Diversifying Metal–Ligand Cooperative Catalysis in Semi‐Synthetic [Mn]‐Hydrogenases

Abstract The reconstitution of [Mn]‐hydrogenases using a series of MnI complexes is described. These complexes are designed to have an internal base or pro‐base that may participate in metal–ligand cooperative catalysis or have no internal base or pro‐base. Only MnI complexes with an internal base or pro‐base are active for H2 activation; only [Mn]‐hydrogenases incorporating such complexes are active for hydrogenase reactions. These results confirm the essential role of metal–ligand cooperation for H2 activation by the MnI complexes alone and by [Mn]‐hydrogenases. Owing to the nature and position of the internal base or pro‐base, the mode of metal–ligand cooperation in two active [Mn]‐hydrogenases is different from that of the native [Fe]‐hydrogenase. One [Mn]‐hydrogenase has the highest specific activity of semi‐synthetic [Mn]‐ and [Fe]‐hydrogenases. This work demonstrates reconstitution of active artificial hydrogenases using synthetic complexes differing greatly from the native active site.


Introduction
Artificial enzymes incorporating non-native metal cofactors while exhibiting native activities can serve as important tools for mechanistic studies of enzymes. [1] They also have potential applications in biocatalysis by increasing availability of hard-to-access enzymes as well as overcoming substrate specificity. [1a,2] Whereas many artificial enzymes have been developed to bring abiological reactivities, [3] only few of them catalyze the native reactions of wild-type enzymes. [2,4] [Fe]-hydrogenase catalyzes the hydrogenation of methenyl-tetrahydromethanopterin (methenyl-H 4 MPT + )t of orm methylene-H 4 MPT and its reverse reaction. Thef orward reaction is ap art of the pathways of microbial methanogenesis. [5] [Fe]-hydrogenase hosts the FeGP cofactor (Figure 1A)w here an Fe II center is coordinated to ap yridone derivative,two cis-CO ligand, acysteine residue,and awater molecule. [6] We recently reconstituted an active [Mn]-hydrogenase by incorporating aM n I mimic of the active site (1a) with the apo-enzyme of [Fe]-hydrogenase. [1d] The[ Mn]hydrogenase exhibited slightly higher specific activity per active metal site than that of as emi-synthetic [Fe]-hydrogenase incorporating an Fe II complex analogue (2a). [1e] The synthesis of acatalytically active [Mn]-hydrogenase paved the way for the current structure-activity study,where the active site is precisely modified via small molecule synthesis.A n analogous study using Fe II mimics has been challenging due to their lower stability compared to the Mn I ones. [7] Them echanisms of H 2 activation in previously reported [Fe]-and [Mn]-hydrogenases all involve metal-ligand cooperation, where ab asic 2-O À group generated by deprotonation of the 2-OH group of an pyridone derivative deprotonates the H 2 molecule coordinated to the metal ion (Figure 1B). [1d,e,6a] In this study,wedesign and synthesize aseries of Mn I complexes with or without an internal base or pro-base ( Figure 1C). Among the former complexes,t he nature and position of the internal base or pro-base are different. Only Mn I complexes containing an internal base or pro-base are active for H 2 activation. And only such complexes could be used to reconstitute an active [Mn]-hydrogenases.B ased on the nature and position of the internal base or pro-base of their Mn I cofactors,c ertain [Mn]-hydrogenases operate by anon-native metal-ligand cooperation mode.One such [Mn]hydrogenase exhibits the highest activity among semi-synthetic [Mn]-and [Fe]-hydrogenases.T he work demonstrates the construction of functional hydrogenases using synthetic complexes that operate via an ew-to-nature reaction mechanism. This strategy might be used to expand the functions of Hmd hydrogenases beyond native reactions using novel substrates other than methenyl-/methylene-H 4 MPT and H 2 .

Results
Subsequent to the work with 1a,w ed esigned as eries of new Mn I models (1b-1e)t hat could be used to probe the influence of an internal base for H 2 activation. Complex 1b is an analogue of complex 1a in which the two acidic protons on methylene group are replaced by two methyl groups.Like 1a, this complex contains a2 -OH group that can serve as an internal base upon deprotonation. Complex 1c has a2 -OMe group which cannot serve as an internal base,b ut the methylene moiety at the 6-position is enolizable and can potentially serve as an internal base upon deprotonation. Complex 1d has acarbamoyl group where the tertiary amide moiety can serve as an internal base.C omplex 1e has no potential basic site and servers as ar eference.
Thes ynthesis of the Mn complexes is summarized in Figure 2A.A na ppropriate pyridine derivate was firstly deprotonated by 1.0 equivalent of n-BuLi to give al ithiated salt, which was directly treated, without purification, with 1.0 equivalent of Mn(CO) 5 Br to yield Mn complexes 1c-1g. Them ethylene groups in 1f and 1g were methylated twice through ad eprotonation/methylation sequence to give complexes 1e and 1h.D eprotection of the MOM (MOM = methoxymethyl) group in 1h using 3.0 equivalents of BBr 3 afforded complex 1b.Deprotonation of the methylene group in 1c by reacting with 1.0 equivalent of KH gave complex 3, which contains an enolate ligand ( Figure 2A). By addition of 1.1 equivalents of 18-crown-6 to 3, 3(18-crown-6) was formed and isolated.
We estimated pK a of relevant OH, CH, and amide acids in DMSO using NMR titration experiments (Supporting Information, section 1.4). ThepK a of the 2-OH group in 1a and 1b were 9.0 AE 0.1 and 8.8 AE 0.1, respectively,that of the C(6)-Hmoiety in 1c was 13.7 AE 0.1, and that of the conjugated acid of 1d was between 1.6-3.6.
TheIRspectra of 1a, 1b and 1c exhibit four n(CO) peaks (Supporting Information, Table S1), consistent with the complexes having four CO ligands.Inthe IR spectra of 1dand 1e, however, only 3 n(CO) peaks are observed due to peak overlapping.I nt he 13 CNMR spectra of 1a-1e,t he four CO ligands give only 3peaks (Supporting Information, Table S2), indicating that the two trans-orienting CO ligands are equivalent on the NMR scale.O verall, the CO ligands in 1a-1e have similar n(CO) frequencies and 13 CNMR shifts, indicating as imilar electronic property among these complexes.T he substituents at the 2-position of the pyridine ligand have only aminor influence on the electron density at the Mn center. Upon deprotonation of the methylene group in 1c to form an enolate group in 3(18-crown-6), the 1 HNMR chemical shift of the remaining proton on the methylene carbon changed from 3.91 ppm to 5.19 ppm, consistent with the formation of an alkene moiety.T his interpretation is also supported by IR spectra, which indicates amore electron-rich Mn center in 3(18-crown-6) than in 1c (Supporting Information, Table S1).
X-ray crystallography confirms 1b-e and 3(18-crown-6) as Mn I tetra(carbonyl) complexes ligated by ab identate N-C ligand derived from pyridine ( Figure 2B). [8] Thecoordination geometry of Mn ions is best described as pseudo-octahedral. TheN -C ligand forms an on-planar five-membered metallocycle with the Mn ion. Among the four CO ligands,t he one trans to the pyridinyl Nligand forms the shortest M-CO bond (Mn-C4;S upporting Information, Table S3), consistent with N(pyridine) being the weakest trans ligand among N-(pyridine), C(acyl), and C(CO). TheM -CO distances for the other three CO ligands are rather similar (Supporting Information, Table S3). Mn ion and the ortho-O atom are confirmed metal-ligand cooperation sites.T heir distances in 1a and 1b are about 3.2 .N otably,t he distances of the Mn ion to the C6 or N2 atom in 1c and 1d,respectively,are only about 3.0 .T his short distance substantiates the possibility of the C6 and N2 atoms serving as an alternate metal-ligand cooperative site.U pon deprotonation of 1c to form 3(18crown-6), the C5-C6 bond shortened from 1.533(9) to 1.396(2) ,a nd the C5-O5 bond elongated from 1.216(7) to 1.250 (2) .T hese changes confirm the presence of acoordinating enolate moiety in 3(18-crown-6).
Only irreversible oxidation and reduction were observed in the cyclic voltammograms of 1a-e (Supporting Information, Figure S40). Judging from the peak potentials for the oxidation, the electron density of the Mn I ion follows the following order: 1a% 1b% 1c> 1e> 1d.
An H 2 /D 2 exchange assay was employed to probe the activities of 1b-1e and 3(18-crown-6) in H 2 activation (Table 1; for more information, see the Supporting Informa- tion, Section 1.3). Ther eactions were conducted under am ixed atmosphere of D 2 (8 bar) and H 2 (12 bar). Thef ormation of HD indicated H 2 activation. Thei nitial Mn complexes remained the only identifiable Mn species during and after the reaction;n oM n-H intermediates were observed. All complexes except 1e catalyzed H 2 /D 2 exchange in the presence of an external base (Table 1). Ther eactivities have the following order: 1a> 1b% 1c> 1d when N-methylpyrrolidine (MP) was used as the external base (entries 1-3, Table 1). Thel ower activity of 1b compared to 1a suggests that the two methyl groups at the C6 position of the ligand hinders H 2 activation (entries 1a nd 2, Table 1). Ther eactivity of 1c depended

Angewandte Chemie
Forschungsartikel strongly on the basicity as well as the quantity of the external base.I nt he presence of one equivalent of as trong base KO t Bu (entry 6, Table 1), no H 2 /D 2 exchange was observed.
Increasing the equivalent of KO t Bu to 2equivalents led to arapid H 2 /D 2 exchange (entry 7, Table 1). This result could be rationalized by considering that the first equivalent of KO t Bu deprotonated the methylene group in 1c to form an enolate complex similar to 3,w hich could catalyze H 2 /D 2 exchange only in the presence of an external base (the second equivalent of KO t Bu). To probe this hypothesis, 3(18-crown-6) was used as catalyst. Without an external base,n oH 2 /D 2 exchange took place (entry 11, Table 1). With 1.0 equiv of KO t Bu, there was still no rapid H 2 /D 2 exchange (entry 12, Table 1). This result was rationalized by the lack of proton in the system, which is necessary to mediate H/D exchange after H 2 /D 2 activation. Indeed, when 1.0 equiv of HO t Bu was added to the reaction system, fast H 2 /D 2 exchange was observed (entry 13, Table 1). For 1c,t he rate of H 2 /D 2 exchange was much higher in the presence of KO t Bu over MP.T his rate difference was attributed to an incomplete formation of the enolate complex in the presence of MP. Indeed, the pK a of the methylene proton in 1c was about 13.7 in DMSO,higher than that of MPH + (10.6). For 1a,asimilar rate of H 2 /D 2 exchange was observed in the presence of KO t Bu or MP (entries 1and 14, Table 1). This result could be rationalized considering that the 2-OH proton in 1a has apK a of 9.0 in DMSO so that both MP and KO t Bu are strong enough to fully deprotonate. We reconstituted semi-synthetic [Mn]-hydrogenases incorporating complexes 1a-1e as the metal co-factors following ap reviously established protocol. [1d] Due to the potential instability of Mn complexes,t he reconstitution experiments were performed as soon as possible after the chemical synthesis (see the Supporting Information). Tw o equivalents of aM nc omplex was dissolved in am ethanol solution containing 1% acetic acid. This solution was then mixed with asolution of one equivalent of the [Fe]-hydrogenase apoenzyme from Methanocaldococcus jannaschii heterologously produced in Escherichia coli in the presence of 2mMg uanosine monophosphate (GMP). We reported earlier that external GMP slightly increased the specific activity of semi-synthetic [Fe]-hydrogenase and [Mn]hydrogenase. [1d,e] Ther esulting [Mn]-hydrogenases were named as jHmd (wild)-1a,j Hmd (wild)-1b,j Hmd (wild)-1c,j Hmd (wild)-1d,j Hmd (wild)-1e,r espectively. Mutant enzymes were prepared in the same way but with the His14A and Cys176A apo-enzymes.T hey are named as jHmd (H14A)-1a,j Hmd (H14A)-1b,j Hmd (H14A)-1c,j Hmd (H14A)-1d,j Hmd (H14A)-1e,j Hmd (C176A)-1a,j Hmd (C176A)-1b,j Hmd (C176A)-1c,j Hmd (C176A)-1d,and jHmd (C176A)-1e.
Thea ctivities of [Mn]-hydrogenases were measured photometrically for both the forward (reduction of methenyl-H 4 MPT + with H 2 )a nd the reverse (H 2 production from methylene-H 4 MPT) reactions (Table 2). Ther econstituted jHmd (wild) enzymes other than jHmd (wild)-1e exhibited the Hmd activity.I nn ative [Fe]-hydrogenase,t he FeGP cofactor is covalently bound to the protein via the Cys176-Smetal coordination. In our previous report as well as this work, ad ramatic loss of activity was detected for jHmd (C176A)-1a (0.06 AE 0.07 Umg À1 )c ompared to jHmd (wild)-1a (7.8 AE 0.7 Umg À1 ). Thesmall residual activity suggests that other interactions such as hydrogen bonding between ametal complex and protein might lead to incorporation of am etal complex in the active site,a lbeit less efficiently.A dditional Cys176Ala mutation studies,h owever, suggest Cys176-metal coordination is required for most Mn complexes.Thus,jHmd (C176A)-1b,j Hmd (C176A)-1c,j Hmd (C176A)-1d and jHmd (C176A)-1e were totally inactive in both forward and reverse reactions ( Table 2, entries 12-15). Based on these properties,w ea ssumed that through reconstitution, the Mn complexes were specifically incorporated into the active site, in which Mn complex coordinated with Cys176-S.W ecannot exclude the possibility that Mn complexes bind at the active site without Cys176S-Mn bonding though. Infrared spectrum of the reconstituted enzymes were similar to those of the free Mn complexes without protein, which indicated that sub-

Angewandte Chemie
Forschungsartikel stantial Mn complex was non-specifically adsorbed on other parts of the protein (Supporting Information, Figure S41). Previously we fortuitously obtained as ample where the majority of adsorbed Mn complex was incorporated into the active site.T he frequency of the two CO bands are similar to those of the native [Fe]-hydrogenase,w hich supported the binding mode of the Mn complex to the protein by Cys176-Smetal bonding.B ased on the intensities of v(CO) band and the amide II band of the protein, we were able to estimate the occupancyofthe active site of samples of jHmd (wild)-1a. [1d] Previous study shown that non-specifically adsorbed Mn complexes had no contribution to the detected enzymatic activities. [1d] In this study,f or [Mn]-hydrogenases incorporating other Mn complexes,wewere not able to obtain samples with only specifically adsorbed complexes.T hus,w ed eveloped an alternative and more general method for estimating the occupancyo ft he active site (Supporting Information, Section 2.2). We assume that jHmd (C176A) cannot bind aM n complex in the active site as the sulfur ligand from the Cys176 is removed. Thev(CO) peaks in reconstituted jHmd (C176A) samples could be attributed to non-specifically bound Mn complexes.M eanwhile,t he v(CO) peaks in reconstituted jHmd (wild) samples were due to both specifically and nonspecifically adsorbed Mn complexes.B ys ubtracting the amount of non-specifically bound complexes,w eo btained the occupancies of the reconstituted jHmd (wild) samples (Scheme S2 for methods and Table 2a nd the Supporting  Information, Table S5 for results). We acknowledge as ubstantial level of uncertainty in this analysis due to,f or example,alimited spectral resolution or the possibility that jHmd (C176A) and jHmd (wild) would unspecifically bind adifferent amount of agiven Mn complex. Nevertheless,this analysis shall give correct qualitative trends.The occupancyof jHmd (wild) 1aobtained by this method (20.3 %) is nearly the same as that estimated by the previous method (20 %). [1d] The occupancies of jHmd (wild) 1a, 1c and 1d were similar (from about 15 %to25%). Notably,reconstitution using complexes 1b and 1e bearing two methyl groups a-to the acyl donor had much higher occupancies (70.9 %a nd 58.8 %, respectively).
Themeasured specific activities were divided by the active site occupancies to give the actual specific activities,w hich could be compared among different enzymes.
[ Mn]-hydrogenases reconstituted with 1b-1d,a long with 1a,w ere all catalytically active for at least the forward reaction. The[Mn]hydrogenase reconstituted with 1e was inactive for both forward and reverse reactions.
Thea ctivity of jHmd (wild)-1a determined on samples prepared in this study (7.8 AE 0.7 for the forward reaction) was nearly identical to that of previously reported sample. [1d] The activity of jHmd (wild)-1b (1.1 AE 0.17 Umg À1 for the forward reaction) was about 14 %ofthat of jHmd (wild)-1a,suggesting that the steric hindrance due to the two additional methyl groups at the C6 position of the ligand slows down the reaction. jHmd (wild)-1d,w hich contains at ertiary amide basic site,h ad similar activity (9.3 AE 1.6 Umg À1 for the forward reaction) as jHmd (wild)-1a.The most active enzyme was jHmd (wild)-1c whose precursor complex has an enolate basic site upon deprotonation. Its actual specific activities for the forward reaction and reverse reaction were 37 AE 17 Umg À1 and 37 AE 16 Umg À1 ,w hich are about 4a nd 37 times higher than those of jHmd (wild)-1a,respectively.T he activities of jHmd (wild)-1c were about 5t imes and 8t imes higher than those of semi-synthetic [Fe]-hydrogenase reconstituted with Fe complex 2a. [1e] These activities were about 10 %ofthose of native [Fe]-hydrogenase which has the FeGP cofactor. Note that semi-synthetic [Fe]-hydrogenase reconstituted with an analogous Fe complex 2b ( Figure 1A)w as inactive.W ea ttributed the difference in activity to the instability of 2b upon deprotonation of the methylene(acyl) group as well as the generally lower activity of Fe mimics compared to Mn mimics. [7a, 9] In the native [Fe]-hydrogenase,H is14 is proposed as ap roton acceptor from 2-OH of pyridinol of the FeGP cofactor. [6a] To probe the function of His14 as abase and apart of the proton relay (Supporting Information, Figure S45), we conducted reconstitution studies using His14Ala mutant of the apoenzyme.T he resulting [Mn]-hydrogenases,j Hmd (H14A)-1n (n = a-d)h ad essentially no detectable activity for both forward and reverse reactions except for jHmd (H14A)-1d (Table 2). This result suggests that His14 is an ecessary base to deprotonate not only 2-OH but also methylene(acyl) groups to form an internal O À or enolate basic sites in the Mn complexes.Incontrast, jHmd (H14A)-1d retained substantial activity of jHmd (wild)-1d.T his result could be understood by considering that the Mn complex 1d has already ab asic amide site,a nd once protonated, the proton on it is rather acidic and can be transferred by H 2 O molecule.

Discussion
Ther esults in Table 1i ndicate that an internal base in aM nc omplex is necessary for H 2 activation. TheH 2 /D 2 exchange was observed only in the presence of an external base because this exchange requires aH + /D + exchange in addition to H 2 activation.
It was previously shown, based on DFT computations, that the mechanism of H 2 activation on 1a involved first deprotonation of the 2-OH group,followed by dissociation of aCOligand trans to the acyl ligand, H 2 coordination, and then intramolecular heterolytic H 2 cleavage ( Figure 3A). [1d] A same mechanism is expected for H 2 activation on 1b,but not on 1c and 1d where the internal bases (C6 or N2) are located at the 6-position of the pyridine moiety.F or the latter two catalysts,wepropose an analogous mechanism. In the case of 1c ( Figure 3B), deprotonation of the catalyst gives ac oordinated enolate anion. Replacement of aC Ocis to the acyl ligand by H 2 then yields aM n-H 2 complex where the H 2 can be heterolytically cleaved by the cooperative action of the O anion of the enolate and the Mn center. In support of this mechanism, we found significant Di ncorporation at the methylene position during both H 2 /D 2 exchange assay and H + /D 2 exchange assay experiments (Supporting Information, Figure S14). As imilar mechanism for H 2 activation was previously proposed for Mn [10] as well as Fe and Ru complexes. [11] Note that deprotonation of methylene(acyl) Angewandte Chemie Forschungsartikel group by ab ase was also reported for Fe models of [Fe]hydrogenase,b ut the resulting enolate anion did not have ac atalytic role. [9] In the case of 1d (Supporting Information, Figure S13), no deprotonation is needed to create abasic site. Thus,t he reaction occurs by replacement of aC Oby H 2 followed by heterolytic H 2 cleavage via aMn-O cooperation.
Ther esults in  Table S6 for estimated TOFs). Thes mallest difference was found for jHmd (wild)-1c and 1c (with 2.0 equiv KO t Bu, entry 7T able 1) but the difference is still large;t he former has aT OF of about 24 s À1 ,w hile the latter has aT OF about 0.03 s À1 .T he differences presumably highlight the essential role of protein environment for efficient catalysis.
Ther eactivity trend of [Mn]-hydrogenases does not correlate with that of the Mn complexes alone.D ifferent multi-step reactions and different conditions are involved in enzyme and H 2 /D 2 exchange assays.T he observed reaction rates do not reflect the reactivity towards the same reaction step,i.e., hydrogen activation, because energetics and kinetics of other steps such as deprotonation or hydride transfer also contribute to the overall rate.I nterestingly,t he most active [Mn]-hydrogenase contains aM nc omplex (1c)t hat is most active for H 2 /D 2 exchange under astrong base.Once activated (deprotonated), this complex has the strongest internal base. We hypothesize that as trong internal basic site is beneficial for metal-ligand cooperative H 2 activation as long as this basic site is easily accessible in the catalytic cycle.T he reactivity trend of [Mn]-hydrogenases,h owever, does not linearly correlate with the basicity of internal bases.F or example, jHmd (wild)-1a and jHmd (wild)-1d have similar activity,but the internal bases at their active sites have quite different basicities.W eh ypothesize that for these cases the overall  ( Figure 4), the acyl ligand of aM n complex is distal to His14, which is the orientation found for FeGP in the native [Fe]-hydrogenase.T he 2-OH group is close to His14 so that proton relay from 2-OH to His14 to form ab asic O À group is facile.H 2 is activated through amechanism described in Figure 3A.This model can explain the reactivity of jHmd (wild)-1a and jHmd (wild)-1b.I n model B ( Figure 4), the acyl/carbamoyl ligand of aM n complex is proximal to His14. Proton relay from X(X= Cor N) to His14 as well as from Ot oXis again facile.H 2 is activated through am echanism described in Figure 3B.T his model can explain the reactivity of jHmd (wild)-1c and jHmd (wild)-1d.The pK a of the 2-OH group in 1a and 1b,aswell as that of the methylene group in 1c,i se xpected to be much lower in water than in DMSO.Asareference,acetic acid has ap K a of 12.3 in DMSO but only 4.8 in water. Thus,p roton relay of these groups with His14 in water appears feasible.
The[ Fe]-hydrogenases containing the FeGP cofactor or an iron mimic, as well as jHmd (wild)-1c,catalyze the forward and reverse reactions in asimilar rate. [1e] However,for jHmd (wild)-1a,j Hmd (wild)-1b,a nd jHmd (wild)-1d,t he reverse reaction was either much slower than the forward reaction or could not be detected. Ther eactions were conducted under  the same conditions for both [Mn] and [Fe]-hydrogenases: the pH for the forward reaction was 7.5 while that for the reverse reaction was pH 6.0. We suspect that the much faster forward reactions for jHmd (wild)-1a,j Hmd (wild)-1b,a nd jHmd (wild)-1d under the employed assay conditions originates from the non-optimal conditions to generate the internal proton relay for the reverse reactions on jHmd (wild)-1a,jHmd (wild)-1b,and jHmd (wild)-1d.The internal bases in 1a, 1b,a nd 1d are weaker than that in 1c.I ti s possible that at pH 6.0, the former basic sites are nearly always in the basic forms,i .e., 2-O À for 1a and 1b and C(O)N(Me) for 1d,w hich limits the turnover rate for the reverse reaction.

Conclusion
We have developed as eries of Mn I mimics of the active site of [Fe]-hydrogenase.T hese complexes are designed to have either a2 -OH group,a ne nolizable CH 2 group,o r atertiary amide group that can serve as an internal base,orno internal base.Only the complexes with internal bases are able to heterolytically cleave H 2 .I ncorporation of the Mn complexes into the apo-enzyme of [Fe]-hydrogenases result in semi-synthetic [Mn]-hydrogenases.T he Mn complexes are predicted to bind to the protein via the Cys176 residue.Only the [Mn]-hydrogenases containing Mn complexes with internal bases are catalytically active for the native reactions of [Fe]-hydrogenase.H 2 activation by both Mn complexes alone and [Mn]-hydrogenases involves metal-ligand cooperation, where ab asic ligand moiety serves as an internal base to deprotonate coordinated H 2 .I na ddition to the cooperation mode found in [Fe]-hydrogenase which involves the 2-O À group of the pyridone ligand of the active site,n ew cooperation modes involving either an enolate or an amide at the 6-position of the pyridone ligand as the internal base are operative.M oreover,t he [Mn]-hydrogenase where an enolate moiety serves as the internal base is about 5times and 8t imes more active (forward reaction) than its Mn and Fe counterparts where a2 -O À group serves as the internal base, respectively.T his result represents ag ood example where an ew-to-nature reaction mechanism [12] results in substantial activity for the native reaction of an enzyme.T he work indicates the possibility to incorporate synthetic complexes distinct from the native active site to reconstitute active Hmd hydrogenases,w hich can potentially open up new areas of applications such as activation of other small molecules and organic synthesis.