with a Zeiss AxioCam MRm camera with the microscopeoperating software and Image Analysis Software AxioVision 4.7.2 (CarlZeiss Imaging Solutions GmbH). The fluorescence intensities were mea-sured with Nikon NIS-Elements software version AR 3.1.

Statistical analysisAll the experiments were performed three times, and data are presented asmeans ± SD. Statistical differences between untreated and treated samples

www.SC

were analyzed with Student’s t test; P < 0.05 was considered statisticallysignificant. In the case of RT-PCR and Western blotting assays, histo-chemistry, immunohistochemistry, and immunofluorescence staining,one representative set of data is shown. In the case of experimentswithmice,each group consists of six animals.

REFERENCES AND NOTES1. G. K. Hansson, A. Hermansson, The immune system in atherosclerosis. Nat. Immunol.

12, 204–212 (2011).2. P. Libby, P. M. Ridker, G. K. Hansson, Progress and challenges in translating the

biology of atherosclerosis. Nature 473, 317–325 (2011).3. H. C. Stary, A. B. Chandler, R. E. Dinsmore, V. Fuster, S. Glagov, W. Insull Jr.,

M. E. Rosenfeld, C. J. Schwartz, W. D. Wagner, R. W. Wissler, A definition ofadvanced types of atherosclerotic lesions and a histological classification ofatherosclerosis. A report from the Committee on Vascular Lesions of the Council onArteriosclerosis, American Heart Association. Circulation 92, 1355–1374 (1995).

4. S. M. Sendobry, J. A. Cornicelli, K. Welch, T. Bocan, B. Tait, B. K. Trivedi, N. Colbry,R. D. Dyer, S. J. Feinmark, A. Daugherty, Attenuation of diet-induced atherosclerosisin rabbits with a highly selective 15-lipoxygenase inhibitor lacking significant antioxidantproperties. Br. J. Pharmacol. 120, 1199–1206 (1997).

5. T. Cyrus, J. L. Witztum, D. J. Rader, R. Tangirala, S. Fazio, M. F. Linton, C. D. Funk,Disruption of the 12/15-lipoxygenase gene diminishes atherosclerosis in apo E–deficientmice. J. Clin. Invest. 103, 1597–1604 (1999).

6. D. Harats, A. Shaish, J. George, M. Mulkins, H. Kurihara, H. Levkovitz, E. Sigal, Over-expression of 15-lipoxygenase in vascular endothelium accelerates early athero-sclerosis in LDL receptor–deficient mice. Arterioscler. Thromb. Vasc. Biol. 20,2100–2105 (2000).

7. M. Mehrabian, H. Allayee, J. Wong, W. Shi, X. P. Wang, Z. Shaposhnik, C. D. Funk,A. J. Lusis, Identification of 5-lipoxygenase as a major gene contributing to athero-sclerosis susceptibility in mice. Circ. Res. 91, 120–126 (2002).

8. S. Ylä-Herttuala, W. Palinski, M. E. Rosenfeld, S. Parthasarathy, T. E. Carew, S. Butler,J. L. Witztum, D. Steinberg, Evidence for the presence of oxidatively modified low densitylipoprotein in atherosclerotic lesions of rabbit and man. J. Clin. Invest. 84, 1086–1095(1989).

9. S. Ylä-Herttuala, J. Luoma, H. Viita, T. Hiltunen, T. Sisto, T. Nikkari, Transfer of15-lipoxygenase gene into rabbit iliac arteries results in the appearance of oxidation-specific lipid-protein adducts characteristic of oxidized low density lipoprotein.J. Clin. Invest. 95, 2692–2698 (1995).

10. T. Cyrus, D. Praticò, L. Zhao, J. L. Witztum, D. J. Rader, J. Rokach, G. A. FitzGerald,C. D. Funk, Absence of 12/15-lipoxygenase expression decreases lipid peroxidationand atherogenesis in apolipoprotein E–deficient mice. Circulation 103, 2277–2282(2001).

11. D. J. Benz, M. Mol, M. Ezaki, N. Mori-Ito, I. Zelán, A. Miyanohara, T. Friedmann,S. Parthasarathy, D. Steinberg, J. L. Witztum, Enhanced levels of lipoperoxidesin low density lipoprotein incubated with murine fibroblast expressing high levelsof human 15-lipoxygenase. J. Biol. Chem. 270, 5191–5197 (1995).

12. A. R. Brash, Lipoxygenases: Occurrence, functions, catalysis, and acquisition ofsubstrate. J. Biol. Chem. 274, 23679–23682 (1999).

13. H. Kühn, J. Barnett, D. Grunberger, P. Baecker, J. Chow, B. Nguyen, H. Bursztyn-Pettegrew,H. Chan, E. Sigal, Overexpression, purification and characterization of human recombi-nant 15-lipoxygenase. Biochim. Biophys. Acta 1169, 80–89 (1993).

14. H. S. Potula, D. Wang, D. V. Quyen, N. K. Singh, V. Kundumani-Sridharan, M. Karpurapu,E. A. Park, W. C. Glasgow, G. N. Rao, Src-dependent STAT-3-mediated expression ofmonocyte chemoattractant protein-1 is required for 15(S)-hydroxyeicosatetraenoic acid-induced vascular smooth muscle cell migration. J. Biol. Chem. 284, 31142–31155(2009).

15. K. R. Chava, M. Karpurapu, D. Wang, M. Bhanoori, V. Kundumani-Sridharan, Q. Zhang,T. Ichiki, W. C. Glasgow, G. N. Rao, CREB-mediated IL-6 expression is requiredfor 15(S)-hydroxyeicosatetraenoic acid–induced vascular smooth muscle cellmigration. Arterioscler. Thromb. Vasc. Biol. 29, 809–815 (2009).

16. P. Henriksson, M. Hamberg, U. Diczfalusy, Formation of 15-HETE as a major hydroxy-eicosatetraenoic acid in the atherosclerotic vessel wall. Biochim. Biophys. Acta834, 272–274 (1985).

17. T. C. Simon, A. N. Makheja, J. M. Bailey, Formation of 15-hydroxyeicosatetraenoicacid (15-HETE) as the predominant eicosanoid in aortas from Watanabe HeritableHyperlipidemic and cholesterol-fed rabbits. Atherosclerosis 75, 31–38 (1989).

18. T. Hiltunen, J. Luoma, T. Nikkari, S. Ylä-Herttuala, Induction of 15-lipoxygenasemRNA and protein in early atherosclerotic lesions. Circulation 92, 3297–3303 (1995).

19. N. T. Aggarwal, S. L. Pfister, W. B. Campbell, Hypercholesterolemia enhances15-lipoxygenase–mediated vasorelaxation and acetylcholine-induced hypotension.Arterioscler. Thromb. Vasc. Biol. 28, 2209–2215 (2008).

IENCESIGNALING.org 17 September 2013 Vol 6 Issue 293 ra83 12

Page 14

R E S E A R C H A R T I C L E

on Septem

ber 17, 2013 stke.sciencem

ag.orgD

ownloaded from

20. M. E. Hatley, S. Srinivasan, K. B. Reilly, D. T. Bolick, C. C. Hedrick, Increasedproduction of 12/15 lipoxygenase eicosanoids accelerates monocyte/endothelialinteractions in diabetic db/db mice. J. Biol. Chem. 278, 25369–25375 (2003).

21. S. K. Chakrabarti, B. K. Cole, Y. Wen, S. R. Keller, J. L. Nadler, 12/15-lipoxygenaseproducts induce inflammation and impair insulin signaling in 3T3-L1 adipocytes. Obesity17, 1657–1663 (2009).

22. H. Yoshida, K. Sasaki, Y. Hirowatari, H. Kurosawa, N. Sato, N. Furutani, N. Tada,Increased serum iron may contribute to enhanced oxidation of low-density lipoproteinin smokers in part through changes in lipoxygenase and catalase. Clin. Chim. Acta345, 161–170 (2004).

23. V. J. O’Leary, V. M. Darley-Usmar, L. J. Russell, D. Stone, Pro-oxidant effects oflipoxygenase-derived peroxides on the copper-initiated oxidation of low-densitylipoprotein. Biochem. J. 282 (pt. 3), 631–634 (1992).

24. J. A. Leopold, J. Loscalzo, Oxidative risk for atherothrombotic cardiovascular disease.Free Radic. Biol. Med. 47, 1673–1706 (2009).

25. J. D. Lambeth, Nox enzymes, ROS, and chronic disease: An example of antagonisticpleiotropy. Free Radic. Biol. Med. 43, 332–347 (2007).

26. E. Galkina, K. Ley, Immune and inflammatory mechanisms of atherosclerosis.Annu. Rev. Immunol. 27, 165–197 (2009).

27. I. Tabas, C. K. Glass, Anti-inflammatory therapy in chronic disease: Challenges andopportunities. Science 339, 166–172 (2013).

28. G. R. Drummond, S. Selemidis, K. K. Griendling, C. G. Sobey, Combating oxidativestress in vascular disease: NADPH oxidases as therapeutic targets. Nat. Rev. DrugDiscov. 10, 453–471 (2011).

29. C. E. Berry, J. M. Hare, Xanthine oxidoreductase and cardiovascular disease: Molec-ular mechanisms and pathophysiological implications. J. Physiol. 555, 589–606(2004).

30. R. A. Oeckler, P. M. Kaminski, M. S. Wolin, Stretch enhances contraction of bovinecoronary arteries via an NAD(P)H oxidase–mediated activation of the extracellularsignal–regulated kinase mitogen-activated protein kinase cascade. Circ. Res. 92,23–31 (2003).

31. N. K. Tonks, Redox redux: Revisiting PTPs and the control of cell signaling. Cell 121,667–670 (2005).

32. I. Nakashima, K. Takeda, Y. Kawamoto, Y. Okuno, M. Kato, H. Suzuki, Redox controlof catalytic activities of membrane-associated protein tyrosine kinases. Arch. Biochem.Biophys. 434, 3–10 (2005).

33. N. Yamamoto, K. Takeshita, M. Shichijo, T. Kokubo, M. Sato, K. Nakashima, M. Ishimori,H. Nagai, Y. F. Li, T. Yura, K. B. Bacon, The orally available spleen tyrosine kinaseinhibitor 2-[7-(3,4-dimethoxyphenyl)-imidazo[1,2-c]pyrimidin-5-ylamino]nicotinamidedihydrochloride (BAY 61-3606) blocks antigen-induced airway inflammation in rodents.J. Pharmacol. Exp. Ther. 306, 1174–1181 (2003).

34. L. Buckbinder, D. T. Crawford, H. Qi, H. Z. Ke, L. M. Olson, K. R. Long, P. C. Bonnette,A. P. Baumann, J. E. Hambor, W. A. Grasser, L. C. Pan, T. A. Owen, M. J. Luzzio,C. A. Hulford, D. F. Gebhard, V. M. Paralkar, H. A. Simmons, J. C. Kath, W. G. Roberts,S. L. Smock, A. Guzman-Perez, T. A. Brown, M. Li, Proline-rich tyrosine kinase 2regulates osteoprogenitor cells and bone formation, and offers an anabolic treat-ment approach for osteoporosis. Proc. Natl. Acad. Sci. U.S.A. 104, 10619–10624(2007).

35. A. K. Bajpai, E. Blaskova, S. B. Pakala, T. Zhao, W. C. Glasgow, J. S. Penn,D. A. Johnson, G. N. Rao, 15(S)-HETE production in human retinal microvascularendothelial cells by hypoxia: Novel role for MEK1 in 15(S)-HETE–induced angio-genesis. Invest. Ophthalmol. Vis. Sci. 48, 4930–4938 (2007).

36. T. Zhao, D. Wang, S. Y. Cheranov, M. Karpurapu, K. R. Chava, V. Kundumani-Sridharan,D. A. Johnson, J. S. Penn, G. N. Rao, A novel role for activating transcription factor-2 in15(S)-hydroxyeicosatetraenoic acid-induced angiogenesis. J. Lipid Res. 50, 521–533(2009).

37. R. Ross, Atherosclerosis—An inflammatory disease. N. Engl. J. Med. 340, 115–126(1999).

38. C. Morgantini, S. Imaizumi, V. Grijalva, M. Navab, A. M. Fogelman, S. T. Reddy, Apo-lipoprotein A-I mimetic peptides prevent atherosclerosis development and reduceplaque inflammation in a murine model of diabetes. Diabetes 59, 3223–3228 (2010).

39. D. Poeckel, K. A. Zemski Berry, R. C. Murphy, C. D. Funk, Dual 12/15- and 5-lipoxygenasedeficiency in macrophages alters arachidonic acid metabolism and attenuates peritonitisand atherosclerosis in ApoE knock-out mice. J. Biol. Chem. 284, 21077–21089 (2009).

40. M. Okigaki, C. Davis, M. Falasca, S. Harroch, D. P. Felsenfeld, M. P. Sheetz,J. Schlessinger, Pyk2 regulates multiple signaling events crucial for macrophagemorphology and migration. Proc. Natl. Acad. Sci. U.S.A. 100, 10740–10745 (2003).

41. A. Katsume, M. Okigaki, A. Matsui, J. Che, Y. Adachi, E. Kishita, S. Yamaguchi,K. Ikeda, T. Ueyama, S. Matoba, H. Yamada, H. Matsubara, Early inflammatoryreactions in atherosclerosis are induced by proline-rich tyrosine kinase/reactive

www.SC

oxygen species–mediated release of tumor necrosis factor-a and subsequentactivation of the p21Cip1/Ets-1/p300 system. Arterioscler. Thromb. Vasc. Biol.31, 1084–1092 (2011).

42. P. Andre, T. Morooka, D. Sim, K. Abe, C. Lowell, N. Nanda, S. Delaney, G. Siu,Y. Yan, S. Hollenbach, A. Pandey, H. Gao, Y. Wang, K. Nakajima, S. A. Parikh,C. Shi, D. Phillips, W. Owen, U. Sinha, D. I. Simon, Critical role for Syk in responsesto vascular injury. Blood 118, 5000–5010 (2011).

43. R. Gadepalli, N. K. Singh, V. Kundumani-Sridharan, M. R. Heckle, G. N. Rao, Novelrole of proline-rich nonreceptor tyrosine kinase 2 in vascular wall remodeling afterballoon injury. Arterioscler. Thromb. Vasc. Biol. 32, 2652–2661 (2012).

44. J. M. Kyriakis, J. Avruch, Mammalian MAPK signal transduction pathways activatedby stress and inflammation: A 10-year update. Physiol. Rev. 92, 689–737 (2012).

45. J. Y. Altarejos, M. Montminy, CREB and the CRTC co-activators: Sensors for hormonaland metabolic signals. Nat. Rev. Mol. Cell Biol. 12, 141–151 (2011).

46. E. Benito, A. Barco, CREB’s control of intrinsic and synaptic plasticity: Implications forCREB-dependent memory models. Trends Neurosci. 33, 230–240 (2010).

47. B. V. Khan, S. S. Parthasarathy, R. W. Alexander, R. M. Medford, Modified lowdensity lipoprotein and its constituents augment cytokine-activated vascular celladhesion molecule-1 gene expression in human vascular endothelial cells. J. Clin.Invest. 95, 1262–1270 (1995).

48. Y. Huo, L. Zhao, M. C. Hyman, P. Shashkin, B. L. Harry, T. Burcin, S. B. Forlow,M. A. Stark, D. F. Smith, S. Clarke, S. Srinivasan, C. C. Hedrick, D. Praticò, J. L. Witztum,J. L. Nadler, C. D. Funk, K. Ley, Critical role of macrophage 12/15-lipoxygenase foratherosclerosis in apolipoprotein E–deficient mice. Circulation 110, 2024–2031 (2004).

49. S. Taleb, M. Romain, B. Ramkhelawon, C. Uyttenhove, G. Pasterkamp, O. Herbin,B. Esposito, N. Perez, H. Yasukawa, J. Van Snick, A. Yoshimura, A. Tedgui, Z. Mallat,Loss of SOCS3 expression in T cells reveals a regulatory role for interleukin-17 inatherosclerosis. J. Exp. Med. 206, 2067–2077 (2009).

50. E. Smith, K. M. Prasad, M. Butcher, A. Dobrian, J. K. Kolls, K. Ley, E. Galkina, Blockadeof interleukin-17A results in reduced atherosclerosis in apolipoprotein E–deficient mice.Circulation 121, 1746–1755 (2010).

51. M. S. Madhur, S. A. Funt, L. Li, A. Vinh, W. Chen, H. E. Lob, Y. Iwakura, Y. Blinder,A. Rahman, A. A. Quyyumi, D. G. Harrison, Role of interleukin 17 in inflammation,atherosclerosis, and vascular function in apolipoprotein E–deficient mice. Arterioscler.Thromb. Vasc. Biol. 31, 1565–1572 (2011).

52. K. Fogh, C. G. Larsen, L. Iversen, K. Kragballe, Interleukin-8 stimulates the formationof 15-hydroxy-eicosatetraenoic acid by human neutrophils in vitro. Agents Actions 35,227–231 (1992).

53. L. H. Zhang, V. S. Kamanna, S. H. Ganji, X. M. Xiong, M. L. Kashyap, Pioglitazoneincreases apolipoprotein A-I production by directly enhancing PPRE-dependenttranscription in HepG2 cells. J. Lipid Res. 51, 2211–2222 (2010).

54. N. K. Singh, D. Wang, V. Kundumani-Sridharan, D. Van Quyen, J. Niu, G. N. Rao,15-Lipoxygenase-1-enhanced Src-Janus kinase 2-signal transducer and activatorof transcription 3 stimulation and monocyte chemoattractant protein-1 expressionrequire redox-sensitive activation of epidermal growth factor receptor in vascularwall remodeling. J. Biol. Chem. 286, 22478–22488 (2011).

55. X. Y. Tian, W. T. Wong, A. Xu, Y. Lu, Y. Zhang, L. Wang, W. S. Cheang, Y. Wang,X. Yao, Y. Huang, Uncoupling protein-2 protects endothelial function in diet-inducedobese mice. Circ. Res. 110, 1211–1216 (2012).

Acknowledgments:We are thankful to J. L. Nadler of Eastern Virginia Medical School forproviding us with ApoE−/−:12/15-LO−/− mice. Funding: This work was supported by grantsHL064165 and HL074860 from the National Heart, Lung, and Blood Institute of the NIH (toG.N.R.). Author contributions: S.K. performed monocyte migration and adhesion as-says, Western blotting, immunoprecipitations, ELISA, ROS measurements, RT-PCR, mousefeeding with a high-fat diet, and isolation of peritoneal macrophages; M.R.H. performedWestern blotting and macrophage isolation; N.K.S. performed monocyte migration andadhesion assays, Western blotting, immunoprecipitations, ROS measurements, isolationof peritoneal macrophages, and immunofluorescence staining of aortic arch sections;G.J.T. participated in project discussions and wrote the manuscript; and G.N.R. conceived theoverall project, designed the experiments, interpreted the data, and wrote the manuscript.Competing interests: The authors declare that they have no competing interests.

Submitted 4 May 2013Accepted 3 September 2013Final Publication 17 September 201310.1126/scisignal.2004214Citation: S. Kotla, N. K. Singh, M. R. Heckle, G. J. Tigyi, G. N. Rao, The transcriptionfactor CREB enhances interleukin-17A production and inflammation in a mouse modelof atherosclerosis. Sci. Signal. 6, ra83 (2013).

IENCESIGNALING.org 17 September 2013 Vol 6 Issue 293 ra83 13