Peptides,Vol. 13, pp. 1193-1199, 1992
0196-9781/92 $5.00 + .00 Copyright© 1992PergamonPressLtd.
Printedin the USA.
Characterization of Amylin Binding Sites in a Human Hepatoblastoma Cell Line S. SHERIFF, J. E. F I S C H E R A N D A. B A L A S U B R A M A N I A M l Division o f G. I. Hormones, Department of Surgery, University of Cincinnati Medical Center, Cincinnati, O H 4526 7-0558 Received 4 March 1992 SHERIFF, S., J. E. FISCHER AND A. BALASUBRAMANIAM.Characterization of amylin binding sites in a human hepatoblastoma cell line. PEPTIDES 13(6) 1193-1199, 1992.--Amylin binding sites in a human hepatoblastoma cell line (HepG2) have been characterized in detail. ~25I-Amylin(rat) bound to HepG2 cells with high affinity. Binding was reversibleand selective, and dependent on time and temperature. Scatchard analysis revealed the presence of high (Kd = 0.11 + 0.04 nM) and low (Kd = 1.3 _+0.4 #M) affinitybinding sites for t25I-amylinin HepG2 cells. The dissociationexperimentsalso showed that n~I-amylindissociated from high- and low-affinitysites. The association data, however, indicated the presence of only one binding site. Rat amylin was more potent than human amylin and rat calcitonin gene-related peptide (CGRP) in displacing ~251-amylinbound to HepG2 cells. Nonhomologous peptides did not displace J2~l-amylin.Rat amylin was, however, less potent than rat CGRP in displacing ~251[Tyr°]CGRPfrom HepG2 cells. Pretreatment of HepG2 cells with rat amylin (10 nM) reduced the specific binding of J25Iamylin by 75%, whereas rat CGRP (10 nM) pretreatment had no effect on amylin binding. Calcitonin gene-related peptide, as well as rat and human amylin, stimulated the adenylate cyclaseactivity of HepG2 cell membrane preparation in a dose-dependent manner, with an order of potency of CGRP > rat amylin > human amylin. A CGRP antagonist, CGRP(8-37), significantly attenuated the stimulatory effect of both amylin and CGRP on adenylate cyclase activity. These investigationsshow that distinct receptors of amylin and CGRP are present in HepG2 cells, and that amylin stimulates adenylate cyclase activity through CGRP receptors. This system could now be exploited for studying amylin receptors and amylin-mediated signal transduction. Amylin
Human hepatoblastoma cell line
Calcitonin gene-related peptide
AMYLIN is a 37-residue peptide originally isolated from the amyloid-rich pancreas of non-insulin-dependent diabetic (NIDD) (10) and insulinoma (29) patients. Subsequently, amylin has also been isolated from the normal pancreas of rats (1). Amylin is now regarded as a normal component of the pancreas, being costored and cosecreted with insulin from the islet secretory granules (16). Recent investigations suggest that amylin inhibits insulin-stimulatedglucose uptake and glycogen synthesis by soleus muscles (19). The peripheral effects of amylin have also been demonstrated in dogs (24) and rats (17,20) using euglycemic clamp technique. These in vivo investigations revealed that amylin also causes hepatic insulin resistance and that the liver is more sensitive to amylin than other peripheral tissues (17). Adipocyte glucose metabolism, however, was not altered by amylin (10). These observations have led to the speculation that amylin may play a role in non-insulin--dependentdiabetes mellitus (NIDDM). Calcitonin gene-related peptide (CGRP), which exhibits >40% homology with amylin, has also been shown to cause peripheral and hepatic insulin resistance (19,20). Furthermore, hCGRP(8-37), an antagonist of CGRP, inhibited the stimulatory adenylate cyclase activity of CGRP as well as that of amylin (21,30). These observations and the failure to demonstrate spe-
cific binding ofamylin on various mammalian membrane preparations (15,30) have led investigators to speculate that amylin actions on liver and muscles are mediated by CGRP receptors. However, recent findings that amylin causes peripheral insulin resistance through a cAMP-independent pathway ( 1 1) and that N-a-Ac-amylin(8-37) is a more selective antagonist to amylin than CGRP (12) suggest that specific receptors of amylin may mediate its actions. This investigation was therefore undertaken to develop a model system to study amylin receptors and amylinmediated signal transductions. The human liver carcinoma cell line (HepG2) was chosen for this study because amylin has previously been shown to increase gluconeogenesis and glycogenolysis in HepG2 (9). This investigation shows that distinct receptors of amylin and CGRP are present in HepG2. METHOD
Materials Human and rat amylin, rat amylin(8-37), rat CGRP I, and human CGRP(8-37) were synthesized in our laboratory (3). The synthetic peptides were characterized by sequence and mass spectral analyses, and were found to be greater than 97% pure by analytical reversed-phase chromatography. [Tyr°]Rat CGRP
Requests for reprints should be addressed to A. Balasubramaniam.
1193
1194 I was obtained from Peninsula Laboratories (Belmont, CA). Insulin and glucagon were gifts from Eli Lilly (Indianapolis, IN). Human hepatoblastoma cell line (HepG2) was obtained from the American Type Culture Collection. All the cell culture reagents (GIBCO, Grand Island, NY), protease inhibitors, and buffer reagents (Sigma, St. Louis, MO) and ~25I (Amersham, Arlington, IL) were obtained commercially.
Cell Culture The cells were cultured at 37°C in a humidified 5% CO2 atmosphere. Cells were plated in 75-cm z flasks at a density of 1 × 106 cells per flask and maintained in Minimum essential medium (MEM) supplemented with Earle's salts, nonessential amino acids, and sodium pyruvate, and containing 10% fetal bovine serum and 16 m M HEPES. When the cells became confluent (5-6 days), the growth medium was aspirated and the cells were rinsed with Dulbecco's phosphate-buffered saline (DPBS) with Ca 2÷ and Mg 2+. The cells were then dispersed by incubation with 10 ml of 0.25% trypsin in Hanks' balanced salt solution (HBSS), pH 7.4, without Ca 2+ or Mg 2÷ for 10 min at 37°C. The dispersed cell suspension was washed twice with 20 ml of growth medium by centrifugation at 1000 × g for 10 min. The final cell pellet thus obtained was suspended in the corresponding assay buffers and used in binding and adenylate cyclase experiments.
lodination of Amylin Synthetic rat amylin was iodinated by the chloramine-T method. Briefly, the following reagents were added sequentially into a glass tube (12 × 75 mm): 50 ~1 phosphate buffer, 0.4 M, pH 7.6, 5 ttg rat amylin in 10 tA distilled water, 0.25 nmol of Na~ZSI (5 tA), 30 #g chloramine-T in 10 ~tl of phosphate buffer, 0.04 M, pH 7.6. After incubating the mixture for 45 s at room temperature, the reaction was arrested by the addition of 80 ~g of sodium metabisulfite in 20/A of phosphate buffer, 0.04 M, pH 7.6. The reaction mixture was immediately mixed with 100 ~1 of phosphate buffer, 0.04 M, pH 7.6, loaded onto a C4 (Vydac) reversed-phase column, and purified as described in the legend to Fig. 1. The peak fractions were pooled and aliquoted in quantities sufficient for one assay; they were stored at - 2 0 ° C and were used within 3-4 weeks. Similar protocol was used to iodinate [Tyr°]rat CGRP. Human amylin could not be iodinated by this method, even by prolonging the incubation time with chloramine-T to 5 min.
SHERIFF, FISCHER AND BALASUBRAMANIAM To determine the effects of competing peptides on ~25I-amylin binding to HepG2 cells, experiments were carried out as above by incubating cells (0.8 × 106/tube) with ~25I-amylin (0.40 nM) for 45 min in the presence of increasing concentrations of competing compounds. Association kinetic experiments were performed by incubating cells (0.8 × t06/tube) with 0.4 nM ~2SI-amylin for various time periods. The association reaction was initiated at 24°C by the addition of 125I-amylin, and samples were removed at various times and the bound and free ~25I-amylin was separated as described previously. Nonspecific binding for association curves was determined in the presence of 10 #M amylin. Dissociation experiments were performed by preincubating the cells with ~25Iamylin (0.4 nM) for 45 min at 24°C. Dissociation rates were measured by the addition of 3 ~zM amylin (final concentration). The binding was terminated as described previously over a period of 90 min.
Adenylate CyclaseActivity The final cell pellet obtained and described previously was homogenized by hand in 2 m M Tris-HCl buffer (pH 7.4) containing 1 m M EDTA and 0.2 m M dithiothreitol in a glass homogenizer. The membrane pellet obtained by centrifugation at 40,000 × g for 30 min was used in adenylate cyclase experiments. Adenylate cyclase activity was determined in 100 t~g membrane protein (50 ul) as reported earlier (2).
Desensitization Experiments The HepG2 cells (1 × 106/ml) were incubated in binding buffer at 37°C for 10 rain in a shaking water bath with rat amylin (10 nM) or rat CGRP (10 riM). At the end of incubation, equal volumes of ice-cold assay buffer were added and centrifuged at 1000 × g for 10 rain. This procedure was repeated once and the final cell pellet was reconstituted in assay buffer. The cell density was adjusted to 8 × 106 cells/ml in the assay buffer, and the binding study was carried out as described.
Data Analysis All points in the binding and adenylate cyclase experiments are the mean + SEM of three experiments performed in duplicate. The binding data were analyzed using the LIGAND program (23) obtained from Biosoft (Release 2) or NIH (Mac Ligand, Version 4.5 plus). RESULTS
Binding Studies t25I-Amylin binding to HepG2 cells was investigated in polypropylene tubes in a total volume of 0.25 ml HBSS buffer, pH 7.4, containing 20 m M HEPES, 1% bovine serum albumin, 0.01% bacitracin, and 300 a M phenylmethylsulfonyl fluoride. In a standard binding assay, 0.8 × 10~ cells/tube (12 × 75 mm) were incubated at 24°C for 45 rain in a shaking water bath with ~25I-amylin(0.40 nM) in the absence (total binding) and presence (nonspecific binding) of 10 a M unlabeled amylin to determine the specific binding. At the end of the incubation period, 1 ml of ice-cold assay buffer was added and the unbound ~25I-amylin was separated by centrifugation at 1000 × g for 10 rain, followed by aspiration of the supernatant. The tubes containing the cell pellet were counted for bound radioactivity in a micromedic gamma counter. Under these conditions, the nonspeeilic binding remained less than 35%. Viability of the cells was 85-90% as determined by trypan blue exclusion method before and after incubation.
The reversed-phase chromatography of amylin-iodination mixture resulted in a sharp single radioactivity peak (Fig. 1). ~25I-Amylin was stable up to 4 weeks when stored at - 2 0 ° C and had a specific activity in the range of 225-300 Ci/mmol. The ~25I-amylin thus obtained coeluted with that prepared by the nonreductive iodobead method, and exhibited good binding (3540%) to specific antibodies of amylin (Peninsula Laboratories). These observations, and the finding that dithiotheritol-treated ~25I-amylin eluted at an earlier time than the untreated tracer (not shown), suggest that amylin, especially the disulfide bond, is intact and has not undergone any modification during chloramine-T treatment. ~25I-[Tyr°]CGRP was p r ~ m ~ similarly and had a specific activity of 200-300 Ci/mmol. Investigations to identify the optimal conditions for binding studies revealed that ~25I-amylin binding is dependent on time and temperature. Specific binding reached a maximum (6.8% of the total radioactivity added) in 15 min at 30°C, but declined rapidly to 2.1% in the next 15 min (Fig. 2A). At 24°C, however,
AMYLIN BINDING SITES
1195
tn o )< In
50
a.
o >I--
4O
I--
o
p-
5
T
A O
24
E
19
~a t~
14
II
oo
I -8.0
'~
-10.0
PEPTIDES,
I -6.0
LOG[M]
B >. I,-
40
=
0.0006
p
Z w
=
0.0024
,
0196-9781/92 $5.00 + .00 Copyright© 1992PergamonPressLtd.
Printedin the USA.
Characterization of Amylin Binding Sites in a Human Hepatoblastoma Cell Line S. SHERIFF, J. E. F I S C H E R A N D A. B A L A S U B R A M A N I A M l Division o f G. I. Hormones, Department of Surgery, University of Cincinnati Medical Center, Cincinnati, O H 4526 7-0558 Received 4 March 1992 SHERIFF, S., J. E. FISCHER AND A. BALASUBRAMANIAM.Characterization of amylin binding sites in a human hepatoblastoma cell line. PEPTIDES 13(6) 1193-1199, 1992.--Amylin binding sites in a human hepatoblastoma cell line (HepG2) have been characterized in detail. ~25I-Amylin(rat) bound to HepG2 cells with high affinity. Binding was reversibleand selective, and dependent on time and temperature. Scatchard analysis revealed the presence of high (Kd = 0.11 + 0.04 nM) and low (Kd = 1.3 _+0.4 #M) affinitybinding sites for t25I-amylinin HepG2 cells. The dissociationexperimentsalso showed that n~I-amylindissociated from high- and low-affinitysites. The association data, however, indicated the presence of only one binding site. Rat amylin was more potent than human amylin and rat calcitonin gene-related peptide (CGRP) in displacing ~251-amylinbound to HepG2 cells. Nonhomologous peptides did not displace J2~l-amylin.Rat amylin was, however, less potent than rat CGRP in displacing ~251[Tyr°]CGRPfrom HepG2 cells. Pretreatment of HepG2 cells with rat amylin (10 nM) reduced the specific binding of J25Iamylin by 75%, whereas rat CGRP (10 nM) pretreatment had no effect on amylin binding. Calcitonin gene-related peptide, as well as rat and human amylin, stimulated the adenylate cyclaseactivity of HepG2 cell membrane preparation in a dose-dependent manner, with an order of potency of CGRP > rat amylin > human amylin. A CGRP antagonist, CGRP(8-37), significantly attenuated the stimulatory effect of both amylin and CGRP on adenylate cyclase activity. These investigationsshow that distinct receptors of amylin and CGRP are present in HepG2 cells, and that amylin stimulates adenylate cyclase activity through CGRP receptors. This system could now be exploited for studying amylin receptors and amylin-mediated signal transduction. Amylin
Human hepatoblastoma cell line
Calcitonin gene-related peptide
AMYLIN is a 37-residue peptide originally isolated from the amyloid-rich pancreas of non-insulin-dependent diabetic (NIDD) (10) and insulinoma (29) patients. Subsequently, amylin has also been isolated from the normal pancreas of rats (1). Amylin is now regarded as a normal component of the pancreas, being costored and cosecreted with insulin from the islet secretory granules (16). Recent investigations suggest that amylin inhibits insulin-stimulatedglucose uptake and glycogen synthesis by soleus muscles (19). The peripheral effects of amylin have also been demonstrated in dogs (24) and rats (17,20) using euglycemic clamp technique. These in vivo investigations revealed that amylin also causes hepatic insulin resistance and that the liver is more sensitive to amylin than other peripheral tissues (17). Adipocyte glucose metabolism, however, was not altered by amylin (10). These observations have led to the speculation that amylin may play a role in non-insulin--dependentdiabetes mellitus (NIDDM). Calcitonin gene-related peptide (CGRP), which exhibits >40% homology with amylin, has also been shown to cause peripheral and hepatic insulin resistance (19,20). Furthermore, hCGRP(8-37), an antagonist of CGRP, inhibited the stimulatory adenylate cyclase activity of CGRP as well as that of amylin (21,30). These observations and the failure to demonstrate spe-
cific binding ofamylin on various mammalian membrane preparations (15,30) have led investigators to speculate that amylin actions on liver and muscles are mediated by CGRP receptors. However, recent findings that amylin causes peripheral insulin resistance through a cAMP-independent pathway ( 1 1) and that N-a-Ac-amylin(8-37) is a more selective antagonist to amylin than CGRP (12) suggest that specific receptors of amylin may mediate its actions. This investigation was therefore undertaken to develop a model system to study amylin receptors and amylinmediated signal transductions. The human liver carcinoma cell line (HepG2) was chosen for this study because amylin has previously been shown to increase gluconeogenesis and glycogenolysis in HepG2 (9). This investigation shows that distinct receptors of amylin and CGRP are present in HepG2. METHOD
Materials Human and rat amylin, rat amylin(8-37), rat CGRP I, and human CGRP(8-37) were synthesized in our laboratory (3). The synthetic peptides were characterized by sequence and mass spectral analyses, and were found to be greater than 97% pure by analytical reversed-phase chromatography. [Tyr°]Rat CGRP
Requests for reprints should be addressed to A. Balasubramaniam.
1193
1194 I was obtained from Peninsula Laboratories (Belmont, CA). Insulin and glucagon were gifts from Eli Lilly (Indianapolis, IN). Human hepatoblastoma cell line (HepG2) was obtained from the American Type Culture Collection. All the cell culture reagents (GIBCO, Grand Island, NY), protease inhibitors, and buffer reagents (Sigma, St. Louis, MO) and ~25I (Amersham, Arlington, IL) were obtained commercially.
Cell Culture The cells were cultured at 37°C in a humidified 5% CO2 atmosphere. Cells were plated in 75-cm z flasks at a density of 1 × 106 cells per flask and maintained in Minimum essential medium (MEM) supplemented with Earle's salts, nonessential amino acids, and sodium pyruvate, and containing 10% fetal bovine serum and 16 m M HEPES. When the cells became confluent (5-6 days), the growth medium was aspirated and the cells were rinsed with Dulbecco's phosphate-buffered saline (DPBS) with Ca 2÷ and Mg 2+. The cells were then dispersed by incubation with 10 ml of 0.25% trypsin in Hanks' balanced salt solution (HBSS), pH 7.4, without Ca 2+ or Mg 2÷ for 10 min at 37°C. The dispersed cell suspension was washed twice with 20 ml of growth medium by centrifugation at 1000 × g for 10 min. The final cell pellet thus obtained was suspended in the corresponding assay buffers and used in binding and adenylate cyclase experiments.
lodination of Amylin Synthetic rat amylin was iodinated by the chloramine-T method. Briefly, the following reagents were added sequentially into a glass tube (12 × 75 mm): 50 ~1 phosphate buffer, 0.4 M, pH 7.6, 5 ttg rat amylin in 10 tA distilled water, 0.25 nmol of Na~ZSI (5 tA), 30 #g chloramine-T in 10 ~tl of phosphate buffer, 0.04 M, pH 7.6. After incubating the mixture for 45 s at room temperature, the reaction was arrested by the addition of 80 ~g of sodium metabisulfite in 20/A of phosphate buffer, 0.04 M, pH 7.6. The reaction mixture was immediately mixed with 100 ~1 of phosphate buffer, 0.04 M, pH 7.6, loaded onto a C4 (Vydac) reversed-phase column, and purified as described in the legend to Fig. 1. The peak fractions were pooled and aliquoted in quantities sufficient for one assay; they were stored at - 2 0 ° C and were used within 3-4 weeks. Similar protocol was used to iodinate [Tyr°]rat CGRP. Human amylin could not be iodinated by this method, even by prolonging the incubation time with chloramine-T to 5 min.
SHERIFF, FISCHER AND BALASUBRAMANIAM To determine the effects of competing peptides on ~25I-amylin binding to HepG2 cells, experiments were carried out as above by incubating cells (0.8 × 106/tube) with ~25I-amylin (0.40 nM) for 45 min in the presence of increasing concentrations of competing compounds. Association kinetic experiments were performed by incubating cells (0.8 × t06/tube) with 0.4 nM ~2SI-amylin for various time periods. The association reaction was initiated at 24°C by the addition of 125I-amylin, and samples were removed at various times and the bound and free ~25I-amylin was separated as described previously. Nonspecific binding for association curves was determined in the presence of 10 #M amylin. Dissociation experiments were performed by preincubating the cells with ~25Iamylin (0.4 nM) for 45 min at 24°C. Dissociation rates were measured by the addition of 3 ~zM amylin (final concentration). The binding was terminated as described previously over a period of 90 min.
Adenylate CyclaseActivity The final cell pellet obtained and described previously was homogenized by hand in 2 m M Tris-HCl buffer (pH 7.4) containing 1 m M EDTA and 0.2 m M dithiothreitol in a glass homogenizer. The membrane pellet obtained by centrifugation at 40,000 × g for 30 min was used in adenylate cyclase experiments. Adenylate cyclase activity was determined in 100 t~g membrane protein (50 ul) as reported earlier (2).
Desensitization Experiments The HepG2 cells (1 × 106/ml) were incubated in binding buffer at 37°C for 10 rain in a shaking water bath with rat amylin (10 nM) or rat CGRP (10 riM). At the end of incubation, equal volumes of ice-cold assay buffer were added and centrifuged at 1000 × g for 10 rain. This procedure was repeated once and the final cell pellet was reconstituted in assay buffer. The cell density was adjusted to 8 × 106 cells/ml in the assay buffer, and the binding study was carried out as described.
Data Analysis All points in the binding and adenylate cyclase experiments are the mean + SEM of three experiments performed in duplicate. The binding data were analyzed using the LIGAND program (23) obtained from Biosoft (Release 2) or NIH (Mac Ligand, Version 4.5 plus). RESULTS
Binding Studies t25I-Amylin binding to HepG2 cells was investigated in polypropylene tubes in a total volume of 0.25 ml HBSS buffer, pH 7.4, containing 20 m M HEPES, 1% bovine serum albumin, 0.01% bacitracin, and 300 a M phenylmethylsulfonyl fluoride. In a standard binding assay, 0.8 × 10~ cells/tube (12 × 75 mm) were incubated at 24°C for 45 rain in a shaking water bath with ~25I-amylin(0.40 nM) in the absence (total binding) and presence (nonspecific binding) of 10 a M unlabeled amylin to determine the specific binding. At the end of the incubation period, 1 ml of ice-cold assay buffer was added and the unbound ~25I-amylin was separated by centrifugation at 1000 × g for 10 rain, followed by aspiration of the supernatant. The tubes containing the cell pellet were counted for bound radioactivity in a micromedic gamma counter. Under these conditions, the nonspeeilic binding remained less than 35%. Viability of the cells was 85-90% as determined by trypan blue exclusion method before and after incubation.
The reversed-phase chromatography of amylin-iodination mixture resulted in a sharp single radioactivity peak (Fig. 1). ~25I-Amylin was stable up to 4 weeks when stored at - 2 0 ° C and had a specific activity in the range of 225-300 Ci/mmol. The ~25I-amylin thus obtained coeluted with that prepared by the nonreductive iodobead method, and exhibited good binding (3540%) to specific antibodies of amylin (Peninsula Laboratories). These observations, and the finding that dithiotheritol-treated ~25I-amylin eluted at an earlier time than the untreated tracer (not shown), suggest that amylin, especially the disulfide bond, is intact and has not undergone any modification during chloramine-T treatment. ~25I-[Tyr°]CGRP was p r ~ m ~ similarly and had a specific activity of 200-300 Ci/mmol. Investigations to identify the optimal conditions for binding studies revealed that ~25I-amylin binding is dependent on time and temperature. Specific binding reached a maximum (6.8% of the total radioactivity added) in 15 min at 30°C, but declined rapidly to 2.1% in the next 15 min (Fig. 2A). At 24°C, however,
AMYLIN BINDING SITES
1195
tn o )< In
50
a.
o >I--
4O
I--
o
p-
5
T
A O
24
E
19
~a t~
14
II
oo
I -8.0
'~
-10.0
PEPTIDES,
I -6.0
LOG[M]
B >. I,-
40
=
0.0006
p
Z w
=
0.0024
,
