Purification and partial characterization of lysosomal neuraminidase from human placenta

Lysosomal neuraminidase and beta-galactosidase are present in a complex together with a 32-kDa protective protein. This complex has been purified and the different components have been dissociated using potassium isothiocyanate (KSCN) treatment. beta-Galactosidase remains catalytically active, but neuraminidase loses its activity upon dissociation. The inactive dissociated neuraminidase was purified by removing the remaining non-dissociated beta-galactosidase/protective protein complex using beta-galactosidase-specific affinity chromatography. The dissociated neuraminidase material shows two major polypeptides on SDS-PAGE with an apparent molecular mass of 76 kDa and 66 kDa. Subsequently the 32-kDa protective protein was dissociated from the beta-galactosidase/protective protein complex, and purified. Antibodies raised against the dissociated inactive neuraminidase preparation specifically immunoprecipitate the active neuraminidase present in the complex with beta-galactosidase and protective protein. By immunoblotting evidence is provided that the 76-kDa protein is a subunit of neuraminidase which, in association with the 32-kDa protective protein, is essential for neuraminidase activity.     

Identification and in vitro reconstitution of lysosomal neuraminidase from human placenta

Lysosomal neuraminidase from human placenta has been obtained in its active form by association of an inactive neuraminidase polypeptide with beta-galactosidase and the protective protein. Using a specific antiserum, we have now identified a 66-kDa protein as the inactive neuraminidase polypeptide. It is specifically recognized on immunoblots only in its nonreduced state, and it coprecipitates with neuraminidase activity. The 66-kDa polypeptide is substantially glycosylated (38-kDa protein core with 7-14 N-linked oligosaccharide chains), a feature characteristic of lysosomal integral membrane proteins. Specific removal of the 66-kDa neuraminidase polypeptide from glycoprotein preparations prevents the generation of neuraminidase activity. Removal of beta-galactosidase or destruction of the protective protein also hinders the formation of active neuraminidase. Reconstitution of neuraminidase activity is observed after mixing glycoprotein preparations, depleted in different components of the beta-galactosidase-neuraminidase-protective protein complex, indicating that all three components of the complex are required for neuraminidase activity. Association of the neuraminidase polypeptide and the protective protein generates unstable neuraminidase activity, whereas association with beta-galactosidase is required for stability.     

Human lymphocyte membrane proteins treated with neuraminidase

Human peripheral blood lymphocytes were surface-iodinated, treated with neuraminidase from Vibrio cholerae and lysed with non-ionic detergent. In addition, surface membrane fractions were isolated from surface-iodinated cells in the absence of detergents and treated with neuraminidase after membrane isolation. The effect of neuraminidase treatment on the membrane proteins was studied by two-dimensional gel electrophoresis. One surface-labelled protein of 45 000 molecular weight which is characterized by its association with the detergent-resistant matrix of the cells and by its specific enrichment in an isolated membrane fraction, was found to be particularly sensitive to neuraminidase treatment both of intact cells and isolated membranes. A prominent labelled protein of apparent molecular weight of 60 000 is observed in the soluble fraction after neuraminidase treatment of intact cells. The analogous protein is detected when isolated membrane fractions are treated with neuraminidase.     

An enzymatic assay reveals that proteins destined for the apical or basolateral domains of an epithelial cell line share the same late Golgi compartments.

The expression of viral envelope proteins on the plasma membrane domains of the epithelial cell line, MDCK, is polar. Influenza virus infection of these cells leads to expression of the viral haemagglutinin and neuraminidase glycoproteins on the apical domain of the plasma membrane while vesicular stomatitis virus (VSV) infection yields basolateral expression of the sialic acid-bearing G protein. We have exploited the ability of the influenza neuraminidase to desialate the G protein of VSV to test for contact between these proteins during their intracellular transport to separate plasma membrane domains. We were able to select for VSV-G protein expression in doubly-infected cells because VSV protein production was accelerated in cells pre-infected with influenza virus. During double infection the envelope proteins of both viruses displayed the same polar localization as during single infection but the VSG-G protein was undersialated due to the action of the influenza neuraminidase. Incubation of singly-infected cells at 20 degrees C blocked the transport of VSV-G protein to the cell surface and resulted in increased sialation of the protein over that seen at 37 degrees C. This suggests that G protein is held in contact with the sialyl transferase at this temperature. 20 degrees C incubations of doubly-infected cells also produced the undersialated G protein characteristic of interaction with the neuraminidase. We conclude that most of the newly synthesised basolaterally-directed G protein is in physical contact with the majority of the neuraminidase through the terminal steps of Golgi processing.     

Transport of human lysosomal neuraminidase to mature lysosomes requires protective protein/cathepsin A.

Human lysosomal N-acetyl-alpha-neuraminidase is deficient in two lysosomal storage disorders, sialidosis, caused by structural mutations in the neuraminidase gene, and galactosialidosis, in which a primary defect of protective protein/cathepsin A (PPCA) leads to a combined deficiency of neuraminidase and beta-D-galactosidase. These three glycoproteins can be isolated in a high molecular weight multi-enzyme complex, and the enzymatic activity of neuraminidase is contingent on its interaction with PPCA. To explain the unusual need of neuraminidase for an auxiliary protein, we examined, in transfected COS-1 cells, the effect of PPCA expression on post-translational modification, turnover and intracellular localization of neuraminidase. In pulse-chase studies, we show that the enzyme is synthesized as a 46 kDa glycoprotein, which is poorly phosphorylated, does not undergo major proteolytic processing and is secreted. Importantly, its half-life is not altered by the presence of PPCA. However, neuraminidase associates with the PPCA precursor shortly after synthesis, since the latter protein co-precipitates with neuraminidase using anti-neuraminidase antibodies. We further demonstrate by subcellular fractionation of transfected cells that neuraminidase segregates to mature lysosomes only when accompanied by wild-type PPCA, but not by transport-impaired PPCA mutants. These data suggest a novel role for PPCA in the activation of lysosomal neuraminidase, that of an intracellular transport protein.     

Soluble recombinant influenza vaccines.

Soluble, recombinant forms of influenza A virus haemagglutinin and neuraminidase have been produced in cells of lower eukaryotes, and shown in a mouse model to induce complete protective immunity against a lethal virus challenge. Soluble neuraminidase, produced in a baculovirus system, consisted of tetramers, dimers and monomers. Only the tetramers were enzymatically active. The immunogenicity decreased very considerably in the order tetra > di > mono. Therefore, we fused the head part of the neuraminidase gene to a tetramerizing leucine zipper sequence; the resulting product was enzymatically active, tetrameric neuraminidase. The protective immunity induced by this engineered neuraminidase, however, remained fairly strain-specific. A third influenza A virus protein, the M2 protein, has only 23 amino acids exposed on the outer membrane surface. This extracellular part, M2e, has been remarkably conserved in all human influenza A strains since 1933. By fusing the M2e sequence to hepatitis B virus core protein, we could obtain highly immunogenic particles that induced complete, strain-independent, long-lasting protection in mice against a lethal viral challenge. Native M2 is a tetrameric protein and this conformation of the M2e part can also be mimicked by fusing this sequence to a tetramerizing leucine zipper. The potential of the resulting protein as a vaccine candidate remains to be evaluated.     

In vitro activation of neuraminidase in the beta-galactosidase-neuraminidase-protective protein complex by cathepsin C

Neuraminidase can be activated by incubation of crude glycoprotein fractions at acidic pH for 90 minutes at physiological temperature. This activation is inhibited by leupeptin. Incubation of the purified neuraminidase-beta-galactosidase-protective protein complex under the same conditions used for crude glycoprotein fractions did not lead to enhanced neuraminidase activity, but incubation in the presence of exogenous Cathepsin C at 4 degrees C resulted in marked enhancement of neuraminidase activity. This activation was again inhibited by leupeptin. Cathepsin D treatment resulted in destruction of neuraminidase under the same conditions and this effect was again inhibited by leupeptin. beta-galactosidase in crude glycoprotein fractions and in the complex was resistant to both Cathepsin C and D, while homogeneous beta-galactosidase was inactivated by these enzymes. We suggest that in vitro activation of neuraminidase may mimic the in vivo intralysosomal conversion of the neuraminidase precursor into the mature form of the enzyme.     

Characterization of the sites of proteolytic activation of Newcastle disease virus membrane glycoprotein precursors

The F1- and F2-polypeptide components of the fusion proteins and the hemagglutinin/neuraminidase proteins of the avirulent Queensland (V4) and virulent Australia-Victoria (AuV) strains of Newcastle disease virus have been isolated and subjected to extensive primary structural analysis including amino-terminal sequence analysis and fast atom bombardment-mass spectrometry mapping. Nucleotide sequence analysis was performed on the gene which encodes the V4 hemagglutinin/neuraminidase protein. Signal peptidase cleavage was found to have occurred at the Ser31-Leu32 peptide bond of the primary translation products of the fusion protein genes. Activation cleavage of the V4 fusion protein precursor generated a sequence of -Gly-Lys-Gln-Gly84 at the carboxyl terminus of the F2-polypeptide and an amino-terminal sequence of the F1-polypeptide commencing with 86Leu-Ile-Gly-. The V4 hemagglutinin/neuraminidase protein gene was found to encode a primary translation product 45 amino acids longer at the carboxyl terminus than obtainable from the corresponding gene of the AuV strain (McGinnes, L. W., and Morrison, T. G. (1986) Virus Res. 5, 343-356). However, post-translational proteolytic processing, exclusive to the primary translation product of the V4 hemagglutinin/neuraminidase protein gene, was found to have removed the last 42 residues of this carboxyl-terminal appendage.     

Further Evidence for an Intrinsic Neuraminidase in CNS Myelin

An intrinsic neuraminidase activity in rat brain CNS myelin has been demonstrated and compared with the neuraminidase activity in rat brain microsomes. With use of ganglioside GM3 as a substrate, the myelin-associated neuraminidase exhibited a shallow pH curve with an optimum at pH 4.8 whereas the microsomal activity had a marked optimum at pH 4-4.3. Neuraminidase activity in both fractions was optimized in 0.3% Triton CF-54 but activation was much greater in the microsomes. When the neuraminidase activities were examined at 60 degrees C, the myelin neuraminidase activity was more than sevenfold of that observed at 37 degrees C and was linear for at least 2 h; the microsomal activity increased only fivefold initially and exhibited a continual loss in activity. Addition of excess microsomes to the total homogenate prior to myelin isolation resulted in no change in myelin neuraminidase activity. When the two membrane fractions were examined at equivalent protein concentrations in the presence of additional cations or EDTA (1 mM), similar but not identical effects on neuraminidase activity were seen. The microsomal neuraminidase was considerably more susceptible to inhibition by divalent copper ion. Activity in both fractions was markedly inhibited by Hg2+ and Ag+ whereas EDTA had no effect on either activity. The myelin-associated neuraminidase activity was the highest in cerebral hemispheres, followed by brainstem, cerebellum, and spinal cord and was extremely low in sciatic nerve. In fact, the myelin neuraminidase activity was higher than the microsomal enzyme activity in the cerebral hemispheres.(ABSTRACT TRUNCATED AT 250 WORDS)     

Interaction of influenza virus proteins with planar bilayer lipid membranes. I. Characterization of their adsorption and incorporation into lipid bilayers

Alterations in the surface potential difference (delta U) of asolectin planar bilayer lipid membranes were measured following the adsorption of isolated matrix protein (M-protein) or neuraminidase of influenza virus. The method used was based upon measurement of the bilayer lipid membrane capacitance current second harmonic. The delta U dependence on the M-protein and neuraminidase concentration indicates different mechanisms of adsorption of these viral proteins by the lipid bilayer. The conductance (G0) dependence of the bilayer lipid membrane with different compositions on the concentration of isolated surface glycoproteins, hemagglutinin and neuraminidase, M-protein or neuraminidase was investigated. The change in G0 for M-protein was observed only after adsorption saturation had been achieved. Neuraminidase alone does not affect the membrane conductivity. The surface charge and lipid composition of the lipid bilayer influences the adsorption and incorporation of influenza virus M-protein and surface glycoproteins. The reversibility of protein incorporation into the bilayers was investigated by a perfusion technique. The results show reversibility of surface glycoprotein incorporation while M-protein binding appears to be irreversible.     

Comparison of desialylation of rat transferrin by cellular and non-cellular methods.

We have previously shown that the liver endothelium can desialylate the glycoprotein transferrin (Tf). In the present work we provide evidence that asialotransferrin obtained by this means behaves differently on Ricinus communis agglutinin (RCA120) lectin affinity chromatography from asialotransferrin obtained by either neuraminidase treatment or acid hydrolysis. Purified rat transferrin was radiolabelled either with 125I (protein moiety) or with 3H (sialyl residues), and subsequently saturated with iron. It was then passed through an RCA120-agarose column to isolate the fully sialylated component. Sialylated Tf was then desialylated either by incubation with purified rat liver endothelium or, in vitro, by neuraminidase treatment or by acid hydrolysis. The protein was again subjected to RCA120 column chromatography. Although both neuraminidase treatment and acid hydrolysis almost completely desialylated the glycoprotein (as evidenced by near absence of 3H label), the glycoprotein was not retained by the RCA120-agarose column. By contrast, liver endothelium partially desialylated the glycoprotein, but this desialylated fraction was retained by the RCA120-agarose column. These results suggest that desialylation with neuraminidase or acid hydrolysis may be inadequate for functional studies of asialotransferrin.     

Isolation and Purification of the Envelope Proteins of Newcastle Disease Virus

A procedure has been developed for the isolation of Newcastle disease virus (NDV) envelope proteins. The two surface glycoproteins and the non-glycosylated membrane protein were solubilized with 2% Triton X-100 and 1 m KCl. Removal of the KCl by dialysis yielded by precipitation a pure preparation of the non-glycosylated membrane protein, which is insoluble in solutions of low ionic strength. The soluble fraction consisting of the two glycoproteins possessed full neuraminidase and hemagglutinating activities. The two glycoproteins could be separated by rate zonal sedimentation in a sucrose gradient containing 1% Triton X-100 and 1 m KCl. Under these conditions, the sedimentation coefficient of the larger glycoprotein, virus protein 1, was 9.3s, and that of the smaller, virus protein 2, was 6.1s. Both hemagglutinating and neuraminidase activities were associated with virus protein 1; virus protein 2 had neither activity. The results suggest that both activities reside on a single NDV glycoprotein. Similar results were obtained previously with another paramyxovirus, simian virus 5. These findings suggest that the association of hemagglutinating and neuraminidase activities with one glycoprotein is a general property of the paramyxovirus group.     

The production of neuraminidase and fucosidase by Helicobacter pylori: Their possible relationship to pathogenicity

Abstract The pathogenicity of enterobacteria often correlates with their production of neuraminidase (sialidase). Forty-nine Helicobacter pylori isolates have therefore been examined for their production of neuraminidase and other glycosidases. All 49 isolates produced considerable neuraminidase (median 228 IU/μgmg protein, interquartile range 121–370), pH optimum 7.5. Nine of the 49 also produced fucosidase (median 23 IU/μgmg protein, interquartile range 12–39), pH optimum 7.0. Production of these enzymes did not correlate with bacterial Cag A expression or duodenal ulceration. Neutrophils exposed to neuraminidase show increased adherence to endothelium so the neuraminidase production by H. pylori could partly explain the predominant neutrophil inflammatory infiltrate seen in H. pylori -associated gastritis. Inhibition of this enzyme by use of neuraminidase-inhibitors could be a useful therapeutic approach.     

Photoaffinity labeling of the lysosomal neuraminidase from bovine testis

ASA-NeuAc2en, a photoreactive arylazide derivative of sialic acid, is shown to be a powerful competitive inhibitor of lysosomal neuraminidase from bovine testis (Ki approximately 21 microM). Photoaffinity labeling and partial purification of preparations containing this lysosomal neuraminidase activity result in specifically and non-specifically labeled polypeptides. Only labeling in a 55 kDa polypeptide is found to be specific, since it could be prevented by the competitive neuraminidase inhibitor NeuAc2en. We conclude that the 55 kDa polypeptide in the bovine testis beta-galactosidase/neuraminidase/protective protein complex contains the catalytic site of neuraminidase.     

Influence of the human parainfluenza virus 3 attachment protein's neuraminidase activity on its capacity to activate the fusion protein

In order to examine functions of the hemagglutinin-neuraminidase (HN) protein that quantitatively influence fusion promotion, human parainfluenza virus 3 (HPIV3) variants with alterations in HN were studied. The variant HNs have mutations that affect either receptor binding avidity, neuraminidase activity, or fusion protein (F) activation. Neuraminidase activity was regulated by manipulation of temperature and pH. F activation was assessed by quantitating the irreversible binding of target erythrocytes (RBC) to HN/F-coexpressing cells in the presence of 4-GU-DANA (zanamivir) to release target cells bound only by HN-receptor interactions; the remaining, irreversibly bound target cells are retained via the fusion protein. In cells coexpressing wild-type (wt) or variant HNs with wt F, the fusion promotion capacity of HN was distinguished from target cell binding by measuring changes with time in the amounts of target RBC that were (i) reversibly bound by HN-receptor interaction (released only upon the addition of 4-GU-DANA), (ii) released by HN's neuraminidase, and (iii) irreversibly bound by F-insertion or fusion (F triggered). For wt HN, lowering the pH (to approach the optimum for HPIV3 neuraminidase) decreased F triggering via release of HN from its receptor. An HN variant with increased receptor binding avidity had F-triggering efficiency like that of wt HN at pH 8.0, but this efficiency was not decreased by lowering the pH to 5.7, which suggested that the variant HN's higher receptor binding activity counterbalanced the receptor dissociation promoted by increased neuraminidase activity. To dissect the specific contribution of neuraminidase to triggering, two variant HNs that are triggering-defective due to a mutation in the HN stalk were evaluated. One of these variants has, in addition, a mutation in the globular head that renders it neuraminidase dead, while the HN with the stalk mutation alone has 30% of wt neuraminidase. While the variant without neuraminidase activity triggered F effectively at 37 degrees C irrespective of pH, the variant possessing effective neuraminidase activity completely failed to activate F at pH 5.7 and was capable of only minimal triggering activity even at pH 8.0. These results demonstrate that neuraminidase activity impacts the extent of HPIV3-mediated fusion by releasing HN from contact with receptor. Any particular HN's competence to promote F-mediated fusion depends on the balance between its inherent F-triggering efficacy and its receptor-attachment regulatory functions (binding and receptor cleavage).     

Renal neuraminidase. Characterization in normal rat kidney and measurement in experimentally induced nephrotic syndrome.

Several lines of evidence suggest that increased neuraminidase activity may be responsible for the loss of glomerular N-acetylneuraminic acid (AcNeu) observed in various glomerular diseases. However, virtually no information is available on the activity of neuraminidase in glomeruli or the potential role of this enzyme in glomerular pathophysiology. Utilizing 2'-(4-methylumbelliferyl)-alpha-D-N-acetylneuraminic acid (4MU-AcNeu) as substrate, we defined optimal assay conditions and characterized neuraminidase activity in glomeruli and, for comparison, in other renal fractions and liver. Neuraminidase activity in glomeruli, cortex and tubules was maximal at pH 4.4. The Km for 4MU-AcNeu was estimated to be 195 microM for glomeruli and 226 microM for cortex. Glomerular neuraminidase was inhibited by AcNeu (90% at 25 mM) and high concentrations of Triton X-100 (26% at 0.5%), but unaffected by CaCl2, EDTA or N-ethylmaleimide (each 1 mM). Neuraminidase activity (nmol/h per mg of protein; mean +/- S.E.M.) in normal rat kidney was: cortex, 14.47 +/- 0.76; medulla, 7.85 +/- 0.64; papilla, 2.64 +/- 0.11; tubules, 13.79 +/- 0.70; glomeruli, 5.57 +/- 0.28. In comparison, neuraminidase activity in rat liver was 2.58 +/- 0.14. Puromycin aminonucleoside (PAN)-induced nephrotic syndrome is a model of glomerular disease in which the loss of glomerular AcNeu is well documented. In two separate studies, we observed no change in the specific activity of neuraminidase in either glomeruli or cortex isolated from rats treated with PAN (15 mg/100 g, intraperitoneally) and killed at either the onset or the peak of proteinuria. Results were similar whether neuraminidase activity was expressed per mg of protein or per microgram of DNA.     

Purification, microheterogeneity, and stability of human lipid transfer protein

A method for the purification of lipid transfer protein (LTP) from human plasma was developed with the aid of succinylated low density lipoprotein-Sepharose affinity column chromatography. The purified LTP exhibited a single main band on sodium dodecyl sulfate-polyacrylamide gel electrophoresis. However, upon isoelectric focusing on polyacrylamide gel, the preparations consistently showed nine bands with isoelectric points ranging from 4.6 to 5.4. The treatment of LTP with Clostridium perfringens neuraminidase shifted these multiple bands toward higher pH regions due to the release of sialic acid. Extensive treatment with neuraminidase resulted in the appearance of a major band with the isoelectric point of 5.6. The purified LTP was rapidly inactivated upon incubation at 37 degrees C due to the denaturation at the "air"-water interface. Various factors promoting or preventing this interfacial denaturation were elucidated. When purified LTP was stored at 4 degrees C, plasma neuraminidase co-purified with LTP became activated, resulting in the gradual desialylation of LTP. It seemed that the LTP preparations of apparent homogeneity are associated with a trace amount of an inactive form of plasma neuraminidase. The inclusion of 4 mM 2-mercaptoethanol or 0.2% EDTA in the storage media completely prevented the activation of plasma neuraminidase. These agents, however, did not significantly inhibit the already activated neuraminidase. When LTP was stored at -20 degrees C in very low ionic strength media, such as 0.001% EDTA (pH 7.4) and at high protein concentrations, the loss of the activity was minimal even after prolonged storage.     

Sequence and structure alignment of paramyxovirus hemagglutinin-neuraminidase with influenza virus neuraminidase.

A model is proposed for the three-dimensional structure of the paramyxovirus hemagglutinin-neuraminidase (HN) protein. The model is broadly similar to the structure of the influenza virus neuraminidase and is based on the identification of invariant amino acids among HN sequences which have counterparts in the enzyme-active center of influenza virus neuraminidase. The influenza virus enzyme-active site is constructed from strain-invariant functional and framework residues, but in this model of HN, it is primarily the functional residues, i.e., those that make direct contact with the substrate sialic acid, which have identical counterparts in neuraminidase. The framework residues of the active site are different in HN and in neuraminidase and appear to be less strictly conserved within HN sequences than within neuraminidase sequences.     

Characterization of influenza virus neuraminidase with hemagglutinin activity and its comparison with that of viral neuraminidase

The neuraminidase associated with the bifunctional protein, hemagglutinin-neuraminidase, of influenza virus has been characterized. The enzyme has a pH optimum of 4.5, does not require Ca2+ and is inactivated (98%) by incubation at 50 degrees C. The enzyme has a Km of 2.00 X 10(-3) M and 0.06 X 10(-3) M with the substrates 2-(3-methoxyphenyl)-N-acetylneuraminic acid and fetuin, respectively. The Ki is 400 X 10(-6) with the inhibitor 2-deoxy-2,3-dehydro-N-acetylneuraminic acid. The incorporation of labeled cysteine, valine and leucine in the hemagglutinin-neuraminidase protein is different from that of viral neuraminidase. A comparison of the properties of the neuraminidase associated with protein hemagglutinin-neuraminidase with that of viral neuraminidase or sialidase showed that the former is biochemically different and an antigenically distinct enzyme. The unique feature of the new enzyme is that it has the hemagglutinin activity as well. The two biological activities could not be separated from each other in all systems used. Apparently, protein hemagglutinin-neuraminidase is genetically transferable and it is detectable in a laboratory recombinant virus E-2971 (H3 Aichi X N7). These results suggest that protein hemagglutinin-neuraminidase is a unique surface protein of the influenza virus A/Aichi/2/68 (H3N2).     

The anti-influenza virus agent 4-GU-DANA (zanamivir) inhibits cell fusion mediated by human parainfluenza virus and influenza virus HA

4-GU-DANA (zanamivir) (as well as DANA and 4-AM-DANA) was found to inhibit the neuraminidase activity of human parainfluenza virus type 3 (HPF3). The viral neuraminidase activity is attributable to hemagglutinin-neuraminidase (HN), an envelope protein essential for viral attachment and for fusion mediated by the other envelope protein, F. While there is no evidence that HN's neuraminidase activity is essential for receptor binding and syncytium formation, we found that 4-GU-DANA prevented hemadsorption and fusion of persistently infected cells with uninfected cells. In plaque assays, 4-GU-DANA reduced the number (but not the area) of plaques if present only during the adsorption period and reduced plaque area (but not number) if added only after the 90-min adsorption period. 4-GU-DANA also reduced the area of plaques formed by a neuraminidase-deficient variant, confirming that its interference with cell-cell fusion is unrelated to inhibition of neuraminidase activity. The order-of-magnitude lower 50% inhibitory concentrations of 4-GU-DANA (and also DANA and 4-AM-DANA) for plaque area reduction and for inhibition in the fusion assay than for reducing plaque number or blocking hemadsorption indicate the particular efficacy of these sialic acid analogs in interfering with cell-cell fusion. In cell lines expressing influenza virus hemagglutinin (HA) as the only viral protein, we found that 4-GU-DANA had no effect on hemadsorption but did inhibit HA2b-red blood cell fusion, as judged by both lipid mixing and content mixing. Thus, 4-GU-DANA can interfere with both influenza virus- and HPF3-mediated fusion. The results indicate that (i) in HPF3, 4-GU-DANA and its analogs have an affinity not only for the neuraminidase active site of HN but also for sites important for receptor binding and cell fusion and (ii) sialic acid-based inhibitors of influenza virus neuraminidase can also exert a direct, negative effect on the fusogenic function of the other envelope protein, HA.