Pattern Recognition Molecule Mindin Promotes Intranasal Clearance of Influenza Viruses¹

Influenza A virus infection causes serious morbidity and mortality in nonpandemic years (1). Recent outbreaks of human infection by avian influenza virus A raise the possibility of a pending pandemic and point to an urgent need for the development of various strategies to encounter influenza virus infection (2, 3). Influenza virus is an enveloped negative-stranded RNA virus belonging to the Orthomyxoviridae family (4). The envelope surface protein hemagglutinin mediates virus attachment to host cells during infection and another surface protein, neuraminidase, is involved in cleaving sialic acid to release virus (5, 6). Due to antigenic shift and drift by influenza virus, prior immunization with influenza vaccines is often rendered ineffective. Thus, innate immunity provides critical protection against influenza virus infection in the first 6–7 days before adaptive immunity is developed to a new viral strain (7).

Previous studies have demonstrated that various elements of the innate immune system play roles in protection against influenza virus infection (7). For example, the collectin family of mammalian C-type lectins including surfactant protein D, surfactant protein A, mannose-binding protein, and conglutinin can bind to influenza virus, act as opsonins, and inhibit influenza virus hemagglutination (HA) activity (8, 9, 10, 11, 12). Defensins can direct inactivate influenza A virus (13) and the θ-defensins retrocyclin 2 inhibit influenza virus fusion and entry through cross-linking host cell membrane glycoproteins (13, 14). It remains to be determined whether other innate immune elements also participate in immune protection against influenza virus infection.

Mindin is a highly conserved extracellular matrix (ECM)³ protein belonging to the mindin-F-spondin family (15), which includes Drosophila M-spondin (16), zebra fish F-spondin, zebra fish mindin 1 and mindin 2 (17), rat, mouse, and human F-spondin and mindin (18, 19, 20, 21). At the amino acid level, mouse mindin demonstrates 97, 85, and 60% identity with rat, human, or zebra fish mindin, respectively (20). All mindin-F-spondin family molecules have F-spondin domain 1, F-spondin domain 2, and thrombospondin type 1 repeats. Although ECM molecules were classically thought to serve as colonization sites for pathogens during infection (22, 23), these molecules can also function as pattern recognition receptors to activate innate immune responses. We have demonstrated that mindin functions as a pattern recognition receptor during bacterial infection and is essential for the initiation of innate immune response to bacteria (20). Besides its role as a pattern recognition molecule for bacterial pathogens, mindin also serves as a ligand for integrins and is critical for recruiting neutrophils and macrophages to inflamed sites (24). Although it is clear that mindin plays a critical role in the innate response to bacterial infection, its role in viral infection has not been determined.

In this study, we examined the role of mindin in influenza virus infection. We show that mice lacking mindin exhibit defective clearance of influenza virus infection. Mindin directly binds to influenza virus particle. Furthermore, recombinant mindin can promote viral clearance in wild-type mice. Collectively, our data demonstrate that mindin plays a critical role in intranasal clearance of influenza virus infection and suggest that mindin may be used as an immune-enhancing agent in influenza infection.

Madin-Darby canine kidney (MDCK) cells were obtained from American Type Culture Collection and cultured in 5% FCS-DMEM with 50 μg/ml gentamicin, 100 U/ml penicillin, and 100 μg/ml streptomycin. The following are the sources of the listed reagents: chicken RBC (cRBC; Charles River Laboratories); mouse IL-6 and TNF-α ELISA detection kits (eBioscience); H1N1 Ab (QED Bioscience); HRP-labeled anti-mouse Ab (Jackson ImmunoResearch Laboratories); tetramethylbenzidine peroxidase enzyme immunoassay substrate kit (Bio-Rad); FITC, glucose, sucrose and mannose, fibronectin, low endotoxin human serum albumin (HSA); tert-butanol, cholesterol and rhodamine-avidin (Sigma-Aldrich); and 1,2-dimyristoyl-3-trimethylammonium propane (DMTAP; Avanti Polar Lipids).

Mindin-deficient (mindin^−/−) (20) and wild-type (mindin^+/+) control mice in a C57BL/6 background were maintained in a specific pathogen-free facility at the Duke University Medical Center Vivarium (Durham, NC). All animal experiments were performed according to protocols approved by the Duke University Institutional Animal Care and Use Committee. Influenza virus H1N1 strain A/PR/8/34 (PR8) and H3N2 strain Hx31 were provided by Dr. D. J. Topham (University of Rochester, Rochester, NY) and Dr. T. Ross (University of Pittsburgh, Pittsburgh, PA), respectively. Five- to 8-wk old sex- and age-matched mindin^−/− and mindin^+/+ mice were anesthetized with ketamine hydrochloride (100 mg/kg) and xylazine (3 mg/kg) by i.p. injection. For intranasal infection, anesthetized mindin^−/− and mindin^+/+ control mice were held on their back. The pipette tip was pointed to the mouse nostril and 1.36 × 10³ (except as indicated otherwise in the figure legends) 50% tissue culture infection dose (TCID₅₀) H1N1 influenza virus fluids in a 20-μl volume were absorbed into the nasal cavity. At day 5 after infection, the lungs of the infected mice were removed and washed with PBS once, and then the lung homogenates were prepared in 0.5 ml of PBS. The homogenates were centrifuged at 6000 rpm in a microcentrifuge at 4°C for 5 min. The virus titers in lung homogenate were measured as described below. For intratracheal infection, the trachea of an anesthetized mouse was exposed and 1.19 × 10³ TCID₅₀ H1N1 in 50 μl was inoculated. At day 5, the virus titers in lung homogenate were determined.

Nasal wash fluids were collected at 18 h after intranasal infection with 1.19 × 10⁵ TCID₅₀ H1N1. Nasal wash was performed according to a described method (25). Briefly, mice were euthanized with CO₂ suffocation. The lower jaw of mice was excised. A 27-gauge needle on a 1-ml syringe with PBS was inserted into the posterior of the nasopharynx and 0.5 ml of cold PBS was injected into the nasal cavity. The wash fluids were collected from the nostrils into a 1.5-ml Eppendorf tube. The nasal wash was repeated three times for each mouse with the same fluids.

Influenza virus titers expressed as TCID₅₀ were detected using MDCK cells according to a standard protocol (26). Briefly, 2 × 10⁴ MDCK cells/well in 100 μl were seeded in a U-bottom cell culture plate. The virus samples were 1/10 serially diluted. Each diluted sample was used to infect five wells by incubating 100 μl of the sample with MDCK cells overnight. The supernatant was replaced with fresh medium containing 0.0002% trypsin. At day 5, the supernatants from all wells were transferred to a V-bottom plate. cRBC (50 μl at 0.5%) were added and incubated for 1 h at 4°C. The agglutination wells were counted and TCID₅₀ were calculated.

Sex- and age-matched mindin-deficient and wild-type mice were sacrificed to expose the trachea. A small hole was cut below the hyoid bone. A plastic tube connected with a 3-ml syringe, which contained cold PBS, was inserted into the trachea and the syringe was gently pushed to inflate the lungs with PBS. Then PBS was collected by pulling the plunger of the syringe. The lavage process was repeated three times. The bronchoalveolar lavage fluids were centrifuged to collect alveolar macrophages. To label H1N1 virus particles, H1N1 virus particles harvested from infected chicken egg embryos were purified using continuous sucrose gradient (20–60%) and ultracentrifugation. Purified H1N1 viruses were labeled with FITC as described by others (27). Purified H1N1 virus particles (0.5 ml) were mixed with 50 μl of 1 mg/ml FITC (Sigma-Aldrich) in 1 M sodium carbonate buffer (pH 9.6) and incubated and rotated at 4°C for 1 h in the dark. The mixture was dialyzed against PBS for 18 h at 4°C. Mindin-deficient and wild-type control alveolar macrophages were incubated with H1N1-FITC for 10 min at 37°C and treated with 40 μl of 0.1% trypan blue to quench extracellular florescence. After washing, the macrophages were fixed in 1% paraformaldehyde and analyzed by flow cytometry.

Macrophages were seeded in a 48-well-plate at 5 × 10⁵ cells/well and infected with purified H1N1 (1 × 10⁸ TCID₅₀/ml) in 100 μl of PBS at 37°C for 1 h. The cells were cultured in complete RPMI 1640 medium for another 18 h at 37°C. Supernatants were collected for IL-6 and TNF-α production. All cytokines were detected with ELISA kits.

The binding of H1N1 influenza viruses to the mindin matrix was analyzed using ELISA. Recombinant mindin generated as described previously (20) (100 μl/well at 10 μg/ml) was coated on ELISA plates in coating buffer at 4°C overnight. After washing three times, the coated plate was blocked with 3% BSA in PBS for 1 h at room temperature (RT). Purified H1N1 influenza virus particles were mixed with 10 mM EDTA, glucose, sucrose, or mannose and then add to the coated plate. The plate was incubated for another hour at RT. After washing, mouse anti-H1N1 Ab was added to the plate and incubated for 1 h. The HRP-labeled anti-mouse Ig was added, followed by tetramethylbenzidine substrate to visualize the reaction.

H1N1 influenza virus particles were 1/2 serially diluted in PBS with 5 μg/ml recombinant mindin, fibronectin, or BSA. One hundred microliters of each diluted virus particles was added to a V-bottom plate and mixed with 50 μl of 0.5% cRBC. The plate was incubated for 2 h at 4°C. The highest dilutions of agglutinated wells were recorded as HA units.

Nasal tissue frozen sections of mindin^−/− and mindin^+/+ mice were prepared. After blocking with 3% BSA-PBS, the sections were incubated with biotin-labeled mouse anti-mindin mAb (20) at RT for 30 min. After washing, the sections were incubated with rhodamine-avidin for another 30 min. The sections were examined under a fluorescence microscope (Zeiss Axiovert 200M).

Nasal septa were gently removed from sex- and age-matched mindin^−/− and mindin^+/+ mice and fixed in 4% glutaraldehyde in 0.1 M sodium cacodylate buffer for at least 4 h. The samples were rinsed in 0.1 M cacodylate buffer containing 7.5% sucrose three times for 15 min each and fixed in 1% osmium in cacodylate buffer for 1 h. After being washed three times in 0.11 M veronal acetate buffer for 15 min each, the samples were incubated with 0.5% uranyl acetate in veronal acetate buffer for 1 h at RT. Specimens were then dehydrated in an ascending series of ethanol (35, 70, 95% and two changes of 100%) for 10 min each, followed by two changes of propylene oxide for 5 min each. The samples were incubated with a 1:1 mixture of 100% resin and propylene oxide for 1 h, followed by two changes of 100% resin, each for 30 min. Finally, the samples were embedded in resin and polymerized at 60°C overnight. Thick sections (0.5 μm) were cut and stained with toluidine blue for light microscopy selection of the appropriate area for ultrathin sections. Thin sections (60–90 nm) were cut, mounted on copper grids, and poststained with uranyl acetate and lead citrate. Micrographs were taken with a Philips LS 410 electron microscope.

We first generated liposome-encapsulated mindin and control HSA according to a published protocol (28). Equal molar ratios of DMTAP and cholesterol were dissolved in tert-butanol at a final concentration of 22.5 mg/ml. The lipid solution was frozen in −80°C for 30 min and then dried in vacuum pump overnight to form lipid films. The lyophilized DMTAP/cholesterol was stored in −80°C before using. When protein liposome was prepared, the DMTAP/cholesterol film were rehydrated in PBS and warmed in a 40°C water bath to dissolve completely. Equal volumes of purified recombinant mindin protein (or control protein HSA) and DMTAP/cholesterol solution were mixed vigorously to form protein-liposome. The weight ratio of lipid:protein is 300:1. Wild-type mice were anesthetized. Liposome-mindin, liposome-HSA, or liposome alone (5 μg of protein in a 10-μl volume/nostril) was inoculated intranasally into mouse nostrils. Twenty minutes after liposome-protein treatment, mice were intranasally infected with H1N1 (TCID₅₀, 9.48 × 10²). The virus titers in lung homogenates were detected at day 5 after infection.

Statistical significance was analyzed with the two-tailed Student t test. Values of p < 0.05 were considered significantly different between comparing groups.

The extracellular matrix protein mindin functions as a pattern recognition molecule and is essential for the initiation of innate immune responses to bacterial infection in vivo (20). Given that mindin is expressed alveolar macrophages in lungs (20) and airway epithelial cells (see below), we wondered whether mindin also plays a role in innate immune responses to influenza virus infection. Mindin-deficient and wild-type control mice were intranasally infected with influenza virus H1N1. At day 5 after infection, we determined the virus titers in the lung homogenates of mindin^−/− and control mice. The virus titers in the lungs of mindin-deficient mice were >10-fold higher than those in control mice (Fig. 1 A). This result suggests that mindin is critically involved in clearance of influenza virus during innate immune responses after intranasal infection.

During infection, intranasally delivered influenza virus particles travel through the upper respiratory tract that includes the nasal cavity and the lower respiratory tract before reaching the lungs. To further define the regions in which mindin plays a role in influenza innate immunity, we examined the virus titers in nasal cavities of infected mindin-deficient and control mice. As shown in Fig. 1 B, the virus titers in nasal wash fluids of intranasally infected mindin-deficient mice 18 h after infection were ∼10-fold higher than those in control mice. These results demonstrate that the clearance of influenza viruses in the nasal cavity of mindin-deficient mice is defective.

A high level of influenza virus and associated inflammation may cause enhanced mortality in infected hosts, as has recently been demonstrated in avian influenza-infected patients (29). We next determined the survival of mindin-deficient mice after intranasal influenza virus infection. Mindin-deficient and control mice were infected and monitored for 14 days. Surprisingly, we did not observe an obvious difference in the survival of mindin-deficient and control mice (Fig. 1,C). Given that the influenza virus titer in the lungs of infected mindin-deficient mice was >10-fold higher than that in control mice (Fig. 1 A), the normal survival of mindin-deficient mice suggests that mindin may regulate other aspects of host innate immune responses to influenza infection (see below).

Influenza virus naturally infects the upper respiratory tract. However, if the infection is not cleared or restricted, influenza can spread to the lower respiratory tract and lungs. Given that mindin is highly expressed in the lung (20), the dramatically higher virus titers in the lungs of mindin-deficient mice could result from defective viral clearance in both the nasal cavity and lung of these animals. To test this, we intratracheally infected mindin-deficient and control mice with influenza virus. In contrast to the higher virus titers in the nasal cavity and lung of mindin-deficient mice after intranasal infection, the virus titers in the lungs of mindin-deficient mice were comparable to those in control mice after intratracheal infection (Fig. 2,A). We then examined the survival of mindin-deficient and control mice after intratracheal infection. Surprisingly, the survival of mindin-deficient mice after intratracheal infection was significantly better than that of control mice (Fig. 2 B).

The similar lung virus titers yet distinct survival patterns in mindin-deficient and control mice after intratracheal infection further support the idea that mindin also regulates other aspects of immune responses to influenza infection. Thus, we measured the amount of the proinflammatory cytokines IL-6 and TNF-α in lung homogenates of these mice at day 5 after intratracheal infection. The levels of IL-6 and TNF-α in mindin-deficient mice were reduced by 85 and 50%, respectively, when compared with those in control mice (Fig. 2 C).

Macrophages are the major cells that produce proinflammatory cytokines during microbial infections (30). We first examined influenza virus-induced TNF-α production by alveolar macrophages from mindin-deficient and control mice. Mindin-deficient alveolar macrophages produced only 10% of TNF-α when compared with control cells after in vitro infection (Fig. 2,D). Furthermore, bone marrow-derived macrophages from mindin-deficient mice also produced lower levels of IL-6 and TNF-α than control macrophages after influenza virus infection in vitro (Fig. 2,E). The reduced levels of IL-6 and TNF-α in the lungs of infected mindin-deficient mice were not due to any impairment in the development of alveolar macrophages as the number of resident macrophages in the lungs of mindin-deficient mice was comparable to that in control mice (Fig. 2 F). Together, these results suggest that mindin is required for efficient activation of macrophages upon influenza virus infection. The reduced production of proinflammatory cytokines may result in less tissue damage and enhanced survival of mindin-deficient mice.

Influenza virus infection of host cells is mediated through interactions between viral surface protein HA and sialic acid on cell surface glycosylated proteins (5, 6). Our previous data show that mindin is a pattern recognition molecule that directly binds to various bacteria through carbohydrate recognition (20). Because mindin also plays an important role in influenza infection, we examined whether mindin directly interacts with influenza virus particles. Highly purified influenza virus particles (H1N1) were incubated with plate-bound mindin or BSA and detected by anti-influenza virus Ab. Influenza viruses readily bound to coated mindin but not BSA (Fig. 3,A). In contrast to mindin-bacteria interactions (20), the mindin-virus interaction was not inhibited by different monosaccharides or EDTA (Fig. 3 A), suggesting that this mindin-virus interaction is not mediated through carbohydrate recognition and does not depend on calcium.

To further assess the interaction between mindin and influenza virus, we determined the effect of recombinant mindin on virus-induced agglutination of cRBC. HA protein on influenza virus particles binds with sialic acid on the surface of cRBC and agglutinates cRBC (31). As shown in Fig. 3,B, mindin protein dramatically inhibited influenza virus-induced agglutination of cRBC while another ECM protein, fibronectin, and BSA had no obvious inhibitory effect (Fig. 3,B). This result further supports the idea that mindin interacts with influenza virus particles. To determine whether mindin interacts with other subtypes of influenza A virus, we incubated Hx31 viral particles (H3N2) with coated mindin, fibronectin, and laminin (Fig. 3,C). Mindin binds equally well to both the H1N1 and H3N2 strains, whereas the other two ECM proteins did not exhibit specific binding to these viral particles (Fig. 3 C). Together, these data demonstrate that mindin specifically interacts with influenza viruses and suggest that the interaction between mindin and influenza may account for the observed defective viral clearance in intranasal infection and defective proinflammatory cytokine production in intratracheal infection.

Since virus clearance in the nasal cavity is defective in mindin-deficient mice, we next examined whether mindin is expressed in nasal tissues by immunofluorescence staining. As shown in Fig. 4 A, mindin is readily detected in the epithelial layer and submucosal tissues of wild-type but not mindin-deficient mice.

The epithelial layer in the nasal cavity plays a critical role in protective mucosal immunity (32). To rule out that the defective nasal viral clearance in mindin-deficient mice is due to impaired development of nasal epithelium, we examined nasal epithelium in these animals. The epithelial cilia and epithelium layer in the nasal cavity of mindin^−/− mice and mindin^+/+ mice were similar in transmission electron microscope analysis (Fig. 4 B), indicating that mindin deficiency does not obviously affect the development of epithelial cilia and epithelial layer in the nasal cavity.

Influenza virus particles bind to sialic acid on the cell surface and then enter cells through endocytosis (5, 6). Given that mindin interacts with influenza virus particles, we examined whether endocytosis of influenza virus by macrophages depends on mindin. Purified influenza virus particles were labeled with FITC and incubated with alveolar or peritoneal macrophages from mindin-deficient and control mice. After quenching the cell surface-bound influenza virus particles, endocytosis of FITC-labeled influenza virus particles by these macrophages was analyzed by flow cytometry. Endocytosis of influenza virus particles by mindin-deficient macrophages was comparable to that in control cells (Fig. 4 C), indicating that the endocytosis of influenza viruses in macrophages does not depend on mindin.

The defective intranasal clearance of influenza virus in mindin-deficient mice suggested that mindin might be used as an immune-enhancing agent to boost innate responses to influenza infection. To test this, we inoculated liposome-encapsulated mindin or a control protein, HSA, into wild-type mice intranasally. Fifteen minutes later, we intranasally infected these mice with influenza viruses. Five days later, we examined influenza virus titers in the lungs of these mice. As shown in Fig. 5,A, mice receiving the carrier liposome or HSA-liposome had 10-fold higher virus titers in their lungs than those receiving recombinant mindin-liposome. These results demonstrate that mindin can promote viral clearance intranasally in innate responses to influenza virus infection. Because the above data from mindin-deficient mice suggested that endogenous mindin in the respiratory tract and lung may regulate multiple aspects of immune responses to influenza infection, we further examined the effect of exogenously administered rMindin on weight loss and survival of wild-type hosts after influenza virus infection. Mice treated with intranasally administered rMindin did not exhibit a significant weight loss while controls suffered a significant weight loss (Fig. 5,B). Furthermore, rMindin-treated mice survived significantly better than control-treated mice (Fig. 5 C). Taken together, these results suggest that rMindin may be used as an innate-enhancing agent against influenza virus infection.

Although mindin has been identified as a novel pattern recognition molecule for bacterial pathogens, its role in viral infection was not established. Results in this study demonstrate that mindin also plays a critical role in host innate defense to influenza virus infection. Two important points have emerged from our study. First, mindin promotes influenza virus clearance in the nasal cavity. Second, efficient activation of macrophages by influenza virus depends on mindin.

The natural infection route for influenza virus is through the upper respiratory tract, starting with the nasal cavity, which includes the nasal sinuses and the nasopharynx. If the virus cannot be cleared in the nasal cavity, it will spread to the lower respiratory tract, which begins at the larynx and continues to the trachea, and reach the lung of the hosts. A layer of epithelial cells constitutes the luminal surface of the airway and protects from infection by inhaled microbial organisms (32). Several lines of evidence suggest that the dramatically elevated virus titers in the lungs of intranasally infected mindin-deficient mice are due to defective clearance of viruses in the nasal cavity. First, influenza virus titers in the nasal cavity of mindin-deficient mice were >10-fold higher than those in control mice. Second, intrachacheal infection of mindin-deficient mice did not result in enhanced influenza viral load in the lungs of these mice. Third, intranasal administration of recombinant mindin enhanced influenza virus clearance and survival in wild-type mice. The role of mindin in intranasal clearance of influenza virus is further supported by its expression in nasal tissues. The precise mechanisms by which mindin mediates influenza virus clearance in the nasal cavity remain to be established. Our transmission electron microscope analysis of nasal tissues revealed an apparently normal epithelial cilia and epithelium layers, ruling out a role of mindin in nasal epithelial cell development.

As an ECM protein, mindin may sequester influenza virus particles by directly binding to them. This is supported by our data showing that mindin can specifically bind to different subtypes of influenza virus particles and inhibit virus-induced agglutination of cRBCs. The binding of influenza virus particles by mindin may result in more efficient interaction and activation of airway epithelial cells given that mindin also functions as an integrin ligand (24, 33). Thus, mindin may interact with integrins expressed on nasal epithelial cilia and/or epithelial cells and regulate their activation in response to influenza infection. As airway epithelial cells not only provide a physical barrier against pathogens but also secret various chemical antimicrobial agents, including secretory leukoprotease inhibitor, uric acid, peroxidase, NO, and β-defensins (reviewed in Ref. 32). Importantly, proinflammatory cytokines can also be induced in airway epithelial cells (32). Our observation that alveolar and bone marrow macrophages from mindin-deficient mice exhibit defective production of IL-6 and TNF-α upon influenza virus infection is consistent with the notion that airway epithelial cells in the nasal cavity of mindin-deficient mice cannot be activated to produce antimicrobial agents. It should be noted that although nasal-associated lymphoid tissue (NALT) is involved in intranasal responses to virus infection (34, 35), it is unlikely that the role of mindin in intranasal clearance of influenza virus is limited to the NALT. This is due to the fact that mindin expression in nasal tissues is not limited to NALT and the removal of NALT does not affect the development of protective immunity and viral clearance in the upper respiratory tract upon influenza infection (36). Additional experiments are needed to determine the activation of airway epithelial cells from mindin-deficient mice to influenza virus infection.

An important finding from our results is that recombinant mindin can promote clearance of intranasally infected influenza viruses and enhance the survival of wild-type mice. We have used liposome-encapsulated recombinant mindin to enhance its stability in the nasal cavity (28). The constant changes in the antigenicity of influenza virus surface proteins HA and neuraminidase require repeated immunizations with new strains of influenza viruses. In the case of a pandemic such as an avian influenza outbreak, human populations will completely lack protective adaptive immunity (2). Innate immunity against avian influenza virus infection will be critical to protect individuals during the initial 6–7 days of infection. Our results suggest that mindin may be used as an immune- enhancing agent to boost host innate immunity against influenza virus infection. The ∼10-fold reduction in influenza virus titer upon mindin administration may be sufficient to improve host overall response to virus infection. Thus, it will be important in future investigations to optimize the regimen of mindin administration in influenza virus infection.

The comparable survival rates between mindin-deficient and control mice after intranasal infection were surprising given that mindin-deficient mice had much higher influenza virus levels in their lungs. This is contradictory to a recent study showing a correlation between high virus titers and high mortality among avian influenza-infected patients (29). This result suggests that mindin regulates other aspect of host responses to influenza virus infection. For example, endogenous mindin may enhance lung airway responses during inflammation and lack of mindin may be beneficial for their survival. Indeed, intratracheal infection of mindin-deficient mice with influenza viruses resulted in comparable virus titers to those in control mice and a significantly better survival than control mice. The distinct death patterns in intranasally and intratracheally infected mindin-deficient mice are correlated well with the viral titers in the lungs of these mice. Furthermore, mindin regulates the severity of allergic airway inflammation as mindin-deficient mice have less severe allergic airway disease with fewer eosinophils (Z. Li, S. Garantziotis, W. Jia, E. Potts, S. Lalani, Z. Liu, Y.-W. He, and J. W. Hollingsworth, unpublished observation). Thus, the overall improved survival of mindin-deficient mice may be due to a lowered production of inflammatory cytokines including IL-6 and TNF-α as well as less airway and lung damages during influenza virus infection. This is consistent with the defective activation of both alveolar and bone marrow macrophages from mindin-deficient mice upon influenza virus infection in vitro.

How does mindin mediate influenza virus-induced activation of innate cells? The interaction of mindin and influenza virus particles suggests that mindin may function as an intermediate for efficient interaction between viral particles and macrophages. Mindin interaction with bacterial pathogens is inhibited by monosacchrides and EDTA. In contrast, mindin interaction with influenza viruses is not affected by these monosacchrides and EDTA, suggesting that the interaction is distinct from that between mindin and bacteria. Furthermore, mindin inhibits cRBC HA induced by influenza virus, suggesting that the interaction affects HA binding to sialic acid. Our recent data show that mindin functions as a ligand for integrins CD18, CD11b,and CD49d on neutrophil and CD29, CD49d, and CD49e on dendritic cells (24, 33). Mindin may provide a costimulatory signal by activating integrins or other unidentified receptors on innate immune cells upon influenza virus recognition. Further investigation of the roles of integrins in influenza virus infection may provide important insights into the innate mechanisms of host immunity to this virus.

We thank Januario C. Estrada and Dr. Sara E. Miller at the Electron Microscope laboratory (Department of Pathology, Duke University Medical Center) for professional expert assistance with the electron microscope sample preparation and photography, and Dr. Nu Zhang, Dr. Ivan Dzhagalov, and Heather H. Pua for critical reading of this manuscript.

Duke University plans to file a provisional patent application on the role of mindin as an innate immune enhancing agent.