The structure of a family 168 glycoside hydrolase from the marine bacterium Muricauda eckloniae

The genome of the marine bacterium Muricauda eckloniae sp. DK169 contains an extensive polysaccharide-utilization locus that targets fucoidan from brown algae. Within this locus is a gene that encodes a putative fucoidan-degrading glycoside hydrolase (locus tag AAY42_01205) assigned to glycoside hydrolase family 168, which we call MeGH168. We present the 2.0 Å resolution X-ray crystal structure of MeGH168, demonstrating a (β/α)₈-barrel fold. The eight loop regions joining each α-helix and β-strand surround the catalytic groove. A comparison with the structure of a GH168, Fun168A, in complex with a fragment of fucoidan (PDB entry 8ya7) revealed conservation of key residues in the catalytic site. However, structural variation in positive-subsite loop regions may recontour the active site to create differences in substrate specificity between the two GH168s. The present data provide additional structural insights into the GH168 family, particularly expanding on sequence and structure conservation (and the lack thereof) in relation to substrate interactions.

1. Introduction

Species in the Muricauda genus are distributed in diverse marine environments ranging from coastal ecosystems to deep-sea hydrothermal fields (Cao et al., 2023). M. eckloniae sp. DK169 was isolated from the rhizosphere of Ecklonia kurome, a marine alga found on the coast of Dokdo Island, Korea (Bae et al., 2007). Accordingly, M. eckloniae primarily colonizes macroalgal surfaces and has garnered increasing interest due to the presence of genomic regions which encode unique carbohydrate-active enzymes (CAZymes) capable of degrading algal polysaccharides (Tran et al., 2022; Mikkelsen et al., 2023). Algal polysaccharides often possess chemically diverse and sophisticated structures; therefore, algal-specific CAZymes often have structural characteristics that impart high substrate specificity to effectively process polysaccharides (Bäumgen et al., 2021; Arnosti et al., 2021; Sarkar et al., 2024). As a result, chemically complex algal polysaccharide degradation has yet to be thoroughly characterized, particularly that of sulfated fucans such as fucoidan.

Fucoidan is a complex polysaccharide that is found in several brown algae species. It is primarily composed of an α-1,3-linked L-fucose backbone or an alternating α-1,3/1,4 backbone, depending on the source species (Mikkelsen et al., 2023; Ale et al., 2011; Chen et al., 2024; Buck-Wiese et al., 2023; Kusaykin et al., 2015). Additionally, heavy sulfation is present primarily on C2 and/or C4 for most or all fucose residues. Along with varying backbone linkage and sulfation patterns, the complexity of fucoidans is further established by species-dependent branching patterns, occasional acetylation and the presence of varying sulfated monosaccharides in its backbone. As a result, different fucoidan-targeting enzymes, or fucoidanases, must accommodate a wide variety of substrate configurations to process the complex structure of fucoidan. However, relatively few fucoidanases have been characterized to date, and thus further investigation of their unique abilities to recognize a structurally heterogenous substrate is required.

The fucoidan-degradation locus of M. eckloniae sp. DK169 encodes 25 proteins associated with fucoidan depolymerization, transport and assimilation (Tran et al., 2022; Mikkelsen et al., 2023). This fucoidan-utilizing locus is mostly comprised of glycoside hydrolases and sulfatases, revealing a wide variety of substrate specificities required for depolymerization. Amongst these is a gene encoding a member of glycoside hydrolase family 168, which we call MeGH168, which is of particular interest due to its potential in initiating fucoidan degradation. In general, the inferred catalytic mechanism of GH168s is a two-step retaining mechanism utilizing two carboxylic acid-containing amino acids as nucleophile and acid–base residues (Shen et al., 2020; Silchenko et al., 2023; Chen et al., 2024). Recently, GH168 enzymes have garnered particular interest, with an increasing number of structures characterized. Of these, the structure of FunGH168 from Wenyingzhuangia fucanilytica, which is trapped in complex with a fucoidan-hydrolysis product, has demonstrated the molecular basis of its endo-hydrolytic activity on α-1,3-L-fucose linkages (Shen et al., 2020; Chen et al., 2024). Here, we report the structure of MeGH168 and use a comparison of the structure with the FunGH168 product complex to reveal the conservation of catalytic residues and some subsite-binding motifs. However, major structural differences suggest potential differences in substrate specificity.

2. Materials and methods

2.1. Macromolecule production

The gene encoding MeGH168 from M. eckloniae sp. DK169 (locus tag AAY42_01205; GenBank KQC28672.1) was obtained as a synthetic gene comprising an in-frame fusion with an N-terminal His₆-tag to generate pET-28a-MeGH168 (Hobbs et al., 2019). The predicted secretion signal peptide was omitted, resulting in the amino-acid boundaries 19–375. Protein production and purification were performed as described previously, utilizing Ni²⁺–NTA immobilized metal-affinity chromatography followed by size-exclusion chromatography using a Sephacryl S-200 column (GE Healthcare) pre-equilibrated with 20 mM Tris–HCl pH 8.0 containing 500 mM sodium chloride (Hobbs et al., 2019). The purified protein was concentrated to ∼10–11 mg ml⁻¹ using a stirred-cell ultrafiltration device (Amicon) with a 10 000 Da molecular-weight cutoff membrane (Millipore) for crystallization. Macromolecule-production information is summarized in Table 1.

Table 1
Macromolecule-production information

Source organism	Muricauda eckloniae DK169 (strain DOKDO 007T)
DNA source	Synthetic DNA construct
Expression vector	pET-28a
Expression host	Escherichia coli
Complete amino-acid sequence of the construct produced†	MGSSHHHHHHSSGLVPRGSHMQQEYYPNFSWDKVPVAFHFGKRDGLMTKDEAKFVTSRSNFIVLEKAHGAPDYEYTEDAIAKEARKLKKLNPGMKVIFYWNSFLDYNMYKAHEVYQNHPQWWLRKQDGELDFKNKGLKRYDLSNPKVRDWWTDVAKNEIVNGSTDGIFMDAFIQVSNPANIKLWGQKKYNDIQQGLKDLIKETREKLGDDKLIVYNGIRSTFQRNVGNNFPDYTDVVMIEHFGHFASTSKESMLTDIQEMEKAGKSGKIVVFKAWPGFAWIDKEAMSKPYVEKQKIAKNSITFPLAAFLAGAQEHSYFIYNWGYRMEMGCLEWYPEFDKPLGKPLNDMVINGWVLTREYEHALVWVNLETNEAKINWK

†The His₆-tag sequence is shown in bold.

2.2. Crystallization

Crystallization screening was conducted with the commercially available screens Index, PEG/Ion and Crystal Screen (Hampton Research) by sitting-drop vapour diffusion at 291 K with a 1:1 ratio of protein solution (11 mg ml⁻¹ in 20 mM Tris–HCl pH 8.0, 500 mM sodium chloride) and crystallization solution (0.5 µl each). A crystal grown in 20%(w/v) PEG 3350 with 8%(v/v) Tacsimate pH 7.0 was harvested from the screen and used to collect diffraction data. Crystallization information is summarized in Table 2.

Table 2
Crystallization information

Method	Vapour diffusion, sitting drop
Temperature (K)	291
Protein concentration (mg ml^−1)	11
Buffer composition of protein solution	20 mM Tris–HCl pH 8.0, 500 mM sodium chloride
Composition of reservoir solution	20%(w/v) PEG 3350, 8%(v/v) Tacsimate pH 7.0
Volume and ratio of drop	1 µl, 1:1
Volume of reservoir (µl)	50

2.3. Data collection and processing

The crystal was cryoprotected in crystallization solution with 20%(v/v) ethylene glycol and immediately cooled in liquid nitrogen for storage and shipping. Diffraction data were collected on the CMCF-ID beamline at the Canadian Light Source (CLS). Data were integrated, scaled and merged with XDS and AIMLESS (Kabsch, 2010; Evans & Murshudov, 2013; Agirre et al., 2023).

Data-collection and processing statistics are summarized in Table 3.

Table 3
Data collection and processing

Values in parentheses are for the outer shell.

Diffraction source	CLS
Wavelength (Å)	0.954
Temperature (K)	100
Detector	Dectris EIGER X 9M
Rotation range per image (°)	0.2
Total rotation range (°)	360
Exposure time per image (s)	0.5
Space group	P22121
a, b, c (Å)	47.27, 120.93, 163.85
α, β, γ (°)	90, 90, 90
Mosaicity (°)	0.1
Resolution range (Å)	97.31–2.00 (2.05–2.00)
Total No. of reflections	493807
No. of unique reflections	64544
Completeness (%)	100.0 (100.0)
Multiplicity	7.6 (8.0)
〈I/σ(I)〉	12.8 (2.4)
Rp.i.m.	0.051 (0.364)
Rmeas	0.142 (1.045)
CC1/2	0.997 (0.764)
Overall B factor from Wilson plot (Å^2)	23.59

2.4. Structure solution and refinement

The crystal structure of MeGH168 was solved by molecular replacement using Phaser (McCoy et al., 2007) and an AlphaFold-generated model of MeGH168 (Jumper et al., 2021; Varadi et al., 2022). Coot was used to manually correct the model and phenix.refine was used for refinement of the experimental model and the addition of water molecules (Liebschner et al., 2019; Emsley & Cowtan, 2004). Refinement procedures were monitored by flagging 5% of all observations as `free' (Brünger, 1992). Model validation was performed with MolProbity (Chen et al., 2010). A summary of the model statistics is shown in Table 4.

Table 4
Structure solution and refinement

Values in parentheses are for the outer shell.

Resolution range (Å)	48.650–2.000 (2.071–2.000)
Completeness (%)	99.97
No. of reflections, working set	64518 (6356)
No. of reflections, test set	3186
Final Rwork	0.1989 (0.2001)
Final Rfree	0.2327 (0.2966)
No. of non-H atoms
Protein	2952
Ligand	1 [EDO]
Water	842
Total	6767
R.m.s. deviations
Bond lengths (Å)	0.002
Angles (°)	0.49
Average B factors (Å^2)
Protein	25.93
Ligand	31.71 [EDO]
Water	35.82
Ramachandran plot
Favoured regions (%)	98.45
Allowed (%)	1.55
Outliers (%)	0.00
PDB code	9nhf

3. Results and discussion

3.1. Overall structure of MeGH168

MeGH168 crystallized in space group P22₁2₁ with unit-cell parameters a = 47.27, b = 120.93, c = 163.85 Å. Analysis using MATTHEWS_COEF (Kantardjieff & Rupp, 2003) showed two protein molecules in the asymmetric unit with a Matthews coefficient of 2.65 ų Da⁻¹ and a solvent content of 53.67%. PISA analysis indicated that dimers of the protein are unlikely to form in solution (Krissinel & Henrick, 2007). Recombinant protein residues 20–375 were modelled and refined; no gaps were present in the backbone.

The structure of MeGH168 is composed of a (β/α)₈ triosephosphate isomerase (TIM) barrel fold, with eight α-helices and eight parallel β-strands alternating at the core of the enzyme (Fig. 1a). Connecting each α-helix and β-strand are eight unique loop regions. Additionally, an antiparallel β-sheet fold, consisting of four β-strands, is present in the C-terminal region. Based on an estimation of the active-site length (approximately 25 Å), determined by cleft topology (Fig. 1b), in combination with the distribution of aromatic residues (Fig. 1a), we expect approximately five subsites to be present, with either two or three subsites on either side of the scissile bond.

Figure 1 Overall structure of MeGH168. (a) Cartoon representation of MeGH168. Each β–α region is shown in different colours, with the (β/α)8 barrel sequentially shown in purple to orange and the C-terminal β-sheet fold in red. Aromatic and predicted catalytic residues in the active-site cleft are shown in blue stick representation. (b) Surface representation of MeGH168. Arrows in both panels indicate the catalytic groove.

Structural similarity results from a DALI server search yielded Fun168A from W. fucanilytica as the top hit [PDB entry 8ya7; Z-score and root-mean-square deviation (r.m.s.d.) of 40.5 and 2.3 Å, respectively; 29% amino-acid sequence identity]. The general structure of Fun168A has the same (β/α)₈ TIM-barrel fold. The structures of MeGH168 and Fun168A both have eight loop regions connecting the barrelling α-helices and parallel β-strands. Compared with Fun168A (PDB entry 8ya7), MeGH168 shows some structural conservation in loop regions near the front of the catalytic groove (for reference: the loop regions to the left of the Fun168A fucoidan ligand in Fig. 2a). However, the loop regions at the back of the catalytic groove have high variability, suggesting some possible contributions to differences in substrate recognition. In particular, an antiparallel β-hairpin in the predicted positive-subsite region of MeGH168 is likely to have a profound effect on reshaping the active site of this enzyme (Fig. 2a). It may promote the recognition of a different fucoidan conformation resulting from an alternate backbone structure, such as the presence of α-1,4-linkages. Notably, the β-hairpin could convert this particular enzyme to an exo-activity, rather than the endo-activity observed for Fun168A, by potentially preventing the accommodation of larger substrates.

Figure 2 Structural alignment of MeGH168 with the Fun168A product complex provides insight into its active site. (a) Complete alignment of MeGH168 (cyan) and Fun168A (PDB entry 8ya7, pink) shown in cartoon representation. MeGH168 loop regions that are structurally variable in comparison to Fun168A are shown in navy. (b) Active-site catalytic residues of MeGH168 (cyan) compared with Fun168A (PDB entry 8ya7, pink). (c) Residues in negative subsites (−1, −2) of MeGH168 (cyan) compared with Fun168A (PDB entry 8ya7, pink). (d) Residues in suggested positive subsites (+1, +2, +3) of MeGH168 (cyan) compared with Fun168A (PDB entry 8ya7, pink). The fucoidan tetrasaccharide ligand is represented in green sticks.

3.2. Structure of the catalytic groove

The structure of Fun168A was determined as a product complex bound to α-L-Fucp-1,3-α-L-Fucp2,4(OSO₃⁻)-1,3-α-L-Fucp2(OSO₃⁻)-1,3-α-L-Fucp2(OSO₃⁻) with ligand inter­actions through the −1 to −2 subsites (Chen et al., 2024). The −3 and −4 subsites displayed no ligand interactions as these fucosyl residues did not make direct interactions with the enzyme, making it unclear whether there truly are −3 and −4 subsites; therefore, we opted to omit analysis of these subsites. A comparison of MeGH168 with this structure revealed conservation of the catalytic residues (Asp167 and Glu237 in MeGH168) in structurally conserved loop regions in the central base of the catalytic groove (Fig. 2b). Based on the retaining mechanism of GH168 enzymes, Asp167 is predicted to function as the nucleophile and Glu237 as the acid/base.

Ligand binding in Fun168A involves salt bridges and hydrogen-bond interactions with sulfate groups specifically, allotting importance to the residues involved in these interactions (Chen et al., 2024). As demonstrated in Fig. 2(c), the Fun168A −1 and −2 subsite residues are generally well conserved within MeGH168. Nonconserved residues in MeGH168 still possessed chemically similar side chains, which could facilitate the same CH–π stacking and salt-bridge interactions as observed in Fun168A.

Putative +1, +2 and +3 subsites in Fun168A were identified by molecular-docking studies (Chen et al., 2024). It should be noted that the authors experimentally confirmed +1 site specificity in Fun168A; however, the +2 and +3 subsite interactions remain putative. The potential positive subsite residues in MeGH168 were identified based on their spatial proximity to corresponding subsites in Fun168A. In Fun168A, +1 subsite specificity was attributed to the residues Arg170, Tyr266 and Gln207, which appears to impart strict binding for nonsulfated fucose (Chen et al., 2024). The MeGH168 residues in the closest proximity are Lys130, Arg216 and Gln171, respectively (Fig. 2d). This suggests that the interactions of these residues would likely be similar; however, the presence of Arg216 in comparison to Tyr266 in Fun168A may reflect a somewhat altered set of interactions in this subsite. Residues in the +2 subsite (Arg170, Tyr132 and Asn267 in Fun168A) were not conserved in MeGH168, with the exception of Asn213. Similarly, Asn267, His209 and Tyr266, present in the suggested +3 subsite, were not present in MeGH168. The lack of structural and sequence conservation in the anticipated positive subsites of MeGH168, combined with presence of a large β-hairpin loop insertion (Fig. 2a), suggests that this enzyme may be able to accommodate substrates with alternative chemical characteristics in comparison to Fun168A. However, further investigations into the activity of this enzyme are required to determine the exact substrate specificity.

4. Conclusion

The structure of MeGH168 from the marine bacterium M. eckloniae sp. DK169 belongs to the sparsely characterized GH168 family. The conservation of catalytic and negative-subsite residues suggests similarities to the retaining hydrolysis mechanism of similar fucoidan substrates observed in the GH168 family. However, structural variation in positive subsites located in loop regions demonstrating high structural variability suggests potential differences in substrate specificity in MeGH168 compared with Fun168A.