NRF1 binding sites contribute to multiple myeloma pathogenesis. (A) Schematic representation of the workflow followed in the building of the MM-specific master list of NRF1-binding sites by combining consensus NRF1 binding peaks with MGUS-specific accessible regions. (B) UpSet plot depicting the overlap of NRF1-binding sites across varied MM cell lines (Kms18, MM196, MM217, and Kms27). The bar plot above represents the number of common peaks in each intersection group, with the total number of peaks in each cell line shown in the horizontal bar chart (left). The highlighted bar indicates the intersection of consensus peaks shared by all 4 cell lines (n = 5749). (C) A Venn diagram showing the number of MM-specific NRF1 binding peaks emerged by intersecting the cumulative MGUS ATAC-seq peaks (blue) with the MM NRF1-binding sites (red). (D) Unsupervised clustering heat map showing normalized scaled accessibility counts (z score of CPM) of NRF1 MM-specific binding sites (n = 699, left bar in blue) and randomly selected NRF1-binding sites common between MM and MGUS samples (n = 300, light blue). Top bar annotations represent the disease type (MM, purple; MGUS, green), MM disease status (NDMM, orange; treated, light purple), and the percentage of malignant PCs (red-to-white gradient, 100%-0%). Accessibility data come from ATAC-seq of MM (n = 55) and MGUS (n = 11) samples. Each column corresponds to a patient sample, and the entire cohort was clustered using the Ward D method with Euclidean distance to assess similarity. (E) Violin plots comparing chromatin accessibility at NRF1 MM-specific binding sites in matched NDMM and treated samples (n = 11), based on ATAC-seq data. Each panel represents a matched pair of samples, with corresponding Wilcoxon P values indicating significant differences in chromatin accessibility of selected sites (∗∗P < .01; ∗∗∗P < .005). The y-axis of each panel shows the log-transformed CPM of ATAC-seq signal intensity. Sample MM_046 is highlighted for its peculiarity in being CD138+dim (low). (F) Heat maps (bottom) and aggregate profiles (top) showing ATAC-seq signal intensity at NRF1 MM-specific binding sites across MM cell lines (U266, RPMI8266, Kms27, and MM196) and merged MGUS (n = 11) samples. The aggregate profiles display signal intensity from 0 (weak) to 1 (strong), whereas the heat maps use a dark red to white gradient to indicate high to low accessibility. (G) Heat maps (bottom) and aggregate profiles (top) displaying ChIP-seq signal intensity at NRF1 MM-specific binding sites across NDMM (n = 5) and MGUS (n = 4) samples. The aggregate profiles display signal intensity from 1 (weak) to 15 (strong), whereas the heat maps use a dark purple to white gradient to indicate high to low NRF1 occupancy. CPM, counts per million; TMM, trimmed mean of M-values.

NRF1 binding sites contribute to multiple myeloma pathogenesis. (A) Schematic representation of the workflow followed in the building of the MM-specific master list of NRF1-binding sites by combining consensus NRF1 binding peaks with MGUS-specific accessible regions. (B) UpSet plot depicting the overlap of NRF1-binding sites across varied MM cell lines (Kms18, MM196, MM217, and Kms27). The bar plot above represents the number of common peaks in each intersection group, with the total number of peaks in each cell line shown in the horizontal bar chart (left). The highlighted bar indicates the intersection of consensus peaks shared by all 4 cell lines (n = 5749). (C) A Venn diagram showing the number of MM-specific NRF1 binding peaks emerged by intersecting the cumulative MGUS ATAC-seq peaks (blue) with the MM NRF1-binding sites (red). (D) Unsupervised clustering heat map showing normalized scaled accessibility counts (z score of CPM) of NRF1 MM-specific binding sites (n = 699, left bar in blue) and randomly selected NRF1-binding sites common between MM and MGUS samples (n = 300, light blue). Top bar annotations represent the disease type (MM, purple; MGUS, green), MM disease status (NDMM, orange; treated, light purple), and the percentage of malignant PCs (red-to-white gradient, 100%-0%). Accessibility data come from ATAC-seq of MM (n = 55) and MGUS (n = 11) samples. Each column corresponds to a patient sample, and the entire cohort was clustered using the Ward D method with Euclidean distance to assess similarity. (E) Violin plots comparing chromatin accessibility at NRF1 MM-specific binding sites in matched NDMM and treated samples (n = 11), based on ATAC-seq data. Each panel represents a matched pair of samples, with corresponding Wilcoxon P values indicating significant differences in chromatin accessibility of selected sites (∗∗P < .01; ∗∗∗P < .005). The y-axis of each panel shows the log-transformed CPM of ATAC-seq signal intensity. Sample MM_046 is highlighted for its peculiarity in being CD138+dim (low). (F) Heat maps (bottom) and aggregate profiles (top) showing ATAC-seq signal intensity at NRF1 MM-specific binding sites across MM cell lines (U266, RPMI8266, Kms27, and MM196) and merged MGUS (n = 11) samples. The aggregate profiles display signal intensity from 0 (weak) to 1 (strong), whereas the heat maps use a dark red to white gradient to indicate high to low accessibility. (G) Heat maps (bottom) and aggregate profiles (top) displaying ChIP-seq signal intensity at NRF1 MM-specific binding sites across NDMM (n = 5) and MGUS (n = 4) samples. The aggregate profiles display signal intensity from 1 (weak) to 15 (strong), whereas the heat maps use a dark purple to white gradient to indicate high to low NRF1 occupancy. CPM, counts per million; TMM, trimmed mean of M-values.

This Feature Is Available To Subscribers Only

or Create an Account

Close Modal
Close Modal