A Drosophila Disease-Model for Transthyretin-associated Amyloidosis

Sammanfattning: Amyloidoses comprise a group of gain-of-toxic function protein misfolding diseases, in which normally soluble proteins in their functional state undergo conformational changes into highly organized and generally intractable thread-like aggregates, termed amyloid fibrils. These structures accumulate predominantly in the extracellular space but growing evidence suggests that amyloids may start to form intracellularly. At least 26 different human proteins, intact or in fragmented form, are known to form amyloid, which is linked with many debilitating neurodegenerative diseases such as Alzheimer’s disease (AD), Creutzfeldt-Jakob disease, and transthyretin (TTR)-related amyloidosis (ATTR).In this work, we focus on ATTR, which is one of the most frequent systemic amyloid diseases. A functional link was established between hereditary ATTR, a severe and fatal disorder and the enhanced propensity of the human plasma protein transthyretin (TTR) to form aggregates, caused by single point mutations in the TTR gene. The disease is heterogeneous and clinical symptoms vary from cardiomyopathy to progressing sensorimotor polyneuropathy dependingon TTR variant involved and the amyloid deposition site. Despite the fact that TTR-derived amyloid accumulates in different organs such as heart, kidney, eyes, and predominantly in the peripheral nerves of ATTR patients, the exact mechanism of the disease development is not understood. In contrast to the case of AD, it has been difficult to generate an animal model for ATTR in transgenic mice that would be useful in understanding TTR aggregation processes and the mechanisms of the associated toxicity as these mice did not develop any neuropathic phenotype besides amyloid deposits. Therefore, we created a disease-model in Drosophila due to its huge repertoire of genetic techniques and easy genotype – phenotype translation, as well as its success in modeling human neurodegeneration.We have generated transgenic flies that over-express the clinical amyloidogenic variant TTRL55P, the engineered variant TTR-A (TTRV14N ? V16E), and the wild-type protein. All TTR variants were found in the secreted form in the hemolymph where misfolding occurred and depending on the pool of toxic species, the fate of the fly was decided. Within a few weeks, both mutants (but not the wild-type TTR) demonstrated a time-dependent aggregation of misfolded molecules in vivo. This was associated with neurodegeneration, change in wing posture, attenuation of locomotor activity including compromised flying ability, and shortened life span. In contrast, expression of the wild-type TTR had no discernible effect on either longevity or fly behavior.In this work, we also addressed the correlation between TTR transgene dosage and thus, protein levels, with the severity of the phenotypes observed in TTR-A flies which developed a “dragged wing” phenotype. Remarkably, we established that degenerative changes such as damage to the retina strictly correlated with increased levels of mutated TTR but inversely with behavioral alterations and the dragged wing phenotype. We characterized formation of aggregates in the form of 20 nm spherules and amyloid filaments intracellularly in the thoracic adipose tissue and brain glia (both tissues that do not express the transgene). Moreover, we detected a fraction of neurotoxic TTR-A in the hemolymph of young but not old flies. We proposedthat these animals counteract formation and persistence of toxic TTR-A species by removal from the circulation into intracellular compartments of glial and fat body cells and this is part of a mechanism that neutralizes the toxic effects of TTR.We validated the fly model for ATTR by applying a genetic screen during study of modifier genes. We found Serum amyloid P component (a product of the APCS gene) as a potent modifier of TTR amyloid-induced toxicity that was effective in preventing the apoptotic response in cell culture assay and capable of reducing the dragged wings when co-expressed in TTR-A flies.Finally, we optimized this fly model in order to screen for therapeutic compounds effective against ATTR. Feeding assays showed the effectiveness of several compounds among known native-state kinetic stabilizers of TTR against its aggregation. We described several early endpoints in this model, which can be used as a rapid and cost-effective method for optimizing concentrations and pre-screening of drug candidates. As the proof of principle, by feeding flies with increasing doses of diflunisal analogue (an FDA-approved Non-Steroidal Anti-Inflammatory Drug) a dose-dependent reduction of the dragged wings was observed.