Charting the structure-sequence landscape of light chain amyloids.

Area of Science:

• Biochemistry
• Structural Biology
• Computational Biology

Background:

• Light chain amyloidosis involves misfolded antibody light chains (LCs) forming toxic amyloid fibrils, causing organ damage.
• While LC overproduction is common, only some individuals develop the disease, indicating specific LC properties drive amyloid formation.
• Understanding LC fibril formation is limited by data scarcity and challenges in studying amyloid structures.

Purpose of the Study:

• To develop a computational tool for comparing unknown LC sequences with known structures.
• To predict which LCs are likely to form novel amyloid conformations.
• To guide experimental investigations into light chain amyloidosis.

Main Methods:

• Utilized Hidden Markov Models (HMMs) to generate similarity scores between LC sequences and Protein Data Bank (PDB) templates.
• Modeled LC amyloid structures based on their similarity to known templates.
• Made the HMMSTUFF tool available as a PyPI package and open-source code.

Main Results:

• HMMSTUFF generates similarity scores, enabling the modeling of LC amyloid structures.
• The tool expands understanding of LC amyloid fibril conformations.
• It also identifies gaps in the current knowledge of LC structural diversity.

Conclusions:

• HMMSTUFF provides a method to analyze and predict LC amyloid structures.
• The tool aids in understanding the structural basis of light chain amyloidosis.
• It facilitates the exploration of novel amyloid conformations and highlights areas for future research.