The synergistic effects of mechanical ventilation and intrauterine inflammation on cerebral inflammation in preterm fetal sheep

Abstract

Background: Intrauterine inflammation and the requirement for mechanical ventilation independently increase the risk of perinatal brain injury and adverse neurodevelopmental outcomes. We aimed to investigate the effects of mechanical ventilation for 24 h, with and without prior exposure to intrauterine inflammation, on markers of brain inflammation and injury in the preterm sheep brain.

Methods: Chronically instrumented fetal sheep at ~115 days of gestation were randomly allocated to receive a single intratracheal dose of 1 mg lipopolysaccharide (LPS) or isovolumetric saline, then further randomly allocated 1 h after to receive mechanical ventilation with room air or no mechanical ventilation (unventilated control + saline [UVC, n = 7]; in utero mechanical ventilation + saline [VENT, n = 8], unventilated control + intratracheal LPS [UVC + LPS, n = 7]; in utero ventilation + intratracheal LPS [VENT + LPS, n = 7]). Serial fetal blood and plasma samples were collected throughout the experimental protocol for assessment of blood biochemistry and plasma interleukin (IL)-6 levels. After 24 h of mechanical ventilation, fetal brains were collected for RT-qPCR and immunohistochemical analyses.

Results: LPS exposure increased numbers of microglia and upregulated pro-inflammatory related genes within the cortical gray matter (GM) and subcortical white matter (SCWM) (p_LPS < 0.05). Mechanical ventilation alone increased astrocytic cell density in the periventricular white matter (PVWM) (p_VENT = 0.03) but had no effect on pro-inflammatory gene expression. The combination of ventilation and LPS increased plasma IL-6 levels (p < 0.02 vs. UVC and VENT groups), and exacerbated expression of pro-inflammatory-related genes (IL1β, TLR4, PTGS2, CXCL10) and microglial density (p < 0.05 vs. VENT).

Conclusion: This study demonstrates that 24 h of mechanical ventilation after exposure to intrauterine inflammation increased markers of systemic and brain inflammation and led to the upregulation of pro-inflammatory genes in the white matter. We conclude that 24 h of mechanical ventilation following intrauterine inflammation may precondition the preterm brain toward being more susceptible to inflammation-induced injury.

Keywords: chorioamnionitis; intrauterine inflammation; neuroinflammation; preterm brain; ventilation.

Figure 1

Schematic indicating fields sampled for histological assessment. Representative image for histological assessment of fields of views (FOV) sampled within the cortical gray matter (GM; yellow) and subcortical white matter (SCWM; green) regions within the first and second parasagittal gyri and the periventricular white matter (PVWM; red). Scale bar is 5 mm.

Figure 2

Blood gas and Metabolite measurements. Blood gas measurements of fetal arterial (A) pH, (B) partial pressure CO₂ (PaCO₂), (C) partial pressure O₂ (PaO₂), (D) blood glucose and (E) lactate taken pre-LPS and pre-ventilation (dotted line). Time shown relative to ventilation in min and then h. Data are mean ± SD. Unventilated control (UVC; n = 7), LPS unventilated (UVC + LPS; n = 7), ventilated vehicle (VENT; n = 8), and ventilated LPS (VENT + LPS; n = 7) fetuses. Two-way ANOVA and Tukey’s multiple comparisons. Significant differences between UVC vs. VENT + LPS indicated as ^{^}p < 0.05; ^{^^}p < 0.01; ^{^^^}p < 0.001; VENT vs. VENT + LPS indicated as *p < 0.05, **p < 0.01, ***p < 0.001; UVC + LPS vs. VENT + LPS indicated as ^†p < 0.05.

Figure 3

Plasma IL-6 levels. Plasma IL-6 levels taken at baseline prior to LPS and ventilation (dotted line). Time shown relative to ventilation in h. Data are mean ± SD. Unventilated control (UVC; n = 7), LPS unventilated (UVC + LPS; n = 7), ventilated (VENT; n = 8), and ventilated LPS (VENT + LPS; n = 7) fetuses. Two-way ANOVA and Tukey’s multiple comparisons. Significant differences between UVC vs. VENT + LPS indicated as ^{^}p < 0.05; VENT vs. VENT + LPS indicated as *p < 0.05.

Figure 4

Fold change of mRNA expression of genes relating to inflammation. mRNA expression (relative to UVC) of genes relating to inflammation measured within (A) the cortical gray matter (GM), (B) periventricular white matter (PVWM) and (C) subcortical white matter (SCWM) in unventilated control (UVC; n = 7), LPS unventilated (UVC + LPS; n = 7), ventilated vehicle (VENT; n = 8), and ventilated LPS (VENT + LPS; n = 7) fetuses. Data are mean ± SD. Two-way ANOVA and Tukey’s multiple comparisons, *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001.

Figure 5

Iba-1 positive immunohistochemistry. (A) Iba-1 immunopositive cell density indicating microglia population and (B) ameboid or activated microglia cell density. Data are means ± SD. Two-way ANOVA and Tukey’s multiple comparisons, *p < 0.05. Unventilated control (UVC; n = 7), LPS unventilated (UVC + LPS; n = 7), ventilated vehicle (VENT; n = 8), and ventilated LPS (VENT + LPS; n = 7) fetuses. (C) Representative images of Iba-1-positive cells indicating microglia in the subcortical white matter (SCWM), cortical gray matter (GM), putamen and caudate. Arrowheads indicate ameboid microglia morphology. Insert are zoomed images of dashed box. Scale bar represents 100 μm.

Figure 6

GFAP positive immunohistochemistry. (A) Numbers of GFAP immunopositive cells and (B) area coverage of GFAP staining. Data are mean ± SD. Two-way ANOVA and Tukey’s multiple comparisons, *p < 0.05. Unventilated control (UVC; n = 7), LPS unventilated (UVC + LPS; n = 7), ventilated vehicle (VENT; n = 8), and ventilated LPS (VENT + LPS; n = 7) fetuses. (C) Representative images of GFAP-positive cells. Scale bar represents 100 μm.

Figure 7

Olig2 and NeuN positive immunohistochemistry. (A) Numbers of Olig-2 and (B) NeuN immunopositive cells. Data are mean ± SD. Two-way ANOVA. Unventilated control (UVC; n = 7), LPS unventilated (UVC + LPS; n = 7), ventilated vehicle (VENT; n = 8), and ventilated LPS (VENT + LPS; n = 7) fetuses. (C) Representative images of Olig-2 and NeuN-positive staining in the periventricular white matter (PVWM) and cortical gray matter (GM). Scale bar represents 100 μm.