Supplementary Figure 3: The effect of Pax5 deletion and reintroduction on pro-B cells and genome organization.

(a), Flow cytometry of wild type and Pax5^−/− pro-B cells stained with antibodies against IgM and CD19. Pro-B cells are isolated at IgM⁻. Wild type also shown in Supplementary Fig. 2a. (b), DNA content examination by flow cytometry of ethanol fixed wild type and Pax5^−/− pro-B cells stained with propidium iodide. The largest peak represents cells in G0/1 of the cell cycle, the second peak G2/M, and the intervening region is made up of cells in S phase. (c), Percentage of DNA loops in wild type and Pax5^−/− pro-B cells with anchors overlapping at least one CTCF binding site. (d), Multi-dimensional scaling plot constructed from the interaction intensities in wild type and Pax5^−/− pro-B cells. Distances between samples represent the leading log₂-fold change between samples for the top 50000 bin pairs with the largest log₂-fold changes. (e), Using the change-point detection algorithm in TADbit, we determined the number and (f), size of TADs in wild type and Pax5^−/− pro-B cells. Data derived from pool of two independent experiments. Box plot shows median and 5^th and 95^th percentiles. (g), Multi-dimensional scaling plot constructed from the interaction intensities in wild type and Pax5^−/− pro-B cells and LSK cells. Distances between samples represent the leading log₂-fold change between samples for the top 50000 bin pairs with the largest log₂-fold changes. (h), As shown in Fig. 3c and f, percentage of DI bin pairs strengthened or weakened between follicular B cells versus pro-B cells and plasmablasts versus follicular B cells with anchors containing either pro-B cell specific, follicular B cell specific or common high (MACS2 peak threshold of 50) or low (MACS2 peak threshold of 10) stringency Pax5 binding sites. Numbers of unclustered DIs are shown. Data from duplicate libraries derived from two independent experiments were compared using Pearson’s chi-squared test with Yates’ continuity correction was used to determine significance. (i), As shown in Fig. 4c, d, percentage of DI bin pairs between Pax5^−/− versus wild type pro-B cells and treated Pax5^−/− Pax5:ER pro-B cells versus untreated with anchors unbound or Pax5 bound (high and low stringency) that are strengthened or weakened in the absence or 24 h reintroduction of Pax5. Numbers of unclustered DIs shown. Data from duplicate or triplicate libraries derived from two or three independent experiments were compared using Pearson’s chi-squared test with Yates’ continuity correction was used to determine significance. (j), Flow cytometry of Pax5^−/− Pax5:ER untreated and 24 h β-estradiol treated pro-B cells stained with antibodies against Flt3 and CD19. Untreated Pax5^−/− pro-B cells were isolated as Flt3^high CD19⁻. 24 h treated pro-B cells (Pax5 reintroduced) were Flt3^low CD19⁺. (k), DNA content examination by flow cytometry of ethanol fixed Pax5^−/− Pax5:ER untreated and 24 h β-estradiol treated pro-B cells stained with propidium iodide. (l), Percentage of DI bin pairs between treated Pax5^−/− Pax5:ER pro-B cells versus untreated with anchors unbound or Pax5 bound that are strengthened or weakened after 6 h reintroduction of Pax5. Numbers of unclustered DIs are shown. Pearson’s chi-squared test with Yates’ continuity correction was used to determine significance. (m), Flow cytometry of Pax5^−/− Pax5:ER pro-B cell cultures with or without β-estradiol (βES) and/or α-amanitin. Numbers indicate the percentage of viable cells in each gate. (n), In situ Hi-C contact matrices of regions around the Pax5 target genes Spi1 and Dntt in untreated, 6 h Pax5 reintroduced and 6 h Pax5 reintroduced with α-amanitin treatment pro-B cells. Color scale indicates number of reads per bin pair