Caffeine is associated with improved alveolarization and angiogenesis in male mice following hyperoxia induced lung injury

Abstract

Background: Caffeine therapy for apnea of prematurity reduces the incidence of bronchopulmonary dysplasia (BPD) in premature neonates. Several mechanisms, including improvement in pulmonary mechanics underly beneficial effects of caffeine in BPD. As vascular development promotes alveologenesis, we hypothesized that caffeine might enhance angiogenesis in the lung, promoting lung growth, thereby attenuating BPD.

Methods: C57Bl/6 mice litters were randomized within 12 h of birth to room air (RA) or 95%O₂ to receive caffeine (20 mg/kg/day) or placebo for 4 days and recovered in RA for 12wks. The lung mRNA and protein expression for hypoxia-inducible factors (HIF) and angiogenic genes performed on day 5. Lung morphometry and vascular remodeling assessed on inflation fixed lungs at 12wks.

Results: Caffeine and hyperoxia in itself upregulate HIF-2α and vascular endothelial growth factor gene expression. Protein expression of HIF-2α and VEGFR1 were higher in hyperoxia/caffeine and angiopoietin-1 lower in hyperoxia. An increase in radial alveolar count, secondary septal count, and septal length with a decrease in mean linear intercept indicate an amelioration of hyperoxic lung injury by caffeine. An increase in vessel surface area and a significant reduction in smooth muscle thickness of the pulmonary arterioles may suggest a beneficial effect of caffeine on vascular remodeling in hyperoxia, especially in male mice.

Conclusions: Postnatal caffeine by modulating angiogenic gene expression early in lung development may restore the pulmonary microvasculature and alveolarization in adult lung.

Keywords: Caffeine; Hyperoxia; Hypoxia-inducible factors; Lung; Mice; Newborn; VEGF.

Fig. 1

Body weights (grams) in adult mice at sacrifice in all the four groups (Open box plots – Room air; shaded box plots – hyperoxia; RAS – room air saline; RAC – room air caffeine; HS – hyperoxia saline; HC – hyperoxia caffeine). Caffeine treated mice had significantly higher body weight at 12wks in hyperoxia (#p < 0.01 vs. HS group, a). Hyperoxic male mice treated with caffeine had significantly higher body weight compared to other groups (**p < 0.001 vs. RAS, RAC & HS groups, b). Caffeinated male mice had significantly higher body weight compared to RA or hyperoxia controls (†p < 0.05 vs. RAS & HS groups). No significant difference in body weight noted in female mice (c). # ** † Fisher’s post-hoc test, ANOVA (n = ten mice per group) (RAS – 6 males, 5 females; RAC – 6 males, 4 females; HS – 5 males, 5 females; HC – 6 males, 4 females)

Fig. 2

Histopathology of lungs by H&E at 12wks (100x - a,c,e,g; 400x - b,d,f,h in all the four groups (room air saline (RAS) – a/b; room air caffeine (RAC) – c/d; hyperoxia saline (HS) – e/f; hyperoxia caffeine (HC) – g/h). Alveoli were numerous in both the RA groups (a/b; c/d); hyperoxia impaired alveolar development with fewer, distended alveoli (E/F) and caffeine restored lung alveolarization (g/h; i/j) (Scale bar: 200 μm (100x), 60 μm (400x)). The radial alveolar count was significantly lower (i) and mean linear intercept was significantly higher (j) in the HS group in both male and female mice suggesting alveolar simplification (Open box plots – RA groups; shaded box plots – hyperoxia groups). *p < 0.001 vs. RAS, RAC & HC groups; Fisher’s post-hoc test, ANOVA (n = ten mice/group)

Fig. 3

Representative sections of elastin (left panel – a/d/g/j), Von Willebrand factor (vWF) (Middle panel – b/e/h/k) and trichrome (right panel – c/f/i/l) staining of lung sections (200x) in all the four groups (room air saline (RAS) – a/b/c; room air caffeine (RAC) – d/e/f; hyperoxia saline (HS) – g/h/i & hyperoxia caffeine (HC) – j/k/l). Septal count (m) and septal length (n) were studied in elastin sections; vessel count (o) and vessel surface area (p) were evaluated by vW Factor immunohistochemistry and trichrome sections (Open box plots – RA groups; shaded box plots – hyperoxia groups). Lower septal count (m) and reduced septal length (n) in the hyperoxia group was significantly augmented by caffeine in both male and female mice (*p < 0.001 vs. RAS, RAC & HC groups, Fisher’s post-hoc test, ANOVA). Vessel count was significantly higher in the HC group, especially in male mice (o, **p < 0.0001 vs. RAS, RAC & HS groups, †p < 0.01 vs. RAC & HS groups; Fisher’s post-hoc test, ANOVA). Blood vessel surface area at 200x (237,600μm²) was higher in the hyperoxia group administered caffeine in both male and female mice (p, **p < 0.0001 vs RAS, RAC & HS groups, Fisher’s post-hoc test, ANOVA (n = ten mice/group)

Fig. 4

Assessment of vessel wall thickness at 12wks by α-smooth muscle actin immunohistochemistry of lung sections (100x: a/c/e/g; 400x: b/d/f/h; room air saline (RAS): A/B; room air caffeine (RAC): c/d; hyperoxia saline (HS): e/f; hyperoxia caffeine (HC): g/h). Scale: 200 μm (100x), 60 μm (400x). Thick arrows (400x) correspond to the same histologic location as thin arrows (100x) of the same group. Demonstrated vessels are of a 30-50 μm diameter in all the four groups. Pulmonary arterial vessel wall thickness was greater in the hyperoxia group suggestive of smooth muscle hypertrophy (i; *p < 0.001 vs. RAS; Fisher’s post-hoc test, ANOVA). Caffeine administration attenuated vessel wall thickness, especially in male mice at 12 weeks of age (**p < 0.0001 vs HS group, Fisher’s post-hoc test, ANOVA; i), suggesting it may have beneficial effects on the pulmonary vasculature (Open box plots – RA groups; shaded box plots – hyperoxia groups) (n = ten mice/group)

Fig. 5

Illustration of potential modulation of HIF regulation by caffeine. Stabilization of HIF1α/2α determined by several factors including oxygen, the activity of PHD complex and its co-factors and reactive oxygen species. HIF-1α/2α activates multiple genes involved in glycolysis, angiogenesis, vascular remodeling, cell proliferation, and erythropoiesis via nuclear transcription. Higher expression of VEGFR1 and angiopoietin-1 along with vascular remodeling following caffeine administration may normalize lung histology seen in hyperoxia-induced lung injury. Hydroxylation of prolyl hydroxylase complex inactivates HIF-1α/2α followed by ubiquitination and proteasome degradation. HSL: hormone-sensitive lipase; 2OG: 2oxyglutarate; ROS: reactive oxygen species; PHD: prolyl hydroxylase; HIF: hypoxia inducible factor; HRE: hypoxia response element; VEGFR1: vascular endothelial growth factor receptor-1; Ang-1: angiopoietin-1; BPD: bronchopulmonary dysplasia; AOE – antioxidant enzyme activity; ETC – electron transport chain; PTP: permeability transition pore (copyright – Vasantha HS Kumar, MD)