In vivo imaging of the pathophysiological changes and neutrophil dynamics in influenza virus-infected mouse lungs

Abstract

The pathophysiological changes that occur in lungs infected with influenza viruses are poorly understood. Here we established an in vivo imaging system that combines two-photon excitation microscopy and fluorescent influenza viruses of different pathogenicity. This approach allowed us to monitor and correlate several parameters and physiological changes including the spread of infection, pulmonary permeability, pulmonary perfusion speed, number of recruited neutrophils in infected lungs, and neutrophil motion in the lungs of live mice. Several physiological changes were larger and occurred earlier in mice infected with a highly pathogenic H5N1 influenza virus compared with those infected with a mouse-adapted human strain. These findings demonstrate the potential of our in vivo imaging system to provide novel information about the pathophysiological consequences of virus infections.

Keywords: in vivo imaging; influenza virus; lung; neutrophil dynamics; two-photon excitation microscopy.

Fig. 1.

Quantitative microscopic analysis of virus-infected cells in living mouse lung. (A) In vivo imaging of influenza virus-infected lung. Mice were intranasally infected with MA-Venus-H5N1 or MA-Venus-PR8 virus and observed on different days postinfection (dpi). Green indicates virus-infected cells. At the indicated time points, fluorescent dextran (blue) and PE-conjugated anti-mouse Ly-6G antibody (red) were i.v. administered to visualize lung architecture and neutrophils, respectively. The lung architecture was used for correcting respiratory motion artifacts in the movies. (Scale bar, 50 µm.) See also Movie S1. (B) Median numbers of influenza virus-infected cells (Venus-positive cells) in the fields of view at the indicated time points. Each dot represents the data obtained from an individual animal. See also SI Appendix, Figs. S1 and S2. Horizontal bars indicate the median. Asterisks indicate P < 0.05 compared with naïve (uninfected) mice (Dunnett’s test).

Fig. 2.

Pulmonary permeability in virus-infected lungs. (A) Time-lapse imaging of dextran leakage in MA-Venus-H5N1– or MA-Venus-PR8–infected mice 4 d postinfection at different time points after dextran injection. Green indicates virus-infected cells. Fluorescent dextran (red) was i.v. injected into the mice during image acquisition. (Scale bar, 50 µm.) See also Movie S2. (B) Area and fluorescent signal intensity of leaked dextran in lung infected with MA-Venus-H5N1 or MA-Venus-PR8 virus. The plot shows the percentage of extravascular area positive for the dextran signal in each field of view. Each data point is color-coded to reflect the intensity level of the extravascular dextran. (C) Quantification of pulmonary edema in virus-infected mice at different time points. Each dot represents the data obtained from an individual animal. Horizontal bars indicate the median. The asterisk indicates P < 0.05 compared with naïve (uninfected) mice (Dunnett’s test).

Fig. 3.

Altered blood perfusion speed in the pulmonary capillaries of virus-infected mice. (A) Perfusion speeds of fluorescent beads in mice infected with MA-Venus-H5N1 or MA-Venus-PR8. Fluorescent beads (red) were i.v. injected into the mice during image acquisition and tracked to measure their speeds. Trajectories (cyan lines) of all beads from the same time-lapse series are superimposed on the same image. White indicates autofluorescence. (Scale bar, 50 µm.) See also Movie S3. (B) Blood perfusion speeds in influenza virus-infected lung at different time points. Each dot represents the data obtained from an individual animal. Horizontal bars indicate the median. Asterisks indicate P < 0.05 compared with naïve (uninfected) mice (Dunnett’s test).

Fig. 4.

Recruitment and motility of pulmonary intravascular neutrophils in response to influenza virus infection. (A) The neutrophil population in the lungs at different time points of influenza virus infection. The number of neutrophils was counted in each field of view for 250 s. Each dot represents the data obtained from an individual animal. Horizontal bars indicate the mean. Asterisks indicate P < 0.05 compared with naïve (uninfected) mice (Dunnett’s test). See also Movies S4 and S5. (B) Tracking of neutrophils in time-lapse image series. Mice were injected with PE-conjugated anti-mouse Ly-6G antibody to visualize the neutrophils (white). Two neutrophils were tracked as examples (shown in red and green), and their trajectories were plotted in the same color. (Scale bar, 20 µm.) See also Movie S6. (C) Instantaneous changes in motility of pulmonary neutrophils. The speeds of neutrophils shown in B are plotted in the corresponding colors. (D) Assessment of neutrophil behaviors defined in terms of slow and rapid motion. All neutrophils observed under the same conditions were arranged along the y axis according to the duration of their slow motion. The amount of time each neutrophil was engaged in slow (blue) or rapid (red) motion was plotted along the x axis. The numbers of mice (N_mice) and total neutrophils (N_Neu) assessed for each condition are indicated in the corresponding plot. (E and F) The average percentage of neutrophils with rapid movement (E) and the average time for which neutrophils engaged in slow movement (F) in influenza virus-infected lung. Each dot represents the data obtained from an individual animal. Asterisks indicate P < 0.05 compared with naïve (uninfected) mice (Dunnett’s test). See also SI Appendix, Fig. S3.

Fig. 5.

Dynamics of the pathophysiological changes and corresponding neutrophilic reactions that occur upon influenza virus infection. These graphs summarize the time course of the virological and pathophysiological features measured upon infection with H5N1 (Left) or PR8 (Right) virus. The line graphs of each feature were generated by connecting the mean or median numbers at different time points originating from Fig. 1B for the number of virus-infected cells, Fig. 2C for pulmonary permeability, Fig. 3B for blood flow speed, Fig. 4A for the number of recruited neutrophils, and Fig. 4F for the duration of slow neutrophil movement.