Lysosomal Activation Is Required for Priming of Quiescent Hematopoietic Stem Cells

Despite their immense in vivo repopulating capacity, hematopoietic stem cells (HSCs) are largely quiescent at the steady-state. However, mechanisms that regulate HSC quiescence/cycling remain incompletely understood. Using mitochondrial membrane potential (MMP) to dissect the heterogeneity of HSCs (LSKCD150+CD48-), we find that HSCs within 25% lowest MMP (MMP-low) fractions are almost entirely (~95% ±2.65) in G₀ as measured by Pyronin Y/Hoechst staining (p<0.05, n=3). In contrast, HSCs within 25% highest MMP (MMP-high HSCs) are in majority in cycling (see abstract 129099). To elucidate mechanisms implicated in the regulation of HSC cycling at the single cell level in quiescent MMP-low versus primed MMP-high HSCs we used single-cell RNA-Seq (scRNA-Seq) analysis. Cycling analysis in silico in each cell by CYCLONE further confirmed that over 80% of MMP-low HSCs are within G₀/G₁, as compared to less than 40% of MMP-high HSCs that are mostly in the S/G₂/M phase. Notably, GO enrichment analysis related to protein degradation through lysosomal- and proteasomal-mediated pathways were significantly enriched in MMP-low HSCs (p=0.002). Strikingly, and in agreement with our scRNA-seq analysis, a greater abundance of lysosomes was observed in MMP-low relative to -high HSCs (p=0.002). Higher expression of lysosomal genes was further confirmed by qRT-PCR in MMP-low relative to -high HSCs. Analysis of lysosomal content by immunofluorescence staining showed that while the lysosomal specific marker LAMP2 was barely detectable in MMP-high HSCs, LAMP2 was readily found in MMP-low HSCs, results further confirmed by additional markers LAMP1 and LysoTracker Green. Lysosomes are, among others, a major component of organelle degradation through autophagy, which is required for the maintenance of HSCs however, whether lysosomes are implicated in regulating HSC beyond autophagy is unknown. To address this we examined the effect of the suppression (and not activation that is required for autophagy) of lysosomal activation on in vitro HSC maintenance. Treatment with concanamycin-A (ConA), a specific inhibitor of lysosomal acidification via inhibition of the vacuolar H⁺ -adenosine triphosphatase ATPase (v-ATPase) led to 3 fold improved frequency of phenotypically defined HSCs from optimally cultured lineage-negative cells in 24 hours (p<0.05, n=4). This was associated with 4-fold greater retention of the MMP-low HSC fraction (p<0.05, n=4). Cell divisions of single MMP-low and -high GFP⁺ HSCs treated with ConA or vehicle control was tracked up to 60 hours in culture. Over 70% of control treated MMP-low GFP⁺ HSCs did not divide during this time, whereas the majority (>85%) of MMP-high GFP⁺ HSCs divided at least once (p=0.001, n=5). While ConA treatment had only a slight effect on non-dividing MMP-low HSCs in culture, it significantly increased the frequency of non-dividing MMP-high GFP⁺ HSCs (p=0.007). Priming of MMP-low to -high HSCs was associated with lysosomal recruitment, and activation of mTOR signaling in MMP-high HSCs (p=0.001, n=5). Importantly, ConA-treatment led to the repression of mTOR expression and activity in MMP-high HSCs (p<0.001). In addition, a 48-hours ConA treatment led to enhanced frequency of LTC-ICs recovered in limiting dilution analysis of both MMP-low (p=0.023) and -high (p=0.004) HSCs ex vivo. To further investigate the role of suppression of lysosomal activation in vivo, FACS-purified MMP-low and -high HSCs were treated with vehicle control or ConA ex vivo for 4 days before 50 ConA- or control-treated MMP-low or -high HSCs were mixed with CD45.2 (2x10⁵) competitor cells and injected into lethally irradiated mice (n=7) in a competitive repopulation assay. Reconstitution levels were consistently more robust in ConA-treated populations of MMP-low (p= 0.001) and -high (p=0.001) HSCs after 18 weeks as compared to control. Importantly, HSC-derived lineage output was balanced in its composition up to 18 weeks in recipients of MMP-low HSC regardless of ConA treatment as well as in ConA-treated MMP-high HSCs, while control MMP-high HSC was myeloid-biased. Overall our results, based on HSC mitochondrial heterogeneity, suggest that lysosomal -content and activity participate in the maintenance of HSC quiescence. Based on these findings, we propose a model that stipulates that lysosomal activation primes HSCs (G_0⇒G₁) while lysosomal suppression maintains HSC quiescence.