Non‐Oxidative Enzymatic (De)Carboxylation of (Hetero)Aromatics and Acrylic Acid Derivatives

Abstract The utilization of carbon dioxide as a C1‐building block for the production of valuable chemicals has recently attracted much interest. Whereas chemical CO2 fixation is dominated by C−O and C−N bond forming reactions, the development of novel concepts for the carboxylation of C‐nucleophiles, which leads to the formation of carboxylic acids, is highly desired. Beside transition metal catalysis, biocatalysis has emerged as an attractive method for the highly regioselective (de)carboxylation of electron‐rich (hetero)aromatics, which has been recently further expanded to include conjugated α,β‐unsaturated (acrylic) acid derivatives. Depending on the type of substrate, different classes of enzymes have been explored for (i) the ortho‐carboxylation of phenols catalyzed by metal‐dependent ortho‐benzoic acid decarboxylases and (ii) the side‐chain carboxylation of para‐hydroxystyrenes mediated by metal‐independent phenolic acid decarboxylases. Just recently, the portfolio of bio‐carboxylation reactions was complemented by (iii) the para‐carboxylation of phenols and the decarboxylation of electron‐rich heterocyclic and acrylic acid derivatives mediated by prenylated FMN‐dependent decarboxylases, which is the main focus of this review. Bio(de)carboxylation processes proceed under physiological reaction conditions employing bicarbonate or (pressurized) CO2 when running in the energetically uphill carboxylation direction. Aiming to facilitate the application of these enzymes in preparative‐scale biotransformations, their catalytic mechanism and substrate scope are analyzed in this review.

Abstract: Theu tilization of carbon dioxide as aC 1building block for the production of valuablec hemicals has recently attracted much interest. Whereas chemical CO 2 fixation is dominated by C À Oa nd C À Nb ondf orming reactions,t he development of novel conceptsf or the carboxylation of C-nucleophiles, which leads to the formation of carboxylic acids,i s highly desired. Beside transitionm etal catalysis, biocatalysis hase merged as an attractive method for the highly regioselective (de)carboxylation of electronrich (hetero)aromatics,w hich has been recently further expanded to include conjugated a,b-unsaturated (acrylic) acid derivatives.D epending on the type of substrate,d ifferent classeso fe nzymes have been explored for (i)t he ortho-carboxylation of phenols catalyzed by metal-dependent ortho-benzoic acid decarboxylases and (ii)t he side-chainc arboxylation of para-hydroxystyrenes mediated by metal-independent phenolic acid decarboxylases.J ust recently,t he portfolio of bio-carboxylation reactions was complementedb y( iii)t he para-carboxylation of phenols and the decarboxylationo fe lectron-rich heterocyclic and acrylic acid derivatives mediated by prenylated FMN-dependent decarboxylases,w hich is the main focus of this review.B io(de)carboxylation processes proceedu nder physiological reactionc onditions employingb icarbonate or (pressurized) CO 2 when running in the energetically uphill carboxylation direction. Aiming to facilitate the applicationo ft hese enzymes in preparative-scale biotransformations,t heir catalytic mechanism and substrate scope are analyzed in this review.

1I ntroduction
Although carbon dioxide (CO 2 )i sp redominantly regarded as an undesired greenhouse gas,i ts utilization in carboxylation reactions as aC 1 -building block for the synthesis of value-added compounds has become highly attractivei nr ecent years.B eing aw eak electrophile, carbon dioxide can react with N-, O-, and Cnucleophiles in af ormal addition reaction, leading to carbamates,u reas,c arbonate esters [1] and carboxylic acids. [2] Thef actt hat these reactions proceedw ith 100% atom economy andt he low (no) cost of the re-agent renders CO 2 an attractiver esource for industrial processes. In particular,t he production of urea (~95 Mt/a), organic (poly)carbonates (~150 kt/a) and salicylic acid (~70 kt/a)p roceeds on av ery large scale,i na ddition to the (chemical) reductiono fC O 2 to (predominantly) methanol (~60 Mt/a). Despite these impressive large-scale processes,t he contribution of chemical CO 2 fixationd uring the synthesiso f organic molecules is only marginal ( 1%,~200 Mt/a) comparedt oi ts massive generation in combustion processes( 37 Gt/a) [3] for the generation of energy, which proceeds with am odest efficiency of 30-35%, dictatedb yt he second law of thermodynamics. [4] Consequently,c arboxylation for the synthesis of organic compounds is irrelevant in the balance of the carbon cycle. [5] In contrast to industrial processes, carboxylation is only scarcelyu sed in small molecule synthesis, owingt ot he low reactivity and kinetic and thermodynamic inertness of CO 2 ,w hich requires significant catalytic activation and/or substantial energy input.T he harsh reactionc onditions requiredi nt raditional carboxylation processes [6] are often the cause for limited selectivities.O ver the last decades,v arious strategies have been developed in order to overcome these limitations. Tr aditionally,b eside strong nucleophilic organometallic (organolithium or Grignard) reagents, [7] which often suffer from poor chemoselectivity,h ighly straineds ubstrates (such as epoxides,a ziridines,e tc.), which are restrictedt oC À Oa nd C À Nb ond-forming reactions,w eree mployed to overcome the energetic barrier to yield the corresponding linear or cyclic (poly)carbonates/carbamates. [2] Thee normous progress in transitionm etal catalysis andc ross-coupling reactions has paved the way for the most desired  (2005)(2006)(2007) with Prof.N icholas J. Tu rner, she returned to Graz as scientist and wasl ater appointed as senior scientist within the Austrian Centre of IndustrialB iotechnology( acib) and the University of Graz. She currently holds ap osition as university assistantinthe Department of Chemistry at the University of Graz. Her research interests focus on biocatalytic synthesis as alternative to chemical systems, in particular, (de)carboxylation and asymmetric hydration reactions. carbon dioxide fixation via C À Cb ondf ormation to furnish the corresponding carboxylica cids. [2] The latter is the second most abundantf unctional group occurringi ns mall molecules producedb yc hemical synthesis. [8] Them ost commonly usedp re-activated substrates for transitionm etal-catalyzed carboxylations include allylstannanes,o rganoboronic esters,o rganozinc reagents and aryl halides. [9] Unsaturated compounds (olefins,a llenes,a lkynes) can be reductively carboxylated in aformal hydrocarboxylation [10] at the expense of ac o-reactant providing hydride species. [2d,11] Thep rogress in organocatalytic [12] and electrochemical [13] methods furthere xpands the scope of carbon dioxidef ixationr eactions in organic synthesis. In this context, modern reaction/process engineering tools (such as continuous flow technologies, [14] etc.) are being developed for process design.
In nature,f our major pathways for biological CO 2 fixationh ave been evolved: [8,15] (i)t he Calvin-Benson-Bassham cycle,(ii)the Arnon-Buchanan (reductive TCA)c ycle,( iii)t he Wood-Ljungdahl (reductive AcetylCoA)c ycle,a nd the (iv) acetyl-CoA pathways.C ommon to all these pathways is the fact that they belong to the primary metabolism, hence the enzymes involveda re highly specialized (evolved) for one (or af ew) substrate(s). [16] Consequently,t hey are generally of limited use for the biotransformation of non-natural organic compounds.I nc ontrast, enzymes involved in defence andd etoxification -t he secondary metabolism -a re generalists,a st hey act on ab road varietyo fs ubstrates andh ence are more useful as biocatalysts for organic synthesis. [8,17,18] Them ajor goal of detoxification is making (lipophilic) toxins more polar to assist their removal from the cell.A mong several pathways (such as oxidation, glycosylation, phosphorylation,s ulfation, peptide conjugation), carboxylation is av iable option to convert lipophilic aromaticsi nto water-soluble carboxylic acids.
Arenes are widely distributed in nature ands erve as substrates for aerobic and anaerobic organisms. Most natural aromaticc ompounds are derivedf rom secondary plant metabolism and often contain phenolic groups,s uch as productsf rom lignin degradation, tannins and flavonoids,w hich predominantly consist of substituted phenols,b enzaldehydes,b enzoic and cinnamic acids.Whereas oxidative biodegradation of aromatics mainly involves oxygenases,a naerobic bacteria apply reductive pathways [19] or redox-neutral carboxylation. [20] Theu tilization of naturest ools -( de)carboxylases -toestablish biocatalytic concepts for the (de)carboxylation of (hetero)aromatic substrates andc onjugated a,b-unsaturated carboxylic (acrylic) acids as as ustainable alternative to chemical methods (such as the Kolbe-Schmitt process) has been intensively investigated over the past years.I nt his review,v arious methodsa re described together with the mechanism of the respective enzymes and their substrate tolerance with particularf ocus on the very recentlye xplored reversible (de)carboxylationr eactions mediated by prenylated FMN-dependent decarboxylases. Thel atter are applicablet ot he decarboxylation of acrylic acid derivatives as well as the para-carboxylation of phenols.F urthermore,ageneralo verview is given for the ortho-a nd side-chain (de)carboxylation of phenol-and hydroxystyrene-type substrates by metal-dependent and cofactor-independent decarboxylases,b oth summarized in ar ecent comprehensive review by I. C. Tommasi. [21] 2E nzymatic (De)Carboxylation Enzymesh ave developed ad iverse set of strategies for the attachment or releaseo fC O 2 to or from various substrates by exploiting metal ions or cofactors, like pyridoxal phosphate( PLP), thiamine diphosphate (ThDP,v itamin B6) or prenylated FMN (prFMN). [22] Thef ollowing requirements to facilitate the carboxylation of aC À Hb ond to install ac arboxylate group need to be fulfilled (Scheme 1): (i)a bstraction of a proton (commonly with ap K a of [15][16][17][18] to generate an intermediary carbanion equivalent, (ii)s tabilization of the accumulating negative charge through delocalization within the substrate structure (mostly as metal ion-chelated enolate) or ac ofactorc onjugate, and (iii)a ctivation of carbon dioxide toward nucleophilic attack of the carbanion equivalent. [23] Theq uestiono fw hether CO 2 or bicarbonate serves as au niform co-substrate in enzymatic carboxylation is stillu nder debate. [24] Although the majority of carboxylases utilize the electrophilic, yet poorly watersoluble carbon dioxide,f ewer enzymes like phosphorenol pyruvate carboxylase and biotin-dependent carboxylases exploit the considerably less reactive( but water-soluble) bicarbonatei on for carboxylation. [23] These enzymes are proposedtopossess adual carboxylase/carbonic anhydrase activity,w hich allowst hem to interconvert bicarbonatea nd CO 2 and use the latter as actualc o-substrate for carboxylation. On the other hand, as pecific motif responsible for binding Scheme1.Genericc arboxylation of aC À Hb ond:d eprotonation forms ac arbanion, which attacks aC O 2 electrophile. CO 2 in carboxylases in analogy to the CO 2 binding site of phosphoenol pyruvate carboxykinase, [25] has not been identified yet.
3B iocatalytic (De)Carboxylationo f (Hetero)Aromatics and a,b-Unsaturated (Acrylic) Acids Compared to carboxylases and decarboxylases (carboxylyases,E C4 .1.1.X) acting on amino acids,c arbo-hydrates (including oxalate and pyruvate) and other aliphatic substrates,w hich are widespread in primary metabolic pathways,l ess enzymes are known to catalyze the (de)carboxylation of aromatic substrates, found in secondary metabolism. Among them, ar emarkable number of enzymes for potential biocatalytic applications has been identifiedwhich can be classified into three major categoriesb ased on mechanistic aspects( Figure 1): (i)d ivalent metal-dependent decarboxylases from the amidohydrolase superfamily,( ii) cofactor-andm etal-independent phenolic acid decarboxylases,a nd (iii)p renylated FMN-dependent (prFMN) decarboxylases from the UbiD superfamily. Thel atter display activity for aw ide array of structurally diverse substrates [ Figure 1, (iii)] and are associated with the reversible (de)carboxylationo fp henols (highlighted in orange), heteroarenes (green), a,b-unsaturated( acrylic) acids (yellow), and other arenes (blue), respectively.S ince prFMNw as discovered only recently the biocatalytic characterization of prFMN-dependent decarboxylasesi sa ti ts early stage assuming that the identification of furthere nzyme candidates is most likelya nd their use for biocatalytic applicationi sa ni ntriguing future aim. Several3 ,4- [26] and 4,5-dihydroxyphthalate [27] decarboxylases have been identified, however, it is not cleary et whether these enzymesa re members from the UbiD superfamily.
Owing to its stability and straightforward reconstitution with the prFMN cofactor, the Fdc subfamily can be regarded as the most versatile and applicable decarboxylation biocatalyst of the UbiD family as of now.

Divalent Metal-Dependent Decarboxylases
All of the metal-dependent decarboxylases [ Figure 1, (i)] identified so far are members of the amidohydrolase superfamily (AHS). They share significant structural and mechanistic similarities,i np articular the characteristic (b/a) 8 -barrelf old harbouring one catalytically relevant divalent metal ion in the active site. [30] Whereast he overall sequence similarity between distinct subclassesi sr atherl ow (around 30%), [31,32] several amino acid residues relevant for catalysis are conserved. Although members of the AHS commonly catalyze the hydrolysis of ester and amide bonds attachedt oe ither ac arbon or phosphorus atom on aw ide range of structurally diverse substrates, [30] some members obviously evolved to cata-lyze the reversible decarboxylation of benzoic acid derivatives and nitrogen-heterocyclic derivatives thereof.
Density functional theory (DFT) calculations employinglargeactive site models based on crystal structures of ortho-benzoic acid decarboxylases (o-BDCs) strongly support ag eneralm echanistic proposal, which resembles a( reverse) electrophilic aromatic substitution by feasible energy barriers and bears a strong resemblance to the Kolbe-Schmitt reaction. [31,32] In more detail, this general mechanismi nvolvest he metal ion (predominantly manganese or zinc) chelating the carboxylate andp henolate group of the arene substrate,t hereby stabilizing their negative charge.C oordination arranges the nucleophilic phenolate for protonation at the carboxylate ipsocarbon by an earbyc onserved catalytic acid (Asp), which is accompanied by dearomatization of the arene.T he latter is restoredu ponl oss of carbon dioxide,w hich dissociates together with the decarboxylated phenolf rom the metal center (Scheme2). [31,32,33] Thes tability and broad substrate tolerance of o-BDCs [8,21] (such as 2,3-dihydroxybenzoica cid decarboxylases from Aspergillus and Fusarium species,s alicylic acid decarboxylase from Trichosporon moniliiforme,a nd 2,6-dihydroxybenzoic acid/g-resorcylate decarboxylases from Agrobacterium, Rhizobium, Pandoraea, Rhodococcus and Polaromonas species) [29a,34] recommended them as excellent biocatalystsf or the regioselective ortho-carboxylation of phenols also on ap reparative scale.T his also applies for 5-carboxyvanillate decarboxylases (LigWs,f rom Sphingomonas and Novosphingobium species), althought hey exhibit am ore restricted substrate tolerance. [21,35] Them inimal structural requirements (shown in Figure 2) are characterized by ap henolic motif,i nw hich the aromatic systems upportst he resonance stabilization of the carbanion intermediate.T he phenolic OH group seems to be mandatory,s ince NH 2 (aniline) and SH variants (thiophenol) are not accepted due to inaccurate electronic (lower or significantly higher pK a of the SH or NH protons, respectively) and/or structural (atomic diameter) properties.C arboxylationi si nevitably associated to af ree ortho-position. [21,33a] Regarding the substitution pattern,t he meta-position (m 2 , Figure 2) opposite to the carboxylation site is most flexible tolerating weakly e À -withdrawing (halogens) and in particulare À -donating groups (alkyl, alkoxy, hydroxy,a nd amino functionalities), which can be even extended to (conjugated) aromatic systems to encompass remarkably large polyphenols,s uch as resveratrol. [33a,34b] The meta-position (m 1 )a djacent to the carboxylation site is less flexible,o nly OH and CH 3 substitution is reported.T he para-position favours weak e À -donorsa nd -acceptors ands eemst ob e uniquei nt he acceptance of strongly e À -withdrawing carbonyl/carboxyl moieties as wella se longated (un)saturatedpropionic/acrylic acids. [35] Substituents in the non-reactive ortho-position (o 2 )a re well accepted as long as they are small. Thet olerance of multiple functionalized substratesi sd iverse and depends on the electronic ands teric nature of the substituents.I n general, the electronic properties of the functionalization seem to play am ore significant rolet han steric effects.
In contrast, the biocatalytic characterizationo fa nother member of this class of enzymes -iso-orotate decarboxylase (IDCase)-acting on ah eterocyclic substrate analogue,r evealed av ery narrow substrate specificity and only acts in the energetically favoured decarboxylation direction. [31] Process engineering,s uch as the addition of quaternary ammonium salts to induce the precipitation of the corresponding carboxylatedp roducts as ion pairs [36] as wella se nzyme engineering, e.g.,f or paraaminosalicylic acid production [34d] significantly improvedt he efficiency and operability of biocatalytic ortho-carboxylations,w hich is particularly important for their industrial application.
PA Ds strictly require af ully conjugateds ystemb etween the C b carbon atom andt he mandatory parahydroxy group to facilitate as ubstrate-based resonance stabilization of the negative charge via aq uinone methide intermediate in the acid-base-catalyzed (de)carboxylation reaction (Scheme 3). [41,42] Theo verall robustness and substrate tolerance of PA Ds is more limited comparedt ot hose of o-BDCs. [21,43] Thei ntroductiono ff unctionalization is mainly restricted to the ortho-positions (o 1 and/or o 2 ) tolerating e À -donating (alkyl, alkoxy) and-withdrawing groups (halogens), whereas meta-substitution (m 1 , m 2 )l ed to unstable products. Variations in the a-o r b-position of the styrene side chain as wella sr eplacement of the para-OH group (e.g.,b yC l, OMe,N H 2 ) is prohibited and causes at otal loss of enzymatic activity.I ng eneral, the steric properties of the substitu- Scheme 3. Generalc atalytic acid-base mechanism for the side chain (de)carboxylation by cofactor-independent PA Ds. ents seem to play am ore crucial role than electronic effects (Figure 3).

prFMN-Dependent Decarboxylases
Ther emarkably broad substrate tolerance,e xcellent regioselectivity,a nd the robustness of most ortho-benzoic acid decarboxylases facilitate their widespread utilization in biocatalytic applications for the orthospecific( de)carboxylationo fp henolic substrates towards the corresponding aromaticc arboxylic acids as abiocatalytic equivalenttothe Kolbe-Schmitt [44] reaction. An enzymatic route to the regiocomplementary para-carboxylation of phenols is highly desired since the chemical counterpart suffers from unsatisfactory regioselectivities.
para-Selective CO 2 fixationw as described for phenylphosphate carboxylases (PPC), which require ATP-dependent activation of the phenol substrate via phosphorylation prior to the carboxylation step (Scheme 4). [45] Twoe nzymes from Pseudomonas strain K172 (PsPPC) [45a,46] and Thauera aromatica (TaPPC,M n 2+ + dependent) [45d,47] have been purified and characterized. Although the biocatalytic applicability of TaPPCw ith aT ON of up to 16000 was demonstrated after stabilization of the oxygen-sensitive enzymesb yi mmobilization on Agar beads, [45d] the scope of this enzyme class is limited to phenylp hosphate andc atechyl phosphate [48] substrates.B oth the narrow substrate specificity andt he dependence on expensiveA TP limit the usability of these enzymes for biocatalytic applications.
Just recently, an alternative to the ATP-dependent para-carboxylation was discovered. With the identification ands ubsequent characterizationo fanovelc ofactor, first described as modifiedF MN by Marsh et al. [49] and later identifieda sp renylated flavin mononucleotide (prFMN) by Leys et al., [50] new concepts for biocatalytic (de)carboxylation strategies emerged. As ac onsequenceo ft hese seminal studies,n umerous decarboxylases of the UbiD superfamily [named after UbiD involved in the ubiquinone( co-enzyme Q) biosynthesisp athway in prokaryotes [51] ]d epending on this modified flavin cofactor were identifieda nd partially characterized. [22a] Biosynthetically,t he prFMN cofactori sp rovided by an associated prenyltransferase (UbiX) under anaerobic conditions.T his enzyme uses the unusualC 5 -metabolite g,g-dimethylallyl monophosphate( DMAP) as co-substrate to build up the prenylm oiety resulting in af ourth six-membered ring between N-5 and C-6 of the isoalloxazine ring systemo ft he reducedf lavin[ Scheme 5, (b)].I nc ontrast, the fungal UbiX analogue Pad1 uses the (metabolicallym ore common)d iphosphate DMAPP instead of DMAP for producing prFMN. [52] In ordert o obtain the catalytically active iminiums pecies of the cofactor( prFMN iminium ), oxidative maturationo fr educed prFMN (prFMN reduced )b ym olecular oxygen in the presence of apo-decarboxylase appears to be crucial [Scheme 5, (b)]. [53] Although the exact oxidation mechanism of the reducedU biX product to prFMN iminium is unknown as of yet, aG lu-Arg-Glu motif as catalytic acid in the active site of the decarboxylase seems to assist this process. [54] Notably,t he prenylm odification in its catalytically active iminium form entirely eliminates the chemical properties of FMN as ar edox mediator for hydride transfer and confers a1 ,3-dipolar azomethine ylide and electrophilic iminium ion character,r espectively [Scheme 5, (b)]. [55]
Overall, the minimum structural substrate requirements are marked by the obligatory conjugation of the E-configured a,b-C=Cb ondi nvolved in the 1,3dipolarc ycloaddition to (at least) one other C=C bond, which may be part of al arger (aromatic) psystem.
Attempts to carboxylate alkenesmatching the minimal substrate requirements (using bicarbonate or CO 2 as co-substrate) were unsuccessful. [62] Figure 6. Substrate scope of AnFdc, ScFdc and CdFdc (applied as purified enzyme or lyophilized E. coli whole-cell preparation containingthe heterologously expressed decarboxylase).
[d] Oxygen-sensitive protein and/or anaerobic strain.
[f] Carboxylation was tested, but no product acid was detected.
Besides the biocatalytically characterized prFMNdependent AroY enzymes,r elated UbiD-like (and hence putatively prFMN dependent) para-carboxylases from various microbials ources can be found in the literature.T able 2g ives an overview of (partially characterized) para-carboxylases and their substrate and non-substrate scope.T he preference for ac ertain substrate allows classification into 4-hydroxybenzoate, protocatechuate (including AroY), gallate and vanillate decarboxylases.H owever, the oxygens ensitivity of some of these proteinss everely impedes their applicationinb iotransformations.
Recent discoveries in the organization of ubiD/ ubiX genes and other associated genese ncoding for proteinso fy et unknown function, [22a,65] as well as the knowledge of in vitro activation with the catalytically active cofactoro utlined above opens newp ossibilities for prFMN-dependent phenol para-carboxylases for biocatalytic purposes.
Polycyclic aromatich ydrocarbons (PAHs): The degradation of polycyclic aromatich ydrocarbons (PAHs) in polluted soil and aquifers under anoxic conditions proceeds via enzymatic carboxylation as one of the initial steps towards mineralization of these environmental hazards. Meckenstock et al. provide ac omprehensive overview of the recent developments in this field, with as pecial focuso nm icrobiology,e cology,a nd biochemistry. [82] Although research is still in its infancy,t hese enzymes pose highly interesting biocatalystsg iven their potential ability to directly functionalize chemically non-activated aromatics under mild, anoxic conditions to yield the corresponding carboxylic acids without the need for ATPo r other expensive co-substrates.C arboxylated metabolites of benzene, [83] biphenyl, [84] naphthalene [85] and phenanthrene [86] were detected in supernatants of anaerobic sulfate-reducing enrichment cultures [ Figure 7, (a)].I namore detailed study with crude cell extractso fc ultureN 47, the carboxylation of naphthalene in the 2-position was demonstrated to involve incorporation of 13 Cb icarbonatei nto the product. Dynamic enzyme-catalyzed carboxylate 13 C-isotope label exchange suggests reversibility of the process [ Figure 7, (b)]. [87] How exactly carboxylation of non-activated aromatics lacking electron-donating substituents is accomplishedi nt he active site of these enzymesi sasubjecto fo ngoing research. Ac rude mechanistic proposal suggestsn ucleophilic attack at the 2-position of the arene with CO 2 as electrophile, while the resulting carbocation is delocalized in the (annulated) aromatic system( S E Ar) [ Figure 7, (c)]. [82] In any case,t he carboxylation of non-activated aromatics is a difficult reaction, as domonstrated by the very low reaction rates. [82] Genetics tudies of the benzene-induced putative anaerobic benzene carboxylases reveal ac luster organization with two subunits (abcA and abcD) and a gene sequence similar to that of ATP-dependent phenyl phosphatec arboxylases from A. aromaticum of the UbiD family. [88] In addition,g enes similar to the ubiD/ubiX system were highly transcribed in benzene-degrading nitrate-reducing enrichment cultures, [89] which suggestsapossible role of prFMN in the carboxylation of PA Hs.
Heteroarene decarboxylases with sequences related to other UbiDm embers [22a,99] [see Figure 1, (iii)],a nd hence putative dependency on prFMN,h aveb een discovered but the mechanisticr ole of this cofactor in the decarboxylation of heterocyclic substrates is unknown so far.
An UbiD-like decarboxylase,Y clC from Bacillus amyloliquefaciens,i si nvolved in the decarboxylation of indole-3-pyruvate in the biosynthesis of the phytohormone indole-3-acetic acid [ Figure 9, (g)]. [100] Another enzyme purified from Alcaligenes sp. strain UK21 is involved in the highly substrate specific non-oxidative decarboxylation of 6-hydroxyquinolinate (85a)t o6 -hydroxypicolinic acid (85b)[ Figure 9, (h)-(i)].T he decarboxylase shows some similarity to phthalate decarboxylases of the UbiD family,a lthoughn or equirement for cofactors was reported. Attempts to run the process in the reverse carboxylation direction were unsuccessful. [101] Recently,a lso the decarboxylation of phenazine-1carboxylic acid (99a)b yt he UbiD-like enzyme PhdA from Mycobacterium fortuitum associated to the
In the presence of elevated concentrations of CO 2 [1 Mb icarbonatea nd/or pressurized CO 2 (32 bar)] the reversedc arboxylation reactiono f2 -furoic acid (100b)t o100a waso bserved employingF DCA (2,5furandicarboxylica cid) decarboxylase (HmfF/HmfG) from Pelotomaculumt hermopropionicum albeit with low yield. [104] In retrospect, little is known about biocatalytic heteroarene( de)carboxylation as of yet, but with the advent of prFMN-dependent catalysis andt he knowledge available for providing active reconstituted enzymes,t he conversion of these substrates by UbiDfamily members poses an interesting research aim.

4C onclusion
Overall, enzyme-catalyzed (de)carboxylation concepts have been established as suitable synthetic tools, using carbon dioxidea sC 1 -building block for the productiono fv aluablec hemicals.T he biocarboxylation of electron-rich (hetero)arenesp roceeds under mild reactionc onditions in ah ighly regioselective fashion. As such, these methodso ffer particularb enefits over chemical methodsl ike the Kolbe-Schmitt reaction, which often require high pressure and temperature and suffer from incomplete regioselectivities.
Metal-dependent ortho-benzoic acid decarboxylases and cofactor-independent phenolic acid decarboxylases are remarkably stable under biotransformation conditions,a nd have been exploited for the regioselective carboxylation of ar ange of electronically and structurally diverse phenolic and coumaric acid derivatives,r espectively.D espite that, process engineering needst ob ef urther developedt od rive the equilibrium in the thermodynamically disfavoured carboxylation direction in order to facilitate large-scale applications.
Ever more prenylated FMN-dependent decarboxylases of the widespread UbiDf amily are beingd iscovered, which catalyze highly interesting preparative reactions.E xamples include the (de)carboxylation of (hetero)aromatic substrates and extensiono ft he biocatalytic toolbox towards the decarboxylation of cin-namic and 2,4-hexadienoic acid derivatives.N onetheless,d ue to the recent discovery of the modified FMN cofactort he investigation of the chemistry impliedb y the prFMNs tructure is as ubjecto fo ngoing research. Hence, at present several limitations for prFMN-dependent decarboxylases need to be addressed in order for this highly interesting enzyme class to become a widelya pplicable biocatalytic tool for organic synthesis.T hese limitations include,b ut are not limited to, the light sensitivity of prFMN iminium ,t he O 2 sensitivity of some UbiD family members,c o-expression of the prenylt ransferase UbiX, as well as efficient in vitro reconstitutiono fapo-decarboxylases as well as the role of other allosteric factorsduring catalysis.