Ellenbroek JH, van Dijck L, Töns HA, Rabelink TJ, Carlotti F, Ballieux BE, de Koning EJ. Long-term ketogenic diet causes glucose intolerance and reduced ␤-and ␣-cell mass but no weight loss in mice. Am J Physiol Endocrinol Metab 306: E552-E558, 2014. First published January 7, 2014 doi:10.1152/ajpendo.00453.2013.-High-fat, low-carbohydrate ketogenic diets (KD) are used for weight loss and for treatment of refractory epilepsy. Recently, short-time studies in rodents have shown that, besides their beneficial effect on body weight, KD lead to glucose intolerance and insulin resistance. However, the long-term effects on pancreatic endocrine cells are unknown. In this study we investigate the effects of long-term KD on glucose tolerance and ␤-and ␣-cell mass in mice. Despite an initial weight loss, KD did not result in weight loss after 22 wk. Plasma markers associated with dyslipidemia and inflammation (cholesterol, triglycerides, leptin, monocyte chemotactic protein-1, IL-1␤, and IL-6) were increased, and KD-fed mice showed signs of hepatic steatosis after 22 wk of diet. Long-term KD resulted in glucose intolerance that was associated with insufficient insulin secretion from ␤-cells. After 22 wk, insulin-stimulated glucose uptake was reduced. A reduction in ␤-cell mass was observed in KD-fed mice together with an increased number of smaller islets. Also ␣-cell mass was markedly decreased, resulting in a lower ␣-to ␤-cell ratio. Our data show that long-term KD causes dyslipidemia, a proinflammatory state, signs of hepatic steatosis, glucose intolerance, and a reduction in ␤-and ␣-cell mass, but no weight loss. This indicates that long-term high-fat, lowcarbohydrate KD lead to features that are also associated with the metabolic syndrome and an increased risk for type 2 diabetes in humans. pancreatic islet; ketogenic diet; glucose intolerance; ␤-cell; ␣-cell

SummaryClinical studies have shown that plasma insulin levels are closely related to plasma plasminogen activator inhibitor 1 (PAI-1) levels. To investigate a possible involvement of hepatocytes we have studied the effect of insulin on PAI-1 production by primary cultures of human hepatocytes. We have isolated human hepatocytes from seven left liver lobes. PAI-1 activity measured in 24 hours conditioned medium varied considerably between the various hepatocyte preparations (from 2.9 to 8.5 units per 5 cm2of cells) possibly as a result of interindividual variability in basal PAI-1 production by hepatocytes from different donors. In all cases, however, the relative extent, time profile and dose-dependency of the insulin-induced increase in PAI-1 synthesis were consistent. Up to about 7 nM, insulin dose-dependently increased both PAI-1 activity and PAI-1 antigen production. The increase in PAI-1 synthesis became measurable between 4 and 8 hours after addition of the hormone, and maximally reached twofold control values. The increase in PAI-1 synthesis could be fully explained by a concomitant increase in PAI-1 mRNA levels. The effect of insulin seems fairly specific for the synthesis of PAI-1: overall protein synthesis and mRNA levels of some control proteins (albumin and fibrinogen) did not markedly change after insulin addition. These results, obtained with primary cultures of human hepatocytes, are fully comparable with those obtained with the hepatocellular carcinoma cell line Hep G2. They strengthen the suggestion that the elevated level of PAI-1 in high insulin plasma might be the result of increased hepatic synthesis of PAI-1.

Aims/hypothesis Incretin-based therapies improve glycaemic control in patients with type 2 diabetes. In animal models of diabetes, glucagon-like peptide-1 receptor agonists (GLP-1RAs) increase beta cell mass. GLP-1RAs are also evaluated in nondiabetic individuals with obesity and cardiovascular disease. However, their effect on beta cell mass in normoglycaemic conditions is not clear. Here, we investigate the effects of the GLP-1RA liraglutide on beta cell mass and function in normoglycaemic mice. Methods C57BL/6J mice were treated with the GLP-1RA liraglutide or PBS and fed a control or high-fat diet (HFD) for 1 or 6 weeks. Glucose and insulin tolerance tests were performed after 6 weeks. BrdU was given to label proliferating cells 1 week before the animals were killed. The pancreas was taken for either histology or islet isolation followed by a glucose-induced insulin-secretion test. Results Treatment with liraglutide for 6 weeks led to increased insulin sensitivity and attenuation of HFD-induced insulin resistance. A reduction in beta cell mass was observed in liraglutide-treated control and HFD-fed mice at 6 weeks, and was associated with a lower beta cell proliferation rate after 1 week of treatment. A similar reduction in alpha cell mass occurred, resulting in an unchanged alpha to beta cell ratio. In contrast, acinar cell proliferation was increased. Finally, islets isolated from liraglutide-treated control mice had enhanced glucose-induced insulin secretion. Conclusions/interpretation Our data show that GLP-1RA treatment in normoglycaemic mice leads to increases in insulin sensitivity and beta cell function that are associated with reduced beta cell mass to maintain normoglycaemia.

AimsBeta cells adapt to an increased insulin demand by enhancing insulin secretion via increased beta cell function and/or increased beta cell number. While morphological and functional heterogeneity between individual islets exists, it is unknown whether regional differences in beta cell adaptation occur. Therefore we investigated beta cell adaptation throughout the pancreas in a model of high-fat diet (HFD)-induced insulin resistance in mice.MethodsC57BL/6J mice were fed a HFD to induce insulin resistance, or control diet for 6 weeks. The pancreas was divided in a duodenal (DR), gastric (GR) and splenic (SR) region and taken for either histology or islet isolation. The capacity of untreated islets from the three regions to adapt in an extrapancreatic location was assessed by transplantation under the kidney capsule of streptozotocin-treated mice.ResultsSR islets showed 70% increased beta cell proliferation after HFD, whereas no significant increase was found in DR and GR islets. Furthermore, isolated SR islets showed twofold enhanced glucose-induced insulin secretion after HFD, as compared with DR and GR islets. In contrast, transplantation of islets isolated from the three regions to an extrapancreatic location in diabetic mice led to a similar decrease in hyperglycemia and no difference in beta cell proliferation.ConclusionsHFD-induced insulin resistance leads to topologically heterogeneous beta cell adaptation and is most prominent in the splenic region of the pancreas. This topological heterogeneity in beta cell adaptation appears to result from extrinsic factors present in the islet microenvironment.

When studying histological characteristics of human donor-pancreata, a remarkably high number of hyperemic islets (HIs) were encountered. The abnormalities in these HIs ranged from single/multiple dilated vessels to hemorrhages extending into the exocrine tissue. We aimed to determine the relevance of the presence of HIs in human donor-pancreata for isolation outcome and to identify donor and procurement factors associated with the occurrence of HIs. The presence of HIs was scored semi-quantitatively (HI-, HI+) in 102 human donor-pancreata. Islet isolation was performed in 40 cases. Donor and procurement factors were retrospectively analyzed in 94 donors. HIs were found in 54.6% of all donor-pancreata. However, only 4.5% of all islets in the affected pancreata was hyperemic. The affected pancreata contained slightly more endocrine tissue, but produced significantly lower yields. When corrected for other factors known to influence isolation outcome, the presence of HIs and endocrine content were the only factors significantly influencing isolation outcome. Prolonged ICU stay and pre-procurement hypertension were associated with the presence of HIs. This study is a first indication that the presence of HIs in human donor-pancreata are associated with reduced isolation outcomes and suggest an impact of the procurement procedure and pre-procurement hemodynamic status of the donor on the islet quality. It is tempting to speculate that this contributes to the generally experienced difficulties in obtaining sufficient amounts of human islets.