Abstract
Cataract, the leading cause of blindness worldwide, results from age-related misfolding and aggregation of long-lived crystallin proteins in the eye lens. The cytoplasm of fiber cells in the lens core becomes increasingly oxidizing with age, allowing non-native disulfides to drive light-scattering aggregation of γ-crystallins. Despite this vulnerability to non-native disulfides, and despite lacking any native-state disulfides, γ-crystallins are unexpectedly Cys-rich. To understand this paradox, we investigated how replacing all four Cys residues in the aggregation-prone N-terminal domain of γD-crystallin affects its stability and aggregation. Cys removal precludes the disulfide-driven aggregation pathway we reported previously. Here, we characterize two full-length human γD-crystallin variants: C18S/C32S/C41S/C78S ("NCS") and C18T/C32A/C41A/C78A ("NCA/T"). Thermodynamic and kinetic stability measurements indicate the N-terminal domain was greatly destabilized in both variants relative to WT, with NCS more destabilized than NCA/T. Upon mild heating or partial denaturation, both variants formed light-scattering aggregates, which were amorphous by transmission electron microscopy. Surprisingly, the aggregation proceeded exclusively from a native-state dimer held together by a C-terminal disulfide bridge. Aggregation was strongly suppressed by the lens's native chemical chaperone, myo -inositol. The aggregation rate depended linearly on protein concentration, indicating that the rate limiting step was a transformation of the native-state dimer to a misfolded dimer. The native-state dimer forms readily even in the WT protein, and evidence of it has been found in the lens. We propose that many age-related chemical modifications could destabilize the native fold of human γD-crystallin, favor misfolding within its native-state dimer, and thereby cause aggregation.