Activation-induced cytidine deaminase prevents pro-B cell acute lymphoblastic leukemia by functioning as a negative regulator in Rag1 deficient pro-B cells

Abstract

Activation-induced cytidine deaminase (AID) is essential for somatic hypermutation and class switch recombination in mature B-cells, while AID was also shown to play a role in developing pre-BCR/BCR-positive B-cells of the bone marrow. To further elucidate a potential function of Aid in the bone marrow prior to V(D)J-recombination, we utilized an in vivo model which exerts a B-cell developmental arrest at the pro-B cell stage with low frequencies of pro-B cell acute lymphoblastic leukemia (pro-B ALL) development. Therefore, p19Arf^-/-Rag1^-/- (AR) mice were crossed with Aid-deficient mice (ARA). Surprisingly, loss of Aid expression in pro-B cells accelerated pro-B ALL incidence from 30% (AR) to 98% (ARA). This effect was Aid dose dependent, since Aid^+/- animals of the same background displayed a significantly lower incidence (83%). Furthermore, B-cell-specific Aid up-regulation was observed in Aid-competent pro-B ALLs. Additional whole exome/sanger sequencing of murine pro-B ALLs revealed an accumulation of recurrent somatic Jak3 (p.R653H, p.V670A) and Dnm2 (p.G397R) mutations, which highlights the importance of active IL7R signaling in the pro-B ALL blast cells. These findings were further supported by an enhanced proliferative potential of ARA pro-B cells compared to Aid-competent cells from the same genetic background. In summary, we show that both Aid and Rag1 act as a negative regulators in pro-B cells, preventing pro-B ALL.

Keywords: Rag1 deficiency; activation induced cytidine deaminase; acute lymphoblastic leukemia; pro-B cells.

Figure 1. Aid deficiency accelerates pro-B ALL development in p19Arf^-/-Rag1^-/- mice

A. Pro-B-ALL-dependent survival curve of p19Arf^-/-Rag1^-/- (AR), p19Arf^-/-Rag1^-/-Aid^+/- (ARa) and p19Arf^-/-Rag1^-/-Aid^-/- (ARA) mice. Curves show a significant difference in pro-B-ALL-dependent survival (Log-rank test p < 0.0001). B. Disease distribution of pro-B ALL in AR, ARa and ARA mice. C. Representative splenomegaly of a diseased ARa and ARA mouse, compared to a C57BL/6J wildtype mouse. D. Hematoxylin/Eosin staining from leukemic ARA spleens, showing loss of their architecture due to blast cell infiltration. E. Representative blot of hematopoietic subsets in diseased AR, ARa and ARA mice, showing an accumulation of CD19⁺IgM^- pro-B cells. F. Representative FACS analysis of a non-irradiated C57BL/6J wildtype recipient mouse that was transplanted with leukemic total BM from a diseased ARA donor mouse. Nine days after BM transplantation, blast cells (CD19⁺ckit⁺) are visible in bone narrow, spleen and lymph nodes (n = 2). G. Immunoglobulin V(D)J-recombination in ARa (lanes 1,2) and ARA (lanes 3,4) tissues infiltrated with leukemic blast cells, as analyzed by PCR. Thymocytes (lane 6) serve as negative control and sorted CD19⁺ B-cells (lane 5) from the spleens of healthy C57BL/6 wildtype mice serve as a control for polyclonal V(D)J-recombination. Infiltrated tissues show only the cμ heavy chain.

Figure 2. B-cell-specific Aid expression detectable in p19Arf^-/-Rag1^-/- mice

A. Quantitative Real-time-PCR analysis showing Aid mRNA expression levels in p19Arf^-/-Rag1^-/- (AR), p19Arf^-/-Rag1^-/-Aid^+/- (ARa) and p19Arf^-/-Rag1^-/-Aid^-/- (ARA) tumors. Stimulated B220⁺ splenic B-cells (LPS/IL-4) from wildtype mice serve as positive control. Aid expression is detectable in AR and ARa pro-B ALLs, while it is absent in the respective T-ALLs and myeloproliferative diseases (ML) (n = 3). B. Immunoblot analysis showing the presence of the Aid protein in AR tumors. Leukemic blast cells of ARA mice serve as a negative control, while stimulated B220⁺ splenic B-cells from wildtype mice were used as a positive control. Beta-Actin serves as a loading control (n = 3). C. Microarray analysis results comparing the gene expression between AR and ARA tumor samples, which were visualized in a volcano plot. Red dots have a fold change greater than two and a corrected p-value smaller than 0.05. Out of the 18465 data points, 823 are lower and 1001 higher in ARA tumor samples relative to AR tumors.

Figure 3. Sequencing of Aid-deficient tumors reveals activating mutations influencing IL-7 signaling

A. From Tumor/reference pairs of leukemic p19Arf^-/-Rag1^-/- (AR), p19Arf^-/-Rag1^-/-Aid^+/- (ARa) and p19Arf^-/-Rag1^-/-Aid^-/- (ARA) mice, CNV profiles were calculated using EXCAVATOR2. A representative plot for each genotype is shown. CNV profiles display recurrent gains (red) of regions on chromosome 14. B. Presentation of WES analysis from three ARA pro-B ALLs. 5-11 of the somatic variants were deemed tumor-specific by MUTECT analysis, with Jak3 and Dnm2 as recurrent genes. C. Representative chromatogram of Jak3 (p.R653H/p.V670A) and Dnm2 (p.G359E/p.G397R/p.R361P/p.S357P) mutation validation by Sanger sequencing. D. Sanger sequencing results of Jak3 Exon15 show an accumulation of somatic Jak3 p.R653H and p.V670A variants in AR (n = 4), ARa (n = 13) and ARA (n = 14) tumors. E. Sanger sequencing identifies recurrent mutations in the Dnm2 gene affecting the amino acids p.G359R/E, p.R361P, p.S357P and p.G397R/A in AR (n = 4), ARa and ARA pro-B ALLs (n = 14). F. Immunoblot analysis showing Stat5 phosphorylation in AR, ARa and ARA pro-B ALLs. IL-3 depleted BaF3 cells serve as negative (-) and BaF3 cells overexpressing murine Jak3 V670A as positive (+) control. Stat5 and beta-Actin were used as loading controls (n = 3).

Figure 4. In vitro cultured pro-B cells of healthy p19Arf^-/-Rag1^-/-Aid^-/- mice are highly IL-7 dependent and show an increased proliferative potential compared to their Aid-competent counterparts

A. Representative flow cytometric analysis of hematopoietic subsets in healthy p19Arf^-/-Rag1^-/-Aid^+/- (ARa) and p19Arf^-/-Rag1^-/-Aid^-/- (ARA) mice compared to wildtype mice. B. FACS analysis depicting the percentage of cells at a late apoptotic stage (7-AAD⁺Annexin⁺) from healthy p19Arf^-/-Rag1^-/- (AR), ARa and ARA pro-B cells cultured in vitro after 24 hours of IL-7 withdrawal. All groups show significant late apoptosis induction as calculated by student's t-test (AR p = 0.0444; ARa p = 0.0049; ARA p = 0.05; n = 3 for AR/ARa; n = 4 for ARA). C. FACS analysis showing the relative apoptosis ratio of healthy AR, ARa and ARA pro-B cells after 24 hours of IL-7 depletion. Therefore, the ratio between 7-AAD/Annexin double positive cells between the groups in IL-7+ and IL-7- conditions was calculated and the AR ratio set to 1. No significant differences in IL-7 sensitivity were observed between the groups (n = 3 for AR/ARA; n = 4 for ARA). D. Proliferation curve of healthy AR, ARa and ARA pro-B cells. Compared to AR cells, ARA cells show significantly increased proliferation on day 3 (p = 0.0098) and day 4 (p = 0.0244) as calculated by student's t-test. E. Representative FACS analysis and blot showing cell cycle distribution of healthy AR, ARa and ARA pro-B cells after 1 hour of BrdU pulse labeling. ARA pro-B cells display increased accumulation of cells in the S-phase compared to AR pro-B cells (student's t-test, p = 0.0318; n = 3 for AR/ ARA; n = 4 for ARA).