Awardees 2009

PI: Elizabeth M. Bradshaw, Ph.D.
Type 1 autoimmune diabetes (T1D) is characterized by the presence of lymphocytic nfiltrates composed of activated T cells, monocytes/macrophages, dendritic cells and B cells that accumulate in pancreatic islets as a precursor to disease onset and thus may be involved in disease initiation leading to the destruction of β cells. However, we do not know what cell type drives this activated autoimmune state. Antigen-presenting cells secrete cytokines, express co-stimulation molecules and contribute to the amplification of the immune response. While examining peripheral blood mononuclear cells (PBMC) from subjects with T1D, we observed a striking activation state of a subset of monocytes isolated without stimulation from diabetic subjects that secreted IL-6 and IL-1β; this was not observed from control subject’s monocytes.  This is of particular interest as IL-1β with IL-6 is a potent inducer/expander of Th17 cells, a distinct helper T cell subset implicated in the pathogenesis of autoimmune diseases. Unstimulated monocytes from T1D subjects induced a larger number of IL-17 secreting cells from central memory T cells derived from healthy controls in comparison to monocytes derived from healthy control subjects. We hypothesize that activated monocytes secreting a combination of IL-1β/IL-6 make a distinct contribution to the induction of pathogenic Th17/Th1 effector cells, which possibly mediate tissue destruction in patients with autoimmune diabetes. We will identify the phenotype of the subpopulation of monocytes secreting IL-1β/IL-6 from T1D subjects and we will determine the mechanism of how these
in vivo activated monocytes drive auto-antigen reactive T cells to an inflammatory effector response. These experiments are extremely relevant as there are ongoing clinical trials using IL-1R antagonists in recent-onset T1D subjects.


PI:Mark W. Feinberg, M.D.
Emerging evidence suggests that in healthy individuals, circulating endothelial progenitor cells (EPCs) represent a population of bone marrow-derived stem and progenitor cells responsible for repairing damaged vascular endothelium and initiating neovasculogenesis. In type 2 diabetes, the enhanced production of proinflammatory cytokines, advanced glycation endproducts, and specific signaling pathways (e.g. protein kinase C, PKC) alter EPC function. Indeed, reduced levels of circulating EPCs and diminished EPC function have been reported and correlated with disease severity. Consequently, diabetic patients exhibit an accelerated and aggressive form of micro-and macro-vascular arteriopathy. KLF10 is a member of the Kruppel-like family of zinc-finger transcription factors that we have found robustly expressed ~60-fold in bone marrow-derived and circulating EPCs in comparison to hematopoietic cells lacking EPC markers. KLF10 expression in EPCs is further induced with the anti-inflammatory growth factor, TGF-β1, whereas it is inhibited upon treatment with pro-inflammatory cytokines, such as TNF-α, PKC activation, or S100b treatment. Our preliminary observations suggest potent defects in EPC differentiation from KLF10–/– bone marrow-derived stem cell progenitors as measured by EPC surface markers (ie. CD34, KDR, CD133), abilityto uptake acLDL/U. lex lectin, and incorporation into tube-like networks. Conversely, overexpression of KLF10 in bone marrow progenitors promoted EPC differentiation, KDR expression, and matrigel angiogenesis assays. On the basis of these observations we hypothesize that (1) KLF10 is a key regulator of EPC differentiation under physiologic states, (2) reduction in KLF10 expression by pathologic stimuli is a critical event in the development of diabetic vascular disease, and (3) modulation of KLF10 expression can regulate the development and progression of diabetic vasculopathy. In Aim1 of the proposal we will investigate the molecular mechanisms underlying the ability of pro-inflammatory stimuli and PKC activation to inhibit KLF10 expression in EPCs in vitro. In Aim2, we will explore the effect of KLF10 overexpression and deficiency on EPC differentiation and function in response to PKC activation. Finally, in Aim3, we will assess in vivo the effect of KLF10 deficiency on EPC function and vascular arterial injury in diabetic mouse models. The results of these studies are of considerable scientific interest and may serve as the basis for novel therapeutic strategies to modulate EPC differentiation, endothelial regeneration, and vascular repair after injury.


PI:Anna Greka, M.D., Ph.D.
Diabetes is a growing global public health problem, due to the ongoing obesity epidemic. A significant portion of diabetic patients go on to develop kidney disease which is manifest as progressively severe proteinuria, often leading to kidney failure. While the mechanisms of acquired proteinuric kidney disease remain elusive, the most important insights havebeen derived from analyses of hereditary forms of the disease, such as Familial Focal Segmental Glomerulosclerosis (FSGS). Interestingly, recently developed, specialized biochemical and computational techniques have enabled the study of the human “metabolome,” the collection of chemicals produced and processed by the body in health and disease states. This is accomplished by taking advantage of tandem mass spectrometry technology to reliably identify changes in metabolites in human biological samples. This technique has been successfully used to detect changes in metabolic profiles in patients undergoing exercise stress testing or experiencing a myocardial infarction. Little is knownabout the metabolome in the setting of diabetes and kidney disease, and, in fact, the content of the urine metabolome remains largely uncharted. We propose to:
Specific Aim 1: Identify metabolic profiles in urine samples from patients with diabetic nephropathy as compared to diabetics with no proteinuria, and healthy controls, in order to define the urine metabolome and its perturbations in diabetic kidney disease.
Specific Aim 2: Identify changes in plasma metabolites before and after a dialysis session in patients with diabetic nephropathy. This will identify candidate “markers” for diabetic nephropathy in the plasma.
Specific Aim 3: Use the information from profiles identified in Specific Aim 2 to perform an informed, targeted screen of plasma derived from patients with diabetic nephropathy as compared to diabetics with no proteinuria, and healthy

In summary, we will harness the tools recently established in the field of metabolite profiling (or “metabolomics”), to study blood and urine derived from patients with diabetic kidney disease and FSGS, as compared to healthy controls. This pilot study will allow us to chart the urine metabolome, and to discover novel risk markers for diabetic and proteinuric kidney disease.

One of the most difficult to treat and feared complications of diabetes is diabetic kidney disease, which is characterized by the spillage of protein into the urine and the eventual onset of kidney failure. We hope to use a new technology called metabolic profiling or “metabolomics,” to study the blood and urine of patients with diabetic kidney disease, in order to identify new molecules or markers of kidney injury. This may allow us to slow or prevent the emergence of kidney damage through early detection and may guide us to novel therapies.


PI: James F. Markmann, M.D., Ph.D.
Pancreatic islet transplantation holds potential as a safer form of beta cell replacement than whole organ pancreas transplantation. Although the field has experienced dramatic improvements since 2000, two formidable barriers remain that must be overcome for the procedure to gain traction as a viable therapy for Type I diabetes: 1) inefficient engraftment, and 2) poor durability of insulin independence.

In the current proposal, we will explore solutions to these two problems in a single clinical trial, taking advantage of discrete markers of engrafted islet mass and long-term insulin independence as a measure of sustained stable function. Initial studies will be conducted at MGH to establish the success of the isolation process and facility (n=4) as mandated by the IRB before proceeding to a second phase in which we conduct an experimental trial of an agent designed to improve engraftment (n=12).  To improve engraftment, we will target the immediate blood islet inflammatory response in a randomized trial using recombinant Activated Protein C (Xigris). In parallel, we will investigate an immunosuppressive regimen that in preliminary studies appears superior to the Edmonton protocol in sustaining long-term islet graft function. Success in these studies would provide a major advance in establishing islet transplantation as an effective therapy for labile Type I diabetes.

A secondary objective of our proposal is to develop an islet transplant program that will span the local academic institutions. Already our human islet production program has been active as a resource for isolated human islets for basic research purposes within the Harvard system and across the country. We expect to expand this effort to collaborations on clinical islet transplant trials with interested local centers. The goals of the current proposal are tightly aligned with those of the BADERC and should provide the opportunity for varied synergistic collaborations.


PI: Gabriella Orasanu, M.D.
Objective: The heterodimeric retinoid X receptor (RXR)-peroxisome proliferator-activated receptor-gamma (PPAR!) nuclear receptor transcriptional complex helps govern the expression of many key players in adiposity, diabetes, and overall energy balance. RXR, as an obligate partner for many nuclear receptors, including all three PPAR isoforms, is a particularly important but poorly understood player in these pathways. In this regard, the endogenous modulation of RXR activity and the role of RXR in diabetes remain largely obscure. In investigating how the putative endogenous RXR ligand 9 cis retinoic acid (9cRA) is generated, we uncovered that the main enzymes controlling the formation of retinaldehyde (Rald) and Rald conversion to retinoic acid (RA) are present in fat and differentially regulated. Retinol is converted to Rald by alcohol dehydrogenases (ADH) and Rald is catabolized to RA by retinaldehyde dehydrogenases (RALDH1). Surprisingly, we found high fat-fed mice lacking RALDH1 (RALDH1-/-) were completely protected against high fat diet (HFD)-induced obesity and insulin resistance - phenotypes that were directly related to Rald concentrations in fat in vivo. In metabolic cage studies, RALDH1-/- mice had increased body temperatures but similar food/water intake. Although Rald inhibits RXRPPAR! !activity, the mechanism underlying the RALDH1-/- hypermetabolic state remains unclear, as does whether the increased insulin sensitivity seen in RALDH1-/- mice derived from their lower body weight or a direct Rald effect. These issues will be addressed here.

Hypothesis: Rald directly modulates insulin sensitivity.
Aim 1: Investigate Rald’s effects in weight-matched Raldh1-/- versus wild type (WT) control mice. Rationale: Although the profound weight differences seen in RALDH1-/- vs. WT mice on HFD (45% kcal, 12 weeks) may account for differences in insulin sensitivity noted above, preliminary data in younger, weight-matched mice on normal chow diet suggest direct Rald effects on glucose metabolism.
Experimental Plan: Glucose, insulin tolerance, and hyperinsulinemic-euglycemic clamp testing will be done in weight-matched RALDH1-/- vs. WT mice on HFD (45% kcal, 4 weeks). Rald concentrations will be measured as before (MS-HPLC) in plasma and relevant tissues (fat, muscle, liver). Additional characterization (insulinstimulated glucose uptake, plasma markers, expression profiling) and mechanistic analyses will consider Rald-specific effects in distinct tissues (liver, muscle) that occur separately from weight changes.

Hypothesis: The RALDH1-/- metabolic phenotype derives from Rald effects on brown adipose tissue (BAT).
Aim 2: Characterize BAT in vitro and in vivo in weight-matched RALDH1-/- vs. WT mice. Rationale: RALDH1-/- mice have higher energy expenditure and increased uncoupling protein 1 (UCP1) mRNA and protein levels consistent with a BAT phenotype. The role of RXR in BAT is largely unexplored. Experimental Plan: BAT (interscapular, ectopic) and primary brown adipocytes will be isolated from RALDH1-/- and WT mice fed HFD (45% kcal, 4 weeks). BAT will be characterized (mass, micro-array, mitochondrial density, morphology, lipid droplet size by electron microscopy).
Cellular responses (profiling, total mitochondrial oxygen consumption) in primary BAT pre-adipocytes and immortalized BAT cell lines before and after stimulation (cyclic AMP, forskolin, with and without Rald, RA, PPAR! !agonist rosiglitazone) will be studied in the presence or absence of established RXR siRNA.


PI: Pavios Pissios, Ph.D.
Type 2-diabetes is one of the most common endocrine disorders that affects over 20.8 million individuals in the US. Therefore, the discovery of novel and therapeutically relevant factors that regulate glucose homeostasis is urgently needed.

We have analyzed microarray data from livers of mice subjected to different diets (ad libitum chow, caloric restriction, ketogenic diet and high fat diet) to identify potential novel regulators of glucose homeostasis. Nicotinamide N-methyltransferase (NNMT) was one of the most regulated transcripts. NNMT is highly expressed in the liver where it methylates nicotinamide leading to its clearance from the body. We hypothesize that NNMT-dependent modulation of intracellular nicotinamide levels regulates the activity of SIRT1 deacetylase, which is potently inhibited by free nicotinamide. Since Sirt1 regulates PGC1a and Foxo1A, two important factors for the expression of PEPCK and G6pase, liver NNMT may have a central role in the control of hepatic gluconeogenesis.
We have found that adenoviral overexpression of NNMT in primary mouse hepatocytes increased PEPCK, G6pase expression and glucose output. In contrast, adenoviral-mediated NNMT knockdown caused decreased PEPCK, G6pase expression and decreased glucose output. We have also found substantially increased levels o NNMT protein in livers of diabetic ob/ob mice compared to wild-type but not in euglycemic ob/ob mice, suggesting that NNMT may contribute to increased hepatic glucose output in diabetes.

We propose to identify the molecular mechanisms by which NNMT regulates gluconeogenesis by examining effects of NNMT on intracellular nicotinamide levels and activity of Sirt1, PGC1a and Foxo1A. We will also use the adenoviral vectors to assess the in vivo effects of NNMT overexpression and knockdown on gluconeogenesis in mouse models of type 2 diabetes. The proposed experiments will establish the role of NNMT as a new regulator of gluconeogenesis and a potential new candidate gene for the treatment for type 2 diabetes. Successful completion of these aims will provide a basis for an R01 application.

Lay Abstract
Diabetes and its complications are on the rise in the US and worldwide. There is great need to understand the underlying pathways that contribute to the regulation of glucose metabolism. Recently, it has been shown that activation of the Silent Information Regulator 1 (Sirt1) improves glucose homeostasis in rodent models of type 2 diabetes. This proposal investigates the pathways that regulate Sirt1 activity. Successful completion of this work might provide novel targets for therapeutic interventions in type 2 diabetes.


PI: Leileata M. Russo, M.D., Ph.D.
Diabetes can result in numerous changes in organ function and may manifest serious complications such as nephropathy, neuropathy and retinopathy as well as having links to cardiovascular disease and obesity. To date we have limited biomarkers to detect early changes in organ function in diabetes and we have no way of analyzing changes in protein and receptor expression at the transcriptional level without a biopsy.  The grant proposed here will investigate an innovative new tool that will allow us to examine the
transcriptional profile of various organs via a blood sample. This will be carried out through the analysis of the RNA transcriptional profile of small vesicles called exosomes that are released from cells of the body into the circulation and that contain mRNAs from their cell of origin. Exosomes will be isolated from the serum of control and diabetic humans via differential centrifugation and the RNA contained within them analyzed via
deep-sequencing. Furthermore, we will also use deep-sequencing to determine the RNA profile of exosomes arising from β-cell primary cultures and compare their profile to i) cultured β-cells and ii) β-cells from fresh frozen tissue using laser-capture microdissection. Analysis of exosomal RNA content and changes in their transcriptional profile may have potential diagnostic applications in humans for examining early changes in organ function in the diabetic milieu.
Specific Aim 1: To analyze the transcriptional profile of RNA present in human serum-derived exosomes of control and diabetic patients.
Specific Aim 2: To determine the exosomal transcriptional profile of insulin producing pancreatic β-cells.

Lay Description:
Cells of the body produce and release into the blood small packages that contain within them information about the cell from which they originate. Normally this information is only obtainable by taking an invasive biopsy of the organ. We would like to investigate if by purifying the cell derived packages from the blood and examining their content whether we are able to obtain insight about whether diabetes is affecting organ function and whether this could be used as a potential diagnostic test.


PI: Melissa K Thomas, M.D.
Obesity is a major public health epidemic in the United States, with increasing prevalence and associated comorbidities that confer substantial economic and human costs. Obesity in the context of the metabolic syndrome, in which insulin resistance and increased visceral fat are prominent features, confers risks of diabetes mellitus and cardiovascular disease. As new molecular mechanisms that govern metabolic switches between energy storage and energy expenditure are elucidated, additional opportunities will emerge to identify novel targets for therapeutic intervention for the treatment of obesity. We propose to evaluate the hypothesis that dysregulation of signaling through the coactivator Bridge-1 contributes to obesity, visceral fat accumulation, and metabolic dysfunction. Our preliminary data indicate that disruption of Bridge-1 signaling in mammals and worms leads to increased visceral and intestinal fat. We envision that Bridge-1 functions as an energy sensor and anti-obesity modulator to negatively regulate accumulation of fat stores. This pilot and feasibility application is intended to explore regulatory mechanisms for Bridge-1-dependent fat accumulation in C. elegans, and translate those findings into mammalian model systems to clarify the strategic experimental approaches for future applications for external funding. We will address the following specific aims: aim 1, to identify regulatory mechanisms by which C44B7.1, the homolog of the coactivator Bridge-1, regulates fat stores in C. elegans, and aim 2, to identify regulatory loci through which the coactivator Bridge-1 modulates fat stores in mice.

Relevance of this research to public health: Identifying molecules that regulate fat storage and understanding signals that determine normal or increased amounts of fat may lead to new approaches to treat obesity and diabetes and prevent associated complications.


PI: Quinchun Tong
The hypothalamus is a critical region in the brain responsible for controlling energy homeostasis. In particular, the arcuate nucleus (Arc) of the hypothalamus responds to changes of nutritional status and conveys these signals to other brain areas to control food intake and energy expenditure. However, the current understanding of neurocircuitry in the hypothalamus is incomplete. In addition to POMC and Agrp/NPY neurons, the Arc has majority of other unidentified neurons and the function of those neurons is currently unknown. Furthermore, the function of GABA, as a principle inhibitory neurotransmitter in the brain, in the neurocircuitry in the hypothalamus, is not well understood. Thus, the goal of this proposal is to investigate the role of GABA release from hypothalamic GABAergic neurons in the regulation of energy balance. Using a combination of Cre-LoxP technology and bilateral brain site specific AAV-Cre vector delivery, I will
1) determine the role of GABA release from a subset of GABAergic neurons in the hypothalamus in the regulation of energy balance;
2) determine the role of GABA release from in adult arcuate GABAergic neurons in the regulation of energy balance. 

The proposed study will most likely reveal an important neural circuit in the brain controlling energy homeostasis, identify new groups of hypothalamic neurons important for the maintenance of energy/glucose homeostasis, and shed light on new mechanisms underlying the pathology of obesity and diabetes. Therefore, results from this proposal will ultimate help better design an efficient strategy to cure diabetes and its associated complications


PI: Brian Wilson, M.D., Ph.D.
CD1d-restricted T cells (or “iNKT cells”) have been reported to regulate an extremely diverse set of immunologic responses and diseases. (1)  Dysfunction of these T cells is clearly correlated with the development of autoimmunity, and in particular autoimmune diabetes.  Despite the importance of CD1d-restricted T cells in this disease, how these T cells function to regulate the transition from innate to adaptive immunity and the exact nature of the disease-associated defects remains unclear. In this regard, potential regulatory functions that would be predicted to have significant impact on type 1 diabetes include recently described critical interactions of iNKT cells with dendritic cells and the activation-induced secretion of Th1 and Th2 cytokines. (2) (3) (4) Th2 cytokine secretion by iNKT cells has been associated with protection from autoimmune diabetes in murine models. (5)  Conversely, iNKT cells cloned from patients with type 1 diabetes were found, amongst other defects, to have an extreme Th1 cytokine bias. (6) (7) Moreover, there are several reports that demonstrate that the activation of iNKT cells with the superagonist a-GalCer prevents the development of diabetes in NOD mice. (2) (5) (8) It is now known that a-GalCer can induce the recruitment of endogenous tolerogenic DC to the regional PLN in a process that is dependent on the expansion of CD4+ iNKT cells. (2) (4)  In contrast, DN iNKT cells accelerate disease and this effect is associated with a selective secretion of IL-17 and –22 (unplublished).  Hence, it is proposed that the ratio of immunoregulatory CD4+ to proinflammatory DN iNKT cells affects autoimmune propensity in individuals at risk for type 1 diabetes by regulating the trafficking and effector function of myeloid DC. Strikingly, we have found that by using a combination of iNKT cell activation with a-GalCer and transfer of antigen-loaded DC, >70% of aggressive Cy-induced diabetes can be reversed. This recovery is in part due to suppression of the autoimmune process coupled to the ability of residual islets to replicate.  We have been able to demonstrate that after recovery of glycemic control there is a residual capacity for beta cells to regenerate as assessed by BrdU+, insulin + beta cells in these mice.  Perhaps more importantly, we have taken the genetic fate-mapping model developed by Melton and co-worker’s to congenic status in the NOD mouse. (9) (10)  Currently, we have the rip-CreER, Ngn3-Cre, and Z/AP reporter lines.  These mice faithfully develop spontaneous and Cy induced diabetes and are >99.9% NOD as assessed by the Jackson Laboratories SNP facility. In addition, immunohistochemical studies in Ngn3-Cre/ZAP and tamoxifen treated rip-CreER/ZAP reveal high fidelity tissue expression. Thus, we believe that we are in the unique position of being able to reverse diabetes using specific cellular targets and determine the precursor source and fate of newly replicated beta cells.


BADERC Home DERC home page