SUPPLEMENTARY FIGURE LEGENDS

Supplementary Figure S1. Transcript analysis of empty pSPL3 vector. RT-PCR fragment resulting from transfection of HeLa and COS-7 cells with empty pSPL3 (confirmed by sequencing).

Supplementary Figure S2. Antisense oligonucleotides sequence and hybridization site. Both LNAs were targeted to the cryptic splice site, sterically blocking its access to the spliceosome. LNAs present slight differences in sequence coverage, with LNA1 distancing 10 nucleotides from the canonical GT and LNA2 distancing only 5 nucleotides.

Supplementary Figure S3. Antisense nucleotides do not affect the splicing pattern of wild-type minigene. Transcript analysis of HeLa cells expressing the wild-type minigene and treated with antisense oligonucleotides confirmed that both LNA oligonucleotides (herein depicted as LNA1) did not affect the normal splicing mechanism, at any concentration (herein represented at 0.5 µM). LNA2 showed identical transcript profile, also subsequently confirmed by sequencing.

Supplementary Figure S4. Panel A, immunoblotting analysis of purified recombinant wild-type (lane 1, WT) and variant (lane 2, p.D274Gfs*17) GALT proteins; Panel B, sequence alignment between wild-type and p.D274Gfs*17 obtained with Clustal X; red box highlights the His184-Pro185-His186 active site; orange spheres highlight the conserved mononuclear iron ligands; light blue box highlights the missing residues in the p. D274Gfs*17; Panel C, structural models of wild-type and p. D274Gfs*17 GALT obtained with Swiss-Modeller with the Escherichia coli GALT (PDB code 1GUP) as the structural template; grey cartoon depicts the overlapping structure of wild-type and p. D274Gfs*17, whereas the light blue cartoon depicts the extra residues in wild-type GALT; orange and purple spheres depict respectively the iron and zinc ions in the bacterial enzyme; orange ribbons depict the opposing monomer from the bacterial GALT PDB; figure generated with PyMOL [41].

Supplementary Figure S5. Variant GALT is more prone to aggregation. Biophysical methodologies employed to analyze the structural impact of the mutation c.820+13A>G on the recombinant GALT protein. Panel A, far-UV circular dichroism spectra and thermal denaturation profiles (Inset) for wild-type (a) and p. D274Gfs*17 (b); Panel B, thermal denaturation profiles obtained by differential scanning fluorimetry for wild-type (a) and p. D274Gfs*17 (b); Panel C, thermal denaturation analyzed by dynamic light scattering for wild-type (a) and p. D274Gfs*17 (b); Panel D, aggregation kinetics analyzed by dynamic light scattering at 42°C for wild-type (a) and p. D274Gfs*17 (b).

1