In This Issue of Diabetes

Impairment of indoleamine 2,3 dioxygenase-1 (IDO1) expression in β-cells may play a role in their eventual destruction and the subsequent development of type 1 diabetes, according to Anquetil et al. (p. 1858). As a result, the authors suggest that IDO1 “could potentially emerge as a promising therapeutic target” for type 1 diabetes. The findings come from a study that used pancreatic tissue sections obtained from donors without diabetes, autoantibody-positive donors with prediabetes, patients with varying lengths of type 1 diabetes, and patients with type 2 diabetes. Sections were then subjected to double-indirect immunofluorescence staining for IDO1 and insulin. Initial investigations showed that IDO1 and insulin signals mostly overlapped, which the authors say indicates that IDO1 was expressed in β-cells. Depending on diabetes status, IDO1 was expressed at different rates, with significantly less IDO1 expression in insulin-containing islets from donors with type 1 diabetes compared to other groups. Expression of IDO1 was also heterogenous with samples having reduced expression (in comparison with donors with no diabetes) when they came from donors with prediabetes or recent diagnosis of type 1 diabetes, but not as much as was seen in samples from donors with much longer duration of diabetes. Autoantibody status (i.e., single or multiple) also influenced IDO1 expression. The authors recommend that further studies should focus on isolated human islets to better understand the role of IDO1 in type 1 diabetes. Author Matthias G. von Herrath told Diabetes: “It is remarkable to me that we now see more and more changes that occur in human islets prior to diagnosis of type 1 diabetes. In this case, we do not know yet what the loss of IDO in β-cells means. It could result in impaired immune protection or it could have something to do with β-cell survival. It’s exciting times.”

Anquetil et al. Loss of IDO1 expression from human pancreatic β-cells precedes their destruction during the development of type 1 diabetes. Diabetes 2018;67:1858–1866

A mouse model of atherosclerosis promises to close a major gap in mouse-based studies of vascular health in diabetes. According to Yuan et al. (p. 1880), they have introduced a method to study the reversal of atherosclerosis with an approach to increase aldose reductase (AR) expression in mice to levels nearing that of humans. Previously naturally low levels of AR expression in mice meant it could not be accounted for in studies, even though it is linked strongly to a number of diabetes complications. The authors describe a series of experiments with various mouse models where initially Ldlr^-/- mice were fed a high-fat diet for 16 weeks to induce atherosclerosis. Following that, aortas from those original mice were then transplanted into other Ldlr^-/- mice as a baseline and various other mice with different genetic/phenotypic backgrounds representing atherosclerosis regression or diabetes with or without human levels of AR (hAR). After 4 weeks of feeding with regular diet, the aortas were harvested to investigate the effects of raising AR levels on atherosclerotic regression. In terms of atherosclerotic lesions, when glucose and lipid levels were normal, lesions reduced significantly after lipid lowering. However, when hyperglycemia was present, lesions decreased at a much lower rate. In the mice with diabetes and hAR expression, lesions actually increased despite normal lipid levels. Although noting the utility of the model in terms of preclinical research, the authors point out that a number of new AR inhibitors are apparently in preclinical testing and suggest that, assuming their mouse data reflect reality in humans, these inhibitors might well exhibit cardiovascular disease benefits. Author Edward A. Fisher commented: “These studies may explain in part why patients with diabetes do not reduce their cardiovascular risk after statin therapy as well as patients without diabetes. It also provides a model for further investigations of potential mechanisms involved in atherosclerosis.”

Yuan et al. Human aldose reductase expression prevents atherosclerosis regression in diabetic mice. Diabetes 2018;67:1880–1891

Insulin-producing β-cells may transition between distinct phases of high and low insulin biosynthesis and to phases of recovery from cellular stress, according to Xin et al. (p. 1783), who provide a detailed analysis of the transcriptome and gene changes involved. According to the authors, cellular stress can arise from proinsulin protein misfolding, which is a fairly common occurrence at times of high insulin demand. To counteract this stress, an unfolded protein response (UPR) occurs to enhance protein folding, reduce endoplasmic reticulum load, and enhance clearance of misfolded proteins. Using large-scale single-cell RNA sequencing and a bioinformatics approach focused on pseudotime analysis, the authors detail how gene expression changes occur in human β-cells from donors without diabetes, according to activity and resting states. They reveal that β-cells can have three major states: low levels of both insulin and UPR gene expression, low UPR but high insulin gene expression, and high UPR but low insulin gene expression. Respectively, the three states represent rest, active insulin gene expression, and UPR gene expression, i.e., recovery. They go on to suggest that pseudotime analysis could provide useful information if it was applied to single β-cells from patients with type 2 diabetes. They speculate, for example, that β-cells might spend more time in high insulin gene expression and less in the recovery phase, potentially explaining why such patients have considerably reduced β-cell mass. Author Jesper Gromada said: “Our present findings show at the single-cell level how UPR activation helps in the resolution of ER [endoplasmic reticulum] stress caused by high insulin production in β-cells. Our data confirm the importance of UPR in stress resolution in β-cells and suggest that the maintenance of proteostasis governed by UPR is critical for β-cell health and renewal. The study of these functional states with regard to their influence on disease development and progression offers new and exciting avenues of investigation.”

Xin et al. Pseudotime ordering of single human β-cells reveals states of insulin production and unfolded protein response. Diabetes 2018;67:1783–1794

The role of T cells in the development of type 1 diabetes (T1D) is explored by Baker et al. (p. 1836), with the suggestion that there is a chance of using them as biomarkers of disease development. The key, the authors report, lies with MHC class II tetramers that can be used to track antigen-specific T cells and specifically to investigate the relative contributions of hybrid insulin peptide (HIP)-reactive T cells in the progression of T1D. With a focus on NOD mice, the authors compare frequency and phenotypes of different T-cell populations in pancreas, lymph nodes, and blood of mice at different stages of disease development. Mainly using different types of cytometry, they explore how different cell types are involved in different disease stages and ultimately conclude that HIP-reactive T cells are disease indicators, at least in the NOD mouse model. They demonstrate the presence of T cells reactive to two types of HIP in the pancreas of NOD mice and report that it appears they are activated very early in the disease process in the pancreatic lymph node. After undergoing alterations in phenotype, the T cells reportedly then exit into peripheral blood where they can be detected at increasing frequency through disease progression before entry into islets. This then seems to lead to an immune response resulting in β-cell destruction. Author Kathryn Haskins commented: “The NOD mouse has provided a highly valuable animal model for the discovery of HIPs and investigating the role of CD4 T cells reactive to these new autoantigens in autoimmune diabetes. HIP-reactive T cells have been found in the islets of T1D donors, and we are now testing whether T cells from T1D patients respond to HIPs to answer the question of whether these cells can also serve as biomarkers of disease activity in humans.”

Baker et al. CD4 T cells reactive to hybrid insulin peptides are indicators of disease activity in the NOD mouse. Diabetes 2018;67:1836–1846