Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Engineering of the glycan-binding specificity of Agrocybe cylindracea galectin towards α(2,3)-linked sialic acid by saturation mutagenesis

Engineering of the glycan-binding specificity of Agrocybe cylindracea galectin towards... Abstract Sialic acid represents a critical sugar component located at the outermost position of glycoconjugates, playing important roles in extensive biological processes. To date, however, there have been only few probes which show affinity to α(2,3)-linked sialic acid-containing glycoconjugates. Agrocybe cylindracea galectin is known to have a relatively high affinity towards Neu5Acα(2,3)Galβ(1,4)Glc (3′-sialyl lactose), but it significantly recognizes various β-galactosides, such as Galβ(1,4)GlcNAcβ (LacNAc) and Galβ(1,3)GalNAcα (T-antigen). To eliminate this background specificity, we focused an acidic amino acid residue (Glu86), which interacts with the glucose unit of 3′-sialyl lactose and substituted it with all other amino acids. Carbohydrate-binding specificity of the derived 14 mutants was analysed by surface plasmon resonance, and it was found that E86D mutant (Glu86 substituted with Asp) substantially lost the binding ability to LacNAc and T-antigen, while it retained the high affinity for 3′-sialyl lactose. Further, frontal affinity chromatography analysis using 132 pyridylaminated oligosaccharides confirmed that the E86D mutant had a strong preference for α(2,3)-disialo biantennary N-linked glycan. However, it showed the large decrease in the affinity for any of the asialo complex-type N-glycans and the glycolipid-type glycans. Thus, the developed mutant E86D will be of practical use in various fields relevant to cell biology and glycotechnology. Agrocybe cylindracea galectin, frontal affinity chromatography, hydrogen bond, saturation mutagenesis, sialo-binding lectin Sialic acid-containing glycoconjugates are known to interact with several lectins (1–3). Sialic acids consist of a family of acidic, nine-carbon sugars that are typically located at the termini of glycoconjugates. Increased sialylation of cell surface glycoconjugates is among the key molecular changes associated with malignant transformation (4) or cancer progression (5–8). In previous studies, lectin column chromatography (9) and lectin immunosorbent assay (10) were used for clinical investigation of cancer-associated carbohydrate alterations towards prostate-specific antigens (PSAs). For diagnosis of prostate cancer, it is essential that the differential binding of free serum PSAs to lectins, which recognize α(2,3)-linked sialic acids, between prostate cancer and benign prostate hypertrophy (11). Several lectins have been reported to interact with α(2,3)-linked sialic acid-containing glycoconjugates; Maackia amurensis leucoagglutin (MAL) (12), Maackia amurensis haemagglutin (MAH) (13) and Agrocybe cylindracea galectin (ACG) (14). ACG shows a high affinity for the NeuAcα(2,3)Galβ unit but recognizes also N-acetylneuraminic acid (NeuAc) and lactose, indicating that it has a broad range specificity for N-glycans and gangliosides (14). Ser44, Arg77 and Trp83 are important for interactions with the sialic acid residue, whereas Ser44, Pro45 and Asn46 in the unique loop region are necessary for the recognition of NeuAcα(2,3)Lac and lactose. Further, side-chains of His62, Asn46, Arg66, Asn75 and Glu86 form hydrogen bonds with the galactose moiety of lactose in the ACG/lactose complex (15). In order to improve the substrate specificity of ACG, we constructed ACG mutants where Glu86 in the galactose-binding site was substituted by all other amino acids and examined their binding specificity by surface plasmon resonance (SPR) and frontal affinity chromatography (FAC) (16). Here, we report that E86D mutant exhibits a specific binding to complex type N-glycans with the bi-terminal α(2,3)-linked sialic acid compared with the wild-type ACG (wtACG) and MAL. Materials and Methods Construction of expression vector for A. cylindracea galectin ACG gene was designed by back-translation from its amino acid sequence (17) and synthesized by Integrated DNA Technologies (Coralville, IA, USA). The ACG-coding region flanked by the Nhel and Xhol sites was amplified by means of the polymerase chain reaction (PCR) using a pair of primers (5′-GGTAGCTAGCGATGAAGTCGATACCACTTCTGCGGTTAACATTTAC-3′ and 5′-ATGCCTCGAGTCAAGCCAGACCAGTGTACGTAA-3′), and subcloned into pET-28a(+) (Novagen) to yield pET-28a/ACG, which can express ACG with the N-terminal hexa-His tag followed by the thrombin cleavage site. Constructions of ACG mutants ACG mutants were constructed by using KOD-Plus-Mutagenesis Kit (Toyobo) and a pair of primers (5′-NNNCAGCGTGTTTCTAACGTAGCAAACCAGTTCATTGG-3′ and 5′-GACCAGCCAAGGAGCGTTCGGCTGACGGGAGTTGAACACGAT-3′). NNN indicates a codon for the position 86 and the wild-type codon (GAA for Glu) was substituted by GCT (Ala), AGA (Arg), AAC (Asn), GAT (Asp), TGC (Cys), CAG (Gln), GGT (Gly), CAT (His), ATC (Ile), CTT (Leu), AAG (Lys), ATG (Met), TTC (Phe), CCA (Pro), TCT (Ser), ACA (Thr), TGG (Trp), TAT (Tyr) or GTC (Val) to yield pET-28a/ACG-E86A, -E86R, -E86N, -E86D, -E86C, -E86Q, -E86G, -E86H, -E86I, -E86L, -E86K, -E86M, -E86F, -E86P, -E86S, -E86T, -E86W, -E86Y and -E86V, respectively. PCR was carried out with KOD-Plus-DNA polymerase in a thermal cycling process (94°C for 2 min; 12 cycles at 94°C for 10 s, 68°C for 1 min/kb of plasmid DNA). Resultant products were treated with DpnI at 37°C for 1 h and used for transformation of Escherichia coli JM109 (dam-) cells after ligation with T4 polynucleotide kinase and ligase. The nucleotide sequence of the entire ACG coding region in mutant plasmids was confirmed by sequencing. Expression and purification of ACG mutants Mutant plasmids were introduced into E. coli strain Rosetta 2 (Novagen) cells and transformants were selected on a LB plate containing 50 µg/ml kanamycin sulphaate and 30 µg/ml chloramphenicol. The E. coli cells harbouring the wild-type or mutant plasmid were grown to the mid log phase at 37°C in LB medium, and then at 30°C in the presence of 0.1 mM isopropyl-β-d-thiogalactoside for 2 h. The E. coli cells were collected by centrifugation, and 2 g of the pellet (wet weight) was resuspended in 35 ml of PBS (pH 7.4) containing 1% Triton X-100. After one freeze–thaw, the mixture was sonicated for 5 min and then centrifuged at 10,000g for 20 min at 4°C. The supernatant was loaded onto a 1-ml His GraviTrap column (GE Healthcare). The His-tagged wild-type and mutant ACGs were purified according to the manufacturer's instructions, dialysed twice against 1-l volumes of PBS (pH 7.4) at 4°C overnight and stored at 4°C until use. Protein concentration was determined by BCA Protein Assay Kit (Thermo Scientific). Purity of the isolated wtACG and mutant ACGs was analysed by SDS–polyacrylamide gel electrophoresis at a constant current of 50 mA/gel using a 5–20% (w/v) gradient acrylamide gel under reducing conditions, followed by silver staining with Bio-Rad Silver Stain. SPR analysis SPR experiments were carried out at 25°C using a BIAcore 3000 system (GE Healthcare) at a flow rate of 20 µl/min with PBS (pH 7.4) as the running buffer. Immobilization of 2.5 µg/ml of multivalent biotinylated saccharide polymers (GlycoTech, MD, USA) on a flow cell of a SA pioneer sensor chip (BIAcore) was done at 5 µl/min with the amine coupling kit (BIAcore) until the desired amount of the polymers was captured. ACG proteins (100 µg/ml in PBS) were injected over the SA-captured surface for 90 s for association, followed by a 300-s-long running buffer injection for dissociation. Parallel injections of analytes over a control surface were always performed for background correction. The protein-captured surface was regenerated for subsequent injections by two successive injections of 100 µl of 50 mM NaOH. Response units (RU) were calculated according to the BIA evaluation software (version 3.1). FAC assay FAC assay was performed on a lectin-immobilized column, using an automated FAC system, FAC-1 (Shimadzu), as described previously (18, 19). wtACG, E86D and MAL were immobilized on N-hydroxysuccinimide-activated Sepharose 4 Fast Flow resins via N-terminal and lysyl amino groups of the polypeptides according to the manufacturer’s manual. The lectin-immobilized Sepharose gel was packed into a capsule-type miniature column (2 mm × 10 mm, bed volume of 31.4 µl) and equilibrated with 10 mM Tris–HCl (pH 7.4) containing 0.8% NaCl. An excess volume (0.5–1.0 ml) of 2.5 or 5.0 nM PA oligosaccharides in 10 mM Tris–HCl (pH 7.4) containing 0.8% NaCl was successively injected into the column by an autosampling system. Elution was performed at a flow rate of 0.125 ml/min in an oven (25°C) and the elution profile was monitored by fluorescence of the PA oligosaccharide (emission at 380 nm/excitation at 310 nm). The volume (V) of the elution front of each PA oligosaccharide was automatically determined. The retardation volume (V − V0) of the elution front compared with that of a negative control (V0), which had no affinity to the column was determined. The retardation volume was used to estimate a dissociation constant (Kd) based on the basic equation of FAC; Kd = Bt/(Vf − V0) − [A]0, where Bt is the effective ligand content in mol, Vf is the frontal volume and [A]0 is the initial molar concentration of each PA oligosaccharide that was applied to the column. The equation can be simplified to Kd = Bt/(Vf−V0), where [A]0 is negligibly small compared with Kd. The Bt value is a specific parameter for the affinity column used, and that of the lectin-immobilized column was determined by dose-response analysis using various concentrations (100–750 nM) of PA oligosaccharides. The association constant (Ka) was obtained as the inverse of Kd (Ka = 1/Kd). Results Expression and purification of mutant ACGs The wild-type and mutant His6-ACG fusion proteins were expressed in E. coli Rosetta 2 and identified as 17.4 and 34.0 kDa polypeptides by SDS–PAGE analysis of randomly chosen clones. These bands are assumed to be a monomer and dimer, respectively, and slightly smaller than 19.8 kDa deduced from the nucleotide sequence. We confirmed by sequencing that plasmid DNAs extracted from mutant clones contained the desired mutations. The wild-type ACG and mutant ACGs, except E86N, E86I, E86L, E86F and E86V were recovered in the soluble fraction and purified to the homogeneity by Ni2+-affinity chromatography (Fig. 1). The yield of purified proteins was 4–20 mg from 1-l culture. Fig. 1 View largeDownload slide SDS–polyacrylamide gel electrophoresis analysis of purified mutant ACGs. Mutant proteins were analysed by electrophoresis on 5–20% SDS–polyacrylamide gradient gel, following by silver staining. An arrowhead indicates the purified monomer and dimer ACG mutant proteins. Fig. 1 View largeDownload slide SDS–polyacrylamide gel electrophoresis analysis of purified mutant ACGs. Mutant proteins were analysed by electrophoresis on 5–20% SDS–polyacrylamide gradient gel, following by silver staining. An arrowhead indicates the purified monomer and dimer ACG mutant proteins. Screening of Glu86 mutants with the higher specificity towards α(2,3)-linked sialic acids Real-time interactions of the mutant ACGs with multivalent biotinylated saccharide polymers (3′-sialyl lactose (Neu5Acα(2,3)Galβ(1,4)Glc), 3′-sialyl LacNAc (Neu5Acα(2,3)Galβ(1,4)GlcNAcβ), LacNAc (Galβ(1,4)GlcNAcβ), 3′-sialyl T (Neu5Acα(2,3)Galβ(1,3)GalNAcα), T-antigen (Galβ(1,3)GalNAcα), 3′-sialyl galactose (Neu5Acα(2,3)Gal) and 6′-sialyl glactose (Neu5Acα(2,6)Galβ)) were analysed by SPR. Most of the mutant ACGs interacted with 3′-sialyl lactose, which are known to binds the native ACG (14). These results indicate that the recombinant ACG mutants have the lectin activities like the native ACG. Their binding properties were compared with those of MAL, MAH, Erythrina cristagalii agglutinin (ECA) (20) and Sambucus nigra agglutinin (SNA) (21), which are known to bind terminal sialic acids (MAL, MAH and SNA) and terminal LacNAc (ECA). The wtACG showed a high affinity for the immobilized 3′-sialyl lactose (6,241 RU) and 3′-sialyl LacNAc (4360 RU), as well as LacNAc (4881 RU), 3′-sialyl T (4858 RU) and T-antigen (5512 RU) (Fig. 2). This indicates that wtACG interacts with not only the terminal sialic acid but β-galactosides like LacNAc and T-antigen and has a poor binding specificity to terminal sialic acids. In contrast, MAL showed a lower affinity for 3′-sialyl lactose (908 RU), 3′-sialyl LacNAc (962 RU) and LacNAc (384 RU), and did not interact with 3′-sialyl T and T-antigen. Most of the mutant ACGs lost the affinity for T-antigen, and many mutants exhibited weakened or undetectable affinity for 3′-sialyl T. Though the affinity for LacNAc was weakened, several mutants kept a strong affinity for 3′-sialyl lactose and 3′-sialyl LacNAc. Especially E86D mutant retained a high affinity for 3′-sialyl lactose (4925 RU) and 3′-sialyl LacNAc (2722 RU) but had reduced the affinity for LacNAc (825 RU), 3′-sialyl T (147 RU) and T-antigen (not detected), indicating the MAL-like carbohydrate-binding specificity. Fig. 2 View largeDownload slide Binding acitivity of mutant ACGs and native lectins for immobilized multivalent biotinylated polysaccharides. Poly[N-(2-hydroxyethyl) acrylamide]-biotinylated polysaccharides used are 3′-sialyl lactose, 3′-sialyl LacNAc, LacNAc, 3′-sialyl T, T-antigen, 3′-sialyl galactose, and 6′-sialyl galactose. RU values were classified into eight tones. Fig. 2 View largeDownload slide Binding acitivity of mutant ACGs and native lectins for immobilized multivalent biotinylated polysaccharides. Poly[N-(2-hydroxyethyl) acrylamide]-biotinylated polysaccharides used are 3′-sialyl lactose, 3′-sialyl LacNAc, LacNAc, 3′-sialyl T, T-antigen, 3′-sialyl galactose, and 6′-sialyl galactose. RU values were classified into eight tones. Sugar-binding specificity of wtACG, E86D and MAL Binding features of wtACG, E86D and MAL with the representative asialo complex-type N-glycans, sialo-complex type N-glycans and glycolipid type glycans are summarized in Fig. 3 and Table I. To find differences in the binding features of E86D from wtACG and MAL, we performed FAC analysis using 132 PA oligosaccharides, including 63 N-glycans (001–017, 053, 056–058, 101–108, 201–203, 301, 302, 304–308, 313, 314, 323, 401–406, 410, 418–420, 501–510, 602), 39 glycolipid-type glycans (701–713, 715–739, 743), and 30 other glycans (901–903, 905–911, 913–915, 918–921, 927–933, 942–947) (Fig. 3 and Supplementary Fig. S1). Fig. 3 View largeDownload slide FAC analysis of wtACG, E86D and MAL with 132 PA oligosaccharides. Glycan types are classified into seven colours. The schematic representation of each oligosaccharide structure is the same as described previously (22, 23). Symbols corresponding to each monosaccharide are shown in the inset of the panel C. Thin and thick bars represent the α- and β-linkage, respectively. Linkage positions between monosaccharides are shown by numbers in the right side figure. Fig. 3 View largeDownload slide FAC analysis of wtACG, E86D and MAL with 132 PA oligosaccharides. Glycan types are classified into seven colours. The schematic representation of each oligosaccharide structure is the same as described previously (22, 23). Symbols corresponding to each monosaccharide are shown in the inset of the panel C. Thin and thick bars represent the α- and β-linkage, respectively. Linkage positions between monosaccharides are shown by numbers in the right side figure. Table I. Comparison of sugar-binding affinity of wtACG, E86D and MAL for representative glycans determined by FAC.     Kd in µM (affinity relative to 307)   Sugar number (types)  Structural features  wtACG  E86D  MAL  Asialo complex-type N-glycans              307  biantennary (type 2)  28 (1)  247 (1)  417 (1)      313  triantennary (type 2)  8 (3.5)  87 (2.8)  166 (2.5)      323  tetraantennary (type 2)  7 (4)  82 (3.0)  163 (2.6)      405  biantennary (type 2)/core Fuc  26 (0.9)  239 (1.0)  546 (0.8)      410  triantennary (type 2)/core Fuc  8 (3.5)  97 (2.5)  196 (2.1)      418  tetraantennary (type 2)/core Fuc  6 (4.7)  97 (2.5)  167 (2.5)  Sialo complex-type N-glycans(biantennary)              501  α(2,6/1-3 branch)/asialo (1-6 branch)  70 (0.4)  921 (0.3)  724 (0.6)      502  asialo (1-3 branch)/α(2,6/1-6 branch)  58 (0.5)  909 (0.3)  >1550      503  α(2,6/1-3 branch)/α(2,6/1-6 branch)  >110  >1440  >1550      505  α(2,6/1-3 branch)/α(2,3/1-6 branch)  5 (5.6)  85 (2.9)  18 (23)      507  α(2,3/1-3 branch)/asialo (1-6 branch)  3 (9.3)  41 (6.0)  22 (19)      508  asialo (1-3 branch)/α(2,3/1-6 branch)  4 (7.0)  43 (5.7)  15 (28)    A mean value of 507 and 508  3.5 (8.2)  42 (5.9)  18.5 (23)      509  α(2,3/1-3 branch)/α(2,6/1-6 branch)  5 (5.6)  74 (3.3)  33 (13)      510  α(2,3/1-3 branch)/α(2,3/1-6 branch)  2 (14)  7 (35)  13 (32)      602  α(2,3/1-3 branch)/α(2,3/1-3 branch)/core Fuc  2 (14)  9 (27)  12 (35)  Glycolipid-type glycans              710  GD1a  5 (5.6)  654 (0.4)  >1550      712  GT1b  4 (7.0)  654 (0.4)  >1550      713  GQ1b  3 (9.3)  314 (0.8)  >1550      737  β3Gal3Lac  6 (4.7)  400 (0.6)  >1550      738  β3Gal4Lac  3 (9.3)  237 (1.0)  >1550      Kd in µM (affinity relative to 307)   Sugar number (types)  Structural features  wtACG  E86D  MAL  Asialo complex-type N-glycans              307  biantennary (type 2)  28 (1)  247 (1)  417 (1)      313  triantennary (type 2)  8 (3.5)  87 (2.8)  166 (2.5)      323  tetraantennary (type 2)  7 (4)  82 (3.0)  163 (2.6)      405  biantennary (type 2)/core Fuc  26 (0.9)  239 (1.0)  546 (0.8)      410  triantennary (type 2)/core Fuc  8 (3.5)  97 (2.5)  196 (2.1)      418  tetraantennary (type 2)/core Fuc  6 (4.7)  97 (2.5)  167 (2.5)  Sialo complex-type N-glycans(biantennary)              501  α(2,6/1-3 branch)/asialo (1-6 branch)  70 (0.4)  921 (0.3)  724 (0.6)      502  asialo (1-3 branch)/α(2,6/1-6 branch)  58 (0.5)  909 (0.3)  >1550      503  α(2,6/1-3 branch)/α(2,6/1-6 branch)  >110  >1440  >1550      505  α(2,6/1-3 branch)/α(2,3/1-6 branch)  5 (5.6)  85 (2.9)  18 (23)      507  α(2,3/1-3 branch)/asialo (1-6 branch)  3 (9.3)  41 (6.0)  22 (19)      508  asialo (1-3 branch)/α(2,3/1-6 branch)  4 (7.0)  43 (5.7)  15 (28)    A mean value of 507 and 508  3.5 (8.2)  42 (5.9)  18.5 (23)      509  α(2,3/1-3 branch)/α(2,6/1-6 branch)  5 (5.6)  74 (3.3)  33 (13)      510  α(2,3/1-3 branch)/α(2,3/1-6 branch)  2 (14)  7 (35)  13 (32)      602  α(2,3/1-3 branch)/α(2,3/1-3 branch)/core Fuc  2 (14)  9 (27)  12 (35)  Glycolipid-type glycans              710  GD1a  5 (5.6)  654 (0.4)  >1550      712  GT1b  4 (7.0)  654 (0.4)  >1550      713  GQ1b  3 (9.3)  314 (0.8)  >1550      737  β3Gal3Lac  6 (4.7)  400 (0.6)  >1550      738  β3Gal4Lac  3 (9.3)  237 (1.0)  >1550  View Large The wtACG (505, 5 µM; 507, 3 µM; 508, 4 µM; 509, 5 µM; 510, 2 µM; 602, 2 µM), E86D (505, 85 µM; 507, 41 µM; 508, 43 µM; 509, 74 µM; 510, 7 µM; 602, 9 µM) and MAL (505, 18 µM; 507, 22 µM; 508, 15 µM; 509, 33 µM; 510, 13 µM; 602, 12 µM: the values written next to the sugar chain numbers is the Kd value) exhibit the high affinity binding to terminal Siaα(2,3)-Gal type N-glycans, but do not bind high mannose type N-glycans (001–017), agalacto-complex type N-glycans (101–108, 201–203) and Siaα(2,6)-Gal type N-glycans (501–504). The wtACG also binds asialo complex-type N-glycans (313, 8 µM; 323, 7 µM; 410, 8 µM; 418, 6 µM) and gangliosides with α(2,3)-linked NeuAc, GD1a (710, 5 µM), GT1b (712, 4 µM), GQ1b (713, 3 µM), and some β-galactans (737, 6 µM; 738, 3 µM). A loss of the terminal α(2,3)-linked Neu5Ac from GD1a decreased the binding affinity to 29 µM (709 = GM1) and a further loss of β(1,3)-linked galactose in GM2 (708, 110 µM) completely abolished the binding affinity. If the numbers of β1–β3-linked galactose increased, the binding affinities of β-galactan were significantly higher (701, 96 µM; 735, 37 µM; 736, 14 µM, 737, 6 µM; 738, 3 µM) (Supplementary Table S1). These results indicate that glycans with the terminal α(2,3)-linked Neu5Ac or β(1,3)-linked galactose enhance the affinity for wtACG. As these glycans’ binding features were also shown in native ACG (24), the recombinant wtACG was considered to fold correctly in E. coli soluble fraction. In contrast, MAL does not bind most glycolipid type glycans (701–713, 715–723, 725–733, 735–739, 743, >1550 µM) including α(2,3) NeuAc-linked glycans (705, 707–713) and other glycans (901, 905–911, 913–915, 919–921, 927–933, 942–947, >1550 µM). Thus, MAL binds strictly α(2,3)-Gal-linked sialo-complex-type N-glycans but has a weak galectin-like activity since it binds asialo complex-type N-glycans with the Galβ(1,4)GalNac unit (313, 166 µM; 323, 163 µM; 410, 196 µM; 418, 167 µM). The E86D mutation in the glycan-binding site of ACG resulted in the large decrease in the binding affinity for asialo complex-type N-glycans (313, 87 µM; 323, 82 µM 410, 97 µM; 418, 97 µM), sialo-complex type N-glycans except α(2,3)-disialo biantennary N-linked glycan (505, 85 µM; 507, 41 µM; 508, 43 µM; 509, 74 µM), glycolipid-type glycans (710, 654 µM; 711, 1296 µM; 712, 654 µM; 713, 314 µM; 737, 400 µM; 738, 237 µM) and other glycans (902, 680 µM; 903, 350 µM; 918, 470 µM) (Supplementary Table S1). Thus, E86D exhibits a rather strict binding to the α(2,3)-disialo biantennary N-linked glycans (510, 7 µM; 602, 9 µM). Binding properties of MAL are similar to those of E86D but MAL has a higher binding affinity for mono-terminal α(2,3)-Gal linked glycans (505, 507–509) (Table I). Discussion FAC analysis of native ACG using 41 PA oligosaccharides (24) and of recombinant ACG using 132 PA oligosaccharides (this study) showed that masking of the terminal galactose with α(2,6)-linked NeuAc (e.g. 501–504, 704) completely abolished the binding affinity. The affinity of ACG for gangliosides was attributed to the presence of the Galβ(1,3)GalNAc unit (i.e. T-antigen) (24). This study showed that wtACG-bound glycans with the GalNAcα(1,3)Gal (719, 23 µM; 720, 12 µM) or GlcNAcβ(1,3)Gal unit (724, 53 µM; 725, 18 µM; 728, 48 µM; 732, 19 µM; 733, 16 µM; 734, 15 µM; 902, 43 µM; 903, 26 µM; 905, 15 µM), indicating that GalNAcα(1,3)Gal and GlcNAcβ(1,3)Gal are the additional recognition units of ACG. Thus, wtACG has a broad substrate specificity in the recognition of the galactose-containing disaccharide units. In fact, it is now evident that galectins recognize diverse disaccharides defined by the Galβ-(syn)-gauche configuration (25). Ban et al. (15) determined the X-ray structures of the ACG/lactose and ACG/3′-sialyl lactose complexes. They revealed that ACG is composed of a β-sandwich of two anti-parallel six-stranded sheets with a shallow concavity. ACG has a conserved carbohydrate recognition domain (CRD), which shares a β-galactoside recognition pattern with highly conserved side-chains, and was categorized as prototype galectins, like mushroom Coprinus cinereus galectin II (Cgl-II) (26), human galectin 1 (hGal-1) (27) and 7 (hGal-7) (28), chicken galectin 16 (CG-16) (29) and conger eel congerin I (ConI) (30). Interestingly, ACG has two unique residues (Ser44 and Asn46) inserted in the CRD domain, which serve as the recognition of α(2,3)-linked sialic acid (15). In the ACG/3′-sialyl lactose complex, O4 of the sialic acid residue forms a H-bond with the main-chain O atom of Ser44, O8 forms a H-bond with NE1 of Trp83 and O9 forms a H-bond with NH2 of Arg77, while O3 of the glucose residue forms H-bonds with Arg66 (NH2) and Glu86 (OE1). The Arg residue in the S4 strand and the Glu residue in the S6 strand were conserved in other prototype galectins (15), suggesting that these two amino acids are important in forming H bonds with O3 of the glucose. Arg66 (NH1) also forms a H-bond with the galactose residue O4 and O5, and Glu86 (OE2) and Asn75 (ND2) with galactose residue O6 (Table II and Fig. 4). Here, we performed saturation mutagenesis of Glu86 in the S6 strand to decrease interactions with the glucose moiety of 3′-sialyl lactose, in the hope that Asn75 can keep the molecular architecture of CRD with the NeuAcα(2,3)Gal unit. In fact, many galectins are known to preserve the significant affinity for lactosamine-type disaccharides via the recognition of galactose O4 and O6 (31–33). Fig. 4 View largeDownload slide X-ray structures of the ACG/lactose and ACG/3′-sialyl lactose complex highlighting the glycan binding site. The 3D structures of the ACG/lactose complex (A; PDB 1WW6) and the ACG/3′-sialyl lactose (B; PDB 1WW4) were drawn with the Cn3D program (38). Oxygen and carbon atoms of disaccharides are indicated in red and grey, respectively. The numbers of oxygen atoms indicate glucose O3, galactose O4, galactose O5 and galactose O6 in (A) and glucose O3, NeuAc O4, NeuAc O8 and NeuAc O9 in (B). Fig. 4 View largeDownload slide X-ray structures of the ACG/lactose and ACG/3′-sialyl lactose complex highlighting the glycan binding site. The 3D structures of the ACG/lactose complex (A; PDB 1WW6) and the ACG/3′-sialyl lactose (B; PDB 1WW4) were drawn with the Cn3D program (38). Oxygen and carbon atoms of disaccharides are indicated in red and grey, respectively. The numbers of oxygen atoms indicate glucose O3, galactose O4, galactose O5 and galactose O6 in (A) and glucose O3, NeuAc O4, NeuAc O8 and NeuAc O9 in (B). Table II. Hydrogen bonds in the ACG/lactose (PDB 1WW6) and ACG/3′-sialyl lactose (PDB 1WW4) complexes. Ligand  Protein atom  Ligand atom  Lactose        Asn46 OD1  Gal O3 and O4    His62 NE2  Gal O4    Arg66 NH1  Gal O4 and O5    NH1  Glc O3    NH2  Glc O3    Asn75 ND2  Gal O6    Glu86 OE1  Gal O6, Glc O3    OE2  Glc O3  3′-Sialyl lactose        Ser44 O  Sia O4    Asn46 OD1  Gal O3 and O4    His62 NE2  Gal O4    Arg66 NH1  Gal O4 and O5    NH2  Glc O3    Asn75 ND2  Gal O6    Arg77 NH2  Sia O9    Trp83 NE1  Sia O8    Glu86 OE1  Glc O3    OE2  Gal O6  Ligand  Protein atom  Ligand atom  Lactose        Asn46 OD1  Gal O3 and O4    His62 NE2  Gal O4    Arg66 NH1  Gal O4 and O5    NH1  Glc O3    NH2  Glc O3    Asn75 ND2  Gal O6    Glu86 OE1  Gal O6, Glc O3    OE2  Glc O3  3′-Sialyl lactose        Ser44 O  Sia O4    Asn46 OD1  Gal O3 and O4    His62 NE2  Gal O4    Arg66 NH1  Gal O4 and O5    NH2  Glc O3    Asn75 ND2  Gal O6    Arg77 NH2  Sia O9    Trp83 NE1  Sia O8    Glu86 OE1  Glc O3    OE2  Gal O6  View Large In the ACG/lactose complex, Glu86 (OE1) forms a H-bond to glucose O3 of lactose (24). The substitution of Glu86 by Asp lacking one methylene carbon resulted in a dramatic decrease in the binding affinity for the asialo complex-type N-glycans, gangliosides and other glycans. Our results suggest that Galβ(1,4)GlcNAc in N-glycans and T-antigen seems not to make a H-bond with OD1 or OD2 of Asp86 via their reducing end sugar. It should be noted that E86Q exhibits a higher affinity than other mutants for 3′-sialyl T-antigen and T-antigen (Fig. 2). This indicates that OE1 or NE2 of Gln86 can form a H-bond with GalNAc of T-antigen in place of Glu86 OE1 in wtACG. The binding affinity of wtACG, E86D and MAL for N-glycans involving non-reducing bi-terminal α(2,3)-linked NeuAc (510, 602) was in the order of WT > E86D > MAL, whereas, that for 505, 507–509 was WT > MAL > E86D (Table I). In this regard, an obvious enhancing effect by α2-3 sialylation should be noted: affinity enhancement in E86D towards disialo forms relative to monosialo forms was as high as six times, when relating biantennary glycans, 507, 508 (monosialo-) and 510 (disialo-) were compared, whereas that in wtACG was merely additive (1.8 times, Table I). On the other hand, the effect of disialylation in the biantennary glycans was rather suppressive in MAL (1.4 times), though that exhibited high affinity enhancement by mono α2-3 sialylation, when asialo-form glycan (307) and monosialo-form glycans (507,508) were compared (23 times, Table I). Thus, the E86D mutant lectin could be regarded as ‘novel’ in terms of affinity-enhancing performance towards multiple sialylation. For this proof, however, further study including FAC analysis using triantennary and tetraantennary silaoglycans is necessary. Previous random mutagenesis approaches such as the overlap-extension method (34), error-prone PCR (35) and phage display system (36), and error-prone PCR and ribosome display system (37) yielded sialic acid-binding lectins (MAH, earthworm 29-kDa galactose-binding lectin) with the novel substrate binding specificity. These gene-engineered lectins are used to distinguish the cell lineage or the glycoform of proteins by lectin profiling, but their sugar specificities are not characterized in details (34–36). Here, we carried out the saturation mutagenesis of ACG Glu86 in the galactose-binding site and characterized the sugar-binding specificity of mutants by FAC with a variety of glycans. We succeeded in the identification of E86D with a narrow glycan-binding specificity. We hope that the developed mutant E86D will be of practical use in various fields relevant to cell biology and glycotechnology. Funding This work was supported in part by Kanto Bureau of Economy, Trade and Industry in Japan. Conflict of interest None declared. Acknowledgements We are indebted to Prof. T. Irimura (University of Tokyo) for his excellent support and thank Dr T. Mogi (MBL, University of Tokyo) for his critical reading of the manuscript. We also thank Prof. K. Yamamoto (University of Tokyo) for providing BIAcore 3000 for SPR analysis. Abbreviations Abbreviations ACG Agrocybe cylindracea galectin CRD carbohydrate recognition domain ECA Erythrina cristagalii agglutinin FAC frontal affinity chromatography LacNAc Galβ (1,4)GlcNAcβ MAH Maackia amurensis haemagglutinin MAL Maackia amurensis leucoagglutinin NeuAc N-acetylneuraminic acid PA pyridylaminated PSA prostate-specific antigen RU response unit 3′-sialyl galactose Neu5Acα(2,3)Gal 6′-sialyl glactose Neu5Acα(2,6)Galβ 3′-sialyl LacNAc Neu5Acα(2,3)Galβ(1,4)GlcNAcβ 3′-sialyl lactose Neu5Acα(2,3)Galβ(1,4)Glc 3’-sialyl T Neu5Acα(2,3)Galβ(1,3)GalNAcα SNA Sambucus nigra agglutinin SPR surface plasmon resonance T-antigen Galβ(1,3)GalNAcα References 1 Mandal C,  Mandal C.  Sialic acid binding lectins,  Experientia ,  1990, vol.  46 (pg.  433- 441) Google Scholar CrossRef Search ADS PubMed  2 Zeng FY,  Gabius HJ.  Sialic acid-binding proteins: characterization, biological function and application,  Z. Naturforsch. ,  1992, vol.  47c (pg.  641- 653) 3 Lehmann F,  Tiralongo E,  Tiralongo J.  Sialic acid-specific lectins: occurrence, specificity and function,  Cell Mol. Life Sci. ,  2006, vol.  63 (pg.  1331- 1354) Google Scholar CrossRef Search ADS PubMed  4 Michele DE,  Campbell KP.  Dystrophin-glycoprotein complex: post-translational processing and dystroglycan function,  J. Biol. Chem. ,  2003, vol.  278 (pg.  15457- 15460) Google Scholar CrossRef Search ADS PubMed  5 Tabares G,  Radcliffe CM,  Barrabes S,  Ramirez M,  Aleixandre RN,  Hoesel W,  Dwek RA,  Rudd PM,  Peracaula R,  de Llorens R.  Different glycan structures in prostate-specific antigen from prostate cancer sera in relation to seminal plasma PSA,  Glycobiology ,  2006, vol.  16 (pg.  132- 145) Google Scholar CrossRef Search ADS PubMed  6 Shah MH,  Telang SD,  Shah PM,  Patel PS.  Tissue and serum α2-3- and α2-6-linkage specific sialylation changes in oral carcinogenesis,  Glycoconj. J. ,  2008, vol.  25 (pg.  279- 290) Google Scholar CrossRef Search ADS PubMed  7 Wang FL,  Cui SX,  Sun LP,  Qu XJ,  Xie YY,  Zhou L,  Mu YL,  Tang W,  Wang YS.  High expression of a α2,3-linked sialic acid residues is associated with the metastatic potential of human gastric cancer,  Cancer Detect. Prev. ,  2009, vol.  32 (pg.  437- 443) Google Scholar CrossRef Search ADS PubMed  8 Cui H,  Lin Y,  Yue L,  Zhao X,  Liu J.  Differential expression of the α2,3-sialic acid residues in breast cancer is associated with metastatic potential,  Oncol. Rep. ,  2011, vol.  25 (pg.  1365- 1371) Google Scholar PubMed  9 Janković MM,  Kosanović MM.  Glycosylation of urinary prostate-specific antigen in benign hyperplasia and cancer: assessment by lectin-binding patterns,  Clin. Biochem. ,  2005, vol.  38 (pg.  58- 65) Google Scholar CrossRef Search ADS PubMed  10 Meany DL,  Zhang Z,  Sokoll LJ,  Zhang H,  Chan DW.  Glycoproteomics for prostate cancer detection: changes in serum PSA glycosylation patterns,  J. Proteome Res. ,  2009, vol.  8 (pg.  613- 619) Google Scholar CrossRef Search ADS PubMed  11 Ohyama C,  Hosono M,  Nitta K,  Oh-eda M,  Yoshikawa K,  Habuchi T,  Arai Y,  Fukuda M.  Carbohydrate structure and differential binding of prostate specific antigen to Maackia amurensis lectin between prostate cancer and benign prostate hypertrophy,  Glycobiology ,  2004, vol.  14 (pg.  671- 679) Google Scholar CrossRef Search ADS PubMed  12 Knibbs RN,  Goldstein IJ,  Ratcliffe RM,  Shibuya N.  Characterization of the carbohydrate binding specificity of the leukoagglutinating lectin from Maackia amurensis,  J. Biol. Chem. ,  1991, vol.  266 (pg.  83- 88) Google Scholar PubMed  13 Konami Y,  Yamamoto K,  Osawa T,  Irimura T.  Strong affinity of Maackia amurensis hemagglutinin (MAH) for sialic acid-containing Ser/Thr-linked carbohydrate chains of N-terminal octapeptides from human glycophorin A,  FEBS Lett. ,  1994, vol.  342 (pg.  334- 338) Google Scholar CrossRef Search ADS PubMed  14 Yagi F,  Miyamoto M,  Abe T,  Minami Y,  Tadera K,  Goldstein IJ.  Purification and carbohydrate-binding specificity of Agrocybe cylindracea lectin,  Glycoconj. J. ,  1997, vol.  14 (pg.  281- 288) Google Scholar CrossRef Search ADS PubMed  15 Ban M,  Yoon HJ,  Demirkan E,  Utsumi S,  Mikami B,  Yagi F.  Structural basis of a fungal galectin from Agrocybe cylindracea for recognizing sialoconjugate,  J. Mol. Biol. ,  2005, vol.  351 (pg.  695- 706) Google Scholar CrossRef Search ADS PubMed  16 Kasai K,  Oda Y,  Nishikawa M,  Ishii S.  Frontal affinity chromatography: theory, for its application to studies on specific interaction of biomolecules,  J. Chromatogr. ,  1986, vol.  376 (pg.  33- 47) Google Scholar CrossRef Search ADS PubMed  17 Yagi F,  Hiroyama H,  Kodama S.  Agrocybe cylindracea lectin is a member of the galectin family,  Glycoconj. J. ,  2001, vol.  18 (pg.  745- 749) Google Scholar CrossRef Search ADS PubMed  18 Nakamura S,  Yagi F,  Otani K,  Ito Y,  Hirabayashi J.  Comparative analysis of carbohydrate-binding properties of two tandem repeat-type Jacalin-related lectins, Castanea crenata agglutinin and Cycas revoluta leaf lectin,  FEBS J. ,  2005, vol.  272 (pg.  2784- 2799) Google Scholar CrossRef Search ADS PubMed  19 Tateno H,  Nakamura-Tsuruta S,  Hirabayashi J.  Frontal affinity chromatography: sugar–protein interactions,  Nat. Protoc. ,  2007, vol.  2 (pg.  2529- 2799) Google Scholar CrossRef Search ADS PubMed  20 Lis H,  Joubert FJ,  Sharon N.  Isolation and properties of N-acetyllactosamine-specific lectins from nine erythrina species,  Phytochemistry ,  1985, vol.  24 (pg.  2803- 2809) Google Scholar CrossRef Search ADS   21 Shibuya N,  Goldstein IJ,  Goldstein WF,  Nsimba-Lubaki M,  Peeters B,  Peumans WJ.  The elderberry (Sambucus nigra L.) bark lectin recognizes the Neu5Ac(α2-6)Gal/GalNAc sequence,  J. Biol. Chem. ,  1987, vol.  262 (pg.  1596- 1601) Google Scholar PubMed  22 Nakamura-Tsuruta S,  Kominami J,  Kamei M,  Koyama Y,  Suzuki T,  Isemura M,  Hirabayashi J.  Comparative analysis by frontal affinity chromatography of oligosaccharide specificity of GlcNAc-binding lectins, Griffonia simplicifolia lectin-II (GSL-II) and Boletopsis leucomelas lectin (BLL),  J. Biochem. ,  2006, vol.  140 (pg.  285- 291) Google Scholar CrossRef Search ADS PubMed  23 Yabe R,  Tateno H,  Hirabayashi J.  Frontal affinity chromatography analysis of constructs of DC-DIGNR and LSECtin extend evidence for affinity to agalactosylated N-glycans,  FEBS J. ,  2010, vol.  277 (pg.  4010- 4026) Google Scholar CrossRef Search ADS PubMed  24 Hirabayashi J,  Hashidate T,  Arata Y,  Nishi N,  Nakamura T,  Hirashima M,  Urashima T,  Oka T,  Futai M,  Muller WE,  Yagi F,  Kasai K.  Oligosaccharide specificity of galectins: a search by frontal affinity chromatography,  Biochim. Biophys. Acta. ,  2002, vol.  1572 (pg.  232- 254) Google Scholar CrossRef Search ADS PubMed  25 Iwaki J,  Tateno H,  Nishi N,  Minamisawa T,  Nakamura-Tsuruta S,  Itakura Y,  Kominami J,  Urashima T,  Nakamura T,  Hirabayashi J.  The Galβ-(syn)-gauche configuration is required for galectin-recognition disaccharides,  Biochim. Biophys. Acta. ,  2011, vol.  1810 (pg.  643- 651) Google Scholar CrossRef Search ADS PubMed  26 Walser PJ,  Haebel PW,  Kunzler M,  Sargent D,  Kues U,  Aebi M,  Ban N.  Structure and functional analysis of the fungal galectin CGL2,  Structure ,  2004, vol.  12 (pg.  689- 702) Google Scholar CrossRef Search ADS PubMed  27 López-Lucendo MF,  Solís D,  André S,  Hirabayashi J,  Kasai K,  Kaltner H,  Gabius HJ,  Romero A.  Growth-regulatory human galectin-1: crystallographic characterisation of the structural changes induced by single-site mutations and their impact on the thermodynamics of ligand binding,  J. Mol. Biol. ,  2004, vol.  343 (pg.  957- 970) Google Scholar CrossRef Search ADS PubMed  28 Leonidas DD,  Vatzaki EH,  Vorum H,  Celis JE,  Madsen P,  Acharya KR.  Structural basis for the recognition of carbohydrates by human galectin-7,  Biochemistry ,  1998, vol.  37 (pg.  13930- 13940) Google Scholar CrossRef Search ADS PubMed  29 Varela PF,  Solís D,  Díaz-Mauriño T,  Kaltner H,  Gabius HJ,  Romero A.  The 2.15 Å crystal structure of CG-16, the developmentally regulated homodimeric chicken galectin,  J. Mol. Biol. ,  1999, vol.  294 (pg.  537- 549) Google Scholar CrossRef Search ADS PubMed  30 Shirai T,  Mitsuyama C,  Niwa Y,  Matsui Y,  Hotta H,  Yamane T,  Kamiya H,  Ishii C,  Ogawa T,  Muramoto K.  High-resolution structure of the conger eel galectin, congerin I, in lactose liganded and ligand-free forms: emergence of a new structure class by accelerated evolution,  Struct. Fold. Des. ,  1999, vol.  7 (pg.  1223- 1233) Google Scholar CrossRef Search ADS   31 Leffler H,  Barondes SH.  Specificity of binding of three soluble rat lung lectins to substituted and unsubstituted mammalian β-galactosides,  J. Biol. Chem. ,  1986, vol.  261 (pg.  10119- 10126) Google Scholar PubMed  32 Ahmed H,  Pohl J,  Fink NE,  Strobel F,  Vasta GR.  The primary structure and carbohydrate specificity of a β-galactosyl-binding lectin from toad (Bufo arenarum Hensel) ovary reveal closer similarities to the mammalian galectin-1 than to the galectin from the clawed frog Xenopus laevis,  J. Biol. Chem. ,  1996, vol.  271 (pg.  33083- 33094) Google Scholar CrossRef Search ADS PubMed  33 Walser PJ,  Haebel PW,  Künzler M,  Sargent D,  Kües U,  Aebi M,  Ban N.  Structure and functional analysis of the fungal galectin CGL2,  Structure ,  2004, vol.  12 (pg.  689- 702) Google Scholar CrossRef Search ADS PubMed  34 Yim M,  Ono T,  Irimura T.  Mutated plant lectin library useful to identify different cells,  Proc. Natl. Acad. Sci. USA. ,  2001, vol.  98 (pg.  2222- 2225) Google Scholar CrossRef Search ADS   35 Maenuma K,  Yim M,  Komatsu K,  Hoshino M,  Takahashi Y,  Bovin N,  Irimura T.  Use of a library of mutated Maackia amurensis hemagglutinin for profiling the cell lineage and differentiation,  Proteomics ,  2008, vol.  8 (pg.  3274- 3283) Google Scholar CrossRef Search ADS PubMed  36 Maenuma K,  Yim M,  Komatsu K,  Hoshino M,  Tachiki-Fujioka A,  Takahashi K,  Hiki Y,  Bovin N,  Irimura T.  A library of mutated Maackia amurensis hemagglutinin distinguishes putative glycoforms of immunoglobulin A1 from IgA nephropathy patients,  J. Proteome Res. ,  2009, vol.  8 (pg.  3617- 3624) Google Scholar CrossRef Search ADS PubMed  37 Yabe R,  Suzuki R,  Kuno A,  Fujimoto Z,  Jigami Y,  Hirabayashi J.  Tailoring a novel sialic acid-binding lectin from a ricin-B chain-like galactose-binding protein by natural evolution-mimicry,  J. Biochem. ,  2007, vol.  141 (pg.  389- 399) Google Scholar CrossRef Search ADS PubMed  38 Wang Y,  Geer L,  Chappey C,  Kans J,  Bryant S.  Cn3D: sequence and structure views for Entrez,  Trends Biochem. Sci. ,  2000, vol.  25 (pg.  300- 302) Google Scholar CrossRef Search ADS PubMed  © The Authors 2011. Published by Oxford University Press on behalf of the Japanese Biochemical Society. All rights reserved http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png The Journal of Biochemistry Oxford University Press

Engineering of the glycan-binding specificity of Agrocybe cylindracea galectin towards α(2,3)-linked sialic acid by saturation mutagenesis

Loading next page...
 
/lp/oxford-university-press/engineering-of-the-glycan-binding-specificity-of-agrocybe-cylindracea-oPzhePm7cl

References (40)

Publisher
Oxford University Press
Copyright
© The Authors 2011. Published by Oxford University Press on behalf of the Japanese Biochemical Society. All rights reserved
ISSN
0021-924X
eISSN
1756-2651
DOI
10.1093/jb/mvr094
pmid
21813503
Publisher site
See Article on Publisher Site

Abstract

Abstract Sialic acid represents a critical sugar component located at the outermost position of glycoconjugates, playing important roles in extensive biological processes. To date, however, there have been only few probes which show affinity to α(2,3)-linked sialic acid-containing glycoconjugates. Agrocybe cylindracea galectin is known to have a relatively high affinity towards Neu5Acα(2,3)Galβ(1,4)Glc (3′-sialyl lactose), but it significantly recognizes various β-galactosides, such as Galβ(1,4)GlcNAcβ (LacNAc) and Galβ(1,3)GalNAcα (T-antigen). To eliminate this background specificity, we focused an acidic amino acid residue (Glu86), which interacts with the glucose unit of 3′-sialyl lactose and substituted it with all other amino acids. Carbohydrate-binding specificity of the derived 14 mutants was analysed by surface plasmon resonance, and it was found that E86D mutant (Glu86 substituted with Asp) substantially lost the binding ability to LacNAc and T-antigen, while it retained the high affinity for 3′-sialyl lactose. Further, frontal affinity chromatography analysis using 132 pyridylaminated oligosaccharides confirmed that the E86D mutant had a strong preference for α(2,3)-disialo biantennary N-linked glycan. However, it showed the large decrease in the affinity for any of the asialo complex-type N-glycans and the glycolipid-type glycans. Thus, the developed mutant E86D will be of practical use in various fields relevant to cell biology and glycotechnology. Agrocybe cylindracea galectin, frontal affinity chromatography, hydrogen bond, saturation mutagenesis, sialo-binding lectin Sialic acid-containing glycoconjugates are known to interact with several lectins (1–3). Sialic acids consist of a family of acidic, nine-carbon sugars that are typically located at the termini of glycoconjugates. Increased sialylation of cell surface glycoconjugates is among the key molecular changes associated with malignant transformation (4) or cancer progression (5–8). In previous studies, lectin column chromatography (9) and lectin immunosorbent assay (10) were used for clinical investigation of cancer-associated carbohydrate alterations towards prostate-specific antigens (PSAs). For diagnosis of prostate cancer, it is essential that the differential binding of free serum PSAs to lectins, which recognize α(2,3)-linked sialic acids, between prostate cancer and benign prostate hypertrophy (11). Several lectins have been reported to interact with α(2,3)-linked sialic acid-containing glycoconjugates; Maackia amurensis leucoagglutin (MAL) (12), Maackia amurensis haemagglutin (MAH) (13) and Agrocybe cylindracea galectin (ACG) (14). ACG shows a high affinity for the NeuAcα(2,3)Galβ unit but recognizes also N-acetylneuraminic acid (NeuAc) and lactose, indicating that it has a broad range specificity for N-glycans and gangliosides (14). Ser44, Arg77 and Trp83 are important for interactions with the sialic acid residue, whereas Ser44, Pro45 and Asn46 in the unique loop region are necessary for the recognition of NeuAcα(2,3)Lac and lactose. Further, side-chains of His62, Asn46, Arg66, Asn75 and Glu86 form hydrogen bonds with the galactose moiety of lactose in the ACG/lactose complex (15). In order to improve the substrate specificity of ACG, we constructed ACG mutants where Glu86 in the galactose-binding site was substituted by all other amino acids and examined their binding specificity by surface plasmon resonance (SPR) and frontal affinity chromatography (FAC) (16). Here, we report that E86D mutant exhibits a specific binding to complex type N-glycans with the bi-terminal α(2,3)-linked sialic acid compared with the wild-type ACG (wtACG) and MAL. Materials and Methods Construction of expression vector for A. cylindracea galectin ACG gene was designed by back-translation from its amino acid sequence (17) and synthesized by Integrated DNA Technologies (Coralville, IA, USA). The ACG-coding region flanked by the Nhel and Xhol sites was amplified by means of the polymerase chain reaction (PCR) using a pair of primers (5′-GGTAGCTAGCGATGAAGTCGATACCACTTCTGCGGTTAACATTTAC-3′ and 5′-ATGCCTCGAGTCAAGCCAGACCAGTGTACGTAA-3′), and subcloned into pET-28a(+) (Novagen) to yield pET-28a/ACG, which can express ACG with the N-terminal hexa-His tag followed by the thrombin cleavage site. Constructions of ACG mutants ACG mutants were constructed by using KOD-Plus-Mutagenesis Kit (Toyobo) and a pair of primers (5′-NNNCAGCGTGTTTCTAACGTAGCAAACCAGTTCATTGG-3′ and 5′-GACCAGCCAAGGAGCGTTCGGCTGACGGGAGTTGAACACGAT-3′). NNN indicates a codon for the position 86 and the wild-type codon (GAA for Glu) was substituted by GCT (Ala), AGA (Arg), AAC (Asn), GAT (Asp), TGC (Cys), CAG (Gln), GGT (Gly), CAT (His), ATC (Ile), CTT (Leu), AAG (Lys), ATG (Met), TTC (Phe), CCA (Pro), TCT (Ser), ACA (Thr), TGG (Trp), TAT (Tyr) or GTC (Val) to yield pET-28a/ACG-E86A, -E86R, -E86N, -E86D, -E86C, -E86Q, -E86G, -E86H, -E86I, -E86L, -E86K, -E86M, -E86F, -E86P, -E86S, -E86T, -E86W, -E86Y and -E86V, respectively. PCR was carried out with KOD-Plus-DNA polymerase in a thermal cycling process (94°C for 2 min; 12 cycles at 94°C for 10 s, 68°C for 1 min/kb of plasmid DNA). Resultant products were treated with DpnI at 37°C for 1 h and used for transformation of Escherichia coli JM109 (dam-) cells after ligation with T4 polynucleotide kinase and ligase. The nucleotide sequence of the entire ACG coding region in mutant plasmids was confirmed by sequencing. Expression and purification of ACG mutants Mutant plasmids were introduced into E. coli strain Rosetta 2 (Novagen) cells and transformants were selected on a LB plate containing 50 µg/ml kanamycin sulphaate and 30 µg/ml chloramphenicol. The E. coli cells harbouring the wild-type or mutant plasmid were grown to the mid log phase at 37°C in LB medium, and then at 30°C in the presence of 0.1 mM isopropyl-β-d-thiogalactoside for 2 h. The E. coli cells were collected by centrifugation, and 2 g of the pellet (wet weight) was resuspended in 35 ml of PBS (pH 7.4) containing 1% Triton X-100. After one freeze–thaw, the mixture was sonicated for 5 min and then centrifuged at 10,000g for 20 min at 4°C. The supernatant was loaded onto a 1-ml His GraviTrap column (GE Healthcare). The His-tagged wild-type and mutant ACGs were purified according to the manufacturer's instructions, dialysed twice against 1-l volumes of PBS (pH 7.4) at 4°C overnight and stored at 4°C until use. Protein concentration was determined by BCA Protein Assay Kit (Thermo Scientific). Purity of the isolated wtACG and mutant ACGs was analysed by SDS–polyacrylamide gel electrophoresis at a constant current of 50 mA/gel using a 5–20% (w/v) gradient acrylamide gel under reducing conditions, followed by silver staining with Bio-Rad Silver Stain. SPR analysis SPR experiments were carried out at 25°C using a BIAcore 3000 system (GE Healthcare) at a flow rate of 20 µl/min with PBS (pH 7.4) as the running buffer. Immobilization of 2.5 µg/ml of multivalent biotinylated saccharide polymers (GlycoTech, MD, USA) on a flow cell of a SA pioneer sensor chip (BIAcore) was done at 5 µl/min with the amine coupling kit (BIAcore) until the desired amount of the polymers was captured. ACG proteins (100 µg/ml in PBS) were injected over the SA-captured surface for 90 s for association, followed by a 300-s-long running buffer injection for dissociation. Parallel injections of analytes over a control surface were always performed for background correction. The protein-captured surface was regenerated for subsequent injections by two successive injections of 100 µl of 50 mM NaOH. Response units (RU) were calculated according to the BIA evaluation software (version 3.1). FAC assay FAC assay was performed on a lectin-immobilized column, using an automated FAC system, FAC-1 (Shimadzu), as described previously (18, 19). wtACG, E86D and MAL were immobilized on N-hydroxysuccinimide-activated Sepharose 4 Fast Flow resins via N-terminal and lysyl amino groups of the polypeptides according to the manufacturer’s manual. The lectin-immobilized Sepharose gel was packed into a capsule-type miniature column (2 mm × 10 mm, bed volume of 31.4 µl) and equilibrated with 10 mM Tris–HCl (pH 7.4) containing 0.8% NaCl. An excess volume (0.5–1.0 ml) of 2.5 or 5.0 nM PA oligosaccharides in 10 mM Tris–HCl (pH 7.4) containing 0.8% NaCl was successively injected into the column by an autosampling system. Elution was performed at a flow rate of 0.125 ml/min in an oven (25°C) and the elution profile was monitored by fluorescence of the PA oligosaccharide (emission at 380 nm/excitation at 310 nm). The volume (V) of the elution front of each PA oligosaccharide was automatically determined. The retardation volume (V − V0) of the elution front compared with that of a negative control (V0), which had no affinity to the column was determined. The retardation volume was used to estimate a dissociation constant (Kd) based on the basic equation of FAC; Kd = Bt/(Vf − V0) − [A]0, where Bt is the effective ligand content in mol, Vf is the frontal volume and [A]0 is the initial molar concentration of each PA oligosaccharide that was applied to the column. The equation can be simplified to Kd = Bt/(Vf−V0), where [A]0 is negligibly small compared with Kd. The Bt value is a specific parameter for the affinity column used, and that of the lectin-immobilized column was determined by dose-response analysis using various concentrations (100–750 nM) of PA oligosaccharides. The association constant (Ka) was obtained as the inverse of Kd (Ka = 1/Kd). Results Expression and purification of mutant ACGs The wild-type and mutant His6-ACG fusion proteins were expressed in E. coli Rosetta 2 and identified as 17.4 and 34.0 kDa polypeptides by SDS–PAGE analysis of randomly chosen clones. These bands are assumed to be a monomer and dimer, respectively, and slightly smaller than 19.8 kDa deduced from the nucleotide sequence. We confirmed by sequencing that plasmid DNAs extracted from mutant clones contained the desired mutations. The wild-type ACG and mutant ACGs, except E86N, E86I, E86L, E86F and E86V were recovered in the soluble fraction and purified to the homogeneity by Ni2+-affinity chromatography (Fig. 1). The yield of purified proteins was 4–20 mg from 1-l culture. Fig. 1 View largeDownload slide SDS–polyacrylamide gel electrophoresis analysis of purified mutant ACGs. Mutant proteins were analysed by electrophoresis on 5–20% SDS–polyacrylamide gradient gel, following by silver staining. An arrowhead indicates the purified monomer and dimer ACG mutant proteins. Fig. 1 View largeDownload slide SDS–polyacrylamide gel electrophoresis analysis of purified mutant ACGs. Mutant proteins were analysed by electrophoresis on 5–20% SDS–polyacrylamide gradient gel, following by silver staining. An arrowhead indicates the purified monomer and dimer ACG mutant proteins. Screening of Glu86 mutants with the higher specificity towards α(2,3)-linked sialic acids Real-time interactions of the mutant ACGs with multivalent biotinylated saccharide polymers (3′-sialyl lactose (Neu5Acα(2,3)Galβ(1,4)Glc), 3′-sialyl LacNAc (Neu5Acα(2,3)Galβ(1,4)GlcNAcβ), LacNAc (Galβ(1,4)GlcNAcβ), 3′-sialyl T (Neu5Acα(2,3)Galβ(1,3)GalNAcα), T-antigen (Galβ(1,3)GalNAcα), 3′-sialyl galactose (Neu5Acα(2,3)Gal) and 6′-sialyl glactose (Neu5Acα(2,6)Galβ)) were analysed by SPR. Most of the mutant ACGs interacted with 3′-sialyl lactose, which are known to binds the native ACG (14). These results indicate that the recombinant ACG mutants have the lectin activities like the native ACG. Their binding properties were compared with those of MAL, MAH, Erythrina cristagalii agglutinin (ECA) (20) and Sambucus nigra agglutinin (SNA) (21), which are known to bind terminal sialic acids (MAL, MAH and SNA) and terminal LacNAc (ECA). The wtACG showed a high affinity for the immobilized 3′-sialyl lactose (6,241 RU) and 3′-sialyl LacNAc (4360 RU), as well as LacNAc (4881 RU), 3′-sialyl T (4858 RU) and T-antigen (5512 RU) (Fig. 2). This indicates that wtACG interacts with not only the terminal sialic acid but β-galactosides like LacNAc and T-antigen and has a poor binding specificity to terminal sialic acids. In contrast, MAL showed a lower affinity for 3′-sialyl lactose (908 RU), 3′-sialyl LacNAc (962 RU) and LacNAc (384 RU), and did not interact with 3′-sialyl T and T-antigen. Most of the mutant ACGs lost the affinity for T-antigen, and many mutants exhibited weakened or undetectable affinity for 3′-sialyl T. Though the affinity for LacNAc was weakened, several mutants kept a strong affinity for 3′-sialyl lactose and 3′-sialyl LacNAc. Especially E86D mutant retained a high affinity for 3′-sialyl lactose (4925 RU) and 3′-sialyl LacNAc (2722 RU) but had reduced the affinity for LacNAc (825 RU), 3′-sialyl T (147 RU) and T-antigen (not detected), indicating the MAL-like carbohydrate-binding specificity. Fig. 2 View largeDownload slide Binding acitivity of mutant ACGs and native lectins for immobilized multivalent biotinylated polysaccharides. Poly[N-(2-hydroxyethyl) acrylamide]-biotinylated polysaccharides used are 3′-sialyl lactose, 3′-sialyl LacNAc, LacNAc, 3′-sialyl T, T-antigen, 3′-sialyl galactose, and 6′-sialyl galactose. RU values were classified into eight tones. Fig. 2 View largeDownload slide Binding acitivity of mutant ACGs and native lectins for immobilized multivalent biotinylated polysaccharides. Poly[N-(2-hydroxyethyl) acrylamide]-biotinylated polysaccharides used are 3′-sialyl lactose, 3′-sialyl LacNAc, LacNAc, 3′-sialyl T, T-antigen, 3′-sialyl galactose, and 6′-sialyl galactose. RU values were classified into eight tones. Sugar-binding specificity of wtACG, E86D and MAL Binding features of wtACG, E86D and MAL with the representative asialo complex-type N-glycans, sialo-complex type N-glycans and glycolipid type glycans are summarized in Fig. 3 and Table I. To find differences in the binding features of E86D from wtACG and MAL, we performed FAC analysis using 132 PA oligosaccharides, including 63 N-glycans (001–017, 053, 056–058, 101–108, 201–203, 301, 302, 304–308, 313, 314, 323, 401–406, 410, 418–420, 501–510, 602), 39 glycolipid-type glycans (701–713, 715–739, 743), and 30 other glycans (901–903, 905–911, 913–915, 918–921, 927–933, 942–947) (Fig. 3 and Supplementary Fig. S1). Fig. 3 View largeDownload slide FAC analysis of wtACG, E86D and MAL with 132 PA oligosaccharides. Glycan types are classified into seven colours. The schematic representation of each oligosaccharide structure is the same as described previously (22, 23). Symbols corresponding to each monosaccharide are shown in the inset of the panel C. Thin and thick bars represent the α- and β-linkage, respectively. Linkage positions between monosaccharides are shown by numbers in the right side figure. Fig. 3 View largeDownload slide FAC analysis of wtACG, E86D and MAL with 132 PA oligosaccharides. Glycan types are classified into seven colours. The schematic representation of each oligosaccharide structure is the same as described previously (22, 23). Symbols corresponding to each monosaccharide are shown in the inset of the panel C. Thin and thick bars represent the α- and β-linkage, respectively. Linkage positions between monosaccharides are shown by numbers in the right side figure. Table I. Comparison of sugar-binding affinity of wtACG, E86D and MAL for representative glycans determined by FAC.     Kd in µM (affinity relative to 307)   Sugar number (types)  Structural features  wtACG  E86D  MAL  Asialo complex-type N-glycans              307  biantennary (type 2)  28 (1)  247 (1)  417 (1)      313  triantennary (type 2)  8 (3.5)  87 (2.8)  166 (2.5)      323  tetraantennary (type 2)  7 (4)  82 (3.0)  163 (2.6)      405  biantennary (type 2)/core Fuc  26 (0.9)  239 (1.0)  546 (0.8)      410  triantennary (type 2)/core Fuc  8 (3.5)  97 (2.5)  196 (2.1)      418  tetraantennary (type 2)/core Fuc  6 (4.7)  97 (2.5)  167 (2.5)  Sialo complex-type N-glycans(biantennary)              501  α(2,6/1-3 branch)/asialo (1-6 branch)  70 (0.4)  921 (0.3)  724 (0.6)      502  asialo (1-3 branch)/α(2,6/1-6 branch)  58 (0.5)  909 (0.3)  >1550      503  α(2,6/1-3 branch)/α(2,6/1-6 branch)  >110  >1440  >1550      505  α(2,6/1-3 branch)/α(2,3/1-6 branch)  5 (5.6)  85 (2.9)  18 (23)      507  α(2,3/1-3 branch)/asialo (1-6 branch)  3 (9.3)  41 (6.0)  22 (19)      508  asialo (1-3 branch)/α(2,3/1-6 branch)  4 (7.0)  43 (5.7)  15 (28)    A mean value of 507 and 508  3.5 (8.2)  42 (5.9)  18.5 (23)      509  α(2,3/1-3 branch)/α(2,6/1-6 branch)  5 (5.6)  74 (3.3)  33 (13)      510  α(2,3/1-3 branch)/α(2,3/1-6 branch)  2 (14)  7 (35)  13 (32)      602  α(2,3/1-3 branch)/α(2,3/1-3 branch)/core Fuc  2 (14)  9 (27)  12 (35)  Glycolipid-type glycans              710  GD1a  5 (5.6)  654 (0.4)  >1550      712  GT1b  4 (7.0)  654 (0.4)  >1550      713  GQ1b  3 (9.3)  314 (0.8)  >1550      737  β3Gal3Lac  6 (4.7)  400 (0.6)  >1550      738  β3Gal4Lac  3 (9.3)  237 (1.0)  >1550      Kd in µM (affinity relative to 307)   Sugar number (types)  Structural features  wtACG  E86D  MAL  Asialo complex-type N-glycans              307  biantennary (type 2)  28 (1)  247 (1)  417 (1)      313  triantennary (type 2)  8 (3.5)  87 (2.8)  166 (2.5)      323  tetraantennary (type 2)  7 (4)  82 (3.0)  163 (2.6)      405  biantennary (type 2)/core Fuc  26 (0.9)  239 (1.0)  546 (0.8)      410  triantennary (type 2)/core Fuc  8 (3.5)  97 (2.5)  196 (2.1)      418  tetraantennary (type 2)/core Fuc  6 (4.7)  97 (2.5)  167 (2.5)  Sialo complex-type N-glycans(biantennary)              501  α(2,6/1-3 branch)/asialo (1-6 branch)  70 (0.4)  921 (0.3)  724 (0.6)      502  asialo (1-3 branch)/α(2,6/1-6 branch)  58 (0.5)  909 (0.3)  >1550      503  α(2,6/1-3 branch)/α(2,6/1-6 branch)  >110  >1440  >1550      505  α(2,6/1-3 branch)/α(2,3/1-6 branch)  5 (5.6)  85 (2.9)  18 (23)      507  α(2,3/1-3 branch)/asialo (1-6 branch)  3 (9.3)  41 (6.0)  22 (19)      508  asialo (1-3 branch)/α(2,3/1-6 branch)  4 (7.0)  43 (5.7)  15 (28)    A mean value of 507 and 508  3.5 (8.2)  42 (5.9)  18.5 (23)      509  α(2,3/1-3 branch)/α(2,6/1-6 branch)  5 (5.6)  74 (3.3)  33 (13)      510  α(2,3/1-3 branch)/α(2,3/1-6 branch)  2 (14)  7 (35)  13 (32)      602  α(2,3/1-3 branch)/α(2,3/1-3 branch)/core Fuc  2 (14)  9 (27)  12 (35)  Glycolipid-type glycans              710  GD1a  5 (5.6)  654 (0.4)  >1550      712  GT1b  4 (7.0)  654 (0.4)  >1550      713  GQ1b  3 (9.3)  314 (0.8)  >1550      737  β3Gal3Lac  6 (4.7)  400 (0.6)  >1550      738  β3Gal4Lac  3 (9.3)  237 (1.0)  >1550  View Large The wtACG (505, 5 µM; 507, 3 µM; 508, 4 µM; 509, 5 µM; 510, 2 µM; 602, 2 µM), E86D (505, 85 µM; 507, 41 µM; 508, 43 µM; 509, 74 µM; 510, 7 µM; 602, 9 µM) and MAL (505, 18 µM; 507, 22 µM; 508, 15 µM; 509, 33 µM; 510, 13 µM; 602, 12 µM: the values written next to the sugar chain numbers is the Kd value) exhibit the high affinity binding to terminal Siaα(2,3)-Gal type N-glycans, but do not bind high mannose type N-glycans (001–017), agalacto-complex type N-glycans (101–108, 201–203) and Siaα(2,6)-Gal type N-glycans (501–504). The wtACG also binds asialo complex-type N-glycans (313, 8 µM; 323, 7 µM; 410, 8 µM; 418, 6 µM) and gangliosides with α(2,3)-linked NeuAc, GD1a (710, 5 µM), GT1b (712, 4 µM), GQ1b (713, 3 µM), and some β-galactans (737, 6 µM; 738, 3 µM). A loss of the terminal α(2,3)-linked Neu5Ac from GD1a decreased the binding affinity to 29 µM (709 = GM1) and a further loss of β(1,3)-linked galactose in GM2 (708, 110 µM) completely abolished the binding affinity. If the numbers of β1–β3-linked galactose increased, the binding affinities of β-galactan were significantly higher (701, 96 µM; 735, 37 µM; 736, 14 µM, 737, 6 µM; 738, 3 µM) (Supplementary Table S1). These results indicate that glycans with the terminal α(2,3)-linked Neu5Ac or β(1,3)-linked galactose enhance the affinity for wtACG. As these glycans’ binding features were also shown in native ACG (24), the recombinant wtACG was considered to fold correctly in E. coli soluble fraction. In contrast, MAL does not bind most glycolipid type glycans (701–713, 715–723, 725–733, 735–739, 743, >1550 µM) including α(2,3) NeuAc-linked glycans (705, 707–713) and other glycans (901, 905–911, 913–915, 919–921, 927–933, 942–947, >1550 µM). Thus, MAL binds strictly α(2,3)-Gal-linked sialo-complex-type N-glycans but has a weak galectin-like activity since it binds asialo complex-type N-glycans with the Galβ(1,4)GalNac unit (313, 166 µM; 323, 163 µM; 410, 196 µM; 418, 167 µM). The E86D mutation in the glycan-binding site of ACG resulted in the large decrease in the binding affinity for asialo complex-type N-glycans (313, 87 µM; 323, 82 µM 410, 97 µM; 418, 97 µM), sialo-complex type N-glycans except α(2,3)-disialo biantennary N-linked glycan (505, 85 µM; 507, 41 µM; 508, 43 µM; 509, 74 µM), glycolipid-type glycans (710, 654 µM; 711, 1296 µM; 712, 654 µM; 713, 314 µM; 737, 400 µM; 738, 237 µM) and other glycans (902, 680 µM; 903, 350 µM; 918, 470 µM) (Supplementary Table S1). Thus, E86D exhibits a rather strict binding to the α(2,3)-disialo biantennary N-linked glycans (510, 7 µM; 602, 9 µM). Binding properties of MAL are similar to those of E86D but MAL has a higher binding affinity for mono-terminal α(2,3)-Gal linked glycans (505, 507–509) (Table I). Discussion FAC analysis of native ACG using 41 PA oligosaccharides (24) and of recombinant ACG using 132 PA oligosaccharides (this study) showed that masking of the terminal galactose with α(2,6)-linked NeuAc (e.g. 501–504, 704) completely abolished the binding affinity. The affinity of ACG for gangliosides was attributed to the presence of the Galβ(1,3)GalNAc unit (i.e. T-antigen) (24). This study showed that wtACG-bound glycans with the GalNAcα(1,3)Gal (719, 23 µM; 720, 12 µM) or GlcNAcβ(1,3)Gal unit (724, 53 µM; 725, 18 µM; 728, 48 µM; 732, 19 µM; 733, 16 µM; 734, 15 µM; 902, 43 µM; 903, 26 µM; 905, 15 µM), indicating that GalNAcα(1,3)Gal and GlcNAcβ(1,3)Gal are the additional recognition units of ACG. Thus, wtACG has a broad substrate specificity in the recognition of the galactose-containing disaccharide units. In fact, it is now evident that galectins recognize diverse disaccharides defined by the Galβ-(syn)-gauche configuration (25). Ban et al. (15) determined the X-ray structures of the ACG/lactose and ACG/3′-sialyl lactose complexes. They revealed that ACG is composed of a β-sandwich of two anti-parallel six-stranded sheets with a shallow concavity. ACG has a conserved carbohydrate recognition domain (CRD), which shares a β-galactoside recognition pattern with highly conserved side-chains, and was categorized as prototype galectins, like mushroom Coprinus cinereus galectin II (Cgl-II) (26), human galectin 1 (hGal-1) (27) and 7 (hGal-7) (28), chicken galectin 16 (CG-16) (29) and conger eel congerin I (ConI) (30). Interestingly, ACG has two unique residues (Ser44 and Asn46) inserted in the CRD domain, which serve as the recognition of α(2,3)-linked sialic acid (15). In the ACG/3′-sialyl lactose complex, O4 of the sialic acid residue forms a H-bond with the main-chain O atom of Ser44, O8 forms a H-bond with NE1 of Trp83 and O9 forms a H-bond with NH2 of Arg77, while O3 of the glucose residue forms H-bonds with Arg66 (NH2) and Glu86 (OE1). The Arg residue in the S4 strand and the Glu residue in the S6 strand were conserved in other prototype galectins (15), suggesting that these two amino acids are important in forming H bonds with O3 of the glucose. Arg66 (NH1) also forms a H-bond with the galactose residue O4 and O5, and Glu86 (OE2) and Asn75 (ND2) with galactose residue O6 (Table II and Fig. 4). Here, we performed saturation mutagenesis of Glu86 in the S6 strand to decrease interactions with the glucose moiety of 3′-sialyl lactose, in the hope that Asn75 can keep the molecular architecture of CRD with the NeuAcα(2,3)Gal unit. In fact, many galectins are known to preserve the significant affinity for lactosamine-type disaccharides via the recognition of galactose O4 and O6 (31–33). Fig. 4 View largeDownload slide X-ray structures of the ACG/lactose and ACG/3′-sialyl lactose complex highlighting the glycan binding site. The 3D structures of the ACG/lactose complex (A; PDB 1WW6) and the ACG/3′-sialyl lactose (B; PDB 1WW4) were drawn with the Cn3D program (38). Oxygen and carbon atoms of disaccharides are indicated in red and grey, respectively. The numbers of oxygen atoms indicate glucose O3, galactose O4, galactose O5 and galactose O6 in (A) and glucose O3, NeuAc O4, NeuAc O8 and NeuAc O9 in (B). Fig. 4 View largeDownload slide X-ray structures of the ACG/lactose and ACG/3′-sialyl lactose complex highlighting the glycan binding site. The 3D structures of the ACG/lactose complex (A; PDB 1WW6) and the ACG/3′-sialyl lactose (B; PDB 1WW4) were drawn with the Cn3D program (38). Oxygen and carbon atoms of disaccharides are indicated in red and grey, respectively. The numbers of oxygen atoms indicate glucose O3, galactose O4, galactose O5 and galactose O6 in (A) and glucose O3, NeuAc O4, NeuAc O8 and NeuAc O9 in (B). Table II. Hydrogen bonds in the ACG/lactose (PDB 1WW6) and ACG/3′-sialyl lactose (PDB 1WW4) complexes. Ligand  Protein atom  Ligand atom  Lactose        Asn46 OD1  Gal O3 and O4    His62 NE2  Gal O4    Arg66 NH1  Gal O4 and O5    NH1  Glc O3    NH2  Glc O3    Asn75 ND2  Gal O6    Glu86 OE1  Gal O6, Glc O3    OE2  Glc O3  3′-Sialyl lactose        Ser44 O  Sia O4    Asn46 OD1  Gal O3 and O4    His62 NE2  Gal O4    Arg66 NH1  Gal O4 and O5    NH2  Glc O3    Asn75 ND2  Gal O6    Arg77 NH2  Sia O9    Trp83 NE1  Sia O8    Glu86 OE1  Glc O3    OE2  Gal O6  Ligand  Protein atom  Ligand atom  Lactose        Asn46 OD1  Gal O3 and O4    His62 NE2  Gal O4    Arg66 NH1  Gal O4 and O5    NH1  Glc O3    NH2  Glc O3    Asn75 ND2  Gal O6    Glu86 OE1  Gal O6, Glc O3    OE2  Glc O3  3′-Sialyl lactose        Ser44 O  Sia O4    Asn46 OD1  Gal O3 and O4    His62 NE2  Gal O4    Arg66 NH1  Gal O4 and O5    NH2  Glc O3    Asn75 ND2  Gal O6    Arg77 NH2  Sia O9    Trp83 NE1  Sia O8    Glu86 OE1  Glc O3    OE2  Gal O6  View Large In the ACG/lactose complex, Glu86 (OE1) forms a H-bond to glucose O3 of lactose (24). The substitution of Glu86 by Asp lacking one methylene carbon resulted in a dramatic decrease in the binding affinity for the asialo complex-type N-glycans, gangliosides and other glycans. Our results suggest that Galβ(1,4)GlcNAc in N-glycans and T-antigen seems not to make a H-bond with OD1 or OD2 of Asp86 via their reducing end sugar. It should be noted that E86Q exhibits a higher affinity than other mutants for 3′-sialyl T-antigen and T-antigen (Fig. 2). This indicates that OE1 or NE2 of Gln86 can form a H-bond with GalNAc of T-antigen in place of Glu86 OE1 in wtACG. The binding affinity of wtACG, E86D and MAL for N-glycans involving non-reducing bi-terminal α(2,3)-linked NeuAc (510, 602) was in the order of WT > E86D > MAL, whereas, that for 505, 507–509 was WT > MAL > E86D (Table I). In this regard, an obvious enhancing effect by α2-3 sialylation should be noted: affinity enhancement in E86D towards disialo forms relative to monosialo forms was as high as six times, when relating biantennary glycans, 507, 508 (monosialo-) and 510 (disialo-) were compared, whereas that in wtACG was merely additive (1.8 times, Table I). On the other hand, the effect of disialylation in the biantennary glycans was rather suppressive in MAL (1.4 times), though that exhibited high affinity enhancement by mono α2-3 sialylation, when asialo-form glycan (307) and monosialo-form glycans (507,508) were compared (23 times, Table I). Thus, the E86D mutant lectin could be regarded as ‘novel’ in terms of affinity-enhancing performance towards multiple sialylation. For this proof, however, further study including FAC analysis using triantennary and tetraantennary silaoglycans is necessary. Previous random mutagenesis approaches such as the overlap-extension method (34), error-prone PCR (35) and phage display system (36), and error-prone PCR and ribosome display system (37) yielded sialic acid-binding lectins (MAH, earthworm 29-kDa galactose-binding lectin) with the novel substrate binding specificity. These gene-engineered lectins are used to distinguish the cell lineage or the glycoform of proteins by lectin profiling, but their sugar specificities are not characterized in details (34–36). Here, we carried out the saturation mutagenesis of ACG Glu86 in the galactose-binding site and characterized the sugar-binding specificity of mutants by FAC with a variety of glycans. We succeeded in the identification of E86D with a narrow glycan-binding specificity. We hope that the developed mutant E86D will be of practical use in various fields relevant to cell biology and glycotechnology. Funding This work was supported in part by Kanto Bureau of Economy, Trade and Industry in Japan. Conflict of interest None declared. Acknowledgements We are indebted to Prof. T. Irimura (University of Tokyo) for his excellent support and thank Dr T. Mogi (MBL, University of Tokyo) for his critical reading of the manuscript. We also thank Prof. K. Yamamoto (University of Tokyo) for providing BIAcore 3000 for SPR analysis. Abbreviations Abbreviations ACG Agrocybe cylindracea galectin CRD carbohydrate recognition domain ECA Erythrina cristagalii agglutinin FAC frontal affinity chromatography LacNAc Galβ (1,4)GlcNAcβ MAH Maackia amurensis haemagglutinin MAL Maackia amurensis leucoagglutinin NeuAc N-acetylneuraminic acid PA pyridylaminated PSA prostate-specific antigen RU response unit 3′-sialyl galactose Neu5Acα(2,3)Gal 6′-sialyl glactose Neu5Acα(2,6)Galβ 3′-sialyl LacNAc Neu5Acα(2,3)Galβ(1,4)GlcNAcβ 3′-sialyl lactose Neu5Acα(2,3)Galβ(1,4)Glc 3’-sialyl T Neu5Acα(2,3)Galβ(1,3)GalNAcα SNA Sambucus nigra agglutinin SPR surface plasmon resonance T-antigen Galβ(1,3)GalNAcα References 1 Mandal C,  Mandal C.  Sialic acid binding lectins,  Experientia ,  1990, vol.  46 (pg.  433- 441) Google Scholar CrossRef Search ADS PubMed  2 Zeng FY,  Gabius HJ.  Sialic acid-binding proteins: characterization, biological function and application,  Z. Naturforsch. ,  1992, vol.  47c (pg.  641- 653) 3 Lehmann F,  Tiralongo E,  Tiralongo J.  Sialic acid-specific lectins: occurrence, specificity and function,  Cell Mol. Life Sci. ,  2006, vol.  63 (pg.  1331- 1354) Google Scholar CrossRef Search ADS PubMed  4 Michele DE,  Campbell KP.  Dystrophin-glycoprotein complex: post-translational processing and dystroglycan function,  J. Biol. Chem. ,  2003, vol.  278 (pg.  15457- 15460) Google Scholar CrossRef Search ADS PubMed  5 Tabares G,  Radcliffe CM,  Barrabes S,  Ramirez M,  Aleixandre RN,  Hoesel W,  Dwek RA,  Rudd PM,  Peracaula R,  de Llorens R.  Different glycan structures in prostate-specific antigen from prostate cancer sera in relation to seminal plasma PSA,  Glycobiology ,  2006, vol.  16 (pg.  132- 145) Google Scholar CrossRef Search ADS PubMed  6 Shah MH,  Telang SD,  Shah PM,  Patel PS.  Tissue and serum α2-3- and α2-6-linkage specific sialylation changes in oral carcinogenesis,  Glycoconj. J. ,  2008, vol.  25 (pg.  279- 290) Google Scholar CrossRef Search ADS PubMed  7 Wang FL,  Cui SX,  Sun LP,  Qu XJ,  Xie YY,  Zhou L,  Mu YL,  Tang W,  Wang YS.  High expression of a α2,3-linked sialic acid residues is associated with the metastatic potential of human gastric cancer,  Cancer Detect. Prev. ,  2009, vol.  32 (pg.  437- 443) Google Scholar CrossRef Search ADS PubMed  8 Cui H,  Lin Y,  Yue L,  Zhao X,  Liu J.  Differential expression of the α2,3-sialic acid residues in breast cancer is associated with metastatic potential,  Oncol. Rep. ,  2011, vol.  25 (pg.  1365- 1371) Google Scholar PubMed  9 Janković MM,  Kosanović MM.  Glycosylation of urinary prostate-specific antigen in benign hyperplasia and cancer: assessment by lectin-binding patterns,  Clin. Biochem. ,  2005, vol.  38 (pg.  58- 65) Google Scholar CrossRef Search ADS PubMed  10 Meany DL,  Zhang Z,  Sokoll LJ,  Zhang H,  Chan DW.  Glycoproteomics for prostate cancer detection: changes in serum PSA glycosylation patterns,  J. Proteome Res. ,  2009, vol.  8 (pg.  613- 619) Google Scholar CrossRef Search ADS PubMed  11 Ohyama C,  Hosono M,  Nitta K,  Oh-eda M,  Yoshikawa K,  Habuchi T,  Arai Y,  Fukuda M.  Carbohydrate structure and differential binding of prostate specific antigen to Maackia amurensis lectin between prostate cancer and benign prostate hypertrophy,  Glycobiology ,  2004, vol.  14 (pg.  671- 679) Google Scholar CrossRef Search ADS PubMed  12 Knibbs RN,  Goldstein IJ,  Ratcliffe RM,  Shibuya N.  Characterization of the carbohydrate binding specificity of the leukoagglutinating lectin from Maackia amurensis,  J. Biol. Chem. ,  1991, vol.  266 (pg.  83- 88) Google Scholar PubMed  13 Konami Y,  Yamamoto K,  Osawa T,  Irimura T.  Strong affinity of Maackia amurensis hemagglutinin (MAH) for sialic acid-containing Ser/Thr-linked carbohydrate chains of N-terminal octapeptides from human glycophorin A,  FEBS Lett. ,  1994, vol.  342 (pg.  334- 338) Google Scholar CrossRef Search ADS PubMed  14 Yagi F,  Miyamoto M,  Abe T,  Minami Y,  Tadera K,  Goldstein IJ.  Purification and carbohydrate-binding specificity of Agrocybe cylindracea lectin,  Glycoconj. J. ,  1997, vol.  14 (pg.  281- 288) Google Scholar CrossRef Search ADS PubMed  15 Ban M,  Yoon HJ,  Demirkan E,  Utsumi S,  Mikami B,  Yagi F.  Structural basis of a fungal galectin from Agrocybe cylindracea for recognizing sialoconjugate,  J. Mol. Biol. ,  2005, vol.  351 (pg.  695- 706) Google Scholar CrossRef Search ADS PubMed  16 Kasai K,  Oda Y,  Nishikawa M,  Ishii S.  Frontal affinity chromatography: theory, for its application to studies on specific interaction of biomolecules,  J. Chromatogr. ,  1986, vol.  376 (pg.  33- 47) Google Scholar CrossRef Search ADS PubMed  17 Yagi F,  Hiroyama H,  Kodama S.  Agrocybe cylindracea lectin is a member of the galectin family,  Glycoconj. J. ,  2001, vol.  18 (pg.  745- 749) Google Scholar CrossRef Search ADS PubMed  18 Nakamura S,  Yagi F,  Otani K,  Ito Y,  Hirabayashi J.  Comparative analysis of carbohydrate-binding properties of two tandem repeat-type Jacalin-related lectins, Castanea crenata agglutinin and Cycas revoluta leaf lectin,  FEBS J. ,  2005, vol.  272 (pg.  2784- 2799) Google Scholar CrossRef Search ADS PubMed  19 Tateno H,  Nakamura-Tsuruta S,  Hirabayashi J.  Frontal affinity chromatography: sugar–protein interactions,  Nat. Protoc. ,  2007, vol.  2 (pg.  2529- 2799) Google Scholar CrossRef Search ADS PubMed  20 Lis H,  Joubert FJ,  Sharon N.  Isolation and properties of N-acetyllactosamine-specific lectins from nine erythrina species,  Phytochemistry ,  1985, vol.  24 (pg.  2803- 2809) Google Scholar CrossRef Search ADS   21 Shibuya N,  Goldstein IJ,  Goldstein WF,  Nsimba-Lubaki M,  Peeters B,  Peumans WJ.  The elderberry (Sambucus nigra L.) bark lectin recognizes the Neu5Ac(α2-6)Gal/GalNAc sequence,  J. Biol. Chem. ,  1987, vol.  262 (pg.  1596- 1601) Google Scholar PubMed  22 Nakamura-Tsuruta S,  Kominami J,  Kamei M,  Koyama Y,  Suzuki T,  Isemura M,  Hirabayashi J.  Comparative analysis by frontal affinity chromatography of oligosaccharide specificity of GlcNAc-binding lectins, Griffonia simplicifolia lectin-II (GSL-II) and Boletopsis leucomelas lectin (BLL),  J. Biochem. ,  2006, vol.  140 (pg.  285- 291) Google Scholar CrossRef Search ADS PubMed  23 Yabe R,  Tateno H,  Hirabayashi J.  Frontal affinity chromatography analysis of constructs of DC-DIGNR and LSECtin extend evidence for affinity to agalactosylated N-glycans,  FEBS J. ,  2010, vol.  277 (pg.  4010- 4026) Google Scholar CrossRef Search ADS PubMed  24 Hirabayashi J,  Hashidate T,  Arata Y,  Nishi N,  Nakamura T,  Hirashima M,  Urashima T,  Oka T,  Futai M,  Muller WE,  Yagi F,  Kasai K.  Oligosaccharide specificity of galectins: a search by frontal affinity chromatography,  Biochim. Biophys. Acta. ,  2002, vol.  1572 (pg.  232- 254) Google Scholar CrossRef Search ADS PubMed  25 Iwaki J,  Tateno H,  Nishi N,  Minamisawa T,  Nakamura-Tsuruta S,  Itakura Y,  Kominami J,  Urashima T,  Nakamura T,  Hirabayashi J.  The Galβ-(syn)-gauche configuration is required for galectin-recognition disaccharides,  Biochim. Biophys. Acta. ,  2011, vol.  1810 (pg.  643- 651) Google Scholar CrossRef Search ADS PubMed  26 Walser PJ,  Haebel PW,  Kunzler M,  Sargent D,  Kues U,  Aebi M,  Ban N.  Structure and functional analysis of the fungal galectin CGL2,  Structure ,  2004, vol.  12 (pg.  689- 702) Google Scholar CrossRef Search ADS PubMed  27 López-Lucendo MF,  Solís D,  André S,  Hirabayashi J,  Kasai K,  Kaltner H,  Gabius HJ,  Romero A.  Growth-regulatory human galectin-1: crystallographic characterisation of the structural changes induced by single-site mutations and their impact on the thermodynamics of ligand binding,  J. Mol. Biol. ,  2004, vol.  343 (pg.  957- 970) Google Scholar CrossRef Search ADS PubMed  28 Leonidas DD,  Vatzaki EH,  Vorum H,  Celis JE,  Madsen P,  Acharya KR.  Structural basis for the recognition of carbohydrates by human galectin-7,  Biochemistry ,  1998, vol.  37 (pg.  13930- 13940) Google Scholar CrossRef Search ADS PubMed  29 Varela PF,  Solís D,  Díaz-Mauriño T,  Kaltner H,  Gabius HJ,  Romero A.  The 2.15 Å crystal structure of CG-16, the developmentally regulated homodimeric chicken galectin,  J. Mol. Biol. ,  1999, vol.  294 (pg.  537- 549) Google Scholar CrossRef Search ADS PubMed  30 Shirai T,  Mitsuyama C,  Niwa Y,  Matsui Y,  Hotta H,  Yamane T,  Kamiya H,  Ishii C,  Ogawa T,  Muramoto K.  High-resolution structure of the conger eel galectin, congerin I, in lactose liganded and ligand-free forms: emergence of a new structure class by accelerated evolution,  Struct. Fold. Des. ,  1999, vol.  7 (pg.  1223- 1233) Google Scholar CrossRef Search ADS   31 Leffler H,  Barondes SH.  Specificity of binding of three soluble rat lung lectins to substituted and unsubstituted mammalian β-galactosides,  J. Biol. Chem. ,  1986, vol.  261 (pg.  10119- 10126) Google Scholar PubMed  32 Ahmed H,  Pohl J,  Fink NE,  Strobel F,  Vasta GR.  The primary structure and carbohydrate specificity of a β-galactosyl-binding lectin from toad (Bufo arenarum Hensel) ovary reveal closer similarities to the mammalian galectin-1 than to the galectin from the clawed frog Xenopus laevis,  J. Biol. Chem. ,  1996, vol.  271 (pg.  33083- 33094) Google Scholar CrossRef Search ADS PubMed  33 Walser PJ,  Haebel PW,  Künzler M,  Sargent D,  Kües U,  Aebi M,  Ban N.  Structure and functional analysis of the fungal galectin CGL2,  Structure ,  2004, vol.  12 (pg.  689- 702) Google Scholar CrossRef Search ADS PubMed  34 Yim M,  Ono T,  Irimura T.  Mutated plant lectin library useful to identify different cells,  Proc. Natl. Acad. Sci. USA. ,  2001, vol.  98 (pg.  2222- 2225) Google Scholar CrossRef Search ADS   35 Maenuma K,  Yim M,  Komatsu K,  Hoshino M,  Takahashi Y,  Bovin N,  Irimura T.  Use of a library of mutated Maackia amurensis hemagglutinin for profiling the cell lineage and differentiation,  Proteomics ,  2008, vol.  8 (pg.  3274- 3283) Google Scholar CrossRef Search ADS PubMed  36 Maenuma K,  Yim M,  Komatsu K,  Hoshino M,  Tachiki-Fujioka A,  Takahashi K,  Hiki Y,  Bovin N,  Irimura T.  A library of mutated Maackia amurensis hemagglutinin distinguishes putative glycoforms of immunoglobulin A1 from IgA nephropathy patients,  J. Proteome Res. ,  2009, vol.  8 (pg.  3617- 3624) Google Scholar CrossRef Search ADS PubMed  37 Yabe R,  Suzuki R,  Kuno A,  Fujimoto Z,  Jigami Y,  Hirabayashi J.  Tailoring a novel sialic acid-binding lectin from a ricin-B chain-like galactose-binding protein by natural evolution-mimicry,  J. Biochem. ,  2007, vol.  141 (pg.  389- 399) Google Scholar CrossRef Search ADS PubMed  38 Wang Y,  Geer L,  Chappey C,  Kans J,  Bryant S.  Cn3D: sequence and structure views for Entrez,  Trends Biochem. Sci. ,  2000, vol.  25 (pg.  300- 302) Google Scholar CrossRef Search ADS PubMed  © The Authors 2011. Published by Oxford University Press on behalf of the Japanese Biochemical Society. All rights reserved

Journal

The Journal of BiochemistryOxford University Press

Published: Aug 3, 2011

Keywords: Agrocybe cylindracea galectin frontal affinity chromatography hydrogen bond saturation mutagenesis sialo-binding lectin

There are no references for this article.