Aberrant TCRδ rearrangement underlies the T-cell lymphocytopenia and t(12;14) translocation associated with ATM deficiency

As a master regulator of the DNA damage response, ataxia telangiectasia mutated (ATM) kinase is rapidly activated by DNA double-strand breaks and phosphorylates substrates involved in both DNA repair and checkpoint control. Loss of ATM causes ataxia-telangiectasia (A-T) syndrome characterized by primary immunodeficiency and greatly increased risk for lymphoid malignancies, especially of T-cell lineage.¹  Moreover, >50% of T-cell prolymphoblastic leukemia (T-PLL) patients carry a somatic mutation of ATM.²  Both T-PLL and A-T patients carry inv(14;14) that involves the T-cell receptor α/δ (TCRα/δ) and immunoglobulin H loci on human chromosome 14. However, the molecular origin of the seemingly T-cell–specific lymphocytopenia and inv(14;14) remains unknown.

ATM^−/− mice recapitulate the T-cell lymphocytopenia of A-T patients and succumb to thymic lymphomas with TCRα/δ-related chromosome 14 amplification and t(12;14) translocations, syntenic with inv(14;14) in humans.³  Reduced αβ T-cell numbers and blockade at the double-positive (DP) to single-positive (SP) transition is characteristic of ATM-deficient T cells.^4,5  Both human and mouse TCRδ loci reside between Vα and Jα segments.⁶  As such, TCRα rearrangement deletes the TCRδ gene, leading to irreversible commitment to the αβ T-cell lineage. This stage-specific rearrangement of the TCRδ in double-negative (DN) and TCRα in DP T cells is driven by the loci-specific enhancer elements Eδ and Eα, respectively.^6,7  However, Eα deletion does not affect the kinetics or translocation patterns of ATM-deficient thymic lymphomas,³  suggesting that TCRδ might be involved. In this context, TCRδ-related translocations are found in the majority of T-ALL and >90% of TAL1 translocated cases.⁸  Using ATM^−/−Eδ^−/− mice, here we identified aberrant TCRδ rearrangements as the origin for both αβ T-cell development defects and the recurrent t(12;14) translocations in ATM-deficient mice.

All animal work was conducted with proof from the institutional Animal Care and Use Committee of Columbia University. ATM^−/−9  and Eδ^−/−10  mice were described previously. T-cell development was characterized in 4- to 6-week-old mice as previously described.¹¹  Briefly, single-cell suspensions of thymus and spleen were counted and stained with fluorophore-conjugated antibodies against CD4, CD8, CD3, TCRβ, or TCRγ. For DN staining, thymocytes were counterstained with phycoerythrin-conjugated anti-CD4, CD8, B220, TCRγδ, Ter119 and CD19 antibodies, then with fluorophore-conjugated anti-CD44 and anti-CD25 antibodies. Flow cytometry data were collected on FACSCalibur (BD) and analyzed with FlowJo. For cytogenetic analyses, single cell suspensions were incubated with colcemid (10 µg/mL) for 4 to 10 hours and fixed and stained³  using ASI applied spectral imaging paints. Images were acquired and analyzed with Isis and Metafer 4 (MetaSystems, Newton, MA). Comparative genomic hybridization (CGH) was performed with the Agilent 244K mouse array. Reverse-transcription polymerase chain reaction (PCR) methods are detailed in Figure 2F and supplemental Table 1 (available on the Blood Web site).

ATM^−/−Eδ^−/− mice were born at the expected ratio and undistinguishable from ATM^−/− littermates. In thymuses of ATM^−/−Eδ^−/− mice, γδ T-cell frequency was reduced, but not abolished, consistent with the role of Eα in residual γδ T-cell development¹⁰  (Figure 1A-B). Although thymus cellularity was reduced by ∼50% in ATM^−/− mice, both thymus cellularity and the number of mature SP T cells were significantly rescued in ATM^−/−Eδ^−/− mice (Figure 1C-D). This is unexpected, given that >90% of thymocytes are αβ T cells. Further analyses showed that this rescue was mostly due to moderate increases of DNIV (CD25⁻CD44⁻) percentage and marked increases of DP thymocyte number without changes in the SP to DP ratio (Figure 1C-D and supplemental Figure 1A-B). TCRδ undergoes V(D)J recombination in DNII-DNIII (CD25⁺).¹⁰  ATM is particularly important for inversional V(D)J recombination,¹²  and Vδ5 is located downstream of Cδ and requires inversional rearrangement.¹³  The configuration of the TCRα/δ locus also determines that if DNA breaks in the TCRδ loci are unrepaired, then Eα would separate from Vαs and disable TCRα V(D)J recombination (Figure 2C). These findings reveal an important role of genomic instability at the TCRδ locus in αβ T-cell development and T-cell lymphocytopenia associated with ATM deficiency, and they suggest that suppression of TCRδ rearrangement might improve T-cell counts in A-T patients.

Despite improved T-cell counts, the kinetics of succumbing to thymic lymphomas was similar in ATM^−/−Eδ^−/− mice and ATM^−/− controls (P = .1), with a median survival of 112 days (vs 99 for ATM^−/− controls) (Figure 2A). ATM^−/−Eδ^−/− thymic lymphomas consisted of immature (surface TCRβ^low) DP T cells similar to ATM^−/− controls (supplemental Figure 1C). Southern blot identified clonal TCRβ rearrangements and focal amplification centromeric to the TCRα/δ loci (chromosome 14) in all ATM^−/−Eδ^−/− (or ATM^−/−Eδ^+/−) lymphomas as documented for ATM^−/− lymphomas (supplemental Figures 1D and 2A).³  ATM^−/− thymic lymphomas were derived from immature T cells that had not yet undergone TCRα arrangement, evidenced by retention of the Cδ.³  All ATM^−/−Eδ^+/− and 2 of 5 ATM^−/−Eδ^−/− lymphomas (545 and 2280) retained at least 1 Cδ, suggesting that the chromosome 14 amplifications most likely occurred during TCRδ rearrangement in those tumors, consistent with the ability for Eα to drive residual TCRδ rearrangements in the absence of Eδ.¹⁰  Notably, the other 3 ATM^−/−Eδ^−/− lymphomas showed homozygous deletion of Cδ, indicative of biallelic TCRα rearrangements (Figure 2C). Southern blot further confirmed that all 3 tumors (497, 713, and 745) rearranged the proximal Jα segments (Figure 2C). Thus, TCRδ rearrangement is not required for chromosome 14 amplification and T-cell lymphomagenesis in ATM-deficient mice. In the absence of the TCRδ rearrangement, aberrant TCRα rearrangement is able to promote chromosome 14 amplifications and oncogenic transformation of ATM-deficient T cells.

We next investigated the origin of the t(12;14) translocations characteristic of ATM deficiency.³  Chromosome paint identified clonal t(12;14) translocations in all ATM^−/−Eδ^+/− lymphomas and the 2 Cδ-positive ATM^−/−Eδ^−/− lymphomas (545 and 2280), but not in the 3 ATM^−/−Eδ^−/− lymphomas (497, 713, and 745) with biallelic TCRα rearrangements (Figure 2B,E). CGH analyses of 2 Cδ-positive (2280 and 545) and 2 Cδ-negative (497 and 745) ATM^−/−Eδ^−/− lymphomas confirmed focal amplification upstream of the TCRα/δ locus (Figure 2D and supplemental Figure 2B) and also revealed a unique hemizygous deletion of chromosome 14 downstream of the TCRα/δ locus only in Cδ-negative (497 and 745) tumors (Figure 2D and supplemental Figure 2B). CGH probe mapping showed that the deletion started at the proximal Jα (Figure 2C-D). Together with the absence of the t(12;14) translocation, this result suggests that when chromosome 14 amplification occurs in DP T cells during TCRα arrangement, the break in chromosome 12 is no longer available to form the t(12;14) translocation, causing the telomeric portion of chromosome 14 to be lost and pointing to concurrent immunoglobulin H and TCRδ rearrangement in DN T cells as the molecular origin for t(12;14). This further implied that t(12;14) and the syntenic inv(14,14) are not essential for T-cell transformation. Despite the lack of t(12;14), both Cδ-negative (497 and 745) tumors displayed hemizygous deletion of chromosome 12 telomeric regions, supporting the existence of potential tumor-suppressor genes (eg, Bcl11b).^3,14-16  In this context, SKY analyses identified t(12;15) translocations involving chromosome 15 in tumor 745 (Figure 2B and supplemental Figure 2C). Finally, CGH also indicated that murine ATM^−/− thymic lymphomas, regardless of Eδ status, display genetic changes that have been associated with human T-cell acute lymphoblastic leukemia (T-ALL), including trisomy of chromosome 15 containing c-Myc, amplification and overexpression of Notch1, deletion of Pten, and overexpression of IL7R (Figure 2F), highlighting ATM-deficient murine thymic lymphomas as a bona fide animal model for human T-ALL. In summary, our study identifies aberrant TCRδ rearrangement in the absence of ATM together with the unique configurations of the TCRα/δ locus as one of the major causes to both T-cell–specific lymphocytopenia and the t(12;14) and related inv(14;14) associated with ATM deficiency and potentially in other T-ALL.

This work was supported in part by National Institutes of Health (NIH) National Cancer Institute grants 5R01CA158073, 1R01CA184187, and 1P01CA174653-01 and American Cancer Society Research Scholar Grant RSG-13-038-01 DMC (S.Z.) and by NIH National Cancer Institute grant P01CA109901 (F.W.A.). S.Z. was a St Baldrick’s Scholar for Pediatric Cancer and is a Leukemia Lymphoma Society Scholar. W.J. was supported in part by NIH National Cancer Institute grant T32-CA09503. F.W.A. is a Howard Hughes Medical Institute investigator.

Contributions: W.J., F.W.A., and S.Z. designed experiments; S.Z. wrote the paper; M.G. performed the SKY analyses for tumor 745; and W.J., B.J.L., C.L., R.L.D., and S.Z. performed the rest of the experiments.

Conflict-of-interest disclosure: The authors declare no competing financial interests.

Correspondence: Shan Zha, Institute for Cancer Genetics, Columbia University Medical Center, 1130 St. Nicholas Ave, Room 503B, New York, NY 10032; e-mail: sz2296@columbia.edu.