Engraftment of Human MDS Samples in Xenotransplanted Mice Depends on Human Cytokines but Not Mesenchymal Stem Cells (MSC)

Background: Myelodysplastic syndromes (MDS) are a heterogeneous group of diseases with limited treatment options. Development of a patient derived xenotransplantation model for pre-clinical studies is a priority goal in the field. It has been hypothesized that patient derived mesenchymal stem cells and/or human cytokines are necessary to establish MDS engraftment in immunocompromised mice. In this study we evaluated the ability of mesenchymal stem cells (MSC) to facilitate the engraftment process having as a goal to generate a robust patient-derived xenograft (PDX) mouse model for MDS. We also assessed the contribution of human cytokines as expressed in transgenic mice.

Methods: MSC from normal donors and MDS patient samples were generated and characterized using standard culture methods. NOD.Cg-Prkdcscid
Il2rgtm1Wjl/SzJ (NSG) and NOD.Cg-Prkdcscid Il2rgtm1Wjl Tg(CMV-IL3,CSF2,KITLG)1Eav/MloySzJ (NSG-S) mice were used as recipients. Mice were injected with 1 x 10⁶ bone marrow mononuclear cells (MNC) alone or in combination with 5 x 10⁵ MSC. MSC were from normal donors, from same patient as MNC (autologous), from other MDS patients or normal donors (allogeneic). All cells were injected intrafemorally. Mice were assessed for engraftment by bone marrow aspiration at 6-8, 10 weeks and at necropsy at 12-32 weeks (depending on initial assessment). Long-term engrafted samples were further assessed for presence of stem cells by secondary transplantation into NSG-S mice. 1 x 10⁶ human CD45+ cells were passaged without or with 5 x 10⁵ MSC. Secondary recipients were assessed at 12 weeks for engrafment. MDS samples represented both high risk (refractory anemia with excess blasts, RAEB)) and low risk (refractory anemia (RA), refractory cytopenia with myelodysplasia (RCMD), chronic myelomonocytic leukemia (CMML)) disease. The overall degree of engraftment was assessed by bone marrow aspiration and measurement of human CD45+ cells by means of FACS. Lymphoid, myeloid and erythroid engraftment was assessed similarly. 5 x 10⁵ MSCs from normal donors were transduced with GFP (pELNS.CBR-T2A-GFP) and intrafemorally injected into NSG mice and the presence of human MSC was evaluated by in vivo Imaging (Xenogen Spectrum system and Living Image Version 4.3 software.).

Results: 12 out of 12 injected MDS samples showed persistent human cells when assessed at 6-8 weeks. For most samples, engraftment at that time-point was low (<2%) and was not consistently influenced by the presence or absence of MSC. On week 10, 4 out of 12 injected MDS patient samples showed increased engraftment and only one showed higher engraftment levels in the mice co-transplanted with MSC. Two of these were from patients with RAEB and CMML and showed both myeloid and erythroid engraftment. Molecular analysis showed consistent engraftment of the malignant clone in 2 out of 2 patient samples tested. Overall, only 3 out of 12 patients showed long-term engraftment (> 12 weeks) and these samples were transferred to secondary animals. Secondary transplanted mice injected with selected hCD45+ cells with or without MSC showed variable engraftment levels on week 10 after injections. One of them reached long-term engraftment (>12 weeks), whereas the second gradually lost engraftment after week 10 and the third is ongoing. MSC tracking showed a gradual loss of MSC after intrafemoral injections to the point that no MSC were detected 3 weeks later. Comparison of engraftment in NSG mice to NSG-S mice (the latter producing human cytokines) showed a slight increase in engraftment at week 6 which continued at later time points.

Conclusions: Passive transfer of MDS hematopoietic cells as assessed at 6-8 weeks after intrafemoral injection is highly consistent and not dependent on MSC when total MNC fraction is used. The latter may be explained since MSC are lost within the first three weeks after injections. Direct comparison of NSG vs. NSG-S mice showed that the presence of hGM-CSF, hIL-3 and h-SCF in the NSG-S mice increased the engraftment levels. Successful long-term engraftment of MDS cells in xenotransplanted mice was achieved for one patient sample so far, indicating that MDS initiating cells can be maintained in NSG-S mice. However, establishing long-term engraftment can only be achieved with a subset of patient samples. MDS engraftment is influenced by human cytokines but not with MSC cells which may suggest that MDS does not depend on MSC's for long term growth.