The extracellular matrix protein laminin-10 promotes blood-brain barrier repair after hypoxia and inflammation in vitro

Abstract

Background: The blood-brain barrier (BBB) of the central nervous system (CNS) is essential for normal brain function. However, the loss of BBB integrity that occurs after ischaemic injury is associated with extracellular matrix (ECM) remodelling and inflammation, and contributes to poor outcome. ECM remodelling also contributes to BBB repair after injury, but the precise mechanisms and contribution of specific ECM molecules involved are unknown. Here, we investigated the mechanisms by which hypoxia and inflammation trigger loss of BBB integrity and tested the hypothesis ECM changes could contribute to BBB repair in vitro.

Methods: We used an in vitro model of the BBB, composed of primary rat brain endothelial cells grown on collagen (Col) I-, Col IV-, fibronectin (FN)-, laminin (LM) 8-, or LM10-coated tissue culture plates, either as a single monolayer culture or on Transwell® inserts above mixed glial cell cultures. Cultures were exposed to oxygen-glucose deprivation (OGD) and/or reoxygenation, in the absence or the presence of recombinant interleukin-1β (IL-1β). Cell adhesion to ECM molecules was assessed by cell attachment and cell spreading assays. BBB dysfunction was assessed by immunocytochemistry for tight junction proteins occludin and zona occludens-1 (ZO-1) and measurement of trans-endothelial electrical resistance (TEER). Change in endothelial expression of ECM molecules was assessed by semi-quantitative RT-PCR.

Results: OGD and/or IL-1 induce dramatic changes associated with loss of BBB integrity, including cytoplasmic relocalisation of membrane-associated tight junction proteins occludin and ZO-1, cell swelling, and decreased TEER. OGD and IL-1 also induced gene expression of key ECM molecules associated with the BBB, including FN, Col IV, LM 8, and LM10. Importantly, we found that LM10, but not FN, Col IV, nor LM8, plays a key role in maintenance of BBB integrity and reversed most of the key hallmarks of BBB dysfunction induced by IL-1.

Conclusions: Our data unravel new mechanisms of BBB dysfunction induced by hypoxia and inflammation and identify LM10 as a key ECM molecule involved in BBB repair after hypoxic injury and inflammation.

Fig. 1

Primary rBEC cultures on Col I-coated Transwell® inserts were subjected to normoxia or OGD for 2.5 h. Immunocytochemical images for occludin and ZO-1 (a) were analysed with ImageJ to measure cell circularity (b), total intensity of fluorescence (c), and ratio of mean cytoplasmic fluorescence intensity to mean membranous ZO-1 fluorescence intensity (d). Data are shown as mean ± standard deviation from three independent experiments carried out on separate cultures. Data were statistically analysed by one-way ANOVA and Bonferroni’s post hoc tests to compare all data sets. The scale bar in a represents 100 μm

Fig. 2

Primary rBEC cultures on Col I-coated Transwell® inserts were subjected to normoxia, IL-1β treatment for 4 h, 2.5 h OGD and 4 h reoxygenation (OGD + R), 2.5 h OGD, and 4 h reoxygenation in the presence of IL-1β (OGD + R + IL-1). Immunocytochemical images for occludin (a) were analysed with ImageJ to measure cell circularity (b), total intensity of occludin fluorescence (c), and ratio of mean cytoplasmic fluorescence intensity to mean membranous occludin fluorescence intensity (d). Data are shown as mean ± standard deviation from three independent experiments carried out on separate cultures. Data were statistically analysed by one-way ANOVA and Bonferroni’s post hoc tests to compare all data sets. *p < 0.05, **p < 0.01, and ***p < 0.001 vs. normoxia. The scale bar in a represents 100 μm

Fig. 3

Primary rBEC cultures on Col I-coated Transwell® inserts were subjected to normoxia, IL-1β treatment for 4 h (IL-1), 2.5 h OGD and 4 h reperfusion (OGD + R), 2.5 h OGD, and 4 h reperfusion in the presence of IL-1β (OGD + R + IL-1). Immunocytochemical images for ZO-1 (a) were analysed with ImageJ to measure total intensity of ZO-1 fluorescence (b) and ratio of mean cytoplasmic fluorescence intensity to mean membranous ZO-1 fluorescence intensity (c). TEER measurements were carried out at the end of each treatment (d). Data are shown as mean ± standard deviation from three independent experiments carried out on separate cultures. Data were statistically analysed by one-way ANOVA and Bonferroni’s post hoc tests to compare all data sets. *p < 0.05, **p < 0.01, and ***p < 0.001 vs. normoxia. The scale bar in a represents 100 μm

Fig. 4

Primary rBEC cultures on Col I-coated Transwell® inserts were subjected to normoxia, or OGD (2.5 h) with (2.5 + 2, OGD + R) or without (2.5 + 0) reoxygenation for 2 h, and in the absence or the presence of IL-1β for 2 h. Total RNA was extracted, and RT-PCR was carried out for lmα4 (a), lmα5 (b), lmβ1 (c), lmγ1 (d), fn (e), col4 (f), and the housekeeping gene Rpl13a (ribosomal protein L13A). Products were resolved on an ethidium bromide-containing 1.5 % (w/v) agarose gel by electrophoresis and then quantified using GE healthcare ImageQuant™. Data are shown as mean ± standard deviation from three independent experiments carried out on separate cultures. Data were statistically analysed by one-way ANOVA and Bonferroni’s post hoc tests to compare all data sets. *p < 0.05, **p < 0.01, and ***p < 0.001 vs. normoxia

Fig. 5

Primary rBECs were seeded onto tissue culture plastic pre-coated with BSA (TcP) or increasing concentrations of Col I, Col IV, FN, LM8, or LM10 for 90 min, and cell attachment was measured and expressed as percentage of cell number seeded (a). Primary rBECs were seeded onto wells coated with 100 nM BSA (TcP), Col I, Col IV, FN, LM8, or LM10 for 120 min, and cell spreading was measured and expressed as percentage of cell number seeded (b). Primary rBECs seeded onto Col I-, Col IV-, FN-, or LM10-coated (100 nM) Transwell® inserts were cultured for 1 to 4 days, after which TEER was measured (c). Data are shown as mean ± standard deviation from three independent experiments carried out on separate cultures. Data were statistically analysed by one-way ANOVA and Bonferroni’s post hoc tests to compare all data sets. For b, **p < 0.01 and ***p < 0.001 vs. TcP; ^††† p < 0.001 vs. Col I. For c, *p < 0.05 vs. Col I

Fig. 6

Primary rBECs seeded onto Col I-, Col IV-, FN-, or LM10-coated (100 nM) Transwell® inserts were cultured for 4 days, after which cultures were treated with vehicle or IL-1β for 4 h. TEER measurements were carried out at the end of each treatment (a). Immunocytochemical images for occludin (b) were analysed with ImageJ to measure total intensity of occludin fluorescence (c), ratio of mean cytoplasmic fluorescence intensity to mean membranous occludin fluorescence intensity (d), and cell circularity (e). Data are shown as mean ± standard deviation from three independent experiments carried out on separate cultures. Data were statistically analysed by one-way ANOVA and Bonferroni’s post-hoc tests to compare all data sets. *p < 0.05, **p < 0.01, and ***p < 0.001 vs. vehicle. The scale bar in a represents 10 μm