Sirtuin 5 Deficiency Does Not Compromise Innate Immune Responses to Bacterial Infections

Sirtuin 5 (SIRT5) is a member of the family of NAD+-dependent lysine/histone deacetylases. SIRT5 resides mainly in the mitochondria where it catalyzes deacetylation, demalonylation, desuccinylation, and deglutarylation of lysine to regulate metabolic and oxidative stress response pathways. Pharmacologic inhibitors of SIRT5 are under development for oncologic conditions, but nothing is known about the impact of SIRT5 on antimicrobial innate immune defenses. Using SIRT5 knockout mice, we show that SIRT5 deficiency does not affect immune cell development, cytokine production and proliferation by macrophages and splenocytes exposed to microbial and immunological stimuli. Moreover, preclinical models suggest that SIRT5 deficiency does not worsen endotoxemia, Klebsiella pneumoniae and Streptococcus pneumoniae pneumonia, Escherichia coli peritonitis, listeriosis, and staphylococcal infection. Altogether, these data support the safety profile in terms of susceptibility to infections of SIRT5 inhibitors under development.


INTRODUCTION
Innate immune cells such as monocytes/macrophages, neutrophils and dendritic cells (DCs) express pattern recognition receptors (PRRs) that mediate the sensing of microbial associated molecular patterns (such as endotoxin, lipoproteins, peptidoglycans, glucans, mannans, and nucleic acids) and danger-associated molecular patterns released by injured or infected cells. PRRs encompass Toll-like receptors (TLRs), C-type lectin receptors, NOD-like receptors, RIG-I-like receptors, and cytosolic DNA sensors (1,2). Upon ligand sensing, PRRs initiate intracellular signaling cascades remodeling host transcriptome to promote cytokine/chemokine production and the development of antimicrobial effector mechanisms. Innate immune responses have to be tightly regulated to avoid imbalanced life-threatening immune responses.
The impact of SIRT5 on antimicrobial host defenses is unknown; which is an important missing piece considering the clinical development of SIRT5 inhibitors. Using SIRT5 knockout mice, we show that SIRT5 deficiency has no major impact on immune cell development and on the response of macrophages and splenocytes to microbial stimulation. Going well along with these observations, preclinical models revealed that SIRT5 knockout mice are not particularly sensitive to endotoxemia, Klebsiella pneumoniae and Streptococcus pneumoniae pneumonia, Escherichia coli peritonitis, listeriosis and staphylococcal infection. Up to now, these data support the assumption that SIRT5 inhibitors should not increase patients' susceptibility to infections.

Ethics Statement
Animal experimentation was approved by the Service de la Consommation et des Affaires Vétérinaires of Canton de Vaud (Epalinges, Switzerland) under authorizations n • VD 3287, 876.8, 876.9, 877.8, and 877.9 and performed according to Swiss and ARRIVE guidelines.

Mice, Cells and Reagents
Experiments were performed using 8 to 12-week-old C57BL/6J mice (Charles River Laboratories, Saint-Germain-sur-l'Arbresle, France) and SIRT5 knockout mice (kindly provided by Prof Johan Auwerx, Ecole Polytechnique Fédérale de Lausanne, Lausanne, Switzerland) backcrossed 7 times on a C57BL/6J background (23). Mice were housed (12 h light/dark cycle, 22 • C, 70% humidity) under specific pathogen-free conditions in the animal facility of the Centre des Laboratoires d'Epalinges (CLE, Epalinges, Switzerland, license number VD-H04). Colonies were free of mouse norovirus and mouse hepatitis virus infections. Mice were fed with γ-irradiated food (Global Rodent XP 18, Provimi Kliba AG, Kaiseraugst, Switzerland) and water ad libitum. Mice were transferred in a BSL2 unit to perform in vivo models of infection.

Flow Cytometry Analyses
Single cell suspensions from thymus and spleen were enumerated and incubated with 2.4G2 monoclonal antibody (mAb) (30). Cells were stained using mAbs listed in Table S1. Data were acquired using a LSR II flow cytometer (BD Biosciences) and analyzed using FlowJo Version 10.2 software (FlowJo LLC, Ashland OR) (31).

Metabolic Activity Measurements
The oxygen consumption rate (OCR, in pmole O 2 /minute) and the extracellular acidification rate (ECAR, in mpH/minute) were analyzed using a 96-well format Seahorse XFe R system, the Seahorse XF Cell Mito Stress Test Kit and the Seahorse XF Glycolysis Stress Test Kit (Agilent Technologies, Santa Clara, CA). Four × 10 4 BMDMs were plated in 96-well plates in complete IMDM medium. The next day, cells were rested one hour in Seahorse medium with or without glucose. Mitochondrial respiration was assessed by measuring OCR following the addition of 1 µM oligomycin (OM), 1 µM FCCP and 2 µM antimycinA/1 µM rotenone (AA/Rot). Glycolytic function was assessed by measuring ECAR following the addition of 10 mM glucose, 1 µM oligomycin and 50 mM 2-deoxy-glucose (2-DG).

RNA Analyses
Total RNA was isolated, reverse transcribed (RNeasy and QuantiTect reverse transcription kits, Qiagen, Hilden, Germany) and used in real-time PCR using Fast SYBR R Green Master Mix and a QuantStudio TM 12K Flex system (Life Technologies, Carlsbad, CA) as reported (24,37). Samples were tested in triplicate. Gene specific expression was normalized to actin expression. Primers are listed in Table S2. Sirt5 mRNA expression levels in organs were extracted from the BioGPS resource (http://biogps.org).

Statistical Analyses
Comparisons between different groups were performed by analysis of variance followed by two-tailed unpaired Student's t-test. In vivo bacteria and cytokine data were analyzed using the Mann-Whitney test. Survival curves were built using the Kaplan-Meier method 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.
To address further whether SIRT5 deficiency affected the response of immune cells, the proliferation of SIRT5 +/+ and SIRT5 −/− splenocytes exposed to LPS, CpG, Pam 3 CSK 4 , TSST-1, and anti-CD3/CD28 was assessed by 3 H-thymidine incorporation, (Figure 4A), while the production of IL-2 by SIRT5 +/+ and SIRT5 −/− splenocytes exposed to TSST-1, anti-CD3/CD28 and PMA plus ionomycin was measured by ELISA ( Figure 4B). Neither proliferation nor IL-2 production was modified by SIRT5 deficiency. Altogether, the results argued against an important role of SIRT5 in controlling cytokine production by macrophages exposed to TLR ligands and the response of splenocytes to microbial and immune stimuli.  Figure 5A). Accordingly, the mortality rates of SIRT5 +/+ and SIRT5 −/− mice were not significantly different (75% vs. 90%; P = 0.4; Figure 5B). To mimic clinical situations, we then explored the impact of SIRT5-deficiency on host defenses in models of infections induced by challenging mice with K. pneumoniae and S. pneumoniae i.n., E. coli i.p. and L. monocytogenes and S. aureus i.v.

DISCUSSION
This is the first report of the impact of SIRT5 on antimicrobial host defenses. In vitro studies using macrophages and splenocytes and preclinical models of endotoxemia and Gram-positive and Gram-negative bacteria infections suggest that SIRT5 deficiency has no major impact on antibacterial defenses. These observations are particularly relevant in light of the development of pharmacological inhibitors of SIRT5 for clinical applications (42). FIGURE 3 | SIRT5 deficiency does not affect macrophage response to microbial stimulation. SIRT5 +/+ and SIRT5 −/− BMDMs were exposed for 1 h to LPS (10 ng/ml), Pam 3 CSK 4 (10 ng/ml) and CpG (2 µg/ml). Gene expression levels were quantified by RT-PCR, normalized to actin levels, and expressed relative to SIRT5 +/+ control set at one. Data are means ± SD of triplicate samples from one experiment performed with three mice. P > 0.05 for all conditions.   (43,44). In SIRT1-deficient mice, CD4 + , CD8 + , and CD4 + CD8 + thymic subpopulations were normal but highly sensitive to DNAdamaging ionizing radiation (45). Circulating T cell, B cell and monocyte counts were normal in 5 months old SIRT7 −/− mice that developed inflammatory cardiomyopathy (46). SIRT6deficent mice developed, after 2 weeks of life, a progeroid syndrome associated with decreased lymphocyte counts in thymus and spleen. However, lymphocyte flaw was not cellintrinsic but linked to systemic defects (47). Overall sirtuins do not seem to affect the development of adaptive immune cells. Yet, SIRT1 was reported to influence T-helper (Th) 2, Th9, Th17 and T-regulatory (Treg) responses and SIRT3 to sustain the suppressive function of Tregs (48)(49)(50)(51)(52). Thus, it would be interesting to define whether SIRT5 shapes T cell responses.
SIRT5 influences diverse metabolic pathways in cardiac and hepatic cells, including urea cycle, amino acid metabolism, the TCA cycle, FAO, glycolysis and oxidative stress response (7)(8)(9)(10)(11)(12)(13)(14)(15)(16). SIRT5 −/− BMDMs showed a modest increased mitochondrial activity and decreased glycolytic activity, suggesting that SIRT5 may be less influential in macrophages than in heart and liver (8,15). Correlatively, SIRT5 mRNA and protein were expressed at much lower levels in immune organs [our data and (9)] and in primary macrophages than in liver and heart (5 and 10fold less SIRT5 mRNA in BMDMs than in liver and heart, respectively). SIRT5 deficiency had no major impact on LPS-induced cytokine production by macrophages and circulating TNF and IL-6 levels were similar in SIRT5 +/+ and SIRT5 −/− endotoxemic mice. SIRT5-deficient mice under high fat diet, a condition inducing inflammation and oxidative stress, showed normal metabolic parameters and signs of inflammation attested by Tnf, Cd68 (a monocyte/macrophage marker) and Cd36 (a scavenger receptor) gene expression in the liver (23). Two recent studies analyzed the impact of SIRT5 deficiency on mouse macrophage response to LPS, leading to opposite conclusions. SIRT5 −/− peritoneal macrophages produced reduced levels of TNF, IL-6, and MCP-1 (monocyte chemoattractant protein-1/CCL2). SIRT5 competed with SIRT2 to interact with NF-κB p65. Since SIRT2 deacetylates p65 to inhibit its transduction activity, SIRT5 indirectly promoted p65 acetylation and activity (53). In sharp contrast, LPS-stimulated SIRT5 −/− BMDMs expressed increased levels of Tnf, Il1b, and Il6 mRNA but not Il10 mRNA. SIRT5 desuccinylated PKM2 (pyruvate kinase M2), promoting tetramer-to-dimer transition and inhibiting pyruvate kinase activity of PKM2. In that study, SIRT5 deficiency protected from DSS-induced colitis. The inconsistency of the impact of SIRT5 on inflammatory responses echoes those reported for SIRT1, SIRT2, SIRT3, and SIRT6 in vitro and in vivo [discussed in (43,44)]. Differences in experimental conditions (BMDMs vs. peritoneal macrophages, germline vs. cell-type specific gene knockout, use of si/shRNA and pharmacological modulators of sirtuins) and subtle variations in qualitative and quantitative caloric input and NAD + availability may explain these differences. Additionally, the length of stimulation and the doses of stimulus [10 ng/ml of ultra-pure LPS here vs. 100 ng/ml of crude LPS in (17) and (53)] may have affected the results. It should also be stressed that SIRT5 deficiency was obtained by disruption of exon 4 in the case of the mice used in this study (23), while exons 2-5 were deleted in the SIRT5 knockout mice available from the Jackson Laboratory used in other studies (17,53). Nonetheless, even in these studies, the background of the animals may have differed substantially considering that commercial knockout mice are of 85% 129 and 15% C57BL/6 backgrounds and that mice were backcrossed 10 times on a BL/6J background in one study (17) while SIRT5 +/+ and SIRT5 −/− littermates were derived from the SIRT5 +/− heterozygote mice in the other study (53). Of note, all broad screening proteomic analyses identified metabolic pathways as the most targeted pathways by SIRT5, while pathways commonly associated with immune/inflammatory responses (such as NF-kB, interferon-response, cytokine, cell migration and inflammation pathways) were not evidenced (8,11,12,15,16).
Endotoxemia reflects pathological situations such as fulminant meningococcemia characterized by high blood loads of endotoxin, but does not reproduce the complex hostpathogen interactions generally taking place during bacterial infections. Therefore, we sought to define the impact of SIRT5 in preclinical models of infections mimicking common clinical situations. SIRT5 deficiency did not sensitize mice to severe S. pneumoniae pneumonia, rapidly lethal E. coli peritonitis, listeriosis and staphylococcal infection. In the most stringent models, SIRT5 deficiency did not protect from lethal infection, as foreseen if SIRT5 would amplify cytokine response. SIRT5 deficiency also did not render mice particularly susceptible to bacterial infections as suggested by the results obtained using models of sub-lethal/mild infection with K. pneumoniae, E. coli, L. monocytogenes, and S. aureus. Considering the diversity of the agents (Grampositive and Gram-negative and intracellular and extracellular bacteria) and of the routes of infection tested (i.n., i.p. and i.v.), these results so far support the assumption that SIRT5 has no dramatic influence on host defenses against bacterial infections and the clinical development of SIRT5 inhibitors for oncologic purposes (18,22). This contrasts with inhibitors of HDAC1-11 which impaired innate immune defenses against infections in mouse models and have been associated with episodes of severe infection when infused into cancer patients (37,(54)(55)(56)(57)(58). Further work will be required to test the efficacy of potential SIRT5 inhibitors (19)(20)(21) in models of cancer (18,22) and of infections and sepsis, then to define whether these inhibitors may predispose to infections in the setting of comorbidities, e.g., in elderly patients and patients with chronic inflammatory disorders like for example colitis and diabetes mellitus. Overall, SIRT5 does not worsen host defenses to bacterial infections under the conditions tested here. Since sirtuins are linked to metabolism, age-associated dysfunctions and lifespan, it would be of interest to investigate the role of SIRT5 under metabolic stress conditions and in older mice. To conclude, our results support the development of SIRT5 inhibitors for clinical purposes, as they suggest that these drugs would not increase patients' susceptibility to infections.

AUTHOR CONTRIBUTIONS
TH, EC, CT, AP, and DLR performed in vitro experiments. JH participated to flow cytometry analyses. TH, EC, CT, and DLR performed in vivo experiments. TR conceived the project, designed the experiments and wrote the paper. All the authors revised the paper.

ACKNOWLEDGMENTS
We thank Filipe Pé-Curto for technical help.
Table S1 | Antibodies used for flow cytometry analyses.