ADENOVIRALLY MEDIATED LONG-TERM EXPRESSION OF THROMBOPOIETIN - A NEW MODEL FOR MYELOPROLIFERATIVE SYNDROME AND OSTEOMYELOFIBROSIS

Using a new adenoviral vector (Ad) construct, we expressed human throm boprotein (TPO) cDNA (AdTPO) in mice with various inherited immune def iciency syndromes such as nude, SCID and NOD-SCID mice. Immune normal Balb/c mice and a vector construct without TPOcDNA (AdNull), respectiv ely, were used for controls. All animals (3 per group) were treated wi th a single application of 10(9) PFU (plaque forming unit) of Ad (AdTP O or AdNull) intraperitoneally on day 0. Four to 5 weeks following AdT PO administration, SCID and NOD-SCID mice demonstrated peak concentrat ion of PLT of 12- to 14-fold normal value simultaneously with maximum concentration of PMNs (10- to 12-fold normal value). Later on these an imals had a chronic thrombocytosis. In contrast, Balb/c mice and nude mice experienced PLT peak concentration of 4- to 6-fold normal value w ithout granulocytosis 1 to 2 weeks following AdTPO treatment. Only nud e mice had chronically elevated PLTs. In contrast, Balb/c mice develop ed thrombocytopenia due to cross-reacting anti-TPO antibodies. Animals with chronic thrombocytosis revealed increased content of CFUG/GM, CF U-GEMM and CFU-Meg in bone marrow compared with controls. In contrast, Balb/c mice showed decreased content of CFUs if anti-TPO-antibodies w ere present. Histologically, only SCID mice developed severe osteomyel ofibrosis and osteomyelosclerosis, hepato-splenomegaly, extramedullary hematopoiesis in liver and lung and ultimately suffered of progressiv e pancytopenia, anisocytosis, fragmentocytosis and a lethal wasting sy ndrome. In contrast, NOD-SCID mice which demonstrated similar extent o f TPO overexpression and in addition to the B- and T-cellular immune d eficiency harbour defective monocytes and macrophages, did not develop fibrotic changes of the bone marrow. From these results, we conclude (1) chronic TPO overexpression in vivo may lead to thrombocytosis and granulocytosis with expansion of CFU-GM, -GEMM and -Meg; (2) in vivo e xpression of adenovirally mediated TPOcDNA depends on immune competenc y of the host; (3) functionally normal monocytes and macrophages are i ndispensable for development of secondary osteomyelofibrosis and (4) a denovirally mediated expression of xenogeneic transgenes may brake imm une tolerance for the respective self protein leading to autoimmune ph enomena. Our in vivo model might provide further insights into the pat hophysiology of secondary osteomyelofibrosis and may prove useful in d esigning new strategies for immune therapies of cancer.