Projects

The in vivo function and dysfunction of human α-cells derived from stem cells or in pseudo T1D-islets

Xiaojuan Chen, MD, PhD., Assistant Professor of Surgical Sciences. Vagelos College of Physicians and Surgeons, Columbia University

The beneficial paracrine signaling of α-cells in enhancing glucose-stimulated insulin secretion from β-cells is an important consideration in the field of stem cell (SC)-derived islet replacement therapy for diabetes. While the SC derived β-cells have been thoroughly characterized in vitro and in vivo, limited information is available regarding the generation and the glucagon-secretion characteristics of SC derived αcells. The regulated glucagon secretion by the SC-α-cells remained to be further studied. We propose to characterize the phenotype of human SC-derived α-cells in vivo after transplant, a step known to promote maturation of the SC-derived islet cells, and comparing it to that of primary human islet α-cells. In addition, although pancreatic α-cells persist in T1D islets, they are often incapable of mounting an appropriate glucagon response in the event of hypoglycemia induced by insulin injection. Recent studies also suggest that α-cell dysfunction contributes to chronic hyperglycemia in patients with T1D due to hyperglucagonemia. For a better glycemic control of patients with diabetes, especially those with reoccurring hypoglycemia, it is important to understand the potential mechanisms involved in the α-cell pathophysiology in diabetes. We have used combined approaches to study glucagon secretion by human α-cells in a pseudo T1D islet model which contains all islet endocrine cells except β-cells after transplant in immune deficient NSG mice without endogenous glucagon production (GCG-KO). Preliminary results have demonstrated successful subcutaneous engraftment of the pseudo-T1D islets in the NSG-GCG-KO mice. The graft glucagon secretion was dose-dependent and increased in response to insulin-induced hypoglycemia. Such graft α-cell function was compromised after a chronic (3 weeks) hyperglycemia exposure, leading to a situation of hyperglucagonemia in the presence of hyperglycemia and a lack of increased glucagon secretion in response to insulin-induced hypoglycemia. We propose to utilize the NSG GCG-KO mice to 1) characterize the function and gene expression of human pluripotent SC (hPSC)-derived α-cells after transplant and compare it to that of primary islet α-cells transplanted; and 2) perform time course studies of gene/protein expression in α-cell grafts exposed to chronic hyperglycemia. These studies will facilitate the efforts in hPSC-islet-cell generation, establish a method for future studies of hPSC-α-cells in a personalized manner, and provide valuable information regarding the mechanisms involved in the hyperglycemia-induced islet α-cell dysfunction.

Regulation of Akt conformation and function

(Haeusler/Glasgow)
Rebecca Haeusler, Ph.D., Associate Professor of Pathology and Cell Biology Associate Professor of Medical Sciences (in Medicine), Vagelos College of Physicians and Surgeons, Columbia University

Anum Glasgow, Ph.D., Assistant Professor of Biochemistry and Molecular Biophysics, Vagelos College of Physicians and Surgeons, Columbia University

The kinase Akt is a central mediator of insulin signaling. In the inactive state, Akt exhibits a closed conformation whereby the kinase domain is inaccessible (Fig 1A). After insulin signaling, Akt is phosphorylated at two key sites, T308 in the kinase domain and S473 in the disordered C-terminal tail. This causes a conformational change that exposes the kinase domain and activates Akt. Using mice and primary murine hepatocytes, we have found unexpectedly that the essential trace element manganese (Mn) potently activates Akt activity without inducing Akt phosphorylation. Insulin causes an additive

effect. Moreover, we have found that hepatocyte Mn efflux is regulated by fasting and feeding, suggesting that Mn is a physiologic modulator of insulin-Akt signaling. We have identified putative Mn-coordinating residues in Akt, and in this grant we propose to test the hypothesis that Mn activates Akt by causing a conformational change to a “primed-active” state that exposes the kinase domain and allows phosphorylation of Akt targets (Fig 1B). We will pursue this goal in two aims: Aim 1-Determine effects of Mn on Akt conformation. We will use hydrogen/deuterium exchange mass spectrometry, isothermal titration calorimetry, inductively coupled plasma mass spectrometry, and intramolecular fluorescence resonance energy transfer of Akt itself to determine the effects of Mn and the putative Mn-binding residues on Akt conformation. Aim 2-Determine effects of Mn-binding on Akt function. We will use fluorescence resonance energy transfer and western blot of Akt substrates in Akt-deficient cells where we’ve expressed wildtype versus Mn-binding-site-mutants of Akt. Success of this project would (i) reveal that Mn determines the conformation and activity of Akt, and (ii) suggest the possibility that micronutrient transport is an unexplored, physiologic mechanism to regulate signal transduction and metabolism.

Novel microRNA therapy for Type 2 diabetes

Ke Cheng, PhD, Professor of Biomedical Engineering, Fu Foundation School of Engineering and Applied Science, Columbia University

We have been working on cell and exosome therapies. While our initial focus was on lung injury, we serendipitously found that inhaled stem cell exosomes decreased blood glucose levels and controlled obesity-induced hyperglycemia in high-fat diet-induced T2D mice. Evidence pointed to exosomal cargo miR-23 as the potential active principle in exosome-mediated T2D control. In this proposal, we aim to determine the therapeutic efficacy and safety of inhaled miR-23 therapy in T2D mice. The Aims are: Aim 1. Formulation of lipid nanoparticle (LNP) based miR-23 delivery for in vivo studies; Aim 2. Determine the safety and therapeutic efficacy of LNP-miR-23 in a mouse model of T2D; Aim 3. Explore potential mechanism, i.e. molecular targets of miR-23 in T2D treatment.

Assessing the effects of placental insulin signaling on the offspring

(Leduc/Thaker)
Charles LeDuc, PhD, Assistant Professor of Pediatrics, Vagelos College of Physicians and Surgeons, Columbia University

Vidhu Thaker, MD, Assistant Professor of Pediatrics, Vagelos College of Physicians and Surgeons, Columbia University

Gestational hyperglycemia is associated with offspring obesity, hyperglycemia, insulin resistance and T2D. Circulating maternal insulin is a growth factor with a role in placental development, nutrient transport, and fetal metabolic programming. Maternal circulating insulin co-varies physiologically with circulating glucose and other metabolites. Thus, it is difficult to assess the independent role of maternal gestational insulin on fetal pancreatic and metabolic development. However, because insulin is the preferred treatment for gestational hyperglycemia, it is important to understand the molecular mechanisms by which elevated gestational insulin affects the fetus. In analysis of data from the Hyperglycemia and adverse pregnancy outcome follow up study (HAPO FUS), we identified associations of maternal C-peptide concentrations with future adolescent offspring C-peptide and peripheral insulin sensitivity at 10-14 years of age after adjusting for glucose, body mass index, and pubertal stage. However, it is hard to dissociate the effect of glucose and insulin in human studies. The proposed study utilizes the ability of mouse models to isolate the effect of placental insulin signaling independent of the circulating maternal glucose concentrations during gestation on the offspring. We hypothesize that circulating maternal insulin has an independent role in nutrient transfer of glucose, lipids and amino acids impacting fetal development and future metabolic health. To prove this hypothesis, we propose to modulate placental insulin signaling in two complementary mouse models while maintaining euglycemia and/or normoinsulinemia throughout pregnancy. The placenta-specific Ada-cre bred with the Insr-flox mice will produce animals with reduced placental insulin signaling while the same mice bred with Insr flox-stop-flox will have increased placental insulin receptor activity. Specifically, we will demonstrate that the two mouse models alter fetal physiology, and we will visualize and quantify islets and islet innervation. The findings of such a study may have profound impact on the management of diabetes during pregnancy.

Investigating the roles of beta-cell Golgi stress in Type 1 Diabetes

Jody Ye, Ph.D., AECOM

Evidence indicates that islet autoimmunity in Type 1 diabetes arises from the failure of beta cells to withstand repeated episodes of stress signals, resulting in the loss of beta cell specific identify, function, and increased immunogenicity. The Golgi apparatus is an important site for protein post-translational modification, transport, and trafficking. The PI found that in human islets, suppression of the T1D susceptibility gene PGM1 triggers Golgi stress, highlighting its potential importance in mediating the development of T1D. However, unlike Endoplasmic Reticulum stress, the role of Golgi stress in beta cells has never been studied in detail. The hypothesis of this proposal is that Golgi stress impairs proinsulin processing, facilitates enteroviral replication, and that blocking GBF1/ARF prevents enteroviral infection in beta cells.

Role of Hyperinsulinemia in NAFLD: Pancreatic Clamp Pilot & Feasibility Study

Joshua R. Cook, M.D., Ph.D. Columbia University

Metabolic dysfunction-associated fatty liver disease (MAFLD) is an under-appreciated complication of lipid dysmetabolism in the insulin resistance (IR) that underlies type 2 diabetes (T2DM). We propose that the hyperinsulinemia that accompanies IR drives the excessive hepatic de novo lipogenesis (DNL) that characterizes MAFLD. Our objective is therefore to observe the impact of lowering insulin levels on DNL in patients with IR (prediabetic state plus hyperinsulinemia) by using the somatostatin analogue- (octreotide-) assisted "pancreatic clamp" technique. To optimize clamp conditions, we must first perform a pilot & feasibility study in which we assess maintaining hyperinsulinemia (MH protocol) at roughly endogenous levels versus a stepwise decrease in insulin levels by 10%, 25%, and 40% ("reduction toward euinsulinemia", RE protocol). All participants will be tested with both MH and RE protocols, in random order, separated by 2-4 weeks. We will evaluate changes in levels of glucose, insulin, and various lipid-metabolic parameters in order to gauge the insulin dose-glycemic response of insulin lowering and select the appropriate insulin infusion rate to apply to the RE protocol in the main study.

Investigate the function of C2cd4a in metabolism
Submitter: Diana Kuo, PhD Assistant Professor, UC Davis

Type 2 diabetes (T2D) is caused by insulin resistance in peripheral tissues and pancreatic beta-cell dysfunction. Insulin resistance precedes beta-cell failure, and the beta-cell’s inability to keep up with the increased demand of insulin production and secretion leads to glucose intolerance and hyperglycemia. The human C2CD4B-C2CD4A-VPS13C locus harbors a pancreatic beta-cell super-enhancer and is heavily decorated by T2D risk-associated GWAS SNPs from virtually every ethnic group studied to date. There are only ~20 publications on “C2cd4a” in PubMed, the majority of which are association studies linking this locus to human diabetes susceptibility. Through a multi-omics approach followed by functional analysis in mice, we found that beta cell-specific C2cd4a ablation impairs insulin secretion. In this proposal, we will delve into the mechanism of C2cd4a-regulated beta cell function, and build a pathway centered on C2cd4a. Two aims are envisioned: in Aim 1, we will investigate the mechanism of exercise-induced hypoglycemia in beta cell-specific C2cd4a knockout mice and delve into metabolic flux in beta cells lacking C2cd4a. In Aim 2, we will identify transcription factor binding partners of C2cd4a in gene regulation to define the mechanistic action of C2cd4a in the nucleus. These aims will advance our understanding of C2cd4a as a human diabetes susceptibility gene and provide a blueprint to leverage human genetics data into biological insight that will eventually benefit patients.

Molecular Mechanisms of Diabetic Neuropathy
Submitter: Amy Rumora, PhD, Assistant Professor of Neurology

This project proposes to examine molecular mechanisms that contribute to peripheral neuropathy in diet-induced obesity. I previously showed that dietary saturated fatty acids (SFAs) and monounsaturated fatty acids (MUFAs) differentially regulate nerve function in murine models of diet-induced obesity and prediabetes. Murine models fed a high-fat diet rich in SFAs develop peripheral neuropathy. Switching these mice from an SFA-rich high-fat diet to a MUFA-rich high-fat diet restores nerve function. We recently found that these changes in nerve function correlate with alterations sphingolipid levels within the sciatic nerve. Herein, we propose to evaluate the impact of SFAs and MUFAs on neuroinflammation and mitophagy in the peripheral nervous system of mice with prediabetes and neuropathy. We hypothesize that a high-fat diet rich in SFAs will induce neuroinflammation and impair mitophagy while a high-fat diet rich in MUFAs will reduce neuroinflammation, restore mitophagy, and nerve function in mice with prediabetes.

Treatment of diabetic foot ulcers via Live yeast secreting wound-healing factors
Submitter: Virginia Cornish, PhD, Helen Rubinstein Professor of Chemistry, Department of Chemistry

The objective of this study is to develop and assess engineered live yeast secreting wound-healing protein factors as a new topical treatment modality for Diabetic Foot Ulcers (DFUs) and other topical wounds. Diabetic foot ulcers (DFUs)—one of the most severe and common

diabetic complications—are a significant financial burden for patients and the healthcare system and a major cause of patient discomfort and surgical amputations. There is an ongoing need for more effective treatments for the DFUs. The technology we present here will be more effective—while cheap and affordable—for the treatment of DFUs and other chronic wounds. We hypothesize that (i) the generally recognized as safe (GRAS) yeast strains can be engineered to secrete wound-healing protein factors in a bioactive form and high enough titers needed for therapeutic applications. (ii) the live engineered yeast can be formulated for topical application on ulcers on the wound bed. (iii) the live engineered yeast will actively secrete and deliver wound-healing factors on the wound site. (iv) the continuous and prolonged secretion and delivery of the wound-healing factors by yeast on the wound site increases the bioavailability of bioactive factors, thus will facilitate the healing process. This project involves (i) optimizing the yeast secretory pathway for the secretion of pharmaceutical proteins in high titers. (ii) engineering optimized yeast for secretion of wound-healing factors of interest. (iii) assessing the secretion levels and the functionality and bioactivity of the secreted protein factors. (iv) formulating a minimalistic hydrogel dressing that contains the engineered live yeast as the active component. (v) assessing the physical/biological properties of the hydrogel formulation and performing optimizations if needed. And (vi) testing the wound healing properties of the engineered live yeast (hydrogel dressing) on an animal model (diabetic mice).

Submitter: Yousin Suh, PhD, Professor of Reproductive Sciences

The vast majority of genetic risk variants associated with T2D GWAS is in non-coding regions, suggesting that they impose risk by altering regulatory elements that control gene expression. Interpretation of identified variants from these extensive studies is currently limited because the clear identity of causative variants in regulatory elements and their regulation of the coding target genes remain unknown. To address these challenges, we are systematically identifying causal enhancer T2D variants and their dysregulated target genes. The human genome is estimated to encode ~1 million enhancer elements, with distinct sets of approximately 40,000–70,000 enhancers being active in a particular cell type. By defining the role of a specific enhancer implicated by a T2D risk association, it is possible to infer the identities of regulatory factors and environmental signals the cell is receiving that might impact T2D. The objective of this work is to identify functional enhancers altered by regulatory T2D variants and uncover their causal impact by high-throughput molecular assays in human iPS-derived beta cells. To this end, we will identify a set of candidate causal variants in enhancers that contribute to the genetic risk of T2D by computational methods that integrate a large number of diverse genomic and epigenomic annotations. We will then perform ultra-high resolution chromatin conformation capture (Tri-HiC) of iPS-derived beta cells to identify the enhancers’ target genes and perform single cell RNA-seq analysis to establish links between the causal enhancer variants and transcriptional outputs. The proposed study will identify enhancer-promoter interaction networks, elucidating whether and to what extent T2D-associated variants alter enhancer function, thereby playing a functional role in T2D genetic risk.

Submitter: Eunhee Choi, PhD, Assistant Professor, Pathology and Cell Biology

Alterations of insulin signaling have been linked to diabetes. Insulin is a potent mitogen, and hyperinsulinemia can promote tumor growth. Thus, understanding insulin action in tumor progression is of paramount importance. The spindle checkpoint ensures the fidelity of chromosome segregation. The key checkpoint protein MAD2 forms a mitotic checkpoint complex (MCC) that arrests cells in mitosis. We discovered a non-canonical role of spindle checkpoint proteins in insulin signaling. MAD2 binds to the insulin receptor (IR), recruits the clathrin adaptor complex by assembling an MCC-like complex, and promotes IR endocytosis. Mice in which MAD2 binding to IR is prevented by point mutations (IR4A) display increased T-cell aneuploidy and T-cell lymphoma. These results have opened up new areas of research on how nutrient signaling affects chromosome stability. We will pursue this observation with 2 aims: in Aim 1 we will determine the function and mechanism of IR in chromosome stability by testing whether IR4A/4A mice develop aneuploidy in vivo using single-cell sequencing and karyotyping.

In Aim 2, we will establish the role of the IR-MAD2 interaction in metabolism. These experiments will advance our understanding of the function, regulation, and mechanism(s) of insulin action in cell proliferation.

Submitter: Gabrielle Page-Wilson, MD, Assistant Professor

Gestational diabetes mellitus (GDM) is now the most common medical disorder of pregnancy. It is associated with adverse outcomes during gestation, and can lead to long-term health complications including Type 2 Diabetes in mothers and glucose intolerance, obesity, and metabolic syndrome in offspring. Targeted interventions to prevent the onset of GDM may mitigate adverse metabolic consequences. Elucidating the central drivers of glucose metabolism in pregnancy will facilitate the identification of effective biomarkers for predicting GDM and may give rise to mechanism based interventions. The central melanocortin system is a key regulator of appetite and metabolism. It is comprised of POMC and AgRP neurons and their receptors. We have recently established CSF measurements of POMC and AgRP as reliable markers of brain melanocortin activity in humans. Using these markers, we have demonstrated that a state of gestational leptin resistance characterized by a decrease in leptin transport into the brain and resistance to the suppressive effects of leptin on AgRP. Our preliminary data also show that plasma AgRP – a peripheral biomarker of the hypothalamic neuropeptide AgRP, is markedly elevated in pregnant women. The goal of this project is to define pregnancy-specific adaptations in leptin, glucocorticoid, and melanocortin physiology in women with gestational diabetes and to determine whether peripheral concentrations of AgRP can be used to predict GDM early in gestation.

Submitter: Magdalena Bogun, MD, Assistant Professor of Medicine

Continuous glucose monitoring (CGM) is widely used in the outpatient setting but its use is still investigational in the inpatient setting. Several studies showed improvement in detection of hyperglycemic and hypoglycemic episodes with the use of CGM when compared to fingerstick point of care (POC) glucose testing in both the ICU and non-ICU setting. The FDA has allowed use of DEXCOM G6 CGM in hospitalized patients during the COVID-19 pandemic. In a recent pilot study performed during the COVID-19 pandemic done in hospitalized patients in the perioperative period, DEXCOM G6 CGM showed a mean absolute relative difference (MARD) of 9.4% compared to point of care (POC) glucose measurements. Use of continuous glucose monitoring has not been reported in hospitalized patients with DKA. In this project the PI plans to evaluate outcomes of DKA patients treated with CGM in the stepdown unit compared to treatment of DKA in the ICU using hourly POC glucose testing. She also intends to study providers and nurses’ satisfaction with remote glucose monitoring. Furthermore, she will study potential cost savings when DKA patients have glucose levels monitored via DEXCOM G6 CGM compared to hourly fingerstick glucose testing in the ICU setting.

Henry M. Colecraft, PhD, Professor of Physiology and Cell Biology

Ca2+ influx through voltage-gated calcium channels (VGCCs) is necessary for glucose-stimulated insulin secretion in pancreatic β-cells (PβCs). VGCCs are thus a potential locus for both PβC-dependent pathophysiology and therapy. CaVβ subunits (β1-β4) are powerful regulators of VGCCs by controlling α1 subunit trafficking and tuning channel gating. A central unresolved question is: how does Ca2+ influx via VGCCs give rise to divergent functions in PβCs such as insulin secretion and excitation-transcription coupling? The PI hypothesized that in PβCs distinct CaVβs are instrumental in organizing VGCCs into discrete macromolecular complexes with specialized functions. A significant barrier to rigorously assess the functional roles of CaVβ molecular diversity in excitable cells, including PβCs, is the inability to inhibit VGCCs based on the identity of their resident CaVβ. The PI will develop novel genetically-encoded CaV channel blockers that enable inhibition of CaVβ-specific VGCC complexes, and apply them to decipher signaling functions of CaVβs in PβCs. The approach exploits a bioengineering method to generate genetically-encoded VGCC inhibitors by pairing the specificity of single-domain antibodies (or nanobodies) with the consequential catalytic activity of an E3 ubiquitin ligase. Using Cavβ isoform-selective nanobodies with molecular biology, electrophysiology, flow cytometry, fluorescence resonance energy transfer (FRET), ion channel engineering, and biochemistry, the PI will develop and engineer nanobodies to selectively inhibit VGCCs on the basis of their resident CaVβ, and elucidate the functions of CaVβ specific VGCC complexes in pancreatic β-cells.

Joshua Cook, MD, PhD, Instructor in Medicine

Bile acids (BA) have emerged as promising and manipulatable mediators of the dual threats posed by insulin resistance (IR) underlying type 2 diabetes: chronic hyperglycemia and cardiovascular disease. Levels of 12ahydroxylated bile acids (12-HBA) strongly correlate with measures of IR in humans and depletion of 12-HBA in animals improves glucose tolerance. However, hepatic IR also appears to cause increased 12-HBA in animals. Better understanding the mechanisms behind this apparent pathologic cycle in humans is essential
for developing novel cardioprotective diabetes treatments. We propose short-circuiting this cycle with two complementary patient-oriented research studies of growth hormone (GH) excess, a secondary cause of insulin resistance featuring both clear pathogenesis and reversibility, and its consequences for BA metabolism. First, we will prospectively study patients with acromegaly (“GH-oma”), measuring IR indices and BA profiles before and after surgical and/or medical treatment. Second, we will conduct a randomized, placebo-controlled prospective clinical trial in which we administer recombinant human GH (rhGH) to healthy volunteers in order to evaluate the impact of temporarily induced IR on BA metabolism – especially 12-HBA content. We will also assess the ability of the BA sequestrant colesevelam to mitigate rhGH-induced IR by offsetting adverse BA reprogramming. In each case we will perform our BA analysis using advanced mass spectrometric techniques capable of identifying >150 individual BA species, most as yet unstudied. These patient-oriented studies will yield new mechanistic insights into the direction(s) of causality in the relationship between IR and BA, particularly 12-HBA, with major implications for drug design.

Andrea Califano, PhD, Professor of Systems Biology

The aim of this proposal is to identify driver networks of pancreatic β-cell failure in humans, and perform an unbiased drug screen to test whether it is reversible by pharmacological treatment. In preliminary data, using a combinatorial approach of single-cell RNA sequencing (scRNA-Seq) with human T2D islets and computational regulatory network analysis, the PI identified potential master regulatory networks associated with diabetes-unique b-cell sub-populations. Using this information, the PI proposes to functionally test the effects of candidate master regulators using

a candidate gene perturbation screen in primary islets. This information will be used to carry out an unbiased screen of a limited subset of FDA-approved drugs in primary islets, using an integrated high-throughput screening (HTS)/RNA-seq platform (PlateSeq) for the high-multiplex, low-cost RNA-seq profiling of compound perturbations in cellular systems. The findings will increase our understanding of cellular heterogeneity in human T2D islets, determine driver networks that control cell fate transition, and begin to define actionable therapeutic pathways to intervene on the disease process.

Lina Sui, PhD, Assistant Professor of Pediatrics

The human genome contains a large number of genetic variants associated with disease risk, including for diabetes. However, for most variants, the functional consequence has not been demonstrated in human cells, and functional assays to evaluate the consequences in animal models are limited to conserved regions. The PI has collected a set of novel variants predicted to alter the function of proteins implicated in familial diabetes. She proposes to study variants in two genes TRIB3 and SEL1L for detailed characterization, because of their likely role in endoplasmic reticulum (ER) function. The ER is critical for proinsulin folding and insulin secretion from beta cells. There is increasing evidence that the presence of unfolded proteins is of central importance to the failure of beta cells in all forms of diabetes, including type 1 and type 2 diabetes. The PI will introduce mutations identified in patients into human pluripotent stem cells, and test their developmental and functional properties of iPS-derived beta cells in vitro as well as after grafting into mice in vivo. She has established an efficient experimental system for beta cell differentiation and grafting without the formation of teratomas that allows the functional study of human genotypes in a physiologically relevant environment. The proposed studies have the potential to identify two novel diabetes genes, and enable the study of mechanisms by which protein folding stress causes beta cells to fail. The proposed work is therefore relevant to all forms of diabetes.

Gordana Vunjak-Novakovic, PhD Professor of Biomedical Engineering

Goal of this project is to develop small 3D microphysiological systems devices that surpass currently available systems in physiological relevance using compartmentalized beta cells and circulating immune cells separated by a layer of endothelial cells. This system will be used to define the conditions that determine the recruitment of immune cells that mediate the sequential phases of autoimmune assault on beta cells. The PI’s hypothesis is that the earliest signals leading to the recruitment of immune cells originate form genome instability resulting from replication-transcription conflicts at highly transcribed genes, such as insulin, in highly proliferating cells. The PI proposes that activation of the cGAS-STING and TLR9 pathways by damaged DNA results in the production of chemokynes. Streamlined production of cellular components into miniature and personalized microphysiological devices will provide a powerful and affordable precision medicine platform to evaluate drugs and perform diagnostic testing.

Long non-coding RNAs in beta cell function 

Luis Arnes, PhD, Assistant Professor of Pathology and Cell Biology

Considerable effort has been directed towards deciphering the molecular mechanisms that regulate β-cell formation, maturation and function. Past studies of pancreas development and β-cell function have largely focused on coding genes. In recent years, integrated analyses of islet-specific transcription factor binding sites, epigenetic modifications and gene expression profiles have uncovered the existence of chromatin domains and long non-coding RNAs (lncRNAs) with largely unknown functions. Notably, many of them are located in the genome nearby islet-specific genes, conserved across species and dynamically regulated by glucose concentration and throughout pancreas development. Although these features suggest a regulatory role in β-cell specification and function, few of them have been examined. The PI hypothesized that lncRNAs affect gene regulation in pancreas development and β-cell function. To test this hypothesis, he proposed to define the transcriptome profile of pancreatic lineages at single cell resolution using in vitro differentiation of human embryonic stem cells. Moreover, he will study the regulation and function of positionally-conserved lncRNAs in the specification of insulin-producing β-cells. The results from these studies will help generate alternative sources for β-cell replacement therapies.

Role of TTC39B in Mediating NAFLD in Insulin-Resistant States

Joanne Hsieh, PhD, Assistant Professor of Medicine

Non-alcoholic fatty liver disease (NAFLD) is very common in diabetes and insulin resistance and one feature is increased de novo lipogenesis (DNL). However, the mechanism for upregulated hepatic DNL is not well-understood. Deficiency in the scaffolding protein TTC39B results in protection from NAFLD in mice, associated with decreased hepatic lipogenic gene expression. TTC39B regulates hepatic lipogenic gene expression by interacting with and promoting the degradation of RB1. TTC39B also interacts with 14-3-3γ, which may sequester from the nucleus. Based on these observations, the PI hypothesizes that insulin resistance activates a phosphatase that releases TTC39B from 14-3-3γ binding, allowing TTC39B to enter the nucleus to degrade RB1, resulting in lipogenic gene expression. Aim 1 will use hepatic RB1-deficient mice to determine its contribution to hepatic lipogenic gene expression and NAFLD pathogenesis in vivo. RNA-Sequencing and ATAC-Sequencing analysis will be used to identify promoter and enhancer motifs that are affected in the TTC39B/RB1-dependent pathway. Aim 2 will use unbiased approaches to identify how TTC39B’s interaction with kinases and phosphatases change as insulin resistance and NAFLD develops. This study has the potential to identify a novel pathway that explains the upregulated DNL in insulin resistant individuals, and inform the development of new therapies to treat

Transcriptional regulation of pancreatic beta cell failure by single cell RNA-Seq
Submitter: James C. Lo, MD, PhD, Assistant Professor of Medicine, Weill-Cornell Medical College

This project examines the gradual decline of beta cell function in diabetes. There is evidence that transcriptional dysregulation of beta cells leads to beta cell failure. Pancreatic islets are composed of a heterogeneous group of specialized endocrine cells. Hence, traditional strategies of RNA profiling on bulk islet cells have likely not captured the complexity of individual pancreatic islet cells in normal physiology and especially the stresses of T2DM. The PI will assess single cell transcriptomes of ~10,000 pancreatic islet cells from healthy control and diabetic mice. He will use these data to map transcriptional networks perturbed in T2DM with an eye on identifying pathways disrupted in beta cell failure.

The Role of ER-Mitochondria Contacts in the AKT Signaling Pathway
Submitter: Estela Area-Gomez, PhD, Assistant Professor of Neurology

This project examines the link between insulin resistance and AKT-dependent mitochondrial signaling. AKT can translocate to the outer mitochondrial membrane (OMM), where it activates hexokinase-II (HK-II), the rate-limiting step of glycolysis, promotes its binding to the mitochondrial voltage-dependent anion channel (VDAC). Thus, localization of AKT to OMM regions bridges ER and mitochondria to regulate glycolysis and calcium transfer between ER and mitochondria. Areas of the ER in close apposition to mitochondria are referred to as mitochondria-associated ER membranes, or MAMs. MAMs are functional domains that regulate lipid metabolism and have been shown to modulate hepatic insulin action and AKT activation. This project will examine how targeting of AKT and its kinase mTORC2 to MAMs regulates hepatic insulin sensitivity, HK-II function, and calcium fluxes.

Hepatic Dach1 regulates lipid metabolism
Submitter: Lale Ozcan, MD, Assistant Professor of Medical Sciences, Department of Medicine

Individuals with type 2 diabetes have a 2–4-fold increase in their lifetime risk of cardiovascular diseases, arising from lipid abnormalities. The PI has identified a pathway activated by calcium/calmodulin-dependent protein kinase II (CaMKII) that inhibits insulin signaling. She went on to characterize the corepressor Dachshund homolog 1 (Dach1) as a key effector of the calmodulin-kinase pathway in liver. Hepatic Dach1 levels increase in obese mice and humans, and Dach1 inhibition protects against hyperinsulinemia and hyperglycemia. In addition, hepatocyte-specific Dach1 deletion leads to a decrease of plasma cholesterol levels. In this PF grant, the PI is investigating whether Dach1-mediated co-repression links insulin resistance with abnormal lipid and cholesterol metabolism.

The diabetic environment perpetuates Staphylococcus aureus infection
Submitter: Dane Parker, PhD, Assistant Professor of Microbial Pathogenesis [in Pediatrics]

The hypothesis of this study is that hyperglycemia supports S. aureus infection. To address this question, the PI established a diabetic mouse model of subcutaneous skin infection and will investigate whether diabetes affects (i) clearance, (ii) pathogenesis, and (iii) immune response to S. aureus skin infection. He has shown that diabetic mice have increased dermal necrosis and delayed resolution of infection in response to S. aureus infection. This work establishes a model to investigate S. aureus skin infection in the context of diabetes. Using this model, the PI will determine the effects of glucose on immune cell function and test the hypothesis that hyperglycemia inhibits the phagocytic and killing function of macrophages and neutrophils as well as their ability to properly invoke an inflammatory response. By using cells exposed to varying levels of glucose as well as cells isolated from control and diabetic mice he will investigate whether phagocytes have diminished bacterial killing capacity and reduced ability to induce inflammatory cytokines.

PPARγ Accumulation and Metabolic Decline in Aging
Submitter: Li Qiang, PhD, Assistant Professor of Pathology and Cell Biology

The association of aging with the metabolic syndrome remains unexplained. The PI proposes to investigate a novel mechanism linking insulin resistance with aging, with the aim to identify novel and safer interventions for age-associated metabolic diseases. PPARγ is a master regulator of adipocyte formation and function, and plays a pivotal role in metabolism and insulin action, but its role in aging is unknown. The PI demonstrated a remarkable increase of PPARγ in tissues from aging individuals and aging experimental animals, along with stark differences in the response to the PPARγ ligand thiazolidinedione (TZD) in young vs. aging animals. He hypothesized that the accumulation of PPARγ contributes to the metabolic decline during aging. He proposed to establish the physiological significance of PPARγ accumulation in aging and elucidate its molecular basis. The ultimate goal of this work is to identify new mechanisms connecting aging and metabolic syndrome and provide novel therapeutic candidates for delaying the onset of metabolic diseases in aging.

Redox regulation of diabetic vascular remodeling by s-glutathiolation
Submitter: Ziad Ali, MD, DPhil, Assistant Professor of Medicine Cardiology

Among the many potential catalysts for the pathophysiological manifestations of diabetes, overproduction of reactive oxygen species (ROS) appears to be a common upstream event. Protein S-glutathiolation, the reversible covalent addition of glutathione to cysteine residues on target proteins, is a candidate mechanism by which changes in intracellular redox state and generation of ROS can modulate protein function. The PI hypothesized that the increased ROS generation in type 2 diabetes promotes protein S-glutathiolation, in turn leading to neointimal hyperplasia, accelerated atherosclerosis, and revascularization failure. He identified two highly glutathiolated proteins, annexin A2 and Beta-actin. These proteins are involved in cellular processes critical to vascular remodeling, such as proliferation and migration. Moreover, recent data suggest that these proteins interact dynamically with one another. Thus, the PI is studying the physiological and cellular effects of S-glutathiolation in type 2 diabetes; identifying the pathways involved in S-glutathiolation-mediated phenotypes; and 3) establishing the role of S-glutathiolation as a therapeutic target to prevent diabetes-induced premature revascularization failure.

Differential Effects of Bile Acid Species in Intestine
Submitter: Rebecca A. Haeusler, PhD, Assistant Professor of Pathology and Cell Biology

Bile acids (BA) are produced in the liver from cholesterol to regulate several biological processes. A key site of BA signaling is the intestine, a tissue that secretes a multitude of peptide hormones that act on the pancreas, brain, and liver. At least two gut hormones are secreted in response to BAs: glucagon-like peptide-1 (Glp1), and fibroblast growth factor-19 (Fgf19). It’s unknown whether different BA species affect secretion of these hormones differently, or whether there are additional intracellular or secreted proteins targeted by BAs in intestine. The PI first demonstrated that different BA species have distinct effects on insulin sensitivity and lipid metabolism. In this work, she hypothesized that alterations in BA levels and composition affect gene expression and secretion of proteins in the intestine. She tested this hypothesis by determining the effects of BA levels and composition on intestinal gene expression using intestinal explants, oral gavage, and intestinal perfusion in mice, followed by tissue collection and transcriptional profiling. Moreover, she investigated the effects of BA levels and composition on the gut secretome using intestinal perfusion in mice, followed by mesenteric blood collection and proteomic analysis.

Deciphering and Visualizing Epitope Spreading in Autoimmune Diabetes
Submitter: Remi J. Creusot, PhD, Assistant Professor of Medical Sciences (in Medicine)

Type 1 diabetes results from T cell-mediated immune response against multiple Beta-cell antigens. Although few self-antigens appear to be involved initially, the autoimmune response spreads to other epitopes. Epitope spreading is a common but unexplained mechanism of autoimmune diseases. Mechanisms regulating T cell clonal selection and responsivity to tissue damade are unclear. The PI investigated whether successfully activated T cell clones can boost the immunogenic function of dendritic cells to stimulate low-affinity T cells in the context of “antigen linkage”. Using a novel polyclonal adoptive transfer model with traceable diabetogenic T cell clones, he tested whether linked cooperation can support low-affinity T cells activation. He evaluated how these diabetogenic T cells, with varying degrees of antigen responsiveness, are stimulated in vivo when transferred alone or as a mixed population and allowed to influence one another within DC clusters. The functional analysis of the activation of these different T cells will help design therapeutic strategies aimed at reversing this phenomenon using linked suppression by regulatory T cells.

The PI has published a paper using data from this PF grant. He has obtained 3 new grants, and is a co-investigator on another one. Data obtained through the PF grant were instrumental to the successful competition for two R21 grants. When asked about the role of the PF grant in his budding career, he wrote that: “The PF grant was instrumental to develop an in vivo system to co-express multiple beta cell antigens by antigen-presenting cells (APCs) and analyze the response of multiple antigen-specific T cell clones. We generated constructs producing up to 8 epitopes that can be recognized by T cell receptor-transgenic T cells. As mRNA, we used these constructs to study how signals delivered by APCs upon antigen engagement can cooperate to induce regulatory T cells. This led directly to an R21 grant. As DNA vaccines, we have delivered these constructs by various routes and under different formulations to improve in vivo delivery and tolerogenic presentation. This help has been invaluable in my career development.”

The AF10/DOT1 Complex, O-GlcNAc, and Transcription: From Nematodes to Mammals
Submitter: Alla Grishok, PhD, Assistant Professor of Biochemistry and Molecular Biophysics

Excess flux through the hexosamine biosynthesis pathway increases protein O-linked GlcNAcylation and contributes to insulin resistance. This reaction is catalyzed by O-GlcNAc transferase (OGT) and reversed by O-GlcNAcase (OGA). Mutations of OGA have been associated with type 2 diabetes. The PI showed that the C.elegans chromatin-binding protein ZFP-1 (AF10 in mammals) acts to reduce transcription of metabolic genes. O-GlcNAc modification on chromatin at active gene promoters has been implicated in transcriptional activation. In C. elegans, ZFP-1 co-localizes with and inhibits O-GlcNAc. The PI investigated whether O-GlcNAc modification at active promoters mediates glucose-responsive transcription and whether ZFP-1 and its major interacting partner DOT-1.1 provide feedback inhibition of O-GlcNAc. She evaluated levels of O-GlcNAc modification on chromatin in zfp-1 mutant by genome-wide ChIP-seq and the effect of O-GlcNAc on gene expression by RNA-seq analyses of zfp-1; ogt-1 and zfp-1; oga-1 double mutant worms. She also tested chromatin localization of AF10 and O-GlcNAc in 3T3-L1 murine adipocytes by ChIP-seq.

The PI has two papers in preparation describing the findings. She has applied for two NIH grants (1DP1OD023968-01 and 1R01GM124175-01) that did not receive a fundable score on first submission but have been resubmitted.

Stem cell-based model for HNF1A deficiency in human pancreatic beta cells
Submitter: Lina Sui, PhD, Associate Research Scientist, Department of Pediatrics

Stem cells can reflect disease-relevant phenotypes after in vitro differentiation and transplantation into mice. Though monogenetic forms of diabetes are rare (1-5% of all diabetics), they provide a model to test the utility of stem cells as an approach to Beta-cell replacement, and to better understand the mechanisms of Beta-cell dysfunction, as the same or related genes are likely to contribute to the risk of developing Beta-cell failure. Insight into the roles of MODY-causing transcription factors in the formation, survival, and function of beta cells can advance strategies for the prevention and mitigation of more prevalent forms of diabetes. Unlike humans, mice haploinsufficient for HNF1A show no diabetic phenotype, indicating that there are species-specific differences in genotype-phenotype relationship, and pointing to the importance of using human cells for these studies. The PI generated pluripotent stem cells from MODY3 patients and converted them into Beta-cells, then assessed HNF1A function using insulin production, storage, and secretion in vitro and in vivo. She also determined rates of proliferation and cell death and analyzed the transcriptome of MODY3 Beta-cells. These studies have contributed to establishing a platform to improve generation and testing of iPS and ESC-derived Beta-cells.

The PI has not yet published the findings of this PF project nor obtained new grants.

Enteroendocrine Specification in Drosophila and Vertebrates
Submitter: Benjamin Ohlstein, MD, PhD, Associate Professor of Genetics & Development and of Medicine

The intestine is the largest endocrine organ in the human body. Scattered throughout the human gut are enteroendocrine cells that secrete hormones involved in the regulation of various physiological processes such as digestion, intestinal motility, and glucose metabolism. The Drosophila adult intestine, like that of vertebrates, contains enteroendocrine cells that can be distinguished by staining with antibodies to the various hormones that they secrete. By analyzing RNA obtained from wild-type intestines, intestines lacking enteroendocrine cells, and intestines producing enteroendocrine cells in excess, the PI generated a set of candidate genes that regulate enteroendocrine differentiation and function. Remarkably, among these candidates are homologs of important determinants of pancreatic endocrine cell fate: Ngn3, Hes, Pax, and Ptf1. In addition to these homologs, the PI identified additional candidate mediators of enteroendocrine cell differentiations. He is currently involved in functional studies of genes so identified.
The PI has published 2 original papers and one review based on data obtained from this PF grant. The PF grant was also instrumental to the successful award of a new NIDDK R01 grant.

Intelligent glucagon delivery vehicle as adjunct for exogenous insulin therapy
Submitter: Milan Stojanovic, PhD, Associate Professor of Medical Sciences and Biomedical Engineering

This project pursued the development of a novel adjunct to exogenous insulin therapy for the management of hyperglycemia accompanying beta cell insufficiency associated with type 1 and type 2 diabetes. The PI developed nanoscale particles that deliver glucagon or insulin in a glucose- responsive manner and have therapeutically useful in vivo half-lives. Ultimate goal of this proposal is the production of a self-operating molecular machine with medical applications. The approach draws inspiration from the PI’s previous work in DNA nanotechnology and molecular computing. The hypothesis is that nano-objects can be used to sequester functions and that a structural change can be used to reveal new functions. As proof-of-principle the PI investigated glucose control in response to insulin. He created nano-eggs (neggs) composed of two half-shells that open and close in response to environmental glucose concentrations. Once opened, neggs release glucagon in response to falling glucose levels to help maintain euglycemia. This delivery vehicle, added to current insulin formulations, has the potential of circumventing the problem of hypoglycemia and can result in tighter glycemic control in diabetic patients.

The PI has published 2 papers based on this PF grant. He has obtained 2 new grants, and is a co-investigator on 2 more with other DRC members. Most importantly, this PF grant has triggered an ongoing, extremely successful collaboration with two additional PF recipients, Dr. Paul Harris and Dr. Qiao Lin. With Harris, Stojanovic is pursuing imaging methods for the functional visualization of pancreatic islets. With Dr. Lin, they are pursuing glucose sensors. These interactions were entirely catalyzed by the DRC, and by the program enrichment activities in which all PF recipients participate.