Identification of Old HSCs with Preserved Self-Renewal and Long-Term Reconstitution Potential

Aging hematopoiesis is characterized by increased numbers of immunophenotypic HSCs that exhibit impaired self-renewal and long-term reconstitution potential, both in competitive and noncompetitive settings. We previously demonstrated that normal young mouse HSCs (CD34-CD150+LSK) can be fractionated into subsets based on expression of c-Kit surface expression, with c-Kit^hi HSCs exhibiting reduced self-renewal and megakaryocytic biased differentiation (Shin et al., 2014). We therefore hypothesized that the expansion of c-Kit^hi HSCs in old mice could potentially explain the age-related decline in immunophenotypically defined old HSC function. Evaluation of the bone marrow of 24-month-old C57Bl/6 mice revealed that the frequency of c-Kit^hiHSCs (out of total HSCs) is 1.5-fold higher in old mice than in 3-month old mice (P=0.04), while the frequency of c-Kit^lo HSCs was 1.5-fold lower in old mice (P=0.007; Fig 1A). This finding is consistent with our previous observation of a megakaryocytic-bias in c-Kit^hiHSCs, since peripheral blood analysis of old mice revealed a 2.1-fold increase in platelets compared to young mice (p<0.01) (Fig 1B).

To test the long-term reconstitution potential of aging HSCs, we competitively transplanted 400 c-Kit^loor c-Kit^hi HSCs from 24-month old mice, along with 300,000 competitor bone marrow cells, into lethally irradiated young recipients. Sixteen weeks post-transplantation, mice receiving old c-Kit^hi HSCs exhibited significantly lower donor peripheral blood chimerism levels compared to old c-Kit^lo HSC recipients (9.4% vs 57.1%, P=0.02) (Fig 1C). Both old c-Kit^hiand old c-Kit^lo HSCs exhibited similar myeloid-reconstituting potential (Fig 1D). Furthermore, mice transplanted with old c-Kit^lo HSCs exhibited 78% donor HSC chimerism, achieving 6.4-fold higher chimerism levels than mice transplanted with old c-Kit^hi HSCs, this was comparable to the differences observed with young c-Kit^lo and c-Kit^hi transplanted HSCs (Fig 1E). To quantify the self-renewal capacity of old HSCs, we calculated the "self-renewal quotient" (Challen et al., 2010). This analysis showed that the self-renewal potential in old c-Kit^hi and c-Kit^lo HSCs were 0.8 and 7.8 respectively, indicating higher self-renewal potential in c-Kit^lo than c-Kit^hi HSCs (Fig 1F). Collectively, these data suggest that myeloid-biased differentiation is an age-associated change in hematopoiesis that may not be associated with decreased self-renewal in all HSCs.

To gain mechanistic insights underlying these qualitative differences, we interrogated transcriptional profiles of microarray data from c-Kit^lo and c-Kit^hi HSCs, to identify potential pathways critical for HSC maintenance. Gene Ontology and pathway analyses showed several differentially expressed pathways between c-Kit^hiand c-Kit^loHSCs, of which genes related to protein translation and mitochondrial activity was significantly enriched in c-Kit^hi HSCs (Fig 1G). Given the underrepresentation of translation-related genes in c-Kit^lo HSCs, we tested whether they exhibit reduced global translation using OP-Puro incorporation assays. These studies confirmed that old c-Kit^lo HSCs show lower global translation levels than c-Kit^hiHSCs (Fig 1H).

Overall, our studies demonstrate functional heterogeneity among old HSCs and identify a novel strategy to identify old HSCs with preserved self-renewal and long-term reconstitution capacity. The ability to identify and prospectively fractionate old HSCs offers a novel approach to investigate the molecular mechanisms underlying HSC aging.

Figure legend. (A) Frequency of c-Kit^hior c-Kit^lo HSCs was assessed by flow cytometry. (B) Circulating platelet numbers were assessed using a Hemavet counter. Competitive transplants of old c-Kit^lo and c-Kit^hi HSCs into lethally irradiated recipients (C-F). Donor chimerism (C) and lineage potential (D) was evaluated in the peripheral blood of primary recipients. Bone marrow was analyzed at 16 weeks, for donor-derived HSC chimerism (E) and self-renewal quotient (F). (G) Enrichment plots comparing microarray data generated from c-Kit^hiand c-Kit^lo HSCs, using pathways translation-related gene sets. (H) OP-Puro incorporation assays in 24-month old mice. Results are representative of three independent experiments, and shown as mean ± SEM. n = 4-5 mice. *, P < 0.05; **, P < 0.01.

Figure 1

View largeDownload slide

Figure 1

View largeDownload slide

Close modal