 |
Home | |
|
Jeffery W. Kelly | |
|
Group Members | |
 |
Research Projects | |
|
Publications | |
|
Group Meeting Schedule | |
|
Links to Other Websites | |
|
Contact | |
 |
||
Transthyretin is a 55 kDa homotetrameric protein (Figure 1) that transports L-thyroxine and holo-retinol binding protein in the serum and cerebrospinal fluid of humans. We discovered that conformational changes alone enable transthyretin aggregation. Transthyretin amyloid formation in vivo is initiated by dissociation of the native tetramer under a denaturation stress of unknown origin. Subsequently, the resulting monomers can partially denature and misassemble into amyloid fibrils and other amorphous aggregates. Aggregation by transthyretin monomers under acidic conditions (conditions that render transthyretin amyloidogenesis fast on a laboratory time scale) occurs via a downhill polymerization mechanism, which means that every step along the amyloid formation pathway is energetically favorable and fast relative to tetramer dissociation. Thus, tetramer dissociation is rate limiting for transthyretin amyloid formation in the case of the wild type protein and for the vast majority of mutants.
Stabilizing the native tetrameric state of transthyretin should ameliorate transthyretin amyloid diseases. This hypothesis is supported by the observation of trans-suppression, in which compound heterozygotes expressing both a disease-associated mutation (e.g., V30M) and a trans-suppressor mutation (e.g., T119M) do not develop transthyretin amyloid disease. We have shown that the trans-suppressor mutation T119M inhibits amyloid formation by kinetically stabilizing (i.e., dramatically slowing the dissociation of) mixed transthyretin tetramers.
We have designed numerous small molecules based on the crystal structures of transthyretin that are now established to avidly bind to the unoccupied L-thyroxine binding sites within transthyretin. We have shown that small molecule binding to these sites inhibits amyloid formation in vitro by selectively stabilizing the native tetrameric state over the dissociative transition state, thus raising the energetic barrier for tetramer dissociation, dramatically slowing the rate-limiting step in the aggregation pathway. Over the past decade, we have identified and synthesized over 1000 small molecule inhibitors of transthyretin amyloid formation that group into half a dozen distinct families. The inhibitors are typically composed of two differentially-substituted aromatic rings connected by linkers of variable chemical composition. These small molecule kinetic stabilizers either just bind to transthyretin or bind and then react chemoselectively with only one of eight lysine ε-amino groups within transthyretin. In collaboration with Dr. Ian Wilson (The Scripps Research Institute, Department of Molecular Biology), we systematically ranked a myriad of possibilities for the three substructures composing a typical transthyretin kinetic stabilizer. We used these data in a substructure combination strategy to develop very potent and selective transthyretin kinetic stabilizers. In collaboration with Dr. Joel Buxbaum (The Scripps Research Institute, Department of Molecular and Experimental Medicine), these and other compounds are being tested in cell and mouse models. These data should allow us to be able to predict the structures of potent and selective transthyretin amyloidogenesis inhibitors that are largely devoid of characteristics undesirable for a clinical candidate.
In a recently completed placebo-controlled, double-blind clinical trial carried out by FoldRx Pharmaceuticals (a company that Kelly cofounded), benzoxazoles discovered by the Kelly Laboratory and developed by FoldRx Pharmaceuticals proved safe and effective at halting the progression of familial amyloid polyneuropathy by a myriad of metrics. This transthyretin kinetic stabilizer, now named Tafamidis, provides the first pharmacological evidence that the process of amyloid fibril formation causes the transthyretin amyloid diseases—reinvigorating efforts of other investigators and companies to do the same in other amyloid diseases. In collaboration with Dr. Martha Skinner and Dr. John Berk at Boston University, another kinetic stabilizer discovered by the Kelly Laboratory, diflunisal, is being tested in a second placebo-controlled human clinical trial for familial amyloid polyneuropathy that is currently enrolling patients.
In collaboration with Dr. Bill Balch (The Scripps Research Institute, Department of Cell Biology) and Dr. Luke Wiseman (The Scripps Research Institute, Department of Molecular and Experimental Medicine), we investigated the relationship between the secretion of destabilized transthyretin variants and the pathology of the disease they cause. The earliest onset (onset at 20-30 years of age) and most severe transthyretin amyloid diseases are generally associated with mutations that strongly destabilize the transthyretin tetramer, resulting in facile transthyretin dissociation and misfolded monomer misassembly into aggregates in the peripheral nerves. However, the most destabilized variants characterized to date, A25T and D18G transthyretin, do not cause an early onset systemic amyloid disease because they are intercepted by the degradation component of the proteostasis network within liver cells—the liver is where most of the transthyretin in the blood plasma is produced. Because of endoplasmic reticulum-associated degradation of these highly destabilized transthyretin variants within the secretory pathway of liver cells, the concentration of the destabilized mutant transthyretin in blood plasma would not be high enough to enable the amyloidogenesis responsible for pathology. Interestingly, these mutants lead to a very rare brain disease, with an intermediate age of onset (40-50 years old), because the choroid plexus is more permissive in its ability to secrete highly destabilized transthyretin variants for reasons that we are seeking to understand. These results suggest that endoplasmic reticulum-assisted folding mediated by the proteostasis network determines protein secretion in a tissue-specific manner, and we propose that its competition with endoplasmic reticulum-associated degradation may explain the appearance of tissue-selective amyloid diseases.
We are also interested in determining the mechanism of in trans proteotoxicity in the transthyretin amyloidoses, understanding how transthyretin fibrils are cleared from normal subjects, discerning the reason why only a subset of individuals who have predisposing mutations get transthyretin amyloid diseases, and we seek to understand the basis for tissue-selective degeneration in the transthyretin amyloid diseases. Towards these ends, we are developing C. elegans transthyretin amyloidosis models expressing a range of mutants associated with human peripheral neuropathy, cardiomyopathy or central nervous system selective amyloidosis. These models will be used to investigate the mechanism of proteotoxicity, particularly that associated with the aging process, and will enable us to test the efficacy of our transthyretin kinetic stabilizers and novel therapeutic strategies in an organismal model of the transthyretin amyloidoses having relevant phenotypes.
