Hematopoietic Stem Cells from Mice and Individuals with Sickle Cell Disease Display Premature Senescence and Loss of Function That Is Targetable By Senolytic Therapy

Abstract

Sickle cell disease (SCD) is an inherited anemia caused by mutations in the β-globin gene, resulting in aberrant hemoglobin polymerization driving hemolysis, vasooclusion, and systemic inflammation. Allogeneic hematopoietic stem cell (HSC) transplantation is potentially curative, and novel autologous therapies involving gene-editing of mobilized HSCs are being developed. However, many SCD patients are unable to undergo autologous gene-editing due to failure to mobilize sufficient numbers of HSCs. Further, the damaging effects of SCD on the HSC pool remain uncharacterized. Therefore, we interrogated young and aged cohorts of humanized SCD mice and non-SCD controls for HSC frequency and function. Immunophenotypic HSC frequency was mildly elevated (1.3-fold, p=0.00316) in 2-month-old SCD mice, relative to controls (n=5-6/group). In contrast, HSC numbers were halved in 6 and 12-month-old SCD mice (n=8-10/group, p=0.00358 and p=0.0148, respectively). Competitive transplantation of whole bone marrow (WBM) from 2 and 6-month-old SCD mice displayed a profound loss of peripheral blood repopulation, relative to age-matched control mice. Quantitative limiting dilution transplantation of WBM from 2 and 6-month-old SCD mice revealed a 4 and 6-fold loss of functional HSCs, relative to control mice. Thus, despite elevated numbers of phenotypic HSCs in young SCD mice, the HSC pool is depleted of functional repopulating cells during SCD. To understand the loss of functional HSCs, we performed bulk RNA-sequencing of purified HSCs from 6-month-old SCD mice. We observed downregulation of 37 genes encoding ribosomal and histone proteins, suggesting impaired histone/ribosome biogenesis, a phenotype associated with senescence. Consistently, pathway enrichment analysis also revealed perturbed p53 activity and a high degree of association with published molecular signatures of senescence. To explore this further, HSCs isolated from 6-month-old SCD mice or age-matched controls were interrogated for senescence associated β-galactosidase (SA-β-gal) activity. On average 19.5% of SCD HSCs were SA-β-gal+ versus only 7.06% of HSCs isolated from controls (n=10-11/group, p=0.004). HSCs isolated from SCD mice also displayed elevated ROS, DNA damage, and expression of molecular markers of senescence (e.g. p21, BCLXL). Importantly, 23.14 ± 5.61% of Lin-CD34+CD38- bone marrow (BM) hematopoietic stem and progenitor cells (HSPCs) isolated from pediatric individuals with SCD were SA-β-gal+. In contrast, only 4.97 ± 2.97% of HSPCs were SA-β-gal+ in age-matched, non-SCD control individuals (n=8 control and n=5 SCD individuals, p=<0.0001). Further, 47.46 ± 17.49% of Lin-CD34+CD38-CD90+CD45RA- BM cells, which are highly enriched for human HSCs, were SA-β-gal+ in young individuals with SCD, relative to only 4.41 ± 2.37% in the BM of age-matched control individuals (n=4 control and n=5 SCD individuals, p=0.0019). We also detected elevated expression of p16, p21, and BCL2 in BM HSPCs isolated from pediatric individuals with SCD, relative to controls (n=6 control and n=9 SCD individuals). In total, these data reveal that SCD pathophysiology aberrantly invokes senescence programs and damages HSC function in mice and humans. Clearance of senescent HSCs from mouse BM has previously been shown to restore aging-related HSC dysfunction. We therefore tested if clearance of senescent HSCs from the BM could restore function to the SCD HSC pool. 2-month-old SCD and control mice were treated with two weekly cycles of 75 mg/kg/day of the senolytic, ABT-263, via oral gavage. 14 days after treatment, WBM was isolated and tested for HSC function and frequency. Senolytic therapy significantly increased HSC numbers in SCD mice BM, relative to vehicle-treated controls (n=5/group). Importantly, the in vivo hematopoietic repopulating activity of WBM from ABT-263-treated SCD mice was comparable to non-SCD controls, revealing a correction of our previously observed severe transplantation defect (n=12/group). In sum, these data reveal premature onset of senescence as a previously unappreciated phenotype of the HSC pool during SCD. This work provides the first proof-of-principle that treatment of individuals with SCD using senolytics may restore function to the HSC pool, and ongoing preclinical work aims to test if senolytic therapy enhances the yield and function of SCD HSCs mobilized for autologous therapy.