Electrophilic characteristics and aqueous behavior of fatty acid nitroalkenes

Abstract

Fatty acid nitroalkenes (NO₂-FA) are endogenously-generated products of the reaction of metabolic and inflammatory-derived nitrogen dioxide (^.NO₂) with unsaturated fatty acids. These species mediate signaling actions and induce adaptive responses in preclinical models of inflammatory and metabolic diseases. The nitroalkene substituent possesses an electrophilic nature, resulting in rapid and reversible reactions with biological nucleophiles such as cysteine, thus supporting post-translational modifications (PTM) of proteins having susceptible nucleophilic centers. These reactions contribute to enzyme regulation, modulation of inflammation and cell proliferation and the regulation of gene expression responses. Herein, focus is placed on the reduction-oxidation (redox) characteristics and stability of specific NO₂-FA regioisomers having biological and clinical relevance; nitro-oleic acid (NO₂-OA), bis-allylic nitro-linoleic acid (NO₂-LA) and the conjugated diene-containing nitro-conjugated linoleic acid (NO₂-cLA). Cyclic and alternating-current voltammetry and chronopotentiometry were used to the study of reduction potentials of these NO₂-FA. R–NO₂ reduction was observed around −0.8 V (vs. Ag/AgCl/3 M KCl) and is related to relative NO₂-FA electrophilicity. This reduction process could be utilized for the evaluation of NO₂-FA stability in aqueous milieu, shown herein to be pH dependent. In addition, electron paramagnetic resonance (EPR) spectroscopy was used to define the stability of the nitroalkene moiety under aqueous conditions, specifically under conditions where nitric oxide (^.NO) release could be detected. The experimental data were supported by density functional theory calculations using 6–311++G (d,p) basis set and B3LYP functional. Based on experimental and computational approaches, the relative electrophilicities of these NO₂-FA are NO₂-cLA >> NO₂-LA > NO₂-OA. Micellarization and vesiculation largely define these biophysical characteristics in aqueous, nucleophile-free conditions. At concentrations below the critical micellar concentration (CMC), monomeric NO₂-FA predominate, while at greater concentrations a micellar phase consisting of self-assembled lipid structures predominates. The CMC, determined by dynamic light scattering in 0.1 M phosphate buffer (pH 7.4) at 25 °C, was 6.9 (NO₂-LA) 10.6 (NO₂-OA) and 42.3 μM (NO₂-cLA), respectively. In aggregate, this study provides new insight into the biophysical properties of NO₂-FA that are important for better understanding the cell signaling and pharmacological potential of this class of mediators.

Highlights

• •Nitro-fatty acids (NO₂-FA) are electrophilic mediators of redox-regulated signaling reactions.
• •The relative electrophilicity of 18 carbon chain length NO₂-FA is NO₂-cLA » NO₂-LA > NO₂-OA.
• •Critical micellar concentrations at pH 7.4 were 10.6, 6.9 and 42.3 μM for NO₂-OA, NO₂-LA and NO₂-cLA, respectively.
• •Nitrogen dioxide (^.NO₂) rapidly reacts with unsaturated FA and yields more stable bioactive NO₂-FA byproducts.

Abbreviations

ACV

alternating-current voltammetry

cLA

conjugated-linoleic acid

CMC

critical micellar concentration

CPSA

chronopotentiometric stripping analysis

CV

cyclic voltammetry

DFT

density functional theory

DLS

dynamic light scattering

DTCS

N-(dithiocarboxy)sarcosine

HMDE

hanging mercury drop electrode

EPR

electron paramagnetic resonance

FA

fatty acids

LA

linoleic acid

MM

molecular modelling

NO₂-cLA

nitro-conjugated-linoleic acid

NO₂-LA

nitro-linoleic acid

NO₂-OA

nitro-oleic acid

OA

oleic acid

PGE

pyrolytic graphite electrode

PMT

post-translational modification

QM

quantum mechanics

RS

reactive species

TCEP

tris(2-carboxyethyl)phosphine hydrochloride.

1. Introduction

Elevated levels of nitrated amino acids, nucleotides and lipids have been detected in tissues and fluids of patients undergoing inflammatory conditions [1,2]. These derivatives are generated in vivo and in vitro upon exposure to nitric oxide (·NO)- and nitrite (NO₂⁻)-derived RS that yield nitrogen dioxide (NO₂). For example, ·NO and NO₂⁻ are precursors for the generation of peroxynitrous acid (ONOOH), nitrosoperoxocarbonate (ONOOCO₂) and dinitrogen trioxide (N₂O₃). Also, NO₂⁻ can be oxidized by heme peroxidases to yield ^.NO₂ [[3], [4], [5]]. When ·NO₂ reacts with the double bond of an unsaturated acyl chain, forming a nitroalkene substituent, the products have generically been termed nitro-fatty acids (NO₂-FA). NO₂-FA found in living organisms are a result of the nitration of oleic (C18:1) [6], linoleic (C18:2) [7], linolenic (C18:3) and arachidonic (C20:4) [8] acids. Current data supports that NO₂-FA display anti-inflammatory and adaptive signaling actions. The administration of synthetic natural and non-natural nitroalkenes also can mitigate pathogenic responses in preclinical models of cardiovascular, ischemic, metabolic, inflammatory and fibrotic diseases [[9], [10], [11], [12]].

NO₂-FA undergo reversible Michael addition [13] with cellular nucleophiles, predominantly Cys, that frequently reside in functionally-significant cell signaling, metabolic and transcriptional regulatory proteins. The nitro group on the alkene moiety has a strong electron-withdrawing behavior that confers electron deficiency to the β-carbon and responsiveness to nucleophilic attack by thiolates. In nitro-conjugated linoleic acid, both the β- and δ-carbons with respect to the nitro group are susceptible to Michael addition [14]. The β-adduct is the kinetic product while the δ-adduct is the thermodynamic product. This Michael addition is reversible and kinetically rapid, which differentiates the NO₂-FA from other endogenous signaling electrophiles (such as aldehydic fatty acid oxidation products and cyclopentenone prostaglandins) that react slowly and irreversibly with nucleophiles [15,16].

NO₂-FA show NO donor activities in vitro [[17], [18], [19]]. The mechanism(s) underlying aqueous NO release by nitroalkenes has not been definitively established, but may include a modified Nef reaction. A possible mechanism is that under neutral aqueous conditions, the nitroalkene facilitates the formation of the nitronate anion; then followed by protonation and deprotonation steps resulting in a nitroso intermediate. This nitroso intermediate has a weak C–N bond that yields ·NO [20]. The presence of the nitrite radical was supported by chemiluminescence measurements and EPR-based methodologies [21]. It is noted that in these aqueous-based NO donor studies and buffer-perfused vascular rings that claim a vasodilatory action of NO₂-FA, experimental results might have been complicated by a) contaminating nitrite in NO₂-FA preparations or b) unique reactions that occur in pure aqueous milieu as opposed to those operative in more complex biological fluids and tissues. For example, extensive preclinical pharmacokinetics and toxicology studies, as well as human Phase 1 and Phase 2 trials do not show an acute effect of intravenous or oral NO₂-FA on blood pressure or pulse rate, suggesting that NO and cGMP-dependent signaling mechanisms have not been engaged (unpublished). In any event, these previous studies motivated the present more detailed analysis of the kinetics of NO generation by fatty acid nitroalkenes and mitigating factors.

The endogenous formation of NO₂-FA [[3], [4], [5]] will be strongly influenced by diet, which may contain pre-formed NO₂-FA or a combination of precursor fatty acids and sources of the organic nitrates required for fatty acid nitration (i.e., NO₂⁻ coming from animal and plant-derived products or NO₂⁻-based preservatives) [22]. Clinical studies have revealed that dietary supplementation of organic nitrates, with and without co-administration of conjugated linoleic acid (cLA), resulted in a sustained level of NO₂-cLA only when cLA was co-administered [23]. This indicates that cLA is rate limiting and its simultaneous presence in the gastric compartment together with NO₂⁻ and its precursor nitrate (NO₃⁻) is needed for the formation of the NO₂-cLA. Related to this, the cardiovascular health-promoting properties attributed to the Mediterranean diet may be due in part to these dietary constituents that could react and upregulate NO₂-FA generation [[24], [25], [26]]. Thus, interest in characterizing the biophysical and biochemical properties of NO₂-FA has grown [27].

The detailed formation, metabolism, distribution and excretion of NO₂-FA, as well as the mechanisms accounting for net signaling responses to NO₂-FA are still not fully resolved [7,20,28]. From a methodological point of view, NO₂-FA are complex analytes because of aqueous insolubility, an amphiphilic nature, a net negative charge and different chemical reactivities in aqueous and lipophilic biological microenvironments. MS-based approaches [29] as well as spectrophotometry and crystallographic analysis [21] have been used for structural characterization and quantification of unesterified NO₂-FA, with mass spectrometry using electrospray ionization (ESI) coupled with liquid chromatography (LC-MS) being the most suitable method [20,30].

Herein, we expand on a previous study [31] and compare the biophysical properties of NO₂-OA, NO₂-LA and NO₂-cLA (Scheme 1) using multiple analytical and computational approaches. Native FA without NO₂ substitution (OA, LA and cLA) served as negative controls in all experiments. These analyses define the reactivity, stability, NO release characteristics, conformational properties and micellarization of key NO₂-FA.

Scheme 1.

Structures and identification of NO₂-FA used in this study.

2. Experimental

2.1. Chemicals

Chemicals were purchased from Merck KGaA (Darmstadt, Germany) or BioRad Laboratories (Hercules, CA, USA). 10-nitrolinoleate (NO₂-LA) as pure positional isomer was from Cayman Chemical (Ann Arbor, MI, USA). NO₂-OA as an equimolar mixture of the 9- and 10-nitrooctadec-(9)-enoic acid and NO₂-cLA as an equimolar mixture of the 9- and 12-nitrooctadeca-(9,11)-dienoic acid were synthesized as previously [32]. The purity of NO₂-FA was higher than 98% for all samples. The stock solutions were prepared using methanol. With exception of NO₂-LA, NO₂-OA and NO₂-cLA were used throughout the study as an appropriate mixture of both positional isomers (Scheme 1). The unmodified fatty acids (LA, cLA and OA) were purchased from Merck KGaA (Darmstadt, Germany). All solutions were prepared using Milli-Q water (18.2 MΩ cm^{– 1}), Millipore, Bedford, MA, USA. All other used chemicals were of analytical grade.

2.2. CMC determination by DLS

The micelles were prepared by dissolving individual fatty acids and their nitro derivatives in 1 mM NaOH solution [33]. Each solution was diluted to the final concentration of 100 μM fatty acid (as sodium salt) with 0.1 M phosphate buffer (pH 7.4). Samples were diluted to several lower concentrations before performing the experiments. Dynamic light scattering (DLS) measurements were performed using Zetasizer Nano ZSP (Malvern Panalytical Ltd, Malvern, UK) at 25 °C. Finally, light scattering intensities (Derived Count Rate, Kcps) were plotted against solute concentration. Critical micellar concentration (CMC) was determined as the intersection between the straight line coming from the light scattering intensity values below CMC with the straight line derived from the light scattering intensities above CMC (region of rapid intensity increase) [34].

2.3. Electrochemistry

All electrochemical measurements were performed at room temperature with a μAutolab III analyzer (EcoChemie, Utrecht, Netherlands) in a three-electrodes cell with an Ag/AgCl/3 M KCl electrode as the reference and platinum wire as the auxiliary electrode. Two types of working electrodes were used: HMDE (hanging mercury drop electrode; area 0.4 mm²) for constant-current chronopotentiometric stripping analysis (CPSA) and alternating-current voltammetry (ACV), and PGE, a basal-plane pyrolytic graphite electrode (electrode area 9 mm², Momentive Performance Materials, USA) for cyclic voltammetry (CV). Individual settings for electrochemical experiments, as well as concentrations of the compounds, are given in the Fig. legends. The electrochemical analyses were performed using two supporting electrolytes: 0.1 M phosphate and Britton-Robinson buffers [35]. For the voltammetry of NO₂-FA, all electrolytes were purged with argon. Degassing was not employed with CPSA because the dioxygen does not interfere with the technique. The pH measurements were performed with a HI 2211 pH/ORP Meter (HANNA instruments, IT).

2.4. Stability of NO₂-FA

The stability of 8 μM NO₂-FA solutions was investigated following the decrease in intensity of the NO reduction peak for more than 24 h by using CPSA. The compounds investigated were dissolved in Britton-Robinson buffer as supporting electrolyte at pH 5, 7.4 and 9 in the presence of atmospheric oxygen. Incubation and CPSA were performed at 25 °C (controlled by thermostat) and room temperature, respectively.

2.5. Detection of ·NO release from NO₂-FA

In order to study whether NO₂-FA are able to release ·NO, the Fe(II) (DTCS)₂ complex consisting of Fe²⁺ and N-(dithiocarboxy)sarcosine, DTCS (Chemos Cz, Prague, Czech Republic) was used as an ·NO- trapping agent. For all three of NO₂-FA, 150 mM ethanol stock solutions were prepared. DTCS was dissolved in water and bubbled with N₂ for at least 30 min to remove O₂. For the purpose of EPR measurements, water solutions of FeSO₄ × 7H₂O (final concentration 4 mM) and DTCS (final concentration 5.5 mM) were added to Britton-Robinson buffer (pH 7.4), followed by adding small volume of NO₂-cLA, NO₂-LA or NO₂-OA ethanol stock solutions (final concentration 25 mM). Final samples were drawn into gas-permeable Teflon tubes (Zeus Industrial Products, Orangeburg, SC, USA), and placed into the SHQE EPR resonator cavity. Between the measurements, all of the samples were kept at 25 °C. Since ·NO production from NO₂-OA has already been inspected [31], previous work was used as a reference.

2.6. EPR data acquisition and processing

EPR spectra were recorded using a Bruker Elexsys II E540 EPR X-band spectrometer (Germany) at 9.85 GHz and 10 mW microwave power. For all measurements, the modulation amplitude was 2 G, modulation frequency 100 kHz, conversion time 14.65 ms and sweep width was 77 G.

2.7. Theoretical calculations

Since the finding of global energy minimum for OA by systematical torsion search was found to be too time consuming and impractical in our previous work [31], a limited set of twenty energetically favorable structures was generated by program FROG2 [36]. This program allows the generation of conformation ensembles of small molecules using a two stage Monte Carlo approach in the dihedral space from initial structure. The conformational search was limited to desired (E-, Z-) isomers of anionic form of studied FA. The retrieved conformers were further reoptimized at DFT level of theory employing 6–311++G (d, p) basis set and B3LYP functional. Analysis of frontier molecular orbitals and molecular electrostatic potentials were performed using the program Avogadro [37]. The role of solvent (water, methanol and n-octanol) was described in implicit way using PCM model [38]. All quantum mechanical calculations were treated in Gaussian 16 program [39].

3. Results and discussion

3.1. CMC determination by DLS

Fatty acids are amphiphilic molecules with an aliphatic tail which can be either saturated or unsaturated and have a polar headgroup that can be protonated (COOH) or deprotonated (COO⁻). When FA are present in aqueous solution in sufficient amounts near or above their respective critical micellar concentrations (CMC), they can assemble to form micelles, bilayer membranes and liquid-crystalline phases [40]. The formation of these structures is influenced by pH, CO₂ concentration and temperature. The CMC for oleic acid/oleate and linoleic acid/linoleate in aqueous solution are 6 and 13 μM, respectively [41]. Depending on pH, the nature of self-assembled lipid structures is affected. Sodium oleate forms vesicles at low pH and, with increasing pH the formation of micelles predominates [42] (Table 1). The CMC for sodium cLA varies from 420 μM to 200 μM when pH decreases from 13 to 8.6 [43]. These characteristics are accounted for the interpretation of the chemical reactivities and biochemical properties of NO₂-FA.

Table 1.

Schematic representation of the self-assembly of fatty acids with respect to critical micellar concentration (CMC). CMC values for free FA and NO₂-FA were determined using density light scattering.

First, we measured the CMC of native FA and compared it with the corresponding NO₂-FA derivative (Table 1). The nitro substitution increases the CMC for all NO₂-FA investigated herein. Low CMCs were found for OA and LA and their nitro derivatives; while cLA and NO₂-cLA, had significantly greater CMCs, i.e. 22.5 and 42.3 μM, respectively. While the CMC of members of the NO₂-FA class have not yet been reported, it was reported that when NO₂-FA are partitioned in micelles or liposomes. This disposition stabilizes and apparently sequesters the nitro-alkene functional group, as this and limits acid-base catalyzed reactions and concomitant ·NO release and Michael addition of thiols [20,30,44].

3.2. Redox behavior of NO₂-FA

In next experimental work we focused on the redox behavior of NO₂-cLA and NO₂-LA; the redox transformations of NO₂-OA were described recently [31], thus the redox behavior of NO₂-cLA and NO₂-LA was evaluated and compared with NO₂-OA. The redox behavior of NO₂-cLA, an equimolar mixture of the 9- and 12-NO₂ regioisomers was evaluated by cyclic voltammetry (CV) with a basal-plane pyrolytic graphite electrode (PGE) (Fig. 1). Native conjugated-linoleic acid (cLA) was used as a control. The reduction peak (peak NO) of the modified fatty acid is visible at ca. −0.8 V in 0.1 M phosphate buffer at pH 7.4 while for the native fatty acid no peak appears (Fig. 1a).

Fig. 1.

Cyclic voltammetry (CV) of NO₂-cLA at pyrolytic graphite electrode. (a) CV of 20 μM NO₂-cLA and cLA in 0.1 M phosphate buffer at pH 7.4. CV parameters: initial potential 0 V, first vertex potential −1.85 V, second vertex potential +1 V. (b) CV of 20 μM NO₂-cLA in 0.1 M phosphate buffer at pH 7.4. CV parameters: initial potential 0 V, first vertex potentials −1.85 V, −1.50, −1.15, −0.87, −0.77, −0.5 V, second vertex potential +1 V. (c) The pH effect on the reduction NO peak coming from 5 μM NO₂-cLA was recorded in Britton-Robinson buffer at pH 3–12. CV parameters: initial potential 0 V, first vertex potential −1.85 V, second vertex potential +1 V. Inset: dependence of the peak NO potential on pH. Prior to the experiments, the electrolytes were deaerated for 10 min with a stream of argon. Time of accumulation t_A = 30 s (E_A = 0 V), step potential 5 mV and scan rate 1 V/s were applied in all cases. (For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this article.)

This reduction process is irreversible and of an adsorptive nature [31]. The final reduction product for the nitro moiety has not yet been established but it has been proposed to proceed via a hydroxylamine to an amine derivative via a complex reduction cascade [20]. For example, the reduction of cis-nitrostilbene in acetonitrile occurs by initial formation of the radical anion followed by irreversible dimerization [45]. When the trans isomer was studied, a different mechanism was operative, where the initial radical anion rapidly isomerizes to the cis form followed by a partial oxidation to cis-nitrostilbene. This process was then followed by the second-order dimerization [45]. The oxidation of individual reduction products of NO₂-cLA can be demonstrated by switching the first vertex potential in CV experiments (Fig. 1b). After reduction of the NO₂ moiety, the anodic peak appears at ca.+0.3 V, most probably corresponding to the oxidation of the first reduction product in the cascade. If the potential of −0.5 V (or a less negative one) is applied, the anodic peak was not observed in reverse branch of cyclic voltammograms, (black line, Fig. 1b). The NO reduction peak for the NO₂-cLA acid is also pH dependent, as reported for NO₂-OA (Fig. 1c). The peak NO potential is shifted toward negative values with increasing of pH. In the neutral pH range the maximum of peak current intensity for the reduction process was obtained.

The electrochemical behavior of NO₂-cLA was also studied by constant-current chronopotentiometric stripping analysis (CPSA) with a hanging mercury drop electrode (HMDE). These measurements showed the presence of a second peak (Ads) at negative potential ca. −1.3 V (Fig. 2a). Alternating current voltammetry (ACV) at HMDE was used for a more detailed view of the interfacial behavior of NO₂-cLA [46]. Comparison of the out-of-phase (sensitive to adsorption properties) and in-phase (sensible to both diffusion control and adsorption processes) AC voltammograms suggests that peak Ads is most probably due to adsorption/desorption or reorientation processes [31,47]. The out-of-phase ACV shows only a tensametric Ads peak (or double peak) for NO₂-cLA (or cLA). In contrast to in-phase, out-of-phase ACV measurement is not as sensitive to peak NO in agreement with previous findings [31] (Fig. 2b and c). In the same way, the reduction of NO₂-LA was tested and quite similar redox behavior as for NO₂-cLA was found (Fig. 3).

Fig. 2.

Chronopotentiometric stripping analysis (CPSA) and alternating current (AC) voltammetry of NO₂-cLA at hanging mercury drop electrode. All measurements were in 0.1 M phosphate buffer at pH 7.4. (a) CPS record of 10 μM NO₂-cLA (blue line) and cLA (red line); i_str = −35 μA, time of accumulation t_A = 30 s (E_A = 0 V). (b) In-phase and (c) out-of-phase AC voltammograms of 30 μM NO₂-cLA and cLA; initial (0 V) and end (−1.95 V) potentials, frequency: 66.2 Hz, amplitude: 5 mV, phase angle: 0° (in-phase) and 90° (out-of-phase). For AC voltammetry the electrolyte was degassed for 10 min with argon. (For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this article.)

Fig. 3.

Chronopotentiometric stripping analysis (CPSA) and cyclic voltammetry (CV) of NO₂-LA. All analyses were performed in 0.1 M phosphate buffer at pH 7.4. (a) CPSA record of 10 μM NO₂-LA at hanging mercury drop electrode; (i_str −35 μA). (b) CV of 20 μM NO₂-LA at pyrolytic graphite electrode. CV parameters: initial potential 0 V, first vertex potential −1.85 V, second vertex potential +1 V, step potential 5 mV, scan rate 1 V/s. Time of accumulation t_A = 30 s (E_A = 0 V) was used for both panel a and b. (For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this article.)

NO₂-cLA redox properties were compared with the corresponding methylene-interrupted diene species (i.e. NO₂-LA) and to the monoenoic nitroalkene NO₂-OA by using CPSA. The NO₂-cLA reduction peak NO appears at a less negative potential compared with NO₂-LA and NO₂-OA (Fig. 4a, inset). The same behavior was observed for PGE using the CV method (Fig. 4b). CPS and CV measurements indicated that both conjugated and methylene-interrupted diene containing nitro-linoleic acids have stronger electrophilic character compared with NO₂-OA. This is consistent with the concept that the electrophilic (electron acceptor) properties of NO₂-FA increase with the shift of the NO peak towards less negative potentials.

Fig. 4.

Chronopotentiometric stripping analysis (CPSA) and cyclic voltammetry (CV) of NO₂-cLA (red line), NO₂-LA (blue line) and NO₂-OA (black line). All experiments were performed in 0.1 M phosphate buffer at pH 7.4. (a) CPS records of 10 μM NO₂-FA at HMDE; i_str −35 μA. Inset table: NO reduction peak potentials and heights. (b) CVs of 20 μM NO₂-FA at PGE. CV parameters: initial potential 0 V, first vertex potential −1.85 V, second vertex potential +1 V, step potential 5 mV, scan rate 1 V/s. Time of accumulation t_A = 30 s (E_A = 0 V) was used for both panel a and b. (For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this article.)

3.3. Evaluation of NO₂-FA stability

When NO₂-FA electroreduction was inspected over the time, it was apparent that depending on pH, that under aqueous buffer conditions NO₂-FA can undergo decomposition reactions, with a range of pH values (5, 7.4 and 9) showing that the greatest degradation rate was in acidic conditions (Fig. 5). The degradation rate was greater for NO₂-cLA than for NO₂-OA and NO₂-LA, a response consistent with the greater electron acceptor capacity of NO₂-cLA (Fig. 4).

Fig. 5.

Time-dependent stability of NO₂-FA. Data is represented by the decrease of NO peak height of 8 μM NO₂-LA, NO₂-cLA and NO₂-OA in Britton-Robinson buffer at pH (a) 5, (b) 7.4 and (c) 9 at HMDE. CPSA at i_str −35 μA was used in all experiments after a pre-concentration step, time of accumulation t_A = 30 s (E_A = 0 V). (For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this article.)

EPR spectroscopy, using Fe(II) (DTCS)₂ as the spin-trapping agent, revealed rates and extents of NO₂-cLA and NO₂-LA aqueous decay and generation of ·NO, and allowed comparison with NO₂-OA (Fig. 6). By comparing the EPR signal intensities of spin-adducts, NO₂-cLA generated a significantly greater yield of ·NO than NO₂-LA and NO₂-OA. The amount of ·NO produced by NO₂-LA and NO₂-OA was almost equivalent. The rate constant of ·NO release over 24 h could be calculated for all three NO₂-FA by applying a consecutive reaction fitting model to the kinetic data (Fig. 7, Table 2). This is a simplified model, since multiple simultaneous processes are contributing to the overall kinetics. Namely, this simplified model takes into account the ·NO release from the NO₂-FA, the rapid reaction of ·NO with Fe(II) (DTCS)₂, and the decay of the EPR-active complex, resulting in a loss of the EPR signal.

Fig. 6.

EPR spectra of NO–Fe(DTCS)₂ adducts produced by NO₂-FA. Data was obtained after incubation for 2 min and 3 h. (a) NO₂-cLA, (b) NO₂-LA and (c) NO₂-OA.

Fig. 7.

Kinetics of NO–Fe(DTCS)₂ EPR signal intensities. Data was obtained over 24 h. For kinetic measurements, the intensity of the middle EPR peak from the signal triplet was measured. For more details, see the main text.

Table 2.

Rate constants for NO release from three different fatty acid nitroalkenes maintained under aqueous conditions.

NO donor	k1 (arbitrary EPR units/h)
NO2-cLA	1.3·106
NO2-LA	5.5·105
NO2-OA	5·105

One of the causes for the EPR signal decay of the NO–Fe(II) (DTCS)₂ complex may be the reaction between this complex and another ·NO molecule, yielding EPR-silent dinitrosyl species, thus affecting the ·NO trapping efficiency [48]. Also, in spite of all the solutions used for spin trapping procedure were N₂-equilibrated, it is possible that a small fraction of iron from Fe(II) (DTCS)₂ was quickly oxidized, giving the Fe(III) (DTCS)₃, which is also able to complex with ·NO. Notably, NO–Fe(III) (DTCS)₃ can also be converted to NO–Fe(II) (DTCS)₂ [49,50]. Accordingly, the extent of Fe²⁺ oxidation to Fe³⁺ and the rate of subsequent steps resulting in the formation of NO–Fe(II) (DTCS)₂ complex could also determine the reduction rate of the EPR signal. Finally, dithiocarbamate (DTC)-Fe complexes can also be employed for indirect detection of NO₂/N₂O₄ [51] via the following reactions:

Fe(DTC)₃ + N₂O₄ → Fe(NO₃) (DTC)₂ + NO + DTC

Fe(DTC)₃ + NO → NO–Fe(DTC)₂ + DTC

Thus, the NO–Fe(II) (DTCS)₂ complex may arise from both ·NO and ·NO₂ generation by NO₂-FA, especially under aerobic conditions.

3.4. Conformation of NO₂-FA

The MM and QM calculations performed on fatty acids showed that the conformation of aliphatic chain [[52], [53], [54], [55], [56]] is different for saturated versus unsaturated compounds. In saturated chains the anti orientation of the C–C–C–C torsion angles (180°) represents the global minimum, whereas its gauche orientation (60°) is a local minimum, being less stable than the global one by approx. 2 kJ/mol. The skew orientation (120°) belongs to the transition state with a barrier height around 14 kJ/mol.

When a double bond is present in the system, the global minimum is represented by a skew, skew conformation of two C–C–C C dihedrals (i.e. like in 3-hexene), also being the case for dienes. In 2,5-heptadiene, a common model for the flexibility of aliphatic chains of FA [53,55], the +skew, +skew (or -skew, -skew) orientation is a global minimum. A positive value of the torsion angle corresponds to counterclockwise rotation of the furthest substituent with respect to the central bond. In this conformation, double-bond directions are aligned in parallel, which leads to the extended aliphatic chain. Almost the same stability occurs for the +skew, -skew (or -skew, +skew) orientation with the double-bond directions situated perpendicularly, providing more compact (clasped) structure [52].

A nitro substituent does not significantly affect the geometry of aliphatic FA. The most stable conformers of NO₂-FA resemble those of non-nitrated FA [56]. For NO₂-OA and NO₂-LA derivatives, the conformations with +skew, -skew combination of C–C–C C dihedrals are slightly preferred over those with +skew, +skew combinations. This could be due to the favorable dispersion interaction between two parts of the aliphatic chains in clasped structures, in comparison with the extended one, which compensates for the energetic superiority of +skew, +skew conformation in small model systems. The most visible differences in the structure of NO₂-OA compared to OA are smaller absolute values of C–C–C(NO₂) C and C–C–C C(NO₂) dihedral angles in comparison with C–C–C C torsion angle in FA: the first being 90° and the second 111°. An optimal value 116° was found for native FA [55]. The bulky nitro group probably also increases the barrier height between energy minima for both of the aforementioned conformers.

For NO₂-cLA the conformation in the vicinity of double bounds is different, since one cis- and one trans-C C bond are in conjugation. This leads at first to slight non-planarity of these two bonds (deviation 5–10°) with an orientation of C–C–C–C (any bond order, starting from the carboxylate, with dihedrals around double bond depicted in bold).

For 9-NO₂-cLA:

-skew	cis	-gauche	trans	-skew	(more stable)
+skew	cis	-gauche	trans	cis

For 12-NO₂-cLA the conformational variability was found even bigger:

-skew	trans	-gauche	cis	-gauche
cis	trans	-gauche	cis	+skew
cis	trans	+gauche	cis	-skew
+skew	trans	-gauche	cis	-skew

The two most stable energy conformers of NO₂-FA and their physical-chemical properties (relative stability, electrostatic potential, HOMO and LUMO) are depicted in Fig. 8a, Fig. 8b, Fig. 9, Fig. 10b, Fig. 10a and Table 3. From this, it is evident that for all NO₂-FA the shape of HOMO and LUMO orbitals is similar, the HOMO orbital is spread around the carboxylic group, whereas the LUMO is composed from contributions in the π* orbitals of the nitro group and of C C bond(s). The energies of HOMO and LUMO in any given FA are insensitive to the position of the nitro group and its molecular geometry. The increase of polarity of the solvent enlarges the HOMO–LUMO energy gap (by ca. 0.35 eV comparing the n-octanol and water solvents). The lowest HOMO-LUMO energy gap resulting to the highest reactivity (by 0.2 eV in comparison with other NO₂-FA) was found for NO₂-cLA what can be caused by the presence of two conjugated C C bonds in this compound.

Fig. 8a.

Physico-chemical properties of 9-NO₂-OA presented for two most stable conformers (I, II). From top to bottom: molecular electrostatic potential (ESP); the highest occupied molecular orbital (HOMO); the lowest unoccupied molecular orbital (LUMO).

Fig. 8b.

Physico-chemical properties of 10-NO₂-OA presented for two most stable conformers (I, II). From top to bottom: molecular electrostatic potential (ESP); the highest occupied molecular orbital (HOMO); the lowest unoccupied molecular orbital (LUMO).

Fig. 9.

Physico-chemical properties of 10-NO₂-LA presented for two most stable conformers (I, II). From top to bottom: molecular electrostatic potential (ESP); the highest occupied molecular orbital (HOMO); the lowest unoccupied molecular orbital (LUMO).

Fig. 10b.

Physico-chemical properties of 12-NO₂-cLA presented for two most stable conformers (I, II). From top to bottom: molecular electrostatic potential (ESP); the highest occupied molecular orbital (HOMO); the lowest unoccupied molecular orbital (LUMO).

Fig. 10a.

Physico-chemical properties of 9-NO₂-cLA presented for two most stable conformers (I, II). From top to bottom: molecular electrostatic potential (ESP); the highest occupied molecular orbital (HOMO); the lowest unoccupied molecular orbital (LUMO).

Table 3.

Relative energies and HOMO-LUMO characteristics of two most stable conformers of studied NO₂–FAs in different solvents. All calculations were performed at B3LYP/6–311++G (d, p) level of theory. For geometry of conformers see Fig. 8a, Fig. 8b, Fig. 9, Fig. 10b, Fig. 10a.

	Water				Methanol				n-Octanol
	ΔEa	EHOMOb	ELUMOc	LUMO-HOMOd	ΔEa	EHOMOb	ELUMOc	LUMO-HOMOd	ΔEa	EHOMOb	ELUMOc	LUMO-HOMOd
12-NO2-cLA	0.00	−8.287	−0.953	7.334	0.00	−8.191	−0.956	7.235	0.00	−7.813	−0.851	6.962
	2.15	−8.289	−0.981	7.308	2.06	−8.192	−0.954	7.238	1.84	−7.816	−0.851	6.965
9-NO2-cLA	0.00	−8.288	−0.893	7.395	0.23	−8.187	−0.863	7.324	0.58	−7.807	−0.749	7.058
	0.01	−8.302	−0.970	7.332	0.00	−8.197	−0.947	7.250	0.00	−7.811	−0.828	6.983
10-NO2-LA	0.00	−8.297	−0.737	7.560	0.00	−8.197	−0.708	7.489	0.00	−7.817	−0.590	7.227
	7.48	−8.297	−0.804	7.493	7.54	−8.197	−0.776	7.421	8.02	−7.814	−0.665	7.149
9-NO2-OA	0.00	−8.286	−0.774	7.512	0.00	−8.188	−0.745	7.443	0.00	−7.808	−0.621	7.187
	0.02	−8.285	−0.774	7.511	0.02	−8.187	−0.746	7.441	1.08	−7.807	−0.636	7.171
10-NO2-OA	0.00	−8.287	−0.758	7.529	0.00	−8.190	−0.731	7.459	0.00	−7.816	−0.627	7.189
	1.02	−8.286	−0.771	7.515	1.00	−8.190	−0.744	7.446	0.96	−7.815	−0.640	7.175

^a relative energy of the conformers (in kJ/mol).

^b,c,d HOMO, LUMO energy and their difference (all in eV).

3.5. Biological relevance of reported data

Herein three nitro-FA with biological signaling capabilities were biophysically characterized under aqueous conditions. The C-10 isomer of NO₂-OA (CXA-10) is being evaluated in Phase 2 clinical trials as a potential new drug candidate [11]. In addition, we have evaluated NO₂-LA and its conjugated diene conformer, the latter of significant biological relevance because of its endogenous levels [57]. Comparative redox and EPR analysis indicate NO radical release is a spontaneous decay mechanism for NO₂-FA, yet likely does not activate soluble guanylate cyclase to exert physiological responses because of low yields and a kinetically-preferential reaction with nucleophilic amino acid residues [58]. In particular, NO₂-FA form adducts with Cys residues of signaling proteins, redox sensors, proteins involved in stress responses and redox-sensitive transcription factors. The important driving factor for these low molecular weight nucleophile (i.e., GSH) or protein alkylation reactions will be steric hindrance, specifically the availability and reactivity of particular Cys residues with NO₂-FA.

The structural data from this study can facilitate the study and prediction of downstream NO₂-FA reactions. For example, the pKa of Cys targets is crucial factor [58], since thiolate anions are so readily nitroalkylated. The involvement of vicinal basic residues that promote Cys deprotonation is also critical and under investigation. Another key parameter that defines downstream signaling responses will be the stability of NO₂-FA in plasma and at cellular and tissue levels. At the same time, excessive protein target accumulation of NO₂-FA can become toxic, compromising their action as signaling molecules. In this regard, enzymatic reduction of the nitroalkene, mitochondrial β-oxidation, NO₂-FA incorporation into complex lipids such as triglycerides and the concerted GSH alkylation and export by multi-drug resistant proteins can limit toxic accumulation of this class of lipid electrophiles [[59], [60], [61], [62], [63]].

Depending on buffer, media or in vivo conditions, NO₂-FA are stable for several minutes to several hours [11,64]. This time perspective corresponds very well to the pharmacokinetics that would support activation of the expression of adaptive signaling-related genes and the modulation of both acute and chronic inflammatory responses [65]. The relationship between the degradation of NO₂-FA and the aqueous ·NO release phenomenon is documented by the comparison of 24 h CPSA-degradation profiles in Fig. 5, which fits the EPR-degradation kinetics shown in Fig. 7. It is viewed that the generation of NO by all nitroalkenes will be a consequence of aqueous decay via a modified Nef reaction, a phenomenon reported for in nitroalkene derivatives of bis-allylic linoleic acid [20]. Herein, we suggest that conjugated diene-containing FA nitroalkenes are also capable of undergoing the same protonation-deprotonation cycle followed by reaction with the hydroxide anion always present in aqueous milieu. Because of the adjacent conjugated double bonds in NO₂-cLA, it is viewed that the nitroso intermediate would have a weaker C–N bond and thus be more susceptible to scission to yield NO radical. Prior mass spectrometric detection of the expected oxidized fatty acid products and the present direct detection of NO formation support this pathway of nitroalkene decay. This acid/base chemistry may potentially be employed by enzymes that might also catalyze NO release from the multiple lipid nitroalkene derivatives that have been detected biologically. It is noted in the latter context however that the main physiological reaction pathway for NO₂-cLA is more likely to be a) stabilization by esterification into complex lipids, b) incorporation into hydrophobic compartments such as hydrophobic pockets of proteins, micelles, chylomicrons, lipid droplets or membranes that will in turn limit protonation-deprotonation of the nitroalkene and subsequent NO formation and c) the alkylation of proteins that in turn also inhibits the chemistries associated with the Nef reaction.

NO₂-FA are generated by the reactions of unsaturated FA with ·NO, and NO₂-derived nitrating species (Scheme 2). The life-time of these species are all very short, e.g. for ONOO-, a t_1/2 of 0.15 s was reported [66]. The generation of these species can be upregulated by inflammation, metabolism and other physiological and pathological manifestations [67]. One common denominator is that these reactive species act vicinal to the location where they are generated, with further distribution limited by a low t_1/2 and high reactivity. In contrast, the nitration of unsaturated FA to a more stable product results in broader distribution and more systemic signaling potential. Thus, the transformation of various nitrogen oxides to ·NO₂ and the formation of NO₂-FA transduces the effects of often toxic mediators of inflammation into a species that induces adaptive metabolic and gene expression responses. The relationship between the stability and biological reactivity of NO₂-FA in the context of their origin, distribution and storage is depicted in Scheme 2.

Scheme 2.

Relationships between the stability and reactivity of NO₂-FA in the context of their origin and distribution. Nitrogen dioxide radical, generated by diverse metabolic and inflammatory reactions (ref. [68]) reacts with bis-allylic or conjugated diene-containing unsaturated fatty acids to yield electrophilic NO₂-FA. These nitroalkenes and many of their metabolites are more stable than the reactive oxygen or nitrogen oxide species that serve as their precursors. Finally, the reversible esterification of NO₂-FA into complex lipids and micellar, lipoprotein, lipid droplet or membrane structures stabilizes these species.

4. Conclusions

This study affirms that NO₂-FA biological actions are pleiotropic and that understanding the physicochemical properties of NO₂-FA is crucial for clearer insight into the mechanisms of action of these electrophilic products of inflammation and metabolism. Despite a number of pharmacological and clinical studies [68], only limited studies describe the structure, stability and reactivity of NO₂-FA.

NO₂-FA are electrochemically reducible around −0.8 V, which corresponding to reduction of R–NO₂ group. Based on the potential of reduction peak NO we can evaluate electrophilic capacity or reactivity of NO₂-FA, which decreases, with NO₂-cLA >> NO₂-LA > NO₂-OA. This electrochemical data is in good agreement with DFT calculations. In this respect, the correlation of LUMO energy and the peak NO potential is an important parameter. The LUMO energy of the tested derivatives ranges from −0.7 to −0.95 eV and, the electron deficient region is located at the R–NO₂ group for all compounds studied. Furthermore, the mechanism of NO release [69] was investigated by EPR spectroscopy. The formation of ·NO or ·NO₂ radicals is due to the spontaneous degradation of NO₂-FA. A highly efficient donor of NO radical(s) is NO₂-cLA, for which a rate constant for radical(s) formation of 1.3·10⁶ was observed. This value differs from the rate constants for NO₂-LA and NO₂-OA by one order of magnitude. From theoretical studies [70] as well as from pharmacological studies of adipose tissue [71], it is evident that NO₂-FA exhibit different behavior when they are an integral part of organized lipid structures such as membranes, micelles, chylomicrons and lipid droplets or if they are freely available in solution (Scheme 2). This has to be taken into account with respect to the stability of the NO₂-FA in various model systems.

When working in an aqueous environment, the concentration of NO₂-FA also plays an important role, because of self-assembly processes [43]. If the concentration of NO₂-FA is lower than the CMC, monomeric fractions are present in the solution. Conversely, when the CMC is exceeded, the NO₂-FA monomers are transformed into self-assembled structures, most often liposome-like micelles. In the native (non-pathological) cellular environment, free NO₂-FA can be found at nM, or sub-nM, concentrations [64]. Thus, when conducting an in vitro experiment with μM NO₂-FA concentrations, it is important to take into account micellization, vesicle formation, and aggregation phenomena.

Defining the endogenous or pharmacological actions of the heterogeneous and pleiotropically-acting NO₂-FA class requires a comprehensive and multi-disciplinary approach. In particular, conformational polymorphism is an inseparable property of NO₂-FA where, in addition to the positional isomers, we have to consider cis/trans isomerism and the formation of “open” and “closed” conformer populations [54]. These will achieve a dynamically changing equilibrium in the aqueous solution depending on the experimental conditions. In general, the presence of the nitro group has a rather negligible effect on the conformation of NO₂-FA as compared to unsubstituted FA, but a profound effect on chemical reactivity and downstream effects.

Declaration of interests

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Declaration of competing interest

B.A.F. acknowledges financial interest in Complexa, Inc. and Creegh Pharmaceuticals.