Repetitive Busulfan Administration Induces Emergence of Dominant and Expanding Hematopoietic Clones with Retroviral Vector Insertion in Rhesus Macaques

Retroviral vector insertion induced mutagenesis has been demonstrated to be associated with the development of leukemias and other tumors in human and animals. In a steady-state, any proliferative advantage of “pre-malignant” cells harboring retroviral insertions takes years to manifest. We hypothesized that, under hematopoietic stress, behavior of some potentially abnormal clones may rapidly become evident. Busulfan (Bu) is known to have profound effects on stem cell behavior, specifically stem cell numbers, thus may function as hematopoietic stressor. In this study, we analyzed the impact of repetitive doses of Bu on hematopoietic clones with retroviral vector insertion in vivo in two rhesus macaques, RQ2297 and RQ2314. Eight years ago, RQ2297 and RQ2314 underwent autologous transplantation of CD34+ cells transduced with standard MLV-derived retrovirus vectors (G1Na and LNL6) carrying bacterial Neo^R gene marker. The two animals received a single dose of Bu 4mg/Kg approximately 2.5 years after transplant, and three to four additional monthly Bu injections (4mg/kg, 4 mg/kg, 6mg/kg and 6mg/kg for RQ2297, and 4mg/kg, 4 mg/kg and 6mg/kg for RQ2314) were administered 4 and 7 years post transplant for RQ2297 and RQ2314, respectively. LAM-PCR and Gene Scanning were employed to profile retroviral integration sites (RIS) in peripheral white blood cells. The results showed that repetitive Bu administration resulted in prolonged (15–20 weeks) decrease in three-lineage blood counts followed by complete recovery to baseline approximately four months after the last infusion of Bu in both animals except for persistent thrombocytopenia in RQ2297. RQ2297 harbored high polyclonality prior to Bu with 68, 75 and 63 independent integrations in granulocytes, T and B cells, respectively, including two independent clones with retroviral insertions in the Mds1/Evi1 locus, previously implicated in clonal expansion and preferential engraftment in granulocytes. Following repetitive Bu, the clonal diversity decreased markedly, the Mds1- Evi1 clones disappeared, but instead a highly dominant clone emerged, characterized by a vector insertion within intron 1 of the HDAC7 gene, which was not dominant prior to repetitive Bu, accounting for 3% of shotgun cloned LAM-PCR insertions, and constituted 71% and 94% of the cloned sequences, 2.5–2.8 years post Bu, respectively. The overall number of independent clones had decreased to only 3–9 in granulocytes at the same time points. Surprisingly, the Neo^R level increased up to 11%–28.2% 2.5 and 2.8 years post Bu compared to 0.1% prior to Bu (p =0.006 and 0.003) only in granulocytes whereas the Neo^R level was not significantly changed in T and B cells (1%–2%), indicating that the clone with retroviral insertion within HDAC7 gene became highly dominant and dramatically expanded following Bu-induced stress. Compared with non-transplanted animals, HDAC7 expression measured by quantitative RT-PCR was 4.3-fold increased in CFU enriched for the HDAC7 clone by G418 selection, but a similar increase was also detected in CFU not enriched for the HDAC7 clone, indicating no upregulation or downregulation of HDAC7 expression. While there has been no report about direct relevance of HDAC7 to cancers, its association with various oncogenes and tumor suppressor genes has been well established. No evidence of malignancy and other hematopoietic disorders has been detected in RQ2297 to date. Those data indicated immortalization or preferential survival and engraftment of cells with retroviral insertion within HDAC7A locus. For RQ2314, the number of clones with RIS decreased to 12 six months post Bu administration from 80 prior to Bu in granulocytes, however no highly dominant clones have emerged yet. The Neo^R level in granulocyes 6 months post Bu increased to 7.5% from 2.7% prior to Bu (p=. 0.038), indicating that some clones expanded. The data suggest that repetitive busulfan administration induced emergence of dominant and consequently expanding hematopoietic clones with retroviral insertion in rhesus macaques. This model could both accelerate analysis of genotoxicity related to vector insertions, as well as help assess the safety of administering myelotoxic chemotherapy in patients previously engrafted with vector-containing cells.