Pentacyanoferrate(II) complex of pyridine‑4‑
and pyrazine‑2‑hydroxamic acid as source of HNO: investigation of anti‑tubercular and vasodilation activities
Edinilton Muniz Carvalho1,2,3 · Tercio de Freitas Paulo1,2,3 · Alix Sournia Saquet1 · Bruno Lopes Abbadi4,5 ·
Fernanda Souza Macchi4,5 · Cristiano Valim Bizarro4,5 · Rafael de Morais Campos6 · Talles Luann Abrantes Ferreira6 · Nilberto Robson Falcão do Nascimento6 · Luiz Gonzaga França Lopes3,5 · Remi Chauvin1,2 ·
Eduardo Henrique Silva Sousa3,5 · Vania Bernardes‑Génisson1,2
Received: 4 May 2020 / Accepted: 5 July 2020
© Society for Biological Inorganic Chemistry (SBIC) 2020
Abstract
A pharmacophore design approach, based on the coordination chemistry of an intimate molecular hybrid of active metabo- lites of pro-drugs, known to release active species upon enzymatic oxidative activation, is devised. This is exemplified by combining two anti-mycobacterial drugs: pyrazinamide (first line) and delamanid (third line) whose active metabolites are pyrazinoic acid (PyzCOOH) and likely nitroxyl (HNO (or NO.)), respectively. Aiming to generate those active species, a hybrid compound was envisaged by coordination of pyrazine-2-hydroxamic acid (PyzCONHOH) with a Na3[FeII(CN)5] moi- ety. The corresponding pentacyanoferrate(II) complex Na4[FeII(CN)5(PyzCONHO−)] was synthesized and characterized by several spectroscopic techniques, cyclic voltammetry, and DFT calculations. Chemical oxidation of this complex with H2O2 was shown to induce the release of the metabolite PyzCOOH, without the need of the Mycobacterium tuberculosis (Mtb) pyrazinamidase enzyme (PncA). Control experiments show that both H2O2- and N-coordinated pyrazine FeII species are required, ruling out a direct hydrolysis of the hydroxamic acid or an alternative oxidative route through chelation of a metal center by a hydroxamic group. The release of HNO was observed using EPR spectroscopy in the presence of a spin trapping agent. The devised iron metal complex of pyrazine-2-hydroxamic acid was found inactive against an actively grow- ing/non-resistant Mtb strain; however, it showed a strong dose-dependent and reversible vasodilatory activity with mostly lesser toxic effects than the reference drug sodium nitroprussiate, unveiling thus a potential indication for acute or chronic cardiovascular pathology. This is a priori a further indirect evidence of HNO release from this metal complex, standing as a possible pharmacophore model for an alternative vasodilator drug.
Keywords Blood vessel vasodilation · Hybrid pro-drug activation · Metallodrug · Pyrazinamide · Sodium nitroprusside derivative · Tuberculosis
Electronic supplementary material The online version of this article (https://doi.org/10.1007/s00775-020-01805-z) contains supplementary material, which is available to authorized users.
* Eduardo Henrique Silva Sousa [email protected]
* Vania Bernardes-Génisson [email protected]
Extended author information available on the last page of the article
Introduction
Tuberculosis, a millennial disease due to the infection by Mycobacterium tuberculosis (Mtb), remains today a world- wide health problem causing more than 1 million deaths annually and being responsible for latent infections in about one third of the global population [1].
Although the cure of tuberculosis can be achieved by a therapeutic regimen involving four frontline drugs, isonia- zid (INH), rifampicin, ethambutol and pyrazinamide (PYZ) (Scheme 1), without treatment, tuberculosis becomes lethal. Moreover, the first-line therapy is inefficient against multi
and extensively drug-resistant Mtb strains, which are increasing importance. FSecond- and third-line drugs should thus be used against F these virulent Mtb strains. However, these drugs displaynumerous toxic side effects, and even for them, resistant Mtb O
strains are emerging. O-
Recently, a novel antitubercular drug, delamanid [2] (Scheme 2), was approved as a third-line drug of the anti- tuberculosis therapy. This molecule has two particular interests: on the one hand, only a few cases of resistance have been yet reported, and, on the other, it acts upon both dormant and replicating Mtb strains [3, 4]. Delamanid, a
DELAMANID = PRO-DRUG N+
O
Ddn activation
HNO / NO = ACTIVE METABOLITE
Required parameters are missing or incorrect.
nitro-dihydro-imidazole derivative, is reported as a pro-drug and thus requires an enzymatic activation step. The active metabolite is described as being nitroxyl, i. e. HNO (proto- nated one-electron reduced form of nitric oxide, NO.), result- ing from a reductive decomposition of the nitroimidazole group by a mycobacterial nitroreductase enzyme (Ddn) [5, 6]. However, a possible action of the NO. radical cannot be entirely ruled out. On the other hand, no organic metabolite arisen from this bio-reductive activation step was until now identified as a bacterial growth inhibitor.
It is interesting to note that while HNO/NO. may exhibit anti-tuberculosis activities, it is also known to play a very important role on other biological processes as an anti- oxidant [7], anti-angiogenic [8], and also as a vasorelaxant
Scheme 2 Structure of the delamanid anti-tubercular pro-drug and its active metabolite(s)agent [9]. Indeed, cardiovascular diseases are among the major causes of deaths around the world, for which there is an urgent need for new therapeutic agents, including HNO and nitric oxide (NO.) donors. Sodium nitroprusside, a nitrosylpentacyanoferrate(II) complex, is the only clini- cally approved metal-based nitric oxide donor, which has been used in cardiovascular emergency procedures for over 50 years. In contrast, there is no HNO donor agent clinically available so far. Together these examples show that HNO First-line anti-tubercular (pro-)drugs and reactive species generated from the isoniazid and pyrazinamide pro-drugs.
donor molecules can be of great interest in the medical field as potential therapeutic agents.
Lately, some of the present authors have studied and developed a biomimetic system, as an alternative non enzyme-dependent way to activate the INH anti- tuberculosis pro-drug (KatG = activation enzyme of INH), based on the redox reactivity of a metal complex of INH (called IQG607) in the presence of oxidizing agents such as H2O2 [10–12]. The anti-mycobacterial activity (in vitro MICMtb = 1.56 μg/mL ~ 3.5 μM) [13], chemical stabil- ity, and absence of toxicity (selectivity index SI > 4000, LD50 > 2000 mg/kg) [14] of the [FeII(CN)5(INH)]3− com- plex motivated us to extend this activation approach to heter- oaryl hydroxamic acids used as ligands for the development of new HNO donor complexes. Thereby, to prepare HNO donors from [FeII(CN)5(heteroaryl hydroxamic acid)] com- plexes with wide spectra of biological activities, including antituberculosis, the pyrazine-type nitrogen aryl ring was naturally first selected and used in this study. The pyra- zine moiety is already present in the structure of a first-line antitubercular drug, pyrazinamide (PYZ, see Scheme 1). Pyrazinamide, like isoniazid and delamanid, is also a pro- drug and is activated by pyrazinamidase (PncA), an enzyme of Mtb [15, 16]. This reaction converts the pro-drug into the corresponding carboxylic acid (pyrazinoic acid = Pyz- COOH) that is assigned to be the effective toxic agent for Mtb (Scheme 1). The sensitivity of Mtb to this drug is partly due to an inefficient efflux of PyzCOOH, that is found accu- mulated in the interior of the cell [17, 18].
Considering that the major route for PYZ resistances relies on point mutations on the PncA enzyme that is respon- sible for the conversion of PYZ into pyrazinoic acid [17–19], the synthesis of the Na3[FeII(CN)5(PyzCONHOH)] complex, could be judicious because the proposed complex can be considered as a hybrid source of the active metabolites of PYZ and delamanid that, under oxidative conditions, could
generate, in an opportunistic way, two antitubercular agents, HNO and pyrazinoic acid. Both metabolites could be ideally released without the need of mycobacterial enzymes (over- coming resistance phenomena) and could act upon multiple targets simultaneously, reducing the chances of selection of resistant bacterial cells.
This paper mainly describes the synthesis, characteri- zation, and chemical–biological studies of pyrazine-2-hy- droxamic acid (2) and its pentacyanoferrate(II) complex 3 (Scheme 3). As proposed above, the ligand 2 is actually a functional hybrid of two different pharmacophores corre- sponding to the active metabolites of PYZ (PyzCOOH) and delamanid (HNO). After chemical activation of the proto- type complex 3 (Fe(II) converted to Fe(III) upon oxidation), and independently of the functional state of the activating enzyme of Mtb (mutated or not), the released pharmacoph- ores are expected to recognize different targets and act by different mechanisms, killing growing and, more particu- larly, non-growing Mtb strains (Scheme 3).
Isonicotinohydroxamic acid has also been synthesized in view of the fact that putative isonicotinic acid, formed after the activation step, can be considered as a bio-isostere of pyrazinoic acid.
Material and methods
Chemicals
Available reagents and solvents were purchased from commercial suppliers and used without further purifica- tion: methyl isonicotinate, methyl pyrazine-2-carboxylate (Alfa Aesar); hydroxylamine solution (NH2OH, 50% in water) from Sigma-Aldrich, pyrazinoic acid, isonicotinic acid and potassium ferricyanide (III) (K3[Fe(CN)6]·3H2O) from Sigma-Aldrich; isoniazid (INH) from ACROS
Instruments
1H NMR spectra were recorded on a Bruker spectrom- eter at 300 and 400 MHz using D2O and DMSO-d6 as solvents, with increasing chemical shifts at low field vs the tetramethylsilane 1H nucleus reference derived from the deuterium lock signal of the solvent. 13C NMR spec- tra were recorded on a Bruker spectrometer at 101 MHz using D2O and DMSO-d6 as solvents, with increasing chemical shifts at low field vs the tetramethylsilane 13C nucleus reference derived from the deuterium lock sig- nal of the solvent. 15N NMR spectra were recorded on a Bruker spectrometer at 50.7 MHz and external sample of CH3NO2 was used as the reference. Electrospray mass spectra (ESI) were obtained on a Perkin–Elmer SCIEX API 365 instrument high resolution and desorption chem- ical ionization (DCI) mass spectra were acquired on a Finnigan TQS 7000 spectrometer. Infrared spectra were obtained using a Perkin–Elmer Spectrum One spectrom- eter. UV–Vis spectra were recorded on a Perkin–Elmer UV/Vis/NIR Spectrometer—Lambda 950. Normal Raman (NR) spectra of the samples were acquired by means of an Xplora (Horiba) microspectrometer: the excitation radia- tion was the 785 nm line and the laser beam was focused on the sample by a × 50 long distance objective. Elemental analyses were obtained using a PERKIN ELMER 2400 series II instrument. Electrochemical measurements were performed using a BAS Epsilon E2 818 potentiostat/gal- vanostat system within a conventional three-electrode cell. Vitreous carbon, platinum, and calomel electrodes were used as working, auxiliary and reference electrodes, respectively. Voltammograms were recorded in phosphate buffer solution (PBS) 0.1 M, pH 7.4. Electron paramag- netic resonance (EPR, or electron spin resonance = ESR) spectra were recorded at room temperature (ca. 291 K) on a Bruker Elexsys-II E500 (X-Band) spectrometer using a
modulation amplitude of 0.5 G, a microwave power of 5.15 mW, with an attenuation of 16 dB with repeated number of 2 scans. The samples were transferred into a capillary (Hirschmann, Duran, Ringcaps, 50 μL), which was then placed inside a larger quartz tube enabling the sample to be accurately positioned inside the resonator. All samples were prepared as 40 mM solutions in phosphate buffer pH 7.4, at 25 °C, by mixing suitable amounts of the studied compound and spin trap (cPTIO or PBN). The reactions were initiated by the addition of the H2O2 and the spectra immediately recorded.
Synthetic procedures
General procedure for hydroxamic synthesis
To a solution of methyl heteroaryl-carboxylate derivative (3.62 mmol) in methanol (10 mL) at room temperature and under stirring, 3.9 mL of hydroxylamine 50% (65.2 mmol) was slowly added. The solution was stirred for 72 h. The solvent was evaporated under vacuum and the solid was resuspended in dichloromethane, filtered, and thus washed from traces of pyrazine-2-carboxylate. Then, the solid was resuspended in water (4 mL) to remove traces of hydroxy- lamine: the pyrazine-2-hydroxamic acid being poorly solu- ble in water, the mixture was cooled down in an ice bath for 3 h to assist the product precipitation. Then, the solid was filtered, washed with cold acetone, and dried under vacuum.
Pyrazine‑2‑hydroxamic acid (2)
Yield = 85% (0.43 g), white solid. 1H NMR (400 MHz, DMSO-d6) δ (ppm): 11.62 (s, 1H), 9.28 (s, 1H), 9.12
(d, J = 1.5 Hz, 1H), 8.84 (d, J = 2.5 Hz, 1H), 8.68 (dd, J = 2.5, 1.5 Hz, 1H). 13C NMR (101 MHz, DMSO-d6) δ (ppm): 160.65 (C=O), 147.75 (CH), 145.52 (Cq), 143.85
(CH), 143.67 (CH). 15N NMR (50.7 MHz, D2O) δ (ppm):
319.30 (N1), 323.36 (N4). IR symmetric stretching ( s), anti-symmetric stretching ( as), symmetric bending (δs) and twisting (τ) (cm−1): 3212 ( s O–H), 3061 ( s N–H), 2823 ( s C–H), 1662 ( s C=O), 1585 – 1524 C=N,
s C=C), 1417 (δs N–H), 1386 (δs C–H), 1023 ( as C=N), 916 (τ C–H). UV–Vis (H2O) max/nm ( /M−1 cm−1) = 211 (8169), 272 (7248), 312 (876). MS (DCI/CH4) m/z: 140.04 [M + H+], 124.05 [(M + H+)-16]. HRMS (DCI/CH4) m/z: for {(C5H5N3O2) + H+} calcd.: 140.0460 found: 140.0464.
Elemental analysis calcd. for C5H5N3O2: C, 43.17; H, 3.62; N, 30.21. Found: C, 42.91; H, 3.51; N, 29.81. TLC (dichlo-
romethane/methanol 90:10) Rf = 0.42. M.p. = 165 °C. Elec- trochemistry in 0.1 M phosphate buffer, pH 7.4, Epa = 0.684 and 1.119 V vs NHE.
Pyridine‑4‑hydroxamic acid (5)
and 753 (π C–H), 652 ( s C=N). UV–Vis (H2O) ß
/nm (ε/
−1 −1
max
M cm ) = 216 (18,820), 270 (6228), 486 (3638). Elemen-
Yield = 98% (0.49 g), white solid. 1H NMR (300 MHz, DMSO-d6) δ (ppm): 11.50 (s, 1H), 9.32 (s, 1H), 8.70
(d, J = 6.1 Hz, 2H), 7.67 (d, J = 6.1 Hz, 2H). 13C NMR
(101 MHz, DMSO-d6) δ (ppm): 162.48 (Cq), 150.68 (CH),
140.31 (Cq), 121.40 (CH). IR symmetric stretching ( s), anti-symmetric stretching ( as), symmetric bending (δs) and twisting (τ) (cm−1) max (cm−1): 3324 ( s O–H), 3132 ( s N–H), 1640 ( s C=O), 1540–1520 ( s C=N, s C=C), 1410 (δs N–H), 1316 (δs C–H), 1025 ( as C=N), 908 (τ C–H).
tal analysis calcd. for C10H4FeN8O2Na4·3H2O: C, 25.55; H, 2.14; N, 23.84. Found: C, 25.93; H, 2.16; N, 23.94. Electro-
chemistry in phosphate buffer (0.1 M, pH 7.4): Epa = 0.722 and 1.091 V vs NHE.
Na4[FeII(CN)5(PyCONHO−)] (Na46)
(PyCONHO− = pyridine‑4‑hydroxamate)
Yield = 79% (0.173 g), yellow solid. 1H NMR (400 MHz,
UV–Vis (H2O) ß /nm (ε/M−1 cm− 1) = 235 (3760), 260
D2O) δ (ppm): 9.05 (d, J = 4.96, 2H), 7.39 (d, J = 4.76, 2H).
max
+ 13C NMR (101 MHz, D O) δ (ppm): 180.64 (CN
), 176.39
(3710). HRMS (DCI/CH4): for {(C6H6N2O2) + H } calcd.:
2 eq
139.0508 found: 139.0515. Elemental analysis calcd. for C6H6N2O2: C, 52.17; H, 4.38; N, 20.28. Found: C, 52.12;
H, 4.21; N, 20.25. TLC (dichloromethane/methanol 90:10) Rf = 0.24. M.p = 154 °C. Electrochemistry in 0.1 M phos- phate buffer, pH 7.4, Epa = 0.767 and 1.207 V vs NHE.
General procedure for preparation
(CNax), 164.02 (Cq), 157.34 (CH), 141.63 (Cq), 120.52
(CH). IR symmetric stretching ( s), antisymmetric stretching ( as), symmetric bending (δs), antisymmetric bending (δas), wagging (π), rocking (ρ) and twisting (τ) (cm−1): 3483 ( s N–H), 3382–3047 ( s C–H), 2052 ( s C≡N), 1647 ( s C=O),
1546 ( s C=N, C=C), 1492—1414 (δs C–H), 1313 (π C–H),
1227 (τ C–H), 1165 ( s C–C). 908 (δas C–H), 846 (π C–H),
of Na [FeII(CN) (HeteroarylCONHO−)]compounds
698 (π N–H). UV–vis (H2O) ß /nm (ε/M−1 cm−1) = 233
4 5 max
(14,400), 266 (5123), 439 (4052). Elemental analysis calcd.
To a solution of Na3[FeII(CN)5(NH3)]·3H2O (0.150 g,
0.46 mmol, 1 equiv.) in water (3 mL), it was slowly added a solution of the ligand (pyrazine-2-hydroxamic acid, 0.077 g,
0.55 mmol, 1.2 equiv. or pyridine-4-hydroxamic acid,
0.076 g, 0.55 mmol, 1.2 equiv.) in water (2 mL). The mixture was stirred at room temperature, under argon atmosphere and protected from light for 3 h. Then 150 mL of a cold solu- tion of ethanol containing an excess of sodium iodide (30 equiv.) was added dropwise. The resulting precipitate was kept overnight at −20 °C before it was collected by filtration, washed with cold ethanol, and dried in a vacuum desiccator. CAUTION: since these complexes are moderately light and oxygen sensitive, they must be stored in a vacuum desiccator protected from light to extend their lifetime.
Na4[FeII(CN)5(PyzCONHO−)] (Na43)
(PyzCONHO− = pyrazine‑2‑hydroxamate)
Yield = 87% (0.180 g), orange solid. 1H NMR (400 MHz, D2O) δ (ppm): 9.46 (s, 1H), 9.18 (d, J = 3.3 Hz, 1H), 8.27 (d, J = 3.2 Hz, 1H). 13C NMR (101 MHz, D2O) δ (ppm): 178.69 (CNeq), 173.75 (CNax), 160.69 (C=O), 153.77
(CH), 150.66 (CH), 145.50 (Cq), 142.33 (CH). 15N NMR
(50.7 MHz, D2O) δ (ppm): 301.10 (N4), 308.56 (N1). IR
symmetric stretching ( s), antisymmetric stretching ( as), symmetric bending (δs), antisymmetric bending (δas) and wagging (π) (cm−1): 3418 ( s N–H), 3452–3250 ( s C–H), 2053 ( s C≡N), 1609 ( s C=O), 1577 ( s C=N, C=C),
1391—1173 (δs C–H), 1033 (τ C–H). 908 (δas C–H), 854
for C11H5FeN7O2Na4·3.5H2O: C, 27.64; H, 2.53; N, 20.51.
Found: C, 27.66; H, 2.21; N, 20.76. Electrochemistry in
0.1 M phosphate buffer (pH 7.4): Epa = 0.551 and 1.099 V vs NHE.
Computational details
All calculations were carried out using the density functional theory (DFT) with the B3LYP [21–23] functional as imple- mented in the Gaussian 09 program package, Revision D.01 (Gaussian Inc., Wallingford, CT) [24]. The 6-311++ G(d,p) basis set was used for non-metal atoms, while the LAN- L2DZ relativistic effective core potential basis set was used for Fe atoms. Vibrational frequency analyses were carried out to confirm convergence to a minimum energy geometry by the absence of imaginary frequencies. A scaling factor of 0.9679 for the calculated harmonic vibrational wavenumbers considering the 6-311++ G(d,p) basis set and the B3LYP functional. The vertical excitation energies were determined by the time-dependent density functional theory protocol (TD-DFT) using the B3LYP functional and mixed basis sets mentioned above. The polarizable continuum model (PCM)
[25] was used in the case of vertical excitation energies to take into account the solvent effect, where the dielectric con- stant of water was considered. The FTIR, UV–Vis spectra were extracted from output files, while Raman spectra were calculated with the GaussSum 3.0 program [26]. This pro- gram was used to calculate the Raman intensity considering an exciting radiation of 785 nm.
Antimycobacterial activity
In vitro activity of the compounds 2, 3, 5 and 6 was evalu- ated against the Mycobacterium tuberculosis H37Rv ATCC 27294 reference strain (American Type Culture Collection, Rockville, Md.), by performing minimum inhibitory con- centration assays (MIC), as described elsewhere [27]. This strain was grown up to mid-log phase (OD600 0.8–1.0) in Middlebrook 7H9 broth (Difco™), supplemented with 10% ADC (albumin, dextrose and catalase—BD BBL™) and 0.05% Tween 80 (Sigma-Aldrich), with agitation (100 rpm), at 37 °C. Following that, a mycobacterial suspension was prepared using sterile glass beads (4 mm), which was ali- quoted and stored at -80 °C until use. Compounds were first solubilized in ultrapure water (4 mg/mL) and diluted in 7H9 broth (200 µg/mL). Serial twofold dilutions (100 µL) of each compound were performed directly in 96-well plates in 7H9 broth, giving a concentration range of 100–0.2 µg/mL. An aliquot of mycobacterial suspension was thawed and diluted in 7H9 broth to a theoretical OD of 0.006, which was added to each well (100 µL). Compound and culture-free wells were used as positive and negative growth controls, respec- tively. Isoniazid (INH, ACROS Organics) was used as an anti-TB drug reference. Plates were sealed with Parafilm M® and incubated inside plastic bags in a bacteriological incuba- tor (37 °C, 5% CO2) for 7 days, before the addition of 60 µL of 0.01% (w/v) resazurin (Sigma-Aldrich) solution to each well. Minimum inhibitory concentration (MIC) was consid- ered as the lowest compound concentration that prevented a color conversion from blue (no growth) to pink (growth). MIC values reported for each compound were the most fre- quently occurring value observed among three independent assays. INH was used as a positive control of MIC assays.
Vasodilation assay
Rats were sacrificed by overdose of sodium thiopental and the thoracic aorta was carefully removed and cut into rings of about 5 mm in length. The aortic rings were mounted in a 5 mL organ bath containing Krebs–Henseleit medium with the following composition: 120 mM NaCl, 4.7 mM KCl, 1.8 mM CaCl2, 1.43 mM MgCl2, 25.0 mM NaHCO3,1.17 mM KH2PO4, glucose and maintained at 37 °C. The rings were attached to a force transducer (TRI202P, Panlab, Barcelona, Spain) coupled to a Powerlab data acquisition systema (ADInstruments, Sydney, Australia) and data were recorded and analyzed using the Labchart 7.0 software. After equilibration, the rings were pre-contracted with 1 μM phenylephrine (PE), and once a stable plateau was achieved, cumulative concentration–response curves were constructed using the ligand 2 and 5 and the metal complexes 3 and 6 in a concentration range from 0.1 nM to 100 μM. Sodium nitro- prusside (SNP, Na2[Fe(CN)5(NO)]) was used as positive
control. This assay was performed using seven independent measurements prepared with different animals for statisti- cal analysis. All procedures were performed according to the ethics committee of State University of Ceará number 2897836/15.
Systemic hemodynamicsMale WistarKyoto (WKY) or spontaneously hypertesive rats (SHR) (250–300 g) were anesthetized with sodium pentobarbital (50 mg/kg) and a polyethylene catheter (PE 240) was inserted into the trachea to help with spontaneous breathing. The left femoral artery was isolated and canulated with a polyethylene catheter (PE 10) and this catheter was coupled to a pressure transducer (MLT844, ADInstruments, Sydney, Australia) for continuous recording of blood pres- sure by means of a data acquisition system (Powerlab, ADIn- struments) and a Labchart 7.0 software. The right carotid artery was canulated with a microtip pressure–volume cath- eter (SPR-869, Millar Instruments; Houston, TX) and this catheter was inserted into the left ventricle for continuous measurement of cardiac parameters such as cardiac output.
Results and discussion
Synthesis
The ligand moiety, pyrazine-2-hydroxamic acid (2) (Pyz- CONHOH) [28, 29], of the targeted hybrid complex 3, was directly prepared from methyl pyrazinoate (1) in NH2OH solution with 85% yield (Scheme 4) and fully character- ized. Then, the coordination of 2 with Na3[Fe(CN)5NH3] was carried out in aqueous medium to afford Na43 in 87% yield (Scheme 4). This compound was analyzed by 1H/13C/15N NMR, IR, UV–Vis and Raman spectroscopy, along with cyclic voltammetry, elemental analysis and DFT calculations. Attempts to obtain crystals of the com- plex Na43 for X-ray diffraction analysis were unsuccessful. Additionally, procedures to exchange the sodium cations by other more crystallogenic/solubilizing cations (e.g. PPN+, NEt4+) also failed. (Protocol a: to a solution of salt (PPN+ or NEt4+) in ethyl acetate, the iron complex was added. The mixture was stirred overnight in the shadow. The solution was filtered to remove the unsolubilized complex. The filtered solution was rotoevaporated and the crude product was characterized by 1H-NMR. Protocol b: an aqueous solution containing the iron complex was added to a solution of salt cation (PPN+ or NEt4+) in ethyl acetate. This mixture was stirred overnight in the shadow. Then the organic and aqueous phases were separated. The organic phase was rotoevaporated and the solid obtained was characterized by 1H-NMR. No signal corresponding
5 Synthesis of pentacyanoferrate(II) complexes of pyridine-4-hydroxamic acid and sodium pyrazine-2-carboxylate
that the coordination of Fe(CN)5 moiety occurs at this center. The high resolution of the NMR spectra of 3, without signal broadening vs the spectra of 2, provides a strong evidence that the iron atom of 3 is indeed in a low-spin + 2 oxidation state (Figure S1).
The cyanide ligands of 3 are revealed by two 13C
NMR signals at 178.7 and 173.8 ppm, characteristic of sp 13CN nuclei in equatorial and axial positions vs the Fe–N axis. These results are consistent with those pre- viously obtained for the pyrazine hydrazide complex Na3[Fe(CN)5(PyzCONHNH2)] [32].
Vibrational spectroscopy
The cyanide ligands were also characterized by infrared (IR) and Raman spectroscopy, where two bands in the range 2050–2100 cm−1 are assigned to the FeC≡N stretch- ing frequencies ν(C≡N) [33]. The most relevant infrared and Raman vibration modes and wavenumbers can be found in the Supporting information (Table S1).
Unexpectedly, the experimental normal Raman spectrum of 3 in the solid state (Fig. 1) was indeed much more com- patible with the corresponding DFT-simulated spectrum, than with the DFT-simulated spectrum of the O-protonated form (3H+, Scheme 4). The O-deprotonated form of the ligand 2 in the tetraanionic complex 3 was also supported by elemental analysis (see experimental section). IR data reveal that the C=O band at 1662 cm−1 in the protonated free ligand 2 is shifted to ca. 1580 cm−1 in the complex 3, in agreement with the anionic hydroxamate form of 2 in 3 (Scheme 4). The formation of 3 is a consequence of the basic medium (pH >> 9) used for the preparation of this complex.
The pKa value of 3 in water solution, determined by a titri- metric method using 0.025 M HCl, is indeed as low as 8.5.
Cyclic voltammetry
Cyclic voltammograms of the ligand 2 and complex 3 in a phosphate buffer solution (0.1 M, pH 7.4) were recorded using a glassy carbon working electrode (Figure S2). For the free ligand 2, two oxidative waves are observed at
0.68 V and 1.12 V vs NHE as irreversible electrochemical processes. In the cyclic voltammogram of the complex 3H+ (Figure S2 red line), the oxidation wave is observed in the potential range of 0.50–0.90 V vs NHE. This wave can be attributed to two processes corresponding to the local oxida- tion of the FeII – > FeIII and coordinated ligand. Additionally, a second irreversible electrochemical process is observed at 1.091 V vs NHE and assigned to the second ligand oxi- dation. The oxidation of the organic moiety likely initiates via an intramolecular process involving the formation of a Fe(III) center, although an intermolecular reaction cannot be strictly ruled out. Furthermore, there are kinetic evidences of an intramolecular electron transfer during the chemical oxidation of the [Fe(CN)5(INH)]3− complex (IQG607), a process which might be similar in the chemical oxidation of 3H+ [10a]. In the cathodic scanning, only one wave at
0.54 V vs NHE is observed and attributed to the FeIII – > FeII
reduction process.
UV–Vis absorption spectroscopy
The electronic spectra of the protonated free ligand 2 and complex 3H+ were recorded in slightly acidic aqueous
Experimental normal Raman (a) and IR (b) spectra of the complex 3 under tetraanionic form, Na4[Fe(CN)5L], L = PyzCONHO– (plain line), in the solid state, and DFT-simulated spectra for [Fe(CN)5L]4− (dotted line)solution (pH = 6.8). The free ligand 2 exhibits two main bands at 211 and 272 nm, and a shoulder at 312 nm, while the complex 3H+ exhibits three main electronic transitions with absorption maxima at 216, 270, and 486 nm (Figure S3). The two first ones are consistent with bands relative to the ligand moiety, while the last band is characteristic of a metal-to-ligand charge transfer transition (MLCT). This type of transition occurs from an electron-rich Fe center to π-accepting ligands, which is in agreement with the + 2 oxi- dation state of the iron center. TD-DFT calculations further validate the character of these transitions (see Supporting information, Table S2 and Figure S4).
Chemical oxidation of the Fe(II) complex under physiological pH conditions
and characterization of drug metabolites active against Mtb
As argued earlier, the complex Na43 has been devised as a precursor of the anti-Mtb metabolites of pyrazinamide and delamanid, i.e. pyrazinoic acid (PyzCOOH) and nitroxyl (HNO/NO.), respectively. The search for evidences of the formation of the later organic and inorganic products upon chemical oxidation with H2O2 is addressed below:
Formation of PyzCOOH upon oxidation of 3
Chemical oxidation of 3 or its protonated form 3H+ by action of H2O2 in a phosphate buffer solution (40 mM, pH = 7.4) can be considered as a mimic of the biologi- cal activation of PYZ by the pyrazinamidase (PncA)
enzyme, which is often found mutated in PYZ-resistant Mtb strains. In these conditions, oxidation of 3 under its major hydroxamic acid form 3H+ and the release of PyzCOO−, without assistance of PncA, was thus investi- gated by UV–vis and 1H NMR spectroscopy.
Before addition of H2O2, the trianionic complex 3H+ (buffer solution of 3 at pH = 7.4) showed an absorption band with a maximum at 489 nm, characteristic of the MLCT transition as previously discussed. This band dis- appeared almost completely within 2 h (Fig. 2, red line) after addition of H2O2, indicating modification of the metal moiety. This result supports the hypothesis that this complex is activated by H2O2 via FeII – > FeIII oxidation (Fig. 2).
1H NMR analysis of 3 in deuterated buffer solution
and in the presence of 2.5 equivalents of H2O2, after 52 h, revealed that the signals relative to the aromatic protons of the trianionic complex 3H+ had completely disap- peared, while new signals corresponding to free pyrazine carboxylate (8) were observed. This was evidenced by a perfect matching of the new aromatic 1H NMR signals (Fig. 3, spectrum C) with those of an authentic sample of sodium pyrazinoate (8) (Fig. 3, spectrum E). Remarkably, after oxidative activation, the organic metabolite 8 was found to be spontaneously released from the metal center, while only very little ligand remains coordinated to the FeII(CN)53– unit (Fig. 3, spectrum C and D)
It is worth noting that the free hydroxamic acid 2 in the presence of H2O2 is unable to produce pyrazinoic acid even after 52 h. In addition, in the presence of Na3[FeIII(CN)6] (2.5 equiv.), the hydroxamic group of 2 is only very slowly UV–Vis monitoring of the reaction of 3 under the protonated main form 3H+(171 µM) with H2O2 (427.5 µM)in phosphate buffer solution,40 mM, pH 7.4, 22 °C. Reaction at t = 0 min, before addition of H2O2, (blue line), and after 2 h (red line). The black curve cor- responds to a solution of sodium 2-pyrazinoate (110 µM). Inset shows the kinetic curve based on changes at 489 nm
1H-NMR spectra at 400 MHz: of the complex 3 under its protonated form 3H+ (20 mM) at 0 h (a) and at 52 h
(b) without addition of H2O2, at 52 h after reaction of 3H+ with H2O2 (50 mM) (c), and of authentic samples of the complex 9 (20 mM) (d) and pyrazinoate salt (8) (20 mM) (e). Solutions in 40 mM phos- phate buffer, pH 7.4, at 25 °C partially oxidized, and converted to the corresponding carboxylic acid PyzCOOH (even after 64 h), after hydroly- sis of the putative dimer N,O-dipyrazinoylhydroxylamine (PyzC(O)NHOC(O)Pyz) formed as intermediate [34]. The latter intermediate has never been observed for the ligand of 3 in oxidative conditions. All together, these results show that both H2O2 and the N-coordinated FeII unit are needed for oxidative activation of the hydroxamic moiety, ruling out a random mechanism passing through a direct hydrolysis of the hydroxamic acid function (PyzCONHOH + H2O → Pyz- COOH + NH2OH) or an alternative activation route where the hydroxamic acid group would coordinate metal ions occasionally present in buffer solutions.
Formation of HNO/NO upon oxidation of 3
The inorganic metabolite of delamanid, i.e. the weak acid nitroxyl (HNO) or radical nitric oxide (NO), was tar- geted. Discrimination between these possible products was envisaged by EPR spectroscopy in the presence of the 2-(4-carboxyphenyl)-4,4,5,5-tetramethylimidazoline-1-oxyl- 3-oxide potassium salt (cPTIO). This compound is a water- soluble nitronyl-nitroxide known to be readily reduced by NO, of which it is a specific scavenging agent, to give the corresponding cPTI nitroxide and the NO2 radical (Scheme S1). This species (cPTI) presents characteristic EPR
signatures different from the quintuplet signal of cPTIO (a septet for cPTI) [34]. cPTIO also reacts with HNO, by which it is reduced to the EPR-silent nitronyl-hydroxylamine, while releasing the NO radical (Scheme S1) [35, 36].
The first control of cPTIO with the free hydroxamic acid 2 or complex 3 showed that the EPR signal of cPTIO was not altered, even after 15 min (Fig. 4b, d). The same sta- bility of cPTIO was observed in the presence of H2O2 and with the mixture of FeCl2 + H2O2, a Fenton-reaction system generating hydroxyl radicals (Fig. 4c, e). Moreover, EPR monitoring did not show any evidence of reaction of cPTIO with Na3[FeIII(CN)6] (Fig. 4f). However, a net decrease of the cPTIO EPR signal was observed upon treatment of the FeII complex 3 with H2O2, indicating neutralization of the nitroxide radical, which would be a priori due to the release of HNO (Fig. 4a).
Our previous studies on the activation of Na3[Fe(CN)5(INH)] (IQG607) complex in the presence of H2O2 gave evidences (by EPR spectroscopy) for the formation of the isonicotinoyl radical as an intermediate towards isonicotinoic acid [31, 37]. On this basis, it was decided to check whether the hydroxamate complex 3 or its acidic form 3H+ could also generate pyrazinoic acid via a transient radical intermediate what could be an indirect evidence for the formation of NO.. By employing the same methodology used before for the investigation of the INH a EPR signal of the cPTIO radical in the presence of the complex 3 under its protonated form 3H+ and H2O2 (t0 = black line and t10min = red line) in phosphate buffer solution, 40 mM, pH 7.4. b Control of the stability of the cPTIO EPR signal in the presence of Na3Fe(CN)5-PyzCONHOH] (t0 = black line and t15min = red line). c Control of the stability of the cPTIO EPR signal in the presence of H2O2 (t0 = black line and t15min = red line). d Control of the stabilitycomplex, EPR spectra were recorded using N-tert-butyl-α- phenylnitrone (PBN) as the spin trapping agent (Scheme S1B). In contrast to the parent iron complexes of isoniazid and of pyrazinoic acid hydrazide [30], Na3[Fe(CN)5(INH)] and Na3[Fe(CN)5(PyzCONHNH2)], respectively, no radical could be trapped upon treatment of the pyrazine and even of the pyridine hydroxamic acid complexes (3 and 6, respec- tively), with H2O2 (Schemes 4 and 5).
To explain the latter results, two hypotheses can be proposed. In the case of the hydroxamic acid com- plexes, possibly formed HNO along with the aroyl radi- cal (by analogy with the INH complex, Scheme 3) might quench the latter thus shortening its half-life time and preventing its detection by EPR. However, the processof the cPTIO EPR signal in the presence of the ligand PyzCONHOH and H2O2 (t0 = black line and t20min = red line). e Control of the sta- bility of the cPTIO EPR signal in Fenton reaction conditions [FeCl2 (5 mM) and H2O2 (200 mM)] (t0 = black line and t15min = red line). f Control of the stability of the cPTIO EPR signal in the presence of [Fe(CN)6]3− and H2O2 (t0 = black line and t20min = red line)
HNO + ArC•=O → NO• + PyzCHO, would yield an alde- hyde, which was not evidenced by 1H NMR spectroscopy. Moreover, the benzile-like product of the radical coupling process 2 ArCO• → ArC(O)–C(O)Ar also failed to be iso- lated and even detected.
Another possibility is thus that, in the hydroxamic series, the carboxylic acid does not arise from an aroyl radical, but from another intermediate instead. This intermediate could be the undissociated aroyl-nitroso compound, gener- ated upon an oxidation step, which could undergo a rapid heterolytic cleavage by the surrounding water molecules (Scheme 6), releasing HNO without any radical species, as previously suggested to occur from other hydroxamic acids [38, 39].
Inhibitory activity against Mtb
The antimycobacterial activity of the ligand 2 and complex 3 was investigated against the Mtb H37Rv reference strain, using INH as positive reference compound. The ligand 5 and the complex 6 were also introduced in these biological assays for comparison purposes. In this anti-mycobacterial assay, resazurin dye was employed to measure the lowest concentration of the tested compounds that could prevent any color change, which indicates a disruption of the growth. Those values were reported as MIC (minimum inhibitory concentration), which, for compounds 2 and 3, 5 and 6 were, respectively, 100 and > 100 μg/mL, and for INH was
0.39 μg/mL. These results indicate that even at considerably high concentrations, there is no measurable inhibition of Mtb growth under these experimental conditions. However, putative activity against of Mtb strains within activated mac- rophage, or even in the non-replicating Mtb state, deserves further investigations.
Vasodilatory and antihypertensive activity
Aiming to test the hypothesis that metal complex 3 or 3H+ can release HNO/NO, vasodilation assays were performed. Again, the ligand 2 and 5 and the complex 6 were also tested for comparison. The NO species is indeed known to induce vasodilation of blood vessels, and NO carrier molecules, such as sodium nitroprusside (SNP, Na2[FeII(CN)5(NO)), have an important therapeutic effect for hypertensive crises [40]. At the same time, HNO, the reduced and protonated form of NO., has also been reported to exhibit vasodila- tion activity, and shares many similar biological features of NO.[38, 39]. The structural analogy of the complexes 3 and 6 with SNP was thus a natural argument for such inves- tigations, along with our further experimental evidences for HNO production (Fig. 4). Indeed, an assay carried out side-by-side with SNP and complexes 3 and 6 showed an interesting profile of vasodilation. While the free ligands 2 and 5 exhibited an expressively higher EC50 value for vaso- dilation activity at 19 μM, the metal complex 3 had an EC50 value of 250 nM only, a remarkable enhancement of 76-fold.
complex 6 exhibits lower EC50 (64 nM) than 3 and SNP is only a fivefold more potent than 6. The beneficial role of the iron complex moiety for vasodilation properties can be promptly noted by comparing the EC50 values of 2 and 3 and also 5 and 6 (Fig. 5). The concentration–response curve for the vasodilation experiments showed that the complex 3 as compared to SNP has the same efficacy and similar potency, which is a priori promising for the design of an alternative drug. Although SNP has several indications in cardiology such as heart failure, angina, after cardiac surgeries, hyper- tensive crisis, hypertensive encephalopathy, cardiogenic pulmonary oedema, it also has several contraindications and limitations such as liver toxicity, reflex tachycardia, and electrocardiographic alterations.
Although the water-soluble complexes 3 and 6, able to release HNO/NO, are less potent than SNP, this can be an advantage because SNP, releasing NO rapidly, can cause major drops in blood pressure, thus requiring a careful con- trol through titration. Relaxation effects on aortic rings pre-contracted with 0.1 µM phenylephrine: [Fe(CN)5(pyrazine-2-hydroxamic acid)]3− (3) (blue), Fe(CN)5(pyridine-4-hydroxamic acid)]3− (6) (red), pyrazine-2-hy- droxamic acid (2) (green), isonicotinoic hydroxamic acid (5) (purple) and SNP (black). *p < 0.05 vs. pyrazine-2-hydroxamic acid and isoni-
Additionally, EC50 values for the complex 3 is 19-fold higher
cotinoic hydroxamic acid. The IC50 and respective 95% confidence
interval were calculated only for 3 (250 nM, 91–690 nM) and 6 (64
Proposed outcome of the chemical oxidation of complex 3 under its trianionic form 3H+ with H2O2
The complexes 3 and 6 can be all the more interesting since their efficacy (maximum relaxation effect) is quite similar to that of SNP. We should also point out that, in contrast to SNP, whose release of NO also promotes cyanide dissociation, the process is greatly reduced for new pentacy- anoferrate complexes. Actually, a quick comparison of the lethal dose of SNP and the pentacyanoferrate(II)-INH com- plex (IQG607) illustrates this topic (mouse orally adminis- tered with SNP and IQG607 LD50 = 61 mg/kg and 2970 mg/ kg, respectively), where SNP is over 48-fold more lethal [14, 41]. It is also noteworthy that those values correspond to oral dosages, but considering that pentacyanoferrate(II)complexes are particularly instable in acidic conditions— occurring in the rats’ stomach— leading to decomposition and cyanide release—the real activity level might be even higher with other administration routes.
All these results indicate that the complex 3 or 6 might be a suitable model for designing alternative anti-hypertensive drugs. Modulation of the electronic effects on the metal center and changes on the aromatic hydroxamic ligand moi- ety might lead to more efficient tuning of their properties to achieve adjustable HNO/NO release. For instance, the compound 6, which exhibited the lower EC50 (64 nM), was demonstrated to decrease blood pressure more efficiently in hypertensive rats (SHR) than its normotensive matched con- trols (Fig. 6). The doses of 1, 5, and 10 mg/kg promoted a
drop in blood pressure equivalent to 45.6 ± 1.9%, 62.1 ± 2.5%nd 65.1 ± 3.1% in SHR rats and 30.8 ± 2.3, 50.9 ± 4.2 an53.3 ± 3.3 in normotensive rats, respectively. This compound did also increase cardiac output significantly, probably sec- ondary to a decreased afterload and increased coronary flow.
Conclusion
The proposed strategy for metal-mediated oxidative acti- vation of hybrid prodrugs, has been investigated for the anti-Mtb metabolites of pyrazinamide and delamanid in the hydroxamic hybrid series. The results unveil biological prospects of Fe(II) coordination complexes in pharmaco- phore design, while spanning a bridge between far-related medicinal challenges, from anti-bacterial to vasodilation effects. The complex 3 in the presence of H2O2 gave direct and indirect evidences for the release of pyrazinoic acid and HNO, respective active metabolites of pyrazinamide and delamanid. Although this complex did not exhibit anti-Mtb activity on actively growing non-resistant strains, further investigations may be warranted employ- ing either/both single-condition whole-cell models (e.g., fatty acid carbon sources, hypoxia, low pH, biofilm, and carbon starvation) or/and multistress condition whole-cell models that try to recapitulate nonreplicating persistent bacilli, upon which pyrazinamide is active. Furthermore,
Fig. 6 Antihypertensive (a) and increased cardiac output (b) induced by compound 6 in normotensive (WKY) rats or spontaneously hyper- tensive rats (SHR). *p < 0.05 vs. control (administration of the vehi- cle)promising advances have been established for antihyper- tensive therapy, where the hydroxamic complexes 3 and 6 have been shown to be globally as active as the SNP drug, with a more regular action and significantly lower toxicity. Additionally, comparison of vasodilation properties of 3 and 6 with those of 2 and 5 respectively, clearly show that the pentacyanoferrate moiety plays a beneficial role in the efficiency of the release of HNO from hydroxamic acids. Investigation of the expected longer duration of vasodila- tory action of 3 and 6 will be performed in due course by systematic pharmacodynamic (PD) studies.
Acknowledgements The authors would also like to acknowledge financial support given by CNPq/FAPERGS/CAPES/BNDES to the National Institute of Science and Technology on Tuberculosis (INCT- TB), Brazil [grant numbers: 421703-2017-2/17-1265-8/14.2.0914.1; LGFLopes: CNPq 303355/2018-2; EHSS: CNPq 308383/2018-4, Uni-versal 403866/2016-2) and COFECUB Project n° Ph-C 883/17. This study was financed in part by the Coordenação de Aperfeiçoamento de Pessoal de Nível Superior—Brasil (CAPES)—Finance Code 001.
Additionally, we are also in debt with Iury A. Paz, Renata Oliveira Santiago and Ariana Gomes da Silva for their assistance on the pre- liminary biological studies.
Compliance with ethical standards
Conflict of interest The authors declare have no conflict of interest.
Ethical approval All procedures were performed according to the ethics committee of State University of Ceará number 2897836/15.
References
1. Health World Organization annual report 1. https://www.who.int/tb/ 1. publications/global_report/en/ Accessed 30 April 2020.
2. Ryan NJ, Lo JH (2014) Delamanid: first global approval. Drugs 74:1041–1045
3. Rustomjee R, Zumla A (2015) Delamanid expanded access novel treatment of drug resistant tuberculosis. Infect Drug Resist 8:359–366
4. Matsumoto M, Hashizume H, Tomishige T, Kawasaki M, Tsub- ouchi H, Sasaki H, Shimokawa Y, Komatsu M (2006) OPC-67683, a nitro-dihydro-imidazooxazole derivative with promising action against tuberculosis in vitro and in mice. PloS Med 3:e466
5. Singh R, Manjunatha U, Boshoff HIM, Ha YH et al (2008) PA-824 Kills Nonreplicating Mycobacterium tuberculosis by Intracellular NO Release. Science 322:1392–1395
6. Laborde J, Deraeve C, Bernardes-Génisson V (2017) Update of antitubercular prodrugs from a molecular perspective: mecha- nism of action, bioactivation pathways and associated resistance. ChemMedChem 12:1657–1676
7. Lopez BE, Shinnyashiki M, Han TH, Fukuto JM (2007) Anti- oxidant actions of nitroxyl (HNO). Free Radical Biol Med 42:482–491
8. Norris AJ, Sartippour MR, Lu M, Park T, Rao JY, Jackson MI, Fukuto JM, Brooks MN (2008) Nitroxyl inhibits breast tumor growth and angiogenesis. Int J Cancer 122:1905–1910
9. Nelli S, McIntosh L, Martin W (2001) Role of copper ions and cytochrome P450 in the vasodilator actions of the nytroxyl anion generator, Angeli’s salt, on rat aorta. Eur J Pharmacol 412:281–289
10. Sousa EHS, Basso LA, Santos DS, Diógenes ICN, Longhinotti E, Lopes LGF, Moreira IS (2012) Isoniazid metal complex reactivity and insights for a novel anti-tuberculosis drug design. J Biol Inorg Chem 17:275–283
11. Basso LA, Schneider CZ, Santos AJAB, Santos AA Jr, Campos MM, Souto AA, Santos DS (2010) An inorganic complex that inhibits mycobacterium tuberculosis enoyl reductase as a proto- type of a new class of chemotherapeutic agents to treat tubercu- losis. J Braz Chem Soc 21:1384–1389
12. Sousa EHS, Pontes DL, Diogènes ICN, Lopes LGF, Oliveira JS, Basso LA, Santos DS, Moreira IS (2005) Electron transfer kinet- ics and mechanistic study of the thionicotinamide coordinated to the pentacyanoferrate(III)/(II) complexes: a model system for the in vitro activation of thioamides anti-tuberculosis drugs. J Inorg Biochem 99:368–375
13. Abbadi BL, Villela AD, Rodrigues-Junior VS, Subtil FT, Dalberto PF, Pinheiro APS, Santos DS, Machado P, Basso LA, Bizarro CV (2018) Revisiting activation of and mechanism of resistance to compound IQG-607 in Mycobacterium tuberculosis. Antimicrob Agents Chemother 62:e2222–e2317
14. Abbadi BL, Rodrigues-Junior VS, Dadda AS, Pissinate K, Villela AD, Campos MM, Lopes LGF, Bizarro CV, Machado P, Sousa EHS, Basso LA (2018) Is IQG-607 a potential metallodrug or metallopro-drug with a defined molecular target in Mycobacte- rium tuberculosis? Front Microbiol 9:880
15. Zhang Y, Mitchison D (2003) The curious characteristics of pyrazinamide: a review. Int J Tuberc Lung Dis 7:6–21
16. Vandal OH, Nathan CF, Ehrt S (2009) Acid resistance in Myco- bacterium tuberculosis. J Bacteriol 191:4714–4721
17. Zhang Y, Mitchison D, Shi W, Zhang W (2014) Mechanisms of pyrazinamide action and resistance. Microbiol. Spectr. 2: MGM2-0023-2013
18. Njire M, Tan Y, Mugweru J, Wang C, Guo J, Yew W, Tan S, Zhang T (2016) Pyrazinamide resistance in Mycobacterium tuber- culosis: review and update. Adv Med Sci 61:63–71
19. Stehr M, Elamin AA, Singh M (2015) Pyrazinamide: the impor- tance of uncovering the mechanisms of action in mycobacteria. Expert Rev Anti Infect Ther 13:593–603
20. Pires BM, Jannuzzi SAV, Formiga ALB, Bonacin JA (2014) Prus- sian blue films produced by pentacyanidoferrate(II) and their application as active electrochemical layers. Eur J Inorg Chem 2014:5812–5819
21. Becke AD (1993) Density-functional thermochemistry. III. The role of exact exchange. J Chem Phys 98:5648–5652
22. Lee C, Yang W, Parr RG (1988) Development of the Colle-Salvetti correlation-energy formula into a functional of the electron den- sity. Phys Rev B 37:785–789
23. Stephens PJ, Devlin FJ, Chabalowski CF, Frisch MJ (1994) Ab Ini- tion calculation of vibrational absorption and circular dichro- ism spectra using density functional force fields. J Phys Chem 98:11623–11627
24. Frisch MJ, Trucks GW, Schlegel HB, Scuseria GE, Robb MA, Cheeseman JR, Scalmani G, Barone V, Mennucci B, Petersson GA, Nakatsuji H, Caricato M, Li X, Hratchian HP, Izmaylov AF, Bloino J, Zheng G, Sonnenberg JL, Hada M, Ehara M, Toyota K, Fukuda R, Hasegawa J, Ishida M, Nakajima T, Honda Y, Kitao O, Nakai H, Vreven T, Montgomery JA Jr, Peralta JE, Ogliaro F, Bearpark M, Heyd JJ, Brothers E, Kudin KN, Staroverov VN, Kobayashi R, Normand J, Raghavachari K, Rendell A, Burant JC, Iyengar SS, Tomasi J, Cossi M, Rega N, Millam JM, Klene M, Knox JE, Cross JB, Bakken V, Adamo C, Jaramillo J, Gomperts R, Stratmann RE, Yazyev O, Austin AJ, Cammi R, Pomelli C, Ochterski JW, Martin RL, Morokuma K, Zakrzewski VG, Voth GA, Salvador P, Dannenberg JJ, Dapprich S, Daniels AD, Farkas O, Foresman JB, Ortiz JV, Cioslowski J, Fox DJ (2013) Gaussian 09, Revision D.01. Gaussian Inc, Wallingford
25. Tomasi J, Mennucci B, Cammi R (2005) Quantum mechanical continuum solvation models. Chem Rev 105:2999–3093
26. O’Boyle NM, Tenderholt AL, Langner KM (2008) Cclib: a library for package-independent computational chemistry algorithms. J Comput Chem 29:839–845
27. Giacobbo BC, Pissinate K, Rodrigues-Junior V, Villela AD, Grams ES, Abbadi BL, Subtil FT, Sperotto N, Trindade RV, Back DF, Campos MM, Basso LA, Machado P, Santos DS (2017) New insights into the SAR and drug combination synergy of 2-(quino- lin-4-yloxy)acetamides against Mycobacterium tuberculosis. Eur J Med Chem 126:491–501
28. Kushner S, Dalalian H, Sanjurjo JL, Bach FL Jr, Safir SR, Smith VK Jr, Williams JH (1952) Experimental chemotherapy of tuber- culosis. II. The Synthesis of pyrazinamides and related com- pounds. J Am Chem Soc 74:3617–3621
29. Zhang YY, Gao W-X, Lin Y-J, Mi L-W, Jin G-X (2017) Syntheses, structures, and solution studies of multicomponent macrocycles and cages based on versatile ligands. Chem Eur J 23:11133–11140
30. Hackley BE Jr, Plafinger R, Stolberg M, Wagner-Jauregg T (1955) Accelaration of the hydrolysis of organic fluoro-phosphates and
fluorophosphonates with hydroxamic acids. J Am Chem Soc 77:3651–3653
31. Laborde J, Deraeve C, Vieira FGM, Sournia-Saquet A, Rechignat L, Villela AD, Abbadi BL, Macchi FS, Pissinate K, Lopes LGF, Sousa EHS, Pratviel G, Bernardes-Génisson V (2018) Synthesis and mechanistic investigation of iron(II) complexes of isoniazid and derivatives as a redox-mediated activation strategy for anti- tuberculosis therapy. J Inorg Biochem 179:71–81
32. Grandjean F, Samain L, Long GJ (2016) Characterization and utilization of Prussian blue and its pigments. Dalton Trans 45:18018–18044
33. Carvalho EM, Rechignat L, Sousa EHS, Lopes LGF, Chauvin R, Bernardes-Génisson V (2020) Mechanistic insights into the in vitro metal-promoted oxidation of (di)azine hydroxamic acids: evidence of HNO release and N, O-di(di)azinoyl hydroxylamine intermediate. New J Chem 44:11965–11973
34. Samuni U, Samuni Y, Goldstein S (2010) On the distinction between nitroxyl and nitric oxide using nitronyl nitroxides. J Am Chem Soc 132:8428–8432
35. Bobko AA, Ivanov A, Khramtsov VV (2013) Discriminative EPR detection of NO and HNO by Delamanid encapsulated nitronyl nitroxides. Free Radic Res 47:74–81
36. Bobko AA, Khramtsov VV (2015) Redox properties of the nitro- nyl nitroxide antioxidants studied via their reactions with nitroxyl and ferrocyanide. Free Radic Res 49:919–926
37. Sousa EHS, Vieira FGM, Butler JS, Basso LA, Santiago DS, Santiago DS, Diógenes ICN, Lopes LGF, Sadler PJ (2014) [Fe(CN)5(isoniazid)]3 −: an iron isoniazid complex with redox behavior implicated in tuberculosis therapy. J Inorg Biochem 140:236–244
38. Maimon E, Lemer A, Samuni A, Goldstein S (2018) Direct obser- vation of acyl nitroso compounds in aqueous solution and the kinetics of their reactions with amines, thiols, and hydroxamic acids. J Phys Chem A 122:7006–7013
39. Maimon E, Lemer A, Samuni A, Goldstein S (2018) Nitrogen dioxide reaction with nitroxide radical derived from hydroxamic acids: the intermediacy of acyl nitroso and nitroxyl (HNO). J Phys Chem A 122:3747–3753
40. Silva Filho PM, Paz IA, Nascimento NRF, Santos CF, Araújo VR, Aquino CP, Ribeiro TS, Vasconcelos IF, Lopes LGF, Sousa EHS, Longhinotti E (2019) Incorporation of nitroprusside on silica nanoparticles—a strategy for safer use of this no donor in therapy. Mol Pharm. 16:2912–2921
41. RTECS—Register of toxic effects of chemical substances. https:// www.cdc.gov/niosh/docs/97-119/default.html. Accessed 30 Apr 2020
Publisher’s Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Affiliations
Edinilton Muniz Carvalho1,2,3 · Tercio de Freitas Paulo1,2,3 · Alix Sournia Saquet1 · Bruno Lopes Abbadi4,5 ·
Fernanda Souza Macchi4,5 · Cristiano Valim Bizarro4,5 · Rafael de Morais Campos6 · Talles Luann Abrantes Ferreira6 · Nilberto Robson Falcão do Nascimento6 · Luiz Gonzaga França Lopes3,5 · Remi Chauvin1,2 ·
Eduardo Henrique Silva Sousa3,5 · Vania Bernardes‑Génisson1,2
1 CNRS, Laboratoire de Chimie de Coordination, LCC, UPR 8241, 205 Route de Narbonne, BP 44099, 31077 Cedex 4 Toulouse, France
2 Université de Toulouse, Université Paul Sabatier, UPS, 118 Route de Narbonne, 31062 Cedex 9, Toulouse, France
3 Grupo de Bioinorgânica, Departamento de Química Orgânica E Inorgânica, Universidade Federal Do Ceará, Campus Pici, Fortaleza, CE 60455-760, Brazil
4 Centro de Pesquisas Em Biologia Molecular E Funcional (CPBMF), Pontifícia Universidade Católica Do Rio Grande Do Sul (PUCRS), Porto Alegre, Brazil
5 Instituto Nacional de Ciência E Tecnologia Em Tuberculose (INCT-TB), Porto Alegre, Brazil
6 Laboratório de Farmacologia Cardiovascular E Renal, Universidade Estadual Do Ceará, Campus do Itaperi, Fortaleza, CEP 60714-903, Brazil