Efficiency of Homing and Engraftment Is Higher in VEGFR-3⁺CD34⁺CD38^- Cells Than in VEGFR-3^-CD34⁺CD38^- Cells in Leukemic Patients

Surviving leukemic stem cells (LSCs) after chemotherapy lead to relapses in acute myeloid leukemia (AML). Because LSCs will not be eradicated after standard chemotherapy, inhibiting the propagation of AML cells by LSCs is a major part in AML treatment. Although CD34⁺CD38^- cells in leukemia are representative LSCs, it is still difficult to distinguish them apart from normal hematopoietic stem cells (HSCs). So far, many studies have shown rapid advances in phenotypic characterization. However, heterogeneous diversity in AML patients may not allow effective eradication of LSCs. Vascular endothelial growth factor receptors (VEGFR)-3 is expressed in AML blasts and in the bone marrow (BM) environment including sinusoidal vessels. In particular, VEGFR-3 is strongly correlated with poor prognosis, leukemic cell proliferation and survival in AML. Based on previous reports, we were able to hypothesize that VEGFR-3 is expressed in LSCs and functions as a LSC marker. Here, we observed high expressions of VEGFR-3 on CD34⁺CD38⁻ cells in AML patients who received chemotherapy and further showed low homing and engraftment capacity of CD45^dim blast cells in the BM by a VEGFR-3 antagonist. In order to determine the expression of VEGFR-3 on LSCs, data was collected from 64 AML patients, 12 of whom were after complete remission (CR), as well as from 14 healthy volunteers. MNCs were first isolated and were then subjected to FACS analysis and immunocytochemistry. NOD-Scid IL2Rγ^Null mice were used for homing efficiency and engraftment in vivo. MAZ51 was used as a VEGFR-3 antagonist. FACS analysis showed that VEGFR-3 was increased on CD34⁺CD38⁻ LSC cells. (Normal vs. AML vs. CR, VEGFR-3 on CD34+CD38- cells: 8.33 ± 4.37% vs. 25.62 ± 2.46% vs. 23.46 ± 5.47%, P < 0.05). Similarly, immunocytochemistry clearly displayed the co-expression of VEGFR-3 on isolated CD34⁺CD38⁻ LSC cells, suggesting the possibility of it as a LSC marker. We checked the ability of LSCs to use colony forming units assay. VEGFR-3⁺CD34⁺ cells showed unarguably enhanced colony forming ability compared to that of VEGFR-3^-CD34⁺ cells from patients. To test whether VEGFR-3^-CD34⁺ cells are not on apoptotic procedure, annexin-V and proliferation assay with Ki67 were performed, and there was no difference in apoptotic and proliferative movement in both cells. Intending to determine homing efficiency and engraftment, CD34 and CD45 markers were used in the BM and it was discovered that sorted VEGFR-3⁺CD34⁺CD38^- cells showed significantly increased homing and engraftment efficiency compared to those of VEGFR-3^-CD34⁺CD38^- cells (by FACS, homing capacity: 1.93 ± 0.19% vs. 0.13 ± 0.06%, P = 0.02; engraftment: 26.4 ± 9.42% vs. 5.30 ± 2.31%, P < 0.05), implying that VEGFR-3 can serve as a marker for LSCs. We demonstrated that VEGFR-3 was highly expressed and enriched on CD34⁺CD38^- cells in CR status as well as in the initial diagnosis of AML. Therefore, the targeting of VEGFR-3 may diminish LSC function in human AML. These findings could suggest some clues to develop therapeutic strategies targeting VEGFR-3⁺ LSCs with favorable tumor microenvironments.