Identification of structural determinants of Iight chain amyloidosis

Details

Request a copy
Serval ID
serval:BIB_3C9116AAC4B7
Type
PhD thesis: a PhD thesis.
Collection
Publications
Institution
Title
Identification of structural determinants of Iight chain amyloidosis
Author(s)
Garofalo Maura
Director(s)
Michielin Olivier
Codirector(s)
Cavalli Andrea
Institution details
Université de Lausanne, Faculté de biologie et médecine
Publication state
Accepted
Issued date
2020
Language
english
Abstract
ln systemic immunoglobulin light chain amyloidosis (AL) disease, pathogenic immunoglobulin light chains (LCs) form toxic species and amyloid fibrils in target tissues, leading to organ failure and death. Prompt diagnosis is crucial, to avoid permanent organ damage, however, delays are common with consequent high mortality rates, as symptoms usually appear only after strong organ involvement. Predicting the onset of AL is highly challenging as each patient carries a different pathogenic LC, which is generated by genetic rearrangement and by a unique set of somatic mutations acquired during B cell affinity maturation. Due to such disease complexity, the molecular mechanism of AL amyloidosis and the determinants of LCs proteotoxicity, still need to be uncovered, further complicating the disease diagnosis. Consequently, the development of specific prediction tools would be a crucial step to anticipate AL diagnosis and improve patients' prognosis.
In this thesis, we aimed at identifying determinants of AL amyloidosis analyzing LC protein sequences, starting from the hypothesis that somatic mutations may be an important driving force in the development of the disease and may be exploited as discriminative factors to classify LCs according to their clinical phenotype . To this aim, we developed LICTOR (Light Chain TOxicity predictoR), a machine leaming approach predicting LC toxicity in AL, starting from LC protein sequences. LICTOR uses somatic mutations, exploited in sequence and structural features, to automatically classify previously unseen LC sequences as either taxie or non-taxie in AL. LICTOR achieves a specificity and a sensitivity of 0.82 and 0.76, respectively, with an AUC of0.87, making it a valuable tool for early AL diagnosis. Taking advantage of LICTOR, we in-silico reverted the toxic phenotype of a LC performing only two mutations. These data were validated in a well-established Caenorhabditis elegans in-vivo model used to evaluate LC toxicity.
In conclusion, LICTOR represents an unprecedented method able to accurately predict LC toxicity in AL. Hence, LICTOR may allow a timely identification of high-risk patients, paving the way for early treatment and higher survival rates.

Create date
23/02/2021 9:44
Last modification date
24/02/2021 6:26
Usage data