br Funding This work was

Funding This work was supported by the grants from MSD, Terumo Life Science Foundation, Takeda Science Foundation, and Japan Diabetes Foundation.
Disclosures
Acknowledgments The authors thank C. Morimoto (Juntendo University, Tokyo, Japan) and K. Takeda (Immunology Frontier Research Center, Osaka University, Osaka, Japan) for generously providing CD26−/− mice and STAT3flox/flox mice, respectively. The authors thank Y. Ohtsuki, M. Furuyama, and Y. Shimizu for their excellent technical assistance.
Introduction The story leading to the development of inhibitors of the enzyme dipeptidyl peptidase-4 (DPP-4) as a successful therapy for type 2 diabetes (T2DM) could almost be described as a textbook example of how an understanding of basic physiology/endocrinology can be exploited to lead to rational prospective drug design based on a prior knowledge of both the target and the underlying mechanism of action. In this respect, it contrasts to the development of some other multidrug resistant of oral anti-hyperglycaemic agents, such as the biguanides, sulphonlylureas and thiazolidinediones, whose glucose-lowering activities were discovered before it was understood how they worked. In the example of the DPP-4 inhibitors, the finding that the incretin hormone, glucagon-like peptide-1 (GLP-1) was particularly susceptible to degradation by DPP-4 prompted the suggestion that targeting this route of metabolism may be a novel therapeutic strategy for managing T2DM [1], spurring research into the role of DPP-4 in GLP-1 biology and glucose metabolism and ultimately ending with the development and launch of a successful medicine, the DPP-4 inhibitor (a.k.a. gliptin) class of antihyperglycaemic drugs. GLP-1 is a gut peptide which was discovered following cloning of the gene encoding the pancreatic hormone, glucagon. As well as the sequence of glucagon itself, the proglucagon gene was found to code for two additional sequences which bore strong homology to glucagon, and the existence of two related peptides, GLP-1 and GLP-2, was predicted [2]. GLP-1 was subsequently identified in intestinal extracts from pigs [3] and rats [4], and found to be highly potent in stimulating insulin [5], [6] and inhibiting glucagon secretion [7] from the perfused pancreas. Later, GLP-1 was shown to be released in response to food ingestion [8], [9] and to act as an incretin hormone, markedly enhancing meal-stimulated insulin secretion [10]. The incretin effect (i.e. the greater insulin response to oral – as opposed to isoglycaemic intravenous infusion of – glucose) was known to be impaired in T2DM [11], so it was of particular interest to find that GLP-1 was insulinotropic in patients with T2DM, and could normalise fasting glucose levels when given intravenously [12]. This raised interest in the possibility of using GLP-1 therapeutically. It came as some disappointment then, when the effects of a single subcutaneous injection of GLP-1 had only a short-lived effect on insulin secretion and failed to bring fasting glucose levels back into the normal range [13]. Nevertheless, repeated subcutaneous injections were effective [13], and when given by continuous subcutaneous infusion, continuous exposure to GLP-1 over 6 weeks improved glycaemic control and caused a small drop in body weight in subjects with T2DM [14], establishing GLP-1 as a viable drug target. At the time, the transient nature of the effect of a single subcutaneous injection of GLP-1 on blood glucose was entirely unexpected. Plasma levels of GLP-1 seemed to be raised significantly [15], and attained levels well above those which were associated with a glucose-lowering effect during an intravenous infusion of GLP-1 [12]. There therefore appeared to be nothing to suggest that the peptide was not stable after the subcutaneous injection, and the reason for the short-lasting effect of GLP-1 remained unexplained.
The role of DPP-4 in incretin hormone metabolism