Xenografting immunodeficient mice has been employed as a surrogate human hematopoietic stem cell (HSC) assay, however, erythroid output has not been reliably reported, limiting the usefulness of this model for erythroid disorders. We have previously demonstrated that busulfan preconditioning is sufficient to produce stable, high level engraftment of human cells in NOD/SCID/IL2Rγ null mice. Importantly, this high level engraftment can be achieved with low mortality, substantially reducing the number of animals required for experiments requiring long-term follow-up. Supplementation with human holo-transferrin (Tf) allows the detection human erythrocytes in this chimeric mouse assay at low levels, providing a potential model for the study of disorders affecting human red blood cells (2007 ASH meeting #3594). In the current work, we extend our observations and establish practical in vivo erythroid assay system of human HSCs in the humanized mouse model by the addition of an in vitro culture.

Bone marrow (BM) from humanized mice containing 24.8±8.7% human cells (n=6) was first exposed to recombinant human (rHu) SCF+IL3 for 3 days in order to specifically enrich for human cells in the mixed chimera. After 3 days, human cells comprised 68.1±8.5% of the culture. The enriched cells were then cultured with rHu EPO+SCF+TGF-β for 7 additional days and then in rHu EPO+SCF for 7 days. After culture, and 98.9±1.47% of cells were of human origin. After centrifugation, the pellets were visibly red. Cells were assayed with both human CD71+ and GPA+ by flow cytometry: 64.0±7.44% were CD71+ and 69.2±7.02% were GPA+, and human α, β, and γ globin were confirmed by hemoglobin electrophoresis and mass spectrometry.

In order to determine the utility of this approach, we tested 3 possible applications of this methodology: gene marking erythroid progeny, modeling of human hemoglobinopathies, and modeling of hemoglobin switching. We first transplanted human cord blood (CB) CD34+ cells after lentiviral transduction with a vector encoding GFP following busulfan conditioning. Three months post-transplant, bone marrow was harvested and placed in the in vitro culture. After in vivo culture, 98.9% of cells were of human origin and 60.7% were CD71+ and 72.3% were GPA+. The majority of CD71+ or GPA+ cells were GFP+ (82.3% and 87.7%, respectively).

We subsequently transplanted human PB CD34+ cells derived from individuals with sickle cell trait (SCT) as we have previously demonstrated that these cells, unlike those from individuals with sickle cell disease, can be safely mobilized and processed. Further, the percentage of HbS expressed can be reliably measured. Three months post transplant, HbS was easily detectable by hemoglobin electrophoresis.

Finally, we sought to address whether this model accurately reflects human erythropoiesis by examining hemoglobin switching after transplanting either CB expressing HbF or PB HSCs expressing HbA and monitoring the output of HbF and HbA over time. Early post-transplant, bone marrow derived from CB recipients expressed predominantly HbF after culture whereas that derived from PB HSC recipients expressed predominantly HbA. HbF declined during follow up and was replaced by HbA over 6 months of follow up from CB recipients, whereas HbA expression remained stable from PB HSC recipients. The time course of hemoglobin switching is similar to human ontogeny.

In summary, our practical approach to model human erythropoiesis in the xenograft mouse should prove useful in the both the study of human erythroid disorders as well as therapeutic interventions.

Disclosures: No relevant conflicts of interest to declare.

Author notes

Corresponding author