Optimization of stem cell fitness and mobilization using moderate transfusion and an oral anti-sickling agent in the sickle mouse model. Free

Ex vivo gene therapy (GT) can transform the lives of people with sickle cell disease (SCD). However, GT relies on mobilization of sufficient quality and quantity of hematopoietic stem cells (HSCs). HSC quality is compromised by stress hematopoiesis and inflammation of the sickle bone marrow (BM), and HSC mobilization is reduced by the disordered BM stroma and vasculature. Only the weaker mobilizer, plerixafor, can be used in SCD; granulocyte colony-stimulating factor cannot be safely used due to risk of inducing sickle-related complications. Currently, disease modification with transfusions to a hemoglobin S (%HbS) goal of <30% is required to optimize stem cell collection but exposes individuals to significant number of blood units in a population at high risk for alloimmunization. Therefore, we assessed if a moderate chronic transfusion therapy (CTT) regimen with a HbS% goal of <50% or an oral anti-sickling agent osivelotor can modify stress hematopoiesis and HSC fitness in the Townes sickle mouse model. Further, we assessed the impact of osivelotor on stem cell mobilization as a potential alternative to CTT for individuals with SCD who cannot be safely transfused and would otherwise be excluded from GT.

We treated 7-to-9-week-old HbSS mice with 0.4% osivelotor chow for 4-5 weeks or provided CTT (%HbS goal <50%) for 8 weeks and quantified the hematopoietic stem progenitor cell (HSPC) population from the BM by flow cytometry. Compared to untreated mice, osivelotor-treated mice had higher Hb (median 16 vs 8.3 g/dL, p<0.001). BM cellularity was similar between the osivelotor-treated and untreated groups. Lin-Sca-1+c-Kit+ (LSKs) (3471 vs 6679, p=0.05) and multipotent progenitor (MPP)-3 numbers (737 vs 2046, p=0.06) trended lower in osivelotor-treated mice while MPP4 (816 vs 1520, p=0.04), short-term (72 vs 120, p=0.03), and long-term HSCs (LT-HSCs) (176 vs 391, p=0.03) were significantly lower compared to untreated controls suggesting a decrease in stress hematopoiesis. In the CTT arm, the median Hb and %HbS over 8 weeks were 9.9 g/dL and 41.2%. There were no significant differences in BM cellularity, LSKs or HSPC subpopulations between the transfused mice and controls.

We then performed competitive repopulation assays to assess the fitness of these disease-modified HSPCs. Mononuclear BM cells were pooled from HbSS mice (CD45.2) treated with transfusions or osivelotor and mixed with competitor BM cells from non-sickle JaxBoy mice (CD45.1) and transplanted into lethally irradiated JaxBoy recipients. CD45.2% engraftment rates were measured by flow and compared to mice similarly transplanted with untreated sickle BM cells. CD45.2% engraftment rates were similar between groups transplanted with osivelotor-treated BM (84.5%) and untreated sickle BM (92.4%) at 25 weeks post-transplant (p=0.42). In the CTT arm, the engraftment rates were significantly higher at 100% compared to 67% in the controls (p=0.01). When considering the number of LT-HSCs transplanted, osivelotor HSCs were 1.33-fold (p=0.03) more efficient in engraftment and CTT HSCs were 1.75-fold (p=0.01) more efficient than HSCs from untreated controls.

Lastly, we assessed whether treatment with osivelotor improved HSC mobilization. HbSS mice were treated with 0.4% osivelotor chow for 8-weeks and mobilized with plerixafor. White blood cell counts and LSKs from peripheral blood (PB) were higher in plerixafor mobilized mice (30e3/ul p=0.23, and 2.13e3/ul p=0.49), and highest when conditioned with osivelotor (40e3/ul p=0.002, and 5.57e3/ul p=0.01) compared to untreated controls (15.27e3/ul and 0.45e3/ul). We then cultured mononuclear cells isolated from PB and found colony forming units were higher with cells mobilized from osivelotor-conditioned mice compared to controls (25.75 vs 4.5, p=0.002).

Disease modification with osivelotor improved anemia, stress hematopoiesis, and fitness of the HSCs. Further, the addition of osivelotor conditioning to plerixafor mobilization improved the number of mobilized LSKs in the PB. Although moderate CTT did not significantly change HSPC populations compared to controls, likely due to persistent anemia, CTT did modify the fitness of the HSCs resulting in higher engraftment efficiency. A weakness of our study is the lack of comparison to standard CTT goal of <30%. Less aggressive transfusion goals and an oral non-myelosuppressive disease modifying agent are potential alternatives to traditional CTT to optimize ex-vivo GT for SCD.