Search

Bacterial superinfections are a major cause of morbidity and mortality during influenza A virus (IAV) epidemics. Depression of phagocyte functions resulting from attachment of the IAV hemagglutinin (HA) to cell surface sialo-glycoproteins is a likely contributory cause of these infections. We have proposed that the group of collagenous lectins (termed collectins) present in blood and pulmonary surfactant play a role in initial host defense against IAV. We used here several recombinant human surfactant protein D (RhSP-D) preparations to determine the mechanism through which opsonization of IAV with collectins protects neutrophils against the deactivating effects of IAV on cellular respiratory burst responses in vitro. RhSP-D was markedly more potent than antibodies that inhibited viral hemagglutination activity (anti-HA antibodies) at protecting neutrophils in this assay. Unlike the anti-HA antibodies, RhSP-D was protective at concentrations that minimally inhibited viral hemagglutination activity. Two related features of SP-D--the degree of multimerization and the ability to cause aggregation of IAV particles--were critical determinants of the ability of SP-D to protect neutrophils against deactivation. Similarly SP-D-induced viral aggregate formation resulted in enhanced IAV binding to neutrophils and potentiated the ability of the virus itself to trigger neutrophil respiratory burst responses. In contrast to the case of IAV-antibody complexes, SP-D-IAV complexes attached to and activated neutrophils through a neuraminidase-sensitive mechanism (ie, similar to unopsonized IAV). These results indicate that collectin-mediated viral aggregation per se may be an important host defense mechanism not only by virtue of reducing the number of infectious viral particles, but also by promoting phagocyte responsiveness.

Bacterial superinfections are the most common cause of mortality during influenza epidemics. Depression of phagocyte functions by influenza A viruses (IAVs) is a likely contributory cause of such infections. We used an in vitro model of viral depression of neutrophil respiratory burst responses to FMLP and PMA to examine the mechanism of IAV-induced phagocyte deactivation. Respiratory burst responses or intracellular calcium mobilization were triggered by the virus itself, but these were not causally related to deactivation. By treating neutrophils with neuraminidase, and by use of purified IAV hemagglutinin (HA) preparations, cross-linking of sialic acid-bearing neutrophil surface components by the IAV HA was shown to be responsible for deactivation. IAV competed for binding to neutrophils with Abs directed against CD43, sialyl-Le(x), CD45, and gangliosides. Deactivation could be reproduced by treating neutrophils with anti-CD43 or -sialyl-Le(x) Abs in the absence of IAV. However, treatment of neutrophils with elastase markedly reduced CD43 expression, without affecting overall IAV binding or the ability of IAV to cause deactivation. Hence, although IAV binding to CD43 can account for deactivation, other IAV-binding proteins exist (e.g., those bearing sialyl-Le(x)) that can independently mediate functional depression.

We tested the hypothesis that pulmonary surfactant-associated lectins--surfactant proteins A and D (SP-A, and -D)--contribute to initial protective mechanisms against influenza A viruses (IAVs). SP-D potently inhibited hemagglutination activity of several strains of IAV as well as causing viral aggregation. SP-D enhanced neutrophil binding of IAV and neutrophil respiratory burst responses to the virus. Neutrophil dysfunction resulting from IAV exposure was diminished when the virus was pre-incubated with SP-D. Each of these effects was mediated by the calcium-dependent carbohydrate-binding property of SP-D. Native SP-D preparations of both human and rat origin, as well as recombinant rat SP-D, had similar activity. SP-A also inhibited IAV hemagglutination activity. We have previously reported that related mammalian serum lectins (mannose-binding lectin [MBL] and conglutinin) have similar effects. SP-D was at least 10-fold more potent at causing hemagglutination inhibition than were SP-A or MBL. SP-D was shown to contribute to potent anti-IAV activity of human bronchoalveolar lavage fluid. These results suggest that SP-D--alone, and in conjunction with SP-A and phagocytic cells--constitutes an important component of the natural immune response to IAV infection within the respiratory tract.

We have studied in detail the in vitro interactions of influenza A viruses (IAVs) with human neutrophils to clarify why these cells become dysfunctional during IAV infection. Unosponized IAV elicited a respiratory burst response in neutrophils which, like that triggered by formylmethionyl-leucyl-phenylalanine (fMLP), involved mediation of signal-transducing GTP-binding proteins and tyrosine kinase activation. The IAV-induced response differed from that provoked by fMLP in that H2O2 was produced without concomitant O2- release. IAV also did not cause extracellular release of granule enzymes in cytochalasin B-treated neutrophils. Using chemiluminescence assays, the respiratory burst response to IAV was found to occur at an intracellular location. These findings may, in part, explain the anomalous nature of the respiratory burst response elicited by IAV and suggest strategies for determining the mechanism of IAV-induced neutrophil deactivation.

Since the 1940's, non-Ig inhibitors of influenza A virus (IAV) hemagglutination activity and infectivity have been recognized in mammalian sera. Recently, the heat labile (beta) inhibitor of this type was identified by indirect methods as the lectin, conglutinin. In support of this hypothesis, we found that purified conglutinin strongly inhibited hemagglutination activity and infectivity of IAV. By using IAV strains with specific variations in glycosylation of the hemagglutinin molecule, we showed these effects to be mediated by binding of conglutinin to high mannose carbohydrate attachments on the viral hemagglutinin. Through the same mechanism conglutinin caused aggregation of IAV particles. Human neutrophils produce hydrogen peroxide upon exposure to IAV. Also, after a brief period of exposure to IAV, neutrophils exhibit depressed responsiveness (deactivation) upon exposure to other stimuli (e.g., chemotactic peptides). These phenomena may be related to the in vivo inflammatory response during IAV infection, and to the propensity of IAV-infected subjects to suffer bacterial superinfection. Pre-incubation of IAV with conglutinin markedly potentiated human neutrophil hydrogen peroxide production in response to the virus. This effect correlated with the ability of conglutinin to aggregate the virus. IAV treated with conglutinin also caused significantly less neutrophil deactivation than did the unopsonized virus. These enhancements of neutrophil respiratory burst responses by conglutinin were again mediated by binding of the lectin to viral carbohydrates. The mammalian C-type lectin family includes conglutinin, mannose-binding protein, and surfactant proteins A and D. These lectins may be important constituents of the initial host response to IAV, by inhibiting IAV infectivity directly, causing viral aggregation, and acting as opsonins to enhance phagocyte responses to the virus.

Influenza A viruses (IAVs) cause substantial morbidity and mortality in yearly epidemics, which result from the ability of the virus to alter the antigenicity of its envelope proteins. Despite the rapid replication of this virus and its ability to infect a wide variety of cell types, viremia is rare and the infection is generally limited to the upper respiratory tract. The preimmune host defense response against IAV is generally, therefore, successful. We have previously provided (and summarized) evidence that neutrophils contribute to defense against IAV, although neutrophil dysfunction and local tissue damage may be less salutory byproducts of this response. Here we provide evidence that the serum lectin mannose-binding protein directly inhibits hemagglutinin activity and infectivity of several strains of IAV. In addition mannose-binding protein acts as an opsonin, enhancing neutrophil reactivity against IAV. Opsonization of IAV by mannose-binding protein also protects the neutrophil from IAV-induced dysfunction. These effects are observed with physiologically relevant concentrations of mannose-binding protein. Two different allelic forms of recombinant mannose-binding protein are found to have similar effects. We believe, on the basis of these data, that mannose-binding protein alone and in conjunction with phagocytic cells is an important constituent of natural immunity (i.e., preimmune defense) against IAV.

Although neutrophils are not viewed as a principal defense against influenza A virus (IAV) infection, their interactions are both complex and clinically relevant. Activation of the neutrophil is distinctive from that described for chemoattractants. To more fully characterize the pathway by which IAV stimulates the human neutrophil, we have examined its binding characteristics. First, inhibition studies with various sialic acid-containing and sialic-free sugars showed that IAV binds to sialic acid residues and activates receptors distinct from those used by Concanavalin-A (Con-A) and formyl-methionyl-leucyl-phenylalanine (FMLP) and that overlap those bound by wheat germ agglutinin (WGA). That viral hemagglutinin (HA) mediates viral binding and activation was shown by preincubating neutrophils with purified monovalent bromelain-released HA (BHA) and showing that IAV-induced membrane depolarization and hydrogen peroxide (H2O2) production were inhibited approximately 95%. However, binding inhibition required significantly higher concentrations of purified HA, suggesting that binding and cell activation have different interactive requirements. Desialation of the neutrophil surface membrane by neuraminidase treatment resulted in a 90.6% +/- 4.4% and 53.1% +/- 8.7% inhibition of IAV activation of neutrophils and viral binding, respectively. Resialation with ganglioside GT1b totally restored viral binding, but did not reverse the inhibition of activation. Thus, although HA was shown to mediate binding and neutrophil activation, viral binding per se was insufficient to stimulate the cell. Having demonstrated the functional role of HA, we sought to establish the mechanism of stimulation. HA in three different forms (BHA, HA-rosettes, and HA-liposomes) failed to activate the cell, although H2O2 production evoked by IAV stimulation was reduced in competitive inhibition studies with each preparation. Upon cross-linking with a monoclonal antibody to HA, activation comparable to that of intact virus was observed. The requirement for cross-linking of functional receptors, as opposed to activation through the neutrophil Fc receptor, was confirmed in experiments using staphylococcal A protein. These studies have shown the chemical specificity of IAV binding to the human neutrophil, the character of the receptor(s) stimulated to activate the IAV-evoked response, and the activation requirement for cross-linking those receptors responsible for stimulating functional responses.

Influenza A virus (IAV) causes both activation and deactivation of the human neutrophil, which may, respectively, contribute to host defense against the virus and enhanced susceptibility to bacterial superinfection. We have shown that certain features of neutrophil activation by IAV are distinctive compared with activation by chemoattractants in terms of both the stoichiometry of the respiratory burst response and the signal transduction events that precede it. We here demonstrate that related myxoviruses as well as sialic acid-binding lectins elicit a respiratory burst response similar to that induced by IAV, in which hydrogen peroxide is formed with minimal accompanying superoxide generation. Brief preincubation of neutrophils with these agents fully inhibits subsequent activation by IAV, implying that they are binding to the same surface membrane components as IAV. Preincubation with Limax flavus agglutinin (LFA) does, in fact, substantially reduce binding of radiolabeled IAV to the neutrophil. This lectin, like IAV, both activates and deactivates the neutrophil. As in the case of IAV, LFA-induced activation (1) is mediated via stimulation of phospholipase C, (2) is pertussis toxin insensitive, and (3) entails a lesser contribution of calcium influx than is the case for chemoattractants.

Influenza A virus (IAV) activates the human neutrophil, but induces a dysfunctional state as well. Cell activation may contribute to the containment of the virus and/or cause local tissue damage. Certain features of the neutrophil activation response elicited by IAV are distinctive when compared with that triggered by formyl-methyl-leucyl-phenylalanine (FMLP). An atypical respiratory burst response occurs in which hydrogen peroxide, but no superoxide, is formed. This unusual respiratory burst stoichiometry persists despite marked priming of the IAV-induced response. A comprehensive examination of the activation cascade initiated by these stimuli failed to show an explanation for these differences. Both IAV and FMLP comparably stimulate inositol trisphosphate and phosphatidic acid production. The subsequent increase in intracellular calcium (Ca2+i) upon FMLP stimulation was more dependent on extracellular Ca2+ than with IAV activation, but both stimuli induced Ca2+ influx. FMLP and IAV exhibited equal susceptibility to inhibition by protein kinase inhibitors in eliciting the respiratory burst, and actin polymerization occurred in response to each agonist. A possible explanation for the anomalous respiratory burst induced by IAV is that O2- is generated at an intracellular site inaccessible to assay, and/or virus binding to sialic acid constituents of the plasma membrane alters the O2- generating capacity of the respiratory burst oxidase; evidence for each mechanism is offered.

We have previously demonstrated that influenza A virus (IAV) stimulates the human neutrophil through phospholipase C activation. With the use of the fluorescent indicator 2',7'-bis(carboxyethyl)-5,6-carboxyfluorescein (BCECF), cytoplasmic acidification and subsequent alkalinization are shown to accompany this activation. These responses are not inhibited by pertussis toxin (PT). The alkalinization is mediated largely *but not entirely) by the Na(+)-H+ antiporter and is not initiated, or modulated, by the IAV-induced cytosolic Ca2+ (Cai2+) rise. Rather, protein kinase C (PKC) is likely the mediator of cell alkalinization, based on studies using the PKC inhibitor 1-(5-isoquinolinesulfonyl)-2-methylpiperazine (H-7). The acidification can be dissociated from the alkalinization response, which is also independent of Cai2+ fluxes and of PKC. Both pHi responses can be dissociated from the respiratory burst. Cytosolic alkalinization and acidification seem to reflect two independently mediated responses of the activated neutrophil, the former resulting ultimately from phospholipase activation and the latter from other activities that are not yet fully characterized.

Natural immune reactions are mediated by lymphocytes, macrophages/monocytes, and neutrophils. The latter have been implicated in a variety of self-surveillance models, i.e., activity against malignant host cells, participation in wound repair, and infliction of damage in postischemic perfusion injury. Better characterized are the interactions with unopsonized pathogens through lectinophagocytosis mechanisms, where the lectin resides either on the phagocyte or on the microorganism. This review examines the infection by influenza A virus (IAV) of the human neutrophil, which results in the vigorous metabolic response of the cell to generate toxic oxygen species. This response is not necessarily characteristic of response to unopsonized particles, as the neutrophil exhibits no such activity to unopsonized zymosan or chlamydia. The virus elicits calcium mobilization from intracellular stores through a pertussis toxin-insensitive mechanism, and in its particulars the activation cascade is unique in comparison to any other characterized agonist. The putative receptor for the IAV binding protein, hemagglutinin (HA), contains the sialic acid residues; identification of specifically linked protein receptors will allow characterization of this stimulation pathway and will define the molecular biology of this activation sequence. Insight into this particular pathway may allow definition of a primitive recognition system that represents a fundamental basis for discernment of self and nonself entities.

Neutrophil dysfunction consequent to influenza A virus infection has been described in vivo and in vitro and may contribute to the serious bacterial sequelae which occur in influenza-infected hosts. On the premise that such dysfunction may represent a form of "deactivation," we sought to characterize neutrophil activation by the virus in comparison with other agonists. The virus induces a respiratory burst in which H2O2 (but not O2-) are formed. Preceding the respiratory burst, a rise in intracellular calcium (Ca2+i) is noted, but both responses are nearly independent of extracellular Ca2+, unlike those elicited by the other well-characterized Ca2+-dependent agonists, formyl-methyl-leucyl-phenylalanine (FMLP), or Concanavalin-A (Con-A). The Ca2+ increase is paralleled by IP3 generation, implying that it is the result of phospholipase C (PLC) activation. The virus also elicits neutrophil membrane depolarization, which is independently mediated from the Ca2+ increase and respiratory burst and may reflect protein kinase C (PK-C) activation. Virus-induced responses are insensitive to pertussis toxin (PT); cholera toxin does inhibit these responses but in a nonspecific manner. Thus, although influenza virus activates PLC in neutrophils, it does so in a PT-insensitive manner and does not elicit or require a discernible Ca2+ influx to generate a respiratory burst response. In aggregate, the data indicate that influenza A virus activates neutrophils in a manner distinct from that of other well-described neutrophil agonists. These results illustrate the diversity of neutrophil activation mechanisms and support the notion that further characterization of this pathway may facilitate understanding of neutrophil dysfunction induced by the virus.

Bacterial superinfection in influenza A virus-related illness may in part be explained by virus-induced neutrophil dysfunction. We here provide evidence that this effect is related to abnormal calcium metabolism of virus-infected cells. Neutrophils exposed to influenza virus for 0.5 h at 37 degrees C showed depressed O2- generation and release of radiolabeled arachidonic acid upon stimulation with FMLP. The peak cytosolic Ca2+ level achieved by virus-infected neutrophils after FMLP stimulation was significantly depressed as is efflux of 45Ca2+. This deficient Ca2+ mobilization could not be attributed to alterations of inositol phosphate production or Ca2+ influx in response to FMLP, both of which were unaffected by prior virus infection. Given these findings, the immediate effects of influenza virus on neutrophil Ca2+ metabolism were examined. The virus itself caused a rise in cytosolic Ca2+ and an efflux of 45Ca2+ without any corresponding 45Ca2+ influx. Total cell Ca2+ however was not depleted as measured by atomic absorption. Influenza virus, therefore, causes neutrophil activation leading to significant perturbations in Ca2+ metabolism and later to impaired mobilization of Ca2+ stores. This system offers a model for phagocyte deactivation and an opportunity to define control mechanisms of signal transduction.

Influenza A virus has the ability to activate human neutrophils and in turn reduce its subsequent oxidative, chemotactic, and degranulation responses to other stimuli, an effect that may lead to adverse clinical sequelae. Current literature on the mechanism of this effect is reviewed, and preliminary data regarding its biochemical basis presented.