|Home | About | Journals | Submit | Contact Us | Français|
It is possible to model the progression of influenza virus from the upper respiratory tract to the lower respiratory tract in the mouse using viral inoculum delivered in a restricted manner to the nose. In this model, infection with the A/Udorn/307/72 (Udorn) strain of virus results ultimately in high viral titers in both the trachea and lungs. In contrast, the A/Puerto Rico/8/34 (PR8) strain causes an infection that is almost entirely limited to the nasal passages. The factors that govern the progression of virus down the respiratory tract are not well understood. Here, we show that, while PR8 virus grows to high titers in the nose, an inhibitor present in the saliva blocks further progression of infection to the trachea and lungs and renders an otherwise lethal dose of virus completely asymptomatic. In vitro, the salivary inhibitor was capable of potent neutralization of PR8 virus and an additional 20 strains of type A virus and two type B strains that were tested. The exceptions were Udorn virus and the closely related H3N2 strains A/Port Chalmers/1/73 and A/Victoria/3/75. Characterization of the salivary inhibitor showed it to be independent of sialic acid and other carbohydrates for its function. This and other biochemical properties, together with its virus strain specificity and in vivo function, indicate that the mouse salivary inhibitor is a previously undescribed innate inhibitory molecule that may have evolved to provide pulmonary protection of the species from fatal influenza virus infection.
IMPORTANCE Influenza A virus occasionally jumps from aquatic birds, its natural host, into mammals to cause outbreaks of varying severity, including pandemics in humans. Despite the laboratory mouse being used as a model to study influenza virus pathogenesis, natural outbreaks of influenza have not been reported in the species. Here, we shed light on one mechanism that might allow mice to be protected from influenza in the wild. We show that virus deposited in the mouse upper respiratory tract will not progress to the lower respiratory tract due to the presence of a potent inhibitor of the virus in saliva. Containing inhibitor-sensitive virus to the upper respiratory tract renders an otherwise lethal infection subclinical. This knowledge sheds light on how natural inhibitors may have evolved to improve survival in this species.
Influenza A virus (IAV) is a respiratory pathogen of global significance, responsible for 250,000 to 500,000 deaths annually. Innate immunity is critical to the host in the early stages of IAV infection, restricting the escalation of viral loads prior to the induction of virus-clearing responses mediated by the adaptive immune system. As part of the innate immune system, a variety of soluble molecules are present in respiratory secretions to provide an initial barrier to infection. They include collectins, pentraxins, mucins, and salivary scavenger cysteine-rich glycoprotein 340 (gp340), all of which interact specifically with the virus to inhibit its function (1,–9).
An understanding of the mechanisms of action of these molecules has been sought with a view to exploiting their therapeutic and prophylactic potential. Collectins, which are a group of soluble mammalian proteins found in serum and pulmonary fluids, include mannose binding protein (MBP), conglutinin, and pulmonary surfactant protein D (SP-D). They neutralize virus infectivity through calcium-dependent binding of their lectin domains to carbohydrate side chains located on the head of the influenza hemagglutinin (HA) of susceptible strains (1, 10, 11) and are classified as β-inhibitors of IAV. In vivo, these molecules may directly block access to the sialic acid-containing receptors for virus on host cells, as demonstrated for SP-D (8, 9). An additional function of MBP, which resembles the complement component C1q, is in mediation of complement-dependent inactivation of virus (12). The γ-inhibitors of IAV, such as α2-macroglobulin (13, 14), act via a different mechanism. Pulmonary surfactant protein A (SP-A) mediates inhibition of IAV by calcium-independent binding of the viral HA to terminal sialic acid expressed on the SP-A molecule (15), blocking the receptor-binding site of the HA so that it can no longer access cellular receptors. Mucins, gp340, and PTX3 also inhibit influenza virus by providing decoy sialic acid ligands to which the virus binds (4, 16). In general, these soluble innate inhibitors and their mechanisms of action are well conserved in evolution.
The majority of research into the in vivo effectiveness of these inhibitors has considered their impact on the pathogenesis of influenza virus infection in the lower respiratory tract (LRT), but very little is known about innate control of viral replication in the upper respiratory tract (URT) and particularly the factors that govern progression of virus from the URT to the LRT. The study of influenza in animal models usually involves intranasal inoculation of virus in anesthetized mice, which results in virus deposition throughout the respiratory tract, and modulation of viral titers in the lungs is used as the primary endpoint. This method is not a good mimic of the natural route of infection, since large amounts of virus are rapidly introduced into the lungs and establish a vigorous infection at that site; in human disease, the infection is generally confined to the URT and large airways, while primary viral pneumonia is rare. Also, any immune effectors in the URT that may act to clear virus at that site and prevent further spread of infection are bypassed.
Though not widely used, a protocol that may more closely resemble the infection process most commonly seen in humans has been described. In this URT delivery model, the virus inoculum is restricted to the nasal epithelium of mice by administration to the external nares in a very small volume in the absence of anesthetic (17, 18). Using URT-restricted delivery, the A/Udorn/307/72 (Udorn; H3N2) strain of influenza A virus establishes an infection in the upper respiratory tract that gradually progresses to the trachea and lungs (19), whereas the A/Puerto Rico/8/34 (PR8; H1N1) strain of influenza A virus causes an infection that remains almost entirely in the nasal passages (17). In this study, we investigated the mechanism responsible for inhibiting the progression of certain IAV strains from the upper to the lower respiratory tract of mice. We show that mouse saliva contains a potent inhibitor of PR8 and other susceptible strains that impedes the descent to the lungs. This study provides insight into a novel antiviral defense mechanism that may have evolved, in this mammalian host at least, to inhibit spread of virus from the upper respiratory tract to protect the trachea and lungs.
To compare the growth kinetics of influenza viruses using the established total respiratory tract (TRT) infection model, anesthetized BALB/c mice were inoculated intranasally (i.n.) with 50 μl of suspension containing 104.5 PFU of Udorn or Mem71-Bel virus or 50 PFU of the more virulent PR8 virus, which is lethal to mice at doses as low as 100 PFU when delivered as a TRT infection. Titers of infectious virus in the lungs were then determined daily in different groups of 5 mice for 1 to 8 days postinfection. All three viruses established high-titer pulmonary infections (Fig. 1A). The kinetics of Udorn and Mem71-Bel viruses were similar; viral titers peaked at approximately 105 PFU and were maintained for the first 4 or 5 days before declining at day 5 or 6 and eventually being cleared by day 7. Despite the lower dose, PR8 virus peaked at 106 PFU in the lungs and was maintained for 6 days postinfection before declining at day 7. Low levels of virus could still be detected in the lungs of all PR8 virus-infected mice at day 8.
The growth kinetics were then compared to those in a similar experiment in which virus was delivered in a manner that restricted the initial deposition to the URT (17, 18), a protocol that may more closely resemble the initiation of the infection process most commonly seen in humans. BALB/c mice were infected i.n. with the same doses of virus described above except the volume was decreased to 10 μl and the mice were not anesthetized. Lung viral titers were determined daily in different groups of mice from days 1 to 8 as described above. After URT inoculation, Udorn virus was not present in the lungs on days 1 and 2 but had progressed down the respiratory tree to reach the site by day 3 and replicated to high titers (Fig. 1B). In stark contrast, Mem71-Bel virus showed very limited progression to the lungs and replicated to detectable titers only beyond day 5, while PR8 did not progress at all and could not be detected in the lungs of mice at any time point examined (Fig. 1B).
To investigate these differences more thoroughly and to determine whether lack of virus in the lungs of PR8 virus-infected mice stemmed from the lower dose (50 PFU) of virus used, mice were again inoculated via the URT with either PR8 or Udorn virus, using an equivalent dose of 104.5 PFU for each virus (Fig. 2). On this occasion, the nasal turbinates (Fig. 2A), trachea (Fig. 2B), and lungs (Fig. 2C) were sampled daily and assayed for the presence of infectious virus up to 8 days postinfection. Both PR8 and Udorn viruses grew to high titers in the nose, peaking at day 2 postinfection and then gradually decreasing between days 3 and 7 and more sharply on day 8. In the trachea (P < 0.0001) and lungs (P = 0.0018), however, the kinetics differed markedly between the two viruses. Udorn virus progressed to the trachea by day 1 or 2 postinfection, reaching peak viral levels similar to those seen in the nose by days 3 and 4. Udorn virus was undetectable in the lungs on day 1 postinfection, but the infection had progressed to this site by days 2 to 4 postinfection in individual mice, peaking at day 6, and was cleared from the lungs of most mice by day 7. In contrast, PR8 virus was recovered from the trachea and lungs of mice only sporadically, and titers of recovered virus were very low (Fig. 2B and andC).C). This was despite the much higher dose of PR8 virus used in this experiment, which represented 300 times the lethal dose when delivered as a TRT infection. Importantly, the PR8-infected mice remained completely active throughout the experiment with no overt signs of clinical infection.
These data indicate that following URT inoculation, PR8 virus is almost entirely limited to the nasal fossa, whereas Udorn virus is capable of causing a progressive infection to the trachea and lungs. It should be noted that the absence of pulmonary PR8 infection after URT delivery of virus was not due to a higher susceptibility of the virus to innate inhibitors in the LRT, since PR8 virus grew to high titers in the lungs after TRT infection. In addition, the inability of PR8 virus to progress to the lungs is independent of receptor specificity, as both viral strains have the capacity to infect cells bearing sialic acid in α2-3 linkage (20,–22) to galactose, which is found on the receptors for influenza virus throughout the mouse respiratory tract and is used when these viruses grow in the lungs after TRT delivery (Fig. 1). The data are compatible with the presence of an inhibitor of influenza virus in the URT that is active against PR8 but not Udorn virus.
To examine the generality of the finding across different mouse strains, a range of wild-type mice were infected with 50 PFU of PR8 virus by the TRT and URT delivery routes (Fig. 3). In all cases, the virus replicated in the trachea and to high titers in the lungs after TRT, but not after URT, inoculation. The same phenomenon was also demonstrated in SCID mice, which lack B and T cell responses, indicating that any inhibitor is likely to be a product of the innate rather than the adaptive immune system.
The experiments described above show that PR8 virus grows to high titers in the nose but the infection is held back in its spread down the respiratory tract after URT delivery. This is suggestive of an inhibitory factor that acts somewhere between the nose and the trachea of mice to prevent the progression of PR8 virus. To identify the location of the inhibitor, mouse nasal washings and saliva were tested for the ability to neutralize 30 to 100 PFU PR8 and Udorn viruses in vitro using a plaque assay in Madin-Darby canine kidney (MDCK) cells. Neither virus showed significant neutralization after incubation with mouse nasal washings (Fig. 4A) (P > 0.05), whereas incubation of PR8 virus with a 1 in 8 dilution of mouse saliva (Fig. 4B) resulted in a greater than 50% reduction in the amount of infectious virus recovered (P = 0.004). Similar treatment of Udorn virus with saliva did not result in reduction in the number of PFU recovered compared to the control treatment (P > 0.05).
To confirm the presence of a PR8 virus-specific inhibitor in murine saliva and to achieve some indication of its potency, 500 PFU of PR8 and Udorn viruses were incubated with different dilutions of saliva, and the amount of nonneutralized virus remaining was assessed by plaque formation. In this experiment, saliva was mixed with the virus at a saliva/virus ratio of 9:1 (vol/vol) to more closely mimic, at the initial dilution, the in vivo situation where the inhaled virus would contact neat saliva (Fig. 4C). While saliva could neutralize both viruses in vitro, the inhibition of PR8 virus was significantly greater than that of Udorn virus. The capacity to neutralize this amount of PR8 virus was lost after an approximately 100-fold dilution. In addition, different amounts of PR8 and Udorn viruses, ranging from 50 to 104.5 PFU, were pretreated with undiluted mouse saliva at a saliva/virus ratio of 9:1 (vol/vol). The degree of neutralization was determined by comparing the titer of recovered virus to that of mock-treated virus using the plaque assay (Fig. 5). Undiluted saliva efficiently neutralized the infectivity of even large amounts of PR8 virus (≥90%; P < 0.0001 with 104.5 PFU) but failed to provide any significant reduction of 5,000 or greater PFU of Udorn virus (P > 0.05). Smaller amounts of Udorn virus (50 and 500 PFU), however, could be reduced by about 50% in the presence of saliva (both P < 0.01). The almost total neutralization of PR8 virus by undiluted mouse saliva is in accordance with the dramatic inhibition of PR8 virus progression seen following URT infection of mice and supports the initial observation indicating that mouse saliva is the source of the inhibitory factor.
Further experimentation (data not shown) revealed that the level of PR8 virus infection of monolayers of MDCK cells pretreated with saliva was no different than that of monolayers pretreated with the control RPMI plus BSA (see Materials and Methods) (P < 0.05), indicating that saliva does not act on host cells to prevent viral infection in a manner similar to interferon or sialic acid receptor blockade agents, but rather, that a direct interaction of the inhibitory factor with the virus itself occurs.
Not all human influenza virus isolates grow well in mice, so many more viral strains could be tested using the in vitro neutralization assay. Table 1 summarizes the susceptibilities of a panel of different natural and reassortant influenza viruses to inhibition by saliva. Over 20 strains of type A virus, including an H1N1pdm09 strain, and the two type B strains tested were inhibited, with Udorn virus and the closely related A/Port Chalmers/1/73 and A/Victoria/3/75 viruses the only resistant strains identified. Selected strains from this panel that do establish infection of mice were tested, using 104.5 PFU of virus, for the ability to progress to the lungs after URT and TRT infection. The day 4 mean titers of virus (log10 PFU) in the lungs of mice infected with Mem71-Bel, Mem71-Bel/BS, Mem72-Bel, and WSN viruses by URT delivery were on average 1.40 ± 0.9, <1, 1.2 ± 0.5, and 1.2 ± 0.5, respectively (where mice with undetectable virus were scored as 1.0) and by TRT delivery were 4.81 ± 0.2, 5.11 ± 0.3, 3.03 ± 0.1, and 6.27 ± 0.1. These data confirmed the association between strong inhibition by saliva in vitro and an inability to progress to the lungs after URT inoculation, despite an ability to grow in the lungs when delivered directly to that site.
Of particular note from the virus strains analyzed in Table 1 is the fact that mouse saliva could neutralize virtually all of the in vitro infectivity of a mutant of the Mem71-Bel (H3N1) virus selected for resistance to the γ-inhibitors in horse serum, suggesting that the mode of action of the salivary inhibitor does not involve the provision of a sialic acid decoy to block the receptor-binding site on the HA, as used by this class of classical inhibitors.
To further investigate this, mouse saliva was treated with bacterial neuraminidase (receptor-destroying enzyme [RDE]) prior to in vitro assay against 1,000 PFU of PR8 and Udorn viruses. Central to our ability to successfully test various treatments of saliva, including RDE treatment, was the ability of the inhibitor to withstand lyophilization and reconstitution without significant loss of activity (P > 0.05) against PR8 virus (Fig. 6A). Likewise, the inhibitor displayed remarkable resistance to heat treatment, including 56°C for 30 min (Fig. 6B), which is necessary to inactivate the bacterial neuraminidase prior to addition of the saliva to the virus.
Figure 6C shows that treatment with RDE from Vibrio cholerae did not eliminate the significant inhibition of PR8 virus (P < 0.0001 for both untreated and RDE-treated saliva compared to the control). This was also true of treatment of saliva with RDE from Arthrobacter ureafaciens (data not shown). However, the small but significant (P < 0.001) amount of inhibition of Udorn virus seen in vitro with saliva (42%) was completely abrogated by prior treatment of saliva with RDE (P < 0.0001 for RDE treated compared to untreated saliva). These results indicated that mouse saliva possesses two types of influenza virus inhibitor, a sialic acid-dependent (RDE-sensitive) inhibitor that is responsible for weak neutralizing activity against Udorn virus and a sialic acid-independent (RDE-resistant) inhibitor that is highly potent against PR8 virus but has no activity against Udorn virus. The salivary inhibitor's function against PR8 virus was independent not only of sialic acid, but also of carbohydrate in general, as evidenced by treatment of saliva with periodate (data not shown), which, like RDE, failed to effect inhibition of PR8 (P = 0.85 for untreated versus periodate-treated saliva) but abolished the weak inhibition of Udorn virus (P < 0.0001 for untreated versus periodate-treated saliva; P = 0.41, control versus periodate-treated saliva).
Using the mouse model of respiratory influenza virus infection, we have shown that the site of deposition of the inoculum following intranasal delivery of virus can have a significant impact on the outcome of infection. Following standard TRT delivery, where the inoculum is initially spread throughout the respiratory tract, PR8 virus replicates to high titers in the nose, trachea, and lungs. Under these conditions, a virus dose of 50 PFU in BALB/c mice establishes a severe infection accompanied by weight loss, and a dose of 100 PFU is invariably fatal. However, when the inoculum is restricted to the nose, doses >300 times this otherwise lethal dose provide subclinical infection. Virus grows to high titers in the nose but fails to progress over time to the lungs. We showed that the mediator of this effect is an innate inhibitor present in saliva that essentially functions as a “molecular barrier” to protect the vulnerable lower respiratory tract. As in humans, saliva in the mouse is secreted into the oral cavity and can bathe the oropharynx. This provides the opportunity for the inhibitor to stop the virus from infecting epithelial cells in the oropharynx, but not in the nasal cavity. However, virus replicating in the nasal epithelium would come into contact with saliva in the process of spreading along the mucosa toward the trachea and lungs.
To our knowledge, this salivary inhibitor has not been previously described. Its ability to effectively neutralize PR8 virus indicates that it is not equivalent to the β-inhibitors, MBP and SP-D, that neutralize this strain poorly, if at all, due to the lack of carbohydrate side chains on the PR8 HA head (23). Similarly, the classical γ-inhibitors are RDE sensitive, so the salivary inhibitor is unlikely to be one of them. Additional support for the novelty of the salivary inhibitor comes from the ability of mouse saliva to neutralize virtually all of the in vitro infectivity of mutants of the Mem71-Bel (H3N1) virus selected for resistance to the β-inhibitors in bovine serum or to the γ-inhibitors in horse serum.
The RDE-resistant inhibitor may be one that is not shared by humans. Many different components of human saliva have been shown to have antibacterial and antiviral properties (24,–26); one that is particularly active against influenza virus is the scavenger receptor-rich glycoprotein gp340, originally referred to as salivary agglutinin (4, 7, 27). This molecule, which acts as a γ-inhibitor presenting sialic acid to block the receptor-binding site of HA, may have evolved to perform a function in humans similar to that of the mouse salivary inhibitor, but its potency in halting the progression of infection down the respiratory tree in humans or animal models remains unknown. Our preliminary studies with the saliva of very young children serologically naive to influenza virus showed that, although inhibitors of influenza virus are present, they do not have the same properties or specificity as the murine salivary inhibitor described here (data not shown).
The importance of this RDE-resistant salivary inhibitor to mice is unknown. To our knowledge, influenza A virus has not been isolated from wild mice (28), despite the sharing of an ecological niche with wild aquatic birds, the natural host of the virus, and the fact that they bear appropriate receptors for infection by avian viruses. Still, it can be inferred that influenza or influenza-like viruses may be important players in the natural history of the species. There is evidence of selection in wild mice for heterozygosity in the gene expressing the Mx protein, an interferon-inducible protein showing strong antiviral activity against influenza A virus (28). Perhaps the salivary inhibitor is another mechanism to limit the impact of the virus in this species. The presence of a potent inhibitor of PR8 but not Udorn virus in mouse saliva may explain our previous data on contact transmission in the mouse model (29). We showed that the ability of an influenza virus-infected mouse to spread virus to cohoused mice was correlated with the titer of infectious virus in the saliva of the donor mouse. Contact transmission readily occurred with Udorn virus but was not observed with PR8 virus, despite high titers of the virus in the noses of the donor mice. These data are consistent with the specificity of the virus-neutralizing function of the inhibitor and indicate a role, not only for protection of the infected individual against inhibitor-sensitive influenza viruses, but also for other individuals that they may come into contact with.
The breadth of reactivity of the salivary inhibitor against different virus strains of type A (H1N1 and H3N2) and B viruses and its apparent potency against lethal infection justify further investigation. Attempts to isolate or enrich for the RDE-resistant inhibitor from mouse saliva by ultrafiltration, affinity chromatography, or reverse-phase high-performance liquid chromatography (HPLC) have not yet been successful in yielding a defined molecule. Nevertheless, we have also been focusing on identifying the target of the inhibitor on the virus, and these findings are reported in the accompanying paper (30).
MDCK cells were maintained in RPMI 1640 without glutamine (Sigma-Aldrich, Castle Hill, New South Wales, Australia) supplemented with 10% (vol/vol) heat-inactivated fetal calf serum (FCS) (HyClone, Logan, UT, USA), benzylpenicillin (100 IU/ml; CSL Ltd., Parkville, Victoria, Australia), streptomycin sulfate (180 μg/ml; MP Biomedicals, Seven Hills, Victoria, Australia), gentamicin sulfate (24 μg/ml; Pfizer, West Ryde, New South Wales, Australia), l-glutamine (2 mM; Sigma-Aldrich), and sodium pyruvate (2 mM; MP Biomedicals) (RF10). Similar medium without FCS, l-glutamine, and sodium pyruvate (RPMI-anti) was used for dilutions, and RPMI-anti containing 1 mg/ml of bovine serum albumin (BSA) (RPMI plus BSA) was used as a control for virus neutralization assays.
Influenza viruses were propagated in 10-day-old embryonated hen's eggs at 35°C for 2 days for influenza A viruses or 33.5°C for 3 days for influenza B viruses. Allantoic fluid was then harvested and stored at −80°C. The following viruses were used in this study: A/WSN/33 (H1N1), A/Puerto Rico/8/34 (PR8; H1N1), A/Brisbane/59/07 (H1N1), A/California/7/09 (H1N1), A/England/51/02 (H1N2), A/Northern Territory/60/68 (H3N2), A/England/878/69 (H3N2), A/Memphis/1/71 (H3N2), A/Udorn/307/72 (H3N2), A/Port Chalmers/1/73 (H3N2), A/Victoria/3/75 (H3N2), A/Bangkok/1/79 (H3N2), A/Philippines/2/82 (H3N2), A/Beijing/353/89 (H3N2), A/Sydney/5/97 (H3N2), B/Hong Kong/8/73, and B/Panama/45/90. HKx31 (X-31; H3N2) influenza virus is a laboratory-derived high-yielding reassortant consisting of the internal proteins of PR8 with the surface proteins of A/Aichi/2/68 (H3N2) (31, 32). BKx73 (H3N2) is a laboratory-derived high-yielding reassortant consisting of the internal proteins of PR8 with the surface proteins of A/Bangkok/1/79 (H3N2). BJx109 (H3N2) is a laboratory-derived high-yielding reassortant consisting of the internal proteins of PR8 with the surface proteins of A/Beijing/353/89 (H3N2) (33). Mem71-Bel (H3N1) is a genetic reassortant of A/Memphis/1/71 (H3N2) bearing the NA of A/Bellamy/42. Mem71-Bel/HS, a mutant of Mem71-Bel selected for resistance to the γ-inhibitor in horse serum (34); Mem71-Bel/BS, a mutant of Mem71-Bel selected for resistance to the β-inhibitor in bovine serum (8); and the plaque-purified Mem71-Bel virus from which they were derived were all kindly provided by E. Margot Anders.
Inbred 6- to 8-week-old BALB/c mice (H-2d) were used for the majority of experiments in this study. Six- to 8-week-old CBA, C57BL/10, and SCID mice were used in one experiment. All the mice were bred in the Animal Facility of the Department of Microbiology and Immunology, University of Melbourne, under specific pathogen-free conditions. All experiments were conducted with the approval of the University of Melbourne Animal Ethics Committee.
To establish a URT infection, unanesthetized mice were inoculated i.n. with 104.5 PFU, or less as indicated, of the specified virus in 10 μl of phosphate-buffered saline (PBS) delivered directly to the nostrils with a pipette. To establish a TRT infection, mice were lightly anesthetized by isoflurane inhalation, and 50 μl of virus was delivered to the external nares to be breathed in deeply by the animal. At different times postinfection, supernatants from homogenates of the nasal turbinates, trachea, and lungs were prepared, and viral titers (PFU per organ) were determined by plaque assay on MDCK monolayers (35). The distribution of the inoculum to the required sites was initially confirmed in test animals using Evan's Blue dye (data not shown).
Mice were anesthetized using isoflurane and injected intraperitoneally (i.p.) with 200 μl of 20 μg/ml Carbachol (carbamycholine chloride; Sigma, St. Louis, MO, USA) in PBS. Approximately 200 μl of saliva was collected from each mouse over a 5-min period with a pipette, and aliquots were stored at −20°C. To obtain nasal washings, mice were killed by cervical dislocation, and the nasal passages were flushed with 0.5 to 1 ml of RPMI plus BSA through a 26-gauge needle inserted into the trachea, which was clamped to prevent backflow to the lungs. Outflow from the nose was collected in a petri dish and then transferred to tubes and kept on ice until it was used.
Plaque reduction in MDCK cell monolayers (35) cultured in 6-well tissue culture (TC6) plates was used to measure neutralization of virus infectivity. Unless otherwise stated, dilutions of virus were prepared in RPMI medium to give approximately 50 to 104.5 PFU per 10 μl and then mixed with untreated or treated saliva, nasal washings, or RPMI plus BSA as a control at a virus/saliva ratio of 1:9 (vol/vol). The mixtures were then incubated at 37°C for 30 min before being added to confluent monolayers of MDCK cells and allowed to adsorb for 45 min at 37°C. The cells were overlaid with 9 mg/ml agarose with 2 μg/ml l-(tosylamido-2-phenyl) ethyl chloromethyl ketone (TPCK)-treated trypsin (Worthington Biochemical) in Leibovitz L-15 medium with glutamine (Gibco) at pH 6.8, containing 0.4 mM HEPES buffer, 0.028% NaHCO3, 120 μg/ml penicillin, and 200 μg/ml streptomycin. After 3 days of incubation at 37°C in 5% CO2, plaques were counted without staining. The virus neutralized by preincubation with mouse saliva was calculated as a percentage of the number of plaques present after incubation with the RPMI-plus-BSA-treated control.
For RDE treatment, 1 volume of saliva in 8 volumes of Ca-Mg saline (0.24 mM CaCl2, 0.8 mM MgCl2, 20 mM boric acid, 0.14 mM sodium tetraborate, 0.15 M NaCl) was treated with 1 volume of V. cholerae or A. ureafaciens RDE (Sigma, St. Louis, MO, USA) for 30 min at 37°C. The final concentration of RDE was 50 mU/ml. The RDE was then inactivated by incubation at 56°C for 30 min. Saliva was also treated with 3 volumes of 0.011 M sodium periodate for 15 min at room temperature (RT), followed by inactivation of periodate with 6 volumes of 0.22% (wt/vol) glycerol in PBS. Treated and mock-treated samples were dialyzed against NH4HCO3 overnight and then freeze-dried and reconstituted to their original volume with PBS.
Statistical analyses of in vitro saliva titration curves and in vivo viral kinetics experiments used two-way analysis of variance (ANOVA), and the comparison of saliva treatments in vitro used one-way ANOVA, all with Sidak's multiple-comparison test. The data from other in vitro experiments were analyzed using an unpaired t test. Analyses were performed using GraphPad Prism software version 6.0e.
The work was supported by Project Grant 509281 and Program Grant 567122 from the National Health and Medical Research Council of Australia. The funders had no role in study design, data collection and interpretation, or the decision to submit the work for publication.
We thank Jane Ryan and Biota Holdings Ltd. for supporting our efforts to isolate the inhibitor.
For a companion article on this topic, see https://doi.org/10.1128/JVI.00145-17.