Spatial covariance reveals isothiocyanate natural products adjust redox stress to restore function in alpha-1-antitrypsin deficiency

Abstract

Alpha-1 antitrypsin (AAT) deficiency (AATD) is a monogenic disease caused by misfolding of AAT variants resulting in gain-of-toxic aggregation in the liver and loss of monomer activity in the lung leading to chronic obstructive pulmonary disease (COPD). Using high-throughput screening, we discovered a bioactive natural product, phenethyl isothiocyanate (PEITC), highly enriched in cruciferous vegetables, including watercress and broccoli, which improves the level of monomer secretion and neutrophil elastase (NE) inhibitory activity of AAT-Z through the endoplasmic reticulum (ER) redox sensor protein disulfide isomerase (PDI) A4 (PDIA4). The intracellular polymer burden of AAT-Z can be managed by combination treatment of PEITC and an autophagy activator. Using Gaussian process (GP)-based spatial covariance (SCV) (GP-SCV) machine learning to map on a residue-by-residue basis at atomic resolution all variants in the worldwide AATD clinical population, we reveal a global rescue of monomer secretion and NE inhibitory activity for most variants triggering disease. We present a proof of concept that GP-SCV mapping of restoration of AAT variant function serves as a standard model to discover natural products such as the anti-oxidant PEITC that could potentially impact the redox/inflammatory environment of the ER to provide a nutraceutical approach to help minimize disease in AATD patients.

Keywords: AATD; COPD; Gaussian process machine learning; PDIA4; alpha-1 antitrypsin deficiency; chronic obstructive pulmonary disease; genetic variation; isothiocyanate; nutraceutical; oxidative stress; protein aggregation; protein disulfide isomerase A4; proteostasis.