Oscillating glucose induces microRNA-185 and impairs an efficient antioxidant response in human endothelial cells

Primary pooled human umbilical vein endothelial cells (HUVECs) and growth factors were purchased from Lonza (Lonza Bioresearch LBS, Basel, Switzerland). Cells were maintained for 3 weeks in endothelial basal medium, supplemented with low fetal bovine serum (2 %), hydrocortisone (1 µg/mL), basal fibroblastic growth factor (5 ng/mL), epidermal growth factor (5 ng/mL), heparin (0.75 units/mL) and gentamicin/amphotericin (GA-1000, 0.1 %) in a humidified incubator with 5 % carbon dioxide added.

2 × 10⁵ HUVECs/well were seeded in 6-well plates (Corning, NY, USA) and exposed for 7 days to three different glucose concentrations: normal glucose (NG; 5 mmol/l), oscillating glucose (OG; 5/25 mmol/l) and high glucose (HG; 25 mmol/l). OG condition was obtained changing glucose concentration (from 5 to 25 mmol/l) every day. Experimental control was performed incubating the cells with mannitol at the same concentration of glucose.

8-OH-dG, a marker of oxidative stress, was determined in the HUVECs using Bioxytech 8-OHdG-EIA Kit (OXIS Health Products, Portland, OR, USA).

Total RNA was extracted using an RNA purification kit (NorgenBiotek, Thorold, ON, Canada). One microgram of total RNA was reverse transcribed using the SuperScript III reverse transcriptase and random hexamers (Invitrogen, Life Technologies, Grand Island, NY, USA). q-PCR was performed using the ABI 7900 HT thermo-cycler (Applied Biosystems, Life Technologies, Grand Island, NY, USA), in a reaction buffer using Taqman Gene expression Master Mix, with pre-optimized primers and probes obtained from Applied Biosystems, and using SYBR green ready mix (SYBR^® Premix Ex Taq™ II, Taqara, Japan). The list of primers used is reported in Table 1. All q-PCR were normalized to actin-beta, as a housekeeping gene.

miR-185 expression was examined with the TaqMan MicroRNA Assay Kit (Applied Biosystems, Life Technologies, Grand Island, NY, USA). MultiScribe Reverse Transcriptase was used for RT-PCR, and TaqMan primers for hsa-miR-185 (assay ID 002271) were used to monitor miR-185 expression. RNU44/48 or SNORD-44/48 (assay ID 001094/ID 001006) was used as endogenous miRNA controls (all purchased from Applied Biosystems). has-miR-185-5p mirVana^® miRNA mimic (MC12486), Anti-miR™miRNA-185 inhibitor (AM12486), an antisense miR-185, and scrambled Anti-miR™miRNA inhibitor negative control (AM17010) was purchased from Ambion (Foster City, CA, USA). Transfections of miRNA-185 inhibitors were performed at least three times in triplicate using INTERFERin^® transfection reagent according to the manufacturer’s protocol (POLYPLUS-transfection, NY, USA).

The target gene predictions of human miRNAs have been gathered from a publicly available database for miRNAs target predictions (TargetScan 5.2, http://​www.​targetscan.​org, for poorly conserved sites [14]). Sequence for miRNA was obtained from the miRNA database, miRBase (Faculty of Life Sciences, University of Manchester). RNA hybrid tool [15] was used to predict the resulting secondary structure formed by interacting mRNA and miRNA and calculate ΔG minimum free energy. RNA hybrid is available at https://​bibiserv.​techfak.​unibiekefeld.​de/​rnahybrid.

HUVEC 5 × 10⁴ cells passage 4 (p4) were transiently co-transfected with 3′-UTR-GPx-1 expression vector firefly luciferase reporter assay (Origene, MD, USA), containing the entire 217 bp GPx-1 3′-UTR together with miRNA-185 mimic sequence, using jetPRIME co-transfection reagent following manufacturer’s instructions (POLYPLUS). As controls, cells were transfected with empty vector (pMIR) alone, and with pMIR with miRNA-185 mimic sequence. Cells were processed for lysis and collected 72 h after co-transfection and luciferase activity of total cell lysates measured using an established luciferase reporter assay kit (Dual Luciferase Reporter System, Promega, USA). Luciferase values were normalized calculating RLU/μg protein.

Whole cell lysates were prepared using RIPA buffer (Sigma-Aldrich, St. Louis, MO, USA) containing a protease and phosphatase inhibitor cocktail. Protein contents were determined using the Bradford Reagent (Sigma-Aldrich, St. Louis, MO, USA). Whole cell lysates and chromatin-bound nuclear extracts were subjected to 4–20 % Tris–glycine gradient (SDS-PAGE) gels (Lonza Bioresearch LBS, Basel, Switzerland) in reducing conditions and blotted onto a polyvilynidene fluoride membrane. After blocking with 5 % non-fat dry milk in 20 mM Tris–HCl (pH 7.5), 135 mM NaCl, and 0.1 % Tween-20, blots were incubated with polyclonal antibodies against phospho-histone2AX (γ-H2AX) obtained from Cell Signaling (Beverly, MA, USA), washed with 20 mM Tris–HCl (pH 7.5), 135 mM NaCl, and 0.1 % Tween-20, and incubated with a horseradish peroxidase-conjugated secondary antibody. Proteins were detected using the ECL system (Pierce Chemical, Rockford, IL, USA), according to the manufacturer’s instructions and revealed using a CCD camera (ImageQuantLAS4000, GE Healthcare, UK). Antibodies against SOD-1, CAT, GPx-1 (Cell Signaling, Beverly, MA, USA), and SOD-2 (Santa Cruz Biotechnology, CA, USA) were blotted in the same conditions. β-Actin (Sigma-Aldrich, St. Louis, MO, USA) was used as loaded control, whereas histone-H3 (Abcam) was used for a chromatin-bound nuclear extract loaded control. Protein content quantification was performed using computer-assisted densitometry (http://​www.​imagej.​nih.​gov, ImageJ, NIH, Bethesda, MD, USA).

Chromatin-bound nuclear extracts were prepared using Subcellular Protein Fractionation Kit for Cultured Cells (Pierce Chemical, Rockford, IL, USA) according to the manufacturer’s description. Thus, HUVECs (2 × 10⁶) cells were fractionated, equal amounts of chromatin-bound nuclear extracts (30 µg) were separated in 4–20 % Tris–glycine gradient gels (Lonza Bioresearch LBS, Basel, Switzerland), and then run on SDS-PAGE in reducing conditions.

HUVECs (1 × 10⁶) were subjected to lysis and sonication (2 × 30 s) on ice. Supernatants were collected and assayed for protein content using Bradford assay. GPx protein activity was measured by GPx activity colorimetric assay kit (Biovision, Milpitas, CA, USA) according to the manufacturer’s instructions and normalized by protein content.

Results are expressed as mean ± SEM. Statistical analysis was performed using GraphPrism5.0^® (http://​www.​graphpad.​com). Differences between groups were carried out by one-way ANOVA, followed by the Tukey’s post hoc test. Significant differences were assumed at p < 0.05. Three different experiments were performed in triplicate to ensure reproducibility.

The endogenous expression of SOD-1, SOD-2, CAT and GPx-1 were assessed with q-PCR and western blot analysis (Fig. 1a–c). In HG in respect to NG, SOD-1 [mRNA (p < 0.01) and protein (p < 0.01)] and GPx-1 [mRNA (p < 0.05) and protein (p < 0.01)] were increased (Fig. 1a–c). In OG in respect to NG, SOD-1 [mRNA (p < 0.05) and protein (p < 0.05)] was increased, while GPx-1 (mRNA and protein) did not change (Fig. 1a–c). This profile of GPx-1 was also confirmed for protein activity (Fig. 1d). After exposure to HG, GPx activity was significantly higher vs NG and OG (p < 0.001). GPx activity was not different in OG vs NG (p = ns). SOD-2 and CAT were unchanged in both conditions, even though there was a tendency for CAT to increase (p = 0.09), (Fig. 1a–c).

The 8-OHdG content and γ-H2AX levels in HUVECs, as markers of oxidative damage to nucleic acids, were investigated. Our results shown increased 8-OHdG in both OG (p < 0.001) and HG (p < 0.01) compared to NG (Fig. 2a). However, OG induced more damage on DNA (8-OHdG) than HG (OG vs HG, p < 0.05). Similarly, HG (p < 0.05) and OG (p < 0.001) promoted increased double strand breaks in the cells, demonstrated by γ-H2AX formation (Fig. 2b). OG induced more γ-H2AX than HG (OG vs HG, p < 0.001).

It was first addressed whether miR-185 expression was regulated in response to different glucose levels. In OG, miR-185 expression significantly increased vs NG (p < 0.001) and HG (p < 0.01), while it was unchanged in NG and HG (Fig. 3a).

The short length of the seed region predicts multiple potential target genes in the genome. Acting as the specificity components of ribonucleoprotein silencing complexes, miRNAs pair with target mRNAs at sites complementary to the miRNA 5′ region. Most effective sites map to 3′ untranslated regions (3′ UTRs) and pair perfectly with the miRNA seed (nucleotides 2–7), with an additional pair at nucleotide 8 and/or an A across from nucleotide 1. Using in silico analysis (TargetScanHuman v.5) to predict the target gene of miR-185, we identified two binding sites into 3′-UTR of the GPx-1 gene (Fig. 3b). Although the target prediction was based on poorly conserved site analysis, the mature miR-185 stranded with the seed region in a perfect alignment between nucleotides 2–7 (Fig. 3b). This region is a key predictor of direct binding to the target gene, as suggested by Lewis BP et al. [16]. The ability of an miRNA to translationally repress a target mRNA is largely dictated by the free energy of binding of the first eight nucleotides in the 5′ region of the miRNA. RNA hybrid analysis of GPx-1 3′-UTR offered further support for direct binding of miR-185 to the site; the predicted binding of miR-185 on GPx-1 3′-UTR had highly favorable predicted minimum free energy scores of ΔG (Gibbs free energy) = −23.4 kcal/mol, consistent with known miRNA targeting [17].

In the present study we confirmed the Targetscan in silico predictions demonstrating for the first time a specific and significant interaction between GPx-1 3′UTR and miR-185 using luciferase assay. Co-transfection of HUVEC with GPx-1 3′UTR and miR-185 resulted in a significant down regulation of the luciferase light emission (RLU, Relative Light Unit) compared to cells co-transfected with pMIR alone (p < 0.05), and with pMIR plus mimic miR-185 (p < 0.05) (Fig. 3c). Three independent experiments and two replicated transfections were performed.

Because GPx-1 expression, with respect to NG, was unchanged in OG and increased in HG, we explored the effects of miR-185 on GPx-1 target gene. OG-induced upregulation of miR-185 was associated with unchanged levels of GPx-1, as detected by q-PCR and Western blot (Fig. 1a–c). To define the effects of miR-185 on GPx-1 gene regulation, miR-185 expression was silenced using the miR-185 Anti-miR™ miRNA inhibitor. We transfected 1 nmol/l of anti-miR-185 in HUVECs at the end of culture in NG, OG and HG (Fig. 4a). The transfection efficiencies, assessed by measuring miR-185 expression in q-PCR (Fig. 4a), were ≥80 %. No difference in NG compared with scrambled miRNA transfection was found. Our results demonstrated that the intracellular levels of miR-185 in transfected samples (Fig. 4a) were reduced in a significant manner (OG vs NG, p < 0.01; OG vs HG, p < 0.001; Fig. 4a). Trypan blue exclusion dye also demonstrated no cellular loss during transfections (Fig. 4b; Additional file 1:Figure S1). In anti-miR-185 inhibition transfection to knockdown miR-185, we noticed a significant inverted tendency in GPx-1 protein levels (Fig. 4c), suggesting that GPx-1 was modulated specifically by OG and could be a real target for miR-185. We also measured GPx activity during inhibition of miR-185 and found an increased activity in OG in respect to HG (p < 0.05, Fig. 4d).