FLT3 ligand facilitates long-term ex vivo expansion of human hematopoietic stem cells by maintaining lymphoid reconstitution potential

Abstract

Abstract Hematopoietic stem cells (HSCs) are blood-forming stem cells that can reconstitute the entire blood system after transplantation, which provides curative options for both blood and non-blood diseases. Despite their known utility, broader applications of human HSCs are hampered by the lack of a stable culture protocol. Previously, we developed a long-term mouse HSC expansion protocol using a synthetic polymer to replace serum albumin, thus mitigating inflammatory responses that trigger HSC differentiation. Although this approach, with ongoing improvements, has been applied to human HSC cultures, achieving long-term expansion of human HSCs remains challenging. To address this, we optimized a previously published cytokine-free culture for human HSCs, which utilizes butyzamide, a thrombopoietin (TPO) receptor agonist (TPO-RA), and 740 Y-P, a PI3K activator, as replacements for TPO and stem cell factor (SCF), respectively. In a short-term (7-day) culture, we found that supplementing SCF in addition to 740 Y-P, or substituting butyzamide with lusutrombopag, an FDA-approved TPO-RA, enhanced the ex vivo expansion of phenotypic hematopoietic stem and progenitor cells (lineage⁻CD41⁻CD34⁺) from CD34+ umbilical cord blood (UCB) cells. In xenograft transplantation model using immunodeficient NSG mice, a mid-term (14-day) physioxic (5% O₂) culture supplemented with lusutrombopag and SCF, named “5LS culture”, resulted in robust expansion of engraftable human HSCs (mean human chimerism: 22.9% [14-day expanded] vs. 0.96% [unmanipulated], with multilineage reconstitution of CD33⁺ myeloid and CD19⁺ B cells; N = 7 mice, 16 weeks post-transplantation). Molecularly, SCF consistently upregulated phosphorylation of AKT, STAT5, and p38 in phenotypic multipotent progenitors/HSCs (pMPPs/pHSCs: lineage⁻CD41⁻CD34⁺EPCR⁺) across different samples, whereas 740 Y-P did not, suggesting a mechanistic insight why SCF addition improves human HSC expansion. We next extended the culture duration to 28 days. Interestingly, despite successful 14-day expansion in the 5LS culture, CD34⁺ UCB cells cultured for 28 days predominantly reconstituted CD33⁺ myeloid cells but lost their lymphoid reconstitution potential in NSG mice. Further optimization identified that FLT3 ligand (FLT3-L) supplementation in a 5LS culture (resulting in the “5LSF culture”) improved both the expansion and the maintenance of phenotypic HSCs (pHSCs: lineage⁻CD41⁻CD34⁺EPCR⁺CD90⁺ITGA3⁺) over 28-day culture of CD34+ UCB cells. Similar results were observed using CD34⁺ mobilized peripheral blood cells. Single-cell RNA sequencing with UMAP visualization and clustering analyses revealed that while HSC/MPP clusters from 5LS cultures were skewed toward granulocyte-monocyte progenitors (GMPs), this skewing was absent in the 5LSF condition. Notably, xenotransplantation demonstrated partial restoration of CD19⁺ B cell reconstitution from 28-day expanded CD34⁺ UCB cells in the 5LSF culture (mean B cell contribution among human cells: 0% [5LS culture] vs. 32.1% [5LSF culture], N = 7 mice, 16 weeks post-transplantation). These results highlight that FLT3-L is critical for maintaining multipotency of HSCs in long-term culture.To assess whether the 5LSF culture supports genetic modification, we performed lentiviral transduction and CRISPR-Cas9/AAV6-mediated targeted gene editing in CD34⁺ UCB cells. Lentiviral vectors successfully transduced the pHSC population. Targeted editing of a reporter cassette in a HLF (a transcription factor specifically expressed in HSCs) locus confirmed efficient editing in HLF⁺ HSCs, underscoring the utility of 5LSF culture for HSC genetic engineering. Finally, a limiting dilution assay using unmanipulated, 14-day-expanded, and 28-day-expanded CD34+ UCB cells enabled estimation of functional HSC frequencies and fold expansion over a 28-day culture. The 5LSF culture exhibited the highest fold expansion of functional HSCs (109.0-fold expansion) after 28 days. In contrast, the 5LS culture demonstrated greater fold expansion of functional HSCs after 14 days (60.6-fold expansion) compared to the 5LSF culture (23.1-fold expansion), which is consistent with the results from the phenotypic characterization. These findings provide evidence that human functional HSC can stably expands long-term in the 5LSF culture. In conclusion, we have developed a long-term expansion protocol for human HSCs that facilitates both clinical and research applications of these functional stem cells.