Phosphatidylinositol-4-Phosphate-5-Kinase ? Deficiency Alters Dynamics of Glucose-Stimulated Insulin Release to Improve Glucohomeostasis and Decrease Obesity in Mice.

2 MB·English

by  Huang, Ping

ORIGINALARTICLE
Phosphatidylinositol-4-Phosphate-5-Kinase a Deficiency
Alters Dynamics of Glucose-Stimulated Insulin Release to
Improve Glucohomeostasis and Decrease Obesity in Mice
Ping Huang,1,2 Oladapo Yeku,1,3,4 Haihong Zong,2 Phyllis Tsang,1,2 Wenjuan Su,1,3 Xiao Yu,1,2
Shuzhi Teng,6 Mary Osisami,1,5 Yasunori Kanaho,7 Jeffrey E. Pessin,2 and Michael A. Frohman1,2,3,5
OBJECTIVE—Phosphatidylinositol-4-phosphate-5-kinase(PI4P5K)
has been proposed to facilitate regulated exocytosis and specifi- F
cally insulin secretion by generating phosphatidylinositol-4,5- ailure of pancreatic b-cells to release adequate
bisphosphate (PIP ). We sought to examine the role of the a amounts of insulin contributes to the onset of
2
isoform of PI4P5K in glucohomeostasis and insulin secretion. type 2 diabetes and obesity (1). Elevated serum
RESEARCH DESIGN AND METHODS—The response of glucose transported into pancreatic b-cells is
PI4P5Ka2/2 micetoglucosechallengeandatype2-likediabetes- metabolized to increase cytosolic ATP levels, which then
inducing high-fat diet was examined in vivo. Glucose-stimulated promote closure of ATP-sensitive K+ (KATP) channels,
responses and PI4P5Ka2/2 pancreatic islets and b-cells were causing membrane depolarization. Membrane depolar-
characterizedin culture. ization triggers opening of L-type Ca2+ channels, influx of
RESULTS—We show that PI4P5Ka2/2 mice exhibit increased Ca2+, and exocytosis. The first phase of exocytosis entails
first-phaseinsulinreleaseandimprovedglucoseclearance,andre- fusionofprimedinsulingranulespredockedattheplasma
sist high-fat diet-induced development of type 2-like diabetes and membrane (2). A second phase involving mobilization of
obesity. PI4P5Ka2/2 pancreatic islets cultured in vitro exhibited distal insulin vesicles occurs after approximately 10 min
decreased numbers of insulin granules docked at the plasma (3), preceded by actin cytoskeletal reorganization (4–6)
membrane and released less insulin under quiescent conditions, and generation of lipid second messengers (7,8). The ma-
butthensecretedsimilaramountsofinsulinonglucosestimula- jority of type 2 diabetes is only weakly associated with
tmioenm.bSrtaimneuloatfiothne-dPeIp4ePn5dKean2tP/2IPp2adnecpreleattiiocnbo-cceclulsr,readccoonmthpeanpileadsmbay specificgeneticdefectsandischaracterizedbyinadequate
release of insulin, in addition to insulin resistance exhibi-
anear-totallossofcorticalF-actin,whichwasalreadydecreasedin
thePI4P5Ka2/2b-cellsunderrestingconditions. tedbyfatandmuscletargetcells.Prolongedstimulationof
b-cells by elevated levels of glucose, such as is encoun-
CONCLUSIONS—OurfindingssuggestthatPI4P5Kaplaysacom- tered in the typical western-style high-fat diet, eventually
plex role in restricting insulin release from pancreatic b-cells suffices to trigger changes in insulin secretion in many
through helping to maintain plasma membrane PIP2 levels and individuals with otherwise seemingly normal physiology
integrity of the actin cytoskeleton under both basal and stimula-
and genetics.
tory conditions. The increased first-phase glucose-stimulated re-
Lipid kinases and their phosphoinositide products play
lease of insulin observed on the normal diet may underlie the
important roles in secretory vesicle trafficking (9). Type I
partialprotectionagainstthe elevatedserumglucoseandobesity
seenintype2diabetes-likemodelsystems.Diabetes60:454–463, phosphatidylinositol-4-phosphate-5-kinases (PI4P5Ks) a,
2011 b, and g generate the signaling lipid phosphatidylinositol-
4,5-bisphosphate (PIP ). Elegant studies in neurons and
2
neuroendocrine cells on the role of PI4P5Kg (10,11) and
PIP (12–14) have revealed that PIP generation at the
Fromthe1CenterforDevelopmentalGenetics,StonyBrookUniversity,Stony 2 2
Brook,NewYork;the2DepartmentofPharmacology,StonyBrookUniver- plasma membrane is critical during regulated exocytosis.
sity,StonyBrook,NewYork;the3PrograminMolecularandCellularPhar- PIP directly facilitates some types of Ca++ signaling
2
macology, Stony Brook University, Stony Brook, New York; the 4Medical (15,16), recruits proteins that facilitate the fusion process
Scientist Training Program, Stony Brook University, Stony Brook, New
York; the 5Program in Genetics, Stony Brook University, Stony Brook, (13,14), and is required for docked secretory vesicles to
NewYork;the6DepartmentofMedicalOncology,Dana-FarberCancerIn- undergo priming to become part of the ready-releasable
stitute,Boston,Massachusetts;andthe7DepartmentofPhysiologicalChem-
pool and then fuse into the plasma membrane (10,11).
istry,GraduateSchoolofComprehensiveHumanSciencesandInstituteof
BasicMedicalSciences,UniversityofTsukuba,Tsukuba,Japan. Although less well understood mechanistically, decreased
Correspondingauthor:MichaelA.Frohman,[email protected]. levels of PIP have also been shown to inhibit insulin se-
Received28April2010andaccepted10November2010. 2
DOI:10.2337/db10-0614 cretion in pancreatic b-cell model systems (7,17–19).
This article contains Supplementary Data online at http://diabetes. However,PIP alsocarriesoutfunctionsthatpotentially
2
diabetesjournals.org/lookup/suppl/doi:10.2337/db10-0614/-/DC1. oppose regulated exocytosis (20). First, PIP supports the
P.H.andO.Y.contributedequallytothiswork. 2
P.H.iscurrentlyaffiliatedwiththeDepartmentofPediatrics,HarvardMedical openstateoftheKATPchannel(21,22);thus,becauseitis
School,Boston,Massachusetts. ATP-driven closure of the KATP channel that triggers se-
H.Z.andJ.E.P.arecurrentlyaffiliatedwiththeDepartmentofMolecularPhar- cretion, PIP deficiency might be anticipated to decrease
X.Ym.aicsocluorgrye,nAtllybearftfiEliaintesdteiwnitChotlhleegeDeopfaMrtemdeicnitneo,fBBriooncxhe,mNeiswtryY,oIrnkd.ianaUni- K+ efflux, tr2iggering membrane depolarization and thus
versity,Bloomington,Indiana. increasing Ca++ currents, resulting in increased secretion
(cid:1)2011bytheAmericanDiabetesAssociation.Readersmayusethisarticleas (23–25).Insupportofthismodel,expressionofadominant-
longastheworkisproperlycited,theuseiseducationalandnotforprofit,
negative isoform of PI4P5K to lower levels of PIP changes
and the work is not altered. See http://creativecommons.org/licenses/by 2
-nc-nd/3.0/fordetails. theresponsivenessofmutantKATPchannelswithdecreased
454 DIABETES,VOL.60,FEBRUARY2011 diabetes.diabetesjournals.org
P.HUANGANDASSOCIATES
ATP sensitivity. However, expression of the dominant- RESEARCHDESIGNANDMETHODS
negativePI4P5Kdoesnotalterfunctionofwild-type(WT) Animals.PI4P5Ka2/2malemicebackcrossedtoC57BL/6weregenotypedper
KATP channels, suggesting that under normal physiologic Sasaki and colleagues (36). Animal experiments were performed in accor-
conditions, the level of PIP on the plasma membrane is dance with the rules and regulations of the Stony Brook University Animal
2
Care Committee. For high-fat diet, 12-week-old mice were switched from
not high enough to strongly affect KATP channel activity
normalchowto45%highfat(ResearchDiets,Inc.,NewBrunswick,NJ)for8
(26). PIP has also been suggested to restrain fusion of
2 weeks(38).
docked vesicles by inhibiting SNARE complex function Intraperitonealglucoseandinsulintolerancetests.Micefastedfor14h
(14,27), with the restraint being alleviated through se- wereinjectedintraperitoneally with1 g/kg25% D-glucosein saline. Blood
questrationofthePIP bySyntaxin-1(14)orCa2+-triggered (5–10 mL) was collected from the tail. Blood glucose was determined by
2
PIP destruction (28). aglucometerandinsulinbyELISAkit(CrystalChemInc.,DownersGrove,IL).
2
For insulin tolerance tests, mice were fasted for 6 h before being injected
Taken together, the action of PIP is complex, and both
2 intraperitoneally with insulin (0.85 units/kg of Novolin R; Novo Nordisk,
itssynthesisanditsturnoverarerequiredatdifferentsteps
Bagsværd,Denmark).Bloodglucoselevelsweremonitoredasaboveimme-
in the fusion process. diatelybeforeinjectionandat15,30,45,60,and90minafterinjection.
Another function undertaken by PIP is to promote as- Isolationofpancreaticislets.Pancreatictissuefrom8-week-oldmalemice
2
sembly of actin filaments (F-actin) (29,30). F-actin at the was dissolved in Hank’s balanced salt solution (HBSS) containing Liberase
periphery of the cell (cortical F-actin) has also been pro- (Roche 11815032001) and DNase I (Roche 10104159001) at 37°C for 40–50
min;digestionwasstoppedwithHBSS/10%FBSandtwowasheswithHBSS/
posed to have both positive and negative regulatory
2%FBS.ThetissuewaspassedthroughacellstrainerandappliedtoaPoly-
functions in exocytosis. Cortical F-actin has been pro-
sucrose(Cellgrow61196RO)gradientof25,23,20,and11%layerstoisolate
posed to act as a barrier to block access of undocked se- theisletsthatwereaspiratedfromtheinterfacesbetweenthe11and20%and
cretory vesicles to the plasma membrane; consistent with the20and23%gradients,washed,andpickedunderadissectingmicroscope.
this model, cortical F-actin is disassembled during regu- IsletswereusedimmediatelyorculturedinRoswellParkMemorialInstitute
latory exocytosis events (31), and pharmacologic agents medium/10%FBS,Pen/Strep,andglutamate.
Insulin secretion via static incubation. Pancreatic islets (n = 10) were
that disassemble F-actin enhance movement of insulin
incubatedin250mLofKrebs-RingerbicarbonateHEPESbufferfor30minat
granules to the plasma membrane and insulin release,
37°C,andaliquotsweretakenasbasalsamples.Theglucosewasincreasedto
whereas agents that stabilize the actin cytoskeleton de- 20mmol/L,andthecultureswereincubatedforanother30min.Sampleswere
creaseinsulinrelease(4,18,32–34).F-actinmayalsoaffect placedoniceorstoredat220°C.Todeterminetotalinsulincontent,theislets
SNARE complex function by binding to and inhibiting weredissolvedin70%ethanoland0.18mol/LHClovernight,sonicated,and
Syntaxin-4 function in a glucose stimulation-relieved centrifuged before samples were taken for analysis using an ultrasensitive
murine insulin ELISA (Mercodia: 10115010) and BioRad plate reader. Total
manner(32,33).Tofurthercomplicatematters,F-actincan
protein concentrations were determined with a BioRad Bradford protein
also undertake a positive role in regulated exocytosis by
assay.
mediatingtranslocationofmoreinternalsecretoryvesicles Immunostaining, electron microscopy, and morphologic analysis of
to the periphery, particularly in poorly granulated or re- islets. Pancreatic islets were fixed in 4% paraformaldehyde for 20 min and
cently degranulated cells (4,35). Thus, dynamic regulation permeabilizedin0.1%TritonX-100for20min.Immunofluorescencestaining
of the actin cytoskeleton is also important in the pro- forinsulinandglucagonwasperformedwithrabbitanti-insulin(SantaCruz
BiotechnologyInc.,SantaCruz,CA)andmurineanti-glucagon(Sigma,St.Louis,
gression of regulated exocytosis and can play positive or
MO),followedbyfluoresceinisothiocyanate-conjugatedanti-rabbitandAlexa
negative roles depending on the setting.
546-conjugated anti-mouse antibodies (Molecular Probes, Eugene, OR). For
Changes in PIP2 and cortical F-actin can have positive actincytoskeletonstaining,fixedandpermeabilizedisletswerestainedwith
and negative effects on the secretory process, making it Rhodamine-phalloidin,andthefluorescenceintensitywasacquiredfromthe
difficult a priori to predict the outcomes of their physio- toptothebottomoftheisletsusingcontinuousxyz-scanmodeontheLeica
logic and experimental manipulation. Critically, individual confocalmicroscope.Allisletswerescannedatthesamegainintensity,andthe
three-dimensionalserieswasassembledusingconfocalimagingsoftware.
PI4P5K isoforms may generate subpools of PIP that reg-
2 For electron microscopy analysis, islets were incubated in Krebs-Ringer
ulatedistinctcomponentsofthefusionprocess.Deletionof
bicarbonateHEPESbuffercontaining2.8mmol/Lglucoseat37°Cfor30min
PI4P5Kg markedly decreases levels of PIP2 at the plasma (nonstimulatedislets),andsomewerefurtherincubatedin20mmol/Lglucose
membrane in resting neurons and neuroendocrine cells, for30min(stimulatedislets).Theisletswerefixedin2%paraformaldehyde/2%
and inhibits secretion at the stage of vesicle priming and electronmicroscopygradeglutaraldehydein0.1mol/Lphosphatebuffer,pH
fusion without notable effect on the actin cytoskeleton 7.4for15min,postfixedin2%OsO4inPBSfor1h,dehydrated,infiltrated,and
embeddedinplasticresin.Ultra-thinsectionswerestainedwithuranylacetate
(10,11). In contrast, deletion of PI4P5Ka increases mast
and lead citrate before examination using a FEI BioTwinG2 transmission
cell degranulation triggered by cross-linking of the IgE re-
electron microscope (FEI Company, Hillsboro, OR). b-cell size and insulin
ceptor, accompanied by minor decreases in cellular PIP2, granuleswereanalyzedusingImageJsoftware(NationalInstitutesofHealth;
but significantly increased levels of Ca2+ signaling and de- http://rsb.info.nih.gov/ij/).
creased levels of total F-actin (36). PI4P5Ka knockdown Statistics.Resultsareexpressedasmean6SDunlessotherwisestatedin
using RNAi has also been reported to alter Ca2+ signaling, Figs. 1–7. Statistical differences between datasets were assessed using the
unpairedStudentttest.
disrupt F-actin, and affect insulin release in a pancreatic
b-cell line (37).
This report examines insulin secretion in PI4P5Ka2/2
mice and finds that first-phase insulin release is aug- RESULTS
mented, and on a high-fat diet, fasting and stimulated se- Enhanced insulin secretion and improved glucose
rum insulin levels are even more elevated, conferring tolerance in PI4P5Ka2/2 mice. PI4P5Ka2/2 (KO) mice
faster glucose clearance and resistance to the develop- fed on normal mouse chow for 8–12 weeks were visually
ment of obesity. In this setting, K+ and Ca++ signaling is indistinguishablefromage-andsex-matchedWTlittermate
seemingly normal; in contrast, the determinative factor mice but weighed significantly less (26 vs. 28 g, P = 0.03,
seemstobeadramaticstimulation-dependentlossofPIP n = 16, Fig. 1A). Glucose tolerance tests revealed that the
2
attheplasmamembranethatleadstoreorganizationofthe fastingserumglucoselevelsweresignificantlylowerinthe
actin cytoskeleton resulting in a near-total loss of cortical KOmicethanintheWTmice(86621mg/dLvs.126634
F-actinanddecreasednumbersofinsulingranulesdocked mg/dL,respectively)andthattheKOmiceclearedglucose
at the plasma membrane. faster(Fig.1B).Seruminsulinlevelswereincreasedinthe
diabetes.diabetesjournals.org DIABETES,VOL.60,FEBRUARY2011 455
PI4P5KaDEFICIENCYIMPROVESGLUCOHOMEOSTASIS
FIG.1.PI4P5KaKOmiceexhibitimprovedglucohomeostasisonnormalmurinechowandresistdevelopmentoftype2-likediabetesresponseson
ahigh-fatdiet.A:BodyweightsofWTandKOmice(n=16)raisedonnormalmurinechow.Age-matchedmalemicewereusedforallexperiments.B:
SerumglucoselevelsforWTandKOmice(n=9)infastingmice(0mintimepoint)andafter1g/kgbodymassintraperitonealinjectionof25%
glucose(intraperitonealglucosetolerancetest)(*P<0.04).C:Seruminsulinlevelsmeasuredinparallel(*P<0.05;#P<0.1).D:Insulintolerance
testperformedonWTandKOmice(n=5;*P<0.05).E:BodyweightsofWTandKOmice(n=10)after8weeksonahigh-fatdiet.F:Intraperitoneal
glucosetolerancetestresultsonhigh-fatdiet(*P<0.03;#P<0.06).G:Seruminsulinlevelsmeasuredinparallel(*P<0.05,n=7).Statistical
significancedeterminedbyStudentttest.Representativeofthreeexperiments.ITT,insulintolerancetest;GTT,glucosetolerancetest.
456 DIABETES,VOL.60,FEBRUARY2011 diabetes.diabetesjournals.org
P.HUANGANDASSOCIATES
FIG.2.First-phaseGSIRandL-argininepotentiationareincreasedinabsenceofPI4P5Ka.A:WTandKOmicereceivedabolusof20%glucoseto
increaseserumglucoselevels>250mg/dLandwerethenmaintainedatthatlevelbyintravenousinfusionadjustedforthechangeinserumglucose
levelsovertime.B:Insulinsecretionduringthehyperglycemicclamp.Bloodsamplesweretakenattheintervalsshownandassayedforserum
insulinlevels.Atmin80,abolusof20%L-arginineandthenanL-arginineinfusionwereadded.Eachexperimentgroupwascomposedoffourage-
matchedmalemice.*SignificantPvaluesasindicated.StatisticalsignificancedeterminedbyStudentttest.
KO mice at 15–30’ post glucose challenge (Fig. 1C), and chow(Fig.1Evs.Fig.1A).However,thePI4P5KaKOmice
total insulin secretion, calculated as areaunder the curve, gained30%lessweightthantheWTmice(Fig.1E).Fasting
was significantly increased (P , 0.04) from 67.8 6 5.5 ng/ serum glucose levels were elevated in both the WT and
mL/120minto76.266.3ng/mL/120min,suggestingabasis PI4P5Ka KO mice, although the levels remained higher in
for the improved glucose clearance. An insulin tolerance the WT mice (Fig. 1F). Moreover, whereas WT mice
test did not reveal increased peripheral glucose utilization exhibitedprolongedandhigherelevationofserumglucose
forPI4P5KaKOmiceincomparisonwithWTmice(Fig.1D), levels,thePI4P5KaKOmiceexhibitedamorerapidreturn
ruling out the possibility of increased insulin sensitivity by to euglycemia (Fig. 1F). Fasting and stimulated serum in-
target tissues as a mechanism. sulin levels were elevated in the WT mice (Fig. 1G).
PI4P5Ka KO mice resist developing type 2 diabetes- However,insulinlevelsinthePI4P5KaKOmicewereeven
like symptoms on a high-fat diet. C57BL/6 mice placed further elevated under fasting (3.01 6 0.40 vs. 1.59 6 0.44
on high-fat diets develop hyperglycemia and obesity (39). ng/mL for the WT mice, P , 0.05) and glucose-stimulated
Age- and sex- matched WT and KO mice generated and conditions.
maintained on the C57BL/6 background were placed on By using the hyperglycemic clamp technique, glucose
high-fat chow for 8 weeks. The mice exhibited a sub- levels were clamped at $250 mg/dL (Fig. 2A) and insulin
stantialgaininweightincomparisonwithmiceonnormal levels were measured at frequent intervals. Elevated
diabetes.diabetesjournals.org DIABETES,VOL.60,FEBRUARY2011 457
PI4P5KaDEFICIENCYIMPROVESGLUCOHOMEOSTASIS
FIG. 4. PI4P5Ka KO islets secrete insulin with altered kinetics on
glucose stimulation. WT and KO islets (10 islets per experimental
sample,experimentperformedsixtimes)wereisolatedfromage-and
sex-matchedmice,preincubatedinnonstimulatoryculturemedia(2.8
mmol/Lglucose),andchallengedwith20mmol/Lglucose.Sixmicewere
usedforeachgenotype.Incubationtimeswere30mininsuccessionfor
each condition. Stimulated insulin release is presented after sub-
tractionofthebasalrelease.Dataarepresentedasmean6SEM.
Taken together with the results shown in Fig. 1, these
findingssuggestthatdeletionofthePI4P5Kageneresults
in increased first-phase glucose-dependent insulin secre-
tion that counters the development of the type 2 diabetic-
like phenotype.
PlasmamembranePIP isnotmaintainedinKOislets
2
on stimulation. The availability of PIP at the plasma
2
membrane plays critical roles in many cell signaling pro-
cesses (40). Total cellular PIP levels in cells from
2
PI4P5Ka KO mice were examined previously (36) and
found to be decreased only modestly (90% of WT levels).
However, PIP exhibits rapid turnover and resynthesis at
2
the plasma membrane on glucose-triggered b-cell stimu-
lationasaconsequenceofCa2+-mediatedphospholipaseC
activation (28). The signaling pathway leading to PIP
2
turnover can also be initiated by elevating levels of ex-
tracellular K+ to depolarize the b-cell, thereby triggering
the same downstream effects(18). Weused this approach
to examine whether there were differences in basal or
stimulatedlevelsofaccessiblePIP attheplasmamembrane
2
of isolated pancreatic b-cells as reported by a fluorescent
FIG.3.PlasmamembranePIP2isnotmaintainedinKOb-cellsonde- PIP sensor based on fusion of EGFP to the PIP -binding
polarization.A:b-cellsisolatedfromWTandKOisletsweretransfected PH2domain of PLCd1 (PLCd1-PH-EGFP). Prim2ary pan-
with a plasmid expressing the PIP sensor PLCd1-PH-EGFP, cultured
for2days,depolarizedwithextrac2ellularK+for5min,fixed,andim- creatic b-cells were transfected in vitro with a reporter
agedusingconfocalmicroscopy.Pancreaticb-cellswereidentifiedus- plasmid expressing PLCd1-PH-EGFP, with a transfection
ing an anti-insulin antibody in combination with a red fluorescent frequency of 80–85% (Supplementary Fig. 1). Under basal
secondary antibody. B: KO b-cells expressing a mutant PIP sensor
(wPeLrCedp1ro-PcHes-Rse4d0Las-EinGFAP.I)mtahgaetsdwoeersetnaoktenexthhriboiutghPItPh2embiinddpionign2tascotfivtihtey bcoranndeitiinonbso,tPhItPh2ewWaTsarnedadKilOy cimelalsge(Fdiga.t3tAh,eropwlassm1aanmde3m)-,
cellsandarerepresentativeofthetypicaltransfected,insulin-positive although it should be noted that this observation demon-
cellsobservedintheexperiments(n>100).Theexperimentwasre-
peatedatleastthreetimesusingtwoanimalseachtime(n=sixmice), strates only that the minimal threshold for sensor re-
and similar outcomes were obtained. Greater than 90% of the cells cruitment is in place, not necessarily that the levels of
examinedrespondedasshownintheimagesineachcondition.(Ahigh- accessible plasma membrane PIP are quantitatively iden-
qualitycolorrepresentationofthisfigureisavailableintheonlineissue.) tical. However, differences were2observed on K+-elicited
depolarization. Whereas no change was observed for WT
b-cells, indicating that the rates of resynthesis of PIP
2
insulin secretion (P , 0.01) in KO mice was observed at compensated for the depolarization-triggered turnover,
5 min after the start of the glucose infusion (Fig. 2B), but relocalization of the sensor from the plasma membrane to
the levels of release were subsequently similar to those of the cytoplasm was observed for the PI4P5Ka KO cells. A
WT mice. Addition of L-arginine, a potentiator of glucose- mutant sensor, PLCd1-R40L-PH-EGFP, which lacks the ca-
stimulated insulin release (GSIR), resulted in increased pabilitytobindPIP ,distributeduniformlyinthecytoplasm
2
insulinreleaseinbothWTandKOmice,buttheeffectwas underbothbasalandstimulatedconditions(Fig.3B).There
substantially greater for the KO mice. were no obvious changes in localization of the insulin
458 DIABETES,VOL.60,FEBRUARY2011 diabetes.diabetesjournals.org
P.HUANGANDASSOCIATES
FIG.5.Morphologyofislets,b-cellsize,andinsulingranuledensityarenotnoticeablydifferentinPI4P5KaKOmice,butdistanceofgranulesfrom
cortexis.A:ConfocalmicroscopyimageofWTandKOisletsstainedwithanti-insulinandantiglucagonprimaryantibodiestovisualizeb-cellsand
a-cells,respectively.B:QuantitationofWTandKOb-cellsizeusingtransmissionelectronmicroscopy.C:Quantitationofthenumberofinsulin
granulescontainedinWTandKOisletsnormalizedforthesizeoftheeachislet.D:TransmissionelectronmicroscopyimagesofWTandKOislets.
KO b-cells show less insulin vesicles docked at the plasma membrane in comparison with WT b-cells (magnified lower panels; bar, 500 nm).
Asterisksindicateboundariesbetweenadjacentb-cells.E:DistanceofsubgranularpoolsofinsulintotheplasmamembraneinWTandKOcells
underbasalconditions.Insulingranulesweregroupedintothreecategorieswithrespecttodistancefromplasmamembrane:<50,50–300,and
>300nm.Granuleslocatedwith50nmfromplasmamembranewereconsideredtobedockedgranules.Granulesweretabulatedfrom10sections
foreachgenotype.Dataarepresentedasmean6SEM.(Ahigh-qualitycolorrepresentationofthisfigureisavailableintheonlineissue.)
granules in the isolated PI4P5Ka KO cells under basal or Fig. 2). However, quantitative analysis of the subcellular
stimulated conditions in comparison with the WT cells. distribution of insulin granules by electron microscopy
These results indicate that plasma membrane PIP de- (41,42) revealedthat decreasednumbersofgranuleswere
2
creases after depolarization (stimulation) of the PI4P5Ka docked at the plasma membrane in the PI4P5Ka KO cells
KO b-cells, but because of a reduced capacity for resyn- (i.e., were localized within 50 nm of the plasma mem-
thesis,the recoveryof PIP levelsatthe plasmamembrane brane)(Fig.5DandE),suggestingabasisforthedecrease
2
is blunted. in basal insulin secretion in KO mice in the ex vivo ex-
PI4P5Ka KO islets exhibit alterations in ex vivo periments.Onglucosestimulation,thenumberofgranules
insulin secretion. Surprisingly, release of insulin by localized at 50–300 nm from the plasma membrane de-
PI4P5Ka KO islets cultured ex vivo was significantly less creased in both WT and KO b-cells, whereas the number
under basal conditions, but similar to WT on glucose of granules localized within 50 nm of the plasma mem-
stimulation(Fig.4).Wemeasuredthetotalinsulincontent brane increased; no significant differences were seen at
per islet (112 ng/islet, WT, vs. 94 ng/islet, KO; NS), exam- that point between WT and KO b-cells for insulin granule
ined islet appearance using anti-insulin and glucagon distribution within 50 nm of the plasma membrane, in-
immunostaining to visualize b- and a-cells, respectively dicating that the readily releasable pool was rapidly
(Fig. 5A), and examined b-cell morphology by electron replenished once the glucose stimulatory pathway had
microscopy (Fig. 5D) to estimate b-cell size (Fig. 5B) and been activated.
the number of insulin granules per cell (Fig. 5C). No sig- WT and KO islets exhibit comparable K+ and Ca++
nificant differences were found between the WT and signaling on b-cell stimulation. Key steps in the glucose-
PI4P5Ka KO islets for any of these parameters. We also stimulated insulin secretion pathway involve KATP channel
used quantitative RT-PCR to examine expression levels closingandmembranedepolarizationthattriggersincreased
for the other two PI4P5K isoforms, b and g, but did intracellular Ca2+, and disassembly of cortical F-actin, all of
not observe compensatory upregulation (Supplementary which could be regulated by PIP availability. To examine
2
diabetes.diabetesjournals.org DIABETES,VOL.60,FEBRUARY2011 459
PI4P5KaDEFICIENCYIMPROVESGLUCOHOMEOSTASIS
FIG.6.PI4P5KaKOisletssecreteinsulinwithalteredkineticsonK+depolarizationanddisplaycomparableCa2+influxonglucosestimulation.A:
WT and KO islets (10 islets per experimental sample, experiment was performed nine times) were isolated from age- and sex-matched mice,
preincubatedinnonstimulatoryculturemedia(2.8mmol/Lglucose),andchallengedwith60mmol/LK+.Incubationtimeswere30mininsuccession
foreachcondition.Stimulatedinsulinreleaseispresentedaftersubtractionofthebasalrelease.Dataarepresentedasmean6SEM.B:Ca2+influx
was measured by fluorescent quantification in WT and KO islets before and after glucose stimulation. The cytosolic Ca2+ concentration was
measuredusingFluo-4dye,andtherelativeintensitychange(F-F0)/F0vs.timewasplotted,whereF0andFstandforrestingfluorescenceand
fluorescenceafter20mmol/Lglucosestimulation,respectively.Tracesofindividualisletsrepresentativeofthreeexperimentsareshown.
whether differences in KATP channel closing differentially b-cells, the amount of cortical F-actin was greatly di-
triggeredinsulinsecretion,webypassedthechannelclosing minished, with much of the residual fluorescence appear-
stepusingelevatedextracellularK+todirectlydepolarizethe ing diffuse or relocalized to internal sites suggestive of
cells. As with glucose stimulation, the PI4P5Ka KO islets vesicles or stress fibers (Fig. 7A). On glucose stimulation
exhibited lower levels of insulin release under basal con- for 5 min, much of the cortical F-actin in the WT b-cells
ditions, and then levels of release on depolarization that was disassembled, with relocalization to stress fibers
weresimilartothatobservedfortheWTislets(Fig.6A).We noted in some cells (Fig. 7B); in contrast, F-actin was
also performed patch clamping on isolated b-cells to exam- virtuallycompletelyabsentinglucose-stimulatedPI4P5Ka
ine KATP channel inactivation, but no differences were KO b-cells. Quantitation of the F-actin fluorescence
notedbetweenWTandPI4P5KaKOcells(datanotshown). revealed a 60% decrease in the PI4P5Ka KO cells in com-
Theseresultssuggestedthatthestepaffectedbythelossof parison with WT islets under basal conditions and an 80%
PI4P5Ka-generated PIP lies downstream of KATP channel decreaseafterstimulationwithglucosefor5min(Fig.7C).
2
depolarization.WenextmeasuredintracellularCa2+levelsin No difference in total actin in the KO cells was observed
both WT and KO islets on glucose stimulation, but did not (Fig. 7D).
observe significant differences between the genotypes on
glucosestimulation(Fig.6B).Takentogether,thesefindings
indicate that the metabolic signaling pathway that proceeds DISCUSSION
fromglucoseimporttoincreasedintracellularCa2+isnormal Both positive and negative roles have been described for
in PI4P5Ka KO b-cells. PIP and the actin-cytoskeleton network in regulation of
2
F-actin organization is diminished at the cell cortex exocytosis and in particular insulin secretion. PIP is re-
2
membrane of PI4P5Ka KO islets. Under basal con- quired for vesicle fusion at the plasma membrane (10,11),
ditions, most of the F-actin in WT b-cells localized to the butalsoinhibitsSNAREcomplex-mediatedfusion(14)and
plasma membrane (Fig. 7A). However, in PI4P5Ka KO facilitates KATP channel activity (21–23), which opposes
460 DIABETES,VOL.60,FEBRUARY2011 diabetes.diabetesjournals.org
P.HUANGANDASSOCIATES
FIG.7.PI4P5KaKOisletsexhibitadramaticdeficiencyincorticalF-actinunderbothbasalandstimulatedconditions.WTandKOisletswere
stainedforF-actinusingrhodamine-phalloidinbefore(A)andafter(B)glucosestimulationfor5minandfixation.Pancreaticb-cellswerevi-
sualizedusinganti-insulinprimaryantibodywithconfocalmicroscopy.Shownareindividualopticalsections.Compiledsetsoftheopticalsections
arepresentedintheSupplementaryData.Scarebar,50mm.C:QuantificationofF-actinfluorescence.Fluorescenceintensitywasacquiredfrom
thetoptothebottomoftheisletsusingcontinuousxyz-scanmodeontheLeicaconfocalmicroscope.Allisletswerescannedatthesamegain
intensity,andthethree-dimensionalserieswasassembledusingtheconfocalimagingsoftware.Totalfluorescenceintensitywasdeterminedfor10
islets,andaveragefluorescenceandstandarddeviationweredetermined.**P<0.01.D:WesternblotforactinfromlysatesofWTandKOcells,
representativeoffourexperiments.(Ahigh-qualitycolorrepresentationofthisfigureisavailableintheonlineissue.)
insulin release. Cortical F-actin serves as a barrier that to the plasma membrane from regions deeper in the cell
preventssecretoryvesicleaccesstotheplasmamembrane (4).Insummary,bothturnover(14,18,20,28)andsynthesis
(18,31,43) and inhibits SNARE complex-mediated fusion (7,10,11,13,14,17–19)ofPIP arerequiredatdifferentsteps
2
(33),butitalsopromotesmovementofsecretorygranules in the secretory process, and dynamic reorganization of
diabetes.diabetesjournals.org DIABETES,VOL.60,FEBRUARY2011 461
PI4P5KaDEFICIENCYIMPROVESGLUCOHOMEOSTASIS
the actin cytoskeleton is also important (4,32,34,35). One other step in the distal part of the process is facilitated in
way that this is achieved is through the use of different the KO cells.
PIPK isoforms; PI4P5Kg and the type 2 PIP -synthesizing Regardless of the precise mechanistic relationship be-
2
enzyme, PI5P4Kb, generate a pool of PIP required for tween insulin secretion and cortical F-actin, the relatively
2
vesiclefusioninneuronsandpancreaticb-cells(7,10,11,19) subtle changes in insulin secretion seem to have a sub-
that does not substantially regulate the actin cytoskeleton. stantial effect on the obesity and hyperglycemia in the
In contrast, PI4P5Ka appears to generate a pool of PIP C57BL/6high-fatanimalmodel.Wewouldsuggestthatthe
2
that inhibits exocytosis in mast cells, potentially through elevatedfirst-phasereleaseofinsulinleadstoamorerapid
stabilizing F-actin (36). Plasma membrane PIP exists in return to euglycemia under conditions of exaggerated
2
microdomains, some of which colocalize with syntaxin nutrient uptake, decreasing the prolonged stimulation of
and promotes exocytosis (44), and others which coloc- b-cells that is associated with eventual aberrant signaling,
alize with lipid rafts and regulate cortical actin assembly blunted insulin release, and cell death. It should be noted
(44,45). PI5P4Kb colocalizes with and regulates the syn- that glucose homeostasis is a complex physiologic pro-
taxin-associated PIP microdomains (44); it is possible cess. b-cells and GSIR play a central role in the process,
2
that PI4P5Ka may localize and regulate the lipid-raft as- but other hormones and the liver, muscle, and adipose
sociated PIP microdomains. tissue are also involved. It is possible that some of these
2
This study reports alterations in insulin release by pan- other physiologic processes might contribute to the phe-
creatic b-cells in mice lacking PI4P5Ka. The KO mice ex- notype we observed. High-fat murine models are not nec-
hibit increased first-phase insulin secretion and improved essarily predictive of outcomes for type 2 diabetes in
glucose clearance under normal diet chow, and elevated humans,butthisworksuggeststheideathatdevelopment
insulin under fasting and GSIR conditions on a high-fat of therapeutics that increase early release of insulin as
diet.ChangesinK+andCa++signalingwerenotobserved, described in this report might be valuable as adjuncts for
which is consistent with the report that under normal current therapies targeted toward increasing insulin re-
physiologic conditions, the level of PIP on the plasma lease over prolonged periods of time or toward improving
2
membrane is not high enough to strongly affect KATP peripheral glucose utilization, because both of the latter
channelactivity(26),andthusdecreasingitviaablationof are associated with the side effects characteristic of ex-
PI4P5Ka has no effect on glucose-stimulated changes in cess insulin action on nondesired target tissues, such as
K+ currents. In contrast, a dramatic decrease of cortical smooth muscle vesicular endothelium, which undergoes
F-actinwasobserved,accompaniedbydecreasednumbers hypertrophyleadingtohypertensionandvasculardisease.
of insulin granules docked at the plasma membrane and
less insulin secretion under basal ex vivo conditions.
Unexpectedly differing results were observed between ACKNOWLEDGMENTS
the in vivo and ex vivo GSIR model systems. Multiple This work was supported by an award from the American
reasons may underlie the findings, but we would suggest HeartAssociationtoP.H.andNationalInstitutesofHealth
thatdifferencesintheassayconditionscouldbethemajor awards GM-071520 and DK-64166 to M.A.F. and DK-55811
factor.Normalphysiologyinvolvesserumglucoselevelsin to J.E.P. O.Y. was supported by a National Institute of
mice of ;120 mg/dL, or 6.6 mmol/L glucose, which is General Medical Sciences Medical Scientist Training Pro-
a modestly stimulatory level under ex vivo conditions. In gram T32 training award, an individual F31 National
the ex vivo assay, however, islets were placed in low Research Service Award from the National Institute of
glucose, 2.8 mmol/L, or ;50 mg/dL, a nonstimulatory DiabetesandDigestiveandKidneyDiseases(NIDDK),and
condition,before assayingbasalinsulinrelease.Itislikely the Turner Foundation. M.O. was supported by an indi-
thatunder“basal”conditionsinvivo(6.6mmol/Lglucose), vidualF31NationalResearchServiceAwardfromNIDDK.
alowchroniclevelofb-cellstimulationleadstocontinued P.T. was supported by an NIDDK postdoctoral fellow T32
transport of insulin granules to the plasma membrane. In training award.
pancreaticb-cellsfromPI4P5KaKOmicewithdiminished No potential conflicts of interest relevant to this article
cortical F-actin, the barriers to fusion are decreased, and were reported.
exocytosis rapidly occurs, leading to elevated insulin re- P.H. and O.Y. researched data and wrote the article.
lease, sustained by the continued replacement of the H.Z., P.T., W.S, X.Y., S.T., and M.O. researched data. Y.K.
granulesviamicrotubuletransport.Thislow-level,chronic andJ.E.P.reviewedandeditedthearticleandcontributed
releaseofinsulinonadlibnormalchowcouldaccountfor to discussion. M.A.F. wrote the article.
the decreased serum glucose levels observed, although The authors thank S. Vanhorn (Stony Brook University)
insulinlevelswerenotsignificantlyincreasedunderfasting for electron microscopy assistance, and J. Burchfield
conditions. Under high-fat diet conditions, insulin levels (Garvan Institute of Medical Research) and Z. Ma (Mount
were increased under fasting conditions in mutants, pre- Sinai School of Medicine) for experimental advice.
sumably reflecting the elevated serum glucose that pro-
videsagreaterchronicstimulationtotheb-cells.Intheex
vivo setting (2.8 mmol/L), however, the lower level of
REFERENCES
b-cell stimulation may fail to suffice to promote the gran-
1. MuoioDM,NewgardCB.Mechanismsofdisease:molecularandmetabolic
ule transport, leading to decreased release once the
mechanismsofinsulinresistanceandbeta-cellfailureintype2diabetes.
cortical population of granules has been sufficiently di- NatRevMolCellBiol2008;9:193–205
minished, as suggested by our morphologic analysis. On 2. MichaelDJ,RitzelRA,HaatajaL,ChowRH.Pancreaticbeta-cellssecrete
elevatedglucosestimulation,however,transportresumes, insulininfast-andslow-releaseforms.Diabetes2006;55:600–607
leading to similar levels of insulin release as barriers to 3. HuangCN,ChouWC,LinLY,etal.Firstphasereleasecoefficientofinsulin
insubjectswithnormalglucosetoleranceonglucoseinfusionanalyzedby
fusion are decreased in the WT pancreatic b-cells as well.
computersimulation.Biosystems2008;91:146–157
The increased first-phase in vivo release may also suggest
4. LiG,Rungger-BrändleE,JustI,JonasJC,AktoriesK,WollheimCB.Effect
that granule transport to the plasma membrane or some of disruption of actin filaments by Clostridium botulinum C2 toxin on
462 DIABETES,VOL.60,FEBRUARY2011 diabetes.diabetesjournals.org
P.HUANGANDASSOCIATES
insulinsecretioninHIT-T15cellsandpancreaticislets.MolBiolCell1994; 26. ShyngSL,BarbieriA,GumusbogaA,etal.Modulationofnucleotidesen-
5:1199–1213 sitivity of ATP-sensitive potassium channels by phosphatidylinositol-4-
5. LiJ, LuoR, Kowluru A,LiG.NovelregulationbyRac1ofglucose- and phosphate5-kinase.ProcNatlAcadSciUSA2000;97:937–941
forskolin-inducedinsulin secretion inINS-1beta-cells.AmJPhysiol En- 27. VicogneJ,VollenweiderD,SmithJR,HuangP,FrohmanMA,PessinJE.
docrinolMetab2004;286:E818–E827 Asymmetric phospholipid distribution drives in vitro reconstituted
6. WaselleL,CoppolaT,FukudaM,etal.InvolvementoftheRab27binding SNARE-dependentmembranefusion.ProcNatlAcadSciUSA2006;103:
proteinSlac2c/MyRIPininsulinexocytosis.MolBiolCell2003;14:4103–4113 14761–14766
7. WaselleL,GeronaRR,VitaleN,MartinTF,BaderMF,RegazziR.Roleof 28. ThoreS,WuttkeA,TengholmA.Rapidturnoverofphosphatidylinositol-
phosphoinositide signaling in the control of insulin exocytosis. Mol En- 4,5-bisphosphateininsulin-secretingcellsmediatedbyCa2+andtheATP-
docrinol2005;19:3097–3106 to-ADPratio.Diabetes2007;56:818–826
8. BarkerCJ,LeibigerIB,LeibigerB,BerggrenPO.Phosphorylatedinositol 29. Hilpelä P, Vartiainen MK, Lappalainen P. Regulation of the actin cyto-
compoundsinbeta-cell stimulus-responsecoupling.AmJPhysiolEndo- skeletonbyPI(4,5)P2andPI(3,4,5)P3.CurrTopMicrobiolImmunol2004;
crinolMetab2002;283:E1113–E1122 282:117–163
9. HeathCM,StahlPD,BarbieriMA.Lipidkinasesplaycrucialandmultiple 30. HaoM,BoganJS.Cholesterolregulatesglucose-stimulatedinsulinsecre-
roles in membrane trafficking and signaling. Histol Histopathol 2003;18: tionthroughphosphatidylinositol4,5-bisphosphate.JBiolChem2009;284:
989–998 29489–29498
10. DiPaoloG,MoskowitzHS,GipsonK,etal.ImpairedPtdIns(4,5)P2syn- 31. VitaleN,CaumontAS,Chasserot-GolazS,etal.PhospholipaseD1:akey
thesisinnerveterminalsproducesdefectsinsynapticvesicletrafficking. factorfortheexocytoticmachineryinneuroendocrinecells.EMBOJ2001;
Nature2004;431:415–422 20:2424–2434
11. GongLW,DiPaoloG,DiazE,etal.Phosphatidylinositolphosphatekinase 32. Thurmond DC, Gonelle-Gispert C, Furukawa M, Halban PA, Pessin JE.
typeIgammaregulatesdynamicsoflargedense-corevesiclefusion.Proc Glucose-stimulatedinsulinsecretioniscoupledtotheinteractionofactin
NatlAcadSciUSA2005;102:5204–5209 with the t-SNARE (target membrane soluble N-ethylmaleimide-sensitive
12. Hay JC, Fisette PL, Jenkins GH, et al. ATP-dependent inositide phos- factor attachment protein receptor protein) complex. Mol Endocrinol
phorylationrequiredforCa(2+)-activatedsecretion.Nature1995;374:173– 2003;17:732–742
177 33. JewellJL,LuoW,OhE,WangZ,ThurmondDC.Filamentousactinregu-
13. GrishaninRN,KowalchykJA,KlenchinVA,etal.CAPSactsataprefusion latesinsulinexocytosisthroughdirectinteractionwithSyntaxin4.JBiol
stepindense-core vesicleexocytosisasaPIP2bindingprotein. Neuron Chem2008;283:10716–10726
2004;43:551–562 34. OlofssonCS,HåkanssonJ,SalehiA,etal.Impairedinsulinexocytosisin
14. JamesDJ,KhodthongC,KowalchykJA,MartinTF.Phosphatidylinositol neural cell adhesionmolecule-/-mice due todefective reorganization of
4,5-bisphosphate regulates SNARE-dependent membrane fusion. J Cell thesubmembraneF-actinnetwork.Endocrinology2009;150:3067–3075
Biol2008;182:355–366 35. Bittner MA, Holz RW. Phosphatidylinositol-4,5-bisphosphate: actin dy-
15. GamperN,ReznikovV,YamadaY,YangJ,ShapiroMS.Phosphatidyli- namicsandtheregulationofATP-dependentand-independentsecretion.
nositol [correction] 4,5-bisphosphate signalsunderlie receptor-specific MolPharmacol2005;67:1089–1098
Gq/11-mediatedmodulationofN-typeCa2+channels.JNeurosci2004; 36. SasakiJ,SasakiT,YamazakiM,etal.Regulationofanaphylacticresponses
24:10980–10992 byphosphatidylinositol phosphatekinasetypeIalpha.JExpMed2005;
16. DelmasP,CosteB,GamperN,ShapiroMS.Phosphoinositidelipidsecond 201:859–870
messengers: new paradigms for calcium channel modulation. Neuron 37. ZhangJ,LuoR,WuH,WeiS,HanW,LiG.RoleoftypeIalphaphospha-
2005;47:179–182 tidylinositol-4-phosphate5-kinaseininsulinsecretion,glucosemetabolism,
17. IshiharaH,WadaT,KizukiN,etal.Enhancedphosphoinositidehydrolysis andmembranepotentialinINS-1beta-cells.Endocrinology2009;150:2127–
viaoverexpressionofphospholipaseCbeta1ordelta1inhibitsstimulus- 2135
induced insulin release in insulinomaMIN6 cells. BiochemBiophys Res 38. MoriokaT,AsilmazE,HuJ,etal.Disruptionofleptinreceptorexpression
Commun1999;254:77–82 inthepancreasdirectlyaffectsbetacellgrowthandfunctioninmice.J
18. LawrenceJT,BirnbaumMJ.ADP-ribosylationfactor6regulatesinsulinse- ClinInvest2007;117:2860–2868
cretion through plasma membrane phosphatidylinositol 4,5-bisphosphate. 39. FearnsideJF,DumasME,RothwellAR,etal.Phylometabonomicpatterns
ProcNatlAcadSciUSA2003;100:13320–13325 of adaptation to high fat diet feeding in inbred mice. PLoS One 2008;3:
19. OlsenHL,HoyM,ZhangW,etal.Phosphatidylinositol4-kinaseservesas e1668
a metabolic sensor and regulates priming of secretory granules in pan- 40. OudeWeerninkPA,SchmidtM,JakobsKH.Regulationandcellularroles
creaticbetacells.ProcNatlAcadSciUSA2003;100:5187–5192 ofphosphoinositide5-kinases.EurJPharmacol2004;500:87–99
20. Hammond GR, Dove SK, Nicol A, Pinxteren JA, Zicha D, Schiavo G. 41. GomiH,MizutaniS,KasaiK,ItoharaS,IzumiT.Granuphilinmolecularly
Eliminationofplasmamembranephosphatidylinositol(4,5)-bisphosphate docksinsulingranulestothefusionmachinery.JCellBiol2005;171:99–109
isrequiredforexocytosisfrommastcells.JCellSci2006;119:2084–2094 42. OhE,ThurmondDC.Munc18cdepletionselectivelyimpairsthesustained
21. BaukrowitzT,SchulteU,OliverD,etal.PIP2andPIPasdeterminantsfor phaseofinsulinrelease.Diabetes2009;58:1165–1174
ATPinhibitionofKATPchannels.Science1998;282:1141–1144 43. IvarssonR,JingX,WaselleL,RegazziR,RenströmE.Myosin5acontrols
22. Shyng SL, Nichols CG. Membrane phospholipid control of nucleotide insulin granule recruitment during late-phase secretion. Traffic 2005;6:
sensitivityofKATPchannels.Science1998;282:1138–1141 1027–1035
23. LinCW,YanF,ShimamuraS,BargS,ShyngSL.Membranephosphoino- 44. AoyagiK,SugayaT,UmedaM,YamamotoS,TerakawaS,TakahashiM.
sitidescontrolinsulinsecretionthroughtheireffectsonATP-sensitiveK+ Theactivationofexocytoticsitesbytheformationofphosphatidylinositol
channelactivity.Diabetes2005;54:2852–2858 4,5-bisphosphatemicrodomainsatyntaxinclusters.JBiolChem2005;280:
24. FanZ,MakielskiJC.AnionicphospholipidsactivateATP-sensitivepotas- 17346–17352
siumchannels.JBiolChem1997;272:5388–5395 45. LauxT,FukamiK,ThelenM,GolubT,FreyD,CaroniP.GAP43,MARCKS,
25. Zhang L, Marcu MG, Nau-Staudt K, Trifaró JM. Recombinant scinderin andCAP23modulatePI(4,5)P(2)atplasmalemmalrafts,andregulatecell
enhances exocytosis, an effect blocked by two scinderin-derived actin- cortexactindynamicsthroughacommonmechanism.JCellBiol2000;149:
bindingpeptidesandPIP2.Neuron1996;17:287–296 1455–1472
diabetes.diabetesjournals.org DIABETES,VOL.60,FEBRUARY2011 463