Substrate Promiscuity of Thermoplasma acidophilum Malic Enzyme for CO2 Fixation Reaction

CO2 fixation technology has gained attention as a method to effectively utilize the abundant CO2 in the atmosphere by converting it into useful chemicals. However, since CO2 is a highly stable molecule, many of the currently developed methods for chemical CO2 fixation require harsh conditions and reactive reagents. The establishment of efficient and sustainable processes is eagerly awaited. In this study, we investigated a biocatalytic process and achieved a carboxylation reaction under mild conditions (37 °C, 0.1 MPa CO2) using a biocatalyst, Thermoplasma acidophilum NADP+-malic enzyme (TaME), and gaseous CO2 by coupling enzymatic coenzyme regeneration. We also demonstrated for the first time that the carboxylation reaction by ME proceeds not only with pyruvate, a natural substrate, but also with 2-ketoglutarate.

I n recent years, due to the urgent need for decarbonization, Carbon Capture and Utilization (CCU) technology has garnered attention for its effective utilization of CO 2 , one of the primary contributors to global warming.−5 The attractiveness of utilizing CO 2 as a reactant stems not only from its ability to promote CO 2 utilization and produce value-added chemicals but also from its safety and ease of handling.For example, some carboxylic acids, such as acrylic acid and methacrylic acid, which are important compounds of pharmaceuticals and chemical synthesis, can be produced by carboxylation of olefins. 6owever, given the unreactive nature of CO 2 , many of the currently developed methods for carboxylation reactions with CO 2 as a reactant require high temperatures, pressures, and unstable reagents, posing challenges in terms of energy efficiency and sustainability.−13 Hence, there have been reports of some biocatalytic reactions involving CO 2 fixation. 14For instance, 2,3-dihydroxybenzoic acid decarboxylase has been noted for its role in catalyzing the carboxylation reaction of catechol with gaseous CO 2 , coupling with simultaneous amine-mediated conversion of CO 2 to bicarbonate. 15−18 For example, T. acidophilum isocitrate dehydrogenase (TaIDH) was employed in the carboxylation of 2-ketoglutarate to isocitrate under 10 MPa CO 2 . 18Additionally, malic enzyme (ME), known as a catalyst for the oxidative decarboxylation reaction of L-malate, producing pyruvate along with the reduction of NAD(P) + , also catalyzes the thermodynamically very unfavorable reverse reaction, the reductive carboxylation.This thermodynamic constraint was overcome by combining electrochemical, photochemical, and enzymatic reaction systems.−21 However, only a handful of studies have reported ME-catalyzed carboxylation using only gaseous CO 2 as CO 2 source; 22 most biocatalytic carboxylations use bicarbonate such as NaHCO 3 and KHCO 3 .Furthermore, researchers have explored few substrates other than malate 1b and pyruvate 1a, which are natural substrates for decarboxylation and carboxylation reactions, respectively, to determine if they serve as substrates for ME.In this study, our focus was on the NADP + -malic enzyme from Thermoplasma acidophilum (TaME) (Scheme 1).
We chose this enzyme hoping for its robustness and ease of handling, as other enzymes from T. acidophilum, such as isocitrate dehydrogenase (TaIDH) and glucose dehydrogenase (TaGDH), were reported to have high thermal and CO 2pressure stabilities. 18Additionally, TaGDH was employed to address the thermodynamic constraint for carboxylation.The objective of our study was to develop a TaME-catalyzed carboxylation reaction using only gaseous CO 2 as a CO 2 source and to investigate the substrate specificity of TaME for both carboxylation and decarboxylation.
Initially, the carboxylation of pyruvate 1a was conducted using TaME and NADPH under 0.1 MPa of CO 2 .While previous studies often relied on carbonates such as NaHCO 3 or KHCO 3 as CO 2 sources (Table S2), our study exclusively used gaseous CO 2 .The resulting yield of the reaction by TaME and a molar equivalent of NADPH to the substrate was 3.8%, as shown in Figure 1a.To address this low yield, we explored a cofactor regeneration system, which involved the addition of TaGDH and D-glucose.This addition led to a substantial improvement, with the yield increasing to 67%, an 18-fold increase (Figure 1a).This result indicated that the TaGDH reaction enhanced the thermodynamic feasibility of the carboxylation reaction, aside from cofactor regeneration, as expected from the case using TaIDH and TaGDH. 18hen, we investigated the reaction conditions for the reductive carboxylation of 1a by TaME, gaseous CO 2 , and TaGDH.The effects of pH, divalent metal ion, substrate concentration, and CO 2 pressure on the carboxylation reaction were investigated.As shown in Figure S2a, the optimal yield was achieved at an initial pH of 7.5−8.5, with a decline observed at pH 7.0.Notably, this pH range differs from the reported optimal pH for the reductive carboxylation reaction by ME from Pseudomonas diminuta IFO 13182 (pH 6.0) 20 and Thermococcus kodakarensis (pH 6.5) 23 as well as the reported TaGDH activity for NADPH production (pH 6.5). 24The differences in optimal pH values may be attributed to the likelihood that the actual pH during the reaction in our experiment was lower than the initial pH due to the formation of H 2 CO 3 and subsequent pH reduction.Considering this factor, the optimum pH for carboxylation with TaME was set at 7.5.
Next, the effect of divalent metal ions was examined using chloride salts of Mg 2+ , Mn 2+ , Co 2+ , Ca 2+ , and Ni 2+ , as these ions (mainly Mn 2+ and Mg 2+ ) are crucial for ME catalysis.−27 Consequently, we found that the highest yields for carboxylation with TaME were achieved with Mg 2+ , Mn 2+ , and Co 2+ (Figure S2b).Therefore, we selected Mg 2+ , which has the lowest environmental impact among these metals, as the optimal divalent metal ion for carboxylation.
Then, the effect of substrate concentration (10, 20, 30, and 40 mM) was examined.The product concentration increases with increasing substrate concentration but not the yield, as shown in Figure S2c.Therefore, considering both the final product concentration and yield, we set 20 mM as the optimal concentration.These conditions (pH 7.5, Mg 2+ , and 20 mM substrate) were used thereafter in TaME carboxylation reactions.
Furthermore, the effect of CO 2 pressure on the reductive carboxylation of 1a by TaME was investigated, and no significant difference in yield was observed between 0.1 to 5 MPa gaseous CO 2 , as shown in Figure 1b.The solubility of CO 2 at 303 K, which is close to the experimental temperature (310 K), was reported to be 28.6 mM (0.1 MPa), 280 mM (1 MPa), and 1080 mM (5 MPa). 28On the other hand, it was reported that K m(CO2) was 4.2 mM for the wheat germ preparation and 1.1 mM for the C4 plant enzyme. 29,30herefore, the insignificant difference in yield between 0.1 and 5 MPa was attributed to the fact that even at 0.1 MPa CO 2 , the concentration of CO 2 was assumed to be sufficiently higher than the K m(CO2) values of TaME.Lastly, the time course of the reductive carboxylation of 1a by TaME under 0.1 MPa CO 2 was examined.Figure 1c shows that the yield increased steadily over time and reached a plateau after 5 h, with a remarkable yield of 72%, which was the highest of any carboxylation reaction with ME developed in the past (Table S2).
Next, we investigated substrate specificity of TaME for decarboxylation and carboxylation.First, the decarboxylation activity was determined using malate 1b, isocitrate 2b, citrate 3b, 3-isopropylmalate 4b, tartrate 5b, tartornate 6b, and lactate 7b each at a concentration of 1 mM by monitoring the increase in NADPH concentration, setting the activity toward 1b, the natural substrate of TaME, as 100% (Table 1).Surprisingly, TaME exhibited higher relative activity (190%) toward 2b compared to 1b.To explain these findings, Michaelis−Menten kinetic parameters for the decarboxylation of 1b and 2b were determined (Table 2).The results revealed that k cat for 2b (1.3 s −1 ) was about twice that of 1b (0.69 s −1 ).Additionally, the K m for 2b (23.7 μM) was significantly lower than the K m for 1a (879 μM).Since the substrate concentration used to determine the activity was 1 both higher k cat and lower K m of 2b likely contributed to its higher relative activity of 2b than that of 1b.It was surprising that the k cat /K m(substrate) values differed by 2 orders of magnitude, with superiority in the unnatural substrate 2b.
To further investigate the differences in the relative activity and Michaelis−Menten kinetic parameters of TaME between 1b and 2b, docking simulations were performed.The X-ray structure of ME from Bdellovibrio bacteriovorous (Bd1833/ MaeB), containing NADP + and Mg 2+ (6ZN7.pdb), 31was used first to dock 1b and 2b using Autodock Vina due to the high amino-acid sequence homology with TaME (Figure S3).Then, the resulting structures were aligned with the structure of TaME predicted by ColabFold using PyMOL (Figure S4a, b), and the interactions between the substrates (1b and 2b) and the TaME were investigated (Figure 2).Consequently, 2b formed three hydrogen bonds (defined within 3.5 Å) with amino acid residues Asn304 and Thr46* (the asterisk indicates residues from another subunit of the dimer) of TaME (Figure 2b), whereas 1b formed only one hydrogen bond with Asn304 (Figure 2a).Therefore, it is assumed that the difference in the number of hydrogen bonds contributed to the difference in k cat /K m .−34 Therefore, Asn304 and Thr46* of TaME may also be important residues in determining the substrate specificity.However, there are critical differences between the 3D arrangements of ME and IDH, which are enantiomeric.The active site of ME is the mirror image of that of IDH, and this is consistent with the opposite stereochemistry at the C α alcohol of L-malate, the natural substrate of ME, and D-isocitrate, the natural substrate of IDH. 35ncouraged by the higher activity of TaME toward unnatural substrate 2b, finally, reductive carboxylation of 2ketoglutarate 2a was investigated to synthesize 2b with high added value. 36This involved the reaction of 2a in the presence of TaME, 0.1 MPa CO 2 , Mg 2+ , NADP + , TaGDH, and glucose in HEPES buffer for 3 h, resulting in the successful synthesis of isocitrate 2b with a yield of 27%.This low yield could be attributed to the accelerated rate of the decarboxylation reaction of 2b, as can be seen in the notably larger k cat /K m of 2b in comparison to 1b.Future study to examine reaction conditions and time is necessary to improve the yield.The absolute configuration of the product 2b, which will be determined in the future study, may be L-(S), the opposite enantiomer of what is produced by the carboxylation using TaIDH. 18This is because TaME, belonging to the "malic enzymes" family, accepts (S)-hydroxy acids as substrates, while IDH, belonging to the "β-decarboxylating dehydrogenases" family, accepts (R)-hydroxy acids as substrates. 35t is noteworthy that TaME was found to exhibit the substrate promiscuity for both decarboxylation and carboxylation in this study, as it belongs to the "malic enzymes" family. 35,37While enzymes in the "β-decarboxylating dehydrogenases" family (IDH, isopropyl malate dehydrogenase, homoisocitrate dehydrogenase, D-malate dehydrogenase, and tartrate dehydrogenase) generally have narrow substrate specificities, 33,38 some enzymes in this family have been reported to catalyze secondary reactions. 17,34However, very few investigations have been performed on the substrate specificity of decarboxylation and carboxylation for the "malic enzymes" family.Thus, this study, which demonstrates substrate promiscuity not only for decarboxylation but also for carboxylation, opens up the possibilities that other enzymes in the "malic enzymes" family, complementary to those in the Table 1.Substrate Specificity for TaME-Catalyzed Decarboxylation a a The activity was determined by the method in Supporting Information section 1.3.The activity toward malate 1b was set to be 100%.Kinetics assays were done under standard assays conditions shown in the Supporting Information section 1.3.1 using 0.05−1.0mM of 1b or 0.03−0.60 mM of 2b.The TaME activity toward 2b reached a plateau after 0.6 mM up to 2.0 mM."β-decarboxylating dehydrogenases" family, would also possess catalytic promiscuity and could be utilized for wide applications in carboxylation.
In conclusion, we found that TaME was a significant biocatalyst for carboxylation under mild conditions, contributing to Carbon Capture and Utilization (CCU).We demonstrated that TaME could catalyze the carboxylation of not only pyruvate 1a, a natural substrate, but also unnatural substrate 2-ketoglutarate 2b.Since the hydrophilic amino acid residues such as Asn304 and Thr46* of TaME were suggested to be important in determining substrate specificity, it is expected that introducing site-directed mutations in these residues will further expand the substrate specificity in the future studies.Additionally, TaME may potentially catalyze the selective synthesis of (S)-hydroxy acids, which could be complementary to our previously developed carboxylation reaction with TaIDH. 18ASSOCIATED CONTENT * sı Supporting Information The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/jacsau.4c00290.Additional experimental details, materials and methods, and additional results, including the SDS-PAGE of TaME (Figure S1), purification fold table of TaME (Table S1), effect of initial pH, divalent metal ion, and substrate concentration on TaME-catalyzed carboxylation of 1a (Figure S2), amino acid sequence alignment of TaME and Bd1833/MaeB (Figure S3), overlapping of TaME ColabFold structure with docking simulation results of Bd1833/MaeB crystal structure with malate 1b and isocitrate 2b (Figure S4), and an example of reductive carboxylation of 1a by ME (Table S2