|
Figure 1. Characterization of full-length cDNA for Xenopus IRBP. A) cDNA XenB1 isolated under low stringency conditions [40], was used here to screen under high stringency a Xenopus stage 45 swimming tadpole cDNA library. B) This screen led to the isolation of Xen10a, a full-length Xenopus IRBP cDNA whose sequence is shown in Figure 2. Arrows represent individual sequencing runs. The solid bars indicate coding regions. E, EcoR1; X, XbaI.
|
|
Figure 2. Translated amino acid sequence of Xenopus IRBP. The first 22 amino acid residues (bold) comprise the signal sequence. The four homologous modules of the protein are boxed. The black boxes with white text are glycosylation consensus sequences. The shaded regions are segments of the recombinant protein that were verified by LC-MS/MS. The 3'-UTR contains two polyadenylation signal sequences (boxed and in bold). This cDNA sequence is available through the European Bioinformatics Information, Genbank and DDBJ Nucleotide Sequence databases under accession number X95473 and sequence identification XLIRBP.
|
|
Figure 3. Comparison of Xenopus IRBP with itself and known IRBPs. A) Dotplot of Xenopus IRBP, comparing the protein sequence against itself. The boxes on the ordinate and abscissa are schematic diagrams of modules 1 through 4. The numbering on the axes correspond with the amino acid residues in Figure 2. The diagonal lines indicate regions of internal similarity, and hence the presence of the four modules. B) Distance tree showing the relationship between Xenopus, human, bovine, goldfish and zebrafish IRBPs. The branch lengths are drawn to scale and the values at the nodes indicate the number of times a grouping occurred in a set of 100 bootstrap values. The long distance separating the teleosts from amphibians is due in part to the teleosts having only two modules.
|
|
Figure 4. Phylogenetic relationship of the IRBP modules. A) Phylogenetic distances between various modules of Xenopus IRBP protein. The numbers at the junctions are the number of times a branch point occurred out of 100 bootstrap reiterations. The tree was rooted with module 4, the ancestral module. B) An unrooted distance tree showing the relationships between the various IRBP modules from different animals. The tree was constructed as in panel A. Even numbered modules are more closely related to each other than to odd numbered modules.
|
|
Figure 5. Expression and purification of full-length Xenopus IRBP expressed in E. coli as a soluble thioredoxin/histidine-patch fusion protein (arrow). Coomassie blue stained 8% polyacrylamide gels, showing over expression of the recombinant IRBP. m = molecular weight markers; i = insoluble fraction; s = soluble fraction; "Pre" and "Post" refer to the bacterial fractions before and after induction with IPTG; p = purified protein.
|
|
Figure 6. Protection of all-trans retinol by full-length Xenopus IRBP. Three μl of an ethanolic solution of all-trans retinol was added to 400 μl of 2.7 μM full-length Xenopus IRBP in PBS. The final concentration of all-trans retinol in solution was 3.2 μM. The degradation of all-trans retinol was monitored by measuring its absorbance at 325 nm as a function of time. For each sample, absorbance measurements were made every 2 min for 72 minutes. Full-length Xenopus IRBP (filled circles) is able to protect all-trans retinol from degradation as compared to a PBS control (unfilled circles).
|
|
Figure 7. Fluorescence-binding studies of full-length Xenopus IRBP to all-trans retinol. The concentration of Xenopus IRBP was 0.65 μM in each panel. A) Titrations of IRBP with all-trans retinol as followed by monitoring the increase in retinol fluorescence (excitation, 330 nm; emission, 480 nm). Retinol fluorescence in the presence of IRBP (â,â, Î) is compared with that in the presence of an fluorescence matched solution of N-acetyl-L-tryptophanamide (â¡). The difference between these two curves, the fluorescence enhancement (-â-), represents all-trans retinol bound to the protein. The curve is a nonlinear least squares fit of Equation 1 to the binding data. Error bars are too small to be visualized. The number of binding sites per molecule of protein (N) was 3.19 ± 0.10 with Kdall-trans = 0.30 ± 0.05 μM (standard error of the mean). B). Emission spectra of apo- and holo-IRBP (curves a and b, respectively) were obtained upon excitation at 280 nm in the presence of a 10 fold excess of all-trans retinol. The drop in emission at 340 nm represents quenching of the protein's intrinsic fluorescence. The emission at 480 (arrow) represents energy transfer to the bound retinol. C) Titration monitoring quenching of intrinsic protein fluorescence by bound retinol. Excitation and emission wavelengths were 280 and 340 nm, respectively. The inner filter effect has been accounted for graphically as previously described (50) (-â-, actual meausurements; -â-, after correction). Calculated binding parameters: N = 1.93 ± 0.41; Kdall-trans = 0.66 ± 0.14 μM. D) Titration monitoring energy transfer (increase in fluorescence at 480 (arrow). Calculated binding parameters: N = 3.72 ± 0.20; Kdall-trans = 0.29 ± 0.12.
|
|
Figure 8. 11-cis retinaldehyde binding to full-length Xenopus IRBP. Since retinaldehydes are nonfluorescent compounds their binding cannot be followed by ligand fluorescence enhancement or energy transfer. Here, the binding of 11-cis retinaldehyde is followed by monitoring quenching of endogenous protein fluorescence, and indirectly by competition with the efficient fluorophore all-trans retinol. A) Representative titrations monitoring quenching of intrinsic protein fluorescence by bound 11-cis retinaldehyde. Excitation and emission wavelengths were 280 and 340 nm, respectively. The inner filter effect was accounted for by graphical correction as previously described [56] (-â-, before correction for inner filter effect; -â-, after correction). The binding parameters were calculated to be: N = 1.81 ± 0.15; Kd11-cis = 0.28 ± 0.05 μM. B. Representative competition titration. From each titration the emission of an 0.D.280 matched solution of N-acetyl-L-tryptophanamide was subtracted. Binding parameters: N = 3.53 ± 0.19 with Kdall-trans = 0.22 ± 0.13 μM; Kd11-cis = 0.21 ± 0.10 μM.
|
|
Figure 9. Competitive inhibition by 11-cis retinaldehyde of all-trans retinol binding to the individual modules of Xenopus IRBP. Binding of all-trans retinol to the individual modules was measured by fluorescence enhancement in the presence of varying amounts of 11-cis retinal. Panels A through D correspond to modules 1 through 4 respectively. Assuming both all-trans retinol and 11-cis retinal share the same binding sites, all-trans retinol binding to IRBP is competitively inhibited. Three dimensional nonlinear regression was used to determine the number of binding sites, the dissociation constant of all-trans retinol, and the dissociation constant of 11-cis retinal for each of the individual modules. The binding parameters are summarized in Table 3. The concentrations of the proteins used in the titrations were as follows: module 1 (0.99 μM; panel A); module 2 (1.10 μM; panel B); module 3 (1.10 μM; panel C); module 4 (1.00 μM; panel D).
|
|
Figure 10. Retinol binding to Xenopus IRBPs consisting of two contiguous modules. Binding of double modules IRBPs with all-trans retinol as followed during titrations by monitoring the increase in retinol fluorescence (excitation, 330 nm; emission, 480 nm). The concentration of the IRBP double module was: modules 1&2, 0.67 μM; modules 3&4, 0.57 μM. A) Xen IRBP modules 1&2 showed N = 2.45 ± 0.11 with Kd = 0.049 ± 0.023 μM. B) Xen IRBP modules 3&4 showed N = 1.43 ± 0.21 with Kd = 0.19 ± 0.05 μM.
|
|
Figure 11. Homology modeling of the modules of Xenopus IRBP. A) Superimposition of the predicted structures of X1-, X3- and X4IRBPs with the X-ray crystal structure of X2IRBP. Ribbon diagram: X1IRBP, orange; X2IRBP, green; X3IRBP, magenta; X4IRBP, yellow. Molecular docking studies predict two binding sites for all-trans retinol. The best scoring poses for the conformational search are shown with the retinol molecule colored grey in site I, and blue in site II. Close-up views of the docked all-trans retinol molecule in (B) site I, and (C) site II. All side chains are shown by thin lines in the colors of the backbones. The amino acid residues within contact distances to the ligand are tabulated in Table 5.
|