In This Issue of Diabetes

Fasting ketone body levels are strongly and positively associated with incident type 2 diabetes in the general population independently of several recognized risk factors, according to Szili-Torok et al. (p. 1187). The findings suggest that ketone bodies act as early biomarkers for type 2 diabetes risk, but also add a cautionary note on ketogenic diets. The findings come from the Prevention of Renal and Vascular End-stage Disease (PREVEND) prospective longitudinal cohort study, which is investigating the predictive value of various biomarkers on the incidence of cardiovascular and renal disease. From the wider cohort, just over 3,300 participants from the general population who were specifically free of diabetes and impaired fasting glucose at baseline were included in the study. Fasting ketone body levels as well as a range of other clinical characteristics were determined at baseline. In the follow-up of ∼7 years, 126 individuals, or 3.8%, went on to develop type 2 diabetes; higher baseline ketone body levels entailed higher type 2 diabetes incidence. Crucially, adjustment for a series of risk factors for type 2 diabetes did not materially change the association. Among others, these factors included age, sex, metabolic syndrome indicators, lipids, BMI, weight, and renal function. Based on their findings and those of a limited number of similar studies, the authors note that the relationship between ketones and diabetes risk is currently not totally consistent. As a result, they call for more research, ideally with samples from participants that were collected under fasting and nonfasting conditions. They also caution about the use of ketogenic diets, noting that high ketone body levels, which the diet induces, may well have an “opposite outcome than is intended.” “The biological role of ketone bodies is still only beginning to unfold,” said author Uwe J.F. Tietge. “It will be exciting to build on the present work and gather more mechanistic insights in future studies tailored to improving diabetes prevention and treatment.”

A long noncoding RNA called LOC100132249 may have a role in inducing endothelial dysfunction in diabetic retinopathy (DR), according to Hu et al. (p. 1307). Specifically, they report identifying the long noncoding RNA, its transmission by exosomes, and the mechanisms that might link it to proliferative DR. The findings, they suggest, might lead to the identification of novel biomarkers for diagnosis, treatment, and prevention strategies for DR. Current treatments have limitations in terms of effectiveness and adverse effects for patients, prompting the search for better approaches to the disease. The findings come from a mix of experiments with vitreous samples collected from patients with proliferative DR or with macular hole (as control participants), a mouse model, and a series of cell cultures. They also use a mix of molecular approaches to investigate the roles of exosomes loaded with long noncoding RNAs and the potential mechanisms involved in angiogenesis in DR. They found that exosomes from patients with proliferative DR promote angiogenesis in terms of proliferation, migration, and tube formation in retinal endothelial cells. Using high-throughput sequencing, they then identified LOC100132249 as the most significantly upregulated long noncoding RNA in exosomes from patients with proliferative DR, with levels 10 times those found in equivalent exosomes from patients with macular hole. Further experiments then demonstrated that LOC100132249 specifically contributed to angiogenesis in cells and the mouse model. Moving on to mechanisms, the authors determined that the long noncoding RNA likely acts as a “sponge” for microRNA-199a-5p. Accordingly, the authors propose that this regulates the endothelial-mesenchymal transition promoter SNAI1 via the Wnt/β-catenin pathway, which ultimately leads to endothelial disruption. Commenting further, author Zizhong Hu said, “Our findings underscore the pathogenic role of endothelial-derived exosomes in DR angiogenesis via a newly found long noncoding cRNA LOC100132249 and may provide new insights into the development of novel therapeutic targets in the future.”

Loss of the gene for carboxypeptidase E (Cpe) in pancreatic β-cells appears not to contribute to the spontaneous development of obesity and hyperglycemia in mice, according to Chen et al. (p. 1277). Conversely, elevated proinsulin levels in the model are likely to change β-cell glucose metabolism and increase their susceptibility to stress-induced dysfunction and, ultimately, the development of diabetes. CPE facilitates the conversion of proinsulin into insulin and C-peptide prior to secretion from β-cells; thus, failure of the process leads to insufficient insulin release, according to the authors. Indeed, individuals with Cpe mutations have elevated levels of proinsulin and often display characteristics of obesity, hyperglycemia, diabetes, and mental disorders. According to the authors, the findings and the model illuminate some of the very earliest changes in β-cells that likely occur in the development of diabetes. However, they also suggest that further studies will be needed in other prediabetes models and human islets to gain further insights into these early events. Ultimately, they suggest, these studies might uncover therapeutic targets to preserve β-cell function prior to the onset of diabetes. The findings come from a series of studies centered around a mouse model with pancreatic β-cell–specific knockout of Cpe that is characterized by a lack of mature insulin granules and increased levels of proinsulin. Surprisingly, however, they found that Cpe deletion that was specific to β-cells did not lead to the development of obesity and hyperglycemia even after a high-fat diet, in complete contrast to whole-body deletion of Cpe, where it does develop. They did find that β-cell proliferation in islets and β-cell area in the mice increased and that the mice remained with normoglycemia, hinting at some sort of “metabolic rewiring.” Genomic analysis revealed elevated gene expression related to glycolysis that hints at reduced mitochondrial activity and increased reactive oxygen species, among other changes.

Elevated levels of HbA_1c are significantly associated with liver fibrosis progression in patients with nonalcoholic fatty liver disease (NAFLD) and diabetes, according to Sako et al. (p. 1297). Conversely, elevated BMI only tended to be associated with liver fibrosis progression. The authors propose that this represents “diabetic steatohepatitis,” where hyperglycemia increases liver fibrosis independently of weight gain. The findings come from an extension of previous studies by the authors in which they look at serial liver biopsies from patients with and without diabetes with various stages of liver issues. Their aim was to identify clinical parameters, apart from obesity, that determine the progression of liver fibrosis—features that currently remain unclear. The authors analyzed 342 serial liver biopsies from 118 participants with NAFLD that were taken over a mean duration of just under 4 years. They found that overall, the fibrosis-4 index and components at baseline examination predicted future fibrosis progression in a trend test. Next, they examined changes in clinical parameters and liver pathology progression over time and found that patients with NAFLD and diabetes had elevated HbA_1c, but not BMI, that was associated with liver fibrosis progression. Conversely, this was not the case in patients with NAFLD and no diabetes. As such, the authors suggest hyperglycemia (i.e., HbA_1c) might be a valuable therapeutic target to prevent fibrosis progression in such patients. Addressing hepatic gene expression, the authors examined 33 of the patients’ biopsy samples and conclude that diabetes-induced hypoxia and oxidative stress most likely injure liver sinusoidal endothelial cells near the central vein, which in turn might lead to inflammation and stellate cell activation and eventually liver fibrosis. “We have found the independent contribution of hyperglycemia to liver fibrosis in Japanese people whose type 2 diabetes does not always accompany obesity,” said author Toshinari Takamura. “However, it may be challenging to observe it in Western people whose obesity and type 2 diabetes are hardly separated.”