Gelsolin Attenuates Neonatal Hyperoxia-Induced Inflammatory Responses to Rhinovirus Infection and Preserves Alveolarization

Abstract

Prematurity and bronchopulmonary dysplasia (BPD) increase the risk of asthma later in life. Supplemental oxygen therapy is a risk factor for chronic respiratory symptoms in infants with BPD. Hyperoxia induces cell injury and release of damage-associated molecular patterns (DAMPs). Cytoskeletal filamentous actin (F-actin) is a DAMP which binds Clec9a, a C-type lectin selectively expressed on CD103+ dendritic cells (DCs). Co-stimulation of Clec9a and TLR3 induces maximal proinflammatory responses. We have shown that neonatal hyperoxia (a model of BPD) increases lung IL-12+Clec9a+CD103+ DCs, pro-inflammatory responses and airway hyperreactivity following rhinovirus (RV) infection. CD103+ DCs and Clec9a are required for these responses. Hyperoxia increases F-actin levels in bronchoalveolar lavage fluid (BALF). We hypothesized that the F-actin severing protein gelsolin attenuates neonatal hyperoxia-induced Clec9a+CD103+ DC-dependent pro-inflammatory responses to RV and preserves alveolarization. We exposed neonatal mice to hyperoxia and treated them with gelsolin intranasally. Subsequently we inoculated the mice with RV intranasally. Alternatively, we inoculated normoxic neonatal mice with BALF from hyperoxia-exposed mice (hyperoxic BALF), RV and gelsolin. We analyzed lung gene expression two days after RV infection. For in vitro studies, lung CD11c+ cells were isolated from C57BL/6J or Clec9a^gfp-/- mice and incubated with hyperoxic BALF and RV. Cells were analyzed by flow cytometry. In neonatal mice, gelsolin blocked hyperoxia-induced Il12p40, TNF-α and IFN-γ mRNA and protein expression in response to RV infection. Similar effects were observed when gelsolin was co-administered with hyperoxic BALF and RV. Gelsolin decreased F-actin levels in hyperoxic BALF in vitro and inhibited hyperoxia-induced D103lo DC expansion and inflammation in vivo. Gelsolin also attenuated hyperoxia-induced hypoalveolarization. Further, incubation of lung CD11c+ cells from WT and Clec9a^gfp-/- mice with hyperoxic BALF and RV, showed Clec9a is required for maximal hyperoxic BALF and RV induced IL-12 expression in CD103+ DCs. Finally, in tracheal aspirates from mechanically ventilated human preterm infants the F-actin to gelsolin ratio positively correlates with FiO2, and gelsolin levels decrease during the first two weeks of mechanical ventilation. Collectively, our findings demonstrate a promising role for gelsolin, administered by inhalation into the airway to treat RV-induced exacerbations of BPD and prevent chronic lung disease.

Keywords: BPD; dendritic cells; f-actin; gelsolin; lung inflammation; prematurity; rhinovirus.

Figure 1

In neonatal mice, gelsolin blocks hyperoxia-induced pro-inflammatory responses to RV infection. Two-day old wild type mice were exposed to hyperoxia or normoxia for 10 days and treated with recombinant human plasma gelsolin (GSN, 0.5 mg/kg) or equal volume BSA-PBS, administered intranasally under anesthesia daily. On day of life 14 the mice were inoculated with RV. Whole lung mRNA expression of Il12p40, Tnfa and Ifng (A) and protein expression of Il12p70, TNF-α and IFN-γ (B) were measured 2 days later. (N = 4-6, mean ± SEM, *p < 0.05, **p < 0.01, NS, nonsignificant, ANOVA). One of two independent experiments is shown.

Figure 2

Gelsolin inhibits hyperoxia-induced CD103+ DC expansion and inflammation. Two-day old wild type mice were exposed to hyperoxia or normoxia for 4 or 10 days and treated with recombinant human plasma gelsolin (GSN, 0.5 mg/kg) or equal volume BSA-PBS, administered intranasally under anesthesia daily. Lungs were enzymatically digested, and a single cell suspension was incubated and stained with specific cell-surface antibodies. Lung CD103+ DCs were distinguished from other lung cells based on expression of CD45, F4/80, CD11c, CD103 and CD11b. (A) Gating strategy to identify lung DCs. (B) Conventional lung DC populations are distinguished based on the expression of CD103 and CD11b. Following hyperoxia, two distinct populations of CD103+ DCs, CD103lo and CD103hi DCs are observed. GSN treatment during hyperoxia decreases CD103lo, but not CD103hi DCs. (C) Quantification of the CD103lo and CD103hi DCs. (D) Both CD103lo and CD103hi DC populations are absent in the lungs of 4-day old Batf3 null mice compared to age-matched wild type (WT) mice. *p < 0.05, **p < 0.01, ***p < 0.001, NS, nonsignificant (ANOVA). These results are representative of three independent experiments. (E) RNA was extracted from the whole lung tissue on day of life 16 after 10 days exposure to normoxia or hyperoxia with and without daily GSN treatment. GSN attenuated hyperoxia-induced mRNA expression of Il12p40, Myd88, Cd207, Cd103 and Clec9a. *p < 0.05, **p < 0.01 (ANOVA). N = 4 per groups. One of three independent experiments is shown.

Figure 3

Gelsolin attenuates hyperoxic BALF-induced inflammatory response to RV. Two-day old mice were exposed to hyperoxia or normoxia for 14 days. BALF was collected on day of life 16. F-actin and gelsolin (GSN) protein levels were measured in BALF supernatant using ELISA and the ratio of F-actin to GSN was calculated and compared between samples from hyperoxia- and normoxia-exposed mice, unpaired t-test. *p < 0.05 (A). (B) 14-day old mice were exposed to hyperoxia for 4 days and BALF was collected (H-BALF). Cell-free H-BALF was incubated with GSN (5mg) in vitro for 10min and F-actin levels were analyzed by ELISA (paired t-test p < 0.001). (C) 14-day old mice were inoculated with H-BALF, RV, GSN, or appropriate controls and whole lung mRNA expression of Il12p40, Ifng and Tnfa were analyzed 2 days later. *p < 0.05, ***p < 0.001 (ANOVA). N = 3-6 per group. One of two independent experiments is shown. ns, not significant

Figure 4

Clec9a is required for maximal hyperoxic BALF-induced pro-inflammatory activation of lung CD103+ DCs. CD11c+ immune cells were purified from wild type (WT) or Clec9a^gfp-/- mouse lungs. Equal number of cells were co-cultured with medium and hyperoxic BALF (H-BALF) or H-BALF and RV. Subsequently the cells were stained with antibodies for flow cytometry analysis. (A) Gating strategy to identify CD103+ DCs. (B) IL-12 expression in CD103+DCs was examined. (C) Compared to control (C), the percentage of wild type IL-12+CD103+ DCs was increased after co-culture with H-BALF (H) and further increased after co-culture with H-BALF and RV (H+RV). These responses were attenuated in Clec9a null cells. N = 4 per group. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001 (ANOVA).

Figure 5

Gelsolin prevents neonatal hyperoxia-induced hypoalveolarization. Two-day old wild type mice were exposed to hyperoxia or normoxia for 10 days and treated with recombinant human plasma gelsolin (0.5 mg/kg) or equal volume BSA-PBS, administered intranasally under anesthesia daily. Lung histology was assessed on day of life 16. (A) Representative lung sections were stained with hematoxylin and eosin. Hyperoxia induced enlargement of the alveolar spaces (hypoalveolarization) in BSA-PBS-treated mice. In contrast, gelsolin treatment preserved the alveolar architecture of hyperoxia-exposed mice. (B) Alveolar chord length was significantly increased in hyperoxia-exposed, BSA-PBS-treated mice, consistent with hypoalveolarization. In contrast, alveolar chord length of hyperoxia-exposed, gelsolin-treated mice was similar to that of normoxia-exposed mice, indicating protective effect of gelsolin on alveolarization during neonatal hyperoxia. ****p < 0.0001 (ANOVA). N = 5-12 per group.

Figure 6

F-actin/Gelsolin ratio in tracheal aspirates of mechanically ventilated human preterm infants with respiratory distress and FiO2 on day of sample collection. Tracheal aspirates were collected in the first week of life. F-actin and gelsolin levels were measured by ELISA in the supernatant. The association between the ratio of F-actin to gelsolin and the FiO2 on the day of sample collection was determined by Pearson’s correlation analysis.

Figure 7

In tracheal aspirates of human preterm infants with respiratory distress, gelsolin concentrations decrease during the first two weeks of mechanical ventilation. For preterm infants who remain endotracheally intubated receiving mechanical ventilation tracheal aspirates were obtained during week 1 and week 2 of mechanical ventilation and gelsolin levels in the supernatant were quantified. Statistical significance was pointed by paired t-test.