TY - JOUR
AU - Matsuda,, Akio
AB - Abstract Endo-β-N-acetylglucosaminidases are enzymes that hydrolyze the N,N′-diacetylchitobiose unit of N-glycans. Many endo-β-N-acetylglucosaminidases also exhibit transglycosylation activity, which corresponds to the reverse of the hydrolysis reaction. Because of these activities, some of these enzymes have recently been used as powerful tools for glycan remodeling of glycoproteins. Although many endo-β-N-acetylglucosaminidases have been identified and characterized to date, there are few enzymes that exhibit hydrolysis activity toward multibranched (tetra-antennary or more) complex-type N-glycans on glycoproteins. Therefore, we searched for novel endo-β-N-acetylglucosaminidases that exhibit hydrolysis activity toward multibranched complex-type N-glycans in this study. From database searches, we selected three candidate enzymes from Tannerella species—Endo-Tsp1006, Endo-Tsp1263 and Endo-Tsp1457—and prepared them as recombinant proteins. We analyzed the hydrolysis activity of these enzymes toward N-glycans on glycoproteins and found that Endo-Tsp1006 and Endo-Tsp1263 exhibited hydrolysis activity toward complex-type N-glycans, including multibranched N-glycans, preferentially, whereas Endo-Tsp1457 exhibited hydrolysis activity toward high-mannose-type N-glycans exclusively. We further analyzed substrate specificities of Endo-Tsp1006 and Endo-Tsp1263 using 18 defined glycopeptides as substrates, each having a different N-glycan structure. We found that Endo-Tsp1006 preferred N-glycans with galactose or α2,6-linked sialic acid residues in their nonreducing ends as substrates, whereas Endo-Tsp1263 preferred N-glycans with N-acetylglucosamine residues in their nonreducing ends as substrates. complex-type N-glycan, endo-β-N-acetylglucosaminidase, glycopeptide, glycoprotein, Tannerella Introduction Protein glycosylation is a major type of posttranslational modification, and it can play important roles in protein folding, stability, activity and intermolecular interactions. There are two types of glycosylation: N-glycosylation, which occurs on the asparagine residue in the N-glycosylation consensus sequence NxS/T, where x denotes any amino acids except proline; and O-glycosylation, which occurs on specific serine or threonine residues in the protein. In vertebrates, all N-glycans used in N-glycosylation contain a common trimannosyl N,N′-diacetylchitobiose unit that adheres to the asparagine residue (the so-called trimannosyl core) and are grouped into three major types according to their structure and sugar composition: high-mannose-, complex- and hybrid-type N-glycans. High-mannose-type N-glycans are mainly composed of mannose residues, whereas complex-type N-glycans can contain N-acetylglucosamine, galactose, fucose and sialic acid residues in addition to the trimannosyl core. Hybrid-type N-glycans have both high-mannose- and complex-type features on each side of the N-glycan branches. Endo-β-N-acetylglucosaminidases (ENGases: EC 3.2.1.96) are enzymes that hydrolyze the N,N′-diacetylchitobiose unit of N-glycans. In addition, transglycosylation activity of ENGases, which corresponds to the reverse reaction of the hydrolysis, has been widely reported (Fairbanks 2017). ENGases are classified into two families on the basis of their amino acid sequences in the carbohydrate-active enZymes database (CAZy): (i) glycoside hydrolase family 18 (GH18), which also includes chitinases; and (ii) glycoside hydrolase family 85 (GH85) (Henrissat 1991; Lombard et al. 2014). In general, the GH18 and GH85 family enzymes have a conserved “DxxDxDxE” motif and “NxE” motif (where x denotes any amino acids) in their active sites, respectively (Collin and Olsén 2001; Gloster and Vocadlo 2010; Tzelepis et al. 2015; Fairbanks 2017). Amino acid substitutions in these conserved motifs often depress the hydrolysis activity, which may also lead to increased transglycosylation activity. Thus, mutant enzymes exhibiting strong transglycosylation activity have been used as glycosynthases for N-glycan remodeling to convert heterologously glycosylated proteins to homogenously glycosylated proteins (Fairbanks 2017). Many ENGases have been found in various organisms, from bacteria to mammals, and the enzymatic properties of these enzymes have been characterized to date. Some of these ENGases are widely used for N-glycan analysis and/or N-glycan remodeling on the basis of their substrate specificity. For example, bacterial GH18 family ENGases, such as EndoH from Streptomyces plicatus (Trumbly et al. 1985) and EndoF1 from Elizabethkingia meningoseptica (formerly named Flavobacterium meningosepticum or Chryseobacterium meningosepticum) (Trimble and Tarentino 1991; Tarentino et al. 1993; Kim et al. 2005), hydrolyze high-mannose-type and hybrid-type N-glycans, whereas EndoF2 and EndoF3 from E. meningoseptica preferentially hydrolyze complex-type N-glycans (Fairbanks 2017). GH85 family ENGases, such as Endo-M from the fungus Mucor hiemalis and Endo-CC1 from the basidiomycete fungus Coprinopsis cinerea, exhibit broad substrate specificity toward high-mannose-type, hybrid-type and complex-type N-glycans, whereas Endo-A from Arthrobacter protophormiae and Endo D from Streptococcus pneumoniae exhibit limited substrate specificity toward high-mannose-type and hybrid-type N-glycans, and truncated core structures (paucimannose structures), respectively (Takegawa et al. 1997; Muramatsu et al. 2001; Fujita et al. 2004; Eshima et al. 2015). Their mutants, such as Endo-M N175Q, Endo-CC1 N180H, Endo-A N171A and Endo D N322A/Q have been used as glycosynthases for N-glycan remodeling (Huang et al. 2009; Umekawa et al., 2010; Fan et al. 2012; Eshima et al. 2015). Although various ENGases with different substrate specificities have been reported, there are few enzymes that exhibit hydrolysis activity toward multibranched (tetra-antennary or more) complex-type N-glycans on glycoproteins (Ito et al. 1993; Miyagawa et al. 2007; Ito 2014; Liu et al. 2018). As so many glycoproteins have multibranched complex-type N-glycans, including bisecting-GlcNAc structures, lack of the hydrolysis activity toward such N-glycans represent a weakness in the use of ENGases for remodeling. To improve this situation, we searched for novel ENGases that exhibit hydrolysis activity toward multibranched complex-type N-glycans. In this study, we analyzed three candidate ENGases from Tannerella species, namely, Endo-Tsp1006, Endo-Tsp1263 and Endo-Tsp1457. We found that Endo-Tsp1006 and Endo-Tsp1263 exhibited hydrolysis activity toward complex-type N-glycans, including multibranched N-glycans, preferentially, whereas Endo-Tsp1457 exhibited hydrolysis activity toward high-mannose-type N-glycans exclusively. Both Endo-Tsp1006 and Endo-Tsp1263 had unique substrate specificities, in which Endo-Tsp1006 preferred N-glycans with galactose or α2,6-linked sialic acid residues in their nonreducing ends as substrates, whereas Endo-Tsp1263 preferred N-glycans with N-acetylglucosamine residues in their nonreducing ends as substrates. Results Database searches for novel ENGase candidates Genome analyses of many microorganisms have been performed to date, and these data have been deposited in public databases. We comprehensively searched databases for novel proteins that have features similar to known ENGases and/or have been annotated as a putative ENGase. We identified a group of proteins, most of which had a type IX secretion system C-terminal target domain (Lasica et al. 2017) in their C-terminus, and some of which had previously been annotated as a putative ENGase. The Prevotella ENGases (Ito 2014; Liu et al. 2018), some of which exhibited hydrolysis activity toward multibranched complex-type N-glycans during our study, were also included in this group. We arbitrarily selected some of these protein sequences and generated a phylogenetic tree using the MEGA7 program (Kumar et al. 2016), which led to the discovery of three subgroups of putative bacterial ENGase candidates. The amino acid sequence lengths of the first subgroup members were around 1000 residues, whereas those of the second subgroup members were around 1200–1300 residues, and those of the third subgroup members were over 1300 residues (Figure 1A). Among these ENGase candidates, we selected one member from each group to be analyzed: a protein consisting of 1006 amino acids from Tannerella species CAG:118, which we tentatively designated as Endo-Tsp1006; a protein consisting of 1263 amino acids from Tannerella species CAG:51, which we tentatively designated as Endo-Tsp1263; and a protein consisting of 1457 amino acids from Tannerella species CAG:51, which we tentatively designated as Endo-Tsp1457. Tannerella are gram-negative bacteria that exist in anaerobic environments such as the gut and oral cavity. The Tannerella species CAG:51 and CAG:118 are derived from human gut metagenome. The structural features of the above proteins are shown schematically in Figure 1B. These proteins commonly have a putative signal peptide sequence at their N-terminus, which is involved in the translocation of the protein through the inner membrane to the periplasm. Then, the C-terminal target domain directs the protein for further translocation across the outer membrane through the type IX secretion system. In general, these secreted proteins are either secreted into the extracellular milieu or modified by attachment of anionic lipopolysaccharide resulting in the anchorage of cargo protein to the cell surface (Lasica et al. 2017). This suggests that Tannerella ENGases are secreted or membrane-bound proteins. These proteins seem to belong to the GH85 family, as they contain the putative NxE motif in their amino acid sequences (Figure 1C). Among previously characterized ENGases, the overall amino acid sequence of Endo-Tsp1006 showed 52.2% sequence similarity with a Prevotella ENGase Pme_A7285/EndoPMα, and the sequence of Endo-Tsp1263 showed 37.7 and 29.5% sequence similarities with Pme_A7283/EndoPMγ and Pme_A7284/EndoPMβ, respectively (Liu et al. 2018; Ito 2014). In contrast, the overall amino acid sequence of Endo-Tsp1457 showed lower similarity (19.0–20.9%) with these enzymes. The overall sequence similarity between Tannerella enzymes ranged between 21.2 and 27.7%, whereas in the putative GH85 family catalytic domain this reached up to 34.0%. However, Tannerella and Prevotella ENGase candidates showed only limited similarities with existing representative GH85 family ENGases, such as Endo-M and Endo-CC1 (Fujita et al. 2004; Eshima et al. 2015), suggesting that novel ENGase candidates might evolve from a different ancestral protein than that of known GH85 family members. Fig. 1 Open in new tabDownload slide Structural features of novel ENGase candidates. (A) The phylogenetic tree of novel bacterial ENGase candidates that have a type IX secretion system C-terminal target domain in their C-terminus. Amino acid sequences of representative novel bacterial ENGase candidates were arbitrarily retrieved from public databases. Their accession numbers and amino acid lengths are shown. The phylogenetic tree was generated by the neighbor-joining method using the MEGA7 program (Kumar et al. 2016) with obtained sequences. The enzymes from Tannerella species analyzed in this study are underlined in red. (B) Schematic representation of novel ENGase candidates from Tannerella species. The putative signal peptide sequences are shown by yellow rectangles. The putative GH85 family catalytic domains are shown by blue rectangles. The type IX secretion system (T9SS) C-terminal target domains are shown by red rectangles. (C) Alignment of amino acid sequences of Endo-Tsp1006, Endo-Tsp1263, Endo-Tsp1457, Endo-M and Endo-CC1 around the putative active site is shown. The highly conserved residues are shown in red letters and the putative NxE motifs are indicated with asterisks. This figure is available in black and white in print and in color at Glycobiology online. Fig. 1 Open in new tabDownload slide Structural features of novel ENGase candidates. (A) The phylogenetic tree of novel bacterial ENGase candidates that have a type IX secretion system C-terminal target domain in their C-terminus. Amino acid sequences of representative novel bacterial ENGase candidates were arbitrarily retrieved from public databases. Their accession numbers and amino acid lengths are shown. The phylogenetic tree was generated by the neighbor-joining method using the MEGA7 program (Kumar et al. 2016) with obtained sequences. The enzymes from Tannerella species analyzed in this study are underlined in red. (B) Schematic representation of novel ENGase candidates from Tannerella species. The putative signal peptide sequences are shown by yellow rectangles. The putative GH85 family catalytic domains are shown by blue rectangles. The type IX secretion system (T9SS) C-terminal target domains are shown by red rectangles. (C) Alignment of amino acid sequences of Endo-Tsp1006, Endo-Tsp1263, Endo-Tsp1457, Endo-M and Endo-CC1 around the putative active site is shown. The highly conserved residues are shown in red letters and the putative NxE motifs are indicated with asterisks. This figure is available in black and white in print and in color at Glycobiology online. Expression and characterization of novel ENGase candidates To characterize the enzymatic properties of Endo-Tsp1006, Endo-Tsp1263 and Endo-Tsp1457, we synthesized artificial genes encoding these proteins and constructed their expression vectors. Endo-Tsp1006, Endo-Tsp1263 and Endo-Tsp1457 were produced in Escherichia coli as glutathione S-transferase (GST)-tagged fusion proteins, and affinity-purified from soluble fractions of cell extracts. The GST tags were then removed from the GST fusion proteins by site-specific proteolysis to obtain the purified enzymes (Figure 2A). The yields of Endo-Tsp1006, Endo-Tsp1263 and Endo-Tsp1457 were 590, 380 and 160 μg per 100 mL culture, respectively, under the culture conditions described in the Materials and methods section below. Fig. 2 Open in new tabDownload slide Hydrolysis activity of Tannerella ENGases toward N-glycans on glycoproteins. Tannerella ENGases and glycoproteins were incubated in 50 mM sodium citrate buffer (pH 5.0) at 45 °C for up to 20 h. As controls, EndoH was used for ribonuclease B assays and PNGase F, which is an amidase capable of releasing almost all types of N-glycans on glycoproteins, was used for ovomucoid and α1-acid glycoprotein assays. When using PNGase F, the substrate glycoproteins were denatured according to the manufacturer’s instructions before enzymatic reactions. The purified recombinant Tannerella ENGases (A) and the reaction mixtures (B) were subjected to SDS–PAGE and gels were stained with Coomassie Brilliant Blue. The apparent molecular weight of ribonuclease B is about 17,000, and the addition of EndoH to a solution of ribonuclease B, which released high-mannose-type N-glycans on glycoproteins (Trumbly et al. 1985; Fairbanks 2017), resulted in a decrease of its molecular weight to 15,000. A similar result was obtained by the addition of Endo-Tsp1457 to a solution of ribonuclease B, indicating that Endo-Tsp1457 could release high-mannose-type N-glycans on ribonuclease B efficiently, whereas Endo-Tsp1263 and Endo-Tsp1006 hydrolyzed N-glycans from ribonuclease B only slightly. The apparent molecular weight of ovomucoid is about 35,000, and the addition of PNGase F to the ovomucoid solution resulted in the decrease of its molecular weight to 27,000. Among the examined enzymes, Endo-Tsp1263 could hydrolyze N-glycans from ovomucoid efficiently and decrease the molecular weight of ovomucoid to 25,000. On the other hand, Endo-Tsp1457 and Endo-Tsp1006 did not apparently release N-glycans from ovomucoid. The apparent molecular weight of α1-acid glycoprotein is about 43,000, and the addition of PNGase F to the α1-acid glycoprotein solution resulted in the release of N-glycans and the decrease of its molecular weight to the major form of 23,000 and the minor form of 27,000. A similar result was obtained by the addition of Endo-Tsp1006 to the α1-acid glycoprotein solution, suggesting that Endo-Tsp1006 could release most of complex-type N-glycans on α1-acid glycoprotein, whereas Endo-Tsp1263 released them only slightly, and Endo-Tsp1457 did not release them to a noticeable degree. Lane M, molecular weight marker. Fig. 2 Open in new tabDownload slide Hydrolysis activity of Tannerella ENGases toward N-glycans on glycoproteins. Tannerella ENGases and glycoproteins were incubated in 50 mM sodium citrate buffer (pH 5.0) at 45 °C for up to 20 h. As controls, EndoH was used for ribonuclease B assays and PNGase F, which is an amidase capable of releasing almost all types of N-glycans on glycoproteins, was used for ovomucoid and α1-acid glycoprotein assays. When using PNGase F, the substrate glycoproteins were denatured according to the manufacturer’s instructions before enzymatic reactions. The purified recombinant Tannerella ENGases (A) and the reaction mixtures (B) were subjected to SDS–PAGE and gels were stained with Coomassie Brilliant Blue. The apparent molecular weight of ribonuclease B is about 17,000, and the addition of EndoH to a solution of ribonuclease B, which released high-mannose-type N-glycans on glycoproteins (Trumbly et al. 1985; Fairbanks 2017), resulted in a decrease of its molecular weight to 15,000. A similar result was obtained by the addition of Endo-Tsp1457 to a solution of ribonuclease B, indicating that Endo-Tsp1457 could release high-mannose-type N-glycans on ribonuclease B efficiently, whereas Endo-Tsp1263 and Endo-Tsp1006 hydrolyzed N-glycans from ribonuclease B only slightly. The apparent molecular weight of ovomucoid is about 35,000, and the addition of PNGase F to the ovomucoid solution resulted in the decrease of its molecular weight to 27,000. Among the examined enzymes, Endo-Tsp1263 could hydrolyze N-glycans from ovomucoid efficiently and decrease the molecular weight of ovomucoid to 25,000. On the other hand, Endo-Tsp1457 and Endo-Tsp1006 did not apparently release N-glycans from ovomucoid. The apparent molecular weight of α1-acid glycoprotein is about 43,000, and the addition of PNGase F to the α1-acid glycoprotein solution resulted in the release of N-glycans and the decrease of its molecular weight to the major form of 23,000 and the minor form of 27,000. A similar result was obtained by the addition of Endo-Tsp1006 to the α1-acid glycoprotein solution, suggesting that Endo-Tsp1006 could release most of complex-type N-glycans on α1-acid glycoprotein, whereas Endo-Tsp1263 released them only slightly, and Endo-Tsp1457 did not release them to a noticeable degree. Lane M, molecular weight marker. To examine whether these purified enzymes had hydrolysis activity toward N-glycans on glycoproteins, we used several glycoproteins as substrates. The enzymatic reactions were performed with purified enzymes and the above substrates in 50 mM citrate buffer (pH 5.0) for 20 h at 45 °C. After the enzymatic reactions were stopped, the reaction mixtures were subjected to sodium dodecyl sulfate–polyacrylamide gel electrophoresis (SDS–PAGE) and gels were stained with Coomassie Brilliant Blue (Figure 2B). Bovine pancreatic ribonuclease B has an N-glycosylation site, where high-mannose-type N-glycans are attached (Fu et al. 1994). Among the examined enzymes, addition of Endo-Tsp1457 to a solution of ribonuclease B resulted in its clear band shift, suggesting that high-mannose-type N-glycans on ribonuclease B were released by Endo-Tsp1457. Mass spectrometric analysis of the reaction products detected signals from several species of high-mannose-type N-glycans (M5–M9 type; Supplementary Figure S1), indicating that Endo-Tsp1457 actually released high-mannose-type N-glycans on ribonuclease B. Chicken ovomucoid has five N-glycosylation sites where multibranched complex-type N-glycans are attached (Yamashita et al. 1982, 1983). Among the examined enzymes, addition of Endo-Tsp1263 to a solution of ovomucoid resulted in its clear band shift, suggesting that multibranched complex-type N-glycans on ovomucoid were released by Endo-Tsp1263. Mass spectrometric analysis of the reaction products detected signals from several species of multibranched complex-type N-glycans, including bisecting GlcNAc-containing di-, tri-, tetra- and penta-antennary N-glycans (Supplementary Figure S1), indicating that Endo-Tsp1263 actually released multibranched complex-type N-glycans on ovomucoid. It should be noted that there was a discrepancy between the apparent molecular weights of Endo-Tsp1263 treated- and peptide:N-glycosidase F (PNGase F) treated-ovomucoid, which the former should have a higher molecular weight than the latter because of the residual GlcNAcs on N-glycosylation sites. This suggested that there were residual N-glycans on ovomucoid after PNGase F treatment. However, there were no residual N-glycans and this discrepancy was probably caused by the irreversible changes of the molecular forms of ovomucoid by the denaturing pretreatment when using PNGase F, as the completely linearized ovomucoid by reductive alkylation did not show such a discrepancy (Supplementary Figure S2). Human α1-acid glycoprotein has five N-glycosylation sites where multibranched complex-type N-glycans are attached (Treuheit et al. 1992). Among the examined enzymes, addition of Endo-Tsp1006 to a solution of α1-acid glycoprotein resulted in its clear band shifts of major and minor molecular species, suggesting that multibranched complex-type N-glycans on α1-acid glycoprotein were released by Endo-Tsp1006. Mass spectrometric analysis of the reaction products detected signals from several species of multibranched complex-type N-glycans, including sialylated di-, tri- and tetra-antennary N-glycans (Supplementary Figure S1), indicating that Endo-Tsp1006 actually released multibranched complex-type N-glycans on α1-acid glycoprotein. Similar results were obtained for enzyme reactions toward complex-type N-glycan-containing glycoproteins such as human serum transferrin, which mainly contains bi-antennary sialylated complex-type N-glycans (Nagae et al. 2014); bovine fetuin, which mainly contains tri- and bi-antennary sialylated complex-type N-glycans (Green et al. 1988; Lin et al. 2018); and human prostate-specific antigen, which mainly contains tri- and bi-antennary complex-type N-glycans, some of which are α1,6-core-fucosylated (Tabarés et al. 2006; White et al. 2009) (Supplementary Figure S3). Taken together, these results suggest that Endo-Tsp1457 prefers high-mannose-type N-glycans as substrates, whereas Endo-Tsp1263 and Endo-Tsp1006 prefer complex-type N-glycans as substrates, although their detailed structural preferences toward complex-type N-glycans were different from each other. It should be noted that N-glycans on some glycoproteins, such as immunoglobulin G (Trastuzumab), which mainly contains bi-antennary complex-type N-glycans (Sanchez-De Melo et al. 2015), and ovalbumin, which mainly contains high-mannose-type and hybrid-type N-glycans (Harvey et al. 2000), could not be hydrolyzed efficiently by these enzymes without the denature treatment (Supplementary Figure S3), suggesting that enzyme activities of Tannerella ENGases are affected by the three-dimensional structure of glycoproteins and the position of N-glycosylation sites. Using α1-acid glycoprotein, ovomucoid and ribonuclease B as substrates for Endo-Tsp1006, Endo-Tsp1263 and Endo-Tsp1457, respectively, we determined enzymatic properties of these enzymes. The optimal pH values for Endo-Tsp1006, Endo-Tsp1263 and Endo-Tsp1457 were pH 5.0, 5.0–6.0 and pH 5.0, respectively. Endo-Tsp1006, Endo-Tsp1263 and Endo-Tsp1457 were stable at least between pH 4.0 and 11.0, pH 4.0 and 11.0 and pH 4.0 and 12.0, respectively, at 4 °C for 24 h. The optimal temperatures of Endo-Tsp1006, Endo-Tsp1263 and Endo-Tsp1457 were ~45, ~45 and 45–65 °C, respectively. Endo-Tsp1006, Endo-Tsp1263 and Endo-Tsp1457 were stable up to 45, 45 and 60 °C for at least 1 h without substrates, respectively (Supplementary Figure S4). Substrate specificities of Endo-Tsp1006 and Endo-Tsp1263 toward glycopeptides Because Endo-Tsp1006 and Endo-Tsp1263 exhibited different substrate specificities toward complex-type N-glycans on glycoproteins, we examined their substrate specificities further using glycopeptides as substrates. We enzymatically prepared 17 species of glycopeptides with different structures of N-glycans using hen egg yolk sialyl glycopeptide (SGP) as starting material in combination with a set of recombinant glycosyltransferases, as described in the Materials and methods section; the glycopeptides are listed in Supplementary Figure S5. The hydrolysis activities of Endo-Tsp1006 and Endo-Tsp1263 toward N-glycans on these glycopeptides are summarized in Figure 3. Endo-Tsp1006 exhibited the highest activity (>300 mU/mg) toward α2,6-sialylated tetra-antennary complex-type N-glycan (α2,6A(2–4)G4) and bisecting GlcNAc-containing galactosylated bi-antennary complex-type N-glycan (G2B) on the SGP-derived peptide. It also exhibited relatively high activity (200–300 mU/mg) toward mono-galactosylated bi-antennary complex-type N-glycan (G1b), α2,6-sialylated bi-antennary complex-type N-glycan (α2,6A2) and galactosylated tri-antennary complex-type N-glycan (G3GN3b) on the SGP-derived peptide, and moderate activity (100–200 mU/mg) toward galactosylated tetra-antennary complex-type N-glycan (G4GN4), galactosylated bi-antennary complex-type N-glycan with core-fucose (G2F), galactosylated bi-antennary complex-type N-glycan (G2) and galactosylated tri-antennary complex-type N-glycan (G3GN3a) on the SGP-derived peptide. On the other hand, Endo-Tsp1006 exhibited low activity (25–100 mU/mg) toward mono-galactosylated bi-antennary complex-type N-glycan (G1a) and agalacto bi-antennary complex-type N-glycan with core-fucose (G0F) on the SGP-derived peptide, and trace or no activity toward agalacto bi-antennary complex-type N-glycan (G0), agalacto tri-antennary complex-type N-glycan (GN3a and GN3b), agalacto tetra-antennary complex-type N-glycan (GN4), bisecting GlcNAc-containing agalacto bi-antennary complex-type N-glycan (G0B), α2,3-sialylated bi-antennary complex-type N-glycan (α2,3A2) and tri-mannose core structure (M3) on the SGP-derived peptide. Fig. 3 Open in new tabDownload slide Hydrolysis activity of Endo-Tsp1006 and Endo-Tsp1263 toward N-glycans on glycopeptides. The hydrolysis activity toward N-glycans on glycopeptides was measured in 50 mM sodium citrate (pH 5.0), 0.25 mM glycopeptide and enzyme preparation (0.25 μg) in a total volume of 10 μL, and incubated at 45 °C. For measurements of specific activity, the enzyme reactions were quenched after 15 min with heating at 100 °C for 3 min. One unit (μmol/min) was defined as the enzyme activity that catalyzes the conversion of 1 μmol of substrate per min. The detailed N-glycan structures of glycopeptide substrates are shown in Supplementary Figure S5. Monosaccharide symbols follow the Symbol Nomenclature for Glycans system (PMID 26543186, Glycobiology 25: 1323–1324, 2015) details at NCBI. Fig. 3 Open in new tabDownload slide Hydrolysis activity of Endo-Tsp1006 and Endo-Tsp1263 toward N-glycans on glycopeptides. The hydrolysis activity toward N-glycans on glycopeptides was measured in 50 mM sodium citrate (pH 5.0), 0.25 mM glycopeptide and enzyme preparation (0.25 μg) in a total volume of 10 μL, and incubated at 45 °C. For measurements of specific activity, the enzyme reactions were quenched after 15 min with heating at 100 °C for 3 min. One unit (μmol/min) was defined as the enzyme activity that catalyzes the conversion of 1 μmol of substrate per min. The detailed N-glycan structures of glycopeptide substrates are shown in Supplementary Figure S5. Monosaccharide symbols follow the Symbol Nomenclature for Glycans system (PMID 26543186, Glycobiology 25: 1323–1324, 2015) details at NCBI. Endo-Tsp1263 exhibited the highest activity (214.6 mU/mg) toward bisecting GlcNAc-containing agalacto bi-antennary complex-type N-glycan (G0B) on the SGP-derived peptide. It also exhibited moderate activity toward agalacto bi-antennary complex-type N-glycans with and without core-fucose (G0F, G0), mono-galactosylated bi-antennary complex-type N-glycan (G1a), agalacto tri-antennary complex-type N-glycan (GN3a) and agalacto tetra-antennary complex-type N-glycan (GN4) on the SGP-derived peptide. Endo-Tsp1263 exhibited low activity toward agalacto tri-antennary complex-type N-glycan (GN3b), mono-galactosylated bi-antennary complex-type N-glycan (G1b) and galactosylated bi-antennary complex-type N-glycan with core-fucose (G2F) on the SGP-derived peptide, and trace or no activity toward galactosylated bi-antennary complex-type N-glycan (G2), galactosylated tri-antennary complex-type N-glycans (G3GN3a and G3GN3b), galactosylated tetra-antennary complex-type N-glycan (G4GN4), galactosylated bisecting GlcNAc-containing bi-antennary complex-type N-glycan (G2B), sialylated complex-type N-glycans (α2,3A2, α2,6A2 and α2,6A(2–4)G4) and tri-mannose core structure (M3) on the SGP-derived peptide. It should be noted that like Endo-Tsp1006 and Endo-Tsp1263, Endo-Tsp1457 also exhibited no activity toward tri-mannose core structure on the SGP-derived peptide (data not shown). For the structural isomers, Endo-Tsp1006 preferred G1b bi-antennary complex-type N-glycan, which is mono-galactosylated on the α1,3-arm side of tri-mannose core, to G1a bi-antennary complex-type N-glycan, which is mono-galactosylated on the α1,6-arm side of tri-mannose core, whereas Endo-Tsp1263 preferred G1a bi-antennary complex-type N-glycan to G1b bi-antennary complex-type N-glycan. Endo-Tsp1006 also preferred G3GN3b tri-antennary complex-type N-glycan, which is branched on the α1,3-arm side of tri-mannose core, to G3GN3a tri-antennary complex-type N-glycan, which is branched on the α1,6-arm side of tri-mannose core, whereas Endo-Tsp1263 preferred G3GN3a tri-antennary complex-type N-glycan to G3GN3a tri-antennary complex-type N-glycan. Altogether, it can be said that the substrate specificities of Endo-Tsp1006 and Endo-Tsp1263 are complementary in a sense. Hydrolysis activity toward N-glycans on galactosylated ovomucoid As shown in Figure 2, Endo-Tsp1263 could hydrolyze N-glycans on ovomucoid efficiently, whereas Endo-Tsp1006 could not. This was likely because most of the nonreducing ends of the N-glycans on ovomucoid were neither α2,6-sialylated nor galactosylated (Yamashita et al. 1982, 1983), and were not preferable substrates for Endo-Tsp1006, judging from the above results. Therefore, we prepared galactosylated ovomucoid using a human galactosyltransferase B4GalT1 (Figure 4A) and examined whether N-glycans on galactosylated ovomucoid could be released by Endo-Tsp1006. As shown in Figure 4B, Endo-Tsp1006 could release N-glycans on galactosylated ovomucoid efficiently, whereas Endo-Tsp1263 could not. Fig. 4 Open in new tabDownload slide The hydrolysis activity of Endo-Tsp1006 and Endo-Tsp1263 toward N-glycans on galactosylated ovomucoid. Galactosylated ovomucoid was prepared as described in the Materials and methods section. Endo-Tsp1006 or Endo-Tsp1263 and galactosylated ovomucoid were incubated in 50 mM sodium citrate buffer (pH 5.0) at 45 °C for 23 h. As a control, PNGase F treatment was performed toward galactosylated ovomucoid. The reaction mixtures were subjected to SDS–PAGE and the gel was stained with Coomassie Brilliant Blue. Lane M, molecular weight marker. (A) Preparation of galactosylated ovomucoid. (B) The hydrolysis activity of Tannerella ENGases toward N-glycans on galactosylated ovomucoid. Fig. 4 Open in new tabDownload slide The hydrolysis activity of Endo-Tsp1006 and Endo-Tsp1263 toward N-glycans on galactosylated ovomucoid. Galactosylated ovomucoid was prepared as described in the Materials and methods section. Endo-Tsp1006 or Endo-Tsp1263 and galactosylated ovomucoid were incubated in 50 mM sodium citrate buffer (pH 5.0) at 45 °C for 23 h. As a control, PNGase F treatment was performed toward galactosylated ovomucoid. The reaction mixtures were subjected to SDS–PAGE and the gel was stained with Coomassie Brilliant Blue. Lane M, molecular weight marker. (A) Preparation of galactosylated ovomucoid. (B) The hydrolysis activity of Tannerella ENGases toward N-glycans on galactosylated ovomucoid. Comparison of the hydrolysis activity toward N-glycans on glycoproteins among several ENGases To examine whether the substrate specificities of Endo-Tsp1006 and Endo-Tsp1263 are unique, we compared hydrolysis activity toward N-glycans on native glycoproteins among several ENGases (Figure 5). In this assay, we used commercially available ENGases such as Endo-M, Endo-CC1, EndoH, EndoF1, EndoF2, EndoF3 and Tannerella ENGases cloned in this study. As a control, PNGase F was employed to release N-glycans on glycoproteins under both undenatured and denatured conditions. All of these ENGases, except Endo-Tsp1006 and Endo-Tsp1263, efficiently released high-mannose-type N-glycans on ribonuclease B (Fu et al. 1994). Among the examined ENGases, Endo-M, Endo-CC1, EndoF2 and EndoF3 are known to have hydrolysis activity toward complex-type N-glycans, and Endo-M, Endo-CC1 and Endo-Tsp1263 partially released complex-type N-glycans on prostate-specific antigen (Tabarés et al. 2006; White et al. 2009), whereas EndoF2, EndoF3 and Endo-Tsp1006 efficiently released them. However, against ovomucoid, most of whose N-glycan structures are asialo-agalacto complex-type N-glycans containing a bisecting-GlcNAc structure (Yamashita et al. 1982, 1983), only Endo-Tsp1263 was able to display hydrolysis activity. With α1-acid glycoprotein, most of whose N-glycan structures are multibranched sialylated complex-type N-glycans (Treuheit et al. 1992), Endo-M, Endo-CC1, EndoF2 and EndoF3 partially released the glycans and only Endo-Tsp1006 efficiently released them. Fig. 5 Open in new tabDownload slide Comparison of the hydrolysis activity toward N-glycans on glycoproteins among several ENGases. Each ENGase and glycoprotein substrates (ribonuclease B, prostate specific antigen, ovomucoid and α1-acid glycoprotein) were incubated under the reaction conditions described in the Materials and methods section. As a control, PNGase F treatments were performed toward glycoprotein substrates with and without denatured conditions. The reaction mixtures were subjected to SDS–PAGE and gels were stained with Coomassie Brilliant Blue. Lane M, molecular weight marker. The protein bands indicated by asterisks correspond to the ENGases used in the assay. Fig. 5 Open in new tabDownload slide Comparison of the hydrolysis activity toward N-glycans on glycoproteins among several ENGases. Each ENGase and glycoprotein substrates (ribonuclease B, prostate specific antigen, ovomucoid and α1-acid glycoprotein) were incubated under the reaction conditions described in the Materials and methods section. As a control, PNGase F treatments were performed toward glycoprotein substrates with and without denatured conditions. The reaction mixtures were subjected to SDS–PAGE and gels were stained with Coomassie Brilliant Blue. Lane M, molecular weight marker. The protein bands indicated by asterisks correspond to the ENGases used in the assay. Discussion In this study, we identified three novel ENGases from Tannerella species and characterized their enzymatic properties. Endo-Tsp1006 and Endo-Tsp1263 preferentially hydrolyzed complex-type N-glycans, but their substrate specificities were different from each other, whereas Endo-Tsp1457 exhibited hydrolysis activity toward high-mannose-type N-glycans exclusively. It appears that Tannerella species produce several ENGases with different substrate specificities to efficiently hydrolyze a range of N-glycans through their cooperative activity. The hydrolysis activity of Endo-Tsp1263 toward N-glycans on ovomucoid and that of Endo-Tsp1006 toward N-glycans on α1-acid glycoprotein and galactosylated ovomucoid indicated that these enzymes are capable of releasing multibranched complex-type N-glycans, including bisecting GlcNAc-containing penta-antennary N-glycans, on native glycoproteins (Figures 2 and 4, Supplementary Figure S1). As shown in Figure 5, many commercially available ENGases cannot efficiently release multibranched N-glycans from glycoproteins. To our knowledge, this is the first report of ENGase activity that can release bisecting GlcNAc-containing penta-antennary N-glycans from a native glycoprotein. We also found unique substrate specificities of these enzymes, as Endo-Tsp1006 preferred galactosylated and α2,6-sialylated complex-type N-glycans as substrates, whereas Endo-Tsp1263 preferred complex-type N-glycans with N-acetylglucosamine residues in their nonreducing ends as substrates. It seems that Endo-Tsp1006 and Endo-Tsp1263 strictly recognize the nonreducing end structures of complex-type N-glycans. As far as we know, such properties have not been reported for most of the previously characterized ENGases with hydrolysis activity toward complex-type N-glycans. Therefore, these enzymes can be used for complex-type N-glycan release from glycoproteins in a nonreducing end structure-specific manner. It can be also said that if these enzymes are used together, most complex-type N-glycans, except α2,3-sialylated N-glycans, can be released from glycoproteins efficiently. On the other hand, Tannerella ENGases analyzed in this study have weak points that they cannot efficiently release oligosaccharides such as tri-mannose core structure (Figure 3) and its β1,2-xylosylated and α1,3-fucosylated form on horseradish peroxidase (Du et al. 2015; data not shown), and probably hybrid-type N-glycans (Supplementary Figure S3). Among ENGases that have hydrolysis activity toward complex-type N-glycans, some exhibit specific preference toward core-fucosylated complex-type N-glycans. For example, core-fucosylated complex-type N-glycans are preferable substrates for EndoF3, Endo-CoM, Endo-BB and Endo-SB-ORF1188, whereas noncore-fucosylated complex-type N-glycans are very poor substrates for these enzymes (Giddens et al. 2016; Huang et al. 2018). In contrast, Endo-M does not show hydrolysis activity toward core-fucosylated complex-type N-glycans, but it shows hydrolysis activity toward noncore-fucosylated complex-type N-glycans (Katoh et al. 2016). For Endo-Tsp1006 and Endo-Tsp1263, it seems that they can hydrolyze both core-fucosylated and noncore-fucosylated complex-type N-glycans, judging from their hydrolysis activity toward glycopeptides such as G0-P, G0F-P, G2-P and G2F-P (Figure 3). As the core-fucosylation levels are different for each glycoprotein, it seems that this substrate specificity of Tannerella ENGases is advantageous for use in releasing complex-type N-glycans from any glycoprotein. These enzymatic properties suggest that Endo-Tsp1006 and Endo-Tsp1263 may become powerful tools for N-glycan analysis and/or N-glycan remodeling of glycoproteins. Concerning this issue, we replaced the asparagine residue in the GH85 family-conserved NxE motif of Tannerella ENGases with glutamine or alanine and generated putative glycosynthase mutants of Endo-Tsp1457, Endo-Tsp1263 and Endo-Tsp-1006. However, as far as we examined, none of these mutants exhibited transglycosylation activity toward glycoprotein acceptors with sugar oxazolines as donor substrates. As Tannerella ENGases are thought to be evolutionally different from the conventional GH85 family ENGases such as Endo-M, active glycosynthase mutants may not be easily obtained by simply mutating the NxE motif. In this study, we grouped novel bacterial ENGase candidates that have a type IX secretion system C-terminal target domain into three subgroups (Figure 1). In addition to the Tannerella ENGases analyzed in this study, we investigated the substrate specificities of candidate ENGases in each subgroup and found that they showed similar substrate specificity to the Tannerella ENGases analyzed in this study from the same subgroup (data not shown). Specifically, the members of subgroup 1, whose amino acid lengths were around 1000 residues, exhibited hydrolysis activity toward galactosylated and α2,6-sialylated complex-type N-glycans; the members of subgroup 2, whose amino acid lengths were around 1200–1300 residues, exhibited hydrolysis activity toward complex-type N-glycans with N-acetylglucosamine residues in their nonreducing ends and the members of subgroup 3, whose amino acid lengths were over 1300 residues, exhibited hydrolysis activity toward high-mannose-type N-glycans preferentially. The type IX secretion system, also known as the Por secretion system, is a complex translocon found only in some species of the Bacteroidetes phylum (Lasica et al. 2017). In human oral pathogens, such as Porphyromonas gingivalis, Prevotella intermedia and Tannerella forsythia, which are major causative agents of periodontitis, this system is involved in virulence factor secretion. Although the physiological functions of Tannerella ENGases analyzed in this study are unknown at present, they may be involved in the host-pathogen interactions as they have a type IX secretion system C-terminal target domain. It is not certain why Tannerella species lack a single ENGase that exhibits hydrolysis activity toward all types of complex-type N-glycans with regard to their nonreducing end structures. Although detailed structural mechanisms have yet to be elucidated, it is possible that different types of ENGases are required for efficient hydrolysis of bulky multibranched complex-type N-glycans, such as bisecting GlcNAc-containing tetra- or penta-antennary complex-type N-glycans. As Endo-Tsp1263 is 257 residues longer than Endo-Tsp1006, this extra region of Endo-Tsp1263 may have an important role for recognition of GlcNAc residues at the nonreducing end of complex-type N-glycans. Structural analyses concerning why Tannerella ENGases exhibit different substrate specificities and how to obtain active glycosynthases are underway. Materials and methods Materials Hen egg yolk SGP was obtained from Fushimi Pharmaceutical Co., Ltd. (Kagawa, Japan). Anti-HER2 monoclonal antibody (trastuzumab) as an immunoglobulin G was obtained from Hoffmann-La Roche (Basel, Switzerland). Fetuin from fetal calf serum, ribonuclease B from bovine pancreas, transferrin and α1-acid glycoprotein from human, and ovomucoid (trypsin inhibitor) and albumin (ovalbumin) from chicken egg white were purchased from Sigma-Aldrich (Milwaukee, WI). Prostate-specific antigen was obtained from BBI solutions (Crumlin, UK). Horseradish peroxidase was purchased from FUJIFILM Wako (Osaka, Japan). EndoH, β1,4-galactosidase, α2–3,6,8 neuraminidase and N-acetylglucosaminidase S were obtained from New England Biolabs (NEB: Ipswich, MA). Endo-M was purchased from Tokyo Chemical Industry Co., Ltd. (Tokyo, Japan). Glycopeptidase F (peptide:N-glycosidase F, PNGase F) was purchased from Takara Bio Inc. (Shiga, Japan) and NEB. HEK293 cells were obtained from RIKEN BioResource Research Center. Dulbecco’s Modified Eagle’s Medium (DMEM) was purchased from Thermo Fischer Scientific Inc. (Waltham, MA). Fetal bovine serum was purchased from Nichirei Biosciences Inc. (Tokyo, Japan). G418 was purchased from AdipoGen (San Diego, CA). The DDDDK-tagged protein purification gel was purchased from Medical and Biological Laboratories Co., Ltd. (MBL: Nagoya, Japan). Uridine 5′-diphospho-α-D-N-acetylglucosamine disodium salt (UDP-GlcNAc, 2Na), uridine 5′-diphospho-α-D-galactose disodium salt (UDP-Gal, 2Na), guanosine 5′-diphospho-β-L-fucose disodium salt (GDP-Fuc, 2Na) and cytidine 5′-monophospho-β-D-N-acetylneuraminic acid disodium salt (CMP-NeuAc, 2Na) were purchased from Toyobo (Osaka, Japan). Database search for novel ENGases Amino acid sequences of novel ENGase candidates were retrieved from the National Center for Biotechnology Information database, UniProt Knowledgebase and CAZy database. A phylogenetic tree was generated by the neighbor-joining method using the MEGA7 program (Kumar et al. 2016). Cloning, expression and purification of recombinant Tannerella ENGases Three candidate ENGases—GenBank accession no. CCY37287 from Tannerella species CAG:118, GenBank accession no. CDD88945 from Tannerella species CAG:51 and GenBank accession no. CDD89351 from Tannerella species CAG:51—were selected from databases and tentatively designated as Endo-Tsp1006, Endo-Tsp1263 and Endo-Tsp1457, respectively. The artificial genes for Tannerella ENGases Endo-Tsp1006, Endo-Tsp1263 and Endo-Tsp1457 were synthesized and cloned into pGEX-6P-1 vector (GE Healthcare Japan, Tokyo, Japan). E. coli strain BL21 (DE3) cells were transformed with these expression vectors. The transformants of each expression vector were cultured in 100 mL of LB broth (Lennox) (Sigma-Aldrich) containing 50 μg/mL ampicillin at 37 °C until the cells reached an absorption at 600 nm of 0.5–0.8. Then, 0.1 mM isopropyl-β-D-thiogalactopyranoside was added to the cultures to induce GST-tagged recombinant Tannerella ENGase overexpression. After further incubation at 25 °C for 3 h, the cells were harvested by centrifugation and lysed by sonication in 5 mL of phosphate buffered saline (PBS). Triton X-100 (Sigma-Aldrich) was then added to the lysates to a final concentration of 1% and mixed gently at room temperature for 30 min to solubilize proteins. The lysates were clarified by centrifugation, and the appropriate amount of Glutathione Sepharose™ beads (GE Healthcare) was added to the supernatants and mixed gently for 16 h at 4 °C to adsorb the GST-tagged proteins. Afterward, the beads were washed 10 times with PBS, and PreScission™ protease (GE Healthcare) treatment was performed to remove the GST tag from the adsorbed GST fusion proteins according to the manufacturer’s instructions. Finally, the affinity-purified recombinant Tannerella ENGases were obtained and concentrated using Vivaspin 500 MWCO 30 K centrifugal concentrators (Sartorius AG, Goettingen, Germany). Construction of expression vectors of human glycosyltransferases For production of FLAG-tagged (corresponding to the amino acid sequence of DYKDDDDK) enzymes, the DNA fragments that encode the luminal domains of human glycosyltransferases (N-acetylglucosaminyltransferases GnT-III, GnT-IVa, GnT-V, β1,4-galactosyltransferase B4GalT1, α1,6-fucosyltransferase FUT8, α2,3-sialyltransferase ST3Gal-IV and α2,6-sialyltransferase ST6Gal-I) were amplified by PCR with KAPA HiFi HotStart ReadyMix PCR Kit (KAPA Biosystems, Boston, MA) or KOD FX DNA polymerase (Toyobo) and cloned into p3xFLAG-CMV-9 vector (Sigma-Aldrich) using restriction enzyme sites in the following PCR primers. For amplification of the GnT-III gene DNA fragment, 5′-TCCAAGCTTCACTTCTTCAAGACCCTGTCC-3′ (forward primer) and 5′-CATGGATCCCTAGACTTCCGCCTCGTCCAGTTT-3′ (reverse primer) were used. For amplification of the GnT-IVa cDNA fragment, 5′-ACTAAGCTTCAAAATGGGAAAGAAAAACTG-3′ (forward primer) and 5′-CTCGGTACCTCAGTTGGTGGCTTTTTTAATATG-3′ (reverse primer) were used. For amplification of the GnT-V cDNA fragment, 5′-ATGGCGGCCGCGCACTTTACCATCCAGCAGCGA-3′ (forward primer) and 5′-AGGTCTAGACTATAGGCAGTCTTTGCAGAGAGC-3′ (reverse primer) were used. For amplification of the B4GalT1 cDNA fragment, 5′-CTGAAGCTTCGCGACCTGAGCCGCCTGCCC-3′ (forward primer) and 5′-ACCTCTAGACTAGCTCGGTGTCCCGATGTC-3′ (reverse primer) were used. For amplification of the FUT8 cDNA fragment, 5′-TGCGGCCGCGAATTCACGAGATAATGACCATCCTGAT-3′ (forward primer) and 5′-CCGGGATCCTCTAGATTATTTCTCAGCCTCAGGATAT-3′ (reverse primer) were used. For amplification of the ST3Gal-IV cDNA fragment, 5′-TCCAAGCTTCGGGAAGACAGGTACATCGAG-3′ (forward primer) and 5′-TTGGGATCCTCAGAAGGACGTGAGGTTCTT-3′ (reverse primer) were used. For amplification of the ST6Gal-I cDNA fragment, 5′-TGTAAGCTTAAGGAAAAGAAGAAAGGGAGT-3′ (forward primer) and 5′-AGCGGATCCTTAGCAGTGAATGGTCCGGAA-3′ (reverse primer) were used. Transfection, establishment of stable transfectants producing soluble human glycosyltransferases, and preparation of soluble glycosyltransferases HEK293 cells were cultured in DMEM with 10% (v/v) fetal bovine serum in 5% CO2 at 37 °C and transfected with the expression vectors constructed above using FuGENE® 6 transfection reagent (Promega, WI). After 48 h, cells were trypsinized and diluted in the above medium containing 500 μg/mL G418 and cultured for 3 weeks. The stable transfectants producing soluble human glycosyltransferases were isolated and propagated. The culture media of stable transfectants were collected and centrifuged to remove cell debris, and then DDDDK-tagged protein purification gels (Medical & Biological Laboratories Co., Ltd., Nagoya, Japan) were added to media and mixed gently for 16 h at 4 °C to adsorb FLAG-tagged soluble glycosyltransferases. The gels were collected by centrifugation and washed 10 times with PBS, and adsorbed FLAG-tagged proteins were eluted according to the manufacturer’s instructions. Finally, the affinity-purified soluble glycosyltransferases were obtained and concentrated using Vivaspin 500 MWCO 30 K centrifugal concentrators (Sartorius AG, Goettingen, Germany). Preparation of glycopeptides Glycopeptides used in this study were prepared from SGP (α2,6A2-P) consisting of α2,6-disialylated bi-antennary N-glycan (A2) with a specific amino acid sequence (KVANKT). The glycopeptides containing asialo-bi-antennary N-glycan (G2-P) and asialo-agalacto-bi-antennary N-glycan (G0-P) were prepared as described previously (Kurogochi et al. 2015). The glycopeptides containing mono-galactosylated bi-antennary N-glycan (G1a-P and G1b-P) were prepared from SGP by treatments with mild acid, β-galactosidase and α2–3,6,8 neuraminidase (Tsukimura et al. 2017; Aoyama et al. 2019). To obtain the glycopeptide containing tri-mannosyl core N-glycan (M3-P), G0-P was treated with N-acetylglucosaminidase S (NEB) according to the manufacturer’s instructions. To obtain further modified glycopeptides, appropriate amount of glycosyltransferases (0.04–100 μg; FUT8, GnT-III, GnT-IVa, GnT-V, B4GalT1, ST3Gal-IV and ST6Gal-I), 2.5 mM acceptor glycopeptide and 10–40 mM nucleotide sugar were added in the reaction buffer (25 mM MOPS-NaOH (pH 7.4), 5 mM MnCl2, 25 mM NaCl, 0.05–1 mg/mL bovine serum albumin, in a total of 0.02–2 mL) and incubated at 37 °C. The reactions were monitored by mass spectrometric analysis and stopped when the reactions had completed. To obtain glycopeptides containing α1,6-core-fucosylated asialo-agalacto-bi-antennary N-glycans (G0F-P), G0-P was treated with FUT8 and GDP-Fuc. To obtain glycopeptides containing α1,6-core-fucosylated asialo-bi-antennary N-glycans (G2F-P), G0F-P was treated with B4GalT1 and UDP-Gal. Similarly, G2-P was also prepared by B4GalT1 from G0-P when necessary. To obtain glycopeptides containing α2,3-sialylated-bi-antennary N-glycans (α2,3A2-P), G2-P was treated with ST3Gal-IV and CMP-NeuAc. To obtain glycopeptides containing bisected asialo-agalacto-bi-antennary N-glycans (G0B-P), G0-P was treated with GnT-III and UDP-GlcNAc. To obtain glycopeptides containing bisected asialo-bi-antennary N-glycans (G2B-P), G0B-P was treated with B4GalT1 and UDP-Gal. To obtain glycopeptides containing asialo-agalacto-2,4-branched tri-antennary N-glycans (GN3b-P), G0-P was treated with GnT-IVa and UDP-GlcNAc. To obtain glycopeptides containing asialo-2,4-branched tri-antennary N-glycans (G3GN3b-P), GN3b-P was treated with B4GalT1 and UDP-Gal. To obtain glycopeptides containing asialo-agalacto-2,6-branched tri-antennary N-glycans (GN3a-P), G0-P was treated with GnT-V and UDP-GlcNAc. To obtain glycopeptides containing asialo-2,6-branched tri-antennary N-glycans (G3GN3a-P), GN3a-P was treated with B4GalT1 and UDP-Gal. To obtain glycopeptides containing asialo-agalacto-tetra-antennary N-glycans (GN4-P), GN3a-P was treated with GnT-IVa and UDP-GlcNAc. To obtain glycopeptides containing asialo-tetra-antennary N-glycans (G4GN4-P), GN4-P was treated with B4GalT1 and UDP-Gal. To obtain glycopeptides containing α2,6-sialylated-tetra-antennary N-glycans (α2,6AnG4-P [n = 2–4]), G4GN4-P was treated with ST6Gal-I and CMP-NeuAc, respectively. Structures of N-glycans attached to glycopeptides were confirmed by mass spectroscopy, as described previously (Kurogochi and Amano 2014; Kurogochi et al. 2015; Aoyama et al. 2019). Examples for the preparation procedures of glycopeptides are summarized in Supplementary Figure S6. Preparation of galactosylated ovomucoid To obtain galactosylated ovomucoid, B4GalT1 (4 μg), ovomucoid (32 μg) and 10 mM UDP-Gal were added in the reaction buffer [25 mM MOPS-NaOH (pH 7.4), 1 mM MnCl2, 25 mM NaCl, in a total of 20 μL] and incubated at 37 °C for 48 h. The reaction was monitored by SDS–PAGE and confirmed by mass spectroscopy, as described previously (Kurogochi and Amano 2014; Kurogochi et al. 2015; Aoyama et al. 2019). Preparation of ENGases Endo-CC1 was prepared as described previously, with some modifications (Eshima et al. 2015). EndoF1 and EndoF2 were expressed as the maltose-binding protein fusion proteins in E. coli (Reddy et al. 1998), and recombinant enzymes were prepared using the pMAL™ Protein Fusion and Purification System (NEB) according to the manufacturer’s instructions. EndoF3 was prepared in a similar method, as described previously (Aoyama et al. 2019). Enzyme activity assay Mass spectrometric analyses of N-glycans attached to glycopeptides or released from glycoproteins were performed as described previously (Kurogochi and Amano 2014; Kurogochi et al. 2015; Aoyama et al. 2019). The hydrolysis activity toward N-glycans on glycopeptides was measured in 50 mM sodium citrate (pH 5.0), 0.25 mM glycopeptide, and enzyme preparation (0.25 μg) in a total volume of 10 μL. The enzyme reaction was performed at 45 °C for up to 2 h. The reaction was terminated by boiling the reaction mixture for 3 min. Without purification, the glycopeptides (substrates) in the reaction mixture were directly measured and the specific activity was determined by quantitative analysis of glycopeptides using a liquid chromatography electrospray ionization-mass spectrometry (LC ESI-MS). The samples before and 15 min after the addition of enzyme were subjected to the analysis by LC ESI-MS using LC systems (Dionex, Sunnyvale, CA) with normal-phase column (InertSustain Amide 3 μm, 1.5 × 150 mm, GL Sciences Inc. Tokyo, Japan) and absorbed with 75% acetonitrile +20 mM ammonium formate (0–3 min), eluted with gradients of 75–45% acetonitrile (3–13 min), washed with gradients of 45–25% acetonitrile (13–15 min) and 25% acetonitrile (15–20 min), and equilibrated with 75% acetonitrile +20 mM ammonium formate (20–35 min) at a constant flow of 125 μL/min (column oven at 45 °C). Electrospray MS data were collected using the HESI-II probe ion source on an LTQ-Velos Pro instrument (Thermo Fisher Scientific, Waltham, MA) in positive and negative mode. The spray voltage was set at 3.5 kV for positive mode and 2.3 kV for negative mode, with a temperature of 150 °C, and Sheath gas flow rate and Aux gas flow rate were set at 20 and 5 arb, respectively. The peaks and masses were integrated using annotation procedures (Qual Browser) in the X calibur 2.2 SP 1.48 software. The change in molar amount of the targeted glycopeptide (decreasing of glycopeptide) was quantitatively calculated from two-point analysis (initial velocity during the progression of the enzyme reaction) and expressed as the enzyme activity. Examples for the time-course of enzyme reaction monitored by LC ESI-MS are shown in Supplementary Figure S7. The specific activity (mU/mg) for a substrate was determined by the calculation of enzymatic activity (mU) per the amount of enzyme (mg). One unit (μmol/min) was defined as the enzyme activity that catalyzes the conversion of 1 μmol of substrate per minute. The hydrolysis activities of Tannerella ENGases toward N-glycans on glycoproteins (ribonuclease B, ovomucoid, α1-acid glycoprotein, transferrin, fetuin, prostate-specific antigen, immunoglobulin G and ovalbumin) were measured in 50 mM sodium citrate (pH 5.0), glycoprotein substrate (1–3 μg) and enzyme preparation (0.4 μg) in a total volume of 10 μL. The enzyme reaction was performed at 45 °C for up to 20 h. The reaction was terminated by the addition of SDS–PAGE loading buffer and boiled for 3 min. The reaction mixtures were directly subjected to SDS–PAGE, and the gels were stained with Coomassie Brilliant Blue for analysis. N-Glycans released from glycoproteins were observed using LC ESI-MS as indicated above; basically, neutral N-glycans were detected as a singly charged ion, whereas acidic N-glycans were observed as multiple charged ions. For comparison of the hydrolysis activity toward N-glycans on glycoproteins among several ENGases, enzyme reactions were performed with glycoprotein substrate (1–3 μg) and enzyme preparation (0.2–0.5 μg) in a total volume of 10 μL at the optimal conditions for each enzyme for 18 h. For Endo-Tsp1006, Endo-Tsp1263 and Endo-1457, reactions were performed in 50 mM sodium citrate (pH 5.0) at 45 °C. For Endo-M, Endo-CC1, EndoH and EndoF1, reactions were performed in 50 mM sodium citrate (pH 6.0) at 37 °C. For EndoF2 and EndoF3, reactions were performed in 50 mM sodium citrate (pH 4.5) at 37 °C. PNGase F was used according to the manufacturer’s instructions. Acknowledgements The authors gratefully acknowledge the help and support of Dr. Takashi Kinoshita (Fushimi Pharmaceutical Co., Ltd., Kagawa, Japan) in constructing Endo-CC1 expression system. SGP was a kind gift from Fushimi Pharmaceutical Co., Ltd. Conflict of interest statement The authors declare that they have no conflicts of interest with the contents of this article. Abbreviations CAZy, carbohydrate-active enZymes; DMEM, Dulbecco’s modified Eagle’s medium; ENGase, endo-β-N-acetylglucosaminidase; GH, glycoside hydrolase; GST, glutathione S-transferase; LC ESI-MS, liquid chromatography electrospray ionization-mass spectrometry; PBS, phosphate buffered saline; PNGase F, peptide:N-glycosidase F; SDS–PAGE, sodium dodecyl sulfate–polyacrylamide gel electrophoresis; SGP, sialyl glycopeptide References Aoyama M , Hashii N, Tsukimura W, Osumi K, Harazono A, Tada M, Kiyoshi M, Matsuda A, Ishii-Watabe A. 2019 . Effects of terminal galactose residues in mannose α1-6 arm of fc-glycan on the effector functions of therapeutic monoclonal antibodies . mAbs. 11 : 826 – 836 . Google Scholar PubMed OpenURL Placeholder Text WorldCat Collin M , Olsén A. 2001 . EndoS, a novel secreted protein from Streptococcus pyogenes with endoglycosidase activity on human IgG . EMBO J. 20 : 3046 – 3055 . Google Scholar PubMed OpenURL Placeholder Text WorldCat Du YM , Xia T, Gu XQ, Wang T, Ma HY, Voglmeir J, Liu L. 2015 . Rapid sample preparation methodology for plant N-glycan analysis using acid-stable PNGase H+ . J. Agric Food Chem. 63 : 10550 – 10555 . Google Scholar PubMed OpenURL Placeholder Text WorldCat Eshima Y , Higuchi Y, Kinoshita T, Nakata S, Takegawa K. 2015 . Transglycosylation activity of glycosynthase mutants of endo-β-N-acetylglucosaminidase from Coprinopsis cinerea . PLoS One. 10 : e0132859 . Google Scholar PubMed OpenURL Placeholder Text WorldCat Fairbanks AJ . 2017 . The ENGases: Versatile biocatalysts for the production of homogeneous N-linked glycopeptides and glycoproteins . Chem Soc Rev. 46 : 5128 – 5146 . Google Scholar PubMed OpenURL Placeholder Text WorldCat Fan SQ , Huang W, Wang LX. 2012 . Remarkable transglycosylation activity of glycosynthase mutants of Endo-D, an endo-β-N-acetylglucosaminidase from Streptococcus pneumonia . J Biol Chem. 287 : 11272 – 11281 . Google Scholar PubMed OpenURL Placeholder Text WorldCat Fu D , Chen L, O'Neill RA. 1994 . A detailed structural characterization of ribonuclease B oligosaccharides by 1H NMR spectroscopy and mass spectrometry . Carbohydr Res. 261 : 173 – 186 . Google Scholar PubMed OpenURL Placeholder Text WorldCat Fujita K , Kobayashi K, Iwamatsu A, Takeuchi M, Kumagai H, Yamamoto K. 2004 . Molecular cloning of Mucor hiemalis endo-β-N-acetylglucosaminidase and some properties of the recombinant enzyme . Arch Biochem Biophys. 432 : 41 – 49 . Google Scholar PubMed OpenURL Placeholder Text WorldCat Giddens JP , Lomino JV, Amin MN, Wang LX. 2016 . Endo-F3 glycosynthase mutants enable chemoenzymatic synthesis of core-fucosylated triantennary complex type glycopeptides and glycoproteins . J Biol Chem. 291 : 9356 – 9370 . Google Scholar PubMed OpenURL Placeholder Text WorldCat Gloster TM , Vocadlo DJ. 2010 . Mechanism, structure, and inhibition of O-GlcNAc processing enzymes . Curr Signal Transduct Ther. 5 : 74 – 91 . Google Scholar PubMed OpenURL Placeholder Text WorldCat Green ED , Adelt G, Baenziger JU, Wilson S, Van Halbeek H. 1988 . The asparagine-linked oligosaccharides on bovine fetuin . J Biol Chem. 263 : 18253 – 18268 . Google Scholar PubMed OpenURL Placeholder Text WorldCat Harvey DJ , Wing DR, Küster B, Wilson IBH. 2000 . Composition of N-linked carbohydrates from ovalbumin and co-purified glycoproteins . J Am Soc Mass Spectrom. 11 : 564 – 571 . Google Scholar PubMed OpenURL Placeholder Text WorldCat Henrissat B . 1991 . A classification of glycosyl hydrolases based on amino acid sequence similarities . Biochem J. 280 : 309 – 316 . Google Scholar PubMed OpenURL Placeholder Text WorldCat Huang Y , Higuchi Y, Kinoshita T, Mitani A, Eshima Y, Takegawa K. 2018 . Characterization of novel endo-β-N-acetylglucosaminidases from Sphingobacterium species, Beauveria bassiana and Cordyceps militaris that specifically hydrolyze fucose-containing oligosaccharides and human IgG . Sci Rep. 8 : 246 . Google Scholar PubMed OpenURL Placeholder Text WorldCat Huang W , Li C, Li B, Umekawa M, Yamamoto K, Zhang X, Wang LX. 2009 . Glycosynthases enable a highly efficient chemoenzymatic synthesis of N-glycoproteins carrying intact natural N-glycans . J Am Chem Soc. 131 : 2214 – 2223 . Google Scholar PubMed OpenURL Placeholder Text WorldCat Ito K . 2014 . Enzyme having activity for releasing glycoprotein sugar chains and method for producing same, and method for releasing sugar chains using said enzyme . World patent number: WO/2014/080991 . Ito K , Okada Y, Ishida K, Minamiura N. 1993 . Human salivary endo-β-N-acetylglucosaminidase HS specific for complex type sugar chains of glycoproteins . J Biol Chem. 268 : 16074 – 16081 . Google Scholar PubMed OpenURL Placeholder Text WorldCat Katoh T , Katayama T, Tomabechi Y, Nishikawa Y, Kumada J, Matsuzaki Y, Yamamoto K. 2016 . Generation of mutant Mucor hiemalis endoglycosidase that acts on core-fucosylated N-glycans . J Biol Chem. 291 : 23305 – 23317 . Google Scholar PubMed OpenURL Placeholder Text WorldCat Kim KK , Kim MK, Lim JH, Park HY, Lee ST. 2005 . Transfer of Chryseobacterium meningosepticum and Chryseobacterium miricola to Elizabethkingia gen. nov. as Elizabethkingia meningoseptica comb. nov. and Elizabethkingia miricola comb. nov . Int J Syst Evol Microbiol. 55 : 1287 – 1293 . Google Scholar PubMed OpenURL Placeholder Text WorldCat Kumar S , Stecher G, Tamura K. 2016 . MEGA7: Molecular evolutionary genetics analysis version 7.0 for bigger datasets . Mol Biol Evol. 33 : 1870 – 1874 . Google Scholar PubMed OpenURL Placeholder Text WorldCat Kurogochi M , Amano J. 2014 . Relative quantification of glycopeptides based on stable isotope labeling using MALDI-TOF MS . Molecules. 19 : 9944 – 9961 . Google Scholar PubMed OpenURL Placeholder Text WorldCat Kurogochi M , Mori M, Osumi K, Tojino M, Sugawara S, Takashima S, Hirose Y, Tsukimura W, Mizuno M, Amano J, et al. 2015 . Glycoengineered monoclonal antibodies with homogenous glycan (M3, G0, G2, and A2) using a chemoenzymatic approach have different affinities for FcγRIIIa and variable antibody-dependent cellular cytotoxicity activities . PLoS One. 10 : e0132848 . Google Scholar PubMed OpenURL Placeholder Text WorldCat Lasica AM , Ksiazek M, Madej M, Potempa J. 2017 . The type IX secretion system (T9SS): Highlights and recent insights into its structure and function . Front Cell Infect Microbiol. 7 : 215 . Google Scholar PubMed OpenURL Placeholder Text WorldCat Lin YH , Franc V, Heck AJR. 2018 . Similar albeit not the same: In-depth analysis of proteoforms of human serum, bovine serum, and recombinant human fetuin . J Proteome Res. 17 : 2861 – 2869 . Google Scholar PubMed OpenURL Placeholder Text WorldCat Liu CP , Tsai TI, Cheng T, Shivatare VS, Wu CY, Wu CY, Wong CH. 2018 . Glycoengineering of antibody (Herceptin) through yeast expression and in vitro enzymatic glycosylation . Proc Natl Acad Sci U S A. 115 : 720 – 725 . Google Scholar PubMed OpenURL Placeholder Text WorldCat Lombard V , Golaconda Ramulu H, Drula E, Coutinho PM, Henrissat B. 2014 . The carbohydrate-active enzymes database (CAZy) in 2013 . Nucleic Acids Res. 42 : D490 – D495 . Google Scholar PubMed OpenURL Placeholder Text WorldCat Miyagawa K , Matsumoto M, Yabuno S, Kawakami N, Hamaguchi T, Iizuka M, Minamiura N, Ito K. 2007 . Transglycosylation of asparagine-linked complex-type oligosaccharides from glycoproteins by endo-β-N-acetylglucosamindase HS . J Appl Glycosci. 54 : 139 – 146 . Google Scholar OpenURL Placeholder Text WorldCat Muramatsu H , Tachikui H, Ushida H, Song X, Qiu Y, Yamamoto S, Muramatsu T. 2001 . Molecular cloning and expression of endo-β-N-acetylglucosaminidase D, which acts on the core structure of complex type asparagine-linked oligosaccharides. 2001 . J Biochem. 129 : 923 – 928 . Google Scholar PubMed OpenURL Placeholder Text WorldCat Nagae M , Morita-Matsumoto K, Arai S, Wada I, Matsumoto Y, Saito K, Hashimoto Y, Yamaguchi Y. 2014 . Structural change of N-glycan exposes hydrophobic surface of human transferrin . Glycobiology. 24 : 693 – 702 . Google Scholar PubMed OpenURL Placeholder Text WorldCat Reddy A , Grimwood BG, Plummer TH, Tarentino AL. 1998 . High-level expression of the endo-β-N-acetylglucosamindiase F2 gene in E. coli: One step purification to homogeneity . Glycobiology. 8 : 633 – 636 . Google Scholar PubMed OpenURL Placeholder Text WorldCat Sanchez-De Melo I , Grassi P, Ochoa F, Bolivar J, García -Cózar M, Durán-Ruiz MC. 2015 . N-glycosylation profile analysis of Trastuzumab biosimilar candidates by normal phase liquid chromatography and MALDI-TOF MS approaches . J Proteomics. 127 : 225 – 233 . Google Scholar PubMed OpenURL Placeholder Text WorldCat Tabarés G , Radcliff CM, Barrabés S, Ramírez M, Aleixandre RN, Hoesel W, Dwek RA, Rudd PM, Peracaula R, de Llorens R. 2006 . Different glycan structures in prostate-specific antigen from prostate cancer sera in relation to seminal plasma PSA . Glycobiology. 16 : 132 – 145 . Google Scholar PubMed OpenURL Placeholder Text WorldCat Takegawa K , Yamabe K, Fujita K, Tabuchi M, Muta M, Izu H, Watanabe A, Asada Y, Sano M, Kondo A, et al. 1997 . Cloning, sequencing, and expression of Arthrobacter protophormiae endo-β-N-acetylfglucosaminidase in Escherichia coli . Arch Biochem Biophys. 338 : 22 – 28 . Google Scholar PubMed OpenURL Placeholder Text WorldCat Tarentino AL , Quinones G, Changchien LM, Plummer TH Jr. 1993 . Multiple endoglycosidase F activities expressed by Flavobacterium meningosepticum endoglycosidases F2 and F3. Molecular cloning, primary sequence, and enzyme expression . J Biol Chem. 268 : 9702 – 9708 . Google Scholar PubMed OpenURL Placeholder Text WorldCat Treuheit MJ , Costello CE, Halsall HB. 1992 . Analysis of the five glycosylation sites of human α1-acid glycoprotein . Biochem J. 283 : 105 – 112 . Google Scholar PubMed OpenURL Placeholder Text WorldCat Trimble RB , Tarentino AL. 1991 . Identification of distinct endoglycosidase (endo) activities in Flavobacterium meningosepticum: Endo F1, endo F2, and endo F3. Endo F1 and endo H hydrolyze only high mannose and hybrid glycans . J Biol Chem. 266 : 1646 – 1651 . Google Scholar PubMed OpenURL Placeholder Text WorldCat Trumbly RJ , Robbins PW, Belfort M, Ziegler FD, Maley F, Trimble RB. 1985 . Amplified expression of Streptomyces endo-β-N-acetylglucosaminidase H in Escherichia coli and characterization of the enzyme product . J Biol Chem. 260 : 5683 – 5690 . Google Scholar PubMed OpenURL Placeholder Text WorldCat Tsukimura W , Kurogochi M, Mori M, Osumi K, Matsuda A, Takegawa K, Furukawa K, Shirai T. 2017 . Preparation and biological activities of anti-HER2 monoclonal antibodies with fully core-fucosylated homogenous bi-antennary complex-type glycans . Biosci Biotech Biochem. 81 : 2353 – 2359 . Google Scholar OpenURL Placeholder Text WorldCat Tzelepis G , Dubey M, Jensen FD, Karlsson M. 2015 . Identifying glycoside hydrolase family 18 genes in the mycoparasitic fungal species Clonostachys rosea . Microbiology . 161 : 1407 – 1419 . Google Scholar PubMed OpenURL Placeholder Text WorldCat Umekawa M , Li C, Higashiyama T, Huang W, Ashida H, Yamamoto K, Wang LX. 2010 . Efficient glycosynthase mutant derived from Mucor hiemalis endo-β-N-acetylglucosaminidase capable of transferring oligosaccharide from both sugar oxazoline and natural N-glycan . J Biol Chem. 285 : 511 – 521 . Google Scholar PubMed OpenURL Placeholder Text WorldCat White KY , Rodemich L, Nyalwidhe JO, Comunale MA, Clements MA, Lance RS, Schellhammer PF, Mehta A, Semmes OJ, Drake RR. 2009 . Glycomic characterization of prostate specific antigen and prostate acid phosphatase in prostate cancer and benign disease seminal plasma fluids . J Proteome Res. 8 : 620 – 630 . Google Scholar PubMed OpenURL Placeholder Text WorldCat Yamashita K , Kamerling JP, Kobata A. 1982 . Structural study of the carbohydrate moiety of hen ovomucoid . J Biol Chem. 257 : 12808 – 12814 . Google Scholar OpenURL Placeholder Text WorldCat Yamashita K , Kamerling JP, Kobata A. 1983 . Structural studies of the sugar chains of hen ovomucoid . J Biol Chem. 258 : 3099 – 3106 . Google Scholar PubMed OpenURL Placeholder Text WorldCat © The Author(s) 2020. Published by Oxford University Press. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/open_access/funder_policies/chorus/standard_publication_model)
TI - Novel endo-β-N-acetylglucosaminidases from Tannerella species hydrolyze multibranched complex-type N-glycans with different specificities
JF - Glycobiology
DO - 10.1093/glycob/cwaa037
DA - 2020-10-21
UR - https://www.deepdyve.com/lp/oxford-university-press/novel-endo-n-acetylglucosaminidases-from-tannerella-species-hydrolyze-d13LyREGLE
SP - 923
EP - 934
VL - 30
IS - 11
DP - DeepDyve
ER -