To study the kinetics of TNF release in endotoxemia following VNS, we subjected rats to vagus nerve stimulation (VNS) for 60 s at 10 Hz, or sham surgery. The rats were then injected intraperitoneally with bacterial endotoxin (lipopolysaccharide, LPS) at various time points after VNS. Serum TNF levels were significantly lower in VNS-treated rats compared to sham treated animals when endotoxin was administered 0, 2, 24, and 48 h after stimulation (Figure 1A). The VNS-mediated inhibition of endotoxin-induced TNF was not significant at 72 h after nerve stimulation.

Next, mice were treated with 60 s of VNS at 10 Hz, or sham surgery, and subjected to endotoxemia 24 h after treatment. Endotoxin-induced TNF levels were significantly lower in VNS- treated mice as compared to sham treated animals (Figure 1B). We previously observed that a single pulse, equivalent to 0.1 s stimulation, is sufficient to significantly reduce TNF release in experimental endotoxemia (8). To investigate if a single pulse confers a sustained reduction in TNF release, mice treated with 0.1 s of VNS were subjected to endotoxemia, 24 h after treatment. Endotoxin-induced TNF levels were significantly lower in VNS treated mice compared to sham treated animals (Figure 1B).

Together, these results suggest that a short activation of the inflammatory reflex reduces TNF production during endotoxemia for ≥24 h.

Prior work established that the α7 nicotinic acetylcholine receptor subunit (α7nAChR) expressed by macrophages is required for the functional integrity of the inflammatory reflex (14, 15, 26). To study the role of α7nAChR in the persistent inhibition by VNS of TNF release in endotoxemia, we next studied α7nAChR deficient and wild type mice. Endotoxin-induced TNF was significantly reduced at 24 h after VNS in wild type mice (Figure 1C, left panel), but not in α7nAChR deficient mice (Figure 1C, right panel). Thus, α7nAChR is required for the persistent inhibition of TNF release by transient electrical stimulation of the vagus nerve.

α7nAChR subunits form ion channels on neuronal cells (20). Although ion flux has been observed in macrophages exposed to cholinergic agonists, this effect is reportedly unlikely to affect cytokine production (21).

In neurons, α7nAChR can interact with adenylyl cyclase 6 (AC6) and promote cAMP formation (27, 28). To investigate the possibility that a similar interaction occurs in macrophages, bone marrow derived macrophages from α7nAChR ^+/+ and α7nAChR ^−/− mice, respectively, were exposed in vitro to the selective α7nAChR agonist, N-(3R)-1-Azabicyclo[2.2.2]oct - 3-yl - 2,3 - dihydro - 1,4-benzodioxin - 6 - carboxamide fumarate (PHA568487) (29). Addition of PHA568487 significantly increased cAMP levels in α7nAChR ^+/+, but not in α7nAChR ^−/− derived macrophages (Figure 2A), indicating that that α7nAChR signaling regulates cAMP in murine macrophages.

To study the mechanism of cholinergic control of TNF release we utilized macrophage-like RAW 264.7 cells. First, we investigated whether cholinergic activation confers prolonged inhibition of endotoxin-induced TNF release in RAW 264.7 cells. Cells were exposed to 10 μM ACh or vehicle for 1 h, and then washed and re-suspended in fresh medium. Cells were then rested for 24 h and subsequently exposed to endotoxin for 4 h. TNF levels in culture medium were significantly lower in cultures treated with ACh (Figure 2B) compared to cultures treated with vehicle. To investigate whether blocking of cAMP synthesis prevents cholinergic inhibition of endotoxin-induced TNF expression, RAW 264.7 cells were cultured in the presence of the adenylyl cyclase-inhibitor [cis-N-(2-Phenylcyclopentyl)azacyclotridec-1-en-2-amine.HCl] (MDL 12,330A) and exposed to the selective α7nAChR agonist choline for 30 min. The cells were then incubated with endotoxin for 3 h and TNF mRNA was measured in cell lysates by qPCR. Choline-mediated suppression of TNF mRNA was significantly reduced in the presence of MDL 12,330A (Figure 2C). Based on the observations that α7nAChR can interact with AC6 (27) and that blocking cAMP synthesis reduced cholinergic inhibition of endotoxin-induced TNF expression, RAW 264.7 cells were transfected with siRNA targeting AC6 or scrambled siRNA. AC6 protein knock-down was confirmed by Western blot (Figure 2D, inset). Transfected cells were subsequently incubated with choline and cultured in the presence of endotoxin for 3 h. Knock down of AC6 cells abolished the choline-dependent reduction of TNF mRNA (Figure 2D), indicating that AC6 is required for the α7nAChR-mediated reduction of endotoxin-induced TNF mRNA. Together, these data indicate that cholinergic inhibition of endotoxin-induced TNF expression requires AC6-dependent cAMP production.

It was previously established that increased cAMP levels promote expression and phosphorylation of c-FOS, which regulates macrophage TNF expression (24). Accordingly, we measured c-FOS protein in endotoxin-exposed RAW 264.7 cells in the presence of choline or vehicle (Figure 3A). Addition of choline augmented the production of c-FOS, supporting involvement of c-FOS in α7nAChR-mediated regulation of TNF. Based on these observations, we hypothesized that increased c-Fos expression in response to selective α7nAChR agonists contributes to the cholinergic inhibition of TNF production in macrophages. To test this, RAW 264.7 cells were transfected with siRNA targeting c-Fos or scrambled siRNA. Transfected cells were subsequently incubated with choline and cultured in the presence of endotoxin for 3 h. In cells transfected with scrambled siRNA, choline significantly inhibited endotoxin-induced TNF. In contrast, the inhibitory effect on endotoxin-mediated TNF expression was completely abolished in cells transfected with siRNA targeting c-Fos (Figure 3B). These data demonstrate that c-Fos is required for cholinergic suppression of endotoxin-induced TNF expression.

In neurons, signal transduction via α7nAChR leads to increased cAMP levels and phosphorylation of the transcription factor CREB (22), a pathway that is essential for long-term retention of information, i.e., memory (23). CREB has also been implicated in cAMP-induced inhibition of NFκB-mediated transcription of TNF (25) and activation of CREB is known to promote expression of c-Fos (30, 31). RAW 264.7 cells and human macrophages cultured in the presence of ACh or vehicle showed increased CREB phosphorylation in cultures exposed to ACh as measured by western blot (data not shown). Accordingly, we hypothesized that CREB phosphorylation underlies sustained cholinergic reduction of endotoxin-induced TNF release in macrophages. To study this, we first established whether cholinergic stimulation of human differentiated macrophages confers sustained inhibition of TNF release. Human macrophages in vitro were exposed to ACh or vehicle for 1 h, the ACh washed away, and the cells rested for 24 h. Subsequently, cultures were exposed to endotoxin for 4 h and TNF levels were measured in culture medium. ACh exposure significantly reduced endotoxin-induced TNF release to the medium 24 h after ACh exposure (Figure 4A), demonstrating that ACh exposure confers persistent reduction in endotoxin-induced TNF release in human macrophages. Next, human macrophages were transduced with a phosphorylation defective CREB (ACREB), using an engineered adenovirus (32). TNF production by human primary macrophages transduced with control virus was significantly inhibited at 24 h after exposure to the selective α7nAChR agonist 3-(2,4-dimethoxybenzylidene)anabaseine (GTS21; Figure 4B). However, GTS21 failed to elicit prolonged deactivation of macrophages transduced with the ACREB containing adenovirus (Figure 4B), indicating that CREB phosphorylation is required for the sustained inhibition of cytokine release.

Experimental in vivo protocols were approved by the Institutional Animal Care and Use Committee (IACUC) at the Feinstein Institute for Medical Research, Northwell Health, which follows the National Institutes of Health guidelines for the ethical treatment of animals. Animals were housed under a standard 12 h light/dark cycle and acclimatized for at least 1 week before conducting experiments. Water and regular rodent chow were available ad libitum. Human PBMCs from blood donors derive from completely anonymized, excess leukocytes, which do not require ethical approval.

Male Sprague-Dawley rats were anesthetized using ketamine (100 mg/kg) and xylazine (10 mg/kg; IP). The left carotid sheath containing the vagus nerve was isolated and stimulated for 60 s at 1.0 mA, 10 Hz frequency, 0.25 ms biphasic pulse using a bipolar electrode (MicroProbes for Life Science). In sham-operated animals, the vagus nerve was visualized but not manipulated.

Male C57Bl/6, wild type and α7nAChR deficient, mice were anesthetized with ketamine (100 mg/kg) and xylazine (10 mg/kg), IP. A midline cervical incision was made, the salivary glands separated, and the left carotid sheath containing the vagus nerve was isolated between the sternomastoid and sternohyoid muscles and secured within a silastic covered bipolar platinum/iridium electrode (MicroProbes for Life Science). Charge-balanced biphasic stimulation was delivered at 0.75 mA, 0.20 ms pulse width, 10 Hz for 60 s by a custom stimulator (SetPoint Medical). In some animals, the nerve was stimulated with a single biphasic pulse, 0.1 s, at 1.0 mA. In sham-operated animals the vagus nerve was visualized but not manipulated. Skin was closed with surgical staples and the mice recovered.

At various times post-VNS and sham-VNS (0, 2, 24, 48, and 72 h in rat, 24 h in mouse) VNS, endotoxin (0111:B4; Sigma-Aldrich) was injected intraperitoneally (IP; 1.5 mg/kg in rats; 8 mg/kg in mice). Rodents were euthanized after 90 min and serum TNF concentrations were measured using ELISA (R&D Systems).

RAW264.7 cells were incubated with 0 or 10 μM acetylcholine in serum-free Opti-Mem media (Gibco) containing 1 mM pyridostigmine bromide (Sigma-Aldrich), for 60 min. Media was collected 4 h after endotoxin (10 ng/ml LPS) exposure and analyzed for TNF by ELISA (R&D Systems).

Human blood was collected from healthy subjects. Peripheral blood mononuclear cells (PBMC) were collected and monocytes isolated using the Monocyte Isolation Kit II (Miltenyi Biotec). Human monocyte and macrophage cultures were exposed to 10 μM acetylcholine chloride (Sigma-Aldrich) in serum-free Opti-Mem media (Gibco) for 60 min in the presence of 1 mM pyridostigmine bromide. Cultures were washed to remove the cholinergic agonist and challenged with endotoxin in serum-free Opti-Mem media (Gibco) 24 h later. Media was collected 4 h after endotoxin addition and analyzed for TNF by ELISA (R&D Systems).

Bone-marrow derived macrophages from wild type (n = 2) and α7nAChR^−/− (n = 3) mice were obtained by flushing the femurs and tibia with sterile PBS. After 6 days of culture in the presence of 20 ng/ml M-CSF (PeproTech), cells were scraped, re-plated, and exposed to the α7nAChR selective agonist PHA568487 (Tocris) for 15 min. Wells were then washed in ice cold PBS, and cell lysis buffer (R&D systems) added on dry ice. Competitive cAMP ELISAs were conducted according to manufacturer instructions (R&D systems). Samples were freeze-thawed twice and centrifuged before assay. Protein concentration was determined by the Bradford assay (Bio-Rad) and used for normalization of cAMP values.

RAW264.7 cells were incubated with the indicated concentrations of [cis-N-(2-Phenylcyclopentyl)azacyclotridec-1-en-2-amine.HCl] (MDL 12,330A; Sigma Aldrich) for 30 min, followed by 50 mM choline chloride for 30 min. Endotoxin was added for 3 h. RNA was then isolated, cDNA generated, and analyzed by qPCR. Data shown (normalized to GAPDH) are representative of three (3) independent experiments and expressed as fold increase relative to unstimulated controls.

siRNA (125 pmol, Dharmacon) directed against adenylyl cyclase 6 (M-043463-01) or c-Fos (L-041157-00), or scrambled siRNA (D-001206-14-20), were transfected into RAW264.7 cells. 72 h later, cells were incubated with 50 mM choline for 30 min, followed by 100 ng/ml LPS for 3 h. RNA levels were analyzed by qPCR. Data shown (normalized to GAPDH) are representative of three independent experiments and expressed as fold increase relative to unstimulated controls.

Protein was extracted from cell lysates and Western blot was performed using the anti-AC5/6 (sc-590, Santa Cruz Biotechnology), anti-c-Fos (sc-52, Santa Cruz Biotechnology), goat-anti-human CREB (AF2989, R&D Systems), rabbit anti-phospho-CREB (AF2510, R&D Systems), and anti-tubulin (05-661, Millipore) antibodies.

Human macrophage cultures were transduced (Lipofectamine® 2000, ThermoFisher) with a replication deficient adenovirus (35) encoding green fluorescent protein (GFP) and a phosphorylation defective CREB (ACREB) (32) or a control virus encoding GFP. Three days after exposure to virus, cells were incubated with 100 μM 3-[(3E)-3-[(2,4-dimethoxyphenyl)methylidene]-5,6-dihydro-4H-pyridin-2-yl]pyridine (GTS21), a stable and selective α7nAChR agonist (36, 37), for 2 h, washed, and then cultured for 24 h. Endotoxin was added to the cultures for 4 h, then media collected for measurement of TNF by ELISA (R&D systems).

Differences between experimental groups were evaluated using two-tailed Student's t-test or ANOVA followed by Tukey post-hoc analysis as appropriate. p < 0.05 was considered significant.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.