Infrared spectroscopy reveals metal-independent carbonic anhydrase activity in crotonyl-CoA carboxylase/reductase

The conversion of CO2 by enzymes such as carbonic anhydrase or carboxylases plays a crucial role in many biological processes. However, in situ methods following the microscopic details of CO2 conversion at the active site are limited. Here, we used infrared spectroscopy to study the interaction of CO2, water, bicarbonate, and other reactants with β-carbonic anhydrase from Escherichia coli (EcCA) and crotonyl-CoA carboxylase/reductase from Kitasatospora setae (KsCcr), two of the fastest CO2-converting enzymes in nature. Our data reveal that KsCcr possesses a so far unknown metal-independent CA-like activity. Site-directed mutagenesis of conserved active site residues combined with molecular dynamics simulations tracing CO2 distributions in the active site of KsCCr identify an ‘activated’ water molecule forming the hydroxyl anion that attacks CO2 and yields bicarbonate (HCO3−). Computer simulations also explain why substrate binding inhibits the anhydrase activity. Altogether, we demonstrate how in situ infrared spectroscopy combined with molecular dynamics simulations provides a simple yet powerful new approach to investigate the atomistic reaction mechanisms of different enzymes with CO2.


3.) Figures S1-S11 4.) References
Electronic Supplementary Material (ESI) for Chemical Science.This journal is © The Royal Society of Chemistry 2024 eq. of the corresponding anhydride (crotonic anhydride: 65 µl, 0.4 mmol butyric anhydride: 65 µl, 0.4 mmol) was added and the solution was stirred on ice for 45 min and then directly injected into the HPLC-MS for purification.
Ethylmalonyl-CoA was synthesized by chemical coupling of crotonic anhydride with CoA followed by the addition of an ECR variant of C. crescentus (CcrC PAG ), NADPH, and KHCO 3 as described previously. [5]10 mg (0.0125 mmol, 1 eq.)CoA was dissolved in 1 mL 0.5 M NaHCO 3 .The mixture was cooled on ice and crotonic anhydride (3.2 µl, 0.0081 mmol, 1.6 eq) was added.After completion (around 30 min, the reaction mixture was added to 8 mL of 250 mM Tris/HCl pH 7.5 containing 30 mM NADPH, 250 mM KHCO 3 , and 100 nM CcrCc IA/CP/FG.The reaction was incubated for 30 min at 30°C, quenched with final a concentration of 5% formic acid and directly used for HPLC purification. All synthesized CoA-thioesters were purified using a 1260 Infinity LC system (Agilent) using a Gemini 10 mm NX-C18 110 Å, 100 x 21.2 mm, AXOA packed column (Phenomenex) as previously described. [5]Method: flow rate: 25 ml/min starting with 5 min of 5% MeOH in 50 mM ammonium bicarbonate pH 8.2, gradient 5-40 % MeOH in 15 min, 2 min wash at 95% MeOH, a re-equilibration 3 min at 5% MeOH.Purified CoA-esters were lyophilized and concentrations were determined by Uv/vis absorbance using extinction coefficients of ε 260 = 16.4 mM -1 cm -1 for saturated CoA esters and ε 260 =22.4 mM -1 cm -1 for unsaturated CoA esters. [4]rbonic anhydrase assay.In solution, CO 2 hydration activity was probed by UV/vis spectroscopy according to established protocols. [6]All experiments were performed in the CO 2free atmosphere of a N 2 -filled glovebox (Coy Lab), exploiting an USB spectrometer (Ocean Optics Optics USB2000) with a halogen/deuterium light source (Ocean Optics DH-2000).The cuvette holder was set to a constant temperature of 4°C via a circulation pump (Julabo F20/HC) operating outside the glovebox.Ahead of the experiment, 400 µl Tris/HCl buffer (25 mM, pH 8) was mixed with 100 µl bromothymol blue solution (10 mg/l) and 100 µl dilute protein solution (10 mg/l EcCa or KsCcr) or 100 µl buffer.All samples were store on ice.A glass vial containing 20 mL ice-cold buffer was purged with CO 2 for 2 h, sealed with a rubber stopper afterwards, and transferred into the glovebox.For the experiment, 600 µl sample was injected into a 1 mL quartz cuvette (d = 1 cm).Electronic spectra were recorded with a temporal resolution of 0.9 s.Approximately 60 s after the acquisition of spectra was started, 400 µl carbonated buffer was injected into the cuvette for a total volume of 1 mL.The cuvette was capped directly afterwards.The decrease of the characteristic bromothylmol blue band at 618 nm could then be followed over time to quantify acidification of the medium as a result of CO 2 hydration.The solution dropped from pH 8 to pH 6.2 over the course of the experiment.For the analysis of data, all spectra were fitted with a single Lorentz function (peak position = 618 nm, fwhm = 83 nm) and a polynomial background accounting for stray a light and baseline shifts.The "peak area" was then plotted as a function of time.
Infrared spectroscopy.The FTIR spectrometer (Tensor27, Bruker Optik, Germany) was equipped with a triple-reflection ZnSe/Si crystal ATR cell (Smith Detection, USA) and placed in an anaerobic chamber (Coy Laboratories, USA).The atmosphere was adjusted to 99% N 2 and 1% H 2 with no O 2 , CO 2 , or H 2 O. Infrared spectra were recorded with 80 kHz scanning velocity at a spectral resolution of 2 cm -1 .Under these conditions, the time-resolution of data acquisition is in the range of 1 s (one interferometer scan in the forward/backward direction).
To prepare a protein film, 1 µL protein solution (100-200 µM) was pipetted onto the silicon crystal of the ATR cell and enclosed by a custom-made gas titration cell to concentrate the sample under a stream of dry N 2 'carrier' gas (1 L/min), as reported earlier. [7]Once a partially de-hydrated protein film was formed, the gas was sent through a reservoir of 150 mL aqueous buffer (H 2 O or D 2 O), creating an aerosol that was used to re-hydrate the hygroscopic protein film.The reaction with CO 2 was started by adding 1-100% CO 2 to the carrier gas while recording spectra with a time resolution of 20 s.The reaction with CO 2 isotopes was probed by pressuring the closed gas cell with 1 bar 13 CO 2 .

Computational Methods
Alchemical Free Energy Calculations and Molecular Dynamics Simulations.The computational models were based on the crystal structure of the ternary complex of KsCcr (PDB ID 6NA4).This corresponds to a tetrameric structure that adopts a pair of dimers geometry.
Each dimer is comprised by a closed and an open subunit (subunits A/C and B/D), where the closed state is the catalytically competent active site conformation.Each subunit contains the cofactor NADPH and the two closed subunits bear the substrate analogue butyryl-CoA (subunits A and B).The latter was modified to restore the original substrate crotonyl-CoA (i.e., one hydrogen atom was removed from the β-carbon and the α-carbon).
To establish if H365 could act as a base in bicarbonate formation we performed alchemical free energy calculations to determine its pK a shift compared to a histidine residue in a fully solvated peptide with methylated carboxy and amino terminals.The histidine peptide was added to the simulation box containing the protein, cofactor, substrate, and solvent.We applied alchemical free energy calculations to obtain the change in free energy , when a proton is transferred ∆∆º from the H365 residue in the open or closed active site to the fully solvated histidine residue.
The pK a shift of H365 in the open and closed active site results from: Topologies for the protonated and deprotonated states of the two residues were created with the PMX toolset [8,9] in GROMACS 2019.3 [10] software package using the CHARMM36m [11] force field and the CHARMM TIP3P water model [12] in combination with the CHARMM general Force Field CGenFF [13,14] to describe CO 2 substrate and cofactor.Specific parameters for NADPH and the CoA fragment of crotonyl-CoA were taken from Pavelites et al. [15] and Aleksandrov et al. [16] , respectively.
Starting with the protonated H365 either in the open or closed active site, we placed the protein tetramer in a dodecahedric water box using gmx solvate, such that the smallest distance between any atom of the protein and box boundary was larger than 1.5 nm.A methylated peptide NME-His-ACE was placed at least 3.5 nm away from H365 replacing water molecules.Potassium and chloride ions were added to neutralize the system, reaching a physiological concentration of 125 mmol/L.Energy minimization was performed using the steepest descent algorithm for 50000 steps for each system followed by a 1 ns equilibration applying position restraints on the protein backbone, peptide backbone and ligands in the NVT ensemble, and subsequently for 5 ns in the NPT ensemble at 1 atm pressure and a temperature of 298K using an integration time step of 2 fs.The velocity rescaling thermostat [17] and Berendsen [18] pressure coupling were used with coupling coefficients of = 0.1 ps and = 2 ps, respectively.All bond lengths of the   protein, peptide and ligands were constrained using LINCS with an expansion order of 4.
Electrostatic interactions were calculated using Particle-Mesh-Ewald [19] , with a real space cutoff of 12 Å and a Fourier spacing of 1.6 Å.For the van-der-Waals interactions a cutoff of 12 backbone and ligands at 298K with stochastic dynamics.The free energy differences (Tab.S1) were estimated by Bennet Acceptance Ratio (BAR) [20] included in GROMACS as gmx bar.
Table S1.Free energy difference (ΔΔG°) between the protonated H365 of wild-type KsCcr in the open or closed active site and E171 in KsCcr H365N and the model peptide with unprotonated histidine of glutamate in aqueous solution.Error estimation results from BAR free energy estimation.The pK a shift of H365 and E171 was calculated according to the equation above.for the systems was done in GROMACS 2019.3 software package [10] .Protonation states of titratable residues were assigned using Propka 3.1 [21] .The interactions in the molecular system were described with the CHARMM36m [11] force field and the CHARMM TIP3P water model [12] in combination with the CHARMM general Force Field CGenFF [13,14] to describe CO 2 , substrate and cofactor.Parameters for CO 2 were validated with hexane/water partition coefficients and free energy calculations.Specific parameters for NADPH and the CoA fragment of crotonyl-CoA were taken from Pavelites et al. [15] and Aleksandrov et al. [16] ,

ΔΔG°
respectively.The protein system was embedded in a rectangular box of water molecules leaving a 10 Å buffer region between the protein and box edges.Potassium and chloride ions were added to neutralize the system, reaching a physiological concentration of 125 mmol/L.Energy minimization was performed using the steepest descent algorithm for 50000 steps for each system followed by a 1 ns equilibration applying position restraints on the protein backbone, peptide backbone and ligands in the NVT ensemble, and subsequently for 5 ns in the NPT ensemble at 1 atm pressure and a temperature of 298K using an integration time step of 2 fs.
The velocity rescaling thermostat [17] and Berendsen [18] pressure coupling were used with coupling coefficients of = 0.1 ps and = 2 ps, respectively.All bond lengths of the protein,

𝜏
and ligands were constrained using LINCS with an expansion order of 4. Electrostatic interactions were calculated using Particle-Mesh-Ewald [19] , with a real space cutoff of 12 Å and a Fourier spacing of 1.6 Å.For the van-der-Waals interactions a cutoff of 12 Å and a switching function starting at 10 Å was used.
After the initial equilibration, 100 CO 2 molecules were added replacing water molecule.To enhance the sampling the simulations presented ~60 times the experimental concentration [1] of CO 2 (c= 55mM).Energy minimization was performed using the steepest descent algorithm for 50000 steps for each system followed by a 1 ns equilibration applying position restraints on the protein backbone, peptide backbone and ligands in the NVT ensemble, and subsequently for 5 ns in the NPT ensemble at 1 atm pressure and a temperature of 298K using an integration time step of 2 fs.After the pressure is equilibrated, production dynamics in the NVT ensemble were performed keeping position restraints on the protein alpha carbons and NADPH heavy atoms to keep the protein conformation fixed.In the NADPH system the distance between R276 and NADPH's adenine moiety was restraint to its initial value.For the NADPH/C-CoA, distance between residue R99 and crotonyl-CoA's phosphate group, R303 and C-CoA's adenine moiety, the reactive Crotonyl-CoA's C ß with the NADPH's C 4 and H409 with the NADPH's phosphate group for the closed subunits were restraint to its initial value to explore CO 2 distributions in the defined states of the catalytic cycle of the enzyme.Multiples replicas were performed generating a total of 20 µs of simulation time for each system, the coordinates were saved every 100ps resulting in 2x10 5 configurations available for analysis.
CO 2 concentration and binding sites in the active site.To analyze carbon dioxide distribution and binding sites around the protein and in its active sites, we use the groma s [22] analysis  tools: The position of the atoms is expanded by 3-D Gaussians to represent atomic densities (widths of the gaussians equal experimentally determined atomic scattering factors).This procedure is carried around at every frame of the trajectory on a grid with a resolution of 0.1 nm and time averaged over trajectories.
We want to compare the CO

〈𝑆𝐴𝑉〉
To estimate the SAV we expand the heavy atoms positions for the protein, NADPH and C-CoA as atomic densities for all frames at 100 ns time intervals.This density is time averaged and accounts for different conformations of protein's side chains.Once the density map is obtained, we set a threshold isosurface value of 0.1 [a.u], which recovers the overall shape of the residue side chains, cofactors and substrates in the protein complex.Finally, to get the SAV we count all grid points (g i ) with a density lower than the threshold, which corresponds to the volume not occupied for the protein, cofactor and substrate atoms, thus accessible for the solvent and CO 2 .
With the CO 2 concentration in the bulk/solution we focused on the local concentration in the active site.We define the active site as the volume inside a box containing all key residues for the reaction (black box Figure 5A), the nicotinamide ring and the crotonyl-CoA fragment until the pyrophosphate.To obtain the local concentrations at the active sites, the number of CO 2,AS molecules and the inside the boxes must be calculated.The number of carbon dioxide 〈〉  molecules can be calculated from the atomic density inside the active site box.For the active site we followed the same procedure described for the whole protein but considering

〈𝑆𝐴𝑉〉 𝐴𝑆
only the volume of the active site box.
To estimate the number of CO 2 molecules we start from the density map obtained by groma s,  this is a 3D discretization of space with density values for CO 2 at every grid point, then we space average all densities for the active site, summing up all the grid points and dividing them by the value of one CO 2 molecule.This quantity gives us the number of CO 2 molecules inside the active site, which is then divided by the active site's SAV to get concentration values: This concentration is derived as the time average of different configurations of CO 2 molecules inside the active site.This volume is corrected accounting only for the accessible volume of the solvent, which is common to the CO 2 molecules.With both active site concentrations and the solution concentration, we can calculate the enrichment in the active site values from which the binding free energy is calculated as described in the manuscript: To define binding sites, we use the CO 2 density maps obtained from the analysis with groma s  in the active sites.We first identified intervals of 100 ns where the average CO 2 concentration in the active is at least twice the solvent concentration to select configurations where carbon dioxide is bound in the active site.Binding sites were defined using the volumes enclosed by    absorbance . (A) FTIR absorbance spectra of liquid water (H 2 O) and hydrated protein film (KsCcr).The H 2 O spectrum comprises the asymmetric and symmetric OH stretching modes (ν 3 , ν 1 ), the HOH bending mode (ν 2 ), and the so-called 'combination band' (cyan).The envelope is shown as a dashed line.The contributions of liquid water can be found in the spectrum of hydrated KsCcr as well; however, the fit is not shown for clarity.The protein spectrum comprises amide contributions A, B, I, II, and III (green).The envelope is shown as a dashed line.Note that the carbohydrate bands (CH x ) are not included in the fit.(B) Representative 'CO 2 -N 2 ' FTIR difference spectrum of KsCcr, including the fits for unspecific changes, i.e., a minor decrease of liquid water (cyan traces, negative) that is accompanied by a small increase of protein concentration (green traces, positive).Water and protein fits are derived from data in panel A. The accumulation of bicarbonate (magenta) in the film is followed via the sum of peak area including bands at 1298 cm -1 (ν 4 ), 1358 cm -1 and 1618 cm -1 (ν 2 and ν 3 ).The 'special' water band is centered at 3000 cm -1 (νOH, blue).(C) Plot of the peak area for bicarbonate, νOH, bulk water, and protein against time.The atmosphere above the hydrated KsCcr film is exchanged from 100% N 2 to 10% CO 2 (yellow area) in five consecutive steps, emphasizing the reversibility of the process.The peak area for bicarbonate and νOH is plotted 'as observed' (full symbols) and corrected for changes in protein concentration (open symbols).Due to the smaller spectral overlap, changes in humidity were neglected.(D) Plot of the peak area for bicarbonate, νOH, bulk water, and protein against time.The atmosphere above the hydrated KsCcr film is exchanged from 100% N 2 to 1-100% CO 2 (yellow area) in five consecutive steps, emphasizing the connection between band intensity and CO 2 partial pressure.The peak area for bicarbonate and νOH is plotted 'as observed' (full symbols) and corrected for changes in protein concentration (open symbols).Due to the smaller spectral overlap, changes in humidity were neglected.The decrease of the HCO 3 -signal after accumulation under 1-100% is plotted against "Δtime".
The first three data points are fitted linearly to obtain the initial velocity v -1 * of the HCO 3 - dehydration reaction (R 2 > 0.99).(D) Initial velocity plotted against HCO 3 -concentration.The latter was obtained from reference experiments (Figure S7).The data do not agree with linear regression (R 2 < 0.9), which indicates a more complex reaction order.See Figure S10 for the determination of experimental variation.(yellow area) and decreases when the atmosphere is swept back to N 2 .Between 75% and 45% water content, the initial reaction velocities decrease (inset) while the steady-state levels of HCO 3 -at t = 150 s are not significantly affected by the humidity level.Below 45% water the protein solution forms a 'cake' and becomes largely impenetrable to gas [7] , resulting in a large decrease of CO 2 conversion activity.These data demonstrate that the observed kinetics are systematically affected by the humidity.A direct comparison with aqueous reaction assays is not advised.Carefully controlling the humidity level, however, different FTIR experiments can be compared both qualitatively and quantitative.hydration activity at alkaline pH, which is expected for the catalyst-free system. [25]Based on these data and similar experiments with KsCcr at pH 5-9, the inset plots the carbonate formation after 60 s.Note that CO 2 hydration with KsCcr is basically pH-independent in this pH range.
that H365 is monoprotonated (neutral) in absence of substrate for the open or closed states of the active site under the experimental condition of pH=8.Both conformations showed a negative pK a shift, related to a more stable configuration for the monoprotonated state in the active site compared to the fully solvated peptide in solution.For the closed state with C-CoA, a positive pK a shift indicates H365 being doubly protonated (charged) compared to the fully solvated peptide in solution.Based on the most probable protonation state of H365 in the closed or open active site we build new systems to analyze the local concentration of CO 2 in the active sites.We started from the reported crystal structure for KsCcr (PDB ID: 6NA4) as previously mentioned and we generated two different configurations.The NADPH only complex with all four H365 monoprotonated and the NADPH/C-CoA ternary complex, which has one Crotonyl-CoA molecule and the H365 doubly protonated in each of the closed active sites.Hydrogen addition and topology generation

isosurfaces of CO 2
concentration selecting an isosurface value 20 times larger than the bulk concentration (see Figure S10 below).The centers of the volumes define the center of the binding sites shown in Figure 6 in the main script.

Figure S2 .
Figure S2.(A) Directly measuring the pH of the sample film in the ATR FTIR experiments is impossible.To probe the stability of pH in the presence of 10% CO 2 the following representative experiment was attempted: We purged 40 mL A. bidest or buffer solution (200 25 mM Tris/HCl pH 8) with 100 mL/min N 2 for 15 minutes at 22°C and a stirring speed of 300 rpm).This removed atmospheric CO 2 from the solution.Then, the experiment was started by measuring the pH with a tabletop pH electrode (Thermo Scientific, Orion Star A111).(B) Plot of pH value as a function of time.At t 0 = 0 min, the gas was switched from 100% N 2 to 90% N 2 and 10% CO 2 .In unbuffered solution (A. bidest, black traces), the bulk pH dropped by approximately 1.7 units over the course of 25 minutes.No such acidification was observed with buffer solution (blue traces).These data show that 20 mM buffer is sufficient to keep the pH value of the protein solution constant in the presence of 10% CO 2 for at least 25 min.(C) In ATR FTIR spectroscopy, much smaller sample volumes are probed, typically 1-3 µl.Therefore, any direct kinetic comparison should be avoided.The figure shows a schematic depiction of the silicon crystal with a drop of liquid (yellow) or a protein film (green).(D) ATR FTIR difference spectra of A. bidest in the presence of 10% CO 2 .No HCO 3 -bands are observed although a strong signal at 2343 cm -1 (CO 2 in aqueous solution) is visible.This is due to the acidification in pure water, as demonstrated in panel (B).(D) ATR FTIR difference spectra of 25 mM Tris/HCl (pH 8) in the presence of 10% CO 2 .While the signal at 2343 cm -1 indicates a similar concentration of CO 2 , bands at 1620, 1360, and ~1300 cm -1 evidence residual HCO 3 - formation.
Figure S4.(A) Same data as in Figure S3D.The increase of the HCO 3 -signal under 1-100% is plotted against "Δtime" for visual comparison.The first three data points are fitted linearly to obtain the initial velocity v 1 * of the CO 2 hydration reaction.(B) Initial velocity plotted against CO 2 concentration.The linear dependence suggests (pseudo-) first order kinetics.See Figure S8 for the determination of experimental variation.(C) Same data as in Figure S3D.

Figure
Figure S5.(A) FTIR spectra of liquid H 2 O (dashed line), the empty ATR cell (magenta line) and a KsCcr protein films at five different levels of humidity.The humidity was adjusted by manually mixing water-free N 2 carrier gas ('dry') and N 2 aerosol ('wet').All spectra show steady-state conditions after 5-10 min of equilibration.Inset: The 3350 cm -1 absorbance value for liquid H 2 O (100%) and the empty cell (0%) define the calibration curve of water content that allows calculating the humidity level of the five KsCcr protein films (see legend in panel (B)).Note that the decrease in humidity is associated with an increase in protein concentration (* amide I, ** amide II).(B) CO 2 conversion kinetics of the five KsCcr protein films.The plot shows the HCO 3 -peak area as a function of time that increases in the presence of 10% CO 2

Figure S6 .
Figure S6.(A) Exemplary data set of CO 2 hydration via KsCcr as probed by UV/vis spectroscopy.The spectrum shows the decrease of the baseline-corrected 618 nm band of bromothylmol blue band at increasingly acidic pH.Here, the red spectra (t < 0 s) indicate an initial drop of intensity when the solution was diluted with carbonated buffer ad 1 ml.The spectral transition (blue to yellow) depicts the CO 2 hydration reaction between 0-120 s as catalyzed by KsCcr.(B) Time traces for the decrease of the 618 nm band of bromothylmol blue band for EcCA (black), KsCcr (red), and buffer (blue).The inset shows how raw data was fitted with a single Lorentz function and a polynomial background.The arrow at t = 0 s marks the injection of carbonated buffer, leading to a drop in intensity due to dilution.CO 2 hydration contributed to the spectra afterwards.These data highlight (i) the superior activity of EcCA and (ii) show that KsCcr catalyzes CO 2 hydration clearly faster than the buffer.See Experimental Methods for experimental details.

3  / cm - 1 FigureFigure
Figure S7.(A) Five films of KsCcr under N 2 at similar humidity.Inset: subtraction of the weighted averaged from each spectrum suggest less than 10% deviation in humidity.(B) The increase and decrease of bicarbonate over time from five KsCcr protein films in the presence and absence of 10% CO 2 is shown in a stacked plot.Linear fits of the first three data points for the N 2 →CO 2 and CO 2 →N 2 step yield apparent reaction velocities v 1 * and v -1 *, respectively, that describe the velocity of the system to converge toward steady-state conditions.(C) According to these data, the rate constants are v 1 * = 0.046 s -1 (experimental deviation 10.9%) and v -1 * = 0.019 s -1 (experimental deviation 9.1%).These values are used to compare data in Figure 2D of the main script.(D) The bicarbonate peak area at the end of each CO 2 step (dashed line in panel B) is shown.These data yield an experimental deviation of 7.5%, which is used to compare data in Figure 4A of the main script.

Figure S10 .
Figure S10.(A) Comparison of CA-like activity in KsCcr protein films under conditions of stable humidity in the presence of various substrates (B-CoA, C-CoA) and/or cofactors (NADPH, NADP + ).60 s after addition of 10% CO 2 to the atmosphere (dashed line), the data facilitate a comparison among the different compositions (see Figure 4 in the main script).(B) Comparison of CA-like activity in KsCcr protein films under conditions of stable humidity for wild-type KsCcr and five amino acid variants.60 s after addition of 10% CO 2 to the atmosphere (dashed line), the data facilitate a comparison among the different variants (see Figure 4 in the main script).The arrow highlights the decrease of CO 2 hydration velocity from KsCcr H365N (cyan) to E171A (blue) to H365N/E171A (magenta).
Figure S11.(A) The surface in red show regions in the active site where the CO 2 concentration is 2, 5, 10, 15, or 20 times larger than in the bulk obtained as time average from groma s, (i)- (v) respectively.The position of the binding sites in dark red spheres shown in the manuscript match the center of the 20 times bulk CO 2 concentration lobules (vi).(B) The nucleophilic attack of CO 2 in the active site via bridging water species µW may involve additional water moieties, according to MD simulations.
2 concentration in the active site in the closed and open state and the absence or presence of substrate compared to the bulk concentration.We start defining the bulk concentration as the solution concentration (C solution ).This concentration corresponds to the number of carbon dioxide molecules inside the simulation box divided by the Solvent Accessible Volume (SAV).The SAV is a magnitude that accounts for the points in space that carbon dioxide molecules can access and represents the volume of the simulation box excluding the volume of the Protein, NADPH and C-CoA atomic densities.SAV is modulated by the movement of all atoms in the molecular dynamics simulation.To account for this effect, we