Sirtuin 1 Promotes Th2 Responses and Airway Allergy by Repressing Peroxisome Proliferator-Activated Receptor-γ Activity in Dendritic Cells

CD4⁺ Th cells play a central role in the orchestration of adaptive immune responses and are important for protective immunity to many infectious agents. In response to Ag stimulation, naive Th cells may differentiate into at least four types of effector Th cells, namely Th1, Th2, Th17, and inducible regulatory T cells, which differ from each other in their cytokine secretion profile and function (1). Th2 cells, which produce IL-4, IL-5, and IL-13, are indispensable for host immunity to extracellular parasites, such as helminths (2). However, it has also been demonstrated that inappropriate Th2 responses to innocuous environmental Ags are responsible for the development of allergic diseases (3).

Dendritic cells (DCs) are professional APCs required for efficient stimulation of naive Th cells and initiation of primary immune responses (4). They are also needed for optimal restimulation of effector Th cells. Although Th1 induction by DCs is well documented, and rapid progress is being made on Th17 stimulation, the molecular mechanisms by which DCs direct Th2 responses remain more elusive (3). Yet, a better understanding of these mechanisms could help in the design of new therapeutic approaches for allergic disorders such as atopic asthma, a chronic inflammatory condition of the airways that is increasing in severity and prevalence in Western countries (5, 6).

The mammalian sirtuin family of proteins comprises seven NAD⁺-dependent protein deacetylases (Sirt1–7) homologous to the yeast Sir2 protein (7–9). The seven members of the family share a conserved catalytic domain but show differences in their enzymatic activity (7–9). Although Sirt1, Sirt2, Sirt3, and Sirt5 deacetylate histone and nonhistone protein substrates, Sirt4 and Sirt6 mainly act as mono–ADP-ribosyl transferases. The enzymatic activity of Sirt7 has not yet been characterized.

The sirtuins have been implicated in a variety of essential biological functions including cell cycle progression, apoptosis, cellular senescence, stress response, neuronal protection, adaptation to caloric restriction, metabolism, tumorigenesis, and transcription of rRNA genes (8–13). In vitro studies have shown that Sirt1 may also limit the inflammatory process by inhibiting NF-κB and AP-1 (14–18), two transcription factors crucially involved in the expression of proinflammatory cytokines such as TNF-α. Consistent with these findings, lung cells from patients with chronic obstructive pulmonary disease (COPD) and from rats exposed to cigarette smoke display reduced expression of Sirt1 associated with increased NF-κB activity and matrix metalloproteinase-9 expression as compared with lung cells from healthy controls (15, 17). Another argument supporting an anti-inflammatory role for Sirt1 is that transgenic Sirt1 mice under a high-fat diet show lower lipid-induced NF-κB–dependent inflammation than their wild-type (WT) counterparts (19). Moreover, many independent studies have shown that deletion of Sirt1 leads to metabolic diseases associated with inflammatory responses (20–23). Sirt6, like Sirt1, is able to attenuate NF-κB–dependent gene expression (24). However, a recent study revealed that Sirt6 also has the surprising capacity to upregulate the expression of TNF-α by acting at a posttranscriptional level (25), indicating that regulation of the inflammatory response by sirtuins is subtler than previously thought. Consistent with this observation, Lee and collaborators (26, 27) reported that treatment with sirtinol, an inhibitor of Sirt1 and Sirt2, reduces airway inflammation and hyperreactivity in a mouse model of atopic asthma, through mechanisms that yet remain to be formally identified (28).

In this study, we show that mice treated with pharmacological sirtuin inhibitors (i.e., sirtinol or cambinol) are impaired in their ability to mount adaptive Th2 responses and therefore allergic airway inflammation, due to decreased maturation, migration, and pro-Th2 accessory activity of DCs. Using conditional models of genetic depletion, we further demonstrate that differentiation of DCs toward a pro-Th2 phenotype requires inhibition of peroxisome proliferator-activated receptor (PPAR)-γ by Sirt1. This study reveals a previously unappreciated role for Sirt1 in modulating DC function and Th2 responses, thus shedding new light on our current knowledge of the regulation of inflammatory processes by sirtuins.

Female BALB/c and C57BL/6 mice were purchased from Harlan Laboratories. OVA-specific, MHC class II (MHC II)-restricted, TCR-transgenic DO11.10 (H-2^d; BALB/c background) and OT-II (H-2^b; C57BL/6 background) mice, CD11c-CRE transgenic mice (C57BL/6 background), as well as Ppar-γ–floxed (Ppar-γ^fl/fl) mice (C57BL/6 background) and Sirt1^fl/fl mice (C57BL/6 background) were from The Jackson Laboratory. Sirt1^fl/fl mice were crossed with CD11c-CRE transgenic mice for generating Sirt1 conditional knockout (cKO) mice. All mice were housed under specific pathogen-free conditions and used at 8–12 wk of age. All experiments were conducted with Institutional Animal Care and Use Committee approval.

Cambinol and sirtinol were purchased from Calbiochem and Alexis Biochemicals, respectively. OVA grade III and grade V and methacholine were purchased from Sigma-Aldrich. Imject Alum was from Pierce. FITC-OVA was from Molecular Probes Invitrogen. Recombinant murine GM-CSF was produced in-house. PE-conjugated anti-F4/80 (BM8), PE-Cy7–conjugated anti-CCR7 (4B12), allophycocyanin-eFluor780 anti-CD11c (N418), eFluor450–anti-CD4 (RM4-5) and anti-CD19 (1D3) were from eBioscience. Biotinylated anti-MHC II I-E^k (17-3-3) and I-A^b (AF6-120.1), PE–anti-CD3ε (145-2C11), and FITC–anti-CD8α (53-6.7) were from BD Biosciences. FITC-conjugated anti-CD40 (3/23), anti-CD80 (RM80), and anti-CD86 (PO3) were from Serotec. 2.4G2 Fc receptor Abs were produced in-house. Allophycocyanin- and FITC-streptavidin were from eBioscience. Isotype controls were from the same manufacturer as their corresponding Abs.

BALB/c mice were sensitized on days 0 and 14 by i.p. injection of 10 μg OVA (grade V) dissolved in 100 μl PBS and mixed with 100 μl Imject Alum (Pierce). Sham-immunized mice received 100 μl PBS mixed with 100 μl Imject Alum (Pierce). On days 28–32, mice were challenged by exposure to an aerosol of 1% OVA (w/v; grade III) in PBS for 1 h. Twenty-four hours after the last challenge (day 33), airway hyperresponsiveness (AHR) was measured, the mice were killed, and the severity of allergic airway inflammation was assessed.

To assess the effects of specific inhibition of sirtuins on pulmonary allergy, cambinol or sirtinol were injected i.p. to OVA-sensitized mice 1 h prior to each OVA inhalation. Cambinol was administered at the dose of 100 mg/kg in 500 μl 10%/10% ethanol/Cremophore solution (29). Sirtinol was given at the dose of 50 mg/kg in 50 μl 100% DMSO solution. Control mice received vehicle alone.

The trachea was catheterized, and the lungs were lavaged with 1 ml ice-cold Mg- and Ca-free PBS containing 0.6 mM EDTA. Cell density in bronchoalveolar lavage fluid (BALF) was assessed by the use of a hemocytometer. Cell differentials were performed on cytospin preparations stained with Diff-Quick (Dade Behring).

Lungs were fixed in 10% formalin, paraffin embedded, cut in 5-μm sections, and stained with H&E. Intracytoplasmic and luminal mucin was assessed by periodic acid-Schiff (PAS) stains. Mucus production was quantified as the percentage of PAS-stained goblet cells per total epithelial cells in randomly selected bronchi. Seven randomly selected sections were analyzed per murine lung.

Bronchial lymph node (BLN) cells (2 × 10⁵ cells in a 96-well plate) were cultured in Click’s medium supplemented with 0.5% heat-inactivated mouse serum and additives, with or without OVA (50 μg/ml, grade V; Sigma-Aldrich). The proliferation was measured as [³H]thymidine incorporation during the last 16 h of a 3-d culture. Culture supernatants were assayed for IL-4, IL-5, IL-13, and IFN-γ by ELISA (Biosource International).

Mice were anesthetized by i.p. injection (200 μl) of a mixture of ketamine (10 mg/ml; Merial) and xylazine (1 mg/ml; VMD). A tracheotomy was performed by insertion of a 20-gauge polyethylene catheter into the trachea. A ligature was performed around the catheter to avoid leaks and disconnections. Mice were ventilated with a flexiVent small animal ventilator (SCIREQ) at a frequency of 450 breaths per minute and a tidal volume of 10 ml/kg. A positive end expiratory pressure was set at 2 h Pa. Measurement started after 2 min of mechanical ventilation. A sinusoidal 1-Hz oscillation was then applied to the tracheal tube and allowed a calculation of dynamic resistance, elastance, and compliance of the airway by multiple linear regressions. Following measurement of baseline lung function, mice were exposed to a saline aerosol (PBS) followed by aerosols containing increasing doses (3, 6, 9, and 12 g/l) of methacholine. Aerosols were generated by the mean of an ultrasonic nebulizer (SCIREQ) and delivered to the inspiratory line of the flexiVent (SCIREQ) using a bias flow of medical air following the manufacturer’s instructions. Each aerosol was delivered for 10 s, and periods of measurement as described above were made at 1-min intervals following each aerosol. Dynamic resistance was the main parameter measured during the challenge (30).

Staining reactions were performed at 4°C. Cells were incubated with 2.4G2 FcR Abs to reduce nonspecific binding. Analyses were performed on an FACSCanto II flow cytometer (BD Biosciences).

To obtain single lung cell suspensions, lungs were perfused with 20 ml PBS through the right ventricle, cut into small pieces, and digested for 1 h at 37°C in 1 mg/ml collagenase A (Roche) and 0.05 mg/ml DNaseI (Roche) in HBSS. Lung DCs were sorted by flow cytometry (FACSAria; BD Biosciences) based on their differential expression of CD11c and MHC II. Alveolar macrophages, which are also positive for CD11c and MHC II, were excluded based on their autofluorescence and high expression of F4/80 (31). DO11.10 and OT-II CD4⁺ T cells were isolated from the spleen using magnetic bead purification (CD4⁺ T Cell Isolation Kit 130-090-860; Miltenyi Biotec). Isolated cells were either directly used for experiments or cultured in RPMI 1640 medium supplemented with 10% FCS, 2 mM l-glutamine, 1 mM sodium pyruvate, 0.1 mM nonessential amino acids, 50 μM 2-ME, 50 μg/ml streptomycin, and 50 IU/ml penicillin (all from Life Technologies-Invitrogen).

Total RNA was extracted from the cells using TRIzol Reagent (Invitrogen). Poly(A) RNA was primed with oligo(dT) (Roche) and reverse transcribed with AMV reverse transcriptase (Roche) for 1 h at 42°C. cDNA products were amplified by PCR using primers specific for mouse Sirt1 (5′ primer, 5′-CAG ACC CTC AAG CCA TGT TT-3′; 3′ primer, 5′-ACA CAG AGA CGG CTG GAA CT-3′), Sirt2 (5′ primer, 5′-TGG AGA GGC AGA GAT GGA CT-3′; 3′ primer, 5′-GTC ACT CCT TCG AGG GTC AG-3′), Sirt3 (5′ primer, 5′-TAC AGG CCC AAT GTC ACT CA-3′; 3′ primer, 5′-GGA TCC CAG ATG CTC TCT CA-3′), Sirt4 (5′ primer, 5′-CAT CCA GCA CAT TGA TTT CG-3′; 3′ primer, 5′-GTT GGG TTG GTG AGA GGA GA-3′), Sirt5 (5′ primer, 5′-CAT CAC CCA GAA CAT TGA CG-3′; 3′ primer, 5′-CAG TGC CAC ACG AGG TAC AC-3′), Sirt6 (5′ primer, 5′-AGC TGA GAG ACA CCA TTC TGG-3′; 3′ primer, 5′-CCA GGG TGA CAG ACA GGT CT-3′), Sirt7 (5′ primer, 5′-AGC TTC GGG ATA CCA TTG TG-3′; 3′ primer, 5′-CAG GAT TGT GTC TGC TTT GC-3′), and Ppar-γ (5′ primer, 5′-GTC ACG TTC TGA CAG GAC TGT GTG AC-3′; 3′ primer, 5′-TAT CAC TGG AGA TCT CCG CCA ACA GC-3′). All primers were purchased from Eurogentec. A 50 μl PCR reaction was set up containing 5 μl cDNA, 10 mM Tris-HCl, 25 pmol each primer, 1.5 mM MgCl₂, 0.2 mM 2'-deoxynucleoside 5'-triphosphate, and 2.5 U AmpliTaq DNA polymerase (PerkinElmer). Amplification consisted of 30 (sirtuins) or 40 (PPAR-γ) cycles of denaturation at 95°C for 20 s, annealing at 55°C (sirtuins) or 60°C (PPAR-γ) for 30 s, and extension at 72°C for 1 min. Amplification products were electrophoresed on 1.2% agarose gels and visualized by ethidium bromide staining.

In vivo migration of lung DCs was assessed by injecting 800 μg FITC-labeled OVA in the trachea of mice that had been treated with vehicle, cambinol, or sirtinol 18 and 3 h before. Twenty-four hours later, BLNs were collected, and single-cell suspensions were analyzed by flow cytometry for the presence of FITC⁺ DCs (CD11c⁺MHC II⁺F4/80⁻ cells). Dead cells were excluded using DAPI (Molecular Probes Invitrogen). Cambinol was given i.p. 18 and 3 h before intratracheal (i.t.) injection of FITC-OVA at the dose of 100 mg/kg in 500 μl 10%/10% ethanol/Cremophore solution. Sirtinol was administered i.p. at the same time points at the dose of 50 mg/kg in 50 μl 100% DMSO solution. Control mice received the vehicle alone.

Cells were assayed for apoptosis using dual-color annexin V-FITC/propidium iodide staining (Annexin V-FLUOS staining kit; Roche Applied Sciences) and flow cytometry analyses.

Bone marrow cells were collected from naive mice, depleted of RBCs using ammonium chloride lysis buffer, and grown for 8 d in DMEM supplemented with 10% FCS, 1% l-glutamine, 1% sodium pyruvate, 0.1% 2-ME, 50 μg/ml streptomycin, 50 IU/ml penicillin, and 20 ng/ml recombinant murine granulocyte/M-CSF (32). At day 8, >90% of bone marrow cells were positive for the DC marker CD11c and considered as bone marrow-derived DCs (BMDCs). In all experiments, BMDCs were stimulated at day 8.

FACS-sorted lung DCs (CD11c⁺MHC II⁺F4/80⁻ cells; 10⁶ cells/ml) or BMDCs from naive BALB/c or C57BL/6 mice were stimulated for 16 h with 125 μg/ml OVA in the presence of vehicle alone (0.1% DMSO), 50 μM cambinol, or 25 μM sirtinol. DCs were then extensively washed, irradiated, and cocultured for 72 h with freshly isolated DO11.10 or OT-II CD4⁺ T cells (1 × 10⁴ DCs were used to stimulate 2 × 10⁵ DO11.10 or OT-II T cells). Unpulsed, vehicle-treated DCs were used as controls. The proliferation of DO11.10 or OT-II T cells was measured as [³H]thymidine incorporation during the last 16 h of the 72-h culture. Culture supernatants were assayed for IL-4, IL-5, and IL-13 by ELISA (Biosource International).

At day 8 of culture, BMDCs were stimulated with 125 μg/ml OVA in the presence of vehicle, cambinol, or sirtinol. At day 9, cells were collected, washed, and resuspended in PBS. A total of 10⁶ cells were then injected i.t. to anesthetized naive recipients. Unpulsed, vehicle-treated BMDCs were used as controls. Ten days after i.t. immunization, mice were challenged with OVA (1% w/v in PBS; grade III) aerosol during a daily 30 min challenge on 5 consecutive d. Twenty-four hours after the last challenge, the mice were killed, and airway allergy was evaluated as described above.

Nuclear protein extracts were prepared as previously described (33). Cytoplasmic buffer (pH 7.9) contained 10 mM HEPES, 10 mM KCl, 2 mM MgCl₂, 0.1 mM EDTA, 0.2% (v/v) Nonidet P-40, 1.6 mg/ml protease inhibitors (Complete; Roche), and 3 mM serine protease inhibitor diisopropyl fluorophosphate (Sigma-Aldrich). Pelleted nuclei were resuspended in 20 mM HEPES (pH 7.9), 1.5 mM MgCl₂, 0.2 mM EDTA, 0.63 M NaCl, 25% (v/v) glycerol, 1.6 mg/ml protease inhibitors, and 3 mM DFP (nuclear buffer), incubated for 20 min at 4°C, and centrifuged for 30 min at 14,000 rpm. Protein amounts were quantified with the Micro BCA protein assay reagent kit (Pierce).

Binding reactions were performed for 30 min at room temperature with 5 μg nuclear proteins in 20 mM HEPES (pH 7.9), 10 mM KCl, 0.2 mM EDTA, 20% (v/v) glycerol, 1% (w/v) acetylated BSA, 3 μg poly(deoxyinosinic-deoxycytidylic) acid (Amersham Pharmacia Biotech), 1 mM DTT, 1 mM PMSF, and 100,000 cpm ³²P-labeled double-stranded oligonucleotide probes. Probes were prepared by annealing the appropriate single-stranded oligonucleotides (Eurogentec) at 65°C for 10 min in 10 mM Tris, 1 mM EDTA, and 10 mM NaCl, followed by slow cooling to room temperature. The probes were then labeled by end-filling with the Klenow fragment of Escherichia coli DNA polymerase I (Roche), with (α-[³²P])–2′-deoxyadenosine triphosphate and (α-[³²P])–2′-deoxycytidine triphosphate (Dupont-New England Nuclear Life Science Products). Labeled probes were purified by spin chromatography on Sephadex G-25 columns (Roche). DNA–protein complexes were separated from unbound probe on 4% native polyacrylamide gels at 150 V in 0.25 M Tris, 0.25 M sodium borate, and 0.5 mM EDTA (pH 8). Gels were vacuum-dried and exposed to Fuji x-ray film (Fujifilm) at −80°C for 12 h. To confirm specificity, competition assays were performed with a 50-fold excess of unlabeled WT probes and with mutated probes. The sequences of the oligonucleotides used in this work were as follows: WT PPAR-γ probe, 5′-TTG GCA AAA CTA GGT CAA AGG TCA CA-3′ (forward) and 5′-TTG GTG TGA CCT TTG ACC TAG TTT TG-3′ (reverse); and mutated PPAR-γ probe, 5′-TTG GCA AAA CTA GCA CAA AGC ACA CA-3′ (forward) and 5′-TTG GTG TGT GCT TTG TGC TAG TTT TG-3′ (reverse) (34).

A lentiviral construct encoding the Cre recombinase under the phosphoglycerate kinase (PGK) promoter was generated by subcloning the PGK promoter from pPGKpuro and the Cre recombinase from pBOB-CAGGS-iCre-SD into the pCSC-SP-PW lentiviral vector [kind gifts from Dr. Inder M. Verma, Salk Institute for Biological Studies, La Jolla, CA (35)]. Lentiviral particles were produced using a third-generation lentiviral packaging system, concentrated, and titrated as previously described (35).

The BMDC culture medium was replaced at day 4 by medium containing recombinant viral particles at a multiplicity of infection of 15 and 5 μg/ml protamine sulfate (36). Six hours later, infecting medium was replaced by fresh culture medium. At day 8, the cells were treated or not with cambinol or sirtinol and pulsed or not with OVA. They were collected and used at day 9.

Data are presented as means ± SDs. Normality of the data distributions was assessed using Anderson-Darling tests. The differences between values were subsequently estimated using an ANOVA test followed by a Fisher’s protected least SD test. A value <0.05 was considered significant. All of the experiments were repeated at least three times.

Recent evidence indicates that inhibition of sirtuins using the broad-spectrum inhibitor sirtinol decreases eosinophilic airway inflammation in a mouse model of allergic asthma (28). To obtain more certainty about the specificity of these effects, we repeated these experiments using sirtinol (Supplemental Fig. 1), but also cambinol (Fig. 1), another pharmacological inhibitor of sirtuins (29). Sirtuin inhibition was achieved by i.p. administration of the inhibitors to mice previously sensitized with OVA 1 h before each repeated airway challenge with aerosolized OVA. OVA-sensitized mice exposed to aerosolized OVA and treated with sirtinol or cambinol are hereafter referred to as OVA/sirtinol-OVA and OVA/cambinol-OVA mice, respectively. Control OVA-sensitized mice exposed to aerosolized OVA that were treated with vehicle alone are called OVA/vehicle-OVA mice. Sham-immunized (PBS in Imject Alum) mice that were injected with vehicle during OVA challenge served as negative controls (PBS/vehicle-OVA mice). As expected in this model, OVA/vehicle-OVA mice, but not PBS/vehicle-OVA mice, developed peribronchial and perivascular lung tissue eosinophilia and lymphocytosis, accompanied by goblet cell hyperplasia and increased mucus production (Fig. 1A, 1B). These modifications were congruently reflected in the increased total and eosinophil counts measured in the BALF of OVA/vehicle-OVA mice compared with PBS/vehicle-OVA animals (Fig. 1C). All of these inflammatory signs were significantly reduced when sensitized mice were treated with cambinol or sirtinol during OVA challenge (Fig. 1A–C, Supplemental Fig. 1A–C). Moreover, Th2 cytokine production and T cell proliferation in BLNs were significantly attenuated in OVA/sirtinol-OVA and OVA/cambinol-OVA mice compared with OVA/vehicle-OVA counterparts (Fig. 1D, 1E, Supplemental Fig. 1D, 1E). Finally, AHR to nonspecific stimuli, a cardinal feature of allergic asthma, was significantly decreased in OVA/cambinol-OVA and OVA/sirtinol-OVA mice compared with OVA/vehicle-OVA animals (Fig. 1F, Supplemental Fig. 1F).

Together, these results support that pharmacological inhibition of sirtuins during allergen provocation may suppress the cardinal features of airway allergy and therefore support a crucial role for these proteins in the development of allergic airway inflammation and AHR.

Th2 responses were significantly downregulated in OVA/sirtinol-OVA and OVA/cambinol-OVA mice, raising the possibility that they were not optimally activated. Lung DCs are powerful APCs that have been found to be necessary and sufficient for Th2 cell stimulation during ongoing allergic airway inflammation (37). We therefore hypothesized that cambinol and sirtinol could exert anti-inflammatory effects by interfering with the function of these cells. To support our assumption, we first assessed by RT-PCR the expression of sirtuins in lung DCs. As shown in Fig. 2A, lung DCs expressed Sirt1, Sirt2, and Sirt7 mRNA, whereas the transcripts of Sirt3, Sirt4, Sirt5, and Sirt6 were only barely detectable.

Migration of lung DCs to the BLNs following allergen challenge is required for efficient activation of Th2 cells and, therefore, for initiation and maintenance of airway allergic inflammation (37, 38). We thus examined whether treatment with sirtuin inhibitors affected the migratory properties of lung DCs. First, the total number of DCs in draining BLNs was counted in PBS/vehicle-OVA, OVA/vehicle-OVA, OVA/cambinol-OVA, and OVA/sirtinol-OVA mice 24 h following the last challenge with aerosolized OVA. As previously described (38), OVA/vehicle-OVA mice displayed a significant increase in the number of DCs in BLNs compared with PBS/vehicle-OVA counterparts (Fig. 2B, 2C). Interestingly, the influx of DCs in BLNs was strongly reduced when sensitized mice were treated with either cambinol or sirtinol prior to each OVA challenge (Fig. 2B, 2C). Of note, the number of circulating DCs, T cells, and B cells, as well as the distribution of these cell types in nondraining axillary and inguinal LNs, were unaffected in cambinol- and sirtinol-treated mice compared with OVA/vehicle-OVA animals (data not shown).

To unambiguously demonstrate an effect of cambinol and sirtinol on the migratory capacity of lung DCs, we adapted an assay developed by Vermaelen et al. (38). In this assay, 800 μg FITC-labeled OVA was given i.t. to vehicle-, cambinol-, or sirtinol-treated mice, and migrating lung DCs were counted in BLNs 24 h later by flow cytometry. As shown in Supplemental Fig. 2A and 2B, the number of Ag-loaded DCs was significantly lower in the BLNs of cambinol- and sirtinol-treated mice than in those of vehicle-treated controls, confirming an inhibitory effect of these drugs on lung DC migration.

Induction of robust Th2 immune responses in the airways requires not only the migration of allergen-loaded lung DCs to the BLNs, but also their maturation (39). We thus examined whether treatment with cambinol and sirtinol could alter the maturation process in lung DCs. For that, lung DCs were FACS sorted and stimulated ex vivo with OVA in the presence or absence of sirtuin inhibitors. Of note, the commercial batch of OVA we used in the current study contained enough endotoxin to induce full DC maturation. As expected, OVA-stimulated lung DCs, compared with freshly isolated lung DCs, displayed increased expression of the maturation markers CD40, CD80, and CD86 (Fig. 2D). Expression of these markers was substantially reduced when the cells were treated with cambinol or sirtinol before OVA stimulation (Fig. 2D). Consistent with our results on DC migration, expression of CCR7, a chemokine receptor playing a pivotal role in DC trafficking (40), was significantly decreased when OVA-stimulated lung DCs were treated with sirtuin inhibitors (Fig. 2D). To ascertain that these observations were not due to toxic effects of cambinol and sirtinol, apoptosis was measured in treated and untreated lung DCs. Fig. 2E shows that neither cambinol nor sirtinol induced significant apoptosis in these cells.

Our data on DC maturation implied that sirtuin inhibition was able to compromise the ability of lung DCs to stimulate T cells. To further address this point, we measured the effects of cambinol and sirtinol on lung DC-induced proliferation of OVA-specific T cells in vitro. For this purpose, we used CD4⁺ T cells from OVA-specific, TCR-transgenic DO11.10 mice (41). Fig. 3A shows that treatment of OVA-pulsed lung DCs with sirtuin inhibitors significantly reduced their ability to induce DO11.10 T cell proliferation. Moreover, DO11.10 T cells stimulated with cambinol- or sirtinol-treated, OVA-pulsed lung DCs secreted lower amounts of Th2 cytokines than those stimulated with vehicle-treated control DCs (Fig. 3B), clearly demonstrating that sirtuins are required for efficient activation of Th2 cells by lung DCs.

To definitively prove that sirtuin inhibition prevents DC-driven Th2 responses in the lung, we used a model in which i.t. transfer of OVA-pulsed BMDCs leads to allergic airway inflammation upon subsequent OVA aerosol challenge (39). BMDCs were used because the number of DCs obtained from the lungs of mice was too small to perform these large-scale experiments. This model is relevant in the study of the biology of DCs in airway allergy, because it was recently shown that BMDCs display a phenotype very similar to that of lung DCs upon allergenic challenge of the airways (42). Further supporting that this model is relevant to the current study, we demonstrated that the ability of BMDCs to mature and stimulate DO11.10 T cells in vitro was abolished upon cambinol and sirtinol treatment (Supplemental Fig. 3A–D), as shown for lung DCs (Fig. 3). Moreover, BMDCs expressed the same sirtuins as lung DCs, namely Sirt1, Sirt2, and Sirt7, along with low levels of Sirt3, Sirt5, and Sirt6 (Supplemental Fig. 4). As previously shown by Lambrecht et al. (39), mice that were immunized with OVA-pulsed, vehicle-treated BMDCs developed BALF and peribronchial eosinophilia following OVA challenge, an effect not observed in mice that were injected with unpulsed, vehicle-treated BMDCs (Fig. 4A, 4B). The pretreatment of OVA-pulsed BMDCs with either cambinol or sirtinol significantly reduced their capacity to induce the development of allergic airway inflammation (Fig. 4A, 4B). These effects of cambinol and sirtinol were accompanied by a reduced capacity of BLN T cells to proliferate and produce Th2 cytokines following OVA restimulation (Fig. 4C, 4D).

Altogether, these results demonstrate that sirtuin inhibition impairs maturation and migration of lung DCs, thereby dampening DC-driven Th2 priming and allergic airway inflammation.

Sirtuins have been shown to be capable of repressing PPAR-γ (43, 44), a nuclear receptor mainly involved in the regulation of carbohydrate and lipid metabolism (45). Given that activation of PPAR-γ in DCs renders them unable to prime Th2 responses in the airways (46), we hypothesized that abrogation of sirtuin-mediated PPAR-γ repression could be the molecular mechanism by which sirtuin inhibitors exert their anti-inflammatory effects in lung DCs.

To test our hypothesis, we first assessed the activity of PPAR-γ by EMSA in vehicle-, cambinol-, and sirtinol-treated, OVA-pulsed DCs. As shown in Fig. 5A (lanes 1–5), treatment of OVA-pulsed BMDCs with sirtuin inhibitors led to a strong increase in PPAR-γ DNA-binding activity, whereas treatment of unpulsed or OVA-pulsed DCs with vehicle alone had no effect on basal PPAR-γ activity. DNA-binding competition experiments using 50-fold excess of unlabeled WT and mutated probes confirmed the specificity of PPAR-γ binding (Fig. 5A, lanes 6 and 7).

To demonstrate that PPAR-γ activation was responsible for the anti-inflammatory effects of sirtuin inhibitors, it was necessary to show that cambinol and sirtinol were no longer able to inhibit DC function in the absence of PPAR-γ. We thus generated PPAR-γ–deficient DCs by transducing Ppar-γ–floxed (Ppar-γ^fl/fl) BMDCs with an advanced generation lentiviral vector carrying the Cre recombinase cDNA under the control of the PGK promoter. WT BMDCs transduced with the same lentiviral vector were used as controls. Efficient Cre-mediated recombination in Ppar-γ^fl/fl cells results in deletion of exons 1 and 2 of the Ppar-γ gene and therefore in shortening of the PPAR-γ transcript [300 bp for the recombinant form versus 700 bp for the floxed form (47)]. The short transcript was predominant in Cre recombinase-transduced Ppar-γ^fl/fl BMDCs, whereas only the full-length transcript was detected in control cells, confirming that this lentiviral transduction method allowed the generation of Ppar-γ^−/− BMDCs (Fig. 5B). Moreover, the DNA binding activity of PPAR-γ did not increase in Cre recombinase-transduced Ppar-γ^fl/fl BMDCs following treatment with cambinol or sirtinol, demonstrating a loss of PPAR-γ function in these cells (Fig. 5C). It has to be noted that no shifted bands were detected when nuclear extracts from Ppar-γ^−/− BMDCs, treated or not with cambinol or sirtinol, were used (Fig. 5C), confirming the specificity of the EMSA complexes.

The functional experiments described above for WT DCs were repeated with Ppar-γ^−/− DCs to prove the involvement of this NF in the anti-inflammatory effects of sirtuin inhibitors. As shown in Fig. 6, cambinol and sirtinol failed to diminish the capacity of OVA-pulsed Ppar-γ^−/− BMDCs to stimulate OVA-specific T cells in vitro, whereas they did in control experiments. Moreover, pretreatment of OVA-pulsed Ppar-γ^−/− BMDCs with either cambinol or sirtinol did not reduce their ability to induce allergic airway inflammation following i.t. transfer (Fig. 7).

These data show that sirtuin inhibitors impair DC function and therefore the development of asthmatic responses by derepressing the anti-inflammatory factor PPAR-γ.

We finally sought to determine which of the sirtuins is required for the pro-Th2 activity of DCs in allergic conditions. Sirtinol has first been reported to selectively inhibit Sirt2 NAD-dependent deacetylase activity in vitro with an IC₅₀ value of 58 μM (26). However, further studies demonstrated that sirtinol also antagonizes Sirt1, although the IC₅₀ value in this case was higher (131 μM) (27). Cambinol inhibits Sirt1 and Sirt2 with similar IC₅₀ values of 56 and 59 μM, respectively (29). Therefore, although cambinol has been reported to have a very weak inhibitory activity against Sirt5 (42% inhibition at 300 μM) (29), the activity of sirtinol and cambinol seems to be principally restricted to Sirt1 and Sirt2. This, coupled with the fact that Sirt1 is the main sirtuin known to repress PPAR-γ (43), made Sirt1 the most likely candidate implicated in DC-mediated induction of Th2 responses in the airways.

We therefore investigated the possible involvement of Sirt1 in the pro-Th2 activity of DCs. Sirt1^−/− mice display an elevated postnatal lethality attributable to developmental problems (48, 49), which precludes their use in models of induced allergy. We therefore generated mice conditionally deficient for Sirt1 in DCs (Sirt1 cKO) by crossing mice with a floxed Sirt1 gene (Sirt1^fl/fl mice) (50) with transgenic mice expressing the Cre recombinase under the dependence of the Cd11c promoter (CD11c-Cre mice) (51). Sirt1^fl/fl and CD11c-Cre mice were used as control. Because both controls gave comparable results, only those in Sirt1^fl/fl are shown. In the model of induced airway allergy, Sirt1 cKO mice displayed significantly reduced parameters of airway inflammation compared with control mice (Fig. 8). This correlated with decreased migratory activity of lung DCs upon Ag exposure of the airways (Fig. 9A). Finally, when assessing whether Sirt1 intervenes in the repression of PPAR-γ in DCs under pro-Th2 conditions, EMSAs revealed that PPAR-γ activity was significantly upregulated in Sirt1 cKO DCs compared with control DCs (Fig. 9B).

These results thus identify Sirt1 as a sirtuin implicated in the repression of PPAR-γ in lung DCs and the subsequent promotion of a pro-Th2 phenotype of these cells in allergic conditions.

The sirtuins are a highly conserved family of NAD⁺-dependent enzymes that are known to regulate diverse biological functions, including aging, cellular stress resistance, tumorigenesis, and energy metabolism (7–9). In vitro studies have demonstrated that sirtuins may also dampen the inflammatory response by inhibiting NF-κB and AP-1 (14–18), two proinflammatory transcription factors. Given that NF-κB and AP-1 are both critically involved in the effector phase of allergic airway inflammation (52, 53), it may have been expected that the asthmatic reaction would be aggravated by the inhibition of sirtuins. Yet, our data, in agreement with a recent report (28), indicate that treatment of mice with sirtuin inhibitors decreases the parameters of allergic airway inflammation in a mouse model of allergic asthma.

In their report, Kim et al. (28) suggested that the anti-inflammatory effects of sirtinol on airway allergy were due to the impairment of the activation by Sirt1 of hypoxia-inducible factor (HIF)-1α, a transcription factor activated during, and regulator of, inflammatory responses (54). Yet, the authors only used a pharmacological approach and a single sirtuin inhibitor, sirtinol, which is known to have higher inhibitory potential on Sirt2 than on Sirt1 (27). The sirtuin implicated in the observed anti-inflammatory effects of sirtinol thus was not formally identified. Furthermore, the evidence that sirtinol downregulates HIF-1α activity in this model was purely correlative, and it cannot be ruled out that the observed reduction in HIF-1α activity was a collateral effect of the downregulation of inflammation by other mechanisms.

Consistent with the existence of such HIF-1α–independent mechanisms, our data indeed indicate that sirtuin inhibition impairs the optimal reactivation of Th2 responses upon allergen challenge of the airways through PPAR-γ–dependent mechanisms. Because Th2 responses are the essential drivers of allergic airway inflammation, it is thus most likely that inhibition of these adaptive responses is a major mechanism by which cambinol and sirtinol suppress all of the characteristic features of asthmatic inflammation.

We also formally identify Sirt1 as a sirtuin promoting Th2 responses through its inhibitory effect on PPAR-γ activity in DCs. Yet, the inhibition of airway inflammation observed in Sirt1 cKO mice was slightly lower than the one observed with pharmacological sirtuin inhibition (compare Fig. 1A–C with 8A–D). Our data thus support that Sirt1 activity in DCs promotes adaptive Th2 responses in allergic conditions and that this pathway is an important contributor to allergic airway inflammation, but do not rule out that Sirt1 and/or Sirt2 may have additional proinflammatory effects in other immune cells implicated in this type of inflammation. For instance, we observed that sirtinol and cambinol inhibit proliferation and cytokine secretion of Th2 cells polarized and activated in vitro, without affecting cell survival (A. Mayer and F. Andris, unpublished observations). Sirtuins might thus also regulate T cell responses directly, although this hypothesis will require further testing.

As illustrated by the above considerations, the role of sirtuins in immune cells remains largely unknown. In particular, whether sirtuins are expressed in DCs and whether they modulate DC function had never been investigated. In the present report, we showed that lung DCs, which are necessary and sufficient for Th2 cell stimulation during ongoing allergic airway inflammation (37, 55), express mainly Sirt1, Sirt2, and Sirt7. Moreover, we demonstrated that pharmacological inhibition of sirtuins impairs allergen-induced maturation and migration of lung DCs, thereby downregulating Th2 cell activation. Using conditional genetic deletion, we furthermore showed that the effects of sirtuin inhibitors on adaptive Th2 responses depended on Sirt1 activity in DCs. Finally, we could demonstrate that Sirt1 inhibits PPAR-γ activity in DCs. PPAR-γ is a key regulator of carbohydrate and lipid metabolism (45), but growing evidence indicates that PPAR-γ also plays an important role in controlling immune and inflammatory responses, including allergic airway inflammation. Indeed, the PPAR-γ agonists ciglitazone, rosiglitazone, and GI 262570 display beneficial effects in various animal models of allergic asthma at least in part due to the capacity of PPAR-γ to reduce maturation, migration, and immunogenicity of DCs (56–60). Thus, given that most of the effects of sirtuin inhibitors were demonstrated on highly purified DCs and were similar to the effects observed in mice conditionally deficient for Sirt1 in their DCs, it is evident that Sirt1 acts in a cell-autonomous manner to control DC function. Our study mainly focused on the role of sirtuins in the induction of Th2 responses by lung DCs. We nevertheless also observed that treatment of keyhole limpet hemocyanin-loaded BMDCs with cambinol or sirtinol prior to footpad inoculation in naive recipient mice inhibits the induction by these cells of both Th1 and Th2 responses in the draining lymph nodes (T. Marichal, unpublished observations). Sirtuin activity might thus be generally required for the activity of DCs as APCs. Therefore, it will be worthwhile studying whether sirtuins are also expressed in other DC populations and whether and how these enzymes participate in the induction of Th1 and Th17 responses by DCs in vivo.

An intriguing question is how sirtuins exert both pro- and anti-inflammatory functions. As mentioned above, Sirt1 seems to play an anti-inflammatory role in COPD (15, 17). COPD is a chronic inflammatory disease of the lungs mainly caused by repeated exposure to cigarette smoke, and there is accumulating evidence that accelerated aging of lungs in response to cigarette smoke-induced oxidative stress is critically implicated in the pathogenesis and progression of this disease (6, 61). It has been demonstrated that sirtuins confer cellular resistance to oxidative stress (9, 12). This increased stress resistance comes from the inter-action between sirtuins and members of the forkhead box class O (FOXO) family of transcription factors. Sirt1 can bind and deacetylate FOXO3a, leading to a selective upregulation of FOXO-regulated stress resistance genes (62, 63). In addition, Sirt1 can also protect cells against oxidative stress by regulating the heat shock response and inhibiting p53 transcriptional activity (64–66). It is therefore probable that sirtuins prevent the development of COPD not only by antagonizing NF-κB but also by attenuating the deleterious effects of cigarette smoke-induced oxidative stress. Sirtuins have also been reported to play an anti-inflammatory role in metabolic diseases. For example, transgenic Sirt1 mice under a high-fat diet display lower lipid-induced inflammation than WT controls (19). Moreover, sirtuins improve insulin sensitivity in adipocytes through their anti-inflammatory properties (18). It thus appears that sirtuins are protective in inflammatory diseases related to aging, oxidative stress, and metabolic dysfunction. Unlike allergic airway inflammation, these inflammatory conditions do not essentially depend on Ag-specific T cell responses and rather are mainly driven by innate inflammatory mechanisms. As we have demonstrated that sirtuins are required for optimal adaptive Th2 responses, one may speculate that sirtuins could promote T cell-dependent inflammations, such as airway allergy, but rather dampen more T cell-independent inflammatory conditions, such as COPD.

The underlying pathogenic mechanisms in asthma and airway allergy are only partly understood, a situation reflected in the variable and incomplete responses by patients to current therapeutics. The emerging concept that targeting airway DC function might be a powerful method to efficiently treat airway allergy has recently been validated by the group of Lambrecht (67, 68). The present study demonstrates for the first time, to our knowledge, that sirtuins (i.e., Sirt1) promote allergic airway inflammation by boosting lung DC function. We therefore suggest that blocking sirtuin activity in lung DCs may have therapeutic potential in the treatment of airway allergy.

We thank Dr. Inder M. Verma (Salk Institute for Biological Studies, La Jolla, CA) for generously providing the lentiviral vectors, Dr. Philippe Gosset (Institut Pasteur de Lille, Lille, France) for advice, Sandra Ormenese and the Cell Imaging and Flow Cytometry Interdisciplinary Cluster for Applied Genoproteomics Technological Platform for FACS analyses, Dr. Gilles Gaudray for animal management, and Valérie Conrath, Sébastien Denanglaire, Luc Duwez, Cédric François, Monique Henket, and Ilham Sbai for excellent technical and secretarial assistance.

The authors have no financial conflicts of interest.