Indeed, DLD bound FDX1 more strongly than FDXR in cells, and this interaction was lost with the D136R/D139R mutants. DLD could also replace FDXR in an FDX1-dependent lipoylation assay in vitro .

But there was a twist. Cells with D136R or D139R mutants were also completely deficient in lipoylation. Yet in vitro these mutants stayed fully functional — supporting LIAS-mediated lipoylation and also binding FDXR/LIAS with similar affinities as WT.

This analysis highlighted two residues — D136 and D139 on FDX1’s α-helix 3.

Mutating them to alanine left cuproptosis intact, but flipping their charge (Asp → Arg) made cells resistant to elesclomol–Cu–induced death.

This let us filter out mutations predicted to disrupt overall folding. We focused on variants that blocked FDX1-driven cuproptosis while predicted to remain structurally intact.

We paired the DMS data with an in silico FoldX analysis, which predicts the mean free-energy change (ΔΔG) for each residue in FDX1.

To find out, we performed deep mutational scanning (DMS), mutating every amino acid in FDX1 and testing its ability to promote cuproptosis in two cell line models.