Aims/hypothesis: Diabetes is a major burden on Australia's Indigenous population, with high rates of disease and vascular complications. Diabetic vascular complications are associated with impaired ischaemia-driven angiogenesis. MicroRNAs (miRNAs) are key players in the regulation of angiogenesis. HDL-cholesterol (HDL-c) levels are inversely associated with the risk of developing diabetic complications and HDL can carry miRNAs. HDL-miRNA profiles differ in disease states and may present as biomarkers with the capacity to act as bioactive signalling molecules. Recent studies have demonstrated that HDL becomes dysfunctional in a diabetic environment, losing its vasculo-protective effects and becoming more pro-atherogenic. We sought to determine whether HDL-associated miRNA profiles and HDL functionality were predictive of the severity of diabetic vascular complications in Australia's Indigenous population.
Methods: HDL was isolated from plasma samples from Indigenous participants without diabetes ('Healthy'), with type 2 diabetes mellitus ('T2DM') and with diabetes-associated macrovascular complications (specifically peripheral artery disease, 'T2DM+Comp'). To assess HDL angiogenic capacity, human coronary artery endothelial cells were treated with PBS, reconstituted HDL (rHDL, positive control) or isolated HDL and then exposed to high-glucose (25 mmol/l) conditions. The expression levels of two anti-angiogenic miRNAs (miR-181c-5p and miR-223-3p) and one pro-angiogenic miRNA (miR-27b-3p) were measured in the HDL fraction, plasma and treated human coronary artery endothelial cells by quantitative real-time PCR. In vitro endothelial tubule formation was assessed using the Matrigel tubulogenesis assay.
Results: Strikingly, we found that the levels of the anti-angiogenic miRNA miR-181c-5p were 14-fold higher (1454 ± 1346%) in the HDL from Aboriginal people with diabetic complications compared with both the Healthy (100 ± 121%, p < 0.05) and T2DM (82 ± 77%, p < 0.05) groups. Interestingly, we observed a positive correlation between HDL-associated miR-181c-5p levels and disease severity (p = 0.0020). Under high-glucose conditions, cells treated with rHDL, Healthy HDL and T2DM HDL had increased numbers of tubules (rHDL: 136 ± 8%, p < 0.01; Healthy HDL: 128 ± 6%, p < 0.01; T2DM HDL: 124 ± 5%, p < 0.05) and branch points (rHDL: 138 ± 8%, p < 0.001; Healthy HDL: 128 ± 6%, p < 0.01; T2DM HDL: 127 ± 5%, p < 0.01) concomitant with elevations in mRNA levels of the key hypoxia angiogenic transcription factor HIF1A (rHDL: 140 ± 10%, p < 0.01; Healthy HDL: 136 ± 8%, p < 0.01; T2DM HDL: 133 ± 9%, p < 0.05). However, this increase in angiogenic capacity was not observed in cells treated with T2DM + Comp HDL (tubule numbers: 113 ± 6%, p = 0.32; branch points: 113 ± 5%, p = 0.28; HIF1A: 117 ± 6%, p = 0.43), which could be attributed to the increase in cellular miR-181c-5p levels (T2DM + Comp HDL: 136 ± 7% vs PBS: 100 ± 9%, p < 0.05).
Conclusions/interpretation: In conclusion, HDL from Aboriginal people with diabetic complications had reduced angiogenic capacity. This impairment is associated with an increase in the expression of anti-angiogenic miR-181c-5p. These findings provide the rationale for a new way to better inform clinical diagnosis of disease severity with the potential to incorporate targeted, personalised HDL-miRNA intervention therapies to prevent further development of, or to reverse, diabetic vascular complications in Australian Aboriginal people.
Keywords: Angiogenesis; HDL functionality; Peripheral artery disease; miRNAs.