Visual Abstract

NAD+ is an essential cofactor in reduction-oxidation metabolism with impact on metabolic and inflammatory diseases. However, data elucidating the effects of NAD+ on the proinflammatory features of human primary monocytes are scarce. In this study, we explored how NAD+ affects TLR4 and NOD-like receptor with a PYD-domain 3 (NLRP3) inflammasome activation, two key innate immune responses. Human primary monocytes were isolated from buffy coats obtained from healthy individuals. Intracellular NAD+ was manipulated by nicotinamide riboside and the NAMPT inhibitor FK866. Cells were primed with LPS with or without subsequent NLRP3 activation with ATP or cholesterol crystals to analyze the effects of NAD+ levels on TLR4-mediated NF-κB activation and NLRP3 activity, respectively. Cytokine release was quantified, and the downstream signal pathway of TLR4 was investigated with Western blot and proteomic analysis. The impact of sirtuin and PARP inhibition was also explored. Our main findings were: 1) elevated NAD+ enhanced IL-1β release in LPS-primed human monocytes exposed to ATP in vitro, 2) both NLRP3-dependent and -independent inflammatory responses in LPS-exposed monocytes were inhibited by NAD+ depletion with FK866, 3) the inhibition was not caused by suppression of sirtuins or PARP1, and 4) phosphorylation of several proteins TLR4 signal pathway was inhibited by FK866-mediated NAD+ depletion, specifically TAK1, IKKβ, IkBα, MEK 1/2, ERK 1/2, and p38. Hence, we suggest a novel mechanism in which NAD+ affects TLR4 signal transduction. Furthermore, our data challenge previous reports of the interaction between NAD+ and inflammation and question the use of nicotinamide riboside in the therapy of inflammatory disorders.

Nicotinamide adenine dinucleotide (NAD+) is an essential cofactor in reduction-oxidation metabolism (1). More recent studies suggest that NAD+ also has a central role in protection against metabolic disturbances and inflammation. This is mediated through sirtuins, a family of NAD+-dependent deacetylases that maintain metabolic homeostasis, reduce cellular damage, and dampen inflammation (2). NAD+ is also a substrate for poly(ADP-ribose) polymerases (PARPs), which has been reported to favor proinflammatory processes (3). Because of a high cellular turnover, regeneration of NAD+ from nicotinamide (NAM) through the NAD+ salvage pathway is necessary to maintain ample cellular levels of NAD+. The key step in this process is the conversion of NAM to NAM mononucleotide (NMN) through the rate-limiting activity of NAM phosphoribosyltransferase (NAMPT) (4). Low cellular levels of NAD+, due to high consumption and/or insufficient regeneration, have been implicated in a range of biological and pathological processes such as aging, metabolic and inflammatory disorders like diabetes, fatty liver disease, and atherosclerosis, and in neurodegenerative disorders like Alzheimer disease (5). Moreover, treatments that enhance cellular NAD+ levels (i.e., NMN or nicotinamide riboside [NR] supplements) have been shown to mitigate such processes and disorders experimentally (6).

Monocytes are an essential part of the innate immune system. They stay in the blood circulation for surveillance and migrate to the insulted tissue during inflammation, maturing into macrophages to clear microbes, dead cells, and debris (7). However, monocytes are also central in the pathogenesis of several inflammatory disorders ranging from autoimmune disorders to atherosclerosis. Several of the pathological effects of monocytes are related to enhanced and persistent secretion of upstream inflammatory cytokines like TNF and IL-1, and anti-inflammatory therapies targeting these cytokines are now in clinical use (815).

NOD-like receptor with a PYD-domain 3 (NLRP3) inflammasomes are intracellular multiprotein complexes consisting of the receptor NLRP3, the scaffolding protein apoptosis-associated speck-like protein (ASC), and the effector protein caspase-1 (16). NLRP3 inflammasomes are tightly regulated and are activated in a two-step manner. First, a priming signal indicative of infection or tissue damage (signal 1) activates the proinflammatory transcription factor NF-κB, which increases NLRP3 and pro–IL-1β protein levels. TLR4, a pattern recognition receptor that can recognize LPS and a range of endogenous danger signals (17), is a classic mediator of signal 1. Then, a second signal (signal 2), indicative of cellular damage such as extracellular ATP or crystals precipitated from urate or cholesterol (18), induces an assembly of the inflammasome, leading to an autolytical processing of procaspase-1 to active caspase-1. Caspase-1 then cleaves pro–IL-1β into active IL-1β, which is subsequently secreted (16).

Intracellular NAD+ has been reported to negatively regulate NLRP3 inflammasome activation in mouse bone marrow–derived macrophages (19). It also seems that NAD+ precursors, such as NMN, NAM, and NR could have anti-inflammatory effects in various mouse disease models (2022). Another study has also shown that NR treatment could reduce inflammasome activation in human PBMC in vitro (23). However, data elucidating the effects of NAD+ on the proinflammatory features of human primary monocytes are scarce. The aim of this study was to investigate to what extent NAD+ affects TLR4 signaling and NLRP3 inflammasome activation in human monocytes.

Ultrapure LPS from Escherichia coli (O111:B4) was purchased from Invivogen, (Carlsbad, CA). NR was obtained from ChromaDex (Irvine, CA). FK866, sirtinol, EX5257, resveratrol, 3-aminobenzamide (3-ABA), buthionine sulfoximine (BSO), and ATP were all purchased from Sigma-Aldrich. Cholesterol crystals (CC) were kindly provided by professor T. Espevik, (Department of Clinical and Molecular Medicine, Norwegian University of Science and Technology, Trondheim, Norway). All Abs are listed in Table I.

Monocytes were isolated from peripheral PBMC, purified from freshly prepared buffy coat obtained from Oslo University Hospital blood bank by gradient centrifugation with Lymphoprep (STEMCELL Technologies, Cambridge, U.K.), and subsequently subjected to plastic adherence as previously described (24). The monocytes were seeded into culture dishes (Nunclon Delta surface; Thermo Fisher Scientific, Waltham, MA) at the density of 300,000 cells/ml and cultured in RPMI 1640 medium containing stable glutamine, 25 mM HEPES (Biowest), 10% heat-inactivated FBS, 5 U penicillin/ml, and 50 μg/ml streptomycin (Sigma-Aldrich). To manipulate NAD+ levels, cells were cultured with or without NR (500 μM) or FK866 (100 nM) for 16 h prior to LPS stimulation (10 ng/ml). For NLRP3 inflammasome activation, monocytes were incubated with ATP (3 mM) or CC (200 μg/ml, preserved in 0.05% human serum albumin/PBS, prepared as the described before) (25). For manipulation of sirtuins or PARP1, monocytes were incubated with sirtinol (10 μM), EX5257 (10 μM), resveratrol (10 μM), or 3-ABA (1 mM), respectively, for 16 h, then stimulated with LPS for 7 h prior to analysis. For glutathione depletion, primary human monocytes were incubated with 500 μM BSO for 16 h, then incubated with LPS (10 ng/ml) for 7 h prior to analysis. Cell-free supernatants and cell pellets were harvested and stored at −80°C until analyses.

After experimental interventions, cell culture medium was collected and centrifuged at 300 × g for 10 min. The supernatant was used to measure IL-1β, TNF, and IL-6 secretion by using DuoSet ELISA Kits (R&D Systems, Minneapolis, MN). The supernatant was also used to examine cytotoxicity by using ToxiLight BioAssay Kit (Lonza). The ATP was extracted from monocytes and measured by using the ENLITEN ATP Assay System (Promega). All assays were performed according to the manufacturer’s instructions.

NAD+ extraction and subsequent quantitative analyses were performed according to Yoshino and Imai method (26). Briefly, monocytes were incubated with 5 mM EDTA PBS solution for 20 min on ice for detachment. A total of 20% of cells were used to determine protein concentration for normalization. The remaining cells were lysed by 10% HClO4 followed by one-third volume of 3 M K2CO3. After centrifuging at 21,300 × g, the supernatant was subjected to HPLC using a 20 × 3.9 mm Sentry Guard column (Nova-Pak C18 bonded silica) connected to a 150 × 4.6-mm Atlantis T3 silica-based, reversed-phase C18 columns (Waters, Milford, MA). NAD+ was detected by UV detector, and UV absorbance was monitored at 260 nm. Elution of cellular NAD+ was verified and quantified by coelution with known amounts of NAD+ standard (Sigma-Aldrich). Results were normalized to total protein.

Total glutathione was quantified with the EnzyChrom GSH/GSSG Assay Kit (BioAssay Systems, Hayward, CA) with some modifications of the manufacturer’s protocol. After the cell samples were deproteinized with 5% metaphosphoric acid, the suspension was frozen in liquid nitrogen and thawed in a 37°C water bath twice and left for 5 min at 2–8°C. The extract was centrifuged at 10,000 × g for 10 min. A total of 5 μl of supernatant was added to 300 μl of 1 × assay buffer (provided in the kit). A total of 200 μl of this suspension was added to a 96-well plate together with 100 μl of working reagent (provided in the kit). The procedure was then continued according to the kit protocol.

RNA was extracted by using RNeasy Mini Kit (QIAGEN). cDNA was synthesized by using qScript cDNA SuperMix (Quantabio, Beverly, MA). Real-time RT-PCR was performed with Brilliant III Ultra-Fast SYBR Green QPCR Master Mix (Agilent Technologies, Santa Clara, CA) on 7900HT Fast Real-Time PCR System (Thermo Fisher Scientific). Primer sequences are listed in Table II.

Predesigned SIRT1 small interfering RNA (siRNA) (identification no. 136457) and Silencer Negative Control siRNA (catalog no. AM4611) was purchased from Ambion, Thermo Fisher Scientific. Monocytes were transfected with the Amaxa Human Monocyte Nucleofector Kit (Lonza, Alpharetta, GA) and the Nucleofector 2b Device (Lonza) in accordance with manufacturer’s protocol. Briefly, freshly isolated PBMCs from buffy coats (107 cells) were incubated with 30 pmol siRNA in Nucleofector Solution at room temperature in certified cuvettes and processed in the Nucleofector 2b Device. A total of 1 ml of prewarmed RPMI without FBS was then added to the cuvettes and the cells gently transferred to 12-well plates. After 1 h incubation (37°C, 5% CO2) nonadherent cells were gently removed and monocytes were cultured in 1 ml of RPMI with 10% FBS for 16 h (37°C, 5% CO2) prior to further interventions.

Total cellular protein was extracted by M-PER Mammalian Protein Extraction Reagent containing Halt Protease and Phosphatase Inhibitor, and the protein concentration was determined by Pierce BCA Protein Assay Kit (all from Thermo Fisher Scientific). Proteins were reduced and denatured, separated by SDS-PAGE, and transferred to PVDF membrane. Abs are listed in Table I. The membranes were developed by SuperSignal West Dura Extended Duration Substrate (Thermo Fisher Scientific). The images were captured by LAS-4000 (GE Healthcare) and quantified by Image Studio Lite (version 5.2; Li-Cor, Lincoln, NE).

Three biological replicates were used for both total proteome analysis and phosphoproteome analysis. Total cellular protein extraction was performed in the same way described in the section above. The proteins were precipitated with 4 vol of acetone in −20°C overnight. The precipitated proteins were dissolved with 6 M urea in 100 mM ammonium bicarbonate, reduced with DTT, and alkylated with iodoacetamide. For total proteome analysis, the proteins were in-solution digested by diluting the urea concentration to 1 M followed by digestion with trypsin overnight at 37°C. The resulting peptides were desalted and concentrated before mass spectrometry by the stop-and-go-extraction tip method using a C18 resin disk (3M Empore). For phosphoproteome analysis, the proteins were in-solution digested with LysC for 2 h in 37°C after which the urea concentration was reduced to 1 M by adding 100 mM ammonium bicarbonate buffer followed by digestion with trypsin overnight in 37°C. The phosphopeptides were enriched using TiO2 material (Titansphere TiO 10 μm; GL Science) as previously described (27).

Each peptide mixture was analyzed by a nEASY-LC coupled to QExactive Plus (Thermo Electron, Bremen, Germany) with EASY Spray PepMap RSLC column (C18, 2 μl, 100 Å, 75 μm × 50 cm). For total proteome samples 120-min liquid chromatography separation gradient and for phosphoproteome samples 60-min liquid chromatography separation gradient was used. The resulting mass spectrometry raw files were submitted to the MaxQuant software version 1.6.1.0 for protein identification and label-free quantification. Carbamidomethyl was set as a fixed modification, and acetyl (protein N-term), carbamyl (N-term), and oxidation were set as variable modifications. First-search peptide tolerance of 20 ppm and main-search error 4.5 ppm were used. Trypsin without proline restriction enzyme option was used, with two allowed miscleavages. The minimal unique+razor peptide number was set to 1, and the allowed false discovery rate was 0.01 (1%) for peptide and protein identification. Label-free quantitation was employed with default settings. The UniProt database with human entries (October 2017) was used for the database searches. Known contaminants as provided by MaxQuant and identified in the samples were excluded from further analysis, and Perseus software 1.6.1.3 was used for the statistical analysis of the total proteome MaxQuant results. The proteins with statistically significant expression-level differences between treated and control cells (i.e., LPS versus control) were submitted to Ingenuity Pathway Analysis (IPA) (QIAGEN Redwood City, https://www.qiagenbioinformatics.com/products/ingenuity-pathway-analysis/), which calculates the activities of upstream regulators in the form of activation Z-score (28). Phosphoproteome data were analyzed with MaxQuant using phosphorylation (STY) as variable modification activated.

The use of human monocytes was approved by the local ethical committee (regional ethics committee of Helse Sør-Øst; permit number S-05172) and conducted according to the ethical guidelines outlined in the World Medical Association’s Declaration of Helsinki for use of human tissue and subjects.

Perseus software 1.6.1.3 was used for the statistical analysis of total proteome data. GraphPad Prism 7.04 was used in the rest of the statistical analysis. A two-tailed paired t test was used to test for significant changes between two experimental conditions. For multiple comparisons, a one-way ANOVA was performed followed by Dunnett multiple comparisons test (when all other means were compared with the same mean) or Sidak multiple comparisons test (when a predetermined selection of means was compared). Which specific statistical tests performed are explained in the figure legends. A p value <0.05 was considered as statistically significant.

To investigate if intracellular NAD+ levels change during NLRP3 inflammasome activation, IL-1β secretion and NAD+ levels were measured at different time points after addition of ATP to LPS-primed human monocytes. IL-1β secretion increased profoundly after ATP was added, but there was no significant decrease in intracellular NAD+ levels (Fig. 1A). To investigate how elevated cellular NAD+ affected NLRP3 inflammasome activation, we next incubated cells with NR overnight (16 h) to boost intracellular NAD+, resulting in a 2-fold increase (Fig. 1B). We expected that the elevated intracellular NAD+ levels would attenuate inflammasome activation. In contrast, however, incubation with NR resulted in mildly increased ATP-induced IL-1β secretion from LPS-primed human monocytes (Fig. 1C). Moreover, pretreatment with NR also seemed to enhance IL-1β secretion when using CC as signal 2, although the increase did not reach statistical significance (Fig. 1D). Also, NR induced a moderate and significant increase in LPS-induced IL-1β mRNA levels with no effect on TNF transcription and secretion, a cytokine that is not regulated by NLRP3 inflammasome (Fig. 1E–G). Corresponding to elevated IL-1β mRNA and secretion, NR treatment also increased LPS-induced pro–IL-1β expression (Fig. 1H, 1I). Hence, increasing intracellular NAD+ by NR did not inhibit NLRP3 inflammasome activation but enhanced the ATP-triggered IL-1β secretion by increasing IL-1β transcription and translation in human monocytes in vitro.

FIGURE 1.

NLRP3 inflammasome activation does not reduce monocyte NAD+ levels. Increased NAD+ enhances IL-1β expression and release from human primary monocytes. (A) Monocytes isolated from buffy coats were primed by LPS (10 ng/ml) for 6 h, followed by ATP (3 mM) activation for 5/15/30/60 min (n = 4). IL-1β release was quantified with ELISA and levels normalized to 100% at 60 min. NAD+ and protein were isolated and measured by HPLC and BCA assay, respectively. NAD+ data were normalized to the corresponding amount of protein. (B) Monocytes were incubated with or without NR (500 μM) overnight (16 h), then primed with LPS (10 ng/ml) for 7 h. NAD+ and protein were isolated and measured by HPLC and BCA assay, respectively. NAD+ data were normalized to the corresponding amount of protein. (C and D) Monocytes were incubated with or without NR (500 μM) overnight (16 h), then primed with LPS (10 ng/ml) for 7 h. ATP was added after 6.5 h of LPS priming (C), and CC (200 μg/ml) was added after the first hour of LPS priming (D). Secreted IL-1β was quantified in conditioned media by ELISA (n = 9). (E) TNF release from monocytes incubated with or without NR (500 μM) overnight (16 h), then primed with LPS (10 ng/ml) for 7 h. (F and G) The corresponding gene transcription of pro–IL-1β and TNF was determined by real-time RT-PCR (n = 5). (H and I) Pro–IL-1β protein expression was quantified by Western blot and normalized to corresponding β-actin (n = 4). All columns are normalized to control = 100%. Columns are shown as mean ± SEM. *p < 0.05 versus control determined by two-tailed paired t test.

FIGURE 1.

NLRP3 inflammasome activation does not reduce monocyte NAD+ levels. Increased NAD+ enhances IL-1β expression and release from human primary monocytes. (A) Monocytes isolated from buffy coats were primed by LPS (10 ng/ml) for 6 h, followed by ATP (3 mM) activation for 5/15/30/60 min (n = 4). IL-1β release was quantified with ELISA and levels normalized to 100% at 60 min. NAD+ and protein were isolated and measured by HPLC and BCA assay, respectively. NAD+ data were normalized to the corresponding amount of protein. (B) Monocytes were incubated with or without NR (500 μM) overnight (16 h), then primed with LPS (10 ng/ml) for 7 h. NAD+ and protein were isolated and measured by HPLC and BCA assay, respectively. NAD+ data were normalized to the corresponding amount of protein. (C and D) Monocytes were incubated with or without NR (500 μM) overnight (16 h), then primed with LPS (10 ng/ml) for 7 h. ATP was added after 6.5 h of LPS priming (C), and CC (200 μg/ml) was added after the first hour of LPS priming (D). Secreted IL-1β was quantified in conditioned media by ELISA (n = 9). (E) TNF release from monocytes incubated with or without NR (500 μM) overnight (16 h), then primed with LPS (10 ng/ml) for 7 h. (F and G) The corresponding gene transcription of pro–IL-1β and TNF was determined by real-time RT-PCR (n = 5). (H and I) Pro–IL-1β protein expression was quantified by Western blot and normalized to corresponding β-actin (n = 4). All columns are normalized to control = 100%. Columns are shown as mean ± SEM. *p < 0.05 versus control determined by two-tailed paired t test.

Close modal
Table I.
Abs used
AbSource/CompanyApplication
Anti–β-actin (A5441) Sigma-Aldrich Western blot 
Anti-ERK1/2 (9102) Cell Signaling Western blot 
Anti–phospho-ERK1/2 (9101) Cell Signaling Western blot 
Anti-IκBα (4814) Cell Signaling Western blot 
Anti–phospho-IκBα (2859) Cell Signaling Western blot 
Anti–IL-1β (MAB601) R&S Systems Western blot 
Anti-IRAK1 (4504) Cell Signaling Western blot 
Anti–phospho-IRAK1 (SAB4504246) Sigma-Aldrich Western blot 
Anti-IRAK4 (4363) Cell Signaling Western blot 
Anti–phospho-IRAK4 (11,927) Cell Signaling Western blot 
Anti-MyD88 (D80F5) Cell Signaling Western blot 
Anti-SIRT1 (ab110304) Abcam Western blot 
Anti-TRAF6 (8028) Cell Signaling Western blot 
Anti-TLR4 (IMG-5031A) Novus Biologicals Western blot 
Anti-rabbit IgG (7074) Cell Signaling Western blot 
Anti-mouse IgG (7076) Cell Signaling Western blot 
AbSource/CompanyApplication
Anti–β-actin (A5441) Sigma-Aldrich Western blot 
Anti-ERK1/2 (9102) Cell Signaling Western blot 
Anti–phospho-ERK1/2 (9101) Cell Signaling Western blot 
Anti-IκBα (4814) Cell Signaling Western blot 
Anti–phospho-IκBα (2859) Cell Signaling Western blot 
Anti–IL-1β (MAB601) R&S Systems Western blot 
Anti-IRAK1 (4504) Cell Signaling Western blot 
Anti–phospho-IRAK1 (SAB4504246) Sigma-Aldrich Western blot 
Anti-IRAK4 (4363) Cell Signaling Western blot 
Anti–phospho-IRAK4 (11,927) Cell Signaling Western blot 
Anti-MyD88 (D80F5) Cell Signaling Western blot 
Anti-SIRT1 (ab110304) Abcam Western blot 
Anti-TRAF6 (8028) Cell Signaling Western blot 
Anti-TLR4 (IMG-5031A) Novus Biologicals Western blot 
Anti-rabbit IgG (7074) Cell Signaling Western blot 
Anti-mouse IgG (7076) Cell Signaling Western blot 
Table II.
Primer sequences
TargetSequenceAcc.nr
18S-forward 5′-CGGCTACCACATCCAAGGAA-3′ NR_003286 
18S-reverse 5′-GCTGGAATTACCGCGGCT-3′  
IL-1β–forward 5′-CCCTAAACAGATGAAGTGCTCCTT-3′ NM_000576 
IL-1β–reverse 5′-GGTGGTCGGAGATTCGTAGCT-3′  
TNF-forward 5′-CCAGGCAGTCAGATCATCTTCTC-3′ M10988 
TNF-reverse 5′-GGAGCTGCCCCTCAGCTT-3′  
IL-6–forward 5′-AGCCCTGAGAAAGGAGACATGTA-3′ M14584 
IL-6–reverse 5′-CATCTTTGGAAGGTTCAGGTTGT-3′  
TargetSequenceAcc.nr
18S-forward 5′-CGGCTACCACATCCAAGGAA-3′ NR_003286 
18S-reverse 5′-GCTGGAATTACCGCGGCT-3′  
IL-1β–forward 5′-CCCTAAACAGATGAAGTGCTCCTT-3′ NM_000576 
IL-1β–reverse 5′-GGTGGTCGGAGATTCGTAGCT-3′  
TNF-forward 5′-CCAGGCAGTCAGATCATCTTCTC-3′ M10988 
TNF-reverse 5′-GGAGCTGCCCCTCAGCTT-3′  
IL-6–forward 5′-AGCCCTGAGAAAGGAGACATGTA-3′ M14584 
IL-6–reverse 5′-CATCTTTGGAAGGTTCAGGTTGT-3′  

To further investigate if intracellular NAD+ levels affect the inflammatory response mediated through TLR4 and NLRP3 inflammasome activation, human monocytes were pretreated overnight with the NAMPT inhibitor FK866 with or without NR (Fig. 2, Supplemental Fig. 1). NAMPT is the rate-limiting enzyme in the salvage pathway for NAD+ generation, and FK866 treatment profoundly reduced intracellular NAD+ levels as well as LPS-induced synthesis and secretion of IL-1β, TNF, and IL-6 (Fig. 2A–H). NR restored NAD+ levels in FK866-treated cells in a dose-dependent manner (Supplemental Fig. 1A), and 500 μM NR was the most efficient dose for restoring cytokine synthesis (i.e., IL-1β, TNF, and IL-6) and release with no extra gain from higher doses (Supplemental Fig. 1B–D). Furthermore, 500 μM NR significantly restored NAD+ levels, improved mRNA synthesis, and almost completely rescued IL-1β, TNF, and IL-6 secretion from FK866-treated cells (Fig. 2B–H). Importantly, although FK866 markedly reduced cellular NAD+ levels, cell viability, as determined by adenylate kinase release, and cellular levels of ATP were not affected, indicating that the monocytes were still healthy (Fig. 2I, 2J). Because NAD+ reduction might affect global reduction-oxidation status, and thereby also potentially the activation of NF-κB, we also investigated whether glutathione (the most important and abundant intracellular antioxidant) scavenging with BSO could reproduce the effects of FK866. However, although BSO reduced glutathione levels by >60% (Supplemental Fig. 2A), this did not translate into any effect on LPS-induced cytokine synthesis or release (Supplemental Fig. 2B, 2C). In summary, these findings suggest that intracellular NAD+ is essential for the LPS-induced inflammatory response in human primary monocytes involving both NLRP3-dependent and -independent cytokines.

FIGURE 2.

Intracellular NAD+ is essential for the inflammatory response to LPS in human primary monocytes. (A) Monocytes from buffy coats were incubated with FK866 (100 nM) with or without NR (500 μM) overnight (16 h), and intracellular NAD+ and protein were quantified with HPLC and BCA assay, respectively. (B and C) Monocytes treated as in (A) were then primed with LPS (10 ng/ml) for 7 h and subsequently activated with CC (200 μg/ml) after the first hour of LPS priming (B) or ATP after 6.5 h of LPS priming (C), and IL-1β was quantified in the conditioned media with ELISA. (DH) Monocytes were treated as in (A) and subsequently primed with LPS (10 ng/ml) for 7 h. TNF and IL-6 were quantified in the conditioned media with ELISA, and corresponding gene transcriptions of IL-1β, TNF, and IL-6 were determined by real-time PCR. (I) Cell viability was determined right before LPS priming by adenylate kinase release. Relative luminescence is shown, higher luminescence indicates lower cell viability. (J) Intracellular ATP was measured right before LPS priming with ATP assay kit. Columns are mean ± SEM of three to five independent biological repeats. **p < 0.01, ***p < 0.001, ****p < 0.0001 versus control and #p < 0.05, ##p < 0.01, ###p < 0.001, ####p < 0.0001 versus FK866-treated monocytes. All p values were determined by one-way ANOVA with the Greenhouse–Geisser correction and Fisher least significant difference.

FIGURE 2.

Intracellular NAD+ is essential for the inflammatory response to LPS in human primary monocytes. (A) Monocytes from buffy coats were incubated with FK866 (100 nM) with or without NR (500 μM) overnight (16 h), and intracellular NAD+ and protein were quantified with HPLC and BCA assay, respectively. (B and C) Monocytes treated as in (A) were then primed with LPS (10 ng/ml) for 7 h and subsequently activated with CC (200 μg/ml) after the first hour of LPS priming (B) or ATP after 6.5 h of LPS priming (C), and IL-1β was quantified in the conditioned media with ELISA. (DH) Monocytes were treated as in (A) and subsequently primed with LPS (10 ng/ml) for 7 h. TNF and IL-6 were quantified in the conditioned media with ELISA, and corresponding gene transcriptions of IL-1β, TNF, and IL-6 were determined by real-time PCR. (I) Cell viability was determined right before LPS priming by adenylate kinase release. Relative luminescence is shown, higher luminescence indicates lower cell viability. (J) Intracellular ATP was measured right before LPS priming with ATP assay kit. Columns are mean ± SEM of three to five independent biological repeats. **p < 0.01, ***p < 0.001, ****p < 0.0001 versus control and #p < 0.05, ##p < 0.01, ###p < 0.001, ####p < 0.0001 versus FK866-treated monocytes. All p values were determined by one-way ANOVA with the Greenhouse–Geisser correction and Fisher least significant difference.

Close modal

Sirtuins are NAD+-dependent enzymes involved in inflammation (2, 19). To examine whether sirtuins are involved in the FK866-mediated inhibition of LPS-induced cytokine release, human monocytes were treated with EX527 (SIRT1-3 and SIRT6 inhibitor) (29, 30), sirtinol (SIRT1-2 inhibitor) (31), and resveratrol (SIRT1 agonist) (32) in the presence or absence of FK866 and/or NR overnight (16 h), then stimulated with LPS for 7 h (Fig. 3A, 3B, Supplemental Fig. 2D, 2E). However, neither EX527, sirtinol, nor resveratrol could reproduce or counter the effect of FK866. Surprisingly, the SIRT1 agonist resveratrol repeatedly showed a trend toward enhancing TNF and IL-6 release from cells either not treated with FK866 or treated with FK866 together with NR. However, this was not statistically significant when corrected for multiple comparisons, and the data should therefore be interpreted with caution (Fig. 3A, 3B). To investigate the effect of sirtuins on NLRP3 inflammasome activation, monocytes were incubated with or without FK866 and/or NR overnight (16 h), then primed with LPS for 6 h, then incubated with EX527, sirtinol, or resveratrol for 30 min prior to NLRP3 activation with ATP for 30 min. IL-1β was quantified in conditioned media with ELISA (Fig. 3C). Again, EX527, sirtinol, and resveratrol could not reproduce or counter the effects of FK866. Finally, SIRT1 knockdown in primary human monocytes with siRNA, resulting in an average reduction in SIRT1 protein levels of 39% (33–46%, n = 4), had simply no effect on LPS-induced TNF, IL-6, or IL-1β release (Fig. 3D–H). Thus, our data suggest that the inhibitory effect of FK866 on TLR4- and NLRP3-mediated cytokine release is independent of the sirtuins SIRT1-3 and SIRT6, although SIRT1 activation may potentially favor LPS-mediated TNF and IL-6 release. NAD+ is also as substrate for PARPs. In particular, PARP1 has been linked to TLR4-mediated NF-κB activation (33). Hence, we also explored the effect of the PARP1 inhibitor 3-ABA (Fig. 3I, 3J). However, 3-ABA had no effect on LPS-mediated TNF or IL-6 release, suggesting that the anti-inflammatory effect of NAD+ depletion is not mediated through reduced PARP1 activity.

FIGURE 3.

Sirtuin inhibition does not suppress the LPS- or NLRP3-induced inflammatory response. (A and B) Monocytes isolated from buffy coats were incubated with EX527 (10 μM), sirtinol (10 μM), or resveratrol (10 μM) overnight (16 h), then primed by LPS (10 ng/ml) for 7 h. TNF and IL-6 were quantified in conditioned media with ELISA. (C) Monocytes were primed with LPS (10 ng/ml) for 6 h before incubation with EX527 (10 μM), sirtinol (10 μM), or resveratrol (10 μM) for 30 min, then activated with 3 mM ATP for 30 min. IL-1β was quantified in conditioned media with ELISA. (DH) SIRT1 protein levels in monocytes were reduced with siRNA (D and E), the cells primed with LPS (10 ng/ml) for 7 h, and cytokines quantified in the conditioned media and compared with scramble siRNA–treated control cells (E–G). For IL-1β, monocytes were activated with 3 mM ATP after 6.5 h of LPS priming. (I and J) Monocytes were treated with 3-ABA (1 mM) overnight (16 h), then stimulated with LPS (10 ng/ml) for 7 h. TNF and IL-6 were quantified in the conditioned media. Columns are mean ± SEM of five (A–C) or four (D–G) biological repeats. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001 versus control. The p values are determined by one-way ANOVA with the Greenhouse–Geisser correction and Dunnet multiple comparison test (A–C) or two-tailed paired t test (D–G).

FIGURE 3.

Sirtuin inhibition does not suppress the LPS- or NLRP3-induced inflammatory response. (A and B) Monocytes isolated from buffy coats were incubated with EX527 (10 μM), sirtinol (10 μM), or resveratrol (10 μM) overnight (16 h), then primed by LPS (10 ng/ml) for 7 h. TNF and IL-6 were quantified in conditioned media with ELISA. (C) Monocytes were primed with LPS (10 ng/ml) for 6 h before incubation with EX527 (10 μM), sirtinol (10 μM), or resveratrol (10 μM) for 30 min, then activated with 3 mM ATP for 30 min. IL-1β was quantified in conditioned media with ELISA. (DH) SIRT1 protein levels in monocytes were reduced with siRNA (D and E), the cells primed with LPS (10 ng/ml) for 7 h, and cytokines quantified in the conditioned media and compared with scramble siRNA–treated control cells (E–G). For IL-1β, monocytes were activated with 3 mM ATP after 6.5 h of LPS priming. (I and J) Monocytes were treated with 3-ABA (1 mM) overnight (16 h), then stimulated with LPS (10 ng/ml) for 7 h. TNF and IL-6 were quantified in the conditioned media. Columns are mean ± SEM of five (A–C) or four (D–G) biological repeats. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001 versus control. The p values are determined by one-way ANOVA with the Greenhouse–Geisser correction and Dunnet multiple comparison test (A–C) or two-tailed paired t test (D–G).

Close modal

Manipulating sirtuin function did not reproduce our findings under low cellular NAD+ levels. We therefore next performed label-free quantitative proteome analysis of monocytes to gain global information of changes in protein expression in human monocytes stimulated with NR only, LPS only, LPS + FK866, or LPS + FK866 + NR, compared with unstimulated control cells. In total, 4445 different proteins were identified (Supplemental Table I). After two-tailed paired t test (p < 0.05) analysis, 146 proteins with significant expression changes were found in the LPS versus control group, 113 proteins in LPS+FK866 versus control group, and 68 proteins in LPS+FK866+NR versus control group. Based on these proteins and their expression changes, we used IPA software to find their upstream regulators and calculate their probable activities. From the predicted upstream regulators, we then compiled a list of proteins known to be involved in the TLR4 signaling pathway (Fig. 4). Two known TLR4-induced second-wave cytokines, IFN-β and TNF, as well as some proteins in their respective intracellular pathways, were also found. The activation Z-scores of these potential regulators are shown in the heat map in Fig. 4. The result suggests a general inactivation of LPS-induced inflammatory regulators in FK866-treated monocytes. NR alone had no effect on these identified upstream regulators (Fig. 4, left heat map column), predicting no effect of NR on the TLR4 signaling pathway independently of FK866 and LPS. However, NR partly restored the predicted activities inhibited by FK866 (Fig. 4, right heat map column).

FIGURE 4.

Low intracellular NAD+ induces proteomic changes in LPS-treated monocytes that predict reduced TLR4 signal transduction. (A) Monocytes from buffy coats were incubated with FK866 (100 nM) with or without NR (500 μM) overnight (16 h), then primed with LPS (10 ng/ml) for 7 h (n = 3). The total protein was isolated for label-free quantitative proteomics analysis. Proteins with significant expression-level change as compared with nontreated control cells were submitted to the IPA software to predict upstream regulators, which might cause these changes. The activation Z-scores of upstream regulators known to be involved in TLR4 signaling are shown as a heat map. Two known TLR4-induced second-wave cytokines, IFN-β and TNF, as well as proteins in their respective pathways, were also found. High Z-score indicates elevated activity and vice versa. (B) The predicted upstream regulators in (A) are highlighted in this illustration of the TLR4 signal pathway. This illustration is based on previously published reports (44, 45). Of note, the MyD88-independent pathway is depicted on the plasma membrane for the sake of simplicity, although this pathway is thought to be mediated from early endosomes after TLR4 internalization. Second-wave cytokines (IFN-β and TNF) and their respective pathways are not included.

FIGURE 4.

Low intracellular NAD+ induces proteomic changes in LPS-treated monocytes that predict reduced TLR4 signal transduction. (A) Monocytes from buffy coats were incubated with FK866 (100 nM) with or without NR (500 μM) overnight (16 h), then primed with LPS (10 ng/ml) for 7 h (n = 3). The total protein was isolated for label-free quantitative proteomics analysis. Proteins with significant expression-level change as compared with nontreated control cells were submitted to the IPA software to predict upstream regulators, which might cause these changes. The activation Z-scores of upstream regulators known to be involved in TLR4 signaling are shown as a heat map. Two known TLR4-induced second-wave cytokines, IFN-β and TNF, as well as proteins in their respective pathways, were also found. High Z-score indicates elevated activity and vice versa. (B) The predicted upstream regulators in (A) are highlighted in this illustration of the TLR4 signal pathway. This illustration is based on previously published reports (44, 45). Of note, the MyD88-independent pathway is depicted on the plasma membrane for the sake of simplicity, although this pathway is thought to be mediated from early endosomes after TLR4 internalization. Second-wave cytokines (IFN-β and TNF) and their respective pathways are not included.

Close modal

Although the activation patterns predicted from global proteome analysis in Fig. 4, further support that the inflammatory response is NAD+-dependent, the underlying mechanism was still unclear. Phosphorylation is the main mode of signal transduction in the cascade. Thus, we investigated LPS-induced phosphorylation of ERK1/2 and IκBα, central downstream components of the TLR4 signaling pathway, and observed markedly attenuated phosphorylation in monocytes pretreated with FK866 (Fig. 5A–C). Moreover, the peak levels of phosphorylation appeared later in FK866-treated cells. This indicates that the TLR4 priming signal cascade was attenuated and delayed under conditions with low cellular NAD+. Further supporting this, FK866 inhibited IκBα phosphorylation in a dose-dependent manner (Fig. 5F, 5G). Interestingly, ERK1/2 phosphorylation was suppressed with only 10 nM FK866 (Fig. 5D, 5E), whereas at least 100 nM was needed to obtain maximal suppression of IκBα phosphorylation. Finally, we investigated whether NR had the opposite effect of FK866. However, no significant changes nor suggestive patterns of change in ERK1/2 or IκBα phosphorylation were seen in NR-treated cells (Fig. 5H–J).

FIGURE 5.

FK866 suppresses the phosphorylation of ERK1/2 and IκBα. No significant effect of NR. (AC) Monocytes from buffy coats were incubated with or without FK866 (100 nM) overnight (16 h), then primed with LPS (10 ng/ml) for 5, 15, 30, 45, or 60 min. Levels of phosphorylated ERK1/2, IκBα, and their corresponding unphosphorylated form were determined by immunoblot analysis (n = 3). Data are normalized to corresponding β-actin, then normalized to the highest amplitude set as 100% (LPS 15 minutes for p-ERK1/2, LPS 45 min for p-IκBα). (DG) FK866 dose responses. Monocytes were incubated with or without FK866 in the indicated doses, then incubated with LPS (10 ng/ml) for 15 min for ERK1/2 phosphorylation analysis (D and E) and 45 min for IκBα phosphorylation analysis (F and G). (HJ) Monocytes were incubated with or without NR (500 μM) overnight (16 h), then primed with LPS (10 ng/ml) for 5, 15, 30, 45, or 60 min. Levels of phosphorylated ERK1/2, IκBα, and their corresponding unphosphorylated form were determined by immunoblot analysis (n = 3). All data are shown as mean ± SEM of three biological repeats. *p < 0.05 versus corresponding time point without FK866 (Sidak multiple comparisons test).

FIGURE 5.

FK866 suppresses the phosphorylation of ERK1/2 and IκBα. No significant effect of NR. (AC) Monocytes from buffy coats were incubated with or without FK866 (100 nM) overnight (16 h), then primed with LPS (10 ng/ml) for 5, 15, 30, 45, or 60 min. Levels of phosphorylated ERK1/2, IκBα, and their corresponding unphosphorylated form were determined by immunoblot analysis (n = 3). Data are normalized to corresponding β-actin, then normalized to the highest amplitude set as 100% (LPS 15 minutes for p-ERK1/2, LPS 45 min for p-IκBα). (DG) FK866 dose responses. Monocytes were incubated with or without FK866 in the indicated doses, then incubated with LPS (10 ng/ml) for 15 min for ERK1/2 phosphorylation analysis (D and E) and 45 min for IκBα phosphorylation analysis (F and G). (HJ) Monocytes were incubated with or without NR (500 μM) overnight (16 h), then primed with LPS (10 ng/ml) for 5, 15, 30, 45, or 60 min. Levels of phosphorylated ERK1/2, IκBα, and their corresponding unphosphorylated form were determined by immunoblot analysis (n = 3). All data are shown as mean ± SEM of three biological repeats. *p < 0.05 versus corresponding time point without FK866 (Sidak multiple comparisons test).

Close modal

We next performed a mass spectrometry–based phosphoproteomics analysis to identify the proteins regulated by phosphorylation (Supplemental Table I). The analysis showed that low intracellular NAD+ suppressed LPS-induced phosphorylation of several proteins involved in the TLR4 signaling pathway, including TAK1, NEMO, MEK1/2, ERK2, and p38 (Fig. 6A, Supplemental Table I). IRAK1 and IRAK4 were not detected in the phosphoproteomic analysis and were instead analyzed with Western blot. A trend of reduced LPS-induced phosphorylation of IRAK1 and IRAK4 in FK866-treated cells compared with control cells were seen, although this was not statistically significant (Fig. 6B–D). Furthermore, label-free quantitative proteome analysis showed that NAD+ depletion with FK866 did not alter the quantitative expression of the proteins in the TLR4 signal pathway (Supplemental Fig. 3A, Supplemental Table I). TLR4 and MyD88 were not reliably quantified with the proteomic analysis, however, and were quantified with Western blot analysis together with IRAK1, IRAK4, and TRAF6 (Supplemental Fig. 3B–G). Taken together, our data show that NAD+ depletion suppresses the TLR4 signal transduction by inhibiting the phosphorylation of several key mediators without affecting the protein expression of the components in the signal pathway (summarized in Fig. 7).

FIGURE 6.

Low intracellular NAD+ inhibits phosphorylation of several key proteins involved in the TLR4 signaling pathway. (A) Monocytes from buffy coat were incubated with FK866 (100 nM) overnight (16 h), then primed with LPS (10 ng/ml) for 15 min (n = 3). The total protein was isolated and phosphoproteomic analysis performed. Phosphorylated proteins known to be involved in the TLR4 signaling pathway were selected, and the ratios of LPS+FK866–treated cells to LPS-treated cells were calculated from their phosphopeptide signal intensity quantified in MaxQuant software (PhosphoSTY sites-sheet in Supplemental Table I). The phosphorylation ratios from each experiment, as well as the mean (right column), are shown in the heat map. A ratio of 0 means that the phosphorylated protein was detected only in LPS-treated samples. N/A means the phosphorylated protein was not identified in neither LPS+FK866–treated nor LPS-treated samples. (BD) Monocytes from buffy coat were incubated with FK866 (100 nM) overnight (16 h), then primed with LPS (10 ng/ml) for 15 min (n = 4). Phosphorylated IRAK1 and IRAK4 were quantified with Western blot analysis (C and D). Columns are mean with SEM. LPS-treated cells were compared with LPS+FK866–treated cells with a two-tailed paired t test.

FIGURE 6.

Low intracellular NAD+ inhibits phosphorylation of several key proteins involved in the TLR4 signaling pathway. (A) Monocytes from buffy coat were incubated with FK866 (100 nM) overnight (16 h), then primed with LPS (10 ng/ml) for 15 min (n = 3). The total protein was isolated and phosphoproteomic analysis performed. Phosphorylated proteins known to be involved in the TLR4 signaling pathway were selected, and the ratios of LPS+FK866–treated cells to LPS-treated cells were calculated from their phosphopeptide signal intensity quantified in MaxQuant software (PhosphoSTY sites-sheet in Supplemental Table I). The phosphorylation ratios from each experiment, as well as the mean (right column), are shown in the heat map. A ratio of 0 means that the phosphorylated protein was detected only in LPS-treated samples. N/A means the phosphorylated protein was not identified in neither LPS+FK866–treated nor LPS-treated samples. (BD) Monocytes from buffy coat were incubated with FK866 (100 nM) overnight (16 h), then primed with LPS (10 ng/ml) for 15 min (n = 4). Phosphorylated IRAK1 and IRAK4 were quantified with Western blot analysis (C and D). Columns are mean with SEM. LPS-treated cells were compared with LPS+FK866–treated cells with a two-tailed paired t test.

Close modal
FIGURE 7.

Summary: NAD+ depletion attenuates the TLR4 signal pathway by inhibiting phosphorylation of several proteins in the MyD88-dependent signaling cascade. (A) In human monocytes, NR increases intracellular NAD+ levels, which favors LPS-induced synthesis of pro–IL-1β. In contrast, FK866-mediated depletion of NAD+ effectively inhibits LPS-induced TLR4 signaling, resulting in reduced synthesis of proinflammatory cytokines in general. (B) In more detail, NAD+ depletion attenuates phosphorylation of several proteins in the MyD88-dependent signaling cascade. The affected proteins are highlighted in red color. This cartoon depicting the TLR4 signal pathway is based on previous reports (44, 45).

FIGURE 7.

Summary: NAD+ depletion attenuates the TLR4 signal pathway by inhibiting phosphorylation of several proteins in the MyD88-dependent signaling cascade. (A) In human monocytes, NR increases intracellular NAD+ levels, which favors LPS-induced synthesis of pro–IL-1β. In contrast, FK866-mediated depletion of NAD+ effectively inhibits LPS-induced TLR4 signaling, resulting in reduced synthesis of proinflammatory cytokines in general. (B) In more detail, NAD+ depletion attenuates phosphorylation of several proteins in the MyD88-dependent signaling cascade. The affected proteins are highlighted in red color. This cartoon depicting the TLR4 signal pathway is based on previous reports (44, 45).

Close modal

In this study, we showed that whereas NR increased NAD+ levels in human monocytes, it enhanced rather than attenuated IL-1β transcription and release in LPS-primed monocytes activated with ATP. In contrast, reducing NAD+ levels by the NAMPT inhibitor FK866 markedly decreased the release of both NLRP3-dependent (IL-1β) and -independent cytokines (TNF and IL-6) from LPS-primed monocytes. These suppressive effects were reversed by NR-induced NAD+ generation and seem not to involve sirtuins or PARP1. However, NAD+ depletion profoundly inhibited the TLR4 signaling pathway. Our data may question the role of NR supplementation as a therapeutic option in human inflammatory disease.

Previously, NAD+ has been reported to negatively correlate with IL-1β secretion through reduced deacetylation of microtubulin in mouse bone marrow–derived macrophages (19). Moreover, elevated NAD+ has been shown to compromise IL-1β and TNF release in human PBMCs (23). Surprisingly, in our study, elevated intracellular NAD+ by NR treatment enhanced IL-1β transcription, synthesis, and secretion from LPS-primed monocytes activated with ATP. In contrast, diminishing intracellular NAD+ by the NAMPT inhibitor FK866 markedly inhibited both NLRP3-dependent (IL-1β) and NLRP3-independent (TNF) cytokine release in LPS-stimulated monocytes. Notably, this inhibition was rescued by refurnishing intracellular NAD+ by NR treatment. This strongly suggests that the effects of FK866 are NAD+ dependent. The reasons for these apparently discrepant results are at present not clear, but some explanation may exist. First, our findings are also supported by previous studies. In 2011, TLR4 stimulation was reported to upregulate NAMPT expression and subsequently promote cellular NAD+ synthesis (34). Moreover, elevated NAMPT and NAD+ have been suggested to promote intestinal inflammation (35). This indicates that NAD+ is essential in the TLR4-induced inflammatory response. Hence, boosting cellular NAD+ with NR might just fuel the inflammatory response. Second, studies in mouse models may not necessarily mimic the situation in human monocytes. Last but not least, most experiments addressing the effect of NAD+ levels on TNF and IL-1β release were performed in the human monocytic THP-1 cell line, differentiated into macrophages by PMA and PBMC, and dominated by nonmonocytic cells (23). These cells may differ from primary human monocytes. Finally, although the anti-inflammatory effects of FK866 were reversed by NR, we cannot exclude that some of the effects of FK866 could be mediated by NAMPT through mechanisms not related to NAD+ generation (36).

Of note, although our data suggest that NR favors LPS-induced pro–IL-1β synthesis and effectively opposes the inhibitory effect of FK866-mediated NAD+ depletion on TLR4 signaling, NR alone did not activate TLR4 or other TLRs in our experiments, as no TNF or IL-6 could be detected in supernatants from monocytes incubated with NR alone overnight. Thus, the effect of NR on TLR4 signaling is indirect, and its opposing effect on FK866 strongly suggests that the proinflammatory effect of NR is mediated through increased intracellular NAD+ levels.

Sirtuins have been shown to be an important link between NAD+ and inflammation (2, 3739). In our study, however, the inhibitory effect of low intracellular NAD+ levels on LPS-induced inflammatory responses seems not to be linked to sirtuins, as neither sirtuin inhibitors (EX527, Sirtinol) nor agonist treatment (i.e., resveratrol, SIRT1 agonist) blunted the LPS-induced cytokine production, nor countered the effect of FK866. Interestingly, resveratrol is reported as an anti-inflammatory agent in vitro (40) and it can inhibit LPS-induced inflammation in RAW 264.7 and HEK293 cell lines (41), but in this study, we observed a trend suggesting that this SIRT1 agonist may actually enhance TNF and IL-6 secretion in LPS-exposed human monocytes (Fig. 3A, 3B). However, these trends were not statistically significant when corrected for multiple analyses, which may be due to low sample size and thus a false-negative statement. The possibly opposite effect of resveratrol in human primary monocytes as compared with RAW 264.7 and HEK293 cell lines needs further investigation. We also performed a siRNA-mediated SIRT1 knockdown in primary human monocytes, which had no effect on TNF, IL-6, or IL-1β secretion. Although a statistically significant protein knockdown of 33–46% was observed, this may not be enough to observe any SIRT1-mediated effects, and lack of complete knockdown weakens these findings. However, the observed effect of SIRT1 knockdown is in line with the observed effects of two pharmacological SIRT1 inhibitors (EX5257 and sirtinol). In summary, our data support that pharmacological SIRT1 activation may further enhance LPS-induced TNF and IL-6 release when NAD+ is available. However, our data do not support that inhibition of sirtuins is mediating the attenuation of LPS/TLR4 signaling seen with NAD+ depletion.

Because PARPs also are known to consume NAD, we explored whether the PARP inhibitor 3-ABA could mimic NAD+ depletion. PARP1 has been linked to innate immune responses in some studies. Of note, in 1999, PARP1 knockout mice were reported to be resistant to LPS-induced shock (33). In the same study, reduced serum TNF levels were observed in LPS-treated PARP1 knockout mice compared with wild-type mice, whereas serum IL-6 was not significantly changed. 3-ABA is a competitive PARP1 inhibitor (42) that has been shown to inhibit PARP1 activity in cell culture at 20 μM (43). However, in the current study, 1 mM 3-ABA had no significant effect on neither TNF nor IL-6 release from LPS-treated human monocytes. The effect of short term (16-h) PARP1 inhibition in vitro may not mimic lifelong PARP1 knockout in vivo. Nonetheless, our data do not support that the effect of NAD+ depletion is mediated through PARP1 inhibition.

To generate a new hypothesis about the underlying mechanisms, we next performed label-free quantitative proteome analysis comparing the effects of NAD+ depletion on the proteome of LPS-treated cells. Based on the observed proteome changes, we identified several upstream regulators that could cause these changes. The patterns clearly suggested that NAD+ depletion had an inhibitory effect on TLR4 signaling, possibly at the level of several proteins in the signal cascade. Furthermore, this pattern could, for the most part, be restored with NR supplementation. Indeed, our subsequent phosphoproteomics and Western blot analyses showed that several key regulators in the LPS-induced TLR4 signaling pathway had reduced or no phosphorylation in the presence of FK866, indicating a blunted signal transduction due to low intracellular NAD+.

In conclusion, our study shows that the LPS-induced inflammatory response in human primary monocytes is downregulated by NAD+ depletion. This attenuating effect was seen for both NLRP3-dependent and NLRP3-independent cytokines and seems to be mediated through inhibition of several proteins in the TLR4 signaling pathway. Our data challenge previous reports of the interaction between NAD+ and inflammation, and the discrepancies mentioned above should be clarified before designing treatment protocols targeting regulation of intracellular NAD+ levels in inflammatory disorders.

We thank Azita Rashidi for excellent technical assistance.

This work was supported by the Norwegian Research Council (Grant 240099/F20), the Helse Sør-Øst Regional Health Authority of Norway (Grant 2018084), and the National Association for Public Health (Grant 1444). Funding sources were not involved in analysis/study design, nor in writing of the manuscript.

The online version of this article contains supplemental material.

Abbreviations used in this article:

3-ABA

3-aminobenzamide

BSO

buthionine sulfoximine

CC

cholesterol crystal

IPA

Ingenuity Pathway Analysis

NAM

nicotinamide

NAMPT

NAM phosphoribosyltransferase

NLRP3

NOD-like receptor with a PYD-domain 3

NMN

NAM mononucleotide

NR

nicotinamide riboside

PARP

poly(ADP-ribose) polymerase

siRNA

small interfering RNA.

1
Cantó
,
C.
,
K. J.
Menzies
,
J.
Auwerx
.
2015
.
NAD(+) metabolism and the control of energy homeostasis: a balancing act between mitochondria and the nucleus.
Cell Metab.
22
:
31
53
.
2
Vachharajani
,
V. T.
,
T.
Liu
,
X.
Wang
,
J. J.
Hoth
,
B. K.
Yoza
,
C. E.
McCall
.
2016
.
Sirtuins link inflammation and metabolism.
J. Immunol. Res.
2016
:
8167273
.
3
Brady
,
P. N.
,
A.
Goel
,
M. A.
Johnson
.
2018
.
Poly(ADP-ribose) polymerases in host-pathogen interactions, inflammation, and immunity. [Published erratum appears in 2019 Microbiol. Mol. Biol. Rev. DOI: 10.1128/MMBR.00010-19.]
Microbiol. Mol. Biol. Rev.
DOI: 10.1128/MMBR.00038-18.
4
Garten
,
A.
,
S.
Schuster
,
M.
Penke
,
T.
Gorski
,
T.
de Giorgis
,
W.
Kiess
.
2015
.
Physiological and pathophysiological roles of NAMPT and NAD metabolism.
Nat. Rev. Endocrinol.
11
:
535
546
.
5
Johnson
,
S.
,
S. I.
Imai
.
2018
.
NAD + biosynthesis, aging, and disease.
F1000 Res.
7
:
132
.
6
Yoshino
,
J.
,
J. A.
Baur
,
S. I.
Imai
.
2018
.
NAD+ intermediates: the biology and therapeutic potential of NMN and NR.
Cell Metab.
27
:
513
528
.
7
Jakubzick
,
C. V.
,
G. J.
Randolph
,
P. M.
Henson
.
2017
.
Monocyte differentiation and antigen-presenting functions.
Nat. Rev. Immunol.
17
:
349
362
.
8
Ridker
,
P. M.
,
T. F.
Lüscher
.
2014
.
Anti-inflammatory therapies for cardiovascular disease.
Eur. Heart J.
35
:
1782
1791
.
9
Wang
,
J.
,
L.
Tan
,
H. F.
Wang
,
C. C.
Tan
,
X. F.
Meng
,
C.
Wang
,
S. W.
Tang
,
J. T.
Yu
.
2015
.
Anti-inflammatory drugs and risk of Alzheimer’s disease: an updated systematic review and meta-analysis.
J. Alzheimers Dis.
44
:
385
396
.
10
Machowska
,
A.
,
J. J.
Carrero
,
B.
Lindholm
,
P.
Stenvinkel
.
2016
.
Therapeutics targeting persistent inflammation in chronic kidney disease.
Transl. Res.
167
:
204
213
.
11
Bua
,
S.
,
L.
Di Cesare Mannelli
,
D.
Vullo
,
C.
Ghelardini
,
G.
Bartolucci
,
A.
Scozzafava
,
C. T.
Supuran
,
F.
Carta
.
2017
.
Design and synthesis of novel nonsteroidal anti-inflammatory drugs and carbonic anhydrase inhibitors hybrids (NSAIDs-CAIs) for the treatment of rheumatoid arthritis.
J. Med. Chem.
60
:
1159
1170
.
12
Ridker
,
P. M.
,
B. M.
Everett
,
T.
Thuren
,
J. G.
MacFadyen
,
W. H.
Chang
,
C.
Ballantyne
,
F.
Fonseca
,
J.
Nicolau
,
W.
Koenig
,
S. D.
Anker
, et al
CANTOS Trial Group
.
2017
.
Antiinflammatory therapy with canakinumab for atherosclerotic disease.
N. Engl. J. Med.
377
:
1119
1131
.
13
Chou
,
R. C.
,
M.
Kane
,
S.
Ghimire
,
S.
Gautam
,
J.
Gui
.
2016
.
Treatment for rheumatoid arthritis and risk of Alzheimer’s disease: a nested case-control analysis.
CNS Drugs
30
:
1111
1120
.
14
Stankovic Stojanovic
,
K.
,
Y.
Delmas
,
P. U.
Torres
,
J.
Peltier
,
G.
Pelle
,
I.
Jéru
,
M.
Colombat
,
G.
Grateau
.
2012
.
Dramatic beneficial effect of interleukin-1 inhibitor treatment in patients with familial Mediterranean fever complicated with amyloidosis and renal failure.
Nephrol. Dial. Transplant.
27
:
1898
1901
.
15
Bresnihan
,
B.
,
R.
Newmark
,
S.
Robbins
,
H. K.
Genant
.
2004
.
Effects of anakinra monotherapy on joint damage in patients with rheumatoid arthritis. Extension of a 24-week randomized, placebo-controlled trial.
J. Rheumatol.
31
:
1103
1111
.
16
Mangan
,
M. S. J.
,
E. J.
Olhava
,
W. R.
Roush
,
H. M.
Seidel
,
G. D.
Glick
,
E.
Latz
.
2018
.
Targeting the NLRP3 inflammasome in inflammatory diseases. [Published erratum appears in 2018 Nat. Rev. Drug Discov. 17: 688.]
Nat. Rev. Drug Discov.
17
:
588
606
.
17
De Nardo
,
D.
2015
.
Toll-like receptors: activation, signalling and transcriptional modulation.
Cytokine
74
:
181
189
.
18
Walsh
,
J. G.
,
D. A.
Muruve
,
C.
Power
.
2014
.
Inflammasomes in the CNS.
Nat. Rev. Neurosci.
15
:
84
97
.
19
Misawa
,
T.
,
M.
Takahama
,
T.
Kozaki
,
H.
Lee
,
J.
Zou
,
T.
Saitoh
,
S.
Akira
.
2013
.
Microtubule-driven spatial arrangement of mitochondria promotes activation of the NLRP3 inflammasome.
Nat. Immunol.
14
:
454
460
.
20
Gomes
,
A. P.
,
N. L.
Price
,
A. J.
Ling
,
J. J.
Moslehi
,
M. K.
Montgomery
,
L.
Rajman
,
J. P.
White
,
J. S.
Teodoro
,
C. D.
Wrann
,
B. P.
Hubbard
, et al
.
2013
.
Declining NAD(+) induces a pseudohypoxic state disrupting nuclear-mitochondrial communication during aging.
Cell
155
:
1624
1638
.
21
Fang
,
E. F.
,
H.
Kassahun
,
D. L.
Croteau
,
M.
Scheibye-Knudsen
,
K.
Marosi
,
H.
Lu
,
R. A.
Shamanna
,
S.
Kalyanasundaram
,
R. C.
Bollineni
,
M. A.
Wilson
, et al
.
2016
.
NAD+ replenishment improves lifespan and healthspan in ataxia telangiectasia models via mitophagy and DNA repair.
Cell Metab.
24
:
566
581
.
22
Kaneko
,
S.
,
J.
Wang
,
M.
Kaneko
,
G.
Yiu
,
J. M.
Hurrell
,
T.
Chitnis
,
S. J.
Khoury
,
Z.
He
.
2006
.
Protecting axonal degeneration by increasing nicotinamide adenine dinucleotide levels in experimental autoimmune encephalomyelitis models.
J. Neurosci.
26
:
9794
9804
.
23
Traba
,
J.
,
M.
Kwarteng-Siaw
,
T. C.
Okoli
,
J.
Li
,
R. D.
Huffstutler
,
A.
Bray
,
M. A.
Waclawiw
,
K.
Han
,
M.
Pelletier
,
A. A.
Sauve
, et al
.
2015
.
Fasting and refeeding differentially regulate NLRP3 inflammasome activation in human subjects.
J. Clin. Invest.
125
:
4592
4600
.
24
Bennett
,
S.
,
S. N.
Breit
.
1994
.
Variables in the isolation and culture of human monocytes that are of particular relevance to studies of HIV.
J. Leukoc. Biol.
56
:
236
240
.
25
Samstad
,
E. O.
,
N.
Niyonzima
,
S.
Nymo
,
M. H.
Aune
,
L.
Ryan
,
S. S.
Bakke
,
K. T.
Lappegård
,
O. L.
Brekke
,
J. D.
Lambris
,
J. K.
Damås
, et al
.
2014
.
Cholesterol crystals induce complement-dependent inflammasome activation and cytokine release.
J. Immunol.
192
:
2837
2845
.
26
Yoshino
,
J.
,
S.
Imai
.
2013
.
Accurate measurement of nicotinamide adenine dinucleotide (NAD+) with high-performance liquid chromatography.
Methods Mol. Biol.
1077
:
203
215
.
27
Jersie-Christensen
,
R. R.
,
A.
Sultan
,
J. V.
Olsen
.
2016
.
Simple and reproducible sample preparation for single-shot phosphoproteomics with high sensitivity.
Methods Mol. Biol.
1355
:
251
260
.
28
Krämer
,
A.
,
J.
Green
,
J.
Pollard
Jr.
,
S.
Tugendreich
.
2014
.
Causal analysis approaches in ingenuity pathway analysis.
Bioinformatics
30
:
523
530
.
29
Napper
,
A. D.
,
J.
Hixon
,
T.
McDonagh
,
K.
Keavey
,
J. F.
Pons
,
J.
Barker
,
W. T.
Yau
,
P.
Amouzegh
,
A.
Flegg
,
E.
Hamelin
, et al
.
2005
.
Discovery of indoles as potent and selective inhibitors of the deacetylase SIRT1.
J. Med. Chem.
48
:
8045
8054
.
30
Kokkonen
,
P.
,
M.
Rahnasto-Rilla
,
P.
Mellini
,
E.
Jarho
,
M.
Lahtela-Kakkonen
,
T.
Kokkola
.
2014
.
Studying SIRT6 regulation using H3K56 based substrate and small molecules.
Eur. J. Pharm. Sci.
63
:
71
76
.
31
Zhang
,
M.
,
Y.
Pan
,
R. G.
Dorfman
,
Y.
Yin
,
Q.
Zhou
,
S.
Huang
,
J.
Liu
,
S.
Zhao
.
2017
.
Sirtinol promotes PEPCK1 degradation and inhibits gluconeogenesis by inhibiting deacetylase SIRT2.
Sci. Rep.
7
:
7
.
32
Chao
,
S. C.
,
Y. J.
Chen
,
K. H.
Huang
,
K. L.
Kuo
,
T. H.
Yang
,
K. Y.
Huang
,
C. C.
Wang
,
C. H.
Tang
,
R. S.
Yang
,
S. H.
Liu
.
2017
.
Induction of sirtuin-1 signaling by resveratrol induces human chondrosarcoma cell apoptosis and exhibits antitumor activity.
Sci. Rep.
7
:
3180
.
33
Oliver
,
F. J.
,
J.
Ménissier-de Murcia
,
C.
Nacci
,
P.
Decker
,
R.
Andriantsitohaina
,
S.
Muller
,
G.
de la Rubia
,
J. C.
Stoclet
,
G.
de Murcia
.
1999
.
Resistance to endotoxic shock as a consequence of defective NF-kappaB activation in poly (ADP-ribose) polymerase-1 deficient mice.
EMBO J.
18
:
4446
4454
.
34
Liu
,
T. F.
,
B. K.
Yoza
,
M.
El Gazzar
,
V. T.
Vachharajani
,
C. E.
McCall
.
2011
.
NAD+-dependent SIRT1 deacetylase participates in epigenetic reprogramming during endotoxin tolerance.
J. Biol. Chem.
286
:
9856
9864
.
35
Gerner
,
R. R.
,
V.
Klepsch
,
S.
Macheiner
,
K.
Arnhard
,
T. E.
Adolph
,
C.
Grander
,
V.
Wieser
,
A.
Pfister
,
P.
Moser
,
N.
Hermann-Kleiter
, et al
.
2018
.
NAD metabolism fuels human and mouse intestinal inflammation.
Gut
67
:
1813
1823
.
36
Chen
,
H.
,
S.
Wang
,
H.
Zhang
,
E. C.
Nice
,
C.
Huang
.
2016
.
Nicotinamide phosphoribosyltransferase (Nampt) in carcinogenesis: new clinical opportunities.
Expert Rev. Anticancer Ther.
16
:
827
838
.
37
Imai
,
S.
,
L.
Guarente
.
2014
.
NAD+ and sirtuins in aging and disease.
Trends Cell Biol.
24
:
464
471
.
38
Yang
,
S. R.
,
J.
Wright
,
M.
Bauter
,
K.
Seweryniak
,
A.
Kode
,
I.
Rahman
.
2007
.
Sirtuin regulates cigarette smoke-induced proinflammatory mediator release via RelA/p65 NF-kappaB in macrophages in vitro and in rat lungs in vivo: implications for chronic inflammation and aging.
Am. J. Physiol. Lung Cell. Mol. Physiol.
292
:
L567
L576
.
39
Brown
,
K. D.
,
S.
Maqsood
,
J. Y.
Huang
,
Y.
Pan
,
W.
Harkcom
,
W.
Li
,
A.
Sauve
,
E.
Verdin
,
S. R.
Jaffrey
.
2014
.
Activation of SIRT3 by the NAD+ precursor nicotinamide riboside protects from noise-induced hearing loss.
Cell Metab.
20
:
1059
1068
.
40
Poulsen
,
M. M.
,
K.
Fjeldborg
,
M. J.
Ornstrup
,
T. N.
Kjær
,
M. K.
Nøhr
,
S. B.
Pedersen
.
2015
.
Resveratrol and inflammation: challenges in translating pre-clinical findings to improved patient outcomes.
Biochim. Biophys. Acta
1852
:
1124
1136
.
41
Kim
,
M. H.
,
D. S.
Yoo
,
S. Y.
Lee
,
S. E.
Byeon
,
Y. G.
Lee
,
T.
Min
,
H. S.
Rho
,
M. H.
Rhee
,
J.
Lee
,
J. Y.
Cho
.
2011
.
The TRIF/TBK1/IRF-3 activation pathway is the primary inhibitory target of resveratrol, contributing to its broad-spectrum anti-inflammatory effects.
Pharmazie
66
:
293
300
.
42
Purnell
,
M. R.
,
W. J.
Whish
.
1980
.
Novel inhibitors of poly(ADP-ribose) synthetase.
Biochem. J.
185
:
775
777
.
43
Wang
,
H.
,
C.
Lu
,
Y.
Tan
,
J.
Xie
,
J.
Jiang
.
2014
.
Effect of adriamycin on BRCA1 and PARP-1 expression in MCF-7 breast cancer cells.
Int. J. Clin. Exp. Pathol.
7
:
5909
5915
.
44
Gay
,
N. J.
,
M. F.
Symmons
,
M.
Gangloff
,
C. E.
Bryant
.
2014
.
Assembly and localization of Toll-like receptor signalling complexes.
Nat. Rev. Immunol.
14
:
546
558
.
45
Menon
,
M. B.
,
M.
Gaestel
.
2016
.
TPL2 meets p38MAPK: emergence of a novel positive feedback loop in inflammation.
Biochem. J.
473
:
2995
2999
.

The authors have no financial conflicts of interest.

Supplementary data