AMP-activated Protein Kinase Activation by 5-Aminoimidazole-4-carbox-amide-1-β-d-ribofuranoside (AICAR) Reduces Lipoteichoic Acid-induced Lung Inflammation

Background: AMPK is a highly conserved energy homeostasis-regulating kinase. Results: Activation of AMPK by AICAR in vitro reduced cytokine production in alveolar macrophage cell line and in vivo reduced LTA-induced neutrophil influx, protein leak and cytokine/chemokine levels. Conclusion: AMPK activation inhibits LTA-induced lung inflammation in mice. Significance: AICAR reduces LTA inflammation. Adenosine monophosphate-activated protein (AMP)-activated kinase (AMPK) is a highly conserved kinase that plays a key role in energy homeostasis. Activation of AMPK was shown to reduce inflammation in response to lipolysaccharide in vitro and in vivo. 5-Aminoimidazole-4-carbox-amide-1-β-d-ribofuranoside (AICAR) is intracellularly converted to the AMP analog ZMP, which activates AMPK. Lipoteichoic acid (LTA) is a major component of the cell wall of Gram-positive bacteria that can trigger inflammatory responses. In contrast to lipopolysaccharide, little is known on the effects of AMPK activation in LTA-triggered innate immune responses. Here, we studied the potency of AMPK activation to reduce LTA-induced inflammation in vitro and in lungs in vivo. Activation of AMPK in vitro reduced cytokine production in the alveolar macrophage cell line MH-S. In vivo, AMPK activation reduced LTA-induced neutrophil influx, as well as protein leak and cytokine/chemokine levels in the bronchoalveolar space. In conclusion, AMPK activation inhibits LTA-induced lung inflammation in mice.


Adenosine monophosphate-activated protein (AMP)-activated kinase (AMPK) is a highly conserved kinase that plays a key role in energy homeostasis. Activation of AMPK was shown to reduce inflammation in response to lipolysaccharide in vitro and in vivo. 5-Aminoimidazole-4-carbox-amide-1-␤-D-ribofuranoside (AICAR) is intracellularly converted to the AMP analog ZMP, which activates AMPK. Lipoteichoic acid (LTA) is a major component of the cell wall of Gram-positive bacteria that can trigger inflammatory responses. In contrast to lipopolysaccharide, little is known on the effects of AMPK activation in LTA-triggered innate immune responses. Here, we studied the potency of AMPK activation to reduce LTA-induced inflammation in vitro and in lungs in vivo. Activation of AMPK in vitro reduced cytokine production in the alveolar macrophage cell line MH-S. In vivo, AMPK activation reduced LTA-induced neutrophil influx, as well as protein leak and cytokine/chemokine levels in the bronchoalveolar space. In conclusion, AMPK activation inhibits LTA-induced lung inflammation in mice.
Gram-positive bacteria are a frequent cause of pneumonia, among which Staphylococcus aureus represents a serious and emerging threat (1). Pneumonia is a leading cause of mortality in the United States (2). The abundant Gram-positive cell wall component lipoteichoic acid (LTA) 2 is the predominant driving force of the host inflammatory response to this type of bacteria (3)(4)(5).
Adenosine monophosphate-activated protein (AMP)-activated kinase (AMPK) is a highly conserved kinase that classically is known for its key role in energy homeostasis. AMPK can be activated by AMP, serine/threonine kinase 11 (LKB1), and calmodulin-dependent protein kinase kinase (CaMKK) (6,7). Activated AMPK has a strong influence on metabolic processes by virtue of its capacity to induce glucose uptake, glycolysis, fatty acid oxidation, and mitochondrial biogenesis and to inhibit fatty acid/cholesterol synthesis, gluconeogenesis, and glycogen and protein synthesis (6). Apart from these metabolic effector functions, AMPK signaling was recently shown to influence the inflammatory response: activation of AMPK was reported to have anti-inflammatory properties both in vitro and in vivo in response to lipopolysaccharide (LPS)-induced lung inflammation (8). Two well known small molecular kinase activators can modulate AMPK signaling: metformin and 5-aminoimidazole-4-carbox-amide-1-␤-D-ribofuranoside (AICAR) (9). Metformin, a well known drug used in patients suffering from type 2 diabetes, activates AMPK by shifting the AMP:ATP ratio. AICAR is converted intracellularly to ZMP, an AMP analog, thus activating AMPK (9).
Here, we studied the potency of AMPK activation by AICAR to reduce inflammation in vitro and in vivo in a model of LTAinduced lung injury. Activation of AMPK in vitro reduced cytokine production in an alveolar macrophage cell line independent of mTOR signaling. In vivo, we validated AMPK activation through enhanced phosphorylation of acetyl-CoA carboxylase (ACC). This increased AMPK activity was accompanied with reduced cell influx and inflammatory mediator release in the pulmonary compartment. With this study we demonstrated the potency of AMPK activation to diminish inflammatory responses in LTA-induced lung inflammation in vivo.

EXPERIMENTAL PROCEDURES
Cell Line Experiments-The effect of AICAR on cytokine responses of resident macrophages and lung epithelium was tested as follows: 1 ϫ 10 5 MH-S (alveolar macrophage cell line; American Type Culture Collection) cells were seeded in a 48-well plate (10 5 cells/well) (Millipore). After 24 h of culture, cells were stimulated with 10 g/ml LTA (purified from S. aureus; endotoxin level: Ͻ1.25 enzyme units/mg; Invivogen). Simultaneously, cells were treated with 1 mM AICAR, 100 nM rapamycin, 1 mM AICAR ϩ 100 nM rapamycin or vehicle (0.3% dimethyl sulfoxide/PBS). At 6 and 24 h, supernatant was harvested for ELISA.
Mice-For all experiments female C57BL/6 mice (aged 10 -11 weeks) were purchased from Charles River. The Animal Care and Use Committee of the University of Amsterdam approved all experiments.
Induction of Lung Inflammation and AICAR Administration-Acute lung inflammation was induced as described previously (4,10). Briefly, mice were anesthetized with isoflurane (Upjohn), and 100 g of LTA (Invivogen) diluted in 50 l of sterile saline was instilled intranasally. 500 mg/kg AICAR in 200 l of saline or 200 l of saline (vehicle) was administered intraperitoneally at the start of the experiment. After 6 and 24 h mice were anesthetized with Domitor (Pfizer Animal Health Care; active ingredient medetomidine) and Nimatek (Eurovet Animal Health, Bladel, The Netherlands; active ingredient ketamine) and sacrificed by cardiac puncture followed by cervical dislocation.
Bronchoalveolar Lavage (BAL)-Through a midline incision the trachea and lungs were exposed; the right lung was isolated from the airways via a suture. The trachea was cannulated with a 22G Abbocath-T catheter (Abbott) and the left lung was instilled with two times 0.4 ml of sterile PBS. The fluid was retrieved and weighed, and total cell counts were determined with a Coulter cell counter (Beckman Coulter). Differential cell counts were determined by counting 100 cells on Giemsastained cytospin preparations. BAL was centrifuged at 1500 rpm for 10 min at 4°C. Supernatant was stored at Ϫ20°C until assays were performed, and the remaining cell pellet was used for Western blot analysis.
Western Blotting-Samples for Western blotting were boiled at 95°C for 5 min in Laemmli buffer and loaded onto SDSpolyacrylamide gels. After electrophoresis the content of the gels was transferred onto Immobilon-PVDF membranes (Millipore). The membranes were blocked in 5% BSA (Roche Applied Science) in TBS-T at room temperature for 60 min.
Statistical Analysis-Data are expressed as mean Ϯ S.E. In vitro analysis was performed by ANOVA and Bonferroni's multiple comparison test post tests. For in vivo data, two sample comparisons were performed by Mann-Whitney U tests using Prism version 5.01 (GraphPad Software). Comparisons between multiple groups were done using the Kruskall-Wallis test. Overall significant individual groups were assessed by Mann-Whitney U tests. p Ͻ 0.05 was considered to be statistically significant.

RESULTS
AICAR Reduces Cytokine Production in Vitro, Independent of mTOR-AMPK activation through AICAR pretreatment was previously shown to reduce TNF-␣ production in LPS-stimulated RAW264.7 murine macrophages (11). We tested the antiinflammatory property of AICAR in the presence or absence of rapamycin in response to LTA-stimulated MH-S (alveolar macrophages) cells. Cells were stimulated with 10 g/ml LTA, and cytokine production was measured after 6 and 24 h.
In MH-S cells treated with AICAR, TNF-␣, and IL-6 production were strongly reduced after both 6 and 24 h of stimulation (p Ͻ 0.01, Fig. 1, A-D). Treatment with mTOR antagonist rapamycin did not alter cytokine levels compared with untreated cells. The combination of AICAR and rapamycin resulted in an effect similar to that of AICAR-only treatment.
To assess the state of the AMPK/mTOR pathway we determined phosphorylation states of ACC, AMPK, p70/S6K, and mTOR in the MH-S cell lysates. AICAR treatment strongly enhanced phosphorylation of ACC and slightly enhanced AMPK phosphorylation (Fig. 1E) whereas the phosphorylation of mTOR or p70/S6K was not altered. As expected, rapamycin reduced phosphorylation of mTOR and p70/S6K; however, ACC and AMPK phosphorylation was unchanged. The effect on Western blotting was more evident after 6 h than after 24 h. Cell viability was not altered by AICAR or rapamycin treatment (data not shown).
AICAR Enhances ACC Phosphorylation in Vivo-Next we set out to assess the inflammatory effects of treatment with AICAR in LTA-induced pulmonary inflammation. Inflammation was induced by intranasal instillation of 100 g of LTA (4, 10). Simultaneously, 500 mg/kg AICAR in 200 l of saline or 200 l of saline (vehicle) was administered intraperitoneally. Cells present in the airways were obtained through BAL. In the cell fraction of BAL fluid, AMPK activation was determined by measuring the phosphorylation level of ACC, a downstream substrate of AMPK (8). ACIAR treatment enhanced the levels of phosphorylated ACC statistically significantly at 6 h (p Ͻ 0.05) compared with vehicle. After 24 h this effect of AICAR treatment was no longer present, and pACC levels were back to base line (Fig. 2, A and B).
AICAR Reduces LTA-induced Lung Inflammation-Treatment with AICAR resulted in reduced cellular counts in BAL fluid 6 h after LTA administration (p Ͻ 0.05, Fig. 3A). Cellular differentiation showed that this reduction was based on a reduction of polymorphonuclear cell (p Ͻ 0.01, Fig. 3C) and lymphocyte (p Ͻ 0.01, Fig. 3C) numbers, whereas macrophage counts were not influenced by AICAR. At 24 h no differences in BAL fluid cellular composition were present between treatment and vehicle groups.
As a measure of vascular leak, total protein levels were determined in BAL fluid. In the ACIAR-treated group protein levels were lowered by 45% relative to vehicle controls after 6 h ( Table  1, p Ͻ 0.01). At 24 h, protein levels in BAL fluid were similar in both treatment groups. To further assess lung damage we measured soluble RAGE in BAL fluid. Soluble RAGE was shown to be a marker of lung epithelial injury based on animal studies in rats and mice of lung injury and clinical measurements of acute lung injury patients (12)(13)(14). Soluble RAGE levels were high at 6 h and decreasing thereafter; no differences were detected between AICAR and vehicle treatment.

DISCUSSION
A potent pulmonary inflammatory response is crucial to host defense against invading pathogens. However, unadapted inflammation can be detrimental to the outcome of infection (15). The Gram-positive cell wall component LTA is a major contributor to the inflammation triggered by Gram-positive bacteria, such as S. aureus, an emerging pathogen in pulmonary infections (1,3).
Here, we studied the effects of AMPK activation by AICAR on pulmonary inflammation induced by administration of LTA via the airways. AICAR was previously shown to reduce LPSinduced lung inflammation (8). We show that AMPK activation by ACIAR diminished cellular influx into the bronchoalveolar space as well as protein leak and inflammatory mediator production at the early phase of LTA elicited lung inflammation.
Previously, ACIAR pretreatment (4 h) was shown to inhibit TNF-␣ production in RAW264.7 murine macrophages in response to LPS (11). Moreover, silencing of AMPK in murine bone marrow-derived macrophages resulted in enhanced TNF-␣, IL-6 and cyclooxygenase-2 mRNA levels in response to LPS stimulation (16). In accordance, transfection of a constitutively active form of AMPK decreased TNF-␣ and IL-6 production in LPS-stimulated bone marrow-derived macrophages     (16). We assessed the impact of direct (no pretreatment) AICAR treatment on an alveolar macrophage cell line. We observed a strong reduction in TNF-␣ and IL-6 levels in alveolar macrophages (MH-S) due to AICAR treatment. This is in accordance with the previously described results on RAW264.7 cells and bone marrow-derived macrophages (11). To assess whether the effects of AICAR were dependent on mTOR signaling, we combined AICAR treatment with rapamycin. As AICAR (unlike rapamycin) did not reduce mTOR or p70/S6K phosphorylation and rapamycin did not affect cytokine production, the effects of AICAR seem to be mediated independent of mTOR.
Our observation on the lack of AICAR effects on p70/S6K phosphorylation are in line with previous findings (17). Our in vivo results are based on applying AICAR treatment concurrently with LTA administration. Other studies applied AICAR in several different models of inflammation. In murine lung inflammation, 500 mg/kg AICAR, given 4 h prior to intratracheal administration of 1 mg/kg LPS, reduced neutrophil accumulation and TNF-␣ and IL-6 protein levels (8). In an OVA and poly(I:C)-based asthmatic exacerbation model, repeated treatment with AICAR (3 ϫ 100 mg/kg) lowered macrophage and eosinophil influx as well as BAL fluid protein levels of IL-5, IL-13 and TNF-␣ (18). Moreover, intravenous infusion of AICAR (0.2 mg/kg/min) reduced BAL fluid protein levels and edema scores and improved survival in porcine mechanical ventilation chest trauma (19). Apart from usage in lung inflammation models, AICAR was also applied in murine trinitrobenzene sulfonic acid-induced colitis. In this model, daily administration of 500 mg/kg AICAR reduced body weight loss and TNF-␣, IFN-␥, IL-17 levels, and partially prevented colon length shortening (20). Although diverse in experimental settings, inflammation types and dosing strategies, all of these studies described anti-inflammatory effects of AICAR treatment and AMPK activation. The anti-inflammatory effects of AICAR reported here in LTA-induced lung inflammation are in accordance with these earlier studies. In addition, we show that pretreatment is not necessary for AICAR to exert its effects.
Different pro-inflammatory (i.e. LPS, LTA, poly(I:C)) stimuli have been used to assess the effects of AMPK activation in inflammation, each acting on different Toll-like receptors (TLRs). LTA signals via TLR2, with myeloid differentiation primary response gene 88 (MyD88) mediating downstream signaling (21). TIR domain-containing adapter-inducing interferon-␤ (TRIF) is the adaptor protein for TLR3 (triggered by poly(I:C) used in Ref. 18). TLR4 stimulation via LPS (used in Refs. 8, 11) acts upon both MyD88 and TRIF (21). The exact mechanism of AMPK modulation on inflammation is unclear. However, given our in vitro results, the effect of AICAR seems to be driven by AMPK in an mTOR-independent manner.
Given the differences in upstream signaling cascades and similar AICAR-induced anti-inflammatory outcomes, it seems that the effect of AICAR and AMPK activation is based on a downstream target of inflammatory signaling cascades. It has been described that AICAR inhibits IB␣ degradation and thereby reduces NFB translocation in LPS-stimulated neutrophils (8). Moreover, AMPK activation may also reduce translo-cation to the cytosol of ELAVL1 (embryonic lethal, abnormal vision, Drosophila)-like 1), which regulates translation of proinflammatory gene mRNAs (e.g. TNF-␣, IL-6) by binding the 3Ј-UTR (22). Another mechanism by which AMPK activation may contribute to inhibition of pulmonary inflammation is by reducing eNOS-dependent leukocyte rolling and adhesion to endothelium (23).
In the current study we showed that AMPK activation by AICAR treatment in vivo reduced LTA-induced murine lung inflammation. We observed reduced pulmonary cell influx and diminished inflammatory mediator production at the early phase of TLR2-dependent lung inflammation (6 h). Based on the current and previous data, activated AMPK immune inhibition is present in a broad range of inflammatory settings and thus may represent an effective strategy in reducing pulmonary inflammation.