Refining anti-inflammatory therapy strategies for bronchopulmonary dysplasia

Abstract

Bronchopulmonary dysplasia (BPD) is a severe lung disease of preterm infants, which is characterized by fewer, enlarged alveoli and increased inflammation. BPD has grave consequences for affected infants, but no effective and safe therapy exists. We previously showed that prophylactic treatment with interleukin-1 receptor antagonist (IL-1Ra) prevents murine BPD induced by perinatal inflammation and hyperoxia. Here, we used the same BPD model to assess whether an alternative anti-inflammatory agent, protein C (PC), is as effective as IL-1Ra against BPD. We also tested whether delayed administration or a higher dose of IL-1Ra affects its ability to ameliorate BPD and investigated aspects of drug safety. Pups were reared in room air (21% O₂ ) or hyperoxia (65% or 85% O₂ ) and received daily injections with vehicle, 1200 IU/kg PC, 10 mg/kg IL-1Ra (early or late onset) or 100 mg/kg IL-1Ra. After 3 or 28 days, lung and brain histology were assessed and pulmonary cytokines were analysed using ELISA and cytokine arrays. We found that PC only moderately reduced the severe impact of BPD on lung structure (e.g. 18% increased alveolar number by PC versus 34% by IL-1Ra); however, PC significantly reduced IL-1β, IL-1Ra, IL-6 and macrophage inflammatory protein (MIP)-2 by up to 89%. IL-1Ra at 10 mg/kg prevented BPD more effectively than 100 mg/kg IL-1Ra, but only if treatment commenced at day 1 of life. We conclude that prophylactic low-dose IL-1Ra and PC ameliorate BPD and have potential as the first remedy for one of the most devastating diseases preterm babies face.

Keywords: IL-1 receptor antagonist; hyperoxia; inflammation; neonatal lung disease; protein C.

Figure 1

Overview of the BPD model and experimental set‐up. Pregnant dams were injected i.p. with lipopolysaccharide (LPS) at day 14 of gestation (embryonic day 14, E 14). Within 24 hrs after birth, pups were randomly allocated into five treatment groups and were either housed in hyperoxia (65% O₂ or 85% O₂) or room air (21% O₂). The five treatment groups were 1) daily subcutaneous (s.c.) injections with 1200 IU/kg PC for 28 days (1a) or 3 days (1b); 2) daily s.c. injections with 10 mg/kg IL‐1Ra for 28 days; 3) daily s.c. injections with 100 mg/kg IL‐1Ra for 28 days; 4) daily s.c. injections with saline until day 6, then daily s.c. injections with 10 mg/kg IL‐1Ra until day 28 (referred to as late onset treatment); and 5) daily s.c. injections with volume‐matched saline for 28 days (5a) or 3 days (5b). Pups were humanely killed at day 3 for pulmonary cytokine analysis or at day 28 for analysis of lung and brain structure. As described in the study of Backstrom E et al. (2011), our experimental set‐up comprised four of the five stages of the murine lung development. The embryonic stage, not depicted here, occurs in the mouse between E 9 and 11.5 and is equivalent to 3–7 gestational weeks (GW) in humans. Depicted are the pseudoglandular (E 11.5–16.5 and GW 5–17), canalicular (E 16‐18 and GW 16–26), saccular (E 17.5‐ post‐natal day 5 and GW 24–38) and alveolar phase (post‐natal days 5–28 and >GW 36).

Figure 2

Lung histology and analysis of lung morphology in 28‐day‐old pups after perinatal LPS and post‐natal hyperoxia at 65%. At day 14 of gestation, pregnant dams were injected intraperitoneally with 150 μg/kg LPS. After delivery, newborn pups were exposed to 21% O₂ (room air) or 65% O₂ (hyperoxia) and received daily s.c. injections of either vehicle, 10 mg/kg IL‐1Ra or 100 mg/kg IL‐1Ra from day 1 to day 28. Another group received daily s.c. injections of vehicle from day 1 to day 5 followed by s.c. injections of 10 mg/kg IL‐1Ra from day 6 to day 28 (late onset). For all pups, lung morphology was assessed at day 28, n = 5–27 per group. (A) One representative image per group is shown. Scale bars 100 μm, ×200 magnification. (B‐D) Scans of whole lungs were analysed for alveolar size in μm² (B), number of alveoli per mm² (C) and the surface area‐to‐volume ratio (μm²/μm³) (D). Data are shown as mean ± S.E.M. ***P < 0.001 for air vehicle versus hyperoxia vehicle; #P < 0.05 and ##P < 0.01 for hyperoxia vehicle versus hyperoxia 10 mg/kg IL‐1Ra. LPS: lipopolysaccharide, IL‐1Ra: interleukin‐1 receptor agonist.

Figure 3

Activation of human PC in mice. Healthy 28‐day‐old mice were injected s.c. with 1200 IU/kg PC or vehicle (volume‐matched). 60 min. after injection, human aPC was measured in the plasma, n = 9–14 per group. Data are shown as means ± S.E.M. **P < 0.01 for vehicle versus PC. PC: protein C, aPC: activated protein C.

Figure 4

Lung histology at day 28 of life after antenatal LPS and exposure to hyperoxia. Pregnant dams were injected with 150 μg/kg LPS at day 14 of gestation. Newborn pups were subjected to 21% O₂ (room air), 65% O₂ or 85% O₂ and received daily s.c. injections of vehicle, IL‐1Ra (10 mg/kg, early onset) or PC (1200 IU/kg). Vehicle‐ and IL‐1Ra‐injected animals in air and 65% O₂ are identical with those shown in Fig. 2 as experiments depicted in Fig. 2 and Fig. 4 were carried out in parallel. The effects of IL‐1Ra treatment on animals reared in 85% hyperoxia are published in the study of Nold MF et al. 15 and thus are not depicted here. Lungs were analysed on day 28, n = 4–27 per group. (A) One representative slide per group is depicted. Scale bars 100 μm, ×200 magnification. (B‐D) Quantitative analysis of lung histology at 65% O₂. Depicted are alveolar size in μm² (B), number of alveoli per mm² (C) and the surface area‐to‐volume ratio (μm²/μm³) (D). Data are represented as means ± S.E.M. ***P < 0.001 for air vehicle versus hyperoxia 65% vehicle; #P < 0.05 for hyperoxia 65% vehicle versus hyperoxia 65% IL‐1Ra. LPS: lipopolysaccharide, IL‐1Ra: interleukin‐1 receptor antagonist, PC: protein C.

Figure 5

Immune profiling on day 3 of the murine BPD model with hyperoxia at 65% O₂. Following the same experimental protocol as in Fig. 4, cytokines were detected in the lungs at day 3. Experiments were performed in parallel with those published in Fig. 4 in Nold MF et al. 15; thus, control animals (air vehicle and hyperoxia vehicle animals) are identical to those in Nold MF et al. 15. (A,B,D) Semi‐quantitative protein analysis of cytokines (A), other mediators (B) and chemokines (D) was performed by cytokine protein array, n = 3 per group. Data are shown as optical density (OD) normalized to the positive control spots on each membrane in arbitrary units ±S.E.M. (c,e) IL‐6 (c) and MIP‐2 (E) were measured by ELISA, n = 8–20 per group. Graphs show means of cytokine abundance normalized to total protein (t.p.) ±S.E.M. (A‐E) *P < 0.05, **P < 0.01 and ***P < 0.001 compared to air vehicle; #P < 0.05, ##P < 0.01 and ###P < 0.001 compared to hyperoxia vehicle. MIP‐2: macrophage inflammatory protein.

Figure 6

Hippocampus morphology at day 28. Pregnant dams were injected with 150 μg/kg LPS at day 14 of gestation. Newborn pups were reared in 21% O₂ (room air) or 65% O₂ (hyperoxia) and received daily s.c. injections of vehicle, PC (1200 IU/kg) or IL‐1Ra (10 mg/kg or 100 mg/kg). Brains were H&E‐stained and analysed on day 28, n = 3–24 per group. One representative slide per treatment group showing the pyramidal layer is depicted at low (×40) and high (×200) magnification. Scale bars: 400 μm for ×40 magnification, 100 μm for ×200 magnification. LPS: lipopolysaccharide, PC: protein C, IL‐1Ra: interleukin‐1 receptor antagonist.