| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Original Studies |
Department of Medicine (C.K., H.U.H., S.M.), University of Tübingen, D 72076 Tübingen, Germany; the Department of Medicine, University of Heidelberg (A.H.), Heidelberg, Germany; the Departments of Medicine (P.A.) and Pathology (A.N.), University of Hamburg, Hamburg, Germany; the Diabetes Research Institute, University of Dusseldorf (I.U., J.E.), Düsseldorf, Germany; the Department of Medicine, University of Cologne (D.M.W.), Cologne, Germany; the Department of Pharmacology, University of Aachen (H.G.J.), Aachen, Germany; the Department of Diabetes and Metabolism, Krankenhaus Bethanien (M.D.), Hamburg, Germany; and the Department of Medicine, University of Vienna (H.W.R.), Vienna, Austria
Address all correspondence and requests for reprints to: Stephan Matthaei, M.D., Department of Medicine IV, Endocrinology, Metabolism, and Pathobiochemistry, University of Tübingen, Otfried Müller Strasse 10, D 72076 Tübingen, Germany. E-mail: snmattha{at}med.uni-tuebingen.de
 |
Abstract |
|---|
 Top Abstract IntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
In fibroblasts derived from VH, GLUT1 glucose transporters were
overexpressed by 8-fold in plasma membranes (PM) and by 5-fold in high
density microsomes, respectively. Exofacial photolabeling revealed that
only 14% of the overexpressed PM-GLUT1 transporters were properly
inserted into the plasma membrane. The complementary DNA structure of
the patient’s insulin receptor and the GLUT1 glucose transporter, the
intrinsic activity of plasma membrane glucose transporters, the
tyrosine phosphorylation, as well as the protein expression of insulin
receptor substrate-1/2 and p85 
/ß- and p110 /ß-subunits of PI
3-kinase were normal. However, insulin-stimulated association of the
p85 subunit of PI 3-kinase was defective in fibroblasts derived from VH
compared to those from controls, and this defect was associated with a
reduced IRS-1-dependent activation of PI 3-kinase by 50.2% and 63.6%
after incubation for 5 and 10 min with 100 nmol/L insulin,
respectively. Furthermore, immunodetection of small GTP-binding Rab
proteins in subcellular membrane fractions indicated a decreased
expression of Rab4 in total cellular homogenates as well as in high
density microsomes by 70% and 58%, respectively. After preparation of
GLUT1-containing vesicles, Rab4 was not detected to be a component of
these vesicles. Analysis of the PI 3-kinase in GLUT1-containing
membrane vesicles revealed insulin-dependent targeting of the p85
subunit to the vesicles immunoadsorbed from VH and control fibroblasts.
Importantly, the association of the p85 subunit as well as the
p85-immunoprecipitable PI 3-kinase activity were markedly reduced in
GLUT1-vesicles derived from the patient.
In conclusion, impaired PI 3-kinase activity in GLUT1-containing membrane vesicles derived from fibroblasts of VH is associated with a defective docking and/or fusion process of glucose transporters with the plasma membrane and thus might contribute to the molecular defect causing insulin resistance in this patient.
|
Introduction |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
After insulin stimulation and autophosphorylation of the insulin receptor, the major intracellular substrates of the activated receptor tyrosine kinase are the insulin receptor substrates-1 (IRS-1) to -4 (19, 20, 21). The central role of IRS-1 as a signal-transducing protein is based on its 22 potential tyrosine phosphorylation sites (22, 23). YMXM and homologous YXXM phosphotyrosine motifs on IRS-1 serve as docking sites for additional signaling proteins with SH2 (Src homology 2) domains after insulin stimulation (15, 22, 23, 24). Such a protein, which contains two SH2 domains in its p85 regulatory subunit, is the phosphatidylinositol (PI) 3-kinase. After insulin stimulation, binding of p85 to IRS-1 leads to activation of the p110 catalytic subunit (25, 26, 27). Several lines of evidence suggest that the IRS-1-dependent activation of PI 3-kinase plays an essential role in the insulin-mediated stimulation of glucose transport by promoting glucose transporter translocation to the plasma membrane (28, 29, 30, 31, 32, 33). Although significant advances have been made in under-standing the insulin signaling cascade from the insulin receptor to the glucose transport system since the original observation by Cushman and Wardzala (34) and Suzuki and Kono (35) of glucose transporter translocation, the exact molecular sequence causing translocation, docking, and fusion of transporter proteins with the plasma membrane remains unresolved (for reviews, see Refs. 3, 4, 36).
The current study was undertaken to gain further insight into the mechanisms of glucose transporter translocation and insertion of GLUT1-containing vesicles into the plasma membranes of fibroblasts derived from a patient with a genetic syndrome of insulin resistance. Clinically, the patient shows features of Werner syndrome with lipodystrophy, scleroderma-like alterations of the skin, alterations of the skeleton, and contractures of joints. Furthermore, the patient has insulin-resistant diabetes mellitus, with fasting insulin levels about 10-fold above normal values. The fasting blood glucose levels are also increased by approximately 2-fold. In a previous report it was shown that insulin resistance in this patient is associated with a defective insulin-stimulated glucose transport, whereas insulin binding to the receptor was normal (37).
We now report the results of studies that further examined the molecular cause of the defective insulin-stimulated glucose transport in fibroblasts of this patient. The data obtained provide evidence that the defective glucose transport is associated with impaired insulin-stimulated association of the p85 subunit of PI 3-kinase with IRS-1, which may lead to reduced activity of PI 3-kinase in GLUT1-containing membrane vesicles.
|
Subjects and Methods |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Insulin-resistant patient. The patient VH was born in 1965 of a nonconsanguineous married couple. Both parents are in good health and phenotypically normal. The patient is the oldest of three affected siblings, who all show clinical features of Werner syndrome. An extensive description of the clinical characteristics and laboratory data has been previously published (37). The patient has a body mass index of 19.5 kg/m2. Laboratory examinations revealed hyperinsulinemia and diabetes mellitus only in VH with a fasting insulin level of 102 µU/mL (5–10 µU/mL) and a fasting blood glucose level of 193 mg/dL (60–110 mg/dL).
Normal controls. Cultured fibroblasts obtained from normal weight males (n = 7) with a negative family history of type 2 diabetes were used as controls. Informed consent was obtained from all subjects tested.
Chemicals and antibodies
Cell culture medium, supplements, and reagents were purchased
from Life Technologies, Inc. (Gaithersburg, MD). Human
insulin (Actrapid) was obtained from Novo Nordisk (Mainz, Germany).
Avian myoblastosis virus reverse transcriptase and
Taq-DNA-polymerase were obtained from Pharmacia Biotech (Uppsala, Sweden) and Roche Molecular Biochemicals (Mannheim, Germany). Sequence-specific
5'/3'-oligonucleotide primers were synthesized by TIB Molbiol (Berlin,
Germany). Buffer components were purchased from Sigma (St.
Louis, MO) or Merck & Co., Inc. (Darmstadt, Germany)
unless otherwise stated. Reagents for SDS-PAGE were obtained from
Bio-Rad Laboratories, Inc. (Hercules, CA). Protein A/G
immobilized to agarose beads was purchased from Pierce Chemical Co. (Rockford, IL). [125I]Protein A was
supplied by NEN Life Science Products-DuPont (Boston, MA);
[-32P]deoxy (d)-CTP and
[-35S]dATP were supplied by Amersham Pharmacia Biotech (Aylesbury, UK);
[
-32P]ATP was supplied by ICN Biochemicals, Inc. (Costa Mesa, CA), respectively. A polyclonal
rabbit antiserum against the C-terminus of rat brain GLUT1 was a
product of Diagnostic International (Karlsdorf, Germany). Polyclonal
anti-IRS-1/2 antibodies and monoclonal/polyclonal anti-p85 antibodies
corresponding to the N-terminal SH2 domain as well as the full-length
85-kDa subunit of rat PI 3-kinase were obtained from Upstate Biotechnology, Inc. (Lake Placid, NY). Polyclonal
antiphosphotyrosine (PY20) as well as monoclonal anti-p85 antibodies
were obtained from Transduction Laboratories, Inc.
(Lexington, KY); a PI 3-kinase anti-p85ß antibody was obtained from
Serotec (Oxford, UK). Anti-p110/p110ß antibodies as
well as all isoform-specific Rab antibodies were purchased from
Santa Cruz Biotechnology, Inc. (Santa Cruz, CA).
Cell culture
Fibroblast cultures were grown from skin biopsies and maintained in DMEM (4.5 mg/mL glucose) supplemented with 10% FBS, 100 IU/mL penicillin, 100 µg/mL streptomycin, 100 µg/mL neomycin, 2 mmol/L glutamine, and 20 mmol/L HEPES. The cultures were kept at 37 C in 5% CO2 and 95% humidity and were stored in liquid nitrogen after the fifth to seventh passage. A sample was thawed for each set of experiments.
Labeling of GLUT1 messenger RNA (mRNA)
Total cellular RNA was extracted from cultured fibroblasts by a modification of the method described by Chomczynski et al. (38) using the RNAzol protocol of Cinna/Biotecx-Laboratories (Houston, TX). For Northern blot analysis 10–15 µg total RNA were separated on a 1% agarose gel containing approximately 2% formaldehyde and blotted onto Magna-Graph/nylon membranes (Micron Separations, Inc., Westboro, MA) (39). After transfer, blots were probed with a GLUT1 complementary DNA (cDNA). 32P Labeling of the DNA template was performed using the Random Primed DNA Labeling Kit from Roche Molecular Biochemicals. The hybridization and washing procedures were previously described (39). After autoradiography at -80 C, labeled bands were quantitated by scanning densitometry.
RNA RT, PCR amplification, and GLUT1 cDNA sequencing
Total RNA extracted from fibroblasts was used as template for
making a complementary copy DNA (cDNA) according to standard
methodology (40). The reaction was performed in 20 µL RT mixture
containing 1–5 µg total cellular RNA, 1 x RT buffer (10
x RT buffer = 500 mmol/L KCl, 1% gelatin, 20 mmol/L
MgCl2, and 100 mmol/L Tris-HCl, pH 8.3), 4 mmol/L
MgCl2, 1 mmol/L of each dNTP, and 0.1–1 µmol/L
22-mer RT primer complementary to the 3'-coding region. After
denaturation at 65 C for 5 min 0.5 µL 100 mmol/L dithiothreitol, 0.5
µL 40 U/µL RNasin, and 1.2 µL 17 U/µL avian myoblastosis virus
reverse transcriptase were added. The reaction mix was incubated for 50
min at 42 C, heated to 95 C for 5 min, and chilled on ice. PCR was
directly performed with 10 µL RT mixture in a final volume of 100
µL 1 x RT buffer buffer containing 250 µmol/L dNTPs, 0.1–2
µmol/L sequence-specific oligonucleotide primer pairs (18- to 25-mer,
one biotinylated/one nonbiotinylated), and 0.025 U/µL Thermus
aquaticus DNA polymerase (Taq-DNA-polymerase). After an
initial denaturation at 95 C for 5 min, the samples were subjected to
28–32 amplification cycles under the following conditions:
denaturation at 94 C for 1 min, annealing at 55 C for 1 min, and
extension at 72 C for 2 min. cDNA templates were prepared using
Dynabeads M-280 streptavidin (DynAl, Oslo, Norway). The
nucleotide sequences of the resulting single stranded DNAs were
analyzed by the dideoxy chain termination method as described by Sanger
et al. (41) according to the T7
sequencing kit protocol of Pharmacia Biotech, using
GLUT1-specific 5'/3'-oligonucleotide primers (18- to 24-mer) and
[-35S]dATP. The resulting DNA fragments were
separated on a 8% denaturing polyacrylamide gel containing 8 mol/L
urea and visualized by autoradiography.
Preparation of cellular fractions
The subcellular fractions were prepared either from basal or insulin-stimulated confluent grown fibroblasts. Cells were serum starved for 16 h before treatment with 100 nmol/L insulin for the times indicated.
Membrane fractions. Total cellular as well as subcellular membrane fractions were prepared as recently described (42).
Total cellular lysates. After incubation of fibroblasts with or without insulin, monolayers were washed twice with phosphate-buffered saline (PBS) containing 1 mmol/L CaCl2, 1 mmol/L MgCl2, and 100 µmol/L Na3VO4 on ice. Cells were immediately lysed in ice-cold lysis buffer composed of 50 mmol/L HEPES (pH 7.5), 137 mmol/L NaCl, 1 mmol/L MgCl2, 1 mmol/L CaCl2, 2 mmol/L Na3VO4, 10 mmol/L Na4P2O7, 10 mmol/L NaF, 2 mmol/L ethylenediamine tetraacetate (EDTA), 10% glycerol, 2 µg/mL aprotinin, 10 µg/mL antipain, 5 µg/mL leupeptin, 0.5 µg/mL pepstatin, 1.5 mg/mL benzamidine, 34 µg/mL phenylmethylsulfonylfluoride (PMSF), and 1% Nonidet P-40 (Calbiochem, San Diego, CA). The lysis buffer was described by Folli et al. (43). After incubation for 45 min at 4 C while rotating end over end, insoluble material was removed by centrifugation at 16,000 x g for 10 min at 4 C. The resulting supernatant was either used immediately for immunoblotting or used for immunoprecipitation after determination of protein content.
Analytical procedures. 5'-Nucleotidase and rotenone-insensitive NADH-cytochrome c reductase were used as marker enzymes for plasma membranes and high density microsomes. Enzyme assays in the homogenate and in each membrane fraction were previously described by Avruch and Hoelzl-Wallach (44) and Dallner et al. (45), respectively.
The protein content of cellular fractions were determined by a modification of the Bradford method using the Bio-Rad Laboratories, Inc., protein reagent and BSA as standard.
Immunohistochemical localization of GLUT1 proteins
Fibroblasts were seeded at a density of 2 x 104 cells/mL onto glass slides. After preincubation for 1–2 days, cells were fixed with 4% buffered formaldehyde. Subsequently, cells were permeabilized by the addition of PBS containing 0.01% Triton X-100. Endogenous peroxidase was blocked with 1% H2O2 diluted in CH3OH. The GLUT1 glucose transporter proteins were detected according to the protocol of the standard ABC kit (Vector Laboratories, Inc., Burlingame, CA) using an anti-C-terminal-GLUT1 antiserum.
Determination of intrinsic activity of plasma membrane glucose transporters
After reconstitution of glucose transporter proteins into artificial liposomes by sonication, the intrinsic activity of the plasma membrane GLUT1 transporter proteins was determined as described by Suzuki and Kono (46).
ATB-BMPA labeling
2-N-[4-(1-Azi-2,2,2-trifluoroethyl)benzoyl]-1,3-bis-
(D-mannose-4-yloxy)-2-propylamine (ATB-BMPA)
and ATB-[2-3H]BMPA (SA, 
10 Ci/mmol) were
prepared as reported by Clark and Holman (47). Surface photolabeling of
GLUT1 glucose transporters was carried out as described in detail
previously (48).
Immunoadsorption of GLUT1-containing membrane vesicles
For the isolation of GLUT1-enriched membrane vesicles, a method was developed according to the procedure for immunoadsorption of GLUT4 vesicles as previously described (49) with minor modifications. Briefly, 200-1000 µg basal or insulin-stimulated cellular membrane fractions were resuspended in ice-cold buffer A (PBS, pH 7.4, added with 0.1 mmol/L PMSF, 2.6 mmol/L dithiothreitol, 1 mmol/L EDTA, and 0.4% BSA). The resulting membrane suspension was precleared with protein G-agarose beads for 16 h at 4 C. After centrifugation, anti-GLUT1-antiserum was added to the supernatant, and the mixture was further incubated for 6 h at 4 C while rotating end over end. The resulting antigen-antibody complex was pelleted, resuspended in buffer A, and incubated with the immunobeads for an additional 16 h at 4 C. After centrifugation the GLUT1-containing beads were washed four times with PBS, pH 7.4, eluted in Laemmli sample buffer, and subjected to SDS-PAGE. Nonspecific adsorption was monitored by identical treatment with a nonimmune antiserum.
Measurement of PI 3-kinase activity in GLUT1 vesicles
After serum starvation for 16 h, fibroblasts were incubated with 100 nmol/L insulin for the times indicated. All of the following steps were carried out at 4 C. Cell monolayers were washed twice with PBS supplemented with 100 µmol/L Na3VO4 and scraped in TES buffer, pH 7.4, containing 1 mmol/L PMSF, 2 mmol/L Na3VO4, 10 mmol/L Na4P2O7, and 10 mmol/L NaF. Total cellular membranes were prepared and resuspended in PBS, pH 7.4, supplemented with the protease/phosphatase inhibitors as described above. Total membrane proteins (500–1000 µg) were subjected to immunoadsorption of GLUT1-vesicles, except that buffer A was supplemented with 2 mmol/L Na3VO4, 10 mmol/L Na4P2O7, and 10 mmol/L NaF. The resulting protein G-vesicle complex was washed four times with PBS, pH 7.4, added with 1 mmol/L PMSF, 2 mmol/L Na3VO4, 10 mmol/L Na4P2O7, and 10 mmol/L NaF again. Beads were either eluted in Laemmli sample buffer and subjected to immunoblotting with p85 PI 3-kinase antibodies or were solubilized in lysis buffer B [50 mmol/L HEPES (pH 7.5), 137 mmol/L NaCl, 1 mmol/L MgCl2, 1 mmol/L CaCl2, 2 mmol/L Na3VO4, 10 mmol/L Na4P2O7, 10 mmol/L NaF, 2 mmol/L EDTA, 10% glycerol, 2 µg/mL aprotinin, 10 µg/mL antipain, 5 µg/mL leupeptin, 0.5 µg/mL pepstatin, 1.5 mg/mL benzamidine, 1 mmol/L PMSF, and 2% Nonidet P-40) and incubated for 2–4 h while rotating end over end. After preclearing, a polyclonal anti-p85 PI 3-kinase antibody was added to the supernatant, and incubation was continued for 16 h. The resulting immunoprecipitate was linked to protein A/G beads for an additional 2 h. After centrifugation, the bead pellet was washed and resuspended in 10 mmol/L Tris (pH 7.5), 100 mmol/L NaCl, and 1 mmol/L EDTA as described for the PI 3-kinase assay. Subsequently, PI 3-kinase activity was determined as outlined below.
Immunoprecipitation
After insulin incubation, fibroblasts were washed twice with ice-cold PBS containing 1 mmol/L CaCl2, 1 mmol/L MgCl2, and 100 µmol/L Na3VO4; lysed; and precleared in the same matter as described for the preparation of total cellular lysates. The protein content of the supernatant was determined, and 500-1500 µg were used for immunoprecipitation with either anti-p85 PI 3-kinase or anti-IRS-1 polyclonal antibodies. After an overnight incubation with antibodies at 4 C, protein A/G was added to the lysates and allowed to bind for an additional 2 h at 4 C. The pelleted conjugates were washed either with lysis buffer or (for PI 3-kinase assay) under the following conditions: three times in PBS, 1% Nonidet P-40, and 100 µmol/L Na3VO4; three times in 100 mmol/L Tris (pH 7.5), 500 mmol/L LiCl, and 100 µmol/L Na3VO4; and twice in 10 mmol/L Tris (pH 7.5), 100 mmol/L NaCl, 1 mmol/L EDTA, and 100 µmol/L Na3VO4. For the PI 3-kinase assay, the precipitates were resuspended in 10 mmol/L Tris (pH 7.5), 100 mmol/L NaCl, and 1 mmol/L EDTA.
PI 3-kinase assay
PI 3-kinase activity was measured by in vitro
phosphorylation of PI according to the protocol described by Ruderman
(50) and Folli et al. (43) using
L--phosphatidylinositol (Avanti, Alabaster,
AL) as a substrate and 20 µCi/tube
[-32P]ATP (4500 Ci/mmol). Unspecific PI
3-kinase reaction was controlled by the addition of 100 nmol/L
wortmannin. The radiolabeled lipid products were separated by TLC on
20 x 20-cm silica gel TLC plates (Merck & Co.,
Darmstadt, Germany) and visualized by autoradiography. The PI 3-P spots
were quantitated by scanning densitometry (Easy Win 32, Herolab,
Germany).
Gel electrophoresis and immunoblotting
Equal amounts of protein samples (50–100 µg/200 µg for
detection of IRS-1 and phosphotyrosines) were solubilized in denaturing
sample buffer, heated in a boiling water bath for 5 min, and subjected
to SDS-PAGE (8–12.5% resolving gel) according to the method of
Laemmli (51). Separated proteins were transferred to nitrocellulose
membranes by electroblotting as described by Towbin et al.
(52). To enhance the elution of high molecular mass proteins, 0.02%
SDS was added to the transfer buffer. The membranes were blocked with
PBS, pH 7.4, containing 4% BSA for 1 h at 20 C and immediately
incubated in the same buffer with the indicated primary antibodies for
2 h at room temperature. After washing (five times, 5 min each
time) in PBS, pH 7.4, added with 0.5% BSA and 0.2% Tween-20,
nitrocellulose sheets were incubated in blocking buffer with
[125I]protein A (0.5 µCi/mL) for 1 h at
20 C. Blots were extensively washed again in PBS, 0.5% BSA, and 0.2%
Tween-20 six times for approximately 10 min each time, and
immunocomplexes were visualized by autoradiography. For quantification,
nitrocellulose pieces corresponding to the specific molecular weight of
immunodetected proteins were excised and measured in a -counter for
60 s.
Statistical analysis
The significance of differences (patient vs. controls) for normally distributed data was assessed by Student’s t test for unpaired samples.
|
Results |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Insulin receptor cDNA sequencing
Analysis of the cDNA sequence encoding for the insulin receptor revealed no mutation in VH fibroblasts (data not shown).
Characterization of the GLUT1 glucose transport system in fibroblasts from patient VH: GLUT1 glucose transporter gene and protein expression
To investigate whether the defective insulin-stimulated glucose
uptake was due to a decreased GLUT1 glucose transporter gene
expression, total cellular RNA was isolated from VH and control
fibroblasts and hybridized with a 32P-labeled
GLUT1 cDNA. As depicted in Fig. 1
, GLUT1
mRNA was increased by 6.3-fold in fibroblasts from VH compared to that
in seven healthy controls (10,723 ± 669 vs. 1,695
± 758 scanning units; P < 0.001).
|
Localization of GLUT1 glucose transporters in subcellular membrane fractions
To examine whether the defective glucose uptake in fibroblasts from the patient was associated with a different subcellular distribution of GLUT1 glucose transporters, total cellular homogenate (HOM), PM, as well as high density microsomes (HDM) and low density microsomes (LDM) were prepared. The purity of the membrane fractions and the degree of intermembraneous contamination were determined by measuring 5'-nucleotidase as well as rotenone-insensitive NADH-cytochrome c reductase. Plasma membranes were enriched by 25.6-fold in 5'-nucleotidase activity, whereas HDM and LDM showed 1.8- and 0.6-fold of the activity in total homogenate. Cytochrome c reductase activity was increased by 3-fold in HDM compared to that in HOM.
The subcellular distribution of GLUT1 glucose transporters in VH and
control fibroblasts was analyzed by immunodetection using an anti-GLUT1
antiserum followed by incubation with
[125I]protein A. A representative autoradiogram
is shown in Fig. 2A. GLUT1 was detected
as a broad protein band with a molecular mass of approximately 46 kDa.
Quantification (Fig. 2B) of the corresponding GLUT1 signals revealed an
increase in GLUT1 transporter proteins by 8-fold in plasma membranes as
well as by 5-fold in high density microsomes and 2.5-fold in low
density microsomes in fibroblasts derived from VH compared to those
from controls (P < 0.001).
|
In fibroblasts derived from the patient, immunohistochemistry
studies showed an accumulation of GLUT1 glucose transporters in the
plasma membrane as well as in the perinuclear endoplasmatic reticulum
compared to that in control cells (Fig. 3), thus confirming the data obtained by
immunodetection.
|
To examine the possibility that a structural GLUT1 defect is
responsible for the defective insulin-stimulated glucose transport with
consecutive up-regulated GLUT1 gene expression, the entire GLUT1 cDNA
of VH was sequenced as outlined in Materials and
Methods. Apart from two silent polymorphisms in codon 15
(Ala-GCT
GCC) and 476 (Asp-GATGAC), a substitution of
phenylalanine to leucine in codon 152 (Phe/TTTLeu/CTT) was found
compared to the published GLUT1 cDNA sequence (53) (Fig. 4). This amino acid sequence was also
detected in the two other siblings, both parents, and five healthy
controls.
|
The intrinsic activity of glucose transporters at the plasma membrane was determined after reconstitution of glucose transporters into artificial liposomes by sonication according to the method originally described by Suzuki and Kono (46). Measurement of glucose uptake into these liposomes revealed that the intrinsic activity of GLUT1 plasma membrane glucose transporters derived from patient VH was not significantly different compared to the control value (data not shown).
Exofacial photolabeling of GLUT1 with ATB-BMPA
To examine whether the proper insertion of GLUT1 glucose
transporters into the plasma membrane of VH fibroblasts is defective,
fibroblasts from VH and controls were photoaffinity labeled using
ATB-BMPA. This plasma membrane-impermeant bis-D-mannose
derivative can be used as a marker for identifying and quantifying the
proportion of glucose transporter proteins functionally proper exposed
at the plasma membrane, as it only labels those transporters that are
present at the cell surface of intact cells (47). The data presented in
Fig. 5 indicate that in the patient’s
fibroblasts only 14% of the 8-fold overexpressed GLUT1 glucose
transporters were properly inserted into the plasma membrane.
|
To investigate whether the defective insulin-stimulated glucose transport could be due to defects intrinsic to the insulin signaling cascade, we analyzed components of the transduction pathway from the insulin receptor through the IRS-1-dependent PI 3-kinase to the glucose transport system in cells derived from VH compared to control cells. As insulin receptor binding, affinity, and cDNA structure were normal, we focused our studies on the major receptor substrates, IRS-1 and IRS-2, as well as on PI 3-kinase.
Immunoblotting with anti-IRS-1/2, antiphosphotyrosine, and anti-PI 3-kinase antibodies
Experiments were performed to determine whether a difference in
the protein expression of these signaling elements or defective
tyrosine phosphorylation is responsible for the impaired glucose
transport detected in VH cells. Thus, total cellular lysates were
prepared from basal or insulin-stimulated (100 nmol/L) fibroblasts,
subjected to SDS-PAGE and analyzed by immunoblotting with antibodies
against IRS-1/2, phosphotyrosine, or the p85 and p110 subunits of PI
3-kinase. To summarize, the total amount of IRS-1/2, insulin-stimulated
tyrosine phosphorylation, as well as the protein levels of the p85
/ß- and p110 /ß-subunits of the PI 3-kinase in
fibroblasts of VH were normal compared to those in control cells (data
not shown). Furthermore, the expression of the PI 3-kinase in total
cellular homogenates as well as in subcellular membrane fractions (Fig. 9A) immunodetected with a monoclonal anti-p85 antibody was not
significantly different in cells derived from VH compared to that in
control cells. However, in the patient’s cells insulin did not
significantly increase IRS-1/p85 association, whereas in control cells
insulin induced a 2.9-fold increase in the amount of IRS-1/p85
precipitates (Fig. 6).
|
|
PI 3-kinase activity was measured by in vitro
phosphorylation of PI as described in Materials and Methods.
To determine the total PI 3-kinase activity in fibroblasts derived from
VH and controls, cells were incubated for 0, 5, and 10 min with 100
nmol/L insulin. The PI 3-kinase assay using PI as substrate was
performed after immunoprecipitation of total cellular lysates with
polyclonal anti-p85 PI 3-kinase antibodies in the absence or presence
of 100 nmol/L wortmannin. The results indicate that the total
p85-immunoprecipitable PI 3-kinase activity was not different in VH
compared to control fibroblasts (data not shown). In contrast, as
represented in Fig. 7, after
immunoprecipitation with anti-IRS-1, the PI 3-kinase activity was
reduced in total cellular lysates isolated from fibroblasts of VH
compared to control cells. A typical autoradiogram of TLC-separated
lipid products is shown in Fig. 7A. Whereas the basal PI 3-kinase
activity was not significantly different, the quantification of
32P-labeled PI 3-phosphate spots from five
independent experiments revealed a reduction in the IRS-1-dependent
enzyme activity by 50.2% and 63.6% after stimulation with 100 nmol/L
insulin for 5 and 10 min (P < 0.05), respectively
(Fig. 7B). The IRS-2-immunoprecipitable PI 3-kinase activity was
unaffected in cells derived from the patient.
|
To examine a possible defective vesicular docking and/or fusion of glucose transporters at the plasma membrane, GLUT1-containing membrane vesicles were prepared and characterized from VH and control fibroblasts.
Immunoadsorption of GLUT1 vesicles from human fibroblasts
GLUT1 vesicles were prepared from total cellular membranes,
microsomal (HDM/LDM) membrane fractions, or PM isolated from VH and
control cells using a polyclonal anti-C-terminal-GLUT1 antiserum and
protein G-agarose beads as outlined in Materials and
Methods. Nonspecific adsorption was monitored by identical
treatment, except that the anti-GLUT1 antiserum was replaced by a
nonimmune polyclonal antiserum. SDS eluates of the immunobeads were
resolved by SDS-PAGE on a 10% gel, transferred to nitrocellulose, and
immunoblotted with anti-GLUT1 antiserum followed by incubation with
[125I]protein A. Figure 8 illustrates uniform and distinct GLUT1
signals (+) in the range of about 46 kDa corresponding to the
postulated molecular mass of GLUT1 glucose transporter proteins. In
contrast, there are only faint bands in the (-)-control precipitates
immunoadsorbed with an unspecific antiserum. These results suggest that
this method is suitable for the preparation of GLUT1-containing
membrane vesicles.
|
To determine whether the PI 3-kinase is directly associated with
the glucose transporter-containing membrane vesicles, GLUT1 vesicles
were prepared from fibroblasts of VH and controls and analyzed by
immunoblotting either with monoclonal or polyclonal anti-p85 PI
3-kinase antibodies. The resulting bands with a molecular mass of 85
kDa, shown in Fig. 9B
, suggest an
association of the p85 subunit of PI 3-kinase with vesicles derived
from VH and control cells. In contrast to the equivalent subcellular
distribution (Fig. 9A), the p85 PI 3-kinase was markedly reduced in
GLUT1 vesicles derived from plasma membranes as well as microsomal
membranes of VH fibroblasts. Furthermore, as depicted in Fig. 10A, the data demonstrate an
insulin-dependent targeting of the p85 PI 3-kinase subunit to GLUT1
vesicles immunoadsorbed from total cellular membrane fractions after
incubation with 100 nmol/L insulin for 10 min. Figure 10B represents
differences in the p85 association with the GLUT1 vesicles prepared
from VH compared to control cells. Summarizing the results from four
independent experiments, the PI 3-kinase association was 50% (5 min,
100 nmol/L insulin) and 30.5% (10 min, 100 nmol/L insulin),
respectively, of the values determined in control fibroblasts
(P < 0.05). These data also suggest that the p85 PI
3-kinase binds specifically to GLUT1 vesicles, as no immunoreactivity
was detectable in the (-)-control precipitates. Although there is high
variability in the association of p85 with GLUT1 vesicles, as
demonstrated in Fig. 10A, the differences between VH and control cells
are significant.
|
Next we examined whether PI 3-kinase activity is detectable in
GLUT1-enriched membrane vesicles derived from human fibroblasts. Thus,
GLUT1 vesicles were prepared from basal and insulin-stimulated total
cellular membrane fractions, solubilized in a Nonidet P-40-containing
lysis buffer, immunoprecipitated with polyclonal anti p-85 PI
3-kinase antibodies, and subjected to PI 3-kinase assay. The
labeled PI 3-phosphate products were visualized by autoradiography. The
results presented in Fig. 11 indicate
an accumulation of 3'-phosphoinositides in GLUT1 vesicles derived from
control cells after stimulation with 100 nmol/L insulin for 10 min. In
contrast, only a very small amount of reaction products was seen in
vesicles prepared from VH fibroblasts. The presence of 100 nmol/L
wortmannin, a specific inhibitor of PI 3-kinase, completely abolished
the enzyme reaction, suggesting the specificity of detected PI 3-P
spots.
|
Low molecular mass GTPases of 20–30 kDa are known to be involved in the regulation of intracellular vesicle-trafficking processes, including glucose transporter translocation. Thus, we analyzed the expression and vesicle association of Rab proteins in cells derived from the patient compared to those in control cells.
Immunodetection of Rabs in total cellular HOM revealed that Rab1, Rab4,
Rab5, Rab6, and Rab8 are expressed in human fibroblasts (Fig. 12A). In contrast to the other
immunodetected Rab proteins, which show equal amounts in VH and control
fibroblasts, the protein level of Rab4 was decreased by 70% in total
cellular homogenates of cells derived from the patient. After
preparation of subcellular membrane fractions, Rab4 was decreased in
HDM by 58% in the VH fibroblasts (P < 0.001; n =
3). A typical autoradiogram is shown in Fig. 12B.
|
|
Discussion |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
To examine whether the insulin resistance in this patient is the result of alterations in the glucose transport system, we first analyzed the GLUT1 glucose transporter mRNA and protein. In contrast to our assumption that the defective glucose transport in fibroblasts from VH is associated with a reduced or unaltered GLUT1 gene and/or protein expression, we detected a marked overexpression of GLUT1 transporter proteins in the plasma membrane as well as in the perinuclear endoplasmatic reticulum. A posttranslational GLUT1 modification is unlikely to be present, as no detectable shift was seen in the electrophoretic mobility on SDS-PAGE, although due to the GLUT1 overexpression this cannot be completely excluded. Furthermore, the functional defect of the insulin-stimulated glucose transport in fibroblasts of VH cannot be accounted for by variations in the GLUT1-coding sequence or in the intrinsic activity of glucose transporters. As determined by the plasma membrane-impermeable reagent ATB-BMPA, which binds only to the exofacial glucose transporter domains (47), there is a discordance between the amount of immunodetected plasma membrane GLUT1 proteins and those that are functionally correct at the extracellular surface in fibroblasts derived from the patient. Thus, although GLUT1 glucose transporters are overexpressed by 8-fold in plasma membranes derived from fibroblasts of VH, those transporter proteins are not functionally active when inserted into the plasma membrane and consequently are not capable of transporting glucose into the cell.
To investigate the molecular cause of the defective insulin-stimulated
glucose transport accompanied by a defective incorporation of GLUT1
into the plasma membrane, we next analyzed the insulin signal
transduction pathway leading to activation of PI 3-kinase as well as
the possibility that PI 3-kinase could be a component of glucose
transporter vesicles. Importantly, the results presented in Fig. 6
demonstrate that in cells from the patient insulin failed to
significantly increase IRS-1/p85 association, whereas in control cells
insulin caused a 2.9-fold increase in the IRS-1/p85 association.
Furthermore, as depicted in Fig. 10, insulin-stimulated association of
the p85 subunit of PI 3-kinase with GLUT1-containing membrane vesicles
as well as PI 3-kinase activity in GLUT1-containing membrane vesicles
(Fig. 11) is reduced in fibroblasts from VH. Although the molecular
signaling sequence from the insulin receptor to the glucose transport
system is not completely understood, there is an increasing set of
evidence suggesting that the IRS-1-dependent activation of PI 3-kinase
plays an essential role in the insulin stimulation of glucose uptake
(4, 14, 18, 36). In this pathway, PI 3-kinase seems to be involved in
both insulin signal transduction as well as vesicular trafficking. The
evidence that PI 3-kinase is necessary for insulin action on glucose
transport is partially based on experiments using wortmannin and
LY294002, which specifically inhibits the catalytic activity of this
enzyme. It has been shown that these inhibitors also reduce
insulin-stimulated GLUT4 and GLUT1 translocation in rat adipocytes,
3T3-L1 cells, and Chinese hamster ovary (CHO) cells (16, 36). In
addition, a mutant of the -p85 PI 3-kinase subunit (
p85) lacking
the binding site for the p110 catalytic subunit disrupted the
translocation of GLUT1 to the plasma membrane in transfected CHO cells
(54). On the other hand, the p110 catalytic subunit of PI 3-kinase is
homologous to the yeast VPS34, which has been shown to be required for
protein sorting and vesicle targeting (55). This sequence similarity
support the idea that PI 3-kinase plays a pivotal role in mammalian
vesicular glucose transporter trafficking.
A major finding of this study is that due to deficient insulin-induced
IRS-1/p85 association, PI 3-kinase activity is markedly reduced in
cells from the patient VH. It is tempting to speculate that the
consequently impaired insulin-stimulated p85-dependent PI 3-kinase
activation in GLUT1-containing vesicles could contribute to the
defective insulin-stimulated glucose transport observed in the
patient’s fibroblasts. O’Rahilly et al. (56) described a
markedly reduced insulin-stimulated PI 3-kinase activity in cultured
fibroblasts from pseudoacromegalic patients. However, in those patients
these researchers showed that the association of p85 with IRS-1 was
not significantly different compared to that in control cells.
It is generally accepted that the activation of PI 3-kinase by insulin alone is not sufficient for its specific function in glucose transporter recycling processes. Many studies have shown that PDGF and other receptor systems also stimulate PI 3-kinase activity, but not glucose transporter translocation from tubulo-vesicular trans-Golgi compartments to the plasma membrane. The molecular mechanism by which PI 3-kinase promotes insulin stimulation of glucose transport is presently unknown. Kelly and Ruderman first showed that activation of PI 3-kinase by insulin involves a tyrosine-phosphorylated IRS-1/PI 3-kinase complex in a low density microsomal membrane fraction (57). Some recent experimental results support the hypothesis that after insulin stimulation the PI 3-kinase migrates to specific intracellular compartments and that this membrane localization is mediated by IRS-1 or other insulin receptor substrates. Hypothetically, from these intracellular membrane fractions PI 3-kinase and its 3'-phosphorylated lipid products might enhance the membrane movement to the cell surface or facilitate docking and fusion with the plasma membrane (36).
The data presented in this report suggest an association of the p85 subunit of PI 3-kinase with the GLUT1-containing membrane vesicles isolated from VH and control fibroblasts and show that this association was markedly reduced in the patient cells. Recently, Heller-Harrison et al. reported that insulin directs the association of PI 3-kinase with GLUT4 vesicles (58). However, to the best of our knowledge, the current study represents the first demonstration of an association of PI 3-kinase with GLUT1-containing vesicles. Furthermore, as previously demonstrated in GLUT4 vesicles, our data demonstrate an insulin-dependent targeting of the p85 subunit of PI 3-kinase to the GLUT1-containing membrane vesicles. The measurement of PI 3-kinase activity revealed an accumulation of 3'-phosphoinositides in GLUT1 vesicles isolated from total cellular membrane fractions of the control fibroblasts, but not in vesicles derived from VH. These results support the idea that one function of PI 3-kinase in the mechanism of insulin-stimulated glucose transport may be based on its association with glucose transporter-containing membrane vesicles and that this process requires sufficient tyrosine-phosphorylated IRS-1/PI 3-kinase com-plexes.
The defective insulin-stimulated IRS-1/p85 association in the patient’s cells, which may lead to the impaired PI 3-kinase activation in GLUT1-containing vesicles, could contribute to the defective insulin-stimulated glucose transport in fibroblasts from VH. The exact molecular sequence of vesicular glucose transporter targeting as well as their docking and fusion with the plasma membrane is currently unknown. Several lines of evidence suggest the involvement of synaptic vesicular proteins to be essential for these mechanisms (3, 4, 36). To analyze whether a defect at this level could explain the diminished plasma membrane insertion of glucose transporter proteins in fibroblasts of VH, we examined the expression of synaptobrevin-2 (VAMP-2), SNAP-25, syntaxin, synaptophysin, synaptotagmin, and Munc18/nSEC-1 in total cellular homogenates as well as in plasma membrane fractions prepared from VH and control fibroblasts. Preliminary data indicate that the v-SNARE protein VAMP-2, the t-SNARE protein SNAP25, nSec-1, as well as synaptotagmin 1, which may play a regulatory role, are expressed in human fibroblasts. We identified two of these proteins, VAMP-2 and nSec-1, in GLUT1-containing vesicles isolated from either plasma membranes or total cellular membrane fractions. Preliminary data indicate no detectable differences in these vesicle proteins in the patient’s cells compared to those in control cells. However, further experiments are necessary to characterize the protein composition of GLUT1 vesicles in more detail.
To examine whether G protein-mediated regulation of intracellular GLUT1 vesicle trafficking (59) could be defective, we examined small GTP-binding Rab proteins in human fibroblasts. Interestingly, the results suggest that Rab4, which seems to be involved in insulin-stimulated GLUT4 glucose transporter translocation, is decreased in HDM isolated from the VH cells. In contrast to Cormont et al., who observed an association of Rab4 in GLUT4 vesicles (60), this guanosine triphosphatase as well as the other immunodetected isoforms were not found to be a component of the GLUT1-enriched membrane vesicles. Further studies are necessary to elucidate the functional significance of the different subcellular Rab4 expressions for the impaired glucose uptake in cells derived from this patient.
In conclusion, impaired PI 3-kinase activity in GLUT1-containing membrane vesicles derived from fibroblasts of VH is associated with a defective docking and/or fusion process of glucose transporters with the plasma membrane and thus might contribute to the molecular defect causing the defective insulin-stimulated glucose transport in fibroblasts of this patient. Further study is necessary to analyze whether the molecular defect causing the defective plasma membrane insertion of GLUT1 might also be present in the process of docking and fusion of GLUT4 vesicles with the plasma membrane and T tubules in skeletal muscle of this patient, producing his insulin-resistant state. In addition, it remains to be examined whether defects in the process of functionally proper insertion of glucose transporters into cell surface membranes might contribute to the defective skeletal muscle glucose transport typically observed in patients in more common insulin-resistant diseases such as type 2 diabetes mellitus.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
Received March 3, 1999.
Revised August 12, 1999.
Accepted October 27, 1999.
|
References |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| 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 |
