Lung development and the host response to influenza A virus are altered by different doses of neonatal oxygen in mice

Abstract

Oxygen exposure in preterm infants has been associated with altered lung development and increased risk for respiratory viral infections later in life. Although the dose of oxygen sufficient to exert these changes in humans remains unknown, adult mice exposed to 100% oxygen between postnatal days 1-4 exhibit alveolar simplification and increased sensitivity to influenza virus infection. Additionally, two nonlinear thresholds of neonatal oxygen exposures were previously identified that promote modest (between 40% and 60% oxygen) and severe (between 80% and 100% oxygen) changes in lung development. Here, we investigate whether these two thresholds correlate with the severity of lung disease following respiratory viral infection. Adult mice exposed to 100% oxygen at birth, and to a lesser extent 80% oxygen, demonstrated enhanced body weight loss, persistent inflammation, and fibrosis following infection compared with infected siblings exposed to room air at birth. In contrast, the host response to infection was indistinguishable between mice exposed to room air and 40% or 60% oxygen. Interestingly, levels of monocyte chemoattractant protein (MCP)-1 were equivalently elevated in infected mice that had been exposed to 80% or 100% oxygen as neonates. However, reducing levels of MCP-1 using heterozygous Mcp-1 mice did not affect oxygen-dependent changes in the response to infection. Thus lung development and the host response to respiratory viral infection are disrupted by different doses of oxygen. Our findings suggest that measuring lung function alone may not be sufficient to identify individuals born prematurely who have increased risk for respiratory viral infection.

Fig. 1.

Neonatal hyperoxia disrupts lung structure and the immune response to viral infection in adult mice. A: representative hematoxylin-eosin stained lung tissues from adult mice depicting the doses of oxygen that cause no [room air (RA) - 40% oxygen], modest (60–80% oxygen), or severe (100% oxygen) changes in lung structure and function. Scale bar, 100 μM. Exposure to 100% oxygen also causes increased sensitivity to influenza virus. B: cartoon model depicting the 2 experimental paradigms used to test whether the high (80% oxygen vs. 100% oxygen) and/or modest (40% oxygen vs. 60% oxygen) thresholds of oxygen exposure increase sensitivity to influenza A virus infection.

Fig. 2.

Exposure to the high threshold of neonatal oxygen disrupts viral clearance and leukocyte recruitment in adult mice upon infection with influenza virus. Adult female mice exposed to RA, 80% oxygen, and 100% oxygen or RA, 40% oxygen, and 60% oxygen at birth were intranasally infected with a nonlethal dose of influenza A virus. Lung virus titers in serially diluted whole lung homogenates and total number of leukocytes in the bronchoalveolar lavage (BAL) fluid were measured on postinfection day 7. A: viral titers were significantly increased in infected mice exposed to the high threshold of neonatal oxygen compared with infected siblings exposed to RA at birth. Each dot represents the virus titer in an individual mouse (n ≥ 4 mice per group, *P < 0.05). VFU, viral foci units. B: total number of leukocytes in the BAL fluid of infected mice exposed to 100% oxygen at birth was significantly elevated compared with infected mice exposed to RA at birth (n ≥ 4 mice per group, *P < 0.05).

Fig. 3.

Exposure to the high threshold of neonatal oxygen promotes fibrotic scarring in infected adult mice. Adult female mice exposed to RA, 80% oxygen, and 100% oxygen at birth were intranasally infected with a nonlethal dose of influenza A virus. A: lungs were harvested on postinfection day 14 and were stained with either Gomori's Trichrome stain or antibodies against α-smooth muscle actin (α-SMA; red), followed by counterstaining with 4′,6-diamidino-2-phenyl-indole (DAPI) (blue). Each image is representative of 5 mice examined per group. B: total collagen protein deposition in the lungs of infected adult mice exposed to RA, 80% oxygen, and 100% oxygen at birth was measured by the Sircol assay on postinfection day 14 (n = 5 mice per group, *P < 0.05). C: whereas all mice were similar in body weight before infection, the percentage of mean body weight lost was often significantly greater in infected adult mice exposed to 100% oxygen at birth compared with infected siblings exposed to RA at birth (n ≥ 14 mice per group, *P < 0.05).

Fig. 4.

Levels of monocyte chemoattractant protein (MCP)-1 are increased in infected mice exposed to 80% and 100% oxygen at birth. Adult female mice exposed to RA, 80% oxygen, and 100% oxygen or RA, 40% oxygen, and 60% oxygen at birth were intranasally infected with a nonlethal dose of influenza A virus. Levels of MCP-1 in the BAL fluid were determined on postinfection days 5 and 7. Significantly elevated levels of MCP-1 were equivalently observed on postinfection day 7 in mice exposed to 80% and 100% oxygen at birth (n = 5–8 mice per group, *P < 0.05).

Fig. 5.

Levels of MCP-1 are reduced in Mcp-1 heterozygous mice exposed to the high threshold of neonatal oxygen upon influenza virus infection. Mcp-1 wild-type (+/+) and heterozygous (+/−) mice exposed to 100% oxygen at birth were intranasally infected with a nonlethal dose of influenza A virus. A: PCR genotyping of Mcp-1^+/+, Mcp-1^+/−, and knockout mice (−/−) [(−) = negative control]. B: mean percent body weights of Mcp-1^+/+ and Mcp-1^+/− mice were similar through postinfection day 7. C: levels of MCP-1 in the BAL fluid were determined on postinfection days 5 and 7. Levels of MCP-1 were significantly reduced on postinfection day 5 in Mcp-1^+/− mice (n = 4–6 mice per group, *P < 0.05).

Fig. 6.

Reducing levels of MCP-1 using heterozygous Mcp-1 mice does not affect oxygen-dependent changes in the host response to infection. Mcp-1 wild-type (+/+) and heterozygous (+/−) mice exposed to 100% oxygen at birth were intranasally infected with a nonlethal dose of influenza A virus. Lung virus titers in serially diluted whole lung homogenates and total number of leukocytes in the BAL fluid were measured on postinfection days 5 and 7. Viral titers (each dot represents the virus titer in an individual mouse) (A) and total number of leukocytes (B) were similar between infected Mcp-1^+/+ and Mcp-1^+/− mice exposed to the high threshold of neonatal oxygen (n = 4–6 mice per group).

Fig. 7.

Reducing levels of MCP-1 using heterozygous Mcp-1 mice does not attenuate oxygen-dependent changes in fibrotic scarring. Mcp-1 wild-type (+/+) and heterozygous (+/−) mice exposed to 100% oxygen at birth were intranasally infected with a nonlethal dose of influenza A virus. A: lungs were harvested on postinfection day 14 and were stained with either Gomori's Trichrome stain or antibodies against α-SMA (red), followed by counterstaining with DAPI (blue). Each image is representative of 5 mice examined per group. B: total collagen protein deposition in the lungs of infected mice exposed to 100% oxygen at birth was measured by the Sircol assay on postinfection day 14 (n = 5 mice per group). C: percentage of mean body weight lost was not different between infected Mcp-1^+/+ and Mcp-1^+/− mice exposed to 100% oxygen at birth (n ≥ 7 mice per group).