Ron Piran

Pancreatic β‐Cell Neogenesis by Direct Conversion from Mature α‐Cells

Because type 1 and type 2 diabetes are characterized by loss of β‐cells, β‐cell regeneration has garnered great interest as an approach to diabetes therapy. Here, we developed a new model of β‐cell regeneration, combining pancreatic duct ligation (PDL) with elimination of pre‐existing β‐cells with alloxan. In this model, in which virtually all β‐cells observed are neogenic, large numbers of β‐cells were generated within 2 weeks. Strikingly, the neogenic β‐cells arose primarily from α‐cells. α‐cell proliferation was prominent following PDL plus alloxan, providing a large pool of precursors, but we found that β‐cells could form from α‐cells by direct conversion with or without intervening cell division. Thus, classical asymmetric division was not a required feature of the process of α‐ to β‐cell conversion. Intermediate cells coexpressing α‐cell‐ and β‐cell‐specific markers appeared within the first week following PDL plus alloxan, declining gradually in number by 2 weeks as β‐cells with a mature phenotype, as defined by lack of glucagon and expression of MafA, became predominant. In summary, these data revealed a novel function of α‐cells as β‐cell progenitors. The high efficiency and rapidity of this process make it attractive for performing the studies required to gain the mechanistic understanding of the process of α‐ to β‐cell conversion that will be required for eventual clinical translation as a therapy for diabetes. STEM CELLS 2010; 28:1630–1638.

COP9 signalosome components play a role in the mating pheromone response of S. cerevisiae

A family of genetically and structurally homologous complexes, the proteasome lid, Cop9 signalosome (CSN) and eukaryotic translation initiation factor 3, mediate different regulatory pathways. The CSN functions in numerous eukaryotes as a regulator of development and signaling, yet until now no evidence for a complex has been found in Saccharomyces cerevisiae. We identified a group of proteins, including a homolog of Csn5/Jab1 and four uncharacterized PCI components, that interact in a manner suggesting they form a complex analogous to the CSN in S. cerevisiae. These newly identified subunits play a role in adaptation to pheromone signaling. Deletants for individual subunits enhance pheromone response and increase mating efficiency. Overexpression of individual subunits or a human homolog mitigates sst2‐induced pheromone sensitivity. Csi1, a novel CSN interactor, exhibits opposite phenotypes. Deletants also accumulate Cdc53/cullin in a Rub1‐modified form; however, this role of the CSN appears to be distinct from that in the mating pathway.

The COP9 signalosome-like complex in S. cerevisiae and links to other PCI complexes

The COP9 signalosome (CSN), the lid subcomplex of the proteasome and translational initiation factor 3 (eIF3) share structural similarities and are often referred to as the PCI family of complexes. In multicellular eukaryotes, the CSN is highly conserved as an 8-subunit complex but in Saccharomyces cerevisiae the complex is rather divergent. We further characterize the composition and properties of the CSN in budding yeast and its interactions with these related complexes. Using the generalized profile method we identified CSN candidates, four with PCI domains: Csn9, Csn10, Pci8/Csn11, and Csn12, and one with an MPN domain, Csn5/Rri1. These proteins and an additional interactor, Csi1, were tested for pairwise interactions by yeast two-hybrid and were found to form a cluster surrounding Csn12. Csn5 and Csn12 cofractionate in a complexed form with an apparent molecular weight of roughly 250kDa. However, Csn5 migrates as a monomer in Deltacsn12 supporting the pivotal role of Csn12 in stabilizing the complex. Confocal fluorescence microscopy detects GFP-tagged Csn5 preferentially in the nucleus, whereas in absence of Csn12, Csn10, Pci8/Csn11, or Csi1, Csn5 is delocalized throughout the cell, indicating that multiple subunits are required for nuclear localization of Csn5. Two CSN subunits, Csn9 and Csi1, interact with the proteasome lid subunit Rpn5. Pci8/Csn11 has previously been shown to interact with eIF3. Together, these results point to a network of interactions between these three structurally similar, yet functionally diverse, complexes.

Pharmacological induction of pancreatic islet cell transdifferentiation: relevance to type I diabetes

Type I diabetes (T1D) is an autoimmune disease in which an immune response to pancreatic β-cells results in their loss over time. Although the conventional view is that this loss is due to autoimmune destruction, we present evidence of an additional phenomenon in which autoimmunity promotes islet endocrine cell transdifferentiation. The end result is a large excess of δ-cells, resulting from α- to β- to δ-cell transdifferentiation. Intermediates in the process of transdifferentiation were present in murine and human T1D. Here, we report that the peptide caerulein was sufficient in the context of severe β-cell deficiency to induce efficient induction of α- to β- to δ-cell transdifferentiation in a manner very similar to what occurred in T1D. This was demonstrated by genetic lineage tracing and time course analysis. Islet transdifferentiation proceeded in an islet autonomous manner, indicating the existence of a sensing mechanism that controls the transdifferentiation process within each islet. The finding of evidence for islet cell transdifferentiation in rodent and human T1D and its induction by a single peptide in a model of T1D has important implications for the development of β-cell regeneration therapies for diabetes.

A Molecular Cryptosystem for Images by DNA Computing

Image encryption by immobilized DNA molecules on chips could offer significant advantages, which include vast parallelism, immense information density, high chemical stability, and energy efficiency. Several theoretical or computer simulated models for encryption and steganography of texts have been proposed on the basis of DNA molecules, and a few DNA-based models relevant to alphanumeric information have been realized. DNA methods were also proposed for commercial encryption of signatures 12] and for information storage. In contrast, image encryption has received very little attention. Although DNA computing procedures were employed in theoretical schemes, no molecular encryption of images has been tested experimentally, and certainly not by methods that involve molecular computing. Herein, we report the use of parallel computing with molecular finite-state automata and fluorescently labeled DNA molecules for deciphering two different images. Both logos of the Technion and The Scripps Research Institute were encrypted on a single DNA chip. To decipher any of these images, a mixture of input molecules was processed by a molecular finite-state automaton, which led to image visualization by fluorescent, surface-bound output molecules. The huge diversity of potentially encrypted images that is offered by this strategy stems from current DNA microarray technology, with millions of printed pixels per chip, along with a plethora of possible input molecules and a variety of DNAbased automata. Our approach to meet the challenge of image encryption was based on our two-symbol-two-state finite automata as a mathematical model for information retrieval by using DNA molecules (Figure 1A). We prepared two input molecules In1 and In2 in the form of a linear double-stranded (ds)DNA (Figure 1B) that contain a string of symbols, which are six base pairs (bp) each. In1 comprises the symbols ab, whereas In2 contains the string aa. Each symbol-string was followed by a six bp terminator segment (t) and a long singlestranded (ss)DNA tail (48–56 bases). This tail was designed to be uniquely complementary to one of the connecting molecules Con1 and Con2, to avoid self-hybridization, secondary structures, and cross-hybridization with incorrect connecting molecules. The connecting molecules, in the form of ssDNA, were intended to localize the output molecules on the microarray glass slide. Again, an important part of their design considered the need to prevent self-hybridization, secondary structures and cross-hybridization with incorrect molecules. The two-symbol-two-state automata were programmed by the choice of subsets from the complete library of eight transition molecules. Each such molecule, which represents a transition rule, is comprised of three parts: a recognition site for the type II endonuclease FokI, a four-base sticky end, and a spacer of between one and five bp between them (Figure 1C). Automaton 1 (A1, Figure 1A) includes transition molecules TM2, TM3, TM6, and TM7, whereas automaton 2 (A2), comprises TM1, TM4, TM6, and TM8. The two dsDNA detection molecules DM0 and DM1 (Figure 1 D) contain a four-base sticky end, which is complementary to one of the two four-base sticky ends that are generated upon restriction of the terminator domain at the final step of the computation. Each deciphering mixture was generated by a computing process, which was carried out by mixing both input molecules In1 and In2, the appropriate set of transition molecules (software), and a buffered solution that contained adenosine triphosphate (ATP), FokI, and ligase (hardware). Mixing the above components results in an autonomous cascade of chemical events that includes repetitive cycles of DNA restriction, hybridization, and ligation. Restriction of the input by FokI represents reading an input symbol and determining the new internal state of the automaton. Each of the six bp symbols, as well as the terminator, can be cleaved by this endonuclease either at the beginning of the symbol or two bp deeper into that domain. The two restriction modes represent the two internal states, S0 and S1 respectively. Each mode is dictated by the distance between the recognition site of FokI and the cleavable symbol, as instructed by the newly incorporated TM. After all symbols are digested, the six bp terminator segment is restricted either at its beginning or two bases deeper, which creates one of two possible four-base sticky ends. One sticky end, 5’-ATAC, represents the final state S0, whereas the other, 5’-ACAT, represents S1. Thus, each of these sticky ends can ligate to one of the detection molecules DM0 or DM1. For example, processing of In1 with [*] S. Shoshani, Dr. R. Piran, Prof. E. Keinan Schulich Faculty of Chemistry, Technion–Israel Institute of Technology, Technion City, Haifa 32000 (Israel) E-mail: keinan@technion.ac.il

Algorithm of myogenic differentiation in higher-order organisms.

Cell fate determination is governed by complex signaling molecules at appropriate concentrations that regulate the cell decision-making process. In vertebrates, however, concentration and kinetic parameters are practically unknown, and therefore the mechanism by which these molecules interact is obscure. In myogenesis, for example, multipotent cells differentiate into skeletal muscle as a result of appropriate interplay between several signaling molecules, which is not sufficiently characterized. Here we demonstrate that treatment of biochemical events with SAT (satisfiability) formalism, which has been primarily applied for solving decision-making problems, can provide a simple conceptual tool for describing the relationship between causes and effects in biological phenomena. Specifically, we applied theŁ ukasiewicz logic to a diffusible protein system that leads to myogenesis. The creation of an automaton that describes the myogenesis SAT problem has led to a comprehensive overview of this non-trivial phenomenon and also to a hypothesis that was subsequently verified experimentally. This example demonstrates the power of applying Łukasiewicz logic in describing and predicting any decision-making problem in general, and developmental processes in particular.

Induction of β-cell replication by a synthetic HNF4α antagonist

Increasing the number of β cells is critical to a definitive therapy for diabetes. Previously, we discovered potent synthetic small molecule antagonists of the nuclear receptor transcription factor HNF4α. The natural ligands of HNF4α are thought to be fatty acids. Because obesity, in which there are high circulating levels of free fatty acids, is one of the few conditions leading to β‐cell hyperplasia, we tested the hypothesis that a potent HNF4α antagonist might stimulate β‐cell replication. A bioavailable HNF4α antagonist was injected into normal mice and rabbits and β‐cell ablated mice and the effect on β‐cell replication was measured. In normal mice and rabbits, the compound induced β‐cell replication and repressed the expression of multiple cyclin‐dependent kinase inhibitors, including p16 that plays a critical role in suppressing β‐cell replication. Interestingly, in β‐cell ablated mice, the compound induced α‐ and δ‐cell, in addition to β‐cell replication, and β‐cell number was substantially increased. Overall, the data presented here are consistent with a model in which the well‐known effects of obesity and high fat diet on β‐cell replication occur by inhibition of HNF4α. The availability of a potent synthetic HNF4α antagonist raises the possibility that this effect might be a viable route to promote significant increases in β‐cell replication in diseases with reduced β‐cell mass, including type I and type II diabetes. Stem Cells 2013;31:2396–2407

Identification of Alverine and Benfluorex as HNF4α Activators

The principal finding of this study is that two drugs, alverine and benfluorex, used in vastly different clinical settings, activated the nuclear receptor transcription factor HNF4α. Both were hits in a high-throughput screen for compounds that reversed the inhibitory effect of the fatty acid palmitate on human insulin promoter activity. Alverine is used in the treatment of irritable bowel syndrome, while benfluorex (Mediator) was used to treat hyperlipidemia and type II diabetes. Benfluorex was withdrawn from the market recently because of serious cardiovascular side effects related to fenfluramine-like activity. Strikingly, alverine and benfluorex have a previously unrecognized structural similarity, consistent with a common mechanism of action. Gene expression and biochemical studies revealed that they both activate HNF4α. This novel mechanism of action should lead to a reinterpretation of previous studies with these drugs and suggests a path toward the development of therapies for diseases such as inflammatory bowel and diabetes that may respond to HNF4α activators.

PAR2 regulates regeneration, transdifferentiation, and death.

Understanding the mechanisms by which cells sense and respond to injury is central to developing therapies to enhance tissue regeneration. Previously, we showed that pancreatic injury consisting of acinar cell damage+β-cell ablation led to islet cell transdifferentiation. Here, we report that the molecular mechanism for this requires activating protease-activated receptor-2 (PAR2), a G-protein-coupled receptor. PAR2 modulation was sufficient to induce islet cell transdifferentiation in the absence of β-cells. Its expression was modulated in an islet cell type-specific manner in murine and human type 1 diabetes (T1D). In addition to transdifferentiation, PAR2 regulated β-cell apoptosis in pancreatitis. PAR2s role in regeneration is broad, as mice lacking PAR2 had marked phenotypes in response to injury in the liver and in digit regeneration following amputation. These studies provide a pharmacologically relevant target to induce tissue regeneration in a number of diseases, including T1D.