Here, we summarize our recent progress on the research projects of the lab that are most relevant to the etiology and treatment of childhood cancers.

We have sought to elucidate mechanisms through which histone H2AX functions downstream of ATM, DNA-PKcs, and ATR to promote normal DSB repair, maintain genomic stability, and suppress transformation. These kinases phosphorylate H2AX to form γ-H2AX in chromatin around DSBs, which creates binding sites for DDR factors. We hypothesized that H2AX functions by stabilizing and holding together broken DNA strands in G1 phase cells and upon G1/S transition to prevent chromosome translocations. To test this hypothesis, we exploited V(D)J recombination in pre-B cell lines and primary thymocytes to induce DSB intermediates at specific locations in G1 phase cells. We used ChIP to quantify H2AX and γ-H2AX densities along Igκ DNA strands as they were cleaved by the RAG1/RAG2 endonuclease and repaired or not un-repaired. Our data revealed that 1) γ-H2AX forms at high densities near DSBs and at diminishing densities farther away from DSBs, 2) ATM regulates γ-H2AX formation through MDC1-dependent and MDC1-independent mechanisms, and 3) reduced H2AX expression impairs the density but not the distance of γ-H2AX along broken DNA strands. We also monitored γ-H2AX formation using 3-dimensional immunofluorescence and found evidence that γ-H2AX forms on both broken and un-broken DNA strands within a limited volume around DSBs. We demonstrated that although H2AX is not required for DNA end-joining within V(D)J recombination substrates, γ-H2AX formation along RAG-cleaved DNA strands prevents their separation in G1 phase cells and their progression into chromosome breaks and translocations following G1/S transition. Our data indicated that H2AX prevents translocations during V(D)J recombination by stabilizing disrupted DNA strands and preventing their irreversible dissociation. Such H2AX-dependent mechanisms likely function throughout the genome to ensure the correct joining of DNA ends generated by other types of DSBs. In the decade before our studies, potential H2AX functions and mechanisms that regulate γ-H2AX formation had been investigated following the induction of DSBs at random locations. Due to inherent limitations of these approaches, none of the mechanisms that we discovered had been found or known previously. Thus, our work had a substantial impact upon the field and altered dogma about mechanisms that regulate γ-H2AX formation.

Like most human cancers, T-cell acute lymphoblastic leukemia (T-ALL) arises through accumulation of mutations in somatic cells. One human T-ALL subtype involves TCRβ+ cells with activating NOTCH1 mutations, inactivation of the p53 pathway, deletions that span H2AX and ATM, and genomic instability. We had shown that germline H2ax/p53-deficient mice rapidly succumb to TCRβ– thymic lymphomas with genomic instability and Notch1 mutations. Although H2ax/p53-deficient lymphomas harbor lesions characteristic of one human T-ALL subtype, these tumors are TCRβ–, while the T-ALL tumors that develop in humans with somatic inactivation of H2AX and p53 express TCRβ. This phenotypic difference suggests that loss of H2AX (and/or ATM) and p53 during embryogenesis and/or prior to lineage commitment influences the developmental stage (pre versus post TCRβ-selection) at which T-cell tumors arise. To investigate this notion, we analyzed mice with conditional deletion of H2ax and p53 initiating in T-lineage committed cells. These mice succumbed at later ages to TCRβ+ thymic lymphomas harboring translocations and Notch1 mutations, as is found in human T-ALLs with somatic inactivation of H2AX and p53. Our data revealed that the developmental timing of H2ax and p53 loss impacts the rate of mortality from and phenotype of T-cell tumors. Notably, our study revealed that conditional deletion of tumor suppressor genes in lymphocytes could provide more appropriate physiologic models for human lymphoid cancers than germline deficient mice. We also found that combined somatic inactivation of H2ax and the Artemis DDR factor in p53-deficient T-lymphocytes increased cellular genomic instability and reduced the rate and incidence of thymic lymphoma. Our findings also proved that compound conditional mutant mice can be utilized to elucidate complementary DSB repair mechanisms required for lymphoma growth, thereby identifying drug targets to exploit tumor-intrinsic DDR phenotypes.

We also hypothesized that ATM-independent H2AX functions downstream of DNA-PKcs and ATR are essential for normal DSB repair, prevention of genomic instability, and suppression of cancer. To test this hypothesis, we sought to generate Atm/H2ax-deficient mice and use their primary lymphocytes and Abelson pre-B cell lines created from these mice to investigate the role of ATM-independent H2AX functions in stabilizing broken DNA strands during V(D)J recombination and in ensuring error-free repair of replication-associated DSBs. We found that ATM/H2AX deficiency caused early embryonic lethality in mice and dramatically increased cellular genomic instability as compared to ATM or H2AX deficiency alone. The severe genomic instability in Atm/H2ax-deficient embryonic cells was associated with requirement for H2AX to repair oxidative DNA damage resulting from ATM deficiency. We also used conditional Atr deletion in H2ax-deficient mouse embyronic fibroblasts to demonstrate that H2AX functions downstream of both ATM and DNA-PKcs at stalled replication forks enable DNA re-replication and prevent the aberrant repair of replication-associated DSBs. Due to the early embryonic lethality of Atm/H2ax-deficient mice, we had to use a proposed alternative method and test our hypothesis through conditional inactivation of H2ax in Atm-deficient T-cells. Through this approach, we have found that ATM/H2AX deficiency leads to a severe impairment in TCRβ-mediated thymocyte expansion and substantially elevated genomic instability in T-cells as compared to ATM and H2AX deficiency alone, but no difference in the rate of mortality from lymphomas despite substantially increased genomic instability in Atm/H2ax-deficient tumors. In contrast to embryonic cells, increased genomic instability in Atm/H2ax-deficient T-cells was due to a mechanistic requirement for H2AX function(s) downstream of DNA-PKcs in the repair of replication-associated DSBs that arise in cells lacking ATM. Thus, our data has revealed that ATM-dependent and independent DDR mechanisms are required for TCRβ-mediated thymocyte proliferation.

Since availability of human cancer samples is frequently limiting, we hypothesized that analysis of Atm-deficient mouse thymic lymphomas would provide insights into the etiology of human T-ALL. ATM deficiency has long been thought to predispose immature thymic T-cells to aberrant DSB repair during TCRα V(D)J recombination and resultant TCRα;TCL1 translocations. However, we found that the translocations of Atm−/− tumors actually involve Tcrδ genes and delete Tcl1 and Bcl11b. Bcl11b encodes a transcription factor that regulates αβ T-cell development and the cellular DDR in response to DNA replication stress and functions as a haplo-insufficient tumor suppressor in mice. Our study revealed previously underappreciated mechanisms that likely help drive transformation of ATM-deficient T-cells and provided proof-of-principle that analyses of mouse tumors can provide novel insights into the etiology of human cancer. Notably, BCL11B has been found inactivated in ~16% of human T-ALLs since our study was published.

Over the next few years, we plan to utilize Atm-deficient mouse lymphocytes as models to elucidate mechanisms by which genomic instability arises from aberrant DSB repair and how resultant genetic changes further impair the DDR in these cells to drive their malignant transformation through cellular proliferation. We will use the knowledge gained and mouse models created by our studies to design and test patient-tailored combinations of DNA-damaging agents and DDR factor inhibitors for their potential as more effective and less toxic cancer therapies.