Effect of the Human Amniotic Membrane on the Umbilical Vein Endothelial Cells of Gestational Diabetic Mothers: New Insight on Inflammation and Angiogenesis

Abstract

One of the most relevant diabetes complications is impaired wound healing, mainly characterized by reduced peripheral blood flow and diminished neovascularization together with increased inflammation and oxidative stress. Unfortunately, effective therapies are currently lacking. Recently, the amniotic membrane (AM) has shown promising results in wound management. Here, the potential role of AM on endothelial cells isolated from the umbilical cord vein of gestational diabetes-affected women (GD-HUVECs), has been investigated. Indeed, GD-HUVECs in vivo exposed to chronic hyperglycemia during pregnancy compared to control cells (C-HUVECs) have shown molecular modifications of cellular homeostasis ultimately impacting oxidative and nitro-oxidative stress, inflammatory phenotype, nitric oxide (NO) synthesis, and bioavailability, thus representing a useful model for studying the mechanisms potentially supporting the role of AM in chronic non-healing wounds. In this study, the anti-inflammatory properties of AM have been assessed using a monocyte-endothelium interaction assay in cells pre-stimulated with tumor necrosis factor-α (TNF-α) and through vascular adhesion molecule expression and membrane exposure, together with the AM impact on the nuclear factor kappa-light-chain-enhancer of activated B cell (NF-kB) pathway and NO bioavailability. Moreover, GD-HUVEC migration and tube formation ability were evaluated in the presence of AM. The results showed that AM significantly reduced TNF-α-stimulated monocyte-endothelium interaction and the membrane exposure of the endothelial vascular and intracellular adhesion molecules (VCAM-1 and ICAM-1, respectively) in both C- and GD-HUVECs. Strikingly, AM treatment significantly improved vessel formation in GD-HUVECs and cell migration in both C- and GD-HUVECs. These collective results suggest that AM positively affects various critical pathways in inflammation and angiogenesis, thus providing further validation for ongoing clinical trials in diabetic foot ulcers.

Keywords: HUVECs; amniotic membrane (AM); angiogenesis; diabetes; inflammation; wound healing.

FIGURE 1

AM reduces the expression of genes involved in inflammation. Gene expression analysis of VCAM-1 (A), ICAM-1 (B), CCL2 (C), and SELE (D), from C- and GD-HUVECs untreated and after TNF-α stimulation of cells that had been pretreated or not with AM for 24 h. Quantitative data in the histograms (A–D) show the mRNA relative expression (target gene expression normalized with GAPDH expression using the 2-ΔΔCt method). Each measurement is defined as the mean ± SD using three different strains for C-HUVECs and three different strains for GD-HUVECs (n = 3). ANOVA and Tukey’s multiple comparison test: *p < 0.05, **p < 0.01, ***p < 0.001 and ****p < 0.0001.

FIGURE 2

Effect of AM on VCAM-1 and ICAM-1 membrane exposure and TNF-α-induced monocyte adhesion in C- and GD-HUVECs analyzed by flow cytometry. (A) VCAM-1 and ICAM-1 membrane exposure in C- and GD-HUVECs untreated (basal) and treated for 16 h with 1 ng/ml of TNF-α with or without AM incubation (24 h). Quantitative data in the histograms show the mean fluorescence intensity (MFI) ratio between the signal and the background noise. The experiments were performed using five different cellular strains for C- and five different cellular strains for GD-HUVECs (n = 5). ANOVA and Tukey’s multiple comparison test: *p < 0.05 TNF-α + AM vs. TNF-α; ***p < 0.001 TNF-α+AM vs. TNF-α; ****p < 0.0001 TNF-α vs. basal and vs. AM; ****p < 0.0001 TNF-α + AM vs. TNF-α. (B) Monocyte-endothelial cell interaction on C- and GD-HUVECs untreated (basal) and incubated with AM (24 h) after TNF-α stimulation for 16 h. Representative pictures of C- and GD-HUVECs for each experimental condition. Bar represents 100 µm. Quantitative data, in the histogram (C), show the number of adherent U937 cells obtained by analyzing the microscope images of four different cellular strains for both C-HUVECs and GD-HUVECs (n = 4), each including 12 counts per condition. Each measurement is defined as the mean ± SD of adherent monocytes. ANOVA and Tukey’s multiple comparison test: *p < 0.05 basal GD-HUVEC vs. basal C-HUVEC and AM GD-HUVEC vs. basal GD-HUVEC; **p < 0.01 TNF-α + AM vs. TNF-α; ****p < 0.0001 TNF-α vs. basal and vs. AM, TNF-α + AM vs. TNF-α.

FIGURE 3

Amniotic membrane attenuates TNF-⍺-induced NF-κB phosphorylation and nuclear translocation. Sub-confluent C-HUVECs and GD-HUVECs were serum-starved, treated with AM for 24h, then TNF-α was added at 1 and 3 h. (A) Protein extract was analyzed for WB for the mentioned antibodies. (B) Histograms show Western blot quantification of NF-κB phosphorylation experiments (n = 3); *p< 0.05 vs basal; #p< 0.05 vs TNF-α 1 h; §p< 0.05 vs TNF-α 3 h. (C) Samples were immunostained with a specific antibody against NF-κB and co-staining with Hoechst-33258 and phalloidin to reveal nuclei and he actin cytoskeleton, respectively. This experiment was repeated at least three times. Representative images are shown.

FIGURE 4

Effect of AM on intracellular cGMP levels measured by ELISA in untreated (basal) and TNF-α-stimulated C- and GD-HUVECs after preincubation for 24 h with AM. Ionomycin (2 μM, 24 h) was used as a positive control. l-NAME (1 mmol/L) was added 45 min before AM and Ionomycin. Data are expressed as pmol/well and results by mean ± SEM, n = 4. ANOVA and Tukey’s multiple comparison test: *p < 0.05 TNF-α vs. basal; TNF-α+AM vs. TNF-α; TNF-α +AM+L-NAME vs. TNF-α+AM; **p < 0.01 IONO vs. basal; ***p < 0.001 IONO+L-NAME vs. IONO.

FIGURE 5

Effect of AM on tube-like structure formation capacity on Matrigel. (A) Representative images of C- and GD-HUVECs for both experimental conditions. (B–G) Different angiogenic parameters analyzed in C- and GD-HUVECs under basal conditions and after 6 h of AM incubation. Each value is expressed as mean ± SD using three different cellular strains for C-HUVECs and three different cellular strains for GD-HUVECs (n = 3). ANOVA and Tukey’s multiple comparison test: *p < 0.05, **p < 0.01.

FIGURE 6

AM induces cell migration in C- and GD-HUVECs. (A) HUVECs were wounded and allowed to migrate for the indicated time in the presence or absence of AM. Representative scratch assay from each experimental condition are shown. The wound edges are outlined in yellow. (B) Quantification of the difference of migration between 0 and 24 h. ANOVA and Tukey’s multiple comparison test: ****p < 0.0001.

FIGURE 7

AM induces an active remodeling of focal adhesions (FAs). (A) Sub-confluent C-HUVECs and GD-HUVECs were serum-starved cultured for 24 h in presence or absence of AM. Then, samples were immunostained with a specific antibody against paxillin and co-stained with Hoechst-33258 and phalloidin to reveal nuclei and the actin cytoskeleton, respectively. All images were acquired using a confocal microscope and processed using ZEN software. These experiments were repeated at least three times. Representative images are shown. Bar represents 50 µm. (B) Focal adhesion (FA) quantification of Paxillin. All images were analyzed through ImageJ software. The plots represent the mean of several technical replicates of three C-HUVEC and three GD-HUVEC strains (n = 3). The number of FAs as the region of interests (ROIs) was selected and measured. *p < 0.05 and ***p < 0.001.

FIGURE 8

Schematic representation of the main findings from the study. Gestational diabetes endothelial cells (GD-HUVECs) have shown increased inflammation leading to endothelial dysfunction as well as impaired angiogenesis and wound healing. By using this cellular model that reproduces a typical diabetic foot ulcer endothelium, the amniotic membrane (AM) was able to restore endothelial function, enhance angiogenesis, and improve wound healing.