Analysis of neuraminidase isozyme phenotypes in mammalian tissues: an electrophoretic approach

Antibody-mediated sialidase activity in blood serum of patients with multiple myeloma

Cell surface sialylation is known to be tightly connected with tumorigenicity, invasiveness, metastatic potential, clearance of aged cells, while the sialylation of IgG molecules determines their anti-inflammatory properties. Four sialidases - hydrolytic enzymes responsible for cleavage of sialic residues - were described in different cellular compartments. However, sialidases activity in body fluids, and specifically in blood serum, remains poorly studied. Here, we characterize first known IgG antibodies possessing sialidase-like activity in blood serum of multiple myeloma (MM) patients. Ig fractions were precipitated with ammonium sulfate (50% of saturation) from blood serum of 12 healthy donors and 14 MM patients, and screened for the presence of sialidase activity by using 4-MUNA (2'-(4-methylumbelliferyl)-α-D-N-acetylneuraminic acid) as substrate. High level of sialidase activity was detected in the MM patients, but not in healthy donors. Subsequent antibody purification by protein-G affinity chromatography and HPLC size exclusion chromatography at acidic conditions demonstrated that sialidase activity was attributable to IgG molecules. Sialidase activity was also specific for (Fab)(2) fragment of IgG and blocked by sialidase inhibitor DANA. Sialidase activity of IgG molecule was also confirmed by in gel assay for cleavage of sialidase substrate. Kinetic parameters of the catalysis reaction were described by Michaelis-Menten equation with K(m)  = 44.4-108 µM and k(cat) = 2.7-23.1 min(-1). The action of IgG possessing sialidase-like activity towards human red blood cells resulted in a subsequent increase in their agglutination by the peanut agglutinin, that confirms their desialylation by the studied IgG. This is the first demonstration of the intrinsic sialidase activity of IgG isolated from blood serum of MM patients.

Mobilization of sialidase from intracellular stores to the surface of human neutrophils and its role in stimulated adhesion responses of these cells

Desialation of cell surfaces has been associated with the initiation or modification of diverse cellular functions. In these studies we have examined the subcellular distribution of sialidase (SE) in human neutrophils as well as the mobilization of this enzyme following neutrophil activation. Separation of subcellular fractions by density gradient centrifugation showed that SE is present not only in neutrophil primary and secondary granule populations, like lysozyme, but also in plasma membrane fractions. Neutrophil activation was associated with a redistribution of SE from secondary granule-enriched fractions to the plasma membrane. Furthermore, SE activity detected on the surface of intact neutrophils with a fluorescent SE substrate increased rapidly after activation with kinetics that matched both the loss of total cell-associated sialic acid and release of free sialic acid from the cells. These activation-dependent events were in each case blocked by incubation of neutrophils with the SE inhibitor, 2-deoxy-N-acetyl-neuraminic acid. Aggregation responses of neutrophils as well as adhesion responses to nylon and plastic surfaces were also inhibited by 2-deoxyNANA. Our findings indicate that the activation-dependent desialation of the neutrophil surface is associated with mobilization of an endogenous SE to the plasma membrane and has a role in stimulated adhesion responses of these cells.

Biochemical characteristics and subcellular localizations of rat liver neuraminidase isozymes: a paradox resolved

A striking discrepancy in the abilities of two analytical approaches (fluorometric and electrophoretic) to detect the effect of a gene, Neu-2, on rat liver neuraminidase phenotypes led us to examine the biochemical and physical properties of the liver isozymes NEU-1 and NEU-2 that might be responsible for this difference. Cell fractionation via Percoll gradient centrifugation revealed NEU-1 activity almost exclusively in the lysosomal cell fraction, while NEU-2 was strictly cytosolic in distribution. The two isozymes were also found to differ in pH activity curves and optima (optima: 4.6-4.8 and 5.4-5.8 for NEU-1 and NEU-2, respectively) and in solubility characteristics (NEU-2 highly soluble; NEU-1 relatively insoluble but solubilized by freezing/thawing). Both isozymes were found to be freeze-thaw stable in crude, whole-cell extracts, but NEU-1 was destabilized in the enriched (partially purified) lysosomal subcellular fraction. Consideration of these properties relative to those described previously for unidentified cytosolic and membrane bound (lysosomal) rat liver neuraminidases (Tulsiani, D. R. P., and Carubelli, R., J. Biol. Chem. 245:1821, 1970) leads us to believe that NEU-2 also is destabilized by partial purification and that NEU-1 and NEU-2 have very different relative abundances within the cell. The biochemical and physical differences between NEU-1 and NEU-2 can account for the discrepant abilities of the fluorometric and electrophoretic approaches to detect the effects of Neu-2. Ways to increase the sensitivity of the fluorometric approach for quantitative assays of specific NEU-1 and NEU-2 activity are discussed.

Cellular localization and substrate specificity of isoelectric forms of human liver neuraminidase activity

The four major isoelectric forms of human liver neuraminidase (with pI values between 3.4 and 4.8) have been isolated by preparative isoelectric focusing and characterized with regard to their substrate specificity using glycoprotein, glycopeptide, oligosaccharide and ganglioside natural substrates. All forms exhibited a rather broad linkage specificity and were capable of hydrolyzing sialic acid glycosidically linked alpha 2-3, alpha 2-6 and alpha 2-8, although differential rates of hydrolysis of the substrates were found for each form. The most acidic form 1 (pI 3.4) was most active on sialyl-lactose, whereas form 2 (pI 3.9) and 3 (pI 4.4) were most active on the more hydrophobic ganglioside substrates. Form 4 (pI 4.8) was most active on the low-Mr hydrophilic substrates (fetuin glycopeptide, sialyl-lactose). Each form was less active on the glycoprotein fetuin than on a glycopeptide derived from fetuin. Organelle-enriched fractions were prepared from fresh human liver tissue and neuraminidase activity on 2'-(4-methylumbelliferyl)-alpha-D-N-acetylneuraminic acid was recovered in plasma membrane, microsomal, lysosomal and cytosolic preparations. Isoelectric focusing of the neuraminidase activity recovered in each of these preparations resulted in significantly different isoelectric profiles (number, relative amounts and pI values of forms) for each preparation. The differential substrate specificity of the isoelectric forms and the different isoelectric focusing profiles of neuraminidase activity recovered in subcellular-enriched fractions suggest that specific isoelectric forms with broad but defined substrate specificity are enriched at separate sites within the cell.

Electrophoretic analysis of liver neuraminidase-1 variation in mice and additional evidence concerning the location of NEU-1

Neuraminidase-1 (NEU-1) is one of two neuraminidase isozymes which can be detected electrophoretically in mouse liver extracts. The inheritance of variation in NEU-1 and the linkage relationships of the gene controlling this variation were studied through a backcross analysis involving the SM/J and MA/MyJ inbred strains, and by examination of NEU-1 phenotypes in three congenic strains: B10.SM, B10.SM(22R) and B10.RVB. The data indicate that NEU-1 is controlled by Neu-1, a gene previously identified by its effect on total liver neuraminidase activity in whole tissue homogenates. Analysis of the congenic strains revealed identical low activity (SM/J-type: Neu-1a/Neu-1a) NEU-1 phenotypes in all three strains. This indicates that Neu-1 lies in the segment of the SM/J-derived H-2 region that is common to all three strains: H-2E alpha to H-2D. In addition, we examined the relationship between NEU-1 and phenotypic variation in liver acid phosphatase (AP; for which a new typing method is described) and linkage order among several other enzyme-coding genes linked to H-2. In all animals that could be scored confidently for AP, the NEU-1 and AP phenotypes were concordant, adding support to the hypothesis that both phenotypes are controlled by Neu-1. Recombination rates among six H-2-linked marker loci were unexpectedly low, but were sufficient to verify the position of Upg-1 as the telomeric flanking marker relative to Glo-1, H-2 (C4), Neu-1 (Apl), Ce-2 and Pgk-2.