Typically, a high number of human CD34+ cells is required to achieve sufficient engraftment of human hematopoietic cells in various organs of immunodeficient mice via total-body irradiation during humanized mouse generation [10,11,12]. Lang et al. reported that at least 50,000 umbilical cord blood-derived human CD34+ cells were required to reconstitute an adequate population of human hematopoietic cells in irradiated BALB/c-Rag2null Il2rgnull mice [13]. However, the optimal number of CD34+ cells varies depending on the specific immunodeficient mouse strain, indicating a strain-dependent variability in the engraftment capacity of human hematopoietic cells. In this study, we aimed to identify the optimal number of CD34+ HSCs for reconstituting human hematopoiesis in NOGW mice, with or without total-body irradiation. Initially, we assessed the sensitivity of NOGW mice to irradiation without transplantation. The survival ratios of NOGW mice up to a 20-day period was 69% for females and 57% for males after exposure to 1.5 Gy irradiation (Figure S1). In contrast, most NOG mice survived at the same irradiation dose. However, all NOGW mice survived exposure to 1 Gy irradiation. Therefore, 1 Gy was determined to be the optimal irradiation dose for the NOGW strain. Irradiated NOGW mice exhibited sufficient engraftment of human CD45+ cells into PB after the transfer of at least 10,000 CD34+ cells. Efficient hematopoiesis was also observed in non-irradiated NOGW mice following the transfer of at least 20,000 CD34+ cells (Fig. 1A). Subsequently, we compared human hematopoiesis in the BM of irradiated NOG and non-irradiated NOGW mice. The frequency of human CD45+ cells was significantly higher in the BM of NOGW mice than in that of NOG mice (Fig. 1B). In addition to the high frequency of hCD45+ cell engraftment, the BM of NOGW mice showed a marked increase in hCD66b+ granulocytes and hCD41+ megakaryocytes compared to the BM of NOG mice (Fig. 1B). Although the frequencies of CD34+CD38+ and CD34+CD38− cells in the lineage (lin)-negative cell population remained unaltered in the BM of NOG and NOGW mice, the cell numbers of these populations were significantly higher in NOGW mice than in NOG mice (Fig. 1C). Stem cell factor (SCF), also known as the c-kit ligand, plays an essential role in promoting self-renewal and development and maintenance of HSC in the BM [14]. Owing to the high homology between humans and rodents, murine SCF binds to human c-kit and may promote the self-renewal of human HSC [15]. We compared the protein levels of murine SCF in the serum and BM fluid of NOG and NOGW mice without humanization. NOGW mice exhibited significantly higher levels of murine SCF in both serum and BM fluids than NOG mice (Fig. 1D). This observation suggests that murine SCF remains present in the blood and BM fluid, given its inability to bind to mutant c-kit. Importantly, these findings indicate that NOGW mice may have a greater potential for human cell engraftment, supporting the sustained presence of HSC or progenitor cells. Therefore, similar to previously established W41 strains [7, 8], the NOGW mouse model offers an irradiation-free approach to achieve efficient engraftment of human hematopoietic cells and reduces the number of CD34+ HSCs required for transplantation.

Human hematopoiesis in NOGW mice. (A) HSC-limiting dilution assay in NOGW mice. Either 40,000, 20,000, 10,000, or 5,000 CD34+ HSCs were transferred into 1-Gy-irradiated or non-irradiated NOGW mice (n = 4–5). The chimeric ratio of human CD45+ cells in the PB was analyzed by flow cytometry at indicated time points. (B) Frequencies of human CD45+ cells, CD19+ B cells, CD33+ myeloid cells, CD66b+ granulocytes, and CD41+ megakaryocytes in the BM of humanized NOG and NOGW mice 16 weeks after HSC transplantation (n = 5). (C) Frequencies of human CD34+CD38− cells and CD34+CD38+ cells in lineage−CD45+ cells, and number of total leukocytes, lineage−CD45+ cells, CD34+CD38− cells, and CD34+CD38+ cells in BM of humanized NOG or NOGW mice. (D) Mouse SCF protein levels in serum (NOG; n = 19, NOGW; n = 18) or BM fluid (NOG; n = 13, NOGW; n = 22) of non-humanized NOG or NOGW mice. *p < 0.05; **p < 0.01; ****p < 0.001

To achieve the differentiation of multiple cell lineages with advanced engraftment abilities, we backcrossed NOGW mice with NOG-EXL mice to generate NOGW-EXL mice. Then, we compared the reconstitution capacity of human hematopoietic cells in NOG, NOG-EXL, NOGW, and NOGW-EXL mice. At 12 and 20 weeks after CD34+ cell transplantation, non-irradiated NOGW-EXL mice exhibited higher rates of human CD45+ cell engraftment than irradiated NOG, NOG-EXL, and non-irradiated NOGW mice (Fig. 2A). NOGW-EXL mice showed higher reconstitution of human CD66b+ granulocytes, which are human eosinophils in most CD66+ cells [4], and CD41+ platelets than the other strains (Fig. 2A). Compared to NOG and NOGW mice, NOGW-EXL and NOG-EXL mice showed an increase in human CD33+14+ monocytes (Fig. 2A). We analyzed the sex-based differences in human cell engraftment in each strain of humanized mice. In most cases, no differences in human cell engraftment were observed between male and female mice. However, female NOGW-EXL and NOG-EXL mice exhibited significantly higher levels of CD45+ cells compared to their male counterparts at 12 and 20 weeks, respectively. This result is consisted with a previous report that the female NSG humanized mice show higher engraftment capacity of human CD45+ cells than male humanized mice [16]. In the spleen, the chimeric ratio of human CD45+ cells was higher in NOG-EXL, NOGW, and NOGW-EXL mice than that in conventional NOG mice, and total human myelopoiesis was significantly enhanced in NOG-EXL and NOGW-EXL mice (Fig. 2B). Tissue-resident human CD68+ macrophages exhibited significant engraftment in the liver and lungs of NOGW-EXL mice (Figures S3A and S3B). Next, we examined the development of human megakaryocytes in the NOGW-based strains. Previously, NSG-W41 and NBSGW mice displayed enhanced reconstitution of human megakaryocytes in the BM, with a detectable level of human platelets in PB [17, 18]. Similar to the previous W41 strain, NOGW mice displayed increased differentiation of human megakaryocytes compared to NOG mice. Among these strains, human megakaryocyte differentiation was significantly elevated in NOGW-EXL mice (Fig. 2C, D). Based on immunohistochemical analysis, the highest engraftment of human CD61+ megakaryocytes was observed in the BM of NOGW-EXL mice, which was further supported by flow cytometry studies (Fig. 2E). IL-3 plays a crucial role in megakaryopoiesis and platelet production in humans [19]. Considering the systemic production of human IL-3, NOGW-EXL mice plausibly exhibited accelerated human megakaryopoiesis compared with NOGW mice, leading to increased platelet production. The minimum number of HSCs required for humanization of non-irradiated NOGW-EXL mice was evaluated. Based on a limiting dilution assay of human HSCs, the transfer of 5,000 CD34+ HSCs achieved over 20% human CD45+ cell chimerism from 8 weeks onward (Figure S4). Additionally, the transfer of 10,000 HSCs yielded approximately 40% human CD45+ cell chimerism at 8 weeks in NOGW-EXL mice. The transfer of both 5,000 and 10,000 HSCs significantly reconstituted human granulocytes and platelets in NOGW-EXL mice but not NOGW mice (Figure S4). Thus, among the HSC-transferred humanized mouse models, the humanized NOGW-EXL mouse model exhibited greater hematopoiesis and engraftment capacity than NOG-EXL and NOGW mice.

Generation of NOGW-EXL mice. (A) Frequencies of human CD45+ cells, CD19+ B cells, CD3+ T cells, CD66b+ granulocytes, CD33+14+ monocytes, CD56+ NK cells, and CD41+ platelets in the PB of humanized NOG, NOG-EXL, NOGW, and NOGW-EXL mice at 12 or 20 weeks after HSC transplantation (n = 6–16). (B and C) Frequencies of human CD45+ cells and other human cell lineages in the spleen (B) and BM (C) of humanized NOG, NOG-EXL, NOGW, and NOGW-EXL mice at 20 weeks after HSC transplantation (n = 6–16). (D) Representative flow cytometry analysis of human macrophages (Mac) and megakaryocytes (MK) in the BM of humanized NOG, NOGW, and NOGW-EXL mice 20 weeks after HSC transplantation. (E) Immunohistochemistry images of human CD61+ megakaryocytes in the bone tissue of humanized NOG, NOG-EXL, NOGW, and NOGW-EXL mice at 20 weeks after HSC transplantation. Each bone image is shown at low magnification (left) or high magnification (right) for each strain. Images are representative of three independent experiments. *p < 0.05; **p < 0.01; ***p < 0.005; ****p < 0.001

We investigated the maintenance of human HSCs in various humanized mouse strains. In the flow cytometry analysis, the frequencies of human CD34+CD38− multipotent hematopoietic progenitors and CD34+CD38+ hematopoietic progenitors in the BM of NOG-EXL and NOGW-EXL mice were significantly lower than those in the BM of NOGW mice (Fig. 3A). Additionally, CD201+, a critical marker of multipotent HSCs [20], was absent in CD34+CD38− cells in NOG-EXL and NOGW-EXL mice (Fig. 3A). This finding suggests that human IL-3 and/or GM-CSF may induce HSC exhaustion and accelerate hematopoietic maturation. We assessed the functional capacity of the engrafted human HSCs by performing serial BM transplantation in humanized mice (Fig. 3B). At all time points, the transfer of BM cells from NOGW mice to NOG mice resulted in a higher engraftment rate of human CD45+ cells than cell transfer from the other three strains (Fig. 3C). Although multilineage analysis of human hematopoietic cells showed no significant differences in leukocyte populations among the strains, significant levels of human platelets were detected 4 weeks after NOGW BM transplantation. This finding suggests that NOGW mice can more effectively preserve mature human megakaryocytes and progenitor cells than the other strains. In the BM of secondary recipients, the engraftment of CD45+ cells was the highest in the NOGW BM–transplanted group (Fig. 3D, E). This high engraftment rate may be attributed to the effective maintenance of human HSCs in NOGW mice (Fig. 3A). Conversely, secondary BM recipients who received BM transplantation from NOG-EXL and NOGW-EXL exhibited limited CD45+ cells and no CD34+CD38+ cells (Fig. 3D, E). These findings indicated that the long-term maintenance of human HSCs was more effective in NOGW mice than in human cytokine-expressing NOG-EXL and NOGW-EXL mice, despite the high rate of human myelopoiesis in these strains. Similar to this finding, previous studies have shown that NSG mice [21], which express three human cytokines—stem cell factor, GM-CSF, and IL-3—could not support the maintenance of transplanted human HSCs in the BM, owing to HSC exhaustion induced by the supraphysiological expression of these cytokines [22]. Nevertheless, these mice showed the long-term maintenance of mature lymphoid and myeloid cells, although most human HSCs were eliminated. Accordingly, maintenance of HSCs in the BM niche may not be necessary to reconstitute mature lymphoid and myeloid cells in humanized mice. Highly maintained lymphoid and myeloid progenitor cells with self-renewal capabilities may be responsible for this phenomenon in the cytokine-expressing humanized mice. However, further research is required to comprehensively clarify the implications of these findings and develop strategies to overcome the maintenance of any progenitor cells and HSC exhaustion induced by cytokine expression.

Maintenance of human HSCs in humanized mice. (A) Frequencies of human CD34+CD38+ cells and CD34+CD38− cells in lineage−CD45+ cells and the frequency of CD201+ cells in CD34+CD38− cells in the BM of humanized NOG, NOG-EXL, NOGW, and NOGW-EXL mice at 16 weeks after HSC transplantation. (B) Schema of the procedure for serial BM transplantation from humanized NOG, NOG-EXL, NOGW, and NOGW-EXL mice. (C) Frequencies of human CD45+ cells and other human lineage cells in the PB of secondary-recipient NOG mice after transfer of BM cells from humanized NOG, NOG-EXL, NOGW, and NOGW-EXL mice 4, 12, and 20 weeks after HSC transplantation (n = 4–8). (D) Frequencies of human CD45+ cells and CD34+CD38+ cells in the BM of secondary-recipient NOG mice 24 weeks after transfer of BM cells from humanized NOG, NOG-EXL, NOGW, and NOGW-EXL mice (n = 4). (E) Representative flow cytometric data of human CD45+ cells and CD34+CD38+ progenitor cells in D. *p < 0.05; **p < 0.01; ***p < 0.005; ****p < 0.001