- Email: [email protected]
Structure of an endogalactosylceramidase from Rhodococcus hoagii 103S reveals the molecular basis of its substrate specificity
Journal of Structural Biology xxx (xxxx) xxx–xxx
Contents lists available at ScienceDirect
Journal of Structural Biology journal homepage: www.elsevier.com/locate/yjsbi
Structure of an endogalactosylceramidase from Rhodococcus hoagii 103S reveals the molecular basis of its substrate specificity Liuqing Chena,b, Qing Changa, Quande Yana, Guangyu Yanga, Yong Zhanga, , Yan Fenga, ⁎
a b
⁎
State Key Laboratory of Microbial Metabolism, School of Life Sciences and Biotechnology, Shanghai Jiao Tong University, Shanghai 200240, China Research Center for Computer-Aided Drug Discovery, Shenzhen Institutes of Advanced Technology, Chinese Academy of Sciences, Shenzhen 518055, China
ARTICLE INFO
ABSTRACT
Keywords: Glycoside hydrolase Endogalactosylceramidase Substrate specificity Crystal structure
Endoglycoceramidases (EGCs) are family 5 glycoside hydrolases that catalyze hydrolysis of the glycosidic linkages between the oligosaccharide and ceramide moieties of glycosphingolipids. Three orthologs of EGCs, each with distinct substrate specificity, have been identified to date, including EGC-I, EGC-II, and EGALC. Although the structures of EGC-I and EGC-II have been reported, the substrate preference mechanism of EGC enzymes remains unclear. Here, we determined the crystal structure of EGALC from Rhodococcus hoagii 103S at a resolution of 1.20 Å. Distinct from EGC-I and EGC-II, which both have a tunnel-like substrate binding site, the structure of EGALC accommodates substrates in a long groove. Further, the oligosaccharide-binding region of groove could be divided into two small pockets that separately bind to the Gal1 and to the Gal3/Gla3 present in 6-gala series substrates. Structural and sequence comparisons of EGC enzymes revealed that the conformation and length of their Nβ8-Lα1 regions are crucial in determining the architectures of their specific substrate binding sites. Importantly, molecular docking analyses indicate that the substrate specificity of each EGC is mainly derived from the complementarity of its active site groove/tunnel with substrates adopting particular conformations. Our study provide insights for understanding the catalytic mechanism of EGALC, which will help protein engineering for improving the substrate preference and catalytic efficiency of EGC enzymes toward important glycosphingolipid substrates.
1. Introduction Glycosphingolipids (GSLs) are ubiquitous cell membrane glycolipids consisting of glycosidically bound carbohydrates and ceramide moieties. GSLs play key roles in acting as secondary messengers or modulators of signal transduction by affecting several events, ranging from cell adhesion, cell growth, to regulation of apoptosis (Regina Todeschini and Hakomori, 2008). Abnormal neuronal accumulation of GSLs, including of galactosylceramide (GalCer) and glucosylceramide (GlcCer), has been reported to cause Krabbe and Gaucher disease (Lingwood, 2011). Due to their strong bioactivity, several GSLs also have been developed into therapeutic molecules against cancer, such as lyso-GM3 (Murozuka et al., 2007). The vast majority of GSLs are classified into distinct “series” according to their saccharide moiety composition. Ganglio, globo, and neolacto series mainly occur in vertebrates, whereas mollu and arthro series predominate in invertebrates (Schnaar et al., 2015). In mammals, GSLs can be degraded by exo-glycosidases, which mainly cleave monosaccharides from the non-reducing end. By contrast,
⁎
endoglycoceramidases (EGCs) from non-mammalian cells have been shown to cleave intact oligosaccharide moieties from GSLs (Li et al., 1987; Ito and Yamagata, 1989; Ashida et al., 1992; Horibata et al., 2000; Horibata et al., 2004). Three different EGC enzymes have been identified in Rhodococccus sp. M−750, including EGC-I, EGC-II, and EGC-III (this was later renamed as endogalactosylceramidase, EGALC) (Ito and Yamagata, 1989). Biochemical characterization shown that EGC-I has the broadest substrate scope amongst these three enzymes; it can hydrolyze ganglio-, lacto-, globo-series GSLs and fucosyl-GM1a (Ishibashi et al., 2012), while EGC-II mainly hydrolyzes ganglio- and globo-series GSLs (Ito and Yamagata, 1989; Izu et al., 1997). In contrast, EGALC only hydrolyzes 6-gala series GSLs, each of which contains a common R-Galβ1-6GalβCer structure that is not recognized by EGC-I or EGC-II (Ishibashi et al., 2007a,b). EGCs can be used for both GSLs analysis and semi-synthesis. In fact, both EGC-II and EGALC can be converted to glycolsynthase by mutation of the catalytic nucleophile residue, which has been successfully used for the synthesis of lyso-GM3 ganglioside and psychosine (Vaughan
Corresponding authors. E-mail addresses: [email protected] (Y. Zhang), [email protected] (Y. Feng).
https://doi.org/10.1016/j.jsb.2019.09.010 Received 2 May 2019; Received in revised form 20 September 2019; Accepted 22 September 2019 1047-8477/ © 2019 Elsevier Inc. All rights reserved.
Please cite this article as: Liuqing Chen, et al., Journal of Structural Biology, https://doi.org/10.1016/j.jsb.2019.09.010
Journal of Structural Biology xxx (xxxx) xxx–xxx
L. Chen, et al.
et al., 2006; Goddard-Borger et al., 2016). Further, engineering EGC-II has improved its synthesis efficiency (Hancock et al., 2009). EGCs are also able to catalyze transglycosylation reactions. For instance, EGALC has been applied to catalyze transglycosylation reaction to product fluorescent glycosylsphingolipids and neoglycoconjugates (Y. Ishibashi et al., 2007). Notably, an EGC from jellyfish has both transglycosylation and reverse hydrolysis activity, which can be used to condense Lac and Cer to form LacCer (Horibata et al., 2001). Naturally, the yield of this transglycosylation reaction is low, as the existed so-called secondary hydrolysis. As the first reported EGCs structure, the EGC-II structure uncovered that the active site residues of this enzyme are located at a wide funnellike polar cavity and a narrow hydrophobic tunnel for binding, respectively, to the oligosaccharide and ceramide moieties of GSLs (Caines et al., 2007). Previously, we reported the crystal structure EGCI from R. equi 103S (Han et al., 2017). Even though we have the structures of EGC-I and EGC-II, the substrate recognition mechanism of EGCs remains unclear. Further, the structure of 6-gala series GSLs are very different from ganglio-series GSLs, and the sequence identity between EGALC and those structure solved EGCs is low (< 30%), suggesting that EGALC may have a distinct substrate recognition feature. Here, we draw on structural and biochemical data to show that EGALC contains a long ‘boat’ shape surface groove for its substrates binding and hydrolysis. By exploring active site interactions during GSL substrates docking, we suggest a model for EGALC substrate recognition that is quite different from the previously reported binding employed by EGC-I and EGC-II. Indeed, our structural and computational results shed light on how EGALC generates its strict substrate specificity toward 6-gala series GSLs. EGALC uses its linear substrate binding pocket to sterically exclude GSLs with branched sugar moieties, thereby facilitating its specific recognition of approximately linear substrates. Our study offers insight into mechanistic details of this EGALC catalysis and guidance for future protein engineering for improving catalytic properties of ECG enzymes.
procedure. Then proteins were exchanged into buffer A using PD-10 desalting columns. EGC-I and EGC-II were also expressed and purified following a previously reported method (Han et al., 2017). 2.2. Enzymatic activity and kinetic analysis Enzymatic activity was examined at 37 °C using substrates 4-nitrophenol-β-D-galactopyranoside (pNP-Gal) or 4-nitrophenol-β-D-glucopyranoside (pNP-Glc) (Raygood Bio, Shanghai, China) and releasing pNP product was measured absorbance at wavelength 400 nm. For pH-profiles, EGALC activity was determined in a range of pH 5.1–9.0 using 50 mM buffers: Sodium cacodylate (5.1, 5.5, 6.0, 6.5, and 7.2) and Tris-HCL (7.2, 7.5, 8.0, 8.5, and 9.0). Metal ion effects were examined with 5 mM Ca2+, Li2+, Mg2+, Ni2+, Zn2+, Mn2+, and Cu2+. For steady state kinetic of EGALC, pNP-Gal was dissolved to concentrations of 1.6, 0.8, 0.4, 0.2, 0.1, and 0.05 mM in 50 mM sodium cacodylate buffer, pH 6.5. The kinetic parameters were obtained by fitting initial velocity data to the Michaelis-Menten equation using GraphPad Prism 5.0 (GraphPad Inc.). GalCer and psychosine (Avanti Polar Lipids) were also utilized to test the hydrolysis activity of EGALC. A 40-µl mixture of 0.5 mM GalCer/psychosine with EGALC/mutants (final concentration ~ 0.5 mg/ml) in 50 mM sodium cacodylate buffer, pH 6.5, 0.1% Triton X-100 was incubated at 37 °C for 12 h. After reaction, the mixture was dried, dissolved in chloroform/methanol (2/1, v/v), and then applied to TLC plates, with chloroform/methanol/0.02% CaCl2 (5/4/1, v/v/v) developing. The visualization of GSLs was performed with orcinol. 2.3. Thermal stability analysis Thermal stability of EGCs and variants was determined by heat-induced inactivation. Enzymes in appropriate concentrations were incubated at various temperatures (30 to 70 °C) for 10 min and then cooled down to 4 °C. The residual activities in the supernatant were measured, and Tm was determined by fitting these data to the equation Vo = Vmax/(1 + exp(m × (T − Tm))) (Wyganowski et al., 2013).
2. Materials and methods 2.1. Expression and purification of recombinant EGALC The EGALC (residues 34–488) from R. hoagii 103S was codon-optimized, synthesized and subcloned into NdeI/XhoI restriction sites of pET28a vector (Novagen, Madison, WI). The expression vectors were transformed into Escherichia coli strain BL21 (DE3) and then colonies was inoculated into 50 mL LB medium containing 100 µg/ml kanamycin and grew overnight at 37 °C. The cultures were inoculated 1.0 L autoinduction medium (Studier, 2005) containing 100 µg/ml kanamycin. Once OD600 reached 2.5, the culture was transferred to 16 °C for inducing additional 20 h. After cells harvest, cells were suspended in lysis buffer (20 mM Tris-HCl, pH 8.0, 500 mM NaCl and 20 mM imidazole) and disrupted by a homogenizer. The recombinant proteins were affinity purified using a Ni-NTA (Ni2+-nitrilotriacetate) column (Smart-lifesciences, Changzhou, China) and eluted with buffer (200 mM imidazole, 100 mM NaCl, 100 mM Tris/HCl (pH 7.4). The eluted fraction was diluted into buffer A (20 mM Tris-HCl, pH 7.4, 50 mM NaCl) directly, and purified by anion-affinity column chromatography. The proteins were eluted using a linear gradient of NaCl from 50 to 500 mM at a flow rate of 4 ml·min−1. Proteins were further purified by gel filtration using a Superdex-200 column (GE Healthcare) equilibrated against buffer (25 mM Tris/HCl pH 7.4, 100 mM NaCl). Protein concentration was determined by BCA method. Finally, the pure proteins were concentrated with a 30 kDa Vivaspin-20 concentrator (GE Healthcare) to ~ 20 mg/ml. Site-directed mutagenesis was performed using QuickChange method, with the plasmid encoding the wild type EGALC as a template. All mutations were confirmed by sequencing. EGALC mutants were obtained by following the above-described expression and purification
2.4. Crystallization and structure determination of EGALC Crystallization experiments were conducted in 48-well plates by the hanging-drop vapour diffusion method at 293 K, and each hanging drop was prepared by mixing 1.0 μl each of protein solution and reservoir solution. Small broom-shaped crystals were obtained in the presence of 0.2 M magnesium chloride, 0.1 M HEPES/Sodium hydroxide pH 7.5 and 30% (v/v) PEG 400. Due to poor diffraction of these initial crystals, microseeding was applied to improve crystal quality. Finally, diffraction quality crystals grew in 2 µl -drops (1 µl of protein, 0.8 µl of precipitant, and 0.2 µl of seed) by equilibration against 200 µl reservoir solution (0.1 M sodium citrate tribasic dihydrate pH 5.6, 20% (v/v) 2-propanol, and 20% (w/v) PEG 4000). Crystals were cryoprotected using a reservoir solution containing 25% (v/v) ethylene glycol. The X-ray data were collected on 17U1 beamline at the Shanghai Synchrotron Radiation Facility. Data was indexed, integrated and scaled using xia2-3dii (Winter et al., 2013). The structure was solved by molecular replacement methods using the crystal structure of EGC-I (PDB code 5J7Z) as template. The model was further built with BUCCANEER (Cowtan, 2006) and yield a high quality model. Subsequently, the structure was refined using REFMAC5 (Murshudov et al., 2011) and inspected with the program COOT (Emsley and Cowtan, 2004). The final model was evaluated using the MolProbity program (Chen et al., 2009). Data collection, processing, and refinement statistics were summarized in Table S1. Graphic representations of structures were generated in Chimera (Pettersen et al., 2004). 2
Journal of Structural Biology xxx (xxxx) xxx–xxx
L. Chen, et al.
2.5. Molecular docking and structure analysis
Table 1 Kinetic constants of wild-type EGALC and mutants.*
Ligands were prepared use Schrödinger suite (http://www. schrodinger.com/ligprep). Ten conformations of CDS, TGC, CTS, CTeS, FGC, CFS, as well as GM4, 5 conformations of GSC591 and sulfatide were generated using LigPrep in Schrödinger suite. The carbon chain of ceramide moiety was truncated in favor of docking. Docking of the varied GSLs to EGALC active site were performed using autodocking vina (Trott and Olson, 2010). During the docking process, most parameters were default, but the global search exhaustiveness and the maximum mode number were set to 8 and 10, respectively. A docking grid with a size of 34.7 × 21.2 × 36.6 Å3 over the substrate-binding groove was used. The oxygen atom of glycosidic bond of ligands at a reasonable distance (< 4 Å) to E234 of EGALC was selected for further analysis. Sequence alignments were performed using MAFFT server (Katoh and Standley, 2013) and presented with ESPript3.
WT R185A R185F R185W W372A H382A H382G H382S H382E H382Q
kcat (min−1)
Km (mM)
0.399 0.167 0.397 0.357 0.600 0.144 0.262 0.099 0.109 0.127
0.052 0.039 0.054 0.055 0.055 0.039 0.053 0.044 0.047 0.048
± ± ± ± ± ± ± ± ± ±
0.004 0.002 0.004 0.006 0.007 0.002 0.002 0.001 0.003 0.001
± ± ± ± ± ± ± ± ± ±
kcat/ Km (mM-1min−1) 0.002 0.003 0.003 0.004 0.003 0.004 0.002 0.004 0.008 0.003
7.672 4.195 7.251 6.434 10.780 3.610 4.901 2.238 2.280 2.622
* Enzymatic activities were examined with pNP-Gal as substrate.
ions we tested, Mg2+ had a considerable promoting effect on EGALC's activity. On the other hand, Mn2+, Ni2+, and Li+ had no significant impact on the activity, whereas Ca2+ had a 40% inhibition effect, and Zn2+ and Cu2+ completely inhibited the enzyme activity (Fig. 1D). EGALC's kinetic parameters toward pNP-Gal were also determined; Km and kcat value is 0.052 ± 0.002 mM, and 0.399 ± 0.004 min−1, respectively (Table 1, Fig. S1). Notably, the catalytic efficiency of EGALC toward pNP-Gal, is ~210 fold lower than Re-EGALC's activity for TGC (Galβ1-6Galβ1-6Galβ1-1′Cer) (Ishibashi et al., 2007a,b). We also tested EGALC activity with GalCer and psychosine as substrates; EGALC can hydrolyze both of them, yet its activity toward psychosine was higher (Fig. S2).
3. Results and discussion 3.1. Biochemical characterization of EGALC The EGALC from R. hoagii 103S shares a high sequence similarity (residue identity 92.27%) with the previously characterized EGALC from R. equi (Re-EGALC) (Ishibashi et al., 2007a,b). Thus, R. hoagii EGALC is expected to have a similar substrate specificity to Re-EGALC, which can hydrolyze 6-gala series GSLs, such as CTeS (Galα1-6Galα16Galβ1-6Galβ1-1′Cer), FGC (Galβ1-Galβ1-6Galβ1-6Galβ1-1′Cer), CDS (Galβ1-6Galβ1-1′Cer), TGC (Galβ1-6Galβ1-6Galβ1-1′Cer), CTS (Galα16Galβ1-6Galβ1-1′Cer) and GalCer and pyschosine, but not GSLs with Glcβ1-1′Cer linkages (Ishibashi et al., 2007a,b). Owing to difficulties in obtaining EGALC's natural substrate (Fig. 1A), we used a chromogenic substrate pNP-Gal to characterize EGALC (Fig. 1B). The recombinant EGALC exhibited maximal activity around pH 6.5 with pNP-Gal as a substrate (Fig. 1C), while Re-EGALC has an optimal activity around pH 5.5 (Ishibashi et al., 2007a,b). This optimal pH variation may result from the use of different protein purification methods. Among the metal
3.2. The overall structure of EGALC To facilitate crystallization, we cloned residues 34–488 for heterologous overexpression, thusly discarding a predicted signal peptide region (first 33 residues). The crystal structure of EGALC was solved in its native form at 1.20 Å resolution. The asymmetric unit consists of one molecule. EGALC structure comprises two domains (Fig. 2A). The Nterminal catalytic domain of EGALC adopts a TIM (β/α) 8 barrel fold.
Fig. 1. Biochemical characterization of EGALC. (A) Schematic of natural substrate 6-gala series GSLs. (B) Schematic of a chromogenic substrate pNP-Gal. (C) pH profiles of EGALC activity toward pNP-Gal. Sodium cacodylate buffer (●, pH 5.1–7.2) and Tris-HCl buffer (■, pH 7.2–9.0) in 50 mM were used. (D) Effects of metals on EGALC activity. The pNP-Gal was used as a substrate. 3
Journal of Structural Biology xxx (xxxx) xxx–xxx
L. Chen, et al.
Fig. 2. Structure of EGALC and site-directed mutagenesis analysis. (A) Overall structure of EGALC. It comprises two distinct domains: N-terminal TIM barrel domain from residue Arg55 to Leu396 (sky blue) and C-terminal β-strand domain from Tyr421 to Leu488 (light green). A structural element (residue Leu37 to Val52) before TIM barrel and a linker (residue Asn397 to Ala420) between N- and C-terminal domains are colored in tan and plum, respectively. (B) Topology of EGALC. Secondary structures are presented in same color in 3D structure. (C) Architecture of EGALC substrate binding site. The surface representation of catalytic groove (orange, hydrophobic, blue, polar) in the EGALC structure. (D) Close-up view of the substrate-binding groove. The Gal is docked into the structure of EGALC by superimposing with GALC (PDB entry 4cce). (E) The groove appearing as a boat-shape are divided into two small pockets. (F) Relative activities of EGALC variants. The pNP-Gal was used as a substrate and the activity of WT EGALC was set up as 100%.
Three cap-like small β-strand sheets at the N-terminus seal the TIM barrel. The C-terminal domain comprises six β-strands. Two small βstrand sheets and one α-helix link the N- and C-terminal domain (Fig. 2A and B). A search for structural homology of EGALC using the DALI server identified a large number of glycoside hydrolase family 5 (GH5) members with high structural similarity. As expected, the highest Zscores were for the two other structurally characterized endoglycoceramidases EGC-II (Z-score 38.9; sequence identity 25%; PDB entry 2osx (Caines et al., 2007)), and EGC-I (Z-score 37.4; sequence identity 24%; PDB entry 5dvg (Han et al., 2017)). Though EGC-I remains unclassified to date, EGC-II and EGALC has been classified into
the GH5-28 and GH5-29 subfamily, respectively. 3.3. The active site architecture of EGALC By structural comparison with other GH5 family members, especially in light of EGC-I and EGC-II complex structures, we predict surface groove in a length of ~27.2 Å for the substrate binding site (Fig. 2C). As a GH5-29 member, EGALC is predicted to use a doubledisplacement mechanism in which an anomeric configuration is retained after glycosidic bond cleavage (Davies and Henrissat, 1995). EGALC catalytic residues comprise E341 (nucleophile) and E234 (acid/ base), which are located in this long groove (Fig. 2D). The distance 4
Journal of Structural Biology xxx (xxxx) xxx–xxx
L. Chen, et al.
Fig. 3. Core active sites alignment of EGCs and substrates docking analysis of EGALC (A) The BGC (β-glucose) from EGC-I-GM1 complex structure and critical residues are in stick representation. (B) CTeS docking and (C) FGC docking into EGALC catalytic groove. The docked ceramide moieties are obtained by superimposing EGC-I-GM1 to EGALC. (D) Substrate binding mode of EGALC for 6-gala series GSLs. The solid lines in the saccharide moiety represent β-1,6 glycosidic bond, while the dashed lines are α-1,6 glycosidic bond.
between the carboxylate groups of E234 and E341 is around 5.2 Å, which supports the plausibility of a double-displacement retaining mechanism. The active groove appears as a ‘boat’ shape (Fig. 2C and E). Hydrophilic residues dominate the active groove and several hydrophobic residues located at one side of the groove. The long groove could be divided into two small pockets (core pocket and pocket-2) (Fig. 2E). Residues H137, D139, R185, N233, E234, Y306, E341, W372, and H382 produce the core pocket; residues W372, S373, N376, Q393, and N395 form the pocke-2. Residues Q184, I310, F345, K385, D386, and N388 are located in the rim of the active groove (Fig. 2E).
and 52% activity of wild-type, respectively, indicating these two glutamate acid residues are not absolutely required for the hydrolysis of pNP-Gal. Though mutagenesis is helpful to identify catalytic nucleophile or general base/acid candidates, early studies involved mutating the catalytic nucleophile (Asp 20) of T4 lysozyme also showed the D20C and D20A mutants have very similar levels of activity of wild-type enzyme (Hardy and Poteete, 1991; Rennell et al., 1991). Indeed, replacement of the glutamic acids with the smaller alanine must leave more space in the cavity of active site which may be filled by some reshuffling of the protein structure or by water molecules. If the latter is the case, filled water molecules are likely to directly attack the anomeric center of the activated sugar moiety resulting in hydrolysis of the glycosidic linkage (Withers et al., 1992). Furthermore, the chemical structure of pNP-Gal substrate is smaller than natural substrates of EGALC, this could feasibly affect its interactions in the active pocket, potentially mediating bond-breaking reactions. Among these mutants, a W372A mutant also showed a 1.4-fold increase in activity, with a similar Km to the WT, but an increased kcat (Table 1). The corresponding residue to EGALC’s R185 in EGC-I and EGC-II is F163 and W178, respectively (Fig. 3A). We thus mutated R185 into these residues and examined enzymatic activity with pNP-Gal as substrate, and both of these mutants showed similar catalytic efficiency with the WT (Table 1). We also tested the hydrolysis activity of the aforementioned R185 mutants toward psychosine. The R185A, R185F, and R185W mutants all had very weak activity, while other mutants (D139A, E234A, E341A, W372A, and H382A) showed no detectable activity (Fig. S4). Based on the docking result, H382 of EGALC may form a polar interaction with O4 of Gal. We also introduced several mutations in this site and found that H382E, H382Q, H382G, and H382S each resulted in a reduced catalytic efficiency (Table 1),
3.4. Substrate binding model of EGALC We tried to co-crystallize mutants harboring mutations for the general acid/base and nucleophile residues with pNP-Gal, D-galactal, psychosine, GalCer, and sphingosine, but failed to obtain any such complex structures. To better understand its substrate-binding mode, we docked Gal into the EGALC active site by superimposing EGALC with the galactocerebrosidase-Gal complex structure (Hill et al., 2013) (PDB entry 4cce). The space of the active groove is apparently large enough to accommodate 6-gala series GSLs (Fig. S3). Gal fits very well with the core pocket of the active groove (Fig. 2D). The 1-OH group of Gal is within hydrogen bonding distance to the two catalytic residues (E234 and E341), supporting the plausibility of this conformation for Gal. Moreover, Gal can apparently form polar interactions with residues D139, R185, and H382. A π-CH stacking interaction was also noted between Gal and W372 (Fig. 2D, Fig. S3). To investigate roles of these residues, we conducted alanine-scanning and tested activity with pNPGal as substrate. At pH 6.5, D139A, R185A, E234A, E341A, and H382A mutants retained 15–76% activity compared to WT EGALC (Fig. 2F). Intriguingly, we noted that E234A and E341A mutant still retained 76% 5
Journal of Structural Biology xxx (xxxx) xxx–xxx
L. Chen, et al.
Fig. 4. Substrate binding site of EGALC differs from that of EGC-I and EGC-II. The middle panel showing the overall substrate binding site of EGALC, EGC-I, and EGCII. The top panel showing the hydrophobic surface representation of ceramide moiety binding site of EGALC, EGC-I, and EGC-II (orange, hydrophobic region; blue, polar region). Hydrophobic residues in the ceramide binding site are indicated. The bottom panel showing the saccharide moiety-binding site of EGALC, EGC-I, and EGC-II.
suggesting that both the side chain length and the charge of this His residue is essential for EGALC activity. To get insight into its substrate preference toward 6-gala series GSLs, we also docked CTeS and FGC into EGALC substrate binding groove (Fig. 3B and C). It showed that several residues from the core pocket stabilize Gal1, and an aromatic F345 stabilizes Gal2 via a π stacking interaction. The pocket-2 of EGALC can accommodate Gal3 or Gla3, having polar interactions with residues S373, Q376, and N395. The residue D386 of EGALC likely stabilizes Gla4 via polar interactions. At the end of the EGALC active site groove, two hydrophobic residues, P237 and L279, appear to recognize the ceramide moiety of substrate, producing fewer interactions to the ceramide moiety. In fact, 6-gala series GSLs showed clearly different ceramide composition (Aoki et al., 2004). Notably, a small Nα8-helix (P381-D386) from the Nβ8-Lβ1 loop (residue W372-V410) might provide potential polar interactions with the O6 of Gal1 and also sterically occlude any substitution at the O3, O4 atoms of Gal1 (Fig. 3B and C). Meanwhile, reasonable docking modes can also be generated for potential substrates CDS, TGC, CTS, CTeS, FGC and CFS (Galα1-Galβ16Galβ1-6Galβ1-1′Cer), in the substrate-binding groove of EGALC (Fig. S3A-F), but no reasonable docking modes were identified for GM4, GSC519 and sulfatide (Fig. S3 G-I). The shape of the saccharide moiety of 6-gala series GSLs share similarity with the highly curved U-shaped structure exhibited by 1,6-β-glucans (Lowman et al., 2011), which matches very well with the boat-shaped surface catalytic groove of EGALC. Therefore, we propose a model for 6-gala series GSLs binding by EGALC (Fig. 3D). In this model, saccharide moieties of 6-gala series GSLs mainly fit into two small pockets in the surface catalytic groove and are stabilized by several crucial residues of EGALC.
3.5. Insight into substrate specificity of EGCs Structural alignment of the EGCs’ core active sites showed that the O4 of BGC (glucose unit of GM1) apparently sterically clashes with H382 of EGALC (Fig. 3A). This steric hindrance can explain EGALC's inability to accept GSLs substrates with glucose as their first sugar unit. In contrast, 4-OH group of galactose is in an axial position, which could avoid such steric hindrance. All H382 mutations of EGALC failed to hydrolysis GlcCer or psychosine (data not shown). The presence of H382 and the lack of residue analogous to K61 and Q298 of EGC-I might collectively contribute to the specific recognition of EGALC toward Galβ1-1′Cer. In the EGC-II structure, it lacks a residue corresponding to Q298 of EGC-I, which might be a reason that EGC-II is unable to hydrolyze GlcCer (Izu et al., 1997; Ishibashi et al., 2012). These results reinforce the view that distinguishing between Galβ11′Cer and Glcβ1-1′Cer linkage in EGCs results from several critical residues. Based on the structural and sequence comparative analysis, we speculate the molecular basis for EGCs’ substrate specificity. For EGALC, an extended long N8 region (loop and helix between the TIM barrel β8 and linker β-strand or α-helix) shapes as a catalytic groove and apparently favors 6-gala series GSL (Fig. 4A). This groove sterically occludes branched-GSLs like GM1 and Gb4. In EGC-I, the N1 region (loop and helix between TIM barrel β1 and β2) constitutes a small pocket (Fig. 4B), which may accommodate the additional sugar units of fucosyl-GM1 (Han et al., 2017). Moreover, the Nα3a-Nα3b loop (loop between TIM barrel α3a-helix and α3b-helix) presents in EGC-II (Fig. 4C, Fig.S5) may affect product release (Ishibashi et al., 2012; Han et al., 2017). These variable structural elements form the different architectures of the active pocket of EGCs, contributing to their diverse substrate specificities. Notably, these three influential regions are 6
Journal of Structural Biology xxx (xxxx) xxx–xxx
L. Chen, et al.
Acknowledgements
Table 2 The melting temperature of EGCs and mutants. Enzyme
Tm (°C)
EGALC EGALC-del EGC-I EGC-I-del EGC-II
52.72 ± 0.24 52.10 ± 0.21 41.1 ± 0.29 49.10 ± 1.13 51.79 ± 0.36
We thank staffs of beamlines BL19U1 and 17U at Shanghai Synchrotron Radiation Facility for their assistance in diffraction data collection. We also thank the Dr. John Hugh Snyder for scientific discussion and manuscript preparation. Appendix A. Supplementary data
highly conserved among GH5-29 subfamily members (Fig. S6).
Supplementary data to this article can be found online at https:// doi.org/10.1016/j.jsb.2019.09.010.
3.6. Structural elements that distinguish EGCs
References
Both EGALC and EGC-I have small β-strand sheets at the N-terminus to seal the TIM barrel, and no such structural elements are apparent in the EGC-II structure. Given that additional structural elements at the Nterminus of GH5 cellulases stabilize the structure of the proteins (Pereira et al. 2010), (Santos et al., 2012), we also constructed EGALC and EGC-I mutants with deleted 46 and 52N-terminal residues, respectively, and measured the melting temperature Tm of these mutants. The results showed that Tm value of these deletion mutants were similar to those of WT EGALC and EGC-I (Table 2), indicating these additional N-terminal structural elements apparently have little effect in stabilizing EGALC or EGC-I. In addition, these elements are located on the opposite face of the substrate binding groove/channel in EGALC and EGC-I, they are not likely to function in directly determining substrate specificity. Therefore, it is still not clear what roles these additional structural elements of the N-terminal domain may serve in EGCs.
Aoki, K., Uchiyama, R., Yamauchi, S., Katayama, T., Itonori, S., Sugita, M., Hada, N., Yamada-Hada, J., Takeda, T., Kumagai, H., et al., 2004. Newly discovered neutral glycosphingolipids in aureobasidin A-resistant zygomycetes identification of a novel family of gala-series glycolipids with Core Galα1-6Galβ1-6Galβ sequences. J. Biol. Chem. 279 (31), 32028–32034. https://doi.org/10.1074/jbc.M312918200. Ashida, H., Yamamoto, K., Kumagai, H., Tochikura, T., 1992. Purification and characterization of membrane-bound endoglycoceramidase from Corynebacterium sp. Eur. J. Biochem. 205 (2), 729–735. https://doi.org/10.1111/j.1432-1033.1992. tb16836.x. Caines, M.E.C., Vaughan, M.D., Tarling, C.A., Hancock, S.M., Warren, R.A.J., Withers, S.G., Strynadka, N.C.J., 2007. Structural and mechanistic analyses of endo-glycoceramidase II, a membrane-associated family 5 glycosidase in the Apo and GM3 ganglioside-bound Forms. J. Biol. Chem. 282 (19), 14300–14308. https://doi.org/10. 1074/jbc.M611455200. Chen, V.B., Arendall, W.B., Headd, J.J., Keedy, D.A., Immormino, R.M., Kapral, G.J., Murray, L.W., Richardson, J.S., Richardson, D.C., 2009. MolProbity: all-atom structure validation for macromolecular crystallography. Acta Crystallogr. Section D: Biol. Crystallogr. 66 (1), 12–21. Cowtan, K., 2006. The Buccaneer software for automated model building. 1. Tracing protein chains. Acta Crystallogr. D Biol. Crystallogr. 62 (Pt 9), 1002–1011. https:// doi.org/10.1107/S0907444906022116. Davies, G., Henrissat, B., 1995. Structures and mechanisms of glycosyl hydrolases. Structure 3 (9), 853–859. https://doi.org/10.1016/S0969-2126(01)00220-9. Emsley, P., Cowtan, K., 2004. Coot: model-building tools for molecular graphics. Acta Crystallogr. Section D: Biol. Crystallogr. 60 (12), 2126–2132. Goddard-Borger, E.D., Tysoe, C., Withers, S.G., 2016. Glycosynthase mediated synthesis of psychosine. Carbohydr. Res. 435, 97–99. https://doi.org/10.1016/j.carres.2016. 09.013. Han, Y.-B., Chen, L.-Q., Li, Z., Tan, Y.-M., Feng, Y., Yang, G.-Y., 2017. Structural insights into the broad substrate specificity of a novel endoglycoceramidase I belonging to a new subfamily of GH5 glycosidases. J. Biol. Chem. 292 (12), 4789–4800. https://doi. org/10.1074/jbc.M116.763821. Hancock, S.M., Rich, J.R., Caines, M.E.C., Strynadka, N.C.J., Withers, S.G., 2009. Designer enzymes for glycosphingolipid synthesis by directed evolution. Nat. Chem. Biol. 5 (7), 508–514. https://doi.org/10.1038/nchembio.191. Hardy, L.W., Poteete, A.R., 1991. Reexamination of the role of Asp20 in catalysis by bacteriophage T4 lysozyme. Biochemistry 30 (39), 9457–9463. https://doi.org/10. 1021/bi00103a010. Hill, C.H., Graham, S.C., Read, R.J., Deane, J.E., 2013. Structural snapshots illustrate the catalytic cycle of β-galactocerebrosidase, the defective enzyme in Krabbe disease. PNAS 110 (51), 20479–20484. https://doi.org/10.1073/pnas.1311990110. Horibata, Y., Higashi, H., Ito, M., 2001. Transglycosylation and reverse hydrolysis reactions of endoglycoceramidase from the jellyfish, Cyanea nozakii. J. Biochem. 130 (2), 263–268. Horibata, Y., Okino, N., Ichinose, S., Omori, A., Ito, M., 2000. Purification, characterization, and cDNA cloning of a novel acidic endoglycoceramidase from the Jellyfish, Cyanea nozakii. J. Biol. Chem. 275 (40), 31297–31304. https://doi.org/10.1074/jbc. M003575200. Horibata, Y., Sakaguchi, K., Okino, N., Iida, H., Inagaki, M., Fujisawa, T., Hama, Y., Ito, M., 2004. Unique Catabolic Pathway of Glycosphingolipids in a Hydrozoan, Hydra magnipapillata, Involving Endoglycoceramidase. J. Biol. Chem. 279 (32), 33379–33389. https://doi.org/10.1074/jbc.M401460200. Ishibashi, Y., Kiyohara, M., Okino, N., Ito, M., 2007a. Synthesis of fluorescent glycosphingolipids and neoglycoconjugates which contain 6-gala oligosaccharides using the transglycosylation reaction of a novel endoglycoceramidase (EGALC). J. Biochem. 142 (2), 239–246. https://doi.org/10.1093/jb/mvm125. Ishibashi, Y., Kobayashi, U., Hijikata, A., Sakaguchi, K., Goda, H.M., Tamura, T., Okino, N., Ito, M., 2012. Preparation and characterization of EGCase I, applicable to the comprehensive analysis of GSLs, using a rhodococcal expression system. J. Lipid Res. 53 (10), 2242–2251. https://doi.org/10.1194/jlr.D028951. Yohei, Ishibashi, Nakasone, T., Kiyohara, M., Horibata, Y., Sakaguchi, K., Hijikata, A., Ichinose, S., Omori, A., Yasui, Y., Imamura, A., et al., 2007b. A Novel Endoglycoceramidase hydrolyzes oligogalactosylceramides to produce galactooligosaccharides and ceramides. J. Biol. Chem. 282 (15), 11386–11396. https://doi.org/ 10.1074/jbc.M608445200. Ito, M., Yamagata, T., 1989. Purification and characterization of glycosphingolipid-specific endoglycosidases (endoglycoceramidases) from a mutant strain of Rhodococcus
4. Conclusion In summary, we have biochemically and structurally characterized EGALC. Based on the comparison of this structure with two other homologous structures, we suggest that a long surface groove in EGALC accommodates 6-gala series GSL substrates and that shape complementarity between this catalytic groove and linear substrates likely dominates the substrate specificity. These results expand our scope in understanding the catalytic mechanism of EGC enzymes and are of vital importance in further protein engineering for improving the substrate preference and catalytic efficiency of ECG enzymes toward important glycosphingolipid substrates. 5. Accession number The structure factor and atomic coordinate of EGALC have been deposited in the RCSB Protein Data Bank under the accession code 6JYZ. Author contributions L.Q.C performed most assays. Q.C., Q.D.Y purified mutant proteins and examined enzymatic activities. L.Q.C., G.Y.Y., Y.Z. and Y.F. designed the experiments, analyzed the data and prepared the manuscript. 7. Funding sources We acknowledge the support from National Natural Science Foundation of China (Grant No. 31770846). Declaration of Competing Interest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper. 7
Journal of Structural Biology xxx (xxxx) xxx–xxx
L. Chen, et al. sp. Evidence for three molecular species of endoglycoceramidase with different specificities. J. Biol. Chem. 264 (16), 9510–9519. Izu, H., Izumi, Y., Kurome, Y., Sano, M., Kondo, A., Kato, I., Ito, M., 1997. Molecular cloning, expression, and sequence analysis of the endoglycoceramidase II gene from Rhodococcus species strain M-777. J. Biol. Chem. 272 (32), 19846–19850. https:// doi.org/10.1074/jbc.272.32.19846. Katoh, K., Standley, D.M., 2013. MAFFT multiple sequence alignment software version 7: improvements in performance and usability. Mol. Biol. Evol. 30 (4), 772–780. https://doi.org/10.1093/molbev/mst010. Li, Y.-T., Ishikawa, Y., Li, S.-C., 1987. Occurrence of ceramide-glycanase in the earthworm, Lumbricusterrestris. Biochem. Biophys. Res. Commun. 149 (1), 167–172. https://doi.org/10.1016/0006-291X(87)91619-6. Lingwood, C.A., 2011. Glycosphingolipid functions. Cold, Spring Harb. Perspect. Biol. 3 (7). https://doi.org/10.1101/cshperspect.a004788. Lowman, D.W., West, L.J., Bearden, D.W., Wempe, M.F., Power, T.D., Ensley, H.E., Haynes, K., Williams, D.L., Kruppa, M.D., 2011. New insights into the structure of (1→3,1→6)-β-D-glucan side chains in the Candida glabrata cell wall. PLOS ONE 6 (11). https://doi.org/10.1371/journal.pone.0027614. e27614. Murozuka, Y., Watanabe, N., Hatanaka, K., Hakomori, S., 2007. Lyso-GM3, its dimer, and multimer: their synthesis, and their effect on epidermal growth factor-induced receptor tyrosine kinase. Glycoconj J. 24 (9), 551–563. https://doi.org/10.1007/ s10719-007-9051-2. Murshudov, G.N., Skubák, P., Lebedev, A.A., Pannu, N.S., Steiner, R.A., Nicholls, R.A., Winn, M.D., Long, F., Vagin, A.A., 2011. REFMAC5 for the refinement of macromolecular crystal structures. Acta Crystallograp. Section D Biol.Crystallograp. 67 (4), 355–367. Pettersen, E.F., Goddard, T.D., Huang, C.C., Couch, G.S., Greenblatt, D.M., Meng, E.C., Ferrin, T.E., 2004. UCSF Chimera—a visualization system for exploratory research and analysis. J. Comput. Chem. 25 (13), 1605–1612. https://doi.org/10.1002/jcc. 20084. Regina Todeschini, A., Hakomori, S., 2008. Functional role of glycosphingolipids and
gangliosides in control of cell adhesion, motility, and growth, through glycosynaptic microdomains. Biochimica et Biophysica. Acta (BBA) – Gener. Subjec. 1780 (3), 421–433. https://doi.org/10.1016/j.bbagen.2007.10.008. Rennell, D., Bouvier, S.E., Hardy, L.W., Poteete, A.R., 1991. Systematic mutation of bacteriophage T4 lysozyme. J. Molecul. Biol. 222 (1), 67–88. https://doi.org/10. 1016/0022-2836(91)90738-R. Santos, C.R., Paiva, J.H., Sforça, M.L., Neves, J.L., Navarro, R.Z., Cota, J., Akao, P.K., Hoffmam, Z.B., Meza, A.N., Smetana, J.H., et al., 2012. Dissecting structure–function–stability relationships of a thermostable GH5-CBM3 cellulase from Bacillus subtilis 168. Biochem. J. 441 (1), 95–104. https://doi.org/10.1042/BJ20110869. Schnaar, R.L., Kinoshita, T., 2015. Glycosphingolipids. In: Varki, A., Cummings, R.D., Esko, J.D., Stanley, P., Hart, G.W., Aebi, M., Darvill, A.G., Kinoshita, T., Packer, N.H., Prestegard, J.H. (Eds.), Essentials of Glycobiology, 3rd ed. Cold Spring Harbor Laboratory Press, Cold Spring Harbor (NY). Studier, F.W., 2005. Protein production by auto-induction in high-density shaking cultures. Protein Expr. Purificat. 41 (1), 207–234. https://doi.org/10.1016/j.pep.2005. 01.016. Trott, O., Olson, A.J., 2010. AutoDock Vina: Improving the speed and accuracy of docking with a new scoring function, efficient optimization, and multithreading. J. Comput. Chem. 31 (2), 455–461. https://doi.org/10.1002/jcc.21334. Vaughan, M.D., Johnson, K., DeFrees, S., Tang, X., Warren, R.A.J., Withers, S.G., 2006. Glycosynthase-mediated synthesis of glycosphingolipids. J. Am. Chem. Soc. 128 (19), 6300–6301. Winter, G., Lobley, C.M.C., Prince, S.M., 2013. Decision making in xia2. Acta Cryst. D. 69 (7), 1260–1273. https://doi.org/10.1107/S0907444913015308. Withers, S.G., Rupitz, K., Trimbur, D., Warren, R.A.J., 1992. Mechanistic consequences of mutation of the active site nucleophile Glu 358 in Agrobacterium.beta.-glucosidase. Biochemistry 31 (41), 9979–9985. https://doi.org/10.1021/bi00156a017. Wyganowski, K.T., Kaltenbach, M., Tokuriki, N., 2013. GroEL/ES buffering and compensatory mutations promote protein evolution by stabilizing folding intermediates. J. Mol. Biol. 425 (18), 3403–3414. https://doi.org/10.1016/j.jmb.2013.06.028.
8
Contents lists available at ScienceDirect
Journal of Structural Biology journal homepage: www.elsevier.com/locate/yjsbi
Structure of an endogalactosylceramidase from Rhodococcus hoagii 103S reveals the molecular basis of its substrate specificity Liuqing Chena,b, Qing Changa, Quande Yana, Guangyu Yanga, Yong Zhanga, , Yan Fenga, ⁎
a b
⁎
State Key Laboratory of Microbial Metabolism, School of Life Sciences and Biotechnology, Shanghai Jiao Tong University, Shanghai 200240, China Research Center for Computer-Aided Drug Discovery, Shenzhen Institutes of Advanced Technology, Chinese Academy of Sciences, Shenzhen 518055, China
ARTICLE INFO
ABSTRACT
Keywords: Glycoside hydrolase Endogalactosylceramidase Substrate specificity Crystal structure
Endoglycoceramidases (EGCs) are family 5 glycoside hydrolases that catalyze hydrolysis of the glycosidic linkages between the oligosaccharide and ceramide moieties of glycosphingolipids. Three orthologs of EGCs, each with distinct substrate specificity, have been identified to date, including EGC-I, EGC-II, and EGALC. Although the structures of EGC-I and EGC-II have been reported, the substrate preference mechanism of EGC enzymes remains unclear. Here, we determined the crystal structure of EGALC from Rhodococcus hoagii 103S at a resolution of 1.20 Å. Distinct from EGC-I and EGC-II, which both have a tunnel-like substrate binding site, the structure of EGALC accommodates substrates in a long groove. Further, the oligosaccharide-binding region of groove could be divided into two small pockets that separately bind to the Gal1 and to the Gal3/Gla3 present in 6-gala series substrates. Structural and sequence comparisons of EGC enzymes revealed that the conformation and length of their Nβ8-Lα1 regions are crucial in determining the architectures of their specific substrate binding sites. Importantly, molecular docking analyses indicate that the substrate specificity of each EGC is mainly derived from the complementarity of its active site groove/tunnel with substrates adopting particular conformations. Our study provide insights for understanding the catalytic mechanism of EGALC, which will help protein engineering for improving the substrate preference and catalytic efficiency of EGC enzymes toward important glycosphingolipid substrates.
1. Introduction Glycosphingolipids (GSLs) are ubiquitous cell membrane glycolipids consisting of glycosidically bound carbohydrates and ceramide moieties. GSLs play key roles in acting as secondary messengers or modulators of signal transduction by affecting several events, ranging from cell adhesion, cell growth, to regulation of apoptosis (Regina Todeschini and Hakomori, 2008). Abnormal neuronal accumulation of GSLs, including of galactosylceramide (GalCer) and glucosylceramide (GlcCer), has been reported to cause Krabbe and Gaucher disease (Lingwood, 2011). Due to their strong bioactivity, several GSLs also have been developed into therapeutic molecules against cancer, such as lyso-GM3 (Murozuka et al., 2007). The vast majority of GSLs are classified into distinct “series” according to their saccharide moiety composition. Ganglio, globo, and neolacto series mainly occur in vertebrates, whereas mollu and arthro series predominate in invertebrates (Schnaar et al., 2015). In mammals, GSLs can be degraded by exo-glycosidases, which mainly cleave monosaccharides from the non-reducing end. By contrast,
⁎
endoglycoceramidases (EGCs) from non-mammalian cells have been shown to cleave intact oligosaccharide moieties from GSLs (Li et al., 1987; Ito and Yamagata, 1989; Ashida et al., 1992; Horibata et al., 2000; Horibata et al., 2004). Three different EGC enzymes have been identified in Rhodococccus sp. M−750, including EGC-I, EGC-II, and EGC-III (this was later renamed as endogalactosylceramidase, EGALC) (Ito and Yamagata, 1989). Biochemical characterization shown that EGC-I has the broadest substrate scope amongst these three enzymes; it can hydrolyze ganglio-, lacto-, globo-series GSLs and fucosyl-GM1a (Ishibashi et al., 2012), while EGC-II mainly hydrolyzes ganglio- and globo-series GSLs (Ito and Yamagata, 1989; Izu et al., 1997). In contrast, EGALC only hydrolyzes 6-gala series GSLs, each of which contains a common R-Galβ1-6GalβCer structure that is not recognized by EGC-I or EGC-II (Ishibashi et al., 2007a,b). EGCs can be used for both GSLs analysis and semi-synthesis. In fact, both EGC-II and EGALC can be converted to glycolsynthase by mutation of the catalytic nucleophile residue, which has been successfully used for the synthesis of lyso-GM3 ganglioside and psychosine (Vaughan
Corresponding authors. E-mail addresses: [email protected] (Y. Zhang), [email protected] (Y. Feng).
https://doi.org/10.1016/j.jsb.2019.09.010 Received 2 May 2019; Received in revised form 20 September 2019; Accepted 22 September 2019 1047-8477/ © 2019 Elsevier Inc. All rights reserved.
Please cite this article as: Liuqing Chen, et al., Journal of Structural Biology, https://doi.org/10.1016/j.jsb.2019.09.010
Journal of Structural Biology xxx (xxxx) xxx–xxx
L. Chen, et al.
et al., 2006; Goddard-Borger et al., 2016). Further, engineering EGC-II has improved its synthesis efficiency (Hancock et al., 2009). EGCs are also able to catalyze transglycosylation reactions. For instance, EGALC has been applied to catalyze transglycosylation reaction to product fluorescent glycosylsphingolipids and neoglycoconjugates (Y. Ishibashi et al., 2007). Notably, an EGC from jellyfish has both transglycosylation and reverse hydrolysis activity, which can be used to condense Lac and Cer to form LacCer (Horibata et al., 2001). Naturally, the yield of this transglycosylation reaction is low, as the existed so-called secondary hydrolysis. As the first reported EGCs structure, the EGC-II structure uncovered that the active site residues of this enzyme are located at a wide funnellike polar cavity and a narrow hydrophobic tunnel for binding, respectively, to the oligosaccharide and ceramide moieties of GSLs (Caines et al., 2007). Previously, we reported the crystal structure EGCI from R. equi 103S (Han et al., 2017). Even though we have the structures of EGC-I and EGC-II, the substrate recognition mechanism of EGCs remains unclear. Further, the structure of 6-gala series GSLs are very different from ganglio-series GSLs, and the sequence identity between EGALC and those structure solved EGCs is low (< 30%), suggesting that EGALC may have a distinct substrate recognition feature. Here, we draw on structural and biochemical data to show that EGALC contains a long ‘boat’ shape surface groove for its substrates binding and hydrolysis. By exploring active site interactions during GSL substrates docking, we suggest a model for EGALC substrate recognition that is quite different from the previously reported binding employed by EGC-I and EGC-II. Indeed, our structural and computational results shed light on how EGALC generates its strict substrate specificity toward 6-gala series GSLs. EGALC uses its linear substrate binding pocket to sterically exclude GSLs with branched sugar moieties, thereby facilitating its specific recognition of approximately linear substrates. Our study offers insight into mechanistic details of this EGALC catalysis and guidance for future protein engineering for improving catalytic properties of ECG enzymes.
procedure. Then proteins were exchanged into buffer A using PD-10 desalting columns. EGC-I and EGC-II were also expressed and purified following a previously reported method (Han et al., 2017). 2.2. Enzymatic activity and kinetic analysis Enzymatic activity was examined at 37 °C using substrates 4-nitrophenol-β-D-galactopyranoside (pNP-Gal) or 4-nitrophenol-β-D-glucopyranoside (pNP-Glc) (Raygood Bio, Shanghai, China) and releasing pNP product was measured absorbance at wavelength 400 nm. For pH-profiles, EGALC activity was determined in a range of pH 5.1–9.0 using 50 mM buffers: Sodium cacodylate (5.1, 5.5, 6.0, 6.5, and 7.2) and Tris-HCL (7.2, 7.5, 8.0, 8.5, and 9.0). Metal ion effects were examined with 5 mM Ca2+, Li2+, Mg2+, Ni2+, Zn2+, Mn2+, and Cu2+. For steady state kinetic of EGALC, pNP-Gal was dissolved to concentrations of 1.6, 0.8, 0.4, 0.2, 0.1, and 0.05 mM in 50 mM sodium cacodylate buffer, pH 6.5. The kinetic parameters were obtained by fitting initial velocity data to the Michaelis-Menten equation using GraphPad Prism 5.0 (GraphPad Inc.). GalCer and psychosine (Avanti Polar Lipids) were also utilized to test the hydrolysis activity of EGALC. A 40-µl mixture of 0.5 mM GalCer/psychosine with EGALC/mutants (final concentration ~ 0.5 mg/ml) in 50 mM sodium cacodylate buffer, pH 6.5, 0.1% Triton X-100 was incubated at 37 °C for 12 h. After reaction, the mixture was dried, dissolved in chloroform/methanol (2/1, v/v), and then applied to TLC plates, with chloroform/methanol/0.02% CaCl2 (5/4/1, v/v/v) developing. The visualization of GSLs was performed with orcinol. 2.3. Thermal stability analysis Thermal stability of EGCs and variants was determined by heat-induced inactivation. Enzymes in appropriate concentrations were incubated at various temperatures (30 to 70 °C) for 10 min and then cooled down to 4 °C. The residual activities in the supernatant were measured, and Tm was determined by fitting these data to the equation Vo = Vmax/(1 + exp(m × (T − Tm))) (Wyganowski et al., 2013).
2. Materials and methods 2.1. Expression and purification of recombinant EGALC The EGALC (residues 34–488) from R. hoagii 103S was codon-optimized, synthesized and subcloned into NdeI/XhoI restriction sites of pET28a vector (Novagen, Madison, WI). The expression vectors were transformed into Escherichia coli strain BL21 (DE3) and then colonies was inoculated into 50 mL LB medium containing 100 µg/ml kanamycin and grew overnight at 37 °C. The cultures were inoculated 1.0 L autoinduction medium (Studier, 2005) containing 100 µg/ml kanamycin. Once OD600 reached 2.5, the culture was transferred to 16 °C for inducing additional 20 h. After cells harvest, cells were suspended in lysis buffer (20 mM Tris-HCl, pH 8.0, 500 mM NaCl and 20 mM imidazole) and disrupted by a homogenizer. The recombinant proteins were affinity purified using a Ni-NTA (Ni2+-nitrilotriacetate) column (Smart-lifesciences, Changzhou, China) and eluted with buffer (200 mM imidazole, 100 mM NaCl, 100 mM Tris/HCl (pH 7.4). The eluted fraction was diluted into buffer A (20 mM Tris-HCl, pH 7.4, 50 mM NaCl) directly, and purified by anion-affinity column chromatography. The proteins were eluted using a linear gradient of NaCl from 50 to 500 mM at a flow rate of 4 ml·min−1. Proteins were further purified by gel filtration using a Superdex-200 column (GE Healthcare) equilibrated against buffer (25 mM Tris/HCl pH 7.4, 100 mM NaCl). Protein concentration was determined by BCA method. Finally, the pure proteins were concentrated with a 30 kDa Vivaspin-20 concentrator (GE Healthcare) to ~ 20 mg/ml. Site-directed mutagenesis was performed using QuickChange method, with the plasmid encoding the wild type EGALC as a template. All mutations were confirmed by sequencing. EGALC mutants were obtained by following the above-described expression and purification
2.4. Crystallization and structure determination of EGALC Crystallization experiments were conducted in 48-well plates by the hanging-drop vapour diffusion method at 293 K, and each hanging drop was prepared by mixing 1.0 μl each of protein solution and reservoir solution. Small broom-shaped crystals were obtained in the presence of 0.2 M magnesium chloride, 0.1 M HEPES/Sodium hydroxide pH 7.5 and 30% (v/v) PEG 400. Due to poor diffraction of these initial crystals, microseeding was applied to improve crystal quality. Finally, diffraction quality crystals grew in 2 µl -drops (1 µl of protein, 0.8 µl of precipitant, and 0.2 µl of seed) by equilibration against 200 µl reservoir solution (0.1 M sodium citrate tribasic dihydrate pH 5.6, 20% (v/v) 2-propanol, and 20% (w/v) PEG 4000). Crystals were cryoprotected using a reservoir solution containing 25% (v/v) ethylene glycol. The X-ray data were collected on 17U1 beamline at the Shanghai Synchrotron Radiation Facility. Data was indexed, integrated and scaled using xia2-3dii (Winter et al., 2013). The structure was solved by molecular replacement methods using the crystal structure of EGC-I (PDB code 5J7Z) as template. The model was further built with BUCCANEER (Cowtan, 2006) and yield a high quality model. Subsequently, the structure was refined using REFMAC5 (Murshudov et al., 2011) and inspected with the program COOT (Emsley and Cowtan, 2004). The final model was evaluated using the MolProbity program (Chen et al., 2009). Data collection, processing, and refinement statistics were summarized in Table S1. Graphic representations of structures were generated in Chimera (Pettersen et al., 2004). 2
Journal of Structural Biology xxx (xxxx) xxx–xxx
L. Chen, et al.
2.5. Molecular docking and structure analysis
Table 1 Kinetic constants of wild-type EGALC and mutants.*
Ligands were prepared use Schrödinger suite (http://www. schrodinger.com/ligprep). Ten conformations of CDS, TGC, CTS, CTeS, FGC, CFS, as well as GM4, 5 conformations of GSC591 and sulfatide were generated using LigPrep in Schrödinger suite. The carbon chain of ceramide moiety was truncated in favor of docking. Docking of the varied GSLs to EGALC active site were performed using autodocking vina (Trott and Olson, 2010). During the docking process, most parameters were default, but the global search exhaustiveness and the maximum mode number were set to 8 and 10, respectively. A docking grid with a size of 34.7 × 21.2 × 36.6 Å3 over the substrate-binding groove was used. The oxygen atom of glycosidic bond of ligands at a reasonable distance (< 4 Å) to E234 of EGALC was selected for further analysis. Sequence alignments were performed using MAFFT server (Katoh and Standley, 2013) and presented with ESPript3.
WT R185A R185F R185W W372A H382A H382G H382S H382E H382Q
kcat (min−1)
Km (mM)
0.399 0.167 0.397 0.357 0.600 0.144 0.262 0.099 0.109 0.127
0.052 0.039 0.054 0.055 0.055 0.039 0.053 0.044 0.047 0.048
± ± ± ± ± ± ± ± ± ±
0.004 0.002 0.004 0.006 0.007 0.002 0.002 0.001 0.003 0.001
± ± ± ± ± ± ± ± ± ±
kcat/ Km (mM-1min−1) 0.002 0.003 0.003 0.004 0.003 0.004 0.002 0.004 0.008 0.003
7.672 4.195 7.251 6.434 10.780 3.610 4.901 2.238 2.280 2.622
* Enzymatic activities were examined with pNP-Gal as substrate.
ions we tested, Mg2+ had a considerable promoting effect on EGALC's activity. On the other hand, Mn2+, Ni2+, and Li+ had no significant impact on the activity, whereas Ca2+ had a 40% inhibition effect, and Zn2+ and Cu2+ completely inhibited the enzyme activity (Fig. 1D). EGALC's kinetic parameters toward pNP-Gal were also determined; Km and kcat value is 0.052 ± 0.002 mM, and 0.399 ± 0.004 min−1, respectively (Table 1, Fig. S1). Notably, the catalytic efficiency of EGALC toward pNP-Gal, is ~210 fold lower than Re-EGALC's activity for TGC (Galβ1-6Galβ1-6Galβ1-1′Cer) (Ishibashi et al., 2007a,b). We also tested EGALC activity with GalCer and psychosine as substrates; EGALC can hydrolyze both of them, yet its activity toward psychosine was higher (Fig. S2).
3. Results and discussion 3.1. Biochemical characterization of EGALC The EGALC from R. hoagii 103S shares a high sequence similarity (residue identity 92.27%) with the previously characterized EGALC from R. equi (Re-EGALC) (Ishibashi et al., 2007a,b). Thus, R. hoagii EGALC is expected to have a similar substrate specificity to Re-EGALC, which can hydrolyze 6-gala series GSLs, such as CTeS (Galα1-6Galα16Galβ1-6Galβ1-1′Cer), FGC (Galβ1-Galβ1-6Galβ1-6Galβ1-1′Cer), CDS (Galβ1-6Galβ1-1′Cer), TGC (Galβ1-6Galβ1-6Galβ1-1′Cer), CTS (Galα16Galβ1-6Galβ1-1′Cer) and GalCer and pyschosine, but not GSLs with Glcβ1-1′Cer linkages (Ishibashi et al., 2007a,b). Owing to difficulties in obtaining EGALC's natural substrate (Fig. 1A), we used a chromogenic substrate pNP-Gal to characterize EGALC (Fig. 1B). The recombinant EGALC exhibited maximal activity around pH 6.5 with pNP-Gal as a substrate (Fig. 1C), while Re-EGALC has an optimal activity around pH 5.5 (Ishibashi et al., 2007a,b). This optimal pH variation may result from the use of different protein purification methods. Among the metal
3.2. The overall structure of EGALC To facilitate crystallization, we cloned residues 34–488 for heterologous overexpression, thusly discarding a predicted signal peptide region (first 33 residues). The crystal structure of EGALC was solved in its native form at 1.20 Å resolution. The asymmetric unit consists of one molecule. EGALC structure comprises two domains (Fig. 2A). The Nterminal catalytic domain of EGALC adopts a TIM (β/α) 8 barrel fold.
Fig. 1. Biochemical characterization of EGALC. (A) Schematic of natural substrate 6-gala series GSLs. (B) Schematic of a chromogenic substrate pNP-Gal. (C) pH profiles of EGALC activity toward pNP-Gal. Sodium cacodylate buffer (●, pH 5.1–7.2) and Tris-HCl buffer (■, pH 7.2–9.0) in 50 mM were used. (D) Effects of metals on EGALC activity. The pNP-Gal was used as a substrate. 3
Journal of Structural Biology xxx (xxxx) xxx–xxx
L. Chen, et al.
Fig. 2. Structure of EGALC and site-directed mutagenesis analysis. (A) Overall structure of EGALC. It comprises two distinct domains: N-terminal TIM barrel domain from residue Arg55 to Leu396 (sky blue) and C-terminal β-strand domain from Tyr421 to Leu488 (light green). A structural element (residue Leu37 to Val52) before TIM barrel and a linker (residue Asn397 to Ala420) between N- and C-terminal domains are colored in tan and plum, respectively. (B) Topology of EGALC. Secondary structures are presented in same color in 3D structure. (C) Architecture of EGALC substrate binding site. The surface representation of catalytic groove (orange, hydrophobic, blue, polar) in the EGALC structure. (D) Close-up view of the substrate-binding groove. The Gal is docked into the structure of EGALC by superimposing with GALC (PDB entry 4cce). (E) The groove appearing as a boat-shape are divided into two small pockets. (F) Relative activities of EGALC variants. The pNP-Gal was used as a substrate and the activity of WT EGALC was set up as 100%.
Three cap-like small β-strand sheets at the N-terminus seal the TIM barrel. The C-terminal domain comprises six β-strands. Two small βstrand sheets and one α-helix link the N- and C-terminal domain (Fig. 2A and B). A search for structural homology of EGALC using the DALI server identified a large number of glycoside hydrolase family 5 (GH5) members with high structural similarity. As expected, the highest Zscores were for the two other structurally characterized endoglycoceramidases EGC-II (Z-score 38.9; sequence identity 25%; PDB entry 2osx (Caines et al., 2007)), and EGC-I (Z-score 37.4; sequence identity 24%; PDB entry 5dvg (Han et al., 2017)). Though EGC-I remains unclassified to date, EGC-II and EGALC has been classified into
the GH5-28 and GH5-29 subfamily, respectively. 3.3. The active site architecture of EGALC By structural comparison with other GH5 family members, especially in light of EGC-I and EGC-II complex structures, we predict surface groove in a length of ~27.2 Å for the substrate binding site (Fig. 2C). As a GH5-29 member, EGALC is predicted to use a doubledisplacement mechanism in which an anomeric configuration is retained after glycosidic bond cleavage (Davies and Henrissat, 1995). EGALC catalytic residues comprise E341 (nucleophile) and E234 (acid/ base), which are located in this long groove (Fig. 2D). The distance 4
Journal of Structural Biology xxx (xxxx) xxx–xxx
L. Chen, et al.
Fig. 3. Core active sites alignment of EGCs and substrates docking analysis of EGALC (A) The BGC (β-glucose) from EGC-I-GM1 complex structure and critical residues are in stick representation. (B) CTeS docking and (C) FGC docking into EGALC catalytic groove. The docked ceramide moieties are obtained by superimposing EGC-I-GM1 to EGALC. (D) Substrate binding mode of EGALC for 6-gala series GSLs. The solid lines in the saccharide moiety represent β-1,6 glycosidic bond, while the dashed lines are α-1,6 glycosidic bond.
between the carboxylate groups of E234 and E341 is around 5.2 Å, which supports the plausibility of a double-displacement retaining mechanism. The active groove appears as a ‘boat’ shape (Fig. 2C and E). Hydrophilic residues dominate the active groove and several hydrophobic residues located at one side of the groove. The long groove could be divided into two small pockets (core pocket and pocket-2) (Fig. 2E). Residues H137, D139, R185, N233, E234, Y306, E341, W372, and H382 produce the core pocket; residues W372, S373, N376, Q393, and N395 form the pocke-2. Residues Q184, I310, F345, K385, D386, and N388 are located in the rim of the active groove (Fig. 2E).
and 52% activity of wild-type, respectively, indicating these two glutamate acid residues are not absolutely required for the hydrolysis of pNP-Gal. Though mutagenesis is helpful to identify catalytic nucleophile or general base/acid candidates, early studies involved mutating the catalytic nucleophile (Asp 20) of T4 lysozyme also showed the D20C and D20A mutants have very similar levels of activity of wild-type enzyme (Hardy and Poteete, 1991; Rennell et al., 1991). Indeed, replacement of the glutamic acids with the smaller alanine must leave more space in the cavity of active site which may be filled by some reshuffling of the protein structure or by water molecules. If the latter is the case, filled water molecules are likely to directly attack the anomeric center of the activated sugar moiety resulting in hydrolysis of the glycosidic linkage (Withers et al., 1992). Furthermore, the chemical structure of pNP-Gal substrate is smaller than natural substrates of EGALC, this could feasibly affect its interactions in the active pocket, potentially mediating bond-breaking reactions. Among these mutants, a W372A mutant also showed a 1.4-fold increase in activity, with a similar Km to the WT, but an increased kcat (Table 1). The corresponding residue to EGALC’s R185 in EGC-I and EGC-II is F163 and W178, respectively (Fig. 3A). We thus mutated R185 into these residues and examined enzymatic activity with pNP-Gal as substrate, and both of these mutants showed similar catalytic efficiency with the WT (Table 1). We also tested the hydrolysis activity of the aforementioned R185 mutants toward psychosine. The R185A, R185F, and R185W mutants all had very weak activity, while other mutants (D139A, E234A, E341A, W372A, and H382A) showed no detectable activity (Fig. S4). Based on the docking result, H382 of EGALC may form a polar interaction with O4 of Gal. We also introduced several mutations in this site and found that H382E, H382Q, H382G, and H382S each resulted in a reduced catalytic efficiency (Table 1),
3.4. Substrate binding model of EGALC We tried to co-crystallize mutants harboring mutations for the general acid/base and nucleophile residues with pNP-Gal, D-galactal, psychosine, GalCer, and sphingosine, but failed to obtain any such complex structures. To better understand its substrate-binding mode, we docked Gal into the EGALC active site by superimposing EGALC with the galactocerebrosidase-Gal complex structure (Hill et al., 2013) (PDB entry 4cce). The space of the active groove is apparently large enough to accommodate 6-gala series GSLs (Fig. S3). Gal fits very well with the core pocket of the active groove (Fig. 2D). The 1-OH group of Gal is within hydrogen bonding distance to the two catalytic residues (E234 and E341), supporting the plausibility of this conformation for Gal. Moreover, Gal can apparently form polar interactions with residues D139, R185, and H382. A π-CH stacking interaction was also noted between Gal and W372 (Fig. 2D, Fig. S3). To investigate roles of these residues, we conducted alanine-scanning and tested activity with pNPGal as substrate. At pH 6.5, D139A, R185A, E234A, E341A, and H382A mutants retained 15–76% activity compared to WT EGALC (Fig. 2F). Intriguingly, we noted that E234A and E341A mutant still retained 76% 5
Journal of Structural Biology xxx (xxxx) xxx–xxx
L. Chen, et al.
Fig. 4. Substrate binding site of EGALC differs from that of EGC-I and EGC-II. The middle panel showing the overall substrate binding site of EGALC, EGC-I, and EGCII. The top panel showing the hydrophobic surface representation of ceramide moiety binding site of EGALC, EGC-I, and EGC-II (orange, hydrophobic region; blue, polar region). Hydrophobic residues in the ceramide binding site are indicated. The bottom panel showing the saccharide moiety-binding site of EGALC, EGC-I, and EGC-II.
suggesting that both the side chain length and the charge of this His residue is essential for EGALC activity. To get insight into its substrate preference toward 6-gala series GSLs, we also docked CTeS and FGC into EGALC substrate binding groove (Fig. 3B and C). It showed that several residues from the core pocket stabilize Gal1, and an aromatic F345 stabilizes Gal2 via a π stacking interaction. The pocket-2 of EGALC can accommodate Gal3 or Gla3, having polar interactions with residues S373, Q376, and N395. The residue D386 of EGALC likely stabilizes Gla4 via polar interactions. At the end of the EGALC active site groove, two hydrophobic residues, P237 and L279, appear to recognize the ceramide moiety of substrate, producing fewer interactions to the ceramide moiety. In fact, 6-gala series GSLs showed clearly different ceramide composition (Aoki et al., 2004). Notably, a small Nα8-helix (P381-D386) from the Nβ8-Lβ1 loop (residue W372-V410) might provide potential polar interactions with the O6 of Gal1 and also sterically occlude any substitution at the O3, O4 atoms of Gal1 (Fig. 3B and C). Meanwhile, reasonable docking modes can also be generated for potential substrates CDS, TGC, CTS, CTeS, FGC and CFS (Galα1-Galβ16Galβ1-6Galβ1-1′Cer), in the substrate-binding groove of EGALC (Fig. S3A-F), but no reasonable docking modes were identified for GM4, GSC519 and sulfatide (Fig. S3 G-I). The shape of the saccharide moiety of 6-gala series GSLs share similarity with the highly curved U-shaped structure exhibited by 1,6-β-glucans (Lowman et al., 2011), which matches very well with the boat-shaped surface catalytic groove of EGALC. Therefore, we propose a model for 6-gala series GSLs binding by EGALC (Fig. 3D). In this model, saccharide moieties of 6-gala series GSLs mainly fit into two small pockets in the surface catalytic groove and are stabilized by several crucial residues of EGALC.
3.5. Insight into substrate specificity of EGCs Structural alignment of the EGCs’ core active sites showed that the O4 of BGC (glucose unit of GM1) apparently sterically clashes with H382 of EGALC (Fig. 3A). This steric hindrance can explain EGALC's inability to accept GSLs substrates with glucose as their first sugar unit. In contrast, 4-OH group of galactose is in an axial position, which could avoid such steric hindrance. All H382 mutations of EGALC failed to hydrolysis GlcCer or psychosine (data not shown). The presence of H382 and the lack of residue analogous to K61 and Q298 of EGC-I might collectively contribute to the specific recognition of EGALC toward Galβ1-1′Cer. In the EGC-II structure, it lacks a residue corresponding to Q298 of EGC-I, which might be a reason that EGC-II is unable to hydrolyze GlcCer (Izu et al., 1997; Ishibashi et al., 2012). These results reinforce the view that distinguishing between Galβ11′Cer and Glcβ1-1′Cer linkage in EGCs results from several critical residues. Based on the structural and sequence comparative analysis, we speculate the molecular basis for EGCs’ substrate specificity. For EGALC, an extended long N8 region (loop and helix between the TIM barrel β8 and linker β-strand or α-helix) shapes as a catalytic groove and apparently favors 6-gala series GSL (Fig. 4A). This groove sterically occludes branched-GSLs like GM1 and Gb4. In EGC-I, the N1 region (loop and helix between TIM barrel β1 and β2) constitutes a small pocket (Fig. 4B), which may accommodate the additional sugar units of fucosyl-GM1 (Han et al., 2017). Moreover, the Nα3a-Nα3b loop (loop between TIM barrel α3a-helix and α3b-helix) presents in EGC-II (Fig. 4C, Fig.S5) may affect product release (Ishibashi et al., 2012; Han et al., 2017). These variable structural elements form the different architectures of the active pocket of EGCs, contributing to their diverse substrate specificities. Notably, these three influential regions are 6
Journal of Structural Biology xxx (xxxx) xxx–xxx
L. Chen, et al.
Acknowledgements
Table 2 The melting temperature of EGCs and mutants. Enzyme
Tm (°C)
EGALC EGALC-del EGC-I EGC-I-del EGC-II
52.72 ± 0.24 52.10 ± 0.21 41.1 ± 0.29 49.10 ± 1.13 51.79 ± 0.36
We thank staffs of beamlines BL19U1 and 17U at Shanghai Synchrotron Radiation Facility for their assistance in diffraction data collection. We also thank the Dr. John Hugh Snyder for scientific discussion and manuscript preparation. Appendix A. Supplementary data
highly conserved among GH5-29 subfamily members (Fig. S6).
Supplementary data to this article can be found online at https:// doi.org/10.1016/j.jsb.2019.09.010.
3.6. Structural elements that distinguish EGCs
References
Both EGALC and EGC-I have small β-strand sheets at the N-terminus to seal the TIM barrel, and no such structural elements are apparent in the EGC-II structure. Given that additional structural elements at the Nterminus of GH5 cellulases stabilize the structure of the proteins (Pereira et al. 2010), (Santos et al., 2012), we also constructed EGALC and EGC-I mutants with deleted 46 and 52N-terminal residues, respectively, and measured the melting temperature Tm of these mutants. The results showed that Tm value of these deletion mutants were similar to those of WT EGALC and EGC-I (Table 2), indicating these additional N-terminal structural elements apparently have little effect in stabilizing EGALC or EGC-I. In addition, these elements are located on the opposite face of the substrate binding groove/channel in EGALC and EGC-I, they are not likely to function in directly determining substrate specificity. Therefore, it is still not clear what roles these additional structural elements of the N-terminal domain may serve in EGCs.
Aoki, K., Uchiyama, R., Yamauchi, S., Katayama, T., Itonori, S., Sugita, M., Hada, N., Yamada-Hada, J., Takeda, T., Kumagai, H., et al., 2004. Newly discovered neutral glycosphingolipids in aureobasidin A-resistant zygomycetes identification of a novel family of gala-series glycolipids with Core Galα1-6Galβ1-6Galβ sequences. J. Biol. Chem. 279 (31), 32028–32034. https://doi.org/10.1074/jbc.M312918200. Ashida, H., Yamamoto, K., Kumagai, H., Tochikura, T., 1992. Purification and characterization of membrane-bound endoglycoceramidase from Corynebacterium sp. Eur. J. Biochem. 205 (2), 729–735. https://doi.org/10.1111/j.1432-1033.1992. tb16836.x. Caines, M.E.C., Vaughan, M.D., Tarling, C.A., Hancock, S.M., Warren, R.A.J., Withers, S.G., Strynadka, N.C.J., 2007. Structural and mechanistic analyses of endo-glycoceramidase II, a membrane-associated family 5 glycosidase in the Apo and GM3 ganglioside-bound Forms. J. Biol. Chem. 282 (19), 14300–14308. https://doi.org/10. 1074/jbc.M611455200. Chen, V.B., Arendall, W.B., Headd, J.J., Keedy, D.A., Immormino, R.M., Kapral, G.J., Murray, L.W., Richardson, J.S., Richardson, D.C., 2009. MolProbity: all-atom structure validation for macromolecular crystallography. Acta Crystallogr. Section D: Biol. Crystallogr. 66 (1), 12–21. Cowtan, K., 2006. The Buccaneer software for automated model building. 1. Tracing protein chains. Acta Crystallogr. D Biol. Crystallogr. 62 (Pt 9), 1002–1011. https:// doi.org/10.1107/S0907444906022116. Davies, G., Henrissat, B., 1995. Structures and mechanisms of glycosyl hydrolases. Structure 3 (9), 853–859. https://doi.org/10.1016/S0969-2126(01)00220-9. Emsley, P., Cowtan, K., 2004. Coot: model-building tools for molecular graphics. Acta Crystallogr. Section D: Biol. Crystallogr. 60 (12), 2126–2132. Goddard-Borger, E.D., Tysoe, C., Withers, S.G., 2016. Glycosynthase mediated synthesis of psychosine. Carbohydr. Res. 435, 97–99. https://doi.org/10.1016/j.carres.2016. 09.013. Han, Y.-B., Chen, L.-Q., Li, Z., Tan, Y.-M., Feng, Y., Yang, G.-Y., 2017. Structural insights into the broad substrate specificity of a novel endoglycoceramidase I belonging to a new subfamily of GH5 glycosidases. J. Biol. Chem. 292 (12), 4789–4800. https://doi. org/10.1074/jbc.M116.763821. Hancock, S.M., Rich, J.R., Caines, M.E.C., Strynadka, N.C.J., Withers, S.G., 2009. Designer enzymes for glycosphingolipid synthesis by directed evolution. Nat. Chem. Biol. 5 (7), 508–514. https://doi.org/10.1038/nchembio.191. Hardy, L.W., Poteete, A.R., 1991. Reexamination of the role of Asp20 in catalysis by bacteriophage T4 lysozyme. Biochemistry 30 (39), 9457–9463. https://doi.org/10. 1021/bi00103a010. Hill, C.H., Graham, S.C., Read, R.J., Deane, J.E., 2013. Structural snapshots illustrate the catalytic cycle of β-galactocerebrosidase, the defective enzyme in Krabbe disease. PNAS 110 (51), 20479–20484. https://doi.org/10.1073/pnas.1311990110. Horibata, Y., Higashi, H., Ito, M., 2001. Transglycosylation and reverse hydrolysis reactions of endoglycoceramidase from the jellyfish, Cyanea nozakii. J. Biochem. 130 (2), 263–268. Horibata, Y., Okino, N., Ichinose, S., Omori, A., Ito, M., 2000. Purification, characterization, and cDNA cloning of a novel acidic endoglycoceramidase from the Jellyfish, Cyanea nozakii. J. Biol. Chem. 275 (40), 31297–31304. https://doi.org/10.1074/jbc. M003575200. Horibata, Y., Sakaguchi, K., Okino, N., Iida, H., Inagaki, M., Fujisawa, T., Hama, Y., Ito, M., 2004. Unique Catabolic Pathway of Glycosphingolipids in a Hydrozoan, Hydra magnipapillata, Involving Endoglycoceramidase. J. Biol. Chem. 279 (32), 33379–33389. https://doi.org/10.1074/jbc.M401460200. Ishibashi, Y., Kiyohara, M., Okino, N., Ito, M., 2007a. Synthesis of fluorescent glycosphingolipids and neoglycoconjugates which contain 6-gala oligosaccharides using the transglycosylation reaction of a novel endoglycoceramidase (EGALC). J. Biochem. 142 (2), 239–246. https://doi.org/10.1093/jb/mvm125. Ishibashi, Y., Kobayashi, U., Hijikata, A., Sakaguchi, K., Goda, H.M., Tamura, T., Okino, N., Ito, M., 2012. Preparation and characterization of EGCase I, applicable to the comprehensive analysis of GSLs, using a rhodococcal expression system. J. Lipid Res. 53 (10), 2242–2251. https://doi.org/10.1194/jlr.D028951. Yohei, Ishibashi, Nakasone, T., Kiyohara, M., Horibata, Y., Sakaguchi, K., Hijikata, A., Ichinose, S., Omori, A., Yasui, Y., Imamura, A., et al., 2007b. A Novel Endoglycoceramidase hydrolyzes oligogalactosylceramides to produce galactooligosaccharides and ceramides. J. Biol. Chem. 282 (15), 11386–11396. https://doi.org/ 10.1074/jbc.M608445200. Ito, M., Yamagata, T., 1989. Purification and characterization of glycosphingolipid-specific endoglycosidases (endoglycoceramidases) from a mutant strain of Rhodococcus
4. Conclusion In summary, we have biochemically and structurally characterized EGALC. Based on the comparison of this structure with two other homologous structures, we suggest that a long surface groove in EGALC accommodates 6-gala series GSL substrates and that shape complementarity between this catalytic groove and linear substrates likely dominates the substrate specificity. These results expand our scope in understanding the catalytic mechanism of EGC enzymes and are of vital importance in further protein engineering for improving the substrate preference and catalytic efficiency of ECG enzymes toward important glycosphingolipid substrates. 5. Accession number The structure factor and atomic coordinate of EGALC have been deposited in the RCSB Protein Data Bank under the accession code 6JYZ. Author contributions L.Q.C performed most assays. Q.C., Q.D.Y purified mutant proteins and examined enzymatic activities. L.Q.C., G.Y.Y., Y.Z. and Y.F. designed the experiments, analyzed the data and prepared the manuscript. 7. Funding sources We acknowledge the support from National Natural Science Foundation of China (Grant No. 31770846). Declaration of Competing Interest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper. 7
Journal of Structural Biology xxx (xxxx) xxx–xxx
L. Chen, et al. sp. Evidence for three molecular species of endoglycoceramidase with different specificities. J. Biol. Chem. 264 (16), 9510–9519. Izu, H., Izumi, Y., Kurome, Y., Sano, M., Kondo, A., Kato, I., Ito, M., 1997. Molecular cloning, expression, and sequence analysis of the endoglycoceramidase II gene from Rhodococcus species strain M-777. J. Biol. Chem. 272 (32), 19846–19850. https:// doi.org/10.1074/jbc.272.32.19846. Katoh, K., Standley, D.M., 2013. MAFFT multiple sequence alignment software version 7: improvements in performance and usability. Mol. Biol. Evol. 30 (4), 772–780. https://doi.org/10.1093/molbev/mst010. Li, Y.-T., Ishikawa, Y., Li, S.-C., 1987. Occurrence of ceramide-glycanase in the earthworm, Lumbricusterrestris. Biochem. Biophys. Res. Commun. 149 (1), 167–172. https://doi.org/10.1016/0006-291X(87)91619-6. Lingwood, C.A., 2011. Glycosphingolipid functions. Cold, Spring Harb. Perspect. Biol. 3 (7). https://doi.org/10.1101/cshperspect.a004788. Lowman, D.W., West, L.J., Bearden, D.W., Wempe, M.F., Power, T.D., Ensley, H.E., Haynes, K., Williams, D.L., Kruppa, M.D., 2011. New insights into the structure of (1→3,1→6)-β-D-glucan side chains in the Candida glabrata cell wall. PLOS ONE 6 (11). https://doi.org/10.1371/journal.pone.0027614. e27614. Murozuka, Y., Watanabe, N., Hatanaka, K., Hakomori, S., 2007. Lyso-GM3, its dimer, and multimer: their synthesis, and their effect on epidermal growth factor-induced receptor tyrosine kinase. Glycoconj J. 24 (9), 551–563. https://doi.org/10.1007/ s10719-007-9051-2. Murshudov, G.N., Skubák, P., Lebedev, A.A., Pannu, N.S., Steiner, R.A., Nicholls, R.A., Winn, M.D., Long, F., Vagin, A.A., 2011. REFMAC5 for the refinement of macromolecular crystal structures. Acta Crystallograp. Section D Biol.Crystallograp. 67 (4), 355–367. Pettersen, E.F., Goddard, T.D., Huang, C.C., Couch, G.S., Greenblatt, D.M., Meng, E.C., Ferrin, T.E., 2004. UCSF Chimera—a visualization system for exploratory research and analysis. J. Comput. Chem. 25 (13), 1605–1612. https://doi.org/10.1002/jcc. 20084. Regina Todeschini, A., Hakomori, S., 2008. Functional role of glycosphingolipids and
gangliosides in control of cell adhesion, motility, and growth, through glycosynaptic microdomains. Biochimica et Biophysica. Acta (BBA) – Gener. Subjec. 1780 (3), 421–433. https://doi.org/10.1016/j.bbagen.2007.10.008. Rennell, D., Bouvier, S.E., Hardy, L.W., Poteete, A.R., 1991. Systematic mutation of bacteriophage T4 lysozyme. J. Molecul. Biol. 222 (1), 67–88. https://doi.org/10. 1016/0022-2836(91)90738-R. Santos, C.R., Paiva, J.H., Sforça, M.L., Neves, J.L., Navarro, R.Z., Cota, J., Akao, P.K., Hoffmam, Z.B., Meza, A.N., Smetana, J.H., et al., 2012. Dissecting structure–function–stability relationships of a thermostable GH5-CBM3 cellulase from Bacillus subtilis 168. Biochem. J. 441 (1), 95–104. https://doi.org/10.1042/BJ20110869. Schnaar, R.L., Kinoshita, T., 2015. Glycosphingolipids. In: Varki, A., Cummings, R.D., Esko, J.D., Stanley, P., Hart, G.W., Aebi, M., Darvill, A.G., Kinoshita, T., Packer, N.H., Prestegard, J.H. (Eds.), Essentials of Glycobiology, 3rd ed. Cold Spring Harbor Laboratory Press, Cold Spring Harbor (NY). Studier, F.W., 2005. Protein production by auto-induction in high-density shaking cultures. Protein Expr. Purificat. 41 (1), 207–234. https://doi.org/10.1016/j.pep.2005. 01.016. Trott, O., Olson, A.J., 2010. AutoDock Vina: Improving the speed and accuracy of docking with a new scoring function, efficient optimization, and multithreading. J. Comput. Chem. 31 (2), 455–461. https://doi.org/10.1002/jcc.21334. Vaughan, M.D., Johnson, K., DeFrees, S., Tang, X., Warren, R.A.J., Withers, S.G., 2006. Glycosynthase-mediated synthesis of glycosphingolipids. J. Am. Chem. Soc. 128 (19), 6300–6301. Winter, G., Lobley, C.M.C., Prince, S.M., 2013. Decision making in xia2. Acta Cryst. D. 69 (7), 1260–1273. https://doi.org/10.1107/S0907444913015308. Withers, S.G., Rupitz, K., Trimbur, D., Warren, R.A.J., 1992. Mechanistic consequences of mutation of the active site nucleophile Glu 358 in Agrobacterium.beta.-glucosidase. Biochemistry 31 (41), 9979–9985. https://doi.org/10.1021/bi00156a017. Wyganowski, K.T., Kaltenbach, M., Tokuriki, N., 2013. GroEL/ES buffering and compensatory mutations promote protein evolution by stabilizing folding intermediates. J. Mol. Biol. 425 (18), 3403–3414. https://doi.org/10.1016/j.jmb.2013.06.028.
8