International Journal of Molecular Sciences Article Photocatalytic Hydrogen Production and Carbon Dioxide Reduction Catalyzed by an Artificial Cobalt Hemoprotein Guillermo A. Oliveira Udry 1 , Laura Tiessler-Sala 2 , Eva Pugliese 1 , Agathe Urvoas 3 , Zakaria Halime 1 , Jean-Didier Maréchal 2 , Jean-Pierre Mahy 1, * and Rémy Ricoux 1 1 UMR 8182, CNRS, Institut de Chimie Moleculaire & des Matériaux d’Orsay, University Paris-Saclay, F-91405 Orsay, France 2 Departament de Química, Universitat Autònoma de Barcelona, 08193 Bellaterra, Spain 3 CEA, CNRS, Institute for Integrative Biology of the Cell (I2BC), University Paris-Saclay, F-91198 Gif-sur-Yvette, France * Correspondence: jean-pierre.mahy@universite-paris-saclay.fr Abstract: The covalent insertion of a cobalt heme into the cavity of an artificial protein named alpha Rep (αRep) leads to an artificial cobalt hemoprotein that is active as a catalyst not only for the photo-induced production of H2 , but also for the reduction of CO2 in a neutral aqueous solution. This new artificial metalloenzyme has been purified and characterized by Matrix Assisted Laser Desorption Ionization-Time of Flight Mass Spectrometry (MALDI-TOF MS), circular dichroism, and UltraViolet–Visible spectroscopy. Using theoretical experiments, the structure of this biohybrid and the positioning of the residues near the metal complex were examined, which made it possible to complete the coordination of the cobalt ion by an axial glutamine Gln283 ligand. While the Co(III)–porphyrin catalyst alone showed weak catalytic activity for both reactions, 10 times more H2 Citation: Udry, G.A.O.; Tiessler-Sala, and four times more CO2 were produced when the Co(III)–porphyrin complex was buried in the L.; Pugliese, E.; Urvoas, A.; Halime, hydrophobic cavity of the protein. This study thus provides a solid basis for further improvement of Z.; Maréchal, J.-D.; Mahy, J.-P.; these biohybrids using well-designed modifications of the second and outer coordination sphere by Ricoux, R. Photocatalytic Hydrogen site-directed mutagenesis of the host protein. Production and Carbon Dioxide Reduction Catalyzed by an Artificial Keywords: artificial metalloenzymes; cobalt porphyrin; proton reduction; CO2 reduction; catalysis Cobalt Hemoprotein. Int. J. Mol. Sci. 2022, 23, 14640. https://doi.org/ 10.3390/ijms232314640 Academic Editors: Angela Lombardi, 1. Introduction Flavia Nastri and Linda Leone One of the greatest challenges facing our society is to replace current fuels with new, Received: 4 November 2022 abundant, and renewable sources of energy. Hydrogen (H2 ) seems to be the ideal energy Accepted: 21 November 2022 carrier because, in addition to its highly energetic H–H bond, its environmentally friendly Published: 24 November 2022 combustion releases only water as a product. H2 production processes are numerous (thermal, photochemical, biological, electrolytic, etc.) and the resources used in these Publisher’s Note: MDPI stays neutral processes are equally as abundant (water, solid waste, cellulose, biomass, etc.). To meet with regard to jurisdictional claims in the fast-growing global demand for H2 , the development of catalysts for the production of published maps and institutional affil- hydrogen is highly sought after. Specifically, new original molecular complexes, most often iations. bioinspired, based on non-noble metals such as nickel, cobalt, or molybdenum have been synthesized and studied to replace platinum, a catalyst that is certainly effective but scarce and expensive [1–4]. Molecular catalysts have been shown to be proficient in the hydrogen Copyright: © 2022 by the authors. evolution reaction (HER), showing increased efficiency and stability. In particular, a variety Licensee MDPI, Basel, Switzerland. of molecular cobalt catalysts have been developed for the production of hydrogen such as This article is an open access article glyoxime and the vitamin B12, which have the disadvantage of being poorly soluble and distributed under the terms and unstable in water [5]. These include mononuclear and dinuclear cobalt polypyridines [6,7]; conditions of the Creative Commons macrocyclic ligands such cobalt corroles [8]; cationic cobalt porphyrins [9]; anionic cobalt Attribution (CC BY) license (https:// porphyrins [10]; cobalt heme-peptides [11]; and more recently, a synthetic mini-enzyme creativecommons.org/licenses/by/ known as mimochrome VI*a [12]. These cobalt complexes underwent a light-induced 4.0/). proton reduction in water when combined with a photosensitizer and a sacrificial electron Int. J. Mol. Sci. 2022, 23, 14640. https://doi.org/10.3390/ijms232314640 https://www.mdpi.com/journal/ijms Int. J. Mol. Sci. 2022, 23, 14640 2 of 15 donor. However, many of these complexes are sparingly soluble in water and their chemical stability is not optimal. To improve the catalytic performance of these systems, their outer coordination sphere can be designed to promote the efficient transfer of electrons, H+ or H2 . To overcome this limitation, efforts have been made to develop artificial metalloenzymes with an unnatural active site integrated into a protein. This strategy has been shown to improve the production of H2 [13] and different artificial Co-artificial metalloenzymes have been obtained by non-covalent insertion of cobaloxime derivatives into apo-hemoproteins such as sperm whale myoglobin or heme oxygenase [14,15] or in a protein carrying out electron transfers such as ferredoxin [16]. In order to improve the catalytic efficiency of those artificial metalloenzymes, another cofactor such as the cobalt protoporphyrin IX was incorporated into a sperm whale myoglobin scaffold by the group of Ghirlanda [17]. In a recent publication, the same group successfully reported the light-driven reduction of CO2 and the generation of H2 using an artificial hemoprotein made from the Cytochrome b562 apoprotein that included Co-Protoporphyrin IX as a cofactor [18]. Inspired by these results, we decided to assemble new artificial metalloenzymes by a covalent-grafting strategy, involving the synthesis of various cobalt cofactors bearing a moiety that would be able to react with the SH group of a cysteinyl residue that could permanently graft them to the protein. Recently, our group has shown that a new class of thermostable artificial proteins named αReps have the potential to be transformed into artificial metalloenzymes by covalent grafting of a synthetic metal complex [19]. These αRep proteins are easily produced in E. coli, are very stable within a large pH range, and exhibit a large hydrophobic core. We have thus previously reported a proof of concept by creating artificial Diels– Alderases by covalent attachment of phenanthroline– or terpyridine–Cu(II) complexes in a specific αRep single-chain asymmetric bidomain variant named (A3A30 )Y26C [20]. This protein has been shown to adopt a closed bivalve shell-like conformation and molecular modeling identified that this cavity is large enough for the anchoring of even larger metal complexes such as meso-tetraphenylporphyrin [21]. Thus, from an enzyme designer’s point of view, the artificial αRep proteins bring together the properties required for the development of artificial metalloenzymes dedicated to H2 production. Accordingly, we here report the production and characterization of a new artificial metalloenzyme obtained by the covalent attachment of a cobalt porphyrin derivative into the αRep (A3A30 )Y26C variant and further shows its remarkable activity to catalyze the photoinduced production of H2 and CO2 reduction. 2. Results and Discussion 2.1. Synthesis of Co(III)Mal-PPIXMME 4 The Co(III)Mal-PPIXMME cofactor 4 was prepared in three steps. In the first step was the hydrolysis of protoporphyrin IX dimethyl ester (PPIXDME) by 6 N HCl for 15 min at room temperature (RT) following the reaction conditions described by Ellsworth R.K. [22] (see Materials and Methods). Protoporphyrin IX monomethyl ester (PPIXMME) 2 was then obtained in a 28% yield, and was characterized by NMR and UV–Visible spectroscopies as well as by ESI MS (see Materials and Methods). The two following steps of the synthesis are shown in Figure 1. The reaction of 2 with a slight excess of N-(2-aminoethyl)-maleimide trifluoroacetate salt (1.1 equiv.) in the presence of triethylamine and PyBOP as a coupling agent in dimethylformamide (DMF) for 30 min at RT afforded the maleimido-ethylamido-protoporphyrin IX monomethyl ester Mal-PPIXMME 3 as a brown solid in a 45% yield. 3 was characterized by NMR and UV–Visible spectroscopies and ESI MS (see Materials and Methods). These characteristics were in good agreement with those of a very similar product, maleimido-ethylamido- protoporphyrin IX monoethyl ester Mal-PPIXMEE, which has already been reported by the group of Y. Watanabe [23]. protoporphyrin IX monoethyl ester Mal-PPIXMEE, which has already been reported by the group of Y. Watanabe [23]. Finally, the addition of 10 equiv. of Co(OAC)2, 4H2O to the obtained compound 3 in anhydrous DMF at 70 °C led, after the removal of the excess of cobalt salt by extraction in a dichloromethane (DCM)/water mixture and the removal of the solvent under reduced Int. J. Mol. Sci. 2022, 23, 14640 3 of 15 pressure, to Co(III)-Mal-PPIXMME as a dark green solid in a 97% yield. 4 was characterized by UV–Visible spectroscopy and ESI MS spectrometry. HOOC NH N N NH MeOOC Protoporphyrin IX dimethyl ester (PPIXMME) 2 O NH3 O PyBOP, Et3N CF3 N 57 % O O O H N N NH N O O N NH MeOOC 3 Co(OAc)2, MeOH reflux, ON or DMF 70°C 97 % O H N N N N O Co N O N MeOOC 4 Figure 1. Figure 1. Synthesis Synthesis of of the the Co(III)Mal-PPIXMME Co(III)Mal-PPIXMME cofactor cofactor 44 from from protoporphyrin protoporphyrin IX IX monomethyl monomethyl ester (PPIX MME) ester (PPIX MME) 1. 1. 2.2. Biohybrid Preparation Finally, the addition of 10 equiv. of Co(OAC)2 , 4H2 O to the obtained compound 3 in anhydrous DMF at 70 ◦ C led, after the removal(A3A3′)Y26C-4) The (A3A3′)Y26C-Co(III)Mal-PPIXMME of the excess of biohybrid cobalt saltwas by extraction prepared in in athree dichloromethane (DCM)/water steps. The single-chain mixture bidomain and the protein removal in which of the solvent a cysteine-free under is domain reduced linked pressure, to Co(III)-Mal-PPIXMME with a domain as a dark((A3A3′)Y26C) bearing the Y26C mutation green solid in was a 97% yield. first 4 was characterized produced and purified, by UV–Visible spectroscopy and ESI MS spectrometry. as recently described [24]. Compound 4 was then synthesized from protoporphyrin IX in three steps, as described above. The biohybrid was finally assembled by incubation of the 2.2. Biohybrid Preparation (A3A3′)Y26C protein with 10 eq. of compound 4 for 3 days at 4 °C in a 50 mM Na- The (A3A3 0 )Y26C-Co(III)Mal-PPIXMME (A3A30 )Y26C-4) biohybrid was prepared in phosphate pH 7.75 coupling buffer containing 150 mM NaCl and 25% dimethyl sulfoxide three steps. The single-chain bidomain protein in which a cysteine-free domain is linked with a domain bearing the Y26C mutation ((A3A30 )Y26C) was first produced and purified, as recently described [24]. Compound 4 was then synthesized from protoporphyrin IX in three steps, as described above. The biohybrid was finally assembled by incubation of the (A3A30 )Y26C protein with 10 eq. of compound 4 for 3 days at 4 ◦ C in a 50 mM Na- phosphate pH 7.75 coupling buffer containing 150 mM NaCl and 25% dimethyl sulfoxide (DMSO). After removal of the excess of compound 4 by extraction with butan-2-one, the (A3A30 )Y26-4 biohybrid was recovered in the aqueous phase. Int.J. J.Mol. Int. Mol.Sci. Sci.2022, 2022,23, 23,x xFOR FORPEER PEERREVIEW REVIEW 4 4ofof1515 Int. J. Mol. Sci. 2022, 23, 14640 4 of 15 (DMSO).After (DMSO). Afterremoval removalofofthe theexcess excessofofcompound compound44by byextraction extractionwith withbutan-2-one, butan-2-one,thethe (A3A3′)Y26-4 biohybrid was recovered in the (A3A3′)Y26-4 biohybrid was recovered in the aqueous phase. aqueous phase. The (A3A3 The 0 )Y26C-4 biohybrid (A3A3′)Y26C-4 biohybridwaswaswas thenthenthen characterized characterized characterized by MALDI-ToF by MALDI-ToF MALDI-ToF mass mass spectrom- The (A3A3′)Y26C-4 biohybrid by mass spectrometry analysis (Figure 2)where0 where (A3A3′)Y26C showed etry analysis analysis spectrometry (Figure 2)(Figure where 2) (A3A3 )Y26C showed ashowed (A3A3′)Y26C major aamajor peak major peak at Mpeak atatMM = 44,502 ==44,502 Da, 44,502 and a Da,and and apeak peakDa 0 )Y26C-4 peak Da, at a45,262 atat45,262 45,262 was Dawas wasobserved observed Da observed for the (A3A3 forthe for the(A3A3′)Y26C-4 (A3A3′)Y26C-4 biohybrid, biohybrid,biohybrid, which which waswhich in wasinin agreement was agreement withthe the covalentattachment attachment 0 )Y26C. with the covalent agreement with attachment covalent of compound ofofcompound 4compound (M = 755.744(M (M=on Da) =755.7 755.7 (A3A3Da) Da) on(A3A3′)Y26C. on (A3A3′)Y26C. Figure2.2.2.Mass Figure Figure Massspectrometry Mass spectrometryanalysis spectrometry analysisofof analysis ofthe the(A3A3′)Y26C-4 the (A3A3′)Y26C-4 (A3A3 biohybrid:Maldi-ToF 0 )Y26C-4biohybrid: biohybrid: Maldi-ToFspectrum Maldi-ToF spectrumof spectrum of of (A3A3′)Y26C 0 (A3A3′)Y26C (blue) (blue) and and (A3A3′)Y26C-4 0 (A3A3′)Y26C-4 (red). (red). (A3A3 )Y26C (blue) and (A3A3 )Y26C-4 (red). Toanalyze To To analyzethe analyze theeffect the effectof effect ofthe of thebinding the bindingof binding of444into of intothe into thecavity the cavityof cavity of (A3A3 0 )Y26Con of(A3A3′)Y26C (A3A3′)Y26C onthe on thefolding the folding folding of of the the protein, protein, circular circular dichroism dichroism (CD) (CD) measurements measurements were of the protein, circular dichroism (CD) measurements were carried out. The far UV were carried carried out. out. The far The UV far UV circular circular dichroism dichroism spectrum spectrum of of the (A3A3 the 0 (A3A3′)Y26C-4 )Y26C-4 biohybrid biohybrid (Figure (Figure 3) 3) showed showed a signala signal with circular dichroism spectrum of the (A3A3′)Y26C-4 biohybrid (Figure 3) showed a signal withtwo two with twonegative negative negative bands bands atat222 at 222 bands 222 and andand 208 208208nm nm nmand and and apositive positive a apositive bandaround band band around190 around 190nm 190 nmthat thatwas that was was characteristicof characteristic characteristic ofproteins of proteinscontaining proteins containingmainly containing mainlyαααhelices mainly helicesas helices asaaasecondary as secondarystructure. secondary structure.This structure. Thissignal This signal signal wasvery was verysimilar very similarto similar tothat to thatof that ofthe of thenon-modified the non-modified(A3A3′)Y26C non-modified (A3A3 0 )Y26C protein (A3A3′)Y26C protein[19,21], [19,21],indicating [19,21], indicatingthat indicating that that was protein its its 3D structure was was not not altered altered after after covalent covalent attachment attachment ofofthe porphyrin the its 3D structure was not altered after covalent attachment of the porphyrin cofactor in its porphyrin cofactor cofactor in its in hydrophobic its hydrophobic pocket, pocket, even even in the in thepresence presence ofof25% 25% DMSO, DMSO, hydrophobic pocket, even in the presence of 25% DMSO, and further treatment with and and further treatment with with butan-2-one. butan-2-one. butan-2-one. Figure3.3. Figure 3.Far-UV Far-UVCD spectrumofofaaa2.2 CDspectrum spectrum 2.2mM 2.2 mMsolution mM solutionof solution ofthe of the (A3A3 0 )Y26C-4 biohybrid the(A3A3′)Y26C-4 (A3A3′)Y26C-4 biohybridinin in20 20mM 20 mM mM Figure biohybrid sodium phosphate sodiumphosphate sodium buffer phosphatebuffer pH bufferpH 7.5. pH7.5. 7.5. 2.3. UV–Visible Characteristics of the (A3A30 )Y26C-4 Biohybrid The addition, under anaerobic conditions, of 1.5 mM sodium dithionite to a 7.5 µM solution of (A3A30 )Y26C-4 in 10 mM HEPES, NaCl 150 mM buffer, pH 8.0 led to the Int. J. Mol. Sci. 2022, 23, x FOR PEER REVIEW 5 of 15 Int. J. Mol. Sci. 2022, 23, 14640 2.3. UV–Visible Characteristics of the (A3A3′)Y26C-4 Biohybrid 5 of 15 The addition, under anaerobic conditions, of 1.5 mM sodium dithionite to a 7.5 µM solution of (A3A3′)Y26C-4 in 10 mM HEPES, NaCl 150 mM buffer, pH 8.0 led to the gradual gradual replacement, replacement, with with an an isosbestic isosbestic point point at at 408 408 nm, nm, of of the the spectrum spectrum characteristic characteristic of of Co(III)-Mal-PPIXMME Co(III)-Mal-PPIXMME 44 with peaks at 423, 535, and 569 nm by a new new one one with with peaks peaks atat 399 and 562 nm, which was characteristic of the reduced form of the biohybrid Co(II)–Mal- 399 and 562 nm, which was characteristic of the reduced form of the biohybrid Co(II)– PPIXMME Mal-PPIXMME4 (Figure 4). 4). 4 (Figure Figure 4. Evolution Figure Evolutionofofthe theUV–Vis UV–Visspectrum (A3A30 )Y26C-4 spectrumofof(A3A3′)Y26C-4 7.57.5 µMµMin 10 in mM HEPES, 10 mM HEPES,NaCl 150 NaCl mM buffer, pH 8.0 after the addition of a 1.5 mM anaerobic solution of sodium dithionite. Spectra 150 mM buffer, pH 8.0 after the addition of a 1.5 mM anaerobic solution of sodium dithionite. Spectra were recorded were recordedevery every55min minduring during22h. h. The characteristics The characteristics of of these these spectra are compared compared in Table Table 11 with with those those obtained obtained after after the incorporation of Co(III)– the incorporation of Co(III)– and Co(II)–protoporphyrin IX into the apoprotein of Co(II)–protoporphyrin IX into the apoprotein of various various hemoproteins including hemoproteins includingmyoglobin myoglobin(Mb)(Mb) [24] [24] and and itsits H64V-V68A H64V-V68A double double mutant [25]; two mutant [25]; two cytochromesP450, cytochromes P450,P450 P450Cam Cam [26] [26] and and CYP119 CYP119 [27]; [27]; two two peroxidases, peroxidases, horseradish peroxidase peroxidase (HRP) (HRP)[24] [24]and anddye dyedecolorizing decolorizing peroxidase peroxidase(DYP) (DYP) [28], aldoxime [28], dehydrogenase aldoxime (Oxd) dehydrogenase [28] (Oxd) and [28] catalase (Cat)(Cat) and catalase [28]. [28]. 0 )Y26C-4 with those of various Co(II)– Table1.1. Comparison Table Comparison of of the the maxima maxima ofof absorption absorption of of (A3A3 (A3A3′)Y26C-4 and Co(III)–PPIX-substituted and Co(III)–PPIX-substituted hemoproteins hemoproteins produced produced either either in in invivo vivo or or in vitro vitro through through reconstitution reconstitution with analogous proteins with analogous proteins with CoPPIX [28]. with CoPPIX [28]. Co(III) λMax (nm) Co(II) λMax (nm) Co(III) λMax (nm) Co(II) λMax (nm) Protein Ligands Soret β α Ligand Soret α/β Ref. Protein Ligands Soret β α Ligand Soret α/β Ref. Mb His/H2O 424 535 572 His 406 558 [24] Mb His/H2 O 424 535 572 406 558 [24] MbH64VV68A His/H2O 428 540His 572 His 406 555 [25] MbH64VV68A His/H2 O 428 540 572 His 406 555 [25] P450Cam Cys/H2 O P450 422 Cam Cys/H 538 2O 422 570 538Cys 570 Cys 404 404 556 556 [26] [26] CYP119 Cys/H2 O CYP119 422 Cys/H 539 2O 422 566 539Cys 566 Cys 410 410 558 558 [27] [27] Oxd His/H2 O 429 Oxd 542 His/H2O 574 429 542His 574 402 His 559 559 402 [28] [28] HRP His/H2 O 421 HRP 533 His/H2O 565 421 533 His 565 401 His 553 553 401 [24] [24] DYP His/H2 O 422 534 567 His 403 564 [28] Cat His/? DYP 427 His/H2O 537 422 570 534His 567 His 403 403 561 564 [28] [28] Cat His/? 427 537 570 His 403 561 [28] This (A3A30 )Y26C-4 ? 423 535 569 ? 399 562 This work (A3A3′)Y26C-4 ? 423 535 569 ? 399 562 work As shown in Table 1, the electronic absorption spectrum of the oxidized Co(III)– (A3A30 )Y26C-4 biohybrid was similar to those obtained through the incorporation of Int. J. Mol. Sci. 2022, 23, x FOR PEER REVIEW 6 of 15 As shown in Table 1, the electronic absorption spectrum of the oxidized Co(III)– Int. J. Mol. Sci. 2022, 23, 14640 (A3A3′)Y26C-4 biohybrid was similar to those obtained through the incorporation 6 of 15 of Co(III)PPIX not only into apoproteins, which bind the cobalt cation through an axial thiolate ligand such as P450 cam [26] and CYP 119 [27], but also to those of some apoproteinsnot Co(III)PPIX thatonly bind theapoproteins, into cobalt cation through which bindantheaxial histidine cobalt cation ligand throughsuch as Mb, an axial HRP thiolate ligand [24], and such DYP as[28]. P450 cama [26] Such and CYP comparison 119not does [27], butfor allow also thetoidentification those of some theapoproteins cobalt axial that ligandbind the cobalt of cobalt cation throughHowever, in (A3A3′)Y26C-4. an axial histidine ligand such when comparing theas Mb, HRPabsorption electronic [24], and DYP spectra[28].of Such a comparison the various reduceddoes not allow complexes, for the that it appears identification the spectral thecharacteristics cobalt axial ligand of the of cobalt in (A3A3 0 )Y26C-4. However, when comparing the electronic absorption spectra Co(II)–(A3A3′)Y26C-4 biohybrid were very different from those observed with P450 cam of the [26] andvarious CYP 119reduced complexes, [27] whereas theyitwere appears that closer thethose than spectral characteristics observed in the caseof of theproteins Co(II)– (A3A3 0 bearing)Y26C-4 an axialbiohybrid were very histidine ligand different with, from those in particular, a veryobserved with P450tocam strong similarity [26]ofand those the CYP Co(II)119 [27] dye whereas they decolorizing were closer peroxidase than [28]. (DYP) thoseThis observed latter in the case of tends comparison proteins to bearing suggest an axial axial histidine ligand coordination of thewith, in particular, cobalt a very strong ion by a histidine ligand similarity to those of the Co(II) in the Co(II)–(A3A3′)Y26C-4 dye decolorizing peroxidase (DYP) [28]. This latter comparison biohybrid. However, in order to clarify the environment of the cobalt in the tends to suggest axial distal pocket coordination of the cobalt ion by a histidine ligand in the Co(II)–(A3A3 0 )Y26C-4 biohybrid. of this artificial metalloenzyme, and in particular, the nature of its axial ligand, However, computational in order to clarify modeling thewere studies environment engaged. of the cobalt in the distal pocket of this artificial metalloenzyme, and in particular, the nature of its axial ligand, computational modeling studies were 2.4. Computational engaged. Modeling Studies In the absenceModeling 2.4. Computational of the X-ray structure of the (A3A3′)Y26C-4 biohybrid, we embarked Studies on a molecular modeling study. It primarily aimed at elucidating the most stable In the absence of the X-ray structure of the (A3A30 )Y26C-4 biohybrid, we embarked on orientation of the non-natural porphyrin into (A3A3′)Y26C and assessing the relevant a molecular modeling study. It primarily aimed at elucidating the most stable orientation interactions between both subsystems including possible coordination bonds between of the non-natural porphyrin into (A3A30 )Y26C and assessing the relevant interactions nearby lying amino acids and the metal. between both subsystems including possible coordination bonds between nearby lying First, protein–ligand dockings were carried out. For the protein structure, we used amino acids and the metal. the representative geometry of the most populated cluster of conformations of a 300 ns First, protein–ligand dockings were carried out. For the protein structure, we used the molecular dynamics (MD) performed with AMBER [29] of the wild type A3A3′ for which representative geometry of the most populated cluster of conformations of a 300 ns molec- Tyr26 ular was mutated dynamics to cysteine (MD) performed withusing AMBER the [29] UCSF Chimera of the wild typeplug-in A3A30 for aminoTyr26 for which acid substitution [30]. The artificial cofactor was accordingly docked using was mutated to cysteine using the UCSF Chimera plug-in for amino acid substitution [30]. the covalent protocol The available artificial cofactorinwas Gold and the GoldScore accordingly docked usingscoring functionprotocol the covalent that weavailable updatedin for the Gold metalloligand recently [31]. and the GoldScore scoring function that we updated for the metalloligand recently [31]. A first A firstdocking docking runrun waswas performed performed with with a rigida protein rigid protein scaffold.scaffold. From theFrom the 50 50 solutions solutions obtained, the lowest energy solution reached 55.1 GoldScore obtained, the lowest energy solution reached 55.1 GoldScore units and can be considered units and can be to considered to be good interaction between both molecular partners. be good interaction between both molecular partners. The porphyrin stands very well The porphyrin stands into veryinter-dominial the well into the inter-dominial pocketgood pocket and displays and displays goodwith interactions interactions withof amino acids amino acids the cavity of the cavity such as Phe318 with the macrocycle, the methyl group such as Phe318 with the macrocycle, the methyl group of the methyl propionate ester side of the methyl propionate chain interactsester withside chain interacts a hydrophobic patchwith a hydrophobic constituted by Trp81,patch Trp310,constituted and Trp342, byand Trp81, the Trp310, cyclic and Trp342, moiety and the of the linker cyclic moiety interacting polarityofwith the linker Arg22 interacting polarity with (Figure 5). Close-lying Arg22 residues (Figure that could5). coordinate Close-lyingare residues that and Gln 61,283 could coordinate 291 or Asp322. are Gln 61,283 and 291 or Asp322. 0 )Y26C-4 biohybrid obtained Figure 5. Representation Figure 5. Representation of of the the lowest lowest energy energy geometry geometry of of the the(A3A3 (A3A3′)Y26C-4 biohybrid obtained with with Gold Gold and and the the Goldscore Goldscore scoring scoring function. function. Based on this structure, further dockings were carried out to see whether nearby lying residues could eventually interact with the metal. This was performed by giving rotational flexibility to the side chain of the amino acid Glu322 and Phe318 as well as Gln283. Based on this structure, further dockings were carried out to see whether nearby lying residues could eventually interact with the metal. This was performed by giving rotational Int. J. Mol. Sci. 2022, 23, 14640 flexibility to the side chain of the amino acid Glu322 and Phe318 as well as 7 of 15 Gln283. From those calculations, it appears that Glu322 and Gln291 may both reach convenient distances for coordination, though the former needs Phe318 to be out of the way to the iron. Additionally, these docking solutions show different degrees of rotation From those calculations, it appears that Glu322 and Gln291 may both reach convenient of the porphyrin around the linker. This highlights a certain amount of flexibility of the distances for coordination, though the former needs Phe318 to be out of the way to the iron. porphyrin that could be related to transient interactions between the metal and the Additionally, these docking solutions show different degrees of rotation of the porphyrin protein. To further analyze this hypothesis, classical molecular dynamics (MD) was around the linker. This highlights a certain amount of flexibility of the porphyrin that could performed. be related to transient interactions between the metal and the protein. To further analyze Classical MDs were carried out with the AMBER suite of programs using the this hypothesis, classical molecular dynamics (MD) was performed. AMBER20 force field [29]. The porphyrin linker was set up by defining a non-standard Classical MDs were carried out with the AMBER suite of programs using the AM- amino acid and by using of the MCPB.py algorithm for the parameterization of the BER20 force field [29]. The porphyrin linker was set up by defining a non-standard amino metallic center and its first coordination sphere (see ESI for further details on the acid and by using of the MCPB.py algorithm for the parameterization of the metallic computational section) [32,33]. Calculations started from the best solution for the initial center and its first coordination sphere (see ESI for further details on the computational series of dockings and ran for 400 section) [32,33]. ns. The trajectory Calculations started showed from thethe beststability solutionand forconvergence the initial series of dockings of the system after this time (Figure S1). The calculations agreed with the and ran for 400 ns. The trajectory showed the stability and convergence flexible dockingof the system after and showed a flexible cofactor in the binding site (Figure 6A). However, some orientations this time (Figure S1). The calculations agreed with the flexible docking and showed a flexi- clearly dominated. The most ble cofactor populated in the one(Figure binding site was similar to the best-docked 6A). However, some orientationsstructure clearly dominated. described at theThe beginning of this section (Figure 6B). One can observe most populated one was similar to the best-docked structure describeda very stable π-at the beginning stacking between Phe318 of this and (Figure section the porphyrin. 6B). OneInterestingly, can observe athe Phe318 very stableside chain placed π-stacking between Phe318 and most of the simulation the porphyrin. Interestingly, the Phe318 side chain placed most of the by just below the metal with a position that would be occupied an simulation just below axial ligand of the metal in natural hemoenzymes. This shows that only one face of the metal with a position that would be occupied by an axial ligand of the metal in natural the porphyrin is accessible hemoenzymes. for catalysis This showsandthat hasonlya one well-defined face of the and asymmetric porphyrin distalfor catalysis and is accessible environment. Alternative orientations has a well-defined corresponddistal and asymmetric to geometries environment. in which the interaction Alternative orientations correspond Phe318-4 is lost.toWhen such anininteraction geometries which the is lost, compound interaction Phe318-4 4 behaves very freely is lost. When such in anthe interaction is lost, interface between compound 4 behaves very freely in the interface between the A3 and A30 subdomains. the A3 and A3’ subdomains. Importantly, one of the most populated geometries of these alternative Importantly, oneorientations showed short of the most populated distances geometries between of these Gln283orientations alternative and showed the cobalt (Figureshort6C), some consistent distances between with Gln283 direct and Co-Gln283 coordination, the cobalt (Figure 6C), some and consistent others with direct with a bridged water Co-Gln283molecule between the coordination, andamino othersacid and with the metal. a bridged Suchmolecule water interaction is between the amino only possible because acid andthe absenceSuch the metal. of the interaction interaction Phe318-4 is only possibleisbecause associated with an the absence of the interaction increase in the accessibility of the metal. Phe318-4 is associated with an increase in the accessibility of the metal. Figure 6. (A) Overlap structures of the main clusters extracted from the 400 ns MD trajectory for the (A3A30 )Y26C-4 systems. (B) MD snapshot of the most populated cluster of the trajectory. (C) Snapshot with the closest distance between the metal and the nitrogen side chain of Gln283. In conclusion, the molecular modeling study showed that: (1) The linkage of com- pound 4 to (A3A30 )Y26C led to a macrocycle located at the interface between both sub- Int. J. Mol. Sci. 2022, 23, 14640 8 of 15 domains and excluded from the solvent; (2) the main orientation of 4 in A3A30 presented a strong contact between Phe318 and the macrocycle that aided in the packing of the porphyrin hydrophobically to one domain and led to an asymmetric environment on the distal side; and (3) that possible coordination of the metal would only appear on transient structures with Gln283. 2.5. Photocatalytic Activity of the (A3A30 )Y26C-4 Biohybrid The ability of the (A3A30 )Y26C-4 biohybrid to catalyze not only the production of H2 , but also to reduce CO2 under photoinduced conditions was explored. [Ru(bpy)3 ]2+ (bpy = 2,20 -bipyridine) was chosen as the photosensitizer and ascorbic acid as the sacrificial electron donor agent, since it has previously been shown that under irradiation at 405 nm, the excited state of [Ru(bpy)3 ]2+ is reduced by ascorbic acid to [Ru(bpy)3 ]+ , which in turn transfers electrons to the catalyst, generating its active form, which is capable of reducing substrates such as protons [34] or CO2 [35]. 2.5.1. Light-Driven H+ Reduction To assay the production of H2 upon the photoinduced reduction of protons, experi- ments were carried out in argon degassed 100 mM ascorbic acid in the presence of 1 mM [Ru(bpy)3 ]2+ as a photosensitizer and either 7 µM (A3A30 )Y26C-4 or 7 µM Co(III)PPIX as a catalyst in 100 mM potassium phosphate buffer at pH 7.0 under irradiation with a white LED light at λ > 405 nm. Figure 7 shows the evolution of the turnover number (TON) related to H2 production over time for 25 h. It appears that in the presence of the A3A30 Y26C-4 biohybrid, the production of H2 increases over time to reach a value of 163 TON after 25 h. In contrast, when the metalloporphyrin CoPPIX was used as a catalyst, only a low production of H2 not exceeding 13 TON was observed after 25 h whereas a negligible H2 production was observed in the presence of the non-modified (A3A30 )Y26C protein or in the absence of a catalyst. These results clearly show that the production of H2 catalyzed by the cobalt(III)–porphyrin was greatly enhanced by a factor of at least 10, when the latter was buried in the hydrophobic cavity of the (A3A30 )Y26C protein. In addition, it is noteworthy that the TON value obtained with the biohybrid (A3A30 )Y26C-4 as a catalyst was similar to that described by Ghirlanda’s group using Co-cytochrome b 562 as the catalyst [30]. It is also important to note that even after 25 h of irradiation time, the A3A30 Y26C-4 did not show any signs of degradation and continued to show a steady production of H2 . 2.5.2. Light-Driven CO2 Reduction The photoinduced reduction of CO2 catalyzed by (A3A30 )Y26C-4 (7 µM) was also stud- ied at pH 7.0 and 20 ◦ C in a CO2 saturated potassium phosphate buffer containing 100 mM ascorbic acid as a sacrificial electron donor and 1 mM [Ru(bpy)3 ]2+ as a photosensitizer. Figure 8 shows that under irradiation with a white LED light at λ > 405 nm, a significant amount of CO could be produced, reaching a plateau with a TON of 80 in about 12 h. The appearance of this plateau is probably due to the deactivation of the cobalt porphyrin through a carbonyl transfer to one of the meso positions of the porphyrin rim as reported by Jiang et al. [36]. The time lag of 3 h observed before CO production as detected by gas chromatography (GC) can probably be rationalized by an initial accumulation of the CO inside the hydrophobic pocket of the protein before being released when its concentration increases. This CO characteristic was reported to be essential for its transport from CO dehydrogenase where it is produced by the acetyl CoA synthase via direct hydrophobic gas channels before being used in a carbonylation reaction [37]. When (A3A30 )Y26C-4 is replaced by the CoPPIX metalloporphyrin, a plateau at TON = 20 is reached after 7 h. In the control experiments, no CO production was observed in the presence of the non- modified (A3A30 )Y26C protein as a catalyst or in the absence of a catalyst. Here also, our results clearly showed that the production of CO catalyzed by the cobalt porphyrin was greatly enhanced by a factor of 4 when the latter was inserted in the hydrophobic cavity Int. J. Mol. Sci. 2022, 23, 14640 9 of 15 of the (A3A30 )Y26C protein. Interestingly, the TON value obtained with the biohybrid Mol. Sci. 2022, 23, x FOR PEER REVIEW (A3A30 )Y26C-4 as a catalyst was twice higher than that described by9 Ghirlanda’s of 15 group using Co-cytochrome b 562 as the catalyst (42 TON) under similar reaction conditions [18]. Figure 7. H2 produced Figure 7.over time fromover H2 produced the time photoinduced reduction of reduction from the photoinduced protons byof(A3A3′)Y26C-4 protons by (A3A30 )Y26C-4 and and CoPPIX at pH 7.0 under argon. The experiments were carried out in 100 mM ascorbic acid, 100 CoPPIX at pH 7.0 under argon. The experiments were carried out in 100 mM ascorbic acid, 100 mM mM potassium phosphate buffer in the presence of 1 mM [Ru(bpy)3]2+ as a photosensitizer, and 7 potassium phosphate buffer in the presence of 1 mM [Ru(bpy)3 ]2+ as a photosensitizer, and 7 µM Mol. Sci. 2022, 23, x FOR PEER µM REVIEW(A3A3′)Y26C-4 (•0), Co(III)PPIX (•), (A3A3′)Y26C (•)0 as the catalyst or without catalyst 10 of(• 15) Int. J. Mol. Sci. 2022,under 23, x FOR PEER (A3A3 )Y26C-4 (•), Co(III)PPIX (•), (A3A3 )Y26C (•) as the catalyst or without REVIEW catalyst (•10 ) under of 15 white LED light, λ > 405 nm irradiation. All experiments were carried out at least twice. white LED light, λ > 405 nm irradiation. All experiments were carried out at least twice. 2.5.2. Light-Driven CO2 Reduction The photoinduced reduction of CO2 catalyzed by (A3A3’)Y26C-4 (7 µM) was also studied at pH 7.0 and 20 °C in a CO2 saturated potassium phosphate buffer containing 100 mM ascorbic acid as a sacrificial electron donor and 1 mM [Ru(bpy)3]2+ as a photosensitizer. Figure 8 shows that under irradiation with a white LED light at λ >405 nm, a significant amount of CO could be produced, reaching a plateau with a TON of 80 in about 12 h. The appearance of this plateau is probably due to the deactivation of the cobalt porphyrin through a carbonyl transfer to one of the meso positions of the porphyrin rim as reported by Jiang et al. [36]. The time lag of 3 h observed before CO production as detected by gas chromatography (GC) can probably be rationalized by an initial accumulation of the CO inside the hydrophobic pocket of the protein before being released when its concentration increases. This CO characteristic was reported to be essential for its transport from CO dehydrogenase where it is produced by the acetyl CoA synthase via direct hydrophobic gas channels before being used in a carbonylation reaction [37]. When (A3A3’)Y26C-4 is replaced by the CoPPIX metalloporphyrin, a plateau at TON = 20 is reached after 7 h. In the control experiments, no CO production was observed in the presence of the non-modified (A3A3′)Y26C protein as a catalyst or in the absence Figure 8. of a catalyst. Production Figureof8. Here 8.CO overalso, Production our timeof COresults from clearly the photoinduced over time from showed that of reduction the COproduction of CO 2 by (A3A3′)Y26C-4 (A3A30 )Y26C-4 Figure Production of CO over time from the the photoinduced photoinduced reduction reduction of of CO CO2 byby (A3A3′)Y26C-4 and CoPPIX catalyzed byatthe pHcobalt 7.0 in aporphyrin CO2 saturated was solution. greatly The experiments enhanced by a were carried factor of 4 out in when 100 the mM latter and CoPPIX and CoPPIX at at pH pH 7.0 in a CO22 saturated saturated solution. solution. The The experiments experimentswere were carried carried out in out in 100 100 mM mM ascorbic was acid, 100 inserted mMhydrophobic inascorbic the potassium phosphate cavity ofsaturated with CO2 inprotein. thephosphate (A3A3′)Y26C the presence of 1 mM [Ru(bpy) Interestingly, the TON 3]2+ ascorbic acid, acid, 100 100 mM mM potassium saturated with saturated with CO CO22 in the presence [Ru(bpy)33]2+ of 1 mM [Ru(bpy) 2+ as a photosensitizer, value obtainedas with and the7 biohybrid µM (A3A3′)Y26C-4 (•), Co(III)PPIX (A3A3’)Y26C-4 as a (•), (A3A3′)Y26 catalyst was twiceC(higher ) as a0catalyst than or that as aa photosensitizer, (A3A3 )Y26C-4 ((• photosensitizer, and 7 µM (A3A3′)Y26C-4 0 •), Co(III)PPIX (••),),(A3A3 (A3A3′)Y26 )Y26 CC(( ) as) as a catalyst a catalyst or without catalyst described by or (•) under white Ghirlanda’s group LEDusing light, Co-cytochrome λ >405 nm irradiation. b>405 562The experiments asirradiation. the catalyst were (42 carried TON)were out in duplicate. without without catalyst catalyst • (•)(under ) underwhite LED white light, LED λ > λ405 light, nmnmirradiation. TheThe experiments experiments carried were out carried under similar out reaction conditions [18]. in duplicate. in duplicate. 3. Materials and Methods 3. Materials and Methods 3.1. Synthesis of Co(III)Mal-PPIX 4 3.1. Synthesis of Co(III)Mal-PPIX 4 The Co(III)Mal-PPIX cofactor 4 was prepared in three steps. The Co(III)Mal-PPIX cofactor 4 was prepared in three steps. Int. J. Mol. Sci. 2022, 23, 14640 10 of 15 3. Materials and Methods 3.1. Synthesis of Co(III)Mal-PPIX 4 The Co(III)Mal-PPIX cofactor 4 was prepared in three steps. 3.1.1. Protoporphyrin IX Dimethyl Ester (PPIX DME) Hydrolysis [23] A total of 66.8 mg of protoporphyrin IX dimethyl ester PPIX DME 1 was dissolved in 6 N HCl and the solution was allowed to stand for 15 min at room temperature. This solution was then added to 200 mL chloroform overlaid with 200 mL of 10% (w/v) aqueous sodium acetate. The resulting mixture was then completely neutralized with solid NaHCO3 to transfer all of the porphyrin into the chloroform organic layer. This layer was then evaporated to dryness and the obtained solid was redissolved in 2 mL of 2,6 lutidine for purification via silica gel column chromatography with dichloromethane/methanol 99/1 as the eluent. PPIX monomethyl ester (PPIX MME) 2 was obtained as a purple solid (18.4 mg, 28% yield). 1 H NMR (CDCl , 400 MHz): δ 9.85–9.98 (m, 4H, meso-H), 8.00–8.16 (m, 2H, -CH=), 6.30, 3 6.26, 6.15, 6.11 (s each, 4H, =CH2 ), 4.23 (m, 4H, -CH2 -), 3.52–3.69 (m, 15H, 2, 7, 12, 18, -CH3 , -OCH3 ), 3.23 (m, 2H, CH2 -Cacide ), 3.14 (m, 2H, CH2 -Cester ), −4.14 (bs, 2H, pyrrole, NH). MS-ESI [M + H+ ]: found. 577.278, calc. 577.280. UV–Vis (acetonitrile) λmax (nm) = 401, 503, 537, 573, 628. 3.1.2. Synthesis of Maleimido-Ethylamido-Protoporphyrin IX Monomethyl Ester Mal-PPIX MME 3 PPIX MME 2 (28.5 mg, 49.42 µmol) was dissolved in 1 mL of anhydrous DMF and 15 µL of triethylamine was added. PyBOB (36 mg, 69.18 µmol) dissolved in 1 mL anhy- drous DMF was added to this solution and the reaction medium was left under stirring for 30 min at room temperature (RT). To this solution, a solution of N-(2-aminoethyl)malemide trifluoroacetate salt (13.8 mg, 54.36 µmol) in anhydrous DMF (1 mL) was added quickly. The resulting mixture was stirred at room temperature overnight and protected from light. The solvent was removed and the crude product was purified by silica gel chromatogra- phy (eluted with 1% MeOH in CHCl3) to give maleimido-ethylamido-protoporphyrin IX monomethyl ester Mal-PPIX MME 3 as a brown solid (15.7 mg, 45%). 1 H NMR (CDCl , 400 MHz): δ 9.89–10.12 (m, 4H, meso-H), 8.17–8.29 (m, 2H, -CH=), 3 6.36, 6.32, 6.18, 6.12 (s each, 4H, =CH2 ), 6.18 (br, 1H, -NH), 4.52, 4.50 (s, 2H, maleimide), 4.27–4.39 (4H, -CH2 -), 2.97–3.03 (M, 2H, CH2 -Cester ) 3.23–3.27 (m, 6H, CH2 Nmaleimide , CH2 NHamide , CH2 Camide ), 3.52–3.66 (m, 15H, 2, 7, 12, 18, -CH3 ; -OCH3 ), −4.14 (bs, 2H, pyrrole, NH). MS-ESI [M + H+ ] = found. 699.327, calc. 699.326. UV–Vis (DCM), λmax (nm) = 400, 503, 536, 570, 628, 662. 3.1.3. Insertion of Cobalt into Maleimido-Ethylamido-Protoporphyrin IX Monomethyl Ester Mal-PPIX MME In a 10 mL round bottom flask, 15.7 mg (22.5 µmol) of compound 3 was dissolved in 3 mL anhydrous DMF and a solution of Co(OAc)2. 4 H2 O (55.96 mg, 225 µmol) in 3 mL anhydrous DMF was added dropwise. The solution was heated at 70 ◦ C until the metalation was completed, as proven by UV–Visible spectroscopy. After the solvent was removed under reduced pressure, the residue was dissolved in 3 mL DCM. The organic layer was then washed three times with 100 mL of water, dried over Na2 SO4 , filtered, and the solvent was removed under reduced pressure to obtain Co(III)-Mal-PPIX MME 4 as a dark green solid (16.5 mg, 97%). MS-ESI [M + H+ ] = found: 755.235, calc. 755.239. UV–Vis (H2 O), λMax (nm) = 411, 527, 561. Int. J. Mol. Sci. 2022, 23, 14640 11 of 15 3.2. Production and Purification of αRep Proteins The production and the purification of the αRep proteins were carried out essentially as reported in a previous paper [24]. 3.2.1. Protein Production TheBL21-Gold(DE3)pLys bacteria strain was transformed with the plasmid encod- ing for the αRep protein and grown on 2YT-agar plates supplemented with ampicillin (100 µg/mL) and 1% glucose. One clone was picked and grown overnight at 37 ◦ C in 2YT media supplemented with ampicillin (100 µg/mL) and 1% glucose. The overnight preculture was used to inoculate 1 L of fresh 2YT medium containing 100 µg/mL ampicillin and distributed into two 2 L flasks. Cells were grown at 37 ◦ C until the optical density at 600 nm reached 0.8. Protein expression was then induced by the addition of 500 µM IPTG and cultures were further incubated at 37 ◦ C for 4 h. Cells were harvested by centrifugation for 20 min at 5000× g at 4 ◦ C and resuspended into 40 mL of washing buffer (50 mM sodium phosphate, 300 mM NaCl, 20 mM imidazole pH 8) supplemented with a tablet of EDTA-free anti-protease cocktail (Roche Diagnostics GmbH, Mannheim, Germany). Suspended cells were stored at −80 ◦ C. 3.2.2. Protein Purification Frozen cells from 1 L of bacterial culture were defrosted and a tablet of EDTA-free anti-protease cocktail (Roche) and 2 µL of DNase I were added. After 30 min of incubation at 4 ◦ C, cells were lysed by five 30 s cycles of sonication on ice and centrifuged for 1 h at 10,000× g at 4 ◦ C. His-tagged proteins were purified from the crude supernatant using nickel-affinity chromatography on a 5 mL Ni-NTA column. The column was washed with 150 mL of washing buffer (50 mM sodium phosphate, 300 mM NaCl, 20 mM imidazole pH 8) and proteins were eluted in six 1 mL fractions in the elution buffer (50 mM sodium phosphate, 300 mM NaCl, 300 mM imidazole pH 8). A typical yield of about 100 mg of protein per liter of culture was obtained. Eluted fractions were analyzed by SDS- PAGE and pure fractions of proteins were pooled. αRep protein was recovered and concentrated, resulting in a 5 mL solution that was loaded on a gel filtration column (HiLoad 16/600 Superdex 75, GE Healthcare, Buc, France) equilibrated with coupling buffer (10 mM HEPES 150 mM NaCl pH 7.75). Elution fractions after gel filtration were pooled and stored at 4 ◦ C until chemical coupling. Purity was checked by SDS-PAGE and concentrations were obtained from the absorbance at 280 nm measured on a Nanodrop spectrophotometer (Thermo Scientific, Wilmington, DE, USA) using an extinction coefficient at 280 nm of 58,900 M−1 cm−1 calculated from the protein sequence (https://web.expasy. org/protparam/ accessed on 28 October 2022). 3.3. Covalent Anchoring of Compound 4 into (A3A30 )Y26C In a 50 mL conical tube, 6.25 mL of a 200 µM Co(III)-mal cofactor solution in DMSO was added (final concentration 50 µM) to a solution of (A3A30 )Y26C in 10 mM HEPES 150 mM NaCl pH 7.75 at a final concentration of 5 µM, in the presence of DMSO (25% in the final coupling solution) for a final volume of 25 mL. The solution was stirred for 3 days at 4 ◦ C. The mixture was then extracted with 25 mL of 2-butanone at 0 ◦ C to remove the excess unreacted Co(III)-mal cofactor, and the remaining solution was cooled in an ice bath. It was then centrifuged at 5000× g for 5 min. The pellet containing the (A3A30 )Y26C- Co(III)Mal-PPIXMME biohybrid, (A3A30 )Y26C-4, was resuspended in 10 mM HEPES, 150 mM NaCl, pH 8. The obtained protein solution was then loaded on a gel filtration column (HiLoad 16/600 Superdex 75, GE Healthcare), which was equilibrated with 10 mM HEPES 150 mM NaCl, pH 8, and the (A3A30 )Y26C-4 cofactor was eluted in a single peak. The purified biohybrid was quantified from its absorbance at 280 nm, measured on a Nanodrop spectrophotometer. Int. J. Mol. Sci. 2022, 23, 14640 12 of 15 3.4. Photocatalysis Experiments To a degassed solution of 1 mM [RuII (bpy)3 ]Cl2 and 100 mM NaAsc, the biohybrid protein was added to lead to a final concentration of 7 µM in a total volume of 3 mL of 0.1 M phosphate buffer, pH 7. The solution was further degassed with Ar and then irradiated with a SugarCube white LED (200 Wm−2 ; filtered at λ > 420 nm) for 25 h. During irradiation, the solution was stirred and thermostated at 20 ◦ C with the RM6 Lauda cooling system. Gas samples from the overhead space were automatically injected into a MicroGC Fusion Gas Analyzer (Inficon, Bad Ragaz, Switzerland) every hour. For CO2 reduction, the cell was saturated with CO2 gas. All experiments were made at least in duplicate. 3.5. Spectroscopic Characterization of the (A3A30 )Y26C-4 Biohybrid Electronic absorption spectra of the oxidized and reduced (A3A30 )Y26C-4 biohybrid were recorded and compared to the spectra of other Co-PPIX–protein complexes previously reported in the literature. The 1 mL samples of a 7.5 µM solution of purified (A3A30 )Y26C-4 were prepared in argon-sparged 10 mM HEPES, NaCl 150 mM buffer, pH 8.0 and placed in a septum-sealed cuvette to ensure anaerobic conditions. The reduction was initiated via the addition of 1.5 mM sodium dithionite. Spectra were recorded every five minutes at RT until the absorbance changes were no longer observed. 3.6. Spectroscopic Characterization of the (A3A30 )Y26C-4 Biohybrid Circular dichroism (CD) spectra were recorded from 185 nm to 260 nm with a data pitch of 0.2 nm, a scan speed of 50 nm min−1 , a response time of 0.5 s, and a bandwidth of 2 nm using quartz cells with a 1 mm path length, on a Jasco J810 dichrograph. Each spectrum was recorded five times and averaged. The CD signal was corrected with buffer subtraction, resulting in the ellipticity (θ) in mdeg. Data were acquired from a 2.2 µM sample of the biohybrid previously desalted on a Zeba spin desalting column (Thermo Scientific, Wilmington, DE, USA) equilibrated with 20 mM sodium phosphate buffer, pH 7.5. 3.7. Computational Methods The coordinates from the modeled A3A30 -Y26C-MnTPP were obtained from a previous study [21]. The covalent MnTPP cofactor was removed and MD simulation was performed to obtain the most representative structure of the native form of A3A30 . Cys26 was mutated using the Dunbrack library available at UCSF Chimera [30]. The cofactor 4 structure (Figure S2) was optimized with Gaussian16 [38] via QM calculations in water solvent (SMD) [39] using B3LYP functional and Grimme’s dispersion D3 with 6-31G(d,p) for CHONS atoms and SDD + F for Co [31]. The oxidation state of Co was set to III and the spin state was singlet. Covalent dockings were performed using GOLD5.8.0 [31], using as the score function Goldscore with parameters adapted for metals [40]. Covalent linkage was performed with a CA-Cys26 Radius search set to 15 Å and 50 runs were performed. Flexibility to certain residues was given by allowing free rotation of the side chains. MD simulations were prepared using xleap [41], using as the force field FF19SB for proteins, TIP3P for water, and GAFF for non-standard residues. MCPB.py was used to parametrize cofactor 4 using the Seminario method [42] for force constants and parameters and RESP for charge, and the RESP method for charges [43]. Optimization, frequency, and RESP calculations were performed at the same level of theory previously specified. The sys- tem was solvated using a cubic water box and Na+ as counter ions were added to neutralize. MD simulations were performed using AMBER20 software [32]. Three different energy minimizations were performed to relax the system, followed by a heating step in which the system equilibrated from 0 K to 300 K. Finally, NVT and NPT equilibrations were performed, followed by a 400 ns production run. Detailed computational [29,31,32,38,39,41–45] can be found in the Supplementary Materials. Int. J. Mol. Sci. 2022, 23, 14640 13 of 15 4. Conclusions In summary, an artificial metalloenzyme was obtained by the covalent coupling of Co(III)Mal-PPIXMME 4 to the mutant (A3A30 )Y26C of an artificial bidomain protein named alpha Rep (αRep). This new artificial metalloenzyme (A3A30 )Y26C-4 was characterized by Maldi-TOF MS, circular dichroism, and UV–Visible spectrophotometry. A comparison of the maxima of absorption of (A3A30 )Y26C-4 with those of various Co(II)– and Co(III)– PPIX-substituted hemoproteins (Table 1) suggested the possible coordination of the cobalt ion by an axial histidine. However, investigation of the structure of this biohybrid and the positioning of the residues near the metal complex using theoretical experiments suggested the possible coordination of the metal by an axial glutamine Gln283, although it would be transient in nature and mutagenic efforts may be needed to improve axial coordination. The (A3A30 )Y26C-4 biohybrid was then shown to be able to catalyze, not only the pho- toinduced production of H2 , but also the reduction of CO2 into CO. While the Co(III)porphyrin catalyst alone showed a low activity with respect to these two reactions, more than 10 times more H2 and 4 times more CO2 were produced in the reaction, whereby the Co(III)porphyrin was buried within the hydrophobic cavity of the protein. Based on these results, we will thus be able in the future to optimize our biohybrid by carrying out different mutants for axial coordination and also to introduce positive charges on the proximal face of the biohybrid, which should increase the local concentration of CO2 in the enzymatic cavity to enhance its conversion into CO. Supplementary Materials: The following supporting information can be downloaded at: https: //www.mdpi.com/article/10.3390/ijms232314640/s1. Author Contributions: Conceptualization, R.R., J.-P.M. and Z.H.; Chemical synthesis and analy- sis G.A.O.U. and R.R.; Protein production and purification, A.U. and R.R.; Biochemical analysis, G.A.O.U., R.R. and A.U.; Catalysis experiments, E.P., Z.H., G.A.O.U. and R.R., Modeling studies, L.T.-S. and J.-D.M.; Writing—original draft preparation, R.R. and J.-P.M.; Writing—review and editing, J.-P.M. and R.R. All authors have read and agreed to the published version of the manuscript. Funding: This research was funded by the French National Research Agency (Carb2zyme, grant no.: ANR-17-CE11-0014-02, LOCO, grant no.: ANR-19-CE05-0020-02 and LABEX CHARMMMAT, grant no.: ANR-11-LABEX0039). Data Availability Statement: Not applicable. Acknowledgments: We thank CNRS, CEA Saclay, ICMMO, and University Paris-Saclay for the financial support. We thank H. Rezaei and INRAe for access to the dichrograph. Conflicts of Interest: The authors declare no conflict of interest. References 1. Zee, D.Z.; Chantarojsiri, T.; Long, J.R.; Chang, C.J. Metal-polypyridyl catalysts for electro- and photochemical reduction of water to hydrogen. Acc. Chem. Res. 2015, 48, 2027–2036. [CrossRef] 2. Landrou, G.; Panagiotopoulos, A.A.; Ladomenou, K.; Coutsolelos, A.G. Photochemical hydrogen evolution using Sn-porphyrin as photosensitizer and a series of Cobaloximes as catalysts. J. Porphyr. Phthalocyanine 2016, 20, 534–541. [CrossRef] 3. Rao, H.; Wang, Z.Y.; Zheng, H.Q.; Wang, X.B.; Pan, C.M.; Fan, Y.T.; Hou, H.Y. Photocatalytic hydrogen evolution from a cobalt/nickel complex with dithiolene ligands under irradiation with visible light. Catal. Sci. Technol. 2015, 5, 2332–2339. [CrossRef] 4. Rao, H.; Yu, W.Q.; Zheng, H.Q.; Bonin, J.; Fan, Y.T.; Hou, H.Y. Highly efficient photocatalytic hydrogen evolution from nickel quinolinethiolate complexes under visible light irradiation. J. Power Sources 2016, 324, 253–260. [CrossRef] 5. Dempsey, J.L.; Brunschwig, B.S.; Winkler, J.R.; Gray, H.B. Hydrogen evolution catalyzed by cobaloximes. Acc. Chem. Res. 2009, 42, 1995–2004. [CrossRef] 6. Tong, L.; Duan, L.; Zhou, A.; Thummel, R.P. First-row transition metal polypyridine complexes that catalyze proton to hydrogen reduction. Coord. Chem. Rev. 2020, 402, 2–20. [CrossRef] 7. Wang, P.; Liang, G.; Reddy, M.R.; Long, M.; Driskill, K.; Lyons, C.; Donnadieu, B.; Bollinger, J.C.; Webster, C.E.; Zhao, X. Electronic and Steric Tuning of Catalytic H2 Evolution by Cobalt Complexes with Pentadentate Polypyridyl-Amine Ligands. J. Am. Chem. Soc. 2018, 140, 9219–9229. [CrossRef] [PubMed] Int. J. Mol. Sci. 2022, 23, 14640 14 of 15 8. Sun, H.; Han, Y.; Lei, H.; Chen, M.; Cao, R. Cobalt corroles with phosphonic acid pendants as catalysts for oxygen and hydrogen evolution from neutral aqueous solution. Chem. Commun. 2017, 53, 6195–6198. [CrossRef] [PubMed] 9. Natali, M.; Luisa, A.; Iengo, E.; Scandola, F. Efficient photocatalytic hydrogen generation from water by a cationic cobalt(II) porphyrin. Chem. Commun. 2014, 50, 1842–1844. [CrossRef] [PubMed] 10. Beyene, B.B.; Mane, S.B.; Hung, C.H. Highly efficient electrocatalytic hydrogen evolution from neutral aqueous solution by a water-soluble anionic cobalt(ii) porphyrin. Chem. Commun. 2015, 51, 15067–15070. [CrossRef] [PubMed] 11. Edwards, E.H.; Jelušić, J.; Chakraborty, S.; Bren, K.L. Photochemical hydrogen evolution from cobalt microperoxidase-11. J. Inorg. Biochem. 2021, 217, 111384. [CrossRef] [PubMed] 12. Edwards, E.H.; Le, J.M.; Salamatian, A.A.; Peluso, N.L.; Leone, L.; Lombardi, A.; Bren, K.L. A cobalt mimochrome for photochem- ical hydrogen evolution from neutral water. J. Inorg. Biochem. 2022, 230, 111753. [CrossRef] [PubMed] 13. Onoda, A.; Hayashi, T. Artificial hydrogenase: Biomimetic approaches controlling active molecular catalysts. Curr. Opin. Chem. Biol. 2015, 25, 133–140. [CrossRef] [PubMed] 14. Bacchi, M.; Berggren, G.; Niklas, J.; Veinberg, E.; Mara, M.W.; Shelby, M.L.; Poluektov, O.G.; Chen, L.X.; Tiede, D.M.; Cavazza, C.; et al. Cobaloxime-based artificial hydrogenases. Inorg. Chem. 2014, 53, 8071–8082. [CrossRef] [PubMed] 15. Bacchi, M.; Veinberg, E.; Field, M.J.; Niklas, J.; Matsui, T.; Tiede, D.M.; Poluektov, O.G.; Ikeda-Saito, M.; Fontecave, M.; Artero, V. Artificial Hydrogenases Based on Cobaloximes and Heme Oxygenase. ChemPlusChem 2016, 81, 1083–1089. [CrossRef] [PubMed] 16. Soltau, S.R.; Niklas, J.; Dahlberg, P.D.; Poluektov, O.G.; Tiede, D.M.; Mulfort, K.L.; Utschig, L.M. Aqueous light driven hydrogen production by a Ru-ferredoxin-Co biohybrid. Chem. Commun. 2015, 51, 10628–10631. [CrossRef] 17. Sommer, D.J.; Vaughn, M.D.; Ghirlanda, G. Protein secondary-shell interactions enhance the photoinduced hydrogen production of cobalt protoporphyrin IX. Chem. Commun. 2014, 50, 15852–15855. [CrossRef] 18. Alcala-Torano, R.; Halloran, N.; Gwerder, N.; Sommer, D.J.; Ghirlanda, G. Light-Driven CO2 Reduction by Co-Cytochrome b 562 . Front. Mol. Biosci. 2021, 8, 609–654. [CrossRef] 19. Di Meo, T.; Ghattas, W.; Herrero, C.; Velours, C.; Minard, P.; Mahy, J.P.; Ricoux, R.; Urvoas, A. αRep A3: A Versatile Artificial Scaffold for Metalloenzyme Design. Chem. Eur. J. 2017, 23, 10156–10166. [CrossRef] 20. Di Meo, T.; Kariyawasam, K.; Ghattas, W.; Valerio-Lepiniec, M.; Sciortino, G.; Maréchal, J.-D.; Minard, P.; Mahy, J.P.; Urvoas, A.; Ricoux, R. Functionalized Artificial Bidomain Proteins Based on an α-Solenoid Protein Repeat Scaffold: A New Class of Artificial Diels–Alderases. ACS Omega 2019, 4, 4437–4447. [CrossRef] 21. Kariyawasam, K.; Di Meo, T.; Hammerer, F.; Valerio-Lepiniec, M.; Sciortino, G.; Maréchal, J.D.; Minard, P.; Mahy, J.P.; Urvoas, A.; Ricoux, R. An Artificial Hemoprotein with Inducible Peroxidase- and Monooxygenase-Like Activities. Chem. Eur. J. 2020, 26, 14929–14937. [CrossRef] [PubMed] 22. Ellsworth, R.K. Chromatographic separation of milligram quantities of protoporphyrin IX monomethyl ester from protoporphyrin IX and its dimethyl ester. Anal. Biochem. 1969, 3, 377–380. [CrossRef] 23. Koshiyama, T.; Yokoi, N.; Ueno, T.; Kanamaru, S.; Nagano, S.; Shiro, Y.; Arisaka, F.; Watanabe, Y. Molecular Design of Heteropro- teinAssemblies Providing a Bionanocup as a Chemical Reactor. Small 2008, 4, 50–54. [CrossRef] [PubMed] 24. Léger, C.; Di Meo, T.; Aumont-Nicaise, M.; Velours, C.; Durand, D.; Li de la Sierra-Gallay, I.; van Tilbeurgh, H.; Hildebrandt, N.; Desmadril, M.; Urvoas, A.; et al. Ligand-Induced Conformational Switch in an Artificial Bidomain Protein Scaffold. Sci. Rep. 2019, 9, 1178. [CrossRef] [PubMed] 25. Wang, Y.; Wen, X.; Cui, X.; Zhang, X.P. Enantioselective radical cyclization for construction of 5-membered ring structures by metalloradical C-H alkylation. J. Am. Chem. Soc. 2018, 140, 4792–4796. [CrossRef] 26. Di Bilio, A.J.; Chang Bo, T.K.; Malmström, G.; Gray, H.B.; Göran Karlsson, B.; Nordling, M.; Pascher, T.; Lundberg, L.G. Electronic absorption spectra of M(II)(Met121X) azurins (MCo, Ni, Cu; XLeu, Gly, Asp, Glu): Charge-transfer energies and reduction potentials. Inorg. Chim. Acta 1992, 198–200, 145–148. [CrossRef] 27. Lu, H.; Dzik, W.I.; Xu, X.; Wojtas, L.; de Bruin, B.; Zhang, P. Experimental evidence for cobalt(III)-carbene radicals: Key intermediates in cobalt(II)-based metalloradical cyclopropanation. J. Am. Chem. Soc. 2011, 133, 8518–8521. [CrossRef] 28. Perkins, L.J.; Weaver, B.R.; Buller, A.R.; Burstyn, J.N. De novo biosynthesis of a nonnatural cobalt porphyrin cofactor in E. coli and incorporation into hemoproteins. Proc. Natl. Acad. Sci. USA 2021, 118, e2017625118. [CrossRef] [PubMed] 29. Case, D.A.; Belfon, K.; Ben-Shalom, I.Y.; Brozell, S.R.; Cerutti, D.S.; Cheatham, T.E.I.; Cruzeiro, V.W.D.; Darden, T.A.; Duke, R.E.; Giambasu, G.; et al. AMBER 2020; University of California: San Francisco, CA, USA, 2020. 30. Shapovalov, M.V.; Dunbrack, R.L., Jr. A Smoothed Backbone-Dependent Rotamer Library for Proteins Derived from Adaptive Kernel Density Estimates and Regressions. Structure 2011, 19, 844–858. [CrossRef] [PubMed] 31. Ehlers, A.W.; Böhme, M.; Dapprich, S.; Gobbi, A.; Höllwarth, V.; Jonas, A.; Köhler, K.F.; Stegmann, R.; Veldkamp, A.; Frenking, G. A set of f-polarization functions for pseudo-potential basis sets of the transition metals Sc/Cu, Y/Ag and La/Au. G. Chem. Phys. Lett. 1993, 208, 111–114. [CrossRef] 32. Li, P.; Merz, K.M., Jr. MCPB.py: A Python Based Metal Center Parameter Builder. J. Chem. Inf. Model. 2016, 56, 599–604. [CrossRef] 33. Chen, J.; Wang, J.; Zhu, W. Zinc ion-induced conformational changes in new Delphi metallo-β-lactamase 1 probed by molecular dynamics simulations and umbrella sampling. Phys. Chem. Chem. Phys. 2017, 19, 3067–3075. [CrossRef] [PubMed] 34. Teets, T.S.; Nocera, D.G. Photocatalytic hydrogen production. Chem. Commun. 2011, 47, 9268–9274. [CrossRef] [PubMed] 35. Morris, A.J.; Meyer, G.J.; Fujita, E. Molecular Approaches to the Photocatalytic Reduction of Carbon Dioxide for Solar Fuels. Acc. Chem. Res. 2009, 42, 1983–1994. [CrossRef] [PubMed] Int. J. Mol. Sci. 2022, 23, 14640 15 of 15 36. Marianov, A.N.; Kochubei, A.S.; Roman, T.; Conquest, O.J.; Stampfl, C.; Jiang, Y. Resolving Deactivation Pathways of Co Porphyrin-Based Electrocatalysts for CO2 Reduction in Aqueous Medium. ACS. Catal. 2021, 11, 3715–3729. [CrossRef] 37. Doukov, T.I.; Blasiak, L.C.; Seravalli, J.; Ragsdale, S.W.; Drennan, C.L. Xenon in and at the End of the Tunnel of Bifunctional Carbon Monoxide Dehydrogenase/Acetyl-CoA Synthase. Biochemistry 2008, 47, 3474–3483. [CrossRef] 38. Frisch, M.J.; Trucks, G.W.; Schlegel, H.B.; Scuseria, G.E.; Robb, M.A.; Cheeseman, J.R.; Scalmani, G.; Barone, V.; Petersson, G.A.; Nakatsuji, H.; et al. Gaussian 16, Revision C.01; Gaussian, Inc.: Wallingford, CT, USA, 2016. 39. Marenich, A.V.; Cramer, C.J.; Truhlar, D.G. Universal Solvation Model Based on Solute Electron Density and on a Continuum Model of the Solvent Defined by the Bulk Dielectric Constant and Atomic Surface Tensions. J. Phys. Chem. B 2009, 113, 6378–6396. [CrossRef] [PubMed] 40. Sciortino, G.; Garribba, E.; Maréchal, J.-D. Validation and Applications of Protein–Ligand Docking Approaches Improved for Metalloligands with Multiple Vacant Sites. Inorg. Chem. 2019, 58, 294–306. [CrossRef] [PubMed] 41. Cornell, W.D.; Cieplak, P.; Bayly, C.I.; Gould, I.R.; Merz, K.M.; Ferguson, D.M.; Spellmeyer, D.C.; Fox, T.; Caldwell, J.W.; Kollman, P.A. A Second-Generation Force Field for the Simulation of Proteins, Nucleic Acids, and Organic Molecules. J. Am. Chem. Soc. 1995, 117, 5179–5197. [CrossRef] 42. Seminario, J.M. Calculation of intramolecular force fields from second-derivative tensors. Int. J. Quant. Chem. 1996, 60, 1271–1277. [CrossRef] 43. Bayly, C.I.; Cieplak, P.; Cornell, W.; Kollman, P.A. A well-behaved electrostatic potential based method using charge restraints for deriving atomic charges: The RESP model. J. Phys. Chem. 1993, 97, 10269–10280. [CrossRef] 44. Ryckaert, J.P.; Ciccotti, G.; Berendsen, H.J.C. Numerical integration of the cartesian equations of motion of a system with constraints: Molecular dynamics of n-alkanes. J. Comput. Phys. 1977, 23, 327–341. [CrossRef] 45. Essmann, U.; Perera, L.; Berkowitz, M.L. A smooth particle mesh Ewald method. J. Chem. Phys. 1995, 103, 8577–8593. [CrossRef]