Emily Tubbs,1 Pierre Theurey,1 Guillaume Vial,1 Nadia Bendridi,1 Amélie Bravard,1

Marie-Agnès Chauvin,1 Jingwei Ji-Cao,1 Fabien Zoulim,2 Birke Bartosch,2 Michel Ovize,1,3

Hubert Vidal,1,4 and Jennifer Rieusset1,4

Mitochondria-AssociatedEndoplasmic ReticulumMembrane (MAM) Integrity IsRequired for Insulin Signalingand Is Implicated in HepaticInsulin ResistanceDiabetes 2014;63:3279–3294 | DOI: 10.2337/db13-1751

Mitochondria-associated endoplasmic reticulum (ER)membranes (MAMs) are functional domains between bothorganelles involved in Ca2+ exchange, through the voltage-dependent anion channel (VDAC)-1/glucose-regulatedprotein 75 (Grp75)/inositol 1,4,5-triphosphate receptor(IP3R)-1 complex, and regulating energy metabolism.Whereas mitochondrial dysfunction, ER stress, and al-tered Ca2+ homeostasis are associated with altered insulinsignaling, the implication of MAM dysfunctions in insulinresistance is unknown. Here we validated an approachbased on in situ proximity ligation assay to detect andquantify VDAC1/IP3R1 and Grp75/IP3R1 interactions atthe MAM interface. We demonstrated that MAM integrityis required for insulin signaling and that induction of MAMprevented palmitate-induced alterations of insulin signal-ing in HuH7 cells. Disruption of MAM integrity by geneticor pharmacological inhibition of the mitochondrial MAMprotein, cyclophilin D (CypD), altered insulin signaling inmouse and human primary hepatocytes and treatmentof CypD knockout mice with metformin improved bothinsulin sensitivity and MAM integrity. Furthermore, ER-mitochondria interactions are altered in liver of both ob/oband diet-induced insulin-resistant mice and improved by

rosiglitazone treatment in the latter. Finally, increasing or-ganelle contacts by overexpressing CypD enhanced in-sulin action in primary hepatocytes of diabetic mice.Collectively, our data reveal a new role of MAM integrityin hepatic insulin action and resistance, providinga novel target for the modulation of insulin action.

Mitochondria and endoplasmic reticulum (ER) networksare interconnected, sharing structural and functionalinteractions essential for the maintenance of cellularhomeostasis. The contacts between ER and mitochondria,known as mitochondria-associated ER membranes (MAMs),play a pivotal role in calcium (Ca2+) signaling, lipid transport,energy metabolism, and cell survival (1). The physical inter-actions between both organelles depend on complementarymembrane proteins, which tether the two organelles to-gether at specific sites. For example, the voltage-dependentanion channel (VDAC) of the outer mitochondrial mem-brane interacts with the inositol 1,4,5-triphosphate receptor(IP3R) on the ER through the molecular chaperone glucose-regulated protein 75 (Grp75), allowing Ca2+ transfer fromthe ER to mitochondria (2). Recently, mitofusin 2 (Mfn2)

1INSERM UMR-1060, CarMeN Laboratory, Lyon 1 University, INRA U1235, INSA ofLyon, Rockefeller and Charles Merieux Lyon-Sud Medical Universities, Lyon,France2INSERM UMR-1052, Cancer Research Center of Lyon, Lyon 1 University, Hos-pices Civils de Lyon, Lyon, France3Hospices Civils de Lyon, Louis Pradel Hospital, Cardiovascular Functional Ex-plorations Service and CIC of Lyon, Lyon, France4Hospices Civils de Lyon, Lyon-Sud Hospital, Endocrinology, Diabetology, andNutrition Service, Pierre-Bénite, France

Corresponding author: Jennifer Rieusset, jennifer.rieusset@univ-lyon1.fr.

Received 18 November 2013 and accepted 30 April 2014.

This article contains Supplementary Data online at http://diabetes.diabetesjournals.org/lookup/suppl/doi:10.2337/db13-1751/-/DC1.

P.T. and G.V. contributed equally to this work.

© 2014 by the American Diabetes Association. Readers may use this article aslong as the work is properly cited, the use is educational and not for profit, andthe work is not altered.

See accompanying article, p. 3163.

Diabetes Volume 63, October 2014 3279

SIG

NALTRANSDUCTIO

N

was discovered as a direct ER-mitochondria tether, regulatingalso the interactions and Ca2+ transfer between both organ-elles (3). Now it appears that MAM architecture involvesa large number of proteins with various functions (4).

Accumulating evidence suggests that both ER andmitochondria are key actors of energy homeostasis ofliver and that hepatic insulin resistance is associated withmitochondrial dysfunction (5), ER stress (6), and alteredlipid and Ca2+ homeostasis (7). Interestingly, the specificloss of Mfn2 in hepatocytes was recently associated withhepatic insulin resistance (8), whereas overexpression ofMfn2 improved diet-induced hepatic insulin resistance (9).In addition, different proteins of insulin-signaling pathwayswere recently identified at the MAM interface, such as pro-tein kinase B (PKB) (10), mammalian target of rapamycincomplex 2 (11), and phosphatase and tensin homolog (12).However, whether perturbation of organelle interactionsparticipates to hepatic insulin resistance is still unknown.

Until now, the visualization of ER-mitochondria con-tacts in situ in pathological contexts was limited by thelack of adapted analytical methods. Conventional imagingapproaches to assess ER-mitochondria interactions rely ontransmission electronic microscopy (13,14), electron to-mography (15,16), or immunocolocalization of ER- andmitochondria-specific fluorophores/fluorescent proteins(17). However, these strategies are not particularly sensi-tive and/or quantitative and not easily amenable to largescreening. Alternative methods have emerged using geneticapproaches, which may, however, limit analysis of interac-tions at endogenous expression levels of proteins (18).

We therefore sought to develop and validate a quanti-tative method to detect and quantify ER-mitochondriainteractions by in situ proximity ligation assay (PLA) andto study MAM implication in insulin signaling and theirpotential contribution to hepatic insulin resistance. Ourdata demonstrate that MAM are an important signalinghub contributing to insulin action and that their dysfunc-tion in liver is associated with hepatic insulin resistance.

RESEARCH DESIGN AND METHODS

Cell CultureHuH7 cells were cultured in DMEM (1 g/L glucose)supplemented with 10% FBS and antibiotics, at 37°Cand 5% CO2. For lipotoxicity studies, HuH7 cells weretreated with BSA or palmitate (200 mmol/L, 18 h) inserum-free culture medium. For mRNA analysis, HuH7cells were cultured for 6 h in serum-free DMEM (1 g/Lglucose) supplemented with forskolin (10 mmol/L) in thepresence or absence of insulin (1027 mol/L).

Primary mouse hepatocytes were isolated by a collage-nase perfusion method as previously described (19) andcultured in DMEM (3 g/L glucose) supplemented with10% FBS and antibiotics at 37°C and 5% CO2. Primaryhuman hepatocytes were isolated as previously described(20). All experiments were performed after 24 h of incuba-tion. For the measurement of glucose production, cells wereincubated for 24 h in Krebs buffer supplemented in lactate

(20 mmol/L), pyruvate (2 mmol/L), and forskolin (10 mmol/L)in the presence or absence of insulin (1027 mol/L). FormRNA analysis, hepatocytes were cultured for 6 h inserum-free DMEM (3 g/L glucose) supplemented in for-skolin (10 mmol/L) in the presence or absence of insulin(1027 mol/L).

DNA ConstructsHuman pcDNA4/TO-Grp75 vector was purchased fromCliniSciences. Human pcDNA3-Mfn2 vector was generatedfrom IMAGE clone (pCMV-SPORT6-MFN2, Life Science) bydigestion with EcoRI and NotI and subcloning into thepcDNA3 vector (Invitrogen). The constructs used for in-ducing tethering between mitochondria and ER (AKAP1-FKBP-YFP and CFP-FRB-ER) were a generous gift of GyorgyHajnoczky and were used as previously described (16,18).Recombinant adenoviral genome carrying cyclophilin D(Ad-CypD) was generated from a vector encoding rat cyclo-philin D (CypD) (21) by homologous recombination in theVmAdcDNA3 plasmid and amplified in HEK293 cells (22).

Modulation of Gene ExpressionHuH7 cells were transfected in six-well plates for 48 hwith 2 mg expression plasmids for specific target or withthe empty vector as control using ExGen 500 transfec-tion reagent (Roche Diagnostics). For silencing experi-ments, HuH7 cells were transfected with specific siRNA(Qiagen) for VDAC1 (25 nmol/L), Grp75 (50 nmol/L),IP3R1 (50 nmol/L), and Mfn2 (25 nmol/L) using HighPerfect transfection reagent (Qiagen) following the techni-cal recommendations.

ImmunofluorescenceHuH7 cells or primary hepatocytes were fixed with 4%paraformaldehyde (10 min at room temperature) andpermeabilized using 0.1% triton 3100 (15 min at roomtemperature). Then saturation was performed using PBS-BSA 1% for 1 h, followed by incubation with primaryantibodies at 4°C overnight: IP3R1 (Santa Cruz, 1/500,rabbit), Grp75 (Santa Cruz, 1/500, mouse), CypD (Abcam,1/500, mouse), VDAC1 (Abcam, 1/100, mouse), OXPHOS(MitoSciences, 1/500, mouse), ANT (MitoSciences, 1/500,mouse), and SERCA2 (Cell Signaling, 1/500, rabbit) (Sup-plementary Fig. 1A). After several washes, the secondaryantibodies coupled to a fluorochrome (Alexa Fluor) wereadded for 1 h. Slides were washed three times in PBS andmounted in VECTASHIELD mounting medium (H-1200).

Duolink Proximity Ligation In Situ AssayDuolink II in situ PLA (Olink Bioscience) enables thedetection, visualization, and quantification of protein inter-actions (,40 nm) as an individual fluorescent dot by mi-croscopy. Cell fixation and permeabilization were performedsimilarly to immunofluorescence. The proximity ligationswere performed according to the manufacturer’s protocol.Preparations were mounted in Duolink II mounting me-dium containing DAPI 18 (Eurogentec). Fluorescence wasanalyzed with a Zeiss inversed fluorescent microscopeequipped with an ApoTome using the AxioVision program.

3280 Disruption of MAM in Hepatic Insulin Resistance Diabetes Volume 63, October 2014

Quantification of signals was done with the BlobFindersoftware (Centre for Image Analysis, Uppsala University)and expressed as percentage of blobs per nucleus com-pared with the controls. Experiments were performed atleast three times, with a minimum of five fields taken percondition.

Animal StudiesAnimal studies were performed in accordance with theFrench guidelines for the care and use of animals andwere approved by a regional ethic committee. MaleC57BL/6J and ob/ob mice (12 weeks old) were purchasedfrom Harlan and adapted to the environment for 1 weekbefore study. Four-week-old C57BL/6J males were eitherfed with a standard diet (SD) or with a high-fat and high-sucrose diet (HFHSD; prepared by the Unit of Experimen-tal Food Preparation, Jouy-en-Josas, France) for 16 weeksas previously described (23). One group of HFHSD micewere treated with rosiglitazone (20 mg/kg in the food)during the last 4 weeks of feeding. CypD knockout (KO)male mice were an initial gift from S.J. Korsmeyer’s lab-oratory (Dana Farber Cancer Institute, Boston, MA) (24),and mice were backcrossed for eight generations on aC57BL/6 genetic background. Wild-type (WT) and CypD-KO mice (24 weeks old) were daily force fed with eithermethylcellulose or metformin (300 mg/kg/day) for 4 weeks.Glucose, insulin, and pyruvate tolerance tests were per-formed on 6 h–fasted mice. Glucose (2 mg/g body weight),insulin (0.75 mU/g body weight), or pyruvate (2 mg/g bodyweight) were injected intraperitoneally, and blood glucoselevels were monitored using a glucometer at the indicatedtime points.

Western BlotPreparation of total lysates, electrophoresis by SDS-PAGE (10, 12, or 4–15% ready gels, Bio-Rad), immunode-tection, and quantification were performed as previouslydescribed (25).

Isolation of MAMIsolation of MAM fractions were performed by differen-tial ultracentrifugation as previously described (26).

Statistical AnalysisData are expressed as the mean 6 SEM. Statistical anal-ysis was performed with the StatView software. Com-parisons among more than two groups were analyzed byone-way ANOVA. When a significant F value was obtained,multiple comparisons were made using a Fisher protectedleast significant difference test. For all other analyses, datawere compared by Student t test. Statistical significancewas defined as a value of P , 0.05.

RESULTS

Monitoring Perturbations of ER-MitochondriaInteractions by In Situ PLA in Fixed CellsWe used in situ PLA to detect and quantify interactionsbetween VDAC1 and IP3R1, two organelle-surface proteinsinvolved in the Ca2+ channeling complex at the MAM

interface. We probed one protein of the outer mitochon-drial membrane and one protein of the MAM with mouseanti-VDAC1 and rabbit anti-IP3R1 primary antibody (Sup-plementary Fig. 1B, panel a) and then with anti-mouse IgGand anti-rabbit IgG that are conjugated to oligonucleotideextensions. If the oligonucleotides are at the requisite dis-tance (,40 nm), they hybridize with the subsequentlyadded connector oligos to form a circular DNA template(Supplementary Fig. 1B, panel b), which is ligated and sub-sequently amplified to create a ssDNA product (Supple-mentary Fig. 1B, panel c). Since the distance between theER and mitochondria at contact sites ranges from 10 to 25nm, we hypothesized that the ER-mitochondria junctionwould enable proximity ligation and subsequent detectionby hybridization of Texas red-labeled oligonucleotidesprobes (Supplementary Fig. 1B, panel d). Each fluorescentdot would represent the formation of one VDAC1/IP3R1interaction, thus allowing quantification of in situ ER-mitochondria interactions in individual cells. Similar insitu PLA experiments were also performed with theGrp75/IP3R1 pair of antibodies. As prerequisites, we vali-dated the primary antibodies by both gene silencing strategyand immunocolocalization experiments in HuH7 cells (Sup-plementary Fig. 2A and B) and confirmed the colocalizationof VDAC1, Grp75, and IP3R1 by conventional immunoflu-orescence in HuH7 cells (Supplementary Fig. 3), indicatingthat these proteins are good candidates to study ER-mitochondria interactions. Then we investigated interac-tions between endogenous VDAC1/IP3R1, VDAC1/Grp75,or Grp75/IP3R1 proteins in fixed HuH7 cells and observedfluorescent signals for all pairs of proteins (SupplementaryFig. 4), indicating that the VDAC1/Grp75/IP3R1 complexcan be visualized by in situ PLA. Appropriate controls con-firmed the specificity of the assay (Supplementary Fig. 4).

Next we challenged organelle coupling by silencing oroverexpressing MAM proteins and quantified thesechanges by in situ PLA in HuH7 cells. Silencing ofVDAC1 (245%; P , 0.05) (Supplementary Fig. 2A) in-duced a significant reduction of VDAC1/IP3R1 interac-tions (299.5%; P , 0.001) (Fig. 1A). Interestingly, thissilencing of VDAC1 also reduced the Grp75/IP3R1 signalcounts (241%; P , 0.001) (Fig. 1A), indicating a reduc-tion of ER-mitochondria interactions. Silencing of Grp75corroborates these results since we observed that down-regulation of Grp75 (264%; P , 0.05) (SupplementaryFig. 2A) reduced VDAC1/IP3R1 (247%; P , 0.05) andGrp75/IP3R1 (257%; P , 0.005) interactions, comparedwith control cells (Fig. 1B). Furthermore, inhibiting eitherIP3R with xestospongin C (1 mmol/L, 18 h) or the IP3production with a PLC inhibitor (U73122, 1 mmol/L,18 h) also reduced VDAC1/IP3R1 interactions (222 and241%, respectively; P , 0.05) (Supplementary Fig. 5A),indicating that pharmacological approach is also an effi-cient strategy to alter MAM integrity. At the opposite, theinduction of Grp75 expression significantly induced theinteractions of both VDAC1/IP3R1 and Grp75/IP3R1 pro-teins (+67 and +70%, respectively; P , 0.05) (Fig. 1C).

diabetes.diabetesjournals.org Tubbs and Associates 3281

Figure 1—Monitoring perturbation of ER-mitochondria interactions linked to genetic modulation of protein of the VDAC1/Grp75/IP3R1 complexby in situ PLA. Representative PLA images (at left,363 and scale bar = 20 mm) and quantitative analysis of both Grp75/IP3R1 and VDAC1/IP3R1interactions (at right) in HuH7 cells transfected either with specific siRNA or transient expressing vectors during 48 h. Genetic invalidation ofVDAC1 (A) and Grp75 (B) significantly reduce Grp75/IP3R1 as well as VDAC1/IP3R1 interactions. Transient overexpression of Grp75 significantlyincreases both Grp75/IP3R1 and VDAC1/IP3R1 interactions (C). Silencing of Mfn2 decreases both VDAC1/IP3R1 and Grp75/IP3R1 interactions(D), whereas overexpression of Mfn2 significantly increases them (E). Linker-mediated tethering between mitochondria and ER following rapa-mycin treatment (100 nmol/L, 2 min) induces VDAC1/IP3R1 interactions in HuH7 cells (F). *P < 0.05 vs. respective control. n = 3 independentexperiments. Co, control.

3282 Disruption of MAM in Hepatic Insulin Resistance Diabetes Volume 63, October 2014

Then we modulated organelle coupling independentlyof the VDAC1/Grp75/IP3R1 complex by modulatingeither the expression of Mfn2, a direct tether of bothorganelles whose invalidation reduced ER-mitochondriainteractions (3), or by overexpressing an artificial tether(16,18). Reduction of Mfn2 expression (258%; P , 0.001)(Supplementary Fig. 2A) using a specific siRNA significantlyreduced the in situ PLA signal for both VDAC1/IP3R1 andGrp75/IP3R1 (262 and 266%, respectively; P , 0.05)(Fig. 1D). In contrast, we observed an increase of bothVDAC1/IP3R1 and Grp75/IP3R1 interactions (+48 and+64%, respectively; P , 0.05) (Fig. 1E) following the tran-sient overexpression of Mfn2 (+593%; P , 0.003) (Supple-mentary Fig. 2A). Alternatively, we transfected HuH7 cellswith CFP-FRB-ER and OMM-FKBP-YFP constructs (16,18)and then added rapamycin (100 nmol/L) for 2 min to in-crease ER-mitochondria contacts. Rapamycin treatmentsignificantly increased the VDAC1/IP3R1 signal counts(+56%; P , 0.04) (Fig. 1F), confirming that in situ PLAis a reliable method to investigate MAM integrity and or-ganelle interactions.

Perturbations of ER-Mitochondria Interactions Impacton Insulin SignalingWe then explored whether genetically induced perturba-tions of ER-mitochondria interactions could affect insulinaction in HuH7 cells. Insulin action was estimated at thelevel of the signaling cascade by measuring phosphoryla-tion of insulin receptor substrate 2 (IRS2), PKB, andglycogen synthetase kinase 3b (GSK3b), as well as at thelevel of gene expression by measuring neoglucogenic geneexpression, such as the glucose-6-phosphatase (G6P) andthe phosphoenolpyruvate carboxykinase (PEPCK). Silenc-ing of Grp75 (Fig. 2A) or Mfn2 (Fig. 2B) reduced insulin-stimulated phosphorylation of IRS2, PKB, and GSK3b andsignificantly prevented the inhibitory action of insulin ofG6P and PEPCK mRNA levels (Fig. 2C). Pharmacologicalinhibition of IP3R, using xestospongin C, also reducedinsulin-stimulated PKB phosphorylation (241%; P , 0.05)(Supplementary Fig. 3B). On the other hand, overexpres-sion of Grp75 or Mfn2 increased both insulin signaling(Fig. 2D and E, respectively) and action (Fig. 2F). Further-more, palmitate treatment of HuH7 cells, which alteredinsulin signaling (Fig. 3A) and action on gene expression(Fig. 3B), was associated with a marked reduction ofVDAC1/IP3R1 interaction (Fig. 3C). Importantly, palmi-tate inhibitory action on insulin signaling could be pre-vented by Mfn2 or Grp75 overexpression, whichincreased organelle interactions (Fig. 3D and E, respec-tively). The opposite modulation of MAM by silencingexperiments did not amplify palmitate-induced altera-tions of insulin signaling (data not shown). Altogether,these results indicate that there is a strong relationshipbetween MAM integrity and insulin action in the humanhepatic cell line.

Using conventional Percoll gradient fractionation, weisolated MAM fractions from liver of fasted mice injected

or not with insulin. Fraction purity was verified byelectronic microscopy (Supplementary Fig. 6). Insulin-stimulated PKB Ser 473 phosphorylation in both liverhomogenates and MAM fractions (Supplementary Fig.7A) suggests that part of phosphorylated PKB localizedat the MAM interface in response to insulin. Interest-ingly, the effect of insulin on PKB phosphorylation ishigher in MAM fractions compared with homogenate(Supplementary Fig. 7A, #P , 0.0001), suggesting thatMAM could contribute to insulin signaling in hepatocytes.We further demonstrated by in situ PLA that both PKBand p(S473)PKB were in close proximity with IP3R1 (Sup-plementary Fig. 7B) and that insulin treatment increasedboth PKB/IP3R1 and p(S473)PKB/IP3R1 interactionswithout changing VDAC1/IP3R1 interactions (Supple-mentary Fig. 7B). These data support a recruitment ofp(S473)PKB at the MAM interface in response to insulin.

Suppression of Organelle Coupling In Vivo Is Linked toHepatic Insulin ResistanceRecently, we identified CypD as a new mitochondrial proteininteracting with the VDAC1/Grp75/IP3R1 Ca2+-channelingcomplex at the MAM interface in cardiomyocytes (27). Weconfirmed that CypD/IP3R1 interactions could be visual-ized by in situ PLA in primary hepatocytes of WT mice.The absence of CypD/IP3R1 signals in CypD-KO hepato-cytes confirmed the specificity of the assay (Supplemen-tary Fig. 8A). We then found a significant reduction ofMAM amount in liver of CypD-KO mice compared withWT mice (Fig. 4A), confirming the implication of CypD inMAM integrity also in liver. Interestingly, CypD-KO micealso showed altered responses to insulin (SupplementaryFig. 8B) and pyruvate (Fig. 4B) tolerance tests, indicatinga state of insulin resistance and an increase of hepaticneoglucogenesis in this model. Furthermore, CypD-KOmice treated during 4 weeks with metformin displayedimproved insulin sensitivity (Supplementary Fig. 8B)and hepatic neoglucogenesis (Fig. 4B) and, concomitantly,an increase in MAM amount in liver (Fig. 4A), indicatingassociation between MAM integrity and hepatic insulinaction in vivo in mice. To define this association, we in-vestigated insulin action and VDAC1/IP3R1 interactions inprimary hepatocytes of these mice. Alterations of insulinsignaling (Fig. 4C) and action, on both gene expression(Fig. 4D) and hepatic glucose production (Fig. 4E), wereclearly associated with a reduction of the VDAC1/IP3R1signal counts (Fig. 4F) in CypD-KO hepatocytes comparedwith WT hepatocytes. Interestingly, primary hepatocytesprepared from metformin-treated CypD-KO mice had im-proved insulin action (Fig. 4C–E) and increased MAMinteractions (Fig. 4F). We further extended these obser-vations to human primary hepatocytes, in which additionof NIM811 (2 mmol/L, 48 h), a specific inhibitor of CypD,significantly reduced CypD/IP3R1 and VDAC1/IP3R1interactions (257 and 234%, respectively; P , 0.05)(Fig. 5A) and decreased insulin-stimulated PKB phosphor-ylation (237%; P , 0.05) (Fig. 5B).

diabetes.diabetesjournals.org Tubbs and Associates 3283

Figure 2—Disruption of MAM integrity alters insulin signaling in HuH7 cells. HuH7 cells were transfected either with specific siRNA ortransient expressing vectors during 48 h, depleted for 3 h in serum, and treated with insulin (1027 mol/L, 15 min for insulin-signalinganalysis, 6 h for mRNA analysis). Representative Western blots (at top) and quantitative analysis (below) of insulin-stimulated IRS2, PKB,and GSK3b phosphorylation following silencing of Grp75 (A), Mfn2 (B) or transient expression of Grp75 (D) or Mfn2 (E). Effect of a 6 h insulintreatment on G6P and PEPCK mRNA levels measured by real-time RT-PCR in HuH7 cells silenced (C ) or overexpressing (F ) Mfn2 or Grp75.*P < 0.05; **P < 0.01; ***P < 0.001 vs. respective control. n = 3. Co, control.

3284 Disruption of MAM in Hepatic Insulin Resistance Diabetes Volume 63, October 2014

Figure 3—Reinforcement of MAM prevents palmitate-induced alterations of insulin action. A: Representative Western blots (at left) andquantitative analysis (at right) of insulin-stimulated IRS2, PKB, and GSK3b phosphorylation in BSA or palmitate-treated HuH7 cells for 24 h.*P< 0.05; **P< 0.01 vs. WT. n = 3. B: Effect of a 6 h insulin treatment on G6P and PEPCK mRNA levels measured by real-time RT-PCR onBSA or palmitate-treated HuH7 cells for 18 h. *P < 0.05. n = 3. C: Representative PLA images (at left, 363 and scale bar = 20 mm) andquantitative analysis (at right) of VDAC1/IP3R1 interactions in HuH7 cells treated with BSA or palmitate during 18 h. ***P < 0.0001. n = 3.D and E: HuH7 cells were transfected with transient expressing vectors for Mfn2 (D) or Grp75 (E), incubated with BSA or palmitate(200 mmol/L) for 24 h, and treated with insulin (1027 mol/L, 15 min). Note that three parts of the same gel are shown. *P < 0.05; #P <0.01. n = 3. Co, control; Palm, palmitate.

diabetes.diabetesjournals.org Tubbs and Associates 3285

Figure 4—CypD-mediated disruption of MAM integrity induces hepatic insulin resistance. WT and CypD-KO mice (24 weeks old) were dailyforce fed with either methylcellulose (as control) or metformin (300 mg/kg/day, CypD-KO + metformin [CypD-KO+Metf.]) for 4 weeks. A:Quantitative analysis of proteins in MAM fractions following cellular subfraction of liver of WT, CypD-KO, and CypD-KO + metformin mice.***P < 0.0005 vs. WT; #P < 0.05 vs. CypD-KO mice. n = 5. B: Pyruvate tolerance tests were performed on WT, CypD-KO, and metformin-treated CypD-KO mice. *P < 0.05 vs. WT mice; #P < 0.05 vs. CypD-KO mice. n = 10. C: Representative Western blots (at top) and

3286 Disruption of MAM in Hepatic Insulin Resistance Diabetes Volume 63, October 2014

Alteration of Organelle Coupling in Liver of Diet-Induced Insulin-Resistant MiceNext we measured organelle coupling in liver of geneti-cally or diet-induced insulin-resistant mice. As expected,both ob/ob and 16-week HFHSD mice are glucose intoler-ant (Supplementary Fig. 9A and C) and insulin resistant(Supplementary Fig. 9B and D) compared with controlmice. Furthermore, ob/ob mice showed increased neoglu-cogenesis during pyruvate tolerance test (Fig. 6A). Inaddition, the treatment of HFHSD mice with the anti-diabetic agent rosiglitazone during the last 4 weeks offeeding improved responses to glucose and insulin tests(Supplementary Fig. 9A and B). Interestingly, the meta-bolic alterations of both ob/ob and HFHSD mice wereassociated with a reduction of ER-mitochondria interac-tions in liver (Figs. 6B and 7A, respectively), whereas rosi-glitazone treatment was associated with an increase ofMAM (+16%; P , 0.05) (Fig. 7A) in HFHSD mice. Fur-thermore, when investigating the expression of MAM-associated proteins in the MAM fractions in liver ofHFHSD-fed mice, we found marked alteration of keyMAM protein (Fig. 7B) when compared with SD mice,with noticeable reduction of VDAC1, CypD, and PACS2amounts, increased levels of Mfn2, and no change ofIP3R1. In addition, basal phosphorylation of PKB was in-creased in MAM of HFHSD mice compared with SD mice(Fig. 7B). Interestingly, rosiglitazone treatment was ableto restore CypD and PACS2 expression, as well as thephosphorylation levels of PKB (Fig. 7B).

We further analyzed insulin action and MAM integrityby in situ PLA in primary hepatocytes of both models ofinsulin resistance. In agreement with in vivo data, thealterations of insulin signaling and/or action are asso-ciated with a significant reduction of VDAC1/IP3R1,Grp75/IP3R1, and/or CypD/IP3R1 interactions in bothob/ob (Fig. 6C–E) and HFHSD (Fig. 7C and D) hepatocytescompared with their respective controls, confirmingdisruption of MAM integrity in hepatocytes of diabeticmice. However, primary hepatocytes prepared fromrosiglitazone-treated mice did not show beneficial ef-fect on insulin-stimulated PKB and GSK3b phosphoryla-tion (Fig. 7C), as well as on protein interactions at theMAM interface (Fig. 7D). These results suggest that theeffect of this antidiabetic agent, well measurable in crudeliver samples (Fig. 7A), was lost after plating cells inculture.

Improvement of ER-Mitochondria InteractionsRestores Hepatic Insulin SignalingTo further confirm the strong relationship between MAMintegrity and insulin action, we sought to restore organ-elle coupling in hepatocytes of diabetic mice and todetermine whether it resulted in enhanced insulin action.As overexpression of MAM proteins increased organellecoupling in HuH7 cells, we reasoned that it would likelyalso be effective in primary hepatocytes of diabetic mice.Therefore, we chose to overexpress CypD in ob/ob and inHFHSD hepatocytes. Firstly, we controlled that the alter-ations of organelle interaction and insulin action in ob/obhepatocytes were conserved after adenovirus infection ofthe cells (Fig. 8A–C). Then overexpression of CypD in-creased the VDAC1/IP3R1 interactions compared withob/ob hepatocytes infected with a green fluorescent pro-tein (GFP)-expressing adenovirus (Fig. 8A), indicatingthat the strategy was successful in reconstituting struc-tural ER-mitochondria interactions ex vivo. Importantly,CypD overexpression in ob/ob hepatocytes also improvedinsulin signaling (Fig. 8B) and the metabolic action of thehormone (Fig. 8C). Similar types of results (Fig. 8D–F)were obtained when overexpressing CypD (Supplemen-tary Fig. 10) in HFHSD primary hepatocytes. Altogether,these data demonstrated that improving MAM integrityin primary hepatocytes of diabetic mice is able to restorehepatic insulin action.

DISCUSSION

In this study, we validated an approach based on in situPLA to detect and quantify ER-mitochondria interactionsand we revealed for the first time that the MAM interfaceis an important novel player in the regulation of hepaticinsulin action and that disruption of MAM could beinvolved in hepatic insulin resistance.

It has been suggested that ER-mitochondria interac-tions can be regulated by drug exposure (28) or by diseasestates (29), but until now, MAM integrity was difficult toaccess in vivo with conventional imaging approaches. Wesought to use in situ PLA to quantify ER-mitochondriainteractions in fixed cells, and to our knowledge, this isthe first application of in situ PLA to detect organelleinteractions. This technique raises several interestingadvantages: 1) in situ PLA confers dual-binder specificityfor the detection of organelle interactions in situ andreveals proximity of proteins in normal cells without

quantitative analysis (below) of insulin-stimulated insulin receptor, PKB, and GSK3b phosphorylation in primary hepatocytes of WT andboth methylcellulose- or metformin-treated CypD-KO mice. Hepatocytes were depleted of serum for 3 h before stimulation with insulin (1027

mol/L, 15 min). *P < 0.01 vs. WT; #P < 0.05 vs. untreated CypD-KO mice. n = 5. D: Inhibitory effect of 6 h insulin treatment (in percentages)on PEPCK and G6P mRNA levels measured by real-time PCR on primary hepatocytes of WT, CypD-KO, and metformin-treated CypD-KOmice. *P < 0.05; **P < 0.01 vs. WT; #P < 0.05 vs. untreated CypD-KO mice. n = 5. E: Inhibitory effect of insulin (in percentages) of glucoseproduction on primary hepatocytes of WT, CypD-KO, and metformin-treated CypD-KO mice. *P < 0.0001 vs. WT; #P < 0.01 vs. untreatedCypD-KO mice. n = 5. F: In situ PLA images (at left, 363 and scale bar = 20 mm) and quantitative analysis (at right) of VDAC1/IP3R1interactions in primary hepatocytes of WT, CypD-KO, and metformin-treated CypD-KO mice. *P < 0.05 vs. WT; #P < 0.0001 vs. untreatedCypD-KO mice. n = 3. HGP, hepatic glucose production; IR, insulin receptor.

diabetes.diabetesjournals.org Tubbs and Associates 3287

being subject to artifacts of overexpression or ectopic ex-pression (30); 2) because the signal produced by in situPLA can be amplified, the method is exquisitely sensitive,ensuring that transient and weak interactions can be vi-sualized as a single spot and quantified; and 3) the systemis rather simple to use and suitable for multiple conditiontesting. In this study, we have carefully validated themethod to ensure that in situ PLA is suitable to visualizeER-mitochondria coupling. In addition, we found very re-producible results when targeting several pairs of proteins

(VDAC/IP3R, Grp75/IP3R, CypD/IP3R) at the MAM in-terface, strongly supporting the robustness of the tech-nique to monitor organelle coupling. However, the factthat the in situ PLA staining involved almost all the cel-lular volume of hepatocytes contrasts with previous datareporting that MAM may represent ;20% of total mito-chondrial network (17). This could rely on the fact thatimages were analyzed by conventional fluorescent micros-copy, allowing simultaneous visualization of protein inter-actions at different focal planes. An important result is

Figure 5—CypD-mediated disruption of MAM integrity alters insulin signaling in human primary hepatocytes. A: In situ PLA images (at left,363 and scale bar = 20 mm) and quantitative analysis (at right) of both CypD/IP3R1 and VDAC1/IP3R1 interactions in human primaryhepatocytes treated or not with NIM811 (2 mmol/L, 18 h), a specific inhibitor of CypD. NIM811 treatment reduces interactions in humanprimary hepatocytes. *P < 0.05; **P < 0.001 vs. control. n = 3 different subjects. B: Three independent Western blots (at left) andquantitative analysis of insulin-stimulated PKB phosphorylation in human primary hepatocytes pretreated or not with NIM811. Hepatocyteswere treated with ethanol (as control) or 2 mm NIM811 for 48 h and depleted of serum for 3 h before stimulation with insulin (1027 mol/L,15 min). NIM811 treatment reduces insulin-stimulated S473 phosphorylation of PKB in human primary hepatocytes. *P < 0.05 vs. control.n = 3 different subjects. Co, control.

3288 Disruption of MAM in Hepatic Insulin Resistance Diabetes Volume 63, October 2014

Figure 6—Disruption of MAM integrity and insulin signaling in liver of ob/ob mice. A: Pyruvate tolerance test (2 mg/g body weight) performedon 6 h–fasted WT and ob/ob mice. *P < 0.001 vs. WT mice. n = 10. B: Quantitative analysis of proteins in MAM fractions following cellularsubfraction of liver of WT and ob/obmice. *P< 0.05 vs. WT. n = 5. C: Representative Western blots (at left) and quantitative analysis (at right) ofinsulin-stimulated insulin receptor, PKB, and GSK3b phosphorylation in primary hepatocytes of WT and ob/ob mice. Hepatocytes weredepleted of serum for 3 h before stimulation with insulin (1027 mol/L, 15 min). *P < 0.05 vs. WT. n = 8. D: Inhibitory effect of insulin(in percentages) on glucose production of WT and ob/ob hepatocytes. *P < 0.01 vs. WT. n = 8. E: In situ PLA images (at top, 363 and scalebar = 20 mm) and quantitative analysis (at bottom) of both CypD/IP3R1 and VDAC1/IP3R1 interactions in primary hepatocytes of WT and ob/obmice. *P < 0.05 vs. WT; **P < 0.0001 vs. WT. n = 5. HGP, hepatic glucose production; IR, insulin receptor.

diabetes.diabetesjournals.org Tubbs and Associates 3289

Figure 7—Disruption of MAM integrity and insulin signaling in liver of diet-induced diabetic mice. A: Quantification of protein amount inMAM fractions following cellular subfraction of liver of SD, HFHSD, and HFHSD + rosiglitazone mice. *P < 0.05 vs. SD mice; #P < 0.05 vs.untreated HFHSD mice. n = 5. B: Representative Western blots (at top) and quantitative analysis (below) of proteins in MAM fractions of liverof SD, HFHSD, and HFHSD + rosiglitazone mice. C: Representative Western blot (at top) and quantitative analysis (below) of insulin-stimulated PKB and GSK3b phosphorylation in primary hepatocytes of SD, HFHSD, and HFHSD + rosiglitazone mice. *P < 0.05;**P < 0.001 vs. respective control; #P < 0.01 vs. SD mice with insulin. n = 3. D: In situ PLA images (at left, 363 and scale bar = 20mm) and quantitative analysis (at right) of CypD/IP3R1, VDAC1/IP3R1, and Grp75/IP3R1 interactions in primary hepatocytes of SD, HFHSD,and HFHSD + rosiglitazone mice. *P < 0.01 vs. SD mice; **P < 0.001 vs. SD mice. n = 3. Rosi, rosiglitazone.

3290 Disruption of MAM in Hepatic Insulin Resistance Diabetes Volume 63, October 2014

Figure 8—Restoration of MAM integrity by CypD overexpression improves insulin signaling in primary hepatocytes of diabetic mice.Primary hepatocytes of WT, ob/ob, or HFHSD mice were infected with adenovirus encoding for GFP (Ad-GFP, as control) or CypD (Ad-CypD) for 36 h. A: Representative in situ PLA images (at top, 363 and scale bar = 20 mm) and quantification (at bottom) of VDAC1/IP3R1interactions in primary hepatocytes of infected WT and ob/ob mice. *P < 0.05 vs. WT Ad-GFP mice; #P < 0.001 vs. CypD-KO Ad-GFPmice. n = 5. B: Representative Western blots (at top) and quantitative analysis (below) of insulin-stimulated insulin receptor, PKB, andGSK3b phosphorylation in primary hepatocytes of infected WT and ob/obmice. Mouse primary hepatocytes were depleted of serum for 3 hbefore stimulation with insulin (1027 mol/L, 15 min). *P < 0.05 vs. WT Ad-GFP mice; #P < 0.05 vs. CypD-KO Ad-GFP mice. n = 6. C:Inhibitory effect of insulin (in percentages) on glucose production of primary hepatocytes of infected WT and ob/ob hepatocytes.*P < 0.005 vs. WT Ad-GFP mice; #P < 0.005 vs. CypD-KO Ad-GFP mice. n = 3. D: Representative in situ PLA images (at top, 363 andscale bar = 20 mm) and quantification (below) of both CypD/IP3R1 and VDAC1/IP3R1 interactions in primary hepatocytes of infected HFHSDmice. *P < 0.05 vs. Ad-GFP mice. n = 4. E: Representative Western blots (at top) and quantitative analysis (below) of insulin-stimulated PKB

diabetes.diabetesjournals.org Tubbs and Associates 3291

that organelle interactions can be modulated by disrup-tion or reinforcement of the interorganellar protein link-age, with different efficiencies depending on the targetedprotein, suggesting that both the nature and the amountof tethering proteins directly control ER-mitochondriacontacts in agreement with previous studies (3,31–33).

Using this method to evaluate ER-mitochondria inter-actions, we found that MAM integrity is required forefficient insulin action in hepatocytes. Indeed, we dem-onstrated in vitro that hepatic insulin signaling andaction could be diminished or strengthened by disruptionor enforcement of organelle interactions, respectively.This finding was confirmed whatever the targeted pro-teins used at the MAM interface (IP3R, Grp75, Mfn2, orCypD). Furthermore, we found that palmitate-inducedalterations of insulin action in HuH7 cells are associatedwith a decrease in MAM integrity, while the inductionof organelle interactions is able to prevent palmitate-induced insulin resistance. These in vitro data wereconfirmed and extended in vivo in mice and in mouseand human primary hepatocytes. Indeed, this closeassociation between MAM integrity and hepatic insulinaction was observed in three different mice models,including CypD-KO mice, ob/ob mice, and HFHSD-fedmice. Importantly, pharmacological improvement of in-sulin sensitivity using metformin or rosiglitazone inCypD-KO and HFHSD insulin-resistant mice improvedER-mitochondria interactions. Reciprocally, increasingMAM amount in primary hepatocytes of diabetic miceby overexpressing CypD enhanced insulin signaling andaction. Altogether, these data strongly support a closefunctional relationship between MAM integrity andhepatic insulin response.

Considering this strong association, it is still, however,difficult to conclude that MAM dysregulation is the causeof hepatic insulin resistance in diabetic state or thataltered insulin action contributes to MAM defects. In-deed, we found parallel alterations in MAM amount andinsulin action in vivo in mice liver and concomitantrestoration upon antidiabetic treatments. The fact thatthe beneficial effects of metformin, but not of rosiglita-zone, were conserved in primary hepatocytes preparedfrom the treated animals suggested the possibility ofdifferent regulatory mechanisms of MAM dynamics,depending on the pharmacological treatment. This mightbe related to the well-known different mechanisms ofaction of these drugs, with metformin known to actprimarily in liver (and potentially directly on mitochon-dria), while rosiglitazone affects hepatic metabolism

indirectly through activation of peroxisome proliferator–activated receptor g in adipose tissue and modulation ofadiponectin production. In line with our data, recent stud-ies demonstrated that loss or mutations of MAM proteinscould be associated with impaired glucose homeostasis inmice (8,34) and that liver-specific overexpression ofa MAM protein, such as Mfn2, is sufficient to improvediet-induced insulin resistance (9). However, whether thisimprovement involves a reinforcement of MAM integritywas not evaluated in these works, and thus additionalstudies are clearly required to demonstrate that an im-provement of MAM integrity in vivo could correct hepaticinsulin resistance.

In line with recent studies (10,11), we also found thatPKB, an important signaling protein mediating the met-abolic effects of insulin, is recruited at the MAM inter-face and phosphorylated on S473 in response to insulin.The fact that interactions between both PKB/IP3R1 andp(S473)-PKB/IP3R1 were found in hepatocytes using insitu PLA strongly supports a true localization of PKB atthe MAM interface. Importantly, we further demon-strated that acute insulin treatment increased theseinteractions without modifying VDAC1/IP3R1 interac-tions, indicating that insulin is likely able to induce a re-cruitment of PKB at the MAM interface. As theinteractions of both phosphorylated and total PKB pro-tein with IP3R1 were increased in response to insulintreatment, it is presently not clear whether phosphory-lation is required for the recruitment of PKB at the MAMinterface or whether PKB is directly phosphorylated inMAM. The presence of mammalian target of rapamycincomplex 2 at the MAM interface (11) might suggest thatPKB could be directly phosphorylated in MAM. Theseresults reinforce the concept that there are differentpools of PKB in hepatocytes, and thus MAM might bea novel, not yet considered, insulin-signaling hub in hep-atocytes. Therefore, loss of MAM integrity could be di-rectly involved in hepatic insulin resistance. Thishypothesis is also supported by the observation thatdiet-induced diabetic mice were characterized by a reduc-tion of VDAC1, CypD, and PACS2 expression and anincrease of PKB phosphorylation at MAM interfaces.These defects were partly restored by rosiglitazone treat-ment, suggesting, therefore, that the relationship be-tween MAM integrity and hepatic insulin sensitivitycould involve the regulation of MAM proteins. In agree-ment, it was reported that PACS2 (31) and CypD (27) arerequired for MAM integrity and that high PKB activity isassociated with reduced Ca2+ flux from the ER through

phosphorylation in primary hepatocytes of infected HFHSD mice. Mouse primary hepatocytes were depleted of serum for 3 h beforestimulation with insulin (1027 mol/L, 15 min). Note that two parts of the same gel originally separated by irrelevant lanes are shown. *P <0.01 vs. Ad-GFP with insulin. n = 4. F: Inhibitory effect (in percentages) of a 6 h insulin treatment (1027 mol/L) on PEPCK and G6P mRNAlevels in primary hepatocytes of infected HFHSD hepatocytes. *P< 0.05 vs. HFHSD Ad-GFP mice. n = 3. HGP, hepatic glucose production;IR, insulin receptor.

3292 Disruption of MAM in Hepatic Insulin Resistance Diabetes Volume 63, October 2014

the IP3R (35,36). On the other hand, we found that theexpression of Mfn2 is increased at the MAM interface ofdiabetic mice, whereas loss (8) or overexpression (34) ofMfn2 was, respectively, associated with hepatic insulinresistance or increased insulin sensitivity, suggestingpotential adaptive mechanisms.

Interestingly, PKB has been reported to phosphorylatethe IP3R channel, reducing its IP3-dependent Ca2+ releasecapacity (35,36). Here we found that inhibition of IP3Ractivity (by xestospongin C or by U73122) altered MAMintegrity. This may indicate a potential link between IP3-dependent ER-calcium release and MAM in liver cells. Thefact that diabetic mice were characterized by increasedphosphorylated PKB in MAM fraction, and also that in-sulin phosphorylates PKB in MAM, could suggest thatregulation of calcium homeostasis is eventually anotherprocess to take into account in the relationship betweenMAM and insulin resistance. Therefore, further studiesare required to prove this hypothesis and to evaluatewhether a disruption of Ca2+ transfer from ER to mito-chondria in conditions of altered MAM integrity couldcontribute to hepatic insulin resistance.

ER stress (6) is a classical hallmark of hepatic insulinresistance, and some studies suggested an interplay be-tween ER stress and mitochondrial dysfunction in thereduction of insulin responsiveness (37). Our data indi-cate that MAM dynamics and regulation could contributeto this interplay. This is in line with recent data showingthat mitochondria-ER coupling was enhanced during earlyphase of ER stress (28,38) but also that alteration of theseinteractions results in a disruption of Ca2+ transfer (3)and subsequent ER stress (31). We also observed thatmodulation of MAM in vitro in HuH7 cells could impactER stress (Supplementary Fig. 11) and that disruption ofMAM in CypD-KO mice was associated with hepatic ERstress (data not shown). We did not investigate in thecurrent study the consequence of MAM modulation onmitochondria function in liver cells. Previous results in-dicated that mitochondria function could either be altered(28,38) or enhanced (11) by MAM disruption, indicatingcomplex relationships. Similarly, the implication of mito-chondrial dysfunction in liver insulin resistance is stilla matter of debate (5,39). Therefore, future studiesshould be designed to define the causal links betweenMAM integrity, ER stress, and mitochondria dysfunctionin relation with insulin resistance. Nevertheless, our studyadds new important information regarding the role of theinteractions between mitochondria and ER in liver, sug-gesting that defective ER-mitochondria contacts couldcontribute to hepatic insulin resistance.

In summary, we validated that in situ PLA offersa reliable method to visualize ER-mitochondria interactionsin a native environment, and we demonstrated that MAMshould be regarded as a novel hub for insulin-stimulatedPKB phosphorylation, its integrity being strongly associ-ated with correct insulin signaling and action in hepato-cytes. Furthermore, defective ER-mitochondria coupling is

closely associated with impaired hepatic insulin sensitivityboth in vivo and in vitro, and increasing organelle couplingimproves insulin signaling in primary hepatocytes. Ourdata clearly indicate, therefore, that the MAM interfaceand interorganelle contacts are novel actors in the mech-anisms of action of insulin in liver and that these inter-organelle domains should be taken into consideration inthe dysregulations associated with insulin resistance.

Acknowledgments. The authors thank Michel Rivoire and Maud Micheletof the INSERM U1052 (Lyon, France) for the preparation of human primaryhepatocytes, Elisabeth Errazuriz for her technical help at the CeCILE ImagingCenter (Lyon, France), Gyorgy Hajnoczky of the Thomas Jefferson University(Philadelphia, PA) for the generous gift of ER-mitochondria linkers, and RomanEliseev of the University of Rochester Medical Center (Rochester, NY) for thegenerous gift of the expression vector encoding rat CypD.Funding. This work was supported by INSERM and the National ResearchAgency (ANR-11-BSV1-033-02 to M.O. and ANR-09-JCJC-0116 to J.R.). E.T. andP.T. were supported by a research fellowship from the French Ministry of HigherEducation and Research.Duality of Interest. This work was supported by Servier Laboratories. Noother potential conflicts of interest relevant to this article were reported.Author Contributions. E.T. and J.R. designed the experiments, re-searched data, contributed to discussion, and wrote the manuscript. P.T., G.V.,N.B., A.B., M.-A.C., J.J.-C., and B.B. researched data. F.Z., M.O., and H.V.contributed to discussion and reviewed/edited the manuscript. J.R. is the guar-antor of this work and, as such, had full access to all the data in the study andtakes responsibility for the integrity of the data and the accuracy of the dataanalysis.Prior Presentation. Some of the data in the article were presented asa poster at the 49th Annual Meeting of the European Association for the Study ofDiabetes, Barcelona, Spain, 23–27 September 2013 (abstract published in Dia-betologia 2013;56:S283) and as an oral presentation at the 74th ScientificSessions of the American Diabetes Association, San Francisco, CA, 13–17June 2014.

References1. Rowland AA, Voeltz GK. Endoplasmic reticulum-mitochondria contacts:function of the junction. Nat Rev Mol Cell Biol 2012;13:607–6252. Szabadkai G, Bianchi K, Várnai P, et al. Chaperone-mediated coupling ofendoplasmic reticulum and mitochondrial Ca2+ channels. J Cell Biol 2006;175:901–9113. de Brito OM, Scorrano L. Mitofusin 2 tethers endoplasmic reticulum tomitochondria. Nature 2008;456:605–6104. Patergnani S, Suski JM, Agnoletto C, et al. Calcium signaling around Mi-tochondria Associated Membranes (MAMs). Cell Commun Signal 2011;9:195. Vial G, Dubouchaud H, Couturier K, et al. Effects of a high-fat diet on energymetabolism and ROS production in rat liver. J Hepatol 2011;54:348–3566. Ozcan U, Cao Q, Yilmaz E, et al. Endoplasmic reticulum stress links obesity,insulin action, and type 2 diabetes. Science 2004;306:457–4617. Fu S, Yang L, Li P, et al. Aberrant lipid metabolism disrupts calcium ho-meostasis causing liver endoplasmic reticulum stress in obesity. Nature 2011;473:528–5318. Sebastián D, Hernández-Alvarez MI, Segalés J, et al. Mitofusin 2 (Mfn2)links mitochondrial and endoplasmic reticulum function with insulin signaling andis essential for normal glucose homeostasis. Proc Natl Acad Sci U S A 2012;109:5523–55289. Gan KX, Wang C, Chen JH, Zhu CJ, Song GY. Mitofusin-2 ameliorates high-fat diet-induced insulin resistance in liver of rats. World J Gastroenterol 2013;19:1572–1581

diabetes.diabetesjournals.org Tubbs and Associates 3293

10. Giorgi C, Ito K, Lin HK, et al. PML regulates apoptosis at endoplasmic re-ticulum by modulating calcium release. Science 2010;330:1247–125111. Betz C, Stracka D, Prescianotto-Baschong C, Frieden M, Demaurex N, HallMN. Feature Article: mTOR complex 2-Akt signaling at mitochondria-associatedendoplasmic reticulum membranes (MAM) regulates mitochondrial physiology.Proc Natl Acad Sci U S A 2013;110:12526–1253412. Bononi A, Bonora M, Marchi S, et al. Identification of PTEN at the ERand MAMs and its regulation of Ca(2+) signaling and apoptosis in a proteinphosphatase-dependent manner. Cell Death Differ 2013;20:1631–164313. Mannella CA. The relevance of mitochondrial membrane topology to mito-chondrial function. Biochim Biophys Acta 2006;1762:140–14714. Cosson P, Schurdi-Levraud V, Le QH, et al. The RTM resistance to poty-viruses in Arabidopsis thaliana: natural variation of the RTM genes and evidencefor the implication of additional genes. PLoS ONE 2012;7:e3916915. Mannella CA, Buttle K, Rath BK, Marko M. Electron microscopic tomographyof rat-liver mitochondria and their interaction with the endoplasmic reticulum.Biofactors 1998;8:225–22816. Csordás G, Renken C, Várnai P, et al. Structural and functional features andsignificance of the physical linkage between ER and mitochondria. J Cell Biol2006;174:915–92117. Rizzuto R, Pinton P, Carrington W, et al. Close contacts with the endo-plasmic reticulum as determinants of mitochondrial Ca2+ responses. Science1998;280:1763–176618. Csordás G, Várnai P, Golenár T, et al. Imaging interorganelle contacts andlocal calcium dynamics at the ER-mitochondrial interface. Mol Cell 2010;39:121–13219. Schwarze E, Tolleshaug H, Seglen O. Uptake and degradation of asialo-orosomucoid in hepatocytes from carcinogen-treated rats. Carcinogenesis 1985;6:777–78220. Lecluyse EL, Alexandre E. Isolation and culture of primary hepatocytes fromresected human liver tissue. Methods Mol Biol 2010;640:57–8221. Eliseev RA, Malecki J, Lester T, Zhang Y, Humphrey J, Gunter TE. Cyclo-philin D interacts with Bcl2 and exerts an anti-apoptotic effect. J Biol Chem 2009;284:9692–969922. Lecomte V, Meugnier E, Euthine V, et al. A new role for sterol regulatoryelement binding protein 1 transcription factors in the regulation of muscle massand muscle cell differentiation. Mol Cell Biol 2010;30:1182–119823. Bonnard C, Durand A, Peyrol S, et al. Mitochondrial dysfunction results fromoxidative stress in the skeletal muscle of diet-induced insulin-resistant mice.J Clin Invest 2008;118:789–80024. Schinzel AC, Takeuchi O, Huang Z, et al. Cyclophilin D is a component ofmitochondrial permeability transition and mediates neuronal cell death after focalcerebral ischemia. Proc Natl Acad Sci USA 2005;102:12005–1201025. Rieusset J, Chauvin MA, Durand A, et al. Reduction of endoplasmic re-ticulum stress using chemical chaperones or Grp78 overexpression does not

protect muscle cells from palmitate-induced insulin resistance. Biochem BiophysRes Commun 2012;417:439–44526. Wieckowski MR, Giorgi C, Lebiedzinska M, Duszynski J, Pinton P. Isolationof mitochondria-associated membranes and mitochondria from animal tissuesand cells. Nat Protoc 2009;4:1582–159027. Paillard M, Tubbs E, Thiebaut PA, et al. Depressing mitochondria-reticuluminteractions protects cardiomyocytes from lethal hypoxia-reoxygenation injury.Circulation 2013;128:1555–156528. Bravo R, Vicencio JM, Parra V, et al. Increased ER-mitochondrial couplingpromotes mitochondrial respiration and bioenergetics during early phases of ERstress. J Cell Sci 2011;124:2143–215229. Area-Gomez E, Del Carmen Lara Castillo M, Tambini MD, et al. Upregulatedfunction of mitochondria-associated ER membranes in Alzheimer disease. EMBOJ 2012;31:4106–412330. Söderberg O, Gullberg M, Jarvius M, et al. Direct observation of individualendogenous protein complexes in situ by proximity ligation. Nat Methods 2006;3:995–100031. Simmen T, Aslan JE, Blagoveshchenskaya AD, et al. PACS-2 controls en-doplasmic reticulum-mitochondria communication and Bid-mediated apoptosis.EMBO J 2005;24:717–72932. Verfaillie T, Rubio N, Garg AD, et al. PERK is required at the ER-mitochondrial contact sites to convey apoptosis after ROS-based ER stress. CellDeath Differ 2012;19:1880–189133. Zampese E, Fasolato C, Kipanyula MJ, Bortolozzi M, Pozzan T, Pizzo P.Presenilin 2 modulates endoplasmic reticulum (ER)-mitochondria inter-actions and Ca2+ cross-talk. Proc Natl Acad Sci U S A 2011;108:2777–278234. Ye R, Ni M, Wang M, et al. Inositol 1,4,5-trisphosphate receptor 1 mutationperturbs glucose homeostasis and enhances susceptibility to diet-induced di-abetes. J Endocrinol 2011;210:209–21735. Marchi S, Marinello M, Bononi A, et al. Selective modulation of subtype IIIIP(3)R by Akt regulates ER Ca(2)(1) release and apoptosis. Cell Death Dis 2012;3:e30436. Szado T, Vanderheyden V, Parys JB, et al. Phosphorylation of inositol1,4,5-trisphosphate receptors by protein kinase B/Akt inhibits Ca2+ release andapoptosis. Proc Natl Acad Sci U S A 2008;105:2427–243237. Lim JH, Lee HJ, Ho Jung M, Song J. Coupling mitochondrial dysfunction toendoplasmic reticulum stress response: a molecular mechanism leading to he-patic insulin resistance. Cell Signal 2009;21:169–17738. Cárdenas C, Miller RA, Smith I, et al. Essential regulation of cell bio-energetics by constitutive InsP3 receptor Ca2+ transfer to mitochondria. Cell2010;142:270–28339. Pospisilik JA, Knauf C, Joza N, et al. Targeted deletion of AIF decreasesmitochondrial oxidative phosphorylation and protects from obesity and diabetes.Cell 2007;131:476–491

3294 Disruption of MAM in Hepatic Insulin Resistance Diabetes Volume 63, October 2014