- Email: [email protected]
Expression and Regulation by Insulin of Glut 3 in UMR 106-01, a Clonal Rat Osteosarcoma Cell Line
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS ARTICLE NO.
218, 789–793 (1996)
0140
Expression and Regulation by Insulin of Glut 3 in UMR 106-01, a Clonal Rat Osteosarcoma Cell Line D. M. Thomas,*,1 F. Maher,† S. D. Rogers,* and J. D. Best* *University of Melbourne Department of Medicine, St Vincents Hospital, 41 Victoria Parade, Fitzroy 3065, Victoria, Australia; and †Department of Pathology, University of Melbourne, Grattan Street, Parkville, Victoria, Australia Received December 20, 1995 Expression of the glucose transporter GLUT 3 is mainly restricted to neuronal tissues, with lower levels in testis and placenta. In addition, GLUT 3 has recently been reported in neonatal rat calvaria by in situ hybridisation. We report the co-expression of GLUT 1 and GLUT 3 mRNA and protein in UMR 106-01, a clonal osteosarcoma cell line. By semi-quantitative analysis we show that GLUT 3 protein is expressed at levels comparable to GLUT 1. Insulin stimulates glucose uptake in UMR 106-01 cells. GLUT 3 mRNA and protein are increased by chronic (16 h) treatment with insulin, and the increase in GLUT 3 mRNA is not inhibited by cycloheximide. Regulation of GLUT 3 mRNA by insulin has not been previously shown. UMR 106-01 represents a useful model for investigating regulation of GLUT 3 expression. © 1996 Academic Press, Inc.
Physiological expression of the glucose transporter GLUT 3 is restricted principally to neuronal tissues, although low levels have been detected in testis, spermatazoa and placenta (1). The role of GLUT 3 in these tissues is unclear. Both GLUT 1 and GLUT 3 have high affinities for glucose which might provide a growth advantage by ensuring a supply of glucose under limiting conditions. Indeed, hypoglycemia induced GLUT 3 mRNA expression in mouse embryo neuronal cultures (2), and expression of GLUT 3 mRNA correlated with cell growth in CaCo-2 cell lines (3). Furthermore, human tumors have increased levels of GLUT 1 and GLUT 3 mRNAs (4). Bone is a highly metabolically active tissue, but little is known of in vivo glucose transporter expression in bone. Evidence that GLUT 3 may be expressed in bone in vivo comes from in situ studies by Bondy et al. (5) in which membranous calvarial bone in neonatal rat brain is found to express high levels of GLUT 3 mRNA. Very little is known regarding the regulation of GLUT 3 mRNA by insulin. Longo et al. (6) found that insulin had no effect on GLUT 1 or GLUT 3 mRNA expression in cultured human fibroblasts, although phorbol esters induced a protein-synthesis dependent increase in both mRNAs. It was noted that this result might be explained by the low level of insulin receptor expression in these cells (1000 receptors/cell). However, it is known that treatment with 10nM IGF-I for 8 h increases GLUT 3 protein in L6 myotubes (7), and that both insulin and IGF-I act via the IGF-I receptor in this cell line (8). IGF-I and insulin bear considerable structural and functional homology, and exert their effects by common pathways involving activation of tyrosine kinase second messenger systems (9). Insulin is an important anabolic hormone in bone, with insulin deficiency in type I diabetic patients resulting in osteoporosis (10). In order to study the expression and regulation by insulin of GLUT 3 in bone cells in vitro we used UMR 106-01, a rat osteogenic sarcoma cell line which possesses phenotypic markers of the mature osteoblast phenotype, and which expresses abundant insulin receptors (z80,000 receptors/cell) (11). Glucose transport is stimulated by prolonged insulin treatment in UMR 106-01 cells and this effect is due in part in part to increased GLUT 1 mRNA expression, accompanied by increased expression of GLUT 1 protein (12). We also studied UMR 201 and UMR 201-10B, clonal rat calvaria-derived cell lines with a preosteoblastic pheno1
To whom correspondence should be addressed. 789 0006-291X/96 $12.00 Copyright © 1996 by Academic Press, Inc. All rights of reproduction in any form reserved.
Vol. 218, No. 3, 1996
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
type (13). This study demonstrates the presence, relative abundances, and co-ordinate regulation by insulin of both GLUT 1 and GLUT 3 mRNA and protein in UMR 106-01 cells. MATERIALS AND METHODS Materials. Crystalline insulin (100U/ml) was purchased from Novo Nordisk, (Sydney, Australia). [1, 2-3H]2-deoxy-Dglucose (2-DOG) was purchased from New England Nuclear (Boston, MA, USA). [a-32P]-dCTP and [g-32P]-ATP were obtained from Amersham Australia Pty., Ltd. (Sydney, Australia). The 2.6kb EcoRI rat GLUT 1 probe was the gift of Dr. M. Birnbaum (Harvard Medical School, Boston, MA), and the 2kb EcoRI GLUT 4 probe, and rabbit polyclonal antibody for rat GLUT1 were generously provided by Dr. D. James (University of Queensland, Australia). The 1.5kb XhoI/XbaI mouse GLUT 3 probe, 1.5kb EcoRI rat GLUT 2 probe and 2.2kb HindIII/XhoI rat GLUT 5 probe were gifts of Dr. G.I. Bell (University of Chicago, Il, USA). Rabbit polyclonal antibody directed against mouse GLUT 3 (amino acids 472–492) was kindly supplied by Hoffman-LaRoche. A 30-mer oligonucleotide probe for rat 18S was synthesized from published sequences. Cell culture. Cells were incubated at 37°C under 5% CO2 in a-MEM containing 10% FCS at a seeding density of approximately 25000 cells/ml. Medium was changed every 48 h prior to use in experiments, and cells were serum-deprived for 24 h prior to harvest, during which time experiments were carried out. 2-DOG uptake. cells were incubated with HBS containing [1, 2-3H] 2-DOG (1.5mCi, specific activity 26.2 Ci/mmol) for 10 min at 22°C. Three of six wells were exposed to 15mM cytochalasin B (Calbiochem, La Jolla, CA) to block facilitated uptake. Uptake was terminated by washing in ice-cold PBS. Cells were lysed by addition 0.1% SDS and samples taken for scintillation counting and DNA assay. Results are presented as counts per minute corrected for DNA content and passive uptake from cytochalasin B blocked wells. RNA extraction and Northern blot. Cells were grown in 175 cm2 flasks until harvest, when experiments were terminated by the addition of ice-cold PBS. Control tissues were obtained from adult Sprague-Dawley rats. Total RNA was extracted by the method of Chomczynski and Sacchi (14), separated in a 1.5% agarose-formaldehyde denaturing gel and transferred to nylon filters (Hybond-N, Amersham, UK). These filters were then hybridised overnight in SSPE (0.15 M NaCl, 0.01 M NaH2PO4, and 0.001 M EDTA) with radio-labelled cDNA probes. After high stringency washing, filters were subjected to autoradiography at −70°C. Western blot. cells were disrupted in homogenisation buffer (250mM sucrose, 5mM NaN3, 2mM EGTA, 10mM NaHCO3, with 100mM fresh phenylmethylsulphonyl fluoride added daily), and centrifuged at 190000g for 60 min (Beckman XL-90 preparative ultracentrifuge using SW41 Ti rotor) to obtain a total membrane pellet which was resuspended in homogenisation buffer. Standard SDS-polyacrylamide gel electrophoresis (SDS-PAGE) with 12% gels was performed and protein transferred to nitrocellulose filters (Schleicher and Schuell, Dassel, Germany). Filters were incubated with antibody to GLUT 1 or GLUT 3 overnight at room temperature, and assayed by a chemiluminescence detection kit (Boehringer Mannheim, Mannheim, Germany). The autoradiographs were quantitated using ImageQuant software (Molecular Dynamics, Model 300A).
RESULTS AND DISCUSSION Expression and regulation of GLUT 1 and GLUT 3 mRNA. We have previously shown expression of GLUT 1 by UMR 106-01 cells (12). To determine the expression of other glucose transporters, we used probes for rat GLUT 2, mouse GLUT 3, rat GLUT 4 and rat GLUT 5. RNA was obtained from UMR 106-01 and UMR 201-10B cells. UMR 201-10B is a clonal rat calvarial cell line, immortalized with the SV40 large T antigen, with a pre-osteoblastic phenotype (13). Positive control rat tissues (liver, brain, gastrocnemius and small intestine, respectively) were included. We were unable to demonstrate expression in UMR 106-01 and UMR 201-10B cells by Northern blot analysis mRNAs for GLUT 2, GLUT 4, or GLUT 5, but GLUT 3 (4.3kb) was clearly present (Fig. 1A). Interestingly, mRNA for GLUT 3 was also detectable in UMR 201 cells, which are nonimmortalized counterparts of UMR 201-10B cells (data not shown). This further suggests that expression of GLUT 3 in UMR 106-01 and UMR 201-10B cells may be a characteristic of bone derived cells. Insulin stimulates glucose uptake by UMR 106-01 cells and we have shown regulation of GLUT 1 mRNA by insulin (11,12). Therefore, we next determined whether GLUT 3 mRNA was regulated by insulin. Treatment of UMR 106-01 cells with 1mM insulin results in a time-dependent increase in both GLUT 3 and GLUT 1 mRNA, with maximal effect within 4 h and a sustained 2-fold increase at 24 h (Fig 1B). We found similar results with 10 nM insulin suggesting that this effect was mediated by the insulin receptor (data not shown). The effect was not dependent on de novo 790
Vol. 218, No. 3, 1996
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
FIG. 1. Glucose transporter mRNA expression and regulation by insulin. A: 20 mg/lane total RNA from UMR 106-01 (01), UMR 201-10B (02), rat liver (L), rat brain (B), rat gastrocnemius (G) and rat jejunum (J) was subjected to Northern blot analysis as described in Materials and Methods. Filters were probed with cDNA probes to rat GLUT 2, GLUT 4 and GLUT 5, and mouse GLUT 3. B: representative time course of insulin (1 mM) action on GLUT 1 and GLUT 3 mRNA in UMR 106-01 cells. 20 mg/lane total RNA was loaded. Densitometric data was normalised for 18S. This experiment was repeated twice with similar results. C: effect of cycloheximide (10 mg/ml) on insulin (10 nM, 8 h) stimulated GLUT 1 and GLUT 3 mRNA expression in UMR 106-01 cells,. B: basal; I: insulin; C: cycloheximide. Densitometric data of GLUT 3 mRNA was obtained from five experiments and normalised with 18S.
protein synthesis, as treatment with 10 mg/ml cycloheximide did not inhibit insulin-stimulated expression of GLUT 3 mRNA, although cycloheximide treatment independently had a small stimulatory effect (Fig 1C). This is of interest, because insulin-stimulation of GLUT 1 mRNA is inhibited by cycloheximide under the same conditions. Expression and regulation of GLUT 1 and GLUT 3 protein. It has previously been shown that expression of mRNA for GLUT 3 does not necessarily imply expression of GLUT 3 protein (15). We used protein derived from cerebellar granular neurone (CGN) membranes with known quan791
Vol. 218, No. 3, 1996
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
tities of GLUT 1 and GLUT 3 protein (3 and 18pmol/mg protein, respectively) (16) to enable estimation of the relative abundances of each isoform in UMR 106-01 cells. Approximately equal amounts of GLUT 1 protein but about one-tenth the amount of GLUT 3 protein were found per mg total membrane protein in UMR 106-01 cells compared to the neuroglial control (Fig 2A). The estimate that approximately 3pmol/mg protein GLUT 1 and 1.7pmol/mg protein GLUT 3 was present in UMR 106-01 cells suggests that both isoforms contribute significantly to total cellular glucose uptake. Treatment of UMR 106-01 cells with 10nM insulin for 16 h induced a co-ordinated increase in cellular GLUT 1 and GLUT 3 protein of approximately 2-fold (Fig 2B). This increment is in proportion to the observed increase in GLUT 3 mRNA, although less than the observed 3-fold increase in cellular uptake of 2-DOG (Fig 2C). The difference may be due to insulin causing translocation of both GLUT 1 and GLUT 3 protein to the plasma membrane, as previously described in L6 cells (12). Conclusions. The clonal bone-derived cell line UMR 106-01 express both mRNA and protein for GLUT 1 and GLUT 3. Unlike neuronal GLUT 3 protein, which is 5- to 8-times more abundant than GLUT 1, similar amounts of GLUT 1 and GLUT 3 protein are expressed in UMR 106-01 cells. We show regulation of both GLUT 1 and GLUT 3 mRNA and protein by insulin. These observations contribute to the notion that GLUT 1 and GLUT 3 have similar physiological roles. Both transporters have high affinities for glucose, and reside principally at the plasma membrane in the resting state (17). Both GLUT 1 and GLUT 3 are found over-expressed in a variety of tumours (4), and both are developmentally regulated (5). However differences are apparent. We have found that whereas insulin stimulated-GLUT 1 mRNA is cycloheximide inhibitable, this is not the case for insulin-stimulated GLUT 3 mRNA. This interesting finding is the subject of further study. A further difference between GLUT 1 and GLUT 3 lies in the tissue distribution, with expression of
FIG. 2. Expression and regulation of GLUT 1 and GLUT 3 protein. A: representative Western blot quantitation of GLUT 3 vs GLUT 1 protein in UMR 106-01 cells. 100 mg/lane rat cerebellar granule neurones (CGN) protein with a known concentration of GLUT 3 (18pmol/mg) and GLUT 1 (3pmol/mg) protein, and 100 mg/lane of UMR 106-01 cell protein were subjected to SDS-12%PAGE in duplicate and Western blotted as described in Materials and Methods. Filters were incubated with antibody to GLUT 3 (1/1000) and GLUT 1 (1/500). This experiment was performed three times with similar results. B: representative Western blot of insulin (INS: 10 nM, 16 h) on GLUT 1 and GLUT 3 protein in UMR 106-01 cells. 50 mg/lane protein was loaded. Densitometric data present the mean and SEM of three experiments. B: basal; I: insulin. C: Effect of insulin on 2-deoxyglucose uptake. UMR 106-01 cells were treated with 10 nM insulin for 16 h under serum-free conditions. 2-deoxyglucose uptake was assayed as described in Materials and Methods. Data are the mean and SEM of 6–9 samples each from three experiments. 792
Vol. 218, No. 3, 1996
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
GLUT 3 being much more restricted than GLUT 1. The previous demonstration of GLUT 3 mRNA in rat calvariae by in situ hybridisation accords with our observation that GLUT 3 mRNA is present in both mature and immature, transformed, immortalised and non-transformed osteoblastic cell lines, and suggests that this is not a feature of in vitro culture. With the paucity of non-neuronal cell lines which express GLUT 3 protein, the expression and regulation of GLUT 3 in UMR 106-01 cells make this cell line a useful tool for future studies. REFERENCES 1. Haber, R. S., Weinstein, S. P., O’Boyle, E., and Morgello, S. (1993) Endocrinology 132, 2538–2543. 2. Nagamatsu, S., Sawa, H., Inoue, N., Nakamichi, Y., Takeshima, H., and Hoshino, T. (1994) Biochem. J. 300, 125–131. 3. Mahroui, L., Rodolosse, A., Barbat, A., Dussaulx, E., Zweibaum, A., Rousset, M., and Brot-Laroche, E. (1994) Biochem. J. 298, 629–633. 4. Yamamoto, T., Seino, Y., Fukumoto, H., Koh, G., Inagaki, N., Yamada, Y., Inoue, K., Manabe, T., and Imura, I. (1990) Biochem. Biophys. Res. Com. 170, 223–230. 5. Bondy, C. A., Lee, W.-H., and Zhou, J. (1992) Mol. Cell. Neurosciences 3, 305–314. 6. Longo, N., Griffin, L. D., Langley, S. D., and Elsas, L. J. (1992) Biochim. Biophys. Acta 1104, 24–30. 7. Bilan, P. J., Mitsumoto, Y., Maher, F., Simpson, I. A., and Klip, A. (1992) Biochem. Biophys. Res. Com. 186, 1129–1137. 8. Maher, F., Clark, S., and Harrison, L. (1989) Mol. Endocrinology 3, 2128–2135. 9. Van Obberghen, E. (1994) Diabetologia 37, S125–S134. 10. Bouillon, R. (1991) Calcified Tissue International 49, 155–160. 11. Ituarte, E. A., Ituarte, H. G., Iida-Klein, A., and Hahn, T. J. (1989) J. Bone Min. Res. 4, 69–73. 12. Thomas, D. M., Rogers, S. D., Sleeman, M. W., Pasquini, G. M., Bringhurst, F. R., Ng, K. W., Zajac, J. D., and Best, J. D. (1995) J. Mol. Endocrinology 14, 263–275. 13. Zhou, H., Hammonds, R. G. Jr., Findlay, D. M., Fuller, P. J., Martin, T. J., and Ng, K. W. (1991) J. Bone Min. Res. 6, 767–777. 14. Chomczynski, P., and Sacchi, N. (1987) Anal. Biochem. 162, 156–159. 15. Maher, F., Vanucci, S., Takeda, J., and Simpson, I. A. (1992) Biochem. Biophys. Res. Com. 182, 703–711. 16. Maher, F., and Simpson, I. A. (1994) Biochem. J. 301, 379–384. 17. Wilson, C. M., Mitsumoto, Y., Maher, F., and Klip, A. (1995) FEBS Letters 368, 19–22.
793
218, 789–793 (1996)
0140
Expression and Regulation by Insulin of Glut 3 in UMR 106-01, a Clonal Rat Osteosarcoma Cell Line D. M. Thomas,*,1 F. Maher,† S. D. Rogers,* and J. D. Best* *University of Melbourne Department of Medicine, St Vincents Hospital, 41 Victoria Parade, Fitzroy 3065, Victoria, Australia; and †Department of Pathology, University of Melbourne, Grattan Street, Parkville, Victoria, Australia Received December 20, 1995 Expression of the glucose transporter GLUT 3 is mainly restricted to neuronal tissues, with lower levels in testis and placenta. In addition, GLUT 3 has recently been reported in neonatal rat calvaria by in situ hybridisation. We report the co-expression of GLUT 1 and GLUT 3 mRNA and protein in UMR 106-01, a clonal osteosarcoma cell line. By semi-quantitative analysis we show that GLUT 3 protein is expressed at levels comparable to GLUT 1. Insulin stimulates glucose uptake in UMR 106-01 cells. GLUT 3 mRNA and protein are increased by chronic (16 h) treatment with insulin, and the increase in GLUT 3 mRNA is not inhibited by cycloheximide. Regulation of GLUT 3 mRNA by insulin has not been previously shown. UMR 106-01 represents a useful model for investigating regulation of GLUT 3 expression. © 1996 Academic Press, Inc.
Physiological expression of the glucose transporter GLUT 3 is restricted principally to neuronal tissues, although low levels have been detected in testis, spermatazoa and placenta (1). The role of GLUT 3 in these tissues is unclear. Both GLUT 1 and GLUT 3 have high affinities for glucose which might provide a growth advantage by ensuring a supply of glucose under limiting conditions. Indeed, hypoglycemia induced GLUT 3 mRNA expression in mouse embryo neuronal cultures (2), and expression of GLUT 3 mRNA correlated with cell growth in CaCo-2 cell lines (3). Furthermore, human tumors have increased levels of GLUT 1 and GLUT 3 mRNAs (4). Bone is a highly metabolically active tissue, but little is known of in vivo glucose transporter expression in bone. Evidence that GLUT 3 may be expressed in bone in vivo comes from in situ studies by Bondy et al. (5) in which membranous calvarial bone in neonatal rat brain is found to express high levels of GLUT 3 mRNA. Very little is known regarding the regulation of GLUT 3 mRNA by insulin. Longo et al. (6) found that insulin had no effect on GLUT 1 or GLUT 3 mRNA expression in cultured human fibroblasts, although phorbol esters induced a protein-synthesis dependent increase in both mRNAs. It was noted that this result might be explained by the low level of insulin receptor expression in these cells (1000 receptors/cell). However, it is known that treatment with 10nM IGF-I for 8 h increases GLUT 3 protein in L6 myotubes (7), and that both insulin and IGF-I act via the IGF-I receptor in this cell line (8). IGF-I and insulin bear considerable structural and functional homology, and exert their effects by common pathways involving activation of tyrosine kinase second messenger systems (9). Insulin is an important anabolic hormone in bone, with insulin deficiency in type I diabetic patients resulting in osteoporosis (10). In order to study the expression and regulation by insulin of GLUT 3 in bone cells in vitro we used UMR 106-01, a rat osteogenic sarcoma cell line which possesses phenotypic markers of the mature osteoblast phenotype, and which expresses abundant insulin receptors (z80,000 receptors/cell) (11). Glucose transport is stimulated by prolonged insulin treatment in UMR 106-01 cells and this effect is due in part in part to increased GLUT 1 mRNA expression, accompanied by increased expression of GLUT 1 protein (12). We also studied UMR 201 and UMR 201-10B, clonal rat calvaria-derived cell lines with a preosteoblastic pheno1
To whom correspondence should be addressed. 789 0006-291X/96 $12.00 Copyright © 1996 by Academic Press, Inc. All rights of reproduction in any form reserved.
Vol. 218, No. 3, 1996
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
type (13). This study demonstrates the presence, relative abundances, and co-ordinate regulation by insulin of both GLUT 1 and GLUT 3 mRNA and protein in UMR 106-01 cells. MATERIALS AND METHODS Materials. Crystalline insulin (100U/ml) was purchased from Novo Nordisk, (Sydney, Australia). [1, 2-3H]2-deoxy-Dglucose (2-DOG) was purchased from New England Nuclear (Boston, MA, USA). [a-32P]-dCTP and [g-32P]-ATP were obtained from Amersham Australia Pty., Ltd. (Sydney, Australia). The 2.6kb EcoRI rat GLUT 1 probe was the gift of Dr. M. Birnbaum (Harvard Medical School, Boston, MA), and the 2kb EcoRI GLUT 4 probe, and rabbit polyclonal antibody for rat GLUT1 were generously provided by Dr. D. James (University of Queensland, Australia). The 1.5kb XhoI/XbaI mouse GLUT 3 probe, 1.5kb EcoRI rat GLUT 2 probe and 2.2kb HindIII/XhoI rat GLUT 5 probe were gifts of Dr. G.I. Bell (University of Chicago, Il, USA). Rabbit polyclonal antibody directed against mouse GLUT 3 (amino acids 472–492) was kindly supplied by Hoffman-LaRoche. A 30-mer oligonucleotide probe for rat 18S was synthesized from published sequences. Cell culture. Cells were incubated at 37°C under 5% CO2 in a-MEM containing 10% FCS at a seeding density of approximately 25000 cells/ml. Medium was changed every 48 h prior to use in experiments, and cells were serum-deprived for 24 h prior to harvest, during which time experiments were carried out. 2-DOG uptake. cells were incubated with HBS containing [1, 2-3H] 2-DOG (1.5mCi, specific activity 26.2 Ci/mmol) for 10 min at 22°C. Three of six wells were exposed to 15mM cytochalasin B (Calbiochem, La Jolla, CA) to block facilitated uptake. Uptake was terminated by washing in ice-cold PBS. Cells were lysed by addition 0.1% SDS and samples taken for scintillation counting and DNA assay. Results are presented as counts per minute corrected for DNA content and passive uptake from cytochalasin B blocked wells. RNA extraction and Northern blot. Cells were grown in 175 cm2 flasks until harvest, when experiments were terminated by the addition of ice-cold PBS. Control tissues were obtained from adult Sprague-Dawley rats. Total RNA was extracted by the method of Chomczynski and Sacchi (14), separated in a 1.5% agarose-formaldehyde denaturing gel and transferred to nylon filters (Hybond-N, Amersham, UK). These filters were then hybridised overnight in SSPE (0.15 M NaCl, 0.01 M NaH2PO4, and 0.001 M EDTA) with radio-labelled cDNA probes. After high stringency washing, filters were subjected to autoradiography at −70°C. Western blot. cells were disrupted in homogenisation buffer (250mM sucrose, 5mM NaN3, 2mM EGTA, 10mM NaHCO3, with 100mM fresh phenylmethylsulphonyl fluoride added daily), and centrifuged at 190000g for 60 min (Beckman XL-90 preparative ultracentrifuge using SW41 Ti rotor) to obtain a total membrane pellet which was resuspended in homogenisation buffer. Standard SDS-polyacrylamide gel electrophoresis (SDS-PAGE) with 12% gels was performed and protein transferred to nitrocellulose filters (Schleicher and Schuell, Dassel, Germany). Filters were incubated with antibody to GLUT 1 or GLUT 3 overnight at room temperature, and assayed by a chemiluminescence detection kit (Boehringer Mannheim, Mannheim, Germany). The autoradiographs were quantitated using ImageQuant software (Molecular Dynamics, Model 300A).
RESULTS AND DISCUSSION Expression and regulation of GLUT 1 and GLUT 3 mRNA. We have previously shown expression of GLUT 1 by UMR 106-01 cells (12). To determine the expression of other glucose transporters, we used probes for rat GLUT 2, mouse GLUT 3, rat GLUT 4 and rat GLUT 5. RNA was obtained from UMR 106-01 and UMR 201-10B cells. UMR 201-10B is a clonal rat calvarial cell line, immortalized with the SV40 large T antigen, with a pre-osteoblastic phenotype (13). Positive control rat tissues (liver, brain, gastrocnemius and small intestine, respectively) were included. We were unable to demonstrate expression in UMR 106-01 and UMR 201-10B cells by Northern blot analysis mRNAs for GLUT 2, GLUT 4, or GLUT 5, but GLUT 3 (4.3kb) was clearly present (Fig. 1A). Interestingly, mRNA for GLUT 3 was also detectable in UMR 201 cells, which are nonimmortalized counterparts of UMR 201-10B cells (data not shown). This further suggests that expression of GLUT 3 in UMR 106-01 and UMR 201-10B cells may be a characteristic of bone derived cells. Insulin stimulates glucose uptake by UMR 106-01 cells and we have shown regulation of GLUT 1 mRNA by insulin (11,12). Therefore, we next determined whether GLUT 3 mRNA was regulated by insulin. Treatment of UMR 106-01 cells with 1mM insulin results in a time-dependent increase in both GLUT 3 and GLUT 1 mRNA, with maximal effect within 4 h and a sustained 2-fold increase at 24 h (Fig 1B). We found similar results with 10 nM insulin suggesting that this effect was mediated by the insulin receptor (data not shown). The effect was not dependent on de novo 790
Vol. 218, No. 3, 1996
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
FIG. 1. Glucose transporter mRNA expression and regulation by insulin. A: 20 mg/lane total RNA from UMR 106-01 (01), UMR 201-10B (02), rat liver (L), rat brain (B), rat gastrocnemius (G) and rat jejunum (J) was subjected to Northern blot analysis as described in Materials and Methods. Filters were probed with cDNA probes to rat GLUT 2, GLUT 4 and GLUT 5, and mouse GLUT 3. B: representative time course of insulin (1 mM) action on GLUT 1 and GLUT 3 mRNA in UMR 106-01 cells. 20 mg/lane total RNA was loaded. Densitometric data was normalised for 18S. This experiment was repeated twice with similar results. C: effect of cycloheximide (10 mg/ml) on insulin (10 nM, 8 h) stimulated GLUT 1 and GLUT 3 mRNA expression in UMR 106-01 cells,. B: basal; I: insulin; C: cycloheximide. Densitometric data of GLUT 3 mRNA was obtained from five experiments and normalised with 18S.
protein synthesis, as treatment with 10 mg/ml cycloheximide did not inhibit insulin-stimulated expression of GLUT 3 mRNA, although cycloheximide treatment independently had a small stimulatory effect (Fig 1C). This is of interest, because insulin-stimulation of GLUT 1 mRNA is inhibited by cycloheximide under the same conditions. Expression and regulation of GLUT 1 and GLUT 3 protein. It has previously been shown that expression of mRNA for GLUT 3 does not necessarily imply expression of GLUT 3 protein (15). We used protein derived from cerebellar granular neurone (CGN) membranes with known quan791
Vol. 218, No. 3, 1996
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
tities of GLUT 1 and GLUT 3 protein (3 and 18pmol/mg protein, respectively) (16) to enable estimation of the relative abundances of each isoform in UMR 106-01 cells. Approximately equal amounts of GLUT 1 protein but about one-tenth the amount of GLUT 3 protein were found per mg total membrane protein in UMR 106-01 cells compared to the neuroglial control (Fig 2A). The estimate that approximately 3pmol/mg protein GLUT 1 and 1.7pmol/mg protein GLUT 3 was present in UMR 106-01 cells suggests that both isoforms contribute significantly to total cellular glucose uptake. Treatment of UMR 106-01 cells with 10nM insulin for 16 h induced a co-ordinated increase in cellular GLUT 1 and GLUT 3 protein of approximately 2-fold (Fig 2B). This increment is in proportion to the observed increase in GLUT 3 mRNA, although less than the observed 3-fold increase in cellular uptake of 2-DOG (Fig 2C). The difference may be due to insulin causing translocation of both GLUT 1 and GLUT 3 protein to the plasma membrane, as previously described in L6 cells (12). Conclusions. The clonal bone-derived cell line UMR 106-01 express both mRNA and protein for GLUT 1 and GLUT 3. Unlike neuronal GLUT 3 protein, which is 5- to 8-times more abundant than GLUT 1, similar amounts of GLUT 1 and GLUT 3 protein are expressed in UMR 106-01 cells. We show regulation of both GLUT 1 and GLUT 3 mRNA and protein by insulin. These observations contribute to the notion that GLUT 1 and GLUT 3 have similar physiological roles. Both transporters have high affinities for glucose, and reside principally at the plasma membrane in the resting state (17). Both GLUT 1 and GLUT 3 are found over-expressed in a variety of tumours (4), and both are developmentally regulated (5). However differences are apparent. We have found that whereas insulin stimulated-GLUT 1 mRNA is cycloheximide inhibitable, this is not the case for insulin-stimulated GLUT 3 mRNA. This interesting finding is the subject of further study. A further difference between GLUT 1 and GLUT 3 lies in the tissue distribution, with expression of
FIG. 2. Expression and regulation of GLUT 1 and GLUT 3 protein. A: representative Western blot quantitation of GLUT 3 vs GLUT 1 protein in UMR 106-01 cells. 100 mg/lane rat cerebellar granule neurones (CGN) protein with a known concentration of GLUT 3 (18pmol/mg) and GLUT 1 (3pmol/mg) protein, and 100 mg/lane of UMR 106-01 cell protein were subjected to SDS-12%PAGE in duplicate and Western blotted as described in Materials and Methods. Filters were incubated with antibody to GLUT 3 (1/1000) and GLUT 1 (1/500). This experiment was performed three times with similar results. B: representative Western blot of insulin (INS: 10 nM, 16 h) on GLUT 1 and GLUT 3 protein in UMR 106-01 cells. 50 mg/lane protein was loaded. Densitometric data present the mean and SEM of three experiments. B: basal; I: insulin. C: Effect of insulin on 2-deoxyglucose uptake. UMR 106-01 cells were treated with 10 nM insulin for 16 h under serum-free conditions. 2-deoxyglucose uptake was assayed as described in Materials and Methods. Data are the mean and SEM of 6–9 samples each from three experiments. 792
Vol. 218, No. 3, 1996
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
GLUT 3 being much more restricted than GLUT 1. The previous demonstration of GLUT 3 mRNA in rat calvariae by in situ hybridisation accords with our observation that GLUT 3 mRNA is present in both mature and immature, transformed, immortalised and non-transformed osteoblastic cell lines, and suggests that this is not a feature of in vitro culture. With the paucity of non-neuronal cell lines which express GLUT 3 protein, the expression and regulation of GLUT 3 in UMR 106-01 cells make this cell line a useful tool for future studies. REFERENCES 1. Haber, R. S., Weinstein, S. P., O’Boyle, E., and Morgello, S. (1993) Endocrinology 132, 2538–2543. 2. Nagamatsu, S., Sawa, H., Inoue, N., Nakamichi, Y., Takeshima, H., and Hoshino, T. (1994) Biochem. J. 300, 125–131. 3. Mahroui, L., Rodolosse, A., Barbat, A., Dussaulx, E., Zweibaum, A., Rousset, M., and Brot-Laroche, E. (1994) Biochem. J. 298, 629–633. 4. Yamamoto, T., Seino, Y., Fukumoto, H., Koh, G., Inagaki, N., Yamada, Y., Inoue, K., Manabe, T., and Imura, I. (1990) Biochem. Biophys. Res. Com. 170, 223–230. 5. Bondy, C. A., Lee, W.-H., and Zhou, J. (1992) Mol. Cell. Neurosciences 3, 305–314. 6. Longo, N., Griffin, L. D., Langley, S. D., and Elsas, L. J. (1992) Biochim. Biophys. Acta 1104, 24–30. 7. Bilan, P. J., Mitsumoto, Y., Maher, F., Simpson, I. A., and Klip, A. (1992) Biochem. Biophys. Res. Com. 186, 1129–1137. 8. Maher, F., Clark, S., and Harrison, L. (1989) Mol. Endocrinology 3, 2128–2135. 9. Van Obberghen, E. (1994) Diabetologia 37, S125–S134. 10. Bouillon, R. (1991) Calcified Tissue International 49, 155–160. 11. Ituarte, E. A., Ituarte, H. G., Iida-Klein, A., and Hahn, T. J. (1989) J. Bone Min. Res. 4, 69–73. 12. Thomas, D. M., Rogers, S. D., Sleeman, M. W., Pasquini, G. M., Bringhurst, F. R., Ng, K. W., Zajac, J. D., and Best, J. D. (1995) J. Mol. Endocrinology 14, 263–275. 13. Zhou, H., Hammonds, R. G. Jr., Findlay, D. M., Fuller, P. J., Martin, T. J., and Ng, K. W. (1991) J. Bone Min. Res. 6, 767–777. 14. Chomczynski, P., and Sacchi, N. (1987) Anal. Biochem. 162, 156–159. 15. Maher, F., Vanucci, S., Takeda, J., and Simpson, I. A. (1992) Biochem. Biophys. Res. Com. 182, 703–711. 16. Maher, F., and Simpson, I. A. (1994) Biochem. J. 301, 379–384. 17. Wilson, C. M., Mitsumoto, Y., Maher, F., and Klip, A. (1995) FEBS Letters 368, 19–22.
793