Abstract
INTRODUCTION: Systemic AL amyloidosis is caused by deposition of monoclonal antibody light chains (LC) as insoluble amyloid fibrils in multiple tissues, leading to irreversible and eventually fatal organ damage. Each patient has a unique LC sequence that appears to define its propensity to aggregate. The complexity and diversity of LC sequences has impeded efforts to understand why some LCs aggregate to cause disease while others do not. METHODS: We investigated residue changes, relative to the inferred precursor germline sequences, in monoclonal LCs associated with AL amyloidosis and multiple myeloma (MM), derived from the AL-Base resource. Consensus matrices, calculated using healthy polyclonal repertoire sequences from Observed Antibody Space (OAS), were used to determine the relative frequency of each residue in the monoclonal LC sequences. RESULTS: A subset of residues observed in AL-associated LCs was uncommon in the healthy repertoire, but these residues were highly diverse and were also observed in MM-associated LCs. We identified multiple positions that more frequently harbor uncommon residues in AL-associated LCs than OAS-derived LCs, including several positions that have previously been identified. However, each individual residue change occurs in only a small fraction of LCs, indicating that many types of residue change can contribute to disease. Furthermore, positions where residue changes occur most frequently were not enriched in amyloidosis-associated residues. DISCUSSION: These data provide a framework for future investigations into sequence determinants of amyloid propensity, supporting efforts towards earlier recognition and diagnosis of AL amyloidosis.