Decoding HSC heterogeneity

In this issue of Blood, Shimazu et al report that some mouse hematopoietic stem cells (HSCs) lack surface expression of CD86, and that these CD86⁻ HSCs are capable of myeloid lineage reconstitution but produce limited to no lymphoid recovery.¹ 

Blood cells derive from rare HSCs, which can self-renew and can also reconstitute all hematopoietic cell lineages, including erythroid, myeloid, and lymphoid cell lineages.²  However, it is now evident that significant heterogeneity exists even among HSCs.^3-5  Transplantation of single HSCs revealed that individual HSCs can differ widely with regard to in vivo differentiation potential and proliferative ability.³  Not only does the number of total mature blood cells reconstituted by individual HSCs show wide variation, but the proportion of myeloid versus lymphoid cells derived from each individual HSC also varies. Some HSCs appear to be myeloid-biased, as on transplantation they repopulate myeloid cell lineages efficiently but are less competent in reconstituting lymphoid cell lineages. The myeloid versus lymphoid lineage differentiation preference of individual HSCs is stably maintained in their HSC progeny after secondary and even tertiary transplants.^3-5  These data suggest that the functional heterogeneity of HSCs is caused by cell-intrinsic mechanisms, with a genetic or epigenetic basis that is presently very poorly understood.

Lymphopoiesis declines with age, during pregnancy, and after chronic infection.^3,6,7  Age-related declines in lymphopoiesis may compromise host protection to pathogens; it is therefore of considerable importance to understand how lymphopoiesis is controlled. Recent studies suggest that in some circumstances, lymphopoiesis might be impaired at the very earliest steps of blood cell development. Indeed, myeloid-biased HSCs accumulate with age and after lipopolysaccharide (LPS) or TGF-β1 treatment.^6,8  HSCs in LPS-treated mice and aged mice also lack expression of CD86 on the cell surface, suggesting that absence of CD86 might be useful in identifying lineage-biased HSCs.⁶ 

In the present study, Shimazu et al compared HSC subsets expressing or lacking surface CD86 expression.¹  They showed that whereas most HSCs were CD86⁺, the minor CD86⁻ HSC subset expressed higher levels of CD150. CD150 is a commonly used surface molecule to enrich HSCs in adult mice, and higher surface expression of CD150 on HSCs was previously shown to correlate with myeloid differentiation bias.³  Transplantation assays were performed to examine the self-renewal and lineage reconstitution activities of CD86⁻ HSCs. Shimazu and colleagues found that CD86⁻ HSCs could reconstitute myeloid lineages; however, their ability to reconstitute lymphoid lineage was diminished. Indeed, some mice receiving CD86⁻ HSCs were devoid of donor lymphoid lineages.

The authors tested the effects of deletion of CD86, and showed that CD86 itself does not affect the lineage-reconstitution ability of HSCs.¹  Conditional ablation of the ets-family transcription factor PU.1 in mice established that the expression of CD86 on HSCs requires PU.1. It remains to be determined whether down-regulation of PU.1 expression contributes to the myeloid bias of CD86⁻ HSCs, or whether other mechanisms are involved. Identifying the responsible mechanisms is an important area of future investigation.

Interestingly, although CD86⁻ HSCs accumulate with age, the authors found that small numbers of myeloid-biased HSCs exist even in young, healthy mice.¹  It is unknown whether this reflects exposure of HSCs to TLR ligands or other products of the normal microbiome, and whether such CD86⁻ HSCs might be absent in germ-free mice. It is also unclear whether accumulation of such lineage-biased or lineage-restricted HSCs might have adaptive significance. For example, age-related declines in lymphopoiesis have been previously suggested to reduce susceptibility to lymphoid-lineage cancers.⁹  It will be interesting to determine whether CD86⁻ HSCs are less susceptible to the transforming effects of oncogenes that drive cancers of B and T cells. The discovery that CD86 can be used as a marker to identify HSCs with reduced lymphoid developmental potential means that these and related questions can now be better addressed.