Mutagenesis of the Ligand Binding Domain of the Human Retinoic Acid Receptor α Identifies Critical Residues for 9-cis-Retinoic Acid Binding
1995; Elsevier BV; Volume: 270; Issue: 35 Linguagem: Inglês
10.1074/jbc.270.35.20258
ISSN1083-351X
AutoresBonnie F. Tate, Joseph F. Grippo,
Tópico(s)Receptor Mechanisms and Signaling
ResumoWe have recently identified a small region (amino acids 405-419) within the ligand binding domain of a truncated human retinoic acid receptor α (Δ419) that is required for binding of 9-cis-retinoic acid (RA), but not all-trans-retinoic acid (t-RA). To probe the structural determinants of this high affinity 9-cis-RA binding site, a series of Δ419 mutants were prepared whereby an individual alanine residue was substituted for each amino acid within this region. These modified receptors were expressed in mammalian COS-1 cells and assayed for their ability to bind 9-cis-RA as well as t-RA. Only two of the mutants, M406A (mutation of methionine 406 to alanine), and I410A (mutation of isoleucine 410 to alanine) exhibit no detectable binding of 9-cis-RA when analyzed using saturation binding kinetics. Substitution of methionine 406 with the amino acids leucine, isoleucine, and valine yields mutant receptors that exhibit decreased binding for 9-cis-RA as the length or hydrophobicity of the R group decreases. Further substitution of methionine 406 with the small polar amino acid, threonine, results in a loss of detectable 9-cis-RA binding. Since amino acids 405-419 on a human RARα (hRARα) are predicted to form a short amphipathic α-helix, modeling of this structure into a helical wheel indicates that these two amino acids, methionine 406 and isoleucine 410, are actually positioned proximal to each other. Data presented here suggest that high affinity 9-cis-RA binding to a hRARα depends on an interaction with the two amino acids methionine 406 and isoleucine 410. We have recently identified a small region (amino acids 405-419) within the ligand binding domain of a truncated human retinoic acid receptor α (Δ419) that is required for binding of 9-cis-retinoic acid (RA), but not all-trans-retinoic acid (t-RA). To probe the structural determinants of this high affinity 9-cis-RA binding site, a series of Δ419 mutants were prepared whereby an individual alanine residue was substituted for each amino acid within this region. These modified receptors were expressed in mammalian COS-1 cells and assayed for their ability to bind 9-cis-RA as well as t-RA. Only two of the mutants, M406A (mutation of methionine 406 to alanine), and I410A (mutation of isoleucine 410 to alanine) exhibit no detectable binding of 9-cis-RA when analyzed using saturation binding kinetics. Substitution of methionine 406 with the amino acids leucine, isoleucine, and valine yields mutant receptors that exhibit decreased binding for 9-cis-RA as the length or hydrophobicity of the R group decreases. Further substitution of methionine 406 with the small polar amino acid, threonine, results in a loss of detectable 9-cis-RA binding. Since amino acids 405-419 on a human RARα (hRARα) are predicted to form a short amphipathic α-helix, modeling of this structure into a helical wheel indicates that these two amino acids, methionine 406 and isoleucine 410, are actually positioned proximal to each other. Data presented here suggest that high affinity 9-cis-RA binding to a hRARα depends on an interaction with the two amino acids methionine 406 and isoleucine 410. INTRODUCTIONThe retinoic acid receptors (RARs) 1The abbreviations used are: RARretinoic acid receptorRAretinoic acidt-RAall-trans-retinoic acidRXRretinoid X receptorhRARαhuman RARα. (1Petkovich M. Brand N.J. Krust A. Chambon P. Nature. 1987; 330: 444-450Crossref PubMed Scopus (1726) Google Scholar, 2Giguere V. Ong E.S. Segul P. Evans R.M. Nature. 1987; 330: 624-629Crossref PubMed Scopus (1527) Google Scholar, 3Brand N. Petkovich M. Krust A. Chambon P. de Thé H. Marchio A. Tiollais P. Dejean A. Nature. 1988; 332: 850-853Crossref PubMed Scopus (809) Google Scholar, 4Krust A. Kastner P. Petkovich M. Zelent A. Chambon P. Proc. Natl. Acad. Sci. U. S. A. 1989; 86: 5310-5314Crossref PubMed Scopus (621) Google Scholar) and the retinoid X receptors (RXRs, 5-8) have a prominent role in cellular differentiation and vertebrate development through their interaction with retinoic acids (RA), which are natural derivatives of vitamin A (retinol, (9Morriss-Kay G. Retinoids in Normal Development and Teratogenesis. Oxford University Press, New York1992Google Scholar) and (10Sporn M.B. Roberts A.B. Goodman DeW.S. The Retinoids: Biology, Chemistry, and Medicine. Raven Press, New York1994Google Scholar)). The RARs and the RXRs are members of a superfamily of ligand-dependent transcriptional regulators which also include the steroid hormone, thyroid hormone, and vitamin D receptors(11Evans R.M. Science. 1988; 240: 889-895Crossref PubMed Scopus (6292) Google Scholar, 12Green S. Chambon P. Trends Genet. 1988; 4: 309-314Abstract Full Text PDF PubMed Scopus (830) Google Scholar). Based on amino acid homologies and functional similarities, six regions of identity (A-F) have been ascribed to members of this superfamily, including the RARs and RXRs, through which modulation of transcription can occur(11Evans R.M. Science. 1988; 240: 889-895Crossref PubMed Scopus (6292) Google Scholar, 12Green S. Chambon P. Trends Genet. 1988; 4: 309-314Abstract Full Text PDF PubMed Scopus (830) Google Scholar). Regions A and B contain a ligand-independent transcription function (AF-1) while region C contains a highly conserved zinc finger DNA-binding domain. The functions of regions D and F are still relatively unknown, however, region E is known to contain a multiplicity of functions, namely, those of heterodimerization, ligand-dependent transactivation (AF-2) and ligand binding(13Leid M. Kastner P. Chambon P. Trends Biochem. Sci. 1992; 17: 427-433Abstract Full Text PDF PubMed Scopus (803) Google Scholar).Both the RARs and RXRs bind endogenous stereoisomers of RA within the ligand binding region and undergo a conformational change prior to their activation as transcriptional regulators(14Allan G.F. Leng X. Tsai S.Y. Weigel N.L. Edwards D.P. Tsai M.-J. O'Malley B.W. J. Biol. Chem. 1992; 267: 19513-19520Abstract Full Text PDF PubMed Google Scholar, 15Leng X. Tsai S.Y. O'Malley B.W. Tsai M.-J. J. Steroid Biochem. Mol. Biol. 1993; 46: 643-661Crossref PubMed Scopus (80) Google Scholar, 16Keidel S. LeMotte P. Apfel C. Mol. Cell. Biol. 1994; 14: 287-298Crossref PubMed Scopus (113) Google Scholar, 17Tate B.F. Allenby G. Janocha R. Kazmer S. Speck J. Sturzenbecker L.J. Abarzúa P. Levin A.A. Grippo J.F. Mol. Cell. Biol. 1994; 14: 2323-2330Crossref PubMed Google Scholar). The RXRs show a binding preference for 9-cis-RA (18Allenby G. Bocquel M.-T. Saunders M. Kazmer S. Speck J. Rosenberger M. Lovey A. Kastner P. Grippo J.F. Chambon P. Levin A.A. Proc. Natl. Acad. Sci. U. S. A. 1993; 90: 30-34Crossref PubMed Scopus (658) Google Scholar, 19Heyman R.A. Mangelsdorf D.J. Dyck J.A. Stein R.B. Eichele G. Evans R.M. Thaller C. Cell. 1992; 68: 397-406Abstract Full Text PDF PubMed Scopus (1556) Google Scholar, 20Levin A.A. Sturzenbecker L.J. Kazmer S. Bosakowski T. Huselton C. Allenby G. Speck J. Kratzeisen Cl. Rosenberger M. Lovey A. Grippo J.F. Nature. 1992; 355: 359-361Crossref PubMed Scopus (1095) Google Scholar, 21Tate B.F. Levin A.A. Grippo J.F. Trends Endocrinol. Metab. 1994; 5: 189-194Abstract Full Text PDF PubMed Scopus (19) Google Scholar) while the RARs bind 9-cis-RA as well as the stereoisomer, all-trans-RA (t-RA), with equal affinity(18Allenby G. Bocquel M.-T. Saunders M. Kazmer S. Speck J. Rosenberger M. Lovey A. Kastner P. Grippo J.F. Chambon P. Levin A.A. Proc. Natl. Acad. Sci. U. S. A. 1993; 90: 30-34Crossref PubMed Scopus (658) Google Scholar). An analysis of ligand binding to mouse RARs using saturation kinetics and Scatchard analyses gave dissociation constants (Kd values) in the range of 0.2-0.7 nM for both 9-cis-RA and t-RA(22Allenby G. Janocha R. Kazmer S. Speck J. Grippo J.F. Levin A.A. J. Biol. Chem. 1994; 269: 16689-16695Abstract Full Text PDF PubMed Google Scholar). Competition of [3H]t-RA or [3H]9-cis-RA by 9-cis-RA and t-RA indicates that the two ligands can compete with one another for binding to the RARs(22Allenby G. Janocha R. Kazmer S. Speck J. Grippo J.F. Levin A.A. J. Biol. Chem. 1994; 269: 16689-16695Abstract Full Text PDF PubMed Google Scholar). These binding analyses led to the seemingly obvious conclusion that a common binding pocket on an RAR exists for both t-RA and its configurational isomer, 9-cis-RA.Recently, however, we found that there are actual differences in the binding pocket for these two ligands and that a distinct region is required for the binding of 9-cis-RA, but not t-RA, to a hRARα(17Tate B.F. Allenby G. Janocha R. Kazmer S. Speck J. Sturzenbecker L.J. Abarzúa P. Levin A.A. Grippo J.F. Mol. Cell. Biol. 1994; 14: 2323-2330Crossref PubMed Google Scholar). Truncation of 43 amino acids to position 419 (removal of the F domain) on a hRARα produced a mutant which had a binding affinity for both t-RA and 9-cis-RA with normal Kd's of 0.2-0.3 nM for both ligands. Truncation of 58 amino acids to position 404, however, produced a mutant which had a normal binding affinity for t-RA (Kd = 0.3 nM), but which demonstrated no detectable binding of 9-cis-RA. We concluded that the binding sites for t-RA and 9-cis-RA on a hRARα were overlapping but distinct and that the region from amino acids 405 to 419 defined a high affinity binding determinant for 9-cis-RA. Interestingly, this region also corresponded with the highly conserved AF-2 domain found in the majority of receptors of the steroid hormone, thyroid hormone, and vitamin D superfamily(23Danielian P.S. White R. Lees J.A. Parker M.G. EMBO J. 1992; 11: 1025-1033Crossref PubMed Scopus (716) Google Scholar). Therefore, this same mutant which lost the ability to bind 9-cis-RA (Δ404) also failed to initiate transcription on a retinoic acid-responsive element after binding t-RA. Addition of these 15 amino acids (405-419) to a Δ404 resulted in a recovery of both 9-cis-RA binding and transactivation ability and yielded a fully functional receptor (Δ419, (17Tate B.F. Allenby G. Janocha R. Kazmer S. Speck J. Sturzenbecker L.J. Abarzúa P. Levin A.A. Grippo J.F. Mol. Cell. Biol. 1994; 14: 2323-2330Crossref PubMed Google Scholar)).Since we had isolated a high affinity 9-cis-RA binding determinant to within 15 amino acids, we were then able to analyze those motif(s) responsible for this high affinity binding. Therefore, we have further characterized the region from amino acids 405 to 419 using alanine scanning mutagenesis on a truncated hRARα and have found that the integrity of two amino acids, methionine 406 and isoleucine 410, appear critical for high affinity 9-cis-RA binding. Based on a Chou and Fasman (24Chou P.Y. Fasman G.D. Adv. Enzymol. 1978; 47: 45-147PubMed Google Scholar) algorithm which predicts that this region forms a short α-helical structure, these two amino acids actually exist in close proximity to each other. From our data we propose a model of 9-cis-RA binding to a truncated hRARα that may involve a contact with these two amino acids, methionine 406 and isoleucine 410.EXPERIMENTAL PROCEDURESMaterialsAll-trans-[11,12-3H]retinoic acid was purchased from DuPont NEN. All-trans retinoic acid, 9-cis-retinoic acid, 9-cis-19-pentyl retinoic acid ((2E,4E,6Z,8E)-3-methyl-7-hexyl-9-(2,6,6-trimethyl-1-cyclohexen-1yl)-2,4,6,8-nonatetraenoic acid), 9-cis-20-butyl-retinoic acid ((2E,4E,6Z,8E)-3-penyl-7-methyl-9-(2,6,6-trimethyl-1-cyclohexen-1-yl)-2,4,6,8-nonatetraenoic acid), and [10-3H]9-cis-retinoic acid were synthesized and purified by the Department of Medicinal Chemistry, Hoffmann-La Roche.Construction of Receptor MutantsMutant RARα receptors were generated by modifying the full-length human RARα cDNA that was cloned into the EcoRI site of the expression vector pSG5 (25Green S. Isseman I. Scheer E. Nucleic Acids Res. 1988; 16: 369Crossref PubMed Scopus (544) Google Scholar). The plasmid pSG5-hRARα was linearized with BamHI and digested with SmaI to remove sequences at the 3′ end of hRARα. Amino acid mutations were incorporated into the design of synthetic single-stranded oligonucleotides. The synthetic single-stranded oligonucleotides were phosphorylated at 37°C for 60 min in the presence of T4 polynucleotide kinase (Life Technologies, Inc.), 70 mM Tris (pH 7.6), 100 mM KCl, 10 mM MgCl2, 5 mM dithiothreitol, and 0.5 mM ATP and then annealed with their corresponding opposite strands by heating to 100°C for 3 min and slowly cooling to room temperature. Each mutant receptor was then generated by insertion of the synthetic double-stranded oligonucleotides encoding the amino acids of interest and a stop codon into the SmaI-BamHI site. The integrity of the mutants and the presence of in-frame TGA codons was confirmed by sequencing.Cell Culture and TransfectionsCOS-1 cells (African green monkey kidney cells derived from SV40-transformed CV-1 cells; ATCC) were maintained in Dulbecco's minimal essential medium (Life Technologies, Inc.) containing 100 units/ml penicillin, 100 μg/ml streptomycin, and 10% heat-inactivated fetal calf serum. COS-1 cells were transiently transfected by electroporation as described previously (20Levin A.A. Sturzenbecker L.J. Kazmer S. Bosakowski T. Huselton C. Allenby G. Speck J. Kratzeisen Cl. Rosenberger M. Lovey A. Grippo J.F. Nature. 1992; 355: 359-361Crossref PubMed Scopus (1095) Google Scholar). Typically, 5 × 105 cells were electroporated with a total of 25 μg of receptor expression vector. The cells were then seeded into 15-cm plates and, after 72 h at 37°C, harvested for the preparation of nucleosol.Nucleosol PreparationsCOS-1 cells transfected with the expression vectors containing cDNAs encoding wild-type or mutant receptors were harvested and nucleosol fractions were prepared essentially as described previously for the use in retinoic acid binding assays(18Allenby G. Bocquel M.-T. Saunders M. Kazmer S. Speck J. Rosenberger M. Lovey A. Kastner P. Grippo J.F. Chambon P. Levin A.A. Proc. Natl. Acad. Sci. U. S. A. 1993; 90: 30-34Crossref PubMed Scopus (658) Google Scholar, 20Levin A.A. Sturzenbecker L.J. Kazmer S. Bosakowski T. Huselton C. Allenby G. Speck J. Kratzeisen Cl. Rosenberger M. Lovey A. Grippo J.F. Nature. 1992; 355: 359-361Crossref PubMed Scopus (1095) Google Scholar, 26Nervi C. Grippo J.F. Sherman M.I. George M.D. Jetten A.M. Proc. Natl. Acad. Sci. U. S. A. 1989; 86: 5854-5858Crossref PubMed Scopus (108) Google Scholar). Briefly, nuclei from transfected cells (4 plates) were disrupted in 4.0 ml of lysis buffer containing 10 mM Tris (pH 8.0), 1.5 mM EDTA, 2 mM dithiothreitol, 10% glycerol, 1 mM phenylmethylsulfonyl fluoride, 10 μg/ml leupeptin, 10 μg/ml aprotinin, and 800 mM KCl. Aliquots were stored at −80°C until further use.Retinoic Acid Binding AssaysBinding assays with [3H]t-RA and [3H]9-cis-RA were performed as detailed previously(20Levin A.A. Sturzenbecker L.J. Kazmer S. Bosakowski T. Huselton C. Allenby G. Speck J. Kratzeisen Cl. Rosenberger M. Lovey A. Grippo J.F. Nature. 1992; 355: 359-361Crossref PubMed Scopus (1095) Google Scholar, 22Allenby G. Janocha R. Kazmer S. Speck J. Grippo J.F. Levin A.A. J. Biol. Chem. 1994; 269: 16689-16695Abstract Full Text PDF PubMed Google Scholar). Aliquots of nucleosol (15-25 μl) were incubated at 4°C for 4 h after the addition of ligand contained in ethanolic solutions that did not exceed 2% of the total assay volume. Receptor bound radioactivity was then separated from free by chromatography over a desalting PD-10 column (Pharmacia LKB Biotechnology Inc.), radioactive eluate was collected and tritium counts were determined as described previously(22Allenby G. Janocha R. Kazmer S. Speck J. Grippo J.F. Levin A.A. J. Biol. Chem. 1994; 269: 16689-16695Abstract Full Text PDF PubMed Google Scholar). Binding of tritiated ligand in the presence of a 100-fold molar excess of unlabeled ligand was defined as nonspecific binding. Specific binding was defined as the total binding minus the nonspecific binding. For total binding assays, the concentration of [3H]t-RA or [3H]9-cis-RA was 5 nM, and %t-RA binding activity was defined as femtomoles (fmol) of 9-cis-RA binding/fmol of t-RA binding × 100, where the amount of nucleosol used in the assays was normalized to yield between 300 and 500 fmol of t-RA binding for each expressed mutant receptor. For saturation kinetics and Scatchard analyses, [3H]t-RA or [3H]9-cis-RA was added at increasing concentrations from 4 × 10-10 to 5 × 10-8M. Affinity of wild-type or mutant receptors for t-RA and 9-cis-RA was determined by the method of Scatchard(27Scatchard G. Ann. N. Y. Acad. Sci. 1949; 51: 660-672Crossref Scopus (17758) Google Scholar). For competitive binding assays, incubations were performed with increasing concentrations of unlabeled competing ligand and a fixed nominal concentration of radioligand (5 nM [3H]t-RA).Analysis of DataData shown in Fig. 3-5 are representative of at least two independent experiments performed in duplicate. The data in Table 1and Table 2are the means of two to four independent experiments performed in duplicate.Tabled 1 Open table in a new tab Tabled 1 Open table in a new tab RESULTSBinding of [3H]9-cis-RA to Mutant ReceptorsTo investigate the contribution of individual amino acids to 9-cis-RA binding within the region from serine 405 to glycine 419 on the truncated receptor, Δ419, each residue (except methionine 413 and leucine 414 which were deleted) was replaced with alanine as depicted in Fig. 1. Substitution of alanine allows an analysis of the importance of the amino acid R groups for 9-cis-RA binding.Figure 1:Schematic of the truncated, mutant human retinoic acid receptor α sequences. A schematic representation of the hRARα receptor truncated to amino acid position 419 (Δ419) is indicated at the top. The dark region corresponds to the ligand binding domain and the black lines indicate the region containing amino acids that have been replaced by alanine (from amino acid 405 to 419). The amino acid sequences are shown below. Deletions are symbolized by Δ and the numbers refer to the positions in the amino acid sequence. The hRARα mutants were made as described under “Experimental Procedures” by ligation of synthetic, double-stranded oligonucleotides.View Large Image Figure ViewerDownload Hi-res image Download (PPT)For each mutant receptor depicted in Fig. 1, nucleosol was isolated from mammalian COS-1 cells transiently transfected with the corresponding receptor expression vector. [3H]t-RA or [3H]9-cis-RA was then incubated with these nucleosol fractions for 4 h at 4°C since little isomerization has been shown to occur under these conditions(22Allenby G. Janocha R. Kazmer S. Speck J. Grippo J.F. Levin A.A. J. Biol. Chem. 1994; 269: 16689-16695Abstract Full Text PDF PubMed Google Scholar). The results are shown in Fig. 2and 9-cis-RA binding is calculated as %t-RA binding activity, where total binding of t-RA has been normalized to yield between 300 and 500 fmol of transfected receptor binding sites for each assay (from 15 to 25 μl of nucleosol).Figure 2:Percent t-RA binding activities of hRARαΔ419 alanine mutants. The mutant receptors were transiently expressed in COS-1 cells and nucleosol was prepared as described under “Experimental Procedures.” Nucleosol was incubated with [3H]t-RA or [3H]9-cis-RA and total binding was calculated in the presence of 100-fold molar excess cold ligand. %t-RA binding was then reported as the total binding with [3H]9-cis-RA/total binding with [3H]t-RA (normalized to 300-500 fmol) × 100. The truncation mutants Δ404 and Δ419 are also indicated.View Large Image Figure ViewerDownload Hi-res image Download (PPT)The alanine substitutions that most strikingly disrupt 9-cis-RA binding are those at methionine 406, isoleucine 410, and proline 407. These substitutions result in only 5.6, 4.0, and 14% of 9-cis-RA binding activity, respectively, relative to that of t-RA. The parent receptor, Δ419, exhibits 89% 9-cis-RA/t-RA binding activity (Fig. 2). Alanine substitution of other amino acids proximal to methionine 406 also reduce 9-cis-RA binding but not as severely. For example, P408A and L409A bind 9-cis-RA at 43 and 53% of t-RA binding capacity, respectively. Interestingly, even though serine 405 is adjacent to methionine 406, S405A has a normal capacity for 9-cis-RA. This may be due to the fact that serine 405 lies N-terminal to the start of the putative amphipathic α-helix.Using saturation kinetics and Scatchard analyses (27Scatchard G. Ann. N. Y. Acad. Sci. 1949; 51: 660-672Crossref Scopus (17758) Google Scholar) we also determined the dissociation constants (Kd values) of select mutant receptors for binding 9-cis-RA and t-RA. The results are summarized in Table 1. Mutant receptors with changes in the amino acids of interest from Fig. 2as well as mutants with changes in each part of the α-helical structure were analyzed. The Kd values for 9-cis-RA binding to the mutant receptors emulate those results seen in Fig. 2where substitutions between amino acids 406 and 410, the amino-terminal portion of the α-helix, most affect 9-cis-RA binding (Table 1). Both M406A and I410A exhibit no detectable binding (NB), while P407A, P408A, and L409A exhibit Kd values of 14, 7.9, and 2.7 nM, respectively (Table 1). Receptors with mutations in the carboxyl-terminal portion of this region exhibit Kd values more in the normal range, 0.9 nM for both the deletion mutant ML413,414Δ and the double mutant E415,418A.Binding of [3H]t-RA to Mutant ReceptorsMutations within the amino-terminal portion of the α-helix that affect 9-cis-RA binding also affect t-RA binding. Binding affinity decreases to 1.8 nM for P408A and to 4.2 nM for M406A in comparison to both Δ419 (0.3 nM) and Δ404 (0.3 nM, Table 1). The fact that these perturbations of t-RA binding are confined to the first half of this α-helical structure demonstrates the critical nature of this region for ligand binding.Representative saturation curves and Scatchard plots shown in Fig. 3lend further support to these interpretations. Both Δ404 (Fig. 3A) and M406A (Fig. 3B) exhibit saturable binding for [3H]t-RA yielding a Kd of 0.3 nM for Δ404, similar to wild-type hRARα(18Allenby G. Bocquel M.-T. Saunders M. Kazmer S. Speck J. Rosenberger M. Lovey A. Kastner P. Grippo J.F. Chambon P. Levin A.A. Proc. Natl. Acad. Sci. U. S. A. 1993; 90: 30-34Crossref PubMed Scopus (658) Google Scholar), and a Kd of 4.2 nM for M406A (Fig. 3C). Scatchard analysis of E415,418A, however, shows a Kd of 0.9 for [3H]9-cis-RA and a similar Kd of 0.8 for [3H]t-RA (Fig. 3, D-F).Amino Acid Substitutions at Methionine 406To further probe the interaction of 9-cis-RA with this region on Δ419, additional mutants were made in which methionine 406 was changed to leucine, isoleucine, valine, or threonine. The results are shown in Fig. 4, where the R groups are also depicted. When normalized to %t-RA binding with methionine at position 406 (represented by Δ419), substitution of leucine, isoleucine, and valine yields 9-cis-RA binding values of 89, 72, and 54%, respectively (Fig. 4). Substitution of methionine with either alanine or threonine yields much lower values of 6 and 3%, respectively. This data suggests that the length, and perhaps the hydrophobicity, of the R group at position 406 is important for high affinity binding of 9-cis-RA.Figure 4:Percent t-RA binding capacities and R group structures of different amino acids at position 406. Percent t-RA activity was calculated for each mutant receptor containing the indicated amino acid at position 406 as described for Fig. 2. The amino acid R group is indicated to the right of each bar, and the percent t-RA binding activity for a wild-type methionine (Δ419) at position 406 has been normalized to 100% as described under “Experimental Procedures.”View Large Image Figure ViewerDownload Hi-res image Download (PPT)Alkyl Substitutions on 9-cis-RASince we lost 9-cis-RA binding by replacing the longer side chain of methionine at position 406 by a methyl group from alanine, we next asked whether 9-cis-RA analogs with alkyl substitutions at positions 9 and 13 could restore high affinity binding of this ligand. Interestingly, a 9-cis-RA derivative which has a methyl to hexyl extension at position 9 competes well with t-RA for binding to M406A with an IC50 of 25 nM (Table 2). However, this same derivative competes with t-RA for binding to Δ419 with an IC50 of only 432 nM. This 17-fold difference in IC50 value suggests that a hydrophobic interaction is occurring between the methyl group at position 9 on 9-cis-RA and the methionine R group at amino acid position 406 on Δ419. A 9-cis-RA derivative which has a pentyl group at position 13, however, competes poorly with t-RA for binding to either M406A or Δ419, further indicating the importance of position 9 on 9-cis-RA. Taken together, the results obtained using these two analogs of 9-cis-RA suggest that a 9-cis-RA analog with an alkyl substitution at position 9 may be capable of restoring a high affinity binding interaction on M406A.DISCUSSIONA number of the functions that allow the RARs to act as hormone-dependent transcription factors are contained within the ligand binding domain, namely, heterodimerization, hormone-dependent transactivation, and ligand binding (reviewed in (13Leid M. Kastner P. Chambon P. Trends Biochem. Sci. 1992; 17: 427-433Abstract Full Text PDF PubMed Scopus (803) Google Scholar)). All of these functions are dependent on the ability of the RARs to bind their cognate ligands. While it is well established that the RARs have a high affinity for the ligands, t-RA and 9-cis-RA(18Allenby G. Bocquel M.-T. Saunders M. Kazmer S. Speck J. Rosenberger M. Lovey A. Kastner P. Grippo J.F. Chambon P. Levin A.A. Proc. Natl. Acad. Sci. U. S. A. 1993; 90: 30-34Crossref PubMed Scopus (658) Google Scholar), little is actually known about critical amino acid interactions that may distinguish between the binding of these two stereoisomers.Recently, we reported that a truncated hRARα contains unique binding determinants for 9-cis-RA within the carboxyl-terminal region of the ligand binding domain(17Tate B.F. Allenby G. Janocha R. Kazmer S. Speck J. Sturzenbecker L.J. Abarzúa P. Levin A.A. Grippo J.F. Mol. Cell. Biol. 1994; 14: 2323-2330Crossref PubMed Google Scholar). In this study, we have determined the specific amino acids within this region (amino acids 405-419) that constitute this high affinity binding of 9-cis-RA by using alanine scanning mutagenesis(28Cunningham B.C. Wells J.A. Science. 1989; 244: 1081-1085Crossref PubMed Scopus (1079) Google Scholar). This region in RARα is predicted to form an amphipathic α-helix that is well conserved among nuclear receptors(23Danielian P.S. White R. Lees J.A. Parker M.G. EMBO J. 1992; 11: 1025-1033Crossref PubMed Scopus (716) Google Scholar). In fact, analyses of the majority of substituted mutants shown, the exceptions being P407A and P408A, using a Chou and Fasman algorithm (24Chou P.Y. Fasman G.D. Adv. Enzymol. 1978; 47: 45-147PubMed Google Scholar) indicates that the predicted secondary structure of this α-helix is not perturbed (data not shown). Overall, alanine scanning of amino acids 405 to 419 on a truncated hRARα demonstrates that the amino-terminal end of this region appears most important for high affinity 9-cis-RA binding. Specifically, amino acids 406, 407, 410, and to a lesser extent, 408, are critical as their replacement with alanine significantly alters the ability of the resulting receptors to bind 9-cis-RA.When we modeled this putative amphipathic α-helical region between amino acids 404 and 420 of the hRARα into a helical wheel(29Landschulz W.H. Johnson P.F. McKnight S.L. Science. 1988; 240: 1759-1764Crossref PubMed Scopus (2521) Google Scholar, 30Zenke M. Muñoz A. Sap J. Vennström B. Beug H. Cell. 1990; 61: 1035-1049Abstract Full Text PDF PubMed Scopus (168) Google Scholar), the importance of methionine 406 and isoleucine 410 becomes clear. Since one turn in an α-helix is approximately 3.6 amino acid residues long, these two amino acids actually lie in close proximity to one another (Fig. 5). In addition, with the ability of the proline residues to initiate and stabilize the formation of the α-helix(24Chou P.Y. Fasman G.D. Adv. Enzymol. 1978; 47: 45-147PubMed Google Scholar), a decrease in binding capacity for 9-cis-RA with the mutation of prolines at positions 407 and 408 is not surprising.Figure 5:Helical wheel analysis of a potential amphipathic α-helix on a hRARα (amino acids 406-419). The amino acid sequence of the carboxyl-terminal end of the ligand binding domain of the hRARα is projected showing the α-helical structure. The helix, as modeled here using a Chou and Fasman algorithm (24Chou P.Y. Fasman G.D. Adv. Enzymol. 1978; 47: 45-147PubMed Google Scholar) and software from IntelliGenetics, Inc., is amphipathic with charged residues to the lower right and hydrophobic residues to the upper left. Those residues found to be most critical for 9-cis-RA binding as well as for maintaining the integrity of the t-RA binding pocket are shown in boxes (methionine 406 and isoleucine 410).View Large Image Figure ViewerDownload Hi-res image Download (
Referência(s)