Fig 1: CBFB::MYH11 protein is cytoplasmic in murine hematopoietic cells.MSCV retroviruses were created with GFP fused in-frame with exons 1–5 of CBFB (CBFB-GFP), CBFB::MYH11 (CBFB:GFP-MYH11), or the portion of MYH11 involved in the CBFB::MYH11 fusion (GFP-MYH11). EV indicates the retrovirus with GFP alone. Lineage-depleted mouse bone marrow cells were transduced with retrovirus and harvested 4–7 days after transduction for immunofluorescence. DAPI staining (identifying the nucleus) is shown in blue, and yellow dotted lines outline the nucleus. GFP (detected directly) is in green, and RUNX1 (detected with antibody staining) is in red. Overlaps between GFP and RUNX1 signals are yellow. In EV-transduced cells, RUNX1 is localized to the nucleus. CBFB-GFP is predominantly localized to the nucleus, where it colocalizes with RUNX1. However, CBFB::MYH11-GFP is predominantly localized in cytoplasmic aggregates; in these cells, RUNX1 is mislocalized from the nucleus to the cytoplasm. GFP-MYH11 forms large aggregates in both the nucleus and the cytoplasm, and does not colocalize with RUNX1. Images shown are representative of 2–4 independent experiments. Scale bars: 10 μm.
Fig 2: The CBFB N104 residue mediates CBFB interaction with RUNX proteins.(A) Immunofluorescence of CBFB::MYH11-TurboID–transduced HSPCs (A) or CBFBN104A::MYH11-TurboID–transduced HSPCs (B) demonstrates cytoplasmic localization of each. DAPI staining (blue) identifies nuclei, which are outlined with yellow dotted lines. TurboID is detected in red. Images are representative of 2 experiments. Scale bars: 10 μm. (C) Normalized spectral counts of CBFB and MYH11 in protein lysates from CBFB::MYH11-TurboID (n = 8) and CBFBN104A::MYH11-TurboID (n = 6) are not significantly different. One-way ANOVA between all samples, ****P < 0.0001, nonsignificant comparisons not labeled. Each point represents an individual sample, bar indicates mean, box indicates 95% confidence interval, and whiskers indicate value range. (D) Interactions between myosin-related proteins are maintained in CBFBN104A::MYH11-TurboID samples. One-way ANOVA between all samples, ****P < 0.0001, nonsignificant comparisons not labeled. (E) CBFBN104A::MYH11 disrupts the interaction with RUNX1. One-way ANOVA between all samples, ****P < 0.0001, nonsignificant comparisons not labeled. (F) CBFBN104A::MYH11 disrupts interactions between CBFB::MYH11 and other nuclear proteins, suggesting that these interactions are also mediated by RUNX-CBFB binding. One-way ANOVA between all samples, **P < 0.01, ***P < 0.001, ****P < 0.0001, nonsignificant comparisons not labeled. (G) Primary murine hematopoietic cells were transduced with retroviruses encoding CBFB::MYH11-GFP (top) or CBFBN104A::MYH11-GFP (bottom). Note that while CBFB::MYH11-GFP and CBFBN104A::MYH11-GFP are both cytoplasmic, RUNX1 relocalizes to the nucleus in cells expressing CBFBN104A::MYH11-GFP. Images are representative of 2–4 experiments. Scale bars: 10 μm.
Fig 3: CBFB::MYH11 is predominantly cytoplasmic in human AML.(A) Primary human CBFB::MYH11 AML immunofluorescence. RUNX1/2/3 is shown in green, MYH11 in red, and overlap in yellow. Yellow dashed lines outline nuclei. In a subset of images, DAPI costaining is shown in blue. Note cytoplasmic MYH11 aggregates with colocalized RUNX1 (yellow arrows) (representative images from 8 different CBFB::MYH11 patients). Scale bars: 10 μm. (B) Quantification of cells with nuclear-only MYH11, cytoplasmic-only MYH11, or both nuclear and cytoplasmic MYH11 (total of 338 cells scored). (C) K562 cells were transfected with a plasmid encoding CBFB or CBFB::MYH11, and Western blotting was performed on protein lysates using an anti-CBFB antibody. Note detection of an approximately 30 kDa band corresponding to CBFB in all lanes but increased with CBFB transfection, and detection of an approximately 79 kDa band corresponding to CBFB::MYH11 only in cells transduced with a CBFB::MYH11 plasmid. (D) ProteinSimple Jess blot on 4 human CBFB::MYH11 AML samples. Equal volumes of nuclear and cytoplasmic lysates were loaded. Anti–lamin A/C and anti-actin antibodies were used to verify nuclear and cytoplasmic purity. N, nuclear fraction; C, cytoplasmic fraction. (E) The percentage of CBFB (left panel) or CBFB::MYH11 (right panel) in the nuclear or cytoplasmic fractions of the samples shown in D. Each point represents an individual sample, bar indicates mean, box indicates 95% confidence interval, and whiskers indicate value range.
Fig 4: PML::RARA, RUNX1::RUNX1T1, and CBFB::MYH11 have distinct protein interactomes in mouse hematopoietic cells.(A) Heatmap showing differentially interacting proteins (DIPs) (edgeR, FDR < 0.05, >2-fold change) with increased detection in PML::RARA-TurboID fusion samples (n = 12) relative to TurboID-alone samples (n = 23); the same proteins detected with the other TurboID fusion samples are plotted “passively.” Data are shown as z-scored normalized spectral counts. (B) Volcano plot of proteins identified in A, with selected proteins labeled. (C) The percentage of proteins in selected nuclear complexes with detectable interactions with PML::RARA-TurboID fusion, relative to TurboID alone. (D) Heatmap showing DIPs with increased detection in RUNX1::RUNX1T1-TurboID fusion samples (n = 6) relative to TurboID-alone samples (n = 23). (E) Volcano plot of proteins identified in D, with selected proteins labeled. (F) The percentage of proteins in selected nuclear complexes with detectable interactions with RUNX1::RUNX1T1-TurboID fusion, relative to TurboID alone. (G) Heatmap showing DIPs with increased detection in CBFB::MYH11-TurboID fusion samples (n = 8), relative to TurboID-alone samples (n = 23). (H) Volcano plot of proteins identified in G, with selected proteins labeled. (I) The percentage of proteins in selected nuclear complexes with increased interaction with CBFB::MYH11-TurboID fusions, relative to TurboID alone.
Fig 5: CBFB and CBFB::MYH11 have distinct protein interactomes.(A) Heatmap showing DIPs with increased detection in CBFB-TurboID fusion samples (n = 12), relative to TurboID alone (n = 23); proteins detected with the CBFB::MYH11-TurboID fusion protein are plotted passively. (B) Volcano plot of proteins identified in A, with selected DIPs labeled. (C) Percentage of proteins in selected nuclear complexes with increased interaction with CBFB-TurboID fusion relative to TurboID alone. (D) Heatmap showing DIPs with increased detection in CBFB::MYH11-TurboID fusion samples (n = 8) relative to TurboID-alone samples, with CBFB-TurboID samples passively plotted. (E) Heatmap showing DIPs between CBFB-TurboID and CBFB::MYH11-TurboID fusion proteins. (F) Volcano plot of DIPs between CBFB- and CBFB::MYH11-TurboID fusions, with key differential interactors labeled. Myosin-related proteins exclusively interact with CBFB::MYH11, while CBFB interacts predominantly with nuclear proteins. (G and H) Gel images of ProteinSimple Jess blot on streptavidin beads from HSPCs expressing the indicated TurboID fusions using an antibody against MYO18A (G) or NCOR2 (H). Note pull-down of MYO18A in CBFB::MYH11-TurboID fusion, and NCOR2 pull-down in RUNX1::RUNX1T1-TurboID fusion. (I and J) ToppFun pathway enrichment for CBFB-TurboID versus CBFB::MYH11-TurboID DIPs enriched in CBFB-TurboID (I) or CBFB::MYH11-TurboID (J). Ratio indicates number of genes identified as DIPs divided by number of genes in gene set. Circle size indicates number of proteins identified. FDR, Benjamini-Hochberg FDR.
Supplier Page from Abcam for Alexa Fluor® 488 Anti-RUNX1 / AML1 + RUNX3 + RUNX2 antibody [EPR3099]