INTRODUCTIONSepsis and septic shock are life-threat-

ening medical conditions, characterizedby uncontrolled or overwhelming sys-temic inflammatory response to an infec-tion and accounts for more than 750,000cases annually in the United States (1).Despite significant advances and exten-sive therapeutic approaches, sepsis re-mains a major health concern and leadingcause of high mortality worldwide (2–4).Human and animal studies have demon-strated that progressive multiple organfailure is the most common cause of death

following sepsis, with the lungs usuallyrepresenting the first organ to fail (5). Pre-disposition of the immune response hasbeen shown to be critical in the develop-ment of an exaggerated inflammatory re-action and ensuing acute lung injury(ALI) (6). The clinical pathology of ALI in-cludes increased vascular permeability, in-flammation, oxidative stress, apoptosis,pulmonary edema and the accumulationof activated neutrophil in lung tissue andeventually cell death (7–11).

AMP-activated protein kinase (AMPK)is a serine/threonine kinase that functions

as an energy sensing enzyme and is acti-vated in response to an increase of theAMP/ATP ratio during hypoxia, glucosedeprivation, heat shock or reduction inmitochondrial oxidative phosphorylation(12,13). AMPK is a heterotrimeric complexconsisting of three subunits, α catalyticsubunit and regulatory β and γ subunits(14). Two isoforms have been identifiedfor both α subunit (α1 and α2) and β sub-unit (β1 and β2), and three isoforms havebeen reported for the γ subunit (γ1, γ2 andγ3) (14). Most recently, several reports in-dicate that the function of AMPK is notonly restricted to the maintenance of en-ergy metabolism, but also to coordinationof several housekeeping mechanisms andinvolvement in modulating oxidativestress and inflammatory mediators(15–18). An adenosine monophosphateanalog, 5-aminoimidazole-4-carboxamideribonucleotide (AICAR), has been used asa pharmacological activator to upregulateAMPK activity (19). Administering

M O L M E D 2 1 : 6 3 7 - 6 4 4 , 2 0 1 5 | M U L C H A N D A N I E T A L . | 6 3 7

Stimulation of Brain AMP-Activated Protein Kinase AttenuatesInflammation and Acute Lung Injury in Sepsis

Nikhil Mulchandani,1 Weng-Lang Yang,1,2 Mohammad Moshahid Khan,2 Fangming Zhang,2

Philippe Marambaud,3 Jeffrey Nicastro,1 Gene F Coppa,1 and Ping Wang1,2

1Department of Surgery, Hofstra North Shore-LIJ School of Medicine, Manhasset, New York, United States of America; 2Center forTranslational Research and 3Litwin-Zucker Research Center for the Study of Alzheimer’s Disease, The Feinstein Institute for MedicalResearch, Manhasset, New York, United States of America

Sepsis and septic shock are enormous public health problems with astronomical financial repercussions on health systemsworldwide. The central nervous system (CNS) is closely intertwined in the septic process but the underlying mechanism is still ob-scure. AMP-activated protein kinase (AMPK) is a ubiquitous energy sensor enzyme and plays a key role in regulation of energy ho-meostasis and cell survival. In this study, we hypothesized that activation of AMPK in the brain would attenuate inflammatory re-sponses in sepsis, particularly in the lungs. Adult C57BL/6 male mice were treated with 5-aminoimidazole-4-carboxamideribonucleotide (AICAR, 20 ng), an AMPK activator, or vehicle (normal saline) by intracerebroventricular (ICV) injection, followed bycecal ligation and puncture (CLP) at 30 min post-ICV. The septic mice treated with AICAR exhibited elevated phosphorylation ofAMPKα in the brain along with reduced serum levels of aspartate aminotransferase, tumor necrosis factor-α (TNF-α), interleukin-1β(IL-1β) and interleukin-6 (IL-6), compared with the vehicle. Similarly, the expressions of TNF-α, IL-1β, keratinocyte-derived chemokineand macrophage inflammatory protein-2 as well as myeloperoxidase activity in the lungs of AICAR-treated mice were significantlyreduced. Moreover, histological findings in the lungs showed improvement of morphologic features and reduction of apoptosiswith AICAR treatment. We further found that the beneficial effects of AICAR on septic mice were diminished in AMPKα2 deficientmice, showing that AMPK mediates these effects. In conclusion, our findings reveal a new functional role of activating AMPK in theCNS to attenuate inflammatory responses and acute lung injury in sepsis.Online address: http://www.molmed.orgdoi: 10.2119/molmed.2015.00179

Address correspondence to Ping Wang, The Feinstein Institute for Medical Research, 350

Community Drive, Manhasset, NY 11030. Phone: 516-562-3411; Fax: 516-562-1022; E-mail:

pwang@nshs.edu.

Submitted July 30, 2015; Accepted for publication July 30, 2015; Published Online

(www.molmed.org) July 30, 2015.

AICAR has been shown to inhibit inflam-mation and to alleviate organ injury in ro-dents (13,20–22).

There is growing evidence that the cen-tral nervous system (CNS) plays an im-portant and functionally relevant role inregulating the inflammatory response.For example, the vagus nerve and cholin-ergic pathway have been demonstratedto be involved in modulating proinflam-matory cytokine-induced injury locallyand systemically in infectious diseases(23–26). However, the molecular mecha-nisms of the CNS in regulating systemicinflammation are still not well character-ized. The hypothalamus, a key region inthe CNS, controls several peripheralphysiological activities through a broadnetwork of hormonal and neuronal com-munication (27). AMPKα is expressed inhypothalamic neurons as a fuel sensorand counteracts energy deficits in thebrain (28). More specifically, the AMPKα1is primarily cytoplasmic, whereasAMPKα2 is predominantly nuclear andplays a role in transcriptional regulation(29). In addition, the α2 catalytic subunitis highly expressed in neurons in com-parison with α1 (30). Thus, we hypothe-sized that activation of AMPK in the CNScould attenuate inflammatory responsesand reduce ALI in sepsis.

In this study, we examined the mecha-nistic and beneficial effect of activatingcentrally located AMPK in septic miceinduced by cecal ligation and puncture(CLP). AICAR administration to thebrain through intracerebroventricular(ICV) injection attenuated the peripheralinflammation, as evidenced by decreasedproduction of cytokines and chemokines,and reduced lung injury in septic mice.We also provided the evidence that theprotective effect of AICAR is mediatedthough the presence of AMPKα2 in thebrain by using AMPKα2 deficient mice.

MATERIALS AND METHODS

AnimalsMale age-matched wild-type C57 BL/6

mice (25-30 g) (Taconic) and AMPKα2knockout (Prkaa2–/–) mice were obtained

from Benoit Viollet (INSERM) and weremaintained at the Feinstein Institute forMedical Research (14). They werehoused in a temperature-controlled roomon a 12-h light–dark cycle and fed a stan-dard mouse diet. All experiments wereperformed in accordance with the Na-tional Institutes of Health guidelines foruse of experimental animals (Guide forthe Care and Use of Laboratory Animals, 8thedition, 2011), and this study was ap-proved by the Institutional Animal Careand Use Committee of the Feinstein In-stitute for Medical Research.

Intracerebroventricular (ICV) InjectionStereotactic and surgical materials

were obtained from Stoelting and Hamil-ton. Animals were lightly anesthetizedwith inhalational isoflurane and the headof the animal was secured on the stereo-tactic apparatus. A thermal pad was usedto maintain core body temperature dur-ing the procedure. The head of the ani-mal was shaved and cleansed with 10%povidone-iodine wash. A midline inci-sion was made, extending from the eyesto the back of the cranium to expose thebregma. Either 20 ng AICAR (Sigma-Aldrich) or 2 μL vehicle (normal saline)was injected into the lateral ventriclearea by using predetermined coordinates(AP, 0.34 mm from bregma; lateral, 1.0 m;and vertical, 2.2 mm).

Animal Model of SepsisThirty minutes after ICV injection, the

animal was removed from the stereotacticapparatus and placed in supine position.The ventral abdomen was shaved andcleansed with 10% povidone–iodine washand cecal ligation and puncture (CLP) wasperformed as previously described (31).One- to 2-cm midline incision was per-formed to allow exposure of the cecumand tightly ligated ~1.0 cm from the tipwith a 3-0 silk suture. A through doublepuncture of the cecum was performedusing a 22-gauge needle. A small amountof feces was expressed from the perforatedsites and returned to the peritoneal cavity.The laparotomy site was then closed with6-0 silk suture. Sham-operated animals

underwent the same procedure with theexception that the cecum was neither lig-ated nor punctured. The CLP animals sub-cutaneously received 1 mL isotonic nor-mal saline immediately after the surgery.

Western BlottingHypothalamus and surrounding tissue

were lysed and homogenized in 300 μLlysis buffer (10 mmol/L Tris-HCL pH 7.5,120 mmol/L NaCl, 1% NP-40, 1% sodiumdeoxycholate and 1% Triton X-100) con-taining protease and phosphatase inhibitorcocktails (Roche) using a sonic dismembra-tor on ice. Samples were centrifuged at14,000g for 20 min at 4°C, and the super-natant were collected. Following measure-ment of sample protein concentration byPierce BCA protein assay kit (PierceBiotechnology), 60 μg samples were sepa-rated on 4% to 12% Bis-Tris gradient gelsand transferred to nitrocellulose mem-branes. Membranes were blocked by incu-bation with 0.1% casein followed by incu-bation with primary antibody againstp-AMPKα, t-AMPKα (Cell Signaling), orβ-actin (Sigma-Aldrich). After washing,membrane was incubated with appropriatefluorescent secondary antibodies. Bandswere detected using the Odyssey FC Dual-Mode Imaging system 2800 (LI-COR).

Measurements of Serum LiverEnzymes and Cytokines

Whole-blood samples were centrifugedat 4,000g for 12 min to collect serum, whichwas then stored at –80°C before use. Theactivity of aspartate aminotransferase(AST) was determined by a commercialassay kit from Pointe Scientific. Serumtumor necrosis factor-α (TNF-α), inter-leukin 1β (IL-1β) and interleukin-6 (IL-6)levels were determined by an enzyme-linked immunosorbent assay kit specificfor mouse (BD Biosciences). The assayswere carried out according to the instruc-tions provided by the manufacturer.

Real-Time Reverse Transcriptase–Polymerase Chain Reaction (RT-PCR)Analysis

Total RNA was extracted from lung tis-sue by TRIzol reagent (Invitrogen) and

6 3 8 | M U L C H A N D A N I E T A L . | M O L M E D 2 1 : 6 3 7 - 6 4 4 , 2 0 1 5

A C T I V A T E B R A I N A M P K T O A T T E N U A T E S E P S I S

reverse-transcribed in to cDNA by usingmurine leukemia virus reverse transcrip-tase (RT) (Applied Biosystems). A PCRreaction was carried out in 25 μL finalvolume containing 0.08 μmol of each forward and reverse primer, 5 μL cDNA,6.5 μL H2O and 12.5 μL SYBR Green PCRMaster Mix (Applied Biosystems). Am-plification was conducted in an AppliedBiosystems 7300 real-time PCR machineunder the thermal profile of 50°C for 2 min and 95°C for 10 min, followed by45 cycles of 95°C for 15 s and 60°C for 1 min. The level of mouse β-actin mRNAwas used for normalization and eachspecific mRNA was conducted in dupli-cate. Relative expression of mRNA wascalculated by the 2–ΔΔCt method, and re-sults were expressed as fold change incomparison with controls. The sequencesof primers for this study are listed inTable 1.

Myeloperoxidase (MPO) ActivityAssay

Lung tissues were sonicated in 50 mmol/L potassium phosphate buffercontaining 0.5% hexadecyl trimethyl ammonium bromide. After the centrifuga-tion, the supernatant was diluted in reac-tion solution containing o-dianisidine hy-drochloride and H2O2. The rate of changein optical density (OD) for 1 min wasmeasured at 460 nm to calculate MPO ac-tivity as described previously (32).

Histological Evaluation and TerminalDeoxynucleotidyl Transferase dUTPNick-End Labeling (TUNEL) Staining ofthe Lungs

Lung tissues were taken from theupper and lower lobes 20 h followingCLP and stored in 10% formalin beforebeing fixed in paraffin. Tissues were then

sectioned to 4-μm cuts and stained withhematoxylin-eosin. For TUNEL staining,fluorescence staining was performedusing a commercially available in situCell Death Detection Kit (Roche). Theassay was conducted according to themanufacturer’s instructions. The nucleuswas stained with propidium iodide. Re-sults were expressed as the averagenumber of TUNEL-positive staining cellsper 10 microscopic fields.

Statistical AnalysisAll data are expressed as a mean ± SE

(n = 6–8/group) and compared by one-way analysis of variance (ANOVA) andthe Student-Newman-Keuls (SNK) test.Differences in values were consideredsignificant if P < 0.05.

RESULTS

AICAR Activates AMPKα in the Brain ofSeptic Mice

We first verified the effect of centrallyadministered AICAR on the activation ofAMPKα. At 20 h after CLP, the hypothal-amus sections of the mice were har-vested and subjected to Western blotanalysis. The phosphorylated level ofAMPKα in the brain remained un-changed after CLP, while its level in-creased by 2.18-fold in the AICAR treat-ment group in comparison with thevehicle treated septic mice (Figure 1).This result indicated that ICV-injection ofIACAR effectively stimulated AMPK ac-tivity in the brain of septic mice.

AICAR Attenuates Organ Injury andSystemic Inflammation in Septic Mice

We then examined the effect of AICARtreatment on the tissue damage and in-flammatory response in sepsis. At 20 h

after CLP, the serum level of AST, anorgan injury marker, increased by 7.4-foldin the vehicle, compared with the sham(Figure 2A). By contrast, AICAR treat-ment significantly decreased the ASTlevel by 36.7% as compared with the vehi-cle group (Figure 2A). Similarly, we alsofound significantly increased serum levelsof proinflammatory cytokines, TNF-α(25.3-fold), IL-1β (24.3-fold) and IL-6 (6.8-fold) in the vehicle group as comparedwith the respective sham groups. Impor-tantly, AICAR treatment significantly de-creased the level of TNF-α by 70.5%, IL-1βby 90.3%, and IL-6 by 42.2% as comparedwith the vehicle (Figures 2B–D).

AICAR Ameliorates Lung Damageand Apoptosis in Septic Mice

The lungs are very vulnerable in sepsis(33). We examined the lung tissue at 20 hafter CLP with hematoxylin and eosinstaining. As shown in Figure 3, CLP in-duced severe lung injury shown with sep-

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

M O L M E D 2 1 : 6 3 7 - 6 4 4 , 2 0 1 5 | M U L C H A N D A N I E T A L . | 6 3 9

Table 1. A list of primer sequences used in this study.

Name GenBank Forward Reverse

TNF-α X_02611 AGACCCTCACACTCAGATCATCTTC TTGCTACGACGTGGGCTACAIL-1β NM_008361 CAGGATGAGGACATGAGCACC CTCTGCAGACTCAAACTCCACKC NM_008176 GCTGGGATTCACCTCAAGAA ACAGGTGCCATCAGAGCAGTMIP-2 NM_009140 CCCTGGTTCAGAAAATCATCCA GCTCCTCCTTTCCAGGTCAGTβ-Actin NM_007393 CGTGAAAAGATGACCCAGATCA TGGTACGACCAGAGGCATACAG

Figure 1. Phosphorylation of AMPK in thebrain. The hypothalamus and surroundingtissue of sham, vehicle- and AICAR-treatedmice were harvested at 20 h after CLP. Thetissue was lysed and subjected to Westernblotting against phosphorylated (p-AMPKα) and total AMPK (t-AMPKα). His-togram shows mean densitometric analy-sis of bands after normalizing with t-AMPKand β-actin. Expression in the sham groupis designated as 1 for comparison. Dataare expressed as mean ± SE (n = 6–8 pergroup). *P < 0.05 versus sham and #P < 0.05versus vehicle.

tal thickening, proteinaceous exudate,pulmonary edema and enhanced inflam-matory infiltrates as compared with thesham. By contrast, the severity of lungdamage was ameliorated by AICAR treat-ment, shown with a better integrity of mi-croscopic structure as compared with thevehicle-treated animals (Figure 3).

Apoptosis plays a major role in thepathogenesis of sepsis-induced organ in-jury (34). We applied TUNEL assay to ex-amine the apoptosis in the lungs by im-munofluorescence. In the vehicle group,the TUNEL-positive cells (green fluores-cence) were well detected, while theywere barely observed in the sham (Fig-ure 4A). On the other hand, the numberof TUNEL-positive cells in the lung tis-sues of the AICAR-treated mice was re-duced significantly (63.3%) as comparedwith the vehicle group (Figure 4B).

AICAR Suppresses Lung Inflammationand Neutrophil Infiltration in SepticMice

Excessive production of proinflamma-tory cytokines is one of the importantcontributing factors for the lung injury

(35). Lung tissue was harvested at 20 hafter CLP and subjected to RT-PCR andELISA analysis. The mRNA level of TNF-αwas increased by 53.1-fold in vehicle-treated animal as compared with thesham. Parallel to mRNA expression, wefound increased protein levels of TNF-α

in vehicle-treated animal as comparedwith the sham (Figures 5A, B). AICARtreatments significantly suppressedmRNA and protein levels of TNF-α by72.5% and 44.2%, respectively, comparedwith the vehicle (Figures 5A, B). Simi-larly, the mRNA and protein levels of IL-1β were elevated significantly in the ve-hicle group as compared the sham, whiletheir levels were reduced by 76.9% and47.4%, respectively, compared with thevehicle (Figures 5C, D).

Chemokines keratinocyte derived-chemokine (KC) and macrophage inflam-matory protein-2 (MIP-2) both have beenshown to participate in the pathogenesisof lung injury (36,37). The mRNA levelsof KC and MIP-2 increased by 748- and1415-fold, respectively, in the vehicle ascompared with the sham (Figures 6A, B).By contrast, a significant reduction of KCmRNA by 76.2% and MIP-2 mRNA by64.2% was observed in lung tissues ofAICAR-treated mice (Figures 6A, B). Wefurther measured the myeloperoxidaseactivity (MPO) in the lungs to evaluatethe neutrophil infiltration. The MPO ac-tivity increased by 3.1-fold in vehicle-treated group as compared with thesham, while AICAR treatment decreasedthe MPO activity by 42.4% as comparedwith the vehicle (Figure 6C).

6 4 0 | M U L C H A N D A N I E T A L . | M O L M E D 2 1 : 6 3 7 - 6 4 4 , 2 0 1 5

A C T I V A T E B R A I N A M P K T O A T T E N U A T E S E P S I S

Figure 2. Alterations in serum levels of organ injury marker and proinflammatory cytokines.Blood of sham, vehicle- and AICAR-treated mice were harvested at 20 h after CLP formeasuring (A) AST, (B) TNF-α, (C) IL-1β, and (D) IL-6. Data are expressed as mean ± SE (n =6–8 per group). *P < 0.05 versus sham and #P < 0.05 versus vehicle.

Figure 3. Morphological changes in the lungs. The lung tissues of sham, vehicle- andAICAR-treated mice were harvested at 20 h after CLP and subjected to histological anal-ysis with hematoxylin–eosin staining. Representative images of lung histology from n = 6–8per group. Original magnification 100× (top panels); 200× (bottom panels).

AICAR Activity Is Mediated byAMPKα2 in the Brain of Septic Mice

To investigate the molecular mechanismby which AICAR treatment to the CNS at-tenuated organ injury and inhibited in-flammation, we undertook a genetic ap-proach. AMPKα2 is highly expressed inthe neurons of the brain. Thus, we per-formed the CLP in AMPKα2 knockout(KO) mice after ICV injection of either ve-hicle or AICAR as described in the wild-type (WT) mice. We first measured theserum AST level and observed its eleva-tion in the vehicle group of AMPKα2 KOmice, which was similar to that in the WTmice (Figure 7A). AICAR treatment onlyreduced AST levels by 13.8% without sta-tistical significance as compared with thevehicle (Figure 7A). Similarly, there was anincrease in serum levels of TNF-α and IL-1β in septic AMPKα2 KO mice, however,AICAR treatment did not significantlychange their levels as compared with thevehicle (Figures 7B, C).

We also examined the mRNA levels ofproinflammatory cytokines andchemokines in the lungs by RT-PCR.There was an increase in mRNA levels ofTNF-α, IL-1β, KC and MIP-2 in the lungsof vehicle-treated AMPKα2 KO mice ascompared with their respective sham(Figures 8A–D). However, AICAR-treatedAMPKα2 KO septic mice did not showany significant changes in the expressionlevels of these cytokines and chemokinesas compared with the vehicle-treatedAMPKα2 KO septic mice (Figures 8A–D).Taken together, these results suggest thatthe protective effects of administered ICVwith AICAR in the WT mice may mainlymediate through AMPKα2-dependentmechanisms in the brain.

DISCUSSIONSepsis is the leading cause of death

among critically ill patients and is themost common risk factor for ALI. Un-governed inflammation, characterized bycytokine storm and subsequent neutrophilsequestration is an underlying componentof sepsis associated organ failure (38).Thus, elucidation of novel molecular andcellular pathways that influence the

pathogenesis of sepsis and its associatedcomplication could provide greater insightinto the mechanisms of disease pathology.

The significance of the brain AMPKstimulation in regulation of inflammationin sepsis is not known. We have used acombination of pharmacological and ge-netic approaches for the study. First, wedemonstrate that ICV injection of AICARmarkedly increases the phosphorylationof AMPK in the brain of CLP-inducedseptic mice. Such AICAR administrationreduced organ injury and inhibited sys-temic inflammation. In more specific or-gans, both the severity of damage and theinduction of apoptosis in the lungs ofseptic mice are attenuated by AICARtreatment, which is associated with a re-duction of cytokine and chemokine pro-duction as well as neutrophil infiltration.On the other hand, AICAR-mediated pro-tective effect on lowering organ injury

and inflammation in septic WT mice isdiminished in septic AMPKα2 KO mice,suggesting that brain AMPKα2 may be akey player to mediate AICAR activity.

Activation of AMPK by systemic ad-ministration of AICAR has been shown tosuppress inflammatory responses in otherstudies. Zhao and coworkers demonstratethat AICAR reduces the proinflammatoryactivity of neutrophils and decreases theseverity of ALI through AMPK activation,linking to cellular responses to metabolicstress (13). Escobar et al. demonstrate thatactivation of AMPK by AICAR amelio-rates liver and renal injury in septic mice,which is associated with a decrease in cir-culating cytokines and tissue inflamma-tion (39). In this study, we provide newevidence of activating the CNS in regulat-ing the systemic responses. By directly in-jecting AICAR into the brain, we have ob-served a significant reduction of serum

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

M O L M E D 2 1 : 6 3 7 - 6 4 4 , 2 0 1 5 | M U L C H A N D A N I E T A L . | 6 4 1

Figure 4. Apoptosis in the lungs after CLP. The lung tissues of sham, vehicle- and AICAR-treated mice were harvested at 20 h after CLP and subjected to TUNEL assay. (A) Repre-sentative images of TUNEL staining (green fluorescent) and nuclear counterstaining (redfluorescent) of lung sections. Original magnification 100×. Scale bar, 50 μm. (B) The num-ber of apoptotic cells in the lung quantified from TUNEL staining. Data are expressed asmean ± SE (n = 6–8 per group). *P < 0.05 versus sham and #P < 0.05 versus vehicle.

levels of proinflammatory cytokines TNF-α, IL-1β and IL-6 in septic mice.

Another important aspect of our studyis to observe an improvement of lungmorphology and inhibition of lung apo-ptosis in septic mice with central AICARinjection. As known, lung tissue damageis observed in 90% of patients dyingfrom sepsis and its associated complica-tion (40). The histopathological featuresof lung injury in septic mice demon-strated here are resembled in the clinicalcondition of patients with sepsis (41).Other studies have indicated that apo-ptosis plays an important role on pro-gression of sepsis (42,43). Oberholzer andcolleagues suggested that targeting sig-naling pathways that lead to apoptosiswould represent a new therapeutic targetfor the patient with sepsis or other re-lated clinical conditions (34). By systemicadministration, AICAR has been shownto decrease the apoptosis in differentcells associated with various disease con-ditions (44–46). Rossi and Lord describethat AICAR treatment decreases the neu-trophil apoptosis through the AMPK ac-

tivation, suggesting an important role ofAMPK in normal function of neutrophils(45). Similarly, Kim et al. demonstratethat AICAR reduces the palmitate-in-duced apoptosis in neuron cells by acti-vation of AMPK (46).

In addition to inhibiting systemic in-flammation, central AICAR administra-tion also decreases local inflammation inthe lungs. We have demonstrated a sig-nificant decrease in expression of cy-

tokines TNF-α.and IL-1β in the lungswith AICAR treatment. Moreover, centralAICAR administration also effectively in-hibits the expression of KC and MIP-2 inthe lungs. It has been reported that ele-vated expression of KC and MIP-2 po-tently stimulate neutrophil influx in thelungs, while suppression of thesechemokines markedly inhibits neutrophilsequestration in the lungs (47). Consis-tent with this mechanism, we have de-tected decreased MPO activity, a markerfor neutrophil infiltration (48). Consid-ered together, our findings suggest thatactivation of the brain AMPK by AICARfor a protective milieu to the lung in sep-sis may confer inhibition of the produc-tion of inflammatory mediators.

We then further identify the specifictarget in the CNS for AICAR’s activity inregulating systemic inflammation. Whencomparing the serum levels of AST, TNF-α and IL-1β in septic AMPKα2 KO micebetween vehicle and AICAR ICV injec-tions, the beneficial effects of AICAR ob-served in the septic WT mice are dimin-ished in the AMPKα2 KO mice.Moreover, the AICAR’s effect on inhibit-ing the expression of cytokines andchemokines in the lungs of the septic WTmice disappeared in the septic AMPKα2KO mice. These results clearly indicatethat AMPKα2 in the brain is majorly re-sponsible for central AICAR stimulation.As it has been indicated that the CNS canmodulate the systemic inflammation andcytokines production by: 1) activation of

6 4 2 | M U L C H A N D A N I E T A L . | M O L M E D 2 1 : 6 3 7 - 6 4 4 , 2 0 1 5

A C T I V A T E B R A I N A M P K T O A T T E N U A T E S E P S I S

Figure 5. Alterations in expression of proinflammatory cytokines in the lungs. The lung tissuesof sham, vehicle- and AICAR-treated mice were harvested at 20 h after CLP for measuringthe mRNA levels of (A) TNF-α and (C) IL-1β by real-time RT-PCR as well as protein levels of(B) TNF-α and (D) IL-1β by ELISA. The mRNA expression in the sham group is designated as 1for comparison. Data are expressed as mean ± SE (n = 6–8 per group). *P < 0.05 versussham and #P < 0.05 versus vehicle.

Figure 6. Alterations in expression of chemokines and myeloperoxidase activity in thelungs. The lung tissues of sham, vehicle- and AICAR-treated mice were harvested at 20 hafter CLP for measuring the mRNA levels of (A) KC and (B) MIP-2 by real-time RT-PCR. (C)Lung myeloperoxidase (MPO) activity was determined spectrophotometrically. The mRNAexpression in the sham group is designated as 1 for comparison. Data are expressed asmean ± SE (n = 6–8 per group). *P < 0.05 versus sham and #P < 0.05 versus vehicle.

the sympathetic nervous system and re-lease of norepinephrine which target theimmune cells expressing adrenergic re-ceptors (49,50); 2) stimulation of thevagus nerve-mediated cholinergic antiin-flammatory pathway (23,51,52); and 3)activation of the hypothalamic-pituitary-adrenal (HPA) axis for glucocorticoids se-cretion in response to adrenocorticotropichormone and immune suppression(53,54). However, which pathway(s)serves as downstream-of-brain AMPK ac-tivation for modulating the systemic and

peripheral tissue inflammation followingsepsis needs further investigation.

By the same token, Giri et al. presentthe evidence that activation of AMPKα2by AICAR attenuates the LPS-mediatedinduction of proinflammatory mediatorsin rat primary astrocytes and microgliacells through inhibiting NF-κB andC/EBP transcription factors (22). A studyby Bernik et al. demonstrates that intrac-erebral administration of a tetravalentguanylhydrazone molecule that inhibitedTNF-α production increases efferent

vagus nerve activity and inhibits inflam-mation outside the CNS (55). Moreover, ithas been shown that activation of hypo-thalamic AMPK can prevent LPS-inducedhypoglycemia in mice liver, suggestingcentrally located AMPK in the regulationof peripheral physiological activities (56).

CONCLUSIONOur results suggest that centrally ad-

ministered AICAR protects against sep-sis-induced pulmonary injury and apo-ptosis, and that this protection isassociated with a decrease in circulatingproinflammatory cytokines. Furthermore,hypothalamic AMPKα2 mediates AICARactivity to regulate systemic inflamma-tory response in sepsis. Thus, further un-derstanding the interaction between theCNS and systemic activity will provideanother direction for sepsis treatment.

ACKNOWLEDGMENTSSupported in part by National Insti-

tutes of Health (NIH) grants GM057468and GM053008 (to P Wang). The authorsthank Benoit Viollet (INSERM, InstitutCochin, Paris, France) for generouslyproviding the AMPKα2 knockout mice.

DISCLOSURE The authors declare that they have no

competing interests as defined by Molecu-lar Medicine, or other interests that mightbe perceived to influence the results anddiscussion reported in this paper.

REFERENCES1. Gaieski DF, Edwards JM, Kallan MJ, Carr BG.

(2013) Benchmarking the incidence and mortalityof severe sepsis in the United States. Crit. CareMed. 41:1167–74.

2. Khan MM, Yang WL, Wang P. (2015) Endoplasmicreticulum stress in sepsis. Shock. 44:294–304.

3. Hu Z, et al. (2015) Ursolic acid improves survivaland attenuates lung injury in septic rats induced bycecal ligation and puncture. J. Surg. Res. 194:528–36.

4. Sharma A, Matsuo S, Yang WL, Wang Z, Wang P.(2014) Receptor-interacting protein kinase 3 defi-ciency inhibits immune cell infiltration and atten-uates organ injury in sepsis. Crit. Care. 18:R142.

5. Neumann B, et al. (1999) Mechanisms of acute in-flammatory lung injury induced by abdominalsepsis. Int. Immunol. 11:217–27.

6. Ayala A, et al. (2002) Shock-induced neutrophilmediated priming for acute lung injury in mice:

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

M O L M E D 2 1 : 6 3 7 - 6 4 4 , 2 0 1 5 | M U L C H A N D A N I E T A L . | 6 4 3

Figure 7. Alterations in serum levels of organ injury marker and proinflammatory cytokinesin AMPKα2 KO mice. Blood of sham, vehicle- and AICAR-treated AMPKα2 KO mice wereharvested at 20 h after CLP for measuring (A) AST, (B) TNF-α and (C) IL-1β. Data are ex-pressed as mean ± SE (n = 6-8 per group). P = nonsignificant (NS).

Figure 8. Alterations in expression of proinflammatory cytokines and chemokines in thelungs of AMPKα2 KO mice. The lung tissues of sham, vehicle- and AICAR-treated AMPKα2KO mice were harvested at 20 h after CLP for measuring (A) TNF-α, (B) IL-1β, (C) KC and(D) MIP-2 by real time RT-PCR. The expression in the sham group is designated as 1 forcomparison. Data are expressed as mean ± SE (n = 6-8 per group). P = nonsignificant (NS).

divergent effects of TLR-4 and TLR-4/FasL defi-ciency. Am. J. Pathol. 161:2283–94.

7. Lomas-Neira JL, Chung CS, Wesche DE, Perl M,Ayala A. (2005) In vivo gene silencing (withsiRNA) of pulmonary expression of MIP-2 ver-sus KC results in divergent effects on hemor-rhage-induced, neutrophil-mediated septic acutelung injury. J. Leukoc. Biol. 77:846–53.

8. Filgueiras LR, Capelozzi VL, Martins JO, JancarS. (2014) Sepsis-induced lung inflammation ismodulated by insulin. BMC Pulm. Med. 14:177.

9. Mannam P, et al. (2014) MKK3 regulates mito-chondrial biogenesis and mitophagy in sepsis- induced lung injury. Am. J. Physiol. Lung Cell.Mol. Physiol. 306:L604–19.

10. Aziz M, Jacob A, Yang WL, Matsuda A, Wang P.(2013) Current trends in inflammatory and im-munomodulatory mediators in sepsis. J. Leukoc.Biol. 93:329–42.

11. Botha AJ, et al. (1995) Early neutrophil sequestra-tion after injury: a pathogenic mechanism formultiple organ failure. J. Trauma. 39:411–7.

12. Hardie DG. (2003) Minireview: the AMP-activatedprotein kinase cascade: the key sensor of cellularenergy status. Endocrinology. 144:5179–83.

13. Zhao X, et al. (2008) Activation of AMPK attenu-ates neutrophil proinflammatory activity and de-creases the severity of acute lung injury. Am. J.Physiol. Lung Cell. Mol. Physiol. 295:L497–504.

14. Viollet B, et al. (2003) The AMP-activated proteinkinase alpha2 catalytic subunit controls whole-body insulin sensitivity. J. Clin. Invest. 111:91–8.

15. Mihaylova MM, Shaw RJ. (2011) The AMPK sig-nalling pathway coordinates cell growth, au-tophagy and metabolism. Nat. Cell Biol. 13:1016–23.

16. Li XN, et al. (2009) Activation of the AMPK-FOXO3 pathway reduces fatty acid-induced in-crease in intracellular reactive oxygen species byupregulating thioredoxin. Diabetes. 58:2246–57.

17. Dolinsky VW, Dyck JR. (2006) Role of AMP- activatedprotein kinase in healthy and diseased hearts. Am. J.Physiol. Heart Circ. Physiol. 291: H2557–69.

18. Salminen A, Hyttinen JM, Kaarniranta K. (2011)AMP-activated protein kinase inhibits NF-κB sig-naling and inflammation: impact on healthspanand lifespan. J. Mol. Med. (Berl). 89:667–76.

19. Sullivan JE, et al. (1994) Inhibition of lipolysis andlipogenesis in isolated rat adipocytes with AICAR,a cell-permeable activator of AMP-activated pro-tein kinase. FEBS Lett. 353:33–6.

20. Bai A, et al. (2010) AMPK agonist downregulatesinnate and adaptive immune responses in TNBS-induced murine acute and relapsing colitis.Biochem. Pharmacol. 80:1708–17.

21. Hoogendijk AJ, Pinhancos SS, van der Poll T,Wieland CW. (2013) AMP-activated protein kinaseactivation by 5-aminoimidazole-4-carbox-amide-1-beta-D-ribofuranoside (AICAR) reduces lipotei-choic acid-induced lung inflammation. J. Biol.Chem. 288:7047–52.

22. Giri S, et al. (2004) 5-aminoimidazole-4-carboxamide-1-beta-4-ribofuranoside inhibits proinflammatory re-sponse in glial cells: a possible role of AMP-activatedprotein kinase. J. Neurosci. 24:479–87.

23. Tracey KJ. (2007) Physiology and immunology ofthe cholinergic antiinflammatory pathway. J. Clin.Invest. 117:289–96.

24. Huston JM, et al. (2007) Transcutaneous vagusnerve stimulation reduces serum high mobilitygroup box 1 levels and improves survival inmurine sepsis. Crit. Care Med. 35:2762–8.

25. Boeckxstaens G. (2013) The clinical importance ofthe anti-inflammatory vagovagal reflex. Handb.Clin. Neurol. 117:119–34.

26. Park J, Kang JW, Lee SM. (2013) Activation of thecholinergic anti-inflammatory pathway by nico-tine attenuates hepatic ischemia/reperfusion in-jury via heme oxygenase-1 induction. Eur. J.Pharmacol. 707:61–70.

27. Cai D, Liu T. (2011) Hypothalamic inflammation:a double-edged sword to nutritional diseases.Ann. N. Y. Acad. Sci. 1243:E1–39.

28. Ronnett GV, Aja S. (2008) AMP-activated proteinkinase in the brain. Int. J. Obes. (Lond). 32 Suppl 4:S42–8.

29. Turnley AM, et al. (1999) Cellular distributionand developmental expression of AMP-activatedprotein kinase isoforms in mouse central nervoussystem. J. Neurochem. 72:1707–16.

30. Amato S, Man HY. (2011) Bioenergy sensing inthe brain: the role of AMP-activated protein ki-nase in neuronal metabolism, development andneurological diseases. Cell Cycle. 10:3452–60.

31. Giangola MD, et al. (2013) Growth arrest-specificprotein 6 attenuates neutrophil migration andacute lung injury in sepsis. Shock. 40:485–91.

32. Hirano Y, et al. (2015) Neutralization of osteopon-tin attenuates neutrophil migration in sepsis- induced acute lung injury. Crit. Care. 19:782.

33. Hudson LD. (1995) New therapies for ARDS.Chest. 108:79S–91S.

34. Oberholzer C, Oberholzer A, Clare-Salzler M,Moldawer LL. (2001) Apoptosis in sepsis: a new tar-get for therapeutic exploration. FASEB J. 15:879–92.

35. Liu D, Zienkiewicz J, DiGiandomenico A, Haw-iger J. (2009) Suppression of acute lung inflam-mation by intracellular peptide delivery of a nu-clear import inhibitor. Mol. Ther. 17:796–802.

36. Schmal H, Shanley TP, Jones ML, Friedl HP,Ward PA. (1996) Role for macrophage inflamma-tory protein-2 in lipopolysaccharide-inducedlung injury in rats. J. Immunol. 156:1963–72.

37. Lomas JL, et al. (2003) Differential effects of macro -phage inflammatory chemokine-2 and keratinocyte-derived chemokine on hemorrhage-induced neu-trophil priming for lung inflammation: assessmentby adoptive cells transfer in mice. Shock. 19:358–65.

38. Chong DL, Sriskandan S. (2011) Pro-inflammatorymechanisms in sepsis. Contrib. Microbiol. 17:86–107.

39. Escobar DA, et al. (2015) Adenosine monophos-phate-activated protein kinase activation protectsagainst sepsis-induced organ injury and inflam-mation. J. Surg. Res. 194:262–72.

40. Torgersen C, et al. (2009) Macroscopic post-mortem findings in 235 surgical intensive carepatients with sepsis. Anesth. Analg. 108:1841–7.

41. Weiss YG, et al. (2001) Adenoviral vector trans-fection into the pulmonary epithelium after cecal

ligation and puncture in rats. Anesthesiology.95:974–82.

42. Aschkenasy G, Bromberg Z, Raj N, DeutschmanCS, Weiss YG. (2011) Enhanced Hsp70 expressionprotects against acute lung injury by modulatingapoptotic pathways. PLoS One. 6:e26956.

43. Chopra M, Reuben JS, Sharma AC. (2009) Acutelung injury: apoptosis and signaling mecha-nisms. Exp. Biol. Med. (Maywood). 234:361–71.

44. Kim JE, et al. (2008) AMPK activator, AICAR, in-hibits palmitate-induced apoptosis in osteoblast.Bone. 43:394–404.

45. Rossi A, Lord JM. (2013) Adiponectin inhibitsneutrophil apoptosis via activation of AMP ki-nase, PKB and ERK 1/2 MAP kinase. Apoptosis.18:1469–80.

46. Kim J, Park YJ, Jang Y, Kwon YH. (2011) AMPKactivation inhibits apoptosis and tau hyperphos-phorylation mediated by palmitate in SH-SY5Ycells. Brain Res. 1418:42–51.

47. Shanley TP, et al. (1997) Requirement for C-X-Cchemokines (macrophage inflammatory protein-2 and cytokine-induced neutrophil chemoattrac-tant) in IgG immune complex-induced lung in-jury. J. Immunol. 158:3439–48.

48. Schmekel B, et al. (1990) Myeloperoxidase inhuman lung lavage. I. A marker of local neu-trophil activity. Inflammation. 14:447–54.

49. Bellinger DL, et al. (2008) Sympathetic modula-tion of immunity: relevance to disease. Cell Im-munol. 252:27–56.

50. Cervi AL, Lukewich MK, Lomax AE. (2014) Neu-ral regulation of gastrointestinal inflammation:role of the sympathetic nervous system. Auton.Neurosci. 182:83–8.

51. Huston JM. (2012) The vagus nerve and the in-flammatory reflex: wandering on a new treat-ment paradigm for systemic inflammation andsepsis. Surg. Infect. (Larchmt). 13:187–93.

52. Ji H, et al. (2014) Central cholinergic activation ofa vagus nerve-to-spleen circuit alleviates experi-mental colitis. Mucosal Immunol. 7:335–47.

53. Smith SM, Vale WW. (2006) The role of the hypothalamic-pituitary-adrenal axis in neuroen-docrine responses to stress. Dialogues Clin. Neu-rosci. 8:383–95.

54. Silverman MN, Pearce BD, Biron CA, Miller AH.(2005) Immune modulation of the hypothalamic-pituitary-adrenal (HPA) axis during viral infec-tion. Viral Immunol. 18:41–78.

55. Bernik TR, et al. (2002) Pharmacological stimula-tion of the cholinergic antiinflammatory path-way. J. Exp. Med. 195:781–8.

56. Santos GA, et al. (2013) Hypothalamic AMPK ac-tivation blocks lipopolysaccharide inhibition ofglucose production in mice liver. Mol. Cell. En-docrinol. 381:88–96.

6 4 4 | M U L C H A N D A N I E T A L . | M O L M E D 2 1 : 6 3 7 - 6 4 4 , 2 0 1 5

A C T I V A T E B R A I N A M P K T O A T T E N U A T E S E P S I S

Cite this article as: Mulchandani N, et al. (2015)Stimulation of brain AMP-activated protein kinaseattenuates inflammation and acute lung injury insepsis. Mol. Med. 21:637–44.