Crystal Structure of the Ergothioneine Sulfoxide Synthase from Candidatus Chloracidobacterium thermophilum and Structure-Guided Engineering To Modulate Its Substrate Selectivity

Abstract

Ergothioneine is a thiohistidine derivative with potential benefits on many aging-related diseases. The central step of aerobic ergothioneine biosynthesis is the oxidative C–S bond formation reaction catalyzed by mononuclear nonheme iron sulfoxide synthases (EgtB and Egt1). Thus far, only the Mycobacterium thermoresistibile EgtB (EgtB_Mth) crystal structure is available, while the structural information for the more industrially attractive Egt1 enzyme is not. Herein, we reported the crystal structure of the ergothioneine sulfoxide synthase (EgtB_Cth) from Candidatus Chloracidobacterium thermophilum. EgtB_Cth has both EgtB- and Egt1-type of activities. Guided by the structural information, we conducted Rosetta Enzyme Design calculations, and we biochemically demonstrated that EgtB_Cth can be engineered more toward Egt1-type of activity. This study provides information regarding the factors governing the substrate selectivity in Egt1- and EgtB-catalysis and lays the groundwork for future sulfoxide synthase engineering toward the development of an effective ergothioneine process through a synthetic biology approach.

Graphical Abstract

Ergothioneine is a thiohistidine derivative. Through an ergothioneine-specific transporter, human and animals absorb ergothioneine from foods, and it accumulates in concentrations as high as 2 mM in erythrocytes, livers, kidneys, lenses, and corneas of eyes.^1–4 In addition, a combination of the two most abundant natural thiols ($E_0^{\prime }=-0.06$ V^5–7 for ergothioneine and $E_0^{\prime }=-0.24$ V⁸ for glutathione) protect cells against reactive oxygen species (ROS) and reactive nitrogen species (RNS) under varying conditions.⁹ Ergothioneine has also been suggested to serve as a protecting agent against several diseases, including cardiovascular disorders,¹⁰ rheumatoid arthritis,^11,12 Crohn’s disease,^13,14 neurodegenerative diseases,^15–18 and diabetes.¹⁹ Because of its potential health benefit, ergothioneine biosynthetic studies have received considerable interest in recent years.

Two aerobic biosynthetic pathways of ergothioneine have been reported (Scheme 1A): the Mycobacterium smegmatis pathway (EgtA-EgtE) and the Neurospora crassa pathway (Egt1/Egt2).^20–30 In these two pathways, the crucial steps are the oxidative C–S bond formation mediated by nonheme iron sulfoxide synthases (EgtB_Msm /Egt1) and the reductive C–S cleavage reaction catalyzed by PLP-dependent lyases (EgtE/Egt2), which differ from other reported sulfur-transfer strategies.^31–34 Recently, the ovothiol biosynthetic pathway has also been reconstituted in vitro (Scheme 1B).^27,35,36 These three sulfoxide synthases (Egt1, EgtB_Msm, and OvoA) differ in at least two aspects: their substrate selectivity and their regioselectivity. First, Egt1 and EgtB_Msm selectively use hercynine as the substrate, while they differ in the sulfur sources [γ-glutamyl-cysteine (γ-Glu-Cys) for EgtB_Msm and L-Cys for Egt1]. On the other hand, OvoA selectively uses L-His and L-Cys as the substrates. Although OvoA can also use of hercynine and γ-Glu-Cys as substrates, sulfoxide 4 is not the major product under this condition.³⁷ Second, for EgtB- and Egt1-catalysis, sulfur was inserted into the ε-position of imidazole side-chain (Scheme 1A), while in OvoA-catalysis, the C–S bond is formed at the imidazole δ-carbon (Scheme 1B).

Scheme 1. Ergothioneine, Ovothiol Biosynthesis, and Distinct Properties of Sulfoxide Synthase EgtB_Cth^a.

^a(A) Two aerobic ergothioneine biosynthetic pathways. (B) Ovothiol biosynthetic pathway. (C) Sequence similarity network analysis of ergothioneine sulfoxide synthases and the link between the sulfoxide synthase EgtB_Cth (in red), Egt1 (in blue), and EgtB (in green). (D) EgtB_Cth exhibits Egt1- and EgtB-type activities.

Given ergothioneine’s many potential health benefits, there is an increasing demand for developing efficient industrial-scale ergothioneine production methods. The pathway involving Egt1-catalysis is preferred because the use of l-Cys as the sulfur source alleviates the competition between ergothioneine and glutathione biosynthesis, in which γ-Glu-Cys is a key biosynthetic intermediate. Among these sulfoxide synthases in ergothioneine and ovothiol biosynthesis, EgtB_Mth from M. thermoresistibile is the only reported structure.²² In this study, a thermophilic Candidatus Chloracidobacterium thermophilum sulfoxide synthase, Cabther_A1318 (EgtB_Cth, Scheme 1C), was identified from a cluster bridging between that of EgtB and Egt1 through sequence similarity network analysis. The separation of cluster of EgtB_Cth from that of EgtB and Egt1 is suggestive of different biochemical properties of EgtB_Cth. Indeed, biochemical characterization shows that EgtB_Cth exhibits both EgtB- and Egt1-type of activities (Scheme 1D). Encouraged by this discovery, based on the EgtB_Cth and EgtB_Mth structural information, we attempted EgtB_Cth activity engineering using Rosetta Enzyme Design calculation.^38,39 We then selected three mutants for biochemical characterization and demonstrated that we can tune the activity of EgtB_Cth toward Egt1-catalysis.

Egt1 from N. crassa exhibited different substrate preference from both EgtB and OvoA (Scheme 1A,B).^28,40 This discovery immediately raised questions as to which factors governed these selectivities (substrate selectivity and product regioselectivity, Scheme 1). Among these enzymes, only EgtB_Mth structure from a thermophilic M. thermoresistibile is available. Our attempts to crystallize N. crassa Egt1 failed. To search for a proper sulfoxide synthase to study factors governing substrate selectivities, we retrieved 21 475 protein sequences containing either DinB_2 domain (Pfam ID: PF12867) or FGE sulfatase domain (Pfam ID: PF03781) from the Pfam protein family database.⁴¹ 3000 sequences out of 21 476 sequences were randomly selected for protein similarity network analysis (Scheme 1C) at an E-value cut off of 10⁻⁵⁰. The protein similarity network was visualized by Cytoscape, which showed some sequences located between EgtB and Egt1 nodes (Scheme 1C and Table S1), implying the possibility of finding different biochemical properties. Among these sequences, Cabther_A1318 (EgtB_Cth) from thermophilic C. thermophilum was chosen for further studies.

The EgtB_Cth gene was overexpressed and purified following a reported procedure.^28,40 The purified EgtB_Cth protein contained ~0.92 iron per monomer (Figure S1). The substrates, L-Cys and γ-Glu-Cys, were characterized before use in reactions (Figure S1). ¹H NMR analysis showed that EgtB_Cth exhibited both EgtB- and Egt1-type activities and accepted both L-Cys and γ-Glu-Cys as the sulfur donor (Scheme 1D, Figures S2 and S3). In addition, the chemical shift (~ 7.11 ppm, the imidazole hydrogen signal of compound 3, Figure 1) is consistent with our previous results on EgtB_Msm and Egt1 studies, suggesting that C–S bond is formed at the imidazole ε-position, instead of the OvoA-type in which the C–S bond is formed at the δ-position.^28,40 EgtB_Cth was characterized kinetically using the oxygen consumption assays as previously reported in Egt1 and OvoA studies (Table 1 and Figures S4 and S5).^28,40 Using hercynine and γ-Glu-Cys as substrates, the k_cat of the reaction was ~18 min⁻¹ with K_M for hercynine of 41.4 ± 3.5 μM and for γ-Glu-Cys of 5.9 ± 0.9 mM. The catalytic efficiency of EgtB_Cth increased by ~42-fold when hercynine and L-Cys were used as the substrates exhibiting a k_cat of ~26 min⁻¹. In this reaction, EgtB_Cth has a K_M of 87.7 ± 7.6 μM and 205 ± 18 μM, for hercynine and L-Cys, respectively (Table 1). The biochemical studies show that EgtB_Cth resembles the activity of EgtB and Egt1 in term of their C–S bond regioselectivity but differ in their substrate specificity as EgtB_Cth shows both Egt1- and EgtB-type of activities (Scheme 1D). The distinct substrate specificity might account for the cluster of EgtB_Cth locating between that of EgtB and Egt1 nodes in the sequence similarity analysis (Scheme 1C).

Figure 1.

¹H NMR analysis of EgtB_Cth reactions. (A) EgtB_Cth reaction under EgtB reaction conditions. The two hydrogens of compound 2 imidazole side chain are labeled as 2, and the hydrogen of compound 3 imidazole hydrogen is labeled as 3. (B) EgtB_Cth reaction under Egt1-conditions. Compound 4 imidazole hydrogen is labeled 4. (C) Ratios of sulfoxide synthase and cysteine dioxygenase activity of EgtB_Cth under EgtB- and Egt1-type of reaction conditions. These two competing pathways are present in all three sulfoxide synthases (Egt1, EgtB, and OvoA). Notably, a significant level of cysteine dioxygenase activity was observed in EgtB_Cth-catalysis.

Table 1.

Kinetic Parameter Characterizations of EgtBCth Wild-Type and Variants

enzyme	substrates	kcat (min−1)	KM, hercynine(μM)	kcat/ KM, hercynine(min−1μM−1)	KM, γ‑Glu‑Cys/Cys (μM)	kcat/ KM, γGluCys/Cys(min−1μM−1)	% of coupling product
wild-type	2 + γ-Glu-Cys	17.5 ± 0.4	41.4 ± 3.5	0.42 ± 0.04	(5.9 ± 0.9)E3	(3.0 ± 0.5)E-3	75%
	2 + L-Cys	26.6 ± 0.7	87.7 ± 7.6	0.30 ± 0.03	205 ± 18	0.13 ± 0.01	72%
A420Y	2 + γ-Glu-Cys	17.4 ± 0.3	13.2 ± 1.3	1.3 ± 0.1	(3.1 ± 0.3)E3	(5.6 ± 0.6)E-3	61%
	2 + L-Cys	27.6 ± 0.5	39.2 ± 1.6	0.70 ± 0.03	28.1 ± 1.8	0.98 ± 0.06	76%
D52L	2 + γ-Glu-Cys	12.9 ± 0.2	19.6 ± 1.4	0.66 ± 0.05	(3.9 ± 0.3)E3	(3.3 ± 0.2)E-3	50%
	2 + L-Cys	27.6 ± 0.5	13.2 ± 1.3	2.1 ± 0.2	47.3 ± 2.4	0.58 ± 0.03	69%
D52L/A420Y	2 + γ-Glu-Cys	18.6 ± 0.4	15.1 ± 1.3	1.2 ± 0.1	(4.2 ± 0.5)E3	(4.4 ± 0.5)E-3	52%
	2 + L-Cys	32.7 ± 0.3	23.3 ± 1.2	1.4 ± 0.07	48.7 ± 3.5	0.67 ± 0.04	75%

In previous studies of EgtB, Egt1, and OvoA, these enzymes also show some cysteine dioxygenase activity.^{25,28,37,42,43} Thus, we analyzed EgtB_Cth reaction for cysteine dioxygenase activity under both Egt1- and EgtB-type of catalytic conditions. We have characterized all the products, and their ¹H NMR signals have been reported.^28,37,42,43 The ratio between sulfoxide and cysteine oxidation activity was analyzed by ¹H NMR (Figures S6 and S7).^25,28 Our analysis indicated that when hercynine and γ-Glu-Cys were the substrates, sulfoxide 3 accounted for ~75% of the total products, and if hercynine and L-Cys were the substrates, sulfoxide 4 accounted for ~72% of the products (Table 1 and Figures S6 and S7). The remaining 25% and 28% were γ-Glu-Cys sulfinic acid 9 or cysteine sulfinic acid 10, respectively (Figures S6 and S7).

With this interesting biochemical information, we also initiated its crystallization studies of EgtB_Cth. Apo-EgtB_Cth crystallized in a cubic form with a space group of P2₁ and diffracted to 2.5 Å in a synchrotron beam source (PDB ID: 6O6M). The single-wavelength anomalous dispersion (SAD) technique was used for de novo phase determination and structure solution using selenomethionine incorporated EgtB_Cth crystals (Table S2). The crystallographic asymmetric unit is composed of four protein molecules (Figure 2A). The overall structure of EgtB_Cth is a tetramer, consistent with its oligomerization state in solution as detected in gel filtration profile. For each monomer, electron densities were resolved for residues 17 to 433, except an interdomain loop of 10 residues (184 to 193) which is missing because of its high flexibility (Figure 2A). Each monomer is composed of an N-terminal helical domain (residue 17 to 183) and a C-terminal domain that consists of an α–ββ–α fold (residue 194 to 433) (Figure 2A). A Dali⁴⁴ search revealed that the N-terminal domain most closely resembles the damage-inducible protein Din-B (PDB ID: 5WK0⁴⁵) and the C-terminal domain is structurally similar to the formylglycine generating enzyme (PDB ID: 5NXL).⁴⁶ The C-terminal domain shares a high structural similarity with the catalytic domain of EgtB_Mth.²²

Figure 2.

Structures of EgtB_Cth and EgtB_Cth·hercynine binary complex. (A) Overall structure of EgtB_Cth in the tetrameric configuration with each monomer labeled. In Monomer A, the N-terminal domain (residue 17 to 183) is shown in blue and the C-terminal domain (residue 194 to 433) is shown in pink. The iron cofactor present at the active site of each monomer is shown as a brown sphere. (B) The 2mF_o-DF_c map of iron coordination site of EgtB_Cth contoured at 1.5σ (blue mesh); the metal ion is shown as a brown sphere and the coordinating residues are represented in sticks. Ordered water molecules coordinating the iron are shown as red spheres. (C) The mF_o-DF_c omit map of the active site of EgtB_Cth cocrystallized with hercynine contoured at 3σ (green mesh). The chemical structure of the substrate hercynine (shown as yellow sticks) was modeled into the positive density. (D) The interaction network between hercynine and EgtB_Cth active-site residues. Residues interacting with the substrate hercynine are shown in sticks with the potential interactions shown in black dash lines. (E) The previously reported structure of EgtB_Mth· dimethyl histidine·γ-Glu-Cys complex (PDB ID 4X8D) superimposed on the EgtB_Cth·herycine complex (shown in green). The side chains of the EgtB_Mth residues interacting with the γ-Glu-Cys are shown in sticks (white) and numbered with a superscript (′), the corresponding residues in the EgtB_Cth structure are shown as blue sticks. (F) The putative γ-Glu-Cys binding mode to EgtB_Cth (shown as yellow sticks). The potential interactions between γ-Glu-Cys and EgtB_Cth active-site residues were depicted as black dashed lines, and the side chains of the interacting residues are shown as blue sticks

The active site of EgtB_Cth is located at the interface between the N- and C-terminal domains for each monomer where a mononuclear nonheme iron is coordinated by His62, His153, His157 and three water molecules in an octahedral arrangement (Figure 2B). Upon soaking or cocrystallizing EgtB_Cth crystals with hercynine, close to the metal ion, strong positive density was observed replacing one of the coordinating water molecules (Figure 2C). Comparison between structures of EgtB_Cth and EgtB_Cth·hercynine binary complex, no significant conformation change was observed (Figure S8). In the EgtB_Cth· hercynine binary complex (PDB ID: 6O6L), hercynine coordinates to the iron center through its imidazole ε-nitrogen (Figure 2D). The hercynine imidazole δ-nitrogen forms hydrogen bonds with Gln156 (3.4 Å) and the hydroxyl group of Tyr93 (3.6 Å). In addition to the dipolar contact with Asn414, the trimethylated amino group of hercynine has cation-π interactions with the side chains of Phe415 (4.7 Å) and Phe416 (5.1 Å).

Despite the lack of overall sequence similarity between EgtB_Cth and EgtB_Mth, a comparison of active sites of the two homologous enzymes reveals a high degree of conservation of residues, suggesting a similar substrate recognition network and catalytic mechanism. The residues used to coordinate the iron center are faithfully conserved in EgtB_Cth and EgtB_Mth (Figure 2E). Additionally, the residues involved in hercynine binding between these two structures are conserved (Gln137, Asn414, and Trp415 in EgtB_Mth vs Gln156, Asn414, and Phe415 in EgtB_Cth, Figure 2E and S9). The binding pocket for the cosubstrate γ-Glu-Cys/L-Cys is located close to the iron center in EgtB_Mth. A similar pocket also exists in EgtB_Cth. As EgtB_Cth·hercynine·Cys or EgtB_Cth·hercynine·γ-Glu-Cys tertiary complexes were resistant to crystallization efforts, we used the structure of M. thermoresistibile EgtB_Mth in complex with dimethyl histidine and γ-Glu-Cys (PDB ID: 4X8D, Figure 2E) to guide the creation of a model of EgtB_Cth·hercynine·γ-Glu-Cys complex (Figure 2F). ²² Upon superimposing, the residues involved in the binding of cysteinyl portion of γ-Glu-Cys are identical for both EgtB_Mth and EgtB_Cth (Figure 2E). They share two conserved Arg residues involving in the binding of 1-carboxylate group of γ-Glu-Cys (Arg87 and Arg90 in EgtB_Mth vs Arg103 and Arg106 in EgtB_Cth (Figure 2E). These pairs of arginine residues also form salt bridges with the carboxylate of L-Cys. The thiol groups of γ-Glu-Cys and L-Cys replace one of the iron-center water ligands.

Based on the structural model in Figure 2F, the regions of active site in EgtB_Cth anchoring the glutamyl portion of γ-Glu-Cys vary between EgtB_Mth and EgtB_Cth. The EgtB_Mth Asp416 and Arg420 residues interacting with γ-Glu-Cys are replaced with Ala420 and Phe416 in EgtB_Cth (Figure 2F). Therefore, some hydrogen bonding and salt bridge interactions present in EgtB_Mth·dimethylhistidine·γ-Glu-Cys do not seem to be present in the modeled EgtB_Cth·hercynine·γ-Glu-Cys complex. However, favorable hydrogen-bond interaction between the γ-Glu-Cys glutamyl groups and EgtB_Cth Gln393 and Asp52 residues led to the binding of γ-Glu-Cys in EgtB_Cth as an alternative rotamer (Figure 2F) relative to that in EgtB_Mth. Overall, EgtB_Cth has a more open active site relative to that of M. thermoresistibile EgtB_Mth²² for cosubstrate recognition, which might account for both Egt1- and EgtB-types of activities in EgtB_Cth.

Based on EgtB_Mth and EgtB_Cth structures, we generated the Egt1 model using the I-TASSER (Figure S9).⁴⁷ In this Egt1 structural model, the iron center histidine ligands (His370, His463, and His467) and the binding pocket for hercynine were conserved (Figure S9). On the contrary, because EgtB_Cth is flexible in making use of either γ-Glu-Cys or L-Cys as the sulfur donor and the fact that EgtB_Cth’s active-site pocket is more open than EgtB_Mth implies that residues next to the γ-Glu-Cys may be turned to modulate EgtB_Cth selectivity (Figure S9). Structural comparison of these three enzymes identified three nonconserved residues (Asp52, Phe416, and Ala420) in the EgtB_Cth active site relative to that of EgtB_Mth and Egt1 (Figure S9). We decided to test this hypothesis by engineering these residues with the goal of tuning the EgtB_Cth activity toward the Egt1-type to facilitate ergothioneine production.

To guide our engineering effort, we employed Rosetta Enzyme Design to optimize the active site environment for substrate binding.³⁸ By using Rosetta energy function and conformational sampling of side chain rotamers, 8000 possible variants were ranked on the basis of the relative energy levels when hercynine and L-Cys are used as the substrates. The top 20 variants listed in Table S3 were further evaluated through Pymol to assess the potential interaction between the variants and L-Cys substrate. Additionally, the variants were further compared with Egt1 sequences retrieved from the protein sequence similarity network analysis in Scheme 1C (Figure S9). The sequence alignment shows conserved Leu360 and Tyr820 among Egt1 homologues; however, these two residues are not conserved in EgtB_Cth (Asp52 and Ala420). Surprisingly, the D52L and A420Y variants were listed among the top 20 variants predicted from Rosetta Enzyme Design, which warrants further examination of the role of these residues. Taking all these factors into account, we started EgtB_Cth engineering with two single mutations: EgtB_Cth A420Y and D52L.

Using hercynine and γ-Glu-Cys as the substrates for EgtB_Cth A420Y variant, the number of turnovers remained similar to that of wild-type EgtB_Cth (k_cat of 17.4 ± 0.3 min⁻¹) with lower Michaelis constant (K_M of 13.2 ± 1.3 μM for hercynine, and a K_M of 3.1 ± 0.3 mM for γ-Glu-Cys, Table 1 and Figure S10). Additionally, ¹H NMR analysis of A420Y mutant shows that the C–S bond is formed at the imidazole side chain ε-position similar to that of wild-type EgtB_Cth (Figure S11). However, when hercynine and L-Cys were the substrates, the K_M for L-Cys was lowered by 10-fold (205 ± 18 μM for wild-type vs 28.1 ± 1.8 μM for the A420Y variant), clearly indicating that this mutation altered substrate selectivity more toward the Egt1-type (Table 1 and Figure S12). Besides this kinetic information, ¹H NMR characterization of EgtB_Cth A420Y variant supports the importance of this position in controlling the reaction selectivity. When hercynine and γ-Glu-Cys were used as the substrates, the amount of sulfoxide 3 decreased from 75% in wild-type EgtB_Cth (Table 1) to 61% in EgtB_Cth A420Y variant (Table 1 and Figure S13), indicating an increased amount of side-reaction (cysteine dioxygenase activity). Interestingly, when hercynine and L-Cys were used as the substrates, the amount of sulfoxide 4 was 76% in EgtB_Cth A420Y, which is slightly improved relative to that of the wildtype EgtB_Cth (Table 1 and Figure S14). Notably, structural analysis shows that this residue does not directly interact with the substrate. However, as A420Y mutant altered the substrate selectivity, this suggested that this mutation may change in the interaction network of the substrate binding site to favor L-Cys binding, which might explain its conservancy among Egt1 homologues. The mutation reduces the active-site pocket, favoring the smaller cysteine.

These studies were also repeated using EgtB_Cth D52L mutant, and the results are shown in Table 1 and Figures S15–S17. In D52L variant, when γ-Glu-Cys was used as the substrate, the amount of sulfoxide 3 decreased to 50% of the product mixture (Figure S18). On the contrary, when L-Cys was used as the substrate, the level of the coupling product 4 was 69%, which was comparable to the wild-type activity (Figure S19). The D52L mutation can possibly disrupt the hydrogen bond between Asp52 and glutamyl group of γ-Glu-Cys, which in turn alters the substrate selectivity of EgtB_Cth. Therefore, EgtB_Cth D52 can play some roles in controlling the partition between Egt1- and EgtB-type of activities.

The promising results from these two variants led us to further characterized the activity of EgtB_Cth D52L/A420Y double mutant (Figure S20). When hercynine and γ-Glu-Cys were the substrates, the kinetic parameters were not significantly altered; however, only 52% of sulfoxide product was observed (Table 1 and Figures S21 and S22). In contrast, using hercynine and L-Cys as the substrates, EgtB_Cth D52L/A420Y exhibits the highest turnover among the wild-type and other mutants (k_cat of 32.7 ± 0.3 min⁻¹) with 75% of sulfoxide product formation (Table 1 and Figures S23 and S24). Therefore, the double mutation also tuned EgtB_Cth activity toward Egt1-type.

In summary, EgtB_Cth exhibited distinct biochemical activities showing both Egt1- and EgtB-type of reactions (Scheme 1). Thus far, only the crystal structure of M. thermoresistibile EgtB_Mth sulfoxide synthase was available. In this study, we have successfully crystallized EgtB_Cth. The EgtB_Cth and EgtB_Mth structural information allows us the opportunity to examine the factors responsible for differentiating the substrate selectivity among these sulfoxide synthases. Guided by computational results using the Rosetta Enzyme Design and information from evolutionary-related sequences (the EgtB-node and the Egt-1 node, Scheme 1C), we selected three mutants for characterizations (EgtB_Cth D52L, A420Y, and D52L/A420Y variants). Indeed, even with a single mutation, we could tune the EgtB_Cth activity more toward Egt1-type of catalysis. For these two mutants, the catalytic proficiency (k_cat/K_m) changes from ~40-fold more favoring Egt1 type in wildtype EgtB_Cth to ~180-fold more favoring Egt1-type.

Because of difficulties faced in chemical synthetic processes in industrial ergothioneine production,^48,49 there is an increasing interest of developing ergothioneine biosynthetic processes. For fermentation-based ergothioneine production, Egt1-type of pathway is preferred relative to that of the EgtB-type of pathway to alleviate the competition between ergothioneine and glutathione biosynthesis. Guided by EgtB_Cth structural information and predictions from Rosetta Enzyme Design calculation, we have successfully demonstrated that EgtB_Cth activities could be tuned toward Egt1-type (D52L, A420Y, and D52L/A420Y). These promising results open up the opportunity of engineering EgtB_Cth enzyme toward Egt1-type, with significantly better thermo-stability, which might benefit the ergothioneine industrial production processes.