A fundamental catalytic difference between zinc and manganese dependent enzymes revealed in a bacterial isatin hydrolase

Scientific Reports, Aug 2018

The catalytic mechanism of the cyclic amidohydrolase isatin hydrolase depends on a catalytically active manganese in the substrate-binding pocket. The Mn2+ ion is bound by a motif also present in other metal dependent hydrolases like the bacterial kynurenine formamidase. The crystal structures of the isatin hydrolases from Labrenzia aggregata and Ralstonia solanacearum combined with activity assays allow for the identification of key determinants specific for the reaction mechanism. Active site residues central to the hydrolytic mechanism include a novel catalytic triad Asp-His-His supported by structural comparison and hybrid quantum mechanics/classical mechanics simulations. A hydrolytic mechanism for a Mn2+ dependent amidohydrolases that disfavour Zn2+ as the primary catalytically active site metal proposed here is supported by these likely cases of convergent evolution. The work illustrates a fundamental difference in the substrate-binding mode between Mn2+ dependent isatin hydrolase like enzymes in comparison with the vast number of Zn2+ dependent enzymes.

A PDF file should load here. If you do not see its contents the file may be temporarily unavailable at the journal website or you do not have a PDF plug-in installed and enabled in your browser.

Alternatively, you can download the file locally and open with any standalone PDF reader:


A fundamental catalytic difference between zinc and manganese dependent enzymes revealed in a bacterial isatin hydrolase

Abstract The catalytic mechanism of the cyclic amidohydrolase isatin hydrolase depends on a catalytically active manganese in the substrate-binding pocket. The Mn2+ ion is bound by a motif also present in other metal dependent hydrolases like the bacterial kynurenine formamidase. The crystal structures of the isatin hydrolases from Labrenzia aggregata and Ralstonia solanacearum combined with activity assays allow for the identification of key determinants specific for the reaction mechanism. Active site residues central to the hydrolytic mechanism include a novel catalytic triad Asp-His-His supported by structural comparison and hybrid quantum mechanics/classical mechanics simulations. A hydrolytic mechanism for a Mn2+ dependent amidohydrolases that disfavour Zn2+ as the primary catalytically active site metal proposed here is supported by these likely cases of convergent evolution. The work illustrates a fundamental difference in the substrate-binding mode between Mn2+ dependent isatin hydrolase like enzymes in comparison with the vast number of Zn2+ dependent enzymes. Introduction The catalytic activity of metallohydrolase enzymes strongly depends on the identity of the metal and on the nature of its binding site. Zn2+ bound hydrolytic enzymes, in particular metallo-β-lactamases1, probably constitute the most extensively studied cases for such a group. Depending on the enzymatic class, hydrolytic enzymes with both a mononuclear or dinuclear Zn2+ sites have been identified as widely distributed throughout all kingdoms of life2. The catalytic mechanism for the mononuclear and dinuclear Zn2+ sites have been debated for years3,4,5, and only recently the collection of accumulated knowledge allowed the proposition of a plausible mechanism6,7,8,9. Despite Zn2+ being the most widely distributed hydrolytic cation, it has been shown that Zn2+ also possesses an inhibitory potential for certain metallohydrolases as, for example, in carboxypeptidase A10. The specialized metallo-enzyme amidohydrolase superfamily (AHS) is able to carry out a diverse range of chemical reactions including the degradation of metabolic precursors like pyrimidine (dihydropyrimidinases) and hydantoin, (hydantoinases)11. The AHS members are often categorised as both Zn2+ and Mn2+ dependent hydrolases. However, amidohydrolases involved with purines degradation in both bacteria and plants are mainly Mn2+ dependent allantoinases12,13, even though Zn2+ dependent allantoinase activity has been reported14. The compound isatin is a metabolic precursors and was originally identified as an intermediate product of the indole-3-acetic acid (IAA) degradation pathway in Bradyrhizobium diazoefficiens15. Hydrolytic activity of isatin hydrolase (IH) was first demonstrated in various symbiotic rhizobial species16. Recent bioinformatics studies indicates that IH are found widespread in bacteria including pathogens specific to the human gut and plants, e.g., B. enterica and R. solanacearum17. Isatin hydrolase A from Labrenzia aggregata (LaIHA) and R. solanacearum (RsIHA) share 59% sequence identity while the two putative orthologues LaIHA and LaIHB in L. aggregata share 51% sequence identity. Both LaIHA and LaIHB contain the central metal binding motif HxG[T/A]HxDxPxH in each protomer of the catalytically active dimer. Structural characterization of LaIHB revealed a novel fold, which later was also attributed to other amidohydrolases such as bacterial kynurenine formamidase B (KynB) (EC The fold is generally described as an α/β hydrolase with a central scaffold resembling the swivelling β/α/β fold18, while the majority of AHS members contains a monomeric (β/α)8-TIM like-barrel structural fold19. A notable difference between LaIHB and kynurenine formamidase from Bacillus anthracis (BaKynB) is the presence of a mononuclear manganese-binding site in LaIHB, while the structure of KynB accommodates a bi-nuclear zinc site20. LaIHB shows a strict specificity for Mn2+ over Zn2+, as reported by kinetic measurements17, and rationalised by computer simulations21. Here we present two crystal structures of LaIHA, one bound to the hydrolytic product isatinate, and one bound to benzyl benzoate, as well as the apo structure of homologous RsIHA. Together with results from enzyme kinetic analysis, the resolved structures are used to describe the molecular basis for the Mn2+-dependent mechanism of hydrolysis, showing that IHA is a true orthologue of IHB. For the first time, the mechanism of a Mn2+ dependent hydrolysis reaction is characterised in detail by computer simulations using multiscale quantum mechanics/molecular mechanics (QM/MM)22 and ab initio molecular dynamics (MD) simulations23. The simulations were crucial to unravelling the essential catalytic role of the conserved residues His79 and His207 for the formation of the product. Based on the identification of mechanistically important residues presented in this work, including the Mn2+ binding residues in the LaIHA active site, we further present evidence that the Mn2+ dependent allantoate amidohydrolase (AAH) (E.C. displays a striking similarity in the active site geometry despite the differences in the overall protein fold. As this likely represents a case of convergent evolution, we propose that the generalised catalytic mechanism described here for the IHs may also apply to the AAH family and other similar Mn2+ dependent hydrolases. Materials and Methods Cloning, expression and purification of LaIHA and RsIHA Chemicals were purchased from Sigma-Aldrich (Norway) unless stated otherwise. The open reading frame encoding LaIHA (UniProtKB: A0P0F0) was amplified by PCR from a boiled colony of L. aggregata IAM12614. DNA fragments were isolated and digested with BamHI and EcoRI and cloned into the T7-RNA polymerase dependent E. coli expression plasmid pT7H624. The 6xHis containing LaIHA was expressed in E. coli BL21 AI cells (Invitrogen) in 2xTYE-medium for 4 hours at 37 °C. Cells were harvested by centrifugation, re-suspended, and lysed by sonication in a buffer containing 50 mM Tris-HCl pH 8.0 and 0.5 M NaCl supplemented with a protease inhibitor tablet (Roche cOmplete, EDTA-free). Insoluble material was removed by centrifugation (4000 g, 10 min.). The soluble protein extract was batch-adsorbed onto 25 mL Ni-NTA agarose resin (Qiagen) per litre of original culture and loaded into liquid chromatography columns. The protein loaded Ni-NTA columns were washed with >20 column volumes (CV) of equilibration buffer (50 mM Tris-HCl pH 8.0, 500 mM NaCl, 50 mM imidazole). The protein was eluted from the resin with elution buffer (50 mM Tris-HCl pH 8.0, 500 mM NaCl, 10 mM Na2EDTA). Fractions containing the recombinant proteins were pooled, and the buffer changed into a low salt buffer (10 mM Tris-HCl pH 8.0, 100 mM NaCl, and 1 mM Na2EDTA) on a Sephadex G-25 column (GE Healthcare). Before crystallisation and kinetic experiments, LaIHA was further treated with 10 mM EDTA and dialysed (500 fold dilution) into a stability buffer (5 mM bis-tris pH 7.0, 100 mM NaCl and 1 mM DTT). The protein solutions were centrifuged at 180,000 g for 10 minutes at 4 °C, and the supernatant could hereafter be stored in a stable condition at 4 °C for at least 12 months. The RsIHA (UniProtKB: Q8XYC3) was ordered from Genscript and inserted into the expression vector pET-M11. The six-histidine containing construct was expressed in E. coli BL21 gold cells, in LB-medium (containing 50 µg/mL kanamycin) overnight at 18 °C. Cells were harvested by centrifugation, re-suspended, and lysed by a bead-beater (Biospec) in a lysis buffer (50 mM Tris-HCl pH 8.0 and 0.1 M NaCl), protease inhibitor tablet (Roche cOmplete, EDTA-free). Insoluble material was removed by centrifugation, and the lysate was supplemented with 20 mM imidazole before loaded onto a pre-equilibrated HisTrap HP (GE). The column was washed with 10 CV with equilibration buffer (50 mM Tris-HCl pH 8.0, 0.1 M NaCl, and 20 mM imidazole) and eluted with elution buffer (50 mM Tris-HCl pH 8.0, 0.1 M NaCl, and 300 mM imidazole). The pooled fractions were supplemented with a molar 1:50 ratio of tobacco etch virus (TEV) protease and dialysed overnight against dialysis buffer (50 mM Tris-HCl pH 8.0, 100 mM NaCl). The RsIHA was again loaded to the HisTrap HP column and the flow-through was collected. RsIHA was concentrated to 20 mg/mL and loaded onto a Superdex 200 10/300 GL size exclusion column (GE) pre-equilibrated in SEC buffer (50 mM Tris-HCl pH 8.0, 100 mM NaCl). Pure RsIHA were pooled, concentrated, and stored at −80 °C. RsIHA was treated like LaIHA before crystallisation and kinetic experiments. Crystallization of LaIHA and RsIHA ligand complexes The pre-treatment of LaIHA and RsIHA with an excess of EDTA and dialysis was mainly carried out to remove any potential trace of metals, leaving an estimated EDTA concentration at 0.02 mM. Prior to crystallization 1 mM MnCl2 was added. LaIHA and RsIHA were crystallized at 22 °C using a protein concentration of 15 mg/mL in 5 mM Bis-Tris pH 7.0, 100 mM NaCl, and 1.0 mM MnCl2. LaIHA was crystallised by vapour diffusion using sitting drops 1 + 1 μL (LaIHA: Reservoir) over 100 μL reservoir solution. The best diffracting crystals were obtained in 20% glycerol and 20% PEG 1500. The LaIHA:isatinate structure was obtained by soaking the crystals for approximately 1 min with 500 µM isatin before flash-freezing the crystals in liquid nitrogen (LN2). This condition also served as cryo-protection. RsIHA was also crystallised with vapour diffusion using hanging drops, 1 + 1 μL (RsIHA: Reservoir) over 400 μL reservoir solution sealed with immersion oil (56822, Sigma). The best diffracting crystals of RsIHA grew in 1 M succinate, 100 mM HEPES pH 6.5. Cryo-protection was achieved by adding 1 μL of 60% w/v PEG400 suspended in reservoir solution to the drop. The crystals were mounted and flash-cooled immediately after. Crystals of LaIHA:benzyl benzoate was obtained in 28% PEG 1500 and 1 mM MnCl2 with a hanging drop setup using 250 μL reservoir and 1 + 1 μL drops. With a protein concentration of 20 mg/mL used. The hanging drop wells were sealed with immersion oil. Crystals appeared within one week. Cryo-protection was obtained by adding glycerol to the reservoir solution to a final concentration of ~12%, which after 12 hours would allow additional vapor diffusion to take place before mounting and flash-cooling in LN225. Data collection and structure determination All datasets were collected at 100 K. For LaIHA:Benzyl benzoate a dataset extending to 1.50 Å resolution was collected at a wavelength of 1.0004 Å. The space group was determined to be P1, and a collection strategy was calculated by iMOSFLM26. The data were processed with XDS27. A test set of 5% was used for Rfree calculation28. Phasing was performed by molecular replacement (MR) using LaIHB (PDB ID: 4J0N) as a search model for Phaser29. Initial model building and refinement were performed in Phenix30. The final model was built using Coot31 with small ligand models and restraints produced using eLBOW30. For LaIHA:Isatinate, a diffraction dataset was collected at beamline P13 at PETRA III (Hamburg, Germany)32. The space group was determined and would only allow a P1 lattice. A complete dataset was collected at 1.79 Å resolution with and oscillation range of Ω = 0.01° and processed with XDS. The structure was solved with MR using a LaIHA: Benzyl benzoate monomer as the search model and performed with Phaser in Phenix. Diffraction data on RsIHA was collected at beamline ID30B at ESRF (Grenoble, France). The space group was determined and allowed P 32 2 1, a complete dataset with at Ω = 0.05° over 126° was collected. The data extended to 2.65 Å and was processed with XDS. The structure was solved with MR using LaIHA:Benzyl benzoate monomer as the search model in Phenix. Model building and refinement was performed using coot33 and phenix.refine within the phenix package30. NCS restraints were applied throughout the refinement, however released in the final round (Number of molecules in the asymmetric of each crystals structure have been added in the Supplementary Table S1)34. All structure validations were performed using MolProbity30. Structure analysis and figure preparation were done using PyMOL Molecular Graphics System, Version, Schrödinger, LLC. The finalised model and structure factors were deposited to Protein Data Bank (PDB) and given the PDB accession codes (LaIHA:Benzyl benzoate PDB ID 5NNA; RsIHA PDB ID 5NMP; LaIHA:isatinate PDB ID 5NNB). Enzymatic assay The measurement of the activity, and its metal dependence, were performed as described by Bjerregaard-Andersen et al.17. The method is based on an increased absorption of the product isatinate at 368 nm. This can be followed using a standard spectrophotometer. Measurements were performed using a JASCO V-630 spectrophotometer with a cuvette light path of 1.0 cm. An extinction coefficient of 4.5 × 103 cm−1mol−1L was used for isatinate. The data were fitted with the Michaelis-Menten equation assuming one binding site using the software GraphPad Prism 6. Sequence analysis Structural homologues of LaIHA and E. coli allantoate amidohydrolase (EcAAH) were identified using the DaliLite v. 335, followed by inspection in PyMoL. The sequences of the identified homologues were then imported in Jalview36 using the fetch from PDB option and combined with sequences of LaIHA, RsIHA, and Rhodococcus rhodochrous HpoH (RrHpoH). Sequences with the following UniProt accession codes (in parentheses) were used: LaIHA (A0P0F0), LaIHB (A0NLY7), RsIHA (Q8XYC3), EcKynB (B4E9I9), PaKynB (Q9I234), BaKynB (Q81PP9), BsAHD (P84132), MjAHD (Q58193), RrHpoH (B5MAD9). For AAH alignment, the following accession codes were used: EcAAH (P77425), AtAAH (Q8VXY9), BvHYD (A4JQA0), BmHYD (A0A0H3KRF1), BcHYD (B4EHA1), GeRAC (Q53389), SkSYN (Q96W94). Sequence alignments were performed using the Clustal algorithm with default settings in Jalview. Phylogenetic trees were calculated in Jalview as average distance using identity percentage and exported to TreeDyn 198.337 for visualisation. The sequences were included in alignment as supplied by the PDB. Molecular modelling The system consisting of LaIHA, isatin, 47331 water molecules, and 19 Na+ ions was placed in a 105 × 126 × 118 Å3 simulation box. The initial geometry of the protein was taken from the X-ray structure of LaIHA:isatinate resolved by us and described in this text. Isatin was introduced in the system by molecular replacement, fitting the aromatic moiety of isatinate. The Mn2+ ion, isatin, the catalytic water, the side chains of the ligating amino acids (His69, His73, Asp75, Gln219), and of the neighbouring His79, His207, Asp193 residues were treated at a quantum-mechanical level (QM) using density functional theory with the Becke and Lee-Yang-Parr (BLYP) approximations for the exchange-correlation functional38,39. The Kohn-Sham orbitals for the valence electrons were expanded over a DZVP Gaussian basis set40 and an auxiliary plane-wave basis set with a cut-off of 240 Ry. The core electrons were integrated out using Goedecker-Teter-Hutter pseudopotentials41. The remainder of the system was described at the molecular-mechanics (MM) level using the Amberff14SB force field42. In a first step, the system was equilibrated by full classical molecular dynamics (MD) simulations in the NpT ensemble (T = 300 K, p = 1 atm), constraining the QM region to its initial positions, and using a time-step of 1.5 fs. Thereafter, QM/MM Born-Oppenheimer MD simulations were performed at 300 K within the NVT ensemble using the Nosé–Hoover thermostat43,44,45 and a time-step of 0.25 fs. The simulations were run using the CP2K package (https://www.cp2k.org/)46,47,48. The mechanism of hydrolysis was sampled by coupling QM/MM MD to metadynamics simulations49,50 using collective variables defined as differences between coordination numbers (Visualized in Fig. S1, in the supporting information): $$\begin{array}{rcl}{\rm{CV1}} & = & {\rm{CN}}\,({{\rm{C}}}_{{\rm{isatin}}}:{{\rm{N}}}_{{\rm{isatin}}})-{\rm{CN}}\,({{\rm{C}}}_{{\rm{istatin}}}:{{\rm{O}}}_{{\rm{water}}});\\ {\rm{CV2}} & = & {\rm{CN}}\,({{\rm{N}}}_{{\rm{His83}}}:{\rm{H}})-{\rm{CN}}\,({{\rm{O}}}_{{\rm{water}}}:{\rm{H}});\\ {\rm{CV3}} & = & {\rm{CN}}\,({{\rm{N}}}_{{\rm{istatin}}}:{\rm{H}})-{\rm{CN}}\,({{\rm{N}}}_{{\rm{His212}}}:{\rm{H}}),\end{array}$$ where CN is the coordination number between two atoms. Each CN is defined as: $$CN(i:j)=\frac{1-{({d}_{ij}/{d}_{0})}^{p}}{1-{({d}_{ij}/{d}_{0})}^{p+q}}$$ (1) where \({d}_{ij}\,\,\) is the distance between atoms \(i\), \(j\), and \({d}_{0}\), p, and q are free parameters. In our simulations, \({d}_{0}\) was set to 1.6 Å for CV1, and 1.1 Å for CV2 and CV3. Values of \(p\) = 12 and \(q\) = 14 were used for all the three CVs. The d0, p, q values where chosen so that the profile of the sigmoidal function (equation 1) did not change significantly during a standard thermal oscillation of the bond, while it signalled the formation or breaking of a chemical bond. Gaussian hills with a height of 2 kcal/mol were spanned every 100 step (i.e., every 25 fs). A Gaussian spread of 0.15 was used for all three collective variables. The free energy profile for the reaction was obtained as the sum of the Gaussians added during the whole metadynamics simulation. The three CVs used in this study were chosen to take into account both the nucleophilic addition/elimination at the carbonyl group and the proton rearrangements necessary to yield the product. Runs using any combination of only two of these CVs did not produce any reactive profile, indicating that all the three CVs are necessary for the determination of the reaction coordinate. With this setup, the system reacted after 1.1 ps of metadynamics simulations, the run was continued until a total time of 1.85 ps to reach convergence in the free energy profile. We assumed convergence of the profile after observation of a second re-crossing of the barrier (Supplementary Fig. S2). Results IHA contains a mononuclear manganese binding site involving an additionally conserved glutamine The structures of LaIHA:benzyl benzoate and LaIHA:isatinate were determined at 1.50 Å and 1.79 Å resolution, respectively, while the one of RsIHA was determined at 2.65 Å resolution. Data collection and model refinement statistics are collected as described in Supplementary Table S1. Both LaIHA structures are highly similar to the one of LaIHB with a root mean square deviation (rmsd) value of ~0.7 Å computed on all main chain atoms. Both LaIHB and LaIHA are dimers with two exchanged regions – a small N-terminal α-helical exchange and a larger β-hairpin exchange (Fig. 1a). A hallmark of this fold is the contribution of two conserved residues of one monomer in the formation of the substrate-binding pocket located in the other monomer. For the IH’s these two residues are two tryptophans (Fig. 1b). Figure 1 (a) Overall structure of the LaIHA dimer found in the asymmetric unit, the protomers are colored green and cyan and shown in two different orientations rotated by 90 degrees. Manganese and highlighted residues a location of active sites in the dimer. (b) Isatinate and benzyl benzoate in the substrate binding pocket of LaIHA. Isatinate is coordinating bidentate directly to the manganese. Parts of the pocket, residues Trp61 and Trp59, are contributed by the opposing monomer via the hairpin swap. Key binding pocket residues of the LaIHB binding pocket (grey) are functionally conserved. (c) Catalytically important residue in the active site of LaIHA: benzyl benzoate. The corresponding site of LaIHA: benzyl benzoate is superposed (transparent). The manganese is found in octahedral coordination similarly to that described in17. Gln219 resides in a double conformation and only partially coordinates to the manganese. Also, W1025 is found in a double conformation (denoted A and B in Fig. 1). Note that Asp75 is coordinating bidentate in LaIHA:isatinate while monodentate in LaIHA: benzyl benzoate. (d) The isatin hydrolysis by LaIHA (circles) and RsIHA (triangles) follows Michaelis-Menten kinetics, and the parameters are collected in f). (e) Both LaIHA and RsIHA display strong manganese dependency. (f) Kinetic parameters of LaIHA and RsIHA collected with previous values from LaIHB and activated mutant LaIHB S225C. All measurements in (d) and (e) were performed in triplets. Full size image The highly conserved Mn2+ binding site is located at the bottom of the substrate-binding pocket. The Mn2+ is found in an octahedral complex coordinated by His69, His73, Asp75, and two water molecules (Fig. 1c). Asp75 forms a bidentate contact in LaIHA:Benzyl benzoate and a mono-dentate electrostatic contact in the LaIHA:isatinate structures. The mono or bi-dentate coordination results in two or one water molecules as additional ligands, respectively, preserving the octahedral geometry. The side chain carbonyl of Gln219 in LaIHA (Ala224 in LaIHB) is found in alternative conformations, one of which coordinates Mn2+. The non-binding conformation allows additional water to approach to Mn2+ (W1025A/B, Fig. 1c). Sequence analysis reveals that Gln219 is part of a motif, conserved as either GLAS (e.g. in LaIHB) or GLQC (e.g. in LaIHA) (Figs 1 and 2a). The short residues motifs AS and QC are conserved as sequence pairs, which could indicate a functional dependency. Interestingly, the residues S or C are found to be the key residue involved in regulating the proton flow through the water channel. A mutation of the serine residue to a cysteine to form a GLAC motif caused a gain in activity in LaIHB17. The GLQC/GLAS sequence, combined with the conserved manganese binding site appears to be a signature motif for isatin hydrolase activity. In KynB, the comparable position holds a conserved I(L/I)E motif, which is also found in the two uncharacterized structures of amidohydrolases (AHD) from Bacillus stearothermophilus (BsAHD) and Methanococcus jannaschii (MjAHD) (Fig. 2a, Full sequence alignment shown in Supplementary Fig. S3). The motif is absent in RrHpoH, which has confirmed cyclase activity. Whether the I(L/I)E motif and perhaps, in particular, the glutamate is indicative of activity on linear amides or a binuclear metal site remains uncertain. Figure 2 (a) Alignment of selected sequences of IHA and AAH homologues with known structures. Conserved residues are highlighted in blue. Residues involved in metal coordination are indicated by a dot, residues presumed to be involved in proton transfer are indicated with an asterisk. All are completely conserved, apart from the LaIHA Asp193 (yellow) which is functionally conserved. The orthologue is defining aromatic binding pocket motif for isatin, WXW/FXW, is highlighted in orange. The key GLQC/GLAS motif of IHA and IHB, indicating isatin hydrolase activity, is highlighted in green, with a similarly conserved motif in KynB homologues highlighted in red. (b) Phylogram of IHA homologues from A indicates IHA-like and KynB-like sequences, RrHpoH proposed to carry a novel cyclase activity seems to belong to neither of these. An example of the fold (dark and light grey) and active site position is shown. (c) Phylogram of AAH homologues. Note the large diversity between EcAHH (prokaryote) and AtAAH (eukaryote). An example of the fold (green and cyan) and active site position is shown, His226 (AtAAH) is delivered from opposing monomer through domain swap of AtAAH. (d) Active site residues of BaKynB (Yellow, black numbers) superposed with LaIHA (Grey, grey numbers), the mononuclear Mn2+ in LaIHA is labelled (MnI) and the binuclear Zn2+ sites in BaKynB are labelled ZnI and ZnII. (e) Active site residues of EcAAH (green and cyan, black numbers) superposed manually with LaIHA (grey, grey numbers), the mononuclear Mn2+ from EcAAH labelled MnII. Full size image Activity measurements confirm that both LaIHA and RsIHA have isatin hydrolase activity (Fig. 1d,f). The determination of the catalytic parameters assuming Michaelis-Menten kinetics yields Km values for LaIHA and RsIHA of 16 μM and 10 μM, respectively. Structural superposition revealed identical metal binding sites in the two homologues. Thus, a metal dependence analysis was only performed on LaIHA. LaIHA shows the highest activity in the presence of Mn2+ confirming that this enzyme is Mn2+ dependent. However, the relative activity of 15.1 ± 2.2% is also observed in the presence of Cd2+. Notably, LaIHA could not be activated in the presence of both Zn2+ and Cu2+ (Fig. 1e). QM/MM calculations reveal the importance of an Asp-His dyad in the proton abstraction In LaIHA, both the carbonyl group of the substrate and the catalytic water coordinate to the Mn2+ ion. Like in other hydrolytic enzymes, the nucleophilic water is activated by a pre-organized base, which can efficiently extract a proton to form hydroxide (OH−). In LaIHA, His207 acts as such base. The positively charged protonated His207-H+ is stabilised by the presence of a hydrogen-bonded partner, i.e., Asp193. The water-His207-Asp193 H-bond network closely resembles the Ser-His-Asp triad common to serine proteases. His79, also H-bonded to the catalytic water, completes a complex His79-water-His207-Asp193 proton-shuttle system (Fig. 2b,d). Water deprotonation in LaIHA is extremely efficient and occurs spontaneously during the QM/MM simulations at room temperature21. Activation of the catalytic water is favoured by its isolation from the water channel that connects the substrate cavity to the exterior. Confinement of the catalytic water has been reported as an important structural requirement in other hydrolytic (metallo)-enzymes51,52,53,54,55. Figure 3 depicts the molecular details of the hydrolytic reaction. The nucleophilic attack of OH− to the carbonyl of isatin leads to the formation of a tetrahedral intermediate. Our model predicts an activation barrier of a few (i.e., less than 6) kcal mol−1. Thus, as in other metallohydrolases (i.e., zinc-β-lactamases), this step is not rate-limiting for the reaction. The tetrahedral intermediate is only metastable and it can easily recombine into the reactant state. In fact, the following separation of the scissile C-N bond constitutes the rate-limiting step of the reaction with a barrier height of ~12 kcal mol−1. Figure 3 Reaction mechanism describing the hydrolysis of isatin. The reaction is visualised from the initial nucleophilic attack of OH− to the carbonyl of isatin until the final substrate isatinate is formed. The final step involves isatinate isomerization, promoted by the strong intra-molecular H-bond between its amino and α-carbonyl groups, as well as the reformation of the hydrogen bond between Asp193 and His207. Free energy differences were estimated from metadynamics simulations. Full size image Our data report that C-N separation occurs through a dissociative pathway. The transition state is stabilised by formation of two hydrogen-bonds (H-bonds), the first between the nucleophilic OH− and the Nδ atom of His79, and the second between the isatin N atom and the Nε-H moiety of His207 (Fig. 3, Supplementary movie). The final product formation of isatinate is accompanied by synchronous proton exchange in both of these two H-bonds. Our simulations indicate that the proton transfers occur after the transition state has been reached and the C-N bond is broken. Kinetic measurements in a deuterated environment did not reveal any observable kinetic isotope effect on the reaction, providing an additional indication that proton transfers are not involved in the rate-limiting step of the enzymatic process (data not shown). The preference of a dissociative pathway over an associative or concerted mechanism may be a consequence of the fact that the N-atom in isatin acts as a poor base. In general, the π conjugation between the lone pair of N and the aromatic ring present in aryl-amines decreases the pKa value by few pH units. As a result, we observe that formation of a stable H-bond with His207, and consequently the proton transfer, occurs only upon a major elongation of the isatin C-N bond, which is associated to increased localization of the electron density at the N atom. Additional stabilization of the moiety into the product is obtained by the deprotonation of the nucleophile OH− group by His79 to produce a carboxylate group chelating the Mn2+ as observed in the crystal structure. The isatinate conformer produced by hydrolysis is not the most stable one. In fact, the α-carbonyl can isomerise via a bicycle motion, similar to that described in excited state isomerisation of retinal in rhodopsin56. This isomerised product found in our simulations is in agreement with the binding geometry of the product, isatinate, described in the crystal structure (Fig. 1b). In particular, the final binding state of isatinate shows bidentate coordination to the manganese ion, and similar stacking interactions with the indole rings of Trp59, Trp61, Trp80 and Tyr204 (Fig. 1b). The strong exothermic isomerisation step (−10 kcal mol−1) yields a structure of isatinate with a strong intra-molecular H-bond between the amino and the α-carbonyl groups. This possible intermediate most likely constitutes the driving force of the reaction and traps the isatinate as a ligand to Mn2+ in its product state before it is replaced by water and released from the binding pocket (Fig. 3). Isatin hydrolase and kynurenine formamidase active-site conservation indicates a similar mechanism The recently determined crystal structure of kynurenine formamidase from Bacillus anthracis (BaKynB) contains a binuclear Zn2+ site and thus is part of the EC 3.5 class that has a similar overall protein fold when compared with LaIHA (sequence identity 24%). Figure 2d shows the superposition of the metal binding sites for LaIHA:isatinate and BaKynB (PDB ID: 4COG). The structural architecture of the metal binding sites is highly conserved between both enzymes (Fig. 2d). The most striking difference is the position of Gln219 in LaIHA, which is occupied by glutamate (Glu172) in BaKynB. This residue is conserved for all verified KynB enzymes. In BaKynB, the His60 (equivalent to His79 in LaIHA) is proposed to be necessary for the hydrolysis reaction to take place, by accepting a proton from the bi-coordinated activated water. In the mechanism for the isatin hydrolysis presented here, both His79 and His207 are involved in the catalytic mechanism (Supplementary Movie M1). In BaKynB, the Zn2+ in position (I) ZnI2+ is octahedrally coordinated and bound in an equivalent position as Mn2+ in LaIHA. In a homologous structure of KynB from Burkholderia cenopacia (BcKynB), this position is occupied by Cd2+. The Zn2+ in position (II), ZnII2+ is also found in an octahedral conformation. Convergent evolution of the active-site geometry in LaIHA and EcAAH Amidohydrolases with known manganese dependency where identified through literature searches and examined manually for structural similarity. This approach revealed two candidates that, upon close inspection, shared metal binding sites residues and residues identified as catalytically important. The allantoate amidohydrolases (AAHs) are part of the purine degradation pathway and hydrolyse the allantoate to ureidoglycolate, carbon dioxide, and ammonium in the process, the allantoin pathway is found in both plants13 and bacteria57 (Fig. 2a,c). The identified structures were manually superposed with the active site of representatives from the AAH’s, (Fig. 2e). The structure for E. coli AAH (EcAAH, PDB ID: 4PXD) was used as input for the DALI server35. The found hits were visually inspected for structural homology and included in alignment to confirm residue conservation (Fig. 2a). Discussion The Baltic sea bacteria Labrenzia aggregata genome includes two homologous open reading frames for isatin hydrolases A and B that share 51% sequence identity (SI). The more distant homolog from Ralstonia solanacearum (RsIHA) shares 59% SI with LaIHA and 51% SI with LaIHB. Unlike the whole sequence, the active site and substrate binding pocket are highly conserved among these proteins. Even though the key substrate binding pocket residues originally identified in the structure of LaIHB - Phe63, Trp65 and Trp84 and Phe20917 do not share equivalent positions in the sequences of LaIHA or RsIHA, their aromatic side chains do adopt an identical structural conformation in the respective binding pockets, thus indicating functional conservation (Fig. 1b). Compared to LaIHB, the Mn2+ binding sites in both RsIHA and LaIHA include Gln219, also responsible for increased metal specificity21. The crystal structure of BaKynB from Bacillus anthracis (PDB ID: 4COG) presents a glutamate (Glu173, Fig. 2d) adopting an equivalent conformation as that of Gln219 in LaIHA. In conjunction with the conserved His161 and Asp56 (BaKynB numbering), it forms a secondary Zn2+ site (ZnII) binding site in the presented binuclear metal binding (Fig. 2d). The remaining metal-coordinating residues are identical to both RsIHA and LaIHA. It cannot be completely ruled out that, in certain conditions, Gln219 would allow the formation of a secondary metal-binding site in LaIHA. This is however unlikely, due to the different Lewis-acid/base properties of an amide group, compared to those of a carboxylate. The structure of LaIHB, which features an alanine residue in the position of Gln219, does not have this option and thus is expected to be strictly mononuclear, with the highest hydrolytic activity measured in the presence of Mn2+, as also observed for LaIHA presented in this manuscript. In the case of the amidohydrolase diaminopimelate desuccinylase (dapE), it was found that the compound L-captopril only binds to the Zn2+ bound form of the enzyme and not to the physiologically relevant Mn2+ bound form, thus stressing the importance of identifying the relevant metal58. Different metallo-β-lactamases can bind and be catalytically active in the presence of both one and two Zn2+ metals in the binding site. In both cases, the reaction mechanisms were characterized by computational modelling in the past (1 Zinc: B1 BcII from Bacillus cereus and B2 CphA from Aeromonas hydrophila, 2 Zinc: B1 CcrA from Bacteroides fragilis)59,60,61, the mono and binuclear reaction scheme for Zn2+ binding beta-lactamases has recently been revised in Lisa MN et al.6. The mechanism of hydrolysis in LaIHA shares similarities with those found in single-Zn2+ enzymes. In particular, the reaction proceeds through a two-step mechanism, where the nucleophilic attack on the amide carbonyl of isatin is followed by the breaking of the C-N bond. On the contrary, in CcrA from B. fragilis, the reaction follows a concerted mechanism where the nucleophilic attack of the hydroxyl takes place simultaneously with the opening of the β-lactam ring. Common to all these enzymes is that the rate-limiting second step of the reaction is constituted by the dissociation of the C-N bond, and that the pathway to the product is facilitated by proton transfer on the N atom that usually occurs at the transition state or immediately after. The proton-shuttling pathways are nonetheless different in the Zn-bound enzymes and IH. The lone pair of the amidic nitrogen of isatin is π-conjugated with both the amidic carbonyl and the aromatic phenyl ring present in the molecular structure. As a consequence of the resulting electronic delocalisation, its basicity remains low even after the initial attack of the OH− and the subsequent formation of the tetrahedral intermediate. In fact, the evolution from the TS to the product requires both the pre-organised His207-Asp193 dyad shuttling one proton on the N atom and the stabilisation of C-N heterolytic breakage by a secondary proton transfer from the carboxylic group to His79. As a consequence, again differently from metallo-β-lactamases, the product formation in LaIHA leaves the protein in a higher protonation state than the initial complex. The deprotonation of the active site after the release of the product likely occurs through the water channel originally described for LaIHB17. Our in silico studies yield further support to the necessity for proton removal, in order to recover the catalyst, and allow further enzymatic cycles. As the calculations produced a slightly endothermic profile (+2 kcal), the excess energy released by deprotonation of the protein may also constitute the driving force for the enzymatic process. Though, we stress that due to high computational costs, the free energy profile was obtained with a resolution that is of the same order as the reaction free energy balance. The sequence alignment and structural superposition confirm that not only the metal coordinating residues (Fig. 2a–c, dots), but also the residues involved in the proton transfer (Fig. 2a–c, asterisk) are highly conserved. Only the position corresponding to LaIHA residue Asp193 (Fig. 2a) alternates between aspartate and glutamate, thus showing anyway functional conservation. The conservation suggests functional importance and, supported by the structural superposition and mechanistic elucidation, points to a common catalytic role. This is particularly interesting in the light of the sequence variation within the IH-like and AAH-like folds as illustrated in Fig. 2b,c,e. Here, the phylogenetic tree of proteins belonging to the respective folds indicates a large sequence variation, perhaps best exemplified by the significant divergence of EcAAH compared to Arabidopsis thaliana AAH (AtAAH), i.e., prokaryotic and eukaryotic. However, both the overall fold, displaying domain-swapped dimers, and the active site geometry are highly conserved (Fig. 2c–e). It remains to be established whether the catalytic mechanism exemplified in LaIHA by our structural and in silico studies is specific to the IH-like enzymes, or could represent a broader case of convergent evolution across a range of AHS metallo-hydrolases. In the case of the AAH homologues, the Mn2+ binding site is structurally completely conserved, including the binding residues in the MnI site (His69, His73 and Asp75 for LaIHA), and the proton transfers triad (His79, His203 and Asp193) in LaIHA compared to Glu127, His382 and His226 in EcAAH, with the peculiar difference that in EcAAH the Mn2+ occupies site II (See comparison in Fig. 2c,e). The IH, KynB and AAH families are not the only ones to have divergent metal site constellation, despite similar metal binding residues, the ureidoglycolate amidohydrolase from Arabidopsis thaliana (AtUAH) has a binuclear Mn2+ site62 (PDB ID 4PXB), while the putative hydantoinases (PDB ID 5I4M, 5THW and 4WJB) and β-alanine synthase from Saccharomyces kluyveri63 (PDB ID 2VL1) have binuclear Zn2+ sites, the L-N-carbamoylase from Geobacillus stearothermophilus has a mononuclear Co2+ site64 (PDB ID 3N5F). The model for hydrolysis presented here would only allow a Mn2+ in site I, and a Zn2+ would act as an inhibitor. In particular, according to previous simulations, Zn2+ is not able to bind both isatin and the water, disrupting the preorganization of the active site and preventing the initial nucleophilic attack21. Thus, it remains an open question whether the Mn2+ in site II also act as an inhibitor or merely adopt an alternative position in the structure of EcAAH due to the crystal conditions. It is a reoccurring problem for metal dependent hydrolases that the majority of the information is collected from in vitro experiments and that conflicting conclusions are reached when it comes to metal specificity. As in the case of KynB were a binuclear Zn2+ is described in the structure20 and Zn2+ dependent activity is presented, however it still remains an open question whether this is the physiological metal, as in other systems such as Ralstonia metallidurans a 15 to 20‐fold activation of kynurenine formamidase was detected after treatment with Mn2+ 65. The major difference between the Michaelis complexes in LaIHA/B and zinc-β-lactamases is constituted by the binding mode of the substrate. β-lactam substrates, bind in the proximity of the metal(s), but their amide moiety do not participate directly to their coordination sphere(s) as is the case for the reacting cyclic amide in the IH. It is carboxylate groups of the carpapenem family, that later during the catalysis interact directly with the Zn2+ ion as proposed is for several beta lactamases6,7,8 including the more recent New Delhi metallo-β-lactamases9,66,67. In LaIHA it is the oxygen of the electrophilic amide carbonyl on isatin that is directly coordinated to the catalytic Mn2+ ion (Fig. 3). The enzyme kinetics analyses, accompanying structural determination of LaIHA and RsIHA, present striking evidence that RsIHA, LaIHA, and LaIHB are true orthologues enzymes. These data support the claim in Bjerregaard-Andersen et al.17 that isatin metabolism is present in bacteria native to the human gut and strengthen the hypothesis that the molecule isatin may be a signalling molecule that links the gut-brain axis under conditions such as stress68,69. Additional information Publisher's note: Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. References 1. Baier, F., Chen, J., Solomonson, M., Strynadka, N. C. J. & Tokuriki, N. Distinct metal isoforms underlie promiscuous activity profiles of metalloenzymes. ACS Chem. Biol. 10, 1684–1693 (2015). ArticlePubMed Google Scholar2. Bebrone, C. Metallo-β-lactamases (classification, activity, genetic organization, structure, zinc coordination) and their superfamily. Biochem. Pharmacol. 74, 1686–1701 (2007). ArticlePubMed Google Scholar3. Llarrull, L. I., Tioni, M. F. & Vila, A. J. M. Content and localization during turnover in B. cereus Metallo-β-lactamase. J. Am. Chem. Soc. 130, 15842–15851 (2008). ArticlePubMed Google Scholar4. Badarau, A. & Page, M. I. The variation of catalytic efficiency of Bacillus cereus metallo-β-lactamase with different active site metal ions. Biochemistry 45, 10654–10666 (2006). ArticlePubMed Google Scholar5. Hu, Z. X., Periyannan, G., Bennett, B. & Crowder, M. W. Role of the Zn-1 and Zn-2 sites in Metallo-β-lactamase L1. J. Am. Chem. Soc. 130, 14207–14216 (2008). ArticlePubMedPubMed Central Google Scholar6. Lisa, M. N. et al. A general reaction mechanism for carbapenem hydrolysis by mononuclear and binuclear metallo-β-lactamases. Nat. Commun. 8, 538 (2017). ADSArticlePubMedPubMed Central Google Scholar7. Wu, S., Xu, D. & Guo, H. QM/MM studies of monozinc β-lactamase CphA suggest that the crystal structure of an enzyme-intermediate complex represents a minor pathway. J. Am. Chem. Soc. 132, 17986–17988 (2010). ArticlePubMedPubMed Central Google Scholar8. Sgrignani, J. et al. On the active site of mononuclear B1 metallo β-lactamases: a computational study. J. Comput. Aided. Mol. Des. 26, 425–435 (2012). ADSArticlePubMed Google Scholar9. Das, C. K. & Nair, N. N. Hydrolysis of cephalexin and meropenem by New Delhi metallo-β-lactamase: the substrate protonation mechanism is drug dependent. Phys. Chem. Chem. Phys. 19, 13111–13121 (2017). ArticlePubMed Google Scholar10. Gomez-Ortiz, M., Gomis-Ruth, F. X., Huber, R. & Aviles, F. X. Inhibition of carboxypeptidase A by excess zinc: analysis of the structural determinants by X-ray crystallography. FEBS Lett 400, 336–40 (1997). ArticlePubMed Google Scholar11. Barba, M., Glansdorff, N. & Labedan, B. Evolution of cyclic amidohydrolases: a highly diversified superfamily. J. Mol. Evol. 77, 70–80 (2013). ADSArticlePubMed Google Scholar12. Ho, Y. Y., Hsieh, H. C. & Huang, C. Y. Biochemical characterization of allantoinase from Escherichia coli BL21. Protein J. 30, 384–394 (2011). ArticlePubMed Google Scholar13. Werner, A. K., Sparkes, I. A., Romeis, T. & Witte, C. P. Identification, biochemical characterization, and subcellular localization of allantoate amidohydrolases from Arabidopsis and soybean. Plant Physiol. 146, 418–430 (2008). ArticlePubMedPubMed Central Google Scholar14. Mulrooney, S. B. & Hausinger, R. P. Metal ion dependence of recombinant Escherichia coli allantoinase. J. Bacteriol. 185, 126–134 (2003). ArticlePubMedPubMed Central Google Scholar15. Jensen, J. B., Egsgaard, H., Van Onckelen, H. & Jochimsen, B. U. Catabolism of indole-3-acetic acid and 4- and 5-chloroindole-3-acetic acid in Bradyrhizobium japonicum. J. Bacteriol. 177, 5762–5766 (1995). ArticlePubMedPubMed Central Google Scholar16. Olesen, M. R. & Jochimsen, B. U. Identification of enzymes involved in indole-3-acetic acid degradation. Plant and Soil 186, 143–149 (1996). Article Google Scholar17. Bjerregaard-Andersen, K. et al. A proton wire and water channel revealed in the crystal structure of isatin hydrolase. J. Biol. Chem 289, 21351–9 (2014). ArticlePubMedPubMed Central Google Scholar18. Herzberg, O. et al. Swiveling-domain mechanism for enzymatic phosphotransfer between remote reaction sites. Proc. Natl. Acad. Sci. USA 93, 2652–7 (1996). ADSArticlePubMed Google Scholar19. Seibert, C. M. & Raushel, F. M. Structural and catalytic diversity within the amidohydrolase superfamily. Biochemistry 44, 6383–6391 (2005). ArticlePubMed Google Scholar20. Diaz-Saez, L., Srikannathasan, V., Zoltner, M. & Hunter, W. N. Structures of bacterial kynurenine formamidase reveal a crowded binuclear zinc catalytic site primed to generate a potent nucleophile. Biochem. J. 462, 581–9 (2014). ArticlePubMedPubMed Central Google Scholar21. Uribe, L., Diezemann, G., Gauss, J., Morth, J. P. & Cascella, M. Structural origin of metal specificity in isatin hydrolase from Labrenzia aggregata investigated by computer simulations. Chem. Eur. J. 24, 5074–5077 (2018). ArticlePubMed Google Scholar22. Warshel, A. & Levitt, M. Theoretical studies of enzymic reactions: dielectric, electrostatic and steric stabilization of the carbonium ion in the reaction of lysozyme. J. Mol. Biol. 103, 227–49 (1976). ArticlePubMed Google Scholar23. Car, R. & Parrinello, M. Unified approach for molecular dynamics and density-functional theory. Phys. Rev. Lett. 55, 2471–2474 (1985). ADSArticlePubMed Google Scholar24. Christensen, J. H., Hansen, P. K., Lillelund, O. & Thogersen, H. C. Sequence-specific binding of the N-terminal three-finger fragment of Xenopus transcription factor IIIA to the internal control region of a 5S RNA gene. FEBS Lett. 281, 181–184 (1991). ArticlePubMed Google Scholar25. Teng, T. Y. & Moffat, K. Cooling rates during flash cooling. J. Appl. Crystallogr. 31, 252–257 (1998). Article Google Scholar26. Battye, T. G., Kontogiannis, L., Johnson, O., Powell, H. R. & Leslie, A. G. iMOSFLM: a new graphical interface for diffraction-image processing with MOSFLM. Acta Crystallogr. D Biol. Crystallogr. 67, 271–281 (2011). ArticlePubMedPubMed Central Google Scholar27. Kabsch, W. Xds. Acta Crystallogr. D Biol. Crystallogr. 66, 125–132 (2010). ArticlePubMedPubMed Central Google Scholar28. Brünger, A. T. Free R-value: a novel statistical quantity for assessing the accuracy of crystal structures. Nature 355, 472–475 (1992). ADSArticlePubMed Google Scholar29. Mccoy, A. J. et al. Phaser crystallographic software. J. Appl. Crystallogr. 40, 658–674 (2007). ArticlePubMedPubMed Central Google Scholar30. Adams, P. D. et al. PHENIX: a comprehensive Python-based system for macromolecular structure solution. Acta Crystallogr. D Biol. Crystallogr. 66, 213–221 (2010). ArticlePubMedPubMed Central Google Scholar31. Emsley, P. & Cowtan, K. Coot: model-building tools for molecular graphics. Acta Crystallogr. D Biol. Crystallogr. 60, 2126–2132 (2004). ArticlePubMed Google Scholar32. Cianci, M. et al. P13, the EMBL macromolecular crystallography beamline at the low-emittance PETRA III ring for high- and low-energy phasing with variable beam focusing. J. Synchrotron Radiat. 24, 323–332 (2017). ArticlePubMedPubMed Central Google Scholar33. Emsley, P., Lohkamp, B., Scott, W. G. & Cowtan, K. Features and development of Coot. Acta Crystallogr. D Biol. Crystallogr. 66, 486–501 (2010). ArticlePubMedPubMed Central Google Scholar34. Chen, V. B. et al. MolProbity: all-atom structure validation for macromolecular crystallography. Acta Crystallogr. D Biol. Crystallogr. 66, 12–21 (2010). ArticlePubMed Google Scholar35. Holm, L. & Rosenstrom, P. Dali server: conservation mapping in 3D. Nucleic Acids Res. 38, W545–W549 (2010). ArticlePubMedPubMed Central Google Scholar36. Waterhouse, A. M., Procter, J. B., Martin, D. M., Clamp, M. & Barton, G. J. Jalview Version 2–a multiple sequence alignment editor and analysis workbench. Bioinformatics 25, 1189–91 (2009). ArticlePubMedPubMed Central Google Scholar37. Dereeper, A. et al. Phylogeny.fr: robust phylogenetic analysis for the non-specialist. Nucleic Acids Res. 36, W465–W469 (2008). ArticlePubMedPubMed Central Google Scholar38. Becke, A. D. Density-functional exchange-energy approximation with correct asymptotic-behavior. Phys. Rev. A 38, 3098–3100 (1988). ADSArticle Google Scholar39. Lee, C. T., Yang, W. T. & Parr, R. G. Development of the Colle-Salvetti correlation-energy formula into a functional of the electron-density. Phys. Rev. B 37, 785–789 (1988). ADSArticle Google Scholar40. VandeVondele, J. & Hutter, J. Gaussian basis sets for accurate calculations on molecular systems in gas and condensed phases. J. Chem. Phys. 127, 114105 (2007). ADSArticlePubMed Google Scholar41. Goedecker, S., Teter, M. & Hutter, J. Separable dual-space Gaussian pseudopotentials. Phys. Rev. B 54, 1703–1710 (1996). ADSArticle Google Scholar42. Maier, J. A. et al. ff14SB: Improving the accuracy of protein side chain and backbone parameters from ff99SB. J. Chem. Theory Comput. 11, 3696–3713 (2015). ArticlePubMedPubMed Central Google Scholar43. Nosé, S. A Molecular-dynamics method for simulations in the canonical ensemble. Mol. Phys. 52, 255–268 (1984). ADSArticle Google Scholar44. Hoover, W. G. Canonical dynamics - equilibrium phase-space distributions. Phys. Rev. A 31, 1695–1697 (1985). ADSArticle Google Scholar45. Martyna, G. J., Klein, M. L. & Tuckerman, M. Nosé-Hoover chains - the canonical ensemble via continuous dynamics. J. Chem. Phys. 97, 2635–2643 (1992). ADSArticle Google Scholar46. Laino, T., Mohamed, F., Laio, A. & Parrinello, M. An efficient linear-scaling electrostatic coupling for treating periodic boundary conditions in QM/MM simulations. J. Chem. Theory Comput. 2, 1370–1378 (2006). ArticlePubMed Google Scholar47. Laino, T., Mohamed, F., Laio, A. & Parrinello, M. An efficient real space multigrid OM/MM electrostatic coupling. J. Chem. Theory Comput. 1, 1176–1184 (2005). ArticlePubMed Google Scholar48. VandeVondele, J. et al. QUICKSTEP: Fast and accurate density functional calculations using a mixed Gaussian and plane waves approach. Comput. Phys. Commun. 167, 103–128 (2005). ADSArticle Google Scholar49. Barducci, A., Bussi, G. & Parrinello, M. Well-tempered metadynamics: a smoothly converging and tunable free-energy method. Phys. Rev. Lett. 100, 020603 (2008). ADSArticlePubMed Google Scholar50. Laio, A. & Parrinello, M. Escaping free-energy minima. Proc. Natl. Acad. Sci. USA 99, 12562–12566 (2002). ADSArticlePubMed Google Scholar51. Piana, S., Bucher, D., Carloni, P. & Rothlisberger, U. Reaction mechanism of HIV-1 protease by hybrid carparrinello/classical MD simulations. J. Phys. Chem. B 108, 11139–11149 (2004). Article Google Scholar52. Cascella, M., Micheletti, C., Rothlisberger, U. & Carloni, P. Evolutionarily conserved functional mechanics across pepsin-like and retroviral aspartic proteases. J. Am. Chem. Soc. 127, 3734–3742 (2005). ArticlePubMed Google Scholar53. De Vivo, M., Ensing, B. & Klein, M. L. Computational study of phosphatase activity in soluble epoxide hydrolase: high efficiency through a water bridge mediated proton shuttle. J. Am. Chem. Soc. 127, 11226–11227 (2005). ArticlePubMed Google Scholar54. De Vivo, M. et al. Proton shuttles and phosphatase activity in soluble epoxide hydrolase. J. Am. Chem. Soc. 129, 387–394 (2007). ArticlePubMedPubMed Central Google Scholar55. Palermo, G., Campomanes, P., Cavalli, A., Rothlisberger, U. & De Vivo, M. Anandamide hydrolysis in FAAH reveals a dual strategy for efficient enzyme-assisted amide bond cleavage via nitrogen inversion. J. Phys. Chem. B 119, 789–801 (2015). ArticlePubMed Google Scholar56. Loppnow, G. R. & Mathies, R. A. Excited-state structure and isomerization dynamics of the retinal chromophore in rhodopsin from resonance Raman intensities. Biophys. J. 54, 35–43 (1988). ArticlePubMedPubMed Central Google Scholar57. Werner, A. K., Romeis, T. & Witte, C. P. Ureide catabolism in Arabidopsis thaliana and Escherichia coli. Nat. Chem. Biol. 6, 19–21 (2010). ArticlePubMed Google Scholar58. Uda, N. R. et al. Zinc-selective inhibition of the promiscuous bacterial amide-hydrolase DapE: implications of metal heterogeneity for evolution and antibiotic drug design. Metallomics 6, 88–95 (2014). ArticlePubMed Google Scholar59. Dal Peraro, M., Vila, A. J., Carloni, P. & Klein, M. L. Role of zinc content on the catalytic efficiency of B1 metallo β-lactamases. J. Am. Chem. Soc. 129, 2808–2816 (2007). ArticlePubMedPubMed Central Google Scholar60. Simona, F. et al. Common mechanistic features among metallo-β-lactamases: a computational study of Aeromonas hydrophila CphA enzyme. J. Biol. Chem. 284, 28164–28171 (2009). ArticlePubMedPubMed Central Google Scholar61. Dal Peraro, M., Llarrull, L. I., Rothlisberger, U., Vila, A. J. & Carloni, P. Water-assisted reaction mechanism of monozinc β-lactamases. J. Am. Chem. Soc. 126, 12661–12668 (2004). ArticlePubMed Google Scholar62. Shin, I., Han, K. & Rhee, S. Structural insights into the substrate specificity of (s)-ureidoglycolate amidohydrolase and its comparison with allantoate amidohydrolase. J. Mol. Biol. 426, 3028–3040 (2014). ArticlePubMed Google Scholar63. Andersen, B., Lundgren, S., Dobritzsch, D. & Piskur, J. A recruited protease is involved in catabolism of pyrimidines. J. Mol. Biol. 379, 243–250 (2008). ArticlePubMed Google Scholar64. Martinez-Rodriguez, S. et al. Mutational and structural analysis of L-N-carbamoylase reveals new insights into a peptidase M20/M25/M40 family member. J. Bacteriol. 194, 5759–68 (2012). ArticlePubMedPubMed Central Google Scholar65. Kurnasov, O. et al. Aerobic tryptophan degradation pathway in bacteria: novel kynurenine formamidase. FEMS Microbiol. Lett. 227, 219–227 (2003). ArticlePubMed Google Scholar66. Feng, H. et al. The mechanism of NDM-1-catalyzed carbapenem hydrolysis is distinct from that of penicillin or cephalosporin hydrolysis. Nat. Commun. 8, 2242 (2017). ADSArticlePubMedPubMed Central Google Scholar67. Tripathi, R. & Nair, N. N. Mechanism of meropenem hydrolysis by New Delhi metallo β-lactamase. ACS Catal. 5, 2577–2586 (2015). Article Google Scholar68. Sommer, T. et al. Enzymatic detection and quantification assay of isatin, a putative stress biomarker in blood. ACS Chem. Neuro. 6, 1353–1360 (2015). Article Google Scholar69. Medvedev, A., Igosheva, N., Crumeyrolle-Arias, M. & Glover, V. Isatin: role in stress and anxiety. Stress 8, 175–83 (2005). ArticlePubMed Google Scholar Download references Acknowledgements We thank Anette Kjems for her help in the cloning and purification of LaIHA. We also would like to thank Diamond Light Source (Didcot, United Kingdom) and the staff of beamline I02 for beam time and excellent assistance during data collection of LaIHA: benzyl benzoate. The data collection of RsIHA was performed on beamline ID30B at the European Synchrotron Radiation Facility (ESRF) (Grenoble, France). We are grateful to Cristoph Müller-Dieckmann at the ESRF for assisting in using the beamline. Data collection of IHA:isatinate was done at PETRA III located at DESY (Hamburg, Germany). We are grateful for beam time and assistance during this data collection. M.C. acknowledges the support of the Research Council of Norway through the CoE Hylleraas Centre for Quantum Molecular Sciences (Grant No. 262695) and by the Norwegian Supercomputing Program (NOTUR) (Grant No. NN4654K). LU acknowledges a fellowship from the Mainz Graduate School of Excellence as well support from the Centre for Theoretical and Computational Chemistry during a stay at the University of Oslo. M.C., G.D. and J.G. are supported by the Deutsche Forschungsgemeinschaft (DFG) via the TRR146 “Multiscale simulation methods for soft matter systems”. J.P.M., K.B.A. and T.S. was financially supported by the Norwegian Research Council (ES486454) and the Lundbeck Foundation (R34-A3616). Author information Author notes Theis Sommer, Kaare Bjerregaard-Andersen and Lalita Uribe contributed equally. AffiliationsNorwegian Center for Molecular Medicine, Nordic EMBL Partnership University of Oslo, Gaustadalléen 21, 0349, Oslo, NorwayTheis Sommer & J. Preben MorthDepartment of Chemistry, University of Oslo, Sem Sælands vei 26, 0371, Oslo, NorwayKaare Bjerregaard-Andersen & Michele CascellaInstitute for Experimental Medical Research, Oslo University Hospital, 0424, Oslo, NorwayJ. Preben MorthDepartment of Molecular Biology and Genetics, Aarhus University, Gustav Wieds vej 10C, 8000, Aarhus, DenmarkMichael EtzerodtInstitut für Physikalische Chemie, Johannes Gutenberg-Universität Mainz, Duesbergweg 10-14, 55128, Mainz, GermanyLalita Uribe, Gregor Diezemann & Jürgen GaussGraduate School Materials Science in Mainz, Johannes Gutenberg-Universität Mainz, Staudinger Weg 9, 55128, Mainz, GermanyLalita UribeHylleraas Centre for Quantum Molecular Sciences, University of Oslo, Sem Saelands vei 26, 0371, Oslo, NorwayMichele CascellaEnzyme and Protein Chemistry, Section for Protein Chemistry and Enzyme Technology, Department of Biotechnology and Biomedicine, Technical University of Denmark, Søltofts Plads, 2800, Kgs. Lyngby, DenmarkJ. Preben Morth AuthorsSearch for Theis Sommer in:Nature Research journals • PubMed • Google Scholar Search for Kaare Bjerregaard-Andersen in:Nature Research journals • PubMed • Google Scholar Search for Lalita Uribe in:Nature Research journals • PubMed • Google Scholar Search for Michael Etzerodt in:Nature Research journals • PubMed • Google Scholar Search for Gregor Diezemann in:Nature Research journals • PubMed • Google Scholar Search for Jürgen Gauss in:Nature Research journals • PubMed • Google Scholar Search for Michele Cascella in:Nature Research journals • PubMed • Google Scholar Search for J. Preben Morth in:Nature Research journals • PubMed • Google Scholar Contributions T.S. and K.B.A. performed most of the wet lab experiments, analysed the data and co-wrote the manuscript with J.P.M. and M.C. L.U. performed the molecular simulations studies supervised by M.C., J.G. and G.D. M.E. designed constructs and performed the initial expression and purification studies. J.P.M. supervised the optimization of protein expression, crystallization and data collection. All analysed the data and edited the manuscript. J.P.M., M.C., G.D. and J.G. obtained the funding. Competing Interests The authors declare no competing interests. Corresponding authors Correspondence to Michele Cascella or J. Preben Morth. Electronic supplementary material Supplementary InformationSupplementary Video Rights and permissions Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/. About this article Publication history Received 14 June 2018 Accepted 15 August 2018 Published 30 August 2018 DOI https://doi.org/10.1038/s41598-018-31259-y

This is a preview of a remote PDF: https://www.nature.com/articles/s41598-018-31259-y.pdf

Theis Sommer, Kaare Bjerregaard-Andersen, Lalita Uribe, Michael Etzerodt, Gregor Diezemann, Jürgen Gauss, Michele Cascella, J. Preben Morth. A fundamental catalytic difference between zinc and manganese dependent enzymes revealed in a bacterial isatin hydrolase, Scientific Reports, 2018, DOI: 10.1038/s41598-018-31259-y