- Email: [email protected]
Structural insights on mouse l -threonine dehydrogenase: A regulatory role of Arg180 in catalysis
Journal of Structural Biology 192 (2015) 510–518
Contents lists available at ScienceDirect
Journal of Structural Biology journal homepage: www.elsevier.com/locate/yjsbi
Structural insights on mouse L-threonine dehydrogenase: A regulatory role of Arg180 in catalysis Chao He a,⇑, Xianyu Huang a,1, Yanhong Liu a,1, Fudong Li b, Yang Yang b, Hongru Tao a, Chuanchun Han c, Chen Zhao b, Yazhong Xiao a, Yunyu Shi b a b c
Anhui Provincial Engineering Technology Research Center of Microorganisms and Biocatalysis and School of Life Sciences, Anhui University, Hefei, Anhui 230601, China Hefei National Laboratory for Physical Sciences at Microscale and School of Life Sciences, University of Science and Technology of China, Hefei, Anhui 230026, China Institute of Cancer Stem Cell, Dalian Medical University, Dalian, Liaoning 116044, China
a r t i c l e
i n f o
Article history: Received 3 July 2015 Received in revised form 1 October 2015 Accepted 19 October 2015 Available online 19 October 2015 Keywords: L-Threonine Nicotinamide adenine dinucleotide (NAD) Dehydrogenase Mouse Crystal structure Arg180 Enzyme kinetics Enzyme mutation
a b s t r a c t Mouse L-threonine dehydrogenase (mTDH), which belongs to the short-chain dehydrogenase/reductase (SDR) superfamily and mediates threonine catabolism, plays pivotal roles in both powerful biosynthesis and signaling in mouse stem cells and has a regulatory residue Arg180. Here we determined three crystal structures of mTDH: wild-type (WT) in the apo form; in complex with NAD+ and a substrate analog, glycerol, or with only NAD+; as well as the R180K variant with NAD+. This is the first description of a structure for mammalian SDR-type TDH. Structural comparison revealed the structural basis for SDR-type TDH catalysis remains strictly conserved in bacteria and mammals. Kinetic enzyme assays, and isothermal titration calorimetry (ITC) measurements indicated the R180K mutation has little effect on NAD+ binding affinity, whereas affects the substrate’s affinity for the enzyme. The crystal structure of R180K with NAD+, biochemical and spectroscopic studies suggested that the R180K mutant should bind NAD+ in a similar way and have a similar folding to the WT. However, the R180K variant may have difficulty adopting the closed form due to reduced interaction of residue 180 with a loop which connects a key position for mTDH switching between the closed and open forms in mTDH catalysis, and thereby exhibited a significantly decreased kcat/Km value toward the substrate, L-Thr. In sum, our results suggest that activity of GalE-like TDH can be regulated by remote interaction, such as hydrogen bonding and hydrophobic interaction around the Arg180 of mTDH. Ó 2015 Elsevier Inc. All rights reserved.
1. Introduction L-Threonine 3-dehydrogenase (EC 1.1.1.103, TDH) is an oxidoreductase, that specifically acting on the >CHOH group of a donor molecule with NAD+ or NADP+ as the acceptor. TDH can catalyze the NAD+-dependent dehydrogenation of the 30 carbon atom of L-threonine
(L-Thr). The product L-2-amino-3-ketobutyrate is nonenzymatically decomposes into aminoacetone and CO2 (Elliott, 1958) or converted to glycine and acetyl-CoA in a CoA
Abbreviations: TDH, L-threonine dehydrogenase; mTDH, L-threonine dehydrogenase from mouse; L-Thr, L-threonine; mESC, mouse embryonic stem cell; GalE, UDP-galactose 4-epimerase; SDR, short-chain dehydrogenase/reductase; MDR, medium-chain dehydrogenase/reductase; WT, wild-type; PRMT5, Protein arginine methyltransferase 5; ASU, asymmetric unit. ⇑ Corresponding author. E-mail address: [email protected] (C. He). 1 These authors contributed equally to this work. http://dx.doi.org/10.1016/j.jsb.2015.10.014 1047-8477/Ó 2015 Elsevier Inc. All rights reserved.
dependent reaction by 2-amino 3-ketobutyrate CoA ligase (Aoyama and Motokawa, 1981; McGilvray and Morris, 1969). Glycine then facilitates one-carbon metabolism via the glycine cleavage system to promote nucleotide synthesis and acetyl-CoA directly enters into the tricarboxylic acid cycle (Kaelin and McKnight, 2013). The reaction catalyzed by TDH is the first and rate-limiting step in the L-Thr degradation pathway. TDH is responsible for the bulk of threonine breakdown and is widely distributed in a broad range of species from bacteria to mammals (Aronson et al., 1989; Kazuoka et al., 2003; Wang et al., 2009). Recent studies revealed that mouse TDH (mTDH) mediatedthreonine catabolism plays a central role in the maintenance of mouse embryonic stem cells (mESCs) self-renewing and the regulation of mouse somatic cell reprogramming (Han et al., 2013; Kaelin and McKnight, 2013; Wang et al., 2009). It has been reported that mESCs depend exclusively on threonine and express exceptionally high levels of mTDH relative to differentiated cells (Wang et al., 2009). Specific chemical inhibitors of the mTDH
C. He et al. / Journal of Structural Biology 192 (2015) 510–518
enzyme selectively killed mouse ES cells but failed to affect the growth of differentiated cells (Alexander et al., 2011). Furthermore, mTDH-mediated threonine catabolism might influence the amount of S-adenosylmethionine, resulting in regulation of histone methylation in pluripotent stem cells (Shyh-Chang et al., 2013). It has also been indicated that mTDH induction also enhances reprogramming of mouse somatic cells into induced pluripotent stem cells. Protein arginine methyltransferase 5 (PRMT5) regulates mTDH enzyme activity through binding with mTDH and/or mediating specific arginine (Arg180) methylation of mTDH and functionally enhanced mTDH-facilitated reprogramming efficiency (Han et al., 2013). Moreover, in Trypanosoma brucei, TDH is one of enzymes bearing on threonine-derived acetate production for lipid biosynthesis (Millerioux et al., 2013). Inhibition of TDH and another protein participating in glucose-derived acetate production is lethal for the procyclic trypanosomes grown. Thus, TDH would be a target protein to design medicines of trypanosomiasis and mTDH is expected to be template to design the medicine which affects trypanosoma but not mammalian TDH in future. In spite of the pivotal role of mTDH, its three-dimensional structure has not been determined. Previous reports mainly focused on the catalytic properties and mechanisms of TDH enzymes from lower prokaryotic organisms. Structures of two types of TDH enzymes have been reported, which are all from bacteria or archaea. One type is classified within the medium-chain dehydrogenase/reductase (MDR) superfamily. MDR-type TDHs are similar to Zn-dependent alcohol dehydrogenases with a Rossmann-fold domain in their C-terminal regions and have members mainly in bacteria (Bowyer et al., 2009; Hedlund et al., 2010; Ishikawa et al., 2007; Persson et al., 2008). The other type is classified within the short-chain dehydrogenase/reductase (SDR) superfamily. Most SDR enzymes are homodimers with a NAD(P)+ binding Rossmann-fold domain in their Nterminal regions (Bray et al., 2009). To date, crystal structures of three SDR-type TDHs identified in the psychrophilic bacterium Flavobacterium frigidimaris KUC-1 (FfTDH) (Yoneda et al., 2010), the thermophilic archaeon Thermoplasma volcanium (TvTDH) (Yoneda et al., 2012) and Cupriavidus necator (CnTDH) (Nakano et al., 2014) have been reported, which are similar to UDPgalactose 4-epimerases (GalEs). Mammalian TDHs belong to the SDR superfamily from sequence. However, structural description of mammalian SDR-type TDH is still lacking. Moreover, it was reported that Arg180 of mTDH plays a regulatory role in catalysis. The enzyme activity decreased when Arg180 was mutated to a lysine and might be increased when Arg180 was methylated by PRMT5 (Han et al., 2013). Interestingly, this residue is not conserved in prokaryotic bacteria or archaea, but conserved in some eukaryotic organisms such as Caenorhabditis elegans, Drosophila melanogaster, zebrafish. It might represent a distinction between higher eukaryotic and lower prokaryotic SDR-type TDH enzymes. How Arg180 plays a role in catalysis remains uncertain. In the present study, we solved three crystal structures of the mTDH enzyme: mTDH wild-type (WT) in complex with NAD+ and a substrate analog, glycerol (mTDH/NAD+/glycerol), or only NAD+ (mTDH/NAD+); WT without NAD+ or any substrate (mTDH apo); and mTDH R180K mutant bound with NAD+ (mTDHR180K/NAD+). This is the first description of a structure for mammalian SDR-type TDH. Our structures showed that the structural basis for the catalytic mechanism of SDR-type TDHs is highly conserved in bacteria and mammals and mTDH also undergoes stepwise dynamical structural changes during catalysis. Moreover, Arg180 of mTDH, which is not conserved in prokaryotic genera, contributes to the stability of its surrounding region and is located distal to the active center. Kinetic enzyme assay and ITC, in combination with the NAD+-complexed R180K crystal structure and biochemical and spectroscopic methods indicated the R180K mutant
511
should bind NAD+ in a similar way and have a similar folding to the WT. However, the R180K mutant exhibited a decreased kcat/ Km value toward the substrate, L-Thr. It is possibly because the R180K variant may have difficulty adopting the closed form upon the substrate binding due to reduced interaction of this residue with its surrounding atoms. In summary, our observations suggest the Arg180 works as a remote switch to convert the closed and open forms during mTDH catalysis.
2. Materials and methods 2.1. Cloning, expression and purification The gene encoding mTDH without the predicted N-terminal 46amino-acid signal sequence was amplified from the mouse cDNA library by PCR and cloned into modified pET-28a (+) (Novagen) vectors, in which the thrombin protease sites were substituted for tobacco etch virus (TEV) cleavage sites. All of the mutants were generated by the Mutant Best kit (Takara). All clones were verified by DNA sequencing. The recombinant proteins were produced in Escherichia coli BL21(DE3) cells (Novagen) grown at 16 °C for 20 h. Protein samples for crystallization were purified using a Nichelating column (Qiagen), treated with TEV to remove His6 tag, and further purified on a Ni-chelating column before sizeexclusion chromatography using a Hiload 16/60 Superdex 200 column (GE healthcare) equilibrated with protein buffer A (20 mM Tris–HCl [pH 8.0], 0.5 M NaCl and 1 mM EDTA).
2.2. Crystallization, data collection and structure determination All the crystals were grown using the sitting drop vapor diffusion method at 10 °C. The mTDH WT enzyme in the apo form was crystallized by mixing equal volumes of 7.5 mg/ml protein and crystallization buffer 1 (0.1 M HEPES [pH 7.0], 10% w/v PEG4000, 10% v/v 2-propanol). Crystals of mTDH WT in complex with NAD+ were obtained by mixing an equal volume of crystallization buffer 2 (0.2 M NaCl, 0.1 M HEPES [pH 7.5], 25% w/v PEG3350) and 7 mg/ml protein pre-incubated with NAD+ at 1:2.5 M ratio. Crystals of mTDH R180K with NAD+ were prepared by incubating 7 mg/ml protein with NAD+ at 1:2.5 M ratio, followed by mixing an equal volume of crystallization buffer 3 (0.2 M trimethylamine N-oxide dihydrate, 0.1 M Tris–HCl [pH 8.5], 20% w/v PEGMME2000). The final concentration of NAD+ was 0.5 mM. All the crystals used for diffraction data collection were quickly soaked in reservoir solution containing 25% glycerol (v/v) as cryoprotectant and cooled to 100 K in a stream of nitrogen gas before they were mounted. Data sets of all the crystals were collected on beam line 17U (BL17U) at Shanghai Synchrotron Radiation Facility (SSRF). The data sets were integrated and scaled with HKL2000/HKL3000 (Minor et al., 2006) or iMosflm (Battye et al., 2011) and SCALA (Evans, 2006) in the CCP4 program suite (Winn et al., 2011). Initial phases of NAD+-complexed mTDH WT were solved by molecular replacement using the program Phaser (McCoy et al., 2007), employing the structure of FfTDH (PDB code: 2YY7) as a template. The structure of mTDH/NAD+ was used as the search model for the other two structures with Phaser. The models were further built and refined using Refmac5 (Murshudov et al., 1997) and COOT (Emsley et al., 2010) by manual model correction. Crystal diffraction data and refinement statistics are presented in Table 1. On the basis of the structure of glycerol, we modeled the L-Thr molecule into the active site of mTDH/NAD+ and then minimized the energy of the complex using SWISS-PDB VIEWER (http:// www.expasy.ch/spdbv/).
512
C. He et al. / Journal of Structural Biology 192 (2015) 510–518 Table 1 Statistics for data collection, phase determination and refinement.
a
mTDH/NAD+
mTDH apo
TDH-R180K/NAD+
Wavelength (Å) Space group
0.979 P22121
0.979 P1211
0.979 P22121
Cell parameters a, b, c (Å) a, b, c (°) Resolution (Å) Rmerge (%) I/rI Completeness (%) Average redundancy
74.15, 160.13, 194.30 90, 90, 90 38.86–2.80(2.95–2.80)a 10.1(58.7) 11.0(2.2) 92.7(88.1) 4.2(4.0)
94.45, 157.31, 162.34 90, 97.004, 90 50.00–2.65(2.74–2.65) 6.2(38.7) 13.9(2.0) 93.9(94.6) 2.1(2.0)
74.24, 154.05, 199.19 90, 90, 90 47.38–3.25(3.43–3.25) 13.4(56.6) 8.6(2.4) 97.4(94.3) 4.0(3.5)
Refinement No. reflections (overall) No. reflections (test set) Rwork/Rfree (%)
50,776 2700 21.55/26.44
118,620 6300 21.96/27.11
33,952 1772 22.76/28.51
No. non-H atoms Protein NAD+/glycerol Water
14,495 264/18 75
27,721 –/– 70
10,592 264/– –
B factors (Å2) Protein NAD+/glycerol Water
50.30 32.64/47.26 27.70
66.50 –/– 45.20
77.30 94.10/– –/–
R.m.s. deviations Bond lengths (Å) Bond angles (°)
0.0047 0.9710
0.0049 0.9511
0.0056 1.0599
Rampage plot % residues Favored Allowed Outliers PDB codes
96.7 3.3 0.0 4YR9
97.1 2.9 0.0 4YRA
93.9 5.5 0.6 4YRB
Values in parentheses are for highest-resolution shell.
2.3. Assay of enzyme activity toward L-Thr and NAD+ Enzyme activity was assayed spectrophotometrically using a Beckman DU800 spectrophotometer equipped with a thermostat. Enzyme activity toward L-Thr and NAD+ was measured using assay buffer 1 (0.1 M Tris–HCl [pH 7.5], 0.15 M NaCl, 10 mM NAD+, and 20.5–410 mM L-Thr) and assay buffer 2 (0.1 M Tris–HCl [pH 7.5], 0.15 M NaCl, 0.3125–7.5 mM NAD+ and 410 mM L-Thr). After preparing the assay buffer and enzyme solution separately and incubating the reaction mixture at 37 °C, the reaction was started by adding the substrate or coenzyme. The time-dependent appearance of NADH was monitored from the absorbance at 340 nm (extinction coefficient = 6.22 mM1 cm1). Initial velocities at different concentrations of L-Thr and NAD+ were calculated and plotted using the ORIGIN program, and the enzyme kinetics parameters kcat, Km,L-Thr and Km,NAD+ were calculated using the Michaelis–Menten equation and applying the non-linear leastsquares method. One unit (U) of activity was defined as the amount of enzyme producing 1.0 lmol of NADH per min at 37 °C under the standard assay conditions, and the specific activity was expressed in U (mg of protein)1. The protein concentration was determined using the Bradford method. To determine the effect of temperature on mTDH WT and R180K stability, the enzyme was incubated for 10 min at different temperature. After incubation, samples were immediately chilled on ice. The residual activity was determined using the standard assay method. Activity assays were performed in triplicate. 2.4. SDS–PAGE of crystals Several R180K crystals were stepwise washed in four droplets of reservoir solution and then dissolved in a droplet of protein buf-
fer A. The crystal wash and dissolving droplets were all loaded for SDS–PAGE. 2.5. Protease digestion of mTDH proteins The WT and R180K mutant mTDH proteins in protein buffer A were subjected to protease digestion at 25 °C for 30 min with trypsin at different enzyme concentrations indicated as protease/substrate (E/S) ratio (w/w). The digestion reaction was stopped by adding the SDS/PAGE loading buffer and immediately boiling the samples at 100 °C for 5 min. The polypeptide fragment products of digestion were analyzed by SDS/PAGE. Proteolysis experiments of WT and R180K proteins were also performed using a lessspecific enzyme proteinase K, incubated at 25 °C for 5 min with increasing protease/substrate (E/S) ratio (w/w). 2.6. Circular dichroism (CD) Far-UV CD spectra were recorded at 25 °C with a JASCO J-810 spectropolarimeter, at wavelengths between 200 and 280 nm, using a 0.1-cm path length cell and approximately 150-lg/ml protein sample in assay buffer 3 (50 mM NaH2PO4/Na2HPO4 [pH 7.5], 0.15 M NaCl). A buffer only reference was subtracted from each curve. 2.7. Intrinsic fluorescence measurements Intrinsic fluorescence measurements of proteins were recorded at 25 °C with a Shimadzu RF 5301 spectrofluorophotometer in a 0.1-cm path length quartz cell using approximately 50-lg/ml protein sample in assay buffer 3. Fluorescence emission spectra were
C. He et al. / Journal of Structural Biology 192 (2015) 510–518
recorded from 300 nm to 400 nm using an excitation wavelength of 280 nm. 2.8. Isothermal titration calorimetry ITC measurements were carried out using a MicroCal ITC200 calorimeter (MicroCal). The purified mTDH WT and R180K proteins were diluted with buffer A to a concentration of 0.04 and 0.08 mM, respectively. The NAD trihydrate powder (Sangon Biotech) was dissolved in buffer A to a concentration of 12 mM. A reference measurement (NAD+ injected into buffer A) was carried out to compensate for the heat of dilution of NAD+. The data were analyzed by fitting a one-site model with the MicroCal Origin software package. 2.9. Accession numbers The atomic coordinates and structure factors for wild-type mTDH bound with NAD+, in the apo form and mTDH R180K mutant with NAD+ have been deposited in the Protein Data Bank (http:// www.rcsb.org/) with accession codes 4YR9, 4YRA and 4YRB, respectively. 3. Results and discussion 3.1. Overall structure of mTDH The crystal structure of NAD+-complexed mTDH with or without a glycerol was determined at 2.80 Å-resolution. The space group is P22121 and there are three homodimers in the asymmetric unit (ASU). The crystal structure of mTDH apo was determined at a resolution of 2.65 Å. The space group is P1211 and the ASU consists of six homodimers. In the mTDH monomer structure, the Nterminal NAD+-binding domain consists of residues 55–226, 266– 294 and 331–351 (green) and the C-terminal catalytic domain is formed by residues 227–265, 295–330 and 352–373 (magenta) (Fig. 2A). The two subunits of mTDH dimer were closely associated through four helices (a5: residues V140-Y161 and a6: residues I194-Y213 in each subunit) and two loops (residues A173-T193 in each subunit) (Fig. 2A). The mTDH enzyme exhibits notable sequence identity with CnTDH (47%), FfTDH (47%) and TvTDH (41%) (Fig. 1), respectively, which implied conservative structural features in bacterial and mammalian SDR-type TDH enzymes. The overall structure of mTDH/NAD+/glycerol was basically the same as those of previously determined SDR-type TDHs: CnTDH (PDB ID: 3WMX), FfTDH (PDB ID: 2YY7) and TvTDH (PDB ID: 3A9W). After superposition, the root mean square deviations (RMSDs) of the Ca atoms of the monomeric structures of mTDH and CnTDH, FfTDH and TvTDH were about 1.5 Å of 286 residues, 1.2 Å of 290 residues and 1.5 Å of 287 residues, respectively. Assembly of the dimers was also the same except a minor difference in intersubunit contacts (Fig. 2B). Besides the hydrophobic interactions as the main force between dimerization helixes as in other reported structures, additional intersubunit hydrogenbonding pairs were observed in mTDH: Asp156–Arg1440 and Arg144–Asp1560 (the primes indicate the neighboring subunit in the dimer). Meanwhile, the intersubunit ion pairs formed by Arg132 and Arg134 on the dimerization loop of TvTDH are conserved in mTDH (corresponding to Arg190 and Arg192). In all, such hydrogen-bonding and ion pair network between each subunit should also be responsible for mTDH dimer formation and thermostability. Moreover, mTDH undergoes stepwise structural and flexibility changes upon binding of NAD+ and glycerol, in a way similar as
513
CnTDH showed. In the structure of mTDH with only NAD+ bound at the active site, the catalytic domain is more flexible in contrast with the NAD+-binding domain, having higher B-factor values (Fig. S1C) and being not superimposed well between each chain in regions of residues 231–237, 245–252 and 317–324 (corresponding to residues 180–186, 194–199 and 267–273 in CnTDH, respectively). By comparison, in the NAD+-binding domain, only the region of residues 132–138 (corresponding to residues 80–87 in CnTDH) was not superimposed well between each chain, having higher B-factor values and indicating moderate flexibility upon binding of NAD+ (Fig. S1C). In agreement, all the corresponding regions in NAD+-bound CnTDH were also assigned as flexible, which was indicated through molecular dynamics simulations (Nakano et al., 2014). Comparison of the structures in the apo and NAD+-bound forms (Fig. 2C) revealed residues 129–138 (corresponding to residues 77– 87 in CnTDH) adopts a more ordered form when NAD+ binds to mTDH. Residues 88–111 and 317–324 of mTDH moved in the direction toward the active center once NAD+ is bound (Fig. 2C). In particular, residues 88–111 (corresponding to residues 38–59 in CnTDH) becomes rigid via interactions between this region and NAD+ (Fig. 3A) as judged from its clear electron density and sharply decreased B-factor value from the apo to NAD+-bound form (Figs. 2C and S1B, C). Moreover, comparison of the structures in the only NAD+ and NAD+/glycerol simultaneously-bound forms showed the regions containing residues 132–138 and 317–324 moved a little more close to the substrate binding pocket once a glycerol, as a substrate analog, is bound (Fig. 2C). Surface representation of the NAD+-complexed mTDH structure with or without a glycerol (Fig. S1E and F) showed that the substrate binding cleft of mTDH/NAD+/glycerol seems to be narrower than that of mTDH/NAD+. Thus, mTDH should mainly adopt the ‘‘open” form as indicated in CnTDH (Nakano et al., 2014) when only NAD+ bound at the active site, leaving a cleft which should serve as a conduit for entry of the substrate and for exit of the product. Once NAD+ is bound, binding of L-Thr to the active site would favor the ‘‘closed” form, which would therefore be regarded as a Michaelis complex structure within SDR-type TDH. 3.2. Active center of mTDH The electron density corresponding to the NAD+ coenzyme bound within the active site was very clear and the ligand was placed in a well-superimposed orientation between different chains. Residues responsible for NAD+ binding in mTDH include Gln66, Asp88, Arg90, Asn106, Tyr195, Lys199, with their side chains forming hydrogen bonds with the NAD+ coenzyme (Fig. 3A). These residues are strictly conserved in GalE-like TDHs (Fig. 1). In the electron density map of NAD+-complexed mTDH, we noticed an extra density within the active site cavity for three mTDH peptide chains in the ASU of the crystal and found that a glycerol molecule could be modeled into that density as the crystal was quickly soaked in reservoir solution containing 25% glycerol (v/v) as cryoprotectant before diffraction data collection. In contrast with NAD+, the orientation of the glycerol coordinated within the active site of mTDH was not the same between the models in different chains including those in homologous FfTDH (Fig. 3B–E). However, in all orientations, part of the oxygen atoms of the glycerol hydroxyl groups are situated within hydrogen bonding distance of atoms in front of them, respectively (Fig. 3B and C). Meanwhile, these bonds together tightly hold the glycerol near the nicotinamide ring of NAD+. In a L-Thr binding model based on the structure of glycerol (Fig. 3D), the b-carbon (C3) position hydrogen of L-Thr is placed in front of the si-face of the nicotinamide ring, enabling pro-S-specific proton transfer to C4 of the
514
C. He et al. / Journal of Structural Biology 192 (2015) 510–518
Fig. 1. Multiple sequence alignment of mTDH with other three reported GalE-like TDHs was performed by using an online program T-Coffee (http://www.tcoffee.org/). FfTDH, TvTDH and CnTDH represent Flavobacterium frigidimaris TDH (PDB code 2YY7), Thermoplasma volcanium TDH (PDB code 3A9W) and Cupriavidus necator TDH (PDB code 3WMX), respectively. Arg180 of mTDH and equivalent positions in homologous proteins are indicated in a red asterisk. Blue cycles mark the conserved Thr/Ser-Tyr-Lys catalytic triad. Residues responsible for substrate recognition and NAD+ binding are labeled in yellow boxes and purple cycles, respectively.
NAD+, which is in agreement with those observed in other GalElike TDHs. Key residues involved in interacting with glycerol include Ser133, Thr170, Tyr195 and Thr237 (Fig. 3B and C), which are also highly conserved in SDR-type TDH enzymes (Fig. 1). We observed that the conserved Thr/Ser-Tyr-Lys catalytic triad and cofactor and substrate binding sites are superimposed well (Fig. 3E) between the structures of FfTDH, TvTDH, CnTDH and that of mTDH. Thus, mTDH should use a common catalytic mechanism among SDR superfamily (Yoneda et al., 2012) and the catalytic triad consists of Thr170-Tyr195-Lys199. The property of the Tyr residue to act as a catalytic acid/base is enhanced by an adjacent Lys residue that together with an oxidized, positively charged cofactor nicotinamide lowers the tyrosine hydroxyl pKa. The lysine e-amino group is also involved in nicotinamide ribose binding, whereas the role of the active site Ser/Thr residue is to stabilize and polarize the carbonyl substrate group (Yoneda et al., 2012). To sum up, the structural basis for catalytic function of GalE-like TDH enzymes remains strictly conserved in bacteria and mammals.
Arg180 of mTDH is in charge of stabilizing its surrounding region. This Arg180 is not conserved in bacteria and archaea. Equivalent loci in other determined SDR-type TDHs include Lys128 of FfTDH, Ala129 of CnTDH and Lys122 of TvTDH, which produce weaker interactions with the loops in their front respectively, due to their shorter side chains and fewer charged groups in comparison with an arginine (Fig. 4C–E). In order to confirm and find out how the R180K mutation affects the enzyme activity, we measured enzyme kinetic parameters of mTDH WT and R180K toward L-Thr and NAD+, respectively. All the enzyme activity assays were performed under physiological conditions at 37 °C. The enzyme kinetics for mTDH WT and mutants are plotted in Fig. 5, and the kinetics parameters are summarized in Table 2. The specific activity toward L-Thr of purified R180K was about 4.65 U mg1, which is much lower than that of WT (about 85.2 U mg1). This is consistent with the earlier observation that the R180K mutation decreased the enzyme activity by addition of mitochondrial protein extract instead of purified recombinant enzyme in the reaction assay. The Km value toward
3.3. The Arg180 site of mTDH and enzyme kinetics of mTDH WT and R180K variant
L-Thr for WT was 72.4 mM (Fig. 5C and Table 2), while the initial velocity toward L-Thr for the R180K variant did not reach maximum (Fig. 5D and Table 2) because of its greatly larger Km,L-Thr value and the solubility limit of L-Thr. The enzyme kinetics of R180K toward L-Thr was similar to the alanine mutants at residues T237 and S133 (Fig. 5F and H), which work as direct substrate binding sites corresponding to Ser74 and Thr179 of TvTDH (Yoneda et al., 2012), respectively. However, the Km,NAD+ for the R180K variant was 2.06 mM, close to the value for WT (1.67 mM) (Fig. 5A, B and Table 2). Moreover, ITC experiments to examine the binding affinities of the NAD+ coenzyme for R180K and WT, respectively, gave fitting dissociation constant (Kd) values for reference (Fig. S4 and Table S1), which were almost identical for R180K
It has been reported that the enzyme activity might be enhanced through methylation on Arg180 of mTDH, therefore we analyzed this site and compared it with equivalent positions of other GalE-like TDH enzymes. Arg180 locates on the back of the active center. It belongs to the NAD+ binding domain and points to the catalytic domain (Fig. 4A). The guanidine group of Arg180 forms hydrogen bonds with its adjacent oxygen atoms, which are located on the loop containing residues 329–333. Additional interaction is achieved by hydrophobic contacts between the Arg180 alkyl chain and the side chain of Met333 (Fig. 4B). Therefore,
C. He et al. / Journal of Structural Biology 192 (2015) 510–518
515
Fig. 2. Overall structure of mTDH. (A) The NAD+-binding and catalytic domains are shown in green and cyan, respectively. The adjacent subunit is shown in gray. The region involved in dimer formation is indicated. NAD+ (magenta) and glycerol (yellow) are shown as stick models. (B) A detailed comparison of intersubunit interactions for mTDH (green), FfTDH (blue), CnTDH (yellow) and TvTDH (pink). The networks of hydrogen bonds and ion pairs are shown as dotted lines. The adjacent subunit is shown in gray. (C) Structural comparison between mTDH apo, mTDH/NAD+ and mTDH/NAD+/glycerol. The apo, only NAD+-bound and NAD+/glycerol simultaneously-bound forms are colored in pink, gray and cyan, respectively. The electron densities for the regions of residues 132–138 and 231–237 of mTDH apo were too low to assign these residues on the map. The regions of residues 88–111 and 317–324 moved in the directions indicated by arrows. All the figures were prepared by using the software PyMOL (http://www.pymol.org/).
(0.47 mM) and WT (0.42 mM). Thus, the R180K mutation has little effect on NAD+ binding affinity, but may affect the enzyme’s substrate recognition. In order to further verify the influence of this mutation on substrate’s affinity for the enzyme, we constructed two more mutations, M333E and M333A, as Met333 and Arg180 are situated side by side and have interactions in the structure. Interestingly, the measured initial velocity of M333A toward L-Thr did not reach maximum (Fig. 5G and Table 2), presenting remarkably increased Km,L-Thr value compared with WT, whereas M333E has a Km,L-Thr value twice as large as WT (Fig. 5E and Table 2), suggesting a moderate decrease in binding affinity for the substrate, but not as obvious as M333A and R180K presented. This was possibly because replacing Met333 with a glutamic acid might weaken the
interaction with Arg180 only a little by introducing some electrostatic interaction with substitution of carboxyl for S-methyl. In sum, based on the structure of mTDH WT, this Arg180 is important, which is responsible for the stabilization of the local structure. Moreover, enzyme kinetics analysis indicated the R180K mutation should decrease the enzyme activity mainly through affecting the substrate recognition, but not NAD+ binding by the enzyme. In addition, we examined the effect of temperature on enzyme activity. The R180K mutant was observed to be less thermostable than WT, losing most of its activity at 50 °C in contrast with WT retaining most of its activity at 50 °C (Fig. S2). This suggested that hydrogen bonds and hydrophobic interaction around the Arg180 were responsible for the thermal stability of mTDH.
516
C. He et al. / Journal of Structural Biology 192 (2015) 510–518
Fig. 3. The active center of mTDH. Stereo view of NAD+ (magenta) (A), the glycerols (yellow) in different chains (B and C) and a proposed model of L-Thr binding (D) coordinated within the active site of mTDH. The Fo–Fc omit electron density maps for NAD+ and glycerols are shown at the 1.5r level. Residues that interact with NAD+ and glycerol are labeled. The networks of hydrogen bonds are shown as dotted lines. Water molecules are shown as cyan sphere models. (E) Superposition of the active site of mTDH/NAD+/glycerol (green) with those of TvTDH/NAD+/L-Thr (blue), FfTDH/NAD+/glycerol (orange) and CnTDH/NAD+/L-Thr (pink). Corresponding residues of mTDH are labeled.
Fig. 4. The Arg180 site of mTDH. (A) Arg180 (magenta), NAD+ (gray) and glycerol (yellow) in the overall structure of mTDH monomer are shown as stick models. The NAD+binding and catalytic domains are shown in green and cyan, respectively. Oxygen and nitrogen atoms are shown in red and blue, respectively. Stereo view of mTDH Arg180 site (B) and the equivalent positions of FfTDH (C), CnTDH (D) and TvTDH (E). The Fo–Fc omit electron density map for Arg180 of mTDH is shown at the 1.5r level. The networks of hydrogen bonds around the Arg180 and the corresponding residues in other determined TDH structures are shown as black dash lines.
C. He et al. / Journal of Structural Biology 192 (2015) 510–518
517
Fig. 5. Enzyme kinetics of mTDH WT and R180K variant. In vitro catalytic activities of purified recombinant mTDH WT and R180K variant toward NAD+ (A, B) and L-Thr (C, D); catalytic activities of four other variants M333E, T237A, M333A and S133A toward L-Thr (E–H).
Table 2 Kinetics parameters for WT and the mutants of mTDH. Variantsa
kcatb s1
Km,L-Thr mM
c
WT R180K M333E M333A T237A S133A
47.3 ± 2.2 n.d.d 46.7 ± 1.8 n.d. n.d. n.d.
72.4 ± 7.3 n.d. 151.9 ± 14.4 n.d. n.d. n.d.
Km,NAD+c mM
kcat/Km,L-Thr s1 mM1
1.67 ± 0.06 2.06 ± 0.18 n.d. n.d. n.d. n.d.
0.7 0.003e 0.3 0.006e 0.004e 0.004e
a Locations of mutated residues in the structure are shown in Fig. 4B. Recombinant enzymes were purified as described in Section 2. b kcat values of the mutants and WT were derived from the data shown in Fig. 5C and E. c Km,L-Thr and Km,NAD+ values represent the Michaelis constant value toward L-Thr and NAD+, respectively. d n.d., not determined. e The initial velocities of the variants (R180K, M333A, T237A and S133A) did not reach maximum value (Fig. 5D, F, G and H) because of their significantly larger Km,L-Thr values and the solubility limit of the substrate, L-Thr. Consequently, the kcat and Km,L-Thr values for these variants could not be determined accurately. Therefore, these parameters are annotated as ‘‘n.d.”, and the kcat/Km, L-Thr value is instead estimated.
3.4. Structural insights into the regulatory role of Arg180 in catalysis In order to explain the mechanism behind the regulatory role of Arg180 in catalysis, we purified the R180K mutant protein and solved its crystal structure bound with NAD+ at 3.2 Å-resolution. The space group is P22121, and there are three homodimers in the ASU. In this structure, the electron densities for the NAD+binding domain and the NAD+ coenzyme bound within the active site were clear (Fig. S3A). The main chain coordinates of NAD+binding domains of mTDH-R180K/NAD+ and mTDH-WT/NAD+ were basically the same, as well as bound NAD+ coenzymes within the active sites (Fig. S3A and B). Meanwhile, stereo view of NAD+ coordinated within the active site of the R180K variant (Fig. S3C) revealed R180K binds NAD+ with similar hydrogen bond networks to the WT, which further demonstrates that the R180K variant should bind NAD+ in a manner similar to that of WT. This was possibly because this mutated residue is relatively distant and points outwards from the NAD+ binding site. Unfortunately, the electron density corresponding to the catalytic domain of mTDH-R180K/NAD+ is unclear and even invisible
(Fig. S3 and Fig. S1D). We examined the low electron density for the catalytic domain of this mutant was not caused by protein degradation. Apparent molecular weights of the R180K mutant and WT crystals shown directly by SDS–PAGE (Fig. S3C) were completely the same, with no degradation bands observed. The structure of the R180K mutant protein has also been probed by limited proteolysis with trypsin and a less-specific enzyme proteinase K, which revealed this mutant is not more susceptible to protease activity compared with the WT (Fig. S5A and B). Moreover, the far-UV CD and tryptophan intrinsic fluorescence spectrums of the R180K, M333A and M333E mutants (Fig. S5C and D) were the same as those of WT, indicating that the overall structure, as well as the local environments of tryptophan residues within the protein should not undergo significant changes. In a word, these observations indicated that R180K should have a similar folding to the WT. However, heat instability and the poor electron densities for the catalytic domain of R180K may rather suggest the structural integrity of this mutant partly impaired. This is possibly because position and directed motion of the local region around the Arg180 may be changed due to these mutations. As we mentioned above, our crystal structures showed mTDH undergoes stepwise structural and flexibility changes upon binding of NAD+ and a substrate analog, glycerol, from the open to semiclosed form in a way similar as CnTDH showed. According to the reference (Nakano et al., 2014), three regions of CnTDH containing residues 80–87, 180–186 and 267–273 are key positions for the enzyme switching between the open and closed forms. Three variants in these regions, L80G, G184A and T186N of CnTDH, which would be difficult to form the closed form because of the steric repulsion, influence enzymatic activity toward the substrate, LThr. Corresponding regions in mTDH are residues 132–138, 231– 237 and 317–324, respectively (Fig. 2C). Although the Arg180 is not located in these regions, its mutation to a lysine may affect position and directed motion of its facing loop containing residues 329–333 (Figs. 4A and B and S3B), which connects the region of residues 317–324 (Fig. 2C), a potentially critical position for converting the open and closed forms in mTDH catalysis. Thus, it is possible that Arg180 of mTDH works as a ‘‘remote switch” for the enzyme changing between the open and closed forms. Consistent with this prediction, we observed that the enzyme kinetics of mTDH R180K is similar to that of the G184A and T186N mutants of CnTDH. The R180K variant might also have difficulty adopting the
518
C. He et al. / Journal of Structural Biology 192 (2015) 510–518
closed form upon the substrate binding, therefore exhibit significantly lower kcat/Km values toward L-Thr than the WT. We suspected when Arg180 of mTDH was methylated by PRMT5, the interaction forces around the Arg180 might be enhanced, which might facilitate the enzyme switching between the open and closed forms, and thus promoting the enzyme activity toward the substrate. To sum up, our results suggested that activity of GalE-like TDH could be regulated by remote interaction, such as hydrogen bonding and hydrophobic interaction around the Arg180 of mTDH.
4. Conclusions Here we described, for the first time, crystal structures of a mammalian SDR-type TDH from mouse. Based on the structures of mTDH WT in the apo, only NAD+ or NAD+/glycerol simultaneously-bound forms, and R180K mutant bound with NAD+, combined with enzyme kinetics analysis to assess the regulatory role of the Arg180 in catalysis, we showed that (1) the structural basis for catalytic function of GalE-like TDH enzymes is highly conserved in prokaryotes and mammals; (2) Arg180 of mTDH, which is not conserved in prokaryotes, contributes to the stability of its surrounding region and is located distal to the active site; (3) mutating Arg180 of mTDH to a lysine has little effect on NAD+ binding affinity for the enzyme and the R180K variant should bind NAD+ in a way similar to the WT; (4) the R180K and its related mutations should not sharply damage the mTDH overall structure, whereas affect the substrate’s affinity for the enzyme. In our structure, the side chain of Arg180 forms adequate hydrogen bonds and hydrophobic interaction with a loop which connects a key position for mTDH switching between the closed and open forms during catalysis. Thus, the R180K and its relevant variants may have difficulty in adopting the closed form and thereby exhibited lower kcat/ Km values toward the substrate than the WT. This idea is based on our experimental results in combination with what was revealed in another GalE-like TDH, CnTDH. In a word, this adjustable residue, the Arg180, which represents a distinction between higher eukaryotic and lower prokaryotic SDR-type TDH enzymes, may work as a remote switch to convert the closed and open forms in mTDH. Our results also suggest that activity of GalE-like TDH can be regulated by remote interaction, which furthers our understanding of the reaction mechanism for this type of enzymes.
Acknowledgements We thank Dr. Yiyang Jiang, Hongyu Bao and the staff at BL17U of Shanghai Synchrotron Radiation Facilities for their kind help in the X-ray data collection and processing. This work was supported by a grant from the Natural Science Foundation of Anhui Province (1408085QC49); the National Natural Science Foundation of China (No. 31400643); the Key Research Program of the Education Department of Anhui Province (KJ2014A008); the Introduction Project of Academic and Technology Leaders in Anhui University (01001770); and the Innovative Research Team Program of 211 Project in Anhui University.
Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.jsb.2015.10.014.
References Alexander, P.B., Wang, J., McKnight, S.L., 2011. Targeted killing of a mammalian cell based upon its specialized metabolic state. Proc. Natl. Acad. Sci. U.S.A. 108, 15828–15833. Aoyama, Y., Motokawa, Y., 1981. L-Threonine dehydrogenase of chicken liver. Purification, characterization, and physiological significance. J. Biol. Chem. 256, 12367–12373. Aronson, B.D., Somerville, R.L., Epperly, B.R., Dekker, E.E., 1989. The primary structure of Escherichia coli L-threonine dehydrogenase. J. Biol. Chem. 264, 5226–5232. Battye, T.G., Kontogiannis, L., Johnson, O., Powell, H.R., Leslie, A.G., 2011. IMOSFLM: a new graphical interface for diffraction-image processing with MOSFLM. Acta Crystallogr. D Biol. Crystallogr. 67, 271–281. Bowyer, A., Mikolajek, H., Stuart, J.W., Wood, S.P., Jamil, F., Rashid, N., Akhtar, M., Cooper, J.B., 2009. Structure and function of the L-threonine dehydrogenase (TkTDH) from the hyperthermophilic archaeon Thermococcus kodakaraensis. J. Struct. Biol. 168, 294–304. Bray, J.E., Marsden, B.D., Oppermann, U., 2009. The human short-chain dehydrogenase/reductase (SDR) superfamily: a bioinformatics summary. Chem. Biol. Interact. 178, 99–109. Elliott, W.H., 1958. A new threonine metabolite. Biochim. Biophys. Acta 29, 446– 447. Emsley, P., Lohkamp, B., Scott, W.G., Cowtan, K., 2010. Features and development of Coot. Acta Crystallogr. D Biol. Crystallogr. 66, 486–501. Evans, P., 2006. Scaling and assessment of data quality. Acta Crystallogr. D Biol. Crystallogr. 62, 72–82. Han, C., Gu, H., Wang, J., Lu, W., Mei, Y., Wu, M., 2013. Regulation of L-threonine dehydrogenase in somatic cell reprogramming. Stem Cells 31, 953–965. Hedlund, J., Jornvall, H., Persson, B., 2010. Subdivision of the MDR superfamily of medium-chain dehydrogenases/reductases through iterative hidden Markov model refinement. BMC Bioinformatics 11, 534. Ishikawa, K., Higashi, N., Nakamura, T., Matsuura, T., Nakagawa, A., 2007. The first crystal structure of L-threonine dehydrogenase. J. Mol. Biol. 366, 857–867. Kaelin Jr., W.G., McKnight, S.L., 2013. Influence of metabolism on epigenetics and disease. Cell 153, 56–69. Kazuoka, T., Takigawa, S., Arakawa, N., Hizukuri, Y., Muraoka, I., Oikawa, T., Soda, K., 2003. Novel psychrophilic and thermolabile L-threonine dehydrogenase from psychrophilic Cytophaga sp. strain KUC-1. J. Bacteriol. 185, 4483–4489. McCoy, A.J., Grosse-Kunstleve, R.W., Adams, P.D., Winn, M.D., Storoni, L.C., Read, R.J., 2007. Phaser crystallographic software. J. Appl. Crystallogr. 40, 658–674. McGilvray, D., Morris, J.G., 1969. Utilization of L-threonine by a species of Arthrobacter. A novel catabolic role for ‘‘aminoacetone synthase”. Biochem. J. 112, 657–671. Millerioux, Y., Ebikeme, C., Biran, M., Morand, P., Bouyssou, G., Vincent, I.M., Mazet, M., Riviere, L., Franconi, J.M., Burchmore, R.J., Moreau, P., Barrett, M.P., Bringaud, F., 2013. The threonine degradation pathway of the Trypanosoma brucei procyclic form: the main carbon source for lipid biosynthesis is under metabolic control. Mol. Microbiol. 90, 114–129. Minor, W., Cymborowski, M., Otwinowski, Z., 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. Murshudov, G.N., Vagin, A.A., Dodson, E.J., 1997. Refinement of macromolecular structures by the maximum-likelihood method. Acta Crystallogr. D Biol. Crystallogr. 53, 240–255. Nakano, S., Okazaki, S., Tokiwa, H., Asano, Y., 2014. Binding of NAD+ and L-threonine induces stepwise structural and flexibility changes in Cupriavidus necator Lthreonine dehydrogenase. J. Biol. Chem. 289, 10445–10454. Persson, B., Hedlund, J., Jornvall, H., 2008. Medium- and short-chain dehydrogenase/ reductase gene and protein families: the MDR superfamily. Cell. Mol. Life Sci. 65, 3879–3894. Shyh-Chang, N., Locasale, J.W., Lyssiotis, C.A., Zheng, Y., Teo, R.Y., Ratanasirintrawoot, S., Zhang, J., Onder, T., Unternaehrer, J.J., Zhu, H., Asara, J. M., Daley, G.Q., Cantley, L.C., 2013. Influence of threonine metabolism on Sadenosylmethionine and histone methylation. Science 339, 222–226. Wang, J., Alexander, P., Wu, L., Hammer, R., Cleaver, O., McKnight, S.L., 2009. Dependence of mouse embryonic stem cells on threonine catabolism. Science 325, 435–439. Winn, M.D., Ballard, C.C., Cowtan, K.D., Dodson, E.J., Emsley, P., Evans, P.R., Keegan, R.M., Krissinel, E.B., Leslie, A.G., McCoy, A., McNicholas, S.J., Murshudov, G.N., Pannu, N.S., Potterton, E.A., Powell, H.R., Read, R.J., Vagin, A., Wilson, K.S., 2011. Overview of the CCP4 suite and current developments. Acta Crystallogr. D Biol. Crystallogr. 67, 235–242. Yoneda, K., Sakuraba, H., Araki, T., Ohshima, T., 2012. Crystal structure of binary and ternary complexes of archaeal UDP-galactose 4-epimerase-like L-threonine dehydrogenase from Thermoplasma volcanium. J. Biol. Chem. 287, 12966–12974. Yoneda, K., Sakuraba, H., Muraoka, I., Oikawa, T., Ohshima, T., 2010. Crystal structure of UDP-galactose 4-epimerase-like L-threonine dehydrogenase belonging to the intermediate short-chain dehydrogenase-reductase superfamily. FEBS J. 277, 5124–5132.
Contents lists available at ScienceDirect
Journal of Structural Biology journal homepage: www.elsevier.com/locate/yjsbi
Structural insights on mouse L-threonine dehydrogenase: A regulatory role of Arg180 in catalysis Chao He a,⇑, Xianyu Huang a,1, Yanhong Liu a,1, Fudong Li b, Yang Yang b, Hongru Tao a, Chuanchun Han c, Chen Zhao b, Yazhong Xiao a, Yunyu Shi b a b c
Anhui Provincial Engineering Technology Research Center of Microorganisms and Biocatalysis and School of Life Sciences, Anhui University, Hefei, Anhui 230601, China Hefei National Laboratory for Physical Sciences at Microscale and School of Life Sciences, University of Science and Technology of China, Hefei, Anhui 230026, China Institute of Cancer Stem Cell, Dalian Medical University, Dalian, Liaoning 116044, China
a r t i c l e
i n f o
Article history: Received 3 July 2015 Received in revised form 1 October 2015 Accepted 19 October 2015 Available online 19 October 2015 Keywords: L-Threonine Nicotinamide adenine dinucleotide (NAD) Dehydrogenase Mouse Crystal structure Arg180 Enzyme kinetics Enzyme mutation
a b s t r a c t Mouse L-threonine dehydrogenase (mTDH), which belongs to the short-chain dehydrogenase/reductase (SDR) superfamily and mediates threonine catabolism, plays pivotal roles in both powerful biosynthesis and signaling in mouse stem cells and has a regulatory residue Arg180. Here we determined three crystal structures of mTDH: wild-type (WT) in the apo form; in complex with NAD+ and a substrate analog, glycerol, or with only NAD+; as well as the R180K variant with NAD+. This is the first description of a structure for mammalian SDR-type TDH. Structural comparison revealed the structural basis for SDR-type TDH catalysis remains strictly conserved in bacteria and mammals. Kinetic enzyme assays, and isothermal titration calorimetry (ITC) measurements indicated the R180K mutation has little effect on NAD+ binding affinity, whereas affects the substrate’s affinity for the enzyme. The crystal structure of R180K with NAD+, biochemical and spectroscopic studies suggested that the R180K mutant should bind NAD+ in a similar way and have a similar folding to the WT. However, the R180K variant may have difficulty adopting the closed form due to reduced interaction of residue 180 with a loop which connects a key position for mTDH switching between the closed and open forms in mTDH catalysis, and thereby exhibited a significantly decreased kcat/Km value toward the substrate, L-Thr. In sum, our results suggest that activity of GalE-like TDH can be regulated by remote interaction, such as hydrogen bonding and hydrophobic interaction around the Arg180 of mTDH. Ó 2015 Elsevier Inc. All rights reserved.
1. Introduction L-Threonine 3-dehydrogenase (EC 1.1.1.103, TDH) is an oxidoreductase, that specifically acting on the >CHOH group of a donor molecule with NAD+ or NADP+ as the acceptor. TDH can catalyze the NAD+-dependent dehydrogenation of the 30 carbon atom of L-threonine
(L-Thr). The product L-2-amino-3-ketobutyrate is nonenzymatically decomposes into aminoacetone and CO2 (Elliott, 1958) or converted to glycine and acetyl-CoA in a CoA
Abbreviations: TDH, L-threonine dehydrogenase; mTDH, L-threonine dehydrogenase from mouse; L-Thr, L-threonine; mESC, mouse embryonic stem cell; GalE, UDP-galactose 4-epimerase; SDR, short-chain dehydrogenase/reductase; MDR, medium-chain dehydrogenase/reductase; WT, wild-type; PRMT5, Protein arginine methyltransferase 5; ASU, asymmetric unit. ⇑ Corresponding author. E-mail address: [email protected] (C. He). 1 These authors contributed equally to this work. http://dx.doi.org/10.1016/j.jsb.2015.10.014 1047-8477/Ó 2015 Elsevier Inc. All rights reserved.
dependent reaction by 2-amino 3-ketobutyrate CoA ligase (Aoyama and Motokawa, 1981; McGilvray and Morris, 1969). Glycine then facilitates one-carbon metabolism via the glycine cleavage system to promote nucleotide synthesis and acetyl-CoA directly enters into the tricarboxylic acid cycle (Kaelin and McKnight, 2013). The reaction catalyzed by TDH is the first and rate-limiting step in the L-Thr degradation pathway. TDH is responsible for the bulk of threonine breakdown and is widely distributed in a broad range of species from bacteria to mammals (Aronson et al., 1989; Kazuoka et al., 2003; Wang et al., 2009). Recent studies revealed that mouse TDH (mTDH) mediatedthreonine catabolism plays a central role in the maintenance of mouse embryonic stem cells (mESCs) self-renewing and the regulation of mouse somatic cell reprogramming (Han et al., 2013; Kaelin and McKnight, 2013; Wang et al., 2009). It has been reported that mESCs depend exclusively on threonine and express exceptionally high levels of mTDH relative to differentiated cells (Wang et al., 2009). Specific chemical inhibitors of the mTDH
C. He et al. / Journal of Structural Biology 192 (2015) 510–518
enzyme selectively killed mouse ES cells but failed to affect the growth of differentiated cells (Alexander et al., 2011). Furthermore, mTDH-mediated threonine catabolism might influence the amount of S-adenosylmethionine, resulting in regulation of histone methylation in pluripotent stem cells (Shyh-Chang et al., 2013). It has also been indicated that mTDH induction also enhances reprogramming of mouse somatic cells into induced pluripotent stem cells. Protein arginine methyltransferase 5 (PRMT5) regulates mTDH enzyme activity through binding with mTDH and/or mediating specific arginine (Arg180) methylation of mTDH and functionally enhanced mTDH-facilitated reprogramming efficiency (Han et al., 2013). Moreover, in Trypanosoma brucei, TDH is one of enzymes bearing on threonine-derived acetate production for lipid biosynthesis (Millerioux et al., 2013). Inhibition of TDH and another protein participating in glucose-derived acetate production is lethal for the procyclic trypanosomes grown. Thus, TDH would be a target protein to design medicines of trypanosomiasis and mTDH is expected to be template to design the medicine which affects trypanosoma but not mammalian TDH in future. In spite of the pivotal role of mTDH, its three-dimensional structure has not been determined. Previous reports mainly focused on the catalytic properties and mechanisms of TDH enzymes from lower prokaryotic organisms. Structures of two types of TDH enzymes have been reported, which are all from bacteria or archaea. One type is classified within the medium-chain dehydrogenase/reductase (MDR) superfamily. MDR-type TDHs are similar to Zn-dependent alcohol dehydrogenases with a Rossmann-fold domain in their C-terminal regions and have members mainly in bacteria (Bowyer et al., 2009; Hedlund et al., 2010; Ishikawa et al., 2007; Persson et al., 2008). The other type is classified within the short-chain dehydrogenase/reductase (SDR) superfamily. Most SDR enzymes are homodimers with a NAD(P)+ binding Rossmann-fold domain in their Nterminal regions (Bray et al., 2009). To date, crystal structures of three SDR-type TDHs identified in the psychrophilic bacterium Flavobacterium frigidimaris KUC-1 (FfTDH) (Yoneda et al., 2010), the thermophilic archaeon Thermoplasma volcanium (TvTDH) (Yoneda et al., 2012) and Cupriavidus necator (CnTDH) (Nakano et al., 2014) have been reported, which are similar to UDPgalactose 4-epimerases (GalEs). Mammalian TDHs belong to the SDR superfamily from sequence. However, structural description of mammalian SDR-type TDH is still lacking. Moreover, it was reported that Arg180 of mTDH plays a regulatory role in catalysis. The enzyme activity decreased when Arg180 was mutated to a lysine and might be increased when Arg180 was methylated by PRMT5 (Han et al., 2013). Interestingly, this residue is not conserved in prokaryotic bacteria or archaea, but conserved in some eukaryotic organisms such as Caenorhabditis elegans, Drosophila melanogaster, zebrafish. It might represent a distinction between higher eukaryotic and lower prokaryotic SDR-type TDH enzymes. How Arg180 plays a role in catalysis remains uncertain. In the present study, we solved three crystal structures of the mTDH enzyme: mTDH wild-type (WT) in complex with NAD+ and a substrate analog, glycerol (mTDH/NAD+/glycerol), or only NAD+ (mTDH/NAD+); WT without NAD+ or any substrate (mTDH apo); and mTDH R180K mutant bound with NAD+ (mTDHR180K/NAD+). This is the first description of a structure for mammalian SDR-type TDH. Our structures showed that the structural basis for the catalytic mechanism of SDR-type TDHs is highly conserved in bacteria and mammals and mTDH also undergoes stepwise dynamical structural changes during catalysis. Moreover, Arg180 of mTDH, which is not conserved in prokaryotic genera, contributes to the stability of its surrounding region and is located distal to the active center. Kinetic enzyme assay and ITC, in combination with the NAD+-complexed R180K crystal structure and biochemical and spectroscopic methods indicated the R180K mutant
511
should bind NAD+ in a similar way and have a similar folding to the WT. However, the R180K mutant exhibited a decreased kcat/ Km value toward the substrate, L-Thr. It is possibly because the R180K variant may have difficulty adopting the closed form upon the substrate binding due to reduced interaction of this residue with its surrounding atoms. In summary, our observations suggest the Arg180 works as a remote switch to convert the closed and open forms during mTDH catalysis.
2. Materials and methods 2.1. Cloning, expression and purification The gene encoding mTDH without the predicted N-terminal 46amino-acid signal sequence was amplified from the mouse cDNA library by PCR and cloned into modified pET-28a (+) (Novagen) vectors, in which the thrombin protease sites were substituted for tobacco etch virus (TEV) cleavage sites. All of the mutants were generated by the Mutant Best kit (Takara). All clones were verified by DNA sequencing. The recombinant proteins were produced in Escherichia coli BL21(DE3) cells (Novagen) grown at 16 °C for 20 h. Protein samples for crystallization were purified using a Nichelating column (Qiagen), treated with TEV to remove His6 tag, and further purified on a Ni-chelating column before sizeexclusion chromatography using a Hiload 16/60 Superdex 200 column (GE healthcare) equilibrated with protein buffer A (20 mM Tris–HCl [pH 8.0], 0.5 M NaCl and 1 mM EDTA).
2.2. Crystallization, data collection and structure determination All the crystals were grown using the sitting drop vapor diffusion method at 10 °C. The mTDH WT enzyme in the apo form was crystallized by mixing equal volumes of 7.5 mg/ml protein and crystallization buffer 1 (0.1 M HEPES [pH 7.0], 10% w/v PEG4000, 10% v/v 2-propanol). Crystals of mTDH WT in complex with NAD+ were obtained by mixing an equal volume of crystallization buffer 2 (0.2 M NaCl, 0.1 M HEPES [pH 7.5], 25% w/v PEG3350) and 7 mg/ml protein pre-incubated with NAD+ at 1:2.5 M ratio. Crystals of mTDH R180K with NAD+ were prepared by incubating 7 mg/ml protein with NAD+ at 1:2.5 M ratio, followed by mixing an equal volume of crystallization buffer 3 (0.2 M trimethylamine N-oxide dihydrate, 0.1 M Tris–HCl [pH 8.5], 20% w/v PEGMME2000). The final concentration of NAD+ was 0.5 mM. All the crystals used for diffraction data collection were quickly soaked in reservoir solution containing 25% glycerol (v/v) as cryoprotectant and cooled to 100 K in a stream of nitrogen gas before they were mounted. Data sets of all the crystals were collected on beam line 17U (BL17U) at Shanghai Synchrotron Radiation Facility (SSRF). The data sets were integrated and scaled with HKL2000/HKL3000 (Minor et al., 2006) or iMosflm (Battye et al., 2011) and SCALA (Evans, 2006) in the CCP4 program suite (Winn et al., 2011). Initial phases of NAD+-complexed mTDH WT were solved by molecular replacement using the program Phaser (McCoy et al., 2007), employing the structure of FfTDH (PDB code: 2YY7) as a template. The structure of mTDH/NAD+ was used as the search model for the other two structures with Phaser. The models were further built and refined using Refmac5 (Murshudov et al., 1997) and COOT (Emsley et al., 2010) by manual model correction. Crystal diffraction data and refinement statistics are presented in Table 1. On the basis of the structure of glycerol, we modeled the L-Thr molecule into the active site of mTDH/NAD+ and then minimized the energy of the complex using SWISS-PDB VIEWER (http:// www.expasy.ch/spdbv/).
512
C. He et al. / Journal of Structural Biology 192 (2015) 510–518 Table 1 Statistics for data collection, phase determination and refinement.
a
mTDH/NAD+
mTDH apo
TDH-R180K/NAD+
Wavelength (Å) Space group
0.979 P22121
0.979 P1211
0.979 P22121
Cell parameters a, b, c (Å) a, b, c (°) Resolution (Å) Rmerge (%) I/rI Completeness (%) Average redundancy
74.15, 160.13, 194.30 90, 90, 90 38.86–2.80(2.95–2.80)a 10.1(58.7) 11.0(2.2) 92.7(88.1) 4.2(4.0)
94.45, 157.31, 162.34 90, 97.004, 90 50.00–2.65(2.74–2.65) 6.2(38.7) 13.9(2.0) 93.9(94.6) 2.1(2.0)
74.24, 154.05, 199.19 90, 90, 90 47.38–3.25(3.43–3.25) 13.4(56.6) 8.6(2.4) 97.4(94.3) 4.0(3.5)
Refinement No. reflections (overall) No. reflections (test set) Rwork/Rfree (%)
50,776 2700 21.55/26.44
118,620 6300 21.96/27.11
33,952 1772 22.76/28.51
No. non-H atoms Protein NAD+/glycerol Water
14,495 264/18 75
27,721 –/– 70
10,592 264/– –
B factors (Å2) Protein NAD+/glycerol Water
50.30 32.64/47.26 27.70
66.50 –/– 45.20
77.30 94.10/– –/–
R.m.s. deviations Bond lengths (Å) Bond angles (°)
0.0047 0.9710
0.0049 0.9511
0.0056 1.0599
Rampage plot % residues Favored Allowed Outliers PDB codes
96.7 3.3 0.0 4YR9
97.1 2.9 0.0 4YRA
93.9 5.5 0.6 4YRB
Values in parentheses are for highest-resolution shell.
2.3. Assay of enzyme activity toward L-Thr and NAD+ Enzyme activity was assayed spectrophotometrically using a Beckman DU800 spectrophotometer equipped with a thermostat. Enzyme activity toward L-Thr and NAD+ was measured using assay buffer 1 (0.1 M Tris–HCl [pH 7.5], 0.15 M NaCl, 10 mM NAD+, and 20.5–410 mM L-Thr) and assay buffer 2 (0.1 M Tris–HCl [pH 7.5], 0.15 M NaCl, 0.3125–7.5 mM NAD+ and 410 mM L-Thr). After preparing the assay buffer and enzyme solution separately and incubating the reaction mixture at 37 °C, the reaction was started by adding the substrate or coenzyme. The time-dependent appearance of NADH was monitored from the absorbance at 340 nm (extinction coefficient = 6.22 mM1 cm1). Initial velocities at different concentrations of L-Thr and NAD+ were calculated and plotted using the ORIGIN program, and the enzyme kinetics parameters kcat, Km,L-Thr and Km,NAD+ were calculated using the Michaelis–Menten equation and applying the non-linear leastsquares method. One unit (U) of activity was defined as the amount of enzyme producing 1.0 lmol of NADH per min at 37 °C under the standard assay conditions, and the specific activity was expressed in U (mg of protein)1. The protein concentration was determined using the Bradford method. To determine the effect of temperature on mTDH WT and R180K stability, the enzyme was incubated for 10 min at different temperature. After incubation, samples were immediately chilled on ice. The residual activity was determined using the standard assay method. Activity assays were performed in triplicate. 2.4. SDS–PAGE of crystals Several R180K crystals were stepwise washed in four droplets of reservoir solution and then dissolved in a droplet of protein buf-
fer A. The crystal wash and dissolving droplets were all loaded for SDS–PAGE. 2.5. Protease digestion of mTDH proteins The WT and R180K mutant mTDH proteins in protein buffer A were subjected to protease digestion at 25 °C for 30 min with trypsin at different enzyme concentrations indicated as protease/substrate (E/S) ratio (w/w). The digestion reaction was stopped by adding the SDS/PAGE loading buffer and immediately boiling the samples at 100 °C for 5 min. The polypeptide fragment products of digestion were analyzed by SDS/PAGE. Proteolysis experiments of WT and R180K proteins were also performed using a lessspecific enzyme proteinase K, incubated at 25 °C for 5 min with increasing protease/substrate (E/S) ratio (w/w). 2.6. Circular dichroism (CD) Far-UV CD spectra were recorded at 25 °C with a JASCO J-810 spectropolarimeter, at wavelengths between 200 and 280 nm, using a 0.1-cm path length cell and approximately 150-lg/ml protein sample in assay buffer 3 (50 mM NaH2PO4/Na2HPO4 [pH 7.5], 0.15 M NaCl). A buffer only reference was subtracted from each curve. 2.7. Intrinsic fluorescence measurements Intrinsic fluorescence measurements of proteins were recorded at 25 °C with a Shimadzu RF 5301 spectrofluorophotometer in a 0.1-cm path length quartz cell using approximately 50-lg/ml protein sample in assay buffer 3. Fluorescence emission spectra were
C. He et al. / Journal of Structural Biology 192 (2015) 510–518
recorded from 300 nm to 400 nm using an excitation wavelength of 280 nm. 2.8. Isothermal titration calorimetry ITC measurements were carried out using a MicroCal ITC200 calorimeter (MicroCal). The purified mTDH WT and R180K proteins were diluted with buffer A to a concentration of 0.04 and 0.08 mM, respectively. The NAD trihydrate powder (Sangon Biotech) was dissolved in buffer A to a concentration of 12 mM. A reference measurement (NAD+ injected into buffer A) was carried out to compensate for the heat of dilution of NAD+. The data were analyzed by fitting a one-site model with the MicroCal Origin software package. 2.9. Accession numbers The atomic coordinates and structure factors for wild-type mTDH bound with NAD+, in the apo form and mTDH R180K mutant with NAD+ have been deposited in the Protein Data Bank (http:// www.rcsb.org/) with accession codes 4YR9, 4YRA and 4YRB, respectively. 3. Results and discussion 3.1. Overall structure of mTDH The crystal structure of NAD+-complexed mTDH with or without a glycerol was determined at 2.80 Å-resolution. The space group is P22121 and there are three homodimers in the asymmetric unit (ASU). The crystal structure of mTDH apo was determined at a resolution of 2.65 Å. The space group is P1211 and the ASU consists of six homodimers. In the mTDH monomer structure, the Nterminal NAD+-binding domain consists of residues 55–226, 266– 294 and 331–351 (green) and the C-terminal catalytic domain is formed by residues 227–265, 295–330 and 352–373 (magenta) (Fig. 2A). The two subunits of mTDH dimer were closely associated through four helices (a5: residues V140-Y161 and a6: residues I194-Y213 in each subunit) and two loops (residues A173-T193 in each subunit) (Fig. 2A). The mTDH enzyme exhibits notable sequence identity with CnTDH (47%), FfTDH (47%) and TvTDH (41%) (Fig. 1), respectively, which implied conservative structural features in bacterial and mammalian SDR-type TDH enzymes. The overall structure of mTDH/NAD+/glycerol was basically the same as those of previously determined SDR-type TDHs: CnTDH (PDB ID: 3WMX), FfTDH (PDB ID: 2YY7) and TvTDH (PDB ID: 3A9W). After superposition, the root mean square deviations (RMSDs) of the Ca atoms of the monomeric structures of mTDH and CnTDH, FfTDH and TvTDH were about 1.5 Å of 286 residues, 1.2 Å of 290 residues and 1.5 Å of 287 residues, respectively. Assembly of the dimers was also the same except a minor difference in intersubunit contacts (Fig. 2B). Besides the hydrophobic interactions as the main force between dimerization helixes as in other reported structures, additional intersubunit hydrogenbonding pairs were observed in mTDH: Asp156–Arg1440 and Arg144–Asp1560 (the primes indicate the neighboring subunit in the dimer). Meanwhile, the intersubunit ion pairs formed by Arg132 and Arg134 on the dimerization loop of TvTDH are conserved in mTDH (corresponding to Arg190 and Arg192). In all, such hydrogen-bonding and ion pair network between each subunit should also be responsible for mTDH dimer formation and thermostability. Moreover, mTDH undergoes stepwise structural and flexibility changes upon binding of NAD+ and glycerol, in a way similar as
513
CnTDH showed. In the structure of mTDH with only NAD+ bound at the active site, the catalytic domain is more flexible in contrast with the NAD+-binding domain, having higher B-factor values (Fig. S1C) and being not superimposed well between each chain in regions of residues 231–237, 245–252 and 317–324 (corresponding to residues 180–186, 194–199 and 267–273 in CnTDH, respectively). By comparison, in the NAD+-binding domain, only the region of residues 132–138 (corresponding to residues 80–87 in CnTDH) was not superimposed well between each chain, having higher B-factor values and indicating moderate flexibility upon binding of NAD+ (Fig. S1C). In agreement, all the corresponding regions in NAD+-bound CnTDH were also assigned as flexible, which was indicated through molecular dynamics simulations (Nakano et al., 2014). Comparison of the structures in the apo and NAD+-bound forms (Fig. 2C) revealed residues 129–138 (corresponding to residues 77– 87 in CnTDH) adopts a more ordered form when NAD+ binds to mTDH. Residues 88–111 and 317–324 of mTDH moved in the direction toward the active center once NAD+ is bound (Fig. 2C). In particular, residues 88–111 (corresponding to residues 38–59 in CnTDH) becomes rigid via interactions between this region and NAD+ (Fig. 3A) as judged from its clear electron density and sharply decreased B-factor value from the apo to NAD+-bound form (Figs. 2C and S1B, C). Moreover, comparison of the structures in the only NAD+ and NAD+/glycerol simultaneously-bound forms showed the regions containing residues 132–138 and 317–324 moved a little more close to the substrate binding pocket once a glycerol, as a substrate analog, is bound (Fig. 2C). Surface representation of the NAD+-complexed mTDH structure with or without a glycerol (Fig. S1E and F) showed that the substrate binding cleft of mTDH/NAD+/glycerol seems to be narrower than that of mTDH/NAD+. Thus, mTDH should mainly adopt the ‘‘open” form as indicated in CnTDH (Nakano et al., 2014) when only NAD+ bound at the active site, leaving a cleft which should serve as a conduit for entry of the substrate and for exit of the product. Once NAD+ is bound, binding of L-Thr to the active site would favor the ‘‘closed” form, which would therefore be regarded as a Michaelis complex structure within SDR-type TDH. 3.2. Active center of mTDH The electron density corresponding to the NAD+ coenzyme bound within the active site was very clear and the ligand was placed in a well-superimposed orientation between different chains. Residues responsible for NAD+ binding in mTDH include Gln66, Asp88, Arg90, Asn106, Tyr195, Lys199, with their side chains forming hydrogen bonds with the NAD+ coenzyme (Fig. 3A). These residues are strictly conserved in GalE-like TDHs (Fig. 1). In the electron density map of NAD+-complexed mTDH, we noticed an extra density within the active site cavity for three mTDH peptide chains in the ASU of the crystal and found that a glycerol molecule could be modeled into that density as the crystal was quickly soaked in reservoir solution containing 25% glycerol (v/v) as cryoprotectant before diffraction data collection. In contrast with NAD+, the orientation of the glycerol coordinated within the active site of mTDH was not the same between the models in different chains including those in homologous FfTDH (Fig. 3B–E). However, in all orientations, part of the oxygen atoms of the glycerol hydroxyl groups are situated within hydrogen bonding distance of atoms in front of them, respectively (Fig. 3B and C). Meanwhile, these bonds together tightly hold the glycerol near the nicotinamide ring of NAD+. In a L-Thr binding model based on the structure of glycerol (Fig. 3D), the b-carbon (C3) position hydrogen of L-Thr is placed in front of the si-face of the nicotinamide ring, enabling pro-S-specific proton transfer to C4 of the
514
C. He et al. / Journal of Structural Biology 192 (2015) 510–518
Fig. 1. Multiple sequence alignment of mTDH with other three reported GalE-like TDHs was performed by using an online program T-Coffee (http://www.tcoffee.org/). FfTDH, TvTDH and CnTDH represent Flavobacterium frigidimaris TDH (PDB code 2YY7), Thermoplasma volcanium TDH (PDB code 3A9W) and Cupriavidus necator TDH (PDB code 3WMX), respectively. Arg180 of mTDH and equivalent positions in homologous proteins are indicated in a red asterisk. Blue cycles mark the conserved Thr/Ser-Tyr-Lys catalytic triad. Residues responsible for substrate recognition and NAD+ binding are labeled in yellow boxes and purple cycles, respectively.
NAD+, which is in agreement with those observed in other GalElike TDHs. Key residues involved in interacting with glycerol include Ser133, Thr170, Tyr195 and Thr237 (Fig. 3B and C), which are also highly conserved in SDR-type TDH enzymes (Fig. 1). We observed that the conserved Thr/Ser-Tyr-Lys catalytic triad and cofactor and substrate binding sites are superimposed well (Fig. 3E) between the structures of FfTDH, TvTDH, CnTDH and that of mTDH. Thus, mTDH should use a common catalytic mechanism among SDR superfamily (Yoneda et al., 2012) and the catalytic triad consists of Thr170-Tyr195-Lys199. The property of the Tyr residue to act as a catalytic acid/base is enhanced by an adjacent Lys residue that together with an oxidized, positively charged cofactor nicotinamide lowers the tyrosine hydroxyl pKa. The lysine e-amino group is also involved in nicotinamide ribose binding, whereas the role of the active site Ser/Thr residue is to stabilize and polarize the carbonyl substrate group (Yoneda et al., 2012). To sum up, the structural basis for catalytic function of GalE-like TDH enzymes remains strictly conserved in bacteria and mammals.
Arg180 of mTDH is in charge of stabilizing its surrounding region. This Arg180 is not conserved in bacteria and archaea. Equivalent loci in other determined SDR-type TDHs include Lys128 of FfTDH, Ala129 of CnTDH and Lys122 of TvTDH, which produce weaker interactions with the loops in their front respectively, due to their shorter side chains and fewer charged groups in comparison with an arginine (Fig. 4C–E). In order to confirm and find out how the R180K mutation affects the enzyme activity, we measured enzyme kinetic parameters of mTDH WT and R180K toward L-Thr and NAD+, respectively. All the enzyme activity assays were performed under physiological conditions at 37 °C. The enzyme kinetics for mTDH WT and mutants are plotted in Fig. 5, and the kinetics parameters are summarized in Table 2. The specific activity toward L-Thr of purified R180K was about 4.65 U mg1, which is much lower than that of WT (about 85.2 U mg1). This is consistent with the earlier observation that the R180K mutation decreased the enzyme activity by addition of mitochondrial protein extract instead of purified recombinant enzyme in the reaction assay. The Km value toward
3.3. The Arg180 site of mTDH and enzyme kinetics of mTDH WT and R180K variant
L-Thr for WT was 72.4 mM (Fig. 5C and Table 2), while the initial velocity toward L-Thr for the R180K variant did not reach maximum (Fig. 5D and Table 2) because of its greatly larger Km,L-Thr value and the solubility limit of L-Thr. The enzyme kinetics of R180K toward L-Thr was similar to the alanine mutants at residues T237 and S133 (Fig. 5F and H), which work as direct substrate binding sites corresponding to Ser74 and Thr179 of TvTDH (Yoneda et al., 2012), respectively. However, the Km,NAD+ for the R180K variant was 2.06 mM, close to the value for WT (1.67 mM) (Fig. 5A, B and Table 2). Moreover, ITC experiments to examine the binding affinities of the NAD+ coenzyme for R180K and WT, respectively, gave fitting dissociation constant (Kd) values for reference (Fig. S4 and Table S1), which were almost identical for R180K
It has been reported that the enzyme activity might be enhanced through methylation on Arg180 of mTDH, therefore we analyzed this site and compared it with equivalent positions of other GalE-like TDH enzymes. Arg180 locates on the back of the active center. It belongs to the NAD+ binding domain and points to the catalytic domain (Fig. 4A). The guanidine group of Arg180 forms hydrogen bonds with its adjacent oxygen atoms, which are located on the loop containing residues 329–333. Additional interaction is achieved by hydrophobic contacts between the Arg180 alkyl chain and the side chain of Met333 (Fig. 4B). Therefore,
C. He et al. / Journal of Structural Biology 192 (2015) 510–518
515
Fig. 2. Overall structure of mTDH. (A) The NAD+-binding and catalytic domains are shown in green and cyan, respectively. The adjacent subunit is shown in gray. The region involved in dimer formation is indicated. NAD+ (magenta) and glycerol (yellow) are shown as stick models. (B) A detailed comparison of intersubunit interactions for mTDH (green), FfTDH (blue), CnTDH (yellow) and TvTDH (pink). The networks of hydrogen bonds and ion pairs are shown as dotted lines. The adjacent subunit is shown in gray. (C) Structural comparison between mTDH apo, mTDH/NAD+ and mTDH/NAD+/glycerol. The apo, only NAD+-bound and NAD+/glycerol simultaneously-bound forms are colored in pink, gray and cyan, respectively. The electron densities for the regions of residues 132–138 and 231–237 of mTDH apo were too low to assign these residues on the map. The regions of residues 88–111 and 317–324 moved in the directions indicated by arrows. All the figures were prepared by using the software PyMOL (http://www.pymol.org/).
(0.47 mM) and WT (0.42 mM). Thus, the R180K mutation has little effect on NAD+ binding affinity, but may affect the enzyme’s substrate recognition. In order to further verify the influence of this mutation on substrate’s affinity for the enzyme, we constructed two more mutations, M333E and M333A, as Met333 and Arg180 are situated side by side and have interactions in the structure. Interestingly, the measured initial velocity of M333A toward L-Thr did not reach maximum (Fig. 5G and Table 2), presenting remarkably increased Km,L-Thr value compared with WT, whereas M333E has a Km,L-Thr value twice as large as WT (Fig. 5E and Table 2), suggesting a moderate decrease in binding affinity for the substrate, but not as obvious as M333A and R180K presented. This was possibly because replacing Met333 with a glutamic acid might weaken the
interaction with Arg180 only a little by introducing some electrostatic interaction with substitution of carboxyl for S-methyl. In sum, based on the structure of mTDH WT, this Arg180 is important, which is responsible for the stabilization of the local structure. Moreover, enzyme kinetics analysis indicated the R180K mutation should decrease the enzyme activity mainly through affecting the substrate recognition, but not NAD+ binding by the enzyme. In addition, we examined the effect of temperature on enzyme activity. The R180K mutant was observed to be less thermostable than WT, losing most of its activity at 50 °C in contrast with WT retaining most of its activity at 50 °C (Fig. S2). This suggested that hydrogen bonds and hydrophobic interaction around the Arg180 were responsible for the thermal stability of mTDH.
516
C. He et al. / Journal of Structural Biology 192 (2015) 510–518
Fig. 3. The active center of mTDH. Stereo view of NAD+ (magenta) (A), the glycerols (yellow) in different chains (B and C) and a proposed model of L-Thr binding (D) coordinated within the active site of mTDH. The Fo–Fc omit electron density maps for NAD+ and glycerols are shown at the 1.5r level. Residues that interact with NAD+ and glycerol are labeled. The networks of hydrogen bonds are shown as dotted lines. Water molecules are shown as cyan sphere models. (E) Superposition of the active site of mTDH/NAD+/glycerol (green) with those of TvTDH/NAD+/L-Thr (blue), FfTDH/NAD+/glycerol (orange) and CnTDH/NAD+/L-Thr (pink). Corresponding residues of mTDH are labeled.
Fig. 4. The Arg180 site of mTDH. (A) Arg180 (magenta), NAD+ (gray) and glycerol (yellow) in the overall structure of mTDH monomer are shown as stick models. The NAD+binding and catalytic domains are shown in green and cyan, respectively. Oxygen and nitrogen atoms are shown in red and blue, respectively. Stereo view of mTDH Arg180 site (B) and the equivalent positions of FfTDH (C), CnTDH (D) and TvTDH (E). The Fo–Fc omit electron density map for Arg180 of mTDH is shown at the 1.5r level. The networks of hydrogen bonds around the Arg180 and the corresponding residues in other determined TDH structures are shown as black dash lines.
C. He et al. / Journal of Structural Biology 192 (2015) 510–518
517
Fig. 5. Enzyme kinetics of mTDH WT and R180K variant. In vitro catalytic activities of purified recombinant mTDH WT and R180K variant toward NAD+ (A, B) and L-Thr (C, D); catalytic activities of four other variants M333E, T237A, M333A and S133A toward L-Thr (E–H).
Table 2 Kinetics parameters for WT and the mutants of mTDH. Variantsa
kcatb s1
Km,L-Thr mM
c
WT R180K M333E M333A T237A S133A
47.3 ± 2.2 n.d.d 46.7 ± 1.8 n.d. n.d. n.d.
72.4 ± 7.3 n.d. 151.9 ± 14.4 n.d. n.d. n.d.
Km,NAD+c mM
kcat/Km,L-Thr s1 mM1
1.67 ± 0.06 2.06 ± 0.18 n.d. n.d. n.d. n.d.
0.7 0.003e 0.3 0.006e 0.004e 0.004e
a Locations of mutated residues in the structure are shown in Fig. 4B. Recombinant enzymes were purified as described in Section 2. b kcat values of the mutants and WT were derived from the data shown in Fig. 5C and E. c Km,L-Thr and Km,NAD+ values represent the Michaelis constant value toward L-Thr and NAD+, respectively. d n.d., not determined. e The initial velocities of the variants (R180K, M333A, T237A and S133A) did not reach maximum value (Fig. 5D, F, G and H) because of their significantly larger Km,L-Thr values and the solubility limit of the substrate, L-Thr. Consequently, the kcat and Km,L-Thr values for these variants could not be determined accurately. Therefore, these parameters are annotated as ‘‘n.d.”, and the kcat/Km, L-Thr value is instead estimated.
3.4. Structural insights into the regulatory role of Arg180 in catalysis In order to explain the mechanism behind the regulatory role of Arg180 in catalysis, we purified the R180K mutant protein and solved its crystal structure bound with NAD+ at 3.2 Å-resolution. The space group is P22121, and there are three homodimers in the ASU. In this structure, the electron densities for the NAD+binding domain and the NAD+ coenzyme bound within the active site were clear (Fig. S3A). The main chain coordinates of NAD+binding domains of mTDH-R180K/NAD+ and mTDH-WT/NAD+ were basically the same, as well as bound NAD+ coenzymes within the active sites (Fig. S3A and B). Meanwhile, stereo view of NAD+ coordinated within the active site of the R180K variant (Fig. S3C) revealed R180K binds NAD+ with similar hydrogen bond networks to the WT, which further demonstrates that the R180K variant should bind NAD+ in a manner similar to that of WT. This was possibly because this mutated residue is relatively distant and points outwards from the NAD+ binding site. Unfortunately, the electron density corresponding to the catalytic domain of mTDH-R180K/NAD+ is unclear and even invisible
(Fig. S3 and Fig. S1D). We examined the low electron density for the catalytic domain of this mutant was not caused by protein degradation. Apparent molecular weights of the R180K mutant and WT crystals shown directly by SDS–PAGE (Fig. S3C) were completely the same, with no degradation bands observed. The structure of the R180K mutant protein has also been probed by limited proteolysis with trypsin and a less-specific enzyme proteinase K, which revealed this mutant is not more susceptible to protease activity compared with the WT (Fig. S5A and B). Moreover, the far-UV CD and tryptophan intrinsic fluorescence spectrums of the R180K, M333A and M333E mutants (Fig. S5C and D) were the same as those of WT, indicating that the overall structure, as well as the local environments of tryptophan residues within the protein should not undergo significant changes. In a word, these observations indicated that R180K should have a similar folding to the WT. However, heat instability and the poor electron densities for the catalytic domain of R180K may rather suggest the structural integrity of this mutant partly impaired. This is possibly because position and directed motion of the local region around the Arg180 may be changed due to these mutations. As we mentioned above, our crystal structures showed mTDH undergoes stepwise structural and flexibility changes upon binding of NAD+ and a substrate analog, glycerol, from the open to semiclosed form in a way similar as CnTDH showed. According to the reference (Nakano et al., 2014), three regions of CnTDH containing residues 80–87, 180–186 and 267–273 are key positions for the enzyme switching between the open and closed forms. Three variants in these regions, L80G, G184A and T186N of CnTDH, which would be difficult to form the closed form because of the steric repulsion, influence enzymatic activity toward the substrate, LThr. Corresponding regions in mTDH are residues 132–138, 231– 237 and 317–324, respectively (Fig. 2C). Although the Arg180 is not located in these regions, its mutation to a lysine may affect position and directed motion of its facing loop containing residues 329–333 (Figs. 4A and B and S3B), which connects the region of residues 317–324 (Fig. 2C), a potentially critical position for converting the open and closed forms in mTDH catalysis. Thus, it is possible that Arg180 of mTDH works as a ‘‘remote switch” for the enzyme changing between the open and closed forms. Consistent with this prediction, we observed that the enzyme kinetics of mTDH R180K is similar to that of the G184A and T186N mutants of CnTDH. The R180K variant might also have difficulty adopting the
518
C. He et al. / Journal of Structural Biology 192 (2015) 510–518
closed form upon the substrate binding, therefore exhibit significantly lower kcat/Km values toward L-Thr than the WT. We suspected when Arg180 of mTDH was methylated by PRMT5, the interaction forces around the Arg180 might be enhanced, which might facilitate the enzyme switching between the open and closed forms, and thus promoting the enzyme activity toward the substrate. To sum up, our results suggested that activity of GalE-like TDH could be regulated by remote interaction, such as hydrogen bonding and hydrophobic interaction around the Arg180 of mTDH.
4. Conclusions Here we described, for the first time, crystal structures of a mammalian SDR-type TDH from mouse. Based on the structures of mTDH WT in the apo, only NAD+ or NAD+/glycerol simultaneously-bound forms, and R180K mutant bound with NAD+, combined with enzyme kinetics analysis to assess the regulatory role of the Arg180 in catalysis, we showed that (1) the structural basis for catalytic function of GalE-like TDH enzymes is highly conserved in prokaryotes and mammals; (2) Arg180 of mTDH, which is not conserved in prokaryotes, contributes to the stability of its surrounding region and is located distal to the active site; (3) mutating Arg180 of mTDH to a lysine has little effect on NAD+ binding affinity for the enzyme and the R180K variant should bind NAD+ in a way similar to the WT; (4) the R180K and its related mutations should not sharply damage the mTDH overall structure, whereas affect the substrate’s affinity for the enzyme. In our structure, the side chain of Arg180 forms adequate hydrogen bonds and hydrophobic interaction with a loop which connects a key position for mTDH switching between the closed and open forms during catalysis. Thus, the R180K and its relevant variants may have difficulty in adopting the closed form and thereby exhibited lower kcat/ Km values toward the substrate than the WT. This idea is based on our experimental results in combination with what was revealed in another GalE-like TDH, CnTDH. In a word, this adjustable residue, the Arg180, which represents a distinction between higher eukaryotic and lower prokaryotic SDR-type TDH enzymes, may work as a remote switch to convert the closed and open forms in mTDH. Our results also suggest that activity of GalE-like TDH can be regulated by remote interaction, which furthers our understanding of the reaction mechanism for this type of enzymes.
Acknowledgements We thank Dr. Yiyang Jiang, Hongyu Bao and the staff at BL17U of Shanghai Synchrotron Radiation Facilities for their kind help in the X-ray data collection and processing. This work was supported by a grant from the Natural Science Foundation of Anhui Province (1408085QC49); the National Natural Science Foundation of China (No. 31400643); the Key Research Program of the Education Department of Anhui Province (KJ2014A008); the Introduction Project of Academic and Technology Leaders in Anhui University (01001770); and the Innovative Research Team Program of 211 Project in Anhui University.
Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.jsb.2015.10.014.
References Alexander, P.B., Wang, J., McKnight, S.L., 2011. Targeted killing of a mammalian cell based upon its specialized metabolic state. Proc. Natl. Acad. Sci. U.S.A. 108, 15828–15833. Aoyama, Y., Motokawa, Y., 1981. L-Threonine dehydrogenase of chicken liver. Purification, characterization, and physiological significance. J. Biol. Chem. 256, 12367–12373. Aronson, B.D., Somerville, R.L., Epperly, B.R., Dekker, E.E., 1989. The primary structure of Escherichia coli L-threonine dehydrogenase. J. Biol. Chem. 264, 5226–5232. Battye, T.G., Kontogiannis, L., Johnson, O., Powell, H.R., Leslie, A.G., 2011. IMOSFLM: a new graphical interface for diffraction-image processing with MOSFLM. Acta Crystallogr. D Biol. Crystallogr. 67, 271–281. Bowyer, A., Mikolajek, H., Stuart, J.W., Wood, S.P., Jamil, F., Rashid, N., Akhtar, M., Cooper, J.B., 2009. Structure and function of the L-threonine dehydrogenase (TkTDH) from the hyperthermophilic archaeon Thermococcus kodakaraensis. J. Struct. Biol. 168, 294–304. Bray, J.E., Marsden, B.D., Oppermann, U., 2009. The human short-chain dehydrogenase/reductase (SDR) superfamily: a bioinformatics summary. Chem. Biol. Interact. 178, 99–109. Elliott, W.H., 1958. A new threonine metabolite. Biochim. Biophys. Acta 29, 446– 447. Emsley, P., Lohkamp, B., Scott, W.G., Cowtan, K., 2010. Features and development of Coot. Acta Crystallogr. D Biol. Crystallogr. 66, 486–501. Evans, P., 2006. Scaling and assessment of data quality. Acta Crystallogr. D Biol. Crystallogr. 62, 72–82. Han, C., Gu, H., Wang, J., Lu, W., Mei, Y., Wu, M., 2013. Regulation of L-threonine dehydrogenase in somatic cell reprogramming. Stem Cells 31, 953–965. Hedlund, J., Jornvall, H., Persson, B., 2010. Subdivision of the MDR superfamily of medium-chain dehydrogenases/reductases through iterative hidden Markov model refinement. BMC Bioinformatics 11, 534. Ishikawa, K., Higashi, N., Nakamura, T., Matsuura, T., Nakagawa, A., 2007. The first crystal structure of L-threonine dehydrogenase. J. Mol. Biol. 366, 857–867. Kaelin Jr., W.G., McKnight, S.L., 2013. Influence of metabolism on epigenetics and disease. Cell 153, 56–69. Kazuoka, T., Takigawa, S., Arakawa, N., Hizukuri, Y., Muraoka, I., Oikawa, T., Soda, K., 2003. Novel psychrophilic and thermolabile L-threonine dehydrogenase from psychrophilic Cytophaga sp. strain KUC-1. J. Bacteriol. 185, 4483–4489. McCoy, A.J., Grosse-Kunstleve, R.W., Adams, P.D., Winn, M.D., Storoni, L.C., Read, R.J., 2007. Phaser crystallographic software. J. Appl. Crystallogr. 40, 658–674. McGilvray, D., Morris, J.G., 1969. Utilization of L-threonine by a species of Arthrobacter. A novel catabolic role for ‘‘aminoacetone synthase”. Biochem. J. 112, 657–671. Millerioux, Y., Ebikeme, C., Biran, M., Morand, P., Bouyssou, G., Vincent, I.M., Mazet, M., Riviere, L., Franconi, J.M., Burchmore, R.J., Moreau, P., Barrett, M.P., Bringaud, F., 2013. The threonine degradation pathway of the Trypanosoma brucei procyclic form: the main carbon source for lipid biosynthesis is under metabolic control. Mol. Microbiol. 90, 114–129. Minor, W., Cymborowski, M., Otwinowski, Z., 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. Murshudov, G.N., Vagin, A.A., Dodson, E.J., 1997. Refinement of macromolecular structures by the maximum-likelihood method. Acta Crystallogr. D Biol. Crystallogr. 53, 240–255. Nakano, S., Okazaki, S., Tokiwa, H., Asano, Y., 2014. Binding of NAD+ and L-threonine induces stepwise structural and flexibility changes in Cupriavidus necator Lthreonine dehydrogenase. J. Biol. Chem. 289, 10445–10454. Persson, B., Hedlund, J., Jornvall, H., 2008. Medium- and short-chain dehydrogenase/ reductase gene and protein families: the MDR superfamily. Cell. Mol. Life Sci. 65, 3879–3894. Shyh-Chang, N., Locasale, J.W., Lyssiotis, C.A., Zheng, Y., Teo, R.Y., Ratanasirintrawoot, S., Zhang, J., Onder, T., Unternaehrer, J.J., Zhu, H., Asara, J. M., Daley, G.Q., Cantley, L.C., 2013. Influence of threonine metabolism on Sadenosylmethionine and histone methylation. Science 339, 222–226. Wang, J., Alexander, P., Wu, L., Hammer, R., Cleaver, O., McKnight, S.L., 2009. Dependence of mouse embryonic stem cells on threonine catabolism. Science 325, 435–439. Winn, M.D., Ballard, C.C., Cowtan, K.D., Dodson, E.J., Emsley, P., Evans, P.R., Keegan, R.M., Krissinel, E.B., Leslie, A.G., McCoy, A., McNicholas, S.J., Murshudov, G.N., Pannu, N.S., Potterton, E.A., Powell, H.R., Read, R.J., Vagin, A., Wilson, K.S., 2011. Overview of the CCP4 suite and current developments. Acta Crystallogr. D Biol. Crystallogr. 67, 235–242. Yoneda, K., Sakuraba, H., Araki, T., Ohshima, T., 2012. Crystal structure of binary and ternary complexes of archaeal UDP-galactose 4-epimerase-like L-threonine dehydrogenase from Thermoplasma volcanium. J. Biol. Chem. 287, 12966–12974. Yoneda, K., Sakuraba, H., Muraoka, I., Oikawa, T., Ohshima, T., 2010. Crystal structure of UDP-galactose 4-epimerase-like L-threonine dehydrogenase belonging to the intermediate short-chain dehydrogenase-reductase superfamily. FEBS J. 277, 5124–5132.