Conjugated Linoleic Acid Modulates Clinical Responses to Oral Nitrite and Nitrate

Abstract

Dietary nitrate (NO₃⁻) and nitrite (NO₂⁻) support nitric oxide (^·NO) generation and downstream vascular signaling responses. These nitrogen oxides also generate secondary nitrosating and nitrating species that react with low molecular weight thiols, heme centers, proteins and unsaturated fatty acids. To explore the kinetics of NO₃⁻ and NO₂⁻ metabolism and the impact of dietary lipid on nitrogen oxide metabolism and cardiovascular responses, the stable isotopes Na¹⁵NO₃ and Na¹⁵NO₂ were orally administered in the presence or absence of conjugated linoleic acid (cLA). The reduction of ¹⁵NO₂⁻ to ¹⁵NO was indicated by electron paramagnetic resonance spectroscopy detection of hyperfine splitting patterns reflecting ¹⁵NO-deoxyhemoglobin complexes. This formation of ¹⁵NO also translated to decreased systolic and mean arterial blood pressures and inhibition of platelet function. Upon concurrent administration of cLA, there was a significant increase in plasma cLA nitration products 9- and 12-¹⁵NO₂-cLA. Co-administration of cLA with ¹⁵NO₂⁻ also impacted the pharmacokinetics and physiological effects of ¹⁵NO₂⁻, with cLA administration suppressing plasma NO₃⁻ and NO₂⁻ levels, decreasing ¹⁵NO-deoxyhemoglobin formation, NO₂⁻ inhibition of platelet activation, and the vasodilatory actions of NO₂⁻, while enhancing the formation of 9- and 12-¹⁵NO₂-cLA. These results indicate that the biochemical reactions and physiologic responses to oral ¹⁵NO₃⁻ and ¹⁵NO₂⁻ are significantly impacted by dietary constituents such as unsaturated lipids. This can explain the variable responses to NO₃⁻ and NO₂⁻ supplementation in clinical trials and reveals dietary strategies for promoting the generation of pleiotropic nitrogen oxide-derived lipid signaling mediators.

Introduction

Inorganic nitrite (NO₂⁻) undergoes reactions that yield vasodilatory products that contribute to blood pressure regulation and hypoxic signaling^{1, 2}. The entero-salivary microbiome-mediated reduction of NO₃⁻ to NO₂⁻ and ^·NO in mammals also promotes vascular relaxation and other pleiotropic pharmacologic responses^3–6. These reactions of NO₃⁻/NO₂⁻/^·NO both complement L-arginine/^·NO synthase/^·NO/cGMP signaling and expand the spectrum of nitrogen oxide intermediates that instigate signaling responses beyond the activation of guanylate cyclase. This new understanding contrasts with the previous perspective that tissue NO₃⁻ and NO₂⁻ represented potentially toxic and relatively stable ^·NO oxidation products.

There are an abundance of metabolic and inflammatory reactions that NO₃⁻ and NO₂⁻ undergo that can differentially impact downstream signaling responses^{1, 2, 4} (Figure 1A). While NO₃⁻ is formed upon the oxidation of ^·NO by oxyhemoglobin, much greater endogenous NO₃⁻ levels are achieved upon the ingestion of dietary NO₃⁻ sources, including leafy green vegetables, roots (e.g., beets) and fruits⁷. Once consumed, NO₃⁻ is readily reduced to NO₂⁻ by commensal oral bacterial nitrate reductase activities⁸. While the majority of NO₃⁻ is reduced in the oral cavity by commensal bacteria, xanthine oxidoreductase can reduce NO₃⁻ to NO₂⁻⁹. Up to 25% of plasma NO₃⁻ can be stored and concentrated by the salivary glands, thus providing a facile route for entero-salivary recycling^10–13. Also acquired from both endogenous ^·NO oxidation reactions and dietary sources, the more intrinsically-reactive NO₂⁻ can undergo multiple reactions such as protonation to nitrous acid (HNO₂) in the low pH environments of the GI tract, inflammatory foci and the mitochondrial intermembrane compartment. Nitrite is also absorbed into the circulation and tissues where it can react with a) deoxyhemoglobin and other heme proteins, as well as molybdopterin-containing enzymes to form ^·NO ^{14, 15,16, 17} or b) heme peroxidases to yield the nitrating species nitrogen dioxide (^·NO₂)¹⁸. During intermediary metabolism and inflammatory responses, partially reduced oxygen species such as superoxide (O₂^·−), hydrogen peroxide (H₂O₂) and lipid peroxyl radicals will also undergo a variety of reactions with ^·NO and NO₂⁻ to accelerate rates of production of nitrosating and nitrating species^{19. ·}Nitric oxide, NO₂⁻ and HNO₂ will also undergo a variety of reactions that lead to the generation of symmetric and asymmetric dinitrogen trioxide (ONONO and OONNO)^20,21. These unstable intermediates rapidly decay to yield secondary nitrosating (addition of ^·NO) and nitrating (addition of ^·NO₂) species, as well as ^·NO. These and other redox reactions are responsible for propagating downstream signaling and pathogenic responses that are a consequence of oxidative, nitrosative and nitrative post-translational protein modification (PTM) reactions.

Figure 1. Trial schemes and NO₂⁻ signaling pathways.

(A) Multiple metabolic and inflammatory reactions yield nitrogen dioxide. Nitration (^·NO₂) and nitrosation (NO⁺) reactions produce an array of bioactive nitrogen oxide products, including NO₂-FA and S-nitrosothiol derivatives that have incompletely-characterized biological activities. (B) To understand the metabolism of single doses of heavy nitrogen labeled sodium NO₃⁻ and NO₂⁻ and to characterize their signaling pathways and physiologic effects, cross-over drug design studies were conducted. Ten healthy volunteers were randomized into one of two subject cohorts to receive a single dose of each study drug, oral ¹⁵N-labeled sodium NO₃⁻ and NO₂⁻, in random order, separated by a washout period. Five of the 10 healthy volunteers who completed Trial 1 returned and were randomized to receive a single dose of each drug, oral ¹⁵N-labeled sodium NO₃⁻ and NO₂⁻ plus conjugated linoleic acid (cLA), in random order, separated by a washout period in Trial 2. (C) To track NO₂⁻ metabolism in vivo, methemoglobin (MetHb), ¹⁵NO bound to the heme of hemoglobin (¹⁵NO-Hb), RS-NO and ¹⁵NO₂-fatty acid (FA) formation were examined.

Human studies and animal models reveal that up to 15 day courses of inorganic and dietary NO₃⁻ and NO₂⁻ administration induce robust responses that include decreased blood pressure (BP)^{1, 2, 11, 22}, hypoxic vasodilation^{1, 23, 24}, modulation of mitochondrial function under hypoxic or exercise stress^{25, 26}, prevention of endothelial dysfunction²⁷ and inhibition of platelet aggregation^{28, 29}. It remains uncertain as to whether these effects are mediated by ^·NO, secondary nitrating and nitrosating products or a combination thereof that will induce functionally-significant PTMs such as thiol S-nitrosation or thiol alkylation^{30, 31}.

One class of of nitrogen oxide metabolites are electrophilic fatty acid nitroalkene derivatives (NO₂-FA)³², formed by radical addition reactions of ^·NO₂ with alkenyl carbons of unsaturated fatty acids. Once formed, these lipid electrophiles rapidly react with the nucleophilic amino acids Cys, and to a lesser extent His, residues via reversible Michael addition³³. Endogenously present at low nM concentrations in healthy human plasma and urine, the rates and extents of production of NO₂-FA can be increased by an array of metabolic and inflammatory-related nitration reactions (Fig. 1C)³⁴. Upon the PTM of functionally-significant hyperreactive Cys residues in critical enzymes and transcriptional regulatory proteins, NO₂-FA influence gene expression and inflammatory responses³⁵. For example, specific pro-inflammatory and blood pressure regulation enzymes are directly inhibited by NO₂-FA, including xanthine oxidoreductase, cyclooxygenase-2 and soluble epoxide hydrolase (sEH)^36–38. Notably, NO₂-FA mediate pleiotropic signaling actions: this includes the PTM of p65 inhibit nuclear factor-κB to inhibit pro-inflammatory cytokine expression, the activation of heat shock factor-1 dependent heat shock protein expression, the partial agonism of peroxisome proliferator-activating receptor-γ and the activation of nuclear factor (erythroid-derived) 2-regulated anti-inflammatory gene expression^39–43.

The most prevalent endogenous NO₂-FA, nitro-conjugated linoleic acid (NO₂-cLA), is detected as both the free acid, complex lipid-esterified and as protein-adducts^{34, 44}. Conjugated linoleic acid (predominantly 18:2, cis-9, trans-11) is a dietary polyunsaturated fatty acid prevalent in dairy products, meats and plants. Endogenous generation of cLA in humans occurs by the isomerization of bis-allylic linoleic acid (LA) to a conjugated diene and by the desaturation of oleic acid to cLA, with both reactions catalyzed by the gut microbiome⁴⁵. Importantly, the external flanking carbons of the conjugated diene of cLA are 3–4 orders of magnitude more reactive than bis-allylic LA⁴⁴. This promotes facile addition reactions of radical species such as ^·NO₂ to yield NO₂-cLA. It will always be a daunting challenge to dissect which products of nitrogen oxide metabolism, such as ^·NO, metal-NO complexes, S-nitrosothiols (RS-NO) or NO₂-FA, are proximally responsible for modulating the physiological actions of nitrogen oxides in humans, such as the regulation of inflammatory responses, BP regulation and platelet function.

We hypothesized that ingested NO₃⁻ is reduced to the redox-active metabolite NO₂⁻ in vivo, which is then metabolized to both ^·NO and NO₂-cLA. Since ^·NO and NO₂-cLA are both signaling molecules known to modulate BP and platelet function^{3, 38, 46–50}, we evaluated nitrogen oxide levels and clinical responses upon consumption of oral¹⁵NO₃⁻ and ¹⁵NO₂⁻ by healthy adults, with and without cLA supplementation. The use of FDA Investigational New Drug (IND)-approved (IND #115926) oral formulations of ¹⁵NO₃⁻ and ¹⁵NO₂⁻ allowed for specific metabolite tracking in vivo, and demonstrated that ¹⁵NO-Hb and ¹⁵NO₂-cLA formation primarily occurred following ¹⁵NO₂⁻ consumption, and that the physiological responses and nitrogen oxide product profiles were strongly influenced by the concomitant presence of cLA.

Material and Methods

Study design: Two pharmacokinetic (PK) studies of ¹⁵NO₃⁻ and ¹⁵NO₂⁻ metabolism were conducted, without (Trial 1) and with cLA (Trial 2) supplementation

In Trial 1, ten subjects were enrolled. To understand the precise mechanism of how the metabolites signal and exert their effects, these subjects were invited to return to participate in a modified PK study, Trial 2, to serve as a direct paired control. Five subjects completed Trial 2. Subjects were randomized in both trials as shown in Figure 1B. All study subjects were ages 18–60 years and had a normal BP defined as systolic BP≤130 and diastolic BP≤85 mmHg. The study was approved by the University of Pittsburgh Institutional Review Board and the U.S. Food and Drug Administration for use of these INDs. Prior to performing any of the research study procedures or interventions, subjects provided written informed consent and procedures were followed in accordance with institutional guidelines.

The selected drug doses were 1 g Na¹⁵NO₃⁻ (11.8 mmol) and 20 mg Na¹⁵NO₂⁻ (0.29 mmol) with and without 3 g cLA. With each drug dose, plasma samples were collected for PK analysis of NO₃⁻ and NO₂⁻ and methemoglobin (MetHb) was assessed using non-invasive co-oximetry (Masimo Corp., Irvine, CA), at times 0 (baseline or trough prior to study drugs), 0.5, 1, 2, 3, 6 and 24 hr post-drug administration. Prior to the start of each PK study visit, subjects fasted for 10–12 hr. BP and mean arterial blood pressure (MAP) were measured for 30–45 min to ensure subjects were at a steady state prior to the time 0 MAP and administration of the study drugs. BP, MAP, respiratory rate and heart rate monitoring were measured every 15 min during the first 2 hr, then at 3, 6, and 24 hr post-drug administration.

To track NO₂⁻ metabolism in vivo (Figure 1C), PK evaluations were utilized to examine total (¹⁵N and ¹⁴N) plasma NO₃⁻, NO₂⁻ and RS-NO (S-nitrosothiols) by gas phase reductive ozone-based chemiluminescence detection. ¹⁵NO was differentiated from endogenous ¹⁴NO in blood using electron paramagnetic resonance spectroscopy (EPR) via the formation of the ¹⁵NO^· ligand to deoxyhemoglobin (¹⁵NO-Hb). This iron-nitrosyl paramagnetic species has a distinctive hyperfine absorbance measured by EPR which produces a doublet for NO-Hb labeled with ¹⁵N. To elucidate the overall metabolic fate of ¹⁵NO₃⁻ and ¹⁵NO₂⁻ without and with cLA, plasma and urinary ¹⁵NO₂-cLA was differentiated from endogenous ¹⁴NO₂-cLA using liquid chromatography-tandem mass spectrometry (LC-MS/MS) as previously described⁴⁴. A detailed description of the urinary measures is published separately³⁴. Platelet activity was monitored at time 0, 6 and 24 hr after ¹⁵N drug with and without cLA. Comprehensive methods are described in Supplemental Material.

Statistical analysis

Detailed statistical analyses for each data set are included in the individual figure legends. Where differences existed between trial drug treatments, post-hoc multiple comparisons were performed using Bonferroni correction. Measurements shown represent mean ± SEM. A p value <0.05 was considered significant.

Results

¹⁵NO₃⁻/¹⁵NO₂⁻ pharmacokinetics with and without cLA

Ten healthy adult volunteers (baseline characteristics in Table S1) who fulfilled the inclusion/exclusion criteria were randomized into one of two subject cohorts of 5 subjects to receive a single dose of each study drug, oral Na¹⁵NO₃⁻ (1 g) and Na¹⁵NO₂⁻ (20 mg) in random order (Figure 1B). PK visits were separated by a 3–7 day washout period to ensure washout of nitrogen oxides and metabolites. Total plasma NO₃⁻ concentrations were significantly increased after ¹⁵NO₃⁻ dosing (peak level of 769 ± 38 μM, p<0.001, Figure 2A). Upon dosing with ¹⁵NO₂⁻, total plasma NO₃⁻ concentrations rose over 3 hr compared to baseline (p<0.001, Figure 2B). When the same subjects received oral ¹⁵NO₃⁻, total plasma NO₂⁻ concentrations increased over 6 hr compared to baseline (p<0.01, Figure 2C). In contrast, when subjects received oral ¹⁵NO₂⁻, total plasma NO₂⁻ concentrations rapidly peaked at 0.5 hr (5.5 ± 0.7 μM, p<0.001, Figure 2D) and returned to basal concentrations after 3 hr.

Figure 2. ¹⁵NO₃⁻ and ¹⁵NO₂⁻ PK without and with cLA.

(A) Following oral ¹⁵NO₃⁻, plasma NO₃⁻ concentrations increase at all time points compared to baseline. (B) Following oral ¹⁵NO₂⁻, plasma NO₃⁻ concentrations rise through 3 hr compared to baseline. (C) Following oral ¹⁵NO₃⁻, plasma NO₂⁻ concentrations increase through 6 hr compared to baseline. (D) Following oral ¹⁵NO₂⁻, plasma NO₂⁻ concentrations increase at 0.5 hr and return to baseline after 6 hr. (E) When plasma NO₃⁻ concentrations were examined following oral ¹⁵NO₃⁻ treatment alone compared to ¹⁵NO₃⁻ with cLA, lower plasma NO₃⁻ concentrations were achieved with ¹⁵NO₃⁻ plus cLA. (F) A trend towards lower plasma NO₂⁻ concentrations was seen through 1 hr with ¹⁵NO₂⁻ plus cLA compared to ¹⁵NO₂⁻ alone. Repeated measures ANOVA with time as the within-subject effect was used to evaluate response to drug treatment for the endpoint measures in A–D. 2×2 repeated measures ANOVA with time as the within-subject effect and trial drug(s) (without vs. with cLA) as the between subject effect was used to compare the endpoint measures between Trial 1 and Trial 2 in E–F.

Five of the 10 healthy volunteers who completed Trial 1 returned for Trial 2 and were randomized to receive a single dose of each study drug, Na¹⁵NO₃⁻ (1 g) and Na¹⁵NO₂⁻ (20 mg), plus 3 g cLA, in random order, separated by a 3–7 day washout period so that metabolic responses could be compared to those from Trial 1 (Figure 1B). Total plasma NO₃⁻ concentrations were lower following ¹⁵NO₃⁻ + cLA at 1 hr (473.2 ± 72.4 μM, Figure 2E closed circles) compared to ¹⁵NO₃⁻ supplementation (743.1 ± 72.2 μM, p<0.01, Figure 2E open circles). Peak plasma NO₃⁻ concentration occurred at 2 hr for ¹⁵NO₃⁻ + cLA administration and 1 hr for ¹⁵NO₃⁻ alone. No significant differences in plasma NO₂⁻ concentrations were observed for ¹⁵NO₃⁻ treatment alone vs. ¹⁵NO₃⁻ + cLA (Figure S1). While total plasma NO₂⁻ concentrations increased over time after ¹⁵NO₂⁻ + cLA (p=0.001), a trend towards lower total plasma NO₂⁻ concentrations was seen at 0.5 and 1 hr (Figure 2F closed circles) when compared to ¹⁵NO₂⁻ alone (p=0.059, Figure 2F open circles). Peak plasma NO₂⁻ concentration was 3.0 ± 0.5 μM for ¹⁵NO₂⁻ + cLA and 4.6 ± 1.0 μM for ¹⁵NO₂⁻ at 0.5 hr.

Formation of MetHb and ¹⁵NO-Hb

Further metabolism of ¹⁵NO₂⁻ was also reflected by MetHb, ¹⁵NO-Hb, plasma RS-NO and ¹⁵NO₂-cLA generation (Figure 1C). Following ¹⁵NO₃⁻ administration, no significant increase in MetHb was observed (Figure 3A). Following ¹⁵NO₂⁻ dosing, a small but significant increase in MetHb occurred from baseline over 1 hr (baseline 1.2 ± 0.1%, peak at 1 hr 1.6 ± 0.1%, p<0.001, Figure 3B) and returned to baseline by 3 hr. cLA supplementation did not impact MetHb levels (not shown). The metabolism of ¹⁵NO₂⁻ to ^·15NO was measured by the formation of ¹⁵NO-Hb in red blood cells sedimented from venipuncture blood samples. A representative EPR spectra (red) shows the hyperfine splitting characteristic of ¹⁵NO-Hb from one subject (raw data in blue, Figure 3C). There was no evidence for the formation of ¹⁴NO-Hb (Figure S2B). ¹⁵NO-Hb formation was only detected between 0.5 and 1 hr after ¹⁵NO₂⁻ supplementation (Figure 3D). ¹⁵NO-Hb derivatives were not detectable after ¹⁵NO₂⁻ + cLA and ¹⁵NO₃⁻ + cLA administrations (not shown).

Figure 3. NO₂⁻ signaling and methemoglobin and ¹⁵NO-Hb formation.

(A) Following oral ¹⁵NO₃⁻, no significant increase in methemoglobin (MetHb) occurs. (B) Following oral ¹⁵NO₂⁻, a significant rise in MetHb occurs through 1 hr. (C) A representative EPR spectra (red) shows the characteristic hyperfine splitting of ¹⁵NO-Hb from one subject following oral ¹⁵NO₂⁻ (raw tracings shown in blue). The fit is composed of 42% pentacoordinate ¹⁵N alphanitrosyl Hb, 21% hexacoordinate alphanitrosyl Hb, and 38% betanitrosyl Hb. Inclusion of pentacoordinate ¹⁴N alphanitrosyl Hb does not substantially improve the fit (Fig. S3B). (D) ¹⁵NO-Hb formation is detected with oral ¹⁵NO₂⁻ treatment only at 0.5 to 1 hours, with no ¹⁵NO-Hb detected with oral ¹⁵NO₃⁻ treatment. NS = not significant, N.D. = not detected. Repeated measures ANOVA with time as the within-subject effect was used to evaluate response to drug treatment for the endpoint measures in A–B.

Formation of RS-NO

Representative plasma RS-NO traces are shown at baseline and for peak concentrations of RS-NO at 0.5 hr after ¹⁵NO₂⁻ dosing, followed by an additional time point after ¹⁵NO₂⁻ treatment (Figure 4A). ¹⁵NO₂⁻ consumption significantly increased plasma RS-NO concentrations (baseline 3.2 ± 1.0 nM, peak concentration of 6.1 ± 1.0 nM at 0.5 hr, p<0.05, Figure 4B) and returned to basal concentrations after 3 hr. In contrast, after ¹⁵NO₃⁻ dosing, there was no significant impact on RS-NO concentration (Figure S3A open circles). There was also no impact of cLA supplementation on plasma RS-NO concentrations after dosing in combination with either ¹⁵NO₃⁻ or ¹⁵NO₂⁻ compared to ¹⁵NO₃⁻ or ¹⁵NO₂⁻ alone (Figure S3A and B).

Figure 4. NO₂⁻ signaling and RS-NO.

(A) Representative plasma RS-NO traces at baseline and the time when peak RS-NO concentrations were detected with oral ¹⁵NO₂⁻ followed by a time point after ¹⁵NO₂⁻ treatment. (B) Following oral ¹⁵NO₂⁻, plasma RS-NO concentrations increase significantly, but after 3 hr, approach baseline. Repeated measures ANOVA with time as the within-subject effect was used to evaluate response to drug treatment for the endpoint measure in B.

Formation of NO₂-cLA

HPLC-MS/MS analysis of the structural regioisomers, β-oxidation metabolites, Michael addition products and pharmacokinetics of ¹⁵NO₃⁻ and ¹⁵NO₂⁻-induced fatty acid nitration in Trials 1 and 2 has been reported³⁴. Endogenous plasma ¹⁴NO₂-cLA concentrations averaged 1.58 ± 0.2 nM after ¹⁵NO₃⁻ and ¹⁵NO₂⁻ and rose significantly to 2.84 ± 0.24 nM upon cLA supplementation. Mean plasma ¹⁵NO₂-cLA concentrations were maximal at 24 hr after ¹⁵NO₃⁻ + cLA (3.06 ± 1.79 nM), and ¹⁵NO₂⁻ + cLA supplementation resulted in maximal ¹⁵NO₂-cLA concentrations that were sustained between 1–6 hr (6.61 ± 4.84 nM to 4.86 ± 2.39 nM). These ¹⁵NO₂⁻ related levels paralleled the levels of free plasma cLA. These data support that free cLA concentrations are limiting for fatty acid nitration. A representative LC-ESI-MS/MS chromatogram of one volunteer’s plasma lipid extract (Figure 5) shows the ¹⁵NO₂-cLA regioisomers generated upon dosing with ¹⁵NO₂⁻ + cLA³⁴.

Figure 5. NO₂⁻ signaling and NO₂-cLA formation.

A representative LC-ESI-MS/MS chromatogram of plasma lipid extract shows ¹⁵NO₂-cLA present in plasma of volunteers treated with oral ¹⁵NO₂⁻ plus cLA (Adapted with permission from Delmastro-Greenwood et al. Nitrite and Nitrate-Dependent Generation of Fatty Acid Nitroalkenes. Free Radic Biol Med. 2015;89:333–341.).

Physiologic responses to ¹⁵NO₃⁻ and ¹⁵NO₂⁻ supplementation with and without cLA supplementation

In Trial 1, oral ¹⁵NO₂⁻ dosing induced vasodilation, decreasing systolic (Δ−12.1±3.3 mmHg, p<0.001), diastolic (Δ−8.1±2.0 mmHg, p=0.002) and mean arterial pressure (Δ−10.4±1.2 mmHg, p=0.001) (Figure 6A and B open circles) with no differences in heart rate (data not shown). In this ¹⁵NO₂⁻ dosed cohort, the reduction in SBP and MAP correlated with the change in plasma NO₃⁻ concentration (Figure S4A and C). Also, the greater the reduction in DBP at 1 hr, the greater the rise in total NO₃⁻ concentration at 0.5 hr (Figure S4B). Notably, the significant decreases in SBP and MAP that persisted for up to 2 hr after subjects were treated with ¹⁵NO₂⁻ were abolished upon co-administration of cLA in Trial 2 (Figure 6A and B closed circles).

Figure 6. Physiologic effects and NO₂⁻ signaling without and with cLA.

Significant decreases in SBP (A) and MAP (B) but not DBP that persisted through 1.5 hr after subjects were treated with oral ¹⁵NO₂⁻ alone (open circles) were abolished by cLA supplementation (closed circles). (C) No differences in the % of platelet activation were detected when comparing oral ¹⁵NO₂⁻ treatment alone (open bars) versus oral ¹⁵NO₂⁻ with cLA (dark bars) over 24 hr. 2×2 repeated measures ANOVA with time as the within-subject effect and trial drug(s) (without vs. with cLA) as the between subject effect was used to compare the endpoint measures between Trial 1 and Trial 2 in A–C.

Inhibition of platelet activation in whole blood was observed 6 hr following ¹⁵NO₂⁻ dosing (baseline 33.2 ± 7.1%, 6 hr post-dosing 7.8 ± 1.6%, p=0.02, Figure 6C open bars), an effect that was eliminated when subjects were dosed with ¹⁵NO₂⁻ + cLA, (Figure 6C dark bars). Following ¹⁵NO₃⁻ dosing in all 10 subjects who completed Trial 1, there was significant inhibition of platelet activation 6 hr after ¹⁵NO₃⁻ administration (17.0 ± 4.1%, 6 hr post-dosing 9.0 ± 1.7%, p=0.029, Figure S5A). With co-administration of ¹⁵NO₃⁻ + cLA in Trial 2, there was no significant inhibition of platelet activation in whole blood at 6 hr (27.5 ± 12.3%, 6 hr post-dosing 20.0 ± 7.5%, Figure S5B).

In subjects given oral ¹⁵NO₃⁻, plasma ¹⁵NO₂⁻ concentrations were significantly increased from 1–2 hr post dosing (Figure S1, open circles), with a maximum of 1.40 μM reached. In Trial 1, oral ¹⁵NO₃⁻ dosing induced a small reduction in DBP at 3 hr compared to baseline (ΔDBP = −6.3±1.2 mmHg, p=0.003, Figure S6B open circles) with no significant change in SBP and MAP (Figure S6A and C open circles) or heart rate (data not shown). No significant correlations were noted between changes in SBP, DBP or MAP and absolute NO₃⁻ or NO₂⁻ concentrations after ¹⁵NO₃⁻ administration (not shown). Following ¹⁵NO₃⁻ + cLA dosing, there were no significant changes in SBP, DBP and MAP (Figure S6A, B and C closed circles) or heart rate (data not shown).

Discussion

There are a panoply of physiological reactions that nitrogen oxides can undergo, all uniquely influenced by factors such as changes in pH, oxygen tension, CO₂/H₂CO₃ levels, rates of production of reactive inflammatory mediators and redox-reactive metalloprotein levels. The relative significance of these mitigating factors will change as basal metabolism transitions to metabolic stress and inflammatory responses. This will critically impact the specific reactions that occur between ^·NO/NO₂⁻, molecular oxygen, partially reduced oxygen species and metal centers. In the context of ^·NO signaling, the physiological outcomes of these reactions are manifested by the diversion of ^·NO away from canonical guanylate cyclase activation and cGMP-dependent signaling responses to the generation of a spectrum of highly reactive and transient secondary oxidizing, nitrosating and nitrating products (Figure 1A)⁵¹. At high concentrations, these reactive products may be pathogenic, but under physiological conditions these chemically-reactive intermediates a) represent a metastable ^·NO reserve that still signals via guanylate cyclase activation and cGMP-dependent mechanisms and b) react with and modify the structure and function of cell targets (e.g. unsaturated fatty acids, proteins). These latter reactions instigate an array of PTMs and non-cGMP-dependent signaling responses³¹. In particular, the nucleophilic amino acid Cys confers proteins with a sensitivity for reduction-oxidation (redox)-induced PTMs that include Cys oxidation, glutathionylation, S-nitrosation, and alkylation upon Michael addition by electrophilic species⁵². These PTMs intimately link metabolic and inflammatory status with changes in cell and organ function, since many enzymes, receptors and transcriptional regulatory proteins that regulate metabolism and inflammation are endowed with functionally-significant hyperreactive Cys moieties. Herein, we report that oral co-administration of 3 g of the polyunsaturated fatty acid cLA with ¹⁵NO₂⁻ significantly redirects the vasodilatory and platelet inhibitory actions of NO₂⁻ to alternative pathways.

Nitrite and its reduction product ^·NO are typically viewed to induce vasodilation via cGMP-dependent mechanisms^{1, 2}. In the human brachial artery, forearm blood flow increases after NO₂⁻ infusion, in concert with a simultaneous rise in ^·NO formation¹. Abundant clinical studies have demonstrated significant reductions in BP in healthy adults following both single ingestion^{11, 53}, repetitive 15 day^{22, 54, 55} courses of NO₃⁻ rich foods⁵⁶ or consumption of a traditional Japanese diet⁵⁷, another rich source of dietary NO₃⁻. Various dosages of pure NO₃⁻ and NO₂⁻ preparations also significantly reduced BP in healthy adults^53,58,59. These human studies have largely been limited to testing single or short-term doses of NO₃⁻ or NO₂⁻, with no evaluation of downstream metabolites. Animal models also recapitulate canonical vasodilatory and BP lowering effects of ^·NO stemming from NO₂⁻ and NO₃^{−60, 61}. There are no statistically significant differences between RS-NO concentrations with ¹⁵NO₂ alone vs. ¹⁵NO₂ + cLA, with peak concentrations of ¹⁵NO-Hb being ~150 times greater than peak concentrations of RS-NO. Thus, these human NO₂⁻ PK data collectively demonstrate elevated plasma ¹⁵NO₂⁻ concentrations (Figure 2D) and ¹⁵NO₂⁻ reduction to ¹⁵NO^· (Figure 3D) in vivo as the most likely mediators of the simultaneous systolic and mean arterial blood pressure reductions upon ¹⁵NO₂ administration (Figure 6A and B, open circles). Oral administration of ¹⁵NO₂ in concert with cLA blunted hemodynamic responses (Figure 6A and B, closed circles), with no detectable formation of ¹⁵NO-Hb adducts. Our lack of blood pressure response with cLA addition does not indicate that our nitrite drug dose is not vasodilating, but that we are reducing the nitrite concentrations enough to reduce the blood pressure lowering extreme. A reduction in blood pressure requires a systemic increase in blood flow (or a decrease in systemic vascular resistance - SVR) that cannot be compensated by an increase in stroke volume and heart rate (MAP - CVP = CO x SVR). We have previously measured vasodilation in the human forearm at nitrite concentrations as low as 150–200 nM (0.1–0.2 uM)¹.

Oral administration of single doses of ¹⁵NO₃⁻ and ¹⁵NO₂⁻ gave detectable increases in plasma nitrogen oxide concentrations, with each nitrogen oxide displaying unique pharmacokinetic and metabolite profiles. Moreover, co-administration of cLA significantly affected these parameters. Oral ¹⁵NO₃⁻ reached its highest plasma NO₃⁻ concentration 1 hr after dosing and yielded a peak concentration of plasma NO₂⁻ 2–6 hr after dosing (Figure 2A and C). This concentration of plasma NO₂⁻ with oral ¹⁵NO₃⁻ was not sufficient to increase MetHb or to yield detectable red blood cell ¹⁵NO-Hb levels (Figure 3A and D) or to induce (Figure S6A and C) or sustain (Figure S6B) reductions in BP parameters. Dosing of ¹⁵NO₃⁻ in concert with cLA suppressed peak plasma NO₃⁻ levels by ~30% (Figure 2E closed circles). For oral ¹⁵NO₂⁻, the highest plasma concentration of NO₂⁻ was measured 0.5 hr after dosing (Figure 2D), giving a modest but significant increase in plasma NO₃⁻ metabolite concentrations 2–3 hr later (Figure 2B). Notably, the co-administration of cLA with ¹⁵NO₂⁻ also suppressed peak plasma NO₂⁻ concentrations by ~30% (Figure 2F closed circles).

Nitrite reacts with oxyHb to form NO₃⁻ and MetHb, and with deoxyhemoglobin to ultimately form NO-Hb¹⁶. Upon NO₂⁻ protonation to HNO₂ and reaction with ^·NO, both symmetric (ONONO) and asymmetric (OONNO) dinitrogen trioxide intermediates are generated, serving as proximal mediators of biomolecule nitrosation and nitration^{20, 62}. After ¹⁵NO₂⁻ administration, evidence of all of these reactions was noted, with a significant 25% increase in MetHb levels, the formation of ~1 μM paramagnetic ¹⁵NO-Hb species and a 3 nM increase in mean plasma protein RS-NO levels 0.5 hr post-dosing (Figures 3B and D, 4A and B). With cLA supplementation, mean plasma NO₂-cLA concentration increased by ~5 nM³⁴. Only a few studies have reported changes in plasma ^·NO concentrations at baseline and following dietary NO₃⁻ or NO₂⁻ supplementation. For example, after a single 5.6 mmol dose of NO₃⁻ rich beetroot juice⁵³, plasma cGMP, an indicator of ^·NO generation, increased within 3 hr. Following a single 24 mmol dose of KNO₃, plasma cGMP increased within 3 hr and remained elevated through 24 hr⁵³.

Following entero-salivary reduction of NO₃⁻ to NO₂⁻, NO₂⁻ is readily protonated (pKa=3.4) to nitrous acid (HNO₂) in the acidic gastric compartment. This promotes a “redirection” of nitrite chemistry where HNO₂ in turn gives rise to secondary N₂O₃ species that mediate nitrosation, nitration and oxidation of susceptible targets. These reactions can be blunted by administering proton pump inhibitors such as esomeprazole, thus inhibiting gastric acid secretion and NO₂⁻ protonation to HNO_2, – a pH response that typifies the pharmacologic action of PPIs⁶³. Esomeprazole also inhibited the blood pressure-lowering effects of nitrite. The more basic conditions in the stomach after esomeprazole administration limited the protonation of nitrite to HNO₂, a species that undergoes both dismutation and nitric oxide reactions to yield the nitrosating and nitrating products symmetric and asymmetric dinitrogen trioxide (ONONO and ONNOO)²⁰. The pathways leading to NO₂-cLA formation first require entero-salivary reduction of NO₃⁻ to NO₂⁻. In the present study, where nitrite and nitrate were orally consumed in the presence of cLA, stomach pH would either remain the same or be lowered by the acidic cLA (pKa = 4.0). The suppression of detectable plasma NO₃⁻ and NO₂⁻ by ~30% upon cLA co-administration indicated that the greater availability and reactivity of cLA favored its nitration, thus supporting the generation of NO₂-cLA derivatives rather than the accumulation of NO₃⁻ and NO₂⁻ in plasma and presumably other tissue compartments. Our present report supports the concept that the addition of cLA and its facile nitration by nitrite-derived species is a consequence of these acid-catalyzed reactions. Our data indicates that this concomitantly leads to an attenuation of NO-forming reactions that would otherwise result in the vasodilation and inhibition of platelet function that was observed when nitrite alone was administered. There may also have been an impact of cLA on gut nitrogen oxide absorption or microbial NO₃⁻ reduction. The small increase in plasma RS-NO derivatives observed in the ¹⁵NO₂⁻ treated cohort in Trial 1 at 0.5 hr post-administration (Figure 4B) was not significantly different compared to Trial 2 with addition of cLA (Figure S3B). Dosing of volunteers with ¹⁵NO₃⁻ with cLA had no impact on plasma RS-NO levels (Figure S3A).

In healthy, hypertensive and high BMI subjects, cLA supplementation (3–6.8g/day, 5–26 wk), compared to control or placebo displays no significant impact on BP in humans^64,65,66,67. Herein, we observed that the decreased BP induced by NO₂⁻ was abrogated by concomitant cLA administration (Figure 6AB). In addition to promoting a shift in nitrosative and nitrative chemistries, cLA might also be modulating NO₂⁻ responses by impacting endogenous ^·NO production^{68,69, 70}. The present data do not diminish the significance of heme proteins such as deoxyHb in mediating reactions of nitrite that lead to changes in vascular function. Rather, our results further affirm the impact that other biomolecules such as conjugated diene-containing fatty acids that readily undergo radical addition reactions, can have on nitrogen oxide reaction pathways and downstream signaling responses. This is manifested by the fact that Fe-NO Hb levels in nitrite-treated subjects are decreased to undetectable levels upon nitrite + cLA administration (EPR analysis has a limit of quantitation of ~500 nM for Fe-NO complexes). These findings indicate that cLA supplementation and other dietary constituents have the ability to modulate the biochemical reactions of NO₂⁻, the trafficking of downstream nitrogen oxide metabolites and the physiological concentrations of this pluripotent signaling mediator. Moreover, a more chronic exposure to elevated cLA and fat in the diet has the potential to also influence the entero-salivary microbiome and its impact on nitrogen oxide metabolism⁷¹.

Along with vasodilatory actions, ^·NO formed by the reduction of NO₂⁻ inhibits platelet function, including platelet adhesion to the endothelium⁴⁷ and platelet aggregation⁷². In contrast, partially reduced oxygen species such as O₂^·− and H₂O₂ propagate platelet activation^{73, 74}. The mechanism by which ^·NO affects platelet activity is by activating cGMP-dependent signaling. While dietary NO₃⁻ sources (beetroot juice or KNO₃⁻) reduce platelet activation or aggregation^{11, 29, 75, 76}, to date only in vitro studies of human platelet function have addressed the effects of NO₂⁻ sources on platelet activation or aggregation^{28, 29, 77, 78}. Herein, the inhibition of platelet activation occurred in whole blood for 6 hr following ¹⁵NO₂⁻ dosing, with the inhibition of platelet activation still evident well beyond the physiologic plasma NO₂⁻ peak and the time of ¹⁵NO-Hb detection. Previous in vitro studies revealed that two synthetic nitro-fatty acids, nitro-linoleic acid and nitro-arachidonic acid (at high concentrations in buffered saline), inhibited platelet aggregation and activation, respectively^{79, 80}. In the more clinically-relevant study herein, ¹⁵NO₂⁻ administration in vivo decreased platelet activation in whole blood at 6 hr. The whole blood basal platelet activation levels of subjects were moderately increased, which are attributed to an artifact of the venipuncture sampling. There may be a threshold plasma nitrite concentration where sufficient ^·NO formation subsequently reduces platelet activation with ¹⁵NO₂⁻ dosing, whereas oral ¹⁵NO₂⁻ and ¹⁵NO₃⁻ in combination with cLA diverted nitrogen oxide metabolites to support the formation of ¹⁵NO₂-cLA, thus consistently attenuating NO₂⁻ and NO₃⁻ inhibition of platelet activation.

Study Limitations and Strengths

The data presented here cannot exclude dietary cLA contribution during Trial 1 and dietary NO₃⁻ and NO₂⁻ contributions also influencing baseline plasma NO₃⁻ and NO₂⁻ concentrations in both trials. Additionally, no formal power calculations were performed as these were exploratory Phase I studies. While this study may be underpowered to detect an effect of ¹⁵NO₃⁻ on BP, given that there were significant reductions in all BP parameters with ¹⁵NO₂⁻, the abolishment of the ¹⁵NO₂⁻ BP effect with the addition of cLA can be assumed to be due to the impact of cLA and not insufficient power. Traditional human NO₃⁻ intakes of 40–100 mg/day^{81, 82}, NO₂⁻ intakes of 0–20 mg/day⁸³ and cLA intakes of 150–200 mg/day⁸⁴ are quite low, but daily intakes of up to 2.6 g of cLA are reported in the literature. Therefore, it may be difficult to achieve all three of these intakes simultaneously through diet on a daily basis, so as to be comparable to those provided in capsule formulations herein. However, our goal with cLA supplementation was to define whether the biochemical reactions and physiologic responses to oral ¹⁵NO₃⁻ and ¹⁵NO₂⁻ were modulated by cLA. Our investigations are strengthened by utilizing paired PK studies in the same subjects so direct comparisons could be made using ¹⁵N sodium NO₃⁻ and NO₂⁻ with and without cLA supplementation to track ^·NO metabolism and the source of ^·NO and NO₂-cLA species formation in vivo with each drug.

Perspectives

The oral administration of cLA altered the metabolic fate of oral ¹⁵NO₃⁻ and ¹⁵NO₂⁻ and modulated the ^·NO-signaling and vasodilatory properties of these nitrogen oxides. The cGMP-dependent signaling actions of NO₃⁻ and NO₂⁻ can be further transformed by digestive reactions, the entero-salivary microbiome and metalloprotein-catalyzed reactions to yield ^·NO and additional nitrogen oxides that elicit redox signaling responses via nitration, nitrosation and oxidation reactions. One would expect that the spectrum of nitrogen oxide products formed will depend on diet, metabolic and inflammatory status, acidic microenvironments and NO₃⁻-reducing salivary bacteria. The concurrent administration of cLA in the diet decreased peak plasma NO₃⁻ and NO₂⁻ concentrations, enhanced the formation of ¹⁵NO₂-cLA, and decreased detectable ^·NO concentrations in the vascular compartment, thus effectively diverting nitrosative reactions to nitration events. This diversion of nitrosative reactions to nitrative chemistries resulted in the increased generation of NO₂-FA, with concomitant attenuation of the inhibition of platelet activation and abrogation of the vasodilatory properties of NO₂⁻. This affirms that the metabolic and the physiological responses to oral NO₃⁻ and NO₂⁻ can be significantly modulated by the composition of the diet, particularly with meat and dairy products that are rich in cLA.

Novelty and Significance.

What is new?

We developed oral capsule formulations of sodium nitrate (NO₃⁻) and nitrite (NO₂⁻) with the stable [¹⁵N] isotope allowing for specific metabolite tracking in vivo and to investigate the human vascular responses.

What is relevant?

• Oral ¹⁵NO₂⁻ consumption resulted in ^·NO formation, vasodilation and acute inhibition of platelet activation

• Conjugated linoleic acid (cLA) with oral ¹⁵NO₃⁻/NO₂⁻ diverted metabolic products to NO₂-cLA

• Addition of dietary cLA with oral ¹⁵NO₂⁻ decreased ^·NO formation and vasodilation and attenuated the acute inhibition of platelet activation

Summary

Concurrent cLA in the diet altered the metabolic fate of oral ¹⁵NO₃⁻ and NO₂⁻ with decreased plasma NO₃⁻/NO₂⁻ levels, enhanced formation of NO₂-FA species and decreased ^·NO formation. This diversion of the downstream reactions of NO₃⁻ and NO₂⁻ to NO₂-cLA attenuated the inhibition of platelet activation and abrogated the vasodilatory properties of NO₂⁻.