Impaired wound healing predisposes obese mice to severe influenza virus infection

Abstract

For the first time, obesity appeared as a risk factor for developing severe 2009 pandemic influenza infection. Given the increase in obesity, there is a need to understand the mechanisms underlying poor outcomes in this population. In these studies, we examined the severity of pandemic influenza virus in obese mice and evaluated antiviral effectiveness. We found that genetically and diet-induced obese mice challenged with either 2009 influenza A virus subtype H1N1 or 1968 subtype H3N2 strains were more likely to have increased mortality and lung pathology associated with impaired wound repair and subsequent pulmonary edema. Antiviral treatment with oseltamivir enhanced survival of obese mice. Overall, these studies demonstrate that impaired wound lung repair in the lungs of obese animals may result in severe influenza virus infection. Alternative approaches to prevention and control of influenza may be needed in the setting of obesity.

Figure 1.

Obese mice have increased mortality during influenza infection. C57BI/6N (n = 28) and diet-induced obese (DIO, n = 28) or genetically obese (ob/ob, n = 28) mice were lightly anesthetized and intranasally infected with (A, B) 10⁵ TCID₅₀ A/California/04/2009 H1N1 (CA/09, pH1N1) or (C) 6.3 × 10⁵ TCID₅₀ A/Hong Kong /68 (HK68, H3N2) influenza virus and monitored for weight loss and mortality. Error bars represent standard deviation. Abbreviations: pH1N1, pandemic influenza A virus subtype H1N1; H3N2, influenza A virus subtype H3N2.

Figure 2.

Changes in immune cell infiltrates in the lungs of obese mice infected with influenza virus. BAL samples were collected on days 3 (A), 6 (B), and 14 (C) post-pH1N1 infection from 2 control and 3 infected mice and flow cytometry performed. Briefly, single-cell populations were collected and single-cell populations were stained with Ly-6G, CD11b, CD11c, F4/80, CD86, NK1.1, CD3, CD4, CD8, and CD69 antibodies directly conjugated to Alexa Fluor 647, Alexa Fluor 700, fluorescein isothiocyanate, phycoerythrin, APC-eFluor combined with LIVE/Dead Fixable Violet Dead Cell Stain Kit. Results are graphed as the number of positive cells per number of lives cells per 1 mL BAL fluid. Error bars represent SD, asterisk (*) indicates significant increase, and (^) indicates significant decrease in cell numbers as compared with C57BI/6 lean infected mice. Number symbol (#) represents significant differences between DIO and ob/ob infected mice. Abbreviations: BAL, bronchoalveolar lavage; pH1N1, pandemic influenza A virus subtype H1N1.

Figure 3.

Decreased cell proliferation in the lungs of obese mice infected with influenza virus. At days 5, 6, and 14 postinfection, lungs were collected from deeply anesthetized and formalin-perfused mice and paraffin-embedded. Sections (4 μm thick) were stained with hematoxylin and eosin (A) or for the cell proliferation marker Ki-67 (B). Representative pictures of each group are shown at 4× (A) and 20× (B) magnification. C, Digital images of the Ki-67 slides were obtained, and the percentage of positive nuclei in 4 random sections of the lung for each animal were determined with ImageScope using a nuclear-based algorithm. Error bars represent SD, and asterisk (*) indicates significant decrease in percent cells positive as compared with C57BI/6N lean infected mice. Number symbol (#) represents significant differences between DIO and ob/ob infected mice. Abbreviations: dpi, days postinfection; DIO, diet-induced obese; ob/ob, genetically obese.

Figure 4.

Influenza virus-infected obese mice have increased lung protein and albumin levels. BAL samples were collected on days 3 and 6 post-pH1N1 infection from 2 control and 3 infected mice total protein concentrations determined by BCA assays (A) or albumin levels by enzyme-linked immunoassay (B). Error bars represent standard deviation, and asterisk (*) indicates significant increase in BAL albumin levels as compared with C57BI/6N lean infected mice. Abbreviations: BAL, bronchoalveolar lavage; pH1N1, pandemic influenza A virus subtype H1N1; BCA, bradford colorimetric assay; DIO, diet-induced obese; ob/ob, genetically obese.

Figure 5.

Enhanced lethality in obese mice is independent of increased viral titers, but mice can still be protected with oseltamivir. A, At 3 and 6 days postinfection, lung homogenates were monitored for viral titers by TCID₅₀ analysis on MDCK cells. B, pH1N1 DIO or ob/ob mice (n = 5 per group) were orally administered either 20 or 100 mg/kg oseltamivir virus 8 hours preinfection, then every 12 hours after infection with 10⁵ TCID₅₀ pH1N1 influenza virus for 5 days, and then monitored for morbidity over 12 days. Error bars represent SD. Abbreviations: TCID₅₀, 50% tissue culture infectious dose; MDCK, Madin-Darby canine kidney; DIO, diet-induced obese; ob/ob, genetically obese; pH1N1, pandemic influenza A virus subtype H1N1.

Figure 6.

Decreased lung histology and increased Ki-67 staining in the lungs of influenza virus–infected mice treated with oseltamivir. A, At days 6 and 14 postinfection, lungs were collected from deeply anesthetized and formalin-perfused mice and paraffin-embedded. Sections (4 μm thick) were stained with hematoxylin and eosin, and representative pictures of each group are shown at 20× magnification. B, Digital images of the Ki-67 slides were obtained, and the percentage of positive nuclei in 4 random sections of the lung for each animal were determined with ImageScope using a nuclear-based algorithm. Error bars represent SD. Abbreviations: dpi, days postinfection; Olestam, oseltamivir; DIO, diet-induced obese; ob/ob, genetically obese.