Molecular cloning and heterologous expression of beta1,2-xylosyltransferase and core alpha1,3-fucosyltransferase from maize

Maize is considered a promising alternative production system for pharmaceutically relevant proteins. However, like in all other plant species asparagine-linked oligosaccharides of maize glycoproteins are modified with beta1,2-xylose and core alpha1,3-fucose sugar residues, which are considered to be immunogenic in mammals. This altered N-glycosylation when compared to mammalian cells may reduce the potential of maize as a production system for heterologous glycoproteins. Here we report the cloning and characterization of the cDNA sequences coding for the maize enzymes beta1,2-xylosyltransferase (XylT) and core alpha1,3-fucosyltransferase (FucT). The cloned XylT and FucT cDNAs were shown to encode enzymatically active proteins, which were independently able to convert a mammalian acceptor glycoprotein into an antigen binding anti-plant N-glycan antibodies. The complete sequence of the XylT gene was determined. Evidence for the presence of at least three XylT and FucT gene loci in the maize genome was obtained. The identification of the two enzymes and their genes will allow the targeted downregulation or even elimination of beta1,2-xylose and core alpha1,3-fucose addition to recombinant glycoproteins produced in maize.     

Molecular cloning and characterization of the Caenorhabditis elegans alpha1,3-fucosyltransferase family

The genome of Caenorhabditis elegans encodes five genes with homology to known alpha1,3 fucosyltransferases (alpha1,3FTs), but their expression and functions are poorly understood. Here we report the molecular cloning and characterization of these C. elegans alpha1,3FTs (CEFT-1 through -5). The open-reading frame for each enzyme predicts a type II transmembrane protein and multiple potential N-glycosylation sites. We prepared recombinant epitope-tagged forms of each CEFT and found that they had unusual acceptor specificity, cation requirements, and temperature sensitivity. CEFT-1 acted on the N-glycan pentasaccharide core acceptor to generate Manalpha1-3(Manalpha1-6)Manbeta1-4GlcNAcbeta1-4(Fucalpha1-3)GlcNAcbeta1-Asn. In contrast, CEFT-2 did not act on the pentasaccharide acceptor, but instead utilized a LacdiNAc acceptor to generate GalNAcbeta1-4(Fucalpha1-3)GlcNAcbeta1-3Galbeta1-4Glc, which is a novel activity. CEFT-3 utilized a LacNAc acceptor to generate Galbeta1-4(Fucalpha1-3)GlcNAcbeta1-3Galbeta1-4Glc without requiring cations. CEFT-4 was similar to CEFT-3, but its activity was enhanced by some divalent cations. Recombinant CEFT-5 was well expressed, but did not act on available acceptors. Each CEFT was optimally active at room temperature and rapidly lost activity at 37 degrees C. Promoter analysis showed that CEFT-1 is expressed in C. elegans eggs and adults, but its expression was restricted to a few neuronal cells at the head and tail. We prepared deletion mutants for each enzyme for phenotypic analysis. While loss of CEFT-1 correlated with loss of pentasaccharide core activity and core alpha1,3-fucosylated glycans in worms, loss of other enzymes did not correlate with any phenotypic changes. These results suggest that each of the alpha1,3FTs in C. elegans has unique specificity and expression patterns.     

Purification, cDNA cloning and homology modeling of endo-1,3-beta-D-glucanase from scallop Mizuhopecten yessoensis

The retaining endo-1,3-β-d-glucanase (LV) with molecular mass of 36 kDa was purified to homogeneity from the crystalline styles of scallop Mizuhopecten yessoensis. The purified enzyme catalyzed hydrolysis of laminaran as endo-enzyme forming glucose, laminaribiose and higher oligosaccharides as products (K_m  600 μg/mL). The 1,3-β-d-glucanase effectively catalyzed transglycosylation reaction that is typical of endo-enzymes too. Optima of pH and temperature were at 4.5 and 45 °C, respectively. cDNA encoding the endo-1,3-β-d-glucanase was cloned by PCR-based methods. It contained an open reading frame that encoded 339-amino acids protein. The predicted endo-1,3-β-d-glucanase amino acid sequence included a characteristic domain of the glycosyl hydrolases family 16 and revealed closest homology with 1,3-β-d-glucanases from bivalve Pseudocardium sachalinensis, sea urchin Strongylocentrotus purpuratus and invertebrates lipopolysaccharide and β-1,3-glucan-binding proteins.The fold of the LV was more closely related to κ-carrageenase, agarase and 1,3;1,4-β-d-glucanase from glycosyl hydrolases family 16. Homology model of the endo-1,3-β-d-glucanase from M. yessoensis was obtained with MOE on the base of the crystal structure of κ-carrageenase from P. carrageonovora as template. Putative three-dimensional structures of the LV complexes with substrate laminarihexaose or glucanase inhibitor halistanol sulfate showed that the binding sites of the halistanol sulfate and laminarihexaose are located in the enzyme catalytic site and overlapped.     

Purification, characterization, cDNA cloning and expression of a novel ketoreductase from Zygosaccharomyces rouxii.

A novel ketoreductase isolated from Zygosaccharomyces rouxii catalyzes the asymmetric reduction of selected ketone substrates of commercial importance. The 37.8-kDa ketoreductase was purified more than 300-fold to > 95% homogeneity from whole cells with a 30% activity yield. The ketoreductase functions as a monomer with an apparent Km for 3,4-methylenedioxyphenyl acetone of 2.9 mM and a Km for NADPH of 23.5 microM. The enzyme is able to effectively reduce alpha-ketolactones, alpha-ketolactams, and diketones. Inhibition is observed in the presence of diethyl pyrocarbonate, suggesting that a histidine is crucial for catalysis. The 1.0-kb ketoreductase gene was cloned and sequenced from a Z. rouxii cDNA library using a degenerate primer to the N-terminal sequence of the purified protein. Furthermore, it was expressed in both Escherichia coli and Pichia pastoris and shown to be active. Substrate specificity, lack of a catalytic metal, and extent of protein sequence identity to known reductases suggests that the enzyme falls into the carbonyl reductase enzyme class.     

High level expression of monomeric and dimeric human alpha1,3-fucosyltransferase V

alpha3/4-Fucosyltransferases play a crucial role in inflammatory processes and tumor metastasis. While several human fucosyltransferases (FucTs) with different acceptor substrate specificities have been identified, the design of specific inhibitors for therapeutic approaches is hampered by the lack of structural information. In this study, we evaluated the expression of different constructs of human fucosyltransferase V to generate the large amounts required for structural studies. The truncated constructs lacking the transmembrane region and the cytosolic N-terminus, were expressed in baculovirus-infected Trichoplusia ni (Tn) insect cells and in two non-lytic expression systems, stably transfected human HEK 293 and T. ni cells. Since secretion of some glycosyltransferases is controlled by formation of dimeric molecules via disulfide bonds, one of the fucosyltransferase V constructs contained the N-terminal cysteine residue 64 for dimerization, whereas this residue was replaced in the other construct by serine. In both human and insect cells dimerization did not prove to be essential for efficient expression and secretion. On the basis of enzymatic activity, the yield of secreted fucosyltransferase V was approximately 10-fold higher in stably transfected insect cells than in HEK 293 cells. In particular the monomeric form of the enzyme provides a valuable tool for structural analyses to elucidate the fine specifity of fucosyltransferase V-mediated fucosylation of Lewis type glycans.     

Purification, cDNA cloning, and recombinant expression of chymotrypsin C from porcine pancreas

Chymotrypsin C is a bifunctional secretory-type serine protease in pancreas; besides proteolytical activity, it also exhibits a calcium-decreasing activity in serum. In this study, we purified activated chymotrypsin C from porcine pancreas, and identified its three active forms. Active chymotrypsin C was found to be different in the length of its 13-residue activation peptide due to carboxydipeptidase (present in the pancreas) degradation or autolysis of the activated chymotrypsin C itself, resulting in the removal of several C-terminus residues from the activation peptide. After limited chymotrypsin C cleavage with endopeptidase Lys C, several purified peptides were partially sequenced, and the entire cDNA sequence for porcine chymotrypsin C was cloned. Recombinant chymotrypsinogen C was successfully expressed in Escherichia coli cells as inclusion bodies. After refolding and activation with trypsin, the comparison of the recombinant chymotrypsin C with the natural form showed that their proteolytic and calcium-decreasing activities were at the same level. The successful expression of chymotrypsin C gene paves the way to further mutagenic structure-function studies.     

Purification, cDNA cloning, and expression of UDP-N-acetylglucosamine: beta-D-mannoside beta-1,4N-acetylglucosaminyltransferase III from rat kidney.

UDP-N-acetylglucosamine: beta-D-mannoside beta-1,4N-acetylglucosaminyltransferase III (GnT-III: EC 2.4.1.144) catalyzes the addition of N-acetylglucosamine in beta 1-4 linkage to the beta-linked mannose of the trimannosyl core of N-linked sugar chains. The enzyme has been purified over 153,000-fold in 1.5% yield from a Triton X-100 extract of rat kidney by fractionation procedures utilizing QAE-Sepharose, Cu(2+)-chelating Sepharose, and affinity chromatography on UDP-hexanolamine and substrate-conjugated Sepharose. The purified protein migrates as one major and one minor band with apparent molecular masses of 62 kDa and 52 kDa, respectively. The purified enzyme was digested with trypsin, and the amino acid sequences of four peptides were determined. Oligonucleotide primers were designed according to those amino acid sequences and used in the polymerase chain reaction. Screening for the cDNA for GnT-III was carried out by plaque hybridization using a rat kidney cDNA library (lambda gt10) and a polymerase chain reaction product as the probe. Rat kidney GnT-III has 536 amino acids and three putative N-glycosylation sites. There is no sequence homology to other previously cloned glycosyltransferases, but the enzyme appears to be a type II transmembrane protein like the other glycosyltransferases. The GnT-III activity in transiently transfected COS-1 cells was found to be about 500-3600-fold as compared to that in non- or mock-transfected cells.     

Purification, characterization, and cDNA cloning of profilin from Phaseolus vulgaris.

Profilin from common bean (Phaseolus vulgaris L.) was purified to homogeneity by poly-L-Pro affinity chromatography and gel filtration. The hypocotyl and symbiotic root nodule protein was detected as a single isoform with a 14.4-kD molecular mass and an isoelectric point of 5.3. Partial amino acid and DNA sequencing of a full-length cDNA clone confirmed its identity as profilin. An antibody generated against the purified protein binds to a protein with the same molecular mass in leaves and nodules. Immunolocalization of the protein showed a diffuse distribution in the cytoplasm of hypocotyls and nodules but enhanced staining at the vascular bundles. The strong identity of the sequence among the profilins of birch, maize, and bean suggests that it may play an important role in the signal transduction mechanism of plant cells and plant-bacterial symbioses.     

Purification of the beta-N-acetylglucosaminide alpha 1----3-fucosyltransferase from human serum.

A beta-N-Acetylglucosaminide alpha 1----3-fucosyltransferase was purified from human serum by ammonium sulfate precipitation, hydrophobic chromatography on phenyl-Sepharose, ion-exchange chromatography on sulfopropyl-Sepharose, affinity chromatography on GDP-hexanolamine-Sepharose, and finally high pressure liquid chromatography gel filtration. Gel filtration chromatography of the native enzyme revealed a Mr of 45,000. Upon sodium dodecyl sulfate-polyacrylamide gel electrophoresis, the purified protein also appeared as a single molecular species of Mr 45,000. In contrast to the multisubunit beta-galactoside alpha 1----2-fucosyltransferases with an apparent Mr of 150,000, present in human serum, the native beta-N-acetylglucosaminide alpha 1----3-fucosyltransferase is a monomer with a Mr of 45,000. The enzyme is glycosylated, as revealed by wheat germ agglutinin binding properties. The alpha 1----3 linkage formed by the enzyme between alpha-L-fucose and the penultimate beta-N-acetylglucosamine by the purified enzyme was confirmed by 1H NMR homonuclear cross-irradiation analysis of the oligosaccharide product. The specificity of the purified enzyme is restricted to type 2 structures, as revealed by its reactivity with different substrates and from the Km values calculated from the initial rate data using various oligosaccharide acceptors. The enzyme has the ability to utilize the N-acetyl-beta-lactosamine determinant (Gal beta 1----4GlcNAc) and the sialylated (NeuAc alpha 2----3Gal beta 1----4GlcNAc) and fucosylated (Fuc alpha 1----2Gal beta 1----4GlcNAc) derivatives of N-acetyl-beta-lactosamine and thus is distinct from both the human Lewis gene-encoded enzyme and the alpha 1----3-fucosyltransferase of the myeloid cell type.     

Novel poly-GalNAcbeta1-4GlcNAc (LacdiNAc) and fucosylated poly-LacdiNAc N-glycans from mammalian cells expressing beta1,4-N-acetylgalactosaminyltransferase and alpha1,3-fucosyltransferase

Glycans containing the GalNAcbeta1-4GlcNAc (LacdiNAc or LDN) motif are expressed by many invertebrates, but this motif also occurs in vertebrates and is found on several mammalian glycoprotein hormones. This motif contrasts with the more commonly occurring Galbeta1-4GlcNAc (LacNAc or LN) motif. To better understand LDN biosynthesis and regulation, we stably expressed the cDNA encoding the Caenorhabditis elegans beta1,4-N-acetylgalactosaminyltransferase (GalNAcT), which generates LDN in vitro, in Chinese hamster ovary (CHO) Lec8 cells, to establish L8-GalNAcT CHO cells. The glycan structures from these cells were determined by mass spectrometry and linkage analysis. The L8-GalNAcT cell line produces complex-type N-glycans quantitatively bearing LDN structures on their antennae. Unexpectedly, most of these complex-type N-glycans contain novel "poly-LDN" structures consisting of repeating LDN motifs (-3GalNAcbeta1-4GlcNAcbeta1-)n. These novel structures are in contrast to the well known poly-LN structures consisting of repeating LN motifs (-3Galbeta1-4GlcNAcbeta1-)n. We also stably expressed human alpha1,3-fucosyltransferase IX in the L8-GalNAcT cells to establish a new cell line, L8-GalNAcT-FucT. These cells produce complex-type N-glycans with alpha1,3-fucosylated LDN (LDNF) GalNAcbeta1-4(Fucalpha1-3)GlcNAcbeta1-R as well as novel "poly-LDNF" structures (-3GalNAcbeta1-4(Fucalpha 1-3)GlcNAcbeta1-)n. The ability of these cell lines to generate glycoprotein hormones with LDN-containing N-glycans was studied by expressing a recombinant form of the common alpha-subunit in L8-GalNAcT cells. The alpha-subunit N-glycans carried LDN structures, which were further modified by co-expression of the human GalNAc 4-sulfotransferase I, which generates SO4-4GalNAcbeta1-4GlcNAc-R. Thus, the generation of these stable mammalian cells will facilitate future studies on the biological activities and properties of LDN-related structures in glycoproteins.     

Molecular cloning and expression of Galbeta1,3GalNAc alpha2, 3-sialyltransferase from human fetal liver.

Based on the sequences of the highly conserved segments in the previously cloned sialyltransferases, a cDNA encoding Galbeta1, 3GalNAc alpha2,3-sialyltransferase (SIATFL) has been isolated from human fetal liver. Expression analysis of the gene has been performed with various carcinoma cell lines, fetal tissues, fetal and adult liver and both hepatoma and the surrounding tissue from the same liver. The SIATFL gene was expressed poorly in fetal liver and in adult liver, slightly in hepatoma and highly in the surrounding tissue of hepatoma. The cDNA encoding the putative active domain was expressed in COS-1, Escherichia coli, and Pichia pastoris. The recombinant protein expressed in COS-1 could catalyse the transfer of NeuAc from CMP-NeuAc to asialo-fetuin. No enzyme activity was detected with a 32-kDa protein in E. coli and both 32-kDa and 41-kDa proteins in P. pastoris. These results suggested that correct glycosylation of the enzyme might play a key role in its folding that may be directly related to the enzymatic activity.     

Rabbit small intestinal trehalase. Purification, cDNA cloning, expression, and verification of glycosylphosphatidylinositol anchoring.

alpha,alpha-Trehalase (EC 3.2.1.28), an intrinsic protein of intestinal brush-border membranes, was purified to homogeneity from rabbits. Partial amino acid sequences were determined. Two degenerate oligonucleotides based on the sequence of a CNBr peptide were employed in a polymerase chain reaction to amplify a 71-base pair fragment of trehalase DNA with rabbit intestine cDNA as a starting template. This fragment was used as a hybridization probe to isolate full length trehalase clones from a rabbit intestine cDNA bank. Sequence analysis revealed that trehalase comprises 578 amino acids, contains at the amino terminus a typical cleavable signal sequence, at the carboxyl terminus a rather hydrophobic region typical of proteins anchored via glycosylphosphatidylinositol, and four potential N-glycosylation sites. Trehalase has no sequence homologies with other sequenced brush-border glycosidases. Northern blot analysis revealed a 1.9-kilobase trehalase mRNA in small intestine and kidney, smaller amounts in liver, and none in lung. Southern blot analysis indicated the gene has a length of 20 kilobase pairs or less. Injection into Xenopus laevis oocytes of mRNA synthesized in vitro from a trehalase template resulted in the expression of trehalase activity several hundredfold above background. The trehalase activity was membrane-bound and could be solubilized upon digestion with phosphatidylinositol-specific phospholipase C from Bacillus thuringiensis. This strongly suggests that rabbit small intestinal trehalase is anchored via glycosylphosphatidylinositol also when expressed in X. laevis oocytes.     

Purification of the secretor-type beta-galactoside alpha 1----2-fucosyltransferase from human serum.

The secretor-type beta-galactoside alpha 1----2-fucosyltransferase from human serum was purified by hydrophobic chromatography on phenyl-Sepharose, ion-exchange chromatography on sulfopropyl-Sepharose, and affinity chromatography on GDP-hexanolamine-Sepharose. Final purification of the enzyme was achieved by high pressure liquid chromatography gel filtration and resulted in a homogeneous protein as determined by sodium dodecyl sulfate-polyacrylamide gel electrophoresis of the radiolabeled protein. The native enzyme appears as a molecule of apparent Mr 150,000 as determined by gel filtration high pressure liquid chromatography. The apparent Mr of the enzyme resolved in the presence of beta-mercaptoethanol by sodium dodecyl sulfate-polyacrylamide gel electrophoresis was determined to be 50,000, indicating a multisubunit structure of the enzyme. Secretor-type alpha 1----2-fucosyltransferase is a glycoprotein as determined by WGA binding properties. A comparison of the Mr of the native blood group H gene encoded with the secretor-type beta-galactoside alpha 1----2-fucosyltransferases as well as comparison of subunit Mr for both enzymes suggests structural similarity. The alpha 1----2 linkage formed between alpha-L-fucose and terminal beta-D-galactose by the purified H- and secretor-type alpha 1----2-fucosyltransferases was determined by 1H NMR homonuclear cross-irradiation analysis of the oligosaccharide products. The substrate specificity and Km values calculated from the initial rate using various oligosaccharide acceptors showed that purified enzymes differ primarily in affinity for phenyl-beta-D-galactopyranoside and GDP-fucose as well as type 1 (Gal beta 1----3GlcNAc), 2 (Gal beta 1----4GlcNAc), and 3 (Gal beta 1----3GalNAc) oligosaccharide acceptors. The secretor-type alpha 1----2-fucosyltransferase shows significantly lower affinity than the H enzyme for phenyl-beta-D-galactopyranoside and GDP-fucose as well as for type 2 oligosaccharide acceptors. On the contrary, type 1 and 3 oligosaccharide acceptors are preferentially utilized by the secretor-type enzyme as compared with the H enzyme. The enzymes also differ in several physicochemical properties, implying nonidentity of the two enzymes (Sarnesto, A., K?hlin, T., Thurin, J., and Blaszczyk-Thurin, M. (1990) J. Biol. Chem. 265, 15067-15075).     

Molecular cloning and expression of rat interleukin-1 alpha cDNA

A cDNA sequence coding for rat interleukin-1 alpha (IL-1 alpha) has been isolated from a cDNA library that was prepared with mRNA derived from LPS-stimulated rat peritoneal macrophages by using human IL-1 alpha cDNA as a probe. The rat cDNA encodes a 270 amino acid residue protein which is homologous (65%) to human IL-1 alpha. The rat cDNA sequence under SV40 early promoter directed the synthesis of biologically active IL-1 in monkey COS-1 cells. Rat IL-1 alpha mRNA is not expressed in spleen, lung, liver or brain, and is also not expressed in these organs of LPS-treated rat except spleen. This suggests that IL-1 alpha is not produced constitutively in various tissues and LPS is not sufficient to induce IL-1 alpha in most tissues. Our data indicate that the IL-1 activities which have been reported to be produced in the brain are not of alpha type. We have constructed a plasmid expressing the carboxy terminal 156 amino acids in Escherichia coli. Recombinant rat IL-1 alpha produced in COS cells or E. coli has cytotoxic activity against the human melanoma cell line A375S1 (GIF activity), which has been reported to be sensitive to human IL-1 alpha and IL-1 beta. This suggests that GIF activity is common to IL-1s derived from various sources.     

A lipopolysaccharide- and beta-1,3-glucan-binding protein from hemocytes of the freshwater crayfish Pacifastacus leniusculus. Purification, characterization, and cDNA cloning

A lipopolysaccharide- and beta-1,3-glucan-binding protein (LGBP) was isolated and characterized from blood cells (hemocytes) of the freshwater crayfish Pacifastacus leniusculus. The LGBP was purified by chromatography on Blue-Sepharose and phenyl-Sepharose, followed by Sephacryl S-200. The LGBP has a molecular mass of 36 kDa and 40 kDa on 10% SDS-polyacrylamide gel electrophoresis under reducing and nonreducing conditions, respectively. The calculated mass of LGBP is 39,492 Da, which corresponds to the native size of LGBP; the estimated pI of the mature LGBP is 5.80. LGBP has binding activity to lipopolysaccharides as well as to beta-1,3-glucans such as laminarin and curdlan, but peptidoglycan could not bind to LGBP. Cloning and sequencing of LGBP showed significant homology with several putative Gram-negative bacteria-binding proteins and beta-1, 3-glucanases. Interestingly, LGBP also has a structure and functions similar to those of the coelomic cytolytic factor-1, a lipopolysaccharide- and glucan-binding protein from the earthworm Eisenia foetida. To evaluate the involvement of LGBP in the prophenoloxidase (proPO) activating system, a polyclonal antibody against LGBP was made and used for the inhibition of phenoloxidase (PO) activity triggered by the beta-1,3-glucan laminarin in the hemocyte lysate of crayfish. The PO activity was blocked completely by the anti-LGBP antibody. Moreover, the PO activity could be recovered by the addition of purified LGBP. These results suggest that the 36-kDa LGBP plays a role in the activation of the proPO activating system in crayfish and thus seems to play an important role in the innate immune system of crayfish.     

Four glucosyltransferases from rice: cDNA cloning, expression, and characterization

Four UDP-dependent glucosyltransferase (UGT) genes, UGT706C1, UGT706D1, UGT707A3, and UGT709A4 were cloned from rice, expressed in Escherichia coli, and purified to homogeneity. In order to find out whether these enzymes could use flavonoids as glucose acceptors, apigenin, daidzein, genistein, kaempferol, luteolin, naringenin, and quercetin were used as potential glucose acceptors. UGT706C1 and UGT707A3 could use kaempferol and quercetin as glucose acceptors and the major glycosylation position was the hydroxyl group of carbon 3 based on the comparison of HPLC retention times, UV spectra, and NMR spectra with those of corresponding authentic flavonoid 3-O-glucosides. On the other hand, UGT709A4 only used the isoflavonoids genistein and daidzein and transferred glucose onto 7-hydroxyl group. In addition, UGT706D1 used a broad range of flavonoids including flavone, flavanone, flavonol, and isoflavone, and produced at least two products with glycosylation at different hydroxyl groups. Based on their substrate preferences and the flavonoids present in rice, the in vivo function of UGT706C1, UGT706D1, and UGT707A3 is most likely the biosynthesis of kaempferol and quercetin glucosides.     

Purification and cDNA cloning of Xenopus liver galectins and their expression

We have characterized galectin family proteins in adult tissues of Xenopus laevis and purified 14-kDa and 36-kDa proteins from the liver. The liver galectins showed comparable hemagglutination activities to those of mammalian galectins. Furthermore, we isolated five galectin cDNAs from a Xenopus liver library. These cDNAs revealed that X. laevis galectins (xgalectins) form a family consisting of at least proto and tandem repeat types based on their domain structures, like the mammalian galectin family. Two proto-type xgalectins, -Ia and -Ib, exhibited a high sequence identity (91%) with each other at the amino acid level and were most similar (49-50% identity) to human galectin-1. From their sequence similarity and ubiquitous tissue distributions, xgalectins-Ia and -Ib both seemed to be Xenopus homologues of mammalian galectin-1. Three tandem repeat-type xgalectins were newly identified. Two of them, xgalectins-IIa and -IIIa, seemed to be homologous to human galectins-4 and -9, respectively, judging from their high sequence similarities (42-50% identity). However, xgalectin-IVa seemed to be a novel type. Distributions of mRNAs of xgalectins were analyzed by northern hybridization. In addition to adult tissues, either of three tandem repeat-type xgalectins were expressed in whole embryos. Moreover, amino acid sequence analysis of liver proteins indicated that xgalectins-Ia, -IIa, and -IIIa are produced as abundant galectins in the adult liver.