Coinfection with Blood-Stage Plasmodium Promotes Systemic Type I Interferon Production during Pneumovirus Infection but Impairs Inflammation and Viral Control in the Lung

Abstract

Acute lower respiratory tract infections (ALRTI) are the leading cause of global childhood mortality, with human respiratory syncytial virus (hRSV) being a major cause of viral ALRTI in young children worldwide. In sub-Saharan Africa, many young children experience severe illnesses due to hRSV or Plasmodium infection. Although the incidence of malaria in this region has decreased in recent years, there remains a significant opportunity for coinfection. Recent data show that febrile young children infected with Plasmodium are often concurrently infected with respiratory viral pathogens but are less likely to suffer from pneumonia than are non-Plasmodium-infected children. Here, we hypothesized that blood-stage Plasmodium infection modulates pulmonary inflammatory responses to a viral pathogen but does not aid its control in the lung. To test this, we established a novel coinfection model in which mice were simultaneously infected with pneumovirus of mice (PVM) (to model hRSV) and blood-stage Plasmodium chabaudi chabaudi AS (PcAS) parasites. We found that PcAS infection was unaffected by coinfection with PVM. In contrast, PVM-associated weight loss, pulmonary cytokine responses, and immune cell recruitment to the airways were substantially reduced by coinfection with PcAS. Importantly, PcAS coinfection facilitated greater viral dissemination throughout the lung. Although Plasmodium coinfection induced low levels of systemic interleukin-10 (IL-10), this regulatory cytokine played no role in the modulation of lung inflammation or viral dissemination. Instead, we found that Plasmodium coinfection drove an early systemic beta interferon (IFN-β) response. Therefore, we propose that blood-stage Plasmodium coinfection may exacerbate viral dissemination and impair inflammation in the lung by dysregulating type I IFN-dependent responses to respiratory viruses.

INTRODUCTION

Acute lower respiratory tract infections (ALRTI) caused by bacteria and viruses are the leading cause of global childhood mortality. Respiratory syncytial virus (RSV) is a major human pathogen. It is estimated that in 2005, 3.4 million severe cases of RSV-associated ALRTI occurred globally in children <5 years of age, causing approximately 66,000 to 199,000 deaths; 99% of these occurred in developing countries (1). In sub-Saharan Africa, hundreds of thousands of children <5 years of age experience severe illnesses each year, not only after infection with RSV, but also due to the Plasmodium species that cause malaria (2, 3). Thus, the combined disease burden for RSV and Plasmodium infections among young children living in sub-Saharan Africa has been very large in the recent past.

However, between 2000 and 2012, the incidence of malaria was reduced by 31% and deaths were reduced by 49% in the WHO African Region through the scaling up of various public health initiatives (3). Despite these great achievements, in 2012, 482,000 children under five died as a result of malaria in this region (3). Estimates for the RSV disease burdens over the same time period in sub-Saharan Africa are currently not available. However, a recent study of RSV in seven different low- to middle-income countries, including Kenya and South Africa, suggests that RSV infection rates among children remain high (4). Together, these studies indicate that despite some clear public health successes since 2000, Plasmodium and RSV infections remain common among children in sub-Saharan Africa.

As the incidence of malaria has dropped over the past 14 years, it has also been noted that the incidence rates for certain other infectious diseases have dropped in sub-Saharan Africa, most notably those caused by nontyphoidal Salmonella (NTS) serovars (5). Indeed, enhanced susceptibility in these Gram-negative bacterial pathogens has also been shown in murine models of malaria (6). These in vivo models also indicate that nonlethal Plasmodium infection elicits a systemic interleukin-10 (IL-10) response, which significantly inhibits subsequent and concurrent immune responses to heterologous pathogens (7–10). In particular, a recent report shows that macrophage-dependent Plasmodium-induced systemic IL-10 regulates immunity to Salmonella spp. in the gastrointestinal tract (7). Whether such regulatory processes might also inhibit immune responses to concurrent respiratory viral infection remains unclear. Interestingly, recent data from a region of Tanzania where malaria is endemic showed that febrile young children infected with Plasmodium were often concurrently infected with respiratory viral pathogens but were 50% less likely to suffer from pneumonia than were non-Plasmodium-infected children (11). These epidemiological data support our hypothesis that blood-stage Plasmodium infection modulates pulmonary inflammatory responses during respiratory viral infection but does not itself facilitate viral control.

In this paper, we present a novel mouse model of RSV-Plasmodium coinfection. Using this model, we demonstrate that blood-stage Plasmodium coinfection reduces clinical symptoms and inflammatory responses associated with viral infection of the lung, but, importantly, also impairs viral control in this organ. We show that Plasmodium-induced immunosuppression in the lung does not depend on IL-10, as was previously reported for systemic coinfection with other Plasmodium species, or gastrointestinal infection with Salmonella but is associated instead with an accelerated systemic beta interferon (IFN-β) response.

MATERIALS AND METHODS

Plasmodium and viral infections.

Specific-pathogen-free female C57BL/6J mice, age 6 to 10 weeks, were used for all experiments. All procedures were approved by The University of Queensland Animal Ethics Committee. To initiate viral respiratory tract infection, each mouse was intranasally (i.n.) inoculated with 5 to 10 PFU of pneumovirus of mice (PVM) (strain J3666) (12) or vehicle (Dulbecco's modified Eagle's medium [DMEM] plus 10% fetal calf serum) under light isoflurane anesthesia. To initiate blood-stage Plasmodium infection, Plasmodium chabaudi chabaudi AS (PcAS) parasites were used after one in vivo passage in wild-type C57BL/6J mice. Each mouse was infected intravenously (i.v.) via a lateral tail vein injection with 10⁵ freshly prepared packed red blood cells (pRBC). To establish coinfection, the mice were first inoculated with Plasmodium, lightly anesthetized, and then inoculated with PVM.

Assessing peripheral blood parasitemia.

Blood parasitemia was measured in Diff Quik-stained (Lab Aids, Narrabeen, NSW, Australia) thin blood smears obtained from tail bleeds only when the passage mice were being assessed for the presence of parasites. For an analysis of parasitemia in the experimental mice, a previously established flow cytometric method (13–16) was employed to measure parasitemia more rapidly. Briefly, a single drop of blood from a tail bleed or cardiac puncture was diluted in 250 μl of RPMI containing 5 U/ml heparin sulfate. The diluted blood was simultaneously stained with 5 μM Syto84 (Life Technologies) to detect RNA/DNA and 10 μM Hoechst 33342 (Sigma) to detect DNA for 30 min in the dark at room temperature. The staining was quenched with 10 volumes of ice-cold RPMI, and the samples were immediately analyzed by flow cytometry using an LSRFortessa fluorescence-activated cell sorter (FACS) analyzer (BD Biosciences) and FlowJo software (Tree Star, CA, USA). pRBC were readily detected as being Hoechst 33342⁺ Syto84⁺. Reticulocytes were readily distinguished from pRBC in PcAS-infected mice, even though they contained slightly more DNA/RNA than that seen in reticulocytes in naive or PVM-infected mice (Fig. 1A).

FIG 1.

PVM coinfection does not impact the course of blood-stage PcAS infection. Wild-type mice (n = 5 per group) were coinfected with PVM (i.n., 10 PFU) and PcAS (i.v., 10⁵ pRBC) or were infected singly with each pathogen as controls. (A) Blood parasitemias were measured by flow cytometric staining of peripheral blood with Hoechst 33342 (for DNA) and Syto84 (for DNA and RNA) in infected and coinfected mice on days indicated (representative FACS plots of RBC shown from day 10 p.i., with gates and numbers indicating percent RBC containing PcAS). (B) Spleen weights were assessed on day 7 p.i. in cohorts of uninfected, singly infected, and coinfected mice (n = 5). The PcAS-infected mice were also assessed on day 14 p.i. NS, nonsignificant (Mann-Whitney test). The data are representative of three independent experiments and show the mean ± standard error of the mean (SEM).

Sample extraction and processing.

After the mice were euthanized by pentobarbitone overdose, blood was obtained by cardiac puncture and centrifuged at 13,000 × g and 4°C for 15 min twice to collect serum, which was stored at −80°C. Bronchoalveolar lavage (BAL) was performed, as described previously (17), the BAL fluid (BALF) was centrifuged at 5,000 × g and 4°C for 5 min, and the supernatant was stored at −80°C until cytokine analysis. Red blood cells were removed from the cell pellet with Gey's lysis buffer. The cells were washed twice in phosphate-buffered saline (PBS) and a cytospin prepared (StatSpin CytoFuge). The cells were air-dried, fixed with 100% (vol/vol) methanol for 15 min, and stained with May-Grünwald Giemsa stain. Typically, 300 to 400 cells were counted for each cytospin sample. The right lobe of the lung was used for histology (fixed in 10% formalin neutral buffer overnight before storage in 70% ethanol). The serum and BALF samples were assessed for IL-6, tumor necrosis factor (TNF), IL-10, monocyte chemoattractant protein-1 (MCP-1), IFN-γ, and IL-12p70 using cytometric bead arrays (CBA) (mouse inflammation kit; BD) or IFN-α or IFN-β using a ProcartaPlex Luminex xMAP technology-based bead array (eBioscience), according to the manufacturers' instructions.

Assessing viral load via lung immunohistochemistry.

Paraffin-embedded sections were prepared, as previously described (12). Briefly, 5-μm-thick tissue sections were pretreated with 10% normal goat serum for 30 min. Primary antibodies against PVM (1:8,000 dilution; the antiserum against PVM protein G was kindly provided by Ulla Buchholz) were added and incubated overnight at 4°C. The sections were washed three times in PBS–0.05% Tween 20 before incubation with biotinylated anti-mouse polyclonal antibody (Invitrogen). This step was then repeated but with streptavidin-alkaline phosphatase. After 60 min of incubation at room temperature, the sections were washed three times and immunoreactivity development was performed with Fast Red (Sigma-Aldrich). The sections were counterstained with hematoxylin before mounting with Glycergel (Dako). The percentage of PVM-positive airway epithelial cells (AEC) was quantified in 5 airways per mouse (ScanScope XT; Aperio). To enumerate the total PVM-infected cells across the whole lung, positive pixel counting was performed across a single lung section for each mouse (Positive Pixel Count version 9 algorithm, ImageScope; Aperio).

In vivo IL-10 receptor blockade.

Monoclonal antibody-secreting hybridomas were grown in 5% (vol/vol) fetal calf serum and RPMI containing 10 mmol/liter l-glutamine, 200 U/ml penicillin, and 200 mg/ml streptomycin. Purified antibody was prepared from the culture supernatants by protein G column purification (Amersham, Uppsala, Sweden), followed by endotoxin removal (Mustang membranes; Pall, East Hills, NY). Anti-IL-10 receptor (anti-IL-10R) blocking monoclonal antibody (1B1.3a) and control IgG were administered in 0.25-mg doses via intraperitoneal (i.p.) injection in 200 μl of 0.9% sodium chloride (Baxter) per mouse on days 0, 3, and 6 postinfection.

Statistical analysis.

Comparisons between the two groups were performed using nonparametric Mann-Whitney tests, unless stated otherwise, in which case Student t tests for Gaussian distributions were employed. A P value of <0.05 was considered significant. All statistical analyses were performed using the GraphPad Prism 6 software.

RESULTS

PVM coinfection does not alter the course of blood-stage PcAS infection.

To model the possible interplay between viral respiratory infection and blood-stage Plasmodium infection in vivo, we established a murine coinfection model in which C57BL6/J mice were first infected intravenously with blood-stage P. chabaudi chabaudi AS (PcAS) parasites and then immediately infected intranasally with pneumovirus of mice (PVM). This combination of pathogens was chosen because both replicate well in mice, and neither causes lethal disease when administered alone, thus offering the possibility of observing increased morbidity during coinfection. We also chose this virus-parasite combination because each pathogen reaches the peak of its infection at approximately the same time in C57BL/6J mice, at around 9 to 10 days postinfection (18). Therefore, we expected neither pathogen to assume dominance during coinfection simply due to an overwhelmingly higher replication rate than that of the other.

After the initiation of coinfection, we noted first that the course of PcAS peripheral blood parasitemia was essentially identical in coinfected mice and in those receiving PcAS alone (Fig. 1A). Moreover, the degree of splenomegaly observed in these two groups of mice, a surrogate marker for an active immune response to blood-borne or systemic pathogens, was equivalent at days 7 and 14 postinfection (p.i.) (Fig. 1B). Together, these data suggest that pulmonary PVM infection had not impacted the ability of the host to respond to and control primary PcAS infection.

PcAS coinfection abrogates PVM-induced weight loss and pulmonary inflammatory responses.

We next assessed the impact of coinfection on PVM-induced pulmonary immune responses and disease symptoms. First, mice infected with PVM alone lost approximately 15% of their starting body weight by day 8 p.i. (Fig. 2A), an outcome consistent with previous published data (12). Strikingly, weight loss was significantly reduced in coinfected mice and was absent in uninfected mice and those infected with PcAS alone (Fig. 2A). Second, inflammatory cytokine levels, particularly those of IL-6, MCP-1, IL-10, TNF, and IFN-γ, were highly upregulated in the BALF samples from the mice infected with PVM alone but were significantly reduced in the coinfected mice (Fig. 2B). Also noteworthy, PcAS infection alone induced almost no BALF cytokine responses, suggesting that the movement of blood-borne parasites through lung microvasculature was not sufficient to drive such local immune responses (Fig. 2B). Finally, while PVM infection alone induced a significant recruitment of immune cells, such as lymphocytes, eosinophils, and neutrophils, into the airways (Fig. 2C), this did not occur during PcAS infection alone or, more importantly, during coinfection (Fig. 2C). Together, these data indicate that PVM-associated weight loss and pulmonary cytokine and cellular inflammatory responses were substantially suppressed by coinfection with blood-stage PcAS parasites.

FIG 2.

PcAS coinfection abrogates PVM-induced weight loss and pulmonary inflammatory responses. Wild-type mice (n = 5 per group) were coinfected with PVM (i.n., 10 PFU) and PcAS (i.v., 10⁵ pRBC) or were infected singly with each pathogen. (A) Weight loss over time was assessed as a percentage of the average weight of the group (n = 5) at the time of infection. The data presented are from one of two independent experiments with similar results. (B) Cytokine levels in BALF at the peak of PcAS infection. The data are pooled from three independent experiments showing similar results (n = 15 per group, n = 5 per group per experiment). (C) Cells present in BALF samples at the peak of PcAS infection (n = 5). The data are representative of one of two independent experiments with similar results and show the mean ± SEM. *, P < 0.05; **, P < 0.01 (Mann-Whitney test).

PcAS coinfection increases PVM dissemination into lung parenchyma.

Since PcAS coinfection had suppressed PVM-dependent lung inflammation, we next assessed the impact of coinfection on viral loads in this organ by immunostaining for the virus. At the peak of PcAS infection, we observed a trend toward reduced PVM infection in airway epithelial cells (Fig. 3A and B). However, we found evidence of more PVM in the lung parenchyma of coinfected mice than in those infected with PVM alone (Fig. 3A and C). These data suggest that coinfection with PcAS parasites facilitated a greater dissemination of PVM into the parenchyma of the lung. We also noted that PVM was undetectable in the lungs of singly or coinfected mice by day 14 p.i., as assessed by immunostaining (data not shown). These data indicate that although pulmonary viral burdens and dissemination were exacerbated during coinfection, the mice were capable of ultimately clearing PVM while experiencing blood-stage PcAS infection.

FIG 3.

PcAS coinfection increases PVM dissemination into lung parenchyma. Wild-type mice were coinfected with PVM (i.n., 10 PFU) and PcAS (i.v., 10⁵ pRBC) or were infected singly with each pathogen. (A) Representative images of immunostaining for PVM in lung tissue at the peak of PcAS infection, with PVM-infected cells shown in red. The boxed regions in the two left images are shown at higher magnification in the images on the right. The images are representative of two independent experiments. (B) Percentage of airway epithelial cells infected with PVM (n = 5 per group per experiment). The data are pooled from two independent experiments (Student's t test, P = 0.0665) and show the mean ± SEM. +ve, positive. (C) Enumeration of total PVM staining throughout entire single section of lung tissue via positive pixel counting (n = 5). The data are representative of one of two independent experiments with similar results and show the median values. *, P < 0.05 (Mann-Whitney test).

Systemic IL-10 during PcAS coinfection does not regulate PVM-induced pulmonary immune responses.

We next sought to determine a mechanism by which Plasmodium infection prevented virally induced lung inflammatory responses. Recent reports indicate that Plasmodium suppresses immune responses to concurrent infections via IL-10, both during systemic infection with a second lethal Plasmodium species (9, 10) and locally in the gastrointestinal tract during nontyphoidal salmonellosis (7, 8). Therefore, we hypothesized that Plasmodium-induced systemic IL-10 inhibits immune responses to respiratory viral coinfection.

First, we found that IL-10 was expressed systemically at moderate levels during coinfection and in PcAS infection alone but not in mice infected with PVM alone (Fig. 4A). Interestingly, this contrasted with the localized IL-10 responses in the lung, which were evident more in PVM-infected mice than in either PcAS-infected or coinfected mice (Fig. 2B). Next, we studied the impact of a systemic blockade of IL-10R signaling upon coinfection (Fig. 4B to D). This treatment significantly reduced splenomegaly (Fig. 4B) and reduced peripheral blood parasitemia (Fig. 4C) compared to those effects in coinfected mice given control IgG. However, we observed no restoration of immune cellular recruitment to the airways in coinfected mice upon IL-10R blockade (Fig. 4D), while it was apparent that this treatment exacerbated BALF responses in the mice infected with PVM alone (Fig. 4D). Finally, we noted no effect of IL-10R blockade on PVM dissemination throughout the lung parenchyma in the coinfected mice (data not shown). These data suggest that coinfection with blood-stage PcAS exacerbates PVM dissemination and suppresses inflammatory immune responses in the lung independently of IL-10.

FIG 4.

Low-level systemic IL-10 produced during PcAS coinfection does not regulate PVM-induced pulmonary responses. Wild-type mice were coinfected with PVM (i.n., 10 PFU) and PcAS (i.v., 10⁵ pRBC) or were infected singly with each pathogen. (A) Serum IL-10 protein levels measured at peak of PcAS infection (n = 5 per group per experiment). The data are pooled from three independent experiments. ***, P < 0.001 (Mann-Whitney test). (B to D) Wild-type mice (n = 5) were coinfected with PVM and PcAS or infected singly with PVM. The infected mice received either anti-IL-10R or control (Ctrl) IgG throughout the infection. At the peak of PcAS infection, assessments of spleen weight (B), PcAS parasitemia (C), and the recruitment of cells into the BALF (D) were made. This experiment was conducted once, and the data show the median values. *, P < 0.05; **, P < 0.01. NS, not significant (Mann-Whitney test).

PcAS coinfection drives an early systemic IFN-β response.

It has been shown that IFN-α/β transcriptional upregulation occurs transiently toward the end of the first week of infection in the lungs of PVM-infected mice (19). This coincides, at least in neonatal mice, with enhanced plasmacytoid dendritic cell (pDC) numbers, which depend upon IFN regulatory factor 7 (IRF7), Toll-like receptor 7 (TLR7), and MyD88 signaling (12). These type I IFN-dependent immune processes mediate PVM control and cellular inflammatory responses in the lung, as well as drive weight loss (12). Therefore, we hypothesized that PcAS coinfection might dysregulate this transient and highly regulated type I IFN response. To test this, we studied IFN-α and IFN-β protein levels during the early stages of coinfection, when the level of PcAS parasitemia was low but detectable (∼0.5%). Although we saw no increase in IFN-α/β levels in the lung tissues of the coinfected mice compared to those in the singly infected mice (data not shown), we did observe an early systemic IFN-β (but not IFN-α) response in the coinfected mice, which was absent from the singly infected mice (Fig. 5). These data provide evidence that coinfection with PcAS drives a unique systemic type I IFN response that does not normally occur during infection with PVM alone.

FIG 5.

Coinfection with PcAS and PVM drives an early systemic IFN-β response. Wild-type mice were coinfected with PVM (i.n., 10 PFU) and PcAS (i.v., 10⁵ pRBC) or were infected singly with each pathogen (n = 3 to 5). The serum IFN-β protein levels were measured when the PcAS parasitemias were at 0.5%. The data are representative of two independent experiments and show the mean ± SEM (one-way analysis of variance [ANOVA], Dunnett's multiple-comparison test). *, P < 0.05; **, P < 0.01.

DISCUSSION

In this study, we established a murine coinfection model with which to explore possible competition and interplay between a respiratory virus and blood-stage Plasmodium parasites in a mammalian host. In designing this model, we specifically chose two pathogens known to infect and replicate efficiently in inbred C57BL/6J mice. We chose these species to avoid infection with a large bolus of a poorly replicative pathogen, which might be expected to stimulate the innate immune system nonphysiologically and thus perturb subsequent pathogenesis. Second, we chose two pathogens with similar growth rates in vivo to avoid the chance that either pathogen would dominate due to a much higher replication rate. We believe our model has been instructive, because both pathogens were clearly able to replicate simultaneously in the host. Indeed, the blood-stage Plasmodium parasite growth appeared unaffected, while PVM growth was moderately increased during coinfection. Therefore, we believe this model may be of general utility for exploring the interactions between respiratory viruses and blood-stage Plasmodium parasites.

A key aspect of our model is the simultaneous nature of the two infections. However, it is clearly more likely that there is a certain time period, perhaps days, between RSV and Plasmodium infections in regions where malaria is endemic. Therefore, it would be of interest in the future to determine if infection timing and/or parasite density affects the capacity of Plasmodium to exert its modulatory effects against pulmonary viral infections. Interestingly, recent data suggest that the density of Plasmodium parasites in the bloodstream does not significantly affect the incidence of coinfection in children (11). Whether parasite density affects susceptibility to pulmonary disease caused by respiratory pathogens remains untested.

In this study, we found that Plasmodium coinfection substantially suppressed inflammatory cytokine production and cellular recruitment to the lung in response to a respiratory virus and furthermore that this suppression did not require IL-10. Importantly, however, our control data demonstrate that IL-10 did suppress parasite control and exacerbated splenomegaly during coinfection. Moreover, our data are consistent with IL-10 playing a protective role against viral bronchiolitis in single-pathogen-infected mice. Our data appear to contrast with recent reports indicating that blood-stage Plasmodium suppresses gastrointestinal immune responses to Salmonella via macrophage-dependent IL-10 (7, 8) and systemic immune responses to a second Plasmodium infection, again via IL-10 (9, 10). Thus, we speculate here that Plasmodium-induced IL-10 might be suppressive in some tissues, such as the gastrointestinal (GI) tract and the spleen, but perhaps not in others, such as the lung.

Given that Plasmodium suppressed viral bronchiolitis independently of IL-10, the exact mechanism by which this occurred remains to be elucidated. We previously showed that the innate immune detection of PVM in the lung is dependent on IRF7-, MyD88-, and TLR7-mediated activation of plasmacytoid dendritic cells (pDC), which promote airway inflammation to control the viral load (12). Given the recent data revealing the transient nature of IFN-α/β production in the lungs of PVM-infected mice (19), we hypothesized that Plasmodium coinfection disrupts type I IFN-dependent innate immune responses to PVM. Consistent with this, we detected an early systemic IFN-β response in the serum of coinfected mice that was absent in the singly infected mice. We speculate here that PcAS-PVM coinfection might boost pDC activation (12, 20, 21) via multiple pattern recognition receptors (e.g., TLR7 by PVM, and TLR9 and STING by Plasmodium-derived hemozoin and AT-rich DNA [12, 20, 22]), which would deplete their numbers in vivo (23) and ultimately disrupt pDC-dependent inflammatory responses in the lung. Given that pDC-dependent processes are required to elicit PVM-specific cytotoxic T lymphocytes in the lung (12), this may explain how Plasmodium coinfection curtailed pulmonary lymphocytic responses in our experiments.

Nevertheless, other mechanisms may also account for the Plasmodium-induced suppression of lung immune responses to PVM. For example, it has been shown that PcAS infection can enhance Foxp3⁺ regulatory T-cell (Treg) responses in peripheral tissues, such as the liver (24). Therefore, it is at least theoretically possible that PcAS coinfection also boosts Treg responses in the lung. Such responses might impair lung inflammation via mechanisms that depend not on IL-10 but on molecules such as cytotoxic-T-lymphocyte-associated antigen 4 (CTLA-4) (25). Future experiments are required to determine whether PcAS coinfection modulates Treg responses in the lungs of PVM-infected mice.

In this study, we chose to examine PVM-PcAS coinfection in a single genetically inbred mouse strain, C57BL/6J. However, it would be of interest in the future to study the interaction between these two pathogens in other genetic backgrounds, for example, in BALB/c mice, in which proinflammatory responses are different from those seen in C57BL/6J mice. Such experiments are clearly of relevance to human coinfections, in which host genetic diversity is likely to play a significant role in the determination of coinfection outcomes.

Given that concurrent infection with respiratory viruses and Plasmodium frequently occurs in children under the age of five in sub-Saharan Africa, we believe that it is of value to model the potential interplay between these two pathogens in mammalian hosts. Our data suggest that under certain circumstances, by harboring Plasmodium parasites in the bloodstream, some individuals might be partially protected from virally induced ALRTI but might experience greater and more disseminated viral infection in the lung. It may be of interest to determine from current or future data sets how disease severity caused by respiratory viruses changes as Plasmodium parasite density in young children decreases across the WHO African Region.