Radiation- and Photo-Induced Oxidation Pathways of Methionine in Model Peptide Backbone Under Anoxic Conditions

International Journal of Molecular Sciences

Article Radiation- and Photo-Induced Oxidation Pathways of Methionine in Model Peptide Backbone under Anoxic Conditions

Tomasz P˛edzinski 1,2 , Katarzyna Grzyb 2 , Konrad Skotnicki 3, Piotr Filipiak 1,2 , Krzysztof Bobrowski 3,* , Chryssostomos Chatgilialoglu 1,4,* and Bronislaw Marciniak 1,2,*

1 Center for Advanced Technology, Adam Mickiewicz University, Uniwersytetu Poznanskiego 10, 61-614 Poznan, Poland; [email protected] (T.P.); [email protected] (P.F.) 2 Faculty of Chemistry, Adam Mickiewicz University, Uniwersytetu Poznanskiego 8, 61-614 Poznan, Poland; [email protected] 3 Institute of Nuclear Chemistry and Technology, Dorodna 16, 03-195 Warsaw, Poland; [email protected] 4 ISOF, Consiglio Nazionale delle Ricerche, Via P. Gobetti 101, 40129 Bologna, Italy * Correspondence: [email protected] (K.B.); [email protected] (C.C.); [email protected] (B.M.); Tel.: +48-22-504-1336 (K.B.); +48-61-829-1885 (B.M.)

Abstract: Within the reactive oxygen species (ROS) generated by cellular metabolisms, hydroxyl radicals (HO•) play an important role, being the most aggressive towards biomolecules. The reactions of HO• with methionine residues (Met) in peptides and proteins have been intensively studied, but some fundamental aspects remain unsolved. In the present study we examined the biomimetic model made of Ac-Met-OMe, as the simplest model peptide backbone, and of HO• generated by ionizing radiation in aqueous solutions under anoxic conditions. We performed the identification and quantification of transient species by pulse radiolysis and of final products by LC-MS and high-resolution MS/MS after γ-radiolysis. By parallel photochemical experiments, using Citation: P˛edzinski,T.; Grzyb, K.; Skotnicki, K.; Filipiak, P.; Bobrowski, 3-carboxybenzophenone (CB) triplet with the model peptide, we compared the outcomes in terms K.; Chatgilialoglu, C.; Marciniak, B. of short-lived intermediates and stable product identification. The result is a detailed mechanistic • Radiation- and Photo-Induced scheme of Met oxidation by HO , and by CB triplets allowed for assigning transient species to the Oxidation Pathways of Methionine in pathways of products formation. Model Peptide Backbone under Anoxic Conditions. Int. J. Mol. Sci. Keywords: methionine; oxidation; pulse and γ-radiolysis; laser flash and steady-state photolysis; 2021, 22, 4773. https://doi.org/ free radicals; high-resolution MS/MS 10.3390/ijms22094773

Academic Editor: Marcus S. Cooke 1. Introduction Received: 22 March 2021 The oxidation of methionine residues (Met) in peptides and proteins is a crucial Accepted: 27 April 2021 Published: 30 April 2021 reaction in the biological environment [1]. Reactions of both one- and two-electron oxidants with Met have been studied in some details. The reactive oxygen species (ROS) network, •− • Publisher’s Note: initiated from superoxide radical anion (O2 ) and nitric oxide (NO ), regulates numerous MDPI stays neutral − with regard to jurisdictional claims in metabolic processes. The production of two-electron oxidants like H2O2, ONOO or HOCl published maps and institutional affil- involves the reaction with Met residues in a site-specific manner with formation S and iations. R epimers of methionine sulfoxide, Met(O) [1]. Interestingly, the two epimeric forms of sulfoxide are repaired enzymatically by methionine sulfoxide reductase Msr-A and Msr-B, respectively [2,3]. Met residues in proteins are not only preserved against oxidative stress, but these transformations play an important role in cellular signaling processes [3,4]. Hydroxyl radicals (HO•) are the most reactive species within the ROS network and Copyright: © 2021 by the authors. have long been regarded as a major source of cellular damage [5]. The main cellular Licensee MDPI, Basel, Switzerland. • This article is an open access article processes that generate HO are the Fenton reaction of H2O2, the reduction of HOCl or •− distributed under the terms and H2O2 by O2 and the spontaneous decomposition of ONOOH [1]. In cells it is estimated • conditions of the Creative Commons that the diffusion distance of HO is very small due to its high reactivity with various types Attribution (CC BY) license (https:// of biomolecules with rates close to diffusion-controlled [6,7]. Scheme1 shows the two-step • creativecommons.org/licenses/by/ reaction of HO with sulfides like Met to give formally one-electron oxidation, i.e., the 4.0/). formation of sulfuranyl radical followed by heterolytic cleavage [8–10].

Int. J. Mol. Sci. 2021, 22, 4773. https://doi.org/10.3390/ijms22094773 https://www.mdpi.com/journal/ijms Int. J. Mol. Sci. 2021, 22, x FOR PEER REVIEW 2 of 16

by O2•– and the spontaneous decomposition of ONOOH [1]. In cells it is estimated that the diffusion distance of HO• is very small due to its high reactivity with various types of biomolecules with rates close to diffusion-controlled [6,7]. Scheme 1 shows the two-step Int. J. Mol. Sci. 2021, 22, 4773 2 of 16 reaction of HO• with sulfides like Met to give formally one-electron oxidation, i.e., the formation of sulfuranyl radical followed by heterolytic cleavage [8–10].

Scheme 1. Methionine (R = CH2CH(NH3)+COO−−) like thioethers reacts with two-electron oxidants Scheme 1. Methionine (R = CH2CH(NH3) COO ) like thioethers reacts with two-electron oxidants • toto givegive thethe correspondingcorresponding sulfoxide, sulfoxide, while while the the reaction reaction with with HO HO• affords affords otherother products. products.

RadiolysisRadiolysis of of water water provides provides a very a very convenient convenient source source of hydroxyl of hydroxyl radicals radicals HO•. Time-HO•. resolvedTime-resolved kinetic kinetic studies studies by pulse by radiolysis pulse radiolysis have expanded have expanded our mechanistic our mechanistic understanding under- ofstanding radical of reaction radical pathways reaction pathways of Met at of various Met at functionalized various functionalized environment environment [11]. Indeed, [11]. neighboringIndeed, neighboring group participation group participation of one-electron of one oxidation-electron of oxidation Met (MetS of•+ Met) reactivity (MetS•+ within) reac- particulartivity within peptides particular and/or peptides proteins and is of/or great proteins importance, is of great due importance, to the presence due of to a manifoldthe pres- ofence possible of a manifold participating of possible functionalities participating (carboxy, functionalities amine, hydroxy (carboxy, and amine, amide groups)hydroxy [ 12and]. Foramide a long groups) time it[12]. was For believed a long that time the it two-centered,was believed that three the electron two-centered, (2c-3e) bonds three between electron the(2c- oxidized3e) bonds sulfurbetween atom the andoxidized the lone sulfur electron atom and pairs, the locatedlone electron on the pairs, nitrogen located atom on the in thenitrogenN-terminal atom aminoin the groupN-terminal and theamino oxygen group atoms and in the the oxygenC-terminal atoms carboxyl in the C group-terminal are responsiblecarboxyl group for the are stabilization responsible offor MetS the stabilization•+ through a of five-membered MetS•+ through or a six-membered five-membered ring or interactionsix-membered [13– rin16].g Subsequentinteraction [13 studies–16]. Subsequent showed that studies heteroatoms showed present that heteroatoms in the peptide pre- bondsent in can the be peptide also involved bond can in be the also formation involved of in similar the formation transient of species similar with transient 2c-3e species bonds withwith the2c-3e oxidized bonds sulfurwith the atom oxidized [14–18 sulfur]. atom [14–18]. ThereThere areare alsoalso aa fewfew radiationradiation chemicalchemical studiesstudies ofof MetMet inin aqueousaqueous solutions,solutions, followedfollowed • byby productproduct characterizationcharacterization and quantification. quantification. In In the the reaction reaction of of HO HO• withwith free free Met Met in ineither either the the absence absence or the or thepresence presence of oxygen, of oxygen, the attack the attack at sulfur at sulfur atom atom accounts accounts of ~90% of ~90%affording affording 3-methylthiopropionald 3-methylthiopropionaldehyde;ehyde; the formation the formation of small of small amounts amounts of the of corre- the correspondingsponding sulfoxide sulfoxide is due is dueto in to situ in situ formation formation of H of2O H22 ratherO2 rather than than to direct to direct oxidation oxidation by • • byHO HO• [19,20].[19,20 Again,]. Again, studying studying the the reaction reaction of of HO HO• withwith tripeptide tripeptide Gly-Met-GlyGly-Met-Gly providedprovided strongstrong evidenceevidence that the the corre correspondingsponding sulfoxide sulfoxide in in the the tripeptide tripeptide derives derives from from the thein situ in situ formed H O . It is relevant to say that the main product of the tripeptide is an formed H2O2. It2 is2 relevant to say that the main product of the tripeptide is an unsymmet- unsymmetrical disulfide (RCH SSCH ) assigned to the chemistry of MetS•+ with parent rical disulfide (RCH2SSCH3) assigned2 3 to the chemistry of MetS•+ with parent compound, compound,while the use while of ae therobic use ofconditions aerobic conditions highlighted highlighted the formation the formation of other ofproducts other products derived derivedfrom peroxyl from peroxylradicalsradicals of the tripeptide of the tripeptide [17]. [17]. • 1 TheThe presentpresent studystudy focusesfocuses onon thethe reactionreaction ofof HO HO• withwith Ac-Met-OMeAc-Met-OMe ((1),), thethe simplestsimplest modelmodel peptidepeptide backbone.backbone. ThisThis reactionreaction waswas previouslypreviously studiedstudied byby pulsepulse radiolysisradiolysis atat thethe pH range 4–5.7 [14]. Herein we extend the identification and quantification of transient pH range 4–5.7 [14]. Herein we extend the identification and quantification of transient species by pulse radiolysis at pH 7 and of final products of γ-radiolysis by LC-MS and species by pulse radiolysis at pH 7 and of final products of γ-radiolysis by LC-MS and high-resolution MS/MS under anoxic conditions. The purpose of acetylation of the N- high-resolution MS/MS under anoxic conditions. The purpose of acetylation of the N-ter- terminal amino group is to eliminate the fast intramolecular proton transfer from the minal amino group is to eliminate the fast intramolecular proton transfer from the amino amino group to the sulfuranyl moiety, which was suggested earlier as the main decay group to the sulfuranyl moiety, which was suggested earlier as the main decay reaction reaction pathway, while the esterification of the C-terminal carboxyl group eliminates its pathway, while the esterification of the C-terminal carboxyl group eliminates its decar- decarboxylation [8,9,17,21,22]. Therefore, the use of 1 allows to study the reaction of HO• boxylation [8,9,17,21,22]. Therefore, the use of 1 allows to study the reaction of HO• with with Met residue with no contribution of N- and C-terminal functional groups. Met residue with no contribution of N- and C-terminal functional groups. Excited triplet states of benzophenones carboxyl derivatives (3CB) were also shown to be very useful for studying one-electron oxidation reactions of Met-containing molecules of biological significance [21,23,24]. The mechanism of primary photo-induced processes occurring during Met oxidation in aqueous solutions is summarized in Scheme2. Int. J. Mol. Sci. 2021, 22, x FOR PEER REVIEW 3 of 16

Excited triplet states of benzophenones carboxyl derivatives (3CB) were also shown Int. J. Mol. Sci. 2021, 22, 4773 to be very useful for studying one-electron oxidation reactions of Met-containing mole-3 of 16 cules of biological significance [21,23,24]. The mechanism of primary photo-induced processes occurring during Met oxidation in aqueous solutions is summarized in Scheme 2.

SchemeScheme 2. 2. KnownKnown mechanism mechanism of of primary primary photochemical photochemical reactions reactions for for sensitized sensitized photooxidation photooxidation of ofmethionine methionine derivatives derivatives in in aqueous aqueous solution. solution.

TheThe quenching quenching of of 33CBCB by by Met Met derivatives derivatives leads leads to to formation formation of of a a complex complex that that can can decaydecay in in three three primary reactions: charge separationseparation leadingleading toto CBCB•−•– and MetS••++;; H H-atom-atom transfertransfer from from alpha alpha carbon carbon atom atom next next to to sulfur sulfur to to form CBH • andand αSαS•• radicals;radicals; and and back back electelectronron transfer transfer leading leading to to formation formation of of reactants reactants in in their ground states. Recently, Recently, the the reactionreaction of of 3 3-carboxybenzophenone-carboxybenzophenone triplet triplet (3CB) (3CB) with with Ac Ac-Met-NHMe-Met-NHMe has has been been studied studied in somein some details details by some by some of us of[25]. us The [25]. mechanism The mechanism of photooxidation of photooxidation of Met moiety of Met showed moiety toshowed involve to mainly involve H- mainlyatom abstraction H-atom abstraction from the two from α-positions the two nextα-positions to sulfur next (see to Scheme sulfur 2).(see HereinScheme we2 ). included Herein we results included of CB results sensitized of CB photooxidation sensitized photooxidation of Ac-Met-OMe of Ac-Met- (1) in aqueousOMe (1) solutions in aqueous applying solutions flash applying photolysis flash photolysisstudies for studies transient for detection transient detectionand continu- and ouscontinuous photolysis photolysis for products for products identification. identification. ApplicationApplication at at the the same same time time of ofradiation radiation and and photochemical photochemical techniques techniques allowed allowed for thefor first the firsttime time to have to have two twocomplementary complementary conditions conditions in order in order to compare to compare in detail in detail the • 3 oxidationthe oxidation mechanisms mechanisms of Met of derivative Met derivative 1 initiated1 initiated either either by HO by• or HO 3CBor leadingCB leading via short via- livedshort-lived intermediates intermediates to stable to stableproducts. products. 2. Results and Discussion 2. Results and Discussion 2.1. Pulse Radiolysis Studies 2.1. Pulse Radiolysis Studies − • • Pulse irradiation of water leads to the primary reactive species e aq, HO , and H , as – • • shownPulse in Reactionirradiation (1). of The water values leads in to brackets the primary represent reactive the radiation species e chemicalaq, HO , yieldand H (G,) as in shown −in1 Reaction (1). The values in brackets represent the− radiation chemical yield (G) in µmol J . In N2O-saturated solution (~0.02 M of N2O), e aq are efficiently transformed into −1 – µmolHO• radicalsJ . In N2 viaO-saturated Reaction (2)solution (k = 9.1 (~0.02× 109 MM −of1 sN−21O),), affording e aq are efficientlyG(HO•) = transformed 0.56 µmol J− 1into[7]. HO• radicals via Reaction (2) (k = 9.1 × 109 M−1s−1), affording G(HO•) = 0.56 µmol J−1 [7]. − • • H2O → eaq (0.28), HO •(0.28),H •(0.06) (1) HO → e (0.28), HO (0.28), H (0.06) (1) − • • − eaq +eN+2 ON+O H+2 HOO→ →HO HO++N N2++ HOHO ((2)2) • TheThe reaction reaction of of HO HO• withwith compound compound 11 waswas investigated investigated in in N N22OO-saturated-saturated solution solution of of 0.20.2 mM mM 1 at natural pHpH (pH(pH valuevalue of of 7.0 7.0 was was recorded). recorded). Transient Transient absorption absorption spectra spectra in thein therange range 270–700 270–700 nm recordednm recorded in the in time the rangetime range of 200 of ns 200 to 400 ns µtos were400 µs collected were collected in Figure in1. FigureThe 1. transient spectrum obtained 1.1 µs after the electron pulse showed a dominant sharp absorption band with λmax = 340 nm. Based on previous studies on methionine derivatives, this band can be assigned to the HO• adduct of the sulfur atom HOS• (cf. Scheme3 for the structure) [13,14,17]. However, as shown in Figure1, the transient spectrum profile changes with time, indicating the overlap of various transient species. The transient spectrum profile can be resolved into contributions from the following components: • • the sulfuranyl radicals (HOS ), the Cα-centered radicals (αC ), the α-(alkylthio)alkyl radicals (αS(1)• and αS(2)•), the inter-molecular sulfur–sulfur radical cations (SS•+) and intramolecular sulfur-nitrogen three-electron-bonded radicals (SN•) (the structures of these intermediates are highlighted in colored boxes in Scheme3; see also Figures S1 and S2 in Int. J. Mol. Sci. 2021, 22, 4773 4 of 16

Int. J. Mol. Sci. 2021, 22, x FOR PEER REVIEW 4 of 16 Supplementary Materials for their individual spectra). These intermediates were previously identiﬁed during HO•-induced oxidation of 1, but at the pH range 4–5.7 [14].

(a) (b)

• FigureFigure 1. 1. AbsorptionAbsorption spectra spectra following following HO oxidationoxidation of of compound compound 11 (0.2(0.2 mM) mM) in in N N22OO-saturated-saturated aqueous aqueous solutions solutions at at pH pH 7.7. Spectra Spectra were were taken taken after after the the following following time time delays: delays: (a ()a 200) 200 ns ns ( (),• ),400 400 ns ns ( (),•), 500 500 ns ns ( (•),), 1.1 1.1 µsµ s( (•),), 3 3 µsµs ( (•),), 5 5 µsµs ( (•) and and inset:inset: short short-time-time profiles representing growthsgrowths atatλ λ == 290290 nmnm ((●),•), 340340 nm nm ( •(●),), 390 390 nm nm ( •()● and) and 490 490 nm nm (• ();●); (b ()b 10) 10µs µs (• ),( 20), 20µs µs (), 50 µs (), 100 µs (), 200 µs (), 400 µs () and inset: long-time profiles representing decays at λ = 290 nm (●), 340 (•), 50 µs (•), 100 µs (•), 200 µs (•), 400 µs (•) and inset: long-time profiles representing decays at λ = 290 nm (•), 340 nm (•), nm (●), 390 nm (●) and 490 nm (●) (Dose per pulse = 11 Gy; optical path = 1 cm). 390 nm (•) and 490 nm (•) (Dose per pulse = 11 Gy; optical path = 1 cm).

TheThe transient spectra recordedspectrum 1.1,obtained 3 and 1.1 6 µµss afterafter the electron electron pulse pulse showed were resolved a dominant into sharpcontributions absorption from band the with same λmax components = 340 nm. Based (HOS •on, α previousC•, αS•, studiesSS•+ and on methionineSN•). Figure de-2 ● ● rivatives,shows the this spectrum band can at be 6 assignedµs, while to Figures the HO S3 adduct and S4 of report the sulfur the atom spectra HOS at 1.1 (cf. and Scheme 3 µs,

3respectively. for the structure) The sum [13,14,17]. of all componentHowever, as spectra shown with in Figure their respective1, the transient radiation spectrum chemical pro- fileyields changes (G-values) with time, resulted indicating in a good the ﬁt overlap (⊗ symbols of various in Figure transient2, Figures species. S3 and The S4) transient to the spectrumexperimental profile spectra. can be resolved into contributions from the following components: the sulfuranylTable 1radicals reports (HOS the radiation●), the Cα- chemicalcentered radicals yields of (αC each●), the radical α-(alkylthio)alkyl and their percentage radicals (contributionαS(1)● and αS(2) to the●), transient the inter spectrum-molecular obtained. sulfur–sulfur radical cations (SS●+) and intramolecular sulfur-nitrogen three-electron-bonded radicals (SN●) (the structures of these inter- Table 1. The radiation chemicalmediates yields are (G ,highlightedµmol J−1) of radicalsin colored and boxes their percentagein Scheme contribution 3; see also Figure (in parenthesis) S1 and S2 to thein Sup- total yield of radicals presentplementary in the reaction Materials of HO• forwith their compound individual1 at different spectra). times These after intermediates electron pulse atwere pH 7.0previouslya. identified during HO●-induced oxidation of 1, but at the pH range 4–5.7 [14]. • • • • •+ • • Time (µs) HOS The spectraαC recordedαS(1) 1.1,+ 3α andS(2) 6 µs after SSthe electron pulseSN were resolvedTotal into R con- 0.39 0.06 0.01 ● 0.04● ● ●+ 0.03 ● 1.1 tributions from the same components (HOS , αC , αS , SS and SN ). Figure 2 shows0.53 the (73.6%) spectrum(11.3%) at 6 µs, while Figures(1.9%) S3 and S4 repor(7.5%)t the spectra at(5.7%) 1.1 and 3 µs, respectively. 0.16 The sum of0.02 all component spectra0.17 with their respective0.11 radiation0.10 chemical yields (G-val- 3 0.56 (28.6%) ues) resulted(3.6%) in a good fit (⊗(30.3%) symbols in Figures(19.6%) 2, S3 and S4) to(17.9%) the experimental spectra. 0.03 0.02 0.26 0.13 0.13 6 0.57 (5.3%) (3.5%) (45.6%) (22.8%) (22.8%) 1400 )

1 a

 For a procedure of G determination see Section3 and [15]. 1200 cm 1

 1000 The calculated total G-value of 0.53 µmol J−1 for the 1.1 µs spectrum is nearly in mol agreement3 800 with the expected G-value of HO• (0.56 µmol J−1) available for the reaction of 1 dm

1 600

at pH 7 and the concentration (0.2 mM) of 1. This small difference in G-values could be • understood,400 since at this time the reaction of HO radicals with 1 via pathways 2 and 3 mol J mol  ( 200 (Scheme 3) is about to be completed. The spectrum showed a dominant sharp absorption

bandxG 0 at λmax = 340 nm. It is worthy to note that the most abundant radical present at • this time300 is HOS350 ,400 which450 constitutes500 550 more600 than 70% of all radicals (Table1). On the other hand, the total G-value [nm] of 0.56 µmol J−1 for 3 µs spectrum is in excellent agreement with Figure 2. Resolution of the spectral components: HOS● (●), αC● (▲), αS(1)● and αS(2)● (■), SS●+ (), SN● (▼) in the transient absorption spectrum recorded 6 µs (●—experimental; ⊗—fit) after the electron pulse in N2O-saturated aqueous solution containing 0.2 mM 1 at pH 7.

Int. J. Mol. Sci. 2021, 22, 4773 5 of 16

the expected yield of HO• radicals. The spectrum after 6 µs (Figure2) was dominated by two distinct absorption bands with the pronounced maxima at λ ≈ 290 and 490 nm. These bands were assigned to αS• radicals and SS•+ radical cations, respectively. The ﬁrst one was obtained by hydrogen abstraction (path 2 in Scheme3), and both of them by a sequence of reactions involving HOS• radicals and sulfur radical cations (S•+). The spectrum at 6 µs was resolved into contributions from the same components (Table1). The total G-value of all radical present (0.57 µmol J−1) is again in excellent agreement with the expected yield of HO• radicals. It is worthy to note that the most abundant radicals present at 6 µs are αS•, SS•+ and SN• radicals which constitute more than 90% of all radicals. Interestingly, the comparison of the radiation chemical yields of HOS• radicals and the sum of radiation chemical yields of αS•, SS•+ and SN• radicals at 3 µs and 6 µs after the pulse (Table1) suggested that the increase of G(αS• + SS•+ + SN•) occurs at the expense of decrease of G(HOS•). This observation can be rationalized by the involvement of the Int. J. Mol. Sci. 2021, 22, x FOR PEER REVIEW•+ 6 of 16 S on Met moiety in two equilibria (an acid-base and a concentration) and an irreversible deprotonation channel presented in Scheme3.

• SchemeScheme 3. 3.ProposedProposed mechanism mechanism for for the the react reactionion of of HO radicals generatedgenerated by byγ γ-irradiation-irradiation of of N 2NO-saturated2O-saturated aqueous aqueous solutionssolutions containing containing 1.0 1.0 mM mM NN-acetyl-acetyl methionine methionine methyl methyl ester (1)) atat naturalnatural pH. pH.

InsetExcellent in Figure material 1a shows balance kinetic of all radicals, traces recorded identiﬁed at in thefour system wavelengths equal to the(290,G -value340, 390 • andof HO490 nm)radicals that available correspond for the to reaction the maxima with 1 of, proved absorption the presence bands ofof all the radical four transientsmost abun- µ dantas the radicals precursors present of end in the products. irradiated At this system, point, i.e., it has αS to●, beHOS stressed●, SN● that and after SS●+ 6. Theses the totalkinetic G µ −1 µ traces-value look of different, the radicals and began they toreached decrease, their reaching maximum the valuesignal of at0.12 variousmol Jtimesat 400after sthe which is only slightly higher than the G-value of αS• radicals (0.10 µmol J−1) (Figure S5). pulse. We assigned the observed buildup at λ = 340 nm to the formation of HOS● radical. This suggests that on this time domain nearly 80% of all radicals formed in the system Because HOS● is decaying in a significant way during its formation (see inset in Figure were consumed in radical-radical processes leading to the ﬁnal products. Moreover, in 1a), one has to account for this decay in order to get the proper value for the rate constant of the formation. Furthermore, based on the reference spectra applied in spectral resolutions (Figure S2), one can expect that the optical absorption bands of at least two radicals (αC● and SN●) overlap with the optical absorption band of HOS●, and thus may “contaminate” formation and decay traces observed at λ = 340 nm. In order to overcome this problem, we extracted the concentration profile of HOS● using spectral resolution of spectra at any desired time delay following the electron pulse ranged from 200 ns to 8 µs (cf. Figure S6A). Similarly, on the basis of the extracted concentration profiles of the other transients (SS●+, αS●, αC● and SN●), it was possible to evaluate their kinetic parameters (cf. Figure S6B–E). Otherwise, this would not be possible based just on “raw” time profiles recorded at the wavelengths of their absorption maxima (see insets in Figure 1). Kinetic parameters of all radical transients are collected in Table 2. The first-order decay of HOS● (kd = 5.6 × 105 s−1) results in radicals with a sulfur radical cationic site (S●+). The S●+ undergo typical reactions expected for such kind of radicals: deprotonation leading to the αS● radicals, the intermolecular formation of SS●+ and intramolecular five-membered cyclic SN● (Scheme 3).

Int. J. Mol. Sci. 2021, 22, x FOR PEER REVIEW 4 of 16

(a) (b)

Figure 1. Absorption spectra following HO• oxidation of compound 1 (0.2 mM) in N2O-saturated aqueous solutions at pH 7. Spectra were taken after the following time delays: (a) 200 ns (), 400 ns (), 500 ns (), 1.1 µs (), 3 µs (), 5 µs () and inset: short-time proﬁles representing growths at λ = 290 nm (●), 340 nm (●), 390 nm (●) and 490 nm (●); (b) 10 µs (), 20 µs (), 50 µs (), 100 µs (), 200 µs (), 400 µs () and inset: long-time proﬁles representing decays at λ = 290 nm (●), 340 nm (●), 390 nm (●) and 490 nm (●) (Dose per pulse = 11 Gy; optical path = 1 cm).

The transient spectrum obtained 1.1 µs after the electron pulse showed a dominant sharp absorption band with λmax = 340 nm. Based on previous studies on methionine derivatives, this band can be assigned to the HO● adduct of the sulfur atom HOS● (cf. Scheme 3 for the structure) [13,14,17]. However, as shown in Figure 1, the transient spectrum profile changes with time, indicating the overlap of various transient species. The transient spectrum profile can be resolved into contributions from the following components: the sulfuranyl radicals (HOS●), the Cα-centered radicals (αC●), the α-(alkylthio)alkyl radicals (αS(1)● and αS(2)●), the inter-molecular sulfur–sulfur radical cations (SS●+) and intramolecular sulfur-nitrogen three-electron-bonded radicals (SN●) (the structures of these intermediates are highlighted in colored boxes in Scheme 3; see also Figure S1 and S2 in Sup- plementary Materials for their individual spectra). These intermediates were previously identified during HO●-induced oxidation of 1, but at the pH range 4–5.7 [14]. The spectra recorded 1.1, 3 and 6 µs after the electron pulse were resolved into con- ● ● ● ●+ ● Int. J. Mol. Sci. 2021, 22, 4773 tributions from the same components (HOS , αC 6, ofαS 16 , SS and SN ). Figure 2 shows the spectrum at 6 µs, while Figures S3 and S4 report the spectra at 1.1 and 3 µs, respectively.

Thethis time sum domain of the all rates component of these termination spectra processes with became their dominant respective compared radiation chemical yields (G-values)to the ratesresulted of transformation in a good of SS• +fitand (⊗SN• symbolsradicals into αinS(1) Figures•, αS(2)• and 2,α S3C• and S4) to the experimental spectra. radicals, respectively.

1400 ) 1  1200 cm 1

 1000 mol

3 800 dm

1 600 

400 mol J mol  ( 200 

G xG 0 300 350 400 450 500 550 600  [nm] • • • • • Figure 2. Resolution of the spectral components: HOS (•), αC (N), αS(1) and αS(2) (), SS + • ● ● ● ● ●+ F(Figure), SN (H 2.) in Resolution the transient absorption of the spectrum spectral recorded 6components:µs (•—experimental; HOS⊗—ﬁt) after (● the), αC (▲), αS(1) and αS(2) (■), SS (), SNelectron● (▼ pulse) in N the2O-saturated transient aqueous absorption solution containing spectrum 0.2 mM 1 at pH recorded 7. 6 µs (●—experimental; ⊗—fit) after the electronInset in pulse Figure1a showsin N kinetic2O-saturated traces recorded aqueous at four wavelengths solution (290, containing 340, 390 and 0.2 mM 1 at pH 7. 490 nm) that correspond to the maxima of absorption bands of the four most abundant radicals present in the irradiated system, i.e., αS•, HOS•, SN• and SS•+. These kinetic traces look different, and they reached their maximum signal at various times after the pulse. We assigned the observed buildup at λ = 340 nm to the formation of HOS• radical. Because HOS• is decaying in a signiﬁcant way during its formation (see inset in Figure1a), one has to account for this decay in order to get the proper value for the rate constant of Int. J. Mol. Sci. 2021, 22, x FOR PEER REVIEWthe formation. Furthermore, based on the reference spectra applied in spectral resolutions7 of 16

(Figure S2), one can expect that the optical absorption bands of at least two radicals (αC• and SN•) overlap with the optical absorption band of HOS•, and thus may “contaminate” Tableformation 2. Kinetic and data decay for growth traces and observed decay of at raλdicals= 340 present nm. Inafter order the reaction to overcome of HO● this with problem, com- • poundwe extracted1 at pH 7.0 the from concentration the pulse radiolysis profile studies. of HOS using spectral resolution of spectra at any desired time delay following the electron pulse ranged from 200 ns to 8 µs (cf. Figure S6A). Similarly,k, s−1 on the basisHOS of● the extractedαC● concentrationαS● profiles of theSS● other+ transientsSN● (SS•+, αS•, αC• and SN2.1•), × it 10 was6 possible1.5 × to 10 evaluate6 4.1 their × 10 kinetic5 b parameters4.7 × 105 (cf. Figure S6B–E). kgrowth 3.7 × 105 Otherwise, this would1.1 × 10 not10 a be possible7.5 × 109 based a just3.6 on× 10 “raw”4 c time2.2 × profiles 109 d recorded at the wavelengthskdecay of their 5.6 × absorption 105 1.4maxima × 106 (see insets 4.0 × in10 Figure3 1). 3.6 Kinetic × 104 parameters 8.2 × 10 of3 all a Τheradical second-order transients rate are constants collected (M− in1s−1 Table) measured2. based on 0.2 mM concentration of 1; b the first- • 5 −1 order rateThe constant first-order of the decayfast growth; of HOS c the first-order(kd = 5.6 × rate10 constants ) results of the slow in radicals growth; with d the asec- sulfur ond-orderradical rate cationic constant site (M (S−•1s+−).1) measured The S•+ undergobased on 0.2 typical mM concentration reactions expected of 1, andfor corrected such kindfor of theradicals: backward deprotonation reaction of the equilibrium, leading to the cf. SchemeαS• radicals, 3. the intermolecular formation of SS•+ and intramolecular five-membered cyclic SN• (Scheme3). •●+ •●+ 9 9 −1 −1−1−1 4 4 −1 ForFor the the equilibrium equilibrium S ++ 11 ⇆ SS ,, the valuesvalues of ofkf k=f 2.2= 2.2× 10× 10M Ms s and andk rk=r = 3.6 3.6× ×10 10s −1 4 4 −1 −1 s givegive thethe equilibriumequilibrium constant constantK K= =k kff//kkrr == 6.1 6.1 ×× 1010 MM, which, which is three-fold is three-fold lower lower than than the the previouslypreviously estimated estimated K forK for analogous analogous dimeric dimeric radical radical cations cations derived derived from from (CH (CH3)23S) 2[26].S[26 ]. The formation of αS● radicals occur via two different mono-exponential processes with k = 4.1 × 105 s−1 and k = 3.6 × 104 s−1 (cf. Figure S6C). The first one is assigned to deprotonation of S●+ that is formed directly from HOS●, and the second to deprotonation of S●+ that is formed via the reverse reaction involving SS●+. These assignments are strongly supported by the fact that the rate constant of the slow formation of αS● radicals (cf. Figure S6C) is equal to the rate constant of the SS●+ decay (cf. Figure S6B). A second-order rate constant of 7.5 × 109 M−1s−1 was obtained for the reaction of HO● radical with 1, via path 3 (Scheme 3), by measuring the rate constant of the pseudo-first- order growth k = 1.5 × 106 s−1 of the αC● concentration at 0.2 mM of 1 (cf. Figure S6D), which was assigned to the direct H-atom abstraction from the α-carbon atom by HO● radicals (see path 3 in Scheme 3). This value is reasonable considering in this case a rather low Cα-H bond energy [27]. A quite surprising result is the short lifetime of αC● radicals which decays via a mono-exponential process with the rate constant k = 1.4 × 106 s−1 (cf. Figure S6D). We tentatively suggest that by-products are formed by unimolecular decay due to β-fragmentation with formation of the acyl radical, CH3C(O)• [28]. The pseudo-first order growth k = 3.7 × 105 s−1 of the SN● concentration can be assigned to the overall reaction with the first step leading to S●+, followed by the concerted cyclization and deprotonation (see Scheme 3). The decay of the SN● radicals is rather slow and occurs with the pseudo-first order rate constant k = 8.2 × 103 s−1. Based on earlier results [14,15], protonation of the SN● radicals provides the most facile mechanistic reaction pathway of their decay. On the basis of the extracted concentration profiles of the SN● radicals in cyclic L-Met-L-Met at various pHs, it was possible to evaluate the rate constant of SN● radicals with protons to be 2.1 × 109 M−1s−1 [15]. At pH 7, these reactions occur with the pseudo-first order rate constant 2.1 × 102 s−1, which are much lower than the pseudo-first order rate constant measured for the decay of the SN● radicals formed from compound 1 (vide supra). Therefore, the most probable pathway responsible for the decay of the SN● involves N-protonation (e.g., by water molecules), followed by deprotonation at the αC- carbon leading to αC● radicals (cf. Scheme 3). These reactions provide an irreversible entry to these C-centered radicals and were previously suggested for linear peptides containing Met residue [14]. Since αC● radicals are formed in a slow process, this fact, combined with their very short lifetime, rationalizes their absence in the resolved absorption spectra recorded at longer times.

2.2. γ-Radiolysis and Product Analysis

In addition to the reactive species e–aq, HO•, and H•, radiolysis of neutral water leads also to H+ (0.28) and H2O2 (0.07); in parenthesis the G in µmol J−1 [7]. In N2O-saturated solution, the G(HO•) = 0.56 µmol J−1, therefore HO• and H• account for 90% and 10%, respectively, of the reactive species (cf. Reactions (1) and (2)).

Int. J. Mol. Sci. 2021, 22, 4773 7 of 16

Table 2. Kinetic data for growth and decay of radicals present after the reaction of HO• with compound 1 at pH 7.0 from the pulse radiolysis studies.

k, s−1 HOS• αC• αS• SS•+ SN• 2.1 × 106 1.5 × 106 4.1 × 105 b 4.7 × 105 k 3.7 × 105 growth 1.1 × 1010 a 7.5 × 109 a 3.6 × 104 c 2.2 × 109 d 5 6 3 4 3 kdecay 5.6 × 10 1.4 × 10 4.0 × 10 3.6 × 10 8.2 × 10 a The second-order rate constants (M−1s−1) measured based on 0.2 mM concentration of 1; b the ﬁrst-order rate constant of the fast growth; c the ﬁrst-order rate constant of the slow growth; d the second-order rate constant (M−1s−1) measured based on 0.2 mM concentration of 1, and corrected for the backward reaction of the equilibrium, cf. Scheme3.

The formation of αS• radicals occur via two different mono-exponential processes with k = 4.1 × 105 s−1 and k = 3.6 × 104 s−1 (cf. Figure S6C). The first one is assigned to deprotonation of S•+ that is formed directly from HOS•, and the second to deprotonation of S•+ that is formed via the reverse reaction involving SS•+. These assignments are strongly supported by the fact that the rate constant of the slow formation of αS• radicals (cf. Figure S6C) is equal to the rate constant of the SS•+ decay (cf. Figure S6B). A second-order rate constant of 7.5 × 109 M−1s−1 was obtained for the reaction of HO• radical with 1, via path 3 (Scheme3), by measuring the rate constant of the pseudo-first- order growth k = 1.5 × 106 s−1 of the αC• concentration at 0.2 mM of 1 (cf. Figure S6D), which was assigned to the direct H-atom abstraction from the α-carbon atom by HO• radicals (see path 3 in Scheme3). This value is reasonable considering in this case a rather low Cα-H bond energy [27]. A quite surprising result is the short lifetime of αC• radicals which decays via a mono-exponential process with the rate constant k = 1.4 × 106 s−1 (cf. Figure S6D). We tentatively suggest that by-products are formed by unimolecular • decay due to β-fragmentation with formation of the acyl radical, CH3C(O) [28]. The pseudo-first order growth k = 3.7 × 105 s−1 of the SN• concentration can be assigned to the overall reaction with the first step leading to S•+, followed by the concerted cyclization and deprotonation (see Scheme3). The decay of the SN • radicals is rather slow and occurs with the pseudo-first order rate constant k = 8.2 × 103 s−1. Based on earlier results [14,15], protonation of the SN• radicals provides the most facile mechanistic reaction pathway of their decay. On the basis of the extracted concentration profiles of the SN• radicals in cyclic L-Met-L-Met at various pHs, it was possible to evaluate the rate constant of SN• radicals with protons to be 2.1 × 109 M−1s−1 [15]. At pH 7, these reactions occur with the pseudo-first order rate constant 2.1 × 102 s−1, which are much lower than the pseudo-first order rate constant measured for the decay of the SN• radicals formed from compound 1 (vide supra). Therefore, the most probable pathway responsible for the decay of the SN• involves N-protonation (e.g., by water molecules), followed by deprotonation at the αC-carbon leading to αC• radicals (cf. Scheme3). These reactions provide an irreversible entry to these C-centered radicals and were previously suggested for linear peptides containing Met residue [14]. Since αC• radicals are formed in a slow process, this fact, combined with their very short lifetime, rationalizes their absence in the resolved absorption spectra recorded at longer times.

2.2. γ-Radiolysis and Product Analysis − • • In addition to the reactive species e aq, HO , and H , radiolysis of neutral water leads + −1 also to H (0.28) and H2O2 (0.07); in parenthesis the G in µmol J [7]. In N2O-saturated solution, the G(HO•) = 0.56 µmol J−1, therefore HO• and H• account for 90% and 10%, respectively, of the reactive species (cf. Reactions (1) and (2)). N2O-saturated solutions containing compound 1 (1.0 mM) at natural pH were irradiated for 400 and 800 Gy under stationary state conditions with a dose rate of 46.7 Gy min−1 followed by LC–MS and high-resolution MS/MS analysis. A representative LC-MS analysis of the 800 Gy irradiated sample is shown in Figure3. Int. J. Mol. Sci. 2021, 22, x FOR PEER REVIEW 8 of 16

N2O-saturated solutions containing compound 1 (1.0 mM) at natural pH were irradiated for 400 and 800 Gy under stationary state conditions with a dose rate of 46.7 Gy min−1 followed by LC–MS and high-resolution MS/MS analysis. A representative LC-MS analysis of the 800 Gy irradiated sample is shown in Figure 3. Int. J. Mol. Sci. 2021, 22, x FOR PEER REVIEW 8 of 16

100 1

80 N2O-saturated solutions containing compound 1 (1.0 mM) at natural pH were irradi- 6 −1 Int. J. Mol. Sci. 2021, 22, 4773 ated for 400 and4 800 Gy under stationary state conditions with a dose rate of 46.7 Gy min8 of 16 followed by LC–MS and high-resolution MS/MS analysis. A representative LC-MS analy- 60 8 sis of the 800 Gy irradiated7 sample is shown in Figure 3. 5 9 10 11 40 100 Intensity,a.u. 1

20 80 2 14 16 18 20 6 22 4 3 0 60 8 7 2 4 6 8 10 12 14 16 18 205 22 924 26 28 10 11 retention time, min 40 Intensity,a.u.

Figure 3. HPLC run of γ-irradiation of N2O-saturated aqueous solutions of 1.0 mM N-acetyl me- −1 thionine methyl ester20 2(1) at natural pH at a dose of 800 Gy (dose rate of 46.7 Gy min ). The consumption of 1 led to the formation of14 10 products.16 18 Inset: 20expansion22 of the chromatogram between

14 and 22 min. 3 0

Eleven compounds2 4were6 detected8 10 12 in14 the16 chromatogram18 20 22 24 26including28 the starting material 1. All peaks were identified andretention their chemicaltime, min structures assigned by examination of their high-resolution mass data and characteristic fragmentation patterns (see Figure S7 FigureFigure 3. 3. HPLCHPLC run of γ--irradiationirradiation of N2OO-saturated-saturated aqueous solutions ofof 1.01.0 mMmMN N-acetyl-acetyl methio-me- and S8). Inthioninenine Schemes methyl methyl 3 ester and ester ( 14) atthe (1 natural) atstructures natural pH at pH aof dose at all a dose ofproducts 800 of Gy 800 (dose areGy (dose ratereported of rate 46.7 ofthat Gy 46.7 min will Gy−1 min).be The de-−1). consumption The con- scribed in somesumptionof 1 led detail to of the 1below. led formation to the formation of 10 products. of 10 products. Inset: expansion Inset: expansion of the chromatogramof the chromatogram between between 14 and Combination1422 and min. 22 of min. the pulse radiolysis results on reactive intermediates and the structural information obtained from the high-resolution MS/MS allows the depiction the mechanistic proposalElevenEleven of compounds compound werewere1 transformation detecteddetected in in the theby chromatogram γchromatogram-radiolysis (Scheme including including 3). theIt the is starting well starting material mate- documentedrial1. from All 1. peaksAll previous peaks were were studies identiﬁed identified on andMet and theirderivatives their chemical chemical that structures the structures sulfoxide assigned assigned 2 is byformed examination by examinationdue of their of to the in-situtheirhigh-resolution generation high-resolution of masshydrogen datamass and peroxidedata characteristic and [17, characteristic19], fragmentationwhile the fragmentation H• addition patterns patternsto (see theFigures sulfur (see S7Figure and S8S7). (k = 1.7 × 10andIn9 M Schemes −1S8).s−1) Inwith Schemes3 and the4 formation the 3 and structures 4 ofthe a structuressulfuranyl of all products ofradical all areproducts intermediate reported are that reported affords will be thatcom- described will be inde- pound 3 andscribedsome CH detail3S in• radical some below. detail (Scheme below. 4) [29,30]. Combination of the pulse radiolysis results on reactive intermediates and the structural information obtained from the high-resolution MS/MS allows the depiction the mechanistic proposal of compound 1 transformation by γ-radiolysis (Scheme 3). It is well documented from previous studies on Met derivatives that the sulfoxide 2 is formed due to the in-situ generation of hydrogen peroxide [17,19], while the H• addition to the sulfur (k = 1.7 × 109 M−1s−1) with the formation of a sulfuranyl radical intermediate affords compound 3 and CH3S• radical (Scheme 4) [29,30].

Scheme 4. The reaction of 1 with• H2O2 and H• atom affords sulfoxide 2 and α-aminobutyric 3 de- Scheme 4. The reaction of 1 with H2O2 and H atom affords sulfoxide 2 and α-aminobutyric 3 derivatives, respectively. rivatives, respectively. Combination of the pulse radiolysis results on reactive intermediates and the struc- From thetural pulse information radiolysis obtained studies described from the above, high-resolution the reaction MS/MS of HO• allows with 1 the(k = depiction1.1 the × 1010 M−1s−1mechanistic) followed one proposal main ofand compound two minor1 transformationpaths. In Scheme by 3,γ -radiolysisthe formation (Scheme of ad-3). It is well duct radicaldocumented (HOS•) is the from main previous path (path studies 1), onwhile Met the derivatives two minor that ones the are: sulfoxide the H-atom2 is formed due • abstractionto from the the in-situ CH2 generation-S-CH3 moiety of hydrogen to give the peroxide intermediate [17,19 αS], while• radicals the (path H addition 2) and to the sul- 9 −1 −1 • the H-atomSchemefur abstraction (k = 4. 1.7 The× fromreaction10 M the of sN 1- withCH) with- COH2 theO 2moiety and formation H• toatom give of affords athe sulfuranyl intermediate sulfoxide radical 2 and αC α intermediate-a minobutyricradical affords 3 de- • (path 3). Tablerivatives,compound 1 shows respectively.3 thatand CH1.1 3µsS aradicalfter the (Scheme pulse the4)[ distribution29,30]. percentages were 73.6, 1.9 and 11.35% forFrom HOS the•, αS pulse• and radiolysis αC•, but at studies 6 µs after described the pulse above, the relative the reaction contribution of HO • with 1 (k = 1.1From× 10the10 pulseM−1s radiolysis−1) followed studies one maindescribed and twoabove, minor the paths.reaction In of Scheme HO• with3, the 1 forma-(k = 1.1 ×tion 1010 of M adduct−1s−1) followed radical one (HOS main•) is and the two main minor path paths. (path 1),In Scheme while the 3, the two formation minor ones of are:ad- • • ductthe H-atom radical abstraction(HOS ) is the from main the path CH2 -S-CH(path 31),moiety while tothe give two the minor intermediate ones are:α theS Hradicals-atom abstraction(path 2) and from the H-atomthe CH2- abstractionS-CH3 moiety from to give the N-CH-CO the intermediate moiety αS to• give radicals the intermediate (path 2) and theαC •Hradical-atom abstraction (path 3). Table from1 shows the N that-CH- 1.1COµ moietys after theto give pulse the the intermediate distribution percentagesαC• radical (pathwere 73.6,3). Table 1.9 and 1 shows 11.35% that for 1.1HOS µs •a,fterαS •theand pulseαC •the, but distribution at 6 µs after percentages the pulse the were relative 73.6, • 1.9contribution and 11.35% changed for HOS to•, 5.3,αS•45.6 and andαC•, 3.5%, but at respectively. 6 µs after the The pulseHOS the relativefollows contribution a ﬁrst-order 5 −1 − decay (kd = 5.6 × 10 s ) by HO elimination to give the sulﬁde radical cation, which is at the crossroad of various possible reactions affording the intermediates SS•+, αS•, SN• and α • • C . We recall from pulse radiolysis section that the radiation chemical yields of HOS and the sum of radiation chemical yields of SS•+, αS• and SN• species at 3 µs and 6 µs Int. J. Mol. Sci. 2021, 22, 4773 9 of 16

after the pulse (Table1) suggest that the formation of SS•+, αS• and SN• species follows the decay of HOS•, as discussed in the previous section. We suggest that the disulfide radical cation (SS•+), which is in equilibrium with sulfide radical cation (S•+) and starting material, fragments and affords the observed disulfide 5 [17]. Moreover, the S•+ is prompt to deprotonation. Evidence from the pulse radiolysis experiments indicated that SN• radical is the progenitor of αC• radical and the proposed mechanism is suggested in Scheme3, including the tentatively suggested unimolecular decay by β-fragmentation with formation of acyl radical [28]. We do not have confirmation of this latter pathway by the present LC-MS data. Lastly, the αS• radicals, which can be both αS(1)• and αS(2)• are shown in Scheme3. Table1 reports the percentage contribution as a function of time, i.e., 1.9, 30.3 and 45.6% for 1.1, 3 and 6 µs after the pulse, which means that at longer time scale the αS• radicals are • • the only remaining reactive species together with CH3S radical derived from the H atom • • • reactivity (cf. Scheme4). The cross-termination of αS(1) and αS(2) with CH3S affords compounds 4 and 6, respectively (see Scheme3). The data from the high-resolution MS/MS showed that compounds 7, 8, 9, 10 and 11 are dimers of αS• radicals (Scheme3). Figure S8 shows the high-resolution MS/MS spectra of the five compounds. The accurate masses of these products, m/z 409.1484, 409.1480, 409.1480, 409.1479, 409.1478, correspond to a molecular weight MH+ equivalent of two αS• radicals. Although the fragmentation patterns are not diagnostic, some information can be extracted. All of them show that initial fragmentations with a loss of CH4O, CH4S, C2H4O and/or C2H2O. Further structural information may be obtained from the analysis of potential dias- teroisomers. What is the ratio of the two αS• radicals? It is expected αS(2)• radical to be in higher concentration than αS(1)• and in line with the higher stability of secondary vs. primary alkyl radical due to favorable deprotonation from the precursor sulfide radical cation. Assuming that the concentration of αS(2)• radical is twice that of the αS(1)• radical, it is expected to having αS(2)–αS(2) and αS(2)–αS(1) from a probability point of view of termination steps. Figure4 shows that αS(2)–αS(2) has four stereocenters, two are from the starting material fixed at the S configuration whereas two new stereocenters generated from the self-termination can be R or S. In total four products, two of them are identical, and therefore we expect to have SSSS, SRSS and SRRS diastereoisomers. Figure4 shows that αS(2)–αS(1) has three stereocenters, two are from the starting material fixed at the S configuration and one stereocenter is generated from the cross- termination of the two radicals, producing the diastereoisomers SSR and SRS. In total, five diastereoisomers that we associated with the five dimeric products detected by LC-MS in γ-radiolysis of the compound 1.

2.3. Photosensitized Oxidation by 3CB Sensitized by triplet CB (3CB) oxidation of Met derivatives leads to numerous transients, well characterized in our earlier studies [23,25,31]. Laser ﬂash photolysis (LFP) results of the 3CB with 1 are presented in Figure S9. The main species observed are 3CB and ketyl radical (CBH•), with a minor contribution of radical anion (CB•−) (Figure5). As expected, the transient species derived from compound 1 could not be directly observed due to the strong absorption overlap of the transients derived from CB photochemistry. However, it is expected that the radical coupling reactions derived from the two αS• radicals are predominant. The Ar-saturated solutions containing CB (4 mM) and compound 1 (20 mM) at natural pH were irradiated for 20 min using a CW 355 nm laser (50 mW, see Experimental Section for details). A representative LC-MS analysis of the irradiated solution is shown in Figure6 . A variety of compounds were detected in the chromatogram, including the starting material 1 and CB. The high-resolution mass data and fragmentation patterns of all peaks were analyzed and information on their chemical structures were extracted. In Scheme5, the Int. J. Mol. Sci. 2021, 22, x FOR PEER REVIEW 9 of 16

changed to 5.3, 45.6 and 3.5%, respectively. The HOS• follows a first-order decay (kd = 5.6 × 105 s−1) by HO− elimination to give the sulfide radical cation, which is at the crossroad of various possible reactions affording the intermediates SS•+, αS•, SN• and αC•. We recall from pulse radiolysis section that the radiation chemical yields of HOS• and the sum of radiation chemical yields of SS•+, αS• and SN• species at 3 µs and 6 µs after the pulse (Table 1) suggest that the formation of SS•+, αS• and SN• species follows the decay of HOS•, as discussed in the previous section. We suggest that the disulfide radical cation (SS•+), which is in equilibrium with sulfide radical cation (S•+) and starting material, fragments and affords the observed disulfide 5 [17]. Moreover, the S•+ is prompt to deprotonation. Evidence from the pulse radiolysis experiments indicated that SN• radical is the progenitor of αC• radical and the proposed mechanism is suggested in Scheme 3, including the tentatively suggested unimolecular decay by β-fragmentation with formation of acyl radical [28]. We do not have confirmation of this latter pathway by the present LC-MS data. Lastly, the αS• radicals, which can be both αS(1)• and αS(2)• are shown in Scheme 3. Table 1 reports the percentage contribution as a function of time, i.e., 1.9, 30.3 and 45.6% for 1.1, 3 and 6 µs after the pulse, which means that at longer time scale the αS• radicals are the only remaining reactive species together with CH3S• radical derived from the H• atom reactivity (cf. Scheme 4). The cross-termination of αS(1)• and αS(2)• with CH3S• affords compounds 4 and 6, respectively (see Scheme 3). The data from the high-resolution MS/MS showed that compounds 7, 8, 9, 10 and 11 are dimers of αS• radicals (Scheme 3). Figure S8 shows the high-resolution MS/MS spectra of the five compounds. The accurate masses of these products, m/z 409.1484, 409.1480, 409.1480, 409.1479, 409.1478, correspond to a molecular weight MH+ equivalent of two αS• radicals. Although the fragmentation patterns are not diagnostic, some information can be extracted. All of them show that initial fragmentations with a loss of CH4O, CH4S, C2H4O and/or C2H2O. Further structural information may be obtained from the analysis of potential dias- teroisomers. What is the ratio of the two αS• radicals? It is expected αS(2)• radical to be in higher concentration than αS(1)• and in line with the higher stability of secondary vs. primary alkyl radical due to favorable deprotonation from the precursor sulfide radical cation. Assuming that the concentration of αS(2)• radical is twice that of the αS(1)• radical, Int. J. Mol. Sci. 2021, 22, 4773 it is expected to having αS(2)–αS(2) and αS(2)–αS(1) from a probability10 of 16 point of view of Int. J. Mol. Sci. 2021, 22, x FOR PEER REVIEWtermination steps. Figure 4 shows that αS(2)–αS(2) has four stereocenters,10 of 16 two are from

the starting material fixed at the S configuration whereas two new stereocenters generated structuresfrom the of self major-termination products are can reported be R or that S. willIn total be described four products, in some detailtwo of in them the are identical, Int. J. Mol. Sci. 2021, 22, x FOR PEER REVIEWFigure 4 shows that αS(2)–αS(1) has three stereocenters, two are from the starting 10 of 16 followingand therefore paragraphs. we expect to have SSSS, SRSS and SRRS diastereoisomers. material fixed at the S configuration and one stereocenter is generated from the cross- termination of the two radicals, producing the diastereoisomers SSR and SRS. In total, five diastereoisomers that we associated with the five dimeric products detected by LC-MS in γ-radiolysisFigure of the4 showscompound that 1. αS(2)–αS(1) has three stereocenters, two are from the starting material fixed at the S configuration and one stereocenter is generated from the cross- 2.3.termination Photosensitized of Oxidationthe two radicals,by 3CB producing the diastereoisomers SSR and SRS. In total, five diastereoisomersSensitized by triplet that CB we (3CB) associated oxidation ofwith Met thederivatives five dimeric leads to products numerous detectedtransi- by LC-MS in ents,γ-radiolysis well characterized of the compoundin our earlier 1 studies. [23,25,31]. Laser flash photolysis (LFP) results of the 3CB with 1 are presented in Figure S9. The main species observed are 3CB and ketyl radical (CBH•), with a minor contribution of radical anion (CB•–) (Figure 5). As ex- 3 pected,2.3. Photosensitized the transient species Oxidation derived by from CB compound 1 could not be directly observed due to the strongSensitized absorption by triplet overlap CB of the (3CB) transients oxidation derived of from Met CB derivatives photochemistry. leads How- to numerous transi- • ever,ents, it iswell expected characterized that the radical in our coupling earlier reactions studies der [23,25,31].ived from the Laser two αSflash radicals photolysis (LFP) re- are predominant. sults of the 3CB with 1 are presented in Figure S9. The main species observed are 3CB and ketyl radical (CBH•), with a minor contribution of radical anion (CB•–) (Figure 5). As expected, the transient species derived from compound 1 could not be directly observed due

to the strong absorption overlap of the transients derived from CB photochemistry. How- Figureever,Figure 4.it 4.The is The expected chemical chemical structures that structures the of αradicalS(2)– ofα αS(2)S(2) couplingand–αS(2)αS(2)– andreactionsαS(1) αS(2)indicating– αS(1)derived the indicating conﬁguration from the the oftwo configuration αS• radicals of stereocenters.stereocenters. are predominant. Figure 5. Structures of the sensitizer (3-carboxybenzophenone, CB) and its respective photo-reduction product (ketyl radical, CBH•).

The Ar-saturated solutions containing CB (4 mM) and compound 1 (20 mM) at natural pH were irradiated for 20 min using a CW 355 nm laser (50 mW, see Experimental Section for details). A representative LC-MS analysis of the irradiated solution is shown in Figure 6. A variety of compounds were detected in the chromatogram, including the starting material 1 and CB. The high-resolution mass data and fragmentation patterns of allFigureFigure peaks 5. 5.Structureswere Structures analyzed of the of sensitizer and the information sensitizer (3-carboxybenzophenone, (3 on-carboxybenzophenone, their chemical CB) andstructures its respective CB) were and photo-reduction extracted. its respective In photo-reduc- Schemeproducttion product (ketyl5, the radical,structures (ketyl CBH radical, •of). major CBH products•). are reported that will be described in some detail in the following paragraphs.

100The Ar-saturated solutions containing CB (4 mM) and compound 1 (20 mM) at natural pH were1 irradiated for 20 min using a CW 355 nm laser (50 mW, see Experimental CB Section80 for details). A representative LC-MS analysis of the irradiated solution is shown in Figure 6. A variety of compounds were detected in the chromatogram, including the starting material 1 and CB. The high-resolution mass data and fragmentation patterns of 60 all peaks were analyzed and information14,15 on their chemical structures were extracted. In Scheme 5, the structures of major products are reported that will be described in some 40 detailIntensity,a.u. in the following paragraphs. 18 8, 9, 10 16,17 12,13 19 20100 1 2 7 11 0 CB 80 2 4 16 18 20 22 24 26 28 30 32 34 36 retention time, min FigureFigure 6. 6. HPLC60HPLC run run of of 355 355 nm nm irradiated irradiated solution solution containing containing CBCB (4(4 mM) mM) and and NN-acetyl-acetyl methionine methionine methylmethyl ester ester (1 ()1 (20) (20 mM). mM). The The peaks peaks are are labelled labelled with with14,15 numbers numbers referring referring to products to products shown shown in in

SchemeScheme 5 and Figu Figurere 77.. 40 Intensity,a.u. 18 8, 9, 10 12,13 16,17 19 20

2 7 11 0

2 4 16 18 20 22 24 26 28 30 32 34 36 retention time, min Figure 6. HPLC run of 355 nm irradiated solution containing CB (4 mM) and N-acetyl methionine methyl ester (1) (20 mM). The peaks are labelled with numbers referring to products shown in Scheme 5 and Figure 7.

Int. J. Mol. Sci. 2021, 22, x FOR PEER REVIEW 12 of 16

Figure 7 shows that αS(2)–CBH has three stereocenters, one is from the starting material fixed at the S configuration and two stereocenters are generated from the cross- termination of the two radicals, producing the diastereoisomers SSS, SRS, SSR and SRR. Regarding the singlet in the shoulder of CB, having also m/z 414.1392, but different fragmentation patterns, it is assigned to αS(1)–CBH (Figure S13). As shown in Figure 7, this compound has two stereocenters, the usual S configuration from the starting material and a new one generated from the cross-termination of the two rad- Int. J. Mol. Sci. 2021, 22, 4773icals, producing the diastereoisomers SS and SR. It is likely that, under our HPLC 11 of 16 Int. J. Mol. Sci. 2021, 22, x FOR PEER REVIEW 11 of 16 conditions, the two diastereoisomers be under the same peak, or one of them overlap with CB.

Based on LFP results for the reactive intermediates and the structural information obtained from the high-resolution MS/MS, the mechanistic proposal of compound 1 transformation by photolysis can be depicted (Scheme 5). The formation of a complex between 3CB and 1 through its sulfur atom is well documented followed by the H-atom abstraction as the main pathway, yielding the two αS(1)• and αS(2)• radicals (path 1) [25,31,32]. Evi- dence that a small portion of the complex undergoes one electron transfer, with the formation of sulfide radical cation (S●+), was obtained (path 2). In analogy with the above- described mechanism in the radiolysis section, we detected traces of compound 5 suggest- ing a small contribution to the formation of αS(1)• and αS(2)• radicals. Moreover, we detected traces of sulfoxide 2. We speculate that the oxidation of sulfur is due the formation of a biradical and its further fragmentation to the corresponding sulfoxide (path 3). The

fate of αS• radicals will depend on the relative concentration of αS(1)• and αS(2)• and the Figure 7. TheFigure chemical 7. The structurespresence chemical of α structuresofS(2)–CBH, a CBH of• αradical. S(1)–CBHαS(2)–CBH, and αS(1) CBH–CBH)–CBH and indicating CBH–CBH) the indicating conﬁguration the ofconfig- stereocenters. uration of stereocenters.

The above analysis suggests that the concentration of αS(2)• radical is much higher than αS(1)•, as expected from the competition of the two H-atom abstraction steps with formation of secondary vs. primary alkyl radical, being the difference in BDE energy 2–3 kcal/mol. The LC run (Figure 6) together with the proposed mechanism (Scheme 5) suggests a ratio of 5–5.5 between the two αS• radicals. In this situation, the intermediates CBH• and αS(2)• will be the main players in the self-termination steps with formation of 18, 19 and 8, 9, 10, respectively, as well as in the cross-termination reaction with formation of 12, 13, 16 and 17 diastereoisomers.

3. Materials and Methods 3.1. Pulse Radiolysis The pulse radiolysis experiments were performed with the LAE-10 linear accelerator at the Institute of Nuclear Chemistry and Technology in Warsaw, Poland with a typical electron pulse length of 10 ns and 10 MeV of energy. A detailed description of the experimental setup has been given elsewhere along with basic details of the equipment and its data collection system [33,34]. The 1 kW UV-enhanced xenon arc lamp (Oriel Instruments, Stratford, CT, USA) was applied as a monitoring light source. The respective wavelengths SchemeScheme 5. 5. ProposedProposed mechanism mechanism for for the the reaction reaction of of 3CB 3CB triplet triplet with with NN--acetylacetyl methionine methionine methyl methyl ester ester ( (11)) at at natural natural pH pH 7. 7. were selected by MSH 301 monochromator (Lot Oriel Gruppe, Darmstadt, Germany) with a resolution of 2.4TheBased nm. HPLC The on LFPrunintensity in results Figure of foranalysing 6 theshows reactive compoundslight intermediateswas measu in threered andretentionby means the structural time of PMT intervals: information R955 (Hamamatsu, Hamamatsu City, Shizuoka, Japan). A signal from detector was digit- (i) obtainedIn the frominterval the high-resolutionof 17.5–19.5 min MS/MS, there are the three mechanistic major peaks proposal that ofcorrespond compound to1 com-trans- ised using a Le Croy WaveSurfer 104MXs-B (1 GHz, 10 GS/s) oscilloscope and then send formationpounds by 8, photolysis9 and 10, and can 2 beminor depicted peaks (Scheme that correspond5). The formation to compounds of a complex 7 and 11 between, which to PC for further3 processing. A water filter was used to eliminate near IR wavelengths. CBall and are1 thethrough dimers its of sulfur αS• radicals atom is observed well documented in γ-radiolysis followed experiments by the H-atom (cf. Scheme abstrac- 3). Absorbed doses per pulse were on the order of 11 Gyα (1 •Gy = 1α J kg−1•). Experiments tionIt as is theworth main underlining pathway, yieldingthat the accurate the two massesS(1) and of theseS(2) productsradicals and (path the 1)fragmenta- [25,31,32]. were performedEvidence with a that continuous a small portionflow of sample of the complex solutions undergoes using a standard one electron quartz transfer, cell with the tion patterns are identical in all sets of experiments (Figure S10). We assigned the with optical lengthformation 1 cm of at sulﬁderoom temperature radical cation (~22 (S °C).•+), Solutions was obtained were (pathpurged 2). for In at analogyleast with the structures 8, 9 and 10 to the 3 diastereoisomers of the αS(2)–αS(2) dimers and 7 and 20 min per 250above-described mL sample with mechanism N2O before in thepulse radiolysis irradiation. section, The weG-values detected were traces calcu- of compound 5 11 (minor peaks) to the 2 diastereoisomers of αS(2)–αS(1) reported in the previous lated from thesuggesting Schuler formula a small (Equation contribution (3)) to [35] the formation of αS(1)• and αS(2)• radicals. Moreover, radiolysis section (see Scheme 5). we detected traces of sulfoxide 2. We speculate that the oxidation of sulfur is due the for- (ii) mationIn the of interval a biradical of 32 and–34 its min further there fragmentation are two peaks to that the correspondingare individuated sulfoxide as compounds(path 3). The18 fate and of 19αS (Figure• radicals 6). willTheir depend accurate on masses the relative (m/z 455.1527, concentration and 455.1522) of αS(1)• correspondand αS(2)• to the MH+ of the dimer• CBH–CBH (Figure S11). Figure 7 shows that CBH–CBH has and the presence of a CBH radical. twoThe stereocenters HPLC run in and Figure a plane6 shows of symmetry compounds that in correspond three retention to erythro time intervals: and threo diastereoisomers. (i) In the interval of 17.5–19.5 min there are three major peaks that correspond to com- (iii) Inpounds the interval8, 9 ofand 2510–29, andmin 2there minor is the peaks major that peak correspond that corresponds to compounds to CB, with7 and two11 , doubletswhich allon arethe theright dimers and left of αsides,S• radicals respectively, observed and in oneγ-radiolysis singlet in experimentsthe shoulder (cf.of CBScheme (Figure3). 6). It In is this worth area underlining of HPLC run that there the accurateare the cr massesoss-coupling of these products products of andαS• andthe CBH fragmentation• radicals. The patterns accurate are masses identical of in two all couples sets of experiments of compounds (Figure named S10). 12, We13 (m/zassigned 414.1391, the structures414.1393) and8, 9 and16, 1710 (tom/z the 414.1390, 3 diastereoisomers 414.1393), as of well the α asS(2)– theirαS(2) fragmen- dimers tation patterns, are identical and assigned to αS(2)–CBH (Scheme 5 and Figure S12).

Int. J. Mol. Sci. 2021, 22, 4773 12 of 16

and 7 and 11 (minor peaks) to the 2 diastereoisomers of αS(2)–αS(1) reported in the previous radiolysis section (see Scheme5). (ii) In the interval of 32–34 min there are two peaks that are individuated as compounds 18 and 19 (Figure6). Their accurate masses ( m/z 455.1527, and 455.1522) correspond to the MH+ of the dimer CBH–CBH (Figure S11). Figure7 shows that CBH– CBH has two stereocenters and a plane of symmetry that correspond to erythro and threo diastereoisomers. (iii) In the interval of 25–29 min there is the major peak that corresponds to CB, with two doublets on the right and left sides, respectively, and one singlet in the shoulder of CB (Figure6). In this area of HPLC run there are the cross-coupling products of αS• and CBH• radicals. The accurate masses of two couples of compounds named 12, 13 (m/z 414.1391, 414.1393) and 16, 17 (m/z 414.1390, 414.1393), as well as their fragmentation patterns, are identical and assigned to αS(2)–CBH (Scheme5 and Figure S12). Figure7 shows that αS(2)–CBH has three stereocenters, one is from the starting material fixed at the S configuration and two stereocenters are generated from the cross-termination of the two radicals, producing the diastereoisomers SSS, SRS, SSR and SRR. Regarding the singlet in the shoulder of CB, having also m/z 414.1392, but different fragmentation patterns, it is assigned to αS(1)–CBH (Figure S13). As shown in Figure7, this compound has two stereocenters, the usual S configuration from the starting material and a new one generated from the cross-termination of the two radicals, producing the diastereoisomers SS and SR. It is likely that, under our HPLC conditions, the two diastereoisomers be under the same peak, or one of them overlap with CB. The above analysis suggests that the concentration of αS(2)• radical is much higher than αS(1)•, as expected from the competition of the two H-atom abstraction steps with formation of secondary vs. primary alkyl radical, being the difference in BDE energy 2–3 kcal/mol. The LC run (Figure6) together with the proposed mechanism (Scheme5) suggests a ratio of 5–5.5 between the two αS• radicals. In this situation, the intermediates CBH• and αS(2)• will be the main players in the self-termination steps with formation of 18, 19 and 8, 9, 10, respectively, as well as in the cross-termination reaction with formation of 12, 13, 16 and 17 diastereoisomers.

where [S] is the HO•-scavenger concentration and with respect to the current work, concentration of Ac-Met-OMe. This form of the Schuler formula gives G(S•) in units of µmol J−1, and with respect to the current work, [S] = 0.2 mM gives G(S•) = 0.557 µmol J−1 where (S•) corresponds to all radicals formed in the system. The dosimetry was based on −2 •− N2O-saturated solutions of 10 M KSCN which following radiolysis, produces (SCN)2 radicals that have a molar absorption coefﬁcient of 7580 M−1cm−1 at λ = 472 nm and are produced with a yield of G = 0.635 µmol J−1 from Equation (3) [36].

3.2. Spectral Resolutions of Transient Absorption Spectra The observed absorption spectra monitored at various time delays following the electron pulse, were transformed from A(λj) to Gε(λj) by multiplying A(λj) by the factor (F) from the dosimetry described in Section 3.1. A(λj) represents the absorbance change of the •− composite spectrum and F = ε472 × G(SCN)2 /A472 where ε472 is the molar absorption •− •− coefﬁcient of (SCN)2 at 472 nm and G(SCN)2 is the radiation chemical yield of the •− SCN)2 (see Section 3.1) and A472 represents the observed absorbance change in the thiocyanate dosimeter. The optical spectra thus converted were resolved into speciﬁc components (representing individual transients) by linear regression according to the following Equation (4) Gε(λi) = ∑ εj(λi)Gj (4) j

where εj is the molar absorption coefficient of the jth species and the regression parameters, Gj, are equal to the radiation-chemical yield of the jth species. The sum in Equation (4) is over all radical species present. For any particular time delay of an experiment, the regression analysis included equations such as Equation (4) for each λi under consideration. Further details of this method were described elsewhere [15]. The reference spectra of these transients were previously collected and applied in the spectral resolutions (cf. Figure S2 in Supplementary Materials) [15,17]. The molar absorption coefficients of the relevant transients, which will be further identified below, • −1 −1 • are provided in the following: HOS , λmax = 340 nm and ε340 = 3400 M cm ; αC , −1 −1 −1 −1 • λmax = 270 nm and ε270 = 6200 M cm and λmax= 370 nm and ε370 = 1800 M cm ; αS , −1 −1 •+ −1 −1 λmax = 290 nm and ε290 = 3000 M cm ; SS , λmax = 480 nm and ε480 = 6880 M cm ; • −1 −1 SN , λmax = 390 nm and ε390 4500 M cm . For the compound (1) studied in this work, it was not possible to generate these radicals selectively since they undergo fast mutual transformation. Based on our earlier experience, however, we can say with certainty that the changes in molar absorption coefficients within the same type of radicals are not significant and are in the limit of 15% error (5% variation in the experimental data and 10% combined error in the reported molar absorption coefficients for the UV-vis spectra of the intermediates under consideration). The fact that the combined yields of the transient species derived from their respective molar absorption coefficients are at the short time delays equal to the expected initial yield of the scavenged •OH radicals by 1, i.e., 0.56 mol J−1 (based on Schuler’s formula (Equation (3) in Section 3.1)), and they also never exceed this value, supports additional validation of the spectral resolutions and eliminates unreasonable fits. Involvement of the H• atom reaction is not reflected in the resolved transient spectra • since the forming CH3S (see Scheme4) is formed with a very low radiation chemical yield (<0.06 mol J−1) and is characterised by the very low molar absorption coefficient (<500 M−1cm−1). However, the involvement of H• atoms is reflected in the formation of final products 4 and 6 (see Scheme3).

3.3. Steady-State γ-Radiolysis Irradiations were performed at room temperature using a 60Co-Gammacell at different dose rates. The exact absorbed radiation dose was determined with the Fricke chemical dosimeter, by taking G(Fe3+) = 1.61 µmol J−1 [37]. Int. J. Mol. Sci. 2021, 22, 4773 14 of 16

3.4. Laser Flash Photolysis The laser flash photolysis (LFP) setup used in this work has been described in detail elsewhere [25,31]. Briefly, this setup employs Nd:YAG laser (Spectra Physics, Mountain View, CA, USA, model INDI 40-10) with 355 nm excitation wavelength as light pump and 150 W pulsed xenon (Applied Photophysics, Surrey, UK) to probe the excited sample. A flash photolysis experiment was performed in oxygen-free environment in 1 × 1 cm rectangular quartz fluorescence cells. Kinetic traces were recorded between 370 and 750 nm at 10 nm intervals. The sample contained the quencher-compound 1 (20 mM) and the sensitizer CB (4 mM) at pH = 7.

3.5. Steady-State Photolysis Steady-state photochemical irradiation experiments were performed in a 1 × 1 cm rectangular cell on an optical bench irradiation system using a Genesis CX355STM OPSL laser from Coherent (Santa Clara, CA, USA) with 355 nm emission wavelength (the output power used was set at 50 mW).

3.6. LC-MS/MS Measurements The LC-MS measurements were carried out using a liquid chromatography Thermo Scientific/ Dionex Ultimate 3000 system equipped with C18 reversed-phase analytical column (2.6 µm, 2.1 mm × 100 mm, Thermo-Scientific, Sunnyvale, CA, USA). The LC method employed a binary gradient of acetonitrile and water with 0.1% (v/v) formic acid. Separation was achieved with a gradient of 7–60% of acetonitrile at a flow rate of 0.3 mL/min for 42 min. This UHPLC system was coupled to a hybrid QTOF mass spectrometer (Impact HD, Bruker Daltonik, Bremen, Germany). The ions were generated by electrospray ionization (ESI) in positive mode. MS/MS fragmentation mass spectra were produced by collisions (CID, collision-induced dissociation) with nitrogen gas in the Q2 section of the spectrometer.

4. Conclusions Summarizing the role of various transient species obtained by the two different time- resolved techniques and their connection with the end-product formation (Schemes3 and5 ), the αS(2)• and αS(1)• radicals play an important role in both oxidation processes, although their mode of formation and relative concentration are quite different: (a) in radiolysis, the one-electron oxidation of sulfide (generated by HO• addition followed by HO− elimination) is followed by α-deprotonation with formation of αS(2)• and αS(1)• radicals, in an approximately 2:1 ratio; (b) in photolysis, the formation of a complex between 3CB and the sulfur moiety is followed by H-atom abstraction, yielding αS(2)• and αS(1)• radicals in an approximately 5:1 ratio. The final products are formed by radical-radical combination. The herein described complete and detailed study of the oxidation mechanisms of Met residue, simulating its position in the interior of long oligopeptides and proteins, represents a significant and original contribution to the understanding of oxidation reactions in real biological systems, i.e., proteins, applicable to research in protein therapeutics. This work offers a benchmark for the identification, quantification and mechanistic determination of products derived from oxidation of methionine derivatives.

Supplementary Materials: The following are available online at https://www.mdpi.com/article/10 .3390/ijms22094773/s1, Figure S1: The structures of six reactive intermediates identiﬁed in the pulse radiolysis experiments. Figure S2: Reference spectra used in the resolutions of the transient absorp- • tion spectra following OH-induced oxidation of CH3C(O)N-Met-OCH3. Figure S3: Resolution of the spectral components in the transient absorption spectrum recorded 1.1 µs after the electron pulse in N2O-saturated aqueous solution containing 0.2 mM AcN-Met-OMe at pH 7.0. Figure S4: Resolution of the spectral components in the transient absorption spectrum recorded 3 µs after the electron pulse in N2O-saturated aqueous solution containing 0.2 mM AcN-Met-OMe at pH 7.0. Figure S5: The sum of all radicals (HOS•, αC•, SS•+, SN•, αS•) taken in spectral resolutions as a function of time. Figure S6: First-order kinetic ﬁts of the growth and decay of radicals HOS• (panel A), SS•+ Int. J. Mol. Sci. 2021, 22, 4773 15 of 16

(panel B), αS• (panel C), αC• (panel D), and SN• (panel E). Figure S7: High-resolution MS/MS spectra of the products 4 (m/z 252.0731) and 6 (m/z 252.0732) derived from the cross-termination of • • • αS(1) and αS(2) with CH3S and product 5 (m/z 238.0578)–a disulfide. Figure S8: High-resolution MS/MS spectra of the five dimeric products 7 (m/z 409.1484), 8 (m/z 409.1480), 9 (m/z 409.1480), 10 (m/z 409.1479) and 11 (m/z 409.1478) derived from the combination of two αS• radicals. Figure S9: Transient absorption spectra following LFP of CB (4 mM) and N-AcMetOCH3 (20 mM) for different time delays at pH 7. Figure S10: High-resolution MS/MS spectra of the dimeric products 8 (m/z 409.1481), 9 (m/z 409.1484) and 10 (m/z 409.1480) derived from the combination of two αS(2)• radicals. Figure S11: High-resolution MS/MS spectra of the two dimeric products 18 (m/z 455.1527) and 19 (m/z 455.1522) derived from of the combination of two CBH• radicals. Figure S12: High- resolution MS/MS spectra of the products 12 (m/z 414.1391, 13 (m/z 414.1393), 16 (m/z 414.1390) and 17 (m/z 414.1393) derived from the cross-termination of αS• and CBH• radicals. Figure S13: High-resolution MS/MS spectra of the product 14 (m/z 414.1392) derived from the cross-termination of αS• and CBH• radicals. Author Contributions: Conceptualization, C.C., T.P., K.B. and B.M.; methodology, K.B., T.P., B.M. and C.C.; radiolysis experiments, K.S. and K.B.; photolysis experiments, K.G. and T.P.; software, P.F.; data analysis, C.C., K.B., T.P., K.G., P.F., K.S. and B.M.; writing—original draft preparation, C.C.; writing—review and editing, C.C., K.B., T.P. and B.M.; and funding acquisition, B.M. All authors have read and agreed to the published version of the manuscript. Funding: This research was funded by the National Science Centre, Poland, project no. UMO- 2017/27/B/ST4/00375 and by statutory funds of the Institute of Nuclear Chemistry and Technology (INCT) (K.B. and K.S.) Data Availability Statement: All data are displayed in the manuscript. Acknowledgments: Authors would like to thank G. L. Hug from the Notre Dame Radiation Labo- ratory (US) for critical comments on the results of time-resolved experiments and Tomasz Szreder (INCT, Poland) for his continuous efforts in implementing improvements to the pulse radiolysis setup in the INCT over the last few years. Conflicts of Interest: The authors declare no conflict of interest. The funders had no role in the design of the study; in the collection, analyses, or interpretation of data; in the writing of the manuscript, or in the decision to publish the results.

References 1. Hawkins, C.L.; Davies, M.J. Detection, identification, and quantification of oxidative protein modifications. J. Biol. Chem. 2019, 294, 19683–19708. [CrossRef] 2. Luo, S.; Levine, R.L. Methionine in proteins defends against oxidative stress. FASEB J. 2009, 23, 464–472. [CrossRef] 3. Lim, J.M.; Kim, G.; Levine, R.L. Methionine in Proteins: It’s Not Just for Protein Initiation Anymore. Neurochem. Res. 2019, 44, 247–257. [CrossRef][PubMed] 4. Javitt, G.; Cao, Z.; Resnick, E.; Gabizon, R.; Bulleid, N.J.; Fass, D. Structure and Electron-Transfer Pathway of the Human Methionine Sulfoxide Reductase MsrB3. Antioxid. Redox Signal. 2020, 33, 665–678. [CrossRef][PubMed] 5. Halliwell, B.; Gutteridge, J.M.C. Free Radicals in Biology and Medicine, 5th ed.; Oxford University Press: Oxford, UK, 2015. 6. Winterbourn, C.C. Reconciling the chemistry and biology of reactive oxygen species. Nat. Chem. Biol. 2008, 4, 278–286. [CrossRef] [PubMed] 7. Buxton, G.V.; Greenstock, C.L.; Helman, W.P.; Ross, A.B. Critical review of rate constants for reactions of hydrated electrons, hydrogen atoms and hydroxyl radicals (•OH/•O−) in aqueous solution. J. Phys. Chem. Ref. Data 1988, 17, 513–886. [CrossRef] 8. Schöneich, C. Radical-Based Damage of Sulfur-Containing Amino Acid Residues. In Encyclopedia of Radical in Chemistry, Biology and Materials; Studer, A.S., Ed.; John Wiley & Sons: Chichester, UK, 2012; Volume 3, pp. 1459–1474. 9. Schöneich, C. Sulfur Radical-Induced Redox Modifications in Proteins: Analysis and Mechanistic Aspects. Antioxid. Redox Signal. 2017, 26, 388–405. [CrossRef][PubMed] 10. Schöneich, C.; Bobrowski, K. Intramolecular hydrogen transfer as the key step in the dissociation of hydroxyl radical adducts of (alkylthio)ethanol derivatives. J. Am. Chem. Soc. 1993, 115, 6538–6547. [CrossRef] 11. Houée-Levin, C.; Bobrowski, K. The use of methods of radiolysis to explore the mechanisms of free radical modifications in proteins. J. Proteom. 2013, 92, 51–62. [CrossRef][PubMed] 12. Glass, R.S.; Hug, G.L.; Schöneich, C.; Wilson, G.S.; Kuznetsova, L.; Lee, T.; Ammam, M.; Lorance, E.; Nauser, T.; Nichol, G.S.; et al. Neighboring Amide Participation in Thioether Oxidation: Relevance to Biological Oxidation. J. Am. Chem. Soc. 2009, 131, 13791–13805. [CrossRef] Int. J. Mol. Sci. 2021, 22, 4773 16 of 16

13. Schöneich, C.; Pogocki, D.; Wisniowski, P.; Hug, G.L.; Bobrowski, K. Intramolecular Sulfur-Oxygen Bond Formation in Radical Cations of N-Acetylmethionine Amide. J. Am. Chem. Soc. 2000, 122, 10224–10225. [CrossRef] 14. Schöneich, C.; Pogocki, D.; Hug, G.L.; Bobrowski, K. Free radical reactions of methionine in peptides: Mechanisms relevant to b-amyloid oxidation and Alzheimer’s disease. J. Am. Chem. Soc. 2003, 125, 13700–13713. [CrossRef][PubMed] 15. Bobrowski, K.; Hug, G.L.; Pogocki, D.; Marciniak, B.; Schöneich, C. Stabilization of sulfide radical cations through complexation with the peptide bond: Mechanisms relevant to oxidation of proteins containing multiple methionine residues. J. Phys. Chem. B 2007, 111, 9608–9620. [CrossRef][PubMed] 16. Hug, G.L.; Bobrowski, K.; Pogocki, D.; Hörner, G.; Marciniak, B. Conformational influence on the type of stabilization of sulfur radical cations in cyclic peptides. ChemPhysChem 2007, 8, 2202–2210. [CrossRef][PubMed] 17. Barata-Vallejo, S.; Ferreri, C.; Zhang, T.; Permentier, H.; Bischoff, R.; Bobrowski, K.; Chatgilialoglu, C. Radiation chemical studies of Gly-Met-Gly in aqueous solution. Free Radic. Res. 2016, 50, S24–S39. [CrossRef][PubMed] 18. Bobrowski, K.; Hug, G.L.; Pogocki, D.; Marciniak, B.; Schöneich, C. Sulfur radical cation peptide bond complex in the one-electron oxidation of S-methylglutathione. J. Am. Chem. Soc. 2007, 129, 9236–9245. [CrossRef][PubMed] 19. Barata-Vallejo, S.; Ferreri, C.; Chatgilialoglu, C. Radiation chemical studies of methionine in aqueous solution: Understanding the role of molecular oxygen. Chem. Res. Toxicol. 2010, 23, 258–263. [CrossRef] 20. Torreggiani, A.; Barata-Vallejo, S.; Chatgilialoglu, C. Combined Raman and IR spectroscopic studies on the radical-based modifications of methionine. Anal. Bioanal. Chem. 2011, 401, 1231–1239. [CrossRef][PubMed] 21. Bobrowski, K.; Houée-Levin, C.; Marciniak, B. Stabilization and Reactions of Sulfur Radical Cations: Relevance to One-Electron Oxidation of Methionine in Peptides and Proteins. Chimia 2008, 62, 728–734. [CrossRef] 22. Filipiak, P.; Bobrowski, K.; Hug, G.L.; Schöneich, C.; Marciniak, B. N-Terminal Decarboxylation as a Probe for Intramolecular Contact Formation in γ-Glu-(Pro)n-Met Peptides. J. Phys. Chem. B 2020, 124, 8082–8098. [CrossRef] 23. Hug, G.L.; Bobrowski, K.; Kozubek, H.; Marciniak, B. Photo-oxidation of Methionine-containing Peptides by the 4- Carboxybenzophenone Triplet State in Aqueous Solution. Competition between Intramolecular Two-centered Three-electron Bonded (S\S)+ and (S\N)+ Formation. Photochem. Photobiol. 2000, 72, 1–9. [CrossRef] 24. Ignasiak, M.T.; Marciniak, B.; Houée-Levin, C. A Long Story of Sensitized One-Electron Photo-oxidation of Methionine. Isr. J. Chem. 2014, 54, 248–253. [CrossRef] 25. P˛edzinski,T.; Grzyb, K.; Ka´zmierczak, F.; Frański,R.; Filipiak, P.; Marciniak, B. Early Events of Photosensitized Oxidation of Sulfur-Containing Amino Acids Studied by Laser Flash Photolysis and Mass Spectrometry. J. Phys. Chem. B 2020, 124, 7564–7573. [CrossRef] 26. Mönig, J.; Goslich, R.; Asmus, K.-D. Thermodynamics of S\S 2σ/1σ* three-electron bonds and deprotonation kinetics of thioether radical cations in aqueous solution. Ber. Bunsenges. Phys. Chem. 1986, 90, 115–121. [CrossRef] 27. Rauk, A.; Yu, D.; Taylor, J.; Shustov, G.V.; Block, D.A.; Armstrong, D.A. Effects of Structure on aC-H Bond Enthalpies of Amino Acid residues: Relevance to H-Transfers in Enzyme mechanisms and in Protein Oxidation. Biochemistry 1999, 38, 9089–9096. [CrossRef][PubMed] 28. Chatgilialoglu, C.; Crich, D.; Komatsu, M.; Ryu, I. Chemistry of Acyl Radicals. Chem. Rev. 1999, 114, 1991–2069. [CrossRef] 29. Chatgilialoglu, C.; Ferreri, C.; Melchiore, M.; Sansone, A.; Torreggiani, A. Lipid geometrical isomerism: From chemistry to biology and diagnostics. Chem. Rev. 2014, 114, 255–284. [CrossRef][PubMed] 30. Chatgilialoglu, C.; Ferreri, C.; Torreggiani, A.; Salzano, A.M.; Renzone, G.; Scaloni, A. Radiation-induced reductive modifications of sulfur-containing amino acids within peptides and proteins. J. Proteom. 2011, 74, 2263–2273. [CrossRef][PubMed] 31. P˛edzinski,T.; Markiewicz, A.; Marciniak, B. Photosensitized oxidation of methionine derivatives. Laser flash photolysis studies. Res. Chem. Intermed. 2009, 35, 497–506. [CrossRef] 32. Ignasiak, M.T.; P˛edzinski,T.; Rusconi, F.; Filipiak, P.; Bobrowski, K.; Houée-Levin, C.; Marciniak, B. Photosensitized Oxidation of Methionine-Containing Dipeptides. From the Transients to the Final Products. J. Phys. Chem. B 2014, 118, 8549–8558. [CrossRef] 33. Bobrowski, K. Free radicals in chemistry, biology and medicine: Contribution of radiation chemistry. Nukleonika 2005, 50 (Suppl. 3), S67–S76. 34. Mirkowski, J.; Wi´sniowski,P.; Bobrowski, K. INCT Annual Report 2000; INCT: Warsaw, Poland, 2001. 35. Schuler, R.H.; Hartzell, A.L.; Behar, B. Track effects in radiation chemistry. Concentration dependence for the scavenging of OH by ferrocyanide in N2O-saturated aqueous solutions. J. Phys. Chem. 1981, 85, 192–199. [CrossRef] − 36. Janata, E.; Schuler, R.H. Rate constant for scavenging e aq in N2O-saturated solutions. J. Phys. Chem. 1982, 86, 2078–2084. [CrossRef] 37. Klassen, N.V.; Shortt, K.R.; Seuntjens, J.; Ross, C.K. Fricke dosimetry: The difference between G(Fe3+) for 60Co gamma-rays and high-energy X-rays. Phys. Med. Biol. 1999, 44, 1609–1624. [CrossRef]