Regulatory Peptides 111 (2003) 137 – 144 www.elsevier.com/locate/regpep
Different role of intracellular loops of glucagon-like peptide-1 receptor in G-protein coupling ¨ lo Langel b, Matjazˇ Zorko a,* Aljosˇa Bavec a, Mattias Ha¨llbrink b, U a
b
Medical Faculty, Institute of Biochemistry, University of Ljubljana, Vrazov trg 2, 1000 Ljubljana, Slovenia Department of Neurochemistry and Neurotoxicology, Arrhenius Laboratories, Stockholm University, Svante Arrhenius va¨g 21A, S-10691 Stockholm, Sweden Received 26 August 2002; received in revised form 6 November 2002; accepted 14 November 2002
Abstract Previous studies revealed the importance of the third intracellular loop of glucagon-like peptide-1 receptor (GLP-1R) in coupling to Gs and Gi1 proteins. In order to further study the signaling mechanisms of GLP-1R, we tested three peptides, corresponding to the sequences of the first (IC1), the second (IC2), and the third (IC3) intracellular loop of GLP-1R, for their interactions with heterotrimeric G-proteins of different types (Gas, Gao, Gai1, and Ga11 plus Gh1g2) overexpressed in sf9 cells. IC3 peptide powerfully stimulates all types of tested Gproteins, whereas IC1 and IC2 peptides show differential effects on G-proteins. Both IC1 and IC2 peptides activate Gs and cooperate with IC3 peptide in its stimulation. Go is not affected by IC1 and IC2. Gi1 and G11 are not affected by IC1, but are activated by IC2, which in activation cooperates with IC3. We suggest that GLP-1R is not coupled only to Gs and Gi1, as shown previously, but also to Go and G11. IC3 loop is the main switch that mediates signaling via GLP-1R to G-proteins, while IC1 and IC2 loops are important in discrimination between different types of G-proteins. D 2002 Elsevier Science B.V. All rights reserved. Keywords: Heterotrimeric G-proteins; Synthetic receptor-loop peptides; GTPgS binding; ADP-ribosylation; sf9 cells
1. Introduction The glucagon-like peptide-1 receptor (GLP-1R), a 64 kDa protein, belongs to the G-protein-coupled receptors (GPCRs), and is classified within secretin/vasointestinal peptide (VIP) receptor family [1]. The physiological ligand of GLP-1R is a 30 – 31-amino-acid-long peptide, glucagonlike peptide-1 (GLP-1), hormone that exerts strong insulinotropic effects in vitro [2] and in vivo [3]. Clinical studies indicate that GLP-1 not only stimulates insulin secretion in normal subjects, but also in those with non-insulin-dependent diabetes mellitus type 2 [4], supporting its therapeutic potential. The active GLP-1R increases cAMP production in the pancreatic h-cells and in the insulinomic cell lines, e.g., Rin m5F [5,6]. In monkey kidney COS 7 cells and rat adipocytes, GLP-1R increases the level of free cytosolic calcium
* Corresponding author. Tel.: +386-1-543-76-43; fax: +386-1-54376-41. E-mail address:
[email protected] (M. Zorko).
[7] and inositol phosphates [8], respectively, implicating activation of phosholiphase C. Both adenylyl cyclase and phospholipase C signaling pathways mediate insulin-like effects of GLP-1 via heterotrimeric G-proteins Gs, Gi1,2, and Gq, but not Gi3 in Chinese hamster ovary (CHO) cells overexpressing GLP-1R [9]. Block deletion mutation studies revealed the importance of IC3 of GLP-1R in activation of adenylyl cyclase [10]. Point mutations in IC1 and IC3 also altered intracellular signaling of adenylyl cyclase system [11]. Mutational scanning study of GLP-1R showed that specific determinants of coupling with Gas are primarily localized at the N-terminus of IC3 intracellular loop [12]. Our recent study on transfected cell line sf9 with baculovirus vectors carrying alpha subunit of heterotrimeric G-proteins Gi1 and Gs together with h1g2 subunit revealed that N-terminal part of IC3 is important in stimulation of cholera toxin-sensitive G-proteins, while C-terminal part of this loop stimulated pertussis toxin-sensitive G-proteins [13]. While the importance of IC3 of GLP-1R in coupling to G-proteins is well documented in the literature [10 –13], the role of IC1 and IC2 in this process is not clear yet. The substitution H180R
0167-0115/02/$ - see front matter D 2002 Elsevier Science B.V. All rights reserved. PII: S 0 1 6 7 - 0 11 5 ( 0 2 ) 0 0 2 8 2 - 3
138
A. Bavec et al. / Regulatory Peptides 111 (2003) 137–144
in IC1 causes a decrease in GLP-1 affinity for GLP-1R and decreased cAMP production [11], indicating that this loop might also be involved in signaling via G-proteins. According to our knowledge, the involvement of the IC2 of GLP-1R in signal transduction has not been reported yet. Several studies have shown that synthetic peptides with the sequence derived from the predicted intracellular loops of G-protein-coupled receptors, e.g., a2-adrenergic [14], D2 [15], follicle-stimulating hormone [16], and GLP-1 receptors [13], can mimic the effect of activation of these receptors acting at the specific receptor – Gprotein interface and not at the receptor binding site. Such synthetic peptides are therefore useful tools for studying receptor – G-protein interaction. We have synthesized three peptides corresponding to the sequences of three intracellular loops of GLP-1 receptor: the first GLP1R(168 – 176), SFRHLHCTR, (IC1); the second GLP1R(252– 264), YTLLAFSVFSEQR, (IC2); and the third GLP-1R(329 – 351), CIVIAKLKANLMCKTDIKCRLAK, (IC3). According to literature data (cf. Refs. [17,18]), the shortest receptor-mimicking peptides derived from different types of GPCRs usually consist of 9– 15 amino acids; peptides used in our work fulfil this requirement. We tested the interactions of these peptides with heterotrimeric G-proteins of different types – Gas, Gao, Gai1, and Ga11 –overexpressed together with h1g2 subunits in sf9 cells.
2. Materials and methods 2.1. Cell cultures Rin m5F cells were grown as monolayer culture in RPMI-1640 medium supplemented with 10% fetal bovine serum, 2 mM L-glutamine, 100 units/ml penicillin, 100 Ag/ ml streptomycin at 37 jC in a 5% CO2 atmosphere. sf9 cells were maintained as monolayer culture in Grace’s insect medium supplemented with 10% heatinactivated fetal bovine serum, 100 units/ml penicillin, 100 Ag/ml streptomycin at 28 jC in a 5% CO2 atmosphere. CHO cells were grown as monolayer culture in DMEM medium supplemented with 10% fetal calf serum, 0.39% NaHCO3, 58 mg/l proline, 53 mg/l aspartic acid, 59 mg/l glutamic acid, 60 mg/l asparagine, 100 units/ml penicillin, 100 Ag/ml streptomycin at 37 jC, 2 mM glutamine in a 5% CO2 atmosphere. 2.2. Overexpression of G-proteins in sf9 cells sf9 cells were cotransfected with recombinant baculoviruses vectors carrying different alpha subunits of heterotrimeric G-proteins (Gas, Gai1, Gao, Ga11) together with h1g2 subunits as described by Na¨ssman et al. [19]
with minor modifications. Briefly, approximately 6 106 cells/75 cm2 flask were infected with high-titer recombinant baculovirus stock solution. After 60 min of incubation at 28 jC, the virus stock was removed, the fresh medium was added, and cells were grown for 3 days at 28 jC. The expression of G-proteins was analyzed by 11% SDS-PAGE which, in contrast to control nontransfected cells, showed strong protein bands with molecular mass of 36 and 40– 45 kDa which corresponded to hsubunits and a subunits of G-proteins, respectively. Each type of the overexpressed G-protein (Gs, Gi1, Go, G11) was further checked by using the corresponding monoclonal antibodies. The yield of the overexpression was assessed by comparison of the rate of [35S]GTPgS binding to the membranes obtained from transfected and nontransfected cells. 2.3. Peptide synthesis Peptides were synthesized in a stepwise manner on 0.1 mmol scale using Applied Biosystem Model 431A peptide synthesizer. Solid synthesis used dicyclohexyl-carbodiimide/hydroxybenzotriazole activation strategy. tertButyloxycarbonyl amino acids were coupled as hydroxybenzotriazole esters to a p-methylbenzylhydrylamine (MBHA) resin (1.1 mmol of amino groups/g, Bachem, Switzerland) to obtain C-terminally amidated peptides. Deprotection of the side chains from formyl and benzyl groups was carried out using the ‘‘low TFMSA’’ method. The protecting groups on histidine (DNP) were removed by treatment for 1 h at room temperature with 20% (v/v) thiophenol/DMF. The peptides were finally cleaved from the resin with liquid HF at 0 jC for 30 min. Deprotection of the side chains, cleavage of the peptides, and purification on HPLC have been described earlier in detail [20]. Peptide purity was >99% as demonstrated with HPLC on an analytical Nucleosil 120-3 C18 0.4 10 cm reverse-phase HPLC column. The molecular mass of each synthetic peptide was determined with a plasma desorption mass spectrometer (Bioion 20, Applied Biosystems), and the calculated values were obtained in each case. 2.4. Plasma membrane preparation Rin m5F cells plasma membranes were obtained according to the protocol of McKenzie [21] with minor modifications described previously [22]. Briefly, monolayer cell cultures were washed and then resuspended in TE buffer (10 mM Tris – HCl, 0.1 mM EDTA, pH 7.5). The detached cells were homogenized in Polytron-type homogenizer (Braun, Germany). Homogenate was centrifuged at 500 g for 15 min, and membranes were collected by centrifugation of the supernatant at 40 000 g for 30 min in Beckman-L8 70 M ultracentrifuge (Beckman Instruments, USA). The whole procedure was undertaken
A. Bavec et al. / Regulatory Peptides 111 (2003) 137–144
at the temperature below 4 jC. The protein concentration in membrane preparations was from 1 to 2.5 mg/ml as determined by the method of Lowry et al. [23]. After expressing Gas, Gai1, Gao, and Ga11 together with h1g2 subunits in sf9 cells, plasma membranes were prepared according to the protocol for the preparation of Rin m5F cells plasma membranes described above. CHO plasma membranes were prepared as described previously [24] with minor modification [25]. Briefly, cells (approximately 2 108) were washed in Hanks’ Balanced Salt Solution (HBSS) without Ca 2 + and Mg2 +, detached with hypotonic buffer, and homogenized in Teflon/glass Potter homogenizer. After centrifugation of homogenate at 300 g, crude membranes were collected from the supernatant by a 20-min centrifugation at 42 000 g. The membranes were further purified by aqueous two-phase system with dextran and polyethylene glycol. All procedures were carried out at 4 jC. The concentration of proteins in plasma membranes was determined with Biorad Protein Assay at 595 nm. 2.5. GTPcS-binding studies The rate of [35S]GTPgS binding to G-proteins from plasma membranes was followed as previously described [21,22] with minor modification. Briefly, the membranes (final protein concentration in the assay mixture in Gi1-, Go-, G11-, Gs-enriched sf9 plasma membranes and in Rin m5F plasma membrane preparation was 390 and 500 Ag/ ml, respectively) were incubated with 5 mM MgCl2, 1 mM dithiothreitol (DTT), 150 mM NaCl, 1 AM GDP, and 0.5 – 1 nM [35S]GTPgS (approximately 140 000 cpm/assay) at 25 jC in TE buffer (pH 7.5), for 2 min with or without peptides. The unbound [35S]GTPgS was washed out by rapid filtration of the reaction mixture through Millipore GF/C glass-fiber filters under vacuum three times with 5 ml of TE buffer. After extraction of the radioactive material overnight in 20 ml of Emulsifier-Safe (Packard, USA) scintillation liquid, radioactivity was determined with LKB 1214 Rackbeta liquid scintillation counter. Blank values were determined by the same procedure in samples in which the membranes were replaced with buffer. 2.6. ADP-ribosylation assay Pertussis and cholera toxin ADP-ribosylations followed the incorporation of [32P]ADP-ribose into the G-protein a subunit and were performed with 10 Ag of CHO plasma membrane preparation. Pertussis toxin was preactivated in 62.5 mM dithiothreitol (DTT) at room temperature for 1 h. The 50 Al of plasma membranes, 13 Ag/ml toxin, 50 mM Tris buffer (pH = 8), 25 mM DTT, 0.5 mM EDTA, 1 mM ATP, 5 106 cpm [32P]NAD, and 2 AM h-NAD were incubated with or without ligand at 37 jC for 1 h. Cholera toxin was preactivated in 50 mM DTT at room
139
temperature for 1 h. The 50 Al of plasma membranes, 10 Ag/ml toxin, 12 mM DTT, 100 mM potassium phosphate buffer (pH = 7.5), 1 mM ATP, 2.5 mM MgCl2, 10 mM thymidine, 5 106 cpm [32P]NAD, and 2 AM h-NAD were incubated with or without ligand at 37 jC for 1 h. The ADP-ribosylation reactions were stopped by diluting the samples in 50 Al Laemmli buffer, boiled, and analyzed by 11% SDS-PAGE followed by electroblot onto nitrocellulose membranes. The radioactivity incorporated into the membrane proteins were detected by Instant Imager Packard [26]. 2.7. CD spectroscopy Circular dichroism (CD) measurements were performed on the automated Aviv CD spectrometer model 62A DS at 25 jC. The parameters used were a bandwidth of 1 nm, 1 s averaging time, and step size of a 0.5 nm. IC3 peptide was dissolved in 5 mM potassium phosphate (pH = 7.4), and measurements at different peptide concentrations were performed in 10 mm cell (1 and 5 AM), 1 mm cell (50 AM), and in 0.1 mm demutable cell (500 AM). Spectra at the 50 and 500 AM concentration were scanned between 185 and 250 nm, whereas spectra at the 5 AM concentration were scanned only between 200 and 250 nm due to strong signal of the 10 mm cell below 200 nm. Single-wavelength measurement was performed at the 1 AM peptide concentration in the 10 mm cell, at local maximum of 218 nm with averaging time of 60 s. The average spectra of four measurements were corrected for the baseline and presented in units of mean residue molar ellipticity. Secondary structure calculations were provided by PROSEC computer program supplied with the instrument [27]. 2.8. Other methods The nonlinear least squares fitting of the curves and statistical analysis (one-way ANOVA) were carried out by PRISM3 computer program (GraphPad Software, USA), which was used also for graphical presentation of the results. In ANOVA analysis, Tukey – Kramer multiple comparison test was used to reveal significant differences between groups of interest, P>0.05 was not considered significant. 2.9. Materials MgCl2, NaCl, and EDTA were from Merck (Germany); [35S]GTPgS and [32P]NAD were from Amersham Pharmacia Biotech (UK); tert-butyloxycarbonyl amino acids were from Bachem and Chemimpex (USA); cell culture sf9 was from ECACC (UK); HBSS, cell culture media, and reagents were from Life Technologies and Sigma (USA); and elecrophoresis reagents were from Eurobio. Nitrocel-
140
A. Bavec et al. / Regulatory Peptides 111 (2003) 137–144
lulose transfer membranes were from Protran. EmulsifierSafe scintillation liquid was from Packard. Baculovirus vectors with the Gai1, Gao, Ga11, Gas, and Gh1g2 subunits of heterotrimeric G-proteins were generous gift of Prof. Tatsuya Haga (University of Tokyo, Tokyo, Japan). All other chemicals were from Sigma and were of analytical grade.
3. Results 3.1. The effect of loop peptides on the rate of [35S]GTPcS binding in plasma membranes from Rin m5F cells The binding of [35S]GTPgS to G-proteins in plasma membranes from Rin m5F cells is time-dependent and onephase exponential association curve was fitted to the experimental data giving maximal [35S]GTPgS binding of 368 fmol/mg and first-order rate constant of 0.12 min 1. The initial rate of [35S]GTPgS binding of 40 fmol/min/mg (basal activity of membranes) was assessed from the apparently liner part (0 – 5 min) of the curve (data not shown). IC3 peptide increased the initial rate of [35S]GTPgS binding as compared to the basal activity. Fitting a sigmoidal dose –response curve to the data obtained for IC3 gave the following parameters: maximal increase of the initial rate of [35S]GTPgS binding = 290.2 F 12.3% over basal, EC50 = 20.6 F 0.1 AM and Hill coefficient = 2.84 F 0.5 (Fig. 1A). IC1 and IC2 peptides (at 25 AM) were not able to influence [35S]GTPgS binding to the plasma membranes, acting separately (Fig. 1A and B, bars B and C) or together (Fig. 1B, bar E). According to one-way ANOVA statistical analysis, the simultaneous presence of all three peptides, IC1, IC2, and IC3 (at 25 AM each), significantly increased the rate of [35S]GTPgS binding to 159% ( P < 0.01) (Fig. 1B, bar H) as compared to the basal level, but decreased the rate of [35S]GTPgS binding for 26.8% ( P < 0.05) in comparison to the action of the IC3 (at 25 AM) alone (Fig. 1B, bar D). Each of peptides IC1 and IC2, when applied together with IC3 peptide, did not decrease the rate of [35S]GTPgS binding relative to IC3 peptide alone (compare bars F and G with bar D in Fig. 1B). 3.2. The effect of loop peptides on the rate of [35S]GTPcS binding in enriched plasma membranes from sf9 cells The binding of [35S]GTPgS to G-proteins in plasma membranes from sf9 cells transiently expressing either Gas, Gai1, Gao, or Ga11 together with Gh1g2 subunit is time-dependent and show classical one-phase exponential association curve with maximal [35S]GTPgS binding of 566 fmol/mg for Gs, 177 fmol/mg for Gi1, 322 fmol/mg for Go, 104 fmol/mg for G11-enriched plasma membranes, and 80 fmol/mg in the control nontransfected sf9 cells
Fig. 1. The effect of peptides derived from the first (IC1), the second (IC2), and the third (IC3) intracellular loop of the glucagon-like peptide-1 receptor on the initial rate of [35S]GTPgS binding to the plasma membranes from Rin m5F cells. (A) Single effect of IC1, IC2, and IC3. Each point with the indicated standard deviation represents the mean value of three independent experiments. The curve for IC3 was obtained by fitting single-phase dose – response equation to the corresponding experimental data. 100% = 40 F 4 fmol/min/mg protein (basal value, no peptides added). (B) Simultaneous effect of IC1, IC2, and IC3 at 25 AM concentration. Each bar with the indicated standard deviation represents the mean value of three independent experiments. 100% = 40 F 4 fmol /min/mg protein (basal value, no peptides added; bar A). ANOVA: F(7,16) = 110.74; P < 0.001. Tukey – Kramer multiple comparison test was used to reveal significant differences between groups of interest. See Results for details.
(data not shown). The apparently linear part of the curve (0 – 5 min) was used to assess the initial rate of [35S]GTPgS binding and the following values were obtained: 36 fmol/mg/min in Gs, 24 fmol/mg/min in Go, 16 fmol/mg/min in Gi1, 11 fmol/mg/min in G11 overexpressed plasma membranes (see Table 1), and 4.5 fmol/mg/min in the membranes from nontransfected sf9 cells. The effect of IC3 peptide on the rate of [35S]GTPgS binding was dose-dependent (Fig. 2). Because of low solubility, the maximal effect of IC3 could not be obtained. The maximal observed increase of the initial
A. Bavec et al. / Regulatory Peptides 111 (2003) 137–144
141
Table 1 Effects of 50 AM concentration of peptides derived from the first (IC1), the second (IC2), and the third (IC3) intracellular loop of the glucagon-like peptide-1 receptor on the initial rate of [35S]GTPgS binding to the plasma membranes from sf9 cells enriched with Gs, Go, Gi1, and G11 type of G-proteins Peptide added (50 AM)
Initial rate of [35S]GTPgS binding to plasma membranes from sf9 cells with the overexpressed heterotrimeric G-proteins of the following types (% of basal) Gs
Go
Gi1
G11
Basal (no peptide added)
100 F 3% (36 F 1 fmol/min/mg) 148 F 9 204 F 3 363 F 19 241 F 6 539 F 5 479 F 4 689 F 25
100 F 9% (24 F 2 fmol/min/mg) 95 F 17 117 F 11 222 F 8 131 F 1 213 F 1 247 F 17 260 F 6
100 F 19% (16 F 3 fmol/min/mg) 98 F 9 116 F 7 224 F 37 176 F 7 211 F 8 304 F 4 362 F 10
100 F 3% (11 F 0.3 fmol/min/mg) 108 F 19 164 F 13 449 F 23 155 F 11 479 F 18 604 F 3 544 F 9
IC1 IC2 IC3 IC1 + IC2 IC1 + IC3 IC2 + IC3 IC1 + IC2 + IC3
rate of [35S]GTPgS binding at IC3 concentration of 100 AM was 730% for Gs, 394% for Gi1, 446% for Go, and 1010% for G11-enriched plasma membranes as compared to the basal activity. Individual and simultaneous effects of IC1, IC2, and IC3 peptides, in the concentration of 50 AM each, on the initial rate of [35S]GTPgS binding to plasma membranes from sf9 cells with the overexpressed heterotrimeric G-proteins of different types are summarized in Table 1. As it can be seen from Table 1, in Gs-enriched plasma membranes IC1 and IC2 peptides in the concentration of 50 AM each increased the initial rate of [35S]GTPgS binding for 148% ( P < 0.01) and 204% ( P < 0.01), respectively, acting separately, and for 241% ( P < 0.01) acting together in comparison to the basal value. Note that the effect of IC1 and IC2 on Gs proteins seems to be additive. The combination of peptides IC1 and IC3 as well as IC2 and IC3 increased the rate of [35S]GTPgS binding for 148% ( P < 0.01) and 132% ( P < 0.05), respectively, in comparison to the action of IC3 peptide itself (363%; P < 0.01). The simultaneous presence of all three peptides IC1, IC2, and IC3 at concentration of 50 AM each increased the rate of [35S]GTPgS binding for 190% ( P < 0.01) as compared to the effect of IC3 alone (assessed from data presented in Table 1). In Go-enriched plasma membranes, IC1 and IC2 peptides did not affect the rate of [35S]GTPgS binding and they were not able to significantly alter the effect of IC3 (Table 1). In Gi1-enriched plasma membrane, IC1 showed no significant effect, while IC2 increased the initial rate of [35S]GTPgS binding in dose-dependent manner to 119% ( P < 0.05) at 50 AM (Table 1) and to 168.2% ( P < 0.01) at 100 AM (data not shown). As it can be calculated from data shown in Table 1, the combination of peptides IC2 and IC3 increased the rate of [35S]GTPgS binding for 135.8% ( P < 0.05) in comparison to the action of IC3 alone. Interestingly, in the presence all three peptides,
IC1, IC2, and IC3, the rate of [35S]GTPgS binding was further increased to 162.7% ( P < 0.01) as compared to the IC3, (calculated from data presented in Table 1) although IC1 alone did not show any effect. Data from Table 1 also revealed that in G11-enriched plasma membranes IC1 showed no effect. IC2 peptide increased the initial rate of [35S]GTPgS binding for 162.8% ( P < 0.01) as compared to basal value and increased the stimulatory effect of IC3 up to 135% ( P < 0.05) as compared to the effect of IC3 alone (assessed from data presented in Table 1). The simultaneous presence of all three peptides, IC1, IC2 and IC3, nonsignificantly affect the rate of [35S]GTPgS binding to 121% ( P < 0.1) as compared to the action of IC3 peptide alone (assessed from data in Table 1). This indicates that IC2
Fig. 2. The effect of peptide derived from the third intracellular loop of the glucagon-like peptide-1 receptor (IC3) on the initial rate of [35S]GTPgS binding to the plasma membranes from sf9 cells enriched with Gs, Gi1, Go, and G11 proteins, respectively. Each point with the indicated standard deviation represents the mean value of three independent experiments. Curves were obtained by fitting single-phase dose – response equation to the corresponding experimental data. Basal value (100%) is 36 fmol/min/mg protein for Gs, 16 fmol/min/mg protein for Gi1, 24 fmol/min/mg protein for Go and 11 fmol/min/mg protein for G11.
142
A. Bavec et al. / Regulatory Peptides 111 (2003) 137–144
activates G11 acting alone or in combination with IC3, but not with IC1. 3.3. The effect of peptide IC3 on pertussis and cholera toxinmediated ADP-ribosylation in plasma membranes from CHO cells Pertussis toxin ADP-ribosylation of plasma membranes from CHO cells showed a protein band with molecular mass of 40 –41 kDa, which corresponded to ai/o subunit of G-proteins. At 50 AM, IC3 decreased ADP-ribosyltransferase activity to 51% of the basal, but it showed no effect at 1 AM concentration (Fig. 3A). The same results Fig. 4. Circular dichroism spectra of the peptide derived from the third intracellular loop of the glucagon-like peptide-1 receptor (IC3) at 5 AM (A), 50 AM (B), and 500 AM (C) concentration.
of pertussis toxin-mediated ADP-ribosylation and effect of IC3 were obtained with rat brain plasma membranes (data not shown). Cholera toxin ADP-ribosylation of CHO plasma membranes showed a protein band with molecular mass of 42 and 45 kDa, which corresponded to two isoforms of as subunits of G-proteins. IC3 showed no significant effect on cholera toxin-mediated ADP-ribosylation at 1 and 50 AM concentrations (Fig. 3B). 3.4. Secondary structure determination of peptide IC3 using CD spectroscopy Fig. 4 shows the CD spectra of peptide IC3 at 5, 50, and 500 AM concentration. Relative amounts of different types of secondary structure of peptide IC3 calculated from these spectra are presented in Table 2. Results clearly show that IC3 adopts higher amount of h-structure with the increased peptide concentration. Due to strong background signal of the 10 mm cell below 200 nm, the complete spectra of IC3 at the lowest concentration (1 AM) could not be obtained. Therefore, only single-wavelength measurement with 3.05j cm2 dmol 1 of mean residue ellipticity (hmrv) was performed at local minimum at 218 nm at this concentration of the peptide. No difference in ellipticity between 1 and 5 AM peptide at wavelength of 218 nm was observed, indicating that the
Fig. 3. The effect of peptide derived from the third intracellular loop of the glucagon-like peptide-1 receptor (IC3) on the pertussis toxin (PTX) and cholera toxin (CTX) mediated ADP-ribosylation in plasma membranes from Chinese hamster ovary cells. (A) PTX mediated ADP-ribosylation. Each bar with the indicated standard deviation represents the mean values of two independent experiments: A = no IC3 (100%), B = 1 AM IC3, C = 50 AM IC3, D = control (no PTX, no IC3). (B) CTX mediated ADPribosylation. Each bar with the indicated standard deviation represents the mean values of two independent experiments: A = no IC3 (100%), B = 1 AM IC3, C = 50 AM IC3, D = control (no CTX, no IC3).
Table 2 Relative proportion of different types of secondary structure of peptide derived from the third intracellular loop of the glucagon-like peptide-1 receptor peptide (IC3) at 5, 50, and 500 AM concentration Secondary structure
a-helix h-structure Random coil h-turn
Relative proportion of different types of secondary structure (%) 5 AM IC3
50 AM IC3
500 AM IC3
8 28 31 33
5 66 29 0
7 63 30 0
A. Bavec et al. / Regulatory Peptides 111 (2003) 137–144
secondary structure of IC3 is not altered at concentration below 5 AM.
4. Discussion Although our results obtained with the membranes from Rin m5F cells point to IC3 as the only loop involved in transmission of signal from GLP-1R to G-proteins (Fig. 1A), apparent inhibitory effect of IC1 and/or IC2 on the activation of G-proteins by IC3 (Fig. 1B) suggests a modulatory role of IC1 and IC2 and more complex interaction between GLP-1R and G-proteins. The observed effect of loop peptides could reflect the action of IC1, IC2, and IC3 only on one type of G-proteins, but it could be also due to a cumulative action of loop peptides on different types of Gproteins present in Rin m5F cells. To clarify this question, we have prepared plasma membranes from sf9 cells cotransfected with different types of heterotrimeric G-proteins, consisting of Gas, G ao, Gai1, and G a11, respectively, together with Gh1g2. Results presented in Fig. 2 demonstrate that IC3 in micromolar concentrations stimulates all types of G-proteins used, but with higher preference for Gs and G11 then for Go and Gi1 proteins. Promiscuous binding of IC3 with apparently similar affinity as obtained in our case to different types of G-proteins was observed also with other loop peptides such as the third intracellular loop peptide of human a2-adrenergic receptor on Gs, Gi, and Go proteins [17]. The interaction of G-proteins with IC3 peptide was monitored also via cholera and pertussis toxin-mediated ADP-ribosylation of plasma membranes from CHO cells in the presence and the absence of IC3 (Fig. 3). Pertussis toxin-mediated ADP-ribosylation showed a protein band with molecular mass of 40 –41 kDa, which corresponded to ai/o subunit of G-proteins. In the presence of 50 AM IC3 peptide, the pertussis toxin ADP-ribosyltransferase activity was decreased to 51% of the basal; at 1 AM concentration, IC3 peptide showed no effect. Gi/o in the heterotrimeric ahg form is the best substrate for pertussis toxin-catalyzed ADPribosylation, whereas free Gai/o subunit is quite poor substrate for this modification. Our results could be interpreted as the consequence of the heterotrimeric complex dissociation into the a and the hg subunit in the presence of IC3. An alternative interpretation of the results would be in terms of sterical hindering of ADP-ribosylation site on Gai/o subunit by bound IC3 peptide. Cholera toxin ADP-ribosylation of plasma membranes from CHO cells showed a protein band with molecular mass of 42 and 45 kDa, which corresponded to two isoforms of as subunits of G-proteins. In contrast to the pertussis toxin, the cholera toxin prefers free Gas subunit, and the heterotrimeric ahg form is a poor substrate. Neither 1 nor 50 AM IC3 were able to affect the ADPribosyltransferase activity. These data could mean that IC3 does not affect Gs proteins. Previous reports [12,13] and
143
data obtained in this study (Fig. 2), however, clearly demonstrate that IC3 affects Gs. It seems possible that the basal concentration of free a subunit of Gs-proteins in CHO membranes is too high for the registration of the difference in ADP-ribosyltransferase activity of cholera toxin in IC3-treated and IC3-nontreated membranes. These results also indicate that the ADP-ribosylation site and IC3 binding site on Gas do not overlap. The results obtained from [35S]GTPgS binding to Gproteins (Fig. 2) together with the results of pertussis toxinmediated ADP-ribosylation (Fig. 3) correlate with the data obtained from CD spectroscopy measurements (Fig. 4, Table 2). It can be seen from Table 2 that below 5 AM concentration IC3 adopts much less h-structure than at 50 AM and above. It would be possible that at higher concentrations, the IC3 structure is changed from ‘‘switch off’’ to ‘‘switch on’’ conformation that stimulates Gs, Go, Gi1, and G11-proteins by increasing the rate of [35S]GTPgS binding and also affects the pertussis-sensitive G-proteins as it has been seen from toxin-mediated ADP-ribosylation experiments. We are well aware, however, that isolated loop peptides can behave much differently from the loops that are integrated into the receptor structure. Additionally, increase of the amount of h-structure might also reflect the aggregation of IC3 peptide at higher concentrations although solubility problems with this peptide occur only above 100 AM concentrations. Both other loops, IC1 and IC2, are also necessary in GLP1R-G-protein interaction. They can act as positive modulators or show no effect on the initial rate of [35S]GTPgS binding in plasma membranes from sf9 cells enriched with Gs, G i1, G0, and G11 heterotrimeric G-proteins, as presented in Table 1. In Gs-enriched plasma membranes, both loops are capable of stimulating Gs-proteins. According to previous reports, IC1 affects the Gs-protein activation, but IC2 does not [12]. The difference between these and our results could be because block deletion and substitution mutations of predicted IC2 region upstream from the A256 of rat GLP1R have not been examined yet. We have used the sequence of IC2 loop as proposed by Swiss-Prot database and this does not fully correspond to the sequence of IC2 used previously by other authors [12]. In Go-enriched plasma membranes, IC1 and IC2 peptides show no significant effect on G-proteins activity. In Gi1-enriched membrane preparation, IC1 peptide shows no effect, while IC2 peptide stimulates Gi1-proteins. In our previous report, we showed that parts of IC3 peptide, but not the complete peptide, could discriminate between cholera and pertussis toxin-sensitive Gs and Gi/o proteins [13]. Our present results suggest that IC2 peptide differentiates also between two closely related types of pertussis toxin-sensitive G-proteins, Gi1 and Go, which show virtually the same apparent affinity and potency for IC3 (Fig. 2). IC2 peptide also stimulates G11 proteins, indicating that the differentiation between Gi1 and G11 proteins is due to different apparent affinities and potencies for IC3 (Fig. 2).
144
A. Bavec et al. / Regulatory Peptides 111 (2003) 137–144
Differences in the apparent affinity of IC3 for G-proteins obtained from Rin m5F membranes and sf9 membranes (Figs. 1 and 2) as well as different effect of IC1 and IC2 on the activation of G-proteins by IC3 in these two membrane preparations could be interpreted as the consequence of the cumulative effect of different types of G-proteins, not only Gs, Go, Gi1, and G11 but also others, for instance, Gq or Gi2 [9], which might also be involved in GLP-1R G-protein coupling in Rin m5F cells. In conclusion, we have demonstrated the ability of peptides derived from the intracellular loops of GLP-1R to activate G-proteins of different types overexpressed in sf9 cells. Not only Gs and Gi1, as shown previously, but also Go and G11 were affected. In this respect, the IC3 peptide was the most effective activator that suggests the possibility that IC3 loop is the main switch that mediates signaling via GLP1R to G-proteins, while IC1 and IC2 loops function as modulators of signal and might be important in discrimination between different types of G-proteins. However, the physiological relevance of our results obtained in the overexpression system is questionable; therefore, further studies are needed to clarify putative promiscuous binding of GLP1R to G-proteins in vivo.
References [1] Thorens B. Expression cloning of the pancreatic beta cell receptor for the gluco-incretin hormone glucagon-like peptide-1. Proc Natl Acad Sci U S A 1992;89:8641 – 5. [2] Mojsov S, Weir GC, Habener JF. Insulinotropin: glucagon-like peptide-1 (7 – 37) co-encoded in the glucagon gene is a potent stimulator insulin release in the perfused rat pancreas. J Clin Invest 1987; 79:616 – 9. [3] Kreymann B, Williams G, Ghatei MA, Bloom SR. Glucagon-like peptide-1 7 – 36: a physiological incretin in man. Lancet 1987;2: 1300 – 4. [4] Gutniak M, Orskov C, Holst JJ, Ahre´n B, Efendic S. Antidiabetogenic effect of glucagon-like peptide-1(7 – 36)amide in normal subjects and patients with mellitus. N Engl J Med 1992;326:1316 – 22. [5] Go¨ke R, Conlon JM. Receptors for glucagon-like peptide-1 (7 – 36) amide on rat insulinoma-derived cells. J Endocrinol 1988;116: 357 – 62. [6] Go¨ke R, Fehmann HC, Richter G, Trautmann M, Go¨ke B. Interaction of glucagon-like peptide-1(7 – 36)amide and somatostatin-14 in RINm5F cells and in the perfused rat pancreas. Pancreas 1989;4: 668 – 73. [7] Wheeler MB, Lu M, Dillon JS, Leng XH, Chen C, Boyd III AE. Functional expression of the rat glucagon-like peptide-1 receptor, evidence for coupling to both adenylyl cyclase and phospholipaseC. Endocrinology 1993;133:57 – 62. [8] Marquez L, Trapote MA, Luque MA, Valverde I, Villanueva-Penacarrillo ML. Inositolphosphoglycans possibly mediate the effects of glucagon-like peptide-1(7 – 36)amide on rat liver and adipose tissue. Cell Biochem Funct 1998;16:51 – 6. [9] Montrose-Rafizadeh C, Avdonin P, Garant MJ, Rodgers BD, Kole S, Yang H, et al. Pancreatic glucagon-like peptide-1 receptor couples to multiple G proteins and activates mitogen-activated protein kinase pathways in Chinese hamster ovary cells. Endocrinology 1999; 140:1132 – 40.
[10] Takhar S, Gyomorey S, Su RC, Mathi SK, Li X, Wheeler MB. The third cytoplasmic domain of the GLP-1(7 – 36 amide) receptor is required for coupling to the adenylyl cyclase system. Endocrinology 1996;137:2175 – 8. [11] Heller RS, Kieffer TJ, Habener JF. Point mutations in the first and third intracellular loops of the glucagon-like peptide-1 receptor alter intracellular signaling. Biochem Biophys Res Commun 1996; 223: 624 – 32. [12] Mathi SK, Chan Y, Li X, Wheeler MB. Scanning of the glucagon-like peptide-1 receptor localizes G protein-activating determinants primarily to the N terminus of the third intracellular loop. Mol Endocrinol 1997;11:424 – 32. ¨ stenson CG, Efendic S, [13] Ha¨llbrink M, Holmqvist T, Olsson M, O ¨ . Different domains in the third intracellular loop of the Langel U GLP-1 receptor are responsible for Galpha(s) and Galpha(i)/Galpha(o) activation. Biochim Biophys Acta 2001;1546:79 – 86. [14] Daaka Y, Luttrell LM, Lefkowitz RJ. Switching of the coupling of the beta2-adrenergic receptor to different G proteins by protein kinase A. Nature 1997;390:88 – 91. [15] Malek D, Munch G, Palm D. Two sites in the third inner loop of the dopamine D2 receptor are involved in functional G protein-mediated coupling to adenylate cyclase. FEBS Lett 1993;325:215 – 9. [16] Leng N, Grasso P, Deziel MR, Reichert Jr LE. A synthetic peptide corresponding to glycoprotein hormone alpha subunit residues 32 – 46 inhibits binding to receptor. Pept Res 1995;8:272 – 7. [17] Okamoto T, Nishimoto I. Detection of G-protein-activator regions in M4 subtype muscarinic, cholinergic, and alpha 2-adrenergic receptors based upon characteristics in primary structure. J Biol Chem 1992; 267:8342 – 6. [18] Grasso P, Deziel MR, Reichert LE. Synthetic peptides corresponding to residues 551 to 555 and 650 to 653 of the rat testicular folliclestimulating hormone (FSH) receptor are sufficient for post-receptor modulation of Sortoli cell responsiveness to FSH stimulation. Regul Pept 1995;60:177 – 83. ˚ kerman KOE. The second intracellular loop [19] Na¨sman J, Jansson CC, A of the a2-adrenergic receptors determinates subtype-specific coupling to cAMP production. J Biol Chem 1997;272:9703 – 8. ¨ , Land T, Bartfai T. Design of chimeric peptide ligands to [20] Langel U galanin receptors and substance P receptors. Int J Pept Protein Res 1992;39:516 – 22. [21] McKenzie FR. Basic techniques to study G-protein function. In: Milligan E. editor. Signal trunsduction. London: Oxford Univ. Press; 1992. p. 33-8 and 51-3. ¨ , Zorko M. Peptitergent PD1 [22] Bavec A, Jure´us A, Cigic B, Langel U affects the GTPase activity of rat brain cortical membranes. Peptides 1999;20:177 – 84. [23] Lowry OH, Rosenbrough NY, Farr AL, Randall RJ. Protein measurement with the folin phenol reagent. J Biol Chem 1951;193:265 – 75. [24] Gettys TW, Sheriff-Carter K, Moomaw J, Taylor IL, Raymond JR. Characterization and use of crude alpha-subunit preparations for quantitative immunoblotting of G proteins. Anal Biochem 1994; 220:82 – 91. [25] Lupi R, Corda D, Di Girolamo M. Endogenous ADP-ribosylation of the G protein beta subunit prevents the inhibition of type 1 adenylyl cyclase. J Biol Chem 2000;275:9418 – 24. [26] Di Girolamo M, D’Arcangelo D, Cacciamani T, Gierschik P, Corda D. K-ras transformation greatly increases the toxin-dependent ADP-ribosylation of GTP binding proteins in thyroid cells. Involvement of an inhibitor of the ADP-ribosylation reaction. J Biol Chem 1992; 267:17397 – 403. [27] Yang JT, Wu C-S, Martinez HM. Calculation of protein conformation from circular dichroism. Methods Enzymol 1986;130:208 – 69.
