Cloning and characterization of a novel mouse Siglec, mSiglec-F: differential evolution of the mouse and human (CD33) Siglec-3-related gene clusters.

A novel mouse Siglec (mSiglec-F) belonging to the subfamily of Siglec-3-related Siglecs has been cloned and characterized. Unlike most human Siglec-3 (hSiglec-3)-related Siglecs with promiscuous linkage specificity, mSiglec-F shows a strong preference for alpha2-3-linked sialic acids. It is predominantly expressed in immature cells of the myelomonocytic lineage and in a subset of CD11b (Mac-1)-positive cells in some tissues. As with previously cloned Siglec-3-related mSiglecs, the lack of strong sequence similarity to a singular hSiglec made identification of the human ortholog difficult. We therefore conducted a comprehensive comparison of Siglecs between the human and mouse genomes. The mouse genome contains eight Siglec genes, whereas the human genome contains 11 Siglec genes and a Siglec-like gene. Although a one-to-one orthologous correspondence between human and mouse Siglecs 1, 2, and 4 is confirmed, the Siglec-3-related Siglecs showed marked differences between human and mouse. We found only four Siglec genes and two pseudogenes in the mouse chromosome 7 region syntenic to the Siglec-3-related gene cluster on human chromosome 19, which, in contrast, contains seven Siglec genes, a Siglec-like gene, and thirteen pseudogenes. Although analysis of gene maps and exon structures allows tentative assignments of mouse-human Siglec ortholog pairs, the possibility of unequal genetic recombination makes the assignments inconclusive. We therefore support a temporary lettered nomenclature for additional mouse Siglecs. Current information suggests that mSiglec-F is likely a hSiglec-5 ortholog. The previously reported mSiglec-3/CD33 and mSiglec-E/MIS are likely orthologs of hSiglec-3 and hSiglec-9, respectively. The other Siglec-3-like gene in the cluster (mSiglec-G) is probably a hSiglec-10 ortholog. Another mouse gene (mSiglec-H), without an apparent human ortholog, lies outside of the cluster. Thus, although some duplications of Siglec-3-related genes predated separation of the primate and rodent lineages (about 80-100 million years ago), this gene cluster underwent extensive duplications in the primate lineage thereafter.

A second uniquely human mutation affecting sialic acid biology.

Siglecs are immunoglobulin superfamily member lectins that selectively recognize different types and linkages of sialic acids, which are major components of cell surface and secreted glycoconjugates. We report here a human Siglec-like molecule (Siglec-L1) that lacks a conserved arginine residue known to be essential for optimal sialic acid recognition by previously known Siglecs. Loss of the arginine from an ancestral molecule was caused by a single nucleotide substitution that occurred after the common ancestor of humans with the great apes but before the origin of modern humans. The chimpanzee Siglec-L1 ortholog remains fully functional and preferentially recognizes N-glycolylneuraminic acid, which is a common sialic acid in great apes and other mammals. Reintroducing the ancestral arginine into the human molecule regenerates the same properties. Thus, the single base pair mutation that replaced the arginine on human Siglec-L1 is likely to be evolutionarily related to the previously reported loss of N-glycolylneuraminic acid expression in the human lineage. Siglec-L1 and its chimpanzee Siglec ortholog also have a different expression pattern from previously reported Siglecs because they are found on the lumenal edge of epithelial cell surfaces. Notably, the human genome contains several Siglec-like pseudogenes that have independent mutations that would have replaced the arginine residue required for optimal sialic acid recognition. Thus, additional changes in the biology of sialic acids may have taken place during human evolution.

Cloning, characterization, and phylogenetic analysis of siglec-9, a new member of the CD33-related group of siglecs. Evidence for co-evolution with sialic acid synthesis pathways.

The Siglecs are a subfamily of I-type lectins (immunoglobulin superfamily proteins that bind sugars) that specifically recognize sialic acids. We report the cloning and characterization of human Siglec-9. The cDNA encodes a type 1 transmembrane protein with three extracellular immunoglobulin-like domains and a cytosolic tail containing two tyrosines, one within a typical immunoreceptor tyrosine-based inhibitory motif (ITIM). The N-terminal V-set Ig domain has most amino acid residues typical of Siglecs. Siglec-9 is expressed on granulocytes and monocytes. Expression of the full-length cDNA in COS cells induces sialic-acid dependent erythrocyte binding. A recombinant soluble form of the extracellular domain binds to alpha2-3 and alpha2-6-linked sialic acids. Typical of Siglecs, the carboxyl group and side chain of sialic acid are essential for recognition, and mutation of a critical arginine residue in domain 1 abrogates binding. The underlying glycan structure also affects binding, with Galbeta1-4Glc[NAc] being preferred. Siglec-9 shows closest homology to Siglec-7 and both belong to a Siglec-3/CD33-related subset of Siglecs (with Siglecs-5, -6, and -8). The Siglec-9 gene is on chromosome 19q13.3-13.4, in a cluster with all Siglec-3/CD33-related Siglec genes, suggesting their origin by gene duplications. A homology search of the Drosophila melanogaster and Caenorhabditis elegans genomes suggests that Siglec expression may be limited to animals of deuterostome lineage, coincident with the appearance of the genes of the sialic acid biosynthetic pathway.

Siglec-7: a sialic acid-binding lectin of the immunoglobulin superfamily.

The Siglecs are a recently discovered family of sialic acid-binding lectins of the immunoglobulin (Ig) superfamily. We report a molecule showing homology to the six first reported Siglecs, with the closest relationship to Siglec-3(CD33), Siglec-5, and Siglec-6(OBBP-1). The extracellular portion has two Ig-like domains, with the amino-terminal V-set Ig domain including amino acid residues known to be involved in sialic acid recognition by other Siglecs. The cytoplasmic domain has putative sites of tyrosine phosphorylation shared with some Siglecs, including an Immuno-receptor Tyrosine-based Inhibitory Motif (ITIM). Expression of the full-length cDNA induces sialic acid-dependent binding to human erythrocytes. A recombinant chimeric form containing the extracellular Ig domains selectively recognizes the sequence Neu5Acalpha2-6Galbeta1-4Glc, and binding requires the side chain of sialic acid. Mutation of an arginine residue predicted to be critical for sialic acid binding abolishes both interactions. Taken together, our findings justify designation of the molecule as Siglec-7. Analysis of bacterial artificial chromosome (BAC) clones spanning the known human genomic location of Siglec-3 indicates that the Siglec-7 gene is also located on chromosome 19q13.3-13.4. Human tissues show strong expression of Siglec-7 mRNA in spleen, peripheral blood leukocytes, and liver. The combination of an extracellular sialic acid binding site and an intracellular ITIM motif suggests that this molecule is involved in trans-membrane regulatory signaling reactions.

Biosynthesis of KDN (2-keto-3-deoxy-D-glycero-D-galacto-nononic acid). Identification and characterization of a KDN-9-phosphate synthetase activity from trout testis.

Although the deaminoneuraminic acid or KDN glycotope (2-keto-3-deoxy-D-glycero-D-galacto-nononic acid) is expressed in glycoconjugates that range in evolutionary diversity from bacteria to man, there is little information as to how this novel sugar is synthesized. Accordingly, biosynthetic studies were initiated in trout testis, an organ rich in KDN, to determine how this sialic acid is formed. These studies have shown that the pathway consists of the following three sequential reactions: 1) Man + ATP --> Man-6-P + ADP; 2) Man-6-P + PEP --> KDN-9-P + P(i); 3) KDN-9-P --> KDN + P(i). Reaction 1, catalyzed by a hexokinase, is the 6-O-phosphorylation of mannose to form D-mannose 6-phosphate (Man-6-P). Reaction 2, catalyzed by KDN-9-phosphate (KDN-9-P) synthetase, condenses Man-6-P and phosphoenolpyruvate (PEP) to form KDN-9-P. Reaction 3, catalyzed by a phosphatase, is the dephosphorylation of KDN-9-P to yield free KDN. It is not known if a kinase specific for Man (Reaction 1) and a phosphatase specific for KDN-9-P (Reaction 3) may exist in tissues actively synthesizing KDN. In this study, the KDN-9-P synthetase, an enzyme that has not been previously described, was identified as at least one key enzyme that is specific for the KDN biosynthetic pathway. This enzyme was purified 50-fold from rainbow trout testis and characterized. The molecular weight of the enzyme was estimated to be about 80,000, and activity was maximum at neutral pH in the presence of Mn(2+). N-Acetylneuraminic acid 9-phosphate (Neu5Ac-9-P) synthetase, which catalyzes the condensation of N-acetyl-D-mannosamine 6-phosphate and phosphoenol-pyruvate to produce Neu5Ac-9-P, was co-purified with the KDN-9-P synthetase. Substrate competition experiments revealed, however, that syntheses of KDN-9-P and Neu5Ac-9-P were catalyzed by two separate synthetase activities. The significance of these studies takes on added importance with the recent discovery that the level of free KDN is elevated in human fetal cord but not matched adult red blood cells and in ovarian cancer cells (Inoue, S., Lin, S-L., Chang, T., Wu, S-H., Yao, C-W., Chu, T-Y., Troy, F. A., II, and Inoue, Y. (1998) J. Biol. Chem. 273, 27199-27204). This unexpected finding emphasizes the need to understand more fully the role that free KDN and KDN-glycoconjugates may play in normal hematopoiesis and malignancy.

Elevated expression of free deaminoneuraminic acid in mammalian cells cultured in mannose-rich media.

Deaminoneuraminic acid (KDN, 2-keto-3-deoxy-D-glycero-D-galacto-nononic acid) is a member of the family of sialic acids in which an acylamino group at the C-5 position of N-acylneuraminic acid (Neu5Acyl) is replaced by a hydroxyl group. It has recently been shown that KDN is synthesized de novo from its precursor, mannose (Man), in trout testis (Angata, T., Nakata, D., Matsuda, T., Kitajima, K., and Troy, F. A. (1999) J. Biol. Chem. 274, in press). In this study, we examined the effect of extracellular free Man on biosynthesis of KDN in mouse melanoma B16 and African green monkey kidney COS-7 cell lines. The following new findings are reported. First, the levels of free and bound forms of KDN increased when the cells were cultured in the presence of 20 mM Man. The level of intracellular free KDN in COS-7 and B16 cells increased 47- and 66-fold respectively, compared with the levels in control cells. Second, the elevated expression of free KDN was proportional to the intracellular concentration of free Man. Third, KDN 9-phosphate (KDN-9-P) synthase, which condenses Man 6-phosphate and phosphoenolpyruvate (PEP), forming KDN-9-P, was detected in cell lysates from both cell lines. Fourth, the de novo synthesis of KDN in both cell lines in the Man-rich media was unaffected by the addition of N-acetylmannosamine (ManNAc), the hexosamine precursor for synthesis of N-acetylneuraminic acid (Neu5Ac). These results show that KDN is synthesized using free Man as its hexose precursor in these mammalian cells. Thus, the KDN biosynthetic pathway utilizes enzymes distinct, at least in part, from those involved in Neu5Ac biosynthesis. This is the first report showing that in vivo synthesis of KDN can be manipulated by growing cells in the presence of Man. This now provides a useful method to study the metabolism and function of the KDN glycotope.

Synthesis of neoglycoconjugates containing deaminated neuraminic acid (KDN) using rat liver alpha2,6-sialyltransferase.

2-Keto-3-deoxy-D- glycero -D- galacto -nononic acid (KDN) was introduced into asialotransferrin and N -acetyllactosamine (LacNAc) from CMP-KDN by using rat liver Galbeta1-->4GlcNAc alpha2, 6-sialyltransferase to form KDN-transferrin and KDN-LacNAc. These structures contain terminal KDNalpha2-->6Gal-residues, a glycotope that has not yet been described in natural glycoconjugates. KDN was transferred to all four Gal residues in asialotransferrin by this enzyme. The incorporation efficiency of KDN from CMP-KDN into asialotransferrin was about half that of Neu5Ac from CMP-Neu5Ac, based on the V max/ K m values for these donor substrates, 0.0527 min-1and 0.119 min-1, respectively. The KDNalpha2-->6Gal linkage was resistant to exosialidase treatment, in contrast to the sensitivity of the Neu5Acalpha2-->6Gal linkage. Interestingly, Sambucus sieboldiana agglutinin (SSA) was shown to prefer KDN-transferrin to the corresponding Neu5Ac-transferrin, as estimated by slot-blot analysis. The use of an alpha2,6-sialyltransferase to synthesize neoglycoproteins containing KDN has not been previously reported. Their facile synthesis using CMP-KDN and sialyltransferases with different specificities offers new possibilities to study the function of neo-KDN-glycoconjugates, and to explore their use in glycotechnology.

Identification, characterization, and developmental expression of a novel alpha 2-->8-KDN-transferase which terminates elongation of alpha 2-->8-linked oligo-polysialic acid chain synthesis in trout egg polysialoglycoproteins.

A novel glycosyltransferase which catalyses transfer of deaminated neuraminic acid, KDN (2-keto-3-deoxy-D-glycero-D-galacto-nononic acid) from CMP-KDN to the non-reducing termini of oligo-polysialyl chains of polysialoglycoprotein (PSGP), was discovered in the ovary of rainbow trout (Oncorhynchus mykiss). The KDN-transferase activity was optimal at neutral pH, and stimulated 2 to 2.5-fold by 2-5 mM Mg2+ or Mn2+. Expression of KDN-transferase was developmentally regulated in parallel with expression of the alpha 2-->8-polysialyltransferase, which catalyses synthesis of the oligo-polysialyl chains in PSGP. Incorporation of the KDN residues into the oligo-polysialyl chains prevented their further elongation, resulting in 'capping' of the oligo-polysialyl chains. This is the first example of a glycosyltransferase that catalyses termination of alpha 2-->8-polysialylation in glycoproteins.

Identification, developmental expression and tissue distribution of deaminoneuraminate hydrolase (KDNase) activity in rainbow trout.

A deaminoneuraminosyl-glycohydrolase (KDNase), which catalyses the hydrolysis of alpha-ketosidic 2-keto-3-deoxy-D-glycero-D-galacto- nononic acid (or naturally occurring deaminated neuraminic acid; KDN) linkages in KDN-glycoconjugates, is required for their structural and functional studies since KDN residues are usually resistant to the action of known sialidases. A search for KDNase was initiated by examining various cells and tissues of rainbow trout because KDN-glycoconjugates were first found in this animal species. Tissue localization studies of KDNase activity showed it to be present in kidney, spleen and ovary. The highest KDNase activity was found in ovarian post-ovulatory follicles obtained from female fish at the time when the reproductive organ was undergoing natural effacement. Little if any activity was found in brain, heart, liver, muscle, mature eggs and testis. Developmentally, higher levels of KDNase were usually expressed 3-4 months before ovulation or spermiation. An exception to this was in the ovary (or ovarian follicles) where the most striking increase in KDNase occurred 1-2 months after the maturation of gamete cells. Enzyme extracts containing KDNase activity also contained sialidase activity. From the data based on a kinetic study using mixed substrates, both KDNase and sialidase activities were indicated to reside on a single enzyme protein. The KDN-sialidase displayed broad specificity, which could possibly limit its usefulness as a probe for KDN-glycoconjugates. Nevertheless, unlike sialidases, KDNase can selectively remove KDN residues, thus making it an important new reagent to identify KDN-glycoconjugates in vivo.(ABSTRACT TRUNCATED AT 400 WORDS)