| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Discovery Research Laboratories II (H.T., I.K., H.H., N.K., K.Y., S.K., Y.F.), Pharmaceutical Research Division, Takeda Chemical Industries Co., Ltd., Tsukuba, Ibaraki 300-4293, Japan; and Department of Biochemistry (H.T., P.F.P.), Boston University School of Medicine, and AdipoGenix, Inc. (E.P., C.C.B., P.F.P.), Boston, Massachusetts 02118
Address all correspondence and requests for reprints to: Hideaki Tojo, Ph.D., Discovery Research Laboratories II, Pharmaceutical Research Division, Takeda Chemical Industries Co., Ltd., 10 Wadai, Tsukuba, Ibaraki 300-4293, Japan. E-mail: Tojo_Hideaki{at}takeda.co.jp.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
As noted above, IRAP has been shown to colocalize with GLUT4 intracellularly, and these proteins move together to the cell surface in response to insulin. Although there are some conflicting data concerning whether IRAP internalizes identically with GLUT4 (11, 12), in the basal state, IRAP is located in an intracellular compartment with GLUT4, and these proteins move to the cell surface identically in response to insulin (3, 4, 5, 6, 7, 8). There is evidence that the N terminus of IRAP, which contains two dileucine motifs and several acidic regions similar to those in GLUT4, functions in the regulation of the intracellular retention and trafficking of GLUT4 vesicles (13). In the cited study, injection of the amino-terminal cytoplasmic domain of IRAP into 3T3-L1 adipocytes caused GLUT4 translocation. A similar effect was observed when the C-terminal cytoplasmic domain of GLUT4 was introduced into adipocytes (14). Furthermore, a chimeric protein containing the intracellular domain of IRAP and the extracellular and transmembrane domains of the transferrin receptor (vpTR) displayed IRAP-like trafficking in 3T3-L1 adipocytes (15). Taken together, these studies provide support for the notion that major GLUT4 vesicle cargo proteins such as IRAP interact with cytosolic/cytoskeletal proteins in a fashion that may be regulated by insulin. Indeed, several recent studies have supported a role for the cytoskeleton in GLUT4 vesicle exocytosis (16, 17, 18, 19, 20, 21, 22). Therefore, a need exists for determining the mechanism of action of IRAP, for identifying novel proteins that interact with IRAP, and for identifying modulators of such molecules for use in regulating a variety of insulin-sensitive cellular responses.
Based on these thoughts, we have performed a two-hybrid screen using the cytoplasmic domain of IRAP as bait, and we have identified one of the formin family of proteins, FHOS (formin homolog overexpressed in spleen) (23), as an IRAP-binding protein. Furthermore, we have found that FHOS has two novel splice variants, including a skeletal muscle-specific isoform. We also observed that FHOS has a distinct affinity for various members of the profilin family. Finally, adenovirus-mediated overexpression of FHOS enhances glucose uptake in L6 myocyte cells. Our data suggest that FHOS acts as a linker protein between GLUT4-containing vesicles and the cytoskeleton.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
). Because preliminary examination revealed that IRAP-(1–109) was a self-activating bait showing significant background in the yeast two-hybrid system, we used IRAP-(55–82) bait throughout this study. Both human skeletal muscle and mouse 3T3-L1 prey cDNA libraries were screened. From the cotransformants grown on minimum media lacking histidine, the reporter lacZ-positive colonies were selected by colony lift and an X-Gal staining assay. Six clones from the differentiated 3T3-L1 library and five from the human skeletal muscle library were obtained as positives of the yeast two-hybrid screen. Among them, one of the positive clones (designated MD36) from the human skeletal muscle cDNA library showed the strongest ß-galactosidase activity measured by a quantitative enzymatic assay. Sequence analysis revealed that MD36 was a formin family protein, FHOS, previously isolated from B cells as an acute myeloid leukemia 1B transcription factor-associated protein (23). Histidine prototrophy and ß-galactosidase activity of the yeast clone harboring both the IRAP and FHOS are shown in Fig. 1, B and C.
|
). In addition, competitive inhibition of this protein-protein interaction was observed. The interaction of IRAP and FHOS was inhibited in a concentration-dependent manner by addition of chemically synthesized IRAP peptide (55–82) (Fig. 1E).
Mammalian Two-Hybrid Assay
To further confirm the protein-protein interaction of IRAP and FHOS, we examined this interaction using the mammalian two-hybrid system. Transient expression experiments using both bait and prey vectors revealed that IRAP-(1–109) and IRAP-(55–82) showed similar affinities for the FHOS C-terminal region (Fig. 1F). IRAP-(1–109) showed no background in the mammalian system, although this sequence displayed significant background in the yeast two-hybrid system. The reporter activities of the two IRAP-baits (1–109 and 55–82) and FHOS were approximately half of that of the p53/SV40-T antigen control interaction.
cDNA Cloning of FHOS and Novel Splicing Variants
To obtain the full-length cDNA of FHOS, we used PCR and primers based on the reported FHOS sequence (23). In the process of PCR cloning, we found that there were at least two novel splicing variants of FHOS. One splicing variant was found in PCR products derived from human skeletal muscle cDNAs and contained a 78-bp (26-amino-acid) insertion. This sequence was inserted between Lys-440 and Ala-441 in the reported FHOS sequence (Fig. 2, A and B). Another truncated-type splicing variant was found in human spleen-derived cDNA. The truncated FHOS possessed a 24-bp insertion containing an in-frame stop codon at the same point that the 78-bp insertion was found. The alternative splicing is summarized in Fig. 2A, and the amino acid sequence of FHOS78 is shown in Fig. 2B.
|
). To verify the alternative splicing of FHOS in human RNA, we examined the tissue distribution of FHOS splicing variants by RT-PCR. As described in Materials and Methods, PCR to detect the FHOS, FHOS24, and FHOS78 transcripts produced 310-, 334-, and 388-bp DNA fragments, respectively. Both FHOS and FHOS24 were detected in all of the examined tissues except for skeletal muscle (Fig. 3B). FHOS78 was detected mainly in skeletal muscle and faintly in heart, and FHOS24 was barely detectable in these tissues. Adipose is another insulin-responsive tissue; however, little expression of FHOS78 was observed.
|
), consistent with its role as a possible tether for GLUT4 vesicles. There is no apparent increase in FHOS in the plasma membrane fraction (PM) due to insulin in rat adipocytes. We also were able to detect a small amount of FHOS in IRAP immunoprecipitates (data not shown).
|
). Profilin IIb mRNA was also widely expressed, but not in leukocyte or skeletal muscle. Profilin IIa was detected mainly in brain, pancreas, placenta, and skeletal muscle. Adipocytes also express profilin IIa; however, the level of transcript was much lower than that of skeletal muscle (Fig. 5B).
|
). These results suggested that the C-terminal 10 amino acids in profilin IIa were essential for binding to both FHOS variants, and that the muscle-specific 78-bp insertion in FHOS78 was not required for the protein-protein interaction with profilin proteins.
|
and 8A). The L6-GLUT4 cell line, an L6 cell line stably expressing GLUT4, was used for this study, and the insulin effect on transport is due to the presence of this transporter isoform (25). There was no effect on basal or insulin-stimulated glucose uptake when the FHOS splicing variants were overexpressed in 3T3-L1 adipocytes (Fig. 7). In L6-GLUT4 cells, the overexpression of FHOS and FHOS78 had a stimulatory effect on basal glucose transport and also enhanced insulin-induced glucose transport activity. The truncated FHOS (FHOS24), which does not contain the IRAP-binding domain (IRAP-BD), caused no enhancement (Fig. 8B). Wortmannin treatment abolished the ability of insulin to stimulate glucose transport, as expected. The enhancement of basal transport by FHOS expression was unaffected by wortmannin, suggesting that FHOS functions downstream of phosphatidylinositol-3 (PI3)-kinase or in a PI3-kinase-independent manner (Fig. 8C).
|
|
). The overexpression of FHOS-CT enhanced glucose uptake in the basal state; however, there was no effect on insulin-stimulated glucose uptake.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
), and by coimmunoprecipitation in cells (data not shown).
FHOS has sequence homology to Diaphanous and formin proteins within the FH1 and FH2 domains (23, 30). FHOS also contains a coiled-coil domain, a collagen-like domain, two nuclear localization signals, and several potential protein kinase A and C phosphorylation sites (23). In this study, we identified two novel alternatively spliced variants of FHOS, FHOS24 and FHOS78. The former is the ubiquitously expressed truncated form (24-bp insertion), and the other is a skeletal muscle-specific form (78-bp insertion). We think the 78-bp insertion is not implicated specifically in insulin-stimulated GLUT4 translocation because FHOS78 is not detected in adipose cells (Fig. 3), and the effect for glucose uptake by FHOS78 overexpression was similar to that of FHOS overexpression in L6-GLUT4 cells (Fig. 8). The truncated variant, FHOS24, did not show any effects on glucose uptake upon overexpression in L6-GLUT4 cells (Fig. 8), suggesting that the N terminus of FHOS does not interact with IRAP.
FHOS contains a proline-rich domain, FH1, which binds the actin-binding protein, profilin. This domain is also found in other known formin family proteins (30). It has been suggested that actin reorganization is involved in the insulin-dependent relocalization of GLUT4 from its intracellular storage site to the cell surface and the subsequent stimulation of glucose uptake (16, 17, 18, 19, 20, 21, 22). Indeed, treatment with cytochalasin D or latrunculin B, which causes actin filament depolymerization, prevents insulin-stimulated glucose transport in 3T3-L1 adipocytes and L6 myotubes (17, 18, 19, 21, 31). Furthermore, it was shown by immunofluorescence that IRAP and GLUT4 concentrate with the actin-rich structures in L6 myotubes and that these are colocalized with PI3-kinase after insulin stimulation (18). It was also shown in these studies that IRAP is concentrated in the detergent-soluble fraction despite its colocalization with actin as detected by immunofluorescence analysis, suggesting that the interaction of IRAP with the cytoskeleton is not direct but is mediated by an unknown actin-binding protein. Taking these findings together with our observation that FHOS has an affinity for profilin IIa as well as with IRAP (Figs. 1 and 6), we hypothesize that FHOS may make an actin-binding complex together with profilin IIa and that these interactions may play an important role in determining the subcellular localization and/or trafficking of IRAP and GLUT4.
Proteins that bind to profilin I, IIa, and IIb with different functional consequences have also been characterized (32, 33, 34). The components of the profilin I complex in brain lysates are clathrin, valosine-containing protein, heat shock protein 70, tubulin, and actin. Proteins included in the profilin IIb complex include dynamin I, hematopoietic protein-2, synapsin IA/B, synapsin IIA/B, Rho-associated coiled-coil kinase, an unknown protein named "a partner of profilin," heat shock protein70, and actin. The affinity of profilin I and IIb for the proline-rich sequence in vasodilator-stimulated phosphoprotein is much lower than the affinity of profilin IIa. Di Nardo et al. (34) showed that profilin IIa, but not profilin IIb, binds to vasodilator-stimulated phosphoprotein; however, their results regarding the affinity for actin are not consistent with other reports (32, 33). Although it is still uncertain whether or not each of these ligand proteins directly binds profilin, these results strongly suggest that each profilin isoform has distinct ligands. It has been reported that both the N- and C-terminal regions of profilin are essential for binding to the poly-L-proline region (35). Based on this, it was suggested that the distinct profile of profilin-binding proteins results from sequence differences in both terminal regions in profilins I, IIa, and IIb. Differences in the sequences between profilin IIa and IIb occur in the C-terminal 32 amino acids, a region that is critical for binding to the proline-rich domain. Indeed, profilin IIp, which is composed of amino acids 1–130 of profilin IIa, did not bind FHOS (Fig. 5).
Recently, it was reported that FHOS also interacts with the Rho family GTPase Rac1 (28). It is known that Rac1 mediates insulin-dependent membrane ruffling in many cell types (36). In addition, dominant negative Rac1 inhibits actin reorganization as well as GLUT4 translocation to the plasma membrane after insulin stimulation (18). Therefore, it is hypothesized that the FHOS/Rac1 interaction may be involved in the mechanism of GLUT4 translocation.
We have shown that overexpression of FHOS significantly enhanced glucose transport in L6-GLUT4 cells (Fig. 8). Overexpression of FHOS may provide additional sites for GLUT4/IRAP vesicles to bind to the cytoskeleton, which then permits additional vesicles to translocate from the intracellular compartment to the plasma membrane upon insulin stimulation. This enhanced glucose uptake caused by FHOS overexpression was not affected by inhibiting PI-3 kinase with wortmannin. Against expectation, no effect on glucose transport in FHOS-overexpressing 3T3-L1 adipocytes was observed. One reason for this result may be the lower level of profilin IIa expression in adipocytes (Fig. 5). Considering that the FHOS/profilin IIa complex may function in trafficking of GLUT4 vesicles, the endogenous profilin IIa protein in adipocytes may be limiting when FHOS is overexpressed. Alternatively, the distinct results in the two cell types may result from other molecular events regulating membrane trafficking. For example, insulin regulates GLUT4 translocation in both fat and muscle, but contraction supplies an additional signal for translocation in the latter tissue that is insulin independent (37).
Overexpression of the FHOS C-terminal IRAP-BD enhanced basal glucose uptake in L6-GLUT4 cells (Fig. 9), consistent with competitive inhibition of endogenous FHOS/IRAP binding (Fig. 1E), and then dissociation of intracellular GLUT4 vesicles from the cytoskeleton. These results are similar to those obtained by injection of the cytoplasmic domain of IRAP (13). It is also possible that the enhancement of glucose uptake by FHOS C-terminal overexpression could result from inhibiting endocytosis of GLUT4/IRAP, and further studies are needed to distinguish between these possible mechanisms.
There are several reports identifying other nonintegral membrane proteins that are associated with GLUT4 vesicles. Tankyrase, a Golgi-associated MAPK substrate, also interacts with the cytoplasmic domain of IRAP (38). However, tankyrase and FHOS may play different roles as IRAP-binding proteins because tankyrase bound to IRAP-(96–101) whereas FHOS bound to IRAP-(55–82). The IRAP cytoplasmic sequences sufficient to translocate GLUT4 vesicles to the cell surface are IRAP-(55–82) or IRAP-(1–52), neither of which contains the tankyrase-binding site. It has been suggested that tankyrase may be involved in the long-term effect of the MAPK cascade on the metabolism of GLUT4 vesicles; however, no evidence was presented for tankyrase function (38). This is in contrast to the present study in which we show an effect of FHOS on transport in Figs. 8 and 9. It has been also proposed that a GLUT4-binding protein functions as a retention protein for GLUT4 vesicles (14). A number of GLUT4-binding proteins, e.g. a myosin-derived partial peptide, aldolase, carboxyl esterase (p62/CE), and Daxx, have been reported to interact with the transporter (39, 40, 41, 42). It has been suggested that, due to its enzymatic activity, p62/CE may play a role in GLUT4 vesicle budding and/or fusion (41). A myosin-derived peptide may function like FHOS as a linker between GLUT4 vesicles and the cytoskeleton (39). In the case of aldolase (40) and Daxx (42), there were no functional consequences reported for their interaction with GLUT4, and it is unclear why an apoptosis-associated adaptor like Daxx would have any physiological role related to glucose transport. Moreover, there is no information about the affinity of these proteins for IRAP or distinct roles of the protein-protein interactions in trafficking of GLUT4 vesicles. It is known that there are several intracellular GLUT4 compartments that are dependent on the distinct stages from endocytosis to exocytosis (1, 2) and, therefore, there may be other proteins interacting with GLUT4 vesicles and involved in the translocation/trafficking pathway. In any case, we show that FHOS binds both IRAP and profilin IIa as illustrated in Fig. 10. Considering the evidence for cytoskeleton involvement in trafficking of GLUT4-containing vesicle, our results suggest that FHOS may play a role as a tethering protein binding GLUT4-containing vesicles to the cytoskeleton.
|
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
Yeast Two-Hybrid Screening to Isolate IRAP-Binding Protein
The human skeletal muscle and differentiated 3T3-L1 cDNA library fused to the yeast GAL4 transcription activation domain (GAL4-AD) was screened for proteins that interact with IRAP-(55–82). Library plasmids expressing interacting proteins were identified by their ability to induce transcription of yeast GAL1 promoter-controlled HIS3 and lacZ. The host yeast cell line used throughout this study was Saccharomyces cerevisiae Y190, carrying these two reporter genes on the chromosomal DNA. Positive clones were determined by histidine prototrophy on a triple dropout SD agar plate (-Trp, -Leu, -His), containing 40 mM of 3-amino-1,2,4,-triazole (3-AT), and colony-lift 5-bromo-4-cloro-3-indolyl-ß-D-galactopyranoside (X-Gal) staining, according to the manufacturer’s manual (CLONTECH Laboratories, Inc.). Both bait and prey plasmids of candidate clones were isolated from yeast cell lysates and were reintroduced into S. cerevisiae Y190 to confirm that the reconstructed double transformants had their reporter activity. Consequently, false positives caused by spontaneous alternation of host yeast cells, prey products interacting directly with GAL4-BD, and those directly increasing GAL1 promoter activity were eliminated.
Quantitative ß-Galactosidase Assays
A chlorophenyl red-ß-D-galactopyranoside (CPRG) assay was performed for quantitative ß-galactosidase analysis of two-hybrid positive clones. Cultured yeast cells having both bait and positive prey plasmids were broken by repeated freeze/thaw cycles using liquid nitrogen. After CPRG was added to the broken cell suspension, ß-galactosidase activity was determined by measuring OD at 578 nm. One unit of ß-galactosidase activity was defined as the amount that hydrolyzes 1 µmol of CPRG to chlorophenol red and D-galactoside per min per cell.
Full-Length cDNA Cloning of FHOS and the Splicing Variants
The full-length cDNA was isolated by PCR based on the reported FHOS sequence (22). The primer sequences for PCR cloning were 5'-TGAGCCGGCCGCAGAGCCATGG-3' (sense primer) and 5'-TGCTCCGTGCGTTCAAGGAGCTCAC-3' (antisense primer). First-strand cDNAs from human spleen and skeletal muscle were synthesized by avian myeloblastosis virus reverse transcriptase (Takara, Shiga, Japan) from mRNA from each tissue (CLONTECH Laboratories, Inc.) and were used as PCR templates. The enzyme used for this PCR was LA-Taq polymerase with GC-buffer (Takara). Thermal cycling parameters were 35 cycles of 98 C for 20 sec, 65 C for 40 sec, and 72 C for 3.5 min. The DNA sequence was analyzed using the ABI prism BigDye sequencing system (Perkin-Elmer Corp., Norwalk, CT). The mutations derived from misreading of Taq polymerase were determined by comparing more than three independent clones from the same PCR bands, and correct cDNA was constructed to the exclusion of PCR mutations using restriction enzyme digestion and ligation.
Northern Hybridization Analysis for FHOS
The human multiple tissue blot membrane (MTN-blot) was purchased from CLONTECH Laboratories, Inc. The 0.68 kb of EcoRI fragment, corresponding to amino acids 993-1190 in Fig. 2B, was labeled with 
-32P-dCTP by random priming and hybridized to the membrane using ExpressHyb solution (CLONTECH Laboratories, Inc.).
RT-PCR to Detect Alternative Splicing of FHOS
Normalized first-strand cDNAs from various human tissues were used for RT-PCR (MTC-Panel; CLONTECH Laboratories, Inc.). Primers to detect alternative splicing of FHOS were as follows: sense, 5'-CCTACCATCTCTGTGGCACCCTCAGCT-3'; antisense, 5'-TTGGGGCTTGCTGGTATCAGTGGCTCC-3'. PCR conditions were 25 cycles of 98 C for 20 sec, 65 C for 40 sec, and 72 C for 60 sec using LA-Taq polymerase with GC buffer (Takara). PCR products were cloned into pCR2.1 cloning vector (Invitrogen, San Diego, CA) and their DNA sequences were confirmed.
cDNA cloning of Human Profilin Family
Profilin I, IIa, and IIb cDNAs were obtained by PCR using primers that amplified each coding region of human profilin I (GenBank accession no. J03191) or IIb (GenBank accession no. L10628). The primers for the profilin I PCR were 5'-ATGGCCGGGTGGAACGCCTACATCGAC-3' (PFN1, 27-mer) and 5'-TCAGTACTGGGAACGCCGAAGGTGGGA-3' (PFN2, 27-mer), and those for the profilin II PCR were 5'-ATGGCCGGTTGGCAGAGCTACGTGGAT-3' (PFN3, 27-mer) and 5'-TTACACATCAGACCTCCTCAGGTATAAAGC-3' (PFN4, 30-mer). Using these primers, each cDNA was amplified from human skeletal muscle cDNA (CLONTECH Laboratories, Inc.) using Pfupolymerase (Stratagene, La Jolla, CA). Cycling parameters for amplification were 95 C for 30 sec, 60 C for 45 sec, 72 C for 60 sec, for 35 cycles. The resulting PCR products were incubated with Taq polymerase at 72 C for 10 min to add dA to the 3'-end, and were then cloned into the TA-cloning vector, pCR2.1, after which their sequences were determined.
Tissue Distribution of Human Profilin Family mRNA
The MTC-panel was used for analyzing the tissue distribution of profilin family mRNAs. PCR detecting the expression of profilin I and both profilin IIa and IIb was performed as described in the procedure for cDNA cloning. For PCR detecting profilin IIa but not IIb, a specific primer for IIa (5'-ACACAAGTTCATACCCCATCACCC-3') was used with the primer PFN3.
Yeast Two-Hybrid Assay for Detecting Protein-Protein Interaction Between FHOS and Profilin Family Members
The protein-protein interaction between FHOS and each profilin isoform was determined using the yeast two-hybrid system. The bait vectors were the GAL4-BD cDNA alone or GAL4-BD-fused profilin I, IIa, or IIb cDNAs, respectively. For the prey vectors, full-length FHOS or FHOS78 or N-terminal sequences corresponding to amino acid residues 1–624 of FHOS or 1–650 of FHOS78 were fused with GAL4-AD. Each sequence contained the FH1 domain and the upstream sequence. Using each set of bait and prey plasmids, S. cerevisiase Y190 was transformed, and the resulting double transformants were streaked on a triple dropout SD agar plate (-Trp, -Leu, -His) containing 40 mM of 3-AT.
Cell Culture
3T3-L1 preadipocytes were obtained from IFO (Institute for Fermentation, Osaka, Japan). Cells were maintained and differentiated into adipocytes as described previously (24). L6-GLUT4 (L6 overexpressing GLUT4) cells (25) were obtained from Dr. D. E. James (University of Queensland, Australia).
Antibodies
Affinity-purified rabbit polyclonal anti-FHOS antibody was generated using a diphtheria toxin-conjugated synthetic peptide, H-RERKRSRGNRKSLRRC-NH2, corresponding to amino acids 1152–1166 of the FHOS protein. Mouse monoclonal anti-GLUT4 (1F8) and rabbit polyclonal anti-IRAP antibodies were described previously (3, 26).
Construction of Recombinant FHOS Expression Viral Vectors
An adenovirus vector expressing FHOS was generated by cloning three types of full-length FHOS cDNAs or the C-terminal IRAP-BD of FHOS (FHOS-CT) into a cosmid vector, pAxCAwt (Takara), which contains the adenovirus type-5 (Ad5) genome, the CAG expression unit [chicken ß-actin promoter, cytomegarovirus enhancer, and the rabbit ß-globin poly(A) additional signal] at the locus substituted for E1. The recombinant cosmid vectors were cotransfected into human HEK-293 cells with DNA-terminal protein complex as described (27). Adenovirus vectors were propagated, purified by CsCl banding, and titered by a standard procedure. An FHOS-encoding virus was confirmed by restriction enzyme digestion of viral DNA and immunoblot analysis of lysates from infected L6-GLUT4 and 3T3-L1 cells. L6-GLUT4 cells or 3T3-L1 adipocytes at d 13–15 were infected 2 h with the purified recombinant adenovirus with multiplicity of infection 200 (L6-GLUT4) or 1,000 (3T3-L1 adipocytes) in serum-free DMEM.
Assay of 2-Deoxy-D-Glucose Transport
3T3-L1 adipocytes at d 13–15 or confluent L6-GLUT4 cells were serum starved for 4 h in DMEM supplement with 0.01% fatty acid-free BSA (Sigma, St. Louis, MO), followed by incubation in KRH buffer (20 mM HEPES, 136 mM NaCl, 4.7 mM KCl, 1.25 mM MgSO4, 1.25 mM CaCl2, pH 7.4) containing 0.1% BSA for 30 min at 37 C. Cells were incubated for 15 min in the presence of 0–1 µM insulin, the final 5 min of which was with 0.1 mM 2-deoxy-D-[1, 2-3H] glucose (1 µCi/ml) in KRH buffer. After stimulation, uptake was stopped by washing cells three times with ice-cold PBS. Cells were solubilized in PBS containing 0.1% sodium dodecyl sulfate (SDS) and the incorporated radioactivity was measured by a scintillation counter.
Expression Vectors and Recombinant Proteins for in Vitro Binding Assay
The Escherichia coli expression vector pGEX-MD36N4 contains the FHOS C-terminal region (FHOS-CT, amino acids 993-1190 in Fig. 2B) followed by glutathione-S-transferase and is under the control of the tac promoter. The parental vector is the pGEX expression vector (AP Biotech, Buckinghamshire, UK). The protease recognition site located at the junction of the fusion protein was eliminated. E. coli BL21 carrying pGEX-MD36N4 was cultured, expression was induced with isopropyl-ß-D-thiogalactopyranoside, and cells were harvested. Cells were disrupted by sonication and the lysate was centrifuged. Most of the recombinant FHOS-CT protein was found in the pellet. Subsequently, inclusion bodies were resuspended in PBS containing 8 M urea and 5 mM dithiothreithol, and the resulting solution was dialyzed against 0.5 M arginine hydrochloride in PBS. pET21-IRAP1–109 contains a 6-amino-acid His-tag fused to the cDNA encoding the IRAP cytoplasmic domain (amino acids 1–109), and expression is driven by the T7 promoter (Novagen, Madison, WI). The recombinant protein [IRAP-(1–109)-His] was expressed in E. coli BL21 (DE3) and purified according to the manufacturer’s instructions.
In Vitro Binding Assay
Recombinant FHOS-CT protein (1 µg/ml) in sodium carbonate buffer [Na2CO3, 2.6 g/liter; NaHCO3, 2.1 g/liter (pH 9.6)] was adhered on 96-well EIA plates by overnight incubation at 4 C. After washing the plates three times with TBS-T (0.05% Tween 20 in PBS), 3% BSA in TBS-T was added to the wells, and then samples were incubated for 1 h at room temperature to block nonspecific binding. IRAP-(1–109)-His (0.1 µg/ml in TBS) was added to the MD36-coated wells, and binding was performed by incubation for 2 h at room temperature. After washing the plates three times, the remaining IRAP-(1–109)-His was quantified using anti-His mouse antibody (QIAGEN, Chatsworth, CA), horseradish peroxidase-conjugated antimouse IgG goat antibody, and the enhanced chemiluminescence system (NEN Life Science Products, Boston, MA). For the binding inhibition assay, biotinylated IRAP-(55–82) was chemically synthesized and added to FHOS-CT-coated wells with IRAP-(1–109)-His.
Mammalian Two-Hybrid Assay
The Mammalian Matchmaker two-hybrid assay kit (CLONTECH Laboratories, Inc.) was used with some modifications. To increase the sensitivity of reporter activity the chloramphenicol acetyltransferase (CAT) reporter gene in the purchased kit was changed to the firefly luciferase gene. In brief, the CAT open reading frame was eliminated from pG5CAT (CLONTECH Laboratories, Inc.) by PCR using Pfupolymerase (Stratagene). The primers used were 5'-AGGCGTTTAAGGGCACCAATAACTGCC-3' and 5'-GGCGAGATTTTCAGGAGCTAAGGAAGCT-3'. The luciferase-coding fragment, which was isolated from pGL3 (Promega Corp., Madison, WI) by PCR, was ligated into pG5CAT lacking the CAT gene. The resulting reporter vector pG5luc+ contained the luciferase gene downstream from five consensus GAL4-binding sites and the minimal promoter of the adenovirus E1b gene. To generate bait vectors, IRAP-(1–109) or IRAP-(55–82) was fused with the GAL4 DNA-BD under the control of the SV40 early promoter. A CHO cell line lacking the dhfr gene was used as a host cell line for the mammalian two-hybrid assay. Cells were cultured in Ham’s F10 for maintenance. For dhfr selection, cells were cultured in DMEM supplemented with 0.02 mg/ml of proline and 10% dialyzed fetal bovine serum. For neo selection, G418 (Invitrogen) was added to the dhfr selection media at a final concentration of 400 µg/ml. Transfection for transient expression was performed using FuGENE transfection reagent (Roche Diagnostics, Indianapolis, IN). CHO (dhfr-) cells were transfected with both pG5luc+ and the Chinese hamster dhfr gene, and stable transfectants carrying the luciferase reporter gene (CHO-luc) were selected from dhfr-positive CHO cells. The obtained cell lines were tested for their reporter activities by introducing pM3-VP16, which is a positive control vector expressing a fusion of the GAL4 DNA-BD to the VP16-AD. For CHO-luc cells cultured in 24-well plates, 0.8 µg each of plasmid DNA was used for transfection by FuGENE reagent (Roche Diagnostics). Luciferase activity was determined using a purchased luciferase assay system (Promega Corp.) at 2 d after transfection.
Subcellular Fractionation of Adipocytes
Subcellular fractions of rat adipocytes from epididymal fat pads of male Sprague Dawley rats or from 3T3-L1 adipocytes at d 12–15 were prepared as described previously (5). In brief, cells isolated by collagenase digestion (rat adipocytes) or harvested after serum starvation (3T3-L1) were homogenized using a Potter-Elvehjem Teflon pestle, and plasma membrane, heavy microsomes, light microsomes, and mitochondria/nuclei fractions were separated by differential centrifugation.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
Abbreviations: 3-AT, 3-Amino-1,2,4,-triazole; CAT, chloramphenicol acetyltransferase; CPRG, chlorophenyl red-ß-D-galactopyranoside; EIA, enzyme immunoassay; FH, formin homology; FHOS, formin homolog overexpressed in spleen; FHOS-CT, FHOS C-terminal protein; GAL4-AD, GAL4 activation domain; GAL4-BD, GAL4 DNA-binding domain; GLUT, glucose transporter; IRAP, insulin-responsive aminopeptidase; IRAP-BD, IRAP binding domain; LM, light microsome; PI3-kinase, phosphatidylinositol-3-kinase; PM, plasma membrane.
Received for publication February 16, 2003. Accepted for publication March 27, 2003.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
-induced insulin resistance in 3T3–L1 adipocytes is accompanied by a loss of insulin receptor substrate-1 and GLUT4 expression without a loss of insulin receptor-mediated signal transduction. J Biol Chem 272:971–976This article has been cited by other articles:
 |
|
 |
|
  G. R. Peck, S. Ye, V. Pham, R. N. Fernando, S. L. Macaulay, S. Y. Chai, and A. L. Albiston Interaction of the Akt Substrate, AS160, with the Glucose Transporter 4 Vesicle Marker Protein, Insulin-Regulated Aminopeptidase Mol. Endocrinol., October 1, 2006; 20(10): 2576 - 2583. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T. Hosaka, C. C. Brooks, E. Presman, S.-K. Kim, Z. Zhang, M. Breen, D. N. Gross, E. Sztul, and P. F. Pilch p115 Interacts with the GLUT4 Vesicle Protein, IRAP, and Plays a Critical Role in Insulin-stimulated GLUT4 Translocation Mol. Biol. Cell, June 1, 2005; 16(6): 2882 - 2890. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. B. Gill, J. Roecklein-Canfield, D. R. Sage, M. Zambela-Soediono, N. Longtine, M. Uknis, and J. D. Fingeroth EBV attachment stimulates FHOS/FHOD1 redistribution and co-aggregation with CD21: formin interactions with the cytoplasmic domain of human CD21 J. Cell Sci., June 1, 2004; 117(13): 2709 - 2720. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |
