2007-pnas-ogawa

?Klotho is required for metabolic activity of fibroblast growth factor 21

Yasushi Ogawa*,Hiroshi Kurosu*,Masaya Yamamoto*,Animesh Nandi*,Kevin P.Rosenblatt*,Regina Goetz ?,Anna V.Eliseenkova ?,Moosa Mohammadi ?,and Makoto Kuro-o*?

*Department of Pathology,University of Texas Southwestern Medical Center,6000Harry Hines Boulevard,Dallas,TX 75390;and ?Department of Pharmacology,New York University School of Medicine,550First Avenue,MSB 425,New York,NY 10016

Edited by Michael S.Brown,University of Texas Southwestern Medical Center,Dallas,TX,and approved March 26,2007(received for review February 21,2007)

Fibroblast growth factor 21(FGF21)is a liver-derived endocrine factor that stimulates glucose uptake in adipocytes.Here,we show that FGF21activity depends on ?Klotho,a single-pass transmem-brane protein whose expression is induced during differentiation from preadipocytes to adipocytes.?Klotho physically interacts with FGF receptors 1c and 4,thereby increasing the ability of these FGF receptors to bind FGF21and activate the MAP kinase cascade.Knockdown of ?Klotho expression by siRNA in adipocytes dimin-ishes glucose uptake induced by FGF21.Importantly,administra-tion of FGF21into mice induces MAP kinase phosphorylation in white adipose tissue and not in tissues without ?Klotho expres-sion.Thus,?Klotho functions as a cofactor essential for FGF21activity.

Klotho ?glucose ?adipocyte ?siRNA ?GLUT1

F

ibroblast growth factor 21(FGF21)was identified based on cDNA sequence homology to other FGFs (1).Phylogenetic and structural analyses have assigned FGF21to the FGF19subfamily,which consists of FGF15(the mouse ortholog of human FGF19),FGF19,FGF21,and FGF23(2).The FGF19subfamily members distinguish themselves from the other 15FGFs in that they function in an endocrine fashion.FGF23is secreted primarily from bone and acts on the kidney to inhibit phosphate reabsorption and vitamin D biosynthesis (3–5).FGF15is expressed by intestinal epithelium and is involved in the negative feedback regulation of bile acid synthesis in the liver (6).FGF21is expressed predominantly in the liver and has emerged as a metabolic regulator of glucose uptake in adipocytes during a search for novel agents with therapeutic potential to treat diabetes mellitus (7).Indeed,administration of recombinant FGF21lowered blood glucose levels in both obese mice and in diabetic mice.Furthermore,transgenic mice that overexpress FGF21were hypoglycemic,sensitive to insulin,and resistant to diet-induced obesity (7).

FGF21can activate FGF receptors (FGFRs)and signaling molecules downstream,including FGFR substrate 2?(FRS2?)and 44/42MAP kinase (ERK1/2),in adipocytes (7–10).How-ever,efforts to demonstrate a direct interaction between FGFRs and FGF21have failed.In addition,various cell types of nonadipocyte origin including 3T3-L1preadipocytes do not respond to FGF21even though they express multiple FGFR isoforms (7).Furthermore,BaF3cells that overexpress FGFRs require suprapharmacological doses of FGF21(200–800nM)to produce a detectable mitogenic response (8–10).These obser-vations suggest that a cofactor(s)may be necessary for FGF21to activate FGF signaling in adipocytes.

We and others identified Klotho,a single-pass transmembrane protein,as an essential cofactor for FGF23to activate FGF23signaling (11,12).Klotho was originally identified as a gene mutated in the klotho mouse that exhibited phenotypes resem-bling human premature-aging syndromes (13).Major phenotypic overlaps were observed between Klotho-deficient mice and Fgf23knockout mice (14,15),suggesting that Klotho and FGF23

may work in a common signal transduction pathway(s).Indeed,Klotho bound to multiple FGFRs and was necessary for FGF23to bind FGFRs and activate FGF signaling in various types of cultured cells (11).In this work we show that ?Klotho,a Klotho family protein,functions as the cofactor required for FGF21signaling.

Results and Discussion

FGF21Requires ?Klotho to Activate FGF Signaling.Neither FGF21

nor FGF23signaled in parental 293cells as evidenced by lack of induction of phosphorylation of FRS2?and ERK1/2(Fig.1).We previously reported that ectopic overexpression of Klotho con-ferred responsiveness to FGF23on 293cells (11).However,these Klotho-overexpressing cells did not respond to FGF21.In contrast,293cells expressing the related ?Klotho protein ac-quired the ability to respond to FGF21(Fig.1).We have shown that Klotho promotes FGF23signaling by forming a ternary complex with FGF23and FGFRs.Hence,we tested whether ?Klotho plays a similar role in FGF21signaling and forms a ternary complex with FGFRs.

?Klotho Binds to Multiple FGFRs.The mammalian FGFRs are encoded by four distinct genes (FGFR1–FGFR4).The ectodo-main of prototypical FGFRs consists of three Ig-like domains (D1–D3).A major alternative mRNA splicing event within the D3of FGFR1–3generates ‘‘b’’and ‘‘c’’isoforms,which have distinct FGF-binding specificities.An additional splicing event generates shorter FGFR1–3isoforms lacking D1and/or D1–D2linker (9).We transiently expressed different FGFR isoforms,depicted in Fig.2,and ?Klotho in 293cells and performed coimmunoprecipitation experiments.As shown in Fig.2,FGFR1c and FGFR4precipitated ?Klotho more efficiently than the other FGFRs.In contrast,the FGFR b isoforms did not pull down ?Klotho under these experimental conditions.The pref-erential binding of ?Klotho to FGFR c isoforms and FGFR4is reminiscent of that of Klotho (11).

Consistent with the strong interaction between ?Klotho and FGFR1c and 4,?Klotho–FGFR1c and ?Klotho–FGFR4com-plexes were able to pull down FGF21more efficiently than FGFRs alone and the other ?Klotho–FGFR combinations,

Author contributions:Y.O.and H.K.contributed equally to this work;Y.O.,H.K.,and M.K.designed research;Y.O,H.K.,M.Y.,A.N.,and M.K.performed research;A.N.,K.P.R.,R.G.,A.V.E.,and M.M.contributed new reagents/analytic tools;Y.O.,H.K.,and M.K.analyzed data;and Y.O.,H.K.,M.M.and M.K.wrote the paper.The authors declare no con?ict of interest.This article is a PNAS Direct Submission.

Freely available online through the PNAS open access option.

Abbreviations:FGFR,?broblast growth factor receptor;FRS,?broblast growth factor receptor substrate;GLUT,glucose transporter;WAT,white adipose tissue.

?To whom correspondence should be addressed.E-mail:makoto.kuro-o@utsouthwestern.

edu.

?2007by The National Academy of Sciences of the USA

7432–7437?PNAS ?May 1,2007?vol.104?no.18

https://www.wendangku.net/doc/c82407072.html, ?cgi ?doi ?10.1073?

pnas.0701600104

demonstrating that FGF21requires ?Klotho to bind to its cognate FGFRs stably (Fig.3).

FGF21Activity Depends on ?Klotho Expression in Adipocytes.Be-

cause FGF21stimulates glucose uptake in differentiated adipo-

cytes but not in preadipocytes (7),we reasoned that differenti-ated adipocytes and not preadipocytes express ?Klotho.Indeed,we could not detect expression of ?Klotho in preadipocytes,and furthermore,FGF21failed to elicit a signal in these cells.The differentiated adipocytes,however,expressed ?Klotho abun-dantly and responded robustly to FGF21(Fig.4A ).To substan-tiate further the dependence of FGF21on ?Klotho,we followed ?Klotho expression as preadipocytes differentiated into adipo-cytes.Expression of ?Klotho was detected as early as day 4and continued to increase up to day 10.This temporal increase in ?Klotho expression correlated with the ability of FGF21to induce phosphorylation of FRS2?and ERK1/2(Fig.4B ).We also studied time course and dose response of FGF21-induced phosphorylation of FRS2?and ERK1/2.As shown in Fig.4C ,phosphorylation of FRS2?and ERK1/2became evident at 3and 5min,respectively,and began to decline at ?60min after cell stimulation.Activation of FGF signaling was detectable with as low as 10ng/ml (0.3nM)FGF21and saturated at 1,000ng/ml (30nM)https://www.wendangku.net/doc/c82407072.html,stly,we showed that endogenous ?Klotho was able to pull down endogenous FGFR1c in differentiated adipo-cytes (Fig.4D ),indicating that ?Klotho and FGFR1c form a complex under physiological conditions.

To provide more compelling evidence for the dependence of FGF21signaling on ?Klotho,we knocked down ?Klotho expression in differentiated adipocytes by using an siRNA approach.Four independent siRNAs against different se-quences in ?Klotho suppressed activation of both FRS2?and ERK1/2by FGF21(Fig.5A )and more importantly,abolished the effect of FGF21on glucose uptake,indicating that ?Klotho is essential for FGF21to exert its metabolic activity on adipocytes (Fig.5B ).

FGF21-induced glucose uptake in adipocytes is accompanied by up-regulation of glucose transporter 1(GLUT1),which is known to regulate insulin-independent glucose uptake,but not glucose transporter 4(GLUT4),which primarily contributes to insulin-dependent glucose uptake (7,16).Therefore,we deter-mined GLUT1and GLUT4protein levels and found that the knockdown of ?Klotho expression by siRNA also attenuated the ability of FGF21to increase GLUT1expression (Fig.5C ).Thus,FGF21–?Klotho signaling may be an important regulator in insulin-independent glucose uptake in adipocytes.

Insulin produced a greater increase in glucose uptake

than

Fig.1.?Klotho is required for FGF21signaling.The 293cells were trans-fected either with mock vector (Mock)or expression vectors for ?Klotho and Klotho,respectively,and then stimulated with vehicle (Control)or 300ng/ml recombinant human FGF21or 300ng/ml FGF23for 10min.Cell lysates were processed for immunoblotting with antibodies against phosphorylated FRS2?(pFRS2?),phosphorylated ERK1/2(pERK1/2),total ERK1/2(ERK1/2),?Klotho,or

Klotho.

Fig.2.?Klotho binds to multiple FGFRs.Lysates of 293cells transfected with expression vectors for ?Klotho and V5epitope-tagged FGFR isoforms were incubated with anti-V5antibody,and immunoprecipitated proteins were analyzed for the presence of ?Klotho (Top )or FGFR (Middle ).Antibodies used for immunoprecipitation (i.p.)and immunoblotting (i.b.)are indicated.Schemes for FGFR isoforms used in this work are shown above the results.The difference between b and c isoforms in FGFR1–3resides in the C-terminal half of the third Ig-like domain (D3)are indicated in the scheme by hatched and black boxes,respectively.Another alternative splicing event occurs within the ?rst Ig-like domain (D1)and acidic box,which generates long (L),middle (M),and short isoforms (S)in FGFR1and FGFR2.A portion of each cell lysate sample was processed for immunoblotting with anti-?Klotho antibody to con?rm even expression of ?Klotho among the cell samples (Bottom

).

Fig.3.FGF21preferentially binds to the ?Klotho–FGFR complex.The 293cells transfected with FGFR alone or cotransfected with FGFR and ?Klotho were lysed,and FGFR or the ?Klotho–FGFR complex was pulled down from cell lysate on agarose beads carrying anti-V5antibody.The beads were then incubated with conditioned medium containing murine FGF21,and bead-bound proteins were analyzed by immunoblotting for the presence of ?Klotho,FGF21,and FGFR.A portion of each cell lysate sample was processed for immunoblot analysis with anti-?Klotho antibody to con?rm even expres-sion of ?Klotho among the cell samples (Bottom ).

Ogawa et al .

PNAS ?May 1,2007?vol.104?no.18?7433

C E L L B I O L O G

Y

FGF21in 3T3-L1adipocytes (data not shown).However,this finding may not necessarily mean that the impact of FGF21on glucose metabolism is insignificant.The effect of FGF21on glucose uptake becomes evident several hours after the stimu-lation and lasts for 24h or longer (7)even though FGF21-induced Akt phosphorylation diminishes within 30min in 3T3-L1adipocytes (17).In contrast,insulin-induced glucose uptake becomes evident within minutes after stimulation in these cells and attenuates within hours.The prolonged effect of FGF21may influence glucose metabolism under physiological settings.In fact,FGF21has a potent and sustained blood glucose-lowering effect when administered into diabetic and obese mice (7).These observations imply that FGF21and insulin play distinct roles in the regulation of glucose metabolism:FGF21induces a moderate and sustained increase in glucose uptake primarily through up-regulating GLUT1expression,whereas insulin induces a strong and transient increase through promoting GLUT4translocation from the intracellular pool to the plasma membrane.

To determine the dependence of FGF21signaling on ?Klotho in an in vivo setting,we injected FGF21into mice and analyzed ERK1/2phosphorylation in white adipose tissue (WAT),skel-etal muscle,and kidney.Importantly,?Klotho was expressed in WAT but not in the skeletal muscle or kidney (Fig.6A ).Consistent with the expression pattern of ?Klotho,FGF21induced ERK1/2phosphorylation only in WAT (Fig.6B ).As a control,injection of FGF23into mice stimulated ERK1/2phos-phorylation only in the kidney where Klotho is expressed (Fig.6C ).These data provide strong in vivo evidence for requirement of ?Klotho and Klotho in FGF21-and FGF23-mediated tissue response,respectively.

The Klotho Gene Family Regulates Metabolic Activities of FGF Hor-mones.In this work,we show that ?Klotho is expressed in

adipocytes and enables FGF21to bind to and activate its cognate

FGFRs,leading to increased glucose uptake.This dependence of FGF21on ?Klotho represents the mechanism by which FGF21acts specifically on adipose tissue.Similarly,we and others have demonstrated that Klotho enables FGF23to bind and activate its cognate FGFRs.?Klotho was isolated based on cDNA sequence homology to Klotho (18).The ?klotho gene encodes a single-pass transmembrane protein that shares 41%amino acid identity with Klotho and in addition to adipose tissue is also expressed in liver and pancreas.Mice deficient in ?Klotho have overlapping phenotypes with mice lacking FGF15or FGFR4(6,19,20).These phenotypes include increased bile acid synthesis and increased expression of two key bile acid synthases,CYP7A1and CYP8B1,in the liver.Based on these genetic data and studies with primary hepatocytes,a model has emerged in which FGF15/FGF19and FGFR4and ?Klotho cooperate to regulate bile acid synthesis negatively (6).Thus,it seems plau-sible that ?Klotho acts also as a cofactor in FGF19/15signaling.In summary,the Klotho gene family may have evolved to confer tissue-specific bioactivity on FGF19subfamily members.Further investigations on Klotho and ?Klotho are expected to provide more insight into the molecular mechanisms by which the FGF19subfamily regulates glucose metabolism,bile acid synthesis,and phosphate/vitamin D metabolism.FGF21and FGF23may play a pathogenetic role in human diseases such as diabetes,obesity,chronic kidney disease,and bone disorders.Klotho and ?Klotho will be important targets of research when exploring therapeutic application of these endocrine FGFs.Materials and Methods

Expression Vectors.Expression vectors for mouse FGFRs with a V5epitope tag at the C terminus and murine FGF23(R179Q)were described previously (11).The murine ?Klotho cDNA was obtained by reverse transcriptase PCR from mRNAs of

differ-Fig.4.Adipocytes expressing ?Klotho respond to FGF21.(A )FGF21activates FRS2?and ERK1/2in differentiated 3T3-L1adipocytes.3T3-L1preadipocytes and differentiated adipocytes were stimulated with conditioned medium containing murine FGF21or murine FGF23.The same volume of conditioned medium from mock-transfected 293cells was used as a negative control (Control).Recombinant FGF1(100ng/ml)was used as a positive control.(B )?Klotho is expressed upon differentiation of preadipocytes to adipocytes.Differentiation of 3T3-L1preadipocytes was induced,and the cells were stimulated for 10min with the FGF21-containing conditioned medium on the indicated days after differentiation.Cell lysate was prepared for immunoblot analysis with the antibodies indicated.(C )Dose response and time course of FGF21signaling in 3T3-L1adipocytes.Differentiated 3T3-L1adipocytes were stimulated with increasing doses of recombinant human FGF21for 10min (Left )or stimulated with 1?g/ml recombinant human FGF21for various time periods (Right ).Cell lysates were subjected to immunoblot analysis by using the antibodies indicated.(D )Endogenous ?Klotho and FGFR1c form a complex in adipocytes.Lysate of 3T3-L1adipocytes was subjected to immunoprecipitation (i.p.)with anti-FGFR1antibody or normal IgG followed by immunoblot (i.b.)analysis with anti-?Klotho antibody or anti-FGFR1antibody.Cell lysate prepared from differentiated 3T3-L1adipocytes (Upper ,representing 5%of the input)or from 293cells transfected with FGFR1c(L)and FGFR1c(S)(Lower ,serving as size markers for these receptors)was directly subjected to immunoblot analysis to verify the presence of ?Klotho and FGFR1c,respectively (Cell lysate).

7434?https://www.wendangku.net/doc/c82407072.html, ?cgi ?doi ?10.1073?pnas.0701600104

Ogawa et al

.

entiated 3T3-L1adipocytes.cDNA encoding murine FGF21(IMAGE Clone;Invitrogen,Carlsbad,CA)or murine ?Klotho was cloned into pEF1expression vector (Invitrogen).Before subcloning,a FLAG epitope tag was added to the C terminus of ?Klotho,and appropriate restriction enzyme sites were added to both ends by using synthetic oligonucleotides and PCR.

Cell Culture.The 3T3-L1preadipocytes (American Type Culture

Collection,Rockville,MD)were maintained in DMEM con-taining 10%calf serum (Mediatech,Herndon,VA).Differen-tiation to adipocytes was induced by culturing the cells for 2days in differentiation medium [DMEM (American Type Culture Collection)/10%FBS/10mM Hepes/MEM nonessential amino acids (NEAA)/penicillin/streptomycin (PC/SM)(all from Invitrogen)/2?M insulin/1?M dexamethasone/0.25mM 3-isobutyl-1-methylxanthine (IBMX)(all from Sigma–Aldrich,St.Louis,MO)]and then culturing in differentiation medium without dexamethasone and IBMX for another 2days.There-after,the medium was changed every 2days with DMEM supplemented with 10%FBS/10mM Hepes/NEAA/PC/SM.Accumulation of lipid droplets was observed in ?95%of cells after 7days,and the cells at day 7–10were used for experiments.

Preparation of FGF21and FGF23.Human recombinant FGF21and

FGF23(R179Q)were expressed in Escherichia coli ,refolded in vitro ,and purified by affinity,ion-exchange,and size-exclusion chromatographies following a previously published protocol (21).Serum-free conditioned medium containing murine FGF21was collected from 293cells transiently transfected with the FGF21expression vector.The activity of FGF21in the cell culture medium was determined by comparing its ability to induce ERK1/2phosphorylation in differentiated 3T3-L1adi-pocytes with that of recombinant human FGF21of known concentration.The activity of murine FGF23(R179Q)present in the cell culture medium was determined by using 293cells stably expressing Klotho as described previously (11).Condi-tioned medium with activity equivalent to that of 2,000ng/ml (67nM)recombinant human FGF21and 300ng/ml (10nM)FGF23,respectively,was used to stimulate cultured cells.The same amount of serum-free conditioned medium from mock-transfected 293cells was used as a negative control.

Immunoprecipitation and Immunoblotting.Subconfluent 293cells

were transfected with expression vectors for ?Klotho and FGFRs 36h before the experiments by using Lipofectamine as carrier (Invitrogen).The cells were lysed in buffer containing inhibitors for phosphatase and proteinase as described previously (22).After saving a portion of each cell lysate sample for immunoblotting with anti-?Klotho antibody,the cell lysates were incubated with anti-V5–agarose beads (Sigma–Aldrich)at 4°C for 3h.The beads were washed four times with Tris-buffered saline containing 1%Triton X-100(TBST);bead-bound proteins were eluted with

Laemmli

Fig.5.FGF21requires ?Klotho to activate FGF signaling,increase glucose uptake,and increase GLUT1expression in adipocytes.(A )Differentiated 3T3-L1adipocytes were transfected with four different siRNAs for ?Klotho (?Klotho1–4)or nontargeting control siRNAs (Control 1,2);2days after transfection,the cells were stimulated for 10min with recombinant human FGF21,FGF1,or vehicle.Cell lysates were immunoblotted (i.b.)with antibodies as indicated.(B )Differentiated 3T3-L1adipocytes transfected with ?Klotho or control siRNAs as in A were incubated for 18h either with recombinant human FGF21or with vehicle and then assayed for glucose uptake.The results are presented as means ?SD error bars (n ?3).*,P ?0.001by Student’s t test.(C )Expression of GLUT1and GLUT4was determined by immunoblot analysis by using the cell lysates from B

.

Fig.6.FGF21activates FGF signaling in WAT in mice.(A )Expression of ?Klotho and Klotho in hindlimb muscle (Muscle),perigonadal fat pad tissue (WAT),and kidney from two wild-type mice was examined by immunoblot analysis.Actin blot is shown as a loading control.(B )Muscle,fat,and renal tissues were excised from mice treated either with recombinant human FGF21(n ?2)or with vehicle (n ?2).Tissue lysates were prepared for immunoblot analysis by using the antibodies indicated.(C )As in B ,except that recombinant human FGF23was injected into mice.

Ogawa et al .

PNAS ?May 1,2007?vol.104?no.18?7435

C E L L B I O L O G

Y

sample buffer,electrophoresed,and then transferred to Hybond C Extra membrane (Amersham Biosciences,Piscataway,NJ).The protein blots were incubated with anti-?Klotho antibody (R&D Systems,Minneapolis,MN)or anti-V5antibody (Invitrogen)fol-lowed by horseradish peroxidase-conjugated anti-goat IgG (Santa Cruz Biotechnology,Santa Cruz,CA)or anti-mouse IgG (Amer-sham Biosciences).Chemiluminescence signals were developed with the SuperSignal West Dura system (Pierce,Rockford,IL).For detection of endogenous interaction between ?Klotho and FGFR1c in differentiated 3T3-L1adipocytes,cell lysate samples were incubated with anti-FGFR1c antibody (Santa Cruz Biotech-nology)and protein G–Sepharose at 4°C for 2h.The Sepharose beads were washed four times with complete lysis buffer and then twice with lysis buffer lacking Triton X-100.Bead-bound proteins were eluted with Laemmli sample buffer and subjected to immu-noblot analysis by using anti-?Klotho or anti-FGFR1c antibody.

FGF21Pull-Down Experiments.Cell lysate samples prepared from

293cells transfected with FGFR alone or from 293cells co-transfected with FGFR and ?Klotho were incubated with anti-V5-agarose beads at 4°C for 3h.The beads were washed four times with TBST and then incubated with serum-free condi-tioned medium containing murine FGF21at 4°C for 3h.Thereafter,the beads were washed three times with Krebs–Ringer–Hepes buffer (118mM NaCl/4.96mM KCl/2.54mM CaCl 2/1.19mM KH 2PO 4/1.19mM MgSO 4/20mM Hepes,pH 7.4)containing 1%Triton X-100followed by three washes with the same buffer lacking Triton X-100.Bead-bound proteins were eluted with Laemmli sample buffer and subjected to immunoblot analysis by using anti-V5antibody,anti-?Klotho antibody,or anti-mouse FGF21antibody (R&D Systems).

Immunoblot Analysis of FGF Signaling.Cells cultured on multiwell

plates were serum-starved overnight and then treated for 10min either with human recombinant FGF21or FGF23(R179Q)or FGF1(FGF1was from Upstate Biotechnology,Lake Placid,NY).The cells were snap-frozen in liquid nitrogen and processed for immunoblot analysis by using anti-phospho-FRS2?antibody (Cell Signaling Technology,Beverly,MA),anti-phospho-44/42MAP kinase (ERK1/2)antibody (Cell Signaling Technology),and anti-ERK antibody (Cell Signaling Technology)as described previously.Wild-type,8-week-old,male inbred 129sv mice were administered either human recombinant FGF21(0.3?g g ?1body weight)or FGF23(R179Q,0.1?g g ?1body weight)or vehicle (10mM Hepes,pH 7.4/150mM NaCl)by injection into the inferior vena cava.Perigonadal fat pads,kidneys,and hind limb muscles were excised 15,17,and 19min,respectively,after protein injection.The tissues were flash-frozen in liquid nitrogen and processed for immunoblot analysis by using anti-phospho-ERK,anti-ERK,and anti-actin (Chemicon International,Temecula,CA)antibodies.All animal experiments were approved by the Institutional Animal Care and Research Advisory Committee of the University of Texas South-western Medical Center at Dallas.

Knockdown of ?Klotho by RNA Interference.3T3-L1adipocytes

were transfected with siRNA duplexes by electroporation as described previously (23).Briefly,differentiated 3T3-L1adipo-cytes were harvested by using 0.5mg/ml collagenase (Sigma–Aldrich),washed twice with PBS,and suspended in PBS (2?107per ml).The cells were mixed with 5nmol per 107cells of siRNA oligonucleotide against four different sequences in ?Klotho (?Klotho 1:sense,CGU GUU UGG UUA UAC GGC CUG GAC U;antisense,AGU CCA GGC CGU AUA ACC AAA CAC G;?Klotho 2:sense,CGC AGU UUA CCG AUC CUC ACC UGU A;antisense,UAC AGG UGA GGA UCG GUA AAC UGC G;?Klotho 3:sense,CAG UUU GCU CUG GAC UGG ACC UCU A;antisense,UAG AGG UCC AGU CCA GAG CAA ACU G;?Klotho 4:sense,GCG GAA GAC ACA GAC UGC ACC AUU U;antisense,AAA UGG UGC AGU CUG UGU CUU CCG C;Invitrogen)or nontargeting random-ized sequences (control 1:sense,CAG UCG UGG UCG UCA CCA GUU UCU A;antisense,UAG AAA CUG GUG ACG ACC ACG ACU G;control 2:sense,GCG CAG AGA CAG UCA CCA CAA GUU U;antisense,AAA CUU GUG GUG ACU GUC UCU GCG C;Invitrogen),and electroporated with a gene pulser system at the setting of 0.18kV and 960?F capacitance (Bio-Rad,Hercules,CA).Immediately after elec-troporation,the cells were mixed with complete medium and incubated for 10min before reseeding onto collagen I-coated 12-well plates (Becton Dickinson Labware,Bedford,MA).Thirty to forty hours after transfection,the cells were serum-starved overnight and then stimulated with FGF21or FGF1and assayed for phosphorylation of FRS2?and ERK1/2and for glucose uptake

Glucose Uptake Assay.Differentiated 3T3-L1cells transfected

with either ?Klotho siRNA or control siRNA were treated with 1?g/ml human recombinant FGF21in DMEM supplemented with 0.1%free fatty acid (FFA)-free BSA (Sigma–Aldrich)for 18h at 37°C.Unstimulated cells served as a negative control.The wells were washed with Krebs–Ringer–Hepes buffer supple-mented with 0.1%FFA-free BSA and then incubated in the same buffer supplemented with 2-deoxy-D -[1-3H]glucose (0.4?Ci,0.1mM;Amersham Biosciences)for 1h.The reaction was stopped by washing the cells twice with ice-cold PBS containing 20?M cytochalasin B (Sigma–Aldrich)followed by snap-freezing in liquid nitrogen.Cell-associated radioactivity was determined by liquid scintillation counting.Nonspecific deoxyglucose uptake was measured in the presence of 50?M cytochalasin B and subtracted from each sample to obtain specific uptake.A portion of each cell sample was saved immediately before the addition of D -[2-3H]glucose and processed for immunoblotting with anti-?Klotho antibody,anti-GLUT1antibody (H-43;Santa Cruz Biotechnology)and anti-GLUT4antibody (H-61;Santa Cruz Biotechnology).

We thank E.C.Friedberg at the University of Texas (UT)Southwestern for critical reading of the manuscript and T.Inagaki,M.Choi,and S.Kliewer at UT Southwestern for helpful discussion.This work was supported in part by grants from The Eisai Research Fund (to M.K.),Ellison Medical Foundation Grant AG-SS-1459-05(to M.K.),The Ted Nash Long Life Foundation (to M.K.),the Irma T.Hirschl Fund (to M.M.),and National Institutes of Health Grants R01AG19712(to M.K.),R01AG25326(to M.K.and K.P.R.),and R01DE13686(to M.M.).

1.Nishimura T,Nakatake Y,Konishi M,Itoh N (2000)Biochim Biophys Acta 1492:203–206.

2.Itoh N,Ornitz DM (2004)Trends Genet 20:563–569.

3.The ADHR Consortium (2000)Nat Genet 26:345–348.

4.Quarles LD (2003)Am J Physiol 285:E1–E9.

5.Schiavi SC,Kumar R (2004)Kidney Int 65:1–14.

6.Inagaki T,Choi M,Moschetta A,Peng L,Cummins CL,McDonald JG,Luo G,Jones SA,Goodwin B,Richardson JA,et al .(2005)Cell Metab 2:217–225.

7.Kharitonenkov A,Shiyanova TL,Koester A,Ford AM,Micanovic R,Gal-breath EJ,Sandusky GE,Hammond LJ,Moyers JS,Owens RA,et al .(2005)J Clin Invest 115:1627–1635.

8.Ibrahimi OA,Zhang F,Eliseenkova AV,Itoh N,Linhardt RJ,Mohammadi M (2004)Hum Mol Genet 13:2313–2324.

9.Mohammadi M,Olsen SK,Ibrahimi OA (2005)Cytokine Growth Factor Rev 16:107–137.

10.Zhang X,Ibrahimi OA,Olsen SK,Umemori H,Mohammadi M,Ornitz DM

(2006)J Biol Chem 281:15694–15700.

11.Kurosu H,Ogawa Y,Miyoshi M,Yamamoto M,Nandi A,Rosenblatt KP,

Baum MG,Schiavi S,Hu MC,Moe OW,Kuro-o M (2006)J Biol Chem 281:6120–6123.

12.Urakawa I,Yamazaki Y,Shimada T,Iijima K,Hasegawa H,Okawa K,Fujita

T,Fukumoto S,Yamashita T (2006)Nature 444:770–774.

7436?https://www.wendangku.net/doc/c82407072.html, ?cgi ?doi ?10.1073?pnas.0701600104Ogawa et al

.

13.Kuro-o M,Matsumura Y,Aizawa H,Kawaguchi H,Suga T,Utsugi T,Ohyama

Y,Kurabayashi M,Kaname T,Kume E,et al.(1997)Nature390:45–51. 14.Shimada T,Kakitani M,Yamazaki Y,Hasegawa H,Takeuchi Y,Fujita T,

Fukumoto S,Tomizuka K,Yamashita T(2004)J Clin Invest113:561–568. 15.Razzaque MS,Sitara D,Taguchi T,St-Arnaud R,Lanske B(2006)FASEB J

20:720–722.

16.Pessin JE,Bell GI(1992)Annu Rev Physiol54:911–930.

17.Moyers JS,Shiyanova TL,Mehrbod F,Dunbar JD,Noblitt TW,Otto KA,

Reifel-Miller A,Kharitonenkov A(2007)J Cell Physiol210:1–6.

18.Ito S,Kinoshita S,Shiraishi N,Nakagawa S,Sekine S,Fujimori T,Nabeshima

Y(2000)Mech Dev98:115–119.19.Ito S,Fujimori T,Furuya A,Satoh J,Nabeshima Y,Nabeshima Y(2005)J Clin

Invest115:2202–2208.

20.Yu C,Wang F,Kan M,Jin C,Jones RB,Weinstein M,Deng CX,McKeehan

WL(2000)J Biol Chem275:15482–15489.

21.Plotnikov AN,Hubbard SR,Schlessinger J,Mohammadi M(2000)Cell

101:413–424.

22.Kurosu H,Yamamoto M,Clark JD,Pastor JV,Nandi A,Gurnani P,McGuin-

ness OP,Chikuda H,Yamaguchi M,Kawaguchi H,et al.(2005)Science 309:1829–1833.

23.Jiang ZY,Zhou QL,Coleman KA,Chouinard M,Boese Q,Czech MP(2003)

Proc Natl Acad Sci USA100:7569–7574.

Ogawa et al.PNAS?May1,2007?vol.104?no.18?7437C E L L B I O L O G

Y