Structure-function analysis of 2-sulfamoylacetic acid synthase in altemicidin biosynthesis

Introduction

Sulfonamide antibiotics, altemicidin (1) and its related compounds, (+)-SB-203207 (2) and (+)-SB-203208 (3), are alkaloid natural products isolated from Streptomyces species [1,2,3] (Fig. 1a). Compound 1 exhibits strong antitumor activity, with IC₅₀ values of 0.84 μg mL⁻¹ against L1210 lymphocytic leukemia and 0.82 μg mL⁻¹ against IMC carcinoma cell lines [4]. Compounds 2 and 3 are strong inhibitors of aminoacyl-tRNA synthetases [4]. Especially, compound 2 shows potent inhibition activity against isoleucyl-tRNA synthetases from Staphylococcus aureus, Pseudomonas fluorescens, and Candida albicans, with IC₅₀ values ranging from 1.4 to 1.7 nM [5, 6].

Compounds 1–3 share a distinctive 5,6-cis-fused azaindane bicyclic structure and a rare sulfonamide side chain [7, 8]. While the structure–activity relationship of the sulfonamide side chain in compounds 1–3 has not yet been elucidated, sulfonamide derivatives are clinically important for treating bacterial infections due to their stability, and broad antibacterial spectra [9]. Furthermore, their interactions with metal ions, amino acid residues, and DNA/RNA moieties make them versatile functional groups in medicine and drug design [10].

The biosynthetic gene cluster (sbz cluster) responsible for 1–3 production was identified in Streptomyces sp. NCIMB 40513 [11, 12]. In the biosynthesis of 1, the 2-sulfamoylacetic acyl group (2-SA) is produced from l-cysteine by two oxidation enzymes, the cupin-type cysteine oxygenase SbzM and the aldehyde dehydrogenase SbzJ (Fig. 1b). Previous in vitro assays showed that SbzM generates 2-sulfamoylacetic aldehyde (4) from l-cysteine, and then SbzJ converts 4 to 2-sulfamoylacetic acid (5) in the presence of NAD⁺. While SbzJ shares moderate sequence similarity (~40%) to biochemically and structurally well-characterized medium chain aldehyde dehydrogenases [13, 14], SbzJ accepts the unique sulfonamide-containing compound 4 as a natural substrate. Although the function of SbzJ has been identified, the substrate promiscuity of SbzJ and the structural basis for its recognition of the sulfonamide moiety remain to be elucidated. In this study, we performed a structure-function analysis of SbzJ. Our in vitro studies demonstrated that SbzJ shows broad substrate specificity toward both its aldehyde substrate and cofactor. Moreover, the complex structure of SbzJ with NAD⁺ and a model structure of SbzJ with substrate 4, combined with subsequent mutagenesis studies, identified the key residues responsible for sulfonamide moiety recognition. These observations also provided detailed insights into the reaction mechanism of SbzJ.

Materials and methods

General

Oligonucleotide primers (Supplementary Table 1) and DNA sequencing services were provided by Eurofins Genomics. The restriction enzymes and PrimeSTAR MAX DNA polymerase were purchased from Takara Bio Inc. Solvents and chemicals were purchased from Wako Chemicals, Ltd. (Tokyo, Japan), Merck KGaA Ltd. (Darmstadt, Germany), and Hampton Research (CA, USA), unless noted otherwise. PCR was performed using a TaKaRa PCR Thermal Cycler Dice® Gradient (Takara Bio Inc.).

Expression and purification of SbzJ and its variants

The pET28a plasmids for the expression of SbzJ and its variants were transformed into Escherichia coli Rosetta (DE3). The resulting strains were cultured in LB medium supplemented with 50 mg L⁻¹ kanamycin sodium at 37 °C, with shaking at 160 rpm. When the OD₆₀₀ reached 0.6, the cell cultures were cooled on ice for 30 min, and then IPTG (0.3 mM) was added to induce the target protein expression and the cell cultures were further incubated at 16 °C, 160 rpm. After 18 h of post-induction incubation, cells were harvested by centrifugation at 5500 × g for 10 min and suspended in lysis buffer, containing 20 mM Tris-HCl (pH 8.0), 100 mM NaCl, 5 mM imidazole, and 10% glycerol. The cell suspension was sonicated for 5 min on ice. After the cell debris was removed by centrifugation at 20,000 × g for 30 min, the supernatant was mixed with 1 mL of Ni-NTA resin and loaded onto a gravity flow column. Unbound proteins were removed with 100 mL lysis buffer containing 20 mM imidazole, and then the His-tagged protein was eluted with lysis buffer containing 300 mM imidazole. For the in vitro assay, the eluted enzymes were concentrated after buffer exchange to 20 mM Tris-HCl (pH 8.0), 100 mM NaCl, 5 mM imidazole, 0.5 mM EDTA, 1 mM DTT, and 10% glycerol, using a 30 kDa Amicon® Ultra-15 filtration unit (Millipore). For crystallization, the His-tag purified enzymes were applied to a 6 mL RESOURCE^TM Q anion exchange chromatography column (4 °C, Cytiva) and a HiLoad 16/60 Superdex 200 prepacked gel filtration column (4 °C, GE Healthcare), and eluted with a solution containing 20 mM HEPES (pH 8.0), and 100 mM NaCl. The resulting eluate was concentrated to 10 mg mL⁻¹, using an Amicon Ultra-15 (MWCO: 30 kDa) filter at 4 °C. The purity of the proteins was monitored by SDS-PAGE, and the protein concentrations were determined with a SimpliNano microvolume spectrophotometer.

In vitro assays of SbzJ

The standard enzymatic reaction of SbzJ was performed in a 50 μL volume, containing 50 mM potassium phosphate buffer (KPi) (pH 8.0), 2.5 mM l-Cys, 5 mM TCEP, 2.5 mM NAD⁺ or NADP⁺, 50 μM SbzM, and 10 μM SbzJ. The reaction mixture was pre-incubated for 1 h at 30 °C in the absence of SbzJ to generate 2-sulfamoylacetic aldehyde (4) from l-Cys and then incubated with SbzJ for 2 h at 30 °C to generate 2-sulfamoylacetic acid (5). For the in vitro analysis of SbzJ with various aldehydes, a 50 μL reaction mixture, containing 50 mM KPi (pH 8.0), 2.5 mM aldehydes, 5 mM TCEP, 2.5 mM NAD⁺, and 10 μM SbzJ was incubated for 2 h at 30 °C. The reaction was quenched by adding 50 μL of acetonitrile, and the products were identified by comparison with authentic standards or MS analysis.

HPLC analysis of labeled compounds

For carboxy group labeling, 50 μL of the reaction mixture was lyophilized and reconstituted with 50 μL of acetonitrile. The reconstituted mixture was treated with 5 μL of 100 mM 4-bromophenacyl bromide (BPA) and 2.5 μL of 100 mM N,N-diisopropylethylamine, and incubated for 60 min at 90 °C. For aldehyde labeling, 50 μL of the reaction mixture was treated with 1.25 μL of 1 M HCl and 20 μL of 10 mM dinitrophenylhydrazine solution, and incubated for 60 min at 37 °C. The samples were injected into a Shimadzu LC-20AD UHPLC system equipped with a COSMOSIL 5C₁₈-AR-II packed column (4.6 × 250 mm, 5 μm; Nacalai Tesque, Inc., Kyoto, Japan). The gradient elution was performed with solvent A (0.1% formic acid) and solvent B (CH₃CN), at a flow rate of 1 mL/min (min, % of B: 0 min, 10%; 5 min, 10%; 20 min, 100%; 25 min, 100%), followed by 5 min of equilibration with 10% solvent B prior to the next analysis. The labeled carboxylic acids and aldehydes were monitored at 254 nm and 350 nm, respectively.

Steady-state kinetic analysis

For the reaction of SbzJ with 4, the reaction was performed in a 50 μL volume, containing 50 mM KPi (pH 8.0), 0.0025–0.1 mM l-Cys, 2.5 mM NAD⁺, 50 μM SbzM, and 0.1 μM SbzJ. The reaction mixture was pre-incubated for 1 h at 30 °C with SbzM to generate 4 from l-cysteine, and then SbzJ was added. The reactions were performed at 20 °C. To determine the enzyme activity, the production of NADH was monitored at 340 nm over 2 min by a Cary 60 UV-Vis spectrophotometer (Agilent Technologies, USA).

The reaction of SbzJ with butyraldehyde (6) was performed in a 50 μL volume, containing 50 mM KPi (pH 8.0), 0.025–2.5 mM 6, 2.5 mM NAD⁺, and 1 μM SbzJ, at 24 °C. The reactions of SbzJ with 2-ethylbutyraldehyde (7), benzaldehyde (8), and 2-phenylpropionaldehyde (9) were performed in a 50 μL volume, containing 50 mM KPi (pH 8.0), 0.05–2.5 mM aldehydes, 2.5 mM NAD⁺, and 10 μM SbzJ, at 24 °C. For the reaction of SbzJ with 3-methylthiobutyraldehyde (11), the reaction was performed in a 50 μL volume, containing 50 mM KPi (pH 8.0), 0.025–2.5 mM 7, 2.5 mM NAD⁺, and 0.1 μM SbzJ, at 24 °C. For all reactions, the production of NADH was monitored at 340 nm over 2 min. Each reaction was performed in triplicate. GraphPad Prism (GraphPad Prism Software Inc., San Diego, CA) was used for statistical data analysis. The kinetic values for cinnamaldehyde (10) could not be determined, due to low efficiency.

Crystallization and structure determination

Crystals of SbzJ were obtained after 1 day at 20 °C. All crystals were obtained by using the sitting-drop vapor-diffusion method with the following reservoir solutions: 8% v/v Tacsimate and 20% w/v polyethylene glycol 3350. The crystals were transferred into cryoprotectant solution (reservoir solution with 25% (v/v) glycerol), and then flash-cooled to −173 °C in a nitrogen gas stream. The X-ray diffraction data sets were collected at BL-1A (Photon Factory, Tsukuba, Japan), using a beam wavelength of 1.1 Å. The diffraction data sets were processed and scaled with the XDS program package [15] and Aimless in CCP4 [16]. The initial phases of the SbzJ structures were determined by molecular replacement, using the structure of YdcW from E. coli (PDBID: 1WNB) as the search model. Molecular replacement was performed with Phaser in PHENIX [17]. The model was further calculated with AutoBuild in PHENIX [18]. The structures were modified manually with Coot and refined with PHENIX.refine [19]. The final crystal data and intensity statistics are summarized in Supplementary Table 2. A structural similarity search was performed, using the Dali program server [20]. All crystallographic figures were prepared with PyMOL (DeLano Scientific, http://www.pymol.org). To construct the docking model, the three-dimensional model structure of compound 4 was generated with the Avogadro software [21], and their geometries were optimized with the eLBOW tool in Phenix [22]. The position of 4 was determined by comparison with a benzyl adduct (PDBID: 5UCD) [23] and the mouse ALDH4 complex with the product glutamate (PDBID: 3V9K) [24]. The clashes between the ligand and the active site residues were removed by PHENIX.refine. The crystallographic data that support the findings of this study are available from the Protein Data Bank (http://www.rcsb.org). The coordinates and structure factor amplitudes for SbzJ were deposited under PDBID 9JU5.

Mutagenesis experiments of SbzJ

The reactions of wild-type SbzJ and its variants were performed in a 50 μL volume, containing 50 mM KPi (pH 8.0), 0.0025–0.1 mM l-Cys, 2.5 mM NAD⁺, 50 μM SbzM, and 1 μM wild-type or mutant SbzJ. The reaction mixture was pre-incubated for 1 h at 30 °C in the absence of SbzJ to generate 4 from l-Cys, and then incubated with wild-type or mutant SbzJ for 30 min at 30 °C. The reaction was quenched by adding 75 μL of acetonitrile. To determine the enzyme activity, the absorption of NADH was measured at 340 nm by an Infinite® 200 PRO M Nano⁺ plate reader (Tecan, Switzerland). The absorption values were corrected by subtracting the blank values of each well. Each reaction was performed in triplicate.

Results and discussion

Substrate specificity of SbzJ

While we previously reported the function of SbzJ with 4 and NAD⁺, its substrate/cofactor specificity had yet to be explored. To investigate the specificity toward the cofactor, enzyme reactions of SbzJ and SbzM were performed with l-cysteine as the substrate and NAD⁺ or NADP⁺ as a cofactor (Fig. S1). As a result, SbzJ generated 5 using NADP⁺ with 74% of the activity compared to that with NAD⁺, indicating that the enzyme does not strongly discriminate against the phosphate group at the C2″ position of the adenosine moiety. Next, to investigate the substrate specificity of SbzJ toward aldehyde compounds, enzyme reactions were performed with various aldehydes, including butyraldehyde (6), 2-ethylbutyraldehyde (7), benzaldehyde (8), 2-phenylpropionaldehyde (9), cinnamaldehyde (10), and 3-methylthiobutyraldehyde (11). Interestingly, SbzJ accepted all these compounds as substrates and generated the corresponding carboxylic acids 12–17 (Fig. 2). We also measured the steady-state kinetics for various aldehyde substrates to further clarify the substrate selectivity of SbzJ. The steady-state kinetics values are summarized in Table 1. The kcat/K_M value of SbzJ for 4 (K_M = 17.6 μM, kcat/K_M = 6.2 × 10⁴ s⁻¹ M⁻¹) is comparable to those of several aldehyde dehydrogenases (ALDHs), including the bacterial ALDH16 from Loktanella sp. (K_M = 21.3 μM, kcat/K_M = 15.5 × 10⁴ s⁻¹ M⁻¹ for hexanal) [25], ALDH-BL21 from E. coli BL21 (DE3) (K_M = 20 μM, kcat/K_M = 2.6 × 10⁴ s⁻¹ M⁻¹ for isobutyraldehyde) [26], and ALDH-11300 from Deinococcus geothermalis DSM 11300 (K_M = 70 μM, kcat/K_M = 3.8 × 10⁴ s⁻¹ M⁻¹ for butyraldehyde) [26].

Table 1 Kinetic analysis of SbzJ

Full size table

While SbzJ showed low efficiency for many substrates (kcat/K_M values less than 15 s⁻¹ M⁻¹), the enzyme exhibited relatively high efficiency for 6 (kcat/K_M = 2.3 × 10³ s⁻¹ M⁻¹) and 11 (kcat/K_M = 2.0 × 10⁴ s⁻¹ M⁻¹). The detailed fitting models and kinetic values are shown in Supplementary Fig. 2. These in vitro and kinetic analyses of SbzJ indicated that the enzyme exhibits broad substrate specificity, with a preference for aliphatic and less bulky aldehydes as substrates.

Structure analysis of SbzJ

To elucidate the structural basis of the enzyme reaction mechanism and the broad substrate specificity, the crystal structure of SbzJ in complex with NAD⁺ was solved at a resolution of 2.5 Å (Fig. S3 and Table S1). The asymmetric unit contains two structurally similar molecules, with an RMSD of 0.2 Å for the Cα-atoms. Monomer A consists of residues 16–461, while monomer B contains residues 12–461. SbzJ forms a functional dimer through interactions in the oligomerization domain (residues 116–136 and 452–461) (Fig. S3). Each monomer adopts the canonical ALDH fold [27, 28], which includes a Rossmann fold for NAD⁺-binding (residues 16–240) and a C-terminal α/β catalytic domain (residues 244–461). The overall structure of SbzJ is similar to those of other classical ALDHs, including the medium chain aldehyde dehydrogenase YdcW from E. coli [13] (PDBID: 1WNB, 37% amino acid sequence identity) and the benzaldehyde dehydrogenase from Pseudomonas putida [23] (PpBADH, PDBID: 5UCD, 24% amino acid sequence identity), with RMSD values of 1.4, and 2.1 Å for Cα-atoms, respectively (Fig. S3).

The cofactor NAD⁺ binds to the N-terminal NAD⁺-binding domain by forming a hydrogen bond network with active site residues (Fig. 3). While only one conformation of NAD⁺ was observed in monomer A, two conformations were observed in monomer B. Of the two conformations in monomer B, conformation A was almost identical to that in monomer A (Figs. 3a and S4), resembled that in the complex structure of benzaldehyde dehydrogenase PpBADH with NADP⁺ and benzoate (PDBID: 5UCD) [23] (Fig. S4). In this conformation, the C1 amino group of the adenine moiety forms hydrogen bonds with Arg206 and Asp227 (Fig. 3). Additionally, the C2″ and C3″ hydroxy groups of the adenosine ribose form hydrogen bonds with Lys170 and the main chain of Val144, while the diphosphate group of NAD⁺ interacts with Trp146 and Ser221. The C2′ and C3′ hydroxy groups of the ribose in the nicotinamide moiety interact with Glu365. The amide group of the nicotinamide forms hydrogen bonds with the main chain of Leu241 and the side chains of Glu240 and His431, positioning the pyridine ring near the catalytic Cys273, which forms a covalent bond with the substrate, at the active center. The other conformation in monomer B (conformation B) is similar to that observed in YdcW from E. coli (PDBID: 1WNB) [13] (Fig. S3). While the adenosine binding mode is the same as in conformation A, the nicotinamide mononucleotide moiety is shifted by 2 ~ 5 Å towards the entrance of the active site, due to interactions between the diphosphate group and Arg202. The distances between the C4 atom of the nicotinamide ring and the thiol group of Cys273 are 3.2 Å in conformation A and 7.0 Å in conformation B (Fig. 3). These observations suggest that conformation A represents the active binding mode during the enzymatic reaction. The adenosine ribose moiety of NAD⁺ is bound on the solvent-exposed surface of SbzJ, with significant space available to accommodate an additional phosphate group at its C2″ position of the adenosine ribose (Fig. S5). The positively charged Lys170 likely plays a key role in binding the negatively charged phosphate group. These observations are consistent with in vitro analyses showing that SbzJ can also utilize NADP⁺ as a cofactor.

A comparison of the aldehyde binding site with those of other aldehyde dehydrogenases revealed that, in addition to Cys273, the key residues Asn147 and Glu240—responsible for forming the oxyanion hole and activating the thiol group of Cys, respectively—are conserved [13, 23, 25] (Fig. 4). SbzJ also contains His431, which is conserved in the Vh-ALDH from Vibrio harveyi, YdcW [29], and PaaZ [30], a bifunctional enzyme in the phenylacetic acid (paa) degradation pathway. This position is conserved as Phe in many ALDHs, such as PpBADH, and presumably contributes to stabilizing the aliphatic chain as part of an aromatic box, which is important for apolar ligand interaction [23]. The functional analysis of Vh-ALDH revealed that His450, which is in the proximity of the thiol group of Cys289, assists in increasing the nucleophilicity of the catalytic Cys269. His431 in SbzJ is located 4.2 Å from the thiol group of Cys273 and forms hydrogen bond interactions with Glu240 and Glu416, suggesting that these residues are also important for activating the catalytic Cys273. In the enzyme reaction of PpBADH, Glu215 acts as the general base that activates the water required for the hydrolysis of the thioacyl-enzyme intermediate [23]. While Glu240 is conserved at the corresponding position of Glu215 in PpBADH, the side chain of Glu240 interacts with the amide group of the nicotinamide moiety and the imidazole ring of His431. These observations suggested that His431 may also activate a water molecule, with assistance from Glu240, facilitating the release of the carboxylate product.

The aldehyde substrate binding site is composed of Tyr148, Met152, Trp155, Ser272, Ala274, Leu423, Gln425, and His431, and is located on the opposite side of the enzyme from the NAD-binding site, suggesting that NAD⁺ and the aldehyde substrate enter from opposite directions. The volume of the aldehyde substrate binding pocket is large enough to accommodate longer aliphatic aldehydes and aromatic aldehyde substrates. Attempts to obtain a structure of SbzJ in complex with the substrate or product were unsuccessful, by co-crystallization or soaking experiments. Therefore, we modeled the Cys-tethered 4 binding structure of SbzJ, which is the first intermediate in the enzyme reaction, based on the benzaldehyde dehydrogenase complex with a benzyl adduct (PDBID: 5UCD) [23] (Fig. 4a). The docking model suggested that the carbonyl group of the thioester is located in the oxyanion hole formed by Asn147 and the main chain of Cys273. The sulfonamide moiety forms hydrogen bonds with the side chains of Tyr148, Ser272, and Gln425, indicating that SbzJ recognizes the sulfonamide group via hydrogen bond interactions. The distance between the thioester and the C4 atom of nicotinamide in this model is 3.4 Å, which is close enough to transfer a hydride from the tetrahedral intermediate to NAD⁺.

Mutagenesis analysis of SbzJ

To confirm the importance of the active site residues for the recognition of compound 4, we performed mutagenesis experiments. The active site residues in the aldehyde binding site Tyr148, Trp155, Glu240, Ser272, Cys273, Gln425, Glu416, and His431 were substituted with alanine (Fig. 5). As expected, the catalytic C273A variant abolished the oxidation activity, as in the cases of other ALDHs. Furthermore, the Y148A, W155A, E240A, S272A, Q425A, and H431A variants also exhibited strikingly curtailed or reduced activities. Notably, the Y148A, E240A, and H431A variants almost completely lost the activity (less than 10%), while the W155A, S272A, Q425A, and E416A variants retained 44%, 66%, 24%, and 22% activities, respectively, compared to the wild type. These results suggested that Glu240 plays a more important role in facilitating the activation of the catalytic Cys273 by His431 while both Glu240 and Glu416 interact with His431. Additionally, Tyr148 is the most important residue for the recognition of substrate 4 through the hydrogen bond interaction.

The mechanism of ALDHs is well-established: the active site cysteine reacts with the aldehyde substrate as a nucleophile, forming a covalently bound thiohemiacetal intermediate [31, 32]. A hydride is then transferred from the tetrahedral intermediate to the C4 atom of the nicotinamide ring, generating a thioester enzyme intermediate and NAD(P)H. This intermediate is rapidly hydrolyzed, releasing the carboxylate product and regenerating the free cysteine. Given the conservation of the key active site residues and the NAD-binding mode, SbzJ likely follows a similar mechanism (Fig. S6). First, Cys273, which is activated by His431, attacks the aldehyde of substrate 4 to form a Cys-bound tetrahedral alkoxide intermediate. Here, the alkoxide is trapped by an oxyanion hole, consisting of Asn147 and the main chain of Cys273. Then, the hydride transfer from the tetrahedral intermediate to the C4 atom of the nicotinamide ring generates a thioacyl-enzyme intermediate and NADH. Finally, the water molecule activated by His431, with assistance from Glu240, attacks to release the carboxylate product.

Conclusion

In conclusion, this study presents a detailed structure-function analysis of SbzJ, a key enzyme involved in the biosynthesis of 2-sulfamoylacetic acid. Through a combination of crystallographic studies, in vitro enzymatic assays, and mutagenesis experiments, we revealed the broad substrate specificity of SbzJ and identified the important residues responsible for its catalytic function, particularly Tyr148, Glu240, and His431. These findings shed light on the unique ability of SbzJ to recognize sulfonamide-containing substrates and its catalytic mechanism. Since SbzJ is the only known ALDH that accepts sulfonamide-containing compound, future substrate specificity analysis of ALDHs for sulfonamide-containing aldehyde compounds, along with protein engineering based on structural comparisons with SbzJ, would pave the way for generation of novel bioactive compounds.