IMIDIA: a precompetitive consortium for the β cell

Bernard THORENS, PhD
Center for Integrative Genomics
University of Lausanne

IMIDIA: a precompetitive consortium for the β cell

by B. Thorens, Switzerland

An absolute or relative deficit in insulin secretion underlies the pathogenesis of type 1 and type 2 diabetes, respectively. Restoring normal pancreatic β-cell insulin secretion capacity and preventing the demise of β cells are, therefore, major goals to improve treatment or prevent development of diabetes. Although β-cell physiology has been investigated for decades, there is still insufficient knowledge of the metabolic, signaling, and differentiation pathways that control β-cell plasticity and failure in adult life. This stems largely from the relative scarcity of β cells, which limits classic cell biological and biochemical investigations, and from the lack of satisfactory imaging techniques to assess their mass and function in living individuals. There is thus an urgent need to develop new tools and experimental approaches to investigate β-cell function and mass. These include fundamental investigations on molecular mechanisms controlling β-cell plasticity, the development of human β-cell lines, the identification of plasma biomarkers of β-cell function and response to drug treatments, and new imaging techniques to monitor β-cell mass and function in vivo. Because of the formidable complexity of the task, no single laboratory or industry can hope to independently achieve these goals. The Innovative Medicine Initiative for DIAbetes (IMIDIA) project is a precompetitive consortium of 15 academic laboratories, 1 small and medium enterprise (SME) and 8 major pharmaceutical companies, which has been set up to work in a highly integrated and coordinated manner to address these key challenges.

Medicographia. 2014;36:384-390 (see French abstract on page 390)

Insulin action on liver, muscle, and fat controls glucose absorption and utilization to generate metabolic energy and precursors for biosynthetic reactions, such as nucleotides synthesis, or for conversion of glucose into glycogen or fat. Insulin is also required to block hepatic glucose production. Insulin resistance, ie, a progressive decline in the efficacy of insulin to control these essential actions, occurs in conditions such as obesity, pregnancy, or aging. Unless pancreatic β cells compensate for insulin resistance, by increasing their insulin secretion capacity, hyperglycemia will develop.1 This compensation process involves not only an increase in insulin secretion capacity by individual β cells, but also an augmentation of their total number,2,3 which can arise from replication of mature β cells, differentiation from progenitors, or transdifferentiation of a or even exocrine cells into β cells.4 This adaptation capacity is, however, limited, and if insulin resistance continues to develop, insulin secretion becomes insufficient, leading to the onset of diabetic hyperglycemia.

Figure 1
Figure 1. β-Cell plasticity and failure in the pathogenesis
of type 2 diabetes.

The β-cell mass present in the pancreas of healthy individuals
can increase to compensate for the development of insulin resistance
and to maintain normoglycemia. As insulin resistance
continues to develop, β cells reach a stage where they can no
longer increase their total insulin secretion capacity. This leads
to the development of hyperglycemia. The combination of hyperglycemia
(and higher plasma free fatty acids released form
insulin-resistant adipocytes) may precipitate a decrease in the
secretion capacity of the β cells and their progressive death by
apoptosis. Unpublished data.

This then further precipitates β-cell dysfunction and induces their death by apoptosis (Figure 1). The insulin resistance threshold at which β cells lose their adaptation capacity is probably dependent on each individual’s genetic architecture. Understanding the molecular control of β-cell plasticity and failure in diabetes is thus a major goal of current diabetes research, with the anticipation that molecular understanding of this plasticity may lead to better therapeutic options in the treatment of type 2 diabetes.4 The main questions to be addressed can be summarized as follows:
♦ What are the intracellular signaling, metabolic, and differentiation pathways that control β-cell secretion activity and cell number in adult mice, in response to insulin resistance and upon metabolic stress?
♦ Can noninvasive monitoring techniques be developed to assess β-cell mass and function in normal and insulin-resistant states, and in response to therapies of diabetes?
♦ Can plasma biomarkers be identified to reliably inform on the functional state of β cells and their response to pharmacological treatments or lifestyle modifications?

Here, I will describe how we have organized, in response to a call from the Innovative Medicine Initiative for DIAbetes (IMIDIA) of the European Union (EU), an academic-pharmaceutical network to address these questions at an unprecedented scale, and with the goal to generate new tools and knowledge as a basis for the development of novel strategies for β cell–centered treatment of type 2 diabetes.

Objectives of the IMIDIA precompetitive network

Recognizing the importance of a β cell–centered investigation program, the EU and the European Federation of Pharmaceutical Industries and Associations (EFPIA) created a public–private partnership (PPP) project to address key bottlenecks in β-cell research. This PPP would bring together academic laboratories and small and medium enterprises (SMEs), in collaboration with the pharmaceutical industry. A call for application was launched in 2009, asking academic networks to propose a specific action plan to address a set of key β cell–related questions. The IMIDIA academic network was selected to complete a full proposal in close interaction with representatives from eight major pharmaceutical companies. The IMIDIA precompetitive network was officially launched on February 1st, 2010. This 5-year project integrates the research activities of more than 100 collaborators in academia, one SME, and the participating pharmaceutical companies with a global budget of ≈25 million euros (Figure 2, page 386; refer to footnote for a full list of participants and general goals).5

Specific issues related to the study of adult b-cell plasticity

Progress towards elucidating the molecular basis of adult β-cell plasticity and failure in type 2 diabetes is faced with several difficulties, due to unique β-cell specificities.

First, is their relative scarcity. Indeed, pancreatic islets represent ≈1% of the total mass of the pancreas and β cells only ≈70% of the islet cells. This represents an important limitation when working with rodent islets and an even greater one when working with human islets, which can only be obtained from autoptic pancreas or infrequently from biopsy material obtained after elective pancreas surgery, usually for tumor removal. Thus, identification of key intracellular signaling, metabolic, and differentiation pathways using classic cell biological and biochemical techniques is very limited, and must be combined with alternative approaches, such as genetic, genomic, and metabolomic techniques.

Second, identification of plasma biomarkers, for monitoring β-cell mass and function in diabetes and during therapeutic treatment, requires the availability of cohorts of individuals who have progressed from a normal state to type 2 diabetes, and who have been well phenotyped during the pathogenic progression.

Third, the distribution of the endocrine pancreas in small islets, consisting of 1-2000 cells, scattered in the exocrine pancreas, imposes particular difficulties in developing contrasting agents and imaging modalities that can specifically and precisely identify these cells and assess their functions.

Figure 2
Figure 2. The IMIDIA consortium and its geographical distribution.

After reference 5: IMIDIA. http://www. © IMIDIA Consortium.

These intricate questions need to be approached in a global, coordinated manner, requiring diverse experimental approaches and technical developments. The IMIDIA network has selected to work along the following lines:
♦ Develop new knowledge about the mechanisms controlling adult β-cell function, expansion, and death.
♦ Establish new tools for β-cell research, in particular new, functionally validated, human β-cell lines.
♦ Identify plasma biomarkers diagnostic of β-cell function and deregulation in the pathogenesis of type 2 diabetes and in response to treatment.
♦ Develop new imaging techniques for in vivo monitoring of β-cell mass and function in diabetes and upon therapeutic treatments.

Specific research projects

Undoubtedly, the major benefit of the IMIDIA consortium was the possibility to address multiple questions and to develop new investigative tools related to b-cell biology, using broadbased technological approaches contributed to by both academic and pharmaceutical partners in a highly synergistic way. Some of the ongoing projects have been presented at a meeting preceding the European Association for the Study of Diabetes (EASD) annual meeting in Barcelona (September 2013), where the three Innovative Medicines Initiative (IMI) diabetes networks (IMIDIA, SUMMIT [SUrrogate markers for Micro- and Macrovascular hard end points for Innovative diabetes Tools], and DIRECT [DIabetes REsearch on patient stratifiCaTion]) highlighted their key progress.6

A systems biology approach to discover novel genes, gene regulatory pathways, and biomarkers of functional β-cell mass
In order to obtain a new view of the functional architecture of β cells in health and in response to metabolic stress, a large Systems Biology approach was undertaken. This consisted of two investigative arms: one focused on mouse islets and the other on human islets. Both arms proceeded independently until the stage of bioinformatics analysis of the collected data, where comparative analysis of mouse and human data reinforced each other in the quest for new essential information about β-cell biology.

In the animal study, mice from different genetic backgrounds were fed a regular or high fat diet for different periods of time (2,10, 30, and 90 days; six different inbred strains of mice were chosen because they differentially adapt to metabolic stress, allowing information on the interaction between genetics and nutrition to be obtained). Each mouse was then fully genotyped for basic parameters such as glycemia and insulinemia, but also for glucose and insulin tolerance, and their β-cell mass was assessed. In parallel, lipidomic analysis was performed to determine the plasma concentration of a few hundreds lipid species and the lipid composition of isolated islets; islet gene expression was determined by RNA sequencing analysis, a technique providing highly accurate and very extensive information about which genes are expressed in islets and at which level.

A critical aspect of these studies was that each piece of data generated for each mouse was deposited, with a unique identification tag, in a database organized and maintained by partners of the Swiss Institute of Bioinformatics.7 These data could be interrogated in a global manner to find correlations between the measured phenotypes and the gene expression profile of islets, or the particular lipid composition of the plasma. Two types of information have been extracted: (i) the composition of new modules (groups) of β-cell genes whose expression correlates with any of the measured phenotypes; and (ii) identification of modules of plasma lipids, which correlate with a particular β-cell phenotype. Modules of gene expression have been identified, which highlight the involvement of so far unsuspected signaling or metabolic pathways with, for instance, insulin secretion capacity or glucose intolerance (Figure 3). These data are now being validated using basic cell biological studies on insulin cell lines or primary islets; these studies confirm that our Systems Biology approach is extremely powerful in generating new discoveries that can hopefully lead to new molecular targets to improve β-cell plasticity.

On the other hand, the plasma lipidomic analysis reveals tight correlation between the concentration of a small group of lipids, with insulin secretion and glycemic control. These studies may lead to identification of novel biomarkers diagnostic of β-cell function. Confirmation of these initial experiments in human plasma samples is now required, a task that is again facilitated by the availability, within the IMIDIA consortium, of longitudinal cohorts of individuals who have developed type 2 diabetes, and for whom plasma samples have been collected before and after the development of diabetes.

Figure 3
Figure 3. Gene modules underlying the adaptation of β cells to
metabolic stress.

A Systems Biology analysis of global gene expression in β cells from mice with
different genetic backgrounds and fed different diets, led to a new view of the
functional organization of β cells. Each dot in the figure represents an individual
gene and the lines connecting the dots represent how their expression is correlated.
The different groups of genes identify functional modules, for instance:
(1) insulin secretion; (2) calcium-mediated responses; (3) cell cycle and mitosis;
(4) metabolism; and (5) insulin secretion and electron transport chain.
Unpublished data: Ibberson M, Liechti R, Xenarios I. Swiss Institute for Bioinformatics.

In the second arm of this Systems Biology approach, a unique European network has been established to isolate and characterize human islets using highly standardized methods. These islet preparations, which come from healthy individuals and type 2 diabetic patients, have been analyzed for gene expression profiling, lipidomic analysis, and the presence of diabetes susceptibility genes.8

Global analysis of gene expression modules in islets from humans and mice is now providing a unique view on the functional organization of β cells in health and disease, and provides an extraordinary new base for the discovery of new targets for β cell–centered therapy for diabetes and for the identification of biomarkers for monitoring β-cell mass and function.

Human β-cell lines
Rodent insulin-secreting cell lines have been available for several decades. Even though they differ in many respects from primary β cells, they have been invaluable for the study of various aspects of β-cell biology, including insulin biosynthesis and the glucose signaling pathway that controls insulin secretion.9-11 In contrast, human insulin cell lines have been much more difficult to obtain, since the transgenic approach used to express proliferation-inducing genes in mouse β cells could not be used in human. An alternative strategy used by IMIDIA researchers led to the first successful development of human cell lines (Figure 4, page 388).12 This approach was based on the transplantation, under the kidney capsules of immunoincompetent mice, of human fetal islet cells transduced with a transforming oncogene. This success was acclaimed as a major breakthrough in human β-cell research, which could have major implications in the development of new drugs controlling insulin secretion and β-cell proliferation. The effort to establish and validate this first cell line as well as the ongoing development of second-generation lines has benefited from a unique combination of academic and private research performed in a SME. IMIDIA is now providing further support for validation of these cell lines by several pharmaceutical partners, to establish them as industryrelevant research tools.

Centralized database
From the planning stage of IMIDIA’s activities, it was recognized that for a successful integration of all network activities, strong database management and bioinformatics supports were essential. Indeed, to identify novel genes and gene networks controlling β-cell plasticity, to find novel biomarkers of β-cell function, and to be able to compare data obtained in animal models and in human tissues, it is essential that all data generated throughout the network are stored in a central database in a format that allows their retrieval for global bioinformatics analysis. This was achieved by creating unique identifiers for each data (individual mice, islet preparation, RNASeq data, phenotype of donors for human islets, plasma lipidomic, glucose tolerance tests, etc).

This initial effort to uniquely label each piece of data has been critical for the bioinformatics analysis. As mentioned above, the first goal was to generate a new view of β-cell gene net- works that are involved in the control of their function, or failure to adapt to metabolic stress. Gene modules have now been discovered that are associated with phenotypes such as insulin secretion, oral glucose tolerance, insulin resistance, body weight, etc.
In each of these modules, genes with highest impact on the module classification have been identified, and in many cases these can be placed in a gene interaction network, allowing identification of genes with important hub positions, and thus probably critical function in β-cell adaptation to metabolic stress.

Figure 4
Figure 4. Characterization of the human β-cell line EndoC-βH1.

Left: Immunofluorescence microscopy analysis insulin (Ins) and C-peptide (C-PEPT), the β-cell–specific transcription factors PDX1 and NKX6-1. Right: Insulin secretion by EndoC-βH1 cells exposed to indicated glucose concentrations and to GLP-1 (glucagon-like peptide 1), glibenclamide (Gli), the phosphodiesterase inhibitor isobutylmethlyxanthine (IBMX), diazoxide (DZ), or leucine (Leu).
After reference 12: Ravassard et al. J Clin Invest. 2011;121:3589-3597. © 2011, American Society for Clinical Investigation.

Figure 5
Figure 5. Different mouse islet imaging modalities.

(A) Immunofluorescence microscopy detection of insulin (Ins; green) and glucagon (Glu; red) in sections of pancreas from a control mouse or from a mouse treated with diphtheria toxin to ablate β cells (these mice express a transgenic diphtheria toxin receptor in β cells). (B) Blood glucose levels from control- and diphtheria toxin–treated mice. (C) Luminescence emission from β cells expressing a transgenic luciferase gene, without (control) or after (DT-treated) diphtheria toxin treatment.
After reference 16: Virostko et al. Proc Natl Acad Sci U S A. 2011;108:20719-20724. © 2011, The authors.

Importantly, crossing mouse and human islets expression data allows the finding of genes with similar hub positions in both species, further supporting their functional importance. These genes are now actively being investigated in a coordinated manner by several IMIDIA laboratories.

Imaging the β cells
One critical shortcoming in the assessment of β-cell plasticity is the lack of suitable in vivo imaging techniques to accurately determine β-cell mass and secretion activity. This precludes the real appreciation of the relative contribution of β-cell number and function in the progression towards type 2 diabetes and the impact of pharmacological or lifestyle interventions on β-cell function.13-15

Figure 6
Figure 6. Analysis
of islet distribution
in the pancreas
of a mouse.

Exendin-4 conjugated to Alexa Fluor 594 was injected through the tail vein of a mouse. After fixation, the pancreas was processed for assessment of fluorescence distribution by optical rojection tomography. This reveals the islet distribution in the pancreas and can be used to quantify β-cell mass.
After reference 19: Ahnfelt-Rønne et al. Diabetologia. 2012;55: 2316-2318. © 2012, Springer-Verlag.

Development of imaging modalities is therefore an area of intense research across the world. Because of the complexity in developing appropriate β-cell imaging markers that can be used with various imaging modalities, and because these techniques need to be applied to humans, the cost of development and the expertise required need a highly integrated approach, which can be best obtained by combining academic and pharmaceutical partners.

Within IMIDIA, innovative imaging markers are currently being developed. For example, chemists are developing new derivatives of sulfonylureas and the gluco-incretin hormone glucagon-like peptide 1 (GLP-1), as well as molecules sensitive to Zn++, an ion secreted by β cells at the time of insulin granule exocytosis.14,16-18 These probes are used for optical and magnetic resonance imaging (MRI), and positron emission tomography (PET); see Figures 5 and 6 for examples.16,19

They are tested in cellular and animal models available in different laboratories, and the successful ones will be tested in humans, as already occurs as part of some of IMIDIA’s partner activities.

General considerations

There is a challenge in establishing a large-scale academicpharmaceutical partnership such as IMIDIA. This requires both parties to be willing to work beyond their usual comfort zones. IMIDIA has demonstrated that this is, however, not only possible, but also highly productive. The examples mentioned above illustrate that completely new avenues of research and tool development are being successfully pursued. The mass of data generated, and the new tools developed will feed years of productive research. The challenge, which is now addressed, is to fully integrate this new knowledge into translational research for improved diabetes treatment. With the planning of a new wave of Innovative Medicine Initiative projects by the EU, we trust that sufficient resources will be made available to fruitfully conduct this translational effort.

Acknowledgments: Support for work performed in the author’s laboratory was from the Swiss National Science Foundation grant 3100A0- 113525 and from the Innovative Medicine Initiative Joint Undertaking under grant agreement number 155005 (IMIDIA), resources of which are composed of financial contribution from the European Union’s Sevenths Framework Programme (FP7/2007-2013) and European Federation of Pharmaceutical Industries and Associations (EFPIA). The author would like to warmly acknowledge the contribution of all IMIDIA partners to the realization of the research goals briefly described in this review.
Footnote: The IMIDIA consortium is coordinated by Sanofi-Aventis (Werner Kramer), Servier (Alain Ktorza), and the University of Lausanne (Bernard Thorens). The other pharmaceutical partners are: AstraZeneca, Boehringer Ingelheim, F. Hoffmann-La Roche, Lilly Deutschland, Novo Nordisk A/S, and Novartis Institute. The academic partners are: CEA/Institut d’Imagerie Biomédicale, CNRS UMR 7091, CNRS-University Paris Diderot, Dresden University of Technology, Hannover Medical School, Imperial College London, INSERM U845, Swiss Institute of Bioinformatics, University of Geneva, University of Pisa, Vrije Universiteit Brussel, and the small and medium enterprise (SME) ENDOCELLS, Paris.

1. Prentki M, Nolan CJ. Islet beta cell failure in type 2 diabetes. J Clin Invest. 2006; 116:1802-1812.
2. Butler AE, Janson J, Bonner-Weir S, Ritzel R, Rizza RA, Butler PC. Beta-cell deficit and increased beta-cell apoptosis in humans with type 2 diabetes. Diabetes. 2003;52:102-110.
3. Henquin JC, Rahier J. Pancreatic alpha cell mass in European subjects with type 2 diabetes. Diabetologia. 2011;54:1720-1725.
4. Thorens B. The required beta cell research for improving treatment of type 2 diabetes. J Int Med. 2013;274:203-214.
5. IMIDIA. Innovative Medicines Initiative for Diabetes: Improving beta-cell function and identification of diagnostic biomarkers for treatment monitoring in Diabetes Web site. Accessed January 13, 2014.
6. IMIDIA. Joint symposium on the occasion of the 49th EASD Annual meeting. Accessed January 13, 2014.
7. Schweizerisches Institut Für Betriebsökonomie. Accessed January 13, 2014.
8. Bonnefond A, Froguel P, Vaxillaire M. The emerging genetics of type 2 diabetes. Trends Mol Med. 2010;16:407-416.
9. Miyazaki J, Araki K, Yamato E, et al. Establishment of a pancreatic beta cell line that retains glucose-inducible insulin secretion: special reference to expression of glucose transporter isoforms. Endocrinology. 1990;127:126-132.
10. Asfari M, Janjic D, Meda P, Li G, Halban PA, Wollheim CB. Establishment of 2-mercaptoethanol-dependent differentiated insulin-secreting cell lines. Endocrinology. 1992;130:167-178.
11. Efrat S, Fusco-DeMane D, Lemberg H, Al Emran O, Wang X. Conditional transformation of a pancreatic b-cell line derived from transgenic mice expressing a tetracycline-regulated oncogene. Proc Natl Acad Sci U S A. 1995;92:3576-3580.
12. Ravassard P, Hazhouz Y, Pechberty S, et al. A genetically engineered human pancreatic beta cell line exhibiting glucose-inducible insulin secretion. J Clin Invest. 2011;121:3589-3597.
13. Di Gialleonardo V, de Vries EF, Di Girolamo M, Quintero AM, Dierckx RA, Signore A. Imaging of beta-cell mass and insulitis in insulin-dependent (type 1) diabetes mellitus. Endocr Rev. 2012;33:892-919.
14. Andralojc K, Srinivas M, Brom M, et al. Obstacles on the way to the clinical visualisation of beta cells: looking for the Aeneas of molecular imaging to navigate between Scylla and Charybdis. Diabetologia. 2012;55:1247-1257.
15. Arifin DR, Bulte JW. Imaging of pancreatic islet cells. Diabetes Metab Res Rev. 2011;27:761-766.
16. Virostko J, Henske J, Vinet L, et al. Multimodal image coregistration and inducible selective cell ablation to evaluate imaging ligands. Proc Natl Acad Sci U S A. 2011;108:20719-20724.
17. Lamprianou S, Immonen R, Nabuurs C, et al. High-resolution magnetic resonance imaging quantitatively detects individual pancreatic islets. Diabetes. 2011; 60:2853-2860.
18. Li D, Chen S, Bellomo EA, et al. Imaging dynamic insulin release using a fluorescent zinc indicator for monitoring induced exocytotic release (ZIMIR). Proc Natl Acad Sci U S A. 2011;108:21063-21068.
19. Ahnfelt-Rønne J, Hecksher-Sørensen J, Schäffer, L, Madsen OD. A new view of the beta cell. Diabetologia. 2012;55:2316-2318.

Keywords: β cells; biomarkers; imaging; lipidomic; signaling pathways; systems biology; transcriptomic; type 2 diabetes