Recombinant human retinol-binding protein refolding, native disulfide formation, and characterization.

Human retinol-binding protein (RBP) is a monomeric 21-kDa protein that is currently the subject of numerous studies owing to its role in the cellular uptake and utilization of retinol. When the RBP gene is overexpressed in Escherichia coli, inclusion bodies of aggregated RBP are found in the cells. These inclusion bodies are solubilized in 5.0 M GdmCl containing 10 mM DTT. Refolding of RBP is carried out in the presence of vitamin A by diluting denatured and reduced RBP into a redox refolding buffer consisting of 3 mM cysteine/0.3 mM cystine at 4 degreesC. Ion exchange chromatography (HPLC) is utilized to purify refolded RBP to homogeneity as demonstrated by SDS-PAGE and electrospray MS. The native structure of refolded RBP was established by its ability to bind to vitamin A and the plasma protein transthyretin. The reconstitution of RBP outlined within affords a 50-60% overall yield, i.e., 73 mg of pure RBP/L of E. coli culture.

Transthyretin quaternary and tertiary structural changes facilitate misassembly into amyloid.

Human transthyretin (TTR) can be transformed into amyloid fibrils by partial acid denaturation to yield a monomeric amyloidogenic intermediate that self-associates into amyloid through quaternary structural intermediates, which are identified by sedimentation velocity methods. The monomeric amyloidogenic intermediate has substantial beta-sheet structure with a nonnative but intact tertiary structure as discerned from spectroscopic methods. Proteolysis sensitivity studies suggest that the C-strand-loop-D-strand portion of TTR becomes disordered and moves away from the core of the beta-sandwich fold upon formation of the monomeric amyloidogenic intermediate over the pH range 5.1-3.9. The single site mutations that are associated with early onset amyloid disease [familial amyloid polyneuropathy (FAP)] function by destabilizing tetrameric TTR. Under mild denaturing conditions, the FAP variants populate the monomeric amyloidogenic intermediate conformation, which assembles into amyloid, whereas wild-type TTR remains tetrameric and nonamyloidogenic. The FAP mutations do not significantly alter the native folded structure; instead, they appear to act by making the thermodynamics and perhaps the kinetics more favorable for formation of the amyloidogenic intermediate. Suppressor mutations have also been characterized that strongly stabilize tetrameric TTR and disfavor the formation of the monomeric amyloidogenic intermediate, thus inhibiting amyloid formation. The mechanistic details characterizing transthyretin amyloid fibril formation available from the biophysical studies outlined within have been utilized to develop a new therapeutic strategy for intervention in human amyloid disease. This approach features small molecules that bind with high affinity to the normal fold of transthyretin, inhibiting the quaternary and tertiary structural changes associated with the formation of the monomeric amyloidogenic intermediate that self-assembles into amyloid. Ligand binding to TTR stabilizes the native tetrameric fold, which is nonamyloidogenic.

Inhibiting transthyretin amyloid fibril formation via protein stabilization.

Transthyretin (TTR) amyloid fibril formation is observed systemically in familial amyloid polyneuropathy and senile systemic amyloidosis and appears to be the causative agent in these diseases. Herein, we demonstrate conclusively that thyroxine (10.8 microM) inhibits TTR fibril formation efficiently in vitro and does so by stabilizing the tetramer against dissociation and the subsequent conformational changes required for amyloid fibril formation. In addition, the nonnative ligand 2,4,6-triiodophenol, which binds to TTR with slightly increased affinity also inhibits TTR fibril formation by this mechanism. Sedimentation velocity experiments were employed to show that TTR undergoes dissociation (linked to a conformational change) to form the monomeric amyloidogenic intermediate, which self-assembles into amyloid in the absence, but not in the presence of thyroxine. These results demonstrate the feasibility of using small molecules to stabilize the native fold of a potentially amyloidogenic human protein, thus preventing the conformational changes, which appear to be the common link in several human amyloid diseases. This strategy and the compounds resulting from further development should prove useful for critically evaluating the amyloid hypothesis--i.e., the putative cause-and-effect relationship between TTR amyloid deposition and the onset of familial amyloid polyneuropathy and senile systemic amyloidosis.

FAP mutations destabilize transthyretin facilitating conformational changes required for amyloid formation.

Functional transthyretin (TTR) can be transformed into amyloid by partial acid denaturation yielding a monomeric amyloidogenic intermediate which self-associates. The amyloidogenic intermediate has substantial beta-sheet structure with non-native but defined tertiary structure. pH-dependent proteolysis sensitivity studies have identified portions of TTR which become disordered and solvent-exposed in the amyloidogenic intermediate. These include the C-strand-loop D-strand portion of TTR which moves away from the core of the beta-sandwich fold. Mutations that are associated with early onset-amyloid disease (familial amyloidotic polyneuropathy; FAP) function by destabilizing tetrameric TTR in favour of the monomeric amyloidogenic intermediate which has a rearranged C-strand-loop D-strand region. In most cases the FAP mutations do not significantly alter the native folded structure, but instead act on the denaturation pathway by a mechanism that is not completely understood. Interestingly, mutations have also been characterized which strongly stabilize tetrameric TTR and make amyloid formation very difficult at pHs accessible in vivo.

Comparison of lethal and nonlethal transthyretin variants and their relationship to amyloid disease.

The role that transthyretin (TTR) mutations play in the amyloid disease familial amyloid polyneuropathy (FAP) has been probed by comparing the biophysical properties of several TTR variants as a function of pH. We have previously demonstrated that the partial acid denaturation of TTR is sufficient to effect amyloid fibril formation by self-assembly of a denaturation intermediate which appears to be monomeric. Earlier studies on the most pathogenic FAP variant known, Leu-55-Pro, revealed that this variant is much less stable toward acid denaturation than wild-type TTR, apparently explaining why this variant can form amyloid fibrils under mildly acidic conditions where wild-type TTR remains nonamyloidogenic. The hypothesis that FAP mutations destabilize the TTR tetramer in favor of a monomeric amyloidogenic intermediate under lysosomal (acidic) conditions is further supported by the data described here. We compare the acid stability and amyloidogenicity of the most prevalent FAP variant, Val-30-Met, along with the double mutant, Val-30-Met/Thr-119-Met, which serves to model the effects of these mutations in heterozygous patients where the mutations are in different subunits. In addition, we have characterized the Thr-119-Met TTR variant, which is a common nonpathogenic variant in the Portuguese population, to further investigate the role that this mutation plays in protecting individuals who also carry the Val-30-Met mutation against the classically severe FAP pathology. This biophysical study demonstrates that Val-30-Met TTR is significantly less stable toward acid denaturation and more amyloidogenic than wild-type TTR, which in turn is less stable and more amyloidogenic than Thr-119-Met TTR. Interestingly, the double mutant Val-30-Met/Thr-119-Met is very similar to wild-type TTR in terms of its stability toward acid denaturation and its amyloidogenicity. The data suggest that the Thr-119-Met mutation confers decreased amyloidogenicity by stabilizing tetrameric TTR toward acid denaturation. In addition, fluorescence studies monitoring the acid-mediated denaturation pathways of several TTR variants reveal that the majority exhibit a plateau in the relative fluorescence intensity over the amyloid-forming pH range, i.e., ca. pH 4.3-3.3. This intensity plateau suggests that the amyloidogenic intermediate(s) is (are) being observed over this pH range. The Thr-119-Met variant does not exhibit this plateau presumably because the amyloidogenic intermediate(s) do(es) not build up in concentration in this variant. The intermediate is undoubtedly forming in the Thr-119-Met variant, as it will form amyloid fibrils at high concentrations; however, the intermediate is only present at a low steady-state concentration which makes it difficult to detect.