(Received for publication, October 3, 1996, and in revised form, January 2, 1997)
From the Hypertension-Endocrine Branch, NHLBI, National Institutes of Health, Bethesda, Maryland 20892 and the Department of Metabolic Diseases, Hoffmann-LaRoche, Inc., Nutley, New Jersey 07110-1199
The protein-tyrosine phosphatases PTP1B and Syp have both been implicated as modulators of the mitogenic actions of insulin. However, the roles of these protein-tyrosine phosphatases in the metabolic actions of insulin are not well characterized. In this study, we directly assessed the ability of PTP1B and Syp to modulate insulin-stimulated translocation of the insulin-responsive glucose transporter GLUT4 in a physiologically relevant insulin target cell. Primary cultures of rat adipose cells were transiently transfected with either wild-type PTP1B (PTP1B-WT), wild-type Syp (Syp-WT), or the catalytically inactive mutants PTP1B-C/S or Syp-C/S. The effects of overexpression of these constructs on insulin-stimulated translocation of a co-transfected epitope-tagged GLUT4 were studied. Cells overexpressing either PTP1B-C/S or Syp-WT had insulin dose-response curves similar to those obtained with control cells expressing only epitope-tagged GLUT4. In contrast, for cells overexpressing PTP1B-WT the level of GLUT4 on the cell surface at each insulin dose (ranging from 0 to 60 nM) was significantly lower than that observed in the control cells. Interestingly, cells overexpressing the dominant inhibitory mutant Syp-C/S also had a small but statistically significant impairment in insulin responsiveness. At a maximally stimulating concentration of insulin (60 nM), cell surface epitope-tagged GLUT4 was approximately 20% less than that of the control cells. It is possible that effects from high level overexpression of Syp and PTP1B constructs may not reflect what occurs under physiological conditions. Nevertheless, our data raise the possibility that PTP1B may be a negative regulator of insulin-stimulated glucose transport, while Syp may have a small role as a positive mediator of the metabolic actions of insulin.
Insulin is an important regulator of growth and metabolism. The pleiotropic actions of insulin are initiated by the binding of insulin to its receptor and the resultant activation of intrinsic receptor tyrosine kinase activity (1). Because tyrosine kinase activity is central to insulin signaling, protein-tyrosine phosphatases (PTPases)1 may be important for modulating insulin signal transduction pathways (2). Although there is good evidence that PTPases regulate mitogenic actions of insulin, the roles of various PTPases in metabolic actions of insulin are not well characterized.
The ubiquitously expressed prototype nontransmembrane PTPase PTP1B was among the first PTPases to be identified, cloned, and characterized (3-8). PTP1B dephosphorylates the insulin receptor both in vitro and in intact cells (9-11). In addition, PTP1B regulates the mitogenic actions of insulin (12, 13). Interestingly, in tissue culture models an increase in the level and activity of PTP1B has been associated with insulin resistance induced by exposure to high glucose levels. In addition, the level and activity of PTP1B in human skeletal muscle is positively correlated with in vivo measures of insulin sensitivity (14-16).
Syp (also known as SH-PTP2, PTP1D, SHPTP3, or PTP2C) is a cytosolic PTPase containing two SH2 domains in addition to a catalytic phosphatase domain (17). Binding of the SH2 domains of Syp to phosphotyrosine motifs on either the insulin receptor or insulin receptor substrate-1 (IRS-1) results in activation of Syp PTPase activity (18, 19). Recently, a number of studies have shown that Syp participates in Ras and mitogen-activated protein kinase-dependent pathways as a positive mediator of mitogenic actions of insulin and other growth factors (20-23). In addition, Hausdorff et al. (24) have investigated the role of Syp in differentiated 3T3-L1 cells (tissue culture cells capable of differentiating into an adipocyte-like phenotype under appropriate conditions). They report that microinjection of either the SH2 domains of Syp or anti-Syp antibodies interfered with the mitogenic actions of insulin, but had no detectable effect on the insulin-stimulated translocation of the insulin-responsive glucose transporter GLUT4 (24).
One of the most important metabolic actions of insulin is to increase glucose transport in tissues such as muscle and fat by recruiting GLUT4 to the cell surface. Previously, we used a transient transfection system for rat adipose cells in primary culture to demonstrate roles for the insulin receptor tyrosine kinase, IRS-1, and phosphatidylinositol 3-kinase in the insulin-stimulated translocation of GLUT4 (25-28). In the present study, we used a similar approach to overexpress wild-type or catalytically inactive mutant forms of PTP1B or Syp to directly test the roles of these PTPases in modulating insulin-stimulated translocation of GLUT4 in a physiologically relevant insulin target cell. Our data suggest that PTP1B may function as a negative regulator of the metabolic actions of insulin, while Syp may mediate a small positive effect on the ability of insulin to recruit GLUT4 to the cell surface.
-An expression vector (pCIS2) that generates high expression levels in transfected rat adipose cells (25) was used as the parent vector for subsequent constructions.
The cDNA coding for human GLUT4 with the influenza hemagglutinin epitope (HA1) inserted in the first exofacial loop of GLUT4 was subcloned into pCIS2 (GLUT4-HA).
An XbaI/SmaI fragment containing the cDNA for human PTP1B (generous gift from Dr. Jonathan Chernoff) was ligated into XbaI/HpaI sites in the multiple cloning region of pCIS2 (PTP1B-WT).
An XbaI/SmaI fragment containing the cDNA for a catalytically inactive mutant PTP1B with a cysteine to serine substitution at position 215 (generous gift from Dr. Jonathan Chernoff) was ligated into XbaI/HpaI sites in the multiple cloning region of pCIS2 (PTP1B-C/S).
An XbaI/XhoI fragment containing the cDNA for human Syp (generous gift from Dr. Benjamin Neel) was ligated into the multiple cloning region of pCIS2 (Syp-WT).
An XhoI/DraI fragment containing the cDNA for a catalytically inactive mutant Syp with a cysteine to serine substitution at position 459 (generous gift from Dr. Benjamin Neel) was ligated into XhoI/HpaI sites in the multiple cloning region of pCIS2 (Syp-C/S).
Milligram quantities of the plasmid DNA vectors described above were obtained using a Magic Megaprep kit (Promega). The wild-type and mutant sequences in the catalytic domain of the respective PTP1B and Syp constructs were confirmed by direct sequencing.
Isolated Rat Adipose Cell PreparationIsolated adipose cells were prepared from the epididymal fat pads of male rats (170-200 g, CD strain, Charles River Breeding Laboratories, Wilmington, MA) by collagenase digestion as described (25, 29).
ElectroporationIsolated adipose cells were transfected by electroporation as described (25-28). Cells from multiple cuvettes were pooled to obtain the necessary volume of cells for each experiment. Table I shows the combinations and concentrations of plasmid DNA as well as the number of cuvettes used for each of the insulin dose-response experiments.
|
20 h after electroporation, adipose cells were processed as described (26-28) and treated with insulin at final concentrations of 0, 0.024, 0.072, 0.3, or 60 nM at 37 °C for 30 min. Cell surface epitope-tagged GLUT4 was determined by using the anti-HA1 mouse monoclonal antibody 12CA5 (Boehringer Mannheim) in conjunction with 125I-labeled sheep anti-mouse IgG as described (26-28). Cells transfected with the empty expression vector pCIS2 were used to determine nonspecific binding of the antibodies. Typically, the nonspecific binding was ~30% of the total binding to cells transfected with GLUT4-HA and maximally stimulated with insulin (26). The actual specific counts were comparable from experiment to experiment (see figure legends). The lipid weight from a 200-µl aliquot of cells was determined as described (30) and used to normalize the data for each sample.
Immunoblotting of PTP1B, Syp, and GLUT4-HAExpression of
recombinant PTP1B, Syp, or GLUT4-HA was confirmed by immunoblotting
extracts of cells that were prepared at the same time and had undergone
transfection in parallel with the cells used for the translocation
assay described above. Whole cell homogenates were prepared from cells
co-transfected with GLUT4-HA (2 µg/cuvette) and either pCIS2,
PTP1B-WT, PTP1B-C/S, Syp-WT, or Syp-C/S (4 µg/cuvette). Cells from 15 cuvettes were pooled for each group. After electroporation and
overnight incubation, the cells were washed once and resuspended in 3 ml of TES buffer (20 mM Tris, 1 mM EDTA, 8.73%
sucrose, 10 µg/ml leupeptin, 10 µg/ml aprotinin, 10 µg/ml soybean
trypsin inhibitor, 10 µg/ml bovine serum albumin, 1 mM
phenylmethylsulfonyl fluoride, pH 7.4, 18 °C). The cells were
homogenized by being passed through a 25-gauge needle three times and
then were centrifuged for 10 min at 400 × g, 4 °C,
to pellet nuclei. The fat cake and pellet were discarded. For detection
of PTP1B and GLUT4-HA, the total membrane fraction was isolated from
the whole cell homogenate by centrifuging for 30 min at 400,000 × g, 4 °C. The pellet containing the total membrane fraction was resuspended in 600 µl of TES buffer and stored at 70 °C until further processing.
For immunodetection of PTP1B constructs, aliquots of the total membrane fractions from each group containing equal amounts of protein (500 µg) were solubilized in Laemmli sample buffer and subjected to SDS-polyacrylamide gel electrophoresis. The contents of the gel were transferred to nitrocellulose, and the PTP1B protein was detected with a polyclonal anti-PTP1B antibody (Upstate Biotechnology, Inc., Lake Placid, NY) and visualized using an antibody against rabbit IgG in conjunction with an enhanced chemiluminescent detection system (ECL, Amersham).
For immunodetection of Syp constructs, EDTA and aprotinin (final concentrations of 2 mM and 1.3 µg/ml, respectively) were added to aliquots of whole cell homogenates containing equal amounts of protein (800 µg). The samples were centrifuged for 20 min at 17,000 × g, and the supernatant was precleared for 30 min at 4 °C by incubating with 20 µl of protein A-agarose (Bio-Rad) that had been prewashed in lysis buffer (phosphate-buffered saline with 1% Nonidet P-40, 2 mM EDTA, and 1.3 µg/ml aprotinin). The samples were then immunoprecipitated by incubating for 45 min at 4 °C with a human-specific polyclonal anti-Syp antibody (Santa Cruz Biotechnology Inc., Santa Cruz, CA) followed by incubation with 20 µl of prewashed protein A-agarose for 45 min at 4 °C. The immunoprecipitated samples were then washed three times in lysis buffer and eluted by adding 60 µl of Laemmli sample buffer and incubating at 98 °C for 5 min. Twenty µl of each sample was loaded per lane and subjected to SDS-polyacrylamide gel electrophoresis. The contents of the gel were transferred to nitrocellulose, and Syp protein was detected with a monoclonal anti-Syp antibody (Transduction Laboratories, Lexington, KY) and visualized using an antibody against mouse IgG in conjunction with an enhanced chemiluminescent detection system (ECL, Amersham).
To determine relative levels of GLUT4-HA in each group of transfected cells, total membrane fractions were prepared as above, and samples were immunoprecipitated with an anti-HA antibody followed by immunoblotting with an anti-GLUT4 antibody as described previously (28).
Assay for PTPase ActivityCells were transfected with the various PTPase constructs as described above. After incubation overnight, cells were stimulated with insulin (60 nM) for 2 min, and cell extracts were prepared exactly as described above for the immunoblotting experiments (except 150 mM NaCl and 1 mM dithiothreitol were added to the TES buffer). PTPase activity in the cell extracts was determined by measuring the hydrolysis of p-nitrophenyl phosphate (pNPP). Equal aliquots of the cell extracts (80 µg of total protein in 50 µl of TES buffer) were added to 450 µl of a reaction mixture containing 50 mM MES, 150 mM NaCl, 2.5 mM EDTA, 0.1% bovine serum albumin, 2 mM dithiothreitol, and 50 mM pNPP and incubated for 10 min at 30 °C. The reaction was stopped by the addition of 500 µl of 2 M KOH, and the amount of product (p-nitrophenyl) produced was measured by determining the absorbance at 405 nM in a spectrophotometer. Nonspecific absorbance was corrected for by subtracting the absorbance at 405 nM determined in the absence of cell extract.
Statistical AnalysisInsulin dose-response curves were compared using multivariate analysis of variance. Paired t tests were used to compare individual points where appropriate. p values of less than 0.05 were considered statistically significant. The insulin dose-response curves were fit to the equation y = a + b [x/(x + k)] using a Marquardt-Levenberg nonlinear least squares algorithm. When plotted on linear log axes, this equation gives a sigmoidal curve where the parameters are associated with the following properties: a = basal response, a + b = maximal response, k = half-maximal dose (ED50), and x = concentration of insulin.
To
directly evaluate the role of PTP1B in insulin-stimulated translocation
of GLUT4, we overexpressed either wild-type or catalytically inactive
mutant forms of human PTP1B in rat adipose cells. We confirmed that
PTP1B-WT and PTP1B-C/S were overexpressed at comparable levels in
our system by immunoblotting total membrane fractions isolated from
transfected cells with an anti-PTP1B antibody (Fig. 1).
In the lane containing cell extracts from control cells transfected
with the empty expression vector pCIS2, there is a faint band
representing endogenous rat PTP1B. The lanes containing extracts from
groups of cells transfected with either PTP1B-WT or PTP1B-C/S show that
the recombinant PTP1B constructs were expressed at levels that are much
higher than the endogenous rat PTP1B levels. Since only ~5% of the
adipose cells that have undergone electroporation are actually
transfected (26), we estimate that there was at least a 100-fold
overexpression of PTP1B-WT and PTP1B-C/S relative to endogenous PTP1B
in the transfected cells. We also tested the PTPase activity of the
recombinant PTP1B constructs by assessing the ability of cell extracts
from insulin-stimulated transfected cells to hydrolyze the substrate
pNPP. PTPase activity in extracts derived from the group of cells
transfected with PTP1B-WT was approximately 3 times higher than that of
the control cells transfected with the empty expression vector pCIS2,
consistent with a high level of overexpression in the 5% of
transfected cells. PTPase activity in extracts derived from the group
of cells transfected with PTP1B-C/S was not significantly different
from that of the control cells (data not shown).
After confirming overexpression of the recombinant PTP1B constructs in
transfected adipose cells, we next determined their effects on the
ability of insulin to recruit a co-transfected epitope-tagged GLUT4 to
the cell surface. The insulin dose-response curve for control cells
co-transfected with the empty expression vector pCIS2 and GLUT4-HA
showed a 2.5-fold increase in cell surface GLUT4-HA upon maximal
insulin stimulation (60 nM) with an ED50 of
0.06 nM. In contrast, the insulin dose-response curve for
cells overexpressing PTP1B-WT demonstrated both decreased sensitivity and decreased responsiveness to insulin (Fig.
2A). In the absence of insulin, the basal
level of cell surface GLUT4 for cells overexpressing PTP1B-WT was
approximately half of that seen in the basal state for the control
cells. Furthermore, at every insulin dose the amount of GLUT4 at the
cell surface of cells overexpressing PTP1B-WT was less than that of the
control cells (ED50 = 0.18 nM). At 60 nM insulin, the maximal insulin response for cells
overexpressing PTP1B-WT was approximately 80% of the maximal response
for the control cells. Thus, the insulin sensitivity of cells
overexpressing PTP1B-WT was decreased approximately 3-fold while the
insulin responsiveness was decreased by about 20%. When analyzed by
multivariate analysis of variance, the difference in the dose-response
curve caused by overexpression of PTP1B-WT was highly significant
(p < 1 × 1010). In contrast to
PTP1B-WT, we were unable to detect any significant difference in the
insulin-stimulated translocation of GLUT4-HA in cells overexpressing
the mutant PTP1B-C/S when compared with the control cells (Fig.
2B).
It is possible that the lower level of cell surface GLUT4-HA we
observed in cells overepxressing PTP1B-WT is due to an effect of
PTP1B-WT on expression of GLUT4-HA rather than to an effect on insulin
sensitivity or responsiveness. Therefore, we measured total GLUT4-HA in
cells co-transfected with pCIS2, PTP1B-WT, or PTP1B-C/S by
immunoprecipitating total membrane fractions with an anti-HA antibody
followed by immunoblotting with an anti-GLUT4 antibody (Fig.
3). We observed comparable levels of GLUT4-HA in all
groups. Taken together, our data suggest that overexpression of
PTP1B-WT has an inhibitory effect on insulin signal transduction pathways related to GLUT4 translocation.
Effects of Overexpression of Syp-WT or Syp-C/S
We confirmed
overexpression of Syp-WT and Syp-C/S by using a human-specific
polyclonal anti-Syp antibody to immunoprecipitate whole cell
homogenates of cells transfected with either the empty expression
vector pCIS2, Syp-WT, or Syp-C/S in conjunction with immunoblotting
with a monoclonal anti-Syp antibody (Fig. 4). The lanes
containing extracts from cells transfected with Syp-WT or Syp-C/S show
that the two recombinant Syp constructs are overexpressed at comparable
levels. A specific band representing endogenous rat Syp is not seen in
cells transfected with pCIS2 because the samples were
immunoprecipitated with a human-specific antibody. We tested the
PTPase activity of the recombinant Syp constructs by assessing the
ability of cell extracts from insulin-stimulated transfected cells to
hydrolyze the substrate pNPP. PTPase activity in extracts derived from
the group of cells transfected with Syp-WT was approximately 2 times
higher than that of the control cells transfected with the empty
expression vector pCIS2, consistent with a high level of overexpression
in the 5% of transfected cells. PTPase activity in extracts derived
from the group of cells transfected with Syp-C/S was not significantly
different from than that of the control cells (data not shown).
To gain insight into the ability of Syp to modulate insulin-stimulated
translocation of GLUT4, we tested the effects of overexpressing either
Syp-WT or Syp-C/S on the ability of insulin to recruit co-transfected
GLUT4-HA to the cell surface (Fig. 5). Control cells
co-transfected with pCIS2 and GLUT4-HA had a 3-fold increase in cell
surface GLUT-HA upon maximal insulin stimulation with an
ED50 of 0.06 nM. Cells overexpressing Syp-WT
had an insulin dose-response curve that was not significantly different
from that of the control cells (p = 0.43) (Fig.
5A). In contrast, cells overexpressing Syp-C/S had decreased
insulin responsiveness when compared with control cells (Fig.
5B). The level of cell surface GLUT4-HA for cells
overexpressing Syp-C/S was approximately 20% lower than that of the
control cells at the maximally stimulating dose of insulin
(p < 0.02). However, there was no significant change
in the estimated value of the half-maximal dose of insulin (ED50) for cells overexpressing Syp-C/S when compared with
the control cells. To help rule out the possibility that the difference in the insulin dose-response curve for cells overexpressing Syp-C/S is
due to an effect of the mutant PTPase on expression of GLUT4-HA, we
evaluated total levels of GLUT4-HA in cells co-transfected with
GLUT4-HA and either pCIS2, Syp-WT, or Syp-C/S. We immunoprecipitated total membrane fractions from each group of transfected cells with an
anti-HA antibody, followed by immunoblotting with an anti-GLUT4 antibody (Fig. 3). The results of this experiment demonstrate that
there is no detectable effect of overexpressing Syp-WT or Syp-C/S on
the total level of GLUT4-HA in transfected cells. Thus, the difference
in the insulin dose-response curve of cells overexpressing Syp-C/S is
most likely due to an effect on signal transduction pathways related to
GLUT4 translocation. Since Syp-C/S is known to behave in a dominant
inhibitory manner, our results suggest that Syp may play a small
positive role in mediating insulin-stimulated translocation of
GLUT4.
A large family of PTPases is thought to be involved in modulating
signal transduction pathways initiated by receptor tyrosine kinases
such as the insulin receptor (31, 32). All known PTPases have a
conserved phosphatase domain containing a cysteine residue that is
critical for catalytic activity. Substitution of serine for this
critical cysteine results in a catalytically inactive molecule.
Particular PTPases appear to show specificity for certain receptor
tyrosine kinases (11). A subgroup of PTPases (of which Syp is a member)
achieve specificity via SH2 domains that interact with phosphotyrosine
motifs on various signaling proteins (32). Other PTPases such as PTP1B
(which is ubiquitously expressed and localized to the endoplasmic
reticulum) appear to have less specificity in their interactions with
various receptor tyrosine kinase signaling pathways (6, 8). Syp, PTP1B,
LAR, and PTP are among the PTPases that have been implicated as
modulators of insulin action (2, 33). However, few studies have
directly tested the ability of particular PTPases to modulate metabolic
actions of insulin.
PTP1B is a good candidate to test for a role in modulating metabolic actions of insulin because it is known to dephosphorylate the insulin receptor in intact cells (9-11). In addition, using a neutralizing antibody against PTP1B, Ahmad et al. (13) have recently shown that PTP1B can influence insulin-stimulated phosphatidylinositol 3-kinase activity in rat hepatoma cells (13). Phosphatidylinositol 3-kinase is a necessary component of the insulin signal transduction pathway leading to the translocation of GLUT4 in adipose cells (28). Therefore, the ability of PTP1B to decrease insulin-stimulated phosphatidylinositol 3-kinase activity suggests that PTP1B may be involved with some of the metabolic actions of insulin. Furthermore, differences in the level and activity of PTP1B have been correlated with differences in both in vitro and in vivo insulin sensitivity with respect to glucose uptake (14-16).
Although only 5% of the cells are transfected in our transient system (26), we are able to study GLUT4 translocation exclusively in the transfected cells by using a co-transfected epitope-tagged GLUT4 (GLUT4-HA) as a reporter. That is, for the purpose of studying GLUT4 translocation, the tagged GLUT4 allows us to distinguish and study transfected cells without interference from non-transfected cells. Because total levels of GLUT4-HA are comparable for all groups of cells, any differences we observe in the insulin dose-response curves presumably reflect alterations in the insulin signaling pathway leading to translocation of GLUT4. In our co-transfection experiments, we used twice as much DNA for the PTPase constructs as we did for GLUT4-HA to increase the likelihood that cells transfected with GLUT4-HA were also transfected with the vector of interest. If some fraction of cells were transfected only with GLUT4-HA, our results would underestimate the differences between control and experimental groups. Based on previous studies (using an identical protocol) where we demonstrated nearly complete inhibition of insulin-stimulated translocation of GLUT4-HA by co-expressing a dominant inhibitory mutant of phosphatidylinositol 3-kinase, we believe that at least 95% of cells expressing GLUT4-HA also express the co-transfected second plasmid under our experimental conditions (28). Unfortunately, the 5% transfection efficiency of our system limits our ability to study the effects of overexpression of PTP1B-WT on functions other than GLUT4 translocation (e.g. the tyrosine phosphorylation state of the insulin receptor, IRS-1, or other substrates). However, we did demonstrate that the overexpressed PTP1B-WT is capable of dephosphorylating the exogenous substrate pNPP.
The levels of overexpression we achieved for the PTP1B constructs in transfected adipose cells are similar to what we have previously observed with other recombinant genes in our system (26, 28, 34). Overexpression of wild-type PTP1B (PTP1B-WT) had the striking effect of decreasing both the insulin sensitivity and responsiveness of insulin-stimulated translocation of GLUT4 when compared with control cells. Interestingly, overexpression of PTP1B-WT caused a decrease in the level of cell surface GLUT4 even in the absence of insulin. Previously, we presented evidence consistent with the possibility that the insulin receptor exhibits a low level of intrinsic tyrosine kinase activity even in the absence of a ligand that is capable of signaling recruitment of GLUT4 in adipose cells (26). Thus, in addition to decreasing signaling in the presence of insulin, it is possible that overexpression of PTP1B-WT is attenuating the small signal generated by unoccupied insulin receptors.
Others have shown that the catalytically inactive mutant PTP1B-C/S can still bind with high affinity to its substrates (8). However, overexpression of PTP1B-C/S in adipose cells did not significantly affect insulin-stimulated translocation of GLUT4. The effect of overexpression of PTP1B-WT on decreasing insulin-stimulated translocation of GLUT4 appears to depend specifically on the presence of PTPase activity (in addition to the ability to bind particular substrates). If endogenous PTP1B is helping to regulate GLUT4 translocation, one might expect that overexpression of PTP1B-C/S would cause an increase in cell surface GLUT4. Our observation that PTP1B-C/S had no effect on insulin-stimulated translocation of GLUT4 may be due to low levels of endogenous PTP1B in adipose cells or the existence of redundant pathways. Taken together, our data suggest that overexpressed PTP1B is capable of functioning as a negative regulator of insulin-stimulated translocation of GLUT4 in the physiologically relevant adipose cell. It is possible that the contribution of PTP1B is relatively small under normal conditions, but that under pathological conditions where levels of PTP1B are increased (14-16), PTP1B may contribute significantly to insulin-resistant states.
Effects of Overexpression of SypAlthough PTPases are usually predicted to be negative regulators of tyrosine kinase dependent pathways, there is strong evidence that Syp functions as a positive mediator of the mitogenic actions of insulin and other hormones (20-23, 35). The SH2 domains of Syp interact directly with phosphotyrosine motifs on both the insulin receptor and IRS-1, resulting in activation of Syp PTPase activity. However, the phosphorylated insulin receptor does not seem to be a substrate for the Syp PTPase. The positive effects of Syp on insulin signaling are presumably due to dephosphorylation of other signaling molecules that have an inhibitory role when they are tyrosine-phosphorylated. As with the PTP1B-WT construct, we were able to demonstrate that the overexpressed Syp-WT construct was capable of dephosphorylating the substrate pNPP after insulin stimulation.
In contrast to the PTP1B-WT construct, overexpression of wild-type Syp (Syp-WT) in adipose cells did not significantly affect the ability of insulin to recruit GLUT4 to the cell surface. However, overexpression of the catalytically inactive mutant Syp-C/S resulted in a small but statistically significant impairment in insulin-stimulated translocation of GLUT4 at high insulin concentrations. Both Syp-WT and Syp-C/S contain identical SH2 domains, and both constructs were expressed at comparable levels. Therefore, the ability of the wild-type and mutant Syp constructs to interact with phosphotyrosine motifs on the insulin receptor, IRS-1 and other proteins is presumably the same. As with PTP1B, the effects we observe on insulin-stimulated translocation of GLUT4 most likely reflect differences in the PTPase activity of the two constructs. Syp-C/S has been shown to function in a dominant inhibitory fashion for the mitogenic actions of insulin in other cell types (20-23, 35). That is, overexpression of Syp-C/S inhibits the functioning of endogenous Syp. Therefore, the decreased insulin responsiveness with respect to GLUT4 translocation observed in adipose cells overexpressing Syp-C/S is consistent with a role for Syp as a positive mediator of metabolic actions of insulin. A recent report showing that transgenic mice heterozygous for a dominant inhibitory mutant allele of Syp have a small impairment in in vivo insulin sensitivity is supportive of our results (36).
Hausdorff et al. (24) investigated the role of Syp in differentiated 3T3-L1 cells by microinjection of anti-Syp antibodies or glutathione S-transferase fusion proteins containing SH2 domains of Syp. In contrast to our results, these experiments were unable to demonstrate an effect of Syp on insulin-stimulated translocation of GLUT4. It is likely that differences between our results and the results of Hausdorff et al. (24) reflect differences in cell type and methodology. Differentiated 3T3-L1 cells are a tissue culture line derived from fibroblasts that serve as a model of adipose cells. However, they are not as insulin-responsive as primary adipose cells. Furthermore, the method used to assess cell surface GLUT4 in 3T3-L1 cells relies on visualizing a plasma membrane sheet that is probed with an antibody against GLUT4 in conjunction with a fluorescently labeled antibody. Since this method is not amenable to precise quantitation, small differences in the amount of GLUT4 present in the plasma membrane may be difficult to detect.
We have previously shown that constitutively active mutants of Ras can signal recruitment of GLUT4 to the cell surface in transfected rat adipose cells (28). Interestingly, we also observed that a dominant inhibitory mutant of Ras had a tendency to decrease the amount of GLUT4 recruited to the cell surface at high insulin concentrations (although this effect was not statistically significant). Syp has been shown to have a positive effect on insulin-stimulated activation of Ras (23). Therefore, it is possible that the small positive effect of Syp on insulin-stimulated translocation of GLUT4 is mediated by Ras-dependent pathways.
It is possible that effects from high level overexpression of Syp and PTP1B constructs may not reflect what occurs under physiological conditions. Nevertheless, our data raise the possibility that PTP1B may be a negative regulator of insulin-stimulated glucose transport, while Syp may have a small role as a positive mediator of metabolic actions of insulin. Thus, in addition to contributing to mitogenic actions of insulin, these PTPases (and others) may play a role in insulin signal transduction pathways related to metabolic functions.
We thank Dr. G. I. Bell for supplying the human GLUT4 cDNA, Dr. C. Gorman for the pCIS2 expression vector, Dr. J. Chernoff for the human PTP1B cDNA, and Dr. B. Neel for the human Syp cDNA. We also thank Dr. S. I. Taylor for thoughtful reading of this manuscript.