Altered pancreas remodeling following glucose intolerance in pregnancy in mice

Gestational diabetes mellitus increases the risk of dysglycemia postpartum, in part, due to pancreatic β-cell dysfunction. However, no histological evidence exists comparing endocrine pancreas after healthy and glucose-intolerant pregnancies. This study sought to address this knowledge gap, in addition to exploring the contribution of an inflammatory environment to changes in endocrine pancreas after parturition. We used a previously established mouse model of gestational glucose intolerance induced by dietary low protein insult from conception until weaning. Pancreas and adipose samples were collected at 7, 30 and 90 days postpartum for histomorphometric and cytokine analyses, respectively. Glucose tolerance tests were performed prior to euthanasia and blood was collected via cardiac puncture. Pregnant female mice born to dams fed a low protein diet previously shown to develop glucose intolerance at late gestation relative to controls continued to be glucose intolerant until 1 month postpartum. However, glucose tolerance normalized by 3 months postpartum. Glucose intolerance at 7 days postpartum was associated with lower beta- and alpha-cell fractional areas and higher adipose levels of pro-inflammatory cytokine, interleukin-6. By 3 months postpartum, a compensatory increase in the number of small islets and a higher insulin to glucagon ratio likely enabled euglycemia to be attained in the previously glucose-intolerant mice. The results show that impairments in endocrine pancreas compensation in hyperglycemic pregnancy persist after parturition and contribute to prolonged glucose intolerance. These impairments may increase the susceptibility to development of future type 2 diabetes.

A mouse model of gestational glucose intolerance through exposure to a low protein diet during fetal and neonatal development

KEY POINTS

Pancreatic β-cell dysfunction is hypothesized to be the significant determinant of gestational diabetes pathogenesis, however pancreatic samples from patients are scarce. This study reports a novel mouse model of gestational glucose intolerance in pregnancy, originating from previous nutrition restriction in utero, in which glucose intolerance was restricted to late gestation as is seen in human gestational diabetes. Glucose intolerance was attributed to reduced β-cell proliferation, leading to impaired gestational β-cell mass expansion in maternal endocrine pancreas, in addition to reduced glucose-stimulated insulin secretion. This model reproduces some of the features of gestational diabetes and is suitable for testing safe therapeutic interventions that increase β-cell mass during pregnancy and prevent or reverse gestational glucose intolerance.

ABSTRACT

Gestational diabetes mellitus (GDM) is an increasingly prevalent form of diabetes that appears during pregnancy. Pathological studies link a failure to adaptively increase maternal pancreatic β-cell mass (BCM) in pregnancy to GDM. Due to the scarcity of pancreatic samples from GDM patients, we sought to develop a novel mouse model for impaired gestational glucose tolerance. Mature female C57Bl/6 mouse offspring (F1) born to dams fed either a control (C) or low-protein (LP) diet during gestation and lactation were randomly allocated into two subsequent study groups: pregnant (CP, LPP) or non-pregnant (CNP, LPNP). Glucose tolerance tests were performed at gestational day (GD) 9, 12 and 18. Subsequently, pancreata were removed for fluorescence immunohistochemistry to assess α-cell mass (ACM), BCM and β-cell proliferation. An additional group of animals was used to measure insulin secretion from isolated islets at GD18. LPP females displayed glucose intolerance compared to CP females at GD18 (P < 0.001). BCM increased threefold at GD18 in CP females. However, LPP females had reduced BCM expansion (P < 0.01) concurrent with reduced β-cell proliferation at GD12 (P < 0.05). LPP females also had reduced ACM expansion at GD18 (P < 0.01). LPP islets had impaired glucose-stimulated insulin secretion in vitro compared to CP islets (P < 0.01). Therefore, impaired glucose tolerance during pregnancy is associated with a failure to adequately adapt BCM, as a result of reduced β-cell proliferation, in addition to lower glucose-stimulated insulin secretion. This model could be used to evaluate novel interventions during pregnancy to increase BCM or function as a strategy to prevent/reverse GDM.

Quantification of cellular proliferation in experimental proliferative vitreoretinopathy

Proliferation of host cells from around the optic nerve head has recently been implicated in the development of experimental proliferative vitreoretinopathy in rabbit eyes injected with homologous fibroblasts. We used liquid scintillation spectrometry to quantitate the tritiated thymidine incorporation into cells in the vitreous, retina, and optic nerve head following intravitreal injection of 250,000 homologous dermal fibroblasts. Cellular proliferation peaked three days after injection of the fibroblasts. The amount of tritiated thymidine incorporation that occurred three days following injection of irradiated homologous fibroblasts (incapable of cellular division) was not significantly different than that following injection of normal homologous fibroblasts, indicating that host cells were responsible for most of the cellular proliferation. Treatment with fluorouracil or triamcinolone acetonide completely arrested cellular proliferation following injection of normal fibroblasts.

The conformation of cross-linked actin.S-1 in the presence and absence of ATP

Electron microscopy studies have shown that the structure of the complex of myosin subfragment 1 (S-1) cross-linked to actin with 1-ethyl-3-[3-(dimethyl-amino) propyl] carbodiimide is very different in the presence and absence of ATP (Craig, R., Greene, L. E., and Eisenberg, E. (1985) Proc. Natl. Acad. Sci. U. S. A. 82, 3247-3251). More recent studies have found that the structure of the cross-linked complex between S-1 modified extensively with N-ethylmaleimide (NEM.S-1) and actin resembles that of the rigor complex both in the presence and absence of ATP, whereas the structure of the cross-linked complex between S-1 modified with N',N'-p-phenylenedimaleimide (pPDM.S-1) and actin resembles that of the cross-linked actin.S-1 complex in the presence of ATP. In the present study, we have obtained biochemical evidence supporting these results. The conformation of the different cross-linked actin.S-1 complexes was determined by studying their effect on the troponin-tropomyosin-actin complex (regulated actin). The basis of this probe for conformation is that S-1.ATP, which is in the weak-binding conformation, interacts very differently with regulated actin than S-1 or S-1.ADP, which are in the strong-binding conformation. We find that both in the presence and absence of ATP, cross-linked NEM.S-1 appears to be in the strong-binding conformation, whereas cross-linked pPDM.S-1 appears to be shifted toward the weak-binding conformation. In contrast, cross-linked unmodified S-1 appears to be in the strong-binding conformation in the presence of ADP and the weak-binding conformation in the presence of ATP. Therefore, in agreement with electron microscopy studies, the cross-linked actin.S-1 complex appears to be able to alternate between the weak-binding and strong-binding conformation during the cross-bridge cycle.

Regulation of the adenosinetriphosphatase activity of cross-linked actin-myosin subfragment 1 by troponin-tropomyosin

Chalovich and Eisenberg [Chalovich, J. M., & Eisenberg, E. (1982) J. Biol. Chem. 257, 2432-2437] have suggested that at low ionic strength, troponin-tropomyosin regulates the actomyosin ATPase activity by inhibiting a kinetic step in the actomyosin ATPase cycle rather than by blocking the binding of myosin subfragment 1 (S-1) to actin. This leads to the prediction that troponin-tropomyosin should inhibit the ATPase activity of the complex of actin and S-1 (acto . S-1) even when S-1 is cross-linked to actin. We now find that the ATPase activity of cross-linked actin . S-1 prepared under milder conditions than those used by Mornet et al. [Mornet, D., Bertrand, R., Pantel, P., Audemard, E., & Kassab, R. (1981) Nature (London) 292, 301-306] is inhibited 90% by troponin-tropomyosin in the absence of Ca2+. At mu = 18 mM, 25 degrees C, the ATPase activity of this cross-linked preparation is only about 2-fold greater than the maximal actin-activated ATPase activity of S-1 obtained with regulated actin in the absence of Ca2+. At physiological ionic strength, the ATPase activity of this cross-linked actin . S-1 preparation is inhibited about 95% by troponin-tropomyosin. Since cross-linked S-1 behaves kinetically like S-1 in the presence of infinite actin concentration, it is very unlikely that inhibition of the ATPase activity of cross-linked actin . S-1 is due to blocking of the binding of S-1 to actin. Therefore, these results are in agreement with the suggestion that troponin-tropomyosin regulates primarily by inhibiting a kinetic step in the ATPase cycle.