Fig 2: Corroboration of Akirin2–IP interactions in HL60 cellsSome of the previously identified Akirin2 IPs [14] were selected for analysis of Akirin2–IP interactions in human HL60 cells. (A) Representation of the pull-down experiment using COOH-Myc-tagged human Akirin2 with c-Myc magnetic beads and HL60 cells lysate. (B) Akirin2 was incubated with HL60 lysate and immunoprecipitated with c-Myc magnetic beads specific for the Akirin2 protein Myc-tag. HL60 lysate and beads incubated only with HL60 lysate were used as positive (C+) and negative (C-) controls, respectively. PBS and recAkirin2 were also used as secondary controls. Akirin2-IPs were detected by mouse or rabbit antibodies specific for each protein and visualized by a dot blot analysis with anti-mouse or anti-rabbit secondary antibodies. (C) Dot blot areas of Akirin2–IPs dots were measured using ImageJ program and compared with the dots of SNRNP70, a protein used as negative control due to its low interaction with Akirin2, by Student’s t test with unequal variance (*P<0.05; n=2 biological replicates).

Fig 3: Differential expression and GO:BP of genes associated only with Akirin2 in response to akirin2 KO in human HL60 cells(A) Heatmap of GO:BP annotations and number of genes with KO/WT -2 < log2foldchange > 2 on each BP. Black arrows indicate the BPs associated only with Akirin2. (B) Heatmap of RNAseq results for genes with KO/WT -2 < log2foldchange > 2 and associated only with Akirin2 (Supplementary Data S5). KO4 and WT4 correspond to KO1–KO3 and WT1–WT3 mean values, respectively. Red and blue arrows indicate genes associated with all four BPs or taxis alone, respectively. (C) Gene representation (n, %) in the four BPs, taxis (TA), cell motility (CM), localization of cell (LC) and regulation of cell adhesion (RCA) associated only with Akirin2. Heatmaps were prepared with average linkage and Euclidean distance measurement method.

Fig 4: Characterization of Akirin2–histone H3 105–124 physical interactions(A) Akirin2–histone H3.1 interaction by recAkirin2 binding to a histone peptide array, which resulted in the identification of histone H3 105–124. Relative fluorescence is shown in RFU. (B) Corroboration of the interactions between recAkirin2 (produced in insect cells and E. coli) and histone H3 105–124 by in vitro protein pull-down and Western blot with Akirin2-specific primary antibodies. Immunoreactive proteins were visualized by chemiluminescence. Streptavidin beads incubated only with Akirin2 and recAkirin2 were included as negative (C–) and positive (C+) controls, respectively. (C) Corroboration of recAkirin2–histone H3 105–124 interactions by ITC. Calorimetric titrations of histone H3 105–124 into Akirin2 (left) and BSA (right). Upper plots show the thermogram and lower plots show the interaction isotherm. Non-linear regression data analysis allowed estimating the interaction parameters for Akirin2 (continuous red line): Ka, association constant; Kd, dissociation constant; ?H, interaction enthalpy; n, fraction of active (binding-competent protein). No interaction was observed for BSA.

Fig 5: Experimental design and rationaleThe effect of akirin2 KO was evaluated on the transcriptome and proteome of human HL60 cells. Results of differential gene expression and protein representation in akirin2 KO and WT cells and analysis of GO annotations for BPs were used to characterize Akirin2 gene/protein targets and to identify Akirin2 functional complements. The BPs that are not related to known Akirin2 IPs and thus associated only with Akirin2 were identified and selected for further analysis (green arrows). Genes in these BPs may be regulated by Akirin2 through unknown IPs, indirect (genetic) interactions or direct chromatin remodeling. To address the possibility of direct chromatin remodeling, Akirin2–histone physical interactions were then investigated.