Biochimica et Biophysica Acta 1853 (2015) 1022–1034

Contents lists available at ScienceDirect

Biochimica et Biophysica Acta journal homepage: www.elsevier.com/locate/bbamcr

Fatty acylated caveolin-2 is a substrate of insulin receptor tyrosine kinase for insulin receptor substrate-1-directed signaling activation Hayeong Kwon 1, Jaewoong Lee 1, Kyuho Jeong 1, Donghwan Jang, Yunbae Pak ⁎ Department of Biochemistry, Division of Applied Life Science (BK21 Plus Program), PMBBRC, Gyeongsang National University, Jinju 660-701, Republic of Korea

a r t i c l e

i n f o

Article history: Received 23 December 2014 Received in revised form 30 January 2015 Accepted 1 February 2015 Available online 7 February 2015 Keywords: Caveolin-2 Fatty acylation Insulin receptor tyrosine kinase Insulin receptor substrate-1 Glucose uptake Cell proliferation and survival

a b s t r a c t Here, we demonstrate that insulin receptor (IR) tyrosine kinase catalyzes Tyr-19 and Tyr-27 phosphorylation of caveolin-2 (cav-2), leading to stimulation of signaling proteins downstream of IR, and that the catalysis is dependent on fatty acylation status of cav-2, promoting its interaction with IR. Cav-2 is myristoylated at Gly-2 and palmitoylated at Cys-109, Cys-122, and Cys-145. The fatty acylation deficient mutants are unable to localize in the plasma membrane and not phosphorylated by IR tyrosine kinase. IR interacts with the C-terminal domain of cav-2 containing the cysteines for palmitoylation. IR mutants, Y999F and K1057A, but not W1220S, fail interaction with cav-2. Insulin receptor substrate-1 (IRS-1) is recruited to interact with the IR-catalyzed phosphotyrosine cav-2, which facilitates IRS-1 association with and activation by IR to initiate IRS-1-mediated downstream signaling. Cav-2 fatty acylation and tyrosine phosphorylation are necessary for the IRS-1-dependent PI3K-Akt and ERK activations responsible for glucose uptake and cell survival and proliferation. In conclusion, fatty acylated cav-2 is a new substrate of IR tyrosine kinase, and the fatty acylation and phosphorylation of cav-2 present novel mechanisms by which insulin signaling is activated. © 2015 Elsevier B.V. All rights reserved.

1. Introduction Insulin receptor (IR) signaling pathways regulating metabolic and mitogenic functions have been well documented [1]. IR signaling diverges into different pathways by activating multiple substrates. Caveolins, lipid rafts-associated integral membrane proteins, are required for caveolar biogenesis and caveolin-1 (cav-1) in caveolae is known to regulate IR signaling pathway [2,3]. IR was reported to phosphorylate cav-1 at Tyr-14 via binding of scaffolding domain of cav-1 (residues 82-101) to a specific motif (residues 1220-1227; WSFGVVLW) [2,4] within kinase domain of IR [2,5,6]. Caveolin-3 (cav-3) was also reported to associate directly with IR and act as an enhancer of insulin signaling in skeletal muscle [3,7]. However, recent structural and bioinformatic analyses argue against such direct physical interactions [8,9]. In a series of investigations, we have demonstrated that caveolin-2 (cav-2) interacts with IR and is phosphorylated at Tyr-19 and Tyr-27

Abbreviations: IR, insulin receptor; Cav-2, caveolin-2; Cav-1, caveolin-1; IRS-1, insulin receptor substrate-1; PI3K, phosphoinositide 3-kinase; PM, plasma membrane; Hirc-B, human insulin receptor-overexpressed rat 1 fibroblast; 2-BP, 2-bromopalmitate; MβCD, methyl-β-cyclodextrin; CHX, cycloheximide; MBP, maltose binding protein; IP, immunoprecipitation; MTT, 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl-tetrazolium bromide; FAE, fatty acyl biotin exchange; 2-NBDG, 2-deoxy-2-[(7-nitro-2,1,3-benzoxadiazol-4yl)amino]-D-glucose; CBM, caveolin-binding motif; CSD, caveolin scaffolding domain ⁎ Corresponding author. Tel.: +82 55 772 1354; fax: +82 55 759 9363. E-mail address: [email protected] (Y. Pak). 1 These authors contributed equally to this work.

http://dx.doi.org/10.1016/j.bbamcr.2015.02.002 0167-4889/© 2015 Elsevier B.V. All rights reserved.

in response to insulin, and that the phospho-tyrosine cav-2 functions as a mitogenic activator for insulin signaling in fibroblasts [10–15]. Recently, we have shown various intracellular distribution of cav-2 [15, 16] and demonstrated that cav-2 forms homooligomeric cav-2 noncaveolar microdomain in the plasma membrane (PM), in which cav-2 facilitates IR recruiting and regulates activation of insulin signaling cascade by direct interaction with IR [16]. However, the underlying molecular mechanisms of their association and the consequent activation of downstream signaling have not been thoroughly investigated. Fatty acylation of proteins is involved in the regulation of numerous cellular processes including membrane targeting and subcellular trafficking of protein, vesicular biogenesis, receptor signaling, and protein stability [17]. Multiple studies have described that both myristoylation and palmitoylation of proteins play a role in their localization to lipid rafts. Many signaling proteins residing inner side of the PM such as Lck, Fyn, and flotillins are both myristoylated and palmitoylated and the fatty acylation mediates their raft membrane association in a regulated fashion [18–20]. Cav-1 was shown to be palmitoylated at Cys-133, Cys-143, and Cys-156. The palmitoylation was not required for its localization to caveolae [21] and c-Src-mediated Tyr-14 phosphorylation of cav-1 was dependent on the palmitoylation at Cys-156 [22]. Cav-2 has putative cysteine residues at the C-terminal domain and a glycine residue at the N-terminus, but actual palmitoylation on the cysteine residues and myristoylation of the glycine are not demonstrated. Thus, the functional consequence of fatty acylation of cav-2 is not known.

H. Kwon et al. / Biochimica et Biophysica Acta 1853 (2015) 1022–1034

Data are presented to show that (i) cav-2 is myristoylated at Gly-2 and palmitoylated at Cys-109, Cys-122, and Cys-145; (ii) the fatty acylation is indispensable for the PM localization of cav-2 to interact with IR; (iii) IR catalyzes Tyr-19 and Tyr-27 phosphorylation of the fatty acylated cav-2; and (iv) IRS-1 binding to the phospho-tyrosine cav-2 facilitates IRS-1 association with and tyrosine phosphorylation by IR for IRS-1directed metabolic and mitogenic insulin signaling activation. 2. Materials and methods

1023

expressing FLAG-tagged IR, IR was moved into the pDS_FLAG-XB vector using GatewayTM Technology (Invitrogen). A construct encoding Y999F, W1220S, and K1057A mutants were generated by using the FLAG-IR (WT) as template via EZchange site-directed mutagenesis kit. All expression vectors were verified by sequencing. The pECFP-Golgi and pECFP-endoplasmic reticulum (ER) plasmids were obtained from Clontech Laboratories. For transient expression of the plasmid constructs, HEK293T cells were transfected using a modified calcium phosphate precipitation protocol as described [10,14]. 3 T3-L1 adipocytes were transfected by electroporation as described [25,26].

2.1. Materials 2.4. Silencing of IRS-1, cav-1 and cav-2 by RNA interference Antibodies and reagents used were purchased as follows: anti-cav-2 (BD 610685), anti-cav-1 (BD 610406), anti-phospho-tyrosine (PY20) (BD 610000), anti-ERK (BD 610124), anti-IRS-1 (BD 611395), and anti-E-cadherin (BD 610182) antibodies from BD Transduction Laboratories; anti-IR (sc-711), anti-pY1162/1163-IR (sc-25103), anti-GFP (sc-9096 AC), anti- maltose binding protein (MBP) (sc-73416), antiα-tubulin (sc-5286), anti-calnexin (sc-11397), anti-Shc (sc-967), antip85α (sc-423), and anti-F-actin (sc-1616) antibodies and imatinibmesylate (sc-202180) from Santa Cruz Biotechnology; anti-pT202/ 204-ERK (#9101) and anti-pS473-Akt (#9271) antibodies from Cell Signaling; anti-pY19-cav-2 (ab3417) and anti-phospho-serine (PS) (ab17465) antibodies from Abcam; anti-pY418-Src (44660G) antibody from Invitrogen; anti-FLAG® M2 (F 1804), FITC-conjugated antimouse (F 0257), anti-rabbit (A 6154) and anti-mouse IgG-peroxidase/ HRP conjugate (A 4416) antibodies, and 2-bromopalmitate (2-BP) (21064), methyl-β-cyclodextrin (MβCD) (C4555), and cycloheximide (CHX) (C7698) from Sigma-Aldrich; PP2 (529573) from Calbiochem; human insulin from Eli Lilly; [9,10-3H]Palmitic acid (30-60ci/mmol), [9,10-3H]Myristic acid (10-60ci/mmol), and [γ-32P]ATP (6000ci/mmol) from PerkinElmer Life Sciences.

Small interfering RNA (siRNA) targets of IRS-1 and cav-1 were purchased from Bioneer Corp. (Daejon, Korea) and cav-2 and scramble control from Dharmacon. The siRNA oligonucleotides were synthesized to the following target sequences: IRS-1; sense (5´-CGGUAUCGUUUCGC AUGGA-3´) and anti-sense (5´-UCCAUGCGAAACGAUACCG-3´), cav-1; sense (5´-CAGUUGUACCAUGCAUUAA-3´) and anti-sense (5´-UUAAUG CAUGGUACAACUG-3´), cav-2; sense (5´-GUAAAGACCUGCCUAAUG GUU-3´) and anti-sense (5´-PCCAUUAGGCAGGUCUUUACUU-3´), scramble control; 5´-GGAAAGACUGUUCCAAAAA-3´. Transfection of the siRNA duplexes was carried out using DharmaFECT Transfection Reagents (Dharmacon) for 48 h. 2.5. Biochemical methods Immunoblotting, immunoprecipitation (IP), densitometry, and quantification of cholesterol were performed as described [15,16]. 2.6. In vitro kinase assay

Human insulin receptor-overexpressed rat 1 fibroblast (Hirc-B) cells [10,23] and cav-2 short-hairpin RNA (shRNA) stable Hirc-B cells established as described [16] were cultured in Dulbecco's modified Eagle's medium (DMEM) containing 5 mM D-glucose supplemented with 10% (v/v) fetal bovine serum (FBS) (Sigma), 100 nM methotrexate (Sigma), and 0.5% gentamycin (Gibco/BRL) in a 5% CO2 incubator at 37 °C. HEK293T and Rat-1 cells were grown in DMEM containing 5 mM D-glucose supplemented with 10% (v/v) FBS and 0.5% penicillin/streptomycin (Sigma) in a 5% CO2 incubator at 37 °C. 3T3L1 preadipocytes were grown in DMEM containing 25 mM D-glucose supplemented with 10% calf serum (Gibco/BRL). Confluent 3 T3-L1 preadipocytes were induced to differentiate into adipocytes as described [24].

Hirc-B cells were treated with or without 100 nM insulin for 10 min and lysed with IP buffer (1% Triton X-100, 150 mM NaCl, 10 mM Tris-HCl (pH 7.4), 1 mM EDTA, 1 mM ethylene glycol tetraacetic acid (EGTA) (pH 8.0), 0.2 mM sodium ortho-vanadate, 0.2 mM phenylmethylsulfonyl fluoride (PMSF), 0.5% Nonidet P-40, and 60 mM n-octylglucoside). Cell lysates were immunoprecipitated with anti-IR antibody and the immunoprecipitates were washed two times in kinase assay buffer (20 mM Hepes, pH 7.6, 20 mM MgCl2, 20 mM β-glycerophosphate, 1 mM dithiothreitol, 10 μM ATP, 1 mM EDTA, 1 mM sodium ortho-vanadate, and 0.4 mM PMSF). The kinase assay was conducted at 30 °C for 30 min in kinase assay buffer containing 5 μCi of [γ-32P]ATP and 20 μg of MBP-cav-2 (WT), MBP-Y19A, MBPY27A, or MBP-Y19/27A. The reaction products were resolved by SDSpolyacrylamide gel electrophoresis (PAGE) and transferred onto polyvinylidene difluoride (PVDF) membrane for autoradiography or immunoblot analysis.

2.3. Plasmid constructs, mutagenesis and transfection

2.7. In vitro binding assay

A full-length cav-2 cDNA was subcloned into pcDNA3 vector (Invitrogen) [11]. Constructs encoding wild type (WT) cav-2 and point mutants including G2A, C109A, C122A, C145A, C109/122A, C122/145A, C109/145A, C109/122/145A (3CA), Y19A, Y27A, Y19/27A, and Y19/27E were generated by PCR mutagenesis using mutated oligonucleotides. Cav-2 domain truncation mutants including Δ1-13, Δ1-46, Δ1-70, Δ47-86 and Δ120-162 were generated by using the cav-2-GFP (WT, residues 1-162) as template via EZchange site-directed mutagenesis kit (Enzynomics, Daejeon, Korea). The resulting entry vectors of WT and mutants were converted into self-constructed FLAG tagging destination expression vector (pDS_FLAG-XB vector), GFP tagging destination expression vector (pEGFP-N1 vector, Clontech Laboratories), or MBP tagging destination expression vector (pMGWA vector). IR (NM_10051.1) plasmid (Gateway PLUS shuttle clone for INSR, GCY2826-CF) was obtained from Genecopoeia. For construction of a vector

HEK293T cells were transfected with FLAG vector, FLAG-IR, FLAG-Y999F, FLAG-W1220S, or FLAG-K1057A for 48 h followed by incubation with 100 nM insulin for 10 min and lysed with IP buffer. Cell lysates (1 mg proteins) were incubated with purified MBP-cav-2, MBPΔ1-13, MBP-Δ1-46, MBP-Δ1-70, MBP-Δ47-86, or MBP-Δ120-162 (100 μg) prebound to amylose resin at 4 °C for 4 h. Amylose resins were washed four times with IP buffer. Bound proteins were analyzed by immunoblotting.

2.2. Cell culture

2.8. Metabolic labeling Hirc-B and cav-2-transfected HEK293T cells were washed with phosphate-buffered saline (PBS) and incubated in the DMEM containing 10 mg/ml fatty acid-free bovine serum albumin (BSA) for 1 h at 37 °C before incubation with 0.5 mCi/ml [9,10-3H]Palmitic acid or

1024

H. Kwon et al. / Biochimica et Biophysica Acta 1853 (2015) 1022–1034

0.4 mCi/ml [9,10-3H]Myristic acid for 4 h. The cells were lysed with IP buffer and immunoprecipitated with anti-cav-2, anti-FLAG, or antiGFP antibodies. Following extensive washing captured proteins were eluted, separated by SDS-PAGE, transferred to PVDF membrane, and exposed to light sensitive film at -80 °C using a KODAK Biomax transcreen LE (Sigma-Aldrich). The same membrane was reprobed with anti-cav-2, anti-GFP or anti-FLAG antibodies. 2.9. Fatty acyl biotin exchange Purification of palmitoylated proteins by the fatty acyl biotin exchange (FAE) was performed as described [27]. Briefly, cells were lysed with buffer A (1% SDS in PBS containing 0.2 mM PMSF, 1 μg/μl aprotinin, and 1 μg/μl leupeptin) and free cysteines alkylated by overnight incubation at 4 °C with 25 mg/ml N-ethylmaleimide (NEM). Excess soluble NEM was removed by chloroform/methanol precipitation. Protein pellets were resolubilized in buffer A, and the samples were divided into two; half was treated with 200 mM hydroxylamine (pH 7.4), and half was treated with 200 mM Tris (pH 7.4). Lysates were treated with 1 mM pyridyldithiol-activated cysteine-reactive biotinylation reagent N-[6-(Biotinamido)hexyl]-3´-(2´-pyridyldithio)-propionamide for 1 h in the dark at room temperature and excess biotinylation reagent was removed by chloroform/methanol precipitation. Protein pellets were resolubilized in buffer B (1% Triton X-100 and 0.2% SDS in PBS containing 0.2 mM PMSF, 1 μg/μl aprotinin, and 1 μg/μl leupeptin). The samples were purified using NeutrAvidin-Agarose resin beads (Thermo scientific) overnight at 4 °C. Beads were washed four times with buffer B and biotinylated proteins were eluted in 5× SDS-PAGE sample buffer. The biotinylated proteins were then resolved, separated by SDS-PAGE, and subjected to immunoblot analysis. 2.10. Confocal microscopy HEK293T cells were co-transfected with cav-2-GFP, G2A-GFP, 3CAGFP, C122/145A-GFP, C109/145A-GFP, or C109/122A-GFP and pECFPGolgi or pECFP-ER, or with cav-2-mCherry and IR-GFP for 48 h. The cells were fixed with 3.7% paraformaldehyde in PBS for 20 min and rinsed with PBS. Fluorescent images were obtained using an Olympus Fluoview 1000 confocal microscope attached to IX-81 inverted microscope equipped with PlanApo 100×/1.35 oil immersion objective lens (Olympus). GFP and CFP signals were excited using an Argon laser at 488 and 434 nm, respectively, and mCherry signal using a He-Ne laser at 543 nm. FV10-ASW software (Olympus) was used to merge the images from GFP and CFP or mCherry. White lines drawn in the merge panels were converted to plot profiles using the Line Profile Tool of Image-ProPlus 6.1 as described [14,15]. 2.11. Cell proliferation and viability assay 3T3L1 preadipocytes or adipocytes (2 × 104) were plated in 96-well plates and transfected with FLAG vector, FLAG-cav-2 (WT), or FLAG-3CA for 24 h. The cells were pretreated with or without 10 mM MβCD for 2 h followed by incubation with 100 nM insulin for 72 h. 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl-tetrazolium bromide (MTT) (M5655, Sigma-Aldrich) stock solution was added to each well at a final concentration of 0.5 mg/ml and incubated at 37 °C for 4 h, followed by lysis with 100 μl of dimethyl sulfoxide (DMSO) (BioShop) to solubilize the final product of MTT metabolism, the formazan precipitate. After 30 min incubation at 37 °C, the optical density at 540 nm was determined using a microplate reader (Model 550, BioRad Laboratories). 2.12. Glucose uptake assay Glucose uptake assay was performed using 2-deoxy-2-[(7-nitro2,1,3-benzoxadiazol-4-yl)amino]-D-glucose (2-NBDG), a fluorescent glucose analog as described [28–30]. The 2-NBDG uptake was measured

by using a Glucose Uptake Cell-based Assay Kit (600470, Cayman Chemical Company) according to the manufacturer’s protocols. 3T3L1 preadipocytes (1 × 104) were seeded in 96-well plates and differentiated into adipocytes. The adipocytes were transfected with FLAG vector, FLAG-cav-2 (WT), or FLAG-3CA for 48 h and pretreated with or without 10 mM MβCD for 2 h, or co-transfected with scramble control or cav-1 siRNA and FLAG vector, FLAG-cav-2 (WT), or FLAG-3CA for 48 h followed by incubation with or without 100 nM insulin for 10 min. Uptake was then initiated by addition of 2-NBDG (150 μg/ml). After 10 min of incubation, the adipocytes were washed and the amount of 2-NBDG was measured at excitation/emission 485/535 nm using a microplate reader (Model 550, BioRad Laboratories). 2.13. Statistical analysis Data are expressed as mean ± S.E. and analyzed in Graphpad Prism 5.0 version. Statistical analysis of more than two groups was performed by two-way ANOVA followed by Bonferroni post hoc test. Analysis comparing two groups was carried out using an unpaired Student’s t test. p values b 0.05 were considered statistically significant. 3. Results 3.1. IR tyrosine kinase catalyzes phosphorylation of cav-2 at Tyr-19 and Tyr-27 Having shown that cav-2 interacts with IR and is phosphorylated at Tyr-19 and Tyr-27 in response to insulin [11–13,16], it was important to find out whether the phosphorylation is catalyzed by IR tyrosine kinase and to confirm the direct and specific binding between cav-2 and IR. We carried out in vitro kinase assay using MBP fusions of WT cav-2 and Y19A, Y27A, and Y19/27A mutants as substrates for IR tyrosine kinase (Fig. 1A). As compared with WT cav-2, 32P incorporation to Y19A and Y27A mutants was decreased by 57.4 and 63.7%, respectively. Simultaneous mutation of both tyrosines (Y19/27A) abolished 32P incorporation, demonstrating that IR catalyzes the phosphorylation of cav-2 at Tyr-19 and Tyr-27 (Fig. 1A). Tyrosine kinase activity of IR in response to insulin was assessed by the IR phosphorylation at Tyr-1162/1163. Since cav-2 was reported to be phosphorylated at Tyr-19 and Tyr-27 in c-Src-overexpressed fibroblasts [31,32], Src-mediated phosphorylation was examined by using PP2, a selective inhibitor for Src-family kinases. PP2 treatment inhibited Akt activation but did not affect insulin-induced phosphorylation of cav-2 at Tyr-19 (Fig. 1B, a, lanes 2 vs. 4), suggesting that cav-2 is not a downstream effector for the c-Src kinase-mediated insulin signaling axis. Since c-Abl kinase was reported to phosphorylate cav-1 at Tyr-14 [33,34], we tested c-Abl kinasemediated phosphorylation of cav-2 by using imatinib-mesylate, a cAbl kinase inhibitor. Imatinib-mesylate treatment inhibited Akt activation (Fig. 1B, b, lanes 2 vs. 4) but did not affect the phosphorylation of cav-2 at Tyr-19 with or without insulin stimulation (Fig. 1B, b, lanes 1 vs. 3 and 2 vs. 4). These data show that cav-2 is directly phosphorylated by IR tyrosine kinase in response to insulin. 3.2. Cav-2 binds to the NPEY motif of IR To identify the specific binding motifs of IR and cav-2 responsible for their interaction in response to insulin, we mutated Tyr-999 within the NPEY motif in the juxtamembrane domain of IR, which was important for IR interaction with IRS-1 and Shc [35], to phenylalanine (Y999F mutant) and Trp-1220 within the reported cav-1 binding motif (residues 1220-1227; WSFGVVLW) [2,4] in tyrosine kinase domain of IR to serine (W1220S mutant). We also generated K1057A mutant, which is known to abolish IR kinase activity and interaction with IR substrates by disrupting the canonical ATP binding site [36]. IR mutant W1220S interacted with cav-2 but Y999F and K1057A did not (Fig. 1C, a). This result was further confirmed by in vitro binding assay (Fig. 1C, b). In

H. Kwon et al. / Biochimica et Biophysica Acta 1853 (2015) 1022–1034

1025

Fig. 1. IR catalyzes phosphorylation of cav-2 at Tyr-19 and Tyr-27 via a specific interaction between the NPEY motif of IR and the C-terminal domain of cav-2. (A) Hirc-B cells were treated with or without 100 nM insulin for 10 min. Whole cell lysates (WCL) were immunoprecipitated with anti-IR antibody and subjected to in vitro kinase assay using purified MBP-cav-2, MBPY19A, MBP-Y27A or MBP-Y19/27A as a substrate. (B) Hirc-B cells were treated with or without 10 μM PP2 for 30 min (a) or 100 μM imatinib-mesylate for 1 h (b) followed by incubation with or without 100 nM insulin for 10 min. WCL were subjected to immunoblot analysis with antibodies against pY19-cav-2, pY1162/1163-IR, pS473-Akt, pT202/204-ERK, pY418-Src, and actin. (C) (a) HEK293T cells were co-transfected with cav-2-GFP and FLAG vector, FLAG-IR, FLAG-Y999F, FLAG-W1220S, or FLAG-K1057A followed by incubation with 100 nM insulin for 10 min. WCL were immunoprecipitated with anti-FLAG antibody and subjected to immunoblot analysis with antibodies against GFP and FLAG. (b) HEK293T cells were transfected with FLAG vector, FLAG-IR, FLAG-Y999F, FLAG-W1220S, or FLAG-K1057A followed by incubation with 100 nM insulin for 10 min. WCL were subjected to in vitro binding assay using purified MBP-cav-2. (D) HEK293T cells were transfected with FLAG-IR followed by incubation with 100 nM insulin for 10 min. WCL were subjected to in vitro binding assay using purified MBP vector, MBP-cav-2, MBP-Δ1-13, MBP-Δ1-46, MBP-Δ1-70, MBP-Δ47-86, or MBP-Δ120-162.

parallel, various domain truncation mutants of cav-2 were generated to find out its IR-binding region. IR interacted with cav-2 mutants except Δ120-162 mutant (Fig. 1D). Unlike cav-1 and cav-3 which are reported to associate with IR thru their scaffolding domains, Δ1-70 and Δ47-86 mutants truncated on the scaffolding domain (residues 67-86) interacted with IR but Δ120-162 mutant containing the intact scaffolding domain did not (Fig. 1D). These data show that the C-terminal domain of cav-2 associates with the NPEY motif of insulin-activated IR. 3.3. Phospho-tyrosine cav-2 mediates IR-IRS-1-directed signaling activation in response to insulin To explore if cav-2 interferes IRS-1 binding to IR or vice versa, effects of cav-2 shRNA and IRS-1 siRNA on their IR association and downstream IR signaling activation were examined. In control cells, IRS-1 was coimmunoprecipitated with pY19-cav-2 and pY1162/1163-IR and tyrosine phosphorylated in response to insulin (Fig. 2A, a, lanes 1 vs. 2). In cav-2 shRNA stable cells, however, IRS-1 association with pY1162/ 1163-IR and tyrosine phosphorylation were abrogated (Fig. 2A, a, lanes 2 vs. 4). And retardation in phosphoinositide 3-kinase (PI3K) interaction with IRS-1 and decrease in PI3K activation in response to insulin were detected with attenuation in downstream Akt and ERK activations as compared to control cells (Fig. 2A, b, lanes 2 vs. 4 and Fig. 2A, c). Shc did not interact with cav-2 and insulin-induced Shc association with IR was unaffected by cav-2 depletion (Fig. 2B). Depletion of IRS-1, which caused inhibition of Akt and ERK activations (Fig. 2C, a), did not affect insulin-induced pY19-cav-2 association with pY1162/1163-IR (Fig. 2C, b). To verify the requirement of cav-2 tyrosine phosphorylation catalyzed by IR for IRS-1 association with IR, their interaction in

response to insulin was examined in the cav-2 tyrosine deficient mutants-expressed cells. Insulin-induced association between IR and IRS-1 was significantly decreased in Y19/27A mutant-expressed cells as compared to the level of their association in WT cav-2-expressed cells (Fig. 2D, lanes 4 vs. 6). These data indicate that cav-2 phosphorylation by IR tyrosine kinase is an upstream signal of the IR-catalyzed IRS-1 tyrosine phosphorylation. As the requirement of cav-2 tyrosine phosphorylation for its interaction with IRS-1 was investigated, insulininduced IRS-1 binding to Y19A, Y27A, or Y19A/27A mutant was decreased as compared to WT cav-2 (Fig. 2E). No interaction of cav-2 with Shc was further confirmed (Fig. 2E). When IRS-1-binding region of cav-2 was examined by in vitro binding assay, IRS-1 interacted with Δ1-13, Δ47-86, and Δ120-162 but not with Δ1-46 and Δ1-70 mutants (Fig. 2F), indicating that IRS-1 interacts with the N-terminal domain (residues 14-46) of cav-2 which contains Tyr-19 and Tyr-27. We then tested effects of cav-2 siRNA and tyrosine deficient Y19/27A mutant on insulin-induced IRS-1 association with and activation by IR in 3T3L1 adipocytes. Both cav-2 siRNA and Y19/27A mutant abrogated the IRS-1 interaction with pY1162/1163-IR and tyrosine phosphorylation (Fig. 3A, a, lanes 2 vs. 4 and Fig. 3B, a, lanes 4 vs. 6) and reduced the downstream activation of Akt and ERK as compared to their scramble control (Fig. 3A, b, lanes 2 vs. 4) and WT cav-2-transfected adipocytes (Fig. 3B, b, lanes 4 vs. 6). Finally, we generated a phosphomimetic cav2 tyrosine mutant, Y19/27E to verify insulin- and/or cav-2 tyrosine phosphorylation-dependent interaction between IR and cav-2 and IRS-1 and enhancement of IRS-1 tyrosine phosphorylation and downstream signaling activation. Like WT cav-2, tyrosine phosphorylation of the IRS-1 interacting with Y19/27E mutant by IR tyrosine kinase and the consequent IRS-1-directed activation of Akt and ERK were

1026

H. Kwon et al. / Biochimica et Biophysica Acta 1853 (2015) 1022–1034

Fig. 2. Cav-2 knockdown and tyrosine deficient mutation attenuate the IRS-1 interaction with IR tyrosine kinase to activate IRS-1-directed PI3K-Akt and ERK signaling. (A-C) Vector control or cav-2 shRNA stable Hirc-B cells (A, B) and scramble control or IRS-1 siRNA-transfected Hirc-B cells (C) were treated with or without 100 nM insulin for 10 min. WCL were immunoprecipitated with anti-IRS-1(A, a), anti-p85α (A, b), anti-Shc (B), or anti-IR (C, b) antibody and/or subjected to immunoblot analysis with antibodies against IR, pY1162/1163IR, cav-2, pY19-cav-2, PY20, IRS-1, p85α, Shc, pS473-Akt, Akt, pT202/204-ERK, ERK, cav-1, and actin (A, c and C, a). The densitometry ratios of p-p85α to total p85α, p-IR to total IR, p-Akt to total Akt, p-ERK to total ERK, cav-2 to actin, and IRS-1 to actin are illustrated (mean ± S.E., n = 3). NS, nonsignificant. (D) HEK293T cells were transfected with FLAG vector, FLAG-cav-2, or FLAG-Y19/27A followed by incubation with or without 100 nM insulin for 10 min. WCL were immunoprecipitated with anti-IR antibody and subjected to immunoblot analysis with antibodies against IRS-1, FLAG, pY1162/1163-IR, and IR. (E) HEK293T cells were transfected with FLAG-cav-2, FLAG-Y19A, FLAG-Y27A, or FLAG-Y19/27A followed by incubation with or without 100 nM insulin for 10 min. WCL were immunoprecipitated with anti-FLAG antibody and subjected to immunoblot analysis with antibodies against IRS-1, Shc, and FLAG. (F) HEK293T cells were treated with 100 nM insulin for 10 min and WCL were subjected to in vitro binding assay using purified MBP vector, MBP-cav-2, MBP-Δ1-13, MBP-Δ1-46, MBPΔ1-70, MBP-Δ47-86, or MBP-Δ120-162.

dependent on insulin stimulation (Fig. 3B, a and b, lanes 3 vs. 4 and 7 vs. 8). These results show that insulin-activated IR tyrosine kinasecatalyzed phospho-Tyr-19 and Tyr-27 cav-2 facilitates IRS-1 recruitment to and association with IR, thereby mediating tyrosine phosphorylation of IRS-1 by IR for the downstream IR signaling activation of the IRS-1-directed both PI3K-Akt and ERK pathways.

3.4. Cav-2 is myristoylated at Gly-2 and palmitoylated at Cys-109, Cys-122 and Cys-145 The C-terminal domain of cav-2 important for its IR binding has cysteine residues for putative palmitoylation. Cys-109, Cys-122, and Cys-145 are predicted by CSS-Palm, a palmitoylation prediction

H. Kwon et al. / Biochimica et Biophysica Acta 1853 (2015) 1022–1034

1027

Fig. 3. Insulin-activated IR-catalyzed phosphorylation of cav-2 at Tyr-19 and Tyr-27 facilitates IRS-1 recruitment to and tyrosine phosphorylation by IR for downstream signaling activation. Scramble control or cav-2 siRNA (A) and FLAG vector, FLAG-cav-2, FLAG-Y19/27A, or FLAG-Y19/27E-transfected (B) 3T3L1 adipocytes were treated with or without 100 nM insulin for 10 min. WCL were immunoprecipitated with anti-IRS-1 (A and B, a) antibody and/or subjected to immunoblot analysis with antibodies against IR, pY1162/1163-IR, cav-2, pY19-cav-2, PY20, IRS-1, FLAG, pS473-Akt, Akt, pT202/204-ERK, ERK, cav-1, and actin (A and B, b). Ecto, ectopic; Endo, endogenous. The densitometry ratios of p-IR to total IR, p-Akt to total Akt, pERK to total ERK, and cav-2 to actin are illustrated (mean ± S.E., n = 3).

algorithm [37] (Fig. 4A). To verify whether Cys-109, Cys-122, and Cys-145 are palmitoylated, we performed 3H-palmitate incorporation and FAE assays with variant cysteine deficient mutants. Both endogenous and ectopic cav-2 were metabolically palmitoylated (Fig. 4B, a and b) but mutation of all three cysteine residues to alanines (3CA) abolished 3H-palmitate incorporation (Fig. 4B, b). Palmitoylation of cav-2, along with cav-1 as a positive control [21], was further

demonstrated by FAE assay in Rat-1 fibroblasts (Fig. 4B, c). Palmitoylation of double cysteine deficient mutants (C109/122A, C122/145A, and C109/145A) was attenuated, and 3CA palmitoylation was completely abolished (Fig. 4B, d). To explore dynamics of cav-2 palmitoylation, effect of 2-BP, a nonmetabolizable analog of palmitate acting as an irreversible inhibitor of palmitoylation, treatment for various time points on the status of endogenous cav-2 palmitoylation was

Fig. 4. Fatty acylation of cav-2. (A) Schematic illustration of cav-2 showing the a.a. sequences of N-terminal residues 1-27 and C-terminal residues 100-162. Asterisks indicate a.a. residues predicted for the myristoylation and palmitoylation of cav-2. (B) Hirc-B (a) and FLAG-cav-2 or FLAG-3CA-transfected HEK293T (b) cells were subjected to metabolic labeling using 3Hpalmitate. WCL were immunoprecipitated with anti-cav-2 (a) or anti-FLAG (b) antibody and subjected to autoradiography or immunoblot analysis with antibody against cav-2 or FLAG. Rat-1 (c) and FLAG-cav-2, FLAG-C109/122A, FLAG-C122/145A, FLAG-C109/145A, or FLAG-3CA-transfected HEK293T (d) cells were subjected to FAE analysis after treatment with hydroxylamine (HA) or Tris as a control. Input shows starting material before the NeutrAvidin affinity purification. (C) (a) Hirc-B cells were treated with 50 μM CHX for 3 h followed by incubation with DMSO or 100 μM 2-BP for the indicated time periods and subjected to FAE analysis. The densitometry ratios of FAE to Input are illustrated (mean ± S.E., n = 3). ⁎⁎p b 0.01, ⁎⁎⁎p b 0.001. (b) Hirc-B cells were treated with 100 μM 2-BP and 50 μM CHX for 15 h, rinsed, and chased for the indicated times in serum-free medium (control) or 100 nM insulin containing medium (insulin) and subjected to FAE analysis. The densitometry ratios of FAE to Input are illustrated (mean ± S.E., n = 3). ⁎⁎⁎p b 0.001. (D) Cav-2-GFP or G2A-GFP-transfected HEK293T cells were subjected to metabolic labeling using 3H-myristate (a) or 3H-palmitate (b). WCL were immunoprecipitated with anti-GFP antibody and subjected to autoradiography or immunoblot analysis with antibody against GFP. The asterisk represents nonspecific bands.

1028

H. Kwon et al. / Biochimica et Biophysica Acta 1853 (2015) 1022–1034

H. Kwon et al. / Biochimica et Biophysica Acta 1853 (2015) 1022–1034

1029

1030

H. Kwon et al. / Biochimica et Biophysica Acta 1853 (2015) 1022–1034

analyzed in CHX-treated Hirc-B cells (Fig. 4C). There was no significant cav-2 palmitoylation turnover observed until 1 h exposure to 2-BP. The irreversible inhibition of cav-2 palmitoylation by 2-BP was started after 1 h exposure with 1.6-3.5 fold reduction in the level of endogenous cav-2 palmitoylation within 3-6 h exposure, suggesting that depalmitoylation of the preexisting palmitoylated cav-2 and repalmitoylation of the depalmitoylated cav-2 take place within the time frame (Fig. 4C, a). The inhibition by 2-BP was completed at 12 h exposure. We then explored if insulin induces cav-2 palmitoylation in the cells which were pretreated with CHX and 2-BP to inhibit de novo synthesis and palmitoylation of cav-2 and chased in serum-free or insulin containing medium for various periods of time (Fig. 4C, b). Palmitoylation of preexisting endogenous cav-2 was completely blocked after the pretreatment with CHX and 2-BP (at 0 time chase) and the inhibition status of cav-2 palmitoylation was not changed in serum-free medium at all the chasing time. However, insulin triggered palmitoylation of newly synthesized cav-2 as early as 15 min of insulin incubation, and insulin-induced cav-2 palmitoylation was maintained for 12 h. Distinct from cav-1 and cav-3, cav-2 displays a putative myristoylation site at Gly-2 in the N-terminus (Fig. 4A). To verify if cav-2 is myristoylated, Gly-2 was substituted by alanine (G2A) and metabolic labeling with 3H-myristate was performed. WT cav-2 was myristoylated, but G2A mutant was neither myristoylated nor palmitoylated (Fig. 4D, a and b). These results show that cav-2 is myristoylated at Gly-2 and the myristoylation is a prerequisite for its palmitoylation. 3.5. Intracellular localization of cav-2 depends on its fatty acylation status Having demonstrated that cav-2 localizes in the ER, Golgi and PM, and translocates from the Golgi to the PM [15,16], we sought to determine whether the subcellular distribution and intracellular translocation of cav-2 are dependent upon its different fatty acylation status. As demonstrated [15,16], cav-2 localized in the Golgi and the PM, and cav-2 colocalized with IR in the PM (Fig. 5A, panels 1 and 2). However, G2A, 3CA, and C122/145A mutants were retained in the ER (Fig. 5A, panels 3-5). Both C109/145A and C109/122A mutants remained mainly in the Golgi and exhibited no PM targeting (Fig. 5A, panels 6 and 7). Line profile analysis verified the colocalization with the appropriate subcellular localization markers. Inhibition of the PM targeting of the mutants was further confirmed with the purified PM fraction (Fig. 5B). These results show that myristoylation at Gly-2 and palmitoylation at Cys-122 or Cys-145 are important for cav-2 translocation from the ER to the Golgi, and that cav-2 myristoylation and palmitoylation at both Cys122 and Cys-145 are required for transport of cav-2 from the Golgi to the PM for its colocalization with IR. 3.6. Fatty acylation is required for cav-2 function as a substrate for IR tyrosine kinase and IR-cav-2-IRS-1-directed activation of mitogenic and metabolic insulin signaling We next explored requirement of cav-2 fatty acylation for its catalysis by IR, IRS-1 interaction with IR, and IRS-1-directed activation of Akt and ERK. G2A or 3CA mutant-expressed cells exhibited retardation of IR-catalyzed phosphorylation at Tyr-19 (Fig. 6A, a, lanes 4 vs. 6 and 8), abrogation of the interaction between IR and IRS-1 (Fig. 6A, b), and inhibition of the IRS-1-directed downstream signaling activation of Akt and ERK (Fig. 6A, a, lanes 4 vs. 6 and 8). These results show that cav-2 myristoylation and palmitoylation facilitate its binding to IR and the

subsequent phosphorylation catalyzed by IR tyrosine kinase to mediate the IRS-1 association with and activation by IR, thereby activating the IRS-1-directed downstream signaling. To signify functional and physiological relevance of cav-2 fatty acylation, 3T3L1 adipocytes were tested for the effect of cav-2 palmitoylation on phospho-tyrosine cav-2mediated IR-IRS-1-directed signaling activation. Ectopically introduced WT cav-2, but not 3CA mutant, was phosphorylated at Tyr-19 (Fig. 6B, a, lanes 4 vs. 6), facilitated association between IR and pY19-cav-2 and IRS-1 (Fig. 6B, b, lanes 4 vs. 6), and enhanced downstream activation of Akt and ERK (Fig. 6B, a, lanes 4 vs. 6) in response to insulin. Ectopic expression of both WT cav-2 and 3CA mutant did not affect endogenous cav-1 level (Fig. 6B, a). Caveolae-dependent IR signaling is well characterized in 3T3L1 adipocytes [38]. To distinguish the cav-2 specific activation in IR signaling from the caveolae- and cav-1-mediated regulation, 3T3L1 preadipocytes and adipocytes were pretreated with MβCD to deplete cholesterol from the PM or subjected to cav-1 depletion before insulin stimulation. We then asked if blocking cav-2 palmitoylation has impact on cell proliferation in the preadipocytes and on cell survival and glucose uptake in the adipocytes. Ectopic introduction of WT cav-2 increased insulin-induced preadipocyte proliferation by 29.1% as compared to vector control (Fig. 6C, a, lanes 1 vs. 2), but 3CA mutant had no effect (Fig. 6C, a, lanes 1 vs. 3). No significant change was observed in the adipocyte viability when WT cav-2 was introduced (Fig. 6C, b, lanes 1 vs. 2), but 3CA mutant decreased the viability by 61.5% as compared to vector control (Fig. 6C, b, lanes 1 vs. 3). The patterns of cav-2-mediated increase in the preadipocyte proliferation and cav-2 palmitoylation defect-induced decrease in the adipocytes viability were not affected from cholesterol depletion by MβCD treatment (Fig. 6C, a and b, lanes 1-3 vs. 4-6). These results show that cav-2 independently of caveolae regulates insulin-induced cell survival and proliferation via the palmitoylated cav-2-mediated IR-IRS-1-directed PI3KAkt- and ERK-dependent signaling. Insulin-stimulated glucose uptake measured by using 2-NBDG showed increased 2-NBDG uptake by 4.6 and 4.7 fold in non- and vector-transfected control adipocytes not treated with MβCD, respectively (Fig. 6C, c, lanes 1 vs. 5 and 2 vs. 6). Ectopic WT cav-2, but not 3CA mutant, enhanced the uptake by 2.0 fold as compared to the uptake level of the control adipocytes (Fig. 6C, c, lanes 6 vs. 7 vs. 8). Similar uptake patterns were observed in the control adipocytes transfected with scramble control (Fig. 6D, a, lanes 1 vs. 5 and 2 vs. 6 and lanes 6 vs. 7 vs. 8). Cholesterol depletion and cav-1 knockdown reduced insulin-induced 2-NBDG uptake by 2.8 and 2.9 fold, respectively in control adipocytes (Fig. 6C, c and D, a, 6 vs. 14). Ectopic WT cav-2 rescued the reduction to the uptake level of control adipocytes but not 3CA mutant (Fig. 6C, c and D, a, lanes 14 vs. 15 vs. 16 and 6 vs. 15), demonstrating the palmitoylated cav-2-specific and caveolae/cav-1-independent glucose uptake. Cholesterol and cav-1 levels were depleted by 70.6 and 91% in MβCD-treated and cav-1 siRNA-transfected adipocytes, respectively (Fig. 6C, d and D, b). These results provide further evidence that fatty acylated cav-2 activates insulin signaling distinct from which regulated by cav-1 and caveolae.

4. Discussion In this study, we demonstrate that cav-2 is a substrate of IR tyrosine kinase to facilitate insulin signaling. We propose a model for the regulatory mechanisms of cav-2 in insulin signaling (Fig. 7). Under basal condition, co-translationally myristoylated cav-2 at Gly-2 is synthesized in the ER. Upon insulin stimulation, myristoylated cav-2 is palmitoylated

Fig. 5. Intracellular localization of cav-2 is dependent on its fatty acylation status. (A) HEK293T cells were co-transfected with cav-2-GFP, G2A-GFP, 3CA-GFP, C122/145A-GFP, C109/145AGFP, or C109/122A-GFP and pECFP-Golgi or pECFP-ER, or cav-2-mCherry and IR-GFP and visualized by confocal microscopy. Scale bars, 10 μm. The white lines in merge panels were converted to line profiles using the Line Profile Tool. (B) HEK293T cells were transfected with FLAG-cav-2, FLAG-3CA, FLAG-C109/122A, FLAG-C122/145A, or FLAG-C109/145A and subjected to PM isolation as described [15,16]. WCL and PM fraction were subjected to immunoblot analysis with antibody against FLAG and detected with antibodies specific for E-cadherin, calnexin and α-tubulin as markers for the PM fraction, ER and cytoplasm, respectively.

H. Kwon et al. / Biochimica et Biophysica Acta 1853 (2015) 1022–1034

1031

Fig. 6. Fatty acylation of cav-2 is required for IR tyrosine kinase-catalyzed phosphorylation of cav-2 and IRS-1 and IR-cav-2-IRS-1-directed activation of mitogenic and metabolic insulin signaling distinct from the caveolae/cav-1-dependent activation. (A, B) HEK293T cells (A) or 3T3L1 adipocytes (B) were transfected with FLAG vector, FLAG-cav-2, FLAG-G2A, or FLAG3CA followed by incubation with or without 100 nM insulin for 10 min. WCL were immunoprecipitated with anti-IR antibody (A and B, b) and/or subjected to immunoblot analysis with antibodies against FLAG, IRS-1, pY1162/1163-IR, IR, pY19-cav-2, pS473-Akt, Akt, pT202/204-ERK, ERK, cav-2, cav-1, and actin (A and B, a). Ecto, ectopic; Endo, endogenous. The densitometry ratios of pY19-cav-2 to FLAG, p-Akt to total Akt, and p-ERK to total ERK are illustrated (mean ± S.E., n = 3). The asterisks represent nonspecific bands. (C) FLAG vector, FLAG-cav-2, or FLAG-3CA-transfected 3T3L1 preadipocytes (a) and adipocytes (b) were pretreated with or without 10 mM MβCD for 2 h followed by incubation with 100 nM insulin for 72 h and subjected to cell proliferation and viability assay, respectively (mean ± S.E., n = 3). (c) FLAG vector, FLAG-cav-2, or FLAG-3CA-transfected 3T3L1 adipocytes were pretreated with or without 10 mM MβCD for 2 h followed by incubation with or without 100 nM insulin for 10 min and subjected to glucose uptake assay (mean ± S.E., n = 3). NT, nontransfected. (d) Cholesterol depletion by 10 mM MβCD treatment for 2 h in 3T3L1 preadipocytes and adipocytes was measured by a cholesterol assay kit (mean ± S.E., n = 3). (D) (a) 3T3L1 adipocytes were co-transfected with scramble control or cav-1 siRNA and FLAG vector, FLAG-cav-2, or FLAG-3CA followed by incubation with or without 100 nM insulin for 10 min and subjected to glucose uptake assay (mean ± S.E., n = 3). NT, nontransfected. (b) Scramble control or cav-1 siRNA-transfected 3T3L1 adipocytes were subjected to immunoblot analysis with antibodies against cav-1 and actin. The densitometry ratios of cav-1 to actin are illustrated (mean ± S.E., n = 3).

at Cys-109, Cys-122, and Cys-145 in the ER. Myristoylation of cav-2 is a prerequisite for its subsequent palmitoylation. Fatty acylated cav-2 is transported to the PM via Golgi. The PM targeted cav-2 forms homooligomeric cav-2 noncaveolar microdomain in the PM, where cav-2 facilitates IR recruiting and regulates activation of insulin signaling cascade by direct interaction with IR [16]. The C-terminal domain of fatty acylated cav-2 associates with the NPEY motif of insulinactivated IR in the noncaveolar microdomain and insulin-activated IR tyrosine kinase catalyzes Tyr-19 and Tyr-27 phosphorylation of cav-2. The phospho-Tyr-19 and Tyr-27 cav-2 facilitates IRS-1 recruitment to and association with IR, thereby mediating tyrosine phosphorylation of IRS-1 by IR for IRS-1-directed metabolic and mitogenic insulin signaling activation. Many signaling proteins are proposed to use conserved caveolinbinding motifs (CBMs) to associate with caveolae via the caveolin

scaffolding domain (CSD) [39]. It has been reported that the scaffolding domains of cav-1 and cav-3 bind to the CBM (ΦXΦXXXXΦ) in the tyrosine kinase domain of the IRβ-subunit [2,3], and that scaffolding domain peptides derived from cav-1 and cav-3 increase tyrosine kinase activity of the IR [3]. Recent structural and bioinformatic analyses, however, have shown that the CSD located very close to the membrane has limited access to interact with the previously proposed proteins containing the CBM [9,40], and that putative CBMs are buried and inaccessible to the CSD [8,9]. Cav-2 differs from cav-1 and cav-3 in its interaction with IR. As the present data show, cav-2 interacts with IR by binding to the NPEY motif in the juxtamembrane domain of IR but not to the cav-1 and cav-3 binding motif in the IR. IR associates with cav-2 by binding to the C-terminal domain of cav-2, where Cys-109, Cys-122, and Cys-145 are palmitoylated, but not to the cav-2 scaffolding domain. The distinct cav-2 binding to the NPEY motif is essential for the IR

1032

H. Kwon et al. / Biochimica et Biophysica Acta 1853 (2015) 1022–1034

Fig. 7. Regulatory mechanisms of cav-2 in IR signaling. Upon insulin stimulation, newly synthesized myristoylated cav-2 is palmitoylated at Cys-109, Cys-122, and Cys-145 in the ER. The fatty acylated cav-2 is transported to the PM through the Golgi. The PM-targeted fatty acylated cav-2 associates with NPEY motif of insulin-activated IR and is phosphorylated at Tyr-19 and Tyr-27 by IR tyrosine kinase in the noncaveolar microdomain. The phospho-tyrosine cav-2 facilitates IRS-1 association with and tyrosine phosphorylation by IR for IRS-1-directed PI3K-Akt and ERK activation responsible for glucose uptake and cell survival and proliferation.

tyrosine kinase-catalyzed phosphorylation of cav-2 at Tyr-19 and Tyr-27. The present data also show that cav-2 recruits IRS-1 to IR through IRS-1 binding to the phospho-tyrosine cav-2 to promote IRS-1 activation for the IRS-1-directed downstream PI3K-Akt and ERK signaling activation. Thus, our findings show that cav-2 acts as a scaffolding molecule facilitating IRS-1 association with IR and regulates the IR-cav-2-IRS-1-directed insulin signaling activation. Cav-2 among caveolin gene family exhibits the most divergent genomic structure and sequence homology. One of the distinctive features is that cav-2 contains the myristoylation consensus site sequence (MetGly) in the N-terminus [41]. The present data demonstrate for the first time that cav-2 is myristoylated at Gly-2, and that the myristoylation is required for the subsequent palmitoylation of Cys-109, Cys-122, and Cys-145. The fatty acylation is not only essential but also a precondition for its tyrosine phosphorylation by IR to activate the IRS-1-directed downstream insulin signaling. Given that cav-2 myristoylation is a prerequisite for its palmitoylation and that the residues 120-162 in the C-terminal domain are important to interact with IR, myristoylation at Gly-2 and palmitoylation at Cys-122 and Cys-145 seem critical to facilitate its binding to IR and the subsequent tyrosine phosphorylation. The present study, therefore, gives the first assessment of a regulatory function of cav-2 fatty acylation in IR signaling. We have previously demonstrated that cav-2 forms homooligomeric microdomain in the PM for initiation of insulin signaling via IR recruitment, and that ectopic cav-2 rescues the defective IR signaling caused by depletion of caveolae and/or cav-1 down regulation by reconstituting the homooligomeric cav-2 microdomain in the PM [16]. The present data show that failure in the cav-2 fatty acylation results in prevention of the PM targeting of cav-2 and its colocalization with IR in the PM. Fatty acylated cav-2 enhances insulin-induced ERK and Akt activations and regulates cell proliferation and viability. And WT cav-2, but not the fatty acylation deficient mutants, ameliorates the retardation of insulin-stimulated glucose uptake in caveolae- or cav-1-depleted adipocytes. The data and our previous findings, thus, suggest that cav-2 fatty acylation is important for the formation of the homooligomeric cav-2 microdomain in the PM and the direct interaction of cav-2 with IR in the microdomain for the IR-cav-2-IRS-1-directed insulin signaling

activation, distinct from those of the IR-Shc-directed and the caveolae/ cav-1-dependent. The present data demonstrate that fatty acylation at specific site is important for the intracellular localization of cav-2. G2A, 3CA, and C122/ 145A mutants remain in the ER whereas C109/145A and C109/122A mutants translocate to and remain in the Golgi. All of the mutants are unable to localize in the PM. These data show that myristoylation at Gly-2 and palmitoylation at Cys-122 or Cys-145 are important for cav-2 translocation from the ER to the Golgi. Cav-2 myristoylation and palmitoylation at both Cys-122 and Cys-145 are required for its colocalization with IR in the PM. Rab6, but not Rab5 and Rab1, interacts with cav-2 and regulates vesicular transport of cav-2 [15,16]. It will be interesting to test if fatty acylation status of cav-2 affects its association with the small GTPases, thereby regulating intracellular vesicular trafficking of cav-2. Palmitoylation is catalyzed by palmitoyl acyltransferases, also known as Asp-His-His-Cys (DHHC)-domain containing proteins. Mammalian DHHC family consists of 23 members and their specific subcellular localization in the ER, Golgi, and PM was reported [42,43]. Two cytosolic acyl thioesterases (APT1 and APT2) have been identified as palmitoyl thioesterases [44,45] but the APTs that control depalmitoylation of soluble and membrane proteins were not defined. Given that the 3CA mutant is retained in the ER and that insulin triggers palmitoylation of newly synthesized cav-2, the present data indicate that the palmitoylation of newly synthesized cav-2 in the ER is catalyzed within min by the ER resident insulin-specific DHHC and suggest the existence of a novel insulin signaling, regulating palmitoylation of cav-2 in the ER. Having shown that palmitoylated cav-2 is transported to the PM and acts as a substrate for IR tyrosine kinase and that palmitoylated cav-2 undergoes its turnover of depalmitoylation and repalmitoylation within hours, the present data suggest that the turnover of cav-2 are processed by the APT and DHHC in the PM. Further studies are needed to identify the enzymes responsible for and address the regulatory mechanisms controlling palmitoylation and depalmitoylation of cav-2 by specific DHHC and APT, respectively, which localize in a particular intracellular compartment. Cav-2 is widely present and its expression is up-regulated in various tumor cells [46–49]. Micro-ribonucleic acid 199a-3p-targeted cav-2

H. Kwon et al. / Biochimica et Biophysica Acta 1853 (2015) 1022–1034

regulates proliferation and survival of tumor cells [50]. Nevertheless, the functional role of cav-2 is not well-defined in tumor growth and metastasis, and the regulatory mechanism of cav-2 in tumor progression has not been fully investigated. Expression level of cav-2 is changed after high-fat feeding in rat white adipose tissue and skeletal muscle [51–53]. A common single nucleotide polymorphism located in the cav-2 gene (rs2270188) associates with dietary fat to affect risk of type 2 diabetes [54]. Micro-ribonucleic acid 29 (miR-29), which regulates Akt-mediated glucose uptake in adipocytes, is up-regulated in skeletal muscles of type 2 diabetic Goto-Kakizaki rats. Cav-2 is found to be one of the target genes of the miR-29, and cav-2 expression is significantly decreased in skeletal muscle and liver tissue of the diabetic rats [55]. Although these results indicate that cav-2 as a candidate gene involved in the nutritional disorder and insulin resistance related to many metabolic alterations, how cav-2 modulates insulin signaling activation in the metabolic alternations has not been explored. Our present findings provide the molecular mechanisms through which cav-2 manifests insulin signaling activation of the IR-cav-2-IRS1-directed metabolic and mitogenic pathways, and demonstrate the regulatory functions of cav-2 controlled by its expression, fatty acylation, oligomerization, and phosphorylation. Transparency document The Transparency document associated with this article can be found, in online version. Acknowledgments We thank Drs. JY Park, EM Hwang, and J Yoo for sharing reagents. This work was supported by the National Research Foundation of Korea (NRF) grants funded by the Korea government (MSIP: No. 2011-0014898 and ME: No. 2010-0007897) to Y. P., (ME: No. 2012R1A1A2041550) to H. K., and by the Next-Generation BioGreen 21 Program (SSAC: No. PJ009495) from the Rural Development Administration of Korea to H. K. J. L. was a recipient of the Graduate Research Fellowship from Gyeongsang National University (2010-2012) and supported by scholarship from the BK21 Plus Program (ME). D. J. is a recipient of the Global Ph. D. Fellowship (ME: 2013H1A2A1034489). References [1] L. Ramalingam, E. Oh, D.C. Thurmond, Novel roles for insulin receptor (IR) in adipocytes and skeletal muscle cells via new and unexpected substrates, Cell. Mol. Life Sci. 70 (2013) 2815–2834. [2] F.H. Nystrom, H. Chen, L.N. Cong, Y. Li, M.J. Quon, Caveolin-1 interacts with the insulin receptor and can differentially modulate insulin signaling in transfected Cos-7 cells and rat adipose cells, Mol. Endocrinol. 13 (1999) 2013–2024. [3] M. Yamamoto, Y. Toya, C. Schwencke, M.P. Lisanti, M.G. Myers Jr., Y. Ishikawa, Caveolin is an activator of insulin receptor signaling, J. Biol. Chem. 273 (1998) 26962–26968. [4] J. Couet, M. Sargiacomo, M.P. Lisanti, Interaction of a receptor tyrosine kinase, EGF-R, with caveolins. Caveolin binding negatively regulates tyrosine and serine/threonine kinase activities, J. Biol. Chem. 272 (1997) 30429–30438. [5] A. Kimura, S. Mora, S. Shigematsu, J.E. Pessin, A.R. Saltiel, The insulin receptor catalyzes the tyrosine phosphorylation of caveolin-1, J. Biol. Chem. 277 (2002) 30153–30158. [6] C.C. Mastick, A.R. Saltiel, Insulin-stimulated tyrosine phosphorylation of caveolin is specific for the differentiated adipocyte phenotype in 3 T3-L1 cells, J. Biol. Chem. 272 (1997) 20706–20714. [7] J. Oshikawa, K. Otsu, Y. Toya, T. Tsunematsu, R. Hankins, J. Kawabe, S. Minamisawa, S. Umemura, Y. Hagiwara, Y. Ishikawa, Insulin resistance in skeletal muscles of caveolin-3-null mice, Proc. Natl. Acad. Sci. U. S. A. 101 (2004) 12670–12675. [8] D.P. Byrne, C. Dart, D.J. Rigden, Evaluating caveolin interactions: do proteins interact with the caveolin scaffolding domain through a widespread aromatic residue-rich motif? PLoS One 7 (2012) e44879. [9] B.M. Collins, M.J. Davis, J.F. Hancock, R.G. Parton, Structure-based reassessment of the caveolin signaling model: do caveolae regulate signaling through caveolinprotein interactions? Dev. Cell 23 (2012) 11–20. [10] S. Kim, Y. Pak, Caveolin-2 regulation of the cell cycle in response to insulin in Hirc-B fibroblast cells, Biochem. Biophys. Res. Commun. 330 (2005) 88–96.

1033

[11] H. Kwon, K. Jeong, Y. Pak, Identification of pY19-caveolin-2 as a positive regulator of insulin-stimulated actin cytoskeleton-dependent mitogenesis, J. Cell. Mol. Med. 13 (2009) 1549–1564. [12] H. Kwon, K. Jeong, E.M. Hwang, J.Y. Park, S.G. Hong, W.S. Choi, Y. Pak, Caveolin-2 regulation of STAT3 transcriptional activation in response to insulin, Biochim. Biophys. Acta 1793 (2009) 1325–1333. [13] H. Kwon, Y. Pak, Prolonged tyrosine kinase activation of insulin receptor by pY27caveolin-2, Biochem. Biophys. Res. Commun. 391 (2010) 49–55. [14] H. Kwon, K. Jeong, E.M. Hwang, J.Y. Park, Y. Pak, A novel domain of caveolin-2 that controls nuclear targeting: regulation of insulin-specific ERK activation and nuclear translocation by caveolin-2, J. Cell. Mol. Med. 15 (2011) 888–908. [15] K. Jeong, H. Kwon, J. Lee, D. Jang, E.M. Hwang, J.Y. Park, Y. Pak, Rab6-mediated retrograde transport regulates inner nuclear membrane targeting of caveolin-2 in response to insulin, Traffic 13 (2012) 1218–1233. [16] H. Kwon, J. Lee, K. Jeong, D. Jang, Y. Pak, A novel actin cytoskeleton-dependent noncaveolar microdomain composed of homo-oligomeric caveolin-2 for activation of insulin signaling, Biochim. Biophys. Acta 1833 (2013) 2176–2189. [17] M.E. Linder, R.J. Deschenes, Palmitoylation: policing protein stability and traffic, Nat. Rev. Mol. Cell Biol. 8 (2007) 74–84. [18] L. Alland, S.M. Peseckis, R.E. Atherton, L. Berthiaume, M.D. Resh, Dual myristylation and palmitylation of Src family member p59fyn affects subcellular localization, J. Biol. Chem. 269 (1994) 16701–16705. [19] C. Neumann-Giesen, B. Falkenbach, P. Beicht, S. Claasen, G. Lüers, C.A. Stuermer, V. Herzog, R. Tikkanen, Membrane and raft association of reggie-1/flotillin-2: role of myristoylation, palmitoylation and oligomerization and induction of filopodia by overexpression, Biochem. J. 378 (2004) 509–518. [20] P. Zlatkine, B. Mehul, A.I. Magee, Retargeting of cytosolic proteins to the plasma membrane by the Lck protein tyrosine kinase dual acylation motif, J. Cell Sci. 110 (1997) 673–679. [21] D.J. Dietzen, W.R. Hastings, D.M. Lublin, Caveolin is palmitoylated on multiple cysteine residues. Palmitoylation is not necessary for localization of caveolin to caveolae, J. Biol. Chem. 270 (1995) 6838–6842. [22] H. Lee, S.E. Woodman, J.A. Engelman, D. Volonté, F. Galbiati, H.L. Kaufman, D.M. Lublin, M.P. Lisanti, Palmitoylation of caveolin-1 at a single site (Cys-156) controls its coupling to the c-Src tyrosine kinase: targeting of dually acylated molecules (GPI-linked, transmembrane, or cytoplasmic) to caveolae effectively uncouples c-Src and caveolin-1 (TYR-14), J. Biol. Chem. 276 (2001) 35150–35158. [23] Y. Pak, C.R. Paule, Y.D. Bao, L.C. Huang, J. Larner, Insulin stimulates the biosynthesis of chiro-inositol-containing phospholipids in a rat fibroblast line expressing the human insulin receptor, Proc. Natl. Acad. Sci. U. S. A. 90 (1993) 7759–7763. [24] A.K. Student, R.Y. Hsu, M.D. Lane, Induction of fatty acid synthetase synthesis in differentiating 3 T3-L1 preadipocytes, J. Biol. Chem. 255 (1980) 4745–4750. [25] E. Gonzalez, E. Fliera, D. Mollea, D. Accilib, T.E. McGrawa, Hyperinsulinemia leads to uncoupled insulin regulation of the GLUT4 glucose transporter and the FoxO1 transcription factor, Proc. Natl. Acad. Sci. U. S. A. 108 (2011) 10162–10167. [26] O. Karylowski, A. Zeigerer, A. Cohen, T.E. McGraw, GLUT4 is retained by an intracellular cycle of vesicle formation and fusion with endosomes, Mol. Biol. Cell 15 (2004) 870–882. [27] L.B. Tulloch, J. Howie, K.J. Wypijewski, C.R. Wilson, W.G. Bernard, M.J. Shattock, W. Fuller, The inhibitory effect of phospholemman on the sodium pump requires its palmitoylation, J. Biol. Chem. 286 (2011) 36020–36031. [28] M. Nakata, S. Nagasaka, I. Kusaka, H. Matsuoka, S. Ishibashi, T. Yada, Effects of statins on the adipocyte maturation and expression of glucose transporter 4 (SLC2A4): implications in glycaemic control, Diabetologia 49 (2006) 1881–1892. [29] W.S. Nam, K.M. Park, J.W. Park, RNA interference targeting cytosolic NADP(+)dependent isocitrate dehydrogenase exerts anti-obesity effect in vitro and in vivo, Biochim. Biophys. Acta 1822 (2012) 1181–1188. [30] K. Yamada, M. Saito, H. Matsuoka, N. Inagaki, A real-time method of imaging glucose uptake in single, living mammalian cells, Nat. Protoc. 2 (2007) 753–762. [31] H. Lee, D.S. Park, X.B. Wang, P.E. Scherer, P.E. Schwartz, M.P. Lisanti, Src-induced phosphorylation of caveolin-2 on tyrosine 19. Phospho-caveolin-2 (Tyr(P)19) is localized near focal adhesions, remains associated with lipid rafts/caveolae, but no longer forms a high molecular mass hetero-oligomer with caveolin-1, J. Biol. Chem. 277 (2002) 34556–34567. [32] X.B. Wang, H. Lee, F. Capozza, S. Marmon, F. Sotgia, J.W. Brooks, R. Campos-Gonzalez, M.P. Lisanti, Tyrosine phosphorylation of caveolin-2 at residue 27: differences in the spatial and temporal behavior of phospho-Cav-2 (pY19 and pY27), Biochemistry 43 (2004) 13694–13706. [33] H. Cao, W.E. Courchesne, C.C. Mastick, A phosphotyrosine-dependent protein interaction screen reveals a role for phosphorylation of caveolin-1 on tyrosine 14: recruitment of C-terminal Src kinase, J. Biol. Chem. 277 (2002) 8771–8774. [34] K. Takeuchi, Y. Morizane, C. Kamami-Levy, J. Suzuki, M. Kayama, W. Cai, J.W. Miller, D.G. Vavvas, AMP-dependent kinase inhibits oxidative stress-induced caveolin-1 phosphorylation and endocytosis by suppressing the dissociation between c-Abl and Prdx1 proteins in endothelial cells, J. Biol. Chem. 288 (2013) 20581–20591. [35] T.A. Gustafson, W. He, A. Craparo, C.D. Schaub, T.J. O'Neill, Phosphotyrosinedependent interaction of SHC and insulin receptor substrate-1 with the NPEY motif of the insulin receptor via a novel non-SH2 domain, Mol. Cell. Biol. 15 (1995) 2500–2508. [36] Y. Ebina, E. Araki, M. Taira, F. Shimada, M. Mori, C.S. Craik, K. Siddle, S.B. Pierce, R.A. Roth, W.J. Rutter, Replacement of lysine residue 1030 in the putative ATPbinding region of the insulin receptor abolishes insulin- and antibody-stimulated glucose uptake and receptor kinase activity, Proc. Natl. Acad. Sci. U. S. A. 84 (1987) 704–708. [37] F. Zhou, Y. Xue, X. Yao, Y. Xu, CSS-Palm: palmitoylation site prediction with a clustering and scoring strategy (CSS), Bioinformatics 22 (2006) 894–896.

1034

H. Kwon et al. / Biochimica et Biophysica Acta 1853 (2015) 1022–1034

[38] A.W. Cohen, T.P. Combs, P.E. Scherer, M.P. Lisanti, Role of caveolin and caveolae in insulin signaling and diabetes, Am. J. Physiol. Endocrinol. Metab. 285 (2003) E1151–E1160. [39] T. Okamoto, A. Schlegel, P.E. Scherer, M.P. Lisanti, Caveolins, a family of scaffolding proteins for organizing “preassembled signaling complexes” at the plasma membrane, J. Biol. Chem. 273 (1998) 5419–5422. [40] R.G. Parton, M. Hanzal-Bayer, J.F. Hancock, Biogenesis of caveolae: a structural model for caveolin-induced domain formation, J. Cell Sci. 119 (2006) 787–796. [41] T.A. Farazi, G. Waksman, J.I. Gordon, The biology and enzymology of protein N-myristoylation, J. Biol. Chem. 276 (2001) 39501–39504. [42] J. Korycka, A. Łach, E. Heger, D.M. Bogusławska, M. Wolny, M. Toporkiewicz, K. Augoff, J. Korzeniewski, A.F. Sikorski, Human DHHC proteins: a spotlight on the hidden player of palmitoylation, Eur. J. Cell Biol. 91 (2012) 107–117. [43] J. Greaves, H. Chamberlain, DHHC palmitoyl transferases: substrate interaction and (patho)physiology, Trends Biochem. Sci. 36 (2011) 245–253. [44] R. Zeidman, C.S. Jackson, A.I. Magee, Protein acyl thioesterases, Mol. Membr. Biol. 26 (2009) 32–41. [45] E. Kong, S. Peng, G. Chandra, C. Sarkar, Z. Zhang, M.B. Bagh, A.B. Mukherjee, Dynamic palmitoylation links cytosol-membrane shuttling of acyl-protein thioesterase-1 and acyl-protein thioesterase-2 with that of proto-oncogene H-ras product and growthassociated protein-43, J. Biol. Chem. 288 (2013) 9112–9125. [46] A. Fong, E. Garcia, L. Gwynn, M.P. Lisanti, M.J. Fazzari, M. Li, Expression of caveolin-1 and caveolin-2 in urothelial carcinoma of the urinary bladder correlates with tumor grade and squamous differentiation, Am. J. Clin. Pathol. 120 (2003) 93–100. [47] S. Lee, H. Kwon, K. Jeong, Y. Pak, Regulation of cancer cell proliferation by caveolin-2 down-regulation and re-expression, Int. J. Oncol. 38 (2011) 1395–1402.

[48] C. Racine, M. Bélanger, H. Hirabayashi, M. Boucher, J. Chakir, J. Couet, Reduction of caveolin 1 gene expression in lung carcinoma cell lines, Biochem. Biophys. Res. Commun. 255 (1999) 580–586. [49] P.E. Scherer, R.Y. Lewis, D. Volonte, J.A. Engelman, F. Galbiati, J. Couet, D.S. Kohtz, E. van Donselaar, P. Peters, M.P. Lisanti, Cell-type and tissue-specific expression of caveolin-2. Caveolins 1 and 2 co-localize and form a stable hetero-oligomeric complex in vivo, J. Biol. Chem. 272 (1997) 29337–29346. [50] T. Shatseva, D.Y. Lee, Z. Deng, B.B. Yang, MicroRNA miR-199a-3p regulates cell proliferation and survival by targeting caveolin-2, J. Cell Sci. 124 (2011) 2826–2836. [51] A. Gómez-Ruiz, C. de Miguel, J. Campión, J.A. Martínez, F.I. Milagro, Time-dependent regulation of muscle caveolin activation and insulin signalling in response to highfat diet, FEBS Lett. 583 (2009) 3259–3264. [52] I.P. López, F.I. Milagro, A. Martí, M.J. Moreno-Aliaga, J.A. Martínez, C. De Miguel, Gene expression changes in rat white adipose tissue after a high-fat diet determined by differential display, Biochem. Biophys. Res. Commun. 318 (2004) 234–239. [53] I.P. López, F.I. Milagro, A. Marti, M.J. Moreno-Aliaga, J.A. Martinez, C. De Miguel, High-fat feeding period affects gene expression in rat white adipose tissue, Mol. Cell. Biochem. 275 (2005) 109–115. [54] E. Fisher, S. Schreiber, H.G. Joost, H. Boeing, F. Döring, A two-step association study identifies CAV2 rs2270188 single nucleotide polymorphism interaction with fat intake in type 2 diabetes risk, J. Nutr. 141 (2011) 177–181. [55] A. He, L. Zhu, N. Gupta, Y. Chang, F. Fang, Overexpression of micro ribonucleic acid 29, highly up-regulated in diabetic rats, leads to insulin resistance in 3 T3-L1 adipocytes, Mol. Endocrinol. 21 (2007) 2785–2794.