View metadata, citation and similar papers at core.ac.uk brought to you by CORE

provided by Elsevier - Publisher Connector Cell Metabolism Previews

Fischer, C., Mazzone, M., Jonckx, B., and Carme- Same´ n, E., Lu, L., et al. (2012). Nature 490, Takamoto, I., Sasako, T., et al. (2011). Cell Metab. liet, P. (2008). Nat. Rev. Cancer 8, 942–956. 426–430. 13, 294–307.

Hagberg, C.E., Falkevall, A., Wang, X., Larsson, E., Karpanen, T., Bry, M., Ollila, H.M., Seppa¨ nen- Poesen, K., Lambrechts, D., Van Damme, P., Huusko, J., Nilsson, I., van Meeteren, L.A., Samen, Laakso, T., Liimatta, E., Leskinen, H., Kivela¨ , R., Dhondt, J., Bender, F., Frank, N., Bogaert, E., E., Lu, L., Vanwildemeersch, M., et al. (2010). Helkamaa, T., Merentie, M., Jeltsch, M., et al. Claes, B., Heylen, L., Verheyen, A., et al. (2008). Nature 464, 917–921. (2008). Circ. Res. 103, 1018–1026. J. Neurosci. 28, 10451–10459.

Hagberg, C.E., Mehlem, A., Falkevall, A., Muhl, L., Kubota, T., Kubota, N., Kumagai, H., Yamaguchi, Samuel, V.T., and Shulman, G.I. (2012). Cell 148, Fam, B.C., Ortsa¨ ter, H., Scotney, P., Nyqvist, D., S., Kozono, H., Takahashi, T., Inoue, M., Itoh, S., 852–871.

Mining in Type 2 Diabetic Islets and Finding Gold

Decio L. Eizirik1,* and Miriam Cnop1,2 1Laboratory of Experimental Medicine, Medical Faculty 2Division of Endocrinology, Erasmus Hospital Universite Libre de Bruxelles (ULB), 1000 Brussels, Belgium *Correspondence: [email protected] http://dx.doi.org/10.1016/j.cmet.2012.10.012

Pancreatic b cell failure is central in the pathogenesis of type 2 diabetes (T2D), but the mechanisms involved remain unclear. Mahdi and colleagues (2012) couple global evaluation of expression with coexpression network analysis of human islets from T2D patients to identify SFRP4 as an early mediator of b cell dysfunc- tion in T2D.

Genome-wide association studies for show that secreted frizzled-related pro- that this metabolic ‘‘T2D-like’’ stress T2D have so far identified 65 susceptibility tein 4 (SFRP4) is highly associated induces a mild inflammatory response, loci for the disease (Morris et al., 2012), with T2D. representing around 5%–10% of the pro- but together these loci account for less Previous work already suggested that inflammatory response of human islets than 10% of the variance in disease IL-1b plays a role in b cell dysfunction exposed to ‘‘T1D-like’’ conditions (Cnop susceptibility. This contrasts with the situ- and death in T1D and T2D. However, et al., 2005; Igoillo-Esteve et al., 2010). ation in type 1 diabetes (T1D), where such whereas a role for inflammation in b cell The role for this ‘‘low-intensity’’ innate studies have identified 50 loci across the loss is well established in the context of immunity-mediated inflammation in b cell associated with T1D that T1D (Eizirik et al., 2009), it has remained dysfunction and death in T2D remains explain nearly 80% of the heritability controversial for T2D (Cnop et al., 2005; unclear. For instance, an IL-1 receptor (Pociot et al., 2010). Other approaches Donath et al., 2008). To explain local antagonist blocked palmitate-induced to identify basic mechanisms of disease IL-1b production in the islets of T2D indi- chemokine expression, but failed to pre- are therefore needed (Taneera et al., viduals, one model proposed that IL-1b vent (Igoillo-Esteve et al., 2010). 2012). Furthermore, direct studies of the production was induced by glucose, Using gene expression topology with diseased human tissue—in the case of leading to upregulation of the apoptotic weighted gene coexpression network T2D, human islets are the gold standard Fas receptor and ligand and b cell analysis (in which coexpressed genes given the central role of b cell dysfunc- ‘‘suicide’’ (Donath et al., 2008), but this are clustered into gene modules based tion in its pathogenesis—are essential to was not confirmed by other groups (re- on their connectivity), Mahdi et al. now further our understanding of human dia- viewed in Cnop et al., 2005). In recent identify a T2D-related gene module en- betes (Cnop et al., 2005; Kahn, 2003). In years a more nuanced view of inflamma- riched for IL-1-related genes. Among the this issue of Cell Metabolism, Mahdi and tion in T2D islets emerged, consisting most connected hub genes, the authors colleagues tackle these challenges by of mild upregulation of cytokines and identify SFRP4 as highly associated with performing microarray analyses of human chemokines in islets from T2D patients, T2D, HbA1c (a measure of average islets isolated from T2D and normoglyce- possibly mediated by increased circu- glucose levels over the past 2 months), mic individuals (Mahdi et al., 2012)(Fig- lating concentrations of the free fatty and insulin secretion (Figure 1). Subse- ure 1). They identify a group of T2D- acid palmitate that induces islet IL-1b quent functional studies then show that associated genes related to interleukin-1 and TNF-a expression (Igoillo-Esteve the SFRP4 is induced by the cyto- (IL-1), a proinflammatory cytokine, and et al., 2010). Experimental findings show kine IL-1b. Furthermore, the authors show

Cell Metabolism 16, November 7, 2012 ª2012 Elsevier Inc. 555 Cell Metabolism Previews

the elevated levels of SFRP4 act in an auto/paracrine manner to inhibit b cell function? In clonal rat b cells, the inhibitory effect of IL-1b on insulin secretion was slightly attenuated following SFRP4 silencing, suggesting that SFRP4 only partially contributes to this inhibitory effect. In human islets, however, IL-1b alone does not inhibit insulin secretion (reviewed in Cnop et al., 2005). Are there other cytokines, besides IL-1b, that affect insulin release in T2D, and is this via SFRP4 or other mechanisms that remain to be discovered? Of note, pharmaceu- tical industry trials with different IL-1 receptor antagonists in T2D have resulted in only small improvements in glycemic control, and most have been discon- tinued. Mahdi et al. suggest crosstalk between SFRP4 and Wnt signaling, but additional studies are required to clarify the underlying gene networks in pancre- Figure 1. The Research Strategy of Mahdi and Colleagues Leading to the Identification of atic b cells. Since SFRP4 plays a role in SFRP4 Based on bioinformatics analyses of microarray studies of human islets from type 2 diabetic and nondia- differentiation in other cell types and betic organ donors, Mahdi et al. identify SFRP4 as a hub gene in the gene expression module associated inhibits proliferation and angiogenesis, it with type 2 diabetes and insulin secretory defects. Functional studies in in vitro and in vivo models then would be of interest to determine whether shed light on the mechanisms by which SFRP4 causes pancreatic b cell dysfunction. Clinical studies in prospective cohorts identified SFRP4 as a potential biomarker for future development of type 2 diabetes. these mechanisms operate in T2D. T2D, type 2 diabetes; ND, nondiabetic. In our opinion, the present study is a beautiful example of the potential that SFRP4 impairs insulin release both mass in T2D is a consequence of meta- power of hypothesis-generating ‘‘mining’’ in vitro in mouse and human islets and bolic stress induced by chronic exposure studies. If well powered and controlled in vivo in SFRP4-treated mice. It does to high glucose and saturated free (including, in this case, replication in inde- not, however, decrease b cell viability. In fatty acids, independently of inflammation pendent samples), based on relevant mechanistic studies, the authors show (Cnop et al., 2005). tissue (i.e., human islets from diseased that SFRP4 inhibits expression of L-type In his novel The Island of Doctor individuals), and analyzed by adequate and P/Q-type Ca2+ channels, possibly Moreau, H.G. Wells provides an interest- bioinformatics tools, they can provide via Wnt signaling, and thereby hampers ing description of the scientific method: a comprehensive view of complex bio- Ca2+ influx, consequently reducing insulin ‘‘I asked a question, devised some logical problems and point to unexpected exocytosis. Importantly, serum SFRP4 methods of getting an answer, and got— pathways to understand disease mecha- levels are elevated in T2D individuals a fresh question.’’ Indeed, the finding of nisms. In other words, these mining and inversely correlated with the disposi- a pathophysiological role for SFRP4 in studies may find gold. tion index (a measure of b cell function). T2D raises many questions. SFRP4 is SFRP4 levels are increased several years subject to alternative splicing, and its REFERENCES before diabetes diagnosis, a key observa- expression is regulated by DNA methyla- tion confirmed in an independent cohort tion of the promoter region. A recent study Cnop, M., Welsh, N., Jonas, J.C., Jo¨ rns, A., (Mahdi et al., 2012). As a whole, these of the DNA methylation profile of T2D Lenzen, S., and Eizirik, D.L. (2005). Diabetes findings suggest that SFRP4 may be human islets identified hypomethylation 54(Suppl 2 ), S97–S107. a potential mediator and early biomarker across more than 200 gene promoters Donath, M.Y., Størling, J., Berchtold, L.A., of b cell dysfunction in T2D. (Volkmar et al., 2012); whether SFRP4 Billestrup, N., and Mandrup-Poulsen, T. (2008). Endocr. Rev. 29, 334–350. These results provide a possible molec- expression in T2D islets is induced via ular link between mild islet inflammation this epigenetic modification remains to Eizirik, D.L., Colli, M.L., and Ortis, F. (2009). Nat. and defective insulin secretion in T2D. be examined. Does the increase in Rev. Endocrinol. 5, 219–226. Taken together with previous findings SFRP4 expression in T2D islets contribute Igoillo-Esteve, M., Marselli, L., Cunha, D.A., (Igoillo-Esteve et al., 2010), the observa- to the increase in circulating levels of Ladrie` re, L., Ortis, F., Grieco, F.A., Dotta, F., tions suggest that the low-intensity islet SFRP4? This is unlikely, given the ubiqui- Weir, G.C., Marchetti, P., Eizirik, D.L., and Cnop, M. (2010). Diabetologia 53, 1395–1405. inflammation in T2D may have a more tous expression of SFRP4 and the rela- relevant role for b cell functional impair- tively low expression levels in human Kahn, S.E. (2003). Diabetologia 46, 3–19. b ment than actual cell death. It is there- islets. Do T2D human islets secrete more Mahdi, T., Ha¨ nzelmann, S., Salehi, A., Muhammed, fore conceivable that the loss of b cell SFRP4 relative to healthy islets, and do S.J., Reinbothe, T.M., Tang, Y., Axelsson, A.S.,

556 Cell Metabolism 16, November 7, 2012 ª2012 Elsevier Inc. Cell Metabolism Previews

Zhou, Y., Jing, X., Almgren, P., et al. (2012). Cell Pociot, F., Akolkar, B., Concannon, P., Erlich, H.A., Vikman, P., Hansson, O., et al. (2012). Cell Metab. Metab. 16, this issue, 625–633. Julier, C., Morahan, G., Nierras, C.R., Todd, J.A., 16, 122–134. Rich, S.S., and Nerup, J. (2010). Diabetes 59, Morris, A.P., Voight, B.F., Teslovich, T.M., Ferreira, 1561–1571. Volkmar, M., Dedeurwaerder, S., Cunha, D.A., T., Segre` , A.V., Steinthorsdottir, V., Strawbridge, Ndlovu, M.N., Defrance, M., Deplus, R., Calonne, R.J., Khan, H., Grallert, H., Mahajan, A., et al. Taneera, J., Lang, S., Sharma, A., Fadista, J., E., Volkmar, U., Igoillo-Esteve, M., Naamane, N., (2012). Nat. Genet. 44, 981–990. Zhou, Y., Ahlqvist, E., Jonsson, A., Lyssenko, V., et al. (2012). EMBO J. 31, 1405–1426.

RIPpingoffGABAReleaseinHypothalamicCircuits Causes Obesity

Sebastien G. Bouret1,2,* 1The Saban Research Institute, Neuroscience Program, Children’s Hospital Los Angeles, University of Southern California, Los Angeles, CA 90027, USA 2Inserm, Jean-Pierre Aubert Research Center, U837, University Lille 2, Lille, 59045, France *Correspondence: [email protected] http://dx.doi.org/10.1016/j.cmet.2012.10.014

Hypothalamic RIP-expressing neurons regulate energy balance, but the precise neural pathways and neurotransmitters mediating this effect remain undetermined. Kong et al. (2012) now demonstrate that RIP neurons regulate energy expenditure and BAT thermogenesis predominantly via a GABAergic arcuate- paraventricular-hindbrain pathway.

The hypothalamus is a critical regulator of Empirical experiments using physical terscapular temperature and UCP1 ex- energy balance and glucose homeo- lesions of specific hypothalamic nuclei pression, which may be partially respon- stasis. Over the past 10–15 years a great and, more recently, genetic and optoge- sible for the reduced energy expenditure deal of attention has been devoted netic studies have shown the importance observed in mice lacking GABA neuro- to hypothalamic neurons that produce of neurons within the ARH, the ventrome- transmission in RIP neurons. In contrast, agouti-related peptide (AgRP) and neu- dial nucleus (VMH), the dorsomedial blockade of glutamatergic release from rons that express proopiomelanocortin nucleus (DMH), the paraventricular RIP-expressing neurons did not affect (POMC) in part because of the importance nucleus (PVH), and the lateral hypotha- energy balance, indicating that GABA is of these neuronal populations in the inte- lamic area (LHA) in the regulation of the major neurotransmitter involved in gration of peripheral blood-born signals, feeding and body weight. Lowell and this process. including endocrine signals and nutrients colleagues used the cre-loxP approach One of the most important functions of (Williams and Elmquist, 2012; Sa´ nchez- to specifically delete GABAergic neuro- RIP neurons is to mediate leptin’s effect Lasheras et al., 2010; Gao and Horvath, transmission in rat insulin promoter on energy balance (Covey et al., 2006). 2007). However, the hypothalamus con- (RIP)-expressing neurons; this neuronal Kong et al. (2012) show that the ability of tains a variety of other neuronal systems, population is distinct from the POMC leptin to reduce body weight requires and their relative contributions to the re- and AgRP neurons and is present in the GABA release from RIP-containing neu- gulation of energy balance remains in- ARH, VMH, and DMH (Kong et al., 2012). rons. In the absence of GABA neurotrans- completely understood. In addition, a Mice lacking synaptic release of GABA mission, leptin does not reduce body better understanding of the complex from RIP-containing neurons (Rip-Cre; weight, likely due to an inability to neuronal network that regulates appetite Vgatflox/flox mice) display higher body increase energy expenditure and BAT and body weight is needed. In a recent weights and a marked increase in fat activity. RIP neurons are primarily present issue of Cell, Kong et al. (2012) mapped mass. The mutant mice also show a in the ARH, VMH, and DMH, and each of the synaptic interactions of a subpopula- greater sensitivity to weight gain on a these nuclei is a direct target for leptin tion of GABAergic neurons from the high-fat diet (HFD). However, the elevated (Patterson et al., 2011). To define the arcuate nucleus (ARH) with other hypo- body weight in chow and HFD-fed mice RIP neurons responsible for the bulk of thalamic and hindbrain pathways and could not be explained by changes in leptin action, Kong et al. (2012) examined demonstrated the importance of these food intake or locomotor activity, but leptin-induced pSTAT3, a marker of leptin neural circuits in the regulation of energy Kong et al. found a marked decrease in receptor activation, in Rip-Cre; lox-GFP expenditure and brown adipose tissue oxygen consumption and altered BAT mice and found cells coexpressing (BAT) activity (Figure 1). thermogenesis, revealed by reduced in- pSTAT3 and Cre mainly in the ARH and

Cell Metabolism 16, November 7, 2012 ª2012 Elsevier Inc. 557