Studies on the sialidoses: properties of human leucocyte neuraminidases

Sialidase substrates for Sialdiase assays - activity, specificity, quantification and inhibition

Sialidases are glycosidases responsible for the removal of sialic acid (Sia) residues (desialylation) from glycan portions of either glycoproteins or glycolipids. By desialylation, sialidases are able to modulate the functionality and stability of the Sia-containing molecules and are involved in both physiological and pathological pathways. Therefore, evaluation of sialidase activity and specificity is important for understanding the biological significance of desialylation by sialidases and its function and the related molecular mechanisms of the physiological and pathological pathways. In addition, it is essential for developing novel mechanisms and approaches for disease treatment and diagnosis and pathogen detection as well. This review summarizes the most recent sialidase substrates for evaluating sialidase activity and specificity and screening sialidase inhibitors, including (i) general sialidase substrates, (ii) specific sialidase substrates, (iii) native sialidase substrates and (iv) cellular sialidase substrates. This review also provides a brief introduction of recent instrumental methods for quantifying the sialidase activity, such as UV, fluorescence, HPLC and LC-MS methods.

Comparative enzymology, biochemistry and pathophysiology of human exo-alpha-sialidases (neuraminidases)

This review summarizes the current research on human exo-alpha-sialidase (sialidase, neuraminidase). Where appropriate, the properties of viral, bacterial, and human sialidases have been compared. Sialic acids are implicated in diverse physiological processes. Sialidases, as enzymes acting upon sialic acids, assume importance as well. Sialidases hydrolyze the terminal, non-reducing, sialic acid linkage in glycoproteins, glycolipids, gangliosides, polysaccharides, and synthetic molecules. Therefore, a variety of assays are available to measure sialidase activity. Human sialidase is present in several organs and cells. Its cellular distribution could be cytosolic, lysosomal, or in the membrane. Human sialidase occurs in a high molecular-mass complex with several other proteins, including cathepsin A and beta-galactosidase. Multi-protein complexation is important for the in vivo integrity and catalytic activity of the sialidase. However, multi-protein complexation, the occurrence of isoenzymes, diverse subcellular localization, thermal instability, and membrane association have all contributed to difficulties in purifying and characterizing human sialidases. Human sialidase isoenzymes have recently been cloned and sequenced. Even though crystal structures for the human sialidases are not available, the highly conserved regions of the sialidase from various organisms have facilitated molecular modeling of the human enzyme and raise interesting evolutionary questions. While the molecular mechanisms vary, genetic defects leading to human sialidase deficiency are closely associated with at least two well-known human diseases, namely sialidosis and galactosialidosis. No therapy is currently available for either disease. A thorough investigation of human sialidases is therefore crucial to human health.

Removal of sialic acid from a glycoprotein in CHO cell culture supernatant by action of an extracellular CHO cell sialidase

We have directly tested the hypothesis that Chinese hamster ovary (CHO) cell-produced glycoproteins are subject to extracellular degradation by a sialidase endogenous to the CHO cell line. Factors important to understanding the potential for extracellular degradation are addressed including the glycoprotein specificity, subcellular source, mechanism of release, and stability of the sialidase activity. The extracellular CHO cell sialidase apparently originates from the cytosol of the cells, and is released to the cell culture supernatant as a result of damage to the cellular membrane. The extracellular sialidase is active toward a variety of CHO cell-produced glycoproteins, and can hydrolyze sialic acid from the recombinant glycoprotein gp120 in the culture supernatant. While measuring the actual degradation of a glycoprotein by extracellular CHO cell sialidase can be difficult, data presented here suggest that the level of degradation can be estimated indirectly by using a more convenient fluorescent substrate, 4-methylumbelliferyl-alpha-D-N-acetylneuraminic acid, to quantify sialidase activity. Degradation by sialidase is minimized through addition of the sialidase inhibitor 2,3-dehydro-2-deoxy-N-acetylneuraminic acid to the culture supernatant. The results in this study suggest additional potential approaches for minimizing degradation by sialidase, including isolation of a sialidase-deficient CHO cell line.

Freeze-stable sialidase activity in human leucocytes: substrate specificity, inhibitor susceptibility, detergent requirements and subcellular localization

Human leucocytes contain a freeze-stable sialidase (neuraminidase; EC 3.2.1.18) activity in addition to the better-characterized lysosomal freeze-labile enzyme. In order to discriminate between the sialidase activities detected with the synthetic fluorimetric substrate 4-methylumbelliferyl-alpha-D-N-acetylneuraminic acid (MU-Neu5Ac), different tritiated sialoglycoconjugate substrates were prepared. Using this sensitive radioactive assay system, leucocyte sialidase activity towards glycoproteins was shown to be labile to repeated freeze-thawing, but a Triton-stimulated activity towards gangliosides was entirely freeze-stable. Assay conditions were optimized for this freeze-stable ganglioside sialidase activity. Subcellular fractionation of mononuclear leucocytes (MNLs) on Percoll-density gradients showed that this ganglioside sialidase activity was entirely associated with the plasma membrane. Study of the detergent requirements showed that MNLs also demonstrated ganglioside sialidase activity when sodium cholate was present in place of Triton. Cholate-stimulated ganglioside sialidase activity was found to be entirely freeze-stable and localized at the plasma membrane. Studies on whole homogenates of MNLs demonstrated that the Triton-stimulated and cholate-stimulated activities showed similar acidic pH optima at < or = 3.9 and were both strongly inhibited by 2-deoxy-2,3-didehydro-N-acetylneuraminic acid and Cu2+, but not by free N-acetylneuraminic acid, N-(4-nitrophenyl)oxamic acid or heparan sulphate. These results suggest that human MNLs contain, in addition to the lysosomal freeze-labile sialidase, a single sialidase activity which is freeze-stable, ganglioside-specific, plasma membrane-associated and stimulated both by Triton and by cholate.

Lysosomal and cytosolic sialidases in rabbit alveolar macrophages: demonstration of increased lysosomal activity after in vivo activation with bacillus Calmette-Guerin

Sialidase activity was assayed in homogenized rabbit alveolar macrophages using a fluorogenic substrate: sodium 4-methylumbelliferyl-alpha-D-neuraminate. After differential centrifugation one acid-active enzyme (optimum pH 4.2) was detected in the 16,000 X g pellet that contained lysosomes, mitochondria and peroxisomes. A second activity, with an optimum pH of 5.4, was found in the cytosolic fraction. The acid-active sialidase accounted for more than 95% of the total sialidase activity in crude homogenate. When alveolar macrophages were collected from rabbits stimulated with bacillus Calmette-Guerin (BCG), the acid-active sialidase specific activity was increased 2.5-fold whereas other lysosomal enzymes such as N-acetylglucosaminidase and beta-galactosidase were stable. The cytosolic sialidase activity did not change.

Human placental neuraminidase. Activation, stabilization and association with beta-galactosidase and its protective protein

Supernatant of homogenized human placenta hardly contains lysosomal neuraminidase activity. It is, however, possible to generate remarkably high activity by concentration of a partially purified glycoprotein fraction. This activity is labile to dilution, but can be stabilized by incubation at 37 degrees C and acid pH. Using beta-galactosidase specific affinity chromatography and immunotitration, we show that the activated and stabilized human lysosomal neuraminidase exists in a complex with beta-galactosidase. Sucrose density gradient centrifugation experiments demonstrate that the neuraminidase activity is exclusively present in a high density multimeric form of beta-galactosidase. The formation of multimeric forms of beta-galactosidase is known to require a 32000-Mr 'protective' protein. Monospecific antibodies against this 'protective' protein were purified from a conventional antiserum containing a mixture of antibodies against the 64000-Mr beta-galactosidase protein and against the 32000-Mr 'protective' protein, using a nitrocellulose blot immunoaffinity purification procedure. Immunotitration experiments with these antibodies show that the 32000-Mr 'protective' protein is present both in association with the beta-galactosidase multimer and with the high-density multimeric form together with neuraminidase. Our data further suggest that association of the 32000-Mr 'protective' protein and another yet unidentified subunit is essential for the catalytic activity of lysosomal neuraminidase. These results explain the absence of neuraminidase activity in the autosomal recessive human lysosomal storage disorder galactosialidosis, where the 32000-Mr 'protective' protein is known to be absent.

Functional lysosomal hydrolase size as determined by radiation inactivation analysis

Electron inactivation analysis with 16 MeV electrons was used to determine the functional target size of a number of commonly studied lysosomal hydrolases. Observed values ranged from a low of 62 000 +/- 4000 Da for beta-galactosidase to a high of 200 000 +/- 17 500 Da (mouse beta-glucuronidase). One group of lysosomal hydrolases (N-acetyl-beta-glucosaminidase, N-acetyl-beta-galactosaminidase, alpha-galactosidase, beta-mannosidase, beta-glucosidase, arylsulphatase A and sphingomyelinase) had target sizes in the range 100 000-120 000 Da, whereas alpha-glucosidase and alpha-fucosidase exist as complex multimers in the 150 000-160 000 Da range. Analysis of freeze-dried cell material showed little evidence of species (mouse versus human) variation in the functional size of most lysosomal hydrolases with the exception of beta-glucuronidase. Our findings suggest the potential usefulness of lysosomal hydrolases as endogenous marker enzymes in studies where the target size of proteins of unknown molecular mass is to be determined.

Radiation inactivation of membrane proteins: molecular weight estimates in situ and after Triton X-100 solubilization

Target size analysis by radiation inactivation is widely used for molecular weight determination of membrane enzymes and receptors in situ without the need for prior solubilization or purification. However, since most molecular weight data available in the literature on membrane proteins involve the use of detergents for solubilization, the target sizes of membrane proteins in situ and after solubilization by detergent treatment have been compared. Using data from the literature and personal results, three different types of behavior of membrane proteins in presence of detergents were found: (i) uncoupling of subunits (electric eel acetylcholinesterase, placental steroid sulfatase, and human nonspecific beta-glucosidase); (ii) coupling of protein molecules (mouse liver neuraminidase, and rat liver insulin receptor regulatory component); and (iii) no major change in quaternary structure (rat liver insulin receptor, kidney gamma-glutamyltransferase, asialoglycoprotein receptor, insulin degrading enzyme, and human leucocyte neuraminidase). For all these proteins, there is a statistically significant increase in target size of about 24% over the value obtained in situ without detergent. A relatively large body of literature data involving a variety of membrane proteins, membrane types, and irradiation conditions (electron accelerators or 60Co sources, and proteins irradiated in lyophilized form or frozen solution) was examined, and it was concluded that target sizes of membrane proteins, irradiated in the presence of Triton X-100, should be diminished by a factor of about 24% to obtain the molecular weight value.

The role of accessory cells in polyclonal T cell activation. III. No requirement for recognition of H-2-encoded antigens on accessory cells

Highly purified murine lymph node T cells were used to test the hypothesis that polyclonal T cell activation requires the recognition of mitogen-modified major histocompatibility complex (MHC) antigens on accessory cells (AC) by the T cells. A variety of tumor cells lines, including macrophage, B and mast cell tumors, as well as thymomas, were shown to function as AC in concanavalin A-induced T cell activation, even if they expressed only one class of MHC antigens or none at all. In contrast to antigen-specific responses, where the Lyt-2+ phenotype is reportedly associated with recognition of class I MHC antigens, T cells enriched for or depleted of Lyt-2+ cells were not preferentially activated in the presence of class I- or class II-positive AC, respectively. In addition, as shown by others in the guinea pig and in the rat systems, T cell proliferation induced by oxidation of cell surface sugars is equally effective if T cells or AC are oxidized. T cell mitogens, therefore, do not seem to act by altering MHC antigens on AC, but rather by providing T cell-AC contact via their agglutinating properties.

Two genetically different MU-NANA neuraminidases in human leucocytes

Human leucocytes contain two different MU-NANA neuraminidases, which can be distinguished by Concanavalin A binding. The Con A binding form is predominant in lymphocytes (more than 80%) and the non-binding form predominates in granulocytes. The pH optima of both these neuraminidases as well as their subcellular localization as determined by Percoll gradient centrifugation suggest that they are both lysosomal. Immunological studies indicate that the Con A binding form is present in a complex with beta-galactosidase whereas the non-binding form is not. Leucocytes from patients with sialidosis or galactosialidosis are deficient in the Con A binding neuraminidase, whereas the non-binding form is normal. In sialolipidosis both forms are normal. These results demonstrate that leucocytes contain at least two genetically different MU-NANA neuraminidases. Thus, the use of leucocytes should be avoided for the diagnosis of sialidosis and galactosialidosis, and isolated lymphocytes should be used to obtain reliable results.

The role of accessory cells in polyclonal T cell activation II. Induction of interleukin 2 responsiveness requires cell-cell contact

It has previously been reported (Hünig, T. et al., Eur. J. Immunol. 1983. 13:1) that highly purified peripheral T cells do not respond to concanavalin A (Con A) even in the presence of Con A-induced spleen cell supernatant as a source of interleukin 2 (IL 2). In the present report, the hypothesis was tested whether this unresponsiveness correlates with the observed inability of Con A to mediate cell-cell contact between highly purified T cells. It was found that T cells, pretreated with neuraminidase to reduce their net negative charge, were both aggregated and rendered IL 2-reactive by Con A. In addition, leukoagglutinin (LA) was found to be able to both agglutinate untreated T cells and to make them IL 2-reactive. Neuraminidase treatment reduced the concentration of LA required to mediate both effects. Neuraminidase treatment did not alter the overall Con A-binding capacity of T cells, nor did it induce reactivity to IL 2 in the absence of lectin. Furthermore, irradiated, neuraminidase-treated T cells served as accessory cells (AC) in the induction of responsiveness to IL 2, but not for production of IL2, which depends on Ia+ AC. Finally, Lyt-2- neuraminidase-treated peripheral T cells responded in the same fashion as whole T cell populations, indicating that at least some T cells of the helper phenotype need not interact with Ia+AC for the induction of IL 2 responsiveness.

Lysosomal sialidase deficiency in sialidosis with partial beta-galactosidase deficiency

We analyzed the subcellular localization of sialidases in human lymphocytes from a patient with adult type sialidosis with partial beta-galactosidase deficiency and normal controls. Sialidase activities were measured with alpha,2 leads to 3 NeuAc-lactitol, 4-methylumbelliferyl-NeuAc and GM3 ganglioside as substrates. Sialidases in the lysosomes were sonication-labile and hydrolyzed mainly hydrophilic substrates such as NeuAc-lactitol and 4-methylumbelliferyl-NeuAc, but hydrolyzed subsidiarily GM3 ganglioside. On the other hand, sialidases in the plasma membrane were sonication-stable and hydrolyzed both hydrophilic substrates and GM3 ganglioside. In sialidosis with partial beta-galactosidase deficiency, the sialidases of the lysosomes showed 3-5% activity toward hydrophilic substrates and 25% activity toward GM3 ganglioside as compared with sialidase activities of the controls. However, there are no differences in the activities of the sialidases in the plasma membrane. These results demonstrate that the essential defect in this disease is the deficiency of a lysosomal sialidase.

Characterization of human placental neuraminidases. Stability, substrate specificity and molecular weight

1. At least two components of neuraminidase can be distinguished on the basis of thermolability and sedimentability by using the artificial fluorogenic substrate 4-methylumbelliferyl N-acetyl-alpha-D-neuraminate. 2. In crude homogenates, thermodenaturation at 25 degrees C showed a biphasic curve corresponding to component A (half-life, 21 min) and B (half-life, 85 min). The two components were partially resolved by centrifugation. A being soluble and B sedimentable. Both had similar pH-activity curves (pH optimum, 4.4), Km values (A, 0.10 mM; B, 0.06 mM) and molecular weight as determined by radiation inactivation (A, 67000; B, 63000). 3. The soluble A form was still aggregated or bound to membranous debris since almost all neuraminidase activity was eluted near or at the void volume of a Sephacryl S-300 column. 4. Both soluble and sedimentable fractions of placenta hydrolysed the GD1A ganglioside and N-acetyl-neuraminyl-D-lactose linearly for 12 h but no fetuin hydrolysis was detected. 5. The neuraminidase activity with the artificial fluorogenic substrate was inhibited by N-acetylneuraminyl-D-lactose but not by the GD1A ganglioside. These preliminary results suggest that there exist two closely related enzymes hydrolysing both the artificial substrate and N-acetylneuraminyl-D-lactose and a third one hydrolysing the GD1A ganglioside exclusively.

Neuraminidase deficiency and accumulation of sialic acid in lymphocytes in adult type sialidosis with partial beta-galactosidase deficiency

Neuraminidase deficiency has been demonstrated in cultured skin fibroblasts of patients who have adult type sialidosis with partial beta-galactosidase deficiency. A substantial amount of residual enzyme activity has been observed in leukocytes, however. To explain this discrepancy, the nature and distribution of the enzyme were studied. Neuraminidase activity was higher in lymphocytes than in granulocytes of normal controls. In patients' lymphocytes, neuraminidase activity was profoundly decreased and total sialic acid contents were increased 2.3-fold. Two neuraminidases, one sonication-labile and the other sonication-stable, were found in lymphocytes; the former was predominant in cultured skin fibroblasts. The defective enzyme in this disorder was found to be the sonication-labile neuraminidase in both cultured skin fibroblasts and lymphocytes.

Neuraminidase activities in sialidosis and mucolipidosis

Defects of neuraminidase activities towards sialyloligosaccharides in fibroblasts and leucocytes and enhanced excretion of sialyloligosaccharides in urine were shown in patients with adult type sialidosis with partial deficiency of beta-galactosidase and cherry red spot-myoclonus syndrome. No differences in their neuraminidase residual levels and urinary excretion patterns on thin-layer chromatography were found between these two disorders. In mucolipidosis II and III patients, the neuraminidase activities towards sialyloligosaccharides were almost normal in leucocytes, although decreased in fibroblasts. The discrepancy of neuraminidase activities towards 2 leads to 3 and 2 leads to 6 sialyloligosaccharide isomers was not noticed in all cases.

beta-Galactosidase-neuraminidase deficiency in adults: deficiency of a freeze-labile neuraminidase in leukocytes and fibroblasts

4-methylumbelliferyl neuraminidase activity was studied in fibroblasts, leukocytes, and frozen tissues from adult patients with beta-galactosidase-neuraminidase deficiency and specific clinical manifestations. This enzyme was almost completely deficient in fibroblasts, but the residual activity was relatively high (20% of the control mean) in the leukocytes from the patients. The frozen liver from one patient showed the enzyme activity as high as controls. This enzyme consisted of two components, freeze-labile and freeze-stable, and it was demonstrated that only the labile enzyme was deficient in fibroblasts and leukocytes. The apparently normal activity of neuraminidase in frozen autopsy tissues of a patient may be explained by the loss of the labile component in control tissues after a long-term freezing. The neuraminidase activity was variable in parents and no definite conclusion was drawn on the hereditary nature of the disease.