Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2014 Mar 25.
Published in final edited form as: Methods Enzymol. 2006;415:31–46. doi: 10.1016/S0076-6879(06)15003-X

Family 47 α-Mannosidases in N-Glycan Processing

Steven W Mast, Kelley W Moremen
PMCID: PMC3964790  NIHMSID: NIHMS559869  PMID: 17116466

Abstract

α-Mannosidases in eukaryotic cells are involved in both glycan biosynthetic reactions and glycan catabolism. Two broad families of enzymes have been identified that cleave terminal mannose linkages from Asn-linked oligosaccharides (Moremen, 2000), including the Class 1 mannosidases (CAZy GH family 47 (Henrissat and Bairoch, 1996)) of the early secretory pathway involved in the processing of N-glycans and quality control and the Class 2 mannosidases (CAZy family GH38 [Henrissat and Bairoch, 1996]) involved in glycoprotein biosynthesis or catabolism. Within the Class 1 family of α-mannosidases, three subfamilies of enzymes have been identified (Moremen, 2000). The endoplasmic reticulum (ER) α1,2-mannosidase I (ERManI) subfamily acts to cleave a single residue from Asn-linked glycans in the ER. The Golgi α-mannosidase I (GolgiManI) subfamily has at least three members in mammalian systems (Herscovics et al., 1994; Lal et al., 1994; Tremblay and Herscovics, 2000) involved in glycan maturation in the Golgi complex to form the Man5GlcNAc2 processing intermediate. The third subfamily ofGH47 proteins comprises the ER degradation, enhancing α-mannosidase-like proteins (EDEM proteins) (Helenius and Aebi, 2004; Hirao et al., 2006; Mast et al., 2005). These proteins have been proposed to accelerate the degradation of misfolded proteins in the lumen of the ER by a lectin function that leads to retrotranslocation to the cytosol and proteasomal degradation. Recent studies have also indicated that ERManI acts as a timer for initiation of glycoprotein degradation via the ubiquitin-proteasome pathway (Hosokawa et al., 2003; Wu et al., 2003). This article discusses methods for analysis of the GH47 α-mannosidases, including expression, purification, activity assays, generation of point mutants, and binding studies by surface plasmon resonance.

Overview

The biosynthesis of glycoproteins containing Asn-linked oligosaccharides is initiated by the transfer of a Glc3Man9GlcNAc2 oligosaccharide from a lipid-linked precursor to nascent polypeptides in the lumen of the endoplasmic reticulum (ER) (Kornfeld and Kornfeld, 1985). After the glycan transfer, a series of processing steps occurs, wherein hybrid or complex oligosaccharides can be formed from the initial high-mannose precursor. The initial steps of this process involve the trimming of glucose residues by glucosidase I and II, enzymes that are localized in the ER. These enzymes are responsible for the excision of two of the three glucose residues to form the GlcMan9GlcNAc2 structure necessary for recognition by the lectin chaperones, calnexin and calreticulin. Further trimming by glucosidase II removes the final Glc residue to result in the release from calnexin/ calreticulin interactions. Subsequent trimming of mannose residues from the precursor oligosaccharide is a prerequisite for the maturation of N-glycans to hybrid and complex-type structures, and in mammalian cells these cleavage reactions are accomplished by the action of ERManI, members of the GolgiManI subfamily of enzymes, and an α1,3/α1,6mannosidase in the Golgi complex, termed Golgi α-mannosidase II (Moremen, 2000). ERManI and the GolgiManI isozymes are Class 1 (GH47) α-mannosidases, whereas Golgi α-mannosidase II is a Class 2 (GH38) mannosidase and has been reviewed elsewhere (Moremen, 2000, 2002). Alternative path-ways for high-mannose glycan maturation have also been proposed involving enzymes termed ER α-mannosidase II and endo α-mannosidase, which can trim Man9GlcNAc2 to Man8GlcNAc2 and Glc3–1Man9GlcNAc2 to Man8- GlcNAc2, respectively. These latter enzymes have also been previously reviewed (Spiro and Spiro, 2000; Weng and Spiro, 1996).

Class 1 α-mannosidases recognize and cleave only α-1,2–linked mannose residues from the Man9GlcNAc2 high-mannose precursor, but each of the subfamilies has unique specificities for recognition of terminal α-1,2-mannose residues in the high-mannose structure. The ERManI subfamily contains a single member in all eukaryotes that have been examined, and the enzyme primarily cleaves a single mannose residue from Man9GlcNAc2 to produce a unique Man8GlcNAc2 isomer (Man8 isomer B; Fig. 1). ERManI is the first mannosidase to act in this pathway, and it has been suggested that it mediates quality control in the ER by creating a specific glycoform that is recognized by members of the putative degradation family of lectins, termed the EDEM proteins. The GolgiManI subfamily of enzymes comprises three isoforms in mammalian organisms: Golgi- ManIA (Lal et al., 1994), GolgiManIB (Herscovics et al., 1994), Golgi-ManIC (Tremblay and Herscovics, 2000). The substrate specificities of each of the isoforms are similar, and they are all complementary to the action of ERManI (Gonzalez et al., 1999). Thus, the GolgiManI isoforms can efficiently trim the Man8GlcNAc2 isomer B down to Man5GlcNAc2, but they poorly cleave the central branch mannose residue that is the target of ERManI action (see Fig. 1) (Lal et al., 1998). Subsequent transfer of a terminal GlcNAc residue by GlcNAc transferase I and mannose trimming by Golgi α-mannosidase II result in the formation of the GlcNAcMan3Glc-NAc2 intermediate that acts as the committed step in the synthesis of complex-type N-glycans.

Fig. 1.

Fig. 1

Open in a new tab

Diagram of the early in vivo N-glycan processing steps of Glc3Man9GlcNAc2-Asn to Man5GlcNAc2-Asn. The processing steps catalyzed by glucosidases I and II remove the three glucose residues. Trimming by ERManI generates the Man8 isomer B, and trimming by the GolgiManI subfamily of enzymes generates Man5GlcNAc2. The GlcNAc residues are indicated by black shaded hexagons. Mannose residues are gray hexagons. Glc residues are white hexagons.

The EDEM subfamily of proteins is composed of three isoforms, and two members (EDEM1 and EDEM2) do not appear to have catalytic activity. Recent data suggest that the third isoform (EDEM3) may have glycosidase activity toward high-mannose oligosaccharides (Hirao et al., 2006). All three of these mannosidase homologs have been shown to participate in the targeting of misfolded glycoproteins from the ER to the cytosol for degradation by the ubiquitin-proteasomal system. The mechanism for this protein disposal function has not been determined, but it is thought that EDEM proteins act as lectins, using the same oligosaccharide binding site employed by the catalytic Class I α-mannosidases (Mast et al., 2005).

ERManI was first isolated, characterized, and cloned fromSaccharomyces cerevisiae (Herscovics, 1999), but subsequent cDNA cloning, expression, and characterization of the human ERManI homolog (Gonzalez et al., 1999) have revealed similarities in structure and specificity between the fungal and human enzymes. Recombinant forms of human (Karaveg et al., 2005; Vallee et al., 2000a) and yeast ERManI (Vallee et al., 2000b) have been expressed in large quantities in Pichia pastoris, and the structures of the catalytic domains of each enzyme have been determined by X-ray diffraction in the presence and absence of bound enzymatic products, small molecule inhibitors, and uncleavable pseudosubstrates.

Members of the GolgiManI subfamily of enzymes have been isolated from a variety of species sources (e.g., rat, rabbit, pig, and calf, reviewed by Moremen et al. [1994]), and recombinant forms of the human (Bause et al., 1993) and mouse enzymes (Herscovics et al., 1994; Lal et al., 1994) have been expressed and characterized. Each of the members of the GolgiManI subfamily has been expressed in P. pastoris, and the structure of mouse GolgiManIA has been solved as a co-complex with a bound glycan enzymatic product (Tempel et al., 2004). In addition, structures have been determined for the secreted GH47 α-mannosidases from Penicillium citrinum (Lobsanov et al., 2002) and Trichoderma reesei (Van Petegem et al., 2001), fungal enzymes that have substrate specificities similar to the GolgiManI enzymes.

The membrane topologies of the ERManI and GolgiManI subfamily members are similar, each containing a short cytoplasmic tail, transmembrane domain, linker region, and luminal catalytic domain (Moremen, 2000). The catalytic domains of both ERManI and GolgiManIA are both barrel-shaped (αα)7 structures composed of two concentric layers of α helices containing a broad open cleft at one end of the barrel structure and a β-hairpin plugging the opposite end of the barrel. The oligosaccharide substrates bind to the open cleft, and extended interactions across the cleft provide specificity for insertion of glycan branches into the barrel core containing the enzyme active site. Structural data indicate that the cores of the catalytic pockets (+1 and −1 substrate-binding subsites) are conserved between the ERManI and GolgiManI subfamily members; however, outside of this area (>+1 residues), including the broad open clefts adjacent to the catalytic sites, the enzymes have distinct oligosaccharide binding sites that provide differences in branch specificities for these enzyme subfamilies (Tempel et al., 2004). The energetic contributions of oligosaccharide substrate binding to the catalytic pocket and the broad open cleft were examined in a series of surface plasmon resonance binding studies between a catalytic mutant of human ERManI (E330Q) and a series of high-mannose N-glycan structures (Karaveg et al., 2005). These studies revealed that the majority of the binding energy was contributed through interactions with the substrate residue in the +1 subsite (~52% of the substrate binding energy). Considerably less energetic contribution was provided by interactions with the −1 subsite residue (~21% of the binding energy). The remainder of the binding energy (~27% of the binding energy) was provided by substrate residues in the >+1 substrate binding subsites. Despite the relatively minor contributions of the >+1 subsite residues to the binding energy, the stereochemistry of the interactions in the broad substrate-binding cleft clearly provides the branch specificity for the selective cleavage of terminal α1,2-mannose residues. Swapping of a key substrate-binding residue from GolgiManIA to ERManI led to broader specificity for glycosidase cleavage but poor catalytic efficiency and a significantly reduced binding affinity. Thus, the catalytic pockets and adjacent substrate-binding clefts for the Class 1 (GH47) α-mannosidases provide for the high-affinity and high-specificity substrate interactions that lead to the unique substrate recognition characteristics for the individual subfamilies. This chapter describes the methods for enzyme expression and characterization for wild-type and mutant forms of the Class 1 mannosidases that can be used as paradigms for similar studies in other enzyme systems.

Overview of Class 1 (GH47) Expression and Purification

Recombinant expression of wild-type and mutant Class 1mannosidases in P. pastoris has proven to be an effective strategy for large-scale production of enzymes for biochemical and structural studies (Karaveg and Moremen, 2005; Karaveg et al., 2005; Tempel et al., 2004). The advantages of this expression system include the ease in generating recombinant constructs and transformants, the short-time course of induction, the high levels of protein expression, the ability to generate recombinant proteins that are secreted into the culture media for easy purification, and the ability to generate recombinant products that contain eukaryotic post-translational modifications (i.e., disulfide bond formation, N-glycosylation). We have generally used commercial expression vectors (Invitrogen pPICZα series or pHIL-S1) that encode an NH2-terminal S. cerevisiae α-factor signal sequence or PHO1 secretion signal followed by a series of restriction sites. cDNA sequences encoding the COOH-terminal catalytic domains of the Class 1 mannosidases (with deletions of the NH2-terminal cytoplasmic tails, transmembrane domains, and stem regions) have been inserted in-frame into the vectors, and transformants were generated in either Mut+ (X-33, GS115) or MutS (KM71H) strains. Following growth of cell mass on glycerol as a carbon source, enzyme expression was initiated by shifting to methanol as a carbon source and inducing agent for the AOX1 promoter. These transformants generate translation products that are targeted to the yeast secretory pathway and released into the culture media as potentially glycosylated products. Both S. cerevisiae ERManI (Vallee et al., 2000b) and mouse GolgiManIA (Tempel et al., 2004) coding regions contain acceptor sequences for N-glycosylation (Table I) while human ERManI (Vallee et al., 2000a) contains no N-glycosylation sites and is secreted as a non-glycosylated protein. In general, affinity tags have not been used for the generation of α-mannosidase recombinant constructs, since purification of the enzyme from the culture media is readily accomplished with two or three chromatography steps (see later). In recent studies, the mannosidase homology domain of human EDEM2 has also been expressed in P. pastoris using pPICZα constructs. These latter constructs have been generated as epitope and/or His-tagged fusions at the NH2-terminus of the α-mannosidase–homology domain, because previous data have indicated that COOH-terminal fusions yield in inactive ERManI constructs. A summary of α-mannosidase constructs generated for expression in P. pastoris is shown in Table I.

TABLE I.

Constructs for Expression of Class 1 α-Mannosidases in Pichia Pastoris

Class 1
α-mannosidasea 		Pichia
host strain		 Expression
vector		 Residues of
coding region
expressed		 Amino acids
in full-length
coding region		 Predicted
N-glycan
acceptor sites
Yeast ERManI GS115, KM71 pHILS-1 34–549 549 3
Human ERManI X33 pPICZ-α 226–699 699 0
Mouse GolgiManIA GS115 pHILS-1 66–655 655 1
Human EDEM2 KM71H pPICZ-α 18–492 578 4
Open in a new tab
a

Information on the constructs and strains for the expression data of yeast ERManI (Lipari and Herscovics, 1996), human ERManI (Karaveg et al., 2005), mouse GolgiManIA (Lal et al., 1998), and human EDEM2 (unpublished data) were obtained from the respective sources.

Generation of Class 1 α-Mannosidase Site-Directed Mutants

A number of site-directed mutants of ERManI have been generated to analyze the effects of altering selected residues on glycan binding and catalytic efficiency (Karaveg and Moremen, 2005; Karaveg et al., 2005). Residues were selected based on their positions in the α-mannosidase structure, and mutations were generally either conservative substitutions or replacements with Ala residues. Mutagenesis was accomplished in the pPICZα vector using the QuikChange mutagenesis kit (Stratagene) using protocols provided by the manufacturer. After the site-directed mutagenesis, the coding regions were fully sequenced to confirm that only the desired mutation was introduced into the recombinant protein product. All mutant proteins were expressed and purified under identical conditions as the wild-type enzyme, and in most instances the protein yields were similar to the wild-type enzyme. For catalytically inactive mutant enzymes, protein purification profiles were followed by immunoblots using antibodies raised to the recombinant wild-type proteins.

Enzyme Expression and Purification

The Class 1 α-mannosidases can be readily expressed in P. pastoris, either in shake flasks (Karaveg et al., 2005) or small fermentor cultures (Tempel et al., 2004). For shake flask cultures, yeast transformants were grown for 2 days in 1L BMGY (400 mg of biotin, 0.5% methanol, 1% glycerol, 1%yeast extract, 2% peptone, 0.1 M potassium phosphate [pH 6], and 1.34% yeast nitrogen base), at 30°, with agitation at 300 rpm to generate cell mass; recombinant protein production was initiated by the addition of 0.5% methanol to the culture. Additional methanol (0.5% final concentration) was then added at 12-h intervals for 3 to 5 days. Alternatively, protein production can be generated in a New Brunswick BioFlow 3000 fermentor (1- or 2.5-l vessels) in BMGY medium supplemented by a limiting feeding strategy using glycerol as a carbon source until a cell density of approximately 200 g/l is attained. The glycerol feed was then stopped, and the culture was maintained at 30° with agitation until oxygen consumption fell, indicating that the carbon source was consumed. Methanol was then pumped into the fermentor vessel and maintained at 0.5% by the use of a methanol probe (Raven Biotech, Vancouver, BC) controlling the methanol feed pump. The culture was then maintained at this methanol concentration for an additional 3 days of culture before harvesting.

Initial purification of the secreted α-mannosidase catalytic domains followed a clarification of the culture supernatant by centrifugation (6000 × g, 20 min) and filtration through a 0.4-µm filter. For GolgiManIA (Tempel et al., 2004), the resulting supernatant was dialyzed against four volumes of 10 mM sodium succinate (pH, 6.5) with two changes of buffer, followed by adjustment of the pH to 5.5 with HCl immediately before chromatography. The human ERManI (Karaveg et al., 2005) culture supernatant was adjusted to pH 6.0 and applied to the cation exchange column after 1:2 dilution with H2O. Both enzyme preparations were loaded on SP Sepharose columns (Amersham-Pharmacia Biotech, 5 × 13 cm column), washed with column buffer, and elution was performed with a linear NaCl gradient (0 to 0.5 M) in the same buffer. For GolgiManIA purification, the column buffer was 10 mM sodium succinate (pH, 5.5); and for the ERManI purification, the column buffer was 10 mM sodium succinate (pH, 6.0), 1 mM CaCl2. ERManI preparations were sufficiently enriched after the SP Sepharose step to proceed directly to gel filtration. GolgiManIA preparations were further purified by adjustment of the solution to 1.0 M (NH4)2SO4, 0.1 M Na-HEPES (pH, 6.5), followed by loading on a Phenyl Sepharose column (Amersham-Pharmacia Biotech; 1.5 × 20 cm). The column was washed with 1.0 M (NH4)2SO4and 10 mM Na-HEPES (pH, 6.5) and eluted using a linear gradient of 1.0 to 0.5 M (NH4)2SO4 in 10 mM Na-HEPES (pH, 6.5). For both enzyme preparations, fractions containing enzyme activity were pooled and concentrated using an Amicon YM10 membrane after the addition of NDSB-201 (Calbiochem) to 0.75 M. Aliquots of 5 ml of concentrated enzyme sample were applied to a Superdex 75 column (Amersham-Pharmacia Biotech, 1 × 100 cm) pre-equilibrated with 20 mM Na-MES (pH, 6.5 for mouse GolgiManIA; pH, 7.0 for human ERManI), 150 mM NaCl, 5 mM CaCl2, and 0.25 M NDSB-201. The final purified enzyme preparations were concentrated to 30 mg/ml in the same buffer and stored at 4°. The inclusion of NDSB-201 in all of the buffers after the Phenyl Sepharose step was an important element in the purification strategies. Significant losses in the recovery of the recombinant enzymes by aggregation and precipitation were avoided by inclusion of NDSB-201 in all of the buffers.

Oligosaccharide Isolation for Enzyme Assays and Binding Studies

Man9GlcNAc2 glycan structures were preparatively isolated from crude soybean agglutinin (SBA) generated from soybean meal by acid extraction and (NH4)2SO4 precipitation (Karaveg and Moremen, 2005; Karaveg et al., 2005). Crude SBA was subsequently denatured, reduced, and carboxyami-domethylated in 8 M guanidine HCl containing 50 mM dithiothreitol and 100 mM iodoacetamide before proceeding to trypsin/elastase digestion. Glycopeptides were recovered by affinity chromatography using concanavalin A-Sepharose (ConA-Sepharose, Amersham-Pharmacia Biotech) and further purified by high performance liquid chromatography (HPLC) on a Cosmosil C18 column for surface plasmon resonance (SPR) studies (Karaveg et al., 2005). Free oligosaccharides were liberated from the peptide by peptide:N-glycosidase F digestion and derivatized with pyridylamine. Man8GlcNAc2-PA isomers, Man6GlcNAc2-PA, and Man5GlcNAc2-PA were generated by digestion with either ERManI or GolgiManIA and isolation by reverse phase HPLC, as described (Karaveg and Moremen, 2005; Karaveg et al., 2005).

α-Mannosidase Assays and Kinetic Analyses

Class I mannosidase enzyme activity can be measured by monitoring the cleavage of Man-α1,2-Man disaccharides or pyridylamine-tagged high-mannose oligosaccharides (Man9GlcNAc2-PA). Assays using tagged high-mannose oligosaccharides as substrates are more sensitive than the disaccharide assays, but they are more labor intensive and not as readily adapted to a high-throughput assay format. An alternative assay using a [3H]mannose-labeled Man9GlcNAc2 substrate followed by detection of cleavage by substrate binding to ConA has also been described (Herscovics and Jelinek-Kelly, 1987), but this requires the generation and handling of hazardous radiolabeled substrates.

Disaccharide assays for α-mannosidase activity make use of a modified glucose oxidase/horseradish peroxidase assay (Lal et al., 1998). Assays were assembled in flat-bottomed 96-well microliter plates containing 10 mM potassium phosphate (pH, 6.0), 2 to 10 mM Manα1,2Manα-O-CH3 (Sigma), and 5 to 15 µl of enzyme solution (25 µl total volume) followed by incubation at 37° for 30 min. Reactions were terminated by adding 25 µl of 1.25 M Tris-Cl (pH, 7.6), and the amount of mannose released was detected by incubation with 250 µl of developing solution containing glucose oxidase (55 U/ml), horseradish peroxidase (1 purpurogallin unit/ml), and o-dianisidine dihydrochloride (70 µg/ml) for 3 h at 37°. The final color intensity was determined by measuring the absorbance at 450 nm. Free mannose was used as a standard. One unit of mannosidase activity is defined as the amount of enzyme that releases 1 nmole of mannose in 1 minute at 37°.

Assays using oligosaccharide substrates (Lal et al., 1998) contained 100 pmol Man9GlcNAc2-PA, 50 mM MES (pH, 7), 150 mM NaCl, and 5 mM CaCl2 (20 µl total volume), and reactions were performed at 37° for varied periods of time, depending on the concentration of the enzyme. Reactions were stopped by the addition of 20 µl 1.25 M Tris-HCl (pH, 7.6), followed by separation of the digested products on a Hypersil APS- 2 NH2 column (4.6 × 250 mm, Alltech), using a gradient from 90% buffer A (80% acetonitrile : 20% 50 mM Na phosphate; pH, 4) to 35% Buffer B (100 mM Na phosphate; pH, 4) run over 50 min at flow rate of 1 ml/min. Elution of the PA-tagged oligosaccharide was monitored by an in-line fluorescence detector (excitation wavelength, 320 nm; emission wavelength, 400 nm), with identification of the cleavage products based on a comparison to the elution positions of Man9–5GlcNAc2-PA standards. One unit of enzyme activity generated 1µmol/min of Man8GlcNAc2 from Man9- GlcNAc2 at 37°. An alternative solvent protocol provided a shorter HPLC separation time. Using 40% acetonitrile in 60 mM Na phosphate buffer (pH, 4) with an isocratic flow of 1.0 ml/min, Man9–5GlcNAc2-PA could be resolved in less than 15 min. Variations in the flow rate (0.7 to 1.4 ml/min) or an increase in the percentage of acetonitrile by 5% to 15% can be used to shift the elution positions of the high-mannose glycans for optimal resolution.

For kinetic analysis (Karaveg and Moremen, 2005; Karaveg et al., 2005), initial rates (v) for the enzyme reactions were determined at various substrate concentrations ranging from 10 to 300 µM. Catalytic coefficient (kcat), and Michaelis constant (Km) values were determined by fitting initial rates to a Michaelis-Menten function by nonlinear regression analysis using SigmaPlot (Jandel Scientific, San Rafael, CA). kcat/Km values could also be derived from reciprocal plots of v and [S]. For kinetic pH rate analysis (Karaveg et al., 2005), values of kcat and Km were determined from initial rates of enzyme reactions in the pH range from 4 to 10 using a 4× universal buffer (80 mMsuccinic acid, 80 mMMES, 80 mMHEPBS, 80 mMHEPES, and 80 mM CHES adjusted to pH with 5 M NaOH) in place of the MES assay buffer described previously. Plots of log(kcat/Km) versus pH could then be used to estimate the pKa values for the acidic and basic limbs of the pH curves for the wild-type and mutant enzymes. In studies on the temperature dependence of catalysis (Karaveg and Moremen, 2005), values for kcat were obtained between 5°and 40° at 5° intervals and were used to calculate activation energies (Ea) from the slopes (−Ea/R) of Arrhenius plots (ln(kcat) as a function of 1/T). The thermodynamic activation parameters can then be calculated from the plots as previously described (Karaveg and Moremen, 2005).

Glycan Binding Studies by SPR

The combination of mutagenesis and substrate binding studies by SPR provided a useful set of tools for investigating the determinants of binding kinetics and glycan affinity for the recombinant α-mannosidases (Karaveg and Moremen, 2005; Karaveg et al., 2005). Among the various strategies for binding studies by SPR, we have chosen to immobilize the enzymes to the sensor chip surface and screen for glycan binding as the mobile phase analyte (Fig. 2).

Fig. 2.

Fig. 2

Open in a new tab

Examples of SPR binding of Man9GlcNAc2, dMNJ, or Kif ligands to wild-type and mutant ERManI. Wild-type (A, B, and F) or mutant (C, D, and E) ERManI forms were immobilized on the SPR chip surface and various concentrations of Man9GlcNAc2 (left panels), dMNJ (B and D), or Kif (F) ligands were tested for binding. Representative SPR sensorgrams in the ligand concentration series are shown. If the on- and off-rates (ka and kd, respectively) were sufficiently slow (B and E), curve fitting of the sensorgrams was performed to determine values for the equilibrium dissociation constants (KD = kd/ka). Where the kinetics for ligand binding were too rapid, equilibrium sensorgram values were used to plot a saturation curve (insets in A and C) and calculate values for KD. In the F659A mutant, binding of Man9GlcNAc2 was minimally affected (C), whereas dMNJ binding was completely abolished, as indicated by the absence of a deflection in the SPR sensorgram trace (D). For panel F, the tight-binding kinetics of Kif resulted in no appreciable dissociation from the chip surface. Reproduced in part from Karaveg and Moremen (2005), with permission of American Society for Biochemistry and Molecular Biology via Copyright Clearance Center.

SPR analyses were conducted using a Biacore 3000 apparatus (Biacore AB), with recombinant enzymes immobilized on the chip surfaces at 25° by amine coupling. The flow cells were activated by injecting a mixture of 50 mM N-hydroxysuccinimide and 200 mM 1-ethyl-3-(dimethylaminopropyl) carbodiimide over the CM5 sensor chip surface for 7 min at 5 µl/min. The recombinant proteins prepared in 10 mM sodium succinic acid (pH, 6.0) were passed through a 0.2-µm filter and diluted in the same buffer to obtain a concentration of 5 µg/ml before injection onto the activated surface. The desired immobilization level can be achieved by specific contact time with the chip surface. The immobilization efficiency for ERManI was about 2500 RU/min at a flow rate of 5 µl/min in HPB-EP buffer (10 mM HEPES; pH, 7.4; 150 mMNaCl; 3.4 mMEDTA; 0.01% polysorbate P20) at 25°. The remaining reactive groups were blocked by injection of 1 M ethanolamine-HCl at pH 8.5 for 7 min at 5 µl/min. Mock derivatized flow cells served as reference surfaces. Binding analyses were generally performed at 10° with a continuous flow (30 µl/min) of running buffer (10 mMMES; pH, 7.0, 300 mM NaCl; and 5 mM CaCl2) except in temperature-dependent interaction studies, which have been performed between 5° and 35° at 5° intervals, consecutively, in an automated method. Analytes were prepared in running buffer by twofold serial dilution to obtain appropriate concentration ranges. A concentration series of Man9GlcNAc2-glycopeptide (0.4 to 400 µM) was analyzed over a low-density immobilization surface of recombinant protein (3000 RU), whereas disaccharides (16 to 1000 µM) and small molecule inhibitors, such as 1-deoxymannojirimycin (dMNJ) (2 to 1000 µM) and kifunensine (Kif), were analyzed over a high-density immobilization surface of recombinant protein (10,000 RU). Samples were injected for 1 min, followed by at least 3 min of dissociation phase. In most cases, the baseline returned to the original response level in 5 min without a further regeneration of the chip surface, except when Kif was used as the analyte (see Fig. 2). This latter inhibitor exhibits tight-binding kinetics and does not dissociate from the chip surface even with extensive washing. SPR data for each concentration of analyte were collected in duplicate and globally fit to a 1:1 Langmuir binding algorithm model to calculate the on-rate (ka), the off-rate (kd), and the equilibrium dissociation constant (kd/ka = KD) using BiaEvaluation version 3.1 software supplied with the instrument. Alternatively, the maximal equilibrium sensorgram values were used to plot a saturation binding curve and calculate values for the equilibrium dissociation constant (KD) directly. KD values measured at different temperatures can also be used to calculate thermodynamic parameters (ΔG, ΔH, TΔS) of binding (Karaveg and Moremen, 2005) using the van’t Hoff equation (Day et al., 2002) to plot the linear relationship of y = ln(KD) versus x = 1/T, which gives a slope of ΔH/R and an intercept of −ΔS/R. The effects of temperature on the individual association rates (ka) and dissociation rates (kd) can also be independently determined using the Eyring equation (Day et al., 2002), which allows thermodynamic parameters to be determined from measured ka and kd values at different temperatures by a linear relationship of y = Rln(hka/kBT) or y = Rln (hkd/kBT) versus x = 1/T, where the slope and the intercept of the Eyring plots are ΔH† and ΔS†, respectively (Karaveg and Moremen, 2005).

Summary

The biochemical characterization of Class 1 α-mannosidases has made use of a combination of approaches including recombinant expression of wild-type and mutant enzymes, purification by classical chromatography approaches, analyses of enzyme kinetics, and binding affinity and kinetic measurements using SPR (Karaveg and Moremen, 2005; Karaveg et al., 2005). In the case of human ERManI, these approaches have been complemented by structural determination of co-complexes between the enzyme and small molecule glycone mimics, uncleavable disaccharide pseudo-substrates, and large glycan enzymatic products (Karaveg et al., 2005; Tempel et al., 2004; Vallee et al., 2000a,b). The result of these studies was a detailed map of the energetic contributions to substrate recognition, binding, and catalysis (Fig. 3) (Karaveg and Moremen, 2005) that will establish a framework for studying the differences in recognition between the different Class 1 α-mannosidase subfamilies. Since these proteins play distinct roles in glycoprotein maturation and ER quality control, further study should lead to a greater understanding of how these common sequences and structural folds can accomplish their unique biological functions. The strategies used for the biochemical characterization of the α-mannosidases can also be applied to the study of many other families of enzymes that play critical cellular functions.

Fig. 3.

Fig. 3

Open in a new tab

Summary of the enzyme kinetics and binding studies for key protein and glycan residues in the ERManI active site. The catalytic and binding components in the ERManI active site are shown, highlighting the protein residues interacting with the −1, +1, and +2 subsite glycan residues. Fold differences for kinetic parameters (values for kcat and Km) and SPR binding parameters (KD values for binding to Man9GlcNAc2 and dMNJ) for the mutant enzymes relative to values for wild-type ERManI are indicated adjacent to each respective amino acid. In addition, the respective proposed functions for each residue based on the binding and kinetic analyses are indicated adjacent to each amino acid. Energetic contributions to binding within the −1 and +1 subsites are also indicated by the dotted lines with shading. Reproduced from Karaveg and Moremen (2005), with permission from American Society for Biochemistry and Molecular Biology via Copyright Clearance Center.

Acknowledgment

This work was supported by National Institutes of Health Research Grants GM047533 and RR05351 (to K.W.M.).

References

  1. Bause E, Bieberich E, Rolfs A, Volker C, Schmidt B. Molecular cloning and primary structure of Man9-mannosidase from human kidney. Eur. J. Biochem. 1993;217:535–540. doi: 10.1111/j.1432-1033.1993.tb18274.x. [DOI] [PubMed] [Google Scholar]
  2. Day YS, Baird CL, Rich RL, Myszka DG. Direct comparison of binding equilibrium, thermodynamic, and rate constants determined by surface- and solution-based biophysical methods. Protein Sci. 2002;11:1017–1025. doi: 10.1110/ps.4330102. [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Gonzalez DS, Karaveg K, Vandersall-Nairn AS, Lal A, Moremen KW. Identification, expression, and characterization of a cDNA encoding human endoplasmic reticulum mannosidase I, the enzyme that catalyzes the first mannose trimming step in mammalian Asn-linked oligosaccharide biosynthesis. J. Biol. Chem. 1999;274:21375–21386. doi: 10.1074/jbc.274.30.21375. [DOI] [PubMed] [Google Scholar]
  4. Helenius A, Aebi M. Roles of N-linked glycans in the endoplasmic reticulum. Annu. Rev. Biochem. 2004;73:1019–1049. doi: 10.1146/annurev.biochem.73.011303.073752. [DOI] [PubMed] [Google Scholar]
  5. Henrissat B, Bairoch A. Updating the sequence-based classification of glycosyl hydrolases. Biochem. J. 1996;316(Pt. 2):695–696. doi: 10.1042/bj3160695. [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Herscovics A. Processing glycosidases of Saccharomyces cerevisiae. Biochim. Biophys. Acta. 1999;1426:275–285. doi: 10.1016/s0304-4165(98)00129-9. [DOI] [PubMed] [Google Scholar]
  7. Herscovics A, Jelinek-Kelly S. A rapid method for assay of glycosidases involved in glycoprotein biosynthesis. Anal. Biochem. 1987;166:85–89. doi: 10.1016/0003-2697(87)90550-1. [DOI] [PubMed] [Google Scholar]
  8. Herscovics A, Schneikert J, Athanassiadis A, Moremen KW. Isolation of a mouse Golgi mannosidase cDNA, a member of a gene family conserved from yeast to mammals. J. Biol. Chem. 1994;269:9864–9871. [PubMed] [Google Scholar]
  9. Hirao K, Natsuka Y, Tamura T, Wada I, Morito D, Natsuka S, Romero P, Sleno B, Tremblay LO, Herscovics A, Nagata K, Hosokawa N. EDEM3, a soluble EDEM homolog, enhances glycoprotein ERAD and mannose trimming. J. Biol. Chem. 2006;281:9650–9658. doi: 10.1074/jbc.M512191200. [DOI] [PubMed] [Google Scholar]
  10. Hosokawa N, Tremblay LO, You Z, Herscovics A, Wada I, Nagata K. Enhancement of endoplasmic reticulum (ER) degradation of misfolded Null Hong Kong alpha1-antitrypsin by human ER mannosidase I. J. Biol. Chem. 2003;278:26287–26294. doi: 10.1074/jbc.M303395200. [DOI] [PubMed] [Google Scholar]
  11. Karaveg K, Moremen KW. Energetics of substrate binding and catalysis by class 1 (glycosylhydrolase family 47) alpha-mannosidases involved in N-glycan processing and endoplasmic reticulum quality control. J. Biol. Chem. 2005;280:29837–29848. doi: 10.1074/jbc.M505130200. [DOI] [PubMed] [Google Scholar]
  12. Karaveg K, Siriwardena A, Tempel W, Liu ZJ, Glushka J, Wang BC, Moremen KW. Mechanism of class 1 (glycosylhydrolase family 47) α-mannosidases involved in N-glycan processing and endoplasmic reticulum quality control. J. Biol. Chem. 2005;280:16197–16207. doi: 10.1074/jbc.M500119200. [DOI] [PubMed] [Google Scholar]
  13. Kornfeld R, Kornfeld S. Assembly of asparagine-linked oligosaccharides. Ann. Rev. Biochem. 1985;54:631–664. doi: 10.1146/annurev.bi.54.070185.003215. [DOI] [PubMed] [Google Scholar]
  14. Lal A, Pang P, Kalelkar S, Romero PA, Herscovics A, Moremen KW. Substrate specificities of recombinant murine Golgi alpha1,2-mannosidases IA and IB and comparison with endoplasmic reticulum and Golgi processing alpha1,2-mannosidases. Glycobiology. 1998;8:981–995. doi: 10.1093/glycob/8.10.981. [DOI] [PubMed] [Google Scholar]
  15. Lal A, Schutzbach JS, Forsee WT, Neame PJ, Moremen KW. Isolation and expression of murine and rabbit cDNAs encoding an alpha 1,2-mannosidase involved in the processing of asparagine-linked oligosaccharides. J. Biol. Chem. 1994;269:9872–9881. [PubMed] [Google Scholar]
  16. Lipari F, Herscovics A. Role of the cysteine residues in the alpha1,2-mannosidase involved in N-glycan biosynthesis in Saccharomyces cerevisiae: The conserved Cys340 and Cys385 residues form an essential disulfide bond. J. Biol. Chem. 1996;271:27615–27622. doi: 10.1074/jbc.271.44.27615. [DOI] [PubMed] [Google Scholar]
  17. Lobsanov YD, Vallee F, Imberty A, Yoshida T, Yip P, Herscovics A, Howell PL. Structure of Penicillium citrinum alpha 1,2-mannosidase reveals the basis for differences in specificity of the endoplasmic reticulum and Golgi class I enzymes. J. Biol. Chem. 2002;277:5620–5630. doi: 10.1074/jbc.M110243200. [DOI] [PubMed] [Google Scholar]
  18. Mast SW, Diekman K, Davis AW, Karaveg K, Sifers RN, Moremen KW. Human EDEM2, a novel homolog of family 47 glycosidases, is involved in ER-associated degradation of glycoproteins. Glycobiology. 2005;15:421–436. doi: 10.1093/glycob/cwi014. [DOI] [PubMed] [Google Scholar]
  19. Moremen K. α-mannosidases in Asparagine-linked oligosaccharide processing and catabolism. In: Ernst B, Hart G, Sinay P, editors. Oligosaccharides in Chemistry and Biology: A Comprehensive Handbook. Vol. II. New York: John Wiley & Sons Inc.; 2000. pp. 81–117. [Google Scholar]
  20. Moremen KW. Golgi alpha-mannosidase II deficiency in vertebrate systems: Implications for asparagine-linked oligosaccharide processing in mammals. Biochim. Biophys. Acta. 2002;1573:225–235. doi: 10.1016/s0304-4165(02)00388-4. [DOI] [PubMed] [Google Scholar]
  21. Moremen KW, Trimble RB, Herscovics A. Glycosidases of the asparagines-linked oligosaccharide processing pathway. Glycobiology. 1994;4:113–125. doi: 10.1093/glycob/4.2.113. [DOI] [PubMed] [Google Scholar]
  22. Spiro MJ, Spiro RG. Use of recombinant endomannosidase for evaluation of the processing of N-linked oligosaccharides of glycoproteins and their oligosaccharide-lipid precursors. Glycobiology. 2000;10:521–529. doi: 10.1093/glycob/10.5.521. [DOI] [PubMed] [Google Scholar]
  23. Tempel W, Karaveg K, Liu ZJ, Rose J, Wang BC, Moremen KW. Structure of mouse Golgi alpha-mannosidase IA reveals the molecular basis for substrate specificity among class 1 (family 47 glycosylhydrolase) alpha1,2-mannosidases. J. Biol. Chem. 2004;279:29774–29786. doi: 10.1074/jbc.M403065200. [DOI] [PubMed] [Google Scholar]
  24. Tremblay LO, Herscovics A. Characterization of a cDNA encoding a novel human Golgi alpha 1,2-mannosidase (IC) involved in N-glycan biosynthesis. J. Biol. Chem. 2000;275:31655–31660. doi: 10.1074/jbc.M004935200. [DOI] [PubMed] [Google Scholar]
  25. Vallee F, Karaveg K, Herscovics A, Moremen KW, Howell PL. Structural basis for catalysis and inhibition of N-glycan processing class I alpha 1,2-mannosidases. J. Biol. Chem. 2000a;275:41287–41298. doi: 10.1074/jbc.M006927200. [DOI] [PubMed] [Google Scholar]
  26. Vallee F, Lipari F, Yip P, Sleno B, Herscovics A, Howell PL. Crystal structure of a class I alpha1,2-mannosidase involved in N-glycan processing and endoplasmic reticulum quality control. EMBO J. 2000b;19:581–588. doi: 10.1093/emboj/19.4.581. [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Van Petegem F, Contreras H, Contreras R, Van Beeumen J. Trichoderma reesei alpha-1,2-mannosidase: Structural basis for the cleavage of four consecutive mannose residues. J. Mol. Biol. 2001;312:157–165. doi: 10.1006/jmbi.2001.4946. [DOI] [PubMed] [Google Scholar]
  28. Weng S, Spiro R. Endoplasmic reticulum kifunensine-resistant a-mannosidase is enzymatically and immunologically related to the cytosolic a-mannosidase. Arch. Biochem. Biophys. 1996;325:113–123. doi: 10.1006/abbi.1996.0014. [DOI] [PubMed] [Google Scholar]
  29. Wu Y, Swulius MT, Moremen KW, Sifers RN. Elucidation of the molecular logic by which misfolded alpha 1-antitrypsin is preferentially selected for degradation. Proc. Natl. Acad. Sci. USA. 2003;100:8229–8234. doi: 10.1073/pnas.1430537100. [DOI] [PMC free article] [PubMed] [Google Scholar]

RESOURCES