<<

biomolecules

Article Oxidation of 5-methylaminomethyl uridine (mnm5U) by Oxone Leads to Aldonitrone Derivatives

Qishun Zhou 1,2, Bao Tram Vu Ngoc 1,2, Grazyna Leszczynska 3 , Jean-Luc Stigliani 1,2 and Geneviève Pratviel 1,2,*

1 CNRS, Laboratoire de Chimie de Coordination, 205 route de Narbonne, 31077 Toulouse CEDEX4, France; [email protected] (Q.Z.); [email protected] (B.T.V.N.); [email protected] (J.-L.S.) 2 Université de Toulouse, Université Paul Sabatier, UPS, 31330 Toulouse, France 3 Institute of Organic Chemistry, Faculty of Chemistry, Lodz University of Technology, Zeromskiego 116, 90-924 Lodz, Poland; [email protected] * Correspondence: [email protected]; Tel: +33-561333146

 Received: 12 October 2018; Accepted: 8 November 2018; Published: 14 November 2018 

Abstract: Oxidative RNA damage is linked to cell dysfunction and diseases. The present work focuses on the in vitro oxidation of 5-methylaminomethyl uridine (mnm5U), which belongs to the numerous post-transcriptional modifications that are found in tRNA. The reaction of oxone with mnm5U in water at pH 7.5 leads to two aldonitrone derivatives. They form by two oxidation steps and one dehydration step. Therefore, the potential oxidation products of mnm5U in vivo may not be only aldonitrones, but also hydroxylamine and derivatives (which may be chemically more reactive). Irradiation of aldonitrone leads to unstable derivatives that are susceptible to isomerization to or to hydrolysis to derivative.

Keywords: RNA oxidation; aldonitone; oxaziridine; KHSO5; water solvent

1. Introduction RNA shows about a hundred different post-transcriptionally modified nucleosides and most of them are in tRNA. They are enzymatically synthesized and are located at specific sites. They play diverse and indispensable roles in the decoding of mRNA and in the gene regulation in all living organisms [1,2]. Modifications in the anticodon stem and loop (ASL) domain of tRNA are the most diverse and the most complex of all modifications in RNA. Because the ASL domain is critical for accurate and efficient translation, the ASL natural modifications directly contribute to the correct decoding of mRNA codons and they constitute an essential key in the regulatory processes of translation [3,4]. Uridine at the wobble anticodon position 34 of tRNAs is often modified with a methylene on C5 of uracil with a variety of modifications such as 5-methylaminomethyl, 5-oxyacetic acid, 5-taurinomethyl. In addition, position 34 uridine may also carry a sulfur or selenium atom on C2 that may be associated to C5 modification and result in hypermodified ribonucleosides, such as 5-methylaminomethyl-2-thiouridine (mnm5s2U) and 5-methylaminomethyl- 2-selenouridine(mnm5se2U) [4–6]. Modifications to position 34 uridines are crucial for the precise decoding of the genetic information and for the tuning of protein synthesis [4,7,8]. The present work deals with the sensitivity to oxidation of the modified ribonucleoside 5-methylaminomethyluridine (mnm5U). The 5-methylaminomethyl group on the C5 position of uracil has been found in bacteria and archaea [5,6]. For instance, the modified uridine mnm5U is present at the wobble anticodon position 34 in E. coli tRNAArg4 isoaceptor [9] and the hypermodified

Biomolecules 2018, 8, 145; doi:10.3390/biom8040145 www.mdpi.com/journal/biomolecules Biomolecules 2018, 8, 145 2 of 18 nucleoside, 5-methylaminomethyl-2-thiouridine also at position-34 in E. coli tRNALys [4,10]. Another related hypermodified nucleoside, 5-methylaminomethyl-2-selenouridine, was found in bacteria [11]. The fact that RNA is sensitive to oxidation is now well documented and oxidative RNA damage has been connected to various forms of cell dysfunction or diseases [12–14]. Apart from general RNA oxidation implying normal nucleobases in particular guanines [15,16], recent focus on natural modified RNA bases have provided interesting insights into their high sensitivity to oxidation [17–21]. , which belongs to the reactive species of oxidative stress, readily reacts with sulfur-containing nucleobases. Oxidation of 2-thiouridine leads to uridine and/or 4-pyrimidinone derivative, depending on the experimental conditions [17,18,21]. Oxidation of mnm5s2U leads mainly to mnm5U[18]. Therefore, the study of the sensitivity of modified RNA nucleosides to oxidative stress appears relevant for the identification of their endogenous oxidative damage, which might prove biologically significant. 5 We report oxidative damage of mnm U in the presence of oxone (KHSO5), which is an asymmetric peroxide soluble in water at physiological pH. It is convenient for in vitro studies, since its reactivity mimics that of while being more reactive. Besides, its reactivity is similar to that of the biologically relevant asymmetric peroxide, peroxynitrite (ONOO−). The identification of chemical modifications due to oxidation on the 5-methylaminomethyl group at the C5 position of uracil may help better understanding of RNA sensitivity to oxidative stress and its consequences in cells. We characterized the oxidation products of mnm5U, namely, the 5-methyl-aldonitrone and the 5-aldonitrone-methyl derivatives, 1 and 2, by UV-visible, mass, and NMR spectroscopy. We also describe the secondary products that may form from these aldonitrones in physiological conditions.

2. Materials and Methods

2.1. Chemicals Ribonucleoside was a gift from Prof. Barbara Nawrot, Polish Academy of Sciences, CMMS, Lodz, Poland and Dr. Elzbieta Sochacka, Technical University, Lodz, Poland. It was prepared according to published procedure [22]. Oxone (triple salt K2SO4, KHSO4, 2 KHSO5) was purchased from Aldrich, Saint-Louis, MO, USA. The mnm5U ribonucleoside was dissolved in ultrapure milliQ water, at a concentration of 10 mM, and it was stored at −20 ◦C.

2.2. Oxidation Reaction Typical oxidation reactions were carried out in a final volume of 100 µL of 50 mM phosphate buffer 5 (pH 6.5–7.5) with 20 µM mnm U and 40 µM or 2 mM KHSO5 and lasted 10 min at room temperature. The addition of 10 µL of 1 M hepes pH 8 buffer stopped the reaction. The reaction was analyzed by High Performance Liquid Chromatography (HPLC).

2.3. Chromatography HPLC analysis was done on a 5 µm C18 reverse phase column (uptisphere 5HDO, 250 × 4.6 mm from Interchim, Montluçon, France) eluted with a gradient of 5 mM ammonium acetate (NH4OAc) pH 6.5 aqueous buffer and methanol at a flow rate of 0.5 mL/min. After 5 min at 100% NH4OAc buffer, the percentage of methanol increased linearly to 30% in 20 min and jt remained at 30% for 5 min, before returning to the initial conditions. A diode array detector recorded the in-line UV-visible spectra. LC/ESI-MS analyses were carried out in the same liquid chromatography conditions.

2.4. Mass Spectrometry Either direct mass analysis or analysis through coupling with liquid chromatography was performed with the mass spectrometers, low resolution (Q TRAP 2000, Applied Biosystems, Foster Biomolecules 2018, 8, 145 3 of 18

City, CA, USA), and high resolution (Q TOF 1er, Waters, Milford, MA, USA) that were operated in the positive and/or negative mode.

2.5. Preparation of Aldonitrone Derivatives 1 and 2 Larger scale oxidation of mnm5U was performed in a final volume of 300 µL of 50 mM phosphate 5 pH 7.4 buffer, 1 mM mnm U was incubated with 4 mM KHSO5. Oxidation lasted 10 min at room temperature. The addition of 30 µL of 1 M hepes pH 8 buffer stopped the reaction. Products 1 and 2 were collected from liquid chromatography on a semi-preparative column and the fractions were lyophilized. The oxidation reaction was repeated three times. The semi-preparative 10 µm C18 reverse phase column (nucleosil, 250 × 10 mm from Interchim, Montluçon, France) was eluted with a gradient of 5 mM ammonium acetate (NH4OAc) pH 6.5 aqueous buffer and methanol at a flow rate of 2.4 mL/min. After 5 min at 100% NH4OAc buffer, the percentage of methanol increased linearly to 30% in 40 min, maintained at 30% for 5 min, and programmed to go back to initial conditions within 5 min. A diode array detector recorded the in-line UV-vis spectra of the products. Aldonitrone 1. UV-visible (H2O): λmax = 245, 263, and 319 nm. − HR-ESI < 0: [M − H] m/z for C11H14N3O7 calcd: 300.0832, obs: 300.0833 amu. + HR-ESI > 0: [M + H] m/z for C11H16N3O7 calcd: 302.0988, obs: 302.0994 amu. The nucleobase fragment C6H8N2O3 was observed at m/z = 170.0586 amu. 1 H-NMR (600 MHz, D2O) (298 K) δ (ppm): 9.79 (s, 1H, H6), 7.78 (s, 1H, H aldonitrone), 5.92 (d, 1H, J = 3.8 Hz, H1’), 4.27 (m, 1H, H2’), 4.12 (m, 1H, H3’), 4.07 (m, 1H, H4’), 3,84 (m, 1H, H5’), 3.76 (m, 1H, H5”), 3.75 (s, 3H, methyl). 13 C-NMR (151 MHz, D2O) δ (ppm): 163.2 (C4), 150.6 (C2), 142.2 (C6), 132.5 (aldonitrone), 105.9 (C5), 89.7 (C1’), 73.9 (C2’), 84.3 (C3’), 69.8 (C4’), 61.8 (C5’), 52.4 (CH3). Aldonitrone 2. UV-visible (H2O): λmax = 236 and 266 nm + HR-ESI > 0: [M + Na] m/z for C11H15N3O7Na calcd: 324.0808, obs: 324.0803 amu. The nucleobase fragment C6H7N2O3Na was observed at m/z = 192.0381 amu. 1 2 H-NMR (600 MHz, D2O) (278 K) δ (ppm): 7.91 (s, 1H, H6), 6.70 (d, 1H, J = 6.1 Hz, = CH2), 6.52 (d, 2 1H, J = 6.1 Hz, = CH2), 5.61 (m, 1H, H1’), 4.50 (under the signal of water, CH2), 4.04 (m, 1H, H2’), 3.93 (m, 1H, H3’), 3.82 (m, 1H, H4’), 3.63 (m, 1H, H5’), 3.52 (m, 1H, H5”). 13 C-NMR (151 MHz, D2O) (278 K) δ (ppm): 164.3 (C4), 151.1 (C2), 142.7 (C6), 134.1 (C = NO), 105.3 (C5), 89.0 (C1’), 74.0 (C2’), 68.7 (C3’), 84.2 (C4’), 60.5 (CH2), 60.2 (C5’). Hydroxylamine derivative 7. UV-visible (H2O): λmax = 265 nm. + HR-ESI > 0: [M + Na] m/z for C10H14N3O7Na calcd: 312.0808, obs: 312.0805 amu. 1 H-NMR (600 MHz, D2O) (278 K) δ (ppm): 7.66 (s, 1H, H6), 5.61 (m, 1H, H1’), 4.04 (m, 1H, H2’), 3.93 (m, 1H, H3’), 3.82 (m, 1H, H4’), 3,63 (m, 1H, H5’), 3.52 (m, 1H, H5”), 3.44 (m, 2H, CH2). 13 C-NMR (151 MHz, D2O) (278 K) δ (ppm): 164.3 (C4), 151.1 (C2), 140.6 (C6), 108.5 (C5), 89.0 (C1’), 74.0 (C2’), 68.7 (C3’), 84.2 (C4’), 49.2 (CH2), 60.2 (C5’).

2.6. Nuclear Magnetic Resonance

NMR samples were prepared by dissolving the samples in 350 µL of D2O using a Shigemi NMR tube from Sigma-Aldrich, Saint-Louis, MO, USA. 1D and 2D 1H and 13C experiments were recorded on a Bruker Avance I 500 MHz or Bruker NEO 600 MHz spectrometers (Bruker, Billerica, MA, USA). All chemical shifts for 1H and 13C are relative to TMS using 1H (residual) chemical shifts of the solvent as a secondary standard. All the 1H and 13C signals were assigned on the basis of the chemical shifts, the spin-spin coupling constants, the splitting patterns, and the signal intensities, and by using 1H-13C Biomolecules 2018, 8, 145 4 of 18

HSQC and 1H-13C HMBC experiments. 1D NOE (Nuclear Overhauser Effect) experiments were realized with a mixing time of 500 ms. Spectra were recorded at 298 K and/or 278 K as indicated.

2.7. Light Irradiation of Aldonitrone 1

Irradiation of 300 µLD2O solution of 1 was performed with a mercury lamp (Mercury lamp, Oriel, 100 W). The sample was kept in ice during irradiation. Irradiation (1.3 mW/cm2) was performed with a glass filter for 1 h and 30 min. The conversion was almost complete. Aldonitrone 1 transformed to two products in 1:1 ratio. The products of photodecomposition of 1, namely oxaziridine derivatives 3 and 4, were not isolated, but characterized as a 1:1 mixture in solution.

Oxaziridine (3 + 4) UV-visible (H2O): λmax = 269 nm. + Oxaziridine (3) HR-ESI > 0: [M + H] m/z for C11H16N3O7 calcd: 302.0988, obs: 302.1002 amu. + The nucleobase fragment C6H8N2O3 was observed at m/z = 170.0575 amu. [M + Na] m/z for C11H15N3O7Na calcd: 324.0808, obs: 324.0800. + Oxaziridine (4) HR-ESI > 0: [M + H] m/z for C11H16N3O7 calcd: 302.0988, obs: 302.1003 amu. + The nucleobase fragment C6H8N2O3 was observed at m/z = 170.0577 amu. [M + Na] m/z for C11H15N3O7Na calcd: 324.0808, obs: 324.0805. 1 Oxaziridine (3 + 4) H-NMR (600 MHz, D2O) (278 K) δ (ppm): 7.76 (s, 1H, H6, 3 or 4), 7.74 (s, 1H, H6, 3 or 4), 5.59 (m, 2H, H1’, 3 and 4), 4.50 (under signal of water, CH oxaziridine 3 and 4), 3.98 (m, 2H, H2’, 3 and 4), 3.90 (m, 2H, H3’, 3 and 4), 3.80 (m, 2H, H4’, 3 and 4), 3.65 (m, 2H, H5’, 3 and 4), 3.51 (m, 2H, H5”, 3 and 4), 2.56 (s, 6H, methyl, 3 and 4). 13 Oxaziridine (3 + 4) C-NMR (151 MHz, D2O) (278 K) δ (ppm) from HSQC and HMBC experiment: 164.5 (C4), 150.8 (C2), 140.6 (C6), 107.9 (C5), 89.3 (C1’), 75.7 (CH oxaziridine), 73.9 (C2’), 68.4 (C3’), 83.6 (C4’), 57.9 (C5’), 46.9 (CH3).

2.8. Modeling The geometry of the structures was optimized at the M06-2X/6-311++G(d,p) model chemistry with the GAUSSIAN 09 program [23]. The bulk solvent effects were described with the integral equation formalism polarizable continuum Model (IEFPCM) with water as solvent [24]. Vibrational frequencies were computed to confirm the convergence to local minima and to calculate the unscaled zero-point-energy (ZPE) and the Gibb’s free energy at 298 K.

3. Results

3.1. Oxidation of 5-methylaminomethyl uridine Leads to Two Oxidation Products Incubation of 20 µM mnm5U, in pH 7.5 phosphate buffer, for 10 min at room temperature, with KHSO5 (2 mM) afforded two oxidation products, 1 and 2. The HPLC analysis of the reaction is shown in Figures1 and2. The starting ribonucleoside eluted at a retention time of 19 min, and 1 and 2 at 24 and 21 min, respectively. According to the decrease of the HPLC area of the peak of initial mnm5U, the reaction resulted in ~50% conversion of mnm5U. Biomolecules 2018, 8, x FOR PEER REVIEW 5 of 18 Biomolecules 2018, 8, x FOR PEER REVIEW 5 of 18 Biomolecules 2018, 8, 145 5 of 18 Biomolecules 2018, 8, x FOR PEER REVIEW 5 of 18

Figure 1. HPLC trace of the oxidation of 5-methylaminomethyl uridine (mnm 5U) (20 µM) with KHSO5 5 5 Figure 1. HPLC trace of the oxidation of 5-methylaminomethyl uridine (mnm5U) (20 µM) with KHSO Figure(2 mM)1. forHPLC 10 min trace in 50 of themM oxidation phosphate of 5-methylaminomethylbuffer at pH 7.5. Conversion uridine 50%. (mnm 5U) (20 µM) with KHSO5 (2Figure mM) 1.for HPLC 10 min trace in 50of themM oxidation phosphate of buffer5-methylaminomethyl at pH 7.5. Conversion uridine 50%. (mnm U) (20 µM) with KHSO5 (2 mM) for 10 min in 50 mM phosphate buffer at pH 7.5. Conversion 50%. (2 mM) for 10 min in 50 mM phosphate buffer at pH 7.5. Conversion 50%.

Figure 2. In-line UV-visible spectra (220–400 nm) of Figure 2: 5-methylaminomethyl uridine (mnm5U) Figure 2. In-lineIn-line UV-visible UV-visible spectra spectra (220–400 (220–400 nm) nm) of of Figure Figure 22:: 5-methylaminomethyl uridine (mnm 5U)U) (Rt = 19 min), 2 (Rt = 21 21 min), min), and and 1 1 (Rt (Rt = = 24 24 min). min). 5 (RtFigure = 19 2. min), In-line 2 (Rt UV-visible = 21 min), spectra and 1 (220–400(Rt = 24 min). nm) of Figure 2: 5-methylaminomethyl uridine (mnm U) (Rt = 19 min), 2 (Rt = 21 min), and 1 (Rt = 24 min). To identifyidentify the the oxidation oxidation products, products, the reactionthe reaction was performedwas performed at higher at higher concentration concentration of mnm 5ofU To identify the oxidation products, the reaction was performed at higher concentration of 5 (1mnm mM)U and(1 mM) KHSO and5 (4KHSO mM)5 in(4 50mM) mM in phosphate 50 mM phosphate pH 7.5 buffer pH 7.5 for buffer 10 min. for Products10 min. Products of the reaction of the mnmTo5U (1identify mM) and the KHSOoxidation5 (4 mM)products, in 50 mMthe reactionphosphate was pH performed 7.5 buffer forat higher10 min. concentration Products of the of werereaction collected were collected separately separately and lyophilized. and lyophilized. reactionmnm5U were(1 mM) collected and KHSO separately5 (4 mM) and in lyophilized. 50 mM phosphate pH 7.5 buffer for 10 min. Products of the reaction were collected separately and lyophilized. 3.2. Characterization of Oxidation Product of mnm5U: Aldonitrone 1 3.2. Characterization of Oxidation Product of mnm5U: Aldonitrone 1 3.2. CharacterizationProduct 1 1 proved proved of Oxidationto to be be stable stable Product during during of isolation. mnm isolation.5U: High-resolutionAldonitrone High-resolution 1 negative negative electrospray electrospray mass analysis mass Product 1 proved to be stable− during isolation. High-resolution negative electrospray mass analysis analysisshowed a showed [M − H] a− [Mion− withH] m/zion = with300.0833m/z =corresponding 300.0833 corresponding to C11H14N3O to7, Cm/z11H (calcd)14N3O 7=, 300.0832m/z (calcd) amu = showedProduct a [M 1− proved H]− ion to with be stable m/z = during 300.0833 isolation. corresponding High-resolution to C11H negative14N3O7, m/z electrospray (calcd) = 300.0832mass analysis amu 300.0832(Figure S1). amu The (Figure molecular S1). Themass molecular of 1 is 301 mass g/mol, of 1 which is 301 g/mol,corresponds which to corresponds an increase toof an 14 increase amu with of (Figureshowed S1). a [M The − H]molecular− ion with mass m/z of= 1300.0833 is 301 g/mol, corresponding which corresponds to C11H14N3 Oto7 ,an m/z increase (calcd) of= 300.083214 amu withamu 14respect amu withto the respect molecular to the mass molecular of the mass initial of themnm initial5U ribonucleoside. mnm5U ribonucleoside. With such With a molecular such a molecular mass, respect(Figure S1).to the The molecular molecular mass mass of of the 1 is initial 301 g/mol, mnm 5whichU ribonucleoside. corresponds Withto an suchincrease a molecular of 14 amu mass, with mass,compound compound 1 might 1 might be either be either an aldonitrone, an aldonitrone, an anoxaziridines or oran an amide amide derivative derivative that that are are three compoundrespect to the1 might molecular be either mass an of aldonitrone, the initial mnman oxaziridines5U ribonucleoside. or an amide With derivativesuch a molecular that are mass,three isomers withwith the the same same molecular molecular formula, formula, and beand susceptible be susceptible to isomerization to isomerization through the through oxaziridine the isomerscompound with 1 mightthe same be either molecular an aldonitrone, formula, anand oxaziridines be susceptible or an amideto isomerization derivative thatthrough are three the derivativeoxaziridine (Scheme derivative1)[ (Scheme25,26]. Indeed,1) [25,26]. the Indeed, oxidation the oxidation of secondary of secondary by amine oxygen by atomoxygen donors atom oxaziridineisomers with derivative the same (Scheme molecular 1) [25,26]. formula, Indeed, and the be oxidation susceptible of secondary to isomerization amine by oxygenthrough atom the (,donors (hydroperoxides, dioxirane, dioxirane, peracids, peracids, oxone) mayoxone) lead may to aldonitrone,lead to aldonitrone, oxaziridine, oxaziridine, and/or and/or amide donorsoxaziridine (hydroperoxides, derivative (Scheme dioxirane, 1) [25,26]. peracids, Indeed, oxone) the oxidationmay lead ofto secondary aldonitrone, amine oxaziridine, by oxygen and/or atom derivative.amide derivative. Nitrones Nitrones are mainly are mainly prepared prepared by oxidation by oxidation of hydroxylamines of hydroxylamines [27,28]. [27,28]. They may They also may be amidedonors derivative. (hydroperoxides, Nitrones dioxirane, are mainly peracids, prepared oxone) by oxidation may lead of tohydroxylamines aldonitrone, oxaziridine, [27,28]. They and/or may obtainedalso be obtained from oxidation from oxidation of of [imines29] and [29] and [amines30–33]. [30–33]. Usual oxidants Usual ox areidants oxygen are atomoxygen donors, atom alsoamide be derivative.obtained from Nitrones oxidation are mainly of imines prepared [29] and by amines oxidation [30–33]. of hydroxylamines Usual oxidants [27,28]. are oxygen They atom may suchdonors, as hydroperoxidessuch as hydroperoxides (including (including H2O2) activated H2O2) activated by metal by ions, metal peracids, ions, andperacids, dioxirane. and dioxirane. However, donors,also be obtainedsuch as hydroperoxides from oxidation of(including imines [29] H2 Oand2) activated amines [30–33]. by metal Usual ions, ox peracids,idants are and oxygen dioxirane. atom oxaziridinesHowever, oxaziridines are the most are common the most oxidation common productsoxidation of products imines with of imines a variety with of a reagentsvariety of [26 reagents,34–37]. However,donors, such oxaziridines as hydroperoxides are the most (including common H oxidation2O2) activated products by metal of imines ions, with peracids, a variety and of dioxirane. reagents On[26,34–37]. the other On hand, the oxidationother hand, of primaryoxidation amines of primary by oxone amines has beenby oxone reported has tobeen afford reported nitrones to [ 38afford–40]. [26,34–37].However, oxaziridinesOn the other are hand, the most oxid ationcommon of primary oxidation amines products by oxoneof imines has with been a reportedvariety of to reagents afford Upnitrones to now, [38–40]. literature Up to reports now, literature deal with reports oxidations deal that with are oxidations not performed that are in water.not performed in water. nitrones[26,34–37]. [38–40]. On the Up other to now, hand, literature oxidation reports of primary deal with amines oxidations by oxone that arehas notbeen performed reported into water. afford nitrones [38–40]. Up to now, literature reports deal with oxidations that are not performed in water.

5 Scheme 1.1.Possible Possible oxidation oxidation products products of5-methylaminomethyl of 5-methylaminomethyl uridine uridine (mnm (mnmU) with5U) a masswith increasea mass Scheme 1. Possible oxidation 5products of 5-methylaminomethyl uridine (mnm5U) with a mass ofincrease 14 amu of with 14 amu respect with to respect mnm U.to mnm5U. increaseScheme of1. 14Possible amu with oxidation respect toproducts mnm5U. of 5-methylaminomethyl uridine (mnm5U) with a mass increase of 14 amu with respect to mnm5U. Biomolecules 2018, 8, 145 6 of 18 Biomolecules 2018, 8, x FOR PEER REVIEW 6 of 18

The observed strong π−π* band at 319 nm on UV-visible spectrum of 1 (Figure 2) is compatible The observed strong π−π* band at 319 nm on UV-visible spectrum of 1 (Figure2) is compatible with an N- aldonitrone function as shown in Scheme 1 [34,41,42]. The aldonitrone structure of 1 with an N-alkyl aldonitrone function as shown in Scheme1[ 34,41,42]. The aldonitrone structure of 1 was unambiguously determined by NMR. The 600 MHz H-NMR1 spectrum of 1 in D2O is shown in 1 was unambiguously determined by NMR. The 600 MHz H-NMR spectrum of 1 in D2O is shown Figure 3. To make it simple, the numbering of atoms corresponds to that of uridine. Oxidation in Figure3. To make it simple, the numbering of atoms corresponds to that of uridine. Oxidation product 1 is a ribonucleoside with the sugar protons showing typical resonances at δ 5.92 ppm for product 1 is a ribonucleoside with the sugar protons showing typical resonances at δ 5.92 ppm for H1’ 3 H1’3 ( JHH = 3.5 Hz) and between 4.27 and 3.76 ppm for H2’, H3’, H4’, H5’, and H5”. One ( JHH = 3.5 Hz) and between 4.27 and 3.76 ppm for H2’, H3’, H4’, H5’, and H5”. One methyl group can be observed at δ 3.75 ppm (accounting for 3 H). Two downfield singlet resonances at δ 7.78 ppm can be observed at δ 3.75 ppm (accounting for 3 H). Two downfield singlet resonances at δ 7.78 ppm and δ 9.79 ppm, each of them integrating for 1 H, were attributed to the H of an aldonitrone function and δ 9.79 ppm, each of them integrating for 1 H, were attributed to the H of an aldonitrone function and to the H6 proton of nucleobase, respectively, from HSQC and HMBC experiments, as detailed and to the H6 proton of nucleobase, respectively, from HSQC and HMBC experiments, as detailed below. From HSQC the protons δ 7.78 ppm (H aldonitrone) and δ 9.79 ppm (H6) correlated through below. From HSQC the protons δ 7.78 ppm (H aldonitrone) and δ 9.79 ppm (H6) correlated through 1 δ δ 1JCH with the 132.5 ppm (C aldonitrone) and 142.2 ppm (C6), respectively. From HMBC JCH with the carbons δ 132.5 ppm (C aldonitrone) and δ 142.2 ppm (C6), respectively. From HMBC data the protons of the methyl group (δ 3.75 ppm) correlated with the carbon δ 132.5 ppm through 3JCH. data the protons of the methyl group (δ 3.75 ppm) correlated with the carbon δ 132.5 ppm through Therefore, we attributed the H signal δ 7.78 ppm, connected to the carbon signal δ 132.5 ppm (1JCH), to the 3J . Therefore, we attributed the H signal δ 7.78 ppm, connected to the carbon signal δ 132.5 ppm CH δ H1 atom of the aldonitrone function. Furthermore, in HMBC experiment both H signals at 7.78 ppm and ( JCH), to the H atom of the aldonitrone function. Furthermore, in HMBC experiment both H signals at δ 3 δ 5 at 9.79 ppm showed a JCH correlation3 with the carbon 163.2 ppm (attributed to the C4 of mnm U δ 7.78 ppm and at δ 9.79 ppm showed a JCH correlation with the carbon δ 163.2 ppm (attributed to the by analogy with reported NMR of thymine as described in the literature) while only the δ 9.79 ppm C4 of mnm5U by analogy with reported NMR of thymine as described in the literature) while only correlated with the carbon δ 150.6 ppm (attributed to the C2 of mnm5U) (Figure S2A). The proton at δ 9.79 the δ 9.79 ppm correlated with the carbon δ 150.6 ppm (attributed to the C2 of mnm5U) (Figure S2A). 3 δ ppm also showed a JCH correlation with the3 carbon 132.5 ppm, which is in accordance with the fact The proton at δ 9.79 ppm also showed a JCH correlation with the carbon δ 132.5 ppm, which is in that it corresponds to the H6 proton of modified uracil. accordance with the fact that it corresponds to the H6 proton of modified uracil.

11 FigureFigure 3.3. H-NMR (600 MHz) spectrum of 1 in D2O at 298 K.

The mechanism of formation of product 1 isis proposedproposed inin SchemeScheme2 2.. TheThe secondarysecondary amineamine ofof 5 mnm5UU is oxidized by KHSO 5 toto the the intermediate intermediate hydroxylamine hydroxylamine derivative. The loss of a molecule of water leads to the imine, which is oxidized byby aa secondsecond moleculemolecule ofof KHSOKHSO55. Biomolecules 2018, 8, 145 7 of 18 Biomolecules 2018, 8, x FOR PEER REVIEW 7 of 18 Biomolecules 2018, 8, x FOR PEER REVIEW 7 of 18

5 Scheme 2. Proposed mechanism of oxidationoxidation ofof 5-methylaminomethyl5-methylaminomethyl uridineuridine (mnm(mnm5U)U) withwith oxone: Scheme 2. Proposed mechanism of oxidation of 5-methylaminomethyl uridine (mnm5U) with oxone: formation of of the the stable stable product product 1. 1. Numbering Numbering of of atoms atoms corresponds corresponds to to the the numbering numbering of of uridine. uridine. formation of the stable product 1. Numbering of atoms corresponds to the numbering of uridine. Only one methyl signal and only one azomethine proton was observed in the 1H-NMR spectrum. Only one methyl signal and only one azomethine proton was observed in the 1H-NMR spectrum. Only one methyl signal and only one azomethine proton was observed in the 1H-NMR spectrum. This probably probably implies implies that that only only one onealdonitrone aldonitrone stereo stereo isomer isomer formed formed in the reaction. in the reaction.The NOE correlation The NOE This probably implies that only one aldonitroneδ stereo isomer formed in the reaction. The NOEδ correlation correlationbetween the betweenmethyl group the methyl (δ 3.75group ppm) and ( 3.75 the ppm)H proton and of the aldonitrone H proton ( ofδ 7.78 aldonitrone ppm) unambiguously ( 7.78 ppm) between the methyl group (δ 3.75 ppm) and the H proton of aldonitrone (δ 7.78 ppm) unambiguously unambiguouslyassigns the Z configuration, assigns the Zwhichconfiguration, is the preferred which conf is theiguration preferred of aldonitron configurationes (Figure of aldonitrones S2B). The assigns the Z configuration, which is the preferred configuration of aldonitrones (Figure S2B). The (Figuredeshielding S2B). effect The deshieldingof aldonitrone effect oxygen of aldonitrone in the Z-geometry oxygen incauses the Z -geometrya lower field causes absorption a lower of field the deshielding effect of aldonitrone oxygen in the Zδ-geometry causes a lower field absorption of the5 absorptionnucleobase H6 of proton the nucleobase of 1, δ 9.79 H6ppm, proton with respect of 1, to9.79 the H6 ppm, of mnm with5U respectδ 8.35 ppm to the[9]. H6 of mnm U δnucleobase H6 proton of 1, δ 9.79 ppm, with respect to the H6 of mnm5U δ 8.35 ppm [9]. 8.35Molecular ppm [9]. modeling confirms the lower energy of the Z as compared to the E isomer. The Z Molecular modeling confirms the lower energy of the Z as compared to the E isomer. The Z conformationMolecular shows modeling a planar confirms geometry the loweroptimizing energy electron of the Zdelocalization as compared between to the E the isomer. aldonitrone The Z conformation shows a planar geometry optimizing electron delocalization between the aldonitrone conformationfunction and the shows aromatic a planar ring geometry of nucleobase optimizing (Figure electron 4). The delocalizationE conformation between is less favorable the aldonitrone due to function and the aromatic ring of nucleobase (Figure 4). The E conformation is less favorable due to functionsteric hindrance and the between aromatic the ring methyl of nucleobase group and (Figure the H64). proton The E conformationof the nucleobase. is less favorable due to steric hindrance between the methyl group and the H6 proton of the nucleobase. stericThe hindrance proposed between structure the methylof N-methyl-aldonitrone group and the H6 proton1 is compatible of the nucleobase. with previously published The proposed structure of N-methyl-aldonitrone 1 is compatible with previously published NMRThe data proposed for aryl-conjugated structure of Naldonitrones-methyl-aldonitrone [38,43,44]. 1 is compatible with previously published NMR dataNMR for data aryl-conjugated for aryl-conjugated aldonitrones aldonitrones [38,43, 44[38,43,44].].

Figure 4. Optimized geometry of (Z) aldonitrone 1 showing a planar arrangement for optimal electron Figure 4. Optimized geometry of ( Z) aldonitrone aldonitrone 1 showing a planar arrangem arrangementent for optimal electron delocalization between aldonitrone moiety and nucleobase. delocalization between aldonitrone moietymoiety andand nucleobase.nucleobase. Oxaziridines (Scheme 1) are often reported in the literature as the products of imine oxidation Oxaziridines (Scheme(Scheme1 )1) are are often often reported reported in in the th literaturee literature as as the the products products of imineof imine oxidation oxidation by by peracids [34,35,37]. During the oxidation of mnm5 5U with oxone, N-methyl-oxaziridine derivative peracidsby peracids [34 ,[34,35,37].35,37]. During During the the oxidation oxidation of mnm of mnmU with5U with oxone, oxone,N-methyl-oxaziridine N-methyl-oxaziridine derivative derivative did did not form as no NMR signal was observed at the typical chemical shift of the CH proton of notdid formnot form as no as NMR no signalNMR wassignal observed was observed at the typical at the chemical typical shiftchemical of the shift CH protonof the ofCH oxaziridines, proton of oxaziridines, namely around δ 4–5 ppm [45]. Previous use of oxone for secondary amine oxidation namelyoxaziridines, around namelyδ 4–5 around ppm [45 δ ].4–5 Previous ppm [45]. use Previous of oxone use for of secondary oxone for amine secondary oxidation amine also oxidation led to also led to nitrone [33,38–40]. nitronealso led [to33 ,nitrone38–40]. [33,38–40]. 3.3. Photochemical CharacterizationCharacterization of Aldonitrone 1 3.3. Photochemical Characterization of Aldonitrone 1 Nitrones are are known known to totransform transform to oxaziridines to oxaziridines when when irradiated irradiated [25,37,41,46–48]. [25,37,41,46 Therefore,–48]. Therefore, isolated Nitrones are known to transform to oxaziridines when irradiated [25,37,41,46–48].λ Therefore, isolated isolatedaldonitrone aldonitrone 1 was exposed 1 was to exposed light in toH2 lightO in an in ice H2 Obath, in anwith ice a bath,glass withfilter a( glass cut-off filter 300 (nm).λ cut-off The 300reaction nm). aldonitrone 1 was exposed to light in H2O in an ice bath, with a glass filter (λ cut-off 300 nm). The reaction Thewas reactionfollowed was by followedHPLC. Aldonitrone by HPLC. Aldonitrone1 transformed 1 transformedto two products to two 3 and products 4 (in 3about and 4equimolar (in about was followed by HPLC. Aldonitrone 1 transformed to two products 3 and 4 (in about equimolar equimolaramount) that amount) eluted thatat Rt eluted = 27 and at Rt 29 = min, 27 and respecti 29 min,vely. respectively. The reaction The was reaction almost was complete almost within complete 30 amount) that eluted at Rt = 27 and 29 min, respectively. The reaction was almost complete within 30 withinmin (Figures 30 min 5 (Figuresand 6). 5Exposure and6). Exposure of 1 to day of 1light to day on lightthe laboratory on the laboratory bench led bench to the led same to the products same min (Figures 5 and 6). Exposure of 1 to day light on the laboratory bench led to the same products products(50% of conversion (50% of conversion of 1 after of1h 1at after room 1h temperature) at room temperature) (not shown). (not shown). (50% of conversion of 1 after 1h at room temperature) (not shown). Biomolecules 2018, 8, x FOR PEER REVIEW 8 of 18

Biomolecules 2018, 8, 145 8 of 18 Biomolecules 2018, 8, x FOR PEER REVIEW 8 of 18

Figure 5. HPLC trace (detection 260 nm) of irradiation of a solution of isolated 1 in water during 30 min showing the two products 3 and 4 at Rt = 27 min and 29 min, respectively. Figure 5. HPLC trace (detection 260 nm) of irradiation of a solution of isolated 1 in water during 30 minmin showing the two products 3 and 4 at Rt = 27 min and 29 min, respectively. showing the two products 3 and 4 at Rt = 27 min and 29 min, respectively.

Figure 6. In-line UV-visible spectra (220–400 nm) corresponding to Figure5: 1 (Rt = 24 min), Figure 6. In-line UV-visible spectra (220–400 nm) corresponding to Figure 5: 1 (Rt = 24 min), 3 (Rt = 27 min), and 4 (Rt = 29 min). 3 (Rt = 27 min), and 4 (Rt = 29 min). TheFigure in-line 6. In-line high resolutionUV-visible massspectra analysis (220–400 of thenm) twocorre compoundssponding to (3 Figure and 4) 5: reveals 1 (Rt = that 24 bothmin), have The3 (Rt in-line = 27 min), high and resolution 4 (Rt = 29 mass min). analysis of the two compounds (3 and 4) reveals that both have the same molecular mass as 1, 301 g/mol, and the same molecular formula, C11H15N3O7 (Figure S3). the same molecular mass as 1, 301 g/mol, and the same molecular formula, C11H15N3O7 (Figure S3). Irradiation of an NMR tube containing 1 in D2O during 1 h and 30 min in ice allowed us to Irradiation of an NMR tube containing 1 in D2O during 1 h and 30 min in ice allowed us to characterizeThe in-line the high mixture resolution of the twomass products analysis (3of andthe two 4) by compounds NMR. The (31 Hand spectrum 4) reveals of that 3 + 4both mixture have 1 characterizethe same molecular the mixture mass of as the 1, two301 productsg/mol, and (3 theand same 4) by molecular NMR. The formula, H spectrum C11H 15ofN 33 O+ 74 (Figuremixture S3). (1:1 (1:1 ratio) in D2O at 278 K is shown in Figure7. The conversion of 1 into 3 + 4 was almost complete, ratio) in D2O at 278 K is shown in Figure 7. The conversi2 on of 1 into 3 + 4 was almost complete, since sinceIrradiation the signals of aldonitronean NMR tube accounted containing for only1 in 10%D O (asduring deduced 1 h fromand 30 comparison min in ice of allowed the methyl us ofto the signals of aldonitrone accounted for only 10% (as deduced from comparison1 of the methyl of the thecharacterize starting product the mixture 1 δ 3.41 of the ppm two with products the methyl (3 and of 4) the bynew NMR. products The H δspectrum2.56 ppm). of 3 + 4 mixture (1:1 starting product2 1 δ 3.41 ppm with the methyl of the new products δ 2.56 ppm). ratio)From in D theO at NMR 278 K analyses is shown 3 in and Figure 4 were 7. The identified conversi ason two of 1 oxaziridineinto 3 + 4 was derivatives. almost complete, Two singlet since the signalsFrom the of aldonitroneNMR analyses accounted 3 and 4for were only identified 10% (as deduced as two oxaziridinefrom comparison derivatives. of the methylTwo singlet of the signals at δ 7.76 ppm and δ 7.74 ppm were assigned to two H6 protons of nucleobases (each integrating signalsstarting at product δ 7.76 1ppm δ 3.41 and ppm δ with7.74 theppm methyl were ofassigned the new to products two H6 δ 2.56protons ppm). of nucleobases (each for 1H). The H1’ proton appeared as a broad signal at δ 5.59 ppm (integration 2H). Ribose protons integratingFrom forthe 1H).NMR The analyses H1’ proton 3 and appeared 4 were asidentified a broad signalas two at oxaziridine δ 5.59 ppm derivatives.(integration 2H).Two Ribosesinglet (10 H) were found between δ 3.91 ppm and δ 3.51 ppm. Finally, a singlet accounting for 6 H was protonssignals (10at δH) 7.76 were ppm found and between δ 7.74 δppm 3.91 wereppm andassigned δ 3.51 toppm. two Finally, H6 protons a singlet of accountingnucleobases for (each 6 H observed at δ 2.56 ppm and it was attributed to the methyl groups. Thus, although 3 and 4 showed wasintegrating observed for at 1H). δ 2.56 The ppmH1’ proton and it appearedwas attributed as a broad to the signal methyl at δ 5.59groups. ppm Thus, (integration although 2H). 3 Riboseand 4 two distinct resonances for H6 and H1’, two separate methyl groups could not be identified, only one showedprotons two (10 H)distinct were resonancesfound between for H6 δ 3.91 and ppmH1’, twando separateδ 3.51 ppm. methyl Finally, groups a singlet could accounting not be identified, for 6 H singlet accounting for 6 H appeared at δ 2.56 ppm. onlywas oneobserved singlet at accounting δ 2.56 ppm for and 6 H itappeared was attributed at δ 2.56 to ppm. the methyl groups. Thus, although 3 and 4 showed two distinct resonances for H6 and H1’, two separate methyl groups could not be identified, only one singlet accounting for 6 H appeared at δ 2.56 ppm. Biomolecules 2018, 8, 145 9 of 18 Biomolecules 2018, 8, x FOR PEER REVIEW 9 of 18

11 Figure 7. H-NMRH-NMR (600 (600 MHz, MHz, 278 278 K) K) spectrum spectrum of of irradiated irradiated 1 1 in D 2O.O. The The tube tube contains 90% of oxaziridines 3 and 4 in a 50:50 ratio and 10% of 1. The solvent peak at δ 4.7 ppm is suppressed.

NOE experiment on the 3 + 4 4 mixture mixture showed showed a a correlation correlation between the H6 protons ( δ 7.76 and 7.74 ppm)ppm) andand some some protons, protons, the the resonances resonances of whichof which appeared appeared under under the signal the signal of water of (twowater singlets (two singletsat δ ~ 4.5 at ppm). δ ~ 4.5 Another ppm). Another NOE correlation NOE correlation was also was noted also between noted between the protons the pr ofotons the methylof the methyl group group(δ 2.56 (δ ppm) 2.56 ppm) and the and same the same protons protons under under the the signal signal of water.of water. These These two two hidden hidden1 H1H resonancesresonances that were located under the water signal were attributed to to the the CH CH proton proton of of the the oxaziridine oxaziridine function. function. The chemical shift shift of of the the CH CH proton proton of of oxaziridin oxaziridineses had had been been previously previously reported reported as as being being around around δ 4–5δ 4–5 ppm ppm [45], [45 which], which is isin inaccordance accordance with with the the present present data data for for 3 and 3 and 4. 4. Furthermore, as deduceddeduced fromfrom thethe correlationcorrelation between between the the methyl methyl and and the the hidden hidden H-atom, H-atom, the the H Hatom atom of oxaziridine,of oxaziridine, and and the methylthe methyl group group are on are the on same the sidesame of side the oxaziridineof the oxaziridine cycle. This cycle. implies This impliesthat among that among the four the possible four possible isomers, isomers, the twothe two observed observed oxaziridines oxaziridines are are in inanti anti conformationconformation (Scheme 33).). A 2D 2D 1H/1H/13C13 HSQCC HSQC NMR NMR experiment experiment assigned assigned the C6 carbons the C6 of carbons the two nucleobases of the two (δ nucleobases 140.6 ppm) connected(δ 140.6 ppm to )H6 connected protons ( toδ 7.76 H6 protons and 7.74 ( δppm),7.76 and the 7.74C1’ carbons ppm), the (δ 89.4 C1’ carbonsppm) connected (δ 89.4 ppm) to H1’ connected protons (toδ H1’5.6 ppm), protons the (δ methyl5.6 ppm), carbons the methyl (δ 46.9 carbons ppm) (connectedδ 46.9 ppm) to connected methyl protons to methyl (δ protons2.56 ppm), (δ 2.56 and ppm), new carbonsand new (δ carbons 75.7 ppm) (δ 75.7 connected ppm) connected to hidden to protons hidden (δ protons ~ 4.5 ppm). (δ ~4.5 ppm). Furthermore, weak 1H/H/1313CC HMBC HMBC correlations correlations could could be be detected detected between between the the protons protons δδ ~~ 4.5 4.5 ppm and the C6 carbons δ 140.6 ppm, the methyl carbons δ 46.9 ppm, and carbons δ 108.0 ppm (C5 of nucleobase). Similarly, Similarly, the carbons δ 75.7 ppm showed HMBC correlations with the H6 and methyl protons. These datadata confirmconfirm the the proposed proposed oxaziridine oxaziridine function function on on the the uridine uridine substituent. substituent. The 13TheC and13C and1H resonances 1H resonances of the of sugarthe sugar moiety moiety were were also also unambiguously unambiguously assigned. assigned. Thus, irradiation of aldonitrone aldonitrone 1 led led to to the the mixture mixture of of two two antianti diastereoisomersdiastereoisomers of of oxaziridines, oxaziridines, S-anti and R-anti (Scheme 33 andand FigureFigure8 8).). SS and and R R refer refer to to the the configuration configuration of of the the carbon carbon atom atom of of oxaziridine. The configuration configuration of the nitrogen atom of oxaziridine is blocked. The The reaction reaction was was almost almost complete (yield 90%) after 1 h and 30 min in ice. Biomolecules 2018, 8, x FOR PEER REVIEW 10 of 18

Biomolecules 2018, 8, 145 10 of 18 Biomolecules 2018, 8, x FOR PEER REVIEW 10 of 18

Scheme 3. Formation of two epimers of oxaziridine, derivatives 3 and 4, in about equimolar amount,

upon irradiation of aldonitrone 1. Numbering of atoms corresponds to the numbering of uridine. SchemeScheme 3. Formation 3. Formation of twoof two epimers epimers of of oxaziridine, oxaziridine, derivatives derivatives 3 and 3 and 4, in 4, about in about equimolar equimolar amount, amount, uponupon irradiation irradiation of aldonitrone of aldonitrone 1. 1. Numbering Numbering ofof atoms corresponds corresponds to tothe the numbering numbering of uridine. of uridine.

Figure 8. Geometry optimization of oxaziridine derivatives 3 and 4. FigureFigure 8. Geometry8. Geometry optimization optimization of oxaziridine oxaziridine derivatives derivatives 3 and 3 and 4. 4. 3.4.3.4. Stability Stability of of Oxaziridi Oxaziridinesnes 3 3 and and 4 4 in in the the Dark Dark 3.4. Stability of Oxaziridines 3 and 4 in the Dark AfterAfterAfter incubation incubation incubation of of the of the theNMR NMR NMR tube tubetube containing containing oxaziroxazir oxaziridinesidines 3 3 + + 4 3 4mixture +mixture 4 mixture in inwater water in in water the in thedark in dark the in the dark in the in fridgethe fridgefridge for three for three threeweeks weeks weeks the theNMR theNMR spectrum NMR spectrum spectrum corresponded corresponded corresponded to thethe one one to of of thepure pure one initial initial of aldonitrone pure aldonitrone initial 1. Therefore, aldonitrone 1. Therefore, 1. ◦ photo-isomerizationTherefore,photo-isomerization photo-isomerization of 1of is1 compleis ofcomple 1 istely completelytely reversible reversible reversible inin the darkdark in the atatdark ~5 ~5 °C. at°C. ~5Thermal ThermalC. Thermal isomerization isomerization isomerization of of oxaziridinesof oxaziridinesoxaziridines to nitrone to nitrone nitrone and/or and/or and/or amide amide amideis isa commona common is a common prop property property ofof these these cyclic cyclic of thesecompounds compounds cyclic [25,49]. compounds [25,49]. Incubation Incubation [25 ,49]. atIncubation a higherat a higher temperature at atemperature higher (37–60 temperature (37–60 °C) °C) of ofthe (37–60 the 3 3+ +4 4mixture◦ mixtureC) of the gagave 3 + riserise 4 mixturenot not only only to gaveto 1, 1,but but rise also also notto aldehyde to only aldehyde to (5) 1, and but (5) and also amideto aldehydeamide (6) derivatives. (6) (5) derivatives. and amideThe The percentage (6)percentage derivatives. of of aldehyde aldehyde The and and percentage amide increased increased of aldehyde with with temperature andtemperature amide (Figures increased (Figures 9 and 9 withand 10).temperature The10). products The products (Figures were were9 ident and ident ified10).ified Theby by LC-ESI-MS products LC-ESI-MS were analysis. analysis. identified by LC-ESI-MS analysis.

Figure 9. HPLC analysis of the transformation of oxaziridines 3 and 4 back to 1 in the dark. Upper trace, at t0 the mixture of 3 and 4 at Rt = 30 min and 32 min, respectively. Middle trace, 1 (Rt = 26 min) is the major product. It is associated with the aldehyde derivative (5) (Rt = 27 min) and the amide FigureFigure 9. HPLC analysis ofofthe the transformation transformation of of oxaziridines oxaziridines 3and 3 and 4 back 4 back to 1 to in 1 the in dark.the dark. Upper Upper trace, at t0 thederivative mixture (6) of (Rt 3 and= 32 4min). at Rt Lower = 30 min trace, and the 32 pr min,oportion respectively. of 5 and 6 Middle becomes trace, more 1 important (Rt = 26 min) with is the trace, respectat t0 the to mixture1 (Rt = 26 of min). 3 and rt: room4 at Rt temperature. = 30 min an d 32 min, respectively. Middle trace, 1 (Rt = 26 min) ismajor the major product. product. It is associated It is associated with the with aldehyde the aldehyde derivative derivative (5) (Rt = (5) 27 min)(Rt = and27 min) the amide and the derivative amide derivative(6) (Rt = 32 (6) min). (Rt = Lower 32 min). trace, Lower the proportiontrace, the pr ofoportion 5 and 6 becomesof 5 and more6 becomes important more withimportant respect with to 1 respect(Rt = 26 to min). 1 (Rt rt: = 26 room min). temperature. rt: room temperature. Biomolecules 2018, 8, 145x FOR PEER REVIEW 11 of 18 Biomolecules 2018, 8, x FOR PEER REVIEW 11 of 18

Figure 10. In-line UV spectra (220–350 nm) corresponding to Figure 9, compounds 1, 5, and 6. Figure 10. In-lineIn-line UV UV spectra spectra (220–350 (220–350 nm) nm) corresponding corresponding to Figure 99,, compoundscompounds 1,1, 5,5, andand 6.6.

The mixture of oxaziridines 3 3 and and 4 4 was incubated in in water water in in the the dark dark at at 37 37 or or 60 60 °C.◦C. The mixture of oxaziridines 3 and 4 was incubated in water in the dark at 37 or 60 °C. Compounds 3 and 4 transformed (principally) (principally) to to aldonitrone aldonitrone 1 1 and and to to two two other other products products (5 (5 and and 6). 6). Compounds 3 and 4 transformed (principally) to aldonitrone 1 and to two other products (5 and 6). Aldonitrone 11 in-linein-line mass mass spectrum spectrum (positive (positive electrospray) electrospray) showed showed signals signals at m/z at m/z= 324.0808 = 324.0808 amu, amu, and Aldonitrone 1 in-line mass spectrum (positive electrospray) showed signals at m/z = 324.0808 amu, andm/z =m/z 170.0565 = 170.0565 amu ,amu, which which correspond correspond to the to sodium the sodium adduct adduct of general of general formula formula C11H C1511NH315ON73NaO7Na (1) and m/z = 170.0565 amu, which correspond to the sodium adduct of general formula C11H15N3O7Na (1)and and to the to nucleobasethe nucleobase fragment fragment C6H8 NC36OH38,N respectively.3O3, respectively. The in-line The massin-line spectrum mass spectrum of 5, the aldehydeof 5, the (1) and to the nucleobase fragment C6H8N3O3, respectively. The in-line mass spectrum of 5, the aldehydederivative, derivative, showed signals showed at m/zsignals= 295.0544 at m/z amu,= 295.0544 and m/z amu,= 141.0299 and m/z amu, = 141.0299 which correspond amu, which to aldehyde derivative, showed signals at m/z = 295.0544 amu, and m/z = 141.0299 amu, which correspondthe sodium adductto the sodium of general adduct formula of general C10H 12formulaN2O7Na C10 andH12N to2O the7Na nucleobase and to the fragmentnucleobase C5 fragmentH5N2O3, correspond to the sodium adduct of general formula C10H12N2O7Na and to the nucleobase fragment Crespectively.5H5N2O3, respectively. Finally, theFinally, in-line the in-line mass spectrummass spectrum of 6, of the 6, the amide amide derivative, derivative, showed showedsignals signals at C5H5N2O3, respectively. Finally, the in-line mass spectrum of 6, the amide derivative, showed signals at m/z == 324.0807 324.0807 amu, amu and, and m/zm/z = =170.0567 170.0567 amu, amu, that that correspond correspond to to the the sodium sodium adduct adduct of of general general formula formula m/z = 324.0807 amu, and m/z = 170.0567 amu, that correspond to the sodium adduct of general formula C11HH1515NN3O3O7Na7Na and and to tothe the nucleobase nucleobase fragment fragment C6H C68NH38ON33, Orespectively3, respectively (Figure (Figure S4). S4).Both Both aldehyde aldehyde and C11H15N3O7Na and to the nucleobase fragment C6H8N3O3, respectively (Figure S4). Both aldehyde and amideand amide are common are common products products of hydrolysis of hydrolysis (aldehyde (aldehyde 5) 5)or or isomerization isomerization (amide (amide 6) 6) of of aldonitrone aldonitrone amide are common products of hydrolysis (aldehyde 5) or isomerization (amide 6) of aldonitrone (Scheme4 4).). (Scheme 4).

Scheme 4. Stability of oxaziridines 3 and 4 in the dark. At low temperature only isomerization to Scheme 4. Stability of oxaziridines 3 and 4 in the dark. At low temperature only isomerization to aldonitrone is observed. Upon heating isomerization to amide 6 and hydrolysis of oxaziridines to the aldonitrone is observed. Upon heating isomerization to amide 6 and hydrolysis of oxaziridines to the aldehyde derivative 5 is observed together with back isomerization to aldonitrone 1.1. aldehyde derivative 5 is observed together with back isomerization to aldonitrone 1. ◦ The kinetic ofof thethe reactionreaction waswas faster faster at at 60 60 C°C (total (total transformation transformation of of 3 +3 + 4 in4 in 1 h)1 h) as as compared compared to ◦ The kinetic of the reaction was faster at 60 °C (total transformation of 3 + 4 in 1 h) as compared to37 37C °C (total (total reaction reaction after after 15 15 h) h) (Figure (Figure 10 ).10). The The ratio ratio between between the the three three products products varied varied depending depending on to 37 °C (total reaction after 15 h) (Figure 10). The ratio between the three products varied depending onthe the temperature temperature with with the the formation formation of 5 of and 5 and 6 being 6 being favored favored at higher at higher temperature temperature in accordance in accordance with on the temperature with the formation of 5 and 6 being favored at higher temperature in accordance◦ withliterature literature [25]. [25]. On the On other the other hand, hand, aldonitrone aldonitrone 1 proved 1 proved stable stable upon upon incubation incubation in water in water at 90 at 90C in°C the in with literature [25]. On the other hand, aldonitrone 1 proved stable upon incubation in water at 90 °C in thedark dark for 5hfor and 5h and stable stable for 1h for in 1h acidic in acidic conditions conditions (0.1M (0.1 acetic M acetic acid inacid water) in water) at room at room temperature temperature in the the dark for 5h and stable for 1h in acidic conditions (0.1 M acetic acid in water) at room temperature indark. the Therefore,dark. Therefore, while aldonitronewhile aldonitrone 1 is a stable 1 is a compound stable compound in the dark, in the oxaziridines dark, oxaziridines 3 and 4 undergo3 and 4 in the dark. Therefore, while aldonitrone 1 is a stable compound in the dark, oxaziridines 3 and 4 undergoisomerization isomerization (to 1 at low (to temperature1 at low temperature and to amide and 6to at amide higher 6 temperature) at higher temperature) and/or hydrolysis and/or undergo isomerization (to 1 at low temperature and to amide 6 at higher temperature) and/or hydrolysis(aldehyde 5).(aldehyde 5). hydrolysis (aldehyde 5). BiomoleculesBiomolecules 20182018,, 88,, x 145 FOR PEER REVIEW 1212 of of 18 18 Biomolecules 2018, 8, x FOR PEER REVIEW 12 of 18 3.5. Characterization of Oxidation Product of mnm5U: Aldonitrone 2 3.5.3.5. Characterization Characterization of of Oxidation Oxidation Product Product of of mnm mnm55U:U: Aldonitrone Aldonitrone 2 2 The second oxidation product of mnm5U ribonucleoside, product 2 (Figure 1), was isolated by LC chromatography.TheThe second second oxidation oxidation Unfortunately, product product of of it mnm was 5not5UU ribonucleoside, ribonucleoside,stable during the product product lyophilization 2 2 (F (Figureigure process.11),), was isolatedisolatedIt partially byby transformedLCLC chromatography. chromatography. into a major Unfortunately, Unfortunately, product (Rt =it it 22was min) not (referred stable during to as 7) the (Figures lyophilization 11 and 12). process. A minor It It partially partiallyproduct wastransformedtransformed also observed into into a a atmajor major Rt = product product22.5 min. (Rt = 22 22 min) min) (referred (referred to to as as 7) 7) (Figures (Figures 1111 andand 12).12). A A minor minor product product waswas also also observed at Rt = 22.5 22.5 min. min.

Figure 11. HPLC trace of isolated 2 (Rt = 21 min) showing one major degradation product (7) and a Figure 11. HPLC trace of isolated 2 (Rt = 21 min) showing one major degradation product (7) and a minorFigure one, 11. HPLC at Rt 22 trace and of 22.5 isolated min, respectively. 2 (Rt = 21 min) showing one major degradation product (7) and a minor one, at Rt 22 and 22.5 min, respectively. minor one, at Rt 22 and 22.5 min, respectively.

Figure 12. In-line UV-visible spectra (220–350 nm) of chromatogram of Figure 11. From left to right: 2 Figure 12. In-line UV-visible spectra (220–350 nm) of chromatogram of Figure 11. From left to right: 2 (Rt = 21 min), degradation product 7 at Rt = 22 min and a minor product at Rt = 22.5 min. (RtFigure = 21 12. min), In-line degradation UV-visible product spectra 7 (220–350at Rt = 22 nm) min of and chroma a minortogram product of Figure at Rt =11. 22.5 From min. left to right: 2 (Rt = 21 min), degradation product 7 at Rt = 22 min and a minor product at Rt = 22.5 min. LC-ESI/MS analysis (high resolution) allowed for us to identify product 2 as an isomer of LC-ESI/MS analysis (high resolution) allowed for us to identify product 2 as an isomer of aldonitrone 1 with a molecular mass of 301 g/mol and the same molecular formula C11H15N3O7. aldonitroneLC-ESI/MS 1 with analysis a molecular (high resolution)mass of 301 allowed g/mol andfor usthe to same identify molecular product formula 2 as an C 11isomerH15N3O of7. Indeed, positive ionization spectrum showed the sodium adduct of the molecular ion [M + Na]+ ion Indeed,aldonitrone positive 1 with ionization a molecular spectrum mass showed of 301 theg/mol sodium and adductthe same of themolecular molecular formula ion [M C 11+ HNa]15N+ 3ionO7. C11H15N3O7Na calcd: 324.0808, obs: m/z = 324.0803 amu (Figure S5). Besides, the molecular mass+ of 7 CIndeed,11H15N 3positiveO7Na calcd: ionization 324.0808, spectrum obs: m/z showed = 324.0803 the sodiumamu (Figure adduct S5). of Besides, the molecular the molecular ion [M + mass Na] ofion 7 was 289 g/mol. It corresponds to a loss of 12 amu as compared to the molecular mass of 2 (Figure S5). wasC11H 28915N 3g/mol.O7Na calcd:It corresponds 324.0808, to obs: a loss m/z of = 12324.0803 amu as amu compared (Figure to S5). the Besides, molecular the mass molecular of 2 (Figure mass ofS5). 7 The observed signal at m/z = 312.0805 amu corresponds to the molecular formula C10H15N3O7Na Thewas observed289 g/mol. signal It corresponds at m/z = 312.0805 to a loss amu of 12 corresponds amu as compared to the molecular to the molecular formula mass C10H 15ofN 23 O(Figure7Na calcd: S5). calcd: 312.0808 and thus to the sodium adduct of the molecular ion [M + Na]+. 312.0808The observed and thus signal to at the m/z sodium = 312.0805 adduct amu of corresponds the molecular to theion molecular[M + Na]+. formula C10H15N3O7Na calcd: 312.0808Compound and thus 2to is the likely sodium to be adduct the terminal of the molecular aldonitrone ion (Scheme[M + Na]5+). and may form by a similar Compound 2 is likely to be the terminal aldonitrone (Scheme5 5) and may form by a similar mechanism with 2 mol. equiv. of KHSO5 with respect to mnm U. The loss of a molecule of water from mechanismCompound with 2 mois likelyl. equiv. to ofbe KHSO the terminal5 with respect aldonitrone to mnm (Scheme5U. The loss5) and of a maymolecule form of by water a similar from the common intermediate hydroxylamine may occur on either side, which will lead to two isomeric themechanism common with intermediate 2 mol. equiv. hydroxylamine of KHSO5 with may respect occur toon mnm either5U. side, The whichloss of willa molecule lead to oftwo water isomeric from imines. A second oxidation step gives rise to either the N-methyl-aldonitrone 1 or the terminal imines.the common A second intermediate oxidation hydroxylamine step gives rise may to occureither on the either N-methyl-aldonitrone side, which will lead 1 orto twothe terminalisomeric aldonitrone 2. aldonitroneimines. A second 2. oxidation step gives rise to either the N-methyl-aldonitrone 1 or the terminal Before any more precise characterization (i.e., NMR), arguments in favor of 2 being the terminal aldonitroneBefore any2. more precise characterization (i.e., NMR), arguments in favor of 2 being the terminal aldonitrone are: (i) the UV-vis spectrum of 2 shows a strong absorption band at 236 nm (Figures2 aldonitroneBefore are:any (i)more the UV-visprecise spectrum characterization of 2 shows (i.e., a strongNMR), absorption arguments band in favor at 236 of nm 2 being (Figures the 2 terminal and 12). and 12). While conjugated aldonitrones (such as 1) show a typical absorption band in the 300 nm Whilealdonitrone conjugated are: (i) thealdonitrones UV-vis spectrum (such ofas 2 1) shows show a stronga typical absorption absorption band band at 236 in nm the (Figures 300 nm 2 andregion, 12). region,non-conjugated aliphatic aldonitrones show a typical band for the aldonitrone function at λmax non-conjugatedWhile conjugated aliphatic aldonitrones aldonitrones (such show as 1) a typicalshow aband typical for theabsorption aldonitrone band function in the at 300 λmax nm c.a. region,236 nm λ [50];non-conjugated (ii) 2 is, not aliphatic surprisingly, aldonitrones less stable show than a typical 1. Its band major for degradation the aldonitrone product function 7, withat max a c.a.molecular 236 nm [50]; (ii) 2 is, not surprisingly, less stable than 1. Its major degradation product 7, with a molecular Biomolecules 2018, 8, 145 13 of 18

Biomolecules 2018, 8, x FOR PEER REVIEW 13 of 18 c.a.Biomolecules 236 nm 2018 [50, 8];, x (ii)FOR 2 PEER is, not REVIEW surprisingly, less stable than 1. Its major degradation product 7,13 with of 18 amass molecular of 289 massg/mol ofmay 289 correspond g/mol may to correspond the hydroxyl toamine the hydroxylamine derivative (Scheme derivative 6); and, (Scheme (iii) terminal6); and, mass of 289 g/mol may correspond to the hydroxylamine derivative (Scheme 6); and, (iii) terminal (iii)aldonitrones terminal aldonitroneshave been previously have been ob previouslyserved in the observed literature in the [28,43]. literature [28,43]. aldonitrones have been previously observed in the literature [28,43].

Scheme 5.5. Mechanism of oxidation of mnmmnm55U. Formation of two different aldonitronealdonitrone derivativesderivatives Scheme 5. Mechanism of oxidation of mnm5U. Formation of two different aldonitrone derivatives (1(1 andand 2).2). (1 and 2).

Scheme 6. Instability of terminal aldonitrone 2 toward hydrolysis (in acidic conditions). Formation of Scheme 6. Instability of terminal aldonitrone 2 toward hydrolysishydrolysis (in(in acidicacidic conditions).conditions). Formation ofof hydroxylamine derivative, referred to as 7 and formaldehyde. Numbering of atoms corresponds to hydroxylamine derivative,derivative, referred referred to to as as 7 and7 and formaldehyde. formaldehyde. Numbering Numbering of atomsof atoms corresponds corresponds to the to the numbering of uridine. numberingthe numbering of uridine. of uridine.

1H-NMR spectrum of isolated 2 (which consists in a mixture of 2 and 7) clearly shows two 1H-NMR spectrum of isolated 2 (which consists in a mixture of 2 and 7) clearlyclearly showsshows twotwo doublets δ 6.70 and 6.52 ppm with a coupling constant of 6.1 Hz for the two non-equivalent protons doublets δδ 6.706.70 andand 6.526.52ppm ppm with with a a coupling coupling constant constant of of 6.1 6.1 Hz Hz for for the the two two non-equivalent non-equivalent protons protons of of the CH2 of the terminal aldonitrone (Figure 13). The methylene CH2, which is directly connected theof the CH CH2 of2 theof the terminal terminal aldonitrone aldonitrone (Figure (Figure 13). The13). methyleneThe methylene CH2, CH which2, which is directly is directly connected connected to the to the C5 of the nucleobase, was evidenced by the HSQC experiment showing that a proton signal C5to the of theC5 nucleobase,of the nucleobase, was evidenced was evidenced by the HSQCby the experimentHSQC experiment showing showing that a proton that a signalproton hidden signal δ 1 hidden under the water signal correlated withδ a carbon 60.5 1ppm through JCH (Figure S6A). underhidden the under water the signal water correlated signal withcorrelated a carbon with60.5 a carbon ppm through δ 60.5 ppmJCH (Figure through S6A). 1JCH Furthermore, (Figure S6A). in Furthermore, in HMBC experiment this proton signal correlated with carbons C4 δ 164.3 ppm, C6 δ HMBCFurthermore, experiment in HMBC this proton experiment signal this correlated proton withsignal carbons correlated C4 δ with164.3 carbons ppm, C6 C4δ 142.7 δ 164.3 ppm, ppm, and C6 C5 δ 142.7 ppm, and C5 δ 105.3 ppm of nucleobase (Figure S6B). Product 2 is a ribonucleoside with the H6 δ142.7105.3 ppm, ppm and of nucleobase C5 δ 105.3 ppm (Figure of nucleobase S6B). Product (Figure 2 is a S6B). ribonucleoside Product 2 withis a ribonucleoside the H6 nucleobase with protonthe H6 nucleobase proton resonance at δ 7.91 ppm, the H1’ at δ 5.61 ppm, and other Hs of sugar between 4.04 resonancenucleobase at protonδ 7.91 resonance ppm, the at H1’ δ 7.91 at δ ppm,5.61 ppm,the H1’ and at otherδ 5.61 Hsppm, of sugarand other between Hs of 4.04sugar and between 3.52 ppm. 4.04 and 3.52 ppm.1 Besides, the 1H-NMR spectrum exhibits two additional resonances that belong to Besides,and 3.52 theppm.H-NMR Besides, spectrum the 1H-NMR exhibits spectrum two additional exhibits resonances two additi thatonal belong resonances to product that 7, thebelong singlet to product 7, the singlet δ 7.66 ppm, and the multiplet δ 3.44 ppm, attributed to the H6 of the nucleobase δproduct7.66 ppm, 7, the and singlet the multiplet δ 7.66 ppm,δ 3.44 and ppm, the multiplet attributed δ 3.44 to the ppm, H6 attributed of the nucleobase to the H6 and of the nucleobase methylene and the methylene CH2 for ribonucleoside 7, respectively. The ratio between 2 and 7 (measured by CHand2 thefor methylene ribonucleoside CH2 7,for respectively. ribonucleoside The 7, ratio respectively. between The 2 and ratio 7 (measured between 2 by and integration 7 (measured of the by integration of the H6 resonances) corresponds to the one observed by UV detection in Figure 11, H6integration resonances) of the corresponds H6 resonances) to the corresponds one observed to by the UV one detection observed in Figureby UV 11detection, about 60%in Figure and 40%, 11, about 60 and 40%, respectively. Other protons of ribonucleoside 7 show the same chemical shifts as respectively.about 60 and Other 40%, protonsrespectively. of ribonucleoside Other protons 7 showof ribo thenucleoside same chemical 7 show shifts the same as those chemical of 2 (Figure shifts 13 as: those of 2 (Figure 13: the H1’ δ 5.61 ppm accounts for 1.6 H). thethose H1’ ofδ 25.61 (Figure ppm 13: accounts the H1’ for δ 5.61 1.6 H).ppm accounts for 1.6 H). Biomolecules 2018, 8, 145 14 of 18 Biomolecules 2018, 8, x FOR PEER REVIEW 14 of 18

11 FigureFigure 13. 13. H-NMRH-NMR (600 (600 MHz, MHz, 278 278 K) K) spectrum spectrum of of isolated isolated 2 2 in in D D22O.O. The The tube tube contains contains a a mixture mixture of of 2 2 (60%)(60%) and and 7 7 (40%). (40%). The The solvent solvent peak peak at at δδ 4.74.7 ppm ppm is is absent. absent.

AsAs deduced deduced from from the the HMBC HMBC experiment, experiment, the the carbon carbon atoms atoms of of ribonucleosides ribonucleosides 2 and and 7 7 are are similar similar exceptexcept the the C5, C5, which which carries carries the the 5- 5-substituent,substituent, and and was was observed observed at at δδ 105.3105.3 ppm ppm and and 108.5 108.5 ppm, ppm, respectively.respectively. A A small small difference difference was also noted between between δ 142.7 ppm for for the the C6 C6 of of 2 2 and and δδ 140.6140.6 ppm ppm forfor the the C6 C6 of of 7 7 (Figure (Figure S6B). S6B). Additionally,Additionally, a NOE correlation exists between one one of of the the protons protons of of the the terminal terminal methylene methylene of of 22 δδ 6.70 ppm and the signal under the solvent peak ( δ 4.5 ppm) attributed to the other CH 2 groupgroup (Figure(Figure S6C). S6C). Not Not surprisingly, the configurationconfiguration ofof aldonitronealdonitrone isis blocked.blocked. AA NOE NOE correlation correlation observed observed between between the the CH CH22 protonsprotons at at C5 C5 of of 7 7 δδ 3.443.44 ppm ppm and and the the H6 H6 of of 7 7 δδ 7.667.66 ppm ppm also also corroborated corroborated the attribution of thes thesee two NMR resonances to product 7 (Figure S6C). Overall,Overall, the NMR data are in accordance withwith thethe proposedproposed structuresstructures for for terminal terminal aldonitrone aldonitrone 2 2and and its its hydroxylamine hydroxylamine hydrolysis hydrolysis product product 7. 7. TheThe stability stability of of 2 2in inthe the dark dark was was tested tested in water in water under under neutral neutral conditions. conditions. Ribonucleoside Ribonucleoside 2 was ◦ incubated2 was incubated at 50 °C at in 50 waterC in for water 4 h. It for proved 4 h. Itstab provedle (not stable shown). (not Therefore, shown). the Therefore, observed the hydrolysis observed (tohydrolysis hydroxylamine (to hydroxylamine 7) during 7)the during lyophilization the lyophilization process was process attributed was attributed to some to transient some transient acidic conditionsacidic conditions due to the due concentration to the concentration of the HPLC of thebuffer HPLC (ammonium buffer (ammonium acetate) in the acetate) collected in fractions the collected [51]. fractions [51]. 4. Discussion 4. Discussion The secondary amine function of the C5 substituent in 5-methylaminomethyl uridine is sensitive to oxidation.The secondary Granted amine that function5-methylaminomethyl of the C5 substituent modified in 5-methylaminomethyl ribonucleosides are present uridine isat sensitivethe key positionto oxidation. 34 in ASL Granted in tRNAs, that 5-methylaminomethyl the present work investigated modified the ribonucleosides oxidation of mnm are5 presentU, since atoxidative the key 5 degradationposition 34 in of ASL mnm in5 tRNAs,U might the have present biological work investigatedrelevance and the subsequently oxidation of mnmmay disturbU, since biological oxidative 5 processesdegradation involving of mnm tRNAs.U might Indeed, have any biological chemical relevance modification and subsequently of the 5-methylaminomethyl may disturb biological group mayprocesses alter involvingthe recognition tRNAs. and Indeed, binding any processes chemical during modification the translation. of the 5-methylaminomethyl Any chemically reactive group functionmay alter appearing the recognition during oxidation and binding may processes lead to fu duringrther chemical the translation. transformation Any chemically of this substituent reactive orfunction to cross-linking appearing reaction during with oxidation biological may leadpartners. to further chemical transformation of this substituent or toOxidation cross-linking of the reaction mnm5U with ribonucleoside biological partners. in water and at pH 7.5 by oxone gave rise to two oxidation 5 products,Oxidation the N-methyl-aldonitrone of the mnm U ribonucleoside 1 and the terminal in water aldonitrone and at pH 2 7.5(Figure by oxone 1 and gave Scheme rise 5), to The two conversionoxidation products, was 50% theafterN -methyl-aldonitronea 10 min reaction at 1room and thetemperature. terminal aldonitrone According 2to (Figure absorbance1 and at Scheme 260 nm,5). whichThe conversion should be wassimilar 50% for after both acompounds, 10 min reaction the ratio at room N-methyl temperature. aldonitrone/terminal According toaldonitrone absorbance is aboutat 260 2/1 nm, (Figure which should1). The beN-methyl-aldonitrone similar for both compounds, 1 is thermodyna the ratiomicallyN-methyl preferred aldonitrone/terminal to the terminal aldonitrone 2. Biomolecules 2018, 8, 145 15 of 18 aldonitrone is about 2/1 (Figure1). The N-methyl-aldonitrone 1 is thermodynamically preferred to the terminal aldonitrone 2. Both compounds arise from two successive oxidation reactions and one dehydration process at the secondary amine function of the C5 substituent of uracil base (Scheme5). The proposed intermediate products, hydroxylamine and imine, were not observed under the experimental conditions used (excess of KHSO5 or 4 mol. equiv. KHSO5). Thus, after the first oxidation step, the reaction goes quickly. Both aldonitrones, 1 and 2, proved stable at various tested temperatures in pH 7.5 aqueous buffer. However, in biological medium, nitrones might react with nucleophiles and/or coordinate transition metal ions leading to cross-links of modified tRNA with other biomolecules [40]. Furthermore, nitrones have drawn special attention due to their spin-trapping properties [52,53]. They can react with transient free radical to form more persistent radical paramagnetic species. The intermediate derivatives (Scheme5), hydroxylamine and imine, are potentially reactive species in physiological conditions. Hydroxylamines are strong nucleophiles while imines are . Various alkylhydroxylamines have been reported to be mutagenic in several bacterial systems [54]. In addition to the direct oxidation products of mnm5U mentioned above, other possible oxidation-derived products might form in vivo, such as the oxaziridine derivatives (for instance 3 and 4) (Scheme3) observed upon light irradiation of aldonitrone 1 (Figure5). Irradiation of 2 was not studied but should also lead to the corresponding oxaziridines since the light-dependent isomerization of aldonitrones is rather classic [37,41,46–48]. As cyclic constrained compounds oxaziridines are less stable than aldonitrones. The half-life of 3 and 4 in water in the dark and at 37 ◦C was about 3 h (not shown). They spontaneously reverse to aldonitrone, or transform to amide (isomerization) and/or aldehyde (hydrolysis) derivatives (Scheme4). The aldehyde derivative (5) may be a chemically reactive lesion in vivo. The known reactivity of oxaziridine function in oxygen atom transfer and nitrogen atom transfer reactions should not apply to the present derivatives, since they are not “activated” with electron withdrawing groups [35,37,55,56]. However, we did not test the reactivity of 3 and 4 with amines or sulfides. Therefore, the present work points out that the oxidation of 5-methylaminomethyl uridine (mnm5U) may be possible in vivo under oxidative stress conditions with subsequent potential cellular dysfunction related to altered tRNAs. The main products of oxidation are two aldonitrone derivatives (1 and 2). If the oxidation to aldonitrone is not complete, hydroxylamine and imine intermediate products (Scheme5) may react with biological molecules. Furthermore, another type of lesion may arise from aldonitrones, like oxaziridines (3 and 4 from 1) (light promoted isomerization), which in turn may transform to amide or aldehyde products. All of these RNA lesions will probably alter tRNA binding and most of them may prove chemically reactive.

5. Conclusions The mnm5U ribonucleoside is oxidized by oxone to two oxidation products (1 and 2), the conversion was 50% in pH 7.5 aqueous buffer and room temperature within 10 min. The N-methyl- aldonitrone 1 and the terminal aldonitrone 2 are two aldonitrones isomers. Both arise from two successive oxidation reactions (and one dehydration process) at the secondary amine function of the C5 substituent of the uracil base. Further evidence for the structure of 1 was obtained by light irradiation experiments showing its conversion to oxaziridine isomers 3 and 4. Both aldonitrones 1 and 2 are relatively stable, but oxaziridines 3 and 4 are not. Oxidation of the initial secondary amine function of mnm5U is thus possible in vivo and it may lead to lesions, such as aldonitrone, hydroxylamine, imine, oxaziridine, amide, and aldehyde derivatives. None of them looks neutral regarding tRNA in biology. Biomolecules 2018, 8, 145 16 of 18

Supplementary Materials: The following are available online at http://www.mdpi.com/2218-273X/8/4/145/s1, Figure S1: HR-ESI-MS analysis of 1, Figure S2: NMR analysis of 1: (A) HMBC, (B) NOESY, Figure S3: HR-ESI-MS analysis of 3 and 4, Figure S4: HR-MS analysis of 5 and 6, Figure S5: HR-MS analysis of 2 and its hydrolysis product 7, Figure S6: NMR analysis of 2 + 7: (A) HSQC, (B) HMBC, (C) NOESY. Author Contributions: G.P. conceived and designed the experiments; Q.Z. and B.T.V.N. performed the experiments; Q.Z. and G.P. analyzed the data; G.L. synthesized 5-methylaminomethyl uridine; J.-L.S. performed modeling; G.P. wrote the paper. Funding: This research was funded by Electricité de France (EDF) and Centre National de la Recherche Scientifique (CNRS). Acknowledgments: Christian Bijani and Yannick Coppel from the NMR facility of LCC, CNRS are acknowledged. ESI-MS data were obtained from the Service de Spectrométrie de Masse, Institut de Chimie de Toulouse, ICT–FR 2599, Toulouse, France. Conflicts of Interest: The authors declare no conflict of interest.

References

1. El Yacoubi, B.; Bailly, M.; de Crecy-Lagard, V. Biosynthesis and function of posttranscriptional modifications of transfer RNAs. Annu. Rev. Genet. 2012, 46, 69–95. [CrossRef][PubMed] 2. Agris, P.F.; Eruysal, E.R.; Narendran, A.; Vare, V.Y.P.; Vangaveti, S.; Ranganathan, S.V. Celebrating wobble decoding: Half a century and still much is new. RNA Biol. 2018, 15, 537–553. [CrossRef][PubMed] 3. Duechler, M.; Leszczynska, G.; Sochacka, E.; Nawrot, B. Nucleoside modifications in the regulation of gene expression: focus on tRNA. Cell. Mol. Life Sci. 2016, 73, 3075–3095. [CrossRef][PubMed] 4. Agris, P.F.; Narendran, A.; Sarachan, K.; Väre, V.Y.P.; Eruysal, E. The Importance of Being Modified: The Role of RNA Modifications in Translational Fidelity. Enzymes 2017, 41, 1–50. [PubMed] 5. Motorin, Y.; Helm, M. RNA nucleotide methylation. Wiley Interdiscip. Rev. RNA 2011, 2, 611–631. [CrossRef] [PubMed] 6. Carell, T.; Brandmayr, C.; Hienzsch, A.; Muller, M.; Pearson, D.; Reiter, V.; Thoma, I.; Thumbs, P.; Wagner, M. Structure and function of noncanonical nucleobases. Angew. Chem. Int. Ed. Engl. 2012, 51, 7110–7131. [CrossRef][PubMed] 7. Väre, V.Y.; Eruysal, E.R.; Narendran, A.; Sarachan, K.L.; Agris, P.F. Chemical and Conformational Diversity of Modified Nucleosides Affects tRNA Structure and Function. Biomolecules 2017, 7, 29. [CrossRef][PubMed] 8. Murphy, F.V., 4th; Ramakrishnan, V.; Malkiewicz, A.; Agris, P.F. The role of modifications in codon discrimination by tRNA(Lys)UUU. Nat. Struct. Mol. Biol. 2004, 11, 1186–1191. [CrossRef][PubMed] 9. Sakamoto, K.; Kawai, G.; Niimi, T.; Satoh, T.; Sekine, M.; Yamaizumi, Z.; Nishimura, S.; Miyazawa, T.; Yokoyama, S. A Modified Uridine in the 1st Position of the Anticodon of a Minor Species of Arginine Transfer-Rna, the ArgU Gene-Product, from Escherichia-Coli. Eur. J. Biochem. 1993, 216, 369–375. [CrossRef] [PubMed] 10. Rozov, A.; Demeshkina, N.; Khusainov, I.; Westhof, E.; Yusupov, M.; Yusupova, G. Novel base-pairing interactions at the tRNA wobble position crucial for accurate reading of the genetic code. Nat. Commun. 2016, 7, 10457. [CrossRef][PubMed] 11. Wittwer, A.J.; Tsai, L.; Ching, W.M.; Stadtman, T.C. Identification and Synthesis of a Naturally-Occurring Selenonucleoside in Bacterial Transfer-RNAs: 5-[(Methylamino)Methyl]-2-Selenouridine. Biochemistry 1984, 23, 4650–4655. [CrossRef][PubMed] 12. Poulsen, H.E.; Specht, E.; Broedbaek, K.; Henriksen, T.; Ellervik, C.; Mandrup-Poulsen, T.; Tonnesen, M.; Nielsen, P.E.; Andersen, H.U.; Weimann, A. RNA modifications by oxidation: A novel disease mechanism? Free Radic. Biol. Med. 2012, 52, 1353–1361. [CrossRef][PubMed] 13. Wurtmann, E.J.; Wolin, S.L. RNA under attack: cellular handling of RNA damage. Crit. Rev. Biochem. Mol. Biol. 2009, 44, 34–49. [CrossRef][PubMed] 14. Nawrot, B.; Sochacka, E.; Duchler, M. tRNA structural and functional changes induced by oxidative stress. Cell. Mol. Life Sci. 2011, 68, 4023–4032. [CrossRef][PubMed] 15. Tomaszewska-Antczak, A.; Guga, P.; Nawrot, B.; Pratviel, G. Guanosine in a Single Stranded Region of Anticodon Stem-Loop tRNA Models is Prone to Oxidatively Generated Damage Resulting in Dehydroguanidinohydantoin and Spiroiminodihydantoin Lesions. Chemistry 2015, 21, 6381–6385. [CrossRef] [PubMed] Biomolecules 2018, 8, 145 17 of 18

16. Alenko, A.; Fleming, A.M.; Burrows, C.J. Reverse Transcription Past Products of Guanine Oxidation in RNA Leads to Insertion of A and C opposite 8-Oxo-7,8-dihydroguanine and A and G opposite 5-Guanidinohydantoin and Spiroiminodihydantoin Diastereomers. Biochemistry 2017, 56, 5053–5064. [CrossRef][PubMed] 17. Sochacka, E.; Kraszewska, K.; Sochacki, M.; Sobczak, M.; Janicka, M.; Nawrot, B. The 2-thiouridine unit in the RNA strand is desulfured predominantly to 4-pyrimidinone nucleoside under in vitro oxidative stress conditions. Chem. Commun. (Camb.) 2011, 47, 4914–4916. [CrossRef][PubMed] 18. Bartos, P.; Ebenryter-Olbinska, K.; Sochacka, E.; Nawrot, B. The influence of the C5 substituent on the 2-thiouridine desulfuration pathway and the conformational analysis of the resulting 4-pyrimidinone products. Bioorg. Med. Chem. 2015, 23, 5587–5594. [CrossRef][PubMed] 19. Huang, W.; Lan, M.D.; Qi, C.B.; Zheng, S.J.; Wei, S.Z.; Yuan, B.F.; Feng, Y.Q. Formation and determination of the oxidation products of 5-methylcytosine in RNA. Chem. Sci. 2016, 7, 5495–5502. [CrossRef][PubMed] 20. Payne, N.C.; Geissler, A.; Button, A.; Sasuclark, A.R.; Schroll, A.L.; Ruggles, E.L.; Gladyshev, V.N.; Hondal, R.J. Comparison of the redox chemistry of sulfur- and selenium-containing analogs of uracil. Free Radic. Biol. Med. 2017, 104, 249–261. [CrossRef][PubMed] 21. Sierant, M.; Kulik, K.; Sochacka, E.; Szewczyk, R.; Sobczak, M.; Nawrot, B. Cytochrome c Catalyzes the Hydrogen Peroxide-Assisted Oxidative Desulfuration of 2-Thiouridines in Transfer RNAs. Chembiochem 2018, 19, 687–695. [CrossRef][PubMed] 22. Bartosik, K.; Leszczynska, G. Synthesis of various substituted 5-methyluridines (xm5U) and 2-thiouridines (xm5s2U) via nucleophilic substitution of 5-pivaloyloxymethyluridine/2-thiouridine. Tetrahedron Lett. 2015, 56, 6593–6597. [CrossRef] 23. Frisch, M.J.; Trucks, G.W.; Schlegel, H.B.; Scuseria, G.E.; Robb, M.A.; Cheeseman, J.R.; Scalmani, G.; Barone, V.; Mennucci, B.; Petersson, G.A.; et al. Gaussian 09, Revision D.01; Gaussian, Inc.: Wallingford, CT, USA, 2013. 24. Tomasi, J.; Mennucci, B.; Cammi, R. Quantum mechanical continuum solvation models. Chem. Rev. 2005, 105, 2999–3093. [CrossRef][PubMed] 25. Spence, G.G.; Taylor, E.C.; Buchardt, O. The photochemical reactions of azoxy compounds, nitrones, and aromatic amine N-oxides. Chem. Rev. 1970, 70, 231–238. [CrossRef] 26. Splitter, J.S.; Calvin, M. Preparation of Oxaziranes by Irradiation of Nitrones. J. Org. Chem. 1958, 23, 651–652. [CrossRef] 27. Cicchi, S.; Corsi, M.; Goti, A. Inexpensive and environmentally friendly oxidation of hydroxylamines to nitrones with bleach. J. Org. Chem. 1999, 64, 7243–7245. [CrossRef] 28. Ali, S.A.; Hashmi, S.M.A.; Siddiqui, M.N.; Wazeer, M.I.M. Regiochemistry of mercury(II) oxide oxidation of unsymmetrical N,N-disubstituted hydroxylamines. Tetrahedron 1996, 52, 14917–14928. [CrossRef] 29. Soldaini, G.; Cardona, F.; Goti, A. Catalytic oxidation of imines based on methyltrioxorhenium/ hydrogen peroxide: a mild and easy chemo- and regioselective entry to nitrones. Org. Lett. 2007, 9, 473–476. [CrossRef][PubMed] 30. Colladon, M.; Scarso, A.; Strukul, G. Mild catalytic oxidation of secondary and tertiary amines to nitrones and N-oxides with H2O2 mediated by Pt(II) catalysts. Green Chem. 2008, 10, 793–798. [CrossRef] 31. Forcato, M.; Nugent, W.A.; Licini, G. A ‘waterproof’ catalyst for the oxidation of secondary amines to nitrones with alkyl hydroperoxides. Tetrahedron Lett. 2003, 44, 49–52. [CrossRef] 32. Murahashi, S.I.; Mitsui, H.; Shiota, T.; Tsuda, T.; Watanabe, S. Tungstate-Catalyzed Oxidation of Secondary-Amines to Nitrones-Alpha-Substitution of Secondary-Amines via Nitrones. J. Org. Chem. 1990, 55, 1736–1744. [CrossRef] 33. Ballistreri, F.P.; Chiacchio, U.; Rescifina, A.; Tomaselli, G.A.; Toscano, R.M. One-Flask Transformation of Secondary-Amines to Nitrones by Oxidation with Hydrogen-Peroxide Mediated by Triscetylpyridinium Tetrakis Oxodiperoxotungsto-Phosphate (Pcwp)-Some Mechanistic Considerations. Tetrahedron 1992, 48, 8677–8684. [CrossRef] 34. Saini, P.; Banerjee, M.; Chattopadhyay, A. Computational Investigation of the Photochemical Reaction Path of Some Synthesized and Experimentally Analyzed Small-Chain Conjugated Nitrones. J. Phys. Chem. A 2016, 120, 396–406. [CrossRef][PubMed] 35. Williamson, K.S.; Michaelis, D.J.; Yoon, T.P. Advances in the chemistry of oxaziridines. Chem. Rev. 2014, 114, 8016–8036. [CrossRef][PubMed] Biomolecules 2018, 8, 145 18 of 18

36. Boyd, D.R.; Coulter, P.B.; Mcguckin, M.R.; Sharma, N.D.; Jennings, W.B.; Wilson, V.E. Imines and Derivatives. Part 24. Nitrone Synthesis by Imine Oxidation Using Either a Peroxyacid or . J. Chem. Soc. Perkin Trans. 1 1990, 1, 301–306. [CrossRef] 37. Davis, F.A.; Sheppard, A.C. Applications of Oxaziridines in Organic-Synthesis. Tetrahedron 1989, 45, 5703–5742. [CrossRef] 38. Gella, C.; Ferrer, E.; Alibes, R.; Busque, F.; de March, P.; Figueredo, M.; Font, J. A metal-free general procedure for oxidation of secondary amines to nitrones. J. Org. Chem. 2009, 74, 6365–6367. [CrossRef][PubMed] 39. Hood, T.S.; Huehls, C.B.; Yang, J. A modular approach to alpha,beta-unsaturated N-aryl ketonitrones. Tetrahedron Lett. 2012, 53, 4679–4682. [CrossRef] 40. Gober, C.M.; Joullie, M.M. From Roquefortine C to Roquefortine L: Formation of a Complex Nitrone with Simple Oxidizing Agents. Isr. J. Chem. 2017, 57, 303–308. [CrossRef] 41. Splitter, J.S.; Calvin, M. Oxaziridines. I. The irradiation products of several nitrones. J. Org. Chem. 1965, 30, 3427–3436. [CrossRef] 42. Kubota, T.; Yamakawa, M.; Mori, Y. The electronic spectra of nitrones and the solvent effect on them. Bull. Soc. Chim. Jpn. 1963, 36, 1552–1563. [CrossRef] 43. Prakash, P.; Gravel, E.; Nguyen, D.V.; Namboothiri, I.N.N.; Doris, E. Direct and Co-catalytic Oxidation of Hydroxylamines to Nitrones Promoted by Rhodium Nanoparticles Supported on Carbon Nanotubes. ChemCatChem 2017, 9, 2091–2094. [CrossRef] 44. Sivasubramanian, S.; Mohan, P.; Thirumalaikumar, M.; Muthusubramanian, S. Synthesis and Separation of the E-Isomer and Z-Isomer of Simple Aldonitrones. J. Chem. Soc., Perkin Trans. 1 1994, 23, 3353–3354. [CrossRef] 45. Jerina, D.M.; Boyd, D.R.; Paolillo, L.; Becker, E.D. Stereospecific chemical shifts and coupling constants in 15N-oxaziridines. Tetrahedron Lett. 1970, 11, 1483–1484. [CrossRef] 46. Saini, P.; Chattopadhyay, A. Spectroscopic features of the low-lying singlet states of some N-alkyl retinylnitrone model systems and their involvement in oxaziridine formation. RSC Adv. 2014, 4, 20466–20478. [CrossRef] 47. Bjorgo, J.; Boyd, D.R.; Campbell, R.M.; Neill, D.C. Photoracemization at a Chiral Pyramidal Nitrogen Center. J. Chem. Soc. Chem. Comm. 1976, 5, 162–163. [CrossRef] 48. Splitter, J.S.; Su, T.M.; Ono, H.; Calvin, M. Orbital symmetry control in the nitrone-oxaziridine system. Nitrone photostationary states. J. Am. Chem. Soc. 1971, 93, 4075–4076. [CrossRef] 49. Hamer, J.; Macaluso, A. Nitrones. Chem. Rev. 1964, 64, 473–495. [CrossRef] 50. Koyano, K.; Suzuki, H. The NMR spectra and molecular geometry of nitrones. Tetrahedron Lett. 1968, 9, 1859–1864. [CrossRef] 51. Emmons, W.D. The Preparation and Properties of Oxaziranes. J. Am. Chem. Soc. 1957, 79, 5739–5754. [CrossRef] 52. Perkins, M.J. Spin trapping. Adv. Phys. Org. Chem. 1980, 17, 1–64. 53. Davies, M.J.; Hawkins, C.L. EPR spin trapping of protein radicals. Free Radic. Biol. Med. 2004, 36, 1072–1086. [CrossRef][PubMed] 54. Marfey, P.; Robinson, E. The genetic toxicology of hydroxylamines. Mutat. Res. 1981, 86, 155–191. [CrossRef] 55. Davis, F.A.; Jenkins, R., Jr.; Yocklovich, S.G. 2-arenesulfonyl-3-aryloxaziridine: a new class of aprotic oxidizing agents (oxidation of organic sulfur compounds). Tetrahedron Lett. 1978, 52, 5171–5174. [CrossRef] 56. Lin, S.; Yang, X.; Jia, S.; Weeks, A.M.; Hornsby, M.; Lee, P.S.; Nichiporuk, R.V.; Iavarone, A.T.; Wells, J.A.; Toste, F.D.; et al. Redox-based reagents for chemoselective methionine bioconjugation. Science 2017, 355, 597–602. [CrossRef][PubMed]

© 2018 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).