+ An incumbent of the Charles H. Hollenberg Professorial Chair

Transcription

1 THE JOURNAL OF BIOLOGICAL CHEMISTRY by The American Society of Biological Chemists, Inc. Vol. 261, No. 32, Issue of November 15, pp ,1986 Printed in U. S. A. Effect of Depletion of Phosphate and Bicarbonate Ions on Insulin Action in Rat Adipocytes (Received for publication, April 14, 1986) Yoram ShechterS and Anat Ron From the Department of Hormone Research, The Weizmann Institute of Science, Rehouot 76100, Israel The effect of insulin on rat adipocytes was studied in isolated adipocytes (8,9). The effects of insulin on ionic fluxes isotonic buffers (ph 7.4) containing NaCl, CaCL are quite rapid and are considered to be independent of its MgS04, KCl, and bovine serum albumin but no phos- effect on glucose transport since they occur in the absence of phate or bicarbonate anions. In phosphate- and bicar- glucose in the medium (reviewed in Ref. 1). bonate-free buffers the dose-response curve to insulin Kissebah et al. (10) have hypothesized that in adipose is shifted to the right, the effects of the hormone on tissue, insulin triggers the release of plasma membrane-bound hexose uptake, glucose metabolism, and inhibition of Ca2+ into the cytoplasm which results in a rapid rise in the lipolysis being observed at much higher (nearly 2 or- intracellular pool of free Ca2+ followed by increased uptake ders of magnitude) concentrations of insulin. The in- sulin binding capacity of the cells is only slightly changed. The dose-response curve for isoproterenol which stimulates lipolysis in the same cell type is almost the same in both Krebs-Ringer bicarbonate buffer and phosphate- and bicarbonate-free buffers. The dose-response curves for agents that mimic the action of insulin such as wheat germ agglutinin or vanadate ions are also shifted to the right. The dose-response plasmic Ca2+ (1). In hepatocytes at least, an increase in cytoplasm Ca2+ due to insulin could not be detected (2). If curve to insulin can be returned to normal by re- cytoplasmic Ca2+ has an essential role in insulin action in addition of either bicarbonate or phosphate. Almost adipocytes it must be recruited from internal pools as most complete recovery is obtained at either 10 mm bicar- or all the actions of insulin in this target tissue are observed bonate or 24 mm phosphate, respectively. External in the absence of Ca2+ in the external medium (1, 2, 11). Ca2+ ions which are not required for the proper action In uitro studies related to insulin s action adipocytes in are of insulin in fat cells maintained in Krebs-Ringer bi- traditionally performed in Krebs-Ringer bicarbonate buffer carbonate buffer, become essential for insulin action (12). The phosphate and bicarbonate content of KRB are 1.2 in bicarbonate-free buffer. The study indicates that and 25 mm, respectively. Phosphate is an essential nutrient depletion of bicarbonate and, to a lesser extent, phos- for cell growth in culture (13), e.g. fibroblasts deprived of phate anions, interferes with an essential insulin-de- phosphate are arrested in the GI phase (13, 14). Insulin acts pendent post-binding event. Also, in bicarbonate-free synergistically with other hormones and growth factors to medium, external Ca2+ ions are essential for insulin- stimulate progression through the cell cycle of the arrested mediated processes. The implications of this study to fibroblasts (13, 14). Short periods of phosphate deprivation, the mode of action of insulin, and to physiological and clinical states of insulin desensitization are discussed. however, do not usually affect cell metabolism and hormone responsiveness. With respect to bicarbonate ions, no data is available concerning effects of insulin on carbonate fluxes or During the last two decades, research has mainly focused on elucidating the effects of insulin on glucose metabolism, while the effects of the hormone on electrolyte entry and balance and the resultant effects on glucose metabolism have remained unexamined. Insulin affects the permeability of plasma membranes to ions resulting in hyperpolarization of the plasma membranes (reviewed in Refs. 1 and 2). In addition, insulin activates Na+,K+-ATPase (in some target tissues) (3) and inhibits high affinity Ca*+-ATPase (4). Early reports documented decreases in serum potassium (5, 6) and inorganic phosphate (7) in response to injected insulin. This has been confirmed in several in uitro systems including * This work was supported by a grant from the Israel Academy of Sciences and Humanities. 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. + An incumbent of the Charles H. Hollenberg Professorial Chair in Metabolic and Diabetes Research, established by the Friends and Associates of Dr. Charles H. Hollenberg of Toronto, Canada. To whom correspondence should be sent into mitochondrial and perhaps endoplasmic reticulum stores. This rise in turn activates insulin-dependent membrane transport events and intracellular enzymes which stimulate glucogenesis and lipogenesis and decrease lipolysis and gly- cogenesis. Many studies indeed suggest that one of the first events after insulin binding to intact cells is a rise in cyto- whether this anion is essential to insulin action. The present study is addressed to investigate this issue. EXPERIMENTAL PROCEDURES Materials-Porcine insulin was purchased from Eli Lilly. D-[ cc] Glucose (4-7 mci/mol, 1 Ci = 37 GBq) and 3-O-[methyl-3H]glucose and NalZ5I were obtained from New England Nuclear and Amersham Corp. Collagenase type I (134 units/mg) was obtained from Worthington. Iodine-labeled insulin was prepared by the chloramine-t method (15). Methods-Fat cells were prepared from male Wistar rats ( g) according to Ref. 12. Unless otherwise stated the collagenase digestion and the biological assays were run in buffer containing 10 mm HEPES (ph 7.4) 120 mm NaC1, 1.4 mm CaCl,, 5.2 mmkc1, 1.2 mm MgS04, and 0.7% BSA. The buffer does not contain phosphate or bicarbonate ions and is referred to as Buffer A throughout this paper. When supplemented with bicarbonate, the ph was corrected to 7.4 with HC1. The following assays and methods were done according to the methods of the references listed uiz the only modification The abbreviations used are: KRB, Krebs-Ringer bicarbonate; WGA, wheat germ agglutinin; HEPES, 4-(2-hydroxyethyl)-l-piperazineethanesulfonic acid EGTA, [ethylenebis(oxyethylenenitrilo)]tetraacetic acid; BSA, bovine serum albumin.

2 ~ -. ~~~~ Bicarbonate Depletion and Insulin Adipocytes Action Rat was in the buffer used; glucose oxidation (12), lipogenesis (16), lipolysis, and anti-lipolysis (17) binding of insulin to rat adipocytes (18), 2 and 3-0-methylglucose transport (19). w o Percent of maximal stimulation was calculated using the equation $ 3 80 (Vri".l-Vbasal)/(Vmar-Vbasal) X 100, when Vim, vbasalt and Vmax are the 8 2 rate of lipogenesis at a given insulin concentration; in the absence of insulin, and at insulin concentration of 100 ng ml-', respectively INSULIN (ng/ml) FIG. 3. Rates of lipogenesis with increasing concentrations of insulin in Buffer A that was supplemented with phosphate. Adipocytes were prepared and assayed for lipogenesis at increasing concentrations of insulin in KRB buffer (O), Buffer A (O), and Buffer A that was supplemented with phosphate to a final concentration of 1.2 mm (D), 12 mm (m), and 24 mm (0). ph levels were corrected to INSULIN (ng/rnl) (0). Buffer A (ph 7.4) consisted of: HEPES, 10 mm; NaCl, 120 mm; TABLE I Rates of basal and maximal stimulation of lipogenesis in KRB buffer and in Buffer A Fat cells were prepared and assayed for lipogenesis in either KRB buffer or Buffer A. Lipogenesis was carried out for 1 h at 37 "C. The final reaction volume of 0.5 ml contained 0.16 mm glucose and about 3 X 10' fat cells. ~ Additions KRB buffer Buffer A cpm incorporated glucose into the fat content of 3 X 10' adipocyteslh None 3,050 f 150 1,530 f 80 Insulin, 100 ng ml" 10,940 f ,800 f 500 -Fold stimulation I INSULIN (nq/ml) from D-[u-"c] FIG. 2. Rates of lipogenesis with increasing concentrations of insulin in Buffer A that was supplemented with bicarbonate. Adipocytes were prepared and assayed for lipogenesis at increasing concentrations of insulin in KRB buffer (0), Buffer A (O), and Buffer A that was supplemented with NaHC03 to a final concentration of 1 mm (o), 3 mm (m), 6 mm (o), and 10 mm (*). ph values were corrected to IO 50 IO X) INSULIN (ng/rnl) FIG. 4. Rates of lipogenesis and glucose oxidation with increasing concentrations of insulin in KRB buffer and Buffer A. Adipocytes were prepared and assayed in KRB buffer (0, 0) or Buffer A (0, m) for lipogenesis (0, 0) and glucose oxidation (0, m) at increasing concentrations of insulin. The final reaction volume of 0.5 ml contained 0.16 mm glucose and about 3 X lo5 fat cells. The cells were incubated for 1 h at 37 "C. The incorporation of labeled glucose into fat and the conversion of D-[U-'4C]glucose to I4CO2 was then determined in the same samples. the results of at least three separate experiments. Each point in the figures represents the means f S.E. All assays were run in duplicate. RESULTS Effects of Phosphate and Bicarbonate Depletion on Lipogenesi.-When adipocytes were prepared and assayed for lipogenesis in Buffer A (which contains BSA and all electrolytes except phosphate and bicarbonate ions) the dose-response curve for insulin was dramatically shifted to the right (Fig. 1). ED,, values were 0.2 f 0.1 and 18 f 2 ng/ml in KRB buffer and Buffer A, respectively. Basal rate of lipogenesis in Buffer A was decreased by about half, while the maximal stimulation (in the presence of 100 ng ml" insulin) remained unaltered in both buffers. Therefore, insulin produced a 3.6- and 7-fold stimulation in KRB buffer and Buffer A, respectively (Table I). Shift of the Dose-response Curve to Normal by Bicarbonate and Phosphate Anions-The role of bicarbonate ions in shifting the dose-response curve to insulin was further evaluated by supplementing Buffer A with increasing concentrations of bicarbonate (Fig. 2). The additions of 1.0, 3.0, 6.0, and 10 mm NaHC03 resulted in stepwise shifts to the left of the insulin dose-response curves. ED,, values were 2 f 0.2, 1.4 f 0.15,

3 Bicarbonate Depletion and Insulin Action in Rat Adipocytes TABLE I1 Rate of 3-O-[methyl-3H]glucose uptake by fat cells maintained in Buffer A at varying concentrations of insulin Adipocyte suspension (1.5 X 105 cells in 0.2 ml of Buffer A in a plastic tube) were incubated with the indicated concentrations of insulin for 30 min at 37 "C. 3-O-[methyl-3H]Glucose was then added (final concentrations of 0.1 mm) and the cells were separated after 20 s through a layer of silicone oil. 3-O-[methyl-3H]Glucose Addition uptake None Insulin 1 ng rnl-lb 10 ng ml" 100 ng ml" 1000 ne ml" pmol/20 ~110% cells" 33f6 35 f 6 43 f 8 85 f a Radioactivity associated with the cells in the presence of 20 p~ cytochalasin B was subtracted from all tubes. I, In KRB buffer, insulin at 1 ng/ml produced % of maximal stimulation in various experiments. " I I I I,,111 I I I I,11111 I 1 I 1 I, IO Insulln(ng/mll FIG. 6. Inhibition of lipolysis by increasing concentrations of insulin, in KRB and Buffer A. Fat cells (-3 X lo6 cells/ml) were prepared and assayed in either KRB buffer (0) or Buffer A (0) and assayed with a fixed concentration of isoproterenol (3 X lo-' M) and increasing concentrations of insulin, for 3 h at 37 "C. Basal and stimulated levels of lipolysis are represented by dashed lines. ISOPROTERENOL (MI FIG. 5. Rate of lipolysis with increasing concentrations of isoproterenol in KRB buffer and Buffer A. Suspension of rat adipocytes (approximately 3 X lo6 cells) were prepared and assayed in either KRB buffer (0) or Buffer A (0). Incubation was performed for 3 h at 37 "C. Glycerol content in the medium was then determined f 0.04, and 0.23 f 0.2 ng ml" at 1, 3, 6, and 10 mm NaHC03, respectively (values derived from Fig. 2). Thus, the Bound (frnoll IO6 cells) addition of 10 mm NaHC03 nearly returned the dose-response FIG. 7. Scatchard analysis of insulin binding to cells precurve of insulin to "normal" (Fig. 2). The decreased basal rate pared in different buffers. Cells were prepared and assayed for of lipogenesis obtained with Buffer A (Table I) remained low insulin binding in KRB buffer (O), Buffer A (O), or Buffer A that even on supplementation with 1-10 mm NaHC03. Therefore, was supplemented with NaHC03 to a final concentration of 1 mm 5-7-fold stimulation of lipogenesis was observed on addition (e). Binding analyses were performed for 45 at min room temperature of insulin in the various experiments (data not shown). with approximately 0.8 x lo6 cells in the absence and presence of 5 p~ In a similar manner, the role of phosphate ions was evaluunlabeled insulin. Nonspecific binding amounted to 30 & 5% of total binding. Binding was terminated by separating the cells through ated. Supplementation of Buffer A with 1.2 mm phosphate a layer of silicone oil (18). (which is the phosphate level in KRB buffer) had only a slight effect on the dose-response curve to insulin (Fig. 3). Further nesis and for glucose oxidation are almost superimposable in supplementation of Buffer A with 12 and 24 mm phosphate shifted the dose-response curve to the left. EDso values were both buffer systems (Fig. 4). In agreement with that, f 1.1, , and 0.6 f 0.05 ng ml" at a phosphate [methyl-3h]glucose uptake into adipocytes that were prepared concentrations of 1.2, 12, and 24 mm, respectively. Thus, it and maintained in Buffer A is not significantly stimulated at seems that the observed shift to the right of the insulin dose ng/ml insulin. However, stimulation occurs at an insulin response curve in phosphate- and bicarbonate-free buffer can concentration of pg ml" (Table 11). be returned to nearly normal values by addition of either Unaltered Dose-response Curve for Isoproterenol in Buffer bicarbonate or high concentrations of phosphate ions. A-In contrast to the effects seen with insulin, the dose- Effect of Anion Deficiency on Glucose Oxidation, and 3-0- response curve for isoproterenol in stimulating lipolysis in the [methyl-3h]glucose Uptake-When glucose oxidation was same cell type was only slightly altered. Half-maximal lipolmeasured in adipocytes that were prepared and assayed in ysis was observed at and 0.03 PM isoproterenol for Buffer A, similar shifts in the insulin dose-response curve lipolysis assayed in KRB buffer and in Buffer A, respectively were observed (Fig. 4). The dose-response curves for lipoge- (Fig. 5).

4 14948 Bicarbonate Depletion and Insulin Action in Rat Adipocytes TABLE I11 Effect of 1.6 mm EGTA on insulin-stimulated lipogenesis in various buffers Cells were prepared and assayed for lipogenesis in the indicated buffers for 1 h at 37 "C. Buffer systems used" Insulin* KRB buffer 10,750 10,950 Buffer A 1,560 11,400 Buffer A, 1.25 mm phosphate Buffer A, 8, mm phosphate 10,200 Buffer 7,700 A, 24 mm phosphate 8,900 Buffer A, 1,900 1 mm bicarbonate 10,800 Buffer 6,000 A, 3 mm bicarbonate 9,900 Buffer A, 9,500 6 mm bicarbonate 10,450 Buffer A. 10 mm bicarbonate Basal Insulin plus EGTA' cpm incorporated from D-[U-"C]g~ucose into the fat content of 3 X 10' adipocytes/h 3,100 1,540 1,900 9,800 2,300 2,600 1,570 1,740 1, a All buffers contained 0.7% BSA. * The final concentration of insulin in the assay is 100 ng ml-'. e EGTA (final concentration 1.6 mm) was added only during lipogenesis. Suppression of insulin-activated lipogenesis by EGTA' % Wheat germ agglutinin(&mi) FIG. 8. Rate of lipogenesis with increasing concentrations of wheat germ agglutinin in KRB buffer and Buffer A. Adipocytes were prepared and assayed for lipogenesis at increasing concentrations of WGA in KRB buffer and Buffer A for 1 h at 37 "C. V BUFFER -A X I I I X X X I TRYPSIN (pg/ml) FIG. 10. Rate of lipogenesis with increasing concentrations of trypsin in KRB buffer and Buffer A. Adipocytes were prepared and assayed for lipogenesis at increasing concentrations of trypsin. Maximal stimulation obtained at 100 ng ml" of insulin is represented by the dashed line in the figure. 0 Sodlum metavanadate hm) FIG. 9. Rate of lipogenesis with increasing concentrations of sodium metavanadate in KRB buffer and Buffer A. Adipocytes were prepared and assayed for lipogenesis at increasing concentrations of sodium metavanadate in KRB buffer and Buffer A for 1 h at 37 "C. Effects of Anion Deficiency on Insulin-mediated Inhibition of Lipolysis-The concentration dependence of the effect of insulin in inhibiting lipolysis was also shifted to the right in Buffer A (Fig. 6). It seems, however, that anion deficiency has more pronounced effects on insulin stimulation of glucose metabolism. Fifty percent inhibition of lipolysis was observed EGTA (mm1 Eternal CO~' Concentralion (PM) FIG. 11. Dose-dependent EGTA inhibition of insulin-activated lipogenesis. A, cells were prepared and assayed for lipogenesis in Buffer A in the presence of insulin (100 ng/ml) and increasing concentrations of EGTA. B, dependency of insulin-activated lipogenesis on external Ca2+ concentrations in Buffer A. Cells were prepared in Buffer A and further washed with Ca2+-free Buffer A. Aliquots of adipocyte suspension were withdrawn before and at different stages of the washing and assayed for lipogenesis in the absence and the presence of insulin (100 ng ml-l). Rate of stimulation obtained in Buffer A was taken as 100%.

5 Bicarbonate Depletion and at 0.2 and 5 ng ml" insulin in KRB buffer and Buffer A, respectively (Fig. 6). Effect of Anion Deficiency on Insulin Binding-When Scatchard analyses of insulin binding were carried out using adipocytes prepared in Buffer A, some reduction in binding capacity (about 20-30%) was observed, as compared to fat cell suspensions maintained and assayed in KRB buffer (Fig. 7). The hormone binding capacity, however,was indistinguishable if the cells were prepared and assayed for binding in either KRB buffer or Buffer A that was supplemented with NaHC03 to a final concentration of 1 mm (Fig. 7). EDSovalues for lipogenesis are 0.2 ng ml" in KRB buffer and 2 ng ml" in Buffer A containing 1 mm NaHC03, respectively (Fig. 1). Therefore, the observed shifts in the dose-response curves to insulin in buffers deficient in phosphate and/or bicarbonate ions is related only in part to impaired binding of insulin in those cells. Effects of Anion Deficiency on Insulinomimetic Agents- The dose-response curves for wheat germ agglutinin (WGA) and for vanadate ions, which have insulin-like effects in rat adipocytes (20-22), were also shifted to the right in phosphate- and carbonate-depleted buffers. EDso values for WGA were 1 and 10 pg ml" in KRB and Buffer A, respectively (Fig. 8), the EDso for vanadate was found at 0.1 and 0.7 mm sodium vanadate in KRB and Buffer A, respectively (Fig. 9). It is important to mention that both insulinomimetic agents at high concentrations produce only about half the maximal effect of the hormone itself (Figs. 8 and 9). Trypsin, at a certain range of concentrations, exhibits insulin-like effects in rat adipocytes (i.e. Ref. 23). As can be seen in Fig. 10, lipogenesis is partially stimulated over a wide range of trypsin concentration in fat cells maintained in KRB buffer. In contrast, no such stimulation could be detected in fat cells maintained and assayed in Buffer A, at any trypsin concentrations ranging from 0.1 to 60 pg/ml (Fig. 10). Effect of EGTA on Lipogenesis in the Various Buffers-The addition of EGTA to KRB buffer under conditions that chelates essentially all external Ca" ions, has no influence on insulin-like effects in rat adipocytes as demonstrated in previous studies (reviewed in Refs. 1 and 2) including a study from this laboratory (11). The results in Table I11 indicate that the addition of EGTA (final concentration 1.6 mm) to lipogenesis carried out in Buffer A has no effect on basal stimulation, although the effect of insulin is completely sup- pressed (Table 111). In contrast, there is no suppression of the action of insulin by EGTA (Table 111) in adipocyte suspensions prepared in Buffer A, supplemented with either phosphate or bicarbonate ions. The stimulating effect of insulin in Buffer A is half-maxi- mally inhibited at EGTA concentration of 1.3 f 0.05 mm (value derived from Fig. 1lA). No inhibition was observed at mm EGTA. This may suggest that low external Ca2+ concentrations are sufficient for the stimulating action of insulin in Buffer A. This was evaluated more directly by preparing adipocytes in Buffer A, and further washing the cells with Ca2+-free Buffer A. Lipogenesis assay was then carried out at known concentrations of Ca2+ in the medium. As shown in Fig. 11B, maximal stimulation by insulin was already observed at 30 pm CaZf or above. Fifty percent of maximal stimulation is estimated to occur at a Ca2+ concentration of 12 f 2 p~ (value derived from Fig. 11B). DISCUSSION The present study demonstrates that depletion of either bicarbonate or phosphate ions from isotonic (ph 7.4) buffers, result in a dramatic shift to the right of the dose-response curves to insulin. Thus, the normal EDso value for the hor- Insulin Action in Rat Adipocytes mone is shifted nearly 2 orders of magnitude towards higher concentrations (Fig. 1). This property is not shared with isoproterenol, a compound that stimulates lipolysis in the same cell type. Further studies have revealed that only one of these anions is required for normal responsiveness and that high phosphate concentrations compensate for relatively low concentrations of bicarbonate ions (Figs. 2 and 3). We have further demonstrated that both insulin-dependent plasma membrane events (hexose uptake) and cytoplasmic events (anti-lipolysis) are altered in Buffer A, while the binding of the hormone is only slightly altered (Fig. 7). Therefore, anion deficiency seems to interfere with post-binding event(s) essential in the action of insulin. This is further confirmed by the rightward shift and reduced responsiveness of insulin-like agents such as WGA (Fig. 8) and vanadate ions (Fig. 9). Another point of interest is that externally located Ca2' ions which are not required for the normal effects of insulin in this cell type (1, 2, 11) are essential in anion-deficient buffer (Table 111). Preliminary studies in this direction have revealed that in Buffer A, the Ca2+ antagonists, sodium ver- apamyl and lanthanum chloride (at p ~ as ) well as the Ca2+ ionophore A (at 1-10 pm), haveno effects on either basal- or insulin-stimulated lipogenesis.2 This indicates that bicarbonate (or phosphate) depletion does not interfere with nonhormonal-dependent Ca2+ channeling (2) nor does it produce a state of insufficient cytoplasmic Ca2+ concentration. Moreover, this indicates that the post-binding event that is "desensitized" and can be "overcome" at high insulin concentrations is a Ca'+-dependent process.2 The present study may assist in understanding the mechanism(s) of action of insulin in particular the post-binding events that are involved in its binding effector chain mechanisms. These are poorly understood at the present time. Several topics which are currently being investigated in our laboratory include: studies on the level of the autophosphorylation of the insulin receptor in the presence and the absence of insulin (reviewed in Ref. 24), in anion-deficient buffers, and theffect of insulin on Ca2+ fluxes as well as on other ion fluxes (2), and on the regulation of intracellular ph (2). It is also of interest to determine the precise composition of the external electrolytes that are required for normal insulin responsiveness in the whole animal model system. This may indicate whether certain states of refractoriness to insulin, related in part to deficiencies of electrolytes. Ackmwledgments-The critical reading of and commenting on the manuscript by Drs. S. Jacobs, N. Sahyoun, and K. J. Chang of Burroughs Wellcome, Research Triangle Park, NC and Dr. Gail Galasko from the Department of Pharmacology, University of the Witwatersrand, Johannesburg, are gratefully acknowledged and we a1so thank B. Zarmi for technical assistance. REFERENCES 1. Czech, M. P. (1977) Annu. Rev. Biochem. 46, Moore, R. D. (1983) Biochim. Biophys. Acta 737, Haddon, J. W., Hadden, E. M., Wilson, E. E., Good, R. A., and Coffey, R. G. (1972) Nat. New Biol. 235, Pershadsingh, H. A., and McDonald, J. M. (1979) Nature 281, Briggs, A. P., Koechig, I., Doisy, E. A., and Weber, C. J. (1924) J. Biol. Chem. 68, Harrop, G. A., Jr., and Benedict, E. M. (1924) J. Biol. Chem. 69, Stadie, W. C. (1944) Yale J. Biol. Med. 16, Touabi, M., and Jeanrenaud, B. (1970) Bwchim. Biophys. Acta 202, Hepp, D., Challoner, D. R., and Williams, R. H. (1968) J. Bid. Chem. 243, Kissebah, A. H., Hope-Gill, H., Videlingum, N., Tullouch, B. R., * Y. Shechter and A. Ron, manuscript in preparation.

6 14950 Bicarbonate Depletion and Insulin Action Rat in Adipocytes Clarke, P. V., and Frazer, T. R. (1975) Lancet 1, Bwchim. Bwphys. Acta Shechter, Y. (1984) Proc. Natl. Acad. Sci. U. S. A. 81, Czech, M. P. (1976) J. Bbl. Chem. 261, '12. Rodbell, M. (1964) J. Biol. Chem. 239, Cuatrecasas, P. (1973) J. Biol. Chern. 248, Strauss, D. (1984) Endocrin. Reu. 5, Dubyak, G. R., and Kleinzeller, A. (1980) J. Biol. Chem. 265, 14. Kamley, D., and Rudland, P. (1976) Nature 260, Hunter, W. M., and Greenwood, F. C. (1962) Nature 194, Shechter, Y., and Karlish, S. J. D. (1980) Nature 284, Kono, T., and Barham, F. W. (1971) J. Biol. Chem. 246, Moody, A. J., Stan, M. A., Stan, M., and Gliemann, J. (1974) 6209 Horm. Metab. Res. 6, Czech, M. P. (1985) Annu. Rev. Physiol. 47, Shechter, Y. (1982) Endocrinology 110, Thomas, A. P., Martin-Requero, A., and Williamson, J. R. (1985) 18. Gliemann, J., Osterlind, K., Vinten, J., and Gammeltoft, S. (1972) J. Biol. Chem. 260,