3-Hydroxyflavones and 3-Hydroxy-4-Oxoquinolines As

molecules

Review 3-Hydroxyﬂavones and 3-Hydroxy-4-oxoquinolines as Carbon Monoxide-Releasing Molecules

Tatiana Soboleva and Lisa M. Berreau *

Department of Chemistry & Biochemistry, Utah State University, 0300 Old Main Hill, Logan, UT 84322-0300, USA; [email protected] * Correspondence: [email protected]; Tel.: +1-435-797-3509

Received: 3 March 2019; Accepted: 27 March 2019; Published: 30 March 2019

Abstract: Carbon monoxide-releasing molecules (CORMs) that enable the delivery of controlled amounts of CO are of strong current interest for applications in biological systems. In this review, we examine the various conditions under which CO is released from 3-hydroxyflavones and 3-hydroxy-4-oxoquinolines to advance the understanding of how these molecules, or derivatives thereof, may be developed as CORMs. Enzymatic pathways from quercetin dioxygenases and 3-hydroxy-4-oxoquinoline dioxygenases leading to CO release are examined, along with model systems for these enzymes. Base-catalyzed and non-redox-metal promoted CO release, as well as UV and visible light-driven CO release from 3-hydroxyflavones and 3-hydroxy-4-oxoquinolines, are summarized. The visible light-induced CO release reactivity of recently developed extended 3-hydroxyflavones and a 3-hydroxybenzo[g]quinolone, and their uses as intracellular CORMs, are discussed. Overall, this review provides insight into the chemical factors that affect the thermal and photochemical dioxygenase-type CO release reactions of these heterocyclic compounds.

Keywords: Gasotransmitter; mechanism; carbon monoxide; oxygen; heme oxygenase

1. Introduction Advancing understanding of carbon monoxide (CO) as a signaling molecule in humans, and evaluating its potential application as a therapeutic agent, are significant current goals in biomedical research [1,2]. Whether endogenously generated or introduced as CO gas, low concentrations of CO produce vasodilation, anti-inflammatory and anti-apoptotic effects that suggest it could be useful in treating cardiovascular disease [3–6]. The delivery of controlled amounts of CO to tumor tissue is also suggested as an approach for treating cancer [7,8]. Providing a precise amount of CO gas to a specific tissue or organ is difficult. Thus, to enhance the spatiotemporal control of CO release, carbon-monoxide releasing molecules (CORMs) have been developed. The vast majority of CORMs reported to date that have been used in in vitro and in vivo studies are metal carbonyl compounds, with the [Ru(CO)3Cl2]2 dimer (CORM-2), and the water soluble molecules [Ru(CO)3Cl(glycinato)] (CORM-3) and [Mn(CO)4(S2CNCH3)(CH2CO2H)] (CORM-401), having been the most extensively employed [9,10]. Notably, the Ru(II) complexes exhibit complicated solution chemistry, including in blood plasma, and interact with proteins [11–16]. Thus, these compounds do not exhibit a single predictable formulation for CO delivery [16,17]. Complicating matters further is the fact that CO release from CORM-2, CORM-3 and CORM-401 occurs spontaneously via ligand exchange. This means that CO release begins when the compound is dissolved in a solution or buffer, and that there is no spatiotemporal control over CO delivery. Notably, recent studies also suggest that some biological effects that have been reported for the Ru(II)-containing CORMs and CORM-401 may not be due to CO release, but instead result from the reactivity of the metal complex or a fragment thereof [18,19].

Molecules 2019, 24, 1252; doi:10.3390/molecules24071252 www.mdpi.com/journal/molecules Molecules 2019, 24, 1252 2 of 26

These limitations provide the impetus for the development of new types of CO-releasing molecules Moleculesthat do not2019, involve 24, x FOR a PEER metal REVIEW carbonyl moiety. 2 of 26 Several research labs have taken up the challenge of preparing transition metal-free CORMs [20]. ExamplesSeveral reported research to labs date have include taken sodium up the boronocarbonatechallenge of preparing (CORM-A1 transition and metal analogs)-free [CORMs21] and [20].norbornadiene-7-one Examples reported derivatives to date include [22], sodium both of boronocarbonate which are spontaneous (CORM-A1 CO-releasing and analogs) [21] CORMs and norbornadiene(Figure1(top)).-7 Visible-one derivatives light-induced [22], organicboth of which photoCORMs are spontaneous reported CO to date-releasing include CORMs unsaturated (Figure 1(top)).cyclic diketones Visible [light23],- xantheneinduced carboxylic organic photoCORMs acid [24], and reportedmeso-carboxy to date BODIPY include [25 ] unsaturated (Figure1(bottom)). cyclic diketoneLimitationss [23], of these xanthene transition carboxylic metal-free acid [24], CORMs and andmeso photoCORMs-carboxy BODIPY include [25] low (Figure yield 1(bottom) multi-step). Limitationssynthetic procedures of these andtransition a lack ofmetal well-deﬁned-free CORMs CO release and photoCORMs products. Thus, include despite low these yield advancements, multi-step syntheticfurther work procedures is needed. and The a development lack of well of-defined a tunable CO structural release products.framework Thus for CO, despite release, these that advancements,enables modulation further of work key propertiesis needed. (e.g.,The d solubility,evelopment trackability of a tunable in structural cells, cellular framework or subcellular for CO releasetargeting,, that triggered enables CO modulation release) via of standardkey properties synthetic (e.g., and solubility, medicinal trackability chemistry in approaches cells, cellular would or subcellularadvance the targeting,ﬁeld. triggered CO release) via standard synthetic and medicinal chemistry approaches would advance the field.

Figure 1. ExamplesExamples of transition transition metal metal-free-free CO CO-releasing-releasing molecules (CORMs). Spontaneous (top) and visible light light-induced-induced (bottom) CORMs.

To design a new family of CORMs, we considered how CO is generated in in biological biological systems. systems. In humans, humans, CO CO is is endogenously endogenously generated generated with with an anequimolar equimolar amount amount of biliverdin of biliverdin in the in first the step first ofstep O2 of-dependent O2-dependent heme hemedegradation degradation in a reaction in a reaction catalyzed catalyzed by heme by oxyg hemeenase oxygenasess (HO-1 (HO-1and HO and-2, SchemeHO-2, Scheme 1(top))1 (top))[26,27]. [ 26The,27 ]. expression The expression of HO of-1 HO-1 protein protein is a response is a response to cellular to cellular stress, stress, with with the productsthe products of heme of heme degradation degradation exhibiting exhibiting notable notable biological biological effects. effects. Specifically, Specifically, CO CO modulates vascular tone tone and and biliverdin/bilirubin biliverdin/bilirubin function function as aspotent potent antioxidants antioxidants [28]. [28 In]. fungi In fungi and andbacteria, bacteria, the Othe2-dependent O2-dependent cleavage cleavage of 3- ofhydroxyflavone 3-hydroxyflavone and and3-hydroxy 3-hydroxy-4-oxoquinoline-4-oxoquinoline derivatives derivatives occurs occurs via enzymevia enzyme-catalyzed-catalyzed dioxygenase dioxygenase-type-type reactions reactions in in which which a a2,4 2,4-peroxo-peroxo species species undergoes undergoes O O-O-O and CC-C-C bond cleavage to give one equivalent each of CO and a carboxylic acid byproduct termed a depside (Scheme (Scheme1 (bottom)) 1(bottom)) [ 29 [29,30].,30]. The The biological biological implications implications of CO of production CO production in fungi in and fungi bacteria and bacteriacould be could for its be use for as its carbon, use as energy, carbon, or energy, electron or sourceelectron [31 source]. Overall [31]. the Overall similar the nature similar of nature the heme of theoxygenase heme oxygenase and bacterial/fungal and bacterial/fungal CO release CO reactions release reactions suggests thatsuggests if controlled that if controlled CO release CO reactivity release reactivityfrom 3-hydroxyflavone from 3-hydroxyflavone and 3-hydroxy-4-oxoquinoline and 3-hydroxy-4-oxoquinoline derivatives can derivatives be harnessed, can compoundsbe harnessed, of compoundsthis class will of provide this class a novelwill provide family ofa novel CORMs. family An attractiveof CORMs. feature An attractive of this strategy feature isof that this flavonols strategy isare that already flavonols under are significant already investigation under significant for their investigation beneficial for health their effects. beneficial For example, health effects. quercetin, For example,a 3-hydroxyflavone quercetin, derivativea 3-hydroxyflavone found in many derivative fruits and found vegetables, in many is fruits noted and for vegetables, its potentially is beneficialnoted for itsanti-inflammatory, potentially beneficial antioxidant anti- andinflamm anti-canceratory, propertiesantioxidant [ 32 and–34 ]. anti Triggered-cancer 3-hydroxyflavone properties [32–34]. or Triggered3-hydroxy-4-oxoquinoline-based 3-hydroxyflavone or 3- CORMshydroxy that-4-oxoquinoline could be targeted-based toCORMs specific that locations could forbe targeted CO release to specificwould be locations especially for attractive CO release for would probing be the especially localized attractive effects of for CO probing delivery the [35 localized]. Using such effects bioactive of CO delframeworks,ivery [35]. itUsing may besuch possible bioactive to deliver frameworks, CO while it may producing be possible additional to deliver beneficial CO while effects. producing additional beneficial effects.

Molecules 2019, 24, 1252 3 of 26 Molecules 2019, 24, x FOR PEER REVIEW 3 of 26

SchemeScheme 1. 1.COCO release release reactions reactions catalyzed catalyzed by by heme heme oxygenase oxygenasess (top) (top) and and 3- 3-hydroxyﬂavonehydroxyflavone and and 3-hydroxy3-hydroxy-4-oxoquinoline-4-oxoquinoline dioxygenas dioxygenaseses (bottom). (bottom).

InIn this this review review we we examine examine enzymatic enzymatic and and non non-enzymatic-enzymatic reaction reaction pathways pathways leading leading to to CO CO releaserelease from from 3- 3-hydroxyflavonehydroxyflavone and and 3- 3-hydroxy-4-oxoquinolinehydroxy-4-oxoquinoline derivatives derivatives to to elucidate elucidate conditions conditions thatthat influence influence CO CO release release reactivity. reactivity. Additionall Additionally,y, w wee discuss discuss recent recent studies studies of of photoinduced photoinduced CO CO releaserelease from from novel novel extended extended 3- 3-hydroxyflavonehydroxyflavone and and 3- 3-hydroxybenzo[hydroxybenzo[g]quinoloneg]quinolone motifs. motifs. Overall, Overall, thesethese combined combined results provide provide insight insight into into the potential the potential to control to control CO release CO from release these from heterocyclic these heterocyclicframeworks frameworks and their potential and their for potential use as nature-inspired for use as nature CORMs.-inspired CORMs.

2. 2.CO CO Production Production from from 3- 3-HydroxyflavonesHydroxyflavones in in Enzyme Enzyme-Catalyzed-catalyzed Reactions Reactions and and Model Model Systems Systems 2.1. Fungal Flavonol Dioxygenases 2.1. Fungal Flavonol Dioxygenases Soil microorganisms have divalent metal-containing enzymes that catalyze CO release from Soil microorganisms have divalent metal-containing enzymes that catalyze CO release from 3-hydroxyflavones. The flavonol dioxygenase from the fungus A. japonicus is a homodimer with 3-hydroxyflavones. The flavonol dioxygenase from the fungus A. japonicus is a homodimer with a a mononuclear Cu(II) center in the N-terminal domain [36]. Two different active site coordination mononuclear Cu(II) center in the N-terminal domain [36]. Two different active site coordination geometries were identified for the Cu(II) center. Approximately 70% of the mononuclear Cu(II) geometries were identified for the Cu(II) center. Approximately 70% of the mononuclear Cu(II) sites sites exhibit a distorted tetrahedral coordination environment comprised of three histidine donors exhibit a distorted tetrahedral coordination environment comprised of three histidine donors and a and a water molecule. The remaining Cu(II) centers (~30%) have an additional glutamate ligand water molecule. The remaining Cu(II) centers (~30%) have an additional glutamate ligand (Glu73), (Glu ), resulting in an overall distorted trigonal bipyramidal geometry. EXAFS studies suggest resulting73 in an overall distorted trigonal bipyramidal geometry. EXAFS studies suggest that that monodentate coordination of the substrate quercetin occurs with displacement of the water monodentate coordination of the substrate quercetin occurs with displacement of the water molecule [37]. As the enzyme exhibits 1000-fold lower activity in a site-directed mutant lacking Glu , molecule [37]. As the enzyme exhibits 1000-fold lower activity in a site-directed mutant lacking73 it has been suggested that this residue acts as an active site base for deprotonation of the flavonol Glu73, it has been suggested that this residue acts as an active site base for deprotonation of the 3-hydroxyl moiety [38]. The consensus reaction mechanism proposed for CO release involves the flavonol 3-hydroxyl moiety [38]. The consensus reaction mechanism proposed for CO release distorted square pyramidal five-coordinate Cu(II) enzyme/substrate complex forming a low-lying involves the distorted square pyramidal five-coordinate Cu(II) enzyme/substrate3 complex forming a excited state Cu(I)-flavonoxy radical species that can undergo reaction with O2 either at the copper low-lying excited state Cu(I)-flavonoxy radical species that can undergo reaction with 3O2 either at center or with the radical on the substrate (Scheme2). Hybrid DFT calculations offer support for both the copper center or with the radical on the substrate (Scheme 2). Hybrid DFT calculations offer pathways [39,40]. The number and location of hydroxyl substituents on the substrate determines the support for both pathways [39,40]. The number and location of hydroxyl substituents on the relative rate of enzymatic oxidation [41]. This relates both to the redox potential of the substrate and to substrate determines the relative rate of enzymatic oxidation [41]. This relates both to the redox secondary hydrogen bonding interactions that influence the positioning of the substrate in the active potential of the substrate and to secondary hydrogen bonding interactions that influence the site [42]. positioning of the substrate in the active site [42].

Molecules 2019, 24, 1252 4 of 26 Molecules 2019, 24, x FOR PEER REVIEW 4 of 26

SchemeScheme 2. Proposed 2. Proposed reaction reaction pathway pathway for for CO CO release release catalyzed catalyzed by by fungal fungal ﬂavonol flavonol dioxygenases dioxygenases [39 ]. RDS[39]. indicates RDS indicates the rate-determining the rate-determining step. step.

2.2.2.2. Bacterial Bacterial Flavonol Flavonol Dioxygenases Dioxygenases ExamplesExamples of of flavonol flavonol dioxygenases dioxygenases from fromBacillus Bacillus subtilis subtilis[ 43[43–46–46]] and andStreptomyces Streptomyces sp. sp. FLAFLA[ [4747––52] have52] beenhave recentlybeen recently characterized. characterized. X-ray X- crystallographicray crystallographic studies studies of theof the iron-containing iron-containing form form of of the Bacillusthe Bacillus subtilis subtilisenzyme enzyme revealed revealed a mononuclear a mononuclear five-coordinate five-coordinate Fe(II) center Fe(II) in center each cupin in each domain cupin of thedomain bicupin of protein the bicupin [45]. Similarprotein to[45]. the SimilarA. japonicus to theenzyme, A. japonicus the metalenzyme, center the metal is ligated center by is three ligated histidine by residues,three histidine a glutamate, residues, and aa waterglutamate, molecule. and a Initial water kinetic molecule. studies Initial of the kineticB. subtilis studiesenzyme of theas B. a subtilis function of metalenzyme ion as provided a function an of activitymetal ion profile provided that hasan activity an approximate profile that match has an to approximate the Irving− Williamsmatch to metalthe ionIrving−Willi series for theams stability metal ion of His- series and for Glu-ligated the stability metal of His complexes- and Glu [-45ligated]. Subsequent metal complexes kinetic studies [45]. showedSubsequent that inclusion kinetic studies of four showed different that metal inclusion ions (Co(II),of four different Mn(II), Cu(II), metal ions and Ni(II))(Co(II), producedMn(II), Cu(II), higher levelsand ofNi(II)) turnover produced than higher the Fe(II)-containing levels of turnover enzyme, than the with Fe(II) the-conta highestining turnover enzyme, inwith the the presence highest of turnover in the presence of Mn(II) [46]. The kcat/KM value and−1 turnover number (~25 s−1) for the Mn(II) [46]. The kcat/KM value and turnover number (~25 s ) for the Mn(II)-containing B. subtilis Mn(II)-containing B. subtilis enzyme is similar to that reported for the CuII-containing A. japonicus enzyme is similar to that reported for the CuII-containing A. japonicus enzyme. Both bind approximately enzyme. Both bind approximately two equivalents of metal ion, while the CoII and FeII-containing B. two equivalents of metal ion, while the CoII and FeII-containing B. subtilis enzyme exhibit lower metal subtilis enzyme exhibit lower metal affinity as well as lower kcat/KM values. affinity as well as lower k /K values. The quercetinase fromcat MStreptomyces sp. FLA is a monocupin protein that exhibits the highest The quercetinase from Streptomyces sp. FLA is a monocupin protein that exhibits the highest turnover in the presence of Ni(II) and Co(II) [48]. Notably, anaerobic absorption spectra of the Ni(II) turnoveror Co(II) in-containing the presence Streptomyces of Ni(II) andsp. FLA Co(II) enzyme [48]. in Notably, the presence anaerobic of the absorption quercetin substrate spectra ofshowed the Ni(II) a orhypsochromic Co(II)-containing shiftsStreptomyces for the flavonol sp. FLA lowesenzymet energy in the absorption presence feature of the quercetin (Band I). substrateThe opposite showed is a hypsochromicobserved for the shifts Cu(II) for-containing the flavonol quercetinase lowest energy of Aspergillus absorption flavus featurewherein (Band a bathochromic I). The opposite shift of is observedthe flavonol for the Band Cu(II)-containing I is observed [49]. quercetinase Considering ofredoxAspergillus potentials, flavus the whereinformation a of bathochromic transient Co(I) shift- ofor the Ni(I) flavonol-flavonoxy Band radicalI is observed species [ 49 is]. unlikely Considering in the redox reaction potentials, pathway the of formation the B. subtilis of transient and Co(I)-Streptomyces or Ni(I)-flavonoxy sp. FLA quercetin radical dixoygenases species is unlikely(QDOs). This in the led reaction to the proposal pathway of the of theactiveB. site subtilis metaland Streptomycesion having sp. a FLA nonquercetin-redox role dixoygenases and instead (QDOs). acting as This a led condui to thet for proposal electron of transferthe active from site metal the ionmetal having-bound a non-redox flavonol rolesubstrate and instead to coordinated acting as dioxygen a conduit [46,48]. for electron Evidence transfer for froman active the metal-boundsite metal flavonolstructure substrate relevant to to coordinated this proposal dioxygen was found [46 ,in48 the]. Evidence X-ray crystallographic for an active site analysis metal of structure a cryotrapped relevant toversion this proposal of the wasNi(II) found-containing in the QDO X-ray from crystallographic Streptomyces sp. analysis FLA in of the a presence cryotrapped of O2 version [50]. While of the exhibiting the typical coordination of three histidines and one glutamate to the Ni(II) center, a Ni(II)-containing QDO from Streptomyces sp. FLA in the presence of O2 [50]. While exhibiting themonodentate typical coordination coordinated of three quercetin histidines (via the and 3- oneOH glutamate which is likely to the deprotonated Ni(II) center, at a the monodentate pH = 8 coordinatedoptimum for quercetin this enzyme) (via theand 3-OHa side- whichon coordinated is likely O deprotonated2 are also present at the(Figure pH 2). = 8 These optimum species for are this proposed to represent flavonoxy and superoxide radicals, respectively, having formed via electron enzyme) and a side-on coordinated O2 are also present (Figure2). These species are proposed to transfer from the flavonolato anion to O2 via the Ni(II) conduit. The O2 ligand is perpendicular to the represent flavonoxy and superoxide radicals, respectively, having formed via electron transfer from the C2-C3 and C3-C4 bonds, which subsequently undergo oxidative cleavage leading to CO release flavonolato anion to O via the Ni(II) conduit. The O ligand is perpendicular to the C2-C3 and C3-C4 from the C3-OH moiety.2 Quantum mechanics/molecular2 mechanics (QM/MM) simulations from two bonds, which subsequently undergo oxidative cleavage leading to CO release from the C3-OH moiety. different laboratories have provided insight into the possible reaction mechanism for the Quantum mechanics/molecular mechanics (QM/MM) simulations from two different laboratories

Molecules 2019, 24, 1252 5 of 26 have provided insight into the possible reaction mechanism for the Ni(II)-containing QDO [51,52]. Key differencesMoleculesMolecules between2019 2019, 24, 24x FOR, thex FOR PEER two PEER REVIEW studies REVIEW concern whether the active site glutamate (Glu74) is protonated5 of5 26 of 26 and the assignment of the rate-determining step. With regard to the former, Li, et al. performed their Ni(II)-containing QDO [51,52]. Key differences between the two studies concern whether the active investigationsNi(II)-containing with a protonated QDO [51,52]. Glu Key74 [ 51differences] whereas between Wang, the et al. two investigated studies concern both whether protonation the active levels sitesite glutamate glutamate (Glu (Glu74) 74is) protonatedis protonated and and the the assignment assignment of theof the rate rate-determining-determining step. step. Wit With regardh regard to to of Glu74 [52]. Both investigations suggest that dioxygen binding to the Ni(II) center results in electron thethe former, former, Li, Liet, al.et al.performed performed their their investigations investigations with with a protonated a protonated Glu Glu74 [51]74 [51] whereas whereas Wang Wang, et, al.et al. transfer from the deprotonated flavonolato anion to O2 resulting in the formation of a Ni(II) flavonoxy investigatedinvestigated both both protonation protonation levels levels of− Gluof Glu74 [52].74 [52]. Both Both investigations investigations suggest suggest that that dioxygen dioxygen binding binding radical/superoxide species with the O2 ligand coordinated in an end-on fashion. The ground state of to to the the Ni(II) Ni(II) center center results results in in electron electron transfer transfer from from the the deprotonated deprotonated flavonolato flavonolato anion anion to to O 2 O2 this complex is an open-shell singlet with a high-spin Ni(II) center antiferromagnetically− coupled to resultingresulting in in the the formation formation of of a Ni(II) a Ni(II) flavonoxy flavonoxy radical/superoxide radical/superoxide species species with with the the O 2 O ligand2− ligand the superoxidecoordincoordinatedated and in inan flavonoxy anend end-on- onfashion. radicals.fashion. The The Inground theground singlet state state of state, ofthis this complex the complex O2 moietyis anis anopen coordinatesopen-shell-shell singlet singlet in with an with end-on a a fashionhighhigh which-spin-spin Ni(II) differs Ni(II) center fromcenter antiferromagnetically theantiferromagnetically observed side-on coupled coupled binding to tothe inthe superoxide the superoxide X-ray and structure and flavonoxy flavonoxy [50 ].radicals. Attackradicals. In of In the terminalthethe singlet superoxide singlet state, state, oxygenthe the O2 Omoiety2 atom moiety coordinates at coordinates C2 of the in flavonoxy inan aendn end-on- radical onfashion fashion leads which which to differs the differs formation from from the the observed of observed a bridging Ni(II)-peroxosideside-on- onbinding species.binding in thein Li, the X et- rayX al.-ray structure suggest structure [50]. that [50]. Attack a Attack conformational of theof the terminal terminal change superoxide superoxide involving oxygen oxygen movementatom atom at C2at C2 of of of the proximalthethe flavonoxy Ni(II)-coordinated flavonoxy radical radical leads leads peroxo to theto the formation oxygen formation atomof aof bridging a closerbridging Ni(II) to Ni(II) the-peroxo C4-peroxo atom species. species. of the Li, flavonolatoLiet, al.et al.suggest suggest ligandthat that is the rate-determininga conformationala conformational step,change change with involving involving a free energy movement movement barrier of of the of the 19.9 proximal proximal kcal/mol Ni(II) Ni(II)- [coordinated51-coordinated]. However, peroxo peroxo the calculations oxygen oxygen atomatom closer closer to theto the C4 C4 atom atom of theof the flavonolato flavonolato ligand ligand is theis the rate rate-determining-determining step, step, with with a free a free energy energy of Wang, et al. instead suggest that this step is not rate-determining and proceeds to two different barrierbarrier of 19.9of 19.9 kcal/mol kcal/mol [51]. [51]. However, However, the the calculations calculations of Wangof Wang, et, al.et al.instea instead suggestd suggest that that this this step step intermediates that differ in terms of the orientation of the glutamate residue. In both calculations, is is not not rate rate-determining-determining and and proceeds proceeds to to two two different different intermediates intermediates that that differ differ in in terms terms of of the the subsequentorientationorientation formation of theof the glutamate ofglutamate a five-membered residue. residue. In bothIn both cyclic calculations, calculations, intermediate subsequent subsequent via formation C4-O formation bond of aof five formation a five-membered-membered enables concurrentcycliccyclic intermediate C2-C3 intermediate and C3-C4via via C4 C4- bondO -bondO bond cleavage, formation formation extrusion enables enables concurrent of concurrent CO, and C2 formationC2-C3-C3 and and C3 ofC3-C4 the-C4 bond depsidebond cleavage, cleavage, product (Schemeextrusionextrusion3). Wang, of of CO, et CO, al. and propose and formation formation that of cleavage of the the depside depside of the product cyclic product peroxide(Scheme (Scheme 3). is 3). rate-determining, Wang Wang, et, et al. al. propose propose with that athat free energycleavagecleavage barrier of oftheof 17.4the cyclic cyclic kcal/mol, peroxide peroxide is which rateis rate-determining, is-determining, similar towith experimentalwith a free a free energy energy kineticbarrier barrier of data 17.4of 17.4 (~15 kcal/mol, kcal/mol, kcal/mol) which which [48 ]. The calculationsis similaris similar to ofto experimental Wang, experimental et al. kinetic also kinetic indicate data data (~15 (~15 that kcal/mol) kcal/mol) if the glutamate [48]. [48]. The The calculations residue calculations is protonated,of of Wang Wang, et, et al.cleavage al. also also of the C2-C3indicateindicate and that O-Othat if theif bonds the glutamate glutamate should residue occur residue is leading protonated,is protonated, to the cleavage formation cleavage of ofthe of the aC2α C2--ketoC3-C3 and acidand O- O product, -bondsO bonds should which should has not beenoccuroccur experimentally leading leading to theto the formation observed formation of [ aof52 α a].- ketoα-keto acid acid product, product, whic which hash has not not been been experimentally experimentally observed observed [52].[52].

FigureFigure 2.FigureRepresentation 2. Representation 2. Representation of the of active ofthe the active site active features site site features features of cryotrapped of ofcryotrapped cryotrappedStreptomyces Streptomyces Streptomyces sp. FLA sp. sp. FLAin FLA the in presencein the the

of O2presence[50presence]. of Oof2 O[50].2 [50].

Scheme 3. Proposed reaction pathway for CO release from Ni(II)-containing QDO from Streptomyces sp. FLA [52]. RDS indicates the rate-determining step.

Molecules 2019, 24, x FOR PEER REVIEW 6 of 26

Molecules 2019, 24, 1252 6 of 26 Scheme 3. Proposed reaction pathway for CO release from Ni(II)-containing QDO from Streptomyces sp. FLA [52]. RDS indicates the rate-determining step. As outlined above, bacterial quercetin dioxygenases (QDOs) catalyze the O2-dependent cleavage of flavonolsAs outlined to release above, CO bacterialand the corresponding quercetin dioxygenases depside. Recently, (QDOs) Farmer catalyze and the co-workers O2-dependent have showncleavage that of the flavonols Mn-containing to release QDO CO from andBacillus the corresponding subtilis will catalyze depside. CO Recently,release from Farme quercetinr and inco- theworkers presence have ofshown a different that the small Mn-containing molecule, nitrosylQDO from hydride Bacillus (HNO) subtilis [53 will]. Thiscatalyze nitroxygenase CO release activityfrom quercetin results in in the the presence incorporation of a different of N and small O atomsmolecule, in nitrosylthe depside hydride product (HNO) (Scheme [53]. This4). Thenitroxygenase substitution activity of dioxygen results in with the incorporation nitrosyl hydride of N isand rationalized O atoms in onthe the depside basis product of the fact(Scheme that the4). The anionic substitution form of of nitrosyl dioxygen exists with asnitrosyl a triplet hydride (3NO is− )rationalized and is isoelectronic on the basis with of the dioxygen fact that [ 54the]. Theanionic Mn-QDO form of nitroxygenase nitrosyl exists reactivity as a triplet was (3 foundNO−) and to be is highly isoelectronic regioselective, with dioxygen producing [54]. only The 2-((3,4-dihydroxyphenyl)(imino)methoxy)-4,6-dihydroxybenzoateMn-QDO nitroxygenase reactivity was found to be highly (Scheme regioselective,4). Although producing dioxygenase only activity2-((3,4-dihydroxyphenyl)(imino)methoxy) is found for the Fe(II)- or Co(II)-containing-4,6-dihydroxybenzoate QDO enzymes, no (Scheme nitroxygenase 4). activity Although was founddioxygenase for these activity enzymes is found [53]. for The the nitroxygenase Fe(II)- or Co(II) reaction-containing pathway QDO has enzymes, been evaluated no nitroxygenase quantum chemicallyactivity was by found Wojdyla for these and enzymes Borowski [53]. [55]. The The nitroxygenase lowest energy reaction reaction pathway sequence has involvesbeen evaluated initial Mn(II)-O-NHquantum chemically complex by formation. Wojdyla Additionand Borowski of the [55]. NH moietyThe lowest of the energy Mn(II)-nitroxyl reaction sequence to the flavonolato involves C2initial center, Mn(II) followed-O-NH by com shiftingplex formation. of the hydrogen Addition bond of involving the NH moiety Glu69 from of the O3 Mn to(II) O4,-nitroxyl enables to C4-O the bondflavonolato formation C2 center, (Scheme followed4). The by rate-determining shifting of the hydrogen step, with bond a barrier involving of 21.5 Glu kcal/mol,69 from O3 involves to O4, cleavageenables C4 of- theO bond five-membered formation (Scheme ring leading 4). The to COrate extrusion.-determining The step, regioselectivity with a barrier of theof 21. reaction5 kcal/mol, is an inherentinvolves property cleavage of of the the reactants. five-membered Specifically, ring the leading larger to electrophilicity CO extrusion. of The the nitrogen regioselectivity atom in of HNO the makesreaction the is C2-N an inherent bond formation property more of feasible the reactants. than C2-O, Specifically, with the C2-N the larger quercetin:HNO electrophilicity adduct of being the >15nitrogen kcal/mol atom lower in HNO in energy makes versus the C2 the-N C2-O bond adduct. formation more feasible than C2-O, with the C2-N quercetin:HNO adduct being >15 kcal/mol lower in energy versus the C2-O adduct.

SchemeScheme 4.4. MechanismMechanism proposed forfor Mn(II)-containingMn(II)-containing bacterial QDO reactivity withwith HNOHNO based on quantumquantum chemicalchemical studiesstudies [[55].55]. 2.3. Model Systems for Cu(II)-Containing Fungal QDOs 2.3. Model Systems for Cu(II)-Containing Fungal QDOs Many synthetic model complexes have been reported for the enzyme/substrate adduct in Many synthetic model complexes have been reported for the enzyme/substrate adduct in fungal copper-containing quercetin dioxygenases, with many notable contributions from Speier fungal copper-containing quercetin dioxygenases, with many notable contributions from Speier and and co-workers [56–58]. These Cu(II) complexes typically contain a supporting bidentate or co-workers [56–58]. These Cu(II) complexes typically contain a supporting bidentate or tridentate tridentate nitrogen donor ligand to mimic histidine ligation and a bidentate-coordinated flavonolato nitrogen donor ligand to mimic histidine ligation and a bidentate-coordinated flavonolato ligand ligand involving a five-membered chelate ring formed via coordination of the deprotonated involving a five-membered chelate ring formed via coordination of the deprotonated 3-OH moiety 3-OH moiety and the 4-keto oxygen atom. This flavonolato coordination mode is different from and the 4-keto oxygen atom. This flavonolato coordination mode is different from the monodentate the monodentate coordination proposed in the enzyme/substrate complex of the A. japonicus coordination proposed in the enzyme/substrate complex of the A. japonicus enzyme, which is enzyme, which is suggested to result from secondary interactions involving the quercetin suggested to result from secondary interactions involving the quercetin substrate hydroxyl substrate hydroxyl substituents [36,37]. The synthetic Cu(II) flavonolato complexes are much less substituents [36,37]. The synthetic Cu(II) flavonolato complexes are much less reactive than the reactive than the enzyme, typically undergoing O2-dependent dioxygenase-type reactivity only at enzyme, typically undergoing O2-dependent dioxygenase-type reactivity only at elevated elevated temperatures (e.g., 80–120 ◦C) in DMF. Enzyme-like products are generated, specifically

Molecules 2019, 24, x FOR PEER REVIEW 7 of 26 temperatures (e.g., 80–120 °C) in DMF. Enzyme-like products are generated, specifically O-benzoylsalicylate (depside) which is often retained on the Cu(II) center (Scheme 5), and CO, which is typically not quantified. A bimolecular rate law for such reactions is –d[Cu(II) flavonolato complex]/dt = k[Cu(II) flavonolato complex][O2]. An example of such a reaction is the O2-dependent CO release reaction of [Cu(idpa)(fla)]ClO4 (Scheme 5; fla = monoanion of 3-hydroxyflavone; idpa = 3,3’-iminoMolecules-bis(2019N, 24,N, 1252-dimethylpropylamine)) [59]. Similar to the proposed enzyme reaction pathway,7 of 26 an idpa-coordinated Cu(I)-flavonoxy radical species is suggested to undergo single electron transfer from Cu(I) to O2 to form a Cu(II) superoxide adduct. The terminal oxygen atom of superoxide OMolecules-benzoylsalicylate 2019, 24, x FOR PEER (depside) REVIEW which is often retained on the Cu(II) center (Scheme5), and7 CO, of 26 moietywhich subse is typicallyquently notcombines quantified.with A the bimolecular flavonoxy rate radical law for to suchform reactions a bridging is –d[Cu(II) peroxo flavonolatospecies. Attack of the Cu(II)-coordinated peroxo oxygen atom at C4 results in the formation of a 2,4-cyclic peroxide complex]/dttemperatures = k (e.g.,[Cu(II) 80 flavonolato–120 °C) complex][O in DMF. E2].nzyme An example-like products of such a reactionare generated, is the O 2-dependent specifically ‡ speciesCOO-benzoylsalicylate from release which reaction CO (depside) extrusion of [Cu(idpa)(fla)]ClO which and depside is often4 (Scheme retainedformation5 ; on fla can the = occur.Cu(II) monoanion centerFor this of (Scheme 3-hydroxyflavone;reaction, 5), Δ andH CO,= 64 ± 5 kJ/molidpawhich and = is 3,3’-imino-bis(Δ typicallyS‡ = −120 not J/mol Nquantified.,N·K.-dimethylpropylamine)) The negativeA bimolecular entropy rate [ 59 oflaw]. activation Similarfor such to reactionsis the consistent proposed is –d[Cu(II) with enzyme the flavonolato O reaction2 activation step pathway,complex]/dtto produce an idpa-coordinateda= kCu(II)[Cu(II)-superoxo flavonolato Cu(I)-flavonoxy species complex][O being radical2]. rate An -exampledetermining. species isof suggested such Enhanced a reaction to undergo electronis the singleO2 -densitydependent electron within the flavonolatotransferCO release from reactionligand Cu(I) to incof O reases[Cu(idpa)(fla)]ClO2 to form the a Cu(II)oxygenation superoxide4 (Scheme rate adduct.5; with fla = themonoanion The Hammett terminal of oxygen 3- hydroxyflavone;value atom being of superoxide −0.29. idpa = moiety3,3’Notably,-imino subsequently - additionbis(N,N-dimethylpropylamine combines of an exogenous with the flavonoxy carboxylate)) [59]. Similar radical donor to to the form ligandproposed a bridging (acetate enzyme peroxo or reaction species. triphenylacetate) pathway, Attack of to thean idpa Cu(II)-coordinated-coordinated Cu(I) peroxo-flavonoxy oxygen atomradical at C4species results is insuggested the formation to undergo of a 2,4-cyclic single electron peroxide transfer species solutions of [Cu(idpa)(fla)]ClO4 results in an acceleration of the CO release reaction [57,60]. This is ‡ from which Cu(I) to CO O extrusion2 to form and a Cu(II) depside superoxide formation adduct. can occur. The For terminal this reaction, oxygen∆ atomH = 64 of ± superoxide5 kJ/mol attributed to Cu(II) coordination of the carboxylate ligand, which results in a shift to monodentate andmoiety∆S ‡subse= −120quently J/mol combines·K. The negative with the entropy flavonoxy of activationradical to isform consistent a bridging with peroxo the O species.activation Attack step coordination of the flavonolato ligand (Scheme 6). The enhanced electron density2 within the toof producethe Cu(II) a-coordinated Cu(II)-superoxo peroxo species oxygen being atom rate-determining. at C4 results in Enhancedthe formation electron of a 2,4 density-cyclic within peroxide the flavonolatoflavonolatospecies frommoiety ligand which at increasesC2 CO makes extrusion the direct oxygenation and electron depside rate transfer formation with the to Hammett canO2 moreoccur. ρ viableForvalue this beingthus reaction, enhancing−0.29. ΔH‡ = the64 ± rate 5 of - formationkJ/mol of and O 2Δ,S which‡ = −120 was J/mol detected·K. The negative in reaction entropy mixtures. of activation is consistent with the O2 activation step to produce a Cu(II)-superoxo species being rate-determining. Enhanced electron density within the flavonolato ligand increases the oxygenation rate with the Hammett  value being −0.29. Notably, addition of an exogenous carboxylate donor ligand (acetate or triphenylacetate) to solutions of [Cu(idpa)(fla)]ClO4 results in an acceleration of the CO release reaction [57,60]. This is attributed to Cu(II) coordination of the carboxylate ligand, which results in a shift to monodentate coordination of the flavonolato ligand (Scheme 6). The enhanced electron density within the flavonolato moiety at C2 makes direct electron transfer to O2 more viable thus enhancing the rate of formation of O2-, which was detected in reaction mixtures.

Scheme 5. Reactivity of [Cu(idpa)(fla)]ClO4 derivatives with O2 to release CO [59]. Scheme 5. Reactivity of [Cu(idpa)(fla)]ClO4 derivatives with O2 to release CO [59]. Notably, addition of an exogenous carboxylate donor ligand (acetate or triphenylacetate) to solutions of [Cu(idpa)(fla)]ClO4 results in an acceleration of the CO release reaction [57,60]. This is attributed to Cu(II) coordination of the carboxylate ligand, which results in a shift to monodentate coordination of the flavonolato ligand (Scheme6). The enhanced electron density within the flavonolato

moiety at C2 makes direct electron transfer to O2 more viable thus enhancing the rate of formation of - O2 , which wasScheme detected 5. Reactivity in reaction of [Cu(idpa)(fla)]ClO mixtures. 4 derivatives with O2 to release CO [59].

Scheme 6. Effect of carboxylate addition on flavonolato coordination [60].

Some Cu(II) flavonolato complexes, such as [Cu(phen)(fla)2] and [Cu(bpy)(fla)2], do not exhibit CO release but instead undergo reaction via a 1,2-dioxetane intermediate to give 2-hydroxyphenylglyoxylate following hydrolysis (Scheme 7) [61,62]. The ligand environment and Lewis acidity of the SchemeCu(II) 6.center Effect ofare carboxylate suggested addition to affect on ﬂavonolatoflavonolato the regioselectivity coordination [[60].of60]. carbon -carbon bond

Some Cu(II) flavonolato complexes, such as [Cu(phen)(fla)2] and [Cu(bpy)(fla)2], do not exhibit

CO release but instead undergo reaction via a 1,2-dioxetane intermediate to give 2-hydroxyphenylglyoxylate following hydrolysis (Scheme 7) [61,62]. The ligand environment and Lewis acidity of the Cu(II) center are suggested to affect the regioselectivity of carbon-carbon bond

Molecules 2019, 24, 1252 8 of 26

Some Cu(II) flavonolato complexes, such as [Cu(phen)(fla)2] and [Cu(bpy)(fla)2], do not exhibit CO release but instead undergo reaction via a 1,2-dioxetane intermediate to give 2-hydroxyphenylglyoxylate following hydrolysis (Scheme7)[ 61,62]. The ligand environment and MoleculesLewis 2019 acidity, 24, x FOR of the PEER Cu(II) REVIEW center are suggested to affect the regioselectivity of carbon-carbon8 of bond26 cleavage by influencing the electrophilicity of the flavonolato carbonyl center [61]. Evidence for the cleavageinvolvement by influencing of a 1,2-dioxetane the electrophilicity are emission of the bands flavonolato at 506, 546, carbonyl and 578 center nm [ 62[61].]. Evidence for the involvement of a 1,2-dioxetane are emission bands at 506, 546, and 578 nm [62].

SchemeScheme 7. O 7.2 rOeactivity2 reactivity of a of phenanthroline a phenanthroline-coordinated-coordinated Cu(II) Cu(II) flavonolato ﬂavonolato complex complex that that does does not not resultresult in CO in COrelease release [61]. [61 ].

2.4.2.4. Model Model Systems Systems for forBacterial Bacterial QDOs QDOs GrubelGrubel,, et al. et synthesized al. synthesized the thefirst first series series of divalent of divalent 3d metal 3d metal flavonolato flavonolato compexes compexes of structural of structural relevancerelevance to bacterial to bacterial quercetin quercetin dioxygenases dioxygenases [63]. [63 These]. These complexes, complexes, supported supported by an by aryl an aryl-appended-appended tris(pyridylmethyl)aminetris(pyridylmethyl)amine ligand, ligand, have have thus far far not not been been investigated investigated for thermal for thermal CO release CO release reactivity. reactivity.Sun, et al. Sun, and et Matuz, al. and et al. Matuz, subsequently et al. subsequently prepared, characterized, prepared, characterized, and examined and the O examined2-dependent the CO release reactivity of two series of divalent metal ion flavonolato complexes (Scheme8)[ 64,65]. Sun, et al. O2-dependent CO release reactivity of two series of divalent metal ion flavonolato complexes studied complexes supported by a carboxylate-containing N O-donor ligand (Scheme8a) [ 64]. A metal (Scheme 8) [64,65]. Sun, et al. studied complexes supported by3 a carboxylate-containing N3O-donor ligandion dependence(Scheme 8(a)) on [64]. the A second-order metal ion dependence rate constant on forthe oxidative second-order cleavage rate (Fe(II)constant > Cu(II)for oxidative > Co(II) > cleavageNi(II) >Zn(II)(Fe(II) > > Cu(II) Mn(II)) > Co(II) correlates > Ni(II) with >Zn(II) the oxidation > Mn(II)) potential correlates of the with coordinated the oxidation flavonolato potential ligand, of thewhich coordinated in turn isflavonolato influenced ligand, by the Lewiswhichacidity in turn of is the influenced metal center. by the These Lewis oxidative acidity cleavage of the reactionsmetal center.occurred These at oxidative lower temperatures cleavage reactions than those occurred reported at for lower synthetic temperatures complexes than having those a neutralreported nitrogen for syntheticsupporting complexes ligand having environment a neutral [ 65nitrogen], likely support due toing the ligand reduced environment Lewis acidity [65], oflikely the due metal to center.the reducedThe yield Lewis of CO acidity produced of the using metal the center. Fe(II) complexesThe yield ofsupported CO produced by carboxylate-containing using the Fe(II) complexes N3O-donor ligand was determined to be 68%. The series of complexes reported by Matuz, et al. (Scheme8b,c supported by carboxylate-containing N3O-donor ligand was determined to be 68%. The series of complexesexhibit similarreported reactivity by Matuz, in et terms al. (Scheme of the products 8(b) and generated8(c)) exhibit [65 similar]. The reactivity ester-appended in terms derivatives of the productsshown generated in Scheme 8[65].c are The more ester reactive-appended that the derivatives propionate shown analogs in (Scheme Scheme8 b),8(c) suggesting are more thatreactive ligand thatexchange the propionate involving analogs the monodentate (Scheme anion 8(b)), is suggestingimportantin that the rate-determining ligand exchange O 2 involactivationving thestep. monodentate anion is important in the rate-determining O2 activation step.

Molecules 2019, 24, 1252 9 of 26 Molecules 2019, 24, x FOR PEER REVIEW 9 of 26 Molecules 2019, 24, x FOR PEER REVIEW 9 of 26

Scheme 8.Scheme Dioxygenase-typeDioxygenase 8. Dioxygenase-type reactivity-type reactivity of divalent of metal divalent complexes metal complexes containing containing a carboxylate a carboxylate or ester or ester appendageappendage [[64,65].64,65]. [64,65].

EnhancingEnhancing the electron the densityelectron withindensity the within carboxylatecarboxyl the carboxylate moietyate of moiety thethe supportingsupporting of the supporting chelatechelate ligandligand chelate ligand producesproduces structural,structural, structural, electronicelectronic electronic andand reactivityreactivity and reactivity effectseffects viavia effects thethe benzoate-M(II)-O(4)=C(27)-C(21)=C(22)benzoate via the benzoate-M(II)-O(4)=C(27)-M(II)-O(4)=C(27)-C(21)=C(22)-C(21)=C(22) conduit inconduit in M(II)-containing M(II) in-containing M(II)-containing flavonolato flavonolato complexes.flavonolato complexes. Specifically, complexes. Specifically, incorporation Specifically, incorporation of incorporation an electron of an donating electron of an electron groupdonating ondonating the group ligand on group benzoatethe ligand on the moiety benzoate ligand produces benzoate moiety a smaller produces moiety torsion produces a smaller angle abetween torsionsmaller the angle torsion flavonol between angle B and between the C the π→π ringsflavonol (Scheme flavonolB and9), C a Brings and (Scheme*C absorption rings (Scheme9), a band π 9),π that* aabsorption π is shiftedπ* absorption toband lower that energy,band is shifted that and is to ashifted lower lower redox toenergy, lower potential andenergy, a and a forlower the redox flavonolatolower potential redox moiety. potential for the These flavonolato for characteristics the flavonolato moiety. result Thesemoiety. in highercharacteristics These dioxygenase characteristics result reactivity in resulthigher in in dioxygenase Co(II), higher Ni(II) dioxygenase andreactivity Fe(II)reactivity in complexes Co(II) ,in Ni Co [66(II)(II)– 68and,]. Ni COFe(II)(II) quantification and complexes Fe(II) complexes [66 studies–68]. wereCO[66– quantification68]. not CO reported quantification forstudies these were studies reactions. not were reported not reported for thesefor reactions. these reactions.

Scheme 9.SchemeElectronic 9. Electronic effects on effects flavonolato on flavonolato ligand reactivity ligand reactivity with O . The with blue O2. highlightedThe blue highlighted portion of portion of Scheme 9. Electronic effects on flavonolato ligand reactivity with O22. The blue highlighted portion of thethe structurestructurethe showsstructureshows thethe shows benzoate-M(II)-O(4)=C(27)-C(21)=C(22)benzoate the -benzoateM(II)-O(4)=C(27)-M(II)-O(4)=C(27)-C(21)=C(22)-C(21)=C(22) conduit [[66 66conduit–6868].]. [66–68]. 2.5. CO-Release Reactivity of Other Metal Flavonolato Complexes 2.5. CO-Release2.5. CO Reactivity-Release Reactivity of Other Metalof Other Flavonolato Metal Flavonolato Complexes Complexes Simple metal flavonolato complexes that lack structural relevance to quercetin dioxygenase Simple metalSimple flavonolato metal flavonolato complexes complexes that lack thatstructural lack structural relevance relevance to quercetin to quercetin dioxygenase dioxygenase but but but exhibit CO release reactivity have also been reported by Speier and co-workers [56,69–71]. exhibit COexhibit release CO reactivityrelease reactivity have also have been also reported been reported by Speier by and Speier co-workers and co- workers[56,69–71]. [56,69 When–71]. When When heated in DMF at 95 ◦C, bis- and tris-flavonolato complexes such as Mn(fla) (py) and heated in DMF at 95 °C, bis- and tris-flavonolato complexes such as 2 Mn(fla)2 2(py)2 and heated in DMF at 95 °C, bis- and tris-flavonolato complexes such as Mn(fla)2(py)2 and Fe(40-MeO-fla) (Figure3) release CO in near quantitative yields (80–90%) resulting in the Fe(4′-MeO3 -fla)3 (Figure 3) release CO in near quantitative yields (80–90%) resulting in the formation Fe(4′-MeO-fla)3 (Figure 3) release CO in near quantitative yields (80–90%) resulting in the formation formation of the corresponding O-benzosalicylate complexes. The rate law for these reactions is of the correspondingof the corresponding O-benzosalicylate O-benzosalicylate complexes. complexes. The rate The law rate for law these for reactions these reactions is – is – d[complex]/dt = k[complex][O2], with the rate constant for the Mn(II) derivative being approximately d[complex]/dt = k[complex][O2], with the rate constant for the Mn(II) derivative being approximately −1 −1 6-fold larger6-fold than larger that than of the that Fe of(II) the complex Fe(II) complex(0.50 and (0.50 0.08 andM−1 ·0.08s−1, respectively M ·s , respectively) at 100 °)C at in 100 DMF. °C in DMF.

Molecules 2019, 24, 1252 10 of 26

–d[complex]/dt = k[complex][O2], with the rate constant for the Mn(II) derivative being approximately Molecules 2019, 24, x FOR PEER REVIEW 10 of 26 6-foldMolecules larger 2019, 24 than, x FOR that PEER ofthe REVIEW Fe(II) complex (0.50 and 0.08 M−1·s−1, respectively) at 100 ◦C in DMF.10 of 26

FigureFigureFigure 3.3. 3. MononuclearMononuclearMononuclear MnMn Mn(II)(II)(II) andand and FeFe Fe(III)(III)(III) flavonolatoflavonolato ﬂavonolato complexescomplexes that that undergo undergo COCO releaserelease in inin DMF DMFDMF [ [69].69[69].].

AnionAnion effects effects on on the the O O2 reactivityreactivity of ofFe(III) Fe(III) flavonolato flavonolato complexes complexes have have been been reported reported [72]. The [72]. Anion effects on the O2 reactivity2 of Fe(III) flavonolato complexes have been reported [72]. The Fe(III)TheFe(III) Fe(III) flavonolatoflavonolato flavonolato complexcomplex complex Fe(fla)(salen)Fe(fla)(salen) Fe(fla)(salen) (fla(fla (fla == anionanion = anion ofof of 33-- 3-hydroxyflavone,hydroxyflavone,hydroxyflavone, SchemeScheme Scheme 10) 10)10 ) undergoes undergoes ◦ reactionreaction with with O O2 atat 100 100–125–125 °CC in in DMF DMF to to give give an an O--benzosalicylatebenzosalicylate complex complex and and CO CO [72]. [72]. The The rate rate reaction with O2 at 100–125 °C in DMF to give an O-benzosalicylate complex and CO [72]. The rate −1 −1 lawlaw for for this this reaction reaction is is–d[Fe(fla)(salen)]/d –d[Fe(fla)(salen)]/dt = tk[Fe(fla)(salen)][O= k[Fe(fla)(salen)][O2] with] withk = 2.07k = ± 2.070.12 ±M−0.121·s−1, MΔH‡ ·=s 76 , law for this reaction is –d[Fe(fla)(salen)]/dt = k[Fe(fla)(salen)][O2] with2 k = 2.07 ± 0.12 M−1·s−1, ΔH‡ = 76 ‡ ‡ ‡ − · kJ/molkJ/mol∆H = 76andand kJ/mol ΔΔSS‡ == and−−9494 J/mol∆J/molS =··KK at94at 373.16373.16 J/mol K.K.K atIncorporationIncorporation 373.16 K. Incorporation ofof anan electronelectron of-- andonatingdonating electron-donating groupgroup atat thethe group parapara positionatposition the para ofof position thethe flavonolflavonol of the resuresu flavonolltslts inin aa results ~2.5~2.5--foldfold in a increaseincrease ~2.5-fold inin increase thethe raterate in ofof the reaction.reaction. rate of A reaction.A significantsignificant A significant 171171--foldfold rate171-foldrate enhancement enhancement rate enhancement is is produced produced is produced upon upon upon addition addition addition of of a a of bulky bulky a bulky carboxylate carboxylate carboxylate anion anion anion (triphenyl (triphenyl (triphenyl acetate, acetate, acetate, − PhPh3CCOOCCOO−, Scheme, Scheme 10). 10 ).The The coordination coordination of of this this anion anion to to the the Fe Fe(III)(III) center center is is proposed proposed t too induce induce a a Ph3CCOO−, Scheme 10). The coordination of this anion to the Fe(III) center is proposed to induce a shiftshift in in the the flavonolato flavonolatoflavonolato coordination coordination mode mode from from bidentatebidentate to to monodentate.monodentate. The The enhanced enhanced electron electron densitydensity within within within the the the flavonolato flavonolato flavonolato ligand ligand ligand is suggested is is suggested suggested to be responsible to to be be responsible responsible for the dramatic for for the the rate dramatic dramatic enhancement. rate rate − enhancement.Theenhancement. overall rate law The The for overall overall the anion-containing rate rate law law reaction for for the the is Rate anion anion = k--[Fe(fla)(salen)][Ocontainingcontaining reactionreaction2][Ph 3CCOO is is Rate Rate] with = = ‡ ‡ k∆[Fe(fla)(salen)][OH = 35 kJ/mol2][Ph and3∆CCOOS = −] 120with J/mol ΔH‡ =· K35 at kJ/mol 313.16 and K. Δ ThisS‡ = lower−120 J/mol activation·K at 313.16 barrier K. resultsThis lower in a k[Fe(fla)(salen)][O2][Ph3CCOO−] with ΔH‡ = 35 kJ/mol and ΔS‡ = −120 J/mol·K at 313.16 K. This lower activationdioxygenase-typeactivation barrier barrier CO results results release in in reactiona a dioxygenase dioxygenase that can--typetype proceed CO CO at release release room reactiontemperature. reaction that that can can proceed proceed at at room room temperature.temperature.

Scheme 10. Effect of bulky carboxylate ligation on the dioxygenase-type CO release reactivity of an SchemeScheme 10.10. EffectEffect ofof bulkybulky carboxylatecarboxylate ligationligation onon thethe dioxygenasedioxygenase--typetype COCO releaserelease reactivityreactivity ofof anan Fe(III) flavonolato complex [72]. Fe(III)Fe(III) flavonolatoflavonolato complexcomplex [72].[72]. 2.6. Summary 2.6.2.6. Summary.Summary. Studies of quercetin dioxygenases and model systems for these enzymes have provided insight StudiesStudies ofof quercetinquercetin dioxygenasesdioxygenases andand modelmodel systemssystems forfor thesethese enzymesenzymes havehave providedprovided insightinsight into the factors that influence the reactivity of metal-flavonolato species with O2 in dioxygenase-type into the factors that influence the reactivity of metal-flavonolato species with O2 in dioxygenase-type CO-releasinginto the factors reactions. that influence In modelthe reactivity complexes of metal containing-flavonolato a flavonolato species with ligand O2 in dioxygenase coordinated-type in a CO-releasing reactions. In model complexes containing a flavonolato ligand coordinated in a bidentateCO-releasing manner reactions. via a five-membered In model complexes chelate containing ring, high a temperatures flavonolato ligandare typically coordinated required in to a bidentatebidentate manner manner via via a a five five--memberedmembered chelate chelate ring, ring, high high temperatures temperatures are are typically typically required required t too induce O2-dependent CO release reactivity. Key to increasing this reactivity is to enhance the electron induce O2-dependent CO release reactivity. Key to increasing this reactivity is to enhance the induce O2-dependent CO release reactivity. Key to increasing this reactivity is to enhance the density within the flavonolato moiety, as this lowers the activation barrier for electron transfer for O2. electronelectron density density within within the the flavonolato flavonolato moiety, moiety, as as this this lowers lowers the the activation activation barrier barrier for for electron electron transfer for O2. Enhanced electron density can be achieved by adding electron-donating substituents transfer for O2. Enhanced electron density can be achieved by adding electron-donating substituents onon thethe flavonolatoflavonolato ligand,ligand, increasingincreasing thethe electronicelectronic donordonor propertiesproperties ofof aa supportingsupporting chelatechelate ligand,ligand, oror byby introducingintroducing additionaladditional bulkybulky anionicanionic donorsdonors thatthat viavia coordinationcoordination toto thethe metalmetal centercenter induceinduce aa shiftshift to to mono monodentatedentate metal metal--flavonolatoflavonolato coordination. coordination. To To date, date, the the systems systems that that exhibit exhibit reactivity reactivity

Molecules 2019, 24, 1252 11 of 26

Enhanced electron density can be achieved by adding electron-donating substituents on the flavonolato ligand,Molecules increasing2019, 24, x FOR the PEER electronic REVIEW donor properties of a supporting chelate ligand, or by introducing11 of 26 additional bulky anionic donors that via coordination to the metal center induce a shift to monodentate closest to room temperature involve monodentate flavonolato coordination similar to that proposed metal-flavonolato coordination. To date, the systems that exhibit reactivity closest to room temperature in the enzymatic reaction pathways. In organic solvents, few of these complexes have been involve monodentate flavonolato coordination similar to that proposed in the enzymatic reaction evaluated in terms of quantification of the amount of CO released. Additionally, 3d metal pathways. In organic solvents, few of these complexes have been evaluated in terms of quantification flavonolato complexes have not been evaluated as CO-releasing molecules in aqueous of the amount of CO released. Additionally, 3d metal flavonolato complexes have not been evaluated environments. Under such conditions, ligand exchange with water would likely influence the CO as CO-releasing molecules in aqueous environments. Under such conditions, ligand exchange with release reactivity. water would likely influence the CO release reactivity.

3.3. CO Production from 3-Hydroxy-4-oxoquinolines3-Hydroxy-4-oxoquinolines viavia Cofactor-FreeCofactor-free Enzyme-CatalyzedEnzyme-catalyzed Reactions Reactionsand in Metal and-c inontaining Metal-Containing Synthetic Systems Synthetic Systems

3.1.3.1. Enzyme-CatalyzedEnzyme-catalyzed Reactions Reactions. TheThe cofactor-free cofactor-free bacterial bacterial 1H-3-hydroxy-4-oxoquinaldine 1H-3-hydroxy-4-oxoquinal 2,4-dioxygenasedine 2,4-dioxygenase (Hod) from Arthrobacter (Hod) from ilicis Rü61aArthrobacter and 1H ilicis-3-hydroxy-4-oxoquinoline Rü61a and 1H-3-hydroxy 2,4-dioxygenase-4-oxoquinoline (Qdo) 2,4-dioxygenase from Pseudomonas (Qdo) putida from33/1 Pseudomonas catalyze theputida insertion 33/1 catalyze of dioxygen the at insertion C2 and ofC4 dioxygen of the 3-hydroxy-4-oxoquinoline at C2 and C4 of the substrates 3-hydroxy- resulting4-oxoquinoline in the releasesubstrates in CO resulting (Scheme in 11 the(top)) release [73– 77 in ]. CO In terms(Scheme of sequence, 11(top)) [73 these–77]. enzymes In terms are of rare sequence, examples these of dioxygenasesenzymes are rare that examples belong to of the dioxygenasesα/β hydrolase that belong family. to Hod the α/β and hydrolase Qdo do not family. contain Hod a and cofactor Qdo ordo metalnot contain ion but a cofactor have a catalyticallyor metal ion requiredbut have histidinea catalytically residue required (His251 histiindine Hod; residue His244 (Hisin Qdo).251 in TheHod; enzyme-bound His244 in Qdo). monoanionThe enzyme of-bound 1H-3-hydroxy-4-oxoquinaldine monoanion of 1H-3-hydroxy is- proposed4-oxoquinaldine to undergo is proposed single 3 electronto undergo transfer singleto electron O2 in thetransfer rate-determining to O2 in the rate step-determining to produce step a triplet to produce superoxide a triplet complex superoxide ( I1, 3 1 1 Schemecomplex 11 ( (bottom)).I1, Scheme An11(bottom)). internal electron An internal transfer electron leads transfer to a closed-shell leads to singleta closed ( -Ishell1). Attack singlet of ( theI1). peroxideAttack of terminal the peroxide oxygen terminal at C4 followed oxygen at by C4 formation followed of by a peroxide formation bridge of a peroxide between C2bridge and C4between leads toC2 the and cyclic C4 leads species to the from cyclic which species CO from extrusion which canCO occurextrusion to give can occurN-acetylanthranilate to give N-acetylanthra anion.nilate This 18 18 reactionanion. This pathway reaction is consistent pathway with is consistentO2 labeling with studies, O2 labeling which showed studies, that which the catalytic showed reaction that the of Hodcatalytic results reaction in incorporation of Hod results of two in labeledincorporation oxygen of atoms. two labeled It should oxygen be noted atoms. that Itthe should substrates be noted for Hodthat the and substrates Qdo exhibit for similarHod and reactivity Qdo exhibit under similar basic, react aerobicivity conditions under basic, (vide aerobic infra). conditions The role of (vide the enzymeinfra). The in these role systems of the mayenzyme be to in provide these systemsa low-dielectric may be reaction to provide environment a low-dielectric that is optimal reaction for positioningenvironment the that substrate is optimal and for electronpositioning transfer the substrate to O2 and and C-O for bond electron formation. transfer to O2 and C-O bond formation.

SchemeScheme 11.11. (top)(top) ReactionReaction catalyzedcatalyzed byby 11HH-3-hydroxy-4-oxoquinaldine-3-hydroxy-4-oxoquinaldine 2,4-dioxygenase2,4-dioxygenase (R(R == -CH-CH33; Hod)Hod) andand 11HH-3-hydroxy-4-oxoquinoline-3-hydroxy-4-oxoquinoline 2,4-dioxygenase2,4-dioxygenase (R(R == -H;-H; Qdo).Qdo). (bottom) Proposed reactionreaction pathwaypathway ofof HodHod basedbased onon kinetic,kinetic, computationalcomputational andand spectroscopicspectroscopic studies.studies.

Molecules 2019, 24, 1252 12 of 26

3.2. Synthetic Systems While not structurally relevant to the active site chemistry of Hod, copper complexes containing a 1H-2-phenyl-3-hydroxy-4-oxoquinolinate ligand have been shown to undergo O2-dependent CO releaseMolecules [201978]., 24 As, x shown FOR PEER in REVIEW Scheme 12, similar to the reactivity of Cu(II) ﬂavonolato complexes,12 these of 26 complexes undergo stoichiometric CO release from each 3-hydroxy-4-oxoquinoline ligand upon heatingcomplexes inDMF. undergo stoichiometric CO release from each 3-hydroxy-4-oxoquinoline ligand upon heating in DMF.

Scheme 12.12. ReactionsReactions ofof aa Cu(II)Cu(II) bisbis 11HH-2-phenyl-3-hydroxy-4-oxoquinolinate-2-phenyl-3-hydroxy-4-oxoquinolinate complexcomplex withwith OO22 toto Scheme 12. Reactions of a Cu(II) bis 1H-2-phenyl-3-hydroxy-4-oxoquinolinate complex with O2 to release CO [78]. release CO [78].

A non-heme iron catalyst has been reported by Speier, et al. to mediate the O2-dependent cleavage A non-heme iron catalyst has been reported by Speier, et al. to mediate the O2-dependent of 3-hydroxyflavoneA non-heme iron and catalyst 1-H-2-phenyl-3-hydroxy-4-oxoquinoline has been reported by Speier, et al. to produce to mediate CO the [79 ]. O At2-dependent 100 ◦C in cleavage of 3-hydroxyflavone and 1-H-2-phenyl-3-hydroxy-4-oxoquinoline to produce CO [79]. At DMFcleavage and of with 3-hydroxyflavone a ratio of 1:25 between and 1- theH-2 complex,-phenyl-3 [Fe(III)(O-bs)(salen)]-hydroxy-4-oxoquinoline (Scheme to 13produce), and the CO substrate, [79]. At 100 °C in DMF and with a ratio of 1:25 between the complex, [Fe(III)(O-bs)(salen)] (Scheme 13), and catalytic100 °C in dioxygenation DMF and with occurs a ratio to of give 1:25 the between products, theO complex,-benzoylsalicylic [Fe(III)(O acid-bs)(salen)] and N-benzoylanthranilic (Scheme 13), and the substrate, catalytic dioxygenation occurs to give the products, O-benzoylsalicylic acid and acidthe substrate, in 78% and catalytic 64% yield, dioxygenation respectively, occurswith concomitant to give the release products, of CO. O- Initialbenzoylsalicylic rate studies acid showed and N-benzoylanthranilic acid in 78% and 64% yield, respectively, with concomitant release of CO. thatN-benzoylanthranilic 3-hydroxyflavone acid is three-times in 78% and more 64% reactive yield, than respectively, 2-phenyl-3-hydroxy-4(1 with concomitantH)-quinoline release toward of CO. Initial rate studies showed that 3-hydroxyflavone is three-times more reactive than O2. As expected, increasing the basicity of the flavonol increases the rate of the O2 reaction. These 2-phenyl-3-hydroxy-4(1H)-quinoline toward O2. As expected, increasing the basicity of the flavonol reactions2-phenyl-3 proceed-hydroxy by-4(1 singleH)-quinoline electron toward transfer O (SET)2. As expected, wherein increasing the metal-coordinated the basicity of deprotonated the flavonol increases the rate of the O2 reaction. These reactions proceed by single electron transfer (SET) substrateincreases provides the rate ofan the electron O2 reaction. for the reduction These reactions of O . proceedSuperoxide by formation single electron was detected transfer using (SET) 2 2 wherein the metal-coordinated deprotonated substrate provides an electron for the reduction of O2. nitroblue tetrazolium (NBT) as a scavenger to produce formazan. Proceeding through the reaction Superoxide formation was detected using nitroblue tetrazolium (NBT) as a scavenger to produce pathway as shown in Scheme 13, an endoperoxide intermediate is formed that undergoes C-C and formazan. Proceeding through the reaction pathway as shown in Scheme 13, an endoperoxide O-O bond cleavage to release CO and generate an enzyme-type organic byproduct. The formation of intermediate is formed that undergoes C-C and O-O bond cleavage to release CO and generate an O-benzoylsalicylate or N-anthranilate as products enhances the rate of reaction through coordination enzyme-type organic byproduct. The formation of O-benzoylsalicylate or N-anthranilate as products to the Fe(III) center as previously described. enhances the rate of reaction through coordination to the Fe(III) center as previously described.

Scheme 13. Oxygenation of 3-hydroxyﬂavone and 3-hydroxy-4-oxoquinoline derivatives with the Scheme 13. Oxygenation of 3-hydroxyflavone and 3-hydroxy-4-oxoquinoline derivatives with the release of CO catalyzed by [Fe(III)(O-bs)(salen)] [79]. release of CO catalyzed by [Fe(III)(O-bs)(salen)] [79].

Molecules 2019, 24, 1252 13 of 26

3.3. Summary Overall these combined studies indicate that 3-hydroxy-4-oxoquinoline derivatives undergo enzyme-catalyzed or metal-promoted reactions to release N-benzoylanthranilic acid and an equivalent of CO. This clean reactivity suggests that 3-hydroxy-4-oxoquinolines may also be useful as a CO-releasing motif. Notably, the slower rate of CO release of 2-phenyl-3-hydroxy-4(1H)-quinoline derivatives versus 3-hydroxyﬂavones upon reaction with O2 indicates that the heteroatom within the pyrone ring is a key factor in determining the rate of CO release.

4. CO Production from 3-Hydroxyﬂavones and 3-Hydroxy-4-oxoquinolines via Base-Catalyzed and Non-Redox Metal Assisted Reactions

4.1. 3-Hydroxyflavones Speier, et al. have previously summarized base-catalyzed flavonol oxygenation that results in CO release in protic solvents as well as non-redox metal-assisted CO release reactions in a variety of solvents [56,80–82]. Briefly, flavonol oxygenation under these conditions (40–90 ◦C) proceeds via one of two mechanistic pathways: single electron transfer (SET) (Scheme 14(Path a)) or a one-step electrophilic reaction of triplet oxygen with the flavonolate ion (Scheme 14(Path b)). The former was found to occur in aprotic solvents via SET from the flavonolate ion to triplet O2, yielding a flavonoxy radical and superoxide ion, which subsequently undergo fast radical-radical coupling to give a 2-hydroxyperoxyflavan-3,4-dionate intermediate. The resulting species then undergoes an intramolecular nucleophilic attack at C4 leading to the formation of an unstable endoperoxide that decomposes into depside (O-benzoylsalicylate) and CO [81]. In the case of the electrophilic reaction between oxygen and the flavonolato ion, a detailed mechanistic study was carried out with 4’-substituted flavonol derivatives in 50% DMSO-H2O in the pH range 6.4–10.8 to map out the reaction pathway [82]. This reaction showed specific base catalysis. A linear Hammett plot yielded a negative reaction constant (ρ = −0.50) implying that a higher electron density on the flavonolate ion makes it more nucleophilic thus facilitating electrophilic attack by O2. Both oxygenation pathways yield stoichiometric formation of CO and depside and/or its hydrolysis byproducts (salicylic acid and benzoic acid). Farmer, et al. have reported the base-catalyzed reaction of HNO with quercetin [83]. Above pH 7, this reaction occurs to give 3,4-dihydroxybenzonitrile, a cleavage product indicating that the regioselectivity observed in the reaction catalyzed by the Mn-containing QDO from Bacillus subtilis (Scheme4) is maintained in the base-catalyzed reaction. Kinetic and thermodynamic studies by Kumar and Farmer have provided evidence that the base-catalyzed reaction between HNO and quercetin occurs via single electron transfer (SET) [83]. It is notable that this reaction occurs orders of magnitude faster than the corresponding dioxygenation. This has been attributed to the difference in driving force for SET in the rate-determining step.

4.2. 3-Hydroxy-4-oxoquinolines For 3-hydroxy-4-oxoquinoline derivatives, a mixture of products resulting from endoperoxide or 1,2-dioxetane species forms, with the ratio depending on the solvent. CO release only occurs via the reaction involving the endoperoxide, with the 1,2-dioxetane undergoing C-C bond cleavage to give a phenylglyoxalic acid derivative (Scheme 14). A long-lived radical is observed in the reaction of 1-H-2-phenyl-3-hydroxy-4-oxoquinoline with O2 providing evidence for a single electron transfer pathway in the oxygen activation step [84]. Molecules 2019, 24, 1252 14 of 26 Molecules 2019, 24, x FOR PEER REVIEW 14 of 26

Scheme 14. 14.Reaction Reaction pathways pathways leading leading to CO release to CO in base-catalyzed release in base and-catalyzed non-redox ametal-promotednd non-redox dioxygenationmetal-promoted reactions dioxygenation of 3-hydroxyflavones reactions of 3-hydroxyflavones and 3-hydroxy-4-oxoquinolines. and 3-hydroxy-4-oxoquinolines. 4.3. Summary 4.3. Summary Overall, the 3-hydroxyflavone derivatives exhibit cleaner reactivity under base-catalyzed Overall, the 3-hydroxyflavone derivatives exhibit cleaner reactivity under base-catalyzed conditions, whereas 3-hydroxy-4-oxoquinoline derivatives exhibit competing formation of 1,2-dioxetane conditions, whereas 3-hydroxy-4-oxoquinoline derivatives exhibit competing formation of species. This suggests that 3-hydroxyflavones may be more reliable CO release agents under 1,2-dioxetane species. This suggests that 3-hydroxyflavones may be more reliable CO release agents biological conditions. under biological conditions. 5. CO Production from 3-Hydroxyflavones via Photochemical Reactions 5. CO Production from 3-Hydroxyflavones via Photochemical Reactions Photooxygenation of flavonols is another reaction pathway that results in CO release. It is Photooxygenation of flavonols is another reaction pathway that results in CO release. It is known that unsubstituted 3-hydroxyflavone (3-HflH) will undergo incorporation of both oxygen known that unsubstituted 3-hydroxyflavone (3-HflH) will undergo incorporation of both oxygen atoms of O2 and expulsion of CO in the presence of a photosensitizer, or via direct illumination atoms of O2 and expulsion of CO in the presence of a photosensitizer, or via direct illumination using using UV light (Scheme 15 (top)) [85,86]. The latter is suggested to proceed via the reaction of UV light (Scheme 15 (top)) [85,86]. The latter is suggested to proceed via the reaction of a tautomeric a tautomeric triplet state (formed via excited state intramolecular proton transfer, ESIPT) with 3 triplet state (formed3 via excited state intramolecular proton transfer, ESIPT) with ground state O2 to ground state O2 to give a 5-membered cyclic endoperoxide intermediate from which CO extrusion give a 5-membered cyclic endoperoxide intermediate from which CO extrusion occurs [87–89]. A occurs [87–89]. A similar reaction pathway has been recently proposed for the photooxygenation similar reaction pathway has been recently proposed for the photooxygenation of of 40-diethylamino-3-hydroxyflavone [90]. Notably, 3-hydroxyflavone is also known to undergo 4′-diethylamino-3-hydroxyflavone [90]. Notably, 3-hydroxyflavone is also known to undergo photoinduced rearrangement in the absence of O2 (Scheme 15 (bottom)) in apolar, polar aprotic and photoinduced rearrangement in the absence of O2 (Scheme 15 (bottom)) in apolar, polar aprotic and polar protic solvents, to give a 3-hydroxy-3-aryl-indane-1,2-dione product [91–94]. In some cases, polar protic solvents, to give a 3-hydroxy-3-aryl-indane-1,2-dione product [91–94]. In some cases, the the indane-1,2-diones releases CO to form a 3-arylphthalide (Scheme 15). indane-1,2-diones releases CO to form a 3-arylphthalide (Scheme 15). Our laboratory has recently demonstrated that divalent metal complexes of the general formula [(L)Zn(3-Hfl)]ClO4, containing a coordinated 3-hydroxyflavonolato ligand and supported by a chelating nitrogen donor ligand (L), undergo clean UV- or visible light-induced reactivity in the presence of O2 to produce one equivalent of CO (determined by GC head space analysis) and a metal carboxylate (depside) complex (Scheme 16)[95–99]. The reaction quantum yield (Table1) for CO release from these complexes depends on the ligand secondary environment of the metal complex or solvent conditions, and on the divalent metal ion present. Divalent zinc 3-hydroxyflavonolato complexes of a variety of supporting chelate ligands exhibit reaction quantum yields for CO release that can be tuned from ~0.0003 to 0.012 [95,98,99]. Encapsulation of the Zn(II)-flavonolato moiety within a hydrophobic microenvironment produced the highest reaction quantum yield whereas positioning the Zn(II) 3-hydroxyflavonolato unit within a hydrogen bond donor pocket produced the least reactive complex. The latter may be explained in that hydrogen-bonding involving either rigid donors in the Scheme 15. UV-light induced reactivity of 3-hydroxyflavone in the presence (top) and absence (bottom) of oxygen.

Molecules 2019, 24, x FOR PEER REVIEW 14 of 26

Scheme 14. Reaction pathways leading to CO release in base-catalyzed and non-redox metal-promoted dioxygenation reactions of 3-hydroxyflavones and 3-hydroxy-4-oxoquinolines.

4.3.Molecules Summary 2019, 24 , x FOR PEER REVIEW 15 of 26

Overall,Our laboratory the 3- hashydroxyflavone recently demonstrated derivatives that exhibit divalent cleaner metal complexes reactivity underof the general base-catalyzed formula conditions,[(L)Zn(3-Hfl)]ClO whereas4, containing 3-hydroxy a -4 coordinated-oxoquinoline 3- hydrderioxyflavonolatvatives exhibito ligand competing and supported formation by of a 1,2chelating-dioxetane nitrogen species. donor This ligandsuggests (L) that, undergo 3-hydroxyflavones clean UV- or may visible be more light -reliableinduced CO reactivity release agents in the underpresence biological of O2 to conditions.produce one equivalent of CO (determined by GC head space analysis) and a metal carboxylate (depside) complex (Scheme 16) [95–99]. The reaction quantum yield (Table 1) for CO 5.release CO Production from these fromcomplexes 3-Hydroxyflavones depends on the via ligand Photochemical secondary environmentReactions of the metal complex or solventPhotooxygenation conditions, and of on flavonols the divalent is another metal reaction ion present. pathway Divalent that results zinc 3 - inhydroxyf CO release.lavonolato It is Molecules 2019, 24, 1252 15 of 26 knowncomplexes that of unsubstituted a variety of supporting 3-hydroxyflavone chelate ligands (3-HflH) exhibit will undergoreaction quantum incorporation yields of for both CO oxygenrelease atomsthat can of Obe2 andtuned expulsion from ~0.0003 of CO toin the0.012 presence [95,98,99]. of a Encapsulationphotosensitizer, of or the via Zn(II) direct-flavonolato illumination moiety using within a hydrophobic microenvironment produced the highest reaction quantum yield whereas UVsecondary light (Scheme environment, 15 (top)) or a[85,86]. hydrogen The bondlatter donoris suggested solvent to (e.g., proceed H2O), via provides the reaction additional of a tautomeric pathways positioning the Zn(II) 3-hydroxyflavonolato unit within a hydrogen bond donor pocket produced3 tripletfor non-radiative state (formed decay via excited thus leading state intramolecular to excited state proton quenching transfer, and ESIPT) a less efficientwith ground CO releasestate O [298 to]. giveSu,the etleast a al.5- membered recentlyreactive reportedcomplex. cyclic endoperoxide aThe series latter of zincmay intermediate flavonolato be explained complexes from in that which hydrogen supported CO extrusion-bonding by tetradentate occurs involving [87 tripodal– 89].either A similarnitrogenrigid donors donor reaction in ligandsthe secondary pathway that are environment, similarhas been to those or recently reporteda hydrogen proposed by ourbond laboratory donor for solvent the and releasephotooxygenation (e.g., H one2O), equivalent provides of 4′ofadditional-diethylamino CO upon pathways visible-3-hydroxyflavone light for illumination non-radiative [90]. [100 Notably, decay]. We thus note 3- hydroxyflavone leading that Protti, to excited et al. is previously state also quenching known reported to and undergo that a less the efficient CO release [98]. Su, et al. recently reported a series of zinc flavonolat+ o complexes supported presencephotoinduced of O2 rearrenhancedangement the photodecomposition in the absence of O of2 (Scheme simple[Zn(3-Hfl)] 15 (bottom))species in apolar, but didpolar not aprotic report and CO productionpolarby tetradentate protic [solvents,101 ].tripodal to give nitrogen a 3-hydroxy donor ligands-3-aryl- indanethat are-1,2 similar-dione to product those reported [91–94]. Inby some our laboratory cases, the indaneand release-1,2-diones one equivalent releases CO of toCO form upon a 3 visible-arylphthalide light illumination (Scheme 15). [100]. We note that Protti, et al. previously reported that the presence of O2 enhanced the photodecomposition of simple [Zn(3-Hfl)]+ species but did not report CO production [101]. The coordination of the 3-hydroxyflavonolato ligand to heavy closed shell metal ions (CdII < HgII, and PbII) produced relatively high quantum yields (>0.20) for CO release [95,96]. Conversely, coordination of the 3-hydroxyflavonolato anion to Mn(II), Co(II), Ni(II), or Cu(II) yielded complexes with low CO release reaction quantum yields (typically ~0.005) due to quenching of the excited state by the open-shell first row metal ion [96]. Despite the low reactivity, these complexes can serve as catalysts for the UV light-induced oxidative degradation of 3-hydroxyflavone to give O-benzoylsalicylic acid and CO [96]. While multiple catalysts for the thermal O2-dependent degradation of 3-hydroxyflavone with CO release have been reported [79,102,103], to our knowledge this is the only light-driven catalytic reaction reported to date. We also note that a Ag(I) 3-hydroxyflavonolato complex, [Ag(3-Hfl)(PPh3)2], was recently reported to undergo visible Scheme 15. UV-light induced reactivity of 3-hydroxyflavone in the presence (top) and absence light-Schemeinduced 15. oxygenationUV-light induced to yieldreactivity a depside of 3-hydroxyflavone complex [104]. in the presence No reaction (top) andquantum absence yield(bottom) or CO (bottom)of oxygen. of oxygen. quantification was reported for this reaction.

Scheme 16. 16.UV- UV or- visible-light-inducedor visible-light-induced CO release CO release reactivity reactivity of divalent of metal divalent flavonolato metal flavonolato complexes. complexes. The coordination of the 3-hydroxyflavonolato ligand to heavy closed shell metal ions (CdII < HgII, and PbII) produced relatively high quantum yields (>0.20) for CO release [95,96]. Conversely, coordination of the 3-hydroxyflavonolato anion to Mn(II), Co(II), Ni(II), or Cu(II) yielded complexes with low CO release reaction quantum yields (typically ~0.005) due to quenching of the excited state by the open-shell first row metal ion [96]. Despite the low reactivity, these complexes can serve as catalysts for the UV light-induced oxidative degradation of 3-hydroxyflavone to give O-benzoylsalicylic acid and CO [96]. While multiple catalysts for the thermal O2-dependent degradation of 3-hydroxyflavone with CO release have been reported [79,102,103], to our knowledge this is the only light-driven catalytic reaction reported to date. We also note that a Ag(I) 3-hydroxyflavonolato complex, [Ag(3-Hfl)(PPh3)2], was recently reported to undergo visible light-induced oxygenation to yield a depside complex [104]. No reaction quantum yield or CO quantification was reported for this reaction. Molecules 2019, 24, x FOR PEER REVIEW 16 of 26

Table 1. Quantum yields of UV- or visible light-induced quantitative CO-releasing reactions of divalent metal 3-hydroxyflavonolato complexes.

Absorption Quantum Yield Reference Compound Molecules 2019, 24, 1252 Max. (nm) for CO Release 16 of 26

[(6-Ph2TPA)Zn(3-Hfl)]ClO4 420 0.09(1) a,c [63,95,97]

Table[(6-Ph 1.2QuantumTPA)Zn(3 yields-Hfl)]ClO of UV-4 or visible light-induced420 0.012(2) quantitative b,c; 0.006(1) CO-releasing b,d reactions[98 of] divalent

metal[(6-Ph 3-hydroxyﬂavonolato2TPA)Cd(3-Hfl)]ClO complexes.4 430 0.28(2) a,c [95,97]

[(6-Ph2CompoundTPA)Hg(3-Hfl)]ClO4 Absorption Max.415 (nm) Quantum0.31(2) Yield a,c for CO Release [95,97Reference] a,c [(6-Ph[(6-Ph2TPA)Zn(3-Hfl)]ClO2TPA)Mn(3-Hfl)]ClO4 4 420415 0.005 0.09(1) a,c [[9663], 95,97] b,c b,d [(6-Ph2TPA)Zn(3-Hfl)]ClO4 420 0.012(2) ; 0.006(1) [98] a,c a,c [(6-Ph[(6-Ph2TPA)Cd(3-Hfl)]ClO2TPA)Co(3-Hfl)]ClO4 4 430430 0.005 0.28(2) [96[95] ,97] a,c [(6-Ph2TPA)Hg(3-Hfl)]ClO4 415 0.31(2)a,c [95,97] [(6-Ph2TPA)Cu(3-Hfl)]ClO4 428 0.005 a,c [96] [(6-Ph2TPA)Mn(3-Hfl)]ClO4 415 0.005 [96] a,c [(6-Ph[(6-Ph2TPA)Co(3-Hfl)]ClO2TPA)Ni(3-Hfl)]ClO4 4 430415 0.008 0.005 a,c [96][ 96] a,c [(6-Ph2TPA)Cu(3-Hfl)]ClO4 428 0.005 [96] a,ca,c [(6-Ph[(6-Ph2TPA)Ni(3-Hfl)]ClO2TPA)Pb(3-Hfl)]ClO4 4 415406 0.21(6) 0.008 [97][ 96] [(6-Ph TPA)Pb(3-Hfl)]ClO 406 0.21(6) a,c [97] 2 4 b,c [(TPA)Zn(3-Hfl)]ClO4 415 0.006(1) b,c [98] [(TPA)Zn(3-Hfl)]ClO4 415 0.006(1) [98] b,c [(bnpapa)Zn(3-Hfl)]ClO[(bnpapa)Zn(3-Hfl)]ClO4 4 401401 0.00027(1)0.00027(1) b,c [98][ 98] b,c {[(bpy)Zn(3-Hfl)]2}(ClO4)2 414 0.004(1) [99] [Ru({[(bpy)Zn(3η6-p-cymene)Cl(3-Hfl)]-Hfl)]2}(ClO4)2 472414 0.004(1)0.001(1) b,c b,c [99[]105 ] a b c d [Ru(η6-p-cymene)Cl(3-Hfl)]300 nm; 419 nm;472 In CH3CN; 1:10.001(1) DMSO:H b,c2O. [105]

a 300 nm; b 419 nm; c In CH3CN; d 1:1 DMSO:H2O. We have also explored the visible light-induced CO release reactivity of the [Ru(η6-p-cymene) We have also explored the visible light-induced CO release reactivity of the (CH3CN)(3-Hfl)]OTf complex (Scheme 17)[105], as such complexes are promising for anticancer [Ru(η6-p-cymene)(CH3CN)(3-Hfl)]OTf complex (Scheme 17) [105], as such complexes are promising applications [106–108]. The combined results from product identification (ESI-MS), CO quantification, for anticancer applications [106–108]. The combined results from product identification (ESI-MS), and 18O-isotope labeling studies upon illumination with UV (300 nm) and visible (419 nm) light in CO quantification, and 18O-isotope labeling studies upon illumination with UV (300 nm) and visible CH3CN show that oxygenation of the flavonolato ligand occurs in a dioxygenase-type manner with (419 nm) light in CH3CN show that oxygenation of the flavonolato ligand occurs in a release of CO. However, the amount of free CO generated was found to depend on the wavelength dioxygenase-type manner with release of CO. However, the amount of free CO generated was found of illumination, with UV light leading to higher yield (0.70 eq CO). This is because the Ru(II) center to depend on the wavelength of illumination, with UV light leading to higher yield (0.70 eq CO). acts as a trap for the released CO and UV light enables photodissociation of the metal carbonyl moiety. This is because the Ru(II) center acts as a trap for the released CO and UV light enables The quantum yield for CO release from [Ru(η6-p-cymene)Cl(3-Hfl)] upon illumination at 419 nm is photodissociation of the metal carbonyl moiety. The quantum yield for CO release from 0.001(1) which is among the lowest values measured to date for divalent metal 3-hydroxyflavonolato [Ru(η6-p-cymene)Cl(3-Hfl)] upon illumination at 419 nm is 0.001(1) which is among the lowest complexes (Table1). values measured to date for divalent metal 3-hydroxyflavonolato complexes (Table 1).

II Scheme 17. Reactivity of Ru 3-hydroxyflavonolato complex with O2 upon illuination with visible Scheme 17. Reactivity of RuII 3-hydroxyflavonolato complex with O2 upon illuination with visible light (λill = 419 nm) [105]. light (λill = 419 nm) [105]. Farmer, et al. have recently synthesized and characterized a series of Ru(II) bipyridine-ligated Farmer, et al. have recently synthesized and characterized a series of Ru(II) bipyridine-ligated 3-hydroxyflavonolato complexes and investigated the mechanistic details of their light-driven 3-hydroxyflavonolato complexes and investigated the mechanistic details of their light-driven reactivity [109,110]. Illumination into distinct excitation bands at low temperature leads to products of reactivity [109,110]. Illumination into distinct excitation bands at low temperature leads to products two different reaction pathways (Scheme 18). Specifically, upon excitation at wavelengths longer than of two different reaction pathways (Scheme 18). Specifically, upon excitation at wavelengths longer 400 nm, the complexes undergo 1,2-addition of dioxygen, generating a 1,2-dioxetane intermediate than 400 nm, the complexes undergo 1,2-addition of dioxygen, generating a 1,2-dioxetane that results in the formation of a RuII-coordinated α-ketoacid complex. This complex is proposed to release CO gas to form a Ru(II) depside complex as determined by ESI-MS and deoxymyoglobin assay. Illumination of the same complex through a 310 nm notch pass optical filter, yields product mixtures derived from 1,3-addition of dioxygen, specifically depside coordinated to Ru(II) and CO. The authors suggest that the observed reactivity patterns are attributed to flavonol tautomeric biradicals [110]. Molecules 2019, 24, x FOR PEER REVIEW 17 of 26

intermediate that results in the formation of a RuII-coordinated α-ketoacid complex. This complex is proposed to release CO gas to form a Ru(II) depside complex as determined by ESI-MS and deoxymyoglobin assay. Illumination of the same complex through a 310 nm notch pass optical filter, yields product mixtures derived from 1,3-addition of dioxygen, specifically depside coordinated to Ru(II) and CO. The authors suggest that the observed reactivity patterns are attributed to flavonol tautomeric biradicals [110].

6. Visible Light-induced CO Release from Extended 3-Hydroxyflavone and 3-Hydroxybenzo[g]quinolone Frameworks We have prepared a series of new flavonols (1-4) with an additional fused aromatic ring (Scheme 19) to shift the absorption maximum further into the visible range. [111]. Compound 1, which ionizes partially in DMSO:PBS buffer at pH = 7.4 [112], has been extensively investigated in terms of its O2-dependent CO release reactivity. In a range of solvents (Table 2), 1 undergoes dioxygenase-type quantitative visible light-induced (λill = 419 or 488 nm) CO release, with the reaction quantum yield (0.006–0.010) depending on the solvent [111,112]. The product generated is an O-benzoylsalicylate derivative (5). Notably, compound 1 has been recently shown to be amenable to two-photon excitation at 800 nm to trigger CO release [113]. Prior to CO release, compound 1 can be visualized in cells via its green emission at ~580 nm (Scheme 19(b)). This emission, which is from a tautomeric species resulting from excited state intramolecular proton transfer (ESIPT), is lost upon release of CO, thus enabling tracking of the intracellular CO release process [111,114]. Compound 1 binds weakly to bovine serum albumin (Ka = 3.2 × 103 M−1) at the warfarin binding Site I [112]. This interaction with a protein reduces the quantum yield for CO release (Table 2). Compound 1 is minimally toxic in A549 (IC50 = 41 μM), HUVECs (IC50 = 82 μM), and RAW 264.7 cells (non toxic up to 100 μM) [111,115,116]. Functionalized derivatives of 1 have been used to probe mitochondrial Molecules 2019, 24, 1252 17 of 26 localization of CO release [115], thiol and H2S sensing [117,118] and H2O2 sensing [113] prior to light-triggered CO release.

II II Scheme 18.SchemePhotoreactivity 18. Photoreactivity of Ru 3-hydroxyﬂavonolato of Ru 3-hydroxyflavonol complexesato complexes with O2 [with109,110 O2 ].[109,110].

6. Visible Light-InducedThe diethylamino CO Release-substituted from Extended2 (Scheme 3-Hydroxyflavone 19(a); Table and 2) exhibits similar quantitative, 3-Hydroxybenzo[g]quinolone Frameworks O2-dependent visible light-induced (λill = 419 nm) CO release to that found for 1. Notably, the Wesulfur have-substituted prepared a 3series (Scheme of 19(a); new flavonolsTable 2) exhibits (1-4) with a much an additionalhigher quantum fused yield aromatic (0.426(3)) ring for CO (Schemerelease 19) to shiftin CH the3CN absorption with λill > maximum 546 nm under further O2 into [111]. the Compound visible range. 4 exhibits [111]. Compound quantitative1, O which2-dependent ionizes partiallyvisible light in DMSO:PBS-induced CO buffer release at pH under = 7.4 aerobic [112], hasconditions been extensively (λill > 546 investigated nm) but also in produces terms of 0.32(7) its O2-dependentequivalents CO of release CO under reactivity. anaerobic In a conditions. range of solvents Both reactions (Table2), of1 undergoes4 lead to a mixture dioxygenase-type of products, with quantitativethe anaerobic visible light-induced CO release reaction (λill = 419 appearing or 488 nm) to involve CO release, photoisomerization with the reaction akin quantum to that yieldobserved for (0.006–0.010)3-hydroxyflavone depending on (Scheme the solvent 15) [92 [111–94].,112 The]. The mechanism product generatedfor visible is-light an O induced-benzoylsalicylate CO release in 1–4 derivative (5). Notably, compound 1 has been recently shown to be amenable to two-photon excitation at 800 nm to trigger CO release [113]. Prior to CO release, compound 1 can be visualized in cells via its green emission at ~580 nm (Scheme 19b). This emission, which is from a tautomeric species resulting from excited state intramolecular proton transfer (ESIPT), is lost upon release of CO, thus enabling tracking of the intracellular CO release process [111,114]. Compound 1 binds weakly to 3 −1 bovine serum albumin (Ka = 3.2 × 10 M ) at the warfarin binding Site I [112]. This interaction with a protein reduces the quantum yield for CO release (Table2). Compound 1 is minimally toxic in A549 (IC50 = 41 µM), HUVECs (IC50 = 82 µM), and RAW 264.7 cells (non toxic up to 100 µM) [111,115,116]. Functionalized derivatives of 1 have been used to probe mitochondrial localization of CO release [115], thiol and H2S sensing [117,118] and H2O2 sensing [113] prior to light-triggered CO release. The diethylamino-substituted 2 (Scheme 19a; Table2) exhibits similar quantitative, O 2-dependent visible light-induced (λill = 419 nm) CO release to that found for 1. Notably, the sulfur-substituted 3 (Scheme 19a; Table2) exhibits a much higher quantum yield (0.426(3)) for CO release in CH 3CN with λill > 546 nm under O2 [111]. Compound 4 exhibits quantitative O2-dependent visible light-induced CO release under aerobic conditions (λill > 546 nm) but also produces 0.32(7) equivalents of CO under anaerobic conditions. Both reactions of 4 lead to a mixture of products, with the anaerobic CO release reaction appearing to involve photoisomerization akin to that observed for 3-hydroxyflavone (Scheme 15)[92–94]. The mechanism for visible-light induced CO release in 1–4 has not been computationally examined but may involve the reactivity of a triplet phototautomer formed via 3 ESIPT reacting with O2 [90,114]. Molecules 2019, 24, x FOR PEER REVIEW 18 of 26 Molecules 2019, 24, x FOR PEER REVIEW 18 of 26 hasMolecules not been2019 ,computationally24, 1252 examined but may involve the reactivity of a triplet phototautomer18 of 26 formedhas not via been ESIPT computationally reacting with 3 Oexamined2 [90,114]. but may involve the reactivity of a triplet phototautomer formed via ESIPT reacting with 3O2 [90,114].

SchemeScheme 19. 19. (a(a) ) Extended Extended 3-hydroxyflavone3-hydroxyflavone derivatives derivatives (1 – (14–).4 ().b , (b left), left) Visible Visible light-induced light-induced CO release CO releasefromScheme 1from.( b 19., 1 right). ((ba,) right) Individual Extended Individual fluorescence3-hydroxyflavone fluorescence microscopy microscopy derivatives images images ( of1– HUVECs4). (ofb, HUVECs left) incubated Visible incubated with light1-induced forwith 4 h.1 for Row CO 4 h.1:release CellsRow from1: exposed Cells 1. (exposedb to, right)1 for4 toIndividual h. 1Row for 4 2: h. Cellsfluorescence Row from 2: Cells first microscopy from row illuminatedfirst rowimages illuminated (488 of HUVECs nm light, (488 incubated withnm light, a light with with density 1 afor lightof4 h. 42,620density Row lx).1: of Cells Cells42,620 exposed were lx). alsoCells to co-stained 1were for 4also h. Row withco-stained 2: Hoechst Cells with from 33342 Hoechst first nuclear row 33342 illuminated dye nuclear (blue) to dye(488 assess (blue) nm celllight, to integrity.assess with a cellSizelight integrity.of density bar =Size 50of µ 42,620ofm. bar =lx). 50 Cells μm. were also co-stained with Hoechst 33342 nuclear dye (blue) to assess cell integrity. Size of bar = 50 μm. ZincZinc complexation complexation of of1–14– provided4 provided the the first first examples examples of of divalent divalent metal metal flavonolato flavonolato complexes complexes (6(–613–13, SchemeZinc, Scheme complexation 20) 20that)that exhibit of exhibit 1 –O42 -provideddependent, O2-dependent, the quantitative first examples quantitative visible of divalent visiblelight-induced light-inducedmetal solidflavonolato-statesolid-state CO complexes releaseCO [119].release(6–13 Similar, Scheme [119]. to Similar 20) the that solution to exhibit the solution reactivity O2-dependent, reactivity of 1– 3quantitative of, these1–3, these zinc visible zinc complexes complexes light - undergoinduced undergo solid reaction reaction-state CO either either release in in solutsolution[119].ion Similaror or inin the the to solid thesolid state solution state to to produce reactivity produce Zn(II)Zn(II) of 1O–O3-benzoylsalicylate, -benzoylsalicylate these zinc complexes complexes complexes undergo (Scheme (Scheme reaction 20) 20 and ) either and one one in equivalentequivalentsolution orof of inCO COthe per persolid flavonolato flavonolato state to produceligand. ligand. The Zn(II) The quantum quantum O-benzoylsalicylate yields yields for for the the complexesreactions reactions of (Scheme of6–69– are9 are 20)enhanced enhanced and one relativerelativeequivalent to to the the of free freeCO flavonols per flavonols flavonolato (Table (Table 2).2ligand.). Limited Limited The solubility quantum of yields thethe Zn(II)Zn(II) for the bis-flavonolatobis reactions-flavonolato of 6 complexescomplexes–9 are enhanced 1010––13 13inrelative in common common to the organic organic free flavonols solvents solvents enabled (Table enabled 2). the Limited use the of use10 solubility ofas 10 a visible as aof visible lightthe Zn(II) driven light bis in driven-flavonolato situ CO in release situ complexes CO compound release 10 – compoundfor13 in palladium-catalyzed common for palladium organic - solventscatalyzed carbonylation enabled carbonylation [119 the]. use [119]. of 10 as a visible light driven in situ CO release compound for palladium-catalyzed carbonylation [119].

Scheme 20. Visible light-induced CO release reactivity of Zn(II) complexes of extended 3-hydroxyflavonesScheme 20. 20. VisibleVisible [119]. light light-induced-induced CO CO release release reactivity reactivity of of Zn(I Zn(II)I) complexes of extended extended 3-hydroxyﬂavones3-hydroxyflavones [[119].119].

Molecules 2019, 24, x FOR PEER REVIEW 19 of 26

Table 2. Quantum yields for photoinduced CO release from extended 3-hydroxyflavones, zinc flavonolato complexes, and 3-hydroxybenzo[g]quinolone.

Absorption Maximum Quantum Yield for Reference Compound (nm) CO Release

Molecules 2019, 241, 1252 409 b 0.007(3) a,b [111,112] 19 of 26 410 c 0.006(3) a,c [112]

Table 2. Quantum yields for photoinduced410 d CO release from0.010(3) extended a,d 3-hydroxyﬂavones,[112] zinc ﬂavonolato complexes, and 3-hydroxybenzo[410 e g]quinolone. 0.0063(1) a,e [112]

Compound Absorption Maximum410 f (nm) Quantum0.0006(1) Yield for COa,f Release Reference[112] 1 2 409 442b b 0.006(1)0.007(3) a,ca,b [[111111],112 ] 410 c 0.006(3) a,c [112] 3 478 b 0.426(3) a,b [111] 410 d 0.010(3) a,d [112] 4 410 e Not0.0063(1) determineda,e [111[112] ] 410 F 0.0006(1) a,f [112] 6 480 b 0.651(2) a,b [119] 2 442 b 0.006(1) a,b [111] 3 7 478 524b b 0.583(4)0.426(3) a,bb,g [119[111] ] 4 8 550 b Not0.951(4) determined b,g [119 [111] ] 6 480 b 0.651(2) a,b [119] 7 9 524 600b b 0.947(7)0.583(4) b,gb,g [119[119] ] b b,g 8 14 550 445 b 0.0045(1)0.951(4) a,b [116[119] ] 9 600 b 0.947(7) b,g [119] a 419 nm; b CH3CN; c 1:1 DMSO:TRIS; d 1:1 DMSO:PBS; e 4% DMSO:PBS + CTAB; f BSA. (40 eq.) in 3.3% 14 445 b 0.0045(1) a,b [116] g aDMSO:TRIS;b Whitec light with 546d nm cut off filters.e f 419 nm; CH3CN; 1:1 DMSO:TRIS; 1:1 DMSO:PBS; 4% DMSO:PBS + CTAB; BSA. (40 eq.) in 3.3% DMSO:TRIS; g TheWhite 3- lighthydroxybenzo[ with 546 nm cutg]quinolone off filters. 14 (Scheme 21) was previously reported as a dye [120]. We recentlyThe demonstrated 3-hydroxybenzo[ that gthis]quinolone compound14 will(Scheme release 21 one) was equivalent previously of CO reported upon illumination as a dye [with120]. Wevisible recently light demonstratedto produce a non that-emissive this compound depside willproduct release [116]. one Similar equivalent to 1, the of CO presence upon illuminationof 14 in cells withcan be visible tracked light via to fluorescence produce a non-emissive prior to CO release. depside Notably, product [14116 has]. Similar a high toaffinity1, the (K presenceSV 2.9 × of10146 Min−1 6 −1 cells(SV = can Stern be tracked-Volmer) via) fluorescencefor bovine serum prior to albumin CO release. (BSA) Notably, protein,14 hasbinding a high at affinity Site I (K(warfarinSV 2.9 × 10bindingM (SVsite). = TheStern-Volmer)) binding affinity for bovine of 14 serum for BSA albumin is 900 (BSA)-fold protein, greater binding than that at Site of 1 I, (warfarin indicating binding that small site). Thestructural binding changes affinity will of 14dramaticallyfor BSA is influence 900-fold greaterprotein than interactions that of 1 in, indicating these bioinspired that small compounds. structural changesUsing the will non dramatically-covalent protein:influence14 proteinadduct interactions for visible in light these-ind bioinspireduced CO compounds. delivery to Using cells, the we non-covalentidentified a lowprotein: IC5014 valueadduct (24 for μ visibleM) for light-induced the eradication CO delivery of cancer to cells,cells we(A549) identified and potent a low ICanti50-inflammatoryvalue (24 µM) effects, for the as eradication measured ofby cancercomplete cells sup (A549)pression and of potent LPS-induced anti-inflammatory TNF-α production effects, asat measured80 nM in RAW by complete 264.7 murine suppression macrophage of LPS-induced cells [116]. TNF- Theα productionlow IC50 value at 80 may nM inbe RAW due in 264.7 part murine to the macrophageknown effect cellsof serum [116]. albumin The low proteins IC50 value in facilitating may be due the in partdelivery to the of knownsmall molecules effect of serum to cancer albumin cells proteinsvia the enhanc in facilitatinged permeability the delivery and ofretention small molecules effects [121]. to cancer Overall, cells these via theinitial enhanced results permeabilityindicate that andthe 3 retention-hydroxybenzo[ effects [g121]quinolone]. Overall, framework these initial has results significant indicate potential that the as 3-hydroxybenzo[ a novel CO-releasingg]quinolone motif frameworkfor applications has significant in biological potential systems. as a novel CO-releasing motif for applications in biological systems.

Scheme 21. CO release reactivity of 14 in the presence of bovine serum albumin (BSA) produces potent anti-cancer and anti-inﬂammatory effects [116]. Molecules 2019, 24, 1252 20 of 26

7. Concluding Remarks and Future Directions In this review we have outlined the various reported reaction pathways whereby 3-hydroxyflavone and 3-hydroxy-4-oxoquinoline compounds undergo CO release. Currently there is a strong interest in the development of carbon monoxide-releasing molecules (CORMs) that are metal free [20,122,123], especially those that can be triggered by visible light, which enables localized CO delivery. An additional desired characteristic for CORMs is a chemical framework that can be structurally modified to tune various factors including solubility, the rate of CO release, toxicity, and subcellular or tissue targeting [124]. The studies outlined herein demonstrate that the 3-hydroxyflavone motif will undergo CO release under a variety of reaction conditions, whether metal-coordinated or metal-free, and by either thermal or light-induced reaction pathways. The 3-hydroxyflavone framework is amenable to a variety of structural modifications to tune or augment chemical properties. In this regard, our lab has recently shown that an extended 3-hydroxyflavone motif can be structurally modified to enable sub-cellular targeting [115] and fluorescence-trackable small molecule sensing properties [117,118] prior to CO release. An additional intriguing feature of using 3-hydroxyflavone derivatives as CORMs is the potential of these compounds to exhibit other advantageous properties such as anti-inflammatory, antioxidant and anti-cancer effects [125,126] in addition to CO release. Future work in our laboratory is focused on identifying and investigating these possible dual activity effects. Overall, the development of CORMs based on 3-hydroxyflavone and 3-hydroxy-4-oxoquinoline structural frameworks is in its early stages. Using the knowledge of the studies summarized herein we anticipate further development of CORMs based on these biologically-relevant structures. In this regard, a key next step for our laboratory will be evaluation of the CO release reactivity of a range of 3-hydroxyflavone and 3-hydroxy-4-oxoquinoline derivatives under visible light-triggered conditions. Currently, very little is known regarding the structure/reactivity relationships for light-driven reactions of these molecules. If compounds with a range of absorption properties and quantum yields can be developed, this family of compounds could enable controlled, triggered release for the delivery of specific boluses of CO.

Funding: This work was financially supported by National Institutes of Health (R15GM124596 to L.M.B.), American Heart Association (18PRE34030099 to T.S.) and the USU Office of Research and Graduate Studies (PDRF Fellowship to T.S.). Conflicts of Interest: The authors declare no conflict of interest.

References

1. Motterlini, R.; Foresti, R. Biological signaling by carbon monoxide and carbon monoxide-releasing molecules. Am. J. Physiol. Cell Physiol. 2017, 312, C302–C313. [CrossRef][PubMed] 2. Motterlini, R.; Otterbein, L.E. The therapeutic potential of carbon monoxide. Nat. Rev. Drug Discov. 2010, 9, 728–743. [CrossRef][PubMed] 3. Kim, H.-H.; Choi, S. Therapeutic aspects of carbon monoxide in cardiovascular disease. Int. J. Mol. Sci. 2018, 19, 2381. [CrossRef][PubMed] 4. Lee, S.R.; Nilius, B.; Han, J. Gaseous Signaling Molecules in Cardiovascular Function: From Mechanism to Clinical Translation. In Reviews of Physiology, Biochemistry, and Pharmacology; Nilius, B., de Tombe, P., Gundermann, T., Jah, R., Lill, R., Peterson, O.H., Eds.; Springer International Publishing: Berlin/Heidelberg, Germany, 2018; Volume 174, pp. 81–156. 5. Otterbein, L.E.; Foresti, R.; Motterlini, R. Heme oxygeanse-1 and carbon monoxide in the heart: The balancing act between danger signaling and pro-survival. Circ. Res. 2016, 118, 1940–1959. [CrossRef][PubMed] 6. Fredenburgh, L.E.; Merz, A.A.; Cheng, S. Haeme oxygenase signalling pathway: Implications for cardiovascular disease. Eur. Heart J. 2015, 36, 1512–1518. [CrossRef][PubMed] 7. Szabo, C. Gasotransmitters in cancer: From pathophysiology to experimental therapy. Nat. Rev. Drug Discov. 2016, 15, 185–203. [CrossRef][PubMed] Molecules 2019, 24, 1252 21 of 26

8. Jin, Q.; Deng, Y.; Jia, F.; Tang, Z.; Ji, J. Gas therapy: An emerging “green” strategy for anticancer therapeutics. Adv. Therap. 2018, 1, 1800084. [CrossRef] 9. Motterlini, R. Carbon monoxide-releasing molecules (CO-RMs): Vasodilatory, anti-ischaemic and anti-inﬂammatory activities. Biochem. Soc. Trans. 2007, 35, 1142–1146. [CrossRef]

10. Crook, S.H.; Mann, B.E.; Meijer, A.J.H.M.; Adams, H.; Sawle, P.; Scapens, D.; Motterlini, R. [Mo(CO)4 {S2CNMe(CH2CO2H)}], a new water-soluble CO-releasing molecule. Dalton Trans. 2011, 40, 4230–4235. [CrossRef] 11. Aki, T.; Unuma, K.; Noritake, K.; Kurahashi, H.; Funakoshi, T.; Uemura, K. Interaction of carbon monoxide-releasing ruthenium carbonyl CORM-3 with plasma ﬁbronectin. Toxicol. In Vitro 2018, 50, 201–209. [CrossRef][PubMed] 12. Chaves-Ferreira, M.; Albuquerque, I.S.; Matak-Vinkovic, D.; Coelho, A.C.; Carvalho, S.M.; Saraiva, L.M.; II Romão, C.C.; Bernardes, G.J.L. Spontaneous CO release from Ru (CO)2-protein complexes in aqueous solution, cells, and mice. Angew. Chem. Int. Ed. 2015, 54, 1172–1175. [CrossRef][PubMed] 13. Seixas, J.D.; Santos, M.F.A.; Mukhopadhyay, A.; Coelho, A.C.; Reis, P.M.; Veiros, L.F.; Marques, A.R.;

Penacho, N.; Gonçalves, A.M.; Romão, M.J.; et al. A contribution to the rational design of Ru(CO)3Cl2L complexes for in vivo delivery of CO. Dalton Trans. 2015, 44, 5058–5075. [CrossRef][PubMed] 14. Santos-Silva, T.; Mukhopadhyay, A.; Seixas, J.D.; Bernardes, G.J.L.; Romão, C.C.; Romão, M.J. Towards improved therapeutic CORMs: Understanding the reactivity of CORM-3 with proteins. Curr. Med. Chem. 2011, 18, 3361–3366. [CrossRef] 15. Santos-Silva, T.; Mukhopadhyay, A.; Seixas, J.D.; Bernardes, G.J.L.; Romão, C.C.; Romão, M.J. CORM-3 reactivity toward proteins: The crystal structure of Ru(II) dicarbonyl-lysozyme complex. J. Am. Chem. Soc. 2011, 133, 1192–1195. [CrossRef] 16. Wang, P.; Liu, H.; Zhao, Q.; Chen, Y.; Liu, B.; Zhang, B.; Zheng, Q. Syntheses and evaluation of drug-like properties of CO-releasing molecules containing ruthenium and group 6 metal. Eur. J. Med. Chem. 2014, 74, 199–215. [CrossRef] 17. Davidge, K.S.; Motterlini, R.; Mann, B.E.; Wilson, J.L.; Poole, R.K. Carbon monoxide in biology and microbiology: Surprising roles for the “Detroit perfume”. Adv. Microb. Physiol. 2009, 56, 85–167. 18. Southam, H.M.; Smith, T.W.; Lyon, R.L.; Liao, C.; Trevitt, C.R.; Middlemiss, L.A.; Cox, F.L.; Chapman, J.A.; El-Khamisy, S.F.; Hippler, M.; et al. A thiol-reactive Ru(II) ion, not CO release, underlies the potent antimicrobial and cytotoxic properties of CO-releasing molecule-3. Redox Biol. 2018, 18, 114–123. [CrossRef] [PubMed] 19. Wareham, L.K.; McLean, S.; Begg, R.; Rana, N.; Ali, S.; Kendall, J.J.; Sanguinetti, G.; Mann, B.E.;

Poole, R.K. The broad-spectrum antimicrobial potential of [Mn(CO)4(S2CNCH3(CH2CO2H)], a water-soluble CO-releasing molecule (CORM-401): Intracellular accumulation, transcriptomic and statistical analyses, and membrane polarization. Antioxid. Redox Signal. 2018, 28, 1286–1308. [CrossRef][PubMed] 20. Abeyrathna, N.; Washington, K.; Bashur, K.; Liao, Y. Nonmetallic carbon monoxide releasing molecules. Org. Biomol. Chem. 2017, 15, 8692–8699. [CrossRef][PubMed] 21. Motterlini, R.; Sawle, P.; Hammad, J.; Bains, S.; Alberto, R.; Foresti, R.; Green, C.J. CORM-A1: A new pharmacologically active carbon monoxide-releasing molecule. FASEB J. 2005, 19, 284–286. [CrossRef] 22. Ji, X.; Wang, B. Strategies toward carbon monoxide prodrugs. Acc. Chem. Res. 2018, 51, 1377–1385. [CrossRef] 23. Peng, P.; Wang, C.; Shi, Z.; Johns, V.K.; Ma, L.; Oyer, J.; Copik, A.; Igarashi, R.; Liao, Y. Visible-light activatable organic CO-releasing molecules (PhotoCORMs) that simultaneously generate ﬂuorophores. Org. Biomol. Chem. 2013, 11, 6671–6674. [CrossRef][PubMed] 24. Antony, L.A.P.; Slanina, T.; Sebej, P.; Solomek, T.; Klán, P. Fluorescein analogue xanthene-9-carboxylic acid: A transition-metal-free CO releasing molecule activated by green light. Org. Lett. 2013, 15, 4552–4555. [CrossRef][PubMed] 25. Palao, E.; Slanina, T.; Muchová, L.; Solomek, T.; Vitek, L.; Klán, P. Transition-metal-free CO-releasing BODIPY derivatives activatable by visible to NIR light as promising bioactive molecules. J. Am. Chem. Soc. 2016, 138, 126–133. [CrossRef][PubMed] 26. Sjöstrand, T. The in vitro formation of carbon monoxide in blood. Acta Physiol. Scand. 1952, 24, 314–332. [CrossRef][PubMed] 27. Sjöstrand, T. Endogenous formation of carbon monoxide in man under normal and pathological conditions. Scan. J. Clin. Lab. Investig. 1949, 1, 201–214. [CrossRef] Molecules 2019, 24, 1252 22 of 26

28. Elbirt, K.K.; Bonkovksy, H.L. Heme oxygenase: Recent advances in understanding its regulation and role. Proc. Assoc. Am. Physicians 1999, 111, 438–447. [CrossRef] 29. Fetzner, S. Ring-cleaving dioxygenases with a cupin fold. Appl. Environ. Microbiol. 2012, 78, 2505–2514. [CrossRef][PubMed] 30. Fetzner, S.; Steiner, R.A. Cofactor-independent oxidases and oxygenases. Appl. Microbiol. Biotechnol. 2010, 86, 791–804. [CrossRef] 31. Conrad, R.; Seiler, W. Role of microorganisms in the consumption and production of atmospheric carbon monoxide by soil. Appl. Environ. Microbiol. 1980, 40, 437–445. [PubMed] 32. Patel, R.V.; Mistry, B.M.; Shinde, S.K.; Syed, R.; Singh, V.; Shin, H.S. Therapeutic potential of quercetin as a cardiovascular agent. Eur. J. Med. Chem. 2018, 155, 889–904. [CrossRef] 33. Rauf, A.; Imran, M.; Khan, I.A.; Ur-Rehman, M.; Gilani, S.A.; Mehmood, Z.; Mubarak, M.S. Anticancer potential of quercetin: A comprehensive review. Phytother. Res. 2018, 32, 2109–2130. [CrossRef] 34. Sharma, A.; Kashyap, D.; Sak, K.; Tuli, H.S.; Sharma, A.K. Therapeutic charm of quercetin and its derivatives: A review of research and patents. Pharm. Pat. Anal. 2018, 7, 15–32. [CrossRef] 35. Steiger, C.; Hermann, C.; Meinel, L. Localized delivery of carbon monoxide. Eur. J. Pharm. Biopharm. 2017, 118, 3–12. [CrossRef] 36. Fusetti, F.; Schröter, K.H.; Steiner, R.A.; van Noort, P.I.; Pijning, T.; Rozeboom, H.J.; Kalk, K.H.; Egmond, M.R.; Dijkstra, B.W. Crystal structure of the copper-containing quercetin 2,3-dioxygenase from Aspergillus japonicus. Structure 2002, 10, 259–268. [CrossRef] 37. Steiner, R.A.; Meyer-Klaucke, W.; Dijkstra, B.W. Functional analysis of the copper-dependent quercetin 2,3-dioxygenase. 2. X-ray absorption studies of native enzyme and anaerobic complexes with the substrates quercetin and myricetin. Biochemistry 2002, 41, 7963–7968. [CrossRef][PubMed] 38. Steiner, R.A.; Kalk, K.H.; Dijkstra, B.W. Anaerobic enzyme-substrate structures provide insight into the reaction mechanism of the copper-dependent quercetin 2,3-dioxygenase. Proc. Natl. Acad. Sci. USA 2002, 99, 16625–16630. [CrossRef][PubMed] 39. Siegbahn, P.E.M. Hybrid DFT study of the mechanism of quercetin 2,3-dioxygenase. Inorg. Chem. 2004, 43, 5944–5953. [CrossRef][PubMed] 40. Fiorucci, S.; Golebiowski, J.; Carbol-Bass, D.; Antonczak, S. Oxygenolysis of flavonoid compounds: DFT description of the mechanism for quercetin. ChemPhysChem 2004, 5, 1726–1733. [CrossRef][PubMed] 41. Oka, T.; Simpson, F.J.; Krishnamurty, H.G. Degradation of rutin by Aspergillus flavus. Studies on specificity, inhibition, and possible reaction mechanism of quercetinase. Can. J. Microbiol. 1972, 18, 493–508. [CrossRef] 42. Fiorucci, S.; Golebiowski, J.; Carbrol-Bass, D.; Antonczak, S. Molecular simulations bring new insights into flavonoid/quercetinase interaction modes. Proteins 2007, 67, 961–970. [CrossRef][PubMed] 43. Barney, B.M.; Schaab, M.R.; LoBrutto, R.; Francisco, W.A. Evidence for a new metal in a known active site: Purification and characterization of an iron-containing quercetin 2,3-dioxygenase from Bacillus subtilis. Protein Expr. Purif. 2004, 35, 131–141. [CrossRef] 44. Bowater, L.; Fairhurst, S.A.; Just, V.J.; Bornemann, S. Bacillus subtilis YxaG is a novel Fe-containing quercetin 2,3-dioxygenase. FEBS Lett. 2004, 557, 45–48. [CrossRef] 45. Gopal, B.; Madan, L.L.; Betz, S.F.; Kossiakoff, A.A. The crystal structure of a quercetin 2,3-dioxygenase from Bacillus subtilis suggests modulation of enzyme activity by a change in the metal ion at the active site(s). Biochemistry 2005, 44, 193–201. [CrossRef] 46. Schaab, M.R.; Barney, B.M.; Francisco, W.A. Kinetic and spectroscopic studies on the quercetin 2,3-dioxygenase from Bacillus subtilis. Biochemistry 2006, 45, 1009–1016. [CrossRef] 47. Merkens, H.; Sielker, S.; Rose, K.; Fetzner, S. A new monocupin quercetinase of Streptomyces sp. FLA: Identification and heterologous expression of the queD gene and activity of the recombinant enzyme towards different flavonols. Arch. Microbiol. 2007, 187, 475–487. [CrossRef] 48. Merkens, H.; Kappl, R.; Jakob, R.P.; Schmid, F.X.; Fetzner, S. Quercetinase QueD of Streptomyces sp. FLA, a monocupin dioxygenase with a preference for nickel and cobalt. Biochemistry 2008, 47, 12185–12196. [CrossRef] 49. Nianios, D.; Thierbach, S.; Steimer, L.; Lulchev, P.; Klostermeier, D.; Fetzner, S. Nickel quercetinase, a “promiscuous” metalloenzyme: Metal incorporation and metal ligand substitution studies. BMC Biochem. 2015, 16, 10. [CrossRef][PubMed] Molecules 2019, 24, 1252 23 of 26

50. Jeoung, J.H.; Nianios, D.; Fetzner, S.; Dobbek, H. Quercetin 2,4-dioxygenase activates dioxygen in a side-on

O2-Ni Complex. Angew. Chem. Int. Ed. Engl. 2016, 55, 3281–3284. [CrossRef][PubMed] 51. Li, H.; Wang, X.; Tian, G.; Liu, Y. Insights into the dioxygen activation and catalytic mechanism of the nickel-containing quercetinase. Catal. Sci. Technol. 2018, 8, 2340–2351. [CrossRef] 52. Wang, W.-J.; Wei, W.-J.; Liao, R.-Z. Deciphering the chemoselectivity of nickel-dependent quercetin 2,4-dioxygenase. Phys. Chem. Chem. Phys. 2018, 20, 15784–15794. [CrossRef][PubMed] 53. Kumar, M.R.; Zapata, A.; Ramirez, A.J.; Bowen, S.K.; Francisco, W.A.; Farmer, P.J. Nitrosyl hydride (HNO) replaces dioxygen in nitroxygenase activity of manganese quercetin dioxygenase. Proc. Natl. Acad. Sci. USA 2011, 108, 18926–18931. [CrossRef] 54. Shafirovich, V.; Lymar, S.V. Spin-forbidden deprotonation of aqueous nitroxyl (HNO). J. Am. Chem. Soc. 2003, 125, 6547–6552. [CrossRef][PubMed] 55. Wojdyla, Z.; Borowski, T. DFT study of the mechanism of manganese quercetin 2,3-dioxygenase: Quest for origins of enzyme unique nitroxygenase activity and regioselectivity. J. Biol. Inorg. Chem. 2016, 21, 475–489. [CrossRef] 56. Kaizer, J.; Balogh-Hergovich, E.; Czaun, M.; Csay, T.; Speier, G. Redox and nonredox metal assisted model systems with relevance to flavonol and 3-hydroxyquinolin-4(1H)-one 2,4-dioxygenase. Coord. Chem. Rev. 2006, 250, 2222–2233. [CrossRef] 57. Pap, J.S.; Kaizer, J.; Speier, G. Model systems for the CO-releasing flavonol 2,4-dioxygenase enzyme. Coord. Chem. Rev. 2010, 254, 781–793. [CrossRef] 58. Kaizer, J.; Pap, J.S.; Speier, G. Copper Dioxygenases. In Copper-Oxygen Chemistry; Karlin, K.D., Itoh, S., Eds.; John Wiley & Sons, Inc.: Hoboken, NJ, USA, 2011; pp. 23–52. 59. Barhács, L.; Kaizer, J.; Pap, J.; Speier, G. Kinetics and mechanism of the stoichiometric oxygenation II of [Cu (fla)(idpa)ClO4 [fla = flavonolate, idpa = 3,3’-imino-bis(N,N-dimethylpropylamine)] and the II [Cu (fla)(idpa]ClO4-catalyzed oxygenation of flavonol. Inorg. Chim. Acta 2001, 320, 83–91. [CrossRef] 60. Balogh-Hergovich, É.; Kaizer, J.; Speier, G. Carboxylate-enhanced reactivity in the oxygenation of copper flavonolate complexes. J. Mol. Catal. A Chem. 2003, 206, 83–87. [CrossRef] 61. Balogh-Hergovich, É.; Kaizer, J.; Pap, J.; Speier, G.; Huttner, G.; Zsolnai, L. Copper-mediated oxygenolysis of flavonols via endoperoxide and dioxetan intermediates. Synthesis and oxygenation of II II [Cu (phen)2(fla)]ClO4 and [Cu (L)(fla)2] [flaH = flavonol; L = 1,10-phenanthroline (phen), 2,2’-bipyridine (bpy), N,N,N’,N’-tetramethylethylenediamine (tmeda)] complexes. Eur. J. Inorg. Chem. 2002, 2002, 2287–2295. 62. Hiatt, R. Organic Peroxides; Swern, D., Ed.; Wiley Interscience: New York, NY, USA, 1971; Volume III, p. 70. 63. Grubel, K.; Rudzka, K.; Arif, A.M.; Klotz, K.L.; Halfen, J.A.; Berreau, L.M. Synthesis, characterization, and ligand exchange reactivity of a series of first row divalent metal 3-hydroxyflavonolate complexes. Inorg. Chem. 2010, 49, 82–96. [CrossRef] 64. Sun, Y.-J.; Huang, Q.-Q.; Tano, T.; Itoh, S. Flavonolate complexes of MII (M = Mn, Fe, Co, Ni, Cu, and Zn). Structural and functional models for the ES (enzyme–substrate) complex of quercetin 2,3-dioxygenase. Inorg. Chem. 2013, 52, 10936–10948. [CrossRef][PubMed] 65. Matuz, A.; Giorgi, M.; Speier, G.; Kaizer, J. Structural and functional comparison of manganese-, iron-, cobalt-, nickel-, and copper-containing biomimic quercetinase models. Polyhedron 2013, 63, 41–49. [CrossRef] 66. Sun, Y.-J.; Huang, Q.-Q.; Zhang, J.-J. Series of structural and functional models for the ES (enzyme–substrate) complex of the Co(II)-containing quercetin 2,3-dioxygenase. Inorg. Chem. 2014, 53, 2932–2942. [CrossRef] [PubMed] 67. Sun, Y.-J.; Huang, Q.-Q.; Zhang, J.-J. A series of NiII-flavonolate complexes as structural and functional ES (enzyme-substrate) models of the NiII-containing quercetin 2,3-dioxygenase. Dalton Trans. 2014, 43, 6480–6489. [CrossRef] 68. Sun, Y.-J.; Huang, Q.-Q.; Zhang, J.-J. Set of Fe(II)-3-hydroxyflavonolate enzyme–substrate model complexes of atypically coordinated mononuclear non-heme Fe(II)-dependent quercetin 2,4-dioxygenase. ACS Omega 2017, 2, 5850–5860. [CrossRef] 69. Kaizer, J.; Baráth, G.; Pap, J.; Speier, G.; Giorgi, M.; Réglier, M. Manganese and iron flavonolates as flavonol 2,4-dioxygenase mimics. Chem. Commun. 2007, 5235–5237. [CrossRef][PubMed] 70. Balogh-Hergovich, É; Kaizer, J.; Speier, G.; Argay, G.; Párkányi, L. Kinetic studies on the copper(II)-mediated

oxygenolysis of the flavonolate ligand. Crystal structures of [Cu(fla)2] (fla = flavonolate) and [Cu(O-bs)2(py)3] (O-bs = O-benzoylsalicylate). J. Chem. Soc. Dalton Trans. 1999, 3847–3854. [CrossRef] Molecules 2019, 24, 1252 24 of 26

71. Balogh-Hergovich, É.; Kaizer, J.; Speier, G.; Fülöp, V.; Párkányi, L. Quercetin 2,3-dioxygenase mimicking ring cleavage of the ﬂavonolate ligand assisted by copper. Synthesis and characterization of copper(I) complexes

[Cu(PPh3)2(fla)] (fla = flavonolate) and [Cu(PPh3)2(O-bs)] (O-bs = O-benzoylsalicylate). Inorg. Chem. 1999, 38, 3787–3795. [CrossRef] 72. Baráth, G.; Kaizer, J.; Speier, G.; Párkányi, L.; Kuzmann, E.; Vértes, A. One metal–two pathways to the carboxylate-enhanced, iron-containing quercetinase mimics. Chem. Commun. 2009, 3630–3632. [CrossRef] [PubMed] 73. Frerichs-Deeken, U.; Ranguelova, K.; Kappl, R.; Hüttermann, J.; Fetzner, S. Dioxygenases without requirement for cofactors and their chemical model reaction: Compulsory order ternary complex mechanism of 1H-3-hydroxy-4-oxoquinaldine 2,4-dioxygenase involving general base catalysis by histidine 251 and single-electron oxidation of the substrate dianion. Biochemistry 2004, 43, 14485–14499. 74. Steiner, R.A.; Janssen, H.J.; Roversi, P.; Oakley, A.J.; Fetzner, S. Structural basis for cofactor-independent dioxygenation of N-heteroaromatic compounds at the alpha/beta-hydrolase fold. Proc. Natl. Acad. Sci. USA 2010, 107, 657–662. [CrossRef]

75. Thierbach, S.; Bui, N.; Zapp, J.; Chhabra, S.R.; Kappl, R.; Fetzner, S. Substrate-assisted O2 activation in a cofactor-independent dioxygenase. Chem. Biol. 2014, 21, 217–225. [CrossRef] 76. Hernández-Ortega, A.; Quesne, M.G.; Bui, S.; Heyes, D.J.; Steiner, R.A.; Scrutton, N.S.; de Visser, S.P. Catalytic mechanism of cofactor-free dioxygenases and how they circumvent spin-forbidden oxygenation of their substrates. J. Am. Chem. Soc. 2015, 137, 7474–7487. [CrossRef][PubMed] 77. Bui, S.; Steiner, R.A. New insight into cofactor-free oxygenation from combined experimental and computational approaches. Curr. Opin. Struct. Biol. 2016, 41, 109–118. [CrossRef] 78. Czaun, M.; Speier, G.; Párkányi, L. Facile copper-mediated activation of the N-H bond and the oxidative cleavage of the C2-C3 bond in 1H-2-phenyl-3-hydroxy-4-oxoquinoline. Chem. Commun. 2004, 1004–1005. [CrossRef][PubMed] 79. Pap, J.S.; Matuz, A.; Baráth, G.; Kripli, B.; Giorgi, M.; Speier, G.; Kaizer, J. Bio-inspired flavonol and quinolone dioxygenation by a non-heme iron catalyst modeling the action of flavonol and 3-hydroxy-4(1H)—quinolone 2,4-dioxygenases. J. Inorg. Biochem. 2012, 108, 15–21. [CrossRef][PubMed] 80. Nishinaga, A.; Tojo, T.; Tomita, H.; Matsuura, T. Base-catalysed oxygenolysis of 3-hydroxyflavones. J. Chem. Soc. Perkin Trans. 1 1979, 2511–2516. [CrossRef] 81. Barhács, L.; Kaizer, J.; Speier, G. Kinetics and mechanism of the oxygenation of potassium flavonolate. Evidence for an electron transfer mechanism. J. Org. Chem. 2000, 65, 3449–3452. [CrossRef] 82. Balogh-Hergovich, É.; Speier, G. Kinetics and mechanism of the base-catalyzed oxygenation of flavonol in

DMSO−H2O solution. J. Org. Chem. 2001, 66, 7974–7978. [CrossRef] 83. Han, X.; Kumar, M.R.; Farmer, P.J. Nitroxygenation of quercetin by HNO. Tetrahedron Lett. 2016, 57, 399–402. [CrossRef] 84. Czaun, M.; Speier, G. The base-catalyzed oxygenation of quinolone derivatives. Tetrahedron Lett. 2002, 43, 5961–5963. [CrossRef] 85. Matsuura, T.; Matsushima, H.; Nakashima, R. Photoinduced reactions—XXXVI: Photosensitized oxygenation of 3-hydroxyflavones as a nonenzymatic model for quercetinase. Tetrahedron 1970, 26, 435–443. [CrossRef] 86. Studer, S.L.; Brewer, W.E.; Martinez, M.L.; Chou, P.T. Time-resolved study of the photooxygenation of 3-hydroxyflavone. J. Am. Chem. Soc. 1989, 111, 7643–7644. [CrossRef] 87. Brewer, W.E.; Studer, S.L.; Standiford, M.; Chou, P.T. Dynamics of the triplet state and the reverse proton transfer of 3-hydroxyflavone. J. Phys. Chem. 1989, 93, 6088–6094. [CrossRef] 88. Martinez, M.L.; Studer, S.L.; Chou, P.T. Direct evidence of the triplet-state origin of the slow reverse proton transfer reaction of 3-hydroxyflavone. J. Am. Chem. Soc. 1990, 112, 2427–2429. [CrossRef] 89. Sisa, M.; Bonnet, S.L.; Ferreira, D.; Van der Westhuizen, J.H. Photochemistry of flavonoids. Molecules 2010, 15, 5196–5245. [CrossRef][PubMed] 90. Szakács, Z.; Bojtár, M.; Drahos, L.; Hessz, D.; Kállay, M.; Vidóczy, T.; Bitter, I.; Kubinyi, M. The kinetics and mechanism of photooxygenation of 40-diethylamino-3-hydroxyflavone. Photochem. Photobiol. Sci. 2016, 15, 219–227. [CrossRef][PubMed] 91. Szakács, Z.; Kállay, M.; Kubinyi, M. Theoretical study on the photooxygenation and photorearrangement reactions of 3-hydroxyflavone. RSC Adv. 2017, 7, 32185–32192. [CrossRef] Molecules 2019, 24, 1252 25 of 26

92. Matsuura, T.; Takemoto, T.; Nakashima, R. Photoinduced reactions. LII. Photorearragement of 3-hydroxyflavones. Tetrahedron 1971, 12, 1539–1540. [CrossRef] 93. Matsuura, T.; Takemoto, T.; Nakashima, R. Photoinduced reactions—LXXI: Photorearragement of 3-hydroxyflavones to 3-aryl-3-hydroxy-1,2-indandiones. Tetrahedron 1973, 29, 3337–3340. [CrossRef] 94. Yokoe, I.; Higuchi, K.; Shirataki, Y.; Komatsu, M. Photochemistry of flavonoids. III. Photorearrangement of flavonols. Chem. Pharm. Bull. 1981, 29, 894–898. [CrossRef] 95. Grubel, K.; Laughlin, B.J.; Maltais, T.R.; Smith, R.C.; Arif, A.M.; Berreau, L.M. Photochemically-induced dioxygenase-type CO-release reactivity of group 12 metal flavonolate complexes. Chem. Commun. 2011, 47, 10431–10433. [CrossRef][PubMed] 96. Grubel, K.; Marts, A.R.; Greer, S.M.; Tierney, D.L.; Allpress, C.J.; Anderson, S.N.; Laughlin, B.J.; Smith, R.C.; Arif, A.M.; Berreau, L.M. Photoinitiated dioxygenase-type reactivity of open-shell 3d divalent metal flavonolato complexes. Eur. J. Inorg. Chem. 2012, 2012, 4750–4757. [CrossRef] 97. Grubel, K.; Saraf, S.L.; Anderson, S.N.; Laughlin, B.J.; Smith, R.C.; Arif, A.M.; Berreau, L.M. Synthesis, characterization, and photoinduced CO-release reactivity of a Pb(II) flavonolate complex: Comparisons to Group 12 analogs. Inorg. Chim. Acta 2013, 407, 91–97. [CrossRef] 98. Anderson, S.N.; Noble, M.; Grubel, K.; Marshall, B.; Arif, A.M.; Berreau, L.M. Influence of supporting ligand microenvironment on the aqueous stability and visible light-induced CO-release reactivity of zinc flavonolato species. J. Coord. Chem. 2014, 67, 4061–4075. [CrossRef] 99. Sorenson, S.; Popova, M.; Arif, A.M.; Berreau, L.M. A bipyridine-ligated zinc(II) complex with bridging flavonolate ligation: Synthesis, characterization and visible-light-induced CO release reactivity. Acta Cryst. 2017, C73, 703–709. [CrossRef] 100. Su, Y.; Yang, W.; Yang, X.; Zhang, R.; Zhao, J. Visible light-induced CO-release reactivity of a series of ZnII-flavonolate complexes. Aust. J. Chem. 2018, 71, 549–558. [CrossRef] 101. Protti, S.; Mezzetti, A.; Lapouge, C.; Cornard, J.-P. Photochemistry of metal complexes of 3-hydroxyflavone: Towards a better understanding of the influence of solar light on the metal-soil organic matter interactions. Photochem. Photobiol. Sci. 2008, 7, 109–119. [CrossRef] 102. Sun, Y.-J.; Huang, Q.-Q.; Li, P.; Zhang, J.-J. Catalytic dioxygenation of flavonol by MII-complexes (M = Mn, Fe, Co, Ni, Cu, and Zn)—Mimicking the MII-substituted quercetin 2,3-dioxygenase. Dalton Trans. 2015, 44, 13926–13938. [CrossRef] 103. Sun, Y.-J.; Li, P.; Huang, Q.-Q.; Zhang, J.-J.; Itoh, S. Dioxygenation of flavonol catalyzed by copper(II) complexes supported by carboxylate-containing ligands: Structural and functional models of quercetin 2,4-dioxygenase. Eur. J. Inorg. Chem. 2017, 2017, 1845–1854. [CrossRef] 104. Vogler, A.; Hischa, B. Synthesis, crystal structure, photoluminescence and photochemistry of bis(triphenylphosphine)siliver(I) flavonolate. Z. Naturforsch. B 2018, 74.[CrossRef] 105. Saraf, S.L.; Fish, T.J.; Benninghoff, A.D.; Buelt, A.A.; Smith, R.C.; Berreau, L.M. Photochemical reactivity of RuII(η6-p-cymene) flavonolato compounds. Organometallics 2014, 33, 6341–6351. [CrossRef] 106. Kurzwernhart, A.; Kandioller, W.; Bartel, C.; Bächler, S.; Trondl, R.; Mühlgassner, G.; Jakupec, M.A.; Arion, V.B.; Marko, D.; Keppler, B.K.; et al. Targeting the DNA-topoisomerase complex in a double-strike approach with a topoisomerase inhibiting moiety and covalent DNA binder. Chem. Commun. 2012, 48, 4839–4841. [CrossRef] 107. Kurzwernhart, A.; Kandioller, W.; Bächler, S.; Bartel, C.; Martic, S.; Buczkowska, M.; Mühlgassner, G.; Jakupec, M.A.; Kraatz, H.-B.; Bednarski, P.J.; et al. Structure-activity relationships of targeted RuII(η6-p-cymene) anticancer complexes with flavonol-derived ligands. J. Med. Chem. 2012, 55, 10512–10522. [CrossRef] 108. Kurzwernhart, A.; Kandioller, W.; Enyedy, É.A.; Novak, M.; Jakupec, M.A.; Keppler, B.K.; Hartinger, C.G. 3-hydroxyflavones vs. 3-hydroxyquinolinones: Structure-activity relationships and stability studies on RuII(arene) anticancer complexes with biologically active ligands. Dalton Trans. 2013, 42, 6193–6202. [CrossRef] 109. Han, X.; Klausmeyer, K.K.; Farmer, P.J. Characterization of the initial intermediate formed during photoinduced oxygenation of the ruthenium(II) bis(bipyridyl)flavonolate complex. Inorg. Chem. 2016, 55, 7320–7322. [CrossRef][PubMed] Molecules 2019, 24, 1252 26 of 26

110. Han, X.; Kumar, M.R.; Hoogerbrugge, A.; Klausmeyer, K.K.; Ghimire, M.M.; Harris, L.M.; Omary, M.A.; Farmer, P.J. Mechanistic investigations of photoinduced oxygenation of Ru(II) bis-bipyridyl flavonolate complexes. Inorg. Chem. 2018, 57, 2416–2424. [CrossRef] 111. Anderson, S.N.; Richards, J.M.; Esquer, H.J.; Benninghoff, A.D.; Arif, A.M.; Berreau, L.M. A Structurally-tunable 3-hydroxyflavone motif for visible light-induced carbon monoxide-releasing molecules (CORMs). ChemistryOpen 2015, 4, 590–594. [CrossRef] 112. Popova, M.; Soboleva, T.; Arif, A.M.; Berreau, L.M. Properties of a flavonol-based photoCORM in aqueous buffered solutions: Influence of metal ions, surfactants and proteins on visible light-induced CO release. RSC Adv. 2017, 7, 21997–22007. [CrossRef]

113. Li, Y.; Shu, Y.; Liang, M.; Xie, X.; Jiao, X.; Wang, X.; Tang, B. A two-photon H2O2 -activated CO photoreleaser. Angew. Chem. Int. Ed. Engl. 2018, 57, 12415–12419. [CrossRef][PubMed] 114. Dick, B.; Ernsting, N.P. Excited-state intramolecular proton transfer in 3-hydroflavone isolated in solid argon: Fluorescence and fluorescence-excitation spectra and tautomer fluorescence rise time. J. Phys. Chem. 1987, 91, 4261–4265. [CrossRef] 115. Soboleva, T.; Esquer, H.J.; Anderson, S.N.; Berreau, L.M.; Benninghoff, A.D. Mitochondrial-localized versus cytosolic intracellular CO-releasing organic photoCORMs: Evaluation of CO effects using bioenergetics. ACS Chem. Biol. 2018, 13, 2220–2228. [CrossRef] 116. Popova, M.; Soboleva, T.; Ayad, S.; Benninghoff, A.D.; Berreau, L.M. Visible-light-activated quinolone carbon-monoxide-releasing molecule: Prodrug and albumin-assisted delivery enables anticancer and potent anti-inflammatory effects. J. Am. Chem. Soc. 2018, 140, 9721–9729. [CrossRef] 117. Soboleva, T.; Esquer, H.J.; Benninghoff, A.D.; Berreau, L.M. Sense and release: A thiol-responsible flavonol-based photonically driven carbon monoxide-releasing molecule that operates via a multiple-input AND logic gate. J. Am. Chem. Soc. 2017, 139, 9435–9438. [CrossRef]

118. Soboleva, T.; Benninghoff, A.D.; Berreau, L.M. An H2S-sensing/CO-releasing flavonol that operates via logic gates. ChemPlusChem 2017, 82, 1408–1412. [CrossRef][PubMed] 119. Anderson, S.N.; Larson, M.T.; Berreau, L.M. Solution or solid—It doesn’t matter: Visible light-induced CO release reactivity of zinc flavonolato complexes. Dalton Trans. 2016, 45, 14570–14580. [CrossRef][PubMed] 120. Bilokin, M.D.; Shvadchak, V.V.; Yushchenko, D.A.; Klymchenko, A.S.; Duportail, G.; Mely, Y.; Pivovarenko, V.G. 3-hydroxybenzo[g]quinolones: Dyes with red-shifted absorption and highly resolved dual emission. Tetrahedron Lett. 2009, 50, 4714–4719. [CrossRef] 121. Sleep, D. Albumin and its application in drug delivery. Expert Opin. Drug Deliv. 2015, 12, 793–812. [CrossRef] 122. Hopper, C. Keto Acids in Carbon Monoxide Therapeutics, 2nd ed.; Independently published via Amazon: Seattle, WA, USA, 2018. 123. Slanina, T.; Sebej, P. Visible-light-activated photoCORMs: Rational design of CO-releasing organic molecules absorbing in the tissue-transparent window. Photochem. Photobiol. Sci. 2018, 17, 692–710. [CrossRef] 124. Ramão, C.C.; Blättler, W.A.; Seixas, J.D.; Bernardes, G.J.L. Developing drug molecules for therapy with carbon monoxide. Chem. Soc. Rev. 2012, 41, 3571–3583. [CrossRef][PubMed] 125. Panche, A.N.; Diwan, A.D.; Chandra, S.R. Flavonoids: An overview. J. Nutr. Sci. 2016, 5, e47. [CrossRef] [PubMed] 126. Butun, B.; Topcu, G.; Ozturk, T. Recent advances on 3-hydroxyflavone derivatives: Structures and properties. Mini Rev. Med. Chem. 2018, 18, 98–103. [CrossRef][PubMed]

© 2019 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/).