Sathish Babu Vasamsetti,1 Santosh Karnewar,1 Anantha Koteswararao Kanugula,1

Avinash Raj Thatipalli,2 Jerald Mahesh Kumar,2 and Srigiridhar Kotamraju1

Metformin Inhibits Monocyte-to-MacrophageDifferentiation via AMPK-Mediated Inhibitionof STAT3 Activation: Potential Role inAtherosclerosisDiabetes 2015;64:2028–2041 | DOI: 10.2337/db14-1225

Monocyte-to-macrophage differentiation is a criticalevent that accentuates atherosclerosis by promotingan inflammatory environment within the vessel wall. Inthis study, we investigated the molecular mechanismsresponsible for monocyte-to-macrophage differentia-tion and, subsequently, the effect of metformin inregressing angiotensin II (Ang-II)-mediated atheroma-tous plaque formation in ApoE2/2 mice. AMPK activitywas dose and time dependently downregulated duringphorbol myristate acetate (PMA)-induced monocyte-to-macrophage differentiation, which was accompa-nied by an upregulation of proinflammatory cytokineproduction. Of note, AMPK activators metformin andAICAR significantly attenuated PMA-induced monocyte-to-macrophage differentiation and proinflammatorycytokine production. However, inhibition of AMPK ac-tivity alone by compound C was ineffective in promot-ing monocyte-to-macrophage differentiation in theabsence of PMA. On the other hand, inhibition ofc-Jun N-terminal kinase activity inhibited PMA-inducedinflammation but not differentiation, suggesting that in-flammation and differentiation are independent events.In contrast, inhibition of STAT3 activity inhibited both in-flammation and monocyte-to-macrophage differentia-tion. By decreasing STAT3 phosphorylation, metforminand AICAR through increased AMPK activation causedinhibition of monocyte-to-macrophage differentiation.Metformin attenuated Ang-II–induced atheromatousplaque formation and aortic aneurysm in ApoE2/2 micepartly by reducing monocyte infiltration. We conclude

that the AMPK-STAT3 axis plays a pivotal role in regulat-ing monocyte-to-macrophage differentiation and that bydecreasing STAT3 phosphorylation through increasedAMPK activity, AMPK activators inhibit monocyte-to-macrophage differentiation.

Substantial evidence implicates macrophages as abun-dantly present at all stages of the atheroscleroticdisease process (1). Macrophages fuel an inflammatoryenvironment in atherosclerotic neointima by exudinga diverse repertoire of inflammatory mediators (2).Continuous production of proinflammatory cytokinesand chemokines augment the influx and retention ofother inflammatory cells’ migration of vascular smoothmuscle cells from media to intima (2,3). All these eventslead to the proximal exacerbation of arterial damage.Restraining monocyte/macrophage recruitment intothe aortic wall may attenuate the risk of atherosclero-sis; hence, strategies to prevent monocyte infiltrationand differentiation comprise an attractive approachfor the treatment of atherosclerosis and other relatedvascular disorders.

Metformin, a widely used antidiabetic drug, hasbeneficial effects in reducing cardiovascular complica-tions in addition to glycemic control (4). The UK Pro-spective Diabetes Study demonstrated that metforminis associated with a significant decrease in the incidenceof myocardial infarction (5). Another clinical study

1Centre for Chemical Biology, Council of Scientific and Industrial Research(CSIR)–Indian Institute of Chemical Technology, Hyderabad, India2CSIR–Centre for Cellular and Molecular Biology, Hyderabad, India

Corresponding authors: Srigiridhar Kotamraju, giridhar@iict.res.in, and JeraldM. Kumar, mahesh73@ccmb.res.in.

Received 7 August 2014 and accepted 20 December 2014.

This article contains Supplementary Data online at http://diabetes.diabetesjournals.org/lookup/suppl/doi:10.2337/db14-1225/-/DC1.

© 2015 by the American Diabetes Association. Readers may use this article aslong as the work is properly cited, the use is educational and not for profit, andthe work is not altered.

See accompanying article, p. 1907.

2028 Diabetes Volume 64, June 2015

SIG

NALTRANSDUCTIO

N

reported that metformin in diabetic patients significantlyattenuates the progression of carotid artery intima-mediathickness, a known index of atherosclerotic progression(6). Earlier studies showed that the pleiotropic effects ofmetformin are partly mediated by the activation ofAMPK (7,8). AMPK functions as a fuel gauge, sensingthe changes in the energy status of the cell and, thus,playing a critical role in regulating systemic energybalance (9,10). AMPK is activated allosterically by anincrease in the intracellular AMP/ATP ratio (10).AMPK is also sensitive to the lipid status of a cell, andits activation is influenced by the availability of intracel-lular fat deposits (11). AMPK-mediated signaling eventsare downregulated by lipopolysaccharides (LPSs), freefatty acids, high-fat diets, and lipid infusion in macro-phages and endothelial cells (12,13). Activation of theAMPK signaling pathway suppresses proinflammatoryresponses and promotes macrophage polarization to ananti-inflammatory functional phenotype in macrophages(14). In addition, decreased AMPK activity with in-creased STAT3 in smooth muscle cells has been shownto promote receptor for advanced glycation end productsignaling–induced neointima formation in response toarterial injury (15).

Monocyte-to-macrophage differentiation is character-ized by the activation of various metabolic and inflam-matory signaling networks. Although the potential role ofAMPK in mediating anti-inflammatory effects againstvarious stimulations is known, its specific role in regulat-ing monocyte-to-macrophage differentiation still remainselusive. In this study, we investigated the effects ofmetformin and AICAR on the monocyte-to-macrophagedifferentiation process using a human monocytic leuke-mia (THP-1) cell line. AMPK activators attenuated themonocyte-to-macrophage differentiation and the pro-inflammatory signaling events associated during thedifferentiation process by a novel mechanism involv-ing AMPK-1a–mediated STAT3 regulation. Furthermore,metformin significantly attenuated angiotensin II (Ang-II)-induced plaque formation and aortic aneurysm (AA) inapolipoprotein E knockout (ApoE2/2) mice, possibly byimpairing monocyte recruitment and its differentiationinto macrophages in the arterial vessel wall.

RESEARCH DESIGN AND METHODS

Cell Culture and DifferentiationCell culture procedures are described in the SupplementaryData. For differentiation of monocytes to macrophages,THP-1 cells were seeded at a density of 2 3 105/mL andstimulated with phorbol myristate acetate (PMA).

Animal ExperimentsExperiments were conducted in 2-month-old male ApoE2/2

mice according to the guidelines formulated for the careand use of animals in scientific research (Indian Councilof Medical Research, India) at a CPCSEA (Committee forthe Purpose of Control and Supervision of Experimentson Animals)–registered animal facility. The experimental

protocols were approved by the Institutional Animal Eth-ical Committee at the Council of Scientific and IndustrialResearch (CSIR)–Indian Institute of Chemical Technology(IICT/CB/SK/20/12/2013/10). Animals were randomly di-vided into three groups of 12: 1) control, 2) Ang-II treat-ment, and 3) Ang-II + metformin treatment. Ang-II andmetformin treatment groups received Ang-II (Sigma) ata dose of 1.44 mg/kg/day as described previously (16–18) for 6 weeks through a subcutaneous route, whereasthe control group received normal saline. The metfor-min treatment group received the drug at a dose of 100mg/kg/day in normal drinking water. All animals were fednormal chow throughout the study. After 6 weeks, animalswere killed per standard protocol. The Supplementary Datadescribe the other methodologies adopted in this study.

Statistical AnalysisData are expressed as mean 6 SD. The significanceof differences between groups was examined using eitherStudent t test or one-way ANOVA as appropriate. P, 0.05was considered statistically significant.

RESULTS

Metformin Inhibits PMA-Induced Monocyte-to-Macrophage Differentiation in THP-1 MonocytesPMA-induced THP-1 monocyte differentiation is a well-accepted in vitro model for studying the monocyte-to-macrophage differentiation process (19). To study theeffect of metformin on PMA-induced monocyte-to-macrophage differentiation, THP-1 monocytes were treat-ed with PMA (100 nmol/L) for 48 h in the presence orabsence of metformin (0–2 mmol/L). The morphologicalobservations by phase contrast microscopy indicated thatmetformin dose dependently inhibited PMA-inducedmonocyte adherence (Fig. 1A). In addition, PMA-inducedincrease in autofluorescence, a distinguished feature ofmacrophage differentiation owing to the increased cellsize by flow cytometry, was significantly inhibited in thepresence of metformin (Fig. 1B). Under these conditions,the mitochondrial content, also a measure of monocytedifferentiation by MitoTracker staining, was greatly reducedupon metformin treatment (Fig. 1C). To rule out the pos-sibility that the inhibition of monocyte-to-macrophage dif-ferentiation by metformin was not due to its cytotoxiceffect, we measured cell viability. Metformin did not elicitany appreciable cell death in monocytes at the indicatedconcentrations (Fig. 1D). Monocyte-to-macrophage differ-entiation is associated with a loss or gain of expression ofan array of genes/proteins destined to perform specializedfunctions. Along these lines, we measured the transcriptand protein levels of SR-A1, LOX-1, and CD-36 duringmonocyte differentiation in the presence or absence ofmetformin. As expected, the levels of CD-36 and LOX-1,known macrophage markers, were increased with PMAstimulation. However, metformin treatment dose depen-dently reduced PMA-induced CD-36, LOX-1, and SR-A1levels (Fig. 1E and F), which are predominantly expressedin macrophages and are associated with the uptake of

diabetes.diabetesjournals.org Vasamsetti and Associates 2029

Figure 1—AMPK activators inhibit PMA-induced monocyte differentiation. THP-1 monocytes were pretreated with various concentrations (0–2mmol/L) of metformin for 2 h followed by stimulation with PMA (100 nmol/L). A: Phase contrast images indicate the inhibition of PMA-inducedmonocyte adherence by metformin treatment after 48 h. B: Same as A except autofluorescence was measured by FACS analysis. C: Same as Aexcept cell size wasmeasured byMitoTracker dye.D: The effect of metformin in the presence or absence of PMA on cell viability measured after 48 husing the trypan blue dye exclusion method. E and F: Expression of differentiation markers: Transcript levels (E) of the indicated genes weremeasured by RT-PCR after 24 h, and protein levels (F) were measured by Western blotting after 48 h. G: Phase contrast images indicating theinhibition of PMA-induced monocyte adherence by AICAR after 48 h. H and I: Expression of differentiation markers showing protein levels after 48 h(H) and transcript levels of the indicated genes after 24 h (I). Data are mean6 SD. *P< 0.05 vs. no treatment; #P< 0.05 vs. PMA control. MFI, meanfluorescence intensity.

2030 Inhibited STAT3 Activation and Atherosclerosis Diabetes Volume 64, June 2015

lipid molecules in these cells, enhancing the foamcell formation. Because metformin clearly abrogatedPMA-induced monocyte-to-macrophage differentiation,we studied whether AICAR, another known AMPK acti-vator, also imposes a similar effect. We found, similarto metformin, that AICAR significantly inhibited mono-cyte differentiation based on morphological observa-tions by phase contrast microscopy (Fig. 1G) andaltered the expression levels of CD-36, SR-A1, andLOX-1 in a dose-dependent fashion (Fig. 1H and I).These results confirm that AMPK activators metforminand AICAR not only inhibit monocyte-to-macrophagedifferentiation but also regulate the expression of scav-enger receptors.

Metformin Inhibits PMA-Induced Inflammation DuringMonocyte DifferentiationBesides monocyte-to-macrophage differentiation per se,inflammatory cascades evoked during the differentiationprocess are critically involved in the acceleration of theatherosclerotic process. To gain insight into the effect ofAMPK activators in regulating proinflammatory signalingpathways during monocyte-to-macrophage differentia-tion, we measured interleukin (IL)-1b, tumor necrosisfactor (TNF)-a, and MCP-1 levels in monocytes stimu-lated with PMA (100 nmol/L). All these cytokines startedto increase from 12 h and reached maximum concentra-tions by 48 h (Fig. 2A–C), suggesting that PMA-inducedmonocyte differentiation is accompanied by an increasedproduction of proinflammatory cytokines. Next, we mea-sured PMA-induced proinflammatory cytokines in thepresence of either metformin (0.5–2 mmol/L) or AICAR(0.5–2 mmol/L) and found that both dose dependentlyreduced PMA-induced IL-1b, TNF-a, and MCP-1 levels(Fig. 2D and E).

Macrophage-derived matrix metalloproteases (MMPs)are highly expressed in atherosclerotic plaques and havebeen implicated in the rupture of plaque structure.Because MMP-9 seems to be a key regulator of vascularcomplications, we studied the effect of PMA on MMP-9activity during monocyte-to-macrophage differentiation.PMA time dependently increased MMP-9 activity (Fig.2F). However, under these conditions, MMP-2 activitywas unchanged (Fig. 2F). Of note, PMA-induced MMP-9activity was dose dependently inhibited by both metfor-min and AICAR treatments (Fig. 2G and H). Next, toknow whether the inhibitory effects of metformin andAICAR on proinflammatory cytokine production andMMP-9 activity were due to their ability to repress thetranscriptional activation of these genes, mRNA levels byRT-PCR were measured after 24 h. Metformin and AICARdose dependently decreased the transcript levels of IL-1b,TNF-a, lipoprotein lipase (LPL), MMP-1, and MMP-9 (Fig.2I and J). All these data suggest that metformin andAICAR along with their ability to inhibit monocyte-to-macrophage differentiation also inhibit the inflammatorypathways associated with this process.

Inactivation of AMPK-1a Phosphorylation PromotesMonocyte-to-Macrophage Differentiation andInflammationBecause metformin and AICAR, the two known activa-tors of AMPK, inhibited monocyte differentiation aswell as proinflammatory cytokine production in PMA-treated monocytes, we investigated whether AMPKphosphorylation status was altered during PMA-inducedmonocyte differentiation. Monocytes were treated withPMA (0–200 nmol/L) for a 48 h. PMA dose dependentlyreduced AMPK-1a phosphorylation (Fig. 3A). The levelsof phospho-AMPK-1a started to decrease by 6 h uponPMA (100 nmol/L) stimulation and completely disap-peared by 24 h (Fig. 3B). Furthermore, to see whetherPKC activation was involved in regulating AMPK, wemeasured phospho-AMPK levels in PMA-stimulatedmonocytes in the presence or absence of calphostin C(0.5–4 mmol/L), a general PKC inhibitor, and found thatcalphostin C treatment could not restore the PMA-inducedinhibition of phospho-AMPK levels (Supplementary Fig.1A). Calphostin C also caused cell death beyond 2 mmol/L(Supplementary Fig. 1B). Additionally, calphostin C treat-ment did not have any effect on either PMA-inducedmonocyte-to-macrophage differentiation (data not shown)or inflammation (Supplementary Fig. 1C and D). Next, tosee whether metformin-induced inhibition of monocyte-to-macrophage differentiation was AMPK dependent, wemeasured phospho-AMPK-1a levels in cells treated withvarious concentrations of metformin. Metformin doseand time dependently increased AMPK-1a phosphoryla-tion in PMA-treated monocytes (Fig. 3C and D). Similarresults were obtained in AICAR-treated cells (Fig. 3E).These findings suggest that phosphorylation of AMPKplays a significant role during monocyte-to-macrophagedifferentiation and that the PKC-mediated pathway isnot involved in metformin-induced AMPK activation. Tofurther understand the role of AMPK phosphorylation inregulating the extent of monocyte differentiation, wetreated cells with compound C, an AMPK inhibitor, andfound that compound C treatment alone dose dependentlydecreased AMPK-1a phosphorylation in monocytes (Fig. 3E).However, compound C failed to promote monocyte-to-macrophage differentiation on its own in the absenceof PMA, as evidenced by morphological observations(Fig. 3F). In agreement with this result, compound Ctreatment alone, unlike PMA, failed to induce IL-1b,TNF-a, and MCP-1 production in monocytes (Fig. 3G).However, coincubation of compound C along with PMAexacerbated PMA-induced monocyte-to-macrophage dif-ferentiation and proinflammatory cytokine production(Fig. 3F and G). These results indicate that althoughAMPK-1a phosphorylation status plays a crucial role dur-ing the PMA-induced monocyte-to-macrophage differenti-ation process, alterations in AMPK-1a activity alone arenot sufficient to regulate the differentiation process. Thisunexpected result led us to search for other signaling path-ways responsible for monocyte differentiation that are

diabetes.diabetesjournals.org Vasamsetti and Associates 2031

Figure 2—Metformin and AICAR inhibit PMA-induced inflammation during monocyte differentiation. THP-1 monocytes were stimulatedwith PMA (100 nmol/L) for 6–48 h, and IL-1b (A), TNF-a (B), and MCP-1 (C) levels were measured in the conditioned medium by ELISA.D: Cells were pretreated with metformin (0–2 mmol/L) for 2 h followed by stimulation with PMA (100 nmol/L) for 48 h, and IL-1b, TNF-a, andMCP-1 were measured in the conditioned medium. E: Same as D except cells were pretreated with AICAR (0–2 mmol/L) for 2 h. F: Same asA–C except MMP-9 and MMP-2 activities were measured using gelatin zymography. G: Same as D except MMP-9 and MMP-2 activitieswere measured. H: Same as G except cells were pretreated with AICAR (0–2 mmol/L). I and J: Same as D except mRNA levels of indicatedinflammatory markers were measured after 24 h in cells pretreated with either metformin (I) or AICAR (J). Data are mean6 SD. *P< 0.05 vs.no treatment; #P < 0.05 vs. PMA control.

2032 Inhibited STAT3 Activation and Atherosclerosis Diabetes Volume 64, June 2015

dependent on or independent of AMPK phosphorylationinduced by PMA.

Inflammation and Differentiation Are IndependentEvents During Monocyte-to-MacrophageDifferentiation: c-Jun N-Terminal Kinase Plays a Rolein Regulating Inflammation but Not DifferentiationInduction of peroxisome proliferator–activated receptor(PPAR)-g expression and LPL activity is a common featureduring monocyte and adipocyte differentiation (20,21).Activation of c-Jun N-terminal kinase (JNK) plays a centralrole in IL-1b induction during monocyte differentiation (22).To study the possible involvement of JNK and other signalingpathways during monocyte-to-macrophage differentiation,

we evaluated the effect of pioglitazone (PPAR-g agonist),GW9882 (PPAR-g antagonist), orlistat (LPL inhibitor), p38inhibitor, SP600125 (JNK inhibitor), U-0126 (extracellularsignal–related kinase 1/2 inhibitor), and Bay-117085 (nuclearfactor-kB inhibitor), with concentrations ranging from 1 to40 mmol/L, on PMA-induced monocyte-to-macrophage dif-ferentiation and the associated inflammation during this pro-cess. Except for U-0126 (2.5 mmol/L) and Bay-117085(2.5 mmol/L), other inhibitors did not cause cell death atthese concentrations in monocytes (data not shown). Ad-ditionally, none of these compounds showed an apprecia-ble effect on monocyte-to-macrophage differentiation(data not shown). However, only SP600125 significantly

Figure 3—Inhibition of AMPK-1a phosphorylation promotes monocyte differentiation and inflammation. A: THP-1 monocytes were treatedwith increasing concentrations of PMA (0–200 nmol/L) for 48 h, and p-AMPK-1a and AMPK-1a protein levels were measured by Westernblotting. B: Same as A except cells were treated with PMA (100 nmol/L) for 6–48 h. C: THP-1 monocytes were pretreated with metformin(0–2 mmol/L) for 2 h followed by stimulation with PMA (100 nmol/L) for 24 h, and p-AMPK-1a and AMPK-1a protein levels were measured.D: Same as C except cells were pretreated with metformin (2 mmol/L) followed by stimulation with PMA (100 nmol/L) for 6–48 h. E: Same asC except cells were pretreated with AICAR (0–2 mmol/L). F: Monocytes were treated with compound C (0–5 mmol/L) alone for 24 h, andp-AMPK-1a and AMPK-1a levels were measured by Western blotting. G: Cells were pretreated with compound C (5 mmol/L) in thepresence or absence of PMA (100 nmol/L) for 48 h, and cell adherence was monitored by phase contrast microscopy. H: Cells werepretreated with compound C (0–5 mmol/L) in the presence or absence of PMA (100 nmol/L) for 48 h, and IL-1b, TNF-a, and MCP-1 levelswere measured in the conditioned medium by ELISA. Data are mean 6 SD. #P < 0.05 vs. PMA control. Thr, threonine.

diabetes.diabetesjournals.org Vasamsetti and Associates 2033

inhibited IL-1b, TNF-a, and MCP-1 concentrations (Fig. 4A)and MMP-9 activity (Fig. 4B) in monocytes stimulatedwith PMA. Along these lines, SP600125 dose dependentlyinhibited proinflammatory cytokine production in PMA-stimulated monocytes (Fig. 4C), suggesting that PMA pro-motes an inflammatory environment by activating theJNK signaling pathway in monocytes. Furthermore, itappears that PPAR-g, LPL, and JNK signaling pathwaysmay not be directly involved in monocyte-to-macrophagedifferentiation. To confirm this result, we measured thetranscript levels of LOX-1 and CD-36 (macrophagemarkers) along with proinflammatory cytokines in mono-cytes treated with JNK inhibitor in the presence of PMA.As expected and in agreement with the results shown inFig. 4A, JNK inhibitor dose dependently decreased mRNAlevels of IL-1b, TNF-a, and MCP-1 but not LOX-1 andCD-36 (Fig. 4D). These observations clearly indicate thatPMA-induced inflammation and differentiation are inde-pendent events. Because metformin inhibited proinflam-matory cytokine production in PMA-treated monocytes,we were interested in its effect on JNK phosphorylationstatus and its substrate c-Jun. For this, we treated

monocytes with metformin (0.5–2 mmol/L) in the pres-ence of PMA (100 nmol/L) and found that metformindose dependently reduced phosphorylation of both JNKand c-Jun (Fig. 4E). This finding suggests that JNKactivation is pivotal in mediating the inflammatory signalingcascade and that metformin exerts its anti-inflammatoryeffects at least in part by inhibiting the activation of theJNK signaling cascade during monocyte-to-macrophagedifferentiation.

Activation of STAT3 Is Involved in Monocyte-to-Macrophage Differentiation and InflammationSTAT3 plays a role in regulating cellular differentiation,and AMPK negatively regulates STAT3 phosphorylation(23,24). In the present study, we saw a downregulation ofAMPK phosphorylation during monocyte-to-macrophagedifferentiation, and although metformin and AICAR treat-ments restored it with concomitant inhibition of monocytedifferentiation, we wanted to know whether a cross-talkexists between AMPK and STAT3 during PMA-inducedmonocyte-to-macrophage differentiation. The resultsshow that PMA treatment alone induced STAT3 phos-phorylation as early as 6 h (Fig. 5A), and under the

Figure 4—Effect of PPAR-g modulators and MAPK inhibitors on PMA-induced inflammation and differentiation in monocytes. A and B:THP-1 monocytes were pretreated with 40 mmol/L pioglitazone, GW9662, orlistat, p38 inhibitor, or SP600125 for 2 h followed by stimu-lation with PMA (100 nmol/L) for 48 h, and IL-1b, TNF-a, MCP-1 levels were measured by ELISA (A) and MMP-9 and MMP-2 activities bygelatin zymography (B) in the conditioned media. C: Monocytes were pretreated with JNK inhibitor SP600125 (0–40 mmol/L) for 2 hfollowed by PMA (100 nmol/L) stimulation for 48 h, and IL-1b, TNF-a, MCP-1 levels were measured. D: Same as C except transcriptlevels of indicated genes were measured by RT-PCR after 24 h. E: Same as C except monocytes were pretreated with metformin (0–2mmol/L) for 24 h, and p-JNK, JNK, p-c-Jun, and c-Jun were measured by Western blotting. Data are mean 6 SD. #P < 0.05 vs. PMAcontrol. Ser, serine; Thr, threonine; Tyr, tyrosine.

2034 Inhibited STAT3 Activation and Atherosclerosis Diabetes Volume 64, June 2015

same conditions, metformin dose dependently inhibitedPMA-induced STAT3 phosphorylation (Fig. 5B). We thenstudied the effect of stattic (STAT3 inhibitor) on PMA-induced monocyte-to-macrophage differentiation. Stattic

dose dependently (5–20 mmol/L) inhibited the adherenceof PMA-treated monocytes (Fig. 5C), which is in line withthe reduction in mRNA levels of macrophage differentiationmarkers (Fig. 5D). These observations support that STAT3

Figure 5—PMA-induced monocyte differentiation alters STAT3 phosphorylation status: effect of metformin. A: THP-1 monocytes weretreated with PMA (100 nmol/L) for 6–24 h, and p-STAT3 and STAT3 were measured by Western blotting. B: Same as A except cells werepretreated with metformin (0–2 mmol/L) for 2 h followed by stimulation with PMA for 24 h. Monocytes were pretreated with stattic (0–20mmol/L) for 2 h followed by stimulation with PMA (100 nmol/L). C: Phase contrast images indicating the inhibition of PMA-inducedmonocyte adherence by stattic treatment after 48 h. D: Same as C except transcript levels of monocyte differentiation marker geneswere measured by RT-PCR after 24 h. E: Same as C except IL-1b, TNF-a, and MCP-1 levels were measured by ELISA. F: MMP-9 andMMP-2 activities were measured by gelatin zymography in the conditioned medium after 48 h. G: Same as C except transcript levels ofproinflammatory cytokine gene expression were measured by RT-PCR after 24 h. H: The effect of stattic in the presence or absence of PMAon cell viability measured after 48 h. I: Monocytes were treated with either SP600125 (40 mmol/L), stattic (20 mmol/L), or compound C (5mmol/L) in either the presence or the absence of PMA stimulation for 24 h, and p-AMPK-1a, AMPK-1a, p-JNK, JNK, p-STAT3, and STAT3were analyzed by Western blotting. Data are mean 6 SD. #P < 0.05 vs. PMA control. Thr, threonine; Tyr, tyrosine.

diabetes.diabetesjournals.org Vasamsetti and Associates 2035

phosphorylation promotes monocyte-to-macrophage differ-entiation. Next, to know whether STAT3 phosphorylationalso plays a role in PMA-induced inflammation, we mea-sured proinflammatory cytokine production and MMP-9activity in the presence of stattic. Stattic dose dependentlyinhibited the secretion of IL-1b, TNF-a, and MCP-1 intothe medium (Fig. 5E) along with a significant inhibition ofMMP-9 activity (Fig. 5F). Furthermore, stattic treatmentdecreased the mRNA levels of IL-1b, TNF-a, and MCP-1(Fig. 5G), suggesting that STAT3 phosphorylation promotestranscription of proinflammatory genes during monocyte-to-macrophage differentiation. The inhibitory effectof stattic on monocyte differentiation and inflammationwere not a result of its cytotoxic effects (Fig. 5H). To gainmore insight into the cross-talk among AMPK, STAT3, andJNK, we measured the phosphorylation statuses of theseproteins under various experimental conditions, usingtheir respective inhibitors in the presence or absence ofPMA as shown in Fig. 5I. First, treatment of cells withstattic inhibited JNK phosphorylation, whereas incuba-tion of cells with SP600125 failed to affect STAT3phosphorylation (Fig. 5I), suggesting that JNK is a down-stream target of STAT3. Second, to understand the effectof AMPK in regulating JNK and STAT3 activation, weexamined the phosphorylation statuses of these proteinsby inhibiting AMPK activation with compound C in boththe presence and the absence of PMA stimulation. We didnot notice any change in the phosphorylation status ofeither STAT3 or JNK with compound C treatment in theabsence of PMA stimulation. However, there was a signif-icant increase in their phosphorylation statuses withcotreatment of compound C and PMA (Fig. 5I). Fromthese results, it appears that although a causal reciprocalrelationship exists between AMPK and STAT3 phosphor-ylation (Fig. 5A and B), it may happen only under PMA-stimulated conditions, which in turn suggests that otheradditional factors may be regulated by PMA stimulation,enabling AMPK-dependent phosphorylation of STAT3 pro-tein. Additionally, we observed that neither SP600125 norstattic was able to rescue PMA-mediated inhibition ofp-AMPK-1a (Fig. 5I). Taken together, these observationsindicate that STAT3 phosphorylation plays a critical role inmediating monocyte-to-macrophage differentiation andinflammation.

Metformin Administration Attenuates the Incidence ofAng-II–Induced Atheromatous Plaque Formation andAA in ApoE2/2 MiceIt is reasonably well known that monocyte-to-macrophagedifferentiation is a prerequisite step in the developmentand progression of the atherosclerotic disease process (2).In this context, we studied the ability of metformin toinhibit monocyte-to-macrophage differentiation and in-flammation during Ang-II–induced atherogenesis in anApoE2/2 mouse model. Morphologically, thoracic and ab-dominal aorta of the Ang-II–treated group showed se-vere incidence of plaque extension, multiple numbersof micro/pseudoaneurysm formation, and maximal aortic

diameters (Fig. 6A–C) at the end of 6 weeks. However,the incidence of plaque extension and AA was significantlyless in the Ang-II + metformin–treated group similar tocontrol animals (Fig. 6A and B). We further analyzed thevascular remodeling in tissue sections of thoracic aortas.Hematoxylin-eosin (H-E) staining of the Ang-II–treatedgroup showed severe atherosclerotic lesions with boththick internal and external walls and intimal plaques(Fig. 6D–F). These changes were greatly reduced inAng-II + metformin–treated animals (Fig. 6D–F), whichwas further confirmed by Masson trichrome staining inwhich thick fibrous capsules comprising mature connec-tive tissue surrounding and in between atheroma wereobserved (Fig. 7A). Additionally, Van Gieson stainingindicated a ruptured medial layer lamella (light brown)with a dark brown nucleus in Ang-II–treated mice (Fig.7B). The collagen tissue in the atheroma, intimal, medial,and external regions appeared as red in the atheromatousregion of Ang-II–treated mice. However, in control andmetformin-treated groups, collagen tissue accumulationin the intimal region was not observed (Fig. 7A, B, E,and F). Along these lines, Ang-II treatment markedly ele-vated macrophage infiltration, as evidenced by Mac-3 im-munofluorescence and H-E staining of the atheromatousregion (Fig. 7C, D, G, and H). Mac-3 is a general marker ofmacrophage abundance often seen under inflammatoryconditions. Metformin treatment significantly inhibitedAng-II–induced macrophage accumulation comparable tocontrol mice (Fig. 7C, D, G, and H). This observation alsocoincides with the increased presence of the foam cellpopulation detected by Sudan black staining in the plaqueregion of Ang-II–treated animals (Fig. 8A and B). To fur-ther corroborate the ability of metformin to inhibitmonocyte-to-macrophage differentiation, we measuredCD-36 and LOX-1 protein levels in total lysates of aorta,and the results are consistent with cell culture studies,showing that metformin treatment significantly inhibitedAng-II–induced CD-36 and LOX-1 (Fig. 8C). Additionally,metformin treatment rescued Ang-II–mediated loss ofAMPK-1a phosphorylation (Fig. 8C). In line with this,metformin-treated mice showed decreased levels of serumproinflammatory cytokines (MCP-1 and TNF-a) comparedwith the Ang-II–treated group (Fig. 8D). On the otherhand, IL-10 (anti-inflammatory cytokine) levels were sig-nificantly increased in metformin-treated mice comparedwith Ang-II–treated and control mice (Fig. 8E). These invivo observations agreed with the ability of metformin toinhibit the proinflammatory cytokine production duringPMA-induced monocyte-to-macrophage differentiation inTHP-1 cells (Figs. 1 and 2). To understand the basis forthe increased incidence of AA in Ang-II–treated mice, wemeasured the transcript levels of MMP-1, -2, and -9 byRT-PCR and found that MMP-1 and -9 were significantlyelevated in Ang-II–treated mouse aortas (Fig. 8F). Metfor-min treatment significantly decreased Ang-II–inducedMMP levels (Fig. 8F). Finally, to extend the vasculopro-tective effects of metformin, we observed that metformin

2036 Inhibited STAT3 Activation and Atherosclerosis Diabetes Volume 64, June 2015

treatment significantly reduced the Ang-II–mediated in-crease in LDL and triglyceride levels while increasing serumHDL levels (Fig. 8G–I). Overall, these results demonstratethat metformin administration significantly attenuatesthe incidence of vascular complications, including plaqueformation and AA.

DISCUSSION

Atherosclerosis is a chronic immune-mediated inflammatorydisease of the arterial vessel wall characterized by thethickening of intima primarily due to monocyte recruitmentinto the subendothelial space (1–4). The current study showsthat metformin, a widely used antidiabetic drug and anAMPK activator, effectively inhibits PMA-induced monocyte-to-macrophage differentiation. Metformin also elicitspotent anti-inflammatory effects by inhibiting IL-1b,

TNF-a, and MCP-1 levels in vitro and in vivo, therebyemphasizing its apparent beneficial effect in the context ofatherogenesis. Similar results were observed with AICAR,a distinct AMPK activator. AMPK activity in the blood ispredominantly located in monocytes and lymphocytes asthe 1a isoform (25), and a reduction in AMPK activity isassociated with various pathophysiological effects (26).Vascular suppression of AMPK stimulates arterial deposi-tion of excess lipids, resulting in the development of ath-erosclerotic lesions (27). AMPK signaling in macrophages issignificantly downregulated by inflammatory stimuli andexternal free fatty acids (13,14). The results of this studyindicate that by activating AMPK-1a phosphorylation,metformin potently inhibits monocyte-to-macrophage dif-ferentiation. This effect of AMPK was counteracted inthe presence of AMPK inhibitor, but compound C failed

Figure 6—Metformin treatment inhibits Ang-II–induced AA and plaque formation in ApoE2/2 mice. A: Thoracic and abdominal aorticdiameters in control, Ang-II–, and Ang-II + metformin–treated groups. B: Percent AA incidence. C: Percent plaque incidence. D: Histo-pathological images of aorta stained with H-E. E: Quantitative analysis of plaque size. F: Quantitative analysis of intima-externa thickness.Data represent mean 6 SD from 3 different animals. *P < 0.05 vs. control. #P < 0.05 vs. Ang-II–treated group.

diabetes.diabetesjournals.org Vasamsetti and Associates 2037

Figure 7—Metformin treatment attenuates Ang-II–induced macrophage accumulation in ApoE2/2 mouse aorta. A and B: Histopathologicalimages of aorta stained with Masson trichrome and Van Gieson for analyzing fibrous and collagen tissue in the vessel wall. C: Confocalmicroscopic images representing Mac-3 immunofluorescence (green fluorescence represents positive staining for macrophages as in-dicated). D: Bright field microscopic images showing macrophage infiltration in the vessel wall. E and F: Quantitative analysis of fibrous andcollagen content in the plaque and externa regions shown in A and B. G: Quantitative analysis of macrophage density in the plaque andexterna regions shown in C. H: Quantitative analysis of macrophage population by counting large cells with a centrally located nucleus asmacrophages in various regions of the vessel wall shown in D. Data are mean6 SD from three different animals. *P< 0.05 vs. control. #P<0.05 vs. Ang-II–treated group.

2038 Inhibited STAT3 Activation and Atherosclerosis Diabetes Volume 64, June 2015

to cause monocyte-to-macrophage differentiation in theabsence of PMA. Inactivation of AMPK-1a triggers inflam-mation only in the presence of external stimuli, like LPS

(14). These findings suggest that inactivation of AMPKalone seems to be inadequate to promote either cellu-lar differentiation or inflammatory signaling pathways.

Figure 8—Metformin treatment attenuates Ang-II–induced foam cell formation and inflammation in ApoE2/2 mice. A: Histopathologicalimages of aorta were stained with Sudan black to detect foam cells (black spots represent foam cells as indicated). B: Quantitative analysisof A. C: Expression levels of p-AMPK-1a and macrophage cell markers CD-36 and LOX-1 proteins in the whole aortic tissue lysate byWestern blotting. D and E: The levels of serum TNF-a, MCP-1, and IL-10 measured by ELISA. F: Quantitative measurement of MMPs in thewhole aortic tissue by RT-PCR. G, H, and I: Serum triglyceride, LDL, and HDL levels. Data are mean 6 SD from at least three independentanimals. *P < 0.05 vs. control; #P < 0.05 vs. Ang-II–treated group. Thr, threonine.

diabetes.diabetesjournals.org Vasamsetti and Associates 2039

Nevertheless, inactivation of AMPK accelerates boththese events in the presence of an external stimulus(e.g., LPS, PMA, macrophage colony-stimulating factor).Of note, JNK inhibition effectively attenuated only PMA-induced inflammation, but not PMA-induced monocyte-to-macrophage differentiation. This result agrees witha finding showing that JNK signaling plays a crucial rolein inducing IL-1b in monocytes (22). Metformin-activatedAMPK inhibited LPS-induced proinflammatory cytokinelevels by inhibiting JNK1 phosphorylation in macro-phages (28). To this end, the present data show thatthe reversal of PMA-induced inhibition of AMPK phos-phorylation by metformin is accompanied by a decreasein JNK1 phosphorylation, indicating that AMPK activa-tors can modulate the inflammatory pathways duringmonocyte-to-macrophage differentiation through targetingJNK signaling and emphasizing that inflammation anddifferentiation are independent events. The results alsoindicate that by negatively regulating STAT3 phosphory-lation through increased activation of AMPK, metformininhibits monocyte-to-macrophage differentiation. Activa-tion of STAT3 signaling is essential for neointima forma-tion in vivo in response to carotid artery angioplasty(15). Metformin- and AICAR-activated AMPK was shownto inhibit proinflammatory gene expression in human livercells by repressing IL-6–stimulated STAT3 phosphoryla-tion (24). The present work shows that by decreasingthe PMA-induced STAT3 phosphorylation, metformin sig-nificantly inhibited monocyte-to-macrophage differentiationand proinflammatory cytokine production.

In support of the observed favorable effects ofmetformin, we also found a complete regression ofatherosclerotic plaque formation in the aorta of Ang-II +metformin–treated mice compared with Ang-II treat-ment alone. Gross pathological investigations revealedextensive plaque reduction in the entire aorta andpseudoaneurysm in abdominal and thoracic aorta ofmice treated with Ang-II + metformin compared withAng-II treatment alone. The aorta of metformin-treatedmice also showed a significant reduction in the recruit-ment of inflammatory macrophages as evidenced byMac-3 staining compared with mice treated with Ang-IIalone. Additionally, in agreement with the effects of met-formin in THP-1 monocytes, Ang-II + metformin–treatedmice showed a significant downregulation of CD-36 andLOX-1 compared with mice treated with Ang-II alone.Apart from the regression of plaque formation, metfor-min treatment also resulted in the reduction of Ang-II–induced AA. This appears to be an important observa-tion, and the molecular basis for reduced AA with met-formin treatment along with changes in blood pressureunder these conditions need to be investigated. One of theother important observations of this study is that metfor-min treatment significantly inhibited an Ang-II–mediatedincrease in LDL and triglyceride levels, which are knownto be associated with proatherosclerotic effects. Moreover,metformin treatment significantly increased HDL levels.

Metformin has been shown to improve the lipid profile byreducing LDL cholesterol as well as triglyceride levels intype 2 diabetic patients (29,30).

In conclusion, we show that AMPK activators caneffectively inhibit monocyte-to-macrophage differentiationand associated inflammatory pathways. It is likely that theantiatherosclerotic effects of metformin are in part medi-ated by perturbing monocyte-to-macrophage differentiationduring Ang-II–mediated atherosclerosis in ApoE2/2 mice.

Funding. This work was supported by grants from the Department of Scienceand Technology, Department of Biotechnology, and CSIR, India, under 12th Five YearPlan projects SMiLE and EpiHeD. S.B.V. and A.K.K. acknowledge CSIR and S.Ka.acknowledges the Indian Council of Medical Research, New Delhi, India, for theaward of research fellowships.Duality of Interest. No potential conflicts of interest relevant to this articlewere reported.Author Contributions. S.B.V. contributed to the experimental design, dataanalysis, and writing of the manuscript. S.Ka. and A.K.K. contributed to the animalexperiments. A.R.T. contributed to the histology studies. J.M.K. contributed to theexperimental design, provision of animals for in vivo experiments, data analysis, andwriting of the manuscript. S.Ko. contributed to the experimental design, provision ofreagents and other material required for performing both in vitro and in vivo experi-ments, data analysis, and writing of the manuscript. S.Ko. is the guarantor of thiswork and, as such, had full access to all the data in the study and takes re-sponsibility for the integrity of the data and the accuracy of the data analysis.

References1. Hansson GK, Libby P. The immune response in atherosclerosis: a double-edged sword. Nat Rev Immunol 2006;6:508–5192. Moore KJ, Tabas I. Macrophages in the pathogenesis of atherosclerosis.Cell 2011;145:341–3553. Li AC, Glass CK. The macrophage foam cell as a target for therapeuticintervention. Nat Med 2002;8:1235–12424. Abbasi F, Chu JW, McLaughlin T, Lamendola C, Leary ET, Reaven GM.Effect of metformin treatment on multiple cardiovascular disease risk factors inpatients with type 2 diabetes mellitus. Metabolism 2004;53:159–1645. UK Prospective Diabetes Study (UKPDS) Group. Effect of intensive blood-glucose control with metformin on complications in overweight patients with type2 diabetes (UKPDS 34). Lancet 1998;352:854–8656. Matsumoto K, Sera Y, Abe Y, Tominaga T, Yeki Y, Miyake S. Metforminattenuates progression of carotid arterial wall thickness in patients with type 2diabetes. Diabetes Res Clin Pract 2004;64:225–2287. Zhou G, Myers R, Li Y, et al. Role of AMP-activated protein kinase inmechanism of metformin action. J Clin Invest 2001;108:1167–11748. Zang M, Zuccollo A, Hou X, et al. AMP-activated protein kinase is requiredfor the lipid-lowering effect of metformin in insulin-resistant human HepG2 cells.J Biol Chem 2004;279:47898–479059. Kahn BB, Alquier T, Carling D, Hardie DG. AMP-activated protein kinase:ancient energy gauge provides clues to modern understanding of metabolism.Cell Metab 2005;1:15–2510. Hardie DG, Ross FA, Hawley SA. AMPK: a nutrient and energy sensor thatmaintains energy homeostasis. Nat Rev Mol Cell Biol 2012;13:251–26211. Lee WJ, Lee IK, Kim HS, et al. Alpha-lipoic acid prevents endothelial dys-function in obese rats via activation of AMP-activated protein kinase. ArteriosclerThromb Vasc Biol 2005;25:2488–249412. Wu Y, Song P, Xu J, Zhang M, Zou MH. Activation of protein phosphatase 2A bypalmitate inhibits AMP-activated protein kinase. J Biol Chem 2007;282:9777–978813. Yang Z, Kahn BB, Shi H, Xue BZ. Macrophage alpha1 AMP-activated proteinkinase (alpha1AMPK) antagonizes fatty acid-induced inflammation throughSIRT1. J Biol Chem 2010;285:19051–19059

2040 Inhibited STAT3 Activation and Atherosclerosis Diabetes Volume 64, June 2015

14. Sag D, Carling D, Stout RD, Suttles J. Adenosine 59-monophosphate-activatedprotein kinase promotes macrophage polarization to an anti-inflammatory func-tional phenotype. J Immunol 2008;181:8633–864115. Yu W, Liu-Bryan R, Stevens S, Damanahalli JK, Terkeltaub R. RAGE sig-naling mediates post-injury arterial neointima formation by suppression of liverkinase B1 and AMPK activity. Atherosclerosis 2012;222:417–42516. Daugherty A, Manning MW, Cassis LA. Angiotensin II promotes athero-sclerotic lesions and aneurysms in apolipoprotein E-deficient mice. J Clin Invest2000;105:1605–161217. Gavrila D, Li WG, McCormick ML, et al. Vitamin E inhibits abdominal aorticaneurysm formation in angiotensin II-infused apolipoprotein E-deficient mice.Arterioscler Thromb Vasc Biol 2005;25:1671–167718. Gopal K, Nagarajan P, Jedy J, et al. b-Carotene attenuates angiotensinII-induced aortic aneurysm by alleviating macrophage recruitment in Apoe-/- mice.PLoS One 2013;8:e6709819. Sakamoto H, Aikawa M, Hill CC, et al. Biomechanical strain induces classa scavenger receptor expression in human monocyte/macrophages and THP-1cells: a potential mechanism of increased atherosclerosis in hypertension. Cir-culation 2001;104:109–11420. Tontonoz P, Nagy L, Alvarez JG, Thomazy VA, Evans RM. PPARgamma promotesmonocyte/macrophage differentiation and uptake of oxidized LDL. Cell 1998;93:241–25221. Namgaladze D, Kemmerer M, von Knethen A, Brüne B. AICAR inhibitsPPARg during monocyte differentiation to attenuate inflammatory responses toatherogenic lipids. Cardiovasc Res 2013;98:479–48722. Tiwari RL, Singh V, Singh A, Barthwal MK. IL-1R-associated kinase-1mediates protein kinase Cd-induced IL-1b production in monocytes. J Immunol2011;187:2632–2645

23. McLemore ML, Grewal S, Liu F, et al. STAT-3 activation is required fornormal G-CSF-dependent proliferation and granulocytic differentiation. Immunity2001;14:193–20424. Nerstedt A, Johansson A, Andersson CX, Cansby E, Smith U, MahlapuuM. AMP-activated protein kinase inhibits IL-6-stimulated inflammatory re-sponse in human liver cells by suppressing phosphorylation of signaltransducer and activator of transcription 3 (STAT3). Diabetologia 2010;53:2406–241625. Lindholm CR, Ertel RL, Bauwens JD, Schmuck EG, Mulligan JD, Saupe KW.A high-fat diet decreases AMPK activity in multiple tissues in the absence ofhyperglycemia or systemic inflammation in rats. J Physiol Biochem 2013;69:165–17526. Steinberg GR, Kemp BE. AMPK in health and disease. Physiol Rev 2009;89:1025–107827. Li Y, Xu S, Jiang B, Cohen RA, Zang M. Activation of sterol regulatory el-ement binding protein and NLRP3 inflammasome in atherosclerotic lesion de-velopment in diabetic pigs. PLoS One 2013;8:e6753228. Woo SL, Xu H, Li H, et al. Metformin ameliorates hepatic steatosis andinflammation without altering adipose phenotype in diet-induced obesity. PLoSOne 2014;9:e9111129. Mullugeta Y, Chawla R, Kebede T, Worku Y. Dyslipidemia associatedwith poor glycemic control in type 2 diabetes mellitus and the protectiveeffect of metformin supplementation. Indian J Clin Biochem 2012;27:363–36930. Musi N, Hirshman MF, Nygren J, et al. Metformin increases AMP-activatedprotein kinase activity in skeletal muscle of subjects with type 2 diabetes. Di-abetes 2002;51:2074–2081

diabetes.diabetesjournals.org Vasamsetti and Associates 2041