Sirtuin 2 Deficiency Increases Bacterial Phagocytosis by Macrophages and Protects from Chronic Staphylococcal Infection

Sirtuin 2 (SIRT2) is one of the seven members of the family of NAD+-dependent histone deacetylases. Sirtuins target histones and non-histone proteins according to their subcellular localization, influencing various biological processes. SIRT2 resides mainly in the cytoplasm and regulates cytoskeleton dynamics, cell cycle, and metabolic pathways. As such, SIRT2 has been implicated in the pathogenesis of neurodegenerative, metabolic, oncologic, and chronic inflammatory disorders. This motivated the development of SIRT2-directed therapies for clinical purposes. However, the impact of SIRT2 on antimicrobial host defense is largely unknown. Here, we address this question using SIRT2 knockout mice. We show that SIRT2 is the most highly expressed sirtuin in myeloid cells, especially macrophages. SIRT2 deficiency does not affect immune cell development and marginally impacts on intracellular signaling and cytokine production by splenocytes and macrophages. However, SIRT2 deficiency enhances bacterial phagocytosis by macrophages. In line with these observations, in preclinical models, SIRT2 deficiency increases survival of mice with chronic staphylococcal infection, while having no effect on the course of toxic shock syndrome toxin-1, LPS or TNF-induced shock, fulminant Escherichia coli peritonitis, sub-lethal Klebsiella pneumoniae pneumonia, and chronic candidiasis. Altogether, these data support the safety profile of SIRT2 inhibitors under clinical development in terms of susceptibility to infections.

Innate immune cells are at the vanguard of host defense against microbial infections. Monocytes/ macrophages and dendritic cells (DCs) sense microbial or danger-associated molecular patterns (MAMPs and DAMPs released by injured or stressed cells) through pattern recognition receptors (PRRs) such as toll-like receptors (TLRs), NOD-like receptors, C-type lectins, scavenger receptors, RIG-I-like receptors, and intra-cytosolic DNA sensors (1,2). The interaction of MAMPs or DAMPs with PRRs triggers the mitogen-activated protein kinases (MAPKs), nuclear factor-κB (NF-κB), and interferon (IFN) response factor signaling pathways that coordinate immune gene expression. The cellular and soluble mediators mobilized upon infection tightly regulate the development of the inflammatory response, the establishment of antimicrobial cellular and humoral responses, and the restoration of homeostasis once pathogens have been contained or eradicated.
The superfamily of histone deacetylases (HDACs) comprises eleven Zn-dependent HDACs (HDAC1-11) and seven NAD +dependent sirtuins (SIRT1-7). HDACs are epigenetic erasers catalyzing histone deacetylation, chromatin compaction, and transcriptional repression. In addition, HDACs target thousands of non-histone proteins affecting many biological processes (3). The subfamily of sirtuins attracted much interest when sirtuins were proposed to promote longevity and represent attractive therapeutic targets for age-related pathologies, such as type 2 diabetes, as well as neurodegenerative, cardiovascular, and oncologic diseases (4)(5)(6). Our knowledge about the impact of sirtuins on innate immune responses is limited. Most studies have focused on SIRT1 and SIRT6, resulting in an overall complex picture attributing both proinflammatory and anti-inflammatory properties to sirtuins (7).
Sirtuin 2 (SIRT2) was originally reported to colocalize with the microtubule network and to deacetylate α-tubulin (8). Indeed, SIRT2 is mainly cytoplasmic, although it can translocate during the G2/M transition phase of cell cycle into the nucleus where it deacetylates histone H4 lysine 16 (9). Of note, SIRT2 possesses a proficient demyristoylation activity, the physiological relevance of which remains to be established (10). By targeting numerous proteins besides histones, SIRT2 regulates cytoskeleton dynamics, cell cycle, and metabolic pathways including inhibition of adipogenesis and promotion of lipolysis and gluconeogenesis. Additionally, SIRT2 can act as a tumor suppressor gene, and is involved in myelogenesis and other brain functions. Accordingly, SIRT2 has been implicated in tumorigenesis and neurodegeneration, and likely plays a role in metabolic disorders, such as obesity and type 2 diabetes (11)(12)(13).
With respect to inflammatory processes, SIRT2 was first shown to deacetylate NF-κB p65, resulting in the expression of a subset of p65-dependent genes in mouse embryonic fibroblasts (14). SIRT2 deficiency sustained brain inflammation in a model of traumatic brain injury and increased the severity of collageninduced arthritis and colitis (15)(16)(17). However, SIRT2 deficiency was also reported to decrease NF-κB p65-mediated inflammatory response, renal tubular inflammation, and ischemia reperfusioninduced hepatocellular inflammation (18)(19)(20). Finally, in an experimental stroke model, SIRT2 deficiency preserved neurological functions without affecting inflammatory parameters (21). Overall, the function of SIRT2 in sterile and chronic inflammatory disorders appears to be context dependent.
Because sirtuins are pleiotropic and in consideration of the development of sirtuin-targeting drugs for clinical conditions, we sought to delineate the role of SIRT2 in the innate immune response. To this end, we used SIRT2 knockout mice to investigate the response of immune cells to immunological and microbial stimuli using in vitro experiments and in vivo preclinical models. Preclinical models included models of shock as well as Gram-negative and Gram-positive bacterial infections and fungal infection. Overall, SIRT2 deficiency does not modulate cytokine production by innate immune cells, but enhances bacterial phagocytosis by macrophages. SIRT2 deficiency protects from chronic staphylococcal infection, while having no impact on toxic shock, endotoxemia, fulminant peritonitis, non-lethal pneumonia, and chronic candidiasis. These data largely support the safety, in terms of susceptibility to infections, of SIRT2 inhibitors developed for clinical applications.

Western Blot analyses
Nuclear and total protein extracts were submitted to PAGE and transferred onto nitrocellulose membranes (30,31). Membranes were incubated with antibodies directed against SIRT2, acetylated α-tubulin, total α-tubulin, total and phosphorylated ERK1/2, p38, JNK, and NF-κB p65 and TATA-box binding protein (used as a control of nuclear extracts) (see antibody description in Table  S2 in Supplementary Material), then with a secondary horseradish peroxidase-conjugated antibody (Sigma-Aldrich) (32). Blots were imaged with the enhanced chemiluminescence Western blotting system (GE Healthcare, Little Chalfont, Royaume-Uni). Images were recorded using a Fusion Fx system (Viber Lourmat, Collégien, France).

Flow cytometry
Single cell suspensions from thymus and spleen, or BMDMs were enumerated and incubated with 2.4G2 monoclonal antibody (mAb). Immune cell subpopulations were tracked by staining performed using mAbs described in Table S2 in Supplementary Material. Splenic CD4 + CD25 + Foxp3 + cells were detected using The Mouse Regulatory T Cell Staining Kit (eBioscience). Data were acquired using a LSR II flow cytometer (BD Biosciences) and analyzed using FlowJo Version 10.2 software (FlowJo LLC, Ashland, OR, USA) (33).

Proliferation assay
The proliferation of 1.5 × 10 5 splenocytes cultured for 48 h in 96-well plates was quantified by measuring 3 H-thymidine incorporation over 18 h (34).

Phagocytosis assays
Fluoresbrite ® Yellow Green Microspheres (Polysciences Inc, Warrington, PA, USA) or FITC-labeled bacteria were added to cells at a ratio of 10 beads or bacteria/cell. After 1 h, cells were washed, incubated for 1 min with trypan blue (0.25 mg/ml) and analyzed by flow cytometry. When specified, beads were opsonized with serum for 30 min at 37°C. To assess phagocytosis of live bacteria, BMDMs (in quadruplicates or sextuplates) were incubated for 1 h with E. coli O18, S. aureus, and GBS (10 bacteria/cell). Non-adherent and extracellular bacteria were removed by washing and killed by a 30-min exposure to 100 µg/ ml gentamicin (Essex Chemie, Luzern, Switzerland; for E. coli and GBS) or 10 µg/ml ciprofloxacin (Fresenius Kabi, Oberdorf, Switzerland; for S. aureus). Serial dilutions of cell lysates were plated on agar plates. Colonies were enumerated to calculate the number of phagocytosed bacteria.

glycolytic activity
The glycolytic activity of BMDMs was analyzed using a 96-well format Seahorse XFe ® system and the Seahorse XF Glycolysis Stress Test Kit (Agilent Technologies, Santa Clara, CA, USA). Briefly, 4 × 10 4 BMDMs were plated in 96-well plates in IMDM medium. The next day, cells were incubated with or without 5 × 10 7 CFU/ml heat-killed S. aureus and rested 1 h in Seahorse medium without glucose. The glycolytic capacity was assessed by measuring the extracellular acidification rate following the sequentially addition of 10 mM glucose, 1 µM oligomycin, and 50 mM 2-deoxy-glucose (2-DG) according to manufacturer's instructions.

statistical analyses
Comparisons between the different groups were performed by analysis of variance followed by two-tailed unpaired Student's t-test. The Kaplan-Meier method was used for building survival curves and differences were analyzed by the log-rank sum test. All analyses were performed using PRISM (GraphPad Software). P values were two-sided, and P < 0.05 was considered to indicate statistical significance. resUlTs sirT2 is highly expressed by Myeloid cells  macrophages and in the spleen, liver, and kidneys ( Figure 1B and data not shown). Western blot analyses confirmed SIRT2 protein expression in BMDMs (see below). Primary osteoblasts, osteoclasts, macrophages, and mast cells expressed 2.6-fold to 7.1-fold higher levels of SIRT2 mRNA than granulocytes, NK cells, T cells, B cells, DCs, and thymocytes ( Figure 1C).
Data are means ± SD of four animals per group expressed as the percentage of total sirT2 Deficiency has no Major impact on the Development of immune cells and host response to TssT-1 SIRT2 −/− mice were described previously (15). These mice were born at the expected Mendelian ratio and developed without abnormalities. SIRT2 mRNA and protein were undetectable in SIRT2 −/− BMDMs (Figures 1D,E). Expression levels of Sirt1 and Sirt3-7 mRNA were unaltered in SIRT2 −/− BMDMs (Figure 1F), suggesting that the lack of SIRT2 was not compensated by an increase in expression of other sirtuins. Compared to SIRT2 +/+ mice, SIRT2 −/− mice expressed normal proportions and absolute numbers of CD4/CD8 double negative (DN1-4), double positive, and single positive (SP) thymocytes ( Table 1). Additionally, SIRT2 −/− mice had normal populations of splenic T cells (DN, SP, naïve, and memory), B cells (immature and mature B cells), DCs (B220 − CD11c + cDCs and B220 + CD11c + pDCs), and Foxp3 + regulatory T cells ( Table 2). Therefore, SIRT2 had no major impact on immune cell development.

sirT2 Deficiency increases Phagocytosis by Macrophages
Macrophages are professional phagocytic cells that play a major role in antimicrobial host defenses. Therefore, we tested whether SIRT2 deficiency had an effect on phagocytosis by BMDMs. SIRT2 +/+ and SIRT2 −/− BMDMs were incubated with fluorescent beads and analyzed by flow cytometry (Figures 3A-C). A higher percentage of SIRT2 −/− than SIRT2 +/+ BMDMs phagocytosed beads (32.4 ± 1.9 vs 24.5 ± 1.2 percent positive cells, P = 0.002; Figures 3A,B), regardless of opsonization ( Figure 3C) BMDMs were additionally exposed to live E. coli, S. aureus and GBS for 1 h before measuring the number of intracellular bacteria by plating cell lysates and enumerating colonies. Confirming the results obtained using inert beads and fluorescent bacteria, the numbers of phagocytosed E. coli, S. aureus and GBS were 1.3-fold to 1.6-fold higher using SIRT2 −/− BMDMs ( Figure 3D).
sirT2 Deficiency Does not affect cytokine response of BMDMs exposed to Microbial ligands and sensitivity of Mice to endotoxemia Sensing of microbial ligands through TLRs initiates MAPK and NF-κB signaling involved in the control of cytokine gene expression by innate immune cells (2). To address whether SIRT2 impacted intracellular signaling, the phosphorylation of ERK1/2, p38, and JNK MAPKs in BMDMs exposed to LPS for 0, 10, 30, and 60 min was analyzed by western blotting. The rate of LPS-induced phosphorylation of ERK1/2, p38 and JNK was very similar in SIRT2 +/+ Impact of SIRT2 on Innate Immune Responses Frontiers in Immunology | www.frontiersin.org August 2017 | Volume 8 | Article 1037
To assess the relevance of these observations in vivo, we developed models of endotoxemia of different severity (Figures 5A-D). In a mild model of endotoxemia (induced by an i.p. challenge with 10 mg/kg LPS), TNF and IL-12p40 concentrations in blood and mortality rates (83 vs 100%, P = 0.3) were comparable in SIRT2 +/+ and SIRT2 −/− mice (Figures 5A,B). In a severe model of endotoxemia (induced by 25 mg/kg LPS), TNF, IL-6, and IL-12p40 concentrations in blood and mortality rates (88% in both groups, P = 0.69) were strongly increased, but remained similar in SIRT2 +/+ and SIRT2 −/− mice (Figures 5C,D). Furthermore, SIRT2 +/+ and SIRT2 −/− mice were equally sensitive to fulminant shock induced by TNF (25 mg/kg i.p. in d-galactosamine sensitized mice), the main driver of the lethal effect of endotoxemia (P = 0.6; Figure 5E). Overall, SIRT2 did not interfere with endotoxemia.

sirT2 Deficiency Protects from chronic staphylococcal infection
Considering that SIRT2 impacted phagocytosis but not cytokine expression, we hypothesized that SIRT2 deficiency should provide some benefit during chronic lethal infection but not fulminant sepsis, and should not sensitize to benign infection. Therefore, we compared the impact of SIRT2 deficiency during rapidly lethal, sub-lethal, and chronic bacterial infections induced by E. coli, K. pneumonia, and S. aureus, three of the most frequent causes of bacterial sepsis in humans (42).

DiscUssiOn
In the present study, we identified a unique role for SIRT2 in host-pathogen interactions. SIRT2 deficiency promoted bacterial phagocytosis by macrophages but not cytokine production. In agreement with these findings, SIRT2 deficiency protected from chronic staphylococcal infection while having no influence on the course of endotoxemia, TNF-induced shock, fulminant bacterial peritonitis, non-severe bacterial pneumonia, and chronic candidiasis. These observations are particularly relevant in light of the development of pharmacological inhibitors of SIRT2 for clinical applications (43), as they suggest that their usage would not increase susceptibility to bacterial and candidal infections.
SIRT2 was the most highly expressed sirtuin in myeloid cells. Macrophages expressed elevated levels of SIRT2, in accordance previous reports describing SIRT2 expression in microglial cells in vivo (16,17). Interestingly, SIRT2 deficiency had no major impact on LPS-induced MAPK activation, NF-κB nuclear translocation,
We analyzed host responses to extracellular bacteria most frequently isolated from septic patients. Strikingly, SIRT2 deficiency enhanced the engulfment of Gram-positive and Gram-negative bacteria by macrophages, an effect apparently unrelated with a differential expression of phagocytic receptors or microtubule polymerization. This observation was surprising considering that on the one hand SIRT2 deacetylates α-tubulin and destabilizes the microtubule network (8,41) and on the other end microtubule depolymerizating agents were reported to inhibit phagocytosis (75,76). Yet, the effects of microtubule depolymerizating agents were tested using an immortalized macrophage-like mouse cell line and human neutrophils that may behave differently than primary BMDMs. As a positive control (77), actin depolymerization efficiently inhibited phagocytosis by BMDMs. Increased glycolysis has been associated with efficient phagocytosis by macrophages (39,40). SIRT2 deficiency reduced HIF-1α deacetylation and destablization (78), and augmented glycolysis in human fibroblasts (79). We observed that SIRT2 deficiency increased glycolysis and that glycolysis inhibition reduced phagocytosis in BMBMs. Albeit preliminary, these results suggest that SIRT2 may influence phagocytosis through metabolic constraints.
The improved control of bacterial burden during chronic staphylococcal infection might be related to improved phagocytosis but also to enhanced autophagy in SIRT2 −/− mice. Autophagy facilitates the clearance of cytoplasmic bacteria (80) and has been involved in host defenses and tolerance to S. aureus infection (81,82). Hyperacetylation of tubulin stimulated autophagy upon nutrient deprivation, and SIRT2 deficiency increased autophagy in a colorectal cancer cell line (83,84). Thus, by regulating tubulin acetylation and metabolic activity, SIRT2 may contribute to modulate phagocytic and autophagy defense mechanisms, though the latter has not been formally demonstrated.
From a translational perspective, it was important to define the impact of SIRT2 in preclinical models of infection. A main observation of this study is that SIRT2 deficiency protected mice from chronic staphylococcal infection, while it neither protected nor sensitized mice to TNF-induced shock, endotoxemia, rapidly lethal E. coli peritonitis and mild K. pneumoniae pneumonia.
Additionally, SIRT2 deficiency did not influence the development of chronic candidiasis as it did for chronic staphylococcal infection. This may not be surprising considering differences in host-pathogen interactions during fungal and bacterial infections. Whether SIRT2 protects from other chronic bacterial infections should be tested in the future. Nonetheless, these results support the clinical development of SIRT2 inhibitors regarding their infection-related safety profile. This contrasts with inhibitors of HDAC1-11 that impaired innate immune defenses, increased susceptibility to infection in preclinical mouse models and have been associated with severe infections in patients (29,(85)(86)(87)(88)(89).
Overall, SIRT2 has a subtle impact on host defense responses to bacterial infections. Considering that sirtuins are intricately linked with metabolism, age-associated dysfunctions and lifespan, it will be important to analyze the impact of SIRT2 on host defenses under metabolic stress conditions and according to age. To conclude, our results are encouraging with respect to developing inhibitors of SIRT2, which are safe in terms of susceptibility to infections, for treating metabolic and neurodegenerative diseases, such as Parkinson's disease and Huntington's disease (90,91).