Diabetes adversely affects multiple organs, including the kidney, eye and nerve, leading to diabetic kidney disease, diabetic retinopathy and diabetic neuropathy, respectively. In both type 1 and type 2 diabetes, tissue damage is organ specific and is secondary to a combination of multiple metabolic insults. Hyperglycaemia, dyslipidaemia and hypertension combine with the duration and type of diabetes to define the distinct pathophysiology underlying diabetic kidney disease, diabetic retinopathy and diabetic neuropathy. Only recently have the commonalities and differences in the metabolic basis of these tissue-specific complications, particularly those involving local and systemic lipids, been systematically examined. This review focuses on recent progress made using preclinical models and human-based approaches towards understanding how bioenergetics and metabolomic profiles contribute to diabetic kidney disease, diabetic retinopathy and diabetic neuropathy. This new understanding of the biology of complication-prone tissues highlights the need for organ-specific interventions in the treatment of diabetic complications.

Peripheral neuropathy (PN) is a complication of prediabetes and type 2 diabetes (T2D). Increasing evidence suggests that factors besides hyperglycaemia contribute to PN development, including dyslipidaemia. The objective of this study was to determine differential lipid classes and altered gene expression profiles in prediabetes and T2D mouse models in order to identify the dysregulated pathways in PN. Here, we used high-fat diet (HFD)-induced prediabetes and HFD/streptozotocin (STZ)-induced T2D mouse models that develop PN. These models were compared to HFD and HFD-STZ mice that were subjected to dietary reversal. Both untargeted and targeted lipidomic profiling, and gene expression profiling were performed on sciatic nerves. Lipidomic and transcriptomic profiles were then integrated using complex correlation analyses, and biological meaning was inferred from known lipid-gene interactions in the literature. We found an increase in triglycerides (TGs) containing saturated fatty acids. In parallel, transcriptomic analysis confirmed the dysregulation of lipid pathways. Integration of lipidomic and transcriptomic analyses identified an increase in diacylglycerol acyltransferase 2 (DGAT2), the enzyme required for the last and committed step in TG synthesis. Increased DGAT2 expression was present not only in the murine models but also in sural nerve biopsies from hyperlipidaemic diabetic patients with PN. Collectively, these findings support the hypothesis that abnormal nerve-lipid signalling is an important factor in peripheral nerve dysfunction in both prediabetes and T2D.

Aim: Podocyte apoptosis is a critical mechanism for excessive loss of urinary albumin that eventuates in kidney fibrosis. Oxidative stress plays a critical role in hyperglycemia-induced glomerular injury. We explored the hypothesis that mammalian target of rapamycin complex 2 (mTORC2) mediates podocyte injury in diabetes. Results: High glucose (HG)-induced podocyte injury reflected by alterations in the slit diaphragm protein podocin and podocyte depletion/apoptosis. This was paralleled by activation of the Rictor/mTORC2/Akt pathway. HG also increased the levels of Nox4 and NADPH oxidase activity. Inhibition of mTORC2 using small interfering RNA (siRNA)-targeting Rictor in vitro decreased HG-induced Nox1 and Nox4, NADPH oxidase activity, restored podocin levels, and reduced podocyte depletion/apoptosis. Inhibition of mTORC2 had no effect on mammalian target of rapamycin complex 1 (mTORC1) activation, described by our group to be increased in diabetes, suggesting that the mTORC2 activation by HG could mediate podocyte injury independently of mTORC1. In isolated glomeruli of OVE26 mice, there was a similar activation of the Rictor/mTORC2/Akt signaling pathway with increase in Nox4 and NADPH oxidase activity. Inhibition of mTORC2 using antisense oligonucleotides targeting Rictor restored podocin levels, reduced podocyte depletion/apoptosis, and attenuated glomerular injury and albuminuria. Innovation: Our data provide evidence for a novel function of mTORC2 in NADPH oxidasederived reactive oxygen species generation and podocyte apoptosis that contributes to urinary albumin excretion in type 1 diabetes. Conclusion: mTORC2 and/or NADPH oxidase inhibition may represent a therapeutic modality for diabetic kidney disease. Antioxid. Redox Signal. 25, 703-719.

Polycystic ovary syndrome (PCOS) is associated with reproductive, endocrine, and metabolic abnormalities. Because hyperandrogenism is the most consistent PCOS feature, we used wild-type (WT) and androgen receptor (AR) knockout (ARKO) mice, together with a mouse model of PCOS, to investigate the contribution of genomic AR-mediated actions in the development of PCOS traits. PCOS features were induced by prenatal exposure to dihydrotestosterone (250 μg) or oil vehicle (control) on days 16-18 of gestation in WT, heterozygote, and homozygote ARKO mice. DHT treatment of WT mice induced ovarian cysts (100% vs 0%), disrupted estrous cycles (42% vs 100% cycling), and led to fewer corpora lutea (5.0±0.4 vs 9.8±1.8). However, diestrus serum LH and FSH, and estradiol-induced-negative feedback as well as hypothalamic expression of kisspeptin, neurokinin B, and dynorphin, were unaffected by DHT treatment in WT mice. DHT-treated WT mice exhibited a more than 48% increase in adipocyte area but without changes in body fat. In contrast, heterozygous and homozygous ARKO mice exposed to DHT maintained comparable ovarian (histo)morphology, estrous cycling, and corpora lutea numbers, without any increase in adipocyte size. These findings provide strong evidence that genomic AR signaling is an important mediator in the development of these PCOS traits with a dose dependency that allows even AR haplosufficiency to prevent induction by prenatal androgenization of PCOS features in adult life.