Screening transthyretin amyloid fibril inhibitors: characterization of novel multiprotein,multiligand complexes by mass spectrometry

Tetrameric transthyretin is involved in transport of thyroxine and, through its interactions with retinol binding protein, vitamin A. Dissociation of these structures is widely accepted as the first step in the formation of transthyretin amyloid fibrils. Using a mass spectrometric approach, we have examined a series of 18 ligands proposed as inhibitors of this process. The ligands were evaluated for their ability to bind to and stabilize the tetrameric structure, their cooperativity in binding, and their ability to compete with the natural ligand thyroxine. The observation of a novel ten-component complex containing six protein subunits, two vitamin molecules, and two synthetic ligands allows us to conclude that ligand binding does not inhibit association of transthyretin with holo retinol binding protein.     

Homology model of thyroxine binding globulin and elucidation of the thyroid hormone binding site.

The tertiary structure of thyroxine binding globulin (TBG) has been modelled on the basis of its close homology to alpha 1-antitrypsin, the archetype of the serine protease inhibitor (serpin) superfamily. Energy minimization was applied to the model to refine the structure further. The putative thyroid hormone binding region suggested in previous labelling studies was found to exist within a beta-barrel structure of complementary dimensions to the thyroid hormones. The model also revealed that the binding cleft provides the hydrophobic environment and specific ionic interaction sites deemed important for thyroid hormone binding. The model is in good agreement with evidence derived from previously reported T3 and T4 binding, stability and isoelectric focussing studies of TBG and TBG variants. Finally, T4 analogue and drug binding studies have enabled us to postulate the orientation and manner of hormone binding to TBG. This may prove to be of assistance in the development of potent and specific, non-thyroidal ligands and also aid in the understanding of physiological thyroid hormone binding interactions.     

Isolation, characterization, cDNA cloning and gene expression of an avian transthyretin. Implications for the evolution of structure and function of transthyretin in vertebrates.

A chicken liver cDNA library was constructed in bacteriophage lambda gt10. A full-length transthyretin cDNA clone was identified by screening with rat transthyretin cDNA and was sequenced. A three-dimensional model of chicken transthyretin was obtained by computer-graphics-based prediction from the derived amino acid sequence for chicken transthyretin and from the structure of human transthyretin determined by X-ray diffraction analysis [Blake, C.C.F., Geisow, M.J., Oatley, S.J., Rérat, B. & Rérat, C. (1978) J. Mol. Biol. 121, 339-356]. The similarity of the amino acid sequences of chicken and human transthyretins was 75% overall and 100% for the central channel containing the thyroxine-binding site. Also, the organization of the transthyretin gene into exons and introns and the tissue specificity of expression of the transthyretin gene were similar in chicken and mammals, despite an evolutionary distance of about 3 x 10(8) years from their common ancestor, the Cotylosaurus. By far the highest levels of transthyretin mRNA were found in choroid plexus. The data suggest a fundamental role for the cerebral expression of transthyretin in all vertebrates. It has been proposed that this role is the transport of thyroxine from the bloodstream to the brain [Schreiber, G., Aldred, A.R., Jaworowski, A., Nilsson, C., Achen, M.G. & Segal, M.B. (1990) Am. J. Physiol. 258, R338-R345].     

Thyroxine-protein interactions. Binding constants for interaction of thyroxine analogues with the thyroxine binding site on human thyroxine-binding globulin.

The binding constants for interaction of various thryoxine analogues with the thyroxine binding site on human thyroxine-binding globulin have been determined. Equilibrium dialysis, at pH 7.4 and 37 degrees C, was used to measure the competitive effects of different iodothyronine compounds on the binding of 125I-labeled thyroxine to highly purified thyroxine-binding globulin. Relative to L-thyroxine, K = 6 . 10(9) M-1, the association constants of some important analogues were D-thyroxine, 1.04 . 10(9) M-1, 3,5-diiodo-3'-isopropyl-L-thyronine, 4.9 . 10(8) M-1; L-triiodothyronine, 3.3 . 10(8) M-1, 3,3',5'-DL-triiodothyronine (reverse triiodothyronine), 3.1. 10(8) M-1; tetraiodothyropropionic acid, 2.7 . 10(8) M-1; tetraiodothyroacetic acid, 2.6 . 10(8) M-1; 3', 5'- diiodo-DL-thyronine, 8.3 . 10(7) M-1; and 3,5-diiodo-DL-thyronine, 7.1 . 10(7) M-1. Calculation of the deltaG0 values for binding of the analogues indicates that a major contribution to the free energy favoring binding is made by the alanine side chain of thyroxine. A change in configuration of the alpha-amino group from the L to D form causes an unfavorable change of 1 kcal/mol in the free energy of binding. Removal of the alpha-amino group as in tetraiodothyropropionic acid causes an unfavorable change of 1.9 kcal/mol in the free energy of binding. With regard to ring substituents, the results indicate that the two inner 3,5-iodines make about the same contribution to binding as the two outer 3', 5'-iodines.     

The role of the S1 binding site of carboxypeptidase M in substrate specificity and turn-over

The influence of the P1 amino acid on the substrate selectivity, the catalytic parameters K(m) and k(cat), of carboxypeptidase M (CPM) (E.C. 3.4.17.12) was systematically studied using a series of benzoyl-Xaa-Arg substrates. CPM had the highest catalytic efficiency (k(cat)/K(m)) for substrates with Met, Ala and aromatic amino acids in the penultimate position and the lowest with amino acids with branched side-chains. Substrates with Pro in P1 were not cleaved in similar conditions. The P1 substrate preference of CPM differed from that of two other members of the carboxypeptidase family, CPN (CPN/CPE subfamily) and CPB (CPA/CPB subfamily). Aromatic P1 residues discriminated most between CPM and CPN. The type of P2 residue also influenced the k(cat) and K(m) of CPM. Extending the substrate up to P7 had little effect on the catalytic parameters. The substrates were modelled in the active site of CPM. The results indicate that P1-S1 interactions play a role in substrate binding and turn-over.     

Role of the low-affinity binding site in electron transfer from cytochrome C to cytochrome C peroxidase

The interaction of yeast iso-1-cytochrome c (yCc) with the high- and low-affinity binding sites on cytochrome c peroxidase compound I (CMPI) was studied by stopped-flow spectroscopy. When 3 microM reduced yCc(II) was mixed with 0.5 microM CMPI at 10 mM ionic strength, the Trp-191 radical cation was reduced from the high-affinity site with an apparent rate constant >3000 s(-1), followed by slow reduction of the oxyferryl heme with a rate constant of only 10 s(-1). In contrast, mixing 3 microM reduced yCc(II) with 0.5 microM preformed CMPI *yCc(III) complex led to reduction of the radical cation with a rate constant of 10 s(-1), followed by reduction of the oxyferryl heme in compound II with the same rate constant. The rate constants for reduction of the radical cation and the oxyferryl heme both increased with increasing concentrations of yCc(II) and remained equal to each other. These results are consistent with a mechanism in which both the Trp-191 radical cation and the oxyferryl heme are reduced by yCc(II) in the high-affinity binding site, and the reaction is rate-limited by product dissociation of yCc(III) from the high-affinity site with apparent rate constant k(d). Binding yCc(II) to the low-affinity site is proposed to increase the rate constant for dissociation of yCc(III) from the high-affinity site in a substrate-assisted product dissociation mechanism. The value of k(d) is <5 s(-1) for the 1:1 complex and >2000 s(-1) for the 2:1 complex at 10 mM ionic strength. The reaction of horse Cc(II) with CMPI was greatly inhibited by binding 1 equiv of yCc(III) to the high-affinity site, providing evidence that reduction of the oxyferryl heme involves electron transfer from the high-affinity binding site rather than the low-affinity site. The effects of CcP surface mutations on the dissociation rate constant indicate that the high-affinity binding site used for the reaction in solution is the same as the one identified in the yCc*CcP crystal structure.     

Thyroxine transport in choroid plexus

The role of the choroid plexus in thyroid hormone transport between body and brain, suggested by strong synthesis and secretion of transthyretin in this tissue, was investigated in in vitro and in vivo systems. Rat choroid plexus pieces incubated in vitro were found to accumulate thyroid hormones from surrounding medium in a non-saturable process. At equilibrium, the ratio of thyroid hormone concentration in choroid plexus pieces to that in medium decreased upon increasing the concentration of transthyretin in the medium. Fluorescence quenching of fluorophores located at different depths in liposome membranes showed maximal hormone accumulation in the middle of the phospholipid bilayer. Partition coefficients of thyroxine and triiodothyronine between lipid and aqueous phase were about 20,000. After intravenous injection of 125I-labeled thyroid hormones, choroid plexus and parts of the brain steadily accumulated 125I-thyroxine, but not [125I]triiodothyronine, for many hours. The accumulation of 125I-thyroxine in choroid plexus preceded that in brain. The amount of 125I-thyroxine in non-brain tissues and the [125I]triiodothyronine content of all tissues decreased steadily beginning immediately after injection. A model is proposed for thyroxine transport from the bloodstream into cerebrospinal fluid based on partitioning of thyroxine between choroid plexus and surrounding fluids and binding of thyroxine to transthyretin newly synthesized and secreted by choroid plexus.     

Kinetic stabilization of the native state by protein engineering: implications for inhibition of transthyretin amyloidogenesis

The amyloidogenic homotetrameric protein transthyretin (TTR) must undergo rate-limiting dissociation to partially denatured monomers in order to aggregate. TTR contains two distinct quaternary interfaces, one of which defines the binding sites for thyroxine and small-molecule amyloidogenesis inhibitors. Kinetic stabilization of the tetramer can be accomplished either by the binding of amyloidogenesis inhibitors selectively to the native state over the dissociative transition state or by the introduction of trans-suppressor subunits (T119M) into heterotetramers to destabilize the dissociative transition state. In each case, increasing the dissociation activation barrier prevents tetramer dissociation. Herein, we demonstrate that tethering two subunits whose quaternary interface defines the thyroxine binding site also dramatically increases the barrier for tetramer dissociation, apparently by destabilization of the dissociative transition state. The tethered construct (TTR-L-TTR)2 is structurally and functionally equivalent to wild-type TTR. Urea is unable to denature (TTR-L-TTR)2, yet it is able to maintain the denatured state once denaturation is achieved by GdnHCl treatment, suggesting that (TTR-L-TTR)2 is kinetically rather than thermodynamically stabilized, consistent with the identical wild-type TTR and (TTR-L-TTR)2 GdnHCl denaturation curves. Studies focused on a construct containing a single TTR-L-TTR chain and two normal monomer subunits establish that alteration of only one quaternary structural interface is sufficient to impose kinetic stabilization on the entire quaternary structure.     

Statistical discovery of site inter-dependencies in sub-molecular hierarchical protein structuring

Background

Much progress has been made in understanding the 3D structure of proteins using methods such as NMR and X-ray crystallography. The resulting 3D structures are extremely informative, but do not always reveal which sites and residues within the structure are of special importance. Recently, there are indications that multiple-residue, sub-domain structural relationships within the larger 3D consensus structure of a protein can be inferred from the analysis of the multiple sequence alignment data of a protein family. These intra-dependent clusters of associated sites are used to indicate hierarchical inter-residue relationships within the 3D structure. To reveal the patterns of associations among individual amino acids or sub-domain components within the structure, we apply a k-modes attribute (aligned site) clustering algorithm to the ubiquitin and transthyretin families in order to discover associations among groups of sites within the multiple sequence alignment. We then observe what these associations imply within the 3D structure of these two protein families.

Results

The k-modes site clustering algorithm we developed maximizes the intra-group interdependencies based on a normalized mutual information measure. The clusters formed correspond to sub-structural components or binding and interface locations. Applying this data-directed method to the ubiquitin and transthyretin protein family multiple sequence alignments as a test bed, we located numerous interesting associations of interdependent sites. These clusters were then arranged into cluster tree diagrams which revealed four structural sub-domains within the single domain structure of ubiquitin and a single large sub-domain within transthyretin associated with the interface among transthyretin monomers. In addition, several clusters of mutually interdependent sites were discovered for each protein family, each of which appear to play an important role in the molecular structure and/or function.

Conclusions

Our results demonstrate that the method we present here using a k- modes site clustering algorithm based on interdependency evaluation among sites obtained from a sequence alignment of homologous proteins can provide significant insights into the complex, hierarchical inter-residue structural relationships within the 3D structure of a protein family.

     

The crystal structure of the transthyretin-like protein from Salmonella dublin, a prokaryote 5-hydroxyisourate hydrolase

The mechanism of binding of thyroid hormones by the transport protein transthyretin (TTR) in vertebrates is structurally well characterised. However, a homologous family of transthyretin-like proteins (TLPs) present in bacteria as well as eukaryotes do not bind thyroid hormones, instead they are postulated to perform a role in the purine degradation pathway and function as 5-hydroxyisourate hydrolases. Here we describe the 2.5 Angstroms X-ray crystal structure of the TLP from the Gram-negative bacterium Salmonella dublin, and compare and contrast its structure with vertebrate TTRs. The overall architecture of the homotetramer is conserved and, despite low sequence homology with vertebrate TTRs, structural differences within the monomer are restricted to flexible loop regions. However, sequence variation at the dimer-dimer interface has profound consequences for the ligand binding site and provides a structural rationalisation for the absence of thyroid hormone binding affinity in bacterial TLPs: the deep, negatively charged thyroxine-binding pocket that characterises vertebrate TTR contrasts with a shallow and elongated, positively charged cleft in S. dublin TLP. We have demonstrated that Sdu_TLP is a 5-hydroxyisourate hydrolase. Furthermore, using site-directed mutagenesis, we have identified three conserved residues located in this cleft that are critical to the enzyme activity. Together our data reveal that the active site of Sdu_TLP corresponds to the thyroxine binding site in TTRs.     

Specific labeling of the thyroxine binding site in thyroxine-binding globulin: determination of the amino acid composition of a labeled peptide fragment isolated from a proteolytic digest of the derivatized protein

[125I] Thyroxine has been covalently bound to the thyroxine binding site in thyroxine-binding globulin by reaction with the bifunctional reagent, 1,5-difluoro-2,4-dinitrobenzene. An average of 0.47 mol of [125I] thyroxine was incorporated per mol protein; nonspecific binding amounted to 8%. A labeled peptide fragment was isolated from a proteolytic digest of the derivatized protein by HPLC and its amino acid composition was determined. Comparison with the amino acid sequence of thyroxine-binding globulin indicated partial correspondence of the labeled peptide with two possible regions in the protein. These regions also coincide with part of the barrel structure present in the closely homologous protein, alpha 1-antitrypsin.     

Receptor-mediated uptake and internalization of transthyretin

Evidence of cellular transthyretin (TTR) binding was sought because of the observation that transthyretin can increase the uptake of its hormonal ligand. Transthyretin was bound by human hepatoma (Hep G2) cells in a time- and temperature-dependent manner, reaching equilibrium within 2 h. Scatchard analysis was consistent with a single class of high affinity binding sites with a Kd of approximately 5 nM at 0 and 4 degrees C and 14 nM at 37 degrees C. These dissociation constants are more than 2 orders of magnitude lower than the concentration of transthyretin in human serum. The apparent capacity at 0 degrees C, corrected for internalized TTR, was approximately 20,000 sites/cell. Saturable, high affinity binding of human transthyretin was also demonstrable with rat primary hepatocytes and human renal adenocarcinoma, neuroblastoma, and transformed lung cells. Rat and human transthyretin were equipotent in displacing isotopically labeled, species-specific transthyretin from human hepatoma cells and rat primary hepatocytes, a finding that is consistent with the strong homology between rat and human transthyretin. Eighty-eight percent of the saturable uptake was internalized as determined by proteolytic removal of surface transthyretin. Internalization was dependent on receptor binding and was more markedly inhibited than surface binding at 0 degrees C. Concentrations of thyroxine within a range that saturated a significant proportion of the primary and secondary TTR iodothyronine binding sites increased the uptake and internalization of transthyretin in a dose-dependent manner. By analogy to the function of receptors for other transport proteins, the interaction between transthyretin and its receptor is likely to affect ligand delivery and may have additional metabolic effects.     

Rat transthyretin (prealbumin). Molecular cloning, nucleotide sequence, and gene expression in liver and brain

Transthyretin cDNA was isolated from a rat liver cDNA library. Analysis of the nucleotide sequence revealed a signal peptide-like sequence preceding a section coding for a full length subunit and an untranslated sequence at the 3' end. The deduced primary structure of rat transthyretin was compared with that of human transthyretin. It was highly conserved at the binding sites for thyroxine and the interfaces and core regions of the subunits. The cDNA for transthyretin was used to measure mRNA levels by hybridization. During acute inflammation, the amount of transthyretin mRNA in liver decreased (reaching a minimum of 25% of the normal level 36 h after inducing inflammation), suggesting regulation of transthyretin synthesis at the mRNA level. Transthyretin mRNA was found only in the liver and in the choroid plexus, but not in other parts of the central nervous system nor in the adrenal glands, kidney, spleen, testes, heart, lung, intestine, and ovaries. One gram of choroid plexus contained about 25 times larger amounts of transthyretin mRNA than 1 g of liver. By synthesizing an important hormone carrier protein, the choroid plexus may be an important link in the chemical communication between the central nervous system and the bloodstream.     

Fluorescence study of the thyroxine-dependent conformational changes in human serum transthyretin.

Fluorescence studies of transthyretin (TTR) were conducted to detect structural changes associated with the environment of its two tryptophans, induced by binding of thyroxine (T4). Non-radiative tryptophans relaxation rate has an activation energy of 6.4 kcal/mol for TTR, which is decreased to 4.4 kcal/mol for TTR-T4 complex. The maximum fluorescence wavelength was red-shifted as the excitation wavelength was increased. T4 changed the magnitude of this shift. T4 binding per se changed the emission maximum reflecting different environments of the tryptophans. Double-quenching experiments also showed that T4 produces changes in the tryptophans environments. These findings were interpreted as the result of structural alterations in the protein matrix induced by T4 which contribute in part to explain the negative cooperativity associated with the occupancy of the second binding site.     

Linkage between the hormone binding site and the reactive center loop of thyroxine binding globulin

Thyroxine binding globulin (TBG) is the major carrier of the thyroid hormones triiodothyronine (T3) and thyroxine (T4) in plasma. TBG is member of the serpin family of proteins although it has no proteinase inhibitory activity. In this study we show that TBG has properties typical of a metastable serpin and provide evidence that occupancy of the hormone binding site alters the conformation of the reactive center loop. After reactive center loop cleavage by endoproteinase Asp-N or neutrophil elastase the protein became more stable to guanidine hydrochloride denaturation compared to the native protein, as a result of loop insertion. In addition, incubation of the native protein with a reactive center loop peptide, caused a change in mobility on a native gel. This is consistent with the idea that thyroxine binding globulin is able to form a binary complex with the peptide as a result of beta-sheet A expansion. To assess the effect of cleavage and loop insertion on the hormone binding site we used the specific binding of a fluorophore, 1,8-anilinonaphthalene sulfonic acid (ANS). Loop insertion itself had no effect on ANS affinity, but cleavage with elastase at the P4'-P5' bond caused a reduction in affinity, presumably because this cleavage site is located within the hormone binding site. These data support the concept that cleavage of TBG by proteinases released in inflammation is a mechanism to deliver thyroid hormones to target tissues. A linkage between the occupancy state of the hormone binding site and the conformation of the reactive center loop was indicated by the observation that binding of T3 to native TBG reduced proteolytic susceptibility by both endoproteinase Asp-N and elastase.     

Interaction of chlorinated phenols with thyroxine binding sites of human transthyretin, albumin and thyroid binding globulin

Previous results (Brouwer and van den Berg, Toxicol. Appl. Pharmacol., 85 (1986) 301) indicated preferential binding of a hydroxylated metabolite of tetrachlorobiphenyl to transthyretin (TTR) a carrier of thyroxine (T4). In the present study it was investigated whether the T4 binding site of TTR could be occupied specifically by hydroxylated chlorinated aromatic compounds using chlorinated phenol congeners as model compounds in a competition assay with [125I]T4. Chlorinated aromatics such as 2,3-dichlorobenzene and 3,4,3',4'-tetrachlorobiphenyl, and phenols such as 4-hydroxybiphenyl and phenol were inefficient competitors. All chlorinated phenols tested were competitors for the T4 binding site of TTR. The ranking in competition was pentachlorophenol (PCP) greater than trichlorophenols greater than dichlorophenols greater than monochlorophenols. Structures with chlorine in both ortho positions to the hydroxyl group were more efficient competitors. The relative affinity of binding of pentachlorophenol (PCP) to TTR was about twice that of T4. Scatchard analysis showed that PCP mainly decreased the affinity constant K11 while the binding capacity R1 was not altered, indicating a competitive type of inhibition. PCP was also able to compete with T4 sites on albumin with a relative affinity of 0.25. T4 binding to thyroid binding globulin (TBG) was much less affected by interference of PCP (relative affinity 0.001). The results indicate a specific interaction of chlorophenols with the T4 binding site of TTR.     

Interaction of arylpiperazine ligands with the hydrophobic part of the 5-HT1A receptor binding site

A flexible docking of a series of arylpiperazine derivatives with structurally different aryl part to the binding site of a model of human 5-HT1A receptor was exercised. The influence of structure and hydrophobic properties of aryl moiety on binding affinities was discussed and a model for ligand binding in the hydrophobic part of the binding site was proposed.     

Allosteric modulation of hormone release from thyroxine and corticosteroid-binding globulins

The release of hormones from thyroxine-binding globulin (TBG) and corticosteroid-binding globulin (CBG) is regulated by movement of the reactive center loop in and out of the β-sheet A of the molecule. To investigate how these changes are transmitted to the hormone-binding site, we developed a sensitive assay using a synthesized thyroxine fluorophore and solved the crystal structures of reactive loop cleaved TBG together with its complexes with thyroxine, the thyroxine fluorophores, furosemide, and mefenamic acid. Cleavage of the reactive loop results in its complete insertion into the β-sheet A and a substantial but incomplete decrease in binding affinity in both TBG and CBG. We show here that the direct interaction between residue Thr(342) of the reactive loop and Tyr(241) of the hormone binding site contributes to thyroxine binding and release following reactive loop insertion. However, a much larger effect occurs allosterically due to stretching of the connecting loop to the top of the D helix (hD), as confirmed in TBG with shortening of the loop by three residues, making it insensitive to the S-to-R transition. The transmission of the changes in the hD loop to the binding pocket is seen to involve coherent movements in the s2/3B loop linked to the hD loop by Lys(243), which is, in turn, linked to the s4/5B loop, flanking the thyroxine-binding site, by Arg(378). Overall, the coordinated movements of the reactive loop, hD, and the hormone binding site allow the allosteric regulation of hormone release, as with the modulation demonstrated here in response to changes in temperature.