INTRODUCTION

Top Abstract Introduction Methods Results Discussion References

THE EXISTENCE OF THE sphingomyelin signaling pathway is well established. In this pathway, activation of sphingomyelinase(s) results in the hydrolysis of sphingomyelin to ceramide and phosphorylcholine. Although no biological effect of phosphorylcholine has been identified as yet, extensive evidence points to an important signaling role of ceramide. The latter substance has been shown to stimulate ceramide-activated protein kinase (19), ceramide-activated protein phosphatase (4), mitogen-activated protein kinase (23), and stress-activated protein kinases (34). In addition, ceramide has been shown to mediate many distinct biological effects in a variety of cell types. Ceramide analogs have been observed to stimulate differentiation of HL-60 cells to monocytes (15), stimulate proliferation of fibroblasts (22), cause apoptosis in Molt-4 T leukemia cells (12), phosphorylate the epidermal growth factor receptor at Thr⁶⁶⁹ in A431 epidermoid carcinoma cells (5), activate nuclear factor kappa B in permeabilized Jurkat cells (27), and attenuate the rate of secretory protein traffic through the Golgi apparatus and out of the cell in vesicular stomatitis virus-infected Chinese hamster ovary cells (25). A membrane-bound ceramidase is able to further convert ceramide to sphingosine. Sphingosine is a potent protein kinase C inhibitor at pharmacological concentrations (7) and has been implicated in proliferation of mesangial cells (3). However, sphingosine may not play as important a signaling role as ceramide since augmentation of the cellular ceramide level with sphingomyelinase has been shown to be associated with only minimal conversion to sphingosine (17) and other studies have demonstrated that exogenous sphingosine is rapidly acylated to ceramide (5). Known activators of the sphingomyelin-signaling pathway include tumor necrosis factor- (TNF-) (15, 36), -interferon (15), interleukin 1 (19), 1,25-dihydroxyvitamin D₃ (20), and 1,2-diacylglycerol (16).

Several studies have linked TNF- and the elevation of cellular ceramide levels with insulin resistance. In animal models of obesity and non-insulin-dependent diabetes mellitus that exhibit insulin resistance, adipose tissue has been found to express elevated levels of TNF- (11). Chronic infusion of low levels of TNF- over a 24-h period induced insulin resistance in otherwise healthy normal rats (18). In genetically obese, insulin-resistant Zucker rats, neutralization of TNF- with a soluble TNF receptor-IgG fusion protein complex dramatically improved the sensitivity of these animals to insulin and partially restored insulin-stimulated peripheral glucose uptake in both adipose tissue and skeletal muscle (9, 11). Studies from this laboratory have also shown that the ceramide concentration in skeletal muscles in vivo remains unchanged during insulin-stimulated glucose uptake by the muscle in normal rats (30). In contrast, elevated ceramide levels have been found in liver and skeletal muscles of genetically obese Zucker rats, which exhibit both hepatic and peripheral insulin resistance, suggesting the possibility that ceramides are involved in the development of insulin resistance (32).

Treatment of cells with exogenous sphingomyelinase provides a useful tool for studies of the sphingomyelin pathway because sphingomyelin is present in both the outer and inner layer of the plasma membrane. Our previous attempt to reproduce tissue insulin resistance by acute elevation of ceramide concentration in muscles in vitro with exogenous sphingomyelinase proved unsuccessful (31). Rat soleus muscles incubated with sphingomyelinase for 1 or 2 h exhibited 140-250% increases in muscle ceramide levels, but under these conditions the muscles showed an augmentation of both basal and insulin-stimulated glucose uptake (31). Subsequent studies revealed that the sphingomyelinase-induced increase in glucose uptake was inhibited by cytochalasin B, suggesting the involvement of glucose transporters. Sphingomyelinase had no effect on the cellular accumulation of L-glucose, which is not transported by glucose transporters (31).

The purpose of the present study was to investigate the mechanisms of sphingomyelinase-stimulated glucose uptake. This project set out to determine whether the sphingomyelinase-stimulated glucose uptake was an effect specific to skeletal muscle or whether it could be reproduced in insulin-responsive cells of a different tissue type, specifically rat adipocytes. The participation of the GLUT-1 and GLUT-4 isoforms of glucose transporters was also investigated. The following sections describe these findings in detail.

	METHODS

Top Abstract Introduction Methods Results Discussion References

Animals and materials. Male Sprague-Dawley rats were purchased from Taconic Farms (Germantown, NY). All experimental procedures were approved by the Institutional Animal Care and Use Committee and the institutional veterinarian. Guiding principles in the care and use of animals approved by the Council of the American Physiological Society were strictly adhered to. Collagenase (Type I) was obtained from Worthington Biochemical (Freehold, NJ). D-[¹⁴C(U)]-glucose and [³H]5'-AMP were purchased from New England Nuclear (Boston, MA). Rabbit antibodies against GLUT-1 and rabbit antibodies against GLUT-4 were obtained from Charles River-East Acres Biologicals (Southbridge, MA). Antiphosphotyrosine antibody (PY-20) and an antibody against the insulin receptor -subunit were from Transduction Laboratories (Lexington, KY). The insulin receptor substrate-1 (IRS-1) antibody (rabbit anti-rat IRS-1 carboxy terminal peptide) was from Upstate Biotechnology (Lake Placid, NY). Anti-mouse Ig, peroxidase-linked sheep antibody, and anti-rabbit Ig, peroxidase-linked donkey antibody, were from Amersham (Arlington Heights, IL). Sphingomyelinase from Staphylococcus aureus and all other chemicals were purchased from Sigma (St. Louis, MO).

Adipocyte preparation. Isolated adipocytes were prepared from epididymal fat pads excised from 150- to 190-g fed male rats by the method of Weber et al. (33). Briefly, whole epididymal fat pads were excised, minced, and digested at 37°C with 3.6 mg collagenase/ml Krebs-Ringer bicarbonate-HEPES buffer (KRBH) [120 mM NaCl, 4 mM KH₂PO₄, 1 mM MgSO₄, 750 µM CaCl₂, 10 mM NaHCO₃, 30 mM HEPES (free acid), 200 nM adenosine, pH 7.4] containing 1% BSA. After ~1 h of digestion at 37°C with rapid agitation, isolated adipocytes were filtered through a 250-µm nylon mesh screen and washed three times with an equal volume of 1% BSA-KRBH and once with 3% BSA-KRBH.

After each wash, cells were quickly packed by centrifugation (250 g). A final wash in 3% BSA-KRBH was followed by a quick centrifugation (1,000 g) to pack the cells in a tight layer. The infranatant was removed and the packed adipocytes were suspended in 3% BSA-KRBH and distributed equally into polyethylene treatment vials.

Uptake of glucose by adipocytes. Isolated adipocytes were suspended in 3% BSA-KRBH at a final concentration of 1.25% (vol/vol) cells and preincubated with 0, 0.1, or 1 mU sphingomyelinase/ml and with 0, 0.01, or 10.0 mU insulin/ml at 37°C for 30 min. At the end of preincubation, [¹⁴C(U)]D-glucose was added at a final concentration of 0.05 µCi/ml (0.2 nM) and the cells were incubated for an additional 30 min. The adipocytes were subsequently separated from the incubation medium with dinonylphtalate and centrifugation using a horizontal rotor. Cell-associated radioactivity was determined after addition of Formula 989 in a scintillation counter.

Lactate dehydrogenase activity. The release of lactate dehydrogenase by adipocytes into the incubation medium was assessed by measurement of NADH oxidation at 340 nm using a lactate dehydrogenase activity diagnostic kit (Sigma).

Subcellular fractionation of adipocytes. Adipocytes were suspended in 3% BSA-KRBH to achieve a 25% (vol/vol) final concentration of cells. The suspension was incubated with 0, 0.1, or 1 mU sphingomyelinase/ml at 37°C with gentle rocking for 30 min. After incubation, cells from each treatment were washed with an equal volume of 17°C TES (20 mM Tris, 1 mM EDTA, 255 mM sucrose, pH 7.4 at 4°C), and plasma membranes (PM) and low-density microsomes (LDM) were prepared according to Weber et al. (33).

5' Nucleotidase activity. 5' Nucleotidase activity was measured using the method of Avruch and Hoelzl-Wallach (1) with modifications. Briefly, known amounts of PM or LDM protein were incubated at 37°C for 30 min in a mixture containing final concentrations of 200 µM 5'-AMP (0.3 µCi [³H]5'-AMP/µmol), 5 µM 2'3'-AMP, 2 mM MgSO₄, 0.05% Triton X-100, and 50 mM Tris, pH 8.0. The reaction was stopped by the addition of 30 mM ZnSO₄ and 30 mM Ba(OH)₂ to each tube. The samples were centrifuged for 1 min to remove the precipitate. The supernatant was transferred to scintillation vials and, after addition of Formula 989, radioactivity was assessed in a scintillation counter.

Gel electrophoresis and immunoblotting for glucose transporters. The total protein concentration of each subcellular membrane fraction was measured with the Bradford assay (Bio-Rad Laboratories, Hercules, CA) using defatted bovine serum albumin as standards. All samples were assayed in triplicate. Aliquots of subcellular membrane fractions containing equal amounts of total protein were mixed with sample buffer (final concentrations: 60 mM Tris, 10% glycerol, 2% SDS, 0.02% bromphenol blue, 5% -mercaptoethanol) and subjected to SDS-PAGE on a 10% resolving gel and a 3% stacking gel under reducing conditions. For assessing GLUT-4, the resolved proteins were electrophoretically transferred onto a pure nitrocellulose membrane and blocked overnight in PBS-Tween (10 mM Na₂HPO₄, 1.75 mM KH₂PO₄, 137 mM NaCl, 2.68 mM KCl, 0.1% Tween-20, pH 7.4) (PBS-T) containing 5% nonfat dry milk (Carnation). The membrane was incubated for 2 h at 25°C in 5% milk-PBS-T containing an antibody directed against GLUT-4 at a dilution of 1:1,000. The membrane was washed with PBS-T, incubated for 1 h at 25°C in PBS-T containing a horseradish peroxidase-linked donkey anti-rabbit Ig (1:4,000), and washed again with PBS-T until protein detection could be performed. To detect GLUT-1, the resolved proteins were electrophoretically transferred onto a pure nitrocellulose membrane and blocked in PBS-T containing 50% equine serum (Hyclone, Logan, UT) overnight. The membrane was incubated for 2 h at 25°C in 1% equine serum-PBS-T containing an antibody directed against GLUT-1 at a dilution of 1:2,000. The membrane was subsequently washed with PBS-T, incubated for 2 h at 25°C in 5% milk-PBS-T containing a horseradish peroxidase-linked donkey anti-rabbit Ig (1:10,000), and washed again with PBS-T. Proteins were detected with enhanced chemiluminescence (Amersham) and quantified by video densitometry.

Preparation of the insulin receptor. Isolated adipocytes diluted to a concentration of 25% (vol/vol) with 3% BSA-KRBH containing 10 mM sodium orthovanadate were incubated with 0 or 1 mU sphingomyelinase/ml and 0 or 100 mU insulin/ml for 5 or 30 min. Cells were washed by the addition of TES containing 10 mM sodium orthovanadate and a quick centrifugation. The packed cells were then subjected to subcellular fractionation to prepare a fraction containing both plasma membrane and nuclear and mitochondrial membranes as described above. A preparation of partially purified insulin receptors was prepared from this crude membrane fraction as described in detail previously (28). Protein concentration was measured with the Bradford assay. All samples were assayed in triplicate. Aliquots of insulin receptor preparation containing equal amounts of total protein were mixed with sample buffer for SDS-PAGE analysis.

Preparation of whole cell homogenates for studies on IRS-1. Isolated adipocytes diluted to a concentration of 25% (vol/vol) with 3% BSA-KRBH containing 10 mM sodium orthovanadate were incubated with 0 or 1 mU sphingomyelinase/ml and 0 or 100 mU insulin/ml for 5 or 15 min. Cells were washed by the addition of TES containing 10 mM sodium orthovanadate and a quick centrifugation (560 g) to pack the cells in a tight layer. Packed cells were homogenized in homogenization buffer (60 mM Tris, 5% SDS, 10% glycerol, 2% bromphenol blue, 10 mM sodium orthovanadate, 100 mM NaF, 10 mM EDTA, 100 mM Na₄P₂O₇, 50 mM dithiothreitol) with a handheld homogenizer at 95°C. The homogenate was centrifuged at 1,500 g for 15 min at 15°C. The fat layer was removed, and the samples were placed in liquid nitrogen until further analysis. The total protein concentration of each sample was measured with the bicinchoninic acid assay (Pierce, Rockford, IL) using defatted BSA standards. All samples were assayed in triplicate. Aliquots of whole cell homogenate containing equal amounts of total protein were subjected to SDS-PAGE analysis.

Gel electrophoresis and immunoblotting for insulin receptor -subunit and IRS-1. SDS-PAGE analysis was performed on a 7% resolving gel and a 3% stacking gel under reducing conditions. For assessing phosphotyrosine, the resolved proteins were electrophoretically transferred onto a pure nitrocellulose membrane and blocked overnight in PBS containing 0.1% Tween-20, 5% nonfat dry milk (Carnation), and 10 mM sodium orthovanadate, pH 7.4 (5% milk-PBS-TV) at 4°C. The membrane was incubated for 4 h at 25°C in 5% milk-PBS-TV containing an antibody directed against phosphotyrosine at a dilution of 1:1,250. The membrane was washed with PBS-T, incubated for 1 h at 25°C in 5% milk-PBS-TV containing a horseradish peroxidase-linked donkey anti-mouse Ig (1:2,000), and washed again with PBS-T. Proteins were detected with enhanced chemiluminescence. The membrane was then stripped and reprobed for the abundance of either insulin receptor -subunit or IRS-1. It was incubated in stripping solution (100 mM -mercaptoethanol, 62.5 mM Tris, 2% SDS, pH 6.8) at 37°C for 30 min, washed in PBS-T, and then blocked in 3% BSA-PBS-T, pH 7.4, for 1 h at 25°C. The membrane was incubated at 25°C for 4 h in 3% BSA-PBS-T containing an antibody directed against either the insulin receptor -subunit at a dilution of 1:250 or against IRS-1 at a dilution of 1:2,500. The membrane was washed with PBS-T, incubated for 1 h at 25°C in PBS-T containing a horseradish peroxidase-linked donkey anti-rabbit Ig (1:2,000 for the insulin receptor -antibody; 1:3,000 for IRS-1), and washed again with PBS-T. Proteins were detected with enhanced chemiluminescence.

Data evaluation. The results are expressed as means ± SE. Statistical analysis was performed using the Bonferroni t-test for multiple samples with the same control, or by ANOVA as appropriate.

	RESULTS

Top Abstract Introduction Methods Results Discussion References

Effect of sphingomyelinase on glucose uptake. Previous work in this laboratory has shown that the addition of 100 or 200 mU sphingomyelinase/ml to the incubation media of rat soleus muscles causes an increase in glucose uptake both under basal conditions and in the presence of submaximal concentrations of insulin (31). To investigate whether this sphingomyelinase-stimulated increase in glucose uptake is tissue specific, the effect of sphingomyelinase on isolated rat epididymal adipocytes was studied. As shown in Table 1, the incubation of adipocytes with 1.0 mU sphingomyelinase/ml augmented basal glucose uptake by 173% (P < 0.007). Incubation of adipocytes with 0.1 mU sphingomyelinase/ml appeared to increase glucose uptake 145%, although this result was not statistically significant (P = 0.156). In the presence of a low physiological concentration of insulin (0.01 mU/ml), glucose uptake by adipocytes was increased 194% (P < 0.001). The addition of 0.1 mU sphingomyelinase/ml to media containing insulin resulted in a 43% increase (P < 0.001) in glucose uptake compared with insulin alone. At a maximal insulin concentration (10 mU/ml), glucose uptake by adipocytes was augmented 8.6-fold (P < 0.001) and the addition of sphingomyelinase caused no further increase in uptake. Incubation of adipocytes at 0.1 or 1.0 mU sphingomyelinase/ml did not cause an increase in lactate dehydrogenase activity in the cell media, indicating that cellular integrity was maintained throughout the experiment (data not shown). The effects of sphingomyelinase in adipocytes are in agreement with previous observations on soleus muscles (31), except that the augmentation in glucose uptake in adipocytes could be demonstrated with 2-3 orders of magnitude lower concentrations of sphingomyelinase than those used in muscle incubations (31). The effect of sphingomyelinase on basal glucose uptake in adipocytes was particularly striking and thus all subsequent experiments were focused on elucidation of its mechanism.

Effect of sphingomyelinase on 5' nucleotidase activity. To validate the subcellular fractionation procedure for adipocytes and to eliminate the possibility that the incubation of cells with sphingomyelinase affects the physical properties of cellular membranes involved in membrane recovery, we assessed 5' nucleotidase activity in membrane fractions of adipocytes exposed to sphingomyelinase. As shown in Table 2, the activity of 5' nucleotidase, a plasma membrane marker, was four times higher in PMs as compared with LDMs. Treatment of adipocytes with up to 1 mU sphingomyelinase/ml had no discernible effect on 5' nucleotidase activity in either membrane fraction, suggesting that incubation of adipocytes with sphingomyelinase does not affect the physical properties related to the recovery of the membrane fractions.

Effect of sphingomyelinase on GLUT-1 translocation. GLUT-1 is involved primarily in the basal glucose uptake of cells and does not play a major role in insulin-stimulated glucose uptake. However, insulin is able to cause a modest translocation of GLUT-1 in adipocytes from LDM to the PM (data not shown).

As shown in Fig. 1A, 1 mU sphingomyelinase/ml caused a 79% (P < 0.001) decrease in the amount of GLUT-1 found in the LDM fraction, indicating the movement of GLUT-1 from this intracellular location. This finding is in agreement with the effect of insulin on GLUT-1 translocation in adipocytes. However, sphingomyelinase had no effect on the abundance of GLUT-1 in the PM fraction of these cells (Fig. 1B). It is possible that there may have been an increase in the amount of GLUT-1 in the PM fraction in response to 1 mU sphingomyelinase/ml, but because of the relatively large amount of GLUT-1 found in the plasma membrane, a slight increase might not be discernible. Because of the much lower amount of GLUT-1 present in the LDM fraction, however, it is conceivable that any slight change in GLUT-1 in this fraction will be detected. Another possible explanation for the disappearance of GLUT-1 from the LDMs without an increase in the PMs might be the translocation of GLUT-1 to a third pool of protein undetected by our fractionation methods (possibly en route to the PM).

View larger version (41K): [in this window] [in a new window]   View larger version (53K): [in this window] [in a new window]   Fig. 1.   Effect of sphingomyelinase on GLUT-1 abundance in low-density microsome (LDM) fractions (A) and plasma membrane (PM) fractions (B). Adipocytes were incubated with 0, 0.1, or 1 mU sphingomyelinase/ml for 30 min. Each value is a mean ± SE from 6-8 experiments. Effect of sphingomyelinase is expressed as a percentage of value for control adipocytes. Representative blots of LDM and PM fractions are shown at top. IOD, integrated optical density. *Statistically significant (P < 0.001).

Effect of sphingomyelinase on GLUT-4 translocation. Figure 2A depicts the effect of the incubation of adipocytes with sphingomyelinase on the abundance of GLUT-4 transporters in the low-density microsomal vesicles. Treatment with sphingomyelinase at 0.1 mU/ml had no effect, but 1 mU sphingomyelinase/ml decreased GLUT-4 abundance by 36% (P < 0.05), indicating movement out of the intracellular LDM pool. This decrease is associated with a 120% increase (P < 0.007) in the abundance of GLUT-4 in PM fractions (Fig. 2B), indicating that 1 mU sphingomyelinase/ml caused translocation of GLUT-4 from the LDMs to the PM.

View larger version (58K): [in this window] [in a new window]   View larger version (58K): [in this window] [in a new window]   Fig. 2.   Effect of sphingomyelinase on GLUT-4 abundance in LDM fractions (A) and in PM fractions (B). Adipocytes were incubated with 0, 0.1, or 1 mU sphingomyelinase/ml for 30 min. Each value is a mean ± SE from 5-8 experiments. Effect of sphingomyelinase is expressed as a percentage of value for control adipocytes. Representative blots of LDM and PM fractions are shown. *Statistically significant (A: P < 0.05; B: P < 0.07).

Phosphorylation and abundance of cellular proteins. Figure 3 shows the tyrosine phosphorylation and abundance of the insulin receptor -subunit in adipocytes incubated with 0 or 1 mU sphingomyelinase/ml and 0 or 100 mU insulin/ml for 5 or 30 min. The results demonstrate that sphingomyelinase had no effect on the phosphorylation of the insulin receptor -subunit at either interval. Insulin increased the phosphorylation of the receptor -subunit, but associated treatment with sphingomyelinase had no additional effect. The treatments had no effect on the abundance of the insulin receptor -subunit.

View larger version (25K): [in this window] [in a new window]   Fig. 3.   A: tyrosine phosphorylation of the insulin receptor -subunit. Adipocytes were incubated with 0 or 1 mU sphingomyelinase/ml and 0 or 100 mU insulin/ml for 5 or 30 min in the presence of 10 mM sodium orthovanadate, a phosphatase inhibitor. Insulin receptor (IR) was partially purified from a PM-containing fraction of the cell homogenate. Equal amounts of protein were subjected to SDS-PAGE and immunoblotted using an antibody against phosphotyrosine. Shown is a blot representative of 3 experiments. B: abundance of the insulin receptor -subunit under the conditions described in A. Shown is a blot representative of 3 experiments. Location of molecular weight markers is indicated at left of each blot. Con, control; SMase, sphingomyelinase; Ins, insulin; S+I, sphingomyelinase and insulin.

Figure 4 depicts observations on whole cell homogenates from adipocytes treated for 5 or 15 min with 0 or 1 mU sphingomyelinase/ml and 0 or 100 mU insulin/ml. No detectable proteins within the molecular weight range of 50-200 were phosphorylated in response to sphingomyelinase alone. IRS-1 was phosphorylated in response to insulin, but not in response to sphingomyelinase alone. The adipocytes that were treated with both sphingomyelinase and insulin showed no difference in IRS-1 phosphorylation as compared with insulin alone. There was no change in the abundance of IRS-1 in response to either insulin or sphingomyelinase (Fig. 4B). The data in Figs. 3 and 4 demonstrate that sphingomyelinase is not acting on glucose uptake via the early components of the insulin signal transduction cascade.

View larger version (60K): [in this window] [in a new window]   Fig. 4.   A: tyrosine phosphorylation of cellular proteins. Adipocytes were incubated with 0 or 1 mU sphingomyelinase/ml and 0 or 100 mU insulin/ml for 5 or 15 min in the presence of 10 mM sodium orthovanadate, a phosphatase inhibitor. Samples of cellular homogenate containing equal amounts of protein were subjected to SDS-PAGE and immunoblotted using an antibody against phosphotyrosine. Shown is a blot representative of 2 experiments. B: abundance of insulin receptor substrate (IRS)-1 under the conditions described in A. Shown is a blot representative of 2 experiments. Location of molecular weight markers is indicated at left of each blot.

	DISCUSSION

Top Abstract Introduction Methods Results Discussion References

The present study demonstrates that the addition of exogenous sphingomyelinase causes an increase in glucose uptake not only in skeletal muscle (31) but in adipocytes as well. The fact that this is not a tissue-specific effect indicates that this sphingomyelinase-induced augmentation of glucose uptake may be a physiologically important effect.

The present study also examined the mechanisms underlying the sphingomyelinase-induced increase in glucose uptake. It shows that the increase in glucose uptake is due to stimulation of translocation of the insulin-responsive glucose transporter, GLUT-4, to the PM with no effect on PM levels of the glucose transporter responsible for basal glucose uptake in adipocytes, GLUT-1. This stimulation of GLUT-4 translocation is not mediated by the activation of early transducers of the insulin signaling pathway, because treatment of adipocytes with sphingomyelinase has no effect on either basal or insulin-stimulated tyrosine phosphorylation of the insulin receptor -subunit or of IRS-1.

Because cellular sphingomyelinase is activated by TNF- (15, 36), it is appropriate to review the actions of TNF- and consider which of these actions, if any, could be mediated by sphingomyelinase. Available evidence indicates that chronic treatment of cultured adipocytes with TNF- decreases the amount of insulin-stimulated tyrosine phosphorylation of the insulin receptor and of IRS-1 (6, 10). Reduced GLUT-4 expression has also been reported as a result of chronic (5 day) TNF- treatment, contributing to the condition of insulin resistance and a corresponding decrease in insulin-stimulated glucose uptake (10). As incubation times are shortened, however, the results reported from various laboratories differ. Studies by Kanety et al. (13) have demonstrated that treatment of Fao hepatoma cells with TNF- for 1 h decreases insulin-stimulated tyrosine phosphorylation of both the insulin receptor and IRS-1, whereas a study by Guo and Donner (6) using 3T3-L1 cultured adipocytes has shown that TNF- causes an increase in insulin-stimulated tyrosine phosphorylation of IRS-1 from 15 to 60 min, with maximal stimulation at 30 min. Guo and Donner (6) also observed a TNF--induced increase of basal tyrosine phosphorylation of IRS-1, with a maximal response at 15-60 min. Kanety et al. (14) reported that the decrease in insulin-stimulated tyrosine phosphorylation of IRS-1 they observed was also detected in cells treated with either exogenous sphingomyelinase for 20 min or a cell-permeable ceramide for 30 min, although the decrease in insulin-stimulated tyrosine phosphorylation of the insulin-receptor was not reproduced in response to extracellular sphingomyelinase. In contrast to both of these reports, the present study did not detect any effect of sphingomyelinase treatment on either basal or insulin-stimulated tyrosine phosphorylation of the insulin receptor -subunit or of IRS-1, indicating that in this system, the increase in glucose uptake and translocation of glucose transporters is not mediated by the initial components of the insulin signal transduction pathway.

When IRS-1 is activated by tyrosine phosphorylation, it associates with and activates phosphatidylinositol 3-kinase (PI3K). Guo and Donner (6) demonstrated that a 15-min treatment of 3T3-L1 adipocytes with TNF- increased the association of IRS-1 with PI3K, which would in turn be expected to lead to translocation of GLUT-4-containing microsomal vesicles to the cell surface. However, Kanety et al. (13, 14) demonstrated a decrease in the insulin-induced association of IRS-1 with PI3K both with a 1-h treatment of Fao cells with TNF- or a 30-min treatment with sphingomyelinase. In contrast to both these studies, Turinsky et al. (31) demonstrated that sphingomyelinase treatment in skeletal muscle did not have any effect on basal or insulin-stimulated PI3K activity, nor did the addition of wortmannin, an inhibitor of PI3K, have any effect on the sphingomyelinase-induced increase in glucose uptake. The observations in the present study that indicate that sphingomyelinase treatment had no effect on tyrosine phosphorylation of either the insulin receptor -subunit or IRS-1 is consistent with the findings in skeletal muscles in vitro.

It is noteworthy that phospholipase C (PLC), the activity of which is also increased in TNF--treated cells (27, 35), exhibits an action similar to sphingomyelinase and dissimilar to TNF-. Studies have shown that exogenous phosphatidylcholine-specific PLC stimulates glucose transport in rat skeletal muscles (8, 29). The effect of PLC on glucose transport is inhibitable by cytochalasin B, indicating that it too is mediated by glucose transporters (8). The stimulatory effect of PLC on glucose transport in muscles involves Ca²⁺, because it is inhibited or abolished when muscles are incubated in a Ca²⁺-free medium (8, 29). This was not found to be the case with sphingomyelinase-stimulated glucose uptake. Neither elimination of CaCl₂ from the incubation media nor the addition of either the calcium chelator EGTA or dantrolene, an inhibitor of Ca²⁺ release from the sarcoplasmic reticulum, to the incubation media was able to prevent sphingomyelinase-stimulated glucose uptake in skeletal muscle (31). Interestingly, the augmentation of glucose transport with sphingomyelinase cannot be reproduced with the degradation products of sphingomyelin hydrolysis. Incubations of soleus muscles with either phosphorylcholine or cell-permeable C2- or C6-ceramides has no effect on glucose uptake (31). The absence of an effect may be due to poor uptake of ceramides by skeletal muscle or to the inability of these substances to reach important regulatory sites within the cell as readily as would endogenously produced metabolites of phospholipids. The latter part is underscored by the identification of three distinct types of sphingomyelinases within the cell that all may be involved in signal transduction.

These include an Mg²⁺-dependent neutral sphingomyelinase, which is localized to the PM (24), an Mg²⁺-independent cytosolic sphingomyelinase (21), and a lysosomal acid sphingomyelinase that is active at pH 5 (2). It appears that some agonists, such as TNF- and interleukin-1, can activate more than one type of sphingomyelinase (26). It has also been reported that the 55-kDa TNF- receptor has two discrete cytoplasmic domains, one mediating the activation of the neutral Mg²⁺-dependent plasma membrane-bound isoform and the other activating the acidic lysosomal enzyme (35). Consequently, some of the differences between the reported actions of TNF- and exogenous sphingomyelinase could be attributed to the fact that exogenous sphingomyelinase mimics solely the activation of the neutral sphingomyelinase in the plasma membrane without the associated activation of the acidic lysosomal isoform induced by TNF-.

Perspectives