Engraftment Of Human Polycythemia Vera CD34+ Cells In hSIRPα-Transgenic-Human-TPO-Expressing RAG2^-/-, IL2Rγ^-/- Immunodeficient Mice

Xenotransplantation of human hematopoietic malignancies into immunodeficient mice represents the most appropriate in vivo model system for human malignant hematopoiesis. While a diversity of immunodeficient mouse strains on the NOD/SCID and RAG2^-/-/IL2Rγ^-/--BALB/C background are described for the most aggressive human malignancies like acute leukemias, xenotransplantation models of less aggressive human hematologic disorders like myeloproliferative neoplasms and myelodysplastic syndromes show only limited engraftment levels. We recently developed next generation mouse strains expressing human cytokines and key factors of xenogeneic cell acceptance (e.g. hSIRPα to inactivate mouse macrophage activation by human cells) and hypothesized that these would represent suitable models for the assessment of human less aggressive hematologic disorders in vivo by providing an optimized “humanized” microenvironment.

Peripheral blood (PB) and bone marrow (BM) samples of polycythemia vera (PV) patients were collected after informed consent at the Division of Hematology, Zurich University Hospital. Human CD34+ cells were isolated by density gradient centrifugation followed by immunomagnetical selection using anti-CD34 coupled beads. Purity of magnetical selection process was confirmed by FACS analysis. Newborn (24h-48h old) hSIRPα-tg-hTPO-knockin mice on the RAG2^-/-/IL2Rγ^-/--BALB/C background received sublethal irradiation (split dose of 2x1.5 Gy) and were transplanted intra-hepatically 24 hours later. Transplanted cell dose was dependent on availability of CD34+ stem and progenitor cells isolated from one phlebotomy sample (∼400ml of PB) of the respective patient. Mice were bled 4 weeks after transplantation and chimerism in peripheral blood was analyzed by flow cytometry using a panel of antigens (mCD45.2, hCD45, hCD33, hCD34, hCD3, hCD19). Mice showing positive chimerism in PB (i.e.>0.1% hCD45+ of total MNCs) at week 4 were sacrificed between week 8-16 and engraftment in BM, spleen and PB was analyzed by flow cytometry. To verify engraftment of human malignant hematopoiesis we quantified allele-burden of JAK2V617F point mutation in mouse BM using allele specific (AS)-PCR for the pathognomonic point mutation of the JAK2 gene.

By transplantation of 4-10x10⁵ CD34+ cells into newborn hSIRPα-tg-hTPO-RAG2^-/-/IL2Rγ^-/- mice we could detect engraftment of hCD45+ cells in PB at week 4 (median 0.68%, range 0.12-23.8%). At week 8, BM engraftment of hCD45+ cells ranged 0.88-54.1% (median 4.43%) with a high proportion of human myeloid cells detected by hCD45/hCD33 co-staining (median 3.07 %, range 0.6-17.6% of total MNCs). We could detect engraftment until week 16, the latest timepoint assessed. Since all transplanted PV patient samples were positive for the common point mutation JAK2V617F, AS-PCR was used to quantify human malignant hematopoiesis. In tested BM samples of engrafted mice we found JAK2V617F positive alleles with a frequency of 2-12% (median 8%). To further assess the clonal composition of the engrafted population we established single cell sorting of primary and engrafted human PV-CD34+ cells in a 96 well format followed by liquid culture expansion and AS-PCR. In pilot studies we could show the clonal composition of a BM engrafted CD34+ population that split into 80% JAK2 WT expressing, 10% JAK2V617F heterozygous and 10 % JAK2V617F homozygous clones. We are currently extending these findings by side by side comparison of the clonal composition of primary vs. xenografted human PV-CD34+ cells of the same patient to test for the influence of a xenogeneic humanized microenvironment on maintenance of malignant cells in vivo.

By using hSIRPα-tg-hTPO-RAG2^-/-/IL2Rγ^-/- mice we could show engraftment of PV-CD34+ cells that extends previous reported engraftment levels in other model organisms. To our knowledge this is the first study assessing the clonal composition of human PV engrafted cells in the xenogeneic environment aiming at identifying components that are critical for the maintenance of human malignant hematopoiesis in vivo. This model will thus be a useful tool to test targeted therapies in vivo.

No relevant conflicts of interest to declare.