The Ataxia Telangiectasia Nude Mouse with No Risk of Thymoma: A Model to Investigate Tumour Development of B Cell and Myeloid Origin Associated with ATM Loss

Abstract 1320 Constitutional inactivation of Ataxia Telangiectasia Mutated (ATM), a principal double strand breaks (DSB) response gene leads to ataxia telangiectasia syndrome characterised by a high risk of development of a wide range of leukaemias and lymphomas. Unfortunately, the existing Atm null mouse model provides the opportunity to study only a single tumour type as affected animals die of thymic T cell lymphoma by age of 3 months. In order to produce an Atm−/− mouse model that recapitulates the human disease, and exhibits a wider tumour range, we bred Atm−/−nu−/− double knockout mice. This led to the absence of thymomas due to defective T cell development caused by the loss of Foxn1 alleles in nude mice. We generated 29 Atm−/−nu−/− double knockout animals exhibiting a severe reduction in T cell compartment. In comparison with single Atm−/− mice or Atm−/−nu+/−displayed a greater than doubling of the median survival time (212 days vs 90 days), delayed onset and a decreased frequency of tumour development. Importantly, these animals also showed a change in tumour phenotype. Of 29 mice, 12 developed lymphomas between 2 and 10 months of age. Importantly, in contrast to the exclusively observed T-cell lymphomas in Atm−/− animals, Atm−/−nu−/−lymphoma appeared to be of B-cell origin. Closer morphological evaluation of tumour nodular proliferation was suggestive of germinal centre derived cells. We performed immunocytochemistry and found tumour cells to be positive for B cell markers B220 and Pax5, GC markers Bcl-6 and AI, weakly positive for IRF4 and negative for T cell marker CD3. Absence of cyclin D1 staining argued against a mantle cell lymphoma phenotype, whereas absence of staining for additional GC markers LMO2, HGAL, GCET1, CD10 and post GC marker Blimp-1 suggested that clonal proliferation has originated from GC cells that had not completed the GC reaction. In addition, the B cell tumour phenotype was also confirmed by FACS analysis that revealed positivity for surface marker B220 and variable positivity for IgM. Interestingly, we observed that tumour nodular formations were surrounded by rare cells that exhibited helper cell phenotype. This was consistent with incomplete abolition of T cell development in Atm−/−nu−/− animals, that might play a role in B lymphoma development. Tumours of non-haematological origin were rarely seen – one Atm−/−nu−/−mouse without a lymphoma developed an aggressive spindle cell sarcoma, topographically associated with the surface of the brain, most likely of meningeal origin. The most intriguing observation however, was the finding of an increased extramedullary haematopoiesis in the Atm−/−nu−/− mice. We have observed the presence of megakaryocytic proliferation as well as their extensive clustering, both in the bone marrow and spleen reminiscent of human myeloproliferative disease (MPD) in 10/22 analysed Atm−/−nu−/− animals. These changes were occasionally accompanied by small foci of haematopoiesis within the liver. The increased number of megakaryocytes in Atm−/−nu−/− animals was not accompanied by bone marrow fibrosis, bone remodelling, increase in blasts or other abnormalities, frequently seen in MPD. Finally, as another indicator of increased myelopoiesis Atm−/−nu−/− mice showed elevated proportions of myeloid progenitors in the spleen. Interestingly, this expansion in the myeloid cell progenitors was not accompanied by an increase in proliferative potential in vitro indicating that this myeloproliferation could be a result of an altered in vivo microenvironment in the Atm−/−nu−/−mice. In conclusion, we have produced an Atm null animal model that recapitulates a further part of the spectrum of tumours observed in humans lacking ATM function and additionally, points to a novel role of ATM in myeloproliferation. This model facilitates elucidation of mechanisms involved in tumorigenesis at different stages of haematopoietic differentiation and in the longer term may assist development of targeted treatments for these aggressive, currently incurable malignancies. Disclosures: No relevant conflicts of interest to declare.