Research in our laboratory focuses on the developmental biology of the pancreas, a vital organ with endocrine and exocrine functions in the vertebrate digestive tract. We seek to understand the mechanisms that govern growth and differentiation of progenitor or stem cells that generate the different cell types comprising the pancreas. Some of the pathways active during embryogenesis also regulate pancreatic growth and function during adulthood, and we are studying the genetics and role of cell interactions in growth control of the mature pancreas. One goal of our work is to translate our studies into novel diagnostic and therapeutic strategies for common pancreatic disease states in humans, particularly diabetes mellitus and pancreatic cancer.

To help us pursue these goals, we are actively collaborating with other Stanford faculty, particularly Drs. Roel Nusse, Gerald Crabtree, Joanna Wysocka and Irving Weissman. In addition, we have regular joint group meetings with laboratories of the UCSF Diabetes Center led by Drs. Matthias Hebrok and Michael German.

1. Genetic dissection of 'islet' cell development and function in Drosophila

Regulation of growth and metabolism by insulin and glucagon-like peptides is a conserved feature in metazoans. Despite this ancient regulatory heritage and a century of knowledge that islet hormones are crucial to human health, our understanding of the genetic basis for pancreatic islet cell differentiation, expansion, and function is surprisingly meager (Kim and Hebrok, 2001; Heit et al 2006a). Recently we discovered endocrine cells that secrete insulin and glucagon-like peptides in Drosophila melanogaster (Figure 1A; Rulifson et al 2002; Kim and Rulifson 2004), an organism superbly suited for discovering genes governing tissue and organ development. Our studies showed remarkable physiological parallels between these Drosophila endocrine cells and pancreatic islet ß- and α-cells, motivating mutant screens to identify conserved regulators of islet development and function. Ablation of insulin-producing cells (IPCs) in flies produced impaired growth and increased circulating glucose levels (Figure 1B-C), phenotypes strikingly reminiscent of impaired growth and diabetes seen in mammals deficient for insulin-producing ß-cells. Glucagon counter-acts insulin activity, and it was known that an insect peptide called adipokinetic hormone (AKH) had glucagon-like functions. We showed that ablation of Drosophila corpora cardiaca (CC) cells that secrete AKH produced systemic hypoglycemia (Kim and Rulifson, 2004), a phenotype seen in mammals deficient for glucagon-secreting α-cells. Moreover, we discovered that CC cells and pancreatic islet cells use identical mechanisms to regulate glucose sensing and hormone secretion. Thus, our work identified Drosophila endocrine cells that are functional orthologs of pancreatic islet α- and ß-cells.

Figure 1. (A) Drosophila insulin-producing cells (IPCs, green) and corpora cardiaca (CC, red) cells viewed by confocal microscopy. (B) Reduced size in flies lacking IPCs. (C) Hyperglycemia in flies after IPC ablation. From Rulifson et al 2002 and Heit et al 2004.

Based on these findings, we are performing genetic screens to identify novel regulators of IPC and CC cell development and function. Our investigations and knowledge of the developmental lineages of IPC and CC cells have accelerated analysis of mutations affecting the differentiation and expansion of IPCs and CC cells. Our expertise in mouse and human endocrine biology (Kim et al 2000; Harmon et al 2004; Smart et al 2006; Sugiyama et al 2006; Rulifson et al 2007) has optimized identification and analysis of mammalian orthologs of Drosophila IPC and CC regulators that control islet cell development and function. Our identification of mutations that promote IPC or ß-cell expansion should therefore accelerate discovery of new signaling pathways and drugs that stimulate generation or expansion of pancreatic islets.

2. Pancreas development

During embryogenesis the pancreas develops from distinct dorsal then ventral evaginations from endoderm (Figure 2, steps A and B ). As yet undefined signals likely “pre-pattern” the endoderm, as indicated by expression of transcription factors (one, called Ipf1 or Pdx1 is shown in Figure 2A) in endoderm destined to become pancreas. The development of the pancreas is then orchestrated by a series of inductive interactions between endoderm and mesoderm-derived tissues, including the notochord, blood vessels, and gut mesoderm (Figure 2A-C). These interactions involve transmission of signals that either permit or direct the differentiation of endoderm to a pancreatic fate. Later, stimulated by other signals (Figure 2C,D), pancreatic epithelial cells proliferate, branch, and differentiate toward one of six major types of cells in the pancreas. These are duct cells, exocrine acinar cells, and the four principal types of islet cell. The accumulated evidence is consistent with the possibility that a unique set of stem/progenitor cells gives rise to all pancreatic cell lineages. Part of the effort in our laboratory is devoted to isolating and characterizing these pancreatic “stem” cells, then elucidating the mechanisms that direct these cells toward specific fates. An understanding of the signals that regulate differentiation of these cells could lead to methods for controlling the expansion or regeneration of cells like islet cells in vivo. They could also lead to an understanding of the molecular changes that provoke diseases like pancreatic cancer.

Figure 2. Diagrams outlining sequential stages of pancreas development in mice. From Kim and MacDonald, 2002.

3. Purifying endoderm and pancreatic progenitor/stem cells

Purification of stem and progenitor cells from an organ is a powerful way to achieve organ replacement. In a celebrated example, successful purification of hematopoietic progenitors and stem cells from bone marrow ultimately produced cell-based therapies for a diverse range of human diseases, and transformed modern medical practice. We have successfully used FACS to purify subsets of multipotent, self-renewing progenitors in the pancreatic islet lineage from mice and humans (Sugiyama et al 2006; Figure 3). Pancreatic progenitor cells provide a natural source for generating new islets. Successful purification of definitive endoderm will also provide a powerful new tool for elucidating genetic and cellular mechanisms controlling development of human organs derived from endoderm like the pancreas and liver, which have otherwise limited experimental accessibility. Since mature islets derive from endoderm and islet progenitors, our FACS strategies also provide unique methods for isolating differentiating human cells in the islet lineage produced from non-pancreatic sources, regardless of origin. Thus, we expect our distinct approaches should accelerate use of embryonic stem (ES) cells to generate endoderm and functional islet cells.

Figure 3. Flow-cytometry-based isolation of NGN3+ endocrine progenitor cells from E15.5 mouse embryonic pancreas. (A) FACS plot of cells sorted following exposure to an antibody that recognizes CD133. Percentages indicated. (B) Subsets of isolated CD133– cells express insulin (white, marked by arrowhead) and glucagon (green, marked by arrows). (C) Subsets of CD133+ cells express NGN3 (red, marked by arrowheads). (D) FACS plot of cells sorted following exposure to an antibody that recognizes CD49f. Percentages indicated. (E) Subsets of isolated CD49f-/low cells express insulin (white, marked by arrowheads) and glucagon (green, marked by arrow). (F) A subset of isolated CD49fhigh cells express CarbA (red). (G) FACS plot of cells sorted following exposure to antibodies recognizing CD133 and CD49f. Fractions labeled I-IV are indicated. (H) CD49flow CD133– cells (fraction III) exclusively contain insulin+ (white, arrowhead) or glucagon+ (green, arrow) cells. (I) CD49flow CD133+ cells (fraction II) exclusively contain NGN3+ cells (red, arrowhead). Adapted from Sugiyama et al 2006.

Our successes with mice led to our discovery of FACS methods for isolating candidate human fetal islet progenitor cells that express Ngn3, an exciting step that creates powerful methods for purifying islet progenitors from virtually any human cell source. Our isolation and culture of human islet progenitor cells (Sugiyama et al 2006) also provides unique opportunities to investigate mechanisms regulating development and expansion of human islets, about which virtually nothing is known. We will now use molecular methods to ask: What genes are required for development of hormone-expressing human islet cells? What signaling pathways direct islet development? What mitogens stimulate expansion of islet progenitors or differentiated hormone-expressing islets? Can isolated islet progenitors be expanded and differentiated into functional islet cells? Answering these questions will require methods to mark cell lineages, inactivate gene products, and prospectively measure cell differentiation and fate in vitro and in vivo.

4. Decoding the basis of physiological and pathological ß-cell proliferation

Once viewed as post-mitotic and incapable of significant proliferation, mature ß-cells in the adult pancreas are now recognized to have a significant capacity to replicate, and thereby maintain ß-cell mass (reviewed in Heit et al 2006a). For this reason, expansion of islets in culture or in the pancreas may become a therapeutic option for diabetes. However, prior attempts to expand cultured islets with mitogens have been bedeviled by the loss of key ß-cell features, like insulin expression, that accompanies proliferation. Thus, it remains elusive how adult ß-cells 'remember' their differentiated fate while proliferating. To decode the mechanisms controlling ß-cell proliferation, we sought to identify genetic and epigenetic pathways that govern expression of hallmark ß-cell factors and cell cycle regulators (Figure 4, from Heit et al 2006a).

Figure 4. Signaling pathways that regulate ß-cell proliferation and differentiation. From Heit et al 2006a.

A. Calcineurin/NFAT signaling regulates ß-cell growth and function

Our studies with the Crabtree group at Stanford revealed that calcineurin/NFAT signaling is a major regulator of ß-cell proliferation and function, and identified this signaling pathway as the basis for developing new therapies for diabetes and endocrine cancers. Diabetes in patients treated with calcineurin inhibitors like cyclosporin A suggested that calcineurin/NFAT signaling might control islet growth or function. In a widely heralded study (Heit et al 2006b), we inactivated calcineurin specifically in ß-cells of juvenile mice and showed that these mice develop age-dependent diabetes. Diabetes in this new model was accompanied by reduced ß-cell mass, a consequence of impaired physiologic ß-cell proliferation. By contrast, conditional NFATc activation in mouse ß-cells stimulated proliferation, leading to increased functional ß-cell mass, insulin content, and hypoglycemia. Ten new targets of NFAT regulation identified by our studies included all 6 genes mutated in hereditary forms of monogenic type 2 diabetes (called MODY), as well as regulators of the ß-cell cycle. Thus, complimentary loss- and gain-of-function strategies demonstrated that calcineurin/NFAT signaling is required and sufficient for maintaining hallmark features of ß-cell growth and function (Heit et al 2006b). We are now testing if attenuation of endogenous NFATc inhibitors can be used to increase ß-cell NFATc activity and promote proliferation without loss of ß-cell features. If so, we envision our work could lead to discovery of drugs that stimulate ß-cell expansion.

B. Menin regulates histone modifications governing ß-cell fate and proliferation

My group has studied a rare human cancer syndrome to gain fundamental insights about mechanisms controlling pathological and physiological ß-cell growth. We have investigated the basis of islet tumor formation in type 1 multiple endocrine neoplasia (MEN1), a familial cancer syndrome I have managed as an oncologist, which results from mutation of the Men1 gene. A key observation in MEN1 motivating our work is that ß-cells in islet tumors often remain “functional”, leading to elevated circulating insulin levels and symptoms from hypoglycemia. Thus, proliferating ß-cells in MEN1 patients often maintain their defining functions. The Men1 gene encodes a protein called menin, and our studies have unveiled a novel epigenetic mechanism of tumor suppression by menin (Karnik et al 2005). Menin associates with MLL, a member of the Trithorax group of proteins, in a nuclear complex that promotes specific covalent histone modifications that regulate expression of target genes. In studies of Men1-deficient mice with features of MEN1 syndrome, we showed that menin promotes histone methylation and expression of genes encoding the cyclin dependent kinase inhibitors p27Kip1, p18INK4C and other cell cycle regulators in islet ß-cells (Karnik et al, 2005). Thus, we showed that menin-dependent histone modifications control islet ß-cell proliferation.

In addition to menin roles in tumor suppression, our recent work shows for the first time that menin also regulates histone modifications and ß-cell proliferation during physiological growth (Karnik et al 2007). Common states like pregnancy and obesity stimulate adaptive ß-cell expansion, and we have discovered that menin levels in ß-cells are strikingly attenuated in each of these states. In pregnant mice, we have shown that reduction of menin levels and function is required for the maternal ß-cell expansion that prevents gestational diabetes. Levels of Men1 expression and menin are regulated by maternal hormones like prolactin and progesterone. This temporary menin attenuation leads to reversible changes in histone methylation and expression of p27Kip1 and p18INK4C by maternal islet ß-cells. Moreover, each of these features is conserved in human islets. Prevention of menin attenuation in pregnant mice impairs ß-cell histone modification and adaptive islet growth, producing gestational diabetes in a unique genetic model (Karnik et al 2007). Thus, excess menin function in human islet could underlie common subsets of type 2 diabetes.

Advances from our studies have opened new frontiers for investigation in islet biology and created new strategies to prognose and treat islet diseases. Deciphering a dynamic histone code that orchestrates the maintenance of hallmark functions in proliferating islet ß-cells could identify a drug-responsive pathway that controls ß-cell proliferation, leading to new treatments for diabetes or neuroendocrine cancers like insulinomas. Likewise, verification that menin or associated proteins in a histone-methylating complex function as an endogenous molecular 'brake' that controls physiologic human ß-cell proliferation would motivate us to search for compounds that attenuate menin levels or activity in ß-cells and could serve as the basis for expanding functional islets in type 1 diabetes.

Summary

Our efforts in just a few years have created unprecedented opportunities for harnessing knowledge about the molecular and cellular basis of pancreatic development and growth to restore pancreas islet function and to treat endocrine cancers. Our work with Drosophila, mice, human islet organogenesis and diseases, cell purification, and chromatin regulation has revealed mechanisms underlying islet development, adaptations and disease pathogenesis. We hope that our discoveries will provide the tools and expertise needed to produce islet regeneration therapies for type 1 diabetes, improve treatments and tests to mitigate or prevent type 2 diabetes, and generate new therapeutic strategies for neuroendocrine cancers.