Compromised barrier integrity of human feto-placental vessels from gestational diabetic pregnancies is related to downregulation of occludin expression

Downregulation of expression of OCLN variant 2 and the fully functional occludin isoform-A protein are a feature of placentas in d-GDM pregnancies. These may be behind the loss of junctional occludin and the increased extravasation of exogenous dextran observed. miR-181a-5p was in part responsible for the downregulation of occludin in placentas from d-GDM pregnancies. Induced overexpression of miR-181a-5p compromised the integrity of the endothelial barrier. Our data suggest that, despite good glucose control, the adoption of lifestyle changes alone during a GDM pregnancy may not be enough to prevent an alteration in the expression of occludin and the subsequent functional consequences in placentas and impaired vascular barrier function in offspring. Graphical abstract.

Term placentas were obtained from normal pregnancies (n = 21), and pregnancies complicated by GDM where glucose levels were controlled by diet (n = 11) or metformin (n = 6). Gene and microRNA (miRNA) expression were determined by quantitative real-time PCR; protein expression by immunoblotting; endothelial junctional occupancy by fluorescence microscopy and systematic sampling; and paracellular leakage by perfusion of placental microvascular beds with 76 Mr dextran. Transfection studies of miRNAs that target OCLN were performed in HUVECs, and the trans-endothelial electrical resistance and tracer permeability of the HUVECs were measured.

All three predicted OCLN gene splice variants and two occludin protein isoforms were found in human placental samples. In placental samples from diet-controlled GDM (d-GDM) pregnancies we found a lower percentage of conduit vessels showing occludin immunoreactivity (12%, p < 0.01), decreased levels of the fully functional occludin isoform-A protein (29%), and differential gene expression of OCLN variant 2 (33% decrease), variant 3 (3.3-fold increase). These changes were not seen in samples from the group with metformin-controlled GDM. In d-GDM placentas, increased numbers of conduit microvessels demonstrated extravasation of 76 Mr dextran (2.0-fold). In d-GDM expression of one of the five potential miRNAs targeting OCLN, miR-181a-5p, expression was 2.1-fold that in normal pregnancies. Experimental overexpression of miR-181a-5p in HUVECs from normal pregnancies resulted in a highly significant downregulation of OCLN variant 1 (69%) and variant 2 (46%) gene expression, with decreased trans-endothelial resistance (78%) and increase in tracer permeability (1.3-fold). Conclusions/interpretation: Downregulation of expression of OCLN variant 2 and the fully functional occludin isoform-A protein are a feature of placentas in d-GDM pregnancies. These may be behind the loss of junctional occludin and the increased extravasation of exogenous dextran observed. miR-181a-5p was in part responsible for the downregulation of occludin in placentas from d-GDM pregnancies. Induced overexpression of miR-181a-5p compromised the integrity of the endothelial barrier. Our data suggest that, despite good glucose control, the adoption of lifestyle changes alone during a GDM pregnancy may not be enough to prevent an alteration in the expression of occludin and the subsequent functional consequences in placentas and impaired vascular barrier function in offspring. Graphical abstract.