The interaction between FHIP2A and Rab1A is dependent on Rab1A membrane association. (A) Schematic of the FHF complex bound to dynein, highlighting the four alternate FHIP subunits. The specific FHIP isoform incorporated into the complex dictates its cargo specificity. (B) AlphaFold 3 predicted structure of the Rab1A–FHF complex (consisting of dimeric Hook2 fragments [amino acids 620–719], monomeric full-length Fts, and monomeric full-length FHIP2A), with accompanying PAE plot. (C) Coomassie-stained gel showing an in vitro–binding assay using beads coated with GST-Rab1A or GST-Rab1B and purified GFP-FHIP2A or GFP control. Both GTP- and GDP-locked Rab1A/B proteins were used as indicated. (D) Helical wheel plot of the N-terminal 17 residues of FHIP2A with hydrophobic residues (yellow) clustered on one face forming an amphipathic helix. (E) GUV-binding assay using GTP- or GDP-locked Rab1A on the GUV, and GFP-FHIP2A. GUVs depicted in magenta and GFP-FHIP2A in cyan. Each large datapoint in the graph depicts the average mean fluorescence intensity of GFP-FHIP2A on a selection of GUV membrane and represents an independent experiment (n = 3), with smaller gray datapoints representing all the technical replicates (AU, arbitrary units). The mean ± SD is shown. **P < 0.01; ***P < 0.001; ****P < 0.0001 (one-way ANOVA with Tukey’s multiple comparisons test). (F) Representative immunoblot of MitoID in HEK293A cells where Rab1A MitoID constructs (detected using anti HA) and 3xFlag-FHIP2A proteins (detected using anti Flag) were transiently expressed. Endogenous CALCOCO1 was used as a positive control. Each datapoint in the graph represents the normalized ratio between the FHIP2A and MitoID construct immunoblot intensities and depicts an independent experiment (n = 3). The mean ± SD is indicated. **P < 0.01; (one-way ANOVA with Dunnett’s multiple comparisons test). Source data are available for this figure: SourceData F2.