Article
Reference
Pancreatic α Cell-Derived Glucagon-Related Peptides Are Required for β Cell Adaptation and Glucose Homeostasis
TRAUB, Shuyang, et al.
TRAUB, Shuyang, et al . Pancreatic α Cell-Derived Glucagon-Related Peptides Are Required for β Cell Adaptation and Glucose Homeostasis. Cell Reports , 2017, vol. 18, no. 13, p. 3192-3203
DOI : 10.1016/j.celrep.2017.03.005 PMID : 28355570
Available at:
http://archive-ouverte.unige.ch/unige:93267
Disclaimer: layout of this document may differ from the published version.
1 / 1
Article
Pancreatic a Cell-Derived Glucagon-Related
Peptides Are Required for b Cell Adaptation and Glucose Homeostasis
Graphical Abstract
Highlights
d a
cell-derived glucagon-related peptides are critical for normal insulin secretion
d
Normally,
acell-derived GLP-1 is not required for sufficient insulin secretion
d
Paracrine GLP-1 in islets is needed for adaptation to aging and metabolic stress
d
Deficiency in
acell-derived GLP-1 can be rescued by inhibition of DPP-4
Authors
Shuyang Traub, Daniel T. Meier,
Friederike Schulze, ..., Pedro L. Herrera, Marianne Bo¨ni-Schnetzler,
Marc Y. Donath
Correspondence
marc.donath@usb.ch
In Brief
Pancreatic
acells may process
proglucagon to glucagon or GLP-1. Traub et al. find that
acell-derived GLP-1 is necessary for glucose homeostasis during aging and metabolic stress.
Deficiency of
acell-derived glucagon- related peptides can be compensated by DPP-4 inhibition.
Traub et al., 2017, Cell Reports18, 3192–3203 March 28, 2017ª2017 The Author(s).
http://dx.doi.org/10.1016/j.celrep.2017.03.005
Cell Reports
Article
Pancreatic a Cell-Derived Glucagon-Related Peptides Are Required for b Cell Adaptation and Glucose Homeostasis
Shuyang Traub,1,2,6Daniel T. Meier,1,2,6Friederike Schulze,1,2Erez Dror,1,2Thierry M. Nordmann,1,2Nicole Goetz,1,2 Norina Koch,1,2Elise Dalmas,1,2Marc Stawiski,1,2Valmir Makshana,1,2Fabrizio Thorel,3,4,5Pedro L. Herrera,3,4,5 Marianne Bo¨ni-Schnetzler,1,2and Marc Y. Donath1,2,7,*
1Endocrinology, Diabetes, and Metabolism, University Hospital Basel, 4031 Basel, Switzerland
2Department of Biomedicine, University of Basel, 4031 Basel, Switzerland
3Department of Genetic Medicine and Development, Faculty of Medicine, University of Geneva, 1211 Geneva, Switzerland
4Institute of Genetics and Genomics in Geneva (iGE3), University of Geneva, 1211 Geneva, Switzerland
5Centre facultaire du diabe`te, University of Geneva, 1211 Geneva, Switzerland
6Co-first author
7Lead Contact
*Correspondence:marc.donath@usb.ch http://dx.doi.org/10.1016/j.celrep.2017.03.005
SUMMARY
Pancreatic
acells may process proglucagon not only to glucagon but also to glucagon-like peptide-1 (GLP-1). However, the biological relevance of para- crine GLP-1 for
bcell function remains unclear. We studied effects of locally derived insulin secreta- gogues on
bcell function and glucose homeostasis using mice with
acell ablation and with
acell-spe- cific GLP-1 deficiency. Normally, intestinal GLP-1 compensates for the lack of
acell-derived GLP-1.
However, upon aging and metabolic stress, glucose tolerance is impaired. This was partly rescued with the DPP-4 inhibitor sitagliptin, but not with glucagon administration. In isolated islets from these mice, glucose-stimulated insulin secretion was heavily impaired and exogenous GLP-1 or glucagon rescued insulin secretion. These data highlight the impor- tance of
acell-derived GLP-1 for glucose homeosta- sis during metabolic stress and may impact on the clinical use of systemic GLP-1 agonists versus stabi- lizing local
acell-derived GLP-1 by DPP-4 inhibitors in type 2 diabetes.
INTRODUCTION
The classical incretin concept views GLP-1 as a hormone pro- duced in the intestinal L cells and acting via the circulation on satiety in the brain, gut motility, and insulin and glucagon secre- tion in the pancreatic islet (Campbell and Drucker, 2013;
Drucker, 2006). However, in contrast to typical hormones, GLP-1 has a very short half-life of less than 2 min (Deacon et al., 1995), and plasma levels are low. Furthermore, GLP-1 is rapidly inactivated by the dipeptidyl peptidase-4 (DPP-4) in the vicinity of L cells within less than a minute after being secreted
(Deacon, 2004; Hansen et al., 1999). This rapid degradation of GLP-1 raises questions about how its effects are mediated on distant target organs such as pancreaticbcells (Donath and Bur- celin, 2013). It was shown that GLP-1 induces its metabolic ac- tions by interacting with its receptor in extra-pancreatic locations such as the gut to activate the submucosal and myenteric ner- vous plexi (Waget et al., 2011; Washington et al., 2010) as well as the brain, which then signals to peripheral tissues (Baraboi et al., 2011; Burcelin et al., 2001; Cabou and Burcelin, 2011; Na- kagawa et al., 2004; Shimizu et al., 1983). Furthermore, small doses of DPP-4 inhibitors improve glucose tolerance without increasing the blood concentration of GLP-1 in a GLP-1 recep- tor-dependent manner (Waget et al., 2011). Other observations suggest that circulating GLP-1 can also directly stimulate insulin secretion from pancreatic b cells without involvement of neuronal circuits (Lamont et al., 2012; Smith et al., 2014).
In intestinal L cells, GLP-1 is cleaved from its precursor proglu- cagon by the enzyme prohormone convertase 1/3 (PC1/3). In contrast, pancreaticacells are viewed to process proglucagon to glucagon by PC2 (Furuta et al., 2001; Rouille´ et al., 1995;
Tucker et al., 1996), and it was believed that adultacells do not produce GLP-1. However, adenovirus-mediated expression of PC1/3 in a cells may increase islet GLP-1 expression and improve insulin secretion (Wideman et al., 2006). Furthermore, several studies have shown that islet-derived GLP-1 can be stimulated by an increased demand of insulin secretion, sug- gesting that locally produced GLP-1 may play a role in long- termbcell adaptation (Ellingsgaard et al., 2011; Hansen et al., 2011; Kilimnik et al., 2010; Marchetti et al., 2012; Nie et al., 2000; Thyssen et al., 2006). Finally, we have shown thatacell hy- perplasia occurs early in response to high-fat diet (HFD) feeding and precedes expansion ofbcell mass required forbcell adap- tation in response to increased secretory demand (Ellingsgaard et al., 2008, 2011). Therefore, the ability ofacells to produce GLP-1 is well documented.
However, the biological relevance of localacell-derived GLP-1 on glucose homeostasis and b cell function in physiological
3192 Cell Reports18, 3192–3203, March 28, 2017ª2017 The Author(s).
This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/).
settings and under metabolic stress remains to be demonstrated.
For this purpose, we used a genetic mouse model of induciblea cell ablation and generated ana cell-specific PC1/3-deficient mouse to prohibit islet-derived GLP-1 production.
RESULTS
Diphtheria Toxin Injection intoGluDTRMice Induces Massive and Long-LastingaCell Ablation
a cells may impact b cell insulin secretion via GLP-1 and glucagon secretion (Huypens et al., 2000; Moens et al., 1998; Sa- mols et al., 1965). Therefore, we first determined the impact ofa
cell ablation usingGluDTRmice, which express the human diph- theria toxin (DT) receptor under the control of the rat glucagon promoter (Thorel et al., 2011). GluDTR mice were fully back- crossed to a C57BL/6N background, and positive mice as well as littermate controls were injected with 1.5mg DT at the age of 6 weeks. 6 weeks later, the percentage ofacells in dispersed islet single cells decreased from 8.9% in controls to 0.4% in cells from GluDTRmice (Figure S1). Immunohistochemical staining for insulin and glucagon in pancreatic sections from GluDTR mice 63 weeks after DT injection showed long-lastingacell abla- tion without evidence for regeneration (Figure 1A). 6 weeks after DT injection, proglucagon mRNA expression in pancreatic islets Figure 1. DT Injection inGluDTRMice Leads to Massive and Long-LastingaCell Ablation
(A) Immunohistochemical staining of pancreatic sections fromGluDTRand wild-type (WT) mice 63 weeks after diphtheria toxin (DT) injection stained for insulin (green) and glucagon (red). Scale bar, 50mm.
(B)GcgandIns2mRNA expression in isolated islets fromGluDTRand WT mice 6 weeks after DT injection. n = 8.
(C) Glucagon und active GLP-1 content of isolated islets from 13-week-oldGluDTRand WT mice. n = 4.
(D) Immunohistochemical staining of transversal colon sections from 31-week-oldGluDTRand WT mice stained for GLP-1 (red). White arrowheads indicate GLP-1-positive L cells. Scale bar, 25mm.
(E) Active GLP-1 content in terminal ileum of 54-week-oldGluDTRand WT mice. n = 6 and 7.
(F) Body weight development ofGluDTRversus WT mice after DT injection. n = 6.
(G) Fasting plasma glucagon levels in 30-week-oldGluDTRand WT mice. n = 6.
(H) Fasting plasma active GLP-1 levels inGluDTRand WT mice. n = 16 and 12.
n indicates the number of independent experiments or number of animals respectively, and error bars represent SEM. *p < 0.05. See alsoFigure S1.
was strongly reduced in GluDTR mice, while preproinsulin expression was similar to control animals (Figure 1B). Accord- ingly, islet protein content of glucagon and active GLP-1 was decreased (Figure 1C). The presence of intestinal L cells after DT injection was confirmed by immunostaining for GLP-1 in co- lon sections (Figure 1D), and active GLP-1 content in terminal ileum was similar inGluDTRand control mice (Figure 1E), con- firming the specificity ofacell ablation. Body weight remained equal inGluDTRand control mice after DT injection (Figure 1F).
Fasting plasma glucagon concentrations were significantly lower in 30-week-oldGluDTRmice compared to controls (Figure 1G), while fasting plasma active GLP-1 concentrations were similar (Figure 1H).
Lack ofaCell-Derived GLP-1 Impairs Glucose Tolerance in Aged Mice
6 weeks after DT injection (at the age of 12 weeks), intraperito- neal (i.p.) glucose tolerance was comparable in GluDTR and control mice (Figure 2A), confirming previously published data (Pedersen et al., 2013; Thorel et al., 2011). Plasma insulin con- centrations were also not different (Figure 2B). At the age of 28 weeks, i.p. glucose tolerance ofGluDTRmice was impaired compared to controls (Figure 2C), fasting insulin secretion was decreased (0.8±0.1 versus 1.2±0.2 ng/mL, p < 0.05), and insu- lin levels were lower during the glucose tolerance test (Figure 2D).
To stabilize systemic active GLP-1, we injected the DPP-4 inhib- itor sitagliptin 30 min prior to i.p. glucose administration, as described previously (Timper et al., 2016). Glucose tolerance was completely restored in 30-week-old GluDTR mice upon DPP-4 inhibition, along with increased insulin secretion (Figures 2E and 2F). Glucagon administration 1 min prior to i.p. glucose tolerance testing resulted in a similar glucose tolerance in both GluDTRand control mice (Figures 2G andS2), whereas insulin levels remained significantly lower inGluDTRanimals (Figure 2H).
Therefore, prolonged a cell ablation leads to impaired insulin secretion, which can be rescued by increasing systemic GLP-1, but not glucagon levels.
YoungaCell-Ablated Mice Show Enhanced Oral Glucose Tolerance due to a Compensatory Increase in Systemic Active GLP-1 Secretion
To understand the relative role of systemic versus islet-derived GLP-1, we performed oral glucose tolerance tests, which lead to a stronger stimulation of intestinal GLP-1 compared to i.p. in- jection. As shown previously (Pedersen et al., 2013), 12-week-old GluDTRmice displayed enhanced oral glucose tolerance (Fig- ure 3A). At this age, the content of active GLP-1 in the terminal ileum was higher inGluDTRmice than in controls (Figure 3B), suggesting compensatory upregulation of GLP-1 production in intestinal L cells. To stabilize systemic active GLP-1, we next in- jected sitagliptin prior to oral glucose tolerance testing. Active GLP-1 secretion was increased in GluDTRmice compared to Figure 2.aCell-Derived GLP-1 Is Required for Normal i.p. Glucose
Tolerance during Aging
(A and B) Intraperitoneal (i.p.) glucose tolerance test (A) and corresponding insulin levels (B) inGluDTRand WT mice at the age of 12 weeks. n = 13.
(C and D) i.p. glucose tolerance test (C) and corresponding insulin levels (D) in GluDTRand WT mice at the age of 28 weeks. n = 20.
(E and F) i.p. glucose tolerance test (E) and corresponding insulin levels (F) inGluDTRand WT mice at the age of 30–32 weeks. Sitagliptin (25 mg/kg body weight) was i.p. injected 30 min prior to glucose injection. n = 13 and 12.
(G and H) i.p. glucose tolerance test (G) and corresponding insulin levels (H) inGluDTRand WT mice at the age of 32–34 weeks. n = 13 and 12.
Glucagon (20mg/kg body weight) was i.p. injected 1 min prior to glucose injection.
n indicates the number of animals, and error bars represent SEM. *p < 0.05,
**p < 0.01, area under the curve (AUC). See alsoFigure S2.
3194 Cell Reports18, 3192–3203, March 28, 2017
control animals (Figure 3C), with concomitantly increased plasma insulin concentration (Figure 3D). To explore the role of glucagon, we administered exogenous glucagon toGluDTRmice prior to oral glucose tolerance testing. Glucagon did not increase glyce- mia compared to vehicle control (Figure 3E), probably due to increased insulin secretion (Figure 3F). This points to an insulin secretagogue effect of glucagon inGluDTRmice, similar to that of GLP-1 or to an effect of glucagon on GLP-1 secretion. How-
ever, active GLP-1 secretion was suppressed by glucagon (Fig- ure 3G), thus pointing to a direct effect of glucagon on insulin secretion. In 28-week-old GluDTRmice, glycemia (Figure 3H) and plasma insulin concentration (Figure 3I) during glucose toler- ance testing were comparable to control mice, suggesting thata cell-derived GLP-1 is not required for maintenance of oral glucose tolerance during aging. Accordingly, administration of the GLP-1 antagonist exendin(9–39) prior to oral glucose Figure 3. Oral Glucose Tolerance Is Improved in YoungaCell-Ablated Mice due to Increased Systemic Active GLP-1 Secretion
(A) Oral glucose tolerance test inGluDTRand WT mice at the age of 12 weeks. n = 15 and 13.
(B) Active GLP-1 content in terminal ileum of 8- to 12-week-oldGluDTRand WT mice. 1 = 154.1±24.7 ng/mL. n = 14 and 12.
(C and D) Plasma active GLP-1 concentration (C) and plasma insulin concentration (D) during oral glucose tolerance test in 13-week-oldGluDTRand WT mice that were injected with sitagliptin (25 mg/kg body weight) 30 min prior to glucose administration. n = 18 and 12.
(E–G) Blood glucose (E), insulin (F), and active GLP-1 (G) levels during an oral glucose tolerance test with or without glucagon inGluDTRand WT mice at the age of 13–15 weeks that were injected with sitagliptin (25 mg/kg body weight) 30 min prior to glucose administration (n = 11). Glucagon (50mg/kg body weight) or vehicle control was i.p. injected 1 min prior to glucose administration.
(H and I) Blood glucose (H) and insulin (I) during an oral glucose tolerance test inGluDTRand WT mice at the age of 28 weeks. n = 18 and 13.
(J) Oral glucose tolerance test with or without exendin(9–39) inGluDTRand WT mice at the age of 28 weeks. n = 9, 10, 6, 7. Exendin(9–39) (70 nmol/kg body weight) or vehicle control was intraperitoneally injected 30 min prior to glucose administration.
n indicates the number of animals, and error bars represent SEM. *p < 0.05, **p < 0.01, ***p < 0.001.
tolerance testing resulted in impairment of glycemia in both GluDTRand control mice (Figure 3J). Therefore, under normal conditions, lack of a cell-derived GLP-1 and glucagon is compensated by enhanced intestinal GLP-1 production.
Glucagon Deficiency Protects against HFD-Induced Glucose Intolerance Despite Obesity
After 22 weeks of HFD feeding,GluDTRmice showed improved i.p. (Figure 4A) and oral (Figure 4C) glucose tolerance, with similar insulin levels (Figures 4B and 4D). This can be explained by the lack of glucagon-mediated insulin resistance. Indeed, in- sulin sensitivity was increased in GluDTR mice (Figure 4E) despite similar body weight (Figure 4F). Finally, fasting systemic active GLP-1 concentrations were comparable (Figure 4G), while as expected,GluDTRmice displayed reduced fasting glucagon levels (Figure 4H). Therefore, the consequences of decreased systemic glucagon levels may mask the potential impairment in bcell function caused by the lack ofacell-derived GLP-1.
aCell-Derived Glucagon-Related Peptides Are Critical for NormalbCell Function
To substantiate the hypothesis thatacell-derived GLP-1 effects are required for normalbcell function, we performed ex vivo ex- periments in isolated islets, where GLP-1 action is limited to
paracrine effects and signaling from peripheral GLP-1 target or- gans to thebcell (such as the nervous system) is missing. First, we performed glucose-stimulated insulin secretion (GSIS) with islets isolated fromGluDTRmice.acell-ablated islets showed a blunted insulin response to glucose (Figure 5A), suggesting thatacell-derived products are necessary for properbcell func- tion. To depict the precise role of GLP-1, we then confirmed the presence of bioactive GLP-1 in normal mouse islets by perform- ing GSIS following incubation with a GLP-1 antagonist. Indeed, in the presence of exendin(9–39), GSIS was decreased by 57%
(Figure 5B). Next, we used the GLP1 receptor activity reporter cell line HTLA (Kroeze et al., 2015) to test whether human islets secrete bioactive GLP-1. Addition of conditioned media from hu- man islets or recombinant GLP-1 as positive control induced a luminescence signal in HTLA cells, suggesting GLP1 receptor signaling occurs (Figure 5C). Co-incubation with exendin(9–39) completely abolished GLP-1 receptor signaling-mediated luminescence. Administration of exogenous GLP-1 or glucagon during GSIS completely restored b cell secretory function (Figure 5D). Cumulative insulin release of cultured islets over 24 hr was significantly reduced in GluDTR islets (Figure 5E), whereas islet insulin content was increased compared to control islets (Figure 5F), as typically observed in situation of an insulin secretion defect. We have shown that the incretin Figure 4. aCell Ablation Protects against HFD-Induced Glucose Intolerance by Increasing Insulin Sensitivity
(A and B) Blood glucose (A) and insulin (B) levels during an i.p. glucose tolerance test in in 27-week-oldGluDTRand WT mice following 21 weeks of high-fat diet (HFD) feeding. n = 13 and 11.
(C and D) Blood glucose (C) and insulin (D) levels during an oral glucose tolerance test inGluDTRand WT mice following 22 weeks of HFD feeding. n = 13 and 11.
(E) Insulin tolerance test inGluDTRand WT mice after 23 weeks of HFD feeding. n = 13 and 11.
(F–H) Body weight (F), fasting plasma active GLP-1 (G), and fasting plasma glucagon (H) concentrations inGluDTRand WT mice following 22 weeks of HFD feeding. DT injections were at 5–6 weeks of age and HFD was started at 6–7 weeks of age.
n indicates the number of animals, and error bars represent SEM. *p < 0.05, **p < 0.01, ***p < 0.001.
3196 Cell Reports18, 3192–3203, March 28, 2017
glucose-dependent insulinotropic polypeptide (GIP) is a strong inducer of active GLP-1 in islets (Timper et al., 2016). Exogenous GIP-induced insulin secretion was impaired inGluDTRversus control islets (Figure S3). Activation of both the GLP-1 and the
glucagon receptor leads to an increase in intracellular cyclic AMP (cAMP) concentration (Seino and Shibasaki, 2005).
Accordingly, the adenylyl cyclase activator forskolin was able to rescue the defective insulin secretion observed inGluDTR Figure 5.a Cell-Derived Glucagon-Related Peptides Are Required for Normal GSIS and bCell Function via cAMP-Mediated bCell Potentiation
(A) Glucose-stimulated insulin secretion (GSIS) of isolated islets fromGluDTRmice that had been injected with DT or saline. n = 3.
(B) GSIS of isolated C57BL/6N mouse islets with or without co-incubation of 100 nM exendin(9–39) during stimulation with 16.7 mM glucose. n = 3.
(C) Relative luminescence of the GLP1R activity luciferase reporter assay. 48 hr after transfection with the GLP1R-Tango vector, HTLA cells were treated with human islet conditioned medium (HI) with or without 100 nM exendin(9–39) or human islet medium containing 100 nM GLP-1 with or without 100 nM ex- endin(9–39). n = 3.
(D) GSIS and corresponding stimulatory index (fold increase stimulated over basal insulin release) ofGluDTRand WT islets with or without co-incubation of 10 nM GLP-1 or 10 nM glucagon during 16.7 mM glucose stimulation. 1 = 7.1±1.5 ng/mL. n = 3.
(E) Cumulative insulin secretion over 24 hr. 1 = 797.6±87.8 ng/mL. n = 3.
(F) Cellular insulin content ofGluDTRand WT islets. 1 = 937.0±109.5 ng/mL. n = 3.
(G) GSIS and corresponding stimulatory index ofGluDTRand WT islets with or without co-incubation of 10mM forskolin during 16.7 mM glucose stimulation.
1 = 100.0±12.2 ng/mL. n = 3.
n indicates the number of independent experiments, and error bars represent SEM. **p < 0.01, ***p < 0.001. See alsoFigure S3.
islets (Figure 5G). Thus,acell-derived glucagon-related peptides are indispensable for normal b cell function, which requires cAMP-mediated potentiation of the glucose effect.
aCell-Specific PC1/3 Knockout Leads to Decreased Active GLP-1 Secretion and Cell Content
Islets fromGluDTRmice lack both GLP-1 and glucagon. Thus, the lack of glucagon-induced insulin resistance (Figure 4) may mask the contribution ofacell-derived GLP-1 effects required forbcell function. To substantiate the relevance ofacell-derived GLP-1, we generated anacell-specific PC1/3 knockout mouse.
For this, we created aPcsk1fl/flmouse (Figure S4A) and crossed it with anacell-specific Gcg-Cre-expressing mouse (Herrera, 2000) to obtain Pcsk1fl/fl;Gcg-Cretg/ (‘‘aPcsk1 / ’’) and Pcsk1fl/fl;Gcg-Cre / (‘‘Pcsk1fl/fl’’) littermate control mice (Fig- ure S4B). To confirm Cre recombinase-targeted recombination ofloxPsites in a tissue with abundant expression of PC1/3, we crossed thePcsk1fl/flmice with PDXcreER mice to generate a tamoxifen-inducible b cell-specific PC1/3 knockout mouse.
Following tamoxifen administration, effective knockdown of PC1/3 inbPcsk1 / mice was confirmed by co-immunostaining for PC1/3 and insulin in pancreatic sections (data not shown) and RNA analysis of isolated islets (5.9%±0.8%Pcsk1expression
relative to control). Isolated islets fromaPcsk1 / mice showed decreased active GLP-1 secretion (65.9% of control;Figure 6A) and protein content (64.5% of control;Figure 6B). The extent of GLP-1 reduction is compatible to previously reported Cre re- combinase activity in 40%–45% of a cells in the Gcg-Cre model (Chow et al., 2014; Lu et al., 2010). Interestingly, glucagon secretion and content was increased inaPcsk1 / compared to control mice, while insulin secretion and content were compara- ble (Figures 6A and 6B). GSIS was similar inaPcsk1 / and con- trol islets and was increased upon stimulation with the incretin hormone GIP (Figure 6C). aPcsk1 / mice had normal body weight development (Figure 6D), and fasting plasma active GLP-1 levels were comparable to control mice (Figure 6E), sug- gesting intact systemic GLP-1 secretion. RNA analysis of iso- lated islets fromaPcsk1 / mice suggested thatbcell identity is not altered compared to controls (Figure 6F). Thus,aPcsk1 / mice show reduced levels of active GLP-1, while glucagon and insulin levels are not reduced.
aCell-Derived GLP-1 Is Required for Adaptation to Metabolic Stress
I.p. glucose tolerance was slightly improved in 11-week-old aPcsk1 / mice compared to controls (Figure 7A), and insulin Figure 6. aCell-Specific PC1/3 Knockout Results in Decreased Islet Active GLP-1 Secretion and Cell Content
(A) Active GLP-1, glucagon and insulin release over 24 hr from isolatedaPcsk1 / and control islets. n = 3.
(B) Active GLP-1, glucagon and insulin content of isolatedaPcsk1 / and control islets. n = 3.
(C) GSIS and corresponding stimulatory index (fold increase of stimulated over basal insulin release) of isolatedaPcsk1 / and control islets with or without co- incubation of 10 nM GIP during 11.1 mM glucose stimulation. 1 = 12.5±2.4 ng/mL. n = 3.
(D) Body weight development ofaPcsk1 / and control mice. n = 9 and 12.
(E) Fasting plasma active GLP-1 concentration in 15-week-oldaPcsk1 / and control mice. n = 16.
(F) RNA profile of isolated islet fromaPcsk1 / and control mice.
n indicates the number of independent experiments or the number of animals, and error bars represent SEM. *p < 0.05, **p < 0.01. See alsoFigure S4.
3198 Cell Reports18, 3192–3203, March 28, 2017
levels tended to be higher (Figure 7B), while oral glucose toler- ance and insulin levels were similar betweenaPcsk1 / and con- trol mice (Figures 7C and 7D). This somewhat paradoxical trend in i.p. glucose tolerance disappeared in 22-week-old chow-fed (Figures S5A and 5B) and 36-week-old HFD-fed (Figures S5E and S5F)aPcsk1 / mice, along with similar body weight and systemic glucagon and active GLP-1 levels as well as ileal active GLP-1 content compared to control animals (Figures S5C, S5D, and S5G–S5I). Next, we tested the adaptive capacity of the aPcsk1 / mice fed a HFD and injected with a single low dose of thebcell toxin streptozotocin at 10 weeks of age. At the age of 15 weeks, body weight as well as active GLP-1 concentrations
in plasma and ileal active GLP-1 content were similar in both groups (Figures 7E–7G). I.p. glucose tolerance was impaired in HFD/streptozotocin-treated aPcsk1 / mice compared to controls (Figure 4H), along with decreased insulin secretion (Fig- ure 7I). Of note, while the difference between the genotype appears mild, it was consistently observed in four independent cohorts of six to ten animals each. In one cohort, the maximal impairment in aPcsk1 / mice was observed at 11 weeks of age, while in the other three, it was observed at 14 weeks. For simplicity, only the average of data of week 14 is shown. Sitaglip- tin injection prior to i.p. glucose tolerance testing partially reversed the impaired glucose tolerance (Figure 7J). However, Figure 7. i.p. Glucose Tolerance Is Improved in Chow-Diet-FedaCell-Specific PC1/3 Knockout Mice and Is Impaired during Metabolic Stress (A and B) Blood glucose (A) and corresponding insulin (B) levels during an i.p. glucose tolerance test inaPcsk1 / and control mice at the age of 11 weeks. n = 16.
(C and D) Blood glucose (C) and insulin (D) levels during an oral glucose tolerance test inaPcsk1 / and control mice at the age of 12 weeks. n = 9.
(E–G) Body weight (E), fasting plasma active GLP-1 concentration (F), and active GLP-1 content (G) in terminal ileum of 15-week-old HFD/streptozotocin (STZ)- treatedaPcsk1 / and control mice. n = 20 and 24.
(H and I) Blood glucose (H) and corresponding insulin (I) levels during an i.p. glucose tolerance test in HFD/STZ-treatedaPcsk1 / and control mice at the age of 14 weeks. n = 26 and 31.
(J and K) Blood glucose (J) and corresponding insulin (K) levels during an i.p. glucose tolerance test in 15-week-oldaPcsk1 / and control mice that were injected with sitagliptin (25 mg/kg body weight) 30 min prior to glucose administration to stabilize active GLP-1. n = 15 and 17.
n indicates the number of animals, and error bars represent SEM. *p < 0.05, **p < 0.01, ***p < 0.001. See alsoFigure S5.
plasma insulin concentrations remained lower inaPcsk1 / mice than in control mice (Figure 7K). To assess the potential rele- vance of these findings in humans, we compared GLP-1 content of human and mouse islets. Human islets showed an even higher active GLP-1 content than mouse islets (29.28±3.48 versus 0.32±0.04 pg/islet, n = 23–27, p < 0.001). To conclude, under metabolic stress and increased secretory demand,bcell adap- tation requiresacell expression of PC1/3 and paracrine action of GLP-1. Finally, we tested whether there is increasedatobcell conversion in aPcsk1 / , which was reported to occur after near-completebcell destruction (Thorel et al., 2010). The level of proinsulin-related peptides (Figure S5J) and the number of insulin-negative but PC1/3-positive cells was identical in aPcsk1 / and control mice, suggesting thatatobcell conver- sion is not increased in our mouse model.
DISCUSSION
In the present study, we show thata cell-derived glucagon- related peptides are mandatory for normal glucose-stimulated insulin secretion. While under normal conditionsacell-derived GLP-1 may be compensated by increased intestinal GLP-1 secretion, paracrine GLP-1 is needed for normal glucose ho- meostasis during aging and adaptation to metabolic stress.
Several in vitro and ex vivo studies have shown the ability ofa cells to produce active GLP-1 (Ellingsgaard et al., 2011; Hansen et al., 2011; Kilimnik et al., 2010; Marchetti et al., 2012; Nie et al., 2000; Thyssen et al., 2006). However, the biological role of localacell-derived GPL-1 in islet function and glucose ho- meostasis remained unknown. We show that during aginga cell-ablated mice develop impaired insulin secretion and glucose homeostasis. Stabilization of systemic GLP-1 with the DPP-4 inhibitor sitagliptin completely rescued this pheno- type, suggesting a role fora cell-derived GLP-1 in aging. In contrast, glucagon administration failed to normalize glycemia inacell-ablated mice. To demonstrate the paracrine effect of local GLP-1 in islets where glucagon production is present, we generated a genetic mouse model ofacell-specific GLP-1 deficiency. TheaPcsk1 / mouse showed decreased active GLP-1 islet content and secretion into culture supernatant, while systemic active GLP-1 concentrations were comparable to control mice, which can be explained by intact GLP-1 secre- tion from intestinal L cells. In young, chow-fedaPcsk1 / mice, i.p. glucose tolerance was slightly improved compared to con- trol animals. However, this effect was lost upon aging. Meta- bolic stress inaPcsk1 / mice impaired glycemic control, while administration of sitagliptin partially rescued this impairment in glucose tolerance.
One limitation of theaPcsk1 / model is the low level of Cre recombinase activity in theGcg-Cre mouse. Consistent with pre- vious publications using theGcg-Cre mouse (Chow et al., 2014;
Lu et al., 2010), only a partial knockout of PC1/3 was achieved in acells with remaining pancreatic active GLP-1 production. This could be the reason why aging did not affect glucose tolerance in aPcsk1 / mice, while it did in the model of a cell ablation.
Despite remaining intra-islet GLP-1 production, the uncovered impaired glucose disposal points to an important role ofacell- derived GLP-1 in adaptation to metabolic stress.
acells were shown to be able to convert to functionalbcells after near-total and partialbcell ablation (Thorel et al., 2010).
Thus, the present findings based on a model using HFD/strepto- zotocin could potentially be explained by conversion of PC1/3- deficientacells tobcells. Suchbcells would be expected to lack PC1/3 and thus would not be able to secrete fully processed insulin. However, this explanation is unlikely, as the levels of pro- insulin (including des-64,65-proinsulin yielded by PC2, but not PC1/3 cleavage) were identical in plasma fromaPcsk1 / and control mice. In addition, the number of insulin-positive but PC1/3-negative cells in pancreata from both genotypes was identical. Altogether, this strongly suggests that, despite use of streptozotocin, a to bcell conversion is not increased in our mouse model.
Administration of the GLP-1 antagonist exendin(9–39) during an oral glucose tolerance test resulted in impaired glucose tolerance in both a cell-ablated and control mice, indicating dependence on systemic GLP-1 signaling. In line with previously published data by Smith and co-workers (Smith et al., 2014), our data indicate that intra-islet GLP-1 is not required for oral glucose tolerance under normal conditions. Indeed, lack of GLP-1 receptors inbcells resulted in impaired i.p. glucose toler- ance compared to controls, while oral glucose tolerance was normal. This suggests indirect action of systemic GLP-1 onb cells during oral glucose challenge (Smith et al., 2014). In young acell-ablated mice, we observed enhanced oral glucose toler- ance, resembling the phenotype of the double glucagon and GLP-1 receptor knockout mouse, which displays normal i.p.
glucose tolerance and improved oral glucose tolerance (Ali et al., 2011). The reason for improved oral glucose tolerance in the double receptor knockout mice is enhanced sensitivity to- ward other incretins. In ouracell-ablated islets, however, equi- molar GLP-1 or GIP caused similar or lower fold increase of GSIS compared to control islets. Thus, we could not observe increased sensitivity ofbcells toward these classical incretins.
Instead, active GLP-1 content in the terminal ileum ofacell-ab- lated mice was increased compared to controls and active GLP-1 secretion was higher during oral glucose tolerance testing. This points to a compensatory increase of peripheral GLP-1, possibly to rescue the lack of local islet-derived GLP-1.
Glucagon administration inacell-ablated mice prior to an oral glucose tolerance test suppressed GLP-1 secretion. In humans, fasting glucagon concentration is a strong determinant of mixed meal-induced GLP-1 secretion (Nauck et al., 2011). One study has reported decreased GLP-1 and GIP plasma levels in response to oral glucose loading in healthy human subjects when glucagon was infused concomitantly (Ranganath et al., 1999). However, highly supraphysiological concentrations of glucagon were infused, thus providing limited value for physio- logical processes. In another human study, where physiological plasma glucagon concentrations were maintained, no effect of exogenous glucagon on GLP-1 secretion after ingestion of a mixed meal was detectable (Meier et al., 2010). Interestingly, upon cessation of glucagon infusion, GLP-1 concentrations increased significantly. Even though our data show a clear decrease of GLP-1 secretion by glucagon, the effect might be also indirect via stimulated insulin secretion. Because glycemia did not change upon glucagon administration, the physiological
3200 Cell Reports18, 3192–3203, March 28, 2017
significance of this hormonal interaction remains to be elucidated.
We further show that glucagon deficiency protects against HFD-induced glucose intolerance in a cell-ablated mice. This phenotype has been described in glucagon-receptor-deficient mice (Conarello et al., 2007; Gelling et al., 2003). However, these studies remained inconclusive about the direct role of glucagon in HFD-induced glucose intolerance, because the mice had a lean phenotype and showed pronounced compensatoryacell hyper- plasia along with highly elevated levels of systemic active GLP-1.
Our HFD-fedacell-ablated mice had body weights and systemic active GLP-1 concentrations comparable to their HFD-induced obese controls. Thus, the development of insulin resistance and glucose intolerance after HFD can be clearly attributed to the ac- tion of glucagon in our model of glucagon deficiency.
We (Ellingsgaard et al., 2011) and others have suggested that a cells may secrete GLP-1, but this is still debated. Using a bioactivity reporter assay, we now show that human islets secrete bioactive GLP-1. In isolatedacell-ablated mouse islets, GSIS was dramatically reduced. This phenotype was fully rescued upon addition of exogenous GLP-1 or glucagon. The adenylyl cyclase activator forskolin also increased GSIS ina cell-ablated islets. Consistent with our findings, isolated islets from glucagon and GLP-1 receptor knockout mice also demon- strated decreased GSIS (Ali et al., 2011). These data show that glucose alone is not sufficient to induce normal insulin secretion from pancreaticbcells and that cAMP-mediatedbcell potentia- tion byacell-produced glucagon-related peptides is mandatory for stimulus-secretion coupling inbcells.
Taken together, our data show that the paracrine action ofa cell-derived glucagon-related peptides are important for normal glucose homeostasis andbcell function.acell-derived GLP-1 is required forbcell adaptation to aging and metabolic stress. It fol- lows that the clinical use of systemic GLP-1 agonists in the treat- ment of type 2 diabetes may be less effective to impact onbcell function compared to stabilizing localacell-derived GLP-1 by DPP-4 inhibitors.
EXPERIMENTAL PROCEDURES Experimental Animals
All animal experiments were performed in accordance with the federal laws of Switzerland and approved by the cantonal and institutional authorities. C57BL/
6N were obtained from Charles River Laboratories.GluDTRmice, which have been described before (Thorel et al., 2010), were backcrossed to a C57BL/6N background using microarray-based speed congenics (Biolytix). BothGluDTR and littermate control mice (‘‘WT’’) received DT injections, except when noted otherwise. To generate conditional Pcsk1 knockout mice (‘‘aPcsk1 /’’), C57BL/6N embryonic stem cells carrying a transgene encoding a FRT-flanked lacZ and neo cassette andloxPsites around exon 5 of the Pcsk1gene (EUCOMM program) were injected into blastocysts. Offspring of chimera with germline transmission were crossed with a ROSA26::FLPe mouse (Farley et al., 2000) to excise the FRT-flanked sequence. The resultingPcsk1fl/flmice (B6-Pcsk1 < tm1Boe > /N) were crossed withGcg-Cre mice (Herrera, 2000), which were previously backcrossed to the C57BL/6N genetic background, to obtainacell-specificPcsk1knockout mice (B6-Tg(Gcg-Cre)1Herr/N x B6- Pcsk1 < tm1Boe > /N) (Figure S4). Male mice were used for all studies. See figure legends for specific ages. Animals were housed in a pathogen-free an- imal facility with a 12-hr light/12-hr dark cycle and were allowed free access to food and water.
DT Injections
DT (D0564; Sigma) dissolved in 0.9% NaCl was administered to 5- to 6-week- old mice in three i.p. injections of 500 ng each on days 1, 3, and 5 as reported before (Thorel et al., 2010). The total amount of injected DT was 1.5mg per mouse.
HFD/Streptozotocin Model
aPcsk1 / andPcsk1fl/fllittermate mice were fed a HFD (D12331, 58 kcal % fat, Research Diets) starting at 4 weeks of age. At 10 weeks of age, animals received an i.p. injection of 130 mg/kg body weight streptozotocin (Sigma).
At 14 weeks of age, an i.p. glucose tolerance test was performed. The glucose tolerance test was repeated at 15 weeks of age with the addition of an injection of 25 mg/kg body weight sitagliptin (S8576, Sigma) 30 min prior to glucose administration
Glucose and Insulin Tolerance Tests
Glucose (2 g/kg body weight) was either administered by i.p. injection or oral gavage after 6 hr of fasting (8:00–14:00). Blood samples were obtained by tail tip bleeding and assayed using a glucometer (Freestyle; Abbott) or mixed with EDTA, aprotinin (A1153; Sigma), or diprotin A (I9759; Sigma) and spun down, and plasma was stored for further analysis. To stabilize systemic active GLP-1, 25 mg/kg body weight sitagliptin dissolved in 0.9% NaCl was i.p.
injected 30 min prior to glucose administration. Synthetic glucagon dissolved in PBS (20mg/kg body weight; G2044; Sigma) was i.p. injected 1 min prior to glucose administration. Synthetic exendin(9–39) dissolved in 0.9% NaCl (236mg/kg body weight; H-8740; Bachem) was i.p. injected 30 min prior to glucose administration. For insulin tolerance tests, mice were fasted 3 hr (8:00–11:00) and injected with 1 U/kg body weight insulin (Novorapid, Novo Nordisk), and measurement of blood glucose was done as described above.
Human Pancreatic Islets
Human islets isolated from pancreata of cadaver organ donors in accordance with the local institutional ethical committee were provided by the islets for research distribution program through the European Consortium for Islet Transplantation, under the supervision of the Juvenile Diabetes Research Foundation (31-2012-783). Islets were cultured in CMRL-1066 medium con- taining 5 mM streptomycin, 100 U/mL penicillin, 2 mM glutamax, and 10% fetal calf serum (FCS) (Invitrogen) on extracellular-matrix-coated 24-well plates (Novamed) in humid environment containing 5% CO2. Islets were cultured for 48 hr with 1 mM diprotin A to stabilize GLP-1. Conditioned media were collected in tubes containing diprotin A (end concentration 100mM) and stored at 80C for the GLP-1R activity luciferase reporter assay.
Mouse Pancreatic Islets
Mouse islet were isolated by collagenase digestion (4189; Worthington) and cleaned by filtration trough a mesh and hand picking. They were either lysed directly after isolation for mRNA extraction or cultured on extracellular-ma- trix-coated 24-well plates in RPMI-1640 medium containing 11.1 mM glucose, 100 U/mL penicillin, 100mg/mL streptomycin, 2 mM glutamax, 50mg/mL gen- tamycin, 10mg/mL fungizone, and 10% FCS.
GSIS Assay
For ex vivo GSIS experiments, mouse islets were plated in quadruplicate and allowed to attach for 2 days. Islets were pre-incubated for 30 min in modified Krebs-Ringer bicarbonate buffer (KRB; 115 mM NaCl, 4.7 mM KCl, 2.6 mM CaCl22H2O, 1.2 mM KH2PO4, 1.2 mM MgSO47H2O, 10 mM HEPES, and 0.5% BSA [pH 7.4]) containing 2.8 mM glucose. KRB was then replaced by KRB 2.8 mM glucose, and supernatants were collected 60 min later, followed by 60-min incubation in KRB containing 16.7 mM glucose with or without addi- tion of indicated treatments. After supernatant collection, islets insulin content was extracted with 0.18 N HCl in 70% EtOH.
Ileal Hormone Content Extraction
To determine active GLP-1 content of the terminal ileum, 3 cm of ileum was excised proximal from the caecum, rinsed with PBS, and extracted with
0.18 N HCl in 70% EtOH. Samples were spun down and supernatants stored for further analysis.
Hormone Measurements
Insulin and active GLP-1 concentrations in plasma and islet supernatants as well as glucagon concentrations in islet supernatants were determined using commercially available kits (Meso Scale Discovery). Plasma glucagon levels were determined using a glucagon ELISA (Mercodia).
RNA Extraction and qPCR
Total RNA of isolated mouse islets was extracted using the NucleoSpin RNA II Kit (Macherey Nagel). cDNA was prepared with random hexamers and Superscript II reverse transcriptase (Invitrogen). For qPCR, the real-time PCR system 7500 (Applied Biosystems) and the following TaqMan assays were used: Gcg - Mm00801714_m1;Ins2 - Mm00731595_gH; Pcsk1 – Mm01345254_m1; Pcsk2 – Mm00500974_m1; Gcgr – Mm00433546_m1;
Glp1r– Mm00445292_m1;Pdx1- Mm00435565_m1;Arx– Mm00545903_
m1;Dpp4 - Mm01329193_m1. Gene expression was analyzed with the comparative 2DDCTmethod.
Immunohistochemical Staining
For tissue sections, formalin-fixed pancreata were embedded in paraffin and cut into 4-mm-thick sections. Tissue sections were deparaffinized, rehydrated, and stained with the indicated antibodies. For staining of single mouse islet cells, mouse islets were dispersed with 0.02% trypsin-EDTA (Invitrogen) followed by cytospin onto glass slides for 3 min at 1,000 rpm. Islet cells were fixed for 30 min with 4% paraformaldehyde, per- meabilized with 0.5% Triton X-100 for 4 min, and stained with the indicated antibodies. The following antibodies were used: guinea pig anti-insulin (1:100; Dako), mouse anti-glucagon (1:1,000; Sigma), goat anti-GLP1 (1:50; Santa Cruz Biotechnology), and rabbit anti-PC1/3 (1:100; Millipore).
Secondary antibodies were coupled to Alexa Fluor 647 dyes (Molecular Probes) and fluorescein isothiocyanate (FITC) (Jackson ImmunoResearch Laboratories). Nuclei were labeled with DAPI (Sigma). Sections were analyzed with Olympus BX61 and BX63 using CellSens software and Im- ageJ (Fiji edition). For determination of insulin+/PC1/3 cells in pancreata from HFD/streptozotocin-treatedaPcsk1 / and control mice, three mice per genotype with a total of 18 and 15 islets (799 and 900 cells, respec- tively) were analyzed.
GLP-1R Activity Luciferase Reporter Assay
The GLP-1R activity reporter vector GLP1R-Tango (Addgene 66295) based on PRESTO-Tango system has been described before (Kroeze et al., 2015). HTLA cells (a HEK293 cell line stably expressing a tTA-dependent luciferase reporter and ab-arrestin2-TEV fusion gene) were a gift from B. L.
Roth (University of North Carolina). Cells were plated at 2.23106cells per 25 cm2cell-culture flasks (day 1), and the reporter vector was transfected on day 2 (Lipofectamine 2000, Thermo Fisher Scientific). On day 3, trans- fected cells were transferred at 30,000 cells per well into 96-well plates con- taining 100mL DMEM medium per well. On day 4, the medium was removed and replaced with human-islet-derived conditioned media containing 1 mM diprotin A. On day 5, medium was removed and 100mL Bright-Glo solution (Promega) diluted 1:1 in RPMI buffer was added. Luminescence was measured 5 min later.
Statistics
Statistical analysis was performed using GraphPad Prism 6.0 (GraphPad Soft- ware). Data are presented as mean±SEM and were analyzed using one- or two-tailed Mann-WhitneyUtests or ANOVA with Sidak’s multiple comparison test when more than two conditions were compared. Differences were consid- ered statistically significant when p < 0.05.
SUPPLEMENTAL INFORMATION
Supplemental Information includes five figures and can be found with this article online athttp://dx.doi.org/10.1016/j.celrep.2017.03.005.
AUTHOR CONTRIBUTIONS
S.T., D.T.M., M.B.-S., and M.Y.D. designed the study and wrote the manuscript. S.T., D.T.M., E. Dror, and M.B.-S. performed and analyzed the ex- periments. F.S., T.M.N., N.G., N.K., E. Dalmas, M.S., and V.M. helped with the experiments. F.T. and P.L.H. provided theGluDTRmice.
ACKNOWLEDGMENTS
This work was financially supported by the MD PhD fellowship and research grants of the Swiss National Science Foundation and by the University of Basel Research Fund for Young Researcher. We thank the Center for Transgenic Models (CTM) of the University of Basel and D. Klewe-Nebenius for supporting the production of thePcsk1fl/flmice and for rederivation of theGluDTRmice.
We thank our technicians Ste´phanie Ha¨uselmann, Marcela Borsigova, and Kaethi Dembinski for their excellent technical assistance.
Received: September 22, 2016 Revised: January 22, 2017 Accepted: March 1, 2017 Published: March 28, 2017
REFERENCES
Ali, S., Lamont, B.J., Charron, M.J., and Drucker, D.J. (2011). Dual elimination of the glucagon and GLP-1 receptors in mice reveals plasticity in the incretin axis. J. Clin. Invest.121, 1917–1929.
Baraboi, E.D., St-Pierre, D.H., Shooner, J., Timofeeva, E., and Richard, D.
(2011). Brain activation following peripheral administration of the GLP-1 recep- tor agonist exendin-4. Am. J. Physiol. Regul. Integr. Comp. Physiol. 301, R1011–R1024.
Burcelin, R., Da Costa, A., Drucker, D., and Thorens, B. (2001). Glucose competence of the hepatoportal vein sensor requires the presence of an acti- vated glucagon-like peptide-1 receptor. Diabetes50, 1720–1728.
Cabou, C., and Burcelin, R. (2011). GLP-1, the gut-brain, and brain-periphery axes. Rev. Diabet. Stud.8, 418–431.
Campbell, J.E., and Drucker, D.J. (2013). Pharmacology, physiology, and mechanisms of incretin hormone action. Cell Metab.17, 819–837.
Chow, S.Z., Speck, M., Yoganathan, P., Nackiewicz, D., Hansen, A.M., Lade- foged, M., Rabe, B., Rose-John, S., Voshol, P.J., Lynn, F.C., et al. (2014).
Gp130 receptor signaling mediates alpha cell dysfunction in a rodent model of type 2 diabetes. Diabetes63, 2984–2995.
Conarello, S.L., Jiang, G., Mu, J., Li, Z., Woods, J., Zycband, E., Ronan, J., Liu, F., Roy, R.S., Zhu, L., et al. (2007). Glucagon receptor knockout mice are resis- tant to diet-induced obesity and streptozotocin-mediated beta cell loss and hyperglycaemia. Diabetologia50, 142–150.
Deacon, C.F. (2004). Circulation and degradation of GIP and GLP-1. Horm.
Metab. Res.36, 761–765.
Deacon, C.F., Nauck, M.A., Toft-Nielsen, M., Pridal, L., Willms, B., and Holst, J.J. (1995). Both subcutaneously and intravenously administered glucagon- like peptide I are rapidly degraded from the NH2-terminus in type II diabetic patients and in healthy subjects. Diabetes44, 1126–1131.
Donath, M.Y., and Burcelin, R. (2013). GLP-1 effects on islets: hormonal, neuronal, or paracrine? Diabetes Care36(Suppl 2), S145–S148.
Drucker, D.J. (2006). The biology of incretin hormones. Cell Metab.3, 153–165.
Ellingsgaard, H., Ehses, J.A., Hammar, E.B., Van Lommel, L., Quintens, R., Martens, G., Kerr-Conte, J., Pattou, F., Berney, T., Pipeleers, D., et al.
(2008). Interleukin-6 regulates pancreatic alpha-cell mass expansion. Proc.
Natl. Acad. Sci. USA105, 13163–13168.
Ellingsgaard, H., Hauselmann, I., Schuler, B., Habib, A.M., Baggio, L.L., Meier, D.T., Eppler, E., Bouzakri, K., Wueest, S., Muller, Y.D., et al. (2011). Interleukin- 6 enhances insulin secretion by increasing glucagon-like peptide-1 secretion from L cells and alpha cells. Nat. Med.17, 1481–1489.
3202 Cell Reports18, 3192–3203, March 28, 2017
Farley, F.W., Soriano, P., Steffen, L.S., and Dymecki, S.M. (2000). Widespread recombinase expression using FLPeR (flipper) mice. Genesis28, 106–110.
Furuta, M., Zhou, A., Webb, G., Carroll, R., Ravazzola, M., Orci, L., and Steiner, D.F. (2001). Severe defect in proglucagon processing in islet A-cells of prohor- mone convertase 2 null mice. J. Biol. Chem.276, 27197–27202.
Gelling, R.W., Du, X.Q., Dichmann, D.S., Romer, J., Huang, H., Cui, L., Obici, S., Tang, B., Holst, J.J., Fledelius, C., et al. (2003). Lower blood glucose, hyper- glucagonemia, and pancreatic alpha cell hyperplasia in glucagon receptor knockout mice. Proc. Natl. Acad. Sci. USA100, 1438–1443.
Hansen, L., Deacon, C.F., Orskov, C., and Holst, J.J. (1999). Glucagon-like peptide-1-(7-36)amide is transformed to glucagon-like peptide-1-(9-36)amide by dipeptidyl peptidase IV in the capillaries supplying the L cells of the porcine intestine. Endocrinology140, 5356–5363.
Hansen, A.M., Bo¨dvarsdottir, T.B., Nordestgaard, D.N., Heller, R.S., Gotfred- sen, C.F., Maedler, K., Fels, J.J., Holst, J.J., and Karlsen, A.E. (2011). Upregu- lation of alpha cell glucagon-like peptide 1 (GLP-1) in Psammomys obesus—
an adaptive response to hyperglycaemia? Diabetologia54, 1379–1387.
Herrera, P.L. (2000). Adult insulin- and glucagon-producing cells differentiate from two independent cell lineages. Development127, 2317–2322.
Huypens, P., Ling, Z., Pipeleers, D., and Schuit, F. (2000). Glucagon receptors on human islet cells contribute to glucose competence of insulin release. Dia- betologia43, 1012–1019.
Kilimnik, G., Kim, A., Steiner, D.F., Friedman, T.C., and Hara, M. (2010). Intra- islet production of GLP-1 by activation of prohormone convertase 1/3 in pancreatica-cells in mouse models of ß-cell regeneration. Islets2, 149–155.
Kroeze, W.K., Sassano, M.F., Huang, X.P., Lansu, K., McCorvy, J.D., Gigue`re, P.M., Sciaky, N., and Roth, B.L. (2015). PRESTO-Tango as an open-source resource for interrogation of the druggable human GPCRome. Nat. Struct.
Mol. Biol.22, 362–369.
Lamont, B.J., Li, Y., Kwan, E., Brown, T.J., Gaisano, H., and Drucker, D.J.
(2012). Pancreatic GLP-1 receptor activation is sufficient for incretin control of glucose metabolism in mice. J. Clin. Invest.122, 388–402.
Lu, J., Herrera, P.L., Carreira, C., Bonnavion, R., Seigne, C., Calender, A., Ber- tolino, P., and Zhang, C.X. (2010). Alpha cell-specific Men1 ablation triggers the transdifferentiation of glucagon-expressing cells and insulinoma develop- ment. Gastroenterology138, 1954–1965.
Marchetti, P., Lupi, R., Bugliani, M., Kirkpatrick, C.L., Sebastiani, G., Grieco, F.A., Del Guerra, S., D’Aleo, V., Piro, S., Marselli, L., et al. (2012). A local glucagon-like peptide 1 (GLP-1) system in human pancreatic islets. Diabetolo- gia55, 3262–3272.
Meier, J.J., Ritter, P.R., Jacob, A., Menge, B.A., Deacon, C.F., Schmidt, W.E., Nauck, M.A., and Holst, J.J. (2010). Impact of exogenous hyperglucagonemia on postprandial concentrations of gastric inhibitory polypeptide and glucagon- like peptide-1 in humans. J. Clin. Endocrinol. Metab.95, 4061–4065.
Moens, K., Flamez, D., Van Schravendijk, C., Ling, Z., Pipeleers, D., and Schuit, F. (1998). Dual glucagon recognition by pancreatic beta-cells via glucagon and glucagon-like peptide 1 receptors. Diabetes47, 66–72.
Nakagawa, A., Satake, H., Nakabayashi, H., Nishizawa, M., Furuya, K., Na- kano, S., Kigoshi, T., Nakayama, K., and Uchida, K. (2004). Receptor gene expression of glucagon-like peptide-1, but not glucose-dependent insulino- tropic polypeptide, in rat nodose ganglion cells. Auton. Neurosci.110, 36–43.
Nauck, M.A., Vardarli, I., Deacon, C.F., Holst, J.J., and Meier, J.J. (2011).
Secretion of glucagon-like peptide-1 (GLP-1) in type 2 diabetes: what is up, what is down? Diabetologia54, 10–18.
Nie, Y., Nakashima, M., Brubaker, P.L., Li, Q.L., Perfetti, R., Jansen, E., Zam- bre, Y., Pipeleers, D., and Friedman, T.C. (2000). Regulation of pancreatic PC1
and PC2 associated with increased glucagon-like peptide 1 in diabetic rats.
J. Clin. Invest.105, 955–965.
Pedersen, J., Ugleholdt, R.K., Jørgensen, S.M., Windeløv, J.A., Grunddal, K.V., Schwartz, T.W., F€uchtbauer, E.M., Poulsen, S.S., Holst, P.J., and Holst, J.J. (2013). Glucose metabolism is altered after loss of L cells anda-cells but not influenced by loss of K cells. Am. J. Physiol. Endocrinol. Metab.304, E60–E73.
Ranganath, L., Schaper, F., Gama, R., Morgan, L., Wright, J., Teale, D., and Marks, V. (1999). Effect of glucagon on carbohydrate-mediated secretion of glucose-dependent insulinotropic polypeptide (GIP) and glucagon-like pep- tide-1 (7-36 amide) (GLP-1). Diabetes Metab. Res. Rev.15, 390–394.
Rouille´, Y., Martin, S., and Steiner, D.F. (1995). Differential processing of proglucagon by the subtilisin-like prohormone convertases PC2 and PC3 to generate either glucagon or glucagon-like peptide. J. Biol. Chem. 270, 26488–26496.
Samols, E., Marri, G., and Marks, V. (1965). Promotion of insulin secretion by glucagon. Lancet2, 415–416.
Seino, S., and Shibasaki, T. (2005). PKA-dependent and PKA-independent pathways for cAMP-regulated exocytosis. Physiol. Rev.85, 1303–1342.
Shimizu, N., Oomura, Y., Novin, D., Grijalva, C.V., and Cooper, P.H. (1983).
Functional correlations between lateral hypothalamic glucose-sensitive neu- rons and hepatic portal glucose-sensitive units in rat. Brain Res.265, 49–54.
Smith, E.P., An, Z., Wagner, C., Lewis, A.G., Cohen, E.B., Li, B., Mahbod, P., Sandoval, D., Perez-Tilve, D., Tamarina, N., et al. (2014). The role ofbcell glucagon-like peptide-1 signaling in glucose regulation and response to dia- betes drugs. Cell Metab.19, 1050–1057.
Thorel, F., Ne´pote, V., Avril, I., Kohno, K., Desgraz, R., Chera, S., and Herrera, P.L. (2010). Conversion of adult pancreatic alpha-cells to beta-cells after extreme beta-cell loss. Nature464, 1149–1154.
Thorel, F., Damond, N., Chera, S., Wiederkehr, A., Thorens, B., Meda, P., Woll- heim, C.B., and Herrera, P.L. (2011). Normal glucagon signaling andb-cell function after near-totala-cell ablation in adult mice. Diabetes60, 2872–2882.
Thyssen, S., Arany, E., and Hill, D.J. (2006). Ontogeny of regeneration of beta- cells in the neonatal rat after treatment with streptozotocin. Endocrinology 147, 2346–2356.
Timper, K., Dalmas, E., Dror, E., R€utti, S., Thienel, C., Sauter, N.S., Bouzakri, K., Be´dat, B., Pattou, F., Kerr-Conte, J., et al. (2016). Glucose-dependent in- sulinotropic peptide stimulates glucagon-like peptide 1 production by pancre- atic islets via interleukin 6, produced by a cells. Gastroenterology 151, 165–179.
Tucker, J.D., Dhanvantari, S., and Brubaker, P.L. (1996). Proglucagon pro- cessing in islet and intestinal cell lines. Regul. Pept.62, 29–35.
Waget, A., Cabou, C., Masseboeuf, M., Cattan, P., Armanet, M., Karaca, M., Castel, J., Garret, C., Payros, G., Maida, A., et al. (2011). Physiological and pharmacological mechanisms through which the DPP-4 inhibitor sitagliptin regulates glycemia in mice. Endocrinology152, 3018–3029.
Washington, M.C., Raboin, S.J., Thompson, W., Larsen, C.J., and Sayegh, A.I.
(2010). Exenatide reduces food intake and activates the enteric nervous sys- tem of the gastrointestinal tract and the dorsal vagal complex of the hindbrain in the rat by a GLP-1 receptor. Brain Res.1344, 124–133.
Wideman, R.D., Yu, I.L., Webber, T.D., Verchere, C.B., Johnson, J.D., Cheung, A.T., and Kieffer, T.J. (2006). Improving function and survival of pancreatic is- lets by endogenous production of glucagon-like peptide 1 (GLP-1). Proc. Natl.
Acad. Sci. USA103, 13468–13473.
