Role of the REST/RE-1 system in beta-cell life, function and differentiation

SUMMARYDiabetes is characterized by insulin deficiency that results from the destruction of insulin-secreting pancreatic beta-cells (Type 1), or in part from beta-cell death and insulin secretion defects (Type 2). Therefore, understanding the mechanisms of beta cell neogenesis (to generate unlimited supply of beta cells for T1D transplantation] or identifying the specific genes that favors insulin secretion or beta-cell survival is of great importance for the management of diabetes. The transcriptional repressor RE-1 Silencing Transcription Factor (REST) restricts the expression of a large number of genes containing its binding element, called Repressor Element-1 (RE-1), to neurons and beta cells. To do so, REST is ubiquitously expressed but in neurons and beta cells. To identify these essential genes and their functional significance in beta cells, we have generated transgenic mice that express REST specifically in beta cells under the control of the rat insulin promoter (RIP-REST mice). This resulted in the repression of the RE-1- containing genes in beta cells, and we analyzed the consequences.We first showed that RIP-REST mice were glucose-intolerant because of a defective insulin secretion. To explain this defect, we identified that a subset of the REST target genes were necessary for insulin exocytosis, such as Snap25, Synaptotagmin (Syt) IX, Complexin II, and Ica512, and we further demonstrated that among the identified REST targets, Syt IV and VII were also involved in insulin release. We next analyzed a novel RIP-REST mouse line that featured diabetes and we showed that this defect was due to a major loss of beta-cell mass. To explain this phenotype, we identified REST target genes that were involved in beta-cell survival, such as Ibl, Irs2, Ica512 and Connexin36, and revealed that another REST target, Cdk5r2 is also involved in beta-cell protection. In a third part, we finally suggest that REST may be important for pancreatic endocrine differentiation, since transgenic mice expressing constitutive REST in pancreatic multipotent progenitors show impaired formation of Ngn3-expressing endocrine- committed precursors, and impaired formation of differentiated endocrine cells. Mapping the pattern of REST expression in wild type animals indicates that it is expressed in multipotent progenitors to become then excluded from endocrine cells. Preliminary results suggest that a downregulation of REST would result in relieved expression of at least the Mytl target, favoring subsequent acquisition of the endocrine competence by endocrine precursor cells.Thus, we propose that the REST/RE-1 system is an important feature for beta-cell neogenesis, function and survivalRESUMELe diabète se caractérise par une déficience en insuline qui résulte d'une destruction des cellules bêta (β) pancréatiques sécrétant l'insuline [Type 1], ou à un défaut de sécrétion d'insuline qui peut être associé à la mort des cellules β (Type 2). La compréhension des mécanismes de néogenèse des cellules β, ainsi que l'identification de gènes impliqués dans leur survie et dans le contrôle de la sécrétion d'insuline est donc importante pour le traitement du diabète. Le facteur de transcription de type répresseur, RE-1 Silencing Transcription Factor [REST], contribue à la spécificité d'expression dans les neurones et les cellules β, d'un grand nombre de gènes portant son motif de fixation, le Repressor Element-1 (RE-1). Pour cela, REST est exprimé dans toutes les cellules, sauf dans les neurones et les cellules β. Afin d'identifier les gènes cibles de REST ainsi que leur fonction au sein de la cellule β, nous avons généré des souris transgéniques qui expriment REST spécifiquement dans ces cellules, sous la dépendance du promoteur de l'insuline (souris RIP-REST]. Cette expression ectopique de REST a permis de diminuer l'expression des gènes contrôlés par REST, et d'en analyser les conséquences. Nous avons montré que les souris RIP-REST étaient intolérantes au glucose et que ceci était du à un défaut de sécrétion d'insuline. Pour expliquer ce phénotype, nous avons mis en évidence le fait que des gènes cibles de REST codent pour des protéines importantes pour l'exocytose de l'insuline, comme SNAP25, Synaptotagmin (Syt) IX, Complexin II ou ICA512. De plus, nous avons découvert deux nouvelles cibles de REST impliquées dans la sécrétion d'insuline, Syt IV et Syt VII. Par la suite, nous avons démontré qu'une nouvelle lignée de souris RIP-REST étaient atteintes d'un diabète sévère à cause d'une perte massive des cellules β. La disparition de ces cellules a été expliquée par l'identification de gènes cibles de REST impliqués dans la survie des cellules β, comme Ibl, Irs2, Ica512 ou la Connexine36. De plus, nous avons découvert qu'une nouvelle cible, Cdk5r2, était aussi impliquée dans la survie des cellules β. Dans une dernière partie, nous suggérons, grâce à l'analyse de nouvelles souris transgéniques exprimant constitutivement REST dans les cellules progénitrices du pancréas embryonnaire, que REST empêche la formation des précurseurs de cellules endocrines ainsi que la différenciation de ces cellules. L'analyse de l'expression de REST au cours du développement embryonnaire du pancréas indique que la diminution de l'expression de REST conduit en partie, à l'induction d'un de ses gènes cible Mytl, qui favorise la formation de précurseurs endocrines. Nous proposons donc que le système REST/RE-1 est important pour la génération, la fonction et la survie des cellules β.

Activation of peroxisome proliferator-activated receptor beta/delta inhibits lipopolysaccharide-induced cytokine production in adipocytes by lowering nuclear factor-kappaB activity via extracellular signal-related kinase 1/2.

OBJECTIVE: Chronic activation of the nuclear factor-kappaB (NF-kappaB) in white adipose tissue leads to increased production of pro-inflammatory cytokines, which are involved in the development of insulin resistance. It is presently unknown whether peroxisome proliferator-activated receptor (PPAR) beta/delta activation prevents inflammation in adipocytes. RESEARCH DESIGN AND METHODS AND RESULTS: First, we examined whether the PPARbeta/delta agonist GW501516 prevents lipopolysaccharide (LPS)-induced cytokine production in differentiated 3T3-L1 adipocytes. Treatment with GW501516 blocked LPS-induced IL-6 expression and secretion by adipocytes and the subsequent activation of the signal transducer and activator of transcription 3 (STAT3)-Suppressor of cytokine signaling 3 (SOCS3) pathway. This effect was associated with the capacity of GW501516 to impede LPS-induced NF-kappaB activation. Second, in in vivo studies, white adipose tissue from Zucker diabetic fatty (ZDF) rats, compared with that of lean rats, showed reduced PPARbeta/delta expression and PPAR DNA-binding activity, which was accompanied by enhanced IL-6 expression and NF-kappaB DNA-binding activity. Furthermore, IL-6 expression and NF-kappaB DNA-binding activity was higher in white adipose tissue from PPARbeta/delta-null mice than in wild-type mice. Because mitogen-activated protein kinase-extracellular signal-related kinase (ERK)1/2 (MEK1/2) is involved in LPS-induced NF-kappaB activation in adipocytes, we explored whether PPARbeta/delta prevented NF-kappaB activation by inhibiting this pathway. Interestingly, GW501516 prevented ERK1/2 phosphorylation by LPS. Furthermore, white adipose tissue from animal showing constitutively increased NF-kappaB activity, such as ZDF rats and PPARbeta/delta-null mice, also showed enhanced phospho-ERK1/2 levels. CONCLUSIONS: These findings indicate that activation of PPARbeta/delta inhibits enhanced cytokine production in adipocytes by preventing NF-kappaB activation via ERK1/2, an effect that may help prevent insulin resistance.

Competition between beta-subunits of cardiac L-type calcium channels

Cardiac L-type Ca (CaV1.2) channels are composed of a pore forming CaV1.2-α1 subunit and auxiliary β- and α2δ-subunits. β-subunits are important not only for surface expression of the channel pore but also for modulation of channel gating properties. Different β-subunits differentially modulate channel activity (Hullin et al., PLOSone, 2007) and thus L-type Ca2+ channel gating is altered when β-subunit expression pattern is changed. In human heart failure increased activity of single ventricular L-type Ca2+-channels is associated with an increased expression of β2-subunits. Interestingly, induction of β2-subunit over-expression in hearts of transgenic mice resembled this heart failure phenotype of hyperactive single L-type Ca2+-channel channels (Beetz et al., Cardiovasc Res. 2009). We hypothesised that competition of less stimulating β-subunits (e.g. β1) with β-subunits causing strong channel stimulation (e.g. β2) might be a means to treat dysfunctional L-type Ca2+-channel activity. To test this hypothesis, we performed whole-cell and single-channel measurements employing recombinant CaV1.2 channels expressed in HEK293 cells together with both β- and β1a2b-subunits. Whole-cell analysis revealed no differences of maximum L-type Ca2+-current densities [pA/pF] with coexpression of either β1a-subunits (-52±3.8), β2b-subunits (-61.5±6.6) or the mixtures of β- and β1a2b-subunits with the plasmid transfection ratio of 2:1 (-60.2±1.6) and 1:1 (-56.7±2.6) respectively. However, steady state inactivation kinetics differed between particular β-subunit and the relative amount of β-subunit presence in the mixtures (β1a1a-subunit (-41.1±1.0), β2b-subunits (-35.1±1.1), mixture 2:1 (-40.3±1.5), and mixture 1:1 (-38.4±2.0); [mV]; p<0.05, students t-test). Using a novel single-channel analysis, switching of gating between β1-like and β2-like modes was monitored on a minute time-scale when both β-subunits were co-expressed in the same cells, but the larger amount of β1a-subunits is required for the effective switching of gating. Our results indicate a model of mutually exclusive binding and effective competition between several β-subunits suggesting that hyperactive channel gating mediated e.g. by β2-subunits can be normalized by β1-subunits. Therefore, competitive replacement between different L-type Ca2+-channel β-subunits might serve as a novel therapeutic strategy for e.g. heart failure.

Vaccination with a recombinant protein encoding the tumor-specific antigen NY-ESO-1 elicits an A2/157-165-specific CTL repertoire structurally distinct and of reduced tumor reactivity than that elicited by spontaneous immune responses to NY-ESO-1-expressing Tumors.

In a recent vaccination trial assessing the immunogenicity of an NY-ESO-1 (ESO) recombinant protein administered with Montanide and CpG, we have obtained evidence that this vaccine induces specific cytolytic T lymphocytes (CTL) in half of the patients. Most vaccine-induced CTLs were directed against epitopes located in the central part of the protein, between amino acids 81 and 110. This immunodominant region, however, is distinct from another ESO CTL region, 157-165, that is a frequent target of spontaneous CTL responses in A2+ patients bearing ESO tumors. In this study, we have investigated the CTL responses to ESO 157-165 in A2+ patients vaccinated with the recombinant protein. Our data indicate that after vaccination with the protein, CTL responses to ESO 157-165 are induced in some, but not all, A2+ patients. ESO 157-165-specific CTLs induced by vaccination with the ESO protein were functionally heterogeneous in terms of tumor recognition and often displayed decreased tumor reactivity as compared with ESO 157-165-specific CTLs isolated from patients with spontaneous immune responses to ESO. Remarkably, protein-induced CTLs used T-cell receptors similar to those previously isolated from patients vaccinated with synthetic ESO peptides (Vbeta4.1) and distinct from those used by highly tumor-reactive CTLs isolated from patients with spontaneous immune responses (Vbeta1.1, Vbeta8.1, and Vbeta13.1). Together, these results demonstrate that vaccination with the ESO protein elicits a repertoire of ESO 157-165-specific CTLs bearing T-cell receptors that are structurally distinct from those of CTLs found in spontaneous immune responses to the antigen and that are heterogeneous in terms of tumor reactivity, being often poorly tumor reactive.

Aggregation of integrins and RhoA activation are required for Thy-1-induced morphological changes in astrocytes.

Thy-1, a cell adhesion molecule abundantly expressed in mammalian neurons, binds to a beta(3)-containing integrin on astrocytes and thereby stimulates the assembly of focal adhesions and stress fibers. Such events lead to morphological changes in astrocytes that resemble those occurring upon injury in the brain. Extracellular matrix proteins, typical integrin ligands, bind to integrins and promote receptor clustering as well as signal transduction events that involve small G proteins and cytoskeletal changes. Here we investigated the possibility that the cell surface protein Thy-1, when interacting with a beta(3)-containing integrin on astrocytes, could trigger signaling events similar to those generated by extracellular matrix proteins. DI-TNC(1) astrocytes were stimulated with Thy-1-Fc immobilized on beads, and increased RhoA activity was confirmed using an affinity precipitation assay. The effect of various inhibitors on the cellular response was also studied. The presence of Y-27632, an inhibitor of Rho kinase (p160ROCK), a key downstream effector of RhoA, significantly reduced focal adhesion and stress fiber formation induced by Thy-1. Similar effects were obtained when astrocytes were treated with C3 transferase, an inhibitor of RhoA. Alternatively, astrocytes were transfected with an expression vector encoding fusion proteins of enhanced green fluorescent protein with either the Rho-binding domain of Rhotekin, which blocks RhoA function, or the dominant-negative N19RhoA mutant. In both cases, Thy-1-induced focal adhesion formation was inhibited. Furthermore, we observed that RhoA activity after stimulation with soluble Thy-1-Fc molecule was augmented upon further cross-linking using protein A-Sepharose beads. The same was shown by cross-linking beta(3)-containing integrin with anti-beta(3) antibodies. Together, these results indicate that Thy-1-mediated astrocyte stimulation depended on beta(3) integrin clustering and the resulting increase in RhoA activity.

Beta-adrenoceptor stimulation increases neuropeptide Y release from sympathetic nerves in intact rats.

This study was undertaken to assess in conscious normotensive rats the effects of beta-adrenoceptor stimulation on plasma neuropeptide Y (NPY) levels. Wistar rats were subjected to adrenal demedullation on the right side and were either adrenalectomized or sham-operated on the left side. Eleven days later, the conscious rats were infused i.v. for 30 min with either isoproterenol (10 ng/min) or its vehicle. Plasma NPY levels were significantly lower (23.8 +/- 2.6 pM, means +/- S.E.M., n = 12, P < 0.01) in vehicle-treated medullectomized rats than in corresponding sham-operated controls (36.7 +/- 4.1 pM, n = 12). The medullectomized rats infused with isoproterenol showed plasma NPY levels (36.7 +/- 3.3 pM, n = 11) comparable to those of sham-operated rats having received the vehicle. These data therefore demonstrate that plasma NPY levels are lower in rats without adrenal medulla and that in these animals isoproterenol increases NPY release, most likely by activating pre-synaptic beta-adrenoceptors.

Interleukins (IL)-1 and IL-2 control IL-2 receptor alpha and beta expression in immature thymocytes.

Functional high-affinity interleukin-2 receptors (IL-2R) contain three transmembrane proteins, IL-2R alpha, beta and gamma. We have investigated the expression of IL-2R alpha and beta genes in immature mouse thymocytes. Previous work has shown that during differentiation these cells transiently express IL-2R alpha on their surface. Stimulation of IL-2R alpha+ and IL-2R alpha- immature thymocytes with phorbol 12-myristate 13-acetate and calcium ionophore induces synthesis of IL-2R alpha and IL-2R beta mRNA. Most of this response depends on autocrine stimulation by IL-2. IL-1 synergizes with IL-2 to induce a 120-fold increase in IL-2R alpha mRNA and a 14-fold increase in IL-2R beta mRNA levels. A large proportion of the stimulated cells contains both transcripts. These interleukins do not induce any differentiation to more mature phenotypes. Collectively, these results show that IL-2 plays a major role in the regulation of IL-2R expression in normal immature thymocyte. We suggest that this response to interleukins may be part of a homeostatic mechanism to increase the production of immature thymocytes during stress.

Silencing of both beta-TrCP1 and HOS (beta-TrCP2) is required to suppress human immunodeficiency virus type 1 Vpu-mediated CD4 down-modulation.

The human immunodeficiency virus type 1 (HIV-1) Vpu protein interacts with CD4 within the endoplasmic reticula of infected cells and targets CD4 for degradation through interaction with beta-TrCP1. Mammals possess a homologue of beta-TrCP1, HOS, which is also named beta-TrCP2. We show by coimmunoprecipitation experiments that beta-TrCP2 binds Vpu and is able to induce CD4 down-modulation as efficiently as beta-TrCP1. In two different cell lines, HeLa CD4+ and Jurkat, Vpu-mediated CD4 down-modulation could not be reversed through the individual silencing of endogenous beta-TrCP1 or beta-TrCP2 but instead required the two genes to be silenced simultaneously.

Signal transduction by the cloned glucagon-like peptide-1 receptor: comparison with signaling by the endogenous receptors of beta cell lines.

Glucagon-like peptide-1 (GLP-1) is a gastrointestinal hormone that potentiates glucose-induced insulin secretion by pancreatic beta cells. The mechanisms of interaction between GLP-1 and glucose signaling pathways are not well understood. Here we studied the coupling of the cloned GLP-1 receptor, expressed in fibroblasts or in COS cells, to intracellular second messengers and compared this signaling with that of the endogenous receptor expressed in insulinoma cell lines. Binding of GLP-1 to the cloned receptor stimulated formation of cAMP with the same dose dependence and similar kinetics, compared with the endogenous receptor of insulinoma cells. Compared with forskolin-induced cAMP accumulation, that induced by GLP-1 proceeded with the same initial kinetics but rapidly reached a plateau, suggesting fast desensitization of the receptor. Coupling to the phospholipase C pathway was assessed by measuring inositol phosphate production and variations in the intracellular calcium concentration. No GLP-1-induced production of inositol phosphates could be measured in the different cell types studied. A rise in the intracellular calcium concentration was nevertheless observed in transfected COS cells but was much smaller than that observed in response to norepinephrine in cells also expressing the alpha 1B-adrenergic receptor. Importantly, no such increase in the intracellular calcium concentration could be observed in transfected fibroblasts or insulinoma cells, which, however, responded well to thrombin or carbachol, respectively. Together, our data show that interaction between GLP-1 and glucose signaling pathways in beta cells may be mediated uniquely by an increase in the intracellular cAMP concentration, with the consequent activation of protein kinase A and phosphorylation of elements of the glucose-sensing apparatus or of the insulin granule exocytic machinery.

Exendin-4 protects beta-cells from interleukin-1 beta-induced apoptosis by interfering with the c-Jun NH2-terminal kinase pathway.

OBJECTIVE: The pro-inflammatory cytokine interleukin-1 beta (IL-1 beta) generates pancreatic beta-cells apoptosis mainly through activation of the c-Jun NH(2)-terminal kinase (JNK) pathway. This study was designed to investigate whether the long-acting agonist of the hormone glucagon-like peptide 1 (GLP-1) receptor exendin-4 (ex-4), which mediates protective effects against cytokine-induced beta-cell apoptosis, could interfere with the JNK pathway. RESEARCH DESIGN AND METHODS: Isolated human, rat, and mouse islets and the rat insulin-secreting INS-1E cells were incubated with ex-4 in the presence or absence of IL-1 beta. JNK activity was assessed by solid-phase JNK kinase assay and quantification of c-Jun expression. Cell apoptosis was determined by scoring cells displaying pycnotic nuclei. RESULTS: Ex-4 inhibited induction of the JNK pathway elicited by IL-1 beta. This effect was mimicked with the use of cAMP-raising agents isobutylmethylxanthine and forskolin and required activation of the protein kinase A. Inhibition of the JNK pathway by ex-4 or IBMX and forskolin was concomitant with a rise in the levels of islet-brain 1 (IB1), a potent blocker of the stress-induced JNK pathway. In fact, ex-4 as well as IBMX and forskolin induced expression of IB1 at the promoter level through cAMP response element binding transcription factor 1. Suppression of IB1 levels with the use of RNA interference strategy impaired the protective effects of ex-4 against apoptosis induced by IL-1 beta. CONCLUSIONS: The data establish the requirement of IB1 in the protective action of ex-4 against apoptosis elicited by IL-1 beta and highlight the GLP-1 mimetics as new potent inhibitors of the JNK signaling induced by cytokines.

Regulation of the transforming growth factor beta-responsive transcription factor CTF-1 by calcineurin and calcium/calmodulin-dependent protein kinase IV.

Transforming growth factor beta (TGF-beta) is a pluripotent peptide hormone that regulates various cellular activities, including growth, differentiation, and extracellular matrix protein gene expression. We previously showed that TGF-beta induces the transcriptional activation domain (TAD) of CTF-1, the prototypic member of the CTF/NF-I family of transcription factors. This induction correlates with the proposed role of CTF/NF-I binding sites in collagen gene induction by TGF-beta. However, the mechanisms of TGF-beta signal transduction remain poorly understood. Here, we analyzed the role of free calcium signaling in the induction of CTF-1 transcriptional activity by TGF-beta. We found that TGF-beta stimulates calcium influx and mediates an increase of the cytoplasmic calcium concentration in NIH3T3 cells. TGF-beta induction of CTF-1 is inhibited in cells pretreated with thapsigargin, which depletes the endoplasmic reticulum calcium stores, thus further arguing for the potential relevance of calcium mobilization in TGF-beta action. Consistent with this possibility, expression of a constitutively active form of the calcium/calmodulin-dependent phosphatase calcineurin or of the calcium/calmodulin-dependent kinase IV (DeltaCaMKIV) specifically induces the CTF-1 TAD and the endogenous mouse CTF/NF-I proteins. Both calcineurin- and DeltaCaMKIV-mediated induction require the previously identified TGF-beta-responsive domain of CTF-1. The immunosuppressants cyclosporin A and FK506 abolish calcineurin-mediated induction of CTF-1 activity. However, TGF-beta still induces the CTF-1 TAD in cells treated with these compounds or in cells overexpressing both calcineurin and DeltaCaMKIV, suggesting that other calcium-sensitive enzymes might mediate TGF-beta action. These results identify CTF/NF-I as a novel calcium signaling pathway-responsive transcription factor and further suggest multiple molecular mechanisms for the induction of CTF/NF-I transcriptional activity by growth factors.

Gene transfer of a soluble IL-1 type 2 receptor-Ig fusion protein improves cardiac allograft survival in rats.

OBJECTIVE: Interleukin-1 (IL-1) mediates ischemia-reperfusion injury and graft inflammation after heart transplantation. IL-1 affects target cells through two distinct types of transmembrane receptors, type-1 receptor (IL-1R1), which transduces the signal, and the non-signaling type-2 receptor (IL-1R2), which acts as a ligand sink that subtracts IL-1beta from IL-1R1. We analyzed the efficacy of adenovirus (Ad)-mediated gene transfer of a soluble IL-1R2-Ig fusion protein in delaying cardiac allograft rejection and the mechanisms underlying the protective effect. METHODS: IL-1 inhibition by IL-1R2-Ig was tested using an in vitro functional assay whereby endothelial cells preincubated with AdIL-1R2-Ig or control virus were stimulated with recombinant IL-1beta or tumor necrosis factor-alpha (TNF-alpha), and urokinase-type plasminogen activator (u-PA) induction was measured by zymography. AdIL-1R2-Ig was delivered to F344 rat donor hearts ex vivo, which were placed in the abdominal position in LEW hosts. Intragraft inflammatory cell infiltrates and proinflammatory cytokine expression were analyzed by immunohistochemistry and real-time reverse transcriptase-polymerase chain reaction (RT-PCR), respectively. RESULTS: IL-1R2-Ig specifically inhibited IL-1beta-induced u-PA responses in vitro. IL-1R2-Ig gene transfer reduced intragraft monocytes/macrophages and CD4(+) cell infiltrates (p<0.05), TNF-alpha and transforming growth factor-beta (TGF-beta) expression (p<0.05), and prolonged graft survival (15.6+/-5.7 vs 10.3+/-2.5 days with control vector and 10.1+/-2.1 days with buffer alone; p<0.01). AdIL-1R2-Ig combined with a subtherapeutic regimen of cyclosporin A (CsA) was superior to CsA alone (19.4+/-3.0 vs 15.9+/-1.8 days; p<0.05). CONCLUSIONS: Soluble IL-1 type-2 receptor gene transfer attenuates cardiac allograft rejection in a rat model. IL-1 inhibition may be useful as an adjuvant therapy in heart transplantation.