Figure 3: Functional and biochemical analysis of PEX19 variants that affect farnesyl recognition.

(a) Hydrophobic interaction chromatography analysis. Proteins were eluted from a Butyl Sepharose FF column using a linear gradient of decreasing (NH₄)₂SO₄ concentrations. For PEX19 CTD wild type and every variant, the farnesylated protein exhibits an increased elution volume compared to the non-farnesylated form (Supplementary Fig. 4d). The plot shows relative differences in hydrophobicity of the farnesylated and non-farnesylated state for each protein, compared to the wild type protein (which was set to 1). (b) Functional complementation of PEX19-deficient fibroblasts by PEX19 variants. PEX19 harbouring single-amino acid substitutions as indicated were introduced into PEX19-deficient fibroblasts by transfection with bicistronic expression vectors coding for full-length PEX19 variants and eGFP-PTS1. Localization of the model peroxisomal protein transport substrate eGFP-PTS1 was monitored by fluorescence microscopy (left column). Endogenous PEX14 as a peroxisomal membrane protein was detected using immunofluorescence microscopy (middle column). A punctate staining pattern and co-localization of eGFP-PTS1 and PEX14 indicate functional peroxisomal protein transport (merge, right column). For better comparison, PEX19-deficient cells expressing only eGFP-PTS1 from the vector are shown in the upper panels. Scale bar: 10 μm. The designation indicated on the right refers to the GFP-localization. (c) Quantitative analysis of ΔPEX19-phenotype complementation by the individual PEX19 variants. Values are obtained from analysing at least 100 cells in three independent transfection experiments. Data shown represent mean±s.d. (d) Immunoblot analysis of the farnesylation mutants shows that all variants are expressed at wild type level and, except for C296S (arrow), are farnesylated in vivo (arrowhead).