Ldb1 is required for Lmo2 oncogene-induced thymocyte self-renewal and T-cell acute lymphoblastic leukemia

Abstract

Prolonged or enhanced expression of the proto-oncogene Lmo2 is associated with a severe form of T-cell acute lymphoblastic leukemia (T-ALL), designated early T-cell precursor ALL, which is characterized by the aberrant self-renewal and subsequent oncogenic transformation of immature thymocytes. It has been suggested that Lmo2 exerts these effects by functioning as component of a multi-subunit transcription complex that includes the ubiquitous adapter Ldb1 along with b-HLH and/or GATA family transcription factors; however, direct experimental evidence for this mechanism is lacking. In this study, we investigated the importance of Ldb1 for Lmo2-induced T-ALL by conditional deletion of Ldb1 in thymocytes in an Lmo2 transgenic mouse model of T-ALL. Our results identify a critical requirement for Ldb1 in Lmo2-induced thymocyte self-renewal and thymocyte radiation resistance and for the transition of preleukemic thymocytes to overt T-ALL. Moreover, Ldb1 was also required for acquisition of the aberrant preleukemic ETP gene expression signature in immature Lmo2 transgenic thymocytes. Co-binding of Ldb1 and Lmo2 was detected at the promoters of key upregulated T-ALL driver genes (Hhex, Lyl1, and Nfe2) in preleukemic Lmo2 transgenic thymocytes, and binding of both Ldb1 and Lmo2 at these sites was reduced following Cre-mediated deletion of Ldb1. Together, these results identify a key role for Ldb1, a nonproto-oncogene, in T-ALL and support a model in which Lmo2-induced T-ALL results from failure to downregulate Ldb1/Lmo2-nucleated transcription complexes which normally function to enforce self-renewal in bone marrow hematopoietic progenitors.

Graphical abstract

Figure 1.

Ldb1 is not essential for T-cell development. (A) Expression of the Rosa26-EGFP^fCre reporter in DN subsets from Rosa26-EGFP^f;Rag1-Cre;Ldb1^fl/fl mice. Numbers are percent of gated cells that express EGFP. Similar subsets from non-Cre transgenic Rosa26-EGFP^f mice are shown as negative control. Shown is 1 representative of 12 mice analyzed from 6 independent experiments. (B) Deletion of Ldb1 in thymocytes mediated by Rag1-Cre. Total thymocyte DNA from the indicated mice was used as a template for polymerase chain reaction (PCR) with primers that amplify the Ldb1⁺ (wt), Ldb1^flox (Ldb1^fl), or Ldb1^Δ (deleted) alleles. (C) Western blot of total thymocyte lysates from the indicated mice with anti-Ldb1, anti-Lmo2, or anti-actin (loading control). (D) Phenotype of Rag1-Cre;Ldb1^fl/fl mice. Upper panels: Flow cytometry (fluorescence-activated cell sorting [FACS]) analysis of total thymocytes (Thy) from the indicated mice stained with fluorochrome-conjugated anti-CD4 and anti-CD8. Lower panels: gated lin^neg thymocytes consisting of immature DN cells stained with fluorochrome-conjugated anti-c-kit and anti-CD25 to delineate DN(1-4) subsets. (E) FACS analysis of total lymph node (LN) cells from the indicated mice stained with fluorochrome-conjugated anti-CD4 and anti-CD8. Numbers in quadrants or gates are percentage of total cells. Results shown in panels B-E are 1 representative of 3 experiments.

Figure 2.

Ldb1 is required for induction of T-ALL in Lmo2-tg mice. (A) Kaplan-Meier survival plot of Ldb1^fl/fl, Ldb1^fl/fl;Lmo2-tg, and Rag1-Cre;Ldb1^fl/fl;Lmo2-tg mice. (B) Ldb1 gene deletion in thymocytes from the 4 Rag1-Cre;Ldb1^fl/fl;Lmo2-tg mice that developed T-ALL. (C) Ldb1 gene deletion in Rag1-Cre;Ldb1^fl/fl;Lmo2-tg thymocytes. Total thymocyte DNA from the indicated mice was used as a template for PCR with primers that amplify the Ldb1⁺ (wt), Ldb1^fl (floxed), or Ldb1^Δ (deleted) alleles. M, molecular weight standard. (D) Western blot of total thymocyte lysates from the indicated mice with anti-Ldb1, anti-Lmo2, or anti-actin (loading control). (E) Kaplan-Meier survival plot of Lmo2-tg and Lck-Cre;Ldb1^fl/fl;Lmo2-tg mice. (F) Ldb1 gene deletion in sorted DN2/3 or DN4 thymocytes from Lck-Cre;Ldb1^fl/fl or Lck-Cre;Ldb1^fl/fl;Lmo2-tg thymocytes. (G) Summary of Rosa26-EGFP^f Cre reporter expression in DN1-4 subsets from Rosa26-EGFP^f;Lck-Cre;Ldb1^fl/fl (n = 10), Rosa26-EGFP^f;Rag1-Cre;Ldb1^fl/fl, (n = 12), and Rosa26-EGFP^f;Il-7ra-Cre;Ldb1^fl/fl (n = 9) mice. Bar graphs show mean and standard deviation of EGFP expression. Panels B, C, and F are 1 representative of 3 experiments.

Figure 3.

Deletion of Ldb1 reverses the abnormal phenotype of Lmo2-tg DN thymocytes. (A) Phenotype of Rag1-Cre;Ldb1^fl/fl;Lmo2-tg mice. Upper panels: flow cytometry (FACS) analysis of total thymocytes from the indicated mice stained with fluorochrome-conjugated anti-CD4 and anti-CD8. Lower panels: gated lin^neg DN thymocytes stained with fluorochrome-conjugated anti-c-kit and anti-CD25 to delineate DN (1-4) subsets. One representative of 10 independent experiments. Right panel, numbers of lin^neg (DN) thymocytes from the indicated preleukemic 2- to 3-month-old mice. (B) Lck-Cre-mediated deletion of Ldb1 fails to normalize the Lmo2-tg phenotype. Upper panels: flow cytometry (FACS) analysis of total thymocytes from the indicated mice stained with fluorochrome-conjugated anti-CD4 and anti-CD8. Lower panels: gated lin^neg thymocytes stained with fluorochrome-conjugated anti-c-kit and anti-CD25 to delineate DN (1-4) subsets. One representative of 10 independent experiments. Right panel, numbers of lin^neg (DN) thymocytes from the indicated preleukemic 2- to 3-month-old mice. (C) Histograms showing surface expression of c-kit (upper) or CD25 (lower) on DN2/3 thymocytes from the indicated mice. One representative of 10 independent experiments. (D) Ratio of DN3:DN4 thymocytes in the indicated preleukemic 2- to 3-month-old mice. P values and number of mice per group are shown.

Figure 4.

Transplantability (self-renewal potential) of Lmo2-tg thymocytes is lost following deletion of Ldb1. (A) Schematic of experimental strategy. (B) Left panels, FACS analysis of total thymocytes from recipient B6 (CD45.1) mice that had been sublethally irradiated and injected intravenously with thymocytes from an Ldb1^fl/fl;Lmo2-tg (CD45.2) donor mouse (upper panel) or a Rag1-Cre;Ldb1^fl/fl;Lmo2-tg (CD45.2) donor mouse (lower panel) 8 weeks before analysis. Right panels, CD45.1 vs CD45.2 staining of total (live gate) or gated DN, DP, CD4 single positive (CD4SP), or CD8SP thymocytes from the indicated recipient mice. Numbers are percentage of donor (CD45.2) thymocytes in each group. One representative of 3 experiments. (C) Percent thymocyte chimerism (total thymocytes, top; DN thymocytes, bottom) in recipient mice 8 weeks after thymocyte transfer. (D) Kaplan-Meier survival plot of recipient mice (n = 3 mice of each genotype).

Figure 5.

Deletion of Ldb1 normalizes gene expression in Lmo2-tg DN thymocytes. (A) Principal component (PC1/2) analysis of RNA-seq gene expression data from B6 (wt, 4 replicates), Lmo2-tg (5 replicates), and Rag1-Cre;Ldb1^fl/fl;Lmo2-tg (5 replicates) DN2/3 and DN4 thymocytes. (B) Venn diagrams depicting number of differentially expressed (DE) genes in the DN2/3 population comparisons. Deletion of Ldb1 lowers the number of Lmo2-induced DE genes from 3369 to 809. (C) Gene set enrichment analysis of RNA-seq data. Genes associated with immune system processes, T-cell differentiation, activation, signaling, or cell proliferation and survival are downregulated in Lmo2-tg DN2/3 thymocytes, whereas genes associated with negative regulation of cell proliferation and growth are upregulated in Lmo2-tg DN2/3 thymocytes. Expression of these genes is normalized in Rag1-Cre;Ldb1^fl/fl;Lmo2-tg DN2/3 thymocytes. (D) Genes previously reported to be up- or downregulated in Lmo2-tg DN thymocytes relative to wt (non–Lmo2-tg; B6) thymocytes are substantially normalized by deletion of Ldb1. (E) Genes previously shown to be positively regulated by Ldb1/Lmo2 complexes in hematopoietic progenitor cells are upregulated in Lmo2-tg DN2/3 thymocytes, and their expression is substantially normalized by deletion of Ldb1. For each of the genes shown in panels C-E, normalization of gene expression in Rag1-Cre;Ldb1^fl/fl;Lmo2-tg thymocytes (ie, trending toward that of wt [B6] compared with Lmo2-tg) was statistically significant (P < .05). FDR, false discovery rate.

Figure 6.

Ldb1 and Lyl1 control expression of a similar cohort of genes that are dysregulated in Lmo2-tg thymocytes. (A-B) Deletion of Ldb1 or Lyl1 has similar effects on gene expression in Lmo2-tg DN thymocytes. Shown are the expression of genes that are upregulated (A) or downregulated (B) in Lmo2-tg thymocytes and that are normalized by deletion of either Lyl1 or Ldb1. (C-D) Deletion of Ldb1 normalizes expression of genes reported to be upregulated (C) or downregulated (D) in human ETP-ALL.^, Results shown are from RNA-seq of DN2/3 thymocytes from B6, Lmo2-tg, or Rag1-Cre;Ldb1^fl/fl;Lmo2-tg mice. For each of the genes shown in panels A-D, normalization of gene expression (ie, trending toward that of wt (B6) in Rag1-Cre;Ldb1^fl/fl;Lmo2-tg thymocytes compared with Lmo2-tg was statistically significant (P < .05).

Figure 7.

Transcription complexes that contain Ldb1 and Lmo2 bind to the promoters of key dysregulated genes in Lmo2-tg thymocytes and human T-ALL cells. (A) Binding of Ldb1 in mouse hematopoietic progenitor cells at the promoters of Lyl1, Hhex, and Nfe2 (genes that are upregulated in Lmo2-tg thymocytes). Shown are mm9 UCSC Genome Browser shots of ChIP-seq (IgG and Ldb1) performed on lin^neg bone marrow cells enriched for hematopoietic progenitors. (B) Ldb1 binding sites identified by ChIP-exo sequencing of anti-Ldb1 ChIP samples from LOUCY human T-ALL cells. (C-D) ChIP-quantitative PCR (ChIP-qPCR) analysis of Ldb1/Lmo2 complex binding sites at the Lyl1, Hhex, and Nfe2 genes (sites are shown in A). Samples for ChIP-qPCR were lineage-depleted (DN) thymocytes pooled from 3 Ldb1^fl/fl;Lmo2-tg (Cre⁻) or 3 Il-7rα-Cre;Ldb1^fl/fl;Lmo2-tg (Cre⁺) mice. ChIP was performed with anti-Ldb1 (C) or anti-Lmo2 (D). Bar height is the mean, and error bars show standard deviation. There was no significant difference in ChIP-qPCR results with Cre⁻ and Cre⁺ samples using control primers and probes located near but outside the binding sites shown in A (data not shown).