Progress in transthyretin fibrillogenesis research strengthens the amyloid hypothesis.

T amyloidoses are diseases of protein conformation, in which a particular soluble innocuous protein transforms and aggregates into an insoluble fibrillar structure that deposits in extracellular spaces of specific organs (reviewed in refs. 1–4). Organ dysfunction accompanies fibrillar deposition, and the amyloid hypothesis proposes a cause and effect relationship between deposition and dysfunction (reviewed in refs. 4–7). Transthyretin (TTR) is one of 20 proteins that are known currently to form fibrillar deposits in human amyloidoses (1, 8). TTR is a tetrameric protein comprised of four identical subunits that contain two antiparallel -sheets packed together in a -sandwich architecture. It is synthesized primarily in the liver and is the primary transport protein for thyroid hormone (L-thyroxine or T4) in cerebrospinal f luid and the secondary L-thyroxine transporter in blood plasma. With the assistance of the retinol binding protein (RBP), TTR also transports vitamin A (all-trans-retinol) in plasma. RBP bound to vitamin A forms multimolecular complexes with TTR, and under physiological conditions, dissociation of vitamin A causes the complexes to disassemble. Both natural sequence TTR and mutated variants of TTR are involved in amyloid disease. In certain elderly individuals, natural sequence TTR is known to transform into amyloid fibrils that deposit in cardiac and other tissues, giving rise to the condition known as senile systemic amyloidosis. The occurrence of mutations in TTR accelerates the process of TTR fibrillogenesis and is the most important risk factor for TTR amyloidosis. Whereas deposition of amyloid fibrils of variant TTR in cardiac tissue produces the condition familial amyloidotic cardiomyopathy, deposition in peripheral nerve tissue produces familial amyloid polyneuropathy. There are more than 80 TTR variants that are known currently to give rise to TTR amyloidoses (8). The involvement of TTR in the pathology of amyloid disease is well established; however, the cause and effect relationship between TTR amyloid deposition and organ dysfunction has not yet been proven. In three papers published recently (9–11), one of which appears in this issue of PNAS (9), Jeffery Kelly and his colleagues from The Scripps Research Institute report results from in vitro studies that provide a biophysical explanation of how diseaseassociated mutations in TTR affect the course of TTR amyloidoses, and thus strengthen the hypothesis that amyloid fibril deposition is the causative agent in these diseases. Although TTR amyloid deposition is known to occur in extracellular spaces, fibrillogenesis may initiate in acidic environments of endosomes or lysosomes (12, 13). In 1992, Colon and Kelly (12) reported that incubation of TTR in low-pH environments is all that is required to initiate the fibrillogenesis reaction. Since then the acid-induced denaturation fibrillogenesis pathway of TTR has been mapped out in great detail with a variety of biophysical and biochemical techniques (Fig. 1). Under mild acidic conditions (pH 5.75), tetrameric wild-type TTR can be induced to partially dissociate into monomers by dilution (14, 15). Hydrogen-deuterium exchange experiments indicate that the dissociated monomers at pH 5.75 retain a native-like structure that is present in the tetramer (15). Further acidification to pH 4.5 induces greater monomer formation; however, the structures of the pH 4.5 monomers show conformational changes indicative of partial unfolding (12, 14, 16). Hydrogen-deuterium exchange experiments indicate that the conformational instability is localized to 13 residues of the CBEF -sheet of TTR; the other -sheet (DAGH) is stable and shows similar protection from hydrogen-deuterium exchange as tetrameric TTR (15). Amyloid fibril formation ensues at this moderately low pH if the temperature and protein concentration are sufficiently high (12, 14, 16). Further reductions in pH reduce the rate of fibril formation and favor the formation of molten globule-like aciddenatured states (A-states) that form low molecular weight aggregates that are not amyloid fibrils (14, 16). The partially denatured monomeric state of TTR that is populated at pH 4.5 appears to be the critical precursor to amyloid fibril formation, and it has been named the amyloidogenic intermediate. The presence of mutations in TTR associated with amyloidosis greatly affects the acid-induced denaturation fibrillogenesis pathway (13, 14, 16), the major effect being an increased tendency to form the amyloidogenic intermediate at higher pH values. In their most recent work on TTR amyloidosis, Kelly and colleagues investigate the V122I variant of TTR (9). This variant is the most common TTR mutation producing familial amyloidotic cardiomyopathy (17). It originated in West Africa and is carried by 3.9% of African Americans and 5% of individuals in some areas of West Africa. The major effect of this mutation, which in chemical terms represents the addition of a single methylene group, is to destabilize the quaternary structure of TTR while leaving the tertiary structure unaffected. Demonstration of this differential destabilization required the engineering of a TTR variant that was monomeric but otherwise behaved similarly to natural sequence TTR. Two conservative mutations in the subunit interfaces proved sufficient to generate a normally folded monomeric variant of TTR (18). The x-ray crystal structure of monomeric TTR revealed that it had near-identical tertiary structure