Cellular Signalling 26 (2014) 1355–1368

Contents lists available at ScienceDirect

Cellular Signalling journal homepage: www.elsevier.com/locate/cellsig

A novel insulin receptor-signaling platform and its link to insulin resistance and type 2 diabetes Farah Alghamdi 1, Merry Guo 2, Samar Abdulkhalek 3, Nicola Crawford 4, Schammim Ray Amith 5, Myron R. Szewczuk ⁎ Department of Biomedical & Molecular Sciences, Queen's University, Kingston, Ontario K7L 3N6, Canada

a r t i c l e

i n f o

Article history: Received 8 February 2014 Received in revised form 23 February 2014 Accepted 23 February 2014 Available online 28 February 2014 Keywords: Neuraminidase-1 Matrix metalloproteinase-9 Insulin receptor

a b s t r a c t Insulin-induced insulin receptor (IR) tyrosine kinase activation and insulin cell survival responses have been reported to be under the regulation of a membrane associated mammalian neuraminidase-1 (Neu1). The molecular mechanism(s) behind this process is unknown. Here, we uncover a novel Neu1 and matrix metalloproteinase-9 (MMP-9) cross-talk in alliance with neuromedin B G-protein coupled receptor (GPCR), which is essential for insulin-induced IR activation and cellular signaling. Neu1, MMP-9 and neuromedin B GPCR form a complex with IRβ subunit on the cell surface. Oseltamivir phosphate (Tamiflu®), anti-Neu1 antibodies, broad range MMP inhibitors piperazine and galardin (GM6001), MMP-9 specific inhibitor (MMP-9i), and GPCR neuromedin B specific antagonist BIM-23127 dose-dependently inhibited Neu1 activity associated with insulin stimulated rat hepatoma cells (HTCs) that overly express human IRs (HTC-IR). Tamiflu, anti-Neu1 antibodies and MMP-9i attenuated phosphorylation of IRβ and insulin receptor substrate-1 (IRS1) associated with insulin-stimulated cells. Olanzapine, an antipsychotic agent associated with insulin resistance, induced Neu3 sialidase activity in WG544 or 1140F01 human sialidosis fibroblast cells genetically defective in Neu1. Neu3 antagonist 2-deoxy-2,3-didehydro-N-acetylneuraminic acid (DANA) and anti-Neu3 antibodies inhibited sialidase activity associated with olanzapine treated murine Neu4 knockout macrophage cells. Olanzapine attenuated phosphorylation of IGF-R and IRS1 associated with insulin-stimulated human wild-type fibroblast cells. Our findings identify a novel insulin receptor-signaling platform that is critically essential for insulin-induced IRβ tyrosine kinase activation and cellular signaling. Olanzapine-induced Neu3 sialidase activity attenuated insulin-induced IGF-R and IRS1 phosphorylation contributing to insulin resistance. © 2014 The Authors. Published by Elsevier Inc. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/3.0/).

Abbreviations: IR, insulin receptor; IRS1, insulin receptor substrate-1; IGF-R, insulin growth factor receptor; EGF, epidermal growth factor; MMP, matrix metalloproteinase; RTK, receptor tyrosine kinase; NGF, nerve growth factor; Neu1, mammalian neuraminidase-1; 4-MUNANA, [2′-(4-methylumbelliferyl)-α-D-N-acetylneuraminic acid]; DANA, 2-deoxy-2,3-didehydro-N-acetylneuraminic acid; IC50, the concentration of a drug that is required for 50% inhibition in vitro; TLR, Toll-like receptor; GPCR, G-protein-coupled receptor; MAPK, mitogen-activated protein kinase; IGF-1, insulin like growth factor-1; GRK, G-protein-coupled receptor kinase; PDGF, platelet-derived growth factor; PDGFβR, PDGF β receptor; PI3K, phosphoinositide 3-kinase; EGF, epidermal growth factor; S1P, sphingosine 1-phosphate; Gi, inhibitory G-protein; Gαi, α subunit of Gi; Grb, growth factor receptor binding protein. ⁎ Corresponding author. Tel.: +1 613 533 2457; fax: +1 613 533 6796. E-mail addresses: [email protected] (F. Alghamdi), [email protected] (M. Guo), [email protected] (S. Abdulkhalek), [email protected] (N. Crawford), [email protected] (S.R. Amith), [email protected] (M.R. Szewczuk). 1 Present address: The King Fahd Armed Forces Hospital, Serology, Jeddah, Saudi Arabia. 2 Present address: Faculty of Medicine, University of Ottawa, Ottawa, ON K1N 6N5, Canada. 3 Present address: Department of Molecular Genetics, Cleveland Clinic Foundation Molecular Genetics, Lerner Research Institute, Cleveland, Ohio, U.S.A. 4 Present address: Department of Family Medicine, McMaster University, Hamilton, Ontario, Canada. 5 Present address: Department of Biochemistry, University of Alberta, Edmonton, Alberta, Canada.

1. Introduction The insulin receptor (IR) is a transmembrane tyrosine kinase receptor that is activated by insulin and insulin growth factors-I and II. Metabolically, insulin-induced IR tyrosine kinase activation plays a key role in the regulation of glucose homeostasis. A dysfunctional process of insulin induced IR activation may result in a range of clinical manifestations including insulin resistance [1], type 2 diabetes (T2DM), obesity, cancer, hypertension, and cardiovascular disorders [2–5]. In particular, insulin resistance occurs when the insulin-sensitive tissues, such as skeletal muscle, adipose tissue and liver, do not have the ability to respond to insulin [6,7], and consequently develop several of these chronic diseases [8–10]. The precise mechanism(s) involved in insulin resistance is not well understood [1]. However, several contributing factors have been proposed to contribute to insulin resistance, and they include increased plasma-free fatty acid levels [11], elastin-derived peptides [12], subclinical chronic inflammation [1], oxidative and nitrative stress, altered gene expression, and mitochondrial dysfunction [1,11].

http://dx.doi.org/10.1016/j.cellsig.2014.02.015 0898-6568/© 2014 The Authors. Published by Elsevier Inc. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/3.0/).

1356

F. Alghamdi et al. / Cellular Signalling 26 (2014) 1355–1368

Several studies have separately demonstrated a connection between insulin resistance and cell membrane sialic acid content. Cytidine 5′monophospho-N-acetylneuraminic acid (CMP-NANA; substrate used for the sialyltransferase sialylation of glycans) was reported to enhance insulin responsiveness by 39% providing support for a role for cell surface sialic acid residues in hepatic insulin action with a concomitant contribution to insulin resistance of diabetes mellitus [13]. Structurally, IR is a heavily glycosylated receptor with 18 glycosylation sites [14,15]. Fourteen of the glycosylation sites are localized on the IRα subunits, while the other sites are on the IRβ subunits [16]. Specifically, the IRα subunits contain only the N-linked carbohydrate, while IRβ subunits contain both O- and N-linked carbohydrates [17]. The importance of glycosylation in IR biosynthesis, insulin binding and activation has been studied in detail using site-directed mutagenesis analyses [16,18]. Leconte et al. reported on the potential contribution of N-linked oligosaccharides of the IRβ subunit in the processing, structure and function of the IR [18]. They specifically targeted and mutated IRβ subunits (IRβ-N1234) which were mutagenized on four potential N-glycosylation sites (ASn-X-Ser/Thr). They compared the mutated IRN1234 with the wild type receptor. There were no differences on the molecular weight of IRα subunits between both receptors. However, a reduction in the molecular weight from 95 kDa to 80 kDa of the IRβ subunit was depicted in the mutated IR form. The data indicated that the mutation of the four glycosylation sites in IRβ subunit had no effect on the α-subunit, but caused a reduction in the β-subunit molecular weight. This IRβ-N1234 mutation had no effect on IR cell surface expression and on insulin binding, but attenuated IRβ tyrosine kinase activation by insulin [18]. These early findings provided supporting evidence for a critical role of oligosaccharide side chains of the IRβ subunit in the molecular events responsible for the IR enzymatic activation and signal transduction. Indeed, the role of membrane glycosylation in the function of insulin receptors and glucose transporters is well known for two decades. A neuraminidase was found to release sialic acid from isolated rat adipocytes [19]. The report disclosed that pretreatment of adipocytes with a neuraminidase resulted in an increase in basal glucose transport, the process of which suggested sialic acids playing multiple roles in the control of glucose-transport activity. Recently, the treatment of rat skeletal L6 myoblast cells with either mouse-derived mammalian Neu1 sialidase or Clostridium perfringens neuraminidase resulted in a desialylation of IR, which coincided with a significant increase of L6 myoblast cell proliferative responses to low concentrations of insulin [20]. In addition, the inhibition of endogenous Neu1 abolished this increase in proliferation of L6 cells induced by 1 and 10 nM of insulin, but amplified the proliferative effect of 100 nM insulin. For IGF receptors, the opposite results were observed where the desialylation process of this receptor coincided with an elimination of the heightened proliferative response of L6 myoblasts to 100 nM insulin. To explain this dichotomy, it was proposed that Neu1 enhanced the mitogenic response of L6 myoblasts to low dose of insulin, but the cytosolic Neu2 sialidase [21] and the cell membrane-bound Neu3 sialidase induced myoblast differentiation, but not proliferation, through the direct modulation of the GM3 ganglioside content in myoblasts [22,23]. In support of this hypothesis, it has been reported that a transient upregulation of Neu3 expression in L6 myocytes caused a significant decrease in IR signaling, the mechanism of which was proposed to be in direct modulation of plasma membrane gangliosides by Neu3 activity and the interaction with the growth factor receptorbound protein 2 (Grb2) [24]. Another possibility to explain this dichotomy of insulin-induced mitogenic responses may involve GPCR-receptor tyrosine kinase (RTK) partners forming novel signaling platforms. Using human embryonic kidney 293 cells, Alderton et al. have clearly demonstrated that the platelet-derived growth factor β receptor (PDGFβR) forms a complex with Myc-tagged endothelial differentiation gene-1, a GPCR whose agonist is sphingosine 1-phosphate, in cells co-transfected with these

receptors [25]. PDGF was shown to stimulate tyrosine phosphorylation of the inhibitory Giα subunit to increase p42/p44 mitogen activated protein kinase (MAPK) activation in the regulation of cell proliferation. Furthermore, both GPCR kinase 2 and β-arrestin-1 were found to associate with the PDGFβR, which play critical roles in the regulation of GPCR signal complexes endocytosis, a requirement for the activation of p42/p44 MAPK. The premise is that PDGFβR signaling is initiated by GPCR kinase 2/β-arrestin-1 complexes that have been recruited to the PDGFβR via tethering to a GPCR(s). These results provide a novel platform for the integrative signaling by these receptors that may account for the co-mitogenic effect of certain GPCR agonists with PDGF on cell proliferation. It is noteworthy that the basally tyrosinephosphorylated Giα subunits do not induce activation of p42/p44 MAPK on their own [25]. The potentiating effects of Giα signaling on the PDGF-stimulated p42/p44 MAPK activation may require the PDGFinduced recruitment to tyrosine-phosphorylated Giα subunits of other intermediates. Indeed, several reports have shown that insulin receptors can interact with Giα subunit [26–28]. The integration of GPCR-RTK partners forming unique signaling platforms in which protein components specific for each receptor are shared to produce a response upon engagement of ligands is eloquently reviewed by Pyne and colleagues [29–32] and Abdulkhalek et al. [33]. Recently, insulin binding to its receptor was reported to rapidly induce the interaction of the insulin receptor with Neu1 sialidase, the activity of which hydrolyzes sialic acid residues of IR and, consequently, induces receptor activation [34]. The report also disclosed that Neu1 deficient mice with ∼ 10% of the normal Neu1 activity exposed to a high-fat diet developed hyperglycemia and insulin resistance twice as fast as the wild-type cohort. They also proposed that endogenous Neu1 sialidase activity is involved in the insulin receptor glycosylation modification. Indeed, Blaise et al. provided additional confirmation to support that Neu1 interacts with IRβ and desialylates the receptor [12]. Other reports have suggested that receptor glycosylation modification may in fact be the connecting link between ligandbinding, receptor dimerization and activation for several other receptors [35–39]. This present report describes a novel organizational IR signaling platform describing the intermediates linked to the insulin-induced receptor activation process in live IR-expressing cells. Here, insulin binding to its receptor induces a Neu1 and matrix metalloproteinase-9 (MMP-9) cross-talk in activating IRs. Central to this process is that Neu1 and MMP-9 cross-talk in alliance with neuromedin B GPCR forms a complex tethered at the ectodomain of IRβ subunits on the cell surface. This signaling paradigm proposes that insulin binding to its receptor on the cell surface induces a conformational change of the receptor to initiate GPCR Giα-signaling and MMP-9 activation to induce Neu1. Activated Neu1 hydrolyzes α-2,3-sialyl residues linked to β-galactosides, which are distant from the insulin binding sites. These findings are consistent with another report [34] that predicts a prerequisite desialylation process by activated Neu1 enabling the removal of steric hindrance to IRβ subunit association and the activation of tyrosine kinases. This Neu1-MMP-9 crosstalk in alliance with neuromedin B GPCR at the ectodomain of IRβ subunits forms the essential signaling platform on the cell surface that is critical for insulininduced receptor activation. Olanzapine, an antipsychotic agent associated with insulin resistance [40–42], induced Neu3 sialidase activity in Neu1-deficient human WG544 or 1140F01 sialidosis fibroblast cells and Neu4 knockout primary murine macrophage cells. Olanzapine-induced Neu3 activity attenuated IRβ and IRS1 phosphorylation associated with insulin-stimulated human fibroblast cells. This study provides confirmation additional evidence that Neu1 sialidase is involved in the regulation of insulin signaling and that a down-regulation Neu1 by Neu3 activity may lead to an impaired insulin receptor signaling and subsequently contributing to insulin-resistant diabetes.

F. Alghamdi et al. / Cellular Signalling 26 (2014) 1355–1368

2. Materials and methods 2.1. Cell lines HTC-WT cell line is the wild-type rat hepatoma cell line and HTC-IR cells, a rat hepatoma cell line that overly expresses the human insulin receptors [43] (kindly provided by Dr. Leda Raptis, of this department). The cell lines were grown in 1 × Dulbecco's Modified Eagle Medium (DMEM, Gibco, Rockville, MD) containing 10% fetal calf serum (FCS) (HyClone, Logan, Utah, USA) and 5 μg/mL plasmocin, and were maintained at 5% CO2 and 37 °C. For HTC-IR cell line, the medium was supplemented with 400 μg/mL of G418 as selection marker for IR expression. WG544 or 1140F01 sialidosis fibroblast cell line is a Neu1 deficient skin fibroblast cells that were isolated from patients with lysosomal storage disorder type 1 sialidosis and immortalized [44]. Wild-type human fibroblast cell line is a normal skin fibroblast cells. Primary bone marrow (BM) macrophage cells were obtained Neu4 knockout mice [45]. These cells were kindly provided by Dr. Alexey Pshezhetsky, Department of Pediatrics and Biochemistry, Montreal University, Service de Genetique, Ste-Justine Hospital, 3175 Cote-Ste-Catherine, H3T1C5, Montreal, QC, Canada. Cells were grown in 1× DMEM medium containing 10% FCS and 5 μg/mL plasmocin. The primary BM macrophage cells were further supplemented with 20% (v/v) of L929 cell supernatant as a source of monocyte colony-stimulating factor (M-CSF). 2.2. Primary mouse bone marrow macrophage cells Bone marrow (BM) cells were flushed from femurs and tibias of mice with sterile Tris-buffered saline (TBS) solution. The cell suspension was centrifuged for 3 min at 900 rpm, and the cell pellet resuspended in red cell lysis buffer for 5 min. The remaining cells were washed once with sterile TBS, and then resuspended in RPMI conditioned medium supplemented with 10% FBS and 20% (v/v) of L929 cell supernatant as a source of monocyte colony-stimulating factor (M-CSF) according to Alatery and Basta [46] and 1 × L-glutamine-penicillin–streptomycin (Sigma-Aldrich Canada Ltd., Oakville, Ontario) sterile solution. The primary BM macrophages were grown on 12 mm circular glass slides in RPMI conditioned medium for 7–8 days in a humified incubator at 37 °C and 5% CO2. Primary BM macrophage cells by day 7 are more than 95% positive for macrophage marker F4/80 molecule as detected by flow cytometry. 2.3. Reagents The active substance in Novolin®ge (Insulin, Human Biosynthetic) is a polypeptide that is structurally identical to natural human insulin. The insulin is produced by recombinant DNA technology in Saccharomyces cerevisiae. Human insulin (Novolin®ge Toronto) is an injectable solution, 3.5 mg (100 IU) of the natural ligand of the insulin receptors, and stored at 4 °C. The injectable insulin solution was used in our experiments at a concentration range of 10–100 nM. This concentration was determined to produce optimal insulin-induced sialidase activity based on the results of our live cell sialidase assay. Incubation times vary between experiments and are indicated. 2-(4-Methylumbelliferyl)-α-D-N-acetylneuraminic acid (4-MUNANA; Biosynth Intl.) is a sialidase substrate that was used at a concentration of 0.318 mM in the sialidase experiments. Olanzapine (OLZ) is an antipsychotic drug sold under the brand name, Zyprexa (Eli Lilly Canada Inc., Toronto, ON, M1N 2E8). 2.4. Inhibitors Oseltamivir phosphate, Tamiflu (99% pure oseltamivir phosphate, Hoffmann-La Roche Ltd.), a broad-range sialidase inhibitor was used at

1357

200–300 μg/mL unless otherwise indicated. DANA (2-deoxy-2,3dehydro-D-N-acetylneuraminic acid) (Sigma-Aldrich) were used at indicated concentrations. Pertussis toxin (PTX, from Bordetella pertussis, in buffered aqueous glycerol solution, Sigma-Aldrich) catalyzes the ADPribosylation of the α subunits of the heterotrimeric G proteins Gi, Go, and Gt. This prevents the G protein heterotrimers from interacting with receptors, thus blocking their coupling and activation. Galardin (GM6001; Calbiochem-EMD Chemicals Inc., Darmstadt, Germany) is a potent, cell-permeable, broad-spectrum hydroxamic acid inhibitor of matrix metalloproteinases (MMPs) (IC 50 = 400 pM for MMP1; IC50 = 500 pM for MMP2; IC50 = 27 nM for MMP3; IC50 = 100 pM for MMP8; and IC50 = 200 pM for MMP9). Galardin was used at a concentration of 12.5 ng/mL to 125 μg/mL for the indicated incubations times. Piperazine (PIPZ; MMPII inhibitor; Calbiochem-EMD Chemicals Inc.) is a potent, reversible, broad range inhibitor of matrix metalloproteinases. PIPZ inhibits MMP1 (IC50 = 24 nM), MMP3 (IC50 = 18.4 nM), MMP7 (IC50 = 30 nM), and MMP9 (IC50 = 2.7 nM). PIPZ was used at a concentration of 12.5 ng/mL to 125 μg/mL for the indicated incubations times. MMP-9-inhibitor (MMP-9i; Calbiochem) is a specific and reversible inhibitor of MMP-9. MMP-9i was used at concentrations of 50 μg/mL or 100 μg/mL. MMP-3-inhibitor (MMP-3i; Calbiochem) is a specific and reversible inhibitor of MMP-3. MMP-3i was used at concentrations of 125 ng/mL to 125 μg/mL. Cyclolignan picropodophyllin (PPP; Calbiochem) was used to discriminate between insulin receptor phosphorylation and insulin-like growth factor-1 receptor phosphorylation by inhibiting IGF-1. PPP is a non-competitive inhibitor that was used at a concentration of 20 nM. BIM-23127 is a specific neuromedin B receptor inhibitor from Tocris Bioscience (Tocris House, IO Centre Moorend Farm Avenue, Bristol, BS11 0QL, United Kingdom).

2.5. Primary antibodies The expression of insulin receptors was determined by using antibody specific for IRβ (C-19; Santa Cruz Biotechnology). IRβ antibody is a rabbit anti-human antibody that is raised against a peptide mapping at the C-terminus of IRβ of human origin. Insulin receptor activation and tyrosine kinase phosphorylation was determined by using two specific antibodies. The anti-phosphorylated insulin receptor β (p-IRβ; MBL International, Woburn, MA 01801) antibody is a rabbit anti-human antibody that is raised against a chemically synthesized phospho-peptide derived from the region of the human insulin receptor that contains tyrosine 972. The anti-phosphorylated insulin receptor substrate-1 (p-IRS-1; Cell Signaling Technology, Inc.) is a rabbit anti-human antibody specific for phosphorylated insulin receptor substrate-1 (p-IRS). In western blot experiments, horse radish peroxidase (HRP) conjugated goat anti-rabbit IgG antibody (Sigma Aldrich) was used as a secondary antibody. In the co-immunoprecipitation experiments, CleanBlot IP Detection Reagent (Pierce, Cell Signaling Technology, Inc.) was used as a secondary antibody. CleanBlot reagent is a conjugated HRP antibody that has the ability to bind to the native antibody and not the denatured antibody fragments from the immunoprecipitation processing step. In the immunocytochemistry and the co-localization experiments, donkey anti-rabbit AlexaFluor 488 or donkey anti-rabbit AlexaFluor 594 antibodies (Molecular Probes, Life Technologies, Carlsbad, CA, 92008. United States) were used as secondary antibodies. Four neutralizing antibodies were used to inhibit sialidase activity in live HTC-IR and primary murine macrophage cell lines. The neutralizing antibodies are rabbit-anti-human Neu1 IgG antibody (Santa Cruz Biotechnology, Inc.), mouse anti-human Neu2 IgG antibody (Santa Cruz), mouse anti-human Neu3 IgG antibody (MBL) and rabbit anti-human Neu4 IgG antibody (Proteintech Group, Inc., Chicago, IL 60612, USA). Goat anti-human MMP9 IgG antibody (Santa Cruz) was used in order to identify the role of MMP9 in insulin receptor activation and its effect on the phosphorylation of the receptor.

1358

F. Alghamdi et al. / Cellular Signalling 26 (2014) 1355–1368

2.6. Sialidase assay in live cells Cells were grown overnight on 12 mm circular glass slides in conditioned medium in a sterile 24-well tissue plate until they reached ~70% confluence as previously optimized in the live cell sialidase assay [47,48]. After removing medium, 0.318 mM 4-MUNANA (2′-(4-methylumbelliferyl)-α-D-N-acetylneuraminic acid; Biosynth Intl.) substrate in Tris-buffered saline (TBS) pH 7.4 was added to each well alone (control), with predetermined dose of insulin, or in combination of insulin and inhibitor or neutralizing antibodies at the indicated concentrations. The stimulated group was treated with insulin at indicated concentrations (10–100 nM) in the presence or absence of PPP (IGF-1R inhibitor). The inhibited group was pretreated with different inhibitors (oseltamivir phosphate, anti-Neu-1, -2, -3, and -4 neutralizing antibodies, piperazine, galardin, MMP9 inhibitor and MMP3 inhibitor) at indicated concentrations followed with insulin stimulation. The sialidase substrate 4-MUNANA at 0.318 mM was added to the cells in order to detect sialidase activity. The substrate is hydrolyzed by sialidase to give free 4-methylumbelliferone, which has a fluorescence emission at 450 nm (blue color) following an excitation at 365 nm. Fluorescent images were taken after 2–3 min using epi-fluorescent microscopy (Zeiss Imager M2, ×40 objective). Sialidase activity of live HTC-IR cells was indicated by the blue fluorescence surrounding the periphery of the cells. The mean fluorescence was calculated using Image J software. 2.7. Immunocytochemistry for phosphorylation of IRβ in insulin-treated HTC-IR cells HTC-IR cells were cultured on 12 mm glass coverslips in a sterile 24 well plate in conditional medium as previously described. After cells reached about 70% confluence, cells were serum starved for 4 h. Cells were either control non-treated cells or secondary antibody only control, or pretreated cells. Pretreated cells were either stimulated by (100 nM) insulin accompanied with PPP for 5 min, or treated with oseltamivir phosphate (350 μg/mL) for 30 min followed by insulin (10-100 nM). Control cells were neither inhibited nor stimulated. Cells were fixed with 4 μg/mL paraformaldehyde for 30 min and permeabilized with 0.2% Triton-X for 5 min. Cells were blocked with 4% bovine serum albumin (BSA) in 0.1% tween-TBS for 40 min on ice. The fixed cells were immunostained with 4 μg/mL rabbit anti-human pIRβ antibody for 1 h at 37 °C, followed with AlexaFluor 488 conjugated donkey anti-rabbit IgG antibody for 1 h at 37 °C. In order to detect the non-specific fluorescence background, control cells were incubated with secondary antibody only. Stained cells were viewed by fluorescence microscopy (40 × magnification). The density of the cell staining (green fluorescence) was measured by Corel Photo Paint 8.0 software. 2.8. Preparation of cell lysates HTC-IR cells were cultured in 75 cm2 flasks at 37 °C until confluent. Cells were treated with insulin, pretreated with different inhibitors followed with insulin, or left untreated as control cells. Cells were removed and re-suspend in 100 μL of RIPA lysis buffer cocktail which contains 1 × RIPA buffer, 1 mM PMSF, 0.2 mg/mL leuptin, 1% β mercaptoethanol and 1 μL protease inhibitor cocktail on ice for 30 min. Cell suspension was then transferred into other Eppendorf tubes and stored at −80 °C. 2.9. Western blot to detect insulin receptor β (IRβ) and human phosphorylated insulin receptor substrate-1 (pIRS-1) Cell lysates were prepared from the following cell lines: HTC-IR, HTC-WT and MiaPaCa-2. Cells (1 × 107 cells) were pelleted and lysed in lysis buffer (50 mM Tris, pH 8, 150 mM NaCl, 1% Nonidet P-40, 0.2 mg/mL leupeptin, 1% β-mercaptoethanol, and 1 mM PMSF). The protein samples were denatured in the lysis buffer at 95–100 °C for

8 min. To ensure the best resolution, samples were mixed before and after heating. Proteins in the cell lysates were resolved by 8% Bis–Tris SDS-PAGE gel. The gel running time was 90 min at 150 V. Semi dry transfer technique was performed to transfer the resolved proteins on PVDF transfer membrane blot. Transfer time was 75 min at 100 mA. The membrane blot was blocked for 2 h at room temperature or overnight at 4 °C using 5% BSA or fat-free skim milk in 0.1% Tween-Tris buffered saline (TBS) to block non-specific background binding. The blot was probed with rabbit anti-insulin receptor β (IRβ) antibody or rabbit anti-human phosphorylated insulin receptor substrate-1 (pIRS-1) antibody (Cell Signaling) as primary antibodies overnight at 4 °C. After incubating, the blot was washed three times followed with horse radish peroxidase (HRP) conjugated secondary goat anti-rabbit IgG antibody for 60 min at room temperature and Western Lightning Chemiluminescence Reagent Plus for 5 min. The chemiluminescence reaction was analyzed with X-ray film. The blot was washed, stripped and re-probed with anti-β-actin antibody (Cell Signaling) as an internal control protein for loading of the cell lysate. Quantitative analysis was done by assessing the density of a band corrected for background in each lane using Corel Photo Paint 8.0 software. Each bar in the graphs represents the mean ratio of IRβ to β-actin of band density ± S.E. (error bars) for 5–10 replicate measurements. 2.10. Co-immunoprecipitation HTC-IR cells were left cultured in medium or in medium containing 100 nM insulin for the indicated time intervals. Cells (1 × 107 cells) were pelleted and lysed in lysis buffer (50 mM Tris, pH 8, 150 mM NaCl, 1% Nonidet P-40, 0.2 mg/mL leupeptin, 1% β-mercaptoethanol, and 1 mM PMSF). For immunoprecipitation, Neu-1, -2 and -4, MMP9 and insulin receptor in cell lysates from HTC-IR cells using 100 μg of cell lysates were immunoprecipitated with either 1 μg of rabbit-antihuman Neu1 antibody, mouse anti-human Neu2 antibody, rabbit antihuman Neu4 antibody or goat anti-human MMP9 antibody overnight at 4 °C. Following immunoprecipitation, complexes were isolated using protein A or G magnetic beads for 90 min, washed three times in buffer (10 mM Tris, pH 8, 1 mM EDTA, 1 mM EGTA, 150 mM NaCl, 1% Triton X-100, and 0.2 mM sodium orthovanadate) and resolved by 8% Bis–Tris gel electrophoresis (SDS-PAGE). The gel's running time was 90 min at 150 V. The proteins were transferred onto PVDF transfer membrane blot via semi-dry blotting technique for 75 min at 100 mA. The blot was blocked with 5% BSA in 0.1% Tween-TBS as blocking buffer for 2 h at room temperature. The blot was incubated with rabbit antihuman IRβ antibody overnight at 4 °C followed with Clean-Blot IP Detection Reagent for immunoprecipitation/western blots (Pierce, Thermo Fisher Scientific) for 60 min at room temperature and Western Lightning Chemiluminescence Reagent Plus. The chemiluminescence reaction was analyzed with X-ray film. HTC-IR cell lysates that were not incubated with the IRβ antibody prior to the immunoprecipitation step were used as a negative control in the co-IP experiments. Blot was washed, stripped and re-probed with anti-insulin receptor antibody in order to determine equal loading. 2.11. Neu1 and MMP9 co-localization with insulin receptor β (IRβ) HTC-IR cells were cultured on 12 mm circular glass coverslips in a sterile 24 well plate in conditional medium as described. After cells reached about 70% confluence, cells were serum starved for 4 h. Cells were fixed with 4 μg/mL paraformaldehyde, permeabilized with Triton-X, and blocked with 4% BSA in 0.1% Tween-TBS for 40 min on ice. Cells were incubated with 4 μg/mL of one of the following combinations of antibodies (IRβ antibody and Neu1 antibody; IRβ antibody and Neu2 antibody; IRβ antibody and Neu3 antibody; IRβ antibody and Neu4 antibody; or IRβ antibody and MMP9 antibody) for 60 min at 37 °C. Co-localization procedure is dependent on the primary antibody combination with respect to animal species. Cells were washed and

F. Alghamdi et al. / Cellular Signalling 26 (2014) 1355–1368

incubated with 4 μg/mL of AlexaFluor 594 and AlexaFluor 488 conjugated secondary antibodies for 60 min at 37 °C. To account for background non-specific fluorescence, control cells with only AlexaFluor 488 or AlexaFluor 594 conjugated secondary antibody were used. Stained cells were mounted on microscope slide using 3 μL of mounting medium (DAKO). Stained cells were viewed using Zeiss M2 epi-fluorescent microscopy (40 × objective). Images were taken under two different channels; red channel and green channel. Overlay and co-localization visualization was done by using Adobe Photoshop software and Image J software. To calculate the amount of colocalization in the selected images, the Pearson correlation coefficient was measured and expressed as a percentage using Image J version 1.38x software. 2.12. Statistical analysis Statistical analysis was performed by using GraphPad Prism 5.0. Comparisons between two groups were made by one-way analysis of variance at 95% confidence using Student's unpaired t test and Bonferroni's multiple comparison test or Dunnett's multiple comparison test for comparisons among more than two groups. 3. Results 3.1. Neu1 sialidase activity is associated with insulin stimulation of insulin receptor (IR)-expressing cells Insulin binding to its receptor has been reported to rapidly induce the interaction of the IR with Neu1 [34]. Activated Neu1 hydrolyzes sialic acid residues of the receptor and, consequently, induces IR activation. The report also disclosed that Neu1 deficient mice with ∼10% of the normal Neu1 activity fed a high-fat diet rapidly developed hyperglycemia and insulin resistance. Collectively, this report and other reports [19,20] suggest that insulin receptor glycosylation modification may be the link connecting insulin-binding and receptor activation, but the mechanism(s) behind this linkage is unknown. Recently, we reported that epidermal growth factor (EGF) binding to EGF receptors potentiates neuromedin B GPCR-signaling and MMP-9 activation to induce Neu1 activity in live EGFR-expressing cell lines [39]. A striking similarity was identified between this novel receptor signaling platform for nerve growth factor (NGF) TrkA receptors [49], cell surface Toll-like receptor (TLR)-4 [48,50,51], and intracellular TLR-7 and TLR-9 receptors [52]. Since IRs belong to the same family of receptor tyrosine kinases like EGFR and TrkA, we initially asked whether insulin-induced IR activation follows this similar receptor signaling paradigm. Using a rat hepatoma cell line overly expressing human IR (HTC-IR), we performed a sialidase assay on the live cells as previously described [48,53,54]. As shown in Fig. 1A, insulin stimulation of live HTC-IR cells dose-dependently induced sialidase activity. This sialidase activity is revealed in the periphery surrounding the cells using a fluorogenic sialidase specific substrate, 4-MUNANA [2′-(4-methylumbelliferyl)-αD-N-acetylneuraminic acid], which fluoresces at 450 nm and caused by the emission of 4-methylumbelliferone. The mean fluorescence of 50 multi-point replicates was quantified using the Image J software and is depicted in the bar graph. We also used specific neutralizing antibodies against the four mammalian sialidases [55], known as lysosomal Neu1 [56–62], cytosolic Neu2, the plasma membrane bound Neu3 [24,63,64] and the fourth sialidase, Neu4, localized to either the mitochondrial [65] compartment or the lysosomal lumen [66]. As shown in Fig. 1B, anti-human Neu1 antibody blocked the sialidase activity associated with insulin-treated live HTC-IR cells comparable to the levels of no insulin treated controls. The anti-Neu1 antibody used here is specific for the epitope corresponding to amino acids 116–415 mapping at the Cterminus of Neu1 of human origin. It also detects Neu1 of mouse, rat and human origin. In contrast, antibodies against the other three human sialidases, Neu-2, -3 and -4 had no blocking effect on sialidase activity associated with insulin stimulated live HTC-IR cells similar to

1359

the mean fluorescence levels of insulin positive control (Fig. 1B). These latter results using neutralizing antibodies are consistent with our previous report with Toll-like receptors [54]. It was not surprising that this insulin-induced sialidase activity was completely blocked by the neuraminidase inhibitor Tamiflu (pure oseltamivir phosphate) at 200 μM as shown in Fig. 1C but also in a dose dependent range of 0.25–250 μg/mL (Fig. 1C). To further elucidate the inhibitory capacity of Tamiflu, the 50% inhibitory concentration (IC50) of the compound was determined by plotting the decrease in sialidase activity against the log of the agent concentration. As shown in Fig. 1C, Tamiflu had an IC50 of 60 μM which is slightly higher than the reported IC50 of 3.876 μM for NGF-TrkA [49] and 1.175 μM for LPS-TLR-4 [54] ligandinduced sialidase activity in TrkA-PC12 and BMC-2 macrophage cells, respectively. For NGF-induced sialidase activity in TrkA-expressing cells, we also reported that other purified neuraminidase inhibitors such as zanamivir (4-guanidino-Neu5Ac2en) and oseltamivir carboxylate had a limited inhibition of NGF-induced sialidase activity in live TrkA-PC12 cells at 1–2 mM compared to the NGF positive control [49]. Other studies using recombinant soluble human sialidases have shown that oseltamivir carboxylate, the active metabolite of Tamiflu, scarcely inhibited the activities of the four human sialidases even at 1 mM [67], while zanamivir significantly inhibited the human Neu2 and Neu3 sialidases in the micromolar range. Zanamivir and DANA (2deoxy-2,3-didehydro-N-acetylneuraminic acid) were found to inhibit endogenous sialidase activity of activated lymphocytes grown in culture, as evidenced by an altered sialylation pattern of cell surface proteins, and that this inhibition of sialidase activity resulted in a reduced production of IFN-γ mRNA and protein in the 1–2 mM range [58]. To confirm Neu1 sialidase activity associated with insulin stimulated live cells at the genetic level, we used human WG544 or 1140F01 type 1 sialidosis fibroblast cell lines with the live cell sialidase assay. Patients with type 1 sialidosis or mucolipidosis-1 have a Neu1 genetic deficiency which is a human autosomal recessive inborn error of metabolic disease. Patients with sialidosis have a build-up of sialylated glycoconjugates in tissues and urine causing neuropathic or nonneuropathic symptoms depending on the onset of the disease. The data in Fig. 1D clearly indicate a lack of sialidase activity associated with insulin stimulation of these live Neu1-deficient cells compared to the wild-type human fibroblast cells. These latter data provide additional supporting evidence for Neu1 involvement with insulin activation of live IR expressing cells. It is noteworthy that the fluorescence associated with ligand treated cells as seen in Fig. 1 is not due to a form of secreted or shed sialidase from the cells as previously described [53,54]. The live cells treated with the substrate alone do not show fluorescence surrounding the cells, consistent with our other reports [49,54]. Taken together, these results suggest that the diffuse fluorescence associated with insulin treated live cells is due to an activation of Neu1 on the cell surface. 3.2. Insulin-induced phosphorylation of insulin growth factor-1 receptor (IGF-1R), insulin receptor substrate-1 (IRS-1) and insulin receptor β (IRβ) is dependent on Neu1 sialidase activity If insulin-induced receptor phosphorylation is dependent on Neu1 activity, then neuraminidase inhibitor like Tamiflu or anti-Neu1 neutralizing antibody should have an inhibitory effect on insulin-induced phosphorylation of IGF-1R, IRS-1 and IRβ in IR-expressing cells. Immunocytochemistry analyses shown in Fig. 2A demonstrate that Tamiflu, anti-Neu1 antibody and Gαi protein inhibitor pertussis toxin (PTX) significantly inhibited insulin-induced pIGF-1R and pIRS-1 in wildtype human fibroblast cells compared to the insulin treated controls. Western blot analyses confirmed that Tamiflu and anti-Neu1 antibodies markedly inhibited insulin-induced pIRS-1 in HTC-IR cells compared to the insulin treated controls (Fig. 2B). To test the inhibitory effect of Tamiflu on insulin-induced phosphorylation of IRβ, HTC-IR cells were left untreated as controls, and either stimulated with 100 nM insulin

1360

F. Alghamdi et al. / Cellular Signalling 26 (2014) 1355–1368

Fig. 1. (A) Insulin induces sialidase activity in live HTC-IR cells in a dose-dependent manner. HTC-IR cells were allowed to adhere on 12 mm circular glass slides in media containing 10% fetal calf sera for 24 h. After reaching ~75% confluence, cells were serum starved for 6 h. After removing media, 0.318 mM 4-MUNANA (4-MU) substrate (2′-(4-methlyumbelliferyl)-α-Nacetylneuraminic acid) in Tris buffered saline pH7.4 was added to live cells alone (control) or with insulin at the indicated doses together with 50 nM PPP. The substrate is hydrolyzed by sialidase enzymes to give free 4-methylumbelliferone, which has a fluorescence emission at 450 nm (blue color) following excitation at 365 nm. Fluorescent images were taken at 2 min after adding substrate using epi-fluorescent microscopy (40× objective). The mean fluorescence of 50 multi-point replicates was quantified using the Image J software. Mean ± error bars, S.E. The data are a representation of one out of three independent experiments showing similar results. (B) Insulin-induced sialidase activity in live HTC-IR cells is blocked by anti-Neu1 and not with anti-Neu-2, -3, or -4 neutralizing antibodies. Insulin-induced sialidase activity in live HTC-IR cells was measured as described in (A). Cells were either treated with 10 nM insulin plus 50 nM PPP alone, in combination with 100 μg/mL of anti-Neu-1, -2, -3, or -4 neutralizing antibodies or left untreated as control cells. The mean fluorescence of 50 multi-point replicates was quantified using the Image J software. Mean ± error bars, S.E. The data are a representation of one out of three independent experiments showing similar results. Significant differences at 95% confidence using the Dunnett's multiple comparison test was used to compare inhibition with the insulin control in each group for n = 50 replicates. (C) Tamiflu inhibits insulin-induced sialidase activity in HTC-IR cells in a dose dependent manner. Insulin-induced sialidase activity in live cells was measured as described in (A). Fluorescent images were taken at 2 min after adding 0.318 mM 4-MU substrate together with insulin plus 50 nM PPP and indicated concentrations of Tamiflu using epi-fluorescent microscopy (40× objective). The mean fluorescence surrounding the cells for each of the images was measured using Image J Software. Mean ± error bars, S.E. The data are a representation of one out of three independent experiments showing similar results. Significant differences at 95% confidence using the Dunnett's multiple comparison test was used to compare inhibition with the insulin control in each group for n = 50 replicates. (D) Insulin induces sialidase activity in live wild-type human fibroblast cells but not in WG544 or 1140F01 Neu1-deficient fibroblast cells derived from type 1 sialidosis patients. Insulin-induced sialidase activity in live human fibroblast cells was measured as described in (A). Cells were treated with 100 nM insulin or left untreated as control cells. The data are a representation of one out of four independent experiments showing similar results.

together with 50 nM cyclolignan picropodophyllin (PPP), or inhibited using 350 μg/mL Tamiflu prior to the stimulation with 100 nM of insulin. PPP is a non-competitive inhibitor and was used to discriminate between insulin receptor and IGF-1R by inhibiting phosphorylation of IGF-1R tyrosine kinase. Immunocytochemistry analyses demonstrated that 100 nM of insulin was able to induce the phosphorylation of IRβ in HTC-IR cells. Tamiflu pretreated cells at 350 μg/mL for 35 min followed by 100 nM insulin stimulation for 5 min together with 50 nM PPP was able to significantly inhibit insulin-induced pIRβ activation in HTC-IR cells (Fig. 2C).

3.3. Neu1 is tethered to naive and insulin-stimulated IR receptors According to the glycosylation model of IR, fourteen of the 18 glycosylation sites are localized on the IRα subunits, while the other sites are on the IRβ subunits [16]. The IRα subunits contain only the N-linked carbohydrate, while IRβ subunits contain both O- and N-linked carbohydrates [17]. Studies on mutational analyses of the four glycosylation sites in IRβ subunit found no effects on the expression of IR on the cell surface and did not interfere with insulin binding to the IRα subunit, but the IRβ tyrosine kinase activation by insulin was attenuated [18].

F. Alghamdi et al. / Cellular Signalling 26 (2014) 1355–1368

1361

Fig. 2. (A) Tamiflu, anti-Neu1 neutralizing antibody and pertussis toxin (PTX) inhibit insulin induced phosphorylation of IGF-1R and IRS-1 in wild-type human fibroblast cells. Cells were grown overnight on glass coverslips in a 24-well tissue culture plate at 37 °C for 24 h or until they reached ~70% confluence. Cells were stimulated with 100 nM insulin for 5 min, pretreated with either 400 μM Tamiflu, 100 ng/mL pertussis toxin (PTX), 1.4 μg/mL anti-Neu1 neutralizing antibodies for 30 min followed with 100 nM insulin for 5 min or were left untreated as no ligand controls. Cells were fixed, permeabilized, blocked with 4% bovine serum albumin (BSA) in 0.1% Tween-Tris buffered saline (TBS) for 20 min on ice and immunostained with rabbit anti-phospho-IGF-1R or rabbit anti-phospho-1RS-1 for 60 min at 37 °C, followed with AlexaFluor 594 goat anti-rabbit secondary antibody for 60 min at 37 °C. Control group had only secondary antibodies with no other treatment. Stained cells were visualized by epi-fluorescence microscopy with a 40× objective. Quantitative analysis was done by assessing the density of cell staining corrected for background in each panel using Corel Photo Paint 8.0 software. Each bar in the figure represents the mean corrected density of culture cell staining ± SEM for equal cell density (5 × 105 cells) within the respective images. The data are a representation of one out of three independent experiments showing similar results. Significant differences at 95% confidence using the Dunnett's multiple comparison test was used to compare inhibition with the insulin control in each group for n = 10–20 stained cells. (B) Tamiflu inhibits insulin induced phosphorylation of IRβ (pIRβ) in HTC-IR cells. Cells were cultured on 12 mm glass coverslips in a sterile 24 well plate in conditioned medium. Cells were stimulated with 100 nM insulin plus 50 nM PPP for 5 min, inhibited with 350 μg/mL Tamiflu for 35 min followed by stimulation with 100 nM insulin plus 50 nM PPP for 5 min, or left untreated as controls. Cells were fixed, permeabilized, blocked and immunostained with 4 μg/mL rabbit anti-human phospho-IRβ antibody for 1 h at 37C° followed with AlexaFluor 488 donkey anti-rabbit IgG secondary antibody for 1 h at 37C°. In order to detect the non-specific fluorescence background, control cells were incubated with secondary antibody only. Stained cells were visualized by epi-fluorescence microscopy using a 40× objective. Quantitative analysis was done by assessing the density of cell staining corrected for background in each panel using Corel Photo Paint 8.0 software. Each bar in the figure represents the mean corrected density of culture cell staining ± SEM for equal cell density (5 × 105 cells) within the respective images. P values represent significant differences at 95% confidence using the Dunnett's multiple comparison test compared to insulin treated cells. (C) Western blot analyses of Tamiflu and anti-Neu1 neutralizing antibody inhibition of insulin-induced pIRS-1R in HTC-IR cell lysates. Cell were treated with 100 nM insulin plus 50 nM PPP for 5 min or pretreated with either 400 μM Tamiflu or 33 μg/mL anti-Neu1 neutralizing antibody for 35 min followed with 100 nM insulin plus 50 nM PPP for 5 min or left untreated as control (media). Cells were pelleted, lysed in lysis buffer and the cell lysates were resolved by SDS-PAGE. The blot was probed with 4 μg/mL rabbit anti-phospho-1RS-1 antibody for 60 min at 20 °C followed with 40 ng/mL horse radish peroxidase-labeled goat anti-rabbit antibody for 75 min at 20 °C and Western Lightning Chemiluminescence Reagent Plus for 5 min. The chemiluminescence reaction was analyzed with X-ray film. The blot was re-probed with β-actin as loading control. Quantitative analysis was done by assessing the density of a band corrected for background in each lane using Corel Photo Paint 8.0 software. Each bar in the graph represents the mean ratio of pIRS1 to β-actin of band density ± S.E. (error bars) for 5–10 replicate measurements. The data are a representation of one out of three independent experiments showing similar results.

These findings suggest that the glycosylation modification of the side chains of the IRβ subunits are critical for the IR tyrosine kinase activation and signal transduction associated with insulin binding to IRα subunits. For cell-surface EGF [39] and nerve growth factor (NGF) TrkA [49] receptors, we reported a receptor signaling paradigm involving a process of receptor ligand-induced G-protein coupled receptor (GPCR) signaling via Gαi-proteins, MMP-9 activation, and the induction of Neu1 activation. To test whether this same organizational signaling platform

is involved with IR activation, we initially asked whether Neu1 forms a complex with IRβ subunits of insulin receptors. Coimmunoprecipitation experiments using cell lysates from HTC-IR cells demonstrated that Neu-1 (Fig. 3A) and not Neu-2 (Fig. 3B) or Neu-4 (Fig. 3C) forms a complex with naive and insulin-stimulated cells. The data also show that Neu-2 and -4 were present in the whole cell lysates from naive HTC-IR cells. However, we were unable to detect Neu3 in the whole cell lysates because the antibody weakly reacts with Neu-3 (data not shown). In support of the co-immunoprecipitation results with

1362

F. Alghamdi et al. / Cellular Signalling 26 (2014) 1355–1368

Fig. 3. (A) Neu1 co-immunoprecipitates with IRβ. HTC-IR cells were treated with 100 nM insulin for 5, 10 and 15 min, or left untreated as control (media). Cells were pelleted, lysed in lysis buffer and the protein lysates were immunoprecipitated with 1 μg of rabbit-anti-human Neu1 antibody, mouse-anti-human Neu2 antibody, or rabbit anti-human Neu4 antibody overnight at 4 °C. Immunocomplexes were isolated using protein G magnetic beads, resolved by SDS-PAGE and the blot probed with either anti-IRβ antibody or indicated anti-Neu1, -2 or -4 antibodies overnight at 4 °C followed by a Clean-Blot IP Detection Reagent for immunoprecipitation/Western blots for 60 min and Western Lightning Chemiluminescence Reagent Plus. HTC-IR cells not immunoprecipitated (no IP Ab) served as negative controls. Western blot of cell lysates from naive HTC-IR cells was run simultaneous with the immunoprecipitation experiments in order to determine detectable levels of IRβ proteins. For the IP blots, the blots were washed, stripped and re-probed with anti-Neu-1, -2 or -4 antibodies in order to determine equal loading. The data are a representation of one out of three independent experiments showing similar results. (B) Neu1 colocalizes with IRβ in naive HTC-IR cells. Cells were fixed, permeabilized, and immunostained with 4 μg/mL of one of the following combinations of indicated antibodies (anti-IRβ with rabbit-anti-Neu1, mouse-anti-Neu2, goat anti-Neu3, or rabbit antiNeu4 antibody) for 60 min at 37 °C. Cells were washed and incubated with 4 μg/mL donkey anti-rabbit AlexaFluor 488, donkey anti-rabbit AlexaFluor 594, donkey anti-goat AlexaFluor 594 antibodies and goat anti-mouse AlexaFluor 594 secondary antibodies for 60 min at 37 °C. To account for background non-specific fluorescence, control cells with only secondary antibody (AlexaFluor 488 or AlexaFluor 594) were used. Stained cells were visualized using a Zeiss Imager M2 fluorescence microscope with 40x objective. Images were processed using Image J version 1.38x software. To calculate the amount of colocalization in the selected images, the Pearson correlation coefficient was measured and expressed as a percentage using Image J. The data are a representation of one of three independent experiments showing similar results.

HTC-IR cells, the data shown in Fig. 3D validated the predicted association of Neu1 with IRβ. Zeiss M2 Imager fluorescence microscopy revealed the colocalization of Neu1 with IRβ (32% overlay) in naive HTC-IR cells. There was no colocalization of Neu2 and Neu4 with IRβ in these cells, which is consistent with the co-immunoprecipitation findings. Taken together, these results indicate that Neu1 forms a complex with IRβ subunits of IR on the cell surface prior to ligand binding, thereby providing confirmation additional to that of previous reports [34]. These findings suggest that Neu1 may be a critical requisite intermediate involved in the glycosylation modification of insulin receptor activation. 3.4. Neu1 and matrix metalloproteinase-9 (MMP-9) cross-talk is essential for insulin-induced IR activation in HTC-IR cells If Neu1 forms a complex with IRβ subunits of insulin receptors, we questioned how insulin binding to its receptor activates Neu1. Insight to this question came from our previous reports demonstrating a novel Neu1 and MMP-9 cross-talk in alliance with neuromedin B GPCR identified by us for EGFR [39] and NGF TrkA [49]. These reports highlight a receptor signaling paradigm involving a receptor-induced G protein coupled receptor (GPCR) signaling process and MMP-9 activation to induce Neu1. This tripartite complex of neuromedin B GPCR, MMP-9 and Neu1 forms an alliance with the receptor tethered at the ectodomain on the cell surface. Active Neu1 in complex with the receptor hydrolyzes α-2,3-sialyl residues, enabling removal of steric hindrance of receptor association and allowing subsequent dimerization, activation, and cellular signaling. We have previously reported the striking similarity between this novel receptor signaling platform for cell surface Toll-like receptor (TLR)-4 [48,50,51,54,68] and intracellular TLR-7 and TLR-9 receptors [52]. Based on these findings, we asked whether Neu1 is also connected with MMP-9 and GPCR for ligand-

induced IRβ subunits. Indeed, insulin was reported to induce MMP-9 via mitogenic signaling pathways in monocytes, whereas the phosphatidylinositol 3-kinase-dependent signaling cascade which is typically altered in insulin resistance was not involved [69]. It was proposed that the induction of MMP-9 by insulin may potentially contribute to a pro-inflammatory state with increased cardiovascular morbidity and mortality in type 2 diabetics [69]. Furthermore, the enhanced proteolytic activity in the microcirculation of spontaneously hypertensive rats (SHRs) was reported to contribute to the pathophysiological mechanism causing cell membrane receptor cleavage [70]. The hypertensive rats had elevated leukocyte counts which were accompanied by degranulation and release of proteases from lysosomal granules of neutrophils [71]. These latter results suggest that elevated MMP activity such as gelatinase B (MMP-9) [72,73] in the microcirculation of these hypertensive rats leads to proteolytic cleavage of cell surface membrane receptors including the cleavage of the IRα subunit binding domain [70]. Accordingly, we propose a Neu1-MMP-9 cross-talk in alliance with GPCR that is required for insulin activation of IRβ. To test this hypothesis, the inhibitory effect of the broad range MMP inhibitors, galardin (GM6001) and piperazine as well as the MMP3 specific inhibitor (MMP3i) on sialidase activity associated with insulinstimulated live HTC-IR cells was examined. The data shown in Figs. 4 and 5 show that piperazine (PIPZ, Fig. 4A), galardin (Fig 4B) and MMP9 inhibitor (MMP9i, Fig. 5A) except for MMP3i (Fig. 4C) dosedependently inhibited the sialidase activity associated insulin-treated live HTC-IR cells compared to the no ligand control levels. Galardin and piperazine had an IC50 of 32 μM and 15 μM, respectively, which is comparable to an IC50 of 4.89 μM for MMP9i. For MMP3i, the IC50 was N146 μM. Western blot analyses using cell lysates from HTC-IR cells clearly showed that MMP9i at 50 and 100 μg/mL inhibited insulin-induced pIRS1 in these cell lysates (Fig. 5B). Together, the additional intracellular

F. Alghamdi et al. / Cellular Signalling 26 (2014) 1355–1368

1363

Fig. 4. Insulin-induced sialidase activity in live HTC-IR cells is inhibited by (A) piperazine (PIPZ), (B) galardin (GM6001) but not by (C) specific MMP-3 inhibitor in a dose-dependent manner. After removing medium, 0.2 mM 4-MUNANA substrate in Tris-buffered saline, pH 7.4, was added to cells alone (control), with 10 nM insulin plus 50 nM PPP, or with insulin plus PPP in combination with piperazine (PIPZ), galardin or MMP-3i at the indicated concentrations. Fluorescent images were taken at 1 min after adding substrate using epi-fluorescent microscopy (40× objective). The mean fluorescence surrounding the cells for each of the images was measured using Image J software. Mean ± error bars, S.E. The data are a representation of one out of three independent experiments showing similar results. Significant differences at 95% confidence using the Dunnett's multiple comparison test was used to compare inhibition with the insulin control in each group for n = 50 replicates.

and cell-surface colocalization of IRβ and MMP-9 validated the predicted cross-talk between Neu1 and MMP-9 crosstalk tethered to IRβ subunits (Fig. 5C). Co-immunoprecipitation experiments using cell lysates from HTC-IR cells demonstrated that MMP-9 forms a complex with naive and insulin-stimulated IRβ (Fig. 5D). The inhibitory effect of neuromedin B GPCR specific antagonist BIM23127 on sialidase activity associated with bombesin-stimulated live HTC-WT cells was examined. The data shown in Fig. 6 show that BIM23127 at 333 μg/mL significantly inhibited the sialidase activity associated bombesin-treated live HTC-WT cells comparable to the no ligand control levels. Co-immunoprecipitation experiment using cell lysates from HTC-WT cells demonstrated that neuromedin B GPCR forms a complex with naive and insulin-stimulated IRβs (Fig. 6B). Collectively, the additional intracellular and cell-surface colocalization of Neu1 and MMP-9 validated the predicted cross-talk between the neuromedin B GPCR–MMP-9–Neu1 tripartite complex tethered to IRβ subunits of insulin receptors. 3.5. Olanzapine-induced Neu3 sialidase activity attenuates insulin-induced IRβ and IRS1 phosphorylation contributing to insulin resistance

the beta cell was unaffected in olanzapine-treated patients. The mechanism(s) behind olanzapine-induced insulin resistance in these patients is unknown. It has been reported that insulin-induced IRS-1 tyrosine phosphorylation in L6 myotubes was increased several fold, and olanzapine inhibited insulin-stimulated IRS-1-associated PI3K activity in a dose-dependent manner [42]. Here, we investigated the molecular mechanism(s) behind olanzapine-induced insulin resistance. As shown in Fig. 7A and B, olanzapine stimulation of live 1140F01 Neu1-deficient cells as well as primary bone marrow (BM) murine macrophage cells derived from Neu4 knockout (KO) mice induced sialidase activity which was inhibited by the Neu3 inhibitor, DANA (2-deoxy-2,3dehydro-N-acetylneuraminic acid) [74] and anti-Neu3 neutralizing antibodies. In addition, immunocytochemistry analyses of wild-type fibroblast pretreated with olanzapine or anti-Neu1 antibodies attenuated phosphorylation of IGF-R and IRS1 associated with insulinstimulated human fibroblast cells. Based on these results, we propose that olanzapine-induced Neu3 sialidase activity down-regulates Neu1, the inhibition of which attenuated insulin-induced IGF-R and IRS1 phosphorylation contributing to insulin resistance. 4. Discussion

Disturbances in glucose homeostasis and metabolism in patients with schizophrenia during antipsychotic treatment with olanzapine are mainly due to insulin resistance [40–42] whereas the function of

The molecular mechanism(s) by which IRs become activated is not well understood. There is substantial evidence to indicate that insulin

1364

F. Alghamdi et al. / Cellular Signalling 26 (2014) 1355–1368

Fig. 5. (A) Specific MMP-9 inhibitor blocks dose-dependently insulin-induced sialidase activity in live HTC-IR cells. After removing medium, 0.2 mM 4-MUNANA substrate in Tris-buffered saline, pH 7.4, was added to cells alone (control), with 10 nM insulin plus 50 nM PPP, or with insulin plus PPP in combination with MMP-9i at the indicated concentrations. Fluorescent images were taken at 1 min after adding substrate using epi-fluorescent microscopy (40× objective). The mean fluorescence surrounding the cells for each of the images was measured using Image J software. Mean ± error bars, S.E. The data are a representation of one out of three independent experiments showing similar results. Significant differences at 95% confidence using the Dunnett's multiple comparison test was used to compare inhibition with the insulin control in each group for n = 50 replicates. (B) Western blot analyses of MMP9i on insulininduced IRS-1R phosphorylation (pIRS-1R) in HTC-IR cells. Cells were left untreated (control), stimulated with 100 nM insulin plus 50 nM PPP for 5 min, or pretreated with 50 μg/mL or 100 μg/mL MMP9i for 35 min followed with 100 nM insulin plus 50 nM PPP for 5 min. Cells were pelleted, lysed in lysis buffer and the cell lysates were resolved by SDS-PAGE. The blot was probed with rabbit anti-human pIR-S1 antibody overnight at 4 °C. The blot was incubated with secondary HRP conjugated goat anti-rabbit IgG antibody for 60 min at room temperature and Western Lightning Chemiluminescence Reagent Plus. The chemiluminescence reaction was analyzed with X-ray film. The blot was re-probed with β-actin as loading control. Quantitative analysis was done by assessing the density of a band corrected for background in each lane using Corel Photo Paint 8.0 software. Each bar in the graph represents the mean ratio of pIRS1 to β-actin of band density ± S.E. (error bars) for 5–10 replicate measurements. The data are a representation of one out of three independent experiments showing similar results. (C) MMP9 colocalizes with IRβ in naive HTC-IR cells. Cells were cultured on 12 mm glass coverslips in a sterile 24 well plate in conditioned medium as described. Cells were fixed, permeabilized and blocked with 4% bovine serum albumin (BSA) in 0.1% tween-TBS. Cells were incubated with 4 μg/mL of rabbit anti-IRβ antibody and goat anti-MMP9 antibody for 60 min at 37 °C. Cells were washed and incubated with 4 μg/mL of AlexaFluor 594 conjugated donkey anti-goat IgG and AlexaFluor 488 conjugated donkey anti-rabbit IgG for 60 min at 37 °C. To account for background non-specific fluorescence, control cells with only secondary antibody (AlexaFluor 488 or AlexaFluor 594) were used. Stained cells were visualized using a Zeiss Imager M2 fluorescence microscope with 40× objective. Images were processed using Image J version 1.38x software. To calculate the amount of colocalization in the selected images, the Pearson correlation coefficient was measured and expressed as a percentage using Image J. The data are a representation of one of three independent experiments showing similar results. (D) IRβ co-immunoprecipitates with MMP9 in naive and insulin-stimulated HTC-IR cells. Cells were left unstimulated as a control or stimulated with 100 nM insulin plus 50 nM PPP for 5,10 or 15 min. Cells were pelleted, lysed in lysis buffer and the protein lysates were immunoprecipitated with 1 μg of goat anti-MMP9 antibody overnight at 4 °C. Immunocomplexes were isolated using protein G magnetic beads, resolved by SDS-PAGE and the blot probed with anti-IRβ antibody overnight at 4 °C followed by a Clean-Blot IP Detection Reagent for immunoprecipitation/western blots for 60 min and Western Lightning Chemiluminescence Reagent Plus. HTC-IR cells not immunoprecipitated (no IP Ab) served as negative controls. For the IP blots, the blots were washed, stripped and re-probed with anti-MMP9 antibodies in order to determine loading. The data are a representation of one out of three independent experiments showing similar results. The chemiluminescence reaction was analyzed with X-ray film. The data are a representation of one out of three independent experiments showing similar results.

receptor glycosylation modification may in fact be the connecting link between ligand-binding and activation [16,18–20,24,34]. However, the parameters controlling interactions between the receptors and insulin to facilitate receptor activation remained poorly defined until now. The findings in this report uncover a molecular organizational GPCR signaling platform of a novel Neu1 and MMP-9 cross-talk in alliance with IRβ subunits on the cell surface of IR-expressing cells that is essential for insulin activation of IRs and subsequent cellular signaling. At the genetic level, the sialidase activity associated with insulin stimulation of human WG544 and 1140F01 type 1 sialidosis fibroblast cell lines was

completely abrogated. These sialidosis fibroblast cells were obtained from patients with type 1 sialidosis or mucolipidosis-1 who have a true Neu1 deficiency [44]. The data further indicate that Neu1 forms a complex with the IRβ subunits of the receptor on the cell surface prior to ligand binding, thereby providing confirmation additional to that of previous reports [12,34]. The report by Dridi et al. provided evidence to demonstrate that Neu1 deficient mice with ∼ 10% of the normal Neu1 activity exposed to a high-fat diet developed hyperglycemia and insulin resistance twice as fast as the wild-type cohort [34]. In addition, mutational analyses targeting IRβ subunits which were mutagenized on

F. Alghamdi et al. / Cellular Signalling 26 (2014) 1355–1368

1365

Fig. 6. (A) Insulin- and bombesin-induced sialidase activity in live HTC-WT cells are inhibited by Tamiflu and selective neuromedin B GPCR antagonist BIM-23127. After removing medium, 0.2 mM 4-MUNANA substrate in Tris-buffered saline, pH 7.4, was added to cells alone (control), with 100 nM insulin plus 50 nM PPP, or with insulin plus PPP in combination with 200 μg/mL Tamiflu or 333 μg/mL BIM-23127. Fluorescent images were taken at 1 min after adding substrate using epi-fluorescent microscopy (40× objective). The mean fluorescence surrounding the cells for each of the images was measured using Image J software. Mean ± error bars, S.E. The data are a representation of one out of three independent experiments showing similar results. (B) Neuromedin B GPCR (NMBR) co-immunoprecipitates with IRβ and verse versa in naive and insulin-stimulated HTC-WT cells. Cells were left unstimulated as a media control or stimulated with 100 nM insulin plus 50 nM PPP for 5 min. Cells were pelleted, lysed in lysis buffer and the protein lysates were immunoprecipitated with 1 μg of anti-IRβ antibody overnight at 4 °C. Immunocomplexes were isolated using protein G magnetic beads, resolved by SDS-PAGE and the blot probed with anti-NMBR antibody overnight at 4 °C followed by a Clean-Blot IP Detection Reagent for immunoprecipitation/western blots for 60 min and Western Lightning Chemiluminescence Reagent Plus. The chemiluminescence reaction was analyzed with X-ray film. Cells not immunoprecipitated (no Ab) served as negative controls. The data are a representation of one out of three independent experiments showing similar results.

Fig. 7. Olanzapine induces sialidase activity in live (A) 1140F01 Neu1-deficient human fibroblast cells derived from type 1 sialidosis patients and (B) primary bone marrow macrophage cells derived from Neu4 knockout mice. Cells were treated with 100 nM insulin, 10 μg/mL olanzapine, pretreated with olanzapine with 200 μg/mL DANA or 62.5 μg/mL anti-Neu3 antibodies or left untreated as control cells. The data are a representation of one out of four independent experiments showing similar results. (C) Olanzapine and anti-Neu1 neutralizing antibody inhibit insulin induced phosphorylation of IGF-1R and IRS-1 in wild-type human fibroblast cells. Cells were grown overnight on glass coverslips in a 24-well tissue culture plate at 37 °C for 24 h or until they reached ~70% confluence. Cells were stimulated with 100 nM insulin for 5 min, pretreated with either 100 μM olanzapine or 1.4 μg/mL anti-Neu1 neutralizing antibodies for 30 min followed with 100 nM insulin for 5 min or were left untreated as no ligand controls. Cells were fixed, permeabilized, blocked with 4% bovine serum albumin (BSA) in 0.1% Tween-Tris buffered saline (TBS) for 20 min on ice and immunostained with rabbit anti-phospho-IGF-1R or rabbit anti-phospho-1RS-1 for 60 min at 37 °C, followed with AlexaFluor 594 goat anti-rabbit secondary antibody for 60 min at 37 °C. Control group had only secondary antibodies with no other treatment. Stained cells were visualized by epi-fluorescence microscopy with a 40× objective. The data are a representation of one out of three independent experiments showing similar results.

1366

F. Alghamdi et al. / Cellular Signalling 26 (2014) 1355–1368

four N-glycosylation sites (IRβ-N1234) attenuated IRβ tyrosine kinase activation by insulin [18]. Collectively, our data provide additional evidence to demonstrate that a tripartite complex of neuromedin B GPCR–MMP-9–Neu1 is tethered to IRβ subunits of insulin receptors. This signaling paradigm proposes that insulin binding to its receptor on the cell surface induces a conformational change of the receptor to initiate GPCR-signaling and MMP-9 activation to induce Neu1. There is substantial evidence strongly to demonstrate that insulin binding to the receptor induces a conformational change (a) at the C terminus domain [75], (b) increases the rate of autophosphorylation rather than increases ATP affinity [76], (c) involves a fast bimolecular binding reaction followed by a slower unimolecular step representing a conformational change in the insulin-receptor complex [77], and (d) favors a basalstate conformation of the activation loop that is more accessible overcoming the intra-sterically inhibitory conformation that encompasses ≥90% of the ligand-free basal state kinase [78]. Based on mutagenesis studies on the four N-glycosylation sites (IRβ-N1234) of the IRβ subunit of the receptor [18], the intra-sterically inhibitory ligand-free basal state kinase is most likely under the control of α-2,3 sialic acid residues of IRβ, and the kinase activation state induced by insulin is regulated by Neu1 sialidase activity in hydrolyzing these sialic residues and removing steric hindrance of the activation loop as shown by our data. Structurally, the insulin receptor is composed of a heterotetramer consisting of two αβ dimers [79]. Disulfide bridges connect the αβ subunits within the dimer as well as the IRα subunits of αβ dimers within the tetramer [80]. The IRα subunit is extracellular and provides the insulin binding site. The IRβ subunit contains a short extracellular region, a transmembrane segment and an intracellular tyrosine kinase domain. Analysis of these structures suggests that an insulin-induced conformational change in the dimeric structure may change the relative orientation of the tyrosine kinases leading to transautophosphorylation and activation of the receptor [81,82]. The molecular mechanism(s) involved in the orientation of the activation loop containing the tyrosine kinases was unknown until now. This report proposes a novel molecular organizational GPCR signaling platform for the IRβ subunits. It is based on the findings that Neu1 and MMP-9 are already in complex with the ligand-free basal state IRβ, and are rapidly induced upon insulin binding to the extracellular IRα subunit. Activated Neu1 specifically hydrolyzes α-2,3-sialyl residues linked to β-galactoside of the ligand-free basal state kinases of the activation loop, which are distant from ligand binding. This process enables removal of steric hindrance for proper orientation of the tyrosine kinases leading to trans-autophosphorylation and activation of the receptor and subsequent IR signaling cascades. The findings in this report also provide evidence for MMP-9 involvement in the insulin activation of IRβ. We have reported that Neu1 in alliance with GPCRsignaling Gαi subunit proteins and MMP-9 is expressed on the cell surface of EGFR- [39], TrkA- [49], and TLR- [51,52] expressing cells. This tripartite alliance makes Neu1 readily available to be induced by ligand binding to the receptor. Our data support this premise. How Neu1 sialidase is rapidly induced by MMP-9 together with GPCR Gαi- subunit proteins still remains unknown. It can be speculated that insulin binding to IRα subunits on the cell surface initiates GPCR-signaling to activate MMP-9. Our data using co-immunoprecipitation of IRβ and neuromedin B GPCR suggest that the neuromedin B GPCR forms a complex with naive and insulin-induced IR activation in HTC-WT cells (see Fig. 6B). Indeed, it has been reported that neuromedin B receptor knockout (NBR-KO) female mice did not change body weight and homeostasis fed a normolipid diet, but they did developed partial resistance to diet-induced obesity [83]. The GPCR signal integration in insulin receptor activation is eloquently reviewed by Patel [84] and Abdulkhalek et al. [33]. Since both MMP-9 and NMBR form a complex with IRβ subunits, they are also associated with each other. Co-immunoprecipitation experiments using cell lysates from RAW-Blue cells demonstrated that the neuromedin B GPCR 80 kDa isoform forms a complex with the active 88 kDa MMP9 isoform from naive or lipopolysaccharide (LPS)-

stimulated cells [68]. These data further validated that neuromedin B GPCR forms a complex with MMP-9 on the cell surface of naive cells. The data in the report also showed that GPCR agonists such as bombesin, lysophosphatidic acid (LPA), cholesterol, angiotensin-1 and -2, and bradykinin binding to their respective GPCR receptors induced Neu1 activity within 1 min and that this activity was blocked by Gαisensitive pertussis toxin, neuraminidase inhibitor Tamiflu, broadrange MMP inhibitors galardin and piperazine, anti-Neu1 and antiMMP-9 antibodies, and siRNA knockdown of MMP-9. The rapidity of the GPCR agonist-induced Neu1 activity suggests that glycosylated receptors like TLRs as well as receptor tyrosine kinases form a functional GPCR-signaling complex. Indeed, Gilmour et al. [39] and Moody et al. [85] have reported that the neuromedin B GPCR regulates EGF receptors by a mechanism dependent on MMP activation. It is well known that agonist-bound GPCRs have been shown to activate numerous MMPs [86], including MMP-3 [87], MMPs 2 and 9 [88,89], including the members of the ADAM family of metalloproteinases [90,91]. We have shown that GPCR agonists can directly activate Neu1 through the intermediate MMP-9 in order to induce transactivation of TLRs and subsequent cellular signaling [52,68]. It is noteworthy that insulin can mediate increases in MMP-9 via IR activation [69], which fits well within our molecular signaling platform of Neu1-MMP-9 cross-talk in regulating insulininduced receptors. That study has also shown that insulin can induce MMP-9 via the mitogenic signaling pathways, whereas the phosphatidylinositol 3-kinase-dependent signaling which is typically altered in insulin resistance is not required [69]. The connection between GPCR and IR has also been demonstrated for β-adrenergic receptors tethering to IR in adipocytes [92–95]. These reports showed that insulin binding IR stimulates the phosphorylation of the β-adrenergic receptor on Tyr350 and this process facilitates IR tethering to β-adrenergic receptor via growth factor receptor-bound protein 2 (Grb-2). This molecular signaling platform integrating the IR/β-adrenergic receptor/Grb-2 tripartite complex is critical for β-adrenergic agonist amplification of insulin-dependent activation of p42/p44 MAPK. The premise behind this molecular signaling platform is to improve IR signaling in our present study and others [92–95], but also in other RTKs like PDGFβR [25] with a GPCR. This latter study demonstrated that PDGF induced a stronger tyrosine phosphorylation of Gαi and a more robust activation of p42/p44 MAPK in cells transfected with both PDGFβR and EDG-1, the GPCR for SIP-1, compared with PDGFβR alone. These RTK–GPCR signaling platforms are eloquently reviewed by Pyne and Pyne [30]. If Neu1 sialidase is playing a major role in the activation of IR as our data suggest and by others [34], logistically Neu1 may be the potential target for insulin resistance. Here, we used a known antipsychotic drug that induces insulin resistance in patients with schizophrenia [40–42], which was not due a malfunction of the β cell. The mechanism(s) behind olanzapine-induced insulin resistance in these patients is unknown. At the genetic level, we have shown that olanzapine induces Neu3 sialidase activity in Neu1-deficient human sialidosis fibroblast cells and in Neu4 knockout primary macrophage cells. Wild-type human fibroblast cells pretreated with olanzapine attenuated phosphorylation of IGF-R and IRS1 associated with insulin-stimulated cells. Collectively, olanzapine-induced Neu3 sialidase activity is proposed to down-regulate Neu1, the inhibition of which attenuates insulininduced IGF-R and IRS1 phosphorylation and contributing to insulin resistance. Indeed, Neu3 has been shown to participate in the control of insulin signaling, most likely via the modulation of gangliosides and the interaction with growth factor receptor-bound protein 2 (Grb2) [24]. That report also showed that the transient transfection of Neu3 into 3T3-L1 adipocytes and L6 myocytes caused a significant decrease in IR signaling. Another report has shown that overexpression of Neu3 sialidase inhibits MMP-9 expression in vascular smooth muscle cells [96]. Recently, it was shown that chow-fed mice given acute and chronic intravenous injections of elastin-binding peptides developed hyperglycemia and insulin resistance [12]. The report proposed that this insulin resistance is due to the interaction of IR with Neu1 subunit of the elastin

F. Alghamdi et al. / Cellular Signalling 26 (2014) 1355–1368

receptor complex triggered by elastin-binding peptides. In another report by these authors, it was shown that following elastin peptide treatment, the cellular GM3 levels decreased while lactosylceramide (LacCer) content increased consistently with a GM3/LacCer conversion [97]. It was suggest that Neu-1-dependent GM3/LacCer conversion is the key event leading to signaling by the elastin receptor complex. Based on these latter results, an alternate interpretation of the role of elastin-binding peptides in insulin resistance may involve the mechanism of direct modulation of plasma membrane gangliosides by increasing Neu3 activity and the interaction of which is facilitated by the growth factor receptor-bound protein 2 (Grb2) as described elsewhere [24]. In the present study, olanzapine-induced Neu3 sialidase activity down-regulates Neu1 by inhibiting MMP9 activation and thereby, disrupting the Neu1–MMP9-neuromedin B GPCR signaling platform tethered to IRβ subunits in the activation of insulin receptors. We propose here that a negative unbalance of this novel IRβ subunit signaling platform critical for the insulin receptor contributes to insulin resistance and type 2 diabetes. Competing interest statement The authors declare no competing financial interests. Authors' contributions F. Alghamdi and M.R. Szewczuk wrote the paper, designed and performed experiments; S. Abdulkhalek assisted in the coimmunoprecipitation and cell cultures; M. Guo performed the neuromedin B experiments; R. Amith performed the sialidase assays on Neu1 deficient human fibroblast cells and Neu4 knockout macrophages; N. Crawford performed the olanzapine studies. M.R. Szewczuk supervised the research design and the writing of the paper. All authors read and commented on the manuscript. Acknowledgments This work was supported in part by a grant to M. R. Szewczuk from the Natural Sciences and Engineering Research Council of Canada (NSERC). F. Alghamdi was the recipient of the King Abdullah Scholarship from the Ministry of Higher Education, Saudi Arabia. S. Abdulkhalek was the recipient of the R.S. McLaughlin Scholarship, the Ontario Graduate Scholarship and the Canadian Institutes of Health Research (CIHR) Doctoral Award: Frederick Banting and Charles Best Canada Graduate Scholarship. References [1] S.M. Hirabara, R. Gorjao, M.A. Vinolo, A.C. Rodrigues, R.T. Nachbar, R. Curi, J. Biomed. Biotechnol. 2012 (2012) 379024. [2] R.A. DeFronzo, E. Ferrannini, Diabetes Care 14 (3) (1991) 173–194. [3] S.E. Kahn, Am. J. Med. 108 (Suppl. 6a) (2000) 2S–8S. [4] F. Mauvais-Jarvis, C.R. Kahn, Diabetes Metab. 26 (6) (2000) 433–448. [5] B.C. Martin, J.H. Warram, A.S. Krolewski, R.N. Bergman, J.S. Soeldner, C.R. Kahn, Lancet 340 (8825) (1992) 925–929. [6] M.A. Abdul-Ghani, R.A. DeFronzo, J. Biomed. Biotechnol. 2010 (2010) 476279. [7] A. Kontrogianni-Konstantopoulos, G. Benian, H. Granzier, J. Biomed. Biotechnol. 2010 (2010). [8] E. Maratou, D.J. Hadjidakis, M. Peppa, M. Alevizaki, K. Tsegka, V. Lambadiari, P. Mitrou, E. Boutati, A. Kollias, T. Economopoulos, S.A. Raptis, G. Dimitriadis, Eur. J. Endocrinol. 163 (4) (2010) 625–630. [9] M. Peppa, E. Boutati, C. Koliaki, N. Papaefstathiou, E. Garoflos, T. Economopoulos, D. Hadjidakis, S.A. Raptis, Metabolism 59 (10) (2010) 1435–1441. [10] M. Peppa, C. Koliaki, P. Nikolopoulos, S.A. Raptis, J. Biomed. Biotechnol. 2010 (2010) 527850. [11] A.R. Martins, R.T. Nachbar, R. Gorjao, M.A. Vinolo, W.T. Festuccia, R.H. Lambertucci, M.F. Cury-Boaventura, L.R. Silveira, R. Curi, S.M. Hirabara, Lipids Health Dis. 11 (2012) 30. [12] S. Blaise, B. Romier, C. Kawecki, M. Ghirardi, F. Rabenoelina, S. Baud, L. Duca, P. Maurice, A. Heinz, C.E. Schmelzer, M. Tarpin, L. Martiny, C. Garbar, M. Dauchez, L. Debelle, V. Durlach, Diabetes 62 (11) (2013) 3807–3816.

1367

[13] A.I. Salhanick, J.M. Amatruda, Am. J. Physiol. 255 (2 Pt 1) (1988) E173–E179. [14] C. Ward, M. Lawrence, V. Streltsov, T. Garrett, N. McKern, M.Z. Lou, G. Lovrecz, T. Adams, Acta Physiol (Oxf) 192 (1) (2008) 3–9. [15] L.G. Sparrow, M.C. Lawrence, J.J. Gorman, P.M. Strike, C.P. Robinson, N.M. McKern, C.W. Ward, Proteins 71 (1) (2008) 426–439. [16] T.C. Elleman, M.J. Frenkel, P.A. Hoyne, N.M. McKern, L. Cosgrove, D.R. Hewish, K.M. Jachno, J.D. Bentley, S.E. Sankovich, C.W. Ward, Biochem. J. 347 (Pt 3) (2000) 771–779. [17] J.P. Mayer, F. Zhang, R.D. DiMarchi, Biopolymers 88 (5) (2007) 687–713. [18] I. Leconte, C. Auzan, A. Debant, B. Rossi, E. Clauser, J. Biol. Chem. 267 (24) (1992) 17415–17423. [19] T.P. Ciaraldi, Diabetes 38 (8) (1989) 951–958. [20] M. Arabkhari, S. Bunda, Y. Wang, A. Wang, A.V. Pshezhetsky, A. Hinek, Glycobiology 20 (5) (2010) 603–616. [21] A. Fanzani, R. Giuliani, F. Colombo, D. Zizioli, M. Presta, A. Preti, S. Marchesini, FEBS Lett. 547 (1–3) (2003) 183–188. [22] L. Anastasia, N. Papini, F. Colazzo, G. Palazzolo, C. Tringali, L. Dileo, M. Piccoli, E. Conforti, C. Sitzia, E. Monti, M. Sampaolesi, G. Tettamanti, B. Venerando, J. Biol. Chem. 283 (52) (2008) 36265–36271. [23] K. Sato, T. Miyagi, Biochem. Biophys. Res. Commun. 221 (3) (1996) 826–830. [24] A. Sasaki, K. Hata, S. Suzuki, M. Sawada, T. Wada, K. Yamaguchi, M. Obinata, H. Tateno, H. Suzuki, T. Miyagi, J. Biol. Chem. 278 (30) (2003) 27896–27902. [25] F. Alderton, S. Rakhit, K.C. Kong, T. Palmer, B. Sambi, S. Pyne, N.J. Pyne, J. Biol. Chem. 276 (30) (2001) 28578–28585. [26] P.L. Rothenberg, C.R. Kahn, J. Biol. Chem. 263 (30) (1988) 15546–15552. [27] L. Luttrell, E. Kilgour, J. Larner, G. Romero, J. Biol. Chem. 265 (28) (1990) 16873–16879. [28] R.M. O'Brien, M.D. Houslay, G. Milligan, K. Siddle, FEBS Lett. 212 (2) (1987) 281–288. [29] N.J. Pyne, S. Pyne, Biochim. Biophys. Acta 1781 (9) (2008) 467–476. [30] N.J. Pyne, S. Pyne, Trends Pharmacol. Sci. 32 (8) (2011) 443–450. [31] N.J. Pyne, C. Waters, N.A. Moughal, B.S. Sambi, S. Pyne, Biochem. Soc. Trans. 31 (Pt 6) (2003) 1220–1225. [32] N.J. Pyne, C.M. Waters, J.S. Long, N.A. Moughal, G. Tigyi, S. Pyne, Adv. Enzyme Regul. 47 (2007) 271–280. [33] S. Abdulkhalek, M. Hrynyk, M.R. Szewczuk, Research and Reports in Biochemistry, 32013. 17–30. [34] L. Dridi, V. Seyrantepe, A. Fougerat, X. Pan, E. Bonneil, P. Thibault, A. Moreau, G.A. Mitchell, N. Heveker, C.W. Cairo, T. Issad, A. Hinek, A.V. Pshezhetsky, Diabetes 62 (7) (2013) 2338–2346. [35] L.J. Slieker, M.D. Lane, J. Biol. Chem. 260 (2) (1985) 687–690. [36] A.M. Soderquist, G. Carpenter, J. Biol. Chem. 259 (20) (1984) 12586–12594. [37] M. Takahashi, T. Tsuda, Y. Ikeda, K. Honke, N. Taniguchi, Glycoconj. J. 20 (3) (2004) 207–212. [38] H. Fernandes, S. Cohen, S. Bishayee, J. Biol. Chem. 276 (7) (2001) 5375–5383. [39] A.M. Gilmour, S. Abdulkhalek, T.S. Cheng, F. Alghamdi, P. Jayanth, L.K. O'Shea, O. Geen, L.A. Arvizu, M.R. Szewczuk, Cell. Signal. 25 (12) (2013) 2587–2603. [40] C.F. Ebenbichler, M. Laimer, U. Eder, B. Mangweth, E. Weiss, A. Hofer, M. Hummer, G. Kemmler, M. Lechleitner, J.R. Patsch, W.W. Fleischhacker, J. Clin. Psychiatry 64 (12) (2003) 1436–1439. [41] A.A. Richards, I.J. Hickman, A.Y. Wang, A.L. Jones, F. Newell, B.J. Mowry, J.P. Whitehead, J.B. Prins, G.A. Macdonald, J. Clin. Psychopharmacol. 26 (3) (2006) 232–237. [42] J. Engl, M. Laimer, A. Niederwanger, M. Kranebitter, M. Starzinger, M.T. Pedrini, W.W. Fleischhacker, J.R. Patsch, C.F. Ebenbichler, Mol. Psychiatry 10 (12) (2005) 1089–1096. [43] M. Li, J.F. Youngren, V.P. Manchem, M. Kozlowski, B.B. Zhang, B.A. Maddux, I.D. Goldfine, Diabetes 50 (10) (2001) 2323–2328. [44] P. Bifsha, K. Landry, L. Ashmarina, S. Durand, V. Seyrantepe, S. Trudel, C. Quiniou, S. Chemtob, Y. Xu, R.A. Gravel, R. Sladek, A.V. Pshezhetsky, Cell Death Differ. 14 (3) (2007) 511–523. [45] V. Seyrantepe, M. Canuel, S. Carpentier, K. Landry, S. Durand, F. Liang, J. Zeng, A. Caqueret, R.A. Gravel, S. Marchesini, C. Zwingmann, J. Michaud, C.R. Morales, T. Levade, A.V. Pshezhetsky, Hum. Mol. Genet. 17 (11) (2008) 1556–1568. [46] A. Alatery, S. Basta, J. Immunol. Methods 338 (1–2) (2008) 47–57. [47] A. Woronowicz, K. De Vusser, W. Laroy, R. Contreras, S.O. Meakin, G.M. Ross, M.R. Szewczuk, Glycobiology 14 (11) (2004) 987–998. [48] S.R. Amith, P. Jayanth, T. Finlay, S. Franchuk, A. Gilmour, S. Abdulkhalek, M.R. Szewczuk, J. Vis. Exp. (43) (2010). [49] P. Jayanth, S.R. Amith, K. Gee, M.R. Szewczuk, Cell. Signal. 22 (8) (2010) 1193–1205. [50] S.R. Amith, P. Jayanth, S. Franchuk, T. Finlay, V. Seyrantepe, R. Beyaert, A.V. Pshezhetsky, M.R. Szewczuk, Cell. Signal. 22 (2) (2010) 314–324. [51] S. Abdulkhalek, S.R. Amith, S.L. Franchuk, P. Jayanth, M. Guo, T. Finlay, A. Gilmour, C. Guzzo, K. Gee, R. Beyaert, M.R. Szewczuk, J. Biol. Chem. 286 (42) (2011) 36532–36549. [52] S. Abdulkhalek, M.R. Szewczuk, Cell. Signal. 25 (11) (2013) 2093–2105. [53] A. Woronowicz, S.R. Amith, K. De Vusser, W. Laroy, R. Contreras, S. Basta, M.R. Szewczuk, Glycobiology 17 (1) (2007) 10–24. [54] S.R. Amith, P. Jayanth, S. Franchuk, S. Siddiqui, V. Seyrantepe, K. Gee, S. Basta, R. Beyaert, A.V. Pshezhetsky, M.R. Szewczuk, Glycoconj. J. 26 (9) (2009) 1197–1212. [55] T. Miyagi, J. Sagawa, K. Konno, S. Tsuiki, J. Biochem. (Tokyo) 107 (5) (1990) 794–798. [56] K.E. Lukong, M.A. Elsliger, Y. Chang, C. Richard, G. Thomas, W. Carey, A. Tylki-Szymanska, B. Czartoryska, T. Buchholz, G.R. Criado, S. Palmeri, A.V. Pshezhetsky, Hum. Mol. Genet. 9 (7) (2000) 1075–1085.

1368

F. Alghamdi et al. / Cellular Signalling 26 (2014) 1355–1368

[57] F. Liang, V. Seyrantepe, K. Landry, R. Ahmad, A. Ahmad, N.M. Stamatos, A.V. Pshezhetsky, J. Biol. Chem. 281 (37) (2006) 27526–27538. [58] X. Nan, I. Carubelli, N.M. Stamatos, J. Leukoc. Biol. 81 (1) (2007) 284–296. [59] G. Yogalingam, E.J. Bonten, D. van de Vlekkert, H. Hu, S. Moshiach, S.A. Connell, A. d'Azzo, Dev. Cell 15 (1) (2008) 74–86. [60] A. Hinek, A.V. Pshezhetsky, I.M. Von, B. Starcher, J. Biol. Chem. 281 (6) (2006) 3698–3710. [61] L. Duca, C. Blanchevoye, B. Cantarelli, C. Ghoneim, S. Dedieu, F. Delacoux, W. Hornebeck, A. Hinek, L. Martiny, L. Debelle, J. Biol. Chem. 282 (17) (2007) 12484–12491. [62] B. Starcher, A. d'Azzo, P.W. Keller, G.K. Rao, D. Nadarajah, A. Hinek, Am. J. Physiol. Lung Cell. Mol. Physiol. 295 (4) (2008) L637–L647. [63] J.A. Rodriguez, E. Piddini, T. Hasegawa, T. Miyagi, C.G. Dotti, J. Neurosci. 21 (21) (2001) 8387–8395. [64] N. Papini, L. Anastasia, C. Tringali, G. Croci, R. Bresciani, K. Yamaguchi, T. Miyagi, A. Preti, A. Prinetti, S. Prioni, S. Sonnino, G. Tettamanti, B. Venerando, E. Monti, J. Biol. Chem. 279 (17) (2004) 16989–16995. [65] K. Yamaguchi, K. Hata, K. Koseki, K. Shiozaki, H. Akita, T. Wada, S. Moriya, T. Miyagi, Biochem. J. 390 (Pt 1) (2005) 85–93. [66] V. Seyrantepe, H. Poupetova, R. Froissart, M.T. Zabot, I. Maire, A.V. Pshezhetsky, Hum. Mutat. 22 (5) (2003) 343–352. [67] K. Hata, K. Koseki, K. Yamaguchi, S. Moriya, Y. Suzuki, S. Yingsakmongkon, G. Hirai, M. Sodeoka, M. von Itzstein, T. Miyagi, Antimicrob. Agents Chemother. 52 (10) (2008) 3484–3491. [68] S. Abdulkhalek, M. Guo, S.R. Amith, P. Jayanth, M.R. Szewczuk, Cell. Signal. 24 (11) (2012) 2035–2042. [69] A. Fischoeder, H. Meyborg, D. Stibenz, E. Fleck, K. Graf, P. Stawowy, Cardiovasc. Res. 73 (4) (2007) 841–848. [70] F.A. DeLano, G.W. Schmid-Schonbein, Hypertension 52 (2) (2008) 415–423. [71] K. Shen, K.L. Sung, D.E. Whittemore, F.A. DeLano, B.W. Zweifach, G.W. Schmid-Schonbein, Biochem. Cell Biol. 73 (7–8) (1995) 491–500. [72] T.M. Camp, S.C. Tyagi, R.M. Senior, M.R. Hayden, S.C. Tyagi, Diabetologia 46 (10) (2003) 1438–1445. [73] T.M. Camp, L.M. Smiley, M.R. Hayden, S.C. Tyagi, J. Hypertens. 21 (9) (2003) 1719–1727. [74] Y. Zou, A. Albohy, M. Sandbhor, C.W. Cairo, Bioorg. Med. Chem. Lett. 20 (24) (2010) 7529–7533. [75] V. Baron, N. Gautier, A. Komoriya, P. Hainaut, J.C. Scimeca, M. Mervic, S. Lavielle, J. Dolais-Kitabgi, E. Van Obberghen, Biochemistry 29 (19) (1990) 4634–4641. [76] J. Lee, P.F. Pilch, S.E. Shoelson, S.F. Scarlata, Biochemistry 36 (9) (1997) 2701–2708.

[77] M. Schlein, S. Havelund, C. Kristensen, M.F. Dunn, N.C. Kaarsholm, J. Mol. Biol. 303 (2) (2000) 161–169. [78] M. Frankel, S.M. Bishop, A.J. Ablooglu, Y.-P. Han, R.A. Kohanski, Protein Sci. 8 (10) (1999) 2158–2165. [79] A. Ullrich, J.R. Bell, E.Y. Chen, R. Herrera, L.M. Petruzzelli, T.J. Dull, A. Gray, L. Coussens, Y.C. Liao, M. Tsubokawa, et al., Nature 313 (6005) (1985) 756–761. [80] L.G. Sparrow, N.M. McKern, J.J. Gorman, P.M. Strike, C.P. Robinson, J.D. Bentley, C.W. Ward, J. Biol. Chem. 272 (47) (1997) 29460–29467. [81] S.R. Hubbard, L. Wei, L. Ellis, W.A. Hendrickson, Nature 372 (6508) (1994) 746–754. [82] N.Q. McDonald, J. Murray-Rust, T.L. Blundell, Structure 3 (1) (1995) 1–6. [83] G.S. Paula, L.L. Souza, A. Cabanelas, F.F. Bloise, V. Mello-Coelho, E. Wada, T.M. Ortiga-Carvalho, K.J. Oliveira, C.C. Pazos-Moura, J. Physiol. 588 (Pt 9) (2010) 1635–1645. [84] T.B. Patel, Pharmacol. Rev. 56 (3) (2004) 371–385. [85] T.W. Moody, M.J. Berna, S. Mantey, V. Sancho, L. Ridnour, D.A. Wink, D. Chan, G. Giaccone, R.T. Jensen, Eur. J. Pharmacol. 637 (1-3) (2010) 38–45. [86] O.M. Fischer, S. Hart, A. Ullrich, Methods Mol. Biol. 327 (2006) 85–97. [87] M.-H. Lee, G. Murphy, J. Cell Sci. 117 (18) (2004) 4015–4016. [88] S.M. Le Gall, R. Auger, C. Dreux, P. Mauduit, J. Biol. Chem. 278 (46) (2003) 45255–45268. [89] S. Murasawa, Y. Mori, Y. Nozawa, N. Gotoh, M. Shibuya, H. Masaki, K. Maruyama, Y. Tsutsumi, Y. Moriguchi, Y. Shibazaki, Y. Tanaka, T. Iwasaka, M. Inada, H. Matsubara, Circ. Res. 82 (12) (1998) 1338–1348. [90] M. Gooz, P. Gooz, L.M. Luttrell, J.R. Raymond, J. Biol. Chem. 281 (30) (2006) 21004–21012. [91] N. Prenzel, E. Zwick, H. Daub, M. Leserer, R. Abraham, C. Wallasch, A. Ullrich, Nature 402 (6764) (1999) 884–888. [92] V. Karoor, L. Wang, H.Y. Wang, C.C. Malbon, J. Biol. Chem. 273 (49) (1998) 33035–33041. [93] V. Karoor, C.C. Malbon, J. Biol. Chem. 271 (46) (1996) 29347–29352. [94] K. Baltensperger, V. Karoor, H. Paul, A. Ruoho, M.P. Czech, C.C. Malbon, J. Biol. Chem. 271 (2) (1996) 1061–1064. [95] V. Karoor, K. Baltensperger, H. Paul, M.P. Czech, C.C. Malbon, J. Biol. Chem. 270 (43) (1995) 25305–25308. [96] S.K. Moon, S.H. Cho, K.W. Kim, J.H. Jeon, J.H. Ko, B.Y. Kim, C.H. Kim, Biochem. Biophys. Res. Commun. 356 (3) (2007) 542–547. [97] A. Rusciani, L. Duca, H. Sartelet, A. Chatron-Colliet, H. Bobichon, D. Ploton, R. Le Naour, S. Blaise, L. Martiny, L. Debelle, PLoS One 5 (11) (2010) e14010.

A novel insulin receptor-signaling platform and its link to insulin resistance and type 2 diabetes.

Insulin-induced insulin receptor (IR) tyrosine kinase activation and insulin cell survival responses have been reported to be under the regulation of ...
2MB Sizes 0 Downloads 3 Views