Identification of CRALBP ligand interactions by photoaffinity labeling, hydrogen/deuterium exchange, and structural modeling

Cellular retinaldehyde-binding protein (CRALBP) functions in the retinal pigment epithelium (RPE) as an acceptor of 11-cis-retinol in the isomerization step of the rod visual cycle and as a substrate carrier for 11-cis-retinol dehydrogenase. Toward a better understanding of CRALBP function, the ligand binding cavity in human recombinant CRALBP (rCRALBP) was characterized by photoaffinity labeling with 3-diazo-4-keto-11-cis-retinal and by high resolution mass spectrometric topological analyses. Eight photoaffinity-modified residues were identified in rCRALBP by liquid chromatography tandem mass spectrometry, including Tyr(179), Phe(197), Cys(198), Met(208), Lys(221), Met(222), Val(223), and Met(225). Multiple different adduct masses were found on the photolabeled residues, and the molecular identity of each modification remains unknown. Supporting the specificity of photo-labeling, 50% of the modified residues have been associate with retinoid interactions by independent analyses. In addition, topological analysis of apo- and holo-rCRALBP by hydrogen/deuterium exchange and mass spectrometry demonstrated residues 198-255 incorporate significantly less deuterium when the retinoid binding pocket is occupied with 11-cis-retinal. This hydrophobic region encompasses all but one of the photo-labeled residues. A structural model of CRALBP ligand binding domain was constructed based on the crystal structures of three homologues in the CRAL-TRIO family of lipid-binding proteins. In the model, all of the photolabeled residues line the ligand binding cavity except Met(208), which appears to reside in a flexible loop at the entrance/exit of the ligand cavity. Overall, the results expand to 12 the number of residues proposed to interact with ligand and provide further insight into CRALBP ligand and protein interactions.     

Mapping the ligand binding pocket in the cellular retinaldehyde binding protein

Retinoid interactions determine the function of the cellular retinaldehyde binding protein (CRALBP) in the rod visual cycle where it serves as an 11-cis-retinol acceptor for the enzymatic isomerization of all-trans- to 11-cis-retinol and as a substrate carrier for 11-cis-retinol dehydrogenase (RDH5). Based on preliminary NMR studies suggesting retinoid interactions with Met and Trp residues, human recombinant CRALBP (rCRALBP) with altered Met or Trp were produced and analyzed for ligand interactions. The primary structures of the purified proteins were verified for mutants M208A, M222A, M225A, W165F, and W244F, then retinoid binding properties and substrate carrier functions were evaluated. All the mutant proteins bound 11-cis- and 9-cis-retinal and therefore were not grossly misfolded. Altered UV-visible spectra and lower retinoid binding affinities were observed for the mutants, supporting modified ligand interactions. Altered kinetic parameters were observed for RDH5 oxidation of 11-cis-retinol bound to rCRALBP mutants M222A, M225A, and W244F, supporting impaired substrate carrier function. Heteronuclear single quantum correlation NMR analyses confirmed localized structural changes upon photoisomerization of rCRALBP-bound 11-cis-retinal and demonstrated ligand-dependent conformational changes for residues Met-208, Met-222, Trp-165, and Trp-244. Furthermore, residues Met-208, Met-222, Met-225, and Trp-244 are within a region exhibiting high homology to the ligand binding cavity of phosphatidylinositol transfer protein. Overall the data implicate Trp-165, Met-208, Met-222, Met-225, and Trp-244 as components of the CRALBP ligand binding cavity.     

Topological and epitope mapping of the cellular retinaldehyde-binding protein from retina.

Cellular retinaldehyde-binding protein (CRALBP) carries 11-cis-retinol or 11-cis-retinaldehyde as endogenous ligands and may function as a substrate carrier protein that modulates interaction of these retinoids with visual cycle enzymes. As a first approach to identifying functional domains and protein recognition sites in CRALBP, a low resolution topological and epitope map has been developed using monoclonal and polyclonal antibodies and limited proteolysis. Fifteen peptides of 8-31 residues spanning 99% of the 316-residue bovine CRALBP were synthesized and used to prepare 13 anti-peptide polyclonal antibodies. Using a competitive ELISA procedure, peptide epitopes were classified as either accessible or inaccessible in the native protein based on the extent of their recognition by these site-specific antibodies. Use of the synthetic peptides to map the epitopes of a polyclonal antibody to intact CRALBP confirmed that the amino terminus and carboxyl terminus are immunodominate regions and hence likely to be exposed, at least in part. Limited tryptic proteolysis of native CRALBP produced three major fragments which were shown by microsequence and Western analysis to be derived from sequential loss of short peptides from the amino terminus. None of these major fragments reacted with four monoclonal antibodies (mAbs) to intact CRALBP although each mAb immunoprecipitated native CRALBP. These results and the lack of mAb recognition of any of the synthetic peptides indicates that the amino terminus of the protein is exposed and contains part of an assembly epitope recognized by the mAbs. Overall this study indicates that residues 1-30, 100-124, and 257-285 contain highly exposed segments in the native protein and therefore constitute potential interaction domains for CRALBP and visual cycle enzymes. Residues 30-99 and 176-229 are inaccessible in the native structure and may be involved with retinoid binding. These results provide a basis for a systematic higher resolution mutagenesis study directed toward correlating CRALBP structural domains with function.     

The complete primary structure of the cellular retinaldehyde-binding protein from bovine retina

Cellular retinaldehyde-binding protein (CRALBP) carries 11-cis-retinol and 11-cis-retinaldehyde as endogenous ligands and may be a functional component of the visual cycle. The complete amino acid sequence of CRALBP from bovine retina has been determined by direct microanalysis of the protein. Bovine CRALBP contains 316 residues in a single amino-terminal-blocked chain corresponding to a molecular weight of 36,421, inclusive of the blocking group. Overlapping peptides were generated by cleavage of lysyl, arginyl, methionyl, glutamyl, and one tryptophanyl bond and sequenced by gas-phase Edman degradation. Analysis of amino-terminal arginyl and methionyl peptides by fast atom bombardment mass spectrometry identified the N alpha-blocking group as an acetyl moiety, and tandem mass spectrometry provided the sequence of the first 9 residues. Comparison of CRALBP with other known protein sequences reveals no significant structural relatedness. The present results provide a basis for relating CRALBP domains with physiological function and for the future development of a more detailed three-dimensional model of the interaction of 11-cis-retinaldehyde with protein.     

Structural and functional characterization of recombinant human cellular retinaldehyde-binding protein.

Cellular retinaldehyde-binding protein (CRALBP) is abundant in the retinal pigment epithelium (RPE) and Müller cells of the retina where it is thought to function in retinoid metabolism and visual pigment regeneration. The protein carries 11-cis-retinal and/or 11-cis-retinol as endogenous ligands in the RPE and retina and mutations in human CRALBP that destroy retinoid binding functionality have been linked to autosomal recessive retinitis pigmentosa. CRALBP is also present in brain without endogenous retinoids, suggesting other ligands and physiological roles exist for the protein. Human recombinant cellular retinaldehyde-binding protein (rCRALBP) has been over expressed as non-fusion and fusion proteins in Escherichia coli from pET3a and pET19b vectors, respectively. The recombinant proteins typically constitute 15-20% of the soluble bacterial lysate protein and after purification, yield about 3-8 mg per liter of bacterial culture. Liquid chromatography electrospray mass spectrometry, amino acid analysis, and Edman degradation were used to demonstrate that rCRALBP exhibits the correct primary structure and mass. Circular dichroism, retinoid HPLC, UV-visible absorption spectroscopy, and solution state 19F-NMR were used to characterize the secondary structure and retinoid binding properties of rCRALBP. Human rCRALBP appears virtually identical to bovine retinal CRALBP in terms of secondary structure, thermal stability, and stereoselective retinoid-binding properties. Ligand-dependent conformational changes appear to influence a newly detected difference in the bathochromic shift exhibited by bovine and human CRALBP when complexed with 9-cis-retinal. These recombinant preparations provide valid models for human CRALBP structure-function studies.     

Crystal structure of the human supernatant protein factor

Supernatant protein factor (SPF) promotes the epoxidation of squalene catalyzed by microsomes. Several studies suggest its in vivo role in the cholesterol biosynthetic pathway by a yet unknown mechanism. SPF belongs to a family of lipid binding proteins called CRAL_TRIO, which include yeast phosphatidylinositol transfer protein Sec14 and tocopherol transfer protein TTP. The crystal structure of human SPF at a resolution of 1.9 A reveals a two domain topology. The N-terminal 275 residues form a Sec14-like domain, while the C-terminal 115 residues consist of an eight-stranded jelly-roll barrel similar to that found in many viral protein structures. The ligand binding cavity has a peculiar horseshoe-like shape. Contrary to the Sec14 crystal structure, the lipid-exchange loop is in a closed conformation, suggesting a mechanism for lipid exchange.     

Amino acids involved in conformational dynamics and G protein coupling of an odorant receptor: targeting gain-of-function mutation

Thousands of different odorants are recognized and discriminated by odorant receptors (ORs) in the guanine nucleotide-binding protein (G protein)-coupled seven-transmembrane receptor family. Odorant-bound ORs stimulate Gs-type G proteins, Galphaolf, which in turn activates cAMP-mediated signaling pathway in olfactory sensory neurons. To better understand the molecular basis for OR activation and G protein coupling, we analyzed the effects of a series of site-directed mutations of mouse ORs, on function. Mutations of conserved amino acid residues in an intracellular loop or the C-terminus resulted in loss of activity without impairing ligand-binding activity, indicating that these residues are involved in Galphas/olf coupling. Moreover, mutation of the serine in KAFSTC, the OR-specific sequence motif, resulted in a dramatic increase in odorant responsiveness, suggesting that the motif is involved in a conformational change of the receptor that regulates G protein coupling efficiency. Our results provide insights into how ORs switch from an inactive to an active state, as well as where and how activated ORs interact with G proteins.     

Cloning of the cDNAs encoding the cellular retinaldehyde-binding protein from bovine and human retina and comparison of the protein structures

A 1173-base pair cDNA encoding bovine cellular retinaldehyde-binding protein (CRALBP) was cloned from a bovine retinal cDNA expression library using as probes both anti-CRALBP polyclonal and monoclonal antibodies. The amino acid sequence deduced from the cDNA corresponds exactly to that determined by direct analysis of NH2-terminally acetylated bovine CRALBP (Crabb, J. W., Johnson, C. M., Carr, S. A., Armes, L. G., and Saari, J. C. (1988) J. Biol. Chem. 263, 18678-18687). Nick-translated bovine CRALBP cDNA probes were then used to clone from a human retinal cDNA library a 1317-base pair cDNA encoding human CRALBP. Bovine and human CRALBP are 92% identical in amino acid sequence and not related to any other known protein sequence. Both the bovine and human proteins contain 316 residues and have calculated molecular weights of 36,378 and 36,347, respectively, exclusive of the NH2-terminal blocking groups. The CRALBP cDNA clones should prove valuable as tools for studying the physiological role of the protein in vision and visual disorders.     

Disease-causing mutations in the cellular retinaldehyde binding protein tighten and abolish ligand interactions

Mutations in the human cellular retinaldehyde binding protein (CRALBP) gene cause retinal pathology. To understand the molecular basis of impaired CRALBP function, we have characterized human recombinant CRALBP containing the disease causing mutations R233W or M225K. Protein structures were verified by amino acid analysis and mass spectrometry, retinoid binding properties were evaluated by UV-visible and fluorescence spectroscopy and substrate carrier functions were assayed for recombinant 11-cis-retinol dehydrogenase (rRDH5). The M225K mutant was less soluble than the R233W mutant and lacked retinoid binding capability and substrate carrier function. In contrast, the R233W mutant exhibited solubility comparable to wild type rCRALBP and bound stoichiometric amounts of 11-cis- and 9-cis-retinal with at least 2-fold higher affinity than wild type rCRALBP. Holo-R233W significantly decreased the apparent affinity of rRDH5 for 11-cis-retinoid relative to wild type rCRALBP. Analyses by heteronuclear single quantum correlation NMR demonstrated that the R233W protein exhibits a different conformation than wild type rCRALBP, including a different retinoid-binding pocket conformation. The R233W mutant also undergoes less extensive structural changes upon photoisomerization of bound ligand, suggesting a more constrained structure than that of the wild type protein. Overall, the results show that the M225K mutation abolishes and the R233W mutation tightens retinoid binding and both impair CRALBP function in the visual cycle as an 11-cis-retinol acceptor and as a substrate carrier.     

Visual cycle impairment in cellular retinaldehyde binding protein (CRALBP) knockout mice results in delayed dark adaptation

Mutations in the human CRALBP gene cause retinal pathology and delayed dark adaptation. Biochemical studies have not identified the primary physiological function of CRALBP. To resolve this, we generated and characterized mice with a non-functional CRALBP gene (Rlbp1(-/-) mice). The photosensitivity of Rlbp1(-/-) mice is normal but rhodopsin regeneration, 11-cis-retinal production, and dark adaptation after illumination are delayed by >10-fold. All-trans-retinyl esters accumulate during the delay indicating that isomerization of all-trans- to 11-cis-retinol is impaired. No evidence of photoreceptor degeneration was observed in animals raised in cyclic light/dark conditions for up to 1 year. Albino Rlbp(-/-) mice are protected from light damage relative to the wild type. These findings support a role for CRALBP as an acceptor of 11-cis-retinol in the isomerization reaction of the visual cycle.     

Molecular modeling of the GABAC receptor ligand-binding domain

We have constructed a molecular model of the ligand-binding domain of the GABA(C) receptor, which is a member of the Cys-loop ligand-gated ion channel family. The extracellular domains of these receptors share similar sequence homology (20%) with Limnaea acetylcholine-binding protein for which an X-ray crystal structure is available. We used this structure as a template for homology modeling of the GABA(C) receptor extracellular domain using FUGUE and MODELLER software. FlexX was then used to dock GABA into the receptor ligand-binding site, resulting in three alternative energetically favorable orientations. Residues located no more than 5 A from the docked GABA were identified for each model; of these, three were found to be common to all models with 14 others present only in certain models. Using data from experimental studies, we propose that the most likely orientation of GABA is with its amine close to Y198, and its carboxylate close to R104. These studies have therefore provided a model of the ligand-binding domain, which will be useful for both GABA(C) and GABA(A) receptor studies, and have also yielded an experimentally testable hypothesis of the location of GABA in the binding pocket. [Figure: see text].     

Excision of a proposed electron transfer pathway in cytochrome c peroxidase and its replacement by a ligand-binding channel

A previously proposed electron transfer (ET) pathway in the heme enzyme cytochrome c peroxidase has been excised from the structure, leaving an open ligand-binding channel in its place. Earlier studies on cavity mutants of this enzyme have revealed structural plasticity in this region of the molecule. Analysis of these structures has allowed the design of a variant in which the specific section of protein backbone representing a previously proposed ET pathway is accurately extracted from the protein. A crystal structure verified the creation of an open channel that overlays the removed segment, extending from the surface of the protein to the heme at the core of the protein. A number of heterocyclic cations were found to bind to the proximal-channel mutant with affinities that can be rationalized based on the structures. It is proposed that small ligands bind more weakly to the proximal-channel mutant than to the W191G cavity due to an increased off rate of the open channel, whereas larger ligands are able to bind to the channel mutant without inducing large conformational changes. The structure of benzimidazole bound to the proximal-channel mutant shows that the ligand accurately overlays the position of the tryptophan radical center that was removed from the wild-type enzyme and displaces four of the eight ordered solvent molecules seen in the empty cavity. Ligand binding also caused a small rearrangement of the redesigned protein loop, perhaps as a result of improved electrostatic interactions with the ligand. The engineered channel offers the potential for introducing synthetic replacements for the removed structure, such as sensitizer-linked substrates. These installed "molecular wires" could be used to rapidly initiate reactions, trap reactive intermediates, or answer unresolved questions about ET pathways.     

The crystal structure of the carbohydrate-recognition domain of the glycoprotein sorting receptor p58/ERGIC-53 reveals an unpredicted metal-binding site and conformational changes associated with calcium ion binding

p58/ERGIC-53 is a calcium-dependent animal lectin that acts as a cargo receptor, binding to a set of glycoproteins in the endoplasmic reticulum (ER) and transporting them to the Golgi complex. It is similar in structure to calcium-dependent leguminous lectins. We have determined the structure of the carbohydrate-recognition domain of p58/ERGIC-53 in its calcium-bound form. The structure reveals localized but large conformational changes in relation to the previously determined metal ion-free structure, mapping mostly to the ligand-binding site. It reveals the presence of two calcium ion-binding sites located 6A apart, one of which has no equivalent in the plant lectins. The second metal ion-binding site present in that class of lectins, binding Mn(2+), is absent from p58/ERGIC-53. The absence of a short loop in the ligand-binding site in this protein suggests that it has adapted to optimally bind the high-mannose Man(8)(GlcNAc)(2) glycan common to glycoproteins at the ER exit stage.     

Prediction of ligand-binding sites of proteins by molecular docking calculation for a random ligand library

A new approach to predicting the ligand-binding sites of proteins was developed, using protein-ligand docking computation. In this method, many compounds in a random library are docked onto the whole protein surface. We assumed that the true ligand-binding site would exhibit stronger affinity to the compounds in the random library than the other sites, even if the random library did not include the ligand corresponding to the true binding site. We also assumed that the affinity of the true ligand-binding site would be correlated to the docking scores of the compounds in the random library, if the ligand-binding site was correctly predicted. We call this method the molecular-docking binding-site finding (MolSite) method. The MolSite method was applied to 89 known protein-ligand complex structures extracted from the Protein Data Bank, and it predicted the correct binding sites with about 80-99% accuracy, when only the single top-ranked site was adopted. In addition, the average docking score was weakly correlated to the experimental protein-ligand binding free energy, with a correlation coefficient of 0.44.     

Identification of a novel cytosolic tocopherol-binding protein: structure, specificity, and tissue distribution

Alpha-tocopherol plays an important role as a lipid-soluble antioxidant. It is present in all major mammalian cell types and shows tissue-specific distribution. This suggests the presence of specific proteins involved in intracellular distribution or metabolism of alpha-tocopherol. A diminution of tocopherol plasma concentrations contributes to the development of diseases such as vitamin E deficiency (AVED), atherosclerosis, and prostate cancer. Further evidence has been obtained for the existence of sites in cellular metabolism and signal transduction where alpha-tocopherol potentially plays a regulatory role. A signal transduction modulation specific for alpha-tocopherol has been described in several model systems. Using radioactively labeled alpha-tocopherol as tracer, we have isolated a new alpha-tocopherol-associated protein (TAP) from bovine liver. This protein has a molecular mass of 46 kDa and an isoelectric point of 8.1. From its partial amino acid sequence, a human gene has been identified with high homology to the newly described protein. Sequence analysis has established that the new TAP has structural motifs suggesting its belonging to a family of hydrophobic ligand-binding proteins (RALBP, CRALBP, alpha-TTP, SEC 14, PTN 9, RSEC 45). Human TAP has been cloned into Escherichia coli, and its tissue-specific expression has been assessed by Northern blot analysis.     

A yeast sensor of ligand binding.

We describe a biosensor that reports the binding of small-molecule ligands to proteins as changes in growth of temperature-sensitive yeast. The yeast strains lack dihydrofolate reductase (DHFR) and are complemented by mouse DHFR containing a ligand-binding domain inserted in a flexible loop. Yeast strains expressing two ligand-binding domain fusions, FKBP12-DHFR and estrogen receptor-alpha (ERalpha)-DHFR, show increased growth in the presence of their corresponding ligands. We used this sensor to identify mutations in residues of ERalpha important for ligand binding, as well as mutations generally affecting protein activity or expression. We also tested the sensor against a chemical array to identify ligands that bind to FKBP12 or ERalpha. The ERalpha sensor was able to discriminate among estrogen analogs, showing different degrees of growth for the analogs that correlated with their relative binding affinities (RBAs). This growth assay provides a simple and inexpensive method to select novel ligands and ligand-binding domains.     

Cloning and Characterization of a Novel, Human Cellular Retinaldehyde-binding Protein CRALBP-like (CRALBPL) Gene

Cellular retinaldehyde-binding protein (CRALBP) plays a role in the vertebrate visual process as a substrate-routing protein. It belongs to a widespread lipid-binding SEC14-like protein family. All the members of the family have the lipid-binding domain called CRAL-TRIO. Here we have isolated a new human CRAL-TRIO domain containing a CRALBP-like (CRALBPL) gene from the cDNA library of human adult brain. The CRALBPL gene consisted of 1,694 bp and had an ORF encoding putatively 354 amino acids with a CRAL-TRIO domain from 118 to 279 aa. The expression pattern in 18 human tissues indicated that CRALBPL gene was mainly expressed in brain. The alignment of CRAL-TRIO domain showed that CRALBPL had 45% identity with human CRALBP. Subcellular location revealed that CRALBPL protein was located in the cytoplasm of HeLa cells. Western blotting indicated that the CRALBPL had a molecular weight of about 40 kDa.     

Oligomerization of DsRed is required for the generation of a functional red fluorescent chromophore

Site-directed mutagenesis was used to map the ligand-binding surface of the type II transforming growth factor-beta receptor extracellular domain (TbetaRII-ECD). Two putative ligand-binding sites were probed, the first being a predicted hydrophobic patch, the second being the finger 1 surface loop. Nine residues were mutated in the context of full-length TbetaRII and the effect of these mutations on ligand-binding and receptor signaling was analyzed. Complementary information was obtained by examining 'natural' evolutionary TbetaRII mutations. Together, the results indicate that residues within the finger 1 region, but not the hydrophobic patch, of the TbetaRII-ECD are required for productive ligand-binding. We conclude that, surprisingly, the ECDs of TbetaRII and type II activin receptor utilize distinct interacting surfaces for binding their respective ligands.     

Artificial protein cavities as specific ligand-binding templates: characterization of an engineered heterocyclic cation-binding site that preserves the evolved specificity of the parent protein.

Cavity complementation has been observed in many proteins, where an appropriate small molecule binds to a cavity-forming mutant. Here, the binding of compounds to the W191G cavity mutant of cytochrome c peroxidase is characterized by X-ray crystallography and binding thermodynamics. Unlike cavities created by removal of hydrophobic side-chains, the W191G cavity does not bind neutral or hydrophobic compounds, but displays a strong specificity for heterocyclic cations, consistent with the role of the protein to stabilize a tryptophan radical at this site. Ligand dissociation constants for the protonated cationic state ranged from 6 microM for 2-amino-5-methylthiazole to 1 mM for neutral ligands, and binding was associated with a large enthalpy-entropy compensation. X-ray structures show that each of 18 compounds with binding behavior bind specifically within the artificial cavity and not elsewhere in the protein. The compounds make multiple hydrogen bonds to the cavity walls using a subset of the interactions seen between the protein and solvent in the absence of ligand. For all ligands, every atom that is capable of making a hydrogen bond does so with either protein or solvent. The most often seen interaction is to Asp235, and most compounds bind with a specific orientation that is defined by their ability to interact with this residue. Four of the ligands do not have conventional hydrogen bonding atoms, but were nevertheless observed to orient their most polar CH bond towards Asp235. Two of the larger ligands induce disorder in a surface loop between Pro190 and Asn195 that has been identified as a mobile gate to cavity access. Despite the predominance of hydrogen bonding and electrostatic interactions, the small variation in observed binding free energies were not correlated readily with the strength, type or number of hydrogen bonds or with calculated electrostatic energies alone. Thus, as with naturally occurring binding sites, affinities to W191G are likely to be due to a subtle balance of polar, non-polar, and solvation terms. These studies demonstrate how cavity complementation and judicious choice of site can be used to produce a protein template with an unusual ligand-binding specificity.     

Sequence annotation of nuclear receptor ligand-binding domains by automated homology modeling

The quality of three-dimensional homology models derived from protein sequences provides an independent measure of the suitability of a protein sequence for a certain fold. We have used automated homology modeling and model assessment tools to identify putative nuclear hormone receptor ligand-binding domains in the genome of Caenorhabditis elegans. Our results indicate that the availability of multiple crystal structures is crucial to obtaining useful models in this receptor family. The majority of annotated mammalian nuclear hormone receptors could be assigned to a ligand-binding domain fold by using the best model derived from any of four template structures. This strategy also assigned the ligand-binding domain fold to a number of C.elegans. sequences without prior annotation. Interestingly, the retinoic acid receptor crystal structure contributed most to the number of sequences that could be assigned to a ligand-binding domain fold. Several causes for this can be suggested, including the high quality of this protein structure in terms of our assessment tools, similarity between the biological function or ligand of this receptor and the modeled genes and gene duplication in C.elegans.