Downregulation of the Netrin-1 Receptor UNC5b Underlies Increased Placental Angiogenesis in Human Gestational Diabetes Mellitus

Abstract

Gestational diabetes mellitus (GDM) is a common metabolic disorder, defined by high blood glucose levels during pregnancy, which affects foetal and post-natal development. However, the cellular and molecular mechanisms of this detrimental condition are still poorly understood. A dysregulation in circulating angiogenic trophic factors, due to a dysfunction of the feto-placental unit, has been proposed to underlie GDM. But even the detailed study of canonical pro-angiogenic factors like vascular endothelial growth factor (VEGF) or basic Fibroblast Growth Factor (bFGF) has not been able to fully explain this detrimental condition during pregnancy. Netrins are non-canonical angiogenic ligands produced by the stroma have shown to be important in placental angiogenesis. In order to address the potential role of Netrin signalling in GDM, we tested the effect of Netrin-1, the most investigated member of the family, produced by Wharton's Jelly Mesenchymal Stem Cells (WJ-MSC), on Human Umbilical Vein Endothelial Cells (HUVEC) angiogenesis. WJ-MSC and HUVEC primary cell cultures from either healthy or GDM pregnancies were exposed to physiological (5 mM) or high (25 mM) d-glucose. Our results reveal that Netrin-1 is secreted by WJ-MSC from healthy and GDM and both expression and secretion of the ligand do not change with distinct experimental glucose conditions. Noteworthy, the expression of its anti-angiogenic receptor UNC5b is reduced in GDM HUVEC compared with its expression in healthy HUVEC, accounting for an increased Netrin-1 signalling in these cells. Consistently, in healthy HUVEC, UNC5b overexpression induces cell retraction of the sprouting phenotype.

Keywords: GDM; HUVEC; Netrin-1; UNC5b; WJ-MSC; angiogenesis.

Figure 1

Netrin-1 expression and secretion do not change in healthy versus GDM WJ-MSC. (A) Scheme of the experimental procedures. WJ-MSC primary cultures were maintained for 10 days at the indicated glucose concentration in 10% FBS and then incubated for up to 2 more days in either 5 mM or 25 mM plus 2% FBS. Samples were collected at indicated times for analysis. The vertical arrow indicates initiation of differential d-glucose treatment. (B) Whole cell lysates were evaluated by Western blot for Netrin-1 expression in WJ-MSC from healthy and GDM cultures. VEGF was used as a positive control and β-actin was used as internal loading control. Exposure to different d-glucose concentration as indicated. (C) Data represent the mean of independent biological replicates ± SEM (healthy n = 12, GDM n = 6). (D) Quantitation of Netrin-1 by Elisa analysis. CM from healthy and GDM WJ-MSC was obtained to evaluate Netrin-1, CM was obtained with different d-glucose treatment at indicated time points (n = 4).

Figure 2

Netrin-1 produced by healthy WJ-MSC promotes angiogenesis in GDM HUVEC independently of d-glucose condition. (A–C) Representative images of Matrigel tubule formation and branch points per field of either normal HUVEC (A) or GDM HUVEC (B) exposed 4 h to DMEM, EGM or CM (48 h) from WJ-MSC grown in d-glucose as indicated and in absence or presence of 2F5 (2 µg/mL). Scale bar = 15 µm. A. Quantified data correspond to the mean ± S.E.M. (CM WJ-MSC and healthy HUVEC, n = 3, * p < 0.05 vs. DMEM, βp < 0.05 vs. EGM, γp < 0.05 vs. CM 25 mM). (B) Quantified data correspond to the mean ± S.E.M. (CM WJ-MSC and GDM HUVEC, n = 3, * p < 0.05 vs. DMEM, Φp < 0.05 vs. CM 5 mM, γp < 0.05 vs. CM 25 mM). (C) Fold change of tubule formation assay in healthy versus GDM HUVEC. Quantified data correspond to the mean ± S.E.M.

Figure 3

The classical anti-angiogenic Netrin-1 receptor UNC5b is decreased in GDM HUVEC. (A) Upper panel shows a representative Western blot for UNC5b and UNC5c protein expression levels in healthy and GDM HUVEC cell lysates. β-actin was used as internal loading control. d-glucose concentration and time points as indicated. Graphs below show quantified results corresponding to the mean ± S.E.M. (healthy n = 3, GDM n = 3). * p < 0.05 vs. healthy. (B) Different HUVEC from fresh umbilical cords randomly selected are indicated as #1-6. Western Blot revealing UNC5b protein expression levels in primary cell cultures (0–1 passages) of |healthy and GDM HUVEC. β-actin was used as internal loading control. Graph below shows receptor expression per replicate, labelled #1, #2, #3 for healthy and #4, #5, #6 for GDM HUVEC.

Figure 4

Overexpression of the anti-angiogenic receptor UNC5b prevents the pro-angiogenic response of Netrin-1 in HUVEC. Healthy HUVEC were transfected with UNC5b O-E or Mock-expressing constructs (1 μg/well) (A) To confirm UNC5b overexpression, cell lysates were analysed by Western Blot. β-actin levels were determined as loading control. Data correspond to the mean ± S.E.M (healthy n = 3). * p < 0.05 vs. internal control. (B) The overexpression of UNC5b was evaluated through a 4 h of Matrigel tubule formation assay in presence of Netrin-1. Representative images of treatments as indicated. Data correspond to the mean ± S.E.M (healthy n = 3). Scale bar = 15 µm. * p < 0.05 vs. internal control.

Figure 5

Working Model. Model integrating the pro-angiogenic effect of Netrin-1 secretion by WJ-MSC. Endothelial cells express both classical UNC5b and UNC5c, as well as non-classical receptors (Integrins) for Netrin-1. Considering that UNC5b is an anti-angiogenic receptor, it is feasible that loss of UNC5b in endothelial cells promotes alterations in placental angiogenesis, as observed in GDM.