JBC

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Originally published In Press as doi:10.1074/jbc.M410227200 on February 10, 2005

J. Biol. Chem., Vol. 280, Issue 15, 14536-14544, April 15, 2005
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow All Versions of this Article:
280/15/14536    most recent
M410227200v1
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Sadagurski, M.
Right arrow Articles by Wertheimer, E.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Sadagurski, M.
Right arrow Articles by Wertheimer, E.
Social Bookmarking
Add to CiteULike   Add to Complore   Add to Connotea   Add to Del.icio.us   Add to Digg   Add to Reddit   Add to Technorati  
What's this?

Insulin Receptor Substrate 2 Plays Diverse Cell-specific Roles in the Regulation of Glucose Transport*

Marianna Sadagurski{ddagger}, Galina Weingarten{ddagger}, Christopher J. Rhodes§, Morris F. White¶, and Efrat Wertheimer{ddagger}||

From the {ddagger}Department of Pathology, Sackler School of Medicine, Tel Aviv University, Tel Aviv, Israel, the §Pacific Northwest Research Institute, Seattle, Washington, and the Division of Endocrinology, Howard Hughes Medical Institute Children's Hospital, Harvard Medical School, Boston, Massachusetts 02115

Received for publication, September 7, 2004 , and in revised form, January 27, 2005.


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 EXPERIMENTAL PROCEDURES
 RESULTS
 DISCUSSION
 REFERENCES
 
The insulin receptor substrate 2 (IRS-2) protein is one of the major insulin-signaling substrates. In the present study, we investigated the role of IRS-2 in skin epidermal keratinocytes and dermal fibroblasts. Although skin is not a classical insulin target tissue, we have previously demonstrated that insulin, via the insulin receptor, is essential for normal skin cell physiology. To identify the role of IRS-2 in skin cells, we studied cells isolated from IRS-2 knock-out (KO) mice. Whereas proliferation and differentiation were not affected in the IRS-2 KO cells, a striking effect was observed on glucose transport. In IRS-2 KO keratinocytes, the lack of IRS-2 resulted in a dramatic increase in basal and insulin-stimulated glucose transport. The increase in glucose transport was associated with an increase in total phosphatidylinositol (PI) 3-kinase and Akt activation. In contrast, fibroblasts lacking IRS-2 exhibited a significant decrease in basal and insulin-induced glucose transport. We identified the point of divergence, leading to these differences between keratinocytes and fibroblasts, at the IRS-PI 3-kinase association step. In epidermal keratinocytes, PI 3-kinase is associated with and activated by only the IRS-1 protein. On the other hand, in dermal fibroblasts, PI 3-kinase is exclusively associated with and activated by the IRS-2 protein. These observations suggest that IRS-2 functions as a negative or positive regulator of glucose transport in a cell-specific manner. Our results also show that IRS-2 function depends on its cell-specific association with PI 3-kinase.


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
EXPERIMENTAL PROCEDURES
 RESULTS
 DISCUSSION
 REFERENCES
 
The insulin-signaling pathway exists in many cells, and is mediated through the insulin receptor (IR).1 However, the effects it has on various tissues are diverse. It is well known that liver, muscle, and adipocytes are the major targets of insulin. In these tissues, insulin-induced signaling pathways regulate major metabolic responses. These include increased glucose transport into muscle and adipocytes, increased glycogen and lipid synthesis, and decreases in gluconeogenesis and glucose release from the liver. However, through its receptor, insulin regulates processes in nonclassical insulin target tissues as well. For example, the lack of IR in pancreatic {beta}-cells results in defective pancreatic development (1), which leads to a selective impairment of glucose-dependent insulin release and diabetes. In the heart (2), insulin signaling plays an important developmental role in regulating postnatal cardiac size, myosin isoform expression, and the switching of cardiac substrate utilization from glucose to fatty acids. It has also been found that deleting IR in neurons results in increased food intake and moderate diet-dependent obesity as well as impaired spermatogenesis and ovarian follicle maturation due to the hypothalamic dysregulation of luteinizing hormone (3). These surprising observations raise a question as to the molecular basis behind the different effects of insulin among the various tissues, especially among the nonclassical insulin target tissues. To answer it, the specific molecules participating in the insulin-induced signaling pathways in each tissue must be identified.

After being activated by insulin, the IR engages and phosphorylates downstream signal proteins, initiating an intracellular signaling cascade. The first known downstream proteins phosphorylated by the activated IR are the insulin receptor substrate (IRS) family of proteins (4). Of the IRS family, the IRS proteins that are considered to play the most important role in mediating insulin signaling are IRS-1 and IRS-2. The IRSs act as docking proteins between the IR and a complex network of intracellular signaling molecules containing Src homology 2 domains. One such signaling protein is the p85 regulatory subunit of phosphatidylinositol (PI) 3-kinase (4, 5), which plays a central role in the metabolic actions of insulin (4).

To understand the role of these individual proteins in the diverse molecular mechanisms activated by insulin, knockout mice carrying null mutations of the genes encoding IRS-1 (6) or IRS-2 (7) were developed. From studies performed on inactivated IRS-1 and IRS-2, it became clear that each of these proteins has a unique role. Disruption of IRS-1 in mice retarded animal growth, but diabetes did not develop (6). In contrast to the IRS-1 KO mice, disruption of IRS-2 impaired both peripheral insulin signaling and pancreatic {beta}-cell function. As a result, IRS-2-deficient mice showed progressive deterioration of glucose homeostasis and developed diabetes (7). From these findings, it is clear that the IRSs are essential in mediating the specific metabolic roles of insulin in the classical insulin target tissues and that they participate in insulin-regulated processes in other tissues as well.

For the last few years, we have investigated the role of insulin signaling in skin. In previous studies, we have shown that functional IR is expressed in cultured murine skin keratinocytes (8). We have also shown that insulin increases the glucose transport rate, induces proliferation, and enhances differentiation in epidermal cells. Moreover, we have found that a lack of IR expression abolishes these insulin-induced processes (9). In fact, there are several skin pathologies associated with impaired insulin signaling, the most severe being impaired wound healing (10). Thus, it is clear that insulin signaling in skin is essential for normal skin function.

Following our understanding that each of the IRS proteins possesses a unique role in the insulin-signaling pathway in different tissues, we focused in the present study on the role of the IRS-2 protein in skin cells. We used the IRS-2 KO mouse model to determine the impact of IRS-2 on the two major components of skin, the dermal compartment and the overlying epidermis. Our results clearly demonstrate that the role of IRS-2 in mediating insulin-induced glucose transport is divergent in the skin epidermal keratinocyte and dermal fibroblasts. Moreover, these results further suggest that impairment in IRS-2 function can lead to an abnormality in cellular functions and may contribute to the development of diabetes-associated skin pathologies.


   EXPERIMENTAL PROCEDURES
TOP
 ABSTRACT
 INTRODUCTION
 EXPERIMENTAL PROCEDURES
RESULTS
 DISCUSSION
 REFERENCES
 
Animals—Homozygous IRS-2 null mice and wild type (WT) littermates were obtained by crossing heterozygous IRS-2 null mice, as previously described (7). Mice were genotyped using the PCR method (7). Animal experiments were approved by the appropriate institutional committee and performed in compliance with the rules for animal care of the National Institutes of Health.

Cell Culture and Protein Lysate Preparation—Primary mouse keratinocytes were prepared and maintained as previously described (8). Briefly, freshly isolated keratinocytes were cultured in Dulbecco's modified Eagle's medium (Biological Industries, Beit Haemek, Israel) with 10% fetal calf serum (Biological Industries), 1% antibiotics, and Ca2+ at a concentration adjusted to 0.05 mM. The cells were grown for 6–7 days in culture, depending on the experimental design. Primary fibroblasts were isolated as described, cultured in Dulbecco's modified Eagle's medium with 10% fetal calf serum, and grown to confluence. Cells were then harvested by scraping them into lysis buffer containing phosphate-buffered saline (PBS), 1% Triton X-100, 1 mM EDTA, 10 mM sodium fluoride, 200 mM sodium orthovanadate, and a protease inhibitor mixture. The lysate was spun down at maximum speed in a microcentrifuge, and the Triton-soluble supernatant was subjected to SDS-PAGE and immunoblotted. Protein concentrations were measured using a modified Lowry assay (DC Protein Assay Kit; Bio-Rad).

Immunoblotting Analysis—The protein samples were separated by SDS-PAGE and transferred to a nitrocellulose membrane. Immunoblotting was performed as described previously (11) using rabbit polyclonal antibodies against the IR {beta} subunit (Santa Cruz Biotechnology, Inc., Santa Cruz, CA), IRS-1 and IRS-2 proteins (Upstate Biotechnology, Inc., Lake Placid, NY), the anti-p85 PI 3-kinase regulatory subunit, which recognizes all the isoforms of the regulatory subunits (Santa Cruz Biotechnology), rabbit polyclonal anti-phospho-Akt antibody recognizing phosphorylated Ser473 of mouse Akt1 (Cell Signaling, Inc., Beverly, MA), and monoclonal antibodies recognizing phosphotyrosine residues (Upstate Biotechnology). Filters were then incubated with the appropriate secondary horseradish peroxidase-conjugated anti-rabbit (Bio-Rad) or anti-mouse (Amersham Biosciences) antibodies. Proteins were detected using an enhanced chemiluminescence protein detection kit (SuperSignal; Bio-Rad).

Immunoprecipitation—Confluent cultures were incubated at 37 °C for 3 or 15 min, in the presence or absence of 10–6 M insulin (Sigma). The reaction was terminated by aspiration of the incubation medium, followed by quick washes with ice-cold PBS on ice. Cells were lysed in cold lysis buffer. The lysates were subjected to immunoprecipitation with the aforementioned antibodies and incubated on protein A/G-agarose beads (Santa Cruz Biotechnology) for 16–20 h at 4 °C. The beads were then washed twice with cold lysis buffer, subjected to immunoblotting, and visualized by using an enhanced chemiluminescence system (Roche Applied Science).

2-Deoxy-[3H]Glucose Uptake—Glucose transport was evaluated by measuring 2-deoxy-[3H]glucose ([3H]dGlc) according to a previously described method (12). Briefly, cells were incubated with insulin (10–6 M; Sigma) or insulin-like growth factor 1 (IGF-1; 10–7 M; PeproTech Inc., Rocky Hill, NJ) and then washed three times with PBS. A solution of 0.1 mM dGlc/PBS with trace amounts of [3H]dGlc (1 µCi/plate; ARC, St. Louis, MO) was added to the cells. Uptake was allowed to continue for 10 min at room temperature. The reaction was stopped by three quick washes with 1 ml of cold PBS on ice. Cells were lysed in 1% Triton X-100. The samples were counted in the 3H window of a Tricarb scintillation counter. Uptake was linear under these conditions for up to 15 min (data not shown). Mean values were determined from measurements of triplicate samples.

Thymidine Incorporation—Cell proliferation was evaluated by measuring [3H]thymidine incorporation into DNA. Cells were pulsed with [3H]thymidine (1 µCi/ml; ICN, Irvine, CA) for 1 h at 37°C. After incubation, cells were washed three times with PBS, incubated for 15 min at room temperature in 5% trichloroacetic acid, and solubilized in 1% Triton X-100. The radioactivity incorporated into the cells was counted in the 3H window of the Tricarb liquid scintillation counter. Mean values were determined from measurements of triplicate samples.

Adenovirus Vectors—The recombinant adenoviral construct, which encodes a WT IRS-2 protein, was constructed as described (13). An adenoviral construct containing a luciferase reporter gene was used as a control to exclude the effects of the infection procedure. For infection, the proliferated cells were cultured in medium containing the adenoviruses for 1 h at 37 °C. Cells were subjected to experiments 48 h after infection. The adenoviruses were applied at the multiplicity of infection indicated in each experiment.

PI 3-Kinase Activity Assay—In vitro phosphorylation of PI was carried out as follows. Total cell extracts were immunoprecipitated as described above, using anti-IRS-1, anti-IRS-2, or anti-phosphotyrosine antibodies (Upstate Biotechnology). The immunocomplexes were washed three times with the following buffers; buffer I (PBS, 1% Nonidet P-40, 0.1 mM sodium orthovanadate), buffer II (100 mM Tris-HCl, pH 7.5, 0.5 M LiCl, 0.1 mM sodium orthovanadate), and buffer III (100 mM Tris-HCl, pH 7.5, 1 mM EDTA, 100 mM NaCl, 0.1 mM sodium orthovanadate). The immunoprecipitates were then incubated with 100 µl of reaction mixture (20 mM Hepes, pH 7.5, 10 mM MnCl2 containing 0.1 mg/ml sonicated PI (Sigma) and 100 µCi/ml [{gamma}-32P]ATP) for 10 min at 22 °C. The reaction was stopped by the addition of 100 µlof1 M HCl, and the phospholipids were extracted with 200 µl of CHCl3/MeOH (1:1, v/v). After centrifugation for 10 min, the organic phase was collected and extracted again with an equal volume of 1 M HCl/MeOH (1:1, v/v). Aliquots (30 µl) from the bottom organic phase were spotted on a silica gel TLC plate (Merck). The plate was developed in CHCl3/CH3OH/H2O/NH4OH (90:70:17:3), dried, and visualized by autoradiography with x-ray film (Eastman Kodak Co.).


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 EXPERIMENTAL PROCEDURES
 RESULTS
DISCUSSION
 REFERENCES
 
The Expression Profile of Insulin-signaling Molecules Is Similar in Both Epidermal and Dermal Skin Cells—The goal of the present study was to compare insulin signaling in two nonclassical insulin target cells composing the skin tissue, epidermal keratinocytes, and dermal fibroblasts. These two cell types are not only proximal to each other; they also work coordinately in many processes regulated by insulin, in particular the wound healing process. In the present study, we focused mainly on the role of one of the insulin-signaling molecules, the IRS-2 protein, by comparing its roles in skin epidermal and dermal cells. For this purpose, we utilized transgenic mice lacking IRS-2 expression.

Initially, we followed the expression and phosphorylation of the first steps of the insulin-signaling pathway, of IR, IRS-1, and IRS-2, in primary keratinocytes and fibroblasts. As expected, Western blot experiments revealed that in the WT cells, these three insulin-signaling molecules are expressed and activated in response to insulin. In the IRS-2 KO cells, IRS-2 protein was not detected in either keratinocytes or fibroblasts (Fig. 1, A and B). IR was similarly expressed and phosphorylated in the WT and IRS-2 KO skin keratinocytes and fibroblasts, regardless of the level of IRS-2 expression. There was only a slight increase in the expression and insulin-induced activation of IRS-1 protein in both IRS-2 KO cell types. These results suggest that there is only a partial compensatory mechanism of insulin-induced IRS-1 activation in skin cells lacking IRS-2 expression.



View larger version (34K):
[in this window]
[in a new window]
  		FIG. 1.
The expression profile of insulin-signaling molecules is similar in skin epidermal and dermal cells. A, WT and IRS-2 KO primary keratinocytes were grown for 5 days in medium containing 0.05 mM Ca2+. Proliferating keratinocytes were acutely stimulated with insulin (10–6 M) for 3 or 15 min at 37 °C. Control keratinocytes (0) were cultured in the absence of insulin. Cells were then lysed, and 700 µg of total protein was immunoprecipitated with anti-IRS-2, anti-IR, and anti-IRS-1 antibodies. The immunoprecipitates were analyzed by Western blotting, using anti-phosphotyrosine, anti-IRS-2, anti-IR, or anti-IRS-1 antibodies, as indicated. IP, immunoprecipitation; IB, immunoblot. B,WT and IRS-2 KO fibroblasts were grown to confluence. The experimental procedure and the immunoblotting were performed as described in A. Results are representative of three separate experiments.

 
IRS-2 Is Not Involved in Regulating the Proliferation of Skin Cells—Insulin is an essential growth factor for the proliferation of many cells, including skin cells (8). This is supported by the reduction in the proliferation rate of skin cells lacking IR (9). We therefore examined whether keratinocytes and fibroblasts lacking IRS-2 expression retain their ability to proliferate. We studied the cellular proliferation of primary cultured cells isolated from WT and IRS-2 KO mice by measuring thymidine incorporation. As can be seen in Fig. 2, A and B, no significant change in thymidine incorporation rate was detected in IRS-2 KO keratinocytes or fibroblasts in comparison with their WT counterparts. We can therefore conclude that IRS-2 is not involved in skin proliferation. In subsequent experiments, we followed the involvement of IRS-2 in another process known to be regulated by insulin, namely skin epidermal differentiation (8, 9), and again, direct regulation by IRS-2 was not apparent (data not shown).



View larger version (17K):
[in this window]
[in a new window]
  		FIG. 2.
IRS-2 is not involved in regulating proliferation of skin cells. For thymidine incorporation studies, primary skin keratinocytes (A) and fibroblasts (B), isolated from WT or IRS-2 KO mice were plated as described under "Experimental Procedures." Black bars, untreated WT and IRS-2 KO cells; gray bars, cells chronically stimulated with insulin (10–6 M) for 16 h. Proliferation was estimated by measuring [3H]thymidine incorporation, as described under "Experimental Procedures," and expressed as disintegrations/min (DPM)/mg of protein. Results are representative of three experiments. Each bar represents the mean ± S.E. of triplicates.

 
Differential Effects of the Lack of IRS-2 Protein on Glucose Transport in Skin Cells—Both skin epidermal keratinocytes and dermal fibroblasts are nonclassical insulin target tissues. Neither of these cell types responds to acute insulin stimulation by rapidly increased glucose transport. However, in previous studies, we have demonstrated the importance of insulin and the IR in chronic regulation of glucose transport into skin cells (9, 14). We thus aimed to evaluate the function of IRS-2 in the regulation of glucose transport into skin cells. We examined basal and insulin-induced glucose uptake in proliferating keratinocytes isolated from IRS-2 KO and WT mice (Fig. 3A). Surprisingly, the lack of IRS-2 resulted in a 2-fold increase in basal glucose transport into skin epidermal keratinocytes relative to the WT mice. There was an even more dramatic increase, 4–10-fold, in the glucose transport rate in response to insulin stimulation, as compared with the response in WT epidermal cells.



View larger version (16K):
[in this window]
[in a new window]
  		FIG. 3.
Differential effects of a lack of IRS-2 protein on glucose transport in skin cells. Proliferating skin keratinocytes (A) and fibroblasts (B) were chronically stimulated with insulin (10–6 M) (gray bars) overnight. Control cells (0) were not stimulated (black bars). dGlc uptake was measured as described under "Experimental Procedures." Values represent mean ± S.D. of triplicate determinations and are shown as disintegrations/min (DPM)/mg of protein per 10 min (*, p < 0.05 WT versus IRS-2 KO control; **, p < 0.05 WT versus IRS-2 KO insulin-stimulated). Results are from one representative experiment of four separate experiments.

 
One possible explanation for the increase in glucose transport in the IRS-2 KO cells may be a change in expression or cellular localization of one of the glucose transporters (15). From the family of glucose transporters, GLUT-1 is the main glucose transporter expressed in primary skin keratinocytes (9).

However, we have studied the expression and localization of the GLUT-1 protein, and there was no relationship between the increase in glucose uptake in cells of the IRS-2 KO mice and the cellular or membranal expression of the GLUT-1 protein (data not shown).

Following these intriguing findings, we studied the effects of a lack of IRS-2 on the glucose transport rate into the second major skin cell type, dermal fibroblasts. In contrast to epidermal cells, we witnessed a decrease in basal glucose transport in IRS-2 KO fibroblasts compared with the WT cells (Fig. 3B). Insulin stimulation resulted in an increase in glucose transport into the IRS-2 KO fibroblasts, but it remained slower than the glucose transport rate into the WT cells.

Therefore, we propose that in skin epidermal cells, IRS-2 plays a unique inhibitory role in the regulation of glucose transport. Consequently, the lack of IRS-2 results in removal of the inhibitory effects, and as a result, there is an increase in the glucose transport rate into the cells. In contrast, in dermal fibroblasts, IRS-2 plays a stimulatory role in the regulation of glucose transport, and the lack of its expression leads to a down-regulation of the glucose transport rate into the cells.

Effects of IRS-2 Overexpression on Glucose Transport Rate in Skin Cells—To further confirm this unexpected finding, we undertook the opposite experiment; IRS-2 was overexpressed in both epidermal keratinocytes and dermal fibroblasts. Overexpression of IRS-2 resulted in a marked decrease in basal glucose uptake into epidermal keratinocytes as well as into epidermal keratinocytes chronically stimulated by insulin (Fig. 4A). In contrast, IRS-2 overexpression in dermal fibroblasts had no such effect (Fig. 4B). These results further support the putative negative regulatory role that IRS-2 plays in glucose transport into skin keratinocytes.



View larger version (16K):
[in this window]
[in a new window]
  		FIG. 4.
Effects of IRS-2 overexpression on glucose transport rate in skin cells. Proliferating keratinocytes (A) and fibroblasts (B) were infected with an adenoviral IRS-2 construct for 1 h. Control cells were infected with a luciferase-containing adenoviral construct (Luc). After 48 h, the cells were treated or not with insulin (10–6 M) for 16 h, after which dGlc uptake was measured as described under "Experimental Procedures." Values represent means ± S.D. from at least three experiments (*, p < 0.05 Luc versus AdexIRS-2 control; **, p < 0.05 Luc versus AdexIRS-2 insulin-stimulated). The results shown are representative of three independent experiments.

 
Differential Activation of PI 3-Kinase in Skin Cells Lacking IRS-2 Expression—To further support our theory, we studied the activation of proteins downstream of IRS-2 in the signaling pathway of glucose transport regulation. Our goal was to determine whether they are similarly activated.

IRS proteins are known to mediate their effects on glucose transport via the PI 3-kinase-Akt/protein kinase B signaling pathway. Thus, we first investigated the effects of a lack of IRS-2 on total PI 3-kinase phosphorylation and activity rates in both skin epidermal keratinocytes and dermal fibroblasts (Fig. 5, A, B, and C and D, E, and F, respectively). As shown in Fig. 5, A–C, overall basal and insulin-induced PI 3-kinase phosphorylation (Fig. 5A) and activation (Fig. 5, B and C) were elevated in the IRS-2 KO epidermal keratinocytes as compared with the WT. On the other hand, a slight decrease in the total cellular PI 3-kinase phosphorylation (Fig. 5D) and activity rates (Fig. 5, E and F) was observed in the IRS-2 KO dermal fibroblasts as compared with the normal cells.



View larger version (27K):
[in this window]
[in a new window]
  		FIG. 5.
Differential activation of PI 3-kinase in skin cells lacking IRS-2 expression. Proliferating skin keratinocytes (A–C) and fibroblasts (D–F) were stimulated with insulin (10–6 M) for 3 or 15 min. Cell lysates were immunoprecipitated with an anti-phosphotyrosine antibody. Western blot analysis was performed with anti-PI 3-kinase antibody (A and D), or immunoprecipitates were immediately used for an in vitro PI 3-kinase assay (B and E), as described under "Experimental Procedures." Results are representative of four experiments. Autoradiograms corresponding to the sum of three independent in vitro PI 3-kinase assay experiments were analyzed by scanning densitometry. Results (C and F) are expressed as arbitrary units and are means ± S.E. PIP, inositol trisphosphate. IP, immunoprecipitation; IB, immunoblot; Ori, origin.

 
Differential Activation of Akt in WT and IRS-2 KO Keratinocytes and Fibroblasts—To further support these results, we assessed the phosphorylation levels of Akt, known to be a key downstream mediator of PI 3-kinase in cellular glucose transport. As can be seen in Fig. 6, A and B, there was an increase in Akt phosphorylation levels in the IRS-2 KO keratinocytes compared with the WT. These results are in agreement with the observed increase in the glucose transport rate in the IRS-2-deficient keratinocytes and are consistent with the proposed inhibitory role of IRS-2 in the IRS-1-PI 3-kinase-Akt signaling pathway. In contrast, in the IRS-2-deficient fibroblasts, a significant insulin-induced decrease in Akt phosphorylation levels was observed (Fig. 6, C and D), consistent with the suggested stimulatory role that IRS-2 plays in mediating glucose transport in the fibroblasts.



View larger version (29K):
[in this window]
[in a new window]
  		FIG. 6.
Differential activation of Akt in WT and IRS-2 KO keratinocytes and fibroblasts. Proliferating keratinocytes (A and B) and fibroblasts (C and D) were treated or not with insulin (10–6 M). Lysates of keratinocytes (A) and fibroblasts (C) were subjected to SDS-PAGE and analyzed by Western blot with anti-phospho-Akt and anti-Akt antibodies, as indicated. Results are representative of three independent experiments. The blots of phospho-Akt were further analyzed by scanning densitometry (B and D). Results are expressed as arbitrary units and are the means ± S.E. (*, p < 0.05 WT versus IRS-2 KO insulin-stimulated). Results are representative of four separate experiments.

 
The IRS-PI 3-Kinase Association Is the Point of Divergence between the Signaling Pathways of Normal Skin Epidermal Keratinocytes and Dermal Fibroblasts—As already mentioned, the lack of IRS-2 expression differentially affects the regulation of glucose transport in skin epidermal keratinocytes and dermal fibroblasts. We further demonstrated that despite the lack of IRS-2 in epidermal keratinocytes, the activity of the PI 3-kinase-Akt signaling pathway, which mediates glucose transport, is elevated. In contrast, we showed that the PI 3-kinase-Akt pathway is inhibited in dermal fibroblasts lacking IRS-2 expression. Thus, we hypothesized that the differences in glucose transport regulation are upstream of PI 3-kinase and that different proteins associate with this kinase. Therefore, our next step was to investigate the mediation of PI 3-kinase association and activation by either IRS-2 or IRS-1.

Cells were stimulated with insulin, and total cell lysates were immunoprecipitated with antibodies against IRS-2 or IRS-1. The IRS-associated PI 3-kinase was identified by immunoblotting with anti-PI 3-kinase antibody or by in vitro PI 3-kinase activity assay. In skin epidermal keratinocytes, PI 3-kinase was exclusively associated with and activated by IRS-1. PI 3-kinase did not associate with IRS-2 (Fig. 7, A–C). In contrast, in fibroblasts, the insulin-stimulated PI 3-kinase was exclusively associated with and activated by IRS-2. PI 3-kinase did not associate with IRS-1 (Fig. 7, D–F). We concluded that the differential association of the IRS-1 and IRS-2 proteins with PI 3-kinase probably indicates a point of divergence between the IRS proteins in the two cell types.



View larger version (23K):
[in this window]
[in a new window]
  		FIG. 7.
The IRS-PI 3-kinase association is the point of divergence between the signaling pathways of normal skin epidermal keratinocytes and dermal fibroblasts. Proliferating skin keratinocytes (A–C) and fibroblasts (D–F) were stimulated with insulin (10–6 M) for 3 or 15 min. Cells lysates were immunoprecipitated with anti-IRS-1 or anti-IRS-2 antibodies and analyzed by Western blot with anti-PI 3-kinase antibody (A and D) or used for an in vitro PI 3-kinase assay, as described under "Experimental Procedures" (B and E). A representative experiment is shown. The autoradiograms corresponding to three independent experiments were analyzed by scanning densitometry (C and F). Results are expressed as arbitrary units of PI 3-kinase activity and are the means ± S.E. PIP, inositol trisphosphate. IP, immunoprecipitation; IB, immunoblot.

 
Insulin-stimulated PI 3-Kinase Activation in IRS-2-deficient Keratinocytes and Fibroblasts—To study whether the IRS-1-PI 3-kinase interaction is affected by the absence of IRS-2, we measured IRS-2 and IRS-1 association with PI 3-kinase as well as the IRS-1-associated PI 3-kinase activity resulting from this interaction, skin cells isolated from the IRS-2 KO mice, and control littermates. A lack of IRS-2 in either epidermal keratinocytes or dermal fibroblasts was not compensated by increased interaction between PI 3-kinase and the IRS-1 protein (Fig. 8, A and C, respectively); nor was there an increase in the IRS-1-associated PI 3-kinase activity in these cells (Fig. 8, B and D, respectively). These findings are intriguing due to the fact that the total PI 3-kinase activity in the IRS-2 KO keratinocytes was increased (Fig. 5), suggesting a non-IRS-1-associated PI 3-kinase activation process.



View larger version (22K):
[in this window]
[in a new window]
  		FIG. 8.
Insulin-stimulated PI 3-kinase activation in IRS-2-deficient keratinocytes and fibroblasts. Proliferating keratinocytes (A and B) and fibroblasts (C and D) isolated from WT or IRS-2 KO mice were stimulated with insulin (10–6 M) for 3 or 15 min. Cells lysates were immunoprecipitated with anti-IRS-1 or anti-IRS-2 antibodies and analyzed by Western blot with anti-PI 3-kinase antibody (A and C) or used for an in vitro PI 3-kinase assay, as described under "Experimental Procedures" (B and D). A representative experiment is shown. PIP, inositol trisphosphate. IP, immunoprecipitation; IB, immunoblot.

 

   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 EXPERIMENTAL PROCEDURES
 RESULTS
 DISCUSSION
REFERENCES
 
Insulin is a central metabolic regulatory hormone, involved in the regulation of glucose transport into muscle, fat, and liver, the classical insulin target tissues. In recent years, however, the significance of other effects of insulin in "nonclassical" tissues has been under investigation. This field of research has markedly expanded due to the availability of tissue-specific KO mice, in which the insulin effects on an individual tissue can be studied without the development of general metabolic consequences (16). The results of several studies using such models have demonstrated that insulin plays a role in the proliferation and differentiation of various tissues and takes part in many processes in tissues that are not considered classical insulin targets. It is also becoming clear that the different IRS molecules, which mediate intracellular insulin signaling, play redundant or selective roles in insulin action in a tissue-dependent manner.

In the present study, we investigated the role of insulin signaling in the nonclassical insulin target tissue, skin. We chose to focus on this tissue because of the skin pathologies associated with conditions of insulin resistance, such as diabetes mellitus, as well as insulin resistance due to mutations in the insulin receptor gene (reviewed in Refs. 10 and 17). More specifically, we studied the role of IRS-2 in two skin cell types, epidermal skin keratinocytes and dermal skin fibroblasts.

Our results demonstrated the involvement of IRS-2 in mediating insulin-regulated glucose transport in skin cells. Surprisingly, we found that in keratinocytes isolated from IRS-2 KO mice, the lack of IRS-2 leads to an enormous increase in glucose transport rate. This increase was observed in both basal and insulin-stimulated cells, indicating that in keratinocytes, IRS-2 plays an inhibitory role in the regulation of glucose transport. When IRS-2 is inactivated, the inhibition is removed, resulting in the activation of various signaling pathways, including an observed increase in PI 3-kinase and Akt activities. This, in turn, leads to a marked increase in glucose transport into the epidermal cells. The results are even more striking in view of the fact that a lack of either IR (9) or IRS-12 in the epidermal cells causes a decrease in the glucose-transport rate. Support for these findings came from experiments that took the opposite approach (i.e. in which IRS-2 was overexpressed in the epidermal keratinocytes); the resultant inhibition of glucose transport further confirmed the inhibitory role of IRS-2 in these cells. However, another interesting observation was made. The effect of IRS-2 absence on glucose transport was tissue-specific. Whereas in skin epidermal keratinocytes, it resulted in increased glucose uptake, in skin fibroblasts, it resulted in a marked decrease in glucose transport.

The increase in glucose transport in the IRS-2 KO cells may have resulted from a change in the expression or cellular localization of one of the glucose transporters. However, there was no correlation between the changes in the glucose transport rate and the expression of GLUT-1, the main glucose transporter expressed in skin keratinocytes. Such a discrepancy between GLUT-1 expression and the glucose transport rate was observed previously in various studies (14, 18–20). This was explained by the hypothesizing of an additional activation step of the glucose transporters, which is essential for its transport activity (21, 22). We assume that a similar activation step is occurring in the skin keratinocytes as well, which could explain our findings.

It is generally accepted that IRS-2 positively regulates glucose transport (23–27), as was observed in the present study in skin fibroblasts. However, there are a few examples of increased IRS-2 being associated with insulin resistance and/or decreased glucose transport, as occurred here in the skin keratinocytes (28–31). Thus, our next step was to identify the molecular mechanism leading to these variations in IRS-2 function in our two cell types.

PI 3-kinase and Akt were activated in a pattern similar to that of glucose transport; in the IRS-2 KO epidermal keratinocytes, PI 3-kinase and Akt activities were up-regulated, whereas in IRS-2 KO dermal fibroblasts, these activities were decreased. Thus, we hypothesized that the point of divergence between keratinocytes and fibroblasts must lie upstream of PI 3-kinase. The only known signaling step between the IRS proteins and PI 3-kinase activation is the formation of IRS-PI 3-kinase complex. We therefore followed this interaction in both cell types. We discovered that the formation of IRS-1-PI 3-kinase or IRS-2-PI 3-kinase complexes was cell-specific. In skin keratinocytes, PI 3-kinase is solely associated with and activated by the IRS-1 protein, whereas in skin fibroblasts, PI 3-kinase is exclusively associated with and activated by the IRS-2 protein. This difference is therefore likely to be the point of divergence between the IRS isoforms in the two cell types.

There are several possible explanations for the dissimilarities between IRS-1 and IRS-2 association with downstream proteins such as PI 3-kinase. (i) Although there is 43% amino acid sequence identity between these two proteins, some domains, such as the COOH-terminal ones, exhibit lower degrees of homology. (ii) IRS-2 is 100 residues longer than IRS-1. (iii) The IRS-2 sequence contains a unique region (amino acids 591–786), not present in IRS-1, which interacts specifically with the kinase regulatory loop binding (KRLB) domain of the {beta}-subunit of IR. (iv) There is a difference in the phosphorylation patterns of IRS-1 and IRS-2. These patterns are important in interactions between the IRS proteins and various downstream Src homology 2 domain-containing effectors, including PI 3-kinase. The IRS-2 protein contains 22 potential tyrosine phosphorylation sites, of which only 13 are conserved in IRS-1. (v) Additionally, the putative binding motifs for PI 3-kinase are not identical. IRS-1 contains two strong and two weak PI 3-kinase-binding sites. IRS-2, on the other hand, contains only two strong PI 3-kinase-binding sites, the amino acid sequences of which vary from those of the IRS-1 sites (reviewed in Refs. 32 and 33). These variations could underlie the differences in the IRS-PI 3-kinase interactions between IRS-1 and IRS-2. In addition, these two substrates may also regulate unique signaling pathways due to distinct subcellular localizations, different kinetics of activation and deactivation, or diverse tissue distribution and expression patterns during fetal development. All of these factors could also contribute to the tissue-specific functions of IRS-1 or IRS-2, as seen in the present study and in other studies.

Another interesting finding was that although in epidermal keratinocytes lacking IRS-2 expression, IRS-1-associated PI 3-kinase activity did not show a compensatory increase, the total cellular PI 3-kinase activity did increase (Figs. 8 and 5, respectively). A possible explanation for these findings could be that in skin keratinocytes, there is an additional protein affecting PI 3-kinase activation. Following this theory, this additional protein would be bound to IRS-2 in WT cells, whereas PI 3-kinase would be associated with IRS-1. However, in cells lacking IRS-2, this additional protein would be free to associate with PI 3-kinase or affect its activity, leading to increased glucose transport into the cells. Note that such non-IRS-associated PI 3-kinase activity has been suggested in other cells. For example, it has been suggested that a monomeric p85 PI 3-kinase subunit that is not bound to IRS plays a physiologically important role in attenuating signaling through PI 3-kinase-dependent pathways (34, 35). There are other possible candidates that could affect the IRS-PI 3-kinase complex formation or PI 3-kinase activity. Shc is known to interact with either IRS or IR. Thus, the balance between IRS-Shc complexes and IRS-IR complexes might determine the formation of the IRS-PI 3-kinase complexes, affecting the biological outcome.

Another possibility is a direct interaction of PI 3-kinase with the IR. Such a shift between IR-PI 3-kinase and IRS-PI 3-kinase complex formation has been demonstrated in skeletal muscle in vivo (36), where certain conditions preferentially increase the pool of PI 3-kinase activity associated with the IR, at the expense of that associated with IRS. One should also consider other proteins that are known to interact with PI 3-kinase in other cells, including PTEN, GRB10, SHP-2, PKC isoforms, and GAB-1. Moreover, the IRS isoforms could be interacting with other unknown proteins. It has been suggested that IRS-1 and IRS-2 assemble into a multiprotein complex, and that this facilitates coupling to the IR (37). Thus, any factor that regulates the binding or release of either of the IRS proteins from this location may affect insulin signaling via the IR-IRS pathway downstream, by way of PI 3-kinase.

In summary, we show here that differences exist in the insulin-signaling pathway, even among the group of non-classical insulin target tissues, in this case epidermal keratinocytes and dermal fibroblasts. Furthermore, we show that IRS-2 has a cell-specific role in glucose transport, and that in some cells, such as skin keratinocytes, it might play an inhibitory role. Furthermore, we propose the existence of a non-IRS-dependent PI 3-kinase signaling pathway. More studies are needed to elucidate the precise functions of each of the proteins in the IR-IRS-PI 3-kinase pathway in regulating cellular processes.


   FOOTNOTES
 
* The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. Back

|| To whom correspondence should be addressed: Dept. of Pathology, Sackler School of Medicine, Tel Aviv University, Tel Aviv 69978, Israel. Tel.: 972-3-640-6111; Fax: 972-3-640-9141; E-mail: effy{at}patholog.tau.ac.il.

1 The abbreviations used are: IR, insulin receptor; dGlc, 2-deoxy-D-glucose; IGF-1, insulin-like growth factor 1; IRS, insulin receptor substrate; IRS-1 KO, IRS-1 knockout; IRS-2 KO, IRS-2 knockout; PI, phosphatidylinositol; WT, wild type; PBS, phosphate-buffered saline. Back

2 E. Wertheimer, unpublished results. Back



   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 EXPERIMENTAL PROCEDURES
 RESULTS
 DISCUSSION
 REFERENCES
 

  1. Kulkarni, R. N., Bruning, J. C., Winnay, J. N., Postic, C., Magnuson, M. A., and Kahn, C. R. (1999) Cell 96, 329–339[CrossRef][Medline] [Order article via Infotrieve]
  2. Belke, D. D., Betuing, S., Tuttle, M. J., Graveleau, C., Young, M. E., Pham, M., Zhang, D., Cooksey, R. C., McClain, D. A., Litwin, S. E., Taegtmeyer, H., Severson, D., Kahn, C. R., and Abel, E. D. (2002) J. Clin. Invest. 109, 629–639[CrossRef][Medline] [Order article via Infotrieve]
  3. Bruning, J. C., Gautam, D., Burks, D. J., Gillette, J., Schubert, M., Orban, P. C., Klein, R., Krone, W., Muller-Wieland, D., and Kahn, C. R. (2000) Science 289, 2122–2125[Abstract/Free Full Text]
  4. White, M. F. (2002) Am. J. Physiol. Endocrinol. Metab. 283, E413–E422[Abstract/Free Full Text]
  5. Giovannone, B., Scaldaferri, M. L., Federici, M., Porzio, O., Lauro, D., Fusco, A., Sbraccia, P., Borboni, P., Lauro, R., and Sesti, G. (2000) Diabetes Metab. Res. Rev. 16, 434–441[CrossRef][Medline] [Order article via Infotrieve]
  6. Tamemoto, H., Kadowaki, T., Tobe, K., Yagi, T., Sakura, H., Hayakawa, T., Terauchi, Y., Ueki, K., Kaburagi, Y., and Satoh, S. (1994) Nature 372, 182–186[CrossRef][Medline] [Order article via Infotrieve]
  7. Withers, D. J., Gutierrez, J. S., Towery, H., Burks, D. J., Ren, J. M., Previs, S., Zhang, Y., Bernal, D., Pons, S., Shulman, G. I., Bonner-Weir, S., and White, M. F. (1998) Nature 391, 900–904[CrossRef][Medline] [Order article via Infotrieve]
  8. Wertheimer, E., Trebicz, M., Eldar, T., Gartsbein, M., Nofech-Mozes, S., and Tennenbaum, T. (2000) J. Invest. Dermatol. 115, 24–29[CrossRef][Medline] [Order article via Infotrieve]
  9. Wertheimer, E., Spravchikov, N., Trebicz, M., Gartsbein, M., Accili, D., Avinoach, I., Nofech-Mozes, S., Sizyakov, G., and Tennenbaum, T. (2001) Endocrinology 142, 1234–1241[Abstract/Free Full Text]
  10. Wertheimer, E. (2004) in Diabetes Mellitus (LeRoith, D., Taylor, S. I., and Olefsky, J. M., eds) pp. 1411–1422, Lippincott Williams & Wilkins, Philadelphia, PA
  11. Wertheimer, E., Barbetti, F., Muggeo, M., Roth, J., and Taylor, S. I. (1994) J. Biol. Chem. 269, 7587–7592[Abstract/Free Full Text]
  12. Wertheimer, E., Sasson, S., and Cerasi, E. (1990) J. Cell. Physiol. 143, 330–336[CrossRef][Medline] [Order article via Infotrieve]
  13. Lingohr, M. K., Dickson, L. M., McCuaig, J. F., and Hugl, S. R., Twardzik, D. R., and Rhodes, C. J. (2002) Diabetes 51, 966–976[Abstract/Free Full Text]
  14. Spravchikov, N., Sizyakov, G., Gartsbein, M., Accili, D., Tennenbaum, T., and Wertheimer, E. (2001) Diabetes 50, 1627–1635[Abstract/Free Full Text]
  15. Joost, H. G., and Thorens, B. (2001) Mol. Membr. Biol. 18, 247–256[CrossRef][Medline] [Order article via Infotrieve]
  16. Kitamura, T., Kahn, C. R., and Accili, D. (2003) Annu. Rev. Physiol. 65, 313–332[CrossRef][Medline] [Order article via Infotrieve]
  17. Wertheimer, E. (2004) Isr. Med. Assoc. J. 6, 287–289[Medline] [Order article via Infotrieve]
  18. Czech, M. P., and Corvera, S. (1999) J. Biol. Chem. 274, 1865–1868[Free Full Text]
  19. Cushman, S. W., Goodyear, L. J., Pilch, P. F., Ralston, E., Galbo, H., Ploug, T., Kristiansen, S., and Klip, A. (1998) Adv. Exp. Med. Biol. 441, 63–71[Medline] [Order article via Infotrieve]
  20. Lemieux, K., Konrad, D., Klip, A., and Marette, A. (2003) FASEB J. 17, 1658–1665[Abstract/Free Full Text]
  21. Harrison, S. A., Clancy, B. M., Pessino, A., and Czech, M. P. (1992) J. Biol. Chem. 267, 3783–3788[Abstract/Free Full Text]
  22. Furtado, L. M., Somwar, R., Sweeney, G., Niu, W., and Klip, A. (2002) Biochem. Cell Biol. 80, 569–578[CrossRef][Medline] [Order article via Infotrieve]
  23. Fasshauer, M., Klein, J., Ueki, K., Kriauciunas, K. M., Benito, M., White, M. F., and Kahn, C. R. (2000) J. Biol. Chem. 275, 25494–25501[Abstract/Free Full Text]
  24. Zhou, L., Chen, H., Lin, C. H., Cong, L. N., McGibbon, M. A., Sciacchitano, S., Lesniak, M. A., Quon, M. J., and Taylor, S. I. (1997) J. Biol. Chem. 272, 29829–29833[Abstract/Free Full Text]
  25. Higaki, Y., Wojtaszewski, J. F. P., Hirshman, M. F., Withers, D. J., Towery, H., White, M. F., and Goodyear, L. J. (1999) J. Biol. Chem. 274, 20791–20795[Abstract/Free Full Text]
  26. Nandi, A., Kitamura, Y., Kahn, C. R., and Accili, D. (2004) Physiol. Rev. 84, 623–647[Abstract/Free Full Text]
  27. Kido, Y., Burks, D. J., Withers, D., Bruning, J. C., Kahn, C. R., White, M. F., and Accili, D. (2000) J. Clin. Invest. 105, 199–205[Medline] [Order article via Infotrieve]
  28. Friedman, J. E., Ishizuka, T., Shao, J., Huston, L., Highman, T., and Catalano, P. (1999) Diabetes 48, 1807–1814[Abstract]
  29. Buren, J., Liu, H. X., Lauritz, J., and Eriksson, J. W. (2003) Eur. J. Endocrinol. 148, 157–167[Abstract]
  30. Buren, J., Liu, H. X., Jensen, J., and Eriksson, J. W. (2002) Eur. J. Endocrinol. 146, 419–429[Abstract]
  31. Laviola, L., Belsanti, G., Davalli, A. M., Napoli, R., Perrini, S., Weir, G. C., Giorgino, R., and Giorgino, F. (2001) Diabetes 50, 2709–2720[Abstract/Free Full Text]
  32. White, M. F. (2002) Am. J. Physiol. Endocrinol. Metab. 283, E413–E422[Abstract/Free Full Text]
  33. Sesti, G., Federici, M., Hribal, M. L., Lauro, D., Sbraccia, P., and Lauro, R. (2001) FASEB J. 15, 2099–2111[Abstract/Free Full Text]
  34. Li, L., Gronning, L. M., Anderson, P. O., Li, S., Edvardsen, K., Johnston, J., Kioussis, D., Shepherd, P. R., and Wang, P. (2004) J. Biol. Chem. 279, 34107–34114[Abstract/Free Full Text]
  35. Barbour, L. A., Shao, J., Qiao, L., Leitner, W., Anderson, M., Friedman, J. E., and Draznin, B. (2004) Endocrinology 145, 1144–1150[Abstract/Free Full Text]
  36. Shao, J., Yamashita, H., Qiao, L., Draznin, B., and Friedman, J. E. (2002) Diabetes 51, 19–29[Abstract/Free Full Text]
  37. Clark, S. F., Molero, J. C., and James, D. E. (2000) J. Biol. Chem. 275, 3819–3826[Abstract/Free Full Text]

Add to CiteULike CiteULike   Add to Complore Complore   Add to Connotea Connotea   Add to Del.icio.us Del.icio.us   Add to Digg Digg   Add to Reddit Reddit   Add to Technorati Technorati    What's this?


This article has been cited by other articles:


Home page
 Mol. Endocrinol.Home page
S. Guo, S. L. Dunn, and M. F. White
The Reciprocal Stability of FOXO1 and IRS2 Creates a Regulatory Circuit that Controls Insulin Signaling
Mol. Endocrinol., December 1, 2006; 20(12): 3389 - 3399.
[Abstract] [Full Text] [PDF]

Home page
 Mol. Cell. Biol.Home page
M. Sadagurski, S. Yakar, G. Weingarten, M. Holzenberger, C. J. Rhodes, D. Breitkreutz, D. LeRoith, and E. Wertheimer
Insulin-like growth factor 1 receptor signaling regulates skin development and inhibits skin keratinocyte differentiation.
Mol. Cell. Biol., April 1, 2006; 26(7): 2675 - 2687.
[Abstract] [Full Text] [PDF]

Home page
 J. Clin. Endocrinol. Metab.Home page
T. P. Ciaraldi, S. A. Phillips, L. Carter, V. Aroda, S. Mudaliar, and R. R. Henry
Effects of the Rapid-Acting Insulin Analog Glulisine on Cultured Human Skeletal Muscle Cells: Comparisons with Insulin and Insulin-Like Growth Factor I
J. Clin. Endocrinol. Metab., October 1, 2005; 90(10): 5551 - 5558.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow All Versions of this Article:
280/15/14536    most recent
M410227200v1
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Sadagurski, M.
Right arrow Articles by Wertheimer, E.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Sadagurski, M.
Right arrow Articles by Wertheimer, E.
Social Bookmarking
Add to CiteULike   Add to Complore   Add to Connotea   Add to Del.icio.us