W700C

Position 700 sits in wolframin's C-terminal lumenal domain (residues 653–869). In the AlphaFold model, W700 is packed against immediate sequence neighbors THR699 (2.5 Å) and THR701 (2.4 Å), and into a distant aromatic-hydrophobic pocket containing PHE825 (3.9 Å) and MET781 (4.8 Å). The wild-type indole ring contributes π-stacking to PHE825 and hydrophobic contact to MET781 — substantial packing density into a defined local environment.

Replacing tryptophan with cysteine here removes the bulky aromatic indole entirely and replaces it with a small thiol. The volume difference is large — roughly 100 ų — but the resulting cavity is partially filled by the surrounding atoms relaxing slightly, and the new free thiol does not introduce charge or strong polarity. This explains the surprisingly small |ΔΔG| of 0.10 kcal/mol: DynaMut2 reports the fold is essentially unperturbed.

The contrast with W700S (same position, serine substitution) is the structural lesson. W700S has |ΔΔG| of 2.49 kcal/mol — Category 2, moderately destabilizing — because the polar hydroxyl introduces unfavorable contacts in the hydrophobic pocket. W700C, with a thiol instead of a hydroxyl, sits closer to the chemistry the pocket can tolerate. Yet both are pathogenic, with W700C carrying the stronger 'Pathogenic' ClinVar classification.

The inference: W700's pathogenic mechanism is not primarily structural destabilization. The lost interaction — likely the π-stacking with PHE825 — is functional rather than fold-critical. The indole at W700 probably mediates a specific protein-protein interaction surface or a fold-locking contact whose loss is functionally severe even when the fold itself remains intact. Additionally, the newly introduced free thiol in the ER lumen's oxidative environment may form aberrant disulfide bonds with nearby cysteines (such as C673 or C690 in the same domain), introducing a misfolding pressure that DynaMut2's predicted ΔΔG cannot fully capture.