Response: The CD150^high compartment is not the exclusive reservoir of LT-HSCs within the bone marrow

Kiel et al respond to our recent article¹ demonstrating that subsets of functional long-term repopulating hematopoietic stem cells (LT-HSCs) can be distinguished based on Hoechst dye exclusion (side population, SP) coupled with CD150 (signaling lymphocytic activation molecule family member 1, SlamF1) staining, with both CD150⁺ and CD150⁻ populations containing substantive numbers of LT-HSCs. We appreciate the opportunity to address these concerns.

In our report, we did not dispute (and, indeed, functionally corroborated) the existence of LT-HSCs defined by CD150⁺CD48⁻CD41⁻. In Figure 1 of our article,¹ our objective was to underscore the necessity of stringent CD150⁺ gating for optimal HSC enrichment. Most investigators use isotype controls to identify “positive” cells. We showed in Figure 1B that such a strategy with the SLAM-family markers would result in populations with relatively low HSC purity, and noted that slight changes in gating lead to quite disparate outcomes. For this very reason, we entreated users desiring superior HSC enrichment to use the stringent CD150^high gates shown in Figure 1C (similar to Figure 4A in Kiel et al²), or to include canonical markers to assist HSC identification (as in Morrison's recent work³). We proposed a CD150 “high” designation for SLAM-identified HSCs to emphasize this point.

Kiel et al also suggest that the observed LT-HSC activity in the CD150⁻ c-Kit⁺Lin⁻Sca-1⁺ (KLS) side population (SP^KLS) might be attributable to contamination by rare CD150⁺ cells. To minimize this possibility, we set exceptionally stringent gates for defining CD150⁻ and CD150⁺ cells (Figure 4A¹), making cross-contamination unlikely. We also reanalyzed sorted samples, to verify purity, showing that fewer than 1 in 100 CD150⁻ SP^KLS cells could be CD150^high contaminants (Figure 4A¹).

While CD150⁻ SP^KLS cells do show a lower contribution to blood chimerism, this deficit is mild (Figure 4B¹). Thus, to attribute the HSC activity of this compartment to a stray CD150^high cell would assume extraordinary potency, as this cell would have to perform comparably to 100 of its CD150⁺ counterparts, a result which seems implausible, particularly since 100% of our animals (14/14) that received transplants of 100 CD150⁻ SP^KSL cells in independent experiments demonstrated full multilineage reconstitution. The 2 studies^4,5 that Kiel et al suggest we overlooked did not investigate whether CD150⁻ bone marrow fractions contained reconstituting activity, but do confirm that SlamF1 transcripts are enriched in HSCs (as do our data⁶).

We agree with Kiel et al that limiting dilution experiments are of value to precisely assess the HSC activity of a given marrow compartment. However, the robust reconstituting ability of only 100 CD150⁻ SP^KLS across all recipients strongly supports substantive LT-HSC activity in the CD150⁻ compartment. Moreover, previous studies on the SP^KLS population, included as a benchmark in our current study, demonstrated that approximately 35% of singly transplanted SP^KLS cells yielded long-term multilineage reconstitution in transplantation assays,⁷ a number comparable with the approximately 40% to 45% of single SLAM cells.² Given that single SP^KLS cells were transplanted without knowledge of their CD150 status, and coupled with our demonstration that the lower SP is only approximately 50% positive for CD150 in young mice (Figure 2B¹), it stands to reason that the proportion of HSCs in the CD150⁻ SP^KLS fraction is not trivial.

As we gain more insight into HSCs, it becomes clear that, like many biologic processes, HSC function is more complex than originally suspected. Recent work points away from a concept of LT-HSCs as a functionally homogenous population toward an emerging picture of HSCs in multiple states,⁸ including greater proliferation than previously thought.⁹ We feel our data help open new avenues, generating novel hypotheses with broad explanatory power. As we discuss more fully,¹ CD150 differences could suggest alternate HSC niches, and help explain distinct differentiation patterns of single HSCs.⁸ Future work may more fully illuminate the enigmatic HSC.

Authorship

Conflict-of-interest disclosure: The authors declare no competing financial interests.

Correspondence: Margaret A. Goodell, Stem Cells and Regenerative Medicine Center, One Baylor Plaza, Houston, TX 77030; e-mail: Goodell@bcm.edu.