This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Request Copyright Permission Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Chow, B. K. C. Articles by Mojsov, S. Search for Related Content PubMed PubMed Citation Articles by Chow, B. K. C. Articles by Mojsov, S. Pubmed/NCBI databases Gene GEO Profiles HomoloGene Protein UniGene Compound via MeSH Substance via MeSH

Identification and Characterization of a Glucagon Receptor from the Goldfish Carassius auratus: Implications for the Evolution of the Ligand Specificity of Glucagon Receptors in Vertebrates

Department of Zoology (B.K.C.C., R.L.C.H., C.-M.Y.), University of Hong Kong, Hong Kong; Department of Biology (T.W.M.), University of Ottawa, Ottawa, Ontario, Canada K1N 6N5; Weill Medical College of Cornell University (P.J.C.) and The Rockefeller University (M.M., S.M.), New York, New York 10021-6399

Address all correspondence and requests for reprints to: Svetlana Mojsov, The Rockefeller University, 1230 York Avenue, New York, New York 10021-6399. E-mail: mojsov{at}mail.rockefeller.edu.

	Abstract

Top Abstract Introduction Materials and Methods Results Discussion References

 
The structural basis of ligand selectivity of G protein-coupled receptors for metabolic hormones has been an area of intense investigation, and yet it remains unresolved. One approach to delineating the mechanism of ligand-receptor interactions is to compare the ligand specificities of receptors expressed in species that emerged at different times within vertebrate evolution. In this paper we describe the isolation, functional, and phylogenetic characterization of the glucagon receptor from the goldfish Carassius auratus (Teleostei, order Cypriniformes), and compare its ligand specificity with that of the homologous rat receptor. Goldfish (gf) glucagon stimulated glucose production in a dose-dependent manner from isolated goldfish hepatocytes, resulting in 5-fold increase at 1 µM. The goldfish glucagon receptor (gfGlucR) shares 56, 51, 50, and 52% amino acid identities with frog Rana tigrina regulosa, mouse, rat, and human glucagon receptors, respectively. In competitive binding experiments, the recombinant gfGlucR displays high affinity toward goldfish, zebrafish, and human glucagons (IC₅₀ = 0.6, 9, and 13 nM, respectively) but not toward goldfish glucagon-like peptide-1 or human glucagon-like peptide-1 (7–36) amide. Whereas both goldfish and human glucagons stimulated dose-dependent increases in intracellular cAMP through the recombinant gfGlucR, the recombinant rat GlucR interacted only with human glucagon, analogous to the specificity of the previously characterized glucagon receptor from the frog R. tigrina regulosa. Our results demonstrate that the binding pocket of gfGlucR can accommodate a broad range of glucagon structures and that in the frogs and mammals, there is a structural switch to a more restrictive conformation of glucagon receptors.

	Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

 
THE FUNCTIONAL AND structural evolution of glucagon and glucagon receptor genes in vertebrates has been an area of intense focus but remains unresolved. In mammals, proglucagon is the precursor for a number of peptides including glicentin-related polypeptide, glicentin, glucagon, oxyntomodulin, and glucagon-like peptides (GLPs)-1 and –2 (1, 2, 3). The proglucagon gene was found to be expressed in intestinal L cells and pancreatic A cells as well as in the brain stem and hypothalamus (4, 5, 6). Differential processing of the proglucagon precursors in the intestine and pancreas (1, 2) gives rise to distinct peptides and, among them, glucagon and GLP-1 are key metabolic hormones that regulate glucose homeostasis (reviewed in Ref.3 , 7). Glucagon and GLP-1 have been found in different species of fish and other nonmammalian vertebrates, including cyclostomes (lamprey (8, 9), holocephalans (ratfish) (10), teleosts (catfish) (11), salmon (12), trout (13), goldfish (14), frogs (15, 16, 17, 18), and chicken (13).

Functionally, the best known and principal functions of glucagon in mammals and fish are its potent glycogenolytic and gluconeogenic actions on liver resulting in hyperglycemia (19, 20). In contrast to glucagon, GLP-1 stimulates different regulatory pathways in fish and mammalian glucose metabolism (20). In mammals, GLP-1 is insulinotropic (21, 22), antagonizing the actions of glucagon, whereas in fish, GLP-1 acts directly on the liver, eliciting glucagon-like effects (23, 24, 25, 26). The apparent functional switch of GLP-1 effects in fish and mammals (20) remains an interesting yet unresolved issue. There is evidence that fish glucagon and GLP-1, although sharing similar functions, act through different receptors (27). Data on receptor binding of glucagon in fish are sometimes inconsistent with their action. For example, glucagon binds to and induces glycogenolysis in hepatocytes isolated from American eel and brown bullhead, but glucagon does not bind to salmonid hepatocytes, even though the rainbow trout hepatic tissue glycogenolysis is responsive to glucagon (27). It is clear that understanding the function of glucagon and GLP-1 in fish and other nonmammalian vertebrates has been hampered to a large degree by the lack of information at the molecular level with respect to glucagon and GLP-1 receptors. We previously characterized a glucagon receptor from the frog Rana tigrina rugulosa that was functionally similar, if not identical, to the human glucagon receptor (28). The frog glucagon receptor, however, was unable to bind fish glucagons (i.e. goldfish and zebrafish). This could be explained by the fact that fish glucagons contain substitutions in several key residues (see Fig. 1) that were previously shown to be critical for the binding of human glucagon to its receptor (29, 30). These findings suggested that fish glucagon receptors would have different ligand specificities than either the frog or human glucagon receptors. In the present study, we isolated and characterized the first glucagon receptor from fish using the goldfish Carassius auratus to test the above hypothesis and to provide further understanding of the structural and functional evolution of glucagon and its receptor in vertebrates.

View larger version (11K): [in this window] [in a new window]   FIG. 1. Amino acid sequences of fish (goldfish, zebrafish), frog (R. tigrina regulosa), and human glucagons and the des-His1-[Nle9-Ala11-Ala16] glucagon antagonist. Shaded areas represent identical residues. The sequence of goldfish glucagon is from Ref.14 and zebrafish glucagon from Ref.44 .

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

 
Goldfish hepatocyte isolation
Goldfish hepatocytes were prepared by collagenase treatment of liver pieces (31). All chemicals were prepared in a basic Hanks’ medium consisting of (in millimoles) 136.9 NaCl, 5.4 KCl, 0.8 MgSO₄·7H₂O, 0.33 Na₂HPO₄·7H₂O, 0.44 KH₂PO₄, 10 HEPES, and 5 NaHCO₃ (pH adjusted to 7.6). The liver was rinsed in a Hanks’-EGTA (1 mM) medium, diced finely with a razor blade, and the Hanks’-EGTA medium replaced by a Hanks’-collagenase (0.65 mg or 3000 U type IV collagenase per milliliter; Sigma Chemical Co., St. Louis, MO) medium to approximately 1 g liver per 20 ml of medium. The pieces were shaken in a 50-ml glass Erlenmeyer flask for 30 min at room temperature, removed, and massaged carefully through two nylon screens (250- and 75-mm mesh sizes). The cells were collected by centrifugation (3 min at 800 rpm in a Sorvall RC-5B at 4 C) and rinsed several times with basic Hanks’. The final rinse contained 50% complete Hanks’ medium (1.5 mM CaCl₂, 2% BSA, and 100 mg bacitracin, 2 ml MEM essential and 1 ml MEM nonessential amino acids in 100 ml). The cells were resuspended in the complete Hanks’ (approximately 100 mg cells/ml) and were allowed to rest on ice for 1.5–2 h before experimentation. Hepatocyte viability was checked with trypan blue and found always to be greater than 90%.

Functional studies of the goldfish glucagon and GLP-1 in the activation of glycogenolysis in goldfish hepatocytes
Cells (50 µl containing approximately 20 mg of cells for GLP-1 or 10 mg for glucagon experiments) were aliquotted into the wells of a 48-well microtiter plate. These wells already contained hormone at concentrations from 10^–10 to 10^–6 M. The final well volume was 100 µl. The microtiter plate was covered and shaken slowly at room temperature for 30 min at which time 17.5 µl of 17.5% perchloric acid was added to each well; the plate was rigorously shaken to ensure mixing. The deproteinized medium was transferred, vortexed, and centrifuged for 3 min at 12,000 x g in a microfuge. Glucose was determined using the hexokinase-glucose-6-P-dehydrogenase enzyme-coupled method adopted for microplate assays (32). Glucose production in the presence of peptide is presented as fold change, compared with baseline values (without hormone addition).

Cloning of the partial receptor cDNAs from several vertebrate species
Partial GlucR-like cDNA clones corresponding to transmembrane domains (TMDs) II-VI of vertebrate receptors for glucagon-like peptides were obtained by a two-step PCR approach essentially following a protocol described earlier (33). The primers used for PCR were designed according to the consensus regions among receptors of the glucagon-secretin receptor family. The sequences of the primers are: G2 (TGCAYTGYACNMGNAAYTAYATYCA), G6 (AGSGGGATSAGSRKNAGNGTGGAYTT), and G7 (TGSACCTCNCCRTTNASRAARCARTA). First-strand cDNAs were prepared from the pancreas and brain of various vertebrates. After the first PCR using G2 and G7 as primers, 1 µl of the reaction mix was reamplified in the second PCR using primers G2 and G6. PCR products between 500 and 600 bp were agarose gel purified and blunt-end cloned into pBlueScript SK+ for DNA sequence analysis using a T7 DNA sequencing kit (Pharmacia, Piscataway, NJ). The PCR contained 50 pmol of primers, 200 µM of deoxynucleotide triphosphates, and 2.5 U of Taq polymerase (Life Technologies, Gaithersburg, MD) in the buffer provided by the manufacturer. Reaction times were 1 min at 94 C, 58 C, and 72 C, respectively, for 30 cycles. A preliminary phylogenetic tree of the partial GlucRs was obtained using the GeneWorks (IntelliGenetics, Mountain View, CA) protein alignment program.

Molecular cloning of a full-length gfGlucR cDNA
The putative partial gfGlucR cDNA fragment was radioactively labeled by the Megaprime DNA labeling system (Amersham, Arlington Heights, IL) and [-³²P]-dATP (3000 Ci/mmol, Amersham) and was used as a probe to screen a combined goldfish brain/pituitary cDNA library (1 million primary clones) as described earlier (33). Three primary positive clones that showed consistent hybridization signals were identified. Sequencing analysis showed that one of them was full length and contained the ATG initiation codon. This clone was isolated and then excised to the phagemid, pBK-gfGluRA, according to instructions from the manufacturer (Stratagene, Cambridge, UK). It was sequenced from both strands using a T7 sequencing kit (Pharmacia) by synthetic primers and subcloning of restriction fragments. The DNA sequences were analyzed by the DNasis version 2.0 program (Hitachi, San Bruno, CA) (nucleotide sequence deposited in GenBank, accession no. bankit608259 AY584244).

Stimulation of intracellular cAMP levels in COS-7 cells expressing the goldfish and rat glucagon receptors
The bacterial lac-promoter was released from the construct pBK-gfGluRA by excising the EcoRI/NheI restriction fragment to produce the eukaryotic expression plasmid pBK-gfGluR. A permanent cell line, COS-gfGlucR, with the gfGlucR expressed was obtained by transfecting the plasmid pBK-gfGlucR (10 µg) into 1 million COS-7 cells using the Lipofectamine reagent (BRL/Gibco) and followed by G418 selection at 500 µg/ml (BRL/Gibco, Gaithersburg, MD) for 2 wk. Functional expression and intracellular cAMP assays were performed as described earlier (33). In summary, 0.2 million cells were seeded onto 6-well plates (Costar, Cambridge, MA) 2 d before cAMP assays. The cells were washed once with MEM containing 1 mg/ml BSA and then incubated with the same medium containing 0.2 mM of 1-methyl-3-isobutylxanthine and the desired concentrations of peptide for 45 min at 37 C. After incubation, the medium was removed and the cells were lysed by the addition of 1 ml iced ethanol. The cell debris was pelleted by centrifugation (10,000 x g for 10 min), and the supernatant was dried using a vacuum concentrator. In the initial characterizations, cAMP levels were quantified using a RIA kit (Amersham). Results from the initial characterization are from six independent peptide stimulations, and the values are represented as means ± SEM; the average value of basal cAMP level was 5.9 pmol/well. All human peptides used in this study were purchased either from Bachem Fine Chemicals, Inc. (Torrance, CA), or Peninsula Laboratories Inc. (Belmont, CA). The goldfish glucagon and GLP-1 peptides were custom synthesized and purified (>90%) by Peninsula Laboratories. Zebrafish glucagon was synthesized by the Protein/DNA Technology Center at Rockefeller University.

In a different set of experiments, intracellular cAMP levels were measured in COS-7 cells that express transiently either the gfGlucR or rat glucagon receptor as described previously (28). Briefly, for the experiments with the gfGlucR, the expression plasmid pBK-CMV-gfGlucR was transfected into COS-7 cells and grown in 100-mm plates, using TransFast transfection reagent (Promega, Madison, WI) according to the manufacturer’s instructions. For the experiments with the rat glucagon receptor, the expression vector pMT5 containing the synthetic glucagon receptor gene (34) (a generous gift of Drs. C. Unson and T. Sakmar, The Rockefeller University, New York, NY) was transfected into COS-7 cells under the same conditions used for the gfGlucR. After 24 h transfected cells were seeded onto 24-well plates (Biocore, Becton Dickinson, Franklin Lakes, NJ), and the dose-dependent stimulation with the test peptides (picomoles-micromoles) was typically performed 24–48 h later. Before the addition of peptides, cells were preincubated with the assay buffer [DMEM, 0.5% BSA, 20 mM HEPES, 0.5 mM 1-methyl-3-isobutylxanthine, 0.1 mM phenylmethylsulfonyl fluoride, pH 7.4)] for 20 min at 37 C. Each peptide concentration was applied to triplicate wells. Incubations were for 20 min at 37 C. Sample work-up was the same as described above. In this set of experiments, intracellular cAMP levels were determined using the enzymatic immunoassay kit (Cayman Chemicals, Ann Arbor, MI). Forskolin (100 nM) was used as a positive control in triplicate wells on each plate. Results presented here for the gfGluc-R are the average of three separate experiments and for the rat GlucR the average of two separate experiments performed with separate transfections several weeks or several months apart. The basal concentration of intracellular cAMP (in the absence of peptides) was typically 11–14 pmol/well and was taken as 1. Results are presented as a fold increase in intracellular cAMP levels over basal.

Experimental conditions for dose-dependent cAMP stimulations with goldfish and human glucagons performed in the presence of the des-His1-[Nle9-Ala11-Ala16] glucagon antagonist (1 µM or 100 nM) (35) (a generous gift of Dr. C. Unson) were the same as described above, except that following the 20-min preincubation with the assay medium at 37 C, cells were preincubated for an additional 10 min with the antagonist before the addition of either goldfish or human glucagons (picomoles-micromoles) to COS-7 cells expressing transiently either goldfish or rat glucagon receptors. In addition, in each experiment des-His1-[Nle9-Ala11-Ala16] glucagon antagonist was added by itself in a dose-dependent manner (picomoles-micromoles) to COS-7 cells transfected with either goldfish or rat glucagon receptors. As before, forskolin (100 nM) was added in triplicate wells to each plate as a positive control. Results represent an average of three different experiments for the gfGlucR and two for the rat GlucR, each experiment representing a separate transfection performed several weeks or months apart.

Competitive binding experiments with the gfGlucR
Competitive binding experiments were performed with the gfGlucR that was expressed transiently in COS-7 cells. Experimental conditions for the competitive binding assays were similar to the ones described for the frog glucagon receptor (28). The ¹²⁵I-goldfish glucagon was custom iodinated by NEN Life Science Products (Boston, MA) with a specific activity of 2200 Ci/mmol. Assays were performed in 24-well plates (Biocore, Becton Dickinson). In brief, COS-7 cells expressing transiently the recombinant gfGlucR were incubated with ¹²⁵I-goldfish glucagon (approximately 100,000 cpm/well) in the presence of increased concentrations of test peptides (picomoles-micromoles) for 16–18 h at 4 C. Stock solutions of peptides (10 µM) were diluted in a binding buffer [Hanks’ balanced salt solution containing 20 mM HEPES (pH 7.4), 0.5% BSA, and 0.1% phenylmethylsulfonyl fluoride] to give a range of concentrations from 10^–11 to 10^–6 M. Each peptide concentration was added to triplicate wells before the addition of ¹²⁵I-goldfish glucagon. After incubation, cells were washed twice with ice-cold PBS, lysed with 1 N NaOH, and radioactivity measured in a -counter. Nonspecific binding was determined in the presence of either 1 µM goldfish or 1 µM human glucagon or 1 µM des-His1 [Nle9-Ala11-Ala16] glucagon antagonist and was typically 0.6%, whereas the specific binding was between 1.5 and 2.3% of the total radioactivity added. Results represent an average of three (goldfish glucagon and des-His1-[Nle 9-Ala11-Ala16] glucagon antagonist) or two [human and zebrafish glucagons, goldfish GLP-1, and hGLP-1 (7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36) amide] separate experiments, each performed with a different batch of COS-7 transfected cells but with the same lot of ¹²⁵I-goldfish glucagon. The competitive binding experiments were also performed with ¹²⁵I-human glucagon (2200 mCi/mmol receptor grade, NEN Life Science Products), goldfish and human glucagons (three separate experiments),and goldfish GLP-1 and hGLP-1 (7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36) amide (data not shown).

Tissue distribution of the gfGlucR mRNA
Distribution of gfGlucR transcripts in various peripheral tissues, pituitary, and brain was investigated by both Northern blot analysis and RT-PCR coupled to Southern blot analysis. Tissues from 50 goldfish were pooled to investigate the tissue distribution of gfGlucR by Northern and Southern blot. A second experiment that used only the RT-PCR approach coupled to Southern blot analysis used pooled tissues from 18 goldfish. Total RNA was prepared using the acid guanidinium thiocyanate-phenol-chloroform method (36). The integrity of total RNA was initially checked by formaldehyde gel electrophoresis. For the Northern blot analysis, 5 and 8 µg of total RNA from pituitary and other tissues, respectively, were used for the preparation of mRNA using the PolyATract isolation system IV (Promega) following the manufacturer’s procedure. mRNA was subjected to electrophoresis using a 1% RNA gel and transferred to nitrocellulose membrane. A 233-bp partial gfGluc-R fragment (from nucleotides 347–569) was used as a probe to detect the transcripts. For the RT-PCR approach coupled to Southern blot analysis, 5 µg of total RNA was used for the preparation of the first-strand cDNA using Superscript first-strand synthesis system (Invitrogen, Carlsbad, CA). PCR was performed using sequence specific primers: F, 5'-ATGTCACAAGTATTCCTC-3'; and R, 5'-CCACACTCACTGTAGTAT-3'. PCR conditions were 1 min at 94 C, 1 min at 55 C, and 1 min at 72 C for 30 cycles. The probe used for Southern blot analysis was a 233-bp partial gfGlucR cDNA fragment (from nucleotides 347–569); its sequence was confirmed by sequence analysis. Actin-specific primers were used as a positive control.

Phylogenetic analysis
Full-length and partial amino acid sequences (corresponding to TMD II to the beginning of TMD VI) of the known vertebrate receptors for glucagon, GLP-1, GLP-2, glucose-dependent insulinotropic polypeptide (GIP), and vasoactive intestinal polypeptide (VIP) were aligned using Clustal W program (37). Phylogenetic analyses of the aligned sequences were performed by PHYLIP 3.572 program package (38), and phylogenetic trees were constructed using corrected protein distances calculated by neighbor-joining methods (SEQBOOT, PRODTIST, NEIGHBOR JOINING, and CONSENSE programs). In addition, trees were also constructed by parsimony methods (PROTPARS programs). In both methods, the confidence level supporting each node of the unrooted tree was calculated by the bootstrap technique (500 replications) using the sequence of goldfish VIP receptor as the outgroup for analysis of the full-length sequences and the sequence of hGLP-1 receptor as an outgroup for analysis of the partial amino sequences of vertebrate glucagon and GLP-1 receptors.

Statistical analysis
Results from the functional characterizations of the gfGlucR and rat GlucR were analyzed using Prism software (39, 40) and plotted using Origin software (41) and are presented as mean ± SEM. Data from the activation of the glycogenolysis in goldfish hepatocytes, competitive binding experiments, and cAMP stimulations performed with the recombinant gfGlucR and rat GlucR were analyzed using a four-parameter logistic sigmoidal curve fit model (39, 40). The IC₅₀ values, representing the concentration of peptides that inhibited the specific binding by 50%, and EC₅₀ values, representing peptide concentrations that elicit 50% of the maximum response, were calculated from the midpoints on the curves (log IC₅₀ or log EC₅₀) and are shown together with 95% confidence intervals, defined as intervals that have 95% confidence that in the long run the intervals constructed in this way will indeed contain the true value. Comparison of the goldfish glucagon displacement curve with displacement curves obtained with zebrafish and human glucagons and the glucagon antagonist in the competitive binding experiments were calculated using the F test, which compares statistically the fitted midpoints of two curves (39). For example, F test compared the fitted midpoint (log IC₅₀) from the data sets in the goldfish glucagon displacement curve with the fitted midpoints (log IC₅₀s) from the data sets in the zebrafish, human, or the glucagon antagonist displacement curves, respectively. Differences between the two curves are represented by P values, where P < 0.05 represents statistically significant differences. The F test was also used to compare the cAMP dose-response curves obtained with goldfish glucagon and human glucagon stimulations in the absence and presence of 100 nM and 1 µM glucagon antagonist.

The functional experiments were also evaluated by two-sample t test and the nonparametric Wilcoxon rank-sum test that compared differences across groups (i.e. responses elicited by goldfish glucagon vs. other peptides tested) and across concentration levels for each peptide tested. All P values are two sided with statistical significance evaluated at 0.05 alpha level. The Intercooled Stata 8.0 program was used for all analysis (42).

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

 
Synthetic goldfish glucagon was used in all experiments; its amino acid sequence was deduced from the nucleotide sequence of goldfish proglucagon cDNA (14). Because goldfish is tetraploid, there are potentially four different glucagon sequences encoded by four potential glucagon genes; because only a single proglucagon cDNA has been isolated so far (14), we characterized the functional properties of gfGlucR with one of the four potential goldfish glucagons and GLP-1s. We have previously shown that synthetic gfGLP-1 and zebrafish glucagon are biologically active (43, 44).

Activation of goldfish hepatic glycogenolysis by goldfish glucagon
Hepatic glycogenolysis was used to characterize the effects of goldfish glucagon on goldfish liver. Basal glucose production by the goldfish hepatocytes was approximately 9.5 µmol/h/g. This value is in the same range as observed for American eel and rainbow trout but nearly 3 times higher than those in brown bullhead hepatocytes (24). Goldfish glucagon increased glucose production by about 5-fold at a concentration of 10^–6 M with an EC₅₀ of 55 nM (95% confidence interval in the concentration range 27 nM to 109 nM) (n = 11) (Fig. 2). The increase in glucose levels obtained upon stimulation with increasing concentrations of goldfish glucagon (10^–9 M to 10^–6 M) was significantly different (P < 0.05) from the glucose level measured after incubation with the lowest concentration of goldfish glucagon tested (10^–10 M). The effect of synthetic goldfish GLP-1 on glycogenolysis was small, with a maximum 2-fold increase detected at a concentration of 10^–6 M (n = 7) (Fig. 2). Glucose levels measured after stimulations with goldfish glucagon (10^–10 M to 10^–6 M) were significantly different (P < 0.0001) from the levels measured after goldfish GLP-1 stimulations (10^–10 M to10^–6 M) starting at a goldfish glucagon and a gfGLP-1 concentration of 10^–9 M.

View larger version (14K): [in this window] [in a new window]   FIG. 2. Activation of glycogenolysis in goldfish hepatocytes by goldfish glucagon () and gfGLP-1 (). Data are presented as fold increase in glucose production above baseline values (no hormone addition) (n = 11 for goldfish glucagon, n = 7 for gf GLP-1). Baseline glucose production rates were 9.3 ± 1.0 µmol/h/g hepatocytes for goldfish glucagon (n = 11) and 9.9 ± 1.1 µmol/h/g hepatocytes for gfGLP-1 (n = 7). Glucose was assayed using an enzymatic method in 10 (for goldfish glucagon) or 20 (for gfGLP-1) mg of cells. Data were analyzed by Prism software (39 40 ) using four-parameter logistic sigmoidal curve fit model. Results are expressed as mean ± SEM and are plotted using Origin software (41 ).

View larger version (121K): [in this window] [in a new window]   FIG. 3. Amino acid sequence alignment of goldfish, frog (R. tigrina regulosa), mouse, rat, and human glucagon receptors. Conserved amino acids are shaded. Putative TMDs are overlined and labeled I-VII. # and *, Potential N-linked glycosylation sites and conserved cysteine residues, respectively. Gaps (represented by –) are introduced to maximize sequence homology.

View larger version (21K): [in this window] [in a new window]   FIG. 4. Functional characterization of the gfGlucR by measurement of intracellular cAMP accumulation. Dose-response cAMP studies of the goldfish glucagon receptor expressed in COS-7 cells upon stimulation with goldfish glucagon (), human glucagon (), gfGLP-1(), hGLP-1 (), or hGIP (). Intracellular cAMP levels were measured and expressed as percent maximum cAMP response relative to the negative control (media alone without peptide addition). Data presented here were from six independent peptide stimulations. Analysis was by Prism software using a four-parameter logistic sigmoidal curve fit (39 40 ) and plotted using Origin software (41 ). Values on the curves represent mean ± SEM; the average value of basal cAMP levels was 5.9 pmol/well.

View larger version (24K): [in this window] [in a new window]   FIG. 5. Recombinant goldfish glucagon receptor binds fish (i.e. goldfish and zebrafish) and human glucagons and des-His-1-[Nle9-Ala11-Ala16] glucagon antagonist but not goldfish GLP-1 or mammalian GLP-1 (7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 ) amide. Competitive binding experiments performed with the recombinant gfGlucR expressed transiently in COS-7 cells in the presence of 125I-goldfish glucagon (2200 mCi/mmol) and increasing concentrations of test peptides (picomoles-micromoles): goldfish glucagon (), human glucagon (), zebrafish glucagon (), des His-1 [Nle9-Ala11-Ala16] glucagon antagonist (), gfGLP-1(), hGLP-1 (7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 ) amide (). Each peptide concentration was added to triplicate wells. Results represent an average of three separate experiments using goldfish glucagon and des-His1-[Nle9-Ala11-Ala16] glucagon antagonist and two with human and zebrafish glucagons, gfGLP-1 and hGLP-1 (7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 ) amide, each performed with different batches of COS-7 transfected cells but same lot of 125I-goldfish glucagon (see Materials and Methods). Data were analyzed by Prism software (39 40 ) using four-parameter logistic sigmoidal curve fit. Results are expressed as mean ± SEM and were plotted using Origin software (41 ).

A second set of experiments used the des-His1-[Nle9-Ala11-Ala16] glucagon antagonist (35) in competitive binding experiments to compare further the ligand specificities of goldfish and frog glucagon receptors. Des-His1-[Nle9-Ala11-Ala16] glucagon antagonist displaced ¹²⁵I-goldfish glucagon bound to the gfGlucR with an IC₅₀ of 7 nM (95% confidence interval in the concentration range 4.4–11 nM) (Fig. 5 and Table 1). This value is similar to the IC₅₀ of 7.8 nM obtained for the frog glucagon receptor (28). Its competitive binding curve is significantly different from the one obtained when goldfish glucagon was used to displace the bound ¹²⁵I-gf glucagon (P < 0.0001) (Fig. 5).

Comparison of the functional properties of the goldfish and rat glucagon receptors
The des-His1-[Nle9-Ala11-Ala16] glucagon antagonist (35) was used to compare directly the functional properties of the goldfish and rat glucagon receptors. This set of experiments measured the effects of the antagonist and goldfish and human glucagons on intracellular cAMP levels in COS-7 cells that express transiently the recombinant goldfish or rat glucagon receptors. Using the same approach, we previously showed that the binding of the antagonist to the frog glucagon receptor (28) did not have any effects on intracellular cAMP levels and, at a concentration of 100 nM, inhibited the human glucagon-stimulated increase in intracellular cAMP.

The cAMP responses mediated through recombinant rat glucagon receptor (Fig. 6A) or gfGlucR (Fig. 6B) were characterized after binding of the glucagon antagonist, goldfish glucagon, or human glucagon. The binding of the antagonist had no effect on intracellular cAMP levels mediated either through the recombinant rat glucagon receptor (Fig. 6A) or the recombinant gfGlucR (Fig. 6B) at all concentrations tested (picomoles-micromoles). These findings are in agreement with the results obtained with the recombinant frog glucagon receptor (28) and the rat glucagon receptor present in rat liver membranes (35). Human glucagon stimulated intracellular cAMP through both rat (Fig. 6A) and goldfish (Fig. 6B) glucagon receptors in the same concentration range and with similar EC₅₀ values (11.6 nM, Fig. 6A, and 9.7 nM, Fig. 6B) (Table 2) and to a similar extent (maximum increase at 1 µM was 5.8-fold, Fig. 6A, and 7.2-fold, Fig. 6B). Results shown in Fig. 6B are in agreement with the results obtained during the initial characterization of the recombinant gfGlucR (Fig. 4). In contrast, goldfish glucagon was able to stimulate in a dose-dependent manner the increase in intracellular cAMP levels only through the recombinant gfGlucR (Fig. 6B) with an EC₅₀ value of 3 nM but not through the recombinant rat glucagon receptor (Fig. 6A and Table 2). These results establish that the goldfish and rat glucagon receptors have different specificities toward goldfish and human glucagons.

View larger version (13K): [in this window] [in a new window]   FIG. 6. Goldfish glucagon does not stimulate an increase in intracellular cAMP levels mediated through the recombinant rat glucagon receptor (A), but human glucagon stimulates in a dose-dependent manner an increase in intracellular cAMP levels mediated through the recombinant gfGlucR (B). Intracellular cAMP levels were measured in COS-7 cells that were transiently transfected with either the rat glucagon receptor (A) or goldfish glucagon receptor (B) after incubation with goldfish glucagon (), human glucagon (), or des-His1-[Nle9-Ala11-Ala16] glucagon antagonist () for 20 min at 37 C. Results represent the mean ± SEM of three separate experiments with gfGlucR (B) and two separate experiments with rat GlucR (A). In each experiment peptides (picomoles-micromoles) were added to triplicate wells in 24-well plates. Results are presented as fold increase over basal (in the absence of peptide), which is taken as 1. The basal concentration of intracellular cAMP was 11–14 pmol/well (see Materials and Methods). Dose-response curves were analyzed by Prism software (39 40 ) using a four-parameters logistic sigmoidal curve fit model and were plotted using Origin software (41 ).

View larger version (19K): [in this window] [in a new window]   FIG. 7. Des-His-1-[Nle9-Ala11-Ala16] glucagon antagonist inhibits in a dose-dependent manner the goldfish glucagon-stimulated increase in intracellular cAMP levels mediated through the recombinant goldfish glucagon receptor. COS-7 cells transiently transfected with the gfGlucR were incubated for 20 min at 37 C with des-His1-[Nle9-Ala11-Ala16] antagonist at 1 µM () or 100 nM () before the addition of increasing concentrations of goldfish glucagon () (picomoles-micromoles) for an additional 20 min at 37 C. Values represent means ± SEM of three separate experiments in which each peptide concentration (picomoles-micromoles) was added to triplicate wells in 24-well plates. In each experiment the inability of des-His1-[Nle9-Ala11-Ala16] glucagon antagonist to stimulate cAMP levels was confirmed by adding the antagonist alone in the concentration range of picomoles to micromoles for 20 min at 37 C to the same batch of COS-7 cells (see Fig. 6B). Results are presented as fold increase over basal concentrations (in the absence of goldfish glucagon), which were 13.5 pmol/well (see Materials and Methods). Results with the goldfish glucagon stimulation alone are identical with those shown in Fig. 6B. Data were analyzed and plotted as described in the legend to Fig. 6.

View larger version (13K): [in this window] [in a new window]   FIG. 8. Des-His-1-[Nle9-Ala11-Ala16] glucagon antagonist inhibits in a dose-dependent manner (, 1 µM; , 100 nM) and to a similar extent the human glucagon ()-stimulated increase in intracellular cAMP levels mediated through the recombinant rat glucagon receptor (A) and recombinant goldfish glucagon receptor (B). Values represent means ± SEM of two separate experiments for rat GlucR (A) and three for gfGlucR (B). Experimental conditions are as described in Fig. 7 legend and noted in Materials and Methods. Results with the human glucagon stimulation alone are the same as those shown in Fig. 6, A and B. Data were analyzed and plotted as described in the legend to Fig. 6.

Tissue distribution of the gfGlucR mRNA
Distribution of the gfGlucR mRNA in goldfish tissues was investigated by both Northern blot analysis and RT-PCR combined with Southern blot analyses. A 233-bp partial gfGlucR cDNA fragment corresponding to nucleotides 347–569 was used to detect gfGlucR transcripts. Signals were detected by both methods in liver, skeletal muscle, gut, kidney, gill, heart, pituitary, brain, gall bladder, and male and female gonadal tissues (data not shown). Detection of gfGlucR transcripts in liver is consistent with the known primary function of glucagon in regulating hepatic glycogenolysis and gluconeogenesis and with our results with goldfish hepatocytes (Fig. 2). Although, this wide tissue distribution of gfGlucR transcripts may seem surprising, it is similar to that described for the expression of glucagon receptor transcripts in rat tissues (46, 47), indicating that in both fish and mammals, glucagon may have effects in tissues that have not been identified in physiological and biochemical experiments as targets for its action. Due to the small size of goldfish, we were unable to isolate mRNA from additional tissues (i.e. pancreatic islets) to fully compare the distribution of gfGlucR transcripts in goldfish and rat tissues.

Detection of gfGlucR transcripts in goldfish brain and gut is consistent with the recently described extrahepatic effects of glucagons (i.e. zebrafish, bovine) in fish (48). For example, zebrafish glucagon stimulates adenylyl cyclase in isolated rockfish (Sebastes caurinus) brain membranes (44), and zebrafish and bovine glucagons increase cAMP levels and uptake of 3-O-methyl-D-glucose in rockfish (S. caurinus) and brown bullhead (Ictalurus melas) enterocytes, respectively (44, 49).

Evolutionary relationship of the gfGlucR to vertebrate glucagon-like receptors
To determine the evolutionary relationship of gfGlucR to the known vertebrate glucagon-like receptors, we aligned the full-length sequences of 15 known receptors for glucagon, GLP-1, GLP-2, and GIP as well as sequences of goldfish, rat, and human VIP receptors. The sequence of the related gfVIP receptor belonging to a different subclass of G proteincoupled receptors was used as the outgroup for analysis by the neighbor-joining and parsimony methods to generate unrooted phylogenetic trees. We reasoned that with the limited number of available sequences for glucagon-like receptors, the unrooted tree would provide the most reliable information about the degree of the phylogenetic relationship among them. The bootstrapping method (500 replications) was used in both methods to assess the strength of support for each branch of the tree. The phylogenetic tree obtained by the neighbor joining method is shown in Fig. 9. The branch lengths are proportional to the number of amino acid changes that have taken place along a given branch. The tree reconstructed by the parsimony method showed an identical relationship among these receptors (data not shown). Both methods established that gfGlucR is the most closely related to the two known fish (i.e. goldfish and zebrafish) GLP-1 receptors. Moreover, the phylogenetic tree indicates that the two known fish GLP-1 receptors originate from the same ancestor as all the vertebrate glucagon receptors. These evolutionary relationships are strongly supported by the bootstrapping calculations, which demonstrated that 92% of replications lead to the vertebrate GlucR branch and that in the GlucR branch goldfish and zebrafish GLP-1 receptors are separated from the gfGlucR in 100% of replications. Both neighbor-joining and parsimony methods generate a tree in which mammalian GLP-1 receptors originate from a different ancestor than vertebrate glucagon receptors. This branch is less well supported (71% of replications), most likely due to limited number of available nonmammalian GLP-1R sequences. The overall relationship among the glucagon-like receptors and especially the place of mammalian GIP receptors in the phylogenetic tree is in a good agreement with the published observations (50) and is well supported by bootstrapping.

View larger version (21K): [in this window] [in a new window]   FIG. 9. Phylogenetic relationship of gf GlucR to the receptors for glucagon-like peptides. The unrooted tree was obtained by the neighbor-joining method using gfVIP receptor sequence as an outgroup for analysis. The length of each branch is proportional to the number of amino acid changes that have taken place along each branch with the scale shown on the lower left. The support for each node was calculated by bootstrapping and is indicated as percent of 500 replications. An identical tree was obtained by the parsimony method. Glucagon receptor sequences from frogs R. tigrina, R. pipens, and X. laevis were from Refs.28 , 50 , respectively; rat, human, and mouse glucagon receptor sequences were from Refs.51 , 62 , 63 , respectively; GLP-1 receptor sequences for goldfish, zebrafish, rat, and human were from Refs.43 , 54 , 64 , 65 , respectively; GIP receptor sequences for human and rat were from Refs.66 , 67 , respectively; the sequence of the rat and human GLP-2 receptors were from Refs.68 ; VIP receptor sequences from goldfish, rat, and human were from Refs.33 , 69 , 70 , respectively.

View larger version (119K): [in this window] [in a new window]   FIG. 10. Relationship between seven known vertebrate glucagon receptors, five known GLP-1 receptors, and three glucagon receptors-like identified in this study assessed by sequence comparisons (A) and unrooted phylogenetic tree (B). Partial amino acid sequences from the FMSFI motif within the TMD II to the RLAK immediately preceding TMD VI were aligned using Clustal W program (37 ). Sequences were obtained as indicated in legend to Fig. 9. The sequence of the mouse GLP-1R was from Ref.71 . Dark- and light-shaded areas (A) represent identical and conserved amino acids, respectively. The numbers to the left and right of the sequences (A) represent the position of amino acids in the full-length sequence. Note that these are absent in the salmon, pigeon, and snake sequences because only their partial sequences were obtained. The sequences from (A) were used to generate unrooted tree by the neighbor-joining method. The sequence of the hGLP-1R was used as an outgroup for the analysis. Tree is characterized by the same parameters as described in legend to Fig. 9. An identical tree was obtained by the parsimony method.

View larger version (22K): [in this window] [in a new window]   FIG. 10A. Continued

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

 
This paper describes the isolation and functional and phylogenetic characterization of the goldfish glucagon receptor, the first glucagon receptor studied to date in fish. Our experiments demonstrate that gfGlucR can recognize with high affinity only fish and human glucagons and not the structurally related gfGLP-1 and hGLP-1 (7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36) amide (Figs. 4 and 5). This finding is in agreement with the described ligand selectivity of the frog (R. tigrina rugulosa) (28) and rat GlucRs (34, 51). However, the binding pocket of gfGlucR has different specificity toward glucagon sequences than the binding pockets of frog R. tigrina regulosa and rat GlucRs. Thus, the recombinant gfGlucR can bind goldfish, zebrafish, and human glucagons (Fig. 5), whereas the recombinant frog glucagon receptor binds only human glucagon but not goldfish or zebrafish glucagons (28). These findings were extended in functional experiments that measured the intracellular effects triggered by binding of goldfish and human glucagons to the recombinant gfGlucR and rat GlucR (Figs. 6–8). Goldfish glucagon failed to stimulate an increase in intracellular cAMP mediated through the recombinant rat GlucR (Fig. 6A), analogous to our earlier findings with the recombinant frog GlucR (28). In contrast, human glucagon stimulated dose-dependent increases of intracellular cAMP mediated by the gfGlucR (Fig. 6B).

These results confirmed our earlier conclusions that frog and mammalian GlucRs have similar if not identical ligand specificities. The switch in ligand specificities of glucagon receptors from a structure that recognizes a broad range of glucagon sequences, like the one found in the gfGlucR, to a structure that recognizes only glucagon sequences found in frogs, avian species (chicken), and mammals correlates with changes in amino acids between teleost fish glucagons and frog, chicken, and mammalian glucagons, respectively, at three key positions in the sequence. These are Asn8 for Ser8; Glu 15 for Asp15; and Thr 16 as in goldfish, zebrafish, catfish, and anglerfish II for Ser 16; Glu 16 as in trout and salmon for Ser 16; or Asp16 as in anglerfish I for Ser16, respectively (Fig. 1 and Refs.20 , 28). These residues (i.e. Ser8, Asp 15, Ser16) have been shown to be critical for the formation of the binding pocket of mammalian glucagon receptors (29, 30). We recently proposed (28) that the frog and mammalian glucagon receptors do not recognize fish glucagons as a result of this triple amino acid substitution in the sequences of teleost fish glucagons.

With the exception of trout II glucagon (13), the known glucagon sequences from teleost fish have identical amino acids up to Tyr 13 (20, 28). Differences in amino acids in the second half of the sequence, after Tyr 13, can vary from six as in anglerfish glucagon I and II, to four as in trout glucagon I and II, or two as in goldfish and zebrafish glucagons (Refs.13 , 20 , and 28 ; and Fig. 1). The most common variation is at the carboxyl terminal residue 29 in which either serine or alanine are found, as in goldfish and zebrafish glucagons (Fig. 1). Because sequences of the other potential goldfish glucagons encoded by the potential four goldfish glucagon genes have yet to be identified, the specificity of the gfGlucR characterized in this study to those gf glucagons can only be inferred from our results. Our functional experiments indicate that all potential goldfish glucagons will likely be recognized by the gfGlucR but not necessarily with the same affinity. These conclusions are based on the following observations. First, results from competitive binding experiments (Fig. 5 and Table 1) demonstrated that despite differences between sequences of goldfish and zebrafish glucagons (glutamic acid at position 24 in zebrafish, compared with alanine in goldfish; alanine at position 29 in zebrafish instead of serine in goldfish) (Fig. 1) (14, 44), the gfGlucR recognizes with high affinity the sequence of zebrafish glucagon. Similarly, sequences of goldfish and human glucagons differ in several amino acids (Fig. 1), yet human glucagon displaces the bound ¹²⁵I-gf glucagon from the gfGlucR with an IC₅₀ of 13 nM (Table 1). Second, even when the ligand specificities of the GlucR toward different glucagon sequences are different as is the case with the specificities toward goldfish and human glucagons (Fig. 5 and Table 1), their intracellular effects are similar, as shown in Figs. 4 and 6B and Table 2. Because the sequences of goldfish and human glucagons differ in seven amino acids (more than found in the known sequences of different fish glucagons), our results suggest that it is likely that the other potential gf glucagons will have similar effects on cAMP levels as the goldfish and human glucagons used in this study.

The ability of goldfish glucagon (Ref.14 and Fig. 1) to activate glycogenolysis from goldfish hepatocytes (Fig. 2) in a similar concentration range as the one described for other fish glucagons (24, 27) further supports our conclusion that the goldfish glucagon used in our functional experiments with the gfGlucR is a physiologically relevant molecule. These conclusions raise the question of how the effects of different goldfish glucagons will be regulated physiologically if they are all able to trigger the same intracellular effects, especially in goldfish liver. One possibility is that control is at the level of gene transcription when different conditions (i.e. availability of nutrients) will induce in a tissue-specific manner the transcription of one of the four potential goldfish glucagon genes. If there are any differences in the amino acid sequences among the potential goldfish glucagons, it is likely that they will be found in residues following Tyr 13, as is the case with the majority of the other known glucagon sequences in teleost fish (i.e. the sequences of anglerfish glucagon I and II). Therefore, it is likely that these goldfish glucagon sequences will not be recognized by either the frog or the mammalian glucagon receptors because they will contain amino acid substitutions in positions 8, 15, and 16.

gfGlucR binds with high affinity the des-His1-[Nle9-Ala 11-Ala16] glucagon antagonist as indicated by the IC₅₀ value of 7 nM (Fig. 5 and Table 1), which is similar to the IC₅₀ value of 7.8 nM determined in the functional experiments with the frog glucagon receptor (28) and in agreement with the results reported for the rat glucagon receptor (35). This confirms the original observations (35, 52, 53) that des-His1-[Nle9-Ala11-Ala16] amide and related glucagon analogs can bind with high affinity to glucagon receptors, even though their sequences contain substitutions of aspartic acid at position 9. These results indicate that the des-His1-[Nle9-Ala 11-Ala16] glucagon antagonist, like human glucagon, may act as a universal glucagon receptor ligand.

Extensive structure-function studies with human glucagon identified Asp-9 as a critical residue needed for the agonist activity of glucagon analogs (53). Replacement of Asp-9, also present in all reported fish glucagons (Ref.20 and Fig. 1), by another amino acid led to antagonists that can also bind with high affinity to the rat glucagon receptor (35). The studies with the gfGlucR confirm and extend these findings, as shown in Figs. 5, 6B, 7, and 8B. Des-His1-[Nle9-Ala 11-Ala16] glucagon does not affect intracellular cAMP levels in COS-7 cells expressing transiently the recombinant gfGlucR, as reported for the frog glucagon receptor (28). More importantly, the binding of the des-His1-[Nle9-Ala11-Ala16] glucagon antagonist to goldfish, frog, and rat glucagon receptors results in a dose-dependent inhibition of goldfish and human glucagon-stimulated increases in cAMP levels (Figs. 6–8 and Ref.28). Collectively, these results establish that Asp-9 is essential for the agonist activity of glucagon mediated through all the known glucagon receptors. In addition, the results demonstrate that the binding of the des-His1-[Nle9-Ala11-Ala16] glucagon antagonist to the gfGlucR interferes with the conformational change(s) needed for its subsequent interaction with the intracellular G proteins, as is the case with the frog and rat glucagon receptors. Therefore, all known glucagon receptors appear to be dependent for their function on similar activation mechanisms.

Our results raise the question of why the binding pocket of gfGlucR can accommodate a wider spectrum of glucagon sequences, whereas the binding pockets of frog R. tigrina regulosa and rat glucagon receptors are restricted to glucagon structured found in frogs and mammals. One plausible explanation is that this feature may have evolved predominantly in fish species to accommodate the binding of different glucagon sequences encoded by multiple fish glucagon genes. Identification of several proglucagons within single teleost species (i.e. anglerfish and trout proglucagon I and II) (20, 28, 13) supports this explanation.

The second related question is how the binding pocket of gfGlucR can accommodate a wider spectrum of glucagon structures than the binding pockets of frog and mammalian GlucRs. One possible explanation is that there are several binding sites within the binding pocket of gfGlucR, possibly overlapping with each other, but each recognizing a different glucagon structure. In other vertebrates, including the frog R. tigrina regulosa and mammals, some of these sites, including the ones recognizing the sequences of fish glucagons, are altered leaving only the sites that recognize the frog and human glucagon sequences (28). A more plausible explanation, however, is that the formation of the binding pocket of glucagon receptors is a multistep process, similar to the one that we proposed recently for the structurally related gfGLP-1 receptor (43). The functional characterization of the recombinant goldfish (43) and zebrafish (54) GLP-1 receptors demonstrated that they could also accommodate a wider range of GLP-1 structures than the hGLP-1 receptor, analogous to the current findings with the goldfish glucagon receptor. We (43) proposed that the initial and critical interaction between GLP-1 and the GLP-1 receptor requires a contact, possibly a salt-bridge formation, between several individual amino acids in GLP-1 and several individual amino acids in the amino terminal domain and the first extracellular loop of the GLP-1 receptor. The second step involves conformational changes in the GLP-1 and GLP-1 receptor structures, leading to the formation of the binding pocket. According to this model, the ability of fish glucagon and GLP-1 receptors to bind a wide range of peptide structures would primarily be defined by their receptor sequences, most likely within the amino terminal domain and the first extracellular loop. This view can also explain in part the differences in the ligand specificities of goldfish, frog, and rat glucagon receptors because the sequence of the gfGlucR has substitutions in some of the key residues that have previously been shown to be necessary for the binding specificity of the human glucagon receptor, i.e. a change of Ser to Ala-75 and Gln to Arg-131 (55).

However, in other parts of the sequence of the gfGlucR, i.e. sequences encompassing the second TMD up to the beginning of TMD VI, there is a high degree of conservation with the sequences of GlucRs isolated from other vertebrates (Fig. 10A). The highest conservation is found in the putative TMDs and intracellular loop between TMD V and TMD VI. This is not surprising because it has been shown in the mammalian glucagon receptors that these regions are key contributors to the proper conformation of the seven transmembrane helices and G protein coupling (56, 57, 58). A highly conserved motif (KLR) is located at the N terminus of the third intracellular loop (Fig. 10A). This motif is found in all glucagon receptors, except in the frog, in which there is a conservative Leu to Met substitution, leading to a KMR sequence. This highly conserved motif is also present in the same location of the GLP-1 and GLP-2 receptors, in which it is represented by the KLK sequence. Previous studies with the rat GLP-1 receptor demonstrated that the KLK motif, and in particular the first lysine residue, is required for the efficient coupling of the receptor to adenylyl cyclase. Alteration of this motif or the lysine residue in the GLP-1 receptor resulted in significantly reduced cAMP production (59). The KLR motif of the glucagon receptors is found in the same position as the KLK motif of GLP-1 receptors (Fig. 10A). This conservation strongly supports an important role for the KLR sequence, presumably the first lysine residue, for G protein coupling and activation of adenylyl cyclase in all known vertebrate glucagon receptors. These conclusions based on sequence comparisons are consistent with the conclusions obtained from the functional experiments with des-His1-[Nle9-Ala11-Ala16] glucagon antagonist, indicating that similar activation mechanisms regulate the intracellular effects of all vertebrate glucagon receptors.

In addition to the conservation of the KLR motif, a second less well conserved motif (S-E_0–1-R/K) is found in the second intracellular loop between TMD III and TMD IV of the frog, mouse, and rat sequences (Fig. 10A). This motif with a conserved serine residue represents a consensus phosphorylation sequence for protein kinase C and may be important for receptor desensitization by phosphorylation on agonist interaction. There is recent evidence indicating that glucagon receptors in rat hepatocytes can be desensitized by a protein kinase C-dependent mechanism. Treatment of rat hepatocytes with the protein kinase C-selective inhibitors, chelerythrine, staurosporine, and calphostine C inhibited the desensitization process (60). In addition, this S-E_0–1-R/K motif is also present in VIP1 (VPAC₁) receptors isolated from eight different vertebrate species (33) and also several other members in the same receptor family, including human secretin, GHRH, GLP-1, GIP, and PTH receptors and the goldfish GHRH receptor (61), indicating a role for this site in the regulation of receptor desensitization in this receptor family. However, this serine residue is not conserved among all known glucagon receptor sequences in vertebrates; the gfGlucR characterized in this study contains the conservative substitution threonine, whereas the human sequence contains a proline residue (Fig. 3).

Phylogenetic trees obtained for 15 full-length sequences of vertebrate glucagon-like receptors (Fig. 9) or 15 partial sequences, consisting of seven partial sequences of the known GlucRs, five partial sequences of the known GLP-1Rs, and three partial glucagon receptor-like sequences identified in this study (Fig. 10B), established that fish (i.e. goldfish) GlucR and fish (i.e. goldfish, zebrafish) GLP-1Rs originate from a common ancestor in the GlucR lineage. Although at first these phylogenetic relationships may seem surprising, they are consistent with the overlapping functional effects of glucagon and GLP-1 in fish glucose metabolism (23, 24, 25). It is very likely that the ancestral receptor had ligand specificities toward both glucagon and GLP-1 sequences, a feature that is still maintained in the goldfish (43) and zebrafish GLP-1Rs (S. Mojsov, unpublished observations) but lost in the gfGlucR that has ligand specificity only toward glucagon structures. These phylogenetic relationships provide evidence in support of the hypothesis proposed by Sivarajah et al. (50) that the specificity of the ligand-receptor pairs evolved much later than the peptides and receptors, and that multiple peptides and receptors originally had similar functions. In the case of glucagon and GLP-1, the divergence of their function in mammalian glucose metabolism (gluconeogenesis and glycogenolysis in liver vs. incretin effects in pancreas) is mirrored in the specificity of mammalian glucagon/GlucR and mammalian GLP-1/GLP-1R interactions. As illustrated in Figs. 9 and 10B, mammalian GLP-1Rs originated from an ancestral receptor that has diverged from the glucagon receptors. This may also explain the difference in ligand specificities between goldfish and zebrafish GLP-1Rs and mammalian GLP-1Rs. Results shown in Fig. 9 establish that glucagon receptors in vertebrates form a distinct subbranch within the larger gene family of GIP/GLP-1/GLP-2 receptors.

In summary, results of these studies together with our recent work with goldfish and zebrafish GLP-1 receptors (43, 54) demonstrate the following: 1) gfGlucR and goldfish and zebrafish GLP-1Rs originate from a common ancestor in the GlucR lineage; 2) the binding pocket of the gfGlucR and goldfish and zebrafish GLP-1 receptors shares a common structural property of recognizing a more broad spectrum of structures than the binding pockets of their frog (for the glucagon receptor) and mammalian (for glucagon and GLP-1 receptors) counterparts (a frog GLP-1 receptor has yet to be cloned or characterized); and 3) the specificity of the biding pocket of the gfGlucR is restricted to glucagon structures and does not bind to GLP-1 structures. Collectively, these findings clarified an important issue: fish glucagon and GLP-1 act through different receptors, as suggested by Navarro and Moon (27).

If these findings can be extended to all fish glucagon and GLP-1 receptors, these results raise a basic question of why there was a switch in the binding specificities of glucagon and GLP-1 receptors during vertebrate evolution. Did this change take place as a random event caused by mutations or, alternatively, as a result of changes in the regulatory mechanisms that control glucose metabolism in teleost fish and other vertebrates?

	Acknowledgments

 
We are grateful to Dr. Cecille Unson for the gift of the glucagon antagonist and Dr. Thomas Sakmar for the vector containing the synthetic rat glucagon receptor gene. We thank Drs. Bruce Merrifield and Cecille Unson for critical reading of the manuscript; Dr. Ralph Steinman for his continuous support; Yang Wei, Li Yang, Xixuan Du, and Angelika Golebiowska for their help with the binding assays and cAMP experiments; Orlee Guttman for assistance in hepatocyte preparation; and Judy Adams for the illustrations.

	Footnotes

 
This work was supported by Grants RGC HKU7181/99M, CRCG 335/026/0053, and 10203410 (to B.K.C.C.); Grant A6944 from the Natural Sciences and Engineering Research Council of Canada (to T.W.M.); National Institutes of Health Grant AI 11942 (to M.M.); and Grant IBN-9904506 from the National Science Foundation (to S.M.).

Abbreviations: gf, Goldfish; GIP, glucose-dependent insulinotropic polypeptide; GLP, glucagon-like peptide; GlucR, glucagon receptor; h, human; TMD, transmembrane domain; VIP, vasoactive intestinal polypeptide.

Received May 16, 2003.

Accepted for publication March 10, 2004.

	References

Top Abstract Introduction Materials and Methods Results Discussion References

 

1. Mojsov S, Heinrich G, Wilson IB, Ravazzola M, Orci L, Habener JF 1986 Preproglucagon gene expression in pancreas and intestine diversifies at the level of post-translational processing. J Biol Chem 261:11880–11889[Abstract/Free Full Text]
2. Orskov C, Holst JJ, Knuhtsen S, Baldissera FG, Poulsen SS, Nielsen OV 1986 Glucagon-like peptides GLP-1 and GLP-2, predicted products of the glucagon gene, are secreted separately from pig small intestine but not pancreas. Endocrinology 119:1467–1475[Abstract/Free Full Text]
3. Kieffer TJ, Habener JF 1999 The glucagon-like peptides. Endocr Rev 20:876–913[Abstract/Free Full Text]
4. Han VK, Hynes MA, Jin C, Towle AC, Lauder JM, Lund PK 1986 Cellular localization of proglucagon/glucagon-like peptide I messenger RNAs in rat brain. J Neurosci Res 16:97–107[CrossRef][Medline]
5. Drucker DJ, Asa S 1988 Glucagon gene expression in vertebrate brain. J Biol Chem 263:13475–13478[Abstract/Free Full Text]
6. Lee YC, Brubaker PL, Drucker DJ 1990 Developmental and tissue-specific regulation of proglucagon gene expression. Endocrinology 127:2217–2222[Abstract/Free Full Text]
7. Holst JJ 1994 Glucagon-like peptide 1: a newly discovered gastrointestinal hormone. Gastroenterology 107:1848–1855[Medline]
8. Conlon JM, Nielsen PF, Youson JH 1993 Primary structures of glucagon and glucagon-like peptide isolated from the intestine of the parasitic phase lamprey Petromyzon marinus. Gen Comp Endocrinol 91:96–104[CrossRef][Medline]
9. Irwin DM, Huner O, Youson JH 1999 Lamprey proglucagon and the origin of glucagon-like peptides. Mol Biol Evol 16:1548–1557[Abstract]
10. Conlon JM, Dafgard E, Falkmer S, Thim L 1987 A glucagon-like peptide, structurally related to mammalian oxyntomodulin, from the pancreas of a holocephalan fish, Hydrolagus colliei. Biochem J 245:851–855[Medline]
11. Andrews PC, Ronner P 1985 Isolation and structures of glucagon and glucagon-like peptide from catfish pancreas. J Biol Chem 260:3910–3914[Abstract/Free Full Text]
12. Plisetskaya E, Pollock HG, Rouse JB, Hamilton JW, Kimmel JR, Gorbman A 1986 Isolation and structures of coho salmon (Oncorhynchus kisutch) glucagon and glucagon-like peptide. Regul Pept 14:57–67[CrossRef][Medline]
13. Irwin DM, Wong J 1995 Trout and chicken proglucagon: alternative splicing generates mRNA transcripts encoding glucagon-like peptide 2. Mol Endocrinol 9:267–277[Abstract/Free Full Text]
14. Yuen TTH, Mok PY, Chow BKC 1997 Molecular cloning of a cDNA encoding proglucagon from goldfish, Carassius auratus. Fish Physiol Biochem 17:223–230[CrossRef]
15. Irwin DM, Satkunarajah M, Wen Y, Brubaker PL, Pederson RA, Wheeler MB 1997 The Xenopus proglucagon gene encodes novel GLP-1-like peptides with insulinotropic properties. Proc Natl Acad Sci USA 94:7915–7920[Abstract/Free Full Text]
16. Conlon JM, Abdel-Wahab YH, O’Harte FP, Nielsen PF, Whittaker J 1998 Purification and characterization of insulin, glucagon, and two glucagon-like peptides with insulin-releasing activity from the pancreas of the toad, Bufo marinus. Endocrinology 139:3442–3448[Abstract/Free Full Text]
17. Yeung CM, Chow BKC 2001 Identification of a proglucagon cDNA from Rana tigrina regulosa that encodes two GLP-1s and that is alternatively spliced in a tissue-specific manner. Gen Comp Endocrinol 124:144–151[CrossRef][Medline]
18. Conlon JM, Kim JB, Johansson A, Kikuyama S 2002 Comparative peptidomics of the endocrine pancreas:islet hormones from the clawed frog Xenopus laevis and the red-bellied newt Cynops pyrrhogaster. J Endocrinol 75:769–777[CrossRef]
19. Burcelin R, Katz EB, Charron MJ 1996 Molecular and cellular aspects of the glucagon receptor: role in diabetes and metabolism. Diabetes Metab 22:373–396[Medline]
20. Plisetskaya EM, Mommsen TP 1996 Glucagon and glucagon-like peptides in fishes. Int Rev Cytol 168:187–257[Medline]
21. Mojsov S, Weir GC, Habener JF 1987 Insulinotropin: glucagon-like peptide I (7–37) co-encoded in the glucagon gene is a potent stimulator of insulin release in the perfused rat pancreas. J Clin Invest 79:616–619
22. Holst JJ, Orskov C, Nielsen OV, Schwartz TW 1987 Truncated glucagon-like peptide I, an insulin-releasing hormone from the distal gut. FEBS Lett 211:169–174[CrossRef][Medline]
23. Mommsen TP, Andrews PC, Plisetskaya EM 1987 Glucagon-like peptides activate hepatic gluconeogenesis. FEBS Lett 219:227–232[CrossRef][Medline]
24. Mommsen TP, Moon T 1990 Metabolic responses in teleost hepatocytes to glucagon-like peptide and glucagon. J Endocrinol 126:109–118[Abstract/Free Full Text]
25. Mommsen TP, Plisetskaya EM 1993 Metabolic and endocrine functions of glucagon-like peptides: evolutionary and biochemical perspectives. Fish Physiol Biochem 11:1–6
26. Plisetskaya E, Duguay S 1993 Pancreatic hormones and metabolism in ectotherm vertebrates: current views. In: Scheibman M, Scanes C, Pang P, eds. The endocrinology of growth, development, and metabolism in vertebrates. New York: Academic Press; 265–287
27. Navarro I, Moon T 1994 Glucagon binding to hepatocytes from two teleost fishes: the American eel and brown bullhead. J Endocrinology 140:217–227[Abstract/Free Full Text]
28. Ngan ESW, Chow LSN, Tse DLY, Du X, Wei Y, Mojsov S, Chow BKC 1999 Functional studies of a glucagon receptor isolated from frog Rana tigrina rugulosa: implications on the molecular evolution of glucagon receptors in vertebrates. FEBS Lett 457:499–504[CrossRef][Medline]
29. Unson CG, Merrifield RB 1994 Identification of an essential serine residue in glucagon: implication for an active site triad. Proc Natl Acad Sci USA 91:454–458[Abstract/Free Full Text]
30. Unson CG, Wu CR, Merrifield RB 1994 Roles of aspartic acid 15 and 21 in glucagon action: receptor anchor and surrogates for aspartic acid 9. Biochemistry 33:6884–6887[CrossRef][Medline]
31. Mommsen TP, Moon T, Walsh P 1994 Hepatocytes: isolation, maintenance and utilization. In: Hochachka PW, Mommsen TP, eds. Biochemistry and molecular biology of fishes: analytical techniques. Vol 3. Amsterdam: Elsevier Science B.V.; 355–373
32. Moon T, Busby ER, Cooper GA, Mommsen TP 1999 Fish hepatocyte glycogen phoshporylase—as sensitive indicator for hormone modulation. Fish Physiol Biochem 21:15–24
33. Chow BK, Yuen TT, Chan KW 1997 Molecular evolution of vertebrate VIP receptors and functional characterization of a VIP receptor from goldfish Carassius auratus. Gen Comp Endocrinol 105:176–185[CrossRef][Medline]
34. Carruthers CJ, Unson CG, Kim HN, Sakmar TP 1994 Synthesis and expression of a gene for the rat glucagon receptor. Replacement of an aspartic acid in the extracellular domain prevents glucagon binding. J Biol Chem 269:29321–29328[Abstract/Free Full Text]
35. Unson CG, Wu CR, Fitzpatrick KJ, Merrifield RB 1994 Multiple-site replacement analogs of glucagon. A molecular basis for antagonist design. J Biol Chem 269:12548–12551[Abstract/Free Full Text]
36. Siebert PD, Chenchik A 1993 Modified acid guanidinium thiocynate-phenol-chloroform RNA extraction method which greatly reduces DNA contamination. Nucleic Acids Res 21:2019–2020[Free Full Text]
37. Thompson JD, Higgins DG, Gibson TJ 1994 CLUSTAL W: improving the sensitivity of progressive multiple sequence alignment through sequence weighting, position-specific gap penalties and weight matrix choice. Nucleic Acids Res 22:4673–4680[Abstract/Free Full Text]
38. Felsenstein J 1993 PHYLIP, Phylogenetic Inference Package, version 3.572, software and documentation. Seattle: University of Washington
39. Motulski H, Christopoulos A 2003 Fitting models to biological data using linear and nonlinear regression. A practical guide to curve fitting. San Diego: GraphPad Software Inc.
40. Miller JR 2003 GraphPad Prism, version 4.0, Step-by-Step Examples, San Diego: Graph Pad Software, Inc.
41. OriginLab Corp. 2003 Scientific graphing and analysis software, version 7.5. Northampton, MA: OriginLab Corp.
42. Stata Corp. 2003 Stata statistical software, release 8.0. College Station, TX: Stata Corp.
43. Yeung CM, Mojsov S, Mok PY, Chow BK 2002 Isolation and structure-function studies of a glucagon-like peptide 1 receptor from goldfish Carassius auratus: identification of three charged residues in extracellular domains critical for receptor function. Endocrinology 143:4646–4654[Abstract/Free Full Text]
44. Mommsen TP, Mojsov S 1998 Glucagon-like peptide-1 activates adenylyl cyclase system in rockfish enterocytes and brain membranes. Comp Biochem Physiol 121:49–56[CrossRef]
45. Pang RT, Ng SS, Cheng CH, Holtmann MH, Miller LJ, Chow BK 1999 Role of N-linked glycosylation on the function and expression of the human secretin receptor. Endocrinology 140:5102–5111[Abstract/Free Full Text]
46. Svoboda M, Tastenoy M, Vertongen P, Robberecht P 1994 Relative quantitative analysis of glucagon receptor mRNA in rat tissues. Mol Cell Endocrinol 105:131–137[CrossRef][Medline]
47. Hansen LH, Abrahamsen N, Nishimura E 1995 Glucagon receptor mRNA distribution in rat tissues. Peptides 16:1163–1166[CrossRef][Medline]
48. Mommsen TP 2000 Glucagon-like peptide 1 in fishes: the liver and beyond. Am Zool 40:259–268[CrossRef]
49. Soengas JL, Moon TW 1998 Transport and metabolism of glucose in isolated enterocytes of the black bullhead Ictalurus melas: effects of diet and hormones. J Exp Biol 201:3263–3273[Abstract]
50. Sivarajah P, Wheeler MB, Irwin DM 2001 Evolution of receptors for proglucagon-derived peptides: isolation of frog glucagon receptors. Comp Biochem Physiol B Biochem Mol Biol 128:517–527[CrossRef][Medline]
51. Jelinek LJ, Lok S, Rosenberg GB, Smith RA, Grant FJ, Biggs S, Bensch PA, Kuijper JL, Sheppard PO, Sprecher CA, O’Hara PJ, Foster D, Walker KM, Chen LHJ, McKernan PA, Kindsvogel W 1993 Expression cloning and signaling properties of the rat glucagon receptor. Science 259:1614–1616[Abstract/Free Full Text]
52. Unson CG, Andreu D, Gurzenda EM, Merrifield RB 1987 Synthetic peptide antagonists of glucagon. Proc Natl Acad Sci USA 84:4083–4087[Abstract/Free Full Text]
53. Unson CG, Macdonald D, Ray K, Durrah TL, Merrifield RB 1991 Position 9 replacement analogs of glucagon uncouple biological activity and receptor binding. J Biol Chem 266:2763–2766[Abstract/Free Full Text]
54. Mojsov S 2000 Glucagon-like peptide-1(GLP-1) and the control of glucose metabolism in mammals and teleost fish. Am Zool 40:246–258[CrossRef]
55. Buggy JJ, Livingston JN, Rabin DU, Yoo-Warren H 1995 Glucagon-like peptide I receptor chimeras reveal domains that determine specificity of glucagon binding. J Biol Chem 270:7474–7478[Abstract/Free Full Text]
56. Unson CG 2002 Molecular determinants of glucagon receptor signaling. Biopolymers 66:218–235[CrossRef][Medline]
57. Cypess AM, Unson CG, Wu CR, Sakmar TP 1999 Two cytoplasmic loops of the glucagon receptor are required to elevate cAMP or intracellular calcium. J Biol Chem 274:19455–19464[Abstract/Free Full Text]
58. Chicchi GG, Graziano MP, Koch G, Hey P, Sullivan K, Vicario PP, Cascieri MA 1997 Alterations in receptor activation and divalent cation activation of agonist binding by deletion of intracellular domains of the glucagon receptor. J Biol Chem 272:7765–7769[Abstract/Free Full Text]
59. Takhar S, Gyomorey S, Su RC, Mathi SK, Li X, Wheeler MB 1996 The third cytoplasmic domain of the GLP-1[7–36 amide] receptor is required for coupling to the adenylyl cyclase system. Endocrinology 137:2175–2178[Abstract]
60. Savage A, Zeng L, Houslay MD 1995 A role for protein kinase C-mediated phosphorylation in eliciting glucagon desensitization in rat hepatocytes. Biochem J 307:281–285
61. Chan KW, Yu KL, Rivier J, Chow BK 1998 Identification and characterization of a receptor from goldfish specific for a teleost growth hormone-releasing hormone-like peptide. Neuroendocrinology 68:44–56[CrossRef][Medline]
62. Burcelin R, Li J, Charron MJ 1995 Cloning and sequence analysis of the murine glucagon receptor-encoding gene. Gene 164:305–310[CrossRef][Medline]
63. Lok S, Kuijper JL, Jelinek LJ, Kramer JM, Whitmore TE, Sprecher CA, Mathewes S, Grant FJ, Biggs SH, Rosenberg GB, Sheppard PO, O’Hara PJ, Foster DC, Kindsvogel W 1994 The human glucagon receptor encoding gene: structure, cDNA sequence and chromosomal localization. Gene 140:203–209[CrossRef][Medline]
64. Thorens B 1992 Expression cloning of the pancreatic ß cell receptor for the gluco-incretin hormone glucagon-like peptide 1. Proc Natl Acad Sci USA 89:8641–8645[Abstract/Free Full Text]
65. Thorens B, Porret A, Buhler L, Deng SP, Morel P, Widmann C 1993 Cloning and functional expression of the human islet GLP-1 receptor. Demonstration that exendin-4 is an agonist and exendin-(9–39) an antagonist of the receptor. Diabetes 42:1678–1682[Abstract]
66. Volz A, Goke R, Lankat-Buttgereit B, Fehmann HC, Bode HP, Goke B 1995 Molecular cloning, functional expression, and signal transduction of the GIP-receptor cloned from a human insulinoma. FEBS Lett 373:23–29[CrossRef][Medline]
67. Usdin TB, Mezey E, Button DC, Brownstein MJ, Bonner TI 1993 Gastric inhibitory polypeptide receptor, a member of the secretin-vasoactive intestinal peptide receptor family, is widely distributed in peripheral organs and the brain. Endocrinology 133:2861–2870[Abstract/Free Full Text]
68. Munroe DG, Gupta AK, Kooshesh F, Vyas TB, Rizkalla G, Wang H, Demchyshyn L, Yang ZJ, Kamboj RK, Chen H, McCallum K, Sumner-Smith M, Drucker DJ, Crivici A 1999 Prototypic G protein-coupled receptor for the intestinotrophic factor glucagon-like peptide 2. Proc Natl Acad Sci USA 96:1569–1573[Abstract/Free Full Text]
69. Ishihara T, Shigemoto R, Mori K, Takahashi K, Nagata S 1992 Functional expression and tissue distribution of a novel receptor for vasoactive intestinal polypeptide. Neuron 8:811–819[CrossRef][Medline]
70. Sreedharan SP, Robichon A, Peterson KE, Goetzl EJ 1991 Cloning and expression of the human vasoactive intestinal peptide receptor. Proc Natl Acad Sci USA 88:4986–4990[Abstract/Free Full Text]
71. Flamez D, Van Breusegem A, Scrocchi LA, Quartier E, Pipleers D, Drucker DJ, Schuit F 1998 Mouse pancreatic ß-cells exhibit preserved glucose competence after disruption of the glucagon-like peptide-1 receptor gene. Diabetes 47:646–652[Abstract]

This article has been cited by other articles:

				J. Wang, Y. Wang, X. Li, J. Li, and F. C. Leung Cloning, Tissue Distribution, and Functional Characterization of Chicken Glucagon Receptor Poult. Sci., December 1, 2008; 87(12): 2678 - 2688. [Abstract] [Full Text] [PDF]

				D. M. Irwin and K. Wong Evolution of New Hormone Function: Loss and Gain of a Receptor J. Hered., May 1, 2005; 96(3): 205 - 211. [Abstract] [Full Text] [PDF]

This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Request Copyright Permission Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Chow, B. K. C. Articles by Mojsov, S. Search for Related Content PubMed PubMed Citation Articles by Chow, B. K. C. Articles by Mojsov, S. Pubmed/NCBI databases Gene GEO Profiles HomoloGene Protein UniGene Compound via MeSH Substance via MeSH