In this study, we used B-NSG mice with an interleukin-2 receptor gamma chain (IL2Rγ) deletion and nearly complete absence of the murine immune system to construct an HLL model, to improve HLL cell engraftment and reduce graft versus host disease [16, 17]. The above experimental results showed that we successfully established a CD34+ hematopoietic stem cell-derived xenograft model of human HLL in B-NSG mice. First, a large number of HLL cells could be detected in blood by transplanting cells into the mice, and leukemic cells were observed based on cell morphology in the peripheral blood and bone marrow of mice. Then, a higher ratio of the immunophenotype of leukemic cells was detected by flow cytometry, which was consistent with the immunophenotype of the patients. Finally, the leukemic cells were severely infiltrated into other tissues, and gene mutations were detected in infiltrating tissues in the CD34+ hematopoietic stem cell-derived xenograft model.
According to previous reports, 1.5–3.5 Gy doses of χ-ray radiation have been used before leukemia cell [18–20] or CD34+ cells [21, 22] transplantation into mice, but only a few studies did not use irradiation [23]; for example, Shafat et al. reported the successful construction of a primary AML model with nonirradiated NSG mice [24]. In this study, we explored the radiation dose in model construction. For the 2.5 Gy χ-ray irradiated group, the body weight loss of mice was faster, and the survival time was shorter than that of the 0 Gy irradiated group, which indicated the 0 Gy irradiated mice could better used for drug research. The body weight of mice did not decrease much, and their body condition was good in the 0 Gy irradiated group; however, the mice in the 2.5 Gy group survived approximately 17 days and were in poor condition, dying easily, which is unfavorable for the study of drug efficacy or the mechanism of HLL development. Importantly, irradiation can briefly result in serious myelosuppression[25] and lead to an inability to clearly observe leukemic cells in blood smears and bone marrow smears, making it difficult to assess whether the model was successful. Moreover, myelosuppression seriously affects routine blood tests and cannot truly allow the evaluation of changes in WBCs [26, 27]. Our results also showed that myelosuppression caused by irradiation could affect the survival time of mice and possibly lead to no significant difference in survival time between the 1.5 Gy irradiated model mice injected with cells derived from the PB and BM of patients.
In our previous reports on drug screening using a PDX model of HLL, B-NSG mice in the 2.5 Gy group survived beyond 20 days when injected with other HLL cells without malignant gene mutations [28], and B-NSG mice without irradiation survived beyond 31 days when injected with human AML cells separated from a hyperleukocytic AML-M5 patient with NPM1 and DNMT3A mutations [29]. In addition, the B-NSG mice with 1.5 Gy irradiation survived beyond 20 days when injected with human AML cells separated from a hyperleukocytic AML-M5 patient with WT1, DNMT3A and FLT3 mutations [30, 31]. The mutational status of AML cells is also a key factor in immunodeficient mice for the construction of a PDX model [32, 33]. In this study, we found that the constructed model mice injected with NRAS gene mutations had shorter survival times than the injected with cells lacking the NRAS gene mutations, and the same results were observed for DNMT3A, FLT3, and NMPM1 gene mutations. The survival time in mice had significant correlation with the survival status of enrolled patients. Therefore, our results suggest that using B-NSG mice for HLL PDX construction is dependent on the use of HLL cells from patients who have malignant gene mutations and poor prognosis and is also related to the injected cell counts and the irradiation dosage in mice.
To establish a CD34+ hematopoietic stem cell-derived xenograft model with human hyperleukocytic AML cells, we found that the use of nonirradiated B-NSG mice injected with CD34+ cells from HLL patients without malignant gene mutations was not successful, and until the 50th day, leukemic cells were not found in blood smears or bone morrow smears. In this study, B-NSG mice were treated with 1.5 Gy χ-ray irradiation and injected with CD34+ cells derived from a HLL patient with WT1 and NRAS mutations, two mutations that frequently occur in AML.[34] The CD34+ cell-derived PDX model of HLL was established successfully, which also indicated that B-NSG mice with low-dose radiation were better used for model construction based on injection of cells with moderate malignant mutations.
However, the HLL model still has some limitations; for example, HLL cells from patients show difficulty proliferating in vitro, and the mice do not present full clinical characteristics, such as certain immunophenotypes, so the PDX model of HLL is still not standardized and requires further investigation, and we are still unable to replicate the complex immune microenvironments in immunodeficiency mice [35], so it is difficult to explore the complex heterogeneity of HLL and the interaction and progression between HLL cells and their microenvironment [14, 36]. Further related experiments will be investigated. Even so, the HLL mouse model is still of great value in the study of the pathogenesis and drug efficacy.
In conclusion, we transplanted hyperleukemic leukemia cells into severe immunodeficient B-NSG mice, which retained most of the original characteristics in the B-NSG mice, and more accurately simulate the characteristics of tumor growth in clinical patients. we assessed the construction methods used for HLL model, optimized in vivo model establishment, and successfully established the CD34+ hematopoietic stem cell-derived xenograft model of HLL in B-NSG mice, and provide a meaningful model for mechanistic research, drug screening, individualized therapy, clinical efficacy assessment and precision medicine in HLL.