Glucose as Regulator of Glucose Transport Activity and Glucose-Transporter mRNA in Hamster p-Cell Line NOBUYA INAGAKI, KOICHIRO YASUDA, GEN INOUE, YOSHIMASA OKAMOTO, HIDEKI YANO, YOSHIMICHI SOMEYA, YASUKAZU OHMOTO, KYOHEI DEGUCHI, KEN-ICHI IMAGAWA, HIROO IMURA, AND YUTAKA SEINO
To investigate the role of glucose in regulating glucose transporters in pancreatic p-cells, we studied the hamster clonal p-cell line HIT-T15, which retains responsiveness to glucose. Northern blot analysis demonstrates that GLUT2 and GLUT1 mRNA are abundant in HIT cells. After a 24-h culture with various concentrations of glucose (0-22.2 mM [0-400 mg/dl]), the GLUT2 mRNA level in HIT cells increased by 40% at 22.2 mM (400 mg/dl) glucose compared with 11.1 mM (200 mg/dl) without a change in mRNA stability. It also decreased proportionally to the reduction of glucose concentration. Glucose deprivation resulted in a decrease of GLUT2 mRNA to an almost undetectable level, with a marked increase in the degradation rate of mRNA. In contrast, the GLUT1 mRNA was not affected by glucose. We show that glucose uptake is highest in HIT cells incubated at 2.8-5.5 mM (50-99 mg/dl) glucose for 24 h, and that levels in cells cultured at 0 mM (0 mg/dl) and 22.2 mM (400 mg/dl) glucose decrease to - 2 0 % of the maximum level. This decrease is consistent with the effects of glucose on glucose-stimulated insulin secretion in HIT cells. Our results indicate that glucose is involved in regulating GLUT2 mRNA and glucose uptake activity and that the glucose responsiveness of the insulin secretion correlates with the glucose-induced change in glucose uptake activity in HIT cells. Diabetes 41:592-97, 1992
From the Second Division, Departments of Medicine, Metabolism, and Clinical Nutrition, Kyoto University Faculty of Medicine, Kyoto; Cellular Technology Institute, and Department of Viral Diseases, Tokushima Research Institute, Otsuka Pharmaceutical Company, Limited, Tokushima, Japan. Address correspondence and reprint requests to Nobuya Inagaki, MD, Second Division, Department of Medicine, Kyoto University School of Medicine, 54 Shogoin Kawahara-cho, Sakyo-ku, Kyoto 606, Japan. Received for publication 23 May 1991 and accepted in revised form 2 January 1992.
592
R
ecent studies revealed the existence of a family of glucose transporters with distinct functional properties but related primary sequences (1,2). Among the five facilitative glucose-transporter isoforms thus far identified, GLUT2 (liver-type glucose transporter) (3-6) is highly expressed in pancreatic p-cells and hepatocytes (3-6). GLUT2 is a unique transporter because it has a low affinity and a high Km for glucose (2,7). Glucose uptake into the p-cells is mediated by this high-K"m glucose-transporter isoform and is increased proportionately to an increase of extracellular glucose concentration. Thus, it has been suggested that GLUT2 plays an important role in glucose sensing by p-cells (8-12). Recently, several studies showed a strong correlation between the reduction of GLUT2 in diabetic islets and the level of hyperglycemia in rat models of diabetes (9-11). Consistent with this, clonal insulinoma cell lines such as RINm5F, which is derived from a rat insulinoma and lacks insulin secretory responsiveness to glucose, express large amounts of low-/11.1 mM (200 mg/dl) of medium glucose concentration was accompanied by a six- and fourfold reduction in 2-DG uptake, respectively (Fig. 4). However, in the other rat p-cell clone, RINr, which expresses GLUT1 but not GLUT2, maximum 2-DG uptake was observed after 24-h culture in a glucose-free medium (data not shown). Figure 5 shows the glucose responsiveness of insulin release of HIT cells after 24-h culture at 0, 5.5, or 22.2 mM (0, 99, or 400 mg/dl) glucose. The maximum glucose responsiveness was observed between 0 and 11.1 mM (0 and 200 mg/dl) glucose in HIT cells cultured at 5.5 mM (99 mg/dl) glucose for 24 h (Fig. 5). The insulin secretion of HIT cells cultured at 0 mM glucose for 24 h was significantly increased at 5.5 mM (99 mg/dl) glucose compared with that at 0 mM glucose, but no significant increase was observed above 5.5 mM (99 mg/dl) glucose (Fig. 5). On the other hand, the HIT cells cultured at 22.2 mM (400 mg/dl) glucose for 24 h were not respon-
DIABETES, VOL. 41, MAY 1992
N. INAGAKI AND ASSOCIATES
TABLE 1 Relative change of GLUT2 and GLUT1 mRNA levels in cultured HIT cells with various glucose concentrations Glucose concentration (mM)
GLUT2 mRNA GLUT1 mRNA
0
2.8
5.5
8.3
11.1
22 .2
6 ± 4* 107 ± 14
59 ± 16t 113±5
81 ± 10 103 ± 13
97 ± 8 105 ± 15
100 100
139 ±: 14t 117± : 16
Quantitative laser scanning densitometry of Northern blot of GLUT2 and GLUT1 mRNA in HIT cells after 24-h culture at 0-22.2 mM glucose. Amount of mRNA is percentage of value at 11.1 mM glucose. Values are means ± SE of 3 experiments. *P < 0.01, tP < 0.05 vs. control (11.1 mM glucose).
sive to glucose stimulation (Fig. 5). In addition, a glucose concentration >11.1 mM (200 mg/dl) depressed the maximum response of HIT cells cultured at 5.5 mM (99 mg/dl) glucose for 24 h, as reported previously (14,18; Fig. 5). DISCUSSION
Our results show that HIT cells that retain responsiveness to glucose stimulation express high levels of GLUT2 and GLUT1 mRNA and that glucose-transport activity in HIT cells is regulated by glucose concentration. However, we could detect no GLUT2 mRNA in RINr cells, which lack insulin secretory responsiveness to glucose. These results suggest that GLUT2 plays an important role in glucose sensing. We studied the role of glucose in the regulation of the amounts of GLUT2 and GLUT1 mRNAs, and glucose uptake activity in HIT cells. The GLUT1 mRNA level was not affected by differing concentrations of glucose. By contrast, glucose altered GLUT2 mRNA level and glucose uptake activity in a different way. The GLUT2 mRNA level in HIT cells incubated at 22.2 mM (400 mg/dl) glucose was higher than that in cells incubated at 11.1 mM (200 mg/dl) glucose, whereas GLUT2 mRNA stability determined with actinomycin D was similar at both glucose concentrations (t1/2 = 7-9 h). This result suggests that the increase of GLUT2 mRNA at 22.2 mM (400 mg/dl) glucose could be caused by increased transcription rates rather than stabilization of GLUT2 mRNA. On
the other hand, the GLUT2 mRNA level decreased proportionately to the reduction of the glucose concentration. The GLUT2 mRNA level in HIT cells was almost undetectable after 24-h glucose deprivation, which may be attributable to the marked increase in the degradation rate of GLUT2 mRNA by glucose deprivation (t1/2 = 2 h). However, whether the rates of GLUT2 mRNA transcription have a role in altering GLUT2 mRNA levels in HIT cells cultured in the medium containing various concentrations of glucose is unclear because direct measurements of GLUT2 mRNA transcription have yet to be performed. Chen et al. (8) showed that the p-cell GLUT2 mRNA level is increased 46% by prolonged hyperglycemic clamping of rats at 11.2 mM (202 mg/dl) for 5 days, whereas p-cells from rats kept hypoglycemic by continuous infusion of insulin lose expression of GLUT2 mRNA and high-Km glucose transport and display a strong reduction in glucose-induced insulin secretion. The similar effect of glucose on the GLUT2 mRNA level in vivo in their experiments and in vitro in ours suggests that glucose regulates GLUT2 mRNA level. c 30
o a E C
TABLE 2 Effects of 0, 11.1, and 22.2 mM glucose on degradation rate of GLUT2 mRNA Time (h)
Glucose concentration (mM) 0 11.1 22.2
0
4
100 100 100
23 ± 1* 66 ± 1 70 ± 5
8 10 ± 1* 54 -t- 5 47 + 3
Degradation rates were determined with actinomycin D treatment. HIT cells cultured in RPM11640 at 11.1 mM glucose were washed and exposed to serum-free RPMI 1640 medium containing 0,11.1, or 22.2 mM glucose for 2 h. Then actinomycin D (5 M-g/ml, final concentration) was added and incubation continued for 8 h. Changes in GLUT2 mRNA were determined at 0, 4, and 8 h by dot-blot analysis and are expressed as percentage of value without actinomycin D. Values are means ± SE of 3 experiments. *P< 0.01 vs. control (11.1 mM).
DIABETES, VOL 41, MAY 1992
20
i "5 a 10CO
a O Q CNI
2.8
5.5
8.3
11.1
22.2
[Glucose], mM FIG. 4. Effects of different glucose concentrations on 2-deoxy [3H]glucose uptake. After maintenance of HIT cells at indicated glucose for 24 h, 2-deoxy[3H]glucose uptake activity was measured as described in METHODS. Values are means ± SE of 3 experiments. *P < 0.01 vs. value at 5.5 mM glucose.
595
GLUCOSE-TRANSPORTER EXPRESSION IN HIT CELLS
mM glucose
0
5.5
22.2
[Glucose], mM of 24-h culture medium FIG. 5. Glucose-stimulated insulin secretion of HIT cells cultured with different concentrations of glucose for 24 h. After 24-h culture at 0, 5.5, or 22.2 mM glucose, Insulin release from HIT cells in response to 0, 5.5,11.1, or 22.2 mM glucose was measured. Values are means ± SE of 6 experiments. *P < 0.02 vs. value stimulated at 0 mM glucose. f P < 0.01 vs. value stimulated at 5.5 mM glucose. i P < 0.01 vs. value stimulated at 11.1 mM glucose.
Because GLUT1 protein was not affected by glucose, and neither GLUT3 nor GLUT4 mRNA was detected in HIT cells (data not shown), it is likely that the alteration of glucose uptake activity in HIT cells reflects change in the GLUT2 protein level. The reason for the discrepancy between the GLUT2 mRNA and the glucose uptake activity at >5.5 mM (99 mg/dl) glucose concentration remains to be explained; it is not known whether the rate of GLUT2 mRNA translation decreases or the rate of GLUT2 protein degradation increases in HIT cells cultured at a high glucose concentration. However, the relationship between the protein level and the glucose uptake activity of GLUT1 and GLUT2 needs to be determined separately. The maximum glucose responsiveness of insulin secretion was observed in 24-h cultured HIT cells at 5.5 mM (99 mg/dl) glucose, in which glucose uptake activity is maximum (Fig. 5). The HIT cells cultured at 22.2 mM (400 mg/dl) glucose, in which glucose uptake activity is decreased, were not responsive to glucose (Fig. 5). Of course, although 2-DG uptake was linear for at least 20 min (data not shown) and glucose transport is rate limiting compared with glucose phosphorylation (13,18) in HIT cells, the situation in islets is quite different. Islets equilibrate glucose in
To investigate the role of glucose in regulating glucose transporters in pancreatic p-cells, we studied the hamster clonal p-cell line HIT-T15, which retains responsiveness to glucose. Northern blot analysis demonstrates that GLUT2 and GLUT1 mRNA are abundant in HIT cells. After a 24-h culture with various concentrations of glucose (0-22.2 mM [0-400 mg/dl]), the GLUT2 mRNA level in HIT cells increased by 40% at 22.2 mM (400 mg/dl) glucose compared with 11.1 mM (200 mg/dl) without a change in mRNA stability. It also decreased proportionally to the reduction of glucose concentration. Glucose deprivation resulted in a decrease of GLUT2 mRNA to an almost undetectable level, with a marked increase in the degradation rate of mRNA. In contrast, the GLUT1 mRNA was not affected by glucose. We show that glucose uptake is highest in HIT cells incubated at 2.8-5.5 mM (50-99 mg/dl) glucose for 24 h, and that levels in cells cultured at 0 mM (0 mg/dl) and 22.2 mM (400 mg/dl) glucose decrease to - 2 0 % of the maximum level. This decrease is consistent with the effects of glucose on glucose-stimulated insulin secretion in HIT cells. Our results indicate that glucose is involved in regulating GLUT2 mRNA and glucose uptake activity and that the glucose responsiveness of the insulin secretion correlates with the glucose-induced change in glucose uptake activity in HIT cells. Diabetes 41:592-97, 1992
From the Second Division, Departments of Medicine, Metabolism, and Clinical Nutrition, Kyoto University Faculty of Medicine, Kyoto; Cellular Technology Institute, and Department of Viral Diseases, Tokushima Research Institute, Otsuka Pharmaceutical Company, Limited, Tokushima, Japan. Address correspondence and reprint requests to Nobuya Inagaki, MD, Second Division, Department of Medicine, Kyoto University School of Medicine, 54 Shogoin Kawahara-cho, Sakyo-ku, Kyoto 606, Japan. Received for publication 23 May 1991 and accepted in revised form 2 January 1992.
592
R
ecent studies revealed the existence of a family of glucose transporters with distinct functional properties but related primary sequences (1,2). Among the five facilitative glucose-transporter isoforms thus far identified, GLUT2 (liver-type glucose transporter) (3-6) is highly expressed in pancreatic p-cells and hepatocytes (3-6). GLUT2 is a unique transporter because it has a low affinity and a high Km for glucose (2,7). Glucose uptake into the p-cells is mediated by this high-K"m glucose-transporter isoform and is increased proportionately to an increase of extracellular glucose concentration. Thus, it has been suggested that GLUT2 plays an important role in glucose sensing by p-cells (8-12). Recently, several studies showed a strong correlation between the reduction of GLUT2 in diabetic islets and the level of hyperglycemia in rat models of diabetes (9-11). Consistent with this, clonal insulinoma cell lines such as RINm5F, which is derived from a rat insulinoma and lacks insulin secretory responsiveness to glucose, express large amounts of low-/11.1 mM (200 mg/dl) of medium glucose concentration was accompanied by a six- and fourfold reduction in 2-DG uptake, respectively (Fig. 4). However, in the other rat p-cell clone, RINr, which expresses GLUT1 but not GLUT2, maximum 2-DG uptake was observed after 24-h culture in a glucose-free medium (data not shown). Figure 5 shows the glucose responsiveness of insulin release of HIT cells after 24-h culture at 0, 5.5, or 22.2 mM (0, 99, or 400 mg/dl) glucose. The maximum glucose responsiveness was observed between 0 and 11.1 mM (0 and 200 mg/dl) glucose in HIT cells cultured at 5.5 mM (99 mg/dl) glucose for 24 h (Fig. 5). The insulin secretion of HIT cells cultured at 0 mM glucose for 24 h was significantly increased at 5.5 mM (99 mg/dl) glucose compared with that at 0 mM glucose, but no significant increase was observed above 5.5 mM (99 mg/dl) glucose (Fig. 5). On the other hand, the HIT cells cultured at 22.2 mM (400 mg/dl) glucose for 24 h were not respon-
DIABETES, VOL. 41, MAY 1992
N. INAGAKI AND ASSOCIATES
TABLE 1 Relative change of GLUT2 and GLUT1 mRNA levels in cultured HIT cells with various glucose concentrations Glucose concentration (mM)
GLUT2 mRNA GLUT1 mRNA
0
2.8
5.5
8.3
11.1
22 .2
6 ± 4* 107 ± 14
59 ± 16t 113±5
81 ± 10 103 ± 13
97 ± 8 105 ± 15
100 100
139 ±: 14t 117± : 16
Quantitative laser scanning densitometry of Northern blot of GLUT2 and GLUT1 mRNA in HIT cells after 24-h culture at 0-22.2 mM glucose. Amount of mRNA is percentage of value at 11.1 mM glucose. Values are means ± SE of 3 experiments. *P < 0.01, tP < 0.05 vs. control (11.1 mM glucose).
sive to glucose stimulation (Fig. 5). In addition, a glucose concentration >11.1 mM (200 mg/dl) depressed the maximum response of HIT cells cultured at 5.5 mM (99 mg/dl) glucose for 24 h, as reported previously (14,18; Fig. 5). DISCUSSION
Our results show that HIT cells that retain responsiveness to glucose stimulation express high levels of GLUT2 and GLUT1 mRNA and that glucose-transport activity in HIT cells is regulated by glucose concentration. However, we could detect no GLUT2 mRNA in RINr cells, which lack insulin secretory responsiveness to glucose. These results suggest that GLUT2 plays an important role in glucose sensing. We studied the role of glucose in the regulation of the amounts of GLUT2 and GLUT1 mRNAs, and glucose uptake activity in HIT cells. The GLUT1 mRNA level was not affected by differing concentrations of glucose. By contrast, glucose altered GLUT2 mRNA level and glucose uptake activity in a different way. The GLUT2 mRNA level in HIT cells incubated at 22.2 mM (400 mg/dl) glucose was higher than that in cells incubated at 11.1 mM (200 mg/dl) glucose, whereas GLUT2 mRNA stability determined with actinomycin D was similar at both glucose concentrations (t1/2 = 7-9 h). This result suggests that the increase of GLUT2 mRNA at 22.2 mM (400 mg/dl) glucose could be caused by increased transcription rates rather than stabilization of GLUT2 mRNA. On
the other hand, the GLUT2 mRNA level decreased proportionately to the reduction of the glucose concentration. The GLUT2 mRNA level in HIT cells was almost undetectable after 24-h glucose deprivation, which may be attributable to the marked increase in the degradation rate of GLUT2 mRNA by glucose deprivation (t1/2 = 2 h). However, whether the rates of GLUT2 mRNA transcription have a role in altering GLUT2 mRNA levels in HIT cells cultured in the medium containing various concentrations of glucose is unclear because direct measurements of GLUT2 mRNA transcription have yet to be performed. Chen et al. (8) showed that the p-cell GLUT2 mRNA level is increased 46% by prolonged hyperglycemic clamping of rats at 11.2 mM (202 mg/dl) for 5 days, whereas p-cells from rats kept hypoglycemic by continuous infusion of insulin lose expression of GLUT2 mRNA and high-Km glucose transport and display a strong reduction in glucose-induced insulin secretion. The similar effect of glucose on the GLUT2 mRNA level in vivo in their experiments and in vitro in ours suggests that glucose regulates GLUT2 mRNA level. c 30
o a E C
TABLE 2 Effects of 0, 11.1, and 22.2 mM glucose on degradation rate of GLUT2 mRNA Time (h)
Glucose concentration (mM) 0 11.1 22.2
0
4
100 100 100
23 ± 1* 66 ± 1 70 ± 5
8 10 ± 1* 54 -t- 5 47 + 3
Degradation rates were determined with actinomycin D treatment. HIT cells cultured in RPM11640 at 11.1 mM glucose were washed and exposed to serum-free RPMI 1640 medium containing 0,11.1, or 22.2 mM glucose for 2 h. Then actinomycin D (5 M-g/ml, final concentration) was added and incubation continued for 8 h. Changes in GLUT2 mRNA were determined at 0, 4, and 8 h by dot-blot analysis and are expressed as percentage of value without actinomycin D. Values are means ± SE of 3 experiments. *P< 0.01 vs. control (11.1 mM).
DIABETES, VOL 41, MAY 1992
20
i "5 a 10CO
a O Q CNI
2.8
5.5
8.3
11.1
22.2
[Glucose], mM FIG. 4. Effects of different glucose concentrations on 2-deoxy [3H]glucose uptake. After maintenance of HIT cells at indicated glucose for 24 h, 2-deoxy[3H]glucose uptake activity was measured as described in METHODS. Values are means ± SE of 3 experiments. *P < 0.01 vs. value at 5.5 mM glucose.
595
GLUCOSE-TRANSPORTER EXPRESSION IN HIT CELLS
mM glucose
0
5.5
22.2
[Glucose], mM of 24-h culture medium FIG. 5. Glucose-stimulated insulin secretion of HIT cells cultured with different concentrations of glucose for 24 h. After 24-h culture at 0, 5.5, or 22.2 mM glucose, Insulin release from HIT cells in response to 0, 5.5,11.1, or 22.2 mM glucose was measured. Values are means ± SE of 6 experiments. *P < 0.02 vs. value stimulated at 0 mM glucose. f P < 0.01 vs. value stimulated at 5.5 mM glucose. i P < 0.01 vs. value stimulated at 11.1 mM glucose.
Because GLUT1 protein was not affected by glucose, and neither GLUT3 nor GLUT4 mRNA was detected in HIT cells (data not shown), it is likely that the alteration of glucose uptake activity in HIT cells reflects change in the GLUT2 protein level. The reason for the discrepancy between the GLUT2 mRNA and the glucose uptake activity at >5.5 mM (99 mg/dl) glucose concentration remains to be explained; it is not known whether the rate of GLUT2 mRNA translation decreases or the rate of GLUT2 protein degradation increases in HIT cells cultured at a high glucose concentration. However, the relationship between the protein level and the glucose uptake activity of GLUT1 and GLUT2 needs to be determined separately. The maximum glucose responsiveness of insulin secretion was observed in 24-h cultured HIT cells at 5.5 mM (99 mg/dl) glucose, in which glucose uptake activity is maximum (Fig. 5). The HIT cells cultured at 22.2 mM (400 mg/dl) glucose, in which glucose uptake activity is decreased, were not responsive to glucose (Fig. 5). Of course, although 2-DG uptake was linear for at least 20 min (data not shown) and glucose transport is rate limiting compared with glucose phosphorylation (13,18) in HIT cells, the situation in islets is quite different. Islets equilibrate glucose in
