β-Cell Identity in Type 2 Diabetes: Lost or Found?

Type 2 diabetes (T2DM) is characterized by deficits of β-cell mass and function. The deficit in β-cell mass in patients with T2DM is up to ∼65% compared with those without diabetes (1). This deficit in β-cell mass has been attributed to increased β-cell apoptosis (1) due to endoplasmic reticulum stress mediated by protein misfolding due to the accumulation of toxic oligomers of islet amyloid polypeptide (2) and/or gluco- or lipotoxicity (3) and disruption of the mitochondrial network (4). Alternatively, decreased β-cell mass in T2DM may reflect insufficient expansion of β-cell mass during fetal and neonatal life (5) or a failure to adaptively regenerate β-cells in response to β-cell loss. Recent studies have shown that loss of β-cell identity by de-differentiation may contribute to the measured β-cell deficit in mouse models of T2DM characterized by impaired leptin signaling, a finding not yet confirmed in humans with T2DM (6).

In this issue of Diabetes, Spijker et al. (7) present data highlighting the loss of β-cell identity in T2DM in humans and nonhuman primates. The authors confirm an increased ratio of α-cells to β-cells reported previously (8) and predicted by the selective loss of β-cells in T2DM. They also confirm an increased frequency of cells coexpressing insulin and glucagon in T2DM (0.52 ± 0.18% vs. 4.05 ± 1.37%; nondiabetic donors vs. T2DM; P < 0.01), as previously reported (0.4 ± 0.1% vs. 3.2 ± 1.4%; nondiabetic donors vs. T2DM; P < 0.05) (9). The authors note that about half of these bihormonal cells lack the β-cell transcription factor Nkx6.1 in both T2DM and control subjects. They report that more than half of cells positive for Nkx6.1 and glucagon do not express insulin. In the majority of T2DM subjects, the transcription factors MafA, Pdx1, and FOXO1 display aberrant islet subcellular localization, reminiscent of their expression pattern in diabetic mice (6) and humans with T2DM (10,11). The data in diabetic nonhuman primates further emphasize that altered endocrine cell identity is a conserved feature associated with T2DM. Interestingly, the authors observe a greater frequency of Nkx6.1⁺glucagon⁺insulin⁻ cells in areas of the pancreas positive for islet amyloid.

Loss of β-cell identity is a potential contributory factor toward β-cell dysfunction in diabetes, although the percentage of cells described in the current study (7) (∼4%) seems too small to induce diabetes. The studies in mouse models have suggested that β-cell de-differentiation is a likely explanation for the presence of endocrine cells with “confused” identity (6), a view also proposed by Spijker et al. As an alternative, these aberrant cells may be an early developmental endocrine cell type and a reflection of attempted compensatory β-cell neogenesis. While neogenesis is rare in mouse models (12,13), some evidence suggests that it may occur in human and primate pancreata (14,15). Spijker et al. cite studies showing that bihormonal cells are present in both obese prediabetic and diabetic states, but β-cell mass is reduced only in diabetes, arguing against bihormonal cells contributing significantly to β-cell loss.

Studies of endocrine identity in T2DM to date, including the work by Spijker et al. (7), have focused on α-cell to β-cell transition. Plasticity of pancreatic cells is well documented, with several recent studies featuring multiple pancreatic cell types, such as δ-cells and ductal cells, as important players in cellular transdifferentiation, possibly toward neogenesis (16–18). It is therefore relevant to examine the contribution of these lineages toward the aberrant endocrine cells in T2DM and also to determine if mechanisms such as neogenesis and de-differentiation operate in parallel, resulting in cells with mixed endocrine identity (Fig. 1). An interesting question posed by Spijker et al. is how metabolic changes in diabetes contribute to altered endocrine identity. Changes in nutrient sensing via the FOXO1 pathway, which also preserves cellular identity, may be one such contributing factor (6). Epigenetic mechanisms, such as DNA methylation that regulates the endocrine cell identity, are also sensitive to the cellular metabolic status (19,20). Metabolic changes in diabetes may lead to dysregulation of DNA methylation, causing DNA damage along with altered identity. Such cells may then undergo apoptosis, in turn triggering further attempted neogenesis to compensate the loss of endocrine mass.

Islets in T2DM are characterized by amyloid deposits (2). Spijker et al. (7) note an increased proportion of cells with impaired endocrine identity in the vicinity of extracellular amyloid. Amyloid, however, was not shown to be the direct cause of loss of β-cell identity in T2DM. The authors cite a study reporting that amyloid accumulation can induce inflammation (21). Inflammatory signals could potentially drive endocrine transdifferentiation or de-differentiation, leading to β-cell dysfunction. Equally likely, stressed β-cells (due to intracellular accumulation of toxic oligomers of islet amyloid polypeptide subsequently shed as extracellular amyloid [2]) may secrete cytokines that trigger neogenesis, producing “confused” endocrine cells.

Understanding the mechanisms that contribute to the failure of endocrine identity in T2DM may be important to developing therapeutic strategies that aim to restore β-cell function. Driving β-cell replication or transdifferentiation from other pancreatic cells could result in cells with mixed endocrine identity, which may not be glucose responsive. Notably, T2DM subjects on incretin therapy were found to have a marked increased β-cell mass and a notable increase in the proportion of insulin/glucagon-expressing cells (∼17%) (9), but as the study subjects still had diabetes, the expanded β-cell mass may not have been functional. Spijker et al. (7) imply that bihormonal cells can directly contribute to the loss of β-cell mass in T2DM. As these cells stain for insulin, they would have been considered β-cells in prior studies of decreased β-cell mass. Thus, if one chooses not to define these bihormonal cells as β-cells, these cells amount to further deficit in β-cell mass, in addition to the loss of β-cell mass previously reported in T2DM. Further studies are necessary to determine the origins and relevance of the small percentage of bihormonal endocrine cells confirmed here as more common in T2DM.