An important goal in diabetes research is to understand the processes that trigger endogenous β-cell proliferation. Hyperglycemia induces β-cell replication, but the mechanism remains debated. A prime candidate is insulin, which acts locally through the insulin receptor. Having previously developed an in vivo mouse hyperglycemia model, we tested whether glucose induces β-cell proliferation through insulin signaling. By using mice lacking insulin signaling intermediate insulin receptor substrate 2 (IRS2), we confirmed that hyperglycemia-induced β-cell proliferation requires IRS2 both in vivo and ex vivo. Of note, insulin receptor activation was not required for glucose-induced proliferation, and insulin itself was not sufficient to drive replication. Glucose and insulin caused similar acute signaling in mouse islets, but chronic signaling differed markedly, with mammalian target of rapamycin (MTOR) and extracellular signal–related kinase (ERK) activation by glucose and AKT activation by insulin. MTOR but not ERK activation was required for glucose-induced proliferation. Cyclin D2 was necessary for glucose-induced β-cell proliferation. Cyclin D2 expression was reduced when either IRS2 or MTOR signaling was lost, and restoring cyclin D2 expression rescued the proliferation defect. Human islets shared many of these regulatory pathways. Taken together, these results support a model in which IRS2, MTOR, and cyclin D2, but not the insulin receptor, mediate glucose-induced proliferation.

Aims/hypothesis We determined whether hyperglycaemia stimulates human beta cell replication in vivo in an islet transplant model Methods Human islets were transplanted into streptozotocininduced diabetic NOD-severe combined immunodeficiency mice. Blood glucose was measured serially during a 2 week graft revascularisation period. Engrafted mice were then catheterised in the femoral artery and vein, and infused intravenously with BrdU for 4 days to label replicating beta cells. Mice with restored normoglycaemia were co-infused with either 0.9% (wt/vol.) saline or 50% (wt/vol.) glucose to generate glycaemic differences among grafts from the same donors. During infusions, blood glucose was measured daily. After infusion, human beta cell replication and apoptosis were measured in graft sections using immunofluorescence for insulin, and BrdU or TUNEL. Results Human islet grafts corrected diabetes in the majority of cases. Among grafts from the same donor, human beta cell proliferation doubled in those exposed to higher glucose relative to lower glucose. Across the entire cohort of grafts, higher blood glucose was strongly correlated with increased beta cell replication. Beta cell replication rates were unrelated to circulating human insulin levels or donor age, but tended to correlate with donor BMI. Beta cell TUNEL reactivity was not measurably increased in grafts exposed to elevated blood glucose. Conclusions/interpretation Glucose is a mitogenic stimulus for transplanted human beta cells in vivo. Investigating the underlying pathways may point to mechanisms capable of expanding human beta cell mass in vivo.

This pilot study compared daytime symptom ratings in primary insomniacs (n = 7) and age-matched controls (n = 8). Participants completed sleep diaries and rated their daytime symptoms using a Daytime Symptom Diary (DSD) 4 times per day for 1 week. DSD responses were collapsed into 4 domains: mood, subjective alertness, energy, and concentration. The level and variability of DSD domains, and correlations between the domains and sleep diary characteristics, were examined. Significant Group x Time of Day interactions were observed in values for each DSD domain, with the most consistent group differences occurring in the morning. Coefficients of variation for DSD domains were greater in the insomnia group. Frequent measures of daytime symptoms may be useful outcomes in insomnia studies.

Calcium, phosphorus, and parathyroid hormone (PTH) levels are routinely measured in patients undergoing chronic hemodialysis. Medications, diet, and dialytic therapies are modified based upon these lab values to achieve specific goal values in the hope of improving outcomes. However, the variability of these values in patients undergoing chronic hemodialysis has only been rarely studied. We prospectively investigated the variability of these measures in 35 patients undergoing chronic hemodialysis as well as the impact of this variability on clinical decision-making in a prospective manner over a month. There is significant session-to-session variability in phosphorus and PTH values (mean coefficient of variations [CV] of 0.19 and 0.31, respectively). Calcium variability is much lower (mean CV of 0.05). Not surprisingly, the CV for all values is increased during the long interdialytic interval. The impact of this variability on clinical decision-making was analyzed. The variability in calcium, phosphorus, and PTH values would lead to a different clinical decision in 23.6%, 41.2%, and 39.7% of different session lab values. We also investigated the variability of these lab measures over a year in these patients and found that the session-to-session variability was very similar to the month-to-month variability. The high degree of variability of these parameters has important implications for clinical decision-making and for implementation of pay-for-performance measures.

No abstract