*3.1. Yeast-Two Hybrid Screenings*

To build the first pair-wise protein–protein interaction database of ZIKV NS proteins encoded by the Brazilian genotype, we performed stringent yeast-two hybrid (Y2H) screenings using ZIKV NS proteins as baits to screen a normalized human brain cDNA library (Figure 1A and Table S1). The Y2H screenings generated a protein–protein interaction network (PIN) ranging from 1 (NS2A) to 56 (NS3) interactions, totaling 99 unique protein hits and 109 bait–prey interactions (Figure 1A,B and Table S2).

To validate our Y2H PIN, a subset of 38 bait–prey interactions (38.4% of the Y2H PIN) was tested for interaction in human HEK293FT cells. All NS coding sequences were cloned in a eukaryotic expression vector in frame with the streptavidin and calmodulin binding peptides (SBP and CBP, respectively). Candidate interactor cDNAs were expressed in frame with glutathione-S-transferase (GST). SBP pulldown assays were performed against GST-tagged preys in HEK293FT cells, and 76.3% of interactions were confirmed (Figure 1C). As some true interactors might not validate in these specific

conditions, Figure 1C retains all interactions, with validated ones indicated. However, further analysis was conducted, retaining only the validated hits.

**Figure 1.** Yeast two-hybrid (Y2H) protein–protein interaction network (PIN). (**A**) Schematic representation of the ZIKA virus (ZIKV) genome and non-structural (NS) constructs used on our protein–protein interaction screenings. Grey boxes represent the GAL4 DNA binding domain (DBD) and the streptavidin binding protein (SBP)–calmodulin binding protein (CBP) tag used on the yeas<sup>t</sup> two-hybrid (Y2H) and tandem affinity purification (TAP) assays, respectively. The number of hits identified by each assay and bait is summarized on the right. UTR, untranslated region. (**B**) Network of the interactions identified by Y2H screens. ZIKV NS nodes are colored in orange, and gray nodes indicate other proteins that were recovered as preys. Edge colors represent results from validation. Edges to all preys are shown and 'not validated' (No), 'validated' (Yes), and 'not tested interactions' (Not tested), and are denoted by blue, purple, and gray edges, respectively. The color legend is depicted on the upper left-hand corner. (**C**) Streptavidin pulldown of TAP-tagged NS constructs from 293FT whole cell lysates followed by Western blotting with the indicated antibodies. The individual percentage of hits validated by bait is depicted on the right. Red dots indicate the expected band size. Red asterisk for Streptavidin pulldown assay.

#### *3.2. Tandem A*ffi*nity Purification Followed by Mass Spectrometry*

To further characterize the NS-mediated protein interactions, we expressed all baits as fusions to SBP and CBP in HEK293FT cells (Figure 1A). We then performed tandem affinity purification coupled to mass spectrometry (TAP-MS), which resulted in a high-confidence PIN with interactions ranging from 8 (NS2B and NS5) to 27 (NS2A), totaling 62 unique protein hits and 89 bait–prey interactions (Figures 1A and 2A,B; Table S3).

**Figure 2.** TAP-MS PIN. (**A**) Protein interaction profile of TAP-MS screenings plotted based on normalized spectral abundance factor (NSAF) [34] (X axis) and specificity based on fold change of spectral counts (Y axis) between TAP-tagged NS proteins and TAP-tagged GFP (negative control). Node size denotes the spectral sum (Spec sum) obtained for each protein. Node color denotes CRAPome PCT score according to the scale. (**B**) Network of the interactions identified by TAP-MS screens. Orange nodes represent ZIKV NS proteins (baits) and gray nodes represent human proteins (preys). Gray nodes with a red circle indicate a prey with high CRAPome score.
