cDNAs for 12 components of the hippo signaling pathway were cloned into into pMK33-CTAP(SG) and stable S2R+ cell lines were generated. Cells were induced overnight with 140uM copper. Experiments were done in triplicate.

Cells were lysed and soluble lysate was incubated with IgG sepharose in the first affinity purification step. Beads were collected, washed and bound proteins eluted by TEV protease cleavage. In a second purification step, TEV-eluted proteins were incubated with Streptavidin Plus resin. Purified proteins were then eluted with 2mM biotin.

Purified proteins were TCA precipitated, trypsin digested in solution, dried and resuspended. Peptides were analysed by LC-MS/MS. MS/MS spectra were matched to peptides using SEQUEST and a composite database of all translated D. melanogaster open reading frames in FlyBase (r5.23).

From twelve baits, 4560 unfiltered interactions among 1518 proteins were identified. SAINT analysis was used to assign confidence scores (PMID: 21131968). Six 'tag-only' purifications, and 8 purifications using 'JAK/STAT' pathway components as bait were used as negative controls. In order to test the performance of the SAINT analysis, a positive reference set of 21 interactions involving Hippo pathway proteins were curated from the literature. Additionally, a random reference set of 1000 interactions was randomly selected from DPiM-1 (FBlc0000184) interactions involving any one of 1348 'non-specific' interactors. Non-specific interactors were defined as proteins that pull down with approximately 1000 experiments in the unfiltered DPiM-1 dataset. A SAINT score threshold of ≥ 0.8 was set to define 204 high confidence interactions among 153 proteins.

23 of 26 interactions tested were validated by coimmunoprecipitation experiments.