High-Fat Feeding Protects Mice From Ventilator-Induced Lung Injury, Via Neutrophil-Independent Mechanisms

Abstract

Objective: Obesity has a complex impact on acute respiratory distress syndrome patients, being associated with increased likelihood of developing the syndrome but reduced likelihood of dying. We propose that such observations are potentially explained by a model in which obesity influences the iatrogenic injury that occurs subsequent to intensive care admission. This study therefore investigated whether fat feeding protected mice from ventilator-induced lung injury.

Design: In vivo study.

Setting: University research laboratory.

Subjects: Wild-type C57Bl/6 mice or tumor necrosis factor receptor 2 knockout mice, either fed a high-fat diet for 12-14 weeks, or age-matched lean controls.

Interventions: Anesthetized mice were ventilated with injurious high tidal volume ventilation for periods up to 180 minutes.

Measurements and main results: Fat-fed mice showed clear attenuation of ventilator-induced lung injury in terms of respiratory mechanics, blood gases, and pulmonary edema. Leukocyte recruitment and activation within the lungs were not significantly attenuated nor were a host of circulating or intra-alveolar inflammatory cytokines. However, intra-alveolar matrix metalloproteinase activity and levels of the matrix metalloproteinase cleavage product soluble receptor for advanced glycation end products were significantly attenuated in fat-fed mice. This was associated with reduced stretch-induced CD147 expression on lung epithelial cells.

Conclusions: Consumption of a high-fat diet protects mice from ventilator-induced lung injury in a manner independent of neutrophil recruitment, which we postulate instead arises through blunted up-regulation of CD147 expression and subsequent activation of intra-alveolar matrix metalloproteinases. These findings may open avenues for therapeutic manipulation in acute respiratory distress syndrome and could have implications for understanding the pathogenesis of lung disease in obese patients.

Figure 1

Time course of peak inspiratory pressure (PIP) change during high stretch ventilation (Panel A), showing that both lean control mice and high fat-fed mice displayed the same initial decrease in PIP, indicative of lung recruitment at very high tidal volumes. While PIP increased dramatically after 120 minutes in control mice, this was much less apparent in fat-fed animals. A number of animals were unable to complete the 180 minute protocol in the control group only, leading to variability in PIP as mice dropped out at later time points, and a significantly worse survival (Panel B). End PIP (Panel C), irrespective of the length of ventilation, was significantly lower in fat-fed mice than controls. This was related to attenuation of changes in both respiratory system elastance (Panel D) and resistance (Panel E), determined by end-inflation occlusion. Arterial oxygenation was well maintained in both groups until 120 minutes, whereupon it fell more dramatically in the control group than in fat-fed mice (Panel F). End pO₂ values, representing either 180 minutes or the point at which mortality surrogates were met, were significantly higher in fat-fed animals. Arterial CO₂ was similarly well maintained in both groups until 120 minutes, after which it tended to increase in control animals as lung injury developed, but did not in fat-fed animals (Panel G). Data within panels A, D, E, F and G were normally distributed, and are shown as mean±SD. Data within panel C were non-parametric, and thus are displayed as a box-whisker plot. N-11-12 / group (until 150 minutes for panel A, as animals dropped out from this point onwards). Panel B shows a Kaplan-Meier survival curve, with significance determined by logrank test. Data in panel C were evaluated by Mann-Whitney U-test, while data in Panels D-G were evaluated by t-test (end points only in panels F and G). *p<0.05, **p<0.01, ***p<0.001.

Figure 2

Alveolar-epithelial barrier permeability assessed by lavage fluid protein (A) and lung wet:dry weight ratio (B) was significantly decreased in fat-fed animals compared to lean controls. Similarly the epithelial stress marker soluble Receptor for Advanced Glycation End-products (sRAGE) in lavage fluid (Panel C) was attenuated in fat-fed mice. There was a tendency for the number of neutrophils (Panel D) and inflammatory Gr1^high monocytes (E) accumulated within the lungs to be reduced in fat-fed mice, but this was not statistically significant. Lung neutrophils (Panel F) and Gr1^high monocytes (G) showed no differences in levels of the activation marker CD11b between groups. N=5 / group for panels B, F and G, and N=7 / group for panels A, C, D and E. Data in panel C are displayed as box-whisker plot and evaluated by Mann-Whitney U-test, while data in all other panels are displayed as mean±SD and assessed by t-test. **p<0.01, ***p<0.001.

Figure 3

High stretch ventilation experiments were carried out in lean and fat-fed TNF receptor 2 knockout mice. Feeding of a high fat diet led to similar attenuation of lung injury as seen with wildtypes, such that only 1/5 fat-fed animals displayed substantial injury, hence the very large variability in PIP towards the end of the experiment and the apparent decrease in PIP at 180 minutes as the only injured fat-fed animal dropped out of the analysis (Panel A). Survival (Panel B), final PIP (Panel C), pO₂ (panel D), pCO₂ (panel E) and lung wet:dry ratio (Panel F) were similarly attenuated with high fat feeding, all of which demonstrate that the high fat diet-induced protection was not dependent on upregulation of soluble TNF receptor 2. As with wildtype mice, fat-diet induced protection was not mediated by decreased leukocyte infiltration as both neutrophils (Panel G) and inflammatory monocytes (Panel H) were not attenuated. Panels A, D, E, G and H are shown as mean±SD, while panels C and F are displayed as box-whisker plots. N=5 / group for all data (until 120 minutes for panel A, as animals dropped out after this point), apart from panels G&H where N=4. Panel B shows a Kaplan-Meier survival curve, with significance determined by logrank test. Data in panels C and F were evaluated by Mann-Whitney U-test, while data in Panels D, E, G and H were evaluated by t-test (end points only in panels D and E). *p<0.05, **p<0.01, ***p<0.001.

Figure 4

Lavage fluid Interleukin-6 (IL-6, Panel A), CXCL1 (Panel B) and tumour necrosis factor-α (TNF, Panel C) were no different following 120 minutes of injurious ventilation in lean or fat-fed mice. In contrast, net MMP activity was significantly lower in lavage fluid of fat-fed animals ventilated with high V_T (left part of Panel D). This was specifically due to attenuated stretch-induced upregulation, as MMP activity was not different in lean and obese mice ventilated with low stretch (right part of Panel D). N=6-7 for all data in panels A-C and 5-6 in Panel D. Data are displayed as mean±SD for Panels A-C (statistical evaluation by t-test), and Box-whisker plots in Panel D (statistical evaluation by Mann-Whitney U-test). **p<0.01.

Figure 5

Surface expression (mean fluorescence intensity, MFI) of CD147 on type 1 (A) and type 2 (B) epithelial cells determined by flow cytometry. Within lean animals expression was increased on both cell types following 1 hour of high V_T ventilation. In a separate set of experiments, CD147 expression was compared after 1 hour of high or low V_T ventilation between lean and fat-fed mice (Panels C&D). Data for panels A&B are displayed as box-whisker plots, while data for panels C&D are displayed as mean±SD, N=4-5 for each dataset. Statistical analysis for Panels C&D was carried out by ANOVA with Bonferroni correction. *p<0.05, **p<0.01,***p<0.001.