Enhanced Parasympathetic Cholinergic Activity Inhibited Lipid-induced Oxidative Stress in Obese African Americans


 Background: African Americans (AAs) are disproportionately affected by cardiovascular disease (CVD), they are 20% more likely to die from CVD than whites, chronic exposure to inflammation and oxidative stress contributes to CVD. In previous studies, enhancing parasympathetic cholinergic activity has been shown to decrease inflammation. Considering that AAs have decreased parasympathetic activity compared to whites, we hypothesize that stimulating it with a central acetylcholinesterase (AChE) inhibitor, galantamine, would prevent lipid-induced oxidative stress.Objective: To test the hypothesis that acute dose of galantamine, an AChE inhibitor, decreases lipid-induced oxidative stress in obese AAs.Methods: Proof-of-concept, double-blind, randomized, placebo-controlled, crossover study that tested the effect of a single dose of 16 mg of galantamine versus placebo on lipid-induced oxidative stress in obese AAs. Subjects were studied on two separate days, one week apart. In each study day, 16 mg or matching placebo was administered before 20% intralipids infusion at doses of 0.8 mL/m2/min with heparin at doses of 200 U/h for 4 hours. Outcomes were assessed at baseline, 2 and 4 hours during the infusion.Main Outcome Measures: Changes in F2-isoprostane (F2-IsoPs), marker of oxidative stress, measured in peripheral blood mononuclear cells (PBMC) and in plasma at baseline, 2, and 4-hrs post-lipid infusion. Secondary outcomes include changes in inflammatory cytokines (IL-6, TNF alpha).Results: A total of 32 obese AA women were screened and fourteen completed the study (age 37.8±10.70 years old, BMI 38.7±3.40 kg/m2). Compared to placebo, 16 mg of galantamine significantly inhibited the increase in F2-IsoPs in PBMC (0.007±0.008 vs. -0.002±0.006 ng/sample, P=0.016), and plasma (0.01±0.02 vs. -0.003±0.01 ng/mL, P=0.023). Galantamine also decreased IL-6 (11.4±18.45 vs. 7.7±15.10 pg/mL, P=0.021) and TNFα levels (18.6±16.33 vs. 12.9±6.16 pg/mL, P=0.021, 4-hrs post lipid infusion) compared with placebo. These changes were associated with an increased plasma acetylcholine levels induced by galantamine (50.5±10.49 vs. 43.6±13.38 during placebo pg/uL, P=0.025).Conclusions: In this pilot, proof-of-concept study, enhancing parasympathetic nervous system (PNS) cholinergic activity with galantamine inhibited lipid-induced oxidative stress and inflammation induced by lipid infusion in obese AAs.Trial registration: ClinicalTrials.gov identifiers NCT02365285

mellitus (T2DM) in this population.(2) However, non-traditional CV risk factors such as chronic subclinical in ammation and oxidative stress also contributes to the pathogenesis of CVD. (3) In this context, previous studies found that AAs have increased activation of nicotinamide adenine dinucleotide phosphate (NADPH) oxidase, which is capable of producing large amounts of reactive oxygen species (ROS) and superoxide in peripheral blood mononuclear cells (PBMC).(4) Further, lipid infusion, which enrich the production of ROS, increased plasma levels of F 2 -isoprostane (F 2 -IsoPs), a biomarker of endogenous oxidative stress, by two-fold in AAs compared with whites. (5) In this regard, previous work by Tracey and collaborators(6) showed that the PNS is an important regulator of in ammation and oxidative stress; its stimulation confers protection against heightened oxidative stress states induced by fulminant hepatitis or ricin poisoning. (7,8) Likewise, central PNS cholinergic stimulation with galantamine, a central acetylcholinesterase (AchE) inhibitor showed neuro protection in rat hippocampal slices stressed with oxygen and glucose deprivation through NADPH oxidase inhibition. (9) Altogether, these studies led us to hypothesize that galantamine could reduce lipid-induced oxidative stress and in ammation, measured with F 2 -IsoPs and cytokine levels in obese AAs.

Study Design
The study was conducted at the Vanderbilt Clinical Research Center (CRC). Thestudy design was a randomized, double-blind, placebo-controlled, 2×2 crossover trial that compared the effect of 16 mg galantamine versus placebo on lipid-induced oxidative stress. Participants were assigned randomly to treatment sequences (galantamine followed by placebo [sequence one], or vice versa [sequence two]). In between treatments, there was a 2-week washout period. Both, study team and study subjects were blinded to the study sequence, gure 1. All studies adhered to the principles of the Declaration of Helsinki and Title 45, U.S. Code of Federal Regulations, Part 46, Protection of Human Subjects. The Vanderbilt Institutional Review Board approved these studies, and they were conducted in accordance with institutional guidelines. All subjects provided informed consent; the studies were registered on ClinicalTrials.gov identi ers NCT02365285.

Study Population
Eligibility criteria included obese women (as de ned as Body Mass Index (BMI) between 30 and 45 kg/m 2 ), aged 18 to 60 years old, and AA; race was self-de ned, but only subjects who reported both parents of the same race were included. We excluded pregnant or breastfeeding women, individuals diagnosed with type 2 diabetes mellitus, hypertension or any cardiovascular disease, impaired renal function (glomerular ltration rate, GFR<60%), impaired hepatic function (abnormal liver function test), or had a history of alcohol or drug abuse. Subjects were also excluded if they used potent inhibitors of Cytochrome P450 (CYP) 3A4, CYP 2D6, AchE inhibitors such as pyridostigmine, bethanechol, or had a signi cant change in weight ≥5% in the previous three months.
After informed consent was obtained, the participants were invited to come to the CRC for a screening visit that consisted of a complete history, physical exam, laboratory assessment (complete blood count, fasting lipid pro le, comprehensive metabolic panel, urine beta-hCG (human chorionic gonadotropin), and an electrocardiogram (ECG). In a separate visit, subjects returned to the CRC to complete a 75g oral glucose tolerance test and body composition measured with dual-energy X-ray absorptiometry (Lunar IDXA, GE Healthcare, CT, USA).
Subjects who met the eligibility criteria were admitted to the CRC for their rst study day (treatment sequence one, see gure 1). Before admission, they were asked to collect 24-hr urine for sodium, creatinine, and F 2 -IsoPs measurements. An intravenous catheter was placed in one arm for blood sampling and another in the contralateral arm for lipid and heparin infusion. Blood samples were collected from non-Esteri ed free Fatty Acids (NEFA), triglycerides, plasma F 2 -IsoPs, prostaglandin F2a (PGF2a), in ammatory cytokines (TNFa, IL-6, IL-10), and acetylcholine levels. PBMC were also isolated for measurements of F 2 -IsoPs. The study nurse administered the blinded medication one hour prior to a continuous infusion of 20% Intralipids® (Baxter Healthcare Crop. Glendale, CA) at a rate of 0.8 mL/m 2 /min as previously described. (5) In addition, the study subject received a heparin bolus of 1,000 units followed by a heparin infusion at a rate of 200U/h for 4-hrs to activate the endothelial lipoprotein lipase and accelerate the hydrolysis of fatty acids from the glycerol backbone of the triglycerides.
Intermittent blood pressures and ECG were measured using the VITAL-GUARD 450c monitor (Ivy Biomedical Systems, Brandford, CT, USA). Contrast-enhanced ultrasonography (CEU) was used to measure changes in microvascular blood volume in the skeletal muscle. All measurements were repeated at 2 and 4-hrs post lipid infusion. We only obtained the CEU measurements and urine samples for sodium, creatinine, and F 2 -IsoPs assessments at 4-hrs. Subjects washed out for two weeks and the subject completed sequence two and underwent all the procedures outlined previously.

Randomization
Subjects were randomly assigned to the treatment sequences using a permuted-block randomization algorithm. The Vanderbilt Investigational Pharmacy was responsible for the randomization, storage, preparation, and labeling of all investigational agents (Intralipids®, heparin, and blinded study drug) and for maintaining accurate drug storage and dispensing logs.

Intervention
Our intervention was 16 mg of galantamine hydrobromide (Razadyne®) versus matching placebo. Galantamine is a competitive central AchE inhibitor that increases the availability of acetylcholine. This drug is FDA-approved for the treatment of Alzheimer's dementia. We chose a dose of 16 mg p.o. based on pharmacokinetic studies in normal volunteers and patients with dementia of Alzheimer's type. The 16 mg dose exhibits linear pharmacokinetics after an oral administration; the oral bioavailability is about 90% 9 and the maximum concentration is achieved at 1-hr post-administration with a short half-life of 7-hrs. Our measurements at 2 and 4-hrs coincide with the peak concentration of galantamine after administration.

Endpoints
The primary endpoint of this study was the change in plasma F 2 -IsoPs with lipid infusion (ΔISO) during placebo versus 16 mg of galantamine.
Secondary endpoints were the changes in F 2 -IsoPs in PBMC and in ammatory cytokines post-lipid infusion during placebo versus galantamine.

Clinical Chemistry
The blood samples were collected in chilled Ethylenediaminetetraacetic acid (EDTA) tubes and were immediately centrifuged to separate the plasma, and stored at −80°C. For serum, the blood was clotted at room temperature for 20 minutes and centrifuged; the serum was removed and stored at −80°C. Plasma glucose was measured at the bedside with a glucose analyzer (YSI Life Sciences, Yellow Springs, OH). Plasma insulin concentrations were determined by radioimmunoassay (Millipore, St. Charles, MO).
Tetrahydrolipstatin was added to the NEFA collection tube to prevent in vitro lipolysis. 13 Serum NEFA were measured as previously published, (10) and Triglycerides (Roche Diagnostics, Indianapolis, IN) by enzymatic colorimetry Cliniqa(11) for microtiter plates.
In ammatory cytokines (TNFa, IL-6, IL-10) were measured with Multiplex Luminex technology using x- Ann Arbor, MI) internal standard. The solution was adjusted to pH 3 with 1N HCl. The sample was then applied to a C-18 Sep-Pak cartridge that has been pre-washed with 5 ml methanol and 5 ml 0.01N HCl.
The cartridge was then washed with 10 ml 0.01N HCl, followed by 10 ml heptane, and compounds were then eluted with 10 ml ethyl acetate: heptane (50:50, v/v). The eluate was applied to a silica Sep-Pak cartridge prewashed with ethyl acetate (5 ml) rinsed with 5 ml ethyl acetate and compounds eluted with 5 ml ethyl acetate: methanol (50:50, v/v). The eluate was dried under nitrogen. Compounds were converted to the penta uorobenzyl (PFB) esters by the addition of 40 μl of a 10% solution of penta uorobenzyl bromide in acetonitrile and 20 μl of a solution of 10% diisopropylethanolamine in acetonitrile and allowed to incubate for 30 min at 37 o C. Reagents were dried under nitrogen and the residue was reconstituted in 30 μl chloroform and 20 μl methanol and chromatographed on a silica Thin-layer chromatography (TLC) plate to 13 cm in a solvent system of chloroform: methanol (93:7, v/v). The R f of PGF2a methyl ester in this solvent system was 0.15. Compounds migrating in the region 1 cm below the PGF2a standard to 1.0 cm above the standard were scraped from the TLC plate, extracted with 1 ml ethyl acetate, and dried under nitrogen. Following TLC puri cation, compounds were converted to trimethylsilyl (TMS) ether derivatives by addition of 20 μl N,O-bis(trimethylsilyl)tri uoroacetamide, and 10 μl dimethylformamide.
The sample was incubated at 37 o C for 10 minutes and then dried under nitrogen. The residue was redissolved for GC/MS analysis in 20 μl undecane that has been stored over a bed of calcium hydride.

Immunoblotting and western blot analysis
Immunoprecipitation and Western blot analysis were performed as previously described. (13) To determine the association of p47phox with gp91phox, protein isolated from PBMC homogenates of patients treated with placebo and galantamine were immunoprecipitated by adding the gp91phox (antibody (2 mL per sample; Abcam, Cambridge, MA, Cat# ab80508) and Protein G Agarose beads (40 ml per sample; ThermoFisher Scienti c, Waltham, MA, Cat# 20398) and incubated at 4 °C overnight. After microcentrifugation and washing the pellet to reduce non-speci c binding and remove excess anti-gp91phox antibody, an equal volume of each denatured immunoprecipitation sample was loaded onto a SDS-PAGE gel (12 -15%). Western blotting was performed using the primary antibodies rabbit anti-p47phox (1:1,000; Millipore, Billerica, MA, Cat# 07-001) and rabbit anti-gp91phox (1:1,000; Abcam, Cambridge, MA, Cat# ab80508) followed by incubation with goat anti-rabbit horseradish peroxidaselabeled IgG (1:2,500; Invitrogen, Rockford, IL, Cat# 62-9520). Next, we analyzed the ratio of p47phox to gp91phox by quantifying relative protein expression to determine NADPH oxidase formation in these cells.
The sample was vortexed for 15 seconds and protein was removed by centrifugation at 10,000g for two minutes. Samples were prepared for mass spectrometry as follows: 5uL of supernatant was added to a 1.5mL microcentrifuge tube. To that was added 10uL each of 500mM NaCO3 (aq) and 2% BZC in acetonitrile. After two minutes, 20uL of isotopically-labeled internal stand solution (acetylcholine-d9 in 20% acetonitrile containing 3% sulfuric acid) was added to the tube. The solution was transferred to an LC/MS vial for analysis.
Chromatographic separation was performed on a 2.0 x 50 mm, 1.7µm particle Acquity BEH C18 column (Waters Corporation, Milford, MA, USA) using a Waters Acquity I-Class UPLC with Sample Organizer. Mobile phase A was 15% aqueous formic acid and mobile phase B was acetonitrile. Samples were separated by a gradient of 98-5% of mobile phase A over 6 min at a ow rate of 450 µl/min prior to delivery to a Waters Xevo TQ-S micro triple quadrupole mass spectrometer. Acetylcholine was monitored using a transition of m/z 146 à m/z 87 (retention time = 0.69 min). Acetylcholine-d9 was monitored using a transition of m/z 155 à m/z 87 (retention time = 0.70 min).

Contrast-Enhanced Ultrasonography
Contrast-enriched images were acquired in the contralateral forearm (brachioradialis muscle) using a linear-array transducer connected to an ultrasound (L9-3 mm transducer, iU22; Phillips). This equipment allowed real-time imaging using low (0.08) and high (1.2) mechanical index as the contrast (microbubbles; De nity; Bristol-Myers Squibb) was infused at a rate of 1.5 ml/min through the Intravenous (IV) access for 10 mins. At steady state (~4 min) the high mechanical index (1.2) used destroyed the microbubbles at the start of video recording. Switching to the low index (0.08) made the microbubbles resonate allowing real-time recording of vascular replenishment. Local temperature was measured using a laser Non-Contact Infrared Skin Thermometer. Data was analyzed in our lab using QLAB ultrasound cardiac.

Sample size and power calculation
The sample size calculation was performed to detect a 30 % difference in D F 2 -IsoPs (primary endpoint) in the same study subjects between placebo versus galantamine treatment. Lopes et al. (5) reported that the mean and SD of D F 2 -IsoPs post lipid infusion was 12.0±2.60 pg/mL in AAs. A sample size of 12 AA women had 85% power to detect a difference of 30 % in D F 2 -IsoPs between treatments in the same individual. We enrolled a total of 14 AA women to account for attrition and loss of follow-up.

Statistics
Standard graphing and screening techniques were used to detect outliers and to ensure data accuracy. The data was assessed for normality. If normality was violated, we applied non-parametric analysis methods. Data were analyzed using R version 3.5.3 software and expressed as mean ± SD throughout the manuscript unless otherwise indicated. Summary statistics for both continuous and categorical variables were provided by randomization groups to describe the study sample. We calculated withinsubject mean differences and 95% con dence intervals for galantamine versus placebo comparison within each racial group and tested for treatment effect using the paired t-test or sign rank test as appropriate. P less than 0.05 was considered statistically signi cant.

Study subjects
A total of 32 obese AA women were screened, 18 subjects were excluded because they met exclusion criteria, or withdrew consent. Fourteen subjects were randomized and completed the two study days, see gure 1, enrollment ow. The baseline characteristics of these patients and the medications they used are presented in the Table 1. Three subjects were on oral contraceptive medications, and two were on inhaled steroids. None of the patients were diagnosed with type 2 diabetes mellitus as per oral glucose tolerance test; 36% of the patients were insulin resistant based on Matsuda index (ISI-M <3), gure 2.
Additionally, we examine the effect of galantamine on lipid-induced assembly of NADPH oxidase, we immunoprecipitated gp9 phox then performed Western blot to detect association with p47 phox . In PBMCs of ve AAs, 20% intralipid® infusion with galantamine at 2-hrs prevented association of p47 phox with gp91 phox compared to placebo (215.2 ±104.80 vs. 11.8 ±4.90 Relative Intensity), gure 6. The data showed a tendency towards a reduction in NADPH oxidase assembly with galantamine treatment.

Effect of Galantamine on Microvascular circulation
Microvascular circulation was measured at baseline and 4-hrs after the lipid infusion. Our results, did not found any difference in the amount of MBV (microvascular blood volume) between placebo and galantamine during acute hyperlipidemia, (0.19±4.01 vs. 1.25±6.77, P=0.15), gure 7.
Acetylcholine levels in plasma as biomarker for galantamine's actions As expected, galantamine signi cantly increased acetylcholine levels two hours post-administration (43.6±13.38 vs. 50.5±10.49 pg/uL with galantamine, P=0.025) which was consistent with the peak effect of the drug, gure 8.

Adverse Events
Eight subjects developed mild adverse events (AEs). There were no serious AEs. In the placebo group, two subjects reported nausea and headaches that resolved after the study ended.
In the galantamine group, the most common side effects were severe nausea, headaches, and abdominal cramps.

Discussion
The main nding of our study was that in obese AA women, galantamine increased PNS cholinergic activity and suppressed lipid-induced oxidative stress as measured by F 2 -IsoPs in PBMC and plasma; this effect was associated with a decrease in the production of in ammatory cytokines, particularly IL-6 and TNFα.
The effect of galantamine on F 2 -IsoPs was evaluated in three different organs (PBMC, plasma, and urine). PBMC is one of the largest sources of ROS; F 2 -IsoPs, an endogenous biomarker of oxidative stress, is produced in vivo by nonenzymatic peroxidation of arachidonic acid esteri ed in membrane phospholipids, and therefore speci cally assess lipid peroxidation. Then, F 2 -IsoPs is released to the circulation by phospholipase A2 activities and excreted in the urine. In AA women, a single dose of galantamine decreased F 2 -IsoPs in PBMC and plasma at 2-hrs during lipid infusion, which coincided with the drug's peak effect and shown by the signi cant increase in Ach levels. Furthermore, at 4-hrs post-lipid infusion, F 2 -IsoPs in plasma returned to baseline values, but it remained elevated in PBMC. In contrast, we did not observe a signi cant decline in F 2 -IsoPs in urine collected at 4-hrs after drug intake, possibly because the maximum effect of galantamine on oxidative stress occurred at 2-hrs after intake.
Previous studies in animal models found that stimulation of the PNS conferred protection against oxidative stress.(6) Stimulation of the PNS decreased malondialdehyde generated in response to various stimuli including ricin poisoning, myocardial infarction and fulminant hepatitis (7,8). This effect was in part mediated through the α7nAChR, (14,15) and vagal afferent nerves.(6) Also, sub-diaphragmatic surgical vagotomy in rats exposed to endotoxin was associated with increased oxidative stress in the brain.(16) Similarly, direct electrical stimulation of the peripheral vagus nerve during lethal endotoxemia inhibited TNF alpha synthesis in the liver. (14) Altogether, these studies suggest that states of heightened ROS production can be targeted by increasing PNS activity either through direct vagus nerve stimulation or with drugs that increase cholinergic transmission.
In the present study, we selected continuous lipid infusion as stimuli to generate oxidative stress because it is safe to use in humans and AAs had a higher F 2 -IsoPs production than whites.(5) This increased susceptibility to heightened oxidative stress production is unknown; previous studies in AA women reported an increased expression of NADPH oxidase, p47 phox subunit, NOX2, and NOX4 in baseline conditions. (17) To understand the mechanism underlying the reduction in lipid-induced F 2 -IsoPs production, we evaluated the activity of NADPH oxidase in PBMC of ve subjects who showed a robust decline in F 2 -IsoPs in response to galantamine. The NADPH oxidase activity depends on the assembly of the cytosolic subunits p40 phox , p47 phox , and p67 phox with the membrane-bound subunits p22 phox and gp91 phox . During activation of the NADPH oxidase, p47 phox is phosphorylated, and the cytosolic subunits assemble and move to the membrane to form a functional complex with the membrane-bound subunits. Three subjects showed reduction in NADPH oxidase activity with galantamine through a substantial decrease in the association of p47 phox with gp91 phox during lipid infusion, gure 6. Further studies with a large sample size are needed to examine the effect of galantamine on activation of the NADPH oxidase.
The relationship between oxidative stress and in ammation is complex, lipid-induced IL-6 and TNF alpha production, a long side F 2 -IsoPs, signi cantly decreased with galantamine. It is noteworthy that F 2 -IsoP and isolevuglandins (IsoLGs) work as neo-antigens leading to activation of antigen-presenting cells, T cells, and release of in ammatory cytokines. This autoimmune state has been associated with the development of salt-sensitive hypertension,(18) which disproportionately affect AAs. In future studies, it would be important to determine if galantamine decreases protein-adduct IsoLGs, antigen-presenting cells, and T cell activation. The present study has several limitations; rst, the sample was small; however, our strong study design (crossover study, which decreases interindividual variability) and the use of lipid infusion to enrich ROS production, allowed us to reach meaningful conclusions. We did not measure other oxidative stress processes like protein and DNA damage derived from oxidation damage, which were shown to be reduced with galantamine in subjects with metabolic syndrome. (20) Nevertheless, our study focused on F 2 -IsoPs measurements in different system, which is a reliable biomarker of endogenous lipid peroxidation.

Conclusions
In summary, our study found that galantamine inhibited lipid-induced oxidative stress as measured by F 2 -IsoPs in PBMC and plasma in AAs; this effect was associated with decreased production of in ammatory cytokines IL-6 and TNF alpha.    Figure 1 Enrollment Flow The study was a double-blind, randomized, placebo-controlled, 2×2 crossover design. A total of 32 subjects were screened, 18 were excluded. Fourteen were randomly assigned to sequence 1 (16 mg galantamine followed by placebo) or sequence 2 (placebo followed by 16 mg of galantamine).     Assessment of NADPH oxidase activation in PBMC with Galantamine We obtained PBMC in ve AAs who showed a robust decline in F 2 -IsoPs in response to galantamine. The NADPH oxidase activity depends on the assembly of the cytosolic subunits p40 phox , p47 phox , and p67 phox with the membrane-bound subunits p22 phox and gp91 phox . There was not statistically signi cant difference in NADPH activity with galantamine.