Deletion of the transcriptional regulator TFAP4 accelerates c-MYC-driven lymphomagenesis

Abstract

Many lymphoid malignancies arise from deregulated c-MYC expression in cooperation with additional genetic lesions. While many of these cooperative genetic lesions have been discovered and their functions characterised, DNA sequence data of primary patient samples suggest that many more do exist. However, the nature of their contributions to c-MYC driven lymphomagenesis have not yet been investigated. We identified TFAP4 as a potent suppressor of c-MYC driven lymphoma development in a previous genome-wide CRISPR knockout screen in primary cells in vivo [1]. CRISPR deletion of TFAP4 in Eµ-MYC transgenic haematopoietic stem and progenitor cells (HSPCs) and transplantation of these manipulated HSPCs into lethally irradiated animals significantly accelerated c-MYC-driven lymphoma development. Interestingly, TFAP4 deficient Eµ-MYC lymphomas all arose at the pre-B cell stage of B cell development. This observation prompted us to characterise the transcriptional profile of pre-B cells from pre-leukaemic mice transplanted with Eµ-MYC/Cas9 HSPCs that had been transduced with sgRNAs targeting TFAP4. This analysis revealed that TFAP4 deletion reduced expression of several master regulators of B cell differentiation, such as Spi1, SpiB and Pax5, which are direct target genes of both TFAP4 and MYC. We therefore conclude that loss of TFAP4 leads to a block in differentiation during early B cell development, thereby accelerating c-MYC-driven lymphoma development.

Fig. 1. Loss of Tfap4 accelerates c-MYC-driven lymphomagenesis.

A Schematic representation of haematopoietic reconstitution of lethally irradiated recipient mice with donor fetal liver cells, a rich source of HSPCs. FLCs from doubly transgenic Eµ-MYC/Cas9 E13.5 donor embryos were lentivirally transduced with vectors encoding BFP or CFP as a marker and sgRNAs targeting Tfap4 (sgTfap4), a positive sgRNA targeting Trp53 (sgTrp53) or a negative non-targeting control sgRNA (sgControl). These transduced donor FLCs were then transplanted into lethally irradiated C57BL/6-Ly5.1 recipient mice. Tumour-free survival of recipient mice was measured as days post-transplantation. Haematopoietic tissues and peripheral blood were harvested from tumour burdened recipient mice for further analysis. B Kaplan–Meier survival curve showing tumour-free survival of mice transplanted with either of two vectors containing a sgTfap4, a positive control sgTrp53 or a negative sgControl. n indicates number of sick mice/number of mice transplanted with transduced HSPCs for each sgRNA. Median survival post-transplantation in days is indicated in brackets. ****p < 0.0001.

Fig. 2. Deletion of Tfap4 removes the selection pressure to acquire defects in the TRP53 pathway during c-MYC-driven lymphoma development.

A Western blot analysis of sgTfap4/Eµ-MYC/Cas9 and sgControl/Eµ-MYC/Cas9 primary lymphomas for TRP53, p19/ARF, TFAP4 and HSP70 (protein loading control). The blot includes positive and negative control cell lysates for mutant (Mut) and wild-type (WT) TRP53. Molecular weight markers are indicated in kDa. B Summary graph representing percentages of sgTfap4/Eµ-MYC/Cas9 (n = 9) and sgControl/Eµ-MYC/Cas9 (n = 9) lymphomas that have defects in the TRP53 pathway as assessed by Western blotting for TRP53 and p19/ARF. TRP53 wt determined as lack of detectable protein expression of both TRP53 and p19/ARF. TRP53 knockout (ko) is determined as no detectable TRP53 protein expression with high expression of p19/ARF. Mutant TRP53 is identified by high level expression of both TRP53 and p19/ARF proteins. Cell viability response curves and corresponding IC50 graphs in sgTfap4/Eµ-MYC/Cas9 and sgControl/Eµ-MYC/Cas9 lymphoma cell lines 24 h after treatment with the indicated doses of nutlin-3A (C) or etoposide (D). Cell viability was determined by flow cytometry; live cells were identified as the Annexin V/PI double negative population. Data represent percentage mean survival of lymphoma cell lines at each dose (sgControl n = 3, sgTfap4 n = 4–9). Data were log transformed and fitted to non-linear regression mean ± SEM. IC50 values were calculated using Prism Graphpad software. Each dot represents an independent lymphoma cell line; error bars represent mean ± SEM. Two-tailed Student’s t test, *p < 0.05.

Fig. 3. Deletion of TFAP4 in pre-leukaemic Eµ-MYC pre-B cells impairs B cell differentiation.

A Representative flow cytometry plot of surface staining for B220, CD19, IgM and IgD of a sgTfap4/Eµ-MYC/Cas9 lymphoma cell line. B Summary graph of surface Ig staining by flow cytometry of cell lines derived from Eµ-MYC/Cas9 primary lymphomas of each genotype; sgTfap4 (n = 10), sgTrp53 (n = 20) and sgControl (n = 5). sIg⁻ represents B220⁺ sIgM⁻ sIgD⁻ lymphomas; sIg⁺ represents B220⁺ sIgM⁺/sIgD⁺ lymphomas; and mixed indicates B220⁺ lymphoma cells with both sIgM⁻ and sIgM⁺ lymphoma cell populations. Lethally irradiated wild-type mice were reconstituted with Eµ-MYC/Cas9 FLCs transduced with either a sgTfap4 or a sgControl vector. The haematopoietic cell subsets in these transplanted mice were analysed at 3 weeks post-transplantation by flow cytometry. The percentages of donor derived (GFP⁺ BFP⁺) B cell subsets; sIgM⁻ (B220⁺ sIgM/sIgD⁻), immature sIgM⁺ B cells (B220⁺ sIgM⁺ sIgD⁻) and mature sIgM⁺/sIgD⁺ B cells (B220⁺ sIgM/sIgD⁺) in the peripheral blood (C), bone marrow (D), spleen (E), and lymph nodes (F) of recipient mice were determined. G Representative flow cytometry dot plot to examine the different B cell subsets, gated on live donor derived lymphoid cells GFP⁺ BFP⁺ CD45.2⁺ B220⁺. Data presents mean ± SEM, each dot represents an individual recipient mouse that had been transplanted with sgTFAP4/Eµ-MYC/Cas9 (n = 8) or sgControl/Eµ-MYC/Cas9 (n = 9) FLCs. Unpaired two-tailed Student’s t test with Welch’s correction, *p < 0.05, **<0.01, ***<0.001.

Fig. 4. Pre-leukaemic Eµ-MYC/Cas9 pre-B cells lacking TFAP4 are transcriptionally distinct from control Eµ-MYC/Cas9 pre-B cells.

Lethally irradiated recipient mice were reconstituted with Eµ-MYC/Cas9 FLCs that had been transduced with either a vector encoding sgTfap4 or a sgControl, and their pre-leukaemic cells were analysed at 3 weeks post-transplantation. A In the bone marrow, the percentages of pro-B cells, pre-B cells and immature B cells derived from donor (GFP⁺ BFP⁺) FLCs were determined by flow cytometric analysis. Data present mean ± SEM, Each dot represents an individual recipient mouse that had been transplanted with sgTfap4 (n = 8) or sgControl (n = 9) transduced Eµ-MYC/Cas9 FLCs. Unpaired two-tailed Student’s t test with Welch’s correction, *p < 0.05, **<0.01, ***<0.001. B Representative flow cytometry plots demonstrating gating strategy to identify the different donor derived (GFP⁺ BFP⁺ CD45.2⁺) B cell subsets in the bone marrow of recipient mice: pro-B (B220⁺ sIgM⁻ c-KIT⁺), pre-B (B220⁺ sIgM⁻ c-KIT⁻), immature B (B220⁺ sIgM⁺) cells. C Mean difference plot presenting log-fold changes (average of each group) with significantly differentially expressed genes highlighted (FDR < 0.05), red = up-regulated, blue = down-regulated. Some genes of interest are labelled from RNA-seq analysis of donor derived pre-leukaemic Eµ-MYC/Cas9 pre-B cells (GFP⁺ BFP⁺ B220⁺ sIgM⁻ c-KIT⁻) from sgTfap4/Eµ-MYC/Cas9 (n = 8) or sgControl/Eµ-MYC/Cas9 (n = 6) cohorts isolated from the bone marrow of recipient mice.

Fig. 5. Genes regulating B cell differentiation are down-regulated in TFAP4 deleted pre-leukaemic Eµ-MYC pre-B cells.

Lethally irradiated recipient mice were transplanted with Eµ-MYC/Cas9 FLCs that had been transduced with a vector containing either a sgTfap4 (n = 8) or a sgControl (n = 6), and the pre-leukaemic pre-B cells from these recipient mice were analysed at 3 weeks post-transplantation by RNA-sequencing. Relative expression (RPKM—reads per kilobase million) of selected gene transcripts regulating apoptosis (A), cell cycling and proliferation (B), or master transcription factors and coordinators of B cell differentiation as well as Erg (C). Statistical significance *adj. p < 0.05, FDR < 0.05. D Spi1, SpiB, Ikzf1, Ikzf3 and Pax5 genomic loci coverage plots of anti-TFAP4 and anti-H3K27ac (a marker of accessible chromatin) binding in pre-leukaemic Eµ-MYC pro/pre-B cells. Y-axis represents counts per million (CPM), gene isoforms pictured, thick bar indicates exons, orange box highlights binding. Data accessed from GSE133514 [20]. E Proposed mechanism: in wild-type mice normal B cell development occurs, whereby pre-B cells differentiate into immature B cells expressing surface IgM. In Eµ-MYC transgenic mice, c-MYC is abnormally overexpressed in the B cell lineage causing excess proliferation of pro-B/pre-B cells and a partial block in their differentiation, thereby producing an abnormally expanded pool of pre-leukaemic pre-B cells and some sIg⁺ B cells. Some of these cells will acquire oncogenic mutations that can collaborate with c-MYC over-expression in neoplastic transformation and consequently give rise to pre-B or B cell lymphoma. The absence of TFAP4 in the Eµ-MYC setting results in an even larger pool of pre-leukaemic pre-B cells arising due to the further restriction of differentiation by downregulation of transcription factors that are critical for B cell differentiation. This even larger pool of highly proliferative pre-leukaemic pre-B cells is thus more likely to acquire additional mutations that drive transformation into malignant surface Ig⁻ pre-B cell lymphoma.