Abstract

Sulfoacetaldehyde reductase (IsfD) is a member of the short-chain dehydrogenase/reductase (SDR) family, involved in nitrogen assimilation from aminoethylsulfonate (taurine) in certain environmental and human commensal bacteria. IsfD catalyzes the reversible NADPH-dependent reduction of sulfoacetaldehyde, which is generated by transamination of taurine, forming hydroxyethylsulfonate (isethionate) as a waste product. In the present study, the crystal structure of Klebsiella oxytoca IsfD in a ternary complex with NADPH and isethionate was solved at 2.8 Å, revealing residues important for substrate binding. IsfD forms a homotetramer in both crystal and solution states, with the C-terminal tail of each subunit interacting with the C-terminal tail of the diagonally opposite subunit, forming an antiparallel β sheet that constitutes part of the substrate-binding site. The sulfonate group of isethionate is stabilized by a hydrogen bond network formed by the residues Y148, R195, Q244 and a water molecule. In addition, F249 from the diagonal subunit restrains the conformation of Y148 to further stabilize the orientation of the sulfonate group. Mutation of any of these four residues into alanine resulted in a complete loss of catalytic activity for isethionate oxidation. Biochemical investigations of the substrate scope of IsfD, and bioinformatics analysis of IsfD homologs, suggest that IsfD is related to the promiscuous 3-hydroxyacid dehydrogenases with diverse metabolic functions.

Introduction

The C2 organosulfonates taurine and isethionate are widespread in the environment and serve as substrates for the metabolism of diverse bacteria [1]. Taurine is highly abundant in the human body as a major intracellular osmolyte and as a component of taurine-conjugated bile salts in the gut [2]. Isethionate is also present in human tissue and is thought to originate as a byproduct of taurine nitrogen assimilation by bacteria in the gut [3].

Although the specific gut bacteria responsible for converting taurine into isethionate have not yet been identified, the pathway for formation of isethionate as a byproduct of taurine nitrogen assimilation has been demonstrated and studied in the environmental bacteria Klebsiella oxytoca TauN1 and Chromohalobacter salexigens DSM3043 (Figure 1A) [4]. In this pathway, taurine is imported by a taurine ABC transporter (TauABC), followed by conversion to sulfoacetaldehyde by taurine:oxoglutarate aminotransferase (Toa), generating glutamate as an intermediate for nitrogen assimilation. Sulfoacetaldehyde is then reduced by the NADPH-dependent sulfoacetaldehyde reductase (IsfD) to isethionate, which is exported as a waste product by the putative isethionate exporter (IsfE). In both K. oxytoca and C. salexigens, genes involved in this pathway are organized in defined gene clusters (Figure 1B).

IsfD-dependent metabolic pathway and genome neighborhood of IsfD.

Figure 1.
IsfD-dependent metabolic pathway and genome neighborhood of IsfD.

(A) Schematic diagram showing the taurine nitrogen assimilatory pathway in bacteria. (B) Genome neighborhood of IsfD in Klebsiella oxytoca and Chromohalobacter salexigens [4].

Figure 1.
IsfD-dependent metabolic pathway and genome neighborhood of IsfD.

(A) Schematic diagram showing the taurine nitrogen assimilatory pathway in bacteria. (B) Genome neighborhood of IsfD in Klebsiella oxytoca and Chromohalobacter salexigens [4].

The isethionate-forming enzyme IsfD has the potential to serve as a marker for the discovery of new isethionate-producing bacteria. IsfD belongs to the SDR family (Pfam PF00106), one of the largest enzyme families, with 152 101 sequences deposited into the EMBL-EBI database to date. The SDR family includes enzymes catalyzing a broad range of reactions, most commonly NAD(P)H-dependent oxidoreductases [5]. These include the insect alcohol dehydrogenase, 3-oxoacyl-(acyl-carrier-protein) reductase and 3-hydroxyacid dehydrogenases, etc.

The structures of SDRs have been extensively studied [5]. Highly conserved structural features of NAD(P)H-dependent oxidoreductases in the SDR family include a Rossmann fold motif for nucleotide binding and an Asn–Ser–Tyr–Lys catalytic tetrad in the active site. A special characteristic of the SDR family is that members with highly divergent primary sequences may share a similar three-dimensional structure and even catalytic activity. Therefore, a detailed understanding of the structural requirements for substrate binding and catalysis is required for the identification of candidate enzymes sharing the same reactivity as IsfD.

In the present study, we determined the crystal structure of recombinantly produced IsfD from K. oxytoca TauN1, in complex with NADPH and isethionate at atomic resolution. We identified the residues required for nucleotide and sulfonate binding by site-directed mutagenesis. We further investigated the substrate scope of IsfD and found its similarity to that of the promiscuous 3-hydroxyacid dehydrogenases. The presence of IsfD homologs in taurine-related gene clusters is also discussed.

Materials and methods

General

Lysogeny Broth (LB) medium was purchased from Oxoid Limited (Hampshire, U.K.). Water used in this work was ultrapure deionized water from Millipore Direct-Q. Acetonitrile was purchased from Concord Technology (Minnesota, U.S.A.). Formic acid was purchased from Merck (New Jersey, U.S.A.). Oligonucleotide primers were synthesized by TSINGKE Biological Technology. TALON resins were purchased from Clontech Laboratories Inc. (California, U.S.A.). All protein purification chromatographic experiments were performed on an ÄKTA pure FPLC system (GE Healthcare, U.S.A.). Isethionate was purchased from Sigma–Aldrich. NAD(P)H and NAD(P)+ were purchased from Solarbio (Beijing, China).

Gene syntheses and cloning

The codon-optimized gene fragment of K. oxytoca TauN1 isfD (Encoded protein Uniprot D3U1D9) was synthesized by General Biosystems Inc. (Anhui, China) and inserted into pET-28a-HMT vector [6] at the SspI restriction site, using the Gibson Assembly® cloning protocol [7] (New England Biolabs (NEB), Ipwich, MA, U.S.A.). The resulting plasmid HMT–IsfD contains in tandem: a His6-tag, maltose binding protein (MBP) and a TEV protease cleavage site, followed by the construct of interest and was verified by sequencing.

Expression and purification of IsfD

Escherichia coli BL21 (DE3) cells (NEB) were transformed with the plasmid HMT–IsfD and plated on LB agar supplemented with 50 µg/ml kanamycin. Transformants were grown in LB medium (typically 1 l in a 2.6 l flask) at 37°C in a shaking incubator at 220 rpm. When OD600 reached ∼0.8, the temperature was decreased to 18°C and isopropyl β-d-1-thiogalactopyranoside (IPTG) was added to a final concentration of 0.5 mM to induce the production of the protein. After 16 h, cells were harvested by centrifugation (8000×g for 10 min at 4°C).

Typically cells (∼8 g wet weight, roughly from 2 l culture) were resuspended in 40 ml of lysis buffer [50 mM Tris–HCl, pH 8.0, 1 mM phenylmethanesulfonyl fluoride, 0.2 mg/ml lysozyme, 0.03% Triton X-100 and 0.02 mg/ml DNase I]. The cell suspension was frozen in a −80°C freezer and then thawed and incubated in a 25°C water bath for 30 min to allow for cell lysis. Eight milliliters of 6% streptomycin sulfate (dissolved in water) was added to the cell lysate followed by gentle mixing and then centrifugation (20 000×g for 10 min at 4°C). The supernatant was filtered through a 0.22 µm filter and loaded onto a 10 ml TALON cobalt column pre-equilibrated with buffer A (20 mM Tris–HCl, pH 7.5 and 5 mM β-mercaptoethanol (BME) and 0.2 M KCl). The column was washed with 10 column volumes (CV) of buffer A and then the protein was eluted with 5 CV of buffer A containing 150 mM imidazole. The eluate (50 ml) was dialyzed against 2 L buffer A for 4 h to remove imidazole and then loaded on a column packed with 40 ml amylose resin (NEB). The column was washed with 5 CV of buffer A and eluted with 2 CV of buffer A containing 10 mM maltose. The eluate was mixed with recombinant His6-tagged TEV protease (1 : 70 molar ratio) and dialyzed overnight against 2 l buffer A. The dialyzed sample was loaded on a 10 ml TALON cobalt column to retain TEV protease and His6-MBP. The flow-through was collected and dialyzed against 2 l buffer B (20 mM Tris–HCl, pH 8.0, 5 mM BME) for 4 h before it was applied to a 10 ml HiTrap Q HP (GE Healthcare). The column was eluted with buffer B containing a linear salt gradient from 100 to 500 mM KCl. A prominent peak containing IsfD, eluted at ∼200 mM KCl, was collected and concentrated to a final volume of 3 ml (5 mg/ml) using a centrifugal concentrator (10K MWCO; Millipore). This protein solution was then injected to a Superdex 200 gel filtration column (∼300 ml) and eluted with buffer C (20 mM Tris–HCl, pH 7.5, 200 mM KCl, 1 mM DTT (dithiothreitol). The protein-containing peak was collected, concentrated and buffer-exchanged with the storage buffer (10 mM HEPES, pH 7.4, 50 mM KCl, 1 mM TCEP (tris(2-carboxyethyl)phosphine) using a centrifugal concentrator. The purified protein was examined on a 12% SDS polyacrylamide gel.

The concentration of purified IsfD was calculated from its absorption at 280 nm (ε280nm = 29 450 M−1 cm−1), measured using a NANODROP ONE (Thermo SCIENTIFIC).

Determination of the oligomeric state of IsfD

A 5 mg/ml solution of IsfD was analyzed by gel filtration as described above. Briefly, a 3 ml protein solution was injected into a Superdex 200 gel filtration column (∼300 ml) and eluted over 150 min with buffer C at 2 ml/min. The same conditions were used to analyze a solution of molecular mass markers including thyroglobulin bovine (669 kDa), horse apoferritin (443 kDa), sweet potato β-Amylase (200 kDa), bovine serum albumin (BSA, 66 kDa) and bovine carbonic anhydrase (29 kDa) (Sigma MWGF 1000-1KT). The molecular mass of IsfD was calculated from its elution volume, using a second-degree polynomial for the relationship between log(molecular mass) and retention time.

Determination of melting temperature of IsfD

IsfD melting curves were measured using fluorescence-based thermal shift assays [8]. Samples contained 0.2 mg/ml protein and 1× Sypro Orange protein gel stain solution (Sigma) in a total volume of 50 µl. Melting curves were obtained in a QuantStudio 6 Flex real-time PCR machine (Life, CA, U.S.A.). The temperature was increased from 10 to 95°C with an increment rate of 0.033°C/s. The melting temperatures were obtained as the midpoint of each transition. The experiments were performed in triplicate.

Site-directed mutagenesis

Twelve single amino acid point mutations in the enzyme active site, T13A, R36A, R37A, S141A, I142A, Y148A, Y154A, I186A, F191A, R195A, Q244A, F249A, were introduced by site-directed mutagenesis using primers listed in Supplementary Table S1 and confirmed by sequencing. A 25 µl of PCR reaction contained 100 ng HMT–IsfD plasmid as template, 0.4 µM forward and reverse primers and the Fast Alteration DNA Polymerase (KM101 from TIANGEN, Beijing, China). The 17-cycle PCR reaction mixture was digested by DpnI to remove the template before transformed into FDM competent cells (TIANGEN).

One-step chromatographic purification of HMT fusion WT and mutant IsfDs

To assess the effects of point mutations on enzymatic activity, HMT fusions of the WT and mutant enzymes were purified using TALON cobalt affinity chromatography. Two hundred milliliters of cell cultures were induced for HMT–IsfD WT and mutant protein production. The proteins were purified as described above, except that instead of using dialysis to remove imidazole, the eluate from the cobalt column was precipitated with ammonium sulfate followed by buffer exchange into buffer C using a G25 column. The HMT–IsfD fusion proteins used for enzymatic assays, including wild-type (WT) and mutants, were less pure due to a simplified one-step chromatographic purification protocol (estimated >80% by gel analyses). Their concentrations were determined using the Bradford method, on a Tecan M200 plate reader with BSA as standard.

Enzyme activity assays and kinetic analyses

In a typical assay, 50 mM 3-(Cyclohexylamino)-2-hydroxy-1-propanesulfonic acid (CAPSO), pH 10.0, 0.1 M isethionate and 0.2 mM NAD(P)+ were premixed and 2.5 µg of IsfD was added to initiate the reaction (total volume 200 µl) in a 96-well plate. The absorbance at 340 nm was monitored using a Tecan M200 plate reader with 15 s intervals for 2–3 min at room temperature. To measure the Michaelis–Menten kinetic constants, the concentration of the substrate was varied in the range of 0–0.2 M for isethionate/l-serine or 0–0.4 M for 3-hydroxypropionate, in the presence of a saturating concentration of 0.2 mM NADP+. In another set of experiments, the concentration of isethionate was fixed at 0.4 M, while the NADP+ concentration was varied from 0 to 0.1 mM. ΔA340 nm and the extinction coefficient of NADPH (6220 M−1 cm−1) were used to calculate the rates of the reactions. Graphpad Prism 6.0 was used to extract the kinetic parameters.

LC–MS detection of sulfoacetaldehyde formation

The sulfoacetaldehyde product was detected by derivatization with 2,4-dinitrophenylhydrazine (DNPH) (J&K, Beijing, China) [9]. A 200 µl reaction mixture, containing 50 mM CAPSO, pH 10.0, 400 mM isethionate, 1 mM NAD(P)+ and 50 µg of HMT–IsfD(WT), was incubated for 5 min at room temperature. One hundred microliters of reaction solution was mixed with 1.1 ml of 0.73 M sodium acetate buffer (pH 5.0), followed by 800 µl of freshly prepared DNPH solution (40 mg dissolved in 100 ml methanol). The mixture was incubated at 50°C for 1 h and then filtered prior to LC–MS analysis.

LC–MS analysis was performed on an Agilent 6420 Triple Quadrupole LC/MS instrument (Agilent Technologies, CA, U.S.A.). LC–MS analysis was carried out with 20 µl sample volume on an Agilent ZORBAX SB-C18 column (4.6 × 250 mm, product number 880975-902). The column was equilibrated with 75% of 0.1% formic acid in H2O, 25% of 0.1% formic acid in CH3CN and developed at a flow rate of 1.0 ml/min from 25 to 65% CH3CN. UV detection was set at 360 nm.

Crystallization, data collection and structure determination of IsfD

Initial screening of IsfD crystals was performed using an automated liquid handling robotic system (Gryphon, Art Robbins, CA, U.S.A.) in 96-well format by the sitting-drop vapor-diffusion method. The screens were set up at 295 K using various sparse matrix crystal screening kits from Hampton Research (CA, U.S.A.) and Molecular Dimensions (OH, U.S.A.). Several crystallization conditions gave hexagonal rod-shaped crystals. After further optimization using the hanging-drop vapor-diffusion method in 24-well plates, we obtained crystals large enough for single crystal X-ray diffraction studies. The best condition yielding large crystals was 0.2 M ammonium acetate, 0.1 M sodium citrate, pH 5.5, 30% W/V PEG 4000 (5 mM NADPH, 0.4 M isethionate). Crystals were flash-cooled in liquid nitrogen using a reservoir solution containing 25% glycerol as a cryoprotectant and same concentrations of NADPH and isethionate. Diffraction data were collected on BL18U1 at Shanghai Synchrotron Radiation Facility (SSRF) to a resolution of 2.8 Å. The data set was indexed, integrated and scaled using HKL3000 suite [10]. Molecular replacement was performed by PHENIX [11] using 2NWQ as a search model. The structure was manually built according to the modified experimental electron density using Coot [12] and further refined by PHENIX [11] in iterative cycles. The statistics for data collection and final refinement are presented in Table 1. All structural figures were generated using UCSF Chimera [13]. The atomic co-ordinates and structure factor amplitudes (accession code 6IXJ) have been deposited into the PDB (http://wwpdb.org).

Table 1
Data collection and refinement statistics for the Klebsiella oxytoca IsfD crystal (PDB ID: 6IXJ)
Crystal Klebsiella oxytoca IsfD 
λ for data collection (Å) 0.9795 
Data collection 
 Space group P 32 
Cell dimension (Å) 
 a, b, c (Å) 125.86, 125.86, 173.92 
 α, β, γ (°) 90 90 120 
 Resolution 37.23–2.80 (2.90–2.80) 
Rmerge 0.150 (0.708) 
 Average I/σ (I10.8 (2.0) 
 Completeness (%) 99.85 (99.43) 
 Redundancy 5.2/5.3 
Z 12 
Refinement 
 Resolution 37.23–2.80 Å 
 No. of reflections 75 893 
Rfactor/Rfree (10% data) 0.207/0.239 
 RMSD length (Å) 0.004 
 RMSD angle (°) 0.77 
No. of atoms 
 Protein 22 418 
 Ligands 660 
 Water 248 
Ramachandran plot (%) 
 Most favored 96.45 
 Additionally allowed 2.61 
Crystal Klebsiella oxytoca IsfD 
λ for data collection (Å) 0.9795 
Data collection 
 Space group P 32 
Cell dimension (Å) 
 a, b, c (Å) 125.86, 125.86, 173.92 
 α, β, γ (°) 90 90 120 
 Resolution 37.23–2.80 (2.90–2.80) 
Rmerge 0.150 (0.708) 
 Average I/σ (I10.8 (2.0) 
 Completeness (%) 99.85 (99.43) 
 Redundancy 5.2/5.3 
Z 12 
Refinement 
 Resolution 37.23–2.80 Å 
 No. of reflections 75 893 
Rfactor/Rfree (10% data) 0.207/0.239 
 RMSD length (Å) 0.004 
 RMSD angle (°) 0.77 
No. of atoms 
 Protein 22 418 
 Ligands 660 
 Water 248 
Ramachandran plot (%) 
 Most favored 96.45 
 Additionally allowed 2.61 

Sequence alignments

MUSCLE [14] was used to construct multiple sequence alignments. Sequence logos were plotted using WebLogo [15].

Results

IsfD is a tetramer in solution

IsfD was purified to near homogeneity through multiple chromatographic steps. The purified protein was examined on a 12% SDS gel (Figure 2A). The gel filtration elution profile of purified IsfD shows a single symmetric peak centered at 201 ml (Figure 2B). The observed molecular mass for IsfD was 101 kDa, whereas the calculated molecular mass for IsfD monomer is 27 kDa. This suggests that IsfD exists as a tetramer in solution, which is consistent with the oligomeric state of many other SDR enzymes [5,16].

SDS–PAGE, gel filtration and thermostability analyses of purified IsfD.

Figure 2.
SDS–PAGE, gel filtration and thermostability analyses of purified IsfD.

(A) 12% SDS gel of purified IsfD with: lane1, molecular mass marker; lane 2–4, 1, 2, 4 µg of IsfD. (B) Elution profile of IsfD using Superdex 200 gel filtration chromatography to determine IsfD molecular mass, estimated to be 101 kDa. (C) Fluorescence-based melting temperature measurement.

Figure 2.
SDS–PAGE, gel filtration and thermostability analyses of purified IsfD.

(A) 12% SDS gel of purified IsfD with: lane1, molecular mass marker; lane 2–4, 1, 2, 4 µg of IsfD. (B) Elution profile of IsfD using Superdex 200 gel filtration chromatography to determine IsfD molecular mass, estimated to be 101 kDa. (C) Fluorescence-based melting temperature measurement.

IsfD is heat tolerant

Thermal stability of K. oxytoca IsfD was measured by fluorescence-based thermal shift assays. The melting temperature was determined to be 58°C (Figure 2C). The observed relatively high thermostability of IsfD is consistent with the previous finding of optimal temperature for IsfD at 40°C, with enzyme activity readily detected at 50°C [17].

IsfD catalyzes the oxidation of isethionate

Sulfoacetaldehyde, the physiological substrate of IsfD, is unstable and previous assays required it to be introduced as a bisulfite adduct [4]. For more quantitative and reproducible routine assays, we established a protocol to assay the reverse reaction, i.e. the NADP+-dependent oxidation of isethionate. In a previous report, no reaction was observed with isethionate concentrations in the range from 0.4 to 7 mM and 1 mM NADP+, at pH 9 and pH 6 [4]. However, we found that the reaction could be promoted by increasing the isethionate concentration and pH. When IsfD was incubated with 100 mM isethionate and 0.2 mM NADP+ at pH 10, a time-dependent linear increase in A340 nm was observed (Figure 3A), indicating the formation of NADPH. LC–MS detection confirmed the production of sulfoacetaldehyde as a product (Figure 3B,C). No reaction was observed in negative controls omitting either IsfD or isethionate. Furthermore, no activity was detected when NAD+ was used in place of NADP+, consistent with previous reports regarding the nucleotide specificity of the forward reaction [4].

Enzymatic assay for IsfD.

Figure 3.
Enzymatic assay for IsfD.

(A) The assays monitor NADPH formation accompanying isethionate oxidation by IsfD. (B) The elution profile of the LC–MS assays monitoring absorbance at 360 nm. (C) The ESI (-) m/z spectrum of the sulfoacetaldehyde-DNPH peak in (B).

Figure 3.
Enzymatic assay for IsfD.

(A) The assays monitor NADPH formation accompanying isethionate oxidation by IsfD. (B) The elution profile of the LC–MS assays monitoring absorbance at 360 nm. (C) The ESI (-) m/z spectrum of the sulfoacetaldehyde-DNPH peak in (B).

The spectrophotometric assay monitoring NADPH formation was used to measure the kinetic parameters of the reverse reaction (Supplementary Figure S1), which were compared with the parameters reported in the literature [4] for the forward reaction in Table 2. The kcat/KM for the sulfoacetaldehyde reduction is ∼1000-fold higher than that of the reverse isethionate oxidation, whereas the KM for the nucleotide substrates in both directions is in the same order of magnitude.

Table 2
Kinetic parameters of IsfD in the reduction in sulfoacetaldehyde and oxidation of isethionate
Reactions kcat (s−1KM (nucleotide, µM) KM (substrate, mM) kcat/KM(substrate) (M−1 s−1
Reduction [170.1 22 0.025 4000 
Oxidationthe present study 0.15 11.6 ± 1.3 50.8 ± 5.8 2.9 
Reactions kcat (s−1KM (nucleotide, µM) KM (substrate, mM) kcat/KM(substrate) (M−1 s−1
Reduction [170.1 22 0.025 4000 
Oxidationthe present study 0.15 11.6 ± 1.3 50.8 ± 5.8 2.9 

Crystal structure of IsfD

The crystal structure of IsfD in complex with NADPH and isethionate was solved at 2.8 Å resolution. The asymmetric unit contains 12 monomers of IsfD, among which there is one intact homotetramer exhibiting 222-point group symmetry (Figure 4A). The core structure of IsfD is a typical Rossmann fold, consisting of a central seven-stranded parallel β-sheet (β1–β7) sandwiched by two arrays of three α-helices (α1, α2, α8 and α3, α4, α5) (Figure 4B), similar to those of previously reported SDR family members. The RMSD between IsfD and four closely related SDR enzymes (3ASV, 2JAH, 3RKU and 3AY6) are 0.70, 0.80, 0.87 and 1.02 Å (over 240 Cα atoms), respectively [1820]. Figure 4C shows the structure-based sequence alignments of IsfD with these four SDR enzymes. The sequence identities between IsfD and 3ASV, 2JAH, 3RKU, 3AY6 are 48.6, 36.0, 37.7 and 22.7%, respectively.

The quaternary, subunit structure of IsfD and the structure-based sequence alignments of IsfD with selected SDR members.

Figure 4.
The quaternary, subunit structure of IsfD and the structure-based sequence alignments of IsfD with selected SDR members.

(A) The quaternary structure of IsfD is shown in ribbon with four subunits in cornflower blue, magenta, orange and dark green, respectively. (B) Subunit structure of IsfD is shown in ribbon. α-helices, β strands and 310 helices in blue, sienna and green, respectively. (C) Structure-based sequence alignments of IsfD with selected SDR members including 2JAH (Clavulanic acid dehydrogenase from S. clavuligerus), 3ASV (YdfG from E. coli), 3AY6 (Glucose 1-dehydrogenase IV from B. megaterium) and 3RKU (NADP+-dependent serine dehydrogenase from S. cerevisiae). Highlighted residues are important for nucleotide binding (blue), substrate binding (red) and catalysis (NSYK catalytic tetrad in box).

Figure 4.
The quaternary, subunit structure of IsfD and the structure-based sequence alignments of IsfD with selected SDR members.

(A) The quaternary structure of IsfD is shown in ribbon with four subunits in cornflower blue, magenta, orange and dark green, respectively. (B) Subunit structure of IsfD is shown in ribbon. α-helices, β strands and 310 helices in blue, sienna and green, respectively. (C) Structure-based sequence alignments of IsfD with selected SDR members including 2JAH (Clavulanic acid dehydrogenase from S. clavuligerus), 3ASV (YdfG from E. coli), 3AY6 (Glucose 1-dehydrogenase IV from B. megaterium) and 3RKU (NADP+-dependent serine dehydrogenase from S. cerevisiae). Highlighted residues are important for nucleotide binding (blue), substrate binding (red) and catalysis (NSYK catalytic tetrad in box).

The conformation of the C-terminal tail formed by β8 and β9 strands differs widely between these five structures (Figure 5). β8–β9 of IsfD protrudes away from the core structure and forms an antiparallel β-sheet with the C-terminal tail from the diagonal subunit (Figure 4A). Such C-terminal cross interactions vary among different SDR family members (Figure 5). IsfD has the most extensive cross-interaction between the diagonal subunits with a buried surface of 2570 Å2. In contrast, the interaction is reduced in other SDRs (2230, 2200, 146 and 79 Å2 for 3AY6, 3ASV, 3RKU, 2JAH in Figure 5A). The significance of this interaction will be discussed later.

Surface and ribbon diagrams revealing representative C-tail structural profiles.

Figure 5.
Surface and ribbon diagrams revealing representative C-tail structural profiles.

(A) Surface diagram. The four subunits in each homotetramer are shown in different colors. Each red dotted box shows the buried interaction surface between the C-terminal tails from a pair of diagonal subunits. (B) Ribbon diagram. Presented structures include IsfD, 3AY6 (Glucose 1-dehydrogenase IV from B. megaterium), 3ASV (YdfG from E. coli), 3RKU (NADP+-dependent serine dehydrogenase from S. cerevisiae) and 2JAH (Clavulanic acid dehydrogenase from S. clavuligerus) in order (from left to right).

Figure 5.
Surface and ribbon diagrams revealing representative C-tail structural profiles.

(A) Surface diagram. The four subunits in each homotetramer are shown in different colors. Each red dotted box shows the buried interaction surface between the C-terminal tails from a pair of diagonal subunits. (B) Ribbon diagram. Presented structures include IsfD, 3AY6 (Glucose 1-dehydrogenase IV from B. megaterium), 3ASV (YdfG from E. coli), 3RKU (NADP+-dependent serine dehydrogenase from S. cerevisiae) and 2JAH (Clavulanic acid dehydrogenase from S. clavuligerus) in order (from left to right).

Nucleotide substrate specificity

The electron density is well defined for NADPH and the coordinating residues, showing that the side chains of T13, R36, R37 form hydrogen bond interactions with the 2′-phosphate group of NADPH (Figure 6). The two arginine residues have been previously shown conserved in some other SDR enzymes contributing to the substrate specificity of NADP(H) over NAD(H) [21]. We further demonstrated the involvement of a T/S residue. SDR enzymes that can use both NAD(H) and NADP(H) lack this T(S)RR motif.

Zoomed-in view of the IsfD active site.

Figure 6.
Zoomed-in view of the IsfD active site.

The substrate-binding pocket is colored in green and the diagonal C-terminal tail contributing to isethionate binding is colored in blue. The 2mFo-DFc electron density maps of both substrates are displayed at 1σ. The side chains of substrate-interacting residues are displayed and labeled. The hydrogen bonds involved in substrate and C-terminal tail stabilization are indicated by dashed lines.

Figure 6.
Zoomed-in view of the IsfD active site.

The substrate-binding pocket is colored in green and the diagonal C-terminal tail contributing to isethionate binding is colored in blue. The 2mFo-DFc electron density maps of both substrates are displayed at 1σ. The side chains of substrate-interacting residues are displayed and labeled. The hydrogen bonds involved in substrate and C-terminal tail stabilization are indicated by dashed lines.

The active site structure

The bound isethionate is oriented with its hydroxyl group facing the tyrosine residue of the catalytic tetrad (Y154) and its sulfonate group forming hydrogen bond network with Y148, R195, Q244 and a water molecule. The side chains of I142, I186 and F191 further stabilize the conformation of the substrate through hydrophobic interactions (Figure 6). The active site structure is consistent with the proposed mechanism general to SDR oxidoreductase, involving deprotonation of the hydroxyl group by the tyrosine residue of the catalytic tetrad (Y154), and a hydride transfer from the substrate to the C4 position of the nicotinamide ring (Figure 6) [16].

A striking feature of the IsfD structure is that its C-terminal tail protrudes away from the diagonal subunit and constitutes part of the isethionate-binding site. The isethionate-bound complex structure shows a detailed picture of how this diagonal C-terminal tail interacts with the substrate. The F249 from the diagonal subunit is in close van der Waals contact with I142; meanwhile, it pushes the side chain of Y148 towards the substrate-binding site to position it in an orientation ready to form a hydrogen bond with the sulfonate group of isethionate.

Most of the SDR members do not exhibit such C-terminal tail cross-interaction from the diagonal subunit with a few exceptions. Figure 5 shows some representative C-terminal tail interactions within the SDR homotetramers. IsfD shows the most extensive cross interactions (Figure 5A,B). The glucose-1 dehydrogenase showing the second largest cross interactions is unique in that its C-terminal carboxylate group interacts with the substrate from the diagonal active site (Figure 5B). Interestingly, a previously annotated E. coli serine dehydrogenase, YdfG, which was later found to be a malonic semialdehyde reductase [22] showed a similar C-terminal interaction as IsfD (Figure 5B). The substrate is absent from this structure, but a phosphate anion likely from the crystallization buffer is present in the active site instead [18]. A yeast SDR member, exhibiting serine dehydrogenase activity in vitro but with unknown physiological function shows a distinct C-terminal structure (Figure 5B). Pro261 in the middle of its C-terminal tail forms a kink and makes the second half of the tail fold back to interact with the active site of the same subunit. Finally, the clavulanic acid dehydrogenase contains a shorter C-terminal tail, which forms a parallel β-sheet with the β-strand from the tail of the neighboring subunit (Figure 5B).

Site-directed mutagenesis

Based on our crystal structure, mutants predicted to affect nucleotide interaction, substrate interaction and catalysis were subjected to activity assays, and their activities were compared with that of the wild-type protein. Relative activities of these mutant enzymes are summarized in Table 3. Enzymatic activity was almost completely abolished in all the mutants except for I142A and I186A, which retained ∼30% of the WT activity.

Table 3
Relative enzyme activities of IsfD mutants in isethionate oxidation
Enzyme Relative activity (%) 
Wild-type 100 
Nucleotide interaction 
 T13A 7.75 
 R36A No activity 
 R37A No activity 
Catalytic center 
 S141A No activity 
 Y154A No activity 
Substrate interaction 
 I142A 33.55 
 Y148A No activity 
 I186A 25.79 
 F191A No activity 
 R195A No activity 
 Q244A No activity 
 F249A No activity 
Enzyme Relative activity (%) 
Wild-type 100 
Nucleotide interaction 
 T13A 7.75 
 R36A No activity 
 R37A No activity 
Catalytic center 
 S141A No activity 
 Y154A No activity 
Substrate interaction 
 I142A 33.55 
 Y148A No activity 
 I186A 25.79 
 F191A No activity 
 R195A No activity 
 Q244A No activity 
 F249A No activity 

To investigate the conservation of the active site residues, we examined the 2400 IsfD homologs in the UniRef cluster UniRef50_Q9KWN1, where each member shares ≥50% sequence identity and ≥80% overlap with the seed sequence of the cluster [23]. Sequence alignments showed that the key residues for substrate and nucleotide binding, are conserved in this cluster (Figure 7A), consistent with the mutagenesis data. This suggests a similar substrate scope for enzymes in this cluster.

Bioinformatics studies of IsfD homologs.

Figure 7.
Bioinformatics studies of IsfD homologs.

(A) Conservations of key active site resides. Sequence logos of the key residues selected for mutation in Table 3 are plotted using WebLogo [15]. (B) Representative bacterial operons putatively responsible for isethionate production.

Figure 7.
Bioinformatics studies of IsfD homologs.

(A) Conservations of key active site resides. Sequence logos of the key residues selected for mutation in Table 3 are plotted using WebLogo [15]. (B) Representative bacterial operons putatively responsible for isethionate production.

Bioinformatics

To investigate the involvement of close homologs of IsfD in pathways for conversion of taurine to isethionate, we focused on the IsfD homologs in UniRef50_Q9KWN1. In addition to K. oxytoca and C. salexigens IsfD, this cluster also contains the YdfG homolog from Salmonella typhimurium (UniProt P69936) and the previously characterized serine dehydrogenase from Rhizobium radiobacter (UniProt Q9KWN1). This indicates additional restraints are needed.

Our one-step forward strategy is to search for the presence of candidate taurine aminotransferases in the genome neighborhoods of IsfD homologs. Using the web-based Enzyme Function Initiative Genome Neighborhood Tool (EFI-GNT) [24], we looked for the occurrence of aminotransferases belonging to the family PF00202, which includes previously reported taurine aminotransferases, within a window of 10 open reading frames surrounding the IsfD homolog. A total of 233 IsfD homologs were found to be associated with an aminotransferase in this family. Out of these aminotransferases, 129 are homologs of K. oxytoca Toa belonging to the cluster UniRef50_A0A0N9XFD2, and 36 are homologs of Bilophila wadsworthia taurine:pyruvate aminotransferase (Tpa) belonging to the cluster UniRef50_Q9APM5.

The IsfD homologs were represented pictorially in the form of a sequence similarity network (Supplementary Figure S2) [25], constructed using the web-based EFI enzyme similarity tool and plotted using Cytoscape [26]. Representative gene clusters are shown in Figure 7B.

Discussion

Determination of the crystal structure of IsfD in complex with isethionate and NADPH, together with mutagenesis of active site residues, enabled us to determine the structural requirements for its substrate binding and catalysis. The overall structure of IsfD monomer exhibits a typical Rossmann fold, in common with other members of the SDR family [5]. IsfD is a homotetramer in both crystal and solution, in consistence with the observation that the homodimer and homotetramer are the most common quaternary structures of the SDR family members. The similarity of the active site of IsfD to those of several previously structurally characterized 3-hydroxyacid dehydrogenases led us to further investigate the substrate scope of IsfD. The activities of NADP+-dependent oxidation of 3-hydroxypropionate and l-serine but not l-threonine were detected. Kinetic parameters were obtained (Supplementary Figure S1) and compared (Table 4).

Table 4
Kinetic parameters of IsfD on different substrates
Substrate Km (mM) kcat (s−1Relative activity (%) 
Isethionate 50.8 ± 5.8 0.15 ± 0.06 100 
l-serine 77.4 ± 12.1 1.9 ± 0.13 1267 
3-Hydropropionate 151.4 ± 11.8 1.1 ± 0.04 733 
Substrate Km (mM) kcat (s−1Relative activity (%) 
Isethionate 50.8 ± 5.8 0.15 ± 0.06 100 
l-serine 77.4 ± 12.1 1.9 ± 0.13 1267 
3-Hydropropionate 151.4 ± 11.8 1.1 ± 0.04 733 

IsfD is the first member of the diverse group of 3-hydroxyacid dehydrogenases to be crystallized with a substrate/product bound in its active site. The most striking feature of IsfD in complex with both NADPH and isethionate is the cross-interaction of the C-terminal tails with the diagonal subunits, which constitutes part of the active site. This C-terminal interaction was also observed in the crystal structure of E. coli YdfG in complex with NADP+ and a phosphate anion (Figure 5B). In contrast, the cross-interaction is absent from the 3-hydroxyacid dehydrogenase from Saccharomyces cerevisiae, where a kink in the C-terminal tail introduced by a proline residue at position 261 causes it to fold back, favoring self-interaction within each subunit (Figure 5B). This diversity in the C-terminal tail interactions of tetrameric 3-hydroxyacid dehydrogenases is an important consideration in future functional predictions of SDRs from their primary sequence.

Our investigation of the genome neighborhood of close homologs of IsfD revealed many gene clusters related to taurine metabolism. These include some organisms that were previously noted containing IsfD, including Klebsiella, Chromohalobacter and Marinimonas [4]. In addition, the discovery of IsfD-containing gene cluster in Pseudomonas sp. is consistent with the previous report of isethionate formation by this bacterium. Some of the clusters contain Tpa and alanine dehydrogenase instead of Toa. However, the vast majority of IsfD homologs are not associated with taurine degradation genes. This includes the E. coli YdfG. Though the in vivo function of YdfG is not fully understood, one of its functions is thought to be as a malonic semialdehyde reductase [27], for detoxification of malonic semialdehyde produced in the RUT pyrimidine degradation pathway, forming 3-hydroxypropionate as a product (Supplementary Figure S4).

Our work provides a basis for further investigations of the prevalence of microbial pathways for taurine and isethionate metabolism, which are important intermediates in sulfur metabolism in the human gut. Considering the structural similarity between isethionate and 3-hydroxypropionate, we hypothesize that the ubiquitous 3-hydroxyacid dehydrogenases are readily co-opted into pathways for isethionate formation, and further investigation may lead to the discovery of new isethionate-forming bacteria.

Abbreviations

     
  • BME

    β-mercaptoethanol

  •  
  • BSA

    bovine serum albumin

  •  
  • CAPSO

    3-(cyclohexylamino)-2-hydroxy-1-propanesulfonic acid

  •  
  • CV

    column volumes

  •  
  • DNPH

    2,4-dinitrophenylhydrazine

  •  
  • DTT

    dithiothreitol

  •  
  • EFI-GNT

    enzyme function initiative genome neighborhood tool

  •  
  • HMT

    His6-tag, maltose binding protein (MBP) and a TEV protease cleavage site

  •  
  • IsfD

    NADPH-dependent sulfoacetaldehyde reductase

  •  
  • IsfE

    isethionate exporter

  •  
  • LB

    Luria Bertani

  •  
  • SDR

    short-chain dehydrogenase/reductase

  •  
  • SSRF

    Shanghai Synchrotron Radiation Facility

  •  
  • TauABC

    Taurine ABC transporter

  •  
  • TCEP

    (tris(2-carboxyethyl)phosphine)

  •  
  • Toa

    taurine:oxoglutarate aminotransferase

  •  
  • Tpa

    taurine:pyruvate aminotransferase

  •  
  • WT

    wild-type

Author Contribution

Y.Z. designed and carried out experiments with IsfD cloning, expression, purification crystallization and enzyme activity assays. Y.W. designed and carried out experiments with bioinformatics and is involved in writing the manuscript. L.L. and T.X. are involved in collecting and analyzing the X-ray diffraction data. E.L.A., H.Z., Z.Y. and Y.Z. are involved in conceptualizing the project, getting grants for the project, overall supervision of the project and writing the manuscript.

Funding

This work was supported by the National Science Foundation of China [Grant 31570060 and 31870049 (to Y.Z.)], National Key Research and Development Program of China [2017YFD0201400, 2017YFD0201403 (to Z.Y.)] and the Agency for Science, Research and Technology of Singapore Visiting Investigator Program (to H.Z.).

Acknowledgements

We thank the staff for assistance in using the in-house X-ray diffraction machines at Tianjin University, and X-ray facility on the beamline BL18U1 at Shanghai Synchrotron Radiation Facility.

Competing Interests

The Authors declare that there are no competing interests associated with the manuscript.

References

References
1
Cook
,
A.M.
and
Denger
,
K.
(
2002
)
Dissimilation of the C2 sulfonates
.
Arch. Microbiol.
179
,
1
6
2
Cook
,
A.M.
and
Denger
,
K.
(
2006
)
Metabolism of taurine in microorganisms: a primer in molecular biodiversity?
Adv. Exp. Med. Biol.
583
,
3
13
3
Fellman
,
J.H.
,
Roth
,
E.S.
,
Avedovech
,
N.A.
and
McCarthy
,
K.D.
(
1980
)
The metabolism of taurine to isethionate
.
Arch. Biochem. Biophys.
204
,
560
567
4
Krejcik
,
Z.
,
Hollemeyer
,
K.
,
Smits
,
T.H.
and
Cook
,
A.M.
(
2010
)
Isethionate formation from taurine in Chromohalobacter salexigens: purification of sulfoacetaldehyde reductase
.
Microbiology
156
,
1547
1555
5
Ladenstein
,
R.
,
Winberg
,
J.O.
and
Benach
,
J.
(
2008
)
Medium- and short-chain dehydrogenase/reductase gene and protein families: structure-function relationships in short-chain alcohol dehydrogenases
.
Cell. Mol. Life Sci.
65
,
3918
3935
6
Nallamsetty
,
S.
and
Waugh
,
D.S.
(
2007
)
A generic protocol for the expression and purification of recombinant proteins in Escherichia coli using a combinatorial His6-maltose binding protein fusion tag
.
Nat. Protoc.
2
,
383
391
7
Gibson
,
D.G.
,
Young
,
L.
,
Chuang
,
R.Y.
,
Venter
,
J.C.
,
Hutchison
, III,
C.A.
and
Smith
,
H.O.
(
2009
)
Enzymatic assembly of DNA molecules up to several hundred kilobases
.
Nat. Methods
6
,
343
345
8
Nettleship
,
J.E.
,
Brown
,
J.
,
Groves
,
M.R.
and
Geerlof
,
A.
(
2008
)
Methods for protein characterization by mass spectrometry, thermal shift (ThermoFluor) assay, and multiangle or static light scattering
.
Methods Mol. Biol.
426
,
299
318
9
Zhang
,
Q.
,
Li
,
Y.
,
Chen
,
D.
,
Yu
,
Y.
,
Duan
,
L.
,
Shen
,
B.
et al.  (
2011
)
Radical-mediated enzymatic carbon chain fragmentation-recombination
.
Nat. Chem. Biol.
7
,
154
160
10
Minor
,
W.
,
Cymborowski
,
M.
,
Otwinowski
,
Z.
and
Chruszcz
,
M.
(
2006
)
HKL-3000: the integration of data reduction and structure solution–from diffraction images to an initial model in minutes
.
Acta Crystallogr. D Biol. Crystallogr.
62
,
859
866
11
Adams
,
P.D.
,
Afonine
,
P.V.
,
Bunkóczi
,
G.
,
Chen
,
V.B.
,
Davis
,
I.W.
,
Echols
,
N.
et al.  (
2010
)
PHENIX: a comprehensive Python-based system for macromolecular structure solution
.
Acta Crystallogr. D Biol. Crystallogr.
66
,
213
221
12
Emsley
,
P.
and
Cowtan
,
K.
(
2004
)
Coot: model-building tools for molecular graphics
.
Acta Crystallogr. D Biol. Crystallogr.
60
,
2126
2132
13
Pettersen
,
E.F.
,
Goddard
,
T.D.
,
Huang
,
C.C.
,
Couch
,
G.S.
,
Greenblatt
,
D.M.
,
Meng
,
E.C.
et al.  (
2004
)
UCSF chimera–a visualization system for exploratory research and analysis
.
J. Comput. Chem.
25
,
1605
1612
14
Edgar
,
R.C.
(
2004
)
MUSCLE: a multiple sequence alignment method with reduced time and space complexity
.
BMC Bioinformatics
5
,
113
15
Crooks
,
G.E.
,
Hon
,
G.
,
Chandonia
,
J.M.
and
Brenner
,
S.E.
(
2004
)
Weblogo: a sequence logo generator
.
Genome Res.
14
,
1188
1190
16
Kavanagh
,
K.L.
,
Jörnvall
,
H.
,
Persson
,
B.
and
Oppermann
,
U.
(
2008
)
Medium- and short-chain dehydrogenase/reductase gene and protein families: the SDR superfamily: functional and structural diversity within a family of metabolic and regulatory enzymes
.
Cell. Mol. Life Sci.
65
,
3895
3906
17
von Rekowski
,
S.
,
Denger
,
K.
,
and Cook
,
K.
and
M
,
A.
(
2005
)
Isethionate as a product from taurine during nitrogen-limited growth of Klebsiella oxytoca TauN1
.
Arch. Microbiol.
183
,
325
330
18
Yamazawa
,
R.
,
Nakajima
,
Y.
,
Mushiake
,
K.
,
Yoshimoto
,
T.
and
Ito
,
K.
(
2011
)
Crystal structure of serine dehydrogenase from Escherichia coli: important role of the C-terminal region for closed-complex formation
.
J. Biochem.
149
,
701
712
19
MacKenzie
,
A.K.
,
Kershaw
,
N.J.
,
Hernandez
,
H.
,
Robinson
,
C.V.
,
Schofield
,
C.J.
and
Andersson
,
I.
(
2007
)
Clavulanic acid dehydrogenase: structural and biochemical analysis of the final step in the biosynthesis of the β-lactamase inhibitor clavulanic acid
.
Biochemistry
46
,
1523
1533
20
Nishioka
,
T.
,
Yasutake
,
Y.
,
Nishiya
,
Y.
and
Tamura
,
T.
(
2012
)
Structure-guided mutagenesis for the improvement of substrate specificity of Bacillus megaterium glucose 1-dehydrogenase IV
.
FEBS J.
279
,
3264
3275
21
Li
,
S.
,
Teng
,
X.
,
Su
,
L.
,
Mao
,
G.
,
Xu
,
Y.
,
Li
,
T.
et al.  (
2017
)
Structure and characterization of a NAD(P)H-dependent carbonyl reductase from Pseudomonas aeruginosa PAO1
.
FEBS Lett.
591
,
1785
1797
22
Kim
,
K.S.
,
Pelton
,
J.G.
,
Inwood
,
W.B.
,
Andersen
,
U.
,
Kustu
,
S.
and
Wemmer
,
D.E.
(
2010
)
The Rut pathway for pyrimidine degradation: novel chemistry and toxicity problems
.
J. Bacteriol.
192
,
4089
4102
23
Suzek
,
B.E.
,
Huang
,
H.
,
McGarvey
,
P.
,
Mazumder
,
R.
and
Wu
,
C.H.
(
2007
)
Uniref: comprehensive and non-redundant UniProt reference clusters
.
Bioinformatics
23
,
1282
1288
24
Zhao
,
S.
,
Sakai
,
A.
,
Zhang
,
X.
,
Vetting
,
M.W.
,
Kumar
,
R.
,
Hillerich
,
B.
et al.  (
2014
)
Prediction and characterization of enzymatic activities guided by sequence similarity and genome neighborhood networks
.
eLife
3
,
e03275
25
Gerlt
,
J.A.
,
Bouvier
,
J.T.
,
Davidson
,
D.B.
,
Imker
,
H.J.
,
Sadkhin
,
B.
,
Slater
,
D.R.
et al.  (
2015
)
Enzyme Function Initiative-Enzyme Similarity Tool (EFI-EST): a web tool for generating protein sequence similarity networks
.
Biochim. Biophys. Acta
1854
,
1019
1037
26
Shannon
,
P.
,
Markiel
,
A.
,
Ozier
,
O.
,
Baliga
,
N.S.
,
Wang
,
J.T.
,
Ramage
,
D.
et al.  (
2003
)
Cytoscape: a software environment for integrated models of biomolecular interaction networks
.
Genome Res.
13
,
2498
2504
27
Fujisawa
,
H.
,
Nagata
,
S.
and
Misono
,
H.
(
2003
)
Characterization of short-chain dehydrogenase/reductase homologues of Escherichia coli (YdfG) and Saccharomyces cerevisiae (YMR226C)
.
Biochim. Biophys. Acta
1645
,
89
94

Author notes

*

These authors contribute to this work equally.