In contrast to the detrimental outcomes most often associated with the resolution of coinfections, the model presented here involving a localized Pneumocystis infection of the lung, followed 2 wk later by an influenza virus infection, results in a significant beneficial outcome for the host. In the week following the influenza infection, immunocompetent coinfected animals exhibited an accelerated rate of virus clearance, an accelerated appearance of higher influenza-specific neutralizing Ab titers in their serum and bronchoalveolar lavage fluid (BALF), significantly reduced inflammatory cytokine levels in their BALF, and reduced levels of morbidity relative to animals infected only with influenza virus. The beneficial outcome observed in coinfected immunocompetent animals was dependent on the ongoing resolution of a viable Pneumocystis infection. No differences in viral clearance were detected between coinfected and influenza-only-infected μMT mice or likewise for SCID mice. The accelerated anti-influenza response did not appear to be associated with influenza-specific CD8 T cell-mediated responses or NK cell responses in the lung. Rather, the increased rate of viral clearance was due to the enhancement of the influenza-specific Ab response, which in turn was transiently dependent upon the resolution of the ongoing Pneumocystis infection.

The generation of Ag-specific immune responses is a critical prerequisite for the effective clearance of most pathogens from a host. Secondary Ag-specific immune responses are accelerated and more intense relative to primary responses and thus are more effective in the clearance of pathogens. The increased efficacy of secondary immune responses is reliant upon the rapid reactivation of existing Ag-specific memory T and B cell clones that were established during the maturation of a prior primary immune response directed against the homologous pathogen. In the absence of established clones of Ag-specific memory cells, individuals must resort to the generation of primary immune responses to clear a pathogen. Although effective, the generation of Ag-specific primary responses is a protracted process, which, in turn, allows for a longer period of pathogen proliferation to proceed in the absence of any effective Ag-specific resistance. The ensuing increases in pathogen burdens most often result in greater host damage and morbidity. This is frequently the case in the event of infection by newly emerging antigenic variants of viral pathogens such as influenza viruses. The objective of current influenza vaccination strategies is to attempt to induce this protracted maturation of the primary immune response well in advance of exposure to anticipated circulating influenza virus strains. However, these vaccination strategies are restricted in their Ag specificity. Infection by emerging antigenic variants will negate the inherent advantages of using secondary Ag-specific memory immune responses and compel the host to generate the more protracted primary immune response to resolve the infection. An intervention strategy that effectively accelerates the appearance and intensity of a mature primary immune response at the site of infection would counter the advantage that pathogens have following the infection of naive hosts.

To accelerate the primary response, we propose that a transient state of immune preparedness be established within a local tissue that is susceptible to infection. This transient state of immune preparedness may involve the recruitment, assembly, and functional interaction of those immune components that are common to the generation of any primary immune response before the commitment to and maturation of Ag specificity. A heightened state of local immune preparedness could be applicable to the acceleration of vaccine-induced responses or to the elicitation of more robust primary protective immunity in the event of infection by an emergent variant of a circulating pathogen. In the infection model presented herein we use two disparate pulmonary pathogens to demonstrate that the resolution of the second infection can be accelerated if exposure to the second pathogen occurs while resolution of the infection by the first pathogen is still in progress.

The pathogens used in our model were Pneumocystis murina followed by influenza virus type A. Although both of these pathogens infect the lung, neither is associated with a predisposition to cause a subsequent infection by the other. This contrasts with other coinfection models in which infection by the first pathogen leads to a predisposition to infection by the second (1, 2). In our case, the possibility of a coinfection of immunocompetent individuals by these two pathogens is premised upon their simultaneous existence in a common environment. Pneumocystis is a ubiquitous opportunistic pathogen that colonizes the alveolar spaces of the lung. Approximately 20% of healthy immunocompetent individuals have detectable Pneumocystis DNA in their oropharyngeal cavity (3). Health-care workers in regular contact with immunocompromised patients are known to be at an increased risk for becoming Pneumocystis carriers. Although immunocompetent individuals develop asymptomatic subclinical responses following infection by this pathogen, they can act as transient reservoirs for the transmission and propagation of Pneumocystis as seen in humans (4) and modeled in murine studies (5, 6). Unlike Pneumocystis, influenza virus infections are capable of inducing symptomatic clinical infections of the lungs of the immunocompetent individuals. This virulent pathogen infects the respiratory and alveolar epithelium of the lung. The effective resolution of influenza infections is dependent upon the local recruitment of Ag-specific humoral and cell-mediated immune responses (7).

In the Pneumocystis/influenza coinfected animals, we observed a significant acceleration of viral clearance, an accelerated production of influenza-specific Ab titers, reduced lung damage, significantly altered cytokine levels in the lung, and evidence of reduced morbidity 1 wk after the influenza infection. If the sequence of the pathogens was reversed, no comparable enhancement of the anti-Pneumocystis response was detected. The ability of the Pneumocystis-induced immune response in an immunocompetent host to augment the protective immunity of a subsequent anti-influenza response suggests that under specific conditions, localized concurrent immune responses in the lung may be exploited for the benefit to the host.

BALB/c and C57BL/6 male and female mice at 6–8 wk of age were either purchased from The Jackson Laboratory or obtained from the breeding colonies maintained at Montana State University, Bozeman, MT. The male and female BALB/c SCID mice were obtained from Montana State University breeding colonies, and C57BL/6 SCID mice and μMT mice at 6–8 wk of age were obtained from The Jackson Laboratory. All of the animals were housed at the Animal Resource Center at Montana State University for the duration of these experiments. The animal facilities and the experimental procedures used throughout these experiments complied with the approved institutional animal care and use committee protocols established at Montana State University.

The Pneumocystis murina used for these experiments was maintained in a Pneumocystis-infected colony of BALB/c SCID mice at Montana State University. The influenza virus used was the A/PR8/8/34 (PR8; H1N1) strain. This virus was prepared at and obtained from the Trudeau Institute (Saranac Lake, NY). The virus stock was grown in the allantoic fluid of 10-day-old chicken embryos that had been infected for 48 h at 35°C. The harvested allantoic fluid was then stored at −80°C.

Lung homogenates from Pneumocystis-infected SCID mice were used to infect the mice in these studies. To administer intratracheal inoculations of Pneumocystis-infected lung homogenates, experimental mice were lightly anesthetized with 5% isoflurane in oxygen. A 100-μl inoculum of the Pneumocystis-infected lung homogenate containing 107Pneumocystis organisms was then directly injected into the lungs of the mice. Control groups in the coinfection experiments were given lung homogenates from SCID mice that had not been infected with Pneumocystis. Aliquots of 100 μl of the uninfected lung homogenates were used. These groups are referred to as the influenza-only control group. The β-glucan preparation was made from a β-glucan stock (Sigma-Aldrich) and was similarly administered.

PR8 influenza inoculations were also done while the mice were under a 5% isoflurane anesthesia. Mice were given a 50 μl intranasal inoculation containing 1500 PFU of PR8 influenza virus. The mice were taken at designated time points up to 10 days after the influenza infection. In those experiments where body weight was monitored, the animals were weighed on the day of the influenza infection and each day afterward until the end of the experiment. The change in body weight was determined by the difference between the body weight at the time of influenza infection and the body weight on the given day after infection.

The Pneumocystis burden in the lungs of infected mice was determined as previously described (8). Briefly, infected lungs were harvested from mice and passed through a mesh screen in 5 ml of HBSS. An aliquot of the suspension was taken to make a cytocentrifuge smear, which was then stained with Diff-Quik (Baxter). The number of Pneumocystis nuclei counted in 30–50 oil immersion fields was used to calculate the total number of Pneumocystis in the lungs. The limit of detection using this method was log10 4.1 nuclei/lung.

Following influenza infection, the mice were sacrificed as previously described (9). Lungs recovered from the mice were snap-frozen in liquid nitrogen and then stored at −80°C until analyzed. At this time the lungs were homogenized and 10-fold serial dilutions of the homogenates were used to inoculate monolayers of Madin-Darby canine kidney cells. Our previously described plaque assay procedure was used in the experiments reported here (10).

Bronchoalveolar lavage fluids (BALF)3 were obtained by washing the lungs with 1.5 ml of 3 mM EDTA in HBSS in two aliquots of 750 μl. A recovery of 1.3 ± 0.1 ml of BALF was reproducibly obtained from each mouse. The cells recovered in the BALF were counted to obtain total cell counts, and an aliquot of the recovered BALF was used to make a Diff-Quik-stained cytospin to examine the differential cell recovery from each animal. The remaining cellular content in the recovered BALF was removed by centrifugation for FACS analysis. The fluid was then stored at −80°C for use in determining the Ab, cytokine, albumin, and lactate dehydrogenase (LDH) content.

The BALF albumin was determined by use of an albumin colorimetric assay (Sigma-Aldrich). Color absorbency was read at 630 nm and reported in mg/ml in the BALF. The level of LDH detected in the lavage fluid was determined by colorimetric assay (CytoTox 96, Promega). The assay was read at an absorbance of 490 nm and reported as U/ml in the BALF. Each of the samples was tested in duplicate for the albumin and LDH assays.

Cytokine levels in the BALF were determined by use of an inflammatory cytokine bead array kit (BD Biosciences). The assays were conducted according to the manufacturer’s instructions. The detection of the cytokines was done using a FACScan cytometer and then analyzed according to the cytometric bead array software (BD Biosciences). The amount of IL-13 recovered in the BALF was determined using a murine IL-13 Quantikine ELISA kit (R&D Systems).

Cells were recovered from the BALF by centrifugation and then stained with the following fluorchrome-conjugated mAbs: PE-conjugated anti-mouse CD8, FITC-conjugated CD49 (BD Pharmingen), PE/Cy5.5-conjugated anti-mouse TCR (Caltag Laboratories), and NP366–374-specific and PA224–233-specific allophycocyanin-conjugated tetramer complexes that recognized H-2Db-restricted virus-specific CD8+ T cells (Trudeau Institute). Analysis of the stained cells was conducted on a FACSCanto (BD Biosciences) and then analyzed using FlowJo software (BD Pharmingen).

PR8-specific Abs were detected by ELISA. These assays were conducted using a PR8 membrane preparation derived from a purified PR8 influenza virus preparation (9). Serum dilutions are designated in the figures, and the BALF samples remained undiluted. Alkaline phosphatase-conjugated anti-mouse IgG (Sigma-Aldrich) and IgA and IgM (Serotec) were used to detect PR8-specific serum Abs in the assays. Absorbance was read at 405 nm. Virus neutralization assays using dilutions of serum or BALF were conducted to determine the extent of their virus neutralization activity. Dilutions of serum or BALF samples were challenged with 150 PFU of influenza virus. The sample/virus mixture was incubated for 90 min at 37°C. Following this, the mixture was separated into aliquots onto Madin-Darby canine kidney cells to determine by plaque assay if any infectious virus remained. Our previously described plaque assay procedure was followed (10). The percentage neutralization was determined by comparing the total number of plaques recovered from nonspecific serum samples vs the total number of plaques recovered from the serum samples obtained from the coinfected and influenza-only infected animals.

The data are expressed as the means ± SD. The results reported here are from one experiment that was representative of at least three independent experiments unless stated otherwise. The sample size of each group is 4–6 mice. Statistical differences between designated groups were determined using nonparametric one-way ANOVA tests with Bonferroni corrections to account for multiple comparisons within given experiments or by t tests when only two experimental groups were involved.

We have previously used the rate of pathogen clearance from an infection site to assess the efficacy of a local adaptive immune response (9). In our present coinfection model we found that virus recovery in the lungs 1 wk after an influenza infection was >100-fold less if a Pneumocystis infection occurred 2 wk earlier relative to that seen in the influenza-only control group (Fig. 1,A). If a Pneumocystis infection was initiated 3 wk earlier, the recovery of virus was only 10-fold less than that seen in the influenza-only control group. If the Pneumocystis and influenza infections were given 6 h apart in either order, we found that the recovery of influenza virus from the lungs was equivalent between the coinfected groups, and that both coinfection groups had equivalent or slightly greater viral recoveries than those seen in the influenza-only group (Fig. 1 B). The recovery of Pneumocystis from the lungs was not altered by an influenza infection given 6 h before or after the Pneumocystis infection (data not shown).

FIGURE 1.

Clearance of influenza virus from the lung following coinfection. A, Viral recovery levels from the lungs of coinfected mice. Mice were infected with a Pneumocystis-infected lung homogenate 2 (▩) or 3 (▦) wk before an influenza infection. The control mice were given the uninfected lung homogenate at 2 wk before the influenza infection (□). The lungs were assessed for influenza virus recovery by plaque assay 1 wk after influenza infection. B, Mice were inoculated with the coinfecting pathogens 6 h apart: Pneumocystis then influenza (▥), influenza then Pneumocystis (▦), and influenza only (□). The lungs were assessed for viral recovery by plaque assay 1 wk later. ∗, p < 0.05 relative to the influenza-only group; n = 5 or 6 mice per group; the dotted horizontal line indicates the limit of assay detection. C, Titration of PR8-specific serum IgG levels from coinfected mice 1 wk after influenza infection. Serum samples from mice at 2- (□) or 3- (▴) wk interval between infections or mice given influenza-only infection (▪) were tested by ELISA and absorbance was read at 405 nm.

FIGURE 1.

Clearance of influenza virus from the lung following coinfection. A, Viral recovery levels from the lungs of coinfected mice. Mice were infected with a Pneumocystis-infected lung homogenate 2 (▩) or 3 (▦) wk before an influenza infection. The control mice were given the uninfected lung homogenate at 2 wk before the influenza infection (□). The lungs were assessed for influenza virus recovery by plaque assay 1 wk after influenza infection. B, Mice were inoculated with the coinfecting pathogens 6 h apart: Pneumocystis then influenza (▥), influenza then Pneumocystis (▦), and influenza only (□). The lungs were assessed for viral recovery by plaque assay 1 wk later. ∗, p < 0.05 relative to the influenza-only group; n = 5 or 6 mice per group; the dotted horizontal line indicates the limit of assay detection. C, Titration of PR8-specific serum IgG levels from coinfected mice 1 wk after influenza infection. Serum samples from mice at 2- (□) or 3- (▴) wk interval between infections or mice given influenza-only infection (▪) were tested by ELISA and absorbance was read at 405 nm.

Close modal

The PR8-specific Ab response in the serum was assessed in the groups of animals 1 wk after the influenza coinfection (Fig. 1 C). The PR8-specific serum IgG response was higher in animals that had received the Pneumocystis infection 2 wk earlier than it was in those animals that had received the Pneumocystis infection 3 wk earlier. The serum PR8-specific IgG response in both of these coinfection groups was greater than that seen in those animals that had received only an influenza infection. As a result of this finding, further coinfection experiments were based on a 2-wk interval between a prior Pneumocystis infection and the influenza infection.

The rate of clearance of influenza virus in the coinfection group and in the influenza-only control group was examined over a 10-day period following the virus challenge (Fig. 2,A). At 3 and 5 days after the influenza infection, viral recoveries in the coinfection and the influenza-only infection groups were high. Although there was no significant difference between these groups, slightly more virus was recovered from the coinfected animals. By 7 days after the influenza infection no virus could be detected in the lungs of the coinfected animals, whereas viral recoveries in the influenza-only control group remained unchanged from those seen earlier in the week. At 10 days postinfluenza infection, significant differences in the level of viral recovery still existed between the coinfection and influenza-only infection groups, although clearance of virus from the influenza-only control group was reduced relative to levels seen at day 7 postinfluenza infection. The wild-type coinfected animals used in these experiments easily cleared the Pneumocystis infections. At the time of the influenza infection (day 0), the Pneumocystis burden in the coinfected animals was >10-fold less than what these animals had been inoculated with 2 wk earlier (Fig. 2,B). Five days after the influenza infection, the Pneumocystis burden was further diminished and was in fact undetectable in 4 of 5 animals. Pneumocystis burdens were below the limit of detection in the coinfected animals at 7 and 10 days postinfluenza infection. If an influenza infection was given 2 wk before a Pneumocystis infection, no alteration in the clearance of the Pneumocystis from the lungs was detected 2 wk later (Fig. 2 C).

FIGURE 2.

Kinetics of influenza and Pneumocystis recovery from coinfected animals. A, Kinetic analysis of the viral recovery was done on mice that were infected with Pneumocystis 2 wk before the influenza infection. Mice from the coinfected group (▩) and influenza-only control group (□) were taken at designated time points following the influenza infection. The recovery of influenza virus in their lungs was assessed by plaque assay. B, The recovery of Pneumocystis from the lungs of coinfected mice was determined at the designated time points using Diff-Quik-stained slides made from the lung homogenates. C, Recovery levels of Pneumocystis 2 wk following Pneumocystis infection of the lungs of mice that had been given an influenza/Pneumocystis coinfection (□) or a Pneumocystis-only infection (▪). n = 5 mice per group; the dotted horizontal line indicates the limit of assay detection; ∗∗∗, p < 0.005 relative to influenza-only control group.

FIGURE 2.

Kinetics of influenza and Pneumocystis recovery from coinfected animals. A, Kinetic analysis of the viral recovery was done on mice that were infected with Pneumocystis 2 wk before the influenza infection. Mice from the coinfected group (▩) and influenza-only control group (□) were taken at designated time points following the influenza infection. The recovery of influenza virus in their lungs was assessed by plaque assay. B, The recovery of Pneumocystis from the lungs of coinfected mice was determined at the designated time points using Diff-Quik-stained slides made from the lung homogenates. C, Recovery levels of Pneumocystis 2 wk following Pneumocystis infection of the lungs of mice that had been given an influenza/Pneumocystis coinfection (□) or a Pneumocystis-only infection (▪). n = 5 mice per group; the dotted horizontal line indicates the limit of assay detection; ∗∗∗, p < 0.005 relative to influenza-only control group.

Close modal

To ascertain the necessity for a viable Pneumocystis infection to achieve this accelerated reduction in viral recovery, mice were dosed with a β-glucan preparation. β-glucan is a subcomponent of the Pneumocystis cell wall and is a well known non-specific immunostimulant (11, 12, 13). This group was treated twice with 250 ng of the β-glucan preparation (1 dose per week) during the 2 wk before the influenza infection. We had previously observed that this dose of β-glucan elicited an inflammatory cell infiltrate in the lungs without incurring any detectable indications of morbidity (data not shown). Despite this treatment regime, we were unable to replicate the accelerated clearance of influenza virus from the lungs of the β-glucan-treated mice that we had observed in the coinfected mice (data not shown).

Cellular infiltrates into the lungs of the coinfected animals were already evident at the time of the influenza infection and were greater during the 5 days after the influenza infection relative to those observed in the influenza-only control animals (Fig. 3). The recovery of macrophages in the BALF of the coinfected mice was significantly greater throughout the resolution of the influenza infection (Fig. 3,A). Macrophage recruitment in the influenza-only infection group did not change from the levels observed at the time of the influenza infection. The level of lymphocyte recovery in the coinfected animals did not show any appreciable change during the week following the influenza infection (Fig. 3,B). By 7 and 10 days after the influenza infection, lymphocyte recruitment in the influenza-only group had reached equivalent or slightly greater levels than those detected in the coinfected animals. Neutrophil recruitment levels peaked at day 3 in the coinfected animals and then diminished thereafter (Fig. 3,C). In contrast, neutrophil recruitment levels did not reach their peak in the influenza-only infected animals until 7 days after the influenza infection. The recovery of eosinophils was significant in the coinfected animals throughout the resolution of the influenza infection, whereas eosinophil recovery in the influenza-only infected animals remained at negligible levels (Fig. 3 D).

FIGURE 3.

Recovery of cells from the BALF following coinfection. The lungs of each animal were lavaged at the designated time points following coinfection with the influenza virus. The BALF cells, macrophages (A), lymphocytes (B), neutrophils (C), and eosinophils (D) were stained with Diff-Quik, and differential cell counts were determined by microscopy. The cell populations recovered from the BALF of the coinfected (▩) and influenza-only infected (□) animals were determined by multiplying the total cell recovery by the percentage of each cell subset observed in the differential cell counts. ∗, p < 0.05; ∗∗, p < 0.01; ∗∗∗, p < 0.001. This is one of three independent experiments.

FIGURE 3.

Recovery of cells from the BALF following coinfection. The lungs of each animal were lavaged at the designated time points following coinfection with the influenza virus. The BALF cells, macrophages (A), lymphocytes (B), neutrophils (C), and eosinophils (D) were stained with Diff-Quik, and differential cell counts were determined by microscopy. The cell populations recovered from the BALF of the coinfected (▩) and influenza-only infected (□) animals were determined by multiplying the total cell recovery by the percentage of each cell subset observed in the differential cell counts. ∗, p < 0.05; ∗∗, p < 0.01; ∗∗∗, p < 0.001. This is one of three independent experiments.

Close modal

To obtain an indication of the general health of the animals as they responded to the influenza infection, we monitored the daily changes in their body weights. During the 10 days following the influenza infection, the coinfected animals did not lose any weight (Fig. 4). In fact, they averaged an additional 1 g weight gain by the end of the experiment. The influenza-only control animals continued to lose weight from day 4 postinfluenza infection to the end of the experiment.

FIGURE 4.

Daily weight changes following the influenza infection. Pneumocystis/influenza coinfected animals (▩) and influenza-only infected animals (□) were weighed on the day of the influenza infection and each day thereafter until the end of the experiment at day 10 postinfluenza infection. The change in weight for each animal is the difference between the starting weight and the weight at each day following the influenza infection. The weight change for each animal in a group was then averaged with the cohort to determine the average weight change for that cohort on the day of interest. This is one of three independent experiments. n = 5 or 6 mice/group; ∗∗∗, p < 0.0001.

FIGURE 4.

Daily weight changes following the influenza infection. Pneumocystis/influenza coinfected animals (▩) and influenza-only infected animals (□) were weighed on the day of the influenza infection and each day thereafter until the end of the experiment at day 10 postinfluenza infection. The change in weight for each animal is the difference between the starting weight and the weight at each day following the influenza infection. The weight change for each animal in a group was then averaged with the cohort to determine the average weight change for that cohort on the day of interest. This is one of three independent experiments. n = 5 or 6 mice/group; ∗∗∗, p < 0.0001.

Close modal

Increased levels of serum albumin and LDH in the BALF have been used as indicators of lung damage following infection (14, 15, 16, 17). Serum albumin (Fig. 5,A) and LDH levels (Fig. 5 B) in the BALF of the coinfected animals were higher at the time of the influenza infection (day 0). Following the influenza infection, serum albumin levels in the BALF of the coinfected animals remained equivalent to the day 0 values, whereas significant increases were detected in the influenza-only infection group. LDH levels in the coinfected animals also remained unchanged following the influenza infection, whereas levels in the influenza-only infection group rose by >3- and 6-fold at 7 and 10 days, respectively, following the influenza infection.

FIGURE 5.

Serum albumin and LDH levels in the BALF of the coinfection and influenza-only infection groups following infection with the influenza virus. Aliquots of 100 μl of the BALF from each coinfected (▩) and influenza-only infected (□) animal were assessed for serum albumin (A) and LDH (B) levels by ELISA. This is one of three independent experiments. n = 5 or 6 mice/group; ∗, p < 0.05; ∗∗, p < 0.01; ∗∗∗, p < 0.001.

FIGURE 5.

Serum albumin and LDH levels in the BALF of the coinfection and influenza-only infection groups following infection with the influenza virus. Aliquots of 100 μl of the BALF from each coinfected (▩) and influenza-only infected (□) animal were assessed for serum albumin (A) and LDH (B) levels by ELISA. This is one of three independent experiments. n = 5 or 6 mice/group; ∗, p < 0.05; ∗∗, p < 0.01; ∗∗∗, p < 0.001.

Close modal

One week after the influenza infection the levels of TNF-α, IL-10, IFN-γ, MCP-1, and IL-6 in the BALF of the coinfected animals were significantly reduced relative to those detected in the influenza-only control animals (Fig. 6). In contrast to this trend, the levels of IL-13 recovered in the BALF of the coinfected animals remained somewhat higher as the influenza infection resolved. IL-5 levels in the BALF following the influenza infection were not significantly altered by the prior Pneumocystis infection (data not shown). By day 10 postinfluenza infection the cytokine levels in the BALF of the influenza-only control group were still generally higher than in the coinfected animals, although the levels were greatly diminished in both groups relative to those detected on day 7. In contrast, IL-13 levels were still higher in the coinfected group at this time.

FIGURE 6.

Cytokine levels in the BALF following influenza infection. Cytokine levels in the BALF of the coinfected (▩) and influenza-only infected mice (□) were determined at 7 and 10 days after the influenza infection. The BALF were assessed for the presence of the selected cytokines using cytokine bead array kits or by ELISA for the IL-13 levels. Cytokine bead array samples were screened using the FACScan, and the data were analyzed by the bead array software. IL-13 ELISA results were calculated from the standard curve as per the manufacturer’s instructions. n = 5 or 6 animals/group; ∗, p < 0.05; ∗∗, p < 0.005; ∗∗∗, p < 0.0001.

FIGURE 6.

Cytokine levels in the BALF following influenza infection. Cytokine levels in the BALF of the coinfected (▩) and influenza-only infected mice (□) were determined at 7 and 10 days after the influenza infection. The BALF were assessed for the presence of the selected cytokines using cytokine bead array kits or by ELISA for the IL-13 levels. Cytokine bead array samples were screened using the FACScan, and the data were analyzed by the bead array software. IL-13 ELISA results were calculated from the standard curve as per the manufacturer’s instructions. n = 5 or 6 animals/group; ∗, p < 0.05; ∗∗, p < 0.005; ∗∗∗, p < 0.0001.

Close modal

We determined the level of influenza-specific CD8 T cells in the BALF by staining with tetramer complexes that recognized MHC class I H-2Db NP366–374-specific and PA224–233-specific CD8 T cells. Before day 7 postinfluenza infection, only background staining for influenza-specific CD8+ T cells was detected in the BALF of the coinfected and influenza-only infected animals (Fig. 7 A). At day 7 we found equivalent but low numbers of influenza-specific CD8+ T cells in both groups of animals. By day 10 postinfluenza infection, the recovery of influenza-specific CD8+ T cells was significantly greater in the influenza-only infection group. In fact, we did not detect any difference in the number of these cells recovered from the coinfected animals at 7 and 10 days after the influenza infection.

FIGURE 7.

Influenza-specific CD8 T cells and NK cells in the BALF. Cells recovered from the BALF of the coinfected (▩) and influenza-only infected (□) mice were stained for FACS analysis. PE-conjugated anti-mouse CD8 and APC-conjugated tetramer complexes specific for NP366–374 and PA224–233 epitopes presented within H-2Db MHC-1 were used to stain influenza-specific CD8 T cells (A). The total number of influenza-specific CD8 T cells was calculated by adding the products from the percentage of specific CD8 T cells for each epitope multiplied by the total cells recovered in the BALF. The total NK cells was determined by staining the cells recovered in the BALF with FITC-conjugated anti-mouse CD49 and PE/Cy5.5-conjugated anti-mouse TCR (B). The percentage of CD49+/TCR cells was multiplied by the total number of cells recovered in the BALF to calculate the total number of recovered NK cells. n = 5 mice. ∗, p < 0.05, ∗∗∗, p < 0.001.

FIGURE 7.

Influenza-specific CD8 T cells and NK cells in the BALF. Cells recovered from the BALF of the coinfected (▩) and influenza-only infected (□) mice were stained for FACS analysis. PE-conjugated anti-mouse CD8 and APC-conjugated tetramer complexes specific for NP366–374 and PA224–233 epitopes presented within H-2Db MHC-1 were used to stain influenza-specific CD8 T cells (A). The total number of influenza-specific CD8 T cells was calculated by adding the products from the percentage of specific CD8 T cells for each epitope multiplied by the total cells recovered in the BALF. The total NK cells was determined by staining the cells recovered in the BALF with FITC-conjugated anti-mouse CD49 and PE/Cy5.5-conjugated anti-mouse TCR (B). The percentage of CD49+/TCR cells was multiplied by the total number of cells recovered in the BALF to calculate the total number of recovered NK cells. n = 5 mice. ∗, p < 0.05, ∗∗∗, p < 0.001.

Close modal

The recovery of NK cells from the BALF of the coinfected animals from the time of the influenza infection until day 7 postinfluenza infection was relatively consistent (Fig. 7 B). NK cells were not found in the BALF of the influenza-only infection group until 3 days after the influenza infection. At this time they were significantly less than the levels observed in the coinfected animals. By day 5 postinfluenza infection, the equivalent amounts of NK cells were detected in the BALF of the coinfected and influenza-only infected animals. NK cell recruitment into the airways of the influenza-only infected animals continued to increase through day 7 postinfluenza infection. The recovery of NK cells from the BALF of the influenza-only infected animals at day 7 postinfluenza infection was significantly greater than from the coinfected animals. Three days later the recovery of NK cells from the BALF of the coinfected and influenza-only infected animals was equivalent and low.

In our initial studies, an ELISA analysis of serum Ab levels revealed that the animals that had a 2-wk interval between the Pneumocystis and influenza infections had greater PR8-specific serum IgG titers 1 wk after the influenza infection than did those animals that had a 3-wk interval between infections (Fig. 1,C). In our subsequent experiments, ELISA analysis of the serum and BALF recovered following the influenza infection indicated that PR8-specific Abs appeared sooner and remained at greater levels in the coinfected animals than in the influenza-only infected animals (Fig. 8). In our analysis of the viral recovery from the lungs of the coinfected animals, we had found that no virus could be detected at 7 and 10 days after the influenza infection, whereas significant levels of viral recovery were still noted in the influenza-only control animals. In view of our viral recovery results we examined the virus-neutralizing activity of the serum and BALF recovered from the coinfected and influenza-only infected animals (Fig. 9). The viral-neutralizing activity in the serum of the coinfected animals was significantly greater at a 1/100 dilution at day 5 and was only slightly better or equivalent to the influenza-only infected animals as the serum became more diluted (Fig. 9,A). At 7 and 10 days after the influenza infection, the serum-neutralizing activity of the coinfected animals tended to be only slightly better than that of the influenza-only infected animals up to a dilution of 1/1000 (Fig. 9, B and C).

FIGURE 8.

Detection of influenza-specific Ab titers in the BALF and serum. ELISA tests were performed on the BALF and serum from the coinfected (▩) and the influenza-only infected (□) animals 7 and 10 days after infection with influenza virus. Influenza-specific serum IgG (A) and IgM (B) levels were measured in a 1/100 dilution of a serum aliquot. Influenza-specific BALF IgG (C), IgA (D), and IgM (E) levels were measured in undiluted BALF aliquots. Absorbance readings were taken at 405 nm. This is one of three independent experiments, n = 5 or 6 animals. ∗, p < 0.05; ∗∗, p < 0.01; ∗∗∗, p < 0.005.

FIGURE 8.

Detection of influenza-specific Ab titers in the BALF and serum. ELISA tests were performed on the BALF and serum from the coinfected (▩) and the influenza-only infected (□) animals 7 and 10 days after infection with influenza virus. Influenza-specific serum IgG (A) and IgM (B) levels were measured in a 1/100 dilution of a serum aliquot. Influenza-specific BALF IgG (C), IgA (D), and IgM (E) levels were measured in undiluted BALF aliquots. Absorbance readings were taken at 405 nm. This is one of three independent experiments, n = 5 or 6 animals. ∗, p < 0.05; ∗∗, p < 0.01; ∗∗∗, p < 0.005.

Close modal
FIGURE 9.

Virus neutralization using serum from days 5 (A), 7 (B), and 10 (C) or BALF from days 7 (D) and 10 (E) postinfluenza infection. Serum or BALF dilutions from the Pneumocystis/influenza coinfected animals (▪) and influenza-only infected animals (□) were incubated with a 150 PFU challenge of influenza virus for 90 min at 37°C. Following this, the entire sample/virus mixture was added to confluent monolayers of MDCK cells in 12-well plates and allowed to adsorb to the cells for 90 min at 35°C. The assay was then incubated for 4 days at 35°C, after which the cells were fixed and stained with crystal violet to enumerate any plaques. n = 4–6 mice per group. ∗, p < 0.05; ∗∗, p < 0.005; ∗∗∗, p < 0.0001.

FIGURE 9.

Virus neutralization using serum from days 5 (A), 7 (B), and 10 (C) or BALF from days 7 (D) and 10 (E) postinfluenza infection. Serum or BALF dilutions from the Pneumocystis/influenza coinfected animals (▪) and influenza-only infected animals (□) were incubated with a 150 PFU challenge of influenza virus for 90 min at 37°C. Following this, the entire sample/virus mixture was added to confluent monolayers of MDCK cells in 12-well plates and allowed to adsorb to the cells for 90 min at 35°C. The assay was then incubated for 4 days at 35°C, after which the cells were fixed and stained with crystal violet to enumerate any plaques. n = 4–6 mice per group. ∗, p < 0.05; ∗∗, p < 0.005; ∗∗∗, p < 0.0001.

Close modal

The accelerated viral clearance in the coinfected animals was evident at 7 days after the influenza infection. At this time, substantial PR8-specific Ab levels were detected in the BALF of the coinfected animals, whereas only trace levels were detected in the influenza-only control animals. To determine whether the accelerated appearance of PR8-specific Abs at the site of infection could be responsible for the accelerated viral clearance observed in the lungs, we conducted viral neutralization assays using the BALF. Viral-neutralizing activity was significantly greater in the BALF recovered from the coinfected animals than from the influenza-only infected animals at 7 days postinfluenza infection (Fig. 9,D). By day 10 postinfluenza infection, the viral-neutralizing capacity of the BALF recovered from the influenza-only control group had become equivalent to that detected in the coinfected animals (Fig. 9 E). No virus-neutralizing activity was detected in the BALF from naive animals or in animals that had received only a Pneumocystis infection (data not shown).

The evidence from our coinfection model with immunocompetent animals suggested that the accelerated appearance of higher serum and BALF levels of influenza-specific Abs and their associated viral-neutralizing activity may be responsible for the enhanced rate of viral clearance. To further examine the necessity of the Ab response for the accelerated clearance of influenza virus in the coinfected immunocompetent mice, we coinfected μMT and SCID mice. At the time of the influenza coinfection, no differences in the cytokine and cellular infiltration levels were detected in the BALF recovered from the immunocompetent and μMT mice as a result of the Pneumocystis infection (data not shown). At this time, Pneumocystis burdens averaged 6.04 ± 0.76 and 7.12 ± 0.09 (log10) nuclei for the immunocompetent and μMT coinfected mice, respectively. Seven days after the influenza infection, viral recovery levels in the μMT coinfected and μMT influenza-only infected control groups were equivalent (Fig. 10) despite the Pneumocystis burden having increased to 7.98 ± 0.18 (log10) nuclei in the coinfected μMT mice. In fact, the viral recovery levels from the coinfected and influenza-only infected μMT groups were also equivalent to those detected in the immunocompetent influenza-only control group. In contrast, 7 days after the influenza infection the viral recovery level in the coinfected immunocompetent animals was significantly reduced and the Pneumocystis burden had been reduced to 4.94 ± 0.74 (log10) nuclei. Viral recoveries from the coinfected and influenza-only infected SCID mice were equivalent and were only slightly greater than those found in both of the μMT infection groups (Fig. 10). Pneumocystis burdens in the coinfected SCID mice at the time of the influenza coinfection were 6.45 ± 0.32 (log10) nuclei and had increased to 7.56 ± 0.21 (log10) nuclei 7 days after the influenza coinfection. At the time of the influenza viral infection, cytokine levels in the BALF of the coinfected SCID mice were lower than in the immunocompetent and μMT coinfected animals or they were only at trace levels (data not shown). Only macrophages were recovered from the BALF of the coinfected SCID mice at this time, and the recovery level of these cells was equivalent to that detected in the coinfected μMT and immunocompetent animals (data not shown).

FIGURE 10.

Viral recovery in coinfected wild-type, μMT, and SCID mice. Lungs were harvested from coinfected wild-type (▩), SCID (▩, lightly shaded), and μMT (▩, shaded) mice and from the influenza-only infected wild-type (□), SCID (▦), and μMT (dark grey) mice for analysis of viral recovery by plaque assay 1 wk after influenza infection. n = 6 animals for analysis of viral recovery; ∗∗, p < 0.001 relative to coinfected wild-type animals.

FIGURE 10.

Viral recovery in coinfected wild-type, μMT, and SCID mice. Lungs were harvested from coinfected wild-type (▩), SCID (▩, lightly shaded), and μMT (▩, shaded) mice and from the influenza-only infected wild-type (□), SCID (▦), and μMT (dark grey) mice for analysis of viral recovery by plaque assay 1 wk after influenza infection. n = 6 animals for analysis of viral recovery; ∗∗, p < 0.001 relative to coinfected wild-type animals.

Close modal

The resolution of the coinfection model presented here results in a beneficial outcome for its host. In the coinfected mice the clearance of influenza virus from their lungs was significantly accelerated relative to that seen in the influenza-only control group. The coinfected mice had at least a 3 day or better headstart on the clearance and resolution of the influenza virus infection. This rapid viral clearance was accompanied by the accelerated appearance of a more intense PR8-specific neutralizing Ab response in the lungs. No corresponding accelerated influenza-specific CD8 T cell- or NK cell-associated response was evident in the lungs of the coinfected animals at the time that the accelerated viral clearance was observed. In the absence of an Ab response, as with the μMT mice, viral recovery from the lungs was not altered by a preceding Pneumocystis infection. The accelerated resolution of the influenza virus coinfection was also associated with reduced indications of morbidity, lung damage, and decreased levels of proinflammatory cytokines recovered in the BALF. Our results indicated that the increased rate of viral clearance observed in the immunocompetent coinfected animals was dependent on an accelerated and enhanced localized influenza-specific Ab response, which, in turn, was dependent on a temporal association with the resolution of an ongoing Pneumocystis infection.

The exposure sequence and the time between the pathogen exposures were found to be critical. An influenza infection 2 wk before a Pneumocystis infection did not alter the clearance of the Pneumocystis infection. If exposure to both pathogens occurred within 6 h of one another in either order, no alteration in the recovery of either pathogen was detected. Since the greatest Pneumocystis burden occurred within 6 h of the influenza infection in these models, this suggested that no Pneumocystis-derived Ags were associated with the enhancement of the influenza response. The inability of the β-glucan preparation to induce an equivalent level of Ab enhancement and viral clearance confirmed that a major component of Pneumocystis could not elicit the same immune mediators or mechanisms as those associated with the response to the viable Pneumocystis infection. This observation is analogous to the inability of a Toxoplasma gondii lysate preparation to elicit the equivalent effect that the response to a viable T. gondii infection had upon the resolution of a Nippostrongylus brasiliensis infection (18). In both coinfection models, a viable infection rather than simply the exposure to a component(s) of the pathogen was required to influence the immune response directed against the second pathogen.

The transient nature of the enhanced anti-influenza response was apparent if changes were made to the interval between the pathogen exposures. If the interval between the Pneumocystis and then the influenza infection was extended to 3 wk, the accelerated effect on viral clearance and the enhancement of influenza-specific serum IgG levels were decreased relative to those observed for a 2-wk interval. This suggested that the potency of the enhancement effect upon the ensuing influenza immune response might be linked to the maturation of an established immunological component of the Pneumocystis-induced immune response.

At the time of the influenza infection, significant levels of cellular recruitment and cytokines were evident in airways of the coinfected animals. Despite this, viral recoveries in the coinfected animals up to 5 days after the influenza infection were still 1000-fold greater than the original viral inoculum and were similar or slightly better than those of the influenza-only control group. This indicated that a comparable influenza virus infection of the respiratory epithelium had been established in the lungs of the coinfected mice rather than the development of an abortive infection of the recruited cells present in the airways of the lung.

In the coinfected animals the accelerated rate of viral clearance was achieved without any additional recruitment of lymphocytes or neutrophils. In contrast, as lung damage and viral loads increased in the influenza-only infection group, neutrophil and lymphocyte recruitment in these animals increased to levels equivalent to or beyond those of the coinfected animals. The absence of any additional increase in the recruitment of inflammatory cells may have contributed to the reduced level of morbidity in the coinfected animals. The presence of eosinophils in the coinfected group and their absence in the influenza-only group were expected, as the recruitment of these cells is typical of the response to a Pneumocystis infection.

The coinfected animals lost the least amount of body weight following the influenza infection, and only this group was able to return to and exceed their starting body weights. The diminishing viral titers in the lungs of the coinfected animals were accompanied by reduced morbidity levels and reduced albumin and LDH levels in their BALF. These observations also comply with the reduced levels of proinflammatory cytokines (TNF-α, IFN-γ, MCP-1, IL-6) detected in the BALF of the coinfected animals following the influenza infection. The diminishing viral burdens in the coinfected mice are likely a primary reason for the reduced levels of morbidity and lung damage observed in these mice. However, the presence of antiinflammatory cytokines (IL-10 and IL-13) detected in the BALF of the coinfected animals at the time of the influenza infection may regulate the production of influenza- induced proinflammatory cytokines. This may have limited any influenza-induced exacerbation of the inflammatory environment in the lungs of the coinfected animals and thereby also have contributed to the reduced levels of morbidity and lung damage in these mice. IL-10 inhibits the production of a variety of CC (including MCP-1) and CXC chemokines, IL-6, IL-12, and subsequently IFN-γ, IL-18, and TNF-α, as well as autoregulating itself (19, 20). Following Pneumocystis infections, IL-10 has been shown to remain at biologically effective inhibitory levels in the lungs of immunocompetent mice up to 14 days after infection (21). IL-13 has been considered to be equivalent to IL-10 in regards to its antiinflammatory capabilities (22). Both IL-10 and IL-13 have been shown to regulate proinflammatory cytokine production by triggering the inhibition of NF-κB nuclear translocation (20, 22). In the absence of IL-10 and IL-13 at the time of the influenza infection, subsequent higher levels of TNF-α, IFN-γ, MCP-1, and IL-6 were observed in association with higher viral recoveries in the influenza-only control group. The lack of any increased cellular recruitment, the reduced levels of proinflammatory cytokines detected in the BALF, and the accelerated viral clearance seen in the coinfected animals support the contention that antiviral immune responses often operate in an excessive inflammatory environment that is beyond what is necessary to achieve viral clearance (15). A moderation of the elicited inflammatory response may be affordable and could be of benefit to the host without compromising viral clearance (23).

Our evidence indicates that the rapid viral clearance from the lungs of the coinfected animals is due to the accelerated appearance of influenza-specific neutralizing Abs at the site of infection. Seven days after the influenza infection, PR8-specific Ab levels in the serum and BALF of the coinfected animals were significantly elevated relative to the influenza-only control animals. At this time, no virus was detected in the lungs of the coinfected animals, and the virus-neutralizing capabilities of the BALF of these animals were significantly greater than those of the influenza-only infected animals. In fact, 5 days after the influenza infection, virus-neutralizing activity could already be detected in the serum of the coinfected animals, whereas it was undetectable in the influenza-only infected animals. It is unlikely that cell-mediated immune responses in the lung are responsible for the rapid viral clearance from the lungs of the coinfected animals. NK cell and influenza-specific CD8 T cell accumulations in the airways of the coinfected and influenza-only infected animals were low and equivalent at the time when the greatest difference in viral burden between the infection groups was observed. The continued viral recovery at day 10 postinfluenza infection in the influenza-only infected animals corresponds with the increased accumulation of influenza-specific CD8 T cells in the airways of these animals. In an analogous manner, the lack of viral recovery at days 7 and 10 postinfluenza infection in the coinfected animals corresponds with the absence of any increase in PR8-specific CD8 T cell accumulation. This coinfection model demonstrates that the coinfected animals had at least a 3 day or better headstart on their Ab-mediated clearance of the influenza infection. Although this headstart was short-lived, it was nonetheless critical in the beneficial resolution of the influenza infection in the coinfected animals. At day 10 postinfluenza infection the level and neutralizing capabilities of these PR8-specific Abs had increased substantially in both groups and had become closer to equivalence. Further long-term experiments will be required to determine whether the neutralizing activity and levels of PR8-specific Abs in both groups would plateau at the same time and level.

Viral recoveries in the influenza-only infected μMT and wild-type and in the coinfected μMT mice were equivalent at day 7 postinfluenza infection and all were more than 100-fold greater than the administered viral inoculum. It did not appear that the Pneumocystis burden, the Pneumocystis-induced cellular recruitment or cytokine production, or any Pneumocystis-induced pathology unique to μMT mice acted to impede the establishment of a viable influenza infection in these mice. The inability of μMT mice to mount Ab-mediated immune responses and the failure of the coinfected μMT mice to duplicate the accelerated viral clearance observed in the immunocompetent coinfected mice support our contention that the accelerated resolution of the influenza infection in these mice is Ab mediated.

The accelerated influenza Ab response may be associated with the generation of a local immune environment within the lung before the influenza infection that would facilitate the accelerated production of influenza-specific Abs. Within the lung, inducible BALT (iBALT) has been shown to play an effective role in the clearance of influenza infections in mice and to endow those animals with the ability to survive higher viral doses while incurring fewer pathological consequences than are associated with systemic immune responses (24). The transient establishment of iBALT structures within the lung before a pathogen insult could facilitate the accelerated appearance of a local Ab response since the cellular components and structural organization of this tertiary lymphatic tissue would already be in place. Additionally, their presence could alleviate the inflammatory component of lung damage that is associated with the recruitment of responding lymphocyte populations from systemic lymphatic tissues via the pulmonary circulation into the interstitial and alveolar spaces of the lung. If local iBALT structures were established within the lungs during the resolution of the Pneumocystis infection, their presence in the lung at the time of the influenza infection could have a significant impact on the resolution of the influenza infection. Further studies assessing the potential involvement of iBALT in the accelerated appearance of enhanced anti-influenza Ab titers and the rapid viral clearance in this coinfection model are ongoing.

We acknowledge the technical assistance of Ann Harmsen, Soo Han, Katie Shampeny, Trenton Bushmaker, Tyronne Markov, and Sara Erikson in conducting the experiments reported herein. We also acknowledge the assistance of Tammy Marcotte and colleagues for animal care expertise at the animal facility at Montana State University. The Idea Network of Biomedical Research Excellence Grant (RR16455) was coordinated by Tim Ford.

The authors have no financial conflicts of interest.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

1

This work was supported by National Institutes of Health Grants HL55002 and RR020185 (to A.G.H.) and from Idea Network of Biomedical Research Excellence Grant RR16455.

3

Abbreviations used in this paper: BALF, bronchoalveolar lavage fluid; LDH, lactate dehydrogenase; iBALT, inducible BALT.

1
Seki, M., K. Yanagihara, Y. Higashiyama, Y. Fukuda, Y. Kaneko, H. Ohno, Y. Miyazaki, Y. Hirakata, J. Kadota, T. Tashiro, S. Kohno.
2004
. Immunokinetics in severe pneumonia due to influenza virus and bacteria coinfection in mice.
Eur. Respir. J.
24
:
143
-149.
2
Alonso, J.-M., A. Guiyoule, M. L. Zarantonelli, F. Ramisse, R. Pires, A. Antignac, A. E. Deghmane, M. Huerre, S. van der Werf, M.-K. Taha.
2003
. A model of meningococcal bacteremia after respiratory superinfection in influenza A virus-infected mice.
FEMS Microbiol. Lett.
222
:
99
-106.
3
Medrano, F. J., M. Montes-Cano, M. Conde, C. de la Horra, N. Respaldiza, A. Gasch, M. J. Perez-Lozano, J. M. Varela, E. J. Calderon.
2005
. Pneumocystis jirovecii in general population.
Emerg. Infect. Dis.
11
:
245
-250.
4
Peterson, J. C., M. T. Cushion.
2005
. Pneumocystis: not just pneumonia.
Curr. Opin. Microbiol.
8
:
393
-398.
5
Gigliotti, F., A. G. Harmsen, T. W. Wright.
2003
. Characterization of transmission of Pneumocystis carinii f. sp. muris through immunocompetent BALB/c mice.
Infect. Immun.
71
:
3852
-3856.
6
Chabe, M., E. Dei-Cas, C. Creusy, L. Fleurisse, N. Respaldiza, D. Camus, I. Durand-Joly.
2004
. Immunocompetent hosts as a reservoir of Pneumocystis organisms: histological and RT-PCR data demonstrate active replication.
Eur. J. Clin. Microbiol. Infect. Dis.
23
:
89
-97.
7
Julkunen, I., T. Sareneva, J. Pirhonen, T. Ronni, K. Melen, S. Matikainen.
2001
. Molecular pathogenesis of influenza A virus infection and virus-induced regulation of cytokine gene expression.
Cytokine Growth Factor Rev.
12
:
171
-180.
8
Wiley, J., A. G. Harmsen.
1995
. CD40 ligand is required for resolution of Pneumocystis carinii pneumonia in mice.
J. Immunol.
155
:
3525
-3529.
9
Wiley, J. A., M. P. Tighe, A. G. Harmsen.
2005
. Upper respiratory tract resistance to influenza infection is not prevented by the absence of either nasal-associated lymphoid tissue or cervical lymph nodes.
J. Immunol.
175
:
3186
-3196.
10
Wiley, J. A., A. Cerwenka, J. R. Harkema, R. W. Dutton, A. G. Harmsen.
2001
. Production of IFNγ by influenza HA-specific CD8 effector T cells influences the development of pulmonary immunopathology.
Am. J. Pathol.
158
:
119
-130.
11
Brown, G. D., J. Herre, D. L. Williams, J. A. Willment, A. S. J. Marshall, S. Gordon.
2003
. Dectin-1 mediates the biological effects of β-glucans.
J. Exp. Med.
197
:
1119
-1124.
12
Vassallo, R., J. E. Standing, A. H. Limper.
2000
. Isolated Pneumocystis carinii cell wall glucan provokes lower respiratory tract inflammatory responses.
J. Immunol.
164
:
3755
-3763.
13
Hahn, P. Y., S. E. Evans, T. J. Kottom, J. E. Standing, R. E. Pagano, A. H. Limper.
2003
. Pneumocystis carinii cell wall β-glucan induces release of macrophage inflammatory protein-2 from alveolar epithelial cells via a lactosylceramide-mediated mechanism.
J. Biol. Chem.
278
:
2043
-2050.
14
Wiley, J. A., A. G. Harmsen.
1999
. Bone marrow-derived cells are required for the induction of a pulmonary inflammatory response mediated by CD40 ligation.
Am. J. Pathol.
154
:
919
-924.
15
Furze, R. C., T. Hussell, M. E. Selkirk.
2006
. Amelioration of influenza-induced pathology in mice by coinfection with Trichinella spiralis.
Infect. Immun.
74
:
1924
-1932.
16
Swain, S. D., N. N. Meissner, A. G. Harmsen.
2006
. CD8 T cells modulate CD4 T-cell and eosinophil-mediated pulmonary pathology in Pneumocystis pneumonia in B-cell deficient mice.
Am. J. Pathol.
168
:
466
-475.
17
Drent, M., N. A. M. Cobben, R. F. Henderson, E. F. M. Wouters, M. van Dieijen-Visser.
1996
. Usefulness of lactate dehydrogenase and its isoenzymes as indicators of lung damage or inflammation.
Eur. Respir. J.
9
:
1736
-1742.
18
Liesenfeld, O., I. R. Dunay, K. J. Erb.
2004
. Infection with Toxoplasma gondii reduces established and developing Th2 responses induced by Nippostrongylus brasiliensis infection.
Infect. Immun.
72
:
3812
-3822.
19
Conti, P., D. Kempuraj, K. Kandere, M. D. Gioacchino, R. C. Barbacane, M. L. Castellani, M. Felaco, W. Boucher, R. Letourneau, T. C. Theoharides.
2003
. IL-10, an inflammatory/inhibitory cytokine, but not always.
Immunol. Lett.
86
:
123
-129.
20
Moore, K. W., R. d. W. Malefyt, R. L. Coffman, A. O’Garra.
2001
. Interleukin-10 and the interleukin-10 receptor.
Annu. Rev. Immunol.
19
:
683
-765.
21
Qureshi, M. H., A. G. Harmsen, B. A. Garvy.
2003
. IL-10 modulates host responses and lung damage induced by Pneumocystis carinii infection.
J. Immunol.
170
:
1002
-1009.
22
Lentsch, A. B., B. J. Czermak, J. A. Jordan, P. A. Ward.
1999
. Regulation of acute lung inflammatory injury by endogenous IL-13.
J. Immunol.
162
:
1071
-1076.
23
Humphreys, I. R., G. Walzl, L. Edwards, A. Rae, S. Hill, T. Hussell.
2003
. A critical role for OX40 in T cell-mediated immunopathology during lung viral infection.
J. Exp. Med.
198
:
1237
-1242.
24
Moyron-Qiroz, J. E., J. Rangel-Moreno, K. Kusser, L. Hartson, F. Sprague, S. Goodrich, D. L. Woodland, F. E. Lund, T. D. Randall.
2004
. Role of inducible bronchus associated lymphoid tissue (iBALT) in respiratory immunity.
Nat. Med.
10
:
927
-934.