β-Hydroxypyruvate: A New Diabetogenic Factor?

The incidence of type 2 diabetes mellitus (T2DM) is increasing at an alarming rate. Insulin secretory dysfunction and insulin resistance are characteristic features of T2DM. However, their relative contributions to the progression from normal glucose tolerance to impaired glucose tolerance (IGT) to T2DM remain unknown. This limits the capacity of researchers to establish therapeutic interventions for T2DM. Glucose-dependent insulinotropic polypeptide (GIP) and GLP-1 are the two primary incretin hormones secreted by intestinal K and L cells, respectively. Secretion of these incretins increases postprandially and they potentiate glucose-stimulated insulin secretion (GSIS). Despite blunted incretin responses in patients with T2DM, insulinotropic responses to exogenously administered GLP-1 remain active in these patients (1). Although controversy exists, a recent study demonstrated that insulin secretory responses to GIP were retained in patients with T2DM (1,2). Additionally, incretin response to GIP is improved by reducing blood glucose levels in patients with T2DM (3). These findings support the rationale for use of incretin-based pharmacotherapies in the treatment and management of T2DM. However, identification of the mechanism by which the function of K cells contributes to the development of T2DM is still elusive and requires a good model of study.

To determine the role of K cells in the regulation of metabolism, K cells were eliminated from mice by expressing the diphtheria toxin A-chain gene in GIP-producing cells (DT mice). Despite a severely impaired incretin response, DT mice on a C57BL/6J background remained normoglycemic and did not develop high-fat diet–induced insulin resistance (4). In this issue of Diabetes, Zhang et al. (5) undertook studies to determine the role of K cells in the regulation of the incretin response and the development of T2DM and to identify a potential mediator of T2DM. When DT mice were backcrossed onto the diabetogenic NONcNZO10/Ltj background and fed with a high-fat diet, they showed hyperglycemia and blunted GSIS due to the lack of incretin response of GIP and the lack of compensatory increase in insulin secretion, mimicking human T2DM (5). Thus, defective K-cell function and/or impaired action of K-cell products contribute to the development of T2DM. Furthermore, these findings qualify the high-fat diet–fed DT mice on this background as a valuable animal model to study the pathophysiology of human T2DM.

Zhang et al. took advantage of this animal model and performed biochemical profiling on plasma 1 week after weaning onto the high-fat diet, well before the occurrence of T2DM. As hyperglycemia occurred as a function of diet and genotype, they focused their analysis on metabolites exhibiting a diet-genotype interaction. Plasma β-hydroxypyruvate levels exhibited a significant diet-genotype interaction and increased in proportion to plasma glucose levels. Thus, increased β-hydroxypyruvate levels may be causative in initiating impairments in GSIS and the development of T2DM. What is the potential mechanism underlying the increased β-hydroxypyruvate levels in these mice? d-amino acid oxidase (DAO) deaminates d-serine to β-hydroxypyruvate and benzoate inhibits the activity of DAO (Fig. 1). Interestingly, benzoate levels were reduced in DT mice compared with wild-type mice. There was a significant inverse correlation between β-hydroxypyruvate and d-serine in humans. Thus, at least some β-hydroxypyruvate is derived from d-serine via DAO.

A somewhat surprising finding is that the ratio of β-hydroxypyruvate to d-serine was lower in individuals with IGT compared with those with normal glucose tolerance and patients with T2DM. If an elevated β-hydroxypyruvate level is the causative change of T2DM, one would expect to see increased β-hydroxypyruvate levels or increased β-hydroxypyruvate/d-serine ratios in individuals with IGT. In addition to GIP, K cells produce other molecules. Xenin-25 is a neurotensin-related peptide produced by a subset of GIP-producing cells and potentiates the incretin effect of GIP possibly via neurotensin receptor 1 and cholinergic enteric neurons (6–9) (Fig. 1). Xenin-25 amplified the effect of GIP on insulin secretion in individuals with IGT, compensating for increased insulin demands. In contrast, xenin-25 had no effect on GIP-induced insulin secretion in patients with T2DM (2). The reciprocal relationship between the β-hydroxypyruvate/d-serine ratio and the response to xenin-25 supports the possibility that changes in DAO activity affect insulin secretion by altering d-serine and β-hydroxypyruvate levels (Fig. 1). How does increased β-hydroxypyruvate cause T2DM? β-Hydroxypyruvate treatment reduced insulin content of islets in mice and reduced excitation of myenteric neurons isolated from guinea pig small intestines (5). Thus, it is likely that increased β-hydroxypyruvate promotes the development of T2DM at least partly by causing impairments in pancreatic enteric neuronal function and reducing insulin production (Fig. 1).

As plasma β-hydroxypyruvate levels increased in proportion to plasma glucose levels regardless of diet or genotype before developing T2DM, increased β-hydroxypyruvate is a potential causative factor for the development of T2DM. If β-hydroxypyruvate plays a significant role in the regulation of pancreatic function, then it is necessary to determine the conditions under which β-hydroxypyruvate levels are elevated in future studies. Plasma β-hydroxypyruvate levels were positively correlated with plasma glucose levels in mice and β-hydroxypyruvate/d-serine ratios were higher in patients with T2DM compared with people with IGT in the study by Zhang et al. (5). A previous study reported increased plasma serine (not necessarily d-serine) levels in diabetic nonhuman primates compared with nondiabetic control subjects (10). Thus, future studies should include the measurements of β-hydroxypyruvate, d-serine, and benzoate levels as well as β-hydroxypyruvate/d-serine ratio and DAO activity. If increased β-hydroxypyruvate levels contribute to the development of T2DM, then reducing β-hydroxypyruvate levels may be beneficial in reversing pancreatic dysfunction and hyperglycemia in diabetes. This strategy may include an inhibition of DAO and an increase in benzoate levels. Interestingly, oral administration of cinnamon increases the serum level of sodium benzoate in mice (11) and consumption of cinnamon lowers blood glucose levels in patients with T2DM (12). These findings further support the idea that reducing β-hydroxypyruvate levels is a worthwhile and testable strategy to restore pancreatic function and glycemic control in patients with T2DM. Thus, the study by Zhang et al. (5) raises many fascinating questions for future analysis and these future studies may open up a new avenue toward the development of therapeutic interventions for the treatment and management of T2DM.